2011-08-01
[18:52:00.0000] <AryehGregor>
/me is annoyed at the new "ayg⊙an via gmail.com", although of course other mail clients would have been saying something like that forever

[18:54:00.0000] <AryehGregor>
Maybe I should just switch all my e-mail over to the ayg⊙an Google account from Simetrical⊙gc?  Probably by setting up the former to read all the latter's contents via IMAP?

[18:54:01.0000] <AryehGregor>
/me wonders if that would be possible and how long it would take, or if there's a better way

[19:23:00.0000] <MikeSmith>
heycam: where you moving to?

[19:26:00.0000] <MikeSmith>
oh, Melbourne, I gues

[19:51:00.0000] <heycam>
MikeSmith, yep

[19:51:01.0000] <MikeSmith>
cool

[19:51:02.0000] <MikeSmith>
I just got back yesterday from a visit to Perth

[19:52:00.0000] <MikeSmith>
surprised how expensive things are there

[19:52:01.0000] <heycam>
oh yeah, all the increased wealth over there from the mining boom

[20:54:00.0000] <bga_>
anybody from Mozilla?

[20:54:01.0000] <bga_>
i have bug

[20:55:00.0000] <bga_>
also miketaylr

[20:55:01.0000] <bga_>
in last opera window.onerror is fake event

[20:55:02.0000] <miketaylr>
bga_: orly

[20:55:03.0000] <bga_>
never fires

[20:56:00.0000] <miketaylr>
i thought we didn't support window.onerror

[20:56:01.0000] <bga_>
but

[20:56:02.0000] <bga_>
'onerror' in window // true!

[20:56:03.0000] <bga_>
plz remove

[20:56:04.0000] <miketaylr>
oh, hmm

[20:56:05.0000] <bga_>
if not support

[20:56:06.0000] <miketaylr>
ok, i'll file a bug if that's the case

[20:56:07.0000] <bga_>
thx :)

[20:57:00.0000] <miketaylr>
np

[21:07:00.0000] <bga_>
miketaylr and opera is one browser which correctly parse { if(foo) _a() else _b() } :)

[21:07:01.0000] <bga_>
other wants ;

[21:07:02.0000] <bga_>
not bug, feature!

[21:09:00.0000] <miketaylr>
:)

[21:22:00.0000] <roc>
bga_: what's your Mozilla bug?

[21:23:00.0000] <bga_>
roc window.addEventListener('error', #(e){ _log(e.message, e.fileName, e.lineNumber) })

[21:24:00.0000] <bga_>
or .filename .lineno as in webkit

[21:24:01.0000] <bga_>
but in Gecko both - null

[21:25:00.0000] <roc>
file a bug I guess

[21:25:01.0000] <roc>
I don't know

[21:25:02.0000] <bga_>
a can get lineNumber only if i attach event via window.onerror = #(message, fileName, lineNumber){}

[21:26:00.0000] <roc>
those are very different

[21:31:00.0000] <roc>
I see that for error events we pass the message, filename and line number as direct parameters to the handler

[21:31:01.0000] <roc>
instead of passing the event object

[21:31:02.0000] <roc>
I don't know why we do that

[21:32:00.0000] <roc>
addEventListener("error", function f(msg, filename, lineno) { ... }, false) should work

[21:36:00.0000] <bga_>
typeof(msg) == 'object' because its event object

[21:37:00.0000] <bga_>
both filename and lineno are undefined

[21:37:01.0000] <bga_>
in 8a1

[21:37:02.0000] <heycam>
the spec says that window.onerror, as well as registering a listener for the error event, is also called for script errors -- and script errors invocations of window.oenrror get called with the three arguments

[21:58:00.0000] <roc>
weird

[23:06:00.0000] <zcorpan>
hi

[23:09:00.0000] <MikeSmith>
zcorpan: hello

[23:09:01.0000] <zcorpan>
hello MikeSmith

[23:10:00.0000] <zcorpan>
what happened last month?

[23:59:00.0000] <hsivonen>
someone filed a bug report as a Word document?

[00:58:00.0000] <zcorpan>
Hixie: i don't mind dropping <datalist>-fallback-to-<select>. most authors use script to implement fallback instead of clever markup

[02:59:00.0000] <mhausenblas>
hey foolip - a new toy to publish HTML5+Schema.org from CSV files https://github.com/mhausenblas/web.instata - in case you wanna play around w/ it ;)

[03:01:00.0000] <foolip>
/me looks

[03:01:01.0000] <foolip>
I don't have any POTD I'm afraid :)

[03:03:00.0000] <mhausenblas>
he he, fair enough

[03:04:00.0000] <mhausenblas>
I just found the templating stuff fun, so I thought I give it a try

[03:04:01.0000] <mhausenblas>
Bottle rocks!

[03:05:00.0000] <mhausenblas>
and, btw, thanks for microdata-live - helped me to debug stuff

[03:05:01.0000] <mhausenblas>
as well as thanks to hsivonen for validator.nu - same applies here

[03:06:00.0000] <foolip>
mhausenblas, do you think vocabulary-specific validation would be useful for validator.nu, or a waste of time?

[03:06:01.0000] <mhausenblas>
hmm

[03:06:02.0000] <mhausenblas>
didn't think about it, tbh

[03:07:00.0000] <mhausenblas>
I guess for widely used vocabs, vCard, iCal, etc. yes

[03:07:01.0000] <mhausenblas>
maybe :)

[03:07:02.0000] <mhausenblas>
how would you do that, foolip?

[03:07:03.0000] <mhausenblas>
ah, you're contributing to validator.nu, right? is it on github?

[03:07:04.0000] <foolip>
mhausenblas, http://bugzilla.validator.nu/show_bug.cgi?id=851

[03:08:00.0000] <foolip>
mhausenblas, I wrote a few patches, but don't expect to be a long-term contributor

[03:08:01.0000] <mhausenblas>
right

[03:09:00.0000] <foolip>
mhausenblas, to validate a vocabulary you "just" check if the properties values follow the syntax required by the vocabulary

[03:09:01.0000] <mhausenblas>
so, let's put it this way: for a couple of common vocabs (that are used throughout microformats, microdata an RDFa) such as license, vCard, etc it would make sense, yes

[03:09:02.0000] <mhausenblas>
makes sense

[03:10:00.0000] <foolip>
are there any vocabularies at all that are used for all 3 syntaxes?

[03:10:01.0000] <mhausenblas>
maybe the first step is to compile a list of vocabs (see above for seeds), including, obviously Schema.org terms

[03:10:02.0000] <mhausenblas>
I guess so, yes - lemme see

[03:11:00.0000] <mhausenblas>
vCard - check, iCal - check, license - check

[03:11:01.0000] <mhausenblas>
Schema.org both for microdata and RDFa or whatever we will have in 3 months time :D

[03:12:00.0000] <foolip>
I think vCard and vEvent are way to complicated and will fail to get adopted, I don't plan to do anything with them

[03:12:01.0000] <mhausenblas>
I agree that they are not that simple, but I think widely used and needed

[03:12:02.0000] <mhausenblas>
but, again, Schema.org to the rescue ;)

[03:13:00.0000] <mhausenblas>
(that's why I'm focusing on Schema.org)

[03:13:01.0000] <foolip>
they're not used in microdata, which is all that I'd be checking

[03:13:02.0000] <mhausenblas>
right

[03:13:03.0000] <mhausenblas>
anyways, I guess a Schema.org validator would be the most interesting thing to have

[03:13:04.0000] <mhausenblas>
I'm about to do it anyway (for web.instata and other projects)

[03:14:00.0000] <mhausenblas>
in which language is validator.nu written?

[03:14:01.0000] <foolip>
Java

[03:14:02.0000] <mhausenblas>
ouch

[03:14:03.0000] <foolip>
the main reason I decided to not do it independently is because it already has all the framework for pointing to the source code location where something was wrong

[03:15:00.0000] <mhausenblas>
/me is chuckling at html5lib/ihatexml.py

[03:15:01.0000] <foolip>
without that, a validator would be useless for all but trivial cases

[03:15:02.0000] <mhausenblas>
roger that

[03:16:00.0000] <mhausenblas>
/me used to do Java ... some 8y ago ... now really only Python and JS

[03:16:01.0000] <mhausenblas>
anyways, guess back to work (though, formally it's a bank holiday here today, I think ;)

[03:17:00.0000] <mhausenblas>
good talking to you foolip - KUTGW!

[03:17:01.0000] <foolip>
mhausenblas, likewise!

[04:24:00.0000] <MikeSmith>
hsivonen: I notice that Hixie's "microsyntaxes-dates" folder is empty but the v.nu behavior for checking "valid date or time string" and "global date and time string" is still not conformant

[04:25:00.0000] <MikeSmith>
specifically, I notice that for date-or-time, v.nu allows, e.g., 1996-01-01T12:05:25 (a date and time with no time-zone information) and 12:05:25Z (a time with no date but with time-zone information)

[04:25:01.0000] <MikeSmith>
but the spec prohibits both of those

[04:26:00.0000] <MikeSmith>
and for global-date-and-time, the spec allows 1996-01-01T12:05Z (a date and time string with no seconds specified), but v.nu prohibits it

[04:37:00.0000] <zcorpan>
i thought v.nu didn't implement the spec but just used an off-the-shelf library for validating dates and times

[04:41:00.0000] <MikeSmith>
zcorpan: from what I can see in the code, it's currently just using regexes

[04:41:01.0000] <zcorpan>
oh

[04:42:00.0000] <zcorpan>
https://bitbucket.org/ms2ger/dom-core/changeset/f92a4bdbb5e6 - how is QName equal to Name? annevk?

[04:43:00.0000] <MikeSmith>
zcorpan: btw, about what happened last month, I have a hard time remembering details about anything further back than two weeks, and I was away most of last week, so my recollections are pretty hazy

[04:43:01.0000] <zcorpan>
k

[04:43:02.0000] <Ms2ger>
https://bitbucket.org/ms2ger/dom-core/changeset/6caa468281a4

[04:43:03.0000] <Ms2ger>
zcorpan, ^

[04:44:00.0000] <MikeSmith>
hsivonen: I don't know if either of those differences are ones that you had sent feedback on, but if so, it seems that Hixie must have already responded

[04:44:01.0000] <zcorpan>
Ms2ger: thanks

[04:58:00.0000] <hsivonen>
MikeSmith: it's quite possible that I've missed Hixie's response(s) or spec edits on that topic

[04:58:01.0000] <MikeSmith>
ok

[05:00:00.0000] <MikeSmith>
hsivonen: btw, I also noticed today that for cases of a time element that has a pubdate attribute, the spec now requires the datetime value can't be just a time

[05:00:01.0000] <MikeSmith>
that is, it must either be a date with a time, or just a date

[05:01:00.0000] <MikeSmith>
so I checked in a change for that today

[05:01:01.0000] <MikeSmith>
but the spec also makes the same requirement for the text content of the time element, if no datetime attribute is specified

[05:02:00.0000] <MikeSmith>
so I can make a change for that too, but I'm wondering if it's worth it, especially given that there seems to be some chance of the time element just being dropped altogether

[05:40:00.0000] <annevk>
What is the latest on mutation listeners?

[06:02:00.0000] <zcorpan>
annevk: grattis

[06:03:00.0000] <annevk>
takk

[06:14:00.0000] <Ms2ger>
annevk, congratulations

[06:18:00.0000] <annevk>
thanks!

[06:26:00.0000] <brucel>
zcorpan "Hixie: i don't mind dropping <datalist>-fallback-to-<select>. most authors use script to implement fallback instead of clever markup" - surely too early to tell? datalist so far has ben implemented in Opera for a while, Chrome recently. So a bit premature to declare the fallback mechanism dead?

[06:27:00.0000] <Ms2ger>
And Firefox for a while

[06:27:01.0000] <Ms2ger>
Don't forget the best implementation in the world ;)

[06:27:02.0000] <jgraham>
/me thinks the fallback is a good idea still. Was there some evidence it isn't

[06:28:00.0000] <annevk>
The fallback mechanism was mostly designed for a time when browsers were not updated

[06:28:01.0000] <jgraham>
?

[06:28:02.0000] <annevk>
Well, IE was not updated

[06:28:03.0000] <jgraham>
That's not evidence it isn't

[06:28:04.0000] <annevk>
That changed and scripts became more prevalent to handle fallback in better ways

[06:28:05.0000] <annevk>
So these days introducing somewhat cleaner design is feasible

[06:29:00.0000] <jgraham>
That sounds a bit like "backwards compatibility is hard. Let's not bother"

[06:29:01.0000] <jgraham>
Or "graceful degradation" if you like

[06:29:02.0000] <annevk>
It's more that backwards compatibility is solved in a different way that allows new features to be designed without clutter

[06:33:00.0000] <jgraham>
What is the clutter in this case? And who does it help if it is already implemented everywhere?

[06:34:00.0000] <annevk>
new implementors, people trying to grasp the platform

[06:34:01.0000] <annevk>
i value those more than legacy implementations

[06:35:00.0000] <jgraham>
Is there any evidence that this is actually hard to implement?

[06:37:00.0000] <annevk>
it's pretty clear it's more complex

[06:37:01.0000] <brucel>
It seems to me that "everything inside datalist is invisible, except the <options>" is beautifully simple, annevk

[06:38:00.0000] <brucel>
(even I understand it)

[06:38:01.0000] <annevk>
brucel, except that doesn't tell you if you need to traverse children or descendants, whether scripts are executed, whether other descendants are submitted, etc.

[06:39:00.0000] <zcorpan>
brucel: maybe, although i base my judgement on other features where i see most authors use script to implement the fallback behavior

[06:40:00.0000] <zcorpan>
brucel: which makes sense since it's dead simple if somebody else writes the script, e.g. modernizr

[06:41:00.0000] <zcorpan>
jgraham: i recall bratell whining about datalist having to consider all descendants and not just the children

[06:42:00.0000] <zcorpan>
it complicates teh implementation

[06:42:01.0000] <brucel>
zcorpan and that's legit  - but how big is the amount of evidence at the moment? Basing it on the decisions of a small group of highly-motivated script-savvy early adopters isn't indicative of the majority of authors who, let's face it, won't be figuring this stuff out until IE10

[06:43:00.0000] <zcorpan>
brucel: the fallback is *for* early adopters. when it's mainstream, you don't bother with fallback at all

[06:44:00.0000] <zcorpan>
if the early adopters don't use the declarative fallback mechanism, it isn't going to do anyone any good

[06:47:00.0000] <brucel>
zcorpan I don't understand you, sorry. The fallback, surely, is so IE6,7,8,9 users get an input mechanism more helpful than a plain text input.

[06:48:00.0000] <zcorpan>
i'm saying that pages that use datalist will use a scripted dropdown if datalist is not supported

[06:48:01.0000] <zcorpan>
if (!input.list) { use jquery dropdown and be done with it }

[06:51:00.0000] <brucel>
feels to me like losing a useful declarative fallback mechanism in order to make it easier for implementors, because in the short time that it's been supported by majority of browsers, not everyone is using it

[06:54:00.0000] <zcorpan>
yeah

[06:55:00.0000] <zcorpan>
i've been disappointed several times where new features have useful fallback mechanisms but people feature check for the feature instead and fallback to jquery or something with script

[06:55:01.0000] <Rik`>
brucel: is it really implemented in Chrome ?

[06:57:00.0000] <brucel>
Rik Not sure - I meant to type Firefox (and made Ms2ger mad by getting it wrong)

[06:57:01.0000] <Rik`>
last time I checked, it was broken because the default CSS had "datalist{ display: none;}"

[06:57:02.0000] <brucel>
ah, think they removed that

[06:58:00.0000] <zcorpan>
without implementing the feature?

[06:58:01.0000] <brucel>
and it's implemented in IE10 although the fallback mechanism is mangled; I reported it to them and was told they were de-mangling that

[06:59:00.0000] <Rik`>
brucel: I removed that ;)

[06:59:01.0000] <zcorpan>
brucel: mangled how?

[06:59:02.0000] <Rik`>
zcorpan: yeah, they had flags for <datalist> implementation but not in the CSS

[06:59:03.0000] <zcorpan>
k

[06:59:04.0000] <brucel>
zcorpan - just let me dig out my tests

[07:05:00.0000] <brucel>
zcorpan <input list=cheese>

[07:05:01.0000] <brucel>
<datalist id=cheese>

[07:05:02.0000] <brucel>
<option>Wensleydale</option>

[07:05:03.0000] <brucel>
<option>Gouda</option>

[07:05:04.0000] <brucel>
</datalist>

[07:05:05.0000] <brucel>
.. works fine in IE10 pp1

[07:06:00.0000] <brucel>
.. but Jeremy's example with <select> fallback  always goes back to the fallback, even though datalist is supported

[07:07:00.0000] <zcorpan>
you mean the select is rendered?

[07:07:01.0000] <brucel>
yes

[07:07:02.0000] <zcorpan>
so chrome stole ie's CSS!

[07:08:00.0000] <Rik`>
brucel: well it's PP1, report it and it should be fixed

[07:08:01.0000] <brucel>
have done, a couple of weeks ago

[07:10:00.0000] <brucel>
but my point is that until we see datalist usable in the Web's Favourite Browser most devs aren't going to be aware of it, so it's too early to call the degrade-to-select pattern a failure

[07:12:00.0000] <zcorpan>
and when we know for sure it'll be too late to change it :)

[07:29:00.0000] <annevk>
so what is left

[07:29:01.0000] <annevk>
* mutations

[07:29:02.0000] <annevk>
* range

[07:30:00.0000] <annevk>
* event handlers

[07:30:01.0000] <annevk>
and evaluation of what can be removed and what needs to be added back in

[07:31:00.0000] <annevk>
some of which depend on either Acid3 changing or browsers giving up on Acid3

[07:51:00.0000] <hsivonen>
I didn't want to think about mutation events today, but I ended up thinking about them anyway.

[07:51:01.0000] <annevk>
sorry

[07:52:00.0000] <Ms2ger>
And it's not even annevk's fault

[07:52:01.0000] <hsivonen>
annevk: oh, not your fault. I blame sicking's patch

[07:54:00.0000] <zcorpan>
so that's the latest on mutation listeners

[08:16:00.0000] <atrigent>
when using HTML5 drag and drop, is it possible to obtain the dom node of the element being dragged on drop?

[08:48:00.0000] <gsnedders>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%20%0A%20%3Cstyle%3E%20%0A%20%20%20div%20%7B%20background-color%3A%20green%3B%20color%3A%20lime%3B%20%7D%20%0A%20%20%20div%3Afirst-line%20%7B%20background-color%3A%20red%3B%20color%3A%20blue%3B%20%7D%20%0A%20%20%20span.one%20%7B%20background-color%3A%20inherit%3B%20color%3A%20inherit%3B%20%7D%20%0A%20%3C%2Fstyle%3E%20%0A%20%3Cdiv%3E%3Cspan%20class%3D%22one%22%3EGreen%3F%3C%2Fs

[08:50:00.0000] <Michael>
That's a cool tool

[09:08:00.0000] <annevk>
gsnedders, should be I think

[09:09:00.0000] <annevk>
gsnedders, the real fun is when you put a break in the <span>

[09:18:00.0000] <gsnedders>
annevk: Why? Why does background-color inherit from one place and color from another?

[09:20:00.0000] <erlehmann>
oh my. google+ javascript routing breaks links containing fragments

[09:20:01.0000] <erlehmann>
or so it seems

[09:20:02.0000] <erlehmann>
srsly.

[09:20:03.0000] <erlehmann>
._.

[09:24:00.0000] <gsnedders>
When is Fx8 release?

[09:25:00.0000] <gsnedders>
14 weeks time?

[09:25:01.0000] <smaug____>
https://wiki.mozilla.org/RapidRelease/Calendar

[09:25:02.0000] <AryehGregor>
gsnedders, your URL was cut off at: Green%3F%3C%2F

[09:27:00.0000] <gsnedders>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1090

[09:27:01.0000] <gsnedders>
is somewhat amusing in Firefox was the comment with it

[09:27:02.0000] <gsnedders>
AryehGregor: so you didn't really miss anything

[09:27:03.0000] <AryehGregor>
Yeah, seems not.

[09:31:00.0000] <erlehmann>
ohne thing is that the release version number stuff seems to work

[09:32:00.0000] <erlehmann>
i fell bad about not having ff $next yet ;)

[09:47:00.0000] <zewt>
gar i need to figure out how to make firefox stop pasting html as html

[09:47:01.0000] <zewt>
it's never ever wanted and i always have to paste text into a text editor and copy it back out to make it stop, heh

[09:48:00.0000] <Philip`>
You should use an OS on which copy-and-paste almost never works reliably so it's bound to degrade to plain text, like Linux

[09:48:01.0000] <zewt>
this feels like one of those things that somebody designed and implemented without really thinking about how annoying and unwanted it is in practice

[09:49:00.0000] <zewt>
eg. pasting text into email and having it end up in a giant font

[09:50:00.0000] <gsnedders>
zewt: It's one of the most common feature requests for Opera, though.

[09:56:00.0000] <erlehmann>
Philip`, sadly, even linux desktops do that right nowadays. bugs me as well.

[10:18:00.0000] <annevk>
gsnedders, that's the current thinking

[10:20:00.0000] <gsnedders>
annevk: citation?

[10:21:00.0000] <annevk>
don't have anything handy

[10:21:01.0000] <annevk>
but see discussion from some years ago

[10:21:02.0000] <annevk>
involved bz

[10:25:00.0000] <gsnedders>
annevk: CSS 2.1 doesn't seem to match that, though

[10:28:00.0000] <gsnedders>
annevk: inherit has a single dfn that is used for everything, which implies constant behaviour

[10:38:00.0000] <annevk>
gsnedders, CSS is a mess

[10:39:00.0000] <Ms2ger>
News at 12

[10:43:00.0000] <gsnedders>
annevk: That doesn't mean we should go against fairly unambiguous parts of the spec. :P

[11:22:00.0000] <gsnedders>
annevk: Also: Happy birthday!

[11:29:00.0000] <dglazkov>
good morning, Whatwg!

[11:30:00.0000] <dglazkov>
annevk: Happy Birthday!

[11:53:00.0000] <Hixie>
hsivonen: it wasn't on my radar, but in general i recommend people send feedback ot the list rather than to a limited-availability social network :-)

[11:54:00.0000] <Hixie>
for those whe were talking about the <datalist>/<select> thing -- as far as i can tell, none of the implementations match the spec or are interoperable with each other, so we don't have enough data to know if the fallback can be used by authors or not

[12:28:00.0000] <AryehGregor>
Okay, so I changed the address in my W3C profile thing to ayg⊙an, but public-html thinks I'm not subscribed?

[12:29:00.0000] <AryehGregor>
Any further magic I need to do?

[12:29:01.0000] <AryehGregor>
/me inquires of MikeSmith

[12:30:00.0000] <AryehGregor>
Also, MikeSmith, can I get a Bugzilla component for my editing spec?  Specifically, are there going to be any political problems based on the fact that the spec isn't hosted at the W3C that might lead people to agitate that it should be shut down after I've come to start relying on it?

[12:30:01.0000] <AryehGregor>
(DOM Range already has one, it's been pointed out)

[12:31:00.0000] <AryehGregor>
(did anyone get my public-html message yet? does it have to be approved by moderators or something in addition to me giving permission to post it?)

[12:45:00.0000] <AryehGregor>
/me wonders why mailing list archives are not conventionally updated in real time

[12:58:00.0000] <Hixie>
annevk: can we get an overload for appendChild() that takes a DOMString and creates a text node for you?

[12:58:01.0000] <AryehGregor>
That would be cool.

[12:58:02.0000] <Ms2ger>
Hixie, I suggest you try to convince implementors directly :)

[12:59:00.0000] <AryehGregor>
Although frankly, if we're going to add shortcuts to the DOM methods, we need to go a heck of a lot further than that.

[12:59:01.0000] <AryehGregor>
They're a massive headache to use.

[13:00:00.0000] <AryehGregor>
As it stands at least they're predictable, so adding a few random shortcuts doesn't seem like it would be a big win.

[13:29:00.0000] <TabAtkins>
Agreed with Aryeh on the "let's just sit down and make the entire DOM less sucky".  ^_^

[13:30:00.0000] <AryehGregor>
Do you have a concrete proposal?  Maybe "clone the most commonly used parts of jQuery"?  :)

[13:30:01.0000] <TabAtkins>
That would be a good start, certainly.

[13:30:02.0000] <TabAtkins>
SVG wants to make a less sucky SVG DOM, so we can coordinate and get everyone together.

[13:31:00.0000] <AryehGregor>
SVG has totally different needs.

[13:31:01.0000] <AryehGregor>
They need some kind of entirely non-DOM API that translates to DOM stuff.

[13:32:00.0000] <TabAtkins>
Likely, yes.

[13:32:01.0000] <TabAtkins>
But some things could be fixed, like automatically handling the namespacing hell.

[13:32:02.0000] <AryehGregor>
SVG is vector graphics shoehorned into XML, HTML is actually a natural fit for the DOM.

[13:32:03.0000] <Michael>
Have you guys seen Adobe Edge?

[13:33:00.0000] <Michael>
It finally became public today

[13:33:01.0000] <TabAtkins>
AryehGregor: Vector graphics are naturally layer-based, which seems like an acceptable fit for a tree-based language.

[13:33:02.0000] <smaug____>
huh, please, no jQuery

[13:34:00.0000] <jgraham>
Hopefully not the sucky magic bits of jQuery

[13:34:01.0000] <TabAtkins>
Depends on your defintion of "sucky magic".

[13:34:02.0000] <AryehGregor>
I was thinking of just looking at things that take five lines of DOM vs. one line of jQuery.

[13:34:03.0000] <jgraham>
Like "oh you entered a string that matched (some regexp) so you probably meant x"

[13:34:04.0000] <AryehGregor>
Not actually trying to copy it.

[13:34:05.0000] <jgraham>
In function parameters

[13:34:06.0000] <smaug____>
though, perhaps I'm against jQuery just because the implementation is ... less-than-perfect

[13:34:07.0000] <AryehGregor>
Heh, probably.

[13:35:00.0000] <miketaylr>
much like the DOM

[13:35:01.0000] <miketaylr>
;)

[13:35:02.0000] <Michael>
smaug____, Have you told the jQuery devs?

[13:35:03.0000] <jgraham>
smaug____: By that token is there any software you ar for ? :)

[13:35:04.0000] <AryehGregor>
Did anyone get my last post to public-html, in the thread about innerText?

[13:35:05.0000] <AryehGregor>
I haven't been able to figure out whether sending from ayg⊙an worked or not.

[13:35:06.0000] <smaug____>
to jresig yes

[13:36:00.0000] <Michael>
What'd he say?

[13:36:01.0000] <smaug____>
not much

[13:36:02.0000] <Michael>
I mean... Did you say "jQuery sucks" or did you pin point areas for improvement?

[13:36:03.0000] <smaug____>
I just CC'ed him to some bugs to show that jQuery causes mem usage to go high

[13:36:04.0000] <smaug____>
very high in some cases

[13:36:05.0000] <Michael>
hmm

[13:37:00.0000] <Michael>
Do other libs like prototype, mootools etc?

[13:37:01.0000] <smaug____>
and jQuery used to have code which slowed down unload of pages a lot, but I think they did fix that

[13:37:02.0000] <smaug____>
don't know about other libs

[13:38:00.0000] <smaug____>
because when profiling, if some slow down has been caused by a js lib, it has usually been jQuery

[13:38:01.0000] <smaug____>
perhaps because it is used so often

[13:39:00.0000] <Michael>
I'd be curious to see the performance benchmarks for jQuery vs prototype etc

[13:39:01.0000] <Michael>
Just because JS itself is a single threaded bottleneck

[13:40:00.0000] <smaug____>
*IIRC*, jQuery also uses some constructs which tend to be hard to optimize in JS JITs (at least in some forms of jits).

[13:42:00.0000] <AryehGregor>
jQuery is an abstraction layer, it aims more for ease of use than performance.

[13:44:00.0000] <AryehGregor>
/me says this as someone who just found that in his own code, he was repeatedly doing getDescendants(document) on a huge document, which recursed through all nodes in the document at a stack depth of like twenty, when he could have written it imperatively in a few more lines and iterated over like twenty nodes instead

[13:44:01.0000] <AryehGregor>
/me just rewrote that and noticed a significant speedup, but there don't seem to be good JS profiling tools last he checked . . .

[13:49:00.0000] <jgraham>
"JS itself is a single-threaded bottleneck" - not really

[13:49:01.0000] <jgraham>
Some concurrency is avaliable via workers

[13:50:00.0000] <jgraham>
and the idea of making the DOM threadsafe or letting random webdevs loose with threads in the browser is... not appealing

[13:51:00.0000] <jgraham>
Also, javascript is generally pretty fast these days. Surprisingly so, even

[13:51:01.0000] <Hixie>
yet in some cases still lacking in what seed like obvious optimisations to me

[13:51:02.0000] <Hixie>
seem

[13:52:00.0000] <Hixie>
i recently optimised a for loop by a factor of two by factoring out some common computation with variables that were not assigned to in the loop

[13:52:01.0000] <Hixie>
the loop had no side-effects, called no user methods

[13:52:02.0000] <jamesr>
it's hard to tell that in JS

[13:53:00.0000] <Hixie>
so are all the other optimisations

[13:53:01.0000] <Hixie>
i was surprised that this one hadn't been done yet

[13:53:02.0000] <Moo-->
AryehGregor: google chrome comes with good performance monitoring tools nowadays

[13:54:00.0000] <AryehGregor>
Moo--, its JS profiler tends to give me useless results.

[13:54:01.0000] <AryehGregor>
I've tried it pretty recently.

[13:54:02.0000] <AryehGregor>
IIRC, IE10 was the only browser that had usable profiling when I tested it out, although I think at the time Firebug wasn't working with my Firefox version.

[13:55:00.0000] <Ms2ger>
Hixie, were you calling into the DOM?

[13:55:01.0000] <Hixie>
no

[13:55:02.0000] <Hixie>
it was a loop on a CanvasPixelARra

[13:55:03.0000] <Hixie>
CanvasPixelArray even

[13:55:04.0000] <Ms2ger>
Same thing, really

[13:55:05.0000] <Hixie>
who is that the same thing?

[13:55:06.0000] <Ms2ger>
/me has a wide definition of "DOM"

[13:56:00.0000] <Ms2ger>
I should have said "native code"

[13:56:01.0000] <Hixie>
CanvasPixelArray shouldn't be any more native than String

[13:57:00.0000] <Ms2ger>
Not sure why it'd be any less native than Node per spec

[13:57:01.0000] <Hixie>
Node has to interact with complicated machinery

[13:57:02.0000] <Hixie>
CanvasPixelArray does not

[13:58:00.0000] <Hixie>
(e.g. what you do to a Node could involve HTTP, CSS, image decoding, memory allocation, mutation events, etc etc etc)

[13:58:01.0000] <Ms2ger>
And people still expect that to be fast :)

[13:59:00.0000] <Hixie>
that doesn't affect my point

[13:59:01.0000] <Hixie>
which is that in the case i was talking about, i don't see why it should be that hard to optimise

[14:00:00.0000] <Hixie>
compared to other optimisations that have already been done

[14:01:00.0000] <Ms2ger>
Yeah, in this case it should

[14:06:00.0000] <jgraham>
Hixie: On all browsers? It's not like everyone has the same optimisations

[14:07:00.0000] <Hixie>
i do not recall on what browsers i tested

[14:08:00.0000] <Ms2ger>
Firefox might be better there, because we implement it as an ArrayBuffer

[14:21:00.0000] <gsnedders>
Hixie: Loop invariant code motion is actually not that easy to do with JS, because such a small amount can actually be proven to be loop-invariant

[14:28:00.0000] <Hixie>
what makes it hard to prove loop-invariancy?

[14:30:00.0000] <AryehGregor>
"But rather than actually calculating the dimensions of that hypothetical box, user agents are free to make a guess at its probable position."

[14:30:01.0000] <AryehGregor>
<3 CSS

[14:34:00.0000] <gsnedders>
Hixie: Doing static analysis of entire lexical scopes is expensive (so any function call becomes expensive), and the fact that objects can change their value in random places.

[14:38:00.0000] <Hixie>
gsnedders: sure but in this case there were no such problems :-)

[14:38:01.0000] <Hixie>
gsnedders: it was just math

[14:38:02.0000] <Hixie>
anyway

[14:42:00.0000] <Philip`>
Rather than proving hard optimisations are perfectly safe and coping with obscure edge cases, browsers ought to run an aggressive optimiser that's safe 99% of the time, and also run a completely independent clone of the page in parallel with a slow but perfectly safe JS engine

[14:43:00.0000] <Philip`>
then if the first one happens to go wrong, the divergence will be detected soon when the second one catches up with it, and the browser can switch to displaying the output of the second one

[14:43:01.0000] <TabAtkins>
And then stay on the slow one forever?

[14:43:02.0000] <Philip`>
so you'll eventually end up with the correct result but in most cases you'll get the result displayed much sooner

[14:44:00.0000] <Philip`>
It's a foolproof plan

[14:44:01.0000] <Hixie>
there's one pretty fatal flaw with this plan

[14:45:00.0000] <Hixie>
no wait, two.

[14:45:01.0000] <Hixie>
though i guess the second one is the same as the first.

[14:45:02.0000] <Hixie>
so one.

[14:45:03.0000] <Hixie>
(namely, that script can have side-effects)

[14:46:00.0000] <Hixie>
(the second one is that the script's side-effects include "taking time" which can be detected from script, but i guess you cuold make new Date() in the second script always return whatever it returned on that occurrence in the first script.)

[14:52:00.0000] <gsnedders>
Hixie: Firefox should do a lot if you're on a traced path. If you're in method JIT, not so much.

[14:53:00.0000] <jarek>
Hi

[14:53:01.0000] <jarek>
why iframe is allowed by default to navigate the parent window?

[14:53:02.0000] <jarek>
isn't this potentialy dangerous?

[14:53:03.0000] <TabAtkins>
It's only allowed to do so when it's same-origin as the parent, right?

[14:54:00.0000] <jarek>
TabAtkins: I'm afraid not

[14:54:01.0000] <jarek>
no… wait, I have same origin policy disabled in browser

[14:55:00.0000] <jarek>
let me check

[14:55:01.0000] <AryehGregor>
Navigating the parent is how things break out of frames, no?

[14:55:02.0000] <jarek>
http://en.wikipedia.org/wiki/Framekiller

[14:55:03.0000] <AryehGregor>
Isn't that one of the things <iframe sandbox> explicitly prevents?

[14:55:04.0000] <AryehGregor>
Basically, if you include an <iframe> you're letting it take over the page, yes.

[14:56:00.0000] <AryehGregor>
sandbox is designed to prevent that.

[14:56:01.0000] <jarek>
it looks like any iframe (even from different origin) can overwrite parent's top.location value

[14:56:02.0000] <AryehGregor>
Sounds right.

[14:56:03.0000] <Hixie>
overriding top.location isn't a security problem

[14:58:00.0000] <jarek>
this permission  could be used to redirect user to malicious website

[14:59:00.0000] <TabAtkins>
Only if you're already embedding the malicious website.

[15:03:00.0000] <Hixie>
jarek: yeah but the user could tell that he was going to a malicious website.

[15:03:01.0000] <Hixie>
jarek: because it would change the location bar

[15:04:00.0000] <Hixie>
jarek: the web's security model assumes that the user is aware of the origin of the current page and can determine if it's safe or not.

[15:04:01.0000] <Hixie>
whether that's a safe assumption or not is open to debate, but if it's not, we have much bigger problems than top.location being settable

[16:00:00.0000] <AryehGregor>
I have a box that's absolutely positioned to the right of the page, then another box inside it.  The inner box is wide and I want it to overflow outside the outer box, but to the left, not the right.

[16:00:01.0000] <AryehGregor>
In LTR.

[16:00:02.0000] <AryehGregor>
Any way to do that?

[16:00:03.0000] <AryehGregor>
/me pings TabAtkins, as for all his n00b CSS questions

[16:01:00.0000] <AryehGregor>
Actually, maybe I don't want that after all . . .

[16:01:01.0000] <TabAtkins>
Yo.

[16:01:02.0000] <AryehGregor>
/me settles for overflow: auto for now

[16:02:00.0000] <TabAtkins>
Is the inner box positioned?

[16:02:01.0000] <AryehGregor>
TabAtkins, click the first "Comments" button on the right: http://aryeh.name/tmp/editing/editing.html#the-forecolor-command

[16:02:02.0000] <AryehGregor>
See the giant table with color values?

[16:02:03.0000] <AryehGregor>
I'd prefer if that somehow stuck out to the left and extended the background or something.

[16:03:00.0000] <TabAtkins>
Ah, I see.

[16:03:01.0000] <TabAtkins>
If it's positioned, you can just position it relative to the right edge.  Since it's not, the only way to influence overflow behavior is through changing the @dir.

[16:03:02.0000] <TabAtkins>
(Or the CSS 'direction', with the same effect.)

[16:03:03.0000] <AryehGregor>
Which will have undesirable side effects, no?

[16:04:00.0000] <TabAtkins>
Yes, definitely.  Horrible side effects.

[16:04:01.0000] <AryehGregor>
Oh well.

[16:04:02.0000] <AryehGregor>
Scroll bar it is.

[16:04:03.0000] <TabAtkins>
Unless you wrap the rest of your content into another element with @dir set correctly.

[16:04:04.0000] <TabAtkins>
Or, wait, float:right will do it too.

[16:05:00.0000] <AryehGregor>
Hmm, interesting thought.

[16:06:00.0000] <AryehGregor>
That looks promising, thanks.

[16:06:01.0000] <AryehGregor>
/me fidgets

[16:08:00.0000] <AryehGregor>
Now if only I could suppress the border where it's on top of the outer box's background.

[16:08:01.0000] <TabAtkins>
/me wishes fervently again for vw and vh units, so you can say "max-width:100vw" in that situation and force scrollbars.

[16:08:02.0000] <TabAtkins>
Put a background color on the float?

[16:09:00.0000] <AryehGregor>
Already did.

[16:09:01.0000] <TabAtkins>
Is the change live?

[16:09:02.0000] <AryehGregor>
Yes.

[16:09:03.0000] <TabAtkins>
Oh!  I see what you mean.

[16:09:04.0000] <TabAtkins>
Misread your comment previously.

[16:10:00.0000] <TabAtkins>
You'd like a border around the "final shape" of the box.

[16:10:01.0000] <AryehGregor>
Yeah.

[16:10:02.0000] <TabAtkins>
Yeah, that'd be cool.

[16:10:03.0000] <AryehGregor>
But it strikes me as probably impossible.

[16:10:04.0000] <TabAtkins>
Well, without diving into a full graphics language, probably.  (SVG will be able to do it with shape-combining and then stroking.)

[16:11:00.0000] <AryehGregor>
Hmm, it might be doable.

[16:11:01.0000] <AryehGregor>
/me adds a couple of extra divs and tries carefully positioning them

[16:12:00.0000] <TabAtkins>
Ah, clever.  Dunno if you can get it done currently, but you'd be able to after Positioning gets done, by attaching the floating div's edges to different box edges.

[16:12:01.0000] <TabAtkins>
Or, wait!

[16:12:02.0000] <TabAtkins>
Another div, positioned on top, with a left:0;right:0;margin-right:100%;

[16:13:00.0000] <TabAtkins>
Or, hm.

[16:13:01.0000] <TabAtkins>
You may need a wrapper div around the float that's a normal block and a positioning root.

[16:13:02.0000] <TabAtkins>
So the margin will resolve its percentage propertly.

[16:21:00.0000] <AryehGregor>
Cool observation by Peter Beverloo: over 50% of web browsers by market share are now open-source (if you count Chrome as open-source).

[16:21:01.0000] <AryehGregor>
Yay.

[16:21:02.0000] <AryehGregor>
Argh.

[16:21:03.0000] <AryehGregor>
My post to public-html has not appeared.

[16:21:04.0000] <AryehGregor>
/me stabs it viciously

[16:22:00.0000] <smaug____>
well, Chrome isn't open source :p

[16:22:01.0000] <AryehGregor>
Pfft, trivial details.

[16:34:00.0000] <annevk>
maybe not appendChild(string), but there's something to say for new Text(...)

[16:35:00.0000] <annevk>
and maybe new Element(localNameWithNamespaceautomaticallyresolvingtoSVGMathOrHTML)

[16:38:00.0000] <TabAtkins>
annevk: Indeed!

[16:39:00.0000] <TabAtkins>
I still wonder if there's any way we could kick everything into a single namespace if we got SVG and MathML to agree to it.

[16:40:00.0000] <annevk>
there is

[16:40:01.0000] <annevk>
by breaking all existing SVG and Math content

[16:40:02.0000] <annevk>
or at least most of it

[16:40:03.0000] <TabAtkins>
I meant "reasonably".

[16:40:04.0000] <TabAtkins>
We *could* do anything if we ignore compat.

[16:41:00.0000] <annevk>
changing XML would be the other path

[16:41:01.0000] <annevk>
seems that would be kind of hard

[16:42:00.0000] <annevk>
apart from the local name clashes

[16:42:01.0000] <TabAtkins>
Are there localname clashes?

[16:42:02.0000] <TabAtkins>
/me didn't think SVG clashed with HTML at all, but isn't sure about MathML.

[16:43:00.0000] <heycam>
annevk, oh, great idea. I did worry that createElement(dosomethingwithnamespacesautomatically) wouldn't be compatible, but doing it for `new Element(...)` only might be a good place to have it

[16:44:00.0000] <TabAtkins>
And I guess it takes the script's current document as the element's document?

[16:44:01.0000] <heycam>
yeah

[16:44:02.0000] <heycam>
I imagine

[16:44:03.0000] <heycam>
I think someone ought to look at how we could resolve the differences between the similar elements of HTML and SVG (like <script>, <a>)

[16:44:04.0000] <TabAtkins>
While we're at it, BAG OF ATTRIBUTES OMG

[16:45:00.0000] <heycam>
:)

[16:45:01.0000] <annevk>
TabAtkins, video, audio, textArea

[16:45:02.0000] <annevk>
TabAtkins, style, script

[16:45:03.0000] <TabAtkins>
annevk: Those are Tiny, right?

[16:45:04.0000] <annevk>
TabAtkins, some are I guess

[16:45:05.0000] <annevk>
TabAtkins, dunno what's implemented in most browsers

[16:45:06.0000] <TabAtkins>
video/audio/textArea are.

[16:45:07.0000] <TabAtkins>
Browsers don't implement Tiny at all.

[16:45:08.0000] <heycam>
Opera does some

[16:46:00.0000] <TabAtkins>
style/script/a are shared, but I suspect are similar enough to unify.

[16:46:01.0000] <heycam>
but yes, we should move away from textArea and towards just CSS boxes full of text

[16:46:02.0000] <TabAtkins>
(And we should unify them anyway, for author understanding.)

[16:46:03.0000] <heycam>
should we add defer/async to svg script, for example? dunno.

[16:46:04.0000] <TabAtkins>
Yes, definitely.

[16:47:00.0000] <TabAtkins>
Any differences between SVG <script> and HTML <script> are platform bugs, imo.

[16:50:00.0000] <annevk>
I do think that would be a good idea

[16:50:01.0000] <annevk>
Someone should probably write up a more coherent plan

[16:51:00.0000] <TabAtkins>
I wrote it on my whiteboard as a todo.  Does that count?

[16:52:00.0000] <heycam>
TODO: Fix the Web platform

[16:52:01.0000] <annevk>
TabAtkins, sorry

[16:53:00.0000] <annevk>
heycam, that's like bug #1 now plugins are "destroyed"


2011-08-02
[17:45:00.0000] <astearns>
Buttons (and other form elements) are non-replaced elements in the current spec. Did they used to be replaced elements?

[17:46:00.0000] <jamesr>
what on earth is SVG <script>?

[17:46:01.0000] <jamesr>
just <script> in the SVG namespace?

[17:50:00.0000] <heycam>
yes

[17:51:00.0000] <heycam>
but slightly different, of course!

[17:51:01.0000] <heycam>
:)

[17:51:02.0000] <heycam>
it uses xlink:href, but we're moving away from xlink for SVG2

[17:54:00.0000] <Dashiva>
astearns: Not as far as I know

[17:54:01.0000] <jamesr>
what's the difference between xlink:href and href?

[17:55:00.0000] <heycam>
nothing practical

[17:56:00.0000] <Dashiva>
Except being different

[17:56:01.0000] <TabAtkins>
It's in the xlink namespace.

[17:57:00.0000] <heycam>
can't remember for <script>, but at least for <a> you have the SVG DOM kind of DOM accessors

[17:57:01.0000] <heycam>
like a.href.baseVal

[18:16:00.0000] <TabAtkins>
Okay, MathML only collides with "<image>".

[18:16:01.0000] <TabAtkins>
SVG collides with <a>, <font>, <image>, <script>, <style>, and <title>.

[18:17:00.0000] <TabAtkins>
<a>, <script>, and <style> should just unify; that's pretty easy to argue.

[18:18:00.0000] <TabAtkins>
<image> is roughly the same.  Unification is possible.

[18:19:00.0000] <TabAtkins>
<font> and <title> very clearly can't unify.

[18:19:01.0000] <TabAtkins>
Darn.

[18:20:00.0000] <TabAtkins>
MathML's <image> is completely different.

[18:21:00.0000] <TabAtkins>
Possibly SVG2 could switch to <img>, leaving <image> for MathML.

[18:21:01.0000] <heycam>
well

[18:21:02.0000] <TabAtkins>
And the magical "new Element()" function could just switch based on that.

[18:21:03.0000] <heycam>
that's not going to help for existing content

[18:21:04.0000] <heycam>
or is this only for SVG-in-text/html?

[18:22:00.0000] <TabAtkins>
I dunno.

[18:26:00.0000] <TabAtkins>
heycam: At the very least, this would help with whatever-in-HTML.  But it would be great if we could drop namespaces in general for SVG2 and MathML.

[18:27:00.0000] <heycam>
"in general" is a big ask. :) would it be XML but all elements in the null namespace?

[18:28:00.0000] <TabAtkins>
That's one possibility.

[18:29:00.0000] <TabAtkins>
(I'm shooting for "in general" because I don't want inconsistance between SVG-in-HTML and SVG-in-HTML-in-SVG.

[18:29:01.0000] <heycam>
I agree that inconsistency would be bad

[18:36:00.0000] <TabAtkins>
heycam: I know that maintaining backwards-compat is a goal for SVG2 (though I'm unsure how strongly you guys feel about it), but if we dropped that, made some breaking changes for SVG2, and then declared that SVG2 and HTML were BFFs, that may help some of the integration issues.

[18:38:00.0000] <heycam>
TabAtkins, I think it's worth raising as a possibility, anyway

[18:48:00.0000] <Hixie>
if you set an expiry date on a resource fetched over http, the browser won't hit the network when using it right?

[18:51:00.0000] <Hixie>
> I wasn't very clear when I say "the main page should not be cached".

[18:51:01.0000] <Hixie>
> I was saying, we should still keep the main page cached,

[18:51:02.0000] <Hixie>
...

[18:58:00.0000] <othermaciej>
TabAtkins: sag:image does have some significantly different attributes and behavior compared to html:img

[18:58:01.0000] <othermaciej>
not sure it is practical

[18:59:00.0000] <othermaciej>
I could imagine giving alternate non-colliding names for the elements that do collied

[18:59:01.0000] <TabAtkins>
othermaciej: It's got x, y, and xlink:href, obviously, and the SVG DOM.  What else is going on with it?

[19:00:00.0000] <TabAtkins>
And then the various SVG-specific global attributes.

[19:00:01.0000] <TabAtkins>
(Which x and y might qualify as?)

[19:01:00.0000] <othermaciej>
in WebKit, SVG image generates a completely different kind of render object

[19:01:01.0000] <TabAtkins>
But in any case, the unification was from the point of (a) merging namespaces, and (b) making "new Element('foo')" work.

[19:01:02.0000] <othermaciej>
its layout behavior is not the same

[19:01:03.0000] <othermaciej>
new Element('foo') seems like a half-baked idea to me

[19:02:00.0000] <TabAtkins>
So having <svg><image/></svg> or new Element('image') make an SVG image, while <html><image /></html> makes an HTML img, is probably okay.

[19:02:01.0000] <othermaciej>
I do not see the advantage over new FooElement, and it is also kind of a lie, because normally it will not give you a direct instance of Element

[19:02:02.0000] <othermaciej>
I can't think of any other case in JS APIs where new FooBar returns something other than a direct instance of the FooBar class

[19:02:03.0000] <TabAtkins>
That's a good point.

[19:03:00.0000] <Hixie>
othermaciej: Image? :-)

[19:03:01.0000] <TabAtkins>
That's because the HTML element is misspelled.  ^_^

[19:03:02.0000] <Hixie>
(new Image() returns an HTMLImageElement which is an <img>)

[19:03:03.0000] <othermaciej>
Image and Option are exceptions in that they are aliases for what was later decided were the true names of those things

[19:03:04.0000] <Duke_>
:P

[19:04:00.0000] <othermaciej>
and I suppose Audio follows their pattern

[19:04:01.0000] <Hixie>
yeah i was gonna say, Audio was added knowingly :-)

[19:04:02.0000] <othermaciej>
unfortunately I don't think that pattern can be extended indefinitely because defining P, I, B and U in the global namespace would surely be a breaking change

[19:05:00.0000] <othermaciej>
however, having a constructor secretly be a factory function that produces instances of different things seems like very poor design to me

[19:05:01.0000] <othermaciej>
so I would still prefer new PElement to new Element('p')

[19:05:02.0000] <TabAtkins>
$('<p>')

[19:06:00.0000] <TabAtkins>
That looks remarkably like an emoji.

[19:06:01.0000] <othermaciej>
and I don't think the slightly more verbose new HTMLParagraphElement is all that bad

[19:06:02.0000] <othermaciej>
I don't really dig strings as APIs

[19:06:03.0000] <jamesr>
what about to make a raw element? new ElementElement()?

[19:06:04.0000] <Hixie>
speaking of emojis

[19:06:05.0000] <TabAtkins>
Nah, I don't either, except that it lets you set attributes on it as well.

[19:06:06.0000] <Hixie>
have you guys checked out the new Apple Colour Emoji font?

[19:06:07.0000] <jamesr>
or do you not do that?

[19:06:08.0000] <TabAtkins>
jamesr: new Element('')

[19:06:09.0000] <othermaciej>
but if you wanted something that short you could define a $p() function

[19:06:10.0000] <othermaciej>
the desire to make a raw element is rare enough that I don't care

[19:06:11.0000] <othermaciej>
Element intrinsically needs a tag name to exist

[19:07:00.0000] <othermaciej>
I'm satisfied with document.createElement() covering that use case

[19:07:01.0000] <Hixie>
U+1F533 WHITE SQUARE BUTTON in Apple's font is a round purple blob.

[19:07:02.0000] <TabAtkins>
If we can fix the constructors to allow setting attributes and content immediately, I'd be a lot happier.

[19:07:03.0000] <othermaciej>
almost always you want an HTMLElement anyway, not an Element

[19:07:04.0000] <TabAtkins>
Hixie: wut

[19:07:05.0000] <Hixie>
and U+1F537 LARGE BLUE DIAMOND is a gray diamond.

[19:07:06.0000] <jamesr>
but wouldn't new HTMLElement() make an <html> in that API?

[19:08:00.0000] <zewt>
since when do unicode characters specify color

[19:08:01.0000] <Hixie>
since the emoji set!

[19:08:02.0000] <othermaciej>
currently you make one with new HTMLHtmlElement

[19:08:03.0000] <othermaciej>
so I would expect new HtmlElement to be the new-style version, assuming we even cared

[19:08:04.0000] <zewt>
i expect almost no font renderers would even be capable of that, since color and glyphs should be fully orthogonal

[19:08:05.0000] <othermaciej>
s/currently/if we followed the interface names literally/

[19:08:06.0000] <Hixie>
zewt: Mac OS X Lion has a colour truetype font

[19:09:00.0000] <Hixie>
zewt: it's pretty sweet

[19:09:01.0000] <othermaciej>
since currently you obviously use createElement if you use anything

[19:09:02.0000] <othermaciej>
if we did want a tag-dispatching version I'd do Element.create(tagName, attributes…)

[19:09:03.0000] <othermaciej>
that seems more obviously a factory method, doesn't lie and claim to be a constructor

[19:10:00.0000] <Hixie>
Element.create() would be nice

[19:10:01.0000] <othermaciej>
and Element.create('p') is not materially more verbose than new Element('p')

[19:10:02.0000] <Hixie>
Element.create(tagName /* picks namespace magically */, {attributes}, [children...])

[19:10:03.0000] <othermaciej>
and if we cared we could add Element.createNS()

[19:11:00.0000] <othermaciej>
that's a neat version, being able to include the children

[19:11:01.0000] <TabAtkins>
We dont' care, and shouldn't do it.  >_<

[19:11:02.0000] <TabAtkins>
Mix in "new Text('foobar')" and I'm sold.

[19:12:00.0000] <TabAtkins>
Or just make strings auto-wrap themselves into textnodes when inserted into the DOM.

[19:12:01.0000] <othermaciej>
it could be an escape hatch for cases of HTML/SVG/MathML namespace conflicts

[19:12:02.0000] <othermaciej>
or give .create() an optional fourth namespace attribute

[19:12:03.0000] <TabAtkins>
othermaciej: I'd rather resolve the conflicts. ^_^

[19:12:04.0000] <Hixie>
i really think we should just make strings work

[19:12:05.0000] <othermaciej>
I dunno how practical it is

[19:12:06.0000] <Hixie>
the only mathml conflict, iirc, is <image>

[19:12:07.0000] <Hixie>
and that's a three-way conflict

[19:13:00.0000] <othermaciej>
at the very least all these elements would need magical rules about what renderer is created depending on whether they arein SVG or HTML content

[19:13:01.0000] <TabAtkins>
Hixie: That's right, I just checked.

[19:13:02.0000] <zewt>
still, unicode specifying colors on glyphs just seems out of scope

[19:13:03.0000] <Hixie>
(i had to look all this stuff up when doing the parser)

[19:13:04.0000] <othermaciej>
hmm, I suppose .create() can handle namespaces by providing xmlns in the list of attributes

[19:13:05.0000] <Hixie>
for the dom api, i'd just make 'image' map to the svg one

[19:13:06.0000] <Hixie>
'a' is more troublesome

[19:13:07.0000] <othermaciej>
it won't actually affect descendants unless you use innerHTML on it

[19:13:08.0000] <Hixie>
what are the other conflicts?

[19:14:00.0000] <TabAtkins>
a, font, image, script, style, title

[19:14:01.0000] <TabAtkins>
a, script, and style should be unified.

[19:14:02.0000] <TabAtkins>
font and title pretty much can't be.

[19:14:03.0000] <Hixie>
font=>svg, image=>svg, script=>html, style=>html

[19:14:04.0000] <Hixie>
(for this api)

[19:14:05.0000] <Hixie>
a and title are tough

[19:14:06.0000] <Hixie>
title kinda doesn't matter

[19:14:07.0000] <TabAtkins>
How often do you actually create <title> elements in HTML via script>

[19:15:00.0000] <Hixie>
nobody generating dynamic stuff this way is going to use either the html or svg one

[19:15:01.0000] <othermaciej>
you can pick either default and use xmlns in the attributes dictionary as an escape hatch

[19:15:02.0000] <Hixie>
a is a bigger problem

[19:15:03.0000] <TabAtkins>
(Probably SVG should drop <title> in favor of global @title.)

[19:15:04.0000] <Hixie>
how bad is merging svg <a> and html <a>?

[19:15:05.0000] <Hixie>
can we make svg just support the html <a> where appropriate?

[19:15:06.0000] <othermaciej>
Elemet.create('a', {'xmlns': svgNS, 'xlink:href', '….'}, new Text('xxx'))

[19:15:07.0000] <TabAtkins>
othermaciej: Or continue to ignore NS, but have a fourth paramter that just accepts "html | svg | mathml", with possible future extensions.

[19:16:00.0000] <Hixie>
or just have the tagname be optionally prefixed by "svg:" to force it into svg

[19:16:01.0000] <TabAtkins>
othermaciej: Since the NS itself is dumb, and we just need to know what language you're talking about.

[19:16:02.0000] <Hixie>
and then have sane defaults

[19:16:03.0000] <othermaciej>
since NS is rarely needed in the first place it would be nice not to be in the business of defining a translation table

[19:16:04.0000] <othermaciej>
just pick sane defaults and have a way to override

[19:16:05.0000] <Hixie>
yeah maybe just default to html and have svg: and mathml: prefixes

[19:16:06.0000] <TabAtkins>
I don't see why the override mechanism should continue to be difficult.

[19:16:07.0000] <Hixie>
or svg: and m:

[19:17:00.0000] <othermaciej>
I like the idea of Element.create()

[19:17:01.0000] <TabAtkins>
But predefined prefixes work too.

[19:17:02.0000] <othermaciej>
I should post to public-webapps

[19:17:03.0000] <Hixie>
Element.create('svg:a', {href: 'example.html'}, ['Click Me! Yay bad link text!']);

[19:18:00.0000] <othermaciej>
your idea is that strings work in the children list?

[19:18:01.0000] <Hixie>
yeah

[19:18:02.0000] <othermaciej>
I kinda like that

[19:18:03.0000] <othermaciej>
the children list could just be rest args instead of an array

[19:18:04.0000] <othermaciej>
would save the need to allocate an array to call this API

[19:18:05.0000] <Hixie>
i'd like strings to implictly create text nodes anywhere nodes are accepted

[19:18:06.0000] <Hixie>
rest args works for me except that often you'll have an array

[19:18:07.0000] <Hixie>
so we'd need to support both

[19:19:00.0000] <othermaciej>
Element.create('svg:a', {href: 'example.html'}, 'Click Me! Yay bad link text!', Element.create('svg:image'….))

[19:19:01.0000] <Hixie>
(think of <ol> where your child list might be an array to which you .push()ed a lot of <li>s)

[19:19:02.0000] <othermaciej>
(actually that may be illegal SVG but you get the idea)

[19:19:03.0000] <othermaciej>
you can let arrays be in the rest args and unpack them to one level

[19:19:04.0000] <othermaciej>
then you can have a fixed list followed by an array

[19:19:05.0000] <Hixie>
lgtm

[19:19:06.0000] <othermaciej>
or array in the middle

[19:20:00.0000] <Hixie>
ship it

[19:21:00.0000] <TabAtkins>
That sounds really good.

[19:23:00.0000] <TabAtkins>
othermaciej: You gonna write it up?  If you don't have time, I'll do it.

[19:24:00.0000] <othermaciej>
writing it up briefly

[19:24:01.0000] <TabAtkins>
kk

[19:30:00.0000] <othermaciej>
sent

[00:21:00.0000] <hsivonen>
:-( looks like manu-db filed the absence of some RDFa misfeatures as bugs against Microdata on Saturday

[00:23:00.0000] <Hixie>
feedback isn't bad

[00:43:00.0000] <hsivonen>
Hixie: in other news, I find it strange that you didn't just go ahead and fix http://www.w3.org/Bugs/Public/show_bug.cgi?id=12267 as a no-brainer

[00:43:01.0000] <annevk>
I guess the other thing we could do is have new Element("html", "img")

[00:43:02.0000] <annevk>
well

[00:43:03.0000] <annevk>
that might as well be new HTMLElement("img")

[00:43:04.0000] <annevk>
so I guess that's not the way to go

[00:44:00.0000] <hsivonen>
annevk: ooh. radical. are factories no longer fashionable for the DOM?

[00:44:01.0000] <Hixie>
hsivonen: that bug is the furthest thing from a no-brainer that i've had to deal with all year

[00:44:02.0000] <Hixie>
hsivonen: you sure you got the right number?

[00:44:03.0000] <annevk>
hsivonen, yeah, somewhat

[00:44:04.0000] <annevk>
hsivonen, e.g. we introduced new Event(...)

[00:46:00.0000] <hsivonen>
Hixie: yes, I got the right number

[00:46:01.0000] <Hixie>
hsivonen: it's not a no-brainer. i don't think anyone else has ever suggested that it is.

[00:46:02.0000] <Hixie>
i mean just look at roc's last comment.

[00:48:00.0000] <annevk>
Oh, there's a mailing list proposal for Element.create()

[00:50:00.0000] <hsivonen>
annevk: is proposing factory patterns for stylistic similarity with legacy?

[00:51:00.0000] <hsivonen>
annevk: now that we no longer believe in Web DOM having to work with separate interfaces and multiple independent backing implementations in the same process, is there a technical reason for needing factories anymore?

[00:59:00.0000] <zcorpan>
dudes, did you check prior art before designing element creation api?

[00:59:01.0000] <zcorpan>
prior art being jquery and other libraries

[01:00:00.0000] <Hixie>
i assumed anne would do that, personally

[01:03:00.0000] <annevk>
hsivonen, the reason given was that it would not actually create an object of type Element

[01:03:01.0000] <hsivonen>
annevk: oh, right

[01:04:00.0000] <hsivonen>
/me needs to wake up properly 

[01:04:01.0000] <annevk>
me too

[01:04:02.0000] <hsivonen>
so Alice and Bob seem to come from A and B. What's the story behind Nadia and Dirk?

[01:05:00.0000] <annevk>
I'm sort of afraid people are only going to implement new features of DOM Core and not care much about aligning on the legacy bits

[01:05:01.0000] <annevk>
That's what happened with XMLHttpRequest

[01:07:00.0000] <hsivonen>
I had assumed Nadia and Dirk were just the TAG's favorite scenario actors, but I just saw Nadia show up in an accessibility scenario, too

[01:08:00.0000] <hsivonen>
are Nadia and Dirk from some seminal book about scenarios that I haven't read?

[01:09:00.0000] <annevk>
Maybe everyone just copies from each other?

[01:23:00.0000] <Hixie>
othermaciej: since nodes get adopted silently now anyway, you basically never need to explictly create a node in another doc

[01:24:00.0000] <othermaciej>
Hixie: hmm, good point

[01:24:01.0000] <annevk>
WebUSB https://bugzilla.mozilla.org/show_bug.cgi?id=674718

[01:25:00.0000] <annevk>
I wonder if that is the right abstraction layer. Does not seem likely USB will be around forever... And there's lots of other ways to connect hardware too...

[01:25:01.0000] <Hixie>
using the word "steal" in the initial description seems like a bad idea in that context ;-)

[01:25:02.0000] <Hixie>
(given the many security implications here)

[01:33:00.0000] <hsivonen>
hmm. the bug description doesn't go in too much detail about use cases, so it's hard to assess if the abstraction is right

[01:49:00.0000] <annevk>
Hixie, so if we do Element.create() I guess we should have appendChild("string"), insertBefore("string", child) etc. as well for consistency?

[02:34:00.0000] <hsivonen>
when I saw the URL http://www.html5la.com/ I thought someone had set up an analog of MPEG-LA

[02:59:00.0000] <annevk>
is there a bug on Gecko on implementing Event constructors?

[02:59:01.0000] <smaug____>
IIRC no

[02:59:02.0000] <annevk>
I'll file one

[03:00:00.0000] <annevk>
would be nice if they got traction so init*Event can die in a fire

[03:00:01.0000] <smaug____>
I won't implement them before I know how to handle mouse events

[03:00:02.0000] <annevk>
What is the problem with mouse events?

[03:01:00.0000] <smaug____>
well, which all properties can be initialized

[03:01:01.0000] <smaug____>
and to which document/window is the event bound to

[03:01:02.0000] <smaug____>
since that affects to pageX/Y values

[03:02:00.0000] <annevk>
all properties can be initialized

[03:02:01.0000] <hsivonen>
doesn't JS |new| always occur in the context of a script global object i.e. window?

[03:02:02.0000] <annevk>
pageX/Y have defaults?

[03:02:03.0000] <annevk>
smaug____, same as for Image()

[03:03:00.0000] <smaug____>
if you initialize only clientX/Y, pageX/Y are calculated from those values

[03:03:01.0000] <hsivonen>
(I suppose there might be uses for a parent document creating event objects associatied with a child window, though)

[03:03:02.0000] <annevk>
Workers might work differently

[03:03:03.0000] <annevk>
smaug____, wow weird

[03:04:00.0000] <smaug____>
annevk: indeed

[03:04:01.0000] <smaug____>
and the web relies on that behavior

[03:04:02.0000] <annevk>
smaug____, for init*Event it does

[03:04:03.0000] <annevk>
smaug____, event constructors can work more sanely

[03:05:00.0000] <smaug____>
that is true

[03:05:01.0000] <annevk>
DOM 3 Events has nothing on this of course

[03:05:02.0000] <smaug____>
though, it would create yet another inconsistent API

[03:05:03.0000] <smaug____>
DOM 3 events doesn't specify pageX/Y

[03:05:04.0000] <annevk>
right

[03:05:05.0000] <smaug____>
it should be in cssom-view if somewhere

[03:06:00.0000] <annevk>
cssom-view just had MouseEvent as a temporary stopgap

[03:06:01.0000] <annevk>
mouse events should really be defined in one place

[03:07:00.0000] <smaug____>
so should Event, but somehow it is now in two places :p

[03:07:01.0000] <annevk>
yeah, I think DOM 3 Events should drop it and rename itself to UI Events or some such

[03:08:00.0000] <annevk>
I think keeping event constructors consistent with themselves is worth more than keeping them consistent with init*Event

[03:08:01.0000] <annevk>
especially with weird legacy proprietary behavior and especially since with event constructors you can set pageX/Y yourself

[03:09:00.0000] <smaug____>
pageX/Y behavior is not proprietary

[03:09:01.0000] <annevk>
which spec?

[03:09:02.0000] <smaug____>
I mean,it is, IIRC, implemented by all

[03:10:00.0000] <smaug____>
though, haven't tested that for years

[03:10:01.0000] <smaug____>
/me should go back to think about mutation event replacement

[03:11:00.0000] <annevk>
sure, it needs to become part of a spec

[03:11:01.0000] <smaug____>
there still isn't a good enough proposal :(

[03:11:02.0000] <annevk>
for mutation events?

[03:11:03.0000] <smaug____>
yeah

[03:11:04.0000] <smaug____>
er, proposal for the replament

[03:11:05.0000] <smaug____>
replacement

[03:12:00.0000] <annevk>
how does the replacement Mozilla initially proposed deal with Comment/ProcessingInstruction modifications?

[03:13:00.0000] <annevk>
kind of a pity progress on mutation event replacement halted a bit with so much conflicting requirements

[03:13:01.0000] <smaug____>
comment would be handled like textnodes

[03:13:02.0000] <annevk>
I can add any agreed on proposal to DOM Core at this point I think

[03:14:00.0000] <smaug____>
unfortunately there isn't such yet

[03:16:00.0000] <smaug____>
in my initial implementation PI would be handled like textnode too

[03:17:00.0000] <annevk>
that would make sense I think

[03:17:01.0000] <annevk>
treat PI as CharacterData except it does not implement any of its methods

[03:18:00.0000] <smaug____>
I wonder if PI could inherit CharacterData

[03:18:01.0000] <annevk>
yes it can, but you was one of those arguing against that

[03:18:02.0000] <annevk>
were*

[03:19:00.0000] <smaug____>
that is how it is implemented in Gecko, we just expose different interface

[03:19:01.0000] <smaug____>
annevk: I was?

[03:19:02.0000] <smaug____>
that is possible

[03:19:03.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=12205

[03:20:00.0000] <annevk>
oh no

[03:20:01.0000] <annevk>
you were not

[03:20:02.0000] <smaug____>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=12205#c10

[03:21:00.0000] <JirkaK>
hi, just noticed <ol reversed> in spec. What are usescases for this bizzare attribute?

[03:21:01.0000] <annevk>
we could make it inherit and add another interface between CharacterData and Text/Comment

[03:22:00.0000] <annevk>
that exposes the modification methods

[03:22:01.0000] <Ms2ger>
JirkaK, lists that go backwards

[03:23:00.0000] <hsivonen>
https://plus.google.com/109022289777044064966/posts/j43r9owxqry ...and the violent language like "punch" enters the discussion about Microdata and RDFa

[03:24:00.0000] <Ms2ger>
Wow, that took long

[03:24:01.0000] <JirkaK>
Ms2ger, well, yes. but is it worth new attribute? It's impossible to render such list progressively and feels more like styling thing anyway

[03:24:02.0000] <hsivonen>
s/ the//

[03:25:00.0000] <Ms2ger>
See http://blog.whatwg.org/reverse-ordered-lists, for example

[03:26:00.0000] <heycam>
Hixie, s/pedancy/pedantry/ in your recent commit message, no? :)

[03:26:01.0000] <Zeddy>
i have a problem conserning the cache manifest, im building a webpage for mobile devices which lets the user download all the required files for offline use, but the content can be in different languages depending on the url mapviewer.php?lang=1 ?lang=2 and so on. This results in double downloading if the user visits both languages, which obviously is the way it should work.. but i would like

[03:26:02.0000] <Zeddy>
to delete all cached files when the user visits another language, otherwise the amount of cached data will exeed the limitations that mobile browsers have

[03:26:03.0000] <Zeddy>
http://www.kokonniemi.fi/guide/index.php/view/map/1/1/1 <- finnish language

[03:27:00.0000] <Zeddy>
http://www.kokonniemi.fi/guide/index.php/view/map/1/2/1 <- english language

[03:27:01.0000] <Zeddy>
visit the first and you have 1.4 Mb cached, then visit the second and you have 2.8 Mb cached

[03:28:00.0000] <Ms2ger>
And they're all comments from foolip

[03:28:01.0000] <erlehmann>
Zeddy, how can it be SO MUCH?

[03:28:02.0000] <erlehmann>
i mean … text?

[03:29:00.0000] <Zeddy>
what text?

[03:31:00.0000] <JirkaK>
Of course, I did some research on <ol reversed>. I think that's more harmful then useful, it's not backward compatible -- it will show wrong numbers in old browsers. I'm thinking about filling bug for removing it. But before that I would like to know whether I haven't miss anything.

[03:38:00.0000] <hsivonen>
I suppose social norms say that I should be OK with Nathan announcing that he's going to punch people who ask about use cases when he says it right after I've asked him about use cases.

[03:39:00.0000] <hsivonen>
yet, I still think it's sad that discussion has come to this (though I don't really expect him to hit me if within reach)

[03:40:00.0000] <foolip>
hsivonen, I read it as "please don't discuss specs with me"

[03:40:01.0000] <foolip>
which is easy enough to not do

[03:40:02.0000] <annevk>
Julian even +1'd it

[03:41:00.0000] <foolip>
but yes, I also interpreted as a response to you, hsivonen, which is not OK

[03:41:01.0000] <zcorpan>
speaking of <ol reversed>, css still doesn't have an equivalent feature yet right?

[03:42:00.0000] <annevk>
zcorpan, don't think so

[03:42:01.0000] <annevk>
zcorpan, reversed counters would be pretty awful

[03:43:00.0000] <Ms2ger>
That makes it rather surprising the CSSWG hasn't added it yet

[03:46:00.0000] <annevk>
ouch

[03:48:00.0000] <jgraham>
hsivonen: "Violence is the last refuge of the incompetent" and all that. I took it as a sign that he can't actually justify anything he wants by reason and so should be ignored.

[04:01:00.0000] <annevk>
so we had Microdata and RDFa and now there's two new proposals plus a bunch of change requests?

[04:01:01.0000] <annevk>
927 is the new 386

[04:02:00.0000] <hsivonen>
annevk: indeed

[04:04:00.0000] <jgraham>
Maybe there should be a short test before you file a bug to ensure that you have read the right comics

[04:05:00.0000] <Ms2ger>
Would that stop them?

[04:05:01.0000] <jgraham>
No, but the idea of testing people on comic knowledge before they can file bugs amuses me

[04:05:02.0000] <hsivonen>
jgraham: see also https://plus.google.com/114490712483753086051/posts/RgEF6RBXH6D

[04:06:00.0000] <jgraham>
hsivonen: Heh

[04:12:00.0000] <foolip>
annevk, the RDF crowd have begun looking at microdata in detail, which I would consider a good thing

[04:12:01.0000] <annevk>
Yeah I guess

[04:13:00.0000] <foolip>
unless all their suggestions are rejected and they just more pissed off...

[04:13:01.0000] <foolip>
get

[04:13:02.0000] <annevk>
I'm with the last paragraph (before the addendum) of http://hsivonen.iki.fi/schema-org-and-communities/

[04:15:00.0000] <foolip>
certainly, I expect that search engines would give little weight to hidden microdata and only use if it seems to match up with everything else

[04:23:00.0000] <annevk>
there is now some Pokemon stuff in my WHATWG spark

[04:23:01.0000] <foolip>
aren't sparks just google news feeds for a keyword?

[04:23:02.0000] <foolip>
they seem extremely useless to me

[04:27:00.0000] <hsivonen>
I think I saw someone saying that sparks are Semantic

[04:28:00.0000] <Zeddy>
did anyone perhaps answer my question while i was away?

[04:29:00.0000] <Ms2ger>
hsivonen++

[04:31:00.0000] <foolip>
hsivonen, I think you mean this: http://semanticweb.com/thoughts-on-google-plus-the-magic-isnt-social-its-semantic_b21852

[04:32:00.0000] <hsivonen>
foolip: yeah, that was what I was thinking about, though I didn't actually read the post fully. just saw the title and someone else's link to it

[04:33:00.0000] <foolip>
hsivonen, https://plus.google.com/102122664946994504971/posts/NRo1igagUGy

[04:42:00.0000] <Zeddy>
what happends if a domain has 9Mb cached, and then it starts caching something else, will the old caches be deleted when the 10Mb limit is reached or will it just stop caching the new stuff? O_oi

[04:42:01.0000] <zcorpan>
Hixie: regarding version history, sometimes it's handy to check the old snapshots in http://www.whatwg.org/specs/web-apps/ and do inline search

[04:43:00.0000] <zcorpan>
Hixie: i think you should publish such snapshots maybe once a year or so and link to them from the spec

[04:43:01.0000] <zcorpan>
Hixie: svn history is too much if you want to know which year canvas was added

[04:44:00.0000] <foolip>
zcorpan, would filtering the commit log per section solve the problem?

[04:45:00.0000] <zcorpan>
foolip: how would you do that?

[04:45:01.0000] <foolip>
zcorpan, by running the spec splitter on each commit and seeing which sections changed

[04:46:00.0000] <foolip>
I was working on a script to do that some time ago, but got bored

[04:46:01.0000] <zcorpan>
oooh. interesting

[04:46:02.0000] <zcorpan>
that would be pretty cool

[04:46:03.0000] <foolip>
It would help me not get surprised by changes to the video section that happened on seemingly unrelated commits

[04:46:04.0000] <zcorpan>
but i'd still want year snapshots

[04:46:05.0000] <jgraham>
Pretty sure I had a similar idea and got similarly bored :)

[04:47:00.0000] <foolip>
ok

[04:47:01.0000] <foolip>
I think I ended up filing bugs on the spec about how the outlining algorithm worked instead... :-/

[04:47:02.0000] <foolip>
(because I tried using it)

[04:48:00.0000] <zcorpan>
maybe the right solution is to have the multipage versions in svn

[04:49:00.0000] <Ms2ger>
There are svn blame snapshots at every thousandth commit somewher

[04:49:01.0000] <Ms2ger>
+e

[04:49:02.0000] <foolip>
I just use my git conversion and bisect on that

[04:52:00.0000] <gsnedders>
zcorpan: So snapshots are a good idea? :P

[04:53:00.0000] <annevk>
they are nice for spec archeology as version history is not always great

[04:53:01.0000] <foolip>
would having web access to every single commit be acceptable?

[04:53:02.0000] <annevk>
yeah

[04:54:00.0000] <annevk>
I was looking into that at some point

[04:54:01.0000] <foolip>
I guess that we can't recreate old versions as the toolchains have changed though, the spec splitter notably

[04:54:02.0000] <annevk>
I guess if you ignore external dependencies changing it might be okay

[04:55:00.0000] <Zeddy>
i just can't believe html5 cache manifest is missing a method for clearing the cache.

[04:59:00.0000] <annevk>
you can use HTTP for that

[05:09:00.0000] <zcorpan>
gsnedders: yes

[05:10:00.0000] <zcorpan>
gsnedders: i don't mind if the snapshotting is automated

[06:07:00.0000] <MikeSmith>
AryehGregor: about bugzilla and e-mail, please ping me when you're around

[06:20:00.0000] <annevk>
whoa MikeSmith da man

[06:20:01.0000] <annevk>
MikeSmith, Traversal can be removed from http://platform.html5.org/

[06:20:02.0000] <MikeSmith>
/me takes a look

[06:23:00.0000] <MikeSmith>
annevk: OK, removed

[06:23:01.0000] <annevk>
cool

[06:23:02.0000] <annevk>
it's in DOM Core now

[06:31:00.0000] <Zeddy>
:((

[06:31:01.0000] <Zeddy>
html5 has failed me

[06:32:00.0000] <cygri>
hi foolip, around? a Live Microdata question

[06:32:01.0000] <foolip>
cygri, always at your service!

[06:32:02.0000] <cygri>
i though that this microdata snippet would produce a property: http://pastebin.com/cr2bz5b2

[06:33:00.0000] <foolip>
it does, if you look at the JSON

[06:33:01.0000] <foolip>
perhaps you wonder why there's no Turtle?

[06:33:02.0000] <cygri>
no, i'm looking at the json and don't see a property

[06:34:00.0000] <cygri>
it shows an item with zero properties

[06:34:01.0000] <cygri>
at http://foolip.org/microdatajs/live/

[06:34:02.0000] <foolip>
I see this: http://pastebin.com/5hDEPzK9

[06:34:03.0000] <Philip`>
Maybe your browser's parser moves <meta> into the <head>?

[06:34:04.0000] <foolip>
cygri, are you using Firefox?

[06:35:00.0000] <foolip>
if so, Philip` is likely correct

[06:35:01.0000] <cygri>
safari. why would this make a difference?

[06:35:02.0000] <cygri>
/me tries other browsers

[06:35:03.0000] <foolip>
some browsers move <meta> during parsing

[06:35:04.0000] <foolip>
they shouldn't if they have an HTML5 parser though

[06:36:00.0000] <foolip>
not sure which "some browsers" are at the moment

[06:36:01.0000] <cygri>
but the markup is just form input in this case, no?

[06:36:02.0000] <foolip>
no, it's parsed in an iframe and then the resulting DOM is used

[06:37:00.0000] <foolip>
since I didn't feel like implementing an HTML parser in JavaScript...

[06:37:01.0000] <foolip>
and for the preview tab of course

[06:37:02.0000] <zcorpan>
hsivonen has already done that

[06:38:00.0000] <cygri>
ah i see. good point

[06:38:01.0000] <cygri>
it works in chrome

[06:38:02.0000] <cygri>
that explains it. good to know that meta in the body isn't quite safe to use yet.

[06:39:00.0000] <cygri>
thanks!

[06:39:01.0000] <foolip>
perhaps I should feature-detect this bug and warn about it

[06:40:00.0000] <Philip`>
Or look for any <meta itemprop>/<link itemprop> in <head>, so you only warn if the input has triggered that bug?

[06:41:00.0000] <cygri>
Philip`++

[07:13:00.0000] <foolip>
cygri, can you try http://foolip.org/microdatajs/live/ now to see if you get a warning about the bug?

[07:13:01.0000] <foolip>
I'll just always warn for it, since I have no way of knowing that the author didn't actually put <meta itemprop> in <head>

[07:14:00.0000] <Zeddy>
Can i use Session variables in the manifest file to get dynamic manifest files but with the same URI ?

[07:16:00.0000] <annevk>
what are session variables?

[07:16:01.0000] <annevk>
if you make your manifest dynamic you lose all benefit it has

[07:18:00.0000] <cygri>
foolip, i can't see any change in safari

[07:18:01.0000] <Zeddy>
annevk, please do explain

[07:19:00.0000] <Zeddy>
that makes no sense whatsoever

[07:19:01.0000] <cygri>
foolip, but in firefox i do see the warning

[07:19:02.0000] <foolip>
cygri, does is <meta> moved to <head> in http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1091 using safari?

[07:19:03.0000] <foolip>
if not, then it must be some other bug

[07:20:00.0000] <Zeddy>
i have a mobile guide for mobile phones, where the user can select in what language they want their guide to be in, i only want to cache the selected language data

[07:20:01.0000] <Zeddy>
each language can have different images and ofcourse text

[07:21:00.0000] <cygri>
foolip, yes the meta is inside head in safari, but i still don't see the warning

[07:21:01.0000] <zcorpan>
hmm. <video></video> is valid but <audio></audio> is not

[07:21:02.0000] <zcorpan>
wonder if that's intentional. Hixie ^

[07:21:03.0000] <foolip>
cygri, very odd, I'll have to find myself a machine with safari to test

[07:21:04.0000] <MikeSmith>
zcorpan: yeah, I wondered about that too

[07:21:05.0000] <foolip>
cygri, which version?

[07:22:00.0000] <cygri>
foolip, Version 5.0.5 (6533.21.1)

[07:22:01.0000] <foolip>
ok, will check tonight

[07:22:02.0000] <cygri>
foolip, wait a sec, migth be a js caching thing

[07:23:00.0000] <foolip>
that would explain it :)

[07:23:01.0000] <annevk>
Zeddy, if you change the manifest the UA will fetch all its listed resources again

[07:26:00.0000] <cygri>
foolip, yes that's it

[07:26:01.0000] <cygri>
it works now

[07:26:02.0000] <foolip>
cygri, yay :)

[07:26:03.0000] <cygri>
had to blow away the cache

[07:26:04.0000] <foolip>
so I guess Safari 5 is the newest?

[07:27:00.0000] <cygri>
i think so

[07:27:01.0000] <foolip>
a big red warning for all, then!

[07:27:02.0000] <cygri>
safari's js error console actually shows this:

[07:27:03.0000] <cygri>
<meta> is not allowed inside <body>. Moving <meta> into the <head>.

[07:27:04.0000] <cygri>
i should have checked there first :-)

[07:28:00.0000] <foolip>
well, looks like they need to update their parser

[07:28:01.0000] <foolip>
file a bug :)

[07:39:00.0000] <jgraham>
I wonder if the html5lib tests have been updated for that case… maybe I missed it

[07:39:01.0000] <cygri>
foolip, i tried to file a bug. it's too hard. looks like i have to file it against webkit, and i have no idea what version of webkit safari is using. i should probably install a safari/webkit nightly build, but that's too much work right now.

[07:39:02.0000] <cygri>
next time i come across some safari issue, i'll figure out how to install a nightly and do the right thing, i promise ;-)

[07:41:00.0000] <brucel>
WHATWG types: has there been any thoughts on extending fragment indentifiers to allow CSS selectors so you can link to arbitrary places on 3rd party documents (like http://simonstl.com/articles/cssFragID.html) ?

[07:44:00.0000] <beverloo>
I think there was a thread about that a year ago

[07:44:01.0000] <annevk>
every other year there's one :)

[07:45:00.0000] <brucel>
does it ever go anywhere? (eg, it it worth going through the logs?)

[07:45:01.0000] <annevk>
nah

[07:45:02.0000] <brucel>
fairy nuff

[07:46:00.0000] <annevk>
not much interest basically

[07:46:01.0000] <annevk>
there's more interesting stuff to work on

[07:48:00.0000] <foolip>
cygri, they probably know that they're not using an HTML5 parser, it's OK

[07:48:01.0000] <asmodai>
annevk: Will Opera ever support more of MathML for HTML 5?

[07:49:00.0000] <annevk>
I don't really know much about our MathML support

[07:49:01.0000] <annevk>
except that I don't think we should have implemented it using some sort of CSS profile that kind of, but not really, works

[07:49:02.0000] <asmodai>
annevk: Seems to parse it and display it roughly, but just plain text, nothing involving a real math look

[07:50:00.0000] <hsivonen>
cygri: Safari 5.0.5 has the old parser

[07:50:01.0000] <hsivonen>
cygri: 5.1 has big internal changes, though I haven't checked if it has the HTML5 parser

[07:51:00.0000] <hsivonen>
cygri: 5.1 is the latest Safari release

[07:51:01.0000] <annevk>
asmodai, that's what I meant, yes

[07:51:02.0000] <asmodai>
annevk: *nod*

[07:52:00.0000] <asmodai>
annevk: It would be so great to have more browsers support it

[07:52:01.0000] <annevk>
asmodai, make sure it is Acid tested

[07:52:02.0000] <hsivonen>
cygri: it seems 5.1 has the same parser as Chrome 8 and newer

[07:52:03.0000] <asmodai>
I know one group of folks that would love to be able to use more math stuff: gamedevs

[07:52:04.0000] <annevk>
asmodai, or design some kind of Math test, dunno

[07:52:05.0000] <asmodai>
annevk: mmm

[07:52:06.0000] <gsnedders>
asmodai: Basically we suppotr http://www.w3.org/TR/mathml-for-css/

[07:53:00.0000] <gsnedders>
*support

[07:53:01.0000] <jgraham>
Well ther are plenty of people who would use maths if they could

[07:54:00.0000] <hsivonen>
maybe Opera will allocate resources to MathML once IE adds MathML support

[07:54:01.0000] <asmodai>
jgraham: Oh of course

[07:54:02.0000] <asmodai>
jgraham: Just that I have seen them pick up on web stuff quickly.

[07:55:00.0000] <asmodai>
A lot of them are looking at how to use recent web tech to replace some tools/utilities they have now that only run on desktop.

[07:55:01.0000] <jgraham>
hsivonen: Maybe. It's not exactly something that people are clamouring for at the moment so it's not hard to see why it doesn't get priority

[07:55:02.0000] <hsivonen>
mathml support looks like burden for browsers as long as e.g. wikipedia doesn't use it

[07:56:00.0000] <jgraham>
I think people are pretty happy to do maths either as images (becase c+p doesn't work all that well anyway) or using e.g. MathJAX

[07:56:01.0000] <asmodai>
You see MathJAX pop up all over the place, yea

[07:56:02.0000] <jgraham>
e.g. I think Kahn Acadamy uses the latter

[07:56:03.0000] <gsnedders>
hsivonen: Yeah, which does make the CSS-expressable subset seem like a reasonable thing to impl

[07:57:00.0000] <asmodai>
although I like the layout stuff you can do with MathML

[07:57:01.0000] <jgraham>
gsnedders: Well it is easy. Which is nice. It kinda asucks compared to a proper implemntation, which is bad.

[07:57:02.0000] <asmodai>
e.g. some of the things I got in http://www.in-nomine.org/~asmodai/3d.html at the beginning

[07:58:00.0000] <hsivonen>
MathML is the kind of thing that is really hard to justify to someone who is making decisions about how to allocate paid effort in browser development

[07:58:01.0000] <asmodai>
/me makes a mental note to expand on that page again in the coming weeks.

[07:58:02.0000] <gsnedders>
jgraham: Indeed.

[07:59:00.0000] <hsivonen>
MathML needs to be contributed by someone who is on a math-specific mission and manages to get the patches accepted before someone stops to think about the ongoing maintenance of that code

[07:59:01.0000] <gsnedders>
hsivonen: Well, unless the paid effort is paid for, which I guess is why we have our impl

[07:59:02.0000] <cygri>
hsivonen, ok thanks, that's great to know. i'm still on os x 10.6.7, seems like they don't make safari 5.1 available for that.

[07:59:03.0000] <hsivonen>
gsnedders: do you mean an Opera embedding customer commissioned the CSS level support?

[08:00:00.0000] <gsnedders>
hsivonen: I don't know, but I guess so, because as you say, it's hard to justify paid development of.

[08:00:01.0000] <hsivonen>
cygri: Safari 5.1 is available for 10.6.8

[08:00:02.0000] <hsivonen>
cygri: and 10.6.8 is available for 10.6.7 :-)

[08:01:00.0000] <cygri>
hsivonen, yes i know. i'm one reboot away from that

[08:01:01.0000] <cygri>
but you know, in 2011, rebooting and restarting all programs is just as much work as reinstalling was in 1998 ;-)

[08:01:02.0000] <wilhelm>
How do mathematicans usually type complicated equations or functions? Are visual editors used, or do people type TeX or MathML directly?

[08:01:03.0000] <asmodai>
A lot use TeX.

[08:02:00.0000] <asmodai>
For MathML I used to use my hand-written code, but switch to firemath extension for firefox

[08:02:01.0000] <asmodai>
If I can take the formulae writing professors at the university I work for as an example of mathematicians.

[08:02:02.0000] <jgraham>
wilhelm: Mathematicians use LaTeX

[08:02:03.0000] <jgraham>
Biologists use MS Word

[08:03:00.0000] <jgraham>
(to make a hopefully-not-too-unfair generalisation)

[08:04:00.0000] <jgraham>
Insane people type MathML directly

[08:04:01.0000] <gsnedders>
hsivonen: Ah, the answer seems to actually be someone implemented it in their spare time.

[08:04:02.0000] <hsivonen>
who typesets in Mathematica?

[08:04:03.0000] <wilhelm>
Right. I've dealt with one biologist and one professor emeritus in mathematics so far regarding this issue.

[08:04:04.0000] <wilhelm>
(That's wearing my web developer hat, not my browser vendor hat.)

[08:04:05.0000] <jgraham>
People typeset in mathematica?

[08:05:00.0000] <asmodai>
The funny part is

[08:05:01.0000] <asmodai>
Microsoft Windows 7 has the math input panel

[08:05:02.0000] <asmodai>
which places data on the clipboard that is mathml apparently

[08:05:03.0000] <asmodai>
http://dpcarlisle.blogspot.com/2010/01/mathml-on-clipboard.html

[08:07:00.0000] <wilhelm>
So from what age can one expect mathematicans to know LaTeX? (Can I just wait for the emeritus to die and be replaced by someone younger, or do I have to implement some fancy editor for these people? :)

[08:07:01.0000] <asmodai>
<?xml version="1.0" encoding="utf-16" ?>

[08:07:02.0000] <asmodai>
<m:math xmlns:m="http://www.w3.org/1998/Math/MathML" xmlns:MicrosoftMathRecognizer="http://schemas.microsoft.com/mathrecognizer" >

[08:08:00.0000] <wilhelm>
(I'm doing some consulting work for an old-fashioned encyclopædia moving from paper to web. :)

[08:09:00.0000] <karlushi>
hixie: Re http://www.w3.org/Bugs/Public/show_bug.cgi?id=13452 itemid

[08:09:01.0000] <karlushi>
It is not about redefining itemid but about clarifying the role. Your statement for closing is true but completely orthogonal to what was suggested.

[08:09:02.0000] <karlushi>
There is also still a false statement in the spec.

[08:09:03.0000] <karlushi>
"The exact meaning of a global identifier is determined by the vocabulary's specification."

[08:09:04.0000] <karlushi>
something along "the specific types of uris usable as a global identifier is determined by the vocabulary's specification." is more exact.

[08:09:05.0000] <jgraham>
wilhelm: In my very limited experience people old enough to not know LaTeX also have secretaries (or grad students) to typeset for them

[08:11:00.0000] <wilhelm>
jgraham: Ah. How convenient.

[08:11:01.0000] <jgraham>
Things I have learnt today: very early Laura Marling has a slight Kate Nash/Lilly Allen twang

[08:20:00.0000] <wilhelm>
jgraham: How do people studying physics or chemistry (or any other relevant subject) type their formulas? Are they more similar to biologists or mathematicans? (c:

[08:22:00.0000] <gsnedders>
wilhelm: I believe my gf's answer to that (chem, FWIW) is "on paper" :)

[08:22:01.0000] <jgraham>
gsnedders: Well she isn't a grad student

[08:22:02.0000] <jgraham>
So isn't writing papers

[08:22:03.0000] <jgraham>
(probably)

[08:23:00.0000] <gsnedders>
True.

[08:23:01.0000] <gsnedders>
/me finds someone older to chase after

[08:24:00.0000] <jgraham>
wilhelm: 435

[08:24:01.0000] <wilhelm>
… 435?

[08:24:02.0000] <jgraham>
Just replace the axis label with "probablity of using LaTeX"

[08:25:00.0000] <jgraham>
And reduce the gap to mathematicians

[08:25:01.0000] <jgraham>
http://xkcd.com/435/

[08:25:02.0000] <wilhelm>
… Ah. (c:

[08:35:00.0000] <jgraham>
http://www.enisa.europa.eu/act/application-security/web-security/a-security-analysis-of-next-generation-web-standards

[08:37:00.0000] <annevk>
yeah not all of those are actual problems

[08:38:00.0000] <annevk>
like those raised on web messaging and CORS are not

[08:38:01.0000] <annevk>
afaict

[08:38:02.0000] <annevk>
I wonder if they are in need of funding

[08:41:00.0000] <MikeSmith>
everybody's in need of funding

[08:43:00.0000] <hsivonen>
wow. site figuring out which parts of a video served by the site the user watched is listed as a threat

[08:44:00.0000] <annevk>
the web messaging issue was at best editorial but is actually just author-facing text vs requirements confusion

[08:45:00.0000] <annevk>
the issue with CORS the issue is that they do not quite get how "bad" the situation with arbitrary content bodies already is

[08:46:00.0000] <karlushi>
everybody's in need of fonding

[08:46:01.0000] <karlushi>
Apple dict is interesting on that topic.

[08:46:02.0000] <karlushi>
fond |fänd|

[08:46:03.0000] <karlushi>
adjective [ predic. ] ( fond of)

[08:46:04.0000] <karlushi>
having an affection or liking for : I'm very fond of Mike

[08:47:00.0000] <jgraham>
Oh, I thought you typoed fondling

[08:47:01.0000] <karlushi>
I should have

[08:47:02.0000] <karlushi>
it would have been more me.

[08:47:03.0000] <karlushi>
but I would have to replace MikeSmith by a dog

[08:47:04.0000] <karlushi>
Apple dict: "stroke or caress lovingly or erotically : the dog came over to have his ears fondled"

[08:48:00.0000] <karlushi>
I had no idea that there was zoophilia in Apple dict

[08:49:00.0000] <MikeSmith>
heh

[08:50:00.0000] <MikeSmith>
dog fondling

[08:50:01.0000] <jgraham>
I'm sure MikeSmith would like having his ears stroked too

[08:51:00.0000] <wilhelm>
/me covers his eyes.

[08:52:00.0000] <MikeSmith>
heh

[08:52:01.0000] <AryehGregor>
/me pings MikeSmith about Bugzilla and e-mail

[08:52:02.0000] <MikeSmith>
aha

[08:53:00.0000] <MikeSmith>
AryehGregor: so I will set up the bugzilla component

[08:53:01.0000] <MikeSmith>
you still not getting mail as expected?

[08:53:02.0000] <Ms2ger>
Oh, if you're doing that, I'd like one too ;)

[08:54:00.0000] <MikeSmith>
Ms2ger: for DOM Range?

[08:55:00.0000] <Ms2ger>
Parsing

[08:55:01.0000] <Ms2ger>
http://html5.org/specs/dom-parsing.html

[08:56:00.0000] <AryehGregor>
jgraham, a math postdoc who taught a summer research thing I was in told me that he had colleagues who used MS Word for typesetting.  Apparently you can do it without driving yourself completely insane, if you memorize all the keyboard shortcuts.

[08:56:01.0000] <AryehGregor>
Although I assume it's much more limited.

[08:56:02.0000] <Ms2ger>
Madness

[08:57:00.0000] <AryehGregor>
(Incidentally, that math postdoc is now some sort of faculty at the University of Tokyo, and just e-mailed me in the last day telling me that the research I did with him that summer has been accepted for publication by the Journal of Number Theory.  Awesome.)

[08:57:01.0000] <AryehGregor>
"Monomial Maps on P^2 and their Arithmetic Dynamics"

[08:58:00.0000] <jgraham>
AryehGregor: Oh. Wow. OK. My experience is admittedly only directly limited to a single astrophysics department which didn't have computers that would even run Word

[08:58:01.0000] <AryehGregor>
/me wonders how much of the resulting paper he'll actually understand

[08:58:02.0000] <jgraham>
Although I don't think it was atypical of the physical sciences as a whole

[08:58:03.0000] <annevk>
DOM Range already has a component

[08:58:04.0000] <AryehGregor>
I'm pretty sure LaTeX is the only format that journals will universally accept, though.  It's definitely the standard one.

[08:58:05.0000] <jgraham>
AryehGregor: Congratulations

[08:58:06.0000] <gsnedders>
AryehGregor: The ECMAScript spec is a Word doc, though published as a PDF

[09:00:00.0000] <jgraham>
Also, I think that arXiv is LaTeX-only

[09:00:01.0000] <hsivonen>
AryehGregor: was your statement about the universal acceptance of LaTeX meant to be qualified my a field like math or physics?

[09:00:02.0000] <AryehGregor>
hsivonen, I was talking about math.

[09:01:00.0000] <AryehGregor>
Physics and computer science seem to be similar, though, in my somewhat limited and mostly indirect experience with them.

[09:01:01.0000] <annevk>
MikeSmith, while you are in Bugzilla "Access Control" can be renamed to "CORS"

[09:01:02.0000] <jgraham>
(arXiv is how physics, CS, and so on route around the brokenness of journals)

[09:01:03.0000] <AryehGregor>
And math.

[09:02:00.0000] <AryehGregor>
Did Perelman ever publish his proof of the geometrization hypothesis in an actual journal?

[09:02:01.0000] <hsivonen>
/me was unaware of CS seriously routing around the brokenness of journals

[09:02:02.0000] <hsivonen>
jgraham: do universities these day routinely give publication credit for arXiv when applying for funding or a raise?

[09:03:00.0000] <AryehGregor>
(apparently it's the geometrization conjecture, not hypothesis, whatever)

[09:03:01.0000] <jgraham>
Well I am not so sure about CS. They are somewhat weird because most publications are "conference proceedings"

[09:03:02.0000] <Ms2ger>
AryehGregor, I'm not sure you could generalize much from what Perelman does

[09:04:00.0000] <MikeSmith>
annevk: OK

[09:04:01.0000] <jgraham>
hsivonen: Since I never did either I wouldn't know. But people would put work that was on arXiv but not yet accepted by a journal on their CV

[09:04:02.0000] <MikeSmith>
Ms2ger, AryehGregor - will get those added and ping you once they're set up

[09:04:03.0000] <Ms2ger>
Ta

[09:04:04.0000] <jgraham>
Of course, they would still *submit* to a journal

[09:04:05.0000] <AryehGregor>
Ms2ger, you mean most world-famous mathematicians don't live on bread-and-water diets in their parents' house in Russia?

[09:05:00.0000] <AryehGregor>
MikeSmith, thanks.

[09:05:01.0000] <Ms2ger>
AFAIK, Wiles doesn't

[09:05:02.0000] <jgraham>
But the reality is that arXiv is how you get other people in the field to read what you have done

[09:05:03.0000] <Ms2ger>
Are there any other world-famous mathematicians?

[09:05:04.0000] <hsivonen>
jgraham: how can one submit to a journal if you've already published in arXiv?

[09:05:05.0000] <AryehGregor>
If by "world-famous" you mean "famous among mathematicians worldwide", then yes, lots.

[09:06:00.0000] <AryehGregor>
Even among non-mathematicians, some people have heard of, say, Terence Tao.

[09:06:01.0000] <AryehGregor>
Probably some others too.

[09:06:02.0000] <jgraham>
hsivonen: There is generally some handwaving about it being OK to distribute preprints

[09:06:03.0000] <Ms2ger>
/me googles

[09:06:04.0000] <AryehGregor>
Among mathematicians, sheesh, anyone who's written a moderately popular textbook.

[09:06:05.0000] <jgraham>
Even though you have to do all the copyright assignment rubbish

[09:07:00.0000] <jgraham>
(handwaving in the agreement you sign I mean)

[09:10:00.0000] <hsivonen>
jgraham: what about in terms of novelty? I've heard worries about even having data on a slide deck on the Web if the data hasn't been published in a journal yet

[09:11:00.0000] <AryehGregor>
MikeSmith, my post yesterday to the thread here "RFC from implementers on Element.innerText", from ayg⊙an, still hasn't appeared in the archives: http://lists.w3.org/Archives/Public/public-html/2011Aug/thread.html

[09:13:00.0000] <AryehGregor>
MikeSmith, I'm currently receiving public-html list mail to ayg⊙an, though.

[09:13:01.0000] <zewt>
the archives seem to randomly totally lose mails

[09:13:02.0000] <zewt>
though i don't recall if it's the w3 archives or whatwg ones that i've seen that on

[09:13:03.0000] <zewt>
probably the latter

[09:14:00.0000] <jgraham>
hsivonen: I have never heard anything like that

[09:14:01.0000] <AryehGregor>
jgraham, really?  I have.  Lots of academics are scared of others stealing their ideas.

[09:15:00.0000] <AryehGregor>
Historically there were lots of mathematicians who wouldn't publish things until they had worked out all the interesting implications themselves, so that they'd get all the credit.

[09:15:01.0000] <jgraham>
Well I have heard paranoia about being scooped of course

[09:15:02.0000] <AryehGregor>
Like Newton, Gauss, etc.

[09:15:03.0000] <jgraham>
Mathematics is a bit different from science here I think

[09:15:04.0000] <AryehGregor>
But if you put it on arXiv, of course, and someone else publishes first and doesn't credit you, you can complain to the journal and prove that you made it public first.

[09:16:00.0000] <AryehGregor>
I mean, you have to give attribution to stuff you find on arXiv too, if you used it.

[09:16:01.0000] <jgraham>
Indeed

[09:17:00.0000] <AryehGregor>
I'm guessing this isn't as much of an issue for experimental stuff, because most of the work is often in the physical work rather than the ideas.

[09:17:01.0000] <AryehGregor>
I'd expect theoretical physics to not be so different from math here, though.

[09:18:00.0000] <annevk>
AryehGregor, did you read the W3C email about posting to the archives and go to the website and hit accept?

[09:18:01.0000] <annevk>
AryehGregor, you should have gotten one

[09:19:00.0000] <AryehGregor>
Yes, I did.

[09:19:01.0000] <MikeSmith>
AryehGregor: I just checked and you message was in the approval queue

[09:19:02.0000] <MikeSmith>
not sure why

[09:20:00.0000] <MikeSmith>
but I made it go forth

[09:20:01.0000] <AryehGregor>
MikeSmith, thanks a lot.

[09:20:02.0000] <MikeSmith>
if it happens again, just let me know and I'll check

[09:20:03.0000] <MikeSmith>
there are many things about this mail system that still confuse me

[09:21:00.0000] <smaug____>
huh, strange things http://blog.chromium.org/2011/08/new-websocket-protocol-secure-and.html

[09:21:01.0000] <smaug____>
"no further breaking changes are expected" o_O

[09:23:00.0000] <MikeSmith>
/me steps out to get some grub

[09:23:01.0000] <annevk>
I like how they point to the W3C edition of the spec and the WHATWG mailing list

[09:25:00.0000] <jgraham>
I like how version -10 of the spec sends a Sec-WebSocket-Version header of 8

[09:29:00.0000] <annevk>
http://www.w3.org/TR/WD-xml-link-970406 did not know about this

[09:29:01.0000] <annevk>
precursor to XLink!

[09:39:00.0000] <AryehGregor>
/me sets up Gmail to send mail via Gmail's SMTP servers instead of Gmail's SMTP servers, so that it appears to come from aryeh.name instead of gmail.com

[09:40:00.0000] <AryehGregor>
Which is not only pointless, it means that Gmail is now storing my password in plaintext.  Hurrah.

[09:46:00.0000] <AryehGregor>
But no more distracting Sender: headers!

[09:46:01.0000] <Philip`>
/me doesn't think he's ever read a CS paper via arXiv

[09:47:00.0000] <dglazkov>
good morning, Whatwg

[09:47:01.0000] <Philip`>
Most people seem to provide their own articles as PDFs on their own site, so you can access most stuff for free

[09:48:00.0000] <jgraham>
/me is worried that dglazkov has delusions of being Robin Williams

[09:48:01.0000] <jgraham>
Philip`: But how do you find new stuff?

[09:49:00.0000] <dglazkov>
jgraham: fine, I'll stop good morningning whatwg

[09:50:00.0000] <Philip`>
jgraham: Look at the sites of relevant conferences, which list all their papers

[09:50:01.0000] <Philip`>
then they either link to a PDF or you Google for the title to find one

[09:53:00.0000] <Philip`>
Journal publication sounds extremely slow (though I have no experience with that), but conferences are usually like six months from first submission until the actual conference, so there's not much incentive to work around the delay by posting not-yet-published papers on other sites

[10:23:00.0000] <manu-db>
hsivonen: :-( looks like manu-db filed the absence of some RDFa misfeatures as bugs against Microdata on Saturday

[10:24:00.0000] <manu-db>
Call them what you will - they were requirements for structured data initiatives

[10:24:01.0000] <manu-db>
and seeing as how people were telling me that somebody should file them, but nobody way, I just took it upon myself to file the bugs so they're at least on record.

[10:25:00.0000] <manu-db>
I didn't expect that they'd result in changes to the spec

[10:26:00.0000] <manu-db>
and that seems to be true since almost all of them are being closed as WONTFIX without waiting for feedback

[10:26:01.0000] <AryehGregor>
Waiting for feedback from whom?

[10:26:02.0000] <AryehGregor>
You can reopen resolved bugs if you have new information you'd like to present in response to the resolver's comments.

[10:27:00.0000] <AryehGregor>
It says that somewhere in the resolution boilerplate.

[10:27:01.0000] <manu-db>
Yes, but that's needlessly combative...

[10:27:02.0000] <zewt>
well, it's very very bad bug-reporter-relations to close a bug without giving a chance to respond

[10:27:03.0000] <manu-db>
exactly

[10:27:04.0000] <zewt>
among the worst things you can do, really

[10:27:05.0000] <manu-db>
that's my point, exactly - thank you zewt

[10:28:00.0000] <scor>
I'd call it rude

[10:28:01.0000] <manu-db>
At least w/ the bugs that hsivonen and Hixie filed against RDFa - we had weeks of debate and discussion - recorded in the public logs - before deciding to close the bugs

[10:28:02.0000] <AryehGregor>
I guess it depends on your assumptions about how the bug tracker is to be used.

[10:28:03.0000] <annevk>
dglazkov, please don't!

[10:28:04.0000] <jgraham>
bugs are a stupid place for this kind of conversation

[10:29:00.0000] <zewt>
doesn't matter. if you close bugs like that, you're *going* to piss off reporters, and that can be avoided

[10:29:01.0000] <AryehGregor>
In HTML's case, open bugs automatically wind up on Hixie's todo list, and remain there until resolved.

[10:29:02.0000] <dglazkov>
annevk: ok.

[10:29:03.0000] <dglazkov>
:)

[10:29:04.0000] <AryehGregor>
So generally "open" = "awaiting action from Hixie".

[10:30:00.0000] <AryehGregor>
RESOLVED NEEDSINFO is supposed to mean "we can't do anything about this until the reporter answers our questions", but maybe it doesn't convey that very well.

[10:30:01.0000] <jgraham>
zewt: That wan't really my point

[10:30:02.0000] <AryehGregor>
I generally use NEEDSINFO if I think there's some reasonable chance that the reporter will be able to clarify, and WONTFIX if I'm confident that we don't want to fix the bug regardless.

[10:30:03.0000] <Philip`>
AryehGregor: RESOLVED doesn't really convey the impression that the issue is currently unresolved and awaiting further discussion

[10:31:00.0000] <zewt>
"resolved" says "we're done here" to reporters

[10:31:01.0000] <AryehGregor>
No, it doesn't, but I don't configure the W3C bug tracker.

[10:31:02.0000] <AryehGregor>
Or the HTMLWG policies that say how to use it.

[10:31:03.0000] <manu-db>
yes, the issue is "RESOLVED", AryehGregor

[10:32:00.0000] <AryehGregor>
I've sometimes avoided RESOLVED NEEDSINFO and just left the bug open when I'm confident the commenter will actually respond with the info, but I don't do that consistently.

[10:32:01.0000] <Philip`>
If people can agree on changes that would significantly improve matters, I assume it wouldn't be too hard to change the tracker and/or policies to cope with that

[10:32:02.0000] <manu-db>
I'd rather have a bug listed as OPEN NEEDSINFO - or BLOCKED NEEDSINFO

[10:32:03.0000] <AryehGregor>
Anyway, please do feel free to reopen any of the bugs if you feel you have further information to provide, regardless of what RESOLVED sounds like it means.

[10:32:04.0000] <Philip`>
so it's probably a useful discussion

[10:33:00.0000] <AryehGregor>
I do agree with your point about the demoralizing effect of the terminology here, but you'd have to complain to MikeSmith or someone about that, I can't add new options for what I can do to bugs.

[10:33:01.0000] <AryehGregor>
Anyway, I'm not going to be doing much bug-resolving anymore, it's taking away too much time from spec editing.

[10:35:00.0000] <manu-db>
It's not necessarily demoralizing for me, AryehGregor - but it really did sound like you were just closing the bug and not listening (I forgot which one precisely), but if you hadn't told me your intent - my base assumption is that you're not interested at all in addressing the concern.

[10:36:00.0000] <manu-db>
that's what RESOLVED means to me...

[10:36:01.0000] <manu-db>
I'll go back and review when I have some time and attempt to explain in more detail.

[10:36:02.0000] <AryehGregor>
Generally when I close NEEDSINFO I say something like "I don't understand, please explain X".  If I close WONTFIX, then yes, I'm basically saying I think I understand what you're saying and I disagree and I don't think you're going to get me to agree.

[10:36:03.0000] <AryehGregor>
(but you're still welcome to try)

[10:37:00.0000] <wilhelm>
Opera's long-dead, horribly forked Bugzilla had a CUSTWAIT status (as in, more information from the customer is needed if we are to proceed with this bug). That prevented such bugs from cluttering developers' todo-lists.

[10:37:01.0000] <AryehGregor>
(or rather you're welcome to try getting Hixie to agree, since I probably won't be resolving bugs much anymore)

[10:37:02.0000] <manu-db>
I have little expectation that I'll convince any of the Microdata proponents about most of those features - but I just wanted to ensure that these issues were documented somewhere.

[10:38:00.0000] <zewt>
a lot of trackers have a "needs info, and the bug will switch to closed after a while otherwise" state

[10:38:01.0000] <manu-db>
at the very least, it protects Microdata from CR bugs... you can show people that there was a discussion about these features and you chose to not support them.

[10:38:02.0000] <manu-db>
I don't think we have that for these features vs. Microdata (except, perhaps, buried across the WHATWG list)

[10:39:00.0000] <wilhelm>
/me wonders what the primary usecase for http://dvcs.w3.org/hg/rdf/raw-file/5187c2033a91/rdf-turtle/built-xhtml/FPWD.html is.

[10:40:00.0000] <manu-db>
I don't think Hixie will agree with many of these bugs - he has a specific idea of what Microdata is for and doesn't seem that interested in supporting some of these other use cases that RDFa supports... which is fine - it's creating uber confusion w/ implementers, but there is only so much that can be done about that.

[10:40:01.0000] <zewt>
wilhelm: dunno, but its title sure makes me think of Logo

[10:41:00.0000] <manu-db>
TURTLE is helpful when expressing RDF in a fairly compact form

[10:41:01.0000] <scor>
wilhelm: you should ask in #swig

[10:42:00.0000] <timeless>
/me finishes ATAG20

[10:42:01.0000] <timeless>
(yuck)

[10:42:02.0000] <manu-db>
we haven't found a use for it at our company - except when explaining the type of output a particular RDF processor should generate

[10:42:03.0000] <manu-db>
We tend to just use JSON-LD: http://json-ld.org/

[10:42:04.0000] <manu-db>
and work in JSON

[10:43:00.0000] <manu-db>
which removes the need to work w/ TURTLE/SPARQL/etc. - it works for us, but other people find TURTLE as a good 'glue' RDF syntax.

[10:45:00.0000] <manu-db>
wilhelm: Go here http://ow.ly/5ThZK and click on the "TURTLE" tab - it's easier to read than the JSON

[10:45:01.0000] <AryehGregor>
MikeSmith, there's no way to change the e-mail address of a Bugzilla account, is there?

[10:46:00.0000] <wilhelm>
manu-db: Interesting. Thanks.

[10:58:00.0000] <AryehGregor>
TabAtkins, problem solved!  http://aryeh.name/tmp/editing/editing.html#the-forecolor-command

[10:58:01.0000] <AryehGregor>
/me is proud of himself

[11:00:00.0000] <annevk>
how can currentColor be true at that point?

[11:00:01.0000] <annevk>
and how do you define valid CSS color?

[11:01:00.0000] <annevk>
or was this about the "comment" boxes on the side?

[11:01:01.0000] <annevk>
I should review this draft at some point...

[11:02:00.0000] <AryehGregor>
And you know what the best part is?  It renders identically in all browsers, and I only tested initially in Chrome.

[11:02:01.0000] <AryehGregor>
<3 <3 standards

[11:02:02.0000] <AryehGregor>
annevk, it was about the first "Comments" box there, namely how to do the thing where it overflows to the left.

[11:02:03.0000] <Philip`>
/me gets confused by the "2." adjacent to the second "Comments" button, until realising that it comes from the <li> on the other side of the screen

[11:02:04.0000] <Philip`>
(in Opera 11.50)

[11:02:05.0000] <AryehGregor>
Yeah, I noticed that too.

[11:02:06.0000] <AryehGregor>
Looks like an Opera bug.

[11:03:00.0000] <AryehGregor>
It's fine in Chrome and Firefox.

[11:03:01.0000] <AryehGregor>
IE has a different bug there, it seems.

[11:03:02.0000] <Philip`>
So by "renders identically in all browsers" you mean except for Opera and IE?

[11:03:03.0000] <AryehGregor>
annevk, value can be currentColor if you do document.execCommand("forecolor", false, "currentColor").  As for defining valid CSS color, I don't (yet).

[11:03:04.0000] <AryehGregor>
Philip`, I was referring the the comment box, with the overflowing-to-left effect.

[11:04:00.0000] <AryehGregor>
Including the one-pixel-high absolutely positioned divs that make it looks like the border stretches out to accommodate the extra width of the table.

[11:05:00.0000] <AryehGregor>
Looks pixel-perfect between browsers.

[11:05:01.0000] <TabAtkins>
AryehGregor: Oh man, covering up the border with extra elements.  Clever, but nasty.

[11:05:02.0000] <AryehGregor>
Notice the extra <div style=clear:right></div> I had to add, too.

[11:05:03.0000] <AryehGregor>
Hey, it works.

[11:06:00.0000] <AryehGregor>
(without the middle div, the last div is still positioned above the float)

[11:06:01.0000] <Ms2ger>
That looks like something Wikipedians would do

[11:06:02.0000] <AryehGregor>
I was never part of that insane-inline-style-hacks Wikipedia crowd, in case you were wonderingg.

[11:06:03.0000] <AryehGregor>
wondering.

[11:06:04.0000] <AryehGregor>
I don't think there's any way to do it without adding extra elements.  ::before and ::after aren't enough, are they?

[11:07:00.0000] <Ms2ger>
I was more of a "use classes!" kind of person back then

[11:09:00.0000] <TabAtkins>
AryehGregor: They... might be.

[11:20:00.0000] <Hixie>
wow

[11:20:01.0000] <Hixie>
the w3c are actually going to fork the whatwg spec of their own accord

[11:21:00.0000] <Hixie>
so much for the "we don't want specs to fork" bullshit

[11:21:01.0000] <Hixie>
(webrtc wg)

[11:22:00.0000] <AryehGregor>
I thought we all agreed long ago that they only didn't want non-W3C organizations to fork W3C specs, and are fine with the W3C forking its own specs or anyone else's.

[11:22:01.0000] <AryehGregor>
At least they're consistent, give them that.

[11:22:02.0000] <AryehGregor>
Is anyone going to pay attention to the fork?

[11:23:00.0000] <Hixie>
AryehGregor: they'll certainly pay attention once i start citing it as precedent for forking while ignoring copyrights being fine

[11:24:00.0000] <Hixie>
(good news, everyone! w3c is saying by their actions that we can fork their specs and ignore the copyright!)

[11:25:00.0000] <AryehGregor>
If they're ignoring the license, it's only insofar as they didn't think to include some boilerplate somewhere, so it's not exactly comparable to ignoring a license that forbids redistribution at all.

[11:25:01.0000] <erlehmann>
i thought specs were permissively licensed?

[11:25:02.0000] <erlehmann>
or is it more like “verbatim copying allowed” ?

[11:25:03.0000] <erlehmann>
(i have no idea, sorry)

[11:26:00.0000] <Hixie>
erlehmann: w3c specs are (c) w3c "all rights reserved"

[11:26:01.0000] <erlehmann>
o.0

[11:26:02.0000] <Hixie>
(whatwg specs are under a licence that allows reuse, though doesn't allow copyright reassignment)

[11:26:03.0000] <zewt>
(c) "we own the web"

[11:27:00.0000] <Hixie>
AryehGregor: i highly doubt they are intending to include any boilerplate

[11:27:01.0000] <Hixie>
AryehGregor: since pubrules requires that the spec says (c) w3c

[11:27:02.0000] <zewt>
even if it's a felony?

[11:27:03.0000] <erlehmann>
so they have replaced “© Copyright 2004-2011 Apple Computer, Inc., Mozilla Foundation, and Opera Software ASA.” ?

[11:27:04.0000] <AryehGregor>
You could make them change it if you complained, you know.

[11:27:05.0000] <Hixie>
they haven't done anything yet as far as i know

[11:27:06.0000] <erlehmann>
oh well

[11:27:07.0000] <Hixie>
AryehGregor: i intend to

[11:27:08.0000] <AryehGregor>
Have fun.

[11:27:09.0000] <Hixie>
once there's something to complain about :-)

[11:28:00.0000] <Hixie>
for now i'm just telling the person who is doing it that doing so violates w3c's policies

[11:28:01.0000] <manu-db>
who is doing that?

[11:28:02.0000] <timeless>
AryehGregor: generally bugzilla lets you change your email address...

[11:28:03.0000] <AryehGregor>
timeless, how?  I didn't see an option.

[11:29:00.0000] <Hixie>
manu-db: webrtc wg

[11:29:01.0000] <manu-db>
the entire WG made a resolution to replace the copyright?

[11:29:02.0000] <timeless>
AryehGregor: typically https://bugzilla/userprefs.cgi?tab=account#new_login_name

[11:29:03.0000] <manu-db>
Link to the resolution?

[11:29:04.0000] <manu-db>
or discussion?

[11:30:00.0000] <zewt>
heh, random dripping sarcasm on webapps

[11:30:01.0000] <AryehGregor>
timeless, I don't see it here: http://www.w3.org/Bugs/Public/userprefs.cgi?tab=account

[11:30:02.0000] <AryehGregor>
Just password and real name, not login name.

[11:30:03.0000] <zewt>
infrequent enough compared to some lists that it stands out when it happens

[11:31:00.0000] <Hixie>
manu-db: nothing public currently

[11:31:01.0000] <manu-db>
I'd be surprised if they did what you are saying - seems like a fairly clear-cut copyright issue that could result in a law suit.

[11:32:00.0000] <timeless>
AryehGregor: then it's been disabled/removed

[11:32:01.0000] <timeless>
the feature exists in bugzilla upstream

[11:32:02.0000] <erlehmann>
manu-db, more like a cease and desist.

[11:32:03.0000] <timeless>
contact Team

[11:32:04.0000] <erlehmann>
right?

[11:32:05.0000] <manu-db>
At least, I haven't seen a WG be that stupid... ever.

[11:33:00.0000] <zewt>
iirc willful copyright misrepresentation is a felony in the US--could get worse than a lawsuit

[11:33:01.0000] <timeless>
AryehGregor: given the way permissions work @w3, it would make sense not to allow users to change accounts directly

[11:34:00.0000] <timeless>
zewt: http://www.lib.purdue.edu/uco/CopyrightBasics/penalties.html

[11:35:00.0000] <Hixie>
manu-db: i was surprised myself. but hopefully my e-mail will help them understand the problem and fix it.

[11:35:01.0000] <timeless>
i'm sure i don't want to know

[11:35:02.0000] <manu-db>
erlehmann, yes - that comes first - but one could argue that just by publishing a copy of that, they've caused irreparable harm to Apple, Opera and Mozilla. It's all lawyer BS at that point, but copyright violations like this can have a number of consequences:

[11:35:03.0000] <timeless>
is this rdfa-in-html?

[11:36:00.0000] <timeless>
/me goes looking for webrtc

[11:36:01.0000] <manu-db>
damages/profits awarded, $200 - $150K per violation, attorneys fees/costs, injunction, impounding of the website, jail time for infringer

[11:37:00.0000] <manu-db>
which is why I think that this isn't going to go anywhere - all that WG needs is to get a whiff of the rammifications and they'll back down. I'd chalk it up to "not knowing how copyright works" rather than willful infringement.

[11:37:01.0000] <AryehGregor>
It's obviously not willful, it was some people who weren't paying attention and/or didn't know what they were doing and will be fixed immediately if someone complains to the right people.

[11:38:00.0000] <manu-db>
timeless, no it's not rdfa-in-html :)

[11:38:01.0000] <timeless>
ok, sorry :)

[11:39:00.0000] <TabAtkins>
manu-db: Some of that only comes into play with criminal copyright infringment, which has much stricter requirements (intended to limit it to people, say, printing books and selling them without agreement from the copyright holder).

[11:39:01.0000] <manu-db>
TabAtkins: yes - that's true. However, lawyers usually aren't shy with pushing the law as far as they can.

[11:40:00.0000] <AryehGregor>
When it comes to criminal law, private lawyers have no say in enforcing it.  You'd have to get the DA to go along with it.

[11:40:01.0000] <AryehGregor>
Private people can't file criminal charges.

[11:40:02.0000] <AryehGregor>
So it's irrelevant unless you think the government would care, which they absolutely would not.

[11:40:03.0000] <zewt>
well, in the US, anyway

[11:40:04.0000] <Hixie>
MikeSmith: is feedback to the -comments list still being turned into bugs?

[11:41:00.0000] <AryehGregor>
Plus none of it is relevant because they'd fix it immediately and clearly have been acting in good faith, so the judge is not going to be pleased with the plaintiff if they bring it to court after it caused them no damage and their complaint was promptly and fully addressed.

[11:41:01.0000] <zewt>
apparently the police enforce copyright in Japan (or some parts of; Kyoto?) on their own

[11:41:02.0000] <zewt>
which to me sounds completely insane

[11:41:03.0000] <AryehGregor>
That happens in the US too for criminal copyright infringement.

[11:41:04.0000] <AryehGregor>
But that's only very severe cases.

[11:42:00.0000] <zewt>
but not to that degree--they're acting against eg. p2p users directly

[11:42:01.0000] <zewt>
which is a strictly civil matter here

[11:47:00.0000] <AryehGregor>
Yes, true.

[11:47:01.0000] <timeless>
zewt =~ s/eg./e.g./g

[11:47:02.0000] <timeless>
(sorry, i'm 99

[11:47:03.0000] <timeless>
% done complaining about ATAG20)

[11:48:00.0000] <hsivonen>
MikeSmith: congrats for http://es5.github.com/ being used to find a bug in the HTML5 parser in Gecko

[11:48:01.0000] <timeless>
in the html5 parser? nice

[11:48:02.0000] <timeless>
bug-url?

[11:48:03.0000] <hsivonen>
timeless: https://bugzilla.mozilla.org/show_bug.cgi?id=675499

[11:51:00.0000] <timeless>
spiffy

[11:52:00.0000] <timeless>
/me chuckles

[11:52:01.0000] <timeless>
someone had:

[11:52:02.0000] <timeless>
.hgrc: [alias]

[11:52:03.0000] <timeless>
push = push --something

[11:52:04.0000] <timeless>
diff = diff --something

[11:52:05.0000] <timeless>
that eventually caused stack overflow

[11:53:00.0000] <timeless>
because the aliases were applied iteratively (once for each invocation of *something*)

[11:55:00.0000] <timeless>
(hg help config says that's a bad idea fwiw, so it was technically documented as don't do that..)

[11:57:00.0000] <hober>
Hixie: their intent to fork was made public a few days ago: http://lists.w3.org/Archives/Public/public-webrtc/2011Jul/0166.html

[11:58:00.0000] <erlehmann>
manu-db, i was of the impression the whatwg are nice guys.

[11:59:00.0000] <sicking>
Hixie: is there a reason why MessageEvents sent by WebSockets doesn't have a .origin?

[11:59:01.0000] <timeless>
Hixie: hey

[11:59:02.0000] <timeless>
http://www.whatwg.org/specs/web-apps/current-work/webrtc.html

[11:59:03.0000] <timeless>
s/specification that define the/specification that defines the/

[11:59:04.0000] <timeless>
Hixie: ^ :)

[12:01:00.0000] <Ms2ger>
File a bug ;)

[12:01:01.0000] <annevk>
Hixie, that's chaired by a Google guy no?

[12:02:00.0000] <annevk>
Hixie, wonder if it'll turn out as lovely as the WebSocket protocol

[12:02:01.0000] <Ms2ger>
Probably

[12:27:00.0000] <annevk>
to be fair

[12:27:01.0000] <annevk>
browsers are implementing WebSocket as is

[12:27:02.0000] <annevk>
though I sort of think that's also because it wasn't really worth the fight

[12:31:00.0000] <zewt>
heh

[12:32:00.0000] <zewt>
posts saying "no, big waste of time" are, ironically, the biggest wastes of time

[12:32:01.0000] <annevk>
I meant more fighting for an alternative simpler version

[12:32:02.0000] <Michael>
Is there a rest api anywhere to query the HTML specs?

[12:33:00.0000] <MacTed>
cygri - latest (regular) Safari is Version 5.1 (6534.50) ... so they may well have already fixed the issue you hit earlier

[12:34:00.0000] <annevk>
Michael, nope

[12:34:01.0000] <cygri>
MacTed, that's right, that's what hsivonen told me as well. looks like 5.1 is the first version that actually includes an HTML5 parser.

[12:34:02.0000] <Michael>
darn.

[12:37:00.0000] <hsivonen>
Safari 5.1 usage share is growing more slowly than I had expected

[12:37:01.0000] <hsivonen>
/me expected System Update to replace Safari versions almost overnight

[12:37:02.0000] <hsivonen>
apparently not

[12:38:00.0000] <hsivonen>
partly the fault of Leopard and 5.1 not being available for Leopard

[12:51:00.0000] <sicking>
Hixie: (i think i asked elsewhere, but I can't find the message now, appologies if it's a repeat) is there a reason why websockets doesn't set .origin on the message events

[12:54:00.0000] <hober>
hsivonen: check 5.0.6 for leopard

[12:59:00.0000] <Hixie>
hober: that e-mail doesn't suggest they're going to use the same text, just fork the functionality

[12:59:01.0000] <Hixie>
sicking: what would the origin be?

[13:00:00.0000] <Hixie>
sicking: by definition if you can connect to a websocket, it's treated as same-origin

[13:01:00.0000] <gsnedders>
hsivonen: Yeah, Apple seem to be having a harder time getting people upgrading now.

[13:01:01.0000] <gsnedders>
hsivonen: I wonder if Leopard will still get security updates.

[13:02:00.0000] <timeless>
AryehGregor: hey

[13:02:01.0000] <AryehGregor>
timeless, hey.

[13:02:02.0000] <timeless>
what kind of caching headers are you using on http://aryeh.name/spec/editing/editing.html

[13:02:03.0000] <timeless>
my browser restarted today

[13:03:00.0000] <timeless>
but it showed me the july 26 version of the doc

[13:03:01.0000] <timeless>
a reload got me the july 29 version

[13:03:02.0000] <timeless>
s/the/a/g

[13:03:03.0000] <AryehGregor>
/me sees Cache-Control:must-revalidate, max-age=0

[13:03:04.0000] <AryehGregor>
July 29 is the most recent published one.

[13:03:05.0000] <Hixie>
if i wanted to add a warning about clickjacking attacks to the html spec, anyone have any suggestions on where i should put it?

[13:03:06.0000] <timeless>
i guess that tabcandy doesn't honor that

[13:03:07.0000] <AryehGregor>
The WIP one is http://aryeh.name/tmp/editing/editing.html, but that's changing as I work on it in real time.

[13:03:08.0000] <timeless>
Hixie: near <iframe>s?

[13:04:00.0000] <AryehGregor>
I'm moving zillions of <!-- comments --> to be viewable without looking at the source, which is taking a while.

[13:04:01.0000] <timeless>
/me loads http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#the-iframe-element

[13:06:00.0000] <Hixie>
timeless: why near iframes? that's just one way to do clickjacking, it's not where an author would look for how to avoid getting p0wned, surely

[13:07:00.0000] <Ms2ger>
Rendering :)

[13:07:01.0000] <timeless>
http://www.whatwg.org/specs/web-apps/current-work/multipage/dom.html#security-document ?

[13:07:02.0000] <Hixie>
hmm

[13:07:03.0000] <Hixie>
i'm thinking maybe the intro section

[13:08:00.0000] <Hixie>
maybe under "A quick introduction to HTML"

[13:08:01.0000] <timeless>
/me shrugs

[13:08:02.0000] <timeless>
i'd search for it under a security section

[13:08:03.0000] <AryehGregor>
Do you have some kind of general security advice section?

[13:08:04.0000] <timeless>
http://www.whatwg.org/specs/web-apps/current-work/multipage/browsers.html#security-nav

[13:08:05.0000] <AryehGregor>
Why provide only this security advice and not other security advice too?

[13:08:06.0000] <Hixie>
AryehGregor: this is basically what this will become, i thikn

[13:08:07.0000] <timeless>
AryehGregor: he has 5 or so

[13:09:00.0000] <Hixie>
AryehGregor: i was going to mention a few other things too

[13:09:01.0000] <Hixie>
if i put it in the intro section

[13:09:02.0000] <timeless>
maybe 20

[13:09:03.0000] <AryehGregor>
It makes sense to include advice about XSS, CSRF, etc. before you start on weird stuff like clickjacking.

[13:09:04.0000] <timeless>
yeah, clickjacking should not be the first thing to draft

[13:09:05.0000] <Hixie>
yeah

[13:09:06.0000] <timeless>
start w/ XSS

[13:10:00.0000] <timeless>
since the version of clickjacking that you seem to mean is actually a derivation from XSS

[13:11:00.0000] <timeless>
I guess after quick introduction wouldn't be a bad place

[13:11:01.0000] <timeless>
`Security Concerns`

[13:14:00.0000] <timeless>
> DOM trees contain several kinds of nodes, in particular a DOCTYPE node, elements, text nodes, and comment nodes.

[13:14:01.0000] <timeless>
Hixie: s/elements/element nodes/ ?

[13:14:02.0000] <Hixie>
*shrug*

[13:14:03.0000] <Hixie>
it's right either way

[13:15:00.0000] <timeless>
the change makes the sentence more consistent..

[13:16:00.0000] <Ms2ger>
I'd file a tracker issue

[13:16:01.0000] <timeless>
does that require a bugzilla account?

[13:16:02.0000] <timeless>
/me isn't sure how to get one of those

[13:17:00.0000] <Hixie>
let's focus on actual bugs and not on whether sentences might be slightly more polished please :-P

[13:17:01.0000] <Hixie>
typos are one thing, i don't mind fixing those

[13:17:02.0000] <Ms2ger>
Hixie, the pedants must win!

[13:17:03.0000] <timeless>
your document is too long! and you're making it longer! ignore the fact that i wanted to make it a word longer ;-)

[13:18:00.0000] <Ms2ger>
Like me

[13:18:01.0000] <Hixie>
but when the sentence is accurate, correct, precise, grammatically sound, spelt correctly, scans well, and merely suffers from a slight inconsistency, let's just let it slide

[13:18:02.0000] <AryehGregor>
timeless, I consistently use all node types as standalone nouns, like "Comment" or "Element", except "Text", where I say "Text node".

[13:19:00.0000] <AryehGregor>
Like "If node is a Comment, do X.  If node is a Text node, do Y."

[13:19:01.0000] <AryehGregor>
Or "If node is a Text, Comment, or ProcessingInstruction node, do Z."

[13:19:02.0000] <AryehGregor>
Because "If node is a Text" sounds silly.

[13:19:03.0000] <timeless>
AryehGregor: that would be a reasonable strategy, but from context, it isn't the one Hixie is using :)

[13:19:04.0000] <Hixie>
yeah if a bug does get filed on that particular thing, i'm just removing the last "node", not adding one after "elements"

[13:19:05.0000] <AryehGregor>
Evidently not.

[13:19:06.0000] <timeless>
Hixie: i'm fine w/ that fix :)

[13:20:00.0000] <Hixie>
cross-site scripting is essentially when you don't escape user input in html; sql injection is the same with sql; what do you call the similar thing where you don't validate user input on the server and so allow out-of-bounds input to be stored in the db?

[13:20:01.0000] <AryehGregor>
Hixie, "not a vulnerability by itself

[13:20:02.0000] <AryehGregor>
"

[13:21:00.0000] <timeless>
lol

[13:21:01.0000] <AryehGregor>
MediaWiki, for instance, doesn't escape anything before storing in the DB, as a rule.

[13:21:02.0000] <AryehGregor>
It trusts text content in the DB as untrusted.

[13:21:03.0000] <timeless>
Hixie: failure to enforce boundary constraints

[13:21:04.0000] <AryehGregor>
All escaping is done as close to output time as possible, so it's easy to audit.

[13:21:05.0000] <AryehGregor>
It's much harder to escape at input time, since input can come from a zillion different places that you have varying degrees of control over.

[13:21:06.0000] <timeless>
AryehGregor: what if there isn't `output` in the normal sense?

[13:21:07.0000] <timeless>
such as a join?

[13:22:00.0000] <AryehGregor>
Then what vulnerability is there?

[13:22:01.0000] <Ms2ger>
/nick Robert'); DROP TABLE Students;--

[13:22:02.0000] <Hixie>
i mean like checking that an "age" field is >=0

[13:22:03.0000] <AryehGregor>
By "output" I mean "use".

[13:22:04.0000] <AryehGregor>
Hixie, store it in the database as an unsigned integer?  :)

[13:22:05.0000] <AryehGregor>
But okay.

[13:22:06.0000] <timeless>
s/0/18/

[13:22:07.0000] <Hixie>
sure there's lots of solutions but what is the vulnerability called?

[13:22:08.0000] <timeless>
or 21 if you agree w/ Nancy Reagan..

[13:22:09.0000] <AryehGregor>
"not validating user input"

[13:23:00.0000] <timeless>
failure to enforce boundary constraints ...

[13:23:01.0000] <Hixie>
yeah that's all i could come up with too

[13:23:02.0000] <Hixie>
i was hoping there was a buzzword like "sql injection" or "xss"

[13:23:03.0000] <timeless>
not everything is cool enough to get a buzzword

[13:25:00.0000] <timeless>
i think `failure to santize <whatever>` is probably the closest to a buzzword, but i'm not sure that matches what you want

[13:25:01.0000] <AryehGregor>
I don't know of a way to phrase it positively (without "failing", "not", etc.)

[13:25:02.0000] <timeless>
you can review cwe to see `failure to santize` appearing a bunch of times

[13:25:03.0000] <timeless>
`improper input validation`

[13:25:04.0000] <timeless>
doesn't technically have a negative :)

[13:25:05.0000] <timeless>
that's from CWE-74

[13:26:00.0000] <AryehGregor>
"Input validation" is definitely the term for the thing you should be doing.

[13:26:01.0000] <AryehGregor>
For the thing you shouldn't be doing, "improper input validation" sounds like as good a term as any.

[13:26:02.0000] <AryehGregor>
It's not an exploit by itself, so it doesn't get a buzzword.

[13:26:03.0000] <AryehGregor>
Well, I guess it can be a minor exploit by itself.

[13:27:00.0000] <AryehGregor>
Like if you have maxlength and users can just remove it in their favorite web development tool and submit stuff that's whatever length they want, that might be classified as a security vulnerability, insofar as it involves users doing something with your site that they aren't supposed to be able to.

[13:27:01.0000] <Philip`>
One occasional problem is storing a user's IP address in a database, when you get it from some function that understands X-Forwarded-For, when not running behind a proxy that sets X-Forwarded-For, so malicious users can set it to whatever they want and trick you into storing the wrong IP address

[13:28:00.0000] <Ms2ger>
So, Bert thinks newline normalization in attributes is the biggest problem when parsing HTML with an SGML parser?

[13:28:01.0000] <Philip`>
which is sort of related to storing wrong information in a database, but separate from constraint validation (which any self-respecting SQL database ought to be enforcing automatically)

[13:29:00.0000] <timeless>
Philip`: in theory you're supposed to validate the source against a list of trusted proxies

[13:30:00.0000] <timeless>
Ms2ger: that means bert doesn't think html as sgml has many problems?

[13:30:01.0000] <timeless>
(like Comments or NetTags ?)

[13:30:02.0000] <Philip`>
In practice people forget to do that

[13:30:03.0000] <Ms2ger>
I wouldn't dare to claim I understand Bert

[13:30:04.0000] <timeless>
heh

[13:31:00.0000] <AryehGregor>
Philip`, Wikipedia was actually hit by a subtle version of that once.

[13:31:01.0000] <timeless>
AryehGregor: nice

[13:31:02.0000] <AryehGregor>
Squid did understand X-Forwarded-For and sanitized it, but the PHP variables used to access it treated X_Forwarded_For the same as X-Forwarded-For, and Squid didn't.

[13:31:03.0000] <timeless>
nice

[13:32:00.0000] <AryehGregor>
MediaWiki has a whitelist of trusted XFF providers that it uses, including various major ISPs in addition to its own internal Squids and Varnishes.

[13:32:01.0000] <timeless>
> Each element in the DOM tree is represented by an object, and these objects have APIs so that they can be manipulated.

[13:32:02.0000] <timeless>
Ms2ger, AryehGregor : what does `manipulated` mean to you?

[13:33:00.0000] <AryehGregor>
It's kind of vague.

[13:33:01.0000] <timeless>
to me, i think `modified`, but perhaps my dictionary is flawed

[13:33:02.0000] <Ms2ger>
Change any kind of internal state, I guess

[13:33:03.0000] <timeless>
/me underlines Ms2ger's use of the word `Change`

[13:33:04.0000] <AryehGregor>
Basically "modified", yeah.

[13:33:05.0000] <AryehGregor>
Very general.

[13:34:00.0000] <Ms2ger>
:)

[13:34:01.0000] <timeless>
but we all selected a word meaning `modify`

[13:34:02.0000] <timeless>
which to me is somewhat exclusive of the concept `to inspect`

[13:34:03.0000] <Ms2ger>
Mm

[13:36:00.0000] <AryehGregor>
I guess.

[13:37:00.0000] <timeless>
i'll take an Mm and an I guess :)

[13:50:00.0000] <timeless>
/me is confused

[13:50:01.0000] <timeless>
annevk: you around?

[13:57:00.0000] <annevk>
not really

[13:59:00.0000] <annevk>
but

[13:59:01.0000] <annevk>
DOM Core defines "element" to mean "Element node"

[14:06:00.0000] <timeless>
annevk: my question to you was different

[14:06:01.0000] <timeless>
(but it seems my irc connect is too flaky for such purposes)

[15:04:00.0000] <AryehGregor>
Can anyone explain to me why in Firefox (7.0a2), list items with the comments button all have an extra line break at the start?  http://aryeh.name/tmp/editing/editing.html#split-the-parent

[15:05:00.0000] <TabAtkins>
I don't see an extra linebreak in current aurora.

[15:05:01.0000] <AryehGregor>
Hmm.

[15:05:02.0000] <AryehGregor>
Oh, Ctrl-F5 fixed it.

[15:05:03.0000] <AryehGregor>
Okay, good.

[15:06:00.0000] <AryehGregor>
/me has no idea why, though

[15:06:01.0000] <AryehGregor>
Opera places the list marker in a quite silly place.

[15:06:02.0000] <TabAtkins>
List marker placement is completely un-interoperable.

[15:07:00.0000] <AryehGregor>
Yeah, but putting it in front of the first child of the list when that child is right-floated is just silly.

[15:07:01.0000] <TabAtkins>
Firefox and Opera generally use one model, with significant differences.  IE and Webkit generally use a second model, with significant differences.

[15:07:02.0000] <gsnedders>
TabAtkins: is this in Opera 12?

[15:07:03.0000] <TabAtkins>
I'm speccing the IE behavior, because it's simple and sane.

[15:07:04.0000] <TabAtkins>
gsnedders: It's in "whatever I have on my windows box, which keeps itself fairly updated".

[15:08:00.0000] <AryehGregor>
Hmm, now IE9 is behaving like Firefox did before I did Ctrl-F5.

[15:08:01.0000] <gsnedders>
TabAtkins: Ah, no, that's not out in a public build yet. Presto 2.9.192 finally fixes the <ul><li></ul> non-marker issue.

[15:08:02.0000] <gsnedders>
TabAtkins: And fixing that caused a lot of change of code that produces markers.

[15:09:00.0000] <gsnedders>
TabAtkins: Also, you defining what CSS properties affect list-markers?

[15:09:01.0000] <gsnedders>
(like font styling)

[15:10:00.0000] <TabAtkins>
gsnedders: All of them.  ::marker is identical to ::before, basically.

[15:10:01.0000] <AryehGregor>
Ah.

[15:14:00.0000] <AryehGregor>
/me now is very confused

[15:14:01.0000] <TabAtkins>
Hm?

[15:14:02.0000] <AryehGregor>
IE9 is still creating the extra line breaks.

[15:14:03.0000] <AryehGregor>
I haven't found a minimal test case yet that distinguishes.

[15:15:00.0000] <AryehGregor>
/me concludes he's too tired to do that right now anyway

[15:15:01.0000] <timeless>
AryehGregor: network gremlins?

[15:15:02.0000] <AryehGregor>
I should really say "the extra vertical whitespace".

[15:15:03.0000] <AryehGregor>
Doesn't seem network-related.

[15:15:04.0000] <TabAtkins>
Ooh, I suspect it's related to the float interfering with phantom line boxes.

[15:16:00.0000] <TabAtkins>
The IE9 model, explained in terms of the Lists spec, which it doesn't actually implement but might as well...

[15:16:01.0000] <TabAtkins>
Means that ::marker is an abspos element using a special positioning scheme.  Abspos elements leave behind a "placeholder" box in their normal position, which is a 0x0 inline element.

[15:17:00.0000] <TabAtkins>
Sometimes, if a linebox would only be generated due to a placeholder, CSS suppresses the linebox.

[15:17:01.0000] <TabAtkins>
It's possible that the float is interfering with this.

[15:17:02.0000] <timeless>
sounds yummy

[15:17:03.0000] <TabAtkins>
s/inline element/inline box/

[15:17:04.0000] <timeless>
/me goes movie hunting

[15:17:05.0000] <timeless>
(seems like a better idea)

[15:18:00.0000] <timeless>
good luck storming the castke

[15:18:01.0000] <timeless>
s/k/l/

[15:21:00.0000] <TabAtkins>
Hmm, but I also can't reproduce anything with that.

[15:24:00.0000] <zewt>
sleep well, and dream of large women


2011-08-03
[17:19:00.0000] <Hixie>
holy cow

[17:19:01.0000] <Hixie>
<option>'s value really is stripped of leading and trailing spaces before submission

[17:19:02.0000] <Hixie>
and newlines replaced by spaces!

[17:20:00.0000] <Hixie>
and multiple spaces collapsed!

[17:21:00.0000] <Hixie>
hm, chrome only does the stripping

[20:07:00.0000] <jcranmer>
Hixie: sigh

[20:07:01.0000] <jcranmer>
why can't people write specifications correctly?

[20:07:02.0000] <jcranmer>
http://www.yenc.org/yenc-draft.1.3.txt

[20:09:00.0000] <jcranmer>
as annoying as HTML5 is to read, that spec is worse

[20:09:01.0000] <jcranmer>
leaving alone the fact that it is a horrible implentation (let's take hithertofore unstructured fields and require them to be structured!)

[22:09:00.0000] <MikeSmith>
Hixie: about feedback to comments list, I have to manually pipe the messages to a script

[22:09:01.0000] <MikeSmith>
and I am way behind on doing that

[22:09:02.0000] <MikeSmith>
I will do them today

[22:11:00.0000] <MikeSmith>
AryehGregor: I think you can change the bugzilla account to a different address and it will change globally/retroactively

[22:11:01.0000] <MikeSmith>
I mean, it will also change the address linked to in any comments you submitted previously, any bugs you reported previously, etc.

[22:12:00.0000] <MikeSmith>
if the UI doesn't let you change it yourself, I am pretty sure I can from the admin UI

[22:40:00.0000] <karlcow>
http://www.reuters.com/article/2011/08/03/us-cyberattacks-idUSTRE7720HU20110803?feedType=RSS&feedName=topNews&dlvrit=59363

[23:06:00.0000] <MikeSmith>
AryehGregor - HTML Editing APIs added in bugzilla under WebAppsWG product

[23:19:00.0000] <annevk>
MikeSmith, still named "Access Control"

[23:19:01.0000] <MikeSmith>
oh

[23:19:02.0000] <MikeSmith>
will fix that now

[23:19:03.0000] <MikeSmith>
how come you awake man?

[23:20:00.0000] <annevk>
little after 7AM here :)

[23:20:01.0000] <MikeSmith>
that's no proper time to be awake, chief

[23:22:00.0000] <MikeSmith>
annevk: http://dev.w3.org/2006/waf/access-control/ is still the current location for the ED, right?

[23:22:01.0000] <MikeSmith>
annevk: name changed now

[23:23:00.0000] <annevk>
yup

[23:23:01.0000] <MikeSmith>
k

[23:23:02.0000] <annevk>
and yeah, ideally I would have slept a little longer

[23:23:03.0000] <MikeSmith>
that URL just keeps getting better with every year

[23:27:00.0000] <MikeSmith>
a lot of reports on sleep research I've seen recently seem to strongly indicate that we really should get 8 hours of sleep every night

[23:27:01.0000] <MikeSmith>
people who don't get 8 hours of sleep perform worse on memory tests and such

[23:28:00.0000] <MikeSmith>
seems like part of what happens with your brain during sleep is a kind of garbage collection

[23:34:00.0000] <zcorpan>
MikeSmith: i sleep at least 8 hours every night, i think

[23:34:01.0000] <MikeSmith>
eh?

[23:34:02.0000] <MikeSmith>
so if you're awake now, what time did you go to sleep last night?

[23:34:03.0000] <zcorpan>
22:00

[23:35:00.0000] <zcorpan>
i was intending to get up at 06:00 but i didn't get up until 06:50

[23:36:00.0000] <MikeSmith>
good god

[23:37:00.0000] <MikeSmith>
I envy you

[23:37:01.0000] <zcorpan>
i guess my night sleep will be fucked when we have kids

[23:38:00.0000] <MikeSmith>
because of timezone differences, 22:00 is time when most of my telcons start

[23:38:01.0000] <MikeSmith>
oh yeah

[23:38:02.0000] <zcorpan>
boo telecons, and *late* telecons to boot :(

[23:39:00.0000] <MikeSmith>
with my daughter, I think for at least the first 4 months, I didn't sleep more than 2 hours straight at time

[23:39:01.0000] <annevk>
i'm open to change that URL

[23:39:02.0000] <annevk>
we can move the whole thing to dvcs

[23:39:03.0000] <MikeSmith>
she was the kind of baby that knew whether I was standing up when I was holding her or not, and at night she'd cry unless I was standing and moving around with her

[23:40:00.0000] <MikeSmith>
annevk: I think it's fine where it is for now, but I'm also happy to move it to dvcs.w3.org if you want. I can do it right now. You want me to?

[23:44:00.0000] <othermaciej>
Hixie: do you have an opinion about text/html-sadboxed not doing what is intended (apparently) in IE6?

[23:44:01.0000] <annevk>
MikeSmith, if you can get me http://dvcs.w3.org/hg/cors/ please yes

[23:44:02.0000] <MikeSmith>
annevk: OK, will do now

[23:44:03.0000] <annevk>
MikeSmith, dev.w3.org is so slow sometimes; nothing beats local hg

[23:44:04.0000] <zcorpan>
othermaciej: it sniffs it as html?

[23:45:00.0000] <MikeSmith>
annevk: yeah, come to think of it, that alone makes it worth moving stuff

[23:45:01.0000] <othermaciej>
zcorpan: at least one Microsoft dude so claims

[23:45:02.0000] <zcorpan>
is there a mime type we could use instead that ie6 wouldn't sniff as html?

[23:45:03.0000] <othermaciej>
IE6 is still over 9% market share, which is pretty bad

[23:46:00.0000] <othermaciej>
I don't know

[23:46:01.0000] <othermaciej>
I haven't even checked the MS guy's claims

[23:46:02.0000] <annevk>
prolly adding a bunch of zero bytes at the start of the document

[23:46:03.0000] <zcorpan>
application/octet-stream; actually=sandboxed-html

[23:46:04.0000] <annevk>
bah

[23:47:00.0000] <othermaciej>
does IE6 actually refrain from sniffing application/octet-stream?

[23:47:01.0000] <zcorpan>
not sure

[23:51:00.0000] <annevk>
so much email

[23:55:00.0000] <MikeSmith>
there have been a lot fewer LC comments than I would have expected at this point

[23:56:00.0000] <MikeSmith>
I think we got something on the order of 375 LC comments filed as bugs in bugzilla

[23:58:00.0000] <MikeSmith>
anyway, email sucks

[00:01:00.0000] <MikeSmith>
I probably should not admit this, but I literally have 7551 unread messages in my inbox

[00:01:01.0000] <MikeSmith>
mostly list mail, but still

[00:01:02.0000] <othermaciej>
it is surprising how low the LC feedback has been

[00:01:03.0000] <othermaciej>
but perhaps this is an artifact of the massive level of ongoing feedback we have gotten

[00:01:04.0000] <othermaciej>
I would have expected > 1000 LC comments

[00:02:00.0000] <MikeSmith>
yep, I had been thinking more like 2000

[00:03:00.0000] <roc>
how many people actually know about the LC deadline who haven't been providing feedback all  along?

[00:03:01.0000] <MikeSmith>
roc: people who only read the TR drafts, I guess

[00:13:00.0000] <MikeSmith>
anyway, givent the spec has been in development for 6-7 years, and been scrutinized and commented on very heavily that whole time, so I guess it's expected we're not going to get a flood of new insights

[00:16:00.0000] <roc>
one might still expect to get a flood of rubbish comments

[00:32:00.0000] <nessy>
oh, I am quite drowning in today's deluge of LC bugs!

[00:32:01.0000] <annevk>
not the only one

[00:50:00.0000] <Hixie>
othermaciej: IIRC, the issue was known at the time the feature was designed, and either there's a workaround or we figured IE6 was so insecure at this point that there wasn't much point worrying about it

[00:50:01.0000] <Hixie>
but i forget the details

[00:50:02.0000] <Hixie>
nobody has raised the issue of IE6 with me recently as far as I know though

[00:51:00.0000] <Hixie>
on another note, does anyone other than AryehGregor and Julian have an opinion on whether we should drop rel=help or add a:link[rel~=help] to the UA style sheet?

[00:51:01.0000] <annevk>
Hixie, I looked on public-webrtc and other than the usual debate didn't see the forking thing you mentioned

[00:51:02.0000] <annevk>
Hixie, was that offlist?

[00:51:03.0000] <Hixie>
(with cursor:hlep, that is)

[00:52:00.0000] <Hixie>
annevk: it was a private e-mail

[00:52:01.0000] <annevk>
cursor:help seems kind of a neat feature

[00:52:02.0000] <annevk>
almost easter egg, but still

[00:53:00.0000] <annevk>
the kind of attention to detail some people will appreciate, like me :)

[00:53:01.0000] <zcorpan>
i recall in windows, stuff that have cursor:help are *buttons*, not links

[00:53:02.0000] <Hixie>
on the web, help is usually found in link form

[00:54:00.0000] <Hixie>
eh, i guess we can try adding this style rule and see how it's taken

[00:54:01.0000] <annevk>
that's a good point, not sure whether that would actually make good UI

[00:54:02.0000] <Hixie>
and if nobody cares, we can drop it

[00:54:03.0000] <annevk>
sorry

[00:54:04.0000] <zcorpan>
we need more easter eggs in html

[00:55:00.0000] <Hixie>
not sure everyone agrees with that

[00:55:01.0000] <Hixie>
but i do!

[00:55:02.0000] <Hixie>
:-)

[00:55:03.0000] <zcorpan>
i'll file a LC comment

[00:55:04.0000] <Hixie>
on more easter eggs?

[00:55:05.0000] <zcorpan>
yeah

[00:55:06.0000] <Hixie>
ok but unless you can name at least 20 existing ones in private e-mail, i'm not adding any for the bug :-P

[00:56:00.0000] <annevk>
Hixie, btw, the writing HTML section does not really mention <svg> and <math> as the elements that trigger special subtree behavior and that <foreignObject> etc. escape out of it again

[00:56:01.0000] <Hixie>
annevk: is that feedback on the normative aspects or on the expository aspects?

[00:56:02.0000] <zcorpan>
Hixie: don't have time to hunt for existing ones, sorry!

[00:57:00.0000] <zcorpan>
maybe instead i should try to come up with a good easter egg to add before filing a bug

[00:58:00.0000] <Hixie>
if it's an easter egg i recommend mailing me the suggestion privately

[00:58:01.0000] <Hixie>
so i can try to slip it in a huge update

[00:58:02.0000] <zcorpan>
k

[01:00:00.0000] <annevk>
Hixie, I guess it is complete, but it's kind of hard

[01:00:01.0000] <Hixie>
we can add a non-normative section explaining it better

[01:00:02.0000] <Hixie>
file a bug saying that?

[01:00:03.0000] <annevk>
http://twitter.com/karlpro/status/98606552990236672 nice

[01:01:00.0000] <annevk>
Hixie, Matt May filed something to that extent, I'll add a comment there

[01:01:01.0000] <Hixie>
cool, thanks

[01:04:00.0000] <annevk>
oh

[01:04:01.0000] <annevk>
I do realize now it does not define e.g. <svg><base></svg> properly

[01:04:02.0000] <annevk>
because in the "SVG model" section it says you can use HTML elements anywhere

[01:04:03.0000] <Hixie>
that's non-conforming, no?

[01:05:00.0000] <Hixie>
i'm not familiar with that section of the spec these days

[01:05:01.0000] <Hixie>
"svg model" doesn't seem to appear in the html spec

[01:06:00.0000] <annevk>
it's called "SVG"

[01:06:01.0000] <annevk>
but it doesn't say that, it says you can use non-HTML elements anywhere

[01:07:00.0000] <Hixie>
doesn't it just defer to the SVG spec?

[01:07:01.0000] <annevk>
not completely, it defines that foreignObject can contain "flow content"

[01:08:00.0000] <Hixie>
it defines that if foreignObject is used to escaped to HTML, it has to contain flow content, yeah

[01:36:00.0000] <annevk>
when adding abort() it seems you did not add a domintro entry

[01:36:01.0000] <annevk>
http://html5.org/tools/web-apps-tracker?from=6348&to=6349

[01:36:02.0000] <annevk>
idea for web-apps-tracker: file bug regarding this change

[01:43:00.0000] <zcorpan>
annevk: yeah that'd be nice

[01:44:00.0000] <zcorpan>
annevk: might be hard to set the right component though

[02:09:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13544

[02:09:01.0000] <annevk>
nessy, so Microsoft I guess, who else?

[02:09:02.0000] <annevk>
they really want to implement TTML?

[02:09:03.0000] <annevk>
is that the IE Team or some other Microsoft division?

[02:10:00.0000] <nessy>
there are guys at Google that want multiple format support in WebKit, too

[02:10:01.0000] <nessy>
I don't think anyone will want to support the full TTML spec

[02:10:02.0000] <annevk>
really, what other formats does Google want?

