2011-07-01
[17:01:00.0000] <boblet>
and browser support for :not is… :p

[17:02:00.0000] <boblet>
will attempt to rephrase more persuasively and email to the list. thanks for the feedback

[17:03:00.0000] <Hixie>
:-)

[17:03:01.0000] <Hixie>
isn't :not pretty widely supported by now?

[17:03:02.0000] <Hixie>
it was in like acid2 wasn't it?

[17:03:03.0000] <jamesr>
:not? no

[17:04:00.0000] <Hixie>
who on earth doesn't support it?

[17:04:01.0000] <Hixie>
i guess it was only acid3

[17:06:00.0000] <jamesr>
:not() in css3-selectors is defined as only taking simple selectors, div.bq is not a simple selector

[17:06:01.0000] <jamesr>
dunno who supports non-simple selectors in :not() these days

[17:06:02.0000] <Hixie>
oh, right

[17:06:03.0000] <Hixie>
it'd just be :not(.bq) > blockquote { }

[18:20:00.0000] <Philip`>
Is Gmail now saying that every email sent by a Gmail user to the WHATWG list is phishing?

[18:20:01.0000] <Philip`>
(as a change from its old approach of saying every email from @google.com was phishing)

[18:24:00.0000] <zewt>
yeah i just started seeing that today

[18:24:01.0000] <zewt>
not every mails, but enough to be problematic

[18:24:02.0000] <zewt>
(every mails? english is hard)

[18:30:00.0000] <Philip`>
(every mail)

[18:30:01.0000] <Philip`>
(english are hard)

[18:31:00.0000] <Philip`>
((Not the latter))

[18:32:00.0000] <zewt>
(i was puzzling at what strange series of edits I must have made to land on "every mails", heh)

[18:33:00.0000] <Philip`>
/me realises he is half asleep, and should go the rest of the way

[18:33:01.0000] <jamesr>
it's pretty hyperactive

[18:34:00.0000] <jamesr>
i'm sure someone is tuning it

[18:34:01.0000] <zewt>
gmail suddenly started cramming way too much info in every from: section, too

[19:29:00.0000] <karlcow>
http://blog.stateless.co/post/6246070973/how-offline-web-apps-should-work

[20:05:00.0000] <roc>
karlcow: I believe a big problem with that proposal is that it doesn't allow for safe updates. How does a Web author ensure that if a user visits a new version of the Web app, manages to download half of the resources and then goes offline, they still have a usable version of the app for offline use?

[20:12:00.0000] <roc>
I'm a bit worried about the increasing use of Function attributes instead of real DOM event listeners

[20:12:01.0000] <karlcow>
roc: in which ways localStorage prevent a partial update? (question, because I do not know)

[20:13:00.0000] <roc>
localStorage a) sucked and b) is nothing to do with this really

[20:13:01.0000] <karlcow>
ooops

[20:13:02.0000] <roc>
the feature he's talking about is the HTML5 "appcache"

[20:13:03.0000] <karlcow>
I meant the appcache

[20:13:04.0000] <karlcow>
23:22 here

[20:14:00.0000] <roc>
I was talking to you at 4am my time last night, so deal with it :-)

[20:14:01.0000] <karlcow>
heh

[20:14:02.0000] <karlcow>
but you are a roc(k)

[20:14:03.0000] <roc>
in my head, anyway

[20:14:04.0000] <roc>
appcache is specifically designed to ensure that you always get a consistent version

[20:14:05.0000] <roc>
it ensures that app resources are always loaded from the cache, not the network

[20:15:00.0000] <roc>
when the manifest changes, it reloads all the resources in the background, while servicing requests from the old version of the app

[20:15:01.0000] <karlcow>
I see

[20:15:02.0000] <roc>
when all the resources for the new version have downloaded, it atomically switches over

[20:15:03.0000] <roc>
that's one reason you need a manifest

[20:16:00.0000] <roc>
There may be ways to improve the appcache --- I'd be glad if there are --- but that guy hasn't quite thought through the problem enough

[20:22:00.0000] <karlcow>
hmm krijn server has hiccup

[20:45:00.0000] <karlcow>
http://ariya.blogspot.com/2011/06/your-webkit-port-is-special-just-like.html

[21:08:00.0000] <Hixie>
according to http://caniuse.com/#search=column css3 column support is pretty widespread

[21:08:01.0000] <Hixie>
but when i try to use it i find a bunch of stuff isn't implemented

[21:08:02.0000] <Hixie>
e.g. column-span

[21:08:03.0000] <Hixie>
what's the story?

[21:08:04.0000] <Hixie>
spec is fiction, or caniuse is fiction?

[21:09:00.0000] <benschwarz>
Hixie, I'm not sure. But even the 'best' support for css columns is pretty lax. Subtle differences between browsers makes it tricky to use too

[21:09:01.0000] <Hixie>
are there any rants about this online i could look at?

[21:12:00.0000] <benschwarz>
Hixie, nothing that comes to mind

[21:12:01.0000] <Hixie>
k

[21:13:00.0000] <benschwarz>
in actuality, I don't think a lot of people are using them right now

[21:13:01.0000] <benschwarz>
columns of text, for the most part, don't really work on screen. better in print

[21:20:00.0000] <MikeSmith>
Hixie: I think you have to use vendor prefixes for the column stuff still

[21:21:00.0000] <roc>
we don't support column-span

[21:21:01.0000] <roc>
and we require vendor prefix

[21:21:02.0000] <MikeSmith>
Hixie: e.g., http://platform.html5.org/style.css

[21:21:03.0000] <Hixie>
yeah caniuse mentions the prefix

[21:21:04.0000] <Hixie>
roc: is it planned? or is the spec in the weeds?

[21:21:05.0000] <roc>
the spec's OK

[21:22:00.0000] <roc>
we need to check our implementation

[21:22:01.0000] <Hixie>
k

[21:22:02.0000] <roc>
and implement column-span, which is not so easy

[21:22:03.0000] <Hixie>
talking of css specs, anyone know how realistic http://dev.w3.org/csswg/css3-background/ is?

[21:23:00.0000] <Hixie>
it looks quite... featureful

[21:25:00.0000] <roc>
it's mostly supported by most browsers

[21:32:00.0000] <karlcow>
the best support for columns is currently in opera, though with -o-

[21:33:00.0000] <Hixie>
in what sense is it better than webkit or gecko's support?

[21:33:01.0000] <Hixie>
it doesn't seem to support span either

[21:33:02.0000] <Hixie>
nor honour -o-column-width?

[21:34:00.0000] <Hixie>
what version should i be using?

[21:40:00.0000] <MikeSmith>
Opera seems to support span, afaict

[21:40:01.0000] <Hixie>
what version?

[21:40:02.0000] <Hixie>
i couldn't get it to work at all

[21:40:03.0000] <MikeSmith>
11.11 and Opera.next

[21:41:00.0000] <Hixie>
i'm on 11.50

[21:41:01.0000] <MikeSmith>
Hixie: I'm using it on http://platform.html5.org/

[21:42:00.0000] <MikeSmith>
Opera also support break-* properties for columns

[21:43:00.0000] <MikeSmith>
e.g., break-inside: avoid and break-after: avoid

[21:43:01.0000] <Hixie>
what am i doing wrong? this works in webkit and gecko but not opera for me http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1053

[21:43:02.0000] <Hixie>
i'm on mac, opera 11.50

[21:43:03.0000] <MikeSmith>
Hixie: span works in my Opera 11.50 on Mac

[21:43:04.0000] <MikeSmith>
/me looks

[21:44:00.0000] <MikeSmith>
Hixie: just do column-count for Opera

[21:44:01.0000] <MikeSmith>
you don't need the vendor prefix

[21:45:00.0000] <Hixie>
oh i thought karl said i had to use -o-

[21:45:01.0000] <Hixie>
k

[21:45:02.0000] <Hixie>
ah, yes, that works

[21:45:03.0000] <MikeSmith>
cool

[21:47:00.0000] <Hixie>
wtf now i can't get -span to work in opera either

[21:47:01.0000] <Hixie>
yet it works in your doc

[21:48:00.0000] <Hixie>
hm, it works in this case

[21:49:00.0000] <Hixie>
weird

[21:50:00.0000] <Hixie>
it works if i use <h1> but not if i use <span>?

[21:50:01.0000] <Hixie>
ah, it only applies to block-level elements

[21:50:02.0000] <Hixie>
i assumed it made them block-level

[22:37:00.0000] <Hixie>
hm, webkit does -span as well

[22:45:00.0000] <zewt>
weirdo layout hack in gmail's mail index header, wonder when they started doing that ... looks really busted when you scroll around and it visibly changes modes

[00:17:00.0000] <hsivonen>
/me wonders what WOW-playing Opera users think of the download progress bar WOW quip

[00:33:00.0000] <zcorpan>
searching for "opera wow" on twitter didn't bring up much about WoW

[00:37:00.0000] <krijn>
karlcow: Yeah, I know, router and/or network cable to the second floor is broken. Will try to fix it tomorrow

[00:43:00.0000] <zcorpan>
hsivonen: http://www.sitepoint.com/forums/html-xhtml-52/link-anchor-doesnt-seem-working-ff4-ff5-766794.html "Someone also pointed out that in about:config, setting "html5.parser.enable" to "false", makes this go away."

[00:43:01.0000] <zcorpan>
/me doesn't know what the bug # is

[01:05:00.0000] <zcorpan>
boblet: what do you mean with "excludes footer"?

[01:07:00.0000] <zcorpan>
boblet: "Content inside a blockquote must be quoted from another source" is a conformance requirement. so if you put content inside blockquote that is not from another source, it makes your document non-conforming

[01:08:00.0000] <zcorpan>
AryehGregor: blockquote is sectioning root

[01:09:00.0000] <zcorpan>
maybe i should reply to teh email instead

[01:09:01.0000] <boblet>
zcorpan: that should have been stated not felt huh

[01:10:00.0000] <boblet>
(took me a while to find where I’d written that :)

[01:10:01.0000] <zcorpan>
boblet: where did i state that?

[01:10:02.0000] <boblet>
however spec for footer says it’s allowed in sectioning root elements, and blockquote is a sectioning root element

[01:10:03.0000] <zcorpan>
indeed, i don't recall stating that blockquote must not contain footer

[01:11:00.0000] <boblet>
I meant in that while footer is able to be quoted in blockquote, it’s not able to be used as a footer for blockquote

[01:11:01.0000] <MikeSmith>
zcorpan: that is, not allowed to contain the attribution, only the quote itself

[01:12:00.0000] <zcorpan>
yeah ok

[01:12:01.0000] <boblet>
I’ve been thinking more about it, and am writing an email about the use cases the current spec makes difficult or excludes

[01:13:00.0000] <boblet>
will look forward to your feedback on it :) hopefully tomorrow…

[01:13:01.0000] <boblet>
biab

[01:46:00.0000] <asmodai>
Any of my regular contacts in here interested in testing Google+?

[01:46:01.0000] <annevk>
I am

[01:48:00.0000] <matjas>
AryehGregor: can you chime in here? https://bugzilla.mozilla.org/show_bug.cgi?id=389348#c6

[01:48:01.0000] <annevk>
it seems I'm in already

[01:48:02.0000] <annevk>
thanks arve!

[01:50:00.0000] <asmodai>
lol

[01:51:00.0000] <asmodai>
Step out for 4 mins to talk to a colleague and annevk already gets invited, haha

[01:51:01.0000] <david_carlisle>
hsivonen:  how many extra names does IE have, are they listed somewhere?

[01:54:00.0000] <hsivonen>
david_carlisle: my vague recollection is that there are more than 5 but fewer than 20.

[01:55:00.0000] <hsivonen>
david_carlisle: the one that I remember is called &pdf; but that ungooglable, so I have trouble finding a list

[01:55:01.0000] <david_carlisle>
hsivonen: grmble, hopefully we'll decide not to do anything, then i can go back to sleep

[01:56:00.0000] <annevk>
we already decided that last time this came up

[01:56:01.0000] <annevk>
not sure what new information there is

[01:56:02.0000] <hsivonen>
annevk: yes we did, but I don't recall the rationale.

[01:56:03.0000] <hsivonen>
annevk: had I recalled the rationale, I'd have WONTFIXed the bug already

[01:56:04.0000] <annevk>
IE has entities for bidirectional characters

[01:57:00.0000] <hsivonen>
annevk: also, I'm not sure I agreed with the rationale last time

[01:57:01.0000] <hsivonen>
still, I don't like flipflopping on stuff like this forever

[01:58:00.0000] <annevk>
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-March/014125.html

[01:58:01.0000] <annevk>
"

[01:58:02.0000] <annevk>
I'm very skeptical about introducing entities for the codes that are

[01:58:03.0000] <annevk>
redundant with dir="" and <bdo> (namely, lre, lro, pdf, rle, rlo).

[01:58:04.0000] <annevk>
I don't know enough about the others to have an educated opinion. I can

[01:58:05.0000] <annevk>
set up a search to examine the data in more detail."

[01:59:00.0000] <annevk>
david_carlisle, in that email you can also find a list, just search for "pdf"

[01:59:01.0000] <hsivonen>
annevk: dir="" and <bdo> don't work inside the title attribute

[01:59:02.0000] <david_carlisle>
annevk: thanks for the link

[02:05:00.0000] <zcorpan>
Hixie: do you look at bugs in their order of priority?

[02:38:00.0000] <asmodai>
hah, awesome: http://grantgalitz.org/get_the_hell_out/LegendOfZelda_Links_Awakening/

[02:40:00.0000] <zcorpan>
i love that game

[02:43:00.0000] <matjas>
is it possible to disable rich text in contenteditable? /inb4 “just use <textarea>”

[02:45:00.0000] <zcorpan>
seems it doesn't have the select glitch

[03:32:00.0000] <hsivonen>
zcorpan: thanks for the sitepoint link

[03:32:01.0000] <hsivonen>
zcorpan: I have no idea what's going on but I trust bz's guess: https://bugzilla.mozilla.org/show_bug.cgi?id=668213#c5

[03:43:00.0000] <bdc>
PM `matjas`

[04:45:00.0000] <annevk>
Ms2ger, any chance I can get commit access to specification-data?

[04:45:01.0000] <annevk>
Ms2ger, I am going to try to do Progress Events with Anolis to see if I can make it work, but I need to add some references

[04:46:00.0000] <annevk>
Ms2ger, currently it does not even have XHR

[04:46:01.0000] <Ms2ger>
I thought you had access already, fixed now

[04:46:02.0000] <Ms2ger>
Btw, you noticed anolis can create those files for you?

[04:47:00.0000] <annevk>
I saw something about it creating the index for terms and such, but not sure how it works

[04:50:00.0000] <Ms2ger>
anolis --dump-xrefs=data/xrefs/dom/xhr.json Overview.src.html /tmp/whatever would update xhr.json

[04:50:01.0000] <Ms2ger>
(You probably need a stub with the URL first)

[04:52:00.0000] <annevk>
interesting

[05:06:00.0000] <annevk>
Ms2ger, I updated references.json with an XHR entry but I still get

[05:06:01.0000] <annevk>
"SyntaxError: Specification not found: xhr."

[05:07:00.0000] <annevk>
does specs.data also need an update?

[05:07:01.0000] <annevk>
hmm

[05:07:02.0000] <Ms2ger>
specs.json is for cross-spec cross-references, references.json is for the references section

[05:17:00.0000] <annevk>
the error reporting is not so great

[05:17:01.0000] <annevk>
lots of times make just says nothing and I get an empty Overview.html

[05:17:02.0000] <zcorpan>
so http://people.opera.com/miket/2011/6/viewer.html is kinda broken if the user uses teh reload button after viewing an image

[05:18:00.0000] <Ms2ger>
It should print an exception at least

[05:18:01.0000] <Ms2ger>
But patches very welcome ;)

[05:19:00.0000] <zcorpan>
would it be worth it to work around this misuse of the api so that if you reload a page that has used pushState, and the new page doesn't have an event listener for popstate, when you navigate back the browser uses the previous page (the one that used pushState) instead of staying on the current one

[05:19:01.0000] <annevk>
I think there are no exceptions for missing reference entries

[05:19:02.0000] <annevk>
it just fails silently

[05:20:00.0000] <annevk>
or maybe once it has thrown an exception it won't throw it again the next time you run make

[05:21:00.0000] <Ms2ger>
Strange, there is "raise SyntaxError, "Reference not defined: %s." % ref" in the code

[05:29:00.0000] <annevk>
http://dev.w3.org/2006/webapi/progress/ now uses Anolis

[05:29:01.0000] <annevk>
I guess I should check in my changes to specification-data

[05:30:00.0000] <Ms2ger>
Please do

[05:31:00.0000] <annevk>
where can you put a global hg file with username info and such?

[05:32:00.0000] <Ms2ger>
~/.hgrc

[05:33:00.0000] <annevk>
https://bitbucket.org/ms2ger/specification-data/changeset/8390f4084a7a

[05:33:01.0000] <Ms2ger>
Lovely

[05:40:00.0000] <hsivonen>
how does Roboform integrate into Opera versions that didn't support extensions?

[05:40:01.0000] <hsivonen>
it claims compat with Opera 7 through 11

[05:40:02.0000] <hsivonen>
does it do some tricky native code injection without API support from Opera?

[05:45:00.0000] <annevk>
Ms2ger, can you use Anolis from Python easily?

[05:48:00.0000] <Ms2ger>
Depends on what you call easily, I guess

[05:48:01.0000] <Ms2ger>
I'd look at http://hg.hoppipolla.co.uk/hgwebdir.cgi/aquarium/file/ebf539280b86/aquarium.py#l74

[05:57:00.0000] <annevk>
I need to do some pre-processing for XMLHttpRequest

[05:57:01.0000] <annevk>
either that needs to move to the Makefile or the Makefile needs to move to Python

[05:57:02.0000] <annevk>
or I guess I can create a bash script that combines them...

[06:01:00.0000] <Ms2ger>
Oh, right

[06:02:00.0000] <annevk>
that last option is prolly the easiest

[06:07:00.0000] <Ms2ger>
annevk, maybe something like this would work? http://pastebin.com/XAMfTyCS

[06:10:00.0000] <annevk>
I actually want to generate both copies at the same time

[06:10:01.0000] <annevk>
but I guess that's a step in that direction

[06:10:02.0000] <annevk>
Ms2ger, how do I do http://dev.w3.org/2006/webapi/XMLHttpRequest-2/#ref-httpverbsec ?

[06:11:00.0000] <Ms2ger>
Wait for me to implement it, I'm afraid

[06:11:01.0000] <Ms2ger>
/me looks

[06:12:00.0000] <annevk>
I sort of like that for XML too

[06:12:01.0000] <annevk>
because you never want XML without XMLNS

[06:12:02.0000] <annevk>
at least not for specifications that matter, afaict

[06:22:00.0000] <Ms2ger>
/me tests

[06:24:00.0000] <Ms2ger>
annevk, { "HTTPVERBSEC": [{first link}, {second link}, {third link}] } should work if you update anolis

[06:25:00.0000] <annevk>
sweet

[06:25:01.0000] <annevk>
I'm adding a bunch of missing references first

[06:29:00.0000] <Ms2ger>
Looks like Vlad doesn't edit Typed Arrays anymore

[06:29:01.0000] <annevk>
Ms2ger, you mean after "href" and "title" right?

[06:29:02.0000] <zcorpan>
seems like some users don't want to update firefox because add-ons and other things break - http://www.sitepoint.com/forums/4908549-post15.html (search for "Actually, two reasons:")

[06:30:00.0000] <annevk>
Ms2ger, I will update that right away

[06:33:00.0000] <Ms2ger>
annevk, http://pastebin.com/C9VbyrWK

[06:37:00.0000] <annevk>
thanks

[06:50:00.0000] <annevk>
matjas, even though whitespace is not allowed it can still be made to work

[06:51:00.0000] <annevk>
matjas, so you should escape it appropriately on http://mothereffingcssescapes.com/ I think besides giving a warning

[06:52:00.0000] <hsivonen>
so IE10 PP2 has msRequestAnimationFrame

[06:52:01.0000] <hsivonen>
what's the Opera situation with that one?

[06:52:02.0000] <annevk>
we don't

[06:53:00.0000] <hsivonen>
sucks for Web authors (again) that this stuff isn't frozen and unprefixed faster

[06:56:00.0000] <Rik`>
hsivonen: the first shipped implementation is only 3 months old

[06:57:00.0000] <hsivonen>
Rik`: it's not like there's so many design decisions here

[06:58:00.0000] <Rik`>
hsivonen: the webkit guys proposed a second argument to determine if the element is displayed

[06:58:01.0000] <hsivonen>
I wonder if MS is interested in supporting MathML. they have pretty cool other layout goodies in 10PP2

[06:58:02.0000] <hsivonen>
Rik`: that kind of change would have been backwards-compatible even if the method had been unprefixed already

[07:00:00.0000] <matjas>
annevk: done, thanks!

[07:00:01.0000] <hsivonen>
in cases like these, prefixing mainly serves to disadvantage browsers that authors didn't bother to support at a given point in time

[07:00:02.0000] <hsivonen>
for example, if Opera adds an o-prefixed version, it will suck that code written for moz and webkit prefixes will go to a fallback setTimout path

[07:01:00.0000] <hsivonen>
even worse if a new engine wanted to break into the market

[07:02:00.0000] <hsivonen>
also, the point of prefixing is defeated if authors anticipate the o-prefixed and unprefixed versions to behave like the existing impls and write code anticipating that

[07:04:00.0000] <Rik`>
yes but the alternative is "first to implement choose the API"

[07:05:00.0000] <gsnedders>
hsivonen: And what we do now is often end up as the first to drop the prefix, which hurts too

[07:05:01.0000] <hsivonen>
Rik`: that happens anyway if Web authors write code that anticipates the unprefixed version working like the first prefixed version

[07:07:00.0000] <matjas>
annevk: wait, not done. Line breaks won’t work when simply escaped through \. Unicoads here I come!!1

[07:07:01.0000] <Rik`>
hsivonen: no, people who used the unprefixed version of the webkit gradients will not see it working

[07:07:02.0000] <hsivonen>
Rik`: when someone with paul_irish's visibility has blogged http://paulirish.com/2011/requestanimationframe-for-smart-animating/ , it's game over for incompatible changes unless browsers are willing to break cool new legacy content

[07:08:00.0000] <david_carlisle>
hsivonen: I wonder about that too:-)

[07:09:00.0000] <smaug____>
browser vendors should always be willing to break prefixed APIs

[07:10:00.0000] <hsivonen>
smaug____: note that copy-pasteable boilerplate in that post anticipates the future existence of compatible unprefixed version, o-prefixed version and ms-prefixed version

[07:10:01.0000] <hsivonen>
so far, guessing that an ms-prefixed version with the same behavior will emerge has paid off!

[07:11:00.0000] <hsivonen>
it will be interesiting to see what happens with flexbox

[07:11:01.0000] <hsivonen>
I changed prefixes to -ms- on code that worked in Gecko and WebKit using -moz- and -webkit- and got different results

[07:12:00.0000] <hsivonen>
(yes, I'm aware that TabAtkins has made actual spec changes after -moz- and -webkit- flexbox impls were made)

[07:12:01.0000] <annevk>
Ms2ger, so I'm trying to populate a newly generated progress-events.json but it does not work

[07:23:00.0000] <hsivonen>
fwiw, I'll be away from the internets next week, so even if someone sorts out the incomplete registrations at http://wiki.whatwg.org/wiki/MetaExtensions , deployment on Validator.nu will have to way until the 11th of July

[07:27:00.0000] <annevk>
Ms2ger, we should change <var> so that it never xrefs

[07:31:00.0000] <hsivonen>
looks like the link that the Webtrends regs should have had is http://product.webtrends.com/WRC/7.0/Documents/TechRef.pdf

[07:53:00.0000] <annevk>
converting XHR is a little more work than expected :/

[07:53:01.0000] <annevk>
it was a feature people have requested for ages though so hopefully it is worth it

[08:26:00.0000] <TabAtkins>
hsivonen: IE10 will have old flexbox behavior.  Webkit is switching to new flexbox behavior.

[08:27:00.0000] <TabAtkins>
Of course, both -moz and -webkit flexbox were sufficiently different/buggy that you couldn't just spam the prefixes and get the same result in non-trivial cases.

[08:31:00.0000] <annevk>
Ms2ger, anolis seems to strip _ from terms

[08:32:00.0000] <annevk>
Ms2ger, this creates problems for e.g. SYNTAX_ERR which becomes syntaxerr but is syntax_err in xrefs

[08:34:00.0000] <annevk>
I could hack Anolis but I have no commit rights

[08:39:00.0000] <annevk>
Ms2ger, I have a fix

[08:39:01.0000] <smaug____>
TabAtkins: just curious, any estimate whether the new flexbox model will be stable, or will it be changed again in few months?

[08:39:02.0000] <annevk>
-non_alphanumeric_spaces = re.compile(r"[^a-zA-Z0-9 \-]+")

[08:39:03.0000] <annevk>
+non_alphanumeric_spaces = re.compile(r"[^a-zA-Z0-9 \-\_]+")

[08:39:04.0000] <annevk>
Ms2ger, ^^ in the xref files

[08:40:00.0000] <smaug____>
so, what is the reason for <command> element?

[08:42:00.0000] <TabAtkins>
smaug____: The underlying model is stable, though I'm re-adding the multiline feature today or next week.  Syntax may still be tweaked, but that's it.

[08:45:00.0000] <smaug____>
I guess there is only one reason for <command>; to have icon with checkbox or radiobutton

[08:45:01.0000] <smaug____>
kind of weak reason

[08:47:00.0000] <AryehGregor>
boblet, zcorpan: Oops.  I should have read the spec before I commented.  :)

[08:48:00.0000] <AryehGregor>
(on blockquote as sectioning root)

[08:50:00.0000] <AryehGregor>
matjas, I'll look.

[09:01:00.0000] <hsivonen>
TabAtkins: apparently the cases I've used (or, well, that Paul Rouget has used and that I've copied) have been simple enough to have the same flex stuff with -moz- and -webkit-

[09:04:00.0000] <paul_irish>
hsivonen: btw i hear ya on rAF.. i wrote that post as a way to gather feedback on the spec primarily.. secondarily i wanted to share a shim that was a bit smarter than existing ones that were circulating. it does put vendors in weird place though as the script assumes certain behavior.. :/

[09:04:01.0000] <paul_irish>
Few devs understand that the existence of a prefix means By Using this You are entering into a Contract where things may break and it is Your Fault if they do. :)

[09:05:00.0000] <paul_irish>
and then i'd also suggest many things stay prefixed too long anyhow, so that implication loses its authority over time

[09:06:00.0000] <paul_irish>
not sure what the right solution is.. but i do get the vibe most devs dont know where/how to submit the community feedback that is desired and implied by the prefix

[09:20:00.0000] <hsivonen>
paul_irish: I think the right solution is to allocate more people-time paid by browser vendors to spec writing

[09:20:01.0000] <hsivonen>
paul_irish: to get specs in order sooner

[09:21:00.0000] <paul_irish>
+1

[09:21:01.0000] <TabAtkins>
+1

[09:21:02.0000] <TabAtkins>
/me could use some more spec writers to help out.

[09:28:00.0000] <smaug____>
not only spec writers, but also spec reviewers

[09:30:00.0000] <MikeSmith>
speaking of spec review, would be nice to get more implementor feedback on the Notifications spec

[09:31:00.0000] <MikeSmith>
we had estimated we could take that to LC this month

[09:31:01.0000] <MikeSmith>
or maybe even it was CR

[09:31:02.0000] <annevk>
John should fix the things Hixie raised I guess

[09:32:00.0000] <MikeSmith>
need to ping him I guess

[09:35:00.0000] <MikeSmith>
whoah, Alastair left Bocoup

[10:14:00.0000] <TabAtkins>
Hm.  Is multipage partially broken?  Looks like http://www.whatwg.org/specs/web-apps/current-work/complete/rendering.html is coming through as an empty doc.

[10:40:00.0000] <annevk>
Ms2ger, if you are still there, need to allow / too

[10:41:00.0000] <annevk>
-non_alphanumeric_spaces = re.compile(r"[^a-zA-Z0-9 \-]+")

[10:41:01.0000] <annevk>
+non_alphanumeric_spaces = re.compile(r"[^a-zA-Z0-9 \-\_\/]+")

[10:41:02.0000] <annevk>
Ms2ger, ^^ is the new diff

[10:45:00.0000] <AnselmBradford>
Does anyone know the current status of the keygen element support in IE? Is this post still the current sentiment of MS? http://blog.whatwg.org/this-week-in-html5-episode-35 e.g. no intention of support in IE?

[10:49:00.0000] <AryehGregor>
AnselmBradford, that's the last I heard, yeah.

[10:49:01.0000] <AryehGregor>
Everyone seems to agree keygen stinks.

[10:49:02.0000] <AryehGregor>
But no one cares enough to replace it.

[10:51:00.0000] <AnselmBradford>
AryehGregor: thanks, seemed to be the sentiment I was reading

[10:51:01.0000] <AnselmBradford>
just wanted to double-check

[10:52:00.0000] <AnselmBradford>
you reckon it will be replaced anytime soon? Or who knows...

[10:53:00.0000] <TabAtkins>
AnselmBradford: It'll be replaced when someone cares enough to.  ^_^  Which no one does currently, so...

[10:53:01.0000] <AnselmBradford>
TabAtkins: heh, fair enough

[11:26:00.0000] <annevk>
http://dev.w3.org/2006/webapi/XMLHttpRequest-2/ and http://dev.w3.org/2006/webapi/XMLHttpRequest/ use Anolis as well now

[13:26:00.0000] <smaug____>
ahaa, accessKey handling is underspecified

[13:26:01.0000] <smaug____>
/me needs to review and file a bug

[14:01:00.0000] <Hixie>
annevk-cloud: yt?

[14:01:01.0000] <Hixie>
oh hm nm

[14:02:00.0000] <Hixie>
annevk-cloud: should http://dvcs.w3.org/hg/from-origin/raw-file/tip/Overview.html just be merged into the Fetch algorithm?

[14:15:00.0000] <Hixie>
i'm going to start making web pages which, isntead of using native controls or canvas or svg or whatever, draw everything pixel by pixel using one radio button per pixel and CSS to set the radio button pixels' colours

[14:15:01.0000] <Hixie>
i demand that we add accessibility apis to support this authoring model!

[14:15:02.0000] <Hixie>
why do you all hate accessibility! you're makign it imposssible for me to write accessible web pages using radio buttons for pixels!

[14:22:00.0000] <The_8472>
Hixie, i prefer my "canvas fallback through 1x1px table cells" approach

[14:23:00.0000] <Hixie>
dude authors would never use table cells for this kind of thing, it's clear that that's bad for accessibility. they will use radio buttons like me! which is why we should have a radio button pixel accessibility api!

[14:24:00.0000] <The_8472>
we could... embed radio buttons in table cells for maximum compatibility AND accessability?

[14:29:00.0000] <Hixie>
TabAtkins: if you could get some traction on http://code.google.com/p/chromium/issues/detail?id=80911&q=bzbarsky&colspec=ID%20Stars%20Pri%20Area%20Feature%20Type%20Status%20Summary%20Modified%20Owner%20Mstone%20OS that'd be fantastic; the difference is masking some potential issues with wf2 api changes

[14:29:01.0000] <jamesr>
google.com had table cell bitmaps for icons at one point briefly

[14:29:02.0000] <jamesr>
with judicious use of rowspan/colspan to cut down on the # of cells

[14:29:03.0000] <Hixie>
the shopping cart guys did that

[14:29:04.0000] <Hixie>
i had words

[14:29:05.0000] <jamesr>
yeah

[14:29:06.0000] <Hixie>
(not that they listened, but still)

[14:29:07.0000] <jamesr>
i think it was only live for a few days

[14:29:08.0000] <hober>
Hixie: :)

[14:30:00.0000] <The_8472>
 <jamesr> with judicious use of rowspan/colspan to cut down on the # of cells <- "let's implement something atrociously slow and horrible. and then optimize it a bit. it'll be perfectly fine!"

[14:30:01.0000] <zewt>
heh

[14:30:02.0000] <zewt>
the firmware update on my old wifi router actually uses a form submit button ... as a progress bar

[14:30:03.0000] <zewt>
o_O

[14:31:00.0000] <jamesr>
The_8472: i think the motivation was mainly to inline the image (to avoid another request) and support browsers that didn't handle data: urls

[14:31:01.0000] <The_8472>
oh. static images. i was thinking of animating stuff with tables!

[14:38:00.0000] <jamesr>
hah

[14:38:01.0000] <jamesr>
you could probably do something decent with paletted tables

[14:49:00.0000] <Ms2ger>
annevk-cloud, I'll look at it tomorrow if I don't forget

[15:58:00.0000] <jamesr>
is anyone here following jason weber's setImmediate proposal discussions closely?

[15:58:01.0000] <jamesr>
i feel like i have a different set of assumptions going in and i'm not exactly sure what they are

[16:06:00.0000] <Hixie>
jamesr: http://www.w3.org/mid/CAD73md+Zp8J2bbdgTqPnK7FYW9A8ZHAEttz_6-H16g6sPKumUA⊙mgc is the crux of the matter as far as i can tell

[16:06:01.0000] <jamesr>
well, yeah

[16:07:00.0000] <jamesr>
that's definitely an issue

[16:07:01.0000] <TabAtkins>
So, hm.  How many of the use-cases for setImmediate can be solved by Workers?

[16:07:02.0000] <Hixie>
jamesr: i don't understand what the proposal achieves other than that, and as you imply, it doesn't even achieve that

[16:07:03.0000] <TabAtkins>
After all, you can just infinite-loop a Worker and go about your business.

[16:08:00.0000] <Hixie>
presumably this is for things that update the dom?

[16:08:01.0000] <Hixie>
but then why not use settimeout-0

[16:08:02.0000] <zewt>
(well, workers still have the problem of not being able to receive events during a processing loop; maybe I should give that thread a kick at some point...)

[16:09:00.0000] <jamesr>
zewt: true

[16:09:01.0000] <jamesr>
we could unclamp settimeout() in workers

[16:10:00.0000] <jamesr>
or rely on authors setting the timer early, like in my example code

[16:10:01.0000] <TabAtkins>
Hixie: You can just send a message to something in the page that'll update the dom for you.

[16:10:02.0000] <TabAtkins>
jamesr: No need to setTimeout at all in workers.

[16:10:03.0000] <jamesr>
TabAtkins: you do in order to process incoming messages

[16:10:04.0000] <Hixie>
TabAtkins: not if the work to update the dom is the expensive work

[16:10:05.0000] <jamesr>
or window.postMessage() yourself to yield

[16:10:06.0000] <zewt>
my proposal was to add a way to explicitly, synchronously receive a message from a MessagePort in workers; don't remember where that left off (just trailed off, iirc)

[16:10:07.0000] <jamesr>
which is equivalent to setImmediate

[16:11:00.0000] <TabAtkins>
jamesr: Ah, yes, recieving messages woudl require that.  If you're just sending them, you don't.

[16:11:01.0000] <TabAtkins>
But yeah, unclamping in a Worker would work, since the primary use-case for the clamp is "keep the UI thread from being hung".

[16:11:02.0000] <jamesr>
it's more to keep really dumb pages from spinning 100%

[16:11:03.0000] <jamesr>
with an unclamped setTimeout() you can still service UI stuff

[16:12:00.0000] <TabAtkins>
Hixie: But it'll get done in separate invocations of the message handler, right?

[16:12:01.0000] <zewt>
jamesr: iirc, an objection to that was that it may be hard to actually guarantee the message loop spins fully when a 0ms timer is looping

[16:12:02.0000] <zewt>
(don't remember clearly, discussion was a while ago)

[16:12:03.0000] <jamesr>
smells like QoI

[16:12:04.0000] <zewt>
er, the event loop

[16:12:05.0000] <Hixie>
TabAtkins: oh, i see what you're saying

[16:13:00.0000] <Hixie>
TabAtkins: yeah, you could do that. it's a bit non-intuitive.

[16:13:01.0000] <jamesr>
the event loop impl in chrome is very carefully designed not to starve any task source (both the HTML-defined ones and the implementation-internal ones)

[16:13:02.0000] <Hixie>
TabAtkins: think of a case where you need to crawl a 5MB document and search for text nodes containing some text

[16:13:03.0000] <Hixie>
TabAtkins: can't really do that in a worker, and you need to yield

[16:13:04.0000] <TabAtkins>
jamesr: Oh, if that's the major use-case, then setImmediate does nothing.  If someone was spinning the CPU with setTimeout(0), there's absolutely no reason they wouldn't spin it with setImmediate too.

[16:13:05.0000] <Hixie>
TabAtkins: (settimeout 0 seems fine for that in practice, to me)

[16:13:06.0000] <TabAtkins>
Hixie: That's true, and yes.

[16:13:07.0000] <jamesr>
TabAtkins: that's my primary objection

[16:14:00.0000] <Hixie>
anyway i don't really see the point of setImmediate()

[16:14:01.0000] <Hixie>
it'll never be possible to keep it unclamped

[16:14:02.0000] <Hixie>
at which point, it's the same as setTimeout()

[16:14:03.0000] <Hixie>
so...

[16:14:04.0000] <TabAtkins>
jamesr: I assumed the big reason was the whole "starving everybody else of events" thing, and because solving that with setTimout(0) can increase your running time by a non-trivial amount if you dont' split up the work large enough.

[16:15:00.0000] <zewt>
TabAtkins: there's an argument against that on that page: setTimeout(0) has never actually spun, since implementations have always clamped it (long before that was specced), and at least some people using setTimeout(0) are depending on that (probably without being aware of it)

[16:15:01.0000] <jamesr>
TabAtkins: then split your work better

[16:15:02.0000] <zewt>
TabAtkins: i certainly wouldn't argue that *some* people wouldn't spin stupidly with setImmediate, of course

[16:15:03.0000] <jamesr>
zewt: that's just not true, though. mozilla was unclamped, they noticed that pages were using 100% CPU, and they clamped

[16:15:04.0000] <zewt>
jamesr: because of people only testing in IE, is my first guess

[16:16:00.0000] <jamesr>
in chrome we had an unclamped setTimeout for a while (before release)

[16:16:01.0000] <jamesr>
couldn't go with it, of course, because too many pages spun the CPU too badly

[16:16:02.0000] <TabAtkins>
jamesr: Indeed.  Your code snippet was simple.  If it's a big problem, we can solve *that* case directly.

[16:16:03.0000] <jamesr>
people's macbook pros were setting their jeans on fire

[16:16:04.0000] <zewt>
(the set of people who would only test in IE and the set of people who would use a setTimeout(0) busy loop probably has a lot of overlap)

[16:16:05.0000] <TabAtkins>
With, like, a "yieldIfTooMuchTimeIsPassing()" function.

[16:16:06.0000] <Hixie>
well you wouldn't want a continuation

[16:17:00.0000] <Hixie>
(well, you would, but vendors wouldn't)

[16:17:01.0000] <TabAtkins>
"yieldIfTooMuchTimeHasPassedSinceTheLastTimeICalledThisFunction()"

[16:17:02.0000] <Hixie>
but the same principle applies

[16:17:03.0000] <Hixie>
hasTooMuchTimeElaphsed()

[16:17:04.0000] <TabAtkins>
Hixie: I certainly do.  And at least some of us webkit people do.

[16:17:05.0000] <Hixie>
really?

[16:17:06.0000] <Hixie>
dude the moment browser vendors add continuations to the platform, i'm speccing features left right and center to use them

[16:18:00.0000] <TabAtkins>
Yeah.  It's even in our Traceur thing.

[16:18:01.0000] <TabAtkins>
We support an "async" keyword that does simple continuations.

[16:18:02.0000] <Hixie>
all the async apis can go away if we have real continuations

[16:18:03.0000] <zewt>
what's the vendor objection to them? just too hard to handle when calling scripts from lots of different exit points?

[16:18:04.0000] <jamesr>
continuations are cool but they're a language level thing

[16:18:05.0000] <jamesr>
dunno if you can just slip them in without a lot of careful design

[16:18:06.0000] <TabAtkins>
Hixie: Indeed, continuations are ridiculously more usable than callbacks.

[16:18:07.0000] <Hixie>
anyway my assumption is that they're not coming any time soon

[16:19:00.0000] <Hixie>
but if that assumption is wrong, i look forward to being proven so :-)

[16:19:01.0000] <TabAtkins>
jamesr: You end up having to duplicate all the async methods into being task-based, but it's not too intrusive.

[16:20:00.0000] <TabAtkins>
Hixie: Tasks-via-continuations are just an inversation of iterator functions, which everyone wants to support (and FF already does).

[16:20:01.0000] <TabAtkins>
s/inversation/inversion/

[16:20:02.0000] <TabAtkins>
It's basically the same "convert the code into a state machine" transformation.

[16:21:00.0000] <Hixie>
i thought iterator functions were very tied to specific objects

[16:22:00.0000] <TabAtkins>
Just to make sure we're talkinga bout the same thing, I'm referring to Python-like functions that use "yield" to return values.

[16:22:01.0000] <Hixie>
yeah

[16:22:02.0000] <Hixie>
you can't just yield from a random function though, you can only yield from one that's known to be an iterator, right?

[16:22:03.0000] <TabAtkins>
Ok, then yeah, it's the same thing but inverted.  When you use an iterator function, the function is converted to a state machine.  When you use a task, *you're* converted to a state machine.

[16:23:00.0000] <jamesr>
TabAtkins: your mom is converted to a state machine

[16:23:01.0000] <TabAtkins>
And you magically transform into producing tasks as well.

[16:23:02.0000] <Hixie>
so what happens if you're in the middle of processing a sync event or one of these new mutation callbacks, and you call yield()?

[16:23:03.0000] <Hixie>
that sounds insanely painful

[16:23:04.0000] <TabAtkins>
I won't think about Jonas's mutation callbacks, because they're not good.

[16:24:00.0000] <TabAtkins>
I dunno how sync events are handled.  What's an example?

[16:24:01.0000] <Hixie>
ok, you're in the middle of a signalling channel callback, and you call yield()

[16:24:02.0000] <Hixie>
does the whole PeerConnection C++ implementation get "converted to a state machine"?

[16:24:03.0000] <zewt>
Hixie: well, there's how Python does it: if you use the yield keyword, then calling the function returns a generator object; which implies that you can only yield if the caller is expecting it, or the rest of the function just won't be executed

[16:25:00.0000] <zewt>
so the exit points that actually support yielding is controllable and not everything necessarily needs to

[16:25:01.0000] <TabAtkins>
Hixie: Dunno how PeerConnection works either.

[16:25:02.0000] <Hixie>
zewt: right, but that yields to the caller. here we're talking about yielding to the event loop. whole different kettle of fish.

[16:26:00.0000] <jamesr>
i think having continuations is nice but it doesn't really reduce the complexity of re-entrancy by that much

[16:26:01.0000] <zewt>
right, but the callback or event could specify whether it supports it

[16:26:02.0000] <TabAtkins>
But the function that uses the async keyword (to magically make an async event look sync in the code) gets transformed to returning a task instead of its normal return value.

[16:26:03.0000] <zewt>
(not sure just how that would work with events)

[16:26:04.0000] <TabAtkins>
And it returns that task as soon as the first "async" is hit.

[16:26:05.0000] <zewt>
tab: yeah, that sounds directly analogous to Python's model

[16:26:06.0000] <TabAtkins>
If the caller doesn't know how to handle tasks, it'll get hurt.

[16:27:00.0000] <TabAtkins>
But everything on the UA level *should* know how to do so.

[16:27:01.0000] <Hixie>
TabAtkins: oh so you're not just yielding to the event loop, you're just yielding to the parent, who then has to yield as well, etc

[16:27:02.0000] <TabAtkins>
Yes.

[16:27:03.0000] <Hixie>
oh well that's not so interesting

[16:27:04.0000] <Hixie>
that's just a generator

[16:27:05.0000] <zewt>
i think lua's model allows yielding all the way out, but it has a full coroutine system for that

[16:27:06.0000] <TabAtkins>
It still lets you write straight-line, callback-less code in many situations that would otherwise be messy.

[16:28:00.0000] <Hixie>
TabAtkins: not really. you have to have your entire callchain expect you to be yielding, it's not much better than having the whole chain use callbacks.

[16:28:01.0000] <TabAtkins>
Hixie: Not quite.  Actually implementing it with a generator involves more fiddliness.

[16:28:02.0000] <Hixie>
in other news, i just did a revamp of http://www.whatwg.org/specs/web-apps/current-work/complete.html#rendering to be organised by element, not property

[16:29:00.0000] <zewt>
Hixie: from my experience with Python, i disagree; having to yield up the call chain is annoying, but *way* better than callbacks or a direct state machine

[16:29:01.0000] <Hixie>
can y'all give it a quick sanity check?

[16:29:02.0000] <Hixie>
(in particular if anyone can reorder the stylesheet at the top of 15.2.5 Phrasing content to make more sense, let me know)

[16:29:03.0000] <Hixie>
(i couldn't come up with a good organisation for it)

[16:30:00.0000] <The_8472>
callbacks and coroutines are nearly the same if you do callbacks with closures and the closure syntax is so thin that you could just as well continue the method...

[16:31:00.0000] <jamesr>
Hixie: wait, 'expected' is a MUST-level requirement?

[16:32:00.0000] <jamesr>
i never realized that (i think i always have skipped to the individual entries and skipped the intro text)

[16:32:01.0000] <Hixie>
jamesr: it's a kind-of-must-requirement for certain classes of UAs

[16:34:00.0000] <The_8472>
iow: if you want to slap a "standards-compliant" label on your UA?

[16:34:01.0000] <zewt>
the: it's a significant difference in readability; being able to write algorithms linearly is, IMO, one of the major advantages of doing stuff in workers instead of with callbacks

[16:35:00.0000] <Hixie>
The_8472: you don't have to implement the rendering rules in general

[16:35:01.0000] <Hixie>
The_8472: you do if you want to be a mainstream browser ("user agent designated as supporting the suggested default rendering")

[16:36:00.0000] <The_8472>
that's very weasel-y language

[16:36:01.0000] <Hixie>
suggestions welcome on making it clearer

[16:36:02.0000] <Hixie>
TabAtkins: what's the opposite of "replaced element"?

[16:36:03.0000] <zewt>
(of course, on the same token I'd rather be able to offload more things into workers than worry about new ways to do complex tasks in the UI thread)

[16:36:04.0000] <Hixie>
TabAtkins: (in css)

[16:36:05.0000] <TabAtkins>
non-replaced element.

[16:36:06.0000] <Hixie>
bummer

[16:36:07.0000] <Hixie>
i was hoping for something more puthy

[16:36:08.0000] <Hixie>
pithy

[16:36:09.0000] <TabAtkins>
We rarely have to talk about that.

[16:37:00.0000] <TabAtkins>
Usually it's just general rules for all elements, and then exceptions for replaced.

[16:37:01.0000] <Hixie>
i have a section on replaced elements, following a section on everything else

[16:37:02.0000] <Hixie>
looking for a heading for that first section

[16:37:03.0000] <The_8472>
zewt, problem with workers is that none of the core components (DOM/layout/javascript) have been designed with multithreading in mind. that's why the webworker api is so useless

[16:37:04.0000] <TabAtkins>
Hixie: Got some suggestion on how to fix 15.2.5 to be more readable.  How you want the feedback?

[16:38:00.0000] <Hixie>
TabAtkins: whatever works for you

[16:38:01.0000] <zewt>
the: they've been designed with multithreading explicitly not in mind, which isn't quite the same thing

[16:38:02.0000] <TabAtkins>
I'll email, then.

[16:38:03.0000] <Hixie>
k, send it straight to ian⊙hc

[16:38:04.0000] <zewt>
(not quite the right way of putting it, but I think you know what I mean)

[16:39:00.0000] <Hixie>
TabAtkins: eta? (just trying to work out if i should check it in now and handle the feedback async, or if i should check google+ then handle your feedback :-) )

[16:39:01.0000] <The_8472>
yeah. the point was is that those design decisions (intentional or not) are what has lead to the web workers being so restricted

[16:39:02.0000] <TabAtkins>
Hixie: 5 min or less

[16:39:03.0000] <Hixie>
cool

[16:40:00.0000] <zewt>
the: yeah; some of those will hopefully eventually be relaxed, some will probably never be (eg. multithreaded access to the same DOM tree)

[16:42:00.0000] <The_8472>
snapshotting

[16:42:01.0000] <The_8472>
snapshot the dom tree, work on a read-only copy

[16:42:02.0000] <The_8472>
could be "good enough" for some things

[16:43:00.0000] <TabAtkins>
Hixie: Sent.  Just some slight reordering, and some redundancy added to make things group better.

[16:43:01.0000] <zewt>
maybe, sounds complex, though

[16:43:02.0000] <The_8472>
e.g. to collect all the data you need to generate a fragment, read the layout properties without risking layout reflows and then attach the generated fragment to the live dom

[16:43:03.0000] <zewt>
(how do you snapshot a WebGL context?)

[16:43:04.0000] <The_8472>
texture

[16:44:00.0000] <The_8472>
you'd need a "get me the live equivalent of this dom node" call of course

[16:44:01.0000] <Hixie>
TabAtkins: thanks

[16:44:02.0000] <nlogax>
STM DOM? :)

[16:45:00.0000] <Hixie>
TabAtkins: you think redundancy is ok? i've been avoiding it like the plague in this section for some reason

[16:45:01.0000] <TabAtkins>
Yes.

[16:45:02.0000] <Hixie>
k

[16:45:03.0000] <TabAtkins>
I always feel icky when I remove redundancy from my stylesheet just for the sake of removing redundancy.

[16:45:04.0000] <TabAtkins>
Selectors should communicate useful information, especially when grouped together.

[16:46:00.0000] <Hixie>
(by redundancy i meant multiple rules applying the same property with the same value to the same element; is that what you meant?)

[16:46:01.0000] <Hixie>
(i don't see any)

[16:47:00.0000] <Hixie>
i think you meant multiple rules giving the same property/value pair

[16:47:01.0000] <Hixie>
which seems ok

[16:47:02.0000] <TabAtkins>
No, I meant splitting up a selector into multiple decl blocks.

[16:47:03.0000] <TabAtkins>
Yeah.

[16:47:04.0000] <Hixie>
looks good

[16:48:00.0000] <Hixie>
regenning

[16:48:01.0000] <The_8472>
/me uses sass at work. the generated CSS can grow to humungous proportions if you don't pay attention... but it makes things so much easier

[16:49:00.0000] <The_8472>
tweak a few variables and the whole color scheme changes. gradients, text, borders... everything

[16:50:00.0000] <The_8472>
can only be a decade until we have that in browsers too :(

[16:50:01.0000] <TabAtkins>
I'm working on it.


2011-07-02
[23:42:00.0000] <annevk>
Hixie, yeah, maybe it should

[23:43:00.0000] <annevk>
Hixie, and at some point CORS too, and then someone should probably try to extract fetching into a separate core spec

[23:43:01.0000] <Hixie>
yeah

[23:43:02.0000] <Hixie>
let's do that all at once

[23:43:03.0000] <annevk>
:)

[23:43:04.0000] <Hixie>
but not right away :-)

[23:43:05.0000] <annevk>
yeah

[23:46:00.0000] <annevk>
anyone wants an invite for IRCCloud?

[23:46:01.0000] <annevk>
while in beta it is free

[23:53:00.0000] <Hixie>
what's IRCCloud?

[23:57:00.0000] <annevk>
https://irccloud.com/about

[23:58:00.0000] <annevk>
I will probably only use it for as long as krijn's server is down

[00:01:00.0000] <Hixie>
oh, i see

[00:01:01.0000] <Hixie>
just get some server somewhere you can put irssi and screen on

[00:04:00.0000] <twisted>
yo :) I got a weird thing going on in Gecko browsers on a page concerning css. It's weird cause the same CSS works on other pages, http://gemetria.bastardoperatorfromhell.org/~jakoury/index.php?p=nieuws is the website, http://gemetria.bastardoperatorfromhell.org/~twisted/skitch/firefox-20110702-025016.png is what it looks like in Firefox 4 and http://gemetria.bastardoperatorfromhell.org/~twisted/skitch/Chrome-20110702-025039.png that's in Chrome what it's s

[00:05:00.0000] <twisted>
someone over in #css helped me track it down to #top-footer:after {}

[00:06:00.0000] <twisted>
it's *white* the moment I swap it to :before and add a dot in the content, but then the other pages break and the website gets 10px longer in Chrome (so does Firefox btw)

[00:07:00.0000] <twisted>
http://browsershots.org/http://gemetria.bastardoperatorfromhell.org/~jakoury/index.php?p=nieuws are some more browser screenshots :)

[00:51:00.0000] <annevk>
is there a way to set a base URL in a bash script so you can run it from any directory without issues?

[00:56:00.0000] <twisted>
annevk: what do you mean? a var?

[00:56:01.0000] <twisted>
annevk: http://pastebin.com/SgyW8kuc that idea?

[01:00:00.0000] <annevk>
my bash script has stuff like python file

[01:01:00.0000] <annevk>
and also "script" where "script" tries to read something from the current directory

[01:01:01.0000] <annevk>
however this currently fails if you call the bash script from a parent directory

[01:01:02.0000] <twisted>
the pwd command shows you what path your currently in

[01:02:00.0000] <twisted>
http://www.manpagez.com/man/1/pwd/

[01:02:01.0000] <annevk>
that's cool, but how I can set a path the bash script should act it is in?

[01:02:02.0000] <twisted>
a basepath var could work

[01:03:00.0000] <twisted>
if it starts with /

[01:03:01.0000] <twisted>
it's an absolute path

[01:03:02.0000] <twisted>
and can never change

[01:03:03.0000] <twisted>
so instead of: project/script/dir/bla you do /home/username/project/script/dir/bla

[01:03:04.0000] <twisted>
same dir, just more specific

[01:03:05.0000] <twisted>
~/project/script/dir/bla could also work

[01:04:00.0000] <twisted>
as long as the executing users homedir is actually /home/username

[01:04:01.0000] <twisted>
just curious btw, did you on accident or on purpose ask this in this channel? :)

[01:10:00.0000] <annevk>
on purpose :)

[01:10:01.0000] <annevk>
/me goes to try basepath

[01:13:00.0000] <zewt>
annevk: er ... cd? heh

[01:13:01.0000] <twisted>
annevk: just try: basepath="dir" \n exec="$basepath/script.py"

[01:14:00.0000] <annevk>
yeah see that does not work

[01:14:01.0000] <annevk>
some of the relatives are assumed by the script

[01:14:02.0000] <annevk>
zewt, yeah I could, but then I have to cd in and out all the time

[01:14:03.0000] <annevk>
I suppose I could move a few other things around instead

[01:15:00.0000] <zewt>
welcome to shell scripting, not exactly the bleeding edge of scripting :P

[01:15:01.0000] <twisted>
annevk: what does not work?

[01:16:00.0000] <zewt>
can do eg. dir=`pwd`; cd wherever; stuff "$dir/file"

[01:16:01.0000] <zewt>
or (cd wherever && stuff) to run in another directory without having to cd back every time

[01:18:00.0000] <annevk>
twisted, I cannot add $basepath for all relative links

[01:18:01.0000] <annevk>
twisted, not all relative links are in the bash script itself

[01:18:02.0000] <annevk>
I'll solve it another way

[01:18:03.0000] <annevk>
no worries

[01:18:04.0000] <twisted>
hmm

[01:19:00.0000] <twisted>
annevk: http://stackoverflow.com/questions/59895/can-a-bash-script-tell-what-directory-its-stored-in

[01:25:00.0000] <annevk>
Hixie, "Constructing the form data set" uses something called "submitter"

[01:25:01.0000] <annevk>
Hixie, can that be made more explicit so I can set that to null or something when I use that algorithm in XMLHttpRequest?

[01:41:00.0000] <annevk>
matjas, you filed a bug on the input event not firing, but there is no specification that requires it fires for contenteditable-areas

[01:41:01.0000] <annevk>
matjas, that is a WebKit extension, though I guess it is one we should adopt

[01:41:02.0000] <twisted>
anyone here btw an idea for my problem?

[01:42:00.0000] <twisted>
I posted it here 1hour 36min ago

[01:43:00.0000] <annevk>
/me looks

[01:46:00.0000] <annevk>
twisted, is it to clear floats?

[01:46:01.0000] <annevk>
twisted, the :after thingie?

[01:46:02.0000] <annevk>
twisted, could try http://annevankesteren.nl/2005/03/clearing-floats

[01:48:00.0000] <twisted>
annevk: well what I don't get is why it works on all the other pages

[01:48:01.0000] <twisted>
same code

[01:53:00.0000] <annevk>
maybe some conflicting CSS?

[01:55:00.0000] <twisted>
hmm

[01:56:00.0000] <twisted>
gonna have to firebug the hell out of it :p

[01:57:00.0000] <twisted>
it's weird cause if I swap it to :before it works

[01:58:00.0000] <twisted>
but then it shifts 10px down in total

[01:59:00.0000] <twisted>
it's suspicious...

[02:00:00.0000] <twisted>
ik trek wel ff de css door de validator

[02:02:00.0000] <twisted>
hmm... I used to have a reallly good css tidy app

[02:02:01.0000] <twisted>
I think it was called Top Style Pro

[02:03:00.0000] <twisted>
but it's windows only :(

[03:08:00.0000] <annevk>
whoa

[03:08:01.0000] <annevk>
it's alive

[03:09:00.0000] <annevk>
krijn, http://krijnhoetmer.nl/irc-logs/ is not updating it seems

[03:09:01.0000] <twisted>
haha funny to see how web dev's themselves usually have boring websites :p

[03:09:02.0000] <twisted>
'boring' in the sense how a designer would look at it

[03:10:00.0000] <annevk>
I was about to say, it's pretty interesting otherwise :)

[03:10:01.0000] <twisted>
for me, the only requirement of a website is that it works properly in lynx/links2

[03:11:00.0000] <twisted>
cause, _IF_ I'm booting into Linux and my X.org #%^@&# up again I can at least browse the web

[03:15:00.0000] <matjas>
annevk: I see. +1 on adopting that. Would you mind leaving a comment on the Fx bug as well, saying just that?

[03:15:01.0000] <krijnh>
Si the ping?

[03:15:02.0000] <krijnh>
Awesome

[03:16:00.0000] <krijnh>
Can someone send me the logs since the 27th?

[03:17:00.0000] <twisted>
very interesting

[03:17:01.0000] <twisted>
http://gemetria.bastardoperatorfromhell.org/~jakoury/index.php?p=portfolio

[03:18:00.0000] <twisted>
on that page I introduced the same bug now

[03:18:01.0000] <twisted>
by changing the padding-top from 1em to 0;

[03:18:02.0000] <twisted>
in #portDetail

[03:19:00.0000] <annevk>
Philip`, krijn needs logs

[03:19:01.0000] <matjas>
annevk: ty

[03:20:00.0000] <krijnh>
Philip`: firstnamelastname @ gmail.com :) Tia! I'm off now, but will update the logs when I get them

[03:20:01.0000] <twisted>
hmm

[03:20:02.0000] <twisted>
well, I 'fixed' it

[03:20:03.0000] <twisted>
but now I see a small... dot

[03:20:04.0000] <twisted>
lol

[03:21:00.0000] <twisted>
isn't there some invisible unicode char that I could use?

[03:21:01.0000] <twisted>
oh of course

[03:21:02.0000] <twisted>
*grr*

[03:22:00.0000] <twisted>
whoo :D

[03:22:01.0000] <twisted>
\o/

[03:22:02.0000] <twisted>
annevk: thanks eventually your http://www.positioniseverything.net/easyclearing.html reminded me to use visibility: hidden;

[03:22:03.0000] <twisted>
it now works

[03:22:04.0000] <twisted>
but, what I still don't understand why Firefox requires a :before

[03:23:00.0000] <twisted>
but until I introduced my new 'bug' into Chrome... Chrome didn't need a :before and now it does too

[03:23:01.0000] <twisted>
so cross-browser weirdness me thinks :p

[03:27:00.0000] <twisted>
is there by chance, a css validator that takes in account browser specific properties?

[03:29:00.0000] <twisted>
hmm maybe someone here can enlighten me, what exactly is the advantage of doing this: http://www.greywyvern.com/code/php/binary2base64

[03:30:00.0000] <twisted>
convert small to medium sized images to base64

[03:32:00.0000] <matjas>
annevk: I’m confused — http://www.whatwg.org/specs/web-apps/current-work/multipage/webappapis.html#event-handlers-on-elements,-document-objects,-and-window-objects lists `oninput` as one of the “event handlers (and their corresponding event handler event types) that must be supported by all HTML elements, as both content attributes and IDL attributes, and on Document and Window objects, as I

[03:32:01.0000] <matjas>
DL attributes”

[03:33:00.0000] <annevk>
event handlers are just that

[03:33:01.0000] <annevk>
event handlers

[03:33:02.0000] <annevk>
they mean nothing with respect to where and when such events are dispatched

[03:34:00.0000] <matjas>
I see

[03:34:01.0000] <matjas>
where does it say when `oninput` should be dispatched?

[03:35:00.0000] <annevk>
it says input is dispatched

[03:36:00.0000] <annevk>
search for "named input" to find instances

[03:37:00.0000] <matjas>
http://www.whatwg.org/specs/web-apps/current-work/multipage/common-input-element-attributes.html#event-input-input

[03:45:00.0000] <twisted>
http://procssor.com/ that is really sweet

[03:46:00.0000] <matjas>
annevk: filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=13118, thanks!

[03:47:00.0000] <annevk>
matjas, one of these days you should teach people about onevent versus event

[03:48:00.0000] <annevk>
matjas, you're not the first to make the mistake

[03:48:01.0000] <matjas>
event handlers vs. events

[03:48:02.0000] <matjas>
inorite

[03:50:00.0000] <matjas>
annevk: onevent = event handler, right? so the bug subject should’ve been “consider firing the input event…”?

[03:51:00.0000] <annevk>
yes

[03:52:00.0000] <annevk>
there's no such thing as onxxx events

[03:53:00.0000] <matjas>
annevk: maybe *you* should write about it, seeing as you blogged about the “ALT tag” and tags vs. elements before </hint>

[03:53:01.0000] <matjas>
/me doesn’t know shit

[03:56:00.0000] <annevk>
i wouldn't say that

[03:56:01.0000] <annevk>
but I guess I can, but not today

[03:56:02.0000] <annevk>
today is another Anolis day

[04:00:00.0000] <matjas>
annevk: all I can come up with is “`onclick` is an event handler. `click` is the event. When the event fires, the respective event handlers are executed.”

[04:06:00.0000] <matjas>
needs moar event handler content attributes

[05:11:00.0000] <smaug____>
what does "When the X binding applies to an Y" mean

[05:11:01.0000] <smaug____>
I mean, where is the CSS binding defined

[05:13:00.0000] <annevk>
I think it means there is some concept of an internal binding

[05:13:01.0000] <annevk>
At least, if you are talking about the HTML rendering section

[05:13:02.0000] <smaug____>
yeah,  I am

[05:14:00.0000] <smaug____>
I just don't see where for example "button" binding is defined

[05:14:01.0000] <smaug____>
ah, it is user-agent-defined value

[05:14:02.0000] <smaug____>
ok

[05:14:03.0000] <Ms2ger>
http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#introduction-15

[05:58:00.0000] <annevk>
what is up with all the validator bugs?

[05:58:01.0000] <annevk>
oh hey Ms2ger!

[05:59:00.0000] <Ms2ger>
Ohai :)

[06:00:00.0000] <annevk>
can you associate my name with my account on specification-data?

[06:01:00.0000] <annevk>
apparently I can't do that with my current rights

[06:02:00.0000] <Ms2ger>
I just did that

[06:04:00.0000] <Ms2ger>
And for anolis, I'd rather not take your patch, but I made xspecxref just ignore that regex

[06:04:01.0000] <annevk>
Ms2ger, also, Anolis seems to mangle the encoding problem

[06:05:00.0000] <annevk>
Ms2ger, Bjoern's name is not coming out right

[06:05:01.0000] <annevk>
(I removed the ascii config line)

[06:07:00.0000] <Ms2ger>
Maybe declaring your encoding would help?

[06:09:00.0000] <Ms2ger>
Also, I'd suggest xreffing case-sensitive/ASCII case-insensitive to domcore

[06:09:01.0000] <Ms2ger>
And fix the xref in Unless otherwise stated...

[06:11:00.0000] <annevk>
Where do you declare the encoding?

[06:15:00.0000] <Ms2ger>
It should pick up <meta charset> in the source

[06:16:00.0000] <annevk>
Can <!-- status --> be made to work or should I copy that somehow?

[06:18:00.0000] <Ms2ger>
What's <!-- status -->?

[06:19:00.0000] <annevk>
http://www.w3.org/2005/07/pubrules?uimode=filter#document-status

[06:20:00.0000] <annevk>
/me copies from CSSOM View

[06:23:00.0000] <Ms2ger>
Oh

[06:23:01.0000] <Ms2ger>
I could probably add that

[06:24:00.0000] <annevk>
a few things depend on the group

[06:24:01.0000] <Ms2ger>
Fun

[06:24:02.0000] <annevk>
I added it manually, not too much trouble

[06:40:00.0000] <annevk>
Ms2ger, if there are no non-normative references the code there does not validate

[06:41:00.0000] <annevk>
<dl></dl> does not work for the W3C

[06:57:00.0000] <Ms2ger>
Fixed, I think

[06:58:00.0000] <annevk>
confirmed

[06:59:00.0000] <Ms2ger>
I'm off now, I'll check back later if you found more bugs :)

[07:29:00.0000] <annevk>
hmm defining garbage collection policies sucks

[08:46:00.0000] <The_8472>
annevk, look at some JVMs. they're usually a good reference point

[08:47:00.0000] <The_8472>
assuming you're talking about GC for memory management and not some other kind of resources

[08:51:00.0000] <annevk>
for the XHR spec

[08:51:01.0000] <annevk>
and XHR in general

[10:56:00.0000] <annevk>
http://rniwa.com/editing/undomanager.html

[10:56:01.0000] <annevk>
via Google+

[11:03:00.0000] <annevk>
AryehGregor, Ms2ger, https://rniwa.com/2011-06-26/position-and-anchor-types/

[11:03:01.0000] <annevk>
Peter`, ^^

[11:31:00.0000] <annevk>
Ms2ger, so should CORS rely on DOM Core for case-sensitive and such?

[11:31:01.0000] <annevk>
or define those terms itself?

[11:32:00.0000] <Ms2ger>
I dunno

[12:08:00.0000] <Hixie>
annevk: file a bug, not around to edit right now

[13:32:00.0000] <zewt>
this floaty status bar fad in browsers seriously needs to end :| browsers leaking their ui into the page's space is terrible

[13:35:00.0000] <Philip`>
Browsers seem to like competing on how little UI space they take, then I guess they realise later they've actually got to display some information to the user so it squeezes out into the page

[15:33:00.0000] <erlehmann>
what is the earliest possible event handler a media element can seek by changing currentTime? canplay does not work.

[15:47:00.0000] <matjas>
Firefox and WebKit seem to convert \r into \n when used in an attribute value. Opera respects the \r. Does the spec say anything about this?

[15:47:01.0000] <matjas>
test: data:text/html;charset=utf-8,<p id%3D"%0D">

[15:47:02.0000] <matjas>
then: document.querySelector('p').id == '\r' // false

[15:47:03.0000] <matjas>
document.querySelector('p').id == '\n' // true

[15:48:00.0000] <bga_>
same with textarea.value iirc

[15:49:00.0000] <matjas>
bga_: yeah, seems to be the case for all attributes and their respective properties

[15:49:01.0000] <matjas>
bga_: it does however seem to respect the \r if you set it through JS

[15:53:00.0000] <gsnedders>
matjas: Everything becomes \n in the tokenizer

[15:53:01.0000] <bga_>
matjas also iirc webkit has issue w/ spaces {'\u2009': 1}['\u200A'] == 1

[15:53:02.0000] <bga_>
but now its fixed

[15:53:03.0000] <matjas>
gsnedders: so is it a bug in Opera then?

[15:55:00.0000] <gsnedders>
matjas: We don't claim to impl the HTML5 parser. :P I guess the HTML5 parser builds should match Fx/WebKit.

[15:56:00.0000] <annevk>
matjas, yes

[15:56:01.0000] <matjas>
I see, thanks for clarifying.

[15:56:02.0000] <annevk>
known bug at that

[16:01:00.0000] <matjas>
gsnedders: so only \r becomes \n, right? Or what do you mean by “everything”?

[16:02:00.0000] <gsnedders>
matjas: \r and \r\n

[16:03:00.0000] <matjas>
gsnedders: \r\n becomes a single \n, interesting

[16:10:00.0000] <matjas>
so when you set an attribute value to \r *through JS*, should it be normalized into \n or not?

[16:10:01.0000] <gsnedders>
matjas: With setAttribute?

[16:10:02.0000] <matjas>
gsnedders: for example

[16:10:03.0000] <gsnedders>
matjas: I think not.

[16:10:04.0000] <matjas>
gsnedders: does setting the property make a difference?

[16:10:05.0000] <gsnedders>
matjas: No

[16:10:06.0000] <matjas>
e.g. foo.id = '\r'

[16:11:00.0000] <matjas>
ok so tokenization doesn’t apply then

[16:11:01.0000] <gsnedders>
Right.

[16:12:00.0000] <bga_>
matjas i guess keep LF CR in attributes is inpractical

[16:12:01.0000] <matjas>
bga_: invalid, too :) was just wondering what’s up

[16:12:02.0000] <matjas>
noticed the “issue” on http://mothereffingcssescapes.com/

[16:17:00.0000] <AlexNRoss>
<a name="warp"></a>Poof! You're now here.<a href="#warp">Warp</a>    Is there an alternate method of doing this? W3C Validator says that the name attribute is obsolete.

[16:19:00.0000] <tomasf>
AlexNRoss: use the id attribute

[16:20:00.0000] <AlexNRoss>
tomasf: I'm not so sure that all browsers will attempt to go to it via id.

[16:24:00.0000] <tomasf>
AlexNRoss: probably not all, but all reasonably modern ones do, afaik.


2011-07-03
[21:11:00.0000] <shetech>
Hi, all. Just wanted to say "hi". I'm one of the (probably many by now) people writing a book on HTML5.

[21:26:00.0000] <Hixie>
shetech: welcome :-)

[21:27:00.0000] <othermaciej>
shetech: neat

[21:38:00.0000] <shetech>
thanks, Hixie and othermaciej!

[23:58:00.0000] <eboyjr>
Can I ask about typed arrays here?

[23:58:01.0000] <eboyjr>
Well should it?

[23:58:02.0000] <eboyjr>
should I*

[23:59:00.0000] <eboyjr>
I can't seem to figure out how to get a Uint32Array (with 64 elements) from an ArrayBuffer that is not large enough

[23:59:01.0000] <eboyjr>
I want it to be padded

[00:02:00.0000] <eboyjr>
Something akin to:

[00:02:01.0000] <eboyjr>
>> var u = new Uint8Array([1, 2, 3]); new Uint32Array(u.buffer, 0, 4);

[01:03:00.0000] <eboyjr>
Figured it out

[01:59:00.0000] <boblet>
hey all, how do you search for example code if you’re trying to document current web usage of an element?

[04:09:00.0000] <annevk>
krijn, /irc-logs/ is still not updating

[04:09:01.0000] <krijn>
Feature, not a bug

[04:26:00.0000] <Philip`>
krijn: (Did you get the copy of the missing logs?)

[04:27:00.0000] <krijn>
Yep, will add them today

[04:27:01.0000] <krijn>
Thanks!

[04:35:00.0000] <annevk>
latest "facts" on http://www.opera.com/ are pretty funny

[04:35:01.0000] <annevk>
didn't know that was still going on

[04:56:00.0000] <krijn>
Philip`: added

[07:58:00.0000] <Ms2ger`>
< annevk> Ms2ger, ...

[07:58:01.0000] <Ms2ger`>
< annevk> oh wait

[07:58:02.0000] <Ms2ger`>
< annevk> you will never read that

[07:58:03.0000] <Ms2ger`>
I did!

[08:18:00.0000] <Ms2ger`>
< hsivonen> paul_irish: I think the right solution is to allocate more people-time paid by browser vendors to spec writing

[08:18:01.0000] <Ms2ger`>
+1

[08:24:00.0000] <boblet>
or to write test cases

[08:28:00.0000] <Ms2ger`>
# [BEEP!]

[08:28:01.0000] <Ms2ger`>
boblet, s/write/submit/

[08:30:00.0000] <boblet>
well… <_< ok that too. but something tells me there’s at least an order of magnitude more unwritten text cases for HTML5 and CSS3 than written but unsubmitted ones

[10:45:00.0000] <AryehGregor>
Dear Firefox user,

[10:45:01.0000] <AryehGregor>
Some time ago you submitted a crash report and asked us to contact you when we had more info about your crash. We're happy to report that this crash has been fixed in Firefox.

[10:45:02.0000] <AryehGregor>
If you haven't already, please download and install the latest version of Firefox from http://mozilla.com to prevent this crash from happening again.

[10:45:03.0000] <AryehGregor>
That's nice, but there's no link to the info about the crash itself?

[10:45:04.0000] <AryehGregor>
Like a bug report?

[11:00:00.0000] <Ms2ger`>
No crash report either?

[11:03:00.0000] <AryehGregor>
No link to any info specific to the crash.

[12:12:00.0000] <zewt>
heh, resisting the urge to jump into the appcache thread to point out how absurd the "hashes break the web" argument is

[12:15:00.0000] <Ms2ger`>
But they do!!!!

[12:15:01.0000] <zewt>
must... not... take... bait


2011-07-04
[04:04:00.0000] <krijnh>
O hai

[04:18:00.0000] <dhx1>
<link rel="tag" ... /> - there doesn't seem to be much information in the spec on how the href attribute is handled. Does the last word of the href attribute become the tag name? (as per http://microformats.org/wiki/rel-tag)

[04:22:00.0000] <annevk>
I think it is invalid for <link>

[04:22:01.0000] <annevk>
or should be

[04:25:00.0000] <dhx1>
annevk: source is http://www.whatwg.org/specs/web-apps/current-work/#link-type-tag

[04:27:00.0000] <annevk>
oh sorry, the URL is the tag

[04:27:01.0000] <annevk>
my bad

[04:28:00.0000] <dhx1>
so in the case of href="http://example.com/example/blue", is the tag "blue" and the page that describes that tag "http://example.com/example/blue"?

[04:29:00.0000] <annevk>
the tag is "http://example.com/example/blue"

[04:31:00.0000] <dhx1>
so a client parsing these tags would probably need to load each URL, grab the <title> of the page (or similar) and display that as the tag name? seems a little vague at the moment

[04:32:00.0000] <annevk>
the tag is the link

[04:32:01.0000] <annevk>
what you display is up to you

[04:35:00.0000] <dhx1>
ok, thanks for clarifying

[04:48:00.0000] <foolip>
hsivonen, what do you make of this RDFa/Microdata TF?

[06:03:00.0000] <smaug____>
how do I activate an accesskey on Opera/Linux

[06:03:01.0000] <smaug____>
alt+shift+<key> doesn't seem to work

[06:03:02.0000] <Philip`>
Shift+escape, I think

[06:04:00.0000] <smaug____>
huh

[06:04:01.0000] <smaug____>
but works

[06:04:02.0000] <smaug____>
thanks

[06:20:00.0000] <annevk>
Can you also download Google web fonts or is it solely for use with their API?

[07:29:00.0000] <bga_>
gsnedders hi

[07:30:00.0000] <bga_>
my friend have found bug in opera 11.50 http://jsfiddle.net/XV8AU/

[07:32:00.0000] <bga_>
in 9.64 - ok

[07:39:00.0000] <annevk>
so per HTTP the following is invalid

[07:39:01.0000] <annevk>
Content-Type:text/html;

[07:40:00.0000] <annevk>
I doubt however anyone follows that

[07:40:01.0000] <annevk>
maybe I should raise an issue on that

[08:33:00.0000] <LoneStar99>
anyone here which can take a look at a canvas? and tell me what is worn?

[08:33:01.0000] <LoneStar99>
wrong?

[08:39:00.0000] <LoneStar99>
why is this image so bloated? http://bobc.in/im/drop/cc/tp/bcqo-0

[08:40:00.0000] <annevk>
your canvas is 150*300 and then you resized that to style="width:768px;height:1024px" with CSS

[08:40:01.0000] <annevk>
only natural that it will look odd

[08:41:00.0000] <annevk>
the width/height attributes of <canvas> are meaningful and set the <canvas> drawing area, use them ;)

[08:41:01.0000] <annevk>
whereas the CSS properties merely set how big the <canvas> should appear on screen

[08:42:00.0000] <timeless>
[15:28] <AryehGregor> It would actually simplify things if I could assume that a selection must be either wholly editable or not.  But it doesn't have

[08:42:01.0000] <timeless>
aryehgregor: as a user, i want to be able to copy what i see, i shouldn't be blocked from copying stuff just because every second letter is content editable...

[08:42:02.0000] <timeless>
zewt: which mail did gmail say might not have been sent by me? all of my mail is actually sent by timeless.bmo2, but gmail should know and trust that....

[08:42:03.0000] <timeless>
tabatkins: if i'm on metered bandwidth your sites which prefetch on mouse hover are incredibly expensive....

[09:02:00.0000] <LoneStar99>
annevk: how do you know my canvas is 150*300?

[09:02:01.0000] <annevk>
because that's the default if you do not set anything

[09:02:02.0000] <LoneStar99>
where do I set them?

[09:03:00.0000] <annevk>
using the width and height attributes

[09:03:01.0000] <annevk>
I said that earlier

[09:04:00.0000] <LoneStar99>
ok i understand, style:width:xx;height:yy; will not do the trick then...

[09:07:00.0000] <annevk>
they size the canvas, not the drawing area

[09:23:00.0000] <LoneStar99>
oh ok thanks

[09:40:00.0000] <LoneStar99>
annevk: i swear nobody every answers a question at #html5 damn mob controlled

[09:42:00.0000] <Philip`>
LoneStar99: HTML5 is obsolete anyway, we've moved on to unversioned HTML now

[09:43:00.0000] <Philip`>
Let the #html5 mob wallow in their outdatedness if they want

[09:45:00.0000] <LoneStar99>
Philip`: right on

[09:53:00.0000] <Kingdutch>
What's wrong with HTML5? :O

[09:53:01.0000] <Kingdutch>
It's only like just here

[09:54:00.0000] <wilhelm>
It's so 2007!

[09:55:00.0000] <annevk>
Kingdutch, read up: http://blog.whatwg.org/html-is-the-new-html5

[10:00:00.0000] <smaug____>
selection handling with :before and  :after is... "interesting" in Gecko and Webkit

[10:00:01.0000] <smaug____>
AryehGregor: are you perhaps spec'ing how selecting text in :before and :after should work?

[10:03:00.0000] <annevk>
when the render and DOM tree collapse is no fun

[10:16:00.0000] <Kingdutch>
Ah right ok

[10:16:01.0000] <Kingdutch>
Hence the <!document html> instead of html5

[10:16:02.0000] <Kingdutch>
I want to make something that can replace HTML

[10:16:03.0000] <Kingdutch>
Or well, more the thing of HTML JS and PHP

[10:17:00.0000] <Kingdutch>
Hence why the prototype would still in the end be serving HTML and JS

[10:17:01.0000] <Kingdutch>
My biggest problem is a lack of starting point =|

[10:17:02.0000] <Kingdutch>
But I guess this is not the right channel to discuss that

[10:43:00.0000] <erlehmann>
Kingdutch, if you write “<!document HTML>”, something is broken.

[10:44:00.0000] <erlehmann>
also, replace HTML. haha. hahahaha. ha.

[10:54:00.0000] <Kingdutch>
I meant <!doctype html>

[10:54:01.0000] <Kingdutch>
I was close :O

[13:58:00.0000] <annevk-cloud>
WHATWG with summer recess?

[13:59:00.0000] <Ms2ger>
'murricans are out celebrating

[13:59:01.0000] <annevk-cloud>
Oh right, their freedom and shit

[14:00:00.0000] <annevk-cloud>
Aren't they like country eighty-something on the freedom-index?

[14:00:01.0000] <Ms2ger>
Sounds rather high

[14:02:00.0000] <annevk-cloud>
I can only find http://en.wikipedia.org/wiki/Democracy_Index#2010_rankings so fast and there they are actually number seventeen

[14:03:00.0000] <annevk-cloud>
Mind you, .18 away from flawed democracy

[14:03:01.0000] <Ms2ger>
Meh, data

[14:04:00.0000] <Ms2ger>
/me prefers uninformed trolling

[14:07:00.0000] <annevk-cloud>
Will you come on Google+?

[14:08:00.0000] <zewt>
antisocial networking, nobody is allowed in

[14:14:00.0000] <annevk-cloud>
A, reddit explains what is going on outside the US: http://24.media.tumblr.com/tumblr_lnt6hbZlwD1qadvboo1_500.png

[14:24:00.0000] <karlcow>
I really wonder how they define democracies

[14:47:00.0000] <Dashiva>
karlcow: Go read the report, it explains it in detail

[14:54:00.0000] <karlcow>
Dashiva just the idea of democracy and King doesn't compute with me… French education :)

[14:57:00.0000] <Dashiva>
Well, much the same way a dictator saying "We're a democracy and staging fake elections" doesn't make a country a democracy, having a person with the title king doesn't necessarily make a country not a democracy

[14:57:01.0000] <Dashiva>
(Move that second " over four words)

[14:58:00.0000] <Kingdutch>
to the left

[14:58:01.0000] <Kingdutch>
you mean

[14:58:02.0000] <Kingdutch>
Cuz I was moving it to the right

[14:58:03.0000] <Kingdutch>
Which makes no sense at all

[15:07:00.0000] <smaug____>
menu/command handling needs to be re-designed, or at least cleaned up

[16:15:00.0000] <annevk-cloud>
smaug____, yeah, I never think that made a lot of sense to be honest

[16:15:01.0000] <annevk-cloud>
smaug____, especially the idea of menu being for system menus rather than typical dropdown menus websites have

[16:16:00.0000] <annevk-cloud>
bit out of touch with what authors needs

[16:16:01.0000] <annevk-cloud>
anyway, bedtime

[16:16:02.0000] <smaug____>
annevk-cloud: also the creation of menus is strange

[16:16:03.0000] <smaug____>
and when to use <command> and when something else etc.

[16:17:00.0000] <smaug____>
and yeah, it is late here too

[16:26:00.0000] <rroyog>
hi

[16:57:00.0000] <dglazkov>
is sXBL dead?

[16:57:01.0000] <heycam>
I think so

[16:58:00.0000] <dglazkov>
k

[16:58:01.0000] <dglazkov>
what's that stuff you're working on, heycam?

[16:59:00.0000] <heycam>
dglazkov, which stuff do you mean? :)


2011-07-05
[17:00:00.0000] <dglazkov>
heycam: the SVG shadow DOM stuff

[17:01:00.0000] <heycam>
dglazkov, so I did implement sXBL once

[17:02:00.0000] <heycam>
but I don't think it got used beyond me writing a couple of simple demos for it

[17:04:00.0000] <dglazkov>
heycam: where?

[17:04:01.0000] <heycam>
the demos themselves are probably lost in time

[17:04:02.0000] <heycam>
the implementation is in Apache Batik, still

[17:10:00.0000] <dglazkov>
/me calls heycam a "crazy person".

[17:11:00.0000] <heycam>
if by "crazy" you mean "naively thinking that sXBL was the future of componentisation of SVG content", then sure :)

[20:45:00.0000] <shetech>
As to democracies, in many ways the US still is, but in many critical ways, it's falling apart. Witness the Supreme Court decision to allow corporations to donate unlimited amounts to candidates. Scary, that. But. No political system is flawless, or 100% what it says it is. So we'll watch the fireworks anyway.

[20:45:01.0000] <shetech>
Enough off-topic stuff for me. I'm still a noob here.

[20:45:02.0000] <shetech>
:)

[20:46:00.0000] <kbrosnan>
#defocus is the general discussion channel

[20:47:00.0000] <shetech>
Heh. Thanks. I so do NOT want to go there and talk politics!

[20:47:01.0000] <shetech>
/me bursts into flames just thinking abou tit

[20:47:02.0000] <shetech>
about it, even

[20:47:03.0000] <shetech>
yoiks

[20:51:00.0000] <shetech>
And with that, I depart to drink wine, watch fireworks and people

[02:04:00.0000] <asmodai>
hsivonen: around by chance?

[02:05:00.0000] <asmodai>
hsivonen: Got a small display nit in Firefox 7 (compared to 5 and 6) and wonder if you might have insight if this is intended or not.

[03:25:00.0000] <annevk>
<progress>.max currently just reflects per the specification. That is a bug right?

[03:25:01.0000] <annevk>
Is it a known bug?

[05:03:00.0000] <smaug____>
annevk: what you mean with "... both have been corrected."

[05:03:01.0000] <smaug____>
does the spec allow link activation with untrusted click?

[05:03:02.0000] <smaug____>
if yes, then that is a very recent change

[05:05:00.0000] <annevk>
no it doesn't

[05:05:01.0000] <annevk>
that's the correction

[05:05:02.0000] <annevk>
you obviously think it should work, but Ian, Jonas and I disagree

[05:06:00.0000] <smaug____>
interesting that you want to spec something that no browser do

[05:07:00.0000] <smaug____>
hmm

[05:07:01.0000] <annevk>
I don't think all browsers interoperate actually

[05:07:02.0000] <smaug____>
<a> element's click handling does allow untrusted click

[05:08:00.0000] <smaug____>
hmm

[05:08:01.0000] <smaug____>
but ok, it is the activation behavior which is strangely spec'ed

[05:10:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=12230

[05:10:01.0000] <annevk>
apparently when I tested before in Gecko it did not work for <a>

[05:11:00.0000] <annevk>
actually, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/866 still does not work in Gecko but does in other browsers

[05:11:01.0000] <smaug____>
IE, Gecko, Webkit and Opera have : synthetic click events on links cause navigation

[05:11:02.0000] <smaug____>
Gecko changed recently

[05:14:00.0000] <annevk>
it's funny how people keep claiming DOM3 Events is ready

[05:14:01.0000] <annevk>
it doesn't seem ready at all

[05:14:02.0000] <annevk>
every time I look it there's some flaw in it

[05:19:00.0000] <annevk>
smaug____, I was testing in today's nightly

[05:19:01.0000] <annevk>
smaug____, and Gecko is not following the link

[05:19:02.0000] <smaug____>
Gecko does follow the link

[05:20:00.0000] <smaug____>
http://mozilla.pettay.fi/moztests/a_click.html

[05:20:01.0000] <smaug____>
click the first button

[05:21:00.0000] <annevk>
smaug____, it needs to be MouseEvent?

[05:21:01.0000] <smaug____>
it needs a mouse event yes

[05:21:02.0000] <annevk>
smaug____, see my link for an example that does not work in Gecko but works elsewhere

[05:21:03.0000] <smaug____>
aha, perhaps we should relax the restriction

[05:21:04.0000] <smaug____>
but that is not the point

[05:22:00.0000] <annevk>
the point is that DOM3 Events is a mess and that this behavior does not make much sense

[05:22:01.0000] <smaug____>
all the browsers follow the link when synthetic mouse click is dispatched

[05:22:02.0000] <smaug____>
this has nothing to do with DOM 3 Event

[05:22:03.0000] <smaug____>
the point is that HTML activation behavior is mess

[05:22:04.0000] <smaug____>
and doesn't reflect reality

[05:22:05.0000] <smaug____>
but does specify something which no browser implement

[05:23:00.0000] <annevk>
what browsers do now makes little sense

[05:23:01.0000] <annevk>
and is not relied upon

[05:23:02.0000] <annevk>
and browser vendors are open to change their impls to see if saner model can be made to work

[05:26:00.0000] <smaug____>
I don't understand why the current behavior doesn't make sense

[05:29:00.0000] <smaug____>
/me was going to think about command API today

[05:29:01.0000] <smaug____>
that has some good things, but also some very weird things in it

[05:58:00.0000] <wilhelm>
Are there any publicly available test suites for the drag and drop API?

[05:59:00.0000] <Ms2ger>
annevk, I believe Gecko doesn't match the spec for progress.max, I doubt it's been filed

[06:06:00.0000] <Ms2ger>
And I sure hope you're following the mutation events thread, because I tuned out a while ago

[06:07:00.0000] <annevk>
Ms2ger, I am sort of following hoping someone comes up with something that everyone agrees to and then I'll write it down

[06:08:00.0000] <Ms2ger>
Or that

[06:13:00.0000] <annevk>
filed a bug on <progress>.max

[06:30:00.0000] <annevk>
document.importNode(new DOMParser().parseFromString('<div xmlns="http://www.w3.org/1999/xhtml">Test</div>', 'text/xml').documentElement, true).tagName

[06:30:01.0000] <annevk>
hmm yeah we cover that

[06:39:00.0000] <Ms2ger>
Should be DIV, I guess?

[06:40:00.0000] <annevk>
yup

[06:45:00.0000] <annevk>
I wonder why there's no feedback from either WebKit or Gecko on <progress>.max

[06:45:01.0000] <annevk>
On setting to a "test" Gecko sets the content attribute to "NaN"; WebKit throws a NOT_SUPPORTED_ERR

[06:45:02.0000] <Rik`>
annevk: I can summon volkmar if you want :)

[06:45:03.0000] <annevk>
a string*

[06:45:04.0000] <annevk>
Rik`, thanks, I added him to the bug

[06:46:00.0000] <annevk>
Rik`, http://www.w3.org/Bugs/Public/show_bug.cgi?id=13139

[06:46:01.0000] <volkmar>
gni ?

[06:47:00.0000] <volkmar>
annevk: i think there is a bug filed

[06:48:00.0000] <Ms2ger>
Is w3c-test.org failing to update again?

[06:48:01.0000] <Ms2ger>
volkmar, wasn't that just about meter?

[06:48:02.0000] <volkmar>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=11939

[06:48:03.0000] <volkmar>
but WONTFIX FTW :)

[06:48:04.0000] <annevk>
sweet

[06:49:00.0000] <volkmar>
annevk: but last I checked, Webkit was following the specs

[06:50:00.0000] <annevk>
are you supposed to throw a NOT_SUPPORTED_ERR?

[06:50:01.0000] <annevk>
oh maybe you are

[06:51:00.0000] <annevk>
since it's NaN

[06:51:01.0000] <annevk>
however, for <progress max=0> WebKit returns 1

[06:51:02.0000] <annevk>
for .max

[06:53:00.0000] <volkmar>
annevk: on setting, Gecko is following the rules for reflecting a double attribute

[06:53:01.0000] <volkmar>
so we do not throw if the value isn't supported

[06:54:00.0000] <annevk>
"Except where otherwise specified, if an IDL attribute that is a floating point number type (double) is assigned an Infinity or Not-a-Number (NaN) value, a NOT_SUPPORTED_ERR exception must be raised."

[06:54:01.0000] <volkmar>
oups :)

[06:57:00.0000] <volkmar>
annevk: then that's a bug in Gecko

[06:57:01.0000] <volkmar>
and that's my fault :)

[06:58:00.0000] <annevk>
still not sure about WebKit's behavior

[06:59:00.0000] <volkmar>
though, Gecko isn't doing that for all double attributes

[06:59:01.0000] <volkmar>
it's not specific to progress.max

[06:59:02.0000] <Ms2ger>
/me expects a patch to reflect.js :)

[07:00:00.0000] <volkmar>
Ms2ger: might be :)

[07:00:01.0000] <annevk>
maybe the IDL attribute should be limited to positive integers?

[07:02:00.0000] <volkmar>
annevk: you mean unsigned long limited to non-negative numbers greater than zero?

[07:03:00.0000] <annevk>
guess that does not work here

[07:04:00.0000] <volkmar>
it could if we change progress.max type

[07:05:00.0000] <annevk>
the problem reported against Opera was <progress max=0>.max returning 0 rather than 1

[07:06:00.0000] <annevk>
it returns 1 in both Gecko and WebKit

[07:06:01.0000] <annevk>
so neither Gecko or WebKit follows the spec

[08:44:00.0000] <AryehGregor>
Ugh, why does Bugzilla have mid-air collisions for stupid stuff like someone adding a CC?

[08:47:00.0000] <AryehGregor>
smaug____, I'm not currently planning on speccing the details of how selecting various things work, at least not for a while to come.  It's really complicated, and it doesn't affect interop nearly as much as some of the other things I'm doing.

[08:47:01.0000] <AryehGregor>
As far as things like execCommand() are concerned, selections are purely DOM-based and generated content is ignored.

[08:47:02.0000] <AryehGregor>
Stringification of the selection should perhaps take account of it, but that's a minefield that I looked into for a while and then abandoned.

[08:49:00.0000] <smaug____>
AryehGregor: ok

[08:49:01.0000] <smaug____>
I was just doing some testing and noticed that selecting generated content works in some cases but not in others

[08:49:02.0000] <smaug____>
both in Gecko and Webkit

[08:50:00.0000] <smaug____>
Opera behaved consistently

[08:50:01.0000] <smaug____>
didn't try IE

[08:50:02.0000] <Ms2ger>
There are cases where that works in Gecko?

[08:51:00.0000] <smaug____>
let me find the testcase....

[08:53:00.0000] <smaug____>
Ms2ger: https://bug552707.bugzilla.mozilla.org/attachment.cgi?id=543790 press ctrl+a in Gecko

[08:54:00.0000] <smaug____>
also try webkit

[08:54:01.0000] <smaug____>
webkit's behavior is even more broken

[08:54:02.0000] <Ms2ger>
Huh

[08:54:03.0000] <Ms2ger>
Oh

[08:54:04.0000] <smaug____>
:)

[08:55:00.0000] <Ms2ger>
Compare with http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1059 in WebKit

[08:56:00.0000] <AryehGregor>
Why is my Aurora build at 6.0a2 even after repeated updates?  Isn't the "2" supposed to increment sometimes?

[08:56:01.0000] <Ms2ger>
No

[08:57:00.0000] <Ms2ger>
a1 == Nightly, a2 == Aurora, b1 == Beta

[08:58:00.0000] <AryehGregor>
Oh.

[08:58:01.0000] <AryehGregor>
Confusing.

[08:58:02.0000] <Ms2ger>
(At least, I think so)

[08:58:03.0000] <AryehGregor>
But I guess you were stuck because "a" and "b" were already used, and there are no letters in between.

[08:59:00.0000] <AryehGregor>
You could call nightlies x.0a, Aurora x.0au, beta x.0b.

[08:59:01.0000] <AryehGregor>
I guess the real info is the date in parentheses after the version.

[08:59:02.0000] <Ms2ger>
Also, I think UA sniffers wouldn't be happy with other letters

[09:05:00.0000] <AryehGregor>
Sigh.

[09:08:00.0000] <annevk>
teehee http://lists.w3.org/Archives/Public/public-html-xml/2011Jul/0000.html

[09:18:00.0000] <Ms2ger>
Hah, XSL:FO

[09:19:00.0000] <linclark>
in microdata, if you use itemref to point to an item that has an itemid, should that item's itemid be attached as an attribute of that item wherever it is referenced?

[09:22:00.0000] <annevk>
Ms2ger, CSS WG still takes it into consideration when naming properties and such

[09:22:01.0000] <annevk>
Ms2ger, it's real sad

[09:23:00.0000] <annevk>
(and potentially harmful)

[09:23:01.0000] <Ms2ger>
What direction do XSL:FO gradients go?

[09:26:00.0000] <annevk>
dunno, don't care

[09:30:00.0000] <AryehGregor>
Opera people: can I import from preferences from Opera to Opera Next?

[09:36:00.0000] <hasather>
AryehGregor: I think you have to do it manually by copying the profile dir

[09:36:01.0000] <AryehGregor>
hasather, k.

[09:51:00.0000] <linclark>
is Philip` foolip?

[09:52:00.0000] <linclark>
oh no

[09:52:01.0000] <foolip>
linclark, no, I'm Jägenstedt, Philip` is Taylor

[09:52:02.0000] <foolip>
we're both awesome, though

[09:52:03.0000] <linclark>
foolip: ha, good to know

[09:52:04.0000] <linclark>
foolip: I have a question about your microdata parser, do you have a minute?

[09:52:05.0000] <foolip>
sure

[09:53:00.0000] <linclark>
so when I use itemref to refer to an item and that item has an itemid, the itemid doesn't show up as an attribute for the item

[09:53:01.0000] <linclark>
foolip: is this part of the spec... or is it not necessarily part of the spec

[09:54:00.0000] <foolip>
are you trying to set itemid by itemref'ing another element that sets itemid?

[09:54:01.0000] <linclark>
yeah, basically

[09:54:02.0000] <foolip>
itemref is for adding properties only, it doesn't influence itemid or itemtype

[09:55:00.0000] <foolip>
of course, you can still have itemid and itemtype on the properties you add, if they are themselves items

[09:55:01.0000] <foolip>
but it doesn't mean "look here for all item* attributes"

[09:56:00.0000] <linclark>
foolip: ok cool, that does make sense... is it stated explicitly in the spec?

[09:56:01.0000] <linclark>
I wasn't able to find it if it is

[09:57:00.0000] <foolip>
linclark, yes, see http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#attr-itemref

[09:57:01.0000] <foolip>
also http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#crawl-the-properties for the full details about exactly how itemref is used

[09:58:00.0000] <foolip>
(note that the last part is likely to change in some way because Opera implemented it differently and asked Hixie to change it)

[09:58:01.0000] <foolip>
Hixie, or, did you interpret the last email as a request to change the spec? there's a testsuite on the way too, that tests for that behavior, if that might work as bait.

[10:00:00.0000] <linclark>
foolip: I would have read "add that element to pending" and then "Remove an element from pending and let current be that element." to mean that the itemref elements are parsed in the same way as the original item

[10:00:01.0000] <linclark>
but then, I've never written a parser ;)

[10:00:02.0000] <linclark>
but the attr-itemref does specify name-value pairs

[10:00:03.0000] <linclark>
so makes sense

[10:01:00.0000] <foolip>
linclark, they're crawled in the same way, but this crawling algorithm is only for finding properties, it doesnt' involve itemid etc

[10:01:01.0000] <linclark>
foolip: ah, cool, thanks for the help :)

[10:01:02.0000] <foolip>
np

[10:01:03.0000] <foolip>
doing anything fun with microdata?

[10:01:04.0000] <foolip>
marking up your cat collection just like in the examples, I presume?

[10:02:00.0000] <linclark>
foolip: haha, indeed... I did start work on http://drupal.org/project/microdata

[10:02:01.0000] <foolip>
linclark, ah, you're involved with drupal?

[10:02:02.0000] <linclark>
but I'm holding off on going into API review until the task force stuff is more clear

[10:02:03.0000] <foolip>
ah yes, I even read so in your Twitter profile just a few days ago

[10:03:00.0000] <foolip>
I wouldn't recommend waiting for the TF to come down with an answer to all of our worries

[10:03:01.0000] <linclark>
foolip: yeah, it's a tough decision... basically, it's hard to try to make all of this understandable to the general community

[10:03:02.0000] <foolip>
Microdata isn't going to change substantially since it's already implemented and shipped

[10:04:00.0000] <foolip>
but, of course drupal could wait it out in case Microdata turns out to be a turd nobody wants to play with

[10:04:01.0000] <foolip>
I'm sure there's lots of fun community politics to be played there as well :)

[10:05:00.0000] <linclark>
foolip: it definitely wouldn't be integrated into core for a while just based on the release cycle... so free to develop freely in contrib

[10:05:01.0000] <foolip>
sounds like fun hacking

[10:05:02.0000] <linclark>
foolip: indeed, though fortunately the Drupal community is generally convivial in their disagreements

[10:05:03.0000] <foolip>
linclark, is this you? https://plus.google.com/100047679106364060080/about

[10:06:00.0000] <linclark>
foolip: I also want to figure out how to parse microdata for use in aggregating data with the Views module

[10:06:01.0000] <foolip>
linclark, is it being parsed server-side?

[10:06:02.0000] <linclark>
foolip: yes, it will be

[10:06:03.0000] <foolip>
python?

[10:06:04.0000] <linclark>
foolip: PHP

[10:07:00.0000] <foolip>
hmm, is there even a HTML5 parser written in PHP yet?

[10:07:01.0000] <Ms2ger>
Boo ;)

[10:07:02.0000] <Ms2ger>
I think there was a PHP port of html5lib

[10:07:03.0000] <foolip>
hey look: http://soyrex.com/php-microdata/

[10:07:04.0000] <linclark>
oh hey, look at that!

[10:07:05.0000] <foolip>
I'd be surprised if it's up to date with the spec though

[10:13:00.0000] <linclark>
foolip: oh yeah, and that google account is mine, though I haven't yet dug into the Google+ stuff

[10:25:00.0000] <jwalden>
random gripe for Opera people here: it is absolutely infuriating that Ctrl+PgUp and Ctrl+PgDn don't switch tabs like they do in every other browser; and while I'm whingeing, let's throw in Ctrl+K not meaning search-bar, too

[10:25:01.0000] <jwalden>
:-)

[10:26:00.0000] <Ms2ger>
Yes

[10:26:01.0000] <Ms2ger>
/me rages

[10:28:00.0000] <AryehGregor>
/me agrees

[10:29:00.0000] <gsnedders>
jwalden: Those shortcuts vary across platform, IIRC

[10:29:01.0000] <AryehGregor>
At least on Linux, I observe the same behavior as jwalden: every browser but Opera supports it.

[10:30:00.0000] <AryehGregor>
I think IE9 on Windows doesn't support it either.

[10:30:01.0000] <jwalden>
pretty sure it does, I don't remember that pain point

[10:30:02.0000] <AryehGregor>
Doesn't on my laptop, just checked.

[10:30:03.0000] <Ms2ger>
It may be a linuxism

[10:30:04.0000] <AryehGregor>
Very possible.

[10:31:00.0000] <gsnedders>
AryehGregor: Unless you change the shortcuts :P

[10:31:01.0000] <AryehGregor>
That's a very Opera response.  :P

[10:32:00.0000] <gsnedders>
(seemingly ctrl+page{up,down} have done scroll viewport left/right for a while in Opera by default, so probably changing it would get people bitching)

[10:34:00.0000] <Philip`>
/me has got used to using 1/2 to switch tabs in Opera

[10:34:01.0000] <Philip`>
(which is fewer keypresses and therefore vastly superior)

[10:35:00.0000] <gsnedders>
/me has got used to using ctrl+page{up,down} to switch tabs in Opera

[10:35:01.0000] <David_Bradbury>
/me has gotten used to using the mouse to navigate the OS :(

[10:36:00.0000] <jwalden>
hum, indeed it doesn't in IE9; shows how much I do any browsing on Windows

[10:37:00.0000] <jwalden>
/me actually doesn't use Firefox much, somewhat out of a sense of "I should be familiar with what's out there", so he uses lots of different things now

[10:44:00.0000] <MikeSmith>
hmm, apparently mobile safari automatically creates hyperlinks for things that look like phone numbers

[10:45:00.0000] <Ms2ger>
I hear it does

[10:45:01.0000] <MikeSmith>
so if you go to, e.g., this part of the spec:

[10:45:02.0000] <MikeSmith>
http://www.whatwg.org/specs/web-apps/current-work/multipage/the-iframe-element.html#sourcing-in-band-text-tracks

[10:45:03.0000] <MikeSmith>
is creates a hyperlink for "4.8.10.12.2"

[10:46:00.0000] <AryehGregor>
Haha.

[10:50:00.0000] <foolip>
jwalden, agreed on both accounts and filed two bugs on Opera Desktop

[10:50:01.0000] <jwalden>
thanks :-)

[10:51:00.0000] <jwalden>
historically I've probably used Opera the least of the non-Firefox browsers, but the last week or two I've been using it a little harder, so running against that more

[10:51:01.0000] <foolip>
let's hope the desktop overlords agree with us

[10:52:00.0000] <jwalden>
shortcut compatibility was one of the big things Phoenix (!) got right, back in the day

[10:53:00.0000] <jwalden>
Safari's shortcuts on Mac were also Think-Differenty, but I used it little enough to not quite feel the pain

[10:57:00.0000] <MikeSmith>
I wonder why is it that developers of Web-platform game engines seem to also like to develop their own languages

[10:58:00.0000] <MikeSmith>
e.g., http://www.playmycode.com/docs/quby

[10:58:01.0000] <MikeSmith>
"JavaScript is missing lots of common features, such as classes. Quby includes these making programming games more straight forward. Its syntax is based on Ruby, a language well known to be easy to learn."

[11:00:00.0000] <zewt>
heh, somehow it's unsurprising that someone who would think it's a good idea to come up with another obscure language would also like ruby

[11:02:00.0000] <zewt>
Philip`: opera's numbers-to-switch-tabs thing is really, seriously obnoxious--it's intruding into keys that should really be standardized as owned by web pages

[11:03:00.0000] <zewt>
i try to use numbers as hotkeys in UI and I get opera users complaining that they collide

[11:09:00.0000] <Philip`>
MikeSmith: Developers of non-web game engines usually seem to like making up their own languages, too - I guess it's part of the general wheel-reinvention mindset, since they can focus on what matches a small team's personal preferences in the short term and it's not like they're going to have to bother supporting it 5 years later as they'll have started from scratch on a new engine or two

[11:10:00.0000] <zewt>
a small team would have to be crazy to make their own language, heh

[11:11:00.0000] <zewt>
it's ... not exactly a small time investment

[11:11:01.0000] <Ms2ger>
...Switzerland's Anti-PowerPoint Party (APPP)...

[11:13:00.0000] <beverloo>
hacking something together does not necessarily take a lot of time, maintainability often is an issue though

[11:14:00.0000] <beverloo>
companies like R* developed own internal languages (in their case, SCM), deployed it to a few (5-10) games and then switch to something different

[11:15:00.0000] <beverloo>
in their case Lua. Other companies use Pawn or Squirrel or something similar. It's just convenient for mission scripting

[11:16:00.0000] <zewt>
well, lua's the major reason hacking together a proprietary language is generally a bad idea :P

[11:16:01.0000] <beverloo>
lua's just another language which syntax you either like or dislike

[11:16:02.0000] <beverloo>
and mission scripting isn't a role for software engineers, apparently

[11:17:00.0000] <zewt>
there isn't much that competes with lua in its niche

[11:17:01.0000] <Philip`>
zewt: Toy languages are pretty easy to make

[11:18:00.0000] <zewt>
Philip`: but you don't want a toy language to develop production software

[11:18:01.0000] <Ms2ger>
Why not?

[11:19:00.0000] <zewt>
what?

[11:19:01.0000] <Philip`>
As long as you end up with a shippable product, who cares?

[11:19:02.0000] <beverloo>
zewt, Squirrel, Pawn, GameMonkey, Angelscript, just to name a few

[11:19:03.0000] <zewt>
beverloo: never heard of any of them

[11:19:04.0000] <beverloo>
Mozilla's David Anderson used to spent time working on Pawn

[11:20:00.0000] <beverloo>
they're used in various games, you just never see them

[11:20:01.0000] <beverloo>
fortunately, often :p

[11:20:02.0000] <zewt>
Philip`: because it's harder to end up with a shippable product with half-baked tools

[11:20:03.0000] <MikeSmith>
Philip`: I wonder how successful most game engines actually are. I mean, the Web-platform-based ones cropping up lately mostly seem like proofs of concept that just get used to develop demos

[11:22:00.0000] <beverloo>
MikeSmith, that depends on the type of game. Engines like Unreal are great for first-person shooters, whereas Rockstar made their own engines (Renderware in the past, now RAGE) for sandbox-like games

[11:22:01.0000] <MikeSmith>
I see

[11:22:02.0000] <beverloo>
http://en.wikipedia.org/wiki/Unreal_Engine

[11:22:03.0000] <Philip`>
zewt: It's easier to write half-baked tools than fully-baked tools though, particularly since you can get a half-baked tool up and running pretty quickly and then your artists/designers learn all the quirks and learn which buttons not press because they crash and then there's no need to improve the tools any further

[11:23:00.0000] <beverloo>
that's a major difference between the web and games; you ship a game and you're done with it. Perhaps an extension later on, but that often gets made by the same team

[11:23:01.0000] <zewt>
it's much easier and quicker to use existing fully-baked tools, especially when the people using them already know how they work

[11:26:00.0000] <zewt>
of course, for web games you're more limited--you really don't want to be layering another scripting language on top of JS, heh

[11:27:00.0000] <MikeSmith>
details about the Twitter backend at http://www.infoq.com/articles/twitter-java-use are interesting

[11:28:00.0000] <zewt>
ew

[11:28:01.0000] <zewt>
if my choices are ruby or java, I think I may choose jumping out the window

[11:29:00.0000] <MikeSmith>
heh

[11:33:00.0000] <Philip`>
zewt: Environments change frequently enough that I imagine it's rare to have fully-baked tools that you can drop in with no integration effort - you might be developing for some unusual CPU architecture or relying on some custom memory management scheme or whatever, and trying to port an existing scripting engine is a more daunting task than starting from scratch

[11:34:00.0000] <zewt>
eh, that's a stretch

[11:36:00.0000] <zewt>
maybe if you're developing for the SNES and you have 128K to work with, but I've used Lua on the PS2 without any difficulty

[11:39:00.0000] <Philip`>
Lua seems rare in its suitability, compared to most other real languages (Python, JS, etc) that are a much bigger pain to embed

[11:40:00.0000] <zewt>
well, lua is designed for it, python isn't

[11:46:00.0000] <Philip`>
Seems like choice is limited if you don't like Lua :-(

[14:39:00.0000] <AryehGregor>
smaug____, did you ever get responses from IE or WebKit on insertNode()?

[14:39:01.0000] <smaug____>
not yet

[14:39:02.0000] <AryehGregor>
k.

[14:40:00.0000] <smaug____>
well, some webkit developer just referred to Hixie's comment abou insertNode

[14:40:01.0000] <smaug____>
but nothing from IE yet


2011-07-06
[22:22:00.0000] <jamesr>
anyone around? looking at a canvas patch for webkit and it looks like firefox does the wrong thing for (stroke|fill)Text when maxWidth is specified as 0

[22:22:01.0000] <jamesr>
it does the draw as if the value was not specified at all

[22:24:00.0000] <jamesr>
wondering if it's just an oversight or if it's difficult to distinguish between not-specified numbers and '0' in that idl implementation

[22:24:01.0000] <roc>
I think it's our bug

[22:24:02.0000] <roc>
    if (aMaxWidth > 0 && totalWidth > aMaxWidth) {

[22:26:00.0000] <jamesr>
opera gets 0 right but not negative numbers

[22:26:01.0000] <jamesr>
roc: guessing aMaxWidth is already coerced to a numerical type at that point?

[22:31:00.0000] <roc>
yeah, but we can get around it

[22:31:01.0000] <roc>
you have to be able to, for drawImage anyway

[22:32:00.0000] <jamesr>
aight

[22:33:00.0000] <jamesr>
negative values throw a weird exception

[22:41:00.0000] <jamesr>
Philip`: yt?

[22:43:00.0000] <jamesr>
Philip`: if you get a chance to see this i think you should add some canvas tests for negative maxWidth, NaN maxWidth, and for globalCompositeOperation + text drawing. opera fails on the first, the second seems unspecified (although i might have missed a step in the conversion routine that fixes it) and the last fails pretty hardcore in webkit

[22:43:01.0000] <jamesr>
or if i can figure out the harness maybe i could write some tests

[22:46:00.0000] <jamesr>
roc: or do you happen to know how a specified value of NaN is supposed to behave when passed in to a function specified as taking  a double in IDL?

[22:47:00.0000] <roc>
That depends on the function

[22:47:01.0000] <jcranmer>
you throw an exception saying "double check your damn math"

[22:47:02.0000] <roc>
canvas generally just ignores the call

[22:47:03.0000] <roc>
we've had some debates about this inside and outside Mozilla

[22:47:04.0000] <roc>
I tend to favour ignoring the call, for a couple of reasons

[22:48:00.0000] <roc>
1) it's the bottom of the "this site 'works' in browser X, but not Firefox, you suck" slippery slope

[22:48:01.0000] <jamesr>
iirc olliej prefers that as well, his argument being that it's easy to get a NaN on an intermediate step of some complex calculation and it's harsh to throw an exception as that'll typically kill all draw calls that would happen afterwards

[22:48:02.0000] <roc>
that was my point #2 :-)

[22:49:00.0000] <roc>
that is especially true for coordinate and matrix parameters

[22:49:01.0000] <jamesr>
when you 'ignore the call' do you completely ignore it or still apply the globalCompositeOperation (i.e. do you clear out everything if the gCO was "copy")?

[22:49:02.0000] <roc>
we ignore it completely

[22:49:03.0000] <roc>
I could go either way on that

[01:26:00.0000] <annevk-cloud>
dglazkov, what is the difference between a DOM element and DOM content?

[01:26:01.0000] <annevk-cloud>
dglazkov, http://wiki.whatwg.org/wiki/Component_Model_Behavior_Attachment could use some examples :)

[02:27:00.0000] <boblet>
crazy q: is it possible to disable Moz’s HTML5 parser via JS for a site?

[02:29:00.0000] <annevk>
no

[02:30:00.0000] <boblet>
cool

[02:40:00.0000] <annevk>
Ms2ger, http://lists.w3.org/Archives/Public/www-archive/2011Jul/0003.html

[02:40:01.0000] <annevk>
... and http://lists.w3.org/Archives/Public/www-archive/2011Jul/0004.html

[02:50:00.0000] <boblet>
annevk: the questions we get at H5D, itellya ;)

[02:52:00.0000] <annevk>
I read a sitepoint discussion this morning where zcorpan was in discussion with a bunch of people believing HTML5 will fail soon

[02:56:00.0000] <Onderhond>
What was their alternative?

[03:01:00.0000] <annevk>
Something would come along...

[03:02:00.0000] <annevk>
The mutation events thread be long, geez

[03:38:00.0000] <annevk>
http://tools.ietf.org/html/draft-weber-iri-guidelines looks somewhat promising

[03:38:01.0000] <annevk>
though replacing the whole set of RFCs with a standard about URLs still makes more sense to me

[03:50:00.0000] <MikeSmith>
hmm, yeah, that doc looks good

[03:50:01.0000] <MikeSmith>
/me is behind on reading the public-iri list

[03:56:00.0000] <MikeSmith>
annevk: I added Cross-Origin Resource Embedding Exclusion to platform.html5.org

[03:57:00.0000] <MikeSmith>
but I think that spec should just be renamed "The From-Origin header"

[03:57:01.0000] <MikeSmith>
hmm, I guess I should also add CSP

[04:01:00.0000] <annevk>
yeah I gues

[04:02:00.0000] <annevk>
s

[04:02:01.0000] <annevk>
The long term plan is to take "fetch" from HTML, CORS, and COREE; and merge them all into a single document about fetching resources

[04:03:00.0000] <MikeSmith>
yeah

[04:03:01.0000] <MikeSmith>
that would be good

[04:50:00.0000] <annevk>
not sure if sylvia is here, but <? does not work in HTML

[04:51:00.0000] <Ms2ger>
How unfortunate.

[06:06:00.0000] <Ms2ger>
"4. Pre-processing Arbitrary Unicode Strings"

[06:06:01.0000] <Ms2ger>
/me wonders if that spec will ever manage to become a RFC

[06:07:00.0000] <annevk>
written by the chair of the group if I'm not mistaken

[06:07:01.0000] <annevk>
should have a chance

[06:39:00.0000] <annevk>
Anyone used http://prizes.org/ for getting people to write specifications?

[06:41:00.0000] <Onderhond>
Long shot (and off-topic).

[06:41:01.0000] <Onderhond>
But I remember a news site that used the right site of the browser to give an indication of content below the fold.

[06:41:02.0000] <Onderhond>
* right side

[06:41:03.0000] <Onderhond>
Any idea if this is a common pattern, what it's generally called and where to find it? :)

[06:42:00.0000] <annevk>
NY Times has/had such an overlay on the bottom right, not sure what it was for anymore

[06:45:00.0000] <Onderhond>
Shame they change their site a lot, I'll check archive.org :)

[08:07:00.0000] <Onderhond>
Is it feasible to use input[type=range] for a fixed set of values? (radius around a point of a map: 1km, 5km, 15km, 25km, 50km, 75km)

[08:08:00.0000] <annevk>
input type=radio and select are for that

[08:09:00.0000] <Onderhond>
Currently there's no way to define a fixed set on values.

[08:09:01.0000] <Onderhond>
Myeah, figured so.

[08:10:00.0000] <Onderhond>
Sadly "those in power" prefer a fancier (and screen estate-reducing) input type.

[08:10:01.0000] <annevk>
need shadow DOM

[08:10:02.0000] <annevk>
or XBL

[08:11:00.0000] <annevk>
or whatever will solve that

[08:11:01.0000] <bga_>
but http://www.whatwg.org/specs/web-forms/current-work/#the-datalist

[08:12:00.0000] <wilhelm>
Your use case makes sense, but you'll need a bit of additional logic. <input type='range' min='0' max='5' step='1'> plus a script that converts indexed values of 0-5 to 1-75km.

[08:14:00.0000] <Onderhond>
Hmmm, sounds reasonable too.

[08:17:00.0000] <annevk>
bga_, that's suggestions, you cannot limit the control to those values

[08:18:00.0000] <bga_>
ah

[08:18:01.0000] <Onderhond>
data-fixedVals="1,5,15,25,50,75" and use js to snap to those values ... I guess that would be an option ?

[08:19:00.0000] <annevk>
data-fixed-vals

[08:19:01.0000] <annevk>
becomes dataset.fixedVals

[08:19:02.0000] <annevk>
and that could work, bit hacky

[08:20:00.0000] <Onderhond>
kay, thanks for the input :)

[08:20:01.0000] <Onderhond>
(hah pun!)

[11:11:00.0000] <karlcow>
http://www.innoq.com/blog/st/2010/03/uddi_rip.html

[11:25:00.0000] <TabAtkins>
I think I just need to write off everyone with "Leif" in their name as bad.

[11:27:00.0000] <othermaciej>
you probably shouldn't judge a person as "bad" based on their participation in a standards group in any case...

[11:27:01.0000] <Ms2ger>
I've only needed to do that with one

[11:27:02.0000] <Ms2ger>
Or I didn't realize there were different people

[11:28:00.0000] <TabAtkins>
othermaciej: I'm not.  I'm judging them by their name.

[11:28:01.0000] <TabAtkins>
And I'm doing it to *all* people with "Leif" in their name.

[11:28:02.0000] <TabAtkins>
'cause I'm racist.

[11:28:03.0000] <TabAtkins>
namist.

[11:29:00.0000] <Ms2ger>
/me fires TabAtkins 

[11:29:01.0000] <TabAtkins>
YOU CAN'T FIRE ME, I QUIT MOZILLA

[11:30:00.0000] <TabAtkins>
/me thinks he remembers that Ms2ger is Moz.

[11:30:01.0000] <annevk>
Is that actually just for W3C purposes or are you meaningfully employed?

[11:30:02.0000] <Ms2ger>
Just W3C red tape

[11:36:00.0000] <miketaylr>
aww TabAtkins not all Leif's are bad

[11:37:00.0000] <miketaylr>
s/Leif's/Leifs/

[11:37:01.0000] <Ms2ger>
leiftaylr?

[11:37:02.0000] <TabAtkins>
HE'S A CRYPTO-LEIF

[11:37:03.0000] <TabAtkins>
GET HIM

[11:38:00.0000] <miketaylr>
indeed leif taylor http://www.flickr.com/photos/11691204@N00/5701898006/in/photostream

[11:38:01.0000] <Ms2ger>
/me was never really fond of the name "Tab" either

[11:38:02.0000] <miketaylr>
/me hides his babby from tab

[11:38:03.0000] <TabAtkins>
I challenge you to find Tabs that are worth vilifying.

[11:38:04.0000] <TabAtkins>
Besides my dad.

[11:42:00.0000] <Hixie>
right, what's next

[11:43:00.0000] <Hixie>
microdata, the command api thing, rtc, and dialogs are on my list

[11:43:01.0000] <Hixie>
am i missing anything?

[11:43:02.0000] <Ms2ger>
Meh, dialogs can wait

[11:44:00.0000] <Hixie>
let's start with microdata

[11:56:00.0000] <Hixie>
you have to love the way the w3c on the one hand cries about how forks are bad and that's why we shouldn't use a free license, and then on the flip side they seem to have no problem making forks.

[11:57:00.0000] <AryehGregor>
I thought we discussed.  Forks of W3C specs are what's bad, regardless of whether the W3C spec happened to come first or second.

[11:57:01.0000] <AryehGregor>
s/discussed/discussed this/

[11:57:02.0000] <Hixie>
clearly

[11:58:00.0000] <Hixie>
and if all the forks occur within w3.org, it's fine too

[11:58:01.0000] <AryehGregor>
Yes, because then they (the good guys) control all of them, so it's okay.

[11:58:02.0000] <Hixie>
so tired of dealing with people who put process ahead of the web

[11:58:03.0000] <AryehGregor>
The only problem is when untrusted evil people make forks of W3C specs.

[11:58:04.0000] <Ms2ger>
Also, generalizations

[12:02:00.0000] <Hixie>
actually i should rephrase that. it's not so much that people put process ahead of the web. It's that they think the process helps the web, despite evidence to the contrary.

[12:03:00.0000] <AryehGregor>
I was going to say something like that, but I've argued with you enough already about the motives of people who disagree with us.

[12:03:01.0000] <AryehGregor>
Lots of people think procedures are good and anarchy is bad, in web standards as in many other things.

[12:04:00.0000] <Ms2ger>
Given that people do care about the process, I'm not sure saying "whatever, publish it as a rec, svg got away with it" is helpful

[12:05:00.0000] <AryehGregor>
It comes up a lot in politics too.  And the administration of large organizations.

[12:06:00.0000] <Hixie>
Ms2ger: the process as a whole isn't being helpful

[12:07:00.0000] <Hixie>
Ms2ger: the right answer is to change it or drop it, but nobody at w3c that i've spoken to seems interested in even discussing considering it

[12:10:00.0000] <AryehGregor>
Because to change the process so massively would be incredibly hard, because of the process.

[12:10:01.0000] <AryehGregor>
Since you'd have to go through the process to change the process.

[12:10:02.0000] <AryehGregor>
So they only tolerate superficial patches that don't really solve the problems.

[12:10:03.0000] <Hixie>
not doing something because it would be hard is possibly the _worst_ reason not to do it

[12:10:04.0000] <AryehGregor>
Hey, you're preaching to the choir here.

[12:11:00.0000] <Hixie>
sorry, just frustrated :-)

[12:11:01.0000] <AryehGregor>
But realistically, the AC is reactionary and will never vote to upturn all the things we need to upturn unless they have a gun held to their head.

[12:11:02.0000] <Hixie>
ok, i'm done reading my e-mail, frustration will now dissipate

[12:11:03.0000] <Hixie>
foolip: you around?

[12:11:04.0000] <Ms2ger>
(which I wouldn't recommend)

[12:12:00.0000] <AryehGregor>
Which wouldn't you recommend, holding a gun to their head?

[12:12:01.0000] <Ms2ger>
Yeah

[12:12:02.0000] <AryehGregor>
Even then they might not be willing to accept all the changes we need, because many of them run directly against their interests as organizations.  E.g., letting implementers run everything.

[12:12:03.0000] <Hixie>
what would you recommend? (personally i've stopped caring enough to even consider finding something to convince them to change, so there's no risk of my doing what AryehGregor suggests)

[12:13:00.0000] <AryehGregor>
As I've said, I don't think they'll ever change, I think the only way out is to abandon ship.  Strikingly similar to TBL's account of how he split from the IETF.

[12:13:01.0000] <AryehGregor>
That was in large part because they spent ages arguing about stupid procedural issues.

[12:13:02.0000] <Ms2ger>
You can't publish specs from a prison, so I'd rather you didn't end up there

[12:14:00.0000] <Hixie>
i meant what would Ms2ger recommend :-)

[12:14:01.0000] <Ms2ger>
I'm fine with convincing MS to join the cabal myself

[12:14:02.0000] <Philip`>
I thought people often wrote books in prison, so why not specs?

[12:15:00.0000] <Hixie>
Philip`: no internet access

[12:15:01.0000] <annevk>
Wait, WebApps is going to patch Ian's drafts and move them ahead independently of the WHATWG drafts?

[12:15:02.0000] <Ms2ger>
Philip`, publish, not write

[12:15:03.0000] <annevk>
That is going to be interesting

[12:15:04.0000] <AryehGregor>
Hixie, who says?  I bet prisoners in low-security prisons get Internet access for good behavior.

[12:15:05.0000] <Hixie>
annevk: "interesting" isn't the word i would pick :-P

[12:15:06.0000] <annevk>
Been on vacation btw?

[12:15:07.0000] <AryehGregor>
Actually, I bet they'd love it if you could work productively from prison and make a decent salary.  You'd think so, right?  Must be better than stamping out license plates or whatever.

[12:15:08.0000] <Philip`>
http://www.ehow.com/how_6629174_write-book-prison-published.html has a handy step-by-step guide

[12:16:00.0000] <Hixie>
annevk: four day weekend in the US

[12:16:01.0000] <Ms2ger>
AryehGregor, sorry, that wouldn't get through the prison's process

[12:17:00.0000] <AryehGregor>
Okay, so are implementers going to come out and say they'll ignore the W3C specs if they patch them independently of Hixie?  That's really what needs to happen to force the W3C to act sane, implementers need to take a stand.

[12:17:01.0000] <annevk>
Hixie, aah, nice

[12:17:02.0000] <annevk>
I'm gonna have one of those starting somewhere tomorrow, in Berlin :)

[12:17:03.0000] <Ms2ger>
AryehGregor, browsers will match the W3C spec more closely than the WHATWG one, no?

[12:17:04.0000] <Hixie>
AryehGregor: in this particular case, the implementations are already doing what the patch is intended to do

[12:17:05.0000] <Hixie>
AryehGregor: i just haven't gotten around to implementing that fix because it's of trivial importance

[12:17:06.0000] <AryehGregor>
Ah, I see.

[12:18:00.0000] <AryehGregor>
Well, I wouldn't say it's trivial, the spec doesn't match reality.  It's just not blocking anything important.

[12:18:01.0000] <Hixie>
the whatwg spec will eventually follow suit; in the meantime, god knows what'll happen if i fix actually important bugs

[12:18:02.0000] <Hixie>
AryehGregor: right

[12:18:03.0000] <AryehGregor>
Well, whatever.

[12:19:00.0000] <Ms2ger>
So, how much time would have been saved if you just did that a couple of weeks ago instead of complaining about the W3C?

[12:19:01.0000] <AryehGregor>
I'd like to see them try that in the HTMLWG.

[12:19:02.0000] <AryehGregor>
/me concurs with Ms2ger that it would be simpler if you had just prioritized this stuff from the beginning to shut people up, since it's a legitimate request

[12:19:03.0000] <AryehGregor>
You prioritize chairs' requests in the HTMLWG, why not WebApps too?

[12:19:04.0000] <Hixie>
i'm not going to start prioritising things in the order that people complain about them

[12:19:05.0000] <Hixie>
that way lies madness

[12:19:06.0000] <Ms2ger>
This way, too

[12:20:00.0000] <AryehGregor>
Well, it might be madness, but it would save people a lot of trouble.

[12:21:00.0000] <Hixie>
what would save people trouble is just not worrying about the process

[12:21:01.0000] <AryehGregor>
You only have to prioritize complaints by chairs and W3C administrative people, and only if they sound threatening.

[12:21:02.0000] <Hixie>
this entire issue only matters because Art has suddenly decided it should be in LC

[12:21:03.0000] <AryehGregor>
Well, yes, but given that we aren't going to get that, no point in worrying about it.

[12:41:00.0000] <annevk>
smaug____, you could allow it only to be called from a top level browsing context

[12:42:00.0000] <smaug____>
annevk: rph ?

[12:42:01.0000] <smaug____>
yeah, that could be one, ugly, solution

[12:45:00.0000] <TabAtkins>
If someone posts a bug asserting something is missing, and it's not missing, is that WORKSFORME or INVALID?

[12:46:00.0000] <AryehGregor>
TabAtkins, http://dev.w3.org/html5/decision-policy/decision-policy-v2.html#bug-resolutions

[12:46:01.0000] <AryehGregor>
WORKSFORME in the HTMLWG.

[12:46:02.0000] <AryehGregor>
INVALID is only for total garbage, or if the filer withdraws it.

[12:46:03.0000] <TabAtkins>
Done, thanks.

[12:49:00.0000] <annevk>
smaug____, doesn't seem so ugly to me

[12:50:00.0000] <smaug____>
annevk: at least it is not backwards compatible, if currently there isn't such limitation

[12:52:00.0000] <annevk>
I doubt you'd even want iframe content to be allowed to ask for such permissions

[12:52:01.0000] <annevk>
E.g. when that happens for geolocation you want to present special UI as well

[13:07:00.0000] <dbaron>
Hixie, fyi, it's a three day weekend in the US, and a 4 day weekend in Google-land

[14:05:00.0000] <AryehGregor>
Hixie, should I resolve WF3 stuff LATER, move it to HTML.next, what?

[14:05:01.0000] <AryehGregor>
Stuff in the "we want this, but not until there's better WF2 support" category.

[14:06:00.0000] <Ms2ger>
/me thinks that's what target milestones should be for

[14:08:00.0000] <AryehGregor>
The latest target milestone is "Rec".  Also, I don't know who's supposed to set those.

[14:08:01.0000] <Ms2ger>
Yeah, they're broken, IMO

[14:09:00.0000] <timeless>
AryehGregor: generally i find RESOLVED LATER = disaster

[14:09:01.0000] <timeless>
it's better to just use a TM to get things off your plate

[14:09:02.0000] <AryehGregor>
TM?

[14:10:00.0000] <Ms2ger>
Target milestone

[14:10:01.0000] <AryehGregor>
Oh.  Well, it all depends on how Hixie's workflow operates here.

[14:10:02.0000] <AryehGregor>
But anyway, we don't have a suitable milestone.

[14:10:03.0000] <AryehGregor>
We'd have to move it to HTML.next.

[14:11:00.0000] <Ms2ger>
That's an even bigger disaster, but oh well :)

[14:12:00.0000] <jamesr>
AryehGregor: can you just leave the bugs open?

[14:12:01.0000] <AryehGregor>
jamesr, that's what I'm asking Hixie.  I'll do whatever he wants.

[14:12:02.0000] <Ms2ger>
That is what you're being paid for after all, no? :)

[14:12:03.0000] <AryehGregor>
I'm pretty sure that HTMLWG procedure requires they be resolved at some point, though, or the component is changed.

[14:12:04.0000] <AryehGregor>
No, I'm not being paid for this, alas.

[14:13:00.0000] <jamesr>
AryehGregor: leaving it open for now != leaving it open forever

[14:14:00.0000] <AryehGregor>
The thing here is that we don't know when exactly we'll want to do it.  It depends when browsers have good WF2 support.

[14:14:01.0000] <Ms2ger>
Never

[14:15:00.0000] <Ms2ger>
:)

[14:25:00.0000] <Hixie>
AryehGregor: either mark them WF3 but otherwise leave them, or mark them WF3 and move them to LATER if you think there's no chance browser vendors would want to work on them soon

[14:25:01.0000] <AryehGregor>
Hixie, okay.

[14:25:02.0000] <Hixie>
if they're _really_ pie in the sky and things that we shouldn't even consider soon, I guess move them to HTML.next

[14:26:00.0000] <Hixie>
but let's not move things into that bucket if there's a chance we might have to solve them sooner rather than later

[14:32:00.0000] <AryehGregor>
Okay.

[14:32:01.0000] <AryehGregor>
I'll leave it open.

[14:32:02.0000] <zewt>
"If the applications you are completing require electronic signature you may be asked to download or update Java version 1.4.2 or higher after you sign in." fffff

[14:34:00.0000] <moo-_->
"may be"

[14:34:01.0000] <moo-_->
zewt: you still have your chances!

[14:34:02.0000] <moo-_->
zewt: run for it

[14:34:03.0000] <zewt>
there will be no survivors

[14:35:00.0000] <moo-_->
zewt: I so love Crockford's Javascript presentation where he dissed Java

[14:37:00.0000] <zewt>
at least if it required Flash i already have that installed ... but almost nobody uses Java on webpages anymore so I don't

[14:38:00.0000] <Hixie>
hsivonen: yt?

[14:44:00.0000] <TabAtkins>
zewt: It's a very bad idea to turn on Java in your browser anyway, due to the attack vectors it opens.

[14:44:01.0000] <TabAtkins>
(I only have Java on my comp at all for Minecraft.)

[14:44:02.0000] <zewt>
that's what VMware is for

[14:44:03.0000] <moo-_->
TabAtkins: I really don't care about attack vectors as Linux user

[14:44:04.0000] <moo-_->
TabAtkins: but I care when my browser freezes for 60 seconds

[14:51:00.0000] <Hixie>
moo-_-: you don't care if your machine gets owned? :-)

[14:52:00.0000] <TabAtkins>
Java attack vectors are not windows-specific.  What do you once you exploit the vector is platform-specific, but the actual holes opened up by Java are often cross-platform.

[14:52:01.0000] <TabAtkins>
s/are not/are often not/

[14:52:02.0000] <jamesr>
write once, owned everywhere

[14:52:03.0000] <Hixie>
write once, debug everywhere

[14:53:00.0000] <jamesr>
actually a number of java sploits are windows-specific, but that's a factor of where the targets are, not where the exploits are

[14:59:00.0000] <Hixie>
anyone know if schema.org implementations support itemref=""?

[15:01:00.0000] <annevk>
AryehGregor, when mutations are sorted out that'll be sorted out

[15:02:00.0000] <Hixie>
looks like google's richsnippet testing tool does, at least

[15:02:01.0000] <AryehGregor>
annevk, which part?

[15:02:02.0000] <Hixie>
though it exposes it in the data model for some reason

[15:02:03.0000] <annevk>
AryehGregor, but basically we don't have to define the methods in terms of each other to add the mutation stuff

[15:02:04.0000] <AryehGregor>
No, but it makes things simpler.

[15:02:05.0000] <annevk>
Hixie, from what I read on schema.org they have proprietary extensions to e.g. the <time> element

[15:02:06.0000] <AryehGregor>
That way there's only one point where we need to define these things.

[15:03:00.0000] <annevk>
AryehGregor, if they all use a single callback you have that too

[15:03:01.0000] <Hixie>
annevk: i think those were mistakes, dunno if they fixed them or not

[15:03:02.0000] <annevk>
Hixie, and they use some kind of extension model that is incompatible with the API

[15:03:03.0000] <AryehGregor>
annevk, that'd be good enough for me.

[15:03:04.0000] <Hixie>
annevk: that, there was discussion of on whatwg

[15:03:05.0000] <annevk>
yeah

[15:03:06.0000] <Hixie>
/me is still thinking of whether to just replace <time> with something else for "computer representation"

[15:03:07.0000] <AryehGregor>
But it would still be more complicated AFAICT.

[15:03:08.0000] <Hixie>
btw speaking of the mutation stuff, what's the status on that? are we close to a solution everyone likes?

[15:04:00.0000] <annevk>
Hixie, no consensus whatsoever it seems

[15:04:01.0000] <Hixie>
k

[15:04:02.0000] <TabAtkins>
Hixie: I end up wanting to use <time> with a unix timestamp anyway, because it's easier to work with that than an rfc timestamp.

[15:04:03.0000] <annevk>
Hixie, there's like very simple mutation listeners versus complex ones that keep some kind of diff-state around

[15:05:00.0000] <Hixie>
TabAtkins: yeah

[15:05:01.0000] <Hixie>
annevk: i've no opinion on the matter, i'm just waiting for someone to make a decision :-)

[15:05:02.0000] <annevk>
AryehGregor, I don't really like defining one method in terms of another

[15:05:03.0000] <annevk>
Hixie, me too

[15:05:04.0000] <annevk>
Hixie, my plan is to put them in DOM Core, whatever "them" is

[15:05:05.0000] <Hixie>
cool

[15:06:00.0000] <AryehGregor>
annevk, then define some abstract steps like "alter the data" that accept the same arguments as replaceData().

[15:06:01.0000] <annevk>
AryehGregor, yeah, could do that I suppose

[15:06:02.0000] <TabAtkins>
Hixie: I do suspect it's useful to have a generic "machine representation".  I'd be afraid of it being misused to embed arbitrary data a la RDFa's @content attribute, but shrug.

[15:07:00.0000] <annevk>
AryehGregor, I haven't looked at that section in a long time, other than minor cleanup

[15:07:01.0000] <TabAtkins>
That's probably not a bad enough fear to really care.

[15:07:02.0000] <Hixie>
TabAtkins: well we already have <meta> for that

[15:07:03.0000] <TabAtkins>
Hixie: For the former or the latter?

[15:07:04.0000] <Hixie>
TabAtkins: but people want to do <span itemprop=a content=b>x</span> instead of <meta itemprop=a content=b>c for some reason

[15:08:00.0000] <Hixie>
(wish i understood why)

[15:08:01.0000] <TabAtkins>
Yeah, I dunno.

[15:08:02.0000] <Hixie>
(i meant for the latter)

[15:09:00.0000] <TabAtkins>
Hixie: However, I still kinda like <time> for its application to the atom extraction algo.  ^_^

[15:11:00.0000] <TabAtkins>
Hixie: Also, would a generic machine-representation element have some way of indicating the datatype?  <time> was useful in that it indicated the data was definitely a datetime, not an arbitrary string.

[15:11:01.0000] <Hixie>
dunno, depends on what hte use cases are

[15:12:00.0000] <Hixie>
for things like microformats or microdata, you don't need the type, since the prop name gives you the type

[15:12:01.0000] <annevk>
only if you know the propname, no?

[15:12:02.0000] <TabAtkins>
I was thinking probably the same thing.

[15:12:03.0000] <TabAtkins>
annevk: If you don't know the propname, what are you doing with the data?

[15:13:00.0000] <TabAtkins>
(Besides just archiving it, where you don't need the datatype.)

[15:13:01.0000] <annevk>
hmm

[15:13:02.0000] <annevk>
not having generic processing supposedly was a problem with microformats

[15:13:03.0000] <TabAtkins>
That's generic *parsing*.

[15:14:00.0000] <Hixie>
what would you do with the type?

[15:14:01.0000] <TabAtkins>
Which was a problem.

[15:14:02.0000] <annevk>
Hixie, store things in different ways

[15:14:03.0000] <TabAtkins>
(In that you couldn't even *extract* microformats without knowing the format.)

[15:14:04.0000] <Hixie>
annevk: well so long as there's a 1:1 mapping from the value to the storage mechanism, you can still do that

[15:14:05.0000] <TabAtkins>
annevk: Possibly good point.  datetimes can be stored much more efficiently than a string timestamp.

[15:14:06.0000] <annevk>
you could do some queries solely based on type, not sure how much meaning you can extract from it

[15:15:00.0000] <Hixie>
annevk: anyway, if there's a concrete use case, i'm open to it

[15:15:01.0000] <Hixie>
i don't know of one

[15:15:02.0000] <annevk>
export to Atom is/was one

[15:16:00.0000] <Hixie>
yeah, pubdate="" would be sacrificed if we did this

[15:16:01.0000] <Hixie>
not sure that's a huge problem

[15:16:02.0000] <Hixie>
we can always redefine the mapping to atom using microdata again

[15:16:03.0000] <Hixie>
btw, at the risk of errecting a bike shed and stepping in a can of paint, the main reason i haven't done this change is that i've no idea what the replacement element would be called

[15:16:04.0000] <Hixie>
<data>?

[15:16:05.0000] <TabAtkins>
Given that I'm already annotating my blogposts with microformats...

[15:16:06.0000] <TabAtkins>
<machine>

[15:16:07.0000] <TabAtkins>
<itemdata>

[15:17:00.0000] <TabAtkins>
<meta>

[15:17:01.0000] <Hixie>
it's not microdata specific, so i'd rather avoid <item*>

[15:17:02.0000] <TabAtkins>
It's not?

[15:17:03.0000] <Hixie>
<meta> is void

[15:17:04.0000] <Hixie>
nah, it could at least be used with microformats as well

[15:17:05.0000] <TabAtkins>
Ah, right.

[15:17:06.0000] <TabAtkins>
Microformats is stuck with html4 syntax anyway.

[15:17:07.0000] <Hixie>
they don't have to be

[15:17:08.0000] <annevk>
I'm not sure what the replacement element would look like

[15:18:00.0000] <TabAtkins>
If they break away, might as well just use microdata syntax entirely.

[15:18:01.0000] <annevk>
If it is solely defined by its attributes maybe just use <span> and <div>

[15:19:00.0000] <annevk>
gives you a reason to like them again ;)

[15:19:01.0000] <TabAtkins>
Having an inline-default and block-default version would be nice.

[15:19:02.0000] <Hixie>
i'd rather not overload them with semantic meaning, it's messy

[15:20:00.0000] <Hixie>
also it complicates the microdata algorithm if now <span> has three ways of having a value

[15:20:01.0000] <Hixie>
bad enough having two

[15:20:02.0000] <Hixie>
(itemscope and textContents)

[15:21:00.0000] <annevk>
Just an idea. Again, I've no idea what this thing would look like

[15:23:00.0000] <Hixie>
it would replace <p><time itemprop=start datetime="2005-01-05">Yesterday</time> it was <meta itemprop="weight" value="5">three weight-units</p> with:

[15:24:00.0000] <Hixie>
<p><data itemprop=start value="2005-01-05">Yesterday</data> it was <data itemprop="weight" value="5">three weight-units</data></p>

[15:24:01.0000] <Hixie>
(and would drop pubdate="")

[15:25:00.0000] <TabAtkins>
I like the consistency somewhat.

[15:26:00.0000] <TabAtkins>
But still think it's basically just for Microdata, and so should have "item" in its name.

[15:27:00.0000] <Philip`>
<annotate itemprop=start ...>

[15:29:00.0000] <Hixie>
TabAtkins: it doesn't have to just be for microdata. e.g. someone could use it to include machine-readable data in the page for their own purposes, much like data-*

[15:29:01.0000] <Hixie>
TabAtkins: in fact this is very similar to data-*

[15:29:02.0000] <Hixie>
*ponders*

[15:30:00.0000] <Hixie>
i guess people could just use data-* if they want their own thing

[15:30:01.0000] <TabAtkins>
Hixie: Indeed.

[15:30:02.0000] <Hixie>
maybe microdata is the only compelling use case for this

[15:30:03.0000] <Hixie>
hmm

[15:31:00.0000] <TabAtkins>
I guess <itemdata> (or whatever) shouldn't be able to accept @itemscope, though, right?

[15:31:01.0000] <TabAtkins>
Since it, by definition, can't represent further microdata items?

[15:32:00.0000] <Hixie>
interesting point

[15:32:01.0000] <Hixie>
this leads me to the thing i wanted to speak to hsivonen about, which came up in the discussion of the command api

[15:32:02.0000] <Hixie>
namely, the global attributes: should we make them less global?

[15:32:03.0000] <Hixie>
it doesn't make much sense to allow tabindex="" or accesskey="" or contenteditable="" on, say, <track>

[15:33:00.0000] <TabAtkins>
Yeah, true.

[15:42:00.0000] <AryehGregor>
On the other hand, <style contenteditable>head, style { display: block }</style> is awesome.

[15:43:00.0000] <jamesr>
heycam: (or anyone else) what's the difference in w3 issue tracker land between RAISED and OPEN?

[15:43:01.0000] <heycam>
jamesr, it's whatever you want it to be :)

[15:43:02.0000] <jamesr>
is OPEN sort of like an accepted state (we think we need to do something but haven't yet)?

[15:43:03.0000] <heycam>
that's how I tend to use it, yeah

[15:44:00.0000] <Hixie>
AryehGregor: yeah, there are some things for which it makes sense even though it's dubious.

[15:44:01.0000] <heycam>
it's probably not that important though; the other specs in the WG haven't been using tracker for their issues

[15:44:02.0000] <AryehGregor>
contenteditable probably should be allowed on any non-void element.

[15:45:00.0000] <AryehGregor>
The way I'm defining it, editing hosts themselves aren't modifiable, only their descendants are, so on a void element it makes no sense.

[15:45:01.0000] <TabAtkins>
AryehGregor: <video contenteditable>?

[15:45:02.0000] <AryehGregor>
Well, that would be a totally different feature.

[15:45:03.0000] <AryehGregor>
Oh, I see.

[15:45:04.0000] <AryehGregor>
You mean that should be invalid too.

[15:45:05.0000] <TabAtkins>
Possibly, yeah.

[15:46:00.0000] <AryehGregor>
Yeah, things whose children have no visible effect should also logically not be allowed to be contenteditable.

[15:47:00.0000] <nlogax>
when using drag and drop, how do you prevent a drop at drop time? if you decide that was not a good place (e.g when sorting stuff by dragging, and you didn't drag it far enough)

[15:47:01.0000] <jamesr>
heycam: aight, i won't worry about it then

[15:47:02.0000] <nlogax>
so that it glides back to whence it came, like when you drop it somewhere non-droppable

[15:47:03.0000] <jamesr>
it seems if i close an issue via the web ui it doesn't generate an email

[15:47:04.0000] <TabAtkins>
nlogax: On dragenter, don't preventDefault?

[15:47:05.0000] <jamesr>
hmm. i suppose that's ok

[15:47:06.0000] <nlogax>
TabAtkins: hmm. but i don't know where it will be dropped when that fires

[15:49:00.0000] <jamesr>
heycam: i also think that email about psychology experiments + knowing the vblank time is important

[15:49:01.0000] <jamesr>
heycam: but before addressing that i really want to propose a time interval API for everything that doesn't suck

[15:49:02.0000] <heycam>
jamesr, I haven't read that mail yet

[15:50:00.0000] <jamesr>
it's good stuff

[15:51:00.0000] <jamesr>
but i want to talk about time intervals on the whatwg list

[15:51:01.0000] <jamesr>
to make sure it gets a broad audience since it's important for the whole platform

[15:51:02.0000] <heycam>
what do you mean by time intervals here?

[15:51:03.0000] <heycam>
jamesr, I saw this recently, also related: http://blog.fishsoup.net/2011/06/30/frame-timing-the-simple-way/

[15:51:04.0000] <jamesr>
you can't tell how much time has elapsed between two callbacks using, f'instance, Date.now()

[15:51:05.0000] <zewt>
timedelta

[15:52:00.0000] <heycam>
jamesr, ah right, yeah we should get on that

[15:52:01.0000] <zewt>
there was some discussion about that recently--don't remember if it was only in here or on the list too

[15:53:00.0000] <jamesr>
one of the timing APIs specs a better clock internally, but doesn't expose it. webaudio has a better clock too, but it's specific to webaudio

[15:53:01.0000] <jamesr>
mozilla had (has?) the animationStartTime global, but i'm not sure how that responds to system clock changes

[15:53:02.0000] <heycam>
jamesr, I am not sure either

[15:53:03.0000] <zewt>
audio can sometimes want to use a clock owned by the audio hardware or drivers, so having that separate might make sense anyway

[15:54:00.0000] <zewt>
afk

[15:54:01.0000] <nlogax>
whatever happens when pressing escape is what i would like to do, but with JS :)

[15:54:02.0000] <nlogax>
but can't find it

[16:15:00.0000] <Hixie>
foolip: http://www.whatwg.org/specs/web-apps/current-work/complete.html#associating-names-with-items look good?

[16:15:01.0000] <Hixie>
foolip: (i am about to start updating the other algorithms to break loops, not done that yet)

[16:23:00.0000] <Hixie>
what's the term meaning the set that contains all items included in the graph if you start at one node and just walk the graph

[16:23:01.0000] <Hixie>
the something something

[16:24:00.0000] <AryehGregor>
The largest connected component containing that node?

[16:24:01.0000] <Hixie>
there's a shorter term for it

[16:24:02.0000] <AryehGregor>
/me doesn't know any graph theory, can't help more than that

[16:25:00.0000] <Hixie>
man i am completely blanking here

[16:25:01.0000] <heycam>
are you thinking of the term "transitive closure"?

[16:25:02.0000] <Hixie>
yes!

[16:25:03.0000] <Hixie>
thank you.

[16:26:00.0000] <heycam>
I think "transitive closure" by itself doesn't necessarily mean just walk the graph; I think it only means something if you give it some relation

[16:26:01.0000] <heycam>
so I'm not exactly sure how to use it correctly

[16:26:02.0000] <Hixie>
yeah my definition wasn't great

[16:26:03.0000] <Hixie>
but it got me the term i was looking for, so good enough!

[16:26:04.0000] <jamesr>
do you want the connected component containing the node?

[16:27:00.0000] <jamesr>
assuming this is an undirected graph

[16:27:01.0000] <Hixie>
the term i was looking for is transitive closure, i was just having a mind blank

[16:30:00.0000] <TabAtkins>
Hixie: "transitive closure" doesn't make sense by itself, for the reason heycam gave - it needs a relation.

[16:31:00.0000] <Hixie>
yes, i am aware :-)

[16:31:01.0000] <TabAtkins>
If your relation is "neighbor of", then taking the transitive closure gives you the largest connected component including your starting node, which is what you want.

[16:33:00.0000] <TabAtkins>
(Via the fact that, after taking the transitive closure, all reachable nodes now return true for the new relation you defined.)

[16:36:00.0000] <jamesr>
i think "largest" is redundant, connected component implies it contains all connected nodes

[16:36:01.0000] <TabAtkins>
No it doesn't.

[16:37:00.0000] <TabAtkins>
The largest connected component probably includes a lot of smaller connected components.

[16:37:01.0000] <Hixie>
tab speaks the truth

[16:37:02.0000] <TabAtkins>
(You're allowed to ignore neighbors if you want, after all.)

[16:37:03.0000] <Hixie>
/me defines a conformance criteria in terms of a constraint rather than steps, for once

[16:37:04.0000] <TabAtkins>
FINALLY

[16:37:05.0000] <TabAtkins>
^_^

[16:37:06.0000] <Hixie>
:-D

[16:38:00.0000] <TabAtkins>
/me keeps defining conformance criteria in terms of "use a cycle detector".

[16:38:01.0000] <Hixie>
heh

[16:39:00.0000] <Hixie>
specifically:

[16:39:01.0000] <Hixie>
"All itemref attributes in a Document must be such that the graph formed from representing each item in the Document as a node in the graph and each property of an item whose value is another item as an edge in the graph connecting those two items does not contain any cycles."

[16:39:02.0000] <Hixie>
that's a pretty horrible sentence, mind you

[16:39:03.0000] <Hixie>
let me rephrase it into english

[16:39:04.0000] <jamesr>
wut

[16:39:05.0000] <TabAtkins>
Yay, I was (barely) able to parse it using my native english facilities!

[16:40:00.0000] <TabAtkins>
(Rather than the deductive variant of english I consciously learned later.)

[16:41:00.0000] <Hixie>
well that's a new error from anolis

[16:42:00.0000] <Hixie>
(some sort of httplib error)

[16:52:00.0000] <Hixie>
if a vcard A has as agent a vcard B and vice versa

[16:52:01.0000] <Hixie>
what should B's AGENT line say?

[16:53:00.0000] <Hixie>
(A's AGENT line is B's entire vCard)

[16:54:00.0000] <Hixie>
AGENT;VALUE=TEXT:ERROR ?

[16:55:00.0000] <annevk>
HTML should start using Ms2ger's Anolis

[16:55:01.0000] <Hixie>
i use whatever jgraham uses

[16:55:02.0000] <Hixie>
pimpmyspec baby

[16:55:03.0000] <Hixie>
what does ms2ger's do that's different?

[16:55:04.0000] <annevk>
xspec xref

[16:55:05.0000] <Hixie>
how?

[16:56:00.0000] <annevk>
using https://bitbucket.org/ms2ger/specification-data

[16:56:01.0000] <annevk>
some kind of repository with references and linkable terms in other specs

[16:57:00.0000] <Hixie>
please tell me that's automatically generated somehow

[16:57:01.0000] <annevk>
and then you use <span data-anolis-ref>HTML</span> for a reference

[16:57:02.0000] <AryehGregor>
It is for all specs except HTML, I think.

[16:57:03.0000] <Hixie>
personally i'd much rather cross-spec xrefs be opt-in rather than automatic... what happens if we have a conflict? or don't we?

[16:57:04.0000] <annevk>
and <code data-anolis-spec=domcore>Node</code>

[16:57:05.0000] <Hixie>
oh

[16:57:06.0000] <AryehGregor>
You have to say which spec you're referring to.

[16:57:07.0000] <annevk>
for terms

[16:58:00.0000] <annevk>
the dictionary maps terms to fragment identifiers

[16:58:01.0000] <annevk>
it's auto-generated most of the time, but for IETF specs that doesn't really work

[16:58:02.0000] <Hixie>
i don't mind listing the terms once, but i'm definitely not going to keep track of what spec things are where all the time

[16:58:03.0000] <AryehGregor>
The makefiles for DOM Core and DOM Range generate the data files for them automatically.

[16:58:04.0000] <Hixie>
can't we make this all automatic?

[16:58:05.0000] <AryehGregor>
I suggested we shouldn't need data-anolis-spec.

[16:58:06.0000] <AryehGregor>
Conflicts should be handled statically somehow when parsing the data files.

[16:59:00.0000] <AryehGregor>
Or just be an error on use, or whatever.

[16:59:01.0000] <annevk>
Yeah, maybe we can get rid of data-anolis-spec

[16:59:02.0000] <annevk>
by having specs use less conflicting terms

[16:59:03.0000] <AryehGregor>
But personally I was too lazy to fix it, so I gave up on typing it out manually and wrote a preprocessor.

[16:59:04.0000] <Hixie>
right now what i do is have a section of the spec that basically "imports" the terms from the other spec

[16:59:05.0000] <Hixie>
this makes it work for print media too

[16:59:06.0000] <AryehGregor>
Even without data-anolis-spec, typing stuff like title=dom-Document-createElement all the time is murderous.

[16:59:07.0000] <AryehGregor>
This works for print media as well.

[16:59:08.0000] <AryehGregor>
It uses generated content.


2011-07-07
[17:00:00.0000] <Hixie>
can we just have a data-foo thingy that we put on a <dfn> that says "this term is actually defined over there"?

[17:00:01.0000] <Hixie>
i guess that's a bit more maintenance

[17:00:02.0000] <Hixie>
i dunno

[17:00:03.0000] <Hixie>
anyway

[17:00:04.0000] <AryehGregor>
Something like @media print { [data-anolis-spec]::after { content: attr(data-anolis-spec); text-transform: uppercase } } or whatever.

[17:00:05.0000] <Hixie>
i'm happy to use whatever if it makes things better

[17:00:06.0000] <AryehGregor>
It's not tenable to do that given the number of terms we have to reference.

[17:00:07.0000] <AryehGregor>
DOM Range uses zillions of terms from DOM Core and HTML.

[17:00:08.0000] <Hixie>
we really should reduce the number of specs, man

[17:01:00.0000] <Hixie>
DOM Range should just be in DOM Core

[17:01:01.0000] <AryehGregor>
HTML isn't so affected because it mostly just defines stuff itself.

[17:01:02.0000] <annevk>
not if DOM Range depends on HTML

[17:01:03.0000] <AryehGregor>
I don't think we need to merge specs if we have good xspec xrefs.

[17:01:04.0000] <Hixie>
merging specs is for sanity of not having a bazillion specs, not for the xrefs

[17:01:05.0000] <Hixie>
we should have like a dozen or so core specs for the web platform

[17:02:00.0000] <AryehGregor>
I think only the Selection part of DOM Range depends on HTML.

[17:02:01.0000] <annevk>
yeah, but we first need to figure out the Web platform

[17:02:02.0000] <annevk>
then we can organize :)

[17:03:00.0000] <annevk>
I mean, we don't even agree on mutation listeners

[17:03:01.0000] <annevk>
:p

[17:04:00.0000] <annevk>
oh, IE10 is removing conditional comments

[17:04:01.0000] <annevk>
about time

[17:04:02.0000] <AryehGregor>
Yay.

[17:04:03.0000] <TabAtkins>
Hixie: Where'd you get the VALUE=TEXT:ERROR thing from?

[17:04:04.0000] <TabAtkins>
/me just read the vcard spec and, as expected, this situation is not addressed.

[17:04:05.0000] <annevk>
ms2ger, ^^ all of the above :)

[17:05:00.0000] <Hixie>
core (dom core, dom events, dom range), i/o (xhr, cors, fetch, from-origin, sniffing, file api), user interaction (mouse events, geo, touch, wheel, device orientation), semantics/api (html, web workers, clipboard, aria), css, svg, mathml, js, unicode, http, woff, webgl, indexeddb...

[17:05:01.0000] <Hixie>
TabAtkins: just made it up

[17:06:00.0000] <Hixie>
TabAtkins: yeah vCard is a spec from the old times

[17:06:01.0000] <TabAtkins>
Hixie: Ah, kk.

[17:06:02.0000] <jamesr>
Hixie: audio

[17:06:03.0000] <Hixie>
jamesr: part of html

[17:06:04.0000] <TabAtkins>
Hixie: I could tell from the fact that it's a Word doc.

[17:06:05.0000] <jamesr>
webaudio isn't (and imo shouldn't be, it's a leaf node)

[17:06:06.0000] <Hixie>
jamesr: along with webrtc, video, etc

[17:06:07.0000] <jamesr>
the <audio> tag is html and insufficient

[17:06:08.0000] <Hixie>
jamesr: i'm saying webaudio should be in the same spec as web workers

[17:07:00.0000] <Hixie>
jamesr: "semantics/api"

[17:07:01.0000] <roc>
webaudio should not be in a silo of its own

[17:08:00.0000] <Hixie>
(note that what i'm saying here doesn't map to who should be working on what)

[17:08:01.0000] <roc>
I guess jamesr hasn't seen my rant about this, because the Audio WG is off in a silo of its own!

[17:08:02.0000] <Hixie>
heh

[17:08:03.0000] <jamesr>
yeah, why is that?

[17:08:04.0000] <Hixie>
why is what?

[17:08:05.0000] <roc>
because that's the way the W3C works

[17:08:06.0000] <jamesr>
hm

[17:09:00.0000] <jamesr>
do we have another process set up other than 'email everything to the whatwg list'?

[17:09:01.0000] <TabAtkins>
Argh, I forget how to get the coordinates of a click relative to the clicked element.  ;_;

[17:09:02.0000] <jamesr>
the web perf WG is another problem, but i don't intend to do anything important enough there for it to matter

[17:10:00.0000] <Hixie>
jamesr: i'm talking about how we publish specs, not how we write them

[17:10:01.0000] <annevk>
TabAtkins, offsetX/Y

[17:11:00.0000] <roc>
unfortunately how we write specs strongly influences how they get published, and how they are designed too

[17:12:00.0000] <Hixie>
yeah

[17:12:01.0000] <annevk>
not so much guidelines for writing specs :(

[17:12:02.0000] <Hixie>
we're in much need of overhaul of much more important stuff before we starting worrying about how we split the specs

[17:12:03.0000] <Hixie>
i'm just blue-skying here

[17:13:00.0000] <annevk>
also need some kind of manual to tell people how to write specs

[17:13:01.0000] <roc>
for example, I think one reason the Audio WG doesn't like my proposal for general media stream processing, is that they're the Audio WG and they're not chartered for or interested in general media stream processing

[17:13:02.0000] <jamesr>
annevk: does anything anywhere define what the DOMTimestamp on events represent, precisely?

[17:13:03.0000] <TabAtkins>
Yay for Conway's Law!

[17:13:04.0000] <Hixie>
annevk: we have a wiki page on whatwg.org, i've helped a bit with it. i'd love to help more but don't want to do it all myself. :-)

[17:14:00.0000] <annevk>
jamesr, http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-event-timestamp

[17:14:01.0000] <annevk>
Hixie, good point, maybe once I'm back from Berlin I'll remember that and fix it up a bit

[17:14:02.0000] <jamesr>
ok, so same as ECMAScript Date.now()

[17:14:03.0000] <annevk>
jamesr, the definition of DOMTimeStamp is in Web IDL btw

[17:14:04.0000] <Hixie>
roc: implementations win

[17:14:05.0000] <jamesr>
yeah, i'm interested in the value not the representation

[17:14:06.0000] <jamesr>
thanks

[17:15:00.0000] <Hixie>
roc: are there implementations of either of the proposals?

[17:15:01.0000] <roc>
jamesr: the relevant thread is here, if you care: http://lists.w3.org/Archives/Public/public-audio/2011AprJun/0102.html

[17:16:00.0000] <roc>
Hixie: yes, Chris Rogers has his in Chrome and I'm working on one

[17:16:01.0000] <jamesr>
oh doug

[17:16:02.0000] <roc>
doug was mostly an innocent bystander in this case

[17:16:03.0000] <jamesr>
with a bad suggestion

[17:17:00.0000] <roc>
Hixie: yes, implementations win, but that's not always a god thing

[17:17:01.0000] <roc>
er, a good thing

[17:17:02.0000] <Hixie>
i wasn't making a value judgement

[17:18:00.0000] <Hixie>
merely observing reality :-)

[17:18:01.0000] <roc>
I thought so

[17:18:02.0000] <Hixie>
looks like i said so in http://lists.w3.org/Archives/Public/public-audio/2011AprJun/0118.html too :-)

[17:19:00.0000] <Hixie>
personally i know nothing about audio so i'm really not in a position to make a proposal

[17:19:01.0000] <roc>
for other people, saying the exact same thing would amount to a value judgment

[17:19:02.0000] <jamesr>
as per your last point audio APIs are one of the main drivers of flash in a lot of cases

[17:19:03.0000] <Hixie>
(that was the case with video conferencing too; after nobody did anything for like a year i finally broke down and taught myself the relevant specs and wrote up the PeerConnection proposal. i don't intend to do that with audio.)

[17:20:00.0000] <Hixie>
roc: true

[17:20:01.0000] <jamesr>
angry birds would be 100% web platform if there was decent audio support on the web

[17:20:02.0000] <roc>
actually

[17:20:03.0000] <roc>
it turns out that most of today's Web games don't need new Web API

[17:21:00.0000] <annevk>
Hixie, btw, is PeerConnection mostly for client-server still?

[17:21:01.0000] <annevk>
Hixie, it doesn't really read like it could be client-client as well

[17:21:02.0000] <Hixie>
annevk: PeerConnection is for peer-to-peer

[17:21:03.0000] <Hixie>
hence the name

[17:21:04.0000] <Hixie>
it does NAT traversal

[17:21:05.0000] <annevk>
Hixie, but the initiation talks about servers

[17:21:06.0000] <annevk>
Hixie, I missed that

[17:22:00.0000] <roc>
they just want to play preloaded sounds with low latency, and would be served just fine with a bunch of <audio preload>s and getElementById(id).clone().play() if that was implemented well

[17:22:01.0000] <Hixie>
annevk: well, yeah, you can't connect two peers without them having some sort of signaling channel to coordinate things

[17:22:02.0000] <roc>
see the comments from the Zynga guys later in the thread

[17:22:03.0000] <Hixie>
annevk: and that means a server to coordinate things

[17:22:04.0000] <roc>
same goes for Angry Birds, AFAIK

[17:22:05.0000] <TabAtkins>
Low-latency audio with <audio> actually isn't *that* hard to hack around.

[17:22:06.0000] <TabAtkins>
http://xanthir.com/demos/collidingcircles demonstrates it, frex.

[17:23:00.0000] <annevk>
Hixie, I see, because the previous idea was that they would connect to each other based on some keys

[17:23:01.0000] <roc>
There are ... implementation issues: http://www.phoboslab.org/log/2011/03/the-state-of-html5-audio

[17:23:02.0000] <annevk>
Hixie, and the way the keys were exchanged was up to the site

[17:23:03.0000] <jamesr>
TabAtkins: is that just keeping a bag of <audio>s around?

[17:23:04.0000] <TabAtkins>
roc: Yes, thus the "hack around" part.

[17:23:05.0000] <Hixie>
annevk: it's basically still like that

[17:23:06.0000] <Hixie>
annevk: you get a callback that gives you the "keys" (actually an SDP payload)

[17:23:07.0000] <TabAtkins>
jamesr: Yeah, I generate 10 <audio>s per sound, and cycle through them everytime I play that particular sound.

[17:23:08.0000] <Hixie>
annevk: and you get it to the other side somehow

[17:24:00.0000] <jamesr>
TabAtkins: doesn't really scale when you are trying to implement quake and have hundreds of sounds that you might want to play and you might need to overlap them dozens of times on top of each other

[17:24:01.0000] <Hixie>
annevk: and then you pass the "keys" to the other browser

[17:24:02.0000] <Hixie>
annevk: and so on back and forth

[17:24:03.0000] <jamesr>
and you need to distance-fade them and adjust some samples to come in one ear more strongly than the other

[17:24:04.0000] <Hixie>
annevk: with all the video and audio and udp data going straight peer-to-peer

[17:24:05.0000] <annevk>
I guess I should read it more closely, thanks

[17:24:06.0000] <jamesr>
or pan the sample from one ear to the other as it plays (and the rocket wooshes by)

[17:24:07.0000] <Hixie>
annevk: i haven't written a good intro

[17:24:08.0000] <Hixie>
annevk: it's on my list

[17:24:09.0000] <jamesr>
which you definitely don't have CPU time to do in javascript

[17:25:00.0000] <TabAtkins>
jamesr: An actual games company came up with another technique where you concatenate all your sounds and seek.  That way you only need to have as many audios as you plan to have *total* sounds playing at once.

[17:25:01.0000] <TabAtkins>
That requires a bit more expertise, though.

[17:25:02.0000] <jamesr>
and a really really big continuous audio sample

[17:25:03.0000] <jamesr>
or rather a few dozen copies of it

[17:25:04.0000] <roc>
what was my last message?

[17:25:05.0000] <TabAtkins>
Yes.

[17:25:06.0000] <jamesr>
roc: There are ... implementation issues: http://www.phoboslab.org/log/2011/03/the-state-of-html5-audio

[17:25:07.0000] <zewt>
jamesr: can't really do it in javascript anyway if you want to support things like hardware surround

[17:26:00.0000] <jamesr>
true

[17:26:01.0000] <zewt>
environmental audio, etc

[17:26:02.0000] <roc>
3D spatialization definitely needs new API

[17:26:03.0000] <roc>
Angry Birds and Zynga games do not

[17:26:04.0000] <zewt>
roc: well, it's debatable whether 3d audio vs. 2d audio should be separate APIs or not

[17:27:00.0000] <roc>
they should be tightly integrated

[17:27:01.0000] <jamesr>
they could use 2d spatialization, it'd be basically free if the API was designed well

[17:27:02.0000] <roc>
but we already have an API for playing sound samples

[17:27:03.0000] <roc>
<audio>

[17:27:04.0000] <zewt>
angry birds could certainly make use of stereo (whether it does or not I don't know)

[17:27:05.0000] <TabAtkins>
zewt: They either use or fake stereo on Android, at least - as you scroll away from pigs, their grunts get quieter.

[17:27:06.0000] <Philip`>
Just wrap OpenAL and call it WebAL

[17:28:00.0000] <jamesr>
Philip`: i think khronos proposed that :/

[17:28:01.0000] <Philip`>
/me 's experience with OpenAL has not been universally positive

[17:28:02.0000] <zewt>
i know environmental audio APIs can be pretty stupidly complex; trying to represent their whole vocabulary in a generic API is probably hopeless

[17:28:03.0000] <roc>
the Zynga guys didn't ask for spatialization

[17:28:04.0000] <roc>
they asked for basic sound playback to work

[17:28:05.0000] <jamesr>
the quake guys did

[17:28:06.0000] <roc>
and it doesn't

[17:29:00.0000] <jamesr>
yes, there are QoI issues with existing <audio>, but there are also API needs

[17:29:01.0000] <Philip`>
(e.g. I work on some game that uses OpenAL, and it randomly hits assertion failures on OS X and nobody has yet figured out why)

[17:29:02.0000] <roc>
there's both, but the QoI issues get less love, because hey, shiny new API!

[17:29:03.0000] <zewt>
roc: heh

[17:30:00.0000] <Philip`>
(and also it breaks with Creative's OpenAL drivers on Windows so we replace it with the OpenAL Soft implementation which defeats the whole point of standards)

[17:30:01.0000] <zewt>
roc: but if they're built on the same underlying framework, in principle both win

[17:30:02.0000] <roc>
indeed

[17:30:03.0000] <zewt>
i seem to remember looking at OpenAL source over a decade ago and it having pages of conditionals 10+ levels deep

[17:30:04.0000] <zewt>
tried to find the code I saw later on but it seemed to have been scoured from the internet

[17:30:05.0000] <roc>
but, and here I reveal my biases, Chris Rogers' API and implementation are much not built on the same underlying framework

[17:31:00.0000] <jamesr>
what do you mean? they use the same backend as our <audio> support

[17:32:00.0000] <Philip`>
zewt: If I remember correctly, the official reference driver was terrible, and OpenAL Soft is a reimplementation from scratch

[17:32:01.0000] <jamesr>
the whole stack is the same except for the API bits, and the frontend support for the webaudio features not there in <audio>

[17:32:02.0000] <zewt>
it was easily some of the worst code I'd ever seen

[17:32:03.0000] <Philip`>
s/driver/implementation/

[17:32:04.0000] <roc>
I mean you play samples using AudioBufferNode, not <audio>

[17:32:05.0000] <zewt>
then or since

[17:32:06.0000] <jamesr>
oh, you mean from the API pov? <audio> can be used as source nodes

[17:33:00.0000] <roc>
that's not very well defined

[17:33:01.0000] <roc>
it's not the way you're supposed to do it

[17:33:02.0000] <jamesr>
how so?

[17:33:03.0000] <jamesr>
http://chromium.googlecode.com/svn/trunk/samples/audio/specification/specification.html#MediaElementAudioSourceNode-section

[17:34:00.0000] <jamesr>
use one of those as a source node linked to your <audio> or <video or whatever, hook it up to an output, note()

[17:34:01.0000] <roc>
for example, if the audio element is paused, what happens?

[17:34:02.0000] <Hixie>
things can definitely be well-defined

[17:35:00.0000] <Hixie>
last i looked at the audio api, the whole thing was rather under-defined

[17:35:01.0000] <Hixie>
that's just how things are before they're mature

[17:35:02.0000] <Hixie>
the question is, is it the right direction at all?

[17:35:03.0000] <Hixie>
if there's more than one way to do things, it probably isn't (e.g. is there another way of getting audio from a file into an audio node?)

[17:36:00.0000] <Hixie>
if it requires a lot of code to do simple things, it probably isn't (e.g. do you have to have six lines of boilerplate code to get from <audio> to an audionode?)

[17:37:00.0000] <Hixie>
if there are things that don't work just because parts of the api have different designs, it probably isn't (e.g. can you go from <audio> to an audio node but not from an audio node to a PeerConnection stream?)

[17:37:01.0000] <Hixie>
(i don't know the answers to those questions)

[17:37:02.0000] <Hixie>
and there's also the question "is there another concrete proposal?"

[17:37:03.0000] <TabAtkins>
Hixie: http://chromium.googlecode.com/svn/trunk/samples/audio/specification/specification.html#AudioElementIntegration-section

[17:38:00.0000] <Hixie>
TabAtkins: i know, i've talked about this with chris in front of you :-P

[17:38:01.0000] <TabAtkins>
Ah, I must not have been listening. ^_^

[17:38:02.0000] <Hixie>
:-)

[17:41:00.0000] <jamesr>
strawpoll on another topic: when calling setTimeout(..., 3000), if the system clock is adjusted before the timer fires all browsers except WebKit fire the timer anyway after 3000ms expire. WebKit-based browsers wait until the system clock has advanced to 3000ms past the time the system clock said when setTimeout() was called

[17:41:01.0000] <jamesr>
is WebKit's behavior here justifiable or just crazy land?

[17:41:02.0000] <jamesr>
i tested against ffx 8.01a, opera 11.something on mac and ie9/win7

[17:42:00.0000] <TabAtkins>
jamesr: So you mean that other browsers use a real time, while webkit goes purely off of Date()?

[17:42:01.0000] <zewt>
sounds like it'd break everything using timers, especially if the time is set back a lot

[17:42:02.0000] <zewt>
(which fortunately doesn't happen often, but)

[17:42:03.0000] <jamesr>
zewt: the delta of Date.now() before the callback and when it fires goes negative in other browsers

[17:42:04.0000] <jamesr>
actually it happens more than most people think

[17:42:05.0000] <jamesr>
TabAtkins: yes, for scheduling purposes

[17:43:00.0000] <TabAtkins>
Webkit's behavior is crazytown.

[17:46:00.0000] <jamesr>
k i'll just fix it

[17:46:01.0000] <zewt>
setTimeout should be defined in terms of the time delta API, once it actually exists

[17:46:02.0000] <jamesr>
yeah, i'm working on a proposal for that too

[17:49:00.0000] <jamesr>
annevk-cloud: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-event-timestamp doesn't say anything about leap seconds

[17:49:01.0000] <jamesr>
annevk-cloud: ECMA-262 defines Date as ignoring leap seconds, although i don't think anyone actually does that

[17:50:00.0000] <zewt>
(nitpick: "number of milliseconds" is more natural than "amount of")

[17:50:01.0000] <jamesr>
i think that means the two differ by either 24 or 34 seconds now (i'm not sure what the wikipedia page is trying to tell me)

[18:05:00.0000] <TabAtkins>
Dammit, these emails I write don't seem so long when I'm writing them.

[18:06:00.0000] <jamesr>
heycam: webperf uses www.w3.org/StyleSheets/TR/W3C-ED.css

[18:06:01.0000] <jamesr>
but it's not https so if you load the webperf spec over https you get mixed content errors

[18:06:02.0000] <jamesr>
is that stylesheet available over https?

[18:07:00.0000] <heycam>
jamesr, it seems it is

[18:07:01.0000] <jamesr>
if i just change the url to https:// the connection times out for me

[18:07:02.0000] <jamesr>
what url are you using?

[18:07:03.0000] <heycam>
https://www.w3.org/StyleSheets/TR/W3C-ED.css

[18:07:04.0000] <nessy>
so, hmm, I wonder if I've been doing this correctly: if I want to point out a bug on a part of the spec that's only in the WHATWG version, can I still go through the w3c bug tracker or should I email the whatwg list?

[18:08:00.0000] <heycam>
maybe the url should be "//www.w3.org/StyleSheets/TR/W3C-ED.css" so that it uses http: or https: depends on how you view the spec. (that works, doesn't it?)

[18:08:01.0000] <jamesr>
yeah

[18:08:02.0000] <jamesr>
(well it should)

[18:09:00.0000] <jamesr>
either that or just always use the https:// version

[18:09:01.0000] <jamesr>
i think // is best practice

[18:09:02.0000] <jamesr>
hm url works for me now, not sure why i was having trouble earlier

[18:11:00.0000] <TabAtkins>
nessy: There's no practical difference between filing a bug and sending an email, so just send an email (that way it avoids anyone being snotty and closing the bug).

[18:11:01.0000] <nessy>
TabAtkins: hehe :-) good idea! thanks!

[18:15:00.0000] <TabAtkins>
nessy: (I'm also prejudiced in that I hate the bugtracker and vastly prefer email.)

[18:16:00.0000] <nessy>
TabAtkins: aha! I don't really care actually - but conversations in the bug tracker tend to be short and to the point, but also often harsh

[18:18:00.0000] <heycam>
hmm, where do you actually download the IE10 PP3 from? http://ie.microsoft.com/testdrive/Info/Downloads/Default.html only seems to have PP2

[18:18:01.0000] <TabAtkins>
Has pp3 beenr elease yet?

[18:19:00.0000] <heycam>
http://windowsteamblog.com/ie/b/ie/archive/2010/06/23/internet-explorer-platform-preview-3-released-today.aspx

[18:30:00.0000] <TabAtkins>
heycam: Hm, I dunno.

[18:44:00.0000] <jamesr>
TabAtkins: don't you see? corporations are -have already- more likely

[18:44:01.0000] <TabAtkins>
jamesr: It's... it's all so obvious now.

[18:44:02.0000] <jamesr>
you have no time to rebut make your accessibility

[18:46:00.0000] <TabAtkins>
Take off all SVG.  For great justice.

[19:03:00.0000] <TabAtkins>
40% of the new email in my inbox from the last 10 minutes is a11y.

[19:04:00.0000] <TabAtkins>
Ah, to be fair, one of those was the public-a11y reminder that I'm not registered and the list hates me.

[19:14:00.0000] <Philip`>
Does 10 minutes provide a statistically significant sample?

[22:23:00.0000] <boblet>
damn, writing specs is hard. I’m trying to think of a good alternative *sentence* to propose… and failing <_<

[22:31:00.0000] <zewt>
heh, g+ feedback button sends a screenshot of the page ... i wonder what they're doing to implement that

[23:53:00.0000] <boblet>
Anyone following along with my problems with the blockquote spec, I wrote an article http://oli.jp/2011/blockquote/ found real-world examples that would be non-conforming with current spec http://oli.jp/example/blockquote-metadata/ and emailed WHATWG list

[23:53:01.0000] <boblet>
so do I file a bug report now or is that overkill? :)

[00:44:00.0000] <annevk>
boblet, email should be sufficient

[00:44:01.0000] <annevk>
boblet, unless you want to escalate it within the W3C if you disagree with Ian, in which case you need to file a bug report

[00:48:00.0000] <annevk>
jamesr_, I mentioned your comment on leap seconds in the source

[00:48:01.0000] <boblet>
annevk: thanks. yeah heard email is enough from Hixie too

[00:48:02.0000] <annevk>
jamesr_, I guess we want to be consistent with new Date()

[00:53:00.0000] <MikeSmith>
it's really great to see the latest IE10 platform preview has an HTML5 that intends to conform with the spec

[00:54:00.0000] <MikeSmith>
a lot of nice surprises in that latest release

[00:55:00.0000] <MikeSmith>
is http://gsnedders.html5.org/html5lib-tests/runner.html up to date?

[00:56:00.0000] <MikeSmith>
and/or it would be great to get some data on how much of the html5lib test suite the IE10 parser is passing at this point

[01:00:00.0000] <MikeSmith>
annevk: dunno if you saw http://enable-cors.org/ yet

[01:01:00.0000] <Rik`>
yeah, happy morning without conditional comments \o/

[01:01:01.0000] <annevk>
Yeah some time ago

[01:01:02.0000] <MikeSmith>
annevk: seems mhausenblas made that page

[01:03:00.0000] <MikeSmith>
Rik`: funny that most responses to that IEblog post were about "no more conditional comments" rather than "IE now has a real HTML5 parser"

[01:04:00.0000] <annevk>
We should sign up for https://plus.google.com/105923173045049725307/posts/E3mVj6nskaX

[01:05:00.0000] <annevk>
http://goo.gl/zq95C

[01:06:00.0000] <Rik`>
MikeSmith: the HTML5 parser is only interesting if you don't write valid code so web developers that care about the IE announcement are not writing invalid code hopefully

[01:06:01.0000] <MikeSmith>
I see

[01:07:00.0000] <annevk>
Rik`, conditional comments are invalid

[01:08:00.0000] <Onderhond>
I'll miss them though.

[01:08:01.0000] <nlogax>
man, i really can't figure out how to cancel a drag&drop action at the time it is dropped

[01:08:02.0000] <nlogax>
like what happens when you press escape mid-d&d

[01:15:00.0000] <MikeSmith>
maybe Google Circles should be called Google Cells instead

[01:16:00.0000] <MikeSmith>
in the sense of terrorist cells where there's only guy who knows who's actually in the cell, and none of the members of the cell necessarily know they are even in it

[01:18:00.0000] <Onderhond>
Wouldn't it still be wise to forsee some kind of standardized mechanism proving fall-back code (css) for unsupporting browsers?

[03:01:00.0000] <foolip>
Hixie, I'm here

[03:01:01.0000] <foolip>
Hixie, I was looking at this just yesterday

[03:01:02.0000] <foolip>
you're probably sleeping, will email you

[03:33:00.0000] <krijn>
(Server will be down a bit today, getting a new fiberglass connection)

[04:02:00.0000] <MikeSmith>
http://groups.google.com/group/es-operating-system/browse_thread/thread/5146403c9623eeec?pli=1

[04:03:00.0000] <MikeSmith>
"Today I pushed a brand-new web browser  source code to the ES operating system's code base. It's been built  from the ground up at Esrille Inc., which I founded in Kyoto just a year ago."

[04:04:00.0000] <MikeSmith>
bravo Shiki

[04:04:01.0000] <MikeSmith>
his company has a great logo too

[04:04:02.0000] <MikeSmith>
http://www.esrille.com/

[04:05:00.0000] <MikeSmith>
that's based on his actual dog

[04:05:01.0000] <MikeSmith>
named Jiminy


2011-07-10
[15:28:00.0000] <Dashiva>
/me blinks

[15:31:00.0000] <Dashiva>
A use-case wiki page that doesn't list use cases. I see.


2011-07-11
[01:46:00.0000] <hsivonen>
MikeSmith: the relevant bit is http://www.whatwg.org/specs/web-apps/current-work/#change-the-encoding

[01:46:01.0000] <hsivonen>
MikeSmith: however, it seems that the way the steps are now ordered, the parser shouldn't get as far as step 3

[01:46:02.0000] <hsivonen>
I wonder if the steps were always in that order...

[01:47:00.0000] <MikeSmith>
yeah, may well have changed

[01:51:00.0000] <MikeSmith>
hsivonen: http://www.w3.org/TR/2008/WD-html5-20080122/#changing

[01:51:01.0000] <MikeSmith>
step number 1: If the new encoding is UTF-16, change it to UTF-8.

[01:52:00.0000] <hsivonen>
MikeSmith: good to see that my initial reading comprehension has worked even if I've failed to keep up with spec changes

[01:53:00.0000] <MikeSmith>
yeah:)

[01:53:01.0000] <MikeSmith>
and incidentally, I guess there's at least one thing that TR drafts can occasionally useful for

[01:54:00.0000] <hsivonen>
we have a use case for TR! :-)

[01:54:01.0000] <MikeSmith>
heh

[01:55:00.0000] <MikeSmith>
fwiw, looks like it changed some time between 2009-02-12 and 2009-04-23

[02:03:00.0000] <MikeSmith>
hsivonen: actually, it looks like this was the exact revision:

[02:03:01.0000] <MikeSmith>
http://html5.org/tools/web-apps-tracker?from=3203&to=3204

[04:35:00.0000] <annevk>
hmm, I have no notes from last week for the WHATWG Weekly

[04:37:00.0000] <annevk>
foolip, the Microdata change means you can now express graph data?

[04:44:00.0000] <tjaytje>
hi guys :)

[04:45:00.0000] <annevk>
hey

[04:46:00.0000] <tjaytje>
this is the official whatwg channel, so if i have any questions what so ever about HTML i can ask them here right ?

[04:50:00.0000] <annevk>
yeah

[04:50:01.0000] <annevk>
also about ponies

[05:00:00.0000] <tjaytje>
Whahaha oke let's talks about ponnies :P

[07:05:00.0000] <annevk>
<aside> is for non-critical content

[07:05:01.0000] <tjay_>
so replace the aside with a section element ?

[07:06:00.0000] <annevk>
actually, that content does look like <aside> would fit

[07:07:00.0000] <volkmar>
i think there were a discussion about <input type='email|url'> that needs to be submitted in ASCII, someone has a link to it by any change?

[07:08:00.0000] <annevk>
volkmar, http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-May/031519.html

[07:09:00.0000] <tjay_>
so everything that is directly related to the content can be coded in a <aside> tag ?

[07:09:01.0000] <Onderhond>
Or in this case, related to the site as a whole.

[07:09:02.0000] <annevk>
it's not directly related

[07:09:03.0000] <annevk>
the content is the text, categories of your blog are not related to that

[07:10:00.0000] <annevk>
only tangentially

[07:10:01.0000] <Onderhond>
It shouldn't be nested inside the content wrap, I believe that's the main difference.

[07:10:02.0000] <annevk>
which fits <aside>

[07:10:03.0000] <annevk>
kind of funny that the same journalists that talk about tablets as post-PC also refer to them as tablet PCs

[07:15:00.0000] <annevk>
quiz: is there a difference between XML's Name and XML Names' QName?

[07:16:00.0000] <annevk>
it appears the answer is no

[07:16:01.0000] <annevk>
I win ten thousand dollar and can continue playing for a fridge

[07:16:02.0000] <tjay_>
En we gaan door voor het gouden pols horloge !

[07:16:03.0000] <tjay_>
:P

[07:30:00.0000] <annevk>
btw

[07:30:01.0000] <annevk>
about those ponies

[07:30:02.0000] <annevk>
https://plus.google.com/112284435661490019880/posts/Qv2bUFsSMR3

[07:31:00.0000] <annevk>
totally related according to Google+

[07:37:00.0000] <timeless>
Ms2ger: ping

[07:39:00.0000] <timeless>
hi smaug____

[07:39:01.0000] <smaug____>
timeless: hi

[07:39:02.0000] <timeless>
/me is reading webidl-3.3.4.4. Named properties

[09:47:00.0000] <hsivonen>
hmm. looks like I'm being nominated to a Task Force: http://lists.w3.org/Archives/Public/www-tag/2011Jul/0011.html

[09:48:00.0000] <hsivonen>
manu-db: btw, I am not a Mozilla employee. I'm a contractor.

[09:48:01.0000] <hsivonen>
Looks like I need to catch up with Microdata/RDFa email that got sent while I was on vacation

[09:48:02.0000] <timeless>
manu-db: fwiw: you probably should favor 'iiuc' over

[09:49:00.0000] <timeless>
'iirc' for such cases

[09:49:01.0000] <manu-db>
hsivonen: Thanks for the clarification, but I don't think it matters - you've had good input on RDFa and Microdata in the past, I think you'd be a good person to be in that group.

[09:49:02.0000] <timeless>
(understand v. recall)

[09:49:03.0000] <timeless>
hsivonen: how / what was your vacation? :)

[09:49:04.0000] <timeless>
And have you been g+'d? :)

[09:50:00.0000] <Ms2ger>
annevk++

[09:50:01.0000] <hsivonen>
timeless: yeah, the Google+ thing seems to have happened while I was away. Not sure what to think of it, but this time I didn't resist it like Facebook

[09:51:00.0000] <timeless>
I'm now in 4 social networks

[09:51:01.0000] <timeless>
Orkut, Facebook, g+, g+

[09:51:02.0000] <hsivonen>
timeless: the vacation was nice. in Saxon Switzerland, Königstein, Dresden and Leipzig

[09:51:03.0000] <timeless>
cool

[09:51:04.0000] <timeless>
places i haven't been

[09:51:05.0000] <timeless>
/me needs to find a WG interested in hosting a meeting near one of those :)

[09:52:00.0000] <hsivonen>
timeless: I'm temporarily staying in Brunswick, so it was easy to travel by train

[09:54:00.0000] <timeless>
/me has issues convincing google-maps to explain  where that is

[09:55:00.0000] <timeless>
Braunschweig ?

[09:55:01.0000] <hsivonen>
manu-db: I didn't mean to imply the distinction mattered for the TF.

[09:55:02.0000] <hsivonen>
timeless: yes

[09:56:00.0000] <timeless>
and did you see /msg?