[02:10:03.0000] <nessy>
but many want to parse the basic format

[02:11:00.0000] <nessy>
I've listed a few :-)

[02:11:01.0000] <annevk>
you listed TTML and WebVTT

[02:11:02.0000] <annevk>
the basic format? sounds like interop hell

[02:11:03.0000] <nessy>
in the example there's also srt and mpsub

[02:11:04.0000] <annevk>
oh

[02:11:05.0000] <nessy>
TTML is already interop hell

[02:11:06.0000] <annevk>
that's going to be even more fun

[02:11:07.0000] <nessy>
I don't really care - I regard anything that goes beyond WebVTT as "specialised" implementations

[02:11:08.0000] <annevk>
if SRT is going to be implemented after all we should just merge WebVTT with it again

[02:12:00.0000] <nessy>
no, that doesn't make sense

[02:12:01.0000] <annevk>
it does, because if SRT is going to be supported by everyone WebVTT makes no chance

[02:12:02.0000] <annevk>
the whole point of doing WebVTT is that we are not doing anything else

[02:12:03.0000] <nessy>
there are too many srt files out there that have specialised extensions that don't fit with webvtt

[02:12:04.0000] <annevk>
if we do something else it's kind of a lost cause

[02:13:00.0000] <nessy>
don't worry about srt

[02:13:01.0000] <nessy>
the key point is that all browsers will implement webvtt, so it becomes the baseline format

[02:13:02.0000] <annevk>
no

[02:13:03.0000] <annevk>
the key point is what the dominant browsers implement

[02:13:04.0000] <nessy>
that' webvtt

[02:14:00.0000] <nessy>
it's why we are creating a webvtt WG, so MS can feel comfortable with it

[02:14:01.0000] <nessy>
well: actually contribute to it

[02:15:00.0000] <zcorpan>
there's no schema.org WG

[02:15:01.0000] <nessy>
I can only say what I was told by MS

[02:16:00.0000] <nessy>
I don't care what they do with other specs

[02:16:01.0000] <nessy>
is schema.org developed by WHATWG?

[02:17:00.0000] <zcorpan>
no

[02:17:01.0000] <nessy>
there's your answer

[02:17:02.0000] <nessy>
unfortunately

[02:18:00.0000] <zcorpan>
what? MS hates WHATWG?

[02:18:01.0000] <zcorpan>
i thought they said it was because WHATWG doesn't have a patent policy, but schema.org doesn't have one either

[02:19:00.0000] <annevk>
nessy, if they also implement SRT I am highly skeptical for WebVTT

[02:19:01.0000] <annevk>
nessy, also, if they implement SRT I want a specification for SRT

[02:20:00.0000] <annevk>
nessy, and don't really see the point in pursuing WebVTT

[02:20:01.0000] <nessy>
no, there's nothing to worry about srt - it's much too simple to support much of the use cases I've come across

[02:20:02.0000] <nessy>
srt and webvtt are worlds apart

[02:20:03.0000] <annevk>
there's plenty to worry

[02:20:04.0000] <annevk>
for one there's no spec

[02:20:05.0000] <nessy>
there's no way you're going to get the webvtt functionality included in srt without breaking many of the srt implementations

[02:21:00.0000] <nessy>
the wikipedia page is as good as it gets for srt

[02:21:01.0000] <nessy>
but I can register a IETF I-D if you prefer

[02:21:02.0000] <annevk>
that's not really acceptable

[02:21:03.0000] <zcorpan>
seems to cover enough use cases to make as widespread as it is

[02:21:04.0000] <annevk>
and if SRT addresses 90% I'm not really sure why we should go to the trouble of implementing WebVTT

[02:22:00.0000] <nessy>
only because people haven't really got a choice - this far, we had the choice between simple and horrendously complex caption formats - webvtt sits in the middle

[02:22:01.0000] <hsivonen>
zcorpan: how do you know that Google, MS and Y! don't have a trilateral patent deal for schema.org?

[02:22:02.0000] <zcorpan>
hsivonen: i don't

[02:22:03.0000] <zcorpan>
maybe they do

[02:26:00.0000] <nessy>
srt does not address 90% of the use cases - it just seems this way because it's all we had available this far on the Web

[02:29:00.0000] <zcorpan>
everyone thought it would be impossible to specify html parsing without breaking compat

[02:29:01.0000] <zcorpan>
until it was done

[02:34:00.0000] <zcorpan>
some day i should take some time to read about ie10. just learned they dropped conditional comments

[02:50:00.0000] <AnselmBradford>
hi whatwg, does anyone know a use-case for using the embed element over the object element, other than to support old versions of FF? Is its inclusion in the spec purely to codify its use in the wild?

[02:54:00.0000] <annevk>
I think with <embed> you do not need the typemustmatch attribute

[02:55:00.0000] <annevk>
and therefore with <embed> you also have the feature of interpreting everything that comes over the wire according to a type under your control

[02:55:01.0000] <annevk>
overall though the features are largely similar

[02:56:00.0000] <annevk>
public-fx is amusing

[02:56:01.0000] <annevk>
"I think, we can safely assume, that there are only experimental decorative projects outside using CSS animation currently and a huge amount of content using SMIL/SVG."

[03:05:00.0000] <smaug____>
that sounds right

[03:06:00.0000] <smaug____>
though, not if the context is the web only

[03:08:00.0000] <AnselmBradford>
annevk: Thanks for the response. There's a typemustmatch attribute? I know "type" is optional on <embed>, but it's this way on <object> too

[03:13:00.0000] <annevk>
yes there is on <object>

[03:13:01.0000] <annevk>
type has different semantics on <embed> and <object>

[03:20:00.0000] <AnselmBradford>
annevk: ahh gotcha, thanks. I was looking at the spec PDF... which is outdated on that attribute.

[03:23:00.0000] <AnselmBradford>
annevk: version I had anyway... got the new one :)

[03:24:00.0000] <hsivonen>
nessy: wait what? are there folks at Google who want to add non-WebVTT formats to WebKit? Why? Why not make sure WebVTT covers enough use cases? Do those other formats have proper specs?

[03:24:01.0000] <nessy>
hsivonen: it's not really decided yet, but some people would like to see other formats supported, too

[03:25:00.0000] <jgraham>
That sounds like every kind of bad

[03:25:01.0000] <hsivonen>
nessy: why on earth? are the people who want it experienced with working on browsers?

[03:25:02.0000] <nessy>
no but they are experienced with captions and have seen many formats float by

[03:25:03.0000] <hsivonen>
nessy: implementing other formats seems like a terrible idea for interop

[03:25:04.0000] <annevk>
I guess you can find people everywhere who want something else, does not mean they need to be catered for (e.g. the UMP people at Google)

[03:25:05.0000] <nessy>
hsivonen: I agree

[03:26:00.0000] <hsivonen>
annevk: so why isn't UMP off the charter yet?

[03:26:01.0000] <annevk>
The W3C and chairs are somehow extra-sensitive to this feedback

[03:27:00.0000] <annevk>
I don't really care either way

[03:27:01.0000] <jgraham>
It also seems like a terrible idea for Google/WebKit. Having to support N > 1 formats is much more code to maintain than supporting N = 1 formats

[03:27:02.0000] <annevk>
If I would work on CORS again it would likely be on a test suite rather than trying to resolve that gap

[03:27:03.0000] <hsivonen>
is this yet another case where non-Chrome folks at Google request stuff and the Chrome team isn't good enough at saying "no"?

[03:27:04.0000] <annevk>
Some fights are not really worth it

[03:27:05.0000] <annevk>
hsivonen, pretty much

[03:28:00.0000] <annevk>
jgraham, well they just hire some more people; if you see the amount of stuff that gets added to WebKit these days...

[03:28:01.0000] <nessy>
hsivonen: let's say I have spend many hours arguing back and forth

[03:29:00.0000] <annevk>
jgraham, see http://peter.sh/

[03:29:01.0000] <jgraham>
annevk: Well yeah having infinte resources isn't good for the soul. But still at some point the complexity will cause issues

[03:29:02.0000] <nessy>
thing is: I know there will be other formats (call them "proprietary" if you like, cause even if specs are open, they are not cross-browser supported) and I think <source> is probably the best solution to deal with this situation

[03:30:00.0000] <hsivonen>
do we have some kind of canned article for explaining why optional features suck for the Web and why should shouldn't propose optional features?

[03:30:01.0000] <nessy>
plus making sure webvtt is baseline format

[03:30:02.0000] <nessy>
in this case I actually think it makes sense - just to avoid the content sniffing problem

[03:30:03.0000] <nessy>
if we had <source> in <img>, we could avoid a lot of pain

[03:31:00.0000] <annevk>
<img> works just fine

[03:31:01.0000] <hsivonen>
I wonder how much the company-internal culture at Google has shifted to it being OK to have Chrome-only optional stuff and optimizing Google services for Chrome

[03:31:02.0000] <Ms2ger>
If only we could just throw out everything and start over, we could avoid even more pain

[03:31:03.0000] <annevk>
unfortunately there's no longer an XHTML WG

[03:32:00.0000] <annevk>
because otherwise I would suggest you join it :p

[03:32:01.0000] <jgraham>
hsivonen: From the outside it doesn't look like it has shifted much. That has always been OK

[03:32:02.0000] <hsivonen>
how would anyone suggest optional non-WebVTT formats without that the assumption that doing non-interoperable Chrome-only stuff is ok

[03:33:00.0000] <hsivonen>
how does Opera work with G+, BTW? I read somewhere that Opera is not "supported" but I don't know what that means

[03:34:00.0000] <annevk>
notifications are not enabled

[03:34:01.0000] <annevk>
the rest seems okayish

[03:34:02.0000] <annevk>
notifications toolbar that is

[03:35:00.0000] <nessy>
<img> still needs content sniffing to determine whether the resource will be decodable

[03:35:01.0000] <annevk>
yeah and it works fine

[03:35:02.0000] <nessy>
I hear from mobile people that it's a real pain

[03:36:00.0000] <zcorpan>
what's the pain?

[03:36:01.0000] <nessy>
delays and extra bandwidth use

[03:37:00.0000] <annevk>
huh?

[03:37:01.0000] <nessy>
can you please add whatever well thought-out objections to supporting multiple formats into the bug, so the discussion is out in the open?

[03:37:02.0000] <zcorpan>
when the page uses image formats that are not supported?

[03:38:00.0000] <nessy>
stuff like this, I guess: http://thomasjaehnel.com/blog/2009/11/attacks-based-on-content-sniffing.html

[03:39:00.0000] <nessy>
(obviously wasn't my argument :-)

[03:39:01.0000] <nessy>
I agreed to take the discussion to the group, so let's have it

[03:40:00.0000] <annevk>
that attach document is not about <img>

[03:40:01.0000] <annevk>
attack8

[03:40:02.0000] <annevk>
*

[03:40:03.0000] <annevk>
damnit

[04:30:00.0000] <gsnedders>
hsivonen: We've not officially supported by any Google stuff.

[04:31:00.0000] <gsnedders>
hsivonen: It's not just +. The only three that have been really painful was Gmail originally, Gmail v.2, and Wave.

[04:49:00.0000] <annevk>
zcorpan, briefly thought about moving html5-diff to Anolis; main problem is that xspec xrefs can only go to a multipage with script enabled (does not link to specific pages automatically)

[04:50:00.0000] <annevk>
zcorpan, so either Anolis would have to be updated, script-based redirects would have to be deemed acceptable, or we link to the singlepage version

[04:54:00.0000] <zcorpan>
annevk: personally i'm fine with either of those except link to singlepage

[04:56:00.0000] <annevk>
does W3C multipage do script-based redirects?

[04:57:00.0000] <annevk>
making Anolis smarter would be ideal of course

[04:58:00.0000] <Philip`>
/me supposes the way to do it automatically in Anolis would be to download http://www.whatwg.org/specs/web-apps/current-work/multipage/fragment-links.js and parse the first line to get the redirects

[05:00:00.0000] <annevk>
that might not be too complicated I guess

[05:00:01.0000] <annevk>
basically https://bitbucket.org/ms2ger/specification-data/src/bd982d2974a1/xrefs/dom/html.json needs to be updated every once and a while

[05:01:00.0000] <zcorpan>
s/and/in/

[05:01:01.0000] <annevk>
generating it from that URL might not be too hard

[05:01:02.0000] <annevk>
could be somewhat standalone

[05:01:03.0000] <annevk>
also has the benefit of adding missing entries

[05:03:00.0000] <annevk>
I can definitely do something like that later today, thanks Philip`!

[05:08:00.0000] <Ms2ger>
Terms with spaces might be annoying

[05:09:00.0000] <annevk>
hmm yeah

[05:10:00.0000] <hsivonen>
what would you call an attribute that gives the localization key for a UI string?

[05:12:00.0000] <annevk>
Ms2ger, we could keep name the current html.json html-base.json, and generate the new copy based on html-base.json and fragment-links.js

[05:13:00.0000] <annevk>
hsivonen, locale?

[05:13:01.0000] <annevk>
hsivonen, assuming you mean "en", "nl" etc.

[05:15:00.0000] <hsivonen>
annevk: no, I mean like isindexPrompt

[05:15:01.0000] <hsivonen>
annevk: the keys for UI string that get locale-specific values

[05:15:02.0000] <annevk>
hsivonen, oh, just isIndexPrompt?

[05:16:00.0000] <annevk>
hsivonen, use some approximation of what it is for / what it will say in English

[05:16:01.0000] <hsivonen>
annevk: I mean keys that you use to indentify localizable UI strings

[05:16:02.0000] <hsivonen>
annevk: localizationkey is a bit long

[05:16:03.0000] <hsivonen>
*identify

[05:17:00.0000] <annevk>
I'm afraid I don't understand

[05:17:01.0000] <hsivonen>
so you have a UI

[05:18:00.0000] <hsivonen>
it has localizable strings

[05:18:01.0000] <hsivonen>
for each string, you have an id

[05:18:02.0000] <hsivonen>
so that for each locale, you have mappings from the ids to strings in the language of the locale

[05:18:03.0000] <annevk>
right

[05:18:04.0000] <hsivonen>
suppose the page title is localizable

[05:19:00.0000] <hsivonen>
<h1>Pagetitle</h1>

[05:19:01.0000] <hsivonen>
and then you want to mark up the id of the string so that localization machinery can fill in the page title in another language

[05:19:02.0000] <hsivonen>
like

[05:20:00.0000] <hsivonen>
<h1 localization-key="pagetitle">Beekeeping</h1>

[05:20:01.0000] <hsivonen>
localization-key is a bit long

[05:20:02.0000] <Philip`>
"localizationkey is a bit long" - how about l13y?

[05:20:03.0000] <annevk>
aaah

[05:20:04.0000] <hsivonen>
Philip`: well, it could be just l10n

[05:20:05.0000] <hsivonen>
Philip`: but are we OK with numbers in attribute names in HTML?

[05:21:00.0000] <Philip`>
They're in element names, so why not?

[05:21:01.0000] <annevk>
which element name has a number?

[05:21:02.0000] <hsivonen>
<h1>!

[05:21:03.0000] <Philip`>
<h1>

[05:21:04.0000] <annevk>
doh

[05:21:05.0000] <hsivonen>
Philip`: good point

[05:21:06.0000] <Philip`>
plus many more

[05:22:00.0000] <annevk>
localize=""

[05:22:01.0000] <annevk>
localizeid=""

[05:22:02.0000] <annevk>
l10nid=""

[05:22:03.0000] <annevk>
localizationid=""

[05:23:00.0000] <annevk>
curious to know where this would be used

[05:23:01.0000] <Ms2ger>
Better than entities, I guess

[05:24:00.0000] <hsivonen>
annevk: l10nid was a possibility I was considering, but I thought it looked a bit ugly

[05:25:00.0000] <annevk>
localizeid="" or just localize="" seems kind of okay to me

[05:25:01.0000] <Philip`>
l10nid makes me think of meteor showers

[05:25:02.0000] <annevk>
just l10n is fine too I guess although a little awkward

[05:25:03.0000] <annevk>
Ms2ger, oh is this for Gecko?

[05:26:00.0000] <annevk>
finally getting rid of DTD madness :)

[05:26:01.0000] <Ms2ger>
No idea

[05:26:02.0000] <annevk>
is RDF next?

[05:26:03.0000] <annevk>
oh

[05:26:04.0000] <annevk>
be back later

[05:27:00.0000] <hsivonen>
annevk: yes, this is for Gecko. the proposal that I'm bikeshedding now says l10n_id, but underscores are obviously uncool for HTML

[05:27:01.0000] <smaug____>
for any l10n proposals, I'd ask what Pike thinks about it

[05:27:02.0000] <annevk>
hyphens are uncool too

[05:27:03.0000] <annevk>
generally

[05:28:00.0000] <hsivonen>
annevk: hyphens are OK, if there's a set of attributes: l10n-id, l10n-foo, l10n-bar, etc.

[05:28:01.0000] <hsivonen>
smaug____: I sort of assumed this has already been past Pike

[05:28:02.0000] <annevk>
i guess, but do we need more than one?

[05:29:00.0000] <hsivonen>
annevk: I don't know yet

[05:29:01.0000] <zcorpan>
hsivonen: can't id="" be used?

[05:29:02.0000] <hsivonen>
zcorpan: I don't know

[05:29:03.0000] <annevk>
pointer to bug report?

[05:30:00.0000] <smaug____>
hsivonen: what you mean "past Pike"

[05:30:01.0000] <smaug____>
hsivonen: he happens to have lots of experience in l10n area

[05:30:02.0000] <hsivonen>
smaug____: I assumed that Pike has already seen this an not vetoed

[05:30:03.0000] <hsivonen>
smaug____: but I don't know if that's the case

[05:31:00.0000] <smaug____>
hsivonen: what is this about btw?

[05:31:01.0000] <smaug____>
localizing what?

[05:31:02.0000] <smaug____>
and where?

[05:31:03.0000] <hsivonen>
smaug____: XUL first, probably HTML later

[05:31:04.0000] <hsivonen>
smaug____: in Firefox chrome first. maybe elsewhere later

[05:32:00.0000] <nessy>
l10n-key ?

[05:32:01.0000] <nessy>
locale-key ?

[05:32:02.0000] <hsivonen>
nessy: locale-key would imply the wrong thing

[05:32:03.0000] <nessy>
yeah, I guess

[05:33:00.0000] <hsivonen>
The bug is https://bugzilla.mozilla.org/show_bug.cgi?id=566906

[05:34:00.0000] <zcorpan>
what's L20n?

[05:34:01.0000] <hsivonen>
zcorpan: I think it's the next level of L10n

[05:35:00.0000] <smaug____>
https://wiki.mozilla.org/L20n has some links

[05:37:00.0000] <zcorpan>
Localizilizziliziation

[05:37:01.0000] <nessy>
did you count this?

[05:38:00.0000] <zcorpan>
my text editor did

[05:38:01.0000] <nessy>
hehe

[05:39:00.0000] <zcorpan>
well then obviously the attribute should read l20n-something rather than l10n-something

[05:44:00.0000] <Ms2ger>
"Tab tries to explain the difference between semantic markup and stylistic presentation to the MS folks"

[05:44:01.0000] <Ms2ger>
TabAtkins++

[05:44:02.0000] <hsivonen>
Ms2ger: is this CSS WG minutes?

[05:44:03.0000] <Ms2ger>
Yes

[05:44:04.0000] <hsivonen>
did the actual explanation get minuted?

[05:46:00.0000] <Ms2ger>
Doesn't look like it

[05:47:00.0000] <hsivonen>
Ms2ger: :-(

[05:47:01.0000] <Ms2ger>
Maybe Tab can write it out

[05:49:00.0000] <gsnedders>
From memory. Word for word.

[05:49:01.0000] <Ms2ger>
He's a smart guy

[05:50:00.0000] <jgraham>
Maybe he did it using drawings of foxes

[06:15:00.0000] <smaug____>
ahaa, a new idea for the mutation events replacement. /me needs to paper and pen

[06:15:01.0000] <smaug____>
s/to/a/

[06:16:00.0000] <gsnedders>
smaug____: Not to a keyboard and a text editor?

[06:16:01.0000] <zcorpan>
replace it with foxes?

[06:18:00.0000] <smaug____>
in some cases it is easier to just use paper

[06:40:00.0000] <smaug____>
and the idea might even work :)

[06:40:01.0000] <zcorpan>
/me was unaware of http://www.ietf.org/mail-archive/web/ietf-types/current/msg01115.html

[06:47:00.0000] <hsivonen>
application/font-* instead of font/* shows how hard it is to work with IANA/IETF

[06:50:00.0000] <zcorpan>
seems that type still isn't registered

[06:55:00.0000] <MikeSmith>
hsivonen: thanks for the heads-up about that bug related to es5.github.com

[06:55:01.0000] <annevk>
hmm smaug left

[06:56:00.0000] <annevk>
I was wondering whether we could call the new stuff modification

[06:56:01.0000] <MikeSmith>
hsivonen: I'm glad for my lack of understanding on how to properly set the encoding on an external CSS stylesheet :)

[06:56:02.0000] <hsivonen>
MikeSmith: :-)

[07:02:00.0000] <MikeSmith>
othermaciej: my earlier LC bug count was way off, btw

[07:02:01.0000] <MikeSmith>
I think I was counting only open bugs

[07:03:00.0000] <MikeSmith>
the actual count of bugs submitted after Oct 1 until today is currently 1502

[07:11:00.0000] <jgraham>
Oh, that explains all the email I have had to ignore

[07:13:00.0000] <annevk>
we had a six months last call?

[07:14:00.0000] <MikeSmith>
kind of

[07:14:01.0000] <MikeSmith>
because decision was that all bugs submitted after Oct 1 would be treated just the same as bugs submitted during the formal LC window

[07:16:00.0000] <annevk>
i see

[07:38:00.0000] <bga_>
hm. we still haven't api to gzip compress "ajax" POST data at clientside. it will sufficient reduce world traffic

[07:39:00.0000] <Ms2ger>
Doubtful

[07:39:01.0000] <bga_>
i can compress using canvas api and toDataURL but its hack

[07:41:00.0000] <annevk>
smaug____, is your new idea going to please everyone?

[07:41:01.0000] <annevk>
smaug____, maybe we can call the new stuff "modification" rather than "mutation"? might make it easier to distinguish between the two

[07:42:00.0000] <bga_>
Ms2ger anyway that api will be usefull

[07:42:01.0000] <bga_>
when i send big json from client to server

[07:43:00.0000] <bga_>
traffic  == money

[07:43:01.0000] <bga_>
you can save my money :)

[07:44:00.0000] <annevk>
write your own impl

[07:44:01.0000] <bga_>
pure js?

[07:44:02.0000] <annevk>
sure

[07:44:03.0000] <bga_>
or spec?

[07:44:04.0000] <annevk>
no

[07:44:05.0000] <bga_>
js too slow

[07:44:06.0000] <Ms2ger>
Doubtful

[07:45:00.0000] <bga_>
https://gist.github.com/519305/51a48139351b3fae048504a6a8d7631c5ae991db

[07:45:01.0000] <bga_>
wrong

[07:45:02.0000] <bga_>
https://gist.github.com/519305

[07:46:00.0000] <bga_>
its gzip like algo for compress js source

[07:46:01.0000] <bga_>
i made it for js1k

[07:47:00.0000] <bga_>
but it compress 70k 1 min

[07:47:01.0000] <bga_>
too slow

[07:47:02.0000] <bga_>
we need native api imho

[07:51:00.0000] <MikeSmith>
annevk: btw, I did request a new dvcs rep for cors and hopefully it'll be available today or tomorrow

[07:53:00.0000] <Philip`>
bga_: That doesn't look much like the gzip algorithm

[07:54:00.0000] <Philip`>
gzip in C usually goes at over 10MB/sec, and a JS port shouldn't be more than maybe ten or a hundred times slower

[07:55:00.0000] <bga_>
Philip` its pattern compression. sorry

[07:56:00.0000] <bga_>
deflate is much simpler

[07:56:01.0000] <Philip`>
If your code takes a minute for 70KB then it's just because your algorithm is incredibly inefficient, not because of any fundamental problem with JS :-)

[07:58:00.0000] <timeless>
jcranmer, `It is frequently desirable to split large binary files into multiple parts` `for transmissio n over the Internet.  Such binaries are often rendered` I like the second word of the second line...

[07:58:01.0000] <jgraham>
Yes, I think jszip (which solves a similar but not identical problem) is much faster than that

[07:58:02.0000] <jgraham>
and using typed arrays or something might be even faster

[07:59:00.0000] <timeless>
MikeSmith: yes, bugzilla doesn't encode email addresses for comment lines, it stores user ids, so when you change your email address, any place where that was used automatically (reporter, cc, assignee, qa, commenter, ...) will be updated

[07:59:01.0000] <MikeSmith>
right

[08:00:00.0000] <MikeSmith>
I just can't remember whether end users can change their account/addresses themselves

[08:00:01.0000] <MikeSmith>
I know I can in the admin UI for the W3C bugzilla

[08:00:02.0000] <Ms2ger>
I think AryehGregor couldn't

[08:01:00.0000] <timeless>
othermaciej: i'm `shocked` to hear that `text/html-sandboxed` isn't doing what someone wanted... oh wait, i think i suggested it was likely to fail :)

[08:01:01.0000] <MikeSmith>
annevk: http://dvcs.w3.org/hg/cors/

[08:01:02.0000] <jgraham>
They can't on W3C bugzilla

[08:01:03.0000] <jgraham>
I thought they could on ohter bugzilla

[08:02:00.0000] <jgraham>
s

[08:02:01.0000] <MikeSmith>
annevk: you want me to copy the source over, or will you?

[08:02:02.0000] <annevk>
open feedback

[08:02:03.0000] <annevk>
MikeSmith, if you can

[08:02:04.0000] <annevk>
i can do it too actually

[08:02:05.0000] <annevk>
let me do it :)

[08:05:00.0000] <MikeSmith>
annevk: ok

[08:06:00.0000] <MikeSmith>
annevk: once you get it set up, I'd suggest also putting a redirect from the old cvs location to the dvcs one

[08:07:00.0000] <annevk>
yeah

[08:13:00.0000] <smaug____>
annevk: the idea won't please Google

[08:14:00.0000] <smaug____>
since they want, for not-clear-to-me reason the almost-async handling

[08:14:01.0000] <timeless>
hsivonen: `term` might work re localization key names

[08:15:00.0000] <hsivonen>
timeless: thanks

[08:15:01.0000] <timeless>
/me is obviously a few hours behind

[08:16:00.0000] <smaug____>
annevk: actually I'm investigating two different approaches now.

[08:18:00.0000] <timeless>
MikeSmith: the w3 bugzilla doesn't allow it

[08:18:01.0000] <timeless>
you need to use https://bugzilla/editusers.cgi to do it as an admin

[08:19:00.0000] <timeless>
normally bugzilla does allow users to update their email addresses

[08:19:01.0000] <MikeSmith>
timeless: OK

[08:20:00.0000] <MikeSmith>
so I will wait to hear back from AryehGregor

[08:20:01.0000] <annevk>
Ms2ger, I will place my script that generates a better html.json (once done) in specification-data, okay?

[08:20:02.0000] <Ms2ger>
K

[08:20:03.0000] <timeless>
> That's much too early. Prepare to fast-forward!

[08:20:04.0000] <annevk>
Ms2ger, you just parse it using json right?

[08:20:05.0000] <annevk>
Ms2ger, so the generated copy can be less nice?

[08:21:00.0000] <timeless>
> We are at now, now

[08:21:01.0000] <Ms2ger>
Sure

[08:21:02.0000] <timeless>
/me hearts ludicrous speed!

[08:21:03.0000] <Ms2ger>
I hope you're generating it with a json serializer as well ;)

[08:22:00.0000] <annevk>
I'm using json from Python 2.6

[08:22:01.0000] <annevk>
and I use some string replacements hacks on the original data

[08:34:00.0000] <annevk>
wow that was easier than expected somehow

[08:35:00.0000] <annevk>
I guess I should add some comments so Ms2ger has less reason to laugh

[08:35:01.0000] <annevk>
and rename my variables from test1 etc.

[08:35:02.0000] <Ms2ger>
I don't need reasons to laugh, thought you knew that ;)

[08:43:00.0000] <annevk>
Ms2ger, who uses "non-conforming element"?

[08:43:01.0000] <annevk>
I'm removing it for now as I cannot find its new links

[08:43:02.0000] <annevk>
link*

[08:44:00.0000] <Ms2ger>
If editcommands doesn't, nobody, I think

[08:46:00.0000] <smaug____>
hsivonen: unfortunately we need to experiment quite a bit with menu/command handling, since the spec is just not quite good

[08:46:01.0000] <smaug____>
hsivonen: I wonder if we could, in gecko, have a pref to enable our experimental menu/command handling

[08:48:00.0000] <annevk>
Ms2ger, https://bitbucket.org/ms2ger/specification-data/changeset/26462f44156d

[08:49:00.0000] <annevk>
html.py has the "fun"

[08:49:01.0000] <hsivonen>
smaug____: does the Experiment Markup Language (XML) not work for experimentation on the DOM level?

[08:50:00.0000] <hsivonen>
smaug____: I'd really like to avoid landing browser-specific HTML parsing algorithm changes on the trunk

[08:50:01.0000] <smaug____>
hsivonen: even with a pref or #ifdef?

[08:51:00.0000] <hsivonen>
smaug____: can we at least have a round of discussion on a multivendor forum before we go there?

[08:51:01.0000] <hsivonen>
smaug____: I supposed a pref is a possibility

[08:51:02.0000] <hsivonen>
*suppose

[08:52:00.0000] <hsivonen>
/me leaves the keyboard to get to places that have closing times

[08:52:01.0000] <smaug____>
does anyone else edit html spec than Hixie?

[08:55:00.0000] <Ms2ger>
No

[09:03:00.0000] <dglazkov>
good morning, Whatwg!

[09:03:01.0000] <dglazkov>
and especially to you, cranky-pants jgraham

[09:03:02.0000] <Ms2ger>
Good evening, dglazkov

[09:03:03.0000] <annevk>
heh

[09:07:00.0000] <jgraham>
I'm not cran…

[09:07:01.0000] <jgraham>
OK I am

[09:08:00.0000] <jgraham>
I can't deny it

[09:08:01.0000] <Ms2ger>
/me pats jgraham on the back

[09:08:02.0000] <annevk>
sweet, #fx logs

[09:18:00.0000] <Ms2ger>
Philip`, what do you think about putting up a JSON file with the date from fragment-links.js?

[09:20:00.0000] <annevk>
are you offended by my code?

[09:21:00.0000] <Ms2ger>
No, by the fact it's necessary

[09:21:01.0000] <Ms2ger>
(And brittle)

[09:22:00.0000] <Ms2ger>
Did you file a bug to remove HTMLElement.{id,className,classList} already?

[09:25:00.0000] <Philip`>
Ms2ger: "date"? (Do you mean data?)

[09:25:01.0000] <Ms2ger>
Yes

[09:26:00.0000] <Philip`>
Is the .js file plus regexps to extract the data not sufficient?

[09:28:00.0000] <annevk>
Ms2ger, I didn't

[09:28:01.0000] <Philip`>
/me looks at the code

[09:29:00.0000] <Philip`>
Oh, JSON doesn't like single quotes? How silly

[09:29:01.0000] <annevk>
going from single to double and not using escapes would already help a lot

[09:31:00.0000] <Ms2ger>
Filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=13610

[09:31:01.0000] <Philip`>
If someone wants to update http://code.google.com/p/html5/source/browse/trunk/spec-splitter/spec-splitter.py#328 then I can deploy the new code

[09:32:00.0000] <Ms2ger>
I'd do that if I had access

[09:33:00.0000] <annevk>
I think there might be a couple of other things HTML has not updated on yet

[09:34:00.0000] <annevk>
Ms2ger, you do now

[09:35:00.0000] <annevk>
Ms2ger, btw, I wonder how controversial that bug is going to be

[09:36:00.0000] <Ms2ger>
We'll see

[09:40:00.0000] <annevk>
the whole section on DOM Trees is rather outdated

[09:43:00.0000] <annevk>
filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=13612 on using cloning steps

[09:45:00.0000] <Ms2ger>
Look ma, I can do SVN

[09:48:00.0000] <othermaciej>
MikeSmith: seems fair to count those

[09:49:00.0000] <othermaciej>
MikeSmith: I wonder how that compares to the prior 9-month period

[09:51:00.0000] <Ms2ger>
Philip`, feel like deploying? :)

[09:56:00.0000] <Philip`>
Ms2ger: Done, and regenerated http://www.whatwg.org/specs/web-apps/current-work/multipage/fragment-links.js

[10:05:00.0000] <Ms2ger>
Ta

[10:11:00.0000] <leaverou>
hi everyone

[10:12:00.0000] <leaverou>
not sure if this is the right place for this, but I don't think inline event handlers are setting a good example here http://www.whatwg.org/specs/web-apps/current-work/#menus-intro

[10:12:01.0000] <Ms2ger>
Meh

[10:13:00.0000] <annevk>
leaverou, sometimes inline event handlers are fine

[10:13:01.0000] <smaug____>
leaverou: why not?

[10:13:02.0000] <smaug____>
I don't see anything bad with those inline event handlers

[10:14:00.0000] <smaug____>
(the menu handling itself will, I hope, change a bit)

[10:14:01.0000] <leaverou>
because the spec example is copied by tutorials all over and new authors will get the impression that inline event handlers are fine

[10:14:02.0000] <zewt>
it sounds like you're under the impression that they're "evil"

[10:14:03.0000] <leaverou>
so it's indirectly (?) promoting bad practices

[10:14:04.0000] <smaug____>
is it a bad practice?

[10:15:00.0000] <leaverou>
in most cases it is, it violates separation of concerns

[10:16:00.0000] <zewt>
do you want to write 3x as much code for no benefit? heh

[10:17:00.0000] <leaverou>
zewt: are you going to praise the <font> tag next? :)

[10:17:01.0000] <zewt>
sarcasm doesn't lead to convincing arguments

[10:18:00.0000] <leaverou>
and it's hardly 3x as much code, just like presentational HTML over CSS, it's actually less code when repetition is involved

[10:18:01.0000] <leaverou>
also, polluting the global namespace with all these functions is a JS bad practice

[10:18:02.0000] <Ms2ger>
Meh

[10:19:00.0000] <zewt>
indeed, meh

[10:19:01.0000] <zewt>
sounds like a typical "xxx considered harmful" article

[10:19:02.0000] <leaverou>
"meh" is even a less convincing argument than sarcasm

[10:20:00.0000] <leaverou>
anyway, I'm not here to convince people for things that are almost universally accepted in the indurstry, just to notify about it in case it was overlooked

[10:20:01.0000] <annevk>
I would be careful with universally acceptedd

[10:20:02.0000] <annevk>
obviously it is not here

[10:20:03.0000] <Ms2ger>
I would challenge "almost universally accepted in the indurstry"

[10:20:04.0000] <Ms2ger>
If I cared

[10:20:05.0000] <Philip`>
The HTML spec even has examples with uppercase tags, which I'm sure every sane person in the world thinks is crazy

[10:21:00.0000] <Philip`>
so it's not trying to be a guide to any set of best practices

[10:22:00.0000] <leaverou>
Philip`: that's a good point.

[10:22:01.0000] <AryehGregor>
Philip`, it has examples with *mixed*-case tags.

[10:22:02.0000] <AryehGregor>
I'm pretty sure it has something like <BloCkquoTE> somewhere.

[10:22:03.0000] <AryehGregor>
The philosophy it adopts is that if something is valid, it wants to use it occasionally in examples so people don't assume it's wrong because they never see it.

[10:23:00.0000] <AryehGregor>
If inline event handlers were always a bad idea, the spec would say they're invalid.

[10:23:01.0000] <AryehGregor>
There's no class of things called "fully conforming but we think they shouldn't be used", that would be silly.

[10:24:00.0000] <AryehGregor>
If we think they shouldn't be used, we should make them at least raise a warning, and if we think it's fine to use them, why not use them in examples?

[10:24:01.0000] <leaverou>
AryehGregor: Ok, that's a valid argument and I agree. Btw, I never said inline event handlers are always bad, just in many cases.

[10:24:02.0000] <karlcow>
AryehGregor: it is a binary reasoning :)

[10:27:00.0000] <jgraham>
/me wonders if anyone has checked the uppercase vs lowercase letters in tags in examples for hidden messages

[10:29:00.0000] <AryehGregor>
MikeSmith, thanks for the Bugzilla component!  And I can't change my e-mail address on Bugzilla, so I'd appreciate it if you changed it to ayg⊙an

[10:30:00.0000] <Ms2ger>
jgraham, last time I checked, they called you a cranky-pants

[10:30:01.0000] <AryehGregor>
(I'm trying to standardize when it comes to publicly-visible addresses, so as not to hopelessly confuse everyone with Simetrical⊙gc vs. AryehGregor⊙gc vs. ayg⊙an vs. all of those with plus-addressing vs. a few others I have hidden up my sleeve)

[10:31:00.0000] <AryehGregor>
(when people show up with multiple addresses in my address book, I never know which to use)

[10:32:00.0000] <jgraham>
Ms2ger: That isn't cat related, so it seems unlikely

[10:32:01.0000] <Ms2ger>
Hah

[10:59:00.0000] <AryehGregor>
Wow, the number of bugs filed today is truly ludicrous.

[10:59:01.0000] <AryehGregor>
Last Call does seem to prompt a lot of feedback.

[11:00:00.0000] <smaug____>
last call of what?

[11:00:01.0000] <AryehGregor>
HTML5.

[11:00:02.0000] <smaug____>
oh, last call of w3c html5

[11:00:03.0000] <smaug____>
/me doesn't care :)

[11:00:04.0000] <AryehGregor>
I don't either, I'm just remarking on the mounds of feedback.

[11:00:05.0000] <AryehGregor>
Lots of people and organizations apparently do care.

[11:01:00.0000] <AryehGregor>
Including, e.g., Microsoft.

[11:01:01.0000] <smaug____>
it is good the get feedback, of course

[11:03:00.0000] <smaug____>
but I don't understand what the last call means in this case

[11:06:00.0000] <AryehGregor>
There's actually loads of high-quality feedback here from Microsoft.

[11:06:01.0000] <AryehGregor>
I have no idea why they put it off to just before Last Call instead of filing as soon as they came up with it.

[11:06:02.0000] <zewt>
"time to give the feedback that you've been holding back for some mysterious reason"

[11:07:00.0000] <AryehGregor>
This Greg Lowney fellow from access-research.org is also providing lots of detailed bugs, although I don't know how many of them will wind up being accepted.

[11:09:00.0000] <AryehGregor>
This is seriously like a hundred bugs filed in the last day or something.

[11:09:01.0000] <AryehGregor>
Or last two days.

[11:10:00.0000] <AryehGregor>
And most of them are fairly good.

[11:10:01.0000] <AryehGregor>
annevk, looks like the cat is out of the bag with dom.spec.whatwg.org, huh?

[11:12:00.0000] <AryehGregor>
Apparently we got some feedback on accessibility from the state of Virginia.

[11:12:01.0000] <zewt>
do people in virginia have specific accessibility needs

[11:13:00.0000] <Ms2ger>
Was that the .doc?

[11:13:01.0000] <AryehGregor>
I think this one was an e-mail, but I archived it without looking too closely.

[11:14:00.0000] <TabAtkins_>
Nah, this was an email.

[11:15:00.0000] <AryehGregor>
Hixie, browsers have to support ES parseInt and parseFloat anyway, so why is their instability or whatever any kind of argument against reusing them?  They exist and we have to deal with them already, and it's extremely confusing to have them work almost the same but not quite.

[11:17:00.0000] <smaug____>
/me agrees with AryehGregor

[11:17:01.0000] <AryehGregor>
smaug____, well, I can't do anything, but if you want to threaten implementer veto, you have my full support.  :)

[11:21:00.0000] <smaug____>
/me hopes bz will just reopen the bug :)

[11:23:00.0000] <AryehGregor>
I forgot that I can sic implementers on him.

[11:23:01.0000] <AryehGregor>
I'll send an e-mail to whatwg once I finish archiving the flood in my spec inbox.

[11:30:00.0000] <AnselmBradford>
does anyone know of a working example of using <object> in form submission? The example on http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-November/017363.html appears to no longer work (on Mac anyway)

[12:03:00.0000] <jgraham>
AryehGregor: Probably the W3C will treat all this last-minute feedback as a vindication of the whole Process rather than evidence that it is a failure because it discourages people from doing ongoing review

[12:06:00.0000] <AryehGregor>
jgraham, I'm not sure they'd be wrong.  Psychologically, if there's no deadline, maybe people will think "meh, I'll get around to it someday" and then never do it.

[12:07:00.0000] <AryehGregor>
I dunno.

[12:07:01.0000] <AryehGregor>
Even if does encourage feedback, though, I think that's massively outweighed by the disadvantages.

[12:09:00.0000] <jgraham>
AryehGregor: If the feedback is coming from implementors they should be providing it as they implement, at the latest. If no one has implemented the feature at LC the feedback it gets won't be very useful

[12:09:01.0000] <AryehGregor>
In theory, CR is the call for implementations.

[12:09:02.0000] <jgraham>
Same reason I don't think that having these "releases" of the testsuite don't make sense

[12:10:00.0000] <jgraham>
People should provide feedback when they run the tests on their own internal system

[12:10:01.0000] <jgraham>
That doesn't happen on a fixed timescale

[12:13:00.0000] <Ms2ger>
I know it sounds unlikely, but non-implementers can make useful comments as well ;)

[12:14:00.0000] <jgraham>
Well, yes

[12:14:01.0000] <jgraham>
Although I think 0 non-implementors have commented on the testsuite so far

[12:15:00.0000] <jgraham>
I know there is like ~1 contributer to the CSS testsuite that is not an implementor

[12:15:01.0000] <Ms2ger>
Oh, who's that?

[12:16:00.0000] <Ms2ger>
(I haven't followed Kris' plans with the test suite, and will probably ignore any releases he does)

[12:17:00.0000] <Ms2ger>
/me wonders why the rsync messages differ between the html and webapps repos

[12:17:01.0000] <timeless>
heh

[12:19:00.0000] <jgraham>
Ms2ger: I believe Gerard Talbot is not an implementor. I don't know if there are others

[12:19:01.0000] <TabAtkins>
Gerard is not.

[12:19:02.0000] <AryehGregor>
jgraham, I'm not an implementer.

[12:19:03.0000] <TabAtkins>
A few other people aren't either.

[12:19:04.0000] <Ms2ger>
Oh, CSS

[12:19:05.0000] <AryehGregor>
Although I happen to be working for an implementer, I have no more contact with the actual implementer people than you do.

[12:19:06.0000] <AryehGregor>
Dunno about CSS.

[12:20:00.0000] <Ms2ger>
I managed to read "CSS" as "HTML"

[12:20:01.0000] <jgraham>
AryehGregor: You are working for Google who are implementors

[12:20:02.0000] <AryehGregor>
(I mean, the actual Google implementer people)

[12:20:03.0000] <AryehGregor>
Yes, but that's random coincidence.

[12:20:04.0000] <AryehGregor>
I wrote the test suite before I was working for them.

[12:20:05.0000] <Ms2ger>
Does Philip` work for an implementer?

[12:20:06.0000] <AryehGregor>
And it wasn't originally part of my job either, I asked for permission to do some extra work on my preexisting reflection tests as part of my job.

[12:20:07.0000] <AryehGregor>
/me doesn't think so

[12:21:00.0000] <jgraham>
Philip` is the counterexample to any general statement

[12:21:01.0000] <AryehGregor>
True.

[12:22:00.0000] <Ms2ger>
Hence, false

[12:23:00.0000] <jgraham>
Well technically the repository shows 3 counterexamples; AryehGregor (who isn't really one), David Carlisle and Philip`

[12:23:01.0000] <AryehGregor>
I am too one.

[12:23:02.0000] <AryehGregor>
I submitted those tests in my spare time.

[12:23:03.0000] <AryehGregor>
I did like 75% of the work unpaid.

[12:24:00.0000] <AryehGregor>
At least.

[12:24:01.0000] <jgraham>
I wrote the test harness in my spare time, but I can't count myself :p

[12:25:00.0000] <gsnedders>
(Though I did the initial review of it as work :P)

[12:26:00.0000] <AryehGregor>
jgraham, you're an actual implementer, though, not just someone who happens to work for an implementer but has no contact with the implementers who work there.

[12:29:00.0000] <timeless>
MikeSmith: ping

[12:29:01.0000] <timeless>
... has w3 ever rescindend any specifications for any reason?

[12:29:02.0000] <jgraham>
AryehGregor: Well whatever. Clearly my initial statement is factually incorrect. But the counterexamples are an editor of a different spec (who works for an implementor), someone who thinks that generating Fortan from xml using xslt is not only sane but enjoyable, and Philip`. Therefore I consider the general thrust of my statement to be correct ;)

[12:29:03.0000] <timeless>
ooh, yes

[12:30:00.0000] <timeless>
http://www.w3.org/TR/2009/PER-xhtml11-20090507/ is rescinded

[12:30:01.0000] <timeless>
cool

[12:30:02.0000] <AryehGregor>
jgraham, on the other hand, very few tests have been submitted period, so those exceptions are fairly significant proportionally.

[12:30:03.0000] <AryehGregor>
Especially if you weight by test number and quality.

[12:30:04.0000] <AryehGregor>
Last I checked, Philip` and I count for a lot extra in that case.

[12:30:05.0000] <Ms2ger>
You beat MS by a huge margin, that's clear :)

[12:31:00.0000] <jgraham>
Well you basically autogenerated a bunch of tests from a table. Which is useful but cheating if we're using number-of-tests as a metric :)

[12:31:01.0000] <Ms2ger>
/me patiently awaits Kris' reply about automated tests

[12:32:00.0000] <jgraham>
Ms2ger: Reply to what?

[12:32:01.0000] <Ms2ger>
My reply about Matheus' test I submitted

[12:34:00.0000] <jgraham>
Oh, about their early test submisions being updated

[12:34:01.0000] <Ms2ger>
Yeah

[12:37:00.0000] <Philip`>
Ms2ger: I don't really work for anyone

[12:38:00.0000] <AryehGregor>
jgraham, I autogenerated a bunch of tests from thousands of lines of carefully-written JavaScript.

[12:39:00.0000] <AryehGregor>
The bulk of my tests is the logic, not the tables.

[12:39:01.0000] <Ms2ger>
That bulk is, unfortunately, unreadable :)

[12:39:02.0000] <AryehGregor>
Sadly.

[12:39:03.0000] <AryehGregor>
To be fair, testharness.js is also unreadable, at least if you ask me.

[12:40:00.0000] <AryehGregor>
/me looked into trying to make the pass/fail checkboxes have a proper <label> so you have a larger click target, but quickly gave up

[12:40:01.0000] <Philip`>
/me still thinks his canvas test .yaml files are reasonably readable

[12:40:02.0000] <Philip`>
(The .py files less so)

[12:40:03.0000] <Ms2ger>
I haven't tried

[12:40:04.0000] <AryehGregor>
Well, my data files are super readable.

[12:40:05.0000] <AryehGregor>
It's only the thousands of lines of logic that's incomprehensible.

[12:41:00.0000] <jgraham>
AryehGregor: I don't see why that would be hard

[12:41:01.0000] <Ms2ger>
Not really, the necessary data is split over two files

[12:41:02.0000] <jgraham>
Although I agree that the templating stuff is "not wonderful"

[12:41:03.0000] <Ms2ger>
So I actually found your data unreadable as well :)

[12:41:04.0000] <AryehGregor>
jgraham, I'm sure it's not hard for you.

[12:41:05.0000] <AryehGregor>
Ms2ger, well, it's actually split over like ten files, because people wanted smaller test files for some reason.

[12:41:06.0000] <AryehGregor>
Originally it was all in one file.

[12:41:07.0000] <AryehGregor>
Along with the logic.

[12:42:00.0000] <AryehGregor>
And the harness.

[12:42:01.0000] <Philip`>
(I suppose it could be argued that mixing YAML and HTML and JS and Python syntax, and the canvas API and the Cairo API, in a single source file, is not an ideal way to make something easy to understand, but I don't mind it myself)

[12:43:00.0000] <Ms2ger>
Yeah, I don't care about the ten files, but about the fact that the data for a single element-attribute combination is split over two files

[12:43:01.0000] <AryehGregor>
It was either that or make the element data much longer and more error-prone.

[12:43:02.0000] <AryehGregor>
Originally I didn't split it, but that involved painful amounts of repetition.

[12:43:03.0000] <AryehGregor>
I could un-split it pretty easily, though.  Maybe format the data files a bit more consistently too.

[12:44:00.0000] <Ms2ger>
OTOH, it would make it easier to review for correctness

[12:44:01.0000] <AryehGregor>
Like have each attribute be one line.

[12:44:02.0000] <AryehGregor>
Yeah, true.

[12:44:03.0000] <Ms2ger>
I think it would reduce my headaches after dealing with reflection, at least ;)

[12:45:00.0000] <Ms2ger>
(Not as much as someone fixing Gecko's integer reflection code would, but still)

[12:51:00.0000] <jgraham>
AryehGregor: Pushed the label fix

[12:51:01.0000] <AryehGregor>
jgraham, thanks.

[12:51:02.0000] <jgraham>
It annoyed me too :)

[12:53:00.0000] <Hixie>
AryehGregor: when it comes to implementation simplicity, browsers are not the concern.

[12:54:00.0000] <AryehGregor>
Hixie, what implementers are you worried about that will need to exactly implement this part of HTML but will not need to implement ES as well?

[12:56:00.0000] <Hixie>
AryehGregor: all the off-the-shelf tools? the vast majority of HTML processors don't support ES, nor are they updated frequently.

[12:56:01.0000] <Hixie>
AryehGregor: should we defer to other specs for everything where we happen to be a subset of another spec's algorithms?

[12:56:02.0000] <AryehGregor>
Hixie, examples?  Ones that need to exactly match how browsers behave?

[12:56:03.0000] <AryehGregor>
Hixie, I don't propose that the algorithm be a subset, I propose that it be made identical.

[12:56:04.0000] <Hixie>
you want to support Infinity?

[12:56:05.0000] <AryehGregor>
Not just an editorial change, a normative change.

[12:57:00.0000] <Hixie>
you want to support a radix attribute on integer attributes?

[12:57:01.0000] <AryehGregor>
Identical with some extra error handling tacked on, okay.

[12:57:02.0000] <Hixie>
subset.

[12:57:03.0000] <Hixie>
or superset, i guess.

[12:57:04.0000] <Hixie>
either way, not identical.

[12:57:05.0000] <AryehGregor>
I already said that it should be the same as parseInt(x, 10), not parseInt(x), so there's no radix issue.

[12:57:06.0000] <Hixie>
that's a subset then.

[12:58:00.0000] <AryehGregor>
And yes, I think that when we have a normative dependency on a spec in practice anyway, we should defer to it where that would make the resulting spec shorter.  In practice we have a normative dependency on ES no matter what, there's no mileage in trying to avoid it.

[12:58:01.0000] <AryehGregor>
But anyway, I posted to whatwg, we'll see what the implementers think.

[12:58:02.0000] <Hixie>
making specs shorter is not a goal.

[12:59:00.0000] <Hixie>
we could make the spec massively shorter, e.g. by dropping all the domintro stuff and introduction sections, all the examples, all the notes

[12:59:01.0000] <Hixie>
every time you have an indirection through anothe spec, it adds significant complexity

[13:00:00.0000] <Hixie>
you have to track the other spec if it changes

[13:00:01.0000] <jgraham>
/me prefers the idea of resusing ES algorithms in general

[13:00:02.0000] <Hixie>
either normatively or just changing the editorial issues like the section or algorithm name

[13:00:03.0000] <AryehGregor>
jgraham, that would be useful to say on whatwg, especially if you're in a position to say whether Opera's willing to do what the spec says.

[13:00:04.0000] <Hixie>
you have to read two specs instead of one

[13:00:05.0000] <Hixie>
etc etc etc.

[13:00:06.0000] <Hixie>
and in this particular case, the ES definition isn't stable, it changes over time

[13:01:00.0000] <Hixie>
which is just plain silly.

[13:01:01.0000] <jgraham>
Not least because there is a non-zero chance that people implementing the feature will just say "oh we have code that does that already" and call that rather than looking for all the subtle differences

[13:01:02.0000] <jgraham>
Which leads to poor interop

[13:01:03.0000] <Hixie>
there's tons of different code to parse integers and floats

[13:01:04.0000] <Hixie>
why should we match ES rather than CSS?

[13:01:05.0000] <Hixie>
or C?

[13:02:00.0000] <jgraham>
In an ideal world HTML and ES and CSS would all match

[13:02:01.0000] <Hixie>
jgraham: the problem you describe is one that validators solve in a hurry.

[13:02:02.0000] <Hixie>
er

[13:02:03.0000] <Hixie>
s/validators/tests/

[13:03:00.0000] <jgraham>
If the problems the tests bring up require lots of new code and only affect edge cases people are unlikley to be in a hurry to fix bugs

[13:03:01.0000] <jgraham>
+those

[13:03:02.0000] <Hixie>
this is not a case that needs lots of new code

[13:03:03.0000] <Hixie>
in fact it requires significantly less code than the ES equivalent

[13:04:00.0000] <Hixie>
and might even require less code than trying to expose the ES function and reference that

[13:04:01.0000] <Hixie>
since there's no guarantee the ES function is actually exposed to other code outside the JS library

[13:04:02.0000] <jgraham>
Well that is true

[13:04:03.0000] <jgraham>
It is also easier to test if the requirements are the same

[13:05:00.0000] <jgraham>
Because you can just reuse the same tests

[13:05:01.0000] <Hixie>
not really

[13:05:02.0000] <Hixie>
how to escape characters is quite different in HTML and JS

[13:06:00.0000] <jgraham>
Well the actual input files will be different. But the range of cases they cover will be identical rather than subtly different

[13:06:01.0000] <AryehGregor>
It's confusing for authors.

[13:06:02.0000] <Hixie>
the range of cases they cover can still be identical, it's just that the tests for HTML wouldn't change over time, whereas in ES' case the tests need to change every time Unicode is updated

[13:07:00.0000] <AryehGregor>
I've seen a case where an author assumed parseInt() would work to translate content attributes to their reflected equivalents.

[13:07:01.0000] <AryehGregor>
I think I linked it on the bug.

[13:07:02.0000] <Hixie>
AryehGregor: that is ludicrous, given the difference is in parts of unicode most authors will never even hear about let alone put in a numeric attribute.

[13:07:03.0000] <Hixie>
parseInt() will work fine on conforming content.

[13:07:04.0000] <AryehGregor>
Yes, the difference is trivial, so it will only fail in bizarre and therefore hard-to-debug corner cases.

[13:07:05.0000] <Hixie>
that's what validators are for.

[13:07:06.0000] <AryehGregor>
On the other hand, since the difference is trivial, there's no reason for it to exist.

[13:07:07.0000] <AryehGregor>
We care about non-conforming content too.  It might be script-generated, too, in which case validators don't help.

[13:08:00.0000] <Hixie>
i've already given half a dozen reasons for the difference to exist, AryehGregor

[13:08:01.0000] <AryehGregor>
I've already said why I disagree with all of them.

[13:08:02.0000] <AryehGregor>
No point in arguing further, either implementers are willing to follow your spec or not.

[13:09:00.0000] <Hixie>
you might disagree with the reasons, but that doesn't mean there are no reasons.

[13:10:00.0000] <AryehGregor>
You gave reasons to avoid the normative dependency on ES, not reasons why the HTML algorithm is better than ES's.  Except for the fact that the ES algorithm changes over time, but you said yourself that the changes are only in cases that most authors will never even hear about.

[13:10:01.0000] <AryehGregor>
I didn't see you give any other reason why the HTML algorithm is better than the ES one, unless I missed something.

[13:11:00.0000] <Hixie>
the only difference between the two is that the ES one changes over time, no?

[13:12:00.0000] <AryehGregor>
The only one I know of, yes.  Or rather, that they use different whitespace definitions.

[13:12:01.0000] <AryehGregor>
But that difference is not important, because the presence of such whitespace is a corner-case and it only matters that behavior be well-defined, not what it actually is.

[13:13:00.0000] <Hixie>
so there's one difference, and it's got a reason (that you think is unimportant).

[13:13:01.0000] <AryehGregor>
Do you think it's important?

[13:13:02.0000] <Hixie>
i think it is important for specs' syntax requirements not to drift over time, yes.

[13:14:00.0000] <AryehGregor>
All specs change over time.

[13:15:00.0000] <Hixie>
and all people eventually die, but that doesn't make death a good thing

[13:15:01.0000] <jwalden>
/me wonders what this hubbub's about, bub

[13:15:02.0000] <Ms2ger>
That's a ludicrous comparison, though

[13:15:03.0000] <cwilso>
when it's a spec's time, it's their time.

[13:15:04.0000] <Ms2ger>
jwalden, integer parsing

[13:16:00.0000] <Hixie>
i'm just saying that the existence of something is not evidence of it not being bad

[13:16:01.0000] <AryehGregor>
Granted, sure.

[13:16:02.0000] <AryehGregor>
I'm not saying there's no disadvantage to matching ES.

[13:16:03.0000] <jwalden>
I see Unicode mentions, and I've been reviewing JS Unicode patchings lately, so it triggered some interest

[13:16:04.0000] <Hixie>
AryehGregor: you did, earlier

[13:16:05.0000] <AryehGregor>
It might well be good if ES changed to match HTML.

[13:16:06.0000] <jwalden>
good luck with that

[13:17:00.0000] <Hixie>
anyway i gotta go. bbiab.

[13:17:01.0000] <AryehGregor>
I didn't mean to imply there were no problems at all with matching ES, I was speaking loosely.  Apologies if it caused any confusion.

[13:22:00.0000] <gsnedders>
Hixie: HTML's syntax requirements will drift over time, though. C.f., self-closing elements.

[13:53:00.0000] <jgraham>
/me remembers why people on the internet suck as a blog post about stress testing memory usage turns into a "my browser is better than yours" pissing competition

[13:53:01.0000] <jgraham>
Well not "why" so much as "that"

[13:59:00.0000] <nlogax>
Hixie: is this correct behavior in chrome? for the "No data" items, only dragstart and end ever fire. for the "Data" items, all the d&d events fire. only difference is the usage of dataTransfer.setData. broke recently-ish in chrome, works in safari (as in all events fire) http://jsfiddle.net/yyb5A/4/

[14:15:00.0000] <AryehGregor>
I count 97 bugs filed in the HTMLWG in the last 24 hours.

[14:15:01.0000] <AryehGregor>
Yikes.

[14:15:02.0000] <Philip`>
/me wonders what happened to the HTML5 Super Friends

[14:16:00.0000] <Philip`>
Did they do anything other than publishing http://www.zeldman.com/superfriends/guide/ (and then not follow up when people tried to get more detail about what kind of "XHTML" syntax checking they wanted)?

[14:19:00.0000] <Ms2ger>
Sounds right

[14:27:00.0000] <jgraham>
AryehGregor: How many from @microsoft.com?

[14:27:01.0000] <AryehGregor>
jgraham, didn't count.

[14:28:00.0000] <MikeSmith>
timeless: here now

[14:28:01.0000] <MikeSmith>
AryehGregor: will change your bugzilla address now

[14:28:02.0000] <AryehGregor>
MikeSmith, thanks!

[14:29:00.0000] <timeless>
MikeSmith: have any other specs been rescinded? or just that one?

[14:29:01.0000] <MikeSmith>
eh?

[14:29:02.0000] <MikeSmith>
timeless: which one?

[14:29:03.0000] <timeless>
http://www.w3.org/TR/2009/PER-xhtml11-20090507/

[14:29:04.0000] <gsnedders>
timeless: You could use SPARQL to find the answer to that question :P

[14:30:00.0000] <MikeSmith>
heh

[14:30:01.0000] <MikeSmith>
where are the super friends when we really need them?

[14:31:00.0000] <timeless>
gsnedders: there's no way i'd find the right sparql service

[14:31:01.0000] <timeless>
and heaven help me find a ua that supports it

[14:33:00.0000] <MikeSmith>
super friends activate! form of a SPARQL query!

[14:33:01.0000] <timeless>
Philip`: i'm amused to see they can't spell `then` correctly and complain about `and`

[14:33:02.0000] <timeless>
> A footer typically contains information about its section such as who wrote it, links to related documents, copyright data, an[d] the like

[14:34:00.0000] <timeless>
> If that is the body than the footer is secondary content to the document as a whole.

[14:34:01.0000] <timeless>
s/than/than [sic]/

[14:35:00.0000] <timeless>
ooh, they want full fledged applications like Bespin to be in the spec!

[14:36:00.0000] <MikeSmith>
timeless: sadly, I did not not even know that that http://www.w3.org/TR/2009/PER-xhtml11-20090507/ had been marked Rescinded

[14:36:01.0000] <MikeSmith>
but I am glad it has been

[14:36:02.0000] <timeless>
MikeSmith: learn something new every day?

[14:36:03.0000] <timeless>
glad to be of help

[14:36:04.0000] <MikeSmith>
yeah

[14:36:05.0000] <MikeSmith>
I guess plh did that

[14:36:06.0000] <timeless>
i don't suppose there's an easy way to find out why it was rescindend

[14:36:07.0000] <timeless>
s/end/ed/

[14:36:08.0000] <Ms2ger>
Michael, there's a Rec at http://www.w3.org/TR/xhtml11/, though

[14:36:09.0000] <MikeSmith>
no, no easy mean

[14:36:10.0000] <Michael>
?

[14:37:00.0000] <Michael>
oh

[14:37:01.0000] <Ms2ger>
"On 19 May 2009 the W3C Director rescinded this document due to process issues."

[14:37:02.0000] <Ms2ger>
MikeSmith, ^

[14:37:03.0000] <Ms2ger>
Sorry Michael

[14:37:04.0000] <MikeSmith>
timeless: the easy mans is to ask Ian Jacobs, probably

[14:38:00.0000] <Ms2ger>
Wasn't that the superman email?

[14:38:01.0000] <Ms2ger>
By Björn

[14:38:02.0000] <MikeSmith>
if that draft has been "rescinded", I would hope that means we can tell the community to ignore XHTML 1.1 completely

[14:39:00.0000] <Ms2ger>
/me finds "Did Lois know that Clark was Superman" problem in www-rdf-interest⊙wo archives

[14:39:01.0000] <MikeSmith>
including the whole misguided mess of XHTML "modularization"

[14:40:00.0000] <MikeSmith>
man, we really needs  for the super friends to get off their asses and actually do something for a change

[14:41:00.0000] <MikeSmith>
I mean, other than issuing proclamations

[14:41:01.0000] <MikeSmith>
well, at least one other proclamation

[14:41:02.0000] <jgraham>
to be honest "super friends" makes then sound like some sort of Japanese escort agency. Which is pretty creepy.

[14:42:00.0000] <Ms2ger>
http://lists.w3.org/Archives/Public/www-archive/2009May/0029.html

[14:42:01.0000] <Ms2ger>
http://lists.w3.org/Archives/Public/www-archive/2009May/0045.html

[14:42:02.0000] <MikeSmith>
based on my personal experience, I kind of like Japanese escorts

[14:42:03.0000] <MikeSmith>
but I really don't like XHTML 1.1

[14:42:04.0000] <Ms2ger>
MikeSmith, thank you for that information

[14:43:00.0000] <MikeSmith>
heh

[14:43:01.0000] <MikeSmith>
well, I need some super friends help on XHTML 1.1

[14:43:02.0000] <Ms2ger>
I'm not sure I'll ever be able to sleep anymore

[14:45:00.0000] <MikeSmith>
super friends should temporarily wake up out of retirement to issue a final proclamation on XHTML 1.1

[14:48:00.0000] <MikeSmith>
along the lines of, "Hey, remember all that stuff we told you about XHTML being the absolute most important thing in the entire universe? well please ignore all that and go back about your business"

[14:49:00.0000] <MikeSmith>
AryehGregor: changed your bugzilla login to ayg⊙an

[14:49:01.0000] <MikeSmith>
please try and let me know if it took

[14:50:00.0000] <MikeSmith>
but as far as I can see, it seems to have

[14:51:00.0000] <annevk>
timeless, that is the PER, not the REC: http://www.w3.org/TR/xhtml11/

[14:51:01.0000] <timeless>
annevk: grumble

[14:52:00.0000] <timeless>
ok, are there *any* Rescinded RECs?

[14:52:01.0000] <MikeSmith>
AryehGregor: because I count 563 bugs where that address appears (including comments)

[14:55:00.0000] <MikeSmith>
I think any time the phrase "family of languages" is used, some sort of automated rescinding mechanism should kick in

[14:55:01.0000] <MikeSmith>
jesus

[14:55:02.0000] <MikeSmith>
"family of current and future document types and modules"

[14:56:00.0000] <MikeSmith>
not just "current", but "future" also

[14:56:01.0000] <MikeSmith>
and not just document types but modules also

[14:57:00.0000] <timeless>
annevk: ok so...

[14:57:01.0000] <timeless>
i was trying to use that document to point out that w3 will rescind things if necessary

[14:57:02.0000] <annevk>
Ms2ger, so since you're not sleeping, fix all bugs

[14:58:00.0000] <timeless>
am i still able to make that claim even if it's only for a weaker PER instead of an actual REC?

[14:58:01.0000] <timeless>
or should i go back to the drawing board, noting that the w3 claims are toothless

[14:58:02.0000] <Ms2ger>
timeless, only if Björn writes a sufficiently sarcastic email

[14:58:03.0000] <timeless>
?

[14:59:00.0000] <annevk>
timeless, rescinding a PER that got later approved with basically the same issues seems pretty toothless to me

[14:59:01.0000] <Ms2ger>
/ignore annevk

[14:59:02.0000] <Ms2ger>
:)

[14:59:03.0000] <timeless>
annevk: grumble

[14:59:04.0000] <timeless>
/me needs a new drawing board

[14:59:05.0000] <annevk>
I do not know any REC

[15:00:00.0000] <annevk>
when I tried to warn implementors about DOM Views being useless Ian Jacobs thought I should use the rescinding process

[15:00:01.0000] <annevk>
I declined

[15:00:02.0000] <timeless>
could we publish something as a REC and go through the complete process of Rescinding it just so we can show it's possible?

[15:00:03.0000] <timeless>
why?

[15:00:04.0000] <timeless>
too much effort?

[15:00:05.0000] <annevk>
right

[15:00:06.0000] <annevk>
/me was about to type exactly that

[15:00:07.0000] <Philip`>
http://www.w3.org/2005/10/Process-20051014/tr#rec-rescind - "W3C MAY rescind an entire Recommendation, for instance when W3C learns of significant errors in the Recommendation, when the Recommendation becomes outdated, ..." - sounds like we should rescind HTML4

[15:01:00.0000] <MikeSmith>
I would really like to rescind XHTML 1.1 completely. it's nominally within my authority to do that now

[15:01:01.0000] <Ms2ger>
Philip`, feel free to lead that

[15:01:02.0000] <timeless>
Philip`: +1

[15:01:03.0000] <annevk>
Philip`, the problem is you have to follow http://www.w3.org/2005/10/Process-20051014/tr#rec-modify

[15:01:04.0000] <Ms2ger>
MikeSmith, want me to loudly proclaim I want you to somewhere? :)

[15:01:05.0000] <MikeSmith>
Ms2ger: tell plh

[15:01:06.0000] <MikeSmith>
XHTML 1.1 is used so rarely on the public web that it's

[15:02:00.0000] <MikeSmith>
XHTML 1.1 is used so rarely on the public web that it's effectively irrelevant anyway

[15:02:01.0000] <timeless>
annevk: do you?

[15:02:02.0000] <annevk>
oh wait you don't

[15:02:03.0000] <timeless>
> To deprecate part of a Recommendation, W3C follows the process for modifying a Recommendation.

[15:02:04.0000] <annevk>
sorry I misread

[15:02:05.0000] <Ms2ger>
Actually, I think I still have an XHTML1.1 site online somewhere

[15:02:06.0000] <timeless>
I think rescinding a *complete* REC skips that

[15:02:07.0000] <timeless>
it's only for partials

[15:02:08.0000] <annevk>
so what is the policy?

[15:02:09.0000] <annevk>
I told Ian Jacobs DOM Views as is is useless

[15:02:10.0000] <annevk>
what's the next step?

[15:02:11.0000] <annevk>
tell someone else within the W3C?

[15:02:12.0000] <MikeSmith>
we need some Jean-Luc Picard "make it happen"

[15:03:00.0000] <timeless>
> 7.7.1 Proposal to Rescind a Recommendation

[15:03:01.0000] <zewt>
itym "make it so"

[15:03:02.0000] <MikeSmith>
annevk: next step is bug me about it

[15:03:03.0000] <MikeSmith>
and then bug me about it again

[15:03:04.0000] <MikeSmith>
until I realize it's serious enough that I need to try to do something about it

[15:03:05.0000] <timeless>
annevk / MikeSmith : would you be ok w/ actually rescinding DOM Views?

[15:04:00.0000] <MikeSmith>
yeah, sure

[15:04:01.0000] <timeless>
i'm willing to help fwiw

[15:04:02.0000] <annevk>
/me follows up with Ian Jacobs

[15:04:03.0000] <timeless>
house cleaning seems like a useful function

[15:04:04.0000] <MikeSmith>
there really is no other policy

[15:04:05.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/0623.html

[15:05:00.0000] <annevk>
thanks for making me look at it again timeless

[15:05:01.0000] <timeless>
glad to

[15:05:02.0000] <timeless>
thanks for being willing to do so

[15:05:03.0000] <MikeSmith>
the W3C Process document actually says very little about publication policy, by design I think

[15:05:04.0000] <annevk>
if this really is the process we should be able to rescind a bunch of stuff

[15:05:05.0000] <annevk>
I wonder if that is appreciated

[15:05:06.0000] <timeless>
i'll certainly appreciate it

[15:05:07.0000] <timeless>
having fewer things that people can reference is a good thing

[15:06:00.0000] <Ms2ger>
MikeSmith, can you change http://www.w3.org/standards/history/domcore to drop the "Web", btw?

[15:06:01.0000] <MikeSmith>
the process for some of this stuff is largely determined by whether you can manage to get IanJ attention about it

[15:06:02.0000] <MikeSmith>
Ms2ger: yeah, thaI can do on my own, I think

[15:07:00.0000] <Ms2ger>
Are you also willing to do it? :)

[15:10:00.0000] <MikeSmith>
Ms2ger: reload

[15:11:00.0000] <Ms2ger>
MikeSmith++

[15:12:00.0000] <Ms2ger>
It still has a Web in the breadcrumbs, though

[15:12:01.0000] <MikeSmith>
hmm

[15:12:02.0000] <MikeSmith>
not sure how to change that but I will try to figure it out

[15:13:00.0000] <Ms2ger>
Great, thanks

[15:15:00.0000] <MikeSmith>
find . | grep -i "Web DOM" in the cvs source for that tree gives me nothing, though

[15:15:01.0000] <MikeSmith>
oops

[15:16:00.0000] <MikeSmith>
needing some xargs

[15:17:00.0000] <MikeSmith>
still nothing, though, even with xargs in their properly

[15:19:00.0000] <Philip`>
That sounds like a convoluted way to do 'grep -ir "Web DOM"'

[15:20:00.0000] <zewt>
it's the unix way, having magic things bolted into particular tools not so much

[15:21:00.0000] <annevk>
maybe it's part of the CMS?

[15:22:00.0000] <timeless>
Philip`: from memory that isn't portable :)

[15:23:00.0000] <timeless>
MikeSmith: note that you need some -print0 and -0s or similar :)

[15:23:01.0000] <MikeSmith>
no idea what that is

[15:24:00.0000] <MikeSmith>
Philip`: what is -r ?

[15:24:01.0000] <TabAtkins>
Recursive, presumably.

[15:25:00.0000] <MikeSmith>
if I'm going to use that I guess I should just go whole hog and use rgrep or whatever abomination GNU has produced

[15:27:00.0000] <timeless>
yeah -r is recursive

[15:27:01.0000] <timeless>
but it isn't portable...

[15:27:02.0000] <timeless>
http://www.linuxquestions.org/questions/solaris-opensolaris-20/recursive-grep-557442/

[15:28:00.0000] <timeless>
> Just pick the gnu grep version which is bundled with Solaris: /usr/sfw/bin/ggrep

[15:28:01.0000] <timeless>
ggrep :)

[15:40:00.0000] <jgraham>
The whole "tools should do one thing and one thing only" idea can be taken too far. Having to learn how to combine large numbers of tools in complex ways is clearly less easy than just using a command line options to a single tool

[15:40:01.0000] <jgraham>
Not least because the documentation of combinations is much worse than the documentation of individual tools

[15:41:00.0000] <Hixie>
clearly?

[15:41:01.0000] <Hixie>
i don't think that's obvious at all

[15:42:00.0000] <Hixie>
heycam: yt?

[15:42:01.0000] <zewt>
i've never heard of "documentation of combinations", i just learned how the tools worked and then it was pretty intuitive how to fit them together

[15:42:02.0000] <heycam>
Hixie, hi

[15:43:00.0000] <Hixie>
heycam: HTMLOptionsElement apparently has an anonymous setter that can be used with any input index

[15:43:01.0000] <Hixie>
heycam: 'setter' only works with known indexes, right?

[15:43:02.0000] <heycam>
Hixie, that's right -- that's the distinction between creator and setter

[15:43:03.0000] <Hixie>
heycam: is there some way to handle this?

[15:44:00.0000] <Hixie>
oh creator

[15:44:01.0000] <Hixie>
duh

[15:44:02.0000] <heycam>
I had wondered whether it is a useful distinction to draw

[15:44:03.0000] <Hixie>
/me is stoopid

[15:44:04.0000] <heycam>
maybe it is

[15:44:05.0000] <heycam>
:)

[15:44:06.0000] <Hixie>
nevermind me

[15:44:07.0000] <heycam>
you can bang "setter creator" on the one operation

[15:44:08.0000] <jgraham>
Well within limits. But it's pretty obvious that if you want to achieve X and you do "man foo" and search for "X" and it says "use the -X switch", that is likly to be easier than imagining a way of piping together foo and bar and baz to achieve the same result. Especially if you are unaware that bar and baz exist.

[15:44:09.0000] <Hixie>
heycam: oh awesome

[15:44:10.0000] <Hixie>
heycam: thanks

[15:44:11.0000] <heycam>
Hixie, np

[15:49:00.0000] <annevk>
hmm https://www.ietf.org/mailman/listinfo/happiana

[15:49:01.0000] <annevk>
anyone here signed up yet?

[15:55:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13640 reference fights

[15:55:01.0000] <annevk>
teehee

[15:55:02.0000] <annevk>
I could learn a thing or two from AryehGregor

[15:57:00.0000] <Hixie>
maybe one day plh will reply to http://lists.w3.org/Archives/Public/public-web-perf/2011Apr/0013.html or http://lists.w3.org/Archives/Public/public-web-perf/2011Jul/0090.html

[15:59:00.0000] <TabAtkins>
Hm.  How do you use innerHTML to replace the contents of an iframe?

[16:02:00.0000] <jgraham>
Which contents?

[16:02:01.0000] <annevk>
http://blog.webmproject.org/2011/08/one-to-one-vp8-video-calling-now.html so that only leaves Apple in a way

[16:02:02.0000] <jgraham>
The child nodes or the framed document?

[16:03:00.0000] <jgraham>
(I would have thought that iframe.innerHTML or iframe.contentDocument.innerHTML would work)

[16:03:01.0000] <jgraham>
(but setting .innerHTML on document might not work everywhere yet)

[16:04:00.0000] <TabAtkins>
The latter doesn't seem to work for me, and I can't get to contentDocument.body

[16:04:01.0000] <jgraham>
contentDocument.documentNode?

[16:04:02.0000] <TabAtkins>
Actually, just calling contentDocument seems to return undefined all the tiem.

[16:04:03.0000] <annevk>
wait for onload

[16:04:04.0000] <jgraham>
Is this with chrome and a data: uri?

[16:05:00.0000] <annevk>
anything but about:blank loads async

[16:05:01.0000] <TabAtkins>
jgraham: Chrome, yes.  data: url, no.  I'm using a blank src.

[16:05:02.0000] <annevk>
iirc

[16:05:03.0000] <jgraham>
The ways of about:blank are mysterious to me

[16:06:00.0000] <jgraham>
Maybe chrome loads that async? But surely taht would break sites...

[16:07:00.0000] <TabAtkins>
annevk: Even creating an iframe within the javascript console doesn't work, though.

[16:07:01.0000] <jgraham>
/me ->sleep++

[16:09:00.0000] <annevk>
wait for onload

[16:09:01.0000] <TabAtkins>
annevk: Again, creating one in the console, then in a separate command asking for contentDocument, return "undefined".

[16:10:00.0000] <TabAtkins>
Given my slow human fingers, I'm quite certain that onload fires between those two actions.

[16:13:00.0000] <annevk>
is it in a document?

[16:13:01.0000] <TabAtkins>
Hm, guess not.

[16:13:02.0000] <annevk>
"When an iframe element is first inserted into a document, the user agent must create a nested browsing context, and then process the iframe attributes for the first time."

[16:13:03.0000] <annevk>
that would explain your problem

[16:14:00.0000] <TabAtkins>
Ah, indeed.  It all works now, thanks!

[16:20:00.0000] <remysharp>
Is this as good a place as any to ask about indexeddb?

[16:20:01.0000] <TabAtkins>
Sure.

[16:21:00.0000] <remysharp>
regarding the transaction method

[16:22:00.0000] <remysharp>
creating new transactions

[16:22:01.0000] <remysharp>
I've seen tutorials offering a third argument - usually 0 (zero) but I can't for the life of me work out what it's supposed to be.

[16:22:02.0000] <remysharp>
any ideas?

[16:22:03.0000] <remysharp>
I did try the spec, but really couldn't work it out from there - not sure if I was being blind or it's just not completely clear

[16:23:00.0000] <remysharp>
for example: db.transaction(['notes'], IDBTransaction.READ_WRITE, 0)

[16:27:00.0000] <TabAtkins>
I can't tell very well, because the spec is almost entirely devoid of examples, but I suspect it's a timeout.

[16:28:00.0000] <remysharp>
yeah - it's not completely clear - glad it's not just me!

[16:28:01.0000] <TabAtkins>
Nah, this spec is *super* horrible for reading.

[16:28:02.0000] <remysharp>
Cheers - searching timeout, I'd say you're right.

[16:29:00.0000] <remysharp>
thanks for that - wasted hours the other night trawling through articles - not sure why I did check in here first!

[16:29:01.0000] <TabAtkins>
^_^

[16:29:02.0000] <annevk>
per spec there's only two arguments

[16:29:03.0000] <annevk>
so that last argument is bogus and will be ignored

[16:29:04.0000] <TabAtkins>
annevk: Where are you finding the description of the transaction() method?

[16:29:05.0000] <annevk>
http://dvcs.w3.org/hg/IndexedDB/raw-file/tip/Overview.html#database-interface

[16:30:00.0000] <remysharp>
cripes - good work finding that

[16:30:01.0000] <TabAtkins>
Ah, there it is.

[16:30:02.0000] <remysharp>
how I didn't see that before....

[16:30:03.0000] <remysharp>
ssheesh

[16:30:04.0000] <jamesr>
Hixie: fwiw on the public-web-perf spec i updated all the specs i'm editor on to reference up-to-date specs

[16:30:05.0000] <Hixie>
nice

[16:30:06.0000] <jamesr>
that does not include page visibility

[16:31:00.0000] <Hixie>
it's really the only way that makes sense, imho

[16:33:00.0000] <annevk>
oh plh replies

[16:33:01.0000] <annevk>
hello W3C log readers

[16:34:00.0000] <Hixie>
he still neatly managed to avoid actually responding to large parts of the e-mail

[16:35:00.0000] <annevk>
http://wiki.csswg.org/spec/reviews/html5


2011-08-04
[17:02:00.0000] <zewt>
<annevk> so that last argument is bogus and will be ignored <- can we apply this to mailing lists

[17:02:01.0000] <Hixie>
it's applied to the whatwg list fwiw

[20:52:00.0000] <Hixie>
anyone got windows handy?

[21:11:00.0000] <roc>
Hixie: yes

[21:12:00.0000] <Hixie>
can you think of a text box somewhere in the native windows UI (not an app and not IE) that is multiline and has some text in it by default?

[21:12:01.0000] <roc>
not off the top of my head

[22:06:00.0000] <hober>
annevk: I'm on the happiana list

[22:08:00.0000] <hober>
annevk: i had tried to subscribe when it was private, and stpeter added me when it opened up

[22:18:00.0000] <MikeSmith>
othermaciej: you around? wanted to ask about moving bugzilla bugs to LC components

[22:20:00.0000] <MikeSmith>
Hixie: I set up LC components for all HTML drafts, and I will need to do a mass-move of them today

[22:20:01.0000] <MikeSmith>
as we did back in October

[22:20:02.0000] <MikeSmith>
wow, it's a blooberry

[22:20:03.0000] <MikeSmith>
blooberry: don't recall seeing you around here for quite a while

[22:38:00.0000] <Hixie>
MikeSmith: k

[22:38:01.0000] <Hixie>
blooberry: sup dude

[23:10:00.0000] <nessy>
MikeSmith: outch! Shame we couldn't avoid that email deluge...

[23:25:00.0000] <MikeSmith>
nessy: yeah, sorry

[23:36:00.0000] <MikeSmith>
ok, mail bomb completed

[23:37:00.0000] <nessy>
uff

[23:37:01.0000] <MikeSmith>
Hixie: I turned your bugmail off about 30 minutes ago, before I launched the mail bomb

[23:37:02.0000] <MikeSmith>
but I just now turned it back on

[23:39:00.0000] <MikeSmith>
othermaciej: all existing bugs are now moved to the LC1 components

[23:39:01.0000] <MikeSmith>
if there are any problems, I'll be around for maybe 2 more hours

[23:40:00.0000] <MikeSmith>
then I have to catch a plane and will be offline for 12 hours or so

[23:54:00.0000] <hsivonen>
the "broadcaster's point of view" of HTML5 is interesting

[23:54:01.0000] <hsivonen>
treating the video element as TV instead of considering the other stuff around it

[23:55:00.0000] <hsivonen>
what if the user doesn't have a <video> playing during a tsunami?

[23:56:00.0000] <hsivonen>
shouldn't the tsunami warning system be a separate daemon that's always in the background?

[00:01:00.0000] <zcorpan>
like Twitter?

[00:06:00.0000] <hsivonen>
zcorpan: I was thinking of something running on the OS level and permission to stick stuff on the screen and to play audio

[00:07:00.0000] <hsivonen>
/me wonders why the CSS WG has reviewed a WD from May when sending feedback in August

[00:08:00.0000] <hsivonen>
"The following references link to Editor's Drafts, not Working Drafts" :-(

[00:11:00.0000] <zcorpan>
HbbTV-- for using a new doctype (and new MIME type) and wanting HTML entities to work

[00:13:00.0000] <zcorpan>
maybe i should instead of trying to fight against it suggest that the HbbTV doctype be added to html5

[00:18:00.0000] <zcorpan>
although other browsers don't support the mime type anyway

[00:32:00.0000] <hsivonen>
zcorpan: what's HbbTV?

[00:34:00.0000] <Hixie>
silly side-effect of the component change is that it de-prioritised everything in the LC components since they've not changed more recently than anything in any other component -_-

[00:36:00.0000] <shetech>
Hey, gang

[00:36:01.0000] <shetech>
question about svg

[00:37:00.0000] <shetech>
hixie: want to take a stab at this? What's the different in and value of embedding svg using <img>, <embed>, <object> or <iframe>? I noticed that the <img> tag doesn't display consistenly across browsers (Chrome sorta barfs on it). Is there a time when one would use one or the other?

[00:38:00.0000] <shetech>
s/different in/difference in/

[00:38:01.0000] <Hixie>
per spec, the only reason you'd not use img is if you want the graphic to be interactive

[00:38:02.0000] <Hixie>
then you'd use object.

[00:39:00.0000] <shetech>
Okay, that's pretty much what I wrote. Chrome's handling of the <img> tag for svg is fugly, though!

[00:40:00.0000] <shetech>
If you're curious, have a look at http://html5.shetech.com/chp_5.html, and look at the "A Circle, using the <img> tag to embed it" under SVG

[00:40:01.0000] <shetech>
FF does it right, Chrome does not.

[00:41:00.0000] <shetech>
Unless I'm doing it wrong?

[00:41:01.0000] <Hixie>
browsers have bugs, news at 11 :-)

[00:41:02.0000] <shetech>
Heh. No doubt.

[00:41:03.0000] <shetech>
Dummies wanna know!

[00:41:04.0000] <shetech>
:D

[00:41:05.0000] <hsivonen>
shetech: img has different security characteristics and doesn't create a browsing context

[00:41:06.0000] <shetech>
splain, please?

[00:41:07.0000] <hsivonen>
shetech: object and iframe have the disadvantage of creating a browsing context

[00:42:00.0000] <hsivonen>
shetech: inlining doesn't create a browsing context

[00:42:01.0000] <shetech>
ah

[00:42:02.0000] <hsivonen>
shetech: I can't remember what <embed> does

[00:42:03.0000] <hsivonen>
shetech: the main annoyance of a browsing context is that non-targeted links cause navigation within the browsing context instead of navigating the parent

[00:42:04.0000] <shetech>
yes

[00:43:00.0000] <shetech>
So for Dummies, maybe I shouldn't mention them as options?

[00:43:01.0000] <shetech>
Or with heavy disclaimers...

[00:44:00.0000] <shetech>
Have I mentioned lately that you guys rock?

[00:44:01.0000] <shetech>
Thanks

[00:47:00.0000] <shetech>
Okay, on that happy note, g'nite, all

[00:47:01.0000] <shetech>
Thanks

[00:55:00.0000] <zcorpan>
hsivonen: "Hybrid Broadcast Broadband TV". a variation of CE-HTML

[00:56:00.0000] <hsivonen>
zcorpan: It's that something that Opera needs to implement in order to please checklist-oriented embedding customers?

[00:56:01.0000] <zcorpan>
yes

[00:57:00.0000] <hsivonen>
my condolences

[01:32:00.0000] <zcorpan>
/me unsubscribes from public-canvas-api

[01:36:00.0000] <hsivonen>
huh? did the same person file a bug requesting content access to accessibility settings and a bug for making it so that content can't even infer the accessibility settigs?

[01:52:00.0000] <hsivonen>
I find it a bit odd that there is any question of "figuring out" if accessibility requirements for MathML and SVG stuff belong in MathML and SVG specs or the HTML spec

[01:55:00.0000] <hsivonen>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13651 embodies the culture differece over what "conforming" means pretty well

[01:56:00.0000] <hsivonen>
I expect that bug to become a total bikeshed about the exact number of words and how to apply the requirement no doubt conceived with English in mind to German, Thai, Chinese, etc.

[01:59:00.0000] <zcorpan>
Documents must not contain more than 1500 words.

[02:02:00.0000] <zcorpan>
it's annoying that whitespace in javascript depends on unicode version

[02:02:01.0000] <zcorpan>
would have been nicer with frozen ascii-only whitespace characters

[02:10:00.0000] <hsivonen>
zcorpan: it seems that TC39 is bad at responding "no way" to pedantic i18n comments

[02:11:00.0000] <zcorpan>
at least ES doesn't do unicode normalization

[02:11:01.0000] <zcorpan>
maybe TC39 only cares when performance is affected

[02:11:02.0000] <hsivonen>
a while ago, I advised another person with generating correct JSON and I was impressed by the correctness of libraries out there

[02:12:00.0000] <hsivonen>
I expected libraries to be wrong when it case to Unicode paragraph breaks but they weren't

[02:12:01.0000] <hsivonen>
s/case/came/

[02:13:00.0000] <hsivonen>
(maybe my expectations about library authors reading specs are too pessimistic)

[02:14:00.0000] <hsivonen>
(or all the libraries that were inspected had gone through the "somebody yells at them" phase already)

[02:15:00.0000] <hsivonen>
oh, and this wasn't just about generating correct JSON but about generating correct JSON that is also correct JS and that doesn't confuse the HTML parser when inlined in <script>

[02:16:00.0000] <hsivonen>
it's also remarkable how the libraries do a better job at dealing with the oddities of ES than the JSON spec

[02:31:00.0000] <annevk>
email overflow

[02:31:01.0000] <annevk>
bah

[02:54:00.0000] <zcorpan>
Hixie: http://html5.org/tools/web-apps-tracker?from=6359&to=6360 - shouldn't it be "If <var title="">n</var> is greater than *zero*" ?

[02:54:01.0000] <annevk>
read-world use cases

[02:54:02.0000] <annevk>
nice

[02:55:00.0000] <hsivonen>
annevk: do you mean the spec itself is a real-world use case for <var>?

[02:57:00.0000] <annevk>
was quoting a typo from Hixie

[02:59:00.0000] <annevk>
zcorpan, no, zero is address later

[03:00:00.0000] <zcorpan>
annevk: in step 4?

[03:02:00.0000] <zcorpan>
let's say you have <select></select> and do select[1] = option;

[03:03:00.0000] <zcorpan>
length is 0, index is 1, n is 1

[03:03:01.0000] <zcorpan>
n is not greater than 1, and it's not zero, so per spec it does nothing

[03:05:00.0000] <zcorpan>
the spec is wrong in another way too

[03:06:00.0000] <zcorpan>
it says to append n new option elements, but it should append n-1 new option elements and then append value

[03:08:00.0000] <annevk>
hmm yeah

[03:09:00.0000] <zcorpan>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1093 - chrome and firefox have empty text nodes in some options

[03:21:00.0000] <hsivonen>
looks like the story today is Gruber's anti-Google bias eclipsing his anti-Microsoft sentiment

[03:23:00.0000] <annevk>
that was somewhat clear earlier on

[03:25:00.0000] <annevk>
http://yaccessibilityblog.com/library/aria-fix-non-standard-images.html :(

[03:30:00.0000] <zcorpan>
ARIA to the rescue!

[03:30:01.0000] <zcorpan>
weird that there's a performance difference

[03:31:00.0000] <annevk>
alright, subscribed to happiana

[03:32:00.0000] <annevk>
thanks hober

[03:46:00.0000] <hsivonen>
http://groups.google.com/group/epub-working-group/browse_thread/thread/ea06deb3500ac246?pli=1

[03:47:00.0000] <hsivonen>
see also http://code.google.com/p/epub-revision/wiki/Telcon20110803 incl. comments below

[04:19:00.0000] <annevk>
CORS is now at http://dvcs.w3.org/hg/cors/raw-file/tip/Overview.html

[04:19:01.0000] <annevk>
yay Mercurial

[04:27:00.0000] <smaug____>
what is "LC1 HTML5 spec" in W3C?

[04:28:00.0000] <smaug____>
I mean in W3C bugzilla

[04:29:00.0000] <annevk>
it means the bug was raised during the Last Call period

[04:29:01.0000] <Philip`>
The first last call period, in particular

[04:29:02.0000] <smaug____>
hmm

[04:30:00.0000] <smaug____>
I haven't cared about last call

[04:30:01.0000] <smaug____>
since I've filed bugs about problems in whatwg html spec

[04:31:00.0000] <smaug____>
well, I guess the lc thing is just some bureaucracy I don't need to care about

[04:32:00.0000] <Philip`>
The spec isn't going to stop improving in response to feedback because of arbitrary deadlines, at least on the WHATWG side

[04:32:01.0000] <asmodai>
sigh: http://blogs.forbes.com/fredcavazza/2011/07/17/why-opposing-html5-and-flash-is-a-non-sense/

[04:32:02.0000] <Philip`>
I guess the main effect it will have is the prioritisation of bugs

[04:32:03.0000] <Philip`>
(since pre-LC bugs should get fixed by some other arbitrary deadline)

[04:32:04.0000] <smaug____>
that is a bit strange

[04:33:00.0000] <annevk>
it's called the W3C

[04:33:01.0000] <smaug____>
:9

[04:33:02.0000] <smaug____>
:)

[04:35:00.0000] <smaug____>
(not sure I really like whatwg process either, but since I don't know how to make things work better, I try to not complain too much :) )

[04:39:00.0000] <zcorpan>
"ok what are you doning" - excellent bug report

[05:03:00.0000] <heycam>
annevk, I wonder if switching those two SVG accessibility bugs from the LC product/component/whatever will mess with the disposition of comments generation when it comes time to do that

[05:03:01.0000] <heycam>
(since they won't be found as easily)

[05:05:00.0000] <annevk>
guess we'll find out

[05:05:01.0000] <heycam>
:)

[05:06:00.0000] <heycam>
/me sleeps

[05:48:00.0000] <zcorpan>
foolip: if you want to hack around with the status boxes, please also fix the mispositioned boxes (they have old IDs e.g. "video" instead of "the-video-element")

[05:48:01.0000] <foolip>
zcorpan, wrong philip?

[05:49:00.0000] <Philip`>
/me certainly doesn't want to do anything about status boxes

[05:51:00.0000] <Workshiva>
Is this old news? http://www.enisa.europa.eu/media/press-releases/web-security-eu-cyber-security-agency-enisa-flags-security-fixes-for-new-web-standards

[05:53:00.0000] <zcorpan>
foolip: no. i saw a bug in which you volunteered to fix status box overlap

[05:53:01.0000] <Philip`>
Workshiva: Yes - http://krijnhoetmer.nl/irc-logs/whatwg/20110802#l-945

[05:54:00.0000] <foolip>
zcorpan, oh, perhaps I did volunteer if someone would just point in the right direction

[05:54:01.0000] <foolip>
can't remember

[05:54:02.0000] <Workshiva>
Weird, it didn't show up in logs search

[05:54:03.0000] <Workshiva>
I guess it isn't indexed yet

[06:02:00.0000] <zcorpan>
foolip: what needs doing is that the IDs need to be updated, and some status boxes need to be removed, but there's no UI to do that (possibly intentionally)

[06:03:00.0000] <annevk>
ms2ger, I would like to move to full name for references

[06:04:00.0000] <annevk>
ms2ger, it's sort of the only place where the first name is abbreviated despite many people going by their first name

[06:04:01.0000] <annevk>
ms2ger, so from now on new entries will have full names and i might slowly migrate older entries

[06:05:00.0000] <annevk>
I really wish JSON had comments

[06:05:01.0000] <zcorpan>
"HTML5: Disabling Click-jacking Protection" is interesting

[06:09:00.0000] <jgraham>
Hmm, thinking "I want a javascript console" and accidentially enabling an ecmascript debugger in a tab with complete.html open seems to be a bad idea

[06:56:00.0000] <annevk>
zcorpan, html5-diff uses Anolis now

[06:57:00.0000] <annevk>
zcorpan, I put the Makefile in CVS, you still need to get https://bitbucket.org/ms2ger/specification-data yourself (and Anolis of course)

[07:01:00.0000] <zcorpan>
annevk: nice!

[07:01:01.0000] <zcorpan>
will look into it another day

[07:12:00.0000] <annevk>
kk

[07:13:00.0000] <zcorpan>
thanks btw

[08:18:00.0000] <annevk>
smaug____, should prolly say that while aDoneCallback is running you cannot invoke ModificationBatch methods

[08:19:00.0000] <smaug____>
annevk: why?

[08:20:00.0000] <annevk>
smaug____, what would happen?

[08:20:01.0000] <smaug____>
the "batch" would listen for some new modifications

[08:21:00.0000] <annevk>
I guess that could work

[08:21:01.0000] <annevk>
also, why not have specialized Modification interfaces?

[08:22:00.0000] <smaug____>
for simplicity

[08:22:01.0000] <annevk>
for text / children / attributes

[08:22:02.0000] <smaug____>
and Modification interface is kind of based on current MutationEvent

[08:23:00.0000] <annevk>
I noticed that

[08:26:00.0000] <smaug____>
annevk: can you see something terribly bad in the approach ?

[08:26:01.0000] <smaug____>
it is quite close to other proposals, of course

[08:29:00.0000] <annevk>
the only thing I think is a little weird is that ModificationBatch is also the argument passed to the callback

[08:29:01.0000] <annevk>
what is the 'this' of the callback?

[08:29:02.0000] <annevk>
and what does .modifications return while not in a callback?

[08:30:00.0000] <smaug____>
I thought I said that .modifications is empty while not in callback

[08:30:01.0000] <smaug____>
empty list

[08:30:02.0000] <annevk>
ah yeah, the comment says so

[08:31:00.0000] <smaug____>
annevk: the callback needs to get some parameter so that { handleBatch: function(b) {}} can be used

[08:31:01.0000] <annevk>
can't you pass "Modification[]" as parameter?

[08:31:02.0000] <annevk>
and not have .modifications

[08:31:03.0000] <annevk>
I think that would make more sense

[08:32:00.0000] <smaug____>
I could, but then one would need to track to which batch the modification is related to

[08:32:01.0000] <smaug____>
I mean if the same callback is used with many ModificationBatches

[08:33:00.0000] <smaug____>
and the callback itself call some of the methods of the ModificationBatch

[08:33:01.0000] <annevk>
is there a use for that?

[08:33:02.0000] <smaug____>
in the callback you may want to call unbatch*()

[08:38:00.0000] <annevk>
what are the ways to make .modifications longer than 1?

[08:38:01.0000] <annevk>
is that stuff like innerHTML, doing things from the callback, inserting DocumentFragment?

[08:41:00.0000] <smaug____>
yeah

[08:43:00.0000] <smaug____>
and probably dom range modifications could utilize that too

[08:43:01.0000] <annevk>
but for setAttribute insertBefore etc. not invoked from the callback it would typically be 1?

[08:43:02.0000] <smaug____>
typically 1

[08:44:00.0000] <annevk>
specifying this would require some very careful algorithms to make it nice

[08:44:01.0000] <annevk>
but seems doable

[08:45:00.0000] <smaug____>
it shouldn't be too hard

[08:46:00.0000] <smaug____>
since, now that I think of it, the batching works quite similarly to gecko's DOMSubtreeModified batching

[08:47:00.0000] <smaug____>
you start a batch and end it, and the outer-most "end" (if there are nested start-ends) fires the callback

[08:47:01.0000] <smaug____>
but anyway, it possible that Google objects the approach

[08:48:00.0000] <smaug____>
and if that is the case, I'm not sure how to proceed with mutation events replacements

[08:50:00.0000] <smaug____>
perhaps someone from Opera could write a new proposal and everyone would be happy with it ;)

[08:54:00.0000] <smaug____>
but let's see what kinds of comments the proposal gets

[08:59:00.0000] <annevk>
fwiw (unrelated to the Modification discussion above) https://plus.google.com/112284435661490019880/posts/eivfrHppgLB

[09:00:00.0000] <annevk>
smaug____, I guess I could write a proposal that splits your Modification interface :p

[09:06:00.0000] <annevk>
smaug____, I guess innerHTML is actually also one modification

[09:06:01.0000] <annevk>
of the childList

[09:14:00.0000] <smaug____>
innerHTML may remove and add several nodes, so it could create several modifications

[09:16:00.0000] <annevk>
oh I see

[09:47:00.0000] <hsivonen>
hmm. so G+ has suspended users from Hong Kong who've used their English name

[09:47:01.0000] <hsivonen>
that sucks

[09:47:02.0000] <TabAtkins>
Man, our "Common Name" policy sucks.  ;_;

[09:48:00.0000] <smaug____>
Does G+ require using a real name?

[09:48:01.0000] <hsivonen>
what does G+ do about e.g. Russian names if a Russian person officially has a Cyrillic name but uses a Romanization in international contexts?

[09:48:02.0000] <TabAtkins>
It requires a "Common Name", but the difference between that and a "Real Name" are nebulous and ill-defined.

[09:48:03.0000] <hsivonen>
smaug____: yes

[09:48:04.0000] <smaug____>
huh

[09:48:05.0000] <TabAtkins>
And I think we're really burning goodwill with this silliness.

[09:49:00.0000] <smaug____>
I think G+ even asked gender when I activated it

[09:49:01.0000] <hsivonen>
I've so far seen a Romaji Japanese name and a Kanji Japanese name

[09:50:00.0000] <smaug____>
but there was some way to not answer to that question

[09:50:01.0000] <hsivonen>
both *so far* unsuspended...

[09:50:02.0000] <TabAtkins>
smaug____: Yeah, you could say "Other", and also I think we hide Gender by default now?

[09:50:03.0000] <zewt>
g+ not supporting gapps sort of makes it hard to take seriously

[09:50:04.0000] <TabAtkins>
I suspect it was just for pronoun use or something.

[09:50:05.0000] <zewt>
as if i'm going to create a separate google account just for one google product

[09:50:06.0000] <dglazkov>
good morning, Whatwg!

[09:51:00.0000] <dglazkov>
it's a lovely component model day for me

[09:51:01.0000] <TabAtkins>
But still, despite the faults, it's working great for me.

[09:51:02.0000] <zewt>
it's like the opposite of single sign-in--multiple signins required for the same site

[09:51:03.0000] <TabAtkins>
Just need a few more people to jump ship from Twitter and I can start winding down over there.

[09:52:00.0000] <dglazkov>
TabAtkins: certainly one thing that makes + more habitable for some folks are extended rage-scussions on how W3C and WHATWG differ. Can't do those in 140 characters.

[09:53:00.0000] <TabAtkins>
dglazkov: Indeed!  And they're actually pretty good discussions!

[09:53:01.0000] <TabAtkins>
Especially anything AryehGregor posts.  +1

[09:53:02.0000] <zewt>
gotta love twitter for training millions of people against writing nontrivial thoughts

[09:54:00.0000] <dglazkov>
zewt: it's more like

[09:54:01.0000] <dglazkov>
"

[09:54:02.0000] <zewt>
"

[09:54:03.0000] <dglazkov>
discovering your natural ability to say nothing in 140 characters"

[09:55:00.0000] <zewt>
it's sort of darkly amusing that people will actually defend it, as if not being able to write complex thoughts is a feature

[09:55:01.0000] <TabAtkins>
It... actually is.

[09:56:00.0000] <zewt>
sure, it's the great equalizer to help people incapable of complex thoughts mask the fact

[09:56:01.0000] <TabAtkins>
I enjoy it as one of my many communications channels.

[09:56:02.0000] <TabAtkins>
zewt: That's indeed one of the benefits.

[09:56:03.0000] <zewt>
that's not a positive. heh

[09:57:00.0000] <hsivonen>
zewt: actually, it takes thought and skill to compress thoughts into tweets

[09:57:01.0000] <TabAtkins>
zewt: Rather than reading too-long blog posts from people who can't think well, you can read pithy and intelligent-sounding tweets!

[09:57:02.0000] <hsivonen>
zewt: it's not at all clear that the compression is equally easy for different people

[09:57:03.0000] <zewt>
hsivonen: or, painfully more frequently, turns the internet into SMS-ese D:

[09:57:04.0000] <TabAtkins>
zewt: You follow the wrong people.

[09:58:00.0000] <zewt>
i don't follow anyone. heh

[09:58:01.0000] <TabAtkins>
Then there you go.

[09:58:02.0000] <TabAtkins>
The good half of Twitter talks in complete sentences and with proper spelling.

[09:58:03.0000] <TabAtkins>
The bad half is full of people I don't care about.

[09:59:00.0000] <dglazkov>
ok, the only thing I need on + is IRC. Then we're set.

[10:00:00.0000] <Philip`>
/me always uses proper capitals and hyphens and semicolons when writing SMS messages, which might be why it usually takes him five minutes to write one

[10:00:01.0000] <TabAtkins>
Philip`: lrn 2 typ fstr

[10:00:02.0000] <zewt>
i'll forgive sms-ese from people on dumbphones, where typing anything at all is pulling teeth ... but most people aren't, anymore, so

[10:00:03.0000] <TabAtkins>
/me had to think for a bit to compress that.

[10:00:04.0000] <dglazkov>
nothing emulates shouting past each other IRL better than IRC

[10:01:00.0000] <zewt>
heh, and it's probably *harder* to type that way on a smartphone, since autocorrect will flip out

[10:01:01.0000] <TabAtkins>
So yeah, zewt, you're making an incorrect generalization.  A significant portion of Twitter (which includes pretty much everyone in tech, as far as I can tell) talks like real people on Twitter.

[10:01:02.0000] <gsnedders>
There's no way I could go back to a phone without a QWERTY keyboard now. So much quicker to type properly on.

[10:01:03.0000] <TabAtkins>
gsnedders: hard or soft keyboard?

[10:02:00.0000] <gsnedders>
TabAtkins: Either.

[10:03:00.0000] <TabAtkins>
kk.  Just wondering if you had a preference.  I used to require hard keyboards due to my gigantic thumbs, until the Nexus S finally had a large enough screen for me to use.

[10:03:01.0000] <TabAtkins>
(My thumbs are the size of a baby's head.)

[10:03:02.0000] <zewt>
i used a G1 initially, and couldn't use the virtual keyboard at all, but eventually switched to nS and crash coursed for a day to learn it

[10:04:00.0000] <Philip`>
1Maybe I wi2l1 write with run-length encoding and s2e1 if the compre2s1ion helps me type faster3.

[10:05:00.0000] <Philip`>
1H2m1, maybe not

[10:05:01.0000] <hsivonen>
Philip`: I WONTFIXed an HTML WG bug about that

[10:05:02.0000] <jgraham>
No doubt G+ will eventually get a "<140 characters" stream. It is obviously trying to merge the feature sets of twitter and facebook

[10:05:03.0000] <TabAtkins>
Hah, the <3br> one?

[10:05:04.0000] <hsivonen>
TabAtkins: yeah

[10:06:00.0000] <zewt>
http://redmine.ruby-lang.org/issues/5054 D:

[10:06:01.0000] <TabAtkins>
wut

[10:06:02.0000] <zewt>
i can't even tell if that's a troll

[10:07:00.0000] <hsivonen>
zewt: Python wins!

[10:07:01.0000] <TabAtkins>
"I'd prefer bad code to look ugly." <-- yes

[10:07:02.0000] <zewt>
well, it's Ruby, that's a given

[10:08:00.0000] <Philip`>
Should have labelled end statements

[10:08:01.0000] <Philip`>
just like labelled break and labelled goto

[10:08:02.0000] <TabAtkins>
Dammit, Philip`, I was *just* typing that.

[10:09:00.0000] <zewt>
<div><div><div><///div>

[10:09:01.0000] <TabAtkins>
 /wrist

[10:09:02.0000] <TabAtkins>
Haha at comment #20

[10:09:03.0000] <Philip`>
TabAtkins: lrn 2 typ fstr

[10:10:00.0000] <TabAtkins>
Philip`: Touche.

[10:11:00.0000] <zewt>
end⑤

[10:12:00.0000] <TabAtkins>
Oh jeez, he was "inspired by Lisp's cdddr".  Those are some of the worst functions in the language.

[10:12:01.0000] <TabAtkins>
Replaced in a ridiculously more readably fashion by NTH and NTH-CDR.

[10:13:00.0000] <Philip`>
/me tries to work out what that's an abbreviation for

[10:13:01.0000] <Philip`>
cod udder?

[10:13:02.0000] <TabAtkins>
"contents of decrement register"

[10:13:03.0000] <Philip`>
Oh, of course

[10:13:04.0000] <TabAtkins>
Originally the address and decrement registers were used to store Lisp conses (2-tuples), thus the names CAR and CDR.

[10:14:00.0000] <TabAtkins>
cdddr is just repeated cdr, equivalent in this case to "throw away the first three elements and gimme what's left"

[10:15:00.0000] <TabAtkins>
And you can mix a and d there.  Most impls support up to three.

[10:15:01.0000] <TabAtkins>
Though I always forget which order they're applied in, which is why I dont' use them.

[10:15:02.0000] <Philip`>
Stick to palindromes so the order doesn't matter

[10:16:00.0000] <TabAtkins>
Sensible idea.

[10:16:01.0000] <Philip`>
though you may have to reorganise your data structures in order to access them in that manner

[10:16:02.0000] <Philip`>
but that's a minor price to pay

[10:16:03.0000] <TabAtkins>
While you're at it, you can haiku-optimize the code too.

[10:22:00.0000] <Philip`>
/me wonders if any languages other than Perl have a poetry chapter in their primary reference book

[10:29:00.0000] <Hixie>
the html spec has poems in it that were specifically written for the spec :-)

[10:29:01.0000] <jgraham>
Hixie: Being like perl is not a positive quality!

[10:30:00.0000] <annevk>
http://blog.chromium.org/2011/08/connecting-web-apps-with-web-intents.html is pretty interesting

[10:30:01.0000] <annevk>
wonder why new bug was filed yet

[10:33:00.0000] <jgraham>
It would be nice if someone at Google or Mozilla would post to the list about that

[10:36:00.0000] <annevk>
TabAtkins, is WebKit going to remove background-position-x/y?

[10:36:01.0000] <TabAtkins>
annevk: I have no idea.  Probably not.

[10:36:02.0000] <annevk>
TabAtkins, so lets add it

[10:36:03.0000] <annevk>
IE has it too

[10:36:04.0000] <annevk>
I think we're planning on adding it some day

[10:36:05.0000] <TabAtkins>
"We" being Opera?

[10:37:00.0000] <annevk>
yeah

[10:37:01.0000] <TabAtkins>
Dammit, people.

[10:37:02.0000] <TabAtkins>
Well, get Elika and Brad to put it into B&B4, then.

[10:37:03.0000] <nimbu>
who has ops in #css?

[10:37:04.0000] <annevk>
I should say damnit to you

[10:37:05.0000] <TabAtkins>
nimbu: I dunno.

[10:38:00.0000] <TabAtkins>
nimbu: Why?

[10:38:01.0000] <nimbu>
some idiot just spammed

[10:38:02.0000] <annevk>
cannot argue against features without removing them first

[10:38:03.0000] <TabAtkins>
nimbu: Oh jeezus.

[10:38:04.0000] <nimbu>
yeah wanna kick/ban but cant without ops :/

[10:42:00.0000] <nimbu>
really nobody has ops :|||

[10:43:00.0000] <TabAtkins>
Could someone go to slide 59 of http://www.slideshare.net/jaffathecake/optimising-where-it-hurts-jake-archibald and tell me which browsers are diverging?  I cant' tell the blue/greens apart.

[10:44:00.0000] <nimbu>
the biggest spike is ff3

[10:44:01.0000] <nimbu>
the one less than that is IE8

[10:44:02.0000] <nimbu>
3rd is saf 3

[10:44:03.0000] <TabAtkins>
Ok, cool.  Thanks!

[10:44:04.0000] <AryehGregor>
MikeSmith, thanks for changing my Bugzilla address (although you're not here and probably won't see this).

[10:44:05.0000] <AryehGregor>
/me wonders why he does that

[10:45:00.0000] <Philip`>
TabAtkins: Optimising where it hurts Jake Archibald sounds like a surprisingly specific optimisation strategy

[10:46:00.0000] <TabAtkins>
And yet, effective!

[10:46:01.0000] <TabAtkins>
It's good to know that scope depth doesn't affect modern browsers.

[10:47:00.0000] <TabAtkins>
Damn you, unlabelled axises!  I have no idea what these numbers are or why I should care.

[10:47:01.0000] <nimbu>
axes

[10:47:02.0000] <nimbu>
FTFY

[10:48:00.0000] <TabAtkins>
NO

[10:48:01.0000] <AryehGregor>
I hate Bugzilla.  Does it really have no way to do mass changes without making me see "Inbox (339)"?

[10:49:00.0000] <AryehGregor>
Oh, and that's only the ones Gmail decided were important.

[10:49:01.0000] <AryehGregor>
in:inbox from:w3.org actually matches 411 conversations.

[10:49:02.0000] <AryehGregor>
Sigh.

[10:49:03.0000] <jgraham>
AryehGregor: Well probably one can access the db directly somehow. I imagine that to be… foolhardy

[10:50:00.0000] <AryehGregor>
I was suggesting maybe there could be a feature that was actually intended to let you do that.

[10:50:01.0000] <AryehGregor>
It seems like a kind of obvious need.

[10:50:02.0000] <AryehGregor>
Like, you know, so people could receive one e-mail for all bugs they're subscribed to?  Would that be so hard?

[10:50:03.0000] <jgraham>
Also, I must remember the 1s of driving === 1 day of charger on standby quote

[10:50:04.0000] <AryehGregor>
. . . what?

[10:50:05.0000] <jgraham>
AryehGregor: See the presentation that TabAtkins couldn't read

[10:51:00.0000] <Philip`>
There should be a "make these changes without emailing people" slider, and each user should have a "actually email me about changes even if someone says not to" slider, and a user only gets emailed if their slider is higher than the changer's slider

[10:51:01.0000] <Philip`>
so people who really want to get emails can opt to do so

[10:51:02.0000] <jgraham>
Or couldn't colour-distinguish, rather

[10:51:03.0000] <AryehGregor>
Philip`, and people should be allowed to choose arbitrary real numbers for the slider, right, without bounds?

[10:51:04.0000] <Philip`>
and people who really really want to get them, even if the changer thinks people who merely really want to get them shouldn't get them, can get them

[10:52:00.0000] <zewt>
what if both people select INF?

[10:52:01.0000] <Philip`>
That's not a real number

[10:54:00.0000] <Philip`>
/me wonders how to implement UI for an unbounded slider

[10:54:01.0000] <AryehGregor>
Philip`, map it bijectively to (0, 1), and have it zoom in as you push the slider toward either edge.

[10:55:00.0000] <AryehGregor>
All we really need is a dense totally ordered set without endpoints here, it doesn't have to be R.

[10:55:01.0000] <zewt>
of course, experience with "importance" flags on email tells us that some people will set all of their changes, no matter how trivial, to the highest value available

[10:55:02.0000] <AryehGregor>
Of course, any two dense totally ordered sets without endpoints are isomorphic if they're of the same cardinality.

[10:55:03.0000] <AryehGregor>
zewt, that's why there is no highest value!

[10:56:00.0000] <AryehGregor>
It will depend on how long people are willing to type digits, or let the slider scroll.

[10:56:01.0000] <zewt>
unless you're willing to send 500-megabyte numbers over the wire, there's always a highest value

[10:56:02.0000] <AryehGregor>
Well, yeah, eventually you're going to hit the server's configured POST size limit.

[10:57:00.0000] <zewt>
and more to the point (as far as this has a point), the fact that some people will always use a really high value regardless of importance would break the system for everyone else

[10:58:00.0000] <zewt>
i suppose that suggests needing a weighting value: if people rate the stuff you're doing as "unimportant" and you're marking it "important", your importance values are reduced

[10:58:01.0000] <Philip`>
I suppose a boringer but maybe more practical solution would be to send a single email on bulk changes, which contains a link you can click that causes the bug tracker to send you an individual email for every bug that was changed, for people who want an email archive of every bug change

[10:59:00.0000] <AryehGregor>
Or how about we just don't care about people who want an e-mail archive of every bug change?

[10:59:01.0000] <AryehGregor>
I mean, why would they?

[10:59:02.0000] <zewt>
ocd

[10:59:03.0000] <Philip`>
The first implementation could make that link point to a 404, then you only bother implementing the multiple-mail feature if enough people complain about it

[10:59:04.0000] <AryehGregor>
Okay, hope I received nothing important from w3.org between 1:00 AM and 1:36 AM, because I'm archiving it all without reading it.

[11:01:00.0000] <Philip`>
/me doesn't know why people would want all that mail but is assuming some people must be objecting to changes that would stop that mail, else surely someone would have implement a bulk-change-with-no-mail feature already

[11:02:00.0000] <AryehGregor>
That seems optimistic.

[11:21:00.0000] <AryehGregor>
Fun fact: when I want to find out what day of the week something is, I now open Google Calendar in Chrome instead of clicking the GNOME clock in the upper-right corner of my desktop, because it's faster to load Google Calendar.

[11:21:01.0000] <AryehGregor>
Seriously.

[11:22:00.0000] <AryehGregor>
If Unity actually takes UI responsiveness seriously, I'm going to switch in 10.10 without looking back.

[11:25:00.0000] <gsnedders>
/me just responded to GNOME3 and Unity by finally moving to a tiling WM

[11:25:01.0000] <AryehGregor>
Heh.

[11:26:00.0000] <AryehGregor>
Unity seems okay so far, although I haven't used it much.

[11:26:01.0000] <AryehGregor>
(I didn't upgrade my desktop to 10.10 because I plan to ditch it soon in favor of my shiny new laptop.)

[11:26:02.0000] <zewt>
"unity" is a good name for a feature when you want it to be confused with 25 other products

[11:26:03.0000] <AryehGregor>
(My desktop has an NVIDIA card anyway, so no shiny stuff for me unless I go the proprietary driver route.)

[11:27:00.0000] <zewt>
(the first thing that "unity" brings to mind for me is VMware's, which is the *worst* thing to bring a comparison to if you're talking about responsiveness)

[11:27:01.0000] <Philip`>
(Why not go the proprietary driver route?)

[11:28:00.0000] <AryehGregor>
Because it caused instability and broke stuff when I last used it.

[11:28:01.0000] <Philip`>
Oh, okay

[11:28:02.0000] <AryehGregor>
It didn't actually crash the system for me anytime after boot completed, to be fair.

[11:28:03.0000] <AryehGregor>
But it did cause a lot of crashes on boot.

[11:29:00.0000] <gsnedders>
/me is wondering whether to care about accelerated 3D on his new system

[11:29:01.0000] <gsnedders>
I mean, I care in Windows for games. But under Linux? Not so much…

[11:29:02.0000] <AryehGregor>
It looks like Linux is moving to actually use accelerated 3D where available.

[11:29:03.0000] <AryehGregor>
For video and such.

[11:29:04.0000] <AryehGregor>
So I'd get it, even if it's just an integrated Intel chip.

[11:30:00.0000] <AryehGregor>
My new laptop has an Intel card that works well enough for stuff like Unity.

[11:30:01.0000] <AryehGregor>
I don't plan to try it with games, though.

[11:31:00.0000] <Philip`>
/me has an oldish laptop with Intel graphics as his main platform for some 3D game development

[11:31:01.0000] <Philip`>
(although admittedly not hugely demanding 3D)

[11:32:00.0000] <Philip`>
(but it's usable, which is good)

[11:32:01.0000] <jgraham>
/me never worked out how tiling wms were supposed to work when you had windows that really wanted to be certain sizes

[11:33:00.0000] <Philip`>
((although sometimes buggy OpenGL code can cause the drivers to get confused so the scrollbars on all my other applications change colour and then everything else gradually change colour until I reboot))

[11:33:01.0000] <jgraham>
(after tring Xmonad for a bit some years ago)

[11:33:02.0000] <gsnedders>
jgraham: Depends if they're floating windows or not

[11:33:03.0000] <Philip`>
((which is unideal)

[11:33:04.0000] <AryehGregor>
Argh, Gmail doesn't believe me when I want to make a filter for "[editing]".  It thinks I mean "editing".

[11:33:05.0000] <gsnedders>
jgraham: Try awesome, it's pretty good out of the box.

[11:33:06.0000] <gsnedders>
Philip`: hah, awesome

[11:33:07.0000] <AryehGregor>
Philip`, that sounds like Linux to me!

[11:33:08.0000] <zewt>
google is starting to ignore +keywords in search :(

[11:33:09.0000] <zewt>
i've always used a lot of +to force it to stop fuzzing searches to death, and now that's starting to not work

[11:34:00.0000] <jgraham>
gsnedders: I doubt I would like it more than compiz grid which basically allows me to tile in the simplistic way that my brain can handle and just dump things in random places when tiling is inconvenient

[11:34:01.0000] <zewt>
building python from source like it's 1996

[11:35:00.0000] <jgraham>
zewt: Can't imagine why that was never a hit single

[11:35:01.0000] <Philip`>
I happened to be searching for some three-character postcode today, and Google was occasionally (depending on search terms) quite insistent that I wanted to find the sixth Harry Potter book

[11:36:00.0000] <zewt>
google needs a "knows how to search" slider, +so +i +don't +have +to +search +like +this

[11:36:01.0000] <jgraham>
Philip`: That is evidence that Google knows best, I would say

[11:37:00.0000] <gsnedders>
jgraham: Well, I have moved to it from grid, which has worked for me.

[11:37:01.0000] <jgraham>
I mean buckinghamshire isn't that exciting really

[11:38:00.0000] <Philip`>
It's got some hills, which are exciting, if you're into hills

[11:40:00.0000] <jgraham>
Those hills even have a wikipedia page: http://en.wikipedia.org/wiki/Category:Hills_of_Buckinghamshire

[11:41:00.0000] <jgraham>
Clearly the are Notable

[11:44:00.0000] <gsnedders>
jgraham: Though each hill in Cambridge has an individual Wikipedia page!

[11:47:00.0000] <jgraham>
Being the only hill in Cambridge is, by definition, quite notable

[12:02:00.0000] <annevk>
"With the host of problems this country is currently facing, the fact that our president is devoting time to the human process of aging is an affront to Americans everywhere," http://www.theonion.com/articles/obama-turns-50-despite-republican-opposition,21061/

[12:03:00.0000] <AryehGregor>
Yeah, that sounds like a typical Onion story.

[12:04:00.0000] <hober>
annevk: you might want to follow along with https://bugs.webkit.org/show_bug.cgi?id=65535 [CSSOM]

[12:06:00.0000] <annevk>
I should really make a proposal

[12:06:01.0000] <annevk>
:(

[12:06:02.0000] <annevk>
That is, make the CSSOM stuff somewhat more concrete for various properties

[12:06:03.0000] <hober>
annevk: please do! :) we're all ears

[12:07:00.0000] <oal>
I'm working on a website where I'm including a website in an iframe, however the iframe tries to forward the user to that website, and navigating away from my site. How'd I handle that?

[12:07:01.0000] <annevk>
I know, I talked to dino the other week; although he was more like, do it now

[12:07:02.0000] <hober>
annevk: hahahhaa, yeah :)

[12:08:00.0000] <AryehGregor>
annevk, PLEASE standardize serialization of CSS values, argh.

[12:08:01.0000] <AryehGregor>
It's been a huge headache in my editing implementation.

[12:08:02.0000] <AryehGregor>
And spec.

[12:08:03.0000] <AryehGregor>
I wish I could just reference some algorithm saying "parse and serialize this CSS value" and be sure that it would come out in some normalized form so I can use string comparison.

[12:09:00.0000] <AryehGregor>
oal, iframes are allowed to navigate the parent page, so in principle there's not much you can do.  Sites are allowed to stop themselves from being framed if they don't want to be.

[12:10:00.0000] <annevk>
serialization is somewhat done

[12:10:01.0000] <annevk>
but is going to be moved from CSSOM to individual CSS modules

[12:11:00.0000] <oal>
AryehGregor: Oh, ok. Thanks :(

[12:11:01.0000] <AryehGregor>
oal, <iframe sandbox> might theoretically give you this ability in some of the newest browsers, but I don't know if it would do exactly what you want.

[12:13:00.0000] <oal>
AryehGregor: Thanks, will try

[12:27:00.0000] <Ms2ger>
Boo, MS

[12:30:00.0000] <gsnedders>
paul_irish: I was guessing that because he already has those patches in the pull queue somebody else had to do something :)

[12:35:00.0000] <annevk>
Ms2ger, oh fun

[12:35:01.0000] <Ms2ger>
They always are

[12:35:02.0000] <Ms2ger>
And plh

[12:37:00.0000] <AryehGregor>
"Boo, MS" why?

[12:38:00.0000] <paul_irish>
gsnedders: yeah i know. :) no worries. rick will stay on 'em for the release cycle

[12:44:00.0000] <Ms2ger>
W3C politics

[12:44:01.0000] <AryehGregor>
Oh, DOM Core?

[12:44:02.0000] <AryehGregor>
To be honest I also think that it's confusing to name specs things like "HTML", "DOM Core", or "DOM Range".

[12:44:03.0000] <AryehGregor>
In the long term it makes sense given lack of versioning, but it's confusing for the interim period.

[12:45:00.0000] <AryehGregor>
Also, the W3C likes to version stuff, so you need a version number at least when you make it into an obsolete snapshot that everyone will ignore.

[12:45:01.0000] <AryehGregor>
The complaint about scope is unreasonable, of course.

[12:45:02.0000] <AryehGregor>
We'll see how it goes.

[12:46:00.0000] <Ms2ger>
Want to reply? I'm afraid I'd offend people if I ded :)

[12:46:01.0000] <Ms2ger>
did*

[12:46:02.0000] <AryehGregor>
Aren't we all?

[12:46:03.0000] <annevk>
I replied

[12:47:00.0000] <AryehGregor>
I was going to leave it to -- yeah, Anne.

[12:47:01.0000] <AryehGregor>
/me doesn't see a reply

[12:47:02.0000] <annevk>
wait for it

[12:47:03.0000] <annevk>
I might be okay with adding a 4

[12:47:04.0000] <AryehGregor>
It only makes sense to add a 4 for the W3C draft.

[12:47:05.0000] <Ms2ger>
So, if "DOM Core" is too confusing for Microsoftees...

[12:48:00.0000] <Ms2ger>
How about "DOM"?

[12:48:01.0000] <AryehGregor>
Googling "DOM Core" still produces useless stuff as the top results.

[12:48:02.0000] <annevk>
yeah, it should be DOM at some point

[12:48:03.0000] <AryehGregor>
"Web DOM Core" used to work fine.

[12:48:04.0000] <AryehGregor>
"DOM" is a bad name, as is "HTML".

[12:48:05.0000] <AryehGregor>
Too short and non-distinctive.

[12:48:06.0000] <AryehGregor>
Maybe if you called it "DOM Standard", "HTML Standard", something like that.

[12:49:00.0000] <annevk>
sure

[12:49:01.0000] <Ms2ger>
Technically, sure

[12:49:02.0000] <Ms2ger>
Politically? I dunno

[12:49:03.0000] <AryehGregor>
I mean from, say, the WHATWG's perspective.

[12:49:04.0000] <AryehGregor>
For the W3C, just call it DOM 4 Core.

[12:50:00.0000] <Ms2ger>
"Document Object Model Level Four Core"

[12:50:01.0000] <Ms2ger>
(tm)(c)

[12:50:02.0000] <annevk>
o_O

[12:50:03.0000] <Ms2ger>
Or Core-XML?

[12:53:00.0000] <jamesr>
if you have to add a number make it 5

[12:54:00.0000] <jamesr>
5 stuff is way cooler than 4 these days

[12:54:01.0000] <annevk>
CSS3, DOM4, HTML5

[12:54:02.0000] <annevk>
makes perfect sense

[12:54:03.0000] <jamesr>
SVG2

[12:54:04.0000] <annevk>
there you go

[12:54:05.0000] <jamesr>
MathML6?

[12:54:06.0000] <Ms2ger>
XHTL1

[12:54:07.0000] <annevk>
XML1

[12:54:08.0000] <jamesr>
or MathML1

[12:54:09.0000] <annevk>
teehee

[12:54:10.0000] <Ms2ger>
/me kicks his "M" key

[12:55:00.0000] <jamesr>
XML1 SVG2 CSS3 DOM4 HTML5, maybe we can get ECMA to make ECMA-262 6th ed

[12:55:01.0000] <scor>
can someone explain to me how I can force HTML snippets inside microdata values? it seems all microdata values can only be plain text

[12:56:00.0000] <annevk>
you cannot have HTML there

[12:56:01.0000] <scor>
s/can/will

[12:56:02.0000] <scor>
the HTML is stripped out

[12:57:00.0000] <scor>
annevk: any idea why such restriction?

[12:57:01.0000] <scor>
annevk: there can be HTML in the page, but the microdata parsing will strip it out (that's how I understand it)

[12:58:00.0000] <jgraham>
Well I guess you can if the vocabulary allows it. I mean it will be plain text but you can always specify that the text should be fed into an HTML parser

[12:58:01.0000] <annevk>
no compelling use case I think scor

[12:58:02.0000] <jgraham>
Just generic tools won't do that

[12:58:03.0000] <jgraham>
HTML in microdata seems like a really bad idea

[12:59:00.0000] <scor>
jgraham: exactly, that's extra processing not specified in the microdata spec

[12:59:01.0000] <scor>
hum, why is it a bad idea jgraham

[12:59:02.0000] <scor>
?

[12:59:03.0000] <jgraham>
It only allows a very limited range of things, but hugely increases complexity

[12:59:04.0000] <scor>
annevk: what do you think of these use cases http://www.w3.org/Bugs/Public/show_bug.cgi?id=13468#c7

[13:00:00.0000] <scor>
jgraham: complexity in the md parsing algo? or?

[13:00:01.0000] <annevk>
scor, seems you're already talking to the right people :)

[13:00:02.0000] <jgraham>
Complexity in the data model, yeah

[13:01:00.0000] <annevk>
/me isn't really sure annotated data is going to work out

[13:01:01.0000] <jgraham>
Having people implement HTML parsers just to understand microdata is a burden

[13:01:02.0000] <scor>
annevk: I only wish there were more than one person I was talking too

[13:01:03.0000] <jgraham>
Having to design APIs that can retrun text-or-markup is hard

[13:01:04.0000] <annevk>
jgraham, you need an HTML parser to get Microdata

[13:01:05.0000] <annevk>
jgraham, and APIs can already change from string to object for other reasons

[13:01:06.0000] <jgraham>
annevk: Not if you consume e.g. JSON

[13:02:00.0000] <scor>
foolip: you were also involved in that thread re HTML snippets in microdata

[13:02:01.0000] <annevk>
jgraham, oh right, but you could expose the markup in an objecty way

[13:02:02.0000] <jgraham>
You infect everyone with the need to process HTML, not just the edges of the system

[13:03:00.0000] <scor>
jgraham: but you also prevent those who need that HTML from using it

[13:03:01.0000] <jgraham>
I'm also not sure that being able to say H<sub>2</sub>O is really that useful

[13:03:02.0000] <scor>
jgraham: H<sub>2</sub>O and H2O are completely different things

[13:04:00.0000] <jgraham>
What if you want to express the chemical structure rather than just the elements?

[13:04:01.0000] <jgraham>
HTML isn't any use as a graph language

[13:05:00.0000] <jgraham>
So "html in microdata" quickly turns into "arbitary markup in microdata"

[13:05:01.0000] <Ms2ger>
There's a Unicode character for <sub>2</sub>, so that's fine

[13:05:02.0000] <scor>
Ms2ger: how do I find it out?

[13:05:03.0000] <annevk>
jgraham, true true, strings just want to become markup typically :)

[13:05:04.0000] <jgraham>
But at that point I think you might as well just make it opaque to generic processors and allow specific vocabs to define processing

[13:06:00.0000] <Ms2ger>
www.google.com/search?q=subscript+2+unicode

[13:06:01.0000] <Ms2ger>
₂

[13:06:02.0000] <jgraham>
I mean it's not like a *generic* processor can do much that is useful with H<sub>2</sub>O as nodes compared to as a literal string

[13:07:00.0000] <jgraham>
It only really helps when the use caseis "format for display"

[13:07:01.0000] <jgraham>
(so there we go, I just conclusively proved that HTML is a presentational langauge :) )

[13:08:00.0000] <Ms2ger>
/me thwaps jgraham 

[13:08:01.0000] <Ms2ger>
Lang-gauge?

[13:08:02.0000] <scor>
Ms2ger: ok. so I suppose I would have to write this HTML snippet using unicode instead of HTML then, but how user friendly is that? (in comparison to the HTML version)

[13:09:00.0000] <jgraham>
Ms2ger: See the google webstats. That's like the most misspelt word in HTML :)

[13:09:01.0000] <jgraham>
scor: Or make up some vocabulary-specific processing

[13:11:00.0000] <jgraham>
I just can't imagine many UAs that display raw microdata as strings to end users, so the fact that H<sub>2</sub>O doesn't make much sense as a string doesn't matter

[13:12:00.0000] <jgraham>
The other problem with generic HTML processing is that H<video src="myvideo"><font size=2>2</font>O isn't much use as a chemical formula

[13:12:01.0000] <jgraham>
so you still need vocab-specific rules

[13:12:02.0000] <Ms2ger>
I hear that kind of formula is used in astrophysics, though

[13:13:00.0000] <jgraham>
Are you insane? It's way too precise.

[13:13:01.0000] <gkellogg>
I don't see why rules for expressing markup should be vocab-specific. As much as I hate @itemvaltype, this is a case where it seems necessary.

[13:13:02.0000] <scor>
jgraham: so how would these vocab-specific rules work wrt to the md parsing? when would these rules be applied?

[13:14:00.0000] <jgraham>
scor: In generic UAs everything would be parsed as strings. The angle brackets would just be angle brackets, not anything magic

[13:14:01.0000] <Ms2ger>
Good point

[13:15:00.0000] <jgraham>
Although astrophysiscis do love a good <video> to go with a theory

[13:15:01.0000] <jgraham>
Or at least an <img> containing an "artist's impression"

[13:15:02.0000] <Ms2ger>
"So all the planets in the solar system kinda follow a pattern, except that we lost one planet"

[13:17:00.0000] <Ms2ger>
/me hopes he summarized the Titius–Bode law correctly

[13:18:00.0000] <annevk>
/me suggests DOM4

[13:22:00.0000] <jgraham>
scor: The other reason to dislike markup in microdata is that it gives the whole feature way more surface area for security problems e.g. if someone sets the value of an item to be <script src=evilscript.js></script> you want to be rather sure that you don't execute it when you don't mean to

[13:22:01.0000] <scor>
jgraham: absolutely, the consumer has to strip out insecure tags

[13:23:00.0000] <Hixie>
i hope you mean strip out anything not known to be secure

[13:23:01.0000] <scor>
I'm not saying HTML snippet should be the default, I'm just wondering why there is not option for allowing it

[13:23:02.0000] <Hixie>
rather than stripping out anything known to be insecure

[13:24:00.0000] <scor>
yes

[13:24:01.0000] <Hixie>
just checking :-)

[13:24:02.0000] <scor>
:)

[13:25:00.0000] <jgraham>
scor: Which is still more risky than just not allowing markup, and doesn't address the use cases where "risky" things make sense as item values

[13:27:00.0000] <scor>
jgraham: not allowing markup makes sense as a default

[13:29:00.0000] <Hixie>
scor: the main reason microdata doesn't have a way to use the DOM Elements as part of the value (as opposed to just the text) is that none of hte use cases it was designed for needed it

[13:29:01.0000] <Hixie>
(though there are certainly additional reasons to avoid it as jgraham has pointed out)

[13:30:00.0000] <scor>
Hixie: when was the list of use cases for designing microdata closed?

[13:30:01.0000] <Hixie>
it wasn't closed

[13:30:02.0000] <Hixie>
if there are new use cases to consider, the design could be augmented accordingly

[13:31:00.0000] <Hixie>
it depends on whether the use cases are deemed important enough

[13:31:01.0000] <scor>
but I guess the use cases provided in the tracker are not important enough

[13:31:02.0000] <dabaR>
can a header contain sections?

[13:32:00.0000] <Hixie>
scor: off-hand, i've no idea

[13:32:01.0000] <scor>
Hixie: so what do you require then? more people asking for these use cases?

[13:32:02.0000] <Hixie>
i don't think i've ever seen a use case for which it makes sense for the element tree to be part of the value, so far

[13:33:00.0000] <Hixie>
but it's quite possible that some have been sent to the whatwg list or on bugzilla that i haven't seen yet

[13:33:01.0000] <Hixie>
the volume of feedback is not important, only the quality matters

[13:33:02.0000] <Hixie>
dabaR: do you mean a <header> or a <h1> header?

[13:34:00.0000] <Hixie>
dabaR: if you mean <header>, then yes. Search the spec for "Little Green Guys With Guns" for an example where a <header> contains a <nav> section.

[13:35:00.0000] <dabaR>
thank you

[13:35:01.0000] <Hixie>
np

[13:37:00.0000] <dabaR>
On here: http://www.w3.org/TR/html5/ ?

[13:37:01.0000] <Hixie>
http://whatwg.org/html

[13:38:00.0000] <dabaR>
Thanks

[13:38:01.0000] <dabaR>
Nope

[13:38:02.0000] <dabaR>
No little green guys with guns

[13:38:03.0000] <dabaR>
In fact, none of those words appear :)

[13:39:00.0000] <Hixie>
oh that's the multipage copy

[13:39:01.0000] <Hixie>
switch to the one-page copy

[13:39:02.0000] <annevk>
if you load whatwg.org/c they will, but beware of crashing your browser

[13:39:03.0000] <Hixie>
there's a link at the top

[13:39:04.0000] <dabaR>
Odd it is there

[13:40:00.0000] <dabaR>
Oh thats why

[13:40:01.0000] <dabaR>
Well that's nav inside header

[13:40:02.0000] <dabaR>
But also section is OK?

[13:41:00.0000] <Hixie>
<nav>, <Section>, <Article>, and <aside> are all sections

[13:42:00.0000] <dabaR>
Oh OK

[13:42:01.0000] <Hixie>
(<section> is just the more generic one)

[13:42:02.0000] <dabaR>
so <section> can go inside <header>

[13:43:00.0000] <dabaR>
You know how on government sites there is often a more web-presence-wide header and footer?

[13:43:01.0000] <dabaR>
Containing links to other sites they have, and also a search bar for all of their sites kinda thingy?

[13:44:00.0000] <dabaR>
I guess that is a <nav>. Is that a good guess?

[13:44:01.0000] <Hixie>
yeah, that's basically that <nav> example

[13:44:02.0000] <dabaR>
OK. Thank you.

[13:46:00.0000] <dabaR>
But a nav gets its own heading?

[13:46:01.0000] <dabaR>
Not in this example

[13:46:02.0000] <dabaR>
A <nav> can have an <h*>

[13:50:00.0000] <karlcow>
/me read the discussion with scor and had difficulties to figure out the benefit of markup.

[13:51:00.0000] <karlcow>
it seems like alt="" and title="" attributes

[13:53:00.0000] <scor>
karlcow: ok, let microdata have the same limitations as @alt and @title then

[13:54:00.0000] <scor>
karlcow: did you check the use cases in the tracker too?

[13:54:01.0000] <karlcow>
scor: link?

[13:54:02.0000] <scor>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13468#c7

[13:54:03.0000] <karlcow>
ah yes what I was talking about

[13:56:00.0000] <karlcow>
years after years of using html, I have grown a feeling that attributes were not the right place for rich markup.

[13:56:01.0000] <karlcow>
attributes most of the type should be more "operational" than "informational".

[13:57:00.0000] <karlcow>
(just a feeling)

[13:57:01.0000] <karlcow>
scor in the bug tracker could you put a full example of code?

[13:58:00.0000] <jgraham>
karlcow: One could of course design it to not have markup in attributes if one wanted that extension to the data model

[13:58:01.0000] <jgraham>
But I think the concept is bad even then

[13:59:00.0000] <karlcow>
/me wonders if jgraham meant "to have markup"

[14:00:00.0000] <karlcow>
I remember all the discussion around description element in RSS feed.

[14:00:01.0000] <scor>
karlcow: you mean a full microdata example?

[14:01:00.0000] <scor>
which I would expect to produce HTML?

[14:02:00.0000] <jgraham>
karlcow: I mean you could define a mechanism where, say, the child nodes of the element with @itemval represented the content

[14:04:00.0000] <jgraham>
But even in the presence of such a mechanism, allowing markup in values would be more harm than good

[14:04:01.0000] <karlcow>
scor: yes

[14:05:00.0000] <karlcow>
jgraham: without counting for what a microdata means inside a microdata

[14:06:00.0000] <jgraham>
That is an example of the additional complexity that allowing markup in values introduces, yes

[14:21:00.0000] <scor>
karlcow: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13468#c12

[14:24:00.0000] <scor>
karlcow: what was the conclusion re. "description element in RSS feed"

[14:25:00.0000] <karlcow>
scor: for RSS, dirty hacks, including escaping markup.

[14:25:01.0000] <karlcow>
Atom introduced a content type

[14:26:00.0000] <karlcow>
<content type="xhtml">…

[14:30:00.0000] <Ms2ger>
The bugs line at http://www.whatwg.org/issues/data.html?period=1 does look interesting

[15:37:00.0000] <jarek>
Hello

[15:37:01.0000] <jarek>
is there any performance difference between element.setAttribute('blah1', 'blah2') and element.dataset.blah1 = 'blah2'?

[15:39:00.0000] <jarek>
I'm currently using custom attributes all over the place and I'm wondering whether switching to HTML5 datasets would make sense

[15:39:01.0000] <Philip`>
Seems very unlikely that there would be any noticeable difference

[15:39:02.0000] <Philip`>
since dataset is just an API to get/set content attributes

[15:40:00.0000] <Philip`>
so it's doing the same amount of DOM manipulation either way

[15:43:00.0000] <jarek>
perhaps setting an attribute directly on node (e.g. element.blah1 = 'blah2')  could be any faster?

[15:46:00.0000] <Hixie>
depends what you're trying to do

[15:46:01.0000] <The_8472>
jarek, you shouldn't set custom attributes directly on dom nodes anyway. it's bad practice for various reasons.

[15:46:02.0000] <Hixie>
are you doing this in a tight loop?

[15:46:03.0000] <Hixie>
if you're not, the performance aspects of this are probably irrelevant anyway

[15:47:00.0000] <AryehGregor>
The_8472, why not?  That's what data-* is meant for.

[15:48:00.0000] <The_8472>
AryehGregor, i mean this: <jarek> perhaps setting an attribute directly on node (e.g. element.blah1 = 'blah2')

[15:48:01.0000] <jarek>
The_8472: I'm using custom XML namespace inside XHTML5 document, so custom attributes should be fully valid in my case

[15:48:02.0000] <The_8472>
attributes may have been the wrong word

[15:48:03.0000] <The_8472>
object properties?

[15:50:00.0000] <Hixie>
the spec uses the terms IDL attributes and content attributes

[15:51:00.0000] <Hixie>
though in your case you're talking abotu object properties, since they're not IDL attributes either

[15:51:01.0000] <Hixie>
they might conflict with IDL attributes in the future though, so it's not good practice :-)

[15:51:02.0000] <jarek_>
oops, I got disconnected

[15:52:00.0000] <jarek_>
yeah, I already learned that extending core objects is considered to be a bad practice: http://perfectionkills.com/whats-wrong-with-extending-the-dom/

[15:52:01.0000] <jarek_>
but I want to do this anyway :)

[15:52:02.0000] <The_8472>
good good

[15:53:00.0000] <jarek_>
btw, why it's possible to use element.dataset on non-html elements?

[15:54:00.0000] <jarek_>
while it's not possible to use e.g. element.style

[15:54:01.0000] <Hixie>
i thought it was the other way around

[15:55:00.0000] <jarek_>
let me double check...

[15:55:01.0000] <jarek_>
yum,

[15:55:02.0000] <jarek_>
element.style always returns null if element is from custom XML namespace

[15:56:00.0000] <jarek_>
s/yum/yup

[15:56:01.0000] <Hixie>
jarek_: and dataset works?

[15:56:02.0000] <jarek_>
element.dataset works fine no matter whether element is from custom namespace or not

[15:56:03.0000] <Hixie>
weird

[15:57:00.0000] <Hixie>
sounds like a bug

[15:57:01.0000] <jarek_>
no, it's a feature

[15:57:02.0000] <jarek_>
please don't fix it

[15:57:03.0000] <The_8472>
it's a bug

[15:57:04.0000] <Hixie>
what if another namespace defines attributes that start with data-?

[15:57:05.0000] <heycam>
move it to Element and into DOM Core

[15:58:00.0000] <heycam>
if we're going to have global id and class attributes, we should do the same for data-

[15:58:01.0000] <Hixie>
i still don't think we should have global id and class :-)

[15:58:02.0000] <The_8472>
id is namespaced in xml documents, isn't it?

[16:00:00.0000] <heycam>
Hixie, if dataset doesn't move to Element, would you be able to stick it in a separate interface? beacuse I'm pretty sure it would be useful to use on SVG elements too, and in that case we'd want to have `SVGElement implements DataSetThingos`

[16:00:01.0000] <Hixie>
i doubt i'll win the id/class battle

[16:00:02.0000] <Hixie>
in which case i'll let anne take data-* to dom core

[16:01:00.0000] <heycam>
ok

[16:01:01.0000] <Hixie>
i'm not sure how anne is planning on defining content attributes though

[16:01:02.0000] <heycam>
/me ducks out for a bit

[16:03:00.0000] <AryehGregor>
Man, Scalia writes awesome opinions.  http://www.supremecourt.gov/opinions/10pdf/08-1448.pdf

[16:04:00.0000] <AryehGregor>
Cogent and extremely sarcastic.

[16:04:01.0000] <AryehGregor>
"One study, for example, found that children who had just finished

[16:04:02.0000] <AryehGregor>
playing violent video games were more likely to fill in the blank letter in “explo_e” with a “d” (so that it  reads “explode”) than with an “r” (“explore”).  App. 496, 506  (internal quotation marks omitted). The prevention of this phenomenon,  which  might have been anticipated with common sense, is not a compelling state interest."

[16:04:03.0000] <Hixie>
hah

[16:07:00.0000] <Hixie>
i also like "Since California has declined to restrict those

[16:07:01.0000] <Hixie>
other media, e.g., Saturday morning cartoons, its video-game regulation is wildly  underinclusive, raising serious doubts about whether

[16:07:02.0000] <Hixie>
the State is pursuing the interest it invokes or is instead disfavoring

[16:07:03.0000] <Hixie>
a particular speaker or viewpoint"

[16:07:04.0000] <Hixie>
man this is full of win

[16:08:00.0000] <AryehGregor>
Scalia's decisions are always like that.

[16:08:01.0000] <AryehGregor>
Roberts is good too.

[16:10:00.0000] <Hixie>
" There is no contention that any of the

[16:10:01.0000] <Hixie>
virtual characters depicted in the imaginative videos at issue here are

[16:10:02.0000] <Hixie>
criminally liabl"

[16:10:03.0000] <Hixie>
that's good to know

[16:13:00.0000] <Hixie>
lol they cite grimm

[16:14:00.0000] <Hixie>
by page number no less

[16:15:00.0000] <The_8472>
writing their opinions probably provides them some fun after dealing with aneurysm-inducing lawyer reasoning.

[16:15:01.0000] <AryehGregor>
I think Alito's concurrence makes sense too.

[16:15:02.0000] <AryehGregor>
But then, I think court interpretations of the First Amendment can be kind of extreme sometimes.

[16:15:03.0000] <AryehGregor>
/me gets to the dissent

[16:16:00.0000] <The_8472>
/me points at corporations having 1st amendment rights in the US.

[16:17:00.0000] <Hixie>
page 11 has a sentence that just says "Who knows?"

[16:17:01.0000] <AryehGregor>
Why shouldn't they have them?  They're basically just collections of people.  Your right to free speech shouldn't be affected by who you're being paid by.

[16:18:00.0000] <Hixie>
AryehGregor: if we could imprison the collection of people, that might have more weight

[16:18:01.0000] <AryehGregor>
You can imprison individual people who individually did things wrong.  Corporations don't have volition, they only have assets.  They're controlled entirely by individual people.

[16:18:02.0000] <The_8472>
http://www.nytimes.com/2010/01/22/us/politics/22scotus.html

[16:18:03.0000] <The_8472>
therefore corporations shouldn't have the same rights as people.

[16:19:00.0000] <AryehGregor>
They don't.  The people they employ do.

[16:19:01.0000] <The_8472>
see... that's how it should be, not how it is.

[16:19:02.0000] <AryehGregor>
Or rather, the rights that have are corollaries of the rights of their employees.

[16:20:00.0000] <AryehGregor>
People who run ads or lobby Congress on behalf of a corporation aren't any less entitled to do so than people who do so on behalf of individuals or themselves.

[16:20:01.0000] <AryehGregor>
Just because you were told to do it by a manager in exchange for a paycheck from a corporate bank account doesn't mean you have any less right to do it.

[16:21:00.0000] <AryehGregor>
/me always gets bored of reading court decisions by the time he gets to the dissents

[16:21:01.0000] <Hixie>
i don't mind them having rights if they have responsibilities to go with them

[16:21:02.0000] <The_8472>
just because you work for a company does not mean they are entitled to represent your political opinion

[16:22:00.0000] <AryehGregor>
Of course they aren't.  But they can pay you to exercise your right of political speech on their behalf.  That's all corporations can ever do -- they can't talk or write anything *themselves*.  They're legal fictions.  Any speech is necessarily speech by some person.

[16:22:01.0000] <wilhelm>
The_8472: They usually don't. They represent the political opinion of the owners, if any.

[16:23:00.0000] <AryehGregor>
Hixie, constitutional rights in America are generally viewed as unalienable, not as being granted in exchange for fulfilling responsibilities.

[16:23:01.0000] <AryehGregor>
Convicted murderers spending a life sentence in prison still have the same right to free speech as anyone.

[16:24:00.0000] <Hixie>
AryehGregor: i'm not saying they're directly linked, just that i would be happy with one if the other existed.

[16:25:00.0000] <AryehGregor>
Hixie, well, corporations basically only have assets.  When it comes to the disposition of those assets, they have similar responsibilities to individuals, often much greater responsibilities (depending on the type of corporation).  That is, they can be held fiscally liable for breaking the law.  But if it's anything non-monetary, you can't really take actions against a corporation, you have to take action against the individuals responsible.

[16:25:01.0000] <Hixie>
AryehGregor: e.g. if someone were to kill all the fish in a sea due to reckless negligence while doing something in that sea, they could be expected to suffer consequences, not be immediately invited to continue doing that thing in the same sea and others, as well as spending large sums of money influencing political decisions that control those consequences

[16:26:00.0000] <The_8472>
/me points at the wallstreet. diffusion of responsibility to the max. they rake in millions and billions by gambling with other people's money (and losing) and get away with a slap on the wrist (to individuals) and practically no regulation at all (towards the companies)

[16:26:01.0000] <The_8472>
they cause far more damage than your lowly criminals that may get many-year sentences...

[16:26:02.0000] <Hixie>
AryehGregor: i'd be fine with it if the individuals were held responsible too

[16:27:00.0000] <AryehGregor>
Hixie, that's orthogonal to the question of corporations.  "Corporation" doesn't mean "big business", it means an entity that can legally act as a person in some respects.  Small charities and whatever are also often incorporated.

[16:27:01.0000] <Hixie>
AryehGregor: unfortunately, in the US (and many other places), corporations act as responsibility shields

[16:27:02.0000] <wilhelm>
Indeed. “because the corporation is legally considered the "person," individual shareholders are not legally responsible for the corporation's debts and damages beyond their investment in the corporation”

[16:27:03.0000] <Hixie>
AryehGregor: for both the matter of responsibility and the matter of free speech, the problem is proportionally bigger as the company gets bigger.

[16:28:00.0000] <AryehGregor>
It's not the corporation that's shielding anything, it's the money and influence.  Corporations are designed to be a good vehicle for concentrating arbitrarily large amounts of money, so it just so happens that any organization of note is going to be incorporated.

[16:28:01.0000] <AryehGregor>
That's not the fault of the incorporation.

[16:28:02.0000] <The_8472>
Hixie, it's not just corporations. political parties or bureaucratic machineries are exactly the same. it's all about diffusing responsibility so much that in the end nobody is responsible at all. or that just some symbolic heads will roll even after biggest fuckups that you can imagine

[16:28:03.0000] <AryehGregor>
If incorporation didn't exist, you'd see the same thing with large private businesses.

[16:29:00.0000] <Hixie>
The_8472: yeah, responsibility diffusion really is the problem. the rights get concentrated, but the responsibilities diffused.

[16:29:01.0000] <wilhelm>
The shielding of individual shareholders is a wonderful tool to encourage the establishment of new businesses, and awfully scary when said businesses grow bigger and more powerful than small nation-states.

[16:29:02.0000] <AryehGregor>
It's simpler than that, it's just that large businesses are powerful enough to have bargaining power with the government.

[16:30:00.0000] <AryehGregor>
In many cases they can credibly threaten the government.

[16:30:01.0000] <AryehGregor>
For instance, a big business could threaten to leave a particular state if it's taxed too heavily.

[16:30:02.0000] <wilhelm>
Or threaten its employees, customers or the general populace.

[16:30:03.0000] <AryehGregor>
That could cause thousands of people to lose their jobs, which would create a big backlash by the general public against the politicians in power.

[16:31:00.0000] <The_8472>
shareholders are another issue when they get dividents. they're syphoning off money from other people's work, even when the company has reached the point of being self-sustaining. so instead of accumulating money within the company to reinvest it gets drained, thus potentially slowing innovation just to fill a few private people's pockets.

[16:31:01.0000] <The_8472>
i know that shares are an important tool to get investments... but in some cases it's just counterproductive

[16:31:02.0000] <The_8472>
the value of the shares themselves should be sufficient

[16:31:03.0000] <AryehGregor>
The_8472, well, yes, that's how capitalism works.  In a capitalist society, it's very easy to start businesses, because investors have the promise of exorbitant returns.

[16:32:00.0000] <AryehGregor>
Most large companies don't pay dividends, though.

[16:32:01.0000] <AryehGregor>
And if they do, it's only a very small fraction of profit, like one or two percent.

[16:32:02.0000] <AryehGregor>
So it's not a big deal in the scheme of things.

[16:32:03.0000] <wilhelm>
The growth of the value of the company is usually sufficient.

[16:32:04.0000] <The_8472>
unless someone buys them up, changes the rules and syphons off money, seen that all too often

[16:33:00.0000] <AryehGregor>
Right, and dividends slow down the company's growth sometimes.

[16:33:01.0000] <AryehGregor>
The_8472, hostile takeovers of that sort are very rare.  A publicly-traded company almost always has far higher market capitalization than assets.  Usually when you have that sort of systematic buy-out, it's an attempt to merge one company into a larger one or such.

[16:34:00.0000] <AryehGregor>
But it's hard to pull off, because trying it causes the stock price to shoot up.

[16:34:01.0000] <The_8472>
<AryehGregor> It's simpler than that, it's just that large businesses are powerful enough to have bargaining power with the government. <- that's one aspect, but not the only one. just look at all the externalties that companies cause and that the taxpayer has to wipe up after them

[16:34:02.0000] <AryehGregor>
Anyway, if you do buy a lot of stock in the company, it's rarely in your interest to try grabbing the assets and running.  You want to hold onto the stock and let it go up in price, then sell it.  That almost always provides better returns.

[16:35:00.0000] <AryehGregor>
Except if the company's stock price has really crashed, then it's sometimes worthwhile to buy it just for the sake of liquidating it.

[16:35:01.0000] <The_8472>
oh, you don't grab the assets. you slowly bleed money from them, then sell it again while it's still somewhat performing

[16:35:02.0000] <AryehGregor>
The_8472, practically any economic activity causes externalities, both positive and negative.  Big businesses can bring a lot of positive economic effects, like greatly reduced costs through economy of scale.

[16:36:00.0000] <The_8472>
i'm talking about negative externalties, obviously.

[16:36:01.0000] <AryehGregor>
They can devote a much larger amount of their budget to R&D and innovation, etc.

[16:36:02.0000] <AryehGregor>
Well, yes, but you can't fairly complain about the negative externalities unless you have evidence that they outweigh the positive externalities.

[16:36:03.0000] <The_8472>
which sometimes exceed even the profit that the companies themselves make (look at nuclear fuel processing and waste storage cycles)

[16:37:00.0000] <The_8472>
it's basically cheap power for the economy in exchange for the govt funding the whole infrastructure and waste management for millions of years to come.

[16:39:00.0000] <wilhelm>
AryehGregor: Sure you can. I spend all day searching for an complaining about the few negative issues in otherweise good software. I don't see why the management of this planet's productive resources should be exempt from any criticism. Quite the contrary.

[16:43:00.0000] <The_8472>
you could go even further... we're putting some of those negative externalties on those who don't see anything of the positive ones (i.e. everything that isn't human)


2011-08-05
[17:01:00.0000] <zewt>
can browsers implement rel=noreferrer already so "anonymous link" sites can die horribly

[17:02:00.0000] <The_8472>
uhm... you know... you could just block cross domain referrers

[17:03:00.0000] <zewt>
uhm, that wouldn't magically make sites using anonymous link proxies stop doing so

[17:03:01.0000] <zewt>
i have GM scripts to undo some of them, but i shouldn't have to, and it doesn't work for all of them (not all of them encode the real URL as a query parameter)

[17:03:02.0000] <The_8472>
if that's your issue... there are addons that clean up the urls as long as the actual target is embedded in the link

[17:04:00.0000] <The_8472>
anyway, noreferrer only solves half of the problem. just look at google. they're linkjacking their own links to track what you've clicked on.

[17:05:00.0000] <zewt>
that's a separate (and roughly equally disgusting) problem

[17:05:01.0000] <The_8472>
it's not just the target sites that harvest data. it's the source sites too

[18:17:00.0000] <rniwa>
annevk: yt?

[18:17:01.0000] <rniwa>
annevk5: ^

[18:21:00.0000] <Hixie>
ironically one of the hardest parts of this <dialog> thing is likely to be working out how to handle the positioning

[18:21:01.0000] <Hixie>
position:fixed or position:absolute?