[10:10:00.0000] <boblet>
anyone know of HTML5 outline support in anything (apart from Firefox and Chrome default styles)?

[10:11:00.0000] <boblet>
afaik nothing on the AT side, but wondering if any browser makers are working on exposing it…

[10:18:00.0000] <TabAtkins>
AryehGregor: Regarding http://www.w3.org/Bugs/Public/show_bug.cgi?id=13176 roc stated my opinions well.

[10:18:01.0000] <AryehGregor>
TabAtkins, where?

[10:18:02.0000] <TabAtkins>
In the scrollback

[10:20:00.0000] <AryehGregor>
Hmm, okay.

[10:21:00.0000] <AryehGregor>
hsivonen, annevk, zcorpan-who-is-never-here-when-I-want-to-ask-him-something, do any of you want to resolve <http://www.w3.org/Bugs/Public/show_bug.cgi?id=13176>?  I don't feel I understand the issue well enough to actually resolve it.  If none of you do either, then I'll raise it to P1 so that they can start writing a Change Proposal if they don't get a response.

[10:21:01.0000] <AryehGregor>
In thirty days.

[10:22:00.0000] <AryehGregor>
(since a lot of the point of Editorial Assistants is to start things moving earlier so there aren't multi-month delays added to the already-long escalation process)

[10:34:00.0000] <MikeSmith>
I see Aaron Leventhal is at RIM now too

[10:48:00.0000] <MikeSmith>
timeless: did you move?

[10:48:01.0000] <MikeSmith>
I mean, move the place where you are living

[10:48:02.0000] <timeless>
yes, i'm in Toronto

[10:49:00.0000] <timeless>
eh, Aaron L joined us?

[10:49:01.0000] <MikeSmith>
I think he's probably been there for a while

[10:50:00.0000] <timeless>
oh, so i joined him?

[10:50:01.0000] <MikeSmith>
heh

[10:50:02.0000] <MikeSmith>
yeah I guess so

[10:51:00.0000] <MikeSmith>
please host some WG meeting so I can go and say Hi to you guys

[10:51:01.0000] <MikeSmith>
I never been to Toronto

[10:51:02.0000] <Ms2ger>
MikeSmith, I assume you're welcome in Mozilla's Toronto office as well ;)

[10:52:00.0000] <MikeSmith>
I hope so :)

[10:52:01.0000] <MikeSmith>
but yeah, that's where mozlabs is headquartered, right?

[13:02:00.0000] <AryehGregor>
If you write two billion lines, you will likely get kicked more than twenty times, assuming the network has automatic flood protection.

[13:03:00.0000] <timeless>
it does, see killer and the kicks i got

[13:04:00.0000] <timeless>
2*1e9/(10*365*24*60*60)

[13:04:01.0000] <timeless>
is 6.3

[13:04:02.0000] <timeless>
i can't remember what the threshold is for getting kicked, but i think it means that at .2billion lines, i wouldn't

[13:04:03.0000] <timeless>
assuming even distribution

[13:05:00.0000] <jcranmer>
so 6.3 lines per second?

[13:05:01.0000] <timeless>
anyway, the point is that the algorithm for determining whether you've been kicked a lot really needs to consider activity

[13:05:02.0000] <timeless>
jcranmer: yeah

[13:05:03.0000] <timeless>
assuming i didn't mess up any of the math

[13:06:00.0000] <zewt>
for all the times people get kicked from here?

[13:06:01.0000] <timeless>
zewt: well, the calculation is 2billion lines / 10 years

[13:06:02.0000] <timeless>
pick a network of your choice

[13:06:03.0000] <timeless>
i was comparing it to killer on moznet

[13:07:00.0000] <timeless>
where i think the threshold is something like 6 lines in 10-15s or something

[13:08:00.0000] <zewt>
irc throttles to one line every two seconds anyway, typically

[13:08:01.0000] <zewt>
(with token bucket throttling to 5-6 lines)

[13:08:02.0000] <timeless>
/me likes how 2 billion lines (a random number pulled out of thin air) is only off by one order of magnitude

[13:08:03.0000] <zewt>
(bursting, rather)

[13:26:00.0000] <zewt>
why does everyone @microsoft posting to webapps post in gigantofont

[13:26:01.0000] <TabAtkins>
It's Outlook's fault.

[13:26:02.0000] <zewt>
obnoxious as hell

[13:26:03.0000] <TabAtkins>
Steve Faulkner has the same problems.

[13:27:00.0000] <timeless>
use thunderbird

[13:27:01.0000] <zewt>
also Travis Leithead

[13:27:02.0000] <timeless>
speaking of which, i need to try thunderbird again now that i found a solution to exchange server

[13:28:00.0000] <zewt>
haha great, it looks like the history api firefox bug is affecting gmail now

[13:28:01.0000] <zewt>
judging from the fact that back/forward are now randomly not working

[15:37:00.0000] <TabAtkins>
jamesr: Oh yeah, also, bikesheddy, "stableTime"?

[15:37:01.0000] <jamesr>
yeah, currentTime() was an idea boris had. i don't have particularly strong feelings about name currently

[15:38:00.0000] <timeless>
heycam: so...

[15:38:01.0000] <jamesr>
was planning to nail down what it does before deciding what it's called :)

[15:38:02.0000] <timeless>
i'm probably missing something obvious

[15:38:03.0000] <timeless>
in ExtendedAttributeList

[15:38:04.0000] <jamesr>
Hixie: it could be. i dunno. implementation wise, this will be consistent across the OS on every system i'm aware of

[15:38:05.0000] <timeless>
what provides for the "Constructor" part of "[Constructor(in float x, in float y)]" ?

[15:39:00.0000] <Hixie>
jamesr: if it's just the same value across the entire platform, then why would it be tied to a window?

[15:39:01.0000] <timeless>
oh

[15:39:02.0000] <timeless>
you allow (Foo)Bar(Baz)

[15:39:03.0000] <jamesr>
doesn't have to be

[15:39:04.0000] <Hixie>
jamesr: just require it to be a random number in a particular range at the start

[15:39:05.0000] <timeless>
ok, strange

[15:39:06.0000] <jamesr>
yeah

[15:39:07.0000] <heycam>
timeless, in fact the Appendix A grammar allows almost anything inside the square brackets

[15:39:08.0000] <jamesr>
Hixie: what is "the start"?

[15:39:09.0000] <heycam>
timeless, as long as it has balanced brackets/parens

[15:39:10.0000] <jamesr>
Hixie: the first access?

[15:39:11.0000] <Hixie>
jamesr: if it's random, doesn't much matter

[15:39:12.0000] <jamesr>
depending on the range :P

[15:39:13.0000] <timeless>
heycam: yeah, there was a note saying you allowed just about everything

[15:39:14.0000] <heycam>
timeless, and then in the #idl-extended-attributes section it defines some more restricted cases

[15:39:15.0000] <heycam>
yeah

[15:40:00.0000] <timeless>
and that the grammar was ambigusous etc

[15:40:01.0000] <jamesr>
if the allowed range is big enough then yeah, it doesn't matter

[15:40:02.0000] <Hixie>
jamesr: make it random within a range that is equivalent to a few billion years or whatever :-P

[15:40:03.0000] <timeless>
heycam, i'm in http://dev.w3.org/2006/webapi/WebIDL/#idl-extended-attributes

[15:40:04.0000] <timeless>
it's just that (foo)bar(baz) is kind strange

[15:40:05.0000] <zewt>
hmm

[15:40:06.0000] <timeless>
and i wonder if it's really useful to allow it

[15:41:00.0000] <jamesr>
Hixie: you've just invented the Yfewbillion disaster!

[15:41:01.0000] <heycam>
timeless, I wanted to allow anything in there just so that people using their own extended attributes weren't restricted too much

[15:41:02.0000] <timeless>
heycam: alright

[15:41:03.0000] <TabAtkins>
JS integers are 53 bits wide, which should be plenty enough for anyone.

[15:42:00.0000] <jamesr>
TabAtkins: these aren't necessarily integral

[15:42:01.0000] <jamesr>
(and in fact probably shouldn't be)

[15:42:02.0000] <TabAtkins>
Because of sub-millisecond timing, right?

[15:42:03.0000] <jamesr>
yeah

[15:44:00.0000] <zewt>
if the seconds value is in the billions, you can't represent nanoseconds with a double, i'm sure someone some day would curse that

[15:45:00.0000] <TabAtkins>
A gigasecond is only 30 years, after all.

[15:45:01.0000] <zewt>
(realistically I hope microseconds is enough)

[15:45:02.0000] <erlehmann>
i am “// check out

[15:45:03.0000] <erlehmann>
i am waiting for comments like „// check out my doubles“ in source code. haven't seen *that* yet.

[15:45:04.0000] <timeless>
??

[15:46:00.0000] <TabAtkins>
?? indeed

[15:46:01.0000] <Philip`>
/me wants a high-res timer so he can do lots of timing attacks and break security

[15:46:02.0000] <zewt>
and then people can blame the timer

[15:47:00.0000] <erlehmann>
timeless, a meme on imageboards. posts are numbered sequentially. it is common to end with double digits, so called “doubles” (every 11th number does). so people make a thread “check out my doubles” with a “doubles guy” pointing towards the post number. if they succeed, hate.

[15:47:01.0000] <Philip`>
/me wonders if anyone has done anything about the canvas image alpha-channel leakage yet

[15:47:02.0000] <jamesr>
quads ftw

[15:48:00.0000] <jamesr>
Philip`: pulling alpha channel alone isn't super interesting

[15:48:01.0000] <erlehmann>
Philip`, the one with the timing attack?

[15:49:00.0000] <zewt>
which one?

[15:49:01.0000] <jamesr>
Philip`: or rather, not the alpha channel by the high-order bit

[15:51:00.0000] <Philip`>
zewt: http://philip.html5.org/demos/canvas/img-timing-1.html or http://philip.html5.org/demos/canvas/img-timing-2.html

[15:51:01.0000] <Philip`>
(Relatedly also http://philip.html5.org/demos/canvas/font-timing.html)

[15:51:02.0000] <Philip`>
(All somewhat browser/OS-dependent)

[15:52:00.0000] <zewt>
but doing what?

[15:52:01.0000] <TabAtkins>
zewt: Extracting image data from a tainted canvas.

[15:52:02.0000] <jamesr>
revealing whether a pixel in a cross-origin image/canvas is fully transparent or not

[15:52:03.0000] <jamesr>
not arbitrary data

[15:52:04.0000] <Hixie>
jamesr: no, that's just the starting point. The up-side of the range should be a few trillion years more.

[15:53:00.0000] <zewt>
webgl dropped support for non-same-origin/CORS images as textures

[15:53:01.0000] <Hixie>
jamesr: and since it's relative, it just means that you can't have one browser run for more than a few trillion years.

[15:53:02.0000] <zewt>
Hixie: er, not with a double; that would limit precision to milliseconds...

[15:53:03.0000] <TabAtkins>
Darn, img-timing-2 doesn't work in chrome/linux.  It's too noisy or something.

[15:53:04.0000] <jamesr>
Hixie: you lose resolution a lot more quickly than that

[15:53:05.0000] <Hixie>
scale my numbers accordingly

[15:54:00.0000] <Hixie>
do i have to do everything here? :-P

[15:54:01.0000] <jamesr>
TabAtkins: the attack depends heavily on implementation details of the gfx libraries involved

[15:54:02.0000] <Philip`>
jamesr: I think it'd be easy to extend to the whole alpha channel - you can tell if the alpha is 0 or non-0, then just draw onto a temporary canvas with 50% globalAlpha then do timing to determine if it's 0-or-1/256 or higher, then repeat until you know all the bits

[15:54:03.0000] <jamesr>
Hixie: well, your numbers scale into ranges that we have to care about :)

[15:54:04.0000] <TabAtkins>
Hixie: Come now, at least be within orders of magnitude.  You can't even reach a million years if you're using JS numbers and counting milliseconds.

[15:54:05.0000] <jamesr>
Philip`: depending on roundoff on the intermediate canvases? that's a lot more interesting

[15:54:06.0000] <jamesr>
Philip`: can you check if that really works?

[15:55:00.0000] <jamesr>
i think we can can afford 1 year of noise and still keep 20 bits around for subpixel precision

[15:55:01.0000] <zewt>
webgl never sets a canvas to non-clean

[15:55:02.0000] <jamesr>
which is easily good enough for exactly representing nanoseconds

[15:56:00.0000] <zewt>
so I'm not sure that it matters anymore whether you can read back a texture with it

[15:56:01.0000] <jamesr>
zewt: it does

[15:56:02.0000] <jamesr>
zewt: timing attacks

[15:56:03.0000] <jamesr>
it's a lot easier to extract arbitrary image data from a webgl canvas than a 2d canvas

[15:56:04.0000] <zewt>
they just changed it a couple months ago, dropping support for non-same-origin/CORS images entirely

[15:56:05.0000] <jamesr>
right ,for this reason

[15:57:00.0000] <zewt>
so ... it doesn't, then :)

[15:57:01.0000] <zewt>
that was the point--this problem already seems fixed

[15:57:02.0000] <zewt>
unless there's something I'm missing

[15:58:00.0000] <zewt>
(it's not a very nice fix, it broke some of my code badly, which is very annoying)

[15:58:01.0000] <zewt>
early adopter woes o/~

[16:00:00.0000] <jamesr>
broke angry birds too

[16:00:01.0000] <jamesr>
it's necessary, tho

[16:02:00.0000] <zewt>
making them even angrier

[16:03:00.0000] <zewt>
i think having a really big noise factor for the base time doesn't matter: the case where that's a problem is where you can take a lot of samples and average out the noise, but in this case by definition you can only get 1 sample, ever

[16:04:00.0000] <zewt>
(it does mean you can tell fairly reliably if the user restarted his browser)

[16:05:00.0000] <zewt>
(or his OS, if that's what the time base uses)

[16:06:00.0000] <zewt>
devil's advocating: it's also a user identifier (so long as he doesn't); even with cookies and everything else disabled, and using eg. Tor, you can probably still track an active user based on his timebase

[16:06:01.0000] <zewt>
(... as he doesn't restart)

[16:08:00.0000] <jamesr>
how could you detect the timebase? juts by the approx range?

[16:08:01.0000] <annevk-cloud>
Is http://lists.w3.org/Archives/Public/public-webrtc/2011Jul/0010.html some indication Mozilla is not planning on working with the WHATWG on this? Kind of weird no email went to the WHATWG list...

[16:09:00.0000] <zewt>
when a user views a page, read the time and send it to the server; when the user views another page do it again; then correllate values

[16:09:01.0000] <zewt>
if it's the same user, the amount of time passed will be roughly the amount of time between the requests; for different users they'll be wildly different

[16:10:00.0000] <jamesr>
yeah

[16:11:00.0000] <zewt>
using Window-specific (or perhaps tab-group-specific or whatever) at least alleviates that...

[16:12:00.0000] <Philip`>
jamesr: http://philip.html5.org/demos/canvas/img-timing-3.html

[16:12:01.0000] <Philip`>
jamesr: In (at least) Opera that reproduces the grey (alpha=0.5ish) and black (alpha=1) lines

[16:13:00.0000] <jamesr>
Philip`: are you testing runs/second or seconds/run?

[16:14:00.0000] <Philip`>
jamesr: It just does 100 drawImage()s and counts the milliseconds

[16:15:00.0000] <jamesr>
yeah it would burn a lot less of my CPU and be more accurate if it tried to do drawImage()s until it hit some time threshhold, then checked how many millis expired

[16:15:01.0000] <Philip`>
then repeats forever and does the shading based on the median timing value for each cell

[16:15:02.0000] <zewt>
jamesr: could alleviate that by going in the other direction: set the timebase so that the timestamp is, say, roughly Epoch-based (according to whatever the system clock is when the browser starts up)

[16:15:03.0000] <jamesr>
zewt: i think having it be unix-epochy is just going to be more confusing

[16:15:04.0000] <Philip`>
It's a proof-of-concept, not a production-quality implementation ;-)

[16:16:00.0000] <zewt>
that way, user timebases would tend to cluster around that timebase, making correllation not useful

[16:16:01.0000] <jamesr>
Philip`: yeah, works pretty well

[16:16:02.0000] <jamesr>
ok, so the whole alpha channel :)

[16:17:00.0000] <Philip`>
You should pick an epoch far in the future, and be mysterious about the reasons for picking it, then people far in the future will think it must be counting down to some doomsday event

[16:17:01.0000] <jamesr>
i doubt webdevs can properly do math with negative numbers

[16:17:02.0000] <zewt>
burn

[16:19:00.0000] <The_8472>
negative margins already confuse them

[16:19:01.0000] <zewt>
i suspect the problem of user trackability is a bigger one than the problem of detecting how long the browser has been running...

[16:19:02.0000] <zewt>
eg. adding noise to the base is causing a worse problem than it solves

[16:20:00.0000] <zewt>
(tracking that way without noise still works sort-of, but much less reliably, I think, since tons of users will have fairly small values)

[16:20:01.0000] <jamesr>
if they are defined as not comparable across contexts, then the tracking problem goes away

[16:21:00.0000] <Philip`>
Could you make the API never return an absolute time, and just return an opaque object with a .getDifference(anotherOpaqueTimeObject) method?

[16:21:01.0000] <jamesr>
thought of that too

[16:21:02.0000] <jamesr>
pain in the butt to use

[16:21:03.0000] <The_8472>
Philip`, btw... your is a nice DoS :P

[16:22:00.0000] <The_8472>
just had to zap my browser

[16:22:01.0000] <zewt>
Philip`: that's also breaking cross-context use implicitly anyway

[16:22:02.0000] <zewt>
assuming the object isn't structured clonable or anything

[16:24:00.0000] <Philip`>
You could do a "window.timeBase = getCurrentTimeObject()" and then use getCurrentTimeObject().getDifference(timeBase) everywhere in your application and get just a number, which would be a bit verbose but no harder to use than a number-return API, unless you want to use it cross-window in which case you just pass the objects instead

[16:24:01.0000] <Philip`>
(and extend the structured clone algorithm to work with it)

[16:25:00.0000] <zewt>
i guess what that essentially does is give each page its own implicit time base, of whenever it initialized a base-time object

[16:26:00.0000] <jamesr>
or define it such that the first call to getMonotonicTime() returns 0 and everything else is relative to that

[16:26:01.0000] <Philip`>
The_8472: Yeah, some browers seem to deal quite inelegantly with it :-(

[16:26:02.0000] <zewt>
a time base which, having an actual representation, can be passed around as needed to different contexts

[16:26:03.0000] <jamesr>
so the time base is explicit and decided by the author

[16:26:04.0000] <zewt>
jamesr: that has the context problem again though

[16:27:00.0000] <zewt>
(if handling multiple contexts is given up on, then that's probably the simplest, though)

[16:27:01.0000] <jamesr>
zewt: only if we want these comparable across contexts, which i don't think is valuable

[16:31:00.0000] <zewt>
messages across threads like "do action when the current time is >= value" can be useful (more precise than saying "do it in 60 seconds"), or "time out when at a specified time" sort of things...

[16:32:00.0000] <jamesr>
that would have to be relative to when they receive a message, i think

[16:32:01.0000] <zewt>
huh? i mean relative times are less precise than calculating a time in advance and passing it down the line, generally speaking

[16:32:02.0000] <jamesr>
yeah, but we don't have shared state between workers

[16:33:00.0000] <zewt>
but if the monotonic timebase is universal across the browser then you can do that

[16:33:01.0000] <Philip`>
I guess hardware is going to become more like a distributed system in the future, so maybe it's better to avoid assuming an architecture that has a universal clock across all windows/threads/processes/etc, given that universal clocks are fundamentally impossible

[16:34:00.0000] <roc>
annevk: no, I think it's just an indication that Anant thought the RTC group was appropriate

[16:35:00.0000] <roc>
I hope that Web apps never have to worry about clock skew

[16:35:01.0000] <jamesr>
Philip`, zewt: yeah, the question really is "how much skew are you willing to put up with?"

[16:35:02.0000] <jamesr>
so i'd rather just not have any promises

[16:37:00.0000] <annevk5>
roc, would prolly be good to send at least one email indicating a fork is in the works, since quite a few people gave feedback on the WHATWG list already

[16:41:00.0000] <roc>
I'll tell him to provide feedback on the WHATWG list

[16:42:00.0000] <roc>
I don't really think of it as a fork, since I don't think Ian's proposal is baked enough for that, but maybe it is


2011-07-12
[18:43:00.0000] <MikeSmith>
and Microdata and RDFa seem to be a case where both could exist and find widespread use in the same way that Relax NG and XML Schema (or whatever) have

[18:47:00.0000] <TabAtkins>
On that note, every time I look at RDFa I become more convinced that it's a technically worse solution in general.

[18:50:00.0000] <MikeSmith>
well, I'm not personally taking any position on that, but I will say that was the view that James Clark and Murata-san expressed about XML Schema, and why they came up with Relax NG

[18:50:01.0000] <MikeSmith>
and I realize I'm harping on about XML Schema and Relax NG and that those are somewhat esoteric

[18:50:02.0000] <MikeSmith>
but to me at least that seems like the most relevant precedent

[21:32:00.0000] <roc>
did Richard Schwerdtfeger ever explain *why* his people are using canvas instead of SVG?

[00:10:00.0000] <hsivonen>
hmm. http://www.xanthir.com/demos/collidingcircles/ doesn't work for me in the Firefox build I happen to have

[00:10:01.0000] <hsivonen>
clicking the canvas does nothing

[00:27:00.0000] <hsivonen>
/me is reading 10 days of public-html email

[00:28:00.0000] <hsivonen>
the accessibility rhetoric does not seem to help with finding solutions

[02:41:00.0000] <annevk5>
XML Schema vs Relax NG is a nice one

[02:42:00.0000] <annevk5>
manu-db, despite what you suggest, people have did think CSS should be removed in favor of XSL-FO

[02:42:01.0000] <annevk5>
manu-db, same with Selectors and XPath

[02:43:00.0000] <annevk5>
manu-db, and again with Media Queries vs some XPath-based syntax

[05:01:00.0000] <Zeddy>
when accessing the page using http://www.kokonniemi.fi/guide/index.php/view/map/1/1 the manifest file saves 0kb

[05:01:01.0000] <Zeddy>
when accessing using http://kokonniemi.fi/guide/index.php/view/map/1/1 it saves 602kb

[05:04:00.0000] <Philip`>
The &lang= would probably be a validator error in the HTML, but shouldn't affect the behaviour of browsers

[05:04:01.0000] <Zeddy>
ok, had the html tag defined twice

[05:04:02.0000] <moo-_->
Zeddy: I suggest you contact your CMS authors and file a bug ;)

[05:04:03.0000] <Zeddy>
that might explain some problems x)

[05:04:04.0000] <Zeddy>
im using codeigniter

[05:06:00.0000] <Zeddy>
god damnit, now that i fixed my html the site isn't caching anything

[05:06:01.0000] <Zeddy>
is it working for you?

[05:06:02.0000] <Zeddy>
chrome again tells me the following: Application Cache Error event: Failed to parse manifest http://www.kokonniemi.fi/guide/index.php/manifest?map=1&lang=2

[05:06:03.0000] <Zeddy>
x)

[05:07:00.0000] <Zeddy>
oh, chrome doesn't like comments it seems

[05:08:00.0000] <Zeddy>
haha, neither did firefox, works in all browsers now :)

[05:08:01.0000] <annevk>
did you have a comment as first line?

[05:08:02.0000] <annevk>
because that is bogus

[05:09:00.0000] <annevk>
read http://www.whatwg.org/specs/web-apps/current-work/complete/offline.html#manifests

[05:10:00.0000] <Zeddy>
hmm, yeah just above Cache manifest x)

[05:10:01.0000] <Zeddy>
moved it beneath the line and it works

[05:13:00.0000] <annevk>
well of course

[05:13:01.0000] <annevk>
CACHE MANIFEST is the file signature

[05:13:02.0000] <annevk>
if you replace that with something random it is no longer a cache manifest

[05:13:03.0000] <Zeddy>
yeah, i realised that

[05:13:04.0000] <Zeddy>
i didn't even notice i had that line beneath the comment

[05:14:00.0000] <Zeddy>
i also noticed that i cannot output time() in the comments, because the manifest file changes during the download process and gets corrupt

[05:15:00.0000] <Zeddy>
wont the browser re-cache everything if a single file is added or deleted to the manifest?

[05:15:01.0000] <annevk>
it will

[05:15:02.0000] <Zeddy>
because i could have some kind of version number in the database which would be updated everytime the admin makes changes to text data

[05:16:00.0000] <Zeddy>
in most cases there will probably be changes in the list of files aswell, but sometimes there might be tiny corrections to the main page, and that won't automatically cause a re-cache, will it?

[05:17:00.0000] <annevk>
no

[05:17:01.0000] <annevk>
everything is loaded from cache

[05:17:02.0000] <annevk>
the manifest is then fetched and compared to the current copy

[05:17:03.0000] <annevk>
if it is changed new stuff is fetched

[05:17:04.0000] <Zeddy>
so whenever an admin makes any changes i just increment a version number which is updated in the database and i should always have everything up to date as long as the user visits the page in online mode

[05:17:05.0000] <annevk>
the next time a refresh happens the new app is loaded

[05:18:00.0000] <annevk>
if the user just keeps a tab open you might have to indicate a new version is available and a refresh is required

[05:18:01.0000] <annevk>
or some such

[05:18:02.0000] <Zeddy>
echo "#version".$map_version."\n";

[05:18:03.0000] <Zeddy>
yeah but thats a bit far fetched since the data wont be updated that often

[05:20:00.0000] <Zeddy>
Another thing i was wondering, since my pages will be dynamic aswell as the manifest file, is it possible to clear everything cached by my domain even though the url is different? Or will it perhaps do that automatically

[05:20:01.0000] <Zeddy>
Since the user might switch from http://www.kokonniemi.fi/guide/index.php/manifest?map=1&lang=2 to http://www.kokonniemi.fi/guide/index.php/manifest?map=1&lang=3

[05:20:02.0000] <Zeddy>
and i don't want it to cache both

[05:24:00.0000] <hsivonen>
now that Zakim supports SIP, what's a good way to get a provider?

[05:26:00.0000] <annevk>
does it work with Skype?

[05:28:00.0000] <Zeddy>
i don't get this, i visit my page with the manifest file, it caches the manifest, i visit it again and it also loads javascript and css not included in the manifest, but if i go offline once and visit the page again the css and javascript are obviously no there, the weird part is that when i go online again and refresh, it doesn't even attempt to load the css and javascript

[05:28:01.0000] <Zeddy>
why does it load the javascript and css when refreshing in online mode in the first place then?

[05:28:02.0000] <annevk>
because it has not seen the manifest yet

[05:29:00.0000] <Zeddy>
Huh, the first time i visit it sees the manifest and download everything, then i press F5, shouldn't it have seen the manifest at that point?

[05:30:00.0000] <annevk>
yes

[05:30:01.0000] <annevk>
oh wait

[05:30:02.0000] <annevk>
not sure what is going on there

[05:33:00.0000] <Zeddy>
you can try it yourself here: http://www.kokonniemi.fi/guide/index.php/view/map/1/1

[05:33:01.0000] <Zeddy>
it seems to be the same with both chrome and firefox

[07:35:00.0000] <hsivonen>
annevk: were we supposed to have an HTML-XML call today?

[07:41:00.0000] <manu-db>
annevk5, annevk: re: CSS should be removed in favor of XSL-FO - I hope that 'removal' of Microdata isn't the intention with the Microdata/RDFa TF. I think that they could be merged successfully (with a number of changes to Microdata or RDFa). I think this discussion could positively impact Microformats as well. I want to see that dialog start. If all of us end up in a place that we can agree...

[07:42:00.0000] <manu-db>
...on, all the better. I think we're not there today because the Microdata, the Microformats and the RDFa camp haven't been communicating as much as they should have been.

[07:43:00.0000] <manu-db>
So, yes - I agree that there was a contingent of people that wanted a black and white decision on CSS and XSL-FO, Selectors and X-Path, etc. - Personally, I'm not looking for that to happen. I want to know if the changes we're proposing to RDFa could find common ground, or if changes to Microdata could find common ground.

[07:44:00.0000] <AryehGregor>
manu-db, I don't think the syntaxes can be merged at this point, realistically.  The data models, maybe, and that would be productive.

[07:44:01.0000] <manu-db>
If the use cases are supported, I don't think that it matters which one ends up "winning".

[07:44:02.0000] <AryehGregor>
I don't see why a committee is going to be remotely useful, though.  It seems like a case of "when in doubt, form a committee".

[07:45:00.0000] <AryehGregor>
It would seem more productive for people to propose specific changes in the form of bugs.

[07:45:01.0000] <manu-db>
AryehGregor: People don't really care that much about the syntaxes - both are equally confusing to some, both are reasonable to others. The data models are important and yes, I do think a discussion about that would be productive.

[07:46:00.0000] <Philip`>
/me imagines the main danger is that merging two different systems will result in greater complexity than either of the systems by itself, since it's easier to argue for features than for simplicity

[07:46:01.0000] <AryehGregor>
I think the syntaxes aren't going to change substantially because both have too much existing content.

[07:46:02.0000] <manu-db>
AryehGregor: I don't see this as a "design by committee" exercise - maybe it'll turn into that, but I want this to primarily be a meeting between people that are in the same area (structured data), but who have not had the chance to talk to each other and express their pain points.

[07:46:03.0000] <Philip`>
(and the complexity of RDFa was one of the main reasons for designing Microdata as a replacement)

[07:47:00.0000] <hsivonen>
manu-db: is there a Web reference for removing CSS to pave way for XSL-FO?

[07:47:01.0000] <manu-db>
Philip`: yes, more complexity is one of the concerns. Now, I'm talking about stripping stuff out of RDFa (just me, not the WG) to see if that could create consensus.

[07:47:02.0000] <AryehGregor>
manu-db, well, good luck.

[07:47:03.0000] <manu-db>
hsivonen: annevk said that, not I. I just trusted that he was correct. :)

[07:48:00.0000] <hsivonen>
manu-db: It sounds plausible, sure.

[07:48:01.0000] <AryehGregor>
I don't think there are going to be strong enough incentives for compromise here.  The W3C can't exert much pressure on RDFa because it has too much entrenched support, and can't exert much pressure on microdata because the microdata people would be happy to tell the W3C to drop dead and only publish at the WHATWG.

[07:48:02.0000] <karlcow>
"do not fear death"

[07:48:03.0000] <hsivonen>
manu-db: in fact, I believe the CSS vs. XSL situation was more combative initially, but I haven't read up on that story enough

[07:49:00.0000] <manu-db>
AryehGregor: I don't think that logging bugs will do much at this point, because there are pockets of people that don't think there is a problem - or that two syntaxes isn't causing confusion. Those folks need to speak to implementers - implementers are getting increasingly agitated that they're having to put both types of markup into their pages.

[07:49:01.0000] <AryehGregor>
Theoretically everyone agrees that having only one solution is better, but in practice both sides think that solution should be theirs.

[07:49:02.0000] <hsivonen>
manu-db: anyway, I suspect it's not correct that XSL-FO has always been an XML print thing

[07:49:03.0000] <karlcow>
hsivonen: if I remember there were a few contentious

[07:49:04.0000] <AryehGregor>
manu-db, you mean authors, I guess, not implementers?  One way to solve that problem is to just decide on one or the other, as schema.org has done.

[07:49:05.0000] <karlcow>
xpath vs css selectors

[07:49:06.0000] <karlcow>
was one

[07:49:07.0000] <hsivonen>
manu-db: why would howcome have written the Considered Harmful essay if there wasn't a threat of FO leaking to the Web?

[07:50:00.0000] <hsivonen>
manu-db: which people don't care about the syntaxes?

[07:50:01.0000] <manu-db>
hsivonen: I didn't mean to say that XSL-FO has always been a "print" thing - I meant that it was /one/ of the targets - which made it a bit different than CSS. If you guys think XSL vs. CSS was a better example of something like this happening in the past, I'd be happy to use that as an example - especially if it resulted in some sort of consensus.

[07:50:02.0000] <AryehGregor>
So I don't think there's a huge problem right now.  Compromising on the syntax seems like a bad idea -- it will just result in something that's worse than either of the existing ones.

[07:50:03.0000] <hsivonen>
manu-db: do we get to kill CURIEs now if syntax isn't cared about? :-)

[07:51:00.0000] <AryehGregor>
I think the only productive thing to do at this point is to make sure you can mechanically transform microdata to RDFa and vice versa to the greatest extent possible.

[07:51:01.0000] <manu-db>
hsivonen: killing CURIEs - it's a possibility (in my mind, at least)

[07:52:00.0000] <AryehGregor>
The microdata spec has a conversion algorithm, but it's not targeted at allowing you to reuse existing RDFa vocabularies, without some translation layer -- e.g., it can only create a very specific sort of vocabulary URL.

[07:52:01.0000] <manu-db>
hsivonen: "which people don't care about syntaxes": Regular web devs (theory). In practice those that I've been able to talk to at the International Press Telecommunications Council, Good Relations, etc.

[07:52:02.0000] <AryehGregor>
So I think the best thing to do is acknowledge that microdata and RDFa will both continue to exist for now, and try to make sure they can interoperate as much as possible.

[07:52:03.0000] <karlcow>
xpath vs css selectors, xml query vs xslt, cassoulet vs choucroute, pepsi vs coke, css and fo, xml vs rdf, xhtml vs html,

[07:53:00.0000] <manu-db>
AryehGregor: The problem isn't interoperability - that's fairly easy *wild hand waving*

[07:53:01.0000] <AryehGregor>
Microformats is a separate thing, that should just be ignored.  It's a legacy hack designed to embed data in pages while still being valid HTML 4.

[07:53:02.0000] <scor>
AryehGregor: that assumes that those consuming this structured data can parse both syntaxes

[07:53:03.0000] <manu-db>
AryehGregor: The problem is that implementers are having to put /both/ Microdata /and/ RDFa into their markup to placate different folks: Facebook OGP and Google schema.org, for example.

[07:53:04.0000] <AryehGregor>
scor, yes, exactly.  Put work into making sure that the same libraries support both interchangeably without extra work, as much as possible.

[07:54:00.0000] <manu-db>
AryehGregor: We're working on that now, with the Structured Data API in the RDF Web Apps WG - it's meant to be syntax agnostic.

[07:54:01.0000] <AryehGregor>
manu-db, yeah, that's a pain.  But a compromise syntax is no more likely to resolve it than just trying to get everyone to converge on one or the other.

[07:54:02.0000] <Philip`>
/me thought every big company made up their own 'standard' vocabularies, so you'd have to insert two separate sets of metadata anyway

[07:54:03.0000] <AryehGregor>
I mean, if you announce a great new compromise tomorrow, Facebook probably isn't going to instantly support it for OGP.

[07:54:04.0000] <manu-db>
AryehGregor, but again - that's not the real problem. The real problem is that Web devs are having to put /both/ types of markup in their pages.

[07:55:00.0000] <AryehGregor>
It would probably be just as easy to get them to support microdata.

[07:55:01.0000] <AryehGregor>
The reason people have to use both isn't because there's no compromise, it's because some organizations have gone with one and some have gone with another.

[07:55:02.0000] <manu-db>
Philip` - yes, the whole NIH vocabularies are a pain. We're trying to get the large companies to converge on vocabularies, there is some progress on that front.

[07:55:03.0000] <AryehGregor>
That's not something that can be resolved in standards-land, we have to talk to the organizations.

[07:56:00.0000] <manu-db>
AryehGregor: I think you would find Google, Facebook, etc. more willing to support a unified way forward than you imagine. At least, that's my read on them.

[07:56:01.0000] <AryehGregor>
Yes, but that unified way forward may as well just be microdata.  It's not like Facebook is using RDFa because they think it's better, they're using it because nobody had heard of microdata when they introduced OGP.

[07:57:00.0000] <linclark>
AryehGregor: just catching up on this... there are some implementers of tools for authoring RDFa/microdata. we're probably a small bunch, but there are some none-the-less ;)

[07:57:01.0000] <karlcow>
hmm the convergence of vocabularies will happen through business practices. I'm pretty sure. Where and when it matters. Community of interests with business needs, requirements aka profitability

[07:57:02.0000] <manu-db>
AryehGregor: That's what I meant - we /are/ talking to the organizations and there seems to be a desire to work together on vocabs.

[07:59:00.0000] <manu-db>
AryehGregor: I think that's a fairly wild claim, to say that you understand why Facebook picked RDFa. I'm just pointing that out - I don't think any one of us can clearly understand why they picked one or the other.

[07:59:01.0000] <AryehGregor>
When was OGP introduced, and when was microdata first published?

[07:59:02.0000] <manu-db>
I think something that does hold true is that all of the companies that have picked one or the other, have not clearly understood /both/ RDFa or Microdata when they adopted it.

[08:00:00.0000] <manu-db>
OGP was announced April 21st 2010

[08:00:01.0000] <AryehGregor>
Oh, sure.  But that's true for basically every web technology.  Almost nobody gets anything except the very most concrete aspects.

[08:00:02.0000] <manu-db>
Yes, exactly

[08:01:00.0000] <manu-db>
Which is why I think that people making the argument from the standpoint of the adopting organizations /completely/ understanding the options is a very difficult argument to make.

[08:01:01.0000] <AryehGregor>
So microdata had been around for a while, but nobody actually used it until schema.org came around, and that might have been related to internal decisions within Google.  (I have no idea, I'm an outside contractor.)

[08:01:02.0000] <AryehGregor>
Whereas RDFa had been gaining traction for years.

[08:02:00.0000] <AryehGregor>
Unless you were a standards aficionado, if you heard of any technology in this area it would be RDF.

[08:03:00.0000] <AryehGregor>
Also, didn't Facebook already use RDFa in the form FBML before OGP?

[08:04:00.0000] <AryehGregor>
Anyway, this really isn't a subject I'm particularly knowledgeable about or interested in.  All I'll say is that no one seems to prefer RDFa to microdata unless they were committed to RDFa before microdata was developed, and at least one major organization (Google) that previously used RDFa is transitioning to microdata, so I'm going to bet the best way to get standardization is just to standardize on microdata.  I think the worst way to proc

[08:04:01.0000] <AryehGregor>
eed is to try introducing yet another markup language that will complicate matters still further.  Regardless, good luck.

[08:05:00.0000] <karlcow>
the walmart effect

[08:05:01.0000] <karlcow>
the elephant in the room changing the business practice for everyone else

[10:13:00.0000] <annevk>
AryehGregor, I once worked on that too btw, in my first summer here in Norway: http://annevankesteren.nl/projects/whatwg/spec

[10:13:01.0000] <annevk>
AryehGregor, I never got really far

[10:14:00.0000] <AryehGregor>
annevk, interesting.

[10:14:01.0000] <AryehGregor>
Never saw that.

[10:14:02.0000] <AryehGregor>
/me will take a look

[10:14:03.0000] <annevk>
be careful, six years ago, I didn't really know how stuff worked

[10:15:00.0000] <AryehGregor>
You'll probably have at least one angle on something that will be interesting to me.

[10:16:00.0000] <AryehGregor>
I've gotten next to no feedback on the spec so far from anyone.

[10:16:01.0000] <annevk>
that always sucks :/

[10:16:02.0000] <annevk>
you basically need to find someone that either implements or writes tests

[10:16:03.0000] <annevk>
for good feedback

[10:16:04.0000] <annevk>
otherwise it just gets stuck until reviewed

[10:17:00.0000] <annevk>
the same is true with lots of the HTML specification

[10:17:01.0000] <annevk>
the parts that were already implemented have had way less review

[10:17:02.0000] <AryehGregor>
Once I'm done polishing up the spec, hopefully within a couple of weeks, I'll publicize it and try to encourage implementers to respond speedily.

[10:17:03.0000] <annevk>
let me know how that goes :)

[10:22:00.0000] <TabAtkins>
AryehGregor: Make sure you pull in Ryosuke (rniwa@) as soon as you're ready.  He's the "editting guy" on Chrome.

[10:22:01.0000] <AryehGregor>
TabAtkins, I know, he's been my WebKit contact.

[10:22:02.0000] <TabAtkins>
Excellent.

[10:23:00.0000] <AryehGregor>
I always CC him when I have specific questions that I ask on whatwg.

[10:23:01.0000] <AryehGregor>
Ehsan is my Mozilla person, and Hallvord is my Opera person, although the latter hasn't given much feedback.

[10:24:00.0000] <AryehGregor>
When I'm ready, I'll post to public-html and CC Adrian Bateman also.

[10:25:00.0000] <TabAtkins>
annevk: When can the Origin header contain multiple origins?

[10:34:00.0000] <AryehGregor>
This is what I love about functional programming:

[10:34:01.0000] <AryehGregor>
/ "Fix disallowed ancestors of each member of sublist."

[10:34:02.0000] <AryehGregor>
sublist.forEach(fixDisallowedAncestors);

[10:34:03.0000] <AryehGregor>
s!^/!//!

[10:35:00.0000] <Philip`>
That looks like you're mutating the list

[10:35:01.0000] <Philip`>
That's not functional :-p

[10:36:00.0000] <AryehGregor>
It's not mutating the list.

[10:36:01.0000] <Philip`>
What's it doing, then?

[10:37:00.0000] <AryehGregor>
It's calling fixDisallowedAncestors on each member of sublist.

[10:37:01.0000] <AryehGregor>
That mutates the DOM, but not the list.

[10:38:00.0000] <Philip`>
/me counts mutating things reachable from the list as mutating the list, in order to keep his statement valid

[10:41:00.0000] <AryehGregor>
In that case, mutating any global variable or anything accessible from a global variable counts as a mutation, because of ownerDocument.defaultView.

[10:42:00.0000] <AryehGregor>
Hey, wait.  Why can't we use instanceof like if (foo instanceof foo.ownerDocument.defaultView.Node) to avoid multiple-global problems, at least for Nodes?

[10:42:01.0000] <AryehGregor>
Our use-case for testharness.js was mostly DOMException, admittedly.

[10:46:00.0000] <TabAtkins>
Who's "we" here?  Because that's pretty verbose and horrible for authors.

[10:46:01.0000] <Philip`>
If you really like functional programming, you should replace all global variables and DOM mutation functions with a monad system

[10:46:02.0000] <AryehGregor>
TabAtkins, I was specifically thinking of testharness.js, where we want to test things like "foo is a Node".

[10:47:00.0000] <AryehGregor>
But can't because of the multiple-global problem.

[10:47:01.0000] <AryehGregor>
But actually the major thing we wanted to test is "foo is a DOMException", and this idea doesn't help for that.

[10:47:02.0000] <AryehGregor>
Philip`, I never said *pure* functional.

[10:47:03.0000] <TabAtkins>
Ah, kk.  Then I dunno, because the DOM is confusing.  But sure, sounds reasonable.

[10:48:00.0000] <Philip`>
Impure functional isn't really functional :-p

[10:48:01.0000] <TabAtkins>
Sure it is.

[10:48:02.0000] <TabAtkins>
Otherwise Lisp isn't functional, and then "functional" isn't a very useful term.

[10:50:00.0000] <AryehGregor>
Sure it is, if you only care about Haskell.

[10:50:01.0000] <AryehGregor>
Anyone who cares about non-Haskell languages doesn't matter.

[10:50:02.0000] <TabAtkins>
Still, any definition of "functional" that excludes Lisp is a silly definition.

[10:51:00.0000] <Philip`>
It's a useful term for computations written in a subset of the language that don't do any mutation, since it means you can happily refactor things or do memoisation etc without worrying about changing the semantics of the program

[10:52:00.0000] <TabAtkins>
That's what the term "pure functional" is for.

[10:53:00.0000] <Philip`>
It doesn't seem a very useful term for referring to any syntax that happens to use functions as first-class objects, like forEach, since pretty much any language can do that

[10:54:00.0000] <TabAtkins>
Many languages still can't do so easily.  Passing around function pointers doesn't really make functions "first-class objects".

[10:59:00.0000] <AryehGregor>
"One of the problems with block-break is that its exact semantics depend on the markup language being used and the nodeName of the editing host. For example, in HTML you want to disable the ability to insert p element nodes inside a h1 element node. For generic XML languages this is almost impossible with using a schema language of some sort."

[10:59:01.0000] <AryehGregor>
annevk, I can tell this is old.  :)


2011-07-13
[18:17:00.0000] <gsnedders>
jamesr: Dragonfly does always seem to be amazingly good at crashing. :P

[18:20:00.0000] <TabAtkins>
Argh, dammit, I cant' reproduce the damned mouseup errors.  >_<

[18:23:00.0000] <TabAtkins>
Also, dammit, Chrome now does XSS protection that hurts live dom viewer.

[18:32:00.0000] <TabAtkins>
Okay, discovered the problem (or at least, one of them).  It was mouse*out* that's being fired too much.  What's worse, it's getting fired directly at the <svg> element (target set to it), even if the mouse never leaves the <svg>.

[20:55:00.0000] <jacobolus>
roc: okay, you're right. i'm not entirely sure what was up, but I cleared out a bunch of lines and now it's working as expected. thanks. :)

[23:38:00.0000] <roc>
hehe

[23:38:01.0000] <roc>
begun, the web-perf setTimeout clamping snarkfest has!

[07:00:00.0000] <gsnedders>
Does anyone (apart from Opera) do modulo for ImageDataArray?

[07:04:00.0000] <mrspeaker>
foolip: Yeah, no getters and setters - I was just trying to make it as close as possible so that when it hits then there will only be minimal changes required (in fact none, because I've added a very thin bridge).  I need it because the dom tree can change at any point and I only want to extract the microdata on a user event targeted at a specific block. And if that's not possible, then I at least need a nice way to link a microdata element to 

[07:04:01.0000] <mrspeaker>
html it came from... is there a way to do that?

[07:16:00.0000] <mtjahangir>
Hi all, I what to parse some xml like tags  (<my:abc>) with html5lib (python), but html5lib converts this to <myU0003Aabc>. I looked into source code, it seems there the tokenizer does not support namespaces. Can anyone help?

[07:17:00.0000] <annevk>
use an XML parser?

[07:22:00.0000] <mtjahangir>
annevk, using html5lib.HTMLParser

[07:25:00.0000] <annevk>
right

[07:25:01.0000] <annevk>
if you want to parse XML, use an XML parser

[07:26:00.0000] <Zeddy>
anyone here that can tell me why my android device isn't responding to the html5 manifest file? It isn't asking me if i want to allow the page to save data for me, and when i go into flight mode and refresh it tells me i need an internet connection? It's working perfectly on browsers, the test url is: http://www.kokonniemi.fi/guide/index.php/view/map/1/1

[07:31:00.0000] <mtjahangir>
Is there a XML Parser in html5lib?

[07:32:00.0000] <gsnedders>
mtjahangir: No

[07:34:00.0000] <mtjahangir>
Is this means that i can't parse tags like <my:abc> with html5lib?

[07:34:01.0000] <mrspeaker>
Zeddy: cache manifest works fine in all the android devices I've used. But it's very tricky (in android and iphone... many times i've pulled my hair out trying tto figure out what hasn't cached. Make sure you have logging on all of the cache events, and start with a super minimal test case.

[07:35:00.0000] <annevk>
mtjahangir, maybe if you use a different serialization format but it does not seem like a good idea

[07:36:00.0000] <mtjahangir>
annevk, The problem is in Parser not Serializer

[07:36:01.0000] <mtjahangir>
I use lxml for converting the tree to string

[07:37:00.0000] <annevk>
hmm

[07:37:01.0000] <annevk>
maybe you should invoke it with a different option or something?

[07:38:00.0000] <gsnedders>
mtjahangir: lxml doesn't support non-namespaced XML, so it doesn't let you have elements called "my:abc", for example.

[07:38:01.0000] <gsnedders>
mtjahangir: Are you actually dealing with HTML and not XML, though?

[07:38:02.0000] <mtjahangir>
Yes

[07:39:00.0000] <mtjahangir>
I want using some few custom tags mixed in html (like <my:for ...>)

[07:40:00.0000] <gsnedders>
mtjahangir: Such elements are invalid, though. If you want html5lib to cope, you'll have to not use lxml (I can't remember what tree-builders do cope)

[07:44:00.0000] <mtjahangir>
thnx

[07:45:00.0000] <Zeddy>
aahh shit, mrspeaker, it does work, its just really confusing when the normal cache caches all kinds of map tiles, i had to clear it, everything seems to be in order no

[07:52:00.0000] <mrspeaker>
Zeddy: good to hear!

[07:53:00.0000] <annevk>
so in Google+

[07:54:00.0000] <annevk>
if I share something someone else is sharing

[07:54:01.0000] <annevk>
what am I sharing?

[07:54:02.0000] <annevk>
RT is a much better interface

[07:54:03.0000] <annevk>
get duplicate info all the time

[07:59:00.0000] <MikeSmith>
annevk: you are making a copy

[07:59:01.0000] <MikeSmith>
it's not a reference

[07:59:02.0000] <MikeSmith>
and if somebody comments on that copy, those comments don't go to the original

[08:00:00.0000] <MikeSmith>
etc.

[08:00:01.0000] <annevk>
so when i share a copy i copy the copy?

[08:00:02.0000] <MikeSmith>
yeah

[08:00:03.0000] <foolip>
mrspeaker, items and properties are represented by HTMLElement, isn't that enough?

[08:00:04.0000] <MikeSmith>
annevk: simple RT is better, I think too

[10:06:00.0000] <timeless>
can't anyone w/ a google account sign up instantly these days?

[10:10:00.0000] <timeless>
heycam|away: ...

[10:10:01.0000] <timeless>
The following IDL fragment defines an interface that has one attribute with the [TreatNullAs] extended attribute, and one operation with an argument that has the extended attribute:

[10:10:02.0000] <timeless>
interface Dog {   attribute DOMString name;   attribute DOMString owner;

[10:10:03.0000] <timeless>
    boolean isMemberOfBreed([TreatNullAs=EmptyString] in DOMString breedName); };

[10:11:00.0000] <timeless>
anyone here see the <one attribute> <with TreatNullAs>? :)

[10:11:01.0000] <danj>
I still feel the duty to spread the word for something I'll probably never use

[10:11:02.0000] <timeless>
danj: good for you

[10:11:03.0000] <timeless>
the world needs enablers

[10:11:04.0000] <timeless>
do you also offer access to controlled substances? :)

[10:12:00.0000] <AryehGregor>
My circles are currently: 1 family, 1 acquaintances, 1 Wikimedia, 16 web standards.

[10:13:00.0000] <timeless>
AryehGregor: i need a way to see other people's web-standards circles

[10:13:01.0000] <timeless>
because my web standards circle is much smaller

[10:14:00.0000] <AryehGregor>
Well, to a first approximation everyone in my circles is web standards, so just look at my profile and add everyone you know.

[10:14:01.0000] <AryehGregor>
Hmm, it doesn't seem to let other people see everyone in my circles, only a limited selection?

[10:15:00.0000] <timeless>
i'm not actually sure how much of you i can see

[10:15:01.0000] <timeless>
or even how to try to see any of you

[10:15:02.0000] <AryehGregor>
You're in my circles, so you can see who's in my circles.

[10:15:03.0000] <AryehGregor>
That's how it's set up.

[10:15:04.0000] <AryehGregor>
/me likes the "view profile as . . ." feature

[10:15:05.0000] <AryehGregor>
You don't seem to get the "View All" link.

[10:16:00.0000] <AryehGregor>
I wonder if that's a bug.

[10:16:01.0000] <timeless>
there's a view all  link under" In Aryeh;'s circles"

[10:16:02.0000] <AryehGregor>
Ah, okay.

[10:16:03.0000] <AryehGregor>
I don't see that when I view it as you.

[10:16:04.0000] <AryehGregor>
Bug.  :(

[10:16:05.0000] <timeless>
or possibly a privacy setting?

[10:17:00.0000] <AryehGregor>
It should still display the same for you as for when I view as you.

[10:17:01.0000] <timeless>
can you give me some urls to look at?

[10:17:02.0000] <timeless>
/me is having trouble parsing this

[10:17:03.0000] <timeless>
/me is also trying to parse webidl

[10:18:00.0000] <AryehGregor>
https://plus.google.com/u/0/100662365103380396132/about

[10:18:01.0000] <AryehGregor>
Can you see who's in my circles there?

[10:18:02.0000] <AryehGregor>
I can only see eight, not the full list.

[10:19:00.0000] <AryehGregor>
Well, just pick someone else if it doesn't work, like: https://plus.google.com/u/0/107429617152575897589/posts

[10:19:01.0000] <AryehGregor>
Then follow the breadcrumbs if that's not enough for you.

[10:19:02.0000] <AryehGregor>
Actually, forget that.  Once you have a few, just go to suggestions.

[10:19:03.0000] <AryehGregor>
It will magically find all the right people for you.

[10:20:00.0000] <timeless>
without being signed in, there's AryehGregor nothing on the left side AryehGregor other than "Report this profile"

[10:20:01.0000] <AryehGregor>
Yes, I have my circles viewable only to my circles.

[10:21:00.0000] <timeless>
ah, and hixie is open

[10:21:01.0000] <timeless>
ooh, for cameron, there's a "people who have cameron in circles"

[10:22:00.0000] <AryehGregor>
Yeah, that's a separate option.

[10:22:01.0000] <AryehGregor>
I turned that off too.

[10:22:02.0000] <timeless>
grr

[10:22:03.0000] <timeless>
tablets kinda suck

[10:22:04.0000] <timeless>
it's too easy to accidentally "click" when you mean to "swipe"

[10:24:00.0000] <timeless>
ok, i need to rename my 'w3' group to 'a-w3' or something

[10:24:01.0000] <timeless>
alphabetical sorting kinda sucks..

[10:26:00.0000] <timeless>
/me ponders

[10:26:01.0000] <timeless>
<Mike|Michael> Smith - isn't such a unique name is it...

[10:27:00.0000] <AryehGregor>
Mike(tm) Smith is.

[10:27:01.0000] <AryehGregor>
Or Michael(tm) Smith.

[10:27:02.0000] <AryehGregor>
As the case may be.

[10:29:00.0000] <AryehGregor>
/me notes that MediaWiki also makes users' gender public if they choose to specify it, but the default is "Unspecified"

[10:31:00.0000] <AryehGregor>
Wow, I managed to waste an amazing amount of time fiddling with Google+ today.

[10:31:01.0000] <AryehGregor>
Back to work for me!

[10:31:02.0000] <timeless>
i reviewed an entire single point of webidl in that time!

[10:31:03.0000] <timeless>
(4.3.13)

[10:32:00.0000] <timeless>
anyone know what happens if someone tries to do: [TreatNullAs=EmptyString] attribute DOMString? confused;

[15:13:00.0000] <timeless>
heycam: current comments for today: http://pastebin.mozilla.org/1271668

[15:13:01.0000] <heycam>
timeless, thanks

[15:13:02.0000] <heycam>
timeless =~ s/m/r/

[15:13:03.0000] <timeless>
it's a lot slower than i was expecting

[15:14:00.0000] <timeless>
/me tries to apply that regexp

[15:14:01.0000] <timeless>
/me fails

[15:15:00.0000] <timeless>
oh!

[15:15:01.0000] <timeless>
yeah, well, i do sleep

[15:15:02.0000] <timeless>
persistent is the best i can offer

[15:15:03.0000] <heycam>
:)

[15:15:04.0000] <timeless>
i even saw Dirty Dancing yesterday

[15:15:05.0000] <timeless>
/me needs to check on the playlist for tongith

[15:17:00.0000] <timeless>
`The Umbrellas of Cherbourg` is on tonight

[15:17:01.0000] <heycam>
at least I've heard of Dirty Dancing…

[15:17:02.0000] <timeless>
oh, i could have seen 'Annie Hall' yesterday instead

[15:17:03.0000] <timeless>
/me has to actually make choices on Tuesdays

[15:18:00.0000] <timeless>
yeah, problem is, i've already seen Dirty Dancing

[15:18:01.0000] <timeless>
i should have gone to see Annie Hall

[15:18:02.0000] <timeless>
<wikipedia> Annie Hall is a 1977 American romantic comedy directed by Woody Allen from a screenplay co-written with Marshall Brickman and co-starring Diane Keaton. One of Allen's most popular and most honored films, it won four Academy Awards including Best Picture. Roger Ebert described it as "just about everyone's favorite Woody Allen movie".[1]

[15:19:00.0000] <timeless>
next week i'll miss `Paris Je T'aime`, but I'll be in Paris then, so it's ok :)

[15:20:00.0000] <timeless>
http://www.blogto.com/film/2011/06/free_outdoor_film_screenings_in_toronto_summer_2011/ fwiw

[15:22:00.0000] <timeless>
the friday films are entirely impractical for me unless i find someone to stay with in that area, and most of those people would object to a friday night movie

[15:23:00.0000] <timeless>
grey gardens seems a bit more doable...

[15:33:00.0000] <timeless>
cool, i managed to get google stree view to color separate

[15:36:00.0000] <timeless>
http://www2.picturepush.com/photo/a/6092230/img/6092230.png

[15:42:00.0000] <TabAtkins>
timeless: You sure that's not the 3d view?

[15:42:01.0000] <timeless>
heh

[15:42:02.0000] <TabAtkins>
'cause it's looks like anaglyph rendering

[15:42:03.0000] <timeless>
i'm sure i didn't ask for such a view

[15:42:04.0000] <TabAtkins>
And it's more pronounced on foreground objects, so that's almost certainly what you're seeing.

[15:42:05.0000] <timeless>
and the make link feature didn't give me the same view

[15:42:06.0000] <timeless>
which was disappointing

[15:44:00.0000] <AryehGregor>
The history APIs work really nicely in Google+.

[15:44:01.0000] <AryehGregor>
Extremely smooth.

[15:44:02.0000] <TabAtkins>
Yeah, I didn't even realize they were ajaxing.  I thought it was just very smooth pageloads.

[15:45:00.0000] <timeless>
didn't you read the 'ask me anything about g+' thing?

[15:46:00.0000] <heycam>
I was hitting Cmd+R the other day while on G+ and getting frustrated that the page wasn't reloading, not seeing the throbber start throbbing, but I guess it was the history apis taking care of it

[15:46:01.0000] <Hixie>
any opera people around?

[15:47:00.0000] <TabAtkins>
timeless: Yeah, that's what let me know.

[15:50:00.0000] <timeless>
heycam: so, i'm @4.5.5

[15:50:01.0000] <timeless>
i think i'm going to dump and run and try to benefit from some TIFF goodies

[15:50:02.0000] <timeless>
if i can remember what they are

[15:51:00.0000] <heycam>
ok thanks very much


2011-07-14
[20:38:00.0000] <roc>
did I kill off the "hit-testing and retained graphics" thread?

[20:41:00.0000] <MikeSmith>
roc: maybe everybody's just regrouping

[20:41:01.0000] <roc>
chilling

[20:41:02.0000] <MikeSmith>
I guess I'm paying more attention to the webrtc thread right now

[21:05:00.0000] <Hixie>
roc: how did you kill it?

[21:06:00.0000] <roc>
I dunno

[21:06:01.0000] <roc>
by posting technical content?

[21:06:02.0000] <roc>
but I wasn't the first to do that

[21:06:03.0000] <roc>
lucky shot, I guess

[23:10:00.0000] <sephr>
this really feels like something that is best handled by CSP

[23:11:00.0000] <roc>
the problem is legacy browsers that don't understand CSP

[23:11:01.0000] <othermaciej>
sephr: text/html-sandboxed is meant to fail to load at all in browsers that don't support sandboxed iframes

[23:11:02.0000] <othermaciej>
CSP does not have that feature - CSP is ignored by browsers that don't support it

[23:12:00.0000] <sephr>
othermaciej: that doesn't answer my question though about how I sandbox an image/svg+xml game

[23:12:01.0000] <othermaciej>
sephr: well I could say two things

[23:12:02.0000] <othermaciej>
1) the fact that this solution doesn't work for SVG doesn't mean that it's ok to instead adopt a solution that doesn't work for HTML either

[23:12:03.0000] <sephr>
though I get what you mean about CSP silently degrading

[23:13:00.0000] <othermaciej>
2) you can embed SVG in HTML now in browers with HTML5 parsers

[23:13:01.0000] <roc>
in fact serving your SVG game as text/html might just work

[23:13:02.0000] <sephr>
if the goal is to make it so that content doesn't load at all in browsers that don't support it, then an application/sandbox media type is more appropriate

[23:13:03.0000] <othermaciej>
market share of brewer versions that support image/svg+xml but not SVG in text/html is going to be pretty low soon

[23:13:04.0000] <sephr>
with 2 parameters: params and type

[23:14:00.0000] <LoneStar99>
Hixie:  one last question, what if I want to crop everything but a swiggly line?

[23:14:01.0000] <sephr>
so Content-Type: application/sandbox; type="image/svg+xml" params="sandbox parameters here"

[23:14:02.0000] <LoneStar99>
sorry clip

[23:14:03.0000] <sephr>
Hixie: what do you think about that example?

[23:14:04.0000] <othermaciej>
putting the sandbox parameters in an HTTP header instead of in the embedding markup defeats the purpose of sandboxing

[23:14:05.0000] <sephr>
ok then just drop the params parameter

[23:15:00.0000] <sephr>
just application/sandbox;type="normal MIME here"

[23:15:01.0000] <othermaciej>
using another content type type as a parameter to a generic type is possible, but makes MIME type parsing and dispatch a bit more complicated than browsers

[23:15:02.0000] <sephr>
?

[23:15:03.0000] <othermaciej>
so I would have to ask, what are the use cases?

[23:15:04.0000] <sephr>
;type="" is completely compliant MIME

[23:16:00.0000] <sephr>
just like many video container formats have a codecs= MIME parameter

[23:16:01.0000] <sephr>
the use cases are user generated content that isn't HTML

[23:16:02.0000] <sephr>
that you want to sandbox

[23:16:03.0000] <othermaciej>
sure, it's fine to have a MIME parameter, it's just that browsers don't normally look at MIME parameters just to decide if content is something they know how to render, if it should form an HTML document or an SVG document, etc

[23:16:04.0000] <othermaciej>
so you are making things a little more complicated

[23:16:05.0000] <othermaciej>
so you have to explain the benefit

[23:16:06.0000] <othermaciej>
give a concrete use case

[23:16:07.0000] <othermaciej>
SVG-based game is already handled by serving SVG in text/html

[23:16:08.0000] <sephr>
what are you talking about? browsers have to specially support text/html-sandboxed too

[23:17:00.0000] <sephr>
also svg can't be served as HTML

[23:17:01.0000] <sephr>
HTML5 can have svg elements in it though

[23:17:02.0000] <othermaciej>
yep, and it is just one thing to add to a list of constants, not a new parsing rule for MIME types

[23:17:03.0000] <sephr>
othermaciej: the charset parameter is used all the time in every website you visit

[23:17:04.0000] <othermaciej>
using HTML5, any SVG content you want can be wrapped in HTML

[23:17:05.0000] <sephr>
text/html;charset=...

[23:17:06.0000] <sephr>
no

[23:17:07.0000] <othermaciej>
the charset parameter is processed at a totally different level than deciding whether it is a supported element

[23:18:00.0000] <sephr>
if I write an SVG document and expect the document to behave as an XML document as it should, it may break in text/html

[23:18:01.0000] <annevk>
we could maybe add application/xml-sandboxed at one point

[23:18:02.0000] <othermaciej>
show me some real SVG content that someone may want to sandbox, but can't be wrapped in enough HTML to serve as text.html

[23:18:03.0000] <annevk>
but it is not clear that it would be that useful

[23:18:04.0000] <Hixie>
sephr: i don't see that there's anything to solve. all the content one may want to sandbox can be sandboxed with the existing type.

[23:18:05.0000] <othermaciej>
but yes, what anne says is true, just adding one type for XML is simpler

[23:18:06.0000] <roc>
annevk: we could. if t's useful

[23:18:07.0000] <Hixie>
sephr: you just make a squiggly shape and clip() that

[23:18:08.0000] <othermaciej>
for most types, making them a subtype of an application/sandboxed the would not be meaningful

[23:18:09.0000] <sephr>
Hixie: myCircle.localName === "circle" ?

[23:18:10.0000] <sephr>
nope, it'll === "CIRCLE"

[23:18:11.0000] <annevk>
yeah, the whole +xml thing was a mistake

[23:19:00.0000] <othermaciej>
and for XML types, using the more specific type is totally unimportant

[23:19:01.0000] <sephr>
which will break my hypothetical SVG document

[23:19:02.0000] <annevk>
no it will be "circle"

[23:19:03.0000] <sephr>
document.createProcessingInstruction() will also instantly break it

[23:19:04.0000] <sephr>
because HTML doesn't support processing instructions

[23:19:05.0000] <sephr>
and an error will be thrown

[23:19:06.0000] <othermaciej>
you are wrong, localName will be "circle"

[23:19:07.0000] <annevk>
lol

[23:19:08.0000] <sephr>
s/local/node/

[23:19:09.0000] <annevk>
who uses processing instructions?

[23:20:00.0000] <sephr>
some guy who wants to?

[23:20:01.0000] <othermaciej>
show me some real sag content that dynamically creates a processing instruction node

[23:20:02.0000] <othermaciej>
*svg

[23:20:03.0000] <othermaciej>
anywhere on the web

[23:20:04.0000] <othermaciej>
then we can talk

[23:20:05.0000] <sephr>
that's not the point

[23:20:06.0000] <sephr>
people will be writing more svg content now that IE supports it

[23:20:07.0000] <sephr>
in the future someone might want to use a processing instruction

[23:20:08.0000] <othermaciej>
you'll find that most people here are interested in solving real problems, not hypothetical problems

[23:21:00.0000] <othermaciej>
I can't even think of a remotely good reason to dynamically create a processing instruction node

[23:21:01.0000] <sephr>
ok here's a better example: user generated content that uses custom XML with XSLT

[23:22:00.0000] <sephr>
that requires an XML document context

[23:22:01.0000] <othermaciej>
that is a little more plausible, but the intersection of content that does that, and content you want to sandbox, seems pretty small

[23:22:02.0000] <roc>
and needs to be sandboxed ...

[23:22:03.0000] <roc>
can you point to such an example?

[23:22:04.0000] <sephr>
othermaciej: XSLT processing instructions

[23:22:05.0000] <Hixie>
/me bows out of the conversation

[23:22:06.0000] <sephr>
the world of warcraft site used to do tht

[23:22:07.0000] <sephr>
that*

[23:22:08.0000] <sephr>
that being using XSLT

[23:22:09.0000] <sephr>
that's a pretty large site

[23:23:00.0000] <sephr>
also I know a few sites that use it

[23:23:01.0000] <othermaciej>
and does anyone have a use case for loading the world of warcraft site in a sandboxed iframe?

[23:23:02.0000] <sephr>
you don't get it

[23:23:03.0000] <sephr>
what if the world of warcraft site wants to allow user submissions

[23:23:04.0000] <sephr>
of minigames or whatever

[23:23:05.0000] <othermaciej>
user submissions that themselves use XSLT?

[23:23:06.0000] <roc>
minigames that use XSLT?

[23:23:07.0000] <sephr>
and they want to make some official minigames

[23:24:00.0000] <sephr>
they won't be able to use what they usually use

[23:24:01.0000] <sephr>
just because they can't sandbox XML

[23:24:02.0000] <othermaciej>
I don't think I've ever seen a Web-based game that used XSLT

[23:24:03.0000] <roc>
I don't think Blizzard's Web developers would have difficulty NOT using XSLT

[23:24:04.0000] <sephr>
and that invalidates my point somehow?

[23:25:00.0000] <sephr>
that's like publishing this in the spec: image/svg+xml is now to be banned from interacting with HTML5 documents

[23:25:01.0000] <sephr>
you may no longer use X Y or Z

[23:25:02.0000] <othermaciej>
well, roc and I both have some responsibility for how the engineering time of particular browser vendors is spent

[23:25:03.0000] <annevk>
that's nonsense

[23:25:04.0000] <roc>
find us a Web developer who wants to build a site with sandboxed user submissions, and a user who wants to submit something that uses XSLT

[23:25:05.0000] <sephr>
yes most people don't use X Y or Z

[23:25:06.0000] <sephr>
so of course it won't matter to you

[23:26:00.0000] <othermaciej>
I at least am usually interested in Safari engineers spending time on things that Web developers *actually* want to do

[23:26:01.0000] <othermaciej>
as opposed to some theoretical thing that you could imagine someone could conceivably do

[23:26:02.0000] <othermaciej>
we have our hands full just meeting the real requests for functionality

[23:26:03.0000] <sephr>
How can I provide current examples of stuff that isn't going to happen *until* you can sandbox SVG?

[23:26:04.0000] <sephr>
for example

[23:26:05.0000] <roc>
just find real developers and users who want to do it on real sites

[23:27:00.0000] <roc>
but currently can't

[23:27:01.0000] <othermaciej>
just provide an example of SVG that in theory could be usefully sandboxed, but can't be served as text/html

[23:27:02.0000] <sephr>
there are no sites that accept SVG submissions because you can't sandbox it

[23:27:03.0000] <othermaciej>
or Web developers who wish they could do it but can't

[23:27:04.0000] <annevk>
they could host it on separate domains at the moment

[23:27:05.0000] <annevk>
which is how we found out about the need for sandboxed HTML

[23:28:00.0000] <othermaciej>
maybe you even have a first-person case of some SVG that you wish you could put in a sandboxed iframe, but alas, it uses something that can't be done in the text/html syntax

[23:28:01.0000] <sephr>
othermaciej: <svg xmlns="http://www.w3.org/2000/svg"><script type="application/ecmascript">if(!document.xmlVersion){throw 'nope'}/*app here*/</script></svg>

[23:28:02.0000] <sephr>
that XML game right there won't work in text/html for example

[23:28:03.0000] <sephr>
shortest example possible though of course

[23:28:04.0000] <othermaciej>
that doesn't look like a very fun game

[23:29:00.0000] <othermaciej>
the idea here is to find real-world examples that show the potential benefits

[23:29:01.0000] <sephr>
since when it is whatwg's choice to decide whether or not I can use xml or not though?

[23:29:02.0000] <annevk>
xmlVersion is going away so that wouldn't work anyway

[23:29:03.0000] <othermaciej>
not make up fake examples that are created solely to prove your point

[23:29:04.0000] <sephr>
annevk: cite please

[23:29:05.0000] <annevk>
and actually, if it didn't go away it would be on all documents

[23:29:06.0000] <sephr>
I use that for detecting if a context is xml

[23:29:07.0000] <annevk>
sephr, http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html

[23:29:08.0000] <othermaciej>
ultimately you are not just asking the whatwg, you are asking browser vendors to implement something

[23:30:00.0000] <sephr>
how will I detect if a document is xml now without testing node case?

[23:30:01.0000] <sephr>
also I still see xmlVersion on http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html

[23:30:02.0000] <othermaciej>
it would be nice if the html-ness vs xml-ness of a document was exposed, but I don't think there is any good clean API for it

[23:30:03.0000] <sephr>
othermaciej: there is: it's !!document.xmlVersion

[23:31:00.0000] <sephr>
it's null in HTML

[23:31:01.0000] <sephr>
1.X in XML documents

[23:31:02.0000] <annevk>
sephr, it is in the section of removed features

[23:32:00.0000] <b0o-supermario>
Hi

[23:32:01.0000] <othermaciej>
I suppose that is less indirect than other ways

[23:33:00.0000] <sephr>
annevk: I don't see a removed features list

[23:34:00.0000] <sephr>
s/list/section/

[23:34:01.0000] <Hixie>
sephr: it's easier to make that example work in text/html-sandboxed than it is for us to introduce a way to do this in xml

[23:34:02.0000] <annevk>
sephr, euh... search for "Historical"

[23:35:00.0000] <sephr>
ok thanks

[23:35:01.0000] <sephr>
Hixie: so there will never be application/xml-sandboxed just because it's unpopular?

[23:35:02.0000] <Hixie>
sephr: no, that's not what i'm saying

[23:35:03.0000] <sephr>
ok

[23:35:04.0000] <Hixie>
sephr: we try to do things in the order that they are important

[23:35:05.0000] <Hixie>
sephr: we first address the things that are most problematic

[23:36:00.0000] <Hixie>
sephr: and we walk down this list

[23:36:01.0000] <Hixie>
sephr: it's possible that in 198 years, we'll be so far down this list that sandboxing svg sent as xml will be the highest priority

[23:36:02.0000] <Hixie>
sephr: but currently, evidence suggests that it's not a high priority

[23:36:03.0000] <Hixie>
sephr: so we can't afford to care

[23:37:00.0000] <sephr>
though your list should also weighed by how easy something is to do

[23:37:01.0000] <othermaciej>
easy to spec, or easy to implement?

[23:37:02.0000] <sephr>
looking at the text/html-sandboxed documentation it looks like it'd be a simple copy and replace

[23:37:03.0000] <sephr>
I mean copy and paste

[23:38:00.0000] <othermaciej>
of spec text or code?

[23:38:01.0000] <sephr>
spec text

[23:38:02.0000] <sephr>
no major changes

[23:39:00.0000] <othermaciej>
since implementation is orders of magnitude more costly than writing the spec text for even the most trivial of features, it seems a bit silly to weight by how easy it is to spec something

[23:39:01.0000] <othermaciej>
furthermore, for anything security-sensitive, if it is really easy to spec *or* to code, you're probably doing it wrong

[23:40:00.0000] <sephr>
the less complex security spec text is the safer

[23:40:01.0000] <sephr>
just because code isn't complex doesn't mean it's more likely to be insecure

[23:40:02.0000] <othermaciej>
indeed, and it takes a lot of time to minimize complexity and to make sure what is left is sound

[01:43:00.0000] <hsivonen>
Hixie: speaking of work licensing, on surface, it seems to me that the work licensing vocab is something you might want to use in addition to another vocab

[01:43:01.0000] <hsivonen>
how do you do that with Microdata

[01:43:02.0000] <hsivonen>
(the other vocabs saying something non-copyright-related about the same images or something)

[01:44:00.0000] <Hixie>
just have two items

[01:49:00.0000] <hsivonen>
Hixie: can you have two overlapping items so that the <img> belongs to both?

[01:50:00.0000] <Hixie>
there are ways, but not ways i would recommend

[02:06:00.0000] <hsivonen>
Hixie: isn't that a bug then? or YAGNI?

[02:07:00.0000] <Hixie>
hsivonen: none of the use cases that people brought up needed it

[02:08:00.0000] <hsivonen>
Hixie: ok

[02:08:01.0000] <Hixie>
hsivonen: if there are concrete use cases that need it, maybe we can revisit that

[02:15:00.0000] <jps>
so, how is microphone audio upload doing?

[02:20:00.0000] <Hixie>
jps: you mean <input type=file accept=audio/*> ?

[02:20:01.0000] <Hixie>
jps: or streaming microphone upload?

[02:20:02.0000] <jps>
both?

[02:20:03.0000] <Hixie>
well they're both in the spec and slowly getting implemented

[02:21:00.0000] <Hixie>
dunno about specifics

[02:21:01.0000] <Hixie>
try it :-)

[02:22:00.0000] <jps>
is there a version of the spec with the annotations saying which parts have been implemented?

[02:22:01.0000] <annevk>
whatwg.org/C

[02:22:02.0000] <annevk>
but not everything is annotated correctly

[02:23:00.0000] <annevk>
Hixie, btw, I'm in favor of going back to one spec

[02:23:01.0000] <jps>
http://www.whatwg.org/specs/web-apps/current-work/multipage/number-state.html#file-upload-state has annotations dated 2009

[02:23:02.0000] <Hixie>
annevk: yeah, i might do that and set up redirects and stuff

[02:23:03.0000] <Hixie>
annevk: trying to catch up a bit on feedback first

[02:24:00.0000] <annevk>
Hixie, also, how easy would it be to host other specs on whatwg.org/specs/ ?

[02:24:01.0000] <annevk>
Hixie, like DOM Range and such

[02:25:00.0000] <annevk>
Hixie, and maybe even DOM Core / XHR for better copyright conditions

[02:25:01.0000] <Hixie>
should be possible to figure something out

[02:25:02.0000] <Hixie>
either in .../specs/ or on separate subdomains

[02:25:03.0000] <Hixie>
(subdomains are easier to set up, but subdirectories should be possible too)

[02:25:04.0000] <Hixie>
not right now though, it's way past my bed time

[02:26:00.0000] <Hixie>
catch me online tomorrow and let's figure something out

[02:27:00.0000] <annevk>
no rush

[02:27:01.0000] <annevk>
and good night

[02:27:02.0000] <Hixie>
k

[02:27:03.0000] <Hixie>
nn

[02:27:04.0000] <jps>
is http://www.whatwg.org/specs/web-apps/current-work/multipage/the-video-element.html#the-audio-element for microphone streaming?  I'm having trouble finding that part

[02:27:05.0000] <annevk>
getUserMedia is

[02:28:00.0000] <jps>
thanks

[11:55:00.0000] <TabAtkins>
Google will do that for you now.

[11:55:01.0000] <timeless>
something to show what the greatest distance you could be from a major city

[11:55:02.0000] <timeless>
TabAtkins: i tried, and it isn't working

[11:55:03.0000] <manu-db>
hsivonen: seems hacky to me - think about it a bit more and let me know if you still think we should be adding attributes for vocabulary terms.

[11:56:00.0000] <hsivonen>
manu-db: I expect it to be common that some people want subitems to be just strings while others want them to be deeper objects

[11:56:01.0000] <hsivonen>
it would be nice to have flattening deeper object to a title into a string

[11:56:02.0000] <hsivonen>
manu-db: suppose you have the concept for an album

[11:56:03.0000] <hsivonen>
manu-db: for some cases, you may want tracks to be just track names

[11:57:00.0000] <cygri>
hsivonen +1

[11:57:01.0000] <TabAtkins>
@itemalt

[11:57:02.0000] <hsivonen>
manu-db: in other cases, you may want tracks themselves to have properties

[11:57:03.0000] <hsivonen>
manu-db: of you may have a list of people be just names or objects with more properties

[11:57:04.0000] <cygri>
rdf was supposed to have that at some point, but they couldn't agree on how to do it properly, and it ended up as the underspecified rdf:value property that no one uses

[11:57:05.0000] <TabAtkins>
(But really, the "I want something to both be a string and have more properties" thing is commonly useful.

[11:58:00.0000] <manu-db>
cygri, hsivonen, TabAtkins: I agree - but that's what we use rdfs:label for - and really, this is application layer stuff.

[11:58:01.0000] <manu-db>
it seems to conflate pushing application logic down into the data model.

[11:59:00.0000] <TabAtkins>
manu-db: Really?  I mean, this pattern is useful for things like the CSSOM.

[11:59:01.0000] <hsivonen>
manu-db: if we had a framework-level way to have a name for things, we could have both forward compat (people find they want to expand a value from string to a subitem later) and a way to generate simpler views generically

[11:59:02.0000] <Philip`>
http://www.wolframalpha.com/input/?i=paris+to+london is nice for finding distances between places

[11:59:03.0000] <manu-db>
I'm not saying it isn't useful - just that that sort of stuff is better handled at the application layer "What string should I use to display this item on a page"

[11:59:04.0000] <cygri>
manu-db, to some extent, *everything* is better handled at the application layer

[12:00:00.0000] <manu-db>
TabAtkins: I know nothing of CSSOM from a spec point of view - so I can't intelligently discuss it.

[12:00:01.0000] <cygri>
manu-db, it's just that some things happen to be needed so often, so why not generlize it. type being the prime example

[12:00:02.0000] <hsivonen>
timeless: was "help" about city distances?

[12:00:03.0000] <timeless>
yeah

[12:00:04.0000] <hsivonen>
timeless: I don't have a better suggestion than Philip`

[12:00:05.0000] <AryehGregor>
manu-db, he's saying that for <span style="color: red">, it's useful to have span.style.color == "red", but also have properties like span.style.color.asHex == "#ffffff" or something.

[12:00:06.0000] <AryehGregor>
(not sure what the exact proposed syntax is)

[12:01:00.0000] <timeless>
hsivonen / Philip` : the problem is that i don't want to find the distance from any city to any other city

[12:01:01.0000] <timeless>
i want a limit on distances between cities :)

[12:01:02.0000] <timeless>
> most of Europe is pretty close to most of the rest of Europe. You probably can't find a point which is >125miles from a major city in continental Europe (this is based on eyeballing France). For the Nordic countries, on a per country basis, and picking an unreasonable definition of a city, that limit is probably closer to 300miles.

[12:01:03.0000] <Ms2ger>
Well, more intersting would be style.color.red

[12:01:04.0000] <manu-db>
AryehGregor: oh - thanks, that helps clarify. Still having a hard time figuring out what the use case for this is...

[12:01:05.0000] <timeless>
does the above statement sound reasonable?

[12:01:06.0000] <AryehGregor>
Ms2ger, . . . what would that do?

[12:01:07.0000] <TabAtkins>
manu-db: Short version is, the CSSOM would like to both expose a property's value as a string (what you inputted) and as a complex object that can be manipulated more easily.

[12:01:08.0000] <Ms2ger>
Give you an int

[12:02:00.0000] <manu-db>
TabAtkins: But that's what Projections do in the RDFa API.

[12:02:01.0000] <Ms2ger>
Like, style.color.red/2

[12:02:02.0000] <Ms2ger>
Like, style.color.red/=2, that is

[12:02:03.0000] <hsivonen>
timeless: does any country have a "reasonable" definition for a "city" anymore?

[12:02:04.0000] <timeless>
hsivonen: doubtful

[12:02:05.0000] <Philip`>
timeless: Presumably restricted to points which are in the country and are not in the sea? :-)

[12:02:06.0000] <hsivonen>
are cities in Britain still defined in terms of bishops?

[12:02:07.0000] <manu-db>
var tab = document.data.getProjection("#tab", projectionTemplate);

[12:02:08.0000] <manu-db>
tab.name === "Tab Atkins Jr."

[12:02:09.0000] <manu-db>
something like that?

[12:03:00.0000] <timeless>
hsivonen: the context is someone claims that European countries allowed the "free market" to provide "internet access"

[12:03:01.0000] <AryehGregor>
Ms2ger, okay, so what would *that* do?

[12:03:02.0000] <Philip`>
hsivonen: http://www.bbc.co.uk/news/magazine-13841482 - "The definition of a city in the UK is a place which has been granted city status by the monarch."

[12:03:03.0000] <timeless>
and got it for everyone without government regulation

[12:03:04.0000] <timeless>
Philip`: nice

[12:03:05.0000] <TabAtkins>
AryehGregor: That would cut the red component of the color in half.

[12:03:06.0000] <AryehGregor>
Philip`, laws are also only approved by the monarch, officially.

[12:03:07.0000] <timeless>
and yeah, iirc there are 2 cities which don't have  a bishop

[12:03:08.0000] <hsivonen>
(surprisingly, defining it in terms of bishops is much better than defining it in terms of "you get the call yourself a city if you can bear the laughter of the neighboring village")

[12:03:09.0000] <AryehGregor>
TabAtkins, oh, I see, yeah.

[12:03:10.0000] <AryehGregor>
That's useful.

[12:04:00.0000] <TabAtkins>
manu-db: Where are projections defined?

[12:04:01.0000] <hsivonen>
Philip`: does that correlate with Anglican dioceses?

[12:05:00.0000] <hsivonen>
Philip`: reading the article, it seems it doesn't anymore

[12:05:01.0000] <manu-db>
TabAtkins: http://www.w3.org/TR/2011/WD-rdfa-api-20110419/#advanced-concepts

[12:05:02.0000] <TabAtkins>
Ah, I was looking in rdfa-api

[12:05:03.0000] <manu-db>
not that we're still re-working through the interfaces heavily

[12:05:04.0000] <TabAtkins>
I mean, rdfa-syntax

[12:05:05.0000] <manu-db>
most everything is going to change in that doc

[12:05:06.0000] <manu-db>
but the basic premise is there.

[12:06:00.0000] <Philip`>
hsivonen: Doesn't seem to, and http://en.wikipedia.org/wiki/City_status_in_the_United_Kingdom says "In the twentieth century, it was explicitly recognised that the status of city in England and Wales would no longer be bound to the presence of a cathedral, and grants made since have been awarded to communities on a variety of criteria, including population size."

[12:06:01.0000] <manu-db>
we're also working through these concepts with JSON for Linked Data (JSON-LD): http://json-ld.org/playground/

[12:06:02.0000] <hsivonen>
timeless: uh, the availability of Internet access in Europe *is* based on govt. regulation

[12:06:03.0000] <manu-db>
The whole concept of taking a graph (which sucks to work with directly) and converting it into a nice JSON object.

[12:06:04.0000] <timeless>
hsivonen: yeah...

[12:06:05.0000] <Philip`>
Cambridge is a city but doesn't have a cathedral and doesn't have much population either (apparently 100K, of which a quarter are students)

[12:10:00.0000] <hsivonen>
manu-db: for JSON, this would mean that someone defines a property to be string-valued first and then you find you need it to be an object. if you haven't defined processor behavior forward-compatibly so that you get to replace a string with an object with a special key in it, you are in trouble

[12:10:01.0000] <TabAtkins>
(That's precisely the problem the CSSOM has, btw.)

[12:11:00.0000] <manu-db>
hsivonen: are you talking about getProjection(), JSON-LD, or something else, Henri (sorry, involved in slightly different conversations w/ different folks on here)

[12:11:01.0000] <annevk>
What is that, TabAtkins?

[12:12:00.0000] <hsivonen>
manu-db: I'm talking about JSON in general. dunno if JSON-LD has solved this

[12:12:01.0000] <TabAtkins>
annevk: That we started with a string-based API, and then want to change to an object-based one.  We can't do an easy upgrade, so we have to instead invent a new access method for the latter api.

[12:12:02.0000] <annevk>
ah yeah

[12:12:03.0000] <hsivonen>
I guess I should read up on JSON-LD before talking about problems that may already be solved

[12:13:00.0000] <annevk>
CSSOM is ugly

[12:13:01.0000] <hsivonen>
(though not solved for microdata and RDFa, AFAICT)

[12:13:02.0000] <Ms2ger>
s/cssom/the web/

[12:14:00.0000] <manu-db>
hsivonen: Could you write out the problem in code... or interpretive dance? (your choice)

[12:14:01.0000] <Philip`>
timeless: Maybe you could use OpenStreetMap's data, then extract places with a sufficiently large population (I have no idea if they provide that data reliably), then just do something like a Voronoi diagram and find the largest cell

[12:14:02.0000] <Philip`>
timeless: Or you could just pluck a number out of nowhere and claim it forcefully and assume nobody is going to check

[12:15:00.0000] <Ms2ger>
That would be the "HTMLWG approach"?

[12:15:01.0000] <manu-db>
Is the problem this: in 2011, my code does this to access an object describing you: if(henri.name instanceof String) { DO_SOMETHING };

[12:16:00.0000] <erlehmann>
manu-db, is there some kind of json pastebin with an API that works cross origin?

[12:16:01.0000] <manu-db>
then in 2012, that code fails because henri.name instanceof Object == true ?

[12:16:02.0000] <erlehmann>
i have a “linked data” project regarding kissology.

[12:16:03.0000] <TabAtkins>
manu-db: Basically, yes.

[12:16:04.0000] <erlehmann>
(no joke.)

[12:16:05.0000] <TabAtkins>
manu-db: And that sounds perfectly reasonable, as "name" may start as a string, then get broken into first/last/etc

[12:17:00.0000] <manu-db>
erlehmann: I don't understand the question - are you asking: "Can I just dump in a URL in the input field and get output?"

[12:17:01.0000] <Philip`>
Strings are objects - can't you just stick more properties (or getters) onto them, and not care about anyone using the APIs in Java?

[12:17:02.0000] <TabAtkins>
Philip`: Apparently augmenting DOMStrings wasn't acceptable for the CSSOM.  Anne would know more.

[12:17:03.0000] <erlehmann>
manu-db, some kind of semi-permanent archive for json data, accessible via script.

[12:17:04.0000] <Philip`>
It would be admittedly quite evil

[12:18:00.0000] <Philip`>
and highly surprising to developers

[12:18:01.0000] <AryehGregor>
Wouldn't it mean typeof foo == "object" instead of "string"?

[12:18:02.0000] <TabAtkins>
Yes.

[12:18:03.0000] <AryehGregor>
Or else you'd have a string that's magical.

[12:19:00.0000] <TabAtkins>
It'd be an object with a special toString, I guess.

[12:19:01.0000] <AryehGregor>
That might work for compat, or so you'd think.

[12:19:02.0000] <hsivonen>
manu-db: at time t_0 I write code that assumes that tracks in an album are strings. At time t_1, someone decides that all tracks become objects that have properties for performers, etc. I want my code to keep working with data that is published with performers so that my code just gets the track title for each track and ignores the performers

[12:19:03.0000] <manu-db>
erlehmann: There is nothing that I know of that fits that description, sorry. Are you looking for some large repository of JSON data?

[12:19:04.0000] <TabAtkins>
Again, you'd have to ask annevk why that wasn't workable for the cssom

[12:19:05.0000] <hsivonen>
manu-db: so at t_0, I need to know what the "must ignore" forward compat rules are

[12:20:00.0000] <erlehmann>
manu-db, no. doesn't matter in the end, i have written my own script for that task. forget about it.

[12:20:01.0000] <annevk>
TC39 did not like magic strings

[12:20:02.0000] <annevk>
and changing typeof to object would cause too much breakage

[12:20:03.0000] <timeless>
/me sighs

[12:20:04.0000] <hsivonen>
does WebKit still have magic strings?

[12:20:05.0000] <TabAtkins>
re hsivonen's case, strings are probably priviledged here in being something that's very likely to be expanded into a complex object in the future, much more so than other datatypes.

[12:21:00.0000] <TabAtkins>
I guess we can't just subclass DOMString to CSSOMPropertyValue or something.

[12:21:01.0000] <manu-db>
hsivonen: So, strictly speaking, in the JSON-LD framing case (which may find its way back into RDFa Projections) - you can specify that you want the value to be a string as a matching rule. That is, you can say: "Make sure that this value is a string, or set the value to 'null'"

[12:22:00.0000] <manu-db>
hsivonen: So, you are somewhat protected against objects that declare objects for a certain property instead of a string.

[12:23:00.0000] <TabAtkins>
manu-db: Only if you explicitly, every time you query for a string property, specify that you want it as a string. :/

[12:23:01.0000] <hsivonen>
manu-db: I don't want to fall back to null. I want to fall back onto what's most like the name of the object

[12:23:02.0000] <manu-db>
but that's kind of annoying because your application won't work with half of the data out there... it's a messy problem. You either fix it in the application layer, or you restrict the application layer to only work with a subset of the data that you deem appropriate. This is our current thinking on the matter, which could change wildly based on implementation feedback/guidance.

[12:24:00.0000] <hsivonen>
manu-db: this is totally an armchair comment--not based in impl. experience :-)

[12:24:01.0000] <hsivonen>
(my comment is an armchair one that is)

[12:24:02.0000] <manu-db>
TabAtkins, hsivonen: Again, this seems like an application layer thing. Different applications require different things to be used when serialized to their "string forms"

[12:24:03.0000] <manu-db>
I agree that it would be nice if everyone could decide on the string to use to refer to an object.

[12:25:00.0000] <hsivonen>
/me 's current chair has neither armrests nor a backrest, though

[12:25:01.0000] <manu-db>
but data is dirty - very difficult for people in the same industry to even agree what an object should be serialized to in string form.

[12:25:02.0000] <TabAtkins>
manu-db: Actually, this now sounds like a reasonably common case, now that I'm thinking of it.  Upgrading a string to an object happens reasonably often in the real world, and having an easy way to specify on the data level what the legacy string form of something is seems potentially useful.

[12:27:00.0000] <TabAtkins>
/me will propose something for Microdata about it.

[12:27:01.0000] <manu-db>
TabAtkins, hsivonen: I don't disagree - just trying to figure out how to make that happen in a generic way. We'd probably implement in JSON-LD and may backport to RDFa Projections. The goal of both of those is to take messy graph data and present it to a Web Developer in a way that is just a simple JSON object.

[12:29:00.0000] <hsivonen>
for a JSON-based thing that has no legacy yet (dunno if JSON-LD fits that criterion), one could reserve the object key "name" for the purpose and decree that all objects that used to be just string have to call the property that's closest to the legacy string 'name'

[12:30:00.0000] <hsivonen>
hmm. that's grammatically ambiguous English, because English doesn't allow me to use commas to mark the scope of "that"

[12:30:01.0000] <manu-db>
hsivonen: That would be fairly easy to do in JSON-LD.

[12:30:02.0000] <TabAtkins>
hsivonen: Context resolves all three "that"s for me.

[12:31:00.0000] <manu-db>
TabAtkins, hsivonen: For an example of what framing looks like in JSON-LD, go here: http://json-ld.org/playground/ and then click on the "Library" example. Then click between the "Normalized" and "Framed" tabs.

[12:32:00.0000] <TabAtkins>
manu-db: Ah yeah, that's very amenable to solving this problem (if it's determined to be worthwhile to solve.)

[12:32:01.0000] <manu-db>
(or the TURTLE and Framed tabs)

[12:32:02.0000] <TabAtkins>
Ah, yeah, Turtle's a lot easier to read than "Normalized". ^_^

[12:32:03.0000] <hsivonen>
I'd expect books to collapse to their dc:title

[12:33:00.0000] <manu-db>
/me nods - normalized is just there for digital signatures in PaySwarm: http://payswarm.com/

[12:33:01.0000] <hsivonen>
manu-db: shades of XML Canonicalization and Signature, eh?

[12:34:00.0000] <manu-db>
/me grins.

[12:34:01.0000] <manu-db>
We hope this mechanism sucks less.

[12:34:02.0000] <manu-db>
Solving this problem was actually worse than XML Canonicalization and Signature... because you have to deterministically normalize a graph, which can have cycles.