[18:21:02.0000] <Hixie>
has to be absolute so you can scroll it if it's too big for the page

[18:21:03.0000] <Hixie>
but

[18:21:04.0000] <Hixie>
do you want to scroll the stuff under it in that case?

[18:21:05.0000] <Hixie>
how do you make sure it is visible when it appears in that case?

[18:22:00.0000] <Hixie>
how do we make the "cover" div under it actually cover everything if the dialog is position:absolute?

[18:22:01.0000] <Hixie>
what an odd mess

[18:22:02.0000] <Hixie>
oh well, for later.

[18:46:00.0000] <jamesr>
Hixie, what's <dialog>? modal dialog?

[18:52:00.0000] <hober>
indeed

[19:00:00.0000] <rniwa>
Hixie: can we get your response for http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2011-June/032078.html ?

[19:01:00.0000] <rniwa>
Hixie: I have a WebKit patch that's sort of blocked by this thread

[20:58:00.0000] <Hixie>
AryehGregor: can you help out rniwa? (see above)

[22:10:00.0000] <Yuhong>
Hixie: I'd love to comment on your posts, but I don't have a Google+ account.

[23:25:00.0000] <Hixie>
Yuhong: what's your e-mail address? i can send you an invite

[23:25:01.0000] <Hixie>
Yuhong: though it's really only fun to use g+ if your close friends and family are on it

[23:26:00.0000] <Yuhong>
I just wish to comment on your posts.

[23:26:01.0000] <Yuhong>
Anyway: yuhongbao_386 at hotmail.com

[23:28:00.0000] <Hixie>
i shared something with you so you should be able to join now

[01:43:00.0000] <hsivonen>
how do these bogus spec bugs that contain the source code of an HTML document come about?

[01:43:01.0000] <hsivonen>
do spammers expect some system somewhere to serve pasted source as an HTML page?

[01:43:02.0000] <hsivonen>
do people accidentally manage to paste view source contents into the comment form?

[01:50:00.0000] <zcorpan>
maybe people go "wtf is this", then paste whatever they have in their clipboard, then click some buttons

[01:52:00.0000] <zcorpan>
maybe the comment form needs an anti-bogo checkbox saying "I understand that submitting this form will file a bug in the W3C Bugzilla"

[02:04:00.0000] <zcorpan>
i don't like the a11yTF bugs that just say current text and suggested new text without mentioning what problem they are trying to solve or highlight what the difference is between the new and old text

[02:05:00.0000] <zcorpan>
so readers have to first carefully read both texts to find out what the difference is, and then make a guess at what problem they're trying to solve

[02:05:01.0000] <zcorpan>
which means that i just don't read those bugs at all

[02:05:02.0000] <zcorpan>
hi john, btw

[02:24:00.0000] <annevk>
Hixie, around now, somewhat

[02:27:00.0000] <annevk>
oh miwa was asking

[02:27:01.0000] <annevk>
I defined superglobal id/class

[02:28:00.0000] <annevk>
not really satisfied with it yet, but it's there

[04:24:00.0000] <matjas>
what’s the (hi)story behind the element@attribute syntax?

[04:25:00.0000] <matjas>
I knew about element/@attribute (XPath) but e.g. https://bugzilla.mozilla.org/show_bug.cgi?id=573322 uses element@attribute

[04:25:01.0000] <matjas>
where does it originate?

[04:28:00.0000] <annevk>
@attribute originates from XPath and people just went with it from there

[04:32:00.0000] <matjas>
I figured as much, thanks for confirming :)

[07:03:00.0000] <zcorpan>
if <button form formaction> etc is a security problem because todays filters use blacklists instead of whitelists, then surely <video src=? onloadstart="..."></video> is more of a security problem

[07:04:00.0000] <zcorpan>
or do blacklists uniformly block attributes that start with "on"?

[07:05:00.0000] <Ms2ger>
I sure hope so

[07:07:00.0000] <annevk>
people raised new event handlers as security issue

[07:09:00.0000] <jgraham>
In other news, using blacklists is a security problem

[07:09:01.0000] <zcorpan>
i wonder if spammers are going to love the hidden attribute

[07:10:00.0000] <zcorpan>
(where site authors/admins use modern browsers and don't see the spam while a non-zero amount of users use legacy browsers that don't support hidden and will see the spam)

[07:10:01.0000] <danj>
and in other news, splitting atoms in your kitchen is a bad idea

[07:11:00.0000] <Ms2ger>
Is it really?!

[07:12:00.0000] <danj>
yup - http://hosted.ap.org/dynamic/stories/E/EU_SWEDEN_NUCLEAR?SITE=AP&SECTION=HOME&TEMPLATE=DEFAULT&CTIME=2011-08-03-11-33-45

[07:15:00.0000] <zcorpan>
/me adds download to html-elements

[07:16:00.0000] <jgraham>
/me notes that atom-splitting in the kitchen happens all the time

[07:18:00.0000] <zcorpan>
hsivonen: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13604 - the dom wouldn't be identical anyway, since in text/html the textContent would contain "<![CDATA[" while in XML it wouldn't

[07:20:00.0000] <hsivonen>
zcorpan: not in foreign content

[07:21:00.0000] <zcorpan>
hsivonen: i thought the bug was discussing html script elements

[07:23:00.0000] <hsivonen>
zcorpan: I clarifyied in case I misunderstood

[07:41:00.0000] <annevk>
In case it comes up or someone is asking, I'm away Mon-Sat

[07:42:00.0000] <zcorpan>
we're not publishing next week right?

[07:43:00.0000] <annevk>
HTML?

[07:43:01.0000] <annevk>
I don't think so

[07:44:00.0000] <zcorpan>
ok

[08:19:00.0000] <Philip`>
hsivonen: Re http://krijnhoetmer.nl/irc-logs/whatwg/20110802#l-459 : I noticed http://www.w3.org/Bugs/Public/show_bug.cgi?id=13508 uses both Nadia and Aidan, and http://www.w3.org/WAI/PF/HTML/wiki/Spec_Review/Interactive_Elements says "Aidan is the opposite of Nadia"

[08:19:01.0000] <Philip`>
which seems an unlikely coincidence so I guess they were intentionally chosen to be palindromic

[08:23:00.0000] <zcorpan>
what's with the "06" in e.g. http://krijnhoetmer.nl/irc-logs/whatwg/20110804#l-761 ?

[08:24:00.0000] <timeless>
--note, the following messages were composed while offline (see comments about using bugzilla while offline...) --

[08:24:01.0000] <timeless>
AryehGregor: mikesmith could read the irc logs :)

[08:24:02.0000] <timeless>
AryehGregor: there have been proposals for dealing w/ mass changes

[08:24:03.0000] <timeless>
sadly, as someone who does mass changes and deals w/ them, it's actually valuable to see each individual message

[08:24:04.0000] <timeless>
since i often screw up say 3% of the bugs i touch in mass changes

[08:24:05.0000] <timeless>
and there's no way you'll spot that while doing the mass change nor in a single email report

[08:24:06.0000] <timeless>
jgraham: in theory one could try to do it directly, but bugzilla keeps track of whom has seen which comment/change

[08:24:07.0000] <timeless>
which means it requires touching each user reference for each bug

[08:24:08.0000] <timeless>
otherwise the next touch to the bug will trigger the unsent changes

[08:25:00.0000] <timeless>
AryehGregor: depending on what you want to do w/ mass changes, you can open the query in a decent version of gmail and use <n> to quickly scan through them

[08:25:01.0000] <timeless>
or you can use select all (in that decent version of gmail) and mark as read

[08:25:02.0000] <timeless>
(there are lame versions of gmail -- in case you're wondering-- where you can't do these things, e.g. the java, html [/h], and mobile [/x])

[08:25:03.0000] <timeless>
zewt: importance also goes the other way, someone will set a very important thing to trivial so that it won't flag a manager :)

[08:25:04.0000] <timeless>
Philip`: i make mass changes to thousands of bugs. having to click thousands links in a bugmail to get back my thousands of bugmails would not be fun

[08:25:05.0000] <timeless>
AryehGregor: there are a couple of reasons to want functional archives of bugzilla:

[08:25:06.0000] <timeless>
1. bugzilla search sucks -- gmail search (free text) is much better

[08:25:07.0000] <timeless>
2. when you make a mistake, it's much easier to undo it if you have the bugmail

[08:25:08.0000] <timeless>
3. bugzilla data reflects current info (e.g. see when AryehGregor renamed his bugzilla account), bugmail reflects data as it was

[08:25:09.0000] <timeless>
4. some people work offline (even gmail supports this)

[08:25:10.0000] <timeless>
AryehGregor: re calendar to find day of week, that's awesome

[08:25:11.0000] <timeless>
/me just hovers over the w7 system notification area and it tells me the day of week

[08:25:12.0000] <timeless>
/me is no longer offline

[08:26:00.0000] <jgraham>
I feel like someone just opened a time capsule

[08:27:00.0000] <jgraham>
But from yesterday, not 1979

[08:27:01.0000] <jgraham>
Or whenever time capsules were mostly buried

[08:27:02.0000] <Philip`>
zcorpan: Probably ANSI escape codes or similar

[08:27:03.0000] <timeless>
i spent yesterday unpacking my shipped apartment

[08:27:04.0000] <timeless>
some of which had boxed from a previous move (5 years ago)

[08:28:00.0000] <timeless>
and some of which have stuff from the move before that (closer to 8 years ago)

[08:28:01.0000] <timeless>
so i know what you mean re time capsules :)

[08:28:02.0000] <timeless>
but, that's also why i wasn't replying in real-time...

[08:29:00.0000] <Philip`>
Oh, actually mIRC colour escapes

[08:29:01.0000] <Philip`>
since it's an 0x03 at the start

[08:29:02.0000] <zewt>
heh, mirc codes are pretty sadly amusing

[08:29:03.0000] <timeless>
can someone ask the blogger of http://blog.chromium.org/2011/08/connecting-web-apps-with-web-intents.html to fix their quotes to use ['] instead of the fancy ones they use?

[08:29:04.0000] <zewt>
apparently the guy invented them because he didn't know that ansi color codes already existed and had been in use on irc for years?

[08:30:00.0000] <timeless>
/me chuckles

[08:30:01.0000] <timeless>
/me isn't surprised

[08:31:00.0000] <zewt>
essentially, reinventing the wheel after you've already been selling cars for years

[08:31:01.0000] <Philip`>
timeless: I think my idea was that you'd get a single bugmail with a single link to click, which would then cause the thousands to be sent to you

[08:31:02.0000] <Philip`>
so it wouldn't be not fun

[08:32:00.0000] <timeless>
Philip`: subscriber preferences make more sense i think

[08:32:01.0000] <timeless>
i haven't met someone who only sometimes wants all bugmail changes

[08:32:02.0000] <timeless>
although, i could be wrong

[08:32:03.0000] <timeless>
the problem is that bugzilla really can't create those emails later

[08:33:00.0000] <timeless>
e.g. if the mass changer is AryehGregor, and i get the email and then he changes his email address

[08:33:01.0000] <timeless>
and then i click the link asking for his mass changes to be split

[08:33:02.0000] <timeless>
the emails bugzilla would give me would have the *wrong* changer address in them

[08:33:03.0000] <timeless>
and if he changes his real name to claim he's on vacation

[08:33:04.0000] <timeless>
then i'd see that he sent them while on vacation

[08:33:05.0000] <timeless>
...

[08:33:06.0000] <timeless>
the result is bugzilla would have to queue the possible email for all possible recipients

[08:33:07.0000] <timeless>
which would be really bad for storage for bugzilla

[08:34:00.0000] <Philip`>
Disk space is cheap, so queueing is easy

[08:34:01.0000] <Philip`>
They'd compress well too :-)

[08:34:02.0000] <timeless>
well, they would compress well yes

[08:34:03.0000] <zewt>
why? save the body of the outbound message, one per change

[08:34:04.0000] <timeless>
but bugzillas tend to use mysql

[08:34:05.0000] <timeless>
which probably sucks at compressing

[08:35:00.0000] <timeless>
they'd only compress well as a complete set, not as individual messages

[08:35:01.0000] <Philip`>
Concatenate all the messages, gzip them, then store them in a BLOB column, and hope no database administrator notices your terrible abuse of SQL

[08:35:02.0000] <timeless>
i suppose you could compose an MBox for each user

[08:35:03.0000] <zewt>
would also allow you to do things like request a full email audit trail of a bug later on

[08:35:04.0000] <gsnedders>
timeless: MySQL can compress parts

[08:36:00.0000] <hsivonen>
Philip`: mindblowing. Is there Krid, too?

[08:36:01.0000] <timeless>
gsnedders: databases are generally designed for pigeon holes, and compressing an individual hole doesn't gain much

[08:36:02.0000] <Philip`>
Alternatively, maybe emails could have some extra header which contains some globally unique mail-set identifier, and mail clients could have a "select all in same set as this message" feature

[08:36:03.0000] <zewt>
well, it would if you were storing a block of text repeated thousands of times, but again I don't see why you need to store them all

[08:36:04.0000] <timeless>
but i suppose if you arranged a portion for inidividual recipients... it might be doable

[08:37:00.0000] <timeless>
Philip`: so...

[08:37:01.0000] <Philip`>
then bulk change mails would get put into the same set

[08:37:02.0000] <timeless>
on that count, there is a mass change header in the mails

[08:37:03.0000] <timeless>
so if your client *isn't* gmail, you already have that!

[08:37:04.0000] <timeless>
it's in x-bugzilla-change-reason: iirc

[08:37:05.0000] <zewt>
gmail is horrifyingly stupid about headers :|

[08:37:06.0000] <Philip`>
Ah, excellent

[08:37:07.0000] <timeless>
zewt: ue[

[08:37:08.0000] <timeless>
yep

[08:37:09.0000] <timeless>
so, we're back to gmail being the only disadvatanged client

[08:38:00.0000] <zewt>
(never mind that it literally has no concept of sorting)

[08:38:01.0000] <timeless>
yeah, which is another problem

[08:38:02.0000] <zewt>
gmail is good at a lot of things but it has these bizarre holes

[08:38:03.0000] <timeless>
i think gmail lists messages by delivery date

[08:38:04.0000] <timeless>
which means if i trigger the email flood after a reply to the flood

[08:38:05.0000] <timeless>
the message sequence will look bizarre

[08:38:06.0000] <timeless>
i already get that in Hg land

[08:38:07.0000] <zewt>
that's the user's problem, i'd say

[08:38:08.0000] <Philip`>
Someone who cares about mail should smuggle themselves into Google and fix Gmail

[08:39:00.0000] <timeless>
where smaller messages sent later are delivered before larger messages sent earlier!

[08:39:01.0000] <timeless>
(and are thus listed closer to 'beginning of time')

[08:39:02.0000] <zewt>
also no way in gmail to view messages *to me* :O

[08:40:00.0000] <timeless>
zewt: ?

[08:40:01.0000] <zewt>
eg. all mail minus mailing list mail

[08:40:02.0000] <timeless>
there's a >> or > indicator on things

[08:40:03.0000] <Ms2ger>
On the subject of email

[08:40:04.0000] <timeless>
although that might only apply to conversations as opposed to messages

[08:40:05.0000] <zewt>
i don't know of any way to search for "no labels"

[08:40:06.0000] <timeless>
perhaps you have to turn that on or off

[08:40:07.0000] <zewt>
(not exactly what I want, but close)

[08:40:08.0000] <timeless>
*that*'s indeed annoying

[08:40:09.0000] <Ms2ger>
Why does MS's Alex use plus-minus signs to quote?

[08:40:10.0000] <timeless>
i really would like <has-no-labels>

[08:41:00.0000] <timeless>
Ms2ger: hrm, good question

[08:42:00.0000] <timeless>
would it be bad if i asked the author of http://blogs.forbes.com/fredcavazza/2011/07/17/why-opposing-html5-and-flash-is-a-non-sense/ to fix his spelling of a Trademark?

[08:42:01.0000] <zewt>
also while randomly ranting about gmail, there's no way to search or sort my message size, so if you have a bunch of attachments using up gigs of space, there's pretty much no way to find them

[08:42:02.0000] <zewt>
(short of downloading the whole lot and loading it in another client)

[08:43:00.0000] <timeless>
zewt: yep

[08:43:01.0000] <timeless>
actually, i think you can try using an imap client

[08:43:02.0000] <timeless>
but perhaps we don't count that :)

[08:43:03.0000] <zewt>
timeless: "is a non sense" heh

[08:43:04.0000] <timeless>
?

[08:43:05.0000] <zewt>
always funny when things like that are edited in the text but left unchanged in the url

[08:43:06.0000] <timeless>
oh

[08:43:07.0000] <timeless>
lol

[08:44:00.0000] <timeless>
cute

[08:44:01.0000] <timeless>
/me kicks firefox find in page for not finding in url

[08:44:02.0000] <zewt>
reminds me of the old ^H^H^H^H^H^H

[08:44:03.0000] <timeless>
yep

[08:44:04.0000] <timeless>
anyway, think he'd respond well to such a request?

[08:46:00.0000] <zewt>
dunno, i skimmed his "points", saw too much non sense(sic) in them and closed the tab

[08:46:01.0000] <zewt>
(we can't replace flash! it's been around for TWELVE YEARS!)

[08:46:02.0000] <timeless>
his statements aren't mostly wrong :)

[08:47:00.0000] <gsnedders>
HTML has been around for TWENTY ONE YEARS. Flash can't replace that!

[08:48:00.0000] <Philip`>
If HTML5 doesn't officially exist because it's not a published W3C standard yet, does Flash not exist either since it's also not one?

[08:49:00.0000] <Ms2ger>
Does that also mean MS Word doesn't exist?

[08:49:01.0000] <jgraham>
Does that also mean that Descartes doesn't exist?

[08:50:00.0000] <reggna>
Do I exist?

[08:51:00.0000] <timeless>
no

[08:52:00.0000] <jgraham>
I don't think

[08:52:01.0000] <zcorpan>
i am not a W3C recommendation, therefore I don't exist

[08:52:02.0000] <timeless>
zcorpan: we're working on rescinding those

[08:53:00.0000] <timeless>
therefore even if you were, you'd be subject to ceasing to exist within ... 4 months?

[08:53:01.0000] <timeless>
annevk: how's that going btw? :)

[08:53:02.0000] <zcorpan>
timeless: murder!

[08:53:03.0000] <Ms2ger>
/me asks ArtB

[08:54:00.0000] <timeless>
zcorpan: planned obsolesence

[08:55:00.0000] <annevk>
timeless, haven't looked too much into it yet

[08:55:01.0000] <timeless>
annevk: i was glad to see ian was responsive

[08:56:00.0000] <timeless>
hopefully we can get it done in closer to 2 months or even 6 weeks :)

[08:56:01.0000] <timeless>
afterwards, i think we can plan to go duck hunting and bag a larger set of lame ducks

[09:08:00.0000] <AryehGregor>
Hixie, I think the right answer here is obviously to make it logical order, so that "character" still moves forward one character and "line" still moves forward a bunch of characters.  But I don't know the first thing about vertical writing, rniwa would probably have a much more informed opinion.

[09:08:01.0000] <AryehGregor>
I still don't really know what the use-cases are for modify().

[09:10:00.0000] <gsnedders>
AryehGregor: one character or one grapheme?

[09:10:01.0000] <gsnedders>
Grapheme is surely what would be expected? How do you put the cursor in the middle of a grapheme.

[09:10:02.0000] <AryehGregor>
gsnedders, most likely a "grapheme cluster", in the terminology of UAX#29.

[09:11:00.0000] <AryehGregor>
That's not what I was talking about, though.

[09:11:01.0000] <gsnedders>
Yeah, I know.

[09:12:00.0000] <AryehGregor>
/me 's eyes glaze over slightly as he reads the definition of "extended grapheme cluster"

[09:12:01.0000] <gsnedders>
Line is probably still what you want with vertical writing, AIUI, just the positioning of the next line is different.

[09:13:00.0000] <AryehGregor>
/me has no idea if the rules make any sense whatsoever for his use-cases, because he has no idea how things like "the spacing (but dependent) vowel signs in Indic scripts" are actually used

[09:13:01.0000] <AryehGregor>
RTL is a piece of cake compared to some of the stuff you see in Unicode.

[09:16:00.0000] <Hixie>
AryehGregor: well, the stuff in question is in your spec now, not mine, so i wash my hands of it :-P

[09:16:01.0000] <AryehGregor>
:)

[09:16:02.0000] <AryehGregor>
I would have replied to rniwa, but he's apparently one of those annoying people who only stays on IRC when he's actually present.

[09:16:03.0000] <AryehGregor>
Rather than idling all the time.

[09:17:00.0000] <Ms2ger>
:(

[09:17:01.0000] <AryehGregor>
/me notes that it's Ms2ger's spec too, and in fact Ms2ger is the only listed editor, so it's only fair to push the problem off on him

[09:18:00.0000] <Hixie>
oh, didn't realise that

[09:18:01.0000] <Ms2ger>
Not anymore

[09:18:02.0000] <Hixie>
hah

[09:18:03.0000] <AryehGregor>
Drat.

[09:18:04.0000] <Ms2ger>
/me removes himself

[09:18:05.0000] <AryehGregor>
https://bitbucket.org/ms2ger/dom-range/changeset/eb238987d64e

[09:18:06.0000] <AryehGregor>
I didn't notice that.

[09:18:07.0000] <AryehGregor>
Oh well.

[09:18:08.0000] <AryehGregor>
We were planning to chop it up and redistribute it among other specs anyway.

[09:19:00.0000] <Ms2ger>
Yeah

[09:19:01.0000] <Ms2ger>
And that question is very much yours :)

[09:19:02.0000] <AryehGregor>
Of course, now Microsoft is complaining about Traversal being in Core, so they'd presumably complain about Range too.

[09:20:00.0000] <AryehGregor>
And I don't want to steal Selection for the editing spec yet because I'm getting lots of review right now and don't want to rock the boat.

[09:25:00.0000] <Hixie>
jgraham: i get 504s quite frequently these days

[09:25:01.0000] <Hixie>
jgraham: even without the annotations enabled

[09:31:00.0000] <jgraham>
Hixie: Hmm. Right now or a few days ago?

[09:31:01.0000] <Hixie>
had one a few minutes ago

[09:31:02.0000] <Hixie>
it's transient

[09:31:03.0000] <Hixie>
if i retry sufficient times it goes away

[09:31:04.0000] <jgraham>
Hmm, OK

[09:31:05.0000] <jgraham>
I will try to look at the problem

[09:32:00.0000] <Hixie>
btw i have the annotation stuff disabled right now because it was failing too often also

[09:47:00.0000] <zewt>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13264 do bug trolls really need a serious response? heh

[09:47:01.0000] <Hixie>
that's not a troll, he's probably right

[09:47:02.0000] <Hixie>
he's certainly right for the html spec

[09:47:03.0000] <Hixie>
dunno what we can add for storage though

[09:48:00.0000] <Ms2ger>
http://rlv.zcache.com/chocolate_chip_cookie_jar_postcard-p239265891586692704trdg_400.jpg

[09:48:01.0000] <zewt>
suggesting pictures in web storage sure sounds 100% trollish to me, heh

[09:48:02.0000] <Philip`>
Trolls can be right

[09:48:03.0000] <AryehGregor>
Okay, I want to post something like this in the public-webapps thread, but I probably shouldn't: "Well, on the other hand, we now have a senior member of the W3C administration demanding that the W3C's editors only reference W3C standards, even if those editors feel parallel standards developed at other standards bodies are superior.  It seems to me that's *definitely* for no reason other than politics, and also undermines and is disrespectfu

[09:48:04.0000] <AryehGregor>
l of the work of the WebApps WG.  But hey, it's your call.  If you alienate editors, it doesn't hurt anyone but the W3C."

[09:48:05.0000] <Ms2ger>
You shouldn't

[09:49:00.0000] <AryehGregor>
I guess I'll just remain silent.

[09:49:01.0000] <Hixie>
Ms2ger: hah, nice picture

[09:49:02.0000] <Ms2ger>
I would appreciate that in the spec ;)

[09:49:03.0000] <zewt>
the "too much reading" in particular is what makes it stand out as trolly to me, heh

[09:49:04.0000] <AryehGregor>
Although the hypocrisy of him saying we have to remove all references to the WHATWG standards and then accusing us of politics is galling.

[09:49:05.0000] <Ms2ger>
Maybe you could photoshop it onto the kitchen sink

[09:50:00.0000] <Hixie>
Ms2ger: hah

[09:50:01.0000] <Hixie>
Ms2ger: is it cc?

[09:50:02.0000] <Ms2ger>
Doubt it

[09:50:03.0000] <Hixie>
the sink is :-)

[09:50:04.0000] <Ms2ger>
I know

[09:50:05.0000] <Hixie>
got an attribution in there and everything

[09:50:06.0000] <Ms2ger>
Even microdata processors know that

[09:50:07.0000] <Hixie>
:-P

[09:51:00.0000] <AryehGregor>
I'm not sure I want to have a Bugzilla component in the W3C.  That gives them leverage.

[09:51:01.0000] <AryehGregor>
Maybe I'll use Google Code instead.

[09:51:02.0000] <smaug____>
then I won't fire any bugs

[09:51:03.0000] <smaug____>
file

[09:51:04.0000] <AryehGregor>
If it's in Google Code, you mean?

[09:51:05.0000] <smaug____>
could whatwg have its own bugzilla?

[09:51:06.0000] <smaug____>
AryehGregor: right

[09:52:00.0000] <AryehGregor>
We could, but then everyone would have to make an account.

[09:52:01.0000] <AryehGregor>
Which is annoying.

[09:52:02.0000] <Ms2ger>
People sure would like that

[09:52:03.0000] <AryehGregor>
Lots of people already have W3C Bugzilla accounts, and practically everyone has a Google account.

[09:52:04.0000] <smaug____>
I'm about to give up my Google account ...

[09:52:05.0000] <Philip`>
http://www.flickr.com/photos/apreche/3442252639/ is CC BY

[09:53:00.0000] <AryehGregor>
I don't want people to have to remember Yet Another Password to file bugs.  I guess we could set up something like the contributor⊙wo system, but that sounds like work.

[09:53:01.0000] <AryehGregor>
Anyway, I'll still accept feedback via e-mail regardless, and file it myself if I can't act on it quickly.

[09:53:02.0000] <Hixie>
i can give you the file-bug.cgi script if you want

[09:53:03.0000] <AryehGregor>
I suppose I'll keep the W3C Bugzilla component for now.

[09:54:00.0000] <AryehGregor>
I assume nobody's going to do anything as drastic as delete all the bugs without warning.

[09:54:01.0000] <AryehGregor>
So it should be safe.

[09:54:02.0000] <Hixie>
the worst that i could see happening is that the bugmail notifications would be turned off and the bugs would all be reassigned or something like that

[09:54:03.0000] <Hixie>
which would make recollecting them a huge pain

[09:54:04.0000] <Hixie>
but not impossible

[09:54:05.0000] <AryehGregor>
I really should avoid getting into political sniping anyway.

[09:55:00.0000] <Hixie>
(or all closed)

[09:55:01.0000] <AryehGregor>
I don't think anyone will do anything hostile without fair warning, at least not for now.

[09:56:00.0000] <shetech>
one hopes

[09:56:01.0000] <Hixie>
the w3c isn't evil

[09:56:02.0000] <Hixie>
they're just using an archaic process

[09:56:03.0000] <shetech>
Heh. I wasn't talking about w3c, but I'm occasionally a cynic. ;-)

[09:57:00.0000] <AryehGregor>
No, but every time something happens like plh demanding that W3C specs only reference W3C specs, that makes me want to have less and less to do with the W3C.

[09:57:01.0000] <gsnedders>
AryehGregor: only reference W3C specs when they publish the document, rather, which doesn't seem that unreasonable

[09:57:02.0000] <shetech>
Ah, politics. It's what makes the world go 'round.

[09:58:00.0000] <Hixie>
AryehGregor: oh the bureaucracy at w3c certainly makes me wish i could just ignore them entirely

[09:58:01.0000] <Hixie>
gsnedders: well except for the part where "publish" is itself unreasonable. :-P

[09:58:02.0000] <Hixie>
bbia

[09:58:03.0000] <Hixie>
b

[09:58:04.0000] <AryehGregor>
gsnedders, it seems unreasonable from my perspective as an editor that a W3C administrator would think he can waste my time telling me where my informative references have to point.

[09:59:00.0000] <dglazkov>
good morning, Whatwg!

[09:59:01.0000] <shetech>
Pardon me, but isn't that the w3c administrator's gig? To *pull* what they want?

[10:00:00.0000] <AryehGregor>
Well, yes.  I kind of object to the existence of a W3C administration, basically.

[10:00:01.0000] <shetech>
/me snorts

[10:00:02.0000] <shetech>
(in a humorous way)

[10:00:03.0000] <AryehGregor>
We get along fine without any administration at the WHATWG.

[10:01:00.0000] <shetech>
yes. I heart the way whatwg is running itself. it sure makes my job easier! Oh, except for the fact that whatever I write is pretty much obsolete next week. :D

[10:01:01.0000] <shetech>
No matter, I like the whatwg approach better.

[10:01:02.0000] <shetech>
Makes more sense

[10:02:00.0000] <jgraham>
What is obsolete next week?

[10:02:01.0000] <shetech>
Not really anything *from* the whatwg... more that comments about browser behavior aren't "current" for very long

[10:03:00.0000] <jgraham>
 Blame browsers for that not the WHATWG

[10:03:01.0000] <shetech>
I ran into one, for example, about differences between how Chrome and FF handle svg files using the <img> tag to place them.

[10:03:02.0000] <shetech>
Yes, browser fault, not whatwg

[10:03:03.0000] <shetech>
:)

[10:03:04.0000] <shetech>
I do think that the "living standard" approach just makes tons more sense in our agile world

[10:03:05.0000] <jgraham>
I mean who wants to live in a world where browsers fix bugs and get new, exciting features, really?

[10:03:06.0000] <jgraham>
;)

[10:03:07.0000] <shetech>
HEH

[10:06:00.0000] <shetech>
besides, jgraham: IE doesn't. :P

[10:06:01.0000] <shetech>
that's what makes it so exciting!

[10:08:00.0000] <jgraham>
Well IE does now compared to 2004

[10:08:01.0000] <jgraham>
Although of course lots of people still use 2004 era IE

[10:08:02.0000] <shetech>
True that. I worked on a team last year that still insisted on IE6.

[10:08:03.0000] <shetech>
Yoiks

[10:09:00.0000] <jgraham>
In related news, Microsoft are the only browser vendor who optimise for large enterprise customers (in pretty much any of their product line)

[10:10:00.0000] <jgraham>
(I mean Apple and Google don't really do "enterprise" except in very narrow ways, even on non-browser products)

[10:10:01.0000] <jgraham>
Possibly this explains the culture at Microsoft that likes numbered versions of things and formal processes and so on

[10:11:00.0000] <shetech>
Well, that's true enough. Apple especially, and Goog to a lesser extent are really intended for consumer level.

[10:11:01.0000] <shetech>
jgraham: would you be willing to provide a little more detail about how IE optimizes for large enterprise? Or am I making an incorrect connection?

[10:11:02.0000] <AryehGregor>
It would be nice if I could tell anolis not to do backlinks to some references.  Specifically the ones in notes and things.

[10:13:00.0000] <jgraham>
shetech: Well I don't know since I neither use IE often nor know much about large enterprises. But I meant that they have very long support lifetimes and integrate with all the windows stuff for doing mass-deployment and locking down settings and so on

[10:13:01.0000] <Ms2ger>
Why?

[10:13:02.0000] <gsnedders>
shetech: Make it easy to deploy to thousands of system, deloy patches and other upgrades in such a way that fits in with standard workflows for testing new software, with long support times.

[10:14:00.0000] <jgraham>
I just said that!

[10:14:01.0000] <AryehGregor>
Why what?

[10:14:02.0000] <AryehGregor>
I don't want backlinks to non-normative stuff because I often use backlinks to figure out what's using an algorithm so I know if I can simplify or get rid of or replace it.

[10:15:00.0000] <AryehGregor>
But I want to use <span> and not manual <a> to link to the algorithm because otherwise the link will rot if I ever remove or rename it.

[10:15:01.0000] <shetech>
Hee.

[10:15:02.0000] <Ms2ger>
Mm

[10:15:03.0000] <shetech>
Thanks guys

[10:16:00.0000] <shetech>
jgraham and gsnedders, thanks. This is helpful (short and sweet).

[10:19:00.0000] <zewt>
AryehGregor: might be more useful to be able to visually separate normative and informative references in the backlinks popup

[10:19:01.0000] <Philip`>
AryehGregor: Would it be sufficient to use <a href=#whatever> if it detected and reported occurrences of #whatever with no corresponding id=whatever?

[10:19:02.0000] <AryehGregor>
zewt, also sounds like more work.

[10:19:03.0000] <zewt>
everything is work :)

[10:19:04.0000] <AryehGregor>
Philip`, that would work, yeah.  It would be more verbose in the source code, but that's no big deal.

[10:20:00.0000] <Philip`>
(The spec splitter will warn about broken fragment links like that, I think)

[10:20:01.0000] <annevk>
AryehGregor, it's even better

[10:21:00.0000] <Philip`>
(http://www.whatwg.org/specs/web-apps/current-work/multipage/stdout.txt - like at the end of that)

[10:21:01.0000] <AryehGregor>
annevk, what is?

[10:21:02.0000] <annevk>
he's blocking publication of Progress Events because of a non-normative reference (used for an example) to the WHATWG copy of HTML

[10:21:03.0000] <AryehGregor>
Yeah, exactly.  And anyone puts up with this nonsense why?

[10:22:00.0000] <annevk>
I might not much longer

[10:22:01.0000] <AryehGregor>
Nice.

[10:22:02.0000] <AryehGregor>
Or, not really nice, but hopefully it will cause the W3C to wake up a bit more.

[10:22:03.0000] <AryehGregor>
More like "necessary" than "nice".

[10:22:04.0000] <timeless>
doubtful

[10:23:00.0000] <jgraham>
Well it's not that surprising. plh's job depends on W3C being accepted as the biggest stanadsrds organisation in town, and companies like the patent policy

[10:23:01.0000] <jgraham>
*standards

[10:23:02.0000] <AryehGregor>
Is Opera willing to back you if you don't want to work anymore in the W3C?

[10:23:03.0000] <timeless>
fwiw, i'm pretty sure the w3 doc policy is pretty clear on referencing stable specs

[10:23:04.0000] <timeless>
and given that the whatwg specs are clearly marked as evolving

[10:23:05.0000] <AryehGregor>
jgraham, he seems not to realize that the way to do that is to make editors want to work there instead of bullying them into toeing the W3C party line.

[10:23:06.0000] <timeless>
i think their complaint is technically valid

[10:24:00.0000] <timeless>
albeit unfortunate and stupid

[10:24:01.0000] <AryehGregor>
timeless, no one was objecting to it on process grounds.

[10:24:02.0000] <AryehGregor>
HTML5 isn't stable either.

[10:24:03.0000] <AryehGregor>
It's still only a Last Call.

[10:24:04.0000] <AryehGregor>
Plus, this is an informative dependency, so stability is not required (right?).

[10:24:05.0000] <timeless>
that i'll have to recheck

[10:24:06.0000] <AryehGregor>
Only normative dependencies have to be as stable as the thing that's referencing them.

[10:24:07.0000] <timeless>
i haven't memorized the requirements maze

[10:24:08.0000] <AryehGregor>
Realistically, no version of HTML is going to be "stable" for a long time.

[10:25:00.0000] <jcranmer>
html 4

[10:25:01.0000] <timeless>
i'm a 3.2 fan

[10:25:02.0000] <Ms2ger>
I'm a 3.0 fan

[10:26:00.0000] <zewt>
i'm a ceiling fan :(

[10:26:01.0000] <jgraham>
HTML+ has been very stable

[10:26:02.0000] <jgraham>
Not implemented, but very stable…

[10:26:03.0000] <timeless>
http://www.w3.org/MarkUp/html3/

[10:26:04.0000] <timeless>
HTML 3.0 Draft (Exprired!) Materials

[10:26:05.0000] <shetech>
/me laughed out loud at zewt's riposte. Randomly, in the middle of a shared office.

[10:26:06.0000] <AryehGregor>
/me also laughed out loud

[10:26:07.0000] <Philip`>
/me is reminded of the quote "There is this special biologist word we use for 'stable'. It is 'dead'."

[10:27:00.0000] <timeless>
Philip`: yeah, that's come up in some contexts here at times

[10:27:01.0000] <timeless>
i think possibly relating to people wanting a stable version of Firefox

[10:27:02.0000] <annevk>
AryehGregor, no idea

[10:27:03.0000] <timeless>
one whose UI and feature set doesn't change

[10:27:04.0000] <jcranmer>
There is a special CS term for "dead"

[10:27:05.0000] <jcranmer>
"Debian stable"

[10:27:06.0000] <timeless>
lol

[10:27:07.0000] <Ms2ger>
jcranmer++

[10:27:08.0000] <annevk>
timeless, non-normative references can in theory reference anything

[10:28:00.0000] <timeless>
jcranmer++

[10:28:01.0000] <AryehGregor>
http://aryeh.name/quotes.html

[10:28:02.0000] <annevk>
timeless, I was willing to change the reference though before he cried foul in Team-only space

[10:28:03.0000] <timeless>
wow

[10:29:00.0000] <annevk>
timeless, now I'm probably still willing, but also less enthusiastic about working in that environment

[10:29:01.0000] <timeless>
annevk: prioritize: 1. kill as many w3 specs as fast as possible. 2. do the minimal work to get your stuff done.

[10:29:02.0000] <Ms2ger>
Hmm?

[10:30:00.0000] <timeless>
that way when you aren't in w3, there's less of w3 left to haunt you, and you don't have to worry about trying to get rid of those pieces when you aren't allow to affect it

[10:30:01.0000] <timeless>
<zcorpan_> going to file:/// gives me an xml parse error in firefox :-(

[10:30:02.0000] <timeless>
was zcorpan using a localized firefox?

[10:30:03.0000] <Ms2ger>
I don't think that was it

[10:30:04.0000] <Ms2ger>
But we fixed it

[10:30:05.0000] <timeless>
or did he have localized filenames which violated the encoding of the xml system

[10:30:06.0000] <timeless>
it's usually one or the other

[10:31:00.0000] <AryehGregor>
I think it was a non-printable ASCII character.

[10:31:01.0000] <timeless>
ah

[10:32:00.0000] <timeless>
/me sighs

[10:33:00.0000] <timeless>
every time i connect my phone, there's an OS update available

[10:33:01.0000] <timeless>
/me ponders

[10:33:02.0000] <timeless>
at nokia i wasn't willing to do updates to my phone more than monthly

[10:33:03.0000] <timeless>
probably because it broke way too many things each time

[10:41:00.0000] <dglazkov>
non-printable character:  an unfortunate outcome of one of the cocaine-laced brainstorms at Marvel.

[10:49:00.0000] <karlcow>
"[12:32]  <annevk> timeless, I was willing to change the reference though before he cried foul in Team-only space"

[10:49:01.0000] <karlcow>
cough cough not very clever from you annevk

[10:53:00.0000] <AryehGregor>
smaug____, why are you getting rid of your Google account, BTW?

[10:53:01.0000] <Philip`>
dglazkov: Maybe the character's costume is covered in those special symbols that are hidden on banknotes, so that printers will detect and refuse to reproduce them

[10:55:00.0000] <karlcow>
/me suddenly imagine the Super Hero costume with a Unicode error message on the shirt.

[10:55:01.0000] <smaug____>
AryehGregor: "big companies have too much influence in Internet"

[10:56:00.0000] <AryehGregor>
smaug____, well, okay.  I mean, there's a reason for that: they usually have the resources to do stuff better than small organizations.

[10:56:01.0000] <smaug____>
and they just happen to do evil things, even if they try not to

[10:57:00.0000] <AryehGregor>
Ideally stuff would be tied together a bit less and federated a bit more, but Google's better at that than most (although it could probably do a lot better than it is).

[10:57:01.0000] <mhausenblas>
annevk available?

[10:58:00.0000] <mhausenblas>
can you help me re a CORS question pls?

[10:58:01.0000] <mhausenblas>
now, a month ago someone pointed out that I have a syntax mistake in http://enable-cors.org/#how-apache

[10:59:00.0000] <mhausenblas>
people pointed out that Access-Control-Allow-Origin: "*" is wrong and it should be Access-Control-Allow-Origin: *

[11:00:00.0000] <mhausenblas>
now, I get a request from someone saying 'I spent a while trying to debug this with our apache servers.' and actually it is the other way round

[11:00:01.0000] <mhausenblas>
so, my inital version was right (?)

[11:00:02.0000] <mhausenblas>
any idea, annevk?

[11:01:00.0000] <mhausenblas>
hmmm ... I guess you're not around annevk  - I leave you a message on G+ then ...

[11:01:01.0000] <Philip`>
http://dvcs.w3.org/hg/cors/raw-file/tip/Overview.html#access-control-allow-origin-response-header looks like it says there shouldn't be any quotes

[11:02:00.0000] <gsnedders>
mhausenblas: there shouldn't be any quotes

[11:02:01.0000] <mhausenblas>
hmmm

[11:02:02.0000] <Philip`>
and later it says "If the Access-Control-Allow-Origin header value is the literal "*" character and the credentials flag is false return pass and terminate this algorithm."

[11:02:03.0000] <mhausenblas>
right

[11:02:04.0000] <Philip`>
which sounds like it's a single character

[11:03:00.0000] <mhausenblas>
so, the first request was right

[11:04:00.0000] <mhausenblas>
I just wonder why the new request (from a guy who claims that he's debugged it with Apache) thinks it's not the case

[11:04:01.0000] <mhausenblas>
I must confess that the spec is rather clear, the BNF says it all

[11:04:02.0000] <mhausenblas>
an Apache bug, maybe?

[11:04:03.0000] <mhausenblas>
nah

[11:05:00.0000] <Philip`>
Maybe 'Header set Access-Control-Allow-Origin *' should be 'Header set Access-Control-Allow-Origin "*"'

[11:05:01.0000] <gsnedders>
mhausenblas: What' she saying it's not the case?

[11:05:02.0000] <gsnedders>
s/What's/Why's/

[11:05:03.0000] <Philip`>
since http://httpd.apache.org/docs/2.0/mod/mod_headers.html indicates it sometimes wants the value to be double-quoted

[11:05:04.0000] <Philip`>
(and it'll strip the quotes off before sending the header, presumably)

[11:05:05.0000] <mhausenblas>
hmmm

[11:06:00.0000] <mhausenblas>
the person who says that the current status at http://enable-cors.org/#how-apache is wrong literally says:

[11:06:01.0000] <mhausenblas>
In the 'For Apache' section, please change :

[11:06:02.0000] <mhausenblas>
Header set Access-Control-Allow-Origin *

[11:06:03.0000] <mhausenblas>
to :

[11:06:04.0000] <mhausenblas>
Header set Access-Control-Allow-Origin "*"

[11:06:05.0000] <mhausenblas>
I spent a while trying to debug this with our apache servers.

[11:06:06.0000] <Philip`>
The Apache documentation sounds pretty vague about syntax

[11:06:07.0000] <mhausenblas>
now I'm a bit in trouble, ain't I?

[11:06:08.0000] <Philip`>
like what strings have to be double-quoted

[11:06:09.0000] <Philip`>
and how you could include a double quote inside a string if you wanted to

[11:07:00.0000] <mhausenblas>
I mean I changed it a month ago cause I was convinced - so blame it on a bad Apache doc? ;)

[11:07:01.0000] <AryehGregor>
There, I think this is productive: https://plus.google.com/100662365103380396132/posts/dzK5yHL1nA3

[11:07:02.0000] <mhausenblas>
:P

[11:07:03.0000] <mhausenblas>
anyways, I guess I have to toy around with it myself to figure what exactly is happening

[11:08:00.0000] <mhausenblas>
thanks for all your help, Philip` and gsnedders!

[11:08:01.0000] <mhausenblas>
now: FAWM

[11:08:02.0000] <Philip`>
Shifting blame to the documentation is always a good idea :-)

[11:09:00.0000] <mhausenblas>
he he

[11:09:01.0000] <mhausenblas>
ok, catch you laters online

[11:09:02.0000] <mhausenblas>
cya

[11:09:03.0000] <zewt>
heh, for htaccess directives i usually end up looking for an example, since the apache docs are ... not so great

[11:09:04.0000] <Philip`>
AryehGregor: Is G+ automatically expanding the link and putting the "Reference to the HTML specification. This message : [ Message body ] [ Respond ] [ More options ] ..." garbage in there?

[11:10:00.0000] <AryehGregor>
Philip`, yes.

[11:10:01.0000] <AryehGregor>
It doesn't notice that's boilerplate.

[11:10:02.0000] <AryehGregor>
Clearly, this would be a great use for everyone to use and support <header> and <article>.

[11:12:00.0000] <TabAtkins>
zewt: Re sorting mail by size, just use the :has(attachment) or whatever criteria.

[11:17:00.0000] <AryehGregor>
smaug____, FWIW, the only reason I thought of using Google Code in this case is because in practice there's no signup system like OpenID or BrowserID or whatever that actually works in practice.  It'd be awesome if BrowserID gets traction.

[11:17:01.0000] <zewt>
google can't even manage single-sign-in for their own products :|

[11:20:00.0000] <Philip`>
Bug trackers shouldn't really need to know people's identities anyway

[11:22:00.0000] <Philip`>
Let anyone file bugs without signing in, and just store some user ID in a permanent cookie so they can come back later and register with an email address and have all their bugs become associated with that, in case they decide they want to follow them by email

[11:22:01.0000] <zewt>
bugzilla is the worst about that

[11:23:00.0000] <Philip`>
(and have efficient spam-removal tools to deal with all the junk)

[11:23:01.0000] <TabAtkins>
And in the meantime you can still see what other bugs are associated with a particular user.

[11:23:02.0000] <zewt>
i signed in once and it was all "you have to change your password!!! and use a capital and a number this time!!!" it's a BUG TRACKER

[11:23:03.0000] <Hixie>
what Philip` describes is more or less what the whatwg file-bug.cgi script does

[11:23:04.0000] <Hixie>
using the ip address as the "unique" identifier

[11:23:05.0000] <zewt>
i am not going to use a super secure password on a bug tracker that i have no special permissions on. heh

[11:24:00.0000] <Philip`>
/me has several IP addresses per day, so that's not a very effective identifier

[11:25:00.0000] <Philip`>
(For a while I appended my initials to the spec bug messages, to make them easy to search for afterwards)

[11:26:00.0000] <Hixie>
hmmm

[11:26:01.0000] <Hixie>
should <dialog> have a close box by default or should it be up to authors to do that

[11:27:00.0000] <Hixie>
(s/by default/added by teh UA if the author asks for one/)

[11:27:01.0000] <timeless>
TabAtkins: that doesn't work

[11:27:02.0000] <Hixie>
i can't work out how we'd style it

[11:27:03.0000] <timeless>
we only want :has(attachment > 100mb)

[11:27:04.0000] <Hixie>
so i guess no close box by default!

[11:28:00.0000] <hober>
agreed

[11:29:00.0000] <AryehGregor>
<dialog>?

[11:29:01.0000] <TabAtkins>
timeless: Unless you get a ton of attachments, you can manually look through them for sizes.

[11:29:02.0000] <hober>
AryehGregor: see the Dialogs page on the WHATWG wiki

[11:29:03.0000] <timeless>
TabAtkins: the users i was looking at have attachments on virtually all messages

[11:30:00.0000] <hober>
/me has until some time tomorrow to write this CP

[11:30:01.0000] <timeless>
(perhaps many have a business card icon)

[11:30:02.0000] <TabAtkins>
timeless: That would be a problem, then.

[11:30:03.0000] <zewt>
searching for has:attachment gives me tons of noise

[11:30:04.0000] <timeless>
so yeah, :has(attachment) is useless

[11:30:05.0000] <zewt>
"smime.p7s" crap

[11:30:06.0000] <timeless>
yeah, either greeting cards or signatures

[11:30:07.0000] <zewt>
"vcf" files

[11:30:08.0000] <Hixie>
hober: pretty much entirely coincidentally, i plan to be working on this today (the mass move of bugs mikesmith did made this the least-recently changed bug!)

[11:30:09.0000] <TabAtkins>
timeless: No, it works great for me.  It may not work for everyone.

[11:30:10.0000] <timeless>
s/greeting/business/

[11:30:11.0000] <timeless>
TabAtkins: oh, it would work fine for timeless@gmail

[11:31:00.0000] <Hixie>
hober: so if you trust me, you can write a CP that just says "do what the whatwg spec does" and i'll try to write it up today :-)

[11:31:01.0000] <timeless>
but that's because 99.999% of mail there is bugmail :)

[11:31:02.0000] <timeless>
a simpler rule is -bugzilla-daemon

[11:31:03.0000] <hober>
Hixie: heh

[11:31:04.0000] <timeless>
that rule gets me anything that might be big :)

[11:31:05.0000] <Hixie>
hober: you can see what i'm thinking of doing in the Ideas section of that page

[11:31:06.0000] <hober>
Hixie: I can pop up there this afternoon for some whiteboarding if you think that would be useful

[11:31:07.0000] <Hixie>
hober: obviously, input is more than welcome

[11:31:08.0000] <hober>
Hixie: I think I have less ambitious goals than you in this case

[11:32:00.0000] <hober>
Hixie: i'm not sure if we should even try addressing the 'fat' tooltip case

[11:32:01.0000] <hober>
Hixie: at least not yet

[11:32:02.0000] <Hixie>
hober: i'm free 2 to 4 if you want, definitely happy to meet up if you want

[11:32:03.0000] <Hixie>
yeah i dunno if it should all go in at once

[11:32:04.0000] <Hixie>
i like to overdesign then cut out, makes me more confident the design can support extension later

[11:32:05.0000] <timeless>
/me grumbles

[11:32:06.0000] <timeless>
http://www.visionmobile.com/research.php#OGI

[11:33:00.0000] <timeless>
they demand my email address

[11:33:01.0000] <Hixie>
s/if you want//g

[11:33:02.0000] <hober>
Hixie: how about I swing by around 2?

[11:34:00.0000] <Hixie>
hober: sounds good to me. easiest is to meet in the building 43 lobby.

[11:34:01.0000] <Hixie>
hober: (i'm meeting with a guy from cisco about webrtc at 1)

[11:35:00.0000] <hober>
ok, see you there/then

[11:35:01.0000] <Hixie>
cool

[11:37:00.0000] <Hixie>
/me ends up interacting with several web dialogs just in trying to book the meeting, heh

[11:38:00.0000] <timeless>
Hixie: this is why you shouldn't allow web dialogs on the web :)

[11:39:00.0000] <Hixie>
wow, i actually managed to book the meeting room closest to the lobby

[11:39:01.0000] <Hixie>
that's not normal

[11:40:00.0000] <Hixie>
probably means the building is being refurbished or something and they just forgot to tell me

[11:40:01.0000] <timeless>
heh

[11:40:02.0000] <timeless>
did you book for friday?

[11:40:03.0000] <timeless>
maybe everyone is going to a party

[11:41:00.0000] <Hixie>
pretty sure i booked for today, yeah

[11:42:00.0000] <Hixie>
the party isn't til 4, why do you think that's when i have to stop the meeting with hober? :-P

[11:42:01.0000] <timeless>
heh

[11:44:00.0000] <Hixie>
hmm

[11:44:01.0000] <Hixie>
some of these dialogs... i guess all of them are the fat tooltips, as hober calls them

[11:44:02.0000] <Hixie>
some of these tooltips open anchored to the cursor position.

[11:44:03.0000] <timeless>
?

[11:45:00.0000] <timeless>
grr

[11:45:01.0000] <timeless>
/me sighs

[11:45:02.0000] <Hixie>
maybe there should be a variant of show() that takes a MouseEvent object

[11:45:03.0000] <timeless>
every time i update my phone's os, Gmail loses its account data

[11:45:04.0000] <Hixie>
hmmmmm

[11:45:05.0000] <timeless>
and i get to reregister it

[11:46:00.0000] <Hixie>
i wonder what i meant by "have to handle showing an element that's descendant of display:none content; does that just not count?"

[11:47:00.0000] <Hixie>
should we use callbacks for dialog.showModal(), or require you to set an event handler?

[11:47:01.0000] <Hixie>
hmm

[11:48:00.0000] <hober>
sorry, meeting

[12:15:00.0000] <Hixie>
hmm, interesting

[12:15:01.0000] <Hixie>
there are some modal dialogs that aren't centered but are instead anchored

[12:15:02.0000] <Hixie>
hmm

[12:15:03.0000] <TabAtkins>
Hixie: I'm rewriting the element() function's text, and I'm not sure how the display of out-of-document elements should be talked about.

[12:15:04.0000] <TabAtkins>
I think it's actually HTML's responsibility to define that images, videos, and canvas can be displayed while out-of-document.

[12:17:00.0000] <hober>
Hixie: yes, you see login dialogs that are anchored to the 'log in' link/button

[12:18:00.0000] <TabAtkins>
hober, Hixie: That will be eventually covered by CSS Positioning.

[12:18:01.0000] <hober>
TabAtkins: indeed, and I will happily defer to that module when it's ready :)

[12:18:02.0000] <TabAtkins>
Yeah, just do something magical for now.  I'll review for reasonableness when necessary.

[12:19:00.0000] <Ms2ger>
hober, I was wondering whether you edited any specs already :)

[12:20:00.0000] <hober>
Ms2ger: I'm technically listed as an editor of the 2d transforms spec, only because i go in and fix spec bugs when working on the transforms test suite

[12:20:01.0000] <Ms2ger>
To me, it sounds like you want to do more :)

[12:20:02.0000] <Hixie>
TabAtkins: cool, thanks (Re positioning)

[12:20:03.0000] <hober>
Ms2ger: heh

[12:21:00.0000] <Hixie>
TabAtkins: about element(), why is <img> special but not, say, <iframe>?

[12:21:01.0000] <Hixie>
TabAtkins: or <svg>?

[12:21:02.0000] <Hixie>
TabAtkins: i don't really understand html's role here. html defines what those elements represent, but beyond that...

[12:21:03.0000] <TabAtkins>
SVG defines its own stuff.

[12:21:04.0000] <Hixie>
TabAtkins: i can have a hook similar to "represent" for this case if you like

[12:22:00.0000] <Hixie>
TabAtkins: and you can use that

[12:22:01.0000] <TabAtkins>
But the host language is the only one that knows whether an element has a "natural" notion of what it measn to be rendered, outside the confines of a document.

[12:22:02.0000] <Hixie>
TabAtkins: "any element that foobars a bitmap..."

[12:22:03.0000] <Hixie>
TabAtkins: and the HTML spec can say "An img element foobars its image" or whatever

[12:23:00.0000] <TabAtkins>
Hmm, yeah, that would work.

[12:23:01.0000] <TabAtkins>
I just need a notion of width, height, and appearance.

[12:25:00.0000] <TabAtkins>
Hixie: Or if it just represents a raster image directly, I can pull what else I need from that.

[12:26:00.0000] <Hixie>
appearance?

[12:26:01.0000] <TabAtkins>
Like, an <img> looks like it's image.

[12:28:00.0000] <gsnedders>
People can live with thereshouldbenored.com being down for a bit? Good. :)

[12:28:01.0000] <timeless>
grrr

[12:28:02.0000] <Hixie>
as opposed to it being what?

[12:28:03.0000] <timeless>
i think google accounts is crashing my browser

[12:28:04.0000] <Hixie>
oh you mean "its image"

[12:29:00.0000] <Hixie>
wow it's not usually confusing to get that one wrong :-P

[12:29:01.0000] <Hixie>
i think all the ones we care about here are bitmaps right?

[12:29:02.0000] <Hixie>
canvas, video, and img

[12:29:03.0000] <TabAtkins>
Shit, sorry.  Damn you, apostrophe-omiting rules!

[12:29:04.0000] <TabAtkins>
Yeah, think so.

[12:29:05.0000] <TabAtkins>
"expose a bitmap" might work.

[12:30:00.0000] <Hixie>
file a bug about me adding that term

[12:30:01.0000] <TabAtkins>
kk

[12:30:02.0000] <Hixie>
link to this log :-)

[12:30:03.0000] <Hixie>
in other news i added a bunch of screenshots from google products to http://wiki.whatwg.org/wiki/Dialogs

[12:31:00.0000] <TabAtkins>
What component do I put this in?

[12:31:01.0000] <timeless>
grr

[12:31:02.0000] <timeless>
/me kicks something

[12:32:00.0000] <timeless>
Hixie: you misspelled Gmail

[12:32:01.0000] <timeless>
s/GMail/Gmail/

[12:32:02.0000] <timeless>
> GMail - Click "more" on the left, then Create New Label.

[12:32:03.0000] <TabAtkins>
s/.*/gmail/

[12:33:00.0000] <Hixie>
TabAtkins: other hixie drafts

[12:39:00.0000] <Ms2ger>
/me omites an apostrophe at TabAtkins 

[12:39:01.0000] <Ms2ger>
Rhymes with "smites"

[12:45:00.0000] <Hixie>
i think http://wiki.whatwg.org/wiki/Dialogs#Proposal covers the bulk of the use cases

[12:49:00.0000] <rillian>
where's the repository for the webvtt spec?

[12:50:00.0000] <Ms2ger>
rillian, svn.whatwg.org/webapps

[12:50:01.0000] <Ms2ger>
It's somewhere in source

[12:50:02.0000] <TabAtkins>
Ms2ger: The double-consonant rules are inconsistent in English.

[12:50:03.0000] <TabAtkins>
Otherwise we'd say "editting".  I sometimes autocorrect away from doubles, precisely because of the word "editting".

[12:51:00.0000] <Ms2ger>
Everything is inconssistent in English

[12:52:00.0000] <AryehGregor>
TabAtkins, a reasonably reliable rule is that you double the last consonant if the word is accented on the last syllable, and don't double it otherwise.

[12:52:01.0000] <AryehGregor>
For when you're adding verb suffixes like -ed or -ing.

[12:52:02.0000] <TabAtkins>
Hmm, never thought of that rule.

[12:52:03.0000] <AryehGregor>
Not entirely reliable, but works okay.

[12:53:00.0000] <AryehGregor>
I think I know it because in a MUD in like 1999, I reported a typo in one of the room descriptions where it said "worshiped", which I thought should be "worshipped", and one of the GMs educated me.

[12:53:01.0000] <AryehGregor>
(I then promptly pointed out another room description where it said "worshipped" instead of "worshiped", which he corrected.)

[12:54:00.0000] <AryehGregor>
(MUDs are cool.  The GMs can just do any random thing by typing commands, since they don't need to provide graphics.)

[12:54:01.0000] <TabAtkins>
Yeah, used to play one.

[12:55:00.0000] <rillian>
Ms2ger, around line 34836. thanks!

[12:56:00.0000] <Ms2ger>
Np

[12:59:00.0000] <Ms2ger>
http://xkcd.com/934/

[12:59:01.0000] <Ms2ger>
Go us

[13:00:00.0000] <AryehGregor>
\o/

[13:00:01.0000] <TabAtkins>
Wait, "someday" have xmonad?  We already have linux in the browser in linux.

[13:00:02.0000] <rillian>
*is* there an xmonad firefox extension?

[13:00:03.0000] <gsnedders>
TabAtkins: Only on little endian hardware, because typed arrays leak harware endianness.

[13:01:00.0000] <kbrosnan>
rillian: https://addons.mozilla.org/en-US/firefox/addon/pentadactyl/ if you are cool with vim

[13:04:00.0000] <rillian>
that one's cute

[13:07:00.0000] <Hixie>
if anyone has a few minutes to review http://wiki.whatwg.org/wiki/Dialogs#Proposal before hober and i discuss it this afternoon that'd be great

[13:07:01.0000] <Hixie>
bbiab

[13:08:00.0000] <smaug____>
I hope that isn't going to the spec before it gets some reviewing

[13:08:01.0000] <AryehGregor>
It can be reviewed once it's in the spec, no?  The WHATWG spec has stability annotations for this reason.

[13:11:00.0000] <smaug____>
what are the stability levels

[13:12:00.0000] <Ms2ger>
The boxes in the left margin that all say "Last Call"

[13:12:01.0000] <smaug____>
I mean, what all can read there

[13:12:02.0000] <smaug____>
"Ready for first implementations" in one

[13:13:00.0000] <AryehGregor>
smaug____, e.g.: http://www.whatwg.org/specs/web-apps/current-work/multipage/video-conferencing-and-peer-to-peer-communication.html#the-data-stream

[13:13:01.0000] <AryehGregor>
"Experimental draft"

[13:13:02.0000] <smaug____>
are those "levels" explained somewhere?

[13:13:03.0000] <AryehGregor>
No, they're pretty informal.

[13:13:04.0000] <AryehGregor>
I think anyone's allowed to change them.

[13:14:00.0000] <Ms2ger>
Yeah

[13:14:01.0000] <smaug____>
there seems to be for example "Ready for first implementations" stuff which is quite stable, and things which will quite likely to change

[13:14:02.0000] <AryehGregor>
They're meant to be just warnings like "don't assume this is stable and you should start implementing it right away".

[13:16:00.0000] <smaug____>
I mean some "Ready for first implementations" is stable, some "Ready for first implementations" is unstable

[13:16:01.0000] <timeless>
/me misses muds

[13:16:02.0000] <timeless>
/me played one in college

[13:16:03.0000] <smaug____>
so it is hard to interpret stability annotations

[13:17:00.0000] <AryehGregor>
You can ask if you're unsure.

[13:17:01.0000] <Philip`>
AryehGregor: It is worshipped, at least in proper English

[13:17:02.0000] <Philip`>
I think http://www.whatwg.org/specs/web-apps/current-work/status-documentation.html defines the stability markers

[13:19:00.0000] <smaug____>
AryehGregor: ask who? If no one has implemented the feature, it is hard to say much about its stability ;)

[13:19:01.0000] <AryehGregor>
smaug____, ask the editor.

[13:19:02.0000] <Ms2ger>
If no one has implemented it, it's unstable :)

[13:19:03.0000] <AryehGregor>
He'll know if he's gotten feedback about it, etc.

[13:20:00.0000] <AryehGregor>
Also, what Ms2ger said.

[13:20:01.0000] <Ms2ger>
At least, I think that's what smaug____ would say ;)

[13:21:00.0000] <timeless>
/me sighs

[13:21:01.0000] <timeless>
hsivonen: it's only 10:30pm where you are... maybe you're awake? :)

[13:21:02.0000] <timeless>
/me needs someone to beat the obvious drum

[13:21:03.0000] <smaug____>
yeah, I'd say all the stuff which hasn't been implemented should be marked "unstable - likely to change" ;)

[13:22:00.0000] <smaug____>
timeless: I think hsivonen is about to start his vacation today or tomorrow

[13:22:01.0000] <timeless>
aww :(

[13:22:02.0000] <timeless>
i hope he enjoys it

[13:22:03.0000] <timeless>
but the world will suffer while he does

[13:22:04.0000] <timeless>
:)

[13:23:00.0000] <Ms2ger>
It won't be reminded of the evilness of doctypes

[13:23:01.0000] <AryehGregor>
smaug____, the annotations also say which browsers have implemented the features.

[13:24:00.0000] <timeless>
i have a bunch of people arguing for APIs to enable individual web  pages to do things which should be a device/platform detail

[13:24:01.0000] <timeless>
and i'm having trouble making the argument because i'm just one person and people have started ignoring me :(

[13:24:02.0000] <smaug____>
AryehGregor: yeah, those annotations are very useful

[13:24:03.0000] <Ms2ger>
If someone made them say that :)

[13:25:00.0000] <smaug____>
AryehGregor: but with my bad English, I can't say if "Work in progress" is more stable than "Experimental draft" for example

[13:26:00.0000] <AryehGregor>
smaug____, . . . to be honest, I'm not sure offhand either.  I think "experimental" is less stable.

[13:26:01.0000] <smaug____>
:)

[13:26:02.0000] <Ms2ger>
The entire spec is a WIP, duh

[13:26:03.0000] <timeless>
i think i'd agree w/ AryehGregor  fwiw

[13:27:00.0000] <timeless>
experimental should probably mean "someone wrote something and has no idea if it works

[13:27:01.0000] <timeless>
and they're trying to get people to experiment with it to see if it does"

[13:27:02.0000] <timeless>
whereas work in progress hopefully means "someone is writing something and iterating with an implementation"

[13:27:03.0000] <The_8472>
today i noticed that multiplebgs + gradients is way more flexible than borders (including border-images).

[13:28:00.0000] <smaug____>
timeless: in this case "Work in progress" certainly doesn't mean that something is implementing the feature

[13:29:00.0000] <timeless>
smaug____: which case is this?

[13:29:01.0000] <smaug____>
but in fact that there are new proposals coming which will replace that feature

[13:29:02.0000] <timeless>
the one where the editor is also making a js based impl

[13:29:03.0000] <smaug____>
Undo handling

[13:29:04.0000] <timeless>
is still an impl

[13:29:05.0000] <timeless>
i'm not sure i'd call undo handling a WIP

[13:29:06.0000] <timeless>
afaict those are just a bunch of proposals

[13:30:00.0000] <moo-_->
is there a way to save references to local files in local storage and retrieve them later on?

[13:30:01.0000] <timeless>
(am i wrong?)

[13:30:02.0000] <moo-_->
not the file data itself, I am talking about photos here

[13:31:00.0000] <timeless>
moo-_-: you want to save a file path?

[13:31:01.0000] <timeless>
so that you can later retrieve the then current underlying data?

[13:31:02.0000] <moo-_->
timeless: yes, for opening the file again if the page is refreshed

[13:31:03.0000] <timeless>
if the file moves, i presume you're happy to get an error?

[13:31:04.0000] <moo-_->
timeless: yes

[13:31:05.0000] <moo-_->
np

[13:31:06.0000] <timeless>
(i'm pretty sure the answer is no, fwiw)

[13:31:07.0000] <moo-_->
timeless: just not lose the user initiated file interaction

[13:31:08.0000] <smaug____>
I sure hope you can get access only to the filename, not to the full path

[13:32:00.0000] <timeless>
so, offhand, i claim it isn't a useful feature

[13:32:01.0000] <timeless>
just because i grant you access to a file on Jan 1

[13:32:02.0000] <timeless>
doesn't mean I'll remember i granted that access on Dec 12

[13:32:03.0000] <timeless>
if you happen to remember the filename and an abstract

[13:32:04.0000] <timeless>
then when you suggest I might want to provide that file again on Dec 12

[13:32:05.0000] <timeless>
i'll either be happy to do so, or decide i've changed my mind

[13:32:06.0000] <timeless>
and that option will make me happy

[13:35:00.0000] <timeless>
moo-_-: in short, if browsers can't make it easy for user's to safely and conveniently initate file interactions,

[13:35:01.0000] <timeless>
... that's a bug in browsers

[13:35:02.0000] <moo-_->
timeless: I am more worried about that references survive browser close/refresh

[13:35:03.0000] <moo-_->
window.URL.createObjectURL()

[13:35:04.0000] <timeless>
and will need to be fixed

[13:35:05.0000] <moo-_->
I need to know if I can reuse result of this call across page loads

[13:36:00.0000] <timeless>
> Browsers will release these automatically when the document is unloaded;

[13:36:01.0000] <timeless>
from MDN [https://developer.mozilla.org/en/DOM/window.URL.createObjectURL]

[13:36:02.0000] <timeless>
in short, i think it's pretty clear that you can't reuse the result across page loads

[13:37:00.0000] <moo-_->
hmmm

[13:37:01.0000] <timeless>
does my logic seem flawed?

[13:37:02.0000] <timeless>
(i need to pack up and go home, so please lemme know soon)

[13:37:03.0000] <timeless>
-- otherwise, have a good weekend

[13:39:00.0000] <moo-_->
http://dev.w3.org/2006/webapi/FileAPI/#lifeTime

[13:39:01.0000] <moo-_->
yep

[13:39:02.0000] <moo-_->
:<

[13:39:03.0000] <moo-_->
need to then upload images to the server first

[13:39:04.0000] <moo-_->
suuuuuuucks

[13:45:00.0000] <moo-_->
I was secretly hoping

[13:45:01.0000] <moo-_->
that if the user is uploading multiple photos

[13:45:02.0000] <moo-_->
he/she can close browser and continue later

[13:45:03.0000] <moo-_->
I wonder if Flash has something to work around the problem

[13:46:00.0000] <zewt>
need to improve file api + structured clone + history api to make that possible

[13:46:01.0000] <zewt>
it's already possible by spec i think, but nobody supports all that together...

[13:46:02.0000] <timeless>
moo-_-: do you understand my argument against that?

[13:46:03.0000] <timeless>
yes, it means that it'd be hard to write a BitTorrent client

[13:47:00.0000] <jgraham>
We should replace the status markers with a simple 1-5 numeric system where 1 means "I pity the fool who implements this" and 5 means "this is stable enough to put on a space mission"

[13:47:01.0000] <zewt>
apis should definitely allow continuing to use files later

[13:47:02.0000] <timeless>
jgraham: how often would we get to 3?

[13:47:03.0000] <moo-_->
timeless: would there be problem to have origin or page URL and having object URLs valid within this origin?

[13:47:04.0000] <jgraham>
Well none of the spec would be a 5 yet, obviously

[13:47:05.0000] <timeless>
moo-_-: ?

[13:47:06.0000] <timeless>
moo-_-: your original question was about local resources, right?

[13:48:00.0000] <zewt>
it's tricky to implement correctly; I think the rule Chrome *tries* to follow is that you can continue to use the file so long as the underlying file hasn't been changed (but I don't see how it can efficiently enforce that), and I'm sure it still doesn't support structured cloning to persistent APIs

[13:48:01.0000] <timeless>
(file:)

[13:48:02.0000] <moo-_->
timeless: so that object URLs generated from files would be bound to page itself, not document session

[13:48:03.0000] <jgraham>
I presume space missions have now come around to the idea of silicon-based-RAM and such things, but the internet is still about a decade too new

[13:48:04.0000] <moo-_->
so if you visit the page again you can access the files previously opened

[13:48:05.0000] <zewt>
for example, it's very annoying that you can't stash open files in History, so that if the session is restored a web app can reopen open files and continue the state it was in

[13:48:06.0000] <timeless>
zewt: i claim that even a file which hasn't changed might turn out to be a problem if you wait long enough

[13:49:00.0000] <timeless>
it could turn out that it has stuff which was reclassified as confidential

[13:49:01.0000] <timeless>
i'd rather the page upon a resume cause the browser to show the user the list of previously authorized objects

[13:49:02.0000] <timeless>
and allow the user to control whether to allow access

[13:49:03.0000] <zewt>
timeless: while I understand the direction you're looking in, I'm not (from the cases I've heard so far) inclined to say that it's enough of a worry to outweigh the uses

[13:50:00.0000] <zewt>
clients are always free to say "restoring this page will continue granting access to these resources, is that OK?", though I'd be surprised if any actually did

[13:50:01.0000] <moo-_->
zewt: the spec says they MUST not http://dev.w3.org/2006/webapi/FileAPI/#lifeTime :)

[13:51:00.0000] <zewt>
moo: not talking about URIs, talking about structured clones of blobs and (more specifically) File

[13:51:01.0000] <moo-_->
zewt: aha

[13:51:02.0000] <moo-_->
File object itself?

[13:51:03.0000] <zewt>
you can store structured clones of objects with apis like indexeddb and History

[13:51:04.0000] <zewt>
and structured clone supports Blob and File

[13:52:00.0000] <timeless>
zewt: so...

[13:52:01.0000] <timeless>
my general feeling is that while I don't support the EUs stupid privacy review stuff (see recent spam)

[13:52:02.0000] <timeless>
is that specs need to be written to favor user's privacy and protection

[13:52:03.0000] <zewt>
(not sure what you're referring to in particular)

[13:54:00.0000] <timeless>
http://www.enisa.europa.eu/act/application-security/web-security/a-security-analysis-of-next-generation-web-standards/

[13:54:01.0000] <timeless>
for lack of a better url

[13:54:02.0000] <timeless>
they triggered a bunch of useless spam to various WGs

[13:54:03.0000] <zewt>
i don't disagree that privacy is important, but I don't have a feeling that this will cause significant real-world issues; and environments which have stronger, stricter requirements (NDAs, corporate policy, government), browsers are free and should be encouraged to expose stricter modes

[13:54:04.0000] <timeless>
often about specs which were nearly dead

[13:54:05.0000] <zewt>
such as like I said, warning before re-granting permissions after a session restore

[13:56:00.0000] <timeless>
zewt: note that i'm not advocating 1990s java style security dialogs

[13:56:01.0000] <zewt>
right

[13:56:02.0000] <zewt>
i'm also not meaning to brush off what you're saying--squinting at how things can go wrong is security 101

[13:57:00.0000] <timeless>
i think chrome's download approach is possible the right way

[13:57:01.0000] <timeless>
basically for any object that's exposed, stick it in the tray with a red color instead of green and an arrow going the other way

[13:57:02.0000] <timeless>
if the use doesn't care about the notification, they can click a hide button at the edge of the tray

[13:58:00.0000] <timeless>
but the problem is that you almost have to let the user interact with the tray before the page is allowed to touch the objects

[13:58:01.0000] <timeless>
otherwise a page is likely to steal all resources before the user makes any decisions

[13:58:02.0000] <timeless>
(which doesn't help the user)

[13:58:03.0000] <zewt>
well, it would discourage leaving permissions active in the first place

[13:58:04.0000] <zewt>
you'd be unlikely to leave a site having access to a file for a year and forgetting about it

[13:58:05.0000] <timeless>
you can do it by having the tray focus when the first resource is accessed by the page

[13:59:00.0000] <timeless>
zewt: depends on whether i visit the site often

[13:59:01.0000] <timeless>
fwiw, i visited gmail recently and got a bar at the top...

[14:00:00.0000] <zewt>
bear in mind, though, that the site can always stash the file somewhere when you first grant access

[14:00:01.0000] <zewt>
if i open a file today, close the site and come back in a month, the site may already (the first time) read it from disk and saved the raw data somewhere else

[14:00:02.0000] <timeless>
http://www.google.com/support/forum/p/gmail/thread?tid=141cc0ce0197cfba&hl=en

[14:01:00.0000] <timeless>
zewt: sure

[14:01:01.0000] <timeless>
but various things to consider

[14:01:02.0000] <timeless>
- the site http://site.example.com/ might have been run by a trusted vendor on Jan 1

[14:01:03.0000] <timeless>
the domain might have expired and might not be run by an evil guy on Dec 12

[14:01:04.0000] <zewt>
well, that's a more broad question

[14:02:00.0000] <timeless>
as an example: http://398west.ca/

[14:02:01.0000] <zewt>
it applies to every sensitive API and so far nobody has come up with an answer, really

[14:02:02.0000] <timeless>
Google still thinks (in Google Maps), that 398 West exists

[14:02:03.0000] <zewt>
(the only possibly-workable approach so far is packaged web apps, and that sucks, badly)

[14:02:04.0000] <timeless>
actually, those are more vulnerable, not less :)

[14:03:00.0000] <timeless>
If the site doesn't properly use HTTPS certs, it's easy for it to communicate to the replacement server

[14:03:01.0000] <zewt>
those you can at least grant permissions to as a unit and know when they change ... but they go fundamentally against the web

[14:04:00.0000] <timeless>
anyway...

[14:04:01.0000] <timeless>
i'm not opposed to a way for a browser to provide a token which is `potentially valid`

[14:04:02.0000] <timeless>
which a site could try to use in the future

[14:05:00.0000] <timeless>
which the user agent would then show the user a request for a resource with the option to select the previously selected resource and a note indicating when it was selected

[14:05:01.0000] <timeless>
in the case where the user wants to continue to share the resource, the cost isn't particularly high

[14:05:02.0000] <timeless>
in the case where the user wants to provide a different resource (e.g. a newer draft of a document)

[14:05:03.0000] <timeless>
the cost isn't particularly high

[14:05:04.0000] <zewt>
that applies to every API that's sensitive enough to require permission, even eg. geolocation

[14:05:05.0000] <timeless>
indeed :)

[14:06:00.0000] <zewt>
(geolocation being something that most people don't consider particularly sensitive, but some people very strongly do)

[14:06:01.0000] <timeless>
and if the user doesn't  want to share anything anymore, the user can do something useful

[14:06:02.0000] <timeless>
yeah

[14:06:03.0000] <timeless>
enisa thing about geolocation

[14:06:04.0000] <timeless>
it was hilarious

[14:06:05.0000] <zewt>
anyway, my point there is that yeah, having the site change out from under you is an attack to worry about, but it's a broader problem that should probably be solved in the broader scope, not specifically for Files (or any other single API)

[14:06:06.0000] <timeless>
they provided a script to do a binary search to find out the user's last geolocation timestamp

[14:07:00.0000] <timeless>
the response was "or you could just get the property from the object directly"

[14:07:01.0000] <timeless>
/me should find that one

[14:07:02.0000] <zewt>
that's not to say I have any particularly good ideas for doing so (that scale to lots of APIs and lots of sites and different types of users)

[14:07:03.0000] <timeless>
http://lists.w3.org/Archives/Public/public-geolocation/2011Aug/0003.html

[14:07:04.0000] <timeless>
is the thread

[14:12:00.0000] <zewt>
heh gmail is still doing the "message may not have been sent by" nonsense

[14:12:01.0000] <zewt>
very poor that they're leaving such a horrible user-mistraining bug unfixed for so long...

[14:12:02.0000] <timeless>
?

[14:12:03.0000] <timeless>
what's wrong w/ that?

[14:12:04.0000] <zewt>
This message may not have been sent by: timeless⊙gc Learn more Report phishing

[14:12:05.0000] <zewt>
showing false positive warnings trains users to ignore warnings

[14:12:06.0000] <timeless>
zewt: well, yeah, err

[14:12:07.0000] <timeless>
it isn't entirely wrong

[14:12:08.0000] <zewt>
I certainly never pay attention to those warnings since every one I've seen so far has been nonsense

[14:12:09.0000] <Philip`>
It used to say it for most mail sent by @google.com people to the WHATWG list

[14:13:00.0000] <Philip`>
Now it says it for many @gmail.com people

[14:13:01.0000] <timeless>
in the case of timeless⊙gc, my mail isn't sent by timeless⊙gc

[14:13:02.0000] <timeless>
it's sent by timeless....⊙gc

[14:13:03.0000] <timeless>
which list are you reading?

[14:13:04.0000] <zewt>
that one's on whatwg

[14:14:00.0000] <timeless>
most likely the list you're reading a list which is not properly dealing w/ magical header things

[14:14:01.0000] <timeless>
i have a domain @last.com which sends to flast⊙gc

[14:14:02.0000] <timeless>
it doesn't do the right thing, so any mail to first⊙lc which is forwarded to flast⊙gc properly gets a warning from gmail

[14:14:03.0000] <timeless>
because the forwarder is broken

[14:16:00.0000] <timeless>
for namecheap, it looks like the solution is to sign up for gapps, and they even wrote a howto: http://community.namecheap.com/forums/viewtopic.php?f=6&t=577

[14:16:01.0000] <timeless>
/me should consider it

[14:16:02.0000] <zewt>
i've used gapps for a long time

[14:17:00.0000] <zewt>
works well for the google products that just don't work with it (like g+)

[14:17:01.0000] <timeless>
right, so SPF records / DKIM signatures.

[14:17:02.0000] <timeless>
yeah yeah

[14:17:03.0000] <timeless>
i'd actually like a list from you of what works some day

[14:17:04.0000] <timeless>
but i need to shave and go home

[14:17:05.0000] <timeless>
(i don't have any mirrors at home yet, so ...)

[14:17:06.0000] <zewt>
g+ is the only thing i've hit that doesn't these days

[14:17:07.0000] <zewt>
they merged auth for google and gapps which fixed a lot of stuff

[14:19:00.0000] <zewt>
used to be some bizarre design bug that caused you two end up with two accounts with the same address, that took them forever to clear up, heh

[14:20:00.0000] <timeless>
does the whatwg mailing list add a footer to the message?

[14:20:01.0000] <timeless>
if it does, that's probably the problem

[14:20:02.0000] <zewt>
though i forward other addresses to it and they work fine for me

[14:20:03.0000] <zewt>
(though I don't post to lists often with my other addresses so)

[14:21:00.0000] <zewt>
nope

[14:21:01.0000] <timeless>
/me ponders

[14:21:02.0000] <timeless>
could you possibly pastebin the source of one of my bad messages?

[14:21:03.0000] <zewt>
one of the dangers of google; when something goes wrong, you sort of have to figure it out yourself :)

[14:21:04.0000] <timeless>
/me is curious

[14:21:05.0000] <timeless>
oh, i have a general idea of what's going wrong

[14:22:00.0000] <timeless>
it's either SPF or DKIM sig being voided

[14:22:01.0000] <timeless>
my bet is the latter, but ..

[14:22:02.0000] <zewt>
http://pastebin.com/6N3y9wA3

[14:23:00.0000] <zewt>
http://mail.google.com/support/bin/answer.py?hl=en&ctx=mail&answer=185812 sounds like it

[14:24:00.0000] <timeless>
h=mime-version:sender:in-reply-to:references:date         :x-google-sender-auth:message-id:subject:from:to:content-type         :content-transfer-encoding;

[14:24:01.0000] <timeless>
is what was hashed

[14:24:02.0000] <timeless>
Sender: whatwg-bounces⊙lwo

[14:24:03.0000] <timeless>
is what was voided

[14:24:04.0000] <timeless>
so yeah.

[14:24:05.0000] <timeless>
sadly i'm not sure what the correct fix is

[14:24:06.0000] <timeless>
the namecheap best solution is to change the mx

[14:24:07.0000] <timeless>
that doesn't work for whatwg of course

[14:25:00.0000] <zewt>
first impression is that Sender shouldn't be authenticated at all, but it's been a long time since I've dug very deeply into smtp

[14:25:01.0000] <timeless>
zewt: well, the signature clearly lists it as one of the fields they're signing

[14:25:02.0000] <timeless>
so you can't do that

[14:25:03.0000] <zewt>
right, I'm saying the signature shouldn't include that, since it's (probably) normal for mailing lists to change it

[14:25:04.0000] <timeless>
a dkim aware mailinglist will need to avoid munging any dkim signed fields

[14:26:00.0000] <timeless>
from google's perspective, the mailinglist is at fault

[14:26:01.0000] <timeless>
from my perspective, i'd like to know what a mailing list can do instead

[14:26:02.0000] <timeless>
sadly finding *that* info is painful

[14:26:03.0000] <timeless>
one approach is to take my message and stick it as a MIME attachment

[14:26:04.0000] <timeless>
(as the only part)

[14:26:05.0000] <zewt>
if mailing lists typically change Sender, then google's signatures are clearly at fault (i don't know, off-hand, whether changing that header is actually typical or if whatwg is doing something unusual)

[14:26:06.0000] <timeless>
and stick a <mailinglist> envelope around it

[14:28:00.0000] <zewt>
if every mailman installation does that, then google is clearly at fault in practice

[14:28:01.0000] <timeless>
i think mailman basically needs to change

[14:28:02.0000] <zewt>
(i'm assuming whatwg is just a standard-issue mailman install like the other billion of them)

[14:28:03.0000] <timeless>
it needs to strip out the DKIM signature if it's valid

[14:28:04.0000] <timeless>
and replace it with its own DKIM signature

[14:28:05.0000] <timeless>
which is correct

[14:28:06.0000] <timeless>
it's claiming sender=whatwg

[14:29:00.0000] <timeless>
so it should just resign it with sender=whatwg

[14:29:01.0000] <timeless>
http://wiki.list.org/display/DEV/DKIM

[14:30:00.0000] <zewt>
timeless: but practically speaking, if every mailman list in the world has this problem, then google is in error showing a warning like this before the problem is solved (or at least in the process of being solved, eg. a mailman update available)

[14:30:01.0000] <timeless>
well

[14:30:02.0000] <zewt>
(i'm assuming there isn't, since it looks like whatwg is already on the latest version)

[14:30:03.0000] <timeless>
could i interest you in working on it w/ me?

[14:30:04.0000] <timeless>
i'm seriously contemplating trying to fix mailman

[14:31:00.0000] <timeless>
i'm fairly certain that no one has fixed it

[14:31:01.0000] <timeless>
by `fixing`, i mean `writing a patch which implements one strategy`

[14:31:02.0000] <timeless>
my proposed strategy is this:

[14:31:03.0000] <timeless>
take the existing signature, if it's valid, store it with a new prefix, munge the message, and resign it for the mailer

[14:32:00.0000] <timeless>
if the signature isn't valid, add a note that it wasn't valid and <dunno, possibly just drop the message on the floor, or optionally send it to moderation, or munge the message and add a broken signature and send it off>

[14:33:00.0000] <timeless>
i'd do the code under the mailman license (whatever that is), and offer it to mailman and any lists i use

[14:34:00.0000] <timeless>
(yes, this has bothered me for a while, but the goal of google here is to get people to implement a fix)

[14:34:01.0000] <timeless>
i think ideally google would have a way to let individuals whitelist individual mailing lists in the interim

[14:34:02.0000] <timeless>
until people like you/me fix the mailing list software to stop breaking signatures

[14:34:03.0000] <zewt>
i'm not familiar with the nuts and bolts of how DKIM works

[14:34:04.0000] <timeless>
zewt: don't worry about it, we'd figure it out

[14:35:00.0000] <timeless>
the documentation seems pretty clear

[14:35:01.0000] <timeless>
and it seems fairly straightforward

[14:36:00.0000] <timeless>
fwiw, the last link i offered indicates that mailman 2.1.9 strips DKIM

[14:36:01.0000] <zewt>
but then that it was reverted by default

[14:37:00.0000] <zewt>
it gives me very low confidence in DKIM that it's being deployed without how it affects mailing lists being very clearly defined and understood

[14:37:01.0000] <timeless>
i think your average user probably doesn't use mailing lists :)

[14:37:02.0000] <zewt>
(which it clearly isn't, given this big list of possible options)

[14:37:03.0000] <timeless>
well, it means that in theory at the risk of using whatever security bugs are in 2.1.9, someone could use it, or flip the toggle

[14:39:00.0000] <timeless>
ok, so mailman is python :)

[14:40:00.0000] <timeless>
http://hewgill.com/pydkim/

[14:40:01.0000] <timeless>
seems like there's a happy library for it

[14:40:02.0000] <timeless>
offhand, it seems like this shouldn't be too hard

[14:41:00.0000] <timeless>
ignoring politics

[14:41:01.0000] <zewt>
here's a mail from you on webapps, which doesn't trigger it: http://pastebin.com/ZNMwPvLK i forget which list software they use

[14:44:00.0000] <timeless>
the problem is the SPF record

[14:44:01.0000] <timeless>
w3 has one

[14:44:02.0000] <timeless>
whatwg does

[14:44:03.0000] <timeless>
n't

[14:44:04.0000] <timeless>
Received-SPF: neutral (google.com: 208.97.161.172 is neither permitted nor denied by best guess record for domain of whatwg-bounces⊙lwo) client-ip=208.97.161.172;

[14:44:05.0000] <timeless>
Received-SPF: pass (google.com: domain of public-webapps-request⊙lwo designates 128.30.52.56 as permitted sender) client-ip=128.30.52.56;

[14:44:06.0000] <timeless>
so whatwg could solve this by adding an SPF

[14:44:07.0000] <timeless>
that seems like a simple thing to fix

[14:44:08.0000] <timeless>
contact hixie or someone and ask them to add the necessary spice

[14:44:09.0000] <timeless>
http://old.openspf.org/dns.html

[14:44:10.0000] <timeless>
/me is amused to see `rim.net` in the examples

[14:46:00.0000] <timeless>
anyway, it'd be great if you poked Hixie about it

[14:46:01.0000] <timeless>
if not, i'll try to remember to do so

[14:47:00.0000] <zewt>
saying Hixie is probably enough to poke him about it :)

[14:48:00.0000] <timeless>
w3.org         text =          "v=spf1 a mx ptr mx:inrialpes.fr ip4:128.30.52.0/22 ip4:133.27.175.5 ip4:193.51.208.64/28 ip 4:212.89.8.176/29 ip4:212.89.8.80/28 mx:fundacionctic.org a:pec.etri.re.kr include:spf.keio.w3.org ~ all"

[14:48:01.0000] <timeless>
is the relevant part of the w3.org dns record :)

[14:48:02.0000] <timeless>
fwiw, dreamhost.com has a spf record

[14:49:00.0000] <timeless>
so if Hixie 's box is in dreamhost (Which

[14:49:01.0000] <timeless>
it probably is), and he's willing to let *anyone* in dreamhost send mail as whatwg

[14:49:02.0000] <zewt>
pretty sure the list is

[14:49:03.0000] <timeless>
he could use a lazy include instead of writing the right thing

[14:49:04.0000] <timeless>
(not a great idea, but...)

[14:50:00.0000] <timeless>
http://wiki.dreamhost.com/SPF

[14:50:01.0000] <timeless>
dreamhost even has a wiki page for it :)

[14:51:00.0000] <timeless>
right, in theory `v=spf1 include:dreamhost.com a mx ptr -all` would do the right thing

[14:51:01.0000] <timeless>
if you have a dreamhost account, you could test that pretty easily :)

[14:51:02.0000] <timeless>
/me really goes home (kinda)

[14:52:00.0000] <gsnedders>
(as in, the next room?)

[15:43:00.0000] <rniwa>
AryehGregor: yt?

[15:43:01.0000] <AryehGregor>
rniwa, yep.

[15:43:02.0000] <rniwa>
AryehGregor: what do you think of edit-action event?

[15:43:03.0000] <AryehGregor>
rniwa, not sure.  Does such a thing exist yet, or is this just a proposal?  What are the use-cases?

[15:43:04.0000] <rniwa>
AryehGregor: no

[15:44:00.0000] <rniwa>
AryehGregor: so, some editors implement their own editing commands

[15:44:01.0000] <AryehGregor>
Right.

[15:44:02.0000] <rniwa>
AryehGregor: or replace the UA's by their own

[15:44:03.0000] <rniwa>
AryehGregor: but there's currently no way of doing this

[15:44:04.0000] <rniwa>
AryehGregor: other than manually intercepting keydown

[15:44:05.0000] <rniwa>
AryehGregor: and other events

[15:45:00.0000] <AryehGregor>
So just an event that fires whenever execCommand() is run?

[15:45:01.0000] <AryehGregor>
Or, no.

[15:45:02.0000] <AryehGregor>
Are you talking about things like "user types text" or "user inserts a line break"?

[15:45:03.0000] <rniwa>
AryehGregor: yes

[15:45:04.0000] <AryehGregor>
Ah, okay.

[15:45:05.0000] <rniwa>
AryehGregor: we should fire an event for both

[15:45:06.0000] <rniwa>
AryehGregor: that's why I said user action

[15:45:07.0000] <AryehGregor>
Right, I see.

[15:45:08.0000] <rniwa>
AryehGregor: could be beforeEditingCommand

[15:45:09.0000] <rniwa>
AryehGregor: or beforeEditingAction

[15:45:10.0000] <rniwa>
AryehGregor: I mean the name could be anything

[15:46:00.0000] <rniwa>
AryehGregor: but the point is that we'll give some standard name for user editing actions

[15:46:01.0000] <AryehGregor>
And the default action would be to run insertText or insertParagraph or whatever.

[15:46:02.0000] <rniwa>
AryehGregor: e.g. inserting new line should be InsertLineBreak / InsertParagraph, etc...

[15:46:03.0000] <rniwa>
AryehGregor: right

[15:46:04.0000] <AryehGregor>
Why is this better than intercepting keydown etc.?

[15:46:05.0000] <rniwa>
AryehGregor: I mean I'm certain UA already does this one way or another

[15:46:06.0000] <rniwa>
AryehGregor: intercepting keydown is problematic because it depends on each platform

[15:47:00.0000] <rniwa>
AryehGregor: e.g. ctrl+y doesn't necessary mean undo

[15:47:01.0000] <AryehGregor>
Hmm, I see.

[15:47:02.0000] <rniwa>
AryehGregor: it's cmd+y

[15:47:03.0000] <AryehGregor>
Oh, right.

[15:47:04.0000] <AryehGregor>
Or Ctrl-Shift-Z.

[15:47:05.0000] <rniwa>
AryehGregor: yeah

[15:47:06.0000] <rniwa>
e.g.

[15:47:07.0000] <rniwa>
etc...

[15:47:08.0000] <AryehGregor>
Okay, that makes sense.

[15:47:09.0000] <AryehGregor>
Also for insertLineBreak.

[15:47:10.0000] <rniwa>
also, the user agent may provide some UI such as context menu that triggers editing commands

[15:48:00.0000] <rniwa>
e.g. Safari provides Format in its context enu

[15:48:01.0000] <rniwa>
menu*

[15:48:02.0000] <rniwa>
and lets user bold, italicize, etc.. .text

[15:48:03.0000] <AryehGregor>
Interesting.

[15:48:04.0000] <rniwa>
AryehGregor: I guess we can talk more when we meet later this month :)

[15:49:00.0000] <AryehGregor>
So it would make sense to just fire an event for every execCommand() invocation, whether internal to the browser or not.

[15:49:01.0000] <AryehGregor>
Sure.  :)

[15:49:02.0000] <rniwa>
AryehGregor: right

[15:50:00.0000] <rniwa>
AryehGregor: but I'm not sure if all browsers go through execCommand internally when executing user editing actions

[15:50:01.0000] <rniwa>
AryehGregor: they might do something slightly different when actions are user triggerd

[15:50:02.0000] <AryehGregor>
Right, that just has to be specced.

[15:50:03.0000] <rniwa>
yeah

[15:51:00.0000] <rniwa>
AryehGregor: but I don't think I should be doing that in my spec

[15:51:01.0000] <AryehGregor>
Right.

[15:51:02.0000] <rniwa>
AryehGregor: your spec is probably a better place to do

[15:51:03.0000] <rniwa>
AryehGregor: if you can give standarized names to each editing action

[15:51:04.0000] <rniwa>
AryehGregor: then I can refer to that in my spec

[15:51:05.0000] <rniwa>
AryehGregor: I really want to tie all of this to new mutation events replacement though

[15:52:00.0000] <AryehGregor>
Sounds good.

[15:52:01.0000] <rniwa>
it seems like the discussion on that thread is bikeshedding to some extent now

[15:52:02.0000] <rniwa>
but ideally, script should be able to get a list of mutations that happened for a given edit action

[15:54:00.0000] <AryehGregor>
Okay, go to go.  See you around.

[15:54:01.0000] <rniwa>
k

[15:54:02.0000] <rniwa>
ttyl


2011-08-06
[18:47:00.0000] <hober>
Hixie othermaciej TabAtkins & others who are interested in this sort of thing: here's the quick first pass on the <dialog modal> CP. I'll revise it tonight & tomorrow based on feedback; feel free to email or discuss it here. http://www.w3.org/html/wg/wiki/User:Eoconnor/ISSUE-133

[18:48:00.0000] <hober>
[or feel free to just make the CP better, for that matter; it's a wiki :)]

[19:13:00.0000] <othermaciej>
hober: neat

[19:15:00.0000] <othermaciej>
hober: I would advise against use of the diaeresis in "coordinated", this ain't the New Yorker

[19:17:00.0000] <othermaciej>
hober: it would be helpful to list what features of other libraries this doesn't cover

[19:19:00.0000] <Hixie>
hober: looks good on a first skim, i'll look more closely later if you mail me a link

[19:25:00.0000] <jamesr>
why is this an element and not a purely script-driven API? it doesn't look like there's any way to use this w/out script initialization, at least

[19:26:00.0000] <nimbu>
Hixie: I am trying to research how document outlines originated, and this is the earliest I could find a cogent reason for their existence: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-May/011481.html Do you recall when it was originally thought of (so I can trawl archives better)?

[19:26:01.0000] <jamesr>
hober, what happens if .show() is invoked when the page doesn't have focus?

[19:26:02.0000] <Hixie>
jamesr: well you need something to display

[19:27:00.0000] <jamesr>
Hixie, granted, but why does it have to be a specific element? why not any block element?

[19:28:00.0000] <Hixie>
why bother with <section>? or <blockquote>?

[19:28:01.0000] <Hixie>
same reason

[19:28:02.0000] <Hixie>
(ease of code maintenance/readability, semantics/accessibility, ease of interpretation by search engines or other tools, etc)

[19:29:00.0000] <zewt>
why does page focus matter? surely this would always be tab-modal

[19:29:01.0000] <zewt>
(not window-modal)

[19:36:00.0000] <jamesr>
zewt, maybe, but hober's proposal says that "the dialog is focused when displayed" which i don't think will really be true

[19:36:01.0000] <jamesr>
Hixie, valid, but <section> and <blockquote> have some meaning in the document of their own

[19:39:00.0000] <Hixie>
so does <dialog>

[19:39:01.0000] <Hixie>
it means "this is going to be used for a dialog"

[19:41:00.0000] <Hixie>
(more importantly, maybe, it's display:none by default, so in non-CSS UAs you don't have to worry about the dialog showing when not active)

[19:43:00.0000] <zewt>
jamesr: well, i'd expect it to be focused within the tab, but i wouldn't expect the tab to be focused within the browser (or the browser within the windowing system), no more than tab-modal alert()s are

[19:44:00.0000] <zewt>
dreaming of the day all window-modal dialogs are eliminated (probably never, file pickers are hard)

[01:50:00.0000] <annevk>
What are we adding to DOM Core?

[01:50:01.0000] <annevk>
nm

[02:25:00.0000] <annevk>
finally removed myself from HyBi

[02:28:00.0000] <annevk>
AryehGregor, not having lots of hooks for the CSS box model is indeed somewhat of a pain

[02:29:00.0000] <annevk>
s/somewhat of//

[02:43:00.0000] <annevk>
It seems I want to get out of happiana too

[04:46:00.0000] <annevk>
can't we set an auto-reply for public-html-comments that tells people to file a bug instead?

[04:46:01.0000] <annevk>
the current process is somewhat annoying in that people reply to emails but then those replies do not get captured in the corresponding bug reports

[04:46:02.0000] <annevk>
MikeSmith, ^^

[04:47:00.0000] <MikeSmith>
I guess we could, yeah

[04:50:00.0000] <Philip`>
Or just shut down the list and tell people to file bugs instead of telling them to post there first?

[04:52:00.0000] <annevk>
well yeah that would be part of it

[07:04:00.0000] <gsnedders>
Ms2ger: You have a testsuite for DOM Core somewhere, no?

[07:04:01.0000] <Ms2ger>
dvcs.w3.org/hg/webapps/DOMCore

[07:06:00.0000] <gsnedders>
How complete is that?

[07:09:00.0000] <Ms2ger>
Not at all

[07:09:01.0000] <Ms2ger>
More tests welcome

[07:27:00.0000] <gsnedders>
Ms2ger: Disclaimer: procrastination may happen.

[07:27:01.0000] <Ms2ger>
Same with me, so that's fine

[07:52:00.0000] <jarek>
Hi

[07:52:01.0000] <jarek>
is it possible to define custom DOM events?

[07:53:00.0000] <jarek>
I mean something like: "document.addEventListener('myCustomEvent', handler() {}, false)"

[07:53:01.0000] <jarek>
is it possible without creating wrapper function on top of document.addEventListener?

[07:55:00.0000] <annevk>
yes

[07:57:00.0000] <jarek>
annevk: what API should I be looking for?

[07:58:00.0000] <jarek>
I'm currently fire and assign custom events with bind() and trigger() functions, but I think in some cases it would be more elegant to use the same syntax as for core DOM events

[07:58:01.0000] <jarek>
s/I'm currently fire and assign/I'm currently firing and assigning

[07:58:02.0000] <jarek>
sorry for typos

[08:02:00.0000] <annevk>
you can do document.createEvent in current browsers

[08:02:01.0000] <annevk>
and target.dispatchEvent()

[08:03:00.0000] <jarek>
awesome, exactly what I needed

[08:30:00.0000] <jarek>
"linkElement.href" always returns absolute path while "linkElement.getAttribute('href')" may return either absolute or relative path

[08:30:01.0000] <jarek>
is this behavior the same on all browsers?

[08:33:00.0000] <jarek>
hmm... looks like it's by design as I can reproduce it on every browser

[08:41:00.0000] <smaug____>
jarek: getAttribute() returns the content attribute value

[08:41:01.0000] <smaug____>
so whatever value the attribute has

[08:42:00.0000] <smaug____>
.href returns what is specified in http://www.whatwg.org/specs/web-apps/current-work/multipage/semantics.html#attr-link-href

[08:43:00.0000] <smaug____>
oh, hmm

[08:43:01.0000] <smaug____>
is there a bug in the spec

[08:44:00.0000] <smaug____>
no

[08:44:01.0000] <smaug____>
"If a reflecting IDL attribute is a DOMString attribute whose content attribute is defined to contain a URL, then on getting, the IDL attribute must resolve the value of the content attribute relative to the element and return the resulting absolute URL"

[08:49:00.0000] <Ms2ger>
smaug____, no bug for once? :)

[09:14:00.0000] <smaug____>
Ms2ger: are you hinting that I occasionally complain about bugs in the HTML spec ;)

[09:14:01.0000] <Ms2ger>
Perhaps once in a while ;)

[09:24:00.0000] <smaug____>
hmm, not much feedback to the new mutation events replacement proposal. Perhaps I should just implement it and write down how the implementation works and call it a spec.

[09:24:01.0000] <Ms2ger>
WFM

[09:25:00.0000] <Ms2ger>
Can you provide it in a format I can c&p into DOM Core? :)

[09:27:00.0000] <smaug____>
Ms2ger: Btw, what kind of template or some such is used for DOM Core

[09:27:01.0000] <Ms2ger>
Plain old HTML

[09:28:00.0000] <Ms2ger>
With some magic for cross-referencing

[09:28:01.0000] <smaug____>
so no special tools which generate right kind of HTML?

[09:28:02.0000] <Ms2ger>
No

[09:28:03.0000] <Ms2ger>
See http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.src.html

[09:29:00.0000] <smaug____>
/me has used some old W3C XSLT templates which work quite ok

[09:29:01.0000] <Ms2ger>
/me runs for the hills

[09:40:00.0000] <Ms2ger>
    if (document.all == null) {

[09:40:01.0000] <Ms2ger>
Good morning, whatwg

[10:51:00.0000] <Hixie>
MikeSmith: if you're doing this bug filing manually anyway, it might be worth making your script give you a way to split e-mails into one-bug-per-issue, even if it's just having it ask you how many issues, and then it filing the same bug N times with an additional comment "This is for issue 3 of 6" or whatever

[10:54:00.0000] <zewt>
what the

[10:54:01.0000] <Hixie>
what?

[10:54:02.0000] <zewt>
oh

[10:55:00.0000] <zewt>
it looked like greader changed their external links to some broken javascripty thing ... but looks like an isolated bug

[11:03:00.0000] <annevk>
smaug____, maybe post it as separate thread with mark 2 or v2 attached to the subject? implementing it wfm though

[11:07:00.0000] <annevk>
www-style should really start recommending people to post use cases instead of feature requests. I mean: "I suggest to consider six new pseudo elements for the Generated Content module, although I’m not entirely sure they would be all that useful: ..."

[13:00:00.0000] <Ms2ger>
Philip`, :(

[13:02:00.0000] <krijnhuman>
/me does foo

[13:02:01.0000] <krijnhuman>
/me does bar

[13:03:00.0000] <Ms2ger>
/me waves at krijnhuman 

[13:03:01.0000] <krijnh>
zcorpan: color codes, apparently, turned them off now

[13:04:00.0000] <krijnh>
o/ Ms2ger

[13:11:00.0000] <jgraham>
Ms2ger: Syntax Error, expected } got EOF

[13:11:01.0000] <Ms2ger>
mm?

[13:12:00.0000] <jgraham>
You startd an if block some time ago

[13:12:01.0000] <Ms2ger>
I did

[13:12:02.0000] <Ms2ger>
}

[13:12:03.0000] <Ms2ger>
There you go

[13:12:04.0000] <jgraham>
Yay, now I can read the scrollback!

[13:14:00.0000] <Ms2ger>
You're welcome(

[13:14:01.0000] <annevk>
trololol

[13:14:02.0000] <gsnedders>
jgraham: You could improve your syntax error messages to note where the opening brace was.

[13:14:03.0000] <Ms2ger>
Better here than on public-webapps, I guess

[13:16:00.0000] <jgraham>
gsnedders: and kitchen design shops could provide price lists. Sadly it doesn't show any signs of happening.

[13:17:00.0000] <Ms2ger>
Getting settled?

[13:17:01.0000] <jgraham>
That remains to be seen

[13:18:00.0000] <gsnedders>
jgraham: Much packing left?

[13:18:01.0000] <jgraham>
gsnedders: Katie is insanely efficient.

[13:19:00.0000] <Ms2ger>
To the point where it's more efficient if you just stay out of her way? :)

[13:20:00.0000] <jgraham>
Erm, no comment? :)

[13:20:01.0000] <Ms2ger>
:)

[13:21:00.0000] <gsnedders>
Letting $girl do packing worked quite well when I moved a couple of months back

[13:22:00.0000] <jgraham>
gsnedders: Your girlfriend is variable?

[13:23:00.0000] <gsnedders>
jgraham: Well, I don't expect to have a single girlfriend for my entire life, so yes.

[13:23:01.0000] <gsnedders>
Otherwise breaking-up would be quite hard, if it had been declared as a constant.

[13:24:00.0000] <Ms2ger>
/me wonders if gsnedders->$girl found out about IRC logs already

[13:24:01.0000] <Hixie>
actually it'd be easier, you'd just overwrite the memory location where the value is kept and you'd instantly assign all previous memories to the new girl.

[13:24:02.0000] <jgraham>
Ah, so functional relationships are like functional languages; single assignment

[13:24:03.0000] <Hixie>
and the compiler would be none the wiser

[13:24:04.0000] <Ms2ger>
Hixie, also the reasons for breaking up? That seems suboptimal

[13:24:05.0000] <Hixie>
well i'm not saying it's optimal

[13:24:06.0000] <Hixie>
just that it wouldn't be hard

[13:24:07.0000] <jgraham>
(alternative worldview: functional relationships are like functional langauges: optimised for concurrency)

[13:24:08.0000] <gsnedders>
jgraham: in most functional langauges you don't assign per-se.

[13:25:00.0000] <annevk>
/me facepalms

[13:25:01.0000] <gsnedders>
Ms2ger: it is possible she has

[13:28:00.0000] <Hixie>
in other news, i have a computer with a dead disk. i can dd(1) to it all i want so long as i don't hit the bad parts, but as soon as i hit a bad bit, /dev/sda becomes completely inaccessible to any program. wtf.

[13:28:01.0000] <jgraham>
(I should maybe point out in the intereests of having a functional relationship that I don't subscribe to the lternative worldview)

[13:28:02.0000] <jgraham>
+A

[13:28:03.0000] <Hixie>
jgraham: ain't nothing wrong with polyamoury.

[13:29:00.0000] <Ms2ger>
Hixie, I await a long series of blog posts on your disk problems :)

[13:29:01.0000] <Hixie>
(so long as everyone involved is ok with it)

[13:29:02.0000] <Ms2ger>
It'll almost feel like 05

[13:29:03.0000] <Hixie>
Ms2ger: it's already begun :-P

[13:29:04.0000] <jgraham>
Hixie: I made no moral statements

[13:29:05.0000] <jgraham>
(and generally agree)

[13:31:00.0000] <gsnedders>
($girl = $girls->pop() then?)

[13:31:01.0000] <jgraham>
gsnedders: You keep your girls on a stack?

[13:32:00.0000] <gsnedders>
jgraham: Yes, otherwise the route needed to visit them all becomes needlessly wrong.

[13:32:01.0000] <gsnedders>
(wrong? long.)

[13:33:00.0000] <jgraham>
/me is very confused

[13:33:01.0000] <gsnedders>
jgraham: I make sure they are on stack in the order needed to minimise the distance to visit them sequentically

[13:34:00.0000] <Philip`>
It's annoying when you pop your girl and have to reinflate her

[13:34:01.0000] <gsnedders>
Philip`: yeah, girl v. girls is an easily made typo

[13:42:00.0000] <jarek>
why "blur" event fires when I click on iframe inside the window?

[13:43:00.0000] <jarek>
is it possible to prevent it from firing in this situation?

[13:47:00.0000] <jgraham>
Generally you can't stop events firing, but there is no real need to do anything with the ones you don't want

[13:47:01.0000] <jgraham>
Well you can preventDefault of course

[13:48:00.0000] <smaug____>
preventDefault() doesn't do anything with blur

[13:51:00.0000] <jarek>
jgraham: but I need to prevent this event only when iframe inside window is clicked, not when window loses focus to some other window

[13:51:01.0000] <jarek>
I have no idea how to do this

[13:51:02.0000] <jarek>
as blur event does not provide any information about what causes the blur

[13:55:00.0000] <jgraham>
jarek: I'm not sure what you are trying to do, but I don't think you can do what you are requesting

[13:56:00.0000] <jgraham>
Fun game: Read http://blog.n01se.net/?p=375 and work out how many of the listed specs are at W3C. Then subtract the ones that were imported wholesale into the W3C without changes

[13:56:01.0000] <jgraham>
(or almost without changes)


2011-08-07
[01:49:00.0000] <Zeddy>
Hey, i have everything else working in my offline webapp, except, i still haven't figured out how to reset a manifest file. The problem is i call my manifest file with parameters manifest.php?map=1, or manifest.php?map=2.. The problem is, it creates two separate caches and the amount of data stored is doubled

[01:49:01.0000] <Zeddy>
can i resolve this issue by using php session variables?

[02:21:00.0000] <SiPlus>
Hai

[02:21:01.0000] <SiPlus>
Any WHATWG members here?

[02:21:02.0000] <SiPlus>
especially websockets standard devs

[02:39:00.0000] <smaug____>
SiPlus: if you have questions, it is better to just ask. Someone may answer :)

[02:40:00.0000] <SiPlus>
Why no UDP Websockets?

[02:42:00.0000] <smaug____>
SiPlus: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13322

[02:42:01.0000] <SiPlus>
smaug____, this is my post XD

[02:43:00.0000] <smaug____>
:)

[02:43:01.0000] <smaug____>
SiPlus: for websockets I'd like to see first at least one stable protocol, before adding more

[02:44:00.0000] <smaug____>
and also, as of now, websocket is for reliable data transfer

[02:44:01.0000] <SiPlus>
smaug___: is GWT Quake 2 multiplayer compatible with C source ports?

[02:45:00.0000] <smaug____>
I don't know what is GWT Quake

[02:48:00.0000] <Philip`>
Raw TCP/UDP sockets will never be allowed (because of security) - you'll only be able to talk to servers that intentionally implement WebSockets (or whatever new protocols are developed) which won't include any existing game servers

[02:51:00.0000] <smaug____>
(I wouldn't say 'never'. Someone may come up with reasonable permission handling ...)

[02:53:00.0000] <Philip`>
(That sounds like it'll always be problematic since the only person you could ask for permission is the user, and they're not going to have a clue whether they're behind a firewall and would be exposing vulnerable services)

[02:55:00.0000] <smaug____>
I don't know how it should work, but something like same-IP-restriction + permission

[02:56:00.0000] <smaug____>
SiPlus: why do you want *WebSockets* to support UDP?

[02:57:00.0000] <SiPlus>
smaug___: for compatibility with C-written Quake clients

[02:58:00.0000] <smaug____>
WebSocket API is something which runs on top of websocket protocol

[02:58:01.0000] <SiPlus>
Philip` Raw TCP/UDP sockets will never be allowed (because of security) - you'll only be able to talk to servers that intentionally implement WebSockets (or whatever new protocols are developed) which won't include any existing game servers

[02:58:02.0000] <SiPlus>
HTML5 sucks

[02:58:03.0000] <smaug____>
and the protocol is defined in IETF

[02:58:04.0000] <smaug____>
if you have problems with the protocol, you could ask IETF hybi group if they have plans to change to protocol

[03:00:00.0000] <SiPlus>
/me has to find a security breach to enable multiplayer in WebQuake

[03:00:01.0000] <SiPlus>
Are there QuakeC parsing specs?

[03:01:00.0000] <smaug____>
SiPlus: there is at least one thread about UDP in hybi http://www.ietf.org/mail-archive/web/hybi/current/msg02100.html

[03:02:00.0000] <annevk>
SiPlus, you could write an intermediary server that translates the calls from one protocol to the other

[03:04:00.0000] <annevk>
hsivonen, I think Opera deals with that email format fairly well, but I'm not a 100% sure

[03:07:00.0000] <SiPlus>
How can I draw dots in 3D space using WebGL?

[03:07:01.0000] <SiPlus>
no

[03:07:02.0000] <SiPlus>
not dots

[03:07:03.0000] <SiPlus>
I don't need dots

[03:07:04.0000] <SiPlus>
Lines

[03:08:00.0000] <SiPlus>
I don't see anything in my canvas

[03:18:00.0000] <smaug____>
Hmm, Websocket API draft points to an ancient unsafe protocol draft :/

[03:28:00.0000] <annevk>
it does?

[03:28:01.0000] <annevk>
seems to refer to the latest draft

[03:37:00.0000] <mhausenblas>
annevk, got a minute re CORS?

[03:39:00.0000] <mhausenblas>
annevk, would you think that what I've got here at http://enable-cors.org/#how-apache is correct (Header set Access-Control-Allow-Origin *)

[03:40:00.0000] <mhausenblas>
someone pointed out recently that it should be Header set Access-Control-Allow-Origin "*"

[03:40:01.0000] <mhausenblas>
with the remark: "I spent a while trying to debug this with our apache servers."

[03:40:02.0000] <mhausenblas>
the spec is rather clear on this, IMO

[03:41:00.0000] <mhausenblas>
while the Apache 2 doc not so, hence I blamed it on the doc ;)

[03:41:01.0000] <mhausenblas>
any advise, annevk?

[03:42:00.0000] <annevk>
I don't know how Apache deals with things

[03:43:00.0000] <annevk>
Can't you test it out on an Apache server and inspect the headers it transmits?

[03:43:01.0000] <mhausenblas>
well, yes, if I'd only knew which Apache version the chap is using :P

[03:43:02.0000] <mhausenblas>
that's what I replied him, btw

[03:43:03.0000] <mhausenblas>
a chap from Pixar

[03:44:00.0000] <annevk>
if you transmit with quotes it should fail

[03:44:01.0000] <mhausenblas>
but you know, I'm lazy, hence I was hoping for some advise rather than doin' it myself ... let's see

[03:45:00.0000] <annevk>
ah, me too ;)

[03:45:01.0000] <mhausenblas>
well, he claims that he has debugged it with 'their apache servers' ... whatever this means ;)

[03:45:02.0000] <mhausenblas>
he he, fair enough

[03:45:03.0000] <smaug____>
annevk: the draft has a link to http://www.whatwg.org/specs/web-socket-protocol/

[03:45:04.0000] <annevk>
smaug____, no it doesn't

[03:45:05.0000] <annevk>
smaug____, what version are you reading?

[03:45:06.0000] <smaug____>
http://dev.w3.org/html5/websockets/

[03:46:00.0000] <smaug____>
"Editor's Draft 5 August 2011"

[03:46:01.0000] <annevk>
oh

[03:46:02.0000] <annevk>
that's only in the SotD it seems

[03:46:03.0000] <annevk>
the actual normative references throughout the draft are to the correct version

[04:11:00.0000] <annevk>
smaug____, filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=13700

[04:26:00.0000] <smaug____>
annevk: thanks

[05:46:00.0000] <Ms2ger>
MikeSmith, could you redirect http://dev.w3.org/2006/webapi/Binding4DOM/?

[05:49:00.0000] <MikeSmith>
Ms2ger: sure, what to?

[05:50:00.0000] <Ms2ger>
http://dev.w3.org/2006/webapi/WebIDL/

[05:58:00.0000] <MikeSmith>
Ms2ger: should be working now

[05:59:00.0000] <Ms2ger>
Works, thanks

[05:59:01.0000] <MikeSmith>
yup

[06:20:00.0000] <SiPlus>
Do WebSockets support IPX?

[08:10:00.0000] <eighty4>
Quick question. Would you say that "HTML5 is as different from HTML4 as coffeescript is to JavaScript" or would you just say that html5 is an evolvement of html4?

[08:12:00.0000] <gsnedders>
eighty4: Definitely the latter.

[08:12:01.0000] <eighty4>
gsnedders: thanks, was starting to think that I was completely off.

[08:12:02.0000] <gsnedders>
eighty4: Except I'd say evolution as a native speaker. :)

[08:13:00.0000] <eighty4>
meh :)

[08:24:00.0000] <Dashiva>
If you need an analogy, maybe say that html5 is like es3 was to earlier javascript

[08:26:00.0000] <Dashiva>
(although es3 had enough problems that it might not be a net positive)

[08:27:00.0000] <gsnedders>
Dashiva: ES5 -> ES3 is a reasonable comparison, on the whole, though

[08:45:00.0000] <llrcombs>
is there any way to force a <progress> to animate when its value changes?

[08:45:01.0000] <llrcombs>
e.g. smoothly expanding the progress bar

[10:43:00.0000] <Hixie>
Zeddy: why are you passing parameters to the manifest?

[11:30:00.0000] <Hixie>
TabAtkins: http://dev.w3.org/csswg/css3-images/ has cvs conflicts in it

[12:50:00.0000] <jgraham>
gsnedders, Dashiva: except that ES3 is a real spec in a way that HTML4 isn't. I mean it's not as good as ES5 but it has testable requirements an stuf

[12:50:01.0000] <jgraham>
*and stuff

[12:51:00.0000] <gsnedders>
jgraham: HTML4 has testable requirements. It's just that a desk can fulfil them.

[12:52:00.0000] <jgraham>
gsnedders: In that sense it is quite unlike ES3

[13:31:00.0000] <annevk>
gsnedders, Dashiva, except we would have defined new Date(...) completely


2011-08-08
[18:32:00.0000] <benjoffe>
I wonder, would it be good to have a media query

[18:33:00.0000] <benjoffe>
'contrast', that is, say a float (0 -- normal) -> (1 - high contrast) that represents the contrast levels the audience is capable of viewing

[00:56:00.0000] <benjoffe>
I guess there's not much discussion going on here...

[01:08:00.0000] <zcorpan>
"It is basically a formal description of the state-of-the art in the Web platform from 3 years ago." - http://blog.n01se.net/?p=375

[01:13:00.0000] <espadrine>
zcorpan: The most important part of HTML5 is JavaScript? Odd.

[02:05:00.0000] <jgraham>
benjoffe_: Well Sunday evening is not the busiest time

[02:06:00.0000] <benjoffe_>
jgraham: Yeah makes sense (I'm GMT+1000 and often forget the reset of the world is far behind)

[02:07:00.0000] <jgraham>
I expect the people who produce the lowest contrast designs would also be the least likely to use that media query

[02:07:01.0000] <jgraham>
Also, I can't really imagine browsers shipping with UI to set the user preference

[02:08:00.0000] <benjoffe_>
jgraham: The idea though is that I would like to provide say red on black links, but it's not good for those with poor eyesight, it would be good if I could selectively make that a different colour only for those who need it

[02:08:01.0000] <jgraham>
Yes, I understand the idea

[02:09:00.0000] <jgraham>
In general I would expect solutions that work independently of the author (e.g. the readability features in various bookmarks and safari) to be bigger wins

[02:10:00.0000] <benjoffe_>
jgraham: sure features like that are great, but their existence is not enough for me to be able to implement these designs that I'd like to

[02:12:00.0000] <benjoffe_>
I can see it would be difficult to introduce this kind of media query though, as authors won't implement it until browsers support it and browsers won't support it (or at least emphasise it) until many sites use it

[02:14:00.0000] <jgraham>
And users won't set the preference unless it affects sites they use

[02:14:01.0000] <jgraham>
and will be annoyed if they set the pref and still get low-contrast text they can't read

[02:15:00.0000] <benjoffe_>
It could be a ui button that appears only if the site is making use of the query perhaps (kind of like how safari only shows 'reader' if it passes some heuristic)

[05:42:00.0000] <MikeSmith_>
Ms2ger: hmm, "no space left on device"

[05:43:00.0000] <MikeSmith_>
I think that host has only 4GB of disk space or so

[05:43:01.0000] <Ms2ger>
"W3C would like to thank Microsoft who donated the server that allows us to run this service."

[05:43:02.0000] <MikeSmith_>
4.3G

[05:44:00.0000] <MikeSmith_>
well, we can still be thankful for that

[05:44:01.0000] <MikeSmith_>
we just need some more diskspace

[05:45:00.0000] <MikeSmith_>
I just cleared off 152M

[05:45:01.0000] <Ms2ger>
/me wonders if enough people have bought HTML5 T-shirts to buy a bigger disk

[05:45:02.0000] <MikeSmith_>
heh

[05:45:03.0000] <MikeSmith_>
trying to see now what else I can clean off that machine

[05:45:04.0000] <MikeSmith_>
it really should not be using 4GB of disk space

[05:46:00.0000] <MikeSmith_>
we have very little running on it

[05:46:01.0000] <MikeSmith_>
what's the best way to get a high-level overview of where the most space is being used?

[05:46:02.0000] <MikeSmith_>
I mean, other than just running du --si or whatever from /

[05:47:00.0000] <jgraham>
MikeSmith_: If I send you a bigger disk, will I get thanked too? :)

[05:47:01.0000] <Ms2ger>
de --su?

[05:48:00.0000] <MikeSmith_>
jgraham: sure

[05:48:01.0000] <MikeSmith_>
OK, 3.0G in /var

[05:48:02.0000] <Ms2ger>
* jgraham sends a 4.4G disk

[05:48:03.0000] <jgraham>
/me doesn't even know where one would find a 4.3Gb diak

[05:49:00.0000] <jgraham>
*disk

[05:49:01.0000] <jgraham>
so I assume it is a small part of something larger

[05:49:02.0000] <MikeSmith_>
2.7G in /var/log

[05:49:03.0000] <MikeSmith_>
almost all of it in apache logs

[05:50:00.0000] <MikeSmith_>
so, I guess we need to set up some sane apache log-maintenance system

[05:50:01.0000] <Ms2ger>
Uh-oh

[05:50:02.0000] <MikeSmith_>
/me wonders if there is a good debian package for that

[05:50:03.0000] <Ms2ger>
/me hopes nobody notices his ddos

[05:51:00.0000] <jgraham>
/me failed to recognise DDOS in lowercase

[05:54:00.0000] <MikeSmith_>
deleted some logs, down to 1.1G now, as far as I can see

[05:55:00.0000] <Philip`>
MikeSmith_: Does logrotate count as sane?

[05:56:00.0000] <MikeSmith_>
Philip`: yeah :)

[05:56:01.0000] <MikeSmith_>
and looks like I have it installed already

[05:57:00.0000] <MikeSmith_>
but I guess we don't have it configured for the logs for the w3c-test.org domain

[05:58:00.0000] <MikeSmith_>
/me looks for logrotate config file

[05:59:00.0000] <MikeSmith_>
I guess I can just drop a new script into /etc/logrotate.d/

[06:04:00.0000] <MikeSmith_>
OK, I think I have it set up

[06:04:01.0000] <MikeSmith_>
Ms2ger: can you please try your push again?

[06:04:02.0000] <MikeSmith_>
if it doesn't work, let me know

[06:09:00.0000] <Ms2ger>
MikeSmith_, works

[06:09:01.0000] <Ms2ger>
Thanks

[06:09:02.0000] <MikeSmith_>
cool

[06:11:00.0000] <MikeSmith_>
Ms2ger: if/when you run into other problems with that w3c-test.org, feel free to ping me if I'm around

[06:11:01.0000] <MikeSmith_>
and/or ping francois on #testing

[06:12:00.0000] <MikeSmith_>
and/or plh

[06:12:01.0000] <MikeSmith_>
but plh is likely to just ping francois or me

[06:12:02.0000] <MikeSmith_>
I have root access to that box

[06:12:03.0000] <MikeSmith_>
for now

[06:13:00.0000] <MikeSmith_>
though the systems team might eventually decide I shouldn't

[06:13:01.0000] <Ms2ger>
/me won't tell anybody :)

[06:13:02.0000] <MikeSmith_>
heh

[06:14:00.0000] <MikeSmith_>
I'm already in trouble for installing packages from Debian testing and unstable there

[06:14:01.0000] <MikeSmith_>
but I don't know what else to do when the software we are running depends on packages that aren't in stable

[06:15:00.0000] <MikeSmith_>
stable stuff is so ancient

[06:17:00.0000] <MikeSmith_>
hmm, "Powered by Sourceforge" on the Roundcube homepage doesn't inspire confidence

[06:17:01.0000] <MikeSmith_>
hsivonen: ↑

[06:18:00.0000] <MikeSmith_>
it seems the state of Web-based e-mail clients kinda sucks

[06:18:01.0000] <Ms2ger>
I could have  told yyou thaat :)

[06:19:00.0000] <Ms2ger>
(With double y, no less)

[06:22:00.0000] <nessy>
I'd buy a html5 t-shirt to support the case

[06:25:00.0000] <jgraham>
MikeSmith_: You mean that no small, open source, project manages to produce a UI as polished as the huge, well funded gmail team in a space that hardly anyone cares about because almost everyoine is happy to just use gmail? What a surprise…

[06:26:00.0000] <MikeSmith_>
heh

[06:27:00.0000] <MikeSmith_>
jgraham: point taken, but I still think there's market opportunity for something better

[06:28:00.0000] <MikeSmith_>
even the big well-funded proprietary gmail alternatives suck

[06:28:01.0000] <Ms2ger>
Hi Zimbra!

[06:36:00.0000] <MikeSmith_>
/me didn't mention any names

[06:40:00.0000] <hsivonen>
MikeSmith_: I think PHP is less confidence-inspiring than sf.net

[06:41:00.0000] <hsivonen>
MikeSmith_: but I'm already a customer of a company that provides Roundcube hosting, so I guess I should check it out

[06:41:01.0000] <hsivonen>
though the screenshots suggest that most likely I don't want to use it

[06:41:02.0000] <MikeSmith_>
true, about PHP

[06:42:00.0000] <MikeSmith_>
I think we need to start a Just Say No campaign for PHP

[06:42:01.0000] <jgraham>
If you want feature parity with gmail, woundcube is not going to make you hppy

[06:42:02.0000] <jgraham>
*roundcube

[06:42:03.0000] <MikeSmith_>
PHP has made a lot of devs and companies/projects lazy

[06:43:00.0000] <MikeSmith_>
/me registers cloudmail.com

[07:17:00.0000] <karlcow>
/me notes that cloudmail.com is already taken, but MikeSmith_ can register nuagecourrier.com

[07:17:01.0000] <MikeSmith_>
heh

[07:53:00.0000] <AryehGregor>
MikeSmith_, for checking disk space I usually use du -m / | sort -nr | head -n 50.

[07:53:01.0000] <AryehGregor>
Prints out the top 50 directories by disk space usage.

[07:54:00.0000] <AryehGregor>
Although it takes a long time if you have a lot of directories.

[07:56:00.0000] <MikeSmith_>
AryehGregor: cool, thanks

[07:57:00.0000] <AryehGregor>
Although one disadvantage is it doesn't print out anything at all until it's done.

[07:57:01.0000] <AryehGregor>
Since sort has to wait for all the input before it can output the first line.

[07:58:00.0000] <MikeSmith_>
only takes 13 seconds on my local machine

[07:59:00.0000] <AryehGregor>
du -mx is probably better than du -m, actually, so it doesn't cross filesystem boundaries.

[07:59:01.0000] <AryehGregor>
(since you only care about one filesystem if you want to know what's using space)

[07:59:02.0000] <AryehGregor>
/me looks hopefully toward a btrfs future where ridiculous issues like that will go away, along with war and world hunger

[07:59:03.0000] <MikeSmith_>
taking a bit longer on my server...

[08:00:00.0000] <MikeSmith_>
/me googles btrfs

[08:00:01.0000] <AryehGregor>
I've seen it take hours on occasion.  But it's just du that's taking that long, it would take that long regardless.

[08:00:02.0000] <AryehGregor>
Because it has to look at every directory in the filesystem.

[08:03:00.0000] <AryehGregor>
btrfs is the cool next-gen filesystem for Linux that's basically a total ripoff of all the good ideas of ZFS, but on an OS that more than three people actually care about.

[08:03:01.0000] <MikeSmith_>
took 3 minutes 45 seconds on my server filesystem of ~16GB

[08:03:02.0000] <MikeSmith_>
heh

[08:03:03.0000] <MikeSmith_>
me dunno much about ZFS either?

[08:03:04.0000] <MikeSmith_>
Sun?

[08:04:00.0000] <AryehGregor>
In particular, it encourages you to put everything on one big filesystem, and divide things up into "subvolumes" that you don't have reserve space for in advance.  So you can divide the space up into subvolumes for different parts of the filesystem and therefore easily track what's using how much space, quickly remove entire subvolumes, etc., but not have to waste space by deciding how much to reserve for each subvolume in advance.

[08:05:00.0000] <AryehGregor>
But you can put limits on how big each subvolume is allowed to get, and track how big each one is, and otherwise mostly handle them as independent filesystems, so you get almost all the benefits of separate filesystems.

[08:05:01.0000] <AryehGregor>
But there are zillions of other cool features, like online resize (shrink as well as grow), online fsck, extremely low-cost snapshotting of subvolumes, improved performance on many workloads, etc. etc. etc.

[08:14:00.0000] <MikeSmith_>
sounds too good to be true

[08:14:01.0000] <MikeSmith_>
there must be a catch

[08:15:00.0000] <espadrine>
AryehGregor: By the way, what is "online resize and fsck"? That is one thing I never understood about Btrfs.

[08:17:00.0000] <AryehGregor>
MikeSmith_, the catch is it's not stable and who knows when it will be stable and until it's stable it will probably lock up and/or eat your data.

[08:18:00.0000] <MikeSmith_>
ah

[08:18:01.0000] <MikeSmith_>
that's definitely a bit of a downside

[08:18:02.0000] <AryehGregor>
Yeah, that's the sort of thing that tends to put a damper on filesystem innovation.

[08:18:03.0000] <AryehGregor>
If your scheduler or something messes up you can always just switch back to an old kernel and big deal, but filesystems messing up is a bit more of a practical problem.

[08:19:00.0000] <espadrine>
Well, ZFS has been stable for some time...

[08:19:01.0000] <AryehGregor>
espadrine, "online resize" means you can increase or decrease the size of the filesystem online, like if you add or remove a disk.  ext4 (and ext3 and ext2) can increase filesystem size but not decrease.

[08:19:02.0000] <AryehGregor>
Online, I mean.

[08:19:03.0000] <AryehGregor>
They can decrease offline, i.e., with the filesystem unmounted.

[08:19:04.0000] <espadrine>
oh! good!

[08:20:00.0000] <AryehGregor>
Online fsck is what it sounds like, btrfs is supposed to be able to check for and repair errors online instead of requiring you to mount your root filesystem read-only and therefore take the the whole system offline if you want to check for or fix errors.

[08:20:01.0000] <AryehGregor>
ZFS has been stable for some time but still reportedly has reliability issues.

[08:21:00.0000] <AryehGregor>
I know of more than one site that used ZFS but then switched away because they had problems with it failing randomly.

[08:21:01.0000] <AryehGregor>
("more than one site" being toolserver.org, and Wikimedia)

[08:21:02.0000] <AryehGregor>
Failing randomly or otherwise not living up to expectations.

[08:21:03.0000] <jgraham>
"Online fsck is what it sounds like" - an excuse for yet more tired /. jokes?

[08:21:04.0000] <AryehGregor>
jgraham, ?

[08:21:05.0000] <espadrine>
That could be the fate of Btrfs when released, unfortunately.

[08:23:00.0000] <jgraham>
AryehGregor: Isn't there a whole genre of jokes based around the small Levenshtein distance between the word "fsck" and the word "fuck"?

[08:23:01.0000] <AryehGregor>
Probably.

[08:24:00.0000] <AryehGregor>
espadrine, yeah, but Linux is much more widely used, so with any luck it will get much better testing.

[08:24:01.0000] <AryehGregor>
I'm not expecting to use it in production anytime within the next few years, though.

[08:25:00.0000] <Philip`>
If quality depends on testing which depends on number of users, that seems to indicate the quality-improvement process is (at least in part) to wait for users to encounter bugs and lose data and then fix those bugs

[08:25:01.0000] <Philip`>
Can't they, like, write it correctly in the first place?

[08:26:00.0000] <Philip`>
/me wishes programming was easier

[08:26:01.0000] <AryehGregor>
If programming was easy, most of the people in this chat room would make a lot less money.  :)

[08:27:00.0000] <jgraham>
If programming was easier we would use it to solve harder problems

[08:27:01.0000] <AryehGregor>
That too.

[08:30:00.0000] <Philip`>
It's only a filesystem, and the API is already designed and implemented many times so they don't have to deal with drastically changing requirements - it doesn't seem like a problem that will always grow until it's slightly too hard to implement

[08:32:00.0000] <AryehGregor>
So you'd think.  But in practice, people want filesystems to do all kinds of things, and people think of ever cleverer ways to do them.

[08:32:01.0000] <AryehGregor>
ZFS and btrfs are a totally different filesystem model from other commonly-used filesystems, they do everything copy-on-write.

[08:32:02.0000] <AryehGregor>
So there are a ton of practical issues to work out.

[08:33:00.0000] <AryehGregor>
Like "what happens when you get low on disk space?"  Both of them have issues then.

[08:33:01.0000] <AryehGregor>
btrfs uses some fancy B-tree-based data structure that some guy made up in like 2005.

[08:35:00.0000] <Philip`>
/me thinks it might be interesting if someone did a book collecting articles from developers of lots of projects, explaining what they see as the fundamentally hard problems they have to face, to show why things are never as easy as they superficially appear

[08:35:01.0000] <Philip`>
(Most things I've seen before are focusing on solutions, which is generally less interesting than problems)

[08:36:00.0000] <jgraham>
Probably no one wants to write a book suggesting that they're not clever enough to trivially solve all problems

[08:37:00.0000] <Philip`>
The developers working on a project have already clearly indicated they can't trivially solve all problems, because if they could then they wouldn't still be working on that project

[08:37:01.0000] <Philip`>
so the book would be a good way for them to offer excuses for not having finished the project yet

[08:37:02.0000] <zewt>
i sure don't want to touch a filesystem that hasn't been stable for many years

[08:38:00.0000] <zewt>
given that it's probably the single most critical piece of code on the entire system

[08:38:01.0000] <zewt>
i expect i'll be sticking with XFS for a long time to come

[08:41:00.0000] <AryehGregor>
I'm okay with using a filesystem once it's been the default in major distros for a version or two and there are no widespread reports of breakage.

[08:41:01.0000] <AryehGregor>
Provided I keep backups of anything important.

[08:42:00.0000] <AryehGregor>
If a million Ubuntu users haven't hit anything really nasty in a year, my desktop isn't important enough that I'm going to worry about being the first.

[08:43:00.0000] <zewt>
well, the default in debian/ubuntu still tends to be ext3/ext4, but i don't trust that

[08:43:01.0000] <AryehGregor>
ext3 has been the default filesystem in Linux for like ten years.

[08:43:02.0000] <Philip`>
Just make sure your backup disks aren't using the new filesystem too

[08:44:00.0000] <zewt>
if a filesystem defaults to doing a lengthy full fsck every N boots by default, that tells me the developers have very little confidence in it

[08:44:01.0000] <espadrine>
zewt: I read somewhere that ext developers had decided never to create ext5

[08:44:02.0000] <espadrine>
they are confident Btrfs will be the future

[08:44:03.0000] <Philip`>
Maybe they'll go straight from ext4 to ext6

[08:44:04.0000] <zewt>
it's very strange that distros still use ext*, since xfs is pretty much better in every way

[08:45:00.0000] <zewt>
for very low-memory systems maybe

[08:45:01.0000] <Ms2ger>
Obviously ext5 would involve writing a spec for each line of code

[08:46:00.0000] <AryehGregor>
espadrine, that was decided a few years ago, yes.  There was a major Linux filesystem summit, and the plan everyone agreed on was to make short-term easy improvements to ext3 in the form of ext4, and consign the long-term future of Linux filesystems to btrfs.

[08:48:00.0000] <espadrine>
AryehGregor: this is going to be a difficult transition... something like Python3

[08:48:01.0000] <AryehGregor>
zewt, what advantage does XFS have over ext4?

[08:48:02.0000] <AryehGregor>
Defragmentation?

[08:48:03.0000] <espadrine>
zewt: XFS looks good, I wonder why I haven't heard of it before

[08:48:04.0000] <espadrine>
AryehGregor: it cannot be shrunk

[08:48:05.0000] <AryehGregor>
espadrine, not at all.  You can do an extremely fast in-place upgrade of ext4 to btrfs, with rollback.

[08:49:00.0000] <AryehGregor>
And the difference is mostly transparent, btrfs does everything important ext4 does as far as applications are concerned.

[08:49:01.0000] <AryehGregor>
So it's just a matter of people waiting till it's stable and reliable.

[08:49:02.0000] <zewt>
my experience with xfs is that it just works: reliable, very consistently fast, and i havn't had to do an fsck of any kind since it was pre-release

[08:49:03.0000] <AryehGregor>
zewt, my experience with ext4 is the same.

[08:49:04.0000] <espadrine>
AryehGregor: I meant, people won't want to touch it unless other people use it

[08:49:05.0000] <espadrine>
which is the issue in Python3

[08:50:00.0000] <zewt>
ext* fscks by default every N boots (which you can turn off, but if a filesystem does something that hugely expensive by default, I assume it's for a reason)

[08:50:01.0000] <AryehGregor>
espadrine, to use Python 3 you have to port your code, which is nontrivial for large codebases, and impossible if you rely on unported libraries.

[08:50:02.0000] <AryehGregor>
Any random person can switch their machine from ext4 to btrfs for the heck of it today if they want.

[08:50:03.0000] <AryehGregor>
And they do.

[08:51:00.0000] <AryehGregor>
zewt, that could speak to a difference in configuration philosophy more than anything.

[08:51:01.0000] <AryehGregor>
The only time I've had nontrivial filesystem corruption on ext4 was when the underlying disk was bad.

[08:51:02.0000] <Philip`>
espadrine: I think the problem with Python 3 is more that it's worse than Python 2 until everybody else in the world has upgraded (so you can be compatible with their applications/libraries), so nobody wants to bother upgrading

[08:51:03.0000] <zewt>
the philosophy i read into it is "we don't trust our code enough"; others may read what they like, of course

[08:51:04.0000] <AryehGregor>
(there were other times that fsck caught something, but it was some trivial thing that it easily fixed)

[08:51:05.0000] <Philip`>
whereas an individual can change their filesystem regardless of what anyone else does

[08:52:00.0000] <zewt>
the problem with python 3 is it doesn't have a sane upgrade path, so it essentially fragments the language

[08:52:01.0000] <AryehGregor>
zewt, or it could be "we know hardware sometimes goes bad or lies regardless of how good our code is, and we aren't overconfident in our code either".

[08:53:00.0000] <zewt>
AryehGregor: has never happened to me, but i sure have had servers offline for long periods of time due to an un-asked-for fsck on reboot

[08:53:01.0000] <AryehGregor>
Yes, that's just stupid configuration.

[08:53:02.0000] <zewt>
ymmv etc etc

[08:53:03.0000] <AryehGregor>
Although ext4 fsck is much, much faster than ext3 fsck.

[08:54:00.0000] <AryehGregor>
Hmm, XFS can't be shrunk?

[08:54:01.0000] <AryehGregor>
Even offline?

[08:54:02.0000] <zewt>
dunno, never tried

[08:54:03.0000] <Philip`>
/me has never encountered the problem of having too much disk space

[08:54:04.0000] <AryehGregor>
I've had to shrink ext* on occasion.  It's really slow and only works offline, but it works.

[08:55:00.0000] <zewt>
really the only particularly interesting feature of newer filesystems is being able to snapshot (without layering another special block layer underneith), to be able to backup databases directly

[08:55:01.0000] <Philip`>
If you don't mind really slow, copy the data to an external disk then repartition and reformat then copy back

[08:55:02.0000] <zewt>
Philip`: well, if you're working with an array, you don't always have that option

[08:55:03.0000] <AryehGregor>
Philip`, only works if you have enough empty storage.

[08:56:00.0000] <zewt>
(+ interesting feature to me, of course)

[08:56:01.0000] <Philip`>
/me wouldn't really want to resize partitions unless he already had a backup copy anyway

[08:56:02.0000] <AryehGregor>
btrfs has lots of interesting features beyond snapshotting.

[08:56:03.0000] <AryehGregor>
In-FS RAID looks like it will be really cool.

[08:56:04.0000] <AryehGregor>
You'll be able to choose different RAID levels for different files.

[08:57:00.0000] <AryehGregor>
And it doesn't have the same resyncing problems as normal software RAID.

[08:58:00.0000] <AryehGregor>
So RAID5 for backups (= almost all reads/writes are serial, take up lots of space) but RAID10 for database files (= lots of random read/write), say.

[08:58:01.0000] <AryehGregor>
Of course, a lot of this stuff is still on the drawing board.

[08:59:00.0000] <zewt>
more interesting would be intelligent RAID for databases across spinnydisks and SSDs

[09:00:00.0000] <AryehGregor>
Oh, and you can copy files instantly, like hard-linking except the file is actually logically totally separate.  That's very cool.

[09:00:01.0000] <zewt>
eg. to be able to get safe writes to disk (which many, especially cheaper, SSDs are very bad at), while being able to use the SSD for reads (where it's an order of magnitude improvement for databases in many cases)

[09:01:00.0000] <AryehGregor>
That's probably best handled at a different layer.

[09:01:01.0000] <AryehGregor>
Somewhere below the filesystem layer.

[09:01:02.0000] <AryehGregor>
There's work in that department in Linux, IIRC including something funded by Facebook.

[09:02:00.0000] <zewt>
it can be done at the FS layer logically: report writes as synced to disk as soon as they've hit at least one disk, without waiting for it to be confirmed to all

[09:02:01.0000] <zewt>
(whether you actually want that or not is a matter of policy)

[09:03:00.0000] <AryehGregor>
md already can do that, with the --write-behind= option.  Although of course that's inflexible.

[09:03:01.0000] <AryehGregor>
In particular, the SSD has to be the same size as the other device, which mostly defeats the point.

[09:03:02.0000] <zewt>
of course, different disk sizes and other things mean it's far from trivial

[09:04:00.0000] <AryehGregor>
The better use for SSDs is probably to use them as a read cache only, not a write buffer.  Cheaper SSDs are very slow to write to randomly anyway.

[09:04:01.0000] <AryehGregor>
Plus, writes can be buffered a lot already in memory or in battery-backed RAID controllers or such.

[09:05:00.0000] <AryehGregor>
Whereas you can't buffer reads, so those are the things you really want to get rid of as much as possible (admittedly depending on workload).

[09:09:00.0000] <timeless>
did eighty4 mean mochascript, or was he poking fun?

[09:09:01.0000] <timeless>
zewt: dialog wouldn't be tab modal, it would only be [[document]] modal.

[09:09:02.0000] <zewt>
i'm not sure what the distinction is

[09:10:00.0000] <timeless>
it's from friday i think :)

[09:10:01.0000] <zewt>
but iirc the point was "not window-modal"

[09:10:02.0000] <timeless>
i think you were worried about dialogs gaining focus when they shouldn't

[09:10:03.0000] <zewt>
i wasn't, i think i was responding to someone else who was

[09:38:00.0000] <timeless>
interesting, +50 Cent

[09:39:00.0000] <timeless>
and +T- Pain

[09:46:00.0000] <smaug____>
annevk5: is http://dev.w3.org/csswg/cssom-view/ the right version of the spec

[09:46:01.0000] <TabAtkins>
smaug____: Yeah.

[09:46:02.0000] <smaug____>
I somehow thought there was a bit different text for http://dev.w3.org/csswg/cssom-view/#dom-document-caretpositionfrompoint

[09:46:03.0000] <timeless>
annevk5: ..

[09:46:04.0000] <timeless>
> Terminology used in this specification is from DOM Core, DOM Range, CSSOM and HTML. [DOMCORE] [DOMRANGE] [CSSOM] [HTML]

[09:47:00.0000] <timeless>
> Content edge, padding edge, border edge, and canvas are defined by CSS.

[09:47:01.0000] <timeless>
shouldn't CSS get a [] ref?

[09:47:02.0000] <timeless>
(preferably listed in the Terminology used ... line)

[09:47:03.0000] <smaug____>
annevk5: anyway, http://dev.w3.org/csswg/cssom-view/#dom-document-caretpositionfrompoint seems to miss the case what should happen with documents without viewport

[09:48:00.0000] <smaug____>
annevk5: returning null seems to be the only reasonable option

[09:50:00.0000] <dglazkov>
good morning, Whatwg!

[09:50:01.0000] <Ms2ger>
Keeps getting later

[10:07:00.0000] <jgraham>
Ms2ger: That's what happens as we move away from midsummer

[10:13:00.0000] <Ms2ger>
Sunset before 10PM :(

[10:13:01.0000] <timeless>
heh

[10:13:02.0000] <timeless>
Ms2ger: where are you again?

[10:13:03.0000] <Ms2ger>
I mean, that's the time I start getting productive

[10:13:04.0000] <timeless>
they move movie in the park showings from 9pm to 8:30pm in august here in Toronto

[10:14:00.0000] <timeless>
(because it gets dark by then..)

[10:14:01.0000] <Ms2ger>
You do always seem to end up pretty far North

[10:14:02.0000] <timeless>
/me ponders

[10:15:00.0000] <timeless>
SJC and DCA seem to be about the same for North

[10:16:00.0000] <timeless>
I've never lived much further south than that, but..

[10:16:01.0000] <timeless>
I've really only lived North for the last bit of my life :)

[10:17:00.0000] <timeless>
and YYZ is really much further south tha HEL :)

[10:18:00.0000] <jgraham>
You can put a bound on where Ms2ger lives from the fact that he experiences post 10pm local time sunsets but they are currently not post 10pm

[10:19:00.0000] <Ms2ger>
Hmm, I would've situated DCA quite a bit north of SJC

[10:19:01.0000] <jgraham>
Unhappily stalkpeopleusingthesun.com doesn't yet exist

[10:19:02.0000] <jgraham>
So it's a bit of a pian to work out

[10:19:03.0000] <jgraham>
*pain

[10:19:04.0000] <timeless>
Ms2ger: i'm using Gmaps to compare..

[10:20:00.0000] <Ms2ger>
Me too :)

[10:20:01.0000] <Ms2ger>
jgraham, that would be .co.uk

[10:20:02.0000] <timeless>
Coordinates  38°51′08″N 077°02′16″W

[10:20:03.0000] <timeless>
Coordinates  37°21′46″N 121°55′45″W

[10:20:04.0000] <timeless>
so, it's 1deg 30mins north

[10:20:05.0000] <Ms2ger>
Also, surely you can calculate that off the top of your head, Dr. Graham? :)

[10:21:00.0000] <jgraham>
This is the point where I wish I had a repitoire of rude gestures in ascii

[10:21:01.0000] <timeless>
/me chuckles

[10:21:02.0000] <timeless>
hold on a sec

[10:21:03.0000] <Ms2ger>
♥

[10:22:00.0000] <timeless>
you can use http://www.happyzebra.com/timezones-worldclock/sunrisesunset.php

[10:23:00.0000] <timeless>
Ms2ger: would you have pictured the delta between DCA and SJC as more than 2degrees?

[10:24:00.0000] <Ms2ger>
Without looking at a map, a couple hundred km

[10:24:01.0000] <timeless>
/me isn't very good at converting degrees to distance...

[10:25:00.0000] <Ms2ger>
360° = 40Mm

[10:25:01.0000] <TabAtkins>
Along a circumference.

[10:26:00.0000] <timeless>
well, 1deg is ~111km

[10:26:01.0000] <timeless>
so... ~150km ?

[10:27:00.0000] <Ms2ger>
/me just stays away from America

[10:28:00.0000] <timeless>
hrm, maybe 170 ?

[10:28:01.0000] <TabAtkins>
Or actually, along a meridian.  Arbitrary circumferences may vary a decent bit from 40Mm

[10:28:02.0000] <timeless>
/me sighs

[10:28:03.0000] <timeless>
are those mega-meters?

[10:28:04.0000] <Ms2ger>
Yes

[10:28:05.0000] <timeless>
people actually use that?

[10:29:00.0000] <TabAtkins>
No, m&ms

[10:29:01.0000] <TabAtkins>
Yes?  Why wouldn't they?

[10:29:02.0000] <timeless>
ooh

[10:29:03.0000] <timeless>
interesting

[10:29:04.0000] <timeless>
Ms2ger: ?

[10:29:05.0000] <jarek>
Hi

[10:29:06.0000] <Ms2ger>
I do, when I'm too lazy to type out three 0s

[10:29:07.0000] <timeless>
do you have firefox handy?

[10:29:08.0000] <timeless>
My find bar has `Mat<u>c</u>h<br>case`

[10:30:00.0000] <jarek>
is it possible to overwrite the values such as width or device-width that are used by media queries?

[10:30:01.0000] <TabAtkins>
An older definition of the kilometer was 1/10000 of a quarter-meridian.  Thus, a full meridian is 40Mm

[10:30:02.0000] <jgraham>
TabAtkins: Depending on what you mean "arbitary circumfrences", "a decent bit" would be "by 100%"

[10:30:03.0000] <TabAtkins>
jgraham: Huh?

[10:30:04.0000] <TabAtkins>
jgraham: I just mean that an equatorial circumference will be a bit larger than a polar one.

[10:31:00.0000] <jgraham>
Well yes, at the pole it will be zero length, right?

[10:31:01.0000] <TabAtkins>
...

[10:31:02.0000] <TabAtkins>
Do I really need to state that it's a great circle?

[10:31:03.0000] <timeless>
yes!

[10:32:00.0000] <TabAtkins>
Nerds.

[10:32:01.0000] <Ms2ger>
Yes!

[10:33:00.0000] <Ms2ger>
Though actually... NO! Geeks! :)

[10:33:01.0000] <timeless>
Yes!

[10:34:00.0000] <jgraham>
/me apologies to the world

[10:37:00.0000] <][DragoNero][>
Un saluto a tutto il chan

[10:39:00.0000] <Ms2ger>
...Hi

[10:54:00.0000] <eighty4>
timeless: no, I mean coffeescript. But it wasn't my quote

[10:55:00.0000] <AryehGregor>
TabAtkins, you have to say that it's a great circle, because latitude lines aren't great circles (except the equator).  So a degree of latitude will generally be less distance than a degree of longitude, up to 100% less.

[10:55:01.0000] <AryehGregor>
. . . I mean the other way around.

[10:55:02.0000] <AryehGregor>
A degree of longitude at our latitude is about 70% as long as a degree of latitude.

[10:56:00.0000] <Ms2ger>
In Texas, all circles are great circles.

[11:02:00.0000] <TabAtkins>
AryehGregor: No, you had it right the first time.

[11:02:01.0000] <TabAtkins>
latitudinal degrees are smaller than longitudinal over most of the world

[11:02:02.0000] <AryehGregor>
No, because moving a degree of latitude is moving north or south, and that's always along a great circle (one that contains the north and south poles).

[11:02:03.0000] <TabAtkins>
Nope.

[11:02:04.0000] <AryehGregor>
Since when you're changing latitude, you're actually moving along a longitude line, and vice versa.

[11:02:05.0000] <TabAtkins>
Meridians are longitudinal lines.

[11:03:00.0000] <TabAtkins>
Oh, wait, I see what you're saying.

[11:03:01.0000] <AryehGregor>
Yes, and moving along a longitudinal line is a change in latitude.  Your longitude remains constant.

[11:03:02.0000] <AryehGregor>
It always confuses me.

[11:04:00.0000] <AryehGregor>
TabAtkins, I just sent you an e-mail, I'd like to know what the headers say.

[11:04:01.0000] <Philip`>
A sphere is a silly shape for a planet, it makes coordinates far too complex

[11:05:00.0000] <TabAtkins>
Philip`: Agreed.  Let's do a 4-torus.

[11:05:01.0000] <AryehGregor>
We should have a flat torus instead.

[11:05:02.0000] <AryehGregor>
Far simpler.

[11:05:03.0000] <TabAtkins>
Haha!  We all agree!

[11:05:04.0000] <AryehGregor>
Well, I said a flat torus, which is actually a sort of 2-torus, but maybe you meant that too.

[11:05:05.0000] <AryehGregor>
(since a flat torus can only be embedded in Euclidean space of four or more dimensions)

[11:05:06.0000] <TabAtkins>
Yes.

[11:05:07.0000] <AryehGregor>
(a 4-torus, strictly speaking, would be a four-dimensional manifold in its own right, not two-dimensional)

[11:05:08.0000] <TabAtkins>
My geometry terminology is all kinds of inconsistent.

[11:06:00.0000] <TabAtkins>
So yeah, a 2-torus embedded in 4d so it's flat in 3d

[11:06:01.0000] <AryehGregor>
Well, to be fair, nobody talks about these things precisely unless you take graduate-level differential geometry or something.

[11:06:02.0000] <Ms2ger>
(* Recognizable by the use of the word "geodetes")

[11:08:00.0000] <Philip`>
/me was thinking more of a discworld (where you just need X and Y coordinates relative to some arbitrary axis) or ringworld (where you need one angular and one linear coordinate)

[11:09:00.0000] <Philip`>
(I suppose ringworlds are inconvenient for drawing maps though, since you need an incredibly long thin piece of paper)

[11:09:01.0000] <TabAtkins>
Are we talking a circular discworld?  Polar coords are still best there.  A square discworld can just use x/y, but you might as well get exotic and go for the torus then.

[11:09:02.0000] <Ms2ger>
What's next, want it on top of an elephant too?

[11:10:00.0000] <Philip`>
Ms2ger: Don't be silly - it'd fall off its back

[11:10:01.0000] <Philip`>
You need at least three elephants for stability

[11:10:02.0000] <TabAtkins>
Philip`: A ringworld doesn't need to be enormous unless you're trying to build it in physical reality.

[11:11:00.0000] <Philip`>
TabAtkins: But polar coordinates make it hard to work out distances between points, which is probably the main value in having coordinates at all

[11:11:01.0000] <TabAtkins>
Philip`: I know, but that doesn't change the fact that polar coords are the most natural for a circular disk.

[11:13:00.0000] <Philip`>
TabAtkins: It still needs to be long and thin, regardless of scale, else it'd be like a Rama rather than a ringworld

[11:14:00.0000] <TabAtkins>
...yes?

[11:14:01.0000] <AryehGregor>
Philip`, not if it's a flat torus.  Then it's just a rectangle with opposite edges identified, so it can be whatever proportions you like.

[11:14:02.0000] <AryehGregor>
It can be a square, no problem.

[11:14:03.0000] <TabAtkins>
Rama's a perfectly fine ringworld as well.

[11:14:04.0000] <Philip`>
If it was somewhere in the middle it'd look silly

[11:14:05.0000] <AryehGregor>
Also, then you can use (x, y) coordinates and calculate using Euclidean distance.

[11:14:06.0000] <AryehGregor>
It's the best of all worlds.

[11:15:00.0000] <AryehGregor>
(in multiple senses, I guess)

[11:15:01.0000] <TabAtkins>
AryehGregor: Though, you still have to do a bit of math.  If you're near the, um, dateline-equivalent, it's shorter to measure across that.

[11:16:00.0000] <AryehGregor>
Well, yes, there are multiple routes to the same point and it's less trivial in some cases to figure out which is shortest.

[11:16:01.0000] <Ms2ger>
Just put your sun right above it

[11:16:02.0000] <AryehGregor>
That's true.

[11:17:00.0000] <TabAtkins>
There's only four routes to check, and you can use straight-line distance for each.

[11:47:00.0000] <Ms2ger>
Potentially of interest to Opera people: http://qdb.us/306411

[12:27:00.0000] <timeless>
http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

[12:27:01.0000] <timeless>
/me rotfl

[12:27:02.0000] <timeless>
There exists an algorithm which transforms names and can be reversed losslessly.  (Yes, yes, you can do it if your algorithm returns the input.  You get a gold star.)