[12:35:00.0000] <manu-db>
I don't think they had to deal with that problem in XML Canonicalization (but I could be entirely mistaken)

[12:35:01.0000] <hsivonen>
at least XML Canonicalization was tricky enough that they had to try again with Exclusive Canonicalization...

[12:36:00.0000] <manu-db>
We've discovered that all of the degenerate cases for graph normalization have to do with graphs that are completely useless in the real world (which is good)

[12:37:00.0000] <manu-db>
but it still opens up the possibility of poisoned messages (for digital signatures) that cause the solution to stray into NP territory (which can be detected and killed)

[12:38:00.0000] <manu-db>
in other words - we think the normalization algorithm we have for JSON-LD will work for useful graphs as well as useless graphs where the number of unlabeled nodes without any properties stay below a developer-defined threshold.

[12:39:00.0000] <manu-db>
and the entire processor can be implemented in less than 1000 lines of Python.

[12:42:00.0000] <manu-db>
/me grins - it never fails - discussing graph normalization algorithms kills the conversation. Every time. :P

[12:42:01.0000] <TabAtkins>
I'm writing an email!

[12:43:00.0000] <manu-db>
/me nice try, Tab.

[12:43:01.0000] <jamesr>
he's normalizing his graph, amirite?!

[12:43:02.0000] <Ms2ger>
So, is svg:script supposed to work in text/html in Chrome?

[12:43:03.0000] <TabAtkins>
I normalized your mom's graph.  No node left blank.

[12:46:00.0000] <manu-db>
oooh!

[12:50:00.0000] <hsivonen>
Ms2ger: it would be rather shocking if it didn't

[12:51:00.0000] <Ms2ger>
That's what I thought

[12:56:00.0000] <AryehGregor>
Could someone explain to me why Aurora 7.0a2 refuses to install Firebug even though I have extensions.checkCompatibility.7.0a2 set to false in about:config?

[12:56:01.0000] <AryehGregor>
Better, could someone explain why there's no extensions.checkCompatibility.aurora or something?

[12:56:02.0000] <AryehGregor>
Aurora is not really usable for testing if extensions break every time you upgrade.

[12:56:03.0000] <jcranmer>
there is now

[12:56:04.0000] <Ms2ger>
There is?

[12:56:05.0000] <jcranmer>
checkCompatibility.<version> doesn't work for nightly/aurora (maybe beta?)

[12:57:00.0000] <AryehGregor>
Apparently not.

[12:57:01.0000] <AryehGregor>
So what should I use instead?

[12:57:02.0000] <AryehGregor>
/me searches Bugzilla

[12:57:03.0000] <timeless>
extensions.checkCompatibility.nightly

[12:58:00.0000] <timeless>
is what i use

[12:58:01.0000] <timeless>
you could just install Add-on Compat thingy

[12:58:02.0000] <AryehGregor>
That works for Aurora?

[12:58:03.0000] <timeless>
no :) you should suffer w/ nightly like everyone else :)

[13:00:00.0000] <AryehGregor>
How stable is nightly in practice?

[13:00:01.0000] <timeless>
i don't have it crash

[13:00:02.0000] <timeless>
but it's unstable in that it pesters me to update every day

[13:00:03.0000] <timeless>
typically twice each day i pay attn

[13:00:04.0000] <Ms2ger>
I only have it crash when I break it myself

[13:01:00.0000] <timeless>
yeah, i only have it die if i terminate the process because i want to save my session and restart my computer

[13:01:01.0000] <timeless>
or if i want to quit it because the *stupid* flash updater demands it not be running in order to update flash

[13:02:00.0000] <AryehGregor>
Oh, I think the right magic incantation is extensions.checkCompatibility.7.0a, not a2.

[13:02:01.0000] <AryehGregor>
Sigh.

[13:02:02.0000] <timeless>
yeah, that's likely

[13:02:03.0000] <AryehGregor>
Well, the add-on compat extension seems to work so far, so it's all good.

[13:02:04.0000] <timeless>
i have:

[13:02:05.0000] <timeless>
extensions.checkCompatibility.7.0

[13:02:06.0000] <timeless>
extensions.checkCompatibility.7.0a

[13:02:07.0000] <timeless>
extensions.checkCompatibility.8.0

[13:02:08.0000] <timeless>
extensions.checkCompatibility.8.0a

[13:02:09.0000] <timeless>
extensions.checkCompatibility.nightly

[13:02:10.0000] <AryehGregor>
Me too.

[13:03:00.0000] <timeless>
plus ones for 60a and older

[13:03:01.0000] <AryehGregor>
. . . no, Firebug is broken.  SIGH.

[13:03:02.0000] <timeless>
try venkman?

[13:04:00.0000] <AryehGregor>
I'm currently looking for a JavaScript profiler.

[13:04:01.0000] <AryehGregor>
Because WebKit's is useless.

[13:04:02.0000] <timeless>
venkman has one..

[13:04:03.0000] <AryehGregor>
/me looks

[13:04:04.0000] <timeless>
don't ask me how to install it

[13:05:00.0000] <timeless>
https://addons.mozilla.org/en-US/firefox/addon/javascript-debugger/

[13:05:01.0000] <timeless>
in theory

[13:06:00.0000] <AryehGregor>
Seems to at least pretend to work.

[13:09:00.0000] <AryehGregor>
Okay, well Venkman apparently wants to save profiling data to a file instead of actually showing it to me.

[13:09:01.0000] <timeless>
save it as html and load it in your favorite web browser

[13:10:00.0000] <timeless>
or save it as csv and then use your favorite spreadsheet program to render your data :)

[13:12:00.0000] <AryehGregor>
Sigh.

[13:12:01.0000] <timeless>
was the data useful?

[13:12:02.0000] <AryehGregor>
It didn't seem worth the hassle.

[13:12:03.0000] <AryehGregor>
I could try again, since I can't get Firebug to work at all.

[13:13:00.0000] <timeless>
saving a file to disk is hard?

[13:13:01.0000] <timeless>
do you not have a working storage medium?

[13:14:00.0000] <AryehGregor>
I guess I'm spoiled, but I'm used to programs not asking me to take gratuitous extra steps like that these days.

[13:14:01.0000] <timeless>
iirc IE has a profiler

[13:15:00.0000] <AryehGregor>
/me looks

[13:15:01.0000] <timeless>
yeah, ie10 does

[13:15:02.0000] <timeless>
press f12

[13:15:03.0000] <timeless>
debug>f12 developer tools

[13:15:04.0000] <Philip`>
Do many JS profilers work without forcing JIT off and making the measurements useless?

[13:15:05.0000] <timeless>
profiler is one of the tabs

[13:15:06.0000] <timeless>
Philip`: dunno

[13:15:07.0000] <timeless>
it's technically possible to do useful performance counting w/ a jit

[13:15:08.0000] <timeless>
e.g. you could go w/ the dtrace model

[13:16:00.0000] <timeless>
which gecko supports

[13:16:01.0000] <timeless>
but i don't know if anyone was nice enough to add in dtrace support to jit generated code

[13:16:02.0000] <timeless>
(probably not)

[13:18:00.0000] <AryehGregor>
Philip`, the measurements wouldn't be useless if you're just looking for stuff like "Is there some function that's taking 40% of my runtime that I could trivially rewrite to cache results?" or such.

[13:18:01.0000] <AryehGregor>
Now IE10PP2 crashes when I go into developer tools.

[13:18:02.0000] <AryehGregor>
The world conspires against me.

[13:18:03.0000] <timeless>
lol

[13:19:00.0000] <timeless>
iirc ie10pp2 crashed on news.google.com [us edition] when i hovered over the picture preview things at the top of news

[13:19:01.0000] <timeless>
i think they were youtube content (fast flip was fine, and so is youtube.com)

[13:19:02.0000] <timeless>
anyway..

[13:19:03.0000] <Philip`>
Native hardware acceleration means it can crash faster than any other browser

[13:20:00.0000] <AryehGregor>
Okay, the Venkman data looks useful.

[13:20:01.0000] <AryehGregor>
Thanks for the tip, timeless.

[13:20:02.0000] <timeless>
:)

[13:23:00.0000] <AryehGregor>
The formatting is terrible, though.

[13:24:00.0000] <AryehGregor>
But it'll do.

[13:34:00.0000] <zewt->
AryehGregor: if it disables JIT and that causes a function to go from 1% to 40% of runtime, that's pretty useless

[13:34:01.0000] <AryehGregor>
zewt, sure.

[13:37:00.0000] <timeless>
AryehGregor: you can use the csv output and format it however you like :)

[13:37:01.0000] <AryehGregor>
timeless, that's what I wound up doing, but having it in the UI would be more handy.

[13:37:02.0000] <timeless>
i think the author assumed someone else would come in and give it some ui love

[13:37:03.0000] <timeless>
and no one did

[13:38:00.0000] <gsnedders>
I believe a lot of debuggers disable most optimization, at least

[13:38:01.0000] <timeless>
afaik, gecko today mostly disables everything when a debugger starts doing stuff

[13:38:02.0000] <timeless>
it isn't strictly required, but yeah, that's generally the first impl since it's easiest

[13:38:03.0000] <gsnedders>
Dragonfly entirely disables JIT (though we only have a type-specializing compiler, so to disable optimization we have to fall back to just the interpreter)

[13:39:00.0000] <timeless>
things get rather annoying when you have a jit that does inlining

[13:40:00.0000] <timeless>
since the person looking at the profiler wants to know which function they wrote is taking time

[13:40:01.0000] <timeless>
but w/o single stepping the optimized code and recognizing which function it came from

[13:40:02.0000] <timeless>
or worse, if two functions get optimized to be the same code and then are shared..

[13:40:03.0000] <zewt>
i'd rather know which optimized function, including inlined calls, is taking the time

[13:41:00.0000] <timeless>
in WPO, potentially that's "your-web-page"

[13:41:01.0000] <zewt>
generally it's fairly obvious, when a profiler is saying a function is taking a lot of time and all it does is call other functions, that it's actually the inlined functions taking the time

[13:41:02.0000] <timeless>
i'd imagine that wouldn't be particularly useful :)

[13:41:03.0000] <AryehGregor>
Okay, now the IE10PP2 profiler works.

[13:41:04.0000] <AryehGregor>
It clearly can't be disabling JIT, since it doesn't take much longer to run.

[13:41:05.0000] <timeless>
zewt: true

[13:41:06.0000] <AryehGregor>
And it produces useful results.

[13:41:07.0000] <AryehGregor>
WINNER: INTERNET EXPLORER.

[13:41:08.0000] <zewt>
... obvious to someone experienced and who understands things like inlining, of course, but novices probably shouldn't be trying to use a profiler anyway

[13:41:09.0000] <timeless>
AryehGregor: yeah, the ms tools teams generally produce the best results..

[13:42:00.0000] <timeless>
my guess is it's using performance counters

[13:42:01.0000] <AryehGregor>
That would be pretty logical.

[13:42:02.0000] <timeless>
but, if they're actually generating CLR stuff, then the hooks available are pretty extensive

[13:42:03.0000] <timeless>
i don't know what the output of their JIT is..

[13:42:04.0000] <AryehGregor>
WebKit's is garbage, AFAICT, its results make no sense.  But at least it's easy to use and looks pretty.

[13:43:00.0000] <timeless>
lol

[13:43:01.0000] <gsnedders>
zewt: It's a lot easier to do profiling if you can accept inlined functions being misreported. Debugging gets fun when you have DCE and loop invariant code motion, for example.

[13:43:02.0000] <gsnedders>
timeless: They aren't doing CLR, it's all custom.

[13:43:03.0000] <AryehGregor>
Interestingly, the IE10 profiler tells me I'm spending the most time in parentNode.  2.5s.

[13:43:04.0000] <zewt>
sure, i've spent way too much time in gdb stepping through optimized C++ code, i know how much fun that is, heh

[13:43:05.0000] <gsnedders>
timeless: CLR people were involved in a consulting way, from what their experience was

[13:43:06.0000] <timeless>
gsnedders: good to know

[13:44:00.0000] <AryehGregor>
Wow, it gives me this great tree and everything.

[13:44:01.0000] <Ms2ger>
That man knows way too much

[13:44:02.0000] <AryehGregor>
/me <3 IE10

[13:44:03.0000] <gsnedders>
Ms2ger: me?

[13:44:04.0000] <jamesr>
AryehGregor: profiling JS? did you try chrome?

[13:44:05.0000] <gsnedders>
Ms2ger: Sorry, I've spent too much time around JS engines. :P

[13:44:06.0000] <Ms2ger>
I gathered that :)

[13:45:00.0000] <timeless>
AryehGregor: so i get points for 2 recommendations :)

[13:45:01.0000] <zewt>
(even more fun is insn stepping through code with no debug symbols, with the source code open in another window, trying to match up the source and assembly while stepping)

[13:45:02.0000] <Ms2ger>
timeless+=2

[13:45:03.0000] <gsnedders>
zewt: I mean, we could go for something confusing to most web developers when implementing it…

[13:45:04.0000] <timeless>
zewt: i've done that a lot

[13:45:05.0000] <timeless>
it's exciting

[13:45:06.0000] <gsnedders>
zewt: code with no debug symbols? Like generated code from a JITing compiler? :)

[13:45:07.0000] <zewt>
timeless: what makes it great is how gdb is so brittle--spend 15 minutes stepping through code to get where you want to be, then crash

[13:45:08.0000] <timeless>
gsnedders: typically taking an optimized release build of an open source product

[13:45:09.0000] <zewt>
= rage

[13:46:00.0000] <timeless>
zewt: yeah um...

[13:46:01.0000] <timeless>
i learned that i was *much* happier w/ any debugger other than gdb

[13:46:02.0000] <gsnedders>
timeless: JITing compilers is where I've dealt with it most

[13:46:03.0000] <gsnedders>
(Which was rather the point)

[13:46:04.0000] <timeless>
windbg/ntsd, adb, dbx, ...

[13:46:05.0000] <gsnedders>
Debugging JS execution bugs often ends up with fun like that. :)

[13:46:06.0000] <timeless>
you name it, if it wasn't gdb or a front end to gdb, i was *much* happier

[13:46:07.0000] <timeless>
since gdb is *incredibly* flaky

[13:47:00.0000] <zewt>
gdb is okay until you suspend and it decides it's just going to keep suspending forever

[13:47:01.0000] <timeless>
/me remembers that

[13:47:02.0000] <timeless>
its interrupt (ctrl-z/ctrl-c) handling is "interesting"

[13:47:03.0000] <timeless>
it also whines and has rather amusing failure handling

[13:52:00.0000] <timeless>
/me returns to 4.7. Platform objects implementing interfaces §

[13:54:00.0000] <AryehGregor>
Hixie, I don't see birthdate anywhere in Google+.  Is it actually there?

[13:55:00.0000] <timeless>
AryehGregor: didn't you have this discussion yesterday?

[13:55:01.0000] <AryehGregor>
timeless, yes, but he responded after I left.

[13:55:02.0000] <AryehGregor>
Saying that it was there.

[13:55:03.0000] <timeless>
/me wonders if it's only present to the extent that it's required for COPA

[13:57:00.0000] <AryehGregor>
/me gives up on profiling, resolves to just suffer

[13:57:01.0000] <AryehGregor>
You'd think they'd still let you display it.

[13:57:02.0000] <timeless>
why?

[13:58:00.0000] <timeless>
who is they, and what are they thwarting?

[13:58:01.0000] <AryehGregor>
Google+.  You'd think they'd let you because it's an obvious thing to want to display.

[14:07:00.0000] <erlehmann>
google+ is a lie.

[14:08:00.0000] <erlehmann>
it sends 404 as a catch-all error message.

[14:08:01.0000] <erlehmann>
great if you get a link from your friends and want to see if it is broken.

[14:08:02.0000] <erlehmann>
s/want/claim/ s/if/that/

[14:08:03.0000] <erlehmann>
sleepytiem, it seems

[14:08:04.0000] <erlehmann>
also, no feeds.


2011-07-16
[03:11:00.0000] <krijnh>
Si le ping

[04:21:00.0000] <Dashiva>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13276 confuses me

[04:47:00.0000] <krijn>
http://www.whatwg.org/specs/web-apps/current-work/multipage/the-button-element.html#the-progress-element - shouldn't that note say "To make a determinate progress" in stead of "To make an determinate progress" ?

[04:56:00.0000] <Philip`>
krijn: Yes

[04:56:01.0000] <krijn>
OMG, HTML5 sucks man!

[04:57:00.0000] <Philip`>
/me thinks "determinate" is a pretty obscure word to use, though

[04:59:00.0000] <annevk>
krijnhoetmer.nl too!

[04:59:01.0000] <krijn>
I know :/

[05:00:00.0000] <krijn>
My server is now located next to my router

[05:00:01.0000] <krijn>
And it still drops connections

[05:04:00.0000] <Philip`>
Maybe you should put the server on top of a stack of books or something, so that the data can flow downhill

[05:05:00.0000] <MikeSmith>
heh

[05:05:01.0000] <krijn>
It actually was 2 floors higher before today :)

[05:05:02.0000] <krijnh>
Yeah!

[05:05:03.0000] <krijn>
Oh shaddap

[05:08:00.0000] <MikeSmith>
we need to set up some kind of annoy mechanism that auto-harrasses krijn any time his site goes down

[05:10:00.0000] <Philip`>
Isn't that what IRC is for?

[05:10:01.0000] <MikeSmith>
heh

[05:10:02.0000] <MikeSmith>
yeah

[05:11:00.0000] <MikeSmith>
we could set up a bot to keep track of whose turn it is to annoy krijn

[05:12:00.0000] <MikeSmith>
anyway, hey, fwiw, I finally managed to finish setting up that common testing-framework Web UI thingey what Francois had started on a while back

[05:12:01.0000] <MikeSmith>
voila:

[05:12:02.0000] <MikeSmith>
http://w3c-test.org/framework/

[05:12:03.0000] <MikeSmith>
you can do absolutely nothing with it yet!

[05:12:04.0000] <MikeSmith>
but hey, it's there

[05:13:00.0000] <MikeSmith>
and before this week I was blessed to not have had to do anything php setup/configuration for a long time

[05:14:00.0000] <MikeSmith>
before this week

[05:15:00.0000] <MikeSmith>
but this week I had to deal with php, which cost me I don't know how many hours

[05:15:01.0000] <MikeSmith>
http://w3c-test.org/framework/ is a php app

[05:16:00.0000] <MikeSmith>
and it was all set up in line with install and config docs that Peter Linss had written up, but it nonetheless wasn't working at all

[05:17:00.0000] <MikeSmith>
and it wasn't logging any errors nor throwing any exceptions nor giving any other indication of what was wrong

[05:17:01.0000] <MikeSmith>
and I still don't know what was wrong

[05:17:02.0000] <MikeSmith>
the only way I fixed it was by upgrading the php version on that host


2011-07-17
[06:38:00.0000] <smaug____>
annevk5: ping

[06:38:01.0000] <smaug____>
annevk5: what kind of license is used for Opera's tests?

[07:10:00.0000] <benschwarz>
Hixie!

[10:10:00.0000] <gsnedders>
Object.create([some cross-origin window object]) should do what? Browsers seem split on this…

[10:10:01.0000] <gsnedders>
(i.e., "opinions?")

[14:29:00.0000] <annevk>
hober, basically with those editorial issues like 150 I would very much prefer someone to set a spec-wide style guide rather than micromanaging individual editors

[14:30:00.0000] <annevk>
micromanaging individual editors is a huge waste of time

[14:38:00.0000] <annevk>
smaug, depends on the tests

[14:39:00.0000] <annevk>
smaug, I let Mozilla use mine under whichever license they want I think per some statement in a bug

[14:39:01.0000] <annevk>
I think some others might be licensed under Apache 2.0 or some such

[14:39:02.0000] <annevk>
not really up to speed with those

[14:40:00.0000] <gsnedders>
annevk: For the HTML WG testsuite ones?

[14:40:01.0000] <annevk>
well those are licensed under whatever the W3C wants I reckon

[14:41:00.0000] <annevk>
I meant http://testsuites.opera.com/

[14:41:01.0000] <gsnedders>
annevk: I mean what tests was smaug asking about

[14:41:02.0000] <gsnedders>
"Unless otherwise noted, all material on testsuites.opera.com is shared under the Apache License, Version 2.0."

[14:41:03.0000] <annevk>
I doubt he would be asking about HTML WG tests

[14:41:04.0000] <annevk>
prolly XHR or EventSource

[15:01:00.0000] <annevk>
smaug____, see logs

[15:02:00.0000] <annevk>
did anyone see any kind of conclusion in the Exceptions thread?

[15:02:01.0000] <annevk>
I still have no idea whether to change DOM Core

[15:02:02.0000] <annevk>
and how

[15:02:03.0000] <annevk>
Hixie, are you around btw?

[15:03:00.0000] <annevk>
Hixie, I was wondering whether you have some kind of list of things that are wrong with DOM Traversal

[15:03:01.0000] <heycam>
annevk, didn't seem to have a conclusion yet, although I think most people seem convinced the "instanceof Blah" checking is bad

[15:07:00.0000] <smaug____>
annevk: so it is ok to take your tests and modify them to work with mochitests ?

[15:08:00.0000] <annevk>
heycam, yeah that's what I got out of it too

[15:08:01.0000] <smaug____>
do I need to add some license text and mention copyright Opera ?

[15:08:02.0000] <annevk>
smaug____, yeah, though preferably that is done via the separate javascript file

[15:09:00.0000] <annevk>
smaug____, i mean the idea is that we use this framework for all tests

[15:09:01.0000] <annevk>
smaug____, no that's not needed

[15:11:00.0000] <smaug____>
annevk: thanks

[15:17:00.0000] <Hixie>
annevk: not specially. it just needs rewriting to actually be defined.

[15:20:00.0000] <annevk>
k

[15:25:00.0000] <heycam>
anyone know if a good way to feature detect svg as a css background image?

[15:25:01.0000] <heycam>
s/if/og/

[15:25:02.0000] <heycam>
/me sighs

[15:28:00.0000] <annevk>
just assume it's supported?

[15:28:01.0000] <heycam>
:)

[15:28:02.0000] <heycam>
maybe I should just do that


2011-07-18
[17:12:00.0000] <benschwarz>
Hixie: !ping

[19:34:00.0000] <rektide_>
are cross site <script type="text/javascript" src="..."></script>'s no longer allowed?

[19:34:01.0000] <rektide_>
i'm not seeing my dev channel chrome attempting the request

[19:35:00.0000] <rektide_>
i thought jsonp was a still used workaround? is it still possible? if so, how? or is there something amiss with my chrome not attempting the above <script> request

[19:37:00.0000] <rektide_>
sorry error elsewhere in my html; my mistake

[03:03:00.0000] <annevk>
The RTC-Web effort (Real Time Collaboration on the World Wide Web) ... http://rtc-web.alvestrand.com/

[03:03:01.0000] <annevk>
The Real Time Communication on the Web (RTC-Web) ... http://rtc-web.alvestrand.com/home

[03:04:00.0000] <annevk>
what is it?

[03:21:00.0000] <hsivonen>
annevk: "What is it?" in what sense?

[03:24:00.0000] <annevk>
Communication or Collaboration

[03:24:01.0000] <hsivonen>
annevk: oh

[03:24:02.0000] <jgraham>
Hmm, I wonder if smaug wrote a testharness.js-to-mochitest adapter. Someone has to do that eventually

[03:25:00.0000] <hsivonen>
annevk: the sites fail to provide a quick elevator pitch like "Chatroulette without Flash"

[03:25:01.0000] <hsivonen>
(Or G+ hangouts without Flash, to give a less creepy usecase)

[03:27:00.0000] <espadrine`>
annevk: I guess it is more about communication, collaboration seems like a particular use case rather than the whole thing

[03:27:01.0000] <hsivonen>
BTW, what's the business case for Adobe for having videoconferencing enablement in Flash Player? Is it only about making Flash Player more sticky so that they can leverage it for other business or does Adobe monetize videoconferencing directly?

[03:28:00.0000] <hsivonen>
espadrine`: seems to me to be more about realtime audio and videa than about either communication or collaboration on a general level

[03:28:01.0000] <Ms2ger>
jgraham, I did :)

[03:29:00.0000] <hsivonen>
Ms2ger++

[03:29:01.0000] <Ms2ger>
Now I just need someone to review it

[03:29:02.0000] <espadrine`>
hsivonen: oh well.

[03:29:03.0000] <jgraham>
Ms2ger: You rock

[03:29:04.0000] <Ms2ger>
Thanks :)

[03:30:00.0000] <hsivonen>
annevk: maybe the C is like X in XBL :-)

[03:30:01.0000] <espadrine`>
"realtime audio/video" is not enterprisy enough

[03:30:02.0000] <Ms2ger>
/me wonders if he could land it with r=jgraham...

[03:31:00.0000] <espadrine`>
/me agrees with jgraham

[03:31:01.0000] <annevk>
Ms2ger, you should probably tell smaug

[03:31:02.0000] <annevk>
Ms2ger, before he rewrites all those tests

[03:31:03.0000] <Ms2ger>
CCd him

[03:35:00.0000] <annevk>
Ms2ger, btw

[03:36:00.0000] <annevk>
Ms2ger, <ol type=TEST> ol[type=test] { background:lime } will style the <ol>

[03:36:01.0000] <annevk>
Ms2ger, in every browser

[03:36:02.0000] <Ms2ger>
Yeah

[03:36:03.0000] <annevk>
which is wrong per HTML at the moment

[03:36:04.0000] <Ms2ger>
Is it really?

[03:37:00.0000] <annevk>
"The exception to the list above is the type attribute on ol elements, which must be treated as case-sensitive."

[03:37:01.0000] <Ms2ger>
Where's that?

[03:38:00.0000] <annevk>
first subsection of "Matching HTML elements using selectors"

[03:39:00.0000] <Ms2ger>
Huh

[03:55:00.0000] <hsivonen>
espadrine`: what kind of enterpriseyness do you perceive there to be about WebRTC and what enterpriseyness do you think it should have?

[04:20:00.0000] <annevk>
it's still window.URL right?

[04:20:01.0000] <annevk>
http://html5-mediasource-api.googlecode.com/svn/trunk/draft-spec/mediasource-draft-spec.html has some confusing URI.createObjectURL() in it...

[04:22:00.0000] <Ms2ger>
Yes

[04:25:00.0000] <annevk>
I wonder if I should bother filing a bug on Mozilla about attribute values and selectors

[04:25:01.0000] <annevk>
Or just file a bug on HTML5 to remove the exception for <ol type>

[04:26:00.0000] <annevk>
Guess I'm going to do the former. The behavior is magical and ugly...

[04:33:00.0000] <annevk>
http://blog.whatwg.org/weekly-rtc-download

[04:51:00.0000] <AnselmBradford>
For the area element, what is meant by a "dead area on an image map" (in reference to one form the element can take)?

[04:51:01.0000] <Ms2ger>
One that doesn't link anywhere?

[04:52:00.0000] <Ms2ger>
Like <area coords=0,0,100,100 href=foo><area coords=25,25,75,75>

[04:54:00.0000] <AnselmBradford>
Ms2ger: gotcha thanks... so corresponds to this text in the spec "If the area element has no href attribute, then the area represented by the element cannot be selected, and the alt attribute must be omitted."

[04:54:01.0000] <Ms2ger>
Sounds right

[04:55:00.0000] <AnselmBradford>
What's a use case for an area without an href?

[04:58:00.0000] <Ms2ger>
Making parts of an image unclickable, I suppose... This was long before my time

[04:59:00.0000] <annevk>
AnselmBradford, you can add a title attribute for instance

[05:00:00.0000] <AnselmBradford>
annevk: ah, I see... so on hover the title shows, but it wouldn't be a hyperlink to anywhere?

[05:00:01.0000] <annevk>
not sure how <area> is layered but I can imagine you have a large clickable layer and then some parts of that should not be clickable maybe

[05:00:02.0000] <annevk>
AnselmBradford, right

[05:02:00.0000] <smaug____>
annevk: would it be possible to have some text in your test that they are Public Domain (or whatever license they use)

[05:08:00.0000] <annevk>
I rather not put that stuff in each test

[05:09:00.0000] <annevk>
last time I contributed tests to Mozilla I just said so in a bug report

[05:11:00.0000] <jgraham>
/me wonders why these tests can't go via W3C

[05:11:01.0000] <jgraham>
Which should solve Mozilla's licensing issue and make them avaliable for everyone

[05:12:00.0000] <annevk>
I think I was waiting for comments before putting them in even more repositories

[05:12:01.0000] <annevk>
Potentially also for the W3C to agree on testharness.js or something else

[05:14:00.0000] <jgraham>
Seems like it might be time to move them then :)

[05:16:00.0000] <annevk>
Still waiting for comments

[05:18:00.0000] <hsivonen>
the enclosure thread shows why you need to bikeshed early and forcefully

[05:18:01.0000] <hsivonen>
who'd have thought that Dave Winer's pick for an RSS element could threaten to leak into HTML rel values one day

[05:23:00.0000] <jgraham>
annevk: My comment is you should move them to some W3C venue before I comment further :)

[05:23:01.0000] <smaug____>
heycam|away: what should happen if some method takes an dictionary, but the object which implements the dictionary has wrong types. Are those properties automatically converted to right type or something?

[05:24:00.0000] <Ms2ger>
smaug____, yes, afaict

[05:25:00.0000] <Ms2ger>
But maybe not if that method is overloaded

[05:28:00.0000] <annevk>
jgraham, not falling for that one :p

[05:52:00.0000] <annevk>
Extra arguments added to methods are ignored right?

[05:53:00.0000] <annevk>
looks like it

[05:53:01.0000] <annevk>
am I glad entityReferenceExpansion is the last argument

[05:54:00.0000] <annevk>
we can just remove it without hassle

[06:24:00.0000] <hsivonen>
I wonder what would happen if an Editorial Assistant WONTFIXed a bug Hixie filed.

[06:24:01.0000] <annevk>
try it!

[06:25:00.0000] <annevk>
although maybe wait for a bit, so I can have some popcorn with that

[06:27:00.0000] <hsivonen>
Another thing I've been wondering about today: Should I write email to Bob Leif expressing that using XSD 1.1 isn't a use case.

[06:29:00.0000] <annevk>
I am hoping the chair will handle it

[06:29:01.0000] <hsivonen>
yeah, maybe I should wait for Norm to deal

[06:42:00.0000] <annevk>
I wonder why the DOM specifications are so verbose

[06:42:01.0000] <annevk>
Well, the pre-DOM Core stuff

[06:43:00.0000] <annevk>
Looking at DOM Traversal it is mostly complicated language for a few simple concepts

[06:48:00.0000] <hsivonen>
annevk: I thought the old DOM specs weren't verbose enough

[06:48:01.0000] <hsivonen>
annevk: maybe they spend words on the wrong things, though

[06:49:00.0000] <annevk>
well yeah, they are not exactly clear

[06:49:01.0000] <annevk>
they explain models by examples rather than well, defining them

[06:59:00.0000] <annevk>
smaug____, why can't withCredentials be a simple attribute on EventSource just as it is with XHR?

[07:00:00.0000] <annevk>
if we need withCredentials at all

[07:14:00.0000] <smaug____>
annevk: EvenSource opens the connections when ctor is called

[07:16:00.0000] <annevk>
no

[07:16:01.0000] <annevk>
it queues a task to do that

[07:17:00.0000] <smaug____>
ah, that is true

[07:17:01.0000] <smaug____>
/me hates the API inconsistency in XHR vs EventSource

[07:17:02.0000] <smaug____>
so, yeah, perhaps withCredentials attribute could work

[07:18:00.0000] <smaug____>
that would give some consistency

[07:20:00.0000] <smaug____>
annevk: I haven't seen any good reason to not have withCredentials

[07:20:01.0000] <annevk>
still think it wasn't the best idea to add it to XHR

[07:22:00.0000] <smaug____>
but to do what?

[07:22:01.0000] <smaug____>
never send credentials?

[07:26:00.0000] <annevk>
always send credentials

[07:28:00.0000] <smaug____>
why

[07:29:00.0000] <annevk>
because that is how cross-origin requests worked already

[09:23:00.0000] <annevk>
Hmm, Traversal depends on mutation quite a bit

[09:49:00.0000] <TabAtkins>
annevk: Hm, I guess that bit about <ol=type> needs to be reworded, then, so that it specifically references the known type values but lets everything else be case-insensitive.

[09:59:00.0000] <smaug____>
Interesting. Per spec calling command.click() ends up firing click event twice

[10:02:00.0000] <annevk>
TabAtkins, that would be even worse

[10:02:01.0000] <annevk>
TabAtkins, also does not match impls afaik

[10:05:00.0000] <TabAtkins>
annevk: Man, really?  so <ol type="A"> is matched by ol[type='a']?

[10:09:00.0000] <annevk>
TabAtkins, always has been

[10:09:01.0000] <annevk>
TabAtkins, the change I am hoping we can make is to treat attribute values as case-sensitive, always

[10:09:02.0000] <annevk>
TabAtkins, I think that would be better than having a magic list of attributes

[10:11:00.0000] <TabAtkins>
annevk: I'm wondering how impls actually apply list styling, then.  I assumed it was done via ua stylesheet.

[10:11:01.0000] <annevk>
UA style sheet has to be special anyway because of XHTML

[10:11:02.0000] <annevk>
e.g. input[type=button] would not work well in XHTML

[10:15:00.0000] <annevk>
smaug____, prolly a bug

[10:48:00.0000] <annevk>
TabAtkins, and it seems browsers have bugs because of their UA style sheet

[10:48:01.0000] <annevk>
e.g. compare data:text/xml,<html xmlns="http://www.w3.org/1999/xhtml"><input type="BUTTON"/><input type="button"/></html> between Gecko / WebKit / Opera

[10:50:00.0000] <TabAtkins>
Yay!

[11:09:00.0000] <rillian>
hey all

[11:09:01.0000] <rillian>
I'm having trouble understanding http://www.w3.org/TR/html5/webappapis.html#provide-a-stable-state

[11:10:00.0000] <rillian>
I understand the idea is to wait for any (asynchronously running?) synchronous sections to complete

[11:10:01.0000] <rillian>
so the attributes available to javascript are coherent

[11:10:02.0000] <rillian>
but isn't that at odds with "A synchronous section never mutates the DOM ... or has any other side-effects."?

[11:11:00.0000] <TabAtkins>
Why do you think it's at odds?

[11:12:00.0000] <rillian>
I suppose because firebug shows global js objects in the DOM tab

[11:12:01.0000] <rillian>
so I don't see how you could update object state without mutating the DOM

[11:18:00.0000] <annevk>
I recommend a) looking at a version of the draft that is not multipage so you can see where the terms are referenced from and b) not using TR/

[11:21:00.0000] <annevk>
rillian, fwiw, mutates the DOM means mutating the Document tree

[11:21:01.0000] <annevk>
rillian, it is not about changing IDL attributes

[11:22:00.0000] <rillian>
so "DOM" there is different from the view in javascript's namespace?

[11:24:00.0000] <annevk>
not sure what you mean by that

[11:24:01.0000] <rillian>
annevk: sorry, by not using 'TR' do you mean http://www.whatwg.org/specs/web-apps/current-work/#provide-a-stable-state ?

[11:24:02.0000] <annevk>
for instance

[11:25:00.0000] <annevk>
TR/ is always out of date

[11:25:01.0000] <rillian>
'k thanks

[11:26:00.0000] <annevk>
DOM mutation is generally understood to mean things like removeChild()

[11:26:01.0000] <rillian>
so things that affect the document object from javascript's point of view?

[11:26:02.0000] <annevk>
things that fire mutation events

[11:27:00.0000] <annevk>
you could ask for that to be made clearer though

[11:27:01.0000] <rillian>
sorry, by "DOM" I pretty much meant "What firebug shows in its DOM tab"

[11:27:02.0000] <rillian>
I'm trying to understand what's spec, what's current implementation, and so on, with the event loop

[11:28:00.0000] <annevk>
as currently phrased you could think setting e.g. <link>.disabled would be such a thing

[11:28:01.0000] <rillian>
*nod*

[11:29:00.0000] <rillian>
I think the spec would also be improved with some less-normative language about what stable-state was for

[11:29:01.0000] <rillian>
right now it feels a bit circular

[12:03:00.0000] <Xeon06_>
How legal is it to create our own attributes on existing elements in HTML5?

[12:03:01.0000] <TabAtkins>
You can use the data-* attributes for whatever you want that's private to the page.  Otherwise, it's invalid.

[12:04:00.0000] <AryehGregor>
Xeon06_, you can make up your own attributes as long as they start with "data-".

[12:04:01.0000] <Xeon06_>
Okay, thanks guys1

[12:04:02.0000] <Xeon06_>
!*

[12:06:00.0000] <Xeon06_>
Wow, this is what jQuery's .data does!

[12:07:00.0000] <TabAtkins>
Yes.

[12:19:00.0000] <Hixie>
hsivonen: if you wontfix a bug i filed i won't notice. but i would be interested in hearing about it if you ever find a bug i filed be something you might consider wontfixing :-)

[12:19:01.0000] <Hixie>
hsivonen: especially since they're generally editorial things :-)

[12:24:00.0000] <janv_>
 Hixie: hi, could you take a look at https://bugzilla.mozilla.org/show_bug.cgi?id=617528#c109 ?

[12:27:00.0000] <janv_>
Hixie: thanks

[12:30:00.0000] <Hixie>
just got their in my bugmail :-)

[12:31:00.0000] <janv_>
:)

[12:31:01.0000] <janv_>
this is the last issue

[12:31:02.0000] <Hixie>
er, there

[12:31:03.0000] <Hixie>
not their

[12:31:04.0000] <janv_>
I see

[12:32:00.0000] <janv_>
I hope we land the patch soon

[14:14:00.0000] <annevk>
Hixie, why does ev = new EV("url"); ev.withCredentials = true; not work?

[14:14:01.0000] <annevk>
new EV; queues a task

[14:15:00.0000] <annevk>
when the task queue is spinned and that task gets done you get the value from withCredentials and all should be fine

[14:48:00.0000] <Hixie>
annevk: workers

[14:49:00.0000] <annevk>
EventSource is defined differently for workers?

[14:50:00.0000] <annevk>
how does onopen get registered on time in workers?

[14:51:00.0000] <Hixie>
workers don't spin the event loop often

[14:51:01.0000] <Hixie>
or, they don't have to spin the event loop often, rather

[14:51:02.0000] <Hixie>
since their scripts can run forever

[14:52:00.0000] <annevk>
either way you need to wait for the next task

[14:52:01.0000] <annevk>
and new ev(); ev.wC is a single operation

[14:52:02.0000] <Hixie>
the EventSource constructor doesn't wait for the event loop to start the connection

[14:57:00.0000] <annevk>
oh I see

[14:57:01.0000] <annevk>
it doesn't queue a task at all

[14:57:02.0000] <annevk>
it does some magic

[14:58:00.0000] <annevk>
should XHR use similar magic?

[14:58:01.0000] <annevk>
hmm

[14:58:02.0000] <Hixie>
not magic, just async

[14:59:00.0000] <annevk>
i thought queue was our async

[14:59:01.0000] <annevk>
not return and do both set of things

[15:00:00.0000] <Hixie>
there's all kinds of different models depending on what the needs are

[15:01:00.0000] <annevk>
i had no idea

[15:01:01.0000] <Hixie>
working on a background thread like this, queuing work on an event loop, using "stable states", hooking in other parts of the event loop, etc

[15:05:00.0000] <annevk>
guess XHR could use this too

[15:05:01.0000] <annevk>
hmm

[15:10:00.0000] <Hixie>
if you use 'fetch' then you already use this

[15:10:01.0000] <Hixie>
since 'fetch' only makes sense on a background thread, not in a task

[15:10:02.0000] <Hixie>
(unless you're doign sync xhr, of course)

[15:10:03.0000] <annevk>
i think it queues a task to fetch atm

[15:11:00.0000] <annevk>
bit sad that nobody actually reviews the spec in detail

[15:11:01.0000] <annevk>
or maybe next to nobody understands those parts

[15:11:02.0000] <jamesr>
the implications can be very subtle

[15:11:03.0000] <annevk>
yeah, way hard

[15:12:00.0000] <annevk>
anyway, got to go

[15:13:00.0000] <rillian>
at least I'm not the only one confused on these points :/

[16:08:00.0000] <AryehGregor>
Hey, someone else actually was using the Live DOM Viewer upload/download feature at the same time as me.  That's a first.

[16:11:00.0000] <TabAtkins>
Did you collide?

[16:11:01.0000] <AryehGregor>
Yes, that's how I knew.

[16:17:00.0000] <AryehGregor>
Can anyone explain to me why the www-html list still exists?

[16:18:00.0000] <TabAtkins>
Looks like just the older version of public-html, from before we switched list naming standards.  Dunno why it's still open.

[16:19:00.0000] <Philip`>
I think the switch to public-html was primarily to ensure subscribers agreed to the patent policy

[16:19:01.0000] <Philip`>
(since those already on www-html hadn't)

[16:20:00.0000] <Hixie>
AryehGregor: i assume you collided with my window.find() thing, sorry :-)

[16:21:00.0000] <Hixie>
also, window.find() is incredibly buggy

[16:21:01.0000] <Hixie>
well, not incredibly

[16:21:02.0000] <Hixie>
but it has six boolean arguments and of those, only one works the same in gecko and webkit

[16:23:00.0000] <jamesr>
i only use the upload/download feature when i am trying to get a permalink and forget which link it is to do that

[16:23:01.0000] <jamesr>
which is pretty much every time i try to get a permalink

[16:23:02.0000] <Philip`>
Isn't it the link that says "permalink"?

[16:24:00.0000] <Hixie>
i think he means the short permalink

[16:24:01.0000] <Hixie>
which is "save"

[16:24:02.0000] <jamesr>
right, that one

[16:24:03.0000] <Philip`>
Ah

[16:24:04.0000] <AryehGregor>
I use it constantly when testing among multiple browsers.

[16:24:05.0000] <Hixie>
yeah me too

[16:24:06.0000] <AryehGregor>
Particularly since one of them is on a different computer.

[16:24:07.0000] <Hixie>
indeed

[16:25:00.0000] <AryehGregor>
Although it seems like in IE9 and IE10, "download" only works once until you restart the browser.

[16:25:01.0000] <AryehGregor>
At least on my machine.

[16:25:02.0000] <Hixie>
actually my first implementation of the upload/download thing was for a clipboard on my portal page

[16:25:03.0000] <Hixie>
IE's wacked

[16:25:04.0000] <AryehGregor>
If you hit it a second time, it just re-downloads the same thing it did last time.  No idea why.

[16:25:05.0000] <AryehGregor>
Also, the permalinks don't work in IE or Chrome because they're XSS.  :(

[16:25:06.0000] <Hixie>
it wants different cache flags, i expect

[16:25:07.0000] <Hixie>
the whole tool is a giant XSS :-P

[16:26:00.0000] <jamesr>
Hixie: send the XSS header then!

[16:26:01.0000] <Hixie>
the XSS header is lame

[16:26:02.0000] <Hixie>
and unspecified

[16:27:00.0000] <Hixie>
i may one day switch to base64-encoding the payload

[16:27:01.0000] <Hixie>
but i'm in no rush

[16:28:00.0000] <Hixie>
so for window.find()

[16:28:01.0000] <Hixie>
should i only specify what's interoperable, or should i specify the superset and just call every implementation buggy?

[16:28:02.0000] <AryehGregor>
Specify the subset that's either interoperable or useful.

[16:28:03.0000] <Hixie>
not clear any of it is useful

[16:29:00.0000] <AryehGregor>
Then just the part that's interoperable, I'd say.

[16:29:01.0000] <Hixie>
and no subset it is 100% interoperable

[16:29:02.0000] <Hixie>
of it

[16:29:03.0000] <TabAtkins>
Your work is done, then.

[16:29:04.0000] <AryehGregor>
Well, then make something up.  :)

[16:29:05.0000] <AryehGregor>
I mean, no subset of execCommand() is 100% interoperable, that's for sure . . .

[16:29:06.0000] <TabAtkins>
Also, this is a horrible function and I hate everyone involved in the creation of it.

[16:30:00.0000] <AryehGregor>
Wow, it only returns a boolean?  I'd have thought it would at least return the location of the first occurrence.

[16:31:00.0000] <AryehGregor>
That seems profoundly useless.

[16:31:01.0000] <TabAtkins>
Yeah, it seems completely useless.

[16:31:02.0000] <Hixie>
it selects the match

[16:31:03.0000] <AryehGregor>
Oh, I see.

[16:31:04.0000] <AryehGregor>
That's . . . also amazingly useless.

[16:31:05.0000] <TabAtkins>
Ah, MDN doesn't mention that.

[16:32:00.0000] <TabAtkins>
Can you query for what parts of the document are selected?

[16:32:01.0000] <AryehGregor>
getSelection()?

[16:32:02.0000] <TabAtkins>
kk, wasn't sure.

[16:32:03.0000] <jamesr>
is it used?

[16:32:04.0000] <TabAtkins>
So I guess that'll return a disconnected selection if there are multiple matches.

[16:32:05.0000] <AryehGregor>
If you couldn't, I wouldn't have had much to do for the last six months.  :)

[16:32:06.0000] <AryehGregor>
I imagine it would just select the first match.

[16:33:00.0000] <AryehGregor>
Or the first one after the current selection.

[16:33:01.0000] <AryehGregor>
Like Ctrl-F.

[16:33:02.0000] <AryehGregor>
Although that doesn't actually select anything.

[16:33:03.0000] <TabAtkins>
Well, my ctrl+f selects all the matches, and then scrolls to the first.

[16:33:04.0000] <AryehGregor>
It doesn't select them, it highlights them, no?

[16:33:05.0000] <Hixie>
window.find() selects the first match after the cursor position

[16:35:00.0000] <AryehGregor>
Apparently someone named Tim Powell used window.find() for bookmarklets as of 2001: https://bugzilla.mozilla.org/show_bug.cgi?id=9550#c14

[16:36:00.0000] <Hixie>
the only reason i'm even looking at this is i got some feedback from oepra saying they're getting feedback from developers asking for it

[16:37:00.0000] <Hixie>
in Gecko, window.find() pops up a dialog

[16:37:01.0000] <Hixie>
but Ctrl+F pops up an inline find bar

[16:38:00.0000] <AryehGregor>
This all sounds amazingly horrible even by web standards standards.

[16:38:01.0000] <Hixie>
these two find UIs in gecko have different features...

[16:38:02.0000] <Hixie>
(neither is a subset of the other)

[16:39:00.0000] <Hixie>
wow

[16:40:00.0000] <Hixie>
i got gecko's to fail with NS_ERROR_INVALID_POINTER

[16:41:00.0000] <AryehGregor>
This sounds like a feature everyone has forgotten about for ten years.

[16:41:01.0000] <smaug____>
window.find() is find, ctrl+f is typeaheadfind

[16:41:02.0000] <smaug____>
who uses window.find() anymore?

[16:41:03.0000] <Hixie>
i don't understand why when i make it pop up the dialog, it comes prepopulated with "cat" as teh search string

[16:41:04.0000] <Hixie>
smaug____: can we drop support for it? that would make my life way easier.

[16:41:05.0000] <TabAtkins>
Presuambly that's the last thing you searched for, Hixie?

[16:41:06.0000] <Hixie>
TabAtkins: no

[16:42:00.0000] <smaug____>
Hixie: dropping window.find(). Yes please.

[16:42:01.0000] <TabAtkins>
It seemed likely.

[16:42:02.0000] <Hixie>
TabAtkins: i searched for any number of other things, but every time the dialog comes back, it says "cat"

[16:42:03.0000] <Hixie>
smaug____: any idea who makes the call on this?

[16:42:04.0000] <smaug____>
I just wonder if it is possible to drop support for that

[16:43:00.0000] <TabAtkins>
Hixie: Mine doesn't do that, so shrug.

[16:43:01.0000] <Hixie>
TabAtkins: oh i'm sure it's something i did

[16:44:00.0000] <Hixie>
wow my gecko is seriously messed up now

[16:44:01.0000] <Hixie>
every time i call it with a non-empty first argument, i get this pointer exception

[16:44:02.0000] <jamesr>
sounds like you need to restart

[16:44:03.0000] <Hixie>
i did!

[16:45:00.0000] <Hixie>
oh, i know what's wrong

[16:45:01.0000] <Hixie>
i'm calling it on a frame i've made display:none

[16:47:00.0000] <Hixie>
webkit has another crazy bug, if yo call it before onload it just returns false regardless

[16:48:00.0000] <smaug____>
so you can't search on long loading page?

[16:48:01.0000] <Hixie>
not in webkit

[16:48:02.0000] <Hixie>
you can in gecko

[16:48:03.0000] <Hixie>
(dunno exactly when it starts being possible in webkit)

[16:49:00.0000] <Hixie>
(it might work progressively, just not immediately)

[16:53:00.0000] <jarek>
Hi

[16:53:01.0000] <jarek>
is there anything wrong in using custom (not defined in specification) tags in HTML5 documents?

[16:53:02.0000] <jarek>
I mean I know they don't validate

[16:54:00.0000] <jarek>
but besides that, are there any other practical drawbacks?

[16:54:01.0000] <Hixie>
yes

[16:54:02.0000] <Hixie>
if we start using them as elements in future revisions of HTML, your pages will break

[16:54:03.0000] <jarek>
I see, I can live with that :)

[16:54:04.0000] <Hixie>
also, today nobody will know what your pages mean, since they won't be following the standard

[16:55:00.0000] <Hixie>
also, there's a high likelihood you will accidentally run into weird behaviour, e.g. <image> parses weirdly even though it's not valid, as does <plaintext> and <xmp>, etc.

[16:55:01.0000] <Hixie>
there's probably a better way to do what you want to do

[16:55:02.0000] <jarek>
I was thinking about using custom XML namespaces, but most of the the new stuff (e.g. contentEditable) does not work with them

[16:57:00.0000] <jarek>
HTML5 is still mostly a document format, I wish there were more XUL-like tags

[16:57:01.0000] <Hixie>
like what?

[16:57:02.0000] <Hixie>
smaug____: filed https://bugzilla.mozilla.org/show_bug.cgi?id=672395

[16:58:00.0000] <jarek>
like <box>, there is not sematics in HTML5 for describing many widgets

[16:58:01.0000] <jarek>
s/not/no

[16:58:02.0000] <Hixie>
<box> is presentational. use CSS for that.

[16:58:03.0000] <Hixie>
what widgets are you missing?

[16:59:00.0000] <jarek>
dialogs, menubars, toolbars, tabs...

[16:59:01.0000] <jarek>
I mean the usual stuff that is used in desktop apps

[16:59:02.0000] <Hixie>
we're looking at dialogs -- if you have any feedback on what you need there, please e-mail the list

[16:59:03.0000] <Hixie>
menubars and toolbars we have - <menu>

[16:59:04.0000] <Hixie>
tabs is a presentational thing - just a way to display multiple sections


2011-07-19
[17:00:00.0000] <Hixie>
i have to run

[17:00:01.0000] <Hixie>
bbiab

[17:02:00.0000] <benschwarz>
Damn. I missed Hixie again

[17:27:00.0000] <realityking>
Anyone know a good IRC channel to discuss WAI-ARIA questions?

[17:27:01.0000] <realityking>
Or would this one be good?

[17:49:00.0000] <TabAtkins>
cssyThis might be good.

[17:49:01.0000] <TabAtkins>
realityking: ^^^ s/cssy//

[17:50:00.0000] <realityking>
thanks

[17:51:00.0000] <realityking>
I wonder what the appropriate roles would be for something like this http://aryweb.nl/projects/mootools-datepicker/Test/

[17:51:01.0000] <realityking>
(month or year picker)

[17:51:02.0000] <realityking>
it's not really a grid, is it?

[17:53:00.0000] <TabAtkins>
I don't think ARIA really has the ability to annotate a complex widget like a datepicker.

[18:05:00.0000] <realityking>
there's a best practive example for the days view

[18:05:01.0000] <realityking>
which is the most complex one

[18:06:00.0000] <realityking>
which is supposed to be done as a grid

[18:06:01.0000] <realityking>
which works because columns and rows actually mean something

[18:06:02.0000] <realityking>
but that isn't true for the months and days views

[18:11:00.0000] <TabAtkins>
Yeah, indeed.

[18:18:00.0000] <jamesr>
good god charles pritchard writes a lot of emails

[18:18:01.0000] <jamesr>
not short ones either

[18:19:00.0000] <heycam>
I find it difficult to reply to him, because he seems to make a lot of irrelevant statements in his emails, which don't really deal with the topics being discussed

[18:21:00.0000] <jamesr>
seems more interested in talking than communicating

[19:34:00.0000] <jamesr>
question for anyone who's implemented canvas' globalCompositeOperation="copy" for text correctly - is there any reasonable way to implement this other than actually constructing a intermediate buffer to render into?

[19:57:00.0000] <roc>
depends on your graphics library

[19:57:01.0000] <roc>
we use a temporary buffer

[19:57:02.0000] <roc>
does that bother you?

[19:58:00.0000] <jamesr>
not necessarily. it's avoidable in some cases for most operations, but there doesn't seem to be any obvious way to avoid the buffer for text

[19:59:00.0000] <jamesr>
i suspect (hope) that people don't use these globalCompositeOperations enough for it to matter too much

[19:59:01.0000] <roc>
I personally would not bother trying to optimize it

[02:02:00.0000] <annevk>
if window.find selects, shouldn't it maybe be in AryehGregor's draft?

[02:02:01.0000] <annevk>
Hixie, I still think we should have something for in-page menus and toolbars

[02:35:00.0000] <jgraham>
annevk: Is there some way to get more context in the webapps tracker diff view?

[02:41:00.0000] <annevk>
not anymore

[02:41:01.0000] <annevk>
see http://code.google.com/p/html5/source/browse/trunk/web-apps-tracker/trackerlib.py

[02:42:00.0000] <annevk>
I suppose you can patch it back in

[02:42:01.0000] <david_carlisle>
should sliders from input type=range have keyboard controls to step the value by default? seems left/right arrow keys work in Opera 11, but not in Chrome dev?

[02:43:00.0000] <hsivonen>
david_carlisle: I'd expect them to be keyboard-accessible without scripting

[02:43:01.0000] <david_carlisle>
so would I but....

[02:43:02.0000] <david_carlisle>
http://dpcarlisle.blogspot.com/2011/07/slinky-canvas.html

[02:43:03.0000] <hsivonen>
david_carlisle: Chrome's HTML5 forms implementation is incomplete to put it politely

[02:44:00.0000] <david_carlisle>
hsivonen: shoul dteh spec say something about keyboard access or is that at another level spec-wise

[02:44:01.0000] <david_carlisle>
the

[02:44:02.0000] <jgraham>
It should be obvious really. But it's a UI issue so the spec can't mandate UI

[02:44:03.0000] <jgraham>
s/so/and/

[02:45:00.0000] <hsivonen>
david_carlisle: in theory, maybe it belongs on the level on intrinsic behavior of native widgets. In practice, I guess it would be good for the spec to say something, because the exact events may be relevant to authors who override stuff with their own event handlers

[02:47:00.0000] <david_carlisle>
hsivonen: or maybe it's best just to moan here until you all implement it the same way

[02:48:00.0000] <hsivonen>
similarly, the spec pretends to delegate default <label> onclick behavior to host OS conventions and the behavior of pressing enter to submit a form, but in practice, browsers need to be consistent

[02:49:00.0000] <hsivonen>
so the spec fails to be useful when browser developers end up reverse engineering other products anyway

[02:49:01.0000] <hsivonen>
for example, the <label> case is being addressed in Gecko based on observations of the behavior of other browsers

[02:51:00.0000] <david_carlisle>
thanks all for the thoughts, good to know I hadn't missed something and that convergence is expected, as implementations catch up with each other

[02:51:01.0000] <hsivonen>
david_carlisle: filing a bug on the chromium bug tracker might be more effective than pointing out the flaw here

[02:51:02.0000] <david_carlisle>
hsivonen: OK, will do

[02:51:03.0000] <david_carlisle>
here is quicker response though:-)

[03:17:00.0000] <annevk>
NodeIterator is a little more complicated than expected

[04:08:00.0000] <AnselmBradford>
Would anyone be able to explain the rationale behind this to me? The id attribute lists the following:  "Identifiers are opaque strings. Particular meanings should not be derived from the value of the id attribute" - which I interpret to mean the semantics of an element should not be conveyed in an id. The class attribute on the other hand includes "authors are encouraged to use values that describe the nature of the content". Why is

[04:08:01.0000] <AnselmBradford>
attaching semantic meaning desirable for classes but not ids?

[04:10:00.0000] <Ms2ger>
Wrong interpretation

[04:10:01.0000] <Ms2ger>
Authors should use values that describe the nature of the content for both classes and ids, but you shouldn't try to derive meaning from them in other people's content

[04:12:00.0000] <annevk>
Ms2ger, if you can figure out a pseudo-algorithm for what moveNext() does let me know

[04:12:01.0000] <Ms2ger>
What's moveNext?

[04:12:02.0000] <annevk>
I mean nextNode()

[04:12:03.0000] <annevk>
I'm pretty close, but it is kind of hairy :(

[04:13:00.0000] <Ms2ger>
Sounds about right, then :)

[04:16:00.0000] <AnselmBradford>
Ms2ger: Ahh thanks! Is that what is meant by "opaque strings", that the values are to be treated as  essentially meaningless to anyone but the author?

[04:16:01.0000] <Ms2ger>
Right

[04:22:00.0000] <AnselmBradford>
Ms2ger: is there a difference between ids and classes in this regard? e.g. why aren't class values identified as opaque strings too? Also shouldn't that term be defined in terminology... or am I just naive to it

[04:23:00.0000] <Ms2ger>
Maybe

[04:26:00.0000] <Ms2ger>
annevk, do you know of any compat issues Opera has by not supporting window.find?

[04:26:01.0000] <annevk>
/me looks

[04:27:00.0000] <annevk>
"Breaks the "find" button in TinyMCE (possibly also the search/replace feature)"

[04:27:01.0000] <annevk>
that seems to be the only thing out there however

[04:28:00.0000] <annevk>
no other report

[04:29:00.0000] <annevk>
I reckon if Gecko drops it that's no longer a problem :)

[04:32:00.0000] <annevk>
sweet

[04:32:01.0000] <annevk>
NodeIterator is not interoperable at all

[04:40:00.0000] <smaug____>
bz was hinting that we should just make window.find() no-op

[04:47:00.0000] <Ms2ger>
krijn, happy birthday!

[05:11:00.0000] <AnselmBradford>
with id attributes ... the spec says "The value must not contain any space characters." and in the next paragraph "user agents must associate the element with the given value (exactly, including any space characters)" is this saying that authors should not use spaces but UAs should honor them if present?

[05:12:00.0000] <smaug____>
annevk: in which way is NodeIterator not interoperable?

[05:12:01.0000] <jgraham>
AnselmBradford: Ye

[05:12:02.0000] <jgraham>
s

[05:13:00.0000] <AnselmBradford>
jgraham: cheers!

[05:14:00.0000] <david_carlisle>
hsivonen: I reported a  chrome bug on slider controls http://code.google.com/p/chromium/issues/detail?id=89698    but they say it's not a bug, I have to tab to focus, not use the mouse, hmmmm

[05:16:00.0000] <hsivonen>
david_carlisle: not having click to focus the control is odd

[05:16:01.0000] <jgraham>
s/odd/buggy/

[05:16:02.0000] <david_carlisle>
hsivonen:  "If you don't like this focus behavior, please file another bug." I guess that's what I'll do...

[05:17:00.0000] <hsivonen>
david_carlisle: probably the best course of action

[06:03:00.0000] <jgraham>
http://blogs.msdn.com/b/ieinternals/archive/2011/07/18/optimal-html-head-ordering-to-avoid-parser-restarts-redownloads-and-improve-performance.aspx#10187680 suggests someone needs to do the "UA requirements" vs "author requirements" talk at Microsoft

[07:22:00.0000] <annevk>
smaug____, e.g. how often a filter is invoked that returns nothing

[07:40:00.0000] <annevk>
this is pretty great: http://my.opera.com/chooseopera/blog/2011/07/19/and-now-for-something-completely-random-your-own-band

[08:22:00.0000] <jgraham>
Hmm, I thought it was required that a charset-declaring <meta> was the first element in the <head>, but it doesn't seem to be

[08:22:01.0000] <jgraham>
+if any

[08:24:00.0000] <Ms2ger>
It might have used to be

[08:35:00.0000] <nlogax>
i think it only needs to fit in the first 512 bytes

[08:38:00.0000] <Philip`>
1024

[08:38:01.0000] <annevk>
smaug____, another difference is when INVALID_STATE_ERR exception is thrown

[08:38:02.0000] <Philip`>
(See http://www.whatwg.org/specs/web-apps/current-work/multipage/semantics.html#charset512 )

[08:38:03.0000] <Philip`>
(which is not at all a confusing link)

[08:39:00.0000] <annevk>
smaug____, in Gecko it happens if the callback invokes detach(); in WebKit you have to invoke the method again

[08:50:00.0000] <dglazkov>
good morning, Whatwg!

[08:53:00.0000] <annevk>
good afternoon, dglazkov!

[08:56:00.0000] <Ms2ger>
Philip`, #charset1024 works too

[08:56:01.0000] <Philip`>
Ms2ger: Less fun though

[09:05:00.0000] <annevk>
As long as the implementor requirements are that you have to look for an encoding forever it is still not really nice though

[09:05:01.0000] <annevk>
That is, not fully deterministic

[09:06:00.0000] <annevk>
Or maybe that changed

[09:22:00.0000] <matjas>
“It's utterly bizarre to me that HTML5 would impose an "authoring requirement" which it doesn't bother to add to its implementation requirements, particularly when its failure to do so could result in a security vulnerability. But I don't work on much HTML5 stuff myself.” — Eric Law http://blogs.msdn.com/b/ieinternals/archive/2011/07/18/optimal-html-head-ordering-to-avoid-parser-restarts-red

[09:22:01.0000] <matjas>
ownloads-and-improve-performance.aspx

[09:23:00.0000] <matjas>
let’s try that again: http://blogs.msdn.com/b/ieinternals/archive/2011/07/18/optimal-html-head-ordering-to-avoid-parser-restarts-redownloads-and-improve-performance.aspx

[09:25:00.0000] <annevk>
he doesn't explain the vulnerability

[09:28:00.0000] <jgraham>
It would be nice if he would

[09:28:01.0000] <jgraham>
I was going to ask but got distracted by charset metas not having to be first in the head per spec

[09:29:00.0000] <jgraham>
I am guessing it it just a "some people might have bad sanitizers" vunerability, but maybe I will be surprised

[09:37:00.0000] <annevk>
How would you describe (1 << (node.nodeType() - 1)) & iterator.whatToShow in prose?

[09:38:00.0000] <annevk>
It is needed to map the constants of http://www.w3.org/TR/DOM-Level-2-Traversal-Range/traversal.html#Traversal-NodeFilter to Node.XXX_Node constants

[09:38:01.0000] <Philip`>
"If the n'th bit (where 0 is the least significant bit) of whatToShow is set, where n is nodeType"?

[09:39:00.0000] <Philip`>
(or nodeType-1)

[09:43:00.0000] <Ms2ger>
/me wonders which implementation that's from

[09:52:00.0000] <Ms2ger>
<annevk> yeah, but we first need to figure out the Web platform

[09:52:01.0000] <Ms2ger>
annevk++

[09:58:00.0000] <annevk>
Ms2ger, context? :)

[09:59:00.0000] <Ms2ger>
http://krijnhoetmer.nl/irc-logs/whatwg/20110707#l-135

[10:01:00.0000] <annevk>
you have quite the backlog

[10:04:00.0000] <Ms2ger>
Yes

[10:04:01.0000] <Ms2ger>
<krijn> (Server will be down a bit today, getting a new fiberglass connection)

[10:04:02.0000] <Ms2ger>
A bit today?

[10:06:00.0000] <jgraham>
Any reason that the web dom core tests aren't on the W3C server? Or am I looking in the wrong place?

[10:07:00.0000] <annevk>
they are in http://dvcs.w3.org/hg/webapps/

[10:08:00.0000] <annevk>
http://dvcs.w3.org/hg/webapps/file/tip/DOMCore/tests/submissions/Ms2ger

[10:08:01.0000] <jgraham>
Oh, I thought they might be in the same place as the spec or something. Would make them easier to find

[10:08:02.0000] <Ms2ger>
w3c-test.org is out-of-date again..

[10:08:03.0000] <annevk>
they used to be

[10:08:04.0000] <Ms2ger>
/me sighs

[10:08:05.0000] <annevk>
no idea why they moved

[10:40:00.0000] <krijnserver>
Ms2ger: if you (or anyone else) have the missing logs, please mail them to me

[10:41:00.0000] <Philip`>
/me has logs

[10:41:01.0000] <Philip`>
What time period?

[10:42:00.0000] <krijnserver>
Ms2ger: heh, and thanks :)

[10:42:01.0000] <krijnserver>
Philip`: a week or something

[10:42:02.0000] <Philip`>
krijnserver: Which week?

[10:43:00.0000] <krijnserver>
http://krijnhoetmer.nl/irc-logs/whatwg/201107

[10:43:01.0000] <krijnserver>
That's all quite empty

[10:43:02.0000] <krijnserver>
Brb, fixing food

[10:44:00.0000] <annevk>
/me wonders what servers eat these days

[10:48:00.0000] <Philip`>
krijnhuman: http://zaynar.co.uk/misc/whatwg-20110701-to-20110719ish.log

[10:49:00.0000] <Philip`>
/me apologises for the bottom half of the file being backwards

[10:58:00.0000] <krijnhuman>
annevk: same stuff as always, http://lockerz.com/s/121710484, keeps servers healthy

[11:00:00.0000] <krijnhuman>
Philip`: thanks, will add them tonightish

[11:06:00.0000] <Ms2ger>
20:34 < annevk> teehee logs offline again

[11:06:01.0000] <Ms2ger>
20:34 < zewt> let's all say mean things

[11:09:00.0000] <annevk>
WHATWG secrets unveiled

[11:10:00.0000] <krijnh>
My idea was to add some..

[11:13:00.0000] <Ms2ger>
01:06 < weinig> Hixie: I am implementing annevk's new Event stuff just to rid my future test cases of it :)

[11:13:01.0000] <Ms2ger>
weinig++

[11:13:02.0000] <weinig>
/me should really land that

[11:17:00.0000] <smaug____>
I wonder when to implement the new event initialization

[11:19:00.0000] <Ms2ger>
Now :)

[11:20:00.0000] <smaug____>
Ms2ger: well, is the spec stable in that case?

[11:20:01.0000] <smaug____>
has it been reviewed?

[11:21:00.0000] <Ms2ger>
As stable as it gets without implementations, I think

[11:24:00.0000] <annevk>
yeah

[11:24:01.0000] <annevk>
need implementations

[11:24:02.0000] <annevk>
and implementor feedback if there are bugs :)

[11:25:00.0000] <weinig>
annevk: bugs?

[11:27:00.0000] <annevk>
in the spec

[11:27:01.0000] <annevk>
I haven't heard of any so far though

[11:33:00.0000] <annevk>
Anyone here have a problem with DOM Core including NodeIterator and TreeWalker?

[11:34:00.0000] <annevk>
In favor?

[11:35:00.0000] <Ms2ger>
meh

[11:35:01.0000] <Ms2ger>
You get to maintain it :)

[11:39:00.0000] <jgraham>
annevk: I like the idea of them being specified somewhere. I don't really care where :)

[12:02:00.0000] <Hixie>
annevk: ooh, i like your thinking (making window.find() aryeh's problem) :-)

[12:03:00.0000] <Ms2ger>
Hixie, hah

[12:05:00.0000] <Ms2ger>
annevk, you realize he's just going to play that trick on you next time, right? :)

[12:05:01.0000] <annevk>
would not be the first time :)

[14:13:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#traversal

[14:13:01.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-traversal

[14:16:00.0000] <TabAtkins>
Argh, I hate numeric constants and bitmasks so bad.

[14:19:00.0000] <annevk>
Thanks for the input

[14:19:01.0000] <Ms2ger>
Do we use <sup>th</sup> consistently?

[14:19:02.0000] <TabAtkins>
Sorry, I know legacy constraints tie our hands here.

[14:19:03.0000] <TabAtkins>
But still.

[14:19:04.0000] <annevk>
Or should I say, film at eleven!

[14:20:00.0000] <Ms2ger>
At eleven? Should hurry up, then

[14:21:00.0000] <annevk>
Ms2ger, not sure, I think we should though

[14:21:01.0000] <Ms2ger>
Should check that one day

[14:22:00.0000] <annevk>
There are a few things about consistency that ought to be done one day

[14:29:00.0000] <annevk>
By the way, the nextNode() and previousNode() algorithms are pretty much identical. Maybe they should use a common algorithm.

[14:35:00.0000] <Ms2ger>
I believe Gecko shares the code there

[14:36:00.0000] <annevk>
Also for maintenance of the specification it would make sense. However, it would be cool if David and others could review it first :)

[14:39:00.0000] <Ms2ger>
sicking, ^

[15:17:00.0000] <Ms2ger>
Got through my #whatwg backlog, yay

[15:17:01.0000] <jgraham>
annevk: Stop 5 of nextNode is wrong

[15:17:02.0000] <jgraham>
Oh, maybe it's not

[15:18:00.0000] <annevk>
substep 5?

[15:18:01.0000] <jgraham>
Yeah

[15:19:00.0000] <jgraham>
It's right but the whole substep/overall steps thing is confusing

[15:19:01.0000] <annevk>
it's a pattern more specs use, not sure how to do it differently

[15:19:02.0000] <jgraham>
Well HTML5 uses it and it can be confusing there. ES5 does something different

[15:19:03.0000] <jgraham>
/me looks for an example

[15:20:00.0000] <Hixie>
i'm not a fan of the terminology

[15:20:01.0000] <Hixie>
not sure how to fix it though

[15:20:02.0000] <Hixie>
ES5's approach isn't one i like either

[15:21:00.0000] <jgraham>
http://es5.github.com/#x10.6 is one example

[15:21:01.0000] <jgraham>
It uses "repeat while" instead of "goto"

[15:22:00.0000] <Hixie>
what if index < 0 at the start of the loop?

[15:22:01.0000] <Hixie>
the html spec does sometimes say "run the following substeps while (some condition)", iirc

[15:22:02.0000] <annevk>
this one is not really goto

[15:22:03.0000] <annevk>
it's more like continue/break

[15:23:00.0000] <jgraham>
Yeah, it's continue/break phrased as goto :)

[15:23:01.0000] <Ms2ger>
goto hell?

[15:23:02.0000] <jgraham>
Also makes it confusing when you say "terminate these steps"

[15:24:00.0000] <jgraham>
Because it could be the inner steps or the overall set of steps

[15:24:01.0000] <Hixie>
yeah the problem i regularly have is with "break"

[15:24:02.0000] <Hixie>
or even worse, "continue"

[15:24:03.0000] <annevk>
this one has both :evil:

[15:25:00.0000] <Ms2ger>
The web :evil:

[15:25:01.0000] <Hixie>
if any of you can come up with better wording for any of hte algorithms in the html spec, don't hesitate to let me know

[15:26:00.0000] <jgraham>
/me should try to work out a nice way to do this to see if he is actually learning anything from "programming pearls"

[15:27:00.0000] <zewt>
Hixie: fwiw, my impression from that is that it sort of sounds like you're supposed to abort the steps immediately if (some condition) ever turns false mid-loop (which I assume is usually not intended)

[15:27:01.0000] <Hixie>
zewt: yeah, that's another problem with that wording

[15:37:00.0000] <jgraham>
Doesn't "return" imply "and terminate these steps"?

[15:37:01.0000] <Hixie>
no

[15:38:00.0000] <Hixie>
there are a number of algorithms that return and continue

[15:39:00.0000] <annevk>
http://wiki.whatwg.org/wiki/How_to_write_a_spec should maybe be updated with instructions on writing algorithms

[15:40:00.0000] <annevk>
and maybe within specs there should be instructions on how to read them?

[15:40:01.0000] <annevk>
maybe after jgraham put some thought into it

[15:40:02.0000] <annevk>
nn

[15:40:03.0000] <Hixie>
as far as reading them, i think it's best if they're just english

[15:41:00.0000] <annevk>
yeah, ideally that's sufficient :)

[15:41:01.0000] <jgraham>
Well they are in some sense. But they contain terms of art

[15:45:00.0000] <roc>
Maybe we should write the algorithms in JS

[15:45:01.0000] <roc>
it would make the control structures clearer

[15:45:02.0000] <roc>
with defined syntax for escaping back into English

[15:47:00.0000] <TabAtkins>
That would be interesting.

[15:47:01.0000] <TabAtkins>
Would probably make flexbox easier to write.

[15:47:02.0000] <heycam>
in the JS algorithm you would construct a string and then say "evalEnglish(theString);" :)

[15:47:03.0000] <Hixie>
then we'd have to define the interaction of english and js

[15:47:04.0000] <roc>
that's exactly what I was thinking

[15:48:00.0000] <Philip`>
Do it as a function call, where the function happens to be defined in English

[15:48:01.0000] <Philip`>
so all the arguments and return values are nice and explicit

[15:48:02.0000] <roc>
AI("Queue a task to fire all applicable queue events");

[15:48:03.0000] <Hixie>
also you'd have to define which parts of the JS semantics you meant to expose and which were side-effects

[15:49:00.0000] <roc>
yeah, it's problematic

[15:49:01.0000] <heycam>
how simple specs could become if the AI function were specified. someone should get on to defining that.

[15:49:02.0000] <roc>
but what we have right now is problematic too: pseudocode in English with gotos

[15:49:03.0000] <Hixie>
i just wish i could have unit tests for english

[15:49:04.0000] <Hixie>
and that my compiler had less latency. and didn't talk back to me. :-P

[15:51:00.0000] <Philip`>
Unit tests for English could just be examples, listing input and output and side-effects

[15:51:01.0000] <Philip`>
The only hard part is executing the unit tests

[15:51:02.0000] <roc>
Google can implement it. Search all the Web's scripts for comments related to the request string, cobble the top 10 hits together in a way that compiles, and eval it

[15:52:00.0000] <TabAtkins>
Anybody know of a half-decent explanation of a Coons patch mesh gradient, so I can implement it?

[15:52:01.0000] <TabAtkins>
Google isn't doing me very good here.

[15:52:02.0000] <TabAtkins>
And I think pdf.js hasn't implemented them yet.

[15:53:00.0000] <Philip`>
Like in http://tavmjong.free.fr/SVG/MESH/Mesh.html ?

[15:53:01.0000] <TabAtkins>
Yes.

[16:05:00.0000] <TabAtkins>
Okay, I think I got it.  First, WLOG assume you're drawing into the unit square.  Assign a color to each corner, and use linear interpolation to assign a gradient to each edge.  Then use bilinear interpolation to fill in the interior.

[16:08:00.0000] <jgraham>
(fwiw I think that defining some terms like in http://es5.github.com/#x5.2 would be helpful, without going all the way to real code)


2011-07-20
[18:29:00.0000] <Hixie>
so apparently the security concern eric is worried about is an attacker smuggling a <base> in the <body> before a <script> with a relative src=""

[18:29:01.0000] <Hixie>
as far as I can tell, IE does indeed ignore <base href> outside <head>

[18:29:02.0000] <jamesr>
all your <base> are belong to us?

[18:29:03.0000] <Hixie>
that surprises me, though

[18:29:04.0000] <Hixie>
does anyone have any more detailed data on what IE does with <base>?

[18:29:05.0000] <Hixie>
my data is all out of date and i don't have a recent IE to test

[18:42:00.0000] <beverloo>
Would someone have a link to Aryeh's contenteditable / execCommand spec at hand?

[18:44:00.0000] <beverloo>
got it: http://aryeh.name/gitweb.cgi?p=editcommands;a=blob_plain;f=editcommands.html

[18:54:00.0000] <Hixie>
testing IE9 it seems the same as IE8, which is to say, in both quirks and non-quieks it ignores <base href> in the body

[18:54:01.0000] <Hixie>
i am surprised

[18:55:00.0000] <AryehGregor>
The blog post said IE has behaved that way since IE7.

[18:55:01.0000] <AryehGregor>
Why are you surprised?

[18:55:02.0000] <Hixie>
i didn't read the blog post

[18:55:03.0000] <AryehGregor>
Also, under what circumstance is an attacker going to be able to smuggle <base> into the body but not any other XSS vector?

[18:55:04.0000] <Hixie>
and i'm surprised because it seems like this would cause them compat issues

[18:56:00.0000] <AryehGregor>
He linked to this: http://blogs.msdn.com/b/ie/archive/2005/08/29/457667.aspx

[18:56:01.0000] <AryehGregor>
Which is, indeed, entitled "All your <base> are belong to us".

[18:56:02.0000] <Hixie>
the idea is that the victim has a blacklist approach and forgets <base>

[18:56:03.0000] <Hixie>
i knew about that blog post but i didn't realise it affected quirks mode as well

[18:56:04.0000] <Hixie>
they're usually so scared of changing quirks mode

[18:56:05.0000] <Hixie>
anyway

[19:05:00.0000] <Hixie>
blimey

[19:05:01.0000] <Hixie>
IE has 5 quirks modes?

[19:05:02.0000] <AryehGregor>
Hixie, maybe aCaseSensitive in WebKit does something along the lines of NFK{C,D}?

[19:06:00.0000] <Hixie>
AryehGregor: apparently it uses the platform user-friendly matching algorithm, or something

[19:06:01.0000] <AryehGregor>
I think IE10 has one quirks mode, three compatibility modes, and one standards mode.

[19:06:02.0000] <Hixie>
i think my wording is more accurate. :-P

[19:06:03.0000] <jamesr>
Hixie: http://hsivonen.iki.fi/doctype/ie8-mode.png

[19:07:00.0000] <jamesr>
but that doesn't include IE9/10 variations

[19:07:01.0000] <Hixie>
blimey

[19:08:00.0000] <Hixie>
their QA team must be... large

[19:08:01.0000] <Hixie>
or ill

[19:08:02.0000] <TabAtkins>
Both.

[19:11:00.0000] <TabAtkins>
Hixie: I thought that IE did the implied-body magic for known elements; it was only unknown elements (like the new sectioning elements) that got swallowed by the implied <head>.

[19:11:01.0000] <Hixie>
some known elements too

[19:11:02.0000] <TabAtkins>
Jeezus.

[19:12:00.0000] <AryehGregor>
IE clearly doesn't do an implied body for, say, <style>.

[19:12:01.0000] <AryehGregor>
In any version.

[19:13:00.0000] <TabAtkins>
Yes, of course not.  That's a <head> element.

[19:13:01.0000] <Hixie>
tab is referring to the mail i just sent

[19:13:02.0000] <TabAtkins>
Hixie's claiming that <form> doesn't trigger implied-body.

[19:13:03.0000] <AryehGregor>
/me didn't see it

[19:13:04.0000] <AryehGregor>
Oh.  That would be . . . insane.

[19:13:05.0000] <Hixie>
i literally just sent it seconds ago

[19:13:06.0000] <Hixie>
tab is just stalking me

[19:13:07.0000] <AryehGregor>
But then, all pre-HTML5 HTML parsers were insane, so . . .

[19:13:08.0000] <Hixie>
yeah well the HTML5 parser isn't that sane either

[19:20:00.0000] <AryehGregor>
Hixie, you just resolved <http://www.w3.org/Bugs/Public/show_bug.cgi?id=12570> in r6314.

[21:13:00.0000] <Hixie>
wow, the patents part of the CLA for W3C Community Groups is surprisingly toothless

[21:14:00.0000] <Hixie>
it only covers your patents on parts of the spec that you actually write yourself

[21:17:00.0000] <Hixie>
also it seems there's a trivial way to get out of the licensing requirements: you just form an independant corporation to which you transfer your patents on the condition of receiving all royalties

[21:24:00.0000] <roc>
Hixie: that "independent corporation" approach actually works around a lot of patent agreements, even ones in which money changes hands

[21:25:00.0000] <roc>
"you have a cross-licensing agreement with us? Fine, we'll spin out our patents to an 'independent' entity to sue you"

[21:25:01.0000] <Hixie>
why would people not require that any license agreement follow the patents?

[21:26:00.0000] <roc>
you'd think

[21:26:01.0000] <roc>
maybe some do

[21:26:02.0000] <Hixie>
 that's the kind of thing i do even when playing monopoly...

[21:27:00.0000] <roc>
This appears to be a case of that: http://thepriorart.typepad.com/the_prior_art/2010/04/mobilemedia-ideas-v-apple.html but I'm not 100% sure

[21:29:00.0000] <roc>
it may involve some kind of workaround where your staff develop some technology but the patent is never actually owned by your company, it goes straight to a holding company

[23:09:00.0000] <Hixie>
so if i add this download="" feature... is it something i shouldn't be adding to the w3c spec?

[23:10:00.0000] <Hixie>
i don't understand whether we're under a feature freeze or not

[23:55:00.0000] <nessy>
difficult call - not sure if an attribute falls under feature freeze - I'd think only big things such as adaptive streaming or an audio api

[01:11:00.0000] <matjas>
hmm, so no browser supports document.innerHTML yet?

[01:11:01.0000] <matjas>
Ms2ger (assuming you read the logs): http://ms2ger.freehostia.com/tests/html5/dynamic-markup-insertion/innerhtml-02.xhtml 404s

[02:38:00.0000] <annevk>
matjas, yeah, it's a new feature

[02:42:00.0000] <jgraham>
annevk: Is your nextNode stuff intended to be correct if the filter function manipulates theDOM?

[02:46:00.0000] <matjas>
annevk: is there an open Opera bug ticket for document.innerHTML or should I file one?

[02:53:00.0000] <annevk>
jgraham, when "iterator collection" is mutated referenceNode and pointerBeforeReferenceNode need to be updated, that is not defined yet

[02:53:01.0000] <annevk>
jgraham, however, once that is defined it should be alright

[02:54:00.0000] <annevk>
matjas, I can't find one so fast, feel free

[02:57:00.0000] <matjas>
annevk: done

[02:57:01.0000] <matjas>
Mozilla: http://bugzil.la/563320 WebKit: http://webk.it/60316 IE: https://connect.microsoft.com/IE/feedback/details/680040/implement-document-innerhtml

[03:04:00.0000] <jgraham>
annevk: Yes, I suppose that would work. And the detached flag. Although if I understand your intent correctly, it doesn't seem to match existing browsers  http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1080

[03:04:01.0000] <jgraham>
In particular it seems they stop if they encounter a node they have seen before, but don't set pointerBeforeReferenceNode

[03:08:00.0000] <annevk>
jgraham, what about the detached flag?

[03:08:01.0000] <jgraham>
annevk: Maybe I don't understand what it is for

[03:09:00.0000] <jgraham>
Why do you check it after the filter function? What causes it to be set?

[03:10:00.0000] <annevk>
jgraham, the filter could set it by invoking detach()

[03:10:01.0000] <annevk>
jgraham, which reminds me, the current algorithm does not handle recursive invocation either

[03:11:00.0000] <annevk>
jgraham, I don't really understand the results from your NodeIterator btw

[03:11:01.0000] <annevk>
jgraham, it has SHOW_ELEMENT set yet Gecko and WebKit (not Opera) are happily returning Text nodes

[03:13:00.0000] <jgraham>
The text node never gets returned, but it does get set as the referenceNode

[03:13:01.0000] <annevk>
doh, I should read better

[03:14:00.0000] <jgraham>
(which also doesn't match the spec I guess)

[03:16:00.0000] <annevk>
Can that situation only occur at the end of the collection and during modifications?

[03:17:00.0000] <annevk>
I could modify step 3.1 to account for it

[03:19:00.0000] <jgraham>
I imagine that internally they update the referenceNode irrespective of whether the node matches the filter

[03:19:01.0000] <jgraham>
So you would just need to move that inside the loop

[03:20:00.0000] <jgraham>
I think it is only observable after modification or at the end though

[03:20:01.0000] <annevk>
that does not work if something throws an exception

[03:22:00.0000] <annevk>
if you update referenceNode all the time and filter throws an exception or sets detach() it will have the wrong value

[03:24:00.0000] <jgraham>
It seems to get set to null in that case

[03:25:00.0000] <jgraham>
Oh, only in gecko

[03:25:01.0000] <jgraham>
In WebKit it seems to be the node that you were trying to filter

[03:26:00.0000] <jgraham>
So you could set referenceNode to node at the start of the loop

[03:28:00.0000] <annevk>
And set it to null if an exception was raised or detach() was called? And that makes the whole thing no longer function?

[03:28:01.0000] <jgraham>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1081 is what I looked at

[03:30:00.0000] <annevk>
It does not even throw an exception in Gecko?

[03:31:00.0000] <annevk>
oh it does

[03:31:01.0000] <annevk>
you were not printing that

[03:32:00.0000] <annevk>
if you just do throw "x" referenceNode does not become null

[03:33:00.0000] <annevk>
it is only null when things are detached

[03:34:00.0000] <annevk>
so invoking detach() just initializes referenceNode to null

[03:34:01.0000] <jgraham>
/me wonders what the point of detach is 

[03:35:00.0000] <annevk>
so you can clear up resources

[03:35:01.0000] <annevk>
silly legacy thing

[03:38:00.0000] <annevk>
thanks for helping out btw

[03:38:01.0000] <jarek>
Hi

[03:38:02.0000] <jarek>
I'm trying to implement CSS parser, but it's rather hard for me to understand some of the regular expressions used in specification

[03:38:03.0000] <jarek>
e.g.: \\[0-9a-f]{1,6}(\r\n|[ \n\r\t\f])?

[03:39:00.0000] <jarek>
is there somewhere a document written in human language that would explain which strings are allowed for a given token?

[03:40:00.0000] <annevk>
other than the CSS specification and its dependencies? I do not think so

[03:40:01.0000] <jarek>
so I will have to read a book on regular expressions in order to understand the specification? :/

[03:42:00.0000] <annevk>
no

[03:42:01.0000] <annevk>
something on yacc

[03:43:00.0000] <jarek>
I have to implement this with javascript

[03:43:01.0000] <jarek>
afaik there is no such tool like yacc that would generate js code

[03:43:02.0000] <jgraham>
Well you will likely need to understand regexp, yes

[03:45:00.0000] <jgraham>
And there are some parser-generators avaliable that will create javascript

[03:45:01.0000] <jgraham>
But your chances of success if you don't understand the input seem rather small

[03:46:00.0000] <jarek>
there is peg.js, but it genarates the whole parser (I just need tokenizer)

[03:48:00.0000] <jgraham>
Well I would read http://flex.sourceforge.net/manual/Patterns.html#Patterns and convert to javascript regexp

[03:48:01.0000] <jarek>
also, CSS2 and CSS3 specifications seem to be using slightly different tokens

[03:48:02.0000] <jarek>
e.g. there is BAD_STRING token in CSS2, but I can't find it in CSS3 specs

[03:49:00.0000] <annevk>
CSS3 is out of date

[03:49:01.0000] <annevk>
ironically enough

[03:52:00.0000] <jarek>
annevk: should I implement the tokenizer according to CSS2 specification and later update it when there is new revision of http://www.w3.org/TR/css3-syntax/?

[03:52:01.0000] <jarek>
or perhaps I should use http://www.w3.org/TR/css3-syntax/ as the only reference?

[03:53:00.0000] <annevk>
jgraham, so I tested again and I think referenceNode only becomes an "invisible" node due to modification

[03:54:00.0000] <annevk>
jarek, you should use CSS2 with modifications

[03:54:01.0000] <annevk>
jarek, css3-syntax is hopelessly out of date

[04:08:00.0000] <annevk>
ah

[04:08:01.0000] <annevk>
recursive calls throw an exception as well

[04:08:02.0000] <annevk>
in WebKit some kind of callstack exception

[04:08:03.0000] <annevk>
in Gecko INVALID_STATE_ERR

[04:09:00.0000] <annevk>
in Opera the same as in Gecko

[04:10:00.0000] <annevk>
in WebKit invoking a different method (e.g. previousNode) does work, it does not in Opera/Gecko

[04:11:00.0000] <annevk>
I guess I'll side with Opera/Gecko

[04:11:01.0000] <jgraham>
simplicity ++

[04:12:00.0000] <Ms2ger>
specs++

[04:13:00.0000] <annevk>
name for this flag?

[04:13:01.0000] <annevk>
"locked flag"?

[04:14:00.0000] <Ms2ger>
wfm

[04:14:01.0000] <jgraham>
yeah, whatever

[04:15:00.0000] <Ms2ger>
For detach(), s/initialize/set/?

[04:19:00.0000] <annevk>
I use initialize for readonly attributes

[04:19:01.0000] <annevk>
because they cannot really be set

[04:20:00.0000] <jgraham>
I'm not sure that is better

[04:21:00.0000] <jgraham>
Initialise implies that it is being set at the point it is created

[04:25:00.0000] <annevk>
better term?

[04:25:01.0000] <annevk>
set just seems wrong

[04:25:02.0000] <annevk>
but maybe we should use it anyway

[04:25:03.0000] <bga_>
lol http://xkcd.com/927/

[04:25:04.0000] <annevk>
for now: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-nodeiterator-nextnode

[04:26:00.0000] <jgraham>
annevk: Better than initialise. I think it is understood that the redonlyness of attributes from the external point of view doesn't matter to internal algorithms

[04:30:00.0000] <AnselmBradford>
hi all, does anyone know of a browser that supports the scoped attribute on the style element?

[04:45:00.0000] <annevk>
jgraham, done

[04:45:01.0000] <annevk>
Anyone an idea how to generalize 3.1 and 3.2 so we can have one algorithm for nextNode() and previousNode()?

[05:00:00.0000] <annevk>
WebKit allows arguments to be omitted when creating a NodeIterator but defaults whatToShow to 0

[05:00:01.0000] <annevk>
oops

[05:00:02.0000] <annevk>
Guess that is because it is the same as undefined for them

[05:35:00.0000] <Ms2ger>
matjas, updated the link to my innerhtml tests in the spec

[05:35:01.0000] <matjas>
Ms2ger: \o/

[05:36:00.0000] <annevk>
I am going to use dl class=switch btw

[05:36:01.0000] <remysharp>
If I wanted to delete localStorage using `delete localStorage.foo` it works, but am I asking for trouble as it's not spec'ed out (or that I've seen)

[05:38:00.0000] <annevk>
it is specified

[05:38:01.0000] <annevk>
"deleter void removeItem(in DOMString key)"

[05:38:02.0000] <annevk>
see Web IDL

[05:39:00.0000] <remysharp>
honestly, I'm not very good at reading Web IDLs - in fact, I don't think I even know what IDL stands for - interface something lang? ?

[05:39:01.0000] <Ms2ger>
description, I guess

[05:39:02.0000] <annevk>
definition I thought

[05:40:00.0000] <Ms2ger>
The "deleter" means that function is called when you do `delete`

[05:40:01.0000] <remysharp>
glad I'm not the only one then :)

[05:40:02.0000] <annevk>
see http://dev.w3.org/2006/webapi/WebIDL/

[05:40:03.0000] <remysharp>
cool - cheers for the clarification annevk & Ms2ger

[05:40:04.0000] <Ms2ger>
Np

[05:41:00.0000] <Ms2ger>
Oh hey, :dir

[05:41:01.0000] <annevk>
what about it?

[05:42:00.0000] <annevk>
though now you mention it, I wonder why that is not :ltr and :rtl

[05:43:00.0000] <Ms2ger>
Aharon filed a Gecko bug to implement it

[05:47:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-nodeiterator-traverse

[05:57:00.0000] <annevk>
does this sound correct:

[05:58:00.0000] <annevk>
nm

[05:58:01.0000] <annevk>
it doesn't

[06:00:00.0000] <matjas>
annevk: to support :dir(upsidedown) ;)

[06:01:00.0000] <annevk>
matjas, isn't dir just for writing direction?

[06:02:00.0000] <annevk>
or whatever the proper term for that is

[06:02:01.0000] <matjas>
</lame-joke>

[06:03:00.0000] <annevk>
vertical text exists of course, but is not a semantic thing (at least that's how it is being approached)

[06:19:00.0000] <MikeSmith>
AnselmBradford: there is a Webkit developer who's been implementing it

[06:19:01.0000] <MikeSmith>
but I don't know if the code landed yet or not

[06:19:02.0000] <MikeSmith>
I think not though

[06:21:00.0000] <AnselmBradford>
MikeSmith: thanks :)

[06:25:00.0000] <AnselmBradford>
MikeSmith: Any idea where I'd be able to find out when it lands ... http://trac.webkit.org/timeline or?

[06:25:01.0000] <MikeSmith>
AnselmBradford: hang on, I will get you the bug number

[06:26:00.0000] <MikeSmith>
https://bugs.webkit.org/show_bug.cgi?id=49142

[06:29:00.0000] <AnselmBradford>
cheers

[06:39:00.0000] <annevk>
mutation feedback requested

[06:39:01.0000] <annevk>
 if the mutated node is not a descendant of root and not an ancestor of the reference node terminate these steps

[06:39:02.0000] <annevk>
 if pointerBeforeReferenceNode is false and there is a node preceding where referenceNode was previously, set referenceNode to that node

[06:39:03.0000] <annevk>
 if pointerBeforeReferenceNode is true and there is a node following where referenceNode was previously, set referenceNode to that node

[06:39:04.0000] <annevk>
 if pointerBeforeReferenceNode is true and there is no node following where referenceNode was previously, set referenceNode to the node preceding where referenceNode was previously and set pointerBeforeReferenceNode to false

[06:39:05.0000] <annevk>
I don't think there are any other conditions

[06:40:00.0000] <jgraham>
When do these steps run? Does it apply irrespective of which node is mutated?