[12:28:00.0000] <AryehGregor>
Is that the one that basically says "you can't assume anything whatsoever about names, up to and including that people actually have one"?

[12:28:01.0000] <timeless>
probably

[12:28:02.0000] <timeless>
but it's possible i haven't seen the one you're referencing

[12:28:03.0000] <timeless>
since i actually got a w3 work item involving names

[12:28:04.0000] <timeless>
please do feel free to pass along your link :)

[12:29:00.0000] <timeless>
btw, does 田中太郎 just map to `Tanaka`?

[12:29:01.0000] <zewt>
transliterating asian names is the worst

[12:29:02.0000] <timeless>
yeah yeah

[12:29:03.0000] <AryehGregor>
translate.google.com says "Taro Tanaka".

[12:30:00.0000] <timeless>
but googling a Chinese name doesn't give me anything i can digest

[12:30:01.0000] <timeless>
ok, i see what i did wrong

[12:31:00.0000] <timeless>
/me is reading <Taro the Space Alien> and didn't properly parse `Taro Tanaka (田中太郎 Tanaka Tarō?)`

[12:31:01.0000] <timeless>
the question mark is amusing btw

[12:31:02.0000] <zewt>
http://rut.org/cgi-bin/j-e/sjis/dosearch?sName=on&H=PW&L=J&T=%91%BE%98Y&WC=none

[12:33:00.0000] <zewt>
fewer matches than a lot of japanese surnames, at least

[12:33:01.0000] <timeless>
zewt: cute

[12:33:02.0000] <timeless>
so it can either be a personal or family name

[12:33:03.0000] <zewt>
er right

[12:33:04.0000] <timeless>
but it has a different <pronunciation?> depending?

[12:34:00.0000] <zewt>
i think those would be pronounced the same

[12:34:01.0000] <timeless>
um, ok

[12:34:02.0000] <timeless>
so other than being different things

[12:34:03.0000] <zewt>
ah right "s" is surname "u" is given name

[12:34:04.0000] <timeless>
and having a secondary presentation which is different

[12:35:00.0000] <timeless>
yeah, i clicked on `u` and read the page for the others

[12:35:01.0000] <timeless>
http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/#comment-4491

[12:35:02.0000] <timeless>
41. If a person has a title such as “Doctor”, they will have at most one such title. (In Germany, the over-educated go by names such as “Prof. Dr. Dr. Hans Schmidt” – a professor with two PhD’s)

[12:35:03.0000] <zewt>
http://rut.org/cgi-bin/j-e/sjis/dosearch?sName=on&H=PW&L=J&T=%93c%92%86&WC=none surnames tend to have a lot more options than given names, i think

[12:36:00.0000] <AryehGregor>
timeless, I know multiple rabbi doctors.

[12:36:01.0000] <timeless>
AryehGregor: yeah, i know some Lawyer Doctors

[12:36:02.0000] <AryehGregor>
(usually Ph.D., not M.D., although probably one or two of the latter also)

[12:36:03.0000] <karlcow>
timeless: except there are names you can't type yet.

[12:36:04.0000] <timeless>
karlcow: ?

[12:36:05.0000] <AryehGregor>
Although most of the rabbis doctors I know go by Mr., because they're modest.

[12:36:06.0000] <timeless>
/me nods

[12:37:00.0000] <timeless>
+1 to them

[12:37:01.0000] <karlcow>
some Japanese names, local, historical variations of persons names not included in Unicode.

[12:37:02.0000] <timeless>
karlcow: oh, sure

[12:37:03.0000] <AryehGregor>
Actually, most of the rabbis I know go by Mr. unless they're actually rabbis of a congregation or religious instructors or something.  Kind of like how most Ph.D.'s don't call themselves "Dr." or "Professor" outside of a professional context, and you get suspicious if they do.

[12:37:04.0000] <timeless>
like en.wikipedia.org/wiki/Worf

[12:38:00.0000] <AryehGregor>
In the vein of: http://robert.ocallahan.org/2011/06/some-advice_14.html

[12:38:01.0000] <timeless>
/me wonders if wikipedia has a link for it

[12:38:02.0000] <timeless>
AryehGregor: indeed

[12:38:03.0000] <timeless>
well...

[12:39:00.0000] <AryehGregor>
Right, Dr. Graham?

[12:39:01.0000] <karlcow>
at least French revolution cut the head of a lot religious names ! that's progress :p

[12:39:02.0000] <timeless>
i guess for chabbad, it'd be <professional-context> as opposed to <congregation> on average

[12:39:03.0000] <karlcow>
s/lot/lot of/

[12:39:04.0000] <timeless>
karlcow: sigh

[12:40:00.0000] <AryehGregor>
As far as I can tell, basically every man who's brought up Lubavitch and wants to spend time proselytizing is given the title of rabbi for professional reasons.

[12:40:01.0000] <karlcow>
I wonder if HTML would have been developer during the revolution if it would have been without <head>

[12:40:02.0000] <karlcow>
s/developer/developed/

[12:41:00.0000] <karlcow>
can't type today

[12:41:01.0000] <timeless>
AryehGregor: to be fair, i think they don't get it immeidately :)

[12:41:02.0000] <AryehGregor>
Probably not.

[12:41:03.0000] <timeless>
e.g. i've met some who are still growing through the basic traveling phases

[12:41:04.0000] <timeless>
and they don't have the title yet

[12:41:05.0000] <timeless>
/me grumbles

[12:41:06.0000] <timeless>
i can't find Worf's klingon glyph

[12:41:07.0000] <timeless>
(s)

[12:42:00.0000] <zewt>
at least if you use a pile of poo to represent your name, you're good to go

[12:42:01.0000] <karlcow>
:D

[12:42:02.0000] <karlcow>
emoji++ \o/

[12:42:03.0000] <timeless>
eep

[12:43:00.0000] <zewt>
ﾍ(ﾟ∀ﾟﾍ)

[12:43:01.0000] <karlcow>
I guess in the list of names which can't be written are all the names of spoken only languages

[12:45:00.0000] <timeless>
> 43 – A person name in country-1 is the same as his name in country-2.  In Holland my name is Gert; In Denmark it is Gerardus; In Germany they refuse to try pronounce my name

[12:48:00.0000] <Philip`>
It's much easier to assign everyone a number and refer to them by that

[12:48:01.0000] <karlcow>
my Japanese Hanko for papers is not the name on my passport

[12:48:02.0000] <timeless>
Philip`: i'm actually supposed to be thinking about a Contacts api

[12:48:03.0000] <zewt>
nevermind addresses :|

[12:49:00.0000] <timeless>
zewt: yeah, that goes after contacts

[12:49:01.0000] <zewt>
t-mobile refused to ship anything to me because my address wouldn't validate in its address thing

[12:49:02.0000] <timeless>
and i need to send some hatemail to the geolocation people about their address api

[12:49:03.0000] <karlcow>
A full contact API should have a touch interface.

[12:49:04.0000] <zewt>
to the point where i had to have stuff shipped to family and drive and pick it up

[12:49:05.0000] <zewt>
haaaaaate

[12:49:06.0000] <timeless>
zewt: hey, O2 UK and IE refused to accept my Credit Card

[12:49:07.0000] <timeless>
because it wasn't a UK or IE credit card at the time

[12:49:08.0000] <timeless>
you'd think the fact that it was a Visa card and that they took Visa should have been sufficient

[12:50:00.0000] <zewt>
lots of stores are dumb about international cards

[12:50:01.0000] <timeless>
but *no*, they *insisted* on validating it against a UK or IE physical address

[12:50:02.0000] <zewt>
but screwing up domestic shipping ... that takes special talent

[12:50:03.0000] <timeless>
because obviously i wanted to use my credit card to pay for a pay as you go sim

[12:50:04.0000] <timeless>
... for a country where i *obviously* lived and had a valid credit card... right

[12:51:00.0000] <timeless>
O2 UK/IE is special

[12:51:01.0000] <timeless>
/me can't wait for someone to put them out of business

[12:51:02.0000] <timeless>
zewt: otoh, you win :)

[12:52:00.0000] <timeless>
what was 'special' about your address?

[12:52:01.0000] <zewt>
usps doesn't ship to my house, i have a po box

[12:52:02.0000] <zewt>
but no shipper should ever care about that

[12:53:00.0000] <timeless>
hrm

[12:54:00.0000] <timeless>
/me will have to get zewt to look at some APIs at some point to see if they work

[12:54:01.0000] <zewt>
i think it's some usps api they use to verify shipping addresses ... but usps is wrong

[12:54:02.0000] <timeless>
presumably USPS will deliver to the (its?!) PO Box

[12:54:03.0000] <zewt>
my street address, with the PO box in the ZIP code, is valid for shipping to me (it's what i use since it works for usps + fedex + ups)

[12:55:00.0000] <zewt>
but i think the usps api only accepts the actual po box, and t-mobile's thing screamed WE DON'T SHIP TO PO BOXES

[12:55:01.0000] <timeless>
have you tried DHL? :)

[12:55:02.0000] <zewt>
dhl is the worst in the US

[12:55:03.0000] <timeless>
yeah, i never used it in the US. But it was common in Finland

[12:55:04.0000] <zewt>
(not that anyone uses it in the US except for people shipping from overseas, and then rarely)

[12:56:00.0000] <zewt>
one time i had $10k in development hardware being shipped here, they left it in the yard

[12:56:01.0000] <zewt>
good job guys

[12:56:02.0000] <timeless>
nice

[12:56:03.0000] <zewt>
another time they left a package at some random house not even on the right street, i had to go pick it up

[12:56:04.0000] <zewt>
"i can't find the house, whatever this one looks good"

[12:57:00.0000] <timeless>
...wow

[12:57:01.0000] <timeless>
http://en.wikipedia.org/wiki/Nicholas_Barbon has a neat middle name...

[12:58:00.0000] <timeless>
s/has/had/ (deceased)

[12:58:01.0000] <Ms2ger>
...the eldest son of Praise-God Barebone...

[12:58:02.0000] <timeless>
read on

[12:59:00.0000] <timeless>
... "Unless-Jesus-Christ-Had-Died-For-Thee-Thou-Hadst-Been-Damned" ....

[12:59:01.0000] <timeless>
great family :)

[13:00:00.0000] <timeless>
> Barbon did this despite long-established restrictions on new buildings associated with various Acts of Parliament and royal declarations in the late 16th century: he often simply disregarded legal and local objections, demolished existing buildings without permission and rebuilt speculatively in search of a quick profit.

[13:00:01.0000] <timeless>
great guy, clearly a pragmatist :)

[13:04:00.0000] <timeless>
anyway...

[13:04:01.0000] <timeless>
my general feeling on Contact cards is:

[13:04:02.0000] <timeless>
support Name[]

[13:05:00.0000] <timeless>
where each element in Name is a full (or not) name, so someone could have Name = ['Dan Smith', 'Daniel Smith', 'Danny

[13:05:01.0000] <timeless>
Smith']

[13:06:00.0000] <timeless>
anything which performs a search should search across all Name[] elements of all Contacts[]

[13:06:01.0000] <timeless>
If someone wants to have a nickname, they would just include it as another entry in Name[]

[13:06:02.0000] <Ms2ger>
As long as it supports ["Ms2ger"], I'm happy with it

[13:06:03.0000] <timeless>
it woul

[13:06:04.0000] <timeless>
d

[13:07:00.0000] <timeless>
My general approach to the other side (transmission) is to encourage all UAs to allow users to modify cards before transmitting

[13:07:01.0000] <timeless>
especially to enable selection of a subset of fields and values

[13:07:02.0000] <timeless>
as well as to provide substitutions and additions

[13:07:03.0000] <timeless>
so you could Name[]+= "Your roommate from freshman year of college"

[13:09:00.0000] <jgraham>
Philip`: Move to Sweden, they have the "everyone should be referred to by number" thing sorted

[13:09:01.0000] <zewt>
google contacts has a "name" and "transliterated name" thing, though that's not really enough

[13:09:02.0000] <Ms2ger>
jgraham, how hard is it to get a number?

[13:09:03.0000] <timeless>
zewt: people complain that google's outgoing behavior is problematic

[13:10:00.0000] <timeless>
but i think that's more because it tries to serialize to forms that have first/last or similar

[13:10:01.0000] <zewt>
eg. if i have a contact with a japanese name and transcription and a japanese person has the same, it's hard for a UI to know whether the user wants a particular contact natively or transliterated

[13:10:02.0000] <gsnedders>
Hi, I'm 199204804497.

[13:10:03.0000] <zewt>
and showing both is just bad noisy ui

[13:10:04.0000] <timeless>
gsnedders: finland has cards

[13:10:05.0000] <timeless>
what's amusing is that i thought it was a numbers only systems

[13:10:06.0000] <zewt>
still, it's a small step forward from contacts that don't have it at all

[13:10:07.0000] <gsnedders>
Ms2ger: You have to have some address for some time, though not always that long.

[13:11:00.0000] <timeless>
it turns out that there's a <dash> between two numbers which isn't always a dash

[13:11:01.0000] <timeless>
and is thus a significant character in the `number`

[13:11:02.0000] <AryehGregor>
gsnedders, what you're missing is that for most purposes, names need be only locally unique.  Your friends would just call you '497.  It would be quite practical.

[13:11:03.0000] <timeless>
yeah, for contacts, your set is generally locally unique

[13:12:00.0000] <timeless>
and when you give out a card to someone, you probably only ever give out a subset of the card

[13:12:01.0000] <timeless>
imagine the card for your s/o has `likes pink roses`

[13:12:02.0000] <gsnedders>
AryehGregor: You don't need the checksum, really :)

[13:12:03.0000] <timeless>
you wouldn't give out that part of the card when you give it to someone

[13:12:04.0000] <Ms2ger>
Calling you 1992', OTOH, would not work all that well

[13:12:05.0000] <AryehGregor>
No, that would be a problem.

[13:12:06.0000] <gsnedders>
I seem to be mistaken, the official form is 920480-4497

[13:12:07.0000] <AryehGregor>
It's like in court cases, where they refer to patents by the last few digits.

[13:13:00.0000] <timeless>
gsnedders: yeah, that's like the finnish form

[13:13:01.0000] <timeless>
they omit the century field

[13:13:02.0000] <timeless>
but have a dash somewhere which isn't insignificant :(

[13:13:03.0000] <gsnedders>
(YYMMDD is the form, but my date is offset as what I have is basically a temporary one where they add 60 to it)

[13:14:00.0000] <gsnedders>
timeless: the dash seems to have meaning in Sweden as well, + being used for people born the 19th cent

[13:14:01.0000] <gsnedders>
timeless: Dunno about difference between 20th/21st

[13:14:02.0000] <Philip`>
<conversation xmlns:497="920480-4497">Hello '497

[13:15:00.0000] <jgraham>
Ms2ger: Not hard to get one if you are an EU citizen and have a job at least

[13:15:01.0000] <jgraham>
I haven't tried any other way

[13:16:00.0000] <jgraham>
Apparently if you are born a Swedish citizen it is also easy ;)

[13:16:01.0000] <timeless>
i have a finnish id

[13:16:02.0000] <timeless>
and a canadian id

[13:16:03.0000] <timeless>
well, not `id` really, but `numbers for tax purposes`

[13:17:00.0000] <timeless>
odd about the +60 bit

[13:17:01.0000] <timeless>
they didn't do that for me in finland

[13:17:02.0000] <jgraham>
Dunno about finland but the Swedish one is for more than just tax

[13:17:03.0000] <jgraham>
Although it is essential for that

[13:17:04.0000] <timeless>
oh, in finland it was basically used everywhere

[13:17:05.0000] <timeless>
all services

[13:17:06.0000] <jgraham>
They get very confused if you try presenting ID that doesn't have a personnummer on it

[13:17:07.0000] <timeless>
in canada you're supposed to be able to limit it to tax associated items

[13:18:00.0000] <jgraham>
Even when it is legally required that the other id is valid

[13:18:01.0000] <timeless>
the only other thing i could use in finland was my passport

[13:18:02.0000] <jgraham>
e.g. because it is an EU passport

[13:18:03.0000] <timeless>
what was the other one?

[13:19:00.0000] <jgraham>
(I believe it is the case that under EU law it muct be possible to use any EU passport in a circumstance where a passport from a single member state is valid)

[13:19:01.0000] <jgraham>
(but I may be misinformed)

[13:20:00.0000] <Ms2ger>
Time for a single EU passport, with a single database to track anything any government knows about you? :)

[13:20:01.0000] <gsnedders>
jgraham: It's more general than that: you're not allowed to discriminate against citizens of other member states compared with your own

[13:21:00.0000] <gsnedders>
(There are a few exemptions for a limited time upon new member states joining)

[13:21:01.0000] <erlehmann>
AFAIK, in germany, the constitutional court banned globally unique numbers. now we have a tax id. haha.

[13:22:00.0000] <jgraham>
gsnedders: Well it's not really clear to what extent that is actually true e.g. they are supposedly introducing some laws here to make it harder to be a teacher if you don't have a Swedish qualification

[13:22:01.0000] <erlehmann>
(globally unique numbers for tracking purposes, i mean)

[13:22:02.0000] <jgraham>
Which seems pretty suspect to me

[13:22:03.0000] <jgraham>
But I guess they could claim that anyone could get such a qualification, or something

[13:22:04.0000] <timeless>
erlehmann: sounds like the us

[13:23:00.0000] <gsnedders>
jgraham: Anyone can get the qualification is the theory there.

[13:23:01.0000] <timeless>
jgraham: iirc Belgium made it pretty much impossible to teach French if you weren't a native French speaker

[13:23:02.0000] <timeless>
(or was that if you weren't a Native and a French speaker?)

[13:24:00.0000] <gsnedders>
jgraham: It is the reason why Scottish universities don't charge tuition fees for anyone from the EU except for the rest of the UK, due to that law and its interaction with the Scotland Act, which forbids the Scottish Goverment to pass an act that would pay for the rest of the UK.

[13:25:00.0000] <erlehmann>
timeless, germany has a quite strong “data protection” culture and laws. fun fact: since the people responsible for data protection and freedom of information are often the same, the latter is lacking considerably.

[13:25:01.0000] <timeless>
/me chuckles

[13:28:00.0000] <AryehGregor>
Heh, Israel is totally different on all of these things.

[13:28:01.0000] <AryehGregor>
Walking through metal detectors is often a multiple-times-per-day occurrence.

[13:28:02.0000] <AryehGregor>
I think everyone has to carry photo ID at all times.

[13:29:00.0000] <timeless>
AryehGregor: sadly i had that in Budapest and other places

[13:29:01.0000] <AryehGregor>
Stuff like that.

[13:29:02.0000] <timeless>
but that was Yom Kippur

[13:29:03.0000] <timeless>
(each time i entered a synagogue for services)

[13:30:00.0000] <erlehmann>
timeless, this wikipedia entry (literally “data protection commissioner”, i think) does not even seem to have an equivalent in any other language. and it is loooong. <https://secure.wikimedia.org/wikipedia/de/wiki/Datenschutzbeauftragter>

[13:30:01.0000] <AryehGregor>
Yeah, Germans seem to be extremely concerned with privacy.

[13:30:02.0000] <TabAtkins>
We'd call them the "privacy czar" in America.

[13:30:03.0000] <timeless>
> Israeli law requires every permanent resident above the age of 16, whether a citizen or not, to carry an identification card called te'udat zehut (Hebrew: תעודת זהות‎) in Hebrew or biţāqat huwīya (بطاقة هوية) in Arabic.

[13:30:04.0000] <erlehmann>
yeah, those hollerith machines seem to have hit a nerve there.

[13:30:05.0000] <AryehGregor>
I'm not concerned about privacy at all, so I'll be a good fit for Israel.  :)

[13:31:00.0000] <AryehGregor>
(not that Israelis aren't concerned with privacy, but they're pragmatic . . .)

[13:31:01.0000] <AryehGregor>
timeless, I've heard it claimed tourists are supposed to carry photo ID too, although no one ever officially told me that when I was a tourist.

[13:31:02.0000] <timeless>
AryehGregor: what do haredi do if they don't have an eruv available?

[13:31:03.0000] <timeless>
/me reads http://en.wikipedia.org/wiki/Teudat_Zehut

[13:32:00.0000] <AryehGregor>
timeless, dati and haredi are the same here.  I assume they just don't carry ID, and are allowed to get away with it.

[13:32:01.0000] <AryehGregor>
My guess is that theoretically the worst they can do to you is make you go home and show them your ID there or something.

[13:32:02.0000] <timeless>
/me unfortunately has a hard time holding the meaning of `dati` and some other one

[13:32:03.0000] <erlehmann>
TabAtkins, even if a company is required to have one? i thought “XXX czar” is reserved for federal overlords?

[13:32:04.0000] <timeless>
/me uses haredi in place since it has a meaning

[13:32:05.0000] <AryehGregor>
I'm sure there's a special exception for religious observance.

[13:33:00.0000] <TabAtkins>
erlehmann: Oh, didn't realize that was a company thing.

[13:33:01.0000] <AryehGregor>
Dati just means religious.

[13:33:02.0000] <timeless>
> Criminal offence carries a 5,000 Old Israeli shekel fine for not carrying an identity card or for misuse of the document (in 1983 prices, which equal about 1,400 NIS today).

[13:33:03.0000] <AryehGregor>
erlehmann, "czar" is a word that newspapers use in headlines to save space.

[13:33:04.0000] <timeless>
yeah, i can't translate `dati [religious]` into my conceptual map

[13:33:05.0000] <timeless>
if i go to services friday night, saturday morning, saturday night, sunday morning, and tonight

[13:33:06.0000] <timeless>
does that make me dati?

[13:34:00.0000] <erlehmann>
TabAtkins, it is both. as i understand it companies have to have people responsible for this. local and federal authorities have to have to.

[13:34:01.0000] <TabAtkins>
I'd say it makes you religious.

[13:34:02.0000] <timeless>
ok, and that strictly translates to dati?

[13:34:03.0000] <timeless>
/me needs to plan for tonight

[13:35:00.0000] <timeless>
there's the place i want to go, the place i'd like to go later, and a place i was asked to go to ensure they have a minyan so people can say kaddish

[13:35:01.0000] <timeless>
s/dd/d/

[13:35:02.0000] <timeless>
(and i haven't unpacked enough boxes to find my canvas shoes)

[13:35:03.0000] <AryehGregor>
"Chiloni" means you don't believe in or care about religion at all.  "Masorti" means you're sympathetic to religion and observe some practices at least when they suit you, maybe including keeping kosher and praying, but aren't fully observant (e.g., don't observe the Sabbath or conjugal laws or something).  "Dati"/"charedi" means you're religious, you theoretically try to keep all of halacha.  The difference between "dati" and "charedi" is so

[13:35:04.0000] <AryehGregor>
cial: datim are a part of the secular world, charedim aren't.

[13:36:00.0000] <AryehGregor>
E.g., datim generally serve in the military, go to college, and get jobs, while charedim don't.

[13:36:01.0000] <timeless>
right

[13:36:02.0000] <timeless>
/me should be able to memorize that

[13:36:03.0000] <erlehmann>
TabAtkins, i believe that clubs and companies that have at least 10 people who work with critical data need a person responsible for data protection.

[13:36:04.0000] <TabAtkins>
erlehmann: Interesting.

[13:36:05.0000] <timeless>
i'm masorti, and i know that, but i haven't yet learned that it actually fits into that map

[13:36:06.0000] <AryehGregor>
So for something like carrying on Shabbat, there's no real difference between charedi and dati.  Anyone who would knowingly carry on Shabbat without an eruv is almost surely masorti.

[13:37:00.0000] <timeless>
<chiloni-masorti-dati-haredi>

[13:37:01.0000] <timeless>
s/har/char/

[13:37:02.0000] <timeless>
erlehmann: wow

[13:37:03.0000] <AryehGregor>
"Chiloni" is from "chol", "secular".  "Masorti" is from "mesorah", "tradition".  "Dati" is from "dat", "religion".  "Charedi" is from "charad", "to tremble".

[13:37:04.0000] <erlehmann>
timeless, read the article. i am no lawyer, but i think it is pretty interesting.

[13:38:00.0000] <AryehGregor>
Germans are crazy.  :(

[13:38:01.0000] <timeless>
yes

[13:38:02.0000] <erlehmann>
even privacy-unrelated stuff is discussed like this in german culture, so it sometimes gets somewhat silly. google street view is pixelated at some places because residents felt uneasy.

[13:38:03.0000] <Philip`>
People are crazy

[13:38:04.0000] <timeless>
AryehGregor: yeah... i should be able to hold that

[13:38:05.0000] <timeless>
but the problem is that i learned these concepts as English-phonetic-sounds long after i stopped using Hebrew for general things

[13:39:00.0000] <erlehmann>
AryehGregor, not necessarily. curtailing specific powers certainly prevents abuse.

[13:39:01.0000] <timeless>
and thus my mapping for them is poor

[13:39:02.0000] <timeless>
whereas i actually do know the hebrew words...

[13:39:03.0000] <AryehGregor>
Dati vs. charedi is like Modern Orthodox vs. yeshivish/ultra-Orthodox.

[13:40:00.0000] <timeless>
yeah, that last one is also a problem

[13:40:01.0000] <erlehmann>
fun fact: a friend of mine, plomlompom, is writing a (post-)privacy book. i'm curious if it will cause a shitstorm in privacy-related discussions when it comes out in one or two months. ;)

[13:40:02.0000] <timeless>
since i never really got an introduction/explanation of what modern-orthodox is

[13:40:03.0000] <timeless>
so i have a hard time drawing/understanding that line

[13:40:04.0000] <timeless>
i can easily understand reform/conservative/orthodox

[13:40:05.0000] <timeless>
because i grew up with those concepts

[13:41:00.0000] <timeless>
erlehmann: presumably in German?

[13:41:01.0000] <Philip`>
People who are concerned about their privacy on Google Street View should print out life-sized cardboard copies of themselves and leave them all over the town, so that nobody can tell where the real them is

[13:41:02.0000] <timeless>
heh

[13:41:03.0000] <timeless>
Philip`: they'd get grafit

[13:41:04.0000] <timeless>
i'd

[13:42:00.0000] <Philip`>
It's pretty much impossible to hide information nowadays, but you can overload people with bogus data so they can't find the real information

[13:42:01.0000] <erlehmann>
timeless, indeed. but it also includes a quite interesting short history of privacy, so it might be usable for other cultures as well.

[13:42:02.0000] <Ms2ger>
They could graffiti themselves

[13:42:03.0000] <erlehmann>
title is „prima leben ohne privatsphäre“

[13:45:00.0000] <erlehmann>
AryehGregor, i think for me the “germans are crazy” thingy can be exemplified by the fact that in sweden, Stuff Your Gov Does™ is public by default until classified otherwise – whereas in germany, it is not until classified otherwise. or so i heard.

[13:45:01.0000] <AryehGregor>
Germany has to have *some* kind of classified information, otherwise they couldn't (for instance) have diplomatic negotiations with other countries.

[13:46:00.0000] <TabAtkins>
erlehmann didn't say Sweden didn't have classified information, just that it's not the default.

[13:46:01.0000] <AryehGregor>
He said Germany didn't, unless I misread him.

[13:47:00.0000] <TabAtkins>
You misread.  He said it's secret until classic otherwise.

[13:47:01.0000] <TabAtkins>
"not" meaning "not public"

[13:47:02.0000] <TabAtkins>
s/classic/classified/

[13:47:03.0000] <erlehmann>
AryehGregor, you misread. sweden has freedom of information. germany has, well, some.

[13:47:04.0000] <AryehGregor>
Oh.

[13:47:05.0000] <erlehmann>
meaning: not much.

[13:47:06.0000] <AryehGregor>
That's backwards from what I'd expect.

[13:47:07.0000] <AryehGregor>
Or, hmm, maybe not.

[13:48:00.0000] <erlehmann>
did i make a grammar error?

[13:48:01.0000] <AryehGregor>
No, I was just confused by context.

[13:48:02.0000] <AryehGregor>
But seriously, someone once sued Wikimedia Deutschland because the English Wikipedia published the names of two convicted German murderers.  Apparently under German law you aren't allowed to publicize the fact that someone committed murder after they've served their sentence.

[13:49:00.0000] <AryehGregor>
It's a violation of their privacy.

[13:49:01.0000] <AryehGregor>
. . .

[13:49:02.0000] <erlehmann>
yeah, i think the reasoning is because it damages their social standing more than intended by the law.

[13:50:00.0000] <AryehGregor>
(they got the judge to make Wikimedia Deutschland take down its wikipedia.de redirect, so it put up a page saying "We had to take down this redirect because of court ruling X.  Please donate!", so the plaintiffs changed their mind and decided they were okay with the redirect after all)

[13:50:01.0000] <AryehGregor>
(this is why no Wikimedia data is hosted anywhere outside the United States, and the only non-data servers that are hosted outside the United States are in the Netherlands)

[13:51:00.0000] <Hixie>
the US has a "default-free" mode too, as far as i can tell, but that doesn't stop the US classifying everything under the sun

[13:51:01.0000] <AryehGregor>
Nowhere in the world has free-speech guarantees close to the U.S.'s, that's for sure.

[13:52:00.0000] <AryehGregor>
Hixie, actually I don't think that much is actually classified gratuitously.  It's a pain to request documents, but if you really want to pursue it you can take it to court and everything, AFAIK.

[13:52:01.0000] <AryehGregor>
Of course, there are still a bunch of exceptions, but you can get a fair hearing by a judge and everything, if you want.

[13:52:02.0000] <Hixie>
yeah maybe the "default-free" is really "default-hidden-until-requested"

[13:52:03.0000] <AryehGregor>
(and are willing to spend the time and money)

[13:53:00.0000] <AryehGregor>
Right, that's probably a better description.

[13:53:01.0000] <TabAtkins>
AryehGregor: Money isn't really an issue - FOIA requests are free, I think.

[13:53:02.0000] <TabAtkins>
Time, definitely, becasue they'll stall sometimes.

[13:53:03.0000] <AryehGregor>
I mean, nobody's going to say you're allowed to wander in to a government office and browse the files on some random computer.

[13:53:04.0000] <AryehGregor>
TabAtkins, it's probably free to file them, but if they deny it and you want to challenge it, I'm pretty sure you'll need a lawyer.

[13:54:00.0000] <AryehGregor>
/me looks it up

[13:54:01.0000] <Hixie>
what's a better word than "patron" or "customer" for the user of a shared worker

[13:54:02.0000] <TabAtkins>
That's true, yes.  But they can't deny unless it's classified.

[13:54:03.0000] <TabAtkins>
Hixie: "leech"

[13:54:04.0000] <Hixie>
context: "inside the shared worker, new ...s are announced using the 'connect' event."

[13:54:05.0000] <TabAtkins>
"followers"

[13:54:06.0000] <erlehmann>
there also was a boxer called “neger-kalle” (literally: “nigger kalle”) who (successfully, i think) sued a newspaper because they had it in the archives and now he says it is racist.

[13:55:00.0000] <erlehmann>
freedom of information is a different thing than free speech.

[13:56:00.0000] <AryehGregor>
Yes, that's true.

[13:56:01.0000] <AryehGregor>
Well, they're related.

[13:56:02.0000] <AryehGregor>
E.g., free speech guarantees in the United States shield people who leak confidential information to at least some extent.

[13:56:03.0000] <Hixie>
someone tell manning that

[13:56:04.0000] <erlehmann>
super cool story, bro.

[13:56:05.0000] <gsnedders>
AryehGregor: Most countries have laws about whistleblowers. Most places will still find some technically legal way to screw you over.

[13:57:00.0000] <erlehmann>
what hixie said.

[13:57:01.0000] <AryehGregor>
Er, sorry.

[13:57:02.0000] <AryehGregor>
By "leak" I really meant "publish".

[13:57:03.0000] <AryehGregor>
The people who actually leak it aren't protected at all, that I know of.

[13:57:04.0000] <erlehmann>
shared workers … applicants!

[13:57:05.0000] <Hixie>
/me went with "patrons"

[13:57:06.0000] <AryehGregor>
I was talking about, e.g., newspapers that republish the leaked material.

[13:57:07.0000] <AryehGregor>
Or websites.

[13:57:08.0000] <erlehmann>
“inside the shared worker, new applicants are announced using the 'connect' event.” :3

[13:57:09.0000] <erlehmann>
oh. okay ._.

[13:57:10.0000] <gsnedders>
AryehGregor: Depends if it is classified or not. If it's not classified, in most places they theoretically do. Just see the above.

[13:58:00.0000] <AryehGregor>
Although I don't know, maybe they could convict them but are afraid of the PR backlash.

[13:58:01.0000] <erlehmann>
Hixie, that also sounds like black magic!

[13:58:02.0000] <AryehGregor>
/me shrugs

[13:58:03.0000] <Hixie>
erlehmann: yeah none of these terms are great

[13:59:00.0000] <tomasf>
how about "clients"?

[13:59:01.0000] <AryehGregor>
Anyway, it would certainly be protected by free speech in some cases.  Look at New York Times Co. v. United States, re the Pentagon papers.

[13:59:02.0000] <Hixie>
AryehGregor: you mean the way that US gov't officials have said assange should be assassinated?

[13:59:03.0000] <Hixie>
tomasf: that's better. thanks.

[13:59:04.0000] <tomasf>
np

[14:00:00.0000] <Ms2ger>
"Magician"

[14:00:01.0000] <AryehGregor>
Hixie, I said it provided some protection.  I don't think anything's going to protect Assange, his leaks were much too indiscriminate.

[14:00:02.0000] <Hixie>
ah, free speech, but only if you discriminate, ok

[14:01:00.0000] <AryehGregor>
Well, no one said free speech is unqualified.  There are limitations.

[14:01:01.0000] <karlcow>
I have a name for the patrons but that will not work in that community

[14:01:02.0000] <Hixie>
really?

[14:01:03.0000] <Hixie>
/me looks at the first amendment

[14:02:00.0000] <TabAtkins>
AryehGregor: The exceptions are very, very limited.  Wikileaks does not fall into any of them.

[14:02:01.0000] <karlcow>
hmm dictionary tells "customer, client, frequenter, consumer, user, visitor, guest; informal regular, habitué."

[14:02:02.0000] <Hixie>
i don't see any limitations in the text of the constitution

[14:02:03.0000] <Hixie>
must be one of those "interpretation" things

[14:03:00.0000] <Hixie>
i guess assassinations wouldn't be covered, since those wouldn't count as laws?

[14:03:01.0000] <TabAtkins>
Basically, if your speech is threatening, libelous, was done knowingly to cause harm, or violates copyright.

[14:03:02.0000] <AryehGregor>
Well, yes, the Constitution wasn't really meant to be interpreted in isolation from the entire tradition of common law that preceded it.

[14:04:00.0000] <AryehGregor>
TabAtkins, under court precedent, the state can impose any restriction on free speech provided it can show that a) there's a compelling state interest, b) the restriction is narrowly tailored to serve that interest.

[14:04:01.0000] <AryehGregor>
It's strict scrutiny, that's all.

[14:04:02.0000] <AryehGregor>
There are certain exceptions where the state isn't even subject to strict scrutiny, based on the traditional application of free speech historically in America, like obscenity.

[14:05:00.0000] <gsnedders>
AryehGregor: So basically they can argue any leak of confidentical material puts state security at risk?

[14:05:01.0000] <Hixie>
AryehGregor: you're saying the constitution is more like HTML4 than HTML5? :-P

[14:05:02.0000] <AryehGregor>
gsnedders, they can argue that.  It's up to the judges to decide on a case-by-case basis.  I'm pretty sure judges are not going to say that Assange's leaks are fully protected, if it came to a trial.

[14:06:00.0000] <zewt>
free speech means "free speech unless someone finds it offensive" these days

[14:06:01.0000] <AryehGregor>
Hixie, sheesh, actual *laws* are more like HTML4 than HTML5.  The *Constitution* is like HTML4 with all the precision removed.

[14:06:02.0000] <Hixie>
hehe

[14:06:03.0000] <AryehGregor>
If you tried to apply the Constitution literally, you'd have all sorts of crazy and totally unintended consequences.

[14:06:04.0000] <jgraham>
/me wonders if Hixie was responsible for the "server/slave" terminology in the Opera test systems

[14:07:00.0000] <AryehGregor>
Example argument: since the right to keep and bear arms shall not be infringed, all American citizens have the right to own and operate nuclear weapons.

[14:07:01.0000] <Hixie>
jgraham: me or allan, i expect

[14:07:02.0000] <jgraham>
Hixie: Well that's what I expect too :p

[14:07:03.0000] <AryehGregor>
Also: since there's a right to freedom of religion, I am free to commit murder as long as it's required by my religion.

[14:07:04.0000] <Hixie>
jgraham: the stargate references, if they still exist, are all allan and tom

[14:07:05.0000] <TabAtkins>
AryehGregor: You'll take my yellowcake when you can pry it from my cold (or rather, hot) dead fingers.

[14:08:00.0000] <AryehGregor>
TabAtkins, shouldn't be too long if you're not wearing protective gear.

[14:08:01.0000] <Hixie>
AryehGregor: yes, that's what the 2nd ammendment says. Also, I think we should repeal that ammendment. :-)

[14:08:02.0000] <gsnedders>
Hixie: Stargate references?

[14:08:03.0000] <zewt>
can we just get obscenity laws applied to the public practice of religion

[14:08:04.0000] <AryehGregor>
And freedom of religion too, no doubt.

[14:08:05.0000] <zewt>
(them's fightin' words)

[14:08:06.0000] <jgraham>
I wouldn't know a stargate reference if it ate my face

[14:08:07.0000] <gsnedders>
Hixie: (Having never seen Stargate I've probably not noticed them)

[14:09:00.0000] <jgraham>
Now we have portal references...

[14:09:01.0000] <TabAtkins>
AryehGregor: Are you kidding?  Freedom of religion is the only thing that stops the dominionist christains from taking over.

[14:09:02.0000] <zewt>
let's use one etching-away-of-constitutional-freedoms to attack another constitutional freedom, just for irony's sake

[14:09:03.0000] <AryehGregor>
Also: since the Thirteenth Amendment bans involuntary servitude, the government is not allowed to make anyone do anything at all, such as go to school.

[14:09:04.0000] <gsnedders>
And Hello Kitty references.

[14:09:05.0000] <Hixie>
AryehGregor: yeah, the lack of a limitation of rights to the extent that they infringe on other people's even more fundamental rights is problematic

[14:09:06.0000] <AryehGregor>
TabAtkins, you could remove it and replace it with something else, like mandating a completely secular state (which America never was).

[14:10:00.0000] <Hixie>
AryehGregor: (i.e. the constitution doesn't say how to resolve the case of two conflicting religions)

[14:10:01.0000] <Hixie>
gsnedders, jgraham: the machine names were all named after stargate characters originally

[14:10:02.0000] <Hixie>
gsnedders, jgraham: for spartan, that is

[14:10:03.0000] <AryehGregor>
Hixie, but this stuff was all *extensively* discussed by philosophers and political thinkers around the time the Constitution was written.  There was a common understanding of what a lot of this stuff meant.

[14:11:00.0000] <AryehGregor>
It has to be understood in that context.

[14:11:01.0000] <jgraham>
Hixie: Oh. Well I think every machine now gets named with a different naming scheme

[14:11:02.0000] <TabAtkins>
AryehGregor: The state *is* mandated to be secular by the first amendment.

[14:11:03.0000] <gsnedders>
Hixie: Ah. Some of them still exist.

[14:11:04.0000] <Hixie>
AryehGregor: just like html4. and we all know how well that went. :-)

[14:11:05.0000] <Hixie>
anyway

[14:11:06.0000] <gsnedders>
Hixie: Like, the really old slaves. :)

[14:11:07.0000] <Hixie>
/me concentrates on bugs

[14:12:00.0000] <AryehGregor>
TabAtkins, is not.  It says that Congress shall pass no law restricting the free exercise of religion.  The government could be explicitly Christian as long as it doesn't stop people from practicing other religions.  In fact, that was basically the status quo until sometime in the mid to late 19th century.

[14:12:01.0000] <AryehGregor>
Plus, the First Amendment never applied to the states at all until after the Civil War.  There were states that said only Christians could vote.

[14:13:00.0000] <gsnedders>
Hixie: Allan ended up using Pokémon when far more slaves were acquired, FWIW. But even those are massively outnumbered.

[14:13:01.0000] <gsnedders>
*now

[14:13:02.0000] <TabAtkins>
AryehGregor: No, it must "make no law respecting an establishment of religion".  An explicitly religious government would, by necessity, have laws respecting the state religion.

[14:13:03.0000] <Hixie>
ah

[14:13:04.0000] <gsnedders>
Hixie: Probably should've just kept that going, now have about ten different naming schemes :(

[14:13:05.0000] <Hixie>
hah

[14:14:00.0000] <jgraham>
Like I said, there is a unique naming scheme per machine

[14:14:01.0000] <jgraham>
Or the unifyinf concepts of "words"

[14:14:02.0000] <jgraham>
s/f/g/

[14:14:03.0000] <AryehGregor>
TabAtkins, well, no, the federal government wasn't allowed to have a state religion along the lines of the Anglican Church.  It still didn't have to be secular in the sense of being neutral with respect to religion, or certainly not neutral with respect to atheism.  When were atheists first allowed to testify in Court in the US?

[14:15:00.0000] <gsnedders>
jgraham: That's unfair. There are only 12.

[14:15:01.0000] <TabAtkins>
There were no federal laws preventing atheists from testifying.

[14:15:02.0000] <gsnedders>
(If I counted correctly.)

[14:15:03.0000] <AryehGregor>
In federal court?

[14:16:00.0000] <AryehGregor>
There were definitely *states* that didn't give anyone but Christians full rights.

[14:16:01.0000] <TabAtkins>
Yes, until the courts began applying the 1st more widely.

[14:16:02.0000] <AryehGregor>
And the whole Bill of Rights.  Incorporation.

[14:17:00.0000] <karlcow>
this is the part of USA I have difficulty to understand. I guess it's education.

[14:18:00.0000] <dbaron>
http://en.wikipedia.org/wiki/Fourteenth_Amendment_to_the_United_States_Constitution#Incorporation

[14:18:01.0000] <karlcow>
The inclusion of religion in many parts of it

[14:19:00.0000] <dbaron>
In the period when the supreme court was most aggressive about enforcing rights, they didn't particularly care about enforcing freedom of religion.

[14:19:01.0000] <dbaron>
In fact, Earl Warren (Chief Justice) previously won a campaign for Governor of California by basically attacking his opponent for athiesm.

[14:20:00.0000] <jgraham>
karlcow: They made a fundamental mistake with the US. What you really want in a country is an official religion that is rather hands off. The church of England is practically perfect in this regard. Only the queen is expected to be religious and the rest of us can happily ignore it

[14:21:00.0000] <Hixie>
speak for yourself :-)

[14:21:01.0000] <jgraham>
By not forcing people to construct their identities by making positive choices around their religion it is much easier to slip into casual atheism

[14:21:02.0000] <zewt>
can we just opt for an athiest government

[14:21:03.0000] <dbaron>
karlcow, btw, many of the things that include religion were done in the early 1950's.

[14:21:04.0000] <AryehGregor>
/me will soon be moving to a country where ~20% of the parliament is pro-theocracy

[14:21:05.0000] <karlcow>
dbaron: ah interesting. I didn't know

[14:22:00.0000] <TabAtkins>
dbaron: Damn the commies!

[14:22:01.0000] <dbaron>
karlcow, things like "In God We Trust" on the money, etc., from the early 1950's.

[14:22:02.0000] <Hixie>
sicking: re bug 538142 -- i would take one implementation as an indication that the direction of implementations was that way, and change the spec accordingly.

[14:22:03.0000] <AryehGregor>
/me cannot easily find any sources about atheists' eligibility for testifying in court

[14:22:04.0000] <Hixie>
sicking: my goal is to have the spec slightly ahead of implementations, and going in the same direction.

[14:23:00.0000] <AryehGregor>
dbaron, on the other hand, things like banning prayer in public schools or banning public government displays that include religion are even more recent, and AFAIK have no precedent before the last few decades.

[14:23:01.0000] <sicking>
Hixie: well, given that currently 4 browsers have said they are not interested, i wouldn't take one chainging their mind as a sign that the 3 others would

[14:23:02.0000] <dbaron>
AryehGregor, Have we actually done the latter?

[14:23:03.0000] <AryehGregor>
"In God We Trust" appeared on US coins since 1864, by the way, according to Wikipedia.

[14:23:04.0000] <sicking>
changing

[14:23:05.0000] <TabAtkins>
AryehGregor: Prayer has never been banned in schools.  Prayer at public events, on the other hand, has.

[14:24:00.0000] <gsnedders>
FWIW, in Scotland, the Church of Scotland is by law a separate entity from the state entirely, and neither can interfere with each other. This does actually come up every so often, and there is even more separation than in England.

[14:24:01.0000] <AryehGregor>
TabAtkins, I mean school-organized prayer, like in the classroom.

[14:24:02.0000] <AryehGregor>
Or school Bible readings.

[14:24:03.0000] <TabAtkins>
AryehGregor: Yes, that's certainly been banned widely in the last few decades.

[14:24:04.0000] <AryehGregor>
My father went to public school and they still had Bible readings.

[14:24:05.0000] <dbaron>
AryehGregor, ah, ok, but it wasn't required on coins until 1956

[14:24:06.0000] <TabAtkins>
Under the (reasonable) doctrine that public school is an accessory of the government, and must obey the same restrictions.

[14:25:00.0000] <gsnedders>
(For the curious, look up the Church of Scotland Act)

[14:25:01.0000] <Hixie>
sicking: typically when one browser does something (especially mozilla) it tends to influence the others, so it's not that simple. :-)

[14:25:02.0000] <AryehGregor>
dbaron, the government can't support public displays of religion in a way that prefers one religion over another, or something like that.  Like here: http://articles.latimes.com/2005/jun/28/nation/na-scotus28

[14:25:03.0000] <Hixie>
sicking: anyway this seems largely academic since you don't want to do it anyway :-)

[14:26:00.0000] <sicking>
Hixie: asking also works sometimes...

[14:26:01.0000] <AryehGregor>
TabAtkins, there were no bans on any government-organized prayer or Bible readings before the last few decades, AFAIK.  Obviously, students who objected could always opt out.

[14:26:02.0000] <AryehGregor>
/me has done *nothing* useful today :(

[14:26:03.0000] <sicking>
Hixie: by the way, i think the SharedWorker privacy bug from MS is them not understanding that workers are only shared same-origin

[14:26:04.0000] <TabAtkins>
AryehGregor: Yeah, I know the doctrine is relatively recent.

[14:27:00.0000] <Hixie>
sicking: asking what?

[14:27:01.0000] <karlcow>
gsnedders: same thing in France for separation church/states

[14:27:02.0000] <Ms2ger>
sicking, sounds like a usual MS bug, then

[14:27:03.0000] <Hixie>
sicking: i think it's more about the same kind of thing as third-party cookies

[14:27:04.0000] <karlcow>
http://en.wikipedia.org/wiki/1905_French_law_on_the_Separation_of_the_Churches_and_the_State

[14:27:05.0000] <Hixie>
sicking: they're talking specifically about iframes from one origin hosted on two sites from two other origins

[14:27:06.0000] <Hixie>
sicking: and preventing those iframes from communicating to each other

[14:27:07.0000] <sicking>
Hixie: ah, could be

[14:27:08.0000] <gsnedders>
karlcow: Scotland is somewhat tarnished by the Church of England having bishops in the House of Lords, though

[14:28:00.0000] <Hixie>
sicking: which seems like it could be a valid point, but mostly seems like closing the barn door long after the horse has bolted

[14:28:01.0000] <sicking>
Hixie: actually, indexedDB doesn't suffer this problem, at least in FF

[14:29:00.0000] <sicking>
Hixie: we don't let you open indexedDB databases unless the whole frame parent chain is same-origin

[14:30:00.0000] <Hixie>
sicking: ah. lame. :-P

[14:30:01.0000] <Hixie>
sicking: so you can't, e.g., have an iframe that provides a service backed by a database?

[14:33:00.0000] <jgraham>
gsnedders: I am much more upset by the strongly religious in the house of commons than in the house of lords

[14:33:01.0000] <jgraham>
(cough cough Tony Blair cough)

[14:36:00.0000] <jgraham>
(because it seems to offer a convenient get out clause from having the humility and introspection to admit when you have made a horrendous mistake)

[14:37:00.0000] <jgraham>
("God will be my judge in Iraq" and all that)

[14:37:01.0000] <jgraham>
*on

[14:42:00.0000] <sicking>
Hixie: currently no

[14:43:00.0000] <Hixie>
sicking: wow, that cuts off a whole set of use cases, interesting

[14:43:01.0000] <Hixie>
sicking: i guess it does mean you don't have to worry abotu privacy issues, at least

[14:43:02.0000] <sicking>
Hixie: you're the first to complain

[14:43:03.0000] <Hixie>
sicking: so if someone wants to provide a shared service to multiple origins, it has to store the data server-side

[14:43:04.0000] <Ms2ger>
And to notice, perhaps?

[14:44:00.0000] <sicking>
Hixie: we'll probably expand what's allowed, but keeping privacy in mind

[14:44:01.0000] <sicking>
Hixie: sure, but server-side doesn't work if 3rd party cookies are disabled

[14:44:02.0000] <Hixie>
(very few people are providing this kind of shared service so far)

[14:44:03.0000] <Hixie>
sicking: well we both know that's a myth in reality, but sure, from a PR standpoint

[14:45:00.0000] <sicking>
Hixie: safari disabled 3rd party by default

[14:45:01.0000] <sicking>
Hixie: we're actually trying to fix the privacy situation, that's why we're avoiding digging deeper

[14:46:00.0000] <Hixie>
you could drop cookies entirely, and it wouldn't help at all

[14:46:01.0000] <gsnedders>
sicking: We disabled 3rd party by default in 10.50, but it broke major sites (mainly in, e.g., Russia, where Safari has almost no marketshare)

[14:46:02.0000] <Hixie>
there's far too many bits of unique data per user to ever stop fingerprinting

[14:50:00.0000] <Hixie>
(i mean in particular in this shared worker case, instead of a shared worker to correlate sessions, you could just do an xhr to a remote host -- the IP address plus the HTTP headers of the request would be more than sufficient to associate the requests as being from the same session)

[14:50:01.0000] <Hixie>
so in other news, apparently mozilla prevents ws:// websockets from being opened from https:// pages. Which I guess makes sense if the socket is to something sensitive used by the page

[14:51:00.0000] <Hixie>
doesn't really make sense if the target websocket is some third-party trivial thing like a clock service or a stock ticker or something

[14:51:01.0000] <Hixie>
should we put that restriction in the spec?

[14:54:00.0000] <Philip`>
Pages might treat the output of the service as HTML and insert it directly into their content, which would be an XSS hole if the output could be tampered with by an attacker, and HTTPS is meant to prevent such tampering

[14:55:00.0000] <Hixie>
oh it's certainly possible to shoot yourself in the foot if we allow https:// to ws:// connections

[14:55:01.0000] <erlehmann>
offtopic: any one of you at the chaos communication camp?

[14:55:02.0000] <Hixie>
the questions is whether we should be protecting them here, given that there are valid and secure use cases for the feature

[14:55:03.0000] <Hixie>
question

[14:56:00.0000] <Hixie>
i guess i'll put the restriction in and see if anyone complains

[14:56:01.0000] <erlehmann>
:)

[14:57:00.0000] <Hixie>
man, i still write "first script" instead of "entry script"

[14:57:01.0000] <Hixie>
all. the. time.

[14:57:02.0000] <Hixie>
yay for my preprocessor's checks

[14:57:03.0000] <timeless>
Hixie: what was wrong with `clients` for `inside the shared worker, new .... are announced using the 'connect' even.`?

[14:57:04.0000] <zewt>
python's backwards "traceback" silliness made me start typing "backflash" instead of flashback

[14:58:00.0000] <Hixie>
timeless: nothing, that's what i used

[14:58:01.0000] <timeless>
oh good, tomasf ++ suggested it

[15:00:00.0000] <smaug____>
annevk5: ping

[15:03:00.0000] <jgraham>
smaug____: I believe annevk5 is away

[15:06:00.0000] <timeless>
/me heads out for the day or something

[15:06:01.0000] <smaug____>
jgraham: ok, thanks

[15:06:02.0000] <smaug____>
there should be some status page for "all" the browser/spec developers

[15:06:03.0000] <Hixie>
Philip`: if you could contribute your insight on the canvas drawimage thread on whatwg that would be awesome

[15:07:00.0000] <Hixie>
Philip`: especially if you can convincingly argue for not changing the spec :-)

[15:07:01.0000] <TabAtkins>
If I had some engineers interested in putting together a "report a problem to the developer API" that could be invoked and then would give the user a browser-generated dialog allowing them to choose to submit a comment, machine details, and screenshots to the page author, where would be best to have them propose this?

[15:07:02.0000] <smaug____>
hsivonen on vacation, annevk5 away, bz soon on vacation ...

[15:07:03.0000] <Hixie>
smaug____: august sucks

[15:07:04.0000] <Hixie>
smaug____: and anne is just a slacker

[15:07:05.0000] <Hixie>
i'll be afk for a week or so near the end of the month

[15:08:00.0000] <Hixie>
i don't recall exact dates

[15:08:01.0000] <gsnedders>
smaug____: to be fair, annevk sent an email internally on Friday

[15:08:02.0000] <Hixie>
TabAtkins: unless you find somewhere else, whatwg would be fine

[15:08:03.0000] <smaug____>
in Finland July is the holiday month, elsewhere August

[15:08:04.0000] <smaug____>
not that I have any vacation

[15:09:00.0000] <smaug____>
and for some reason hsivonen is out now

[15:09:01.0000] <smaug____>
gsnedders: :)

[15:09:02.0000] <gsnedders>
smaug____: It's not like we had much notice either\

[15:19:00.0000] <sicking>
Hixie: what do you mean by "Web Sockets by definition are same-origin with whatever script created them"? I would say that it's cross-origin by definition since it always connects to a different scheme

[15:26:00.0000] <Hixie>
sicking: well i guess it depends what you mean by "origin"

[15:27:00.0000] <Hixie>
sicking: on the web you can only access same-origin resources. The way I look at it, CORS and WebSockets get around this by allowing servers to opt-in to being treated as same-origin in specific cases.

[15:28:00.0000] <sicking>
Hixie: that's not how i see CORS

[15:28:01.0000] <sicking>
Hixie: i see it as a way for a server to say "you can read this data even though you are from a different origin"

[15:28:02.0000] <Hixie>
what's the origin of an <img> obtained using CORS?

[15:28:03.0000] <Hixie>
of the image data of an <img>, i mean

[15:29:00.0000] <Hixie>
it's not cross-origin; if it was, it would still taint canvas

[15:29:01.0000] <sicking>
Hixie: it's the origin of the server it was loaded from

[15:29:02.0000] <Hixie>
not according to the spec

[15:29:03.0000] <sicking>
Hixie: i think that makes the definition of "origin" much more complex

[15:29:04.0000] <Hixie>
what's the origin of a document on http://example.com/ if it's loaded in an <iframe sandbox>?

[15:30:00.0000] <Hixie>
not http://example.com/, according to the spec

[15:30:01.0000] <sicking>
Hixie: that i'm not quite sure of. I agree it's not "http://example.com/"

[15:30:02.0000] <Hixie>
what's the origin of an about:blank document? rarely if ever is it "about:blank"

[15:30:03.0000] <sicking>
agreed

[15:30:04.0000] <Hixie>
anyway my point is the "origin" is a tuple that doesn't necessarily match the url of the resource

[15:31:00.0000] <Hixie>
with websockets there isn't really a concept of origin since there's no origin check after tha handshake

[15:31:01.0000] <Hixie>
the

[15:31:02.0000] <sicking>
the origin check is part of the handshake

[15:31:03.0000] <sicking>
similar to CORS

[15:31:04.0000] <Hixie>
the only origin involved in that check is the script's origin

[15:31:05.0000] <Hixie>
the server just has to echo it

[15:31:06.0000] <sicking>
if you consider CORS being part of the handshake for crosssite xhr

[15:32:00.0000] <gsnedders>
Blocks in the normal flow are normally 100% of the containing block?

[15:32:01.0000] <gsnedders>
*with

[15:32:02.0000] <Hixie>
width?

[15:32:03.0000] <sicking>
my point is that it's good to associate a token to the message so that pages can make its own security decision

[15:32:04.0000] <gsnedders>
*width

[15:32:05.0000] <sicking>
...using that token

[15:32:06.0000] <sicking>
usually we can that token "origin"

[15:32:07.0000] <Hixie>
gsnedders: margin+border+padding+width = containing block width assuming it's a display:block and not replaced

[15:33:00.0000] <sicking>
if you want to call it something else and expose it through the .origin property, i guess i can live with that

[15:33:01.0000] <Hixie>
sicking: well they have event.target.URL

[15:33:02.0000] <gsnedders>
Hixie: Right, okay. Where the hell had I got the thought that wasn't that?

[15:33:03.0000] <Hixie>
or .url

[15:33:04.0000] <gsnedders>
/me is blatantly too tired

[15:33:05.0000] <Hixie>
or whatever we call it this week

[15:33:06.0000] <sicking>
Hixie: sure, but for other message events it's available through .origin

[15:34:00.0000] <sicking>
Hixie: that seems suboptimal

[15:34:01.0000] <Hixie>
sicking: for the other message events, there's a real origin :-)

[15:34:02.0000] <sicking>
Hixie: by your definition of origin, sure

[15:34:03.0000] <sicking>
Hixie: if you want to call the security token something other than "origin" i'm fine with that

[15:34:04.0000] <Hixie>
...hm, well, i guess EventSource doesn't have a real origin either actually

[15:34:05.0000] <sicking>
Hixie: as long as it's consitently exposed

[15:34:06.0000] <Hixie>
ok i'll do it like eventsource

[15:35:00.0000] <Hixie>
what was the bug # again so i can include it in the checkin comment?

[15:35:01.0000] <sicking>
Hixie: ...or possibly you should change your definition of "origin" ;-)

[15:35:02.0000] <sicking>
Hixie: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13525

[15:36:00.0000] <Hixie>
thanks

[15:36:01.0000] <sicking>
Hixie: data can still come from a source with a security label, even when it's readable across such security labels

[15:36:02.0000] <sicking>
Hixie: websites will want to build their own security model on top of the one that UAs use. They already do

[15:37:00.0000] <Hixie>
you've already convinced me

[15:40:00.0000] <sicking>
Hixie: Thanks!

[15:40:01.0000] <Hixie>
done

[16:37:00.0000] <jwalden>
/me sees he missed interesting off-(on)topic discussion a couple hours ago


2011-08-09
[21:23:00.0000] <Hixie>
fantastic. i have a pile of code a mile high and it finds one obscure canvas performance bug in webkit and one obscure canvas clearRect(?) bug in gecko.

[21:23:01.0000] <Hixie>
and i basically have no way to make minimised test cases for them.

[21:23:02.0000] <Hixie>
sigh.

[21:24:00.0000] <shetech>
gotta love obscure finds.

[22:45:00.0000] <Hixie>
nice. webkit has a huge slowdown bug with clip() _and_ doesn't support the last argument to fillText().

[22:46:00.0000] <Hixie>
so basically there's no easy way to clip user-provided text in webkit

[22:46:01.0000] <Hixie>
meanwhile, gecko has some crazy bug with clearRect() where it's not actually clearing the whole rect, unless i first call fillRect() on the whole canvas, then it works fine...

[23:35:00.0000] <Hixie>
i'm all for experimental implementations of css features getting prefixes

[23:35:01.0000] <Hixie>
but when EVERYONE implements the feature, can we move on already?

[23:35:02.0000] <Hixie>
https://developer.mozilla.org/en/CSS/-moz-linear-gradient#Cross-browser_gradients is just ridiculous.

[23:36:00.0000] <zewt>
on the other hand, everyone's probably afraid of being That Guy that releases something unprefixed and ends up preventing some fix or another...

[00:02:00.0000] <Hixie>
zewt: imho, the experimental prefixes should only be used with experimental features, not specced things.

[00:02:01.0000] <Hixie>
zewt: the problem is the csswg iterate for far too long instead of freezing fast like we do with the html spec

[00:03:00.0000] <Hixie>
they probably end up with more well-thought-out APIs eventually

[00:03:01.0000] <Hixie>
but in the meantime, we have to suffer through years of silliness :-)

[00:04:00.0000] <Hixie>
in other news, webkit has much prettier 'inset' and 'outset' border colours than firefox and opera

[00:04:01.0000] <zewt>
well, i've seen plenty of complaining on the lists about releasing APIs unprefixed too soon

[00:06:00.0000] <zewt>
with the way specs are developed these days, it might be worth clarifying somehow when the editor considers a spec stable enough for unprefixed implementations

[00:06:01.0000] <zewt>
seems like that's very fuzzy currently

[00:07:00.0000] <zewt>
the websocket thread a week ago comes to mind

[00:07:01.0000] <Hixie>
websocket had the worst of both worlds

[00:07:02.0000] <Hixie>
it didn't have prefixes, _and_ it wasn't fast-freeze

[00:07:03.0000] <Hixie>
the ietf really screwed that one over

[00:08:00.0000] <zewt>
i remember grumbling about blobs during the blob.slice thread, though to me the blob api seemed stable enough by then that it's hard to blame people for not prefixing it

[01:09:00.0000] <janv_>
hixie: I think textContent should be changed to text in "4.11.5.4 Using the option element to define a command"

[04:51:00.0000] <jgraham>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1098 <-- anyone want to explain the behaviour of that in Gecko/WebKit to me?

[04:53:00.0000] <zcorpan>
jgraham: the "2" script is async?

[04:54:00.0000] <jgraham>
Hmm was this the big thread a while ago?

[04:55:00.0000] <jgraham>
/me concludes that hsivonen should not be allowed to take vacation :)

[04:55:01.0000] <zcorpan>
the thread with LABjs?

[04:56:00.0000] <jgraham>
Yeah

[04:57:00.0000] <zcorpan>
i think gecko had the same behavior for this case before that thread, that appendChilded scripts with src are async

[04:57:01.0000] <jgraham>
I thought it was the other way around, that WebKit had this behaviour first and gecko copied it. But maybe I am wrong

[04:57:02.0000] <zcorpan>
could be

[04:58:00.0000] <jgraham>
Either way it doesn't seem to match the spec afaict

[04:58:01.0000] <jgraham>
But I could easilly be missing something

[04:59:00.0000] <jgraham>
Hmm, maybe the force-async flag is relevant

[05:01:00.0000] <Ms2ger>
Yes, it would be

[05:05:00.0000] <jgraham>
So it would fall into the 5th clause of step 15 of the "prepare a script" algorithm. It's not really clear to me why there would be a big difference between the 4th and 5th clauses here though

[05:07:00.0000] <jgraham>
(the difference seems to be that the 4th clause ensures the relative order of appendChild inserted scripts, but the 5th does not)

[05:07:01.0000] <jgraham>
+external

[05:08:00.0000] <Ms2ger>
That's what @async means, no

[05:08:01.0000] <Ms2ger>
?

[05:09:00.0000] <zcorpan>
it wouldn't make any difference for this case since there's only one appendChilded script, right?

[05:09:01.0000] <jgraham>
Right. But if one didn't have force-async set in this case, you might expect the appendChilded script to execute before the following parser-inserted script

[05:10:00.0000] <Ms2ger>
Expect based on what?

[05:10:01.0000] <Ms2ger>
Locig? See the topic :)

[05:10:02.0000] <Ms2ger>
Locig, really?

[05:11:00.0000] <Ms2ger>
/me blames his keyboard

[05:11:01.0000] <jgraham>
Expect based on the idea that people want a way to control the order of script execution

[05:12:00.0000] <matjas>
.async = false

[05:12:01.0000] <Ms2ger>
And not block the parser

[05:12:02.0000] <jgraham>
Well you can have one or the other

[05:13:00.0000] <jgraham>
Setting .async to false doesn't seem to have an effect

[05:13:01.0000] <jgraham>
Similarly clearing the force-async flag doesn't seem to have an effect in this case

[05:13:02.0000] <jgraham>
(it appears that setting and unsetting the async content attribute will do that)

[06:06:00.0000] <matjas>
jgraham: IIRC setting .async = false on a dynamically inserted script makes sure that the source order is respected as far as dynamically inserted scripts go

[06:07:00.0000] <jgraham>
matjas: Possibly. If that is the case I am missing the spec text that ensures it

[06:07:01.0000] <Ms2ger>
But not relative to parser-inserted scripts

[06:08:00.0000] <matjas>
Ms2ger: exactly

[06:08:01.0000] <matjas>
jgraham: http://hsivonen.iki.fi/script-execution/

[06:09:00.0000] <jgraham>
Oh, yeah, sorry I misread you

[06:11:00.0000] <MikeSmith>
cool to see janv_ working on the HTML5 context-menu feature

[06:11:01.0000] <MikeSmith>
http://hg.mozilla.org/mozilla-central/rev/561821863607

[06:14:00.0000] <jgraham>
matjas: Yeah, I have read that blog entry before, but this kind of thing is easy to not fully understand until you actually need to solve a problem

[06:14:01.0000] <jgraham>
At least for me

[06:14:02.0000] <matjas>
MikeSmith: what’s with the `data://www.mozilla.org/favicon.ico`?

[06:14:03.0000] <MikeSmith>
hmm

[06:14:04.0000] <MikeSmith>
matjas: no idea

[06:16:00.0000] <Ms2ger>
Just something invalid to test error handling?

[06:17:00.0000] <matjas>
Ms2ger: ah, duh

[06:23:00.0000] <matjas>
/me would’ve used data:lolwut

[06:23:01.0000] <MikeSmith>
heh

[06:24:00.0000] <Ms2ger>
lolwut

[06:25:00.0000] <MikeSmith>
so I wonder how well Google Swiffy works

[06:26:00.0000] <MikeSmith>
http://swiffy.googlelabs.com/faq.html says "Swiffy supports most of the features in Flash 5" but does that mean you can take most Flash 5 content and generate usable output from it?

[06:27:00.0000] <MikeSmith>
being able to author with Flash dev tools and generate Web-platform output is a pretty powerful thing, if it actually works

[06:29:00.0000] <robman>
wouldn't being able to author using the Web-platform to generate Web-platform be even betterer MikeSmith? 8)

[06:30:00.0000] <MikeSmith>
robman: no, not necessarily

[06:30:01.0000] <MikeSmith>
not at all, really

[06:31:00.0000] <robman>
really...i hate having to run windows/macosx to run adobe tools

[06:31:01.0000] <MikeSmith>
I hate using windows/macosx to do anything

[06:31:02.0000] <MikeSmith>
linux too

[06:32:00.0000] <MikeSmith>
OSes suck

[06:32:01.0000] <MikeSmith>
all of them

[06:32:02.0000] <jgraham>
MikeSmith: Presumably the flash tools will produce output that is no better than flash in many ways

[06:32:03.0000] <robman>
8)

[06:32:04.0000] <robman>
MikeSmith - well you got me there

[06:33:00.0000] <MikeSmith>
jgraham: well, the point is the tools make things much easier for devs, they are very good tools in terms of the usability, user-friendliness, ease-of-use or however you want to term it

[06:34:00.0000] <jgraham>
Having nice tools is of course good

[06:34:01.0000] <jgraham>
But I imagine it will be easy to produce content that is inaccessible to search engines, screenreaders, etc. and hard to do the reverse

[06:34:02.0000] <MikeSmith>
in comparison as far as similar tools for producing Web-platform output natively, we are at the level of caveman rubbing sticks together to make fire

[06:35:00.0000] <MikeSmith>
jgraham: yeah, point taken

[06:35:01.0000] <MikeSmith>
the platform is full of tradeoffs

[06:35:02.0000] <robman>
/me like fire

[06:35:03.0000] <robman>
/me grunts

[06:35:04.0000] <MikeSmith>
heh

[06:36:00.0000] <MikeSmith>
/me is reminded of the "Fire bad!" scene in Young Frankenstein

[06:36:01.0000] <robman>
8)

[06:38:00.0000] <MikeSmith>
"Ladies and gentlemen of the jury, I'm just a caveman. I fell on some ice and later got thawed out by some of your scientists. Your platform frightens and confuses me"

[09:07:00.0000] <zcorpan>
Ms2ger: you read www-archive?

[09:07:01.0000] <Ms2ger>
I skim it sometimes

[09:07:02.0000] <zcorpan>
i saw an email from björn with tests for XMLSerializer/DOMParser

[09:09:00.0000] <Ms2ger>
Saw those

[09:10:00.0000] <Ms2ger>
Those are somewhere near the bottom of my todo list

[09:10:01.0000] <zcorpan>
k

[09:50:00.0000] <cryptopsy>
is there a printable reference for html5 on whatwg.org ?

[09:51:00.0000] <Ms2ger>
cryptopsy, whatwg.org/html has a link hidden in the header, IIRC

[09:52:00.0000] <cryptopsy>
news, specs, demo, charter ...

[09:53:00.0000] <Ms2ger>
Did you miss the /html?

[09:54:00.0000] <Ms2ger>
PDF print versions:

[09:54:01.0000] <Ms2ger>
A4: http://www.whatwg.org/specs/web-apps/current-work/html-a4.pdf

[09:54:02.0000] <Ms2ger>
Letter: http://www.whatwg.org/specs/web-apps/current-work/html-letter.pdf

[09:55:00.0000] <cryptopsy>
how did you find that?

[09:56:00.0000] <Philip`>
By going to http://whatwg.org/html then reading it, I expect :-)

[09:56:01.0000] <Ms2ger>
And knowing where it was

[09:56:02.0000] <Philip`>
I suppose that helps too

[09:59:00.0000] <jgraham>
cryptopsy: Basically there are three stages: 1) Become a Hixie fanboi and hang on his every edit b) Get asked question about the spec c) Use knowledge gained from a) to answer b) and so impress girls/boys (depending on preference)

[09:59:01.0000] <cryptopsy>
i looked at the header for whatwg.org

[10:00:00.0000] <cryptopsy>
jgraham: i'm new to webdev, don't the html people publish a reference or spec for people to base on?

[10:00:01.0000] <cryptopsy>
thanks

[10:00:02.0000] <cryptopsy>
Ms2ger: ^

[10:00:03.0000] <jgraham>
cryptopsy: Base what on?

[10:01:00.0000] <Ms2ger>
whatwg.org/html

[10:01:01.0000] <cryptopsy>
some sort of document that lists all the possible tags and their usage, like how LaTeX has?

[10:01:02.0000] <jgraham>
The spec is big and complex and user unfriendly because browsers are big and complex and their codebases are unfriendly

[10:01:03.0000] <cryptopsy>
jgraham: where is this spec? i need those additional details

[10:01:04.0000] <zewt>
friendly enough to me

[10:01:05.0000] <cryptopsy>
zewt: now you're just teasing me ..

[10:01:06.0000] <zewt>
perhaps in part due to years of reading specs from ietf and friends

[10:01:07.0000] <jgraham>
And the spec is designed to allow different big unfriendly browsers to get on without having an all out war and destroying all the little people

[10:02:00.0000] <jgraham>
cryptopsy: But for authors all the complexity isn't usually needed

[10:02:01.0000] <zewt>
jgraham: depends heavily on what part of the system you're looking at, though

[10:02:02.0000] <zewt>
some parts are easier to understand clearly from the spec than other docs

[10:02:03.0000] <jgraham>
There is a stylesheet you can apply that will remove implementation requirments from the spec

[10:03:00.0000] <dglazkov>
good morning, Whatwg!

[10:03:01.0000] <jgraham>
Or you can look at http://dev.w3.org/html5/markup/

[10:03:02.0000] <zewt>
in one limited slice of the planet, i suppose

[10:04:00.0000] <cryptopsy>
jgraham: true

[10:04:01.0000] <jgraham>
zewt: "i'm new to webdev" does suggest that the finer points of how to schedule scripts might be a few orders of magnitude too complex at this stage

[10:05:00.0000] <cryptopsy>
not new to browsers and encoding though

[10:05:01.0000] <zewt>
someone new to web development probably wants any 101 documentation and some code to squint at, anyway

[10:05:02.0000] <cryptopsy>
im not using this to make a blog or post pictures of cats and my 95 year old grandmother

[10:05:03.0000] <zewt>
poor cats/grandmothers

[10:06:00.0000] <cryptopsy>
so where's this super detailed doc at?

[10:07:00.0000] <zewt>
i'm not sure we know what you're asking for

[10:08:00.0000] <zewt>
if you want a list of tags, jgraham's link gives that

[10:11:00.0000] <zewt>
the actual spec for html elements is at http://www.whatwg.org/specs/web-apps/current-work/complete/#auto-toc-4

[10:13:00.0000] <zewt>
the stuff in there is really not all that scary, IMO

[10:13:01.0000] <Ms2ger>
zewt, that section does include many of the finer points of script timing ;)

[10:14:00.0000] <zewt>
as with any detailed documentation, you do have to recognize when to skip a section you don't care about :)

[11:06:00.0000] <TabAtkins>
telnet towel.blinkenlights.nl

[11:06:01.0000] <TabAtkins>
do it now

[11:06:02.0000] <TabAtkins>
NOW

[11:37:00.0000] <hober>
classic

[13:12:00.0000] <Hixie>
janv_: agreed; can you file a bug or send e-mail? (filing a bug is really easy; just use the little text field in the bottom right of the spec)

[13:27:00.0000] <jarek>
Hi

[13:28:00.0000] <Ms2ger>
Evening

[13:28:01.0000] <jarek>
Let's say there is an iframe that is 1000px wide and that was scaled down with CSS3 transforms by half with '-webkit-transform: scale(0.5)'

[13:28:02.0000] <jarek>
 should the media queries inside the iframe assume that its width is 1000px or 500px?

[13:29:00.0000] <jarek>
on Safari they assume that the width is 500px, on Chrome they assume 500px

[13:29:01.0000] <jarek>
no, I mean the opposite

[13:30:00.0000] <jarek>
On Chrome scalling an iframe does not change the behavior of media queries

[13:30:01.0000] <jarek>
is this a bug?

[13:31:00.0000] <jarek>
s/ on Safari they assume that the width is 500px, on Chrome they assume 500px/ on Safari they assume that the width is 500px, on Chrome they assume 1000px

[13:32:00.0000] <espadrine>
The spec states that in a <script> with a src="", the content of <script>...</script> can only be comments.

[13:32:01.0000] <espadrine>
Is this implemented in major browsers?

[13:32:02.0000] <Philip`>
espadrine: That is purely an authoring conformance requirement, it doesn't affect browsers

[13:33:00.0000] <Philip`>
(Browsers don't run the content of the script element at all, regardless of what's in there, if it has a src attribute)

[13:34:00.0000] <espadrine>
Philip`: ok. Makes sense.

[13:34:01.0000] <Philip`>
(The requirement is there as a way of warning authors when they think they've written script code that will get executed but won't)

[13:34:02.0000] <espadrine>
Do you still get it through script.text though?

[13:34:03.0000] <Philip`>
(without making loads of existing pages that put comments inside scripts non-conforming)

[13:34:04.0000] <Ms2ger>
Yep

[13:35:00.0000] <smaug____>
hmm, probably too late to change message channels :(

[13:36:00.0000] <espadrine>
ok

[13:36:01.0000] <smaug____>
though, they are not yet implemented everywhere, so web doesn't really depend on them

[14:06:00.0000] <zcorpan>
smaug____: what would you want to change with message channels?

[14:08:00.0000] <smaug____>
still reviewing all the stuff, but I think I'd like to remove MessageChannel :)

[14:09:00.0000] <smaug____>
I guess it would be quite hard to convince others to do that.

[14:10:00.0000] <zcorpan>
why remove?

[14:10:01.0000] <smaug____>
there is no real use for it

[14:12:00.0000] <zcorpan>
hmm

[14:12:01.0000] <smaug____>
I'd rather do something like var newPort = otherWindow.createPort("portName"); newPort.postMessage("foobardata");

[14:18:00.0000] <smaug____>
the whole port transfer is quite awkward

[14:18:01.0000] <zcorpan>
yeah i must say i get confused by the ports thingies

[14:19:00.0000] <zcorpan>
but i don't know what a better api would be

[14:19:01.0000] <zcorpan>
i look forward to your conclusions and proposals

[14:19:02.0000] <smaug____>
well, just create a new port and the other end could get "connect" event which had .port as a property

[14:31:00.0000] <smaug____>
Hixie: ping

[14:54:00.0000] <Hixie>
smaug____: pong

[14:55:00.0000] <smaug____>
Hixie: what is the use case for MessageChannel?

[14:55:01.0000] <smaug____>
why not just create ports on both sides

[14:55:02.0000] <Hixie>
both sides of what?

[14:55:03.0000] <Hixie>
the use case is the capabilities model. bascially the ability to send someone an API endpoint.

[14:55:04.0000] <smaug____>
but why?

[14:56:00.0000] <Hixie>
why what?

[14:56:01.0000] <smaug____>
why you need the ability to send API endpoint

[14:56:02.0000] <smaug____>
why couldn't you just create end point on both sides

[14:56:03.0000] <Hixie>
both sides of what??

[14:56:04.0000] <smaug____>
say, you have two windows...

[14:57:00.0000] <smaug____>
var port = otherWindow.createPort("name-of-the-port"); // this would cause "connect" event in the otherWindow

[14:57:01.0000] <smaug____>
similar could work with workers

[14:57:02.0000] <Hixie>
you don't know how you've got on the other end, that's the whole point

[14:57:03.0000] <Hixie>
say you are providing an API to, I dunno, add someone to a contact list

[14:58:00.0000] <Hixie>
say you're a webmail app

[14:58:01.0000] <Hixie>
and you have a contact list

[14:58:02.0000] <Hixie>
and the contact list has an API, and one of the operations is "add someone to the contact list"

[14:58:03.0000] <smaug____>
and?

[14:58:04.0000] <Hixie>
if another part of the app wants to add someone to the contact list, they could get the local contacts API object, and then ask it to add to the contac tlist

[14:59:00.0000] <Hixie>
but what if the contact list wants to be implemented in another thread?

[14:59:01.0000] <Hixie>
now the contact API has to proxy everything though

[14:59:02.0000] <Hixie>
what if it's in another domain? now you have to proxy through to an iframe and then from there to the worker, etc

[15:00:00.0000] <Hixie>
now say that the contact list wants to grant the original user of this API the right to delete the user they added, but no other

[15:00:01.0000] <Hixie>
how would you do it?

[15:00:02.0000] <Hixie>
with message channels, you can do all this trivially.

[15:00:03.0000] <Hixie>
there's no need to proxy anything -- the messages just end up at the right end point

[15:00:04.0000] <smaug____>
ah, you want to move the port first via window level postMessage and then post it to a worker?

[15:01:00.0000] <Hixie>
there's no need to worry about permissions, you just give back a port and then whoever has that port can just delete that one contact.

[15:01:01.0000] <Hixie>
etc.

[15:01:02.0000] <Hixie>
for example, yeah

[15:01:03.0000] <Hixie>
this is just a specific example of the capabilities programming model

[15:01:04.0000] <Hixie>
it's like functional programming, it's just a style

[15:01:05.0000] <Hixie>
i mean it's like saying "what's the use case for functions being first-class objects"

[15:01:06.0000] <Hixie>
or "what's the use case for object-orientated programming"

[15:02:00.0000] <smaug____>
it is just quite awkward

[15:02:01.0000] <Hixie>
it's just another tool

[15:03:00.0000] <Hixie>
(a very powerful one, much like oop and lambdas)

[15:03:01.0000] <Hixie>
how is it awkward?

[15:03:02.0000] <smaug____>
when for common cases creating ports could be enough

[15:03:03.0000] <smaug____>
no need to transfer anything

[15:03:04.0000] <Hixie>
?

[15:03:05.0000] <smaug____>
but ok, what you explained is something which isn't clear when reading the drafts

[15:04:00.0000] <Hixie>
yeah the drafts assume you know this programming style and that you'll recognise it :-)

[15:04:01.0000] <Hixie>
i should probably have an intro section that explains it or something

[15:04:02.0000] <Hixie>
but i'm not good at explaining it

[15:04:03.0000] <smaug____>
Also, "objects listed in transfer have already been transfered once before, then throw a DATA_CLONE_ERR exception and abort these steps" is easy to misread

[15:05:00.0000] <smaug____>
since I wasn't sure whether one can re-transfer a port

[15:05:01.0000] <Hixie>
that refers to the objects, not the underlying port, but yeah, i see what you mean

[15:06:00.0000] <smaug____>
I still don't see the need to MessageChannel

[15:06:01.0000] <smaug____>
for

[15:06:02.0000] <smaug____>
If one could create ports using createPort or such

[15:06:03.0000] <Hixie>
createPort === MessageChannel

[15:06:04.0000] <smaug____>
very much so

[15:06:05.0000] <Hixie>
MessageChannel is nothing but a port constructor

[15:07:00.0000] <smaug____>
and a bit strange one

[15:07:01.0000] <smaug____>
since I the idea is that the other port is sent to somewhere

[15:07:02.0000] <Hixie>
are you familiar with unix socket programming?

[15:07:03.0000] <smaug____>
yes

[15:07:04.0000] <Hixie>
think of ports as pipes.

[15:07:05.0000] <Hixie>
or rather, sockets

[15:08:00.0000] <Hixie>
when you create a pipe (two sockets) you get both end points

[15:08:01.0000] <smaug____>
sure

[15:08:02.0000] <Hixie>
in exactly the way MessageChannel gives you two ports

[15:08:03.0000] <Hixie>
this is just trying to bring that programming model to JS authors

[15:08:04.0000] <smaug____>
but why do you need both ends

[15:08:05.0000] <smaug____>
when you're going to use one end in one context and the other some other context

[15:11:00.0000] <Hixie>
i don't understand the question

[15:11:01.0000] <smaug____>
MessageChannel creates two ports in the same context

[15:11:02.0000] <smaug____>
and then usually, I assume, the other port is sent to otherWindow or worker or whatever

[15:11:03.0000] <smaug____>
right?

[15:11:04.0000] <Hixie>
it can be, doesn't have to be

[15:11:05.0000] <smaug____>
but that is the basic use case, right?

[15:11:06.0000] <Hixie>
just like when you create a pipe you often pass one socket to somewhere else

[15:12:00.0000] <smaug____>
Why couldn't you just call otherWindow.createPort() which would create one end in the executing context and one port in otherWindow

[15:12:01.0000] <smaug____>
if you really want to, you could call window.createPort()

[15:12:02.0000] <Hixie>
so you'd have otherWindow.createPort, worker.createPort, port.createPort, etc?

[15:12:03.0000] <smaug____>
yes

[15:13:00.0000] <Hixie>
that would be like having Object.addFunction, Array.addFunction, String.addFunction, etc, instead of having a Function constructor

[15:13:01.0000] <Hixie>
i mean i guess it's fine, but it's certainly not what i would describe as neat orthogonal api design

[15:13:02.0000] <smaug____>
that would be like having addEventListener in all the EventTarget objects

[15:14:00.0000] <Hixie>
no, addEventListener() doesn't create anything

[15:15:00.0000] <smaug____>
actually, window, worker and port should all inherit, if possible, some comment base interface

[15:15:01.0000] <smaug____>
s/comment/common/

[15:15:02.0000] <Hixie>
well they all inherit from EventTarget now so that's not an option anymore

[15:15:03.0000] <smaug____>
you could add something between eventtarget and the current interface :)

[15:15:04.0000] <Hixie>
anyway sometimes you just want a standalone message channel because you're goign to be passing it to a callback

[15:16:00.0000] <Hixie>
the fact that you're going to send the port to someone else is not something you should need to care about when you're creating the port

[15:16:01.0000] <Hixie>
the port itself is just a capability endpoint

[15:16:02.0000] <Hixie>
you might keep it in your context, or you might not

[15:16:03.0000] <Hixie>
that's besides the point

[15:17:00.0000] <zewt>
passing them around with structured clone also means you're passing them around with an existing communications interface, rather than defining a separate event

[15:18:00.0000] <smaug____>
I thought the point is to create communication channels between different "contexts"

[15:18:01.0000] <zewt>
eg. so you can receive a message { type: "email", target: "user⊙ec", portForThisInterface: aMessagePort }

[15:19:00.0000] <Hixie>
smaug____: no, the point is to create a capability token.

[15:19:01.0000] <smaug____>
huh

[15:19:02.0000] <Hixie>
smaug____: it just so happens that many times, this will be used in a different context, because in the same context you could get the same effect with a simple function reference

[15:19:03.0000] <smaug____>
anyway, as of now, I don't think we have plans to implement MessageChannels

[15:20:00.0000] <Hixie>
the capability programming people will be sad

[15:23:00.0000] <smaug____>
and, for the common case createPort would make more sense

[15:23:01.0000] <Hixie>
common case of what?

[15:23:02.0000] <zewt>
smaug: not really

[15:23:03.0000] <smaug____>
that you want to enable communication between two different contexts

[15:23:04.0000] <Hixie>
no

[15:23:05.0000] <Hixie>
capability tokens aren't about communicating between different contexts

[15:24:00.0000] <zewt>
it'd separate port creation from other messages, so if you want to create a port with other information attached to it (a description of the port so the other side knows what to do with it), you'd end up duplicating messaging functionality

[15:24:01.0000] <Hixie>
any more than object orientated programming is about, say, allocating memory. or function dispatch.

[15:24:02.0000] <Hixie>
they're related, but that's no the point of the feature.

[15:24:03.0000] <smaug____>
in practice port is all about communication, like sockets

[15:25:00.0000] <Hixie>
i would imagine that most people writing code that creates capability tokens have the code creating the token quite separate from the communication bootstrapping code

[15:29:00.0000] <smaug____>
/me disagrees, but knew that trying get rid of MessageChannel would be hard :)

[15:29:01.0000] <Hixie>
you didn't even know what capability tokens were an hour ago, i don't think you have much credibility in terms of opinions on how caps programming is done :-P

[15:30:00.0000] <smaug____>
that is not the point

[15:30:01.0000] <smaug____>
the point is to have easy to use APIs for web devs

[15:33:00.0000] <Hixie>
i'm sure we can all agree on that point

[15:39:00.0000] <Hixie>
if anyone has opinions on http://www.w3.org/Bugs/Public/show_bug.cgi?id=12306 it would be helpful if you could put them in the bug. Right now I'm leaning towards leaving the MUST in, which makes a lot of <canvas>-based apps non-conforming currently (especially games).

[15:41:00.0000] <Hixie>
hober: any chance we can get the csswg to use bugzilla?

[15:43:00.0000] <jgraham>
Hixie: I think that bug is angels on the head os a pin stuff that will have no effect on any part of the real world. So I wouldn't spend any time on it

[15:43:01.0000] <jgraham>
I also agree that having to pass around message ports seems weirdly cumbersome, but it does sound quite powerful

[15:46:00.0000] <Hixie>
unless you're using their power, you're really not going to need to send them over at all, as far as i can tell. for example workers create them implicitly.

[15:46:01.0000] <jgraham>
(it might alos be nice if the spec used the word "connected" rather than "entangled" since afaict that is the concept that you are expressing)

[15:46:02.0000] <Hixie>
i try to use different words so that you can more easily search for them

[15:46:03.0000] <Hixie>
"connect" is used for far too many concepts already

[15:47:00.0000] <jgraham>
In this case it sounds like you chose a difficult sounding word at random from a physcs textbook :)

[15:47:01.0000] <jgraham>
"Coupled" would have worked too

[15:47:02.0000] <jgraham>
Anyway it is not important

[15:48:00.0000] <jgraham>
There was already some big debate about the order of arguments in 3 argument postMessage wasn't there?

[15:49:00.0000] <Hixie>
well "entangled" in the quantum sense does rather describe what's going on here

[15:49:01.0000] <jgraham>
So it would be unwise of me to say that the current spec seems very strange

[15:49:02.0000] <Hixie>
i do not recall one way or the other

[15:49:03.0000] <Hixie>
if it's implemented i don't intend to change it

[15:50:00.0000] <jgraham>
Sure.

[15:50:01.0000] <jgraham>
I have no idea if this is implemented

[15:50:02.0000] <jgraham>
Presumably it is webkit-only if it is

[15:51:00.0000] <Hixie>
what's the order and what do you want it to do?

[15:52:00.0000] <jgraham>
(quantum entanglement seems rather different because the entangled parts resolve into different states as needed to preserve physical invariants. Whereas here the ports presumably have the same information on both ends)

[15:53:00.0000] <jgraham>
I just mean taht postMessage(message, ports, targetOrigin) seems inconsistent with postMessage(message, targetOrigin)

[15:53:01.0000] <jgraham>
I would expect the additional argument to be last

[15:54:00.0000] <Hixie>
ah well the other way to look at it is postMessage() has two forms, (message) and (message, transferrables), and Window just adds an argument on the end of those

[15:54:01.0000] <jgraham>
(also entanglement suggests there is some kind of randomness which I assume there is not)

[15:54:02.0000] <Hixie>
depends which postMessage() you think of as canonical :-)

[15:54:03.0000] <Hixie>
yeah entanglement certainly isn't an exact metaphor

[15:56:00.0000] <zcorpan>
jgraham: we implement all arguments to postMessage

[15:56:01.0000] <jgraham>
Well I don't find it as easy to view so the current spec makes sense and I would expect people to get it wrong more often this way around. But I don't feel strongly enough to dpend time advocating changing it or anything

[15:57:00.0000] <jgraham>
zcorpan: Oh, we implement the message channels stuff?

[15:57:01.0000] <zcorpan>
yes

[15:57:02.0000] <jgraham>
Well I guess I should stop complaining about it now then ;)

[15:57:03.0000] <Philip`>
Hixie: About the canvas bug: I guess it depends on whether you think training authors to ignore non-machine-checkable conformance requirements is a problem, or if that's already a lost cause and so the spec might as well give some idealistic vision instead

[15:57:04.0000] <zcorpan>
since 10.60

[15:57:05.0000] <jgraham>
Ah, I thought we had just got workers, not the other stuff too

[15:58:00.0000] <zewt>
(i already pretty much ignore those "requirements"; they frankly seem rather nonsense to me)

[15:58:01.0000] <zewt>
(that one being a solid example of why)

[15:58:02.0000] <zcorpan>
it was implemented at the same time as workers

[15:58:03.0000] <jgraham>
Yeah, makes sense

[15:58:04.0000] <Hixie>
Philip`: yeah

[15:58:05.0000] <zewt>
(speaking as a random author :)

[15:58:06.0000] <jgraham>
(it still wouldn't be a big pain to change the argument order if it really helped. I can't image much depends on this yet)

[15:59:00.0000] <jgraham>
*imagine

[15:59:01.0000] <jgraham>
(But liek I say I really don't care all that much)

[15:59:02.0000] <zcorpan>
jgraham: we asked for the spec to change to the order it has now :)

[16:00:00.0000] <jgraham>
zcorpan: Well that was silly :p

[16:00:01.0000] <jgraham>
zcorpan: You mean you did? Why?

[16:00:02.0000] <zcorpan>
jgraham: makes sense to have the optional argument last

[16:01:00.0000] <Hixie>
hey, wait, yeah, what are you saying jgraham

[16:01:01.0000] <Hixie>
the order isn't what you said it was

[16:01:02.0000] <Hixie>
the order you described is what i would expect

[16:01:03.0000] <Hixie>
but yeah, we changed it

[16:02:00.0000] <smaug____>
/me needs to tests message channels and ports in Opera, since Webkit is too buggy

[16:02:01.0000] <Hixie>
smaug____: file bugs :-)

[16:02:02.0000] <jgraham>
Oh, in that case obviously zcorpan is an eminently sensible individual with good taste

[16:02:03.0000] <smaug____>
I'll file a bug to remove MessageChannel :p

[16:02:04.0000] <jgraham>
and I am probably tired enough to not notice when I am reading an old draft

[16:02:05.0000] <jgraham>
:)

[16:02:06.0000] <jgraham>
I think this suggests that sleep might be a good thing

[16:03:00.0000] <Hixie>
in oher news, what is with chrome stripping off a leading "javascript:" when i post a url into the url bar

[16:03:01.0000] <Hixie>
damn silly computers trying to stop me hurting myself

[16:03:02.0000] <zewt>
your browser knows best!

[16:03:03.0000] <zcorpan>
jgraham: actually it was sof and tarquin who pushed for changing the order, but i guess i did the emailing to public lists

[16:04:00.0000] <zcorpan>
smaug____: opera should be compliant as far as i know

[16:04:01.0000] <smaug____>
Hixie: I think other browsers are doing similar silly things. In FF there is at least some hidden pref to get back reasonable javascript: handling

[16:04:02.0000] <zcorpan>
lemme know if you find bugs :)

[16:04:03.0000] <jgraham>
zcorpan: Ah, well obviously Opera is full of people with good taste (as long as you define good taste as "agree with me" :) )

[16:05:00.0000] <zewt>
doesn't everyone?

[16:05:01.0000] <jgraham>
Hixie: A few scams on facebook and this is what happens :(

[16:07:00.0000] <zewt>
if they turn into "open the javascript console and paste this", will browsers start removing javascript consoles? heh

[16:08:00.0000] <Hixie>
"type in the following C code and compile it, then run it with sudo"

[16:08:01.0000] <zcorpan>
yeah try that on facebook

[16:09:00.0000] <Hixie>
(i guess that explains the way iOS prevents the user from having access to the machine)

[16:09:01.0000] <zewt>
well, mobile devices are as interested in protecting the carrier from the user than the user from the internet

[16:14:00.0000] <zcorpan>
maybe webidl can add a way to annotate an attribute so that it gets enumberable:false

[16:16:00.0000] <smaug____>
zcorpan: Opera seems to work well. The API is still bizarre

[16:16:01.0000] <zcorpan>
http://dev.w3.org/2006/webapi/WebIDL/ needs more google juice (compared to TR/)

[16:17:00.0000] <zcorpan>
smaug____: agree it's a bit confusing

[16:17:01.0000] <zcorpan>
smaug____: not convinced we should change it

[16:24:00.0000] <zcorpan>
/me -> sleep


2011-08-10
[17:02:00.0000] <Hixie>
http://en.wikipedia.org/wiki/Non-English_usage_of_quotation_marks is rather humbling.

[17:37:00.0000] <hober>
Hixie: indeed, hence dbaron's bug

[17:38:00.0000] <Hixie>
hm?

[17:38:01.0000] <dbaron>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13718

[17:40:00.0000] <Hixie>
ah, interesting. hadn't seen that one yet.

[17:40:01.0000] <Hixie>
can't really argue with the first point.

[17:40:02.0000] <Hixie>
for the second point, we tried that. people didn't go for it.

[17:41:00.0000] <Hixie>
or rather, we tried making <q> that.

[17:41:01.0000] <Hixie>
there isn't really a use case for the described element.

[17:41:02.0000] <dbaron>
for people who like semantics? :-)

[17:42:00.0000] <dbaron>
alternatively, deprecate <q> in favor of quotation marks

[17:43:00.0000] <Hixie>
semantics isn't a use case on its own

[17:43:01.0000] <Hixie>
the HTML spec already says that there's no reason to use <q> rather than quotation marks

[17:43:02.0000] <Hixie>
"The use of q elements to mark up quotations is entirely optional; using explicit quotation punctuation without q elements is just as correct."

[17:44:00.0000] <Hixie>
there's even an example

[18:03:00.0000] <Hixie>
heycam: are you going to drop [AllowAny]?

[18:03:01.0000] <Hixie>
heycam: in http://www.w3.org/Bugs/Public/show_bug.cgi?id=12870 you suggest i use it but the spec says it might go away

[18:06:00.0000] <heycam>
Hixie, probably not going to drop it

[18:06:01.0000] <Hixie>
k

[18:06:02.0000] <heycam>
Hixie, it's kind of ugly, but... *shrug*

[18:06:03.0000] <Hixie>
is there anything in JS that has a [[Call]] thingy but is not a Function?

[18:07:00.0000] <heycam>
Hixie, in the spec, I don't think so. but isn't there something with RegExp objects?

[18:08:00.0000] <heycam>
(might be completely wrong here)

[18:08:01.0000] <Hixie>
/me tries setTimeout with a RegEcp

[18:08:02.0000] <Hixie>
RegExp

[18:09:00.0000] <Hixie>
...i have no idea what i would test

[18:09:01.0000] <heycam>
maybe they're not going to be callable forever... https://bugzilla.mozilla.org/show_bug.cgi?id=582717

[18:11:00.0000] <Hixie>
Having the domintro blocks and IDL blocks match isn't a goal, btw.

[18:11:01.0000] <heycam>
Hixie, ok, I understand that

[18:11:02.0000] <heycam>
you want to omit the finicky details

[18:11:03.0000] <Hixie>
I mean, they both have to match the same reality, but it's no problem if they use different arguments and stuff

[18:11:04.0000] <Hixie>
yeah

[18:11:05.0000] <Hixie>
(different argument names, that is)

[18:12:00.0000] <Hixie>
gotta love using more fancy webidl syntax though

[18:12:01.0000] <heycam>
except when it looks crazy and distracting

[19:16:00.0000] <gsnedders>
heycam, Hixie: yeah, there's agreement between all major browser vendors to drop [[Call]] from RegExp. Some already have, some haven't, AFAIK.

[19:16:01.0000] <gsnedders>
(IE never had [[Call]] on RegExp)

[19:17:00.0000] <gsnedders>
Ah, gone in Fx5. Probably in Saf5.1, too, though I'm not digging through SVN at this time to see if they have.

[19:18:00.0000] <gsnedders>
Opera 11.50 dropped it too

[19:19:00.0000] <gsnedders>
/me wishes all this stuff happened in public and not emails between JS developers — but it's considered off-topic for es-discuss as it's non-standard

[19:20:00.0000] <AryehGregor>
Use whatwg.  :)

[19:21:00.0000] <Hixie>
yeah you're welcome to use whatwg

[19:21:01.0000] <Hixie>
don't forget to update the relevant whatwg wiki page

[20:10:00.0000] <MikeSmith>
Hixie: I'll split bug 12986

[20:10:01.0000] <MikeSmith>
(the big "Last Call comments to HTML5" one)

[20:11:00.0000] <zewt>
If you're having problems seeing the site, it may be because Grooveshark doesn't support your current zoom level. To remedy this problem, simply press CTRL+0 (CMD+0 for Mac users) to return to your browser's default zoom level.

[20:11:01.0000] <zewt>
gross

[20:22:00.0000] <MikeSmith>
Hixie: 13590 too

[21:47:00.0000] <MikeSmith>
that was fun

[22:26:00.0000] <Hixie>
MikeSmith: thanks dude

[22:27:00.0000] <MikeSmith>
no problem man

[22:27:01.0000] <MikeSmith>
I should have done it from the beginning

[22:27:02.0000] <MikeSmith>
was just being lazy

[22:31:00.0000] <Hixie>
heh

[22:34:00.0000] <MikeSmith>
Hixie: if I missed any others that need splitting, let me know

[22:35:00.0000] <MikeSmith>
btw, hsivonen can read your mind

[22:35:01.0000] <MikeSmith>
you made this change recently:

[22:35:02.0000] <MikeSmith>
http://html5.org/tools/web-apps-tracker?from=6387&to=6388

[22:35:03.0000] <MikeSmith>
about the <link> element for microdata

[22:36:00.0000] <MikeSmith>
but hsivonen already implemented checking for it in validator.nu exactly it that way last year

[22:38:00.0000] <Hixie>
cool

[00:05:00.0000] <Hixie>
does css have a way to remove whitespace nodes these days?

[00:05:01.0000] <Hixie>
e.g. if i have <div><span/> <span/></div> and want to have the two spans be inline-blocks that touch, no space from the space character?

[00:21:00.0000] <benjoffe>
Hixie: 'whitespace: discard;' should do that I think

[00:21:01.0000] <Hixie>
oooh

[00:21:02.0000] <Hixie>
is that implemented anywhere?

[00:21:03.0000] <benjoffe>
Not sure

[00:23:00.0000] <benjoffe>
white-space-collapse: discard;

[00:26:00.0000] <MikeSmith>
Dadhacker: Revolutions That Weren’t - http://www.dadhacker.com/blog/?p=1444  scroll down to the "Ex Em Hell" part

[00:26:01.0000] <MikeSmith>
"What we actually got: Any number of crappy serialization schemes and over-designed and under-implemented replacements for INI files.  Undebuggable configuration files, poorly written attempts at replacements to already perfectly awful tools (yes, ANT and MSBuild, I’m thinking of you), and a lot of other smelly garbage littered with angle brackets."

[00:27:00.0000] <MikeSmith>
"We are still living this nightmare, with no end in sight."

[02:11:00.0000] <zcorpan>
selectors has [foo^=bar], [foo$=bar] and [foo*=bar] but no way to match case-insensitively

[02:12:00.0000] <zcorpan>
[foo=/bar/i]

[02:13:00.0000] <jgraham>
Of course what selectors needs is to look even more like perl

[02:14:00.0000] <zcorpan>
is there a punctuation character that is usually used to indicate case-insensitivity?

[02:15:00.0000] <jgraham>
Well, I don't think using random punctuation would be *better*. I mean selectors is already a giant mess

[02:15:01.0000] <woef>
It would be nicer if html would be case-insensitive in the first place.

[02:16:00.0000] <jgraham>
But giving people /foo/i would make them think that foo was a regexp

[02:16:01.0000] <zcorpan>
woef: html *is* case-insensitive for various attribute values

[02:16:02.0000] <jgraham>
or /bar/i I guess, given your examples

[02:17:00.0000] <zcorpan>
yeah i was thinking maybe we should just allow full regexps. but maybe that's not acceptable

[02:17:01.0000] <woef>
zcorpan: so just css selector matching :)

[02:18:00.0000] <zcorpan>
woef: the problem is that selectors isn't case-insensitive

[02:18:01.0000] <zcorpan>
except there's an ugly hack to make it so for some attributes in text/html

[02:19:00.0000] <jgraham>
zcorpan: Well people could always write very slow regexp

[02:19:01.0000] <zcorpan>
yeah

[02:19:02.0000] <jgraham>
Or, I guess you could limit the syntax to real regexp and expect people to use something like that google library that is supposed to have good time behaviour in all cases

[02:20:00.0000] <jgraham>
But that seems like quite a lot of implementation complexity

[02:21:00.0000] <jgraham>
(re2)

[02:21:01.0000] <zcorpan>
yeah i'd like just a flag for case-insensitive, but would be nice if it could be used together with ^= $= and *=

[02:21:02.0000] <jgraham>
(but the site seems to be down)

[02:21:03.0000] <zcorpan>
and doesn't look like a smiley

[02:21:04.0000] <zcorpan>
^_=

[02:21:05.0000] <zcorpan>
or, what the hell

[02:22:00.0000] <zcorpan>
^:=

[02:23:00.0000] <zcorpan>
that looks ok

[02:24:00.0000] <woef>
http://images2.layoutsparks.com/1/112359/punisher-black-white-face.jpg -> looks a lit like the Punisher logo :p

[02:25:00.0000] <zcorpan>
awesome

[02:25:01.0000] <jgraham>
/me doesn't like the fact that CSS loosk increasingly like a cat walking across the number keys whilst holding down shift

[02:30:00.0000] <woef>
^_= looks like you just punched someone in the face. Which I guess is exactly what you want to do if you have to worry about case sensitivity in html attributes :x

[03:16:00.0000] <Philip`>
zcorpan: Maybe [lc(foo)^=bar] so it's case-sensitive matching a forced-to-lowercase value

[03:22:00.0000] <zcorpan>
wfm

[03:23:00.0000] <zcorpan>
but wouldn't you do [foo^=lc(bar)] ?

[03:25:00.0000] <Philip`>
bar is the string the user typed into the selector and they'll have already written that in lowercase, so no

[03:26:00.0000] <Philip`>
foo is the attribute which they don't know the case of and want to force to lowercase before matching

[03:26:01.0000] <zcorpan>
fair enough

[04:37:00.0000] <gsnedders>
jgraham_: re2 is O(n) when matching, though can be O(n^2) compiling the regexp. It's the tradeoff you make for converting the NFA to a DFA.

[04:45:00.0000] <Philip`>
O(n^2) memory usage too, I think

[04:54:00.0000] <jgraham_>
gsnedders: But you could compile once per document rather than once per match, so it seems much more acceptable

[04:54:01.0000] <jgraham_>
Memory usage might be more of an issue

[04:54:02.0000] <jgraham_>
Not that I am advocating this at all

[04:59:00.0000] <Ms2ger>
MikeSmith, might be good to redirect http://people.w3.org/mike/web-platform/

[04:59:01.0000] <Ms2ger>
Or at least remove D3C :)

[05:01:00.0000] <Ms2ger>
MikeSmith, and add a link to xml-stylesheet?

[05:08:00.0000] <MikeSmith>
Ms2ger: done and done

[05:08:01.0000] <MikeSmith>
thanks

[05:08:02.0000] <Ms2ger>
Thanks

[05:09:00.0000] <Ms2ger>
I think that was the reason I've got "<?xml-stylesheet" written on my desk

[05:09:01.0000] <Ms2ger>
/me gets an eraser

[05:10:00.0000] <MikeSmith>
heh

[05:17:00.0000] <gsnedders>
Philip`: Indeed.

[05:19:00.0000] <zcorpan>
Ms2ger: surely you had it on your desk because you wanted to praise me for having worked on it

[05:20:00.0000] <Ms2ger>
zcorpan, it's a mess of a spec

[05:20:01.0000] <Ms2ger>
zcorpan, I mean, er..

[05:20:02.0000] <Ms2ger>
Huh

[05:20:03.0000] <Ms2ger>
http://html5.org/tools/web-apps-tracker?from=6403&to=6404

[05:22:00.0000] <zcorpan>
Ms2ger: i blame the WG for working against me :P

[05:22:01.0000] <Ms2ger>
I guess that's actually true

[05:24:00.0000] <Ms2ger>
/me shakes his fist at XMLCore

[05:28:00.0000] <MikeSmith>
nice, I just committed revision 666 of the validator.nu schema

[05:29:00.0000] <MikeSmith>
/me breaks out some champagne and upside-down crosses

[05:34:00.0000] <MikeSmith_>
test automation UI that Francois set up is nice:

[05:34:01.0000] <MikeSmith_>
http://w3c-test.org/framework/suite/nav-timing-default/

[05:35:00.0000] <zcorpan>
MikeSmith: what wsa the commit?

[05:35:01.0000] <Ms2ger>
Section "4.3: The"

[05:36:00.0000] <MikeSmith>
zcorpan: https://bitbucket.org/validator/syntax/changeset/51b2578b6d58

[05:38:00.0000] <zcorpan>
evil

[05:39:00.0000] <jgraham_>
MikeSmith: Uh, opinions about that might vary :)

[05:39:01.0000] <MikeSmith>
well, it's a start

[05:39:02.0000] <MikeSmith>
it was totally manual before

[05:39:03.0000] <jgraham_>
I was quite surprised that it auto0submitted my results to the W3C

[05:40:00.0000] <jgraham_>
Maybe it said somewhere that it would, but that makes it pretty toxic for people who don't work entirely in public

[05:41:00.0000] <jgraham>
I should point this out on the mailing list

[05:43:00.0000] <MikeSmith>
oh

[05:43:01.0000] <MikeSmith>
yeah

[05:43:02.0000] <MikeSmith>
true

[05:44:00.0000] <MikeSmith>
needs a warning and/or confirmation dialog of some kind

[05:47:00.0000] <jgraham>
In general I am a test-framework skeptic

[05:47:01.0000] <Ms2ger>
[x] Report test results automatically (when possible)

[05:47:02.0000] <Ms2ger>
/me missed that as well the first time

[05:48:00.0000] <jgraham>
I don't think the problem it is trying to solve ("we need test reports to get to CR") is actually the right problem to focus on

[05:49:00.0000] <jgraham>
The right problem to focus on is "how do we get vendors running all the tests every day"

[05:49:01.0000] <jgraham>
Once you have that, the "we need the results" thing becomes trivial

[05:49:02.0000] <Ms2ger>
Working on that...

[05:50:00.0000] <Ms2ger>
(And IE every three days, I presume)

[05:50:01.0000] <jgraham>
Heh

[05:50:02.0000] <jgraham>
/me is working on it too

[05:51:00.0000] <Ms2ger>
Next up: useful tests

[05:51:01.0000] <jgraham>
Yep. We are getting there, slowly

[05:52:00.0000] <Ms2ger>
And mostly on Opera's back, or so it seems

[05:53:00.0000] <Ms2ger>
  <noscript><p>Enable JavaScript and reload</p></noscript> <-- Is that common in your tests? :)

[05:53:01.0000] <jgraham>
That is in our internal guidelines

[05:53:02.0000] <jgraham>
I have suggested it is silly

[05:54:00.0000] <jgraham>
But apparently it isn't so silly that the people who have been doing it habitually want to stop

[05:55:00.0000] <Ms2ger>
The same guidelines that want window.undefined? :)

[05:56:00.0000] <jgraham>
No, that's just Tarquin :)

[05:56:01.0000] <zcorpan>
maybe i should start writing window.frames.self.undefined

[05:56:02.0000] <zcorpan>
you can almost form a sentence with attributes that return window

[05:59:00.0000] <zcorpan>
hmm there weren't more than those, how disappointing

[05:59:01.0000] <Ms2ger>
Gecko has _content

[05:59:02.0000] <Ms2ger>
And content

[05:59:03.0000] <Ms2ger>
Not sure what they do exactly

[06:00:00.0000] <zcorpan>
huh, didn't know about those

[06:02:00.0000] <Ms2ger>
Interestingly enough, if you set window.content yourself, _content will mirror that

[06:04:00.0000] <zcorpan>
window === content // false

[06:04:01.0000] <zcorpan>
window === _content // false

[06:04:02.0000] <zcorpan>
content is enumerable but _content is not

[06:05:00.0000] <zcorpan>
_content = 1 does nothing

[06:06:00.0000] <Ms2ger>
Nope

[06:07:00.0000] <Ms2ger>
We only handle it when getting

[06:07:01.0000] <Ms2ger>
/me wants all this code to die

[06:08:00.0000] <zcorpan>
is there a bug?

[06:09:00.0000] <Ms2ger>
For getting bindings that don't suck as much? Yes, probably

[06:11:00.0000] <zcorpan>
i meant for dropping content and _content

[06:12:00.0000] <Ms2ger>
Might be something for extensions

[06:23:00.0000] <zcorpan>
MikeSmith: is it possible to tweak the settings for which emails go to public-html-bugzilla so that if only the cc list is changed, don't send an email?

[06:24:00.0000] <Ms2ger>
Should be

[06:25:00.0000] <MikeSmith>
zcorpan: yeah

[06:25:01.0000] <MikeSmith>
will do that in a bit

[06:25:02.0000] <MikeSmith>
but now, breakfast

[08:00:00.0000] <benjoffe_>
I'm surprised that no browser has implemented a special readable stylesheet for html pages that have no styles or scripts

[08:01:00.0000] <Ms2ger>
I'm not

[08:01:01.0000] <Ms2ger>
There is lots of other stuff to do

[08:02:00.0000] <miketaylr>
safari has that reader/readability feature

[08:02:01.0000] <benjoffe_>
This wouldn't be hard to implement, and there's tonnes of pages out there like this, especially academic content

[08:03:00.0000] <benjoffe_>
miketaylr: I'm thinking something like that but with all the links and forms styled well inline too

[08:05:00.0000] <AryehGregor>
You can use a bookmarklet.

[08:05:01.0000] <AryehGregor>
I have one called Readability.

[08:05:02.0000] <AryehGregor>
I use it for pages with horrible contrast or such.

[08:07:00.0000] <benjoffe_>
I know there's ways for me to consume this content, but I'm just surprised none of them have anything decent on by default

[08:15:00.0000] <AryehGregor>
Someone seriously needs to tell Google Maps about the 184th Street entrance to the 181st Street A station.  And the Bennett Avenue entrance to the 190th A.  It seems to think there's only one entrance to each station, and it totally messes up the directions it gives.

[08:16:00.0000] <AryehGregor>
Like thinking it makes as much sense to tell you to walk to 181st and St. Nicholas to get to the 1 as walking to 181st and Fort Washington to get to the A, when you're starting at 184th and Overlook and there's an entrance to the A like fifty feet away.

[08:16:01.0000] <gsnedders>
AryehGregor: Not as bad as Subway station closest to me, it's ignorant of there is a staircase down from bridge to where it is, so makes you go a huge long way around

[08:17:00.0000] <zewt>
as much as I like gmaps, I really wish Garmin would make an Android release

[08:17:01.0000] <AryehGregor>
It does that with the 190th A.

[08:18:00.0000] <AryehGregor>
There's an entrance on Bennett and one on Fort Washington, and to get from one to the other you have to walk several blocks around because Fort Washington is on top of a cliff overlooking Bennett at that point.

[08:18:01.0000] <AryehGregor>
Google Maps only knows about the Fort Washington entrance.

[08:19:00.0000] <gsnedders>
It's really quite amusing around where I used to live: such a random selection of footpaths on it.

[08:19:01.0000] <AryehGregor>
/me tries http://maps.google.com/support/bin/answer.py?answer=162873

[08:21:00.0000] <zewt>
is glenn adams trolling? heh

[08:21:01.0000] <Ms2ger>
zewt, no comment :)

[08:25:00.0000] <jgraham>
Someone should tell him why W3C publishes "technical recommendations"

[08:26:00.0000] <Ms2ger>
De Jure Standards, I'm telling you!

[08:26:01.0000] <gsnedders>
AryehGregor: Not available in the UK, so can't do that for here…

[08:26:02.0000] <jgraham>
SOmeone whould tell me how the fuck I have managed to break virtualbox so that installing guest additions has no effect

[08:27:00.0000] <zewt>
through the power of song

[09:31:00.0000] <dglazkov>
good morning, Whatwg!

[10:55:00.0000] <AryehGregor>
/me does not understand how someone can *consistently* misspell his name as "Areyeh" when communicating in a text medium where his name is presented with the correct spelling on every post

[10:56:00.0000] <gesa>
Someone doesn't know how to properly utilize copy & paste. Clearly

[11:00:00.0000] <Philip`>
AryehGregor: Some people consistently can't even spell "Philip" :-(

[11:00:01.0000] <Ms2ger>
Phillip`: Hmm?

[11:01:00.0000] <Philip`>
/me generally avoids referring to people by name entirely, and sticks with pronouns where possible, else is careful to double-check and/or copy-and-paste their names

[11:04:00.0000] <jgraham>
/me gets Philip wrong at least 50% of the time unless there is tab completion

[11:08:00.0000] <jgraham>
(otoh, people very often forget that the possesive of James is spelt James')

[11:08:01.0000] <Ms2ger>
Jame's or James's?

[11:09:00.0000] <zewt>
james's

[11:09:01.0000] <zewt>
there is not more than one jame :)

[11:09:02.0000] <jgraham>
mmm jam

[11:09:03.0000] <jgraham>
Sorry we were talking about food, right?

[11:10:00.0000] <jgraham>
I believe that the extra s is also acceptable

[11:10:01.0000] <zewt>
people regularly call me "glen", even when writing my name in an email where my name, correctly spelled, is right there on screen

[11:10:02.0000] <zewt>
i'd consider james' flatly incorrect, fwiw

[11:11:00.0000] <Ms2ger>
/me raises a FORMAL COMPLAINT against James'

[11:11:01.0000] <jgraham>
Well maybe I get it wrong then :p But the mistake I had in mind was not having any suffix at all

[11:11:02.0000] <gsnedders>
zewt: What is your name?

[11:12:00.0000] <zewt>
glenn

[11:12:01.0000] <gsnedders>
I see all kinds of variations of Geoffrey. Most commonly "Geoffery".

[11:12:02.0000] <Ms2ger>
Are you the Glenn on webapps?

[11:12:03.0000] <zewt>
there are two Glenns on webapps

[11:12:04.0000] <zewt>
one of which I am not

[11:12:05.0000] <jgraham>
zewt: You should tell the other Gleen that he is brining your name into disrepute

[11:12:06.0000] <Ms2ger>
It would be weird if you were both

[11:12:07.0000] <zewt>
i'm certainly not the one (seemingly) trolling against whatwg today, heh

[11:13:00.0000] <Ms2ger>
Good

[11:13:01.0000] <jgraham>
Not much weirder than the two Philip Taylors on public-html

[11:13:02.0000] <zewt>
it's sad when it's more important to explain clearly who you are not than who you are

[11:13:03.0000] <zewt>
it sure is confusing when hixie refers to "Ian"

[11:13:04.0000] <Ms2ger>
That

[11:13:05.0000] <Philip`>
/me would generally consider "James'" to be probably technically acceptable but old-fashioned and silly, kind of like "an historic" or "the data are"

[11:14:00.0000] <Ms2ger>
/me thinks Philip` is an hero

[11:14:01.0000] <jgraham>
/me thinks that insisting data is plural is... very annoying

[11:14:02.0000] <zewt>
i'd personally consider it technically incorrect because the "s'" construction is for plural possessives

[11:15:00.0000] <zewt>
at least "an historic" and "data are" i know where they come from

[11:15:01.0000] <Ms2ger>
Octopodes, I'm telling you

[11:17:00.0000] <Philip`>
zewt: It's not childrens' etc - I see it as more related to ending with an s rather than being a plural (and rather than being a plural that ends with an s)

[11:17:01.0000] <zewt>
and i assume everyone else over 20 is also righteously infuriated over such nonsense as "on accident"

[11:18:00.0000] <Philip`>
(though I'm just going by what seems right, rather than what is officially right in some sense)

[11:19:00.0000] <jgraham>
Yeah, I would also say "gsnedders'" but "zewt's"

[11:19:01.0000] <zewt>
Philip`: but the "'s" is pronounced as a separate "s", where "s'" is not; for example, we do pronounce two s's in "boss's"

[11:19:02.0000] <jgraham>
(umm, obviously the latter)

[11:19:03.0000] <zewt>
likewise for "james's", it's not pronounced as "james desk"

[11:19:04.0000] <Ms2ger>
!summon timeless

[11:19:05.0000] <Philip`>
zewt: Surely nobody would say something like that by purpose?

[11:19:06.0000] <zewt>
*smack*

[11:20:00.0000] <jgraham>
zewt: You seem to be mistakenly assuming a clean mapping between English spelling and pronounciation

[11:20:01.0000] <zewt>
people under around 20 seem to be saying "on purpose" a lot, which makes me sad (there are some others, though I'm having trouble recalling them off-hand)

[11:21:00.0000] <zewt>
there's a reasonably consistent mapping between "s'" and "s's" and one or two s sounds, in all cases I can think of

[11:21:01.0000] <zewt>
english certainly isn't a consistent language, but that doesn't mean it has no consistencies at all :)

[11:21:02.0000] <Ms2ger>
Those it has are probably accidents

[11:23:00.0000] <MikeSmith>
it would be very cool to actually rescind a Rec

[11:23:01.0000] <MikeSmith>
for a number of reasons

[11:23:02.0000] <MikeSmith>
including having a precedent

[11:23:03.0000] <zewt>
show that it can be done? heh

[11:25:00.0000] <Philip`>
Let's rescind them all and start again from scratch

[11:26:00.0000] <Ms2ger>
Philip`, want to write two paragraphs to rescind Views?

[11:27:00.0000] <Philip`>
Since I don't know what it is, no

[11:27:01.0000] <timeless>
anyone here familiar w/ <input type=number> ?

[11:30:00.0000] <Ms2ger>
Ah, timeless

[11:30:01.0000] <Ms2ger>
timeless, want to write two paragraphs to rescind Views? :)

[11:33:00.0000] <Philip`>
I guess that's a no?

[11:34:00.0000] <Ms2ger>
:(

[12:10:00.0000] <Ms2ger>
So delete window.Node should do what?

[12:18:00.0000] <AryehGregor>
Nothing good, I suspect.

[12:18:01.0000] <Ms2ger>
>.<

[12:19:00.0000] <Ms2ger>
It seems like the answer is "Nothing", but I can't figure out why that is

[12:21:00.0000] <AryehGregor>
/me recommends summoning heycam|away 

[12:22:00.0000] <Ms2ger>
I tried

[12:22:01.0000] <Ms2ger>
He's always away

[12:22:02.0000] <Ms2ger>
I mean, it's past 6AM in NZ

[12:32:00.0000] <gsnedders>
Ms2ger: Well, I guess window.Node.[[Configurable]] = false

[12:33:00.0000] <Ms2ger>
No

[12:33:01.0000] <gsnedders>
Ms2ger: But I guess you mean in IDL terms?

[12:33:02.0000] <Ms2ger>
The property has the attributes { [[Writable]]: true, [[Enumerable]]: false, [[Configurable]]: true }.

[12:34:00.0000] <Ms2ger>
http://dev.w3.org/2006/webapi/WebIDL/#es-interfaces

[12:35:00.0000] <gsnedders>
Ms2ger: The global object is odd and defined nowhere in real terms :(

[12:35:01.0000] <Ms2ger>
/me sighs

[12:36:00.0000] <Ms2ger>
/me files a bug

[12:38:00.0000] <gsnedders>
Ms2ger: Look at self and parent for the *real* gfun.

[12:38:01.0000] <gsnedders>
*fun

[12:38:02.0000] <Ms2ger>
No thanks

[12:38:03.0000] <w3ztb0y>
is there any indonesian people??

[12:39:00.0000] <Ms2ger>
Presumably there are

[12:39:01.0000] <Ms2ger>
AFAIK, Indonesia hasn't been abandoned

[12:39:02.0000] <AryehGregor>
Even if it had, wouldn't there still be Indonesian expatriates?

[12:40:00.0000] <Ms2ger>
Would you call them Indonesian people in that case? Presumably yes.

[12:41:00.0000] <AryehGregor>
Sure.

[12:41:01.0000] <TabAtkins>
Hixie: whitespace-collapse:discard isnt' yet implemented anywhere afaik.  But it's definitely awesome, for precisely the case you mention.

[12:42:00.0000] <AryehGregor>
it does sound awesome.

[12:42:01.0000] <AryehGregor>
w3ztb0y, no, no one here is Indonesian.

[12:42:02.0000] <TabAtkins>
zcorpan, others: Selectors 4 introduced a way to match attributes case-insensitively a few weeks ago.

[12:42:03.0000] <Hixie>
TabAtkins: yeah once i saw it i realised it was what we were talking about back in 2003 or so

[12:42:04.0000] <Hixie>
TabAtkins: i still think we should get rid of the three properties and still only use white-space, with whatever keywords make sense

[12:42:05.0000] <Ms2ger>
Hixie, then it should be implemented in 5 years or so

[12:43:00.0000] <Hixie>
TabAtkins: since the vast majority of the combinations make no sense at all

[12:43:01.0000] <Ms2ger>
Send email to www-style

[12:43:02.0000] <AryehGregor>
TabAtkins, won't that force browsers to keep a separate case-normalized copy of every attribute value?  Do they actually want to do that?

[12:43:03.0000] <Hixie>
Ms2ger: dude, you obviously aren't familiar with the csswg's specs' schedules (especially fantasai's...)

[12:43:04.0000] <Ms2ger>
The CSSWG apparently thinks you don't know about www-style

[12:44:00.0000] <Hixie>
by "the CSSWG" do you mean daniel? :-)

[12:46:00.0000] <TabAtkins>
Hixie: Nah, fantasai was making fun of you for saying that you weren't sure how to ask the CSSWG to handle ::paragraphs.

[12:46:01.0000] <Hixie>
what i wasn't sure about was whether there was some convenient way to just reassign the bug

[12:47:00.0000] <TabAtkins>
No, we don't use the bugtracker for feature requests.

[12:47:01.0000] <Hixie>
lame

[12:47:02.0000] <TabAtkins>
Those are solely handled on the mailing list.

[12:47:03.0000] <Hixie>
where are they tracked?

[12:47:04.0000] <Ms2ger>
The editor's heads

[12:47:05.0000] <Hixie>
i'm going to go back to "lame"

[12:47:06.0000] <Ms2ger>
Hmm

[12:47:07.0000] <TabAtkins>
Various places.  We're consolidating to *tracking* on a bugtracker.

[12:48:00.0000] <TabAtkins>
s/a bugtracker/bugzilla/

[12:48:01.0000] <Ms2ger>
That's a bit of a strange picture

[12:48:02.0000] <TabAtkins>
But discussion shouldn't ever happen on bugzilla.

[12:48:03.0000] <Ms2ger>
*The editors' heads

[12:48:04.0000] <TabAtkins>
 /lunch

[12:49:00.0000] <Hixie>
well where discussion happens is not really my concern

[12:49:01.0000] <Hixie>
i just wanted to get the bug off my list :-P

[12:49:02.0000] <zewt>
when you run out of space, do you have to buy a razor

[14:01:00.0000] <TabAtkins>
Hixie: I think that wrapping and whitespace collapsing and newline collapsing are all distinct, and useful to control separately, so the separate properties are good.

[14:04:00.0000] <jgraham>
/me blinks

[14:05:00.0000] <TabAtkins>
jgraham: Don't stop blinking!

[14:06:00.0000] <jgraham>
TabAtkins: It's OK I think that I will keep blinking for as long as I keep getting email about the W3C test framework thing.

[14:07:00.0000] <jgraham>
It is being built with a set of assumptions that are quite alien to me

[14:07:01.0000] <jgraham>
/me should try to ignore it

[14:28:00.0000] <Hixie>
TabAtkins: i think giving authors switches where the majority of possible setting combinations are bad when we can enumerate the good ones is bad UI

[14:28:01.0000] <matjas>
is there a point in using @defer when you only use a single <script> and it’s at the bottom, right before </body>?

[14:28:02.0000] <Hixie>
not really

[14:30:00.0000] <matjas>
not really or not at all?

[14:30:01.0000] <matjas>
what is the point?

[14:31:00.0000] <Hixie>
there's no point that i can think of

[14:31:01.0000] <Ms2ger>
Being fancy! :)

[14:31:02.0000] <Hixie>
there are some subtle minor differences, but nothing useful i don't think

[14:35:00.0000] <matjas>
Hixie: it seems to make a difference in rendering, at least in Chrome

[14:35:01.0000] <matjas>
compare http://stevesouders.com/cuzillion/?c0=bi1hfff2_0_f&c1=bj1hfff2_0_f&t=1313008395 and http://stevesouders.com/cuzillion/?c0=bi1hfff2_0_f&c1=bj1hfft2_0_f&t=1313008412

[14:36:00.0000] <matjas>
first test is blank until the script is loaded, second test (with @defer) displays the table “immediately”, then shows the last line after the script is loaded

[14:36:01.0000] <Ms2ger>
QoI

[14:36:02.0000] <Hixie>
that's a browser detail

[14:40:00.0000] <Hixie>
jgraham: 504ing

[15:20:00.0000] <jgraham>
Hixie: I really don't know what to do. The only message in the logs is about the ekkepalive connection being closed which afaict from the docs is an effect rather than the cause

[15:21:00.0000] <jgraham>
*keepalive

[15:21:01.0000] <Hixie>
any idea _which_ connection?

[15:21:02.0000] <Hixie>
when does the problem occur?

[15:28:00.0000] <jgraham>
Well I assume that what's happening is that the host has changed their setup to use nginx as a frontend for all requests and then send them to apache if needed. And sometimes the script takes a little bit too long and nginx gets bored waiting for a response

[15:28:01.0000] <Hixie>
aaah

[15:28:02.0000] <Hixie>
interesting

[15:28:03.0000] <Hixie>
no way to change the timeout?

[15:28:04.0000] <Hixie>
maybe make your script send stuff back regularly?

[15:29:00.0000] <Hixie>
i ignore what you send back, iirc

[15:29:01.0000] <Hixie>
oh no wait

[15:29:02.0000] <jgraham>
I'm looking to see if there is a timeout

[15:29:03.0000] <Hixie>
that's for the multipage thing

[15:29:04.0000] <Hixie>
i actually use your output

[15:29:05.0000] <Hixie>
you could send back something in a comment at the top that i then strip out

[15:29:06.0000] <Hixie>
<!-- progress... line 1000... line 2000... line 3000...

[15:29:07.0000] <Hixie>
or whatever

[15:34:00.0000] <jgraham>
I probably can

[15:34:01.0000] <jgraham>
But it is going to be a crazy hack

[15:35:00.0000] <Hixie>
up to you :-)

[15:35:01.0000] <Hixie>
on my end the entire pipeline is one long series of crazy hacks

[15:36:00.0000] <Philip`>
Only a long series, not a complex interconnected DAG?

[15:37:00.0000] <Philip`>
(or even a cyclic one)

[15:38:00.0000] <gsnedders>
Philip`: Then it wouldn't be a DAG

[15:38:01.0000] <jgraham>
gsnedders: That's presumably why he said "or even"

[15:38:02.0000] <Philip`>
I didn't mean to imply it would be

[15:40:00.0000] <jgraham>
Hixie: I think it is too late at night for crazy hacks right now. Hopefully I will wake up with a better idea than I have right now

[16:11:00.0000] <Hixie>
anyone know if the css rules define how to serialise rgba()'s alpha value?

[16:15:00.0000] <tantek>
as in how many decimal places?

[16:15:01.0000] <nimbu>
http://www.w3.org/TR/2003/REC-SVG11-20030114/masking.html#SimpleAlphaBlending ?

[16:16:00.0000] <nimbu>
(from http://www.w3.org/TR/2011/REC-css3-color-20110607/#alpha )

[16:16:01.0000] <TabAtkins>
Hixie: They do not.

[16:16:02.0000] <Hixie>
hmm

[16:16:03.0000] <TabAtkins>
In general, serialization is a big pile of undefined crap.

[16:17:00.0000] <TabAtkins>
;_;

[16:17:01.0000] <Hixie>
i thought anne had largely fixed that using my proposal

[16:18:00.0000] <tantek>
Hixie, he did

[16:18:01.0000] <TabAtkins>
anne hasn't done anything with it, as far as I know.  cssom still just has an issue for both <color> and <number>

[16:18:02.0000] <tantek>
but then we introduced new things into CSS

[16:18:03.0000] <tantek>
like rgba

[16:18:04.0000] <tantek>
/me figures Anne will do something sensible.

[16:19:00.0000] <TabAtkins>
Now that fantasai and I are co-editors on the Values & Units spec, we intend to try and define serialization there.

[16:19:01.0000] <Hixie>
man i wish values & units would be implemented by browsers already

[16:19:02.0000] <Hixie>
i want vh and vw so bad

[16:20:00.0000] <Hixie>
"User agents must express the fractional part of the alpha value, if any, with the level of precision necessary for the value, when reparsed, to be interpreted as representing the same alpha value.

[16:20:01.0000] <Hixie>
"

[16:20:02.0000] <Hixie>
still doesn't quite define what the value should be

[16:21:00.0000] <Hixie>
e.g. if the underlying precision is a 2-bit precision (0%, 33%, 67%, 100%) it doesn't say whether to express 33% as 0.2, 0.3, or 0.4...

[16:22:00.0000] <Hixie>
and it doesn't disallow 0.33333333333333333333

[16:22:01.0000] <Hixie>
well i guess it does disallow it

[16:22:02.0000] <Hixie>
since that level of precision isn't necessary

[16:22:03.0000] <Hixie>
hmm

[16:23:00.0000] <Hixie>
ok well we'll go with that for now

[16:24:00.0000] <TabAtkins>
Hixie: IE implements the viewport-relative uits.

[16:24:01.0000] <Hixie>
wow, really?

[16:24:02.0000] <Hixie>
go IE

[16:24:03.0000] <TabAtkins>
Yup.

[16:24:04.0000] <Hixie>
how about calc()? anyone got that yet?

[16:25:00.0000] <TabAtkins>
(Half-unfortunately, they did so unprefixed, which makes it somewhat more painful to change vm to vmin (so we can allow vmax)).

[16:25:01.0000] <Hixie>
i specced that like half a decade ago already

[16:25:02.0000] <TabAtkins>
Yes, Firefox and IE have it, Webkit is getting it.

[16:25:03.0000] <Hixie>
what's vm?

[16:25:04.0000] <zewt>
i wonder if the algorithm used by python for representing floats as strings is clearly specified

[16:25:05.0000] <TabAtkins>
vm = min(vh,vw)

[16:25:06.0000] <Hixie>
ah

[16:25:07.0000] <Hixie>
just add vb

[16:25:08.0000] <Hixie>
for "big"

[16:25:09.0000] <Hixie>
(if you really want it)

[16:25:10.0000] <Hixie>
(what's the use case for vb?)

[16:26:00.0000] <Hixie>
or just use min()!

[16:26:01.0000] <TabAtkins>
Dunno yet.  I suspect there's one there, but I wouldn't add it yet.

[16:26:02.0000] <Hixie>
calc(min(1vh,1vw))

[16:26:03.0000] <TabAtkins>
Yes, just using min/max is fine too.

[16:26:04.0000] <Philip`>
Clearly max should be abbreviated to "vx"

[16:26:05.0000] <TabAtkins>
And remember, min/max dont' ahve to occur within calc() anymore!

[16:26:06.0000] <Hixie>
ah, cool

[16:26:07.0000] <Hixie>
well, cool once implemented

[16:27:00.0000] <Hixie>
i really wish there was just a single css spec i could keep up with instead of this mass of drafts

[16:27:01.0000] <Hixie>
anyway

[16:27:02.0000] <Hixie>
i should take a week to learn where browsers are with css

[16:28:00.0000] <jamesr>
some 2.1, some 3, and some 4

[16:28:01.0000] <jamesr>
of course

[16:28:02.0000] <TabAtkins>
You can just watch the yearly snapshots.

[16:29:00.0000] <Hixie>
when was the last one of those?

[16:29:01.0000] <TabAtkins>
That's by definition fairly behind, but still.

[16:29:02.0000] <TabAtkins>
last year.  Fantasai's publihsing one per year now.

[16:29:03.0000] <Hixie>
link?

[16:29:04.0000] <TabAtkins>
Or rather, intends to (she just started last year).

[16:29:05.0000] <TabAtkins>
http://www.w3.org/TR/css-2010/

[16:29:06.0000] <TabAtkins>
It's just the specs CR or later.

[16:30:00.0000] <Hixie>
oh

[16:30:01.0000] <Hixie>
well

[16:30:02.0000] <Hixie>
that's no good

[16:30:03.0000] <TabAtkins>
Sure, not right now.  css-2011 will be quite a bit bigger.

[16:30:04.0000] <tantek>
/me entertains the idea of an automatically combined spec of latest less-than-year-old WDs + CRs + PRs + RECs of CSS3 modules

[16:30:05.0000] <Hixie>
what i want to know is what's implemented, and what's still fiction, and apparently what the prefixes are.

[16:30:06.0000] <TabAtkins>
Hixie: Yeah, that's something else.  caniuse.com is a partial example

[16:31:00.0000] <tantek>
Hixie, putting things in one spec won't tell you "what's implemented, and what's still fiction", after all, HTML5 doesn't necessarily tell you what's implemented, and what's still fiction

[16:31:01.0000] <Hixie>
tantek: well we have hte little status boxes that try to tell you that

[16:31:02.0000] <Hixie>
tantek: but sure

[16:31:03.0000] <tantek>
Hixie, yeah, I like the status boxes

[16:31:04.0000] <Hixie>
tantek: it would, however, help me know what was out there

[16:32:00.0000] <tantek>
they're a definite improvement

[16:32:01.0000] <tantek>
I agree, to know "what was out there"

[16:33:00.0000] <tantek>
I think Eric Meyer complained rather vociferously on Twitter a few months ago about how hard it was for him to track down "all" the CSS3 specs and ascertain their relative levels of status (as opposed to claimed status in the document status section)

[16:33:01.0000] <Hixie>
yeah

[16:33:02.0000] <jamesr>
yeah i have no idea how to do that, and i'm pretty sure i've implemented some of it

[16:33:03.0000] <tantek>
I'm personally thinking of just starting to mark features as at risk or not based on implementation status

[16:33:04.0000] <tantek>
in the specs I edit

[16:33:05.0000] <tantek>
even before CR

[16:34:00.0000] <Hixie>
another way we could do this is by expanding the css section of platform.html5.org

[16:34:01.0000] <tantek>
I'm not sure centralized works

[16:34:02.0000] <tantek>
I think we have to upgrade what's expected of spec editors/authors

[16:34:03.0000] <tantek>
that scales better

[16:34:04.0000] <Hixie>
no argument from me there

[16:35:00.0000] <tantek>
is there a "how to" written up somewhere for how a spec editor could add the little status boxes?

[16:35:01.0000] <TabAtkins>
I'm happy to help out here.  I don't think it's appropriate to add an "Implementation Status" section to the specs (it'll be outdated shortly after publication), but maintaining a wiki page or something like that is easy.

[16:35:02.0000] <tantek>
/me is into writing-up how-to's for spec editors. e.g. http://wiki.csswg.org/spec/cvs

[16:36:00.0000] <tantek>
Tab - I'm suggestion per property notes, just like the status boxes

[16:36:01.0000] <TabAtkins>
/me still references that regularly.

[16:36:02.0000] <tantek>
/me does too.

[16:36:03.0000] <Hixie>
TabAtkins: in the case of the html spec the "how to" is "write a server-side CGI script that interfaces with a MySQL database, then write a JS front-end that interfaces with the CGI script"

[16:36:04.0000] <tantek>
/me has run into CVS problems, and googled for the error and found that document. Scary.

[16:36:05.0000] <Hixie>
er

[16:36:06.0000] <Hixie>
tantek: ^

[16:36:07.0000] <Hixie>
tab-completion fail

[16:36:08.0000] <TabAtkins>
tantek: Yes, I understand.  They just can't be part of the spec itself.  Having a script that grabs from something external is fine.

[16:36:09.0000] <Hixie>
oh, a pun! Tab-completion fail!

[16:36:10.0000] <TabAtkins>
Sigh.

[16:37:00.0000] <tantek>
lol

[16:37:01.0000] <zewt>
D:

[16:37:02.0000] <TabAtkins>
/me thinks he should go ahead and write that.

[16:37:03.0000] <tantek>
TabAtkins - that would be great if you added support for status boxes to CSS specs

[16:37:04.0000] <Hixie>
how widespread is CORS support?

[16:37:05.0000] <tantek>
/me would use it in his specs.

[16:37:06.0000] <Hixie>
i don't mind hosting the CGI side of this

[16:38:00.0000] <Hixie>
and then the scripts could just pull from that

[16:38:01.0000] <Philip`>
If there were status boxes in the CORS spec you'd be able to answer that yourself

[16:38:02.0000] <Hixie>
tis true

[16:38:03.0000] <Hixie>
we have a bootstrapping problem

[16:38:04.0000] <tantek>
I'll gladly test it. For now I've been putting in various notes into css3-ui. And yeah, it's ad hoc but I'm capturing it there because there is an expectation of imminent return to LCWD(2) and then CR (thus tracking "at risk" is important)

[16:39:00.0000] <Hixie>
/me wonders where the code for the status cgi script is...

[16:40:00.0000] <TabAtkins>
Hixie: No need for a cgi.  Just a json file would work.

[16:40:01.0000] <Hixie>
ah, here we go

[16:40:02.0000] <Hixie>
TabAtkins: how would you update it?

[16:40:03.0000] <TabAtkins>
...by editting it?

[16:40:04.0000] <Hixie>
oh

[16:40:05.0000] <Hixie>
well that's lame

[16:41:00.0000] <Hixie>
the html spec can be edited by anyone who has sent feedback!

[16:41:01.0000] <Hixie>
crowd-sourcing

[16:41:02.0000] <Hixie>
it's fashionable

[16:41:03.0000] <TabAtkins>
Bah.

[16:41:04.0000] <Hixie>
hah

[16:41:05.0000] <TabAtkins>
If we check it into dev.w3.org, then anyone with CVS access can update it.

[16:42:00.0000] <Hixie>
ok well in that case, no need for me to update the status.cgi script, excellent

[16:42:01.0000] <TabAtkins>
/me is squeamish around SQL these days, anyway.

[16:43:00.0000] <Hixie>
hah

[16:43:01.0000] <Hixie>
did it bite you?

[16:43:02.0000] <Philip`>
How many people have made substantial updates to the HTML spec stat boxes?

[16:43:03.0000] <TabAtkins>
Hixie: I've been bitten by corruption before.  Binary formats are the devil.

[16:43:04.0000] <Hixie>
ah well yeah

[16:43:05.0000] <Hixie>
i make daily text-form backups

[16:44:00.0000] <Hixie>
Philip`: please hold

[16:44:01.0000] <Philip`>
(If it's only like two people then a static file seems good enough)

[16:46:00.0000] <Hixie>
Philip`: how many changes is "substantial"

[16:47:00.0000] <Peter`>
ideally we'd have one central place to track implementation status of web platform features

[16:47:01.0000] <Peter`>
combining the stability of the specification (or rather, chapter in the specification) with the various rendering engines and associated browsers

[16:47:02.0000] <Peter`>
some kind of (semi-)official caniuse.com

[16:48:00.0000] <Philip`>
Hixie: Probably the shape of the distribution is more important than the numbers

[16:50:00.0000] <Hixie>
the answer seems to be "4". http://junkyard.damowmow.com/493

[16:51:00.0000] <TabAtkins>
4 besides you?  I'd rate zcorpan as substantial.

[16:51:01.0000] <Hixie>
4 including me

[16:51:02.0000] <Hixie>
zcorpan == simonp

[16:51:03.0000] <TabAtkins>
Yes.

[16:51:04.0000] <TabAtkins>
Oh, wait, didn't realize he was further down the list as well.

[16:52:00.0000] <Hixie>
(the database stores every change so i can nuke a griefer pretty easily by just deleting their changes and the status just reverts to the previous state)

[16:54:00.0000] <Philip`>
Looks like the long tail of contributors is actually a pretty small tail, so the crowdsourcing probably isn't hugely useful

[16:55:00.0000] <tantek>
/me likes wikis for crowd sourcing that kind of thing.

[16:56:00.0000] <hober>
indeed

[16:56:01.0000] <Hixie>
in this particular instance i think the real trouble is discoverability

[16:56:02.0000] <Hixie>
first you have to send feedback, then you have to log in, then you have to magically guess that you have to alt-double-click

[16:56:03.0000] <Hixie>
it's pretty obscure

[16:56:04.0000] <Hixie>
i've had several people ask me about it

[16:57:00.0000] <Hixie>
including a number of people trying to update the values by submitting bugs

[16:57:01.0000] <Hixie>
i think we could massively improve matters just by making it easier to update values

[16:57:02.0000] <Philip`>
Alt-double-click doesn't even work in my browser (since the window manager interprets it as the start of drag-moving a window), though I think shift-alt-double-click worked okay

[16:58:00.0000] <Hixie>
yeah i worked pretty hard to find a test that worked everywhere with some combination or other

[16:58:01.0000] <Hixie>
but anyway, we could e.g. have non-logged-in be able to submit values that displayed their IP instead of them having to log in first

[16:58:02.0000] <Hixie>
and we could put an "edit" link in the boxes


2011-08-11
[17:07:00.0000] <jamesr>
Philip`: qq - do you have any canvas 2d tests for globalCompositeOperation=copy and fillText()?

[17:07:01.0000] <jamesr>
in the suite?

[17:07:02.0000] <jamesr>
we fail on that currently in WebKit

[17:08:00.0000] <jamesr>
and implementations might use different logic for their text drawing routines than for other draw calls

[17:08:01.0000] <Hixie>
oh hey it turns out anyone can get an account, not just those who have sent feedback

[17:08:02.0000] <Hixie>
wonder when i changed that

[17:10:00.0000] <Philip`>
jamesr: I don't think so, though I have a note in my write-only todo list about that

[17:13:00.0000] <Hixie>
ok i've updated the status boxes to have an edit link

[17:13:01.0000] <Hixie>
also slightly tweaked them

[17:15:00.0000] <Hixie>
any transition experts in the house? how do i transition from height:0 to height:auto?

[17:15:01.0000] <jamesr>
lawl

[17:15:02.0000] <jamesr>
:(

[17:15:03.0000] <jamesr>
you don't

[17:15:04.0000] <hober>
heh

[17:15:05.0000] <jamesr>
you can measure what height:auto would resolve to and then transition from 0 to that value

[17:16:00.0000] <hober>
Hixie: transitions to or from auto don't work

[17:16:01.0000] <Hixie>
ok. that's lame.

[17:16:02.0000] <TabAtkins>
Hixie: Transition max-height from 0 to a mild overestimate of the height.

[17:16:03.0000] <Hixie>
someone fix that.

[17:16:04.0000] <TabAtkins>
Yes, it is totally fucking lame.

[17:16:05.0000] <Hixie>
TabAtkins: hmm, good idea.

[17:17:00.0000] <Hixie>
ok i hid the ID in the status boxes, it now appears on hover

[18:06:00.0000] <TabAtkins>
Cross-origin XHR is completely disallowed, right?  Even just sending the request, not just receiving the response?

[18:09:00.0000] <jamesr>
sicking: you there?

[18:10:00.0000] <sicking>
jamesr: yo

[18:10:01.0000] <sicking>
TabAtkins: how do you mean?

[18:10:02.0000] <sicking>
TabAtkins: you can do cross-origin XHR using CORS

[18:10:03.0000] <TabAtkins>
Right, but outside of that it's disallowed.

[18:10:04.0000] <sicking>
TabAtkins: which means that you can make the request in many cases without the server opting in

[18:11:00.0000] <TabAtkins>
sicking: What do you mean?  What I'm asking about is using XHR to send data from client->server only, not caring about the response back.

[18:11:01.0000] <TabAtkins>
(Actually asking about a different API that effectively accomplishes that.)

[18:12:00.0000] <sicking>
TabAtkins: yup, you can basically do that, as long as you use POST with a limited set of Content-Types

[18:12:01.0000] <TabAtkins>
Without having to invoke CORS?

[18:12:02.0000] <sicking>
well, it invokes CORS

[18:12:03.0000] <sicking>
but CORS doesn't do anything until it start receiving a response

[18:13:00.0000] <sicking>
since CORS is response-header based

[18:13:01.0000] <sicking>
note, that you can only do this if your POST is fairly "plain"

[18:13:02.0000] <TabAtkins>
That's what I thought.  So the actual message from client->server gets through regardless.

[18:13:03.0000] <sicking>
i.e. you can't set any custom headers

[18:13:04.0000] <sicking>
TabAtkins: yup

[18:13:05.0000] <TabAtkins>
kk.

[18:13:06.0000] <sicking>
TabAtkins: just like with <form>

[18:14:00.0000] <TabAtkins>
All right, cool.

[18:15:00.0000] <TabAtkins>
(Context is someone pinging me about the feasibility of a built-in bugreporting function, that lets you specify a url to post to, some custom data to send, and a list of UA-provided data that you'd like, so the browser can let the user decide what to send and then post for you.)

[18:15:01.0000] <jamesr>
sicking: for XHR 'text' can be considered a parsed format

[18:16:00.0000] <jamesr>
sicking: what happens if the text is encoded in utf-8 and only part of the codepoint sequence (or whatever unicode calls it) is in the most recently received packet?

[18:16:01.0000] <sicking>
jamesr: then you don't return that character until in the next progress notification

[18:16:02.0000] <jamesr>
but the idea is the browser could ditch most of the response?

[18:16:03.0000] <sicking>
jamesr: yup

[18:17:00.0000] <sicking>
all but the last few bytes

[18:18:00.0000] <jamesr>
would be nice

[18:18:01.0000] <sicking>
TabAtkins: just make sure to not even return a error code indicating if the submission succeeded or not

[18:18:02.0000] <TabAtkins>
Ah, good point.

[18:19:00.0000] <jamesr>
what happens if you try to set .responseType after receiving a few progress events? it throws an error?

[18:19:01.0000] <sicking>
yes, that's already the case

[00:28:00.0000] <zcorpan>
ooooh http://dev.w3.org/csswg/selectors4/#attribute-case

[00:36:00.0000] <MikeSmith>
Hixie: I don't understand http://html5.org/r/6413

[00:36:01.0000] <MikeSmith>
the spec already states that maxlength on input@type=number is not valid

[00:37:00.0000] <MikeSmith>
so validator.nu currently emits an error for it

[00:37:01.0000] <MikeSmith>
this change would cause it to emit a warning in addition to an error

[00:38:00.0000] <MikeSmith>
unless I'm misunderstanding something

[00:38:01.0000] <MikeSmith>
but http://www.whatwg.org/specs/web-apps/current-work/multipage/number-state.html#number-state lists maxlength under "The following content attributes must not be specified and do not apply to the element"

[00:40:00.0000] <MikeSmith>
and http://www.whatwg.org/specs/web-apps/current-work/multipage/the-input-element.html#input-type-attr-summary also indicates that maxlength is now allowed on input@type=number

[00:41:00.0000] <MikeSmith>
if the intent is to actually make it conforming but obsolete, it seems like those parts of the spec should be changed as well

[00:41:01.0000] <zcorpan>
MikeSmith: "should not" means it's allowed if you have a good reason

[00:42:00.0000] <MikeSmith>
OK

[00:42:01.0000] <MikeSmith>
but if so, those two parts of the spec should be changed

[00:42:02.0000] <zcorpan>
yeah

[00:42:03.0000] <MikeSmith>
I don't see how otherwise we can make any changes to the validation behavior

[00:43:00.0000] <MikeSmith>
I'm post a comment to the bug

[02:52:00.0000] <zcorpan>
so if i'm reading the spec right it allows <link rel="help shortcut icon" ...>

[02:53:00.0000] <zcorpan>
does that work in ie?

[02:54:00.0000] <zcorpan>
i guess it does since hsivonen tested it

[02:55:00.0000] <GlitchMr>
I will check :P

[02:57:00.0000] <GlitchMr>
http://glitchmr.pl/private/favtest.php

[02:57:01.0000] <GlitchMr>
It seems to work under IE9

[02:58:00.0000] <GlitchMr>
It also seems to work under Quirks Mode...

[02:58:01.0000] <GlitchMr>
validator.nu seems to accept such structure and it seems to not cause issues with IE

[03:00:00.0000] <GlitchMr>
But then, semantically it doesn't make sense

[03:00:01.0000] <GlitchMr>
How can be shortcut icon helpful?

[03:03:00.0000] <zcorpan>
heh

[03:04:00.0000] <zcorpan>
well it must be helpful for something else you wouldn't include it in the first place

[03:04:01.0000] <GlitchMr>
You can use <link rel="help" href="..."> for help pages for examples

[03:05:00.0000] <GlitchMr>
example*

[03:05:01.0000] <GlitchMr>
Many browsers show their help on "About"->"Help" or something like this. Some might have reason to insert help into webpage :P

[03:06:00.0000] <GlitchMr>
But favicon.ico is helpful by itself (in visual browsers), but it's not help page...

[03:13:00.0000] <zcorpan>
Hixie: can you make the status boxes increase z-index on hover so you can access ones that are overlapped by other status boxes?

[03:26:00.0000] <Ms2ger>
Oh hey, I'm below zcorpan in status annotation edits...

[03:26:01.0000] <Ms2ger>
/me goes off to fix that

[05:43:00.0000] <Ms2ger>
/me shoots

[09:35:00.0000] <dglazkov>
good morning, Whatwg!

[09:41:00.0000] <TabAtkins>
Alan Gresley just said my head might be filled with crack. ;_;

[09:42:00.0000] <jgraham>
Is this from the "chocolate teapot orbiting the sun" school of "might"?

[09:43:00.0000] <TabAtkins>
Not sure.  Alan's kinda crazy himself.

[09:45:00.0000] <Philip`>
TabAtkins: Sounds like you could make a fortune by selling your head

[09:46:00.0000] <AryehGregor>
TabAtkins, what probability are we talking about here?  I mean, not to put too fine a point on it, but that would m -- drat, Philip` beat me to it.

[09:46:01.0000] <AryehGregor>
I mean, what would that be, a few pounds of crack?  What's the street value of that?

[09:47:00.0000] <dglazkov>
I somehow doubt it's crack.

[09:47:01.0000] <TabAtkins>
I like my head.  If there's crack in there, evidence says that I'm using it to think.

[09:47:02.0000] <dglazkov>
ah. that's where the old "get cracking" saying comes from.

[09:47:03.0000] <Philip`>
Heads are nice, but money is nice too, so you have to balance both sides of the argument

[09:49:00.0000] <Ms2ger>
I mean, half a head would be about as useful, no?

[09:49:01.0000] <TabAtkins>
I use more than half of my head.

[09:50:00.0000] <gsnedders>
int a = 5; int *ptr = &a; size_t foo = ptr; <- foo is now the memory address that ptr points to, right?

[09:50:01.0000] <Ms2ger>
How much of that is just for waste-ink?

[09:50:02.0000] <Ms2ger>
gsnedders, surely that would warn

[09:50:03.0000] <TabAtkins>
Ms2ger: Ooh, good point.

[09:51:00.0000] <Philip`>
gsnedders: I think so, though I'm not sure size_t is guaranteed to be pointer-sized

[09:51:01.0000] <Ms2ger>
/me liked that name

[09:51:02.0000] <Philip`>
intptr_t is better if you want a pointer-sized int

[09:51:03.0000] <Ms2ger>
I think gsnedders actually wants = *ptr

[09:51:04.0000] <Philip`>
That would give foo = 5

[09:52:00.0000] <Ms2ger>
Right

[09:52:01.0000] <Ms2ger>
I knew that! ;)

[09:52:02.0000] <gsnedders>
Ms2ger: I don't want foo = 5; I want the address that ptr points to.

[09:52:03.0000] <Philip`>
/me has no idea what gsnedders actually wants

[09:52:04.0000] <Philip`>
Ah

[09:52:05.0000] <gsnedders>
The memory address of a, effectively.

[09:52:06.0000] <jgraham>
/me doesn't like to think about what gsnedders wants

[09:53:00.0000] <Philip`>
ptr is already the memory address

[09:53:01.0000] <Ms2ger>
Why the hell are you trying to stick that into a size_t?

[09:53:02.0000] <Philip`>
so if you copy it into another variable then it's still that memory address (modulo 2^something)

[09:53:03.0000] <gsnedders>
Philip`: Okay, that's what I thought.

[09:54:00.0000] <Ms2ger>
But you do want intptr_t or uintptr_t

[09:55:00.0000] <jgraham>
Gotta love C and all the million sometime-interchangable, but not safely, types

[09:57:00.0000] <Ms2ger>
Don't you like C++ more? It has namespaces!

[09:57:01.0000] <Philip`>
If you want to be like the cool kids you could store the pointer in the NaN space of a double

[09:57:02.0000] <Ms2ger>
Philip`, ... do you do that?

[09:57:03.0000] <Philip`>
I'm not cool enough :-(

[10:01:00.0000] <gsnedders>
Ms2ger: JS engines do, on 32-bit arches

[10:01:01.0000] <Philip`>
At least SpiderMonkey on 64-bit too

[10:02:00.0000] <Ms2ger>
JS engines are a mess of their own

[10:02:01.0000] <Philip`>
/me thinks nanboxing and nunboxing would be interesting new Olympic sports

[10:03:00.0000] <gsnedders>
Philip`: How? Assuming no pointer will be over 53-bits?

[10:03:01.0000] <Philip`>
gsnedders: 47 bits, I think

[10:04:00.0000] <gsnedders>
Then what about the other six? Obviously they have to reserve one NaN value as NaN, and it can't be zero (because then it's not NaN but Inf)

[10:05:00.0000] <gsnedders>
Ms2ger: (The reason why I was using size_t above was because that's what Cython does… although I know it isn't guaranteed to work…)

[10:09:00.0000] <Philip`>
gsnedders: amd64 only has 48-bit virtual address space, and I think the 48th bit is magic and is always 0 in interesting OSes

[10:09:01.0000] <gsnedders>
Philip`: Ah.

[10:10:00.0000] <Philip`>
(http://en.wikipedia.org/wiki/X86-64#Canonical_form_addresses etc)

[11:13:00.0000] <Hixie>
zcorpan: done (the z-index thing)

[11:18:00.0000] <Ms2ger`>
People might be somewhat interested in http://lists.w3.org/Archives/Member/w3c-ac-members/2011JulSep/0007.html (MO)

[11:53:00.0000] <espadrine>
Ms2ger`: There is no locked-up equivalent to that mailing-list, is there?

[11:54:00.0000] <Ms2ger`>
Hmm?

[12:19:00.0000] <espadrine>
Ms2ger`: We can't access http://lists.w3.org/Archives/Member/w3c-ac-members/2011JulSep/0007.html

[12:19:01.0000] <Ms2ger`>
Right

[12:20:00.0000] <espadrine>
without W3C names

[12:20:01.0000] <Ms2ger`>
Would've linked to a public copy if there was one

[12:21:00.0000] <espadrine>
too bad there isn't one

[12:24:00.0000] <gnarf>
seems to be an error in http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#url-decomposition-idl-attributes regarding the "port" attribute

[12:25:00.0000] <gnarf>
ooo wait

[12:25:01.0000] <gnarf>
nvm

[12:25:02.0000] <gnarf>
read a column header wrong

[12:26:00.0000] <gnarf>
cancel the alarm!!!

[12:26:01.0000] <gnarf>
/me hides

[14:29:00.0000] <AryehGregor>
/me discovers another person who spells his name Areyeh -- interesting

[14:46:00.0000] <KevinMarks>
there's now a .ss TLD coming. Who's going to get c.ss ?

[15:11:00.0000] <karlcow>
http://highscalability.com/blog/2011/8/10/leveldb-fast-and-lightweight-keyvalue-database-from-the-auth.html

[15:11:01.0000] <TabAtkins>
Ooh, that sounds fun.

[15:12:00.0000] <karlcow>
wondering if it will show up in Google line of products

[15:12:01.0000] <karlcow>
ah it's already

[15:12:02.0000] <karlcow>
"LevelDB is being used as the back-end for IndexedDB in Chrome. For designing how to map secondary indices into LevelDB key/values, look at how the IndexedDB support within Chrome is implemented."

[15:14:00.0000] <TabAtkins>
Yeah, sqlite has corruption issues when it runs for a long time (not really its fault - its from occasional misplaced writes in other processes).

[15:14:01.0000] <Hixie>
"misplaced writes"?

[15:15:00.0000] <TabAtkins>
You know, writes to a dangling pointer or something.  Otherwise known as "security vulnerabilities".

[15:15:01.0000] <Hixie>
how can you cause corruption in another _process_?

[15:15:02.0000] <Hixie>
don't modern OSes pretty much make that impossible?

[15:15:03.0000] <TabAtkins>
Dont' ask me.  I'm recalling lunchtime conversation.

[15:16:00.0000] <Hixie>
if the problem is other code in the _same_ process randomly stomping on the space used by the sqlite library, it seems any library would suffer the same problem

[15:16:01.0000] <AryehGregor>
Hixie, maybe when two processes have the same database file open and one writes to it in a way that's not coordinated with the other, so the other winds up reading garbage somehow?

[15:17:00.0000] <Hixie>
that'd be a bug in sqlite's locking

[15:17:01.0000] <AryehGregor>
I thought only one process can open an SQLite database at a time, though.

[15:17:02.0000] <AryehGregor>
Yeah, I dunno.

[15:17:03.0000] <TabAtkins>
AryehGregor: It wasn't sqlite's problem directly; it was definitely memory corruption coming from elsewhere.

[15:18:00.0000] <zewt>
so it's very wrong to say "sqlite has corruption issues"

[15:18:01.0000] <Hixie>
yeah i don't see why sqlite would be any more at risk than leveldb

[15:19:00.0000] <Hixie>
but what do i know :-)

[15:19:01.0000] <zewt>
which I doubt anyone would accept anyway; sqlite should be high on any competent programmer's top-five list of most reliable libraries in the world

[15:20:00.0000] <karlcow>
the storage library the highest on my list is definitely… my fridge!

[15:20:01.0000] <karlcow>
every day, a few times a day.

[15:21:00.0000] <TabAtkins>
zewt: From the conversation I had with a guy working on indexeddb, sqlite made some not-strictly-required architectural decisions that mean that it holds its memory for too long, making it more subject to this kind of corruption.

[15:22:00.0000] <zewt>
sounds like a very contrived argument

[15:22:01.0000] <TabAtkins>
I may be misstating things.

[15:22:02.0000] <TabAtkins>
zewt: I'm not sure what you're trying to imply.  We used sqlite.  We had regular db corruption.  We traced it to a problem with the design of sqlite.

[15:23:00.0000] <jgraham>
It seems to me that "it's not designed to be a k-v store" might be a god enough reason on its own to think about other solutions for indexeddb

[15:23:01.0000] <jgraham>
*good

[15:23:02.0000] <TabAtkins>
I'm not saying sqlite is horrible.  I'm saying it didn't suit our needs, so we designed a new one.

[15:23:03.0000] <jamesr>
i don't think the primary motivation for leveldb was sqlite db corruption bugs

[15:23:04.0000] <jamesr>
although it may have been one

[15:23:05.0000] <TabAtkins>
jgraham: Indeed, that too.

[15:23:06.0000] <jamesr>
jgraham: exactly

[15:23:07.0000] <zewt>
... if library X writes to random memory locations and corrupts memory owned by library Y, it is not in any conceivable sense library Y's fault for "holding that memory for too long"

[15:23:08.0000] <TabAtkins>
jamesr: It was the reason that FileSystem is switching from sqlite to leveldb, at least.

[15:23:09.0000] <zewt>
"i shot some bullets down the street and someone got hit, it's hit fault for standing there for so long"

[15:23:10.0000] <Philip`>
Most databases have a separate server process, so only bugs in that process could scribble over the database, whereas SQLite runs in the address space of the host application so it's vulnerable to bugs in any applications that touch the database

[15:24:00.0000] <TabAtkins>
zewt: Robustness.  If Y can avoid a lot of those problems, but it doesn't, then that's a problem.

[15:24:01.0000] <Philip`>
though LevelDB seems to be embedded in applications too so it'd be just as vulnerable

[15:24:02.0000] <zewt>
avoid problems caused by unrelated code smashing memory? sorry, that's utter nonsense

[15:24:03.0000] <gsnedders>
TabAtkins: Why have there not been issues with corruptions with the insane number of other people using SqLite3?

[15:24:04.0000] <TabAtkins>
zewt: I'm not blaming anybody, so don't treat it as a personal attack.  I'm saying that we needed something that didn't stand in the street quite so long.

[15:24:05.0000] <zewt>
problems caused by code smashing random memory (or whatever) is the fault of that code and that code alone

[15:24:06.0000] <karlcow>
http://google-opensource.blogspot.com/2011/07/leveldb-fast-persistent-key-value-store.html

[15:25:00.0000] <TabAtkins>
gsnedders: No clue.  Again, I'm relaying lunchtime conversation.

[15:25:01.0000] <zewt>
i'm not treating anything as a personal attack; i have nothing to do with sqlite development

[15:25:02.0000] <karlcow>
http://leveldb.googlecode.com/svn/trunk/doc/benchmark.html

[15:25:03.0000] <zewt>
i'm saying it's a ridiculous, nonsensical reason to change libraries.

[15:25:04.0000] <Hixie>
either it's a bug in sqlite that should be fixed, or it's a bug in the surrounding code and leveldb would be equally vulnerable, as far as i can tell

[15:25:05.0000] <jamesr>
no, there were others

[15:25:06.0000] <TabAtkins>
zewt: If you're shooting bullets into the street, it's your fault when you hit somebody.  But if you're relying on something that needs to cross the street, it's better to rely on one that won't spend as long in the crossfire.

[15:26:00.0000] <jgraham>
Quick to the car metaphors!

[15:26:01.0000] <TabAtkins>
Car metaphors explains EVERYTHING.

[15:26:02.0000] <jgraham>
+,

[15:26:03.0000] <zewt>
changing libraries means whoever's firing the bullets is going to be hitting other things, and it'll be some other random piece of code being broken

[15:27:00.0000] <TabAtkins>
zewt: ...yes?  And with luck, it'll be something that doesn't lose long-term data when it errors out.

[15:27:01.0000] <Hixie>
luck seems like a rather crappy design philosophy :-)

[15:27:02.0000] <zewt>
it's just nonsense FUD that'll lead less experienced developers to believe that "leveldb" (whatever that is) is more reliable than sqlite, which is bogus

[15:27:03.0000] <jamesr>
i'm pretty sure that leveldb is no safer to memory corruption in your process than anything else

[15:28:00.0000] <Hixie>
yeah i don't see how it could be

[15:28:01.0000] <TabAtkins>
zewt: Argh, I'm saying, very specifically, that one of *our* teams is running into this problem with sqlite, and is switching to leveldb because it fixes the architectural issue.

[15:28:02.0000] <zewt>
if you really want to protect from corruption, move your sqlite access into a helper process and proxy SQL commands to it

[15:28:03.0000] <jgraham>
/me thinks that having arguments based on half-remembered lunch conversations using metaphors involving violent death is probably not the road to enlightenment

[15:28:04.0000] <zewt>
which seems like a much simpler engineering task than changing backends

[15:29:00.0000] <zewt>
and much more effective

[15:29:01.0000] <TabAtkins>
zewt: I'm not talking about general engineering advice to the world at large.

[15:29:02.0000] <Philip`>
The road to enlightenment is strewn with randomly-shot bullets that luckily missed all the databases

[15:30:00.0000] <TabAtkins>
How true.

[15:30:01.0000] <jamesr>
since it's a piece of software

[15:30:02.0000] <jamesr>
/me agrees with jgraham and goes back to work

[15:30:03.0000] <jgraham>
Ah, well in that case I will know to take that one rather than the one paved with good intentions

[16:12:00.0000] <AryehGregor>
http://html5.org/r/6419 ++

[16:12:01.0000] <AryehGregor>
Way too many people get that wrong.

[16:12:02.0000] <Hixie>
srsly


2011-08-12
[01:31:00.0000] <zcorpan>
<http://www.w3.org/mid/D0BC8E77E79D9846B61A2432D1BA4EAE0341573B⊙Trcmc> - wonder if i should object

[03:04:00.0000] <jgraham>
zcorpan: I swear DOM3 events is happening in some alternate reality

[03:09:00.0000] <Ms2ger>
And it only connects to our reality every few months?

[03:15:00.0000] <Ms2ger>
Do browsers support object.contentWindow?

[03:19:00.0000] <Ms2ger>
Ah, looks like Opera does

[05:40:00.0000] <zcorpan>
oh hey, the splitter ids have been updated! yay

[05:41:00.0000] <zcorpan>
or not? complete/ has video in the-iframe-element.html but multipage/ has video in the-video-element.html

[05:41:01.0000] <zcorpan>
Philip`: ^

[05:43:00.0000] <Philip`>
I think my instance of the splitter is only used for multipage/

[05:43:01.0000] <Philip`>
(and is the latest code from SVN)

[05:44:00.0000] <Philip`>
I don't know where complete/ is generated

[05:44:01.0000] <zcorpan>
Hixie: ^

[07:49:00.0000] <karlcow>
http://norman.walsh.name/2011/08/12/styleThis

[07:49:01.0000] <karlcow>
Styling list numerations with CSS

[09:02:00.0000] <AryehGregor>
Oops! Google Chrome could not connect to software.hixie.ch

[09:02:01.0000] <AryehGregor>
Try reloading: software.­hixie.­ch/­utilities/­js/­live-­dom-­viewer/­

[09:02:02.0000] <AryehGregor>
Nooooooooooo

[09:02:03.0000] <Ms2ger>
And Firefox can't establish a connection to the server at junkyard.damowmow.com either

[09:03:00.0000] <AryehGregor>
But whatwg.org is up.  I thought they were on the same Dreamhost account?

[09:05:00.0000] <smaug____>
Hixie: you were writing down some ideas for <dialog>. I wonder where that page is

[09:05:01.0000] <Ms2ger>
wiki.whatwg.org somewher

[09:05:02.0000] <Ms2ger>
+e

[09:05:03.0000] <smaug____>
(since I wonder why we need dialog when we have showModalDialog)

[09:06:00.0000] <smaug____>
browsers could just implement showModalDialog in a bit different way

[09:28:00.0000] <dglazkov>
good morning, Whatwg!

[09:30:00.0000] <Ms2ger>
Evening

[09:31:00.0000] <manu-db>
Suggestion that the W3C TAG shut down the RDFa/Microdata Task Force initiative: http://lists.w3.org/Archives/Public/www-tag/2011Aug/0050.html

[09:42:00.0000] <jgraham>
Does anyone other than Opera support DOM 3 Load and Save?

[09:43:00.0000] <gsnedders>
jgraham: No

[09:45:00.0000] <karlcow>
jgraham: yes. Let me search, because I do not remember where

[09:46:00.0000] <Ms2ger>
We don't

[09:46:01.0000] <Ms2ger>
Except for some stuff on XHRProgressEvent

[09:46:02.0000] <karlcow>
Xerces-C++

[09:47:00.0000] <karlcow>
http://xerces.apache.org/xerces-c/program-dom-3.html

[09:48:00.0000] <karlcow>
Xerces is incomplete though

[09:48:01.0000] <karlcow>
there is also http://www.doxdesk.com/software/py/pxdom.html

[09:49:00.0000] <karlcow>
http://lists.w3.org/Archives/Public/www-dom/2003OctDec/0064.html

[09:49:01.0000] <karlcow>
and X-Hive, incomplete too

[09:49:02.0000] <karlcow>
http://www.w3.org/2003/10/DOM-Level-3-LS-implementations.html

[09:49:03.0000] <jgraham>
karlcow: That was a long way of saying "no"

[09:50:00.0000] <jgraham>
No one that needs to interoperate with web browsers

[09:50:01.0000] <karlcow>
heh, you didn't ask that question ;)

[09:50:02.0000] <jgraham>
This is #whatwg, it is implied :p

[09:51:00.0000] <karlcow>
haaa if you changed the rules

[09:51:01.0000] <Ms2ger>
There are no rules ;)

[09:51:02.0000] <karlcow>
sense of logic on #whatwg

[09:51:03.0000] <karlcow>
;)

[09:51:04.0000] <jgraham>
Well it would be logical that there was never a sense of logic. Having one sometimes is only illogical

[10:06:00.0000] <TabAtkins>
AryehGregor: http://nooooooooooooooo.com/

[10:09:00.0000] <AryehGregor>
for i in `seq 50`; do echo -n "$i: "; dig +short n`yes o | tr -d "\n" | head -c $i`.com; done

[10:09:01.0000] <AryehGregor>
Quite interesting, actually.

[10:13:00.0000] <AryehGregor>
(for i in `seq 62`; do dig +short n`yes o | tr -d "\n" | head -c $i`.com | tr "\n" " "; echo -e "\t$i"; done) | sort

[10:14:00.0000] <AryehGregor>
(for i in `seq 62`; do dig +short n`yes o | tr -d "\n" | head -c $i`.com; done) | sort | uniq -c | sort -nr

[10:14:01.0000] <AryehGregor>
Most of these point to 72.3.199.7.

[10:14:02.0000] <AryehGregor>
The one with 15 o's is the only one that points to 74.208.14.77.

[10:15:00.0000] <AryehGregor>
Actually, there are only six IPs that have more than one variant pointing to them.

[10:15:01.0000] <TabAtkins>
Hahaha, now I see what you're doing.

[10:15:02.0000] <AryehGregor>
Okay, I really need to do work instead of wasting time.

[10:15:03.0000] <AryehGregor>
Did you run the commands?

[10:15:04.0000] <TabAtkins>
Yes.

[10:15:05.0000] <TabAtkins>
And I was confused by the list of ips.

[10:15:06.0000] <AryehGregor>
It only goes up to 62, apparently no part of the domain name can be longer than 63 bytes.

[10:15:07.0000] <TabAtkins>
Then I man'd some of the commands and figured it out.

[10:15:08.0000] <AryehGregor>
That command is a test of your Unix command line knowledge.

[10:16:00.0000] <AryehGregor>
/me doesn't know if there's any easier way to repeat a string than yes piped to tr and head

[10:20:00.0000] <Ms2ger>
python "print i*\"o"

[10:22:00.0000] <TabAtkins>
Doesn't appear to be on my machine, but "string" has the ability to do that easily.

[10:27:00.0000] <swarren08>
Which browser right now has the most support for HTML5?

[10:27:01.0000] <swarren08>
everything i see is chrome beta 13 and 14

[10:28:00.0000] <smaug____>
swarren08: define HTML5 ;)

[10:28:01.0000] <swarren08>
lol

[10:28:02.0000] <swarren08>
good point

[10:28:03.0000] <smaug____>
and define "support"

[10:29:00.0000] <smaug____>
is support something that browser report to web page, or is it something actually usable

[10:29:01.0000] <swarren08>
Support=  me using html5 and the browser showing it

[10:29:02.0000] <Ms2ger>
But the answer is obviously Firefox

[10:29:03.0000] <smaug____>
:p

[10:29:04.0000] <swarren08>
i use firefox nightly

[10:29:05.0000] <swarren08>
and chrome b13

[10:30:00.0000] <swarren08>
i have not tried internet explorer since ie7

[10:30:01.0000] <gnarf>
b13? isnt 13 stable now?

[10:30:02.0000] <smaug____>
I would guess Firefox,Chrome and Opera are pretty much in the same level

[10:30:03.0000] <smaug____>
they just happen to "support" different things

[10:30:04.0000] <swarren08>
not to my knowledge

[10:31:00.0000] <swarren08>
i just download the beta via google

[10:31:01.0000] <swarren08>
12.0.725.0 beta-m

[10:31:02.0000] <swarren08>
and got that

[10:31:03.0000] <gnarf>
chrome stable: http://cl.ly/1N261S430A403g172M0g

[10:31:04.0000] <gesa>
http://caniuse.com

[10:32:00.0000] <swarren08>
ah

[10:32:01.0000] <swarren08>
correct it is stable

[10:32:02.0000] <swarren08>
i just restarted chrome again

[10:32:03.0000] <swarren08>
and am currently sitting at 14.0.835.35 beta-m

[10:32:04.0000] <swarren08>
/me wishes he could use cloudapp

[10:33:00.0000] <timeless>
karlcow: the list numeration thing is impressively depressing

[10:34:00.0000] <timeless>
i think the way to do start right for the other case is probably: <ol><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li class="hidden"><li>This is a continued list

[10:37:00.0000] <swarren08>
this version of chrome doesnt support Microdata

[10:37:01.0000] <timeless>
fwiw, ie10preview-whatever is pretty good

[10:37:02.0000] <timeless>
if you have windows, you really should grab it

[10:37:03.0000] <timeless>
they're actually implementing various proposals and giving constructive feedback

[10:37:04.0000] <timeless>
(they really aren't the enemy you remember)

[10:38:00.0000] <swarren08>
i remember, i just never liked ie

[10:38:01.0000] <TabAtkins>
Just wait until Windows 8.  I'm betting on the enemy I remember returning.

[10:38:02.0000] <swarren08>
to be honest, i forgot why...

[10:38:03.0000] <timeless>
what do you remember the enemy doing?

[10:38:04.0000] <timeless>
fwiw, ie has gotten a bunch of things right

[10:39:00.0000] <timeless>
among other things, their developer tools (js debugger/profiler) are better than many others

[10:39:01.0000] <timeless>
they also automatically profile addons and tell end users when there are slow ones

[10:39:02.0000] <TabAtkins>
Embracing a platform, extending it with proprietary extensions, releasing dev tools that make it really easy to use said extensions, then taking the platform the way they want now that they ahve a captive audience.

[10:39:03.0000] <timeless>
and ie doesn't automatically enable addons anymore

[10:39:04.0000] <timeless>
you're told when new addons are added and you get to *choose* to enable them

[10:39:05.0000] <TabAtkins>
Basically I'm not afraid of IE.  I'm afraid of Windows.

[10:39:06.0000] <timeless>
so far i haven't seen any addons violent/evil enough to violate that

[10:40:00.0000] <AryehGregor>
swarren08, Firefox/Chrome/Opera are all pretty comparable in standards support.  Firefox/Chrome somewhat more, maybe.  Safari is like Chrome except it lags behind due to release schedule.  IE is worse, but it's rapidly closing the gap with IE9 and now IE10.

[10:40:01.0000] <timeless>
(my bet is microsoft would crack down on such things)

[10:40:02.0000] <AryehGregor>
Yeah, IE9's addon handling is awesome.

[10:40:03.0000] <timeless>
/me nods

[10:40:04.0000] <AryehGregor>
"Do you want to enable Piece-of-Garbage Add-On by Some Random Company You've Heard of But Never Authorized to Install Add-Ons, Inc.?"  "No"

[10:40:05.0000] <timeless>
"and thanks for asking me"

[10:41:00.0000] <timeless>
which reminds me, i need to file a bug, our internal addon doesn't list a vendor

[10:41:01.0000] <gesa>
Or "This add-on created by M$ is doubling launch time. Do you want to disable it?"

[10:41:02.0000] <swarren08>
Like i said, i havent tried ie since ie7... and now that we are on ie9 (and beta 10) i figure ill give it a shot

[10:41:03.0000] <AryehGregor>
"Did you realize that all these stupid add-ons that have enabled themselves are causing the browser to take an extra 4.75 s to start up?"  "Good grief, no, disable them all"

[10:41:04.0000] <AryehGregor>
Firefox should get with the picture on that score.

[10:41:05.0000] <swarren08>
yea

[10:42:00.0000] <timeless>
oh, right, they also have a genuine "disable all" button

[10:42:01.0000] <timeless>
Multiple add-ons are selected

[10:42:02.0000] <AryehGregor>
Add-ons are a huge performance drain in some cases.  Which is why when I used Firefox, I eventually wound up with a grand total of like three extensions, and I was suspicious of those.

[10:42:03.0000] <timeless>
You have selected multiple add-ons in the list above. You can enable or disable all selected add-ons by clicking the appropriate button below.

[10:42:04.0000] <AryehGregor>
Firebug definitely killed performance in some cases, although it's probably gotten better.

[10:43:00.0000] <timeless>
i'm using 8, i don't have firebug

[10:43:01.0000] <AryehGregor>
At one point I actually saw a notification in Gmail saying "Hey, we noticed you're using Firebug, and our statistics show this murders page responsiveness.  Disable it when you're not using it."

[10:43:02.0000] <swarren08>
Firefox use to be pretty fast

[10:43:03.0000] <timeless>
AryehGregor: yeah, i remember that

[10:43:04.0000] <AryehGregor>
Sadly, Firebug is still essential for me on Firefox.

[10:43:05.0000] <timeless>
+1 to gmail team for doing that

[10:43:06.0000] <swarren08>
thats pretty cool

[10:43:07.0000] <AryehGregor>
I think Firebug has gotten better, though, like not enabling the network tab unless you're viewing it.

[10:43:08.0000] <TabAtkins>
AryehGregor: I had like 30 FF extensions.  I just didn't shut it down often, and when I did took it as an opportunity to make myself a snack.

[10:43:09.0000] <timeless>
heh

[10:43:10.0000] <AryehGregor>
TabAtkins, they can also slow down page load.

[10:43:11.0000] <timeless>
/me should get lunch

[10:43:12.0000] <AryehGregor>
And new tab creation time.

[10:44:00.0000] <TabAtkins>
AryehGregor: Apparently not enough for me to have cared at the time.

[10:44:01.0000] <smaug____>
swarren08: Firefox is a lot faster than ever ;)

[10:44:02.0000] <timeless>
AryehGregor: note that IE actually tells you if  an add-on slows down start up or page load

[10:44:03.0000] <timeless>
w/ timings!

[10:44:04.0000] <AryehGregor>
TabAtkins, or you just didn't realize how much slower it was making your browsing.

[10:44:05.0000] <AryehGregor>
Maybe that's part of why Chrome seemed so much faster when you tried it.  :)

[10:44:06.0000] <AryehGregor>
timeless, yes, it's absolutely awesome.

[10:44:07.0000] <AryehGregor>
Firefox needs to just clone the whole feature.

[10:44:08.0000] <smaug____>
AryehGregor: Btw, Skype extension has been the worst extension ever, I think

[10:44:09.0000] <smaug____>
perf-vise

[10:44:10.0000] <smaug____>
it slowed down some DOM operations like 300x

[10:44:11.0000] <swarren08>
yes

[10:44:12.0000] <timeless>
smaug____: well, scanning the entire page for phone numbers...

[10:44:13.0000] <timeless>
constantly

[10:45:00.0000] <swarren08>
thank you to chrome

[10:45:01.0000] <swarren08>
it automatically shuts off my skype extension

[10:45:02.0000] <AryehGregor>
FWIW, I did notice a huge responsiveness difference between Firefox 3.6 (with *no* extensions) and Chrome on Linux when I tried them.  It got a lot better with Firefox 4.

[10:45:03.0000] <TabAtkins>
AryehGregor: I switched to Chrome for the UI, not the speed.

[10:45:04.0000] <AryehGregor>
But you really can't beat Chrome's "nothing ever freezes the UI even briefly".

[10:45:05.0000] <swarren08>
The UI is very impressive

[10:45:06.0000] <timeless>
AryehGregor: have you had the chrome in IE freeze?

[10:45:07.0000] <TabAtkins>
That's not 100% true, but yeah, it's nice.

[10:45:08.0000] <smaug____>
AryehGregor: have you tried printing with Chrome ;)

[10:45:09.0000] <timeless>
i don't think i've seen it freeze in ie9/10 except when they crash

[10:45:10.0000] <AryehGregor>
Versus Firefox's "mostly the UI is responsive but the whole browser can hang sometimes due to long-running script etc.".

[10:45:11.0000] <smaug____>
that is the way to freeze Chrome

[10:45:12.0000] <timeless>
which doesn't count (and yes, i've crashed chrome)

[10:46:00.0000] <swarren08>
i dont think ive creshed chrome

[10:46:01.0000] <AryehGregor>
smaug____, I don't have a printer.  Do people still use paper these days?  I haven't noticed.

[10:46:02.0000] <swarren08>
i crash firefox and nightly constantly

[10:46:03.0000] <timeless>
/me prints

[10:46:04.0000] <AryehGregor>
(actually, I just use my parents' printer if I need one, and Chrome prints fine in my experience)

[10:46:05.0000] <smaug____>
printing has been the way to hang and crash chrome

[10:46:06.0000] <timeless>
swarren08: have you used about:crashes and ensured there are bugs filed?

[10:46:07.0000] <swarren08>
yep

[10:46:08.0000] <AryehGregor>
I haven't seen a whole-browser crash in Chrome for at least a year or two.

[10:46:09.0000] <swarren08>
i always do

[10:46:10.0000] <swarren08>
and i always put my email in

[10:47:00.0000] <AryehGregor>
I've seen plenty of individual tab crashes.

[10:47:01.0000] <AryehGregor>
Sometimes reproducible.

[10:47:02.0000] <dglazkov>
oh cool, a browser bash

[10:47:03.0000] <AryehGregor>
But they're really no big deal, just refresh.

[10:47:04.0000] <TabAtkins>
I suppose I havent' seen a full renderer crash since the %: or whatever bug.

[10:47:05.0000] <dglazkov>
I like curl myself

[10:47:06.0000] <AryehGregor>
(I especially get crashes in Web Inspector)

[10:47:07.0000] <TabAtkins>
dglazkov: I use a shell script that uses curl and smtp so I can read the web in pine.

[10:47:08.0000] <AryehGregor>
dglazkov, curl doesn't give you enough control.  I think telnet is the only tenable option.

[10:47:09.0000] <Ms2ger>
dglazkov, wget, dammit

[10:48:00.0000] <AryehGregor>
Although it's tricky to use for HTTPS.

[10:48:01.0000] <AryehGregor>
You have to be really good at mental arithmetic.

[10:48:02.0000] <dglazkov>
AryehGregor: what's HTTPS?

[10:49:00.0000] <AryehGregor>
dglazkov, that thing that adds extra round-trips to the first page load in exchange for people not being able to play devastating practical jokes on you with Firesheep.

[10:49:01.0000] <dglazkov>
seems like a lot of pain to avoid being pranked.

[10:49:02.0000] <Philip`>
It's the thing that triggers all those warnings about invalid certficates that you click "accept" on

[10:49:03.0000] <swarren08>
sometimes id rather be rick roll'd

[10:49:04.0000] <AryehGregor>
I know, right?  But some people like it.

[10:50:00.0000] <AryehGregor>
Philip`, or that you click through four different menus that try to confuse and dissuade you before the thing will actually let you view the site, in Firefox's case.

[10:50:01.0000] <AryehGregor>
But in Chrome I've seen it not give any option to continue at all.

[10:50:02.0000] <AryehGregor>
That is absolutely unacceptable.

[10:51:00.0000] <timeless>
oh, sob

[10:51:01.0000] <AryehGregor>
It was some MS site that had a revoked cert for some reason, but I didn't care in the slightest because the most an attacker could steal was my MS login, which I don't care about.

[10:51:02.0000] <timeless>
our IE toolbar hasn't been updated since 2008

[10:51:03.0000] <AryehGregor>
/me stabs "security" zealots

[10:52:00.0000] <swarren08>
wow 2008?

[10:52:01.0000] <timeless>
swarren08: i presume it "just works" and doesn't really need maintenance

[10:52:02.0000] <zewt>
yeah chrome's https handling is beyond unacceptably broken

[10:52:03.0000] <timeless>
it's an internal tool

[10:52:04.0000] <zewt>
last i saw, anyway

[10:52:05.0000] <swarren08>
sounds like a bunch of people i know

[10:53:00.0000] <timeless>
yeah, the webkit client i use here uses curl

[10:53:01.0000] <zewt>
firefox's is also unacceptably broken, but not to the degree of actually preventing you from doing things you need to do

[10:53:02.0000] <timeless>
and the errors i get for web sites are terrible

[10:53:03.0000] <zewt>
firefox only goes to the point of ensuring every firefox user is trained to click through every security dialog without reading it

[10:53:04.0000] <TabAtkins>
AryehGregor: Argh, worst is the OMGWTFBBQ about self-signed certs.  Not even a whisper of complaint when you visit a page that does login over HTTP POST, but if you *dare* to try to secure the connection without going through the registrar mafia?  RED PAGE.

[10:53:05.0000] <timeless>
AryehGregor: there's sconnect or something for ssl

[10:54:00.0000] <timeless>
i've used that as a telnet replacement for https

[10:54:01.0000] <zewt>
tab: yeah, i've been ranting for a while about the ridiculousness of browsers training people to think of unsigned encryption as *less* secure than plaintext

[10:54:02.0000] <zewt>
i understand the rationale, but the end result is no less absurd for it

[10:54:03.0000] <timeless>
zewt: well

[10:55:00.0000] <TabAtkins>
I don't even understand the rationale.

[10:55:01.0000] <timeless>
the rationale is:

[10:55:02.0000] <timeless>
if you normally don't get an error

[10:55:03.0000] <TabAtkins>
You don't trust the endpoint, but you trust the connection.

[10:55:04.0000] <timeless>
then when you do get an error, you're probably being attacked

[10:55:05.0000] <zewt>
the rationale is that you don't want people loading their https://bank.com and having an attacker present the site with a self-signed cert

[10:55:06.0000] <TabAtkins>
That's better than trusting neither.

[10:55:07.0000] <timeless>
whereas if you are visiting a site that never pretends to be secure

[10:55:08.0000] <timeless>
then you're supposed to know that it isn't necessarily trustworthy

[10:55:09.0000] <timeless>
but you also know that there's no reason to consider today any more or less different from yesterday

[10:56:00.0000] <timeless>
the big problem is...

[10:56:01.0000] <timeless>
banks offer http://bank.com

[10:56:02.0000] <timeless>
and often have login pages on that page

[10:56:03.0000] <zewt>
the solution, IMO, is/was/should have been to have an shttp:// protocol, which is exactly identical to https but without the certificate check, so loading https URLs can't be compromised in that way, while still allowing people to use encryption without a cert

[10:56:04.0000] <timeless>
which teach users to enter their credentials *there*!

[10:57:00.0000] <timeless>
zewt: for what purpose?

[10:57:01.0000] <zewt>
i've never seen a bank with a login not on https (of course I only use my bank), but my electric or gas company (forget which) has a login on http (with some dumb flash hack to handle the actual encryption)

[10:57:02.0000] <AryehGregor>
TabAtkins, https:// with self-signed cert is, in general, worse than http://.

[10:57:03.0000] <zewt>
timeless: huh? for the purpose of encrypting :)

[10:57:04.0000] <timeless>
zewt: try http://www.discovercard.com/

[10:57:05.0000] <AryehGregor>
/me digs up rationale

[10:57:06.0000] <timeless>
there's a login form on that page

[10:57:07.0000] <timeless>
i've actually done surveys of banks

[10:57:08.0000] <timeless>
and it varies by size and region

[10:57:09.0000] <timeless>
but there are definitely plenty of banks w/ login forms on their front page (which is http)

[10:57:10.0000] <AryehGregor>
TabAtkins, http://lwn.net/Articles/413600/

[10:58:00.0000] <timeless>
and please note that having a phone number or hours/addresses on http is also bad

[10:58:01.0000] <zewt>
if there was an http* protocol meant for encryption-without-certs, every webserver in the world would ship with a default configuration allowing encryption, and the level of security in the world would be a notch higher

[10:58:02.0000] <swarren08>
wow...  people will do anything to hack you

[10:58:03.0000] <timeless>
if i'm using an internet cafe in a foreign country

[10:58:04.0000] <timeless>
i can easily be tricked into going to a fake bank or fake atm

[10:58:05.0000] <timeless>
i wouldn't know the difference..

[10:58:06.0000] <AryehGregor>
/me <3 Google for giving his post as the first result for "lwn.net Simetrical self-signed"

[10:58:07.0000] <zewt>
it would not be a *lot* higher, since it obviously wouldn't protect against MITM, but the world would be substantially less susceptible to passive sniffing

[10:58:08.0000] <TabAtkins>
AryehGregor: Yes, the confusion angle is the only thing.  Presenting the same UI for a self-signed and an EV-signed would be bad.

[10:58:09.0000] <AryehGregor>
zewt, tcpcrypt.org is a very neat technology.

[10:59:00.0000] <AryehGregor>
I hope it takes off.

[10:59:01.0000] <timeless>
passive sniffing is kinda useless as risks go

[10:59:02.0000] <timeless>
if i'm in an internet cafe

[10:59:03.0000] <swarren08>
A good example of what you are talking about is Steam

[10:59:04.0000] <zewt>
spdy is also always encrypted, signed or no

[10:59:05.0000] <zewt>
iirc

[10:59:06.0000] <timeless>
there's really no way to protect me from a MITM

[10:59:07.0000] <AryehGregor>
tcpcrypt covers all TCP, though.

[10:59:08.0000] <timeless>
and that MITM becomes equivalent to passive snooping

[10:59:09.0000] <AryehGregor>
SPDY is always encrypted because it has to tunnel through HTTP proxies that try to meddle with regular HTTP.

[10:59:10.0000] <zewt>
mitm is ... the opposite of passive

[11:00:00.0000] <AryehGregor>
No, MITM can include passive as well as active.

[11:00:01.0000] <timeless>
you might as well get people to do the right thing and have good certs

[11:00:02.0000] <swarren08>
People go thru and do a web site dump of the Official site

[11:00:03.0000] <AryehGregor>
Generally any time you can do passive you can do active, but active is often harder to pull off and/or carries more risk of detection.

[11:00:04.0000] <swarren08>
and adjust it to the site they own, make own certs and send the links to people

[11:00:05.0000] <timeless>
/me notes that there are deployed attacks which actually do this

[11:00:06.0000] <TabAtkins>
AryehGregor: Thanks for reminding me about tcpcrypt.  I meant to install it last time I saw it.

[11:01:00.0000] <timeless>
they've even been described on krebsonsecurity.com

[11:01:01.0000] <swarren08>
they are able to take over what they want and also capture the accounts of people

[11:01:02.0000] <zewt>
active is by nature riskier, which is a deterrant by itself

[11:01:03.0000] <AryehGregor>
tcpcrypt is clever that way: you don't know whether there will be authentication until the encryption is finished.  So a MITM has to decide whether to intercept the connection before knowing if they'll be detected.  If you have auth layered on top of tcpcrypt or similar, the attacker will intercept it fine, but then the authentication will fail and the attacker will be detected.

[11:01:04.0000] <AryehGregor>
In principle, if tcpcrypt were deployed, you could do things like expose the client's session id to JavaScript, say.

[11:02:00.0000] <timeless>
err

[11:02:01.0000] <timeless>
can't the server expose that already?

[11:02:02.0000] <AryehGregor>
So the server could send some code with the server's session id somewhere and check it against the client id.  The attacker could theoretically rewrite the code to prevent that, but not in general -- it's the halting problem.

[11:02:03.0000] <timeless>
oh

[11:02:04.0000] <timeless>
well

[11:02:05.0000] <timeless>
no

[11:03:00.0000] <AryehGregor>
So the attacker could have special hacks to rewrite the checks for known apps, but some random webpage that does it in some slightly different way would still be able to throw up a scary error.

[11:03:01.0000] <timeless>
MITM could pretty trivially include code which tries to capture access to a property

[11:03:02.0000] <zewt>
AryehGregor: don't know how you mean that MITM can be done passively; the definition of MITM is active, sending packets to one or both parties

[11:03:03.0000] <AryehGregor>
timeless, you could make the property a non-replaceable global so that nothing can capture access to it.

[11:03:04.0000] <zewt>
(most often both, hence the term)

[11:03:05.0000] <AryehGregor>
zewt, it can include interception as well, AFAIK.  MITMs can merely eavesdrop.

[11:04:00.0000] <AryehGregor>
Hmm, I see.

[11:04:01.0000] <zewt>
eavesdropping isn't MITM, it's sniffing

[11:04:02.0000] <AryehGregor>
Well, an MITM is perhaps always active to some extent, but not necessarily in a noticeable fashion.

[11:04:03.0000] <timeless>
AryehGregor: at that point, the attacker will just rewrite all streams looking for the keyword, replace `eval`, and rewrite any instance of `window` to an object which does shadowing

[11:04:04.0000] <AryehGregor>
A compromised router is a MITM.

[11:04:05.0000] <zewt>
depends on the particular attack

[11:04:06.0000] <AryehGregor>
But it might just be sniffing traffic, not doing anything discernibly wrong.

[11:04:07.0000] <timeless>
i wonder if someone has a toolkit that does that

[11:04:08.0000] <zewt>
a compromised router is a MITM if you're using it to sit between and manipulate the two endpoints

[11:05:00.0000] <zewt>
ngrep on a compromised router is not MITM

[11:05:01.0000] <AryehGregor>
timeless, hmm, I guess.  You could make it a magic keyword, but then they could sniff for a keyword.

[11:05:02.0000] <AryehGregor>
Anyway, it would be a way to catch attackers who aren't careful.

[11:05:03.0000] <timeless>
yep

[11:05:04.0000] <timeless>
AryehGregor: the problem is...

[11:05:05.0000] <timeless>
attackers will share toolkits

[11:05:06.0000] <AryehGregor>
zewt, definitionally, I disagree, but it's semantics, so whatever.

[11:05:07.0000] <timeless>
which means the X months it takes to deploy a security feature

[11:05:08.0000] <AryehGregor>
timeless, sophisticated ones will.  Script kiddies will use whatever they happen to get off the shelf, which may or may not be competently written depending on their connections and price range.

[11:05:09.0000] <zewt>
mitm is a specific term referring to a specific class of attacks

[11:06:00.0000] <timeless>
will be defeated in X/12 months and then shared

[11:06:01.0000] <timeless>
it will appear in the script kiddy toolkits in 3/7 * X months

[11:06:02.0000] <AryehGregor>
Yeah, it's just a speed bump.  But it will catch some attacks.

[11:06:03.0000] <timeless>
it really isn't a useful speed bump

[11:06:04.0000] <timeless>
and it's somewhat expensive to implement

[11:06:05.0000] <AryehGregor>
It's basically free to implement, given that tcpcrypt is already being used.

[11:06:06.0000] <timeless>
cost benefit there is terrible

[11:07:00.0000] <AryehGregor>
Which is a good idea anyway to stop sniffers.

[11:08:00.0000] <timeless>
what protects tcpcrypt from a mitm that always claims not to support tcpcrypt?

[11:08:01.0000] <timeless>
(standard downgrade attack)

[11:09:00.0000] <timeless>
/me hunts lunch

[11:10:00.0000] <AryehGregor>
timeless, such a MITM could also just intercept the connection directly.  So there's no added threat surface.

[11:11:00.0000] <AryehGregor>
Either way, it forces them to become an active attacker -- they can't just sniff.

[11:11:01.0000] <AryehGregor>
Also, tcpcrypt exposes a very nice, useful primitive (session id) that you can trivially layer authentication on top of.

[11:11:02.0000] <AryehGregor>
So it would allow a lot of code to be unified that currently has to be redone for every different application-level protocol.

[11:11:03.0000] <Ms2ger>
Oh hey, go Microsoft

[11:12:00.0000] <AryehGregor>
?

[11:13:00.0000] <Ms2ger>
They like DOM Views

[11:13:01.0000] <AryehGregor>
So that was sarcastic?

[11:15:00.0000] <Ms2ger>
Apologies

[11:15:01.0000] <Ms2ger>
</sarcasm>

[11:15:02.0000] <AryehGregor>
/me is never quite sure without context, since MS does occasionally do things surprisingly right

[11:16:00.0000] <AryehGregor>
They're okay with a notice, though, so I think that's fine.

[11:16:01.0000] <AryehGregor>
Actually rescinding is going to be a waste of time, and it won't happen if the replacement isn't also a REC.

[11:18:00.0000] <TabAtkins>
I just notice how much they talked about "stability" in that email, and not "correctness".

[11:18:01.0000] <Ms2ger>
Well duh

[11:18:02.0000] <AryehGregor>
Yes, because that's what their corporate customers care about.

[11:18:03.0000] <AryehGregor>
Standards compliance is an item on a checklist, not something they directly care about for any practical reason.

[11:19:00.0000] <AryehGregor>
MS cares about the practical reasons, but most of their customers don't.

[11:19:01.0000] <AryehGregor>
They're the only major browser vendor to actually deal with large corporations to any meaningful extent, remember.

[11:19:02.0000] <AryehGregor>
(as users of their browsers, that is)

[11:37:00.0000] <gsnedders>
AryehGregor: Well, we ship to customers who care about these sorts of things

[11:37:01.0000] <swarren08>
Who is Mr. Lastweek?

[11:37:02.0000] <AryehGregor>
gsnedders, yeah, okay, but you clearly aren't as concerned with them as MS.

[11:37:03.0000] <Ms2ger>
MikeSmith

[11:37:04.0000] <Ms2ger>
Or so I hear

[11:38:00.0000] <gsnedders>
swarren08: That was a big question a few years back. Still unknown, AFIAK.

[11:38:01.0000] <swarren08>
ah

[11:38:02.0000] <swarren08>
im reading a book

[11:38:03.0000] <swarren08>
and its in here

[11:38:04.0000] <gsnedders>
AryehGregor: Well, yeah. Most of our customers care more about CE-HTML and the like than HTML.

[11:38:05.0000] <Ms2ger>
Which book?

[11:39:00.0000] <swarren08>
Addison Wesley HTML5 Guidelines for Web Developers

[11:39:01.0000] <swarren08>
An imaginary character,

[11:39:02.0000] <swarren08>
Mr. LastWeek, comments on the events with sometimes hefty blog entries at

[11:39:03.0000] <swarren08>
http://lastweekinhtml5.blogspot.com in reaction to the publicly accessible IRC

[11:39:04.0000] <swarren08>
protocols at http://krijnhoetmer.nl/irc-logs

[11:40:00.0000] <gsnedders>
AryehGregor: And from a more sane POV, things like HbbTV.

[11:40:01.0000] <Ms2ger>
The IRC protocols?

[11:40:02.0000] <swarren08>
lol thats what the book says

[11:42:00.0000] <swarren08>
Apparently im oblivious to everything... or just didnt know but they are actually working on xhtml2?

[11:43:00.0000] <gsnedders>
swarren08: They were. It never had much interest from browser vendors, and totally stopped a couple of years back.

[11:43:01.0000] <swarren08>
ah

[11:43:02.0000] <Ms2ger>
Well, the WG was disbanded

[11:43:03.0000] <Ms2ger>
And then continued to publish specs for another year

[11:43:04.0000] <swarren08>
i looked up the guy who was doing it and it had info on his site about xhtml

[11:43:05.0000] <swarren08>
xhtml2*

[11:44:00.0000] <gesa>
xhtml made us all write cleaner code. rip.

[11:45:00.0000] <swarren08>
did you not like to write cleaner code?

[11:45:01.0000] <gesa>
i LOVED writing cleaner code. I'll always write cleaner code because xhtml existed.

[11:46:00.0000] <swarren08>
i make my coding look nice

[11:46:01.0000] <gesa>
none of this non-quoted attribute values, unclosed tags nonsense

[11:46:02.0000] <swarren08>
i always close my tags

[11:46:03.0000] <TabAtkins>
Unquoted attributes are the major reason I write my SVG in HTML.  ^_^

[11:47:00.0000] <Ms2ger>
Ugh

[11:47:01.0000] <gesa>
me too. but future web developers have no reason to anymore.

[11:47:02.0000] <Ms2ger>
Next you're going to claim you write <br/>

[11:47:03.0000] <swarren08>
i do actually

[11:47:04.0000] <gesa>
... me too.

[11:47:05.0000] <Ms2ger>
And <script/> as well?

[11:48:00.0000] <gesa>
Though I find I don't use <br /> very often anymore

[11:48:01.0000] <gesa>
no way.

[11:48:02.0000] <gesa>
<script></script>

[11:48:03.0000] <gesa>
<script/> breaks some (most? all?) browsers.

[11:48:04.0000] <Ms2ger>
All

[11:48:05.0000] <swarren08>
lol

[11:48:06.0000] <Philip`>
Ms2ger: Surely not <br/>, it has to be <br /> for compatibility

[11:49:00.0000] <Ms2ger>
/me throws a cake at Philip` 

[11:49:01.0000] <Ms2ger>
And tell me my test updates are correct

[11:49:02.0000] <gsnedders>
Ms2ger: But don't you care about NN! You insensitive sod!

[11:49:03.0000] <AryehGregor>
gesa, <script/> breaks all browsers, it's parsed the same as <script>.

[11:49:04.0000] <AryehGregor>
Except if served as XML.

[11:49:05.0000] <Philip`>
(Also, never forget to comment out your script contents)

[11:50:00.0000] <gesa>
Yeah. Learned that the hard way in a night of obsessive over the top code cleanup

[11:51:00.0000] <GlitchMr>
Browsers don't understand /> in SGML mode. Only while parsing documents as XML. Otherwise, they consider that / is just some random character to ignore.

[11:51:01.0000] <TabAtkins>
s/SGML/HTML/

[11:52:00.0000] <Ms2ger>
Nah

[11:52:01.0000] <TabAtkins>
In SGML mode, the / closes the tag, and the > is part of the text content.  Luckily, no one's been an SGML browser for some time.

[11:52:02.0000] <GlitchMr>
Oh, right. Browsers fail at SHML. They don't know that <br/> in actually <br> and after that &gt; character...

[11:52:03.0000] <Ms2ger>
They acknowledge the trailing solidus

[11:54:00.0000] <Philip`>
TabAtkins: No one except the W3C validator, which will happily parse your document completely differently to any real browser before validating the result

[11:54:01.0000] <GlitchMr>
I remember <!-- -- >The comment is closed :). Browsers implemented it for some time just because of ACID2...

[11:55:00.0000] <GlitchMr>
W3C Validator is lame for me anyways. It doesn't take into account real problems and it blatantly accepts markup like <h1><ins><h2><del><h3></h3></del></h2></ins></h1>.

[11:56:00.0000] <GlitchMr>
And it's not error in validator :P. HTML shouldn't be parsed using DTD.

[12:22:00.0000] <Hixie>
Philip`: in case it comes up again, anne's he one running the splitter for complete.html

[12:40:00.0000] <swarren08>
hmmm i apparently could buy an HTML5 tshirt....

[12:41:00.0000] <swarren08>
How much does the css we use today differ from CSS3

[12:42:00.0000] <gsnedders>
/me facepalms at the sizes of the HTML5 t-shirts — the XL "body width" is the same size as my entire waist. Woops.

[12:45:00.0000] <swarren08>
im glad im not the only one gsnedders

[12:45:01.0000] <gsnedders>
"Profits from the sale of every HTML5 shirt go directly to the development of HTML5 test suite." — uh, who?

[12:45:02.0000] <gsnedders>
(And to those worried, don't worry. I've eaten almost 250g of chocolate this afternoon)

[12:46:00.0000] <gesa>
gsnedders: Women's XL width is way larger than my entire waist. Humans are varied, that's all.

[12:46:01.0000] <swarren08>
it goes to W3C

[12:47:00.0000] <gsnedders>
swarren08: No W3C employee has submitted a single test to the testsuite…

[12:47:01.0000] <swarren08>
lol

[12:47:02.0000] <swarren08>
it was a joke

[12:47:03.0000] <swarren08>
relax

[12:48:00.0000] <gsnedders>
gesa: And I thought my flatmate was thin…

[12:50:00.0000] <gesa>
gsnedders: Not thin. Just petite.

[12:50:01.0000] <annevk>
I wonder, did anyone miss the WHATWG Weekly?

[12:50:02.0000] <swarren08>
/me slowly raises hand

[12:50:03.0000] <swarren08>
ive been stuck in the hospital since last sat for a family member, so ive missed a bunch

[12:51:00.0000] <annevk>
heh

[12:51:01.0000] <gsnedders>
gesa: Well, you can still be thin in generic terms while being petite, and hence not thin relative to the rest of you.

[12:51:02.0000] <annevk>
I mean missed as in "omg why was it not written and published last Monday"

[12:51:03.0000] <gesa>
annevk: in that case, yes.

[12:51:04.0000] <annevk>
swarren08, my best to your family member

[12:52:00.0000] <gesa>
thought i was missing a bigger picture

[12:52:01.0000] <swarren08>
thanks, we hope she will awake soon

[12:52:02.0000] <gsnedders>
/me concludes a girls small HTML5 t-shirt would fit him in terms of width, though not be at all long enough

[12:53:00.0000] <Philip`>
/me guesses the measurements must be in inches, since if they were cm the XL wouldn't quite be wide enough for him

[12:53:01.0000] <swarren08>
i must agree with that one philip

[12:59:00.0000] <Ms2ger>
gsnedders, http://dvcs.w3.org/hg/html/rev/1a50586be65c

[12:59:01.0000] <gsnedders>
Ms2ger: Oh, my bad.

[12:59:02.0000] <Ms2ger>
That makes one :)

[14:36:00.0000] <Hixie>
is there really no spec out there that describes how to parse application/x-www-form-urlencoded data?

[14:37:00.0000] <Hixie>
hey we didn't register application/x-www-form-urlencoded either

[14:37:01.0000] <Hixie>
i guess we should do that huh

[14:38:00.0000] <gsnedders>
Hixie: Not that I could find.

[14:38:01.0000] <Hixie>
we suck

[14:39:00.0000] <TabAtkins>
Holy crap.

[14:39:01.0000] <TabAtkins>
I can't believe that these things go undefined like that.

[14:39:02.0000] <TabAtkins>
For decades.

[14:39:03.0000] <gsnedders>
How many impls are there, though, really?

[14:39:04.0000] <Hixie>
like six bazillion?

[14:40:00.0000] <TabAtkins>
Every HTTP server?

[14:40:01.0000] <Hixie>
and every cgi library

[14:40:02.0000] <gsnedders>
TabAtkins: That's still not a huge number.

[14:40:03.0000] <Hixie>
and every custom cgi script not based on a library

[14:40:04.0000] <gsnedders>
Hixie: Does each CGI library implement its own?

[14:40:05.0000] <TabAtkins>
More impls than there are browsers, by an order of magnitude or two.

[14:40:06.0000] <gsnedders>
TabAtkins: Well, yes.

[14:40:07.0000] <Hixie>
gsnedders: all the ones i've seen do

[14:40:08.0000] <Hixie>
i mean it's trivial to implement

[14:40:09.0000] <gsnedders>
Hixie: Heh.

[14:40:10.0000] <Hixie>
you just cut on &s and then cut on =s and then expand the %xxs

[14:40:11.0000] <gsnedders>
PHP does it all internally.

[14:40:12.0000] <Hixie>
and replace the +s

[14:41:00.0000] <Hixie>
and guess the encoding somehow

[14:41:01.0000] <Hixie>
by MAGIC

[14:41:02.0000] <gsnedders>
Hixie: We managed to do it wrong in one of our systems, which is why I realized there was no spec.

[14:41:03.0000] <Hixie>
heh

[14:41:04.0000] <gsnedders>
/me forgets what exactly the bug was

[14:41:05.0000] <TabAtkins>
We can just define the encoding, hopefully, and shut down one more place where the encoding wars hurt everyone.

[14:42:00.0000] <gsnedders>
TabAtkins: We can't. No way.

[14:42:01.0000] <TabAtkins>
Damn.

[14:42:02.0000] <Hixie>
yeah that's a lost battle

[14:42:03.0000] <Hixie>
_charset_ is the only way out of that one

[14:42:04.0000] <gsnedders>
TabAtkins: What encoding you want to use depends upon what you expect it will be, which is related to the encoding of the page it is sent frm.

[14:42:05.0000] <gsnedders>
*from

[14:43:00.0000] <TabAtkins>
;_;

[14:44:00.0000] <gsnedders>
Hixie: IIRC we were doing something like map(lambda x: x.split("=", 1), foo.split("&")) and then decoding

[14:44:01.0000] <gsnedders>
OH, no!

[14:44:02.0000] <Hixie>
that should work except for +s, no?

[14:44:03.0000] <gsnedders>
We did the decoding too soon.

[14:44:04.0000] <gsnedders>
That's what.

[14:44:05.0000] <Hixie>
ah, yeah

[14:44:06.0000] <TabAtkins>
Oh, heh.

[14:44:07.0000] <Hixie>
rookie mistake :-P

[14:45:00.0000] <TabAtkins>
Argh, figuring out what to do in mathematically degenerate cases is annoying.

[14:45:01.0000] <gsnedders>
Hixie: Sadly not in the code that was originally yours, so I can't blame you :P

[14:45:02.0000] <Hixie>
woot!

[14:46:00.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=12819 look right to you?

[14:46:01.0000] <Hixie>
wow i just mistook a chrome version number for an ip address

[14:46:02.0000] <Hixie>
that was confusing

[14:47:00.0000] <hober>
heh

[14:49:00.0000] <Hixie>
bbiab

[14:52:00.0000] <gsnedders>
Hixie: isindex just gets a single item?

[14:52:01.0000] <gsnedders>
Just the value?

[14:56:00.0000] <Ms2ger>
Hixie, http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A...%3Cscript%3E%0Aw%28%22width%22%20in%20document.createElement%28%22input%22%29%29%0A%3C%2Fscript%3E

[15:00:00.0000] <AryehGregor>
Does Gecko not fire keyup/keydown for things like Ctrl+B, or what?

[15:00:01.0000] <AryehGregor>
/me can't seem to get it to work

[15:01:00.0000] <smaug____>
works fine here

[15:01:01.0000] <AryehGregor>
/me tries again

[15:04:00.0000] <Hixie>
gsnedders: isindex is you skip the &-split step and the =-split step and just assign the whole thing to one 'value'

[15:05:00.0000] <AryehGregor>
I get a keydown for the ctrl, but not the b.

[15:05:01.0000] <Hixie>
Ms2ger: oh you mean the IDL attributes, ok

[15:05:02.0000] <gsnedders>
Hixie: Yeah, that's what I thought.

[15:05:03.0000] <Hixie>
Ms2ger: interesting

[15:05:04.0000] <Hixie>
Ms2ger: (your bug was too terse!)

[15:05:05.0000] <Ms2ger>
Oh, yes, should have said that

[15:06:00.0000] <Hixie>
gsnedders: what do you mean about non-ASCII bytes?

[15:07:00.0000] <gsnedders>
Hixie: If I have a FF byte, what do I do? What happens when I decode?

[15:07:01.0000] <gsnedders>
(pretend we're UTF-8 here)

[15:07:02.0000] <gsnedders>
Per HTML5 what is sent should be ASCII. What if it's not?

[15:08:00.0000] <AryehGregor>
Ah, I see.

[15:08:01.0000] <AryehGregor>
I was just getting confused.

[15:08:02.0000] <AryehGregor>
Oh, it's sending a capital B for keyCode instead of lowercase.

[15:08:03.0000] <AryehGregor>
Fun.

[15:08:04.0000] <Philip`>
Some libraries (e.g. CGI in Perl) split on ';' as well as '&' - should specs require/allow that?

[15:09:00.0000] <AryehGregor>
Don't you love interop?

[15:09:01.0000] <AryehGregor>
/me grumbles

[15:09:02.0000] <smaug____>
AryehGregor: what should it do then?

[15:10:00.0000] <smaug____>
other browsers give the same keyCode

[15:10:01.0000] <TabAtkins>
Philip`: That's part of the original definition of the encoding.  Several libraries expose it as an option.

[15:10:02.0000] <smaug____>
er, charcode

[15:11:00.0000] <AryehGregor>
/me scratches head

[15:11:01.0000] <AryehGregor>
Now I'm confused.

[15:11:02.0000] <smaug____>
keypress gives charcode, keydown/up  give keycode

[15:11:03.0000] <smaug____>
Opera seems to have some strange handling

[15:11:04.0000] <AryehGregor>
Because I could swear WebKit was giving the keycode for lowercase B before (97?), not uppercase.

[15:12:00.0000] <AryehGregor>
But my eyes must have been deceiving me.

[15:12:01.0000] <AryehGregor>
Oh well.

[15:12:02.0000] <smaug____>
keycode 66 for down/up

[15:12:03.0000] <smaug____>
and charcode 98 for press

[15:13:00.0000] <AryehGregor>
Right, I'm looking only at down/up.

[15:13:01.0000] <Ms2ger>
Hixie, should I reopen or will you?

[15:15:00.0000] <Hixie>
Ms2ger: please do

[15:15:01.0000] <Hixie>
gsnedders: ah, error handling

[15:15:02.0000] <Hixie>
gsnedders: yeah

[15:15:03.0000] <Hixie>
gsnedders: dunno

[15:16:00.0000] <Hixie>
gsnedders: interop isn't a big deal on that matter so maybe i just don't spec it

[15:17:00.0000] <gsnedders>
Hixie: Probably good enough to just leave it undefined, on the whole

[15:44:00.0000] <gsnedders>
Hixie: Pff, r6434! I mean, there's already a SHOULD requirement to solve the halting problem!

[15:47:00.0000] <Hixie>
yeah but that one only applies to validators :-P

[15:47:01.0000] <Hixie>
heycam|away: dude, stop changing the terms you use in webidl :-P

[15:50:00.0000] <Hixie>
is there some list i can subscribe to that e-mails me all changes to webidl?

[15:50:01.0000] <Hixie>
and dom core for that matter?


2011-08-13
[17:06:00.0000] <Philip`>
http://chrome.blogspot.com/2011/08/building-better-web-apps-with-new.html - is that NaCl and not PNaCl, so they're not even pretending to be CPU-architecture-independent?

[17:14:00.0000] <gsnedders>
Philip`: NaCL, AIUI

[17:15:00.0000] <Hixie>
seems that way. "Native Client modules are operating system independent, not (yet) processor independent. Therefore, you must compile separate versions of the Native Client module for x86-32-bit, x86-64-bit, and other instruction sets. The manifest file, specified in the src attribute of the <embed> tag, specifies which version of the Native Client module to load, depending on the end-user's processor." - ...

[17:15:01.0000] <Hixie>
... http://code.google.com/chrome/nativeclient/docs/technical_overview.html

[17:15:02.0000] <Hixie>
/me raises an eyebrow at the irssi module he apparently set up

[17:15:03.0000] <Hixie>
neat

[17:16:00.0000] <wilhelm>
splitlong ftw.

[17:58:00.0000] <gsnedders>
http://googleonlinesecurity.blogspot.com/2011/08/fuzzing-at-scale.html?m=1

[20:20:00.0000] <heycam>
Hixie, sorry ;)

[20:21:00.0000] <heycam>
Hixie, you could subscribe to member-webapi-cvs or whatever it's called

[20:33:00.0000] <Hixie>
hm, i thought i had

[20:34:00.0000] <heycam>
ah so it never got added to the spec as [NonConfigurable]

[20:34:01.0000] <Hixie>
oh, right, i didn't because it was so noisy (lots of widget stuff and stuff about stuff i do)

[20:34:02.0000] <heycam>
procmailrc is your friend :)

[20:34:03.0000] <Hixie>
or equivalent, yeah

[20:35:00.0000] <Hixie>
well let's subscribe and add filters as i get tired

[23:40:00.0000] <Hixie>
...you can't take a screenshot of dvd player on mac. wtf.

[23:57:00.0000] <Rik`>
Hixie: vlc should be ok

[23:59:00.0000] <gesa>
HAHAHA You can, it just shows a photoshop-esque transparency grid. WTF.

[00:02:00.0000] <Hixie>
Rik`: i just ended up taking a screen shot of screen sharing on another machine looking at this one...

[00:03:00.0000] <Rik`>
if you want other stupid things with dvd player : http://www.hardmac.com/news/2008/12/19/use-dvdplayer-5-0-3-with-an-external-reader

[00:03:01.0000] <gesa>
Hixie: Now that's solution finding...

[00:04:00.0000] <Rik`>
I like the fact that it's by only using software that Apple ships

[00:04:01.0000] <Hixie>
i was going to screen share to another machine then screen share from that one back, so i could do it all without getting up, but for some reason i can't screen share in that direction at the moment. some configuration issue.

[00:20:00.0000] <Hixie>
wow now i've got dvd player in a completely wacked state where it's in full-screen mode, but the video is playing in a window UI... wtf

[00:20:01.0000] <Hixie>
and it pauses if i switch to the space with the dvd player but works fine if i switch to another

[00:21:00.0000] <Hixie>
jesus

[00:21:01.0000] <Hixie>
time for a reboot?

[00:21:02.0000] <gesa>
so.... window on black background? Wait wait, let me guess, you're on Lion?

[00:21:03.0000] <Hixie>
the background is actually lion's weird fabric background!

[00:21:04.0000] <gesa>
Time for a reboot.

[00:21:05.0000] <gesa>
/me is feeling pretty smug for waiting til 10.7.1 ships for an upgrade

[00:21:06.0000] <Hixie>
ok now the mission control thingy has stopped responding.

[00:21:07.0000] <Hixie>
definitely rebooting.

[00:22:00.0000] <Hixie>
yeah lion is pretty damn buggy

[00:22:01.0000] <gesa>
<-- former Apple employee. I learned things.

[00:22:02.0000] <gesa>
(Patience, that is.)

[01:10:00.0000] <Hixie>
rebooting worked a charm.

[01:10:01.0000] <Hixie>
of course at the same time the trackpad's batteries died.

[01:11:00.0000] <Hixie>
and lion with a mouse isn't pleasant.

[02:15:00.0000] <annevk>
zcorpan, complete/ is generated on html5.org

[02:16:00.0000] <annevk>
I'm not sure if I published the splitter script I use somewhere, but that would have to be updated to whatever Philip` updated it to I suppose

[02:16:01.0000] <annevk>
We should get rid of the /html /C distinction anyway

[03:34:00.0000] <annevk>
http://tools.ietf.org/html/draft-nottingham-http-new-status status code for captive portals...

[03:34:01.0000] <annevk>
I wonder if that'll get traction

[03:47:00.0000] <annevk>
ooh

[03:47:01.0000] <annevk>
almost have my first W3C REC

[03:48:00.0000] <GlitchMr>
I love how XHTML5 cannot be used with text/html :).

[03:49:00.0000] <GlitchMr>
(technically, XHTML1.1 also couldn't be used, but some did pages with XHTML1.1 DOCTYPE and text/html)

[03:50:00.0000] <annevk>
I know I did

[03:51:00.0000] <GlitchMr>
Wait... so <a href=#> is valid HTML5?

[03:52:00.0000] <GlitchMr>
I through that in HTML only [a-zA-Z0-9_.-] could be unquoted... I read specification and I learn new things...

[03:54:00.0000] <annevk>
HTML5 is a bit more flexible

[03:54:01.0000] <Ms2ger>
Hah, I did XHTML1.1 with application/xhtml+xml

[03:55:00.0000] <GlitchMr>
That's good :)

[03:55:01.0000] <GlitchMr>
It annoys me to see XHTML DOCTYPE and text/html

[03:56:00.0000] <GlitchMr>
If sites say text/html, that means they're HTML. But using /> is invalid HTML (I don't count HTML5, it's different thing).

[03:57:00.0000] <GlitchMr>
Hi, Amorphous

[03:57:01.0000] <Ms2ger>
It's conforming WHATWG HTML in any case

[03:57:02.0000] <annevk>
Got to love that DOM Views thread

[03:58:00.0000] <Ms2ger>
And MS

[03:58:01.0000] <annevk>
The WebApps WG decided somewhere in 2009 to add a note to that document and Ian Jacobs back then said it was okay

[03:58:02.0000] <annevk>
Then I finally got around to it and we had to rescind it instead

[03:58:03.0000] <annevk>
Now it is okay again

[03:58:04.0000] <annevk>
And MS is acting silly as a bonus

[03:59:00.0000] <GlitchMr>
IE9 doesn't fail on errors in application/xhtml+xml. Kinda weird...

[04:01:00.0000] <annevk>
"Therefore we have rejected this issue and retained the scroll event as a UIEvent."

[04:01:01.0000] <annevk>
DOM Level 3 Events is such a disaster

[04:02:00.0000] <Ms2ger>
Noticed their dependency on HTML5?

[04:03:00.0000] <annevk>
haha but not Web IDL?

[04:03:01.0000] <annevk>
classic

[04:04:00.0000] <Ms2ger>
But they actually claim it's informative

[04:04:01.0000] <annevk>
of course

[04:11:00.0000] <annevk>
I wonder if it is worth fighting DOM3Events or whether I should focus on something else

[04:12:00.0000] <annevk>
The reasoning Microsoft gives is often very bad, but nobody else seems to care

[04:25:00.0000] <roc>
I think smaug cares

[04:35:00.0000] <annevk>
I think he might attend the teleconferences yes

[04:35:01.0000] <annevk>
I really wonder though how he would agree to things like

[04:35:02.0000] <annevk>
"We agreed that the scroll event fits as a UIEvent from a taxonomy standpoint."

[04:35:03.0000] <annevk>
As he normally is quite opposed to useless cruft

[04:35:04.0000] <annevk>
So either he is not paying much attention or something else is going on

[04:50:00.0000] <annevk>
By the way, thanks Ms2ger and AryehGregor for replying on the DOM Core thread

[04:51:00.0000] <annevk>
AryehGregor, merging in Range is already listed as open issue, I'm not in a hurry to actually do it quite yet though

[04:51:01.0000] <roc>
talk to him

[05:14:00.0000] <annevk>
createElementNS("http://www.w3.org/1999/xhtml", "TABLE")

[05:15:00.0000] <annevk>
HTMLUnknownElement yay

[05:23:00.0000] <Ms2ger>
WebKit implemented HTMLUnknownElement recently

[05:38:00.0000] <annevk>
Ms2ger, any idea on how to implement http://dev.w3.org/2006/webapi/XMLHttpRequest-2/xhr-source using --filter instead?

[05:38:01.0000] <annevk>
Would be kind of nice to just use the Makefile

[05:41:00.0000] <Ms2ger>
...For the last Last Call Working Draft < Heh

[05:46:00.0000] <Ms2ger>
Shouldn't be too hard, I think

[05:46:01.0000] <Ms2ger>
Or you can run your script from the Makefile

[05:51:00.0000] <annevk>
That's an interesting idea

[05:58:00.0000] <gsnedders>
Ms2ger: Interseting. I know othermaciej was against it — though never got around to writing anything objecting to it as he said he would.

[05:58:01.0000] <annevk>
http://www.w3.org/2010/html-xml/2011/08/09-minutes#item05 :)

[06:00:00.0000] <gsnedders>
Still sad the reason why it had to get into the spec. :( (The only mostly standards based code-path on Yahoo Mail is the Gecko one — both the WebKit and IE ones rely on non-standard stuff — so for any other browser to be competitive you have to implement the only-just-being-specced HTMLUnknownElement, XMLParser, and DOMSerializer, and change your UA string to pretend to be Firefox.)

[06:00:01.0000] <gsnedders>
(And people care about Yahoo Mail, so you have to support one of their three browser-specific code-paths)

[06:03:00.0000] <roc>
what non-standard stuff does the Webkit code path use?

[06:03:01.0000] <gsnedders>
roc: I forget — it was quite a long time since I looked at it

[06:14:00.0000] <gsnedders>
roc: Digging through bugs, I think it was relying on things like exact innerHTML serialization, random edge-cases around ranges.

[06:14:01.0000] <gsnedders>
(And, uh, it appears what eventually launched at the new Yahoo Mail supports us, if I understand correctly.)

[06:15:00.0000] <roc>
we should be able to get around those issues as the specs tighten up and Webkit fixes their bugs

[06:15:01.0000] <gsnedders>
roc: Yeah, indeed

[06:15:02.0000] <roc>
I'd be more worried if they were using Webkit-prefixed features or Web SQL DB or something

[06:16:00.0000] <gsnedders>
roc: We support Web SQL DB, so I probably wouldn't have noticed.

[06:16:01.0000] <roc>
*I'd* have been more worried :-)

[06:16:02.0000] <gsnedders>
roc: And it's entirely possible they are using WebKit-prefixed stuff in their WebKit branch. Certainly something Google does.

[06:16:03.0000] <gsnedders>
roc: Even though they give you a codepath for your engine? :)

[06:17:00.0000] <gsnedders>
roc: And I presume the Gecko codepath doesn't have WebKit-prefixed stuff. :)

[06:17:01.0000] <Ms2ger>
gsnedders, we care about the web, not just about Gecko ;)

[06:17:02.0000] <roc>
exacxtly!

[06:17:03.0000] <gsnedders>
Then I should start telling you about Google stuff.

[06:18:00.0000] <Ms2ger>
What, NaCl?

[06:19:00.0000] <gsnedders>
Most things have four code-paths: an IE one relying upon loads of IE bugs; a Gecko one using lots of Gecko-prefixed stuff; a WebKit one using lots of WebKit-prefixed stuff; and an Opera one using lots of Opera-prefixed stuff (and historically relying upon random bugs too, though we've managed to get rid of the worst of that to fix the bugs).

[06:19:01.0000] <gsnedders>
If you want to bring a new engine to market, you're pretty much fucked by Google.

[06:20:00.0000] <Ms2ger>
There is Opera-prefixed stuff? I thought you guys never got around to implementing something before everyone dropped prefixes? ;)

[06:23:00.0000] <annevk>
Wait, is this the reverse joke from we did everything back in 95?

[06:37:00.0000] <roc>
Ms2ger has a mean streak

[06:37:01.0000] <Ms2ger>
/me bows

[08:24:00.0000] <annevk>
oh yay

[08:25:00.0000] <annevk>
XMLHttpRequest was already correct with respect to fetch and asynchronously returning

[08:37:00.0000] <annevk>
Ms2ger, did you move the XMLHttpRequest test suite?

[08:37:01.0000] <annevk>
Ms2ger, http://test.w3.org/webapps/tests/XMLHttpRequest/info.htm no longer being a valid link is annoying

[08:43:00.0000] <Ms2ger>
Yes, I think I did

[08:43:01.0000] <Ms2ger>
I can fix

[08:49:00.0000] <annevk>
Not that anyone complained by the way

[08:49:01.0000] <annevk>
I wonder how many people noticed

[09:04:00.0000] <annevk>
I keep discovering new fun issues in DOM Level 3 Events

[09:04:01.0000] <annevk>
ECMAScript is a reference but never referenced

[09:04:02.0000] <annevk>
The mapping of OMG IDL to ECMAScript is not defined

[09:05:00.0000] <annevk>
It talks about a PostScript version which does not exist, subsection of appendix G is labeled F, etc.

[09:05:01.0000] <annevk>
I think DOM Level 3 Events is a fine example of monkeypatching

[09:11:00.0000] <annevk>
Can someone change http://wiki.whatwg.org/index.php?title=Research&redirect=no so that it does not appear in the Proposals category?

[09:16:00.0000] <Dashiva>
I think it's supposed to be [[:Category:Proposals]]

[09:43:00.0000] <gsnedders>
annevk: Is there a testsuite for DOM Level 3 Events?

[09:43:01.0000] <Ms2ger>
gsnedders, a mostly manual one

[09:44:00.0000] <Ms2ger>
http://dvcs.w3.org/hg/webapps/file/tip/DOMEvents

[09:44:01.0000] <annevk>
thanks Dashiva

[09:44:02.0000] <gsnedders>
Because, although I know it's trolling, I'm so tempted to send feedback saying, "What defines that there is a createEvent property on this object? The IDL states there is something on the interface, but doesn't define that is actually exposed as a property."

[09:45:00.0000] <Ms2ger>
gsnedders, are you prepared to get an email back next year saying that that's intentional?

[09:46:00.0000] <Ms2ger>
Or that it's defined by the informative WebIDL appendix

[09:46:01.0000] <gsnedders>
"Please document that you are relying upon the informative appendix for the testsuite."

[09:47:00.0000] <Ms2ger>
No. Our customers rely on the stability of our tests, not on their correctness.

[09:48:00.0000] <gsnedders>
(Basically following the reducio ad absurdum route of making them call-out their only way to test their spec is by relying upon non-standard behaviour.)

[09:49:00.0000] <Ms2ger>
Sure

[09:49:01.0000] <Ms2ger>
Is it worth your time?

[09:50:00.0000] <gsnedders>
No. Not at all.

[09:50:01.0000] <gsnedders>
There's going to be interop because everyone will follow the informative appendix.

[09:51:00.0000] <Ms2ger>
"OpenClass - HTML 5 Workshop (Supported by Opera Software)"

[09:51:01.0000] <Ms2ger>
What?

[09:51:02.0000] <Ms2ger>
"Use this area to offer a short teaser of your email's content. Text here

[09:51:03.0000] <Ms2ger>
   will show in the preview area of some email clients."

[09:53:00.0000] <gsnedders>
What?

[09:56:00.0000] <Ms2ger>
One isn't expected to read that email in plain text mode

[09:59:00.0000] <gsnedders>
Yeah, I got that much,

[09:59:01.0000] <gsnedders>
Just… what?

[12:52:00.0000] <jarek>
Hi

[12:52:01.0000] <gesa>
Hi.

[12:53:00.0000] <jarek>
do you know of any compatability tables that would show which CSS featues are supported on mobiles?

[12:53:01.0000] <paul_irish>
http://caniuse.com is pretty good at that

[12:53:02.0000] <jarek>
I mean something like http://caniuse.com/, but with mobile browsers

[12:53:03.0000] <paul_irish>
it has mobile browsers

[12:54:00.0000] <gesa>
Yeah, far right column iirc

[12:54:01.0000] <jarek>
oh wait, indeed...

[15:07:00.0000] <jarek>
what's the proper english term for describing the page area that is visible after load witout scrolling?

[15:07:01.0000] <gesa>
above the fold?

[15:09:00.0000] <jarek>
hmm… how about "fold area"? Would this still have the same meaning?

[15:09:01.0000] <gesa>
not entirely. what's the context?

[15:10:00.0000] <jarek>
uhm… it's a big complicated

[15:10:01.0000] <gesa>
"initial view"?

[15:10:02.0000] <jarek>
I have an app that shows a preview of the website, I would like to put a horizontal line in the place where page fold is

[15:11:00.0000] <Philip`>
http://www.motive.co.nz/glossary/fold.php seems to use "first fold"

[15:11:01.0000] <jarek>
now I would like to toggle this feature via menu, but I don't know how to call it :)

[15:13:00.0000] <jarek>
Philip: thanks!


2011-08-14
[17:22:00.0000] <The_8472>
<jarek> I have an app that shows a preview of the website, I would like to put a horizontal line in the place where page fold is <- uh yeah... how about different screen sizes?

[17:23:00.0000] <The_8472>
or even browser viewport sizes

[17:23:01.0000] <The_8472>
ah damn, he quit

[17:23:02.0000] <gesa>
Clearly I have been spending too much time in THE FUTURE because I just assumed it was all adaptive...

[17:25:00.0000] <The_8472>
the point is that it can change, so how can you adapt to future changes?

[17:27:00.0000] <gesa>
I suppose it depends entirely on the use case.

[17:28:00.0000] <The_8472>
probably, yes. but if it's about trying to optimize for any particular size then it would be a bad idea

[17:28:01.0000] <gesa>
I concur. The idea of optimizing for a specific size didn't even occur to me because I forgot people still do that.

[17:30:00.0000] <The_8472>
i had to fight tooth and nail to get *one* dynamic width page into a 1024x786 "optimized" (read: fixed width layout) backend to cram all that information in there

[17:32:00.0000] <gesa>
I don't follow. One site with a fixed width layout and a single dynamic width page? What do you mean 1024x786 backend?

[17:35:00.0000] <gesa>
(bear with me. I've been coding html email all day. brain kaput.)

[17:35:01.0000] <The_8472>
corporate software with a web interface. it's using a fixed width layout, optimized for that resolution. and they wanted an overview page where you can get all the relevant information at one glance (if possible)

[17:35:02.0000] <gesa>
oh jesus.

[17:35:03.0000] <The_8472>
so to cram as much information onto a single page i proposed a dynamic width layout to make use of the screen space on larger monitors

[17:35:04.0000] <The_8472>
of course i was rattling on some sacrosanct "guidelines" there...

[04:09:00.0000] <annevk>
There are three bugs on caretBlinkRate but there is no such feature in the specification? Am I missing something?

[04:19:00.0000] <Ms2ger>
Might be a W3Cism

[04:20:00.0000] <annevk>
http://dev.w3.org/html5/2dcontext/ cannot find it

[04:21:00.0000] <annevk>
oh wait I can

[04:25:00.0000] <annevk>
I guess Microsoft will implement that feature?

[04:25:01.0000] <annevk>
Is anyone planning on implementing it?

[04:27:00.0000] <Ms2ger>
Don't see a bug over here

[04:57:00.0000] <annevk>
Does anyone know if it is easy to file RFC errata?

[04:58:00.0000] <annevk>
http://tools.ietf.org/html/rfc2388 talks about application/x-url-encoded rather than application/x-www-form-urlencoded

[05:02:00.0000] <Philip`>
Does anybody bother to read errata?

[05:03:00.0000] <Philip`>
/me usually forgets about them even if he's aware there's an errata page, so they seem a bit useless :-(

[05:04:00.0000] <annevk>
I guess at some point there's enough bugs to consider issuing a new draft

[05:26:00.0000] <rabbi1>
how can i get "A List Apart" code ?

[05:47:00.0000] <gsnedders>
annevk: AFAIK it's not too hard if the authors of the RFC care at all, if you just follow the normal form

[06:09:00.0000] <asmodai>
Mmm

[06:09:01.0000] <asmodai>
interesting

[06:10:00.0000] <asmodai>
So Safari 5.1 tries to do MathML

[06:10:01.0000] <asmodai>
I wonder what it is trying to use for a font for part of this rendering

[06:10:02.0000] <asmodai>
Giving me boxes for parts of the rendering (mostly braces and the likes)

[06:13:00.0000] <annevk>
Ah I found a form

[06:19:00.0000] <annevk>
http://www.rfc-editor.org/errata_search.php?rfc=2388&eid=2937 done

[07:53:00.0000] <david_carlisle>
asmodai:on  safari mathml, it probably assumes stix fonts (which i gather are distributed with the latest osx, or are freely available otherwise.

[07:53:01.0000] <asmodai>
ah

[07:54:00.0000] <asmodai>
david_carlisle: will install those later. thought it would have sensible defaults to the Win 7 math fonts

[07:54:01.0000] <asmodai>
ff at least does.

[07:55:00.0000] <asmodai>
jeez, they really need to update their site again

[07:56:00.0000] <david_carlisle>
asmodai: it would be even more sensible if the cambria math windows font was distributed with a licence that meant it could be used on other platfoms (like the original ms core fonts)

[08:01:00.0000] <gsnedders>
david_carlisle: There's no way MS is going to do that again, though, sadly. They consider fonts too much of a competitive advantage now…

[08:01:01.0000] <david_carlisle>
gsnedders: probbaly, shame though as it's a nice font.

[08:02:00.0000] <gsnedders>
Indeed. Nobody buys an OS because of what fonts it has, so it's not really an advantage.

[08:06:00.0000] <zewt>
sure they do

[08:06:01.0000] <zewt>
if Joe Consumer sees linux next to windows in a store, and due to having second-rate fonts linux looks unpolished next to Windows, that'll easily nudge decisions a little

[08:34:00.0000] <gsnedders>
zewt: Well, okay, but by-and-large fonts are "good enough".

[08:53:00.0000] <annevk>
Has anyone tallied email on www-tag?

[08:53:01.0000] <annevk>
E.g. useful vs euh what is going on?

[08:57:00.0000] <jarek>
Hi

[08:58:00.0000] <jarek>
is there any standard describing which CSS featuers should be supported in print stylesheets?

[08:58:01.0000] <Ms2ger>
All of them

[08:58:02.0000] <jarek>
but most browsers ignore e.g. outlines and backgrounds

[08:59:00.0000] <jarek>
which of course make sense

[08:59:01.0000] <jarek>
is this behavior non-standard?

[08:59:02.0000] <jarek>
s/make/makes

[08:59:03.0000] <Ms2ger>
"The following are links to the Editor's draft of the geolocation documents, they are not final and should NOT be relied upon by implementors."

[08:59:04.0000] <Ms2ger>
Nice one

[08:59:05.0000] <Ms2ger>
jarek, see the thread on www-style

[09:03:00.0000] <jarek>
Ms2ger: do you mean this thread? http://lists.w3.org/Archives/Public/www-style/2011Aug/0269.html

[09:03:01.0000] <Ms2ger>
Yep

[09:05:00.0000] <jarek>
is there a reference table that would list which CSS properties are not supported by browsers when using print media type?

[09:38:00.0000] <AryehGregor>
TabAtkins, why don't you just make zero-height and zero-width radial gradients invalid syntax, so they get ignored?  Is there any use for them?

[09:46:00.0000] <astearns>
normal, legitimate gradients under transforms approach zero-width and zero-height, so we still need to decide what to do with them

[12:11:00.0000] <webben>
http://lists.w3.org/Archives/Public/public-html-xml/2011Aug/0010.html : because Microsoft is well keen on "portable" file formats. ;)

[12:14:00.0000] <Ms2ger>
He might agree with annevk, but I doubt annevk would agree with him

[12:14:01.0000] <Ms2ger>
Anyway, happy I stayed out of that

[12:22:00.0000] <Philip`>
So Microsoft should fully implement HTML5 (plus CURIEs) in XSD1.1 in order to help solve the US healthcare crisis?

[12:22:01.0000] <Philip`>
Sounds like a good plan to me

[12:22:02.0000] <webben>
Yes that was also ... an interesting suggesting.

[12:22:03.0000] <webben>
*suggestion

[12:23:00.0000] <MikeSmith>
webben: thanks fighting the good fight in bugzilla, man

[12:23:01.0000] <Ms2ger>
/me wonder where webben had a fight

[12:23:02.0000] <webben>
MikeSmith: Hey, yw :)

[12:24:00.0000] <Ms2ger>
/me also wonder why can't tpye

[12:24:01.0000] <MikeSmith>
Philip`: can anybody really claim at this point that a giant red fag was not raised about CURIEs a gazillion years ago?

[12:25:00.0000] <Ms2ger>
A giant red fag?

[12:25:01.0000] <MikeSmith>
um

[12:25:02.0000] <MikeSmith>
FLAG

[12:26:00.0000] <Ms2ger>
Oh

[12:26:01.0000] <Ms2ger>
Yes, that does make more sense

[12:26:02.0000] <MikeSmith>
typing on  iPad is an adventure

[12:28:00.0000] <Philip`>
Choosing a name strongly associated with cancer doesn't seem like a good sign either :-(

[12:28:01.0000] <Ms2ger>
Philip`++

[12:29:00.0000] <MikeSmith>
anyway, the message now appears to be, WE are the ones willing to compromise, and you are not. therefore we are justified in continuing the same bs we havw

[12:29:01.0000] <MikeSmith>
oops

[12:29:02.0000] <MikeSmith>
anyway

[12:30:00.0000] <MikeSmith>
said enough already I guess

[12:30:01.0000] <Ms2ger>
Well, that's better than "Our design is cleaner. therefore we are justified in continuing the same bs"

[12:31:00.0000] <MikeSmith>
dunno

[12:31:01.0000] <Ms2ger>
What about "it's not worse"?

[12:32:00.0000] <MikeSmith>
but really, what pressure for change would there have ever been unless microdata had come along

[12:32:01.0000] <MikeSmith>
I wonder

[12:33:00.0000] <Philip`>
They should have called it an Abbreviated URI Character-Lessening Encoding, for a more ear-related association, or something along those lines

[12:33:01.0000] <Ms2ger>
MikeSmith, meh, wondering... I prefer ranting on public-webapps :)

[12:34:00.0000] <MikeSmith>
yeah, well, microdata was clearly not a mistake

[12:34:01.0000] <MikeSmith>
however dudes might like to try to rewrite history now

[12:34:02.0000] <Ms2ger>
Ah, but the winners get to do that

[12:35:00.0000] <Ms2ger>
So if Hixie ever gets bored of this kind of science fiction, he can write his autobiography

[12:36:00.0000] <MikeSmith>
oh man

[12:36:01.0000] <Philip`>
Rewriting history after you've won is less useful than rewriting history while you're still fighting in order to help you win

[12:37:00.0000] <MikeSmith>
Ms2ger: that would sell like hotcakes for sure

[12:37:01.0000] <Ms2ger>
I'd buy one

[12:38:00.0000] <Ms2ger>
But only if I get an autograph

[12:38:01.0000] <MikeSmith>
Philip`: no winners here, unfortunately

[12:39:00.0000] <MikeSmith>
just an should-have-been-avoidable mess

[12:43:00.0000] <Philip`>
Ms2ger: Traditional autobiographies are obsolete before they've even been printed - it ought to be a continually-updated living autobiography instead

[12:43:01.0000] <Philip`>
I guess you could tell a lot about people's attempts to rewrite history if the history had an SVN log

[14:56:00.0000] <Hixie>
i may have to temporarily disable pdf production unless someone else wants to provide a web service that i can call to get them generated

[14:57:00.0000] <Hixie>
right now the setup i have is spiking ram consumption over 1GB each time they are generated and i don't have the time to go fix it

[15:07:00.0000] <Hixie>
christ the websocket protocol has been messed up

[15:07:01.0000] <Hixie>
i just noticed that they totally broke how the origin headers work and the spec self-contradicts itself all over about it

[15:26:00.0000] <yuhong>
Is hsivonen here?

[15:28:00.0000] <Philip`>
yuhong: No

[15:28:01.0000] <yuhong>
I was going to report an internal error.

[15:29:00.0000] <yuhong>
Validate anandtech.com using validator.nu.


2011-08-15
[17:06:00.0000] <AryehGregor>
Hixie, http://www.w3.org/Bugs/Public/show_bug.cgi?id=13551 was filed by an implementer -- you really want it assigned to contributor⊙wo?

[17:07:00.0000] <Hixie>
assigning to contributor⊙wo just means i think that it would be productive for the "editorial assistants" to look at it first

[17:08:00.0000] <AryehGregor>
k.

[17:08:01.0000] <Hixie>
i'll go through them all eventually too

[17:14:00.0000] <AryehGregor>
TabAtkins, it's *possible* to play a game based on WASD+mouselook without a mouse lock, if the mouse sensitivity isn't really low.  You just have to do some 360s every time the mouse gets too close to the edge of the screen.

[17:15:00.0000] <zewt>
it's ... really not, heh

[17:15:01.0000] <zewt>
like, sure, you can drive a car that only turns left, but ... not really

[17:16:00.0000] <AryehGregor>
I didn't say it would be *fun*.

[17:22:00.0000] <Philip`>
If you don't want it to be fun, just use the arrow keys for turning (and alt+arrows for strafe) then you don't need mouselook at all

[01:06:00.0000] <MikeSmith>
question about the Do Not Track feature/spec, for anybody who knows

[01:06:01.0000] <MikeSmith>
if I send a DNT: 1 opt-out header to "first-party" site, does that mean it also can't do any "collection, retention, and use of all data related  to the request and response"?

[01:07:00.0000] <MikeSmith>
or does it mean only any third-party sites with embedded content on that page can't do collection, retention, and use of all data related to the request and response?

[01:07:01.0000] <MikeSmith>
it's not clear from the spec

[01:07:02.0000] <MikeSmith>
http://tools.ietf.org/html/draft-mayer-do-not-track-00

[01:08:00.0000] <MikeSmith>
the spec says, "In processing a request that includes an OPT-OUT header, a server MUST NOT perform THIRD-PARTY TRACKING"

[01:09:00.0000] <MikeSmith>
but it doesn't say the first-party site can't do "tracking" (= collection, retention, and use of all data related to the request and response)

[01:10:00.0000] <MikeSmith>
and given a request with a DNT: 1 header, how does a site know whether it's a third-party site or a first-party site?

[01:14:00.0000] <MikeSmith>
or does the UA only send the DNT: 1 header to third-party sites but not not first-party sites?

[01:47:00.0000] <zcorpan>
Hixie: you recall why websocket has sec-websocket-origin rather than just origin?

[01:56:00.0000] <jgraham>
zcorpan: Didn't it have something to do with XHR originally

[01:56:01.0000] <jgraham>
Then I have a feeling that the real reason got removed by some other change and the sec-websocket- bit got cargo culted in

[01:56:02.0000] <jgraham>
Or stayed in for cargo-cult reasons

[01:58:00.0000] <zcorpan>
maybe websocket's origin header predated the other origin header

[02:15:00.0000] <MikeSmith>
Ms2ger: do you know if Alex Fowler is around on IRC much?

[02:15:01.0000] <Ms2ger>
Who's that?

[02:15:02.0000] <MikeSmith>
he's involved with the Do Not Track feature

[02:16:00.0000] <jgraham>
Ms2ger: What code did you want to share between the testsuites?

[02:16:01.0000] <Ms2ger>
MikeSmith, not following that

[02:16:02.0000] <MikeSmith>
ok

[02:16:03.0000] <Ms2ger>
jgraham, the DOMTokenList reflection stuff

[02:18:00.0000] <jgraham>
I guess the idea is that one could implement one feature without the other

[02:19:00.0000] <jgraham>
(I am not really opposed to sharing code, but can imagine more fun things to spend time on)

[02:19:01.0000] <Ms2ger>
I just want to have a shared testDOMTokenListReflectionStuff(element, attribute) somewhere

[02:23:00.0000] <annevk>
https://rniwa.com/2011-08-14/apple-style-span-is-gone/ is nice

[02:42:00.0000] <annevk>
making sense of bugs Leif files is hard

[02:51:00.0000] <zcorpan>
heh, the "IDL" labels look different in opera, firefox and chrome

[02:56:00.0000] <annevk>
I'm trying to come up with something to write for the WHATWG Weekly but nothing really comes to mind

[02:56:01.0000] <annevk>
end of Last Call, over a hundred mostly editorial and minor technical fixes were made to HTML, and some attempt at getting input on HTML.next

[02:56:02.0000] <annevk>
that's one sentence...

[03:02:00.0000] <zcorpan>
if nothing much has happened, that's ok to write in a weekly report :)

[03:07:00.0000] <Ms2ger>
Some discussion about the organization of DOM specs

[03:10:00.0000] <annevk>
I wouldn't really call that discussion :(

[03:11:00.0000] <annevk>
Microsoft does not participate and objects when publishing; sucks

[03:13:00.0000] <webben>
annevk: what did they object to publishing?

[03:13:01.0000] <annevk>
DOM Core

[03:14:00.0000] <annevk>
and then when I make a suggestion they do not answer

[03:14:01.0000] <annevk>
further slowing progress

[03:19:00.0000] <annevk>
put that on the record: http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/0869.html

[03:27:00.0000] <annevk>
smaug____, I was wondering, do you agree with all the DOM 3 Events stuff?

[03:27:01.0000] <annevk>
smaug____, because some of those replies to Last Call comments are rather, well, odd

[03:27:02.0000] <smaug____>
which one in particularly

[03:27:03.0000] <annevk>
e.g. on DOM features

[03:28:00.0000] <smaug____>
"odd" replies are the norm in standardization, apparently :)

[03:28:01.0000] <annevk>
or changing scroll to a UIEvent

[03:28:02.0000] <annevk>
which contradicts e.g. CSSOM View

[03:28:03.0000] <annevk>
and implementations

[03:30:00.0000] <annevk>
smaug____, so you support the decisions?

[03:30:01.0000] <smaug____>
Let's say so that I'm ok with the decisions

[03:31:00.0000] <annevk>
wow

[03:31:01.0000] <smaug____>
how does changing scroll to UIEvent contradicts with CSSOM ?

[03:33:00.0000] <annevk>
it says to use Event

[03:33:01.0000] <annevk>
but that event overview is meaningless anyway

[03:33:02.0000] <annevk>
what matters is what interface is used when dispatching

[03:34:00.0000] <annevk>
the way events are defined in DOM Level 3 Events is rather laughable compared to how dispatching events is defined in other specifications

[03:34:01.0000] <smaug____>
(of topic,  HTML spec and Web DOM code contradict with implementations quite often )

[03:34:02.0000] <annevk>
usually with good reason

[03:35:00.0000] <annevk>
not "to make this table look pretty"

[03:35:01.0000] <smaug____>
not very often with  good reason

[03:35:02.0000] <smaug____>
but I'm ok change scroll back to event

[03:35:03.0000] <smaug____>
I just don't really care whether it is UIEvent or Event

[03:35:04.0000] <annevk>
well you disagree with the reason sometimes

[03:36:00.0000] <annevk>
but making it UIEvent because the spec's organization is cleaner then is nonsense

[03:36:01.0000] <annevk>
than*

[03:37:00.0000] <smaug____>
that wasn't shepazu's reason to make scroll UIEvent

[03:40:00.0000] <smaug____>
I might even like if most of the event in browsing context were UIEvents

[03:41:00.0000] <smaug____>
since those give easy access to window

[03:41:01.0000] <smaug____>
s/event/events/

[03:51:00.0000] <annevk>
smaug____, where does DOM Core contradict specifications btw without good reason?

[03:51:01.0000] <annevk>
smaug____, euh, implementations

[03:51:02.0000] <jgraham>
annevk: (I assume that depends what you mean by "good reasons" e.g. removing attr nodes might not be considered "good reasons")

[03:51:03.0000] <annevk>
smaug____, cause I thought people were on board with Node : EventTarget, Attr no longer being a node, etc.

[03:52:00.0000] <smaug____>
those are ok.

[03:53:00.0000] <smaug____>
ÍMO

[03:53:01.0000] <smaug____>
CDATA removal probably isn't

[03:54:00.0000] <smaug____>
at least there hasn't been any good proposal how to fix the problems removing CDATA causes

[03:54:01.0000] <smaug____>
(related to serialization )

[03:56:00.0000] <smaug____>
oh, hmm, normalize() is removed

[03:56:01.0000] <smaug____>
why

[03:56:02.0000] <annevk>
sicking has a proposal and there's a patch implementing CDATA removal in Gecko

[03:57:00.0000] <smaug____>
I know Ms2ger has the patch to remove CDATA from Gecko

[04:02:00.0000] <annevk>
smaug____, presumably normalize() is either not implemented widely or lacks good use cases

[04:15:00.0000] <annevk>
wait

[04:15:01.0000] <annevk>
if CR LF is always normalized on submitting

[04:15:02.0000] <annevk>
text/plain can be parsed reliably

[04:15:03.0000] <annevk>
with respect to form submission

[04:28:00.0000] <zcorpan>
foolip: yt? you have any opinion about http://www.w3.org/Bugs/Public/show_bug.cgi?id=12405 ?

[05:19:00.0000] <rabbi1>
Can you help? i am trying to achieve this http://imagebin.org/167884 in a <ul> the data in green background appear when the user logs in. All is part of a <ul>. <li> has contents for user logged in and user logged out......

[05:59:00.0000] <annevk>
MikeSmith, it seems that "amir arsalan <m_a_s_7⊙yc>" is spamming Bugzilla somehow

[05:59:01.0000] <annevk>
MikeSmith, see http://lists.w3.org/Archives/Public/public-html-bugzilla/2011Aug/1706.html and http://lists.w3.org/Archives/Public/public-html-bugzilla/2011Aug/1707.html

[05:59:02.0000] <MikeSmith>
thanks

[06:00:00.0000] <MikeSmith>
trying to delete his account now

[06:04:00.0000] <annevk>
Google acquires hardware part of Motorola: http://googleblog.blogspot.com/2011/08/supercharging-android-google-to-acquire.html

[06:10:00.0000] <MikeSmith>
annevk: disabled his account

[06:13:00.0000] <foolip>
zcorpan, it seems that if we want people to mix native controls with custom subtitles, etc, then we need to expose events for them showing/hiding together with the size

[06:14:00.0000] <foolip>
or we could abuse the multitrack stuff to "sync" a src-less <video controls> with the <video> elements that have the actual video streams

[06:15:00.0000] <zcorpan>
events could work

[06:16:00.0000] <zcorpan>
i don't follow what you mean with the multitrack stuff

[06:59:00.0000] <annevk>
after publishing I realize I should have mentioned the UndoManager stuff

[06:59:01.0000] <annevk>
noted for next week

[07:34:00.0000] <jarek>
Hi

[07:35:00.0000] <jarek>
what's the definition of "visual", "tactile" and "bitmap" media type as used in http://www.w3.org/TR/css3-mediaqueries/ ?

[07:36:00.0000] <jarek>
does "visual" include "tty"? Or does "bitmap" include print?

[07:37:00.0000] <annevk>
no

[07:37:01.0000] <annevk>
types are exclusive

[07:37:02.0000] <annevk>
see CSS 2.1

[07:37:03.0000] <jarek>
ahh... right...

[07:38:00.0000] <jarek>
I just googled the terms and found CSS2.1 specs: http://www.w3.org/TR/CSS2/media.html

[07:38:01.0000] <annevk>
media queries 3.1 will prolly fix this

[07:38:02.0000] <annevk>
well fix in the sense that it will define both media types and media features and not defer to CSS 2.1 for the former

[07:40:00.0000] <jarek>
why "resolution" property accepts only dpi and dpcm units?

[07:40:01.0000] <jarek>
why not ppi?

[07:40:02.0000] <annevk>
what's the difference with dpi?

[07:40:03.0000] <jarek>
so for screen media type, dpi is treated like ppi?

[07:41:00.0000] <jarek>
I thought that DPI should be only used for printed media, where you have actual dots

[07:42:00.0000] <annevk>
I think so

[07:42:01.0000] <annevk>
maybe suggest on www-style that it is clarified

[07:42:02.0000] <annevk>
although I'm the editor I don't really know much about these units

[07:50:00.0000] <jarek>
CSS3 media queries specification says that resolution may accept only dpi and dpcm units and that resolution property applies to bitmap media types

[07:50:01.0000] <jarek>
now CSS2 specs defines bitmap media types as:

[07:51:00.0000] <jarek>
print, projection, screen and tv

[07:51:01.0000] <jarek>
I guess this means that DPI should be treated as PPI

[07:52:00.0000] <jarek>
at least on screen

[07:52:01.0000] <annevk>
http://en.wikipedia.org/wiki/Dots_per_inch#Computer_monitor_DPI_standards

[07:54:00.0000] <annevk>
yay, WHATWG Blog has >6000 subscribers per Google Reader

[07:54:01.0000] <jarek>
now I'm confused, how windows can <<set default display "DPI" to 96 PPI>>? Isn't this value depended on monitor size and resolution?

[07:54:02.0000] <benjoffe_>
hmm didn't know there's a blog, make that 6001

[07:55:00.0000] <benjoffe_>
well, >6001

[07:57:00.0000] <Philip`>
jarek: They're independent of physical inches

[07:58:00.0000] <benjoffe_>
jarek: I'm pretty sure that's due to a lack of monitor dpi api's

[07:58:01.0000] <annevk>
AryehGregor, there were several comments on your blog post that did not get approved thus far

[07:58:02.0000] <annevk>
AryehGregor, I approved them

[07:58:03.0000] <jarek>
so there are actually three units: DPI, logical PPI and physical PPI?

[07:59:00.0000] <jarek>
nevermind, this thing makes no sense

[08:00:00.0000] <jarek>
I will just hard-code in my code that PPI equals DPI, hopefully no one will notice :P

[08:01:00.0000] <Philip`>
There's the number of RGB cells per inch on your LCD screen, and the number of pixels of data that the computer sends to the display per cell (hence per inch), and the number of pixels of data that the computer uses to represent objects measured in inches (or in fractions of inches, e.g. points)

[08:02:00.0000] <Philip`>
plus any scaling factors the software wants to add

[08:03:00.0000] <Philip`>
Also, the length of an inch will change if you travel at relativistic speeds

[08:03:01.0000] <Philip`>
So it's quite hard to be consistent about measurements

[08:04:00.0000] <jarek>
Philip`: if iPad's specification says that it has 132 PPI screen (http://en.wikipedia.org/wiki/List_of_displays_by_pixel_density), how do I target this or any higher resolution with media queries?

[08:05:00.0000] <jarek>
will this work?

[08:05:01.0000] <jarek>
"@media screen and (min-resolution: 132dpi)"

[08:05:02.0000] <Philip`>
Haven't got a clue

[08:05:03.0000] <benjoffe_>
jarek: why don't you just test against it? Do you have an ipad?

[08:06:00.0000] <jarek>
benjoffe_: no, it's overpriced

[08:06:01.0000] <jgraham>
(only the length of an inch in a different frame of reference. So not really a problem if you are looking at your computer in a spaceship travelling at 0.98c, but perhaps worth thinking about if you are building an intergalactic Oxford Circus and want the adverts to work for passers by)

[08:06:02.0000] <Philip`>
Seems complex enough that the only reliable method is to test it (or copy from someone who's tested it)

[08:06:03.0000] <zewt>
vmware + osx + ios emulator

[08:06:04.0000] <jarek>
but I guess that this will be the only reliable way to verify it

[08:06:05.0000] <benjoffe_>
jarek: do you have xcode? the ipad simulator is available for $5, or free if you have lion

[08:06:06.0000] <benjoffe_>
(requires osx)

[08:07:00.0000] <zewt>
haha what, it was free when i downloaded it

[08:07:01.0000] <Philip`>
jgraham: Might still be a problem if your spaceship is falling into a small black hole at 0.98c

[08:07:02.0000] <Philip`>
since your monitor may accelerate faster than you

[08:08:00.0000] <benjoffe_>
zewt: it's free if you have a dev lincese which costs $99 per year

[08:08:01.0000] <jarek>
great, I have the developer license

[08:08:02.0000] <jgraham>
Yes, I was only really thinking of special relativity. It's always easier to pretend gravity doesn't exist

[08:08:03.0000] <zewt>
i don't, and i'm pretty sure i just downloaded it from apple

[08:08:04.0000] <jarek>
but I can't find it in dev section, do I need to participate in iPhone developer program? I'm currently signed up for Mac dev program

[08:09:00.0000] <benjoffe_>
it's available for me as an ipohne dev, not sure about the mac dev program

[08:09:01.0000] <benjoffe_>
i thought it would have been, considering xcode is used for mac dev too

[08:11:00.0000] <Philip`>
Running an iPad simulator on a desktop PC doesn't sound ideal if you're trying to figure out whether the real physical display resolution has an effect on the browser

[08:11:01.0000] <zewt>
it should be the same

[08:12:00.0000] <benjoffe_>
yes I would be quite surprised if there was a difference there

[08:12:01.0000] <Philip`>
Either the browser ignores the system, or the system lies to the browser, and the latter case wouldn't inspire much confidence that things aren't lying in other ways too

[08:13:00.0000] <benjoffe_>
heck, i have an ipad simulator, maybe i'll go check

[08:13:01.0000] <zewt>
what?

[08:15:00.0000] <benjoffe_>
hmm it keeps crashing, shouldn't have installed the beta.. :/

[08:15:01.0000] <zewt>
better than even odds it just lies about dpi in the same way as every other browser...

[08:16:00.0000] <jarek>
will I get iPhone and iPad emulator if I buy XCode from App Store?

[08:17:00.0000] <benjoffe_>
jarek: yes, it's part of the bundle

[08:17:01.0000] <zewt>
there's just one with a menu option to change modes

[08:17:02.0000] <jarek>
the version of XCode that I downloaded from Mac Dev Program does not ship with any emulators

[08:17:03.0000] <benjoffe_>
jarek: better off just buying lion imo though, $25 more

[08:17:04.0000] <benjoffe_>
jarek: are you sure? it's not easy to find

[08:18:00.0000] <jarek>
benjoffe_: I already have lion, it did not ship with XCode

[08:18:01.0000] <benjoffe_>
jarek: oh, if you have lion you can download xcode from the mac app store for free

[08:19:00.0000] <Philip`>
/me wonders what's the point of charging $5 for development tools, since surely it can't provide any non-trivial amount of revenue

[08:20:00.0000] <benjoffe_>
Philip`: I think it's more about getting people to attach their credit card to the store in the hope they buy other items

[08:21:00.0000] <jarek>
ahh... I have already iOS simulator installed, I just had to create new iOS project

[08:22:00.0000] <benjoffe_>
jarek: you can also launch the simulator directly if you only want to test websites, just make a shortcut, it's in a silly location like /Developer/Platforms/... (instead of the sane location of ~/Applications)

[08:22:01.0000] <jgraham>
Oh, I thought it was so that the platform wasn't tainted with poor users

[08:25:00.0000] <timeless>
:(, my g+ account was suspended

[08:25:01.0000] <timeless>
it seems that the fact that i've been using `timeless` as my name for more than half my life (and my entire professional carrer) including IRL wasn't expected behavior

[08:26:00.0000] <Ms2ger>
Myself, I didn't try

[08:26:01.0000] <zewt>
it's pretty ironic that google said that privacy is a priorty on g+, then immediately violates people's privacy by trying to force them to reveal their name

[08:26:02.0000] <zewt>
"well that was quick"

[08:26:03.0000] <timeless>
they said privacy was a priority?

[08:27:00.0000] <zewt>
everyone says that

[08:33:00.0000] <jarek>
it looks like iPad does not support resolution media query at all

[08:33:01.0000] <timeless>
those are silly anyway :)

[08:34:00.0000] <Ms2ger>
iPads?

[08:34:01.0000] <jarek>
but there is "-webkit-device-pixel-ratio" media query, is it the same thing?

[08:35:00.0000] <benjoffe_>
jarek: no, that's the number of physical pixels per css pixel

[08:35:01.0000] <benjoffe_>
which for ipad/iphone3 is 1, for iphone4 is 2

[08:37:00.0000] <timeless>
jarek: what would you do w/ a resolution media query

[08:38:00.0000] <timeless>
other than poorly support every device you've never met

[08:38:01.0000] <timeless>
/me has only recently found one site which didn't suck wrt discrimination

[08:38:02.0000] <zewt>
cheers to misapplication of the word "discrimination"

[08:39:00.0000] <timeless>
(it was Fandango fwiw, they let me choose Mobile, BB app, or Normal)

[08:39:01.0000] <timeless>
zewt: you don't think that a service which forces users to get subpar content just because it thinks they want it is discriminating against those users?

[08:39:02.0000] <jarek>
timeless: well, I expect resolution media query to be widely supported in the near future, that's why I would like to implement it in my app sooner than later

[08:40:00.0000] <zewt>
no, i think that's using a loaded word to make something sound worse than it is

[08:40:01.0000] <timeless>
it's pretty bad

[08:40:02.0000] <timeless>
if i'm not allowed to participate in something

[08:40:03.0000] <timeless>
then how is that not discrimination?

[08:41:00.0000] <Philip`>
It seems like the usual attitude is "I want my page to do something special on iPhones, how do I write a media query to target iPhones?", which is effectively an obfuscated form of browser sniffing

[08:41:01.0000] <timeless>
further, segregated busing in some cases still allowed you to ride the same buses

[08:41:02.0000] <timeless>
you just got stuck in the back

[08:41:03.0000] <timeless>
getting second or third class service on a bus was still service

[08:41:04.0000] <zewt>
jarek: does anything in media queries account for the fact that the font and layout size you want depends on the device, and not soley the dpi?

[08:41:05.0000] <timeless>
that's what we're getting on the internet

[08:42:00.0000] <timeless>
that sure sounds like discrimination to me

[08:42:01.0000] <timeless>
Philip`: indeed

[08:42:02.0000] <zewt>
eg. a phone is typically much higher resolution than a desktop monitor, much smaller, and viewed much closer

[08:42:03.0000] <zewt>
the resolution is only part of the story

[08:42:04.0000] <timeless>
anyway, as dbaron recently pointed out on www-style (?)

[08:42:05.0000] <Philip`>
(Have any browsers started lying to media queries yet, in order to receive iPhone-targeted content?)

[08:43:00.0000] <timeless>
it's easy to get media queries wrong when you reach boundary conditions

[08:43:01.0000] <timeless>
Philip`: not sure about media queries

[08:43:02.0000] <timeless>
but i know that browsers have lied to receive iPhone content in other cases/ways

[08:43:03.0000] <timeless>
and i certainly expect it to happen soon enough

[08:43:04.0000] <timeless>
..

[08:44:00.0000] <timeless>
zewt: so, which part of my logic is flawed which makes my use of the word offensive to you?

[08:44:01.0000] <zewt>
<timeless> other than poorly support every device you've never met <- you called that discrimination, when it's the opposite of discrimination: behaving generically (even if not always usefully)

[08:45:00.0000] <timeless>
oh

[08:45:01.0000] <timeless>
they don't do that

[08:45:02.0000] <timeless>
they don't behave generically

[08:45:03.0000] <zewt>
that's what i was responding to

[08:45:04.0000] <timeless>
they either give out good content or crap

[08:45:05.0000] <timeless>
and you can only get good content if you pay the <iPhone-x> fee

[08:46:00.0000] <timeless>
`We only allow whites to sit in the front of buses, anyone who we've never met can since in the back`

[08:48:00.0000] <Philip`>
There must be a better of solving the browser-sniffing problem (since clearly people often want to do that rather than querying capabilities) - given N browser/device combinations, authors want to select some subset, then new browsers/devices developed afterwards want to insert themselves into some selection of subsets (while still being uniquely identifiable themselves by later authors)

[08:49:00.0000] <Philip`>
without needing identifiers that perpetually grow in length (like UA strings)

[08:49:01.0000] <timeless>
Philip`: generally speaking `let the user pick` works well for me

[08:49:02.0000] <timeless>
it certainly makes me happy

[08:49:03.0000] <Philip`>
There must be some kind of maths that would support that sort of subsetting

[08:49:04.0000] <timeless>
have you used any of the ClearType configurator?

[08:49:05.0000] <timeless>
s/tor/tors/

[08:49:06.0000] <timeless>
(subpixel-antialiasing)

[08:50:00.0000] <timeless>
you show the user a couple of pictures and let them choose the one they like

[08:50:01.0000] <Philip`>
The thing that asks "which of these three samples looks best"?

[08:50:02.0000] <timeless>
yeah

[08:50:03.0000] <zewt>
i don't want to undergo an eye exam every time i go to a new site, heh

[08:50:04.0000] <timeless>
i'm willing to do it

[08:50:05.0000] <timeless>
i'd certainly be happy to have a button at the bottom of each site which lets me undergo such an exam

[08:51:00.0000] <zewt>
"ask the user" is always a useful fallback but should always be a fallback

[08:51:01.0000] <timeless>
it's fine w/ me for sites to guess wrong as long as they let me choose

[08:51:02.0000] <timeless>
but most of the sites i see don't do that

[08:51:03.0000] <timeless>
my favorite is the *idiotic* news sites which when i am sent directly to an article

[08:51:04.0000] <timeless>
... redirect me to the front of their mobile site

[08:51:05.0000] <timeless>
because *obviously* i didn't want to read the article i visited!

[08:53:00.0000] <zewt>
heh

[08:53:01.0000] <zewt>
generally if i click a link and it doesn't go to what i clicked, i just leave

[08:54:00.0000] <zewt>
most commonly sites that go "well, you clicked on something explicitly, but we're going to ignore you and show you an ad/ask you to sign up/ask you to do the raindance/whatever first"

[08:54:01.0000] <timeless>
i need to get google to let me say `don't ever show content from this service again`

[08:55:00.0000] <timeless>
i think it's actually possible these days

[08:55:01.0000] <timeless>
but ...

[08:55:02.0000] <zewt>
it often will show it in searches, but not always (seemingly at random)

[08:55:03.0000] <timeless>
yeah, sadly i use google news

[08:56:00.0000] <timeless>
i think for that i need to use the special configurator box or the impossible to use on my touch-based-device heisendropdown

[08:56:01.0000] <zewt>
i couldn't stand google news, because it seemed to refresh itself automatically all the time, so it always showed up highlighted in firefox as a "new tab"

[08:56:02.0000] <timeless>
(which insists on not being clickable)

[08:56:03.0000] <timeless>
hrm

[08:56:04.0000] <timeless>
it does refresh automatically

[08:56:05.0000] <timeless>
but you must be using some extension because firefox doesn't do that to me

[08:56:06.0000] <zewt>
i was

[08:56:07.0000] <zewt>
i'm not right now since i stopped using tabmixplus, but it's very useful

[08:57:00.0000] <zewt>
pages should never auto-refresh, that's just obnoxious

[08:57:01.0000] <timeless>
you don't want gmail to autorefresh?

[08:58:00.0000] <timeless>
hrm, you clearly don't use web-irc

[08:58:01.0000] <zewt>
that's ajax updating, it doesn't refresh the entire page

[08:58:02.0000] <timeless>
tomato tomato?

[08:58:03.0000] <zewt>
which is what gnews seemed to do, at least at the time

[08:58:04.0000] <zewt>
... no?

[08:59:00.0000] <zewt>
reloading the entire page is far different than a lightweight ajax update

[08:59:01.0000] <timeless>
how?

[08:59:02.0000] <timeless>
reloading the entire page could mean you could retrieve old versions from cache

[09:00:00.0000] <timeless>
(which can be useful)

[09:00:01.0000] <zewt>
how? it's sort of the ... point of ajax, heh

[09:00:02.0000] <zewt>
(only in part, of course)

[09:00:03.0000] <timeless>
if 99% of a page is content

[09:01:00.0000] <timeless>
then i'm not quite sure i'd buy that using ajax would actually mean much less load than a traditional page load

[09:01:01.0000] <zewt>
then you can update the content unintrusively

[09:01:02.0000] <zewt>
eg. greader

[09:06:00.0000] <annevk>
volkmar, if you specify trim on password, I don't really see why it shouldn't trim

[09:07:00.0000] <timeless>
annevk: why would someone trim a password?

[09:14:00.0000] <annevk>
to make copy and paste work better

[09:14:01.0000] <annevk>
if you don't allow spaces anyway

[09:14:02.0000] <volkmar>
annevk: i don't think we should allow trimming passwords, that's really dangerous

[09:15:00.0000] <annevk>
dangerous?

[09:15:01.0000] <benjoffe_>
google strips whitespace from passwords

[09:20:00.0000] <volkmar>
annevk: if I type " foobar " in a password field and they I want to login and I can't...

[09:20:01.0000] <volkmar>
I'm afraid it could be used without really realizing some side effects

[09:21:00.0000] <annevk>
how could you register such a password to begin with if they trim consistently and do not allow whitespace?

[09:22:00.0000] <zewt>
would seem like a very arbitrary special case to not support it

[09:22:01.0000] <annevk>
<input trim> UI could just mean you cannot type spaces at the beginning and end of your input

[09:22:02.0000] <jgraham>
annevk: My concern might be the *if*

[09:23:00.0000] <annevk>
jgraham, then they would have problems with their JavaScript-based solution as well

[09:23:01.0000] <annevk>
or with any restrictions on password for that matter, which certainly are around

[09:23:02.0000] <Philip`>
If you want to support accidental whitespace included around passwords, it'd be just as easy to do the trimming on the server (and then you could check both the trimmed and untrimmed passwords, in case the spaces really are part of the password)

[09:23:03.0000] <zewt>
conversely (though this is all very contrived and I don't see either ever happening in reality) someone might have their site trim spaces when storing a password, and use <input trim> to trim when entering, and not realize the *lack* of side effects due to arbitrary special cases

[09:25:00.0000] <volkmar>
annevk: in general, altering the passwords value seems a wrong idea to me

[09:26:00.0000] <timeless>
yeah, any sanitization really needs to be enforced server side

[09:27:00.0000] <timeless>
not doing that is just plain stupid

[09:27:01.0000] <timeless>
zewt: fwiw...

[09:27:02.0000] <timeless>
i think it was canada post

[09:27:03.0000] <timeless>
but one of the sites was very anal about its login form `do not remember password`

[09:27:04.0000] <zewt>
well, doing trimming client-side can make it clearer that it's happening (though in practice it probably doesn't matter whether the user is aware of it or not, in the case of passwords)

[09:27:05.0000] <timeless>
the reset account password form was *not* anal

[09:27:06.0000] <timeless>
so you could remember the password there...

[09:28:00.0000] <timeless>
so... yes... it's trivially to be stupid and inconsistent!

[09:28:01.0000] <timeless>
s/ly//

[09:28:02.0000] <timeless>
yeah, it absolutely doesn't matter to users what you do to the password

[09:28:03.0000] <zewt>
quasirelatedly, we need a real password-storage API

[09:29:00.0000] <timeless>
they don't need to know that you're storing a sha256 hash instead of the actual password

[09:29:01.0000] <zewt>
annoying how so many sites break or semi-break browser password storage

[09:29:02.0000] <timeless>
how would a password-storage api help?

[09:29:03.0000] <zewt>
firefox insists on remembering my old bugzilla password

[09:29:04.0000] <timeless>
sites would just demand to disable it, or break it, or flood it with crap

[09:29:05.0000] <zewt>
well, there are two problems, i think: remembering passwords from login forms, and ajax logins

[09:30:00.0000] <zewt>
the former is what usually works, but sometimes does not; don't know why in the particular cases, but it'd be nice for that to be more reliable

[09:30:01.0000] <zewt>
the latter almost never works; it would be nice if it could

[09:30:02.0000] <timeless>
the former doesn't work when sites actively try to break it

[09:30:03.0000] <timeless>
the easiest way is to add the attribute that says `please break password manager`

[09:30:04.0000] <zewt>
i rarely see that, usually it's just unawareness of it

[09:31:00.0000] <zewt>
as far as i'm concerned sites should never have the ability to do that

[09:31:01.0000] <timeless>
the second easiest way is to not have a <form> containing the password field

[09:31:02.0000] <timeless>
(and then use js to shift it into the form/submission)

[09:31:03.0000] <zewt>
google passwords are regularly out of date for me, too

[09:31:04.0000] <timeless>
odd, i've never hit that problem

[09:32:00.0000] <zewt>
it just never offers to store the newly-entered password

[09:32:01.0000] <zewt>
but google is weird and it doesn't happen everywhere

[09:32:02.0000] <timeless>
i'm pretty sure firefox has offered to update my password for google

[09:32:03.0000] <timeless>
/me will have to check sometime

[09:33:00.0000] <zewt>
gmail fills in the password fine, for example, but if i open gapps management, that one's wrong

[09:34:00.0000] <timeless>
you know you can `easily` `fix` it by opening the password list and deleting the one for the gapps site

[09:34:01.0000] <timeless>
(ideally you'd be able to use Larry to fix it, but he never grew that feature)

[09:34:02.0000] <zewt>
anyway, it would be nice for ajaxy logins to have the ability to integrate with password management; right now they don't at all

[09:34:03.0000] <timeless>
well

[09:34:04.0000] <timeless>
all they have to do is use <form> <input> <input type=password>

[09:35:00.0000] <timeless>
as long as they have a handler on the <form> to trap onsubmit and send it where they need

[09:35:01.0000] <timeless>
they can absolutely integrate with password management

[09:35:02.0000] <zewt>
but if the form is never actually submitted as a form, firefox will never offer to store it

[09:35:03.0000] <timeless>
you sure?

[09:35:04.0000] <zewt>
how would it know?

[09:36:00.0000] <zewt>
ajax logins typically cancel form submission entirely (if form submit happens at all), so it can do it some other way

[09:36:01.0000] <benjoffe_>
reddit seems to do login async and has never had problems remembering my login details

[09:36:02.0000] <benjoffe_>
on chrome

[09:37:00.0000] <zewt>
don't know what reddit does

[09:37:01.0000] <timeless>
data:text/html,<form action="data:text/plain," onsubmit="alert(1); return false"><input id=name><input id=pass type=password><input type=submit>

[09:37:02.0000] <timeless>
isn't working 100%, but it's definitely asking to remember my password

[09:38:00.0000] <zewt>
alert(1) in an event handler is probably going to do weird things

[09:38:01.0000] <timeless>
although, perhaps noscript is suppressing js entirely :)

[09:38:02.0000] <zewt>
data:text/html,<form action="data:text/plain," onsubmit="console.log('xxx'); return false"><input id=name><input id=pass type=password><input type=submit>

[09:39:00.0000] <zewt>
doesn't prompt

[09:39:01.0000] <zewt>
(for me, in ff6)

[09:39:02.0000] <timeless>
data:text/html,<form action="data:text/plain," onsubmit="return false"><input id=name><input id=pass type=password><input type=submit>

[09:39:03.0000] <zewt>
it shouldn't--the form submission was cancelled, after all, why would it?

[09:39:04.0000] <timeless>
prompts for me in Nightly (different profile, no noscript)

[09:39:05.0000] <timeless>
/me shrugs

[09:39:06.0000] <timeless>
it is here

[09:40:00.0000] <benjoffe_>
why are you setting 'id' on the form inputs? Should be name

[09:40:01.0000] <timeless>
what version of firefox are you using?

[09:40:02.0000] <timeless>
benjoffe_: because i can :)

[09:40:03.0000] <zewt>
as far as firefox can tell the submission was aborted, it'd be strange for it to offer to remember the password at that point

[09:40:04.0000] <timeless>
fwiw i only have add-on compatibility reporter and https-everywhere

[09:40:05.0000] <benjoffe_>
timeless: forms remember and post the 'name' attribute, if you want to play nice with password managers use it

[09:40:06.0000] <timeless>
benjoffe_: my point is that even this contrived thing *does* play nicely with password managers

[09:41:00.0000] <timeless>
at least, it does w/ my Nighty

[09:41:01.0000] <timeless>
s/y$/ly/

[09:41:02.0000] <timeless>
/me wants to know why it isn't w/ zewt 's

[09:41:03.0000] <benjoffe_>
that's surprising, they must be getting better

[09:41:04.0000] <timeless>
zewt: firefox --no-remote -P

[09:41:05.0000] <zewt>
i'm suggesting that offering to remember the password in the above code is wrong

[09:41:06.0000] <timeless>
well, it is remembering it

[09:41:07.0000] <timeless>
and remember, the offer isn't a modal dialog

[09:41:08.0000] <timeless>
so the user can safely ignore it

[09:42:00.0000] <timeless>
or say `why thank you for offering`

[09:42:01.0000] <timeless>
heh

[09:42:02.0000] <timeless>
Error: uncaught exception: [Exception... "'Can't add a login with a null or empty hostname.' when calling method: [nsILoginManager::addLogin]"  nsresult: "0x8057001e (NS_ERROR_XPC_JS_THREW_STRING)"  location: "JS frame :: resource:///components/nsLoginManagerPrompter.js :: <TOP_LEVEL> :: line 853"  data: no]

[09:42:03.0000] <zewt>
heh i cringe every time i see a prompt that says "no thanks" "yes please"

[09:42:04.0000] <timeless>
ok, well, i should say `it mostly works`

[09:43:00.0000] <zewt>
software forcibly injecting politeness into the user

[09:43:01.0000] <timeless>
since apparently the password manager's store and the ui don't agree on when to make an offer

[09:43:02.0000] <timeless>
which is partially unfortunate

[09:43:03.0000] <zewt>
that just looks like it doesn't care for data:

[09:43:04.0000] <timeless>
that's right

[09:43:05.0000] <zewt>
which wouldn't be surprising

[09:44:00.0000] <timeless>
fwiw, it's the origin data and not the destination data

[09:44:01.0000] <timeless>
(changing the action to http didn't resolve it)

[09:45:00.0000] <zewt>
makes sense

[09:45:01.0000] <timeless>
/me goes to watch a tech talk

[09:45:02.0000] <timeless>
anyway, the point is that ajax should work

[09:45:03.0000] <timeless>
unless you *try* to break password managers

[09:45:04.0000] <timeless>
which sites generally do

[09:45:05.0000] <timeless>
but that's because sites are evil

[09:45:06.0000] <zewt>
RIGHT CLICK IS NOT SUPPORTED

[09:45:07.0000] <timeless>
it isn't because it's hard to support

[09:45:08.0000] <timeless>
/me chuckles

[09:45:09.0000] <timeless>
yeah, tell that to my one button mac mouse :)

[09:57:00.0000] <dglazkov>
good morning, Whatwg!

[10:08:00.0000] <annevk>
/me thanks TabAtkins for making it easier to resolve a bunch of bugs

[10:10:00.0000] <Ms2ger>
TabAtkins, so when are you publishing a Values LCWD? :)

[10:26:00.0000] <annevk>
aah

[10:26:01.0000] <annevk>
gruber will be disappointed http://www.google.com/press/motorola/quotes/

[10:27:00.0000] <AryehGregor>
Of course they're going to *say* they're happy with it, what choice do they have?

[10:30:00.0000] <annevk>
not saying anything

[10:30:01.0000] <annevk>
which would be the usual course of action

[11:55:00.0000] <timeless>
annevk / AryehGregor : what i like is that it sounds like someone sent those guys a PR draft and instructed them to repeat it

[11:56:00.0000] <Ms2ger>
"Please tick the box next to the quote you like most"

[11:57:00.0000] <TabAtkins>
Ms2ger: Near future.  A bit more review and clean-up is needed beforehand, but not more than a month or so.

[11:58:00.0000] <TabAtkins>
AryehGregor: I strive for continuity whenever possible and practical.  Having rounding errors drastically affect rendering is bad if we can avoid it.

[11:58:01.0000] <Ms2ger>
Oh, and the Overview.html fantasai sent to www-archive had the second half of the spec nested in a <dd>, if you hadn't noticed yet

[11:58:02.0000] <TabAtkins>
Yeah, it's already been fixed.

[11:58:03.0000] <TabAtkins>
(We'd left in a </ol> instead of a </dl>, or something like that.)

[11:59:00.0000] <Ms2ger>
Sounds like something I'd do

[12:02:00.0000] <AryehGregor>
/me sees ~value.indexOf("foo") used to mean value.indexOf("foo") != -1 -- freaky but sort of cool

[12:12:00.0000] <AryehGregor>
Is this too snarky?  "This is an excellent opportunity for those who believe that the W3C is a good place to develop specs to show their willingness to improve the web.  All it would take is a modest amount of effort to submit the spec for W3C publication, and of course in their view, this would be beneficial because the W3C is a good place to publish specs.  I wait with interest for one of the many people who have expressed concern about the

[12:12:01.0000] <AryehGregor>
 editing spec not being at the W3C to spend the necessary time themselves to fix the problem, rather than expecting others to do it."

[12:12:02.0000] <AryehGregor>
/me thinks it probably is

[12:13:00.0000] <timeless>
/me doesn't see it as snarky

[12:13:01.0000] <annevk>
I would phrase it more direct

[12:21:00.0000] <AryehGregor>
Okay, I said it.

[12:21:01.0000] <AryehGregor>
Let's see how people react.

[12:21:02.0000] <timeless>
i fully expect them to ignore you :)

[12:23:00.0000] <hober>
AryehGregor: :)

[12:23:01.0000] <AryehGregor>
Either no one will do it, or someone will try briefly and then it will stagnate and die.

[12:23:02.0000] <AryehGregor>
I'm betting on the first one.

[12:24:00.0000] <timeless>
/me too

[12:24:01.0000] <timeless>
well

[12:24:02.0000] <timeless>
i suppose if you were placing money, i'd just do it and take your money :)

[12:24:03.0000] <AryehGregor>
It won't waste much of anyone's time if someone wants to escalate a FIXED resolution.  I can spend twenty minutes to write a counter-CP if someone escalates.

[12:28:00.0000] <MikeSmith>
AryehGregor: not too snarky, IMHO

[12:28:01.0000] <AryehGregor>
Okay, good.

[12:29:00.0000] <timeless>
MikeSmith: fwiw, the w3 gym bag came in handy

[12:29:01.0000] <timeless>
i unpacked it yesterday and started using it today :)

[12:30:00.0000] <timeless>
i also unpacked the w3 kitchen thing...

[12:30:01.0000] <timeless>
and put it in my kitchen..

[12:30:02.0000] <timeless>
s/thing/apron/

[12:30:03.0000] <MikeSmith>
ah

[12:30:04.0000] <timeless>
we'll see if it actually gets used

[12:30:05.0000] <timeless>
so far i have a newly purchased toaster-oven and 0 free counter space (and i haven't unpacked my kitchen items from my move)

[12:31:00.0000] <timeless>
cooking for me will be mostly toasting breakfast and baking dinners :)

[12:31:01.0000] <MikeSmith>
heh

[12:31:02.0000] <MikeSmith>
you can do a lot with a toaster oven man

[12:31:03.0000] <MikeSmith>
I do at least

[12:32:00.0000] <MikeSmith>
I don't have a real oven

[12:32:01.0000] <MikeSmith>
few people in Japan do

[12:39:00.0000] <AryehGregor>
MikeSmith, because there's too little space?

[12:40:00.0000] <MikeSmith>
yeah, in part at least

[12:41:00.0000] <MikeSmith>
but also because how often do you need to roast a whole turkey or whatever?

[12:42:00.0000] <timeless>
when you want to heat your house or just run up your electric, gas or oil bill?

[12:42:01.0000] <timeless>
iirc one of my friend's experiences was indeed the gas side of things

[12:42:02.0000] <MikeSmith>
big-oven cooking, you have to do the whole pre-heating thing, which is nuts

[12:42:03.0000] <timeless>
even toaster ovens need preheating

[12:43:00.0000] <timeless>
(if you're baking)

[12:43:01.0000] <timeless>
(not if you're toasting or broiling)

[12:43:02.0000] <MikeSmith>
yeah, but it takes a lot less time of course

[12:44:00.0000] <timeless>
yeah

[12:44:01.0000] <timeless>
and as i said, it heats your whole house

[12:44:02.0000] <timeless>
(and it's much easier to burn yourself that way)

[12:45:00.0000] <timeless>
otoh, things really are different if you're cooking for 4+ v. 1-2

[12:45:01.0000] <timeless>
in helsinki, i'd have gladly traded my Oven for a toaster-oven and an automatic dishwasher

[12:46:00.0000] <timeless>
(well, really traded the oven for the dishwasher and added a separate toaster-oven)

[12:47:00.0000] <AryehGregor>
Hixie, there's no diff given at <http://www.w3.org/Bugs/Public/show_bug.cgi?id=13024>.

[12:47:01.0000] <timeless>
'13024>' is not a valid bug number.

[12:47:02.0000] <timeless>
i think the diffs are supposedly generated by a bot later

[12:47:03.0000] <AryehGregor>
Use a client that has better URL detection heuristics.  > is not valid in URLs per RFC.

[12:47:04.0000] <timeless>
i've never actually seen them

[12:47:05.0000] <AryehGregor>
Yes, but it involves some manual step that Hixie sometimes messes up.

[12:48:00.0000] <zewt>
it's not customary to <> urls on IRC

[12:48:01.0000] <timeless>
well, i seriously haven't ever seen those diffs appearing

[12:48:02.0000] <timeless>
and i do get the resolved mails

[13:27:00.0000] <Hixie>
zcorpan: back when i was editing the spec, it had Origin: client-to-server, and Sec-WebSocket-Origin server-to-client

[13:28:00.0000] <AryehGregor>
This seems to be broken: http://aryeh.name/tests-root/tests/submission/AryehGregor/reflection/reflection-forms.html

[13:28:01.0000] <Hixie>
zcorpan: because Origin: is only defined for client-to-server, and because it was another way to prevent echo servers getting p0wned

[13:29:00.0000] <Hixie>
AryehGregor: yeah, no diff for css changes

[13:29:01.0000] <AryehGregor>
Oh.

[13:29:02.0000] <Hixie>
anyone interested in taking over the pdf generation?

[13:40:00.0000] <AryehGregor>
So in Chrome 15, when I close all windows and reopen one, I no longer see a way to restore the tabs I had open in any window other than the last one closed.

[13:40:01.0000] <AryehGregor>
Sigh?

[13:40:02.0000] <AryehGregor>
I was relying on that.

[13:40:03.0000] <AryehGregor>
Oh, wait.

[13:40:04.0000] <AryehGregor>
They moved it to the lower right corner.

[13:41:00.0000] <AryehGregor>
Works for me.

[13:55:00.0000] <timeless>
heh

[14:23:00.0000] <jgraham>
I assume the correct answer to "how often do you cook turkey" is "never, turkey is the blandest of meats and therefore not worth wasting your life on"

[14:23:01.0000] <TabAtkins>
jgraham: You have never had a correctly-cooked turkey, I see.

[14:23:02.0000] <TabAtkins>
When done right, turkey is moist and flavorful and great.

[14:24:00.0000] <jgraham>
TabAtkins: I am rather confident that there is nothing you can do to turkey to get over the basic problem that it is exceptionally bland meat

[14:24:01.0000] <TabAtkins>
Look up Alton Brown's turkey recipes - the guy is a genius and his method has 100% success with everyone I know who's tried it.

[14:24:02.0000] <TabAtkins>
jgraham: You're confidently wrong, then.

[14:24:03.0000] <Hixie>
where do you stand on chicken?

[14:24:04.0000] <Philip`>
Standing on chickens is cruel

[14:24:05.0000] <Hixie>
(just trying to calibrate so i know how to respond)

[14:25:00.0000] <jgraham>
Chicken is OK if you find a decent one

[14:25:01.0000] <Philip`>
Oh, chicken, not chickens

[14:25:02.0000] <jgraham>
Standing on chickens is what happens in inensive farming :(

[14:25:03.0000] <Hixie>
ok i'm with tab then

[14:25:04.0000] <jgraham>
*intensive

[14:26:00.0000] <jgraham>
I remain skeptical that one could make turkey as nice as e.g. goose

[14:26:01.0000] <jgraham>
One simply has a better raw ingredient

[14:27:00.0000] <TabAtkins>
Never had goose, but I've had duck.  Are they at all similar?

[14:27:01.0000] <jgraham>
Somewhat

[14:27:02.0000] <jgraham>
It is certainly the closest comparison I can think of

[14:27:03.0000] <TabAtkins>
Then I have personally cooked turkey that is as good as duck in a high-quality restaurant (from a pure meat perspective).

[14:31:00.0000] <jgraham>
Then I will withhold final judgement until I have tried this miracle recipe. But in the meantime I will continue to work under the hypothesis that the anglophone tradition of turkey for celebratory meals is a sad loss for gastronomy

[14:31:01.0000] <TabAtkins>
jgraham: Come to my house for Christmas.

[14:32:00.0000] <TabAtkins>
I'm not disagreeing with your hypothesis, though - most people cook *terrible* turkey.

[14:32:01.0000] <jgraham>
That would make it hard to cook Christmas dinner at my house. Unless Christmas is like pixar releases and happens earlier in the US than anywhere else in the world

[14:33:00.0000] <jgraham>
(I seriously recommend trying goose though. Properly cookd it can be delicious)

[14:35:00.0000] <timeless>
jgraham: the reason for turkey is that it's bigger and cheaper and feeds more

[14:35:01.0000] <timeless>
geese are smaller, more expensive, take more effort and feed fewer

[14:35:02.0000] <jgraham>
timeless: I know

[14:36:00.0000] <jgraham>
Although the "take more effort" is wrong

[14:36:01.0000] <timeless>
i think i've had duck once or twice at most, it's nice, but kind of like fish, rather delicate

[14:36:02.0000] <jgraham>
Unless you are keeping them yourself, perhaps

[14:36:03.0000] <timeless>
jgraham: more effort is wrt total effort to prepare sufficient to feed the same number of people

[14:36:04.0000] <jgraham>
(although I understand geese are not that hard to keep)

[14:37:00.0000] <Ms2ger>
/me wonders if he should read Hixie's latest commit message as "Pretty iffy"

[14:38:00.0000] <jgraham>
Possibly I am just so antisocial that I never need to feed more than about 8-10 people

[14:38:01.0000] <Ms2ger>
8-10? You must be extremely social

[14:38:02.0000] <timeless>
how many geese?

[14:39:00.0000] <jgraham>
timeless: I'm pretty dure I have done 8 people with one largish goose

[14:39:01.0000] <jgraham>
*sure

[14:40:00.0000] <jgraham>
Yes, the FAQ at http://www.clerkesgeese.com/FAQs.html agrees with me there

[14:41:00.0000] <jgraham>
Of course if US home portions are in the same ratio as US restaurant portions compared to european ones, you can probably divide by 1.5-2

[14:42:00.0000] <jgraham>
/me assumes they are not

[14:42:01.0000] <TabAtkins>
I dunno, it's probably closer than not.

[14:42:02.0000] <TabAtkins>
Though I've never eaten a euro home-cooked meal.

[14:46:00.0000] <jgraham>
If you are ever in the same toen as me then we will have to change that

[14:46:01.0000] <jgraham>
*town

[14:46:02.0000] <jgraham>
sigh

[14:47:00.0000] <Hixie>
jgraham: dude if i ever have to feed more than like 5 people i order pizza

[14:49:00.0000] <timeless>
heh

[14:49:01.0000] <jgraham>
Swedish pizza is... interesting. Kebab meat and bearnaise sauce is a popular topping

[14:51:00.0000] <moo-_->
/me in Sweden right now

[14:51:01.0000] <moo-_->
jgraham: I prefer Kebab, bluecheese, pineapple and mayonase

[15:00:00.0000] <jgraham>
ugh

[15:00:01.0000] <jgraham>
:)

[15:01:00.0000] <Hixie>
can you not animate box-shadow with -webkit-transition?

[15:01:01.0000] <Hixie>
can anyone come up with something you might animate based on details[open] ?

[15:05:00.0000] <jgraham>
I assume the height of the bit that appears isn't the answer you are looking for here

[15:05:01.0000] <AryehGregor>
/me votes for height

[15:06:00.0000] <TabAtkins>
Hixie: You should be able to - <shadow> is a transitionable type.

[15:07:00.0000] <TabAtkins>
jgraham: That pizza sounds pretty good, actually.

[15:07:01.0000] <Hixie>
am i doing it wrong? http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1105

[15:08:00.0000] <gsnedders>
TabAtkins: You're just wrong.

[15:09:00.0000] <TabAtkins>
gsnedders: About the pizza or the shadow?

[15:09:01.0000] <gsnedders>
TabAtkins: the pizza.

[15:09:02.0000] <Hixie>
about the shadow as well as far as i can tell :-P

[15:09:03.0000] <gsnedders>
TabAtkins: Why would I care about vaguely on-topic discussion? :)

[15:09:04.0000] <TabAtkins>
Hixie: Ah, I see.  While <shadow> is an animatable type, the spec currently only defines text-shadow as an animatable property.  I assume that's because box-shadow was dropped for a little while.

[15:10:00.0000] <TabAtkins>
/me could probably fix that.

[15:10:01.0000] <Hixie>
lame

[15:10:02.0000] <TabAtkins>
Ah, there's even a note that the list omits box-shadow.

[15:10:03.0000] <Hixie>
you got any other ideas of what to animate? 'height' like AryehGregor suggested seems reasonable but hard to do properly since we can't animate to/from 'auto' and since the whole point of <details> is to change the height so UAs are probably already doing it.

[15:11:00.0000] <TabAtkins>
Hixie: Rotate the marker.

[15:12:00.0000] <Hixie>
isn't that something that UAs should do also?

[15:12:01.0000] <Hixie>
maybe i could animate the background or something

[15:12:02.0000] <Hixie>
fade in the details

[15:12:03.0000] <Hixie>
opacity, that might work

[15:12:04.0000] <annevk>
opacity:0 > 1

[15:12:05.0000] <annevk>
or the other way around

[15:13:00.0000] <annevk>
I always forget what is transparent

[15:13:01.0000] <Hixie>
i guess i have to animate a div inside it

[15:13:02.0000] <TabAtkins>
annevk: Fully opaque things aren't transparent.

[15:15:00.0000] <jgraham>
It can't be said too often: not being able to transition to auto height is lame

[15:16:00.0000] <TabAtkins>
Yus.

[15:16:01.0000] <jgraham>
Also, so is pizza with bearnaise sauce

[15:17:00.0000] <eighty4>
Sorry for the nick changes earlier

[15:34:00.0000] <Hixie>
TabAtkins: dude this don't work neither! http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1106

[15:35:00.0000] <jamesr>
Hixie: that link tries to animate box-shadow

[15:35:01.0000] <Hixie>
no?

[15:36:00.0000] <Hixie>
  details div { -webkit-transition: opacity 0.5s ease; opacity: 0; }   details[open] div { box-shadow: black 0 0 1em; opacity: 1; }

[15:36:01.0000] <Hixie>
it has a shadow appear, sure

[15:36:02.0000] <Hixie>
but the opacity is what is changed dynamically

[15:36:03.0000] <jamesr>
ah, i see

[15:37:00.0000] <annevk>
prolly a bug?

[15:37:01.0000] <Hixie>
yeah i guess you just can't animate with <details> in webkit

[15:37:02.0000] <Hixie>
anyone else implement it?

[15:38:00.0000] <annevk>
don't think so

[15:39:00.0000] <Hixie>
my opera build won't load live dom viewer, wtf

[15:51:00.0000] <TabAtkins>
Hixie: Oh, actually, yeah, that's probably true.  Our <details> impl is pretty hacky.  I think we do some manual rendertree hacking, so it's not surprising if some things break.

[16:20:00.0000] <AryehGregor>
Sigh.

[16:20:01.0000] <AryehGregor>
So Apple wants the spec at the W3C, and people are in fact going to fork it if I don't publish it there.

[16:20:02.0000] <AryehGregor>
All right, I lose.

[16:21:00.0000] <AryehGregor>
Or at any rate, I give up.  W3C it is.

[16:21:01.0000] <AryehGregor>
No arguing with major implementers.

[16:22:00.0000] <annevk>
offlist?

[16:22:01.0000] <annevk>
and does "fork" mean publish?

[16:23:00.0000] <annevk>
doesn't seem so bad to me

[16:27:00.0000] <TabAtkins>
Yeah, they'll do your work for you.  ^_^

[16:30:00.0000] <annevk>
found the relevant bug

[16:30:01.0000] <annevk>
should get ourselves a pp

[16:32:00.0000] <annevk>
AryehGregor, also, have fun getting the WebApps WG recharter to take on that work item :p

[16:45:00.0000] <mercator>
Does HTML5 define somewhere how HTML tags are supposed to interact with context-sensitivity in Unicode characters?

[16:45:01.0000] <mercator>
I.e. http://stackoverflow.com/questions/7069247/css-formating-cursive-word-breaks-it

[16:46:00.0000] <TabAtkins>
mercator: That's a CSS issue, actually.

[16:47:00.0000] <Philip`>
I thought at least Firefox was able to apply colours to characters without breaking apart ligatures etc

[16:48:00.0000] <TabAtkins>
Should be able to, yes.

[16:48:01.0000] <Philip`>
My Firefox on Linux does do that with that example

[16:48:02.0000] <Philip`>
(Opera doesn't)

[16:50:00.0000] <mercator>
TabAtkins: Is it? It does mention things like right-to-left embedding. Isn't this similar? The actual CSS being applied doesn't really matter. Should a tag break ligatures?

[16:51:00.0000] <TabAtkins>
mercator: In general, the answer is no, it shouldn't.  Whether or not implementations support it properly is a different question.

[16:52:00.0000] <annevk>
we don't define things in that detail

[16:52:01.0000] <annevk>
it's implementation quality atm

[16:52:02.0000] <annevk>
much like line-breaking

[16:55:00.0000] <mercator>
Hmmm, I suppose character selection in fonts *is* more CSS than HTML... And indeed, implementations are rather different.


2011-08-16
[17:02:00.0000] <roc>
Anne's right of course, but this situation is a bit different from line breaking in that with line breaking, there often is no clearly "right" algorithm, but it's pretty clear that shaping across element boundaries *should* work

[17:09:00.0000] <mercator>
That might not always be the case, either, if, say, you float one part of the ligature or something... (But that is decidedly CSS.)

[17:11:00.0000] <roc>
sure there are edge cases, but when it's obvious that it should work, it should work :-)

[17:14:00.0000] <annevk>
I wish writing specs worked like that

[17:14:01.0000] <annevk>
Invoking X must do what is obvious for Y

[17:15:00.0000] <zewt>
step 1: just do it

[17:23:00.0000] <roc>
I'm not suggesting specs should be written like that, of course. Just that implementation quality should be judged that way when specs are silent

[17:35:00.0000] <mercator>
Thanks for the insights!

[20:09:00.0000] <boblet>
foolip: know when microdata will be in a public (non-dev) release of Opera? I’m guessing Opera 12, but not sure if that’s imminent or a wee way off yet… (and yes that does mean I’m updating a certain article ;)

[20:12:00.0000] <boblet>
also, any Moz ppl know about the progress of microdata ref: https://bugzilla.mozilla.org/show_bug.cgi?id=591467 ? also guessing WIP and not imminent…

[21:03:00.0000] <roc>
boblet: looks like it just needs review ... that could happen anytime, and then it would show up in nightlies within a day or three

[21:04:00.0000] <boblet>
roc: thanks for the explanation

[02:57:00.0000] <smaug____>
and yet another wontfix/reopen/wontfix/reopen cycle coming...

[03:02:00.0000] <janv_>
you mean bug 12945 ?

[03:03:00.0000] <janv_>
ah

[03:03:01.0000] <janv_>
I see

[03:03:02.0000] <janv_>
just got the bugmail

[03:04:00.0000] <smaug____>
janv_: yep

[03:05:00.0000] <annevk>
prolly more productive to try to get input from other implementors than argue in that bug

[03:06:00.0000] <smaug____>
yeah, getting input from implementors would be great

[03:06:01.0000] <smaug____>
randomly just wontfixing bugs isn't very productive

[03:07:00.0000] <annevk>
I meant that as advice to you

[03:07:01.0000] <smaug____>
:)

[03:07:02.0000] <annevk>
he's not randomly wontfixing, he believes that design is appropriate

[03:07:03.0000] <annevk>
if you can find several implementors that disagree you can resolve it towards a design you find appropriate

[03:25:00.0000] <foolip>
boblet, it ought to be in the next stable release, yes

[03:26:00.0000] <Ms2ger>
AryehGregor, you know why ranges' offsets are signed?

[03:29:00.0000] <annevk>
are they in browsers?

[07:47:00.0000] <AryehGregor>
annevk, the charter says the WebApps WG can take on stuff spun off from HTML5 without rechartering, if I read correctly.

[07:48:00.0000] <AryehGregor>
Ms2ger, I have no idea.  Do you mean specifically .startOffset and .endOffset, or also parameters to various methods like setStart()?

[07:48:01.0000] <AryehGregor>
I can't see how it would be black-box detectable for .startOffset or .endOffset.

[07:49:00.0000] <jgraham>
/me really thinks the WebApps charter should read "any DOM API that two or more browser vendors are interested in implementing"

[07:49:01.0000] <AryehGregor>
Actually, I don't see how it would be black-box detectable for .setStart() either, unless you have a node with about 2G children or character elements.

[07:50:00.0000] <AryehGregor>
Because if it were unsigned, a negative number would wrap around to a large positive number, which winds up being INDEX_SIZE_ERR either way.

[07:50:01.0000] <AryehGregor>
So how about we just make them all unsigned?

[07:50:02.0000] <AryehGregor>
jgraham, I really think there shouldn't be working groups and charters.  :)

[07:51:00.0000] <jgraham>
Well yes, taht would be fine too

[07:52:00.0000] <jgraham>
But seems less likely in the short term

[07:52:01.0000] <AryehGregor>
Sadly.

[07:53:00.0000] <annevk>
does seem like there needs to be AC review

[07:53:01.0000] <annevk>
at the very least

[07:54:00.0000] <annevk>
anyway, all that does not matter that much long term

[07:54:01.0000] <AryehGregor>
Why not?

[07:55:00.0000] <annevk>
well it's annoying, but mostly to people who actually have to sit through it

[07:55:01.0000] <annevk>
it just delays publishing a WD

[07:56:00.0000] <annevk>
http://www.w3.org/QA/2011/08/subject_from_innovation_to_sta.html

[07:57:00.0000] <smaug____>
AryehGregor: in which way should the standards be written if not in some kind of WGs

[07:57:01.0000] <smaug____>
(just curious)

[07:57:02.0000] <annevk>
community groups have better patent protection (at until REC, which hardly any spec reaches these days) and better licensing...

[07:58:00.0000] <AryehGregor>
smaug____, in some format where there's no formal organization or membership at all.  I mean, you'd want separate mailing lists and Bugzilla components and such, but no actual organization divisions are needed.

[07:59:00.0000] <AryehGregor>
annevk, but will real specs be accepted as community groups?  Worth trying . . .

[07:59:01.0000] <smaug____>
AryehGregor: might work, or might not

[08:00:00.0000] <annevk>
AryehGregor, as far as I know moving specs from WGs to CGs is somewhat problematic (though might be possible)

[08:00:01.0000] <AryehGregor>
smaug____, the WHATWG seems to have worked pretty well so far.  We probably wouldn't even have to deal with the W3C at all if not for patent policies.

[08:00:02.0000] <annevk>
AryehGregor, new specs such as DOM Parsing should be fairly trivial however

[08:00:03.0000] <smaug____>
AryehGregor: WhatWG is a WG ;)

[08:00:04.0000] <AryehGregor>
annevk, but what's the point?  Is it going to make Apple or Microsoft happier to have it in a community group instead of some random website?  If so, okay.

[08:01:00.0000] <AryehGregor>
I'm game then.

[08:01:01.0000] <smaug____>
in whatever form the standardization happens, there is always the problem what to do when there are conflicting opinions/interest

[08:01:02.0000] <AryehGregor>
smaug____, not in the sense I was talking about -- there's no charter or membership or anything like that.

[08:01:03.0000] <AryehGregor>
If there are conflicting opinions on implementer requirements, implementers can work it out and specs will follow if they're being written sanely, because people want interop.

[08:01:04.0000] <annevk>
AryehGregor, a) you have more patent protection than when you publish it as W3C WD; b) makes it easier to move it to REC within a WG when finished as I understand it

[08:02:00.0000] <AryehGregor>
If the implementers can't work it out, like storage mutex or whatever, leave it undefined.

[08:02:01.0000] <AryehGregor>
None of this chartering stuff is needed.

[08:02:02.0000] <annevk>
WHATWG has a charter

[08:02:03.0000] <AryehGregor>
annevk, okay, I'll be sure to run that by everyone.  I don't mind having it at the W3C as long as it's in some place where I don't have to deal with the bureaucracy.

[08:02:04.0000] <annevk>
and has some kind of committee

[08:02:05.0000] <AryehGregor>
Not in practice.

[08:03:00.0000] <annevk>
also in practice, just not that visible

[08:05:00.0000] <AryehGregor>
Not in practice, insofar as no one pays any attention to them and most people who participate heavily in the WHATWG haven't even heard of their existence.

[08:05:01.0000] <AryehGregor>
If they have any significance, it's as documentation of what we'd all do anyway even if it weren't officially written down anywhere.

[08:05:02.0000] <Ms2ger>
So the difference is that the WHATWG's committee works in secret and the HTMLWG's in public? :)

[08:06:00.0000] <annevk>
I meant that we don't interfere much as it's not deemed necessary

[08:06:01.0000] <AryehGregor>
You'd have about as much say whether or not you were officially part of some committee.

[08:07:00.0000] <AryehGregor>
The reason you have say is really because you're implementers, not because you're on the steering committee.

[08:07:01.0000] <annevk>
if you say so

[08:07:02.0000] <AryehGregor>
When was the last time Hixie did anything he didn't want to do because the steering committee told him to?

[08:07:03.0000] <AryehGregor>
More to the point, what practical enforcement mechanism is there if the steering committee wants to tell Hixie what to do?

[08:08:00.0000] <Ms2ger>
Kick him out

[08:08:01.0000] <AryehGregor>
He owns the domain name, he can do whatever he feels like.  At the W3C, the Team holds the cards.

[08:08:02.0000] <AryehGregor>
Or whatever, the W3C staff.

[08:08:03.0000] <AryehGregor>
Yeah, right, so we'd have the WHATWG operating without Hixie.  I really see that happening.

[08:10:00.0000] <annevk>
/me goes back to work on mutation events

[08:10:01.0000] <smaug____>
annevk: what are you doing to them?

[08:11:00.0000] <smaug____>
/me goes back to implement the replacement

[08:11:01.0000] <annevk>
sorting out the arguments for end of task and return of outermost method

[08:11:02.0000] <annevk>
and fleshing them out a bit better

[08:12:00.0000] <annevk>
the end of task arguments are somewhat compelling I think, just not that clear currently

[08:14:00.0000] <smaug____>
the end of task suffers badly from the case that one would need to be very careful to not do certain modal stuff in the same task as dom mutations

[08:14:01.0000] <annevk>
is there anything besides showModalDialog?

[08:14:02.0000] <smaug____>
the spec does have some other case

[08:15:00.0000] <smaug____>
IIRC

[08:15:01.0000] <annevk>
the after modification suffers badly from the same problem as the current mutation events, except the problem now becomes that of library authors rather than UA implementors

[08:16:00.0000] <smaug____>
anyway, it should be trivial to change the current implementation from "outermost method" to "end of task"

[08:16:01.0000] <smaug____>
though, it is not even clear what kind of API people want

[08:16:02.0000] <annevk>
yeah, your API would work either way

[08:17:00.0000] <annevk>
yeah, Rafael had that as open issue (6/7 or so?) but never elaborated

[08:17:01.0000] <annevk>
I will try to push a bit

[08:18:00.0000] <smaug____>
so far web devs have managed to work with mutation events quite well. "outermost method" doesn't really make it any worse

[08:20:00.0000] <smaug____>
annevk: also, the proposed <dialog> would cause similar problems as showModalDialog

[08:21:00.0000] <smaug____>
hmm, actually, the proposal isn't clear what showModal does

[08:51:00.0000] <annevk>
smaug____, btw, Rafael says sicking also thinks end of task is better

[08:51:01.0000] <annevk>
and apparently Indexed DB has such an "end of task" concept already?

[08:52:00.0000] <jgraham>
/me now plans to tell someone that annevk claims that Rafael said that sicking thinks end of task is better

[08:53:00.0000] <annevk>
just wondering whether smaug____ knows or not

[08:54:00.0000] <jgraham>
:)

[08:55:00.0000] <Ms2ger>
/me will alert the world that jgraham told someone that... Meh, it's boring already

[08:59:00.0000] <zcorpan>
it wasn't boring until you ruined it

[09:00:00.0000] <Ms2ger>
zcorpan, sorry :(

[09:01:00.0000] <smaug____>
annevk: I do know that

[09:01:01.0000] <smaug____>
I disagree with sicking

[09:01:02.0000] <annevk>
what happened to whatwg.org

[09:02:00.0000] <smaug____>
?

[09:02:01.0000] <smaug____>
works here

[09:02:02.0000] <smaug____>
er, no

[09:02:03.0000] <smaug____>
not anymore

[09:02:04.0000] <annevk>
server disappeared?

[09:03:00.0000] <jgraham>
Gone for me too :(

[09:04:00.0000] <jgraham>
The server bus factor of web standards work is way too low

[09:04:01.0000] <jgraham>
I mean as comforting as it is that Hixie has everything backed up in eleventy bajillion places it doesn't help when his host goes down

[09:05:00.0000] <annevk>
AryehGregor, ECMAScript uses "code unit" and "character"

[09:05:01.0000] <AryehGregor>
Does it?

[09:06:00.0000] <AryehGregor>
"The String type is the set of all finite ordered sequences of zero or more 16-bit unsigned integer values (“elements”)."

[09:06:01.0000] <annevk>
oh only in source text

[09:06:02.0000] <AryehGregor>
It looks like it uses "code unit" to mean something totally different.

[09:06:03.0000] <AryehGregor>
Oh, it uses it for multiple things.

[09:06:04.0000] <AryehGregor>
How confusing.