[06:43:00.0000] <annevk>
whenever the "iterator collection" is mutated

[06:44:00.0000] <annevk>
maybe once mutation is defined a little better we can have more details

[06:45:00.0000] <annevk>
but for now this seems sufficient

[06:45:01.0000] <jgraham>
Oh, so this is supposed to be a seperate algorithm…

[06:45:02.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#iterator-collection

[06:47:00.0000] <jgraham>
What does mutating a collection mean? Is it only operations that change the order of the collection or any operation that mutates nodes in the collection?

[06:48:00.0000] <annevk>
yeah

[06:48:01.0000] <annevk>
I guess it should say "whenever the iterator collection has nodes inserted or removed"

[06:49:00.0000] <jgraham>
Does that work if I remove a node and insert it somewhere else?

[06:49:01.0000] <jgraham>
I mean should the algorithm run once or twice?

[06:50:00.0000] <jgraham>
(also, with replaceChild)

[06:51:00.0000] <annevk>
does it matter if it runs once or twice?

[06:51:01.0000] <jgraham>
I don't know :) I haven't fully understood it yet

[06:52:00.0000] <jgraham>
In fact I'm not sure I understand it at all

[06:52:01.0000] <jgraham>
e.g. the second step says "if there is a node before where referenceNode was previously"

[06:52:02.0000] <jgraham>
But reference node might not have been moved

[06:53:00.0000] <jgraham>
And I assume the intent isn't to move referenceNode in that case

[06:54:00.0000] <annevk>
look at the first clause

[06:55:00.0000] <annevk>
I should add and is not the reference node to the first clause

[06:57:00.0000] <annevk>
jgraham, so the algorithm is terminated if referenceNode is not affected

[06:58:00.0000] <hsivonen>
Hmm. so when you set innerHTML on a title element, the appropriate end tag remains undefined? Firefox sets it to "title". Ragnarök and Chrome don't.

[06:58:01.0000] <hsivonen>
see http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1084

[06:59:00.0000] <annevk>
the behavior of Gecko is not really nice

[07:00:00.0000] <jgraham>
annevk: Yes, OK

[07:01:00.0000] <annevk>
thanks: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#iterator-collection

[07:01:01.0000] <hsivonen>
Hixie: if the Gecko behavior isn't what you meant, it would have been nice to have a note affirming that in the spec

[07:02:00.0000] <hsivonen>
Hixie: for document.createElement("title").innerHTML = "</title>"; that is

[07:02:01.0000] <jgraham>
annevk: "where the referenceNode was was previously" isn't too clear

[07:02:02.0000] <annevk>
yeah

[07:02:03.0000] <annevk>
not sure how to do that

[07:02:04.0000] <annevk>
or do it better, that is

[07:04:00.0000] <jgraham>
What do you actually mean? If I have [a b c * d e] and mutate it to [c * d a b e], where * is the reference node, which node do you have in mind?

[07:04:01.0000] <jgraham>
e?

[07:06:00.0000] <annevk>
* is not the reference node

[07:06:01.0000] <annevk>
it's either [c] or [d]

[07:07:00.0000] <annevk>
you have a reference node

[07:07:01.0000] <annevk>
and an indicator that is either before or after the reference node

[07:07:02.0000] <annevk>
for direction

[07:09:00.0000] <jgraham>
I don't understand :)

[07:09:01.0000] <jgraham>
Why isn't it possible to set up a situation like I described?

[07:10:00.0000] <annevk>
the reference node is an actual node

[07:10:01.0000] <jgraham>
Yes, I meant * to be an actual node

[07:10:02.0000] <hsivonen>
/me filed https://bugzilla.mozilla.org/show_bug.cgi?id=672775

[07:10:03.0000] <jgraham>
Sorry, maybe that wasn't clear

[07:10:04.0000] <annevk>
if you mutate it to that the referenceNode is not affected so nothing happens

[07:11:00.0000] <annevk>
only the list mutates

[07:11:01.0000] <annevk>
but how that happens is clear from the definition of collection

[07:11:02.0000] <annevk>
or follows from

[07:11:03.0000] <hsivonen>
so...

[07:11:04.0000] <jgraham>
Well the mutated node can be a desendant of root and an ancestor of referenceNode

[07:12:00.0000] <jgraham>
in my example

[07:12:01.0000] <jgraham>
If * is a child of c in the actual tree

[07:12:02.0000] <hsivonen>
if innerHTML is being set on style, xmp, iframe, noembed, noframes or script, is there any use in putting the tokenizer in the RAWTEXT or script data state without an appropriate end tag instead of just putting the tokenizer in the PLAINTEXT state?

[07:14:00.0000] <hsivonen>
doh. there actually *is* already a green note in the spec saying all these things

[07:14:01.0000] <hsivonen>
I wonder when that note appeared

[07:14:02.0000] <hsivonen>
could I have missed it when I implemented fragment parsing? hard to believe

[07:14:03.0000] <jgraham>
hsivonen: I seem to remember looking at this at some point. But I don't know if I caused a note to appear

[07:15:00.0000] <annevk>
jgraham, how is c mutated there?

[07:16:00.0000] <annevk>
oh you move c forward

[07:17:00.0000] <hsivonen>
Sadly, Atom ended up like this: http://www.xkcd.com/927/

[07:17:01.0000] <annevk>
jgraham, depending on direction it should be either b or e

[07:17:02.0000] <hsivonen>
/me checks the www-tag archive to see if the xkcd URL has already been posted there

[07:18:00.0000] <hsivonen>
not according to list search

[07:18:01.0000] <annevk>
nope

[07:18:02.0000] <annevk>
you're not subscribed?

[07:18:03.0000] <hsivonen>
annevk: I'm not

[07:23:00.0000] <hsivonen>
annevk: maybe I should subscribe, though. I have a number of other W3C list subscribed and going into a bucket that I don't read actively but occasionally browse and respond to

[07:23:01.0000] <annevk>
jgraham, i would love a better way to have that described though

[07:23:02.0000] <hsivonen>
(in the case of www-tag, I read the archives and then copy and paste when replying)

[07:24:00.0000] <hsivonen>
(I probably end up reading a smaller propertion of www-style messages than www-tag messages even though I subscribe to www-style)

[07:26:00.0000] <jgraham>
annevk: Yeah, it is not quite trivial. I guess I might try talking about the state of the collection before and after the mutation, with some names for the set of elements that is moved during the mutation, the set that is deleted, the set that is inserted, and the rest

[07:27:00.0000] <jarek_>
does CSS3 allow comments between "!" and "important"?

[07:27:01.0000] <jarek_>
according to specification it's not allowed, but for some reason it does validate on http://jigsaw.w3.org/

[07:28:00.0000] <annevk>
jgraham, I guess the only problem is removal of the referenceNode

[07:28:01.0000] <annevk>
if the referenceNode is removed by itself you simply need the prev or next node

[07:29:00.0000] <annevk>
if it's an ancestor you need the ancestor prev or next node

[07:29:01.0000] <annevk>
not sure what else is there

[07:29:02.0000] <annevk>
replaceChild I guess

[07:30:00.0000] <annevk>
should prolly treat that as removal as well

[07:40:00.0000] <annevk>
jgraham, http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#iterator-collection

[07:41:00.0000] <Ms2ger>
jarek_, does the spec say it explicitly? You can usually put comments anywhere between tokens

[07:41:01.0000] <Ms2ger>
(Assuming ! and important are two tokens)

[07:42:00.0000] <jgraham>
annevk: Saying "referenceNode attribute value" everywhere rather than "referenceNode" is bad

[07:42:01.0000] <jarek_>
Ms2ger: CSS3 Syntax module defines IMPORTANT_SYM token as "!{w}important"

[07:42:02.0000] <jarek_>
CSS2 defines it as: "!"({w}|{comment})*{I}{M}{P}{O}{R}{T}{A}{N}{T}

[07:42:03.0000] <jgraham>
It is confusing because of DOM attributes

[07:43:00.0000] <jgraham>
and verbose

[07:43:01.0000] <Ms2ger>
-syntax is irrelevant

[07:43:02.0000] <Ms2ger>
As was said earlier

[07:43:03.0000] <annevk>
jgraham, that style is rather consistently used

[07:43:04.0000] <annevk>
jgraham, and besides the point

[07:43:05.0000] <jgraham>
annevk: OK, that style rather consistently sucks :)

[07:44:00.0000] <annevk>
yeah whatever

[07:46:00.0000] <annevk>
pretty pleased with this new definition

[07:47:00.0000] <jgraham>
annevk: It is still not clear what the "node preceeeding the node that was removed" is

[07:48:00.0000] <jgraham>
A node that isn't in the collection doesn't have a node preceding it

[07:50:00.0000] <annevk>
jgraham, if it isn't in the collection it would already terminate

[07:50:01.0000] <annevk>
jgraham, again, see step 1 of the algorithm

[07:51:00.0000] <jgraham>
annevk: Hmm? If I remove a node from a collection, how can it also be in the collection?

[07:51:01.0000] <jgraham>
It *was* in the collection before it was remove

[07:51:02.0000] <jgraham>
d

[07:52:00.0000] <annevk>
oh right

[07:52:01.0000] <annevk>
this needs to run before they are actually removed

[07:53:00.0000] <jgraham>
OK. It should also be clear what the :mutated node" is. Especially since the algorithm is run "whenever nodes [plural] are removed from the collection"

[07:53:01.0000] <annevk>
jgraham, replacing "was" with "is being" helps you?

[07:54:00.0000] <annevk>
<p>Whenever a <span title=concept-node>node</span> is removed from the

[07:54:01.0000] <annevk>
<span>iterator collection</span>, these steps must be run:

[07:54:02.0000] <annevk>
and that?

[07:54:03.0000] <hsivonen>
hmm. it's an unusual situation that Sam's latest blog post is well over a month old

[07:54:04.0000] <annevk>
add being after is again

[07:55:00.0000] <jgraham>
I thought you wanted to add :before"?

[07:55:01.0000] <jgraham>
s/:/"/

[07:56:00.0000] <annevk>
oh

[07:56:01.0000] <annevk>
like

[07:56:02.0000] <annevk>
<p>Before a <span title=concept-node>node</span> is removed from the

[07:56:03.0000] <annevk>
<span>iterator collection</span>, these steps must be run:

[07:56:04.0000] <annevk>
could do that too

[07:57:00.0000] <annevk>
refresh

[08:00:00.0000] <jgraham>
annevk: Well I can't be sure it is right, but at least I understand it now :)

[08:01:00.0000] <annevk>
yeah, needs more tests and stuff

[08:06:00.0000] <danj>
/me Crys over Vagrant

[08:07:00.0000] <annevk>
TreeWalker.root is fairly meaningless, no?

[08:07:01.0000] <annevk>
I wonder why that exists

[08:07:02.0000] <annevk>
oh never mind

[08:07:03.0000] <annevk>
although

[08:07:04.0000] <annevk>
"

[08:07:05.0000] <annevk>
Moves to and returns the closest visible ancestor node of the current node. If the search for parentNode attempts to step upward from the TreeWalker's root node, or if it fails to find a visible ancestor node, this method retains the current position and returns null."

[08:07:06.0000] <annevk>
does not seem to account for currentNode being outside the TreeWalker root

[08:23:00.0000] <annevk>
http://mxr.mozilla.org/mozilla-central/source/content/base/src/nsTreeWalker.cpp#143 I wonder why that security check is there

[08:33:00.0000] <Ms2ger>
https://bugzilla.mozilla.org/show_bug.cgi?id=156452

[08:34:00.0000] <Ms2ger>
if you hadn't found it yet

[08:41:00.0000] <annevk>
hmm

[08:50:00.0000] <smaug____>
that is really old code

[08:50:01.0000] <gsnedders>
Motherboard with processor/memory given ATX2 power cables plugged in should turn on it's power-on LED, right?

[08:50:02.0000] <smaug____>
and as far as I know, not needed anymore

[09:34:00.0000] <dglazkov>
good morning, Whatwg!

[09:38:00.0000] <AryehGregor>
Good morning, dglazkov!

[09:41:00.0000] <annevk>
I have the feeling scrolling is now the opposite direction in OS X Lion

[09:41:01.0000] <annevk>
That is somewhat confusing

[09:41:02.0000] <zewt>
heh people complaining about that simultaneously in two channels i'm in

[09:43:00.0000] <AryehGregor>
What does "the opposite direction" mean?

[09:44:00.0000] <annevk>
fingers up, content goes up

[09:45:00.0000] <annevk>
well, move fingers from top to bottom and you scroll upwards

[09:45:01.0000] <annevk>
used to be the other way around

[09:46:00.0000] <zewt>
it says a lot about how much a company cares about their users when they're willing to destroy users' hard-baked UI habits and muscle memory approaching a decade and a half of use

[09:46:01.0000] <AryehGregor>
That's the way touchscreens all work, right?

[09:47:00.0000] <AryehGregor>
The idea being that you're sort of pushing the content down while the screen stays in place.

[09:47:01.0000] <jgraham>
That sounds awful

[09:47:02.0000] <annevk>
AryehGregor, yeah, it's like on the iPhone

[09:47:03.0000] <AryehGregor>
As though it's printed on a little piece of paper stuck behind the screen that you're physically pushing.

[09:47:04.0000] <annevk>
except there is no screen under my fingers

[09:50:00.0000] <Philip`>
/me presumes it's futile to ask if the scroll direction is a configurable option

[09:51:00.0000] <annevk>
you can configure it

[09:52:00.0000] <Philip`>
Oh, that's fortunate

[09:52:01.0000] <annevk>
it's a checkbox called "scroll direction: natural\n content tracks finger movement"

[09:55:00.0000] <annevk>
I guess I'll try to get used to it

[09:55:01.0000] <zewt>
and then good luck using any other system, heh

[09:56:00.0000] <annevk>
oh, Safari has two finger sideways scrolling for history

[09:56:01.0000] <annevk>
quite neat

[09:57:00.0000] <annevk>
zewt, they already trip me up with the keys

[09:57:01.0000] <zewt>
i use mousewheel tilt for that

[09:57:02.0000] <annevk>
the way it is animated is quite nice

[09:57:03.0000] <zewt>
annevk: heh i had to use windows in a vm on a macbook for a week or so once, and the buttons that were mapped to alt and windows key were flipped

[09:57:04.0000] <zewt>
drove me absolutely insane for a few days, then as I was getting used to it I went home and was back to a regular keyboard, where I could go insane again

[09:58:00.0000] <annevk>
computers are terrible

[10:05:00.0000] <swarren08>
annevk you can change the way it scrolls in the new Lion

[10:06:00.0000] <swarren08>
under trackpad / scroll & zoom , uncheck the scroll directional

[10:09:00.0000] <annevk>
I know

[10:10:00.0000] <swarren08>
cool. So what do you think of it?

[10:12:00.0000] <annevk>
dunno, I mostly went back to work on DOM Traversal

[10:12:01.0000] <annevk>
looks a bit neater though, and being able to play albums directly from the screen saver is a nice touch

[10:12:02.0000] <zewt>
(except now all node traversal will be backwards)

[10:14:00.0000] <swarren08>
I cant wait to get myself a mac

[10:14:01.0000] <annevk>
can't wait for competition to catch up personally

[10:14:02.0000] <annevk>
but it seems like that might be a while

[10:15:00.0000] <swarren08>
that is true

[10:16:00.0000] <swarren08>
apple seems to come up with this technology faster than others, which is great

[10:16:01.0000] <swarren08>
obviously it is helping them out immensly. Did anyone happen to see the stats on the recent quarter for apple? it was definitly insane to see

[10:26:00.0000] <annevk>
the Gecko and WebKit TreeWalker implementations look remarkably similar

[10:27:00.0000] <Ms2ger>
How do they compare to Opera's?

[10:28:00.0000] <annevk>
dunno and could not tell you either

[10:46:00.0000] <annevk>
writing everything down is a little more complicated than I initially thought

[10:47:00.0000] <annevk>
though I guess three-five days is not that bad

[10:47:01.0000] <annevk>
events was just simpler

[10:50:00.0000] <Hixie>
hsivonen: no idea to what you are referring, but i'm happy to add notes to the spec, drop me a mail or a bug

[10:51:00.0000] <hober>
annevk: you'll get used to the new scrolling in a couple of days, and then you won't want to go back.

[10:55:00.0000] <jgraham>
hober: Spoken like a good mini-Steve

[10:56:00.0000] <jgraham>
:)

[10:56:01.0000] <hober>
jgraham: :)

[11:04:00.0000] <AryehGregor>
Hixie, you resolved <http://www.w3.org/Bugs/Public/show_bug.cgi?id=12570> in r6314.  You should close the bug as FIXED.

[11:05:00.0000] <AryehGregor>
(I don't know why we have an informative reference to PDF but not the zillions of other standards mentioned in non-normative text, though)

[11:10:00.0000] <Hixie>
looks like it's already marked fixed

[11:10:01.0000] <Hixie>
the pdf one is mentioned in normative text

[11:10:02.0000] <Hixie>
or at least, in main prose, as opposed to notes and examples

[11:10:03.0000] <Hixie>
if there's anything else you think should get referneces, let me know

[11:23:00.0000] <jarek_>
CSS3 selectors specification defines NOT token as ":"{N}{O}{T}"(" while CSS3-mediaqueries specification defines a token with the same name as {N}{O}{T}

[11:24:00.0000] <TabAtkins_>
Yes, it's inconsistent.  Been pointed out before, hasn't been fixe dyet.

[11:24:01.0000] <AryehGregor>
Oh, Ms2ger marked it FIXED even though technically only the editor is supposed to.  Oh well, I doubt anyone will care.

[11:25:00.0000] <AryehGregor>
Hixie, I didn't see any reference to PDF in normative text.  It's mentioned in the prose, but only preceded by "e.g." or "for example".

[11:25:01.0000] <Ms2ger>
Meh, I assume the guy won't try to process-troll me over getting his bug fixed

[11:25:02.0000] <AryehGregor>
Well, that's why it's an informative reference, of course.

[11:25:03.0000] <jarek_>
perhaps it would be easier if I had written two tokenizers - one for selectors and another one for everything else (the second parser would call the first one)

[11:25:04.0000] <jarek_>
CSS3 selectors are pretty complex

[11:25:05.0000] <AryehGregor>
But I don't see the difference between entire paragraphs marked as examples or notes, and sentences within normative text marked as examples or notes.

[11:26:00.0000] <AryehGregor>
Still, it makes no difference in the end.

[11:26:01.0000] <AryehGregor>
As I said in the bug, examples of other formats mentioned in examples include H.264, AAC, MP4, MPEG-4,

[11:26:02.0000] <AryehGregor>
AMR, 3GPP, Theora, Vorbis, Ogg, Speex, FLAC, and Dirac, and that's without even trying to find them.

[11:43:00.0000] <MikeSmith>
AryehGregor: yep

[11:44:00.0000] <MikeSmith>
dude who filed that bug is basically trolling, actually

[11:45:00.0000] <MikeSmith>
that seems to be his whole M.O., a

[11:45:01.0000] <Hixie>
AryehGregor: i'm happy to include more references

[11:45:02.0000] <Hixie>
AryehGregor: file a bug or send mail letting me know what you want references for

[11:46:00.0000] <MikeSmith>
Hixie: well, don't feed the trolls man

[11:47:00.0000] <MikeSmith>
I don't think AryehGregor is suggesting more references

[11:49:00.0000] <Hixie>
i'm all for adding more references

[11:49:01.0000] <MikeSmith>
yeah, well

[11:50:00.0000] <MikeSmith>
the point is whether they are actually useful

[11:50:01.0000] <MikeSmith>
as opposed to gratuitous

[11:50:02.0000] <Hixie>
we are _long_ past the point of gratuitous

[11:51:00.0000] <Ms2ger>
Isn't adding gratuitous references a stated goal of the HTMLWG?

[11:52:00.0000] <MikeSmith>
further gratuoisity proliferation us not a win

[11:52:01.0000] <MikeSmith>
or however you spell that, christ

[11:53:00.0000] <Hixie>
i don't think the references make any difference whatsoever :-)

[11:53:01.0000] <MikeSmith>
Ms2ger: good example of trolling-trolling

[11:53:02.0000] <MikeSmith>
clever

[11:54:00.0000] <Ms2ger>
/me likes trolling trolls

[11:54:01.0000] <MikeSmith>
Hixie: good, remove all of them, then

[11:55:00.0000] <Ms2ger>
MikeSmith, agreed, and we should import all other web specs into WA1.0 at the same time :)

[11:55:01.0000] <MikeSmith>
yeah, let's make one giant monolithic spec

[11:56:00.0000] <MikeSmith>
that is clearly a great thing

[11:56:01.0000] <Ms2ger>
It worked for CSS!

[11:57:00.0000] <Hixie>
i'd much rather we had one spec than 100

[11:58:00.0000] <jcranmer>
maybe here's a better place

[11:58:01.0000] <jcranmer>
is there a nice way in CSS to get a height that's the same as the line-height?

[11:59:00.0000] <TabAtkins_>
No.

[11:59:01.0000] <TabAtkins_>
(Unless the line-height is the same as the font-size.)

[11:59:02.0000] <jcranmer>
so my best bet is to set the line-height myself and hope for the best?

[12:01:00.0000] <MikeSmith_>
Hixie: me too

[12:01:01.0000] <MikeSmith_>
honestly

[12:02:00.0000] <MikeSmith>
but some delusional people seem to prefer specs that are less than 5MB

[12:03:00.0000] <Hixie>
i don't mind having a multipage copy

[12:03:01.0000] <Hixie>
but if the concern is just having lots of pages to read, you get fewer total pages if you have 1 spec than 100

[12:04:00.0000] <Hixie>
since each spec has a lot of duplicate boilerplate, cross-spec stuff, etc

[12:04:01.0000] <MikeSmith>
true

[12:05:00.0000] <MikeSmith>
I think CSS as gone to the other extreme

[12:05:01.0000] <Ms2ger>
Splitting specs is fine, IMO, if you can find good split-lines

[12:05:02.0000] <Ms2ger>
These are almost non-existent, unfortunately

[12:05:03.0000] <TabAtkins_>
jcranmer: yes.

[12:06:00.0000] <TabAtkins_>
Ms2ger: That's not true.  For example, different layout algos for CSS can split apart pretty easily.

[12:06:01.0000] <Ms2ger>
I was talking WA1.0 mostly

[12:06:02.0000] <Hixie>
CSS has a number of problems, dunno if splitting is a problem itself there or a solution that's just been swamped by other issues

[12:06:03.0000] <Hixie>
CSS definitely has a "velocity" problem

[12:07:00.0000] <Ms2ger>
Also, editors

[12:07:01.0000] <Hixie>
but i think that's more an issue of the editing style

[12:07:02.0000] <TabAtkins_>
Editors are our largest problem.  Secondary is that we're bound by W3C process, which made 2.1 a glorious mess.

[12:07:03.0000] <Hixie>
yeah

[12:08:00.0000] <Hixie>
i've been speaking with jeff and i have no faith that he's interesting in changing the process

[12:09:00.0000] <jcranmer>
hmm

[12:10:00.0000] <jcranmer>
9pt/13px

[12:10:01.0000] <TabAtkins_>
We should add CSS unit conversion to google calc.

[12:12:00.0000] <jcranmer>
so 9pt = 12px

[12:13:00.0000] <jcranmer>
how about I call it 0.9em and everyone's happy?

[12:13:01.0000] <AryehGregor>
MikeSmith, Hixie: Personally I don't include any references at all in the edit commands spec.  The spec is already going to be horribly broken if you try doing something stupid like printing it out, since all the xrefs will be broken, so I'm just linking to sections of other specs directly when I want to reference something.

[12:13:02.0000] <beverloo>
TabAtkins:  1vm = ?, 1% = ?. Too much depends on the context imo

[12:13:03.0000] <AryehGregor>
beverloo, do it for absolute units only.

[12:13:04.0000] <TabAtkins_>
beverloo: Yeah, only the physical units I mean.

[12:13:05.0000] <jcranmer>
pt, pc, px, in, cm

[12:13:06.0000] <TabAtkins_>
Converting between in, pt, px

[12:14:00.0000] <jcranmer>
mm

[12:14:01.0000] <beverloo>
How does that handle dpi?

[12:14:02.0000] <TabAtkins_>
The ratio is fixed in CSS.

[12:14:03.0000] <TabAtkins_>
1in = 96px

[12:14:04.0000] <jcranmer>
the ratios is essentially fixed

[12:14:05.0000] <jcranmer>
it's way too broken otherwise

[12:14:06.0000] <TabAtkins_>
Not essentially, actually.  ^_^

[12:14:07.0000] <beverloo>
Fair enough

[12:15:00.0000] <jcranmer>
px: pixel units — 1px is equal to 0.75pt.

[12:15:01.0000] <jcranmer>
ah, it is fixed

[12:15:02.0000] <AryehGregor>
As of like a year ago.

[12:15:03.0000] <AryehGregor>
Before that it was unofficially not fixed but mostly fixed in practice.

[12:15:04.0000] <TabAtkins_>
Was it wrong before, jcranmer?

[12:15:05.0000] <jcranmer>
I thought it was only in CSS3

[12:15:06.0000] <TabAtkins_>
Oh, the px/in thing.  Yeah.

[12:15:07.0000] <AryehGregor>
I think he means "fixed" as opposed to "variable", not as opposed to "broken".

[12:15:08.0000] <TabAtkins_>
We set it when we were resolving 2.1 issues.

[12:16:00.0000] <jcranmer>
I should say

[12:16:01.0000] <jcranmer>
people assume 96dpi in too many pages, that to make any other value would break the web

[12:16:02.0000] <jcranmer>
that's what I meant by "it's way too broken otherwise"

[12:16:03.0000] <TabAtkins_>
Yes, exactly.

[12:19:00.0000] <TabAtkins_>
Huh.  I never realized that I had to actually put the </script> for the script to execute.

[12:19:01.0000] <TabAtkins_>
(I'm testing some interfaces before I write the real page.)

[12:20:00.0000] <AryehGregor>
I only realized that when I started using Live DOM viewer.

[12:32:00.0000] <hsivonen>
Hixie: it turned out that the note I was asking for already exists

[12:36:00.0000] <hsivonen>
is there something genuinely relative unusual about Swedish names that I'm not aware of or is this person just throwing an arbitrary collection of country names out there for the sake of illustration? http://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/

[12:38:00.0000] <Hixie>
hsivonen: excellent :-)

[12:39:00.0000] <swarren08>
Whats the best way to learn HTML5. Im going to be honest, watching everyone talk makes me feel like i know nothing.

[12:41:00.0000] <Philip`>
Spend a year reading the spec and the mailing lists and IRC :-)

[12:42:00.0000] <hsivonen>
swarren08: people on this channel develop HTML5 itself and may, therefore, not have a good idea of what's it like to learn for others

[12:42:01.0000] <hsivonen>
swarren08: but what Philip` says should work

[12:42:02.0000] <swarren08>
its amazing watching you guys

[12:43:00.0000] <jarek_>
swarren08: I would from https://developer.mozilla.org

[12:43:01.0000] <jgraham>
hsivonen: Maybe the way that Swedish sorts characters being different to e.g. German?

[12:43:02.0000] <jarek_>
s/I would/I would start

[12:43:03.0000] <jgraham>
Philip`: You are mean :)

[12:44:00.0000] <swarren08>
So you guys actually develop HTML5? or what you call the Living Standard?

[12:44:01.0000] <jgraham>
hsivonen: Although that isn't really about names. I'm not aware of anything common and strange about Swedish names

[12:45:00.0000] <jgraham>
swarren08: Nah Hixie does that. We just hold the palm leaves and feed him grapes

[12:45:01.0000] <AryehGregor>
swarren08, yes.

[12:45:02.0000] <AryehGregor>
Also, what jgraham says.

[12:45:03.0000] <Hixie>
wait, i get grapes?!

[12:45:04.0000] <Hixie>
where is this!

[12:45:05.0000] <swarren08>
ha ha ha

[12:45:06.0000] <AryehGregor>
We're really all Hixie's minions, he does all the work.

[12:45:07.0000] <AryehGregor>
We just do menial labor for him and take some of the credit.

[12:45:08.0000] <swarren08>
Well I do thank you guys that do work on it

[12:46:00.0000] <hsivonen>
AryehGregor: but only you are contractually Hixie's minion :-)

[12:46:01.0000] <Hixie>
hah

[12:46:02.0000] <AryehGregor>
Yes, that's true.

[12:46:03.0000] <Hixie>
i dunno if i'd call what y'all do "menial"

[12:46:04.0000] <AryehGregor>
Other people are his minions for fun.

[12:46:05.0000] <TabAtkins_>
"fun"

[12:46:06.0000] <AryehGregor>
I'm heartless and only do minion work for money.

[12:46:07.0000] <Hixie>
"indispensible" maybe

[12:46:08.0000] <swarren08>
If you guys dont mind, id like to hang around in this channel to watch you guys

[12:47:00.0000] <AryehGregor>
swarren08, go ahead, that's how lots of people get started in standards work.

[12:47:01.0000] <Ms2ger>
You're welcome

[12:47:02.0000] <Hixie>
swarren08: you are welcome to hang around :-)

[12:47:03.0000] <AryehGregor>
Ask any questions you have.

[12:47:04.0000] <Ms2ger>
Within a year or two, you'll be a minion as well :)

[12:47:05.0000] <AryehGregor>
Like, it's how I got started, and TabAtkins_, and others too.

[12:47:06.0000] <swarren08>
Ive been building websites for years now

[12:47:07.0000] <Hixie>
swarren08: if you hang around long enough, you might find yourself being useful, that's how pretty much all of us started :-)

[12:47:08.0000] <wilhelm>
Just remember to leave your sense of logic at the door, please.

[12:47:09.0000] <swarren08>
I finally decided to take a look into HTML5 and CSS3 since its going to be standard soon

[12:47:10.0000] <swarren08>
Logic? whats that? :)

[12:48:00.0000] <Ms2ger>
People who write websites? These really exist? :)

[12:48:01.0000] <swarren08>
apperently im a dieing race now

[12:48:02.0000] <swarren08>
lol

[12:48:03.0000] <swarren08>
i was tought using notepad

[12:48:04.0000] <jgraham>
I'm not sure. As far as I cn tell the world is full of people who write web browsers

[12:48:05.0000] <jgraham>
I though the websites just formed out of the aether

[12:48:06.0000] <swarren08>
taught*

[12:48:07.0000] <Ms2ger>
That's my experience as well

[12:49:00.0000] <Ms2ger>
And fishy aether, usually

[12:49:01.0000] <swarren08>
How long have you all been doing this? as in working on the standard?

[12:50:00.0000] <jgraham>
Well the WHATWG started in 2004 or so

[12:50:01.0000] <swarren08>
yea after the conference

[12:51:00.0000] <jgraham>
Of course some people date from before that

[12:51:01.0000] <AryehGregor>
Some people here have been working on browsers since the Netscape days.

[12:51:02.0000] <Hixie>
i've been working on specs since late 1998, iirc

[12:51:03.0000] <Hixie>
if you could posting on mailing lists "working on specs"

[12:51:04.0000] <swarren08>
Wow

[12:52:00.0000] <Ms2ger>
But Hixie's just old

[12:52:01.0000] <Hixie>
it's true

[12:52:02.0000] <swarren08>
I feel like i may be the youngest one here

[12:52:03.0000] <jgraham>
swarren08: Well if your user name is from your year of birth

[12:52:04.0000] <hober>
swarren08: probably not, unless your nick means you were born in 2008. good work on typing & language acquisition, if so

[12:52:05.0000] <swarren08>
lol

[12:53:00.0000] <swarren08>
actually its my graduating year of high school

[12:53:01.0000] <swarren08>
ive had that sn since about the 6th grade

[12:53:02.0000] <AryehGregor>
We have people pretty close to your age here.

[12:53:03.0000] <AryehGregor>
I graduated high school in 2005.  There are at least a couple people here younger than me.

[12:53:04.0000] <swarren08>
Wow

[12:54:00.0000] <jgraham>
yeah, gsnedders is about 3 or 4 now

[12:54:01.0000] <jgraham>
(OK, more like 19)

[12:54:02.0000] <swarren08>
and you work on the standards also AryehGregor?

[12:54:03.0000] <Ms2ger>
And it's still jgraham who has the infantile humour? :)

[12:54:04.0000] <AryehGregor>
swarren08, yes, currently contracting for Google to write this: http://aryeh.name/spec/editcommands/editcommands.html

[12:55:00.0000] <jgraham>
Ms2ger: Not to mention thee grey hair :)

[12:55:01.0000] <swarren08>
woa ehy that is you, i was looking at that yesterday

[12:55:02.0000] <AryehGregor>
I'm pretty sure the 90th percentile age among regulars in this room is well under 30.

[12:55:03.0000] <swarren08>
Thats awesome

[12:56:00.0000] <AryehGregor>
But really, nobody cares.  You're judged on your contributions, not your age.

[12:56:01.0000] <Hixie>
(i'm 31, was 18 or 19 or so when i started posting to the lists)

[12:56:02.0000] <jgraham>
Hmm, I wonder if that is actually true

[12:56:03.0000] <swarren08>
I bet you all have or are in college

[12:56:04.0000] <Ms2ger>
...Or your degrees

[12:56:05.0000] <swarren08>
-have + have been in

[12:57:00.0000] <AryehGregor>
Right, nobody cares much about degrees either.  I mean, most of the people here have or are pursuing degrees in some technical field, in some cases Ph.D.'s, but it doesn't really matter.

[12:57:01.0000] <AryehGregor>
We just get to make fun of the Ph.D.'s by calling them "Dr." if we want to annoy them.

[12:57:02.0000] <Ms2ger>
Except for Hixie, who's happy with his BSc

[12:57:03.0000] <swarren08>
ah

[12:57:04.0000] <TabAtkins_>
Sure, but my college education has very little effect on my standards contribution.

[12:57:05.0000] <AryehGregor>
Who's the Opera person with a Ph.D. who doesn't like being called Dr.?  jgraham?

[12:58:00.0000] <TabAtkins_>
Yeah, jgraham

[12:58:01.0000] <jgraham>
/me hides

[12:58:02.0000] <jgraham>
AryehGregor: A glance at the stats page suggests that more than 10% of the top 25 people might well be over 30

[12:58:03.0000] <TabAtkins_>
On that note, who's got a math degree and wants to help me implement Coons patch mesh gradients?

[12:58:04.0000] <AryehGregor>
othermaciej also has a Ph.D., but reasonable people would actually believe that based on his behavior, unlike jgraham.

[12:58:05.0000] <TabAtkins_>
I got them working in squares, but bezier-curve sides are a bit more difficult...

[12:59:00.0000] <swarren08>
I know Apple, Opera and Mozilla are the main companies who were behind the Living Standard, is thier others?

[12:59:01.0000] <Ms2ger>
Maciej has a PhD? Really?

[12:59:02.0000] <Ms2ger>
Google

[12:59:03.0000] <swarren08>
Forgot Google

[12:59:04.0000] <AryehGregor>
TabAtkins_, I have a math degree, but it's pure math, so I get to contemptuously reply that such drudgework has nothing to do with the kind of math I studied (i.e., *real* math).

[12:59:05.0000] <TabAtkins_>
AryehGregor: Nuts to your real math.

[12:59:06.0000] <Ms2ger>
AryehGregor++

[12:59:07.0000] <TabAtkins_>
I need a Math BS here.

[13:00:00.0000] <AryehGregor>
Oh, I'm wrong.

[13:00:01.0000] <AryehGregor>
Only a master's.

[13:00:02.0000] <AryehGregor>
His sister has a Ph.D.

[13:00:03.0000] <AryehGregor>
His master's is from MIT, though.

[13:00:04.0000] <swarren08>
dang

[13:00:05.0000] <AryehGregor>
TabAtkins_, I have a math BS.

[13:00:06.0000] <jamesr>
MEng?

[13:00:07.0000] <AryehGregor>
It's just a pure math Bs.

[13:00:08.0000] <AryehGregor>
BS.

[13:00:09.0000] <TabAtkins_>
Also: omigod it's so frustrating being locked out of email.  I have so many threads to respond to, including one where I now have a working counterexample to someone's statement.

[13:00:10.0000] <AryehGregor>
jamesr, http://en.wikipedia.org/wiki/Maciej_Stachowiak

[13:00:11.0000] <swarren08>
you guys are smart people

[13:00:12.0000] <Hixie>
that's debatable

[13:00:13.0000] <jgraham>
BS is the worst name for a type of degree

[13:00:14.0000] <TabAtkins_>
BS

[13:01:00.0000] <Ms2ger>
Hixie is

[13:01:01.0000] <Hixie>
and indeed frequently debated!

[13:01:02.0000] <AryehGregor>
swarren08, nobody cares about the degrees.  They only care that you're smart.  Of course, smart people often get advanced degrees.

[13:01:03.0000] <Ms2ger>
AryehGregor, ... or they do something useful :)

[13:01:04.0000] <AryehGregor>
I said "often".

[13:01:05.0000] <jgraham>
And often people get advanced degrees without being unusually smart

[13:01:06.0000] <AryehGregor>
Also, the two are not mutually exclusive.

[13:01:07.0000] <Ms2ger>
/me went the degree route

[13:02:00.0000] <AryehGregor>
Well, yes, but smart people are more likely to get advanced degrees.  Which is 90% of the reason anyone bothers, to signal that they're likely to be smarter than average.

[13:02:01.0000] <jgraham>
Hmm, I did it because it was fun

[13:02:02.0000] <AryehGregor>
Yeah, people do that too.

[13:02:03.0000] <swarren08>
Does anyone actually work for Google, Apple, Mozilla or Opera? that would be pretty awesome if someone does

[13:02:04.0000] <AryehGregor>
I'd have gone on to get a Ph.D. if it was fun, probably.

[13:02:05.0000] <AryehGregor>
swarren08, basically everyone you're talking to works for one of those.

[13:02:06.0000] <swarren08>
Really?

[13:02:07.0000] <Ms2ger>
/me doesn't

[13:03:00.0000] <swarren08>
Wow

[13:03:01.0000] <Hixie>
Ms2ger: how do i know which interface should be marked 'partial'?

[13:03:02.0000] <AryehGregor>
Hixie, TabAtkins_, and I are Google.  Ms2ger volunteers for Mozilla, although apparently doesn't work for them.  jgraham is Opera.

[13:03:03.0000] <TabAtkins_>
Ms2ger: You contribute to Moz heavily enough that it count.

[13:03:04.0000] <jgraham>
But one of the brightest, most productive people I work with doesn't have a degree so there are certainly exceptions

[13:03:05.0000] <swarren08>
Thats is amazing

[13:03:06.0000] <AryehGregor>
jamesr is Google, hober is Apple.

[13:03:07.0000] <Ms2ger>
Hixie, the one in the obsolete section? :)

[13:03:08.0000] <kbrosnan>
roc dbarron mozilla

[13:03:09.0000] <Ms2ger>
wilhelm is Opera

[13:03:10.0000] <TabAtkins_>
swarren08: Though Aryeh and I were working on this stuff before we were hired by Google.

[13:03:11.0000] <Hixie>
Ms2ger: oh, i got confused

[13:03:12.0000] <swarren08>
Ive always wanted to talk to people who worked for those companies

[13:03:13.0000] <TabAtkins_>
And were in fact hired by Google *because* of the standards work we o.

[13:03:14.0000] <jamesr>
anyone know what "kwac" or "Obigo X10" are? (i'm looking at the view-mode media feature conformance report from public-webapps)

[13:04:00.0000] <AryehGregor>
Etc.

[13:04:01.0000] <gsnedders>
jgraham: I don't think much stuff in general applies to jl, though.

[13:04:02.0000] <Hixie>
Ms2ger: i saw you put it on HTMLElement and forgot that we had a duplicate of that in the microdata section for w3c's nonsense

[13:04:03.0000] <jgraham>
Most people were working on his stuf before they were hired to work on it

[13:04:04.0000] <Ms2ger>
You didn't notice the comment about that in the context? :)

[13:04:05.0000] <jgraham>
gsnedders: If you decapitate him he will die?

[13:04:06.0000] <wilhelm>
For seven years now. Shit, I'm getting old.

[13:04:07.0000] <Ms2ger>
Which reminds me of krijn's quote "What else do I need to do to get hired by Google"

[13:05:00.0000] <swarren08>
Here is an observation, ive only notice you guys speak of W3C 1 time in the past 20 minutes(maybe more)

[13:05:01.0000] <jgraham>
wilhelm: And yet you are still younger than me. Bastard

[13:05:02.0000] <Ms2ger>
We like not to think about the W3C

[13:05:03.0000] <swarren08>
What is the difference between what you guys do and what W3C does?

[13:05:04.0000] <Ms2ger>
And if we do, we rant

[13:05:05.0000] <gsnedders>
wilhelm: You're not the youngest person at Opera anymore! Though there again, we have younger interns than me this summer…

[13:05:06.0000] <swarren08>
im sorry if its a sore spot

[13:05:07.0000] <AryehGregor>
swarren08, roughly, we do the work and the W3C gets in the way.

[13:05:08.0000] <wilhelm>
jgraham: But I have the same level of education as Mr. swarren08. Can I borrow some of your diplomas?

[13:05:09.0000] <Ms2ger>
We write specs, the W3C publishes specs? :)

[13:05:10.0000] <Hixie>
we actually make progress, w3c releases press-releases claiming our progress as theirs

[13:05:11.0000] <Hixie>
wait, was that impolitick

[13:05:12.0000] <TabAtkins_>
swarren08: We're both standards bodies.  The WHATWG was formed originally because the W3C didn't want to work on HTML.

[13:05:13.0000] <Hixie>
crap

[13:05:14.0000] <smaug____>
it really doesn't matter too much where the standardization happens, if it happens

[13:06:00.0000] <swarren08>
yea XHTML is what W3C was interested in

[13:06:01.0000] <AryehGregor>
No, but it matters if the W3C distracts us with bureaucratic nonsense that interferes with our actual work.

[13:06:02.0000] <Ms2ger>
Hixie, I should quote that on your wikipedia page :)

[13:06:03.0000] <jgraham>
wilhelm: I'm not sure they will be of much help to you. There isn't a big market for astrophysicists

[13:06:04.0000] <wilhelm>
Oh.

[13:06:05.0000] <AryehGregor>
And knowingly publishes outdated and incorrect standards that it refuses to update for procedural reasons.

[13:06:06.0000] <AryehGregor>
Etc. etc. etc.

[13:06:07.0000] <TabAtkins_>
swarren08: Later, the W3C requested that HTML be developed in it again, so now it's developed jointly between the two.

[13:06:08.0000] <Hixie>
hey there's no k on the end of impolitic

[13:06:09.0000] <Hixie>
who knew

[13:06:10.0000] <TabAtkins_>
swarren08: With the caveats that other people have mentioned.

[13:07:00.0000] <AryehGregor>
Hixie, there is if you went to school in the 19th century.  How old did you say you are again?

[13:07:01.0000] <swarren08>
So you guys do the work and W3C takes credit?

[13:07:02.0000] <swarren08>
am i getting that right?

[13:07:03.0000] <Ms2ger>
You'll note that AryehGregor is rather anti-W3C

[13:07:04.0000] <Hixie>
i was in school in the 20th century!

[13:07:05.0000] <AryehGregor>
swarren08, we're being sarcastic and bitter.

[13:07:06.0000] <Hixie>
swarren08: the truth is less cynical than that

[13:07:07.0000] <swarren08>
Ah... i feel so dumb

[13:07:08.0000] <Hixie>
also more complicated

[13:07:09.0000] <wilhelm>
swarren08: Many here wear hats from several organizations.

[13:07:10.0000] <AryehGregor>
swarren08, the W3C is an organization, it doesn't do anything but provide some web hosting and things like that.  It doesn't employ the people who write or edit almost any of the actual standards.

[13:07:11.0000] <wilhelm>
Best tool for the job, &c.

[13:08:00.0000] <Hixie>
the w3c is an institution that provides support for standards development

[13:08:01.0000] <swarren08>
i had no idea the standards would have so much "drama" behind it

[13:08:02.0000] <Hixie>
the whatwg is essentially a mailing list

[13:08:03.0000] <AryehGregor>
Plus a wiki.

[13:08:04.0000] <jgraham>
swarren08: It's a bit like a soap opera

[13:08:05.0000] <Ms2ger>
Who described "working group" as "the intersection of web technology and religion"?

[13:08:06.0000] <AryehGregor>
The W3C imposes lots of requirements on how exactly standards are to be developed, as a condition of hosting them.

[13:08:07.0000] <swarren08>
ive begun to notice that jgraham

[13:08:08.0000] <Hixie>
we need the w3c because the w3c has managed to convince a large number of companies to grant licenses for their patents

[13:08:09.0000] <cygri>
Ms2ger: i thought that was RDF

[13:08:10.0000] <Ms2ger>
cygri++

[13:08:11.0000] <jgraham>
At christmas we will have a special episode where it will be revealed that Hixie is actually TimBL's lovechild

[13:08:12.0000] <AryehGregor>
In exchange they provide their brand name, patent licensing, etc.  Plus Microsoft only participates in the W3C, not the WHATWG.

[13:09:00.0000] <Ms2ger>
cygri, but RDF isn't webtech

[13:09:01.0000] <swarren08>
Ive noticed microsoft doesnt do much with HTML5

[13:09:02.0000] <jgraham>
Yes, the patent licesing is a big deal

[13:09:03.0000] <swarren08>
or at least didnt till IE9

[13:09:04.0000] <AryehGregor>
So we're forced to participate there for political reasons, which means we have to obey their rules, which we all think are stupid and a waste of everyone's time.

[13:09:05.0000] <Ms2ger>
They're a couple of years behind

[13:09:06.0000] <AryehGregor>
swarren08, Microsoft is implementing HTML5 like everyone else, and they're catching up at it.  They provide useful feedback on the things they implement, like other browsers.

[13:09:07.0000] <AryehGregor>
They just came somewhat late to the game.

[13:10:00.0000] <AryehGregor>
IE10 likely won't be far behind other browsers' HTML5 support.

[13:10:01.0000] <Hixie>
Ms2ger: what was the stuff in web workers you weren't sure about?

[13:10:02.0000] <Hixie>
it's not so much that the w3c's rules are stupid, so much as they are anachronistic

[13:10:03.0000] <Hixie>
they haven't evolved with the times

[13:10:04.0000] <Ms2ger>
DedicatedWorkerGlobalScope, SharedWorkerGlobalScope, AbstractWorker

[13:10:05.0000] <swarren08>
To be honest, i would have figured Microsoft would be more interested in this stuff since they have 40+% of the world using IE

[13:11:00.0000] <AryehGregor>
They still annoy us because they tend to make opaque institutional decisions based on internal discussions instead of having discussions individually on public lists, and they also refuse to acknowledge the WHATWG, which forces us to deal with it more than we'd like.

[13:11:01.0000] <AryehGregor>
But MS is definitely a part of the picture here.

[13:11:02.0000] <Hixie>
which is highly ironic because the w3c's technologies are largely responsible for moving humanity's culture in the way that its processes now need to be updated to take into account

[13:11:03.0000] <Hixie>
Ms2ger: k

[13:11:04.0000] <AryehGregor>
Yeah, it's kind of ridiculous when the W3C doesn't update to account for changes brought about by the success of the web.

[13:11:05.0000] <Ms2ger>
We don't really hate the W3C as much as we usually make it look

[13:12:00.0000] <Hixie>
Ms2ger: did you leave those as [Supplemental]

[13:12:01.0000] <Ms2ger>
I didn't touch them

[13:12:02.0000] <Hixie>
k

[13:12:03.0000] <Hixie>
cool

[13:12:04.0000] <Hixie>
thanks

[13:12:05.0000] <swarren08>
I definitly see where your coming from AryehGregor, ive wondered the same things at time. Wont it hurt them down the line

[13:12:06.0000] <Hixie>
woot, your patch applied cleanly!

[13:12:07.0000] <AryehGregor>
Like, what point is there in a spec maturity cycle that requires the spec to freeze for years before it's called complete when major browsers are all on release cycles from six weeks to a year?

[13:12:08.0000] <Hixie>
Ms2ger++

[13:13:00.0000] <Hixie>
/me waits for Ms2ger to rename himself Ms3ger

[13:13:01.0000] <Ms2ger>
Old joke, sorry :)

[13:13:02.0000] <Hixie>
:-P

[13:13:03.0000] <AryehGregor>
Hixie, surely "Ms2ger" evaluates to 0 and becomes 1?

[13:13:04.0000] <Hixie>
maybe i just get a type mismatch

[13:13:05.0000] <Ms2ger>
AryehGregor, :(

[13:13:06.0000] <Hixie>
or maybe Ms2ger overloaded the ++ operator

[13:13:07.0000] <AryehGregor>
No, NaN.

[13:14:00.0000] <AryehGregor>
In PHP it would be 1.

[13:14:01.0000] <Ms2ger>
operator++() { return this; }

[13:14:02.0000] <wilhelm>
swarren08: It already has.

[13:15:00.0000] <AryehGregor>
Yeah, the formation and success of the WHATWG put a serious dent in the W3C's preeminence as a web standards body . . .

[13:15:01.0000] <Hixie>
swarren08: eventually if the w3c doesn't adapt there's going to come a point where someone gets annoyed enough that they set up an alternative patent policy

[13:15:02.0000] <AryehGregor>
Anyway, I should get back to working.

[13:15:03.0000] <Hixie>
swarren08: at which point they're basically doomed

[13:15:04.0000] <Hixie>
/me has been trying really hard to convince them to change, without any suggestion of success

[13:15:05.0000] <AryehGregor>
Hixie, you need to have implementers credibly threaten to abandon them.

[13:15:06.0000] <AryehGregor>
That's the only thing that will force them to change.

[13:16:00.0000] <Hixie>
that will happen once there's a patent policy, i expect

[13:16:01.0000] <AryehGregor>
Anything else will be voted down by the AC on reactionary grounds.

[13:16:02.0000] <swarren08>
Then why hasent someone done that? You all basically work for big companies and see where the web is going

[13:16:03.0000] <Ms2ger>
Or, the <dfn title=dom-Ms2ger-operator++>++</dfn> operator must return the <span>context object</span>

[13:16:04.0000] <swarren08>
You all should have more say for whats going on

[13:16:05.0000] <Hixie>
swarren08: it's a huge amount of really painful work

[13:16:06.0000] <AryehGregor>
In practice we do have all the say.

[13:16:07.0000] <AryehGregor>
The W3C annoys us, but doesn't interfere too much in the end.

[13:16:08.0000] <AryehGregor>
Just wastes our time.

[13:17:00.0000] <Hixie>
swarren08: so for someone to do it they have to be _really_ pushed over the edge

[13:17:01.0000] <jgraham>
swarren08: If we did that there wouldn't be anything to complain about

[13:17:02.0000] <Ms2ger>
IME, it's mainly the HTMLWG that wastes our time

[13:17:03.0000] <Hixie>
swarren08: i've gotten close, but not enough

[13:17:04.0000] <AryehGregor>
Yes, other WGs are much saner than the HTMLWG.

[13:17:05.0000] <Hixie>
Ms2ger: i'm getting my time wasted by plenty of other WGs too now

[13:17:06.0000] <swarren08>
so what your telling me

[13:17:07.0000] <Ms2ger>
WebApps wastes some, but not enough for me to whine

[13:17:08.0000] <jgraham>
and the peanut gallery couldn't use the yellow marker of horror to highlight pople saying cynical things

[13:17:09.0000] <Hixie>
webapps, webrtc, websockets (though that one was IETF, not W3C, i guess)

[13:17:10.0000] <swarren08>
is they bother you but yet not enough to get rid of?

[13:17:11.0000] <Ms2ger>
Yeah

[13:18:00.0000] <Hixie>
swarren08: pretty much

[13:18:01.0000] <swarren08>
Ah

[13:18:02.0000] <AryehGregor>
Hixie, what you need to do is lure someone into doing it who doesn't realize how much work it is.  You should start talking about how great it would be if someone just put in a little effort, while of course noting that you yourself already have far too many obligations to fit it into your schedule.

[13:18:03.0000] <swarren08>
I think I am finially understanding

[13:18:04.0000] <Hixie>
AryehGregor: i've been doing just that :-)

[13:18:05.0000] <hober>
AryehGregor: that sounds familiar somehow :)

[13:18:06.0000] <AryehGregor>
The W3C made enough concessions that it's not worth it for us to try too hard to pull out.

[13:18:07.0000] <Hixie>
AryehGregor: nobody has bitten enough to actually do it yet

[13:18:08.0000] <karlcow>
free specs? :) https://github.com/kobolabs/Kobo-Reader/blob/master/fickel/main.cpp#L320

[13:19:00.0000] <jgraham>
The idea of spending lots of time locked in a room wih lawyers is much less appealing than the idea of spending that same time thinking of a new cool feature for the web, or implementing part of the spec

[13:19:01.0000] <jgraham>
So guess which we do

[13:19:02.0000] <AryehGregor>
jgraham, that depends.  I like law, it's fun!

[13:19:03.0000] <Hixie>
swarren08: the other thing that means we're not hugely motivated to set up a patent policy is that we know we can always do it in the future if the w3c does go off the rails, so there's no urgency

[13:19:04.0000] <swarren08>
But then it comes back to the point where you guys to the work and W3C publishes the specs

[13:19:05.0000] <AryehGregor>
Probably related to my being Jewish.

[13:19:06.0000] <Hixie>
swarren08: most of the important specs are developed in a way that means we could fork them if we had to

[13:19:07.0000] <swarren08>
You all are playing the waiting game?

[13:20:00.0000] <jgraham>
No, we're doing fun stuff to improve the web

[13:20:01.0000] <AryehGregor>
swarren08, we could take the specs away from the W3C if we wanted to.  We edit them and we or our employers hold the copyright.  All we'd have to do is refuse to license new changes under the W3C's license.

[13:20:02.0000] <wilhelm>
More like the just-get-the-work-done-game.

[13:20:03.0000] <Hixie>
what wilhelm said

[13:20:04.0000] <swarren08>
oh

[13:20:05.0000] <AryehGregor>
But all this secession stuff would cause even more hassle.

[13:20:06.0000] <swarren08>
I see now

[13:20:07.0000] <Hixie>
heycam|away: ok so there's one thing i still need [Supplemental] for -- please ping me when you're around

[13:20:08.0000] <swarren08>
its just easier to let them play as they say

[13:21:00.0000] <Hixie>
it would cause a giant shitstorm if we were to actually fork

[13:21:01.0000] <AryehGregor>
Plus it would be perceived by the general public as "evil browser implementers try to take control of specs away from beloved W3C", since the W3C has a very positive reputation in most circles.

[13:21:02.0000] <Hixie>
dunno about "most" anymore

[13:21:03.0000] <Hixie>
maybe "many"

[13:21:04.0000] <jarek_>
is there an API that would allow me to detect how hard the user has pressed the stylus?

[13:21:05.0000] <swarren08>
I can only imagine Hixie

[13:21:06.0000] <AryehGregor>
There's some idea that it's open and inclusive and whatever, and not controlled by implementers, which is a joke, but there you have it.

[13:22:00.0000] <jarek_>
e.g. on tablet on touch sensitive device

[13:22:01.0000] <jarek_>
s/on/or

[13:22:02.0000] <Hixie>
i feel bad, i'm marking Ms2ger's huge patch as "editorial" for the tracker

[13:22:03.0000] <Ms2ger>
Hah

[13:22:04.0000] <AryehGregor>
What patch is this?

[13:22:05.0000] <Ms2ger>
That's fine

[13:22:06.0000] <Ms2ger>
Supplemental -> partial

[13:22:07.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13069

[13:23:00.0000] <AryehGregor>
Ah.

[13:23:01.0000] <Hixie>
Ms2ger: thanks a ton for this btw

[13:23:02.0000] <smaug____>
jarek_: touch.force

[13:23:03.0000] <swarren08>
Wow

[13:23:04.0000] <smaug____>
and some vendor specific things

[13:23:05.0000] <smaug____>
like mozPressure

[13:23:06.0000] <Ms2ger>
Happy to do something useful for once :)

[13:24:00.0000] <Hixie>
:-D

[13:24:01.0000] <jarek_>
smaug____: thanks, http://www.w3.org/TR/touch-events/ is exactly what I was thinking about

[13:24:02.0000] <Ms2ger>
Now I believe you owe me a fix for those DOM Core bugs? ;;)

[13:24:03.0000] <Hixie>
Ms2ger: aw man, i knew that was coming

[13:25:00.0000] <Ms2ger>
jarek_, http://dvcs.w3.org/hg/webevents/raw-file/default/touchevents.html, you mean

[13:25:01.0000] <Hixie>
Ms2ger: ok what do we need to do for this

[13:25:02.0000] <Ms2ger>
Hixie, I should form a WG to prioritize your work ;)

[13:25:03.0000] <jarek_>
I was recently playing with http://mrdoob.com/projects/harmony/#sketchy, this could be a killer app if it could support pressure sensitivity

[13:26:00.0000] <Hixie>
Ms2ger: is there some way we can diff the relevant parts of the specs to make sure we're not regressing anything?

[13:26:01.0000] <Hixie>
Ms2ger: or maybe let's just do it one small bit at a time?

[13:26:02.0000] <Ms2ger>
That's fine too

[13:26:03.0000] <Hixie>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html is the latest?

[13:27:00.0000] <Ms2ger>
Yes

[13:27:01.0000] <swarren08>
I noticed someone didnt like webapps, why is that?

[13:27:02.0000] <Hixie>
jgraham: i'm getting 504 gateway timeouts on pms

[13:27:03.0000] <Hixie>
jgraham: not even doing the annotations

[13:27:04.0000] <Hixie>
swarren08: the chair is pushing me to work on things for bureaucractic reasons rather than because they make the web better

[13:27:05.0000] <Ms2ger>
swarren08, webapps mean the Web Applications Working Group in this context, btw

[13:28:00.0000] <Hixie>
Ms2ger: just stuff in Terminology and Exceptions?

[13:28:01.0000] <Ms2ger>
Sounds good

[13:30:00.0000] <jgraham>
Hixie: OK I will look

[13:30:01.0000] <Hixie>
jgraham: it fixed itself

[13:30:02.0000] <Hixie>
jgraham: dunno what that was about

[13:31:00.0000] <jgraham>
Oh. Good

[13:31:01.0000] <jgraham>
I imagine it is just my hosting being rubbish or something

[13:32:00.0000] <Hixie>
Ms2ger: hmm, there's less here than i realised

[13:32:01.0000] <Hixie>
Ms2ger: what exactly are you looking for?

[13:32:02.0000] <Hixie>
Ms2ger: just the string stuff and the microsyntaxes stuff, and domexception?

[13:33:00.0000] <Hixie>
Ms2ger: or even things like "preferred mime type"?

[13:33:01.0000] <Ms2ger>
Dunno about mime type

[13:34:00.0000] <Ms2ger>
The rest was what I had in mind, I think

[13:34:01.0000] <Hixie>
s/codepoint/code point/

[13:34:02.0000] <Ms2ger>
It's been a while :)

[13:34:03.0000] <Hixie>
("codepoint" isn't english)

[13:34:04.0000] <jgraham>
It is if you choose it o be

[13:34:05.0000] <ParadoX->
Hey everyone

[13:34:06.0000] <jgraham>
*to

[13:34:07.0000] <hsivonen>
/me didn't know "HTML Standards" was an "engineering discipline"

[13:35:00.0000] <hsivonen>
(from Maciej's Wikipedia page)

[13:35:01.0000] <Hixie>
Ms2ger: also the dom core spec is missing "compatibility caseless"

[13:37:00.0000] <Hixie>
Ms2ger: the microsyntax section also isn't an exact match

[13:37:01.0000] <jgraham>
Is there some law that all WHATWG people with a wikipedia page have to have a dreadful photo?

[13:37:02.0000] <Hixie>
Ms2ger: which makes sense the specs have different requirements from these algorithms

[13:37:03.0000] <Ms2ger>
We copied just what we needed, I think

[13:37:04.0000] <Hixie>
Ms2ger: maybe it would make more sense for us to share some material and merge it in at publication time

[13:37:05.0000] <Hixie>
Ms2ger: rather than reference each other

[13:38:00.0000] <Hixie>
Ms2ger: i think it would be lame, e.g., for HTML to reference DOM Core for "skip whitespace" but not for "skip White_Space"

[13:38:01.0000] <Ms2ger>
I can pull that in as well

[13:38:02.0000] <Hixie>
(just like many of the split-out versions of WA1 have duplicate terminology)

[13:38:03.0000] <Hixie>
Ms2ger: yeah but then it's lame that DOM Core has stuff it doesn't use

[13:38:04.0000] <AryehGregor>
jgraham, apparently.  Part of the reason I haven't posted photos of myself anywhere is because I'm horrifyingly unphotogenic.

[13:39:00.0000] <Ms2ger>
The potential for divergence is also lame :)

[13:39:01.0000] <Hixie>
Ms2ger: yeah, hence, let's have some sort of common source we can merge together

[13:40:00.0000] <Hixie>
Ms2ger: the common source would be a superset of what we both need, and we'd just take the bits we need automatically when generating the spec

[13:40:01.0000] <jgraham>
AryehGregor: But this isn't even about being unphotogenic. It's just about the photos being bad

[13:40:02.0000] <Hixie>
Ms2ger: if you like we can make the dom core spec be that document, if it has appropriate markers i can just slurp it in

[13:40:03.0000] <Ms2ger>
Hmm

[13:40:04.0000] <AryehGregor>
Wikipedia photos are usually bad.  They prefer bad photos to no photos, because bad photos encourage the subject to contribute better photos under a free license.

[13:41:00.0000] <jgraham>
Obviouly th W3C should blow its budget on hiring a pro photographer for TPAC

[13:41:01.0000] <Hixie>
AryehGregor: i thought it was against the rules to submit your own photo

[13:41:02.0000] <jgraham>
Who will license the photos under a free license

[13:41:03.0000] <AryehGregor>
It often comes up for celebrities, where they post some trash that someone took at a public event with their cheap phone and put on Flickr.

[13:41:04.0000] <AryehGregor>
Then all someone has to do is inform the celebrity's press department and they'll get a freely-licensed photo right away!

[13:42:00.0000] <Ms2ger>
So I add the definitions we don't have yet and add some magic markers?

[13:42:01.0000] <Hixie>
Ms2ger: that would work

[13:42:02.0000] <Hixie>
in the meantime i'm fixing the exceptions stuff

[13:42:03.0000] <AryehGregor>
Hixie, not at all.  You're perfectly entitled to edit your own articles, even, as long as you're careful about it and are making only clear improvements.  See <http://en.wikipedia.org/wiki/WP:COI>.

[13:43:00.0000] <Hixie>
Ms2ger: you don't even need to add the stuff you don't have

[13:43:01.0000] <Hixie>
Ms2ger: just put markers around what you do have and i'll do the rest

[13:43:02.0000] <Hixie>
Ms2ger: i have an infrastructure to import stuff already

[13:43:03.0000] <AryehGregor>
If you have a better photo that's freely licensed, upload it and replace the existing one yourself if you like.

[13:43:04.0000] <jcranmer>
good god that's a bad photo

[13:44:00.0000] <Hixie>
i have a horrible feeling this project is going to involve me learning Yet Another source control versioning system

[13:44:01.0000] <Ms2ger>
But you'd want to put, say, compatibility caseless matches right in the middle, no?

[13:44:02.0000] <Hixie>
AryehGregor: that feels... self-indulgent in a way i'm not comfortable with

[13:44:03.0000] <AryehGregor>
They even have a specific guideline for it: http://en.wikipedia.org/wiki/Wikipedia:COI#Photographs_and_media_files

[13:44:04.0000] <Hixie>
Ms2ger: yeah but that's easy to do

[13:44:05.0000] <Ms2ger>
And if you don't know hg yet, yes

[13:44:06.0000] <AryehGregor>
No it's not, you're only improving Wikipedia.

[13:44:07.0000] <Ms2ger>
OK

[13:44:08.0000] <Hixie>
Ms2ger: just put plenty of markers in

[13:44:09.0000] <AryehGregor>
Uploading high-quality photos of Wikipedia subject matter is encouraged.

[13:44:10.0000] <Hixie>
Ms2ger: and i can just do regexps to slurp the right bits

[13:45:00.0000] <Hixie>
well on the plus side, hg is on this machine already

[13:45:01.0000] <Hixie>
that's lucky

[13:45:02.0000] <Ms2ger>
I hope you have all of this code backed up :)

[13:45:03.0000] <Hixie>
code?

[13:45:04.0000] <jgraham>
/me wonders what the average number of VCSs developers know these days is

[13:45:05.0000] <Ms2ger>
Your preprocessor

[13:45:06.0000] <Hixie>
yes

[13:46:00.0000] <Hixie>
but why would i need that? :-)

[13:46:01.0000] <Hixie>
my plan is to just have a script that splits apart your spec into separate files for eahc bit i need

[13:46:02.0000] <Hixie>
and then use the code i already have for importing files into the spec

[13:46:03.0000] <Hixie>
the same code i use to take examples in whatwg.org/demos and merge them into the workers section, w.g.

[13:46:04.0000] <Hixie>
e.g.

[13:46:05.0000] <Ms2ger>
I'm just afraid of the time lost if anything ever happens to your computer :)

[13:47:00.0000] <jcranmer>
most of my scripts aren't in VCS

[13:47:01.0000] <jcranmer>
I have an entire directory of that stuff

[13:47:02.0000] <jcranmer>
/src/mozilla-tools (which is a bit of a lie)

[13:47:03.0000] <jcranmer>
it's more of a "this isn't a separate repository, but it's a collection of little things which aren't worth crap by themselves"

[13:48:00.0000] <Ms2ger>
So, which sections, and what marker? :)

[13:49:00.0000] <hsivonen>
it seems that http://en.wikipedia.org/wiki/Wikipedia:Article_Feedback_Tool has existed for months, and yet I only learned about it just now

[13:50:00.0000] <karlcow>
hsivonen: is it bad to lag?

[13:51:00.0000] <hsivonen>
hsivonen: I don't know if it is *bad*. I just use wikipedia so often that I'm surprised I haven't seen that before.

[13:52:00.0000] <AryehGregor>
hsivonen, it's being gradually rolled out.

[13:52:01.0000] <AryehGregor>
First it was only a small subset of articles.

[13:52:02.0000] <AryehGregor>
IIRC, current plans are to roll it out to all articles soon, if it hasn't happened already.

[13:57:00.0000] <annevk>
so make stopped working on Lion

[13:57:01.0000] <annevk>
what the fuck

[13:59:00.0000] <Ms2ger>
Lion--

[14:01:00.0000] <jgraham>
annevk: You need to download the new Xcode too

[14:01:01.0000] <karlcow>
mwahah

[14:01:02.0000] <annevk>
ffs

[14:01:03.0000] <jgraham>
Always the best UX with Apple

[14:02:00.0000] <karlcow>
bye Google Labs http://googleblog.blogspot.com/2011/07/more-wood-behind-fewer-arrows.html

[14:02:01.0000] <annevk>
thanks jgraham

[14:03:00.0000] <annevk>
kind of interseting that macports is still around

[14:04:00.0000] <annevk>
but I cannot run upgrade

[14:04:01.0000] <annevk>
oh this is going to take ages too

[14:04:02.0000] <karlcow>
:) I guess annevk was too impatient to go to his (Lion) prom ball.

[14:05:00.0000] <annevk>
well it would have taken ages at some point

[14:06:00.0000] <annevk>
where ages is not really that long

[14:06:01.0000] <annevk>
downloading Lion takes about 30min

[14:06:02.0000] <annevk>
same for Xcode

[14:06:03.0000] <annevk>
installing either takes about half an hour too I guess, but that's still quite long

[14:07:00.0000] <karlcow>
I guess it depends on which type of lines you are.

[14:07:01.0000] <karlcow>
I wonder how long it takes in some less broadband fortunate places in the world.

[14:08:00.0000] <annevk>
actual ages

[14:08:01.0000] <karlcow>
It seems that they will finally release a USB key in August

[14:08:02.0000] <annevk>
this is >100mbit theoretically

[14:08:03.0000] <karlcow>
yup I do not see my parents downloading it through their house communication in the countryside

[14:09:00.0000] <Hixie>
Ms2ger: my computer here is a machine at dreamhost, which i backup via rsync nightly to a machine that is backed up to both time machine to a redundant drobo and via the cloud to an off-site backup system

[14:09:01.0000] <Hixie>
Ms2ger: so i think we're good

[14:09:02.0000] <Hixie>
lunch, bbl

[14:09:03.0000] <Ms2ger>
Good

[14:18:00.0000] <annevk>
I am not really convinced that CSS splitting is that great. Simple global things like parsing and its object model are very much under appreciated with the current design. And I am pretty sure there is a lot more.

[14:25:00.0000] <TabAtkins_>
That's because parsing is BOOOORING.

[14:25:01.0000] <TabAtkins_>
(Not really, please god let's fix it.)

[14:25:02.0000] <Ms2ger>
Yeah, let TabAtkins fix it :)

[14:26:00.0000] <TabAtkins_>
If I had another me, I'd do it right now.

[14:29:00.0000] <annevk>
parsing serializing and the object model are all overlooked in favor of "hey, new features!"

[14:29:01.0000] <annevk>
and "new features" often get prioritized in weird ways

[14:29:02.0000] <annevk>
quite often not in line with what developers and browsers are asking for I have the feeling

[14:30:00.0000] <annevk>
anyway

[14:30:01.0000] <annevk>
Xcode updated

[14:30:02.0000] <annevk>
make still fails

[14:30:03.0000] <annevk>
need to reboot terminal?

[14:30:04.0000] <annevk>
does not appear to work

[14:31:00.0000] <AryehGregor>
Argh, I really wish someone would fix CSS serialization.

[14:31:01.0000] <AryehGregor>
It's a nightmare to deal with.

[14:31:02.0000] <AryehGregor>
No interop at all.

[14:31:03.0000] <annevk>
wtf

[14:31:04.0000] <annevk>
Error: Target org.macports.configure returned: configure failure: shell command failed (see log for details)

[14:31:05.0000] <annevk>
Log for cvs is at: /opt/local/var/macports/logs/_opt_local_var_macports_sources_rsync.macports.org_release_ports_devel_cvs/main.log

[14:31:06.0000] <annevk>
Error: Unable to upgrade port: 1

[14:31:07.0000] <annevk>
I hate everything about this

[14:32:00.0000] <annevk>
Maybe I need to reboot Lion

[14:32:01.0000] <annevk>
hmm

[14:32:02.0000] <jamesr>
you're trying to develop on lion? ahahahahahahahahahahaha

[14:38:00.0000] <annevk>
yeah fuck this

[14:39:00.0000] <annevk>
time to do nothing

[15:10:00.0000] <heycam>
Hixie, here now

[15:16:00.0000] <Hixie>
heycam: dedicated workers and shared workers both have a global scope that is the same except it has different members

[15:16:01.0000] <Hixie>
heycam: the way i do it now is that i have a common ancestor

[15:17:00.0000] <Hixie>
heycam: and the descendant interfaces are [Supplemental, NoInterfaceObject]

[15:17:01.0000] <Hixie>
heycam: and then in the prose i just refer to the descendant interfaces

[15:17:02.0000] <heycam>
Hixie, ok so you want distinct dedicated and shared worker prototype objects, which have some common members, is that right?

[15:17:03.0000] <Hixie>
yeah

[15:17:04.0000] <Hixie>
but they have the same name

[15:18:00.0000] <heycam>
Hixie, the same name?

[15:18:01.0000] <Hixie>
WorkerGlobalScope

[15:21:00.0000] <heycam>
Hixie, ok, could you mail public-script-coord about it? I will try my hardest to avoid adding [Supplemental] but you never know ;)

[15:21:01.0000] <Hixie>
i've no problem solving it another way, i'd just like to avoid awkward prose

[15:21:02.0000] <Hixie>
will e-mail the list

[15:21:03.0000] <heycam>
sure

[15:21:04.0000] <heycam>
thanks

[15:22:00.0000] <heycam>
/me will get back to looking at Web IDL in a couple of weeks

[15:57:00.0000] <Hixie>
heycam: what happens if an object marked NoInterfaceObject inherits from another and is instantiated?

[15:57:01.0000] <jwalden>
huh, ie9 doesn't support wav in its audio element?

[15:57:02.0000] <heycam>
Hixie, it works just as if it didn't have NoInterfaceObject -- the only thing NoInterfaceObject affects is whether the global.InterfaceName property exists

[15:59:00.0000] <Hixie>
heycam: k

[16:00:00.0000] <Hixie>
heycam: mail sent

[16:00:01.0000] <heycam>
thanks

[16:11:00.0000] <jwalden>
/me marks http://krijnhoetmer.nl/irc-logs/whatwg/20110720#l-1335 appropriately

[16:35:00.0000] <annevk>
did Google Maps get a new UI?

[16:35:01.0000] <TabAtkins>
Yes.

[16:35:02.0000] <annevk>
let me the be the first to point it is inconsistent with Calendar and Google+

[16:35:03.0000] <TabAtkins>
Vaguely related: I wish we used a distance-preserving projection on Maps.

[16:36:00.0000] <annevk>
David Bloom refers to this as first world problems

[16:36:01.0000] <TabAtkins>
Canada looks too damn big.

[16:36:02.0000] <annevk>
I say ugly

[16:37:00.0000] <jcranmer>
lol

[16:37:01.0000] <jcranmer>
zoom out to the second notch

[16:37:02.0000] <jcranmer>
hover somewhere on the left

[16:37:03.0000] <jcranmer>
and then zoom in


2011-07-21
[17:01:00.0000] <TabAtkins>
Augh, I need to move off of Aurora.  twice-a-day restarts are killing me.

[19:37:00.0000] <jwalden>
TabAtkins: I thought we only updated once a day

[21:07:00.0000] <Hixie>
there's no shipped way to save a canvas across page loads and sessions, right?

[21:07:01.0000] <Hixie>
since nobody saves imagedata to web storage?

[21:10:00.0000] <beverloo>
could get an image from the canvas and store that, I suppose

[21:11:00.0000] <Hixie>
like the data: URL?

[21:11:01.0000] <Hixie>
i guess

[21:11:02.0000] <beverloo>
Yes

[21:11:03.0000] <Hixie>
that's pretty messed up

[21:11:04.0000] <Hixie>
but i guess it would work

[21:11:05.0000] <Hixie>
i wonder what the perf implications would be

[21:35:00.0000] <Hixie>
if i have two <div>s with similar but not identical content that overlap

[21:36:00.0000] <Hixie>
and i want to fade from one to the other

[21:36:01.0000] <Hixie>
is there a way to do it that doesn't involve the areas that are the same changing coloru at all?

[22:11:00.0000] <othermaciej>
Hixie: layer one over the other, transition opacity of the top one from 100% to 0%

[22:12:00.0000] <zewt>
only works for opaque images

[22:15:00.0000] <othermaciej>
zewt: I assume from Hixie's example that both divs in question have or can be given an opaque background

[22:16:00.0000] <zewt>
i assumed the more complex case since i figured he knew how to use opacity :)

[22:17:00.0000] <othermaciej>
hmm, maybe they overlap but not 100%

[22:17:01.0000] <othermaciej>
and he wants to change which is "on top"

[22:17:02.0000] <othermaciej>
that would be less obvious

[22:48:00.0000] <annevk>
oh yes

[22:48:01.0000] <annevk>
make works again

[22:49:00.0000] <annevk>
macports however is still broken

[22:56:00.0000] <Hixie>
man some of these lion changes are quite buggy

[22:57:00.0000] <Hixie>
e.g. the flickering when you switch from one terminal to another if you have different settings for blur for active vs inactive windows

[23:00:00.0000] <Hixie>
or this crazy sheering effect in the screen saver prefs "shuffle photos" sheet

[23:14:00.0000] <annevk>
http://trac.webkit.org/browser/trunk/Source/WebCore/dom/TreeWalker.cpp#L139 how does the "sibling = sibling->previousSibling();" line matter?

[23:15:00.0000] <annevk>
look like sibling will be set by the for loop right after

[00:00:00.0000] <annevk>
http://trac.webkit.org/browser/trunk/Source/WebCore/dom/TreeWalker.cpp#L139 how does the "sibling = sibling->previousSibling();" line matter?

[00:00:01.0000] <annevk>
look like sibling will be set by the for loop right after

[00:00:02.0000] <annevk>
othermaciej, ^^

[00:01:00.0000] <othermaciej>
annevk: which "sibling = sibling->previousSibling()" line?

[00:02:00.0000] <annevk>
163

[00:02:01.0000] <othermaciej>
(I see at least two lines that say that)

[00:03:00.0000] <othermaciej>
that loop is a little mysterious to me

[00:03:01.0000] <annevk>
glad I am not the only one

[00:03:02.0000] <othermaciej>
in the "NodeFilter::FILTER_REJECT", nothing updates node

[00:04:00.0000] <othermaciej>
so it seems like that would cause an infinite loop

[00:05:00.0000] <othermaciej>
likewise for NodeFilter::FILTER_SKIP in the case where sibling has no children

[00:05:01.0000] <othermaciej>
maybe break actually breaks out of the loop instead of the switch statement

[00:05:02.0000] <othermaciej>
or is intended to

[00:05:03.0000] <othermaciej>
or something

[00:05:04.0000] <annevk>
in both of those cases you want the previousSibling

[00:06:00.0000] <othermaciej>
annevk: oh, nevermind

[00:06:01.0000] <othermaciej>
RefPtr<Node> sibling = node->previousSibling() is the initializer

[00:06:02.0000] <othermaciej>
the for loop has no increment step

[00:06:03.0000] <othermaciej>
and the loop condition is sibling not being null

[00:06:04.0000] <othermaciej>
so it totally makes sense

[00:06:05.0000] <annevk>
but how does line 163 make sense?

[00:07:00.0000] <annevk>
the for loop overwrites that, no?

[00:19:00.0000] <othermaciej>
annevk: no

[00:19:01.0000] <othermaciej>
annevk: the for loop's assignment to sibling is the initializer, not the increment

[00:19:02.0000] <othermaciej>
annevk: so it only sets sibling the first time before entering the loop

[00:20:00.0000] <annevk>
ah thanks

[02:34:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-treewalker-parentnode

[02:34:01.0000] <annevk>
Would be nice to auto feed concept-node-filter with whatToShow and filter somehow

[02:35:00.0000] <annevk>
Maybe besides "active flag" I should just define whatToShow and filter as concepts and then say that the attributes on the objects map to those concepts?

[02:36:00.0000] <annevk>
It has to be something like that; the current way is way verbose

[02:36:01.0000] <Ms2ger>
wfm

[02:46:00.0000] <annevk>
not sure when I'll get to it btw

[02:47:00.0000] <annevk>
got up a little too early and am now gonna relax for a while :)

[02:47:01.0000] <annevk>
but feedback welcome

[02:50:00.0000] <ivan`>
any ideas why Access-Control-Allow-Origin doesn't have subdomain wildcards?

[02:50:01.0000] <ivan`>
hmm, maybe I can just convince the server to send the right header for whichever subdomain

[03:35:00.0000] <hsivonen>
ivan`: the design of CORS assumes that it is applied to dynamic content, so it's OK to require the headers to vary, too. By not allowing wild-carding, the programmer is forced to think more and is assumed to be less likely to declare bad policies

[03:35:01.0000] <hsivonen>
ivan`: the result is, though, that CORS really sucks for static resources like images or fonts

[03:35:02.0000] <hsivonen>
if you want anything more granular than the blanket anything-goes wildcard * that is

[03:36:00.0000] <hsivonen>
this would become a lot less painful if Apache shipped with a mod_cors

[04:06:00.0000] <ivan`>
hsivonen: thank you

[04:07:00.0000] <ivan`>
I was worried about that header getting cached

[04:07:01.0000] <ivan`>
and then later I learned I could also just look at Origin

[04:10:00.0000] <ivan`>
anyway, now I see that it doesn't get cached, obviously

[08:38:00.0000] <dpy>
hi guys

[08:39:00.0000] <jgraham>
hej

[08:39:01.0000] <dpy>
Where would I post bug reports about this (i.e. which bugzilla component): http://validator.nu/?doc=http%3A%2F%2Fjsbin.com%2Fetabil%2F3

[08:42:00.0000] <hsivonen>
dpy: do you mean the spelling bug "Namspaces"?

[08:42:01.0000] <dpy>
no, data-[ is a valid data-attribute name

[08:42:02.0000] <dpy>
and so are data-$, data-%, etc.

[08:43:00.0000] <jgraham>
Oh, I thought you meant that the line numbers were wrong :)

[08:43:01.0000] <dpy>
basically anything that is a non-space character in available unicode that is not a control character and not [\0"'/=]

[08:43:02.0000] <hsivonen>
dpy: that's not what the spec said last time I looked

[08:44:00.0000] <hsivonen>
/me looks again

[08:44:01.0000] <Philip`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/elements.html#embedding-custom-non-visible-data-with-the-data-*-attributes

[08:44:02.0000] <dpy>
http://www.w3.org/TR/html5/syntax.html#attributes-0 <-- ?

[08:44:03.0000] <Philip`>
"▶  A custom data attribute is an attribute in no namespace whose name starts with the string "data-", has at least one character after the hyphen, is XML-compatible, and contains no characters in the range U+0041 to U+005A (LATIN CAPITAL LETTER A to LATIN CAPITAL LETTER Z)."

[08:45:00.0000] <Philip`>
"Attribute names are said to be XML-compatible if they match the Name production defined in XML, they contain no U+003A COLON characters (:), and their first three characters are not an ASCII case-insensitive match for the string "xml"."

[08:45:01.0000] <hsivonen>
dpy: what Philip` said

[08:45:02.0000] <dpy>
What source are you quoting?

[08:45:03.0000] <dpy>
that link?

[08:45:04.0000] <hsivonen>
dpy: http://www.whatwg.org/specs/web-apps/current-work/#embedding-custom-non-visible-data-with-the-data-*-attributes

[08:45:05.0000] <hsivonen>
and http://www.whatwg.org/specs/web-apps/current-work/#xml-compatible

[08:46:00.0000] <dpy>
So the stuff on w3.org is outdated?

[08:46:01.0000] <jgraham>
Well yes, but more to the point it's the wrong part of the spec

[08:47:00.0000] <Philip`>
http://www.w3.org/TR/html5/elements.html#embedding-custom-non-visible-data-with-the-data-attributes is the relevant part

[08:47:01.0000] <jgraham>
In general anything on TR/ is outdated

[08:47:02.0000] <Philip`>
which hasn't changed since the obsolete TR/

[08:47:03.0000] <jgraham>
And, increasingly, it seems that anything on the W3C site will be outdated

[08:47:04.0000] <dpy>
What does TR stand for if I may ask?

[08:47:05.0000] <hsivonen>
dpy: Technical Reports officially

[08:47:06.0000] <jgraham>
Temporally Retarded

[08:47:07.0000] <dpy>
lol

[08:48:00.0000] <Ms2ger>
TRash

[08:49:00.0000] <dpy>
Still, I think it is strange that two parts of the spec (at one point) seems to be ambiguous

[08:49:01.0000] <dpy>
but I probably misread

[08:50:00.0000] <dpy>
oh wait

[08:50:01.0000] <hsivonen>
dpy: you quoted a part about attributes in general. data-* attributes have more specific rules on top of the general stuff

[08:50:02.0000] <dpy>
yeah, the part I read was about the syntax

[08:50:03.0000] <Philip`>
The "Writing HTML documents" section says the form that attributes take, but you're only allowed to use attributes that are explicitly defined in the semantics

[08:50:04.0000] <dpy>
ok, I understand

[08:51:00.0000] <dpy>
too bad,  would've been nice to be able to write things like:  data-$foo=   or data-foo[0]=

[08:51:01.0000] <Philip`>
which is a finite set of attributes plus a particular infinite subset of the ones start with "data-"

[08:52:00.0000] <Philip`>
so the attribute "data-$" is no different to the attribute "$$$"

[08:52:01.0000] <Philip`>
(in terms of conformance, at least)

[08:52:02.0000] <Philip`>
(unless I'm mistaken)

[08:55:00.0000] <hsivonen>
I fixed the typo in version control. Will deploy at some point but not now.

[09:09:00.0000] <dpy>
okay..   show set [:ANY:] on unicode.org was not a good idea

[09:22:00.0000] <TabAtkins>
dpy: Hah, indeed.

[09:22:01.0000] <TabAtkins>
"Please show me 10000 characters."

[09:26:00.0000] <dpy>
Why is data-a:0  not a valid attribute name?

[09:26:01.0000] <TabAtkins>
Because : makes the name not xml-compatible.

[09:27:00.0000] <TabAtkins>
(Philip` gave the description of "xml-compatiable" from the spec about 45min ago.)

[09:27:01.0000] <dpy>
Well, the html specs send me to the #NT-Name anchor of XML

[09:27:02.0000] <TabAtkins>
Just do data-a_0

[09:27:03.0000] <dpy>
and a:0 is a valid name according to that production

[09:28:00.0000] <TabAtkins>
http://www.whatwg.org/specs/web-apps/current-work/#xml-compatible

[09:28:01.0000] <dpy>
it even explicitly states in text: "author should not use the colon in XML names except for namespace purposes, but XML processors must accept the colon as a name character"

[09:28:02.0000] <Philip`>
"... they contain no U+003A COLON characters (:)"

[09:28:03.0000] <Philip`>
(from #xml-compatible)

[09:29:00.0000] <dpy>
I stand corrected

[09:29:01.0000] <dpy>
I overlooked that by clicking through

[09:29:02.0000] <dpy>
seems that some people are trying very hard to not try to put indexed values in a data-attribute name

[09:30:00.0000] <dpy>
and that's exactly what I would like to be able to do

[09:30:01.0000] <TabAtkins>
No, it's an incidental consequence of making the attribute xml-compatible.

[09:30:02.0000] <Philip`>
What's the situation where you're trying to do that?

[09:30:03.0000] <TabAtkins>
Use an underscore.

[09:31:00.0000] <dpy>
the problem with the underscore is that it's typically also allowed in names where I don't expected indexed values

[09:31:01.0000] <dpy>
my software would not be able to distinguish between them

[09:31:02.0000] <Philip`>
I think the usual assumption for structured data is that you can do it with JSON, like data-whatever="[1,2,3]" instead of data-whatever[0]="1" ...

[09:31:03.0000] <TabAtkins>
Make those names use - instead.

[09:31:04.0000] <dpy>
Philip`: yes, but that get's hairy when we're not simply talking 1, 2 3

[09:32:00.0000] <jgraham>
/me wonders what is not hairy munging text from an attibute name but is hairy parsing its value

[09:34:00.0000] <dpy>
I have three independent pieces of JSON, which are much easier to read when stored in three attributes than as a single JSON list crammed into a single attribute

[09:34:01.0000] <annevk>
* Microsoft will be submitting tests for SCA (Structured Cloning Algorithm, Forms, Async Script, Drag & Drop and Sandbox).

[09:34:02.0000] <annevk>
better set aside some reviewing time

[09:34:03.0000] <jgraham>
Really [[1,2,3], [2,,5,6], [7,8,9]] doesn't seem that hard to read

[09:35:00.0000] <jgraham>
And it is even easy to parse

[09:35:01.0000] <dpy>
jgraham: now replace each of the characters by a nested JSON structure

[09:35:02.0000] <jgraham>
dpy: If you have nested structures anyway, having an array of three of them doesn't seem at all bad compared to having three seperately

[09:36:00.0000] <dpy>
like: [{"foo":["X","Y","Z"],"bar":[...]"},{"foo":["X","Y","Z"],"bar":[...]"},{"foo":["X","Y","Z"],"bar":[...]"},{"foo":["X","Y","Z"],"bar":[...]"},{"foo":["X","Y","Z"],"bar":[...]"}]

[09:36:01.0000] <dpy>
and compare this with:

[09:36:02.0000] <dpy>
foo[0]={"foo":["X","Y","Z"],"bar":[...]"}

[09:36:03.0000] <dpy>
foo[1]={"foo":["X","Y","Z"],"bar":[...]"}

[09:36:04.0000] <dpy>
+ '    '

[09:37:00.0000] <dpy>
jgraham: well, it is if you keep in mind that it is likely for one of these to change and you need to resolve conflicts

[09:37:01.0000] <jgraham>
Well if you really care you can write {0:{"foo":["X","Y","Z"],"bar":[...]"},\n1:{"foo":["X","Y","Z"],"bar":[...]"}}

[09:38:00.0000] <dpy>
If your users aren't experiences programmers, I'd rather have them solve 2 attributes that are in conflict, than bore them with the details of JSON

[09:38:01.0000] <jgraham>
I am having great difficulty in seeing how having three attributes containing JSON is easier than having one attribute contaning JSON

[09:39:00.0000] <dpy>
I expect my users to understand html5, not JSON

[09:39:01.0000] <dpy>
now they are required to understand JSON

[09:39:02.0000] <dpy>
to solve a merge conflict

[09:40:00.0000] <jgraham>
I totally don't understand how they could not need to understand JSON if the data they are merging is in JSON

[09:40:01.0000] <dpy>
plus I liked the idea: one data-attribute = one property

[09:40:02.0000] <dpy>
not: some data attributes define multiple properties

[09:41:00.0000] <TabAtkins>
dpy: What happens if your data structure ends up nesting more deeply?

[09:41:01.0000] <jgraham>
Well you can do it your way if you like. But you have to use syntax that doesn't conflict with XML

[09:41:02.0000] <dpy>
jgraham: you don't need to understand the details of a syntax to point out which line you want to keep and which you don't

[09:41:03.0000] <dpy>
like: it's easier to read something, than it is to write it

[09:42:00.0000] <dpy>
TabAtkins: they are always singly lined, and should not get too complicated (otherwise something else is probably smelly)

[09:42:01.0000] <jgraham>
/me hopes you are not planning to use this for anything important

[09:42:02.0000] <jgraham>
Because expecting people to edit data in a format they don't understand

[09:42:03.0000] <jgraham>
That's not gonna work

[09:43:00.0000] <dpy>
jgraham: I'm obviously not able to do much about the status quo...   I'm clearly not allowed to do it the way I like...

[09:43:01.0000] <TabAtkins>
So anyway, point is, you're free to embed things however you like if you use xml-compatible names (which happens to exclude some of the more natural ways to write what you want).  It may be a better idea to just embed json, but whatever.

[09:43:02.0000] <dpy>
TabAtkins: I agree with you

[09:44:00.0000] <dpy>
Though I don't like it, it seems the only viable alternative, given the great lengths the specification has gone to prevent what I want to do

[09:45:00.0000] <TabAtkins>
dpy: You're making it sound like the spec went out of its way to prevent you from making it easy to solve your case.  It didn't.  It just said that custom attributes have to obey the same syntax rules as every other attribute in the language.

[09:45:01.0000] <dpy>
I would have rather seen a more liberal approach: You can use anything, except these...

[09:45:02.0000] <dpy>
Rather then: you can't use anything except these...

[09:45:03.0000] <dpy>
then=than

[09:46:00.0000] <TabAtkins>
If you get an identical set of allowed characters either way, it seems irrelevant whether the set is specified additively rather than subtractively.

[09:46:01.0000] <dpy>
TabAtkins: I don't take it personal. I'm feeling it as though it was a partical design choice. I may not agree with it, but it's fact now, so I might just as wel accept the situation

[09:47:00.0000] <dpy>
TabAtkins: " It just said that custom attributes have to obey the same syntax rules as every other attribute in the language." <-- this is not true. non-data HTML5 attributes are much less restricted

[09:48:00.0000] <TabAtkins>
Huh?  How so?

[09:48:01.0000] <Philip`>
Non-data attributes are entirely restricted because you can't use any of them at all (except the finite set listed in the spec)

[09:48:02.0000] <dpy>
So any future attribute can still be written like: foo[2]=   and current parsers need to accept it

[09:48:03.0000] <TabAtkins>
That's only possible in the case that the spec changes the rules and allows a non-XML-compatible attribute name.

[09:48:04.0000] <dpy>
Philip`: I'm talking about restrictions toward the future

[09:49:00.0000] <dpy>
TabAtkins: Where do you read that a non-data attribute must be xml compatible

[09:49:01.0000] <dpy>
?

[09:49:02.0000] <TabAtkins>
It wouldn't make much sense for the spec to impose restrictions on itself, given that it's the spec and can just change those restrictions later.

[09:49:03.0000] <TabAtkins>
dpy: Every attribute is XML-compatible.

[09:50:00.0000] <TabAtkins>
dpy: Also, informally, keeping the language such that it generates the same DOM in HTML and XHTML as much as possible has been a strong goal.

[09:50:01.0000] <Philip`>
Parsers accept data-foo[2]="..." exactly the same way as foo[2]="...", and expose it as e.dataset["foo[2]"] etc - the only restrictions are for conformance (to discourage dangerous or XML-incompatible syntax))

[09:50:02.0000] <Philip`>
s/)//

[09:52:00.0000] <dpy>
Philip`: I should I read that?  That I can use it safely, or that I can't?

[09:52:01.0000] <TabAtkins>
You, as an author, are required to not do it.

[09:52:02.0000] <TabAtkins>
Browsers are required to accept to it.

[09:52:03.0000] <TabAtkins>
s/to it/it/

[09:53:00.0000] <dpy>
I don't want to have a broken app, if something in the spec changes and the browsers no longer accept my data-attributes

[09:53:01.0000] <zewt->
the usual contradiction. heh

[09:53:02.0000] <TabAtkins>
(Because every possible sequence of bits is an HTML document, though possibly an invalid one, and must have rules on how to process it.)

[09:54:00.0000] <dpy>
Okay, this is the old: be strict in what you produce, be unstrict in what you accept

[09:54:01.0000] <oal>
Is it possible to insert a link <a href="something">..</a> with css? When I do it with :after {content: 'somehtml'} it displays as text

[09:54:02.0000] <TabAtkins>
Basically, yeah.