[09:06:05.0000] <AryehGregor>
Yeah, looks like only for source text.

[09:06:06.0000] <annevk>
does element also mean escapes?

[09:06:07.0000] <annevk>
because you don't want that

[09:07:00.0000] <annevk>
'Throughout the rest of this document, the phrase “code unit” and the word “character” will be used to refer to a 16-bit unsigned value used to represent a single 16-bit unit of text.'

[09:07:01.0000] <annevk>
that seems to be what we want

[09:08:00.0000] <AryehGregor>
So it looks like Steve Faulkner has put up a clone for me: http://dev.w3.org/html5/editing-api/Overview.html

[09:08:01.0000] <AryehGregor>
Now, I find this bit very interesting: "Copyright © 2011 W3C© (MIT, ERCIM, Keio), All Rights Reserved. W3C liability, trademark and document use rules apply."

[09:08:02.0000] <AryehGregor>
That strikes me as actively misrepresenting the copyright status of the document.

[09:09:00.0000] <annevk>
back online

[09:31:00.0000] <david_carlisle>
AryehGregor: Is the original copyright status consistent? It says copyriight google, but released under CC0 which the CC0 page to which you link has the heading "no copyright"

[09:31:01.0000] <AryehGregor>
david_carlisle, AFAIK (IANAL), the copyright holder is Google, but by licensing it CC0 all actual rights are waived.  Where does it say copyright Google, though?  I thought I removed that.

[09:32:00.0000] <david_carlisle>
er I was looking at this which might not be the right thing

[09:32:01.0000] <david_carlisle>
http://aryeh.name/gitweb.cgi?p=editing;a=blob;f=LICENSE;h=85cb5c8689f2c59cd6b40e87c7b608b303d09294;hb=HEAD

[09:32:02.0000] <AryehGregor>
Oh, I missed that.

[09:33:00.0000] <AryehGregor>
david_carlisle, try now: http://aryeh.name/gitweb.cgi?p=editing;a=blob;f=LICENSE;hb=HEAD

[09:33:01.0000] <david_carlisle>
AryehGregor: I think publishing as a w3c spec has a w3c copyright as a rerequisite 9although I must adnit it's a while since I read those rules:-)

[09:33:02.0000] <AryehGregor>
david_carlisle, pubrules require that line to be present.

[09:34:00.0000] <AryehGregor>
But it's not actually submitted anywhere as REC-track just now.

[09:35:00.0000] <david_carlisle>
AryehGregor: yes but presumably the only point of copying it to w3c woul dbe to put it on rec trac (so if you were to object to a copyright change or anything else, better to get those things sorted earlier rather than later?0

[09:35:01.0000] <AryehGregor>
david_carlisle, well, there are now Community Groups at the W3C.

[09:35:02.0000] <AryehGregor>
I'll see if people are okay with using those.

[09:35:03.0000] <AryehGregor>
I mean, it'll have to be REC track eventually if we want the patent policy.

[09:35:04.0000] <AryehGregor>
But doesn't have to be yet.

[09:36:00.0000] <david_carlisle>
well could stay in your git hosting until then;-)

[09:36:01.0000] <AryehGregor>
Except people will fork it if I do that.

[09:36:02.0000] <AryehGregor>
Which is fair enough, since I basically told them to.  :)

[09:39:00.0000] <zewt>
this is ... odd

[09:39:01.0000] <Ms2ger>
What? AryehGregor not being openly hostile to the W3C? :)

[09:40:00.0000] <zewt>
the "quotes" in mark watson's mails show up in a very slightly different color than his text (red vs. black), which I assumed was his mailer being horrible

[09:40:01.0000] <zewt>
... but when I view-original-message, there's no HTML part in it at all--is it Gmail somehow doing that? or is "view original" not actually showing the real original data?

[09:40:02.0000] <AryehGregor>
Ms2ger, well, maybe I am openly hostile, but if people are going to try publishing outdated forks of my spec . . .

[09:41:00.0000] <Ms2ger>
Can't have that

[09:41:01.0000] <zewt>
anyone want to tell me what his mails look like in another mailer and let me know the manner in which I'm going crazy? (eg. the last post to the MediaSource thread)

[09:43:00.0000] <Philip`>
zewt: They look normal colour to me, in Gmail

[09:43:01.0000] <Philip`>
Does it still look red if you zoom in so the text is larger?

[09:43:02.0000] <Philip`>
(Maybe it's an artifact of subpixel font rendering or something)

[09:44:00.0000] <jgraham>
Do you mean quotes from the message he is replying to?

[09:44:01.0000] <zewt>
https://zewt.org/~glenn/weird.png

[09:44:02.0000] <jgraham>
I can't tell which is which

[09:44:03.0000] <zewt>
i have to squint to even tell the difference, but it's there

[09:45:00.0000] <zewt>
... but when I "show original", it's just a regular text/plain message

[09:46:00.0000] <Philip`>
If you mean quote marks then they look normal, if you mean quotations from earlier messages in a purplish colour then Gmail does that by comparing lines of text to lines in previous messages in that conversation

[09:46:01.0000] <zewt>
that's ... horribly disgustingly evil

[09:46:02.0000] <zewt>
and with a nearly-indistinguishable color? even worse

[09:46:03.0000] <Philip`>
The colour difference for quoted text looks clearly distinct to me

[09:47:00.0000] <zewt>
it's very subtle

[09:47:01.0000] <jgraham>
Well so is randomly interspersing your content amonst the content you are replying to with no plain text distinguishing marks

[09:47:02.0000] <Philip`>
Surely it's less evil than displaying the message with no indication of what's a quote and what's a reply?

[09:47:03.0000] <zewt>
well yeah, i was originally sending him a mail to ask him to quote normally when I was confused by that

[09:47:04.0000] <jgraham>
I thought only a11y people did that

[09:48:00.0000] <jgraham>
But it seems I was wrong

[09:49:00.0000] <dglazkov>
good morning, Whatwg!

[09:49:01.0000] <zewt>
Philip`: it encourages people to think that people can read messages that are quoted like that, encouraging people like mark to continue quoting in that horrible way

[09:50:00.0000] <Philip`>
zewt: It doesn't look like it was posted via Gmail, so I would assume the poster isn't going to be influenced by how Gmail renders it

[09:50:01.0000] <Philip`>
(until Gmail users complain at him)

[09:50:02.0000] <zewt>
didn't say he was

[09:51:00.0000] <Philip`>
(and Gmail's colouring will mean a smaller fraction of Gmail users will complain)

[09:51:01.0000] <Philip`>
(which seems like a pretty indirect form of encouragement)

[09:51:02.0000] <zewt>
thus proving my point :)

[09:51:03.0000] <Philip`>
(and likely oughtweighed by the convenience it provides to Gmail users)

[09:51:04.0000] <Philip`>
(since some people will format emails stupidly regardless of how much you complain)

[09:52:00.0000] <zewt>
it's not much use to me; as soon as I try to reply to it the questionable distinction is lost (it doesn't transfer to reply quotes in any way)

[09:52:01.0000] <jgraham>
It appears to be posted from exchange

[09:52:02.0000] <jgraham>
Which is mildly surprising

[09:52:03.0000] <jgraham>
I was expecting lotus notes

[09:52:04.0000] <jgraham>
But only very mildly

[09:52:05.0000] <Ms2ger>
zewt, my reply is colored green

[09:53:00.0000] <zewt>
Ms2ger: heh I hate when people do that--but at least they're distinguishing *somehow*

[09:53:01.0000] <AryehGregor>
I'd actually prefer to use a W3C Community Group than to host this on my website, on reflection, because hopefully it will push the W3C to be less dysfunctional about its process requirements if it sees that people don't mind working there if we don't have to deal with lots of crazy pointless rules.

[09:53:02.0000] <zewt>
(provided that the receiver sees HTML mail, that he hasn't disabled colors in HTML mail, and that green is readable on whatever his background happens to be)

[09:54:00.0000] <Philip`>
Green is an environmentally friendly colour, so I always give emails a solid green background before printing them

[09:54:01.0000] <timeless>
heh

[10:06:00.0000] <AryehGregor>
Okay, I need four people with W3C accounts to support my new Community Group so it can be approved: http://www.w3.org/community/groups/proposed/

[10:06:01.0000] <AryehGregor>
Get to it!

[10:06:02.0000] <annevk>
oh Firefox 6 is out

[10:06:03.0000] <AryehGregor>
/me waits to see how many minutes it will take

[10:07:00.0000] <annevk>
does not work in Opera?

[10:07:01.0000] <zewt>
only difference i've seen in ff6 is ... randomly shuffled menu items breaking a bunch of my menu habits

[10:07:02.0000] <AryehGregor>
What doesn't, the W3C page?

[10:07:03.0000] <AryehGregor>
That's amusing.

[10:07:04.0000] <zewt>
(havn't checked into what API updates they've made yet)

[10:08:00.0000] <jgraham>
Nope, teh page is broken in Opera

[10:08:01.0000] <annevk>
AryehGregor, yeah, the big button does not appear

[10:08:02.0000] <AryehGregor>
lulz.

[10:08:03.0000] <jgraham>
Gotta love that

[10:09:00.0000] <jgraham>
AryehGregor: In firefox the links in the description seem to be missing

[10:09:01.0000] <zcorpan>
w3c hates opera!

[10:10:00.0000] <zcorpan>
/me goes to bbq

[10:10:01.0000] <Ms2ger>
Two more

[10:11:00.0000] <annevk>
I cheated by using a different browser

[10:14:00.0000] <zewt>
wow, ff6 seems much much slower at incrementally decoding big (print resolution) images

[10:15:00.0000] <Hixie>
the server going down earlier was because of the pdf generation

[10:15:01.0000] <Hixie>
happens every day these days because the pdf generation is taking too many resources

[10:15:02.0000] <smaug____>
zewt: really? Please file a bug

[10:15:03.0000] <Hixie>
i need to disable it really

[10:15:04.0000] <smaug____>
zewt: also, do you have a testcase?

[10:16:00.0000] <zewt>
need to reinstall ff5 to re-test first

[10:16:01.0000] <annevk>
Hixie, that would be better than the server going offline

[10:16:02.0000] <jgraham>
Oh, look at that, the problem is an HTML parsing issue

[10:16:03.0000] <annevk>
Hixie, maybe email whatwg⊙wo to ask if anyone wants to volunteer

[10:16:04.0000] <Philip`>
Hixie: You should make the spec smaller

[10:16:05.0000] <jgraham>
Also, who the hell uses wordpress to design a site like this

[10:16:06.0000] <Hixie>
Philip`: :-P

[10:16:07.0000] <Hixie>
annevk: yeah

[10:17:00.0000] <AryehGregor>
jgraham, cool.

[10:17:01.0000] <Philip`>
jgraham: What's wrong with Wordpress for a site like that?

[10:17:02.0000] <Ms2ger>
ö

[10:17:03.0000] <Ms2ger>
That page uses header

[10:18:00.0000] <annevk>
Ms2ger, only TR/ is outdated

[10:18:01.0000] <jgraham>
Philip`: It seems entirely unlike a blog

[10:18:02.0000] <jgraham>
Also, did you look at the source?

[10:18:03.0000] <Philip`>
"The group\'s deliverables" yay escaping

[10:18:04.0000] <AryehGregor>
We need two more people with a W3C account to signal support here: http://www.w3.org/community/groups/proposed/

[10:18:05.0000] <AryehGregor>
/me prods

[10:18:06.0000] <Philip`>
"It will start work with the preliminary specification hosted at <http://aryeh.name/spec/editing/editing.html>" yay escaping again

[10:18:07.0000] <Ms2ger>
Hixie, ^

[10:19:00.0000] <AryehGregor>
WTF.

[10:20:00.0000] <Philip`>
Try XSSing it to get people to automatically approve your group as soon as they open the page

[10:20:01.0000] <jgraham>
<header class="group tMargin header_article"><a href="#" class="h3"><img src="/community/src/img/icon-minus.png" class="icon_plus_minus" width="14" height="14" alt="icon-minus" />HTML Editing APIs</a></header>

[10:20:02.0000] <jgraham>
Yes, it uses <header>

[10:20:03.0000] <jgraham>
No, it doesn't really get the point of <header>

[10:20:04.0000] <smaug____>
AryehGregor: so what does Community Group mean in practice?

[10:21:00.0000] <AryehGregor>
smaug____, AFAICT, the W3C provides some infrastructure but no Process requirements, and it doesn't produce "official" specs like RECs or anything.

[10:21:01.0000] <AryehGregor>
smaug____, http://www.w3.org/community/about/faq/

[10:22:00.0000] <smaug____>
AryehGregor: and why should editing APIs be there and not in WebApps ?

[10:22:01.0000] <smaug____>
(Sorry, I haven't followed all the politics )

[10:23:00.0000] <AryehGregor>
smaug____, because that way I don't have to deal with the Process, which is nice, because I'm not a masochist.  Eventually it will have to be in WebApps for the Patent Policy.

[10:23:01.0000] <smaug____>
ok

[10:26:00.0000] <smaug____>
AryehGregor: now you need only one more to click "support"

[10:26:01.0000] <zewt>
reminds me of petitions, heh

[10:27:00.0000] <Ms2ger>
zewt, please sign my petition! Or do you *want* to kill puppies?

[10:27:01.0000] <zewt>
absolutely

[10:27:02.0000] <smaug____>
"Oil, Gas and Chemicals Business Group " o_O

[10:28:00.0000] <zewt>
"i got thousands of people to sign a piece of paper! i think one or two of them actually read it"

[10:28:01.0000] <Ms2ger>
smaug____, surprised me as well

[10:31:00.0000] <smaug____>
AryehGregor: I hope you'll  use webapps or whatwg mailing lists for editing stuff

[10:35:00.0000] <jgraham>
Why pass up the oppertunity to start a new mailing list?

[10:35:01.0000] <jgraham>
Everyone knows that N+1 lists are strictly better than N, always

[10:36:00.0000] <zewt>
shh we'll end up in an infinite loop

[10:36:01.0000] <Ms2ger>
N+2*

[10:36:02.0000] <AryehGregor>
smaug____, I intend to, yeah.

[10:37:00.0000] <AryehGregor>
Likewise for wikis etc.

[10:37:01.0000] <Ms2ger>
Uh

[10:37:02.0000] <Ms2ger>
N+4*

[10:37:03.0000] <Ms2ger>
"Each group has four mailing lists by default, two public and two non-public."

[10:37:04.0000] <AryehGregor>
. . .

[10:38:00.0000] <jgraham>
See. The only way that could be improved is if they gave you 5 lists by default

[10:39:00.0000] <Ms2ger>
And CGs must use the public lists for technical stuff, Business Groups don't

[10:42:00.0000] <annevk>
from what I heard you can contact ij and get alternative setups

[10:42:01.0000] <annevk>
e.g. use an existing mailing list or repository

[10:43:00.0000] <annevk>
or bug tracker etc.

[10:43:01.0000] <annevk>
at the moment however that requires manual setup

[10:46:00.0000] <AryehGregor>
We still need one more supporter.

[10:48:00.0000] <nimbu>
foolip: Is this true "And: Unfortunately Opera still lack a very important feature in their HTML5 video spec implementation, which is the Buffer API." ?

[10:48:01.0000] <Ms2ger>
nimbu, surely you have a W3C account

[10:49:00.0000] <nimbu>
Ms2ger: I do >_>

[10:49:01.0000] <Ms2ger>
Want to click a button on http://www.w3.org/community/groups/proposed/?

[10:50:00.0000] <timeless>
AryehGregor: what is `as (1) te contenteditable ` ?

[10:50:01.0000] <AryehGregor>
timeless, a typo, possibly my fault and possibly not.

[10:50:02.0000] <AryehGregor>
(they reformatted it)

[10:51:00.0000] <timeless>
AryehGregor: if you fix it,  you can have my vote:)

[10:51:01.0000] <nimbu>
what button Ms2ger

[10:52:00.0000] <Ms2ger>
The one you get to see after you click "HTML Editing APIs", if timeless doesn't beat you to it

[10:52:01.0000] <annevk>
Ms2ger, sounds like you need to find an XSS exploit lol

[10:52:02.0000] <AryehGregor>
timeless, Ian fixed it.

[10:52:03.0000] <Ms2ger>
Heh

[10:52:04.0000] <AryehGregor>
Ian Jacobs, that is.

[10:52:05.0000] <timeless>
>    This group is no longer proposed, it has been

[10:52:06.0000] <timeless>
it has been what?

[10:52:07.0000] <annevk>
"oh come on, just log in and click the button, you'll get a cookie"

[10:53:00.0000] <timeless>
and i can't click the report a bug link

[10:53:01.0000] <timeless>
because the error page doesn't have it!

[10:53:02.0000] <timeless>
          <h3>             Errors           </h3>                  <ul>   <li>This group is no longer proposed, it has been </li> </ul>

[10:53:03.0000] <annevk>
it's a group

[10:53:04.0000] <annevk>
http://www.w3.org/community/editing/

[10:53:05.0000] <timeless>
it has been... what?

[10:53:06.0000] <Ms2ger>
mailto:team-community-process⊙wo?Subject=Problem reported for editing

[10:54:00.0000] <Philip`>
It's a has been

[10:55:00.0000] <timeless>
oh cute

[10:55:01.0000] <timeless>
the group was on the list of /proposed/ with the error for when i tried to vote for it

[10:56:00.0000] <timeless>
but when i tried to vote for it *again*, the error appeared and the group did not reappear

[10:56:01.0000] <timeless>
/me goes to hunt lunch

[10:56:02.0000] <Ms2ger>
Heh

[10:57:00.0000] <Ms2ger>
AryehGregor, joined your own IRC channel already? :)

[10:57:01.0000] <AryehGregor>
We don't want our own IRC channel.

[10:59:00.0000] <AryehGregor>
I'm sad that the functional methods like .filter and such return arrays instead of iterators.

[10:59:01.0000] <AryehGregor>
It makes things gratuitously inefficient when you chain them, like .filter(...).every(...).

[11:01:00.0000] <gsnedders>
AryehGregor: Well, there aren't iterators in ES, so that's be hard.

[11:01:01.0000] <AryehGregor>
That's a potential issue, I grant.

[11:01:02.0000] <AryehGregor>
There should be, they're totally awesome.

[11:01:03.0000] <AryehGregor>
One of my favorite things in Python.

[11:01:04.0000] <gsnedders>
AryehGregor: Could still have lazily evaluated arrays, though.

[11:01:05.0000] <gsnedders>
So use thunks for filters and the like applied to an array

[11:02:00.0000] <gsnedders>
Somewhat evil, but would be a perf gain when stuff is chained

[12:11:00.0000] <timeless>
anyone here familiar w/ <video> ?

[12:12:00.0000] <timeless>
assuming someone had a <source> which was magical, i presume there's a way for a JS associated to a <video> to discover that the <video> has reached the end of the [stream]

[12:13:00.0000] <timeless>
is there also a way to discover if the <video> is dropping frames?

[12:14:00.0000] <timeless>
it looks like there's onstalled

[12:14:01.0000] <timeless>
and onwaiting

[12:14:02.0000] <timeless>
and onemptied

[12:14:03.0000] <timeless>
/me sighs

[12:14:04.0000] <timeless>
too many events

[12:19:00.0000] <timeless>
ok, so onstalled generally covers one half of the problem

[12:20:00.0000] <timeless>
> This specification doesn't define how the user agent achieves the    appropriate playback rate

[12:21:00.0000] <timeless>
so, there isn't actually a way to find out if the player is dropping frames

[12:21:01.0000] <timeless>
hrm, this consumer has a requirement to know roughly that

[12:21:02.0000] <timeless>
(not precisely that, but roughly that)

[12:21:03.0000] <timeless>
(The consumer is trying to [in JS] automatically select the right content bitrate [instead of letting the browser negotiate for it])

[12:22:00.0000] <jamesr>
mozilla had a vendor-prefixed way to collect more stats

[12:22:01.0000] <jamesr>
like frames decoded, frames presented, etc

[12:22:02.0000] <timeless>
yeah

[12:22:03.0000] <timeless>
i'll mention that

[12:22:04.0000] <timeless>
i'm not sure i really want to encourage this

[12:22:05.0000] <timeless>
i'll pastebin my proposal at some point and ask for feedback before i send it out

[12:23:00.0000] <timeless>
[mozBug 580531]

[12:23:01.0000] <timeless>
.mozDroppedFrames

[13:38:00.0000] <zcorpan>
annevk: "DOM4" is not a great name from a google search result standpoint either

[13:39:00.0000] <zcorpan>
(unless SafeSearch is on strict)

[13:41:00.0000] <nimbu>
ROFLZ

[13:42:00.0000] <zcorpan>
/me is pleased to see several instances of DOM cleanup in firefox 6 release notes

[13:43:00.0000] <Ms2ger>
Try guessing how many of these were mine :)

[13:44:00.0000] <zcorpan>
all of them?

[13:44:01.0000] <micheil>
hey, in the WebSockets specification / draft, where is the MessageEvent type defined?

[13:44:02.0000] <zcorpan>
micheil: it's defined in the html spec

[13:44:03.0000] <micheil>
okay...

[13:45:00.0000] <micheil>
is this the comms spec?

[13:45:01.0000] <Ms2ger>
whatwg.org/C/#messageevent

[13:45:02.0000] <micheil>
http://www.w3.org/TR/2008/WD-html5-20080610/comms.html

[13:45:03.0000] <zcorpan>
what Ms2ger said

[13:45:04.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#event-definitions-1

[13:45:05.0000] <Ms2ger>
You certainly don't want to be looking at a 3 years old spec

[13:45:06.0000] <Hixie>
TR/ page strikes again

[13:45:07.0000] <Hixie>
good lord

[13:46:00.0000] <micheil>
okay

[13:46:01.0000] <micheil>
well, I'm just basing some new work so it's a familiar API to those using websockets in the browser

[13:46:02.0000] <micheil>
hence the reason I ask about MessageEvent, which is what the WebSocket object uses on message

[13:47:00.0000] <zcorpan>
micheil: read the whatwg.org/C version, it contains websockets and messageevent and has working xrefs

[13:47:01.0000] <zcorpan>
and it's up-to-date

[13:47:02.0000] <micheil>
is there any plan to extend MessageEvent to include a data type?

[13:47:03.0000] <Hixie>
data type?

[13:48:00.0000] <micheil>
yeah

[13:48:01.0000] <micheil>
so you don't need to do a typeof check

[13:49:00.0000] <Hixie>
...what's wrong with typeof?

[13:49:01.0000] <Ms2ger>
Well, it's annoying from C++

[13:49:02.0000] <micheil>
it's not exactly the best thing

[13:49:03.0000] <Ms2ger>
Why?

[13:50:00.0000] <Hixie>
i agree that it's not the best thing

[13:50:01.0000] <Hixie>
the best thing, is, like, a cat riding a model train or something

[13:50:02.0000] <Hixie>
but i don't see how that would help here

[13:50:03.0000] <micheil>
because, this project is a node.js project, so I wouldn't be using "blob", but rather "buffer"

[13:51:00.0000] <micheil>
I mean, we already know the type, no need to hide it from view

[13:51:01.0000] <micheil>
or obscure it

[13:51:02.0000] <Ms2ger>
Hixie, pics or it didn't happen ;)

[13:51:03.0000] <Hixie>
if you already have a library around this, just make it expose the type however you want it to expose it

[13:51:04.0000] <micheil>
okay

[13:52:00.0000] <micheil>
I like to stay close to standards if possible.

[13:52:01.0000] <micheil>
but that's cool, thanks for pointing me in the right direction.

[13:52:02.0000] <Hixie>
you can also still use typeof with whatever type you're actually using

[13:52:03.0000] <Hixie>
if it's been wrapped

[13:53:00.0000] <Ms2ger>
Microsoft about IndexedDB: "... so we can start getting some adoption on this technology."

[13:53:01.0000] <jgraham>
/me doesn't even see safesearch on google anymore (and just gets DOM4 when searching for DOM4)

[13:53:02.0000] <jgraham>
Also, instant breaks inline search

[13:53:03.0000] <Ms2ger>
I got shepazu's DOM4

[13:53:04.0000] <timeless>
inline?

[13:54:00.0000] <shepazu>
hmmm?

[13:54:01.0000] <jgraham>
the type you get pressing /

[13:54:02.0000] <shepazu>
I don't remember uploading my draft of DOM4...

[13:54:03.0000] <shepazu>
but maybe I did

[13:55:00.0000] <Ms2ger>
You didn't hear Google now searches your local disk as well? :)

[13:55:01.0000] <jgraham>
Ms2ger: Oh, so did I. Having meaningless years in the URL is helpful in to noticing that there is anything wrong

[13:55:02.0000] <Ms2ger>
http://dev.w3.org/2006/webapi/DOM4Core/DOM4Core.html

[13:55:03.0000] <timeless>
jgraham: FAYT :)

[13:56:00.0000] <jgraham>
/me just rejects anything contating /TR/ and tries not to think too hard about the numbers

[13:56:01.0000] <shepazu>
Ms2ger: oh, that was a very early rough draft

[13:56:02.0000] <timeless>
shepazu: the best kind!

[13:56:03.0000] <AryehGregor>
Could we start a wiki page documenting cases where people were confused by /TR/?

[13:56:04.0000] <Ms2ger>
Won't argue with that ;)

[13:56:05.0000] <jgraham>
/me wonders what safesearch has o do with it

[13:56:06.0000] <AryehGregor>
As evidence, you know.

[13:57:00.0000] <zcorpan>
wait, i was editor for that?

[13:57:01.0000] <jgraham>
Does that block unfinished DOM drafts?

[13:57:02.0000] <zcorpan>
/me didn't know

[13:57:03.0000] <jgraham>
Maybe it blocks all non-TR content!

[13:57:04.0000] <jgraham>
Google protecting us from material of questionable stability

[13:57:05.0000] <Ms2ger>
I was going to share DOM Core, but Google doesn't like my name

[13:58:00.0000] <shepazu>
zcorpan: you and I talked about you editing "Web DOM" at W3C, which is why (I think) I started that draft

[13:58:01.0000] <Ms2ger>
Or "+1" it, what the cool kids do these days

[13:58:02.0000] <zcorpan>
shepazu: ah

[13:58:03.0000] <shepazu>
no, Ms2ger, +1 is evil!

[13:58:04.0000] <timeless>
Ms2ger: google killed my name

[13:58:05.0000] <Ms2ger>
I heard

[13:59:00.0000] <shepazu>
zcorpan: but we never closed the loop on it, and now others have picked it up, which I'm happy with

[13:59:01.0000] <Ms2ger>
shepazu, less work for you! ;)

[13:59:02.0000] <zcorpan>
so am i

[13:59:03.0000] <shepazu>
Ms2ger: precisely

[14:00:00.0000] <shepazu>
and now it looks like the DOM extensions I proposed for SVG may make it in as well, with convenience methods like .create() with property bags

[14:00:01.0000] <shepazu>
so, I'm happy

[14:00:02.0000] <zcorpan>
wonder why my email didn't make it to help@ archives. maybe i sent from the wrong address

[14:01:00.0000] <shepazu>
AryehGregor: nice work on the editing spec… I've only skimmed it, but I like what I see

[14:01:01.0000] <AryehGregor>
shepazu, thanks.

[14:01:02.0000] <smaug____>
timeless: Google did what to your name?

[14:02:00.0000] <timeless>
smaug____: they didn't believe that `timeless developer` satisfied their `real name` policy

[14:02:01.0000] <shepazu>
AryehGregor: for v2, it would be interesting to consider adding an SVG/graphics editing API into that

[14:02:02.0000] <timeless>
i claim that `timeless` should satisfy their `common name` requirement

[14:02:03.0000] <smaug____>
um

[14:02:04.0000] <timeless>
given that i've used it in real life for more than half my life

[14:02:05.0000] <timeless>
(and certainly my entire professional life)

[14:02:06.0000] <timeless>
unfortunately the stupid system insisted that i provide 2 names (first, last), so i had to pick a second part

[14:02:07.0000] <AryehGregor>
shepazu, that sounds like a different project, really.  Graphics editing is a whole separate thing from rich text editing.  Obviously it would be a useful addition to the web, of course.

[14:02:08.0000] <timeless>
i had business cards with `timeless developer`

[14:03:00.0000] <AryehGregor>
timeless, send them scans of the business cards and other documentation?

[14:03:01.0000] <shepazu>
a few months ago, I couldn't recall timeless' "real" name and had to ask him :)  I know him almost exclusively as timeless

[14:03:02.0000] <smaug____>
timeless: or just give up with G+ ;)

[14:03:03.0000] <timeless>
shepazu: the majority of mozilla people do too

[14:03:04.0000] <timeless>
(know me almost exclusively as)

[14:03:05.0000] <shepazu>
you gotta admit, though, timeless… it's a little uppity for a name :)

[14:04:00.0000] <timeless>
/me shrugs

[14:04:01.0000] <timeless>
i've used it since around '94 or '95 ...

[14:04:02.0000] <shepazu>
/me contemplates changing nick to "teh_awesome" :P

[14:04:03.0000] <jgraham>
Not more so than chastity

[14:04:04.0000] <jgraham>
Or something

[14:04:05.0000] <jgraham>
Well maybe that is the opposite

[14:05:00.0000] <jgraham>
But still a pretty silly "real" name

[14:05:01.0000] <annevk>
zcorpan, that's because we aren't using it yet

[14:05:02.0000] <annevk>
zcorpan, DOM4 that is

[14:05:03.0000] <Ms2ger>
shepazu, do you have a real name?

[14:06:00.0000] <smaug____>
time to close some tabs. 400+ is getting hard to manage

[14:06:01.0000] <timeless>
heh

[14:06:02.0000] <timeless>
panorama isn't helping?

[14:06:03.0000] <shepazu>
Ms2ger: I have the name my parents gave me, though I don't really care for it

[14:07:00.0000] <timeless>
AryehGregor: i didn't seee any useful place to do anything about it

[14:07:01.0000] <AryehGregor>
timeless, unsurprising.  Google doesn't seem to like dealing with user feedback.

[14:07:02.0000] <smaug____>
well, 400 is enough that panorama just puts them over each others

[14:07:03.0000] <AryehGregor>
(in general)

[14:07:04.0000] <timeless>
AryehGregor: anyway

[14:07:05.0000] <timeless>
yeah well

[14:07:06.0000] <timeless>
they aren't alone

[14:07:07.0000] <shepazu>
AryehGregor: yeah, I could see that being a separate API, though there will be overlap in the text editing

[14:07:08.0000] <timeless>
i've had pretty bad luck dealing w/ various groups

[14:07:09.0000] <timeless>
which reminds me

[14:07:10.0000] <AryehGregor>
Why don't you legally change your name?

[14:07:11.0000] <timeless>
... time to call Chase again and yell at them... again

[14:07:12.0000] <timeless>
heh

[14:08:00.0000] <timeless>
that'd probably mess up my immigration status here

[14:08:01.0000] <zewt>
panorama seems pretty useless

[14:08:02.0000] <zewt>
thumbnails of webpages aren't helpful at all

[14:08:03.0000] <timeless>
i can't even leave ontario for more than 30 days between when i arrived and some day in december

[14:09:00.0000] <smaug____>
panorama works quite well if you have something like <50 tabs per group

[14:09:01.0000] <timeless>
zewt: i basically have 7 windows

[14:09:02.0000] <shepazu>
/me always liked names like "Capability Brown" and the Puritan "virtue names" like Temperance, Lamentation, Redeemed, Be-faithful, The-peace-of-God, etc.

[14:10:00.0000] <timeless>
one of those windows has 4 groups

[14:10:01.0000] <timeless>
`current work`, `<employer>`, `accessibility`, `w3 todo`

[14:11:00.0000] <timeless>
one of those groups is empty, two have 2 items, todo has 7

[14:12:00.0000] <annevk>
400 tabs? wow

[14:12:01.0000] <annevk>
and my friends think I'm crazy with about thirty

[14:12:02.0000] <timeless>
another window has 5 tabs (in a single group)

[14:12:03.0000] <zewt>
i don't find it useful, i don't know what use thumbnails of windows filled with text is supposed to be

[14:12:04.0000] <shepazu>
/me once had almost 1000

[14:12:05.0000] <timeless>
in my case, i don't need the thumbnails

[14:12:06.0000] <timeless>
i need the document titles

[14:12:07.0000] <zcorpan>
annevk: 400 is nothing for bratell

[14:12:08.0000] <timeless>
`the web platform: bro...` `world wide web cons...` `widget packaging and...`

[14:13:00.0000] <zewt>
i wish the address bar completion was faster, which would make searching for tabs by title more useful

[14:13:01.0000] <timeless>
`w3c news archive: 20...` `the network informati...` `deviceorientation eve...`

[14:13:02.0000] <zewt>
but even with the completion delay set to 0 there's an ugly delay between typing and completion

[14:13:03.0000] <annevk>
zcorpan, damn it, and he already was the better man

[14:13:04.0000] <timeless>
i only have one duplicate in that set of 7

[14:13:05.0000] <zewt>
generally i just remember where tabs are in each window, in practice

[14:13:06.0000] <annevk>
I tried joining the Editing group btw

[14:13:07.0000] <annevk>
seems that request is now redirected to Charles

[14:13:08.0000] <timeless>
amusingly, the duplicate is in fact a duplicate (two views of the w3 home page!)

[14:14:00.0000] <timeless>
annevk: requires AC Rep approval?

[14:14:01.0000] <annevk>
yes because of patent commitments

[14:14:02.0000] <timeless>
joy

[14:14:03.0000] <annevk>
it's kind of interesting that you can vouch for a group without AC approval but joining it requires AC approval

[14:14:04.0000] <annevk>
it makes sense that joining it requires that though

[14:14:05.0000] <timeless>
yeah, that seems slightly flawed

[14:14:06.0000] <annevk>
given the patent policy

[14:15:00.0000] <timeless>
yeah, that part seems right

[14:15:01.0000] <annevk>
which until you get to REC is a lot better than what every other WD has

[14:15:02.0000] <timeless>
zewt: anyway, in my case, the previews of the w3 todo items are actually pretty good / usable

[14:15:03.0000] <annevk>
so hopefully the W3C will move normal groups to that as well

[14:15:04.0000] <timeless>
annevk: wait

[14:15:05.0000] <timeless>
CGs require patent grants on current work?

[14:16:00.0000] <timeless>
as opposed to WGs which only apply the grants to REC achieved work?

[14:16:01.0000] <AryehGregor>
Only for your actual contributions.

[14:16:02.0000] <AryehGregor>
IIUC.

[14:16:03.0000] <annevk>
timeless, yes

[14:19:00.0000] <shepazu>
timeless: only for those contributions that go on to Rec track, I think… but you make the commitment up front, as opposed to waiting until Rec (IIRC)

[14:35:00.0000] <annevk>
guess I should read up again on what patent policy protection you get

[14:40:00.0000] <paul_irish_>
who is a good contact at MSFT for webforms ?

[14:41:00.0000] <timeless>
annevk / shepazu : is it unreasonable for me to expect that protection to be clearly linked from the /pending/ page ?

[14:42:00.0000] <shepazu>
timeless: not sure what you mean...

[14:43:00.0000] <timeless>
shepazu: well, there's clearly a patent policy which affects CGs

[14:43:01.0000] <annevk>
paul_irish_, maybe Adrian?

[14:43:02.0000] <shepazu>
timeless: yes… it's linked from the FAQ

[14:43:03.0000] <paul_irish_>
thanks annevk

[14:43:04.0000] <annevk>
paul_irish_, actually, maybe Eliot Graff

[14:44:00.0000] <annevk>
I think he's on the HTML WG at least

[14:44:01.0000] <timeless>
shepazu: is it linked from http://www.w3.org/community/groups/proposed/ ?

[14:44:02.0000] <shepazu>
timeless: doesn't look like it

[14:44:03.0000] <timeless>
what about from http://www.w3.org/community/groups/ ?

[14:45:00.0000] <timeless>
sorry, since i haven't seen/found the page, i'm listing the pages where i'd have needed to see that link

[14:45:01.0000] <timeless>
i think there are only 3

[14:45:02.0000] <timeless>
those two plus the actual groups themselves

[14:45:03.0000] <shepazu>
it's at least on the "About" tab, and every time you say "Join"

[14:46:00.0000] <timeless>
http://www.w3.org/community/editing/ has a link to W3C Community Contributor License Agreement (CLA).  -- http://www.w3.org/community/about/agreements/cla/

[14:46:01.0000] <timeless>
which isn't a terrible place for it

[14:46:02.0000] <timeless>
but i claim it's one page too late on average

[14:47:00.0000] <timeless>
if it could be added to those two other pages, i think that'd be beneficial to some people

[14:47:01.0000] <shepazu>
timeless: file a bug

[14:47:02.0000] <timeless>
"how" :)

[14:47:03.0000] <timeless>
/me is actually in the middle of filing an internal bug (and has already filed one w3 bug today about CG)

[14:48:00.0000] <annevk>
why did I start a conversation on public-html-xml again while hsivonen is away

[14:49:00.0000] <annevk>
now I need to actually follow up myself :(

[14:49:01.0000] <timeless>
heh

[14:49:02.0000] <timeless>
yeah, definitely a bad idea

[14:55:00.0000] <zcorpan>
nobody implements css3 marquee right?

[14:55:01.0000] <shepazu>
timeless: Ian has updated the FAQ about how to report bugs… also see http://www.w3.org/community/about/tool/

[14:56:00.0000] <annevk>
zcorpan, don't think so

[14:56:01.0000] <annevk>
zcorpan, was there -wap-marquee or some such as well?

[14:56:02.0000] <annevk>
zcorpan, we might have that

[15:04:00.0000] <annevk>
david_carlisle, if you can rewrite the HTML parsing algorithm into something schema / functional; I'd be impressed

[15:04:01.0000] <annevk>
david_carlisle, I don't think it has much to do with politics

[15:05:00.0000] <annevk>
david_carlisle, the basic algorithm is already nigh-on impossible I believe, and then document.write() comes on top of that

[15:06:00.0000] <david_carlisle>
annevk:  if I rules the world (which is false for political reasons) i'd do away with document.write and any other inconveniences

[15:07:00.0000] <annevk>
an interesting way to defend your argument

[15:07:01.0000] <david_carlisle>
annevk: maybe my argument wasn't so clear (for others on irc, this has leaked from public-html-xml_

[15:07:02.0000] <timeless>
david_carlisle =~ s/rules/ruled/

[15:07:03.0000] <annevk>
but I'm not sure it holds, you didn't really say it involved making changes

[15:08:00.0000] <gsnedders>
david_carlisle: And then leave document.write etc. undefined?

[15:08:01.0000] <david_carlisle>
timeless: you clearly don't follow xslt-list (otherwise you'd give up correcting my typos0

[15:09:00.0000] <timeless>
s/0/)/ -- and yes, clearly

[15:09:01.0000] <timeless>
as a matter of fact, i don't have any love for xslt :)

[15:09:02.0000] <annevk>
you might be one of the few WHATWG regulars that follows xslt-list :)

[15:09:03.0000] <david_carlisle>
annevk:  I account for around a third of the traffic on that list last time I looked

[15:10:00.0000] <timeless>
but note that i didn't correct rules to ruled as a typo, but as a grammar-o

[15:11:00.0000] <timeless>
on irc i'm less likely to correct typos

[15:11:01.0000] <timeless>
i am impressed that you can mistype ) as both 0 and _ fwiw :)

[15:12:00.0000] <david_carlisle>
timeless: my usual metric is to try to be within one key with the right hand  and two with the left,

[15:13:00.0000] <Philip`>
/me wrote (some of) an HTML5 parser is OCaml, and wonders if that counts as "functional" :-)

[15:14:00.0000] <timeless>
does anyone use it?

[15:14:01.0000] <timeless>
if not, i'd call it non-functional :)

[15:14:02.0000] <david_carlisle>
annevk: but I didn't say that i could make html parsing schema driven, just that it would have been nicer to do that in some other legacy free world, which is more or less what you said in your reply to John, that leniemt xml parsing can be simpler because ther eis less legacy

[15:15:00.0000] <Philip`>
timeless: I used it to generate interesting tokeniser test cases that are now part of the html5lib test suite, so it's very indirectly still used :-)

[15:16:00.0000] <timeless>
i don't count that :)

[15:16:01.0000] <timeless>
s/do/would/

[15:17:00.0000] <annevk>
My name is mentioned in http://www.w3.org/TR/SVG11/ but I forgot what I contributed :(

[15:17:01.0000] <timeless>
heh

[15:17:02.0000] <annevk>
Congratulations nonetheless, shepazu & posse

[15:17:03.0000] <heycam>
oh right, I forgot that was being published today

[15:18:00.0000] <annevk>
heycam, it's on TR/ man, old news :p

[15:18:01.0000] <heycam>
heh

[15:18:02.0000] <timeless>
ooh, i'm in http://www.w3.org/TR/2008/PR-ElementTraversal-20081117/

[15:18:03.0000] <heycam>
TR/ has one thing going for it -- it's 5 billion times faster than dev.w3.org

[15:18:04.0000] <shepazu>
annevk: thanks

[15:19:00.0000] <timeless>
heycam: heh


2011-08-17
[22:28:00.0000] <Hixie>
where's hsivonen when you need him

[22:28:01.0000] <Hixie>
or ms2ger

[22:28:02.0000] <Hixie>
is anyone around who cares about keeping the whatwg and w3c specs in sync?

[22:29:00.0000] <Hixie>
how about any opera people?

[22:29:01.0000] <Hixie>
ok fine

[22:29:02.0000] <Hixie>
/me will just make something up

[22:43:00.0000] <Hixie>
the definition of makework, courtesy of julian and the chairs: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11183

[01:30:00.0000] <annevk>
morning

[01:35:00.0000] <zcorpan>
morning

[01:37:00.0000] <annevk>
hmm, one new email from Microsoft

[01:38:00.0000] <annevk>
apparently they have not read the "Goals" section

[01:40:00.0000] <zcorpan>
r6479 sounds a bit like "if you call within 15 minutes -- you heard right, within 15 minutes -- you will get this semantic dust sweeper WITHOUT EXTRA CHARGE!"

[01:42:00.0000] <annevk>
zcorpan, pointer? it's not on HTML5 Tracker

[01:43:00.0000] <zcorpan>
yeah it is

[01:44:00.0000] <zcorpan>
editorial change

[01:46:00.0000] <annevk>
well editorial changes make it there too normally

[01:48:00.0000] <zcorpan>
http://html5.org/tools/web-apps-tracker?from=6478&to=6479

[01:50:00.0000] <annevk>
that is weird

[01:51:00.0000] <annevk>
I wonder how that happened

[02:10:00.0000] <annevk>
hmm

[02:10:01.0000] <annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=312019 suggests the namespace algorithms in DOM Core are flawed

[02:14:00.0000] <zcorpan>
what, there are bugs with namespace algorithms? not possible

[02:14:01.0000] <annevk>
sssh

[02:14:02.0000] <annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=505178 suggests the algorithm needs changing, but nobody followed up thus far

[02:14:03.0000] <annevk>
/me added a comment asking for feedback

[03:18:00.0000] <MikeSmith>
wow, http://www.html5rocks.com/en/tutorials/internals/howbrowserswork/ is seriously great

[03:19:00.0000] <MikeSmith>
have never before seen anybody actually write all this up

[03:19:01.0000] <MikeSmith>
paul_irish: ↑ kudos to you Tali man

[03:20:00.0000] <MikeSmith>
people are going to get tired of hearing me talk about that article

[03:22:00.0000] <MikeSmith>
hmm, some great stuff in the Resources section also

[03:22:01.0000] <MikeSmith>
http://www.html5rocks.com/en/tutorials/internals/howbrowserswork/#Resources

[03:23:00.0000] <zcorpan>
MikeSmith: ok, great article, now stop talking about it :P

[03:23:01.0000] <MikeSmith>
heh

[03:24:00.0000] <Ms2ger>
Unfortunately it contains nonsense about DTDs

[03:25:00.0000] <MikeSmith>
oh really?

[03:25:01.0000] <MikeSmith>
hmm yeah

[03:25:02.0000] <Ms2ger>
And its definition of "Really lousy HTML" looks rather clean to me :)

[03:26:00.0000] <zcorpan>
just four errors

[03:27:00.0000] <zcorpan>
of which two are parser errors

[03:27:01.0000] <MikeSmith>
it correctly doesn't actually claim that browsers do anything with DTDs, and given that the article is about what browsers do, the stuff about DTDs could (should) just be yanked

[03:27:02.0000] <Ms2ger>
google.com probably hits more than that without trying

[03:28:00.0000] <zcorpan>
MikeSmith: i dunno if the article tries to set straight misconseptions, but if it does in general then it could talk briefly about how browsers *don't* use DTDs

[03:29:00.0000] <MikeSmith>
yeah

[03:29:01.0000] <MikeSmith>
notes for paul_irish

[03:29:02.0000] <MikeSmith>
the mention of DTDs seems to only be there to support the point that there is no context-free grammar for HTML

[03:29:03.0000] <MikeSmith>
it should instead mention the HTML5 parsing algorithm, and link to it

[03:30:00.0000] <Ms2ger>
It does somewhat further down, actually

[03:30:01.0000] <zcorpan>
the stray table example is wrong

[03:30:02.0000] <MikeSmith>
Ms2ger: oh, ok

[03:31:00.0000] <MikeSmith>
it also says, "The vocabulary and syntax of HTML are defined in specifications" but doesn't mention the parsing rules are also defined in a spec now

[03:32:00.0000] <Ms2ger>
The HTML definitions can be found here: www.w3.org/TR/2003/REC-DOM-Level-2-HTML-20030109/idl-definitions.html.

[03:32:01.0000] <MikeSmith>
actually, I see now that this article is mostly from an earlier document from Tali

[03:32:02.0000] <zcorpan>
"www.liceo.edu.mx is an example of a site that achieves a level of nesting of about 1500 tags, all from a bunch of <b>s. We will only allow at most 20 nested tags of the same type before just ignoring them all together." - the spec doesn't have this limit hard-coded right?

[03:32:03.0000] <MikeSmith>
http://taligarsiel.com/Projects/howbrowserswork1.htm

[03:33:00.0000] <zcorpan>
/me wonders if the code snippets are from webkit's old parser

[03:34:00.0000] <jgraham>
The bit about parsing is very confusing

[03:34:01.0000] <jgraham>
"Every format you can parse must have deterministic grammar consisting of vocabulary and syntax rules. It is called a context free grammar"

[03:34:02.0000] <zcorpan>
if (t->isCloseTag(brTag) && m_document->inCompatMode()) { suggests so

[03:34:03.0000] <jgraham>
Well clearly no

[03:34:04.0000] <jgraham>
t

[03:35:00.0000] <MikeSmith>
jgraham: yeah, that bit is self-contradictory

[03:35:01.0000] <annevk>
note that it was not written by paul_irish

[03:35:02.0000] <annevk>
he edited it

[03:35:03.0000] <MikeSmith>
OK

[03:35:04.0000] <MikeSmith>
yeah

[03:35:05.0000] <Ms2ger>
Could have edited a bit more ;)

[03:35:06.0000] <MikeSmith>
yeah, it clearly needs some more editing love

[03:35:07.0000] <annevk>
the original is from Tali Garsiel based on research on Gecko/WebKit

[03:35:08.0000] <annevk>
guess that was a while ago

[03:35:09.0000] <MikeSmith>
for this kind of document, maybe they should move the source to github or somewhere and accept patches

[03:36:00.0000] <jgraham>
In fact, what is the point of all the section describing parser theory?

[03:37:00.0000] <annevk>
jgraham, didn't get that either

[03:37:01.0000] <MikeSmith>
to set some context?

[03:37:02.0000] <jgraham>
MikeSmith: It reads a bit like someone has tried to summarise the wikipedia article on parsing

[03:37:03.0000] <Ms2ger>
s/github/Bitbucket/

[03:38:00.0000] <jgraham>
I would prefer just a link to wikipedia

[03:38:01.0000] <MikeSmith>
jgraham: it is a pretty longish overall, too

[03:38:02.0000] <MikeSmith>
trimming some stuff out would make it an easier read

[03:39:00.0000] <Ms2ger>
MikeSmith, well, browsers are big ;)

[03:39:01.0000] <jgraham>
and yes, the bit about DTDs is very misleading

[03:39:02.0000] <jgraham>
Ms2ger: Yeah, so best not to waste space with background material that is better covered elsewhere

[03:39:03.0000] <zcorpan>
MikeSmith: trimming stuff that have little to do with browser internals would make it more focused on the topic

[03:40:00.0000] <MikeSmith>
Ms2ger: yeah, so big that when writing about how they work, it's best to omit stuff that's not really related at all to how browsers work

[03:40:01.0000] <MikeSmith>
zcorpan: what you said :)

[03:40:02.0000] <jgraham>
There are also a number of spelling/grammar mistakes but I will let timeless point those out

[03:40:03.0000] <MikeSmith>
and what jgraham said

[03:40:04.0000] <MikeSmith>
you dudes type faster than me

[03:40:05.0000] <MikeSmith>
or maybe you think faster

[03:41:00.0000] <zcorpan>
we have faster internets

[03:41:01.0000] <MikeSmith>
heh

[03:41:02.0000] <zcorpan>
fatter tubes

[03:41:03.0000] <MikeSmith>
tmi

[03:42:00.0000] <jgraham>
Is that HTML parsing comment really in the new webkit parser? It seems very pre-HTML5

[03:42:01.0000] <foolip>
anyone know who nimbu is?

[03:43:00.0000] <Ms2ger>
foolip, Opera employee ;)

[03:43:01.0000] <zcorpan>
jgraham: i already concluded that the snippets must be from the old parser

[03:43:02.0000] <zcorpan>
jgraham: if (t->isCloseTag(brTag) && m_document->inCompatMode()) {

[03:44:00.0000] <foolip>
Ms2ger, ah, it's Divya Manian it seems

[03:44:01.0000] <jgraham>
The part of script scheduling is… not complete

[03:44:02.0000] <jgraham>
(on

[03:45:00.0000] <Ms2ger>
jgraham, well, would anybody get through it if it was?

[03:46:00.0000] <jgraham>
Well fair point, but one should at least point out that things are much more complex than they appear

[03:46:01.0000] <jgraham>
<head> goes in the render tree if you set it to display:block or something

[03:49:00.0000] <jgraham>
No ecmascript?

[03:49:01.0000] <jgraham>
/me skimmed most of the style/layout bit

[03:52:00.0000] <jgraham>
Also, "Feynman" is misspelt in the bio. If you are going to compare yourself to an intellectual luminary, you should at least try to get their name right :)

[03:55:00.0000] <jgraham>
annevk: XHR2 uses "subtract 0x20 from each byte" to do ascii case folding?!

[04:00:00.0000] <annevk>
jgraham, method is a sequence of octets at that point, not characters

[04:01:00.0000] <jgraham>
annevk: That makes no sense. Step 2 talks about code points. The same step talks about "case insensitive match" for a string

[04:03:00.0000] <annevk>
you can use case-insensitive for octets

[04:03:01.0000] <annevk>
also, step 2 turns method into a sequence of octets

[04:04:00.0000] <jgraham>
I don't see why you can do a case insensitive match and can't do case transformations

[04:05:00.0000] <jgraham>
Except for via direct manipulation of the values

[04:05:01.0000] <annevk>
presumably you can do case transformations, just nobody has that defined for octet sequences afaik

[04:05:02.0000] <MikeSmith>
Ms2ger: why bitbucket instead of github?

[04:06:00.0000] <jgraham>
OK, then my feedback is that you should find a way to either transform the case before the deflate or define what it means to case transform a sequence of bytes

[04:06:01.0000] <Philip`>
Because hg > git

[04:06:02.0000] <jgraham>
also hg < git. It is a funny algebra

[04:07:00.0000] <annevk>
not sure why I should do any of that

[04:08:00.0000] <jgraham>
Because the current spec relies on implicit knowledge of the layout of ascii to judge intent

[04:08:01.0000] <jgraham>
That's bad

[04:08:02.0000] <annevk>
HTML5 has that too iirc

[04:08:03.0000] <jgraham>
Well it's bad in HTML5 too then

[04:09:00.0000] <jgraham>
If you tell me where I will file a bug

[04:09:01.0000] <annevk>
e.g. http://www.whatwg.org/specs/web-apps/current-work/complete.html#concept-get-attributes-when-sniffing

[04:14:00.0000] <MikeSmith_>
speaking of github, http://es5.github.com/ is now multi-page

[04:14:01.0000] <jgraham>
MikeSmith: Oh, taht's sad

[04:14:02.0000] <MikeSmith_>
thanks to a repurposing of Philip` splitter

[04:14:03.0000] <jgraham>
I hope there is also a single page version

[04:14:04.0000] <jgraham>
and it is the default

[04:14:05.0000] <MikeSmith_>
jgraham: single-page is at http://es5.github.com/spec.html

[04:14:06.0000] <MikeSmith_>
hmm

[04:14:07.0000] <MikeSmith_>
OK, I guess I could make it the default

[04:15:00.0000] <jgraham>
The multipage version isn't very useful is it?

[04:15:01.0000] <jgraham>
I mean that page doesn't tend to kill browsers

[04:15:02.0000] <MikeSmith_>
matjas indicated otherwise

[04:15:03.0000] <MikeSmith_>
source is 1.5MB

[04:15:04.0000] <jgraham>
and the whole point of HTML in this context is to get proper search and hyperlinks

[04:15:05.0000] <MikeSmith_>
OK

[04:16:00.0000] <MikeSmith_>
so I can make the single-page the default again

[04:16:01.0000] <MikeSmith_>
until somebody suggests otherwise

[04:16:02.0000] <MikeSmith_>
would be good to hear what matjas thinks

[04:17:00.0000] <annevk>
http://platform.html5.org/ should probably link to the crypto stuff and abarth's URL API

[04:18:00.0000] <zcorpan>
is URL implemented?

[04:18:01.0000] <Ms2ger>
We're implementing

[04:19:00.0000] <MikeSmith>
what's the link for the URL API?

[04:19:01.0000] <MikeSmith>
annevk: you mean DOMCrypt?

[04:21:00.0000] <annevk>
I think it is https://docs.google.com/document/edit?id=1r_VTFKApVOaNIkocrg0z-t7lZgzisTuGTXkdzAk4gLU&hl=en&pli=1

[04:23:00.0000] <annevk>
hmm, .isInDocument

[04:23:01.0000] <Ms2ger>
Oh, did smaug propose that?

[04:23:02.0000] <annevk>
it's fairly trivial to just check if the node is documentElement or an ancestor is

[04:24:00.0000] <annevk>
he mentioned it

[04:24:01.0000] <Ms2ger>
He's mentioned it before

[04:25:00.0000] <annevk>
MikeSmith, for XMLHttpRequest please link to the -2 edition

[04:25:01.0000] <MikeSmith>
OK

[04:25:02.0000] <Ms2ger>
"I think the PATCH method [1] should be supported in the XMLHttpRequest2 (section 4.6.1)."

[04:26:00.0000] <Ms2ger>
Use cases? What's that?

[04:26:01.0000] <MikeSmith>
annevk: and by crypto stuff did you mean DOMCrypt or the random-number thing?

[04:26:02.0000] <Ms2ger>
Both! :)

[04:27:00.0000] <annevk>
I think it would have been better if XHR uppercased all methods

[04:27:01.0000] <annevk>
but that's too much of a fork of HTTP reportedly

[04:27:02.0000] <annevk>
despite the fact that I don't think anyone will ever mint a lowercase method name at this point

[04:30:00.0000] <annevk>
btw, zcorpan, if you can figure out what exactly XHR should say with respect to garbage collection, I'd appreciate it

[04:33:00.0000] <zcorpan>
annevk: i'll have a look

[04:34:00.0000] <annevk>
you already filed two bugs, I'm now asking you to fix them :p

[04:34:01.0000] <annevk>
last time I took a look at the various sections in HTML I got very confused

[04:35:00.0000] <smaug____>
Ms2ger: annevk: so what do you think about isInDocument

[04:36:00.0000] <smaug____>
it would be trivial to add

[04:36:01.0000] <zcorpan>
annevk: one about navigation and the other about event handlers, right?

[04:36:02.0000] <smaug____>
and quite useful

[04:36:03.0000] <annevk>
smaug____, useful for what?

[04:37:00.0000] <annevk>
zcorpan, http://www.w3.org/Bugs/Public/show_bug.cgi?id=12837 and http://www.w3.org/Bugs/Public/show_bug.cgi?id=12838

[04:37:01.0000] <smaug____>
well, when handling some subtrees

[04:37:02.0000] <smaug____>
/me needs to find some use case

[04:38:00.0000] <zcorpan>
heh, "The readyState attribute changes at some seemingly arbitrary times for historical reasons."

[04:41:00.0000] <matjas>
MikeSmith: yeah, the single-page variant froze my browser regularly

[04:41:01.0000] <MikeSmith>
matjas: ok

[04:41:02.0000] <matjas>
(thanks btw)

[04:41:03.0000] <MikeSmith>
np

[04:41:04.0000] <matjas>
I would be fine with it still being the default though

[04:41:05.0000] <MikeSmith>
ok

[04:41:06.0000] <MikeSmith>
will make that change now, then

[04:42:00.0000] <matjas>
anything to make jgraham happy!

[04:49:00.0000] <MikeSmith>
matjas: btw, chapter 15 is still a 600+ MB file

[04:49:01.0000] <erlehmann>
wat

[04:49:02.0000] <MikeSmith>
um, 600K

[04:49:03.0000] <MikeSmith>
I meant

[04:50:00.0000] <matjas>
phew ;)

[04:50:01.0000] <MikeSmith>
anyway, not surprising it's that big I guess

[04:50:02.0000] <MikeSmith>
I may tweak the splitter further to chop that up into smaller files

[04:50:03.0000] <jgraham>
matjas: So far today everything is conspiring to make me unhappy (more so than usual I mean), so I am greatful for small victories ;)

[04:50:04.0000] <MikeSmith>
jgraham: ok, default is now back to single-page

[04:50:05.0000] <matjas>
MikeSmith: this is chapter 15: http://i.imgur.com/HD83W.png :(

[04:51:00.0000] <MikeSmith>
matjas: point taken :)

[04:51:01.0000] <matjas>
needs moar splitting

[04:51:02.0000] <MikeSmith>
I will take a look at the splitter code now

[04:51:03.0000] <MikeSmith>
yup

[04:51:04.0000] <zcorpan>
annevk: it seems the spec fires the events in the same task as it changes readyState, which is great for specifying the gc policy

[04:52:00.0000] <jgraham>
MikeSmith: Thanks :)

[04:52:01.0000] <MikeSmith>
yup

[04:52:02.0000] <MikeSmith>
thanks for prodding me

[04:53:00.0000] <MikeSmith>
the splitter currently just splits on any <h2> but I guess  a good tweak would to also have it split on anything that has a, say, class=splitme

[04:53:01.0000] <MikeSmith>
/me counts how many ch. 15 subsections he needs to add a class attribute to

[04:57:00.0000] <annevk>
zcorpan, cool I guess :)

[04:57:01.0000] <annevk>
/me wonders where Ms2ger went

[04:58:00.0000] <annevk>
I tried clarifying the goals of the DOM spec: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#goals

[04:59:00.0000] <annevk>
I hope that's clear enough for Microsoft

[05:08:00.0000] <MikeSmith>
matjas: OK, chapter 15 now split up too

[05:09:00.0000] <MikeSmith>
matjas: all the generated files are now less than 130K

[05:09:01.0000] <MikeSmith>
but if you have problems with any others, I can split some further

[05:11:00.0000] <matjas>
yay \o/

[05:11:01.0000] <MikeSmith>
:)

[05:14:00.0000] <zcorpan>
annevk: ok now i need go through the sync case

[05:14:01.0000] <matjas>
seems like github-pages still haven’t been rebuilt fully

[05:15:00.0000] <Ms2ger>
annevk, seems like the HTML5 stuff doesn't really belong below (1)

[05:18:00.0000] <MikeSmith>
matjas: they have for me

[05:18:01.0000] <MikeSmith>
though I had to force-reload

[05:21:00.0000] <MikeSmith>
matjas: e.g., http://es5.github.com/x15.4.html

[05:21:01.0000] <matjas>
yeah works now

[05:21:02.0000] <MikeSmith>
ok

[05:21:03.0000] <matjas>
those URLs already worked but the links just wouldn’t show up, even after a force-reload

[05:22:00.0000] <matjas>
¯\_(ツ)_/¯

[05:23:00.0000] <annevk>
Ms2ger, good point

[05:23:01.0000] <annevk>
I'll add that as a separate point

[05:23:02.0000] <annevk>
4 goals for DOM4

[05:28:00.0000] <smaug____>
annevk: how does http://dev.w3.org/2006/webapi/DOM-Level-3-Events/html/DOM3-Events.html#event-types-list prevent anything?

[05:28:01.0000] <smaug____>
the table is marked non-normative

[05:30:00.0000] <annevk>
" if it is attached to the bubbling phase only, this event listener is required not be triggered"

[05:30:01.0000] <annevk>
DOM3 Events is hopelessly confusing, contradictory and full of holes

[05:30:02.0000] <annevk>
it's beyond me why it's being developed

[05:31:00.0000] <smaug____>
HTML spec is hopelessly confusing and full of holes :p

[05:32:00.0000] <annevk>
no it isn't

[05:32:01.0000] <smaug____>
and yet I think it is a reasonable thing to develop

[05:32:02.0000] <smaug____>
it is

[05:32:03.0000] <zcorpan>
annevk: ok commented on the GC bug. couldn't figure out if the sync case should be different or not

[05:32:04.0000] <zcorpan>
annevk: likely it's not quite right, too. needs review

[05:32:05.0000] <annevk>
smaug____, it's far more detailed and doesn't contain such glaring inaccuracies

[05:32:06.0000] <annevk>
thanks zcorpan

[05:39:00.0000] <bencc>
who is responsible to send keep-alives ping requests in websockets? the application code or the browser?

[05:40:00.0000] <annevk>
prolly app

[05:40:01.0000] <jgraham>
I thought the opposite

[05:41:00.0000] <annevk>
might have changed I guess, haven't followed anything recent

[05:42:00.0000] <bencc>
prolly app?

[05:42:01.0000] <bencc>
what is prolly?

[05:42:02.0000] <annevk>
slang for probably

[05:43:00.0000] <bencc>
ok. thanks

[05:43:01.0000] <jgraham>
Oh well the draft doesn't actually say anything. So who knows

[05:44:00.0000] <bencc>
the browser teams must know

[05:44:01.0000] <bencc>
because they either implement a ping or not

[05:44:02.0000] <jgraham>
Well not by reading the draft they won't

[05:45:00.0000] <bencc>
so it should be fixed

[05:45:01.0000] <bencc>
they probably don't handle it

[05:48:00.0000] <bencc>
how do I send a ping from the client side?

[05:49:00.0000] <jgraham>
AFAIK you can't

[05:50:00.0000] <zcorpan>
the API doesn't expose anything around pings/pongs

[05:50:01.0000] <zcorpan>
so teh app can't do anything with pings/pongs

[05:50:02.0000] <jgraham>
So presumably the browser is supposed to send them?

[05:50:03.0000] <zcorpan>
at least in the browser

[05:50:04.0000] <annevk>
zcorpan, will you do salvaging too?

[05:50:05.0000] <jgraham>
Or are we ignoring them?

[05:50:06.0000] <bencc>
how can the server knows how long to wait to find a dead connection?

[05:50:07.0000] <zcorpan>
annevk: yeah i'll look at that as well

[05:50:08.0000] <annevk>
zcorpan, also, for sync, can you even get into garbage collection?

[05:51:00.0000] <annevk>
zcorpan, while a sync operation is ongoing

[05:51:01.0000] <zcorpan>
annevk: sure

[05:51:02.0000] <annevk>
i guess if people close the browser?

[05:52:00.0000] <zcorpan>
or if the browser is running out of resources and wants to do a hard GC somewhere

[05:53:00.0000] <Ms2ger>
Oh look, D3E now has two automated tests

[05:53:01.0000] <annevk>
zcorpan, that particular scenario is out of scope typically

[05:55:00.0000] <annevk>
Ms2ger, can we change Anolis to just output the names of all authors in the references? or is that forbidden?

[05:55:01.0000] <zcorpan>
websocket talks about what to do if you GC an open websocket

[05:55:02.0000] <Ms2ger>
WFM

[05:55:03.0000] <Ms2ger>
Want to fix? :)

[05:56:00.0000] <annevk>
not now, but I can look into it I guess

[06:19:00.0000] <zcorpan>
"Act as if the user aborted the cancels the request."? wow, wonder how i ended up with that

[06:20:00.0000] <zcorpan>
hmm i guess i typed up to "aborted the", then looked up what the spec called it and then forgot to remove that part

[06:21:00.0000] <zcorpan>
annevk: possibly you need to have a way to kill an XHR without firing events, to use when navigating away

[07:28:00.0000] <timeless>
jgraham: i'd only bother if someone would accept my edits

[07:28:01.0000] <timeless>
otoh, if it were on bitbucket, i'd fork it and make a patch series and send a pull request

[07:33:00.0000] <timeless>
zcorpan: re websocket you can define a per app specific NOP/PING and then use .send() with it

[07:33:01.0000] <timeless>
bencc: interestingly enough, i had the same question yesterday

[07:33:02.0000] <timeless>
/me had poor connectivity and couldn't figure out where the current websocket protocol spec was

[07:34:00.0000] <timeless>
/me reaches now now

[07:37:00.0000] <zcorpan>
timeless: those would be data messages, not ping/pong frames

[07:42:00.0000] <timeless>
sure, data messages w/ app specific ping-pongs :)

[08:39:00.0000] <bencc>
sessionStorage is per browser tab and localStorage lasts between sessions

[08:39:01.0000] <bencc>
is there something in between?

[08:39:02.0000] <bencc>
I need to store data that all the tabs can access but I need it to be per session

[08:39:03.0000] <bencc>
like cookies are

[08:43:00.0000] <AryehGregor>
Shelley: "Just an opinion, but Aryeh, I think you need to have a chat with Google about your work with the W3C, because you seem to be confused about Google's relationship with the W3C and how it impacts on your work."

[08:43:01.0000] <AryehGregor>
My carefully calculated response: "The person at Google who oversees my work is Ian Hickson.  If you think I misunderstand my work obligations, I advise you to take it up with him."

[08:45:00.0000] <jgraham>
My carefuly crafted response to Shelley would be ""

[08:45:01.0000] <jgraham>
Life is just too short

[08:46:00.0000] <AryehGregor>
Yeah, I have an excessively strong tendency toward http://xkcd.com/386/.

[08:51:00.0000] <annevk>
where did zcorpan go?

[08:51:01.0000] <annevk>
hmm

[08:51:02.0000] <annevk>
I think XMLHttpRequest garbage collection might need to be closer to EventSource than WebSocket

[09:00:00.0000] <annevk>
why does "make disappear" not apply to EventSource?

[09:01:00.0000] <AryehGregor>
specification Failed to load resource

[09:01:01.0000] <AryehGregor>
dfn.js Failed to load resource

[09:01:02.0000] <AryehGregor>
My spec now has no styling.

[09:03:00.0000] <annevk>
whatwg.org offline again?

[09:25:00.0000] <dglazkov>
good morning, Whatwg

[09:25:01.0000] <Ms2ger>
Evening

[09:25:02.0000] <annevk>
hi hi

[11:38:00.0000] <AryehGregor>
Is there some way to make text have a color that inverts whatever it lies on top of?  I have some text that's absolutely positioned, and it sometimes obscures other text.  I'd like the place where they overlap to become white instead of black, or something.

[11:38:01.0000] <AryehGregor>
Would I have to use an SVG filter or some craziness like that?

[11:39:00.0000] <AryehGregor>
/me figures he'll have to live with it overtyping for now

[11:39:01.0000] <shepazu>
AryehGregor: you could do it with Canvas, or with an SVG filter

[11:39:02.0000] <shepazu>
oh, wait

[11:39:03.0000] <shepazu>
I misread that

[11:40:00.0000] <shepazu>
I'd have to see an example, really

[11:40:01.0000] <AryehGregor>
foo<span style=position:absolute>{}</span>bar

[11:40:02.0000] <Ms2ger>
color: invert?

[11:40:03.0000] <AryehGregor>
That's the specific example I'm looking at.

[11:40:04.0000] <Ms2ger>
Or was that only for outline?

[11:41:00.0000] <AryehGregor>
/me doesn't see it as available in http://www.w3.org/TR/css3-color/ for color

[11:41:01.0000] <AryehGregor>
http://www.w3.org/TR/CSS21/ui.html#dynamic-outlines

[11:41:02.0000] <AryehGregor>
Looks like only for outline.

[11:41:03.0000] <AryehGregor>
/me tries it with an outline

[11:44:00.0000] <AryehGregor>
Ah, but that outlines the whole box.

[11:44:01.0000] <AryehGregor>
Well, whatever.

[11:44:02.0000] <Ms2ger>
Though at least Gecko dropped support for that years ago

[11:46:00.0000] <Hixie>
AryehGregor: google's relationship with the w3c's impact on your work is nil, as far as i'm concerned

[11:47:00.0000] <AryehGregor>
Hixie, I realize.

[11:47:01.0000] <Hixie>
:-)

[11:47:02.0000] <Hixie>
though of course you can use your judgement in figuring out how to get feedback, get things implemented, etc

[11:47:03.0000] <Hixie>
(so long as you're not committing google to anything, obviously)

[11:48:00.0000] <AryehGregor>
Clearly.

[11:49:00.0000] <rniwa>
AryehGregor: are you aware of http://www.w3.org/community/editing/?

[11:49:01.0000] <AryehGregor>
rniwa, I created it.

[11:49:02.0000] <rniwa>
AryehGregor: ah ok

[11:49:03.0000] <rniwa>
AryehGregor: should I join that?

[11:49:04.0000] <AryehGregor>
If you like.  I don't think the membership will matter.

[11:49:05.0000] <rniwa>
AryehGregor: ok

[11:50:00.0000] <rniwa>
/me likes to be in the loop

[11:50:01.0000] <AryehGregor>
I plan to use it for publishing the spec and nothing else, since we already have wikis/mailing lists/etc. and I don't want to fragment them.

[11:52:00.0000] <Hixie>
annevk: when you get back, see http://www.w3.org/Bugs/Public/show_bug.cgi?id=11204

[12:49:00.0000] <manu-db>
Question about JavaScript APIs spec'd in WebIDL directed at anyone that knows the answer:

[12:49:01.0000] <manu-db>
When processing a part of a DOM or JSON, is it better to return a null value (if there was some sort of processing error), or is it better to raise an exception of some sort?

[12:50:00.0000] <manu-db>
and if it "depends on the application", are there general guidelines?

[12:52:00.0000] <AryehGregor>
manu-db, it depends, and partly it's a matter of taste.  Exceptions cause the whole app to blow up if uncaught, so it's best to only throw them only if the author did something seriously wrong.  Returning null or something is typically a better idea if there's nothing wrong with what the author did, but there's no useful information to return.

[12:52:01.0000] <Hixie>
(of course returing null if the author didn't expect it can quickly turn into an exception too)

[12:53:00.0000] <AryehGregor>
It would be easier to say if you gave a specific example.

[12:53:01.0000] <manu-db>
/me goes to find an example...

[12:53:02.0000] <manu-db>
http://json-ld.org/spec/ED/20110817/#widl-JSONLDProcessor-frame-object-object-input-object-frame-object-options-JSONLDProcessorCallback-callback

[12:53:03.0000] <manu-db>
bleh, that is an awful URL...

[12:54:00.0000] <Ms2ger>
Yes, I was about to mention that :)

[12:54:01.0000] <zewt>
returning null on an error condition isn't very useful--it doesn't say anything about what failed

[12:54:02.0000] <AryehGregor>
It really depends.

[12:55:00.0000] <manu-db>
anyway - the jsonld.frame() call takes a JavaScript object (input) and another JavaScript object (frame) and re-organizes the original input to have a specific structure/hierarchy (based on the frame).

[12:55:01.0000] <Ms2ger>
And what is the error condition you're thinking of?

[12:55:02.0000] <manu-db>
but there are some frames that are just plain invalid.

[12:55:03.0000] <manu-db>
(because keywords were messed up or data is wrong, etc.)

[12:55:04.0000] <AryehGregor>
manu-db, if designing a new API from scratch, I'd usually say to throw an exception if the method was passed invalid/malformed input.

[12:55:05.0000] <Ms2ger>
What he said ^

[12:56:00.0000] <Hixie>
there are varying schools of thought. My main feedback would be that if you pick one, make sure to stick to it so that the API is consistent

[12:56:01.0000] <manu-db>
so, you could answer back with null, which means (you gave us crap... so your output is null) - or you could throw an exception (you gave us crap, your code is busted... fix it)

[12:56:02.0000] <Hixie>
and if people convince you to change from one to the other, make sure to change everything at the same time so it stays consistent :-)

[12:56:03.0000] <AryehGregor>
If the author is passing malformed input to the API, it almost certainly indicates some kind of bug.

[12:57:00.0000] <AryehGregor>
There's no reason you'd want to pass malformed input to this API, right?

[12:57:01.0000] <Hixie>
well it could be user-provided input

[12:57:02.0000] <Hixie>
right?

[12:57:03.0000] <manu-db>
AryehGregor - correct.

[12:57:04.0000] <Ms2ger>
Consistency is the hobgoblin of little minds.

[12:57:05.0000] <Hixie>
that you just didn't check

[12:57:06.0000] <zewt>
returning null for invalid input would be very unusual

[12:57:07.0000] <manu-db>
well, developer provided in this case.

[12:57:08.0000] <Philip`>
If the function just returned null, how would you tell what the error was in the input and how to fix it?

[12:57:09.0000] <AryehGregor>
Right, then you want to do try/catch.

[12:57:10.0000] <Hixie>
Ms2ger: so are cliches :-P

[12:57:11.0000] <manu-db>
I can't think of a use case where it would be user-provided, Hixie...

[12:57:12.0000] <AryehGregor>
Philip`, probably an exception would just be SYNTAX_ERR anyway.

[12:57:13.0000] <Ms2ger>
Hixie, indeed :)

[12:57:14.0000] <zewt>
to little minds, consistency is a hobgoblin

[12:58:00.0000] <AryehGregor>
manu-db, oh, so you'd expect that the correct use of the function would only be to accept input that comes from another function?

[12:58:01.0000] <Ms2ger>
AryehGregor, of course he should invent an exception hierarchy five levels deep

[12:58:02.0000] <AryehGregor>
Or that was typed manually by the developer?

[12:58:03.0000] <AryehGregor>
In that case, definitely throw.

[12:58:04.0000] <AryehGregor>
Invalid input indicates a bug.

[12:58:05.0000] <manu-db>
AryehGregor - correct.

[12:59:00.0000] <AryehGregor>
The only time I'd say not to throw on invalid input is if it's user-provided or something, so it might be invalid even if there are no bugs in the program.

[12:59:01.0000] <zewt>
should throw in that case too, IMO

[12:59:02.0000] <AryehGregor>
Yeah, I'd usually say to throw then too.

[12:59:03.0000] <AryehGregor>
But there's more of a case for not throwing, because it can cause the program to die.

[12:59:04.0000] <AryehGregor>
If you can handle it gracefully, that might be better.

[12:59:05.0000] <manu-db>
Ok, well - that was all very helpful - thanks AryehGregor, Hixie, Ms2ger, Philip`, zewt

[13:00:00.0000] <zewt>
the error result could do that in any case, if the programmer wasn't aware that it can happen

[13:00:01.0000] <manu-db>
I'll change the API to throw exceptions, consistently, on input that is a syntax error.

[13:00:02.0000] <AryehGregor>
E.g., per my current spec, execCommand() and friends only throw if you call a method on some command where it makes no sense, like asking for the value of bold.

[13:00:03.0000] <zewt>
at least, for things returning a value

[13:00:04.0000] <AryehGregor>
zewt, it depends what you return.  If you're normally going to return an object, yes.  If you're normally returning a string, probably not.

[13:00:05.0000] <AryehGregor>
Particularly not if you return an empty string on error instead of null.

[13:01:00.0000] <AryehGregor>
Basically, make it easy to gracefully handle user input errors.

[13:01:01.0000] <AryehGregor>
If possible.

[13:01:02.0000] <zewt>
just seems like a way of masking bugs and making them harder to fix

[13:01:03.0000] <AryehGregor>
zewt, not if the user input is uncommon or unreasonable anyway.  E.g., at one point I had execCommand("insertImage", false, "") (= insert an image with no URL) throw.  Now it just does nothing.

[13:02:00.0000] <AryehGregor>
Thus it has no effect, which makes sense, instead of crashing the program.

[13:02:01.0000] <AryehGregor>
That's a reasonable case of graceful error handling.  Throwing wouldn't really help anyone.

[13:02:02.0000] <AryehGregor>
Since typically, the URL will be user-provided, and if the user provides the empty string as a URL for some reason, it's not helpful to crash.

[13:03:00.0000] <Philip`>
manu-db: By the way: "The 'input' is used to build the framed output and is returned if there are no errors." sounds like it's returning the 'input' value, which seems kind of redundant - is it meant to say it returns the framed output instead?

[13:03:01.0000] <AryehGregor>
Of course a well-designed app will validate the URL to a greater extent than that, but this helps out apps that were written without so much care.

[13:03:02.0000] <zewt>
that's not reporting an error with a mechanism other than exceptions (eg. returning it), though, that's just silently ignoring errors, which is different

[13:04:00.0000] <manu-db>
Philip` the prose is crap right now... gonna make another pass at cleaning it up. But yes, you are right - I'll try to make it more apparent that the input value is being used to generate /something else/ that will be returned by the method.

[13:04:01.0000] <Philip`>
AryehGregor: "crashing the program" seems misleading - more likely it'll just abort the onclick handler of the button the user clicked to insert an image, which isn't a particularly fatal problem

[13:04:02.0000] <AryehGregor>
Philip`, depends what the onclick handler is doing.

[13:05:00.0000] <AryehGregor>
zewt, yes, true.

[13:12:00.0000] <manu-db>
Got another question for anyone that knows how @lang or @xml:lang is processed in HTML5 across most of the browsers... Ivan reported a bug where the language of a document may not be properly detected if there is a bunch of stuff in the <HTML> element - for example if there are a ton of @xmlns: or @prefix declarations. It seemed as if some of the browsers sniff for the language of the...

[13:12:01.0000] <manu-db>
...document, and if they don't find it in the first 256 bytes (or something like that), they just default to a western character set, which hoses the page.

[13:12:02.0000] <AryehGregor>
IIRC, charset detection isn't very standard right now.  Browsers definitely do charset detection only based on the first handful of bytes, though.

[13:12:03.0000] <manu-db>
This came up because we were thinking of limiting @prefix declarations in RDFa 1.1 to the HTML element... but if this bug exists, we may want to allow @prefix on HTML, HEAD and BODY... instead of all elements.

[13:12:04.0000] <AryehGregor>
As a rule.

[13:13:00.0000] <AryehGregor>
So if they're using lang="" as part of charset detection, that can't be too far into the HTML source.

[13:13:01.0000] <AryehGregor>
Same with <meta charset> etc.

[13:13:02.0000] <manu-db>
damn...

[13:13:03.0000] <AryehGregor>
That's because they don't want to get the first packet, start parsing in one charset, then have to throw everything away and start over when they get the second packet.

[13:14:00.0000] <Philip`>
/me wasn't aware of browsers ever using lang for charset detection

[13:14:01.0000] <AryehGregor>
So they only look at a handful of bytes, which will hopefully be in the first packet.

[13:14:02.0000] <manu-db>
is there any effort to spec that in HTML5? This seems to be an issue in any non-western character-set HTML document.

[13:14:03.0000] <AryehGregor>
There's something about it, let me see.

[13:14:04.0000] <manu-db>
hmm... I see.

[13:14:05.0000] <Philip`>
HTML5 says to look in at least the first 1024 bytes for a charset, I think

[13:14:06.0000] <Ms2ger>
I doubt that browsers really use lang

[13:14:07.0000] <manu-db>
Problem is that it totally screws over the non-western charset people.

[13:15:00.0000] <Ms2ger>
Those who still don't use UTF-8?

[13:15:01.0000] <AryehGregor>
Basically, everyone should just use UTF-8, yeah.

[13:15:02.0000] <manu-db>
yeah

[13:15:03.0000] <AryehGregor>
http://www.whatwg.org/specs/web-apps/current-work/multipage/parsing.html#determining-the-character-encoding

[13:16:00.0000] <manu-db>
Ms2ger yes it screws over the non-UTF-8 people... but there are quite a number of them (anecdotal)

[13:16:01.0000] <Ms2ger>
And if they insist on something else, there's still the HTPP header

[13:16:02.0000] <AryehGregor>
HTML5 seems not to mandate a specific number of bytes.

[13:16:03.0000] <AryehGregor>
"The user agent may wait for more bytes of the resource to be available, either in this step or at any later step in this algorithm. For instance, a user agent might wait 500ms or 1024 bytes, whichever came first."

[13:16:04.0000] <AryehGregor>
"Note: The authoring conformance requirements for character encoding declarations limit them to only appearing in the first 1024 bytes. User agents are therefore encouraged to use the preparse algorithm below (part of these steps) on the first 1024 bytes, but not to stall beyond that."

[13:17:00.0000] <AryehGregor>
The upshot is, though, you want to have <meta charset> or whatnot somewhere early in the page.

[13:17:01.0000] <AryehGregor>
Having it after lots of prefixes or whatever is definitely bad, it can cause problems.

[13:18:00.0000] <Philip`>
And you want to have HTTP content-type:text/html;charset=utf-8 so browsers don't have to sniff for the encoding at all

[13:18:01.0000] <AryehGregor>
Or a BOM or something, right.

[13:18:02.0000] <AryehGregor>
But <meta charset> early in the document is also fine.

[13:18:03.0000] <Philip`>
(Having a document specify its encoding in a way that you can only correctly read after decoding the document is a stupid idea :-( )

[13:19:00.0000] <manu-db>
Philip` this is an issue with files on disk... no way to set Content-Type:

[13:19:01.0000] <AryehGregor>
Philip`, yeah, which is why there's a mini-parser to retrieve <meta charset> . . .

[13:19:02.0000] <The_8472>
how does that parser work?

[13:19:03.0000] <The_8472>
pushing the entire thing through a whole array of charset decoders?

[13:20:00.0000] <Philip`>
The_8472: It's a simplified version of the tokeniser algorithm that looks for meta elements with suitable attributes

[13:20:01.0000] <manu-db>
I'm trying to think of language that we could put in the RDFa Core 1.1 spec that would basically say: "If you want to publish conformant RDFa 1.1 - you should be using UTF-8 and you should be specifying the Content-Type:"

[13:20:02.0000] <Philip`>
and assumes the document is encoded in an ASCII-compatible way

[13:20:03.0000] <The_8472>
ah

[13:21:00.0000] <AryehGregor>
manu-db, note that in HTML5, any document with <meta charset> after the first 1024 bytes is already invalid.

[13:21:01.0000] <The_8472>
so utf-32 would break it

[13:21:02.0000] <Hixie>
utf-32 is strongly discouraged

[13:21:03.0000] <AryehGregor>
So presumably it's already invalid in HTML5+RDFa too.

[13:21:04.0000] <Ms2ger>
Hixie, just discouraged?

[13:21:05.0000] <jamesr>
is utf-32 implemented?

[13:21:06.0000] <manu-db>
We're attempting to constrain @prefix to the root element... and do so in a way that doesn't screw the character encoding detection in browsers.

[13:21:07.0000] <jamesr>
supported?

[13:22:00.0000] <Ms2ger>
Gecko dropped support IIRC

[13:22:01.0000] <Philip`>
I don't the use of UTF-8 is relevant (though it's a good idea anyway) - the default decoding would be win-1252, so UTF-8 is no better than any other encoding in terms of failed charset detection

[13:22:02.0000] <Philip`>
s//think/

[13:22:03.0000] <Hixie>
"Authors should not use UTF-32" and "Support for UTF-32 is not recommended"

[13:23:00.0000] <AryehGregor>
manu-db, you can always use a BOM, if you're using UTF-8 or such.

[13:23:01.0000] <AryehGregor>
Although that can be annoying/confusing.

[13:23:02.0000] <manu-db>
So using a BOM would trigger a high confidence rating ("tentative") across all browsers?

[13:23:03.0000] <Philip`>
The_8472: UTF-16 would break it too, as would e.g. EBCDIC (which is the cause of security vulnerabilities in some web software)

[13:23:04.0000] <The_8472>
some editors don't deal well with it

[13:24:00.0000] <manu-db>
Is there anything else that we could use to trigger UTF-8 detection?

[13:25:00.0000] <manu-db>
scratch that last thing I said - it didn't make any sense... rewording...

[13:25:01.0000] <The_8472>
magic charset guessing algos

[13:25:02.0000] <The_8472>
but they are... well... magic. and need enough content to find something that's not ascii

[13:25:03.0000] <zewt>
charset guessing is always horrible, heh

[13:25:04.0000] <The_8472>
yeah

[13:26:00.0000] <zewt>
also browsers preferring the user's local charset as the default is horrible

[13:26:01.0000] <manu-db>
There are no attributes that the browser uses to influence its guess at character encoding??

[13:26:02.0000] <zewt>
countless japanese pages i've hit that don't display correctly, because of not being able to guess it based on content, and it happening-to-work because the author was on a japanese system, changing the default

[13:27:00.0000] <zewt>
noticed less of that lately; whether due to browser improvements, pages improving or chance i don't know

[13:29:00.0000] <Philip`>
manu-db: You could always encourage people to write <!doctype html><html><meta charset=utf-8><html prefix=...> and rely on the parser to hoist the prefix attribute back onto the root element

[13:29:01.0000] <manu-db>
yikes - that sounds like a recipe for disaster :P

[13:30:00.0000] <Ms2ger>
Not at all, it's very well specified ;)

[13:30:01.0000] <Philip`>
A very well specified disaster

[13:30:02.0000] <manu-db>
true story

[13:30:03.0000] <AryehGregor>
manu-db, it does.  It's called <meta charset>.

[13:30:04.0000] <AryehGregor>
Nobody put arbitrary-length attributes on <html> or <head> before, so it worked fine.

[13:30:05.0000] <manu-db>
/me wonders if it would be okay to say that all RDFa documents MUST be encoded in UTF-8...

[13:31:00.0000] <Philip`>
manu-db: I don't think that would solve any problems

[13:31:01.0000] <Philip`>
since browsers don't assume UTF-8 by default

[13:31:02.0000] <AryehGregor>
manu-db, you don't have to specify anything extra.  Anything that has an encoding specification that browsers won't pick up should, in theory, be invalid already anyway.

[13:31:03.0000] <AryehGregor>
If it's not, that's a bug in HTML5.

[13:31:04.0000] <manu-db>
sorry, Philip` - I thought the next few releases of browsers were switching over to do that...

[13:31:05.0000] <AryehGregor>
So if someone does tons of prefixes and then <meta charset>, it's already invalid.

[13:32:00.0000] <AryehGregor>
We all wish we could do UTF-8 as default, but it would horribly break the web.

[13:34:00.0000] <manu-db>
This makes me sad - it means that we probably won't be able to limit @prefix to the root element (because others in the WG feel strongly about not screwing up the meta charset detection)

[13:34:01.0000] <Philip`>
/me wonders if prefixes really need to be on an element, rather than a <meta name=prefixes ...> (where you look up the first such meta element in the first head element of the document, in place of looking up the attribute on the root element)

[13:34:02.0000] <manu-db>
Philip` - we do that for the BASE element... and everybody hated having to pre-process the document for that...

[13:34:03.0000] <manu-db>
although, that's an idea that really hasn't had much time in the spotlight in the RDFa work... I'll raise it again on the next telecon.

[13:35:00.0000] <manu-db>
Philip` - one of the other issues is that RDFa works on all XML dialects, and this isn't a problem in those other XML dialects... just HTML, IIRC

[13:35:01.0000] <manu-db>
s/on/in/

[13:35:02.0000] <manu-db>
(and XHTML)

[13:37:00.0000] <Philip`>
manu-db: Would it require more processing than the attribute, given that <html><body>millions of lines<html prefix=...> will put the prefix attribute on the root element in HTML5 parsers so you presumably need some way to deal with attributes changing later (even without scripting)?

[13:37:01.0000] <Ms2ger>
Philip`, rdfa-using authors don't do that

[13:39:00.0000] <Philip`>
Evil ones might

[13:39:01.0000] <Ms2ger>
You don't use rdfa, do you?

[13:39:02.0000] <manu-db>
We do have a large number of evil authors... wait... no, that's not true at all.

[13:39:03.0000] <manu-db>
I think that would be a tough sell, Philip` - it's not just RDFa-using authors - it's any authors.

[13:39:04.0000] <Ms2ger>
manu-db, that would imply you have a large number of authors ;)

[13:39:05.0000] <Philip`>
Ms2ger: Only when trying to break it :-)

[13:40:00.0000] <manu-db>
Ms2ger - :P - no response... troll feeding and all :)

[13:40:01.0000] <Ms2ger>
/me bows

[13:41:00.0000] <manu-db>
Philip` - telling them to jump through those kinds of hoops because of the way character encoding works in browsers is going to be confusing to authors.

[13:43:00.0000] <manu-db>
Does anybody have any hard numbers on the number of documents on the Web that are non-ASCII, non-UTF8 and are served up without any sort of character encoding information?

[13:43:01.0000] <AryehGregor>
A very significant percentage.

[13:44:00.0000] <Philip`>
http://philip.html5.org/data/charsets.html may be relevant

[13:44:01.0000] <Philip`>
and/or may be confusing

[13:45:00.0000] <MikeSmith>
manu-db: fwiw, thanks for being frank about the task-force idea

[13:46:00.0000] <MikeSmith>
manu-db: you continue to be wrong about one thing, though

[13:46:01.0000] <manu-db>
MikeSmith: np - just trying to be pragmatic.

[13:46:02.0000] <Ms2ger>
/me trolls: Just one?

[13:46:03.0000] <MikeSmith>
yep

[13:46:04.0000] <manu-db>
/me would like to point out that it's far more than one.

[13:47:00.0000] <manu-db>
/me peers at Ms2ger.

[13:47:01.0000] <Ms2ger>
Also, I managed to read "pragmatic" as "pedantic" somehow

[13:47:02.0000] <manu-db>
Well, don't keep me waiting, MikeSmith - I need to add it to my list :P

[13:47:03.0000] <MikeSmith>
manu-db:  I mean, the in-hindsight contention that proceeding with publication of the microdata and HTML+rdfa drafts in parallel was a mistake

[13:47:04.0000] <MikeSmith>
it was not a mistake

[13:48:00.0000] <MikeSmith>
it was pretty much an inevitability, given the circumstances

[13:49:00.0000] <manu-db>
I do personally feel it was a mistake - meaning, the ideal path would have been for a number of us involved to work more closely with one another and get behind some unified approach.

[13:49:01.0000] <MikeSmith>
true of many things

[13:49:02.0000] <MikeSmith>
but the ideal path is usually not available

[13:50:00.0000] <MikeSmith>
and the fact is that microdata is not bad

[13:50:01.0000] <manu-db>
I think it could've been if everybody had been a bit less hard headed about working together.

[13:50:02.0000] <MikeSmith>
well, lots of blame to be spread around there

[13:51:00.0000] <manu-db>
right, and I'm not that interested in spreading blame - more interested in figuring out how to not make the same series of mistakes again (as a community)

[13:51:01.0000] <MikeSmith>
Hixie and Henri made their concerns known years ago

[13:51:02.0000] <MikeSmith>
and they were largely blown off

[13:51:03.0000] <manu-db>
and so did we...

[13:51:04.0000] <manu-db>
and I disagree that they were blown off - I made it a point to discuss their issues in the RDFa 1.1 group... it was a focus.

[13:52:00.0000] <MikeSmith>
you did, yeah

[13:52:01.0000] <manu-db>
I know there is this belief that we didn't... but we spent a very long time discussing the concerns.

[13:52:02.0000] <MikeSmith>
you

[13:52:03.0000] <MikeSmith>
others did not

[13:52:04.0000] <AryehGregor>
IMO, this is a case where people just were going to disagree in the end, and the only way to move forward was to have two approaches and see how it went.  I don't think any compromise would have been acceptable to all the key people.

[13:52:05.0000] <AryehGregor>
I definitely don't think it's worth rehashing at this point.

[13:52:06.0000] <manu-db>
A compromise would have been acceptable to me - and many others. (just for the record)

[13:52:07.0000] <manu-db>
anyway - we have what we have now...

[13:53:00.0000] <manu-db>
let's see how the market responds

[13:53:01.0000] <MikeSmith>
it's worth rehashing in order to make sure we all don't waste massive amounts of time repeating the same mistakes again

[13:53:02.0000] <manu-db>
Prediction: confusion, followed by taking sides, followed by a long-drawn out back and forth and finally, one comes out successful.

[13:54:00.0000] <manu-db>
I think the core of the rift had to do with the old XHTML2 vs. HTML5 stuff carrying over into the RDFa work.

[13:54:01.0000] <MikeSmith>
I think it remains  unclear whether authors and content providers will ever on a large scale take time to mark up their content with any additional semantic markup at all

[13:54:02.0000] <manu-db>
and having not had anything to do with the XHTML2 stuff, the new folks in the RDFa work were kinda thrown into this mess that goes back 10+ years. Lots of bad blood in the Web standards game... most of which the new folks don't care one bit about... but keep getting smacked in the face with it.

[13:55:00.0000] <manu-db>
MikeSmith: I agree - I don't think anyone in the RDFa group believes that just because the spec exists that people are going to suddenly care about data.

[13:55:01.0000] <manu-db>
The whole semantics in HTML bit is what we're trying now because the "publish your data in a separate file" bit wasn't working.

[13:56:00.0000] <manu-db>
and really, people do this sort of stuff because there is something in it for them - SEO is the first carrot.

[13:56:01.0000] <Hixie>
tim does, or at least wouldn't acknowledge that they might not a few years ago when i spoke to him about it

[13:56:02.0000] <manu-db>
and perhaps the Web Payments stuff will be the second carrot.

[13:56:03.0000] <manu-db>
Hixie: TimBL thinks what? (I didn't follow)

[13:56:04.0000] <MikeSmith>
XHTML2 was a very big turn in the wrong direction -- as was the whole XHTML modularization focus. Anybody who still doesn't realize that, well, it's hard to have a discussion with them

[13:57:00.0000] <Hixie>
tim thinks people are going to use semantic markup in their output

[13:57:01.0000] <Hixie>
and that it'll be usable by tools

[13:57:02.0000] <Hixie>
i.e. that it won't be like existing markup, which is 90% syntactically invalid and semantically dubious

[13:57:03.0000] <manu-db>
MikeSmith - yes, and I'm one of those people that is happy about the HTML5 stuff winning out - but keep getting treated as if we're all still gung-ho about XHTML2 (which I'm not - but do appreciate the work that was done on that stuff as well)

[13:58:00.0000] <manu-db>
Hixie: I'm less certain about it - I don't know what will happen... I'm very bad at predicting the future. I think it's an idea worth trying, however... because if people /do/ start publishing more data - some of it may be very useful for the future of the Web.

[13:59:00.0000] <manu-db>
We're specifically interested in it because we had no other way to make PaySwarm truly decentralized...

[13:59:01.0000] <manu-db>
http://manu.sporny.org/2011/better-world/

[13:59:02.0000] <manu-db>
(Building a Better World with Web Payments)

[13:59:03.0000] <Hixie>
i'm just saying that when you say "I don't think anyone in the RDFa group believes that...", my experience is that you are wrong

[13:59:04.0000] <manu-db>
RDFa is a means to an end for us...

[14:00:00.0000] <MikeSmith>
manu-db: you are exceptional in a lot of ways. If more people involved in this took the approach you have, we'd be a lot further along

[14:00:01.0000] <Hixie>
(at least for a definition of "RDFa group" that includes prominent people like tim)

[14:00:02.0000] <manu-db>
Hixie: I've never heard any of them say that they believe that the future is definitely going to be data in pages...

[14:00:03.0000] <Hixie>
then you haven't been listening :-)

[14:00:04.0000] <manu-db>
Hixie: but we do want that to happen... we're just not sure it will happen because it's up to the authors and CMSes now.

[14:01:00.0000] <Hixie>
we haven't even gotten people to use html correctly on any sort of widespread basis

[14:01:01.0000] <manu-db>
I don't include TimBL in the RDFa group... I was being more precise - the RDFa WG members that show up to all the calls.

[14:01:02.0000] <manu-db>
Hixie: I think HTML usage has been getting better over the years... people now care more about standards than they ever have.

[14:01:03.0000] <Hixie>
so the idea that they'll do something even more complicated than html, like microdata, with any sort of consistency or accuracy is imho wildly optimistic

[14:02:00.0000] <Hixie>
manu-db: do you have data backing that up?

[14:02:01.0000] <manu-db>
If you don't believe that, then I don't know why you're doing what you do. :)

[14:02:02.0000] <Hixie>
what do you think we're doing?

[14:02:03.0000] <manu-db>
No, I don't have hard data - just anecdotal evidence about what people are writing about online and conversations I have with web development firms over the past 10+ years.

[14:03:00.0000] <manu-db>
Hixie: Who is "we"? I was talking about /you/ :)

[14:03:01.0000] <manu-db>
That is not to say that people don't get stuff wrong...

[14:04:00.0000] <manu-db>
but there is a greater "moral pressure" to get things right now than there was in the late 1990s.

[14:04:01.0000] <Hixie>
ok what do you think I'm doing?

[14:04:02.0000] <AryehGregor>
Look, is this discussion really any different from the ones everyone has been having for the last few years?  Do you think it's really going to be more useful this time than every time before?

[14:05:00.0000] <Hixie>
i think the moral pressure you're seeing is biased by the circles you go in. I've seen no actual evidence that people are using HTML more correctly than than 5, 10, or 15 years ago

[14:05:01.0000] <AryehGregor>
It's the same people making basically the same arguments over and over again.  Let's just wait and see who's right.

[14:05:02.0000] <manu-db>
I think you're spec'ing what people practice online.

[14:05:03.0000] <manu-db>
and you're attempting to do it in a way that is consistent across all browsers.

[14:06:00.0000] <manu-db>
(even if the practice is bad)

[14:06:01.0000] <Hixie>
then why is what i said above a factor in why i do what i do?

[14:06:02.0000] <manu-db>
(and that happens in very few of the cases)

[14:06:03.0000] <Philip`>
/me guesses perceived XML-related gunghosity may be because RDFa dragged in a load of XML baggage from Namespaces (and partly Canonicalisation via XML literals etc), as well as from RDF itself, so most of the discussion was about problems and complexity induced by those things rather than by the interesting functionality of RDFa

[14:07:00.0000] <manu-db>
"what I said" - what specifically are you referring to?

[14:08:00.0000] <manu-db>
(because you said a number of things above)

[14:08:01.0000] <Hixie>
I said "so the idea that they'll do something even more complicated than html, like microdata, with any sort of consistency or accuracy is imho wildly optimistic" and you said "If you don't believe that, then I don't know why you're doing what you do." but I don't see why

[14:09:00.0000] <manu-db>
My statement wasn't in response to what you wrote on the line above

[14:09:01.0000] <Hixie>
oh

[14:09:02.0000] <Hixie>
what was it in response to?

[14:09:03.0000] <manu-db>
I think HTML usage has been getting better over the years... people now care more about standards than they ever have... If you don't believe that, then I don't know why you're doing what you do.

[14:10:00.0000] <Hixie>
people as in authors or people as in browser vendors?

[14:10:01.0000] <manu-db>
^^^ All I was saying is that I think you think the work that you're doing is having a positive impact - otherwise you wouldn't be doing it... so your belief is that things are getting better.

[14:10:02.0000] <manu-db>
or, you're a masochist. :P

[14:10:03.0000] <Ms2ger>
Or both

[14:11:00.0000] <manu-db>
XOR

[14:11:01.0000] <Hixie>
even if authors entirely ignored the spec, i would still think it's important to have a single spec so that browsers can converge on interoperable behaviour

[14:11:02.0000] <Hixie>
(in fact, i think authors by and large do entirely ignore the spec)

[14:11:03.0000] <manu-db>
I don't think they do - I think there is a bias when you see errors... to only see the errors and not the sea of things that are being done correctly out there.

[14:12:00.0000] <Hixie>
there's no bias. I've done actualy objective studies on the google search corpus.

[14:12:01.0000] <manu-db>
/me wishes those studies were available to the public.

[14:12:02.0000] <Hixie>
looking to see what percentage of pages have conformance errors

[14:12:03.0000] <Philip`>
(Validator developers don't ignore the spec and a non-negligible number of authors don't ignore validators, so I'd assume that's how the spec most influences authors)

[14:12:04.0000] <manu-db>
I don't think just looking at conformance errors are interesting... I think the type of conformance errors are very important.

[14:13:00.0000] <Hixie>
Philip`: right

[14:13:01.0000] <manu-db>
exactly, Philip`

[14:13:02.0000] <Hixie>
manu-db: well, you're welcome to do your own studies like Philip` has :-)

[14:13:03.0000] <Hixie>
but i have not seen any evidence that things are changing

[14:13:04.0000] <Philip`>
(which incidentally means the non-machine-checkable conformance requirements are a waste of everyone's time)

[14:13:05.0000] <Hixie>
and i don't see why that would affect whether i do this work

[14:13:06.0000] <manu-db>
Yeah - we really do need public data on this.

[14:14:00.0000] <manu-db>
That's why I'm pursuing this deal w/ this Web crawl company - so we can get some of this data out in the public and start making more data-driven decisions.

[14:14:01.0000] <Philip`>
Getting reliable data for a single moment in time is hard enough - trying to compare it over years seems pretty much impossible :-(

[14:14:02.0000] <Philip`>
Any real changes will be drowned out by sampling biases

[14:15:00.0000] <Philip`>
(assuming it was even possible to get historic samples)

[14:15:01.0000] <manu-db>
Philip` but it's one of the only ways we can stop these ridiculous "I think X" ... "Oh yeah, well I think Y" discussions

[14:15:02.0000] <manu-db>
/me wonders if you could crawl the Internet archive for historical samples?

[14:16:00.0000] <manu-db>
I really think that we should be doing this sort of data collection as a community... anyway - working on it.

[14:16:01.0000] <Philip`>
Yeah, some data is still better than no data :-)

[14:16:02.0000] <manu-db>
Hixie - I trust that you have done the background research - but without having the data publicly available, it's not science.

[14:16:03.0000] <manu-db>
and I have no idea why Google wouldn't agree to release those figures publicly.

[14:17:00.0000] <Hixie>
manu-db: i don't think anyone would claim this was science

[14:17:01.0000] <manu-db>
There are good bits of Microformats, Microdata and RDFa that are witchcraft - no solid data behind the decisions.

[14:18:00.0000] <manu-db>
Hixie: I don't think it would hurt us to go more toward science than witchcraft. :)

[14:18:01.0000] <Hixie>
well sure

[14:18:02.0000] <Hixie>
that's up to the people doing the language design

[14:18:03.0000] <manu-db>
I think it's up to the community, not just the handful of people doing language design.

[14:19:00.0000] <manu-db>
/me has to run to a telecon.

[14:20:00.0000] <manu-db>
Good chat - thanks all for answering those two questions... much appreciated.

[15:50:00.0000] <AryehGregor>
The fact that we have LLVM-to-JavaScript compilers that work in practice and are efficient enough to use for real-world applications is somehow terrifying: https://github.com/kripken/emscripten/wiki

[15:51:00.0000] <AryehGregor>
Like a speech synthesizer.  Written in C++, compiled to LLVM, recompiled to JavaScript, actually runs in a web browser.

[15:51:01.0000] <AryehGregor>
It actually works.  http://syntensity.com/static/espeak.html

[15:51:02.0000] <AryehGregor>
/me boggles

[15:52:00.0000] <Philip`>
It's possible to write code and run it on a computer? Amazing!

[15:53:00.0000] <AryehGregor>
"The development versions of the top JavaScript engines today can run code compiled from C++ only 3-5X slower than a fast C++ compiler, and getting even better."

[15:53:01.0000] <David_Bradbury>
Any ideas if CSS3 gradients will be able to be used on boarders down the road?

[15:53:02.0000] <zewt>
... uh huh

[15:53:03.0000] <AryehGregor>
David_Bradbury, TabAtkins is the one to ask about that.

[15:54:00.0000] <roc>
depends on how you want to use them

[15:54:01.0000] <David_Bradbury>
Kk - TabAtkins: Any ideas if CSS3 gradients will be able to be used on boarders down the road? :D

[15:54:02.0000] <roc>
current CSS3 drafts already let you use gradients with border-image

[15:54:03.0000] <roc>
although Gecko at least doesn't support that yet AFAIK

[15:55:00.0000] <David_Bradbury>
Don't you have to use an actual pre-rendered image to do that or can you use an ID or something as an image?

[15:55:01.0000] <TabAtkins>
David_Bradbury: In border-image.

[15:55:02.0000] <TabAtkins>
David_Bradbury: No, border-image just takes an image.

[15:55:03.0000] <TabAtkins>
Anything which is defined as part of the <image> type, theoretically.

[15:55:04.0000] <AryehGregor>
But doesn't border-image require an image in a very particular format?

[15:55:05.0000] <TabAtkins>
No.

[15:56:00.0000] <David_Bradbury>
Well, I mean I could use an image to do a background gradient on a div

[15:56:01.0000] <David_Bradbury>
the point of having an api that allows you to do it is so that you don't need to do that

[15:56:02.0000] <AryehGregor>
I mean, you have to provide one single image for all four borders and then slice it up somehow, no?

[15:56:03.0000] <AryehGregor>
You can't use different images for different sides, say.  Or could you fake it by just using complicated gradients with a bunch of extra stops?

[15:56:04.0000] <David_Bradbury>
No, you can use a single image using border-image

[15:57:00.0000] <roc>
yeah that's why I said it depends on *how* you want to use a gradient on the border

[15:57:01.0000] <annevk>
jamesr, UTF-32 is verboten :p

[15:57:02.0000] <roc>
some uses could be done with border-image:linear-gradient or border-image:radial-gradient; others can't

[15:57:03.0000] <David_Bradbury>
but it seems silly to have to use an image for a gradient border if you're going to allow people to define gradients as backgrounds

[15:58:00.0000] <jamesr>
annevk: it's not a terrible encoding for some things, actually

[15:58:01.0000] <roc>
yes it is

[15:58:02.0000] <jamesr>
annevk: constant-time indexing for the full range of unicode codepoints

[15:58:03.0000] <roc>
what are the use-cases for constant-time indexing?

[15:58:04.0000] <jamesr>
if you have a lot of codepoints that are 2^17th and larger it's better than utf-16

[15:58:05.0000] <roc>
David_Bradbury: you need to describe what effect you're trying to get

[15:59:00.0000] <jamesr>
on the web it's useless

[15:59:01.0000] <David_Bradbury>
A one-pixel linear gradient as a boarder?

[15:59:02.0000] <David_Bradbury>
border*

[15:59:03.0000] <AryehGregor>
David_Bradbury, in theory, you can use gradients anywhere you can use images, as I understand it.

[15:59:04.0000] <TabAtkins>
Firefox has the non-standard border-colors property.

[15:59:05.0000] <TabAtkins>
So you can manually implement a gradient.

[16:00:00.0000] <AryehGregor>
jamesr, you basically never actually need constant-time code-point-based indexing, though.  Either you're iterating through the whole string, or you can use byte-based indexing instead.

[16:00:01.0000] <David_Bradbury>
I guess I mean to ask if there is something planned in the spec down the road

[16:00:02.0000] <annevk>
jamesr, you still don't get all characters

[16:00:03.0000] <annevk>
jamesr, well, compound characters or whatever they are called

[16:00:04.0000] <David_Bradbury>
Hi davidb, I'm David B.

[16:01:00.0000] <davidb>
hi David_Bradbury

[16:02:00.0000] <jamesr>
annevk: surrogate pairs? those only exist in utf-16

[16:02:01.0000] <jamesr>
annevk: all unicode codepoints are less than 2^22 or 2^23, so they all take the same space in utf32

[16:02:02.0000] <jamesr>
none of this is important for the web so forget i asked :)

[16:03:00.0000] <TabAtkins>
annevk: jamesr is right.  In utf-32 you just store the codepoint directly.

[16:03:01.0000] <jgraham>
jamesr: I suspect annevk means glyphs that are formed froma character plus a combining character

[16:04:00.0000] <jgraham>
So constant time indexing of codepoints doesn't give you constant time indexing of anything users care about

[16:08:00.0000] <annevk>
jamesr, no, diacritical marks as separate codepoints

[16:08:01.0000] <annevk>
yeah, what jgraham said

[16:09:00.0000] <annevk>
UTF-16 is nonsense too

[16:09:01.0000] <annevk>
but we cannot kill it

[16:09:02.0000] <annevk>
it's baked into fricking ECMAScript and the DOM :(

[16:09:03.0000] <annevk>
maybe fricking baked*

[16:09:04.0000] <zewt>
utf-16 is great: it takes the advantages of utf-8 and the advantages of a fixed-width encoding like ucs4 ... then throws them away and keeps the disadvantages of both

[16:10:00.0000] <jamesr>
annevk: UCS-2 is baked in

[16:10:01.0000] <jamesr>
ECMAScript has basically no UTF16 support

[16:10:02.0000] <jamesr>
if it has any

[16:10:03.0000] <jamesr>
surrogate pairs = you deal with it, sucker

[16:13:00.0000] <jamesr>
fromCharCode() and codePointAt() only handle codepoints <= 2^16

[16:13:01.0000] <annevk>
that's kind of what I meant

[16:13:02.0000] <annevk>
but I should have said 16-bit code units

[16:23:00.0000] <gsnedders>
jamesr: {decode,encode}URL{Component,} deals with surrogates! But I think that's it.

[16:24:00.0000] <gsnedders>
Harmony will introduce some sort of support for full Unicode.

[16:27:00.0000] <kennyluck>
oh really, that sounds interesting.

[16:28:00.0000] <annevk>
hmm something is going wrong with the tracker again

[16:29:00.0000] <annevk>
oh oops

[16:29:01.0000] <annevk>
I had unchecked "Show editorial changes"

[16:29:02.0000] <Hixie>
should add something somewhere that says "23 editorial changes hidden"

[16:31:00.0000] <annevk>
yeah, or attempt to kill that option again over zcorpan's objections :p

[16:32:00.0000] <annevk>
Hixie, btw, what is the difference between EventSource and WebSocket when it comes to garbage collection?

[16:32:01.0000] <annevk>
Hixie, WebSocket is way more complicated

[16:33:00.0000] <Hixie>
the main difference is that websockets speak back to the host

[16:34:00.0000] <Hixie>
also maybe eventsource is buggy? doesn't look like it preserves the object for load/error events...

[16:34:01.0000] <annevk>
e.g. only WebSocket has "make disappear" steps

[16:35:00.0000] <annevk>
EventSource talks about a strong reference instead

[16:35:01.0000] <annevk>
maybe EventSource was done and then later you decided upon another way of doing this?

[16:36:00.0000] <annevk>
I'll file a bug on EventSource so you can take a look, let me know, and maybe I can then finally update XMLHttpRequest

[16:36:01.0000] <Hixie>
the make disappear stuff is entirely because websocket speaks back

[16:37:00.0000] <Hixie>
whereas eventsource is just a one-way connection

[16:37:01.0000] <Hixie>
websockets has a closing handshake, etc

[16:38:00.0000] <annevk>
well eventsource has a request and response

[16:38:01.0000] <Hixie>
yeah but it's http, so the connection can just be dropped arbitrarily

[16:38:02.0000] <Hixie>
whereas websocket has this closing handshake thing

[16:38:03.0000] <annevk>
okay

[16:38:04.0000] <annevk>
so XHR should be modeled after EventSource

[16:39:00.0000] <annevk>
I'll file a bug on the other events

[16:51:00.0000] <annevk>
haha http://imgur.com/xFfzw


2011-08-18
[17:07:00.0000] <heycam>
Hixie, can I give you a perl script rather than a diff to remove the "in"s?

[17:08:00.0000] <heycam>
Hixie, http://pastebin.mozilla.org/1302853

[17:08:01.0000] <heycam>
I didn't check the output extensively, but it looks right on a first glance :)

[17:12:00.0000] <Philip`>
/me notes the $a thing could likely be replaced by "..", like "while (<>) { if (/<pre class="idl/ .. /<\/pre>/) { s/whatever/ } }"

[17:12:01.0000] <Philip`>
which is one of my top 50 favourite Perl features

[17:14:00.0000] <annevk>
yay for removing 'in'

[17:14:01.0000] <annevk>
now modules

[17:14:02.0000] <Hixie>
heycam: i'll try, but it scares me!

[17:14:03.0000] <heycam>
Philip`, wow I don't know about that feature

[17:15:00.0000] <heycam>
Philip`, is it an implicit while loop there?

[17:15:01.0000] <Philip`>
heycam: The "while (<>) {" looks pretty explicit to me :-)

[17:15:02.0000] <Hixie>
what is ".."? is that new?

[17:16:00.0000] <heycam>
oh I see, so it's not an implicit loop, just an implicit variable

[17:16:01.0000] <Hixie>
wow i never knew about the bistable .. operator

[17:16:02.0000] <Philip`>
Scalar ".." is a flip-flop

[17:17:00.0000] <Philip`>
It goes true when the first operand is true, and false when the second goes true

[17:18:00.0000] <Philip`>
It's existed forever, I think

[17:18:01.0000] <heycam>
that's kind of awesome

[17:19:00.0000] <Hixie>
that's pretty sweet

[17:20:00.0000] <Philip`>
(Given http://www.cs.cmu.edu/cgi-bin/perl-man it seems to have existed since before Perl 5)

[17:22:00.0000] <Hixie>
ok well i found some problems with this scrtip

[17:22:01.0000] <Hixie>
e.g. it missed "[Clamp] in..."

[17:22:02.0000] <heycam>
ah!

[17:22:03.0000] <heycam>
do you want me to fix it or can you handle it?

[17:23:00.0000] <Hixie>
i found four. fixed them.

[17:23:01.0000] <heycam>
ah cool. only four extended attributes on arguments? fewer than i thought there'd be.

[17:24:00.0000] <Hixie>
two [Clamp]s and two [AllowAny]s

[17:24:01.0000] <Hixie>
one of the [Clamp]s is movign to the arraybuffer spec soon

[17:26:00.0000] <heycam>
cool

[17:26:01.0000] <Hixie>
no webidl errors

[17:26:02.0000] <Hixie>
if we missed any i guess we'll find out when dom updates the checker

[17:27:00.0000] <Hixie>
is there a bug # for this?

[17:29:00.0000] <Hixie>
couldn't find it again. nevermind, i'll just commit without a bug #.

[17:31:00.0000] <annevk>
there's no bug

[17:31:01.0000] <Hixie>
k

[18:12:00.0000] <Hixie>
did anyone else get invited to this schema.org thing next month? (foolip?)

[18:48:00.0000] <mkanat>
"in which the transparent element finds itself" seems redundant in this paragraph: http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#transparent

[18:48:01.0000] <mkanat>
Should I file a bug?

[18:50:00.0000] <TabAtkins>
Ah, now I understand the scalar .. operator.  It's defining a range implicitly over multiple invocations.

[19:26:00.0000] <yuhong>
Just found this: http://picturoku.blogspot.com/2011/08/diaries-of-vulnerability.html

[22:58:00.0000] <zcorpan>
Hixie: &amp;lt in r6493 should be &lt;

[22:58:01.0000] <Hixie>
can you paste the rest of the line?

[22:59:00.0000] <zcorpan>
+ <strong>&lt;option value=""> Select unit type &lt/option></strong>

[22:59:01.0000] <Hixie>
oh yeah i fixed that already

[22:59:02.0000] <zcorpan>
ok

[23:00:00.0000] <Hixie>
thanks though!

[23:28:00.0000] <zcorpan>
ooh, meta robots, why haven't we thought of that before?

[02:08:00.0000] <asmodai>
hahaha: http://imgur.com/1nZfG

[02:30:00.0000] <rimantas>
yup. Saw the post on HN, did not expect it to be _that_ bad

[02:35:00.0000] <annevk>
Hixie, in the base URL change steps change you used <span> instead of <dfn>

[02:38:00.0000] <annevk>
zcorpan, did you see my comments in the XHR bugs?

[02:39:00.0000] <zcorpan>
nope

[02:41:00.0000] <annevk>
it seems you don't get email for bugs

[02:41:01.0000] <zcorpan>
indeed. i think i've set that up since i usually get email anyway on public-html-bugzilla etc

[02:42:00.0000] <zcorpan>
but i'm not subscribed to public-webapps-bugzilla if there's such a thing

[02:44:00.0000] <zcorpan>
annevk: seems like a bug in eventsource if it can get still get open/error events and be GC

[02:44:01.0000] <annevk>
yeah, filed that bug

[02:45:00.0000] <annevk>
but EventSource also uses strong reference rather than the language WebSocket uses

[02:45:01.0000] <annevk>
heycam, idea I had last night

[02:45:02.0000] <zcorpan>
*shrug*

[02:45:03.0000] <annevk>
heycam, Web IDL defines DOMException

[02:46:00.0000] <annevk>
heycam, other than that we keep exceptions the way they are and if people want new exceptions they define extensions to Web IDL, to be coordinated via the WebApps WG

[02:46:01.0000] <zcorpan>
maybe websockets should use "strong reference" also

[02:46:02.0000] <heycam>
annevk, to solve the coordination problem?

[02:46:03.0000] <annevk>
you can say shrug, but if you don't really know what's going on that's quite confusing

[02:47:00.0000] <annevk>
heycam, yeah, we just consolidate all exceptions into DOMException and just like DOMString and DOMTimeStamp Web IDL takes care of it

[02:47:01.0000] <annevk>
heycam, that way you only need a reference to Web IDL to make use of DOMException too

[02:47:02.0000] <heycam>
annevk, maybe. I'm not too fussed where DOMException itself lives -- I would be fine with it in DOM Core still.

[02:48:00.0000] <zcorpan>
annevk: agree it should use the same language if it means the same

[02:48:01.0000] <heycam>
/me has to run

[02:49:00.0000] <annevk>
heycam, the idea is more that there would be nothing apart from DOMException

[02:49:01.0000] <annevk>
guess I'll add that to some bug somewhere

[02:50:00.0000] <zcorpan>
annevk: did you file a bug about language difference between websocket/eventsource?

[02:52:00.0000] <annevk>
I think I put that in the same bug

[02:52:01.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13812

[02:52:02.0000] <annevk>
could use some elaboration maybe

[02:53:00.0000] <david_carlisle>
hixie: (who isn't here) Henri's parser doesn't like the complete spec: Recoverable error on line 31460 column 13 of complete.html:

[02:53:01.0000] <david_carlisle>
  SXXP0003: Error reported by XML parser: No dt element in scope but a dt end tag seen.

[02:53:02.0000] <david_carlisle>
would it be possible to remove the /dt from   line  31460  which is currently </dt></dt></dt><dd>

[02:55:00.0000] <jgraham>
david_carlisle: File a big?

[02:55:01.0000] <jgraham>
*bug

[02:55:02.0000] <jgraham>
But it could be an error somewhere in the pipeline ofc

[02:55:03.0000] <jgraham>
In fact I'm not really sure how Hixie could cause that

[02:56:00.0000] <david_carlisle>
jgraham: yes i suppose I should, the spurious dts are explict in wget http://www.whatwg.org/specs/web-apps/current-work/complete.html, not in a file I had pre-processed

[02:56:01.0000] <annevk>
also, reported by XML parser?

[02:57:00.0000] <david_carlisle>
oh that was really Henri's html 5 parser, just that I was using its sax interface so saxon thought itwas xml...

[02:58:00.0000] <annevk>
crazy ;p

[02:59:00.0000] <zcorpan>
probably a bug in the html parser that's used in anolis or something

[03:00:00.0000] <jgraham>
david_carlisle: Right, but Hixie's text goes through a parser -> tree -> serializer pipeline

[03:00:01.0000] <zcorpan>
where <dt> doesn't close dt

[03:00:02.0000] <jgraham>
So it seems unlikely that he could do anything that would cause a stray </dt>

[03:00:03.0000] <david_carlisle>
ah but I got a pdf out of xslt and tex with only using a few bytes of memory which seems to be rather less than the current pdf generation:-)

[03:01:00.0000] <annevk>
sweet

[03:01:01.0000] <david_carlisle>
jgraham: you're all using html rather than xml tools,  things going wrong iand appearing in stray places is the expected behaviour isn't it:-)

[03:02:00.0000] <jgraham>
david_carlisle: Well in this case the input is being processed by an xml toolset with a hacked on html parser

[03:02:01.0000] <jgraham>
(libxml2)

[03:02:02.0000] <david_carlisle>
just like me then

[03:02:03.0000] <jgraham>
(specifically lxml)

[03:03:00.0000] <zcorpan>
isn't there a good perf html5 parser these days we can use instead?

[03:04:00.0000] <jgraham>
But I don't really see how that could cause the bug; you could have thought that as long as it generates a tree it can only be the serliaizer that causes ill-formed output

[03:04:01.0000] <zcorpan>
jgraham: if the input is <dl><dt>foo<dt>foo<dt>foo<dd>bar</dl>

[03:04:02.0000] <zcorpan>
jgraham: and the parser creates a tree where teh <dt>s are nested

[03:05:00.0000] <jgraham>
/me wonders if david_carlisle will use his xslt script to generate fortran bindings for the dom apis, for good measure :p

[03:05:01.0000] <zcorpan>
then voilà

[03:05:02.0000] <david_carlisle>
jgraham: want some?

[03:05:03.0000] <jgraham>
zcorpan: Oh, yes, that would do it I suppose

[03:06:00.0000] <jgraham>
david_carlisle: I'm good thanks :)

[03:06:01.0000] <david_carlisle>
jgraham:  yes the file has the same number of <dt> and </dt> it's just that the html5 parser has closed some dt by the time it gets to the /dt/dt bit at that line

[03:06:02.0000] <jgraham>
So, uh, you could file a bug on the spec to use explicit close tags I guess

[03:07:00.0000] <zcorpan>
or flip a pref in the serializer to omit </dt>

[03:08:00.0000] <jgraham>
Or you could help hsivonen get the gecko html parser working for non-gecko uses

[03:10:00.0000] <annevk>
libxml2 guys should get their act together

[03:21:00.0000] <Dimitri87>
There is a failure on the following wiki page. http://wiki.whatwg.org/wiki/HTML_vs._XHTML#HTML_Elements_with_Optional_Tags There stands twice "dt" should one not mean "dd"?

[03:21:01.0000] <annevk>
zcorpan, so I think garbage collection rules can be pretty simple

[03:22:00.0000] <annevk>
if XMLHttpRequest is a) fetching and b) has event listeners for ... don't

[03:22:01.0000] <annevk>
otherwise, terminate any ongoing fetching algorithm and die happily

[03:29:00.0000] <foolip>
Hixie, I was invited, but I doubt Opera will be putting me on a plane to California for it

[03:36:00.0000] <jgraham>
Since Google have all the money they should have more conferences that they have to travel for :)

[03:47:00.0000] <Dimitri87>
Is that the right location for my report or should i do report this wrong info any where else?

[03:48:00.0000] <annevk>
you can fix it yourself ;)

[03:48:01.0000] <annevk>
it's a wiki

[03:54:00.0000] <zcorpan>
annevk: sounds good

[03:55:00.0000] <zcorpan>
annevk: except you can GC if there's just a loadstart listener and loadstart has already been fired

[04:30:00.0000] <Dimitri871>
annevk: ok done!

[04:32:00.0000] <annevk>
thanks!

[04:42:00.0000] <annevk>
zcorpan, I guess I will not mention them

[04:42:01.0000] <annevk>
they will only dispatch if you invoke send()

[04:42:02.0000] <annevk>
and invoke before send() returns

[04:47:00.0000] <zcorpan>
ah

[04:53:00.0000] <Ms2ger>
/me likes how Hixie added another Avenue Q reference

[04:55:00.0000] <zcorpan>
heh

[04:58:00.0000] <annevk>
is there a wiki page documenting the references? :)

[04:59:00.0000] <Ms2ger>
No, that should be hosted at IANA

[05:00:00.0000] <annevk>
damn it Julian

[05:01:00.0000] <MikeSmith>
annevk: I changed the XHR link to point to -2 and added a link to Adam's URL API doc

[05:01:01.0000] <MikeSmith>
but still not sure what crypto stuff you meant

[05:01:02.0000] <MikeSmith>
DOMCrypt or the random-number thing

[05:10:00.0000] <annevk>
MikeSmith, DOMCrypt I guess, but maybe the other too?

[05:10:01.0000] <annevk>
whatever is on window.crypto

[05:10:02.0000] <annevk>
or going to be there

[05:10:03.0000] <MikeSmith_>
hai

[05:17:00.0000] <annevk>
zcorpan, http://dev.w3.org/2006/webapi/XMLHttpRequest-2/#garbage-collection

[05:17:01.0000] <annevk>
oh great

[05:18:00.0000] <annevk>
that needs some XHR1 markers

[05:19:00.0000] <smaug____>
annevk: why its state is HEADERS_RECEIVED, or its state is LOADING ?

[05:19:01.0000] <smaug____>
er, nm

[05:20:00.0000] <smaug____>
hmm

[05:20:01.0000] <smaug____>
annevk: yes, why "its state is HEADERS_RECEIVED, or its state is LOADING"

[05:21:00.0000] <annevk>
because those are states while it is fetching data?

[05:21:01.0000] <smaug____>
ah, I'm reading that wrong

[05:22:00.0000] <smaug____>
would it be easier to say something like "send() flag is set and state is not DONE"

[05:24:00.0000] <annevk>
yeah maybe

[05:24:01.0000] <smaug____>
but that depends on how the spec defines send() flag

[05:24:02.0000] <annevk>
currently the spec sort of associates the send() flag with OPENED

[05:24:03.0000] <smaug____>
I mean in which cases it is cleared

[05:24:04.0000] <annevk>
but only in prose

[05:25:00.0000] <annevk>
it would be either UNSENT or DONE btw

[05:27:00.0000] <zcorpan>
annevk: looks good

[05:29:00.0000] <MikeSmith>
so I see that Adam implemented window.crypto.getRandomFloat32Array(length) and window.crypto.getUint8Array(length) in WebKit but I can't find an actual spec for those anywhere

[05:30:00.0000] <zcorpan>
wonder if an open eventsource should set salvegeable to false. i guess it should

[05:31:00.0000] <annevk>
it didn't need make disappear at least

[05:31:01.0000] <annevk>
and that file in WebKit you said was about salvageable didn't contain references to XHR

[05:32:00.0000] <zcorpan>
oh, so browsers bfcache the page even if there's an ongoing XHR?

[05:33:00.0000] <annevk>
no idea :(

[05:33:01.0000] <annevk>
I don't really feel like diving into that either

[05:34:00.0000] <zcorpan>
shouldn't be so hard to test

[05:34:01.0000] <zcorpan>
/me checks

[05:35:00.0000] <MikeSmith>
ah, found http://wiki.whatwg.org/wiki/Crypto

[05:35:01.0000] <MikeSmith>
hmm, that's not it either

[05:39:00.0000] <Ms2ger>
MikeSmith, that's it, but it's out of date

[05:39:01.0000] <MikeSmith>
ok

[05:40:00.0000] <MikeSmith>
Ms2ger: so you are all implementing window.crypto.getRandomFloat32Array(length) and window.crypto.getUint8Array(length) ?

[05:40:01.0000] <Ms2ger>
I don't know who's implementing what

[05:41:00.0000] <Ms2ger>
Hmm

[05:41:01.0000] <Ms2ger>
I see a patch for getRandomValues in b.m.o

[05:42:00.0000] <Ms2ger>
Do you have a link to the webkit implementation?

[05:42:01.0000] <MikeSmith>
I find https://bugzilla.mozilla.org/show_bug.cgi?id=440046 and https://bugzilla.mozilla.org/show_bug.cgi?id=673432

[05:42:02.0000] <MikeSmith>
yeah, lemme get a link to the bug

[05:43:00.0000] <MikeSmith>
in the mean time, http://trac.webkit.org/browser/trunk/Source/WebCore/page/Crypto.cpp

[05:43:01.0000] <Ms2ger>
That's fine

[05:43:02.0000] <MikeSmith>
bug is https://bugs.webkit.org/show_bug.cgi?id=22049

[05:43:03.0000] <Ms2ger>
Er

[05:44:00.0000] <Ms2ger>
"Crypto::getRandomValues"

[05:45:00.0000] <Ms2ger>
Where did you get getRandomFloat32Array from?

[05:50:00.0000] <MikeSmith>
hmm, I guess that was an earlier patch

[05:50:01.0000] <MikeSmith>
https://bug-22049-attachments.webkit.org/attachment.cgi?id=81287

[05:51:00.0000] <MikeSmith>
but it's now just Crypto::getRandomValues(ArrayBufferView* array, ExceptionCode& ec)

[05:51:01.0000] <MikeSmith>
http://trac.webkit.org/browser/trunk/Source/WebCore/page/Crypto.cpp

[06:16:00.0000] <zcorpan>
annevk: from my testing it seems opera and chrome bfcache even if there's an open XHR

[06:16:01.0000] <zcorpan>
annevk: firefox seems to think my navigation is a redirect and removes the previous history entry

[06:17:00.0000] <zcorpan>
http://simon.html5.org/dump/bfcache-test/

[06:19:00.0000] <zcorpan>
so... yeah, when navigating back you have a dead XHR object

[06:26:00.0000] <zcorpan>
oh, actually

[06:26:01.0000] <zcorpan>
not dead

[06:26:02.0000] <Ms2ger>
Alive?

[06:27:00.0000] <zcorpan>
yeah

[06:27:01.0000] <zcorpan>
but you will miss the readystatechange events if it changes while you're away

[06:27:02.0000] <zcorpan>
at least in opera

[06:28:00.0000] <zcorpan>
/me tests other browsers

[06:29:00.0000] <annevk>
zombie XHR

[06:31:00.0000] <zcorpan>
but it seems chrome kills the XHR (i see it fills the log when navigating away) and does not bfcache anymore

[06:31:01.0000] <zcorpan>
maybe because i have an onreadystatechange listener now

[06:31:02.0000] <smaug____>
zcorpan: does chrome have bfcache?

[06:31:03.0000] <smaug____>
that is news to me

[06:32:00.0000] <zcorpan>
same with firefox

[06:32:01.0000] <zcorpan>
smaug____: it doesn't? http://simon.html5.org/dump/bfcache-test/baseline.html suggests it does

[06:34:00.0000] <smaug____>
I guess it just works quite differently

[06:34:01.0000] <annevk>
that also suggests Opera doesn't

[06:34:02.0000] <smaug____>
in some cases

[06:34:03.0000] <annevk>
at least in my build

[06:35:00.0000] <smaug____>
is there a reason to use sessionStorage in the test?

[06:35:01.0000] <zcorpan>
annevk: my opera says it was cached

[06:35:02.0000] <zcorpan>
annevk: 11.50 mac

[06:36:00.0000] <zcorpan>
smaug____: need a way to check if the script runs for the first time or second time

[06:37:00.0000] <annevk>
same, build 1074 here

[06:37:01.0000] <smaug____>
zcorpan: just set some variable

[06:37:02.0000] <zcorpan>
smaug____: i don't see how that would work

[06:39:00.0000] <smaug____>
if ("somevar" in window) alert("bfcached")

[06:40:00.0000] <zcorpan>
smaug____: if it is cached, the script doesn't run again when navigating back

[06:40:01.0000] <smaug____>
you could add some button which does " if ("somevar" in window) alert("bfcached")"

[06:40:02.0000] <smaug____>
in onclick

[06:40:03.0000] <smaug____>
or use pageshow

[06:40:04.0000] <smaug____>
(I don't know which browsers support pageshow event)

[06:40:05.0000] <zcorpan>
opera doesn't

[06:41:00.0000] <zcorpan>
does sessionStorage invalidate the test?

[06:41:01.0000] <smaug____>
don't know

[06:41:02.0000] <smaug____>
I was just curious why you use that

[06:42:00.0000] <zcorpan>
i don't see how the var would work -- if the page was not bfcached, then the script would run again, and the script would again set the variable

[06:42:01.0000] <zcorpan>
so the button would always say the variable is set

[06:42:02.0000] <jgraham>
Two buttons?

[06:43:00.0000] <jgraham>
One to set the variable, one to check it

[06:43:01.0000] <zcorpan>
ah, yeah

[06:44:00.0000] <zcorpan>
i just try to automate things without thinking about it :)

[06:52:00.0000] <smaug____>
zcorpan: also, IIRC, if load event listener causes a new page load, session history is handled specially

[06:53:00.0000] <smaug____>
so better to load a new page using a timer or something

[06:53:01.0000] <zcorpan>
smaug____: interesting

[06:54:00.0000] <zcorpan>
i noticed a difference in chrome if script navigates there and back

[06:54:01.0000] <zcorpan>
then it would cache

[06:54:02.0000] <zcorpan>
but if i navigate by clicking links, it doesn't cache

[06:54:03.0000] <zcorpan>
/me checks with a timeout

[06:55:00.0000] <annevk>
I suspect navigation is full of bugs

[06:57:00.0000] <smaug____>
zcorpan: session history implementations are full of bugs

[06:57:01.0000] <smaug____>
and the whole thing is underspeficied

[06:57:02.0000] <smaug____>
zcorpan: AFAIK, IE's session history is least buggy

[06:57:03.0000] <zcorpan>
chrome doesn't want to navigate from a timeout

[06:57:04.0000] <zcorpan>
smaug____: i'm here to get it better specified :)

[06:57:05.0000] <smaug____>
zcorpan: even if you do location = "url" ?

[06:58:00.0000] <smaug____>
zcorpan: awesome!

[06:58:01.0000] <smaug____>
(link.click() is just strange way to navigate to another page)

[06:59:00.0000] <zcorpan>
location = "url" worked

[06:59:01.0000] <zcorpan>
i used location first but firefox would just think that it was a redirect and remove the previous history entry

[06:59:02.0000] <smaug____>
you did that in load event listener

[06:59:03.0000] <zcorpan>
yeah

[07:00:00.0000] <zcorpan>
i guess it's better UX for pages that do scripted redirects

[07:01:00.0000] <zcorpan>
maybe we should spec this also

[07:04:00.0000] <annevk>
having to wait for load is not better UX though

[07:05:00.0000] <annevk>
I have pages that redirect using location because of fragment identifiers (that do not go to the server), but they execute directly

[07:07:00.0000] <smaug____>
/me tries to find when the "handle session history differently during load" was added

[07:07:01.0000] <smaug____>
my guess is about 10 years ago

[07:07:02.0000] <annevk>
you actually need something like location.redirect() if you want to handle this properly

[07:07:03.0000] <annevk>
and then maybe grandfather Firefox' use of onload

[07:08:00.0000] <zcorpan>
there's already location.replace()

[07:08:01.0000] <annevk>
or maybe if location was set before any load event handlers return

[07:09:00.0000] <zcorpan>
anyway, i'm not getting any further with XHR and bfcache. seems it's sometimes cached and sometimes not

[07:09:01.0000] <annevk>
(and not involving user action...)

[07:09:02.0000] <zcorpan>
what behavior do we *want* for XHR?

[07:10:00.0000] <smaug____>
2003

[07:11:00.0000] <smaug____>
hmm, nothing too interesting in https://bugzilla.mozilla.org/show_bug.cgi?id=166736

[07:12:00.0000] <smaug____>
zcorpan: have you already started to test the fun parts of session history handling: how to handle "dynamically" added iframes

[07:13:00.0000] <zcorpan>
smaug____: nope. i only intended to test XHR today

[07:14:00.0000] <zcorpan>
though i know iframes and navigation is not without problems

[07:15:00.0000] <smaug____>
yeah

[07:15:01.0000] <smaug____>
it is easy to get active back button which doesn't do anything

[07:15:02.0000] <smaug____>
or needs too many click

[07:15:03.0000] <smaug____>
s

[07:16:00.0000] <smaug____>
or in some cases it is even possible to get wrong document to be load to some iframe

[07:17:00.0000] <smaug____>
that used to be a bad problem in Gecko, but hopefully not anymore. IE and Webkit do, IIRC still suffer from that in some cases. Opera, again IIRC, doesn't have that problem

[07:18:00.0000] <zcorpan>
i recall we've had a number of bugs where teh back button just navigates an ad iframe instead of the whole page. and likely we've had the opposite problem as well.

[07:19:00.0000] <zcorpan>
dunno if it's good now

[07:19:01.0000] <zcorpan>
but i doubt it

[08:40:00.0000] <jgraham>
http://coding.smashingmagazine.com/2011/08/16/html5-and-the-document-outlining-algorithm/ is pretty nice

[09:30:00.0000] <jgraham>
Yay for things appearing in meeting notes that didn't actually get discussed at the meeting

[09:31:00.0000] <Ms2ger>
Where?

[09:34:00.0000] <jgraham>
public-html-testsuite

[09:34:01.0000] <jgraham>
"Microsoft submitted HTML5 Forms Tests (http://w3c-test.org/html/tests/submission/Microsoft/forms/)"

[09:35:00.0000] <jgraham>
I don't remember that coming up and it isn't in the logs…

[09:35:01.0000] <Ms2ger>
Me neither

[09:35:02.0000] <jgraham>
In other news one of me|Microsoft has forgotten how missing value defaults are supposed to work

[09:35:03.0000] <Ms2ger>
Or both :)

[09:36:00.0000] <jgraham>
OK, at least one :)

[09:37:00.0000] <jgraham>
The spec is not entirely clear so I am starting to think that I am wrong

[09:38:00.0000] <jgraham>
If an enumerated attribute foo is in the missing value default state and you do x.foo will you get the default value back, or the empty string?

[09:39:00.0000] <AryehGregor>
http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#reflecting-content-attributes-in-idl-attributes

[09:39:01.0000] <AryehGregor>
If a reflecting IDL attribute is a DOMString attribute whose content attribute is an enumerated attribute, and the IDL attribute is limited to only known values, then, on getting, the IDL attribute must return the conforming value associated with the state the attribute is in (in its canonical case), or the empty string if the attribute is in a state that has no associated keyword value

[09:39:02.0000] <AryehGregor>
So it depends on whether there's a state associated with the missing value default state.

[09:39:03.0000] <AryehGregor>
The enum stuff is confusing.

[09:41:00.0000] <jgraham>
OK, so in this case it looks like it should return the missing value default

[09:41:01.0000] <jgraham>
So I am wrong

[09:41:02.0000] <jgraham>
But in some other cases I might be right :)

[09:41:03.0000] <jgraham>
Gotta love the web

[09:42:00.0000] <Hixie>
Ms2ger: i had to add it, i dropped the earlier one!

[09:43:00.0000] <Ms2ger>
You did? :(

[09:43:01.0000] <Hixie>
it was in the rdf stuff iirc

[09:43:02.0000] <dglazkov>
good morning. Whatwg!

[09:43:03.0000] <Philip`>
Also gotta love algorithms (like reflection) that are heavily dependent on complex details of the context in which that algorithm is called (like whether the attribute is an enum with known values etc etc), instead of having specialised versions and linking to the right one

[09:44:00.0000] <Hixie>
yeah reflection is a bit of a mess

[09:44:01.0000] <AryehGregor>
It would be a lot nicer if there were separate xrefs for the different types of reflection, yeah.

[09:44:02.0000] <AryehGregor>
Some of them were actually quite confusing when I first wrote my tests.

[09:45:00.0000] <AryehGregor>
IIRC, it was unclear in some cases whether some attributes were defined to take a URL or not.

[09:45:01.0000] <jgraham>
I wonder if I can write an irssi plugin that filters lines that are close matches for "good morning. Whatwg!"

[09:45:02.0000] <annevk>
morning dglazkov, and hi grumpy jgraham :p

[09:46:00.0000] <Philip`>
jgraham: "/ignore dglazkov"? :-)

[09:46:01.0000] <Philip`>
Might be kind of rude though :-(

[09:46:02.0000] <Hixie>
if anyone wants to send me a patch that changes reflection to N algorithms and updates all the xrefs to point to the right ones, i'd apply it

[09:46:03.0000] <Hixie>
not gonna spend the time to do it myself though

[09:46:04.0000] <Hixie>
too many bigger fish to fry

[09:47:00.0000] <jgraham>
Philip`: Well yeah but at other times he is not being gratingly cheery :)

[09:47:01.0000] <Hixie>
bbiab

[09:47:02.0000] <jgraham>
I can only imagine it will be even worse in a few months when it has already been dark for 3 hours here

[09:47:03.0000] <Philip`>
Maybe you just need a filter that does s/morning/evening/ ?

[09:48:00.0000] <Philip`>
Then you can be contentedly miserable while everyone else is jollied up by dglazkov's enthusiasm

[09:49:00.0000] <dglazkov>
/me helpfully suggests everyone moving to the Bay Area.

[09:49:01.0000] <dglazkov>
it's a freakishly gorgeous morning here. You just _can't_ not be cheerful.

[09:51:00.0000] <hober>
fsvo "bay area"

[09:51:01.0000] <hober>
it was pretty gross in sf this morning

[09:51:02.0000] <dglazkov>
ok, everyone move to Mountain View :)

[09:52:00.0000] <annevk>
all you have is Castro man

[09:52:01.0000] <Philip`>
Google should just hire the whole of #whatwg

[09:52:02.0000] <annevk>
though that street is kind of amusing, that's it

[09:53:00.0000] <dglazkov>
you should view the bay area as <input type="range">

[09:53:01.0000] <hober>
annevk: indeed. nothing like the hustle and bustle of downtown cupertino

[09:53:02.0000] <hober>
oh, wait

[09:53:03.0000] <hober>
:)

[09:53:04.0000] <dglazkov>
at the top, in SF, it's awesome nightlife and gross mornings

[09:53:05.0000] <dglazkov>
at the bottom, in Saratoga, it's always sunny and really, really quiet

[09:53:06.0000] <annevk>
best of Cupertino is the city council when together with Steve Jobs

[09:53:07.0000] <hober>
annevk: indeed :)

[09:56:00.0000] <othermaciej>
SF has great mornings, if you like fog

[09:58:00.0000] <dglazkov>
fog no

[10:03:00.0000] <annevk>
http://lists.w3.org/Archives/Public/www-style/2011Aug/0503.html interesting perspective at the end on reading the HTML spec

[10:03:01.0000] <annevk>
if a bit ranty

[10:04:00.0000] <zewt>
"read it correctly"? what does that mean? heh

[10:04:01.0000] <zewt>
it's not like one normally reads a spec like that front-to-back

[10:06:00.0000] <Hixie>
have you seen the "how to read" section? :-)

[10:06:01.0000] <Philip`>
I think I once read about the first 2.5 sections from front to back, before getting bored and giving up

[10:07:00.0000] <Hixie>
annevk: it's a valid concern

[10:07:01.0000] <zewt>
i don't read things that tell me how to read, i learned that a long time ago :)

[10:07:02.0000] <Hixie>
annevk: i've been wondering what to do about it for a while

[10:09:00.0000] <Ms2ger>
I managed to get to the alt attribute section, at least

[10:09:01.0000] <Ms2ger>
I distinctly remember that my feedback was "tl;dr"

[10:11:00.0000] <Hixie>
hey it's better than html4's equivalent section

[10:11:01.0000] <annevk>
Hixie, the stability indicators should prolly be more prominent

[10:11:02.0000] <Hixie>
that one was "ts;ntr" ("too short, nothing to read")

[10:11:03.0000] <Hixie>
annevk: yeah

[10:11:04.0000] <Hixie>
annevk: i made them more obviously editable recently

[10:11:05.0000] <jgraham>
It's a valid concern, but the situation with CSS is at least as bad

[10:11:06.0000] <AryehGregor>
Maybe the stability indicators should affect the color scheme of the whole section?

[10:11:07.0000] <Philip`>
If it's too long for anyone to read then it's no more useful than writing nothing

[10:12:00.0000] <AryehGregor>
The way ED vs. WD does at the W3C?  Well, that doesn't affect the color scheme of the whole section, but it's extremely visible.

[10:12:01.0000] <jgraham>
I mean at least HTML doesn't have cases where the leven N spec is right but the level N+1 spec is buggy

[10:12:02.0000] <AryehGregor>
Or N+0.9, as the case may be?

[10:12:03.0000] <Hixie>
Philip`: i'm sort of assuming some pedantic authors will read it all and then in discussions they'll point authors to the right subsections

[10:12:04.0000] <jgraham>
I don't think stability indicators are the right solution per-se

[10:12:05.0000] <Philip`>
The stability indicators seem to give way too little information to be any value to authors

[10:13:00.0000] <MikeSmith>
http://www.w3.org/community/editing/

[10:13:01.0000] <MikeSmith>
oops

[10:13:02.0000] <Ms2ger>
Hi MikeSmith

[10:13:03.0000] <jgraham>
I think per-section implementation reports backed by testsuite results are the ideal solution

[10:13:04.0000] <Hixie>
AryehGregor: we could do something where a section that is marked as implemented by 3 or more UAs gets a thick border or something

[10:13:05.0000] <annevk>
oh nice, an edit link

[10:13:06.0000] <Ms2ger>
jgraham, testsuite? Where? ;)

[10:13:07.0000] <AryehGregor>
Philip`, they're more useful for implementers, I think.

[10:13:08.0000] <jgraham>
Ms2ger: Well yeah

[10:13:09.0000] <jgraham>
Think of the testsuite as icing

[10:14:00.0000] <MikeSmith>
Ms2ger: こんばんは

[10:14:01.0000] <jgraham>
But we should just integrate caniuse.com into the spec

[10:14:02.0000] <Hixie>
that would be interesting

[10:14:03.0000] <Hixie>
not sure how it should work

[10:14:04.0000] <annevk>
Hixie, for sections that do not yet have a status box you should have a create link

[10:14:05.0000] <Hixie>
annevk: are there any? i try to create them for every section

[10:14:06.0000] <Philip`>
It's not useful knowing that features are implemented in trunk versions - I expect authors want to know what caniuse shows (previous/current/next versions of browsers etc) though probably in much more detail, probably with some prose explaining which parts of features are broken or buggy

[10:14:07.0000] <jgraham>
Although it would provide a big incentive for people to release half-baked implementations of things to look good

[10:15:00.0000] <annevk>
Hixie, text/html-sandboxed does not have one

[10:15:01.0000] <Hixie>
if anyone wants to experiment with some scripts btw i'm happy to add them to the spec

[10:15:02.0000] <annevk>
Hixie, no idea how to fix

[10:15:03.0000] <jgraham>
So yeah, maybe the icing is important

[10:15:04.0000] <Hixie>
annevk: alt+double-click

[10:15:05.0000] <Hixie>
annevk: on the section

[10:15:06.0000] <MikeSmith>
heh -- Syd Lawrence has a great bio on his twitter page

[10:15:07.0000] <MikeSmith>
"I am a social web guru using HTML6, CSS4, AJAZ, using twitbook and facedIn. Expert with paint, frontpage and VBScript"

[10:16:00.0000] <MikeSmith>
http://twitter.com/#!/sydlawrence

[10:16:01.0000] <annevk>
Hixie, thx

[10:17:00.0000] <annevk>
Hixie, still not sure whether a little box on the side is helping

[10:17:01.0000] <Hixie>
yeah

[10:17:02.0000] <Hixie>
we could do some sort of background colour thing

[10:17:03.0000] <Hixie>
really no idea what to do

[10:17:04.0000] <jgraham>
But it still isn't the actual information that people want to know

[10:18:00.0000] <Ms2ger>
Which people?

[10:18:01.0000] <Hixie>
authors

[10:18:02.0000] <jgraham>
If, by some miracle, there is a section that is known to be stable and has 0 implementations, no one cares

[10:18:03.0000] <Hixie>
i agree that caniuse information is more helpful

[10:18:04.0000] <Ms2ger>
jgraham, you mean style scoped?

[10:18:05.0000] <jgraham>
Well I'm not sure if that is stable

[10:19:00.0000] <jgraham>
Or, I guess the HTML part is

[10:19:01.0000] <Hixie>
there's pending feedback on style scoped

[10:19:02.0000] <Hixie>
to do with selectors

[10:19:03.0000] <annevk>
authors want implemented and widely deployed

[10:19:04.0000] <Hixie>
don't we all

[10:19:05.0000] <annevk>
:p

[10:19:06.0000] <Ms2ger>
No

[10:19:07.0000] <Hixie>
there's a section status for "implemented and widely deployed"

[10:19:08.0000] <Hixie>
i don't think any part of the spec is marked that way yet

[10:20:00.0000] <jgraham>
Right, which is why we should actually give that information rather than providing some abstract proxy

[10:20:01.0000] <jgraham>
Which doesn't really macth

[10:20:02.0000] <Hixie>
oh, some bits are marked that way actually

[10:20:03.0000] <annevk>
yeah

[10:20:04.0000] <Hixie>
they all seem to also have all browsers marked as buggy!

[10:20:05.0000] <annevk>
acknowledgments is :)

[10:21:00.0000] <Hixie>
actually there's a bunch of sections marked this way

[10:21:01.0000] <Ms2ger>
To send or not to send D3E feedback...

[10:22:00.0000] <Philip`>
Boolean bugginess doesn't seem useful, since all browsers will always be buggy, at least in trivial details that nobody cares about - what matters is whether the bugs are easy to avoid or to work around (and also the bugs need to be documented so people know what to avoid)

[10:22:01.0000] <zewt>
documenting all of the bugs--now that sounds like a fun project

[10:23:00.0000] <Philip`>
All the bugs that authors will regularly hit, at least

[10:23:01.0000] <Ms2ger>
Might as well fix them

[10:23:02.0000] <jgraham>
Well they are already documented

[10:23:03.0000] <annevk>
Hixie, "no support whatsoever" should probably be renamed to "unknown"

[10:23:04.0000] <jgraham>
Just typically not cross-referenced with the spec

[10:23:05.0000] <AryehGregor>
The info that authors want is what other authors say browser support is like in practice.

[10:24:00.0000] <AryehGregor>
If there's a good test suite and a browser passes it, that should be sufficient to say it has good support, but it's far from necessary.

[10:24:01.0000] <jgraham>
AryehGregor: So we should allow authors to vote on whether a browser supports a particular thing in a good enough way?

[10:25:00.0000] <AryehGregor>
jgraham, no, just get one or a few good authors to put in the work.  Like that QuirksMode guy does for some things.

[10:25:01.0000] <AryehGregor>
Or, realistically, just give up and let people figure out by word of mouth.

[10:25:02.0000] <Ms2ger>
jgraham, no, we should have them write tests

[10:25:03.0000] <AryehGregor>
Also, yeah, let's try actually writing tests.  :)

[10:26:00.0000] <jgraham>
Ms2ger: What, as a fee for being allowed to look at the spec?

[10:26:01.0000] <Ms2ger>
wfm

[10:26:02.0000] <jgraham>
Put ads on it and encourage people to become "pro" spec readers by donating tests?

[10:26:03.0000] <Ms2ger>
Ship it!

[10:27:00.0000] <Philip`>
The HTML5 t-shirt money could be shared among those people

[10:27:01.0000] <jgraham>
I look forward to my 25p

[10:27:02.0000] <Ms2ger>
Do you get paid for all of Opera's tests? :)

[10:28:00.0000] <jgraham>
Point. I should revise that estimate down a bit

[10:32:00.0000] <annevk>
so is http://tantek.com/2011/220/b1/web-actions-a-new-building-block the next rel=tag or something we need to standardize?

[10:32:01.0000] <annevk>
I saw some <intent> element come by at some point

[10:33:00.0000] <Ms2ger>
/me wants <semantics>

[10:33:01.0000] <zewt>
wow that page is subtly really hard to read

[10:33:02.0000] <zewt>
and i'm not entirely sure why

[10:33:03.0000] <zewt>
even ignoring the silly inline diffs

[10:34:00.0000] <annevk>
still lobbying for background:unicorn

[10:34:01.0000] <annevk>
or background:unicorn

[10:34:02.0000] <annevk>
euh

[10:34:03.0000] <annevk>
double-rainbow

[11:05:00.0000] <annevk>
http://daringfireball.net/misc/2011/08/white-android-prototype.jpg hahaha

[11:10:00.0000] <Ms2ger>
Anybody who wants to place a bet on when I'll get a reply to my D3E comments?

[11:10:01.0000] <zewt>
after they've made an off-list decision

[11:11:00.0000] <Ms2ger>
Well, that's a given

[11:13:00.0000] <AryehGregor>
How easy would it be to write a tool that converts more or less arbitrary HTML5 into polyglot HTML5/XHTML5?

[11:14:00.0000] <Philip`>
Depending on how strictly you interpret "polyglot", probably impossible

[11:14:01.0000] <Philip`>
since you can't use inline scripts

[11:14:02.0000] <Philip`>
and can't use <pre> if the content starts with a blank line

[11:14:03.0000] <Philip`>
etc

[11:37:00.0000] <AryehGregor>
Philip`, polyglot to the extent of "it more or less looks the same when served as HTML vs. XHTML".

[11:38:00.0000] <AryehGregor>
Basically, I'm talking with Ian Jacobs, and he said that the W3C would be okay with allowing specs to be polyglot HTML5, but aren't so sure about regular HTML5.

[11:39:00.0000] <AryehGregor>
So if someone could just throw together a tool that would take the a text/html spec and output a polyglot one that behaves more or less the same, that would be helpful for progress.

[11:39:01.0000] <annevk>
the fuck?

[11:39:02.0000] <AryehGregor>
s/the//

[11:39:03.0000] <AryehGregor>
?

[11:39:04.0000] <AryehGregor>
Well, what he said maybe wasn't official.

[11:39:05.0000] <AryehGregor>
So perhaps I shouldn't be repeating it in public.

[11:39:06.0000] <annevk>
polyglot is such nonsense

[11:39:07.0000] <AryehGregor>
He made it seem like that would be helpful, though.

[11:40:00.0000] <annevk>
and has been a failure for many years even when experienced web authors tried it

[11:40:01.0000] <AryehGregor>
It's reasonably harmless nonsense, though, and sure beats XHTML 1.0 or whatever the W3C's HTML5 is in.

[11:40:02.0000] <Ms2ger>
HTML4

[11:40:03.0000] <Philip`>
Just set the HTML serialiser to not omit optional tags and to quote attributes, and stick an xmlns on it, and surely that'll be good enough

[11:41:00.0000] <Philip`>
It presumably doesn't need to actually work when parsed as XML (in terms of executing scripts correctly etc), it just needs to look well-formed enough

[11:41:01.0000] <AryehGregor>
Sounds plausible.

[11:41:02.0000] <annevk>
which is why it's wrong

[11:42:00.0000] <annevk>
either go your XML way and require XHTML5 or some such or just don't sit in the way

[11:42:01.0000] <AryehGregor>
However, doing things that are wrong but harmless is better than doing things that are wrong and also harmful, like having it be HTML 4.

[11:42:02.0000] <AryehGregor>
So in the interests of progress, it sounds better than the status quo by a large margin.

[11:42:03.0000] <AryehGregor>
Philip`, you'd have to set the serializer to use only numeric entities, too.

[11:42:04.0000] <Ms2ger>
anolis --quote-attr-values --no-minimize-boolean-attributes --use-trailing-solidus --space-before-trailing-solidus will probably get you a long way

[11:43:00.0000] <annevk>
AryehGregor, HTML4 does not bother me that much at the moment

[11:43:01.0000] <AryehGregor>
It bothers me vis-a-vis my spec, should I get around to publishing it as an ED.

[11:44:00.0000] <AryehGregor>
Ian was explaining to me that the CG's patent policy should really be just as good as the regular one, so maybe that won't be necessary.

[11:49:00.0000] <annevk>
I wonder why the W3C Team gets to decide on this and not the membership

[11:49:01.0000] <annevk>
karlcow always tells me the membership is in charge

[11:49:02.0000] <AryehGregor>
I don't think it's useful to worry about process issues here.

[11:50:00.0000] <AryehGregor>
Make sure the practicalities are in place, including our ability to leave if they decide to change their mind and do something stupid.

[11:50:01.0000] <AryehGregor>
What it says on paper doesn't matter.

[12:07:00.0000] <gsnedders>
annevk: probably for the same reason editors make a number of decisions and not WGs

[12:08:00.0000] <gsnedders>
Most of the time nobody really cares.

[12:11:00.0000] <annevk>
good point, editors should have a say in the format

[12:11:01.0000] <annevk>
this should be discussed on spec-prod

[12:12:00.0000] <annevk>
and the Team should justify their opinions there

[12:12:01.0000] <annevk>
spec-prod is http://lists.w3.org/Archives/Public/spec-prod/ fwiw

[12:13:00.0000] <Ms2ger>
Where are the canonical parsing tests?

[12:14:00.0000] <annevk>
http://code.google.com/p/html5lib/ I thought

[12:21:00.0000] <annevk>
gsnedders, thanks, asked why it's not discussed on spec-prod

[12:23:00.0000] <gsnedders>
annevk: That's really not what I meant.

[12:24:00.0000] <gsnedders>
annevk: The reason why the Team makes decisions and not Members is the same reason as why editors make decisions and not WGs.

[12:28:00.0000] <manu-db>
Does anybody know the history behind why "application/json" was picked over "text/json" - besides Douglas Crockford's rather cryptic response - "JSON is not really JavaScript and not really text"? Trying to figure out if JSON-LD should have an IANA registration for text/json-ld or application/json-ld or application/json+ld (JSON-LD is 100% compatible w/ JSON)

[12:30:00.0000] <AryehGregor>
I'd guess people here mostly aren't interested or knowledgeable about details of MIME type registration.

[12:30:01.0000] <AryehGregor>
(not that there's anything wrong with asking, I just wouldn't expect such a good answer here)

[12:30:02.0000] <manu-db>
alright - just attempting to be thorough in how we pick the name.

[12:30:03.0000] <manu-db>
s/name/mime type/

[12:31:00.0000] <manu-db>
related browser question: Does any browser actually pay attention to the "+" part of a mime type? For example - "text/xml+xhtml" - do browsers go "oh, that's text/xml" and then go "oh, and it's also xhtml" - or do they just do a regular string comparison for "text/xml+xhtml"?

[12:32:00.0000] <annevk>
it's application/xhtml+xml

[12:32:01.0000] <manu-db>
sorry... that's what I meant.

[12:32:02.0000] <annevk>
and Opera pays special attention to +xml

[12:32:03.0000] <annevk>
not everyone does I believe

[12:32:04.0000] <Ms2ger>
There's an old Gecko bug to do that

[12:32:05.0000] <Ms2ger>
Don't think it's fixed yet

[12:32:06.0000] <annevk>
text/* in general is a flawed media type according to the IETF and therefore they encourage everyone to just use the application/* one

[12:32:07.0000] <manu-db>
just wondering what best practice might be from a browser perspective - application/json-ld or application/json+ld ?

[12:33:00.0000] <annevk>
which of course makes the distinction kind of pointless

[12:33:01.0000] <annevk>
we have been registering various new text/* types following what we think makes sense...

[12:33:02.0000] <annevk>
manu-db, doesn't matter

[12:33:03.0000] <manu-db>
/me nods - there doesn't seem to be a clear logical thread through picking MIMEType names...

[12:34:00.0000] <annevk>
though application/ld+json would make more sense

[12:34:01.0000] <annevk>
or is it not actually JSON?

[12:34:02.0000] <manu-db>
it is 100% compatible w/ json.

[12:34:03.0000] <manu-db>
it's just JSON in a particular structure.

[12:34:04.0000] <annevk>
then I guess it ought to be +json

[12:35:00.0000] <annevk>
the goal btw is for browsers to treat all XML media types the same way

[12:35:01.0000] <annevk>
as in you can label your XHTML file text/xml and it's fine

[12:35:02.0000] <annevk>
because in practice XML works based on namespaces, not media types

[12:36:00.0000] <manu-db>
one of the reasons I'm shying away from application/ld+json is because there is no such "application/ld" media type... and because I hate plus signs in MIME Types (although, that can't be a basis for any sort of technical decision).

[12:36:01.0000] <annevk>
manu-db, I guess you want application/ld+json so it's consistent with the original JSON (which should have been text/json but isn't) and the +xml precedent

[12:37:00.0000] <annevk>
there's no application/xhtml either so that first reason is groundless

[12:37:01.0000] <manu-db>
I want to be consistent - but it doesn't seem like any of this is actually consistent...

[12:37:02.0000] <manu-db>
hmm... true.

[12:37:03.0000] <kennyluck>
what about application/json+ld ?

[12:37:04.0000] <annevk>
kennyluck, the format is JSON, not LD

[12:37:05.0000] <manu-db>
kennyluck, that's under consideration as well... (bikeshedding)

[12:40:00.0000] <Philip`>
/me vaguely remembers reading that +s should be avoided in MIME types unless the suffix is defined by some specification

[12:41:00.0000] <Philip`>
(and probably nobody has defined a +json suffix yet)

[12:41:01.0000] <Philip`>
so using that character would be unnecessary hassle

[12:42:00.0000] <manu-db>
So, I just checked w/ the XHTML folks and they said:  the convention for media types based upon XML is applicaiton/foo+xml - it is a convention handed down by the XML working group back in the day...I always thought it was backwards.

[12:42:01.0000] <manu-db>
Seems like Tim Bray created the +xml meme...

[12:42:02.0000] <manu-db>
http://www.ietf.org/rfc/rfc3023.txt

[12:43:00.0000] <manu-db>
This document also standardizes a convention (using the suffix '+xml')

[12:43:01.0000] <annevk>
doesn't matter who did it and whether or not it's backwards

[12:43:02.0000] <annevk>
if you are going to do it the other way around for JSON, people will not be kind

[12:44:00.0000] <manu-db>
which people? (just curious to know who would defend the +xml pattern)

[12:45:00.0000] <annevk>
everyone who wants some consistency in media types, like me I guess

[12:45:01.0000] <annevk>
(though I'd really like them to die, it seems we have to keep a few of them around)

[12:49:00.0000] <Hixie>
the w3c is so confusing

[12:49:01.0000] <Hixie>
half the decisions are not up to them but up to the membership

[12:49:02.0000] <Hixie>
while the other half are up to them

[12:49:03.0000] <Hixie>
and yet they do none of the work

[12:56:00.0000] <manu-db>
hmm... ok - application/ld+json it is

[12:57:00.0000] <manu-db>
Philip` any idea where you read the "don't use +s in MIME Type registrations" bit?

[12:57:01.0000] <Ms2ger>
Trying to figure out which docs one should reference informatively sure is hard work

[12:57:02.0000] <Philip`>
manu-db: No - I might just be imagining it

[12:58:00.0000] <manu-db>
/me wanders off to look, just in case.

[13:00:00.0000] <manu-db>
hmm... +json pattern has been used before: vnd.collection+json, vnd.hal+json, and vnd.oftn.l10n+json

[13:02:00.0000] <jgraham>
Yes, the canonical parsing tests are on google code

[13:03:00.0000] <jgraham>
/me has forgotten who was asking already

[13:03:01.0000] <jgraham>
Ms2ger: ^ seems you were asking

[13:03:02.0000] <jgraham>
Yell if you need access

[13:04:00.0000] <Ms2ger>
Ta

[13:30:00.0000] <annevk>
Ms2ger, apis-in-html-documents needs to move to DOM I suppose

[13:30:01.0000] <Ms2ger>
Probably

[13:31:00.0000] <Ms2ger>
Or I might have left them there because they test SVG parsing as well

[13:32:00.0000] <annevk>
parsing > html5lib

[13:37:00.0000] <karlcow>
annevk: spec-prod and chairs mailing-list. There might be a question of the toolchain used already for dealing with these documents maybe.

[13:37:01.0000] <karlcow>
no harm pushing for it.

[13:38:00.0000] <karlcow>
You might have some pushback from other members who are used to other tools. Community process, dialogs, etc.

[13:39:00.0000] <AryehGregor>
Hixie, did you see above where I said that Ian Jacobs told me that he thought the W3C would be okay with allowing specs to be published as HTML5 as long as it's polyglot?

[13:40:00.0000] <Hixie>
yes

[13:40:01.0000] <karlcow>
AryehGregor: I would be careful on that statement. From Ian Jacobs to W3C generalization.

[13:40:02.0000] <shepazu>
I think that's the general W3C staff sentiment

[13:41:00.0000] <AryehGregor>
karlcow, I made it clear that it was what he told me, not an official statement.

[13:41:01.0000] <AryehGregor>
If that's the sentiment, it should be trivial to set up.

[13:41:02.0000] <shepazu>
personally, I don't see the need for them to even be polyglot

[13:41:03.0000] <AryehGregor>
Nor do we, obviously.  :)

[13:41:04.0000] <AryehGregor>
That's what HTML5 parsers are for.

[13:41:05.0000] <shepazu>
I think Mike feels similarly

[13:41:06.0000] <AryehGregor>
But it's a small concession.

[13:41:07.0000] <AryehGregor>
. . . also, why is non-polyglot HTML5 inferior to HTML4 in any way?  HTML4 is allowed.

[13:42:00.0000] <AryehGregor>
And that's what the spec is now.

[13:42:01.0000] <shepazu>
yup, agreed

[13:42:02.0000] <karlcow>
I can imagine that for the Webmasters it might need to refactor a big part of the XML Toolchain but I'm not even sure. I wonder if XSLT is still used for the pub rules checker

[13:42:03.0000] <shepazu>
dunno

[13:42:04.0000] <shepazu>
I could test it

[13:43:00.0000] <AryehGregor>
karlcow, how could it be, if HTML4 is allowed?

[13:43:01.0000] <shepazu>
yeah

[13:43:02.0000] <karlcow>
shepazu: creating a dummy HTML5 spec and testing it with the webmaster to see how/if it fails would be good.

[13:43:03.0000] <AryehGregor>
Is there some place where people other than W3C staff can discuss this with the powers that be, rather than having to be told people's reasoning by intermediaries?

[13:43:04.0000] <shepazu>
nothing to prevent you from putting in an HTML5 plugin into XSLT, anyway

[13:43:05.0000] <karlcow>
AryehGregor: I imagine tidy is involved

[13:44:00.0000] <Hixie>
the only difference between the spec being html5 and being html4 is the doctype, a couple of examples that are stripped out of the w3c copy, the <meta charset> being stripped, and some attributes being added.

[13:44:01.0000] <shepazu>
AryehGregor: I suggest starting the conversation on spec-prod and seeing if we can get others to chime in

[13:45:00.0000] <Hixie>
i literally have a four-line perl script to make the spec html4-compliant

[13:45:01.0000] <shepazu>
Hixie: I don't think that should be necessary, but we can consider that as a fallback

[13:46:00.0000] <Hixie>
it's only necessary because ij says i have to use html4 and not versionless-html

[13:46:01.0000] <shepazu>
/me doesn't know what that means, but if you mean "HTML5", then I think we can sell that

[13:46:02.0000] <shepazu>
:)

[13:47:00.0000] <Hixie>
because somehow even though html3.2, html4, html4.01, xhtml1, and xhtml1.1 were all allowed to be self-hosting, the new html spec, despite being more closely aligned to reality, is apparently a Compatibility Risk that the w3c isn't willing to take on

[13:47:01.0000] <zewt>
that's called a Vote of Confidence

[13:47:02.0000] <shepazu>
Hixie: I'm not sure we're in the same conversation here

[13:47:03.0000] <karlcow>
Maybe there is a fear about what HTML5 means in terms of markup. (No idea really, just trying to imagine). Pandora box.

[13:48:00.0000] <Philip`>
The spec splitter uses <div> instead of <nav> with the --w3c flag, for HTML4 compatibility

[13:48:01.0000] <Philip`>
or at least my version of the splitter, which isn't used for the W3C copy

[13:49:00.0000] <Philip`>
so I suppose there might be more differences in MikeSmith(?)'s version

[13:50:00.0000] <Ms2ger>
/me wants an id on the </sarcasm> dt

[13:51:00.0000] <gsnedders>
Ms2ger: Where is that in the spec? :P

[13:52:00.0000] <MikeSmith>
my version is optimized for the sole purpose of creating as little as possible work/bs for me

[13:52:01.0000] <Ms2ger>
whatwg.org/html/#sa... Ah, dammit

[13:53:00.0000] <Hixie>
the in-jokes intentionally don't have ids :-)

[13:53:01.0000] <Ms2ger>
:(

[13:54:00.0000] <Ms2ger>
Go fix some bugs ;)

[13:58:00.0000] <AryehGregor>
karlcow, shepazu: http://lists.w3.org/Archives/Public/spec-prod/2011JulSep/0010.html

[14:00:00.0000] <karlcow>
cool. I would not have turned the mail this way. But at least it starts the discussion :)

[14:01:00.0000] <karlcow>
shepazu: what about sending a pointer to this discussion to chairs⊙wo too

[14:01:01.0000] <karlcow>
I'm not on chairs mailing list

[14:18:00.0000] <jgraham>
FWIW I think requiring polyglotness is worse than not allowing HTML(5)

[14:19:00.0000] <jgraham>
One is an obviosuly stupid policy. The other is also a stupid policy but looks slightly like a non-stupid one

[14:19:01.0000] <jgraham>
So people might mistake it for a sensible one

[14:29:00.0000] <shepazu>
way to hardline it, jgraham…let's hope nobody compromises, so we don't have any progress :)

[14:29:01.0000] <shepazu>
/me nominates jgraham for US Congress :)

[14:31:00.0000] <karlcow>
heh

[14:45:00.0000] <jgraham>
shepazu: True polyglotness is a cure worse than the disease. Approximate polyglotness is sort-of acceptable but I don't really see how it helps anyone

[14:46:00.0000] <jgraham>
I mean if the pubrules want to say "thou must close all thy tags" that's silly, but whatever

[14:46:01.0000] <jgraham>
Anything that can't be trivially achieved by setting a few serializer options isn't sensible

[14:57:00.0000] <Hixie>
jgraham: i entirely agree

[14:58:00.0000] <Hixie>
i don't really see what there is to compromise here. What's the "other" position that we're compromising with to get to polyglot?

[14:58:01.0000] <TabAtkins_>
XML forever, everywhere?

[14:59:00.0000] <karlcow>
getting already all heat up :) funny guys. Looks like a boys club before a soccer game. cute

[14:59:01.0000] <Hixie>
we could move to xhtml5 sent as xml, sure

[15:00:00.0000] <AryehGregor>
karlcow, this does not appear to be a constructive approach to take when W3C editors complain that a requirement imposed upon them doesn't seem to make sense.

[15:00:01.0000] <Hixie>
but if that's what the w3c wants, why hasn't anyone even tried suggesting it?

[15:00:02.0000] <AryehGregor>
Rather than giving us substantive answers, you're brushing us off and suggesting we're being unreasonable (without explaining why).

[15:00:03.0000] <karlcow>
AryehGregor: who is the "you" :)

[15:01:00.0000] <karlcow>
who is the "us"

[15:01:01.0000] <AryehGregor>
you = the person currently using the nick karlcow in this channel

[15:01:02.0000] <AryehGregor>
us = me, Hixie, jgraham, inter alia

[15:01:03.0000] <karlcow>
what did I brush off?

[15:02:00.0000] <karlcow>
beautiful rain outside of my window, right now. And wonderful humid smell. The tree leaves are fragrant.

[15:03:00.0000] <Philip`>
/me doesn't think speaking in haiku will help much

[15:04:00.0000] <zcorpan>
annevk: i think we dropped +xml knowledge because it broke sites

[15:05:00.0000] <karlcow>
Philip`: that was not a haiku, that was the description of what I see from the window.

[15:06:00.0000] <karlcow>
I sincerely do not know what AryehGregor is talking about cf brushing us off.

[15:06:01.0000] <AryehGregor>
Oh, one of the lines was shepazu rather than you.

[15:06:02.0000] <AryehGregor>
Well, whatever.

[15:07:00.0000] <AryehGregor>
We were complaining that requiring polyglot made no sense, and shepazu implied that he considered that uncompromising, and you made an amused remark about how we were getting excited, but no one has yet addressed the substantive concern.

[15:07:01.0000] <AryehGregor>
(at least, not here, haven't checked the mailing list yet)

[15:09:00.0000] <karlcow>
yes the discussion is funny, because indeed it looks like yet another little big debate. It's why I find it amusing. Not brushing off, more amused by our childish tendencies to make small issues international issues ;)

[15:09:01.0000] <shepazu>
AryehGregor: my position is that we should simply be able to use HTML5, not only polyglot HTML5, but I'm not the decision maker

[15:10:00.0000] <AryehGregor>
shepazu, who is the decision maker?

[15:10:01.0000] <karlcow>
heh

[15:10:02.0000] <shepazu>
if some people say they are more comfortable with polyglot, and that's all we can get, well, that's suboptimal, but it's a beachhead

[15:10:03.0000] <Philip`>
Who decides who is the decision maker?

[15:10:04.0000] <Philip`>
and who decides who that is?

[15:11:00.0000] <karlcow>
ah finally Philip` is reasonable ;)

[15:11:01.0000] <Philip`>
Surely that gets closer to who is the real decision maker

[15:11:02.0000] <shepazu>
AryehGregor: I'm guessing it would be some combination of Ian Jacobs and TimBL

[15:11:03.0000] <Philip`>
It must be a finite sequence since there's less than seven billion people in the world

[15:12:00.0000] <shepazu>
Philip`: don't forget the alien lizard overlords

[15:13:00.0000] <shepazu>
and it might be time-limited, so by the time you get to the answer, it may have changed

[15:13:01.0000] <AryehGregor>
If that guess could be clarified a bit so we could speak to the actual people responsible instead of being told by people who agree with us that we should compromise with the people who we've never spoken to, whose reasons have not been explained, and who have not even been clearly identified, that would be a step forward.

[15:13:02.0000] <Hixie>
also helpful would be an idea of what we're compromising between

[15:14:00.0000] <shepazu>
AryehGregor: I think that if we are able to get TimBL (who I think is on vacation just now) and IanJ on the conversation, that would be critical mass

[15:14:01.0000] <AryehGregor>
I already spoke with IanJ, and it sounded like he didn't have any problem with publication as straight HTML5.

[15:14:02.0000] <AryehGregor>
But let's see, I CC'd him.

[15:15:00.0000] <shepazu>
Hixie: I think that would be part of the conversation… though maybe there aren't any deal breakers, just clarification needed

[15:15:01.0000] <shepazu>
I can't really say, since I don't see a problem with just using HTML5

[15:15:02.0000] <zcorpan>
which HTML5?

[15:16:00.0000] <Hixie>
doesn't matter, there's no way to distinguish them in practice for the specs we're talking about :-)

[15:17:00.0000] <Hixie>
shepazu: i tried to convince w3c to let us publish as contemporary HTML a few years ago, and got told no quite firmly. I figured it wasn't worth my time to argue and that the W3C would catch up with the world eventually. And indeed that seems to be happening, in part through Aryeh poking the hornet's nest again. :-)

[15:17:01.0000] <zcorpan>
depends on what happens with the HTML5s going forward

[15:17:02.0000] <Hixie>
other than the doctype, i can't really see anything that could change that would affect us

[15:18:00.0000] <shepazu>
Hixie: actually, I'd credit MikeSmith with starting this up again, internally

[15:18:01.0000] <Hixie>
and if the w3c decides to change the doctype, they're going to have bigger problems than the format the spec is in

[15:18:02.0000] <Hixie>
shepazu: ah, k

[15:18:03.0000] <shepazu>
not to dismiss what Anne, SamRuby, or AryehGregor have done

[15:19:00.0000] <zcorpan>
Hixie: well there's an example that uses <p>s in <caption>, i dunno, maybe the html wg decides that teh content model of <caption> shall be changed, or similar

[15:19:01.0000] <jgraham>
I'm pretty suprised that W3C aren't embarassed not to be publishing their own stuff in their own flagship standard language

[15:19:02.0000] <Hixie>
zcorpan: well if they change the content model, the example would have to change too

[15:19:03.0000] <shepazu>
Hixie: you might not have asked in a way that was palatable to whoever you talked to

[15:20:00.0000] <zcorpan>
Hixie: touche

[15:20:01.0000] <karlcow>
jgraham: why should they? :)

[15:20:02.0000] <Hixie>
shepazu: i believe the conversation was something along the lines of pubteam: "Hey, you can't use HTML5 to publish HTML5. You must use HTML4." me: "say what?"

[15:20:03.0000] <jgraham>
karlcow: It sugests a certain lack of belief in their own propoganda

[15:21:00.0000] <karlcow>
jgraham: nope. just a different set of priorities. It doesn't really matter. :) the document being readable.

[15:21:01.0000] <karlcow>
You know if it doesn't break, don't fix it. :)

[15:21:02.0000] <zcorpan>
Hixie: but now when html5 is at LAST CALL, the situation is clearly quite different

[15:22:00.0000] <jgraham>
karlcow: By analogy I would also be surprised to find that the IE team all use IE6

[15:22:01.0000] <shepazu>
Hixie: I wasn't part of the conversation, so I can't comment on the verity of your characterization… but in any case, I think it's a reasonable think to do

[15:22:02.0000] <Hixie>
zcorpan: as karlcow would put it, it's very cute

[15:22:03.0000] <karlcow>
jgraham: I'm not using Opera Next ;) just for testing purpose

[15:23:00.0000] <jgraham>
karlcow: Or, if you prefer, the document would be readable as flash or pdf. But I don't think we should publish using those

[15:23:01.0000] <shepazu>
/me authors almost exclusively in HTML5 for all his non-TR documents on W3C (group sites, etc)

[15:23:02.0000] <shepazu>
and have for over a year

[15:23:03.0000] <karlcow>
and I'm not necessary jumping on new version of OSes or software when they come out. I really don't care about the newness of things :) as long as I can publish and communicate words

[15:23:04.0000] <jgraham>
HTML 5 actually adds features that are useful for writing specs e.g. the new semantic elements

[15:24:00.0000] <Hixie>
yeah <aside> is the main thing i wish i could use

[15:24:01.0000] <Philip`>
jgraham: By "useful" you mean "equivalent to <div class=foo> in practice"?

[15:25:00.0000] <jgraham>
Philip`: Useul to the person writing the document. Especially for e.g. automatic ToC generation using anolis

[15:26:00.0000] <jgraham>
*Useful

[15:26:01.0000] <karlcow>
Hixie: what aside provides as a direct ROI? (being curious. apart of a cleaner semantics)

[15:26:02.0000] <Hixie>
karlcow: nothing

[15:26:03.0000] <Hixie>
jgraham: true, <section>/<h1> would be useful too. it could even be used incrementally, though i'm sure the inconsistency would drive me batty.

[15:26:04.0000] <karlcow>
heh

[15:29:00.0000] <karlcow>
/me time for being off the screen. paper and pen!

[15:29:01.0000] <jgraham>
More haiku?

[15:31:00.0000] <shepazu>
would using something like HTMLShiv.js to ensure it works in older browsers be acceptable?

[15:31:01.0000] <zcorpan>
the spec already doesn't work in older browsers

[15:36:00.0000] <shepazu>
heh

[15:37:00.0000] <shepazu>
maybe the HTML5 spec doesn't, but other specs that use HTML5 might

[15:41:00.0000] <AryehGregor>
You can write HTML4 that doesn't work in old browsers too, if you like.  That seems orthogonal.

[15:42:00.0000] <AryehGregor>
For instance, HTML4 that relies on, I dunno, IndexedDB.

[15:42:01.0000] <AryehGregor>
If we want compat with old browsers, that should be a separate requirement.

[15:43:00.0000] <zcorpan>
i thought pubrules didn't allow scripts at all

[15:44:00.0000] <zcorpan>
not that i can find anything about "script" in pubrules

[15:51:00.0000] <AryehGregor>
Surely some specs use scripts?

[15:51:01.0000] <AryehGregor>
W3C HTML5 doesn't use them?

[15:52:00.0000] <zcorpan>
hmm yeah i think the comment form uses script

[15:54:00.0000] <shepazu>
scripts are allowed… brb

[16:00:00.0000] <TabAtkins_>
/me wouldn't want to use <section>/<h1> for CSS specs.  Being able to write straight-line code is useful.

[16:07:00.0000] <AryehGregor>
I always find this report depressing: http://www.redhat.com/f/pdf/security/RHEL4_RiskReport_6yr_wp_5732067_0311_ma_web.pdf

[16:07:01.0000] <AryehGregor>
Critical flaws in Mozilla products: 194.  Critical flaws in all other programs combined: 58.

[16:07:02.0000] <AryehGregor>
Granted, web browsers are extremely complicated, change rapidly, and are network-exposed, but still . . .

[16:10:00.0000] <AryehGregor>
Hmm, are there any useful points here?  http://www.infoworld.com/print/169665

[16:11:00.0000] <AryehGregor>
Point 11 is surprisingly clueful.

[16:12:00.0000] <Philip`>
Maybe nobody bothers testing other applications much?

[16:13:00.0000] <Philip`>
/me has always wondered how vulnerable multiplayer games are to malicious network packets, since game developers aren't particularly known for focusing on security and clean coding practices

[16:14:00.0000] <AryehGregor>
You don't have nearly as many targets for multiplayer games, though.

[16:14:01.0000] <jamesr>
AryehGregor: that's basically punishing mozilla for disclosing accurately

[16:15:00.0000] <AryehGregor>
jamesr, this is RHEL, everything here is open-source.  I don't think many of the major projects are hiding too many of their security fices.

[16:15:01.0000] <jamesr>
no?

[16:15:02.0000] <AryehGregor>
Well, not everything, but almost everything.

[16:15:03.0000] <AryehGregor>
Maybe to some extent, yeah.  Like the kernel doesn't give details on vulnerabilities, it's true.

[16:15:04.0000] <jamesr>
you don't think they fold multiple bugs into a single report, or downgrade some vulns?

[16:15:05.0000] <jamesr>
i agree that in OSS the problem is not nearly as bad as with closed source

[16:16:00.0000] <AryehGregor>
AFAIU, the severity and so on are determined by RHEL here, not the vendor.

[16:16:01.0000] <jamesr>
but mozilla has a pretty responsible security disclosure policy so they look bad here relative to projects that don't spend as much time+effort

[16:17:00.0000] <AryehGregor>
Also, come to think of it, its userbase is mostly Windows desktop, which gives it about two orders of magnitude usage over almost anything else in the default RHEL install.

[16:17:01.0000] <AryehGregor>
So you'd expect commensurably more testing.

[16:17:02.0000] <AryehGregor>
Okay, that's all fair.  But I wonder how much of these critical vulnerabilities could be effectively mitigated by Chrome-style sandboxing.

[16:18:00.0000] <jamesr>
yeah, a lot would become high severity

[16:18:01.0000] <jamesr>
or at least it would using our terminology, not sure what RHEL's vocabulary is

[16:18:02.0000] <AryehGregor>
I mean, it's fair to say that there's more effort on finding browser bugs, but hopefully there's also more effort in fixing them.  :)

[16:19:00.0000] <AryehGregor>
So one would hope it could cancel out.

[16:19:01.0000] <AryehGregor>
Oh well.

[16:19:02.0000] <yuhong>
http://www.google.com/url?sa=t&source=web&cd=1&ved=0CBwQFjAA&url=http%3A%2F%2Flcamtuf.blogspot.com%2F2010%2F05%2Fvulnerability-databases-and-pie-charts.html&ei=lpBNTo6fO8rTiAKJ7dGCAQ&usg=AFQjCNHwG2qJiz9ugZwdxihAWJu-UV-iUg&sig2=WmNI4ObFu-E7I-bp0Pei8A

[16:19:03.0000] <yuhong>
http://lcamtuf.blogspot.com/2010/05/vulnerability-databases-and-pie-charts.html

[16:19:04.0000] <jamesr>
definitely worthwhile to spend a lot of effort on a binary that people feed a lot of untrusted content from the web into on a daily basis

[16:19:05.0000] <jamesr>
and that tends to manage credentials for their email, banking, etc

[16:19:06.0000] <AryehGregor>
Yep.


2011-08-19
[19:46:00.0000] <roc>
hmm

[19:46:01.0000] <roc>
should we have implemented caretPositionFromPoint with a moz prefix?

[20:04:00.0000] <TabAtkins>
roc: Probably, yeah.

[20:42:00.0000] <Hixie>
roc: depends if the spec author is tracking implementations and guaranteeing to not break them, or if the spec author is doing his own thing and hoping for the best

[20:43:00.0000] <Hixie>
roc: who's the spec author?

[20:43:01.0000] <Hixie>
(which spec is it?)

[21:02:00.0000] <roc>
CSSOM Views

[21:02:01.0000] <roc>
it only just landed, we can get a moz prefix on it

[21:03:00.0000] <roc>
although I notice that Chrome already shipped caretRangeFromPoint (the obsolete previous version) without a prefix :-(

[21:04:00.0000] <jamesr>
i argued with someone about prefixing selectionDirection in webkit

[21:04:01.0000] <jamesr>
i think i lost that argument

[21:04:02.0000] <jamesr>
know when we shipped it?

[21:05:00.0000] <roc>
no idea

[21:05:01.0000] <jamesr>
is it in safari?

[21:05:02.0000] <roc>
no idea

[21:05:03.0000] <jamesr>
that's how you can tell if it's old or not

[21:05:04.0000] <roc>
I suppose I should install Safari here

[21:09:00.0000] <Hixie>
roc: so long as it changes name it doesn't matter if it's missing teh prefix

[21:10:00.0000] <roc>
I think it matters a little bit if Web devs think it's a standard

[21:10:01.0000] <Hixie>
cssom is anne, so speak to him

[21:10:02.0000] <roc>
yeah, I CCed him on the bug

[21:10:03.0000] <Hixie>
fwiw, for specs i edit, i prefer browsers to not use prefixes, and then to just keep me closely in the loop of any implementation experiences

[21:10:04.0000] <roc>
hmm

[21:10:05.0000] <Hixie>
that allows me to freeze the spec much faster and gets widespread implementations quicker

[21:10:06.0000] <Hixie>
it also means more care is needed

[21:11:00.0000] <Hixie>
but that's a fair trade imho

[21:11:01.0000] <roc>
that's fine I guess as long as you don't freeze prematurely

[21:11:02.0000] <roc>
it's a delicate dance we have to play here

[21:12:00.0000] <roc>
Safari 5.1 has caretRangeFromPoint FWIW

[21:12:01.0000] <erlehmann>
oh you.

[21:12:02.0000] <Hixie>
roc: indeed

[21:13:00.0000] <Hixie>
roc: i really don't want to end up doing what css does and have features still vendor prefixed even 5 years after everyone is interoperable modulo prefixes, though

[21:14:00.0000] <roc>
that is partly the result of W3C processes I think

[21:14:01.0000] <roc>
and perhaps slow update cycles

[21:14:02.0000] <jamesr>
there unprefixing is tied to advancing to CR

[21:15:00.0000] <roc>
I'm a conservative sort of person so I tend to fear the damage of premature freezing more than the inconvenience of a year or two of prefixing

[21:15:01.0000] <roc>
on the grounds that the former is forever

[22:41:00.0000] <zewt>
at least prefixing javascript apis is generally less inconvenient than prefixing css, when the prefixed implementations are otherwise compatible

[01:15:00.0000] <zcorpan>
MikeSmith: <audio/> is now valid per spec

[01:16:00.0000] <MikeSmith>
zcorpan: OK

[01:45:00.0000] <zcorpan>
Ms2ger: http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dom-tree-accessors/document.getElementsByClassName-same.html can't fail (unless it throws) right?

[01:46:00.0000] <Ms2ger>
Right

[01:46:01.0000] <Ms2ger>
People didn't want an actual requirement about caching

[01:52:00.0000] <zcorpan>
Ms2ger: http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dom-tree-accessors/document.title-07.html doesn't check "expected title"

[01:53:00.0000] <Ms2ger>
It's tested in http://dvcs.w3.org/hg/webapps/file/bc93d42ccbe7/DOMCore/tests/submissions/Ms2ger/DOMImplementation-createHTMLDocument.html

[01:54:00.0000] <zcorpan>
ah

[01:55:00.0000] <Ms2ger>
(I moved the entire test to DOM Core, and then David Flanagan complained that I was testing HTML)

[01:56:00.0000] <zcorpan>
http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dynamic-markup-insertion/document.open-02.html - is w instanceof Window expected to work? shouldn't it be w instanceof w.Window?

[01:56:01.0000] <Ms2ger>
Oh, good point

[01:57:00.0000] <zcorpan>
http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dynamic-markup-insertion/document.write-02.html has an // XXX

[01:58:00.0000] <Ms2ger>
An incorrect XXX at that

[01:58:01.0000] <Ms2ger>
and -ln-02 as well, I guess

[01:59:00.0000] <zcorpan>
yah

[01:59:01.0000] <Ms2ger>
At least I'm consistent :)

[02:03:00.0000] <zcorpan>
http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dynamic-markup-insertion/innerhtml-03.xhtml shouldn't the span have xmlns?

[02:04:00.0000] <Ms2ger>
Looks like it

[02:05:00.0000] <zcorpan>
also, getting innerHTML in xml isn't defined to return only one possible string

[02:05:01.0000] <zcorpan>
it'd be conforming to use single quotes etc

[02:06:00.0000] <zcorpan>
i think that's a bug in the spec, but until the spec is fixed...

[02:08:00.0000] <zcorpan>
http://dvcs.w3.org/hg/html/file/f8b7e4633037/tests/submission/Ms2ger/dynamic-markup-insertion/innerhtml-01.xhtml could be testing all listed cases that should throw

[02:08:01.0000] <Ms2ger>
It could be, if I had time :)

[02:10:00.0000] <zcorpan>
ok i'll let somebody else review the rest

[02:10:01.0000] <zcorpan>
should i email the above to teh list?

[02:14:00.0000] <jgraham>
/me plans to start the telepathic enhancements for html community group, just so that I can tell people to email teh list

[02:15:00.0000] <jgraham>
/me is apparently incapable of keeping a whole sentence in the third person

[02:15:01.0000] <Ms2ger>
I pushed http://dvcs.w3.org/hg/html/rev/d46205148900

[02:46:00.0000] <zcorpan>
i like the "remove" algorithm in domcore

[02:52:00.0000] <zcorpan>
Ms2ger: "If node is a doctype and either parent has a doctype child that is not child, or an element is preceding child." seems like the paragraph isn't complete. http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-node-replace

[03:17:00.0000] <Ms2ger>
zcorpan, would you like it better with s/If node/If child/?

[03:21:00.0000] <MikeSmith>
foolip, zcorpan - <audio></audio> validator fix checked in

[03:21:01.0000] <MikeSmith>
thanks for the heads-up

[03:25:00.0000] <foolip>
MikeSmith, is it live somewhere?

[03:27:00.0000] <MikeSmith>
foolip: not yet, but I can make it live at w3.org/html/check if you want

[03:27:01.0000] <foolip>
that'd be nice

[03:27:02.0000] <MikeSmith>
OK

[03:27:03.0000] <MikeSmith>
gimme a couple minutes

[03:30:00.0000] <zcorpan>
Ms2ger: i'd like it better if it ended with "throw exception blah blah"

[03:30:01.0000] <Ms2ger>
Eh

[03:30:02.0000] <Ms2ger>
Right

[03:31:00.0000] <Ms2ger>
/me blames annevk

[03:31:01.0000] <zcorpan>
MikeSmith: nice

[03:32:00.0000] <Ms2ger>
Man, how many substeps does this algorithm have?

[03:36:00.0000] <Ms2ger>
zcorpan, fixed, thanks

[03:39:00.0000] <MikeSmith>
foolip: OK, pushed now to http://www.w3.org/html/check

[03:39:01.0000] <foolip>
MikeSmith, seems to work, thanks!

[03:40:00.0000] <MikeSmith>
np

[03:41:00.0000] <MikeSmith>
foolip: btw, when building I'm now getting some messages emitted about schema errors for some microdata stuff

[03:41:01.0000] <MikeSmith>
e.g., Reference to undefined pattern “common.attrs.microdata.itemid”

[03:41:02.0000] <MikeSmith>
oh, I think I know why that is

[03:42:00.0000] <foolip>
MikeSmith, hmm, I didn't add or remove anything with itemid, just tweaked its definition I think

[03:43:00.0000] <MikeSmith>
yeah, I'm getting the same message for all the aria attributes, so it's not related to specific change

[03:44:00.0000] <MikeSmith>
I think it's just because of have some stuff in my workspace that I need to update

[03:44:01.0000] <MikeSmith>
anyway, I'll figure it out

[04:04:00.0000] <zcorpan>
hmm. XMLHttpRequestUpload can also get events

[06:29:00.0000] <ooga-booga>
im all about the hypertext.....

[09:30:00.0000] <dglazkov>
good morning, Whatwg!

[09:33:00.0000] <Hixie>
man, you get up way too early

[09:39:00.0000] <jgraham>
Alternative theory: You are too cheery in the morning

[09:39:01.0000] <Hixie>
well that makes no sense

[09:39:02.0000] <Hixie>
how can you be too cheery

[09:39:03.0000] <jgraham>
Alternative alternative theory: You are actually a sociopath

[09:39:04.0000] <Hixie>
cheery is good!

[09:42:00.0000] <Ms2ger>
jgraham, no, that's me

[09:46:00.0000] <Philip`>
Surely that's jgraham?

[09:46:01.0000] <Ms2ger>
Well no, jgraham is just a doctor

[10:00:00.0000] <dglazkov>
/me missed an excellent opportunity to troll. :(

[10:01:00.0000] <Ms2ger>
dglazkov, very much unlike you

[10:01:01.0000] <dglazkov>
I know, right!

[10:02:00.0000] <dglazkov>
^^^ should probably be "I know, right?!"

[10:43:00.0000] <karlcow>
http://stiern.com/articles/usability/usability-in-icons/

[10:43:01.0000] <zewt>
icons, the bane of usability

[10:48:00.0000] <Ms2ger>
"Suggesting new feature: Accessible Image tag for Visually Impaired"

[10:49:00.0000] <TabAtkins>
Ms2ger: Specifically, they're asking for the ability to specify an alternate embossing-friendly image.

[11:06:00.0000] <dglazkov>
what's a good definition of the "Web Platform" that I could link to? I see lots of use of the term, but not a good definition we can use.

[11:06:01.0000] <nimbu>
http://platform.html5.org/ is the best I know but it is not a definition :P

[11:07:00.0000] <dglazkov>
maybe http://en.wikipedia.org/wiki/Open_Web_Platform

[11:07:01.0000] <dglazkov>
nimbu: that could work

[11:08:00.0000] <karlcow>
http://www.w3.org/wiki/Open_Web_Platform

[11:08:01.0000] <karlcow>
 hmmm I should add one to the wiki

[11:15:00.0000] <karlcow>
be free to improve it

[11:15:01.0000] <karlcow>
The Open Web Platform is the collection of open (royalty-free) technologies which enables the Web. Using the Open Web Platform, everyone has the right and the ability to implement a software component of the Web without requiring any approvals or waiving license fees. — http://www.w3.org/wiki/Open_Web_Platform

[11:16:00.0000] <zewt>
everyone has the ability?

[11:16:01.0000] <zewt>
heh

[11:16:02.0000] <dglazkov>
ability but not capacity

[11:17:00.0000] <zewt>
ability means capacity

[11:17:01.0000] <zewt>
not permission, heh

[11:17:02.0000] <dglazkov>
dur

[11:17:03.0000] <dglazkov>
word smithing is hard

[11:17:04.0000] <Philip`>
"legal ability"?

[11:17:05.0000] <Philip`>
The legal perspective seems like a pretty boring one to view it from, though

[11:18:00.0000] <karlcow>
pardon my French :)

[11:19:00.0000] <zewt>
well, it's boring to people who don't care about "open" to begin with :)

[11:19:01.0000] <karlcow>
maybe cut ability

[11:20:00.0000] <karlcow>
ability is cut

[11:20:01.0000] <dglazkov>
zewt: wow, internet has everything in it: http://www.ehow.com/how_4486601_use-words-capacity-ability-correctly.html

[11:23:00.0000] <zewt>
gmail scrolling header hack is eeeevil

[11:26:00.0000] <Philip`>
There's a lot more stuff that is open but not part of the web platform (like PHP and .NET and MJPEG and PDF and VRML and VML etc), than stuff that is not open but is part of the web platform (Flash, maybe H.264)

[11:27:00.0000] <Philip`>
so the "web platform" part seems more important and distinctive than the "open" part

[11:27:01.0000] <zewt>
at least nobody's calling it the "free open web platform"

[11:27:02.0000] <Ms2ger>
gif :)

[11:28:00.0000] <Philip`>
Surely that'd have to be the Free Libre Open Web Platform

[11:28:01.0000] <Philip`>
FLOWP is a nice enough acronym, too

[11:32:00.0000] <Ms2ger>
Leave the web out: "FLOP"

[12:00:00.0000] <karlcow>
There's a lot more people on the platform but not part of the web platform (like Paul and Dotty and Meg and Peter and Vince and Veronica etc), …

[12:46:00.0000] <dglazkov>
how could I use progress element in Mediawiki. AryehGregor, you know everything about that, right?

[12:47:00.0000] <AryehGregor>
dglazkov, it's not whitelisted.  You'd have to hack includes/Sanitizer.php.

[12:47:01.0000] <dglazkov>
AryehGregor: do I need to first hack into whatwg.org?

[12:47:02.0000] <AryehGregor>
I guess.

[12:48:00.0000] <gsnedders>
dglazkov: That shouldn't be that hard, should it?

[12:48:01.0000] <dglazkov>
gsnedders: :)

[12:48:02.0000] <dglazkov>
I know a cool hacker who can do this

[12:48:03.0000] <dglazkov>
Hixie: ^^^^ :)

[12:51:00.0000] <AryehGregor>
/me could also do it, in theory . . .

[12:51:01.0000] <AryehGregor>
Local hacks are no fun, though.

[12:51:02.0000] <AryehGregor>
They'll break as soon as you upgrade.

[12:51:03.0000] <AryehGregor>
In theory I could write and commit a proper patch, but that would take time.

[12:52:00.0000] <AryehGregor>
As usual, I am Hixie's slave and will do his bidding on this matter.

[12:56:00.0000] <Hixie>
wait, what?

[12:57:00.0000] <Hixie>
what are we talking about

[12:57:01.0000] <Ms2ger`>
Hixie, MediaWiki hacking

[12:58:00.0000] <Hixie>
AryehGregor: hey mad you do whatever you think is best, i don't know nothing about this wiki thing

[12:58:01.0000] <Hixie>
er

[12:58:02.0000] <Hixie>
"hey man"

[12:59:00.0000] <smaug____>
:)

[13:12:00.0000] <shepazu>
I am utterly confused by Powers' stance… I'd think that making HTML more modular, and doing the work in an even more open forum still within W3C and with a good patent policy, would be a clear win… AryehGregor, did you step on her pet gerbil or something?

[13:12:01.0000] <AryehGregor>
shepazu, most of us find Shelley a little hard to deal with at times.

[13:13:00.0000] <Ms2ger`>
shepazu, Utter confusion is pretty much my default feeling when dealing with her

[13:13:01.0000] <shepazu>
personalities aside, just the particulars of the topic seem sensible prima facie

[13:14:00.0000] <shepazu>
oh, well

[13:15:00.0000] <shepazu>
I suppose I should stop "lecturing" people to correct basic misinformation on Community Groups, and get back to real work

[13:15:01.0000] <Ms2ger`>
xkcd.com/386

[13:16:00.0000] <shepazu>
indeed

[13:29:00.0000] <AryehGregor>
/me wonders if WebKit people will get annoyed at all the editing bugs he's updating

[13:42:00.0000] <Hixie>
heycam|away: ping - wanted to ask you if there was an easy way to add an event handler attribute to something that is currently defined as a Foo[]

[13:42:01.0000] <Hixie>
(in webidl, obviously, not in js)

[13:53:00.0000] <AryehGregor>
TabAtkins, explanation for <http://www.w3.org/Bugs/Public/show_bug.cgi?id=13841>?

[14:50:00.0000] <zewt>
http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/0942.html heh again

[14:50:01.0000] <zewt>
should be a faq about that somewhere, it seems to be by a large margin the most common misunderstanding on the platform

[15:14:00.0000] <jgraham>
zewt: Are we looking at the same bug?

[15:23:00.0000] <zewt>
looks like the same old "someone doesn't know that events aren't dispatched until you return to the event loop", which it seems at least one person a month on the list or tracker doesn't get

[15:25:00.0000] <jgraham>
Well I guess. But it is a bit complex and hardly seems like the most common misunderstanding on the platform

[15:25:01.0000] <zewt>
it's one of the very few that I've seen over and over and over again

[15:26:00.0000] <jgraham>
I mean more common misunderstandings are things like "my XHTML doctype is somehow meaningful"

[15:26:01.0000] <jgraham>
Although maybe that particular one is going away a bit these days

[15:27:00.0000] <jgraham>
But sure, write a blog post or add it to the WHATWG FAQ or something

[15:40:00.0000] <gsnedders>
zewt: I know that… But I don't know when you return to the event loop.

[16:11:00.0000] <Hixie>
foolip: yt?

[16:12:00.0000] <Hixie>
should cues that the browser had to skip because of being too busy or because the cue was too short for the UA to get to it before it was already obsolete still fire events?

[16:12:01.0000] <Hixie>
i'm thinking yes

[16:12:02.0000] <Hixie>
especially pause-on-exit cues

[16:16:00.0000] <TabAtkins>
Hixie: Yeah, I think it should.


2011-08-20
[18:44:00.0000] <heycam>
Hixie, no, no easy way to do that

[18:45:00.0000] <heycam>
Hixie, you'd need to turn it into a regular interface with indexed properties

[18:45:01.0000] <heycam>
Hixie, btw I'm happy to try running the pdf generator on my machine if you want

[18:46:00.0000] <heycam>
Hixie, the downside of turning it into an interface is that there's then no way to get Array.prototype in the prototype chain

[18:46:01.0000] <heycam>
Hixie, but there is still a bug open about allowing something like that iirc...

[04:25:00.0000] <Ms2ger>
"Does web work add complex to web?"

[06:46:00.0000] <smaug____>
Grr...

[06:46:01.0000] <smaug____>
does anyone know how to get rid of Gnome instant search for good

[06:47:00.0000] <smaug____>
it tends to enable itself occasionally

[06:51:00.0000] <smaug____>
nm

[06:51:01.0000] <smaug____>
oops

[06:52:00.0000] <smaug____>
I wrote Gnome, when I meant Google. (I was searching for things about Gnome)

[07:10:00.0000] <msucan>
hello

[07:10:01.0000] <msucan>
which interface defines contentEditable?

[07:13:00.0000] <msucan>
found it

[10:21:00.0000] <Hixie>
heycam|away: can't i just inherit from the Foo[] interface or something like that? That would be convenient.


2011-08-21
[01:28:00.0000] <yihua>
這樣就加入了？

[06:30:00.0000] <nlogax>
twss

[10:04:00.0000] <swirrl>
LIST

[10:12:00.0000] <jcranmer>
this isn't an NNTP server ... ;-)

[10:21:00.0000] <Philip`>
jcranmer: Maybe swirrl was telling everybody to lean to one side in an attempt to tip the channel over

[12:25:00.0000] <jarek>
Hello

[12:26:00.0000] <jarek>
does any browser support this? http://dev.w3.org/html5/markup/input.color.html

[12:28:00.0000] <Ms2ger>
Opera, or so I hear

[12:29:00.0000] <nimbu>
yes

[12:29:01.0000] <nimbu>
i think chrome does too now

[12:30:00.0000] <jarek>
http://diveintohtml5.org/forms.html#type-color

[12:30:01.0000] <jarek>
this looks awesome, it uses the native color picker from OSX

[12:31:00.0000] <jarek>
but it seems to be disabled by default on WebKit

[12:35:00.0000] <jarek>
OK, it works on Opera

[12:36:00.0000] <jarek>
but initially it is displaying simple, non-standard color picker

[12:36:01.0000] <jarek>
the native color picker shows only after pressing "Other…" button

[12:36:02.0000] <jarek>
is there a way to display native picker right away?

[12:42:00.0000] <Ms2ger>
Ask Opera nicely? :)

[12:42:01.0000] <nimbu>
:)

[12:43:00.0000] <Ms2ger>
I hear nimbu can fix that ;)

[12:43:01.0000] <nimbu>
would def try!

[12:55:00.0000] <jarek>
but I need this feature on WebKit...

[12:55:01.0000] <jarek>
it doesn't work even on Chrome Canary

[12:59:00.0000] <jarek>
also, there is no alpha slider, is this a bug or by design?

[13:00:00.0000] <jarek>
since we can specify rgba() colors, it would make sense to have alpha slider as well

[13:02:00.0000] <jarek>
"The value attribute, if specified and not empty, must have a value that is a valid simple color." "A  simple color consists of three 8-bit numbers in the range 0..255, representing the red, green, and blue components of the color respectively, in the sRGB color space."

[13:02:01.0000] <jarek>
no alpha support? :/

[13:04:00.0000] <nimbu>
jarek: good question

[13:08:00.0000] <Hixie>
no alpha support, intentionally

[15:27:00.0000] <heycam>
gsnedders, I'm looking at your comment about Web IDL lacking a definition of [[Call]] for normal operations

[15:28:00.0000] <heycam>
gsnedders, and the problem of `alert("hi")` needing to work even in strict mode

[15:28:01.0000] <heycam>
gsnedders, do you know whether all methods of window need to work like this (i.e., have this set to the window)?

[15:29:00.0000] <gsnedders>
heycam: Dunno. I'm on vacation. :)

[15:29:01.0000] <heycam>
gsnedders, it's swapped out of your head then? :)

[15:29:02.0000] <gsnedders>
heycam: It'd be better for it to be consistent, though.

[15:29:03.0000] <heycam>
gsnedders, suppose so. dunno if people want to move to requiring "window." for newer functions though.

[15:30:00.0000] <gsnedders>
I doubt it.

[15:30:01.0000] <heycam>
k

[15:30:02.0000] <heycam>
well I might introduce an extended attribute to have this set correctly, which can then be put on Window in the idl

[15:31:00.0000] <gsnedders>
Yeah. Window is quite the special case, being the global object and all.


2011-08-22
[04:31:00.0000] <annevk>
"i think we dropped +xml knowledge because it broke sites" sad state of affairs that is :(

[04:32:00.0000] <Ms2ger>
Welcome to the web

[05:57:00.0000] <zcorpan>
Hixie: maybe the status boxes could have a field for http://www.w3.org/Bugs/Public/show_bug.cgi?id=13853

[08:52:00.0000] <jgraham>
TabAtkins: "Does web work add complex to web?" is clearly a question about whether it is possible to form multi-protein complexes on spider's webs by carefully controlled application of energy ("work") to the structure ("web")

[08:55:00.0000] <Ms2ger>
jgraham++

[08:56:00.0000] <Ms2ger>
TabAtkins, I was mostly thinking about TR/, though

[09:40:00.0000] <annevk>
Hixie: content type is including parameters arguably so I am not sure that is the correct term and it seems MIME type is somewhat incorrect (at least, thus argue the IETF folk)

[09:52:00.0000] <dglazkov>
good morning, Whatwg!

[09:52:01.0000] <Ms2ger>
Evening

[09:53:00.0000] <Philip`>
May this day bless you and your family, dglazkov!

[09:54:00.0000] <dglazkov>
:)

[10:11:00.0000] <jgraham>
May your loins be blessed with many fruit?

[11:19:00.0000] <karlcow>
annevk: http://backstage.soundcloud.com/2011/08/soundcloud-mobile-proxies/

[11:19:01.0000] <karlcow>
about CORS

[11:21:00.0000] <annevk>
someone should write a test suite

[11:22:00.0000] <Ms2ger>
We probably have one

[11:22:01.0000] <Ms2ger>
If you find it, I can probably get it submitted

[11:25:00.0000] <annevk>
I think the Gecko one uses a JavaScript server framework

[11:42:00.0000] <Ms2ger>
Probably

[11:49:00.0000] <Hixie>
annevk: well media type is just confusing

[11:49:01.0000] <Hixie>
annevk: maybe "contentType string"?

[11:49:02.0000] <Hixie>
annevk: it's not like this has any real meaning anyway

[11:49:03.0000] <Hixie>
annevk: it's not the type from the server, it's only barely the type the browser thinks the resource was, etc

[11:55:00.0000] <annevk>
"resource type"?

[11:55:01.0000] <timeless>
Ms2ger: ah, the hazards of writing long reviews

[11:55:02.0000] <timeless>
/me is reading heycam's response

[11:55:03.0000] <annevk>
could maybe rename MIME type to resource type

[11:56:00.0000] <timeless>
(for the second time actually, the first was using Gmail.java, which hides all quoted bits)

[11:59:00.0000] <Ms2ger>
timeless, mm, sending detailed feedback to D3E works better for that... You won't remember anything about your comments by the time you get a reply

[12:00:00.0000] <timeless>
Ms2ger: oh, heycam managed that too

[12:00:01.0000] <timeless>
i mean, i remembered vaguely being amused (when i read) about objects defying things

[12:00:02.0000] <timeless>
but i didn't really remember any of my comments in any useful level of detail

[12:01:00.0000] <timeless>
of course, the fact that my general buffer is probably at best 3 screens long, and each message i wrote was probably 30 didn't help

[12:01:01.0000] <timeless>
(and there were 5 messages0

[12:01:02.0000] <timeless>
s/0/)

[12:02:00.0000] <Ms2ger>
Yeah, that reply was quite the wall of text

[12:02:01.0000] <timeless>
it was actually amazingly readable in Gmail.java when you consider the input

[12:03:00.0000] <timeless>
(there are minor issues w/ how Gmail.java strips quoted sections, so you can get unrelated paragraphs running into eachother)

[12:03:01.0000] <timeless>
(but really, all in all, amazingly readable)

[12:03:02.0000] <timeless>
but absolutely useless for trying to do DoC

[12:03:03.0000] <timeless>
/me is using notepad for that

[12:03:04.0000] <timeless>
/me used notepad to review the spec too

[12:04:00.0000] <Ms2ger>
Pff, DoC

[12:08:00.0000] <Ms2ger>
Oh, LC ends tomorrow

[12:09:00.0000] <timeless>
on WebIDL?

[12:09:01.0000] <timeless>
grr

[12:09:02.0000] <Hixie>
annevk: resource type is fine by me

[12:09:03.0000] <Ms2ger>
/me raises "I don't understand your grammar"

[12:10:00.0000] <Hixie>
annevk: MIME type is what everyone calls it, like URL. Don't listen to the Julians of the world, their pedancy is not useful.

[12:10:01.0000] <Hixie>
annevk: i agree that .contentType is not the document MIME type though, even though it looks like a MIME type

[12:10:02.0000] <timeless>
interface TimelessParsingWhatWG { DOMString onWebIDL?; };

[12:10:03.0000] <timeless>
grr, no, that fails

[12:10:04.0000] <Ms2ger>
DOMString? onWebIDL

[12:10:05.0000] <timeless>
yeah

[12:11:00.0000] <Ms2ger>
No

[12:11:01.0000] <Ms2ger>
attribute DOMString? onWebIDL

[12:11:02.0000] <timeless>
well, hrm, that's awkward

[12:12:00.0000] <Ms2ger>
Did anybody raise an issue to drop the chapter about Java?

[12:12:01.0000] <timeless>
interface on { attribute DOMString? discussion; }; interface TimelessParsingWhatWg { attribute on WebIDL; };

[12:13:00.0000] <timeless>
we'll skip the discussion about using a the timeless module, because we wouldn't want any other groups to think there's a ghost of a chance of this supporting them

[12:13:01.0000] <timeless>
Ms2ger: but, that'

[12:13:02.0000] <timeless>
s parsable, right? :)

[12:14:00.0000] <Ms2ger>
I... think so

[12:14:01.0000] <Ms2ger>
I mean, I understand the code, but I don't understand the grammar

[12:15:00.0000] <timeless>
/me gets lost

[12:21:00.0000] <timeless>
ooh

[12:21:01.0000] <timeless>
/me reaches part 4

[12:22:00.0000] <annevk>
Hixie, it's not the specified one, but it's the determined one, no?

[12:51:00.0000] <timeless>
/me reaches part 5!

[12:51:01.0000] <timeless>
ooh, it only took me half an hour to read through the response to one section

[12:52:00.0000] <timeless>
that's not bad

[13:02:00.0000] <timeless>
so, wordpad claims my reply is 5 pages

[13:04:00.0000] <timeless>
outlook (which i unfortunately trust a bit more here) claims it's 15

[13:16:00.0000] <Ms2ger>
Who uploaded a contenteditable test to livedom?

[13:27:00.0000] <annevk>
WHATWG Weekly will have to wait till tomorrow

[13:47:00.0000] <Ms2ger>
Hixie, yt?

[13:52:00.0000] <AryehGregor>
Ms2ger, probably me.

[13:55:00.0000] <Ms2ger>
AryehGregor, so do you know about insertAdjacentHTML?

[13:55:01.0000] <AryehGregor>
Ms2ger, not really.

[14:05:00.0000] <timeless>
Ms2ger: ping

[14:06:00.0000] <Ms2ger>
Hi

[14:06:01.0000] <Ms2ger>
Do *you* know about insertAdjacentHTML? :)

[14:07:00.0000] <timeless>
i could

[14:07:01.0000] <timeless>
i'll trade

[14:09:00.0000] <timeless>
right, so, what about it?

[14:17:00.0000] <Hixie>
annevk: vaguely the determined one, but only vaguely

[14:17:01.0000] <Hixie>
annevk: e.g. createDocument() has one

[14:17:02.0000] <Hixie>
annevk: as do error messages

[14:17:03.0000] <Hixie>
Ms2ger: here

[14:19:00.0000] <Ms2ger>
http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#insertadjacenthtml

[14:19:01.0000] <Ms2ger>
Did you consider the case where the parent of target is a DocumentFragment?

[14:19:02.0000] <Hixie>
probably not

[14:20:00.0000] <Hixie>
doesn't seem to break in that case though

[14:20:01.0000] <Ms2ger>
...HTML fragment parsing algorithm. The algorithm optionally takes as input an Element node, referred to as the context element...

[14:21:00.0000] <Ms2ger>
So in that case, the context element isn't actually an element, right?

[14:26:00.0000] <Hixie>
holy cow, the ietf changes to the websocket protocol have introduced a massive security hole

[14:26:01.0000] <Hixie>
the server no longer explicitly lists the origin!

[14:31:00.0000] <AryehGregor>
SIGH: http://lists.w3.org/Archives/Public/spec-prod/2011JulSep/0035.html

[14:31:01.0000] <AryehGregor>
This is what happens when you try to make reasonable suggestions at the W3C.

[14:31:02.0000] <AryehGregor>
Well, let's see how it goes.

[14:32:00.0000] <Ms2ger>
I think HTML4 has a test suite, actually

[14:32:01.0000] <AryehGregor>
It would be hard for it to be a useful one.

[14:33:00.0000] <Ms2ger>
Well, it sure isn't useful

[14:34:00.0000] <Ms2ger>
http://www.w3.org/MarkUp/Group/Test/HTML401/tests/index.html

[14:34:01.0000] <AryehGregor>
This looks promising: https://lists.webkit.org/pipermail/webkit-dev/2011-August/017738.html

[14:34:02.0000] <AryehGregor>
Nice, requires W3C login?

[14:34:03.0000] <AryehGregor>
To be fair, HTML5 has practically no useful tests either, other than in a few tiny areas.

[14:35:00.0000] <timeless>
Hixie: why is that shocking?

[14:35:01.0000] <Ms2ger>
I think there's a public place as well, but I can't find it

[14:35:02.0000] <Hixie>
timeless: i was told that the ietf would bring expert review to the protocol

[14:36:00.0000] <Hixie>
timeless: i knew that that wasn't true, but i didn't realise it would be so bad that they would actually introduce vulnerabilities

[14:36:01.0000] <Ms2ger>
Ah, http://www.w3.org/MarkUp/Test/HTML401/current/tests/index.html

[14:37:00.0000] <Hixie>
Ms2ger: that test suite makes unwarranted assumptions about the conformance requirements in html4

[14:37:01.0000] <Ms2ger>
Well sure

[14:38:00.0000] <Ms2ger>
I only pointed out its existence, I was expecting it to suck

[14:39:00.0000] <Hixie>
i could have sworn i subscribed to member-webapps-cvs or whatever it's called

[14:39:01.0000] <Hixie>
but i don't seem to get e-mail from it

[14:39:02.0000] <Hixie>
wtf

[14:46:00.0000] <timeless>
so, wrt html4 test suite

[14:46:01.0000] <timeless>
if you want to blame someone

[14:46:02.0000] <timeless>
http://www.w3.org/MarkUp/Test/HTML401/current/tests/acknowledgements.html

[14:46:03.0000] <timeless>
"Tantek Çelik, who acted as senior mentor for this effort."

[14:46:04.0000] <timeless>
:)

[14:53:00.0000] <Hixie>
no blame to be assigned

[14:54:00.0000] <Hixie>
tantek's the only one even tried (well, tantek, me, and a few others -- there have been several such efforts over the years)

[14:54:01.0000] <Hixie>
the problem is the spec sucks :-)

[15:07:00.0000] <Hixie>
reminder to anyone who might be interested in giving talks about HTML5 to add yourself to http://wiki.whatwg.org/wiki/Presentations#Speakers

[15:07:01.0000] <Hixie>
i point people to that page pretty regularly

[15:37:00.0000] <Hixie>
what's the right terminology for saying that a list should be sorted first by some key A, then breaking ties on some key B, and breaking further ties on some key C?

[15:39:00.0000] <timeless>
secondary/tertiary sorting?

[15:39:01.0000] <Hixie>
complete this sentence: "Sort the list L "

[15:39:02.0000] <jcranmer>
it should be as if sorted by C then B then A using stable sorts

[15:40:00.0000] <Hixie>
there's no better terminology? bummer.

[15:40:01.0000] <timeless>
yeah, jcranmer 's text is correct

[15:40:02.0000] <timeless>
there's text for hashing

[15:41:00.0000] <jcranmer>
it's more precise but I don't like it that much

[15:41:01.0000] <timeless>
in hashing, it's double hashing where you specify the second hash function

[15:43:00.0000] <karlcow>
AryehGregor: I didn't reply to Liam, because I didn't think it would move forward the discussion.

[15:43:01.0000] <timeless>
Hixie: is your sorting going to specify how to handle unicode garbage? :)

[15:43:02.0000] <timeless>
/me goes to look up the technical term

[15:43:03.0000] <Hixie>
no

[15:43:04.0000] <Hixie>
it's sorting the tasks that get queued during the processing of captions

[15:43:05.0000] <Hixie>
for video

[15:46:00.0000] <timeless>
oh right

[15:46:01.0000] <timeless>
opera mobile *sucks* at dealing w/ large <textarea>s

[15:49:00.0000] <Philip`>
Hixie: Sort by a lexicographic order on (A, B, C)

[15:49:01.0000] <Hixie>
it wouldn't be lexicographic, but more importantly, where is it defined what that means?

[15:49:02.0000] <Philip`>
Why not lexicographic?

[15:49:03.0000] <Hixie>
well A is a time, B is a position in another list, and C is creation order :-)

[15:50:00.0000] <Hixie>
no graphs to lexic, as it were

[15:50:01.0000] <Philip`>
http://mathworld.wolfram.com/LexicographicOrder.html

[15:50:02.0000] <Philip`>
(Doesn't need to be anything to do with alphabets)

[15:50:03.0000] <Hixie>
i have no idea what that page is saying

[15:51:00.0000] <Hixie>
ok the wikipedia page clued me in

[15:51:01.0000] <Hixie>
what a confusing term to reuse

[15:53:00.0000] <timeless>
hi heycam

[15:53:01.0000] <heycam>
morning timeless

[15:53:02.0000] <timeless>
i have a longer response which i'm trying to paste

[15:53:03.0000] <timeless>
but i'm on the phone now

[15:54:00.0000] <timeless>
and i'll send another 1-8 messages

[15:54:01.0000] <heycam>
oh dear 8 messages? :)

[15:54:02.0000] <Philip`>
Hixie: Algebraists aren't necessarily known for non-confusingness

[15:55:00.0000] <Philip`>
but at least that particular term seems to only have a single definition in mathematics, so it's only unobvious and not ambiguous

[15:55:01.0000] <timeless>
main reply, 6 spun off items, and a review of the actual diff commit

[15:56:00.0000] <timeless>
plus an optional review of the current spec

[15:56:01.0000] <Hixie>
Philip`: well it would be ambiguous if one of the keys was a number

[15:56:02.0000] <timeless>
ok, so, maybe 2-9? :)

[15:56:03.0000] <Hixie>
Philip`: since lexical order for numbers (string definition) is not the same as lexical order for numbers (mathematical definition)

[16:00:00.0000] <Philip`>
If it's a lexicographic order over a product of types, you have to say what order to use for each of those types (or assume it's obvious to readers, like numbers are in increasing order) - the lexicographicness doesn't matter to those types

[16:26:00.0000] <timeless>
grr

[16:27:00.0000] <timeless>
Opera is trying to parse mail.google.com/mail/u/0/x/*/ as xml

[16:27:01.0000] <timeless>
and gives me an error w/ a link to reparse as html

[16:27:02.0000] <timeless>
for each link i click through in gmail

[16:29:00.0000] <timeless>
heycam: ok, the big email is outbound

[16:30:00.0000] <heycam>
timeless, thanks

[16:30:01.0000] <timeless>
that just leaves me w/ the diff, the 6 subissues, and re-reviewing the entire document

[16:34:00.0000] <timeless>
heycam: a note on VCS commit blame churn...

[16:34:01.0000] <timeless>
</x:codeblock>

[16:34:02.0000] <timeless>
and similar tags should really be on their own line ;-)

[16:35:00.0000] <heycam>
timeless, that might result in some extra white space in the published document

[16:35:01.0000] <timeless>
really

[16:35:02.0000] <timeless>
?

[16:35:03.0000] <heycam>
since it gets turned into a <pre>

[16:35:04.0000] <timeless>
yuck

[16:35:05.0000] <heycam>
maybe

[16:35:06.0000] <heycam>
can't remember exactly what my xslt does

[16:35:07.0000] <timeless>
get a better preprocessor :)

[16:36:00.0000] <timeless>
http://dev.w3.org/cvsweb/2006/webapi/WebIDL/Overview.xml.diff?r1=1.341;r2=1.342;f=h

[16:36:01.0000] <timeless>
near  Line 5765  results.numbers.length;       <span clas

[16:37:00.0000] <timeless>
i feel bad

[16:37:01.0000] <timeless>
the <must> in there ruins how <take(s)> should be used :(

[16:38:00.0000] <heycam>
timeless, yeah I noticed that and fixed it up in a following change

[16:38:01.0000] <timeless>
oh good

[16:39:00.0000] <timeless>
sorry about that (it happens in a bunch of spots)

[16:39:01.0000] <timeless>
this is where a dvcs is useful

[16:39:02.0000] <timeless>
commit changes for person

[16:39:03.0000] <timeless>
commit fixes for changes for person

[16:39:04.0000] <timeless>
commit further fixes for changes for changes...

[16:39:05.0000] <timeless>
merge

[16:39:06.0000] <timeless>
if you need even more changes, you go back to the premerged version and make them there

[16:40:00.0000] <timeless>
and when you want to show someone what you did for them, you can send either the leaf, or the delta between their origin point and their leaf

[16:40:01.0000] <timeless>
(no one does this)

[16:44:00.0000] <timeless>
isn't this illegal:

[16:44:01.0000] <timeless>
  [TreatUndefinedAs=Null] attribute DOMString? owner;

[16:44:02.0000] <timeless>
oh, no, that's TreatNulllAs

[16:44:03.0000] <timeless>
/me sighs

[16:44:04.0000] <timeless>
terribly confusing :)

[16:44:05.0000] <timeless>
/me should go home and sleep or something

[16:45:00.0000] <heycam>
:)

[16:45:01.0000] <timeless>
(when you get online, it's a hint that i've been at work too long)

[16:50:00.0000] <timeless>
ooh, i'm in the credits!

[16:51:00.0000] <timeless>
right, i've reached the end of the diff

[16:53:00.0000] <timeless>
ok, i'll send my comments about the diff, and then i'm giving up for the day

[16:53:01.0000] <timeless>
that leaves [x1..x6] + rereview

[16:53:02.0000] <timeless>
there's no way i can possibly do the rereview today, or really tomorrow

[16:54:00.0000] <timeless>
given it took me an entire week to do the initial review, and most of today just to review your reply to my review..

[16:55:00.0000] <timeless>
alright, sent

[16:55:01.0000] <timeless>
have a good day


2011-08-23
[17:02:00.0000] <Hixie>
TabAtkins: you still want cssElementMap right?

[17:07:00.0000] <Hixie>
TabAtkins: you still want cssElementMap right?

[17:10:00.0000] <Hixie>
i love, in "Detecting and Resolving Circular Relationships", that in the prose it confidently states "These relationships can be easily and reliably detected and resolved" but in the issue marker below that it says "Someone else needs to review this and make sure that I'm not missing any cycles"

[17:10:01.0000] <Hixie>
(http://dev.w3.org/csswg/css3-images/#element-reference)

[17:31:00.0000] <TabAtkins>
Hixie: Yes, I do.

[17:31:01.0000] <Hixie>
k

[17:32:00.0000] <TabAtkins>
Hixie: The issue will disappear in time, leaving only my confident declaration.

[17:32:01.0000] <Hixie>
any chance i could get you to change "represents" in "represents a paint source" to "provides"?

[17:33:00.0000] <Hixie>
"represents" has a defined meaning in the html spec so it is a little confusing in some of these sentences

[17:33:01.0000] <TabAtkins>
Sure.

[17:33:02.0000] <Hixie>
awesome

[17:35:00.0000] <TabAtkins>
done

[17:35:01.0000] <Hixie>
csselementmap isn't in the w3c copy of html, should "provides a paint source" be?

[17:35:02.0000] <Hixie>
or should i add csselementmap too?

[17:35:03.0000] <Hixie>
(i don't recall why it's not included)

[17:35:04.0000] <TabAtkins>
Yes.  The intention currently is that a display:none <img> provides a paint source.

[17:35:05.0000] <TabAtkins>
And sure, since it's not exposing any API or anything, might as well.

[17:36:00.0000] <Hixie>
/me is confused by the two most recent lines from TabAtkins 

[17:36:01.0000] <Hixie>
come again?

[17:37:00.0000] <TabAtkins>
The clause that invokes paint sources in the Image Values spec is applicable to more than just things assigned to the csselementmap.

[17:37:01.0000] <TabAtkins>
Since "provides a paint source" isn't an API detail (like csselementmap is), it shouldn't arouse any complaints if it's put in the spec.

[17:38:00.0000] <Hixie>
ah, ok. so you're saying paint source in w3c, csselementmap not

[17:38:01.0000] <Hixie>
ok

[17:38:02.0000] <Hixie>
sounds good, that's what i had done

[17:38:03.0000] <TabAtkins>
I don't care if you put csselementmap in.  I'm just saying that paint source is totally fine.

[17:39:00.0000] <Hixie>
ok done: http://html5.org/tools/web-apps-tracker?from=6519&to=6520

[17:40:00.0000] <TabAtkins>
Looks good.

[18:08:00.0000] <Hixie>
AryehGregor: see the widget where you pick people to share with in G+ for a use case of contenteditable="" that might be worth bearing in mind

[18:09:00.0000] <Hixie>
AryehGregor: also re http://www.w3.org/Bugs/Public/show_bug.cgi?id=12834#c10 -- it's not my scripts, i just use a bugzilla query of open bugs, so closed bugs don't get matched by the query :-)

[18:09:01.0000] <Hixie>
that's why i don't see comments on closed bugs

[01:54:00.0000] <nlogax>
Hixie: are you there?

[07:02:00.0000] <krijn>
Ms2ger: back, sorry! :)

[07:03:00.0000] <krijn>
My mom was showing somebody how fuses work.

[07:03:01.0000] <krijn>
She really shouldn't

[07:24:00.0000] <Ms2ger>
Heh

[08:06:00.0000] <jgraham>
Ms2ger: Technically that part of the spec is a SHOULD

[08:06:01.0000] <jgraham>
I'm not sure if we ever come to a conclusion about testing SHOULDs

[08:06:02.0000] <zcorpan>
annevk: wonder how many you've managed to confuse on events

[08:07:00.0000] <Ms2ger>
That's a bug in the spec I should file, then

[08:10:00.0000] <jgraham>
Ms2ger: I suppose one should argue that the spec should allow you to open a text/plain document in an external editor if you like

[08:12:00.0000] <annevk>
zcorpan, per matjas the fault is HTML4: http://www.w3.org/TR/html4/interact/scripts.html#h-18.2.3

[08:14:00.0000] <realityking>
I'm looking for a HTML5 parser for php and the only thing that looks promising so far is html5lib

[08:14:01.0000] <realityking>
but the php port hasn't been updated for 2 years

[08:15:00.0000] <realityking>
anyone know how mature it is or if there's a better alternative?

[08:16:00.0000] <annevk>
I haven't seen anything

[08:19:00.0000] <jgraham>
realityking: I don't really know anything about the PHP port, sorry. It won't have changed to match the current spec perfectly

[08:19:01.0000] <jgraham>
It might be slow depending on what your needs are

[08:19:02.0000] <zcorpan>
annevk: woah

[08:53:00.0000] <Ms2ger`>
annevk, I know... That's why I don't see us running XHR tests any time soon

[09:33:00.0000] <realityking>
jgraham, annevk: thanks for the info

[09:40:00.0000] <dglazkov>
good morning, Whatwg!

[09:41:00.0000] <dglazkov>
and evening, Ms2ger`

[09:41:01.0000] <Ms2ger`>
Good afternoon, dglazkov

[09:41:02.0000] <dglazkov>
doh

[09:42:00.0000] <Ms2ger`>
And may this day bless you and your family, to quote Philip`

[09:42:01.0000] <dglazkov>
and fruit in the pants, to qoute jgraham

[09:46:00.0000] <annevk>
:)

[11:06:00.0000] <gsnedders>
realityking: the PHP parser doesn't perfectly match the current spec, but should work reasonably well on not-overly long documents (parts of PHP's DOM impl have O(n^2) perf, so it hurts)

[11:06:01.0000] <realityking>
gsnedders: usage would be to parse articles to a DOM in a CMS so certain attributes/elements can be filtered on save

[11:06:02.0000] <realityking>
so documents shouldn't be too long

[11:06:03.0000] <realityking>
but I am slightly worried about it not perfectly matching the spec

[11:06:04.0000] <realityking>
I wish the PHP devs would just get a proper parser into the core...

[11:06:05.0000] <realityking>
thanks for your help tough

[11:06:06.0000] <gsnedders>
realityking: Basically, when there were major changes to the spec a couple of years ago, I spent what seemed like an eternity making the tedious changes to the Python version, and never had the motivation to make them in PHP too. Especially seeming I ended up never using it in the end.

[11:06:07.0000] <gsnedders>
(around script parsing, that was)

[11:07:00.0000] <realityking>
btw, is there any open source c library available for html5 parsing?

[11:08:00.0000] <gsnedders>
takkaria wrote one a few years back. No idea if it is maintained.

[11:08:01.0000] <Ms2ger`>
There will be a C++ one once hsivonen gets around to packaging his

[11:08:02.0000] <Ms2ger`>
Maybe even in this decade

[11:08:03.0000] <realityking>
:)

[11:09:00.0000] <Hixie>
foolip: previously, cues that were found to have been missed when the browser next checked for cues were just dropped on the floor

[11:09:01.0000] <Hixie>
foolip: i fixed that

[11:11:00.0000] <gsnedders>
Ms2ger`: You're hopeful. :)

[11:12:00.0000] <Ms2ger`>
One needs to be in this business... There's so little hope for the platform itself already ;)

[11:21:00.0000] <Ms2ger`>
gsnedders, so I hear you're good at that :)

[11:22:00.0000] <gsnedders>
Ms2ger`: Don't believe *her*. She lies, or so I'm told. But there again, testing *is* the future…

[11:29:00.0000] <Hixie>
i love the copyright notice on http://dev.w3.org/2011/webrtc/editor/webrtc.html

[11:29:01.0000] <Hixie>
it's the most explicit statement of hypocrisy i've ever seen the w3c make

[11:30:00.0000] <Ms2ger`>
Wow, that must be quite a thing

[11:30:01.0000] <Ms2ger`>
/me checks

[11:30:02.0000] <Ms2ger`>
>.<

[11:30:03.0000] <Hixie>
(amusingly, all the changes i could find are actually edits that made the spec worse)

[11:31:00.0000] <Ms2ger`>
Following a long tradition of copyrighting junk, then

[11:32:00.0000] <AryehGregor>
karlcow, I'm not clear on who exactly has influence on the decision here, or who's going to make it, so I thought I'd reply to be on the safe side.  How exactly will a decision get made, and who has a say here?  Is there anyone in particular I have to convince or can ignore?

[11:33:00.0000] <AryehGregor>
Hixie, which widget where you pick people to share with?

[11:33:01.0000] <AryehGregor>
Hixie, and okay.

[11:34:00.0000] <Hixie>
when you do a post, the one where you pick the circles

[11:34:01.0000] <karlcow>
I'm trying to not focus on who but what are the requirements and then see what we can propose to achieve these requirements.

[11:36:00.0000] <karlcow>
liam could have had a real proposal by saying we need to identify keywords, code, etc. I do not think the grammar of HTML whatever version is a guarantee, but the documentation of classname on HTML is a lot more for the purpose of the spec is what is needed. At the same time it was never really required in the past.

[11:36:01.0000] <karlcow>
so it's why the argument is a bit MOOT

[11:37:00.0000] <karlcow>
too much talk, not enough how-do-we-do-it

[11:47:00.0000] <AryehGregor>
Hixie, which part is contenteditable?  I don't see it.

[11:47:01.0000] <AryehGregor>
(too much div soup!)

[11:47:02.0000] <Hixie>
oh i don't know if it's actually implemented as contenteditable (though i can't see how else it could be sanely done)

[11:47:03.0000] <AryehGregor>
I don't see any part that looks editable.

[11:48:00.0000] <Hixie>
i just mean the concept of a control that contains both editable text or immutable "chips"

[11:48:01.0000] <Hixie>
let me get you a screenshot

[11:48:02.0000] <AryehGregor>
Which part of the text is editable?

[11:48:03.0000] <AryehGregor>
Oh, I see.

[11:48:04.0000] <AryehGregor>
Right.

[11:48:05.0000] <AryehGregor>
It's an input, actually.

[11:49:00.0000] <AryehGregor>
They somehow cleverly make it not look like one.

[11:49:01.0000] <AryehGregor>
But the non-editable stuff is actually not part of the editable area at all, it's just sitting next to it.

[11:49:02.0000] <AryehGregor>
Still, this is what <span contenteditable=false> is for, yea.h

[11:49:03.0000] <AryehGregor>
yeah.

[11:49:04.0000] <AryehGregor>
Although that's a huge PITA to spec/implement correctly, it adds corner cases to *everything*.

[11:50:00.0000] <Hixie>
crap gotta go. bbiab with screenshots.

[12:56:00.0000] <AryehGregor>
Hmm.  When I customize my sync preferences in Chrome and hit OK, it waits for a long time with a little spinner thing.  I wonder whether it's frozen or actually doing something.

[13:09:00.0000] <annevk>
Hixie, seems to be within their right

[13:12:00.0000] <AryehGregor>
Hixie, yeah, I don't see anything wrong with that copyright notice.

[13:12:01.0000] <AryehGregor>
They're reproducing the original notice and making the copyright status clear.

[13:48:00.0000] <timeless>
/me ponders

[13:49:00.0000] <timeless>
is there an easy way to find out which version of a document @w3 is the one for a given TR?

[13:49:01.0000] <timeless>
or LC or whatever

[13:50:00.0000] <Ms2ger`>
Mm?

[13:50:01.0000] <timeless>
http://dev.w3.org/cvsweb/2006/webapi/WebIDL/Overview.xml?r1=1.353;f=h#rev1.353

[13:50:02.0000] <timeless>
i've decided to compare 1.339 to tip

[13:50:03.0000] <timeless>
but i'm not sure if that's the right choice

[13:51:00.0000] <Ms2ger`>
No way to do that in general, but 339 sounds right

[13:52:00.0000] <karlcow>
timeless: do you mean which revision number has been published into TR/ space?

[13:52:01.0000] <timeless>
karlcow: i'm pretty sure if i did, i wouldn't need to ask :)

[13:52:02.0000] <karlcow>
hey?

[13:53:00.0000] <timeless>
oh

[13:53:01.0000] <timeless>
yes, sorry, i misread your question

[13:53:02.0000] <karlcow>
:)

[13:53:03.0000] <timeless>
yes, that's what i'm trying to get

[13:53:04.0000] <timeless>
/me misread `mean` as `know`

[13:53:05.0000] <timeless>
/me has no idea how

[13:53:06.0000] <karlcow>
no idea if it's tracked, but that could be useful indeed for future version.

[13:54:00.0000] <karlcow>
there could be even a link of the diff between the tip and the published rev number

[13:56:00.0000] <timeless>
Ms2ger`: interface screwy { attribute DOMString getter; };

[13:56:01.0000] <timeless>
there's no harm in that, right?

[13:57:00.0000] <Ms2ger`>
In what?

[13:57:01.0000] <timeless>
that attribute name

[13:57:02.0000] <jgraham>
Don't think so

[13:57:03.0000] <Ms2ger`>
You probably need _getter

[13:58:00.0000] <timeless>
/me will have to give it as input to the spec and see what comes out

[13:59:00.0000] <timeless>
btw, how was the earthquake (to whomever it may concern)

[13:59:01.0000] <nimbu>
it was a GOOD ONE

[13:59:02.0000] <nimbu>
/me lives on the west coast

[13:59:03.0000] <timeless>
nimbu: `the only good earthquake is one too far away from me to be noticed`? :)

[13:59:04.0000] <aho>
so, spilt your coffee? :)

[14:00:00.0000] <nimbu>
timeless: exactly

[14:01:00.0000] <timeless>
…

[14:01:01.0000] <timeless>
...

[14:01:02.0000] <timeless>
/me ponders

[14:01:03.0000] <timeless>
Ms2ger / jgraham : ... / …

[14:02:00.0000] <timeless>
+            <pre class='syntax'><i>return-type</i> <i>identifier</i>([<em>extended-attributes</em>] <i>type</i> <i>identifier</i>, [<em>extended-attributes</em>] <i>type</i> <i>identifier</i>, …);</pre>

[14:02:01.0000] <timeless>
+            <pre class='syntax'><i>return-type</i> <i>identifier</i>(<i>type</i><em>...</em> <i>identifier</i>);

[14:02:02.0000] <timeless>
+<i>return-type</i> <i>identifier</i>(<i>type</i> <i>identifier</i>, <i>type</i><em>...</em> <i>identifier</i>);</pre>

[14:02:03.0000] <timeless>
/me ponders

[14:03:00.0000] <timeless>
i guess the last two lines are actually focused on the ellipsis

[14:03:01.0000] <jgraham>
I think I was wrong

[14:03:02.0000] <jgraham>
But I haven't quite proved it yet

[14:30:00.0000] <timeless>
oh

[14:30:01.0000] <timeless>
interface in {};

[14:30:02.0000] <timeless>
is that likely to cause pain for people? :)

[14:31:00.0000] <timeless>
.. probably not, i guess

[14:45:00.0000] <dglazkov>
posted some stuff to read for y'all: http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/0975.html

[14:47:00.0000] <timeless>
* { visibility: hidden; } code, .syntax { visibility: visible !important; }

[14:47:01.0000] <timeless>
is a good way to proofread stuff :)

[14:48:00.0000] <timeless>
code, code > *, .syntax, .syntax > * { visibility: visible !important; }

[14:48:01.0000] <timeless>
works better :(

[14:52:00.0000] <TabAtkins>
Hehe.

[14:56:00.0000] <timeless>
well, both proved what i needed, but the latter showed me more stuff to review :(

[14:56:01.0000] <timeless>
http://pastebin.mozilla.org/1308553

[14:56:02.0000] <timeless>
TabAtkins: any guesses how i got that?

[14:57:00.0000] <TabAtkins>
A regex that looks for "an [aeiou]"?

[14:57:01.0000] <TabAtkins>
Or I guess the other way round.

[14:57:02.0000] <timeless>
`can i`

[14:58:00.0000] <timeless>
i did it the old fashioned way, for each (letter not in aeiou) find ` an {letter}`

[14:58:01.0000] <timeless>
(using firefox)

[14:58:02.0000] <TabAtkins>
Bah.  Use a regex text editor.  That would have been slightly easier.  ^_^

[14:58:03.0000] <TabAtkins>
 /an [^aeiou]/

[14:59:00.0000] <timeless>
TabAtkins: those fail on ` an <b>car</b>`

[14:59:01.0000] <TabAtkins>
Get one that understands HTML as well.

[14:59:02.0000] <TabAtkins>
(I don't know if one exists.)

[14:59:03.0000] <timeless>
/me chuckles

[14:59:04.0000] <timeless>
emacs might count

[14:59:05.0000] <timeless>
word probably does count

[14:59:06.0000] <timeless>
but one wouldn't typically call either a text editor

[14:59:07.0000] <timeless>
the former is an operating system

[14:59:08.0000] <timeless>
the latter is a word processor

[15:00:00.0000] <TabAtkins>
Alternately: copypasta from the browser to a text file, *then* search.  Markup is lost.

[15:00:01.0000] <timeless>
/me nods

[15:01:00.0000] <timeless>
/me will actually ask Word what it thinks of the document

[15:01:01.0000] <timeless>
(word tends to do a good job of drawing green squiglies)

[15:03:00.0000] <Philip`>
/an (<[^>]*>)*[^b-df-hj-nptv-z]/ ?

[15:03:01.0000] <timeless>
Philip`: that fails `can i` :)

[15:03:02.0000] <Philip`>
/\ban (<[^>]*>)*[^b-df-hj-nptv-z]/ ?

[15:04:00.0000] <timeless>
that might work

[15:04:01.0000] <timeless>
/me thinks it's more trouble than it's worth

[15:05:00.0000] <timeless>
does that fail ` an <em><a>foo` ?

[15:05:01.0000] <timeless>
oh, you have another * there

[15:05:02.0000] <timeless>
it might fail ` an <em> <a> foo` :)

[15:05:03.0000] <Philip`>
Only weirdos would write markup like that :-p

[15:06:00.0000] <timeless>
also, that probably fails any case of <a foo="bar>fp">

[15:06:01.0000] <timeless>
and yes, the author in question *does* use unencoded >'s

[15:07:00.0000] <TabAtkins>
Unencoded > is the best.  Much eaiser to read.

[15:07:01.0000] <timeless>
s/eais/eas/

[15:07:02.0000] <timeless>
s/eais/easi/

[15:08:00.0000] <TabAtkins>
eaiesr

[15:08:01.0000] <timeless>
(it's hard to correct what i can't see..)

[15:08:02.0000] <Philip`>
/\ban (<[^>]*>)*(?!html|http|href|hgroup|rt|rp|li|xml|svg|hour)[b-df-hj-np-tv-z]/i

[15:09:00.0000] <Philip`>
That's less broken and has less false positives

[15:09:01.0000] <Philip`>
and finds a few errors in the HTML5 spec

[15:09:02.0000] <timeless>
<em>an ...

[15:09:03.0000] <timeless>
otoh

[15:09:04.0000] <timeless>
if it finds errors in the html5 spec, then it's a win!

[15:10:00.0000] <Philip`>
/\ban/ matches "<em>an"

[15:11:00.0000] <timeless>
yeah

[15:11:01.0000] <timeless>
but technically you don't want to

[15:11:02.0000] <timeless>
c<em>an

[15:11:03.0000] <timeless>
not sure when you'd hit that, but..

[15:11:04.0000] <Philip`>
"<dt>An <code>hr</code> element</dt>"

[15:11:05.0000] <Philip`>
Only weirdos use markup in the middle of words :-p

[15:11:06.0000] <timeless>
wanna do a search in html5 for that? :)

[15:11:07.0000] <Hixie>
annevk, AryehGregor: nothing legally wrong with it. It's just hypocritical is all.

[15:13:00.0000] <timeless>
code, code > *, .syntax, .syntax > *, .idlstring, .esstring, .grammar { visibility: hidden !important; } * { visibility: default; }

[15:13:01.0000] <timeless>
do a search for `"`

[15:13:02.0000] <Philip`>
Lots of <var>index</var>th and <code>Document</code>s etc

[15:13:03.0000] <timeless>
:)

[15:13:04.0000] <Philip`>
but anyway, false positives don't hurt much :-)

[15:13:05.0000] <timeless>
/me is trying to decide if any of these `"`s should be tagged somehow

[15:14:00.0000] <timeless>
depends how long your document is :)

[15:14:01.0000] <timeless>
a colleague passed around a screen shot of a dialog from MS Word

[15:14:02.0000] <timeless>
the dialog was Word's "I give up, there are too many errors in this spec"

[15:14:03.0000] <Philip`>
HTML5 is pretty long, and my earlier regexp only finds 44 matches

[15:14:04.0000] <timeless>
one conclusion: spec isn't written by native speakers

[15:14:05.0000] <timeless>
alternate: spec is too long

[15:15:00.0000] <Hixie>
AryehGregor: do you have a mac? another example of contenteditable is the native date format editor (http://junkyard.damowmow.com/494)

[15:15:01.0000] <Philip`>
which was few enough that I gave up adding exceptions to the list

[15:15:02.0000] <timeless>
ok :)

[15:15:03.0000] <timeless>
Hixie: i remember that widget

[15:17:00.0000] <Philip`>
Hixie: Do you want bug reports of phrases like "an hostile" and "an platform array object" and "an global" and "an published" and "an JavaScript" and "an SYNTAX_ERROR"?

[15:17:01.0000] <Hixie>
sure

[15:17:02.0000] <Hixie>
one bug for all ideally

[15:17:03.0000] <Philip`>
and "an VideoTrackList"

[15:18:00.0000] <Hixie>
if you could just paste the actual markup in the bug that would be ideal

[15:18:01.0000] <Hixie>
so i can just search for that

[15:18:02.0000] <Hixie>
but if you don't have it that's fine

[15:18:03.0000] <timeless>
code, code > *, .syntax, .syntax > *, .idlstring, .idlvalue, .esstring, .grammar, .sym { visibility: hidden !important; } * { visibility: default; }

[15:18:04.0000] <timeless>
ended up working for my purposes

[15:18:05.0000] <Philip`>
I can paste the regexp :-)

[15:19:00.0000] <timeless>
heh

[15:19:01.0000] <Hixie>
if it's a regexp against the markup that would be perfect

[15:19:02.0000] <Hixie>
i can just stick it in my build script

[15:19:03.0000] <Hixie>
to catch future errors

[15:19:04.0000] <timeless>
woohoo, automation!

[15:20:00.0000] <Hixie>
anyone know if there's a way to renumber a screen in screen(1)?

[15:20:01.0000] <timeless>
http://brainscraps.wikidot.com/screnum

[15:20:02.0000] <timeless>
in short, `yes`

[15:21:00.0000] <timeless>
/me opts not to decode the code

[15:22:00.0000] <Hixie>
yeah that code is weird

[15:22:01.0000] <timeless>
regexps are much saner than shell

[15:22:02.0000] <Hixie>
seems to involve nested screen sessions

[15:22:03.0000] <timeless>
screen -l = list screens more or less iirc

[15:23:00.0000] <timeless>
screen -x = attach and share a screen session in case it's already attached (don't disconnect the currently attached clients)

[15:23:01.0000] <Hixie>
aha, ^A : number <n>

[15:23:02.0000] <timeless>
but yeah, it does do stuff in a screen session

[15:24:00.0000] <timeless>
i suspect that's to avoid damaging your console or something

[15:24:01.0000] <timeless>
or poisoning the session with your dimensions or who knows

[15:24:02.0000] <Hixie>
nice

[15:24:03.0000] <Hixie>
it works

[15:24:04.0000] <Hixie>
^A :number 2

[15:24:05.0000] <Hixie>
swaps the current window with window #2

[15:25:00.0000] <hober>
nice

[15:26:00.0000] <timeless>
heycam: is there a rule you use for <foo attrib="bar"> v. <foo attrib='bar'> ?

[15:26:01.0000] <heycam>
timeless, in the source? no, just depends on my mood.

[15:26:02.0000] <timeless>
ok :)

[15:26:03.0000] <heycam>
most single quotes though I think?

[15:26:04.0000] <timeless>
i believe so

[15:27:00.0000] <heycam>
I would not accept a patch that made them consistent :)

[15:27:01.0000] <timeless>
i think you've reached the point where it wouldn't cost much to standardize on ''

[15:27:02.0000] <timeless>
not?

[15:27:03.0000] <timeless>
that's ok, i haven't sent patches anywhere in a while :)

[15:27:04.0000] <TabAtkins>
My rule is to use " when my hands are on home row, and ' when a hand is on the arrows and it would be inconvenient to hit shift.

[15:27:05.0000] <Philip`>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13879 - I think that's enough exceptions

[15:28:00.0000] <Hixie>
/me hopes that regexp is in a format grep accepts...

[15:28:01.0000] <Hixie>
let's see...

[15:28:02.0000] <timeless>
heh

[15:28:03.0000] <timeless>
w/ Hixie ?! ?

[15:28:04.0000] <Philip`>
'ack' accepts it

[15:29:00.0000] <Philip`>
I expect anything non-Perl would be unhappy with (?!...)

[15:29:01.0000] <timeless>
grr, this irc client catches ctrl-tab as tab as tab-complete

[15:29:02.0000] <Philip`>
I suppose one problem is this doesn't cope with newlines after "an"

[15:30:00.0000] <timeless>
s/ /\s/; s!/i!/im! ?

[15:30:01.0000] <timeless>
or something like that, maybe sprinkle in a + and perhaps escape the backslash..

[15:31:00.0000] <Philip`>
Then you need a grep-like tool that doesn't process line-by-line, but does print enough context for you to find each match

[15:32:00.0000] <timeless>
w/ perl you could print $. or something (there's a line accumulator floating around, i can't remember if that's it)

[15:32:01.0000] <timeless>
actually, that might not do the right thing

[15:32:02.0000] <timeless>
/me ponders

[15:32:03.0000] <timeless>
but yeah, it gets painful

[15:32:04.0000] <timeless>
you can capture (.{200}) before and after and print them :)

[15:33:00.0000] <Hixie>
can you see what's wrong with this other than the newline thing you mentioned?:

[15:33:01.0000] <Hixie>
perl -ne 'print "$.: $_" if /\ban(<[^>]*>)*(?!(http|https|href|hgroup|rt|rp|li|xml|svg|hour|hr|xhtml|xslt|xbl|nntp|m[ions]|mtext|merror|h[1-6]|xmlns|xpath|s|x|sgml|huang|srgb|rsa)\b|html)[b-df-hj-np-tv-z]/i' source | perl -lpe 'print "\nPossible article problems:" if $. == 1'

[15:33:02.0000] <Hixie>
i'm getting way too many false positives

[15:33:03.0000] <Hixie>
e.g. it matches:10:  !    - element, attribute, content model, and interface indexes

[15:33:04.0000] <Philip`>
You need at least a space or a \s+ after the "an"

[15:33:05.0000] <Hixie>
oh oops

[15:33:06.0000] <Hixie>
copy pasta

[15:33:07.0000] <timeless>
yeah, blame your web browser :)

[15:33:08.0000] <Hixie>
that's better

[15:34:00.0000] <Hixie>
awesome

[15:34:01.0000] <Hixie>
thanks

[15:34:02.0000] <Hixie>
i added it to my build script

[15:35:00.0000] <Hixie>
will fix errors later, i'm in the middle of an edit

[15:35:01.0000] <timeless>
what does it say that i correctly remembered $. -- i don't use it, and haven't touched perl in months?

[15:35:02.0000] <timeless>
/me sighs

[15:35:03.0000] <Hixie>
i use $. quite often

[15:36:00.0000] <timeless>
i suspect the last time i did was on my text to html/atom script

[15:36:01.0000] <Hixie>
Philip`: if you figure out a way to make this work for multiline stuff while still being a short script let me know, i could use such magic in other efforts

[15:36:02.0000] <timeless>
which was probably a couple of years ago

[15:36:03.0000] <Philip`>
I'm assuming everyone pronounces HTML as "aitch tee em ell", not as "hotmeal" or similar

[15:36:04.0000] <jgraham>
/me is very suspicious of regexps to prase phonetics using the rule "all vowels plus some listed exceptions"

[15:36:05.0000] <timeless>
Hixie: are you on a memory constrained system?

[15:37:00.0000] <jgraham>
*parse

[15:37:01.0000] <timeless>
if not, you could go the mxr line counting approach

[15:37:02.0000] <jgraham>
/me assumes Hixie doesn't have an infinte amount of ticker tape

[15:37:03.0000] <timeless>
and capture your context and replace all line feeds up to it and then count that replacement

[15:37:04.0000] <Hixie>
timeless: yes, but more importantly i can write the script as a multiline one without help, i just can't work out a way to do a one-liner to do this :-)

[15:37:05.0000] <Hixie>
if i had infinite ram i'd just do the regexps against the whole file

[15:38:00.0000] <timeless>
i can imagine a way that cheats a bit

[15:38:01.0000] <timeless>
takes the file, and each time it encounters an<tags>{linebreak} it swaps the {linebreak} and an<tags>

[15:39:00.0000] <timeless>
and then applies the rule to that resulting stream

[15:39:01.0000] <Hixie>
there are many trivial ways to do it if you can write more than one line of code :-P

[15:39:02.0000] <Hixie>
the point is that this is currently a -ne perl one-liner

[15:40:00.0000] <timeless>
oh, you have access to variables in -ne

[15:40:01.0000] <timeless>
you can easily do:

[15:40:02.0000] <Philip`>
perl -le '@_=<>; for $i (0..$#_) { $_="$_[$i]$_[$i+1]"; print "$i: $_" if /\ban\s+etcetcetc/im }' source

[15:40:03.0000] <timeless>
$_ = $a.$_ if $a; $a = "an" if s/an$//;

[15:40:04.0000] <Philip`>
(Surely not the best way but it seems to work)

[15:41:00.0000] <timeless>
grr, don't forget to clear $a in that first if :)

[15:41:01.0000] <Hixie>
lordy

[15:41:02.0000] <Hixie>
you guys give perl a bad name :-P

[15:44:00.0000] <Philip`>
One of my variables had an alphanumeric name - it could have been worse

[15:44:01.0000] <timeless>
heh

[15:45:00.0000] <timeless>
>  The following changes were made since the previous Windows cumulative time zone update:

[15:45:01.0000] <Philip`>
perl -le '@_=<>; for $i (0..$#_) { $_="$_[$i]$_[$i+1]"; print "$i: $_" if --$| and /\ban\s+etcetcetc/im }' source

[15:45:02.0000] <timeless>
> Turkey Standard Time: A new Windows time zone was created: Turkey Standard Time. The display name is "(UTC+2:00) Istanbul." The government of Turkey has decided to change the 2011 DST schedule. It will start on Monday March 28 instead of Sunday March 27. "Istanbul" was removed from the display name of the "(UTC+2:00) Athens, Bucharest, Istanbul" time zone.

[15:45:03.0000] <Philip`>
That'll stop it reporting each line twice

[15:46:00.0000] <Philip`>
Oh, but I suppose it'll break if 'an' is at the end of a second line, so ignore that

[15:46:01.0000] <timeless>
> Egypt Standard Time [Display Name "(UTC+2:00) Cairo"]: The government of Egypt has canceled daylight saving time.

[15:46:02.0000] <timeless>
/me chuckles

[15:47:00.0000] <timeless>
ooh, Newfoundland is -3:30

[15:48:00.0000] <timeless>
i didn't know i could get a half hour time zone w/o leaving this hemisphere

[15:52:00.0000] <timeless>
oh

[15:52:01.0000] <timeless>
Philip`: ?

[15:52:02.0000] <timeless>
can you ensure that `an [ImplicitThis]` is properly handled? :)

[15:53:00.0000] <timeless>
(or is that too Web IDL specific?)

[15:58:00.0000] <timeless>
ok, i'm done reviewing the diff

[15:58:01.0000] <timeless>
there's a dancing in the dark movie tonight

[15:58:02.0000] <timeless>
/me should figure out dinner, probably skip dancing, 

[16:03:00.0000] <timeless>
heycam: ?

[16:03:01.0000] <timeless>
http://dev.w3.org/2006/webapi/WebIDL/#idl-special-operations

[16:04:00.0000] <timeless>
Special operations are declared by using one or more special keywords keywords in an operation declaration.

[16:04:01.0000] <timeless>
MS Word objects to `keywords keywords`

[16:11:00.0000] <shepazu>
annevk: I don't understand matter's #ffs … any explanation?

[16:12:00.0000] <timeless>
awww

[16:12:01.0000] <timeless>
Word crashed

[16:12:02.0000] <timeless>
heycam: it's all your fault!

[16:13:00.0000] <timeless>
[4.2.20.3]

[16:20:00.0000] <TabAtkins>
shepazu: Do you not understand #ff at all, or just mattur's ones?

[16:20:01.0000] <TabAtkins>
shepazu: Oh wait, I see.  Never mind.

[16:20:02.0000] <TabAtkins>
That's a "for fucks sake", I assume.

[16:21:00.0000] <TabAtkins>
So, I dunno.

[16:21:01.0000] <shepazu>
like, I would use that when I think someone said something extremely stupid

[16:21:02.0000] <shepazu>
so, maybe he thought my statement was really dumb… :)

[16:21:03.0000] <shepazu>
oh, well

[16:21:04.0000] <nimbu>
shepazu: no i think he sez DO IT ALREADY

[16:21:05.0000] <nimbu>
clearly annevk can explain himself

[16:22:00.0000] <nimbu>
#ffs

[16:22:01.0000] <shepazu>
yeah, I reckon that's one interpretation

[16:23:00.0000] <shepazu>
/me will probably get some tsk-tsks for saying that...

[16:25:00.0000] <heycam>
timeless, thanks fixed (locally)

[16:25:01.0000] <astearns>
/me just accidentally created a #ffs channel clicking on shepazu's comment above

[16:26:00.0000] <shepazu>
astearns, what a stupid mistake! #ffs

[16:26:01.0000] <shepazu>
see, that's how you use #ffs :)

[16:26:02.0000] <astearns>
it was blue, I had to click...

[16:49:00.0000] <mattur>
shepazu: I agree with your comment, and meant DO IT ALREADY

[16:49:01.0000] <shepazu>
mattur: ok

[16:50:00.0000] <shepazu>
mattur: to be fair, I don't think there's any great rush, it will happen eventually (I'd guess sooner rather than later)

[16:51:00.0000] <shepazu>
I can actually see both sides of the topic

[16:53:00.0000] <Hixie>
for registerProtocolHandler() and registerContentHandler(), can anyone think of any states the browser could be in beyond these?:

[16:53:01.0000] <Hixie>
http://junkyard.damowmow.com/495

[16:59:00.0000] <mattur>
shepazu: sooner would be good :)

[16:59:01.0000] <shepazu>
mattur: agreed


2011-08-24
[17:00:00.0000] <mattur>
shepazu: the #ffs was short for: gosh, this is a slightly exasperating; the W3C /should/ be eating its own dogfood

[17:01:00.0000] <Hixie>
(and always has before!)

[01:28:00.0000] <annevk>
morning

[01:43:00.0000] <zcorpan>
Hixie: why "+     <li>All <a href=#xml-mime-type title="XML MIME type">XML MIME types</a></li>" ?

[01:46:00.0000] <zcorpan>
Hixie: the idea of supporting all unknown mime types ending with +xml has failed. web compat requires non-support.

[01:53:00.0000] <Ms2ger>
<Hixie> you guys give perl a bad name :-P

[01:53:01.0000] <Ms2ger>
Because it didn't have one yet?

[01:55:00.0000] <jgraham>
Please evreyone stop saying that. Every time I read it to the tune of Bon Jovi. That's practically illegal under the Geneva convention.

[01:56:00.0000] <annevk>
zcorpan, for XHR too?

[01:58:00.0000] <zcorpan>
annevk: for navigation. dunno about xhr, doesn't xhr assume xml if you use an unknown type (even if it doesn't end in +xml)?

[02:00:00.0000] <annevk>
no for missing or malformed content-type only

[02:16:00.0000] <zcorpan>
annevk: in opera i get an XMLDocument for text/foo

[02:24:00.0000] <annevk>
Opera is buggy

[02:24:01.0000] <annevk>
actually causes us issues

[02:25:00.0000] <zcorpan>
do other browsers do the +xml thing for xhr?

[02:25:01.0000] <annevk>
WebKit does at least

[02:25:02.0000] <zcorpan>
ok

[02:25:03.0000] <annevk>
WebKit is pretty compliant

[02:26:00.0000] <zcorpan>
ok, then +xml hasn't completely failed :)

[02:27:00.0000] <zcorpan>
but it's not clear to me why we should ban registration of unknown +xml types

[02:28:00.0000] <zcorpan>
hmm i guess i should file a bug on navigation

[02:42:00.0000] <zcorpan>
turns out google earth was the major +xml abuser

[02:42:01.0000] <AryehGregor>
Hixie, I don't have a Mac, no.

[02:43:00.0000] <annevk>
blocking all XML types btw blocks application/atom+xml

[02:43:01.0000] <annevk>
which I think web pages might legitimately handle

[02:43:02.0000] <AryehGregor>
(also, I'll be in Toronto for the next two days, so if anyone wants to contact me, e-mail is a better bet than IRC)

[02:43:03.0000] <zcorpan>
annevk: yeah

[02:44:00.0000] <annevk>
the web+ scheme convention is interesting

[02:44:01.0000] <annevk>
I wonder how the IETF will react

[02:44:02.0000] <annevk>
prolly late

[02:44:03.0000] <zcorpan>
annevk: although the last bullet point bans atom also

[02:45:00.0000] <annevk>
I wonder if that is for application/pdf and the like

[02:51:00.0000] <zcorpan>
<http://www.w3.org/mid/E1QvsLJ-0007QA-SV⊙jwo> - this is why we should always err on the side of REOPEN instead of WONTFIX+TrackerIssue when somebody does REOPEN+TrackerIssue

[02:55:00.0000] <annevk>
I think the problem was that Murata added TRackerIssue but the i18n WG did not know about it

[03:08:00.0000] <zcorpan>
heh, the spec uses application/vnd.google-earth.kml+xml as an example use case for registerContentHandler but then bans XML types

[03:11:00.0000] <annevk>
whoa

[03:11:01.0000] <annevk>
blog.whatwg.org is slow

[03:13:00.0000] <annevk>
http://blog.whatwg.org/weekly-undomanager

[03:14:00.0000] <annevk>
I'm away the next 6 or 7 days so if someone else wants to write it next week go ahead

[03:22:00.0000] <zcorpan>
http://www.w3.org/2010/html-xml/snapshot/report.html seems to say "parsing HTML with an XML parser is problematic and parsing XML with an HTML parser is problematic. the best solution is to parse XML with an XML parser and HTML with an HTML parser."

[03:24:00.0000] <Ms2ger>
Tell the TAG

[03:25:00.0000] <annevk>
zcorpan, pretty much

[03:26:00.0000] <jgraham>
Hey, it's taken 10 ears to get past the "parsing HTML with an XML parser is problematic so everyone should produce XML always" stage

[03:26:01.0000] <jgraham>
*years

[03:26:02.0000] <jgraham>
The number of ears sacrified is unknown

[03:29:00.0000] <zcorpan>
2.5 How can XML be made more forgiving of errors? is interesting

[03:29:01.0000] <zcorpan>
wonder if xml core will agree to specify lenient xml parsing

[03:30:00.0000] <annevk>
I put that in

[03:30:01.0000] <annevk>
sort of

[03:30:02.0000] <zcorpan>
or if somebody else will successfully do it over xml core's objections

[03:30:03.0000] <annevk>
I would finish my specification if one browser vendor would seriously consider implementing it

[05:04:00.0000] <Kellen`>
is the author of validator.nu here by any chance?

[05:06:00.0000] <Ms2ger>
He's on holiday until early next week, IIRC

[05:07:00.0000] <Kellen`>
okay thanks! is he usually in this channel?

[05:07:01.0000] <jgraham>
yep

[05:07:02.0000] <Kellen`>
awesome, I'll check back then

[05:10:00.0000] <zcorpan>
MikeSmith also fixes stuff in v.nu

[09:11:00.0000] <scor>
foolip: is the live microdata service down? http://foolip.org/microdatajs/live/

[09:12:00.0000] <foolip>
scor, no, looks fine to me

[09:12:01.0000] <foolip>
is it b0rked in some way?

[09:12:02.0000] <foolip>
I pushed some changes a few days ago

[09:12:03.0000] <scor>
"wating for foolip.org"

[09:12:04.0000] <scor>
same goes for http://foolip.org/

[09:13:00.0000] <foolip>
both are fine here

[09:13:01.0000] <foolip>
sounds like a network problem

[09:13:02.0000] <scor>
could it be my network then?

[09:13:03.0000] <scor>
it used to work from here in the past. let me check

[09:13:04.0000] <foolip>
perhaps I was blacklisted :)

[09:15:00.0000] <scor>
my network at work is very picky. sorry to have bugged you foolip, I'll check with my sysadmins…

[09:16:00.0000] <karlcow>
cute http://gsgwebs.com/help/browser_compatibility.html

[09:16:01.0000] <karlcow>
This page lists the discrepancies in HTML tag support between Netscape Navigator and Internet Explorer.

[09:18:00.0000] <karlcow>
had forgotten about dynsrc and lowsrc on images http://msdn.microsoft.com/en-us/library/ms533742(v=vs.85).aspx

[09:35:00.0000] <dglazkov>
good morning, Whatwg!

[09:36:00.0000] <Ms2ger>
Morning

[09:36:01.0000] <dglazkov>
wait what? are you visiting? :)

[09:50:00.0000] <dglazkov>
wiki.whatwg.org is down

[09:51:00.0000] <dglazkov>
Hixie, AryehGregor: ^^^

[10:09:00.0000] <karlcow>
http://www.mnot.net/blog/2011/08/24/distributed_hungarian_notation_doesnt_work

[11:00:00.0000] <shepazu>
in case anyone here uses Thunderbird, and finds themselves needing to share a link to a specific email, I've made an add-on http://schepers.cc/archived-link

[11:02:00.0000] <Ms2ger>
I guess that's easier than ctrl-u and copy it from the header

[11:52:00.0000] <jarek>
Hi!

[11:52:01.0000] <jarek>
does any browser support or is planning to support native rendering of menu element?

[11:52:02.0000] <jarek>
http://www.w3.org/TR/html5/interactive-elements.html#the-menu-element

[11:52:03.0000] <smaug____>
jarek: latest FF builds have support for context menu

[11:53:00.0000] <jarek>
<smaug____> yeah, I have just stumbled upon this news

[11:54:00.0000] <smaug____>
jarek: note, the child elements should be <menuitem> elements, not <command> elements

[11:54:01.0000] <jarek>
<smaug____> but I wonder whether all menus are going to be rendered this way

[11:55:00.0000] <jarek>
is it possible to trigger contextmenu with left mouse button?

[11:55:01.0000] <smaug____>
depends on OS, I guess

[12:02:00.0000] <dglazkov>
annevk: Dimitri “good morning” Glazkov :D

[12:05:00.0000] <smaug____>
jarek_: in which way should the menus be rendered?

[12:08:00.0000] <jarek_>
<smaug____> I mean something like this: http://twitpic.com/68ztwa/full

[12:09:00.0000] <smaug____>
jarek_: isn't that how it is rendered now?

[12:09:01.0000] <jarek_>
<smaug____> I need HTML5 menus that would look exactly as those from native apps

[12:09:02.0000] <smaug____>
(I don't have a mac)

[12:09:03.0000] <jarek_>
<smaug____> I have just checked menu and contextmenu element on Chrome, both are completely unstyled by default

[12:10:00.0000] <smaug____>
contextmenu element?

[12:10:01.0000] <smaug____>
and I have no idea what Chrome does

[12:11:00.0000] <smaug____>
I thought it doesn't support <menu type="context"> yet

[12:11:01.0000] <jarek_>
oopps, it should be <menu type="context">

[12:11:02.0000] <dglazkov>
jarek_: no implementation in WebKit as far as I know, so probably not in Chrome


2011-08-25
[20:29:00.0000] <heycam>
I'm wondering if there's any reason that the Function interface defined in the spec, which takes (any... args), is used for event handler attributes rather than something that takes say (Event event)

[20:29:01.0000] <heycam>
are there cases where those functions will be invoked with other arguments?

[21:22:00.0000] <Hixie>
heycam: yes, e.g. onerror's arguments array has length 3

[21:23:00.0000] <Hixie>
heycam: also you can pass any function to an event handler attribute, not just one with a single argument

[21:32:00.0000] <heycam>
Hixie, ok. maybe onerror should be a different type? maybe it's not worth it.

[21:33:00.0000] <Hixie>
what practical difference does it make?

[21:33:01.0000] <heycam>
none, just helpers readers of the spec know what type of values are going to be passed to the function

[21:34:00.0000] <Hixie>
*shrug*

[21:34:01.0000] <heycam>
you do have some other callback interface function things in the spec

[21:34:02.0000] <Hixie>
true

[21:34:03.0000] <heycam>
anyway, not important

[01:41:00.0000] <asmodai>
Anyone got insight on this (acquaintance of mine asking): http://stackoverflow.com/questions/7183196/negotiating-a-websocket-in-c-chrome-connecting-to-unity-c-tcp-client

[01:51:00.0000] <zcorpan>
asmodai: doesn't chrome implement the new spec these days?

[01:52:00.0000] <zcorpan>
guess it doesn't seem so from the dump

[01:53:00.0000] <zcorpan>
/me wonders why origin is null

[01:54:00.0000] <zcorpan>
asmodai: the request doesn't include a subprotocol but the response does

[01:54:01.0000] <zcorpan>
asmodai: i think that should be fine per spec but maybe chrome has a bug there?

[01:57:00.0000] <zcorpan>
asmodai: i'd recommend testing opera (enable in opera:config) and checking error console

[01:57:01.0000] <asmodai>
zcorpan: Not sure -- I also tossed it to a friend who is a chrome dev nowadays

[01:57:02.0000] <asmodai>
zcorpan: Opera 11.50 ought to be ok?

[01:57:03.0000] <zcorpan>
yeah 11.00 and up

[01:58:00.0000] <zcorpan>
why his dump has origin:null i don't know. is he testing from file:/// or something?

[02:00:00.0000] <asmodai>
"At the moment yes testing from file:/// is that likely to cause issues?"

[02:00:01.0000] <zcorpan>
/me hasn't thought about the security model with opening a websocket from file:///

[02:00:02.0000] <zcorpan>
it could be that chrome doesn't allow opening a connection for null origins

[02:03:00.0000] <zcorpan>
bbiab

[02:06:00.0000] <asmodai>
zcorpan: yea

[04:20:00.0000] <zcorpan>
hmm, how do i reply to the gray messages between the post and the answers in http://stackoverflow.com/questions/7183196/negotiating-a-websocket-in-c-chrome-connecting-to-unity-c-tcp-client ?

[04:21:00.0000] <Ms2ger>
"add comment"

[04:21:01.0000] <Ms2ger>
(You need to be logged in to see that)

[04:22:00.0000] <zcorpan>
i'm logged in. i only see that on the answers

[04:25:00.0000] <zcorpan>
*shrug*

[04:30:00.0000] <Philip`>
Do you need some minimum number of Stack Overflow magic beans to be allowed to do that?

[04:31:00.0000] <Philip`>
http://stackoverflow.com/privileges/comment

[04:34:00.0000] <zcorpan>
Reputation Required

[04:34:01.0000] <zcorpan>
50

[04:41:00.0000] <Ms2ger>
"If the character indicated by position (the first character)"

[04:56:00.0000] <asmodai>
zcorpan: ta

[04:56:01.0000] <asmodai>
zcorpan: upvoted ;)

[05:57:00.0000] <zcorpan>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13891 - who's Annie?

[05:58:00.0000] <Ms2ger>
A woman?! ;)

[06:01:00.0000] <zcorpan>
is she a mozilla guy? er, girl?

[08:03:00.0000] <benjoffe>
was trying to look for good examples of sites that work well in both rtl and ltr, hard to find them (example include google, wikipedia, facebook, reddit (barely), bbc, cnn, un)

[08:03:01.0000] <benjoffe>
anyone know of any list of good examples?

[08:05:00.0000] <Rik`_>
benjoffe: mozilla.com ?

[08:05:01.0000] <Rik`>
or at least, I hope so :)

[08:07:00.0000] <benjoffe>
Rik`: seems their front page works in rtl, but any of the pages with much content seem to revert to english, eg http://www.mozilla.org/ar/firefox/features/

[08:08:00.0000] <Rik`>
benjoffe: well that's because the contributors have not translated it yet

[08:09:00.0000] <benjoffe>
hmm the addons site is okay in arabic

[08:18:00.0000] <jgraham>
TabAtkins: Any idea how up to date http://wiki.csswg.org/test/css2.1/format is? In particular is there a CSS3 version of that page?

[08:19:00.0000] <Ms2ger>
jgraham, up to date, AUIU

[08:19:01.0000] <Ms2ger>
AIUI, even

[08:21:00.0000] <jgraham>
Ms2ger: It doesn't mention the fact the CSS3 tests must be reftests afaict

[08:21:01.0000] <Ms2ger>
Mm, right

[08:21:02.0000] <Ms2ger>
And self-describing reftests, unfortunately

[08:21:03.0000] <jgraham>
Is there some general page I can point people at for "everyhting one needs to know to write a CSS test that will be accepted"

[08:22:00.0000] <astearns>
but it does have the current reference link info for reftests

[08:23:00.0000] <jgraham>
/me notes that the amount of metadata required for CSS tests makes them way more budensome to submit than HTMLWG tests

[08:23:01.0000] <astearns>
how about the first link on that page? http://www.w3.org/Style/CSS/Test/guidelines.html

[08:25:00.0000] <jgraham>
astearns: That link is also wrong

[08:25:01.0000] <astearns>
in what way?

[08:26:00.0000] <jgraham>
Insofar as it doesn't say "Tests accepted by CSSWG will meet the following criteria: a) They will be self describing reftests b) etc."

[08:26:01.0000] <jgraham>
It's basically a document about how to write good visual tests

[08:27:00.0000] <Ms2ger>
a) They will be self describing reftests or b) they will be written by MS?

[08:27:01.0000] <jgraham>
(excuse the oxymoron)

[08:30:00.0000] <astearns>
hmm - there's a page describing how to write a reftest, but I don't see anything that says they're required

[08:31:00.0000] <astearns>
I do believe they are meant to be required. I'll send a note to css-testsuite confirming that

[09:19:00.0000] <karlcow>
mwahaha "The WHATWG does point to a ray of hope for extended namespaces in HTML5 that has been largely blocked by W3C" — http://www.readwriteweb.com/hack/2011/08/whos-validating-the-validators.php

[09:22:00.0000] <zcorpan>
"Fourteen of Alexa's top 25 sites list HTML5 as their doctypes, he noted."

[09:22:01.0000] <zcorpan>
not bad

[09:23:00.0000] <jgraham>
So I was just thinking http://joshduck.com/periodic-table.html was quite cool. Then I noticed the W3 Schools links

[09:24:00.0000] <kennyluck>
sigh

[09:26:00.0000] <zcorpan>
/me pings him on twitter

[09:32:00.0000] <zcorpan>
karlcow: that article seems utterly uninformed

[09:33:00.0000] <Ms2ger>
"a professional technologist who personally contributed to the RSS specification"

[09:33:01.0000] <Ms2ger>
Now that's something I wouldn't admit to

[09:34:00.0000] <miketaylr>
"Today, there is no punishment for non-compliance" maybe time for a Punishment Community Group?

[09:37:00.0000] <karlcow>
zcorpan: yup definitely

[09:37:01.0000] <karlcow>
if tristan louis is on the channel :)

[09:38:00.0000] <jgraham>
Ms2ger: The only question is "which one"?

[09:38:01.0000] <dglazkov>
good morning, Whatwg!

[09:38:02.0000] <Ms2ger>
Mm?

[09:38:03.0000] <jgraham>
10:36 < Ms2ger> "a professional technologist who personally contributed to the RSS specification"

[09:39:00.0000] <jgraham>
Although put like that it sounds like the start of a play

[09:39:01.0000] <jgraham>
Ms2ger - a professional technologist who personally contributed to the RSS specification

[09:39:02.0000] <jgraham>
Hixie - a Capulet

[09:40:00.0000] <Ms2ger>
jgraham - A doctor?

[09:42:00.0000] <jgraham>
Although having dglazkov here I feel it is more like Beckett than Shakespeare

[09:43:00.0000] <jgraham>
Or at the very least Tom Stoppard

[09:46:00.0000] <Ms2ger>
/me curses tests

[10:04:00.0000] <jgraham>
The colour scheme on the blog really is awful

[10:04:01.0000] <jgraham>
Isn't anyone with design sense interested in HTML?

[10:04:02.0000] <jgraham>
Failing that could we just delete the light green bar thing?

[10:07:00.0000] <nimbu>
are we talking about whatwg blog?

[10:08:00.0000] <jgraham>
Yes

[10:12:00.0000] <nimbu>
i can take a stab at it.

[10:12:01.0000] <nimbu>
that ugly neon green line needs to go

[10:12:02.0000] <jgraham>
Yes

[10:12:03.0000] <nimbu>
how can I work on it?

[10:14:00.0000] <nimbu>
i can ask anne when I meet him next week.

[10:15:00.0000] <jgraham>
You just need an account

[10:16:00.0000] <nimbu>
i also want access to the theme I suppose :P

[10:16:01.0000] <jgraham>
Yeah, dunno what level of permissions you need for access to that

[10:17:00.0000] <jgraham>
I can change it

[10:17:01.0000] <jgraham>
Do you have an account at all?

[10:17:02.0000] <dglazkov>
can you make the font more pretty? I really like comic sans. It's friendly.

[10:18:00.0000] <nimbu>
jgraham: def not

[10:18:01.0000] <nimbu>
/me glowers at dglazkov

[10:18:02.0000] <dglazkov>
:D

[10:18:03.0000] <Ms2ger>
I should ban you for that :)

[10:18:04.0000] <dglazkov>
and we need more pictures. Smiling people.

[10:18:05.0000] <nimbu>
i will just make sure you never get access to editing the theme dglazkov

[10:19:00.0000] <Philip`>
Make sure it uses @font-face so the page loads twice as slowly and looks ugly because the font wasn't tested on my platform

[10:19:01.0000] <nimbu>
smiling in face of apocalypse

[10:19:02.0000] <dglazkov>
nimbu: the best way to test for design experience is to ask questions like this and look for eyelid twitches

[10:20:00.0000] <Ms2ger>
Linux Libertine?

[10:20:01.0000] <nimbu>
dglazkov: it would be more than that for me

[10:20:02.0000] <jgraham>
nimbu: I can set you up with an account if you like

[10:20:03.0000] <jgraham>
Alternatively give me some CSS and I can paste it in

[10:21:00.0000] <nimbu>
jgraham: o that would work too. I will tweak in a bit.

[10:21:01.0000] <nimbu>
and link it up here.

[10:22:00.0000] <jgraham>
Perfect

[11:10:00.0000] <timeless>
heycam|away: i'l have more comments, sorry :)

[11:10:01.0000] <timeless>
Word isn't done :)

[11:12:00.0000] <Ms2ger>
Done crashing?

[11:19:00.0000] <nimbu>
jgraham: https://raw.github.com/gist/1171203/3d1dc9d34119ab94092981220252e004dde5895f/style.css

[11:19:01.0000] <nimbu>
(i hope you can test by just linking to that)

[11:25:00.0000] <Ms2ger>
pre { font-family:Verdana, sans-serif }

[11:28:00.0000] <Hixie>
ok seriously

[11:28:01.0000] <Hixie>
how hard can it be to understand that first we figure out what problem we're trying to solve

[11:28:02.0000] <nimbu>
Ms2ger: i just cleaned it up a bit.

[11:28:03.0000] <Hixie>
and THEN we look at possible solutions?

[11:28:04.0000] <nimbu>
Ms2ger: i did not touch the fonts :P

[11:29:00.0000] <Ms2ger>
Hixie, I think we need a "proposal" element for that

[11:30:00.0000] <MacTed>
but I have a hammer!  you *must* be able to frame your problem as a nail...

[11:31:00.0000] <Ms2ger>
MacTed, no, I want browsers to give me a hammer

[12:01:00.0000] <timeless>
Ms2ger: hey, heycam's spec only crashed Word *once*! :)

[12:01:01.0000] <timeless>
oh, i met AryehGregor this morning

[12:02:00.0000] <Ms2ger>
timeless, so are you the mentioned Annie?! :)

[12:03:00.0000] <timeless>
no, but i kinda met (at least saw her)

[12:03:01.0000] <timeless>
i think i even tried to introduce myself

[12:04:00.0000] <Ms2ger>
So, who is she?

[12:04:01.0000] <timeless>
she worked on Google Docs

[12:05:00.0000] <timeless>
and thus is more or less representing their interests

[12:05:01.0000] <timeless>
in things, like e.g. context menus

[12:05:02.0000] <timeless>
or clipboard

[12:05:03.0000] <Ms2ger>
Oh, an Author

[12:05:04.0000] <timeless>
or anything else

[12:05:05.0000] <timeless>
hrm

[12:05:06.0000] <timeless>
sure

[12:05:07.0000] <timeless>
/me goes back to Word

[12:05:08.0000] <timeless>
/me doesn't think google docs has a grammar checker

[12:10:00.0000] <timeless>
ooh, Word is now spell checking the contributors list

[12:10:01.0000] <Ms2ger>
I bet it doesn't like me

[12:10:02.0000] <timeless>
actually, it's fine w/ you

[12:11:00.0000] <timeless>
it didn't mind Shiki but objected to Okasaka fwiw

[12:11:01.0000] <timeless>
(and had issues w/ both non latin words)

[12:13:00.0000] <timeless>
how do i search a mailing list? :)

[12:14:00.0000] <timeless>
/me switches to gmail

[12:14:01.0000] <smaug____>
whaaat, disabling Google instant doesn't disable previews anymore :/

[12:14:02.0000] <timeless>
which preview?

[12:14:03.0000] <timeless>
the one's on the right of a result?

[12:14:04.0000] <smaug____>
yeah

[12:14:05.0000] <smaug____>
the really annoying ones

[12:15:00.0000] <timeless>
/me shrugs

[12:15:01.0000] <timeless>
have you tried Bing?

[12:15:02.0000] <timeless>
if so, lemme know how it is ;-)

[12:15:03.0000] <timeless>
/me hasn't

[12:16:00.0000] <smaug____>
I just changed my default search engine to Bing

[12:16:01.0000] <smaug____>
let's see how well it works

[12:16:02.0000] <smaug____>
at least the UI is better :)

[12:17:00.0000] <timeless>
/me accidentally launches Nightly

[12:17:01.0000] <timeless>
/me gets a painting glitch and has to wait for a ui

[12:17:02.0000] <timeless>
my tabs from last time: FishIE tank

[12:18:00.0000] <timeless>
woohoo

[12:18:01.0000] <timeless>
my tablet got a new toy!

[12:18:02.0000] <timeless>
(and my n900 is getting 13fps w/ 20 fish)

[13:04:00.0000] <karlcow>
nimbi can the logo be bigger? Could it be shinier? And what about more photos? … (for those who know)

[13:04:01.0000] <karlcow>
s/nimbi/nimbu/

[13:04:02.0000] <karlcow>
smaug____: I'm using duckduckgo

[13:05:00.0000] <smaug____>
karlcow: does that use Google and Bing internally ?

[13:08:00.0000] <karlcow>
smaug____: result are given by Bing built with Yahoo but with an extra layer. http://duckduckgo.com/privacy.html

[15:15:00.0000] <zcorpan>
rfc2397 says it imports "urlchar" from rfc2396, but the string "urlchar" doesn't appear in that document

[15:16:00.0000] <zcorpan>
does it mean "uric"?

[15:35:00.0000] <zcorpan>
ok so with rfc3987's reserved and unreserved chars, minus # and &, we could invent a "base79" encoding for use instead of base64 in data urls

[15:36:00.0000] <zcorpan>
and that still uses ' which could be problematic since then it can't be used in single-quoted values, and there's no char to represent what base64 uses "=" for

[15:38:00.0000] <erlehmann>
zcorpan, why would you do that?

[15:38:01.0000] <zcorpan>
erlehmann: to make it more compactly encoded

[15:39:00.0000] <erlehmann>
(._.)

[15:39:01.0000] <zcorpan>
although to make it worthwhile we'd have to go beyond what rfc3987 allows i guess

[15:39:02.0000] <zcorpan>
64 vs 79 (or 78) is a wash

[15:40:00.0000] <erlehmann>
i am very much not interested in that, base64 is okay for the occasional inline icon.

[15:40:01.0000] <erlehmann>
(for me)

[15:40:02.0000] <zcorpan>
sure

[15:45:00.0000] <jamesr_>
are you really saving anything post-gzip?

[15:46:00.0000] <zcorpan>
unlikely :)

[15:47:00.0000] <zcorpan>
maybe if i go with a scheme similar to utf-8 and try encoding several bytes at a time... hmm

[15:49:00.0000] <zcorpan>
that, or i go with "naw it was a stupid idea" and go back to sleep

[15:50:00.0000] <smaug____>
/me doesn't understand what https://twitter.com/#!/slightlylate/status/106512911165435904 tries to say

[15:50:01.0000] <jamesr_>
http://code.google.com/p/webgl-loader/wiki/UtfEight

[15:51:00.0000] <roc>
http://joshduck.com/periodic-table.html

[15:52:00.0000] <roc>
jamesr_: "it just doesn't compare to native code, and the fancy things "real" compression algorithms do" is not really true now

[15:53:00.0000] <jamesr_>
depends on your browser

[15:53:01.0000] <jamesr_>
and what threading model you want

[15:54:00.0000] <roc>
browsers that support WebGL tend to have modern JS engines with typed arrays and specialization that works well for tight bit-banging loops

[15:54:01.0000] <roc>
however, using UTF-8 that way is a cool idea anyway

[15:59:00.0000] <TabAtkins>
smaug____: Alex has retracted his suggestion for this particular API.  In general, though, many people reach for an API when they should be doing an <input>.

[16:00:00.0000] <TabAtkins>
smaug____: Like DAP's Contacts API, or even Geolocation.

[16:00:01.0000] <smaug____>
doing <input> ?

[16:00:02.0000] <smaug____>
huh

[16:01:00.0000] <TabAtkins>
<input type=contacts>, for example.

[16:01:01.0000] <smaug____>
fortunately speech is not going to be bound specifically to input or other form elements

[16:02:00.0000] <smaug____>
<input type=contact> would make sense, if it would read data from "contacts"

[16:03:00.0000] <TabAtkins>
Yes, exactly.

[16:03:01.0000] <smaug____>
but it would be type=contact

[16:03:02.0000] <TabAtkins>
Which would be provided by your phone, or by telling your browser that Google manages your contacts, or whatever.

[16:03:03.0000] <smaug____>
not contacts

[16:03:04.0000] <timeless>
that's being done by dap

[16:03:05.0000] <timeless>
/me hopes

[16:03:06.0000] <TabAtkins>
timeless: No, DAP is doing some pure-API thing.

[16:03:07.0000] <TabAtkins>
Maybe they're switching over?  I dunno.

[16:04:00.0000] <smaug____>
but there are other use cases for API

[16:04:01.0000] <timeless>
well, ideally whatever work is done by dap would be leveragable//

[16:04:02.0000] <smaug____>
the type="contact" would be just an autocomplete like thing

[16:07:00.0000] <smaug____>
binding geolocation to input element would be  rather silly

[16:07:01.0000] <timeless>
well

[16:07:02.0000] <timeless>
<input type=location> or <input type=address>

[16:07:03.0000] <timeless>
is perfectly reasonable

[16:07:04.0000] <TabAtkins>
Exactly.

[16:07:05.0000] <timeless>
but it's much better if your UA can bring up a map

[16:08:00.0000] <timeless>
or let you retrieve an address from your contacts

[16:08:01.0000] <timeless>
the Mozilla GeoLocation impl and its owners are incredibly annoying here

[16:08:02.0000] <TabAtkins>
These are, of course, perfectly fine options for an <input>'s UI.

[16:08:03.0000] <timeless>
as are the other vendors

[16:08:04.0000] <timeless>
"We don't want to standardize on an api to do this"

[16:08:05.0000] <timeless>
but it's stupid and wrong

[16:08:06.0000] <smaug____>
if <input type=location> uses geolocation in the background, that is implementation detail

[16:09:00.0000] <timeless>
if the design of geolocation wasn't "I need to spy on the user's location"

[16:09:01.0000] <timeless>
but instead "I'd like a location from the user, any will do"

[16:09:02.0000] <timeless>
And UAs happened to have "convert CurrentLocation" as an option in addition to

[16:09:03.0000] <timeless>
"enter from contact"

[16:09:04.0000] <timeless>
and "select from map"

[16:09:05.0000] <timeless>
but no,

[16:09:06.0000] <timeless>
/me sighs

[16:09:07.0000] <timeless>
/me curses Nokia

[16:10:00.0000] <timeless>
(it's all their fault that I wasn't allowed at that table when the standard was written)

[16:10:01.0000] <TabAtkins>
You've got it exactly, timeless.  Getting a location/address is a useful thing to have, and to submit in a form.  Where you get that (current loc, manually entered, chosen from a map) is a UI issue.

[16:10:02.0000] <timeless>
thank you!

[16:10:03.0000] <Hixie>
no reason the geo api can't return arbitrary locations too

[16:10:04.0000] <timeless>
but note that sadly vendors need to be encouraged to do right by users

[16:10:05.0000] <timeless>
Hixie: true

[16:10:06.0000] <timeless>
except that vendors suck here

[16:10:07.0000] <timeless>
and are lazy

[16:10:08.0000] <timeless>
and don't want to ship that by default

[16:10:09.0000] <timeless>
and aren't competing to be good to their users

[16:10:10.0000] <timeless>
so we're stuck w/ privacy invasion only

[16:11:00.0000] <TabAtkins>
Hixie: Problem is that, with the script-driven API, you're stuck with either smuggling some options into the permissions dialog, or popping up an intrusive dialog whenever the script makes a request.

[16:11:01.0000] <timeless>
and the vendors say "oh, you can get a [crappy] addon which [hand-waving] will let you do that"

[16:11:02.0000] <TabAtkins>
An <input> lets the dialog be user-driven.

[16:11:03.0000] <timeless>
and yes, i've tried the Opera thing

[16:11:04.0000] <timeless>
it's crappy

[16:11:05.0000] <timeless>
it didn't really work

[16:11:06.0000] <timeless>
and i did try to send feedback

[16:11:07.0000] <timeless>
and i've looked at the mozilla impl

[16:11:08.0000] <timeless>
(repeatedly!)

[16:11:09.0000] <Hixie>
TabAtkins: doesn't have to be intrusive

[16:11:10.0000] <timeless>
and it is *not* friendly

[16:12:00.0000] <smaug____>
<input type="location"> should definitely ask permission if it was filled automatically

[16:12:01.0000] <Hixie>
TabAtkins: just drop down an infobar that says "what location do you wish to report to the page? ((current location)) (pick from map) (old location v) (none)"

[16:12:02.0000] <timeless>
smaug____: it shouldn't fill in automatically

[16:12:03.0000] <timeless>
you just get a thing with a bunch of buttons

[16:12:04.0000] <timeless>
a map, an addressbook, and a location icon

[16:12:05.0000] <smaug____>
timeless: well, then it is just an implementation detail

[16:12:06.0000] <timeless>
if you click the location icon, it fills it in

[16:13:00.0000] <timeless>
smaug____: not exactly

[16:13:01.0000] <timeless>
unimplemented details don't help users

[16:13:02.0000] <timeless>
the precise ui is an implementation detail

[16:13:03.0000] <timeless>
but requirements to let the user select an arbitrary location

[16:13:04.0000] <timeless>
should be *requirements*

[16:13:05.0000] <timeless>
instead of

[16:13:06.0000] <timeless>
[hand waving]

[16:14:00.0000] <timeless>
done right, the UAs would have competed to get decent UE for this stuff

[16:14:01.0000] <timeless>
and we'd probably have had a *better* system that didn't require permission dialogs at all

[16:14:02.0000] <timeless>
since the users could just select a point from a map

[16:14:03.0000] <timeless>
e.g. Paris

[16:14:04.0000] <smaug____>
so, I haven't seen any reasons to bind geolocation to input, but have better UIs in the browsers

[16:14:05.0000] <zewt>
specs should specify the apis, they shouldn't try to require UI features for users

[16:14:06.0000] <timeless>
zewt: do

[16:15:00.0000] <timeless>
so...

[16:15:01.0000] <TabAtkins>
smaug____: Location is manifestly a potentially useful thing to submit in forms.

[16:15:02.0000] <timeless>
an API which only allows for privacy invasion

[16:15:03.0000] <timeless>
is not a good thing

[16:15:04.0000] <timeless>
and a claim that "oh, in theory, someone could lie"

[16:15:05.0000] <timeless>
doesn't help if no one implements it

[16:15:06.0000] <timeless>
and it certainly doesn't help if the api suggests "or you can just abort"

[16:15:07.0000] <TabAtkins>
You could use the scripted API and fill a hidden input, but why do that when you can instead just make an <input> type and script at that?

[16:15:08.0000] <zewt>
sure, and specs should note "there are privacy concerns here and UAs should consider them", and perhaps advise on how, but not try to specify how

[16:15:09.0000] <timeless>
which results in sites discriminating and doing other stupid things

[16:16:00.0000] <timeless>
zewt: right, i'm not saying specify how

[16:16:01.0000] <TabAtkins>
In other words, when gathering information from the user, the impetus should be to justify why it *shouldn't* be a new input type.

[16:16:02.0000] <timeless>
whether it's an input field, or a map, or contacts is an implementation detail

[16:16:03.0000] <timeless>
but requiring that the user be able to enter another value isn't a bad thing

[16:16:04.0000] <timeless>
if a ui doesn't have a keyboard, it can offer the users a couple of random values

[16:16:05.0000] <timeless>
that's still valid

[16:16:06.0000] <timeless>
and it would still help users

[16:17:00.0000] <timeless>
<input type=location onchange=doSomething()>

[16:17:01.0000] <smaug____>
still, only about UI. Nothing requiring geolocation or similar.

[16:17:02.0000] <timeless>
/me fails to parse that

[16:18:00.0000] <smaug____>
browser could know the location in other ways than using geolocation

[16:19:00.0000] <timeless>
right

[16:19:01.0000] <zewt>
i don't generally think specs trying to require user-facing, non-api-facing features seems very useful; if i was an implementor i'd deem that out of scope for the spec and ignore it (except informationally, of course--i'd treat it as a note)

[16:19:02.0000] <timeless>
but unfortunate the geolocation api was designed to be a geolocation api

[16:19:03.0000] <timeless>
instead of a location api

[16:19:04.0000] <timeless>
which is a bug

[16:19:05.0000] <timeless>
(that was standardized!)

[16:19:06.0000] <smaug____>
so saying that type="location" has something to do with geolocation is just an implementation detail

[16:19:07.0000] <zewt>
if it's not useful enough to implement on my (implementor's) own, it'd be surprising if demanding it in the spec would change that

[16:20:00.0000] <timeless>
smaug____: which is fine with us

[16:20:01.0000] <smaug____>
timeless: ah, you just want different API

[16:20:02.0000] <smaug____>
you have different use cases

[16:20:03.0000] <timeless>
smaug____: i want a different outlook

[16:20:04.0000] <timeless>
my use cases are privacy first

[16:20:05.0000] <timeless>
and being able to plan for where i'll be

[16:20:06.0000] <timeless>
or help someone remotely

[16:20:07.0000] <timeless>
or ...

[16:20:08.0000] <timeless>
but still be able to use geolocation services

[16:20:09.0000] <timeless>
nothing prevented geolocation from working well using <input type=location>

[16:20:10.0000] <smaug____>
nothing prevents that now

[16:20:11.0000] <timeless>
and nothing prevents users from being able to use it for lots of other things

[16:21:00.0000] <timeless>
sure things do

[16:21:01.0000] <TabAtkins>
Even ignoring privacy, the ability to get a location from the user (not their current, just a chosen loc) is really useful for, say, timezone calculations.

[16:21:02.0000] <timeless>
none of the browsers did anything friendly for users

[16:21:03.0000] <timeless>
and none of them will

[16:21:04.0000] <TabAtkins>
"I'm going to be HERE, what's the time?"

[16:21:05.0000] <zewt>
TabAtkins: i hate timezones to death

[16:21:06.0000] <timeless>
i'll give you 100 bucks if one does by march

[16:21:07.0000] <smaug____>
why not?

[16:21:08.0000] <zewt>
just throwing that out there

[16:21:09.0000] <timeless>
(and 50 bucks if one does by july)

[16:21:10.0000] <smaug____>
timeless: just write a spec for type="location"

[16:21:11.0000] <TabAtkins>
zewt: Preach it.  If Swatch hadn't screwed up the zero point on their beats I'd be using them.

[16:22:00.0000] <timeless>
heh

[16:22:01.0000] <timeless>
smaug____: would you get it implemented in gecko if i wrote the spec?

[16:22:02.0000] <timeless>
(or webkit, i'm not picky)

[16:23:00.0000] <smaug____>
possibly

[16:23:01.0000] <timeless>
(and would you help get the geolocation api deprecated?)

[16:23:02.0000] <smaug____>
depends on the quality of the spec ;)

[16:23:03.0000] <smaug____>
why should we deprecate geolocation API?

[16:23:04.0000] <smaug____>
it has different use cases

[16:23:05.0000] <timeless>
not really

[16:24:00.0000] <zewt>
yeah deprecate a widely deployed, heavily used API, great idea heh

[16:24:01.0000] <smaug____>
using geolocation web app can query the location, type="location" requires user to actively give the location to web app

[16:24:02.0000] <timeless>
it doesn't

[16:24:03.0000] <smaug____>
huh

[16:24:04.0000] <timeless>
the input could easily let the user say `update location as i move`

[16:25:00.0000] <timeless>
see my example above

[16:25:01.0000] <smaug____>
huh

[16:25:02.0000] <timeless>
<input type=location onchange=doSomething()>

[16:25:03.0000] <smaug____>
terrible

[16:25:04.0000] <timeless>
what's terrible about that?

[16:25:05.0000] <smaug____>
so each web app should depend on the UI <input type="location"> happens to have

[16:25:06.0000] <Philip`>
TabAtkins: I'd expect timezone location-selection would need a specialist UI, because you might be selecting a point near a timezone boundary, in which case you need to be made aware that you're near the boundary and that you need to be very precise in your location

[16:26:00.0000] <Philip`>
whereas if you're somewhere near the middle of a timezone then you don't need to bother with that precision

[16:26:01.0000] <TabAtkins>
Philip`: Sure, it was just an example off the top of my head.

[16:26:02.0000] <timeless>
timezones are disasters :)

[16:26:03.0000] <timeless>
did you guys install MS's recent time zone update?

[16:26:04.0000] <Philip`>
so the UI ought to be giving you feedback instead of being a fully general one

[16:26:05.0000] <TabAtkins>
timeless: Was it an auto-update?

[16:26:06.0000] <timeless>
http://support.microsoft.com/kb/2570791

[16:26:07.0000] <timeless>
it was pushed by Windows Update, yeah

[16:26:08.0000] <zewt>
well, there should be a way to query the timezone from the system, very rarely should it be needed to nag the user about it

[16:27:00.0000] <zewt>
but ... timezones are tricky, heh

[16:27:01.0000] <timeless>
zewt: err

[16:27:02.0000] <zewt>
stupid 1/2 (and even a couple 1/4th) timezones D:

[16:27:03.0000] <timeless>
privacy violation

[16:27:04.0000] <zewt>
zzz

[16:27:05.0000] <zewt>
ask people who want to be asked

[16:27:06.0000] <timeless>
it should be possible to ask the system to show a time in the user's timezone

[16:27:07.0000] <timeless>
or in some other time zone

[16:27:08.0000] <timeless>
but it really isn't ok to automatically retrieve the user's time zone

[16:28:00.0000] <timeless>
(plus, it's not necessarily accurate)

[16:28:01.0000] <zewt>
that's at the level where maybe browsers should have a "paranoid" mode where they ask about it, but 98% of users don't care and don't need to

[16:28:02.0000] <timeless>
<input type=timezone>

[16:28:03.0000] <TabAtkins>
I am all for adding a dozen new inputs.

[16:28:04.0000] <Philip`>
zewt: Can't you determine the user's timezone already from Date().toString() ?

[16:28:05.0000] <TabAtkins>
inputs are great.

[16:28:06.0000] <timeless>
the value can be colored in gray with the current value with a way for the user to accept that value

[16:29:00.0000] <zewt>
is the timezone in that standardized?

[16:29:01.0000] <timeless>
Philip`: if the JS system is good enough and the os is good enough

[16:29:02.0000] <timeless>
kinda

[16:29:03.0000] <timeless>
but one or the other is typically not good enough for 100% :)

[16:29:04.0000] <zewt>
i guess it must be

[16:29:05.0000] <timeless>
/me has read the edge cases there for gecko fwiw, they exist

[16:30:00.0000] <zewt>
heh i wonder if chrome's paranoia mode hides that

[16:30:01.0000] <TabAtkins>
I doubt it.

[16:30:02.0000] <zewt>
nope

[16:30:03.0000] <timeless>
note that you can't necessarily determine if a user will switch in and out of DST

[16:30:04.0000] <timeless>
and certainly not necessarily when

[16:30:05.0000] <timeless>
knowing someone is in +x doesn't tell you about transitions

[16:31:00.0000] <timeless>
and it also won't tell you if the computer they're using has correct transition info :)

[16:31:01.0000] <zewt>
the format is "GMT-0400 (Eastern Daylight Time)"--something I'm not sure is whether the string "Eastern Daylight Time" is a standardized string that can be mapped to a timezone db, or if it's ad hoc and unreliable

[16:31:02.0000] <zewt>
because GMT-0400 doesn't tell you the timezone, it only tells you the offset in the user's timezone currently

[16:32:00.0000] <zewt>
stupid Arizona

[16:33:00.0000] <timeless>
Thu Aug 25 2011 18:36:01 GMT-0400 (US Eastern Daylight Time)

[16:33:01.0000] <TabAtkins>
The current time is Na, in base-60 minutes since UTC midnight.

[16:33:02.0000] <timeless>
Thu Aug 25 2011 18:36:23 GMT-0400 (Eastern Daylight Time)

[16:33:03.0000] <zewt>
yeah, that limits the usefulness of that for anything nontrivial

[16:33:04.0000] <timeless>
both of those times are from Chrome

[16:33:05.0000] <timeless>
anyone want to guess what the difference is?

[16:33:06.0000] <zewt>
guessing regional

[16:33:07.0000] <zewt>
we don't say "US EDT" in the US

[16:33:08.0000] <timeless>
pick a region? :)

[16:34:00.0000] <zewt>
since we're the US, damn it!

[16:34:01.0000] <timeless>
so guess the region?

[16:34:02.0000] <zewt>
why? heh

[16:34:03.0000] <timeless>
it's Indiana fwiw

[16:34:04.0000] <timeless>
are you saying that isn't in the US? :)

[16:34:05.0000] <Hixie>
AryehGregor: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13424

[16:34:06.0000] <TabAtkins>
zewt: http://www.xanthir.com/time-manifesto.php

[16:35:00.0000] <timeless>
Thu Aug 25 18:38:00 2011

[16:35:01.0000] <timeless>
is IE9 fwiw :)

[16:35:02.0000] <Hixie>
AryehGregor: if you're ok with moving the definitions to your spec, please mention in the spec which sections i should remove from the html spec

[16:36:00.0000] <zewt>
we can't even avoid streets having different names across town boundaries, short of a global dictatorship we'll never see a rational timezone system, heh

[16:37:00.0000] <timeless>
so, afaict IE has decided to

[16:37:01.0000] <timeless>
btw, there are more than 4 timezones for the US :)

[16:38:00.0000] <timeless>
ignoring Indiana

[16:38:01.0000] <zewt>
well, it's still giving local time, and (if the user's clock isn't wrong) you can figure out the GMT offset from that

[16:38:02.0000] <Philip`>
Does it do the same for 'new Date().toLocaleString()'?

[16:39:00.0000] <timeless>
Philip`: that gets me a bing search

[16:39:01.0000] <TabAtkins>
timeless: Continental US!

[16:39:02.0000] <timeless>
TabAtkins: aww

[16:39:03.0000] <timeless>
Alaska excluded?

[16:39:04.0000] <timeless>
(it's on the contintent!)

[16:40:00.0000] <timeless>
I think you're looking for `lower 48+dc` fwiw :)

[16:41:00.0000] <TabAtkins>
Is Alaska usually included when you say "continental US"?  I thought that term usually meant lower 48+dc.

[16:41:01.0000] <timeless>
http://en.wikipedia.org/wiki/Contiguous_United_States#Continental_United_States

[16:41:02.0000] <timeless>
Because Alaska is also on the North American continent, the term continental United States, if interpreted literally, should also include that state,[7] so the term is sometimes qualified with the explicit inclusion or exclusion of Alaska to resolve any ambiguity.[8][3][9][10][11]

[16:41:03.0000] <zewt>
if something says "shipping only to continental US" that usually excludes alaska

[16:41:04.0000] <timeless>
they usually explicitly note they exclude it though

[16:42:00.0000] <timeless>
note sure if that's 8, 3, 9, 10, or 11 :)

[16:42:01.0000] <timeless>
s/note/not/

[16:42:02.0000] <zewt>
heh

[16:42:03.0000] <timeless>
/me has no idea why 3 is after 8

[16:42:04.0000] <zewt>
i use stylish to hide footnotes on wikipedia

[16:42:05.0000] <Hixie>
anyone familiar with rniwa's undomanager stuff?

[16:42:06.0000] <zewt>
there are *way* too many of them on a lot of pages, to the point where it's distracting

[16:42:07.0000] <Hixie>
should i be removing window.onundo and window.onredo?

[16:42:08.0000] <Hixie>
i don't see it mentioned in the proposal

[16:42:09.0000] <timeless>
The term was in use prior to the admission of Alaska and Hawaii as states of the United States, and at that time usually excluded outlying territories of the U.S.[12][13] However, even before Alaska became a state, it was sometimes included within the "Continental US".[14]

[16:43:00.0000] <timeless>
anyway, `contiguous` is usually the word you want

[16:43:01.0000] <smaug____>
Hixie: I think he and ehsan and others were discussing  about undomanager yesterday or today

[16:43:02.0000] <timeless>
> The term lower 48 may or may not include the District of Columbia (which is not part of any of the 48 states). The National Geographic style guide recommends the use of contiguous or conterminous United States instead of lower 48 when the 48 states are meant, unless used in the context of Alaska.[17] Otherwise it is avoided as a misnomer, because all the major islands of Hawaii are farther south than the most southern point of the continenta

[16:43:03.0000] <Hixie>
yeah, they had a meeting up in the canadian lands

[16:43:04.0000] <smaug____>
I hope they'll report something

[16:43:05.0000] <zewt>
ugh nvidia installed two plugins into firefox without permission

[16:44:00.0000] <timeless>
they were meting today

[16:44:01.0000] <timeless>
s/met/meet/

[16:44:02.0000] <timeless>
zewt: you using ff9?

[16:44:03.0000] <timeless>
if so, it probably told you when you started..

[16:44:04.0000] <zewt>
6

[16:44:05.0000] <timeless>
change to 9

[16:44:06.0000] <timeless>
you'll get a dialog complaining about added addons/plugins

[16:44:07.0000] <zewt>
i'll pass on non-production builds of my primary browser, heh

[16:45:00.0000] <timeless>
/me is using 9.0a1 2011-08-25

[16:45:01.0000] <timeless>
or "Nightly is up to date" as Asa would want me to say if i visited the about screen

[16:45:02.0000] <timeless>
it's fine as long as you're happy w/ losing your irc logs every day

[16:46:00.0000] <zewt>
if firefox started deleting mirc logs i'd be concerned :)

[16:47:00.0000] <timeless>
you clearly haven't bought into B2G

[16:47:01.0000] <timeless>
/me is using webchat.freenode.net ..

[16:48:00.0000] <nimbu>
/me thanks whoever uploaded the styles to whatwg blog theme

[16:51:00.0000] <timeless>
can someone remind me how <cite> works?

[16:51:01.0000] <timeless>
http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#wai-aria

[16:51:02.0000] <timeless>
has a <cite> tag...

[16:51:03.0000] <karlcow>
cite is the reference of the Work

[16:52:00.0000] <Hixie>
timeless: <cite> means "title of work"

[16:52:01.0000] <timeless>
karlcow: nope

[16:52:02.0000] <Hixie>
timeless: insofar as how it works in browsers, it works the same as <i>, more or less

[16:52:03.0000] <timeless>
Hixie provided the correct answer

[16:52:04.0000] <karlcow>
<cite>Gone With The Wind</cite>

[16:52:05.0000] <timeless>
(of course, since he wrote the spec, that isn't surprising)

[16:52:06.0000] <karlcow>
huh?

[16:52:07.0000] <Hixie>
heh

[16:52:08.0000] <karlcow>
we both gave the same answer

[16:52:09.0000] <timeless>
no

[16:53:00.0000] <karlcow>
/me is scratching his head

[16:53:01.0000] <timeless>
/me is looking for a style guide name

[16:53:02.0000] <timeless>
but roughly a "reference" to a work as in http://honolulu.hawaii.edu/legacylib/mlahcc.html would include more than the work title

[16:54:00.0000] <timeless>
it tends to include an author and a date

[16:55:00.0000] <timeless>
> The cite element is obviously a key part of any citation in a bibliography, but it is only used to mark the title:  <p><cite>Universal Declaration of Human Rights</cite>, United Nations, December 1948. Adopted by General Assembly resolution 217 A (III).</p>

[16:55:01.0000] <karlcow>
yup

[16:55:02.0000] <timeless>
your definition would have meant the <p> to me

[16:55:03.0000] <timeless>
which is what i needed clarified

[16:55:04.0000] <karlcow>
ah ok

[16:56:00.0000] <karlcow>
/me add to html5 <ref></ref>

[16:56:01.0000] <timeless>
sorry for the terse `nope`

[16:56:02.0000] <timeless>
/me was blanking on the spelling of MLA

[16:56:03.0000] <timeless>
(amongst other problems)

[16:56:04.0000] <timeless>
/me wants dinner

[16:57:00.0000] <nimbu>
jgraham: in style.css line 63 (for blog.whatwg.org) there needs to be an overflow: hidden; after the display: block; decl. SRRY I did not notice that.

[16:57:01.0000] <zewt>
i wish browsers let me middle-click on form submit buttons to submit into another tab

[16:57:02.0000] <timeless>
zewt: what happens if the script does something magical instead? :)

[16:57:03.0000] <karlcow>
<ref><cite>title</cite>, <author title="it's here we start to enter into a maze of markup issues">authors list</author>, etc etc etc </ref>

[16:57:04.0000] <zewt>
the same thing as links that do the same, heh

[16:58:00.0000] <timeless>
zewt: e.g. google's instant search

[16:58:01.0000] <timeless>
which generally doesn't load a new page at all

[16:58:02.0000] <Philip`>
zewt: Ctrl+shift+click seems to open buttons in new tabs in Opera

[16:58:03.0000] <timeless>
for pages w/o access to databases or remote resources, you could clone the document tree

[16:59:00.0000] <timeless>
but that fails if there's a local database or some similar remote resource

[16:59:01.0000] <timeless>
/me curses Greyhound

[16:59:02.0000] <Philip`>
(Actually just shift+click)

[16:59:03.0000] <timeless>
their serve wouldn't let me have 2 searches open in distinct windows

[16:59:04.0000] <timeless>
it invalidated all but the most recently used one upon submission

[16:59:05.0000] <zewt>
heh


2011-08-26
[17:00:00.0000] <timeless>
oh, if you want to truly appreciate greyhound...