[09:54:03.0000] <Ms2ger>
oal, only in Opera

[09:54:04.0000] <TabAtkins>
oal: No.  Linking is an HTML behavior that can't be controlled with CSS.

[09:55:00.0000] <Ms2ger>
And Prince, I think

[09:55:01.0000] <oal>
Ok :/

[09:57:00.0000] <Philip`>
If it was argued sufficiently convincingly that allowing non-XML-compatible characters in attribute names would be the best solution to some use cases, and if people were in a fed-up-with-XML mood and didn't mind abandoning the pretence of supporting it as an equal to text/html, then the conformance requirements could be relaxed and nothing else would change

[09:57:01.0000] <dpy>
Hurray!

[09:57:02.0000] <Philip`>
(since browsers will already parse it in the expected way)

[09:57:03.0000] <dpy>
Middle dot to the rescue:

[09:57:04.0000] <dpy>
data-foo·0="bar"

[09:57:05.0000] <dpy>
is allowed

[09:57:06.0000] <Philip`>
That's disgusting :-p

[09:57:07.0000] <zewt>
gross

[09:57:08.0000] <dpy>
now don't you guys dare take this away from me!

[09:57:09.0000] <dpy>
:P

[09:57:10.0000] <TabAtkins>
It seems bad to use non-ascii characters in an attribute, since you can't type them on your keyboard without an OS escape sequence, but whatever.  ^_^

[09:58:00.0000] <TabAtkins>
Also, retrieving the attribute will be more difficult.

[09:58:01.0000] <Ms2ger>
Mm, can you use entities in attribute names?

[09:58:02.0000] <dpy>
TabAtkins: they aren't really meant to be typed, so people can't accidentally type them.. which may even be a good thing

[09:58:03.0000] <TabAtkins>
Ms2ger: Don't think so.

[09:59:00.0000] <Philip`>
dpy: I thought you wanted inexperienced people to edit these things by hand

[09:59:01.0000] <dpy>
Philip: read: delete stuff

[09:59:02.0000] <dpy>
Philip`: but it may be a bad idea tomorrow morning

[10:00:00.0000] <Philip`>
Ms2ger: No but you could always write <script>document.write('<div data-foo\xb7')</script>="...data..."></div> if you want to avoid non-ASCII source

[10:01:00.0000] <Ms2ger>
/me glares at Philip` 

[10:02:00.0000] <Philip`>
Oops, I forgot the 0 in there

[10:02:01.0000] <annevk>
jgraham, congrats on the labs release!

[10:08:00.0000] <annevk>
https://plus.google.com/112095156983892490612/posts/EMADdvHSRbH

[10:08:01.0000] <annevk>
gotta love the rhetoric in the comments

[10:10:00.0000] <TabAtkins>
I'm surprised.  Julian's suggestion seems reasonable.

[10:10:01.0000] <dpy>
Okay, maybe it's just best to go with   data-foo_0, foo.bar_0, foo.bar.xyz_0  and just prohibit my users to use the underscore _  (they'll just have to use the full stop as a delimiter)

[10:11:00.0000] <TabAtkins>
dpy: Or the dash.

[10:11:01.0000] <dpy>
TabAtkins: the dash is interpreted

[10:11:02.0000] <annevk>
His suggestion is basically what is being done, no?

[10:11:03.0000] <annevk>
Bunch of new features do not appear in the W3C draft

[10:11:04.0000] <dpy>
data-foo-0 becomes foo0, right?

[10:11:05.0000] <TabAtkins>
Only if you use the dataSet accessor.

[10:11:06.0000] <TabAtkins>
dataset.foo0

[10:11:07.0000] <dpy>
just as data-foo-bar becomes fooBar

[10:12:00.0000] <TabAtkins>
But if you're using dataSet, you can't access anything with a period in it.

[10:12:01.0000] <dpy>
huh?

[10:12:02.0000] <dpy>
/me checks

[10:12:03.0000] <TabAtkins>
Or rather, you'd need to access it like an array.  dataSet['foo.bar']

[10:13:00.0000] <TabAtkins>
Though, actually... let me check the spec real quick about data-foo_0

[10:13:01.0000] <Philip`>
(dataset, not dataSet)

[10:13:02.0000] <TabAtkins>
dpy: No, data-foo-0 is not changed.

[10:13:03.0000] <dglazkov>
good morning, Whatwg!

[10:13:04.0000] <TabAtkins>
it's still dataset['foo-0']

[10:14:00.0000] <Hixie>
/me nods to dglazkov 

[10:14:01.0000] <TabAtkins>
Or rather, data-foo_0 is still dataset['foo_0']

[10:14:02.0000] <TabAtkins>
Well, both.

[10:15:00.0000] <TabAtkins>
ANYWAY, you only drop the dash if the following character is a lowercase letter (and then you uppercase that letter)

[10:15:01.0000] <dpy>
ok

[10:15:02.0000] <dpy>
I'll report that back to my team

[10:16:00.0000] <TabAtkins>
annevk: Not really.  The HTMLWG hasn't branched properly, it's forced Hixie to branch and then keep the branch in line with trunk.

[10:16:01.0000] <dpy>
TabAtkins: tnx

[10:16:02.0000] <TabAtkins>
If we did a proper branch, a lot would be easier (and the HTMLWG branch would be immediately obsolete).

[10:17:00.0000] <annevk>
and keep maintaining it as Julian suggests?

[10:17:01.0000] <annevk>
sounds like a bad idea

[10:17:02.0000] <TabAtkins>
annevk: If the HTMLWG is maintaining a frozen branch, it's not really important.

[10:18:00.0000] <annevk>
I don't see how it would be better

[10:20:00.0000] <Ms2ger>
Less work for Hixie, more time to do something useful?

[10:20:01.0000] <TabAtkins>
I was just typing that.

[10:21:00.0000] <Ms2ger>
/me finally beat TabAtkins to something

[10:22:00.0000] <hsivonen>
annevk: regarding rhetoric, Leigh Klotz is an XForms guy, right?

[10:22:01.0000] <annevk>
think so, yes

[10:23:00.0000] <TabAtkins>
In other words, "someone who's bitter about the fact that browsers are the only relevant user of HTML".

[10:23:01.0000] <Philip`>
/me guesses one significant danger is more time being wasted by people who read and/or implement the W3C version of the spec without realising it's obsolete

[10:24:00.0000] <annevk>
yeah, it does not seem like a smart thing

[10:24:01.0000] <annevk>
to let that run wild

[10:24:02.0000] <TabAtkins>
The alternative is dealing with the HTMLWG continually trying to an LC.

[10:25:00.0000] <annevk>
if the W3C wants to move on they should focus more on tests and less on process

[10:25:01.0000] <annevk>
or change the process

[10:25:02.0000] <annevk>
tests and implementations are the bottleneck to move HTML5 forward there, nothing else

[10:25:03.0000] <annevk>
they are always the bottleneck

[10:26:00.0000] <Ms2ger>
Weren't they going to hire people to write tests from the HTML5 T-shirt sales?

[10:27:00.0000] <jgraham>
Well they said they would use them to fund testsuite development

[10:27:01.0000] <jgraham>
Not exactly how

[10:27:02.0000] <Philip`>
Maybe nobody has bought them since they're waiting for the Living Standard T-shirts

[10:27:03.0000] <jgraham>
So maybe they are saving it all for a big testsuite-development party

[10:27:04.0000] <jgraham>
With champagne and caviar

[10:28:00.0000] <annevk>
damnit

[10:28:01.0000] <annevk>
this stuff requires subsubsteps

[10:28:02.0000] <annevk>
jgraham is gonna kill me

[10:28:03.0000] <jgraham>
You should do it like ES

[10:28:04.0000] <jgraham>
It's really better :)

[10:29:00.0000] <annevk>
how do they do break?

[10:29:01.0000] <jgraham>
I don't recall. But usually you can avoid it if you structure the loop correctly

[10:29:02.0000] <Ms2ger>
jgraham, when we go to PR?

[10:30:00.0000] <annevk>
example:

[10:30:01.0000] <jgraham>
Ms2ger: Any earlier and we wouldn't be able to afford real champagne

[10:30:02.0000] <annevk>
  while node is not null

[10:30:03.0000] <annevk>
    sibling = node nextSibling/previousSibling

[10:30:04.0000] <annevk>
    if sibling

[10:30:05.0000] <annevk>
      node = sibling

[10:30:06.0000] <annevk>
      break (back to outer loop)

[10:30:07.0000] <Philip`>
Split it into a separate function and do a 'return'?

[10:30:08.0000] <Ms2ger>
while node is not null and (random boolean) is true

[10:30:09.0000] <Ms2ger>
... (random boolean) = false

[10:31:00.0000] <annevk>
eww

[10:31:01.0000] <Ms2ger>
:)

[10:31:02.0000] <jgraham>
Well anything is better than "and terminate these sub sub steps"

[10:31:03.0000] <Ms2ger>
goto

[10:31:04.0000] <hsivonen>
break;

[10:31:05.0000] <jgraham>
Well annevk uses goto as well

[10:31:06.0000] <hsivonen>
well, break with label

[10:32:00.0000] <hsivonen>
it's annoying that C++ lack break with label

[10:32:01.0000] <hsivonen>
but spec English doesn't have to lack break with label

[10:32:02.0000] <Philip`>
/me has occasionally done "for (i = 0; i < 100; ++i) { for (j = 0; j < 100; ++j) { ... if (whatever) i = 100; /* break out of outer loop */ } }"

[10:32:03.0000] <Philip`>
(but then felt unbearably dirty)

[10:32:04.0000] <Ms2ger>
And you should

[10:33:00.0000] <Philip`>
(and usually rewrote it as function with 'return' since that's the least nasty unstructured control flow statement)

[10:36:00.0000] <annevk>
separate algorithms work I guess, but it seems so pointless

[10:37:00.0000] <annevk>
or maybe I should go the Ms2ger route

[10:37:01.0000] <annevk>
with a silly variable

[10:37:02.0000] <Ms2ger>
Silly? That's me :)

[10:37:03.0000] <annevk>
oh wait, that does not work

[10:38:00.0000] <annevk>
well i would have to check it in several other places too, that would suck

[10:41:00.0000] <annevk>
man this sucks

[11:10:00.0000] <annevk>
because I can http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-treewalker-firstchild

[11:29:00.0000] <nmadzharov>
hi

[11:32:00.0000] <nmadzharov>
is  handleEvent( Event evt) { if(evt.isMouseEvent()) MouseEvent me = (MouseEvent) evt; } a common way to implement the function?

[11:39:00.0000] <TabAtkins>
I presume you're talking about something other than JS.

[11:48:00.0000] <AryehGregor>
http://trevorparscal.com/stuff/msiejs/msiejs-diagram.png

[11:51:00.0000] <TabAtkins>
My favorite part is the "Murderer".

[11:51:01.0000] <nmadzharov>
lol

[12:07:00.0000] <versatiletech>
HTML5 job in close to Boston,MA? Relocation expenses paid. Very high pay. Opportunity to work on  networking web app. PM me if you're interested.

[12:23:00.0000] <TabAtkins>
Ooh, got a coons patch working!

[12:23:01.0000] <TabAtkins>
http://www.xanthir.com/etc/scaling-algorithms.html

[12:27:00.0000] <smaug____>
annevk: seems like http://dev.w3.org/csswg/cssom-view/#dom-document-caretpositionfrompoint mentions still caretRange which is already removed from the interface

[12:31:00.0000] <annevk>
smaug____, http://lists.w3.org/Archives/Public/www-style/2011Jun/0604.html

[12:38:00.0000] <smaug____>
ah

[13:25:00.0000] <dglazkov>
TabAtkins: did I capture this right? http://wiki.whatwg.org/wiki/Component_Model_Use_Cases#Rendering_Form_Controls_with_SVG

[13:33:00.0000] <Hixie>
TabAtkins: you know the css in html is normative, right? :-)

[13:36:00.0000] <annevk>
dglazkov, looks good to me!

[13:37:00.0000] <dglazkov>
cool

[13:38:00.0000] <TabAtkins>
Hixie: If I did, I wouldn't've said the opposite.

[13:38:01.0000] <smaug____>
XForms extensions supports svg styled form controls :p

[13:39:00.0000] <othermaciej>
dglazkov: those use cases look much better than the old use cases document

[13:39:01.0000] <smaug____>
s/extensions/extension/

[13:39:02.0000] <othermaciej>
though it is hard to see how the "Desirable Properties" follow from the use cases

[13:42:00.0000] <othermaciej>
dglazkov: it would be nice if the "Desirable Properties" section said which use case requires which property and why, or better yet, list under each use case what this specific use case implies about desirable properties for a solution

[14:34:00.0000] <TabAtkins>
Hmm.  Is it possible to add a function to Number that operates on the value?

[14:34:01.0000] <TabAtkins>
Number.prototype.add(that) { return this+that; } doesn't work.

[14:35:00.0000] <gsnedders>
TabAtkins: this is a Number object in that situation

[14:36:00.0000] <gsnedders>
TabAtkins: this.valueOf() gives you the Number primitive

[14:38:00.0000] <TabAtkins>
Hmm, I'm getting a screwy result here that's probably unrelated...

[14:44:00.0000] <TabAtkins>
Oh jeez, I see what I did wrong.  Never mind.

[14:45:00.0000] <metron>
hi

[14:47:00.0000] <metron>
Iam still reading webdev books but I can't any discussion about integration concepts of HTML5 standards, where I can find that stuff? thx

[14:47:01.0000] <metron>
+find

[14:47:02.0000] <TabAtkins>
What do you mean?

[14:48:00.0000] <metron>
Tab: information integration from old websites maybe

[14:50:00.0000] <TabAtkins>
I still don't understand what you mean.  Could you rephrase in different words?

[14:51:00.0000] <annevk>
hmm

[14:51:01.0000] <annevk>
with Lion my connection is dropped way more often

[14:51:02.0000] <annevk>
bah

[14:51:03.0000] <annevk>
IRC connection at least

[14:52:00.0000] <metron>
Tab: I use to work with html5 but Iam not so experienced with it, still reading books, now friends ask me how those old html4 / xhtml sites will be integrated to the new html5 type of website.

[14:54:00.0000] <AryehGregor>
metron, you can just add features incrementally.  HTML5 features work in HTML4 pages and vice versa.

[14:54:01.0000] <metron>
Tab: just by doing all changes manually? just renaming elements and stuff like that?

[14:54:02.0000] <AryehGregor>
It's all totally compatible.

[14:54:03.0000] <AryehGregor>
Use whatever features are useful to you, you don't have to rewrite your pages if you don't want to.

[14:55:00.0000] <metron>
Aryeh: ok, hmm

[14:55:01.0000] <D^mien>
see metron html5 aims to be backwards compatible with existing content

[14:56:00.0000] <D^mien>
html5 acknoledges previously existing elements even when those elements are no longer in html5

[14:56:01.0000] <D^mien>
With HTML5, anything goes. Uppercase, lowercase, quoted, unquoted, self-closing or not; it’s entirely up to you.

[14:57:00.0000] <D^mien>
i'm just getting started with it, but so far it sure is a beauty to work with :D

[14:58:00.0000] <D^mien>
the easisest way to start switching to html5, change doctype to <!DOCTYPE HTML5>

[14:58:01.0000] <metron>
D^mien: k

[14:58:02.0000] <D^mien>
:)

[14:58:03.0000] <D^mien>
A good book that got me started was HTML5 for webdesigners by Jeremy Keith

[14:59:00.0000] <jamesr>
and find a good validator/linter

[15:00:00.0000] <metron>
I think I'll use netbeans, it was very useful for me with php5.3

[15:01:00.0000] <annevk>
euh D^mien

[15:01:01.0000] <annevk>
doctype is <!DOCTYPE HTML>

[15:01:02.0000] <metron>
D^mien: good to know about, thx

[15:01:03.0000] <annevk>
no 5

[15:01:04.0000] <D^mien>
haha ya

[15:01:05.0000] <D^mien>
duh

[15:01:06.0000] <D^mien>
sorry

[15:01:07.0000] <metron>
^^

[15:01:08.0000] <D^mien>
thanks for clearing that up, forgot no5

[15:01:09.0000] <D^mien>
lol

[15:03:00.0000] <annevk>
Is it just me or is Daring Fireball reading like Fox News when it comes to Google?

[15:04:00.0000] <metron>
D^mien: Maybe I should read that book you suggested

[15:11:00.0000] <annevk>
shit

[15:11:01.0000] <annevk>
flamed on G+

[15:11:02.0000] <annevk>
never did that on Facebook

[15:14:00.0000] <AryehGregor>
You can edit your posts, right?

[15:16:00.0000] <Hixie>
annevk: d'oh!

[15:16:01.0000] <Hixie>
annevk: url!

[15:16:02.0000] <Hixie>
/me grabs popcorn

[15:17:00.0000] <dglazkov>
othermaciej: sounds good!

[15:22:00.0000] <annevk>
not that bad

[15:22:01.0000] <annevk>
https://plus.google.com/112095156983892490612/posts/EMADdvHSRbH

[15:22:02.0000] <annevk>
Julian brought his splitting argument up again, I sneered about HTTP going so fast

[15:23:00.0000] <annevk>
(HTTP is now split into 7-10 drafts or so)

[15:23:01.0000] <annevk>
(a giant mess if you ask me, I can never find what I'm looking for)

[15:23:02.0000] <Hixie>
HTTP is a disaster

[15:25:00.0000] <annevk>
I like how that RDF thread goes on again about how the toolchain will save you

[15:26:00.0000] <Hixie>
RDF equally a disaster

[15:26:01.0000] <Hixie>
RDF might even be worse

[15:26:02.0000] <annevk>
Web developers hardly use tools

[15:26:03.0000] <annevk>
And pretty much requiring them to use tools will just mean less people develop, which seems bad

[15:26:04.0000] <Hixie>
what i like about the rdf tool thing is the way the dude was like "write a script? hello no, you'd write sparql instead!"

[15:26:05.0000] <Hixie>
how is that not a script!

[15:27:00.0000] <The_8472>
that's because the standards are so bad that you have to implement a whole browser before you can start making tools.

[15:27:01.0000] <AryehGregor>
Most web authors use tools of some kind, like CMSes.

[15:27:02.0000] <Hixie>
people totally use tools all the time

[15:27:03.0000] <The_8472>
it's like your server-side code IDE has to integrate the whole client side too

[15:27:04.0000] <Hixie>
but you cannot rely on a tool to save a technology that is too complicated to hang-author

[15:27:05.0000] <Hixie>
hand

[15:27:06.0000] <zewt>
web developers hardly use tools because largely the tools suck, heh

[15:27:07.0000] <The_8472>
works with xml

[15:28:00.0000] <zewt>
been getting better lately but at a snail's pace and still far from a mature environment

[15:28:01.0000] <The_8472>
nobody touches multi-megabyte XML files by hand

[15:28:02.0000] <Hixie>
and XML has been a failure on the web :-)

[15:28:03.0000] <zewt>
you could have stopped at "failure"

[15:28:04.0000] <The_8472>
but not a failure as technology in general

[15:28:05.0000] <gsnedders>
RDF is certainly successful, though, even if it's not successful on the public web.

[15:28:06.0000] <annevk>
zewt, there are perfectly fine XML parsers and yet people seem to do it with regexp every now and then

[15:28:07.0000] <The_8472>
crazy people

[15:28:08.0000] <zewt>
annevk: unfortunately, you can't spec away incompetence

[15:29:00.0000] <annevk>
no

[15:29:01.0000] <annevk>
you just have to take that into account

[15:29:02.0000] <The_8472>
you can. but that means your spec has to include fail-fast, fail-hard provisions. which may be good for quality, but usually not for adoption.

[15:30:00.0000] <Hixie>
no it doesn't

[15:30:01.0000] <Hixie>
CSS is a great example

[15:30:02.0000] <Hixie>
it handles typos fine, without hard-failure

[15:31:00.0000] <Hixie>
only local failure

[15:31:01.0000] <The_8472>
which leads/has lead to different browser behavior depending on how it dealt with your typo

[15:31:02.0000] <hober>
I've been trying to get people to call this 'argumentum ad adminiculum'

[15:32:00.0000] <hober>
because it's a fallacy that comes up over and over and over again

[15:32:01.0000] <hober>
"it's ok that it's crazy-complex, because people don't author things, tools do"

[15:32:02.0000] <TabAtkins>
The_8472: CSS's typo recovery is very well-defined.  Differing behavior is bugs.

[15:32:03.0000] <The_8472>
TabAtkins, tell that to those years of browser hacks relying on inconsistent handling.

[15:33:00.0000] <annevk>
cool, Norm Walsh shows QR codes on printed pages

[15:33:01.0000] <Hixie>
hober: i usually reply "and who authors the tools?"

[15:33:02.0000] <hober>
Hixie: exactly. it's turtles all the way down

[15:33:03.0000] <zewt>
hober: of course, even for formats which are always output by tools (binary formats), you still have people writing tools that do it wrong, heh

[15:33:04.0000] <annevk>
not sure if QR codes are still cool, but it's definitely better than URLs users have to type over

[15:33:05.0000] <Hixie>
The_8472: that's the cause of a bad spec, not a necessary feature of error recovery

[15:33:06.0000] <Dashiva>
QR is totally last year

[15:33:07.0000] <hober>
annevk: were qr codes ever cool?

[15:33:08.0000] <TabAtkins>
The_8472: Um, they're still bugs.  I don't see what point you're trying to make.

[15:33:09.0000] <zewt>
(parsing AVIs in the wild? good luck)

[15:33:10.0000] <hober>
zewt: true, true

[15:34:00.0000] <zewt>
of course, that was never really a well-specified format to begin with

[15:34:01.0000] <The_8472>
TabAtkins, the point? that browser hacks wouldn't have existed if browsers were fail-hard and we might have seen actual feature-detecting emerging much earlier.

[15:35:00.0000] <gsnedders>
The_8472: People would've just done browser hacks based upon other nbugs.

[15:35:01.0000] <gsnedders>
*bugs

[15:35:02.0000] <TabAtkins>
Dude, you kidding?  Fail-hard CSS would have frozen the technology very early on.  Local-fail is a necessary part of how CSS allows new tech.

[15:35:03.0000] <The_8472>
i think java is a good example for tools alleviating *some* complexity. you can write java without IDEs. but having one to back you up is a huge boon if you have to work with other people's code. it being very computer-friendly (static typing allowing lots of analysis) helps of course

[15:36:00.0000] <The_8472>
TabAtkins, i'm talking about syntax. not about permitted values.

[15:36:01.0000] <TabAtkins>
They're one and the same.  An unknown property is the same as a syntax error.

[15:36:02.0000] <The_8472>
permitted values is just a matter of well-defined forwards-compatibility

[15:36:03.0000] <zewt>
writing java without an IDE is a fast path to migraines

[15:37:00.0000] <zewt>
(but only because java code tends to be split into 10x as many files as necessary, often in directory structures ten levels deep)

[15:37:01.0000] <Hixie>
well-defined forwards-compatibility is basically the same as fail-local

[15:38:00.0000] <Hixie>
you can have well-defined forwards-compatibility in syntax, in values, in semantics, etc

[15:38:01.0000] <The_8472>
xml has forwards-compatibility with namespaces and yet is fail-hard on its core syntax (preventing tag-soup)

[15:38:02.0000] <Hixie>
but in each case, it's a matter of handling a current failure in a graceful and predictable but non-fatal way

[15:38:03.0000] <Hixie>
The_8472: right, and that means it doesn't have well-defined forwards-compatibility for syntax changes

[15:39:00.0000] <The_8472>
have there been any major syntax changes in html?

[15:40:00.0000] <gsnedders>
The_8472: Adding more self-closing tags, adding foreign content… IE allowing ` as an attribute quote char, Opera allowing far more whitespace chars in tags.

[15:40:01.0000] <Hixie>
The_8472: yes

[15:40:02.0000] <Hixie>
The_8472: e.g. we added inline svg/mathml

[15:40:03.0000] <The_8472>
see, none of those things are essential features that really improve the language

[15:41:00.0000] <The_8472>
namespaces could have done that in the case of xml. in fact, they do exactly that

[15:41:01.0000] <AryehGregor>
Um, really?

[15:41:02.0000] <AryehGregor>
Never being allowed to add more self-closing tags is a good thing?

[15:41:03.0000] <The_8472>
yes

[15:41:04.0000] <AryehGregor>
Well, I'd have to disagree with that.

[15:41:05.0000] <The_8472>
if by selfclosing you mean things like <p>

[15:41:06.0000] <Hixie>
The_8472: the point is not that there are syntax changes that people want to add to xml but can't in a backwards-compatible way, the point is that if there were, it wouldn't matter, because it would be impossible to add them

[15:41:07.0000] <AryehGregor>
I mean things like <source>.

[15:42:00.0000] <AryehGregor>
Or <img>, for an older example.

[15:42:01.0000] <AryehGregor>
Where there's no closing tag.

[15:42:02.0000] <gsnedders>
The_8472: XML 1.0 Fifth Edition is a backwards incompatible change in XML, likewise XML 1.1

[15:43:00.0000] <The_8472>
1.1 -> goes into the header -> is an extension point

[15:43:01.0000] <Hixie>
not a backwards-compatible one

[15:43:02.0000] <The_8472>
obviously it would be nuts using a 1.0 parser on 1.1 code.

[15:43:03.0000] <Hixie>
so much so that 1.1 was basically killed in favour of 1.0 5th ed

[15:43:04.0000] <Hixie>
not "obviously" at all.

[15:43:05.0000] <gsnedders>
The_8472: And 5th ed?

[15:43:06.0000] <Hixie>
people use html4-era parsers on html5-era pages all the time

[15:44:00.0000] <Hixie>
that's what backwards-compatibility enables

[15:45:00.0000] <The_8472>
except that svg doesn't work. so it fails in unpredictable ways, depending on how complete the environment is in which you run it

[15:46:00.0000] <Hixie>
that's because html4 didn't define parsing

[15:46:01.0000] <Hixie>
when we add new syntax in a future revision, we'll be able to exactly predict the failure modes

[15:47:00.0000] <The_8472>
the thing is you can't assume anything will work at all. user might have javascript off. or it might be an old parser. or it might not support a random feature. so you have to test with all browser versions

[15:47:01.0000] <Hixie>
i'm just saying that fail-hard prevents having well-defined forwards-compatibility

[15:48:00.0000] <Hixie>
i'm not arguing that having well-defined forwards-compatibility makes everything bug-free

[15:48:01.0000] <Hixie>
or that web development is easy because of having well-defined forwards-compatibility

[15:49:00.0000] <The_8472>
well, ok. it does make moving forward easier. but it causes quite some headaches too. and imo in some places fail-hard is useful

[15:49:01.0000] <Hixie>
there are times fail-hard is useful, but they're rare on the web

[15:49:02.0000] <Hixie>
can't speak about headaches

[15:49:03.0000] <Hixie>
q.v. the /topic

[15:53:00.0000] <metron>
Hixie: what you ment with "fail-hard"?

[15:53:01.0000] <annevk>
almost done with DOM Traversal o_O

[15:53:02.0000] <Hixie>
metron: handling an error by entirely refusing to process the resource

[15:53:03.0000] <annevk>
well, except for like introductory text and stuff

[15:53:04.0000] <Hixie>
metron: JS and XML are examples of fail-hard languages at the syntax level

[15:54:00.0000] <annevk>
and a few more methods, and getting killed by jgraham for using hard-to-read-algorithms

[15:54:01.0000] <Hixie>
well here's a shocker

[15:54:02.0000] <Hixie>
the spec doesn't seem to require that browsers let users follow hyperlinks

[15:55:00.0000] <Hixie>
or even encourage it

[15:55:01.0000] <Hixie>
am i missing something i wrote somewhere?

[15:55:02.0000] <Hixie>
it has text for <area> and <link>...

[15:56:00.0000] <annevk>
implemented <a> with a table the other day

[16:01:00.0000] <The_8472>
mhh... replace capital punishment with the task to rewrite the spec to derive everything from tables.

[16:02:00.0000] <annevk>
it's actually a reference to an old joke; the claim was a table (an ordinary non-computer one) with quotation marks around it was a conforming HTML4 user agent

[16:02:01.0000] <Hixie>
not just any able, dbaron's desk at harvard ;-)

[16:03:00.0000] <The_8472>
/me is confused

[16:03:01.0000] <annevk>
:)

[16:05:00.0000] <Hixie>
The_8472: HTML4 includes only one testable requirement on browsers

[16:05:01.0000] <Hixie>
The_8472: which is that <q> elements must be surrounded by quotation marks

[16:05:02.0000] <Hixie>
The_8472: so an ordinary desk with a pair of quotation marks engraved in it is a compliant HTML4 browser

[16:06:00.0000] <The_8472>
ah, semantic lawyering.

[16:06:01.0000] <Hixie>
not just semantic lawyering. it has also made it near impossible to write test suites for html4.

[16:06:02.0000] <Hixie>
(fixing this is one of the greatest achievments of the new html spec)

[16:24:00.0000] <dbaron>
Hixie, The_8472, the other testable requirement was that it treat SGML-equivalent streams the same... but doing nothing all the time is treating them the same

[16:25:00.0000] <Hixie>
ah yes

[16:25:01.0000] <Hixie>
a requirement which, incidentally, no browser ever implemented

[16:37:00.0000] <TabAtkins>
Except for dbaron's desk.

[16:38:00.0000] <dbaron>
not a browser

[16:38:01.0000] <TabAtkins>
Okay, granted.

[16:39:00.0000] <Dashiva>
Dunno, did html4 define what a browser is?

[16:41:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#traversal

[16:41:01.0000] <annevk>
still need TreeWalker previousNode/nextNode

[16:45:00.0000] <metron>
annevk: I thought XLink will do that for us?!

[16:46:00.0000] <annevk>
if XLink is not out already it is on its way

[16:46:01.0000] <annevk>
this is some legacy DOM API that ended up in Acid3 but was not really fully defined

[16:46:02.0000] <annevk>
now it is, presumably with new bugs

[16:47:00.0000] <Hixie>
hober: (fwiw, i filed a bug on that circle ordering thing. i have the same problem around the same number of circles.)

[16:55:00.0000] <metron>
annevk: k, so I'll change my mind with XLink

[16:56:00.0000] <annevk>
metron, rule of thumb: specs where the initial letter is an "X" are not really making it these days


2011-07-22
[17:01:00.0000] <metron>
annevk: hmm, .oO(…she mean Xing too?) ;)

[17:01:01.0000] <TabAtkins>
(Btw, annevk's a dude.)

[17:01:02.0000] <TabAtkins>
(He's dutch, where "anne" is a common male name.)

[17:02:00.0000] <metron>
ups

[17:02:01.0000] <metron>
annevk: sry, wasn't know that

[17:03:00.0000] <TabAtkins>
Don't worry, I'm pretty sure everyone makes that mistake once, unless they see him in person first or catch someone using a gendered pronoun on him.

[17:04:00.0000] <metron>
its late, I think, I'll have to go off

[17:04:01.0000] <metron>
thx for helping me out

[17:04:02.0000] <metron>
cya guys

[17:08:00.0000] <annevk>
at the airports they typically look from me to my passport and back again etc.

[17:09:00.0000] <annevk>
some of these people even ask, "is this really you?"

[17:09:01.0000] <annevk>
wtf are passports for...

[17:10:00.0000] <Philip`>
You could have mixed up your passport with your identical other-gendered twin

[17:10:01.0000] <zewt>
(the Usual Internet Ratio just red-flagged it for me to begin with, so I checked, heh)

[17:10:02.0000] <Philip`>
Presumably if they thought you were really trying to trick them, they'd have to interrogate you more thoroughly than asking "is this really you?"

[17:16:00.0000] <Dashiva>
Nah

[17:16:01.0000] <Dashiva>
They're just waiting for someone to answer "No" sarcastically, so they can arrest you

[00:23:00.0000] <Hixie>
can anyone think of anything that would make a browsing context navigation do something other than change the document, other than:

[00:23:01.0000] <Hixie>
- content-disposition: attachment

[00:23:02.0000] <Hixie>
- unrecognised content-type triggering a download

[00:23:03.0000] <Hixie>
- http 204 or 205 response

[00:24:00.0000] <Hixie>
- network error in a browser that displays a dialog instead of an inline page for network errors

[00:24:01.0000] <Hixie>
- a scheme being one that triggers a helper app (e.g. mailto:)

[00:24:02.0000] <Hixie>
...?

[00:24:03.0000] <wirepair>
security validation error (certs)

[00:24:04.0000] <Hixie>
ooh, good call

[00:25:00.0000] <wirepair>
other than that... hmm

[00:26:00.0000] <wirepair>
maybe plugins taking over the context, like flash or something?

[00:26:01.0000] <Hixie>
that changes the document

[00:27:00.0000] <Hixie>
i can't think of others any either, fwiw

[00:27:01.0000] <Hixie>
didn't even think of the crypto one :-)

[00:27:02.0000] <wirepair>
right. then yeah that should be it hehe

[00:35:00.0000] <Hixie>
/me adds yet another reference

[00:38:00.0000] <JennaBerry>
Any W3C people around?

[00:39:00.0000] <JennaBerry>
I have a question about CSS

[00:39:01.0000] <Hixie>
you probably don't want w3c people then, more css people :-)

[00:39:02.0000] <smaug____>
JennaBerry: just ask and someone may answer

[00:40:00.0000] <JennaBerry>
I'd like to submit a section into CSS that deals with CSS-based image sprites and multiple backgrounds, so that you can reduce code and not work with extra images/libraries/scripts/styles

[00:40:01.0000] <Hixie>
already exists and supported in several browsers

[00:41:00.0000] <JennaBerry>
Oh?

[00:41:01.0000] <Hixie>
http://dev.w3.org/csswg/css3-background/

[00:44:00.0000] <JennaBerry>
Can't find the exact thing I'm looking for.

[00:44:01.0000] <JennaBerry>
I'd like something that places images, repeats or stretches them and does all sorts of things, but not to the whole object, so that you can create a button-element with a single image, cut into three parts (left, middle, right) and positioned accordingly as the background

[00:45:00.0000] <JennaBerry>
even with the current multiple backgrounds, repeat goes from one side to the other, instead of stopping at, say, 10px off the edge

[00:46:00.0000] <Hixie>
http://dev.w3.org/csswg/css3-background/#border-images

[00:47:00.0000] <JennaBerry>
The elements should also have contents

[00:48:00.0000] <Hixie>
the elements can have contents

[00:49:00.0000] <JennaBerry>
Does the border image also allow for backgrounds?

[00:49:01.0000] <Hixie>
the border image also allows for backgrounds

[00:49:02.0000] <JennaBerry>
I mean, without using a second image

[00:49:03.0000] <Hixie>
without using a second image

[00:49:04.0000] <Hixie>
(or with, both work)

[00:49:05.0000] <JennaBerry>
Interesting...

[00:50:00.0000] <Hixie>
this is already implemented in at least chrome

[00:50:01.0000] <JennaBerry>
Does it always stretch?

[00:50:02.0000] <Hixie>
dude just read the spec :-)

[00:50:03.0000] <JennaBerry>
Sorry... I've been working all night :P

[00:50:04.0000] <JennaBerry>
It's easier to ask, but yeah... I will

[00:51:00.0000] <Hixie>
sorry, kinda busy here :-)

[00:51:01.0000] <Hixie>
there's probably some tutorials you can google for

[00:51:02.0000] <Hixie>
the spec isn't the easiest thing to read

[00:51:03.0000] <JennaBerry>
I can read spec :)

[00:51:04.0000] <JennaBerry>
I did that for SVG, yesterday

[00:51:05.0000] <JennaBerry>
And then found out WebKit doesn't have effects :|

[00:53:00.0000] <smaug____>
use gecko :p

[00:54:00.0000] <JennaBerry>
I'm building a Safari Extension's configuration page... so it's not really possible to use Gecko

[02:16:00.0000] <jgraham>
Hmm, isn't overloading functions based on number of arguments rather common in non-DOM js?

[02:16:01.0000] <jgraham>
Like in jQuery?

[04:31:00.0000] <annevk>
holy shit

[04:32:00.0000] <annevk>
ALT+left directional key works in Terminal.app

[04:32:01.0000] <annevk>
*happy*

[04:33:00.0000] <jgraham>
Have they fixed it yet so you can enter characters not on your keyboard when you have the alt-sends-meta option selected?

[04:33:01.0000] <jgraham>
/me thinks not being able to type ä or whatever into terminals is a bit ridiculous

[04:33:02.0000] <smaug____>
Does Terminal.app not hang anymore when using gdb?

[04:34:00.0000] <jgraham>
In general is terminal.app no longer crappy? :)

[04:34:01.0000] <annevk>
jgraham, smaug____, no idea

[04:34:02.0000] <annevk>
it does what I need :)

[04:50:00.0000] <smaug____>
AryehGregor: seems like the range draft doesn't handle ReplaceData nor AppendData cases

[04:52:00.0000] <jgraham>
Firefox has disabled typing javascript: URIs into the address bar?

[04:53:00.0000] <smaug____>
there is a pref to enable it

[04:53:01.0000] <smaug____>
though, IIRC, even then it is run in a sandbox

[04:53:02.0000] <jgraham>
Is there some announcement about the change?

[04:55:00.0000] <smaug____>
looking...

[04:55:01.0000] <jgraham>
(or a bug number)

[04:59:00.0000] <annevk>
we should really define following/preceding

[04:59:01.0000] <smaug____>
jgraham: https://bugzilla.mozilla.org/show_bug.cgi?id=527530 is the meta bug

[05:04:00.0000] <jgraham>
smaug____: Thanks

[05:39:00.0000] <foolip_>
Hixie, ping?

[06:06:00.0000] <smaug____>
AryehGregor: also, I wonder how you ended up with the splitText handling

[06:22:00.0000] <Ms2ger>
<!-- bring it on jgraham -->

[06:32:00.0000] <jgraham>
Ms2ger: Huh?

[06:32:01.0000] <Ms2ger>
http://dvcs.w3.org/hg/domcore/file/9a02cfae1bdb/Overview.src.html#l3873

[06:33:00.0000] <jgraham>
Oh. Clearly payman_ is insane

[06:34:00.0000] <jgraham>
Sorry annevk

[06:34:01.0000] <jgraham>
+,

[06:35:00.0000] <jgraham>
Although payman_ is probably also insane

[06:48:00.0000] <bga_>
<a download="foo.txt" href="data:...">Download generated data</a>

[06:48:01.0000] <bga_>
new attr is very good idea

[06:48:02.0000] <jgraham>
annevk: "Let sibling be the preceding sibling of node" doesn't account for node not having a preceding sibling. I assume sibling is supposed to be null or something there

[06:56:00.0000] <jgraham>
Also, I am missing simething about 3.2.3 in the previouNode algorithm

[06:56:01.0000] <jgraham>
*something

[06:56:02.0000] <jgraham>
Oh wait

[06:58:00.0000] <jgraham>
It's fine

[06:58:01.0000] <jgraham>
(I really wish you made "return" imply "and terminate these steps"

[06:58:02.0000] <jgraham>
we don't have coroutines)

[07:20:00.0000] <annevk>
jgraham, it does not always mean that

[07:20:01.0000] <annevk>
though maybe in DOM Core it does

[07:26:00.0000] <jgraham>
annevk: "return" meaning multiple differrnt thigns seems like a bug that should be fixed

[07:28:00.0000] <annevk>
it does not mean multiple things

[07:28:01.0000] <annevk>
but e.g. XMLHttpRequest.send() can return and still continue running

[07:28:02.0000] <linclark>
so the HTML WG used a lot of data about how people were using classes to create the new semantic elements, right?

[07:29:00.0000] <linclark>
if that is right, does anyone have a link to the data?

[07:29:01.0000] <linclark>
I remember adactio talking about it I think, but can't find link

[07:29:02.0000] <annevk>
the raw data is not available

[07:29:03.0000] <annevk>
a summary is available here: http://code.google.com/webstats/

[07:29:04.0000] <linclark>
ahhhhh, awesome

[07:29:05.0000] <linclark>
thanks!

[07:29:06.0000] <annevk>
it is from 2005 though, hopefully at some point a new study is done

[07:30:00.0000] <adactio>
linclark: Note though, that the research was done *after* the new structural elements had already been named (if I recall correctly). So the data provided *justification* but wasn't necessarily the origin of the element names.

[07:31:00.0000] <linclark>
oh, ok, didn't realize that. thanks for the clarifications

[07:31:01.0000] <adactio>
John Allsopp also did some research 'round about the same time (2005), if I recall.

[07:32:00.0000] <adactio>
Here's John's study (a smaller sample than Google's, obviously): http://westciv.typepad.com/dog_or_higher/2005/11/real_world_sema.html

[07:41:00.0000] <Hixie>
foolip_: pong

[07:41:01.0000] <foolip_>
Hixie, I filed two bugs instead

[07:41:02.0000] <Hixie>
k thanks

[07:41:03.0000] <foolip_>
was just going to ask if putting defaultMuted on HTMLMediaElement was intentional

[07:41:04.0000] <annevk>
/me wonders if Hixie is in California

[07:42:00.0000] <Hixie>
btw the way that the IETF leaves obsolete specs on the web instead of replacing specs when they're updated makes the RFC space _even worse than the TR page_

[07:42:01.0000] <Hixie>
it's like the W3C and the IETF compete for how badly they can do web standards

[07:42:02.0000] <Hixie>
annevk: yes

[07:42:03.0000] <Hixie>
foolip_: yes

[07:42:04.0000] <Hixie>
foolip_: why would it not be?

[07:42:05.0000] <annevk>
so is it morning for you or just late? :)

[07:42:06.0000] <foolip_>
Hixie, like poster="", muted="" does nothing for <audio>

[07:42:07.0000] <Hixie>
morning

[07:43:00.0000] <Hixie>
foolip_: i thought i made muted="" do something on audio?

[07:43:01.0000] <foolip_>
oh, wait a second

[07:43:02.0000] <foolip_>
it says "When a media element is created, if it has a muted attribute specified, the user agent must set the muted IDL attribute to true, overriding any user preference."

[07:44:00.0000] <foolip_>
sloppy reading on my part, I'll close the bug as invalid and just implement the spec

[07:44:01.0000] <Hixie>
k :-)

[07:44:02.0000] <Hixie>
i don't really mind, we can make it video-only, but it seemed simpler to just make things the same everywhere where possible

[07:45:00.0000] <foolip_>
this is what confused me: "The muted attribute on the video element controls"

[07:45:01.0000] <foolip_>
but it's actually defined on both audio and video

[07:46:00.0000] <foolip_>
do you want a bug on that?

[07:47:00.0000] <hsivonen>
hmm. Microdata may get corrupted if someone does <script itemprop=foo> and a sanizer zaps scripts

[07:47:01.0000] <hsivonen>
I guess I'll worry about that if it becomes a Real Problem

[07:47:02.0000] <foolip_>
hsivonen, same thing for <object>, right?

[07:47:03.0000] <hsivonen>
foolip_: oh, good point

[07:47:04.0000] <foolip_>
and attribute whitelists...

[07:47:05.0000] <hsivonen>
foolip_: I'm whitelisting Microdata

[07:48:00.0000] <foolip_>
hsivonen, you're writing a sanitizer?

[07:48:01.0000] <hsivonen>
foolip_: yes

[07:48:02.0000] <hsivonen>
foolip_: well, it's mostly written. I'm just fixing the obvious problem that <link itemprop> and <meta itemprop> got dropped

[07:48:03.0000] <hsivonen>
and now I see further problems

[07:49:00.0000] <foolip_>
hsivonen, what would you expect <script itemprop> to do?

[07:50:00.0000] <hsivonen>
foolip_: make the textContent of the element be the value of the property

[07:50:01.0000] <foolip_>
<script src="foo" itemprop="prop"> isn't reflected by itemValue the way you might expect, though

[07:50:02.0000] <foolip_>
that inconsistency would be a bit iffy, IMO

[07:50:03.0000] <Hixie>
foolip_: sure

[07:51:00.0000] <Philip`>
/me thought the IETF deleting specs after 6 months was a worse problem than them keeping specs around for too long, since it forces you to look in less-official places to find archived copies

[07:51:01.0000] <Hixie>
sanitizers corrupt pages, that's what they're for :-)

[07:51:02.0000] <hsivonen>
dropping <object> worries me now more than dropping <script>

[07:51:03.0000] <Hixie>
Philip`: yeah that's pretty messed up too

[07:51:04.0000] <annevk>
that's fixed by just using tools.ietf.org

[07:51:05.0000] <hsivonen>
Hixie: I could make non-whitelisted elements that have Microdata turn into spans

[07:52:00.0000] <jgraham>
hsivonen: or <meta>

[07:52:01.0000] <jgraham>
Oh, but that doesn't work I guess

[07:52:02.0000] <Philip`>
tools.ietf.org links to updated versions of obsoleted drafts so I suppose it helps with both problems

[07:52:03.0000] <Philip`>
...as long as you find your way to an HTML version, not a text version

[07:52:04.0000] <hsivonen>
Hixie: but I guess this patch has been in bz's review queue for long enough that I'll leave that kind of feature creep for another bug

[07:53:00.0000] <Hixie>
hsivonen: what's the use case for round-tripping microdata through a sanitizer?

[07:53:01.0000] <Hixie>
annevk: in case you're planning on replying to Takeshi, note that i've already said on public-webapps that i don't care if we force the extension on or off, so long as we force it one way or the other

[07:57:00.0000] <hsivonen>
Hixie: the same as round-tripping Content MathML: Not destroying invisible data that doesn't need to be destroyed when someone pastes it into contenteditable

[07:58:00.0000] <Hixie>
that doesn't seem like a realistic use case

[07:58:01.0000] <Hixie>
it's just gonna cause the data to be corrupted when the user fiddles with the content

[08:01:00.0000] <hsivonen>
is switching the twitter avatar into a monkey face a new fad or a twitter bug? I think I've seen two people now use the same monkey face.

[08:08:00.0000] <hsivonen>
hmm. maybe it's just a coincidence that two people have chosen a face of a monkey as their avatars instead of their own face

[08:11:00.0000] <heedly>
What error does the video tag fire when it can't find the requested media?

[08:16:00.0000] <manu`>
Any Opera folks in Oslo, Norway? Everyone okay there?

[08:18:00.0000] <manu`>
Reason I ask is because a bomb has just gone off there - targetting the PM's government office: https://plus.google.com/102122664946994504971/posts/HRjA3xr8wXy

[08:19:00.0000] <jgraham>
That's not good…

[08:21:00.0000] <hsivonen>
BBC is running the story, so I guess the tweets are for real

[08:21:01.0000] <Hixie>
hsivonen, MikeSmith: apologies, i forgot to mark the download="" revision as affecting conformance checkers

[08:22:00.0000] <scor>
http://chatlogs.planetrdf.com/swig/2011-07-22.html#T13-45-08

[08:22:01.0000] <wilhelm>
hsivonen: Yes, there was a blast: http://www.nrk.no/nyheter/norge/1.7722919

[08:22:02.0000] <Hixie>
bbl

[08:23:00.0000] <wilhelm>
Lots of glass, but little structural damage, it seems.

[08:24:00.0000] <hsivonen>
wilhelm: I saw a tweet saying than "An entire block has exploded"

[08:24:01.0000] <manu`>
Opera HQ is 2km away from the bombing location

[08:24:02.0000] <hsivonen>
*that

[08:24:03.0000] <manu`>
Prime Minister's government HQ was attacked - PM is okay.

[08:28:00.0000] <wilhelm>
hsivonen: Well, this is where it hit: http://static01.vg.no/drfront/images/2011-07/22/88-0c8d07aa-11a997f9.jpeg

[08:28:01.0000] <wilhelm>
There are windows blown out more than a kilometer away.

[08:28:02.0000] <wilhelm>
Same building: http://gfx.dagbladet.no/labrador/174/174138/17413824/jpg/active/978x_13197743.jpg

[08:29:00.0000] <AryehGregor>
smaug____ (who isn't here): the draft doesn't have to special-case replaceData/appendData because when it was written, those just called insertData/deleteData.  Now annevk has rewritten that in DOM Core at my request, and I'll rewrite the Range stuff soon to work with the new wording.  As for splitText, see the comments in the spec's HTML source.  IIRC, I copied the behavior from WebKit, who made it up; it makes a lot of sense, and IIRC roc t

[08:29:01.0000] <AryehGregor>
hought it was a good idea.  Without that special-case I'd have to add a bunch of different execCommand() special-cases, so it's no simpler, and authors don't get the improved behavior.

[08:32:00.0000] <foolip_>
heedly, that would be MEDIA_ERR_NETWORK

[08:33:00.0000] <foolip_>
well, it doesn't "fire" MEDIA_ERR_NETWORK, it fires an error event and video.error.code is MEDIA_ERR_NETWORK

[08:40:00.0000] <annevk>
jgraham, defined next/previous sibling now and they include null; thanks!

[08:41:00.0000] <annevk>
jgraham, btw, it actually got worse from that bring it on comment

[08:41:01.0000] <annevk>
jgraham, search for "innermost" o_O

[08:44:00.0000] <jgraham>
annevk: I saw that. You are insane

[08:45:00.0000] <foolip_>
hsivonen, http://www.w3.org/Bugs/Public/show_bug.cgi?id=13329

[08:45:01.0000] <annevk>
I just try to get things done until I have a better way to say things

[08:46:00.0000] <foolip_>
right now validator.nu is saying "Attribute muted not allowed on element audio at this point."

[08:46:01.0000] <foolip_>
where should I file a bug?

[08:46:02.0000] <jgraham>
Well I think it is good that you defined this stuff :)

[08:48:00.0000] <hsivonen>
foolip_: bugzilla.validator.nu is the place, but no need to file a bug, since I can just go ahead and fix it now

[08:48:01.0000] <foolip_>
hsivonen, much appreciated

[08:49:00.0000] <foolip_>
you might want to double-check the spec so that I'm not confused again

[08:49:01.0000] <foolip_>
I'm not 100% sure what the normative part is, but the intention is fairly clear

[08:56:00.0000] <annevk>
http://www.w3.org/TR/from-origin/ WD'd

[08:58:00.0000] <Ms2ger>
annevk++

[08:59:00.0000] <hsivonen>
ouch. looks like my ISP has changed my IP address so now my servers don't let me in

[08:59:01.0000] <hsivonen>
good thing I have alternative permitted IP addresses to connect from

[09:02:00.0000] <annevk>
Ms2ger, back at you for fixing the date :)

[09:02:01.0000] <jgraham>
hsivonen: That sounds like quite an odd security measure. Was it opt-in?

[09:04:00.0000] <hsivonen>
jgraham: it's self-configured

[09:04:01.0000] <hsivonen>
jgraham: maybe my tin-foil hat is too thick

[09:04:02.0000] <hsivonen>
jgraham: but seems potentially useful in case sshd has vulnerabilities

[09:07:00.0000] <Ms2ger>
annevk, so, when do we obsolete it for a fetching spec? :)

[09:08:00.0000] <AryehGregor>
Hmm, so when you share a post on G+ it looks like it's designed so as to encourage people to comment on your share instead of the original post?  That's not what I want.

[09:09:00.0000] <hsivonen>
heh: "The problem is RDFa is horrifically badly designed for those use cases (either because it'd badly designed, or because it wasn't intended for those use cases — I've always thought the latter, but Manu insists it's the former), and has only been getting worse." (Hixie on G+)

[09:09:01.0000] <annevk>
Ms2ger, when I forget why I hate working on networking specs

[09:09:02.0000] <Ms2ger>
:)

[09:10:00.0000] <AryehGregor>
Anyway, this is a potentially interesting discussion: https://plus.google.com/u/0/112095156983892490612/posts/EMADdvHSRbH

[09:10:01.0000] <annevk>
AryehGregor, yeah, people comment on your version of the share, but clicking on share, shares the original share

[09:11:00.0000] <annevk>
I think that is really confusing, but whatever

[09:11:01.0000] <annevk>
euh, shares what is shared

[09:11:02.0000] <AryehGregor>
I'd like something that just copies the original discussion to everyone who's following me.

[09:11:03.0000] <AryehGregor>
Without giving them any new place to leave comments.

[09:11:04.0000] <AryehGregor>
Like "hey, look at this interesting discussion, go and comment there".

[09:12:00.0000] <hsivonen>
AryehGregor: whoa. I thought the comments went onto the original

[09:12:01.0000] <annevk>
just close comments on your share

[09:12:02.0000] <hsivonen>
but now that I checked, it's not the case

[09:12:03.0000] <annevk>
and mention that people should go to the original

[09:12:04.0000] <AryehGregor>
Ah, you can close comments.

[09:12:05.0000] <hsivonen>
for whatever reason, the RDFa discussion is happening on my share instead of Craig Kellog's original

[09:12:06.0000] <AryehGregor>
Is there a link to the original post?  I didn't actually see one.

[09:13:00.0000] <hsivonen>
I don't see a link

[09:14:00.0000] <annevk>
https://plus.google.com/u/0/112095156983892490612/posts/EMADdvHSRbH

[09:14:01.0000] <AryehGregor>
Can I add links to comments somehow?

[09:14:02.0000] <annevk>
oh the RDFa one

[09:15:00.0000] <dylangluck>
Hey guys, I'm new to the group. My name is Dylan Gluck and I am a Front-End Developer at Mars Design in NYC. I am a big fan of what you guys are doing, and would like to support in any way I can

[09:16:00.0000] <annevk>
be careful now ;)

[09:16:01.0000] <annevk>
http://wiki.whatwg.org/wiki/Specifications_TODO

[09:16:02.0000] <annevk>
dylangluck, there is also http://wiki.whatwg.org/wiki/What_you_can_do

[09:17:00.0000] <dylangluck>
Awesome! thanks annevk. I will definitely look over everything

[09:17:01.0000] <foolip_>
hsivonen, it's too bad that you can comment on a share at all, as opposed to commenting on the original...

[09:17:02.0000] <foolip_>
they seem to have optimized for the edge case where you reshare to exclude some people from the original post

[09:18:00.0000] <annevk>
hmm the wiki just went down? Hixie

[09:18:01.0000] <dylangluck>
yes I am getting no response

[09:18:02.0000] <annevk>
"Using an outdated version of a specification isn't comparable to using the old version of bash that ships with the enterprise Linux distro you use, it's more like planning your vacation using a decades-old globe that still has the USSR on it. You're using something that's known to be wrong, and your only hope is that the errors happen not to affect whatever part of the spec you're relying on."

[09:19:00.0000] <annevk>
I like that one AryehGregor :)

[09:19:01.0000] <Ms2ger>
AryehGregor++

[09:19:02.0000] <hsivonen>
foolip_: having a "Share" function on non-Public posts is questionable in the first place

[09:19:03.0000] <foolip_>
hsivonen, indeed, it's too pad it's not more like retweet on twitter

[09:19:04.0000] <dylangluck>
back up after db error

[09:19:05.0000] <AryehGregor>
Yeah, seems weird to me.

[09:19:06.0000] <foolip_>
bad

[09:20:00.0000] <Philip`>
/me wonders if you could write an actual science fiction novel in the form of a specification

[09:20:01.0000] <dylangluck>
haha ^

[09:20:02.0000] <hsivonen>
new-style retweets are also nicer, because they collapse

[09:20:03.0000] <dylangluck>
Thanks for the info guys, I will definitely be back

[09:20:04.0000] <MacTed>
encouraging background image watermarks that proclaim "obsolete" on relevant pages, and perhaps encouraging (creating?) <link @rel="obsoleted by"> and/or <link rel="obsoletes">, could go far with IETF, W3C, and others who want to preserve historical docs at their original URIs...

[09:20:05.0000] <MacTed>
unfortunately, <link rel="canonical"> requires starting from the position that "this doc will change" and few standards are created with their evolution in mind.

[09:20:06.0000] <annevk>
dylangluck, sorry about that; most people typically start out by giving feedback on specs

[09:20:07.0000] <hober>
We MUST niuke it from orbit. It MAY be the only way to be sure.

[09:20:08.0000] <dylangluck>
absolutely

[09:21:00.0000] <hsivonen>
when 5 people reshare a post in my stream, I'd like their commentary to stack onto one item instead of having 5 items

[09:21:01.0000] <dylangluck>
i am not a preogrammer

[09:21:02.0000] <annevk>
dylangluck, but there's also writing tests (which involves giving feedback), and writing specs if you're up for it

[09:21:03.0000] <foolip_>
Philip`, I'd read it!

[09:21:04.0000] <Ms2ger>
Writing tests would be awesome

[09:21:05.0000] <annevk>
hsivonen, you want RT

[09:21:06.0000] <Philip`>
You could hint at lots of background through examples and references and implicit assumptions about what's implementable, so I imagine it wouldn't be impossible to tell a story

[09:21:07.0000] <Ms2ger>
foolip_, more than a page of it? :)

[09:22:00.0000] <scor>
AryehGregor, hsivonen: is that the url you are looking for (original post)? https://plus.google.com/u/0/115239936584020095918/posts/KPNTayJ3eNW but the fact remains, I had to go to Gregg's profile and look for his post (no link from Henri's profile)

[09:22:01.0000] <dylangluck>
just a developer who loves the idea of expanding on what the WHATWG has done and giving some feedback

[09:22:02.0000] <Philip`>
Also you can use deprecated features or weird quirks to indicate the history of your story world

[09:22:03.0000] <scor>
in any case, this is where the conversation is happening, on hsivonen's profile: https://plus.google.com/u/0/115203359751471044302/posts/92VKitpppB4

[09:22:04.0000] <annevk>
dylangluck, sounds good

[09:23:00.0000] <foolip_>
Ms2ger, that depends on how well it manages to incorporate obscure references to spec discussions/wars that make me feel clever by recognizing them

[09:23:01.0000] <AryehGregor>
scor, I was talking about the post on Jeni Tennison's profile.

[09:23:02.0000] <Ms2ger>
Heh

[09:23:03.0000] <AryehGregor>
I wound up posting a link in my share, and disabling comments.

[09:24:00.0000] <foolip_>
is it just me, or is it very difficult to get from a Google+ share to the original post?

[09:24:01.0000] <AryehGregor>
It is.

[09:24:02.0000] <AryehGregor>
We were just saying.

[09:24:03.0000] <AryehGregor>
You have to go to the profile and scroll down until you find it.

[09:24:04.0000] <foolip_>
yeah, that's what I'm doing

[09:24:05.0000] <Ms2ger>
"Note: This requirement is not present in the Foo version of this specification due to a flame war in the 4th century"

[09:24:06.0000] <hsivonen>
s/Craig/Gregg/

[09:24:07.0000] <AryehGregor>
That's why I posted a link: https://plus.google.com/100662365103380396132/posts

[09:25:00.0000] <annevk>
the number of times I type "hg make"...

[09:33:00.0000] <hsivonen>
hmm. looks like Safari 5.1 isn't available for Leopard

[09:33:01.0000] <hsivonen>
I guess that's going to slow down MathML and inline SVG adoption

[09:34:00.0000] <AryehGregor>
Safari 5.1 has MathML support?  How good is it?

[09:34:01.0000] <AryehGregor>
Does Chrome also have it?

[09:34:02.0000] <hsivonen>
AryehGregor: based on hearsay, Chrome doesn't

[09:34:03.0000] <Rik`>
hsivonen: I don't really think so, Apple users are quickly using the latest version

[09:35:00.0000] <hsivonen>
Rik`: do you have Leopard vs. Snow Leopard stats?

[09:35:01.0000] <hsivonen>
I assumed that some users hadn't bothered to update because of "no new features"

[09:36:00.0000] <hober>
which is too bad, since SL is way better than leopard

[09:36:01.0000] <Rik`>
hsivonen: I don't at the moment but I think Leopard users are quite irrelevant

[09:36:02.0000] <Rik`>
maybe less than 1% of all web users

[09:37:00.0000] <Rik`>
and mostly it's PPC users since Snow Leopard is not compatible

[09:37:01.0000] <AryehGregor>
hsivonen, http://stats.wikimedia.org/wikimedia/squids/SquidReportClients.htm

[09:38:00.0000] <hsivonen>
Rik`: http://stats.wikimedia.org/wikimedia/squids/SquidReportOperatingSystems.htm suggests that the numbers for Intel Leopard are quite high relative to Intel Snow Leopard

[09:38:01.0000] <AryehGregor>
If you want OS: http://stats.wikimedia.org/wikimedia/squids/SquidReportOperatingSystems.htm

[09:38:02.0000] <AryehGregor>
Yeah.

[09:38:03.0000] <hsivonen>
Rik`: also, Intel Leopard much higher than PPC Leopard

[09:39:00.0000] <AryehGregor>
Leopard is at 1.78% of total OS market share, Snow Leopard is 5.04%, so Leopard is a really big chunk.

[09:39:01.0000] <Rik`>
oh right

[09:39:02.0000] <AryehGregor>
10.4 is 0.31%.

[09:39:03.0000] <AryehGregor>
Which is to say, apparently, Tiger.

[09:39:04.0000] <hsivonen>
I wonder what the security patch situation for Leopard is now that Leopard + 2 is out

[09:41:00.0000] <Rik`>
there is no security update for Snow Leopard yet

[09:43:00.0000] <hsivonen>
Rik`: security update for what holes?

[09:43:01.0000] <Rik`>
I don't know but it's usual to launch a security update for the n -1 version when you launch the n version

[09:44:00.0000] <hsivonen>
Rik`: they did release a point release of 10.6.x a bit earlier, though

[09:44:01.0000] <Rik`>
oh ok, haven't seen it

[09:45:00.0000] <hsivonen>
anyway, StatCounter's stats suggest that Safari 5.0 users on Snow Leopard aren't applying the 5.1 update immediately

[09:46:00.0000] <miketaylr>
hsivonen: i just tried software update and 5.1 is available to snow leopard, it seems

[09:46:01.0000] <Rik`>
hsivonen: http://support.apple.com/kb/HT1222

[09:46:02.0000] <Rik`>
looks like they launched Safari 5.0.6 for Leopard http://support.apple.com/kb/HT4808

[09:48:00.0000] <AryehGregor>
Why is it not called Safari 6?  Is there a full update of WebKit to something recentish?

[09:48:01.0000] <AryehGregor>
/me hopes there's a full WebKit update and they just didn't change enough UI to call it a major version bump

[09:48:02.0000] <TabAtkins>
hsivonen: I think that shares are separated so that ACLs don't collide.  But I agree that if you sent it to public, and other people reshare it to public, it would be niuce to collapse them.

[09:50:00.0000] <hober>
AryehGregor: http://svn.webkit.org/repository/webkit/branches/

[09:58:00.0000] <annevk>
emailed WebApps about Traversal

[09:58:01.0000] <annevk>
first draft done

[09:59:00.0000] <AryehGregor>
Fun discovery: default link color in WebKit (at least Chrome) is -webkit-link, which appears to compute to rgb(0, 0, 238).

[09:59:01.0000] <AryehGregor>
Not, um, blue?

[09:59:02.0000] <AryehGregor>
Yay interop?

[10:00:00.0000] <AryehGregor>
A.k.a. #00e, I guess.

[10:01:00.0000] <AryehGregor>
This kind of thing does affect interop in some cases . . . grumble grumble.

[10:01:01.0000] <Ms2ger>
Does anybody use actual blue?

[10:02:00.0000] <AryehGregor>
HTML5 says to . . .

[10:02:01.0000] <AryehGregor>
/me tests

[10:02:02.0000] <hsivonen>
foolip_: I deployed <audio muted> to validator.nu

[10:02:03.0000] <AryehGregor>
Hmm.

[10:02:04.0000] <AryehGregor>
Ms2ger, good point, this looks like a spec bug.

[10:02:05.0000] <foolip_>
hsivonen, confirmed working, thanks!

[10:03:00.0000] <AryehGregor>
So Gecko and WebKit are #00e.

[10:06:00.0000] <AryehGregor>
IE10PP2 is rgb(0, 102, 204), which is what?  #06c.  And Opera 11.50 is #00c.

[10:06:01.0000] <AryehGregor>
Oh, wait.

[10:06:02.0000] <AryehGregor>
No, it's fine, that's not visited.

[10:06:03.0000] <AryehGregor>
/me tests :visited, although Gecko will lie about it

[10:06:04.0000] <Philip`>
(Can't you test it trivially with a screenshot and Gimp/etc's colour picker tool?)

[10:06:05.0000] <AryehGregor>
Yes, if I wanted to do that much work.  :)

[10:07:00.0000] <AryehGregor>
For visited, Gecko lies.  Opera is #800080.  IE10PP2 seems to be still reporting #06c, is it lying too?

[10:07:01.0000] <AryehGregor>
/me uses dev tools

[10:07:02.0000] <AryehGregor>
I'd use dev tools for Firefox too, except you know, Firebug doesn't work on Aurora.

[10:07:03.0000] <AryehGregor>
(or are there basic built-in DOM inspection things these days?)

[10:08:00.0000] <AryehGregor>
#551A8B in Chrome, it looks like.

[10:09:00.0000] <AryehGregor>
There's an Inspect tool somewhere, but it doesn't seem to do anything very useful . . .

[10:10:00.0000] <Philip`>
Firefox's about:config says browser.visited_color = #551A8B

[10:10:01.0000] <Rik`>
AryehGregor: I think Firebug 1.8 works

[10:10:02.0000] <AryehGregor>
Rik`, not unless they updated it in the last few days.

[10:10:03.0000] <AryehGregor>
It works with current beta, but not Aurora.

[10:11:00.0000] <AryehGregor>
Ah, they did update it.

[10:11:01.0000] <AryehGregor>
/me tries

[10:11:02.0000] <AryehGregor>
Yay, works again!

[10:12:00.0000] <AryehGregor>
/me makes a mental note never to upgrade Aurora again without checking Firebug compat

[10:12:01.0000] <AryehGregor>
Firebug lies about visited link color too, though.

[10:12:02.0000] <AryehGregor>
So I'll rely on Philip` here.

[10:12:03.0000] <AryehGregor>
/me doesn't bother trying to figure out IE

[10:12:04.0000] <Rik`>
/me makes a mental note never to use Firebug until https://bugzilla.mozilla.org/show_bug.cgi?id=669730 is fixed

[10:12:05.0000] <Philip`>
I might be lying too, of course

[10:13:00.0000] <AryehGregor>
Rik`, I don't use Firefox for anything other than testing, so it can leak all it wants.

[10:13:01.0000] <AryehGregor>
Also, that's why my new laptop has 8G of RAM.

[10:13:02.0000] <Rik`>
I do use Firefox so I do my debugging in other browsers

[10:15:00.0000] <AryehGregor>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13330

[10:18:00.0000] <AryehGregor>
Links that are actually blue do look noticeably bright.

[10:45:00.0000] <david_carlisle>
AryehGregor: Chrome doesn't have MathML, presumably they could pick it up, as it's implemented in the webkit base

[10:53:00.0000] <AryehGregor>
:(

[11:03:00.0000] <dglazkov|away>
MathML --> splosion.

[11:03:01.0000] <dglazkov>
good morning, Whatwg

[11:03:02.0000] <Ms2ger>
Evening, dglazkov

[11:07:00.0000] <dglazkov>
Ms2ger: :)

[11:13:00.0000] <annevk>
used to be that dglazkov checked in in the afternoon

[11:14:00.0000] <annevk>
:p

[11:14:01.0000] <dglazkov>
had to run an errand this morning.

[11:30:00.0000] <sicking>
Hixie: would you mind updating the EventSource spec to add the {withCrededentials: true} syntax? So that we can avoid prefixing it

[11:33:00.0000] <annevk>
just adding it to the spec doesn't exactly make it a done deal

[11:34:00.0000] <annevk>
do we know if WebKit agrees?

[11:37:00.0000] <AryehGregor>
sicking, isn't the theory that standard but experimental stuff is prefixed too?  Or is that only CSS for some reason?

[11:44:00.0000] <TabAtkins>
CSS has a firm policy that everyone voluntarily follows.  JS is pretty loosy-goosy about it.

[11:46:00.0000] <dglazkov>
JS is pretty loosy-goosy about _everything_.

[12:06:00.0000] <AryehGregor>
Wow, it's 102 degrees outside.

[12:07:00.0000] <AryehGregor>
(For those of you who like sane temperature scales, that would be 312 K.)

[12:07:01.0000] <TabAtkins>
Ah, thank you, I can never think in F.

[12:07:02.0000] <TabAtkins>
Or C.

[12:08:00.0000] <Philip`>
I prefer to think of it as 1.78 radians

[12:08:01.0000] <Ms2ger>
Well played

[12:09:00.0000] <scor>
AryehGregor: where is that?

[12:09:01.0000] <AryehGregor>
scor, Manhattan.

[12:10:00.0000] <scor>
only 4 degrees cooler up here in Boston

[12:46:00.0000] <scor>
was there ever a time where the content attribute was allowed in tags other than meta to override textContent in microdata? or has @content always been only allowed in <meta>?

[12:47:00.0000] <annevk>
always

[12:49:00.0000] <scor>
annevk: thanks

[12:58:00.0000] <Hixie>
sicking: are other browsers in agreement?

[13:01:00.0000] <annevk>
http://www.w3.org/2001/tag/2011/07/21-minutes.html#item04 lol the task force conclusions are not in line with what i want, please do it again!

[13:03:00.0000] <Hixie>
i don't even understand those minutes

[13:04:00.0000] <annevk>
armchair troublemakers

[13:16:00.0000] <annevk>
/me replies to an email from November 2009

[13:37:00.0000] <annevk>
Would be pretty nifty if ij updates older specifications in this simple process manner

[13:38:00.0000] <annevk>
HTML4 could use a note :)

[13:40:00.0000] <Ms2ger>
/me lacks context

[13:41:00.0000] <annevk>
public-webapps

[13:47:00.0000] <annevk>
where is mike?

[13:47:01.0000] <annevk>
should remove DOM Traversal from http://platform.html5.org/

[13:49:00.0000] <Ms2ger>
Could try...

[13:49:01.0000] <Ms2ger>
Aww

[13:57:00.0000] <annevk>
Ms2ger, you mean the reply from ij?

[13:57:01.0000] <annevk>
I guess I should not have gotten my hopes up

[14:03:00.0000] <annevk>
oh hey

[14:03:01.0000] <annevk>
Opera invented INVALID_STATE_ERR for recursive traversal invocations

[14:03:02.0000] <annevk>
and Gecko copied it

[14:03:03.0000] <annevk>
you don't see that very often

[14:04:00.0000] <annevk>
thanks for the pointer sicking :)

[14:07:00.0000] <jamesr>
annevk: i thought opera invented everything

[14:08:00.0000] <karlcow>
http://elie.im/blog/security/tracking-users-that-block-cookies-with-a-http-redirect/

[14:08:01.0000] <karlcow>
jamesr: this is a truism

[14:08:02.0000] <annevk>
jamesr, I think that particular rhetoric typically only applies to UI stuff

[14:09:00.0000] <annevk>
jamesr, assuming you are referring to the pointless debates on who did tabs / speed dial / mouse gestures / whatnot first

[14:09:01.0000] <karlcow>
it applies to love, joy, tears, emotions, etc. I mean how could not feel anything at Carmen

[14:10:00.0000] <Ms2ger>
Don't you know that we all use "Opera did it first" to laugh at you guys? :)

[14:11:00.0000] <annevk>
Ms2ger, heh, I didn't

[14:11:01.0000] <annevk>
Ms2ger, but I can certainly see the fun in that :)

[15:08:00.0000] <annevk>
I wonder how Page Visibility went to Last Call

[15:08:01.0000] <annevk>
prolly because all these new WGs have like zero oversight

[15:10:00.0000] <Ms2ger>
The Process only applies to relevant WGs

[15:10:01.0000] <Yuhong>
So, on https://plus.google.com/112095156983892490612/posts/EMADdvHSRbH, describe the "substantial changes" that would be needed in detail.

[15:10:02.0000] <Ms2ger>
/me yawns

[15:11:00.0000] <annevk>
Ms2ger, curb your enthusiasm is back

[15:11:01.0000] <Yuhong>
And won't the substantial changes be limited by compatibility with existing content and implementations?

[15:12:00.0000] <annevk>
I searched for "substantial changes" in that link and it yielded nothing

[15:12:01.0000] <Ms2ger>
/me throws a PARSE_ERR at annevk 

[15:12:02.0000] <annevk>
Ms2ger, search for the first three words and watch it :)

[15:12:03.0000] <jamesr>
i didn't know the w3 had a process to rescind a recommendation

[15:12:04.0000] <Yuhong>
Sorry, look at the mailing list post linked.

[15:13:00.0000] <Ms2ger>
I did! :)

[15:13:01.0000] <Ms2ger>
I also know the history of boolean attributes

[15:13:02.0000] <Ms2ger>
Both pieces of knowledge are equally useless

[15:13:03.0000] <Yuhong>
Sorry, substantive changes .

[15:14:00.0000] <AryehGregor>
Yuhong, Maciej's post starting the thread there is pretty clear about what qualifies as a substantive change.

[15:14:01.0000] <AryehGregor>
It's very broad.

[15:14:02.0000] <AryehGregor>
Basically, any change that anyone other than a bike-shedder has reason to care about.

[15:14:03.0000] <annevk>
Ms2ger, for when your children ask you to tell them once more the story about boolean attributes

[15:15:00.0000] <Yuhong>
BTW, I just recently learned the history of HTML 3.0 and Netscape too.

[15:15:01.0000] <Ms2ger>
"Daddy, what's SGML"?

[15:15:02.0000] <annevk>
Ms2ger, zcorpan meanwhile tells his about SHORTTAG NET and how things were before he figured out legacy color parsing

[15:15:03.0000] <Ms2ger>
Hah

[15:16:00.0000] <TabAtkins>
I'm very glad I didn't have to figure out legacy color parsing myself.

[15:16:01.0000] <TabAtkins>
That shit's crazy.

[15:16:02.0000] <Ms2ger>
TabAtkins, how about trying to figure out CSS parsing in quirks mode? :)

[15:16:03.0000] <TabAtkins>
I'm gonna... I'm gonna not do that.

[15:16:04.0000] <annevk>
Yuhong, you want to read the W3C Process document for a definition of that term

[15:17:00.0000] <annevk>
Yuhong, if you are after "substantive changes"

[15:20:00.0000] <Ms2ger>
Notes, notes, notes

[15:20:01.0000] <Ms2ger>
One for CSS2.1 too? :)

[15:20:02.0000] <TabAtkins>
2.1 is the most recent and accurate reference!

[15:20:03.0000] <Ms2ger>
Still a disaster :)

[15:21:00.0000] <TabAtkins>
And we have a note on CSS2 along those lines.

[15:21:01.0000] <Ms2ger>
But less of a disaster than the rest of CSS, I'll grant you that

[15:22:00.0000] <TabAtkins>
Relatively speaking, 2.1 is one of the best specs the W3C has produced.  We actually have a large (though still not large enough) test suite.

[15:23:00.0000] <Yuhong>
In fact, I learned that the history of CSS1 is related to the history of HTML 3.0.

[15:23:01.0000] <annevk>
XSLT and XML are pretty watertight (and way simpler)

[15:23:02.0000] <Yuhong>
More history: http://www.reddit.com/r/web_design/comments/exnr8/on_netscape_and_css_history_heres_the_history/

[15:23:03.0000] <annevk>
CSS 2.1 has a lot of handwaving still :(

[15:24:00.0000] <TabAtkins>
annevk: I said "relatively" ^_^

[15:30:00.0000] <annevk>
I wonder why you need a whole WG for something like page visiblity

[15:30:01.0000] <annevk>
two new attributes on Document

[15:33:00.0000] <Hixie>
beats me

[15:39:00.0000] <annevk>
jgraham, you look sad on your G+ photo

[15:39:01.0000] <annevk>
well, more disinterested, which might be accurate I suppose :p

[15:40:00.0000] <Yuhong>
So what are examples of "substantive changes" that might be needed in the future?

[15:41:00.0000] <annevk>
Yuhong, http://www.w3.org/2005/10/Process-20051014/tr.html#substantive-change

[15:42:00.0000] <Yuhong>
Sorry, there was no Glossary, making it hard to find.

[15:43:00.0000] <annevk>
glad you looked at least

[15:43:01.0000] <annevk>
:)

[15:44:00.0000] <Ms2ger>
annevk, want to tell them they need a reference to WebIDL as well?

[15:44:01.0000] <annevk>
Ms2ger, can do

[15:45:00.0000] <Yuhong>
So what are examples of "substantive changes" that might be needed in the future, or the HTML5 spec quickly becomes inaccurate?

[15:45:01.0000] <Yuhong>
I am looking for specific examples.

[15:46:00.0000] <Hixie>
how can we know...?

[15:46:01.0000] <annevk>
we might have to remove an attribute that's not implemented

[15:46:02.0000] <annevk>
we might have to change an attribute because it's implemented incorrectly

[15:46:03.0000] <Ms2ger>
/me stays out of this

[15:47:00.0000] <annevk>
good idea

[15:47:01.0000] <annevk>
if Ms2ger is still reading the logs: http://lists.w3.org/Archives/Public/public-web-perf/2011Jul/0080.html

[15:47:02.0000] <Yuhong>
Candidate Recommendation is exactly the place to prove everything is implemented.

[15:47:03.0000] <Yuhong>
If it is not, it will be removed.

[15:48:00.0000] <annevk>
hey man, if you already know, don't ask

[15:49:00.0000] <Yuhong>
My point is that it is an already solved problem.

[15:50:00.0000] <Yuhong>
I mean, the W3C process already solves this problem.

[15:51:00.0000] <Yuhong>
With Candidate Recommendation and the tests etc.

[15:51:01.0000] <annevk>
yes, you have to revert to WD after a substantive change

[15:51:02.0000] <Yuhong>
Yes, but why would it loop forever?

[15:53:00.0000] <Yuhong>
Yes, I am well aware of CSS 2.1, and how it looped several times.

[15:53:01.0000] <annevk>
because with a spec this size you are bound to find something

[15:53:02.0000] <annevk>
CSS 2.1 would have looped way more if the WG fixed all outstanding issues

[15:53:03.0000] <TabAtkins>
2.1 stopped looping because we made several things undefined.

[15:53:04.0000] <TabAtkins>
Unrelated: This is ridiculously good - http://research.microsoft.com/en-us/um/people/kopf/pixelart/supplementary/multi_comparison.html

[15:54:00.0000] <TabAtkins>
The "Ours" scaling algo

[15:56:00.0000] <Yuhong>
Yep, I know. In contrast, CSS3 is modularized and so far it loops less I think.

[15:57:00.0000] <TabAtkins>
Indeed, but that brings its own problems.

[15:57:01.0000] <AryehGregor>
The whole issue of looping is only because the W3C Process demands that specs progress as a unit.

[15:57:02.0000] <AryehGregor>
WHATWG HTML has per-section annotations and it works fine.

[15:57:03.0000] <AryehGregor>
It would be really nice if we could get a usable subset spec that only includes things that are known to be followed more or less completely by at least two major browsers, though, or something like that.

[15:58:00.0000] <AryehGregor>
Or at least leaving out the stuff that's really speculative or cutting-edge.

[15:58:01.0000] <AryehGregor>
(maybe leaving in the stuff that's implemented by everyone but not interoperably)

[15:59:00.0000] <AryehGregor>
Hmm, Maciej never got back to me about innerText yet, did he?

[15:59:01.0000] <AryehGregor>
I'll have to poke him again.

[16:00:00.0000] <Yuhong>
I think being implemented by at least two implementations interoperably is part of the W3C process already.

[16:01:00.0000] <Hixie>
css3 is older than css2.1 :-)

[16:01:01.0000] <Hixie>
so clearly it doesn't loop less :-)

[16:01:02.0000] <TabAtkins>
I don't acknowledge the existince of pre-2.1 css3.

[16:02:00.0000] <Yuhong>
The year-long gap in CSS3 working drafts that existed since 1999 I think is partly because there was no point in doing it when IE6 had a monopoly and did not implement much of CSS2.

[16:03:00.0000] <Yuhong>
And remember that CSS2.1 did not exist when IE6 was released.

[16:06:00.0000] <dglazkov>
TabAtkins: this page made we want to play Mario

[16:06:01.0000] <TabAtkins>
I know!

[16:07:00.0000] <Hixie>
Yuhong: from being in the wg, i assure you your assumptions are wrong

[16:07:01.0000] <Hixie>
Yuhong: it was more because we were focusing on css2.1 and had insufficient editors

[16:07:02.0000] <Hixie>
a problem the wg still has

[16:08:00.0000] <Hixie>
TabAtkins: amazing how some of those pictures scale amazingly well with their algorithm and others look really dumb

[16:08:01.0000] <Hixie>
TabAtkins: does seem to be almost always the best or second-best algo though

[16:09:00.0000] <TabAtkins>
Hixie: Yeah.  It looks like it does worst on the ones that had the most care given to them.  Lost Vikings, frex.

[16:09:01.0000] <TabAtkins>
Lesson: Don't make your sprites too pretty.  Future higher-rez generations will appreciate your highly-contrasting colors.

[16:10:00.0000] <Hixie>
TabAtkins: yeah, or the icons that were meant to have hard edges, like the Setup icon which turns into a clay cartoon

[16:10:01.0000] <Hixie>
kinda funny how bicubic is really just squinting for you

[16:10:02.0000] <Hixie>
i hadn't realised that before

[16:11:00.0000] <Yuhong>
That is why I say partly.

[16:11:01.0000] <TabAtkins>
Yeah, it's very overtly a blur.

[16:12:00.0000] <Hixie>
Yuhong: it wasn't even partly. there was no relationship to IE.

[16:12:01.0000] <Hixie>
hq4x seems to not actually work beyond 4x zoom?

[16:13:00.0000] <TabAtkins>
That would be the origin of the name, yes.

[16:13:01.0000] <Hixie>
aah

[16:13:02.0000] <Hixie>
ok

[16:13:03.0000] <Hixie>
seems unfair to show that one for the higher zooms then

[16:13:04.0000] <TabAtkins>
(There's an hq2x and 3x too.)

[16:13:05.0000] <Hixie>
for 4x it's better than theirs for some of them

[16:13:06.0000] <Hixie>
e.g. setup

[16:13:07.0000] <Hixie>
and control panel

[16:14:00.0000] <Hixie>
basically this algorithm does really well when the underlying graphic has curves, and not so well when it has straights

[16:15:00.0000] <Hixie>
makes sense that one algorithm wouldn't be able to distinguish the two cases, of course

[16:15:01.0000] <Hixie>
bbiab.

[16:16:00.0000] <zewt>
it looks identical to "EPX" (whatever that is), but with some kind of smoothing applied

[16:16:01.0000] <TabAtkins>
Yeah, looks like the two make pretty similar coloring decisions.

[16:16:02.0000] <zewt>
and you can sort of intuitively see what EPX is doing, though i havn't squinted at it hard enough to know exactly

[16:17:00.0000] <TabAtkins>
It's a really easy algorithm.  Wikipedia has details.

[16:18:00.0000] <Yuhong>
Fun fact: One of the modules dates back to a proposal from 1996: http://www.w3.org/Style/CSS/multi-column.html

[16:18:01.0000] <Yuhong>
Also see: http://lists.w3.org/Archives/Public/www-html/1996Jun/0234.html

[16:19:00.0000] <Yuhong>
And consider the fact it will be IE10 before that module will finally get implemented.

[16:22:00.0000] <zewt>
the scaling might not actually work well for animated sprites, though, since the interpolation would probably wobble around, lacking any cohesion across frames

[16:23:00.0000] <TabAtkins>
zewt: I wonder if I can find the algo and implement it myself.

[16:23:01.0000] <zewt>
it looks like it's using EPX in vertex form and just applying some standard smoothing algorithm

[16:23:02.0000] <TabAtkins>
zewt: I think you're right that it's identical to EPX except for the region-shaping algorithm.

[16:25:00.0000] <jamesr>
there's a siggraph paper, i believe

[16:31:00.0000] <annevk>
when are these hangouts coming without plugins?

[16:34:00.0000] <sephr>
What's the reasoning behind limiting crypto.getRandomValues() to integers only? It introduces an extra unnecessary step when you want random floats: var nums = new Float64Array(10); crypto.getRandomValues(new Uint8Array(nums.buffer));

[16:35:00.0000] <sephr>
I don't see the point of having to do new Uint8Array(floats.buffer)

[16:35:01.0000] <Hixie>
annevk: at a minimum, browsers would have to support PeerConnection and getUserMedia() (or their equivalents in other proposals) before that can happen

[16:36:00.0000] <sephr>
and not to mention that it doesn't make sense to not allow crypto.getRandomValues() on an plain ArrayBuffer acting as a binary blob that you don't need indexes on

[16:42:00.0000] <TabAtkins>
zewt: As for animation, http://www.youtube.com/watch?v=o2Fd-4NzB0w

[16:59:00.0000] <fishd>
Hixie: i'm curious why you restricted a@download to require a user gesture


2011-07-23
[17:01:00.0000] <TabAtkins>
Me too, given that an iframe pointint to a C-D resource can be script-inserted.

[17:02:00.0000] <fishd>
this is reference to https://bugs.webkit.org/show_bug.cgi?id=64580#c51

[17:02:01.0000] <fishd>
TabAtkins: exactly

[17:03:00.0000] <jamesr>
sephr: floating point random numbers are really hard

[17:03:01.0000] <jamesr>
for one, the range isn't uniform the way developers will typically expect

[17:03:02.0000] <sephr>
so is randomizing the bytes not sufficiently cryptographically random?

[17:03:03.0000] <jamesr>
it doesn't have the distribution that people will normally expect

[17:04:00.0000] <jamesr>
frex random values in the range [0, 1) will be biased since the epsilon between representable values is non-uniform on that range

[17:04:01.0000] <sephr>
I understand that the distrobution may confuse some, but specifically excluding floats doesn't really help for people who honestly want random floats

[17:04:02.0000] <sephr>
ok thanks

[17:05:00.0000] <jamesr>
well when do you want cryptographically random floats?

[17:05:01.0000] <jamesr>
i don't know of any cryptographic routines that use floating point input values like this

[17:05:02.0000] <sephr>
I don't; I was just wondering

[17:05:03.0000] <sephr>
lol ok :p

[17:05:04.0000] <sephr>
thanks for the explanation

[17:05:05.0000] <jamesr>
not claiming that i know a lot of crypto stuff

[17:05:06.0000] <jamesr>
but i think it'll bite people who want secure numbers

[17:06:00.0000] <sephr>
btw, Hixie: on the whole text/html-sandboxed thing being insufficient: a "sandbox" directive was added to CSP

[17:06:01.0000] <sephr>
so you may or may not want to just remove the text/html-sandboxed registration

[17:06:02.0000] <sephr>
http://dvcs.w3.org/hg/content-security-policy/raw-file/tip/csp-specification.dev.html#sandbox

[17:16:00.0000] <zewt>
including floats doesn't help people who honestly want random floats :)

[17:16:01.0000] <jamesr>
did you mean excluding?

[17:16:02.0000] <zewt>
no, including

[17:17:00.0000] <zewt>
he said excluding doesn't help; i say including doesn't help either

[17:19:00.0000] <sephr>
zewt: whar about plain arraybuffers btw?

[17:20:00.0000] <sephr>
surely there are times when you want a binary blob and you don't care about it being indexed into a uint8array

[17:20:01.0000] <jamesr>
you should only use the crypto API when you care about cryptographic properties of the values

[17:23:00.0000] <sephr>
jacobolus: think truecrypt's hidden volume impl

[17:23:01.0000] <jacobolus>
jamesr: ↑ ?

[17:23:02.0000] <sephr>
just writing a random blob, not actually using the values in a computation

[17:24:00.0000] <zewt>
that's using the data at an array of 8-bit ints

[17:24:01.0000] <sephr>
yeah but the file api doesn't take 8 bit int arrays

[17:25:00.0000] <sephr>
it takes arraybuffers

[17:25:01.0000] <sephr>
I get that getting .buffer is easy enough

[17:25:02.0000] <zewt>
which you can create an 8-bit int view for to create the data--the underlying algorithm is using an array of 8-bit ints

[17:26:00.0000] <zewt>
there's always a data type associated with data, for any algorithm

[17:26:01.0000] <sephr>
well most likely

[17:26:02.0000] <sephr>
might be a 7bit filesys though

[17:26:03.0000] <sephr>
even though it surely isn't

[17:27:00.0000] <sephr>
anyways I get you so ok end of discussion

[17:30:00.0000] <zewt>
that's not exactly how discussions work :)

[18:28:00.0000] <TabAtkins>
heycam: How should I respond to issues in the SVG issue tracker?

[18:30:00.0000] <heycam>
TabAtkins, just send mail on www-svg and include the ISSUE-nnnn in the subject or body of the mail

[18:30:01.0000] <heycam>
subject is good, then repliers don't need to remember to reinclude it in the body

[18:30:02.0000] <heycam>
(so in fact I should have done that in my mail just now)

[18:31:00.0000] <heycam>
(anyway I put a link to the start of the stroke gradient thread in the initial text of the issue, so it doesn't matter too much of other mails in that thread don't get tagged correctly)

[18:33:00.0000] <TabAtkins>
heycam: ...huh.  I'm apparently not subscribed to enough of the SVG lists.  I didn't get any of this thread.

[18:33:01.0000] <heycam>
TabAtkins, you're not on www-svg?

[18:33:02.0000] <TabAtkins>
Apparently not.

[18:33:03.0000] <heycam>
huh

[18:33:04.0000] <TabAtkins>
Anyway, just subscribed.

[18:34:00.0000] <heycam>
so from now on we're trying to do technical discussion on that list rather than public-svg-wg, so it's easier for non-WG members to contribute

[18:34:01.0000] <TabAtkins>
Is public-svg-wg not a public list?

[18:34:02.0000] <TabAtkins>
Or is it just easier to suscribe to for some reason?

[18:35:00.0000] <heycam>
you can read public archives of it, but non WG members can't subscribe to it

[18:35:01.0000] <TabAtkins>
Ah, gotcha.  That's silly.

[18:35:02.0000] <heycam>
yeah, it's a weird half way compared to other WGs

[19:47:00.0000] <Hixie>
anyone have any tests for this crazy filename* syntax the http guys are peddling?

[19:47:01.0000] <Hixie>
julian says everyone implements it

[19:49:00.0000] <TabAtkins>
Link?

[19:49:01.0000] <Hixie>
whatwg e-mail

[19:49:02.0000] <TabAtkins>
Oh, that encoding stuff?

[19:51:00.0000] <Hixie>
6266

[19:55:00.0000] <TabAtkins>
What is this number supposed to refer to?

[19:56:00.0000] <TabAtkins>
Do you mean RFC 2231/5987?

[20:54:00.0000] <Hixie>
TabAtkins: i meant RFC 6266

[01:07:00.0000] <annevk>
Hixie, I think Julian made tests

[01:07:01.0000] <annevk>
Hixie, http://greenbytes.de/tech/tc2231/

[03:10:00.0000] <nessy>
is anyone else enjoying the discussion at http://www.w3.org/Bugs/Public/show_bug.cgi?id=13333 ? :-)

[04:34:00.0000] <doublec>
that looks like a good bug to avoid reading

[04:35:00.0000] <Ms2ger>
annevk5, thanks

[07:54:00.0000] <Stevezau>
Using html5lib i keep getting this error "  File "/usr/lib/pymodules/python2.7/html5lib/inputstream.py", line 50, in seek" "    assert pos < self._bufferedBytes()" "AssertionError"

[07:54:01.0000] <Stevezau>
it's strange as it works on ubuntu 11.04 but not 11.10

[07:55:00.0000] <Stevezau>
same code..

[07:59:00.0000] <Stevezau>
any ideas?

[08:44:00.0000] <jgraham>
Stevezau: Got some sample input?

[08:45:00.0000] <jgraham>
I don't have 11.04 handy so I'm not quite sure how I will track it down.

[08:45:01.0000] <jgraham>
/me is actually trying to make some fixes to html5lib for the first time in a while

[08:52:00.0000] <smaug____>
AryehGregor: ping

[09:05:00.0000] <foolip>
hsivonen, I'm having trouble checking out the validator source because the w3.org DTDs are timing out, is it possible to skip those if I'm just interested in HTML5?

[09:19:00.0000] <Stevezau>
jgraham nevermind, i tracked it to urllib2

[09:19:01.0000] <Stevezau>
issue with python 2.7.2 latest release

[09:36:00.0000] <jgraham>
Stevezau: OK

[09:44:00.0000] <jgraham>
So this sems like a spec bug, but maybe I am not thinking straight

[09:46:00.0000] <jgraham>
<html><table><math><mi>foo</mi>

[09:47:00.0000] <jgraham>
AFAICT the foo causes the treebuilder to go into the in table text insertion mode

[09:47:01.0000] <jgraham>
Which buffers the character tokens

[09:47:02.0000] <jgraham>
But the </mi> is processed in foreign content which doesn't flush the buffer

[09:48:00.0000] <jgraham>
so the foo ends up outside the <mi>

[09:48:01.0000] <jgraham>
Did I miss something?

[10:04:00.0000] <Jack9>
How do you apply a css class to a strokeRect? (not using the strokestyle, which can only be a css color,canvasgradient,canvaspattern)

[10:05:00.0000] <Ms2ger>
You don't

[10:07:00.0000] <Jack9>
is there a standard way to do animated strokeStyles other than looping through drawImage?

[12:33:00.0000] <foolip>
WTF? http://whattf.org/

[12:46:00.0000] <Ms2ger>
Yes, WhatTF

[15:22:00.0000] <annevk>
adding domintro boxes sucks

[15:26:00.0000] <Hixie>
yes

[15:29:00.0000] <annevk>
a lot of it is just duplicate information phrased slightly differently

[15:29:01.0000] <annevk>
i added a few extra to DOM Core though and I guess I will try to complete it

[15:44:00.0000] <abarth>
Hixie: i think you enjoy trolling the RDFa folks too much


2011-07-24
[18:15:00.0000] <sephr>
is there such a thing as `CustomEvent document.customEvent(...)`?

[18:15:01.0000] <sephr>
I just noticed that on MDN https://developer.mozilla.org/en/DOM/customEvent and it looks like someone just made up the page

[18:31:00.0000] <heycam>
sephr, I don't think that page is accurate

[18:31:01.0000] <heycam>
no such thing as document.customEvent in Gecko afaict

[18:31:02.0000] <sephr>
yeah I think someone messed up

[18:31:03.0000] <sephr>
I'm moving the page to Event/CustomEvent fixing the docs now

[18:32:00.0000] <heycam>
cool, thanks

[18:38:00.0000] <sephr>
heycam: speaking of events, does initStorageEvent actually modify dom storage?

[18:38:01.0000] <sephr>
I would think it doesn't

[18:38:02.0000] <sephr>
/me is trying to base the new customevent page off the storageevent one

[18:41:00.0000] <heycam>
sephr, no, it should just initialise the event object

[18:41:01.0000] <heycam>
sephr, shouldn't have any other effect

[18:42:00.0000] <sephr>
k thanks

[20:26:00.0000] <Hixie>
abarth: not trying to troll them. It's just frustrating that they don't understand the difference between RDFa and microdata. The very thing they keep saying is the problem with microdata is the reason microdata is giving them a problem in the first place.

[20:26:01.0000] <abarth>
i like the comparison between the pencil and the industrial laser printer

[20:27:00.0000] <abarth>
"RDFa: A horribly-designed technology that..."

[20:27:01.0000] <abarth>
that's the part that sounds trollish

[20:30:00.0000] <Hixie>
I previously had assumed that RDFa was a well-designed technology to address a problem I didn't understand.

[20:31:00.0000] <Hixie>
But Manu has been insistent in telling me that I am wrong, that RDFa was designed to address a specific set of use cases.

[20:31:01.0000] <Hixie>
well given those use cases, I really don't know how else to characterise RDFa but "horribly-designed".

[22:12:00.0000] <Hixie>
wow: https://plus.google.com/u/0/112095156983892490612/posts/BSShqcMm49B

[22:12:01.0000] <Hixie>
at least on xforms person _still_ doesn't understand to this day what the backwards-compatibility issue is

[02:02:00.0000] <jgraham>
foolip: hg export might be what you are looking for

[02:02:01.0000] <foolip>
jgraham, thanks, looks like it is!

[02:12:00.0000] <foolip>
jgraham, you're not CC:d on the bug, are the mails sent to some list?

[02:14:00.0000] <jgraham>
foolip: No, I just followed your link

[02:14:01.0000] <foolip>
eh, stupid me :)

[02:15:00.0000] <foolip>
1. announce 2. forget 3. surprise at attention

[05:10:00.0000] <Ms2ger>
So, is anybody going to register rel=disclosure or are we going to keep all W3C specs invalid?

[06:34:00.0000] <jgraham>
Ms2ger: Is there a spec for it?

[06:34:01.0000] <Ms2ger>
Not one I managed to find

[06:34:02.0000] <Ms2ger>
Isn't "mentioned in pubrules" enough?

[08:00:00.0000] <AryehGregor>
smaug____, pong.

[08:00:01.0000] <AryehGregor>
[110722 10:32:42] <AryehGregor> smaug____ (who isn't here): the draft doesn't have to special-case replaceData/appendData because when it was written, those just called insertData/deleteData.  Now annevk has rewritten that in DOM Core at my request, and I'll rewrite the Range stuff soon to work with the new wording.  As for splitText, see the comments in the spec's HTML source.  IIRC, I copied the behavior from WebKit, who made it up; it make

[08:00:02.0000] <AryehGregor>
s a lot of sense, and IIRC roc t

[08:00:03.0000] <AryehGregor>
[110722 10:32:42] <AryehGregor> hought it was a good idea.  Without that special-case I'd have to add a bunch of different execCommand() special-cases, so it's no simpler, and authors don't get the improved behavior.

[08:01:00.0000] <AryehGregor>
Double-wrapping :(

[08:04:00.0000] <smaug____>
AryehGregor: yeah, I haven't yet quite decided which behavior for splitText makes most sense

[08:05:00.0000] <AryehGregor>
It's a tradeoff between simplicity and doing what makes sense in the specific case.

[08:05:01.0000] <smaug____>
for consistency the new spec is bad, but I could see the new spec more useful for web devs

[08:05:02.0000] <AryehGregor>
Without a special case for splitText, range mutation will do the wrong thing every time for it.

[08:05:03.0000] <smaug____>
wrong?

[08:05:04.0000] <AryehGregor>
As in unexpected, not what the author intended.

[08:06:00.0000] <smaug____>
it would do what is logical, IMHO

[08:06:01.0000] <AryehGregor>
And as I said, the added special case here removes multiple special cases elsewhere, so IMO it's an overall win for simplicity.

[08:06:02.0000] <AryehGregor>
Only if you view splitText as performing a sequence of DOM operations that involves deleting some text from one node and then creating an entirely unrelated node.

[08:06:03.0000] <AryehGregor>
If you view it as actually splitting an existing node, which is what it does intuitively and what the name implies, the non-special-case behavior makes no sense at all.

[08:07:00.0000] <AryehGregor>
E.g., it would be impossible to predict which node the range remained in, unless you knew which node the method happens to add (first vs. second).

[08:07:01.0000] <AryehGregor>
Which I'm not absolutely confident I remember, in fact, although I'm pretty sure it keeps the first node and adds the second.

[08:08:00.0000] <AryehGregor>
But if I have to think about it, your average web author stands no chance at all.

[08:08:01.0000] <AryehGregor>
(granted that which node is added is already visible in that that's the one it returns, but it's still not the way anyone's going to actually think of it working)

[08:08:02.0000] <smaug____>
it is just very surprising that DOM mutations can change range.start/endNode

[08:09:00.0000] <AryehGregor>
Only if you're thinking about things at a very low level.

[08:09:01.0000] <smaug____>
er, start/endContainer

[08:09:02.0000] <smaug____>
not really low level

[08:09:03.0000] <AryehGregor>
If you're thinking about things from a high level, what's surprising is when calling a method like splitText that has no visual effect at all in normal circumstances makes half the user's selection disappear.

[08:10:00.0000] <smaug____>
you attach the range to some text node, and suddenly it is moved to some other node

[08:10:01.0000] <smaug____>
that is surprising

[08:10:02.0000] <AryehGregor>
No, you select some visible text and suddenly half your selection disappears even though the text is plainly still there.  *That's* surprising.

[08:11:00.0000] <AryehGregor>
It just so happens that the selection of text is implemented in terms of a DOM with Nodes and Ranges and stuff, but that's not what anyone's going to base their expectations on unless they're a real JS guru.

[08:11:01.0000] <smaug____>
in that case the js code shouldn't use splitText, but modify the text data directly

[08:11:02.0000] <smaug____>
but as I said, I'm not yet quite sure which approach I prefer...

[08:12:00.0000] <smaug____>
usually I'm all for consistency, but in this case inconsistent behavior might be ok

[08:13:00.0000] <AryehGregor>
How can it modify the text data directly and achieve the desired effect?

[08:13:01.0000] <AryehGregor>
Manually resetting the selection?

[08:14:00.0000] <AryehGregor>
That's what execCommand() largely does, but it's extra work and it breaks other Ranges people might be using for other purposes (e.g., if some script saved the selection and wants to restore it later).

[08:14:01.0000] <smaug____>
ah, right, the case when insertNode(someelement) is called on text node

[08:15:00.0000] <smaug____>
I was thinking about insetNode(sometextnode)

[08:17:00.0000] <AryehGregor>
What about it?

[08:19:00.0000] <smaug____>
for insertNode(sometext) js should usually just modify the text data

[08:19:01.0000] <smaug____>
s/for/instead of/

[08:20:00.0000] <AryehGregor>
We're talking about splitText, no?  When did the conversation shift to insertNode?

[08:20:01.0000] <AryehGregor>
I assume that if you're insertNode()ing a Text node into a Text node, it's only because you're writing a generic algorithm that's meant to handle multiple different types of nodes.

[08:20:02.0000] <AryehGregor>
It would be pretty silly to not just use insertData(), otherwise.

[08:22:00.0000] <smaug____>
insertNode does splitText

[08:22:01.0000] <AryehGregor>
As do many other things, yeah.

[08:26:00.0000] <smaug____>
I really wish specs had some annotation to indicate why something is defined as it is

[08:27:00.0000] <smaug____>
especially specs which aren't written based on bugs in some tracker/bugzilla

[08:28:00.0000] <espadrine>
smaug____: something like perl6's Apocalypses and Exegesis? Yes, that can help

[08:29:00.0000] <AryehGregor>
smaug____, I put it in comments in the HTML source.

[08:29:01.0000] <AryehGregor>
There are some really long detailed comments in the specs I edit.

[08:29:02.0000] <AryehGregor>
Sometimes a hundred lines long or more.

[08:29:03.0000] <AryehGregor>
(in at least one extreme case)

[08:30:00.0000] <smaug____>
ah, need to just make those visible to the reader

[08:36:00.0000] <AryehGregor>
I could, but they can get very long.

[08:36:01.0000] <AryehGregor>
In plenty of cases I have multiple paragraphs of discussion for a single line of spec text.

[08:36:02.0000] <AryehGregor>
Maybe comments aren't the best way to do that, but having it inline doesn't seem great either.

[08:36:03.0000] <AryehGregor>
Hmm, I should have it inline but collapsed using JS.

[08:36:04.0000] <AryehGregor>
(or <details>, if anyone supports that properly yet)

[08:36:05.0000] <AryehGregor>
That would be ideal.

[08:49:00.0000] <TabAtkins_>
AryehGregor: <details> works in Chrome now.

[08:50:00.0000] <AryehGregor>
TabAtkins_, do you know of a good JS shim?

[08:50:01.0000] <AryehGregor>
Also, it seems like the Chrome implementation forces the <details> to its own line, which I don't want.

[08:50:02.0000] <AryehGregor>
When I do style=display:inline, it doesn't do anything.

[08:50:03.0000] <AryehGregor>
Oh, wait.

[08:51:00.0000] <AryehGregor>
I was trying to put it inside a <p>.

[08:51:01.0000] <AryehGregor>
/me tries again

[08:51:02.0000] <TabAtkins_>
AryehGregor: Also maybe put display:inline on the summary too.

[08:51:03.0000] <AryehGregor>
Seems to work if I put it in an <li> or something.

[08:51:04.0000] <AryehGregor>
That's bad, though, I want to put it in <p>.

[08:51:05.0000] <TabAtkins_>
Blame Hixie.

[08:52:00.0000] <TabAtkins_>
He didn't want <p>'s autoclose behavior to be inconsistent.

[08:52:01.0000] <AryehGregor>
Hmm.

[08:52:02.0000] <AryehGregor>
/me files a bug

[08:54:00.0000] <TabAtkins_>
For JS shim, I guess check out Paul Irish's master list of shims.

[08:54:01.0000] <AryehGregor>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=13345

[08:54:02.0000] <AryehGregor>
/me e-mails Hixie a link to the bug

[08:55:00.0000] <AryehGregor>
(to request prioritization, since it affects existing new implementations and we don't have much time to change the behavior if we're going to do so)

[08:55:01.0000] <TabAtkins_>
Hmm, making it inline by default is interesting.  Makes sense for the spec use-case.

[08:55:02.0000] <TabAtkins_>
/me wants to use it for all his notes and issues now.

[09:02:00.0000] <AryehGregor>
Most notes/issues shouldn't be collapsed, IMO.

[09:02:01.0000] <AryehGregor>
Rationale should be.

[09:42:00.0000] <AryehGregor>
Also maybe some todo's.

[09:46:00.0000] <hsivonen>
foolip: there's no support in the build script for not pulling the w3.org DTD. (hooray for w3.org DTDs!)

[10:26:00.0000] <AryehGregor>
Apparently IE10PP2 isn't using a fully compliant HTML parser . . . <font color=#ABCDEF> becomes <font color=#abcdef> in the DOM.

[10:28:00.0000] <annevk>
oops

[10:51:00.0000] <AryehGregor>
hsivonen, do the parser tests test that <font color=#ABCDEF> doesn't get parsed to <font color=#abcdef>?  If not, could we add such a test?

[10:56:00.0000] <ruediger>
hi

[10:56:01.0000] <AryehGregor>
Hi!

[11:01:00.0000] <AryehGregor>
Actually, IE is much more egregious.

[11:01:01.0000] <AryehGregor>
It turns <span style=color:#aBcDeF> into <span style="color: rgb(171, 205, 239);"> too.

[11:01:02.0000] <AryehGregor>
It's not the only one that does weird stuff.

[11:02:00.0000] <AryehGregor>
Gecko adds tabindex="0" to <video>.

[11:02:01.0000] <AryehGregor>
Always.

[11:02:02.0000] <AryehGregor>
Like, wtf?

[11:02:03.0000] <ruediger>
does the html5 wg accept new named character references? (section 8.5 in the html5 spec) I'd like to propose &Szlig; for capital ß http://en.wikipedia.org/wiki/Capital_%C3%9F

[11:03:00.0000] <AryehGregor>
hsivonen, do the parser tests include basic sanity checks to make sure browsers aren't randomly mangling attributes, or is it mostly focused on ensuring that everything gets the right parents and such?

[11:03:01.0000] <hsivonen>
AryehGregor: IIRC, they don't test that and you are welcome to contribute tests under the MIT license that html5lib uses

[11:03:02.0000] <annevk>
ruediger, the Math WG is in charge of entities

[11:03:03.0000] <annevk>
ruediger, and they have not been expanded since 98 practically

[11:04:00.0000] <annevk>
ruediger, just use the character like you did here

[11:04:01.0000] <AryehGregor>
ruediger, we probably are not going to add new named entities for things that are sufficiently obscure that they were only added in Unicode 5.1.

[11:04:02.0000] <AryehGregor>
Actually, we're probably not going to add many new named entities at all.

[11:04:03.0000] <AryehGregor>
Just encode it as UTF-8.

[11:04:04.0000] <ruediger>
annevk: yeah, normally I don't bother with named characters and use utf8. But for the sake of consistency

[11:05:00.0000] <hsivonen>
https://twitter.com/#!/daaitoulaam/status/94988135091535873

[11:05:01.0000] <ruediger>
lower case ß is supported as &szlig;

[11:05:02.0000] <AryehGregor>
hsivonen, where's the right place to look at and submit new parser tests?

[11:06:00.0000] <annevk>
ruediger, you can raise it as a bug; the relevant math people will prolly cc themselves

[11:06:01.0000] <annevk>
ruediger, he's in this channel as david_carlisle

[11:06:02.0000] <ruediger>
thanks

[11:07:00.0000] <ruediger>
I'll do that

[11:07:01.0000] <hsivonen>
AryehGregor: http://code.google.com/p/html5lib/source/browse/#hg%2Ftestdata

[11:08:00.0000] <annevk>
hsivonen, you retweet that he calls you an utter ass?

[11:09:00.0000] <annevk>
@gruber-style :p

[11:09:01.0000] <AryehGregor>
hsivonen, and those are being made part of the official HTML5 test suite at this point, or what?

[11:09:02.0000] <hsivonen>
annevk: yeah, that was a @gruber-style retweet--not a retweet implying agreement

[11:10:00.0000] <hsivonen>
AryehGregor: dunno about Official, but that's the set of test parser devs pay attention to

[11:10:01.0000] <AryehGregor>
hsivonen, are they runnable cross-browser by just visiting a web page?

[11:10:02.0000] <hsivonen>
AryehGregor: not to my knowledge, but my knowledge may be stale

[11:11:00.0000] <AryehGregor>
Hmm, okay.

[11:12:00.0000] <hsivonen>
AryehGregor: jgraham probably knows more about the cross-browser harness situation

[11:12:01.0000] <AryehGregor>
jgraham, do you know what the story is with cross-browser parser tests, and their status as Official W3C Tests(tm)?

[11:12:02.0000] <annevk>
adding <video>, something with a style attribute and such does make sense

[11:13:00.0000] <AryehGregor>
I'd add at least one example of every element/attribute combination, ideally.

[11:13:01.0000] <AryehGregor>
Of course, I like my tests to be thorough.

[11:13:02.0000] <AryehGregor>
Also, make sure to test corner cases for all of them, including leading and trailing whitespace, special characters, etc.

[11:14:00.0000] <AryehGregor>
There should be no special per-attribute value mangling at the parser level at all, AFAIK.

[11:14:01.0000] <AryehGregor>
Definitely not stuff like reserializing style attributes.

[11:15:00.0000] <annevk>
correct

[11:17:00.0000] <AryehGregor>
Or adding tabindex="0".

[11:17:01.0000] <AryehGregor>
Or stripping leading/trailing whitespace from URL attributes (I caught Gecko doing that even with setAttribute(), but I filed a bug and they fixed it).

[11:18:00.0000] <hsivonen>
AryehGregor: adding tabindex="0" is not a parser bug, FWIW, but it's a bug exposed by parser tests

[11:19:00.0000] <hsivonen>
(of course, if you are testing the whole box, you shouldn't care which part of it gets the blame)

[11:20:00.0000] <AryehGregor>
Hmm, using appendChild creates it too.

[11:20:01.0000] <AryehGregor>
Fun.

[11:27:00.0000] <hsivonen>
I wonder how the "This user" badges for Wikipedia user pages started

[11:28:00.0000] <AryehGregor>
You mean the userboxes?

[11:28:01.0000] <AryehGregor>
Somebody made one up for fun because they were bored, probably in like 2003 or something.

[11:29:00.0000] <hsivonen>
AryehGregor: I meant {{user foo}} boxes

[11:29:01.0000] <AryehGregor>
The Babel boxes, for languages?  Like {{user en}}?

[11:29:02.0000] <AryehGregor>
Oh, other userboxes also often start with "User ".

[11:29:03.0000] <AryehGregor>
Anyway, yeah, those are userboxes.

[11:30:00.0000] <AryehGregor>
Actually, I think Babel boxes may have been first, just for practical reasons.

[11:30:01.0000] <AryehGregor>
Then people had fun with it and made all sorts of stupid random useless ones.

[11:30:02.0000] <hsivonen>
AryehGregor: I mean "This user would like to own one or more dogs."

[11:30:03.0000] <AryehGregor>
Yeah, userboxes.

[11:30:04.0000] <AryehGregor>
At some point there was a purge and the more frivolous ones were moved out of Template: namespace into individual users' namespaces.

[11:31:00.0000] <AryehGregor>
http://en.wikipedia.org/w/index.php?title=Template:Userbox&oldid=18039616

[11:31:01.0000] <AryehGregor>
There were probably earlier uses using other templates, though, or just ad hoc stuff.

[11:32:00.0000] <AryehGregor>
http://en.wikipedia.org/w/index.php?title=Template:User_en&oldid=11184536

[11:33:00.0000] <hsivonen>
AryehGregor: my humor detector is dysfunctional with some of these

[11:34:00.0000] <hsivonen>
AryehGregor: that is, I can't tell which ones are serious and which ones an inside wiki meta humor

[11:35:00.0000] <hsivonen>
s/an/are/

[11:43:00.0000] <david_carlisle>
ruediger:  I  was away from my machine, but as they said we don't normally add new names just on consistency grounds, For example only the letters have entity names for open facem even though Unicode added digits as well in the end, see http://www.w3.org/2003/entities/2007doc/double-struck.html

[11:45:00.0000] <ruediger>
david_carlisle: :-/ but thanks for your reply

[11:49:00.0000] <david_carlisle>
ruediger:  seem the bug http://www.w3.org/Bugs/Public/show_bug.cgi?id=13108#c2 which is a requested name with slightly different issues, but possibly the same outcome (although that bug isn't closed yet)

[12:18:00.0000] <jgraham>
AryehGregor: http://w3c-test.org/html/tests/submission/Opera/html5lib/

[12:18:01.0000] <jgraham>
That needs to be updated a little, but it is only a few commits behind

[12:19:00.0000] <AryehGregor>
jgraham, in Chrome 14 dev, with "Test with document.write" and "Run All Tests", it stops after a little while with "test.js:239Uncaught ReferenceError: template is not defined"

[12:21:00.0000] <jgraham>
Oh, that may be one of the things that needs to be updated :)

[12:22:00.0000] <AryehGregor>
I get security errors when trying to use data: URLs, too.

[12:22:01.0000] <jgraham>
I thought I pushed that change to the W3C copy already, but maybe I only did the Opera-internal copy

[12:22:02.0000] <jgraham>
Yeah, Chrome won't let you read the DOM of a data URI loaded in an iframe

[12:22:03.0000] <jgraham>
AFAICT this is dumb and chrome should allow it

[12:27:00.0000] <AryehGregor>
/me concurs

[12:29:00.0000] <AryehGregor>
abarth, why does Chrome not allow reading the DOM of a data URL loaded in an iframe?

[12:29:01.0000] <annevk>
there's an open bug on that I think

[12:30:00.0000] <TabAtkins_>
AryehGregor: Because we don't handle origins correctly, I believe.

[12:30:01.0000] <TabAtkins_>
(The spec makes it clear that it should be same-origin.)

[12:30:02.0000] <AryehGregor>
So it's not intentional, in other words?

[12:30:03.0000] <AryehGregor>
Ah, okay.

[12:30:04.0000] <AryehGregor>
So just a bug.

[12:33:00.0000] <annevk>
http://lists.w3.org/Archives/Public/public-web-perf/2011Jul/thread.html#msg79 reference mania strikes again

[12:33:01.0000] <abarth>
AryehGregor: because you haven't sent WebKit a patch that fixes that bug yet :)

[12:34:00.0000] <abarth>
AryehGregor: its a tricky bug to fix, which is why no one has sent us a patch yet

[12:48:00.0000] <AryehGregor>
Oh, Opera doesn't implement Object.keys?  Lame.

[12:49:00.0000] <AryehGregor>
As of 11.50.

[12:49:01.0000] <AryehGregor>
Isn't that part of ES5, and haven't all the cool kids implemented ES5 by now?

[12:50:00.0000] <karlcow>
AryehGregor: ES5 is not implemented yet in a released version of Opera

[12:50:01.0000] <AryehGregor>
/me works around it

[12:54:00.0000] <annevk>
I wonder when dev.w3.org will move to hg

[12:55:00.0000] <jgraham>
annevk: Remind me, when is the heat death of the universe again?

[12:57:00.0000] <Dashiva>
Next Thursday

[13:00:00.0000] <annevk>
jgraham, tralala

[13:00:01.0000] <jcranmer>
Dashiva: already?

[13:01:00.0000] <jcranmer>
damn, better hurry up if I want to finish some things before the universe dies

[13:15:00.0000] <karlcow>
hsivonen: re: identity… it is also a bit like leaving your passport for the length of your visit to the local police.

[13:16:00.0000] <TabAtkins_>
Hixie: I can't find the diff where you added the :ltr and :rtl pseudos.  Halp?

[13:26:00.0000] <Hixie>
TabAtkins_: please hold

[13:26:01.0000] <Hixie>
r5672?

[13:26:02.0000] <Hixie>
$ grep :ltr blame.r6000  5672       ianh    <dt><dfn title="selector-ltr"><code>:ltr</code></dfn></dt> 5672       ianh     <p>The <code title="selector-ltr">:ltr</code> pseudo-class must 5672       ianh :ltr { direction: ltr; }

[13:26:03.0000] <Hixie>
well

[13:26:04.0000] <Hixie>
er

[13:26:05.0000] <Hixie>
the spacing in that paste went all to hell

[13:26:06.0000] <Hixie>
but anyway

[13:26:07.0000] <Hixie>
if you need blame files i have files called blame.rXXXX in the spec directory

[13:26:08.0000] <Hixie>
for various values of XXXX:

[13:26:09.0000] <Hixie>
blame.r0005  blame.r0500  blame.r1000  blame.r2200  blame.r2500  blame.r2905  blame.r3000  blame.r4000  blame.r5000  blame.r5600

[13:26:10.0000] <Hixie>
blame.r0050  blame.r0954  blame.r2000  blame.r2216  blame.r2800  blame.r2947  blame.r3075  blame.r4667  blame.r5287  blame.r6000

[13:26:11.0000] <Hixie>
hth

[13:33:00.0000] <AryehGregor>
Sounds like a workaround for svn blame being unreasonably slow.

[13:33:01.0000] <AryehGregor>
git blame takes like thirty seconds max on big repos for me . . . although granted I haven't tried it with a repo that mostly consists of one huge file that's changed in every commit.

[13:34:00.0000] <Philip`>
Does git blame provide any way to discover what revision a line was removed in?

[13:34:01.0000] <AryehGregor>
I'm pretty sure it does, but I haven't ever tried.

[13:37:00.0000] <Philip`>
Ah, looks like it has a --reverse option

[13:37:01.0000] <AryehGregor>
git commands have seldom been faulted for having an insufficient number of options.

[13:38:00.0000] <Philip`>
/me has occasionally wanted a feature like that, since in SVN/etc it's much more annoying to have to do a manual binary search than having to wait a few minutes while a normal blame runs in the background

[13:38:01.0000] <AryehGregor>
"JSON.parse: unexpected keyword"

[13:38:02.0000] <AryehGregor>
What could that possibly mean?

[13:39:00.0000] <AryehGregor>
Firebug is reporting it for the line "test = JSON.parse(test);".

[13:39:01.0000] <AryehGregor>
Oh, it must mean there's an error in the input.

[13:39:02.0000] <AryehGregor>
I.e., the test variable isn't valid JSON.

[13:39:03.0000] <AryehGregor>
That confused me.

[13:49:00.0000] <Hixie>
AryehGregor: that's exactly what it is :-(

[13:50:00.0000] <Hixie>
AryehGregor: last time i heard of someone trying to put the spec in git, the migrate crashed or something.

[13:50:01.0000] <AryehGregor>
Hmm.

[13:50:02.0000] <AryehGregor>
It's true that git doesn't like very large files, but the spec should be small enough, I'd think.

[13:50:03.0000] <Hixie>
5MB * 6000 revisions? :-)

[13:51:00.0000] <AryehGregor>
What's the SVN checkout URL, just http://svn.whatwg.org/webapps/?

[13:51:01.0000] <Hixie>
i believe so

[13:51:02.0000] <jgraham>
I have it in git

[13:51:03.0000] <Hixie>
ah, cool

[13:51:04.0000] <jgraham>
The problem was mirroring it to github

[13:51:05.0000] <Hixie>
aah

[13:52:00.0000] <jgraham>
The import was slow and timed out or something

[13:52:01.0000] <AryehGregor>
How fast is a blame?

[13:52:02.0000] <AryehGregor>
/me tries git svn clone http://svn.whatwg.org/webapps/

[13:54:00.0000] <AryehGregor>
git was designed for use by the Linux kernel, the number of revisions you have is trivial.

[13:55:00.0000] <AryehGregor>
As is overall repo size.

[13:55:01.0000] <AryehGregor>
I doubt 5M in one file is a big problem either.

[13:55:02.0000] <AryehGregor>
I had problems when I tried using it for incremental backup of a database; apparently it doesn't like multi-gigabyte files (tries to read them entirely into memory).

[13:55:03.0000] <AryehGregor>
But 5M should be no problem.

[13:55:04.0000] <AryehGregor>
It's worth switching from svn to git just for the speed . . .

[13:56:00.0000] <jgraham>
Either I got the blame syntax wrong or it is not all that fast in this case

[13:57:00.0000] <AryehGregor>
It's usually not super-fast, it does have to read the whole history off disk.

[13:57:01.0000] <AryehGregor>
Well, not the whole history.

[13:57:02.0000] <AryehGregor>
But it's ridiculously faster than svn blame.

[13:58:00.0000] <jgraham>
Well yes I wouldn't even try this with svn

[13:59:00.0000] <jgraham>
If someone mirrored the spec to something with a reasonable diff viewer (github, google code, etc.) the web apps tracker could just point there

[13:59:01.0000] <AryehGregor>
++

[13:59:02.0000] <Hixie>
AryehGregor: blame is near instantaneous for me, since all i use is grep

[14:00:00.0000] <jgraham>
Rather than doing it badly (no way to get more context)

[14:00:01.0000] <Hixie>
AryehGregor: how is git going to make that better?

[14:00:02.0000] <AryehGregor>
Hixie, it's not, given that you've already spent the effort to work around how broken svn blame is.

[14:00:03.0000] <Hixie>
unless git is also instantaneous, i'd have to work around git too

[14:01:00.0000] <AryehGregor>
It doesn't have to be instantaneous, it just has to be fast enough that it's not worth working around.

[14:01:01.0000] <AryehGregor>
I've never felt git blame was slow enough to bother working around, on any project I've worked on.

[14:01:02.0000] <AryehGregor>
/me has checked out up to r540 or so

[14:05:00.0000] <Hixie>
i've never felt subversion blame was too slow either until i tried using it on a 5mb document with 6000+ revisions :-)

[14:05:01.0000] <jgraham>
AryehGregor: < 8 minutes

[14:06:00.0000] <AryehGregor>
Any reasonably large software project is *way* bigger than that.  MediaWiki's full repo is a gigabyte or two, with the 93010th revision committed about ten minutes ago.

[14:06:01.0000] <AryehGregor>
No individual text files nearly that big, of course.

[14:06:02.0000] <jgraham>
AryehGregor: In most reasonable projects one file doesn't change slightly in every revision

[14:06:03.0000] <AryehGregor>
Yes, that's certainly true.

[14:07:00.0000] <AryehGregor>
This is a bit of a worst-case for blame.

[14:07:01.0000] <jgraham>
Anyway, the spec being in git would be way more convenient

[14:07:02.0000] <jgraham>
Not just for blame, but particularly for diff

[14:08:00.0000] <AryehGregor>
Yeah.

[14:19:00.0000] <jgraham>
AryehGregor: Updated the html5lib tests on w3c-test.org

[14:20:00.0000] <AryehGregor>
jgraham, where should I submit new ones, if I were so inclined?

[14:20:01.0000] <jgraham>
Oh but I forgot to include the tokenizer tests

[14:20:02.0000] <jgraham>
AryehGregor: To the html5lib repository

[14:21:00.0000] <AryehGregor>
k

[14:21:01.0000] <jgraham>
Although you are quite welcome to mention it on public-html-testsuite if you feel so inclined

[14:21:02.0000] <jgraham>
The idea is that the html5lib repository has the canonical version of the test .dat files

[14:22:00.0000] <jgraham>
I'm not sure Kris has entirely realised / agreed with this yet mind you

[15:35:00.0000] <Yuhong>
FYI on base elements and security: http://i8jesus.com/?p=10


2011-07-25
[17:40:00.0000] <AryehGregor>
Hixie, redirects from fragments like <http://www.whatwg.org/C#the-video-element> seem to be broken.

[17:41:00.0000] <Hixie>
Philip`: ^ the js file you're sending me has some sort of syntax error

[17:42:00.0000] <Hixie>
oh it's incomplete

[17:42:01.0000] <AryehGregor>
I love open-source software: http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/0453.html

[17:42:02.0000] <Hixie>
i guess i'll try regenning it

[17:43:00.0000] <annevk>
AryehGregor, I thought that was pretty funny

[17:44:00.0000] <Hixie>
i don't understand that e-mail

[17:44:01.0000] <Hixie>
did you link to the right line?

[17:44:02.0000] <TabAtkins_>
Hixie: Yes, it was pointing out the recursion check.

[17:44:03.0000] <TabAtkins_>
(Which was written by Jonas, who was claiming they don't have one.)

[17:44:04.0000] <Hixie>
the NS_ENSURE_TRUE line?

[17:44:05.0000] <Hixie>
that doesn't get compiled in release builds

[17:45:00.0000] <annevk>
it certainly throws a INVALID_STATE_ERR

[17:45:01.0000] <annevk>
an*

[17:45:02.0000] <Hixie>
oh i guess just the warning doesn't get compiled

[17:46:00.0000] <Hixie>
the actual check is compiled

[17:46:01.0000] <Hixie>
nevermind me

[17:48:00.0000] <Hixie>
my god i hate sites that log me out automatically

[17:50:00.0000] <zewt>
like google? heh

[17:50:01.0000] <Hixie>
when does google do that?

[17:50:02.0000] <zewt>
gmail/greader randomly logging me out drives me nuts

[17:50:03.0000] <Hixie>
sounds like a bug

[17:50:04.0000] <zewt>
seems periodic, like once a week or something

[17:50:05.0000] <Hixie>
you may have checked a checkbox asking for it

[17:50:06.0000] <TabAtkins_>
zewt: Your cookie expires eventually, but it should be longer than a week.

[17:50:07.0000] <zewt>
greader does it the most

[17:51:00.0000] <TabAtkins_>
(Like, with 2factor auth, you can make it remember for 30 days)

[17:51:01.0000] <zewt>
heh, for some reason now, after the google apps transition, youtube feels the need to tell me every single time i log in "this account is managed by zewt.org"

[17:52:00.0000] <zewt>
(yeah, i know, i set it up, thanks again youtube)

[17:52:01.0000] <annevk>
WebKit is going to share the HTML and XML tokenizer: https://bugs.webkit.org/show_bug.cgi?id=65000

[17:52:02.0000] <annevk>
That is kind of interesting...

[17:57:00.0000] <Hixie>
o_O

[17:57:01.0000] <Hixie>
that sounds like a world of pain...

[17:58:00.0000] <Hixie>
wait, it's not going to share the tokeniser

[17:58:01.0000] <Hixie>
it's just using the same interfaces

[18:03:00.0000] <annevk>
oh

[09:19:00.0000] <jgraham>
So… afaict Web DOM Core is wrong about hwo createElement works

[09:20:00.0000] <jgraham>
In particular if I createElement an element from a script in frame A in the context of a different frame B, the element seems to get the namespace of A or null

[09:20:01.0000] <jgraham>
Even if the ownerDocument is the document in B

[09:21:00.0000] <annevk>
createElement always uses the HTML namespace

[09:22:00.0000] <jgraham>
Either that isn't true or I am going crazy

[09:23:00.0000] <jgraham>
In particular I have a script running in a SVG document that tries to createElement some elements in the parent HTML document

[09:23:01.0000] <jgraham>
and they don't seem to end up in the right namespace

[09:25:00.0000] <jgraham>
and doing s/createElement/createElementNS(xhtml_ns, / fixed the problem

[09:25:01.0000] <jgraham>
Although I had to look up the xhtml ns

[09:26:00.0000] <annevk>
well it's not implemented everywhere yet

[09:27:00.0000] <jgraham>
s/everywhere/anywhere/?

[09:43:00.0000] <jgraham>
hsivonen: BTW did you see http://krijnhoetmer.nl/irc-logs/whatwg/20110723#l-397 ?

[09:44:00.0000] <Philip`>
/me learns that PHP 5.3 added "goto" statements

[09:44:01.0000] <Philip`>
(which incidentally happens to break old code that has a function named "goto")

[10:01:00.0000] <annevk>
jgraham, could be

[10:02:00.0000] <annevk>
this is what we decided upon years ago

[10:03:00.0000] <TabAtkins_>
Yo, I'd like to apologize for last night.  I was not fair to you or your arguments, which had a lot more merit than I allowed at the time.  I'm sorry that we had to part somewhat angry at each other.

[10:03:01.0000] <TabAtkins_>
...

[10:03:02.0000] <TabAtkins_>
goddammit

[10:04:00.0000] <annevk>
someone had a fight last night

[10:04:01.0000] <annevk>
teehee

[10:05:00.0000] <TabAtkins_>
I'd like to apologize to the room for two minutes ago.  I was not fair to your eyes, which were a lot cuter than I allowed at the time.  I'm sorry that we couldn't have met somewhere more private, whatwg room.

[10:05:01.0000] <TabAtkins_>
/me hugs whatwg.

[10:06:00.0000] <jgraham>
/me assumes "Yo" is short for "Yo! Sushi"

[10:08:00.0000] <jgraham>
/me realises that might have been a somewhat UK-centric reference

[10:08:01.0000] <TabAtkins_>
Yes, yes it was.

[10:10:00.0000] <Philip`>
When I first saw a YO! Sushi, it looked startlingly Japanese

[10:10:01.0000] <Philip`>
particularly with all the walls seemingly covered in giant anime characters

[10:11:00.0000] <Philip`>
/me guesses that was the intended effect

[10:23:00.0000] <jgraham>
Were the eyes more or less cute than those of the WHATWG room?

[10:23:01.0000] <jgraham>
+of the anime characters

[10:25:00.0000] <jgraham>
Argh

[10:25:01.0000] <jgraham>
Silly cache

[10:28:00.0000] <TabAtkins_>
Nobody has eyes cuter than you guys.  Especially you, jgraham.

[10:29:00.0000] <jgraham>
And nobody needs glasses more than you, TabAtkins_ :)

[10:29:01.0000] <zewt>
(´▽｀)

[10:41:00.0000] <hober>
/me feels the love in #whatwg this morning

[11:13:00.0000] <sicking>
annevk: ping

[11:14:00.0000] <annevk>
hey sicking

[11:14:01.0000] <annevk>
i'm in your timezone

[11:14:02.0000] <annevk>
o_O

[11:16:00.0000] <sicking>
annevk: and you're up this early !?!

[11:17:00.0000] <sicking>
annevk: is there a reason we're forbidding doctypes from getting adopted. It's a very odd exception

[11:18:00.0000] <annevk>
I think it was because doctypes don't have an associated document

[11:18:01.0000] <annevk>
but I am happy to change that

[11:18:02.0000] <annevk>
I wish Acid3 had not tested DOCTYPEs so we could have dropped them on the floor

[11:18:03.0000] <sicking>
annevk: worst, reason, ever

[11:19:00.0000] <annevk>
Acid3 even has

[11:19:01.0000] <annevk>
assertEquals(doctype.ownerDocument, null, "doctype's ownerDocument was wrong after creation");

[11:19:02.0000] <annevk>
omg

[11:19:03.0000] <annevk>
I hate Acid3

[11:19:04.0000] <sicking>
get hixie to fix it

[11:19:05.0000] <sicking>
if it's holding back the web platform, it is counter to its purpose

[11:20:00.0000] <annevk>
so what do we want? keep doctypes but let them have an ownerdocument?

[11:21:00.0000] <annevk>
and let them be adopted

[11:21:01.0000] <sicking>
i was just wanting to let them be adopted for now

[11:21:02.0000] <sicking>
they do have an owner document generally

[11:22:00.0000] <sicking>
it's just that when they're created they don't have one

[11:22:01.0000] <sicking>
i'm not sure how we'd deprecated doctype nodes

[11:22:02.0000] <sicking>
aren't browsers inserting them in the DOM now? during normal parsing

[11:22:03.0000] <annevk>
yeah they are

[11:22:04.0000] <annevk>
so I guess that's not really feasible anymore

[11:23:00.0000] <sicking>
i think it'd be hard yeah

[11:23:01.0000] <sicking>
my concern is that removing them would break sites that do document.childNodes[1] to get the document element

[11:24:00.0000] <annevk>
so adoptNode for doctype would nullify its ownerDocument right?

[11:24:01.0000] <sicking>
no, it would do the same as for other nodes

[11:24:02.0000] <annevk>
okay

[11:24:03.0000] <sicking>
set ownerDocument to the new document

[11:25:00.0000] <annevk>
so then I just need to remove doctype from the adoptNode line

[11:25:01.0000] <annevk>
let me do that right now

[11:25:02.0000] <sicking>
doctypes have a owner document as soon as they are inserted in a doc

[11:27:00.0000] <annevk>
oh and http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-ensure-samedoc is affected as well

[11:27:01.0000] <annevk>
the implicit adoption

[11:29:00.0000] <sicking>
right

[11:29:01.0000] <sicking>
we've supported implicit adoption forever for doctypes, just not explicit adoption

[11:30:00.0000] <sicking>
annevk: oh, and i'd be thrilled to say that doctypes always have an owner document, just like all other nodes

[11:30:01.0000] <AryehGregor>
Yes, please.  Also allow cloning them.

[11:30:02.0000] <AryehGregor>
Down with magic!

[11:30:03.0000] <sicking>
annevk: i'll just be a bit tricky to define what that that ownerdoc is

[11:30:04.0000] <annevk>
getting Acid3 changed is a real pain though, Hixie?

[11:31:00.0000] <AryehGregor>
Has it made SVG fonts optional yet?

[11:31:01.0000] <annevk>
sicking, why is it harder than for an element?

[11:32:00.0000] <sicking>
annevk: because you don't create them from a document, but rather from a implementation

[11:32:01.0000] <sicking>
AryehGregor: nope

[11:33:00.0000] <sicking>
Acid3 is the IE6 of conformance test suites :)

[11:33:01.0000] <annevk>
sicking, not funny :)

[11:34:00.0000] <annevk>
sicking, fixed for adoptNode()

[11:35:00.0000] <sicking>
yay

[11:35:01.0000] <annevk>
still throws for documents

[11:35:02.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-document-adoptnode

[11:38:00.0000] <sicking>
annevk: awesome, thanks!

[11:42:00.0000] <annevk>
AryehGregor, so yes

[11:42:01.0000] <annevk>
cloning of doctypes and documents should be allowed?

[11:50:00.0000] <annevk>
per email debate I just reread that already works

[11:51:00.0000] <annevk>
but importNode should block documents

[11:51:01.0000] <annevk>
oh because a document cannot own another one

[12:16:00.0000] <annevk>
notes:

[12:16:01.0000] <annevk>
* make deep argument optional

[12:16:02.0000] <annevk>
* make clone dfn concept-node-clone

[12:16:03.0000] <zewt>
for feature parity with the rest of the internet

[12:17:00.0000] <AryehGregor>
What are some examples of where people have reported bugs against a WD that are fixed in ED, or implementers have implemented something based on an obsolete WD?

[12:18:00.0000] <TabAtkins_>
annevk has several examples, I believe.

[12:18:01.0000] <annevk>
it happens often, but usually in internal bug reports

[12:21:00.0000] <Hixie>
wow, if ever there was an argument for amalgamting specs rather than splitting them, it has to be the graph on http://greenbytes.de/tech/tc2231/

[12:22:00.0000] <annevk>
so importNode and cloneNode are pretty much identical

[12:22:01.0000] <annevk>
but one throws for document in implementations and the other doesn't

[12:22:02.0000] <annevk>
weird

[12:29:00.0000] <TabAtkins_>
Hixie: I suspect it would be slightly less horrible if they had real names, not RFC numbers.

[12:29:01.0000] <Hixie>
the whole http layer should just be one spec

[12:29:02.0000] <Hixie>
having all this stuff split all over the place is ludicrous

[12:30:00.0000] <annevk>
HTTP itself is becoming 8 or so specs

[12:30:01.0000] <zewt>
rfcs having no xrefs at all doesn't help

[12:30:02.0000] <hsivonen>
jgraham: no, I did not see that. (I was away in the countryside as a wedding guest over the weekend and missed just about everything that went on online)

[12:30:03.0000] <annevk>
I don't really want to fork HTTP though

[12:30:04.0000] <annevk>
too much effort

[12:31:00.0000] <TabAtkins_>
zewt: Indeed.  RFCs in general are nearly the worst possible format to have split specs among.

[12:31:01.0000] <zewt>
seems like just publishing rfcs in a format other than fixed-width 80-column with hard page breaks as if people still print on 66-line printers is a hard enough battle

[12:31:02.0000] <AryehGregor>
They have HTML versions, which are significantly less horrible.

[12:31:03.0000] <zewt>
(and as if anyone actually prints specs)

[12:32:00.0000] <Philip`>
/me wonders if anyone has printed HTML5 recently

[12:32:01.0000] <zewt>
one copy, one rainforest

[12:36:00.0000] <hsivonen>
Philip`: based on tweets, it looked like Jirka printed it recently

[12:36:01.0000] <hsivonen>
I probably should have started a "why RDFa sucks" wiki page 3 years ago so that I wouldn't need to write the same stuff over and over again

[12:47:00.0000] <annevk>
never too late

[12:53:00.0000] <annevk>
applied notes to draft

[13:08:00.0000] <kangax>
Does anyone remember offhand if form's submit event should fire when form fails validation? I see Opera 11.50 fires it; FF, Chrome don't.

[13:09:00.0000] <annevk>
looks like a bug in Opera per the specification

[13:10:00.0000] <annevk>
are you submitting it manually?

[13:10:01.0000] <annevk>
because if you submit it through submit() Opera is correct

[13:10:02.0000] <sicking>
annevk: do you know if opera is doing the HTML5 undomanager?

[13:11:00.0000] <annevk>
i know we don't

[13:11:01.0000] <jgraham>
sicking: So far we looked at the spec and got confused I think

[13:11:02.0000] <sicking>
jgraham: cool, i have an alternative proposal that i'm pitching

[13:12:00.0000] <sicking>
but i don't want to start implementing if other browsers have already set down the HTML5 path

[13:12:01.0000] <smaug____>
annevk: why you made /deep/ optional?

[13:12:02.0000] <jgraham>
sicking: Sounds good

[13:12:03.0000] <Hixie>
sicking: there's been several alternative proposals, i hope we can find something to replace it

[13:12:04.0000] <sicking>
Hixie: do you know if anyone's implemented what's in the spec?

[13:13:00.0000] <sicking>
Hixie: if people already have, then it'll be a lot harder to do something else

[13:13:01.0000] <jgraham>
Hixie: Can you pull it from the spec if it is unimplemented and should not be implemented?

[13:13:02.0000] <Hixie>
sicking: to my knowledge no. I removed it from the w3c copy to try to prevent it from getting implemented.

[13:13:03.0000] <kangax>
annevk: no, submitting by pressing submit button

[13:13:04.0000] <Hixie>
jgraham: well it's not that it shouldn't be implemented, it should get implementation feedback. which it is getting. :-)

[13:14:00.0000] <sicking>
Hixie: is "we don't like it and don't think it should be implemented" enough feedback :)

[13:14:01.0000] <Hixie>
sicking: no :-)

[13:14:02.0000] <Hixie>
sicking: we need _something_

[13:14:03.0000] <sicking>
Hixie: second proposal in https://bugzilla.mozilla.org/show_bug.cgi?id=617532#c12

[13:14:04.0000] <Hixie>
sicking: (though it's not urgent, certainly)

[13:14:05.0000] <kangax>
annevk: oh well. i guess will work around for now with `checkValidity` in submit handler.

[13:15:00.0000] <annevk>
sicking, there's also http://rniwa.com/editing/undomanager.html and http://rniwa.com/editing/undomanager-usecases.html

[13:15:01.0000] <Hixie>
sicking: what anne said

[13:15:02.0000] <Hixie>
sicking: please coordinate with other browser vendors on what proposal to implement :-) (feel free to use the whatwg list for this purpose)

[13:16:00.0000] <annevk>
smaug____, it's optional in some implementations already

[13:16:01.0000] <annevk>
smaug____, and it makes sense in the same way as making the last argument of addEventListener optional makes sense, I think

[13:16:02.0000] <sicking>
smaug____: it seems like an argument that would be useful to make optional

[13:16:03.0000] <smaug____>
it is just not backwards compatible change

[13:17:00.0000] <smaug____>
it should be optional, but if sites start to rely on it being optional...

[13:18:00.0000] <annevk>
if sites rely on <input type=datetime> people need to update their browsers too

[13:18:01.0000] <annevk>
sort of how things work...

[13:19:00.0000] <smaug____>
calling importNode without the last parameter will throw in some browsers

[13:19:01.0000] <annevk>
sure

[13:19:02.0000] <smaug____>
ok, so we don't care about backwards compatibility

[13:19:03.0000] <annevk>
not with browsers

[13:19:04.0000] <annevk>
we care about compatibility with content

[13:20:00.0000] <smaug____>
that is new

[13:20:01.0000] <annevk>
not really

[13:20:02.0000] <smaug____>
"not with browsers"

[13:20:03.0000] <smaug____>
but I'm all for it

[13:20:04.0000] <smaug____>
we can simplify for example command handling significantly

[13:21:00.0000] <annevk>
sounds good to me :)

[13:26:00.0000] <smaug____>
command and menu handling... no need to care about the strange <select> and <input> handling in menu

[14:33:00.0000] <AryehGregor>
The proposals to remove various link relations are harmless, right?

[14:37:00.0000] <AryehGregor>
Another question: if a proposal has security issues, who are some good people to CC, like maybe from Mozilla or Opera?  I'm CCing abarth on this, but if I could get input from another implementer or two that would be nice.

[14:37:01.0000] <abarth>
AryehGregor: there are a bunch of moz folks who might be good to CC

[14:37:02.0000] <abarth>
maybe starting with bz or sid stamm might be good

[14:37:03.0000] <AryehGregor>
Okay.

[14:45:00.0000] <abarth>
AryehGregor: some variation of that idea could work

[14:45:01.0000] <AryehGregor>
Like what?

[14:46:00.0000] <abarth>
AryehGregor: if you think of it like a popup window that's constrained to the parent page's extent

[14:46:01.0000] <AryehGregor>
But whose position and focus you can't control?

[14:46:02.0000] <AryehGregor>
Does that actually satisfy the use-cases (not that I'm totally clear on what those are)?

[14:47:00.0000] <AryehGregor>
I guess the use-case is something like "let me embed stuff without the site being able to break out and take away my revenue from the banner ad I'm plastering on top of other people's sites".

[14:47:01.0000] <AryehGregor>
I mean, okay, it could be more innocuous, like Google Translate.

[14:47:02.0000] <abarth>
there are a couple interesting use cases

[14:47:03.0000] <abarth>
one tricky one is something like facebook connect

[14:48:00.0000] <abarth>
let's assume you're already signed into facebook, so you don't need to type your password

[14:48:01.0000] <abarth>
but you need to approve the connection between the site and facebook

[14:48:02.0000] <abarth>
facebook wants that UI to be displayed "on top" so that it can't be clickjacked

[14:48:03.0000] <abarth>
today, they'd need to use a popup window to get that sort of display

[14:48:04.0000] <AryehGregor>
What's wrong with that?

[14:48:05.0000] <abarth>
but they don't because popup windows are ugly, hard to use, and have trouble with popup blockers

[14:49:00.0000] <abarth>
they want the drawing model of a popup

[14:49:01.0000] <abarth>
but the interaction model of a lightbox

[14:49:02.0000] <AryehGregor>
Hmm, okay.

[14:49:03.0000] <AryehGregor>
Seems reasonable.

[14:49:04.0000] <abarth>
the harder case is when you need to type your facebook password

[14:49:05.0000] <abarth>
because you're not authenticated to facebook yet

[14:50:00.0000] <abarth>
then you need an address bar to tell you which site you're interacting with

[14:50:01.0000] <abarth>
which isn't possible using this mechanism

[14:50:02.0000] <karlcow>
In which spec atob has been moved?

[14:51:00.0000] <annevk>
karlcow, HTML, no?

[14:51:01.0000] <annevk>
http://www.whatwg.org/C#windowbase64

[14:53:00.0000] <karlcow>
hmmm

[14:53:01.0000] <karlcow>
http://dev.w3.org/html5/spec/

[14:54:00.0000] <gsnedders>
karlcow: There was an objection to it being included as it was a new feature after LC

[14:55:00.0000] <gsnedders>
(Despite it having been implemented for almost two decades…)

[14:55:01.0000] <AryehGregor>
It was added before LC.

[14:55:02.0000] <karlcow>
in fact I was looking where the data uris were defined. I wanted to know if there is a requirement for base64 on the specification.

[14:56:00.0000] <karlcow>
but could not find it

[14:56:01.0000] <gsnedders>
AryehGregor: Before? Then my memory must be wrong… Why was it removed then?

[14:56:02.0000] <gsnedders>
/me has found the commit now, but it gives no justification

[14:56:03.0000] <AryehGregor>
Because Sam filed a bug, so Hixie removed it, then I complained, the chairs clarified they didn't technically ask him to remove it, and Hixie re-added it.

[14:57:00.0000] <hsivonen>
Hixie: sharing the HTML tokenizer with the XML code path is not completely nuts. I've been pondering something similar.

[14:57:01.0000] <gsnedders>
karlcow: base64 for data URIs appears to be completely undefined.

[14:57:02.0000] <Hixie>
anyone know anything about RFC 4733? I'm trying to work out if you need a separate audio channel or if it can share the audio channel somehow but i'm not able to work it out

[14:57:03.0000] <Hixie>
hsivonen: really? seems like a world of pain

[14:57:04.0000] <karlcow>
gsnedders: ah interesting

[14:57:05.0000] <karlcow>
I was testing a few things about it

[14:58:00.0000] <hsivonen>
Hixie: the HTML tokenizer is *very* close to what one would want a non-recursive XML tokenizer to look like

[14:58:01.0000] <karlcow>
gsnedders: http://www.la-grange.net/2011/07/25/data-uri-base-test

[14:58:02.0000] <hsivonen>
Hixie: except the XML tokenizer needs to be able to deal with the internal subset to be compliant

[14:58:03.0000] <hsivonen>
hooray for the internal subset

[14:58:04.0000] <Hixie>
half the rfc seems to say that there's just one audio channel that's the same as the main one, and the other half seems to suggest you need to negotiate a telephone-events channels separately. *confused*

[14:59:00.0000] <Hixie>
hsivonen: i'd be worried about edge cases like NULLs and so on, but if you say so

[14:59:01.0000] <abarth>
hsivonen: you're aware that we're starting down that path in webkit, right?

[14:59:02.0000] <Hixie>
abarth: that's why we're discussing it

[14:59:03.0000] <hsivonen>
abarth: I just saw a mention in the IRC log

[14:59:04.0000] <abarth>
its somewhat experimental work

[15:00:00.0000] <annevk>
we have our own HTML and XML parser but have not done that

[15:00:01.0000] <hsivonen>
Hixie: I already added PI tokenization in order to be able to use the HTML tokenizer for syntax highlighting XML for View Source

[15:00:02.0000] <annevk>
maybe now WebKit implements their own XML parser they can implement XML5!

[15:00:03.0000] <hsivonen>
Hixie: it mishighlights the internal subset, though

[15:00:04.0000] <hober>
annevk: :)

[15:01:00.0000] <othermaciej>
our XML parser isn't yet ready to handle XML1

[15:01:01.0000] <hsivonen>
annevk: yeah, if the HTML tokenizer is reused, XML5 is very close

[15:01:02.0000] <annevk>
XML5 is easier :)

[15:02:00.0000] <annevk>
no need to check Name productions and such

[15:03:00.0000] <othermaciej>
our new XML tokenizer is similar to our HTML5 tokenizer and reuses some code, but it's not the same piece of code

[15:04:00.0000] <abarth>
its mostly the data structures / architecture that's shared

[15:04:01.0000] <abarth>
its going to have its own state machine, for example

[15:05:00.0000] <Hixie>
that makes more sense to me

[15:05:01.0000] <hsivonen>
I've been thinking of forking the tokenizer slightly for XML

[15:05:02.0000] <AryehGregor>
othermaciej, have you remembered about providing innerText feedback?  I've had feedback from Microsoft, Mozilla, and Opera since July 14, and am just waiting for WebKit.

[15:05:03.0000] <othermaciej>
AryehGregor: can you ping hober about it?

[15:05:04.0000] <hsivonen>
the main need to fork, AFAICT, is to perform qname splitting on colon in the tokenizer for efficiency

[15:05:05.0000] <hsivonen>
but on the HTML side, you don't want the tokenizer to do that

[15:05:06.0000] <othermaciej>
AryehGregor: I think the short version is that we won't remove it but I dunno our relative feelings about the other alternatives

[15:06:00.0000] <hsivonen>
but the states could be shared

[15:06:01.0000] <hsivonen>
some states like the script stuff would be unused on the XML side

[15:07:00.0000] <hsivonen>
while PI and internal subset states are needed on top of the HTML states

[15:07:01.0000] <AryehGregor>
othermaciej, okay.

[15:09:00.0000] <hober>
AryehGregor: consider me pinged :)

[15:15:00.0000] <jgraham>
/me occasionally hears complaints that the XML parser in Opera is lighting fast but it never appears in benchmarks

[15:15:01.0000] <jgraham>
s/it/XML parsing/

[15:15:02.0000] <jgraham>
So I guess I ought to be promoting XML on the web :)

[15:16:00.0000] <jgraham>
/me hasn't actually tested this though

[15:19:00.0000] <shepazu>
AryehGregor: no, but in SVG 2, I expect SVG fonts will be optional

[15:21:00.0000] <AryehGregor>
jgraham, make an XHR microbenchmark that involves downloading a gigantic XML file and then retrieving one trivial piece of data from it.

[15:21:01.0000] <AryehGregor>
Where the download is cached.

[15:21:02.0000] <gsnedders>
From the end of the document.

[15:24:00.0000] <jgraham>
AryehGregor: And give it a catchy name like "MoonArachnid" and tell people it represents browser performance :)

[15:27:00.0000] <jamesr>
XML parsing should matter for parsing pages served up as XML/XHTML

[15:27:01.0000] <jgraham>
All three of them

[15:27:02.0000] <jamesr>
guess the lesson is "don't optimize codepaths that nobody uses", then :D

[15:28:00.0000] <gsnedders>
jamesr: Until you have a customer who wants a [insane number here] MB SVG file to load quickly on something like a 200 MHz MIPS device, and will pay you for it. :P

[15:29:00.0000] <jamesr>
heh, true

[16:04:00.0000] <annevk>
/me summons Ms2ger

[16:04:01.0000] <annevk>
/me thinks step 2 in http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-ensure-samedoc cannot happen

[16:27:00.0000] <annevk>
hmm

[16:27:01.0000] <annevk>
WHATWG Weekly

[16:43:00.0000] <annevk>
suggestions anyone?

[16:44:00.0000] <annevk>
download="" / from-origin / traversal / window.find() / ???

[16:57:00.0000] <othermaciej>
annevk: end of W3C Last Call is coming up


2011-07-26
[17:00:00.0000] <annevk>
thanks othermaciej

[17:17:00.0000] <annevk>
http://blog.whatwg.org/weekly-end-of-last-call

[17:17:01.0000] <annevk>
also gave myself a week off :)

[18:04:00.0000] <annevk>
/me hopes abarth's URL spec will end up in HTML one day

[18:06:00.0000] <abarth>
annevk: yeah, me too :)

[18:06:01.0000] <abarth>
i have all the info, i just need to find a couple hours to do it

[18:21:00.0000] <annevk>
if Mozilla throws a NS_ERROR_INVALID_POINTER, is that an indication of a problem?

[18:23:00.0000] <jamesr>
there seems to be a porblem

[18:25:00.0000] <annevk>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3Ex%3Cscript%3Etry%7Bdocument.body.replaceChild%28null%2Cdocument.createElement%28%22x%22%29%29%7Dcatch%28e%29%7Bw%28e%29%7D%3C%2Fscript%3E

[18:42:00.0000] <annevk>
AryehGregor, you should maybe make some wiki with W3C Process issues

[18:42:01.0000] <Hixie>
damowmow.com/playground/ideas-for-the-w3c might help if you do that ;-)

[18:42:02.0000] <annevk>
AryehGregor, I mean, would you please make one :)

[18:42:03.0000] <annevk>
Hixie, did you look at chairs⊙wo? o_O

[18:43:00.0000] <Hixie>
what post?

[18:43:01.0000] <Hixie>
i don't read the secret lists anymore

[18:44:00.0000] <annevk>
maybe a good idea

[18:45:00.0000] <annevk>
but e.g. http://lists.w3.org/Archives/Member/chairs/2011JulSep/0053.html (W3C Member-only)

[18:45:01.0000] <annevk>
not that it makes much sense

[18:46:00.0000] <hober>
annevk: good times

[18:46:01.0000] <Hixie>
ah, well, daniel and i often disagree about process things :-)

[18:46:02.0000] <Hixie>
(i'm assuming he disagrees with me, i didn't do more than skim it)

[18:46:03.0000] <wolfman2000>
evening. *checks topic* hmm...remind me to do that more often.

[18:47:00.0000] <wolfman2000>
how goes progress here?

[18:47:01.0000] <annevk>
wolfman2000, there's http://blog.whatwg.org/

[18:48:00.0000] <annevk>
wolfman2000, and http://html5.org/tools/web-apps-tracker

[18:48:01.0000] <annevk>
both should give some indication of progress

[18:48:02.0000] <wolfman2000>
don't recall the blog the last time I was here.

[18:48:03.0000] <annevk>
wolfman2000, we even have http://twitter.com/WHATWG

[18:48:04.0000] <wolfman2000>
...should have known

[18:49:00.0000] <wolfman2000>
everyone is going to twitter

[18:49:01.0000] <wolfman2000>
next thing I know, you guys will say you have a Google+ account

[18:49:02.0000] <annevk>
Google+ does not allow accounts for groups yet

[18:49:03.0000] <wolfman2000>
don't know if that's good or bad

[18:49:04.0000] <Hixie>
i did sign up for the company profile thing for whatwg

[18:49:05.0000] <Hixie>
doubt we'll get in on teh beta though :-)

[18:49:06.0000] <annevk>
ah cool

[18:50:00.0000] <wolfman2000>
well, that covers one of my purposes here.

[18:51:00.0000] <wolfman2000>
The other purpose will depend on whether a particular user is active...and even then, I'd have to take that conversation to PM or a different channel

[18:51:01.0000] <annevk>
alright

[18:52:00.0000] <wolfman2000>
so hopefully it won't be a problem if I idle in here until he is around

[18:52:01.0000] <wolfman2000>
also, now following you guys on twitter to make things easier

[18:52:02.0000] <annevk>
there's no rules here

[18:54:00.0000] <wolfman2000>
strange thing, but I can adapt.

[19:23:00.0000] <Hixie>
annevk5: it's interesting how the responses kinda miss the point

[19:23:01.0000] <Hixie>
e.g. chaals talks about how the process i describe would make the editor "all powerful"

[19:23:02.0000] <Hixie>
but it would actually make the editor as powerless as the whole w3c is now

[19:29:00.0000] <benschwarz>
Hixie!

[19:58:00.0000] <erlehmann>
Hixie, could you suggest readings on community gouvernment that influenced you?

[20:17:00.0000] <Hixie>
benschwarz!

[20:17:01.0000] <Hixie>
erlehmann: community gouvernment?

[20:17:02.0000] <Hixie>
erlehmann: oh you mean like how to run a community?

[20:18:00.0000] <Hixie>
erlehmann: http://video.google.com/videoplay?docid=-4216011961522818645 or http://www.youtube.com/watch?v=-F-3E8pyjFo

[20:18:01.0000] <Hixie>
erlehmann: and similar talks

[20:18:02.0000] <Hixie>
in the first of those i think you can see the back of my head in the audience at one point :-)

[20:18:03.0000] <erlehmann>
Hixie, whatever you call it. i am interested for example, why you prefer descriptive over normative standards and stuff.

[20:19:00.0000] <Hixie>
oh, that's not really about community building

[20:19:01.0000] <Hixie>
hm

[20:19:02.0000] <erlehmann>
also, how to respond to volunteers without driving them off. i think the whatwg did a pretty good job with that.

[20:19:03.0000] <Hixie>
what makes you think i prefer descriptive standards over normative standards?

[20:19:04.0000] <Hixie>
how to respond to volunteers is exactly what those videos are about, so that answers that one

[20:20:00.0000] <erlehmann>
your comments on codecs in particular, like “speccing that so-and-so won't change apple's stance on it“

[20:20:01.0000] <Hixie>
(in fact just watch all the videos that you find in a search for [brian fitzpatrick and ben collins sussman])

[20:21:00.0000] <erlehmann>
ima gonna do that some time this week. thank you.

[20:22:00.0000] <erlehmann>
oh, and context: i maintain the dev branch of a minecraft clone. first project having significant contributions from people i do not personally know.

[20:22:01.0000] <Hixie>
erlehmann: "speccing that so-and-so won't change apple's stance on it" doesn't mean i prefer descriptive over normative standards

[20:22:02.0000] <Hixie>
erlehmann: it's just that you can't have normative standards if you don't have any power over the people who use them

[20:22:03.0000] <erlehmann>
oh. thanks for clearing that up, then.

[20:22:04.0000] <Hixie>
erlehmann: so descriptive standards are the only option on the webd

[20:22:05.0000] <Hixie>
web

[20:23:00.0000] <Hixie>
personally i'd much rather we had normative standards, then i could just spec what i wanted and not listen to any feedback, my life would be way easier :-P

[20:23:01.0000] <zewt>
text/hixie

[20:26:00.0000] <erlehmann>
there was a german joke made about that: „in relation to the browser makers, we are all petitioners“ – only that the german word for someone petitioning “bittsteller” can also be mis-construed as “someone who places bits”

[20:29:00.0000] <erlehmann>
a web dev said that at the end of a panel discussion in april. pretty good ad-hoc humor, in my opinion :)

[21:07:00.0000] <wolfman2000>
zewt: are you around?

[01:10:00.0000] <hsivonen>
I feel so baited to reply to http://www.jenitennison.com/blog/node/160#comment-11185 even though I know I should be reviewing and writing code today

[01:12:00.0000] <erlehmann>
>threads on google+

[01:12:01.0000] <erlehmann>
>deleted a sizable amount of my friends' accounts

[01:12:02.0000] <erlehmann>
do not want ;_;

[01:14:00.0000] <Hixie>
what on earth does it mean to be called a "Jacobin"? My ancient political history apparently fails me

[01:14:01.0000] <Hixie>
erlehmann: why did they delete accounts? did your friends escalate?

[01:15:00.0000] <jacobolus>
Hixie: they were a political club in revolutionary france

[01:15:01.0000] <Hixie>
hsivonen: it baffles me how people think standards have any power

[01:15:02.0000] <Hixie>
jacobolus: that much i gathered from wikipedia

[01:15:03.0000] <jacobolus>
Hixie: calling someone a Jacobin roughly implies someone who cares mainly about ideological purity

[01:16:00.0000] <erlehmann>
Hixie, they used their nick names. i have gone back to a normal feedreader since then.

[01:16:01.0000] <jacobolus>
hah. just looked at wikipedia, which apparently uses the same phrase as description :)

[01:17:00.0000] <Hixie>
erlehmann: you mean they violated the terms of service and then the terms of service were applied? i'm not saying i agree with the policy, but it seems silly to complain about a site doing what it said it would do.

[01:17:01.0000] <Hixie>
erlehmann: anyway, they can just put in the more common names they're actually known by and the accounts will be restored

[01:17:02.0000] <Hixie>
jacobolus: but what does it mean to be called a jacobian? is that good or bad?

[01:17:03.0000] <Hixie>
jacobolus: oh, sorry, missed your answer

[01:17:04.0000] <jacobolus>
it's certainly derogatory

[01:18:00.0000] <erlehmann>
sometimes the common name is just „plomlompom“ or something equally weird.

[01:18:01.0000] <hsivonen>
did the Gmail ToS prohibit pseudonyms, too?

[01:18:02.0000] <jacobolus>
it means you'll toss out allies who don't live up to an impossible extremist standard

[01:18:03.0000] <Hixie>
hsivonen: no (and still doesn't, to my knowledge)

[01:18:04.0000] <hsivonen>
the scariest part of G+ ToS getting applied is people getting their Gmail account closed, too

[01:19:00.0000] <hsivonen>
dunno if that's happened over pseudonyms, but, reportedly, it has happened over age requirement

[01:19:01.0000] <Hixie>
erlehmann: if someone really is called that then it should be easy to demonstrate it. but i can understand people doubting that someone is called that in person.

[01:20:00.0000] <erlehmann>
Hixie, i think you are absolutely right with the tos. i think they should move out of the google namespace.

[01:20:01.0000] <hsivonen>
maybe the age requirement is in the Gmail ToS, too, but it's terrible to close a non-American kid's email account without asking the parents when the U.S. law would allow parental consent and the kid isn't in the U.S. anyway

[01:20:02.0000] <Hixie>
hsivonen: as i understand it from media reports (i have no inside knowledge) there is one case where someone who was already violating the ToS (<13) had their account suspended pending parental approval after they tried to join G+ (which is closed to <18 anyway) and corrected their birthday.

[01:20:03.0000] <jacobolus>
Hixie: "The Jacobins assumed more and more power during the spring of 1793, with the support of the Parisian mob, which overawed the Convention, culminating in a coup at the end of May. They were to hold power until the summer of 1794, and they repeatedly purged the Convention of those they held disloyal to the Republic, ending with a widespread program of execution, the Reign of Terror in their last months. Robesp

[01:20:04.0000] <jacobolus>
ierre, generally the spokesman for the successful faction, had great esteem for his reputation as "the sea-green incorruptible", and set up the slogan of the Republic of Virtue, until the Jacobins' last purge, 9 Thermidor, July 27, 1794, Robespierre attempted suicide and only succeeded in shattering his lower jaw. He was executed the next day on Thermidor 10, July 28, 1794.[7]"

[01:20:05.0000] <Hixie>
hsivonen: but we already have a system in place that lets the parents unsuspend the account in that case, so i dunno why it made the news

[01:21:00.0000] <jacobolus>
there's wikipedia again

[01:21:01.0000] <erlehmann>
unfortunately, without g+ providing feeds and publish APIs, that is problematic. oh well.

[01:21:02.0000] <erlehmann>
walled gardens again.

[01:21:03.0000] <Hixie>
jacobolus: wow, glad he is reconsidering!

[01:21:04.0000] <hsivonen>
Hixie: oh. I guess it wasn't at all clear to the parents that there was such a system

[01:21:05.0000] <erlehmann>
my sister read that article. she's ten. uses google mail. though not for anything important, by her own account.

[01:21:06.0000] <erlehmann>
she's scared now. i showed her IMAP and will help her migrate.

[01:22:00.0000] <Hixie>
hsivonen: either that or they were trying to get attention. I've seen several cases where that's happened.

[01:22:01.0000] <hsivonen>
I've become more scared of having a foreign BigCo manage my identity when I've lately been trying to unlock an old Apple ID of mine

[01:23:00.0000] <erlehmann>
the funniest case in germany is a woman whone last name can be literally translated as „unknown“. of course social networks kill her accounts.

[01:23:01.0000] <Hixie>
/me is a big proponent of open federated standards for this kind of thing

[01:23:02.0000] <Hixie>
any one vendor being in control of anything is a dangerous situation, however benevolent the vendor.

[01:24:00.0000] <erlehmann>
indeed. i don't understand why my friends did that stuff. but then i consider g+ a big imageboard. though lacking features.

[01:24:01.0000] <hsivonen>
so far, iforgot.apple.com doesn't work for my old Apple ID, phone support directed me to iTunes email support and iTunes email support person is having scripted conversations with me and their scripts don't seem to cover dealing with decade-old Apple IDs that have never been used in the iTunes store

[01:24:02.0000] <erlehmann>
and by „did that stuff“, i mean „use it instead of using their own servers“.

[01:24:03.0000] <erlehmann>
oh itunes. the internet explorer 6 of the media players you are.

[01:26:00.0000] <hsivonen>
Are all new Apple IDs required to be email addresses? Is it even possible to register @-less Apple IDs anymore?

[01:26:01.0000] <erlehmann>
regarding google+: i can only hope that people realized they should not whine when stuff breaks. unless they pay for it.

[01:26:02.0000] <hsivonen>
where the @-part is implied to be @mac.com for iChat purposes

[01:26:03.0000] <hsivonen>
the Apple ID page still talks about iChat

[01:27:00.0000] <erlehmann>
it still looks like a PR problem from here. maybe only in germany.

[01:27:01.0000] <erlehmann>
or maybe only with early adopters from hacker culture.

[01:27:02.0000] <hsivonen>
if I register an Apple ID with @whatever and whatever doesn't have the XMPP DNS stuff, how would it work in iChat?

[01:35:00.0000] <hsivonen>
It's so frustrating to go back and forth with a support person who is following a script instead of paying good attention to what I say.

[01:36:00.0000] <erlehmann>
the only good support i ever had via phone was a short talk that amounted to „just remove the yellow disk on the underside with a screw driver and clean it“

[01:36:01.0000] <hsivonen>
If Apple has a policy against letting me revive my old Apple ID, I'd prefer them to tell me instead of spewing scripted answers to me

[01:37:00.0000] <hsivonen>
erlehmann: oh, Apple's phone support was clued but the phone guy wasn't empowered to poke at the Apple ID database

[01:37:01.0000] <erlehmann>
oha!

[01:37:02.0000] <hsivonen>
erlehmann: he pointed me to iTunes Store support and it went downhill from there

[01:38:00.0000] <erlehmann>
why do you need to revive it anyway? can't you just create a new one? or is stuff tied to it?

[01:39:00.0000] <erlehmann>
like .mac trial or something? (i am guessing here)

[01:39:01.0000] <hsivonen>
erlehmann: in case the ID is now or in the future exposed in some communication service (as it used to be exposed in iChat), I want to be known as "hsivonen" as in other services

[01:39:02.0000] <hsivonen>
if it won't be exposed anywhere public, I guess then I shouldn't care

[01:39:03.0000] <erlehmann>
oh well.

[01:39:04.0000] <erlehmann>
namespace issues again ;)

[01:42:00.0000] <hsivonen>
I supposed it's rational for them to by default assume that I'm clueless, but it's still so very annoying

[01:42:01.0000] <hsivonen>
*suppose

[01:45:00.0000] <erlehmann>
funny. one of my acquaintances was not deleted, though he has used his IRL nickname and added „dings“ (like the idiom „foo“)

[01:46:00.0000] <erlehmann>
i wonder if there is something done manually, he is somewhat popular.

[01:46:01.0000] <erlehmann>
oh well.

[01:57:00.0000] <hsivonen>
looks like Darth Vader got kicked out of G+, too

[03:07:00.0000] <hsivonen>
why does Chrome decode this as UTF-8? http://m.gsmweb.cz/

[03:08:00.0000] <hsivonen>
is Chrome applying on-by-default heuristics to detect UTF-8 when there's no declaration?

[03:08:01.0000] <hsivonen>
or is Chrome willfully violating HTML5 and honoring the XML decl?

[03:09:00.0000] <espadrine`>
What is the reason behind requiring btoa input to be below U+00FF? Is it technical, or historical?

[03:19:00.0000] <Philip`>
espadrine`: What would you expect it do for >= U+0100?

[03:21:00.0000] <Philip`>
(Base-64 is an algorithm that's defined over 8-bit values so it's unclear what should happen to larger values)

[03:22:00.0000] <Philip`>
(and you can't just do an implicit UTF-8 encode because then btoa and atob couldn't be exact inverses any more)

[04:45:00.0000] <hsivonen>
hmm. looks like Chrome is better at detecting Polish ISO-8859-2 text as ISO-8859-2 than Firefox, IE and Opera

[04:47:00.0000] <hsivonen>
whoa. Poland and Germany both have high Firefox usage but the usage by Firefox version is *very* different

[04:47:01.0000] <hsivonen>
oops. no matter. I was looking at the wrong numbers

[04:48:00.0000] <espadrine`>
Philip`: ah, ok (I was thinking indeed about a utf-8 encode)

[05:53:00.0000] <karlcow>
ePub previewer http://dl.dropbox.com/u/160781/apps/mac/Murasaki/index.html

[06:02:00.0000] <jgraham>
karlcow: Seems like one should just implement an ePub reader in HTML since they are basically HTML documents

[06:04:00.0000] <jgraham>
Alternative theory: it seems like ePub should just be plain HTML with a few specific conventions

[06:04:01.0000] <karlcow>
It's what is done already by most epubreaders I think. Not sure I understand what you mean :)

[06:05:00.0000] <hsivonen>
I'm feeling an 386 urge to reply to http://www.jenitennison.com/blog/node/160#comment-11185 again. I need to learn to look away.

[06:05:01.0000] <jgraham>
karlcow: Well I mean clearly that uses WebKit for the rendering. But I mean you may as well make the rest of the UI HTML too. Especially for a "preview" reader

[06:06:00.0000] <jgraham>
Although I understand that some people will value the OSX integration

[06:07:00.0000] <Philip`>
/me likes the non-word "unanimosity"

[06:07:01.0000] <jgraham>
hsivonen: You could reply briefly with only the most important points

[06:07:02.0000] <karlcow>
jgraham: yup I found it by looking for an ePub Spotlight importer to just be able to full text search on the files I have.

[06:09:00.0000] <karlcow>
hsivonen: if you want to make constructive points to some things which seem general enough, do not leave a comment, but create a FAQ on your web site and point to it.

[06:09:01.0000] <karlcow>
The FAQ will have the benefits to be able to grow across time and subsequent blog posts

[06:14:00.0000] <Ms2ger>
/me wonders what http://www.w3.org/wiki/DOM/domcore is about

[06:15:00.0000] <hsivonen>
karlcow: maybe I should do that, but having a page that catalogs incorrect allegations and misunderstandings about me, the stuff I'm working on and stuff I don't want to work on would be a rather odd thing to have on my site

[06:15:01.0000] <Philip`>
http://www.legislation.gov.uk/%0C - yay XML

[06:16:00.0000] <hsivonen>
Philip` strikes again

[06:21:00.0000] <Philip`>
Hmm, their inline sitestat script code seems to not execute, because it's <script><!-- ... --></script> but served as XML so the comment gets ignored

[06:23:00.0000] <Philip`>
(which is lucky since http://www.legislation.gov.uk/"+alert(document.cookie)+" would be an XSS hole if it did execute)

[06:24:00.0000] <Philip`>
(Oddly it does seem to prevent "--" being injected into the script code, but nothing else)

[06:25:00.0000] <Philip`>
/me guesses that's the consequence of using an XML templating system that thinks the page name is actually meant to be inserted into a legitimate comment

[06:26:00.0000] <Philip`>
(which is presumably introducing the vulnerability for any users without an XHTML-supporting browser)

[06:29:00.0000] <Philip`>
(Sticking an HTML5 serialiser on the output wouldn't help since that'll serialise the comment node inside the script element as <script><!-- ... and result in it becoming executable too)

[06:30:00.0000] <Philip`>
/me would like it if parsing tree-structured data wasn't so very complex :-(

[06:31:00.0000] <Philip`>
s/parsing/serialising and parsing/

[07:29:00.0000] <gsnedders>
"inserted into a document" means if you follow parentNode you'll reach the Document node, right?

[07:33:00.0000] <Ms2ger>
That's "in a document" in DOM Core/HTML

[07:34:00.0000] <gsnedders>
Right, and inserted into is when it enters that state?

[07:34:01.0000] <gsnedders>
Yay I can understand specs. :3

[07:36:00.0000] <Philip`>
Something is in a document once it's been inserted into a document? Who would've guessed!

[07:36:01.0000] <jgraham>
/me wonders what he wanted to say to Ms2ger 

[07:38:00.0000] <jgraham>
Oh yes

[07:38:01.0000] <Philip`>
/me wonders what the Plain English Campaign would think about the specs

[07:38:02.0000] <jgraham>
Ms2ger: Do you have a link to the testharness.js/Mochitest integration bug?

[07:39:00.0000] <Ms2ger>
https://bugzilla.mozilla.org/show_bug.cgi?id=647323

[07:39:01.0000] <jgraham>
Takk

[07:40:00.0000] <Ms2ger>
/me should figure out at what line length annevk wraps

[07:45:00.0000] <annevk>
uh?

[07:49:00.0000] <jgraham>
/me wonders which text editor annevk uses

[07:50:00.0000] <Philip`>
/me ponders what font annevk picked

[07:50:01.0000] <Ms2ger>
/me ponders Philip` 

[07:51:00.0000] <annevk>
oh I get it now

[07:51:01.0000] <annevk>
76

[07:51:02.0000] <annevk>
TextWrangler, Monaco-12

[07:52:00.0000] <Philip`>
An excellent choice

[07:52:01.0000] <jgraham>
Now you should blog about how that exact setup is the key to writing good specs

[07:53:00.0000] <Ms2ger>
Pff, 80, gedit, Monospace-10 beats that :)

[07:54:00.0000] <annevk>
gedit is nice

[07:55:00.0000] <jgraham>
/me was at least right in thinking annevk isn't an emacs/vi type

[07:56:00.0000] <Ms2ger>
Don't tell Hixie

[08:04:00.0000] <gsnedders>
Philip`: Well, question of whether inserted means in the tree or has ownerDocument.

[08:13:00.0000] <foolip>
does anyone know how to leave a W3C WG? http://www.w3.org/2004/01/pp-impl/42786/join exists but /leave does not

[08:15:00.0000] <jgraham>
/me tries to avoid humming hotel california

[08:16:00.0000] <Philip`>
foolip: Wait a year until you get automatically unsubscribed?

[08:16:01.0000] <Philip`>
/me wonders if that only applies to the HTML WG

[08:16:02.0000] <annevk>
foolip, you should talk to Charles

[08:18:00.0000] <annevk>
foolip, the /join form goes via Charles too

[08:18:01.0000] <foolip>
annevk, k

[08:43:00.0000] <Ms2ger>
"User agents may wish to offer the user more control over the provided media."

[08:43:01.0000] <Ms2ger>
/me wonders if that's an RFC2119 MAY

[08:51:00.0000] <Ms2ger>
<AryehGregor> git commands have seldom been faulted for having an insufficient number of options.

[08:51:01.0000] <Ms2ger>
Hah

[08:51:02.0000] <david_carlisle>
annevk: thanks for closing that bug

[08:53:00.0000] <annevk>
your welcome

[08:57:00.0000] <annevk>
btw

[08:57:01.0000] <annevk>
I learned my lesson

[08:58:00.0000] <annevk>
one Microsoft employee in the water, 500 dollar potential Apple hardware damages

[09:14:00.0000] <Ms2ger>
annevk, step 2 at http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-ensure-samedoc was for sync mutation events, not sure if it was a good idea

[09:15:00.0000] <annevk>
oh because adoptNode removes

[09:15:01.0000] <gsnedders>
Window.prototype.toString doesn't exist per WebIDL, right?

[09:16:00.0000] <annevk>
I think the idea is that the mutation event fire after all mutations happened

[09:16:01.0000] <annevk>
so an implicit call to adoptNode would not fire one

[09:16:02.0000] <annevk>
prolly need to split out the adoptNode algorithm to handle that scenario

[09:16:03.0000] <annevk>
better

[09:20:00.0000] <Ms2ger>
gsnedders, why would it?

[09:22:00.0000] <annevk>
btw

[09:22:01.0000] <gsnedders>
Ms2ger: Because Jeff Walden implied it did :)

[09:22:02.0000] <annevk>
the shadow DOM stuff apparently introduces a new Node type

[09:22:03.0000] <Ms2ger>
Boo, Waldo

[09:26:00.0000] <dglazkov>
good morning, Whatwg

[09:26:01.0000] <dglazkov>
annevk: yes!

[09:53:00.0000] <annevk>
so what is the plan dglazkov?

[09:53:01.0000] <annevk>
is Ian going to edit the specification?

[09:53:02.0000] <annevk>
or you?

[09:55:00.0000] <dglazkov>
It's going to be me and Dominic Cooney, for starters.

[09:55:01.0000] <dglazkov>
"for starters" is here, btw: http://dglazkov.github.com/component-model/

[09:56:00.0000] <dglazkov>
the only section that's more or less spec'd is events: http://dglazkov.github.com/component-model/events.html

[09:56:01.0000] <dglazkov>
I essentially took Hixie's events section from XBL2 and wrote it down in algorithms.

[09:57:00.0000] <dglazkov>
unfortunately, I need to work on something different this quarter (WebKit hygiene related stuff)

[09:58:00.0000] <dglazkov>
so I am 20%-ing on the use cases/spec work at the moment

[09:59:00.0000] <dglazkov>
but I'll be back with the vengeance in October!! :)

[10:00:00.0000] <dglazkov>
annevk: does this make any sense? :)

[10:01:00.0000] <Ms2ger>
The web? None

[10:01:01.0000] <dglazkov>
Ms2ger: all we have to do is slightly adjust the definition of "sense"

[10:02:00.0000] <Ms2ger>
Apply a binding to the word, if you will

[10:02:01.0000] <dglazkov>
Ms2ger: you know me, I would much rather sub-class it

[10:02:02.0000] <dglazkov>
class WebSense : Sense

[10:02:03.0000] <Ms2ger>
{

[10:03:00.0000] <Ms2ger>
  // Reimplement everything

[10:03:01.0000] <dglazkov>
/ override all methods

[10:03:02.0000] <dglazkov>
:)

[10:04:00.0000] <annevk>
/me reads

[10:05:00.0000] <Ms2ger>
/me cleans up annevk's typos

[10:05:01.0000] <annevk>
sounds good dglazkov

[10:05:02.0000] <annevk>
you should probably announce that spec on WebApps

[10:06:00.0000] <dglazkov>
annevk: yes.

[10:13:00.0000] <annevk>
Ms2ger, I had the idea of not <dfn>'ing historical constants

[10:14:00.0000] <annevk>
Ms2ger, Web IDL already requires everything there is to need to know about them

[10:14:01.0000] <annevk>
Ms2ger, and they do not need to be referenced

[10:14:02.0000] <Ms2ger>
That works

[10:15:00.0000] <gsnedders>
/me would like Anolis to automatically semantically mark-up IDL blocks

[10:16:00.0000] <gsnedders>
(and make it possible to xref them, etc.)

[10:16:01.0000] <Ms2ger>
You mean, turn it into ReSpec? :)

[10:18:00.0000] <gsnedders>
ReSpec didn't exist when I started talking about implementing that :)

[10:18:01.0000] <Ms2ger>
Also, patches welcome

[10:19:00.0000] <annevk>
semantically how?

[10:20:00.0000] <annevk>
haven't changed exceptions because of the outstanding bug

[10:21:00.0000] <gsnedders>
annevk: basically making it possible to just type a pre/code block of WebIDL with no elements within it

[10:22:00.0000] <gsnedders>
annevk: And getting the same output as today

[10:22:01.0000] <annevk>
yeah that would be kind of cool

[10:23:00.0000] <Ms2ger>
Someone (who isn't me) should write a python WebIDL parser

[10:24:00.0000] <gsnedders>
Ms2ger: This was what I concluded. :)

[10:24:01.0000] <Ms2ger>
So, is the DOM Tree a tree?

[10:24:02.0000] <Ms2ger>
/me enjoys www-dom

[10:26:00.0000] <annevk>
/me replied with a link to wikipedia

[10:27:00.0000] <dglazkov>
oh good lord

[10:27:01.0000] <dglazkov>
/me feels bad for the poor guy

[10:27:02.0000] <Ms2ger>
gal*

[10:27:03.0000] <dglazkov>
pesone

[10:28:00.0000] <dglazkov>
^^ person that is

[10:28:01.0000] <annevk>
Ms2ger, can http://dev.w3.org/2009/webidl-checker/webidl-check be reused in some way

[10:29:00.0000] <Ms2ger>
Wait, that's python?

[10:29:01.0000] <Ms2ger>
/me is confused

[10:29:02.0000] <annevk>
combined with XSLT

[10:29:03.0000] <dglazkov>
PYTHON IS AWESOME

[10:30:00.0000] <annevk>
oh it looks more like a nice hack

[10:30:01.0000] <annevk>
heycam, is there some Python parser for Web IDL that you know of?

[10:31:00.0000] <heycam>
annevk, not that I know of

[10:31:01.0000] <annevk>
also hi

[10:31:02.0000] <heycam>
hi :)

[10:31:03.0000] <heycam>
you might like to look at that es-operating-system project? to see what parser they're using

[10:32:00.0000] <zewt>
well that's lovely

[10:32:01.0000] <zewt>
"wait 30 seconds to download" sites that stop the countdown if you change tabs

[10:33:00.0000] <zewt>
three cheers for degenerate scripting abuses

[10:35:00.0000] <annevk>
http://code.google.com/p/es-operating-system/source/browse/#svn%2Ftrunk%2Fesidl

[10:39:00.0000] <zewt>
heh www-dom

[10:39:01.0000] <zewt>
should have suggested that a tree refers to the roots

[10:41:00.0000] <zewt>
i'd have assumed it was a troll if it wasn't from a named domain

[10:44:00.0000] <Philip`>
DOM Tree doesn't make any sense because it doesn't photosynthesise

[10:53:00.0000] <Hixie>
so i'm thinking <dialog> should also have a <form> feature: <form method=dialog> which causes form submission to just look for the nearest ancestor dialog and set its return value accordingly

[10:53:01.0000] <Hixie>
so you could do: <dialog> <form method=dialog> <button type=submit value=1>Ok</button> <button type=submit value=0>Cancel</button> </form> </dialog>

[10:54:00.0000] <Hixie>
and then: dialog.showModal(); dialog.onclose = function () { if (dialog.returnValue == '1') { ... } };

[10:54:01.0000] <dglazkov>
Hixie: <dialog> sounds pretty cool

[10:54:02.0000] <annevk>
for some reason I read <details> and got confused

[10:54:03.0000] <Hixie>
(instead of having to hook each button up to the <dialog>, etc)

[10:54:04.0000] <Hixie>
this also gets you form validation for free

[10:54:05.0000] <Hixie>
so you can use <input required>, etc

[10:54:06.0000] <Hixie>
and it all magically works

[10:55:00.0000] <dglazkov>
so returnValue is what?

[10:55:01.0000] <Hixie>
the value of the button the user used to submit the form

[10:55:02.0000] <dglazkov>
what if I have a dialog to ask for a new label name?

[10:56:00.0000] <Hixie>
you can get the value from the form by just grabbing the form control and picking the input element and getting its value

[10:56:01.0000] <Hixie>
dialog.getElementsByTagName('form')[0].newLabelName.value

[10:57:00.0000] <Hixie>
(with <input name=newLabelName>)

[10:57:01.0000] <annevk>
will these dialogs be written in markup though?

[10:57:02.0000] <Hixie>
haven't really done the research yet but i'm guessing yes

[10:57:03.0000] <zewt>
is there a reason to do this in a surrounding dialog element, instead of just using events on the form?

[10:57:04.0000] <Hixie>
(still need to see if it's what people want by studying how it's done today)

[10:58:00.0000] <Hixie>
zewt: how would that look?

[10:58:01.0000] <annevk>
zewt, you need a container that will be the dialog

[10:58:02.0000] <zewt>
akin to onsubmit, i imagine

[10:59:00.0000] <Hixie>
how would you write the example above?

[11:00:00.0000] <zewt>
something like form.method = "none"; /* submitting triggers events but doesn't submit to the server */ form.onaftersubmit = function(e) { form.returnValue ... }

[11:01:00.0000] <Philip`>
Should OK/Cancel buttons get automatically hooked up to keys like Enter/Escape and other UI like 'X' buttons, rather than just being arbitrary submit buttons?

[11:01:01.0000] <Hixie>
zewt: how do you get the dialog to show up in this model?

[11:02:00.0000] <Hixie>
Philip`: ok/cancel are an antipattern in dialog design, i don't think we should be encouraging their use. I was just lazy in my example.

[11:02:01.0000] <zewt>
i don't know anything about <dialog>, i'm looking at it from the javascript-form perspective

[11:02:02.0000] <Hixie>
zewt: ah well if all you want is a form, you can already do that using action="javascript:..."

[11:02:03.0000] <zewt>
where it seems like you sort-of want f.onsubmit = function(e) { e.preventDefault(); }, but you don't want to actually stop the submission entirely--still go through validation and so on, just don't send it to the server

[11:02:04.0000] <TabAtkins_>
Not for much longer, if at all anymore.

[11:02:05.0000] <zewt>
ew

[11:02:06.0000] <Philip`>
Hixie: Keyboard shortcuts and default choices seem important regardless of what the buttons are labelled with

[11:03:00.0000] <Hixie>
Philip`: ah if you mean should it be possible to mark a button as the default button, that's already on the list of features for a future version

[11:04:00.0000] <zewt>
can you do form.action = function in JS? code in a javascript: url is evil

[11:05:00.0000] <nlogax>
not afaik. but why not add an event listener?

[11:05:01.0000] <zewt>
listener on what?

[11:05:02.0000] <Hixie>
zewt: why is it evil?

[11:06:00.0000] <zewt>
how is it not evil? heh

[11:06:01.0000] <Hixie>
what's the difference between onclick="foo()" and action="javascript:foo()" ??

[11:06:02.0000] <Ms2ger>
That the latter is evil :)

[11:07:00.0000] <zewt>
i can say x.click = function() in JS; x.action = "foo()" in JS is nasty

[11:07:01.0000] <zewt>
onclick

[11:07:02.0000] <Hixie>
oh well if you're talking about scripted stuff sure

[11:07:03.0000] <Hixie>
i assumed you were talking about markup

[11:07:04.0000] <nlogax>
zewt: submit listener on the form

[11:07:05.0000] <zewt>
whether prefixed with javascript: or not (that just makes it worse)

[11:08:00.0000] <Philip`>
Scripts in attribute strings are effectively executed using eval, and eval is inherently evil

[11:08:01.0000] <zewt>
nlogax: being able to trigger validation without performing a server submit

[11:08:02.0000] <nlogax>
zewt: you can prevent the default action

[11:09:00.0000] <nlogax>
form.addEventListener("submit", function(e) { e.preventDefault(); do some stuff; }, false)

[11:10:00.0000] <zewt>
i assumed that would prevent form validation as well (havn't used form validation yet)

[11:11:00.0000] <nlogax>
i think an invalid form is not possible to submit anyway.. but only used it once and forgot

[11:11:01.0000] <nlogax>
>_>

[11:11:02.0000] <nlogax>
/me reads

[11:11:03.0000] <zewt>
too tired at the moment to think much harder about it

[11:11:04.0000] <zewt>
two hours of sleep is fun

[11:20:00.0000] <erlehmann>
i am of sleepy as well.