[17:00:01.0000] <zewt>
the worst is broken crsf stuff that breaks if you ever have more than one form open

[17:00:02.0000] <timeless>
load http://greyhound.ca/

[17:00:03.0000] <timeless>
enter `toronto` (select the first one) in the origin

[17:00:04.0000] <timeless>
enter `albany` (select the first one) in the destination

[17:00:05.0000] <timeless>
then search

[17:01:00.0000] <timeless>
now go back and try `toronto` (...) and `vancouver`

[17:01:01.0000] <timeless>
i tried to describe how broken greyhound.ca was to someone because i experienced the first behavior

[17:01:02.0000] <timeless>
but they tried the second set of inputs...

[17:02:00.0000] <timeless>
oh darn

[17:02:01.0000] <timeless>
it isn't behaving nicely for vancouver

[17:02:02.0000] <timeless>
hold

[17:02:03.0000] <timeless>
ok, s/vancouver/waterloo/g

[17:02:04.0000] <timeless>
go figure

[17:03:00.0000] <timeless>
clearly i haven't divined the right rule for when their site breaks

[17:03:01.0000] <timeless>
(for waterloo, be sure to pick ON [the second item])

[17:05:00.0000] <karlcow>
timeless: and the worse for me… is that there is no sane way to do the same thing by train

[17:05:01.0000] <timeless>
karlcow: i'm actually going from Toronto to Albany

[17:06:00.0000] <timeless>
bus is really the best path

[17:06:01.0000] <timeless>
i looked at train (sucky schedules)

[17:06:02.0000] <karlcow>
unfortunately

[17:06:03.0000] <timeless>
and flights (1000cur)

[17:06:04.0000] <karlcow>
yup

[17:06:05.0000] <karlcow>
trains are slow and expensive in Canada :/

[17:06:06.0000] <timeless>
i presume you got the message from your browser?

[17:07:00.0000] <karlcow>
from Montreal to NY the only reasonable way is to take a plane

[17:07:01.0000] <timeless>
(if you're using firefox, you got *my* message actually...)

[17:07:02.0000] <timeless>
(since i wrote it)

[17:07:03.0000] <timeless>
driving shouldn't be bad

[17:07:04.0000] <timeless>
/me doesn't have a license

[17:07:05.0000] <timeless>
my path (toronto/albany) would be pretty good by car

[17:07:06.0000] <timeless>
scenic, etc

[17:08:00.0000] <karlcow>
/me is missing train transportation in Japan and Europe :)

[17:11:00.0000] <timeless>
karlcow: can you drive from montreal to plattsburg?

[17:11:01.0000] <timeless>
68 Adirondack  Departs: 12:35 PM Wed Sep 07 2011 Plattsburgh, NY  (PLB) Arrives: 8:40 PM Wed Sep 07 2011 New York, NY - Penn Station  (NYP) Duration: 8 hr, 5 min

[17:12:00.0000] <timeless>
not terrible

[17:12:01.0000] <timeless>
oh, 62USD

[17:12:02.0000] <karlcow>
8h… :) awful

[17:13:00.0000] <karlcow>
507 km. should be taking max 1h30 to 2h.

[17:15:00.0000] <timeless>
google maps says it's a 5hour drive

[17:17:00.0000] <karlcow>
yup insane. 8h of train…

[17:17:01.0000] <timeless>
boston south station to washington d.c. union station is 8 hours/450miles or a 6 1/2hr train ride

[17:19:00.0000] <timeless>
wierd

[17:19:01.0000] <timeless>
if i start my gmaps trip in montreal, i get KM

[17:19:02.0000] <timeless>
if it's us-us i get Mi

[17:19:03.0000] <timeless>
(please excuse incorrect abbr's)

[17:20:00.0000] <timeless>
note that europe's train system doesn't work perfectly for non major locations

[17:20:01.0000] <timeless>
and you can often end up w/ 3 trains

[17:20:02.0000] <timeless>
i think you're looking at roughly trying to get north of switzerland to south of switzerland

[17:21:00.0000] <timeless>
/me tries to remember which euro train site didn't suck

[17:21:01.0000] <timeless>
http://maps.google.com/maps?q=Stuttgart,+Germany+to+Genova,+Italy&saddr=Stuttgart,+Germany&daddr=Genova,+Italy&hl=en&sll=46.649436,10.524902&sspn=10.181416,23.269043&geocode=FZJH6AIdYRaMACnTj63BNNuZRzHkz5F3HMHVeQ%3BFRaZpQIdZVKIAClpb64h5kPTEjG7Arq1xlP3aQ&vpsrc=0&doflg=ptk&t=h&z=7

[17:21:02.0000] <timeless>
for lack of something else to try

[17:22:00.0000] <timeless>
Stuttgart Hbf Genova Piazza Principe  25.08.11   05:18 dep 15:42 arr    10:24  4

[17:22:01.0000] <timeless>
10 hours, 4 changes

[17:23:00.0000] <timeless>
best is:

[17:23:01.0000] <timeless>
Stuttgart Hbf Genova Piazza Principe  25.08.11   07:58 dep 16:42 arr    8:44  2

[17:23:02.0000] <timeless>
~9 hours, 2 changes

[17:23:03.0000] <timeless>
so, no, europe isn't wonderful either

[17:23:04.0000] <timeless>
you're just used to going from cities which don't have mountains in the way

[17:23:05.0000] <timeless>
and which have heavy traffic

[17:24:00.0000] <timeless>
/me is happy at those two thrown darts

[17:25:00.0000] <timeless>
anyway, that trip is comprable for distance

[17:27:00.0000] <karlcow>
paris-lyon, paris-marseille, paris-london, paris-brussels :) like a charm

[09:15:00.0000] <dglazkov>
good morning, Whatwg!

[09:15:01.0000] <Ms2ger>
Morning

[09:17:00.0000] <nimbu>
:))

[09:18:00.0000] <jgraham>
Well shall we go?

[09:19:00.0000] <dglazkov>
yes! where are we going?

[11:01:00.0000] <AryehGregor>
Hixie, we discussed undo/redo events, and I think we concluded that we'll just have events along the lines of onbeforecommand and onaftercommand that will fire for things like undo/redo/delete/insertText/etc. whether it's the browser or the user triggering them.  So in principle that means we don't need special events for undo/redo.

[11:01:01.0000] <AryehGregor>
Hixie, I'll get back to you on bug 13424, the e-mails should be somewhere in my spec inbox.

[11:18:00.0000] <timeless>
karlcow: that's because paris is well connected and you don't have any mountains to deal with

[11:19:00.0000] <karlcow>
like in Japan? ;)

[11:19:01.0000] <timeless>
/me doesn't know enough about japan

[11:19:02.0000] <timeless>
/me isn't sure if Japan just goes around the mountains

[11:19:03.0000] <karlcow>
mountains land

[11:20:00.0000] <timeless>
http://maps.google.com/maps?saddr=Nagano,+Nagano+Prefecture,+Japan&daddr=Okayama,+Okayama+Prefecture,+Japan&hl=en&sll=34.939985,136.252441&sspn=3.039282,5.817261&geocode=FWY2LwIdQ608CCmZRDTmXYAdYDHIkigHdKko8Q%3BFarLEAIdDnP7BylPoPsvcQVUNTFmV2lgJoDDEA&vpsrc=0&ttype=now&noexp=0&noal=0&sort=time&mra=ls&t=h&z=7&start=0

[11:21:00.0000] <timeless>
every time i try to use gmaps to get transit directions in japan, it sends me to an airport!

[11:21:01.0000] <timeless>
ooh, options has a [] no airplanes box!

[11:22:00.0000] <karlcow>
heh

[11:22:01.0000] <timeless>
http://maps.google.com/maps?saddr=Nagano,+Nagano+Prefecture,+Japan&daddr=Kyoto,+Kyoto+Prefecture,+Japan&hl=en&ll=35.808904,136.977539&spn=3.006745,5.817261&sll=35.808904,136.988525&sspn=3.006745,5.817261&geocode=FWY2LwIdQ608CCmZRDTmXYAdYDHIkigHdKko8Q%3BFTQ8FgId3acXCCnxwzzN1qgBYDE9HbtrNh2WwA&vpsrc=0&ttype=now&noexp=0&noal=0&sort=time&mra=ltm&t=h&z=8&start=0

[11:22:02.0000] <timeless>
ok, there we go

[11:22:03.0000] <timeless>
250miles

[11:22:04.0000] <timeless>
4:30

[11:22:05.0000] <timeless>
(by train)

[11:23:00.0000] <timeless>
but that's only ~400km, and we were using 600km as a benchmark

[11:24:00.0000] <timeless>
http://maps.google.com/maps?saddr=Nagano,+Nagano+Prefecture,+Japan&daddr=Takahashi,+Okayama+Prefecture,+Japan&hl=en&sll=35.621582,135.900879&sspn=3.013818,5.817261&geocode=FWY2LwIdQ608CCmZRDTmXYAdYDHIkigHdKko8Q%3BFYLfEgIdaNP2BykhIvph4LRWNTGhAIJnEygEew&vpsrc=0&ttype=now&noexp=0&noal=0&sort=time&mra=ls&t=h&z=8&start=0

[11:24:01.0000] <timeless>
6-7hrs / 400mile ~640km

[11:25:00.0000] <timeless>
which means that 8hours for a train from montreal to nyc really isn't *much* worse than japan :)

[11:25:01.0000] <karlcow>
tss tss

[11:25:02.0000] <timeless>
and that's with japan's *good* network

[11:25:03.0000] <karlcow>
big cities connection

[11:25:04.0000] <timeless>
vs a pair of cities which don't care to connect

[11:25:05.0000] <AryehGregor>
/me waves to timeless 

[11:25:06.0000] <timeless>
yeah, hi, back @nyc?

[11:25:07.0000] <timeless>
/me presumes AryehGregor flew

[11:26:00.0000] <AryehGregor>
Yep.

[11:26:01.0000] <timeless>
(we're having a long rolling conversation about transit in the NE corridor)

[11:26:02.0000] <AryehGregor>
Heck yeah I flew.  I went to Toronto by bus once, I'm not going to put up with that for a meetup.

[11:26:03.0000] <timeless>
heh

[11:27:00.0000] <timeless>
i'm going to albany by bus

[11:27:01.0000] <timeless>
(that's more or less one of the starting points for this conversation)

[11:27:02.0000] <timeless>
your flight which covers more distance was probably ~200usd?

[11:28:00.0000] <timeless>
priceline is offering 273usd

[11:28:01.0000] <timeless>
so i guess we should call it 300usd

[11:28:02.0000] <AryehGregor>
Well, Google paid for it.  They said just tell them what flight I wanted, so I didn't try fidgeting with departure dates or anything to get the best price.

[11:28:03.0000] <timeless>
flying from here to albany is ~1000

[11:29:00.0000] <timeless>
yeah, i know

[11:29:01.0000] <AryehGregor>
It was something like $700 or $800 round-trip, I think.

[11:29:02.0000] <timeless>
i'm just using it for perspective

[11:29:03.0000] <timeless>
ouch

[11:29:04.0000] <timeless>
business class?

[11:29:05.0000] <AryehGregor>
No, regular.

[11:29:06.0000] <AryehGregor>
Economy or whatever.

[11:29:07.0000] <AryehGregor>
It surprised me

[11:29:08.0000] <timeless>
very short notice?

[11:29:09.0000] <AryehGregor>
.

[11:29:10.0000] <AryehGregor>
Yes, less than a week before IIRC.

[11:29:11.0000] <timeless>
ok

[11:30:00.0000] <timeless>
anyway, the 1000cur is for over a month in advance

[11:30:01.0000] <AryehGregor>
A round trip to Israel is only like $1500, but I guess that's if you buy it well in advance and don't care about the exact dates/times.

[11:30:02.0000] <timeless>
whereas fwiw, in general your roundtrip flight should be 300-400

[11:30:03.0000] <AryehGregor>
That was more like what I expected.

[11:30:04.0000] <AryehGregor>
But I looked at price comparison sites and such and didn't find it.

[11:31:00.0000] <AryehGregor>
Maybe I was booking too late to get a decent price, but there were empty seats on the plane, so I dunno.

[11:33:00.0000] <timeless>
kayak is offering ~900usd for flights from jfk to tlv for um

[11:33:01.0000] <timeless>
nov 16 return dec 14

[11:33:02.0000] <timeless>
(turkish airlines)

[11:33:03.0000] <timeless>
(dates randomly selected by throwing darts into the future at a calendar)

[11:34:00.0000] <timeless>
anyway, point being, it's cheaper for you to fly to israel than for me to fly to albany

[11:35:00.0000] <timeless>
(oh, and priceline will give you those flights for 2$ less than kayak [898 v 900!])

[11:38:00.0000] <AryehGregor>
Heh, we were talking about how Google Docs reimplements everything from scratch -- I just saw a message "Your browser's current zoom level is not supported.  Please reset to the default zoom."

[11:39:00.0000] <AryehGregor>
That's what you get, I guess.

[11:39:01.0000] <timeless>
yeah, i've seent that message

[11:40:00.0000] <AryehGregor>
So it seems like my credit card charges in Canada are taking a long time to post.

[11:41:00.0000] <timeless>
they can easily take over a month

[11:41:01.0000] <timeless>
the same applies for roaming charges

[11:41:02.0000] <timeless>
which is great if you need to get approval for your expenses..

[11:41:03.0000] <timeless>
and are supposed to do it as a single bundle

[11:41:04.0000] <AryehGregor>
Great.

[11:42:00.0000] <AryehGregor>
Well, I don't need approval, I just need to file for reimbursement.

[11:42:01.0000] <timeless>
/me hasn't really had that problem yet, not having a working credit card for a complete trip yet

[11:42:02.0000] <AryehGregor>
Conveniently, I haven't actually lost any money until the charges post, so . . .

[11:42:03.0000] <timeless>
well, if you live month to month, iut

[11:42:04.0000] <timeless>
's the same thing

[11:42:05.0000] <AryehGregor>
No, that's the nice thing about credit cards.  It doesn't come out of my bank account.

[11:42:06.0000] <timeless>
/me isn't living month to month anymore

[11:42:07.0000] <AryehGregor>
(although I don't live month to month or anywhere close to it, I actually still live with my parents)

[11:42:08.0000] <AryehGregor>
:)

[11:43:00.0000] <timeless>
yeah, my parents left toronto this morning

[11:43:01.0000] <timeless>
they're on their way to somewhere near detroit

[11:46:00.0000] <timeless>
btw, what kind of internet connection do you have? fiber? dsl? cable?

[11:46:01.0000] <timeless>
[fiber/fios]

[11:46:02.0000] <timeless>
/me is wondering about costs of providers and wouldn't mind an NYC based reference point

[11:49:00.0000] <karlcow>
cable.

[11:49:01.0000] <timeless>
bandwidth, limit/month, price/month?

[11:51:00.0000] <karlcow>
Download: 30Mb/s Upload: 2Mb/s, 120 Go/month, 66.95 CAD/months +15% Taxes (QC)

[11:51:01.0000] <timeless>
hrm

[11:52:00.0000] <timeless>
i'm looking at .. teksavvy.com

[11:52:01.0000] <AryehGregor>
I pulled a USB device out of Windows while logging out without saying "safely remove hardware" and Windows blue-screened.

[11:52:02.0000] <AryehGregor>
lol windows.

[11:52:03.0000] <AryehGregor>
0x0000008e.

[11:52:04.0000] <AryehGregor>
/me looks it up

[11:52:05.0000] <jcranmer>
ERR_USER_WAS_AN_IDIOT

[11:52:06.0000] <karlcow>
http://www.videotron.com/service/internet-services/internet-access

[11:53:00.0000] <timeless>
they're offering 15 ; 1; 9tb (estimated assuming constant download); 54.95 cad/month +xx% taxes (ON)

[11:53:01.0000] <timeless>
that's cable

[11:53:02.0000] <timeless>
AryehGregor: are you an admin?

[11:53:03.0000] <timeless>
install windbg and pull up the .dmp file

[11:54:00.0000] <timeless>
also did the STOP message finger a driver?

[11:54:01.0000] <AryehGregor>
KERNEL_MODE_EXCEPTION_NOT_HANDLED, apparently, or else STATUS_ACCESS_VIOLATION.

[11:54:02.0000] <AryehGregor>
I didn't look closely.

[11:54:03.0000] <AryehGregor>
I'm an admin but it's my parents' computer, I was just using it for the scanner.

[11:54:04.0000] <timeless>
the first google hit for your stop is rustock rootkit :)

[11:54:05.0000] <timeless>
install windbg

[11:55:00.0000] <timeless>
it's a pretty fast process

[11:55:01.0000] <timeless>
karlcow: the big question i have is whether there's a risk of lower quality service using cable than dsl

[11:55:02.0000] <AryehGregor>
See, but I don't care.  I use Linux.  Not my problem.

[11:56:00.0000] <timeless>
because traditionally cable is shared w/ neighbors

[11:56:01.0000] <timeless>
AryehGregor: you don't care if your parent's banking info is stolen/abused?

[11:56:02.0000] <timeless>
http://en.wikipedia.org/wiki/First_they_came%E2%80%A6

[11:56:03.0000] <karlcow>
How we found the file that was used to Hack RSA — http://www.f-secure.com/weblog/archives/00002226.html

[11:56:04.0000] <AryehGregor>
I'm not going to spend some unknown amount of time debugging their computer without them asking me to fend off that risk, no.

[11:57:00.0000] <AryehGregor>
My father manages their computers, he's a programmer too.

[11:57:01.0000] <karlcow>
timeless: I'm working from home mostly

[11:57:02.0000] <AryehGregor>
I doubt it's virus-related.  Windows System Internals says it's one of the most common blue screens.

[11:57:03.0000] <karlcow>
so the neighbors are mostly not here :)

[11:57:04.0000] <timeless>
karlcow: heh

[11:57:05.0000] <timeless>
AryehGregor: yeah, STOP is

[11:57:06.0000] <timeless>
but it's helpful to know what caused it

[11:57:07.0000] <timeless>
sometimes it just means there's a driver that should be updated

[11:58:00.0000] <AryehGregor>
My parents chose to use Windows, they get to suffer with the consequences of their decision.  :)

[11:58:01.0000] <AryehGregor>
(but it didn't affect them anyway, except for restarting the computer)

[12:00:00.0000] <timeless>
a buggy driver can affect them in the future

[12:00:01.0000] <timeless>
and a virus will affect them in the future :)

[12:00:02.0000] <AryehGregor>
Of course it can.  They use Windows.

[12:00:03.0000] <AryehGregor>
Of course it will.  They use Windows.

[12:00:04.0000] <timeless>
/me chuckles

[12:00:05.0000] <timeless>
buggy drivers exist on other platforms too you know :)

[12:00:06.0000] <timeless>
/me has had panics on just about every paltform imaginable

[12:01:00.0000] <AryehGregor>
They're a heck of a lot less frequent on non-Windows platforms.

[12:01:01.0000] <AryehGregor>
I don't believe I've ever had a kernel panic on Linux after boot on my desktop, although I might be misremembering.

[12:01:02.0000] <AryehGregor>
(I've had X hangs and crashes, though!)

[12:01:03.0000] <timeless>
i've had panic's on my linux systems

[12:01:04.0000] <timeless>
i think some were caused by a nic

[12:01:05.0000] <Ms2ger>
timeless, including emacs? :)

[12:02:00.0000] <timeless>
Ms2ger: i've had that die too, i think even at worldgate

[12:02:01.0000] <timeless>
and jesup used emacs as his os (he's on moznet these days)

[12:02:02.0000] <Ms2ger>
jesup works for Mozilla now

[12:03:00.0000] <timeless>
i've certainly had panic's on my linux based nokia tablets :)

[12:03:01.0000] <timeless>
i've also caused at least one freebsd box to panic :)

[12:03:02.0000] <timeless>
do you remember / have you seen `stress`? :)

[12:04:00.0000] <smaug____>
timeless: you've probably done unusual things with your nokia tablets :)

[12:04:01.0000] <timeless>
(that page has killed all sorts of devices)

[12:04:02.0000] <smaug____>
though, N900 does crash occasionally

[12:04:03.0000] <timeless>
smaug____: to be fair, most of the panics were in code that hadn't shipped

[12:04:04.0000] <AryehGregor>
They also probably run low-quality out-of-tree drivers.

[12:04:05.0000] <timeless>
well, of course

[12:04:06.0000] <AryehGregor>
Heh.

[12:04:07.0000] <timeless>
just like the drivers on your parents' windows system

[12:04:08.0000] <AryehGregor>
Yeah, but the point is, on desktop Linux you usually don't have to use such drivers.

[12:04:09.0000] <timeless>
the odds of that windows system actually panicing because of a bug in MS drivers

[12:05:00.0000] <timeless>
vs a bug in third party drivers

[12:05:01.0000] <timeless>
hrm

[12:05:02.0000] <AryehGregor>
I don't think I use any, unless Ubuntu sneakily enabled them without asking me.

[12:05:03.0000] <timeless>
you don't use nvidia/amd/ati video drivers?

[12:05:04.0000] <timeless>
or whichever

[12:05:05.0000] <AryehGregor>
Nope.

[12:05:06.0000] <timeless>
you're no fun

[12:05:07.0000] <timeless>
and um

[12:05:08.0000] <AryehGregor>
I use nouveau.  Works for me.

[12:05:09.0000] <timeless>
iirc ubuntu planned to change their default ther

[12:05:10.0000] <AryehGregor>
It will even support 3D soon.

[12:05:11.0000] <karlcow>
TabAtkins: why only rotation compared to a full EXIF apis?

[12:07:00.0000] <timeless>
http://www.ubuntu.com/project/about-ubuntu/licensing

[12:07:01.0000] <timeless>
... In addition, we install some hardware drivers that are available only in binary format, but such packages are clearly marked in the restricted component.

[12:13:00.0000] <timeless>
/me sighs

[12:13:01.0000] <timeless>
the portable contacts "spec" has really terrible text

[12:13:02.0000] <timeless>
> Service Provider MAY return a subset of the requested fields if they are not supported.

[12:14:00.0000] <timeless>
/me reads that as service provider may return a subset (including not supported fields, and excluding some supported fields)) of the requested fields

[12:15:00.0000] <timeless>
... if they (the subset) are not (all) supported

[12:15:01.0000] <timeless>
:)

[12:15:02.0000] <jcranmer>
Service Provider may exclude a requested field if said field is not supported

[12:16:00.0000] <timeless>
jcranmer: it didn't say that though

[12:16:01.0000] <timeless>
it should have

[12:16:02.0000] <jcranmer>
s/a/any/

[12:16:03.0000] <jcranmer>
it's more precise and shorter, too

[12:16:04.0000] <Ms2ger>
Service Provider MUST support all fields.

[12:16:05.0000] <timeless>
indeed

[12:16:06.0000] <jcranmer>
and arguably less confusing

[12:16:07.0000] <timeless>
and that

[12:17:00.0000] <timeless>
anyway, i highly recommend *not* reading this `spec`

[12:17:01.0000] <jcranmer>
I read ISO C++ from time to time

[12:17:02.0000] <timeless>
/me nods

[12:17:03.0000] <jcranmer>
and old RFCs

[12:17:04.0000] <timeless>
i have a feeling this is worse than both

[12:17:05.0000] <timeless>
and i'm well aware of both

[12:17:06.0000] <jcranmer>
after those, I doubt you can find anything else that's crap

[12:18:00.0000] <jcranmer>
well, excluding patents

[12:18:01.0000] <timeless>
/me has read some portions of iso c++ (probably proposals as opposed to the final versions)

[12:18:02.0000] <timeless>
/me remembers visiting the patent office as a child

[12:19:00.0000] <timeless>
my grandfather (zlb) wanted to file a patent and was doing due diligence

[12:19:01.0000] <timeless>
(which is something no one seems to do anymore)

[12:20:00.0000] <jcranmer>
stupid State St. decision

[12:20:01.0000] <timeless>
and i

[12:21:00.0000] <timeless>
ve of course read old rfcs

[12:21:01.0000] <timeless>
especially mime, pop, smtp, imap, nntp ..

[12:22:00.0000] <timeless>
>     500:         Internal Server Error (un unexpected error occurred during processing)

[12:22:01.0000] <timeless>
/me ponders the random `un`

[12:23:00.0000] <timeless>
s/sepc/spec/

[12:23:01.0000] <timeless>
/me suspects no one actually reviewed this sepc

[12:26:00.0000] <jcranmer>
it's not even at last call yet, I think

[12:27:00.0000] <timeless>
?

[12:28:00.0000] <Ms2ger>
Is it a Living Standard?

[12:28:01.0000] <timeless>
/me hopes it's an unused standard

[12:29:00.0000] <jcranmer>
oh, I misread

[12:29:01.0000] <timeless>
hrm, it's used by google

[12:29:02.0000] <jcranmer>
I think you were talking about the DevAPI Contacts spec

[12:29:03.0000] <timeless>
no

[12:29:04.0000] <timeless>
http://portablecontacts.net/draft-spec.html if you must read it

[12:29:05.0000] <timeless>
(you were warned)

[12:30:00.0000] <timeless>
ooh, google is getting a new sign in page

[12:31:00.0000] <jcranmer>
egads

[12:31:01.0000] <jcranmer>
heavy use of rules

[12:32:00.0000] <timeless>
?

[12:33:00.0000] <jcranmer>
horizontal rules

[12:37:00.0000] <jcranmer>
in other words, the spec formatting is complete crap

[12:37:01.0000] <jcranmer>
which is totally inexcusable if you're using HTML

[12:42:00.0000] <timeless>
i did try to warn you

[12:43:00.0000] <timeless>
/me is reading the spec in ms word

[12:43:01.0000] <Philip`>
Seems to be the default xml2rfc formatting, so you should feel lucky they didn't just use the ASCII output

[12:45:00.0000] <timeless>
grr

[12:45:01.0000] <timeless>
the web page i'm visiting is showing me <!--[if !mso]>

[12:45:02.0000] <jcranmer>
I actually like the ASCII output

[12:57:00.0000] <Hixie>
AryehGregor: k, i'll nuke onundo and onredo.

[13:48:00.0000] <zcorpan>
so could somebody explain to me what's bad about constants appearing on both the interface object and the prototype?

[13:53:00.0000] <zcorpan>
does ecmascript have a convention for constants?

[13:53:01.0000] <TabAtkins>
Dunno.  I suspect the convention is on the interface.

[13:54:00.0000] <zcorpan>
any example?

[13:58:00.0000] <TabAtkins>
For example, the Node interface.

[13:59:00.0000] <zcorpan>
Node was the example used in the thread where constants are on both :)

[13:59:01.0000] <TabAtkins>
Heh.

[13:59:02.0000] <TabAtkins>
Math.PI

[13:59:03.0000] <TabAtkins>
That's *just* an interface object, of course.

[13:59:04.0000] <zcorpan>
there's no object implementing Math

[13:59:05.0000] <zcorpan>
yeah

[14:02:00.0000] <TabAtkins>
In IDB, IDBRequest, IDBCursor, and IDBTransaction all put constants on the interface.  The only thing with constants on the object is IDBDatabaseException, which matches the general exception pattern.

[14:03:00.0000] <zcorpan>
doesn't IDB just use webidl to declare its constants?

[14:03:01.0000] <TabAtkins>
Yeah.

[14:04:00.0000] <zcorpan>
then webidl makes the constants appear on the prototype also

[14:05:00.0000] <TabAtkins>
Oh, then I'm confused.  What did you mean by "interface object" if not a WebIDL interface?

[14:06:00.0000] <zcorpan>
"interface object" is defined in webidl

[14:06:01.0000] <zcorpan>
window.Node is an interface object

[14:06:02.0000] <zcorpan>
Node.ELEMENT_NODE is defined and Node.prototype.ELEMENT_NODE is also

[14:07:00.0000] <TabAtkins>
Ah, okay.  Then shrug, if you use IDL you just define them on both I guess.

[14:08:00.0000] <zcorpan>
http://lists.w3.org/Archives/Public/public-script-coord/2011JulSep/0314.html

[14:11:00.0000] <TabAtkins>
Ah, then I won't be able to help much.  I don't know much about what happens on the pure-ES side of things.

[15:10:00.0000] <zcorpan>
/me removed onundo and onredo from html-elements

[15:11:00.0000] <zewt>
heh what the

[15:11:01.0000] <zewt>
just installed current opera and apparently the ui is a straight-up chrome mimic now

[15:11:02.0000] <bga_>
:(

[15:12:00.0000] <bga_>
9.64 is best

[15:12:01.0000] <zewt>
uhh

[15:12:02.0000] <zewt>
apparently onbeforeunload is completely broken in opera now?

[15:12:03.0000] <zewt>
compose mail in gmail, close tab, mail lost

[15:13:00.0000] <zcorpan>
i don't think onbeforeunload has ever worked in opera

[15:13:01.0000] <bga_>
opera keep pages during 2 min on memory after you close it

[15:14:00.0000] <bga_>
may be you want reopen it again

[15:14:01.0000] <zewt>
well, i compose a mail, close the tab, control-alt-t and the changes are gone, so

[15:14:02.0000] <zewt>
strange to be missing such a basic thing

[15:18:00.0000] <zewt>
gah opera made itself the default browser without asking

[15:20:00.0000] <zcorpan>
opera knows what you want :)

[15:20:01.0000] <zewt>
opera is telling me what i want, my opinion doesn't seem to enter into it

[15:21:00.0000] <zcorpan>
which OS?

[15:21:01.0000] <zewt>
windows

[15:21:02.0000] <zewt>
just very strange to throw away one of the most basic, well-established things You Have To Ask Permission For

[15:25:00.0000] <zcorpan>
windows 7?

[15:28:00.0000] <zewt>
xp

[15:29:00.0000] <zcorpan>
ok

[15:29:01.0000] <zewt>
it seems like there was a heedless rush for "zero-click installs" (which is actually one-click since there's still EULA crap), without much concern that there are things you have to ask, heh

[15:29:02.0000] <jgraham>
zewt: Are we using the same Opera? I don't think "chrome mimic" is even close

[15:30:00.0000] <jgraham>
And yeah, onbeforeunload doesn't work in opera

[15:30:01.0000] <bga_>
jgraham you just have old config

[15:31:00.0000] <jgraham>
bga_: It would have to be a pretty epic configuration change

[15:32:00.0000] <jgraham>
(possibly you just mean that the theme looks a bit like chrome or something)

[15:32:01.0000] <zcorpan>
maybe opera and chrome look more alike on windows than on mac

[15:32:02.0000] <jgraham>
But ui !== theme

[15:32:03.0000] <jgraham>
except in a very superficial way

[15:32:04.0000] <zewt>
https://zewt.org/~glenn/foo1.png https://zewt.org/~glenn/foo2.png

[15:33:00.0000] <zewt>
tabs even animate almost identically now

[15:33:01.0000] <zewt>
the "opera" button is more like FF, i guess

[15:34:00.0000] <jgraham>
Well you have the same website open in both. That's pretty similar I guess

[15:34:01.0000] <zewt>
don't get me wrong, i'm fine with browsers ganking stuff from other browsers when it's a good idea, just compared to the last time i installed opera it was pretty "wtf?"

[15:35:00.0000] <jgraham>
hen was the last time you installed Opera?

[15:35:01.0000] <zewt>
also i sat there for a second hitting alt-f trying to open the menu and apparently keyboard shortcuts are old and busted :(

[15:35:02.0000] <jgraham>
It is pretty differnet from Opera 5...

[15:35:03.0000] <zewt>
10 maybe? it had a standard menu, heh

[15:38:00.0000] <Dashiva>
Probably 9.x then

[15:39:00.0000] <jgraham>
I thought the single menu was introduced in 10.50

[15:39:01.0000] <zewt>
maybe an earlier 10?

[15:39:02.0000] <zewt>
seems like a big UI change to make without bumping a major version number, though

[15:40:00.0000] <zcorpan>
at opera, .50 is a major version. sometimes.

[15:41:00.0000] <jgraham>
If our version numbering is correlated with something, no one has explained the system to me and I m not bright enough o reverse engineer it

[15:41:01.0000] <zewt>
that sounds like a great way to confuse users heh

[15:41:02.0000] <jgraham>
I'm not saying it *isn't* correlated to anything

[15:41:03.0000] <zewt>
if i'm on x.y and i upgrade to x.z, i'm doing so with the expectation of not seeing huge ui changes

[15:41:04.0000] <TabAtkins>
Quick question: on http://www.xanthir.com/etc/exif/ do you see all three Fs as the same size?

[15:42:00.0000] <jgraham>
Just that I'm not in the desktop team :)

[15:42:01.0000] <zewt>
left squiggly F is smaller

[15:42:02.0000] <zewt>
in ff6

[15:43:00.0000] <TabAtkins>
What. The. Hell.  For some reason FF and only FF is receiving an older version of the first image that's bigger (and thus scales to a smaller size).

[15:43:01.0000] <jgraham>
Same in Opera

[15:44:00.0000] <TabAtkins>
IE and Chrome get the correct image.  wtf

[15:45:00.0000] <TabAtkins>
There, fixed.  Dunno what was up, but deleting the file on the server and re-uploading fixed it.

[15:48:00.0000] <zewt>
there are heisenbugs, schroedinbugs and screwitbugs

[15:54:00.0000] <zcorpan>
MikeSmith: is there an es5.1 html version?


2011-08-27
[17:09:00.0000] <mkanat>
Is there any effort presently to standardize <video> performance information?

[17:17:00.0000] <Hixie>
mkanat: yeah, there's a wiki page on it and discussion on the whatwg list

[17:18:00.0000] <Hixie>
mkanat: it's on my list of things to get to eventually

[17:18:01.0000] <mkanat>
Hixie: Cool. It's probably something I will take an interest in as a user.

[17:18:02.0000] <mkanat>
(In my YouTube capacity, not my Bugzilla capacity. :-) )

[17:20:00.0000] <Hixie>
man, w3.org is down as often as whatwg.org

[17:20:01.0000] <Hixie>
and i'm paying for whatwg.org with loose change

[17:20:02.0000] <mkanat>
Ha.

[17:26:00.0000] <TabAtkins>
Dude, you're paying for whatwg.org yourself?

[17:34:00.0000] <Hixie>
TabAtkins: it's the same box as hixie.ch (and about 60 other domains)

[17:34:01.0000] <Philip`>
Hixie: You haven't spent decades telling the world's software to DDOS your server

[17:36:00.0000] <Hixie>
k well i guess i can't get any work done if i can't get to w3.org

[17:39:00.0000] <Hixie>
did they get hit by the storm maybe?

[17:39:01.0000] <Hixie>
MikeSmith: yt?

[17:39:02.0000] <Hixie>
shepazu: yt?

[17:56:00.0000] <shepazu>
Hixie: here now

[17:56:01.0000] <shepazu>
was out for an evening constitutional

[17:58:00.0000] <shepazu>
hmmm… you're right, looks like w3.org is down… (at least some servers)

[17:58:01.0000] <shepazu>
weird!

[17:58:02.0000] <zewt>
vincent scheib killing off his own threads with these unreadable three-page mails

[18:00:00.0000] <TabAtkins>
zewt: He's responding to feedback, same as Hixie does with his mega emails.

[18:00:01.0000] <zewt>
hixie responds concisely, he's responding with large paragraphs

[18:00:02.0000] <zewt>
i sure don't feel motivated to read all of that

[18:01:00.0000] <shepazu>
seems like w3.org is back up, Hixie, please test

[18:51:00.0000] <TabAtkins>
Where's the statement in HTML that all algorithms described are optimized for simplicity and clarity, not efficiency, so UAs may implement any algorithm that produces the same results?

[20:16:00.0000] <Hixie>
shepazu: cool, thanks. i was actually pinging you to make sure everyone in boston was ok!

[20:16:01.0000] <Hixie>
looks like the storm hasn't hit yet though

[20:17:00.0000] <shepazu>
Hixie: ah, ok… actually, I'm in North Carolina, not Boston

[20:18:00.0000] <Hixie>
aah

[20:18:01.0000] <shepazu>
and the storm isn't due here until late tonight or tomorrow, then another several hours to Boston

[20:18:02.0000] <shepazu>
thanks for asking though

[20:18:03.0000] <shepazu>
NC has a lot of hurricanes, so I'm not that worried

[21:33:00.0000] <MikeSmith>
very cool to see that doublec implemented the media temporal-fragments URI stuff

[00:40:00.0000] <MikeSmith>
node now has support for typed arrays

[00:57:00.0000] <micheil>
yup

[01:06:00.0000] <MikeSmith>
oh hey it's micheil

[01:06:01.0000] <MikeSmith>
long time no chat

[01:06:02.0000] <MikeSmith>
you node KO'ing this weekend I guess?

[01:59:00.0000] <krijn>
brucel: how's your dancing outfit coming along?

[01:59:01.0000] <krijn>
Did you see Stephen and Chris are also going to perform for us? :]

[02:29:00.0000] <Ms2ger>
TabAtkins, http://www.whatwg.org/specs/web-apps/current-work/multipage/infrastructure.html#conformance-requirements

[02:30:00.0000] <micheil>
MikeSmith: I'm judging, and technically I'm also a staff member

[02:30:01.0000] <micheil>
but I'm mainly doing pusher.com stuff.

[03:20:00.0000] <zcorpan>
Philip`: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10830#c45

[04:20:00.0000] <Philip`>
zcorpan: Don't really know how to count that

[04:21:00.0000] <Philip`>
(Also don't really know what the relevance is)

[04:26:00.0000] <zcorpan>
yeah me either

[05:28:00.0000] <karlcow>
http://tantek.com/2011/238/b1/many-ways-slice-url-name-pieces

[05:28:01.0000] <karlcow>
http://www.la-grange.net/2011/08/27/broken-tantek

[05:29:00.0000] <karlcow>
I'm a bit puzzled

[05:29:01.0000] <karlcow>
Opera seems to not display correctly stuff, only me?

[05:34:00.0000] <Dashiva>
There's a CSS rule with i{content:"\"/*"}

[05:38:00.0000] <Dashiva>
I don't understand why it doesn't affect other browsers

[05:38:01.0000] <Ms2ger>
Because they don't implement content on elements

[05:41:00.0000] <Dashiva>
That would explain it I guess

[08:26:00.0000] <karlcow>
http://sethsandler.com/multitouch/interactive-table-list/

[08:27:00.0000] <karlcow>
"timeline of interactive tables. These tables may involve multi-touch, single-touch, audio, objects, or other unique types of table interactions."


2011-08-28
[19:52:00.0000] <karlcow>
http://www.pbs.org/mediashift/2011/08/publishers-doing-an-apple-end-run-to-deliver-to-ipad236.html

[02:30:00.0000] <rabbi1>
working on a site that got, top menu, main menu and search. Very confused with the arrangements and color combination. Tried almost all combinations, no luck. any help/ suggestions please ?

[08:46:00.0000] <Ms2ger>
<html xmlns="http://www.w3.org/Profiles/html4-transitional">

[08:47:00.0000] <Ms2ger>
shepazu++

[09:42:00.0000] <shepazu>
Ms2ger, ?

[09:42:01.0000] <Ms2ger>
Your last email to spec-prod

[10:07:00.0000] <shepazu>
oh… I sent that days ago...

[10:07:01.0000] <shepazu>
/me is tired of the topic, it should simply happen

[10:07:02.0000] <shepazu>
what else usefully has to be said?

[10:08:00.0000] <shepazu>
（╯°□°）╯︵ ɹʇ/ uı 5ןɯʇɥ

[11:02:00.0000] <kennyluck>
shepazu, lol.

[11:03:00.0000] <shepazu>
:D

[11:05:00.0000] <Hixie>
just as a heads-up, i'm going to be only intermittently around for the next few days


2011-08-29
[01:49:00.0000] <jgraham>
Conclusion from the last two days of HTML related email: there is nothing interesting happening.

[03:35:00.0000] <gsnedders>
/me finally gets around to looking at what heycam changed in WebIDL because of him :P

[03:36:00.0000] <gsnedders>
Though trying to load the spec on train wifi isn't working brilliantly

[03:37:00.0000] <gsnedders>
(I guess dev.w3.org being on a sewing machine doesn't help)

[03:42:00.0000] <smaug____>
/me imagines a server getting electricity from an old singer http://upload.wikimedia.org/wikipedia/commons/0/00/Singer_sewing_machine.jpg

[03:43:00.0000] <tomasf>
maybe they let the sewing machine "render" the page on actual fabric and automatically take a photo of it and deliver that as the web page

[03:46:00.0000] <jgraham>
Lack of a CSS media type for fabric suggests that isn't the case

[04:24:00.0000] <smaug____>
Does anyone have tests for canvas objects in "data" documents

[04:25:00.0000] <smaug____>
something like

[04:25:01.0000] <smaug____>
document.implementation.createHTMLDocument("").createElement("canvas").getContext("2d");

[04:25:02.0000] <smaug____>
and then using that context somehow

[04:36:00.0000] <woef>
If they say html5 should be "backwards compatible" ... how far back in time are they talking about?

[04:37:00.0000] <jgraham>
If who says what?

[04:38:00.0000] <jgraham>
For developing HTML5 the tenets were more like "support existing content" and "no mode switching"

[04:38:01.0000] <woef>
I remember a discussion about why a <dl> only has "assumed" list items (no unique container)

[04:39:00.0000] <woef>
It was said that adding one was not backwardds compatible with what existed today.

[04:39:01.0000] <jgraham>
So you can't make changes that break (large enoguh numbers of) existing sites (that browser makers would be reluctant to make the changes). Since some sites are as old as the web, that gives a timescale

[04:40:00.0000] <woef>
Wouldn't it be interesting to make a "intermediate" tracks. Support two modes, until the old way of working gets fazed out?

[04:40:01.0000] <jgraham>
No

[04:40:02.0000] <jgraham>
It's a lot of effort

[04:40:03.0000] <jgraham>
and there is no good idea of what the timescale for the old way fading out is

[04:40:04.0000] <woef>
Yeah ...

[04:40:05.0000] <gsnedders>
woef: How many things have ever been fazed out?

[04:41:00.0000] <woef>
image maps?

[04:41:01.0000] <woef>
flash slowly is going away.

[04:41:02.0000] <jgraham>
(pretty sure phased is spelt wityh a ph and an s)

[04:41:03.0000] <woef>
fazed is the "i'm a cool geek" way of spelling ;)

[04:42:00.0000] <jgraham>
Ah. I never get those memos for some reaswon

[04:42:01.0000] <woef>
I'm afraid we don't really do the whole ph-thing in Belgium :)

[04:42:02.0000] <woef>
I get that wrong a lot.

[04:42:03.0000] <gsnedders>
woef: image maps are still supported by all browsers.

[04:42:04.0000] <jgraham>
woef: Yeah, but gsnedders is scottish. That's almost english, so he doesn't have an excuse

[04:43:00.0000] <woef>
Hah!

[04:43:01.0000] <gsnedders>
woef: An extreme example is the plaintext element, noted as being deprecated in the first ever HTML spec, still special-cased in the parser in HTML 5.

[04:44:00.0000] <woef>
I once tried to read some Irvine Welsh books, my English lessons didn't really help.

[04:44:01.0000] <gsnedders>
jgraham: I just copied what woef said!

[04:44:02.0000] <gsnedders>
jgraham: You know I can't spell. :P

[04:44:03.0000] <gsnedders>
Copying others is normally better than guessing myself :P

[04:44:04.0000] <Philip`>
/me is unphased by homonyms

[04:44:05.0000] <gsnedders>
(It also provides a convenient way to mitigate blame for my inability to spell.)

[04:45:00.0000] <jgraham>
Is someone who is phased by homonyms a homophobe?

[04:46:00.0000] <gsnedders>
woef: Speaking of Welsh, the only book I found hard to read was one with phonetic transcription of dialect from where I grew up. :)

[04:46:01.0000] <gsnedders>
(I can cope with it fine when spoken.)

[04:51:00.0000] <woef>
I think I tried Trainspotting once.

[04:51:01.0000] <woef>
But if I get it right, we'll never have a <dl> list with proper list items?

[04:51:02.0000] <woef>
Or nested <a>-tags?

[04:53:00.0000] <gsnedders>
woef: nested a-elements will break existing sites, so there's no way you'll get browsers to change behaviour. Equally, what does a link within a link *do*?

[04:53:01.0000] <gsnedders>
woef: What's the use-case for changing dl though? What's unclear about dt/dd?

[04:53:02.0000] <woef>
link to it's href?

[04:53:03.0000] <gsnedders>
woef: which href?

[04:53:04.0000] <jgraham>
gsnedders: (pretty obvious what it does, really)

[04:53:05.0000] <woef>
The one with the deepest nesting.

[04:53:06.0000] <jgraham>
Still won't happen ofc

[04:54:00.0000] <Philip`>
(Presumably it does the same as it does if you nest <a> in XHTML)

[04:54:01.0000] <woef>
Like "i block-level-link my blog intro but the contained link to the comment section goes there immediately"

[04:54:02.0000] <gsnedders>
jgraham: (that would require half a second of thought which I'm not doing while on vacation, even if I am responding to spec changes I requested)

[04:54:03.0000] <jgraham>
Or in HTML with script

[04:54:04.0000] <gsnedders>
woef: Still, changing the parser would break existing sites.

[04:54:05.0000] <woef>
gsnedders: how so?

[04:55:00.0000] <woef>
I hope nobody is nesting links now, if you see the dom-crap that ensues :)

[04:56:00.0000] <jgraham>
hahaha

[04:56:01.0000] <jgraham>
:)

[04:56:02.0000] <gsnedders>
woef: Plenty of people are. The vast majority of the web doesn't follow the spec.

[04:56:03.0000] <gsnedders>
Like, approximately 100% of the webhas invalid HTML.

[04:56:04.0000] <woef>
But these sites break by default :p

[04:57:00.0000] <jgraham>
I wonder what the half life of this conversation is

[04:57:01.0000] <gsnedders>
Breaking the web would cause people to not use a browser, so good luck getting any browser to change :P

[04:57:02.0000] <jgraham>
We've had the same discussion with different people since 2004

[04:57:03.0000] <jgraham>
Hopefully it will eventually stop

[04:57:04.0000] <woef>
As for the <dl>, I just don't like that there is no <li> equivalent.

[04:58:00.0000] <jgraham>
That one at least might be changable

[04:58:01.0000] <jgraham>
There was a long thread on WHATWG a while back

[04:58:02.0000] <jgraham>
I don't remembert the conclusions

[04:59:00.0000] <woef>
You just say that to increase the suspense.

[05:00:00.0000] <jgraham>
Yep, it's practically Resident Evil in here

[05:01:00.0000] <gsnedders>
At least it's not the merky darkness of Doom 3.

[05:08:00.0000] <Philip`>
Does it count as suspense when you're 100% certain a monster is going to spring into existence behind you when you enter a room?

[05:09:00.0000] <gsnedders>
Okay, true, once you get beyond the first few levels it is like that.

[05:09:01.0000] <woef>
It killed AvsP2

[05:10:00.0000] <woef>
The first one was creepy as hell though.

[05:10:01.0000] <woef>
Damn motion detector.

[06:20:00.0000] <hsivonen>
looks like ISSUEs are as important as ever http://www.w3.org/html/wg/tracker/issues/170

[06:30:00.0000] <hsivonen>
looking at the number of duplicates on https://bugzilla.mozilla.org/show_bug.cgi?id=162653 I wonder if mattur will be able to find an XML design thread from 1997 or 1998 containing an assertion that HTML authors won't be confused by XML doing <foo/>

[06:41:00.0000] <gsnedders>
Is (char)(int)1 a plausible way to check endianness of integers

[06:41:01.0000] <gsnedders>
?

[06:44:00.0000] <jgraham>
gsnedders: Won't the compiler get upset by that?

[06:44:01.0000] <jgraham>
See also http://stackoverflow.com/questions/1001307/detecting-endianness-programmatically-in-a-c-program

[06:44:02.0000] <gsnedders>
(Yay for writing code that depends on endianness)

[06:48:00.0000] <Philip`>
int i = 1; char c[sizeof(int)]; memcpy(c, &i, sizeof(int)); bool is_littleendian = c[0];

[06:49:00.0000] <Philip`>
I'm not sure the union trick is actually legal C (though I'm not sure it's illegal either)

[06:49:01.0000] <Philip`>
but memcpy is always safe

[06:50:00.0000] <gsnedders>
Philip`: It's illegal, I'm pretty certain. At least people have managed to cite quite obvious bits of the spec about it.

[06:50:01.0000] <gsnedders>
(Well, it's not illgal, it's just undefined.)

[06:50:02.0000] <Philip`>
(char)(int)1 will always just give you 1, and *(char*)&i will give you bogus results depending on what compiler version and optimisation settings you use

[06:52:00.0000] <Philip`>
(The union thing works in all compilers in practice, though)

[06:52:01.0000] <gsnedders>
Philip`: Casting to char always take the low-bits?

[06:53:00.0000] <gsnedders>
Philip`: bool is C99?

[06:54:00.0000] <jgraham>
/me would imageine without knowing that casting to char does an actual type conversion rather than just reading some subset of the bits

[06:54:01.0000] <gsnedders>
jgraham: What else can it do?

[06:55:00.0000] <gsnedders>
jgraham: a char is just a numberic type

[06:55:01.0000] <gsnedders>
*numeric

[06:55:02.0000] <jgraham>
Well yes, I know that

[06:55:03.0000] <jgraham>
It could overflow differently, for example. And be endianness independent.

[06:56:00.0000] <Philip`>
If char is unsigned, then (char)i adds or subtracts 256 to/from i until it's in the range [0, 255] and then returns that

[06:56:01.0000] <Philip`>
If char is signed, and i is in [0, 255] then it returns that, else it's implementation-defined behaviour

[06:56:02.0000] <gsnedders>
Philip`: I could remember there was some magic about char, but I couldn't remember what.

[06:57:00.0000] <Philip`>
The magic thing is that you don't know whether it's signed or unsigned

[06:57:01.0000] <gsnedders>
And about casting as well? Which I guess is the adding/subtracting…

[06:58:00.0000] <Philip`>
(bool isn't in C99; use a BOOL or int or whatever, or use C++)

[06:59:00.0000] <Philip`>
In C99, casting is the same process for any integer types (except _Bool)

[06:59:01.0000] <Philip`>
...except replace "256" by appropriate values etc

[07:00:00.0000] <tomasf>
according to http://en.wikipedia.org/wiki/Stdbool.h , bool expands to _Bool in C99

[07:00:01.0000] <gsnedders>
Philip`: (Yeah, I know what to do)

[07:01:00.0000] <Philip`>
Basically it always does the mathematically obvious thing, except it's unwilling to assume two's-complement representation so any cases of signed overflow are implementation-defined behaviour

[07:02:00.0000] <Philip`>
and it's all endian-independent until you try to read the char representation of a larger value

[07:03:00.0000] <Philip`>
But in any case, it's usually more useful to know endianness at compile-time and not run-time

[07:04:00.0000] <gsnedders>
so reading an int into a short is endian dependent, assuming they're different sizes?

[07:04:01.0000] <Philip`>
since then you can have byte-swap macros that get replaced with noops

[07:04:02.0000] <Philip`>
so you should have a configure script or something to detect it and set a macro

[07:04:03.0000] <Philip`>
What do you mean by "reading into"?

[07:05:00.0000] <gsnedders>
*read the short representation of an int

[07:08:00.0000] <Philip`>
What do you mean by that?

[07:08:01.0000] <gsnedders>
Philip`: (But yes, it makes sense to do this in a saner way once I have all the setup to do that, which is pretty much going to be when I stop playing around with this code as a standalone application

[07:08:02.0000] <Philip`>
(Values have byte repesentations, not short representations)

[07:09:00.0000] <gsnedders>
/me wants to give up on IRC as this internet connection isn't working well

[07:09:01.0000] <gsnedders>
/me is on the WCML being all fancy and stuff in first-class, because Bank Holiday Monday => cheap upgrades.

[07:16:00.0000] <Philip`>
In practice, the solution that works 99.99% of the time is "#define IS_LITTLE_ENDIAN 1"

[07:18:00.0000] <gsnedders>
That's effectively what this code already is. :)

[07:32:00.0000] <karlcow>
http://glandium.org/blog/?p=2180

[07:33:00.0000] <woef>
About the article tag ... why is it the author's responsability to decide what content is syndicable?

[07:37:00.0000] <woef>
Or is the "syndication" part just a way of helping you decide whether to use an article tag?

[09:08:00.0000] <zcorpan>
hey, a spambug reported from http://html5.clearboth.org/spec !

[09:11:00.0000] <timeless>
anyone here have some time to consider some Web IDL?

[09:12:00.0000] <timeless>
X implements Y;

[09:12:01.0000] <timeless>
interface Y: Z;

[09:12:02.0000] <timeless>
Z implements P;

[09:12:03.0000] <timeless>
the contents of the interfaces are omited for brevity and the names have been changed to protect the guilty

[09:13:00.0000] <zcorpan>
the question is?

[09:14:00.0000] <timeless>
err lemme make sure i got that last line right

[09:14:01.0000] <timeless>
yeah, i did

[09:14:02.0000] <timeless>
well, first of all, is one really supposed to be using `implements` on interfaces which are inherited by other interfaces?

[09:14:03.0000] <timeless>
/me thought that `implements` was generally for objects and not interfaces

[09:16:00.0000] <timeless>
note to self: send feedback to list requesting a line "Generally you should use `dictionary Foo` instead of `[NoInterfaceObject] interface Foo`."

[09:17:00.0000] <zcorpan>
yeah implements might be a bad word

[09:18:00.0000] <zcorpan>
though i don't see what's wrong with using it on interfaces that are inherited

[09:21:00.0000] <timeless>
well

[09:21:01.0000] <timeless>
the problem is this:

[09:21:02.0000] <timeless>
ConcreteObject A;

[09:21:03.0000] <timeless>
A implements IFoo;

[09:21:04.0000] <timeless>
A implements IBar;

[09:21:05.0000] <timeless>
A implements IBaz;

[09:21:06.0000] <timeless>
ConcreteObject B;

[09:21:07.0000] <timeless>
B implements IFoo;

[09:22:00.0000] <timeless>
if someone decides that they have a `B` and want to make B's IFoo interface do more

[09:22:01.0000] <timeless>
they might say:

[09:22:02.0000] <timeless>
IFoo implements IMyFooMore;

[09:22:03.0000] <timeless>
expecting it to make B more powerful

[09:22:04.0000] <timeless>
but at the same time, it causes random effects on A

[09:22:05.0000] <timeless>
note that we had this experience in mozilla land

[09:22:06.0000] <timeless>
the rule which we try to enforce is this:

[09:23:00.0000] <timeless>
Only the owner of IFoo is allowed to extend IFoo until IFoo is frozen, at which point, IFoo /may/ but probably /should not/ be extended by others

[09:24:00.0000] <timeless>
roughly that translates into `You, owner of ConcreteObject X are free to use 'X implements IWhatever;'

[09:24:01.0000] <timeless>
but you are not free to use 'IStandardsBodyThing implements IWhatever;'.`

[09:57:00.0000] <dglazkov>
good morning, Whatwg!

[11:07:00.0000] <Philip`>
/me sees HTML5 supports &measuredangle;, which he will make good use of next time he wants to do some dangle measuring

[11:07:01.0000] <timeless>
heh

[11:07:02.0000] <timeless>
i need to send some feedback to some app vendors

[11:07:03.0000] <timeless>
they aren't unescaping entities :(

[11:07:04.0000] <timeless>
(it makes names of people and places really hard to read)

[11:11:00.0000] <ojan>
TabAtkins: ping

[11:25:00.0000] <zewt>
death to entities

[12:58:00.0000] <AryehGregor>
jgraham, this seems to not produce any output anymore.  http://dvcs.w3.org/hg/html/raw-file/tip/tests/submission/AryehGregor/reflection/reflection-onepage.html

[13:04:00.0000] <AryehGregor>
/me can't see why

[13:05:00.0000] <AryehGregor>
It's not just the number of tests, I don't think.

[13:05:01.0000] <timeless>
/me sighs

[13:05:02.0000] <timeless>
maybe i was reading the wrong spec?

[14:39:00.0000] <jgraham>
AryehGregor: So it doesn't. I will look tomorrow

[14:39:01.0000] <AryehGregor>
jgraham, thanks.

[14:47:00.0000] <dglazkov>
Accio Hixie!

[14:47:01.0000] <dglazkov>
Hixie: are you working on the dialog plumbing?

[14:52:00.0000] <ojan>
TabAtkins: in your flex rewriting this week...are you touching flex-flow?

[14:52:01.0000] <ojan>
TabAtkins: what's there now seems...complicated

[14:57:00.0000] <dglazkov>
hey fsamuel!

[14:58:00.0000] <dglazkov>
fsamuel: this is Hixie's natural habitat. You can see a glimpse of him eating a banana here. Be very, very still.

[14:59:00.0000] <fsamuel>
dglazkov, Hixie: Hi dglazkov, hi hixie! :)

[15:00:00.0000] <dglazkov>
Hixie: meet fsamuel. He's been very interested in building user-stylable select menu drop-downs. I wondered if there's an overlap with your thinking on <dialog>

[15:23:00.0000] <smaug____>
AryehGregor: ping

[15:23:01.0000] <AryehGregor>
smaug____, pong.

[15:23:02.0000] <smaug____>
AryehGregor: about http://html5.org/specs/dom-range.html#range-behavior-under-document-mutation and splitText

[15:24:00.0000] <smaug____>
what if textnode's parent is null

[15:24:01.0000] <AryehGregor>
Don't I handle that explicitly?

[15:24:02.0000] <AryehGregor>
"If old node's parent is not null, . . ."

[15:24:03.0000] <smaug____>
AryehGregor: you handle not-null case

[15:24:04.0000] <smaug____>
but where is the is null

[15:25:00.0000] <AryehGregor>
That's a corner case, so nothing special happens.  Regular range mutation rules apply.

[15:25:01.0000] <AryehGregor>
IIRC I used to handle it, but decided it was pointless to worry about it.

[15:25:02.0000] <smaug____>
hmm

[15:25:03.0000] <AryehGregor>
Often you can't do better than the default behavior anyway.

[15:25:04.0000] <AryehGregor>
Since the range can't span the two text nodes.

[15:25:05.0000] <smaug____>
AryehGregor: so what if you split text between the boundary points

[15:26:00.0000] <AryehGregor>
If the parent is null?  Exactly what Gecko does now: any boundary point after the offset you're splitting at is clamped to the end of the node.

[15:26:01.0000] <AryehGregor>
There's no special case then.

[15:27:00.0000] <smaug____>
ah

[15:27:01.0000] <smaug____>
ok

[15:27:02.0000] <smaug____>
a bit hard to read but ok

[15:27:03.0000] <smaug____>
I mean the draft

[15:28:00.0000] <AryehGregor>
What wording would you suggest?

[15:29:00.0000] <smaug____>
well, I was assuming that since there is the "If old node's parent is not null" case, there would be also the "is null" case

[15:30:00.0000] <AryehGregor>
Would you assume that when reading code?  If it doesn't give the case, it's not there.

[15:30:01.0000] <AryehGregor>
This is really just pseudocode.

[15:30:02.0000] <smaug____>
well, there is the other case defined

[15:30:03.0000] <smaug____>
"User agents must run the following steps whenever they execute splitText()"

[15:31:00.0000] <AryehGregor>
Right, just sometimes they do nothing.  Of course, the horrible way this is written is a hack to work around the fact that we haven't merged it into DOM Core yet.

[15:31:01.0000] <smaug____>
well, they need to update the boundary points when parent is null

[15:31:02.0000] <smaug____>
that is not "do nothing"

[15:32:00.0000] <AryehGregor>
It's "do nothing extra".

[15:32:01.0000] <smaug____>
the draft doesn't say that do something, but when splitText() is called, do something extra

[15:33:00.0000] <AryehGregor>
It says UAs have to run the steps as though they were written at the end of splitText().

[15:33:01.0000] <AryehGregor>
Earlier than that, splitText() replaced data and probably inserted a node.

[15:33:02.0000] <AryehGregor>
That already triggered different mutation rules.

[15:33:03.0000] <AryehGregor>
I have a note there that points this out explicitly: "Thus they must be executed after following the range mutation rules for any changes to data or insertions that the method caused."

[15:34:00.0000] <AryehGregor>
Also, one of the steps is "Let original offset equal what the boundary point's offset was before this splitText invocation replaced the data of node."  That doesn't make any sense if you didn't already have range mutation rules running beforehand.

[15:34:01.0000] <AryehGregor>
I'd ask for suggestions to make this clearer, but it will be moot once it's merged into DOM Core.

[15:35:00.0000] <smaug____>
a lot clearer would be to add just the case for null parent

[15:35:01.0000] <smaug____>
or at least a link

[15:36:00.0000] <AryehGregor>
A link to what?

[15:36:01.0000] <smaug____>
to the part of the spec which says what to do when parent is null

[15:36:02.0000] <AryehGregor>
You do nothing extra when the parent is null.

[15:37:00.0000] <smaug____>
this is about readability of the spec

[15:37:01.0000] <smaug____>
not correctness

[15:37:02.0000] <AryehGregor>
Readability is a mess because it's trying to hack extra steps onto DOM Core when DOM Core doesn't provide hooks for it.

[15:37:03.0000] <AryehGregor>
This will be solved when DOM Range is merged into DOM Core, so this will just be an extra step or two in the splitText() algorithm itself.


2011-08-30
[23:45:00.0000] <zcorpan>
does ie support <mark>?

[00:23:00.0000] <annevk>
Hixie, "<td> Whether the element is an editable" seems wrong

[00:31:00.0000] <annevk>
anything happened btw?

[00:32:00.0000] <annevk>
hmm, the new blog styling has a lot of whitespace

[00:36:00.0000] <zcorpan>
whatwg blog?

[00:36:01.0000] <annevk>
yeah

[00:49:00.0000] <zcorpan>
i don't see it being different from a few months ago?

[00:52:00.0000] <annevk>
well nimbu changed it with help from jgraham so you must be missing something or have something cached

[00:56:00.0000] <woef>
As far as analogies go, would you think it's fair to say <article> broadly matches back-end content types?

[00:58:00.0000] <annevk>
depends on what back-end content types are

[00:58:01.0000] <zcorpan>
i reloaded http://blog.whatwg.org/wp-content/themes/org.whatwg.awesome/style.css but still looks the same. i guess i'm missing something, then

[01:05:00.0000] <annevk>
compare with http://web.archive.org/web/20100823233819/http://blog.whatwg.org/

[01:08:00.0000] <zcorpan>
ah. ok

[01:10:00.0000] <woef>
annevk: well, typically article, product, blogpost, event ... structured content.

[01:10:01.0000] <woef>
Usually appears in a list-detail structure.

[01:16:00.0000] <annevk>
sounds about right

[01:17:00.0000] <woef>
Cool, a bit easier to understand than the usual "syndication" stuff imo :)

[01:29:00.0000] <annevk>
oh yes

[01:29:01.0000] <annevk>
http://trac.webkit.org/changeset/93951

[01:29:02.0000] <annevk>
weinig+++

[01:29:03.0000] <asmodai>
wow

[01:29:04.0000] <asmodai>
lovely diginotar fuckup

[01:31:00.0000] <annevk>
?

[01:31:01.0000] <asmodai>
Dutch certificate provider gave a wildcard certificate for *.google.com

[01:31:02.0000] <asmodai>
to some Iranian requester or something along those lines

[01:31:03.0000] <asmodai>
annevk: http://tweakers.net/nieuws/76444/iran-gebruikt-nederlands-certificaat-om-gmail-te-onderscheppen.html

[01:32:00.0000] <asmodai>
http://tweakers.net/nieuws/76445/browsermakers-geven-nieuwe-versies-uit-na-diginotar-blunder.html

[01:45:00.0000] <jgraham>
annevk: I don't think it has anything like excessive whitespace. Indeed ideally I would like it to look more like e.g. diveintomark with no sidebar

[01:48:00.0000] <annevk>
jgraham, feel free to modify the templates

[01:48:01.0000] <annevk>
archives and categories are pretty useless imo

[01:48:02.0000] <annevk>
so we would just need those sidebar links presented differently

[01:50:00.0000] <jgraham>
Yeah, it's a bit non-trivial to modify the actual templates without being able to download them

[01:51:00.0000] <annevk>
I can prolly get you ssh access if I remember all the details

[02:03:00.0000] <annevk>
zcorpan, did you introduce the prefix match for "xmlns" in setAttribute()?

[02:03:01.0000] <annevk>
zcorpan, see thread on www-dom

[02:06:00.0000] <zcorpan>
annevk: don't remember

[02:06:01.0000] <zcorpan>
annevk: seems bogus

[02:07:00.0000] <zcorpan>
annevk: so it must have been you :P

[02:13:00.0000] <annevk>
heh, will take the blame :)

[02:14:00.0000] <annevk>
even though I have no recollection of specifying these methods

[02:14:01.0000] <annevk>
maybe I should blame gsnedders, I hear he is away

[02:14:02.0000] <annevk>
http://www.mnot.net/blog/2011/08/24/distributed_hungarian_notation_doesnt_work didn't we ask mnot for an alternative to Sec-?

[02:16:00.0000] <annevk>
indeed

[02:16:01.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapps/2011JanMar/0623.html

[02:16:02.0000] <annevk>
lovely complaining on your blog while not providing solutions

[02:16:03.0000] <jgraham>
Or blame Ms2ger

[02:16:04.0000] <jgraham>
But do it on Google+ so he can't see

[02:16:05.0000] <Ms2ger>
jgraham, isn't that your job?

[02:17:00.0000] <Ms2ger>
Oh, and that prefix match was you

[02:18:00.0000] <annevk>
me?

[02:18:01.0000] <Ms2ger>
Yep

[02:19:00.0000] <annevk>
i still blame gsnedders

[02:19:01.0000] <annevk>
added a comment to mnot's post

[02:24:00.0000] <zcorpan>
this is why we shouldn't have multiple editors. makes it harder to place blame.

[02:27:00.0000] <Ms2ger>
zcorpan, that's what hg is for :)

[02:29:00.0000] <zcorpan>
what you mean 'the tools will save us'? silly

[03:32:00.0000] <jgraham>
annevk: Isn't document.contains(node) different to node.inDocument in the case of multiple documents?

[03:33:00.0000] <annevk>
yes

[03:34:00.0000] <annevk>
I wonder when you want to know the latter but not the former

[03:35:00.0000] <jgraham>
Well that seems like an important criterion in determining the right API to add

[03:35:01.0000] <gsnedders>
annevk: IT WASN'T ME.

[03:35:02.0000] <gsnedders>
/me looks innocent

[03:36:00.0000] <annevk>
jgraham, yes

[03:37:00.0000] <Ms2ger>
node.ownerDocument.contains(node), then?

[03:38:00.0000] <annevk>
time to get some food

[03:41:00.0000] <zcorpan>
maybe we want node.topMostAncestor instead

[03:42:00.0000] <annevk>
Ms2ger, btw, the DocumentType.ownerDocument change breaks Acid3

[03:42:01.0000] <Ms2ger>
Time to fix Acid3, then

[03:42:02.0000] <roc>
acid3--

[03:42:03.0000] <annevk>
agreed

[03:42:04.0000] <annevk>
changing Acid3 does not seem very simple however :(

[03:43:00.0000] <Ms2ger>
roc, acid2?

[03:43:01.0000] <annevk>
but ignoring it for now probably works and at some point it will have to adapt or die

[03:43:02.0000] <Ms2ger>
Hixie, please fix acid3, thanks

[03:54:00.0000] <annevk>
renameNode semantics sound more like mv in shell

[03:54:01.0000] <annevk>
moveNode

[03:54:02.0000] <annevk>
well, moveElement

[03:55:00.0000] <annevk>
actually, nm

[04:01:00.0000] <asmodai>
annevk: gets better, looks like that root authority has now been hacked by Iranians

[04:01:01.0000] <asmodai>
annevk: http://www.f-secure.com/weblog/archives/00002228.html

[04:22:00.0000] <smaug____>
Ms2ger: what is "node document"

[04:23:00.0000] <smaug____>
strangely named

[04:23:01.0000] <Ms2ger>
/me blames annevk 

[04:23:02.0000] <Ms2ger>
It's just ownerDocument

[04:23:03.0000] <smaug____>
Ms2ger: and, createDocument doesn't work atm

[04:23:04.0000] <smaug____>
"If the doctype's node document is not null, throw a WRONG_DOCUMENT_ERR exception and terminate the overall set of steps."

[04:24:00.0000] <Ms2ger>
Didn't I change that?

[04:24:01.0000] <Ms2ger>
/me checks

[04:25:00.0000] <Ms2ger>
Er, I didn't

[04:27:00.0000] <smaug____>
Ms2ger: so what is your plan? just remove steps 1 and 2

[04:47:00.0000] <annevk>
node document is named after XMLHttpRequest document

[04:52:00.0000] <smaug____>
Ms2ger: so, what is the plan? Allow using dtd nodes even if they are already bound to some other document?

[04:52:01.0000] <Ms2ger>
Auto-adopt, like we do for all nodes

[04:53:00.0000] <smaug____>
k

[04:53:01.0000] <zcorpan>
but doctypes have special powers!!

[05:02:00.0000] <annevk>
and soon they have none

[05:03:00.0000] <annevk>
it's a shame browsers support document.doctype at all

[05:06:00.0000] <Ms2ger>
Acid3?

[05:06:01.0000] <annevk>
I think for Opera it was Acid3 yes

[05:07:00.0000] <smaug____>
yeah, I'm just trying to load acid3 to test whether these changes break it

[05:07:01.0000] <smaug____>
but the site seems to be down

[05:07:02.0000] <annevk>
Acid3 has a check for ownerDocument being null after createDocumentType()

[05:08:00.0000] <smaug____>
interesting

[05:08:01.0000] <smaug____>
someone is going to blame me for breaking it in FF

[05:08:02.0000] <smaug____>
I do remember that acid3 has seemingly random tests about dtd and createDocument

[05:08:03.0000] <annevk>
I am planning on writing a blog post at some point about the wrongs of Acid3

[05:09:00.0000] <smaug____>
we need to change acid3

[05:09:01.0000] <annevk>
might salute you for not caring (and IE for not caring about SMIL)

[05:09:02.0000] <smaug____>
http://acid3.acidtests.org/ is very very slow

[05:10:00.0000] <jgraham>
smaug____: Hixie's server is having issues

[05:10:01.0000] <smaug____>
again :(

[05:13:00.0000] <smaug____>
would be interesting if acid3 was changed so that all the current browsers start to fail in a test

[05:14:00.0000] <jgraham>
Would be nice if ACID3 was retired as a failed experiment (maybe)

[05:15:00.0000] <smaug____>
yes, that would be good

[05:15:01.0000] <hsivonen>
wiki.whatwg.org is apparently on the same server that's having issues :-(

[05:15:02.0000] <annevk>
Hixie uses one server

[05:45:00.0000] <gsnedders>
How frickin' hard can it be to return a broken laptop: obviously very.

[05:47:00.0000] <hsivonen>
Are Apple and Opera zapping the DigiNotar root?

[06:05:00.0000] <smaug____>
Hixie: could you remove assertEquals(doctype.ownerDocument, null, "doctype's ownerDocument was wrong after creation"); from acid3. Per DOM Core dtd nodes have ownerDocument

[06:10:00.0000] <Kellen`>
hsivonen: is there any progress towards a validator.nu library which can be used by external projects?

[06:25:00.0000] <hsivonen>
Kellen`: no real progress. it's part of the long-term plan, but there's other stuff ahead of it in my work queue

[06:26:00.0000] <hsivonen>
Kellen`: if you need it now, I suggest writing a mock servlet request/response objects for running VerifierServletTransaction in a holder that makes it think it lives in a servlet container

[06:26:01.0000] <hsivonen>
s/objects/object pair/

[06:27:00.0000] <Kellen`>
hsivonen: that's one possibility. i need to validate a few gigs of data so it seemed like it could be a bottleneck to send HTTP requests everywhere

[06:32:00.0000] <annevk>
Ms2ger, using concept-node-pre-insert is probably wrong for a bunch of cases

[06:32:01.0000] <Ms2ger>
Tell me about it

[06:32:02.0000] <annevk>
Ms2ger, my idea was that pre-insert would also do modification listener stuff

[06:32:03.0000] <Kellen`>
hsivonen: are you  open to contributions towards making the validator library or would that cause extra problems for you?

[06:33:00.0000] <annevk>
I guess we can cross that bridge later... when it is more clear how modifications listeners are going to work

[06:33:01.0000] <Ms2ger>
Which is fine, because you can't add listeners to a document before it's created, no?

[06:34:00.0000] <annevk>
hmm

[06:34:01.0000] <hsivonen>
Kellen`: I'm open to contributions (under the MIT license) but preferably with design discussion beforehand and not out of the blue

[06:38:00.0000] <hsivonen>
oh well. WebVTT didn't evolve the way I wanted, but at least there's forward motion: https://bugs.webkit.org/show_bug.cgi?id=64132

[06:39:00.0000] <Kellen`>
hsivonen: okay thanks. I'll take a look at the servlet code and see how I feel about it. I'll get back to you before/if I do any serious development work on it.

[06:47:00.0000] <hsivonen>
Kellen`: ok. thanks

[07:56:00.0000] <annevk>
Ms2ger, how is the Attr migration going?

[07:57:00.0000] <Ms2ger>
Nothing much happening

[08:15:00.0000] <annevk>
http://www.w3.org/wiki/TPAC2011

[08:17:00.0000] <jgraham>
Wow

[08:17:01.0000] <jgraham>
That's quite some progress

[08:21:00.0000] <jgraham>
Maybe should propose something testing related if I get to attend.

[08:25:00.0000] <jgraham>
(not sure xactly what has the right level of appeal and isn't covered by the testing IG meeting or so)

[08:25:01.0000] <jgraham>
+e

[08:33:00.0000] <karlcow>
even far to be a niche problem… it's one of the first questions I get when talking about mediaqueries for background images. "OH cool!!! How can I use it for img too?"

[08:36:00.0000] <zewt>
different images for different screen resolutions? uh, no...

[08:37:00.0000] <zewt>
for different screen sizes vs. dpi, perhaps

[08:37:01.0000] <Ms2ger>
<img lowsrc=foo>

[08:39:00.0000] <zewt>
(yuck)

[08:39:01.0000] <karlcow>
lowsrc is not enough

[08:40:00.0000] <karlcow>
you need to be able to select the screen size.

[08:40:01.0000] <zewt>
screen size is meaningless without dpi as well

[08:41:00.0000] <zewt>
(and even then is only meaningful on mobile browsers, not on anything where the window size is variable)

[08:41:01.0000] <zewt>
(technically the window size is variable on android's browser, but)

[08:42:00.0000] <karlcow>
/me has a tendency to resize his windows on desktop to get the mediaqueries for small screens. Better sites usually.

[08:43:00.0000] <zewt>
i always use maximized browsers, myself

[08:44:00.0000] <karlcow>
zewt: never for me. :)

[08:46:00.0000] <zewt>
heh, this is also one of the rare cases where progressive jpeg could have been useful, if it was possible for browsers to request the number of passes to download

[08:47:00.0000] <zewt>
so it could download however much is needed for the resolution it's actually displaying at (and then resume the download to get more, if that changes)

[08:50:00.0000] <annevk>
karlcow, if you are already talking in the context of images it's not surprising the question comes up

[08:50:01.0000] <karlcow>
annevk: not sure how does it relate to what we are saying.

[08:50:02.0000] <karlcow>
contextual images is a **need**, not a niche problem, and a growing frustration from designers.

[08:50:03.0000] <karlcow>
it is already implemented and well used for background-images in CSS

[08:50:04.0000] <karlcow>
people want it for img element

[08:51:00.0000] <annevk>
what's wrong with using a high-resolution image?

[08:51:01.0000] <karlcow>
It has nothing to do with the resolution

[08:53:00.0000] <karlcow>
for example, contextual images are an image where you would have a "Logo + company name" (let's say a 400 pixels banner) then when you reach a smaller screen, you want to display only the logo without the company name, (like a 100px square logo)

[08:53:01.0000] <karlcow>
It is just an example.

[08:54:00.0000] <karlcow>
a thumbnail of an image at a small size is not necessary the full image but a crop of the bigger image.

[08:54:01.0000] <zewt>
i'd think you'd just have two images in that case, toggled on mutually exclusively with css, not a single image with src magic

[08:54:02.0000] <karlcow>
like for example a rectangle image and you show a square crop at smaller size

[08:54:03.0000] <karlcow>
etc etc

[08:54:04.0000] <karlcow>
zewt: you are not forced to limit yourself to two images

[08:55:00.0000] <zewt>
yes, you can have three or four images, again with visibility controlled with css :)

[08:55:01.0000] <karlcow>
because you can precisely target a very large screen, a tablet screen, a giant billboard, a desktop screen, etc.

[08:55:02.0000] <zewt>
(not ideal, since it won't work well if the stylesheet isn't loaded, but "css is optional" doesn't exactly work in practice very often anyway heh)

[08:55:03.0000] <shetech>
/me thinks this would be great for targeting mobile vs. full screen

[08:58:00.0000] <zewt>
karlcow: what i'm saying is, why do this with a complex <img> mechanism, instead of implementing it with multiple, distinct <img> nodes, controlling visibility on each (whether with css or otherwise)

[08:59:00.0000] <annevk>
<img> is already way complex, adding some conditional resource loading on top of it used by a couple of authors is not worth it

[09:00:00.0000] <jgraham>
I think characterising it as :a couple of authors" is not very fair

[09:00:01.0000] <karlcow>
yup

[09:00:02.0000] <shetech>
jgraham+1

[09:00:03.0000] <woef>
Is Apple already conforming to the media="handheld" on css files for it's mobile devices?

[09:01:00.0000] <zewt>
(what use would that be? a "handheld" device might have a display in the same class as a desktop pc)

[09:01:01.0000] <annevk>
jgraham, maybe not, is <source> a success?

[09:01:02.0000] <zewt>
(or a laptop, more practically)

[09:01:03.0000] <woef>
zewt: so it's out of the spec?

[09:02:00.0000] <zewt>
i don't know anything about it; i'm asking since it sounds useless to me on first impression

[09:02:01.0000] <annevk>
woef, you should use media queries instead

[09:02:02.0000] <annevk>
handheld is an old concept, nearly obsolete

[09:02:03.0000] <zewt>
good :)

[09:03:00.0000] <woef>
But media queries are based on browser window size right?

[09:04:00.0000] <woef>
If I reduce the size of my browse I don't necessarily want  to end up with some iPad-enhanced css

[09:04:01.0000] <woef>
*browser

[09:04:02.0000] <woef>
Probably should say *degraded

[09:04:03.0000] <annevk>
they select on capabilities

[09:04:04.0000] <annevk>
viewport size being one of them

[09:06:00.0000] <karlcow>
zewt see http://dev.opera.com/articles/view/an-introduction-to-meta-viewport-and-viewport/

[09:07:00.0000] <woef>
"handheld" just sounded like a good "crap for phones" alternative to me :)

[09:07:01.0000] <zewt>
(i know about meta viewport, doesn't seem relevant here though)

[09:07:02.0000] <karlcow>
and http://dev.w3.org/csswg/css-device-adapt/#the-lsquoresolutionrsquo-property

[09:08:00.0000] <karlcow>
oops I meant for woef

[09:09:00.0000] <zewt>
heh, meta viewport was pretty nastily inconsistent between iOS and android last i used it (three cheers for vendors making crap up as they go, because it's the 90s, right?)

[09:09:01.0000] <karlcow>
about mediaqueries and selection on capabilities

[09:09:02.0000] <karlcow>
/me is impatient to see that implemented

[09:09:03.0000] <karlcow>
@viewport {

[09:09:04.0000] <karlcow>
  width: device-width;

[09:09:05.0000] <karlcow>
}

[09:10:00.0000] <karlcow>
instead of having to tweak the HTML code with meta

[09:10:01.0000] <zewt>
and fun when you need different versions for ios and android

[09:11:00.0000] <zewt>
ended up having to use document.write to handle that :|

[09:11:01.0000] <zewt>
(the only time i can ever remember having had to use that)

[09:13:00.0000] <woef>
Btw, how I came to this question:

[09:13:01.0000] <woef>
I was asked to remove a bg-image for mobile (because of bandwidth issues)

[09:13:02.0000] <woef>
So screen-size is really not an interesting capability here.

[09:14:00.0000] <zewt>
@media 2g? :P

[09:16:00.0000] <zewt>
that isn't mobile-specific, though--there are painfully slow links that aren't mobile

[09:17:00.0000] <zewt>
(ever try viewing a site in China during their prime time?)

[09:17:01.0000] <zewt>
(at least in the US)

[09:17:02.0000] <woef>
As a broad guess, it's probably not bad though.

[09:17:03.0000] <zewt>
oh it's much much worse

[09:19:00.0000] <zewt>
and conversely, many people on mobile have very fast links--it'd be reducing the site unnecessarily for them

[09:19:01.0000] <jgraham>
annevk: <source> seems to be a success, yes. Although it addresses a quite different problem

[09:19:02.0000] <zewt>
it's the wrong metric for adjusting for bandwidth capability

[09:19:03.0000] <woef>
Probably, but it's better than nothing though.

[09:20:00.0000] <woef>
Which is pretty much the perfect summary of 20 years of webdev :p

[09:20:01.0000] <annevk>
jgraham, where is it used?

[09:20:02.0000] <annevk>
jgraham, <source> takes a media query

[09:21:00.0000] <gsnedders>
zewt: Equally, different resources on the same page could be fetched over different connection types

[09:21:01.0000] <zewt>
that's much more niche

[09:22:00.0000] <zewt>
mobile could in theory do some stuff over wifi and some over 3g, but i don't know why they would

[09:28:00.0000] <gsnedders>
zewt: It's not that niche if you're on, e.g., a bus, or a train, or in a car.

[09:28:01.0000] <gsnedders>
zewt: Basically any reasonably quick-moving phone can hit that issue.

[09:28:02.0000] <zewt>
but you're still only using a single connection type at a time

[09:30:00.0000] <gsnedders>
3G -> 2G is possible, esp. seeming sites often cause multiple connections to the server to be created

[11:03:00.0000] <timeless>
zewt: my BlackBerry probably does some stuff over WiFi and some stuff over cellular

[11:04:00.0000] <timeless>
i think it can choose to send corporate VPN stuff over cellular if it needs to

[11:34:00.0000] <gsnedders>
hsivonen: No, we're not [revokeing the DigiNotar root yet], just relying on protection against blocked revocation lists.

[11:36:00.0000] <timeless>
gsnedders: how do you do that?

[11:36:01.0000] <timeless>
if the ocsp responder is down, what then?

[11:36:02.0000] <gsnedders>
timeless: I don't know the detail

[11:36:03.0000] <timeless>
/me hasn't checked the crls they issue to see how long they're valid

[11:36:04.0000] <timeless>
(but generally one can replay an old crl)

[11:37:00.0000] <hsivonen>
I thought Moxie Marlinspike had shown that CRLs don't work

[11:37:01.0000] <timeless>
they don't

[11:37:02.0000] <hsivonen>
or don't work with the defaults of non-Opera browsers? does Opera have different behavior?

[11:37:03.0000] <timeless>
and yeah

[11:37:04.0000] <timeless>
oh, i misread

[11:38:00.0000] <timeless>
gsnedders really did say they were relying on CRLs and not OCSP

[11:38:01.0000] <gsnedders>
timeless: I did?

[11:38:02.0000] <timeless>
`blocked revocation lists` implies `revocation lists` implies `crl`

[11:38:03.0000] <gsnedders>
/me notes it's been a long time since he's looked at a lot of this

[11:38:04.0000] <timeless>
as opposed to OCSP which isn't really a list

[11:39:00.0000] <gsnedders>
/me looks up OCSP, and remembers how that works

[11:39:01.0000] <gsnedders>
Okay, I was grouping OCSP and CRLs as the same thing.

[11:40:00.0000] <timeless>
ok, they're idiots

[11:40:01.0000] <timeless>
the CRL they have right now http://service.diginotar.nl/crl/root/latestCRL.crl next updates 2/3/2012

[11:40:02.0000] <timeless>
wtf

[11:41:00.0000] <timeless>
it also claims it was last updated 2/3/2011

[11:41:01.0000] <gsnedders>
timeless: AFAIK if the OCSP responder is down, we treat the the site as untrusted.

[11:41:02.0000] <gsnedders>
Because we can't guarantee the cert is valid.

[11:43:00.0000] <timeless>
/me kinda wishes this stuff was documented in some easy to review place

[11:43:01.0000] <timeless>
(one place for all browsers)

[11:43:02.0000] <timeless>
(security through distributed obscurity)

[11:44:00.0000] <gsnedders>
timeless: default-secure for OCSP being down seems bad, though

[11:45:00.0000] <timeless>
oh yeah, i'm not arguing that

[11:45:01.0000] <timeless>
i just wish i could look at a single web page to get a list of what each browser does for this stuff

[11:45:02.0000] <timeless>
i shouldn't have to spend 10 minutes reversing how each browser behaves

[11:45:03.0000] <timeless>
(nor should i have to waste anyone else's time for each browser)

[11:45:04.0000] <gsnedders>
as since revoked the digital certificate. This is not a Microsoft security vulnerability; however, the certificate potentially affects...

[11:45:05.0000] <gsnedders>
21 minutes ago · Share

[11:45:06.0000] <gsnedders>
Korin Tom Queen Mine's Opera. Dunno how we're affected.

[11:45:07.0000] <gsnedders>
19 minutes ago

[11:46:00.0000] <gsnedders>
Sam Vennall Might wanna double check your security certificates

[11:46:01.0000] <gsnedders>
18 minutes ago

[11:46:02.0000] <gsnedders>
Sam Vennall The one you should be worried about is DigiNotar

[11:46:03.0000] <gsnedders>
18 minutes ago

[11:46:04.0000] <gsnedders>
Korin Tom Queen Nothing seems to be an issue. Geoffrey Sneddon?

[11:46:05.0000] <gsnedders>
15 minutes ago

[11:46:06.0000] <gsnedders>
Axel Theodor Rolandsson Klingberg Umm, Chrome?

[11:46:07.0000] <timeless>
heh

[11:46:08.0000] <gsnedders>
14 minutes ago

[11:46:09.0000] <gsnedders>
Daniel Callander http://www.reddit.com/r/programming/comments/jya5h/gmailcom_being_mitmd_by_iran_using_this/

[11:46:10.0000] <gsnedders>
Google's domain is being man-in-the-middle'd by the Iranian Government. This doesn't affect us, however. Only people connecting from around Iran.

[11:46:11.0000] <gsnedders>
But if you have a root certifcate for DigiNotar you should probably delete it.Gmail.com being MITM'd by Iran using this certificate : programming

[11:46:12.0000] <gsnedders>
www.reddit.com

[11:46:13.0000] <gsnedders>
reddit: the front page of the internet

[11:46:14.0000] <gsnedders>
14 minutes ago

[11:46:15.0000] <gsnedders>
Sam Vennall Fair enough.

[11:46:16.0000] <gsnedders>
13 minutes ago

[11:46:17.0000] <gsnedders>
Sam Vennall You're fine. Google caught it

[11:46:18.0000] <gsnedders>
13 minutes ago

[11:46:19.0000] <gsnedders>
Michael Dyson No problems. I have a certificate pre-screener for FireFox xD.

[11:46:20.0000] <gsnedders>
12 minutes ago

[11:46:21.0000] <gsnedders>
Uh,. shit.

[11:46:22.0000] <gsnedders>
Totally didn't just paste from the wrong clipboard.

[11:46:23.0000] <gsnedders>
Sorry.

[11:47:00.0000] <gsnedders>
http://my.opera.com/securitygroup/blog/2011/08/30/when-certificate-authorities-are-hacked-2

[11:47:01.0000] <timeless>
it was amusing

[11:47:02.0000] <gsnedders>
That's the right clipboard.

[11:48:00.0000] <timeless>
so..

[11:48:01.0000] <timeless>
minor problem

[11:48:02.0000] <timeless>
revocation url *usually* means CRL

[11:48:03.0000] <timeless>
and their CRL was last updated in February

[11:48:04.0000] <timeless>
at best they've updated their OCSP server

[11:48:05.0000] <gsnedders>
Yeah, it's not the best of posts.

[11:48:06.0000] <gsnedders>
I'd trust if Sigbjørn claims it's fine that they have updated OCSP.

[11:48:07.0000] <timeless>
if you guys means OCSP responder

[11:49:00.0000] <timeless>
.. it'd be good if your post was updated to clarify the text to use "responder"

[11:49:01.0000] <timeless>
instead of "revocation url"

[11:49:02.0000] <gsnedders>
timeless: email sigbjorn⊙oc

[11:49:03.0000] <timeless>
ok

[11:49:04.0000] <timeless>
/me sighs and pulls up blackberry-gmail

[11:50:00.0000] <gsnedders>
/me is about to head off

[11:59:00.0000] <timeless>
gsnedders: email sent... kinda

[11:59:01.0000] <timeless>
/me waits for gmail.java to finish "Sending message..."

[12:00:00.0000] <timeless>
and gah, your my site requires me to log in

[12:00:01.0000] <timeless>
e-way-too-painful

[12:01:00.0000] <gsnedders>
timeless: Should be possible to post an anon comment on myopera

[12:01:01.0000] <timeless>
Write a comment  You must be logged in to write a comment. If you're not a registered member, please sign up.

[12:01:02.0000] <timeless>
i could see if bugmenot can help..

[12:01:03.0000] <gsnedders>
It used to be possible to post anon comments, at least

[12:02:00.0000] <zewt>
heh sites that expect you to sign up for an account to leave a comment are sort of, uh, ... no, heh

[12:02:01.0000] <timeless>
i wouldn't mind so much if i could use google/facebook creds

[12:02:02.0000] <timeless>
maybe

[12:02:03.0000] <timeless>
i still wouldn't do it

[12:02:04.0000] <timeless>
since i don't share those creds w/ work

[12:02:05.0000] <timeless>
but i wouldn't mind so much!

[12:02:06.0000] <zewt>
not as bad as sites that try to make you sign up for an account for no reason at all

[12:03:00.0000] <zewt>
as if people signing up for accounts is an end by itself

[12:04:00.0000] <gsnedders>
zewt: But it is!

[12:04:01.0000] <zewt>
even better, when combined with arbitrary blacklists

[12:04:02.0000] <zewt>
eg. "you can't sign up with a free email"

[12:04:03.0000] <timeless>
http://tech.slashdot.org/comments.pl?sid=2403472&cid=37247406

[12:06:00.0000] <zewt>
win7's file manager is so much worse than xp's, i don't expect it'll recover for many windows generations :|

[12:07:00.0000] <zewt>
applications seem to reach a peak, at which point engineering knows they don't really need further changing--at which point marketing takes over to force further changes, causing a bottomless downwards spiral

[12:09:00.0000] <timeless>
darn, i learned about something useful from /. : http://www.officelabs.com/projects/searchcommands/Pages/default.aspx

[12:09:01.0000] <timeless>
what's the world coming to?

[12:10:00.0000] <zewt>
heh

[12:10:01.0000] <zewt>
a ui that needs a search engine to find menu items is ...

[12:10:02.0000] <timeless>
actually

[12:11:00.0000] <timeless>
have you used win7's start menu?

[12:11:01.0000] <zewt>
yeah it's terrible

[12:11:02.0000] <timeless>
/me loves being able to type into the text box that appears at its bottom

[12:11:03.0000] <timeless>
it's like a command line

[12:11:04.0000] <timeless>
and you can do something similar in control panel

[12:11:05.0000] <timeless>
very handy

[12:11:06.0000] <zewt>
the searching is okay but they threw away a very heavily polished menu ui that we had a decade and a half of experience and muscle memory using

[12:12:00.0000] <timeless>
/me nods

[12:12:01.0000] <timeless>
i've been using windows since 3.0

[12:12:02.0000] <zewt>
shows a serious and complete lack of caring about existing users

[12:12:03.0000] <timeless>
(i've used older versions, but i wasn't using it)

[12:12:04.0000] <timeless>
actually

[12:12:05.0000] <timeless>
they were  very careful

[12:12:06.0000] <timeless>
they studied users and discovered what they could change

[12:12:07.0000] <timeless>
and what they couldn't

[12:12:08.0000] <timeless>
they're *incredibly* careful not to change keystrokes

[12:12:09.0000] <timeless>
because power user muscle memory there is very important

[12:12:10.0000] <zewt>
... but they threw away the entire start menu interface, which is fundamental muscle memory

[12:13:00.0000] <timeless>
e.g. they had yes/no/cancel dialogs when you closed an unsaved document in notepad/paint/etc.

[12:13:01.0000] <timeless>
the dialog now says save/don't save/cancel

[12:13:02.0000] <zewt>
they changed the post display-mode-change confirmation from "yes/no" to like "keep/cancel", breaking the &y shortcut

[12:13:03.0000] <zewt>
pressing backspace in file explorer to go up a level doesn't always work (sometimes it acts like "back")

[12:13:04.0000] <timeless>
but in the old dialog you could use 'n' to mean "no", and you can still use "n" to mean "don't save"

[12:13:05.0000] <timeless>
yes, they broke the "yes" case

[12:13:06.0000] <timeless>
but that's ok, for the yes case, you could have been using <space>

[12:14:00.0000] <timeless>
(and maybe you were)

[12:14:01.0000] <zewt>
what you could have been using isn't really relevant

[12:14:02.0000] <timeless>
worst case there, you hear the beep and press space

[12:14:03.0000] <zewt>
my muscle memory is what i'm using, not what i might have been :)

[12:14:04.0000] <zewt>
(that one's not a major case, just a random example)

[12:14:05.0000] <timeless>
well, they studied large numbers of users to determine which ones people used in high concentrations

[12:14:06.0000] <timeless>
there are some amusing stories about which ones they could change and which they coudn't

[12:14:07.0000] <zewt>
i don't care what "lots of people use", i care about what i use, heh

[12:15:00.0000] <timeless>
you're greedy :)

[12:15:01.0000] <zewt>
win7's file explroer is also mysteriously missing the window icon (system menu icon), which is annoying since i use that to close windows

[12:15:02.0000] <zewt>
(it's still there--you can still double-click where the system menu should be to close the window--but it's not visible)

[12:15:03.0000] <timeless>
yeah

[12:15:04.0000] <zewt>
strange inconsistency since everything else has it

[12:15:05.0000] <timeless>
i think that's because they wanted to get rid of the window proxy

[12:15:06.0000] <timeless>
windows explorer used to have a window proxy there

[12:15:07.0000] <timeless>
which was "special"

[12:16:00.0000] <timeless>
and a rarely used feature

[12:16:01.0000] <zewt>
file explorer has some of the worst breakages, overall

[12:16:02.0000] <timeless>
anyway, i'll be glad to see w7's explorer replaced by the ribbon ui

[12:16:03.0000] <zewt>
eg. no keyboard shortcuts at all for copy/save "overwrite file", etc. prompts, short of tediously tabbing around

[12:16:04.0000] <timeless>
really no shortcuts there?

[12:16:05.0000] <timeless>
/me tries to remember how to get that dialog easily

[12:17:00.0000] <zewt>
previously it was just &yes or &all

[12:18:00.0000] <timeless>
there's &Do this for the next 1 conflicts

[12:18:01.0000] <timeless>
and &Skip

[12:18:02.0000] <timeless>
note that i find the <for the next N conflicts> incredibly valuable

[12:18:03.0000] <timeless>
and i was complaining about the lack of that style feature in Nokia's Maemo platform

[12:18:04.0000] <zewt>
but there's no shortcut for the actual confirmation

[12:18:05.0000] <timeless>
(backup and restore was terrible there

[12:18:06.0000] <timeless>
you could have 100-800 conflicts

[12:18:07.0000] <timeless>
and no indication

[12:18:08.0000] <timeless>
plus no useful grouping!

[12:18:09.0000] <timeless>
)

[12:20:00.0000] <zewt>
no way to get from the location bar to the file list (pressing tab goes to a bunch of other things before hitting the file list again; it even goes to the column headers first); randomly changing view modes regardless of settings; mouseover highlighting is very confusing (two highlighted rows); keyboard prefix-searching is broken (doesn't reset state properly like it used to); i'll stop here :)

[12:20:01.0000] <timeless>
for the conflict dialog, i think they made the concious decision that you actually need to read through the dialog and make a real choice

[12:20:02.0000] <timeless>
which isn't a bad approach

[12:21:00.0000] <timeless>
yes it hurts power users (who could use copy /y or whatever if they were really power users)

[12:21:01.0000] <timeless>
but it protects average users (who could really hurt themselves)

[12:21:02.0000] <zewt>
uh i'm not going to drop to a shell to bypass broken dialogs heh

[12:21:03.0000] <zewt>
making things deliberately cumbersome is never ever ever an acceptable ui policy

[12:21:04.0000] <timeless>
/me disagrees

[12:21:05.0000] <timeless>
security behavior is often that way

[12:21:06.0000] <zewt>
firefox's ssl exemption dialog is one of the worst ui's i've seen in a long time

[12:21:07.0000] <timeless>
have you seen mozilla's xpinstall dialog?

[12:22:00.0000] <timeless>
we intentionally force you to wait 3-5s

[12:22:01.0000] <zewt>
yeah that's horrible and one of the first bits of nonsense i turn off

[12:22:02.0000] <timeless>
because we have no choice

[12:22:03.0000] <timeless>
if you're being gamed by being tricked into playing a game

[12:22:04.0000] <timeless>
you can easily fill those 3s with keystrokes

[12:22:05.0000] <timeless>
some of which go to us instead of the web page

[12:22:06.0000] <timeless>
and then you're screwed

[12:23:00.0000] <timeless>
fwiw, from the address bar, f6 twice is pretty good

[12:23:01.0000] <timeless>
in explorer

[12:23:02.0000] <timeless>
instead of using tab

[12:24:00.0000] <zewt>
the ssl exemption dialog is frustrating since it's training every user in the world not only to click through one-off prompts, but as many prompts as necessary to make it go awayt

[12:24:01.0000] <zewt>
(okay, every firefox user :)

[12:24:02.0000] <timeless>
well

[12:24:03.0000] <timeless>
the hope is that users will instead *maybe* realize there's actually something *wrong* there

[12:24:04.0000] <zewt>
it's as if someone thinks making the flawed-but-best-we've-got ssl quirks more annoying will improve the system

[12:25:00.0000] <timeless>
i'm open to other suggestions

[12:25:01.0000] <zewt>
don't make things willfully annoying :)

[12:25:02.0000] <timeless>
offhand, i think an alternative is probably sending users to a search page which searches for the hash of the server's cert

[12:25:03.0000] <AryehGregor>
Which a MITM will also control.

[12:25:04.0000] <timeless>
and a way for people to add a comment "this bit me, i don't like <write vendor name here>"

[12:25:05.0000] <zewt>
that won't help users who need help

[12:26:00.0000] <timeless>
AryehGregor: it'd be pinned SSL

[12:26:01.0000] <timeless>
zewt: why not?

[12:26:02.0000] <AryehGregor>
In that case, why force the user to search for it?

[12:26:03.0000] <AryehGregor>
Just use something like Perspectives.

[12:27:00.0000] <zewt>
regular users don't know what a hash is, or a certificate, and have no idea what the entire certificate dialog (in any incarnation) means

[12:27:01.0000] <AryehGregor>
Or, like, fix SSL so we don't have so many sites with broken certs.

[12:27:02.0000] <timeless>
sorry

[12:27:03.0000] <timeless>
lemme define "search"

[12:27:04.0000] <timeless>
to mean "where a server takes your input and provides results based on it"

[12:27:05.0000] <timeless>
that could very well *be* perspectives

[12:27:06.0000] <zewt>
aryeh: i really wish we could have "shttp", which is just https without any certificates--just to give people who don't want to pay for a cert the ability to use ssl without causing certificate failures

[12:28:00.0000] <timeless>
/me has yet to see an shttp which wasn't just a simpler way to do MITM

[12:28:01.0000] <zewt>
simpler?

[12:28:02.0000] <timeless>
yes?

[12:28:03.0000] <zewt>
simpler how? heh

[12:28:04.0000] <timeless>
w/ today's MITM, i have to spend time finding a stupid CA

[12:29:00.0000] <timeless>
then i have to trick it into giving me a CERT

[12:29:01.0000] <timeless>
then i have to feed that cert to my server and mitm someone

[12:29:02.0000] <zewt>
shttp wouldn't allow MITM against https sites--that's the entire reason it'd be a different protocol name

[12:29:03.0000] <timeless>
wouldn't shttp allow mitm shttp?

[12:29:04.0000] <zewt>
yes

[12:29:05.0000] <timeless>
?

[12:30:00.0000] <timeless>
and isn't that simpler than mitm https

[12:30:01.0000] <timeless>
/me dismisses a periodic "i don't like your mail server cert" dialog

[12:30:02.0000] <zewt>
shttp should be compared against http, not https

[12:30:03.0000] <timeless>
(the dialog pops up from an otherwise invisible service, because the service is stupid)

[12:31:00.0000] <zewt>
perhaps "shttp" is a bad name, since it should absolutely not be viewed as an alternative to https ("real" security), but as an improvement on http (preventing passive sniffing)

[12:31:01.0000] <timeless>
(the server seems to be round robin with two distinct certificates, which really annoys the service)

[12:31:02.0000] <zewt>
i wonder if spdy is going anywhere ... that encrypts everything anyway, so it's equivalently secure

[12:32:00.0000] <zewt>
havn't heard much about it lately

[12:35:00.0000] <roc>
we are implementing SPDY

[12:36:00.0000] <AryehGregor>
And Chrome already has, right?

[12:36:01.0000] <AryehGregor>
I assume it supports negotiation of some type so you can switch between it and regular HTTP seamlessly.

[12:38:00.0000] <AryehGregor>
Drat it.  I thought of a good reason for why getRangeAt()/addRange() should deal with references (like IE/Gecko) and not copies (like WebKit/Opera), but now I forgot it.

[12:38:01.0000] <timeless>
heh

[12:38:02.0000] <timeless>
grr

[12:38:03.0000] <timeless>
/me gets 7 update notifications

[12:39:00.0000] <timeless>
and i get to enter my password once per notification

[12:55:00.0000] <zewt>
bluh, ff6 (at least) broke structured clone with File (previously it didn't do anything, now it throws)

[12:56:00.0000] <Ms2ger>
Blame the spec

[12:56:01.0000] <zewt>
the spec says it should work

[12:56:02.0000] <zewt>
(and as I'm only using it with History, it's okay to me if it's silently discarded, but throwing is much worse)

[13:00:00.0000] <zewt>
(the spec is horribly vague on how structured clone on File is actually supposed to work, but that's a separate problem...)

[13:07:00.0000] <roc>
what about nightlies?

[13:09:00.0000] <smaug____>
Hixie: ping

[13:16:00.0000] <timeless>
grr

[13:16:01.0000] <timeless>
AryehGregor: so...

[13:16:02.0000] <timeless>
MikeSmith: or maybe you

[13:16:03.0000] <timeless>
i want a way to sign up before a community group is approved

[13:16:04.0000] <timeless>
i.e. i don't want to support a group

[13:16:05.0000] <timeless>
but if the group is created, i do want its spam

[13:16:06.0000] <timeless>
and i don't want to be forced to sign up after it is created

[13:17:00.0000] <timeless>
apparently someone is trying to create a Crypto CG

[13:18:00.0000] <zewt>
because discussing things on a mailing list we're all already on is far too convenient and open

[13:18:01.0000] <timeless>
well

[13:18:02.0000] <timeless>
ignoring that

[13:18:03.0000] <timeless>
because i don't want to *miss* discussions that someone annoyingly moves to a list i'm *not yet* already on!

[13:18:04.0000] <timeless>
and i clearly can't prevent that behavior

[13:19:00.0000] <zewt>
"gotta get away from all these annoying people telling me what i'm doing wrong"

[13:19:01.0000] <timeless>
but i could at least preempt its damage to me :)

[13:19:02.0000] <zewt>
was that the one that started out with that bizarre "crypto + address book" nonsense? heh

[13:19:03.0000] <timeless>
zewt: i didn't look at the guy carefully to figure out if he was one of those

[13:19:04.0000] <timeless>
did you?

[13:19:05.0000] <zewt>
i don't remember, it's been a while since that thread

[13:20:00.0000] <timeless>
from my perspective, reviewing the person takes too much time for me now

[13:20:01.0000] <timeless>
but asking about a way to avoid missing important mail that is hidden behind some nascent CG

[13:20:02.0000] <timeless>
seems generically valuable

[13:20:03.0000] <timeless>
actually, as a gmail user, i'd kinda like a way to tell W3 that i'd like to subscribe to *@cg.w3

[13:20:04.0000] <timeless>
give me all your spam, i can take it!

[13:20:05.0000] <zewt>
if it's the "crypto + address book" one i have little hope of anything useful coming out of it, heh

[13:23:00.0000] <timeless>
/me chuckles

[13:23:01.0000] <timeless>
<anne> The mobile devices are the ones with the high-resolution displays.

[13:23:02.0000] <timeless>
<jreschke> Speak for your own device :-)

[13:24:00.0000] <zewt>
anything modern certainly is higher resolution than any desktop lcd, heh

[13:25:00.0000] <timeless>
i  have 2880x900

[13:25:01.0000] <zewt>
on what size display

[13:25:02.0000] <timeless>
roughly two armspans? (from wrist to elbow x2)

[13:25:03.0000] <zewt>
err what? heh

[13:26:00.0000] <timeless>
/me looks for a ruler

[13:26:01.0000] <timeless>
ooh, i have one at my desk!

[13:26:02.0000] <timeless>
in its original plastic shrink wrap no less!

[13:26:03.0000] <zewt>
my monitor is 1920x1200, 24", resolution around 80 dpi

[13:26:04.0000] <zewt>
my phone is something like 200 dpi?

[13:27:00.0000] <zewt>
much higher resolution

[13:27:01.0000] <timeless>
well, each screen is 1440x900 and seems to be 17 or 18:

[13:27:02.0000] <timeless>
s/:/"

[13:27:03.0000] <timeless>
i think that people have multiple definitions of "resolution" though

[13:27:04.0000] <timeless>
some are talking about pixel dimensions

[13:28:00.0000] <timeless>
and others are talking about ppi

[13:28:01.0000] <zewt>
well anne was obviously talking about dpi

[13:28:02.0000] <timeless>
a coworker has pointed out that there's another concept to consider

[13:28:03.0000] <timeless>
which is roughly css's arc concept

[13:29:00.0000] <timeless>
/me doesn't know if there's a good way to describe it in short hand

[13:29:01.0000] <zewt>
there are many factors that make it tricky: screen size, dpi, and average viewing distance all come into play

[13:29:02.0000] <timeless>
pixels-per-arc-degree?

[13:30:00.0000] <zewt>
also the fact that you can zoom the viewport in mobile browsers

[13:30:01.0000] <zewt>
(can on desktops too with less versitility, though people seem unfortunately content to let anyone using a higher full-page zoom get blurry images)

[13:30:02.0000] <Philip`>
Don't forget that it depends on the velocity of the observer too

[13:31:00.0000] <zewt>
what's the resolution of an iphone falling over the event horizon of a black hole

[13:31:01.0000] <timeless>
Philip`: are you talking about billboards?

[13:31:02.0000] <timeless>
(times square/tokyo)

[13:31:03.0000] <Philip`>
timeless: Just any situation in which relativity matters

[13:35:00.0000] <Philip`>
/me thinks authors think "I want to target all things that would be commonly considered to be mobile devices" and it seems silly to force them to say "target all devices with a display resolution >150dpi" (or whatever) as an inaccurate proxy for what they really want

[13:35:01.0000] <zewt>
i think right now we don't have a solid vocabulary of what's needed, since things have changed so rapidly

[13:37:00.0000] <zewt>
i don't think the first-gen iPhone was >150dpi

[13:37:01.0000] <zewt>
(fwiw)

[13:41:00.0000] <zewt>
ipad is something like 130

[13:49:00.0000] <timeless>
Philip`: soudns right

[13:49:01.0000] <timeless>
s/dn/nd/

[13:49:02.0000] <timeless>
/me sighs

[13:50:00.0000] <timeless>
/me is relearning to use a keyboard

[13:50:01.0000] <timeless>
(kinesis arrived yesterday and today is the first day using it)

[14:22:00.0000] <AryehGregor>
MikeSmith, could you make me the default assignee for DOM Range bugs?

[14:24:00.0000] <Ms2ger>
/me is all for that

[14:27:00.0000] <timeless>
/me suggests not using default assignees this way

[14:27:01.0000] <timeless>
then you don't have to make such requests regularly :)

[14:29:00.0000] <AryehGregor>
How do you suggest using them?

[15:01:00.0000] <timeless>
AryehGregor: have watchable qa contacts per component

[15:01:01.0000] <timeless>
and default assignees of roughly unassigned⊙wo

[15:01:02.0000] <AryehGregor>
timeless, I still want to be the default assignee.

[15:01:03.0000] <AryehGregor>
Not just watch it.

[15:01:04.0000] <timeless>
whe someone starts working on a bug, they take ownership of the bug

[15:01:05.0000] <timeless>
why?

[15:01:06.0000] <timeless>
when you start working on an area, you start watching the qa contact

[15:02:00.0000] <timeless>
when you stop, you stop watching it

[15:02:01.0000] <AryehGregor>
I'm the only one working on all these bugs.

[15:02:02.0000] <timeless>
today

[15:02:03.0000] <AryehGregor>
Granted.

[15:02:04.0000] <timeless>
but what happens when that changes?

[15:02:05.0000] <AryehGregor>
We change all the bugs, obviously.  :)

[15:02:06.0000] <timeless>
you're an average short cited grasshopper

[15:02:07.0000] <AryehGregor>
/me blames Bugzilla for being broken

[15:02:08.0000] <timeless>
i'm speaking as a long lived ant

[15:02:09.0000] <Ms2ger>
cited?

[15:02:10.0000] <timeless>
listen to my words of experience

[15:02:11.0000] <timeless>
sited :(

[15:03:00.0000] <timeless>
?

[15:03:01.0000] <timeless>
sighted!

[15:03:02.0000] <timeless>
/me sighs

[15:03:03.0000] <timeless>
/me sighed

[15:03:04.0000] <Philip`>
/me prefers "psyted"

[15:03:05.0000] <timeless>
<cite>AryehGregor</cite>, a short sighted grasshopper

[15:04:00.0000] <Ms2ger>
Are you calling AryehGregor a piece of work?

[15:04:01.0000] <Philip`>
(which I suppose could also be a telepathic version of SuperTed)

[15:04:02.0000] <timeless>
http://www.youtube.com/user/psyted seens potentially NSFW

[15:04:03.0000] <timeless>
Ms2ger: hey, at least i'm not calling him NSFW :)

[15:05:00.0000] <timeless>
Ms2ger: can i claim he's self published as a collection of all of his quotations?

[15:06:00.0000] <Ms2ger>
You could always try

[15:06:01.0000] <timeless>
/me tries

[15:16:00.0000] <AryehGregor>
Did it work?

[15:30:00.0000] <timeless>
i'm satisfied with it :)

[15:43:00.0000] <karlcow>
AryehGregor: before I say more stupid things on the list, let me check here. ;) could the renameElement thing be used to feed the browser with "<foo><bar/></foo>" and then transform it into the DOM as "<p><img/></p>"

[16:03:00.0000] <TabAtkins>
karlcow: Yes, but it would be a really dumb and fragile thing to do.

[16:06:00.0000] <TabAtkins>
karlcow: You'd have to iterate through the entire document, transforming every individual element.  If JS failed, you'd just have a giant bag of unknown elements that styling wouldn't work on.  Even when JS worked, you'd get significant and unavoidable flashes of unstyled content (multiple, because each swap would almost certainly trigger relayout).

[16:35:00.0000] <karlcow>
TabAtkins: thanks a lot

[16:36:00.0000] <karlcow>
for the dumbness ;) I'm always surprised by what people try to do online ;)

[16:36:01.0000] <llrcombs>
does the <video> element have any way to specify or work with 3D?

[16:36:02.0000] <llrcombs>
afaik it doesn't

[16:37:00.0000] <TabAtkins>
llrcombs: Not explicitly.  If you're doing something like anaglyphic 3d, that's easy to bake into the file.

[16:38:00.0000] <llrcombs>
does anyone think it should be able to hint to the browser that "this video should be displayed in 3D by combining the side-by-side; use the right-side image if no 3D is available?"

[16:38:01.0000] <llrcombs>
(e.g. for graphics cards + displays with native 3D, and stuff like the 3DS)

[16:38:02.0000] <llrcombs>
or a mechanism for providing two separate video files, one for each eye?

[16:39:00.0000] <llrcombs>
also, can WebGL display in 3D on graphics cards that are capible of it?

[16:41:00.0000] <llrcombs>
OHSHI-

[16:42:00.0000] <llrcombs>
which server was that?

[16:52:00.0000] <roc>
llrcombs: Firefox 4 and later can display WebM videos in 3D

[16:52:01.0000] <roc>
with Nvidia's 3D hardware at least

[16:52:02.0000] <llrcombs>
and we're back

[16:52:03.0000] <llrcombs>
roc: does WebM have an internal hinting system for what type of 3D it is?

[16:52:04.0000] <roc>
WebM has a standard way to encode stereo, it just packs left and right frames into a single frame and adds a tiny bit of metadata to tell you it's done that

[16:52:05.0000] <roc>
yes

[16:52:06.0000] <llrcombs>
yeah, not all formats have that

[16:53:00.0000] <roc>
right, so it's up to the container format to define that, then browsers can just do it

[16:53:01.0000] <roc>
no HTML spec changes are required

[16:58:00.0000] <llrcombs>
well, if the user tells the browser that they own a pair of anglyagraph glasses, and a 3D video online is side-by-side (but encoded in H.264), why not allow the site to hint to the browser that it's side-by-side so the browser can display it anglyagraphic instead? I could see this being doable using <canvas>, but definitely not for hardware 3D like parallax or shutter glasses.

[16:59:00.0000] <llrcombs>
Preexisting H.264 or other-format 3D videos could be properly displayed on devices that support that codec

[16:59:01.0000] <llrcombs>
I also like the concept of the browser removing 3D effects if the user says that they don't have any 3D apparatus


2011-08-31
[17:00:00.0000] <llrcombs>
basically, replicating what FireFox does for existing videos and non-WebM files

[17:00:01.0000] <llrcombs>
AGAIN?

[17:00:02.0000] <llrcombs>
good afternoon, netsplit

[17:03:00.0000] <llrcombs>
and back again

[23:33:00.0000] <annevk>
why did undoManager move to Node?

[23:33:01.0000] <annevk>
I thought the plan was to make Node smaller

[23:56:00.0000] <annevk>
AryehGregor, would you be okay with renaming the DOM Range component to Selection and either adding a new DOM Range component and moving range bugs there or moving them to DOM Core?

[01:16:00.0000] <annevk>
Element.replaceElement()

[01:16:01.0000] <annevk>
Does that seem somewhat appropriate? I think a shorter name is somewhat risky and not in the same style as Node.cloneNode()

[01:21:00.0000] <zcorpan>
replaceElement sounds like you should pass in an element node

[01:22:00.0000] <zcorpan>
i dunno, i think rename is fine

[01:23:00.0000] <annevk>
It is much more than renaming though

[01:24:00.0000] <annevk>
But yeah, I guess it is clearer

[01:29:00.0000] <annevk>
zcorpan, you don't think rename() would conflict with custom properties people have out there?

[01:30:00.0000] <annevk>
zcorpan, that was a problem anyway with some of the forms extensions

[01:48:00.0000] <asmodai>
hahaha

[01:48:01.0000] <asmodai>
annevk: http://tweakers.net/nieuws/76475/firefox-vertrouwt-digid-toch-na-verzoek-nederlandse-overheid.html

[01:52:00.0000] <annevk>
ja zag het

[01:53:00.0000] <annevk>
mooie overheid hebben we

[01:53:01.0000] <asmodai>
eh , wth

[01:53:02.0000] <asmodai>
firefox 7 beta now gives me sec_error_ocsp_unknown_cert for bol.com's https. Worked yesterday. o_O *digs into certificates*

[01:56:00.0000] <asmodai>
Mmm, Chrome 14 works

[02:03:00.0000] <annevk>
still one month for Gears of War 3

[02:03:01.0000] <annevk>
it better be awesome

[02:05:00.0000] <asmodai>
That's odd, I don't have the option selected that specifies to treat a certificate as invalid if the OCSP cannot be reached.

[02:05:01.0000] <asmodai>
Yet it does fail on FF 7, unless I am misunderstanding what is going wrong here.

[02:05:02.0000] <asmodai>
Wonder if I trip a bug here or not.

[02:06:00.0000] <asmodai>
annevk: No idea tbh. Just finished DXHR last night, nice story and pretty good game overall.

[02:07:00.0000] <asmodai>
And yet Aurora doesn't fail. Eh?

[02:08:00.0000] <annevk>
/me is not a big fan of RPGs

[02:09:00.0000] <asmodai>
annevk: You're more the FPSer?

[02:11:00.0000] <asmodai>
OK, now I totally don't get it anymore. I temporarily disable OCSP checking. Log into that site, log out again. Make sure I am back on the normal http homepage. Enable OCSP checks again and log in again. No complaints this time. OCSP checks get cached or something?

[02:13:00.0000] <annevk>
platform and some FPS games I guess

[02:14:00.0000] <asmodai>
annevk: From what I heard so far, and I deal with a lot of gamedevs, it seems to be quite ok, worthy successor in the series.

[02:17:00.0000] <annevk>
though nowadays I only play games every month or so

[02:17:01.0000] <annevk>
and occasionally a couple of times a year a little longer if there's a good game

[02:17:02.0000] <zcorpan>
annevk: possible. renameElement is fine also

[02:17:03.0000] <annevk>
oops, that was double :)

[02:19:00.0000] <asmodai>
annevk: Given I am trying to move into the gamedev industry, I play too much ;)

[02:32:00.0000] <Ms2ger>
var is_khtml = (navigator.vendor == 'KDE' || ( document.childNodes && !document.all && !navigator.taintEnabled ));

[03:01:00.0000] <annevk>
so setAttribute can simply overwrite existing xmlns attributes and create xmlns attributes in the null namespace

[03:02:00.0000] <annevk>
I guess it does not matter that much

[03:03:00.0000] <annevk>
xmlns attributes should only matter during parsing

[03:03:01.0000] <annevk>
but the lookup namespace algorithms still use them

[03:03:02.0000] <annevk>
hmm

[04:03:00.0000] <hsivonen>
I noticed that the wiki says nofollow is allowed on <link> but the spec says otherwise

[04:08:00.0000] <karlcow>
annevk: does the Element.replaceElement(), renameElement, etc only change the name?

[04:08:01.0000] <karlcow>
not  a clone. identical copy (two of them after the operation),

[04:09:00.0000] <karlcow>
a mask? change  how it looks but inside still the same. not sure.

[04:09:01.0000] <karlcow>
is it a mutation?

[04:09:02.0000] <karlcow>
/me wonders if there is a Marvel hero for illustrating it. Hulk?

[04:09:03.0000] <karlcow>
Element.hulkElement() - green mutation

[04:10:00.0000] <Ms2ger>
It would create a new element

[04:10:01.0000] <annevk>
you get more or less what cloneNode does

[04:13:00.0000] <annevk>
but the details are not yet entirely clear

[04:14:00.0000] <annevk>
you create a new element, copy attributes and child nodes from the old element and append them to the new element

[04:14:01.0000] <annevk>
then invoke replaceChild on the old element's parent

[04:14:02.0000] <annevk>
open questions are what happens with event handlers, event listeners, and JS properties

[04:15:00.0000] <annevk>
and maybe what happens with cloning steps; but I think that would require a new concept

[04:15:01.0000] <Ms2ger>
JS properties should go away, no?

[04:16:00.0000] <annevk>
I think so, yes

[04:17:00.0000] <annevk>
Personally I favor the most simple approach

[04:17:01.0000] <annevk>
copying event listeners could work, but then people might expect event handlers to work too, which won't always be the case...

[04:20:00.0000] <zcorpan>
/me thinks it should be similar to cloneNode

[04:23:00.0000] <smaug____>
in general event listeners would be broken

[04:23:01.0000] <smaug____>
since the type of 'this' would change

[04:23:02.0000] <smaug____>
and 'this' wouldn't suddenly have some needed properties

[04:24:00.0000] <smaug____>
so, copying event listeners would be very error-prone

[04:24:01.0000] <zcorpan>
sounds like we have an idea about what it should do

[04:24:02.0000] <zcorpan>
so here's another case. what if the element has no parent?

[04:25:00.0000] <zcorpan>
what if it's the root element?

[04:25:01.0000] <smaug____>
root element should work just fine

[04:25:02.0000] <zcorpan>
AryehGregor's suggested algorithm wouldn't work for root

[04:26:00.0000] <smaug____>
cloneAndRename (or whatever to call it) should return the created node, so element without parent should be ok

[04:26:01.0000] <smaug____>
hmm, what could be a good name.

[04:26:02.0000] <smaug____>
it isn't a clone, it isn't rename

[04:27:00.0000] <smaug____>
 morph?

[04:29:00.0000] <Ms2ger>
replaceWithANewElementAndTransferAttributesAndStuff(DOMString localName)

[04:29:01.0000] <zcorpan>
that's more like it

[04:31:00.0000] <jgraham>
Can we have another method that moves the element to a random location in the DOM?

[04:31:01.0000] <jgraham>
We could call it Element.witnessProtection()

[04:32:00.0000] <jgraham>
*renames and moves

[04:32:01.0000] <jgraham>
Sigh

[04:33:00.0000] <annevk>
zcorpan, replaceChild works for root

[04:34:00.0000] <annevk>
jgraham, all mutation methods already move by default

[04:34:01.0000] <Ms2ger>
But not to a random location

[04:35:00.0000] <annevk>
smaug____, I think renameElement(name) is clearer than morphElement(name)

[04:35:01.0000] <annevk>
smaug____, renaming an element is just a complex operation :)

[04:35:02.0000] <smaug____>
renameElement isn't clear at all, since it doesn't rename ;)

[04:35:03.0000] <Ms2ger>
What he said

[04:38:00.0000] <annevk>
so replaceElement(name)?

[04:38:01.0000] <annevk>
zcorpan did not like that because he thought it took a node rather than a string

[04:40:00.0000] <zcorpan>
yeah i think it's too similar to replaceChild

[04:40:01.0000] <smaug____>
I agree

[04:40:02.0000] <jgraham>
I think the problem is trying to wrap two logical operations into a single DOM call

[04:41:00.0000] <zcorpan>
three

[04:41:01.0000] <jgraham>
three?

[04:41:02.0000] <jgraham>
Well clone, rename, replace I guess

[04:42:00.0000] <smaug____>
create ,copy stuff, replace

[04:42:01.0000] <zcorpan>
1. make a new element. 2. replace with old element. 3. transfer children and attributes

[04:42:02.0000] <zcorpan>
maybe that's 4

[04:42:03.0000] <jgraham>
It would be nice if cloneNode could just be extended with an extra argument to specify what to replace

[04:42:04.0000] <zcorpan>
see we can't even agree on the order :P

[04:42:05.0000] <annevk>
jgraham, how does that help?

[04:44:00.0000] <jgraham>
annevk: Well if you can get a reference to the subtree that you want to insert then you can use normal DOM operations to insert the new subtree in place of the old one

[04:45:00.0000] <zcorpan>
renameElement doesn't make a clone

[04:45:01.0000] <annevk>
it does seem similar to file move

[04:45:02.0000] <jgraham>
Well I guess it is different for the children

[04:45:03.0000] <annevk>
move a b

[04:45:04.0000] <annevk>
a.move(b)

[04:45:05.0000] <annevk>
but moveElement is all kind of wrong somehow

[04:46:00.0000] <jgraham>
Well it's not really that similar to a move since nothing moves

[04:46:01.0000] <jgraham>
That's really the distinguishing feature of a move :)

[04:48:00.0000] <karlcow>
it looks like the teletransportation in The Fly

[04:49:00.0000] <karlcow>
Element.jeffGoldblumed

[04:50:00.0000] <zcorpan>
maybe we should call it dunnoLOL()

[04:52:00.0000] <karlcow>
hmm Element.lolCat()

[04:54:00.0000] <annevk>
I still think rename is best

[04:55:00.0000] <annevk>
it's not actually what it does, but it matches what people expect

[04:57:00.0000] <jgraham>
substitute?

[04:57:01.0000] <zcorpan>
renameElement is my vote

[04:58:00.0000] <Ms2ger>
Mm, I think we should paint the bike shed "substitute" as well

[04:58:01.0000] <annevk>
substitute is like replace, as confusing

[05:08:00.0000] <smaug____>
how does rename match what people expect

[05:09:00.0000] <jgraham>
I think that the differnce between rename and substitute is that it sounds like rename preserves old === new whereas it doesn't sound like substitute does

[05:10:00.0000] <smaug____>
substitute is significantly better than rename

[05:13:00.0000] <zcorpan>
i like substituteElement better than replaceElement

[05:14:00.0000] <hsivonen>
when trying to find a bug in the JS of a product that's still being sold, I have a hard time deciding if finding Dreamweaver boilerplate for Netscape 4.x should make me facepalm or think "compatibility FTW"

[05:23:00.0000] <annevk>
smaug____, they expect the element to be renamed and prolly don't care too much about the side effects in the usual scenario

[05:23:01.0000] <annevk>
smaug____, such as renaming <strong> to <b> or some such

[05:24:00.0000] <hsivonen>
...and code for IE 5.0, too...

[05:24:01.0000] <smaug____>
the "doesn't copy event listeners" is such a major side effect, that I assume people will care about it

[05:25:00.0000] <smaug____>
renameElement sounds like  the element is just renamed, but all the properties and listeners are kept

[05:29:00.0000] <annevk>
in the typical case no event listeners are registered though

[05:30:00.0000] <annevk>
well the scenarios sketched so far

[05:30:01.0000] <zcorpan>
changeElement()?

[05:30:02.0000] <Ms2ger>
doSomething()

[05:32:00.0000] <smaug____>
what is wrong with substituteElement() ?

[05:32:01.0000] <smaug____>
(or morph, but others don't seem to like it)

[05:33:00.0000] <smaug____>
(I think substitute is better though)

[05:34:00.0000] <annevk>
substitute sounds like it takes a node

[05:34:01.0000] <annevk>
it's also hard to type

[05:37:00.0000] <zcorpan>
yeah with renameElement it's clear what the expected argument is

[05:40:00.0000] <zcorpan>
annevk: now spec this out! :)

[05:41:00.0000] <smaug____>
but renameElement is still wrong, since it doesn't rename the element

[05:41:01.0000] <smaug____>
/me is being annoying again. Sorry :)

[05:43:00.0000] <annevk>
I think I will write a WHATWG Weekly instead

[05:44:00.0000] <jgraham>
I think the fact that rename breaks the implied invariant is worse than the fact that it isn't obvious what arguments substitute takes

[05:44:01.0000] <jgraham>
Alsom if we extend it to also allow different attributes on the new element then it doesn't make sense to use rename

[05:45:00.0000] <jgraham>
(possible use case: editor that wants to replace a <span> with an <a href> but preserve children)

[05:45:01.0000] <annevk>
renameElement followed by setAttribute()

[05:46:00.0000] <jgraham>
Well sure

[05:46:01.0000] <jgraham>
But this whole thing is a convenience API

[05:47:00.0000] <jgraham>
It seems strange to base your decision of what is convenient on your preferred name

[05:48:00.0000] <annevk>
create/copy/replace is complicated enough as it is, I would not want to overload it more

[05:48:01.0000] <zcorpan>
the use case also involved restoring selection

[05:49:00.0000] <jgraham>
I don't think that adding/removing/changing some attribute values is particularly complex compared to the other operations it does

[05:50:00.0000] <jgraham>
Especially if we get the convenience method for creating elements (I forget the name) that has the syntax whatever(name, {attr_name, attr_value})

[05:50:01.0000] <jgraham>
er

[05:50:02.0000] <jgraham>
whatever(name, {attr_name: attr_value})

[05:50:03.0000] <jgraham>
It would be strange not to support the same syntax

[05:53:00.0000] <zcorpan>
maybe we should solve teh restore selection use case separately

[05:54:00.0000] <zcorpan>
and wait with convenience apis

[05:54:01.0000] <zcorpan>
restore selection is the only thing that was impossible today

[06:02:00.0000] <annevk>
whoa http://upload.wikimedia.org/wikipedia/commons/b/b7/Population_curve.svg

[06:02:01.0000] <hsivonen>
MikeSmith: I redeployed V.nu with the new code from the repos. Thanks

[06:02:02.0000] <annevk>
zcorpan, might be worth looking into

[06:03:00.0000] <hsivonen>
MikeSmith: I'm not ready to express an opinion about <menuitem>, yet, because I haven't discussed it enough with the people who want it

[06:27:00.0000] <hsivonen>
what's the most current location of the spec that defines createContextualFragment?

[06:27:01.0000] <hsivonen>
Ms2ger: ^

[06:28:00.0000] <hsivonen>
http://html5.org/specs/dom-parsing.html this?

[06:29:00.0000] <annevk>
yes

[06:30:00.0000] <hsivonen>
annevk: thanks

[06:34:00.0000] <annevk>
http://blog.whatwg.org/weekly-rich-text-editing

[06:58:00.0000] <annevk>
hsivonen, seems to be source order in Opera/WebKit

[06:59:00.0000] <annevk>
I wish we could turn it into a dictionary though and serialize in alphabetical order

[07:09:00.0000] <hsivonen>
annevk: yeah, I noticed that Opera and WebKit were less surprising

[07:10:00.0000] <hsivonen>
annevk: but I think Firefox and IE are enough to make my point that source order can't be required for Web compat

[07:10:01.0000] <hsivonen>
except for the <embed> element!

[07:10:02.0000] <annevk>
sure, but can be required for sanity :)

[07:11:00.0000] <hsivonen>
annevk: depends on what nice things Gecko and Trident get from what they do now

[07:11:01.0000] <hsivonen>
which is why I CCed bz

[07:11:02.0000] <hsivonen>
I believe bz knows *why* CSS-mapped legacy presentational hint attributes sort differently in Gecko

[07:12:00.0000] <hsivonen>
my hypothesis is that the ordering in IE simply depends on the implementation details of a hashtable somewhere in IE internals

[07:18:00.0000] <hsivonen>
does ubuntu have downgrade manager now or what's the thing with down-pointing arrow in https://launchpadlibrarian.net/78619733/Launcher%2520washout.png

[07:19:00.0000] <zcorpan>
downloads maybe?

[07:25:00.0000] <zcorpan>
hsivonen: does the attribute order for <font> have web compat implications?

[07:29:00.0000] <hsivonen>
zcorpan: I'm not aware of any

[07:30:00.0000] <hsivonen>
zcorpan: since browsers have different orders, I'm guessing it doesn't have Web compat implications

[07:39:00.0000] <annevk>
can ranges ever be outside a document?

[07:47:00.0000] <annevk>
smaug____, could you comment on http://www.w3.org/Bugs/Public/show_bug.cgi?id=12205#c21 maybe?

[07:47:01.0000] <annevk>
smaug____, how likely is it that Gecko would implement it?

[07:47:02.0000] <annevk>
I can probably convince someone at Opera to do it if Gecko has done it

[07:51:00.0000] <smaug____>
looking

[08:07:00.0000] <hsivonen>
btw, now that Zakim supports SIP, has anyone figured out what the least troublesome way of getting SIP access is?

[08:07:01.0000] <hsivonen>
the more I read about SIP, the less I appreciate it, though

[08:07:02.0000] <hsivonen>
and the more it seems that Skype deserves to win

[08:26:00.0000] <AryehGregor>
annevk, sounds good to me.

[08:27:00.0000] <AryehGregor>
TabAtkins, DOM operations never trigger repaints.  Repaints only happen when the event loop spins.  Except in Opera, which is broken.

[08:27:01.0000] <annevk>
"broken"

[08:28:00.0000] <jgraham>
Hmm, why is that broken?

[08:28:01.0000] <AryehGregor>
Because it will display transient states to the user.

[08:28:02.0000] <AryehGregor>
Also, it's different from every other browser, which is broken more or less by default.

[08:28:03.0000] <AryehGregor>
I saw someone from Opera say it caused site compat issues, IIRC.

[08:29:00.0000] <jgraham>
Displaying transient states to the user is normal

[08:29:01.0000] <jgraham>
Think progressive loading

[08:29:02.0000] <timeless>
asmodai: i'm fairly certain ocsp checks are cached

[08:29:03.0000] <asmodai>
timeless: That would explain then

[08:29:04.0000] <AryehGregor>
Depends on your definition of "transient".

[08:29:05.0000] <jgraham>
It can cause slowness if you don't carefully control how often you repaint ofc

[08:29:06.0000] <annevk>
it causes uglyness indeed mostly because other browsers have a silly blocking implementation

[08:30:00.0000] <annevk>
and authors authored against the silly implementations

[08:30:01.0000] <AryehGregor>
In Opera's implementation, there's no way to have the browser not display a half-mutated DOM.  It will display some snapshot at random depending on when it decides to repaint.

[08:31:00.0000] <jgraham>
Roughly the same as every browser and loading off the network really

[08:31:01.0000] <AryehGregor>
As opposed to other browsers, which will never display anything unless the event loop spins, so a bunch of consecutive operations will always be run atomically from the user's perspective.

[08:31:02.0000] <AryehGregor>
Loading off the network is a special case.

[08:31:03.0000] <jgraham>
But there's no intrinsic reason it has to be

[08:31:04.0000] <jgraham>
If you have a very slow running script, displaying something is better than displaying nothing

[08:32:00.0000] <AryehGregor>
Yes, if it's slow-running enough, maybe.

[08:32:01.0000] <AryehGregor>
Particularly if you try to be intelligent about which point to display.

[08:32:02.0000] <jgraham>
That is often the case on lower end devices

[08:35:00.0000] <zcorpan>
hmm. setting textarea.value = '\r\n' it gets normalized to '\n' in webkit and gecko

[08:35:01.0000] <zcorpan>
but spec doesn't say so afaict

[08:39:00.0000] <jgraham>
There is a bug somewhere about normalizing textarea on suibmission

[08:39:01.0000] <jgraham>
I'm not sure if it covers the DOM as well

[09:34:00.0000] <zcorpan>
yeah it seems to me the DOM should be normalized to \n and submission normalized to \r\n

[09:39:00.0000] <annevk>
hmm event handlers

[09:45:00.0000] <dglazkov>
good morning, Whatwg!

[09:45:01.0000] <zcorpan>
afternoon dglazkov

[09:49:00.0000] <Ms2ger>
Night

[09:53:00.0000] <Ms2ger>
Interesting subject: http://lists.w3.org/Archives/Member/tag/2011Aug/0015.html

[10:12:00.0000] <annevk>
http://www.w3.org/TR/DOM-Level-3-Core/introduction.html#ID-Conformance suggests DOM3 Events can be split

[11:11:00.0000] <zcorpan>
matjas: you forgot to RT :)

[11:12:00.0000] <matjas>
no I didn’t! *whistles*

[11:12:01.0000] <zcorpan>
:)

[11:13:00.0000] <annevk>
WHATWG forums looks different

[11:14:00.0000] <zcorpan>
since when?

[11:14:01.0000] <annevk>
since green?

[11:15:00.0000] <annevk>
zcorpan, cannot read news without being logged in?

[11:16:00.0000] <zcorpan>
huh, that should be possible

[11:20:00.0000] <zcorpan>
fixed

[11:20:01.0000] <zcorpan>
thanks

[11:30:00.0000] <zcorpan>
annevk: it has had this design for like a year

[11:31:00.0000] <annevk>
evidently I'm more of an IRC guy

[11:31:01.0000] <annevk>
made a RT from @whatwg

[11:34:00.0000] <zcorpan>
thanks

[11:37:00.0000] <annevk>
so <textarea> submission basically makes text/plain decoding lose

[11:37:01.0000] <annevk>
or is text/plain submission \n separated?

[11:38:00.0000] <zcorpan>
does anyone support text/plain submission?

[11:38:01.0000] <zcorpan>
what's the use case for text/plain again?

[11:40:00.0000] <annevk>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1126

[11:40:01.0000] <annevk>
debugging I guess?

[11:41:00.0000] <annevk>
does not seem to work in WebKit?

[11:41:01.0000] <annevk>
well Chrome

[11:43:00.0000] <zcorpan>
seems to me like it is a dubious use case and it changes security assumptions about form submission

[11:44:00.0000] <annevk>
what assumptions?

[11:44:01.0000] <annevk>
this is not a new feature fwiw

[11:45:00.0000] <zcorpan>
that browsers can only submit in two specific formats

[11:51:00.0000] <annevk>
seems like a false assumption

[11:54:00.0000] <zcorpan>
evidently. but we should be careful about what changing what can be sent from the browser. c.f. websockets

[12:02:00.0000] <timeless>
Ms2ger: grumble, your link was members only

[12:04:00.0000] <Ms2ger>
Grumble at the tag, please

[12:04:01.0000] <timeless>
of note, that email links to http://www.w3.org/2010/html-xml/snapshot/

[12:04:02.0000] <timeless>
which i believe is less confidential

[12:06:00.0000] <timeless>
do they take editorial bug reports?

[12:06:01.0000] <timeless>
> Where HTML goes to great lengths to defined how

[12:06:02.0000] <timeless>
s/defined/define/

[12:07:00.0000] <Philip`>
"It seems that the world at large is unlikely to adopt polyglot markup as the standard way to encode all HTML documents"

[12:07:01.0000] <Philip`>
They seem good at understatement

[12:08:00.0000] <Ms2ger>
I hear they're all Brits

[12:09:00.0000] <timeless>
Philip`: yeah well

[12:10:00.0000] <timeless>
they're wearing diplomatic hats

[12:10:01.0000] <timeless>
a coworker asked me if i spoke diplomat today

[12:10:02.0000] <timeless>
i looked at his problem and threw up my hands :(

[12:10:03.0000] <timeless>
if anyone here happens to have a diplomat hat handy and 10mins, please let me know, i'd be glad to share our problem :)

[12:10:04.0000] <timeless>
> makes all of the XML power available to any HTML document.

[12:11:00.0000] <timeless>
what's this `XML power` of which they speak? :)

[12:11:01.0000] <Philip`>
"How can XML be made more forgiving of errors?" - I think the problem with XML is not that it's unforgiving of errors, it's that there are lots of obscure edge cases that are considered errors

[12:11:02.0000] <zewt->
isn't not being forgiving of errors sort of the point of xml? heh

[12:12:00.0000] <timeless>
/me thought that it was the point of xml

[12:13:00.0000] <Philip`>
Draconically requiring matched tags and quoted attributes is probably okay; requiring that text doesn't contain U+FFFE or "]]>", and that comments don't contain "--", and that xmlns:* does not have the value "http://www.w3.org/XML/1998/namespace", is not okay and trips people up

[12:13:01.0000] <timeless>
> the most successful approach is likely to involve parsing the XML with an XML parser and the HTML with an HTML parser<span class="revision-deleted"><span>, combining the resulting DOMs through some other process.</span></span></p>

[12:13:02.0000] <timeless>
/me likes how they deleted the period from the end of that sentence

[12:14:00.0000] <jcranmer>
http://tools.ietf.org/html/rfc6366.txt

[12:14:01.0000] <AryehGregor>
Philip`, I'd say that draconian enforcement of quoted attributes is fine, but matched tags is already dodgy.  It means well-formedness isn't local.  It's common to want to stitch together a bunch of elements with the start and end tags added in separate places.

[12:15:00.0000] <AryehGregor>
But it's very rare you want to output an attribute other than all at once, so it's relatively easy to vet for well-formedness.

[12:15:01.0000] <timeless>
> Also, conforming trees that have tr elements as children of table elements will be replaced with semantically equivalent but tree-wise different construct where the tr elements gain a tbody parent which is a child of the table.

[12:15:02.0000] <AryehGregor>
MediaWiki has known cases where its parser misnests tags in output, which are very hard to fix, but attribute quoting is just fine.

[12:15:03.0000] <timeless>
is it me, or is `construct` in the above misconjugated?

[12:15:04.0000] <AryehGregor>
s/construct/constructs/

[12:18:00.0000] <annevk>
timeless, prolly best to email public-html-xml

[12:19:00.0000] <timeless>
annevk: i wonder what kind of headache that will give me

[12:19:01.0000] <annevk>
from this channel hsivonen and I are involved

[12:20:00.0000] <timeless>
how many advils has it cost you so far? :)

[12:20:01.0000] <annevk>
I think in particular if hsivonen had not been involved that document could have turned out less realistic

[12:22:00.0000] <hober>
annevk: indeed

[12:23:00.0000] <timeless>
that doesn't at all surprise me

[12:23:01.0000] <timeless>
or perhaps better, that matches my expectations

[12:25:00.0000] <timeless>
heh

[12:25:01.0000] <timeless>
/me is amused by <script type="application/xml"> <![CDATA[ your xml cdata here ? ]]> </script>

[12:26:00.0000] <Ms2ger>
>.<

[12:26:01.0000] <Ms2ger>
Are these the guys who invented RSS?

[12:26:02.0000] <timeless>
ATOM is indirectly referenced repeatedly

[12:27:00.0000] <timeless>
their only *actual* references are to MTOM and XOP

[12:27:01.0000] <timeless>
which is amusing

[12:27:02.0000] <timeless>
since they should formally reference at least DOM (since they reference it in the body)

[12:28:00.0000] <timeless>
> Working out all of the details to assure that the necessary error correction produced expected results in all cases might be tedious

[12:28:01.0000] <timeless>
do they mean `produced` or `produces`?

[12:31:00.0000] <Philip`>
Or 'produce'?

[12:32:00.0000] <Philip`>
Tenses are hard

[12:33:00.0000] <zcorpan>
it should be producez

[12:33:01.0000] <Philip`>
The more fundamental problem with that sentence is that they consider trying to come up with a widely accepted definition of "expected" to be "tedious" instead of "an unending exercise in bikeshedding"

[12:34:00.0000] <timeless>
Philip`: i only deal in grammatical errors when sending editorial comments

[12:34:01.0000] <timeless>
technical details like "

[12:34:02.0000] <timeless>
utter futility"

[12:34:03.0000] <timeless>
are outside my scope :)

[12:35:00.0000] <Philip`>
(Better to come up with something roughly agreeable just for the most obvious cases, and let the other "all cases" be whatever the algorithm happens to produce)

[12:36:00.0000] <Philip`>
You might want to suggest they use a spelling checker too ("gauranteed", "concatentation", etc)

[12:37:00.0000] <timeless>
you're kidding

[12:37:01.0000] <timeless>
/me sighs

[12:37:02.0000] <zewt>
heh

[12:37:03.0000] <Philip`>
/me wonders if a concatentation is when you join two incantations together

[12:37:04.0000] <timeless>
"simulataneously"

[12:37:05.0000] <timeless>
you're not kidding

[12:37:06.0000] <zewt>
am i the only person who assumes that people who say "2" or "u're" are doing that to cover the fact that they can't keep to/two/too and your/you're straight

[12:38:00.0000] <Philip`>
zewt: Yes

[12:38:01.0000] <Ms2ger>
u're? Don't think I've seen that before

[12:38:02.0000] <Philip`>
I assume they're just too lazy to type properly

[12:38:03.0000] <timeless>
Philip`: `yes he's the only person`?

[12:38:04.0000] <zewt>
i saw it recently, i was dimly impressed

[12:38:05.0000] <Philip`>
(or grew up in a peer group where that has become the expected form of communication)

[12:38:06.0000] <timeless>
/me just watched a clock turn backwards by a minute

[12:38:07.0000] <timeless>
/me wonders why

[12:39:00.0000] <zewt>
lsd?

[12:39:01.0000] <Philip`>
ntp?

[12:40:00.0000] <zewt>
by "turn" i pictured an analog clock

[12:40:01.0000] <Philip`>
Analogue clocks could be driven by a synchronised digital time source

[12:40:02.0000] <timeless>
zewt: have you seen android's clock?

[12:40:03.0000] <timeless>
it's the old fashioned clock radio style

[12:40:04.0000] <timeless>
(flip digits)

[12:40:05.0000] <zewt>
i remember my highschool had analog wall clocks that were synchronized to the office, and once in a while you'd see them change as someone adjusted them

[12:41:00.0000] <timeless>
1?#>!@$

[12:41:01.0000] <timeless>
wow

[12:41:02.0000] <zewt>
not on my android (nexus s stock)

[12:41:03.0000] <Philip`>
I remember my school had a clock that took about 10 seconds for the seconds hand to count from 0 to 30, and then about 50 seconds to drag itself all the way back up to 60

[12:41:04.0000] <timeless>
http://www.google.com/imgres?imgurl=http://appstorehq-production.s3.amazonaws.com/senseanalogclockwidget4x2-android-239034.320x480.1275100026.63102.jpg&imgrefurl=http://www.appstorehq.com/senseanalogclockwidget4x2-android-239034/app&usg=__7btBcdHqoaaikl31ObYlzbLzOB0=&h=480&w=320&sz=35&hl=en&start=0&zoom=1&tbnid=RjpJhSBnt3aZrM:&tbnh=140&tbnw=95&ei=JIFeTu2HGILK0AHhxIjmAg&itbs=1&iact=hc&vpx=1216&vpy=320&dur=1109&hovh=275&hovw=183&tx=133&ty=112&pa

[12:41:05.0000] <zewt>
i'm never a fan of digital things that try to pretend they're analog things--digital oscilloscopes are really bad like that

[12:42:00.0000] <Philip`>
which was disturbing when that clock was used for timing exams

[12:42:01.0000] <timeless>
Philip`: wow

[12:42:02.0000] <zewt>
timeless: that's pretty gross, but i havn't seen it--ui looks 1.xish though

[12:43:00.0000] <timeless>
zewt: it's really disturbing when you switch back to the home screen

[12:43:01.0000] <timeless>
because it doesn't update while it isn't visible

[12:43:02.0000] <timeless>
so when you switch back, it immediately does Flips to get to <now>

[12:43:03.0000] <timeless>
and you blink and think "did i just see it do something?"

[12:43:04.0000] <zewt>
a pretty terrible ui--why would you waste half of the very limited space on your home screen for something that's already shown in the notification bar?

[12:44:00.0000] <AryehGregor>
Because lots of people use their phones as clocks, and want to be able to quickly pull them out and glance at the time.

[12:45:00.0000] <zewt>
no, when you do that you're looking at the lock screen, not the home screen

[12:45:01.0000] <zewt>
and the lock screen does have a bigger clock (not that big)

[12:45:02.0000] <timeless>
zewt: i'm not defending it

[12:45:03.0000] <AryehGregor>
My home screen doesn't have a clock on it . . . are we talking about Android here?

[12:45:04.0000] <timeless>
i'm merely noting that it was what i saw when i encountered droids

[12:45:05.0000] <timeless>
it drove me nuts

[12:46:00.0000] <zewt>
timeless: i've just never seen it before--are you sure that isn't some vendor hack or third party widget?

[12:46:01.0000] <timeless>
/me shrugs

[12:46:02.0000] <AryehGregor>
My stock Nexus One doesn't have anything like that.

[12:46:03.0000] <timeless>
i'm not an android owner

[12:46:04.0000] <timeless>
unless you count my n900

[12:46:05.0000] <timeless>
or some other non hardware droid systems

[12:46:06.0000] <timeless>
(and presumably you don't..)

[12:47:00.0000] <timeless>
fwiw, i do like having a large clock (digital please) on a lock screen

[12:47:01.0000] <timeless>
the n8 does that pretty well (although it's disturbing at first since they use an Outline font because the screen is OLED)

[12:47:02.0000] <zewt>
http://images.google.com/search?hl=en&gbv=2&tbm=isch&q=usb+oscilloscope these things are all the most absurd uis

[12:47:03.0000] <timeless>
http://www.google.com/imgres?imgurl=http://4.bp.blogspot.com/_PVz_hhcxiro/TUoC022Sj0I/AAAAAAAAAT0/RYO1jL-VYec/s1600/lockscreen.png&imgrefurl=http://www.nabiltewolde.com/2011/02/symbian3-ui-improvements.html&usg=__u80a9d2JbT3Lqy4UTfpJITPSchE=&h=640&w=360&sz=7&hl=en&start=0&zoom=1&tbnid=UF_W4KVixJzmQM:&tbnh=175&tbnw=103&ei=hYJeTt3mMOHa0QGcrOHRDg&itbs=1&iact=rc&dur=637&page=1&ndsp=31&ved=1t:429,r:16,s:0&tx=57&ty=73

[12:54:00.0000] <timeless>
Philip`: ooh, you missed "arbirary " :)

[12:55:00.0000] <Philip`>
timeless: I wasn't trying to be comprehensive :-)

[12:56:00.0000] <timeless>
Philip` / Ms2ger / annevk : http://pastebin.mozilla.org/1316571

[12:56:01.0000] <Philip`>
/me didn't even try to read the whole thing, because it seemed uninteresting

[12:56:02.0000] <timeless>
/me needs to get into the habit of feeding all documents through MS Word

[12:56:03.0000] <Ms2ger>
Don't want anything to do with it :)

[12:58:00.0000] <Philip`>
(In particular it just seemed to be XML people trying to figure out what HTML5 people have been saying about HTML/XML integration for the past five years)

[13:00:00.0000] <annevk>
timeless, okay if I email that to the list?

[13:01:00.0000] <timeless>
annevk: i was going to subscribe and email it

[13:02:00.0000] <timeless>
step one is figuring out how to subscribe, step 2 is getting the pastebin into my mail client

[13:02:01.0000] <Ms2ger>
You're a brave man

[13:03:00.0000] <timeless>
the word you sought was "foolish"

[13:03:01.0000] <annevk>
timeless, public-html-xml-request with subject subscribe should do

[13:04:00.0000] <timeless>
annevk: yeah, but that's easier to do if you're willing to use your desktop to do it

[13:04:01.0000] <annevk>
heh, gonna be fun meeting you again I can tell :)

[13:04:02.0000] <timeless>
i don't like giving my personal credentials to my corporate desktop computers

[13:04:03.0000] <timeless>
oh right, i need to file an expense claim for the tpac registration (it already hit my credit card statement)

[13:04:04.0000] <timeless>
and i need to book travel

[13:05:00.0000] <timeless>
(oddly, i trust my corporate phone w/ my personal credentials)

[13:05:01.0000] <annevk>
can Mozilla sponsor you Ms2ger?

[13:06:00.0000] <Ms2ger>
Wasn't planning to go

[13:06:01.0000] <timeless>
grr

[13:06:02.0000] <timeless>
annevk: can you poke gmail for me?

[13:06:03.0000] <timeless>
i'm using opera mobile on my n900

[13:06:04.0000] <timeless>
and when i try to do stuff w/ m.gmail.com (?) i get XML parsing failed

[13:06:05.0000] <timeless>
XML parsing failed: syntax error

[13:06:06.0000] <timeless>
Reparse document as html

[13:07:00.0000] <timeless>
which would be fine if it were one document

[13:07:01.0000] <timeless>
but it'

[13:07:02.0000] <timeless>
s each link for a session

[13:07:03.0000] <timeless>
so i have to click that extra intermediate page each time i do anything :(

[13:07:04.0000] <annevk>
works here

[13:07:05.0000] <annevk>
Ms2ger, too bad

[13:08:00.0000] <timeless>
this is https://mail.google.com/mail/u/0/x/ ....

[13:08:01.0000] <timeless>
/me wonders how to check the opera version

[13:08:02.0000] <timeless>
it doesn't *always* happen

[13:08:03.0000] <timeless>
but it does seem to always happen for a given session

[13:09:00.0000] <timeless>
/me goes to see how this list writes its message summaries

[13:10:00.0000] <timeless>
/me doesn't detect any patterns

[13:10:01.0000] <timeless>
/me waits for the subscribe bot

[13:16:00.0000] <timeless>
/me finds all sorts of unread mail :(

[13:19:00.0000] <timeless>
annevk: ok, almost sent

[13:21:00.0000] <timeless>
Ms2ger: http://lists.w3.org/Archives/Public/public-html-xml/2011Aug/0048.html

[13:22:00.0000] <Ms2ger>
Hah, telcon

[13:22:01.0000] <timeless>
hurricane trumps

[13:22:02.0000] <timeless>
:)

[13:24:00.0000] <timeless>
annevk: ok, my email appears in the archives

[13:41:00.0000] <timeless>
Ms2ger: for unknown reasons, i only now saw your email about `you were unhappy about the ordering` :(

[13:43:00.0000] <timeless>
oh odd

[13:43:01.0000] <timeless>
/me got offlist mail about something

[13:43:02.0000] <timeless>
/me wonders why someone expected that to work

[13:45:00.0000] <zewt>
why would anyone expect offlist mail to not work?

[13:58:00.0000] <TabAtkins>
AryehGregor: Did I say "repaint"?  If so, I meant "relayout", which happens in the middle of script execution all the time.

[14:02:00.0000] <AryehGregor>
TabAtkins, you said "Even when JS worked, you'd get significant and unavoidable flashes of unstyled content (multiple, because each swap would almost certainly trigger relayout)."  You did say relayout, but I'm not clear how you'd get FOUCs without repaints.

[14:02:01.0000] <TabAtkins>
Oh, whoops, indeed.  Bad on me.  You'd get a lot of slowdown, rather.

[14:22:00.0000] <timeless>
TabAtkins: http://www.guardian.co.uk/artanddesign/2010/nov/21/sargy-mann-blind-painter-tim-adams is amusing fwiw

[14:23:00.0000] <timeless>
s/amusing/educational+informative/

[14:23:01.0000] <timeless>
you have to watch the video

[14:23:02.0000] <timeless>
but my conclusions from it is that one generally will need an `assistant` who can answer questions like `what color is this point`

[14:24:00.0000] <timeless>
i haven't actually seen ATs that do that, oddly

[14:37:00.0000] <timeless>
it also seems relatively clear that one is unlikely to generally want a list of all arcs which compose a forest

[14:37:01.0000] <timeless>
since most of them will be overlapped by other things and are thus irrelevant to you or anyone who would be at your perspective

[14:38:00.0000] <timeless>
anyway

[14:39:00.0000] <timeless>
it seems that in the end, if someone is going to design something for people who need accessibility support, they will need to ensure that the widgets they provide aren't ambiguous and have sufficient detail to indicate what would happen if they were poked

[14:39:01.0000] <timeless>
a button labeled "button" would thus fail, especially if it were next to 4 others with the same label..

[14:40:00.0000] <timeless>
whereas buttons labeled "floodfill", "stamp", "line", "erase" wouldn't have such a problem, and one wouldn't necessarily need to care where they are so long as one could discover they exist..

[14:41:00.0000] <timeless>
/me has a feeling that a lot of this is already addressed by ARIA's ability to list areas which are subject to change, and identify changes

[14:41:01.0000] <timeless>
/me heads out for the day

[16:06:00.0000] <Moo^_^>
does it taint the canvas if I use images loaded from the same domain, different port? (repeated from #html5)

[16:26:00.0000] <jacobolus>
do any desktop browsers fire events for multitouch trackpad gestures?

[16:32:00.0000] <jacobolus>
desktop safari doesn't seem to fire 'touchstart' or 'gesturestart', but I guess I can get both X and Y movements out of a 'mousewheel' event

[16:32:01.0000] <jacobolus>
any way to get pinches or rotations?

[16:33:00.0000] <jacobolus>
do other browsers support trackpad multitouch?

[16:38:00.0000] <roc>
Moo^_^: yes I think so

[16:39:00.0000] <Moo^_^>
roc: do you know if there exist anything similar to CORS but for images?

[16:39:01.0000] <Moo^_^>
/me has not read the spec

[16:40:00.0000] <jacobolus>
Moo^_^: what's the use case?

[16:40:01.0000] <Moo^_^>
jacobolus: I want to use images from another domain, but not tainting the canvas

[16:40:02.0000] <Moo^_^>
need to mash up images from two different sources

[16:41:00.0000] <roc>
there is CORS for images now

[16:41:01.0000] <roc>
see the "crossorigin" attribute

[16:41:02.0000] <roc>
only implemented in latest versions of Chrome and Firefox AFAIK

[16:41:03.0000] <zewt>
yeah, CORS already does that, for webgl

[16:44:00.0000] <Philip`>
Moo^_^: Tainting is based on origins, which are scheme+host+port

[16:45:00.0000] <Philip`>
so changing port is the same as changing host

[16:45:01.0000] <Moo^_^>
roc: zewt: Philip`: Thanks a lot!