[11:28:00.0000] <annevk>
I wonder how we should organize the Nodes section

[11:28:01.0000] <annevk>
the current order makes little sense

[11:28:02.0000] <annevk>
maybe they should be in nodeType order

[11:29:00.0000] <annevk>
with the exception of comment

[11:29:01.0000] <annevk>
that should just come after Text and not ProcessingInstruction

[11:30:00.0000] <Ms2ger>
Tree order? Doc/PI/doctype/Element/CharData/Text/Comment/DocFrag

[11:31:00.0000] <annevk>
that seems kind of nice actually, again with Comment out of order

[11:32:00.0000] <Ms2ger>
I just want (CharData/Text/Comment) together

[11:36:00.0000] <annevk>
yeah

[12:32:00.0000] <AryehGregor>
So it looks like one of the two bugs I filed in IE Connect was silently fixed at some point.

[12:32:01.0000] <AryehGregor>
Encouraging.  Maybe I'll file some more.

[12:33:00.0000] <zewt>
was it reflexively WONTFIX'd anyway

[12:33:01.0000] <zewt>
heh

[12:34:00.0000] <annevk>
god I wish Mozilla obsoleted Attr already

[12:35:00.0000] <annevk>
and other browsers too of course

[12:35:01.0000] <annevk>
but I mainly want to clean up the DOM spec :p

[12:35:02.0000] <Ms2ger>
We warn already!

[12:36:00.0000] <AryehGregor>
zewt, no, both of my bugs got quick form responses saying: Thank you for your feedback.

[12:36:01.0000] <AryehGregor>
We were able to reproduce the issue and are investigating it.

[12:36:02.0000] <AryehGregor>
Best regards,

[12:36:03.0000] <AryehGregor>
The Internet Explorer Team

[12:36:04.0000] <AryehGregor>
Which is actually impressive by itself.

[12:36:05.0000] <karlcow>
each time a Website is using it, make Firefox sends an email to the spec editor ;)

[12:37:00.0000] <zewt>
i'd say "buy a lottery ticket" if randomness worked that way

[12:37:01.0000] <Ms2ger>
karlcow, to the website author ;)

[12:37:02.0000] <AryehGregor>
Also, they kept them open after shipping IE9.  One is still open, the other is fixed, and I confirmed that the fixed one is actually fixed in 10PP2.

[12:37:03.0000] <annevk>
D/DF/PI/DT/E/T/C is the new order

[12:37:04.0000] <AryehGregor>
So actually my only complaint is that I don't seem to get e-mail for status updates, and there's no explanation of when the fixed one was fixed.

[12:38:00.0000] <Ms2ger>
Weird place for DF

[12:38:01.0000] <annevk>
I also abstracted out the adopting stuff Ms2ger

[12:38:02.0000] <annevk>
it's close to Document

[12:38:03.0000] <karlcow>
Ms2ger, I really wish we had a simple way of having a feedback loop for people in charge of websites. It is hard to impossible to find the right persons, web agencies, departments.

[12:39:00.0000] <Ms2ger>
Same here

[12:39:01.0000] <zewt>
if only there was always actually someone in charge of websites :|

[12:39:02.0000] <Ms2ger>
annevk, great

[12:39:03.0000] <annevk>
i think the new adopt stuff is quite good yes :)

[12:40:00.0000] <annevk>
now I just need to solve xxxhierarchy

[12:40:01.0000] <annevk>
and see how the mutating algorithms overlap

[12:40:02.0000] <annevk>
so there's only a few places to patch eventually for mutation events

[12:41:00.0000] <wolfman2000>
afternoon

[12:42:00.0000] <Ms2ger>
Night

[12:42:01.0000] <annevk>
almost midday

[12:45:00.0000] <karlcow>
http://www.niso.org/topics/ccm/e-book_annotation/

[12:45:01.0000] <karlcow>
Books in Browsers Meeting

[12:45:02.0000] <karlcow>
 October 26, 2011

[12:46:00.0000] <Ms2ger>
How is the hit-testing going?

[12:46:01.0000] <Ms2ger>
(Pushed)

[12:46:02.0000] <annevk>
tantek is finding out it is hard :)

[12:48:00.0000] <Ms2ger>
He's managing to make others write spec text, though

[12:49:00.0000] <annevk>
maybe he can make you do more than fix typos :p

[12:50:00.0000] <Ms2ger>
Noooo

[12:51:00.0000] <annevk>
tantek, btw, will you remove the pseudo-elements from CSS3 UI?

[12:52:00.0000] <tantek>
annevk: lol - what I was finding out is just how many *more* unprefixed CSS properties have been implemented.

[12:52:01.0000] <tantek>
(which may affect hit-testing yes)

[12:53:00.0000] <tantek>
regarding the pseudo-elements in CSS3-UI, I've already placed them all "at risk"

[12:56:00.0000] <Ms2ger>
<ChrisL> can we please drop the 'levels' stuff

[12:56:01.0000] <Ms2ger>
Hear, hear

[13:01:00.0000] <tantek>
and will be keeping ::value and ::choices per implementer interest/intent (Mozilla) - but at this point looking at dropping the ::repeat-* pseudo-elements.

[13:01:01.0000] <zewt>
gar

[13:02:00.0000] <zewt>
next thing browsers need: modal save-as prompts

[13:02:01.0000] <zewt>
which is ... hard :|

[13:03:00.0000] <wolfman2000>
Maybe look at jquery to see how they did their modal stuff?

[13:06:00.0000] <hober>
zewt & wolfman2000: please contribute what you find to http://wiki.whatwg.org/wiki/Dialogs

[13:06:01.0000] <annevk>
tantek, hmm, if they are going to be used for shadow DOM like stuff it might be better to wait for that

[13:08:00.0000] <tantek>
annevk - my understanding is that ::value and ::choices are simple enough that it will fairly easy to map them to shadow DOM like stuff when that is defined.

[13:08:01.0000] <tantek>
and if there is only one implementation of the pseudo-elements then they won't exit CR

[13:08:02.0000] <tantek>
so I'm ok with seeing if browser implement them or not and keeping/dropping accordingly.

[13:08:03.0000] <tantek>
*browserS

[13:10:00.0000] <annevk>
I think for shadow DOM though instead of having a bunch of pre-defined pseudo-elements we want something else

[13:10:01.0000] <annevk>
::shadow(ident)

[13:10:02.0000] <annevk>
and in your shadow tree you have <div pseudo=test> which can then be matched from the outside using ::shadow(test)

[13:11:00.0000] <annevk>
dglazkov might know more

[13:11:01.0000] <annevk>
because otherwise you need to keep updating the selectors grammar and in general it would not be very extensible

[13:17:00.0000] <dglazkov>
annevk: I tried to convince dhyatt to do something like ::shadow(ident), but he insisted that we just use pseudo-elements.

[13:17:01.0000] <dglazkov>
annevk: and allow builders of shadow DOM to define them with "pseudo" attribute.

[13:18:00.0000] <annevk>
but that does not work

[13:18:01.0000] <annevk>
::foobar { } is dropped from a style sheet

[13:19:00.0000] <annevk>
unrecognized pseudo-elements do not make it past parsing

[13:19:01.0000] <dglazkov>
annevk: not in WebKit

[13:19:02.0000] <annevk>
that's a CSS compliance issue then

[13:20:00.0000] <dglazkov>
annevk: right http://lists.w3.org/Archives/Public/www-style/2011Apr/0769.html

[13:20:01.0000] <dglazkov>
WebKit. we're awesome.

[13:21:00.0000] <annevk>
well that extension makes sense

[13:21:01.0000] <annevk>
but that is not about not dropping unrecognized pseudo-elements on the floor

[13:22:00.0000] <dglazkov>
how would you know if a pseudo-element is unrecognized?

[13:22:01.0000] <annevk>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%3Cstyle%3E%3A%3Atesttest%2Cbody{background%3Ared}%3C%2Fstyle%3E

[13:23:00.0000] <annevk>
only red in webkit

[13:23:01.0000] <annevk>
violating forwards compatible parsing rules

[13:24:00.0000] <wolfman2000>
hober: I may have misunderstood your request. I'm looking at the slashdot example on the dialog wiki page, and that looks like jquery to me.

[13:24:01.0000] <annevk>
WebKit. we've awesomely undocumented extensions and do not push for them in standards bodies

[13:24:02.0000] <annevk>
;)

[13:24:03.0000] <dglazkov>
;)

[13:25:00.0000] <dglazkov>
maybe TabAtkins should do it

[13:25:01.0000] <dglazkov>
he's smart

[13:26:00.0000] <Ms2ger>
CSS3WEBKIT?

[13:29:00.0000] <tantek>
annevk - longer term it seems the ::shadow(ident) approach makes more sense. which is why I'm fairly conservative with adding new pseudo-elements. ::value and ::choices have specific uses / use-cases and implementer interest which is why I'm even considering them.

[13:37:00.0000] <gsnedders>
B2G — an entirely memory unsafe OS, on a limited memory platform?

[13:41:00.0000] <Ms2ger>
Limited memory and Gecko in one sentence?

[13:43:00.0000] <zewt>
"gecko causes me to have limited memory"

[13:45:00.0000] <gsnedders>
Ms2ger: Well, isn't that the implication of B2G? Or going down the Maemo/N900 route of needing swap space on microSD or the like?

[13:46:00.0000] <Ms2ger>
Dunno, I'm happy to have skipped that thread

[13:47:00.0000] <zewt>
swap on sd? sounds like a good way to have to replace an sdcard every couple months, heh

[14:06:00.0000] <AryehGregor>
How do you get raw HTML hosted at github, like es5.github.com?  Does it require $$$?

[14:21:00.0000] <karlcow>
AryehGregor: http://pages.github.com/ if you like solutions which are not free, tied to a corporate for hosting ;)

[14:22:00.0000] <AryehGregor>
Yeah, so it's not free.  Figures.

[14:31:00.0000] <Philip`>
Do any modern browsers not have document.images?

[14:32:00.0000] <Ms2ger>
Gecko does, that much my source tells me

[14:34:00.0000] <Philip`>
Hmm, looks like IE4+, NN3+, O3+

[14:34:01.0000] <Philip`>
per http://blooberry.com/indexdot/html/topics/uastring-navobj.htm

[14:39:00.0000] <AryehGregor>
Okay, I'd say my editing spec is ready for implementer review.

[14:39:01.0000] <AryehGregor>
/me writes up posts

[15:23:00.0000] <AryehGregor>
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-July/032630.html

[15:27:00.0000] <jamesr>
AryehGregor: tl;dr

[15:27:01.0000] <AryehGregor>
jamesr, :(

[15:27:02.0000] <jamesr>
AryehGregor: :D

[15:27:03.0000] <Hixie>
man, harsh

[15:29:00.0000] <jamesr>
is the intent that this spec be detailed enough to clean-room from the spec and be compatible with the web?

[15:29:01.0000] <AryehGregor>
jamesr, yes.

[15:29:02.0000] <jamesr>
i'm wondering if anyone would try to do an html5-parser style rewrite+ditch

[15:29:03.0000] <AryehGregor>
I'm pretty sure that's the only feasible strategy here.

[15:29:04.0000] <AryehGregor>
I don't know if it's quite ready for that, though.

[15:29:05.0000] <jamesr>
for interop, yeah

[15:30:00.0000] <AryehGregor>
It still has some known issues, and I'm sure there are loads of problems I didn't think of at all.

[15:30:01.0000] <karlcow>
sometimes I wonder if specs should not just been small index card. linked together

[15:31:00.0000] <AryehGregor>
One problem with this is that interop is so bad right now that sites that rely on execCommand all use massive browser sniffing.

[15:31:01.0000] <AryehGregor>
So any convergence will have to break sites in at least one browser, probably.

[15:31:02.0000] <AryehGregor>
But that's not avoidable.

[15:31:03.0000] <jamesr>
also a certain amount of content is inherently engine-specific

[15:31:04.0000] <AryehGregor>
Yes, we all know WebKit has that problem.

[15:31:05.0000] <AryehGregor>
Well, and IE, but they just gave up and started keeping around the old engines, so it's not an issue for them.

[15:31:06.0000] <jamesr>
it's a significant issue in WebKit, but probably true to some degree for everyone

[15:32:00.0000] <jamesr>
i guess apple could keep another version of WebKit.framework around for Mail.app :/

[15:32:01.0000] <AryehGregor>
The algorithms should be compatible enough in most cases, because usually sites aren't going to depend on the exact DOM produced.

[15:32:02.0000] <AryehGregor>
Or so one hopes.

[15:33:00.0000] <AryehGregor>
Unlike with the HTML parser.

[15:33:01.0000] <AryehGregor>
But there will definitely be some migration pain here.

[15:33:02.0000] <AryehGregor>
Still, no way around it.

[15:35:00.0000] <zewt>
well, you *could* avoid backwards-compatibility, in principle, by speccing a new function ("newExecCommand"), everyone implements that, then declare execComment to be inherently browser-specific for backwards-compat only, as if it was a prefixed vendor extension

[15:35:01.0000] <zewt>
not very pretty, of course

[15:35:02.0000] <jamesr>
if you're going to do that, might as well go ahead and design something that makes more sense

[15:35:03.0000] <AryehGregor>
WebKit can do that if they want to preserve compat with WebKit-specific content, but it's not reasonable for the web in general.

[15:36:00.0000] <AryehGregor>
We need to define something for execCommand() that's basically web-compatible.

[15:36:01.0000] <zewt>
jamesr: well, that way means that the new API gives a pretty simple migration for existing code

[15:36:02.0000] <jamesr>
that's basically saying "let's design a new editing API", which is different from standardizing execCommand()

[15:36:03.0000] <zewt>
whereas a new sparkly API may not

[15:36:04.0000] <AryehGregor>
jamesr, well, a redesign from scratch risks introducing unforeseen design flaws, and it would be harder for authors to port their sites.

[15:36:05.0000] <AryehGregor>
But yeah, this isn't an option, we're standardizing what's out there.

[15:36:06.0000] <zewt>
also a much bigger investment implementation- and design-side

[15:36:07.0000] <AryehGregor>
Most users probably won't break.

[15:36:08.0000] <jamesr>
fingers crossed!

[15:36:09.0000] <AryehGregor>
I mean, most sites that use it.

[15:37:00.0000] <zewt>
breaking users sounds like a fascinating side-effect

[15:37:01.0000] <zewt>
*crunch*

[15:37:02.0000] <AryehGregor>
If they do, we'll figure out what to do at the time.  I'm happy to change the spec to be more compatible if people identify specific sites that break.

[15:55:00.0000] <AryehGregor>
Hixie, was the WebSockets protocol ever part of the HTML5 draft at the W3C, before it moved to the IETF?

[15:56:00.0000] <annevk>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%3Cscript%3Evar%20df%20%3D%20document.createDocumentFragment()%3B%20df.appendChild(document.createElement(%22x%22))%3Bdocument.replaceChild(df%2Cdocument.documentElement)%3C%2Fscript%3E only works in Firefox

[15:56:01.0000] <annevk>
hmm

[15:57:00.0000] <franksalim>
AryehGregor: a predecessor protocol was as part of TCPConnection

[15:57:01.0000] <annevk>
guess expanding DocumentFragment nodes for insertion and append operations isn't really well supported

[15:58:00.0000] <AryehGregor>
Opera 11.50 is weird.

[15:58:01.0000] <Hixie>
AryehGregor: no idea off-hand

[15:59:00.0000] <AryehGregor>
It replaces the element successfully, but it sticks in an empty <head> and <body>.

[15:59:01.0000] <AryehGregor>
Hixie, k.

[16:00:00.0000] <annevk>
AryehGregor, http://www.w3.org/TR/2009/WD-html5-20090212/comms.html#network

[16:01:00.0000] <AryehGregor>
annevk, thanks.

[16:02:00.0000] <annevk>
I guess I will go with what Firefox did

[16:03:00.0000] <annevk>
DocumentFragment, bah

[16:03:01.0000] <zewt>
heh DRM talk on the list--yeah right

[16:05:00.0000] <franksalim>
AryehGregor: I just realized you said "at the W3C." please disregard my last message

[16:19:00.0000] <Hixie>
ok i just replied to the oldest e-mail in my folders, from 2007

[16:20:00.0000] <TabAtkins_>
Hah, awesome.

[16:20:01.0000] <Hixie>
there's some 2008 e-mail about contenteditable that i'll look at eventually but right now it's not included in my counts because aryeh's working on editing

[16:25:00.0000] <Hixie>
is anyone working on speccing the fullscreen api?

[16:32:00.0000] <AryehGregor>
Is there a Bugzilla I could use for the editing API stuff?  If I start getting nontrivial amounts of feedback, it would be a lot more convenient than e-mail.

[16:32:01.0000] <annevk>
Hixie, I planned on doing it

[16:33:00.0000] <annevk>
but then didn't

[16:33:01.0000] <Hixie>
AryehGregor: prod mikesmith to get you a component in the bugzilla

[16:33:02.0000] <Hixie>
annevk: ah, awesome

[16:33:03.0000] <annevk>
heycam was next in line

[16:33:04.0000] <annevk>
well not awesome

[16:33:05.0000] <Hixie>
oh

[16:33:06.0000] <annevk>
so far nothing happened :)

[16:33:07.0000] <AryehGregor>
Nobody will mind that the spec isn't at the W3C?

[16:33:08.0000] <Hixie>
i see

[16:33:09.0000] <Hixie>
AryehGregor: not if you don't make a fuss about it

[16:33:10.0000] <Hixie>
:-)

[16:33:11.0000] <AryehGregor>
The problem is if someone else later does after I started relying on it.

[16:33:12.0000] <annevk>
has not happened so far

[16:34:00.0000] <annevk>
e.g. with DOM Range

[16:34:01.0000] <Hixie>
AryehGregor: the W3C has kindly agreed to allow us to use the bugzilla for non-w3c specs

[16:34:02.0000] <AryehGregor>
Oh, okay then.

[16:34:03.0000] <AryehGregor>
I didn't realize DOM Range had a component.

[16:36:00.0000] <Hixie>
annevk: i'm trying to work out where to send all this feedback

[16:37:00.0000] <annevk>
attach it to a bug report?

[16:37:01.0000] <Hixie>
there's like 50 e-mails

[16:37:02.0000] <annevk>
and assign when it's clear there's an editor or something

[16:37:03.0000] <annevk>
email it to www-archive?

[16:37:04.0000] <annevk>
and add it to Specifications TODO

[16:37:05.0000] <annevk>
on the wiki

[16:38:00.0000] <Hixie>
i'll just move it to a bucket that doesn't count against my numbers i guess

[16:38:01.0000] <Hixie>
and add a note on the wiki

[16:38:02.0000] <annevk>
or you know, just spec it :)

[16:38:03.0000] <Hixie>
not gonna happen

[16:38:04.0000] <Hixie>
:-P

[16:39:00.0000] <Hixie>
annevk: hey do you want my range and traversal buckets?

[16:41:00.0000] <annevk>
traversal yes

[16:41:01.0000] <annevk>
range not yet (maybe AryehGregor wants that for now)

[16:41:02.0000] <annevk>
traversal is now in DOM Core

[16:42:00.0000] <annevk>
and I suspect Range will one day be too

[16:42:01.0000] <AryehGregor>
I could take the Range buckets, but I don't know how I'd actually respond to them in the format you're sending me.

[16:42:02.0000] <annevk>
you just reply to the individuals

[16:42:03.0000] <AryehGregor>
So just lots and lots and lots of copy-paste?

[16:43:00.0000] <annevk>
hixie could resend each of them

[16:43:01.0000] <annevk>
i suppose

[16:43:02.0000] <heycam>
annevk, cpearce is implementing it for us, maybe he would like to do speccing work too? :)

[16:44:00.0000] <cpearce>
hmmm? I working on what?

[16:44:01.0000] <heycam>
full screen

[16:44:02.0000] <cpearce>
ah yes...

[16:44:03.0000] <cpearce>
I've basically implemented https://wiki.mozilla.org/index.php?title=Gecko:FullScreenAPI

[16:44:04.0000] <Hixie>
AryehGregor: i can also just bounce all the mails to you as separate e-mails

[16:44:05.0000] <cpearce>
without all the security checks yet ;)

[16:45:00.0000] <AryehGregor>
Hixie, well, if you want me to spend time systematically responding to all the editing and range stuff, I don't mind.

[16:45:01.0000] <Hixie>
annevk: looks like hte only thing i have about traversal is about what happens if a filter moves its iterator or treewalker

[16:45:02.0000] <AryehGregor>
Probably would be easier for me if you forwarded them all to me in some format where I could reply all to get the individual plus the list.

[16:46:00.0000] <Hixie>
annevk: you defined that?

[16:46:01.0000] <Hixie>
AryehGregor: i'm more interested in making sure the points are addressed than that they get replies at this point

[16:46:02.0000] <Hixie>
AryehGregor: i can send the replies later when i do the review work

[16:47:00.0000] <Hixie>
AryehGregor: as i mentioned in /msg, it would help me learn the spec :-)

[16:47:01.0000] <AryehGregor>
k.

[16:47:02.0000] <AryehGregor>
(got to go now)

[16:47:03.0000] <Hixie>
later

[16:48:00.0000] <annevk>
Hixie, yes

[16:48:01.0000] <annevk>
Hixie, it throws

[16:48:02.0000] <Hixie>
hah, i chave a CSS 2.1 bucket with over 100 e-mails in it

[16:48:03.0000] <Hixie>
annevk: ok, i'll delete this mail then

[16:49:00.0000] <Hixie>
how about dom events, do you want a multipart/digest of those?

[16:49:01.0000] <Hixie>
got some mail from 2004 in this bucket

[16:49:02.0000] <Hixie>
and a whatwg mail from 2005

[16:50:00.0000] <annevk>
Hixie, I guess

[16:50:01.0000] <annevk>
Hixie, kind of depends if they are about the core part or UI events

[16:51:00.0000] <Hixie>
didn't check

[16:51:01.0000] <Hixie>
sent you a pile of feedback

[16:51:02.0000] <Hixie>
see comment in the e-mail

[16:51:03.0000] <Hixie>
i have a folder here named "cssom--anne-says-eta-2015", wonder what that's about

[16:52:00.0000] <annevk>
suspect new features or something

[16:53:00.0000] <Hixie>
TabAtkins_: do you recall if you've gone through all that feedback i sent you about css lists? (i have a bucket here that says i should delete it once you're confident you've dealt with it)

[16:53:01.0000] <Hixie>
/me deletes two e-mails from hyatt dated 2003 about the flexbox model

[16:54:00.0000] <Hixie>
ok seriously, i'm deleting this css 2.1 feedback. i hope it all got dealt with.

[16:54:01.0000] <annevk>
should maybe email to elika

[16:55:00.0000] <Hixie>
it's all ancient stuff

[16:56:00.0000] <annevk>
k

[16:56:01.0000] <Hixie>
oh heck, a websqldb folder

[16:56:02.0000] <Hixie>
/me deletes

[16:57:00.0000] <Hixie>
ok that's some good cleanup


2011-07-27
[17:00:00.0000] <annevk>
I think the switch to relying on DOM Core was intended to be more than a reference switch, but I guess we can file specific bugs

[17:06:00.0000] <Hixie>
annevk: i'm going through http://www.w3.org/Bugs/Public/show_bug.cgi?id=11305 now

[17:08:00.0000] <annevk>
cool

[17:11:00.0000] <Hixie>
annevk: how do i file dom core bugs

[17:12:00.0000] <annevk>
WebApps WG, component DOM Core

[17:13:00.0000] <annevk>
http://www.w3.org/Bugs/Public/enter_bug.cgi?product=WebAppsWG

[17:13:01.0000] <Hixie>
ta

[17:15:00.0000] <annevk>
reading DOM Range it does look like it would be good to have inside DOM Core

[17:20:00.0000] <annevk>
DocumentFragment sucks

[17:21:00.0000] <annevk>
well, insertion methods combined with it suck

[17:21:01.0000] <annevk>
they are build for a single Node, but DocumentFragment can hold a bunch of them

[17:21:02.0000] <timeless>
heycam: ping

[17:21:03.0000] <heycam>
timeless, pong

[17:21:04.0000] <timeless>
heycam: it'd be *incredibly* beneficial for me to be able to link to #whatever-undefined

[17:22:00.0000] <timeless>
to be able to talk about `undefined`

[17:22:01.0000] <timeless>
WebIDL doesn't seem to have any anchors for `undefined`

[17:22:02.0000] <timeless>
it just has <span class="esvalue">undefined</span>

[17:22:03.0000] <heycam>
undefined in what ense?

[17:22:04.0000] <heycam>
the JS value undefined?

[17:22:05.0000] <timeless>
yeah

[17:22:06.0000] <timeless>
i want to reply to a spec which says:

[17:22:07.0000] <heycam>
why do you need to link to it? :)

[17:22:08.0000] <timeless>
foo.STUPID_CONSTANT

[17:22:09.0000] <timeless>
> This attribute is optional.

[17:23:00.0000] <timeless>
"Ok, does that mean it's http://w3/WebIDL/#undefined ?"

[17:23:01.0000] <heycam>
ok. so I don't know if there really should be a #undefined... it's purely a ECMAScript-defined thing

[17:23:02.0000] <timeless>
/me nods

[17:23:03.0000] <timeless>
i know

[17:23:04.0000] <heycam>
just like a don't have a link for Number

[17:23:05.0000] <timeless>
well

[17:24:00.0000] <timeless>
i'm 99% certain CSS has a link for Number :)

[17:24:01.0000] <timeless>
and actually, the same applies for Number..

[17:24:02.0000] <timeless>
it'd be really useful to be able to link to those things

[17:24:03.0000] <heycam>
do you just want an anchor somewhere in the #es-attribute and #es-constant sections?

[17:24:04.0000] <heycam>
no actually I wouldn't even want to say anything about undefined there in webidl...

[17:25:00.0000] <timeless>
defintely not that section

[17:25:01.0000] <timeless>
you could just add <undefined> into http://dev.w3.org/2006/webapi/WebIDL/#es-type-mapping

[17:25:02.0000] <timeless>
it's technically not a type

[17:25:03.0000] <timeless>
but it'd sure make my life simpler

[17:26:00.0000] <timeless>
(stick it near void and explain that it really devolves to void...)

[17:26:01.0000] <heycam>
well, that's only true for operation return values

[17:26:02.0000] <timeless>
/me nods

[17:26:03.0000] <heycam>
I'm not actually sure where it would go, or what the sentence might say

[17:26:04.0000] <timeless>
i don't have a good suggestion

[17:26:05.0000] <timeless>
but basically, this spec is probably supposedly referencing webidl

[17:27:00.0000] <Hixie>
timeless: why don't you link to the ES spec instead?

[17:27:01.0000] <timeless>
Hixie: offhand i don't think webidl links to it

[17:27:02.0000] <timeless>
which means i probably can't find it :)

[17:27:03.0000] <timeless>
oh, it does

[17:27:04.0000] <timeless>
but not where i was looking

[17:27:05.0000] <Hixie>
http://people.mozilla.org/~jorendorff/es5.html

[17:27:06.0000] <timeless>
right, the fact that es is a pdf

[17:28:00.0000] <TabAtkins_>
Hixie: Yes, I've gone through it all.

[17:28:01.0000] <timeless>
is a pretty big reason for me not linking to it...

[17:28:02.0000] <timeless>
i guess http://people.mozilla.org/~jorendorff/es5.html#sec-4.3.9 isn't a terrible link

[17:28:03.0000] <Hixie>
TabAtkins_: awesome, more e-mail i can delete!

[17:28:04.0000] <timeless>
asside from being incredibly unofficial

[17:29:00.0000] <annevk>
timeless, there's also http://es5.github.com/

[17:29:01.0000] <timeless>
/me needs to seriously consider changing the primary display (or at least the one that has the taskbar) to be the right one

[17:29:02.0000] <annevk>
timeless, see http://platform.html5.org/

[17:29:03.0000] <timeless>
/me keeps hovering the mouse over the right edge of the left display (roughly center of screen) and having w7 hide all windows

[17:30:00.0000] <timeless>
annevk: hrm, http://es5.github.com/#x4.3.9 isn't terrible

[17:30:01.0000] <timeless>
not great

[17:30:02.0000] <timeless>
mostly terrible :)

[17:30:03.0000] <timeless>
an anchor that has the word `undefined` in it is much better than a random number :)

[17:31:00.0000] <annevk>
yeah dunno

[17:31:01.0000] <annevk>
standards organizations are generally pretty crappy

[17:31:02.0000] <timeless>
+1

[17:31:03.0000] <annevk>
ECMA guys publish in PDF

[17:31:04.0000] <annevk>
IETF in plain text

[17:31:05.0000] <annevk>
W3C in HTML4

[17:31:06.0000] <timeless>
with a req for XHTML, no? :)

[17:31:07.0000] <annevk>
none of them really seems to care about the web much

[17:31:08.0000] <timeless>
+1

[17:32:00.0000] <Philip`>
WHATWG publishes in pages that crash your browser

[17:32:01.0000] <timeless>
+1

[17:32:02.0000] <timeless>
Philip`: to be fair, i have a number of those myself

[17:32:03.0000] <jcranmer>
RFC in XML

[17:32:04.0000] <timeless>
the sad thing is that my aunt has one

[17:32:05.0000] <Philip`>
and with a zillion different variants so you can never tell which one has which feature

[17:32:06.0000] <jcranmer>
well, the RFCs are writtin in XMLs

[17:32:07.0000] <heycam>
timeless, your aunt has an RFC?

[17:32:08.0000] <jcranmer>
and published as plain text and html

[17:32:09.0000] <timeless>
heycam: no, a page she publishes that crashes a browser

[17:33:00.0000] <timeless>
https://bugs.meego.com/show_bug.cgi?id=18260#c8

[17:33:01.0000] <timeless>
jcranmer: yeah, i know

[17:33:02.0000] <timeless>
my colleague (one door over) is writing an RFC :)

[17:33:03.0000] <timeless>
/me keeps trickle feeding him comments

[17:33:04.0000] <jcranmer>
timeless: if I make a browser that loads every page by doing *(int*)0

[17:33:05.0000] <jcranmer>
everybody has a page that crashes a browser!

[17:34:00.0000] <timeless>
jcranmer: yeah well

[17:34:01.0000] <timeless>
we tried suggesting to her that perhaps such a long page wasn't a good idea

[17:34:02.0000] <timeless>
she didn't buy the argument

[17:34:03.0000] <timeless>
and she definitely wasn't willing to buy the demo'd product

[17:34:04.0000] <timeless>
(it really really FAILed)

[17:35:00.0000] <timeless>
heycam: anyway, i think i'd still rather have something in webidl

[17:35:01.0000] <timeless>
even if it's just a table listing the actual sections in ES

[17:35:02.0000] <timeless>
it'd be a public service :)

[17:36:00.0000] <timeless>
(useful named anchors, what a concept!)

[17:36:01.0000] <heycam>
timeless, send a mail to public-script-coord then, I'll think about it :)

[17:37:00.0000] <timeless>
ok

[17:37:01.0000] <jamesr>
ecmascript 'undefined' is reassignable

[17:37:02.0000] <timeless>
oh, i saw that you posted a list of known emails :)

[17:37:03.0000] <timeless>
jamesr: yeah, i know

[17:37:04.0000] <timeless>
one of my favorite features

[17:38:00.0000] <Hixie>
annevk: is XML doc vs HTML doc affecting "the case-sensitivity of some selectors" no longer true?

[17:38:01.0000] <heycam>
void 0 ftw

[17:38:02.0000] <timeless>
yep

[17:40:00.0000] <annevk>
Hixie, that is still true in general, but we are trying to change a few things

[17:40:01.0000] <Hixie>
k

[17:41:00.0000] <annevk>
Hixie, in HTML type and attribute names are case-insensitive

[17:41:01.0000] <annevk>
Hixie, I'm trying to get the magic list for HTML attributes whose values are matched case-insensitively removed

[17:42:00.0000] <Hixie>
i love how dom core takes all the easy stuff and leaves all the complicated stuff

[17:42:01.0000] <timeless>
Hixie: someone had to claim the easy stuff first!

[17:42:02.0000] <timeless>
(that reminds me of Nokia)

[17:42:03.0000] <Hixie>
i claimed it first!

[17:42:04.0000] <Hixie>
then they stole it!

[17:42:05.0000] <Hixie>
:-P

[17:42:06.0000] <annevk>
Hixie, what complicated stuff?

[17:43:00.0000] <Hixie>
like the document's address, and the reload override flag

[17:43:01.0000] <annevk>
we could do document URL I suppose but you would have to set it correctly etc.

[17:44:00.0000] <Hixie>
the less you take the easier it is for me :-P

[17:44:01.0000] <annevk>
it would just be a concept, just like charset / media type

[17:44:02.0000] <timeless>
aww, this spec didn't have any spelling errors

[17:44:03.0000] <timeless>
that's no fun

[17:44:04.0000] <timeless>
well, they misspelled JavaScript once

[17:44:05.0000] <timeless>
but everyone does that :)

[17:44:06.0000] <annevk>
document URL is an open issue

[17:45:00.0000] <annevk>
I raised it on the WHATWG list

[17:45:01.0000] <timeless>
hrm

[17:45:02.0000] <timeless>
anyone here understand how paste works?

[17:45:03.0000] <Hixie>
annevk: some of this stuff is kinda making the spec harder to understand. e.g. not mentioning the HTML parser is what puts a doc into quirks mode

[17:46:00.0000] <timeless>
whenever i paste something from Notepad into IE/Firefox on pastebin.mozilla.org

[17:46:01.0000] <Hixie>
annevk: and removing the examples in HTMLCollection

[17:46:02.0000] <timeless>
I get doubled lines for lines I write

[17:46:03.0000] <timeless>
but only single lines for lines that have >

[17:46:04.0000] <annevk>
Hixie, I did not move that stuff, but we can certainly add some examples

[17:47:00.0000] <timeless>
http://www3.ttc.ca/

[17:47:01.0000] <timeless>
/me chuckles

[17:47:02.0000] <timeless>
/me wonders why ttc did that

[17:47:03.0000] <annevk>
Hixie, as well as mentioning how quirks mode is set; the reason we moved it is because it's global

[17:47:04.0000] <annevk>
if you feel any of this is on the wrong side we can certainly reevaluate

[17:48:00.0000] <Hixie>
annevk: um, you say the default character encoding is UTF-8

[17:48:01.0000] <Hixie>
but it's UTF-16

[17:48:02.0000] <annevk>
but this allows CSS to hook into the DOM without having to hook into HTML

[17:48:03.0000] <Hixie>
i think it's the right side

[17:48:04.0000] <Hixie>
i just think moving it is making the specs harder to read

[17:48:05.0000] <Hixie>
ideally we would just have one spec and it wouldn't be an issue! :-)

[17:49:00.0000] <annevk>
'If the attribute is present in an XML document, its value must be an ASCII case-insensitive match for the string "UTF-8"'

[17:49:01.0000] <annevk>
"utf-16" (including quotes) yields no results in HTML

[17:50:00.0000] <Hixie>
?

[17:50:01.0000] <Hixie>
the document's character encoding defaults to UTF-16 in WA1 and to UTF-8 in DOM Core

[17:50:02.0000] <Hixie>
i believe UTF-16 is the correct default

[17:50:03.0000] <annevk>
never mind me

[17:52:00.0000] <Hixie>
annevk: should i file a bug?

[17:53:00.0000] <annevk>
w((document.implementation.createHTMLDocument("test")).characterSet) gives UTF-8 in Gecko

[17:56:00.0000] <Hixie>
i'll have to test it more later

[17:56:01.0000] <Hixie>
i gotta go

[17:59:00.0000] <annevk>
k, I tried testing it a bit more, but I'm getting a bunch of undefined things for these attributes

[18:02:00.0000] <annevk>
basically it's not DocumentFragment so much as it is having these methods available on Document

[18:56:00.0000] <tantek>
/me notices discussion about fullscreen API (cc: Hixie annevk cpearce)

[19:09:00.0000] <tantek>
Hixie, I've got it in my queue after CSS3-UI: https://wiki.mozilla.org/Tantek-Mozilla-projects#Full_Screen

[19:10:00.0000] <tantek>
assuming I get permission to publish it with a CC0+OWFa license, where would it go on the whatwg site?

[19:46:00.0000] <tantek>
until there's an actual spec draft - feel free to edit / add to https://wiki.mozilla.org/Gecko:FullScreenAPI and I'll incorporate from there.

[20:30:00.0000] <mahir256>
hello?

[21:16:00.0000] <Hixie>
tantek: we don't have a system for publication yet but we can figure something out

[21:16:01.0000] <Hixie>
tantek: easiest for me is to just give you a subdomain, e.g. css.whatwg.org

[21:19:00.0000] <tantek>
Hixie - I suppose I should research how whatwg specs are organized.

[21:19:01.0000] <Hixie>
there is no pattern

[21:19:02.0000] <Hixie>
right now there's only one spec, and it keeps changing :-)

[21:20:00.0000] <tantek>
for some reason I thought there were multiple specs and editors :)

[21:20:01.0000] <Hixie>
not currently

[21:21:00.0000] <Hixie>
though we keep talking about it

[21:21:01.0000] <tantek>
what about the spec that Aryeh was saying he's working on

[21:21:02.0000] <Hixie>
usually around the same time an editor who mainly works at the w3c gets frustrated with the w3c :-)

[21:21:03.0000] <Hixie>
aryeh's stuff was intended to get merged into the html spec

[21:21:04.0000] <tantek>
oh ok

[21:21:05.0000] <Hixie>
not sure what we'll end up doing

[21:21:06.0000] <tantek>
interesting - so this is a bit of a novel situation :)

[21:21:07.0000] <Hixie>
there's a lot of playing by ear here :-)

[21:22:00.0000] <tantek>
makes sense :)

[21:23:00.0000] <Hixie>
process is for weenies :-P

[21:27:00.0000] <tantek>
heh - I suppose that makes me a bit of a process weenie then: http://microformats.org/wiki/process

[21:27:01.0000] <Hixie>
:-)

[21:28:00.0000] <tantek>
Hixie, in other news, it looks like I've picked up the task to integrate / update your work on hit-testing into the CSS3-UI spec.

[21:28:01.0000] <tantek>
"your work" meaning this email message: http://lists.w3.org/Archives/Public/www-style/2009Feb/0287.html

[21:29:00.0000] <Hixie>
that was mostly the work of someone else at opera iirc

[21:29:01.0000] <Hixie>
but man, that was a while ago

[21:31:00.0000] <tantek>
somewhere I've seen somebody say something about things just taking longer.

[21:31:01.0000] <Hixie>
hehe

[21:32:00.0000] <tantek>
if you have any updates on hit-testing, let me know - I'll be starting from that email message, and then analyzing this set of notes from Leif about interop challenges: http://lists.w3.org/Archives/Public/www-style/2010Aug/0407.html

[21:32:01.0000] <Hixie>
i'm happy to say that i know nothing new

[21:32:02.0000] <tantek>
good to know.

[21:34:00.0000] <tantek>
well at least your assumption that "the CSS specifications will at some future point deal with the issue" is being realized (regarding hit-testing)

[21:34:01.0000] <Hixie>
heh

[21:34:02.0000] <tantek>
CSS3-UI seems like the logical place to document hit-testing, and there's consensus on that from CSS/SVG WGs.

[21:34:03.0000] <Hixie>
cool

[21:35:00.0000] <tantek>
or perhaps just that no-one else wanted to pick it up, and I was foolish enough to do so.

[21:36:00.0000] <Hixie>
i know _that_ feeling

[21:36:01.0000] <tantek>
lol

[21:37:00.0000] <tantek>
I definitely had that feeling of, what, *this* is not defined? Then asked "How is this not an essential part of the open web platform?" (silence)

[21:38:00.0000] <tantek>
And then I asked why it wasn't defined in HTML5 (what I would have expected) which then led to your email message.

[21:38:01.0000] <tantek>
And the subsequent realization that there was only one logical course of action.

[21:38:02.0000] <Hixie>
in a somewhat futile attempt to scope the work i have on my plate i've drawn a line around anything presentational and have left that for, well, apparently, you.

[21:42:00.0000] <tantek>
/me has a feeling he'll be spending more time in #whatwg.

[22:23:00.0000] <jamesr>
annevk: yt? what's the preferred URL to cite for DOMCore (assuming i can w3 bullshit, for the moment)

[22:24:00.0000] <annevk>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html until we reach a more stable state

[22:29:00.0000] <jamesr>
/me global search-replaces links to /TR/2004/REC-DOM-Level-3-Core-20040407/

[22:29:01.0000] <annevk>
which spec is this?

[22:29:02.0000] <annevk>
or is this source code?

[22:29:03.0000] <jamesr>
http://dvcs.w3.org/hg/webperf/raw-file/tip/specs/RequestAnimationFrame/Overview.html

[22:30:00.0000] <jamesr>
will update in a sed

[22:30:01.0000] <jamesr>
sec

[22:33:00.0000] <annevk>
aah, roc's API

[22:33:01.0000] <annevk>
nice

[22:33:02.0000] <jamesr>
yeah it's a good one. just have to write it down concretely enough :)

[22:35:00.0000] <jamesr>
another noob question: when citing something from whatwg HTML, do you link to the multipage version or the single page version?

[22:35:01.0000] <jamesr>
in the spec itself all internal links go to the same version

[22:35:02.0000] <jamesr>
but you can't pull that trick from an external reference. i know hixie would say it should all be in the same spec :)

[22:36:00.0000] <annevk>
I used http://www.whatwg.org/html for now

[22:37:00.0000] <annevk>
you can append #fragments to the end of that URL and redirects will happen automatically

[22:37:01.0000] <jamesr>
ah cool, so since the current default is multipage it'll go to the multipage version?

[22:37:02.0000] <annevk>
yes that will always go to the multipage

[22:37:03.0000] <annevk>
see e.g. on http://dev.w3.org/csswg/cssom-view/

[22:38:00.0000] <annevk>
search for "queue a task" or some such

[22:39:00.0000] <jamesr>
http://www.whatwg.org/html/#queue-a-task. any reason for that over http://www.whatwg.org/html#queue-a-task or http://whatwg.org/html#queue-a-task ?

[22:39:01.0000] <annevk>
no

[22:40:00.0000] <annevk>
it's just what our scripts outputs

[22:40:01.0000] <annevk>
we generate those links automatically

[22:40:02.0000] <annevk>
see https://bitbucket.org/ms2ger/dom-core for the setup

[22:40:03.0000] <annevk>
Anolis is quite nice for HTML specs

[22:45:00.0000] <jamesr>
my problem is i only get a few hours every few weeks to work on actually editing specs, which isn't enough time to get up to speed and stay up to speed with how to do it efficiently

[22:46:00.0000] <annevk>
fair enough

[22:46:01.0000] <annevk>
I will let you know when we made this easy enough so you can be set up in a couple of minutes

[22:47:00.0000] <annevk>
(which probably will be a while)

[22:48:00.0000] <Hixie>
jamesr: that's how i feel about coding :-)

[22:50:00.0000] <jamesr>
annevk: well for this spec there's currently an xml+makefile+xslt setup that heycam got working, and that seems to do something

[22:50:01.0000] <jamesr>
i guess i could just start with the .html output of that and then figure out how to munge it into the input anolis expects

[22:54:00.0000] <annevk>
it's what HTML, XHR, DOM Core, etc. are all using, but if the current setup works than it might not be worth the trouble if you have that little bandwidth

[22:56:00.0000] <jamesr>
is the input format documented anywhere?

[22:56:01.0000] <jamesr>
man, do i have to clone https://bitbucket.org/ms2ger/anolis/src in order to view README.html?

[22:57:00.0000] <annevk>
jamesr, we have https://bitbucket.org/ms2ger/spec-template

[23:00:00.0000] <annevk>
actually, that might be out of date

[23:01:00.0000] <annevk>
easiest is just looking at Overview.src.html files

[23:35:00.0000] <annevk>
what is so hard about specifying commonAncestorContainer ?

[23:35:01.0000] <annevk>
other than that the Range specification looks really nice

[23:56:00.0000] <annevk>
http://oreilly.com/catalog/0636920021360

[23:56:01.0000] <annevk>
via myakura on G+

[23:56:02.0000] <annevk>
book on HTML5 FileSystem API

[23:56:03.0000] <annevk>
(not that there's such a thing in HTML)

[00:28:00.0000] <zewt>
74 pages isn't a "book", heh

[00:59:00.0000] <hsivonen>
http://www.jenitennison.com/blog/node/160#comment-11188

[02:15:00.0000] <hsivonen>
so, it looks like multiple people are so far upset about AryehGregor not submitting the HTML Editing API spec to the W3C and are attributing AryehGregor's email to the "WHATWG"

[02:35:00.0000] <jgraham>
All Fear The Cabal!

[02:44:00.0000] <jgraham>
But seriously, unless people get over the WHATWG-as-conspiracy rhetoric and actually look at real issues (W3C Process having a big impedence mismatch with the needs of authors and implementors, W3C high level goals as expressed through e.g. focus of the TAG being very different to the needs of its core consituents), this kind of conflict is going to continue

[02:46:00.0000] <espadrine`>
hsivonen: I think his choice was reasonable. If you refuse Minix' license, create Linux. If you refuse W3C's license, create stuff your way and choose a different license for it. I guess I don't really understand the fuss around Aryeh Gregor's actions.

[02:49:00.0000] <hsivonen>
jgraham: what I read into AryehGregor's email was that he needs to become an employee (of Google) in order be able to use time to deal with the W3C

[03:09:00.0000] <jgraham>
hsivonen: The reason doesn't seem that important. The fact is that someone has done a bunch of difficult work that has the potential to enhance the web platform and people are getting upset because that person isn't willing to do even more work to deal with Process *even though* he has set things up so that anyone who feels strongly about Process can do that part of the work instead

[03:10:00.0000] <jgraham>
I can't understand the sense of entitlement one needs to have an attitude like that

[05:43:00.0000] <Dashiva>
Is _anything_ really a natural fit for XML?

[05:43:01.0000] <Ms2ger>
Pain.

[05:44:00.0000] <Dashiva>
Obviously you can fit almost anything into XML, but claiming something is a natural fit seems like a strong claim

[05:49:00.0000] <jgraham>
Well if you have data with just the right level of structure (more than key-value pairs but less than a graph), don't care about representation efficiency and don't care that small errors in the input will break everything, I guess it isn't a bad fit

[05:53:00.0000] <Philip`>
The alternatives are generally worse, though

[05:53:01.0000] <jgraham>
For which usecase?

[05:56:00.0000] <Philip`>
/me can't remember what he was thinking now

[05:59:00.0000] <hsivonen>
Dashiva: DocBook?

[06:07:00.0000] <JirkaK>
Highly structured documents, that's for what SGML and XML was invented and is natural fit. Today most known and used formats are DocBook, DITA and TEI

[06:09:00.0000] <Dashiva>
Such documents often have needs that translate into overlapping markup

[06:11:00.0000] <karlcow>
Dashiva: I think there is also a question of culture (habits). I'm more used to XML toolchain, so it is a lot easier for me to benefit from the "strictness" of XML than trying to reinvent it with the DOM.

[06:11:01.0000] <hsivonen>
JirkaK: who actually uses DITA these days (except for Eliot Kimber)?

[06:11:02.0000] <JirkaK>
No, unless you want to do change tracking in elegant way

[06:11:03.0000] <hsivonen>
is there an elegant way to de change tracking?

[06:12:00.0000] <karlcow>
Is there an elegant way? (Reductio ad absurdum)

[06:13:00.0000] <JirkaK>
hsivonen: A lot of software companies are authoring their help systems in DITA. In past years I have seen many companies to migrate from "legacy" systems like FrameMaker to DITA. There was a lot of marketing behind DITA in past years, so it got more traction in industry then DocBook recently.

[06:14:00.0000] <JirkaK>
probably most elegant way of change tracking is to use plain text and some versioning system ;-)

[06:15:00.0000] <hsivonen>
JirkaK: interesting. I was unaware of that kind of usage.

[06:15:01.0000] <hsivonen>
(help system that is. not plain text)

[06:16:00.0000] <JirkaK>
hsivonen: a lot of XML editors now provide out-of-the box configuration from DITA so it is fairly easy to author in DITA

[06:37:00.0000] <david_carlisle>
hsivonen: and as you'll see in Norm W's recent posts dita mindset is pushing docbook developments to a certain extent

[06:39:00.0000] <david_carlisle>
Dashiva:  we use xml for essentially everything here (documentation, code generation, fun, ...)

[06:39:01.0000] <jgraham>
david_carlisle: You also use Fortran :p

[06:39:02.0000] <david_carlisle>
jgraham:  yes but of course we generate it from xml

[06:40:00.0000] <david_carlisle>
why can't the whole world be like this?

[06:40:01.0000] <hsivonen>
david_carlisle: I can't tell if you are joking now. are you?

[06:40:02.0000] <david_carlisle>
no

[06:40:03.0000] <jgraham>
My humor detector is oscillating wildly

[06:40:04.0000] <Ms2ger>
Generating Fortran from XML?

[06:40:05.0000] <jgraham>
I really couldn't tell at all

[06:40:06.0000] <david_carlisle>
yes (and the other way actually)

[06:41:00.0000] <david_carlisle>
we have xml saying what the interface should be and we generate fortran, c, c#, matlab, r, python, whatever from that

[06:41:01.0000] <JirkaK>
using XSLT I guess?

[06:42:00.0000] <david_carlisle>
what else? 9although some of the youngsters use python)

[06:43:00.0000] <Ms2ger>
You make me want to jump into a lake, but I'm afraid annevk5 wouldn't let me out :)

[06:45:00.0000] <david_carlisle>
looking at people transform documents using DOM scripting has a similar effect on me, it works but it's fragile and painful compared to using a language designed for the job

[06:48:00.0000] <hsivonen>
foolip: is it OK to put the email address you are using in the validator.nu bugzilla in the hg user field?

[06:48:01.0000] <foolip>
hsivonen, philip at foolip dot org, yes

[06:48:02.0000] <foolip>
this wasn't work, after all

[06:48:03.0000] <hsivonen>
foolip: thanks

[06:49:00.0000] <foolip>
does that mean I'll have an account on the mercurial server, or just that the commit says so?

[06:49:01.0000] <Ms2ger>
The latter

[06:49:02.0000] <foolip>
ok, like git then

[06:49:03.0000] <karlcow>
/me has the same feeling than david_carlisle… culture I guess.

[06:50:00.0000] <foolip>
it turns out there's a lot of friction when moving from git to hg, all the little tings are different :)

[06:50:01.0000] <foolip>
I imagine the reverse is equally true

[06:50:02.0000] <Ms2ger>
Git? What's that? :)

[06:51:00.0000] <foolip>
Ms2ger, https://twitter.com/#!/agnoster/status/44636629423497217

[06:54:00.0000] <jgraham>
Are those just difficult sounding words or would it actually mean something if I understood the sentence properly?

[06:54:01.0000] <foolip>
jgraham, it's technobabble :)

[06:54:02.0000] <Ms2ger>
A Hilbert space? Come on.

[06:55:00.0000] <jgraham>
Well pure mathematics is often surprising

[06:56:00.0000] <Ms2ger>
To physicists, you mean ;)

[06:56:01.0000] <jgraham>
Possibly :)

[06:58:00.0000] <Ms2ger>
Did people see http://tools.ietf.org/html/draft-freed-media-type-regs-00, btw?

[06:58:01.0000] <jgraham>
It did sound rather like some of the explainations of monads you hear from hardcore haskellers

[06:59:00.0000] <Philip`>
http://tartley.com/?p=1267 too

[07:15:00.0000] <jgraham>
I still want a proper code review system for hg

[07:15:01.0000] <Ms2ger>
Don't we all?

[07:16:00.0000] <jgraham>
Well I'm surprised Mozilla haven't written one already

[07:16:01.0000] <Ms2ger>
We use bugzilla

[07:16:02.0000] <jgraham>
I know

[07:17:00.0000] <jgraham>
I'm surprised Mozilla haven't written one already

[07:18:00.0000] <hsivonen>
jgraham: gecko developers do stuff the way it has been done since the CVS days and Labs, Webdev, etc. use github

[07:19:00.0000] <hsivonen>
these days, b.m.o has a link to something called "splinter reviews" but I've never used the newfangled thing

[07:21:00.0000] <hsivonen>
G+'s real names policy has interesting effects on people whose name doesn't use the Latin script

[07:22:00.0000] <hsivonen>
do you use the native script, the Romanization of your native name or the (if you have one) your English alias?

[07:22:01.0000] <jgraham>
hsivonen: One reason I am surprised is that I occasionally see people complain about you having difficulties retaining new contributers, and the review system is often cited as a reason

[07:22:02.0000] <jgraham>
In particular the need to find the right person to review

[07:23:00.0000] <hsivonen>
jgraham: yeah, Mozilla isn't good at fixing its problems with attracting and retaining contributors

[07:23:01.0000] <jgraham>
But that is a purely technical problem that you could solve with software

[07:24:00.0000] <hsivonen>
/me notices 村田真's "Bragging Rights" https://plus.google.com/u/0/110720658734820520378/about

[07:28:00.0000] <JirkaK>
/me Murata-san?

[07:29:00.0000] <hsivonen>
JirkaK: so it seems

[07:30:00.0000] <Dashiva>
Yes

[07:49:00.0000] <timeless>
jgraham: which is purely technical?

[07:56:00.0000] <jgraham>
timeless: Finding the right reviewer for a change

[07:56:01.0000] <jgraham>
In general making reviews low-pain

[07:56:02.0000] <jgraham>
Both for the reviewer and reviewee

[07:56:03.0000] <timeless>
yeah, i think i outlined how to do that about 8 years ago

[07:57:00.0000] <timeless>
(programmatically)

[07:57:01.0000] <timeless>
it was never implemented

[07:57:02.0000] <timeless>
there's actually more than enough data to figure it out

[07:59:00.0000] <smaug____>
I haven't seen any good way to do reviews

[07:59:01.0000] <smaug____>
s/way/software/

[07:59:02.0000] <timeless>
oh, i'm just talking about how to find the right reviewer

[08:00:00.0000] <timeless>
the software i've seen that tries to actually "enable" reviewing itself gives me headaches

[08:01:00.0000] <timeless>
http://www.usatoday.com/news/nation/2011-07-27-ny-thruway-prayers_n.htm

[08:01:01.0000] <smaug____>
/me needs to close some of his 200+ tabs :/

[08:02:00.0000] <timeless>
smaug____: heh

[08:02:01.0000] <timeless>
/me ponders

[08:02:02.0000] <timeless>
i have no idea how many tabs i have

[08:02:03.0000] <timeless>
i'm using panorama which makes counting tabs *hard*

[08:02:04.0000] <jgraham>
smaug____: The system we have is pretty good. I think Gerrit is a bit similar

[08:03:00.0000] <jgraham>
/me discovers hg-review

[08:03:01.0000] <jgraham>
This looks like it might not obviously be awful

[08:03:02.0000] <timeless>
oh yeah, 've seen that mentioned

[08:03:03.0000] <zewt>
quite a standard to set for software

[08:03:04.0000] <timeless>
i've never tried it

[08:03:05.0000] <timeless>
/me should

[08:04:00.0000] <timeless>
there's actually a hg repo called `all-extensions` iirc

[08:04:01.0000] <timeless>
/me needs to grab it

[08:04:02.0000] <timeless>
https://bitbucket.org/abuehl/allextensions

[08:07:00.0000] <hsivonen>
sadly, reviews don't really scale

[08:08:00.0000] <timeless>
?

[08:08:01.0000] <hsivonen>
you can hire more people to write more code but you can't hire more reviewers

[08:08:02.0000] <timeless>
well

[08:08:03.0000] <timeless>
you could hire the people who used to code for you

[08:08:04.0000] <timeless>
moco chose not to do that

[08:09:00.0000] <hsivonen>
timeless: :-(

[08:09:01.0000] <zewt>
... and you want people experienced with your code to review, but people often aren't going to view being moved from development to reviewing as a promotion, heh

[08:09:02.0000] <smaug____>
browser vendors should hire some more spec reviewers ;)

[08:23:00.0000] <timeless>
smaug____: that happened to me 2 months ago :)

[08:23:01.0000] <timeless>
but yeah

[08:23:02.0000] <timeless>
/me should provide a place where people could suggest specs to review

[08:23:03.0000] <timeless>
hrm, if people want, please feel free to comment on my G+ page with suggestions :)

[08:24:00.0000] <zewt>
too bad g+ is broken : |

[08:24:01.0000] <smaug____>
timeless: specs to review?

[08:25:00.0000] <smaug____>
like all of them?

[08:25:01.0000] <zewt>
any new google product that doesn't work with gapps is just poor and broken

[08:25:02.0000] <timeless>
smaug____: a prioritized list with links is appreciated

[08:25:03.0000] <smaug____>
yeah

[08:25:04.0000] <timeless>
gathering a list of specs is actually moderately painful

[08:25:05.0000] <timeless>
i've tried a couple of times

[08:25:06.0000] <timeless>
i keep giving up

[08:26:00.0000] <timeless>
there are probably hundreds of xml specs that i don't want to look at

[08:26:01.0000] <timeless>
as an example of things to disprove your "all of them"

[08:26:02.0000] <smaug____>
well, sure. All the relevant ones...

[08:27:00.0000] <timeless>
right... see "please give me a prioritized list with links"

[08:27:01.0000] <timeless>
seriously..

[08:27:02.0000] <timeless>
technically bugzilla does a better job of this than standards bodies

[08:27:03.0000] <timeless>
i can go to bugzilla.mozilla.org/request.cgi and get a list

[08:28:00.0000] <timeless>
it isn't perfect

[08:28:01.0000] <timeless>
e.g. it'd be nice if i could sort by deadline

[08:28:02.0000] <timeless>
(which is something bmo doesn't have anyway, but which is definitely relevant for standards work)

[08:29:00.0000] <smaug____>
well, what does deadline mean in whatwg?

[08:30:00.0000] <smaug____>
(but sure, with w3c specs it would be useful)

[08:31:00.0000] <jgraham>
Thoughts on hg-review so far: Who the fuck thought that black on green was a nice colour scheme for displaying code?

[08:32:00.0000] <jgraham>
I guess it's because these were lines added

[08:32:01.0000] <zewt>
someone who uses a screen reader, i'd imagine

[08:32:02.0000] <jgraham>
But seriously it's ugly

[08:33:00.0000] <jgraham>
zewt: That would also explain the fact that it users overflow:hidden at (presumably) 80 columns

[08:33:01.0000] <zewt>
... that would make me stop looking and move on, alone

[08:33:02.0000] <jgraham>
Well these are fixable issues

[08:34:00.0000] <zewt>
it speaks of a complete lack of experience with real code

[08:34:01.0000] <zewt>
80 columns is used by two people: those who havn't changed their coding habits since the 90s, and people who religiously follow PEP-8 (and their overlap)

[08:34:02.0000] <jgraham>
If I can work out how to actually share review comments I might suggest using it

[08:34:03.0000] <zewt>
two classes of people, rather

[08:34:04.0000] <jgraham>
/me likes code that wraps at a sensibly small number of columns

[08:35:00.0000] <zewt>
me too, such as 120 :)

[08:35:01.0000] <jgraham>
It means I can have multiple side-by-side windows

[08:35:02.0000] <jgraham>
I find that people typically use 80 or "however wide my monitor is"

[08:36:00.0000] <zewt>
my terminal windows are around 240x70

[08:37:00.0000] <zewt>
so having two windows would still make 120 columns sane; and even 100 columns is a lot less annoying than 80, which just causes too much line splitting

[08:41:00.0000] <timeless>
/me has at least 4 side by side windows

[08:41:01.0000] <timeless>
smaug____: anyway...

[08:42:00.0000] <timeless>
to prioritize for whatwg, i'd probably ask for things which people think are <small> or "urgent" or which vendors seem likely to try to implement "sooner" than others

[08:42:01.0000] <jgraham>
Oh, I found a scrollbar

[08:42:02.0000] <jgraham>
But still

[09:27:00.0000] <Ms2ger>
timeless, not sure if you saw platform.html5.org? It's not really what you need, but it's about the best list of web-relevant specs I've seen

[09:37:00.0000] <annevk>
didn't know foolip was patching Validator.nu, neat

[09:39:00.0000] <foolip>
annevk, I had nothing to do this weekend :)

[09:40:00.0000] <Ms2ger>
Could've written a spec :)

[09:41:00.0000] <timeless>
Ms2ger: yeah, i've seen it

[09:41:01.0000] <timeless>
i'm pretty sure it's unusable for my purposes

[09:41:02.0000] <timeless>
what i need is a list of things which i should be reviewing

[09:42:00.0000] <timeless>
things which are large and have lots of eyeballs and aren't moving don't count

[09:42:01.0000] <timeless>
(SVG comes to mind)

[09:42:02.0000] <hsivonen>
krijnh: the logs have oddities on /me lines

[09:42:03.0000] <dglazkov>
good morning, Whatwg!

[09:43:00.0000] <hsivonen>
dglazkov: good morning

[09:43:01.0000] <timeless>
/me looks for a definition of `cancelable`

[09:45:00.0000] <foolip>
it seems to me that writing specs on algorithm form Hixie-style is a lot like writing code without test compiling it, and then waiting for a very long time before you get the warnings/errors

[09:45:01.0000] <smaug____>
indeed

[09:46:00.0000] <smaug____>
though, what would be a better way to write specs?

[09:47:00.0000] <timeless>
compiling and testing? :)

[09:48:00.0000] <Ms2ger>
Then feel free to review DOM Core :)

[09:48:01.0000] <timeless>
heh

[09:48:02.0000] <timeless>
i will at some point

[09:49:00.0000] <timeless>
but i'd like to have a priority ordered list

[09:49:01.0000] <timeless>
if there are things i should look at first, i'd rather do that...

[09:49:02.0000] <annevk>
Ms2ger, btw, I'm pretty close to have xxxhierarchy sorted out

[09:49:03.0000] <timeless>
/me is currently looking at deviceorientation

[09:49:04.0000] <Ms2ger>
Yay!

[09:49:05.0000] <annevk>
Ms2ger, in a style that does not make use of the nodes model

[09:50:00.0000] <annevk>
making it much more explicit what actually needs to be done which seems somewhat better

[09:54:00.0000] <timeless>
Ms2ger: wrt seeing that host, it's in my w3 panorama stack (#5/7)

[10:22:00.0000] <AryehGregor>
Sigh, specdrama.

[10:23:00.0000] <AryehGregor>
hsivonen, I assume that if I wanted to use Google time to submit the spec to the W3C or whatever, Hixie would probably let me, but he didn't tell me to and I'm not interested in doing it unless I'm told to.

[10:57:00.0000] <annevk>
AryehGregor, it's the same people that always complain, regardless of what you do

[10:58:00.0000] <annevk>
AryehGregor, if Hixie made the HTML specification forty pages longer they would complain too

[10:58:01.0000] <AryehGregor>
annevk, I thought there have been a couple of very productive Google+ discussions lately.

[10:58:02.0000] <AryehGregor>
I'm trying to continue the trend.  :)

[10:58:03.0000] <AryehGregor>
(but yeah)

[10:58:04.0000] <AryehGregor>
The only that would have made them happy would be to totally ignore the W3C and work only in the WHATWG, so they didn't know about it.

[11:01:00.0000] <annevk>
they would know about it

[11:01:01.0000] <AryehGregor>
They didn't know about it until I just announced it.

[11:01:02.0000] <AryehGregor>
Even though I announced it on the WHATWG list months ago.

[11:02:00.0000] <annevk>
Julian commented on the WHATWG list I thought

[11:03:00.0000] <AryehGregor>
You mean yesterday, or in general?

[11:03:01.0000] <annevk>
yesterday

[11:03:02.0000] <AryehGregor>
Yesterday he commented on public-html.

[11:03:03.0000] <AryehGregor>
On, of course, a purely procedural issue.

[11:04:00.0000] <annevk>
oh you're right

[11:28:00.0000] <AryehGregor>
Megapost written: https://plus.google.com/u/0/105458233028934590147/posts/h7nsT7wuNmX

[11:32:00.0000] <Ms2ger>
/me is perfectly happy to stay out of that discussion

[11:35:00.0000] <annevk>
I think you make some pretty good points in there

[11:35:01.0000] <annevk>
Both about power of editors and process issues with the W3C

[11:40:00.0000] <jgraham>
AryehGregor: Nice commen

[11:40:01.0000] <jgraham>
t

[12:08:00.0000] <hober>
AryehGregor: I'm sad that I'm limited to +1ing that once.

[12:11:00.0000] <annevk>
AryehGregor, you mind if I quote that verbatim on my blog with context?

[12:11:01.0000] <AryehGregor>
annevk, go ahead.

[12:11:02.0000] <AryehGregor>
I'm optimistic that we can win over some of the more reasonable people to a large degree, given a chance.

[12:12:00.0000] <AryehGregor>
Not the W3C itself, though.

[12:14:00.0000] <AryehGregor>
Maybe someone should also copy it to the discussion on the chairs list.  The answer in the end is obviously going to have to be "we're not going to ever change things enough to satisfy him", but I imagine some of the participants there would find it informative.

[12:17:00.0000] <tantek>
/me tries to catch-up on the drama.

[12:17:01.0000] <AryehGregor>
Hixie, I'll respond to some of the DOM Range stuff, at least, since there's not much of it.

[12:19:00.0000] <AryehGregor>
Maybe some of the editing stuff too.

[12:20:00.0000] <annevk>
AryehGregor, was your editing work going to take in Selection in due course?

[12:20:01.0000] <AryehGregor>
annevk, yeah, that's the theory.

[12:20:02.0000] <AryehGregor>
I don't want to do it right now because I just asked for implementer review and don't want to add a massive chunk of spec text just now.

[12:20:03.0000] <AryehGregor>
I'll work on other things for the moment.

[12:23:00.0000] <annevk>
okay

[12:23:01.0000] <AryehGregor>
Unless you want to absorb DOM Range into DOM Core and this is blocking your work.

[12:23:02.0000] <tantek>
AryehGregor - what's the spec you're developing?

[12:23:03.0000] <AryehGregor>
tantek, http://aryeh.name/spec/editing/editing.html

[12:24:00.0000] <tantek>
AryehGregor - seems reasonable to me.

[12:25:00.0000] <timeless>
AryehGregor: so...

[12:25:01.0000] <tantek>
/me is confused by why there is any drama about this at all.

[12:25:02.0000] <timeless>
i'd like to note that there are too many mailing lists

[12:25:03.0000] <timeless>
and no useful announce lists

[12:25:04.0000] <timeless>
if you could have posted to one useful announce list

[12:25:05.0000] <tantek>
timeless - you can shorten that to: there are too many mailing lists

[12:25:06.0000] <tantek>
:)

[12:25:07.0000] <timeless>
and if i could have been subscribed to that one list

[12:25:08.0000] <timeless>
tantek: +1

[12:25:09.0000] <timeless>
i'd probably be reviewing it now

[12:25:10.0000] <tantek>
timeless - I think that's called "choose your feeds wisely" :)

[12:25:11.0000] <timeless>
tantek: i haven't found any useful feeds!

[12:26:00.0000] <AryehGregor>
I could make a post on the WHATWG blog, if that would be helpful.

[12:26:01.0000] <timeless>
i've been soliciting for them all day

[12:26:02.0000] <tantek>
/me commends AryehGregor for his tireless hard work and working in the open with a liberal license. Keep up the good work. And don't waste time on the haters.

[12:26:03.0000] <timeless>
if you did it and got others to do the same, that'd encourage me to read that blog

[12:26:04.0000] <timeless>
so, +1 :)

[12:28:00.0000] <erlehmann>
a bit off-topic: is a „liberal“ a „free“ license (e.g. not proprietary) ? or does it mean MIT/X1-style (not copyleft) ?

[12:28:01.0000] <timeless>
s/X1/X11/ ? :)

[12:29:00.0000] <timeless>
i think it leans to MIT/..

[12:32:00.0000] <timeless>
/me will +1 the actual plus comment later from home..

[12:33:00.0000] <erlehmann>
he, i use „add“ for „+1“ to confuzzle my not-yet-deleted friends.

[12:33:01.0000] <erlehmann>
„i added that“

[12:33:02.0000] <TabAtkins_>
"incremented"

[12:34:00.0000] <Ms2ger>
"inc'd"

[12:36:00.0000] <AryehGregor>
erlehmann, specs should be permissively licensed, like CC0 or MIT-style.  Copyleft for specs is a bad idea, IMO.

[12:36:01.0000] <AryehGregor>
It means projects can't, e.g., include parts of it in their documentation unless their license happens to be compatible.

[12:36:02.0000] <timeless>
right

[12:36:03.0000] <AryehGregor>
Proprietary forks are not a big problem in practice for specs anyway.

[12:36:04.0000] <timeless>
similarly, examples shouldn't be copyleft

[12:37:00.0000] <annevk>
draft: http://annevankesteren.nl/2011/07/editor

[12:37:01.0000] <timeless>
heck, header files generally shouldn't be copy left

[12:37:02.0000] <annevk>
not sure about the title

[12:37:03.0000] <timeless>
(ignoring the minor detail that they generallly don't have sufficient substance to have anything for which one can claim copyright!)

[12:38:00.0000] <timeless>
annevk: is it unreasonable for me to wish your <editing APIs of the web platform> was a link?

[12:38:01.0000] <erlehmann>
AryehGregor, i fully agree with that. differentiation through implementations makes more sense.

[12:38:02.0000] <annevk>
timeless, no

[12:38:03.0000] <timeless>
:)

[12:39:00.0000] <timeless>
annevk: might as well try to get eyeballs for something useful after all :)

[12:40:00.0000] <erlehmann>
timeless, why that stuff with the header files?

[12:41:00.0000] <timeless>
erlehmann: if i want to be compatible with an api

[12:41:01.0000] <timeless>
why should the api be able to influence the license on *my* product?

[12:41:02.0000] <erlehmann>
isn't there at least one program that is free because of readline being GPL?

[12:41:03.0000] <timeless>
generally speaking commercial products get this right and ensure their licenses on headers are non tainting

[12:41:04.0000] <Ms2ger>
Obviously you lack a cite attribute

[12:41:05.0000] <timeless>
i'm sure there are a dozen crappy programs ou tthere

[12:42:00.0000] <timeless>
s/u t/ut /

[12:42:01.0000] <timeless>
but personally i don't believe that's a valid reason for anything

[12:42:02.0000] <timeless>
and besides, readline is a library, not just a header

[12:42:03.0000] <erlehmann>
so you are more talking about stuff susceptible to network effects.

[12:42:04.0000] <timeless>
if you want to require someone to have <foo-license> because they include <foo-library>

[12:42:05.0000] <timeless>
that's ok

[12:43:00.0000] <timeless>
but just for a header file?

[12:43:01.0000] <timeless>
that's not ok in my book

[12:43:02.0000] <annevk>
timeless, added links

[12:43:03.0000] <timeless>
thanks

[12:43:04.0000] <annevk>
Ms2ger, I could link to the thread that way I suppose

[12:43:05.0000] <erlehmann>
well, i think its simply a matter of strategy. if you make it too restrictive, few people will use it.

[12:44:00.0000] <timeless>
/me shrugs

[12:45:00.0000] <erlehmann>
oh well.

[12:46:00.0000] <Ms2ger>
Hmm

[12:46:01.0000] <timeless>
you object?

[12:46:02.0000] <Ms2ger>
Not at all

[12:46:03.0000] <timeless>
do you want me to get -emeritus added?

[12:46:04.0000] <Ms2ger>
I wish it was still true :)

[12:46:05.0000] <timeless>
it's better than a /maemo or /meego hostmask :)

[12:47:00.0000] <timeless>
that'd be really problematic :)

[12:47:01.0000] <timeless>
but yeah, we're a webkit shop :)

[12:47:02.0000] <Ms2ger>
Boo :)

[12:47:03.0000] <timeless>
amusingly the tools team internally has firefox addons

[12:47:04.0000] <timeless>
and they also fix compatibility issues w/ nightly :)

[12:47:05.0000] <timeless>
yay tools teams!

[13:01:00.0000] <tantek>
erlehmann, timeless, yes, liberal as in MIT or CC0 (preferred).

[13:02:00.0000] <tantek>
AryehGregor, please consider also saying the IP in the spec is available under OWFa 1.0 as well (handles patent stuff)

[13:02:01.0000] <AryehGregor>
What does it do about patents?

[13:02:02.0000] <AryehGregor>
I don't own any patents, so I can't see that I could do much.

[13:02:03.0000] <tantek>
it basically is the way to do a patent non-assert

[13:02:04.0000] <tantek>
in an open / portable way

[13:03:00.0000] <tantek>
the same way that CC is the way to do a copyright non-assert

[13:03:01.0000] <tantek>
in an open / portable way

[13:03:02.0000] <AryehGregor>
I have no patents to non-assert, though.

[13:03:03.0000] <tantek>
sure, but people often say "but w3c gets you patent /ip protection etc."

[13:03:04.0000] <tantek>
and using OWFa allows you to do it independently

[13:03:05.0000] <wilhelm>
Your employer may have.

[13:03:06.0000] <tantek>
http://www.openwebfoundation.org/legal/the-owf-1-0-agreements/owfa-1-0

[13:05:00.0000] <tantek>
more on the topic of doing open standards development in general: http://tantek.com/2011/168/b1/practices-good-open-web-standards-development

[13:05:01.0000] <gsnedders>
AryehGregor: In a country with submarine patents, why should I believe you?

[13:05:02.0000] <tantek>
(probably more than anyone here needs, and most of that will just be duh/assumed material for the folks here)

[13:06:00.0000] <tantek>
(hopefully useful as a reference though)

[13:06:01.0000] <AryehGregor>
gsnedders, because if you suspect me of secretly holding patents, you may as well suspect everyone in the world of it, give up on software development, and take up some safe occupation like coal mining instead.

[13:06:02.0000] <AryehGregor>
People who worry about patents are not going to worry about me.

[13:06:03.0000] <AryehGregor>
Also, patent filings are public from day one, I think.

[13:06:04.0000] <AryehGregor>
And submarine patents as such don't exist anymore, at least as the term was formerly used.

[13:09:00.0000] <gsnedders>
AryehGregor: My understanding (which is limited, because, well, I'm not in the US for a start) is that the last submarine patents are still not all published. Regardless of whether you hold patents or not, people have good reason to prefer a promise you won't sue them for violating a patent (which you know they won't, because you don't hold any).

[13:09:01.0000] <AryehGregor>
I think submarine patents only exist if they were filed before, like, 1996.

[13:09:02.0000] <AryehGregor>
When I was eight.

[13:09:03.0000] <gsnedders>
AryehGregor: It's about insurance that the editor of the spec hasn't deliberately put something in the spec so they can sue for it.

[13:10:00.0000] <AryehGregor>
Okay, that's all very well, but I don't think anyone is going to be credibly worried about it.

[13:10:01.0000] <AryehGregor>
It's not high on my to-do list.

[13:10:02.0000] <gsnedders>
AryehGregor: It's about the risk. Sure, I could take you at your word, but then someday someone will lie.

[13:11:00.0000] <AryehGregor>
Find me someone who's actually worried about my secret patent portfolio and I'll consider spending time on the issue.

[13:12:00.0000] <gsnedders>
AryehGregor: Waste as much time on it as you did the text of the spec?

[13:12:01.0000] <AryehGregor>
I'm doing other things, this is a distraction based on a theoretical issue.

[13:16:00.0000] <Hixie>
hey where's innnerText supposed to live?

[13:16:01.0000] <Ms2ger>
In hell

[13:17:00.0000] <Hixie>
dom core? dom parsing? html? editing?

[13:17:01.0000] <Ms2ger>
Not in a spec I edit

[13:17:02.0000] <foolip>
surely not dom core?

[13:17:03.0000] <Ms2ger>
So one of the latter

[13:17:04.0000] <Ms2ger>
It's not clear anybody's interested in a spec, though

[13:17:05.0000] <gsnedders>
HTML, I'd guess

[13:18:00.0000] <annevk>
depends on the details

[13:18:01.0000] <Hixie>
k i'll just continue to ignore this problem for now

[13:18:02.0000] <erlehmann>
academics that deal with patents are interesting. i have no idea how the two cultures i like to think represented by cerf vs. brandenburg (the mp3 guy) are intermingled or opposed to each other.

[13:19:00.0000] <Hixie>
annevk: http://platform.html5.org/ needs updating for dom traversal, right?

[13:19:01.0000] <Hixie>
i wish http://platform.html5.org/ didn't list each section from WA1 separately

[13:19:02.0000] <Ms2ger>
Yes

[13:19:03.0000] <annevk>
yes

[13:20:00.0000] <annevk>
I haven't reached Mike in a while

[13:20:01.0000] <Hixie>
(and i wish it had css more expanded)

[13:21:00.0000] <jgraham>
foolip: Well it sort-of is DOM core if it applies to all namespaces. And if it doesn't apply to all namespaces, that's pretty weird

[13:21:01.0000] <jgraham>
OTOH I widh it would die :)

[13:21:02.0000] <Ms2ger>
Anybody can define stuff on Element

[13:21:03.0000] <foolip>
jgraham, does it exist on SVG element nodes?

[13:21:04.0000] <annevk>
Hixie, some of the feedback you gave me was about way different things

[13:22:00.0000] <Ms2ger>
That doesn't mean it needs to be in Core

[13:22:01.0000] <foolip>
oh, I'm stupid, what else is new

[13:22:02.0000] <jgraham>
Ms2ger: Yes but having gratuitously split interfaces isn't helpful

[13:22:03.0000] <annevk>
Hixie, like disabled attribute for <iframe>

[13:22:04.0000] <foolip>
I thought it was innerHTML :-/

[13:22:05.0000] <jgraham>
*wish

[13:24:00.0000] <Hixie>
annevk: the use case was dragging something, the solution is mouse capture

[13:24:01.0000] <Hixie>
annevk: the iframe bit is a distraction

[13:24:02.0000] <Hixie>
i'll move that thread to a separate bucket on my end though

[13:24:03.0000] <annevk>
the keyboard stuff is also not for DOM Core

[13:25:00.0000] <annevk>
btw, the SVG WG wants to follow the on* pattern for event handlers

[13:25:01.0000] <Hixie>
one e-mail from hallvord?

[13:25:02.0000] <annevk>
several emails on keyboard stuff

[13:25:03.0000] <annevk>
and dblclick

[13:26:00.0000] <Hixie>
k

[13:27:00.0000] <Hixie>
ok after taking out event-specific stuff, i have 16 e-mails left

[13:28:00.0000] <Hixie>
want me to send you that bucket again so you can make sure i got the right ones?

[13:28:01.0000] <annevk>
cancelling keydown is one of those

[13:28:02.0000] <annevk>
sure

[13:28:03.0000] <annevk>
there was some relevant event stuff about Apple-specific extensions to the Event interface

[13:28:04.0000] <annevk>
(though I doubt that will result in spec changes)

[13:29:00.0000] <zewt>
i'm surprised more sites aren't really really evil about cancelling keydown

[13:29:01.0000] <Hixie>
annevk: if there's nothing in what i just sent you that needs any response and it's all been taken into account already, i can just delete the bucket on my end

[13:32:00.0000] <annevk>
Node having an item() method looks like something that might need updating in DOM Core

[13:32:01.0000] <annevk>
"Node item(in long index); // and Node is also indexable as array"

[13:32:02.0000] <Ms2ger>
Who does that?

[13:32:03.0000] <annevk>
this was an email from Maciej

[13:34:00.0000] <annevk>
Safari also has a bunch of stuff on Event apparently

[13:34:01.0000] <annevk>
srcElement - clone of the nonstandard IE extension srcElement, an alias for target

[13:34:02.0000] <annevk>
returnValue - clone of the nonstandard IE extension returnValue, this is a read-write property that lets you get and set whether default is prevented. This lets you un-prevent default, something not allowed in the standard DOM event model.

[13:34:03.0000] <annevk>
cancelBubble - similar to returnValue, but for stopPropagation

[13:34:04.0000] <annevk>
dataTransfer - this appears to be applicable only to drag & drop events, probably a mistake that it is in the Event interface

[13:34:05.0000] <annevk>
clipboardData - similarly, but for clipboard events

[13:35:00.0000] <annevk>
Hixie, you can delete that bucket, I will file a bug on Node.item() and looking at Event extensions

[13:35:01.0000] <Hixie>
cool

[13:35:02.0000] <Hixie>
thanks dude

[13:38:00.0000] <annevk>
hmm I cannot find item in document

[13:42:00.0000] <Hixie>
ok twice now i have explained to plh how the w3c process is broken, he claims it is not, i then respond in detail explaining how it is, and he goes silent

[13:42:01.0000] <Hixie>
http://lists.w3.org/Archives/Public/public-web-perf/2011Jul/0090.html and http://lists.w3.org/Archives/Public/public-web-perf/2011Apr/0013.html

[13:43:00.0000] <erlehmann>
well, egoless discussions are rare. and you are like, the W3Cs nemesis personified? ;)

[13:43:01.0000] <zewt>
at least he didn't say "let's agree to disagree" as if that settles everything

[13:43:02.0000] <AryehGregor>
plh just added me on Google+ after my big post about how I'm not interested in working with the W3C.

[13:43:03.0000] <zewt>
that's pretty much the worst internet-disagreement expression there is

[13:43:04.0000] <Hixie>
i just wish he would either explain why i'm wrong, or stop claiming things that i've shown are wrong

[13:43:05.0000] <erlehmann>
“i think both parties have made their points today. now for something completely different”

[13:45:00.0000] <erlehmann>
such people exist. i regard overly broad criticism that is not told as backstory to narrow issues as trolling anyway.

[13:53:00.0000] <erlehmann>
annevk, not wanting to troll you, but why don't you link directly to the fragment of the comment on the page?

[13:53:01.0000] <erlehmann>
(i know people hate scrolling)

[13:55:00.0000] <Hixie>
AryehGregor: fwiw, the editing stuff looks fantastic.

[13:56:00.0000] <AryehGregor>
Hixie, thanks.

[13:57:00.0000] <timeless>
AryehGregor: i think the current problem with patents is that you can revise claims after filing..

[13:57:01.0000] <AryehGregor>
timeless, I think that in most cases, the current problem with patents is that they exist.

[13:57:02.0000] <Hixie>
so this crypto thread on whatwg -- anyone know what spec it's supposed to affect?

[13:57:03.0000] <AryehGregor>
But that's a futile discussion.

[13:57:04.0000] <AryehGregor>
Hixie, about the new API?  I assumed people were making a new spec.

[13:58:00.0000] <Hixie>
do you know who specifically?

[13:58:01.0000] <AryehGregor>
No clue, haven't been following closely.

[13:59:00.0000] <timeless>
annevk: mike was around on monday iirc

[14:01:00.0000] <timeless>
zewt: ePost.ca seems pretty evil about form input

[14:01:01.0000] <timeless>
i hit a couple of other .ca sites which were pretty evil

[14:02:00.0000] <zewt>
that's one place it's easy to do really stupid things, but not necessarily willfully

[14:02:01.0000] <zewt>
like "preventDefault if the key isn't [0-9]"

[14:02:02.0000] <zewt>
but evil i'm talking about things like preventing escape for browser stop, alt-left/right, f5, etc

[14:02:03.0000] <Ms2ger>
Hixie, ddahl is writing up something he calls a spec :)

[14:02:04.0000] <zewt>
eg. deliberately being obnoxious to users

[14:03:00.0000] <Hixie>
Ms2ger: for crypto?

[14:03:01.0000] <Hixie>
Ms2ger: k

[14:03:02.0000] <timeless>
Hixie: i presume it's the mozilla hosted crypto spec

[14:03:03.0000] <Hixie>
do you have a link?

[14:03:04.0000] <zewt>
there was some proposal a while back but it was sort of a mess and i didn't look very closely

[14:03:05.0000] <Ms2ger>
Somewhere on wikimo, I'll look

[14:03:06.0000] <timeless>
zewt: oh, i think i've hit some of those

[14:03:07.0000] <timeless>
Hixie: yeah, i think

[14:04:00.0000] <timeless>
grumble, it isn't googleable

[14:04:01.0000] <timeless>
lemme ask gmail

[14:04:02.0000] <Ms2ger>
https://wiki.mozilla.org/Privacy/Features/DOMCryptAPISpec/Latest

[14:04:03.0000] <zewt>
the other day i hit a "wait 30 seconds to download" site that stopped the countdown when the window loses focus, to try to force you to sit there and wait :(

[14:04:04.0000] <zewt>
another one of those really gross evil obnoxious things that I'm surprised I hadn't seen before

[14:05:00.0000] <timeless>
oh right, it doesn't have 'o' after 'Crypt'

[14:05:01.0000] <timeless>
it intentionally breaks any google searches i might try

[14:05:02.0000] <timeless>
zewt: those will be *more* common

[14:05:03.0000] <Ms2ger>
timeless, what's the bugzilla component to add new repos to mxr?

[14:05:04.0000] <timeless>
webtools/mxr i think

[14:06:00.0000] <zewt>
those have been possible for a long long time, and those download sites have always been obnoxious, so it's puzzled me that it took so long

[14:06:01.0000] <timeless>
zewt: http://www.w3.org/TR/2011/WD-page-visibility-20110721/

[14:06:02.0000] <Ms2ger>
OK, I thought that was about issues in MXR

[14:06:03.0000] <timeless>
Ms2ger: i don't think we have a second place

[14:06:04.0000] <timeless>
you can use server operations

[14:06:05.0000] <zewt>
it's the sort of thing that just needs browser overrides, really

[14:07:00.0000] <zewt>
or, for my purposes, at least enough functionality in things like GM to kill it

[14:07:01.0000] <Hixie>
looks like ddhal and abarth are coauthoring it

[14:08:00.0000] <Hixie>
abarth: what should i do with crypto-related feedback sent to whatwg?

[14:10:00.0000] <timeless>
Ms2ger: technically you couldmisfile in websites/Other

[14:11:00.0000] <Ms2ger>
I (mis)filed in ServerOps

[14:11:01.0000] <timeless>
/me can't *find* serverops

[14:12:00.0000] <Ms2ger>
mozilla.org

[14:14:00.0000] <Ms2ger>
Huh, Google tells me Hixie +1d dev.w3.org/2006/webapi/FileAPI/

[14:15:00.0000] <Ms2ger>
And that page is completely broken

[14:15:01.0000] <Hixie>
wow, what happened to that page

[14:15:02.0000] <Ms2ger>
It starts with

[14:15:03.0000] <Ms2ger>
  Overview.html

[14:15:04.0000] <Ms2ger>
  Web IDL

[14:15:05.0000] <Ms2ger>
/me looks for heycam

[14:17:00.0000] <Hixie>
wtf is that page

[14:18:00.0000] <Ms2ger>
A botched attempt at using XSLT

[14:18:01.0000] <Hixie>
sicking: File API spec got broken dude

[14:20:00.0000] <Ms2ger>
Probably because Arun's last edit introduced a wellformedness error

[14:20:01.0000] <Ms2ger>
"This note is normative."

[14:21:00.0000] <Ms2ger>
What?

[14:21:01.0000] <erlehmann>
haha

[14:32:00.0000] <annevk>
:)

[14:32:01.0000] <annevk>
erlehmann, give me the fragment

[14:33:00.0000] <erlehmann>
annevk, i believe you are implying g+ is somehow too dense for that. point made.

[14:34:00.0000] <annevk>
I mean I couldn't get to the fragment

[14:36:00.0000] <annevk>
Ms2ger, so I cannot find Node.item() anymore

[14:38:00.0000] <sicking>
Hixie: ugh :( Poke Arun

[14:38:01.0000] <annevk>
that's why people should use Anolis to make specs

[14:38:02.0000] <Ms2ger>
Me neither

[14:38:03.0000] <annevk>
good

[14:43:00.0000] <annevk>
http://trac.webkit.org/browser/trunk/Source/WebCore/dom/Event.idl does seem to have quite some extensions

[14:44:00.0000] <AryehGregor>
Hixie, okay, I responded to all of the stuff you sent me where I could give a useful answer, and that was sent to whatwg (not other places).

[14:44:01.0000] <AryehGregor>
And where I hadn't already responded months ago.

[14:45:00.0000] <AryehGregor>
Some of those threads are things I started.

[14:45:01.0000] <annevk>
http://mxr.mozilla.org/mozilla-central/source/dom/interfaces/events/nsIDOMEvent.idl seems to match the specification

[14:45:02.0000] <annevk>
but Mozilla sometimes additional stuff elsewhere

[14:46:00.0000] <AryehGregor>
Okay, can I post stuff to the WHATWG blog somehow?  I can't remember if I have an account.

[14:46:01.0000] <AryehGregor>
/me doesn't see anywhere to try logging in

[14:47:00.0000] <Ms2ger>
annevk, try nsIDOMNSEvent

[14:47:01.0000] <Ms2ger>
(I'm removing it, but haven't got around to it yet)

[14:50:00.0000] <annevk>
originalTarget

[14:50:01.0000] <annevk>
explicitOriginalTarget

[14:50:02.0000] <annevk>
and a similar list of constants as Safari

[14:50:03.0000] <annevk>
bah

[15:01:00.0000] <timeless>
AryehGregor: ...

[15:01:01.0000] <timeless>
about nested lists

[15:01:02.0000] <AryehGregor>
What about them?

[15:01:03.0000] <timeless>
have you seen http://support.microsoft.com/kb/2489283 ?

[15:01:04.0000] <timeless>
(random kb article)

[15:01:05.0000] <AryehGregor>
What about it?

[15:01:06.0000] <timeless>
just that they tend to have <ul>'s in <li>'s

[15:02:00.0000] <AryehGregor>
It nests <ul> inside <li>.  That's not a problem.

[15:02:01.0000] <AryehGregor>
I never said that should be invalid.

[15:02:02.0000] <timeless>
/me considers trying to reread the message

[15:02:03.0000] <AryehGregor>
But I don't think execCommand() should produce it, under normal circumstances.

[15:02:04.0000] <AryehGregor>
For execCommand(), it's simpler to only have lists nested directly inside lists.

[15:02:05.0000] <timeless>
/me ponders

[15:02:06.0000] <timeless>
so...

[15:02:07.0000] <AryehGregor>
It doesn't have to let you write arbitrary HTML.  If you want to do something fancier, write the HTML yourself.

[15:03:00.0000] <annevk>
emailed www-dom about extensions to Event

[15:03:01.0000] <AryehGregor>
FWIW, I wasn't able to get Word or OpenOffice to produce lists like that one.

[15:03:02.0000] <timeless>
the problem is that i think the kb article is a pretty natural thing to want to write

[15:03:03.0000] <timeless>
really?

[15:03:04.0000] <AryehGregor>
In a WYSIWYG editor?

[15:03:05.0000] <AryehGregor>
I mean, you can get something that looks visually something like that.

[15:04:00.0000] <AryehGregor>
You can do that with my execCommand() spec too.

[15:04:01.0000] <AryehGregor>
As long as you only put the nested list at the end of the <li>.

[15:04:02.0000] <timeless>
but you're saying you couldn't get them to generate content that's logically like that?

[15:05:00.0000] <AryehGregor>
I'm saying we should treat </li><ol>...</ol> logically the same as <ol>...</ol></li>.

[15:05:01.0000] <AryehGregor>
We can make up whatever semantics are convenient.

[15:05:02.0000] <annevk>
Ms2ger, please clean up Gecko http://lists.w3.org/Archives/Public/www-dom/2011JulSep/0039.html

[15:05:03.0000] <AryehGregor>
annevk, how can I post something on the WHATWG blog?

[15:05:04.0000] <Ms2ger>
File a bug

[15:06:00.0000] <annevk>
I guess I can

[15:06:01.0000] <AryehGregor>
Hmm, actually, I guess you'll mention my announcement anyway.

[15:06:02.0000] <AryehGregor>
In the WHATWG Weekly.

[15:06:03.0000] <AryehGregor>
So I don't see any need.

[15:06:04.0000] <timeless>
annevk: my favorite is explicitOriginalTarget

[15:06:05.0000] <annevk>
AryehGregor, please post it separately

[15:06:06.0000] <annevk>
I'll make an account for you

[15:06:07.0000] <AryehGregor>
Okay, why?

[15:06:08.0000] <timeless>
(there actually was one worse than that)

[15:06:09.0000] <timeless>
/me thinks it had `real` in it

[15:06:10.0000] <annevk>
AryehGregor, because six months of work deserves a little more than a note

[15:07:00.0000] <AryehGregor>
MediaWiki has or used to have series of methods in the same class called "X", "doX", "reallyDoX".  In at least one place.

[15:07:01.0000] <AryehGregor>
annevk, k.

[15:08:00.0000] <timeless>
annevk: http://mxr.mozilla.org/mozilla-central/source/dom/interfaces/events/nsIDOMNSEvent.idl#90

[15:08:01.0000] <timeless>
fwiw

[15:08:02.0000] <annevk>
you should have an invite for the blog AryehGregor

[15:08:03.0000] <timeless>
 105   [noscript] readonly attribute nsIDOMEventTarget  tmpRealOriginalTarget;

[15:08:04.0000] <timeless>
is the best part

[15:08:05.0000] <AryehGregor>
annevk, got it.

[15:09:00.0000] <annevk>
timeless, yeah, but it has [noscript] so I decided to not make jokes about it :)

[15:09:01.0000] <timeless>
annevk: yeah well

[15:09:02.0000] <timeless>
what i like is the comment

[15:09:03.0000] <timeless>
http://mxr.mozilla.org/mozilla-central/ident?i=tmpRealOriginalTarget

[15:09:04.0000] <timeless>
100   /* XXX This is TEMPORARY.

[15:10:00.0000] <annevk>
when was it added?

[15:10:01.0000] <annevk>
:)

[15:10:02.0000] <timeless>
hold on

[15:10:03.0000] <timeless>
i need to ask bonsai(cvs)

[15:10:04.0000] <timeless>
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/security/manager/ssl/src/nsSmartCardEvent.cpp&rev=1.1&root=/cvsroot

[15:10:05.0000] <timeless>
around 2005 in that file

[15:10:06.0000] <timeless>
http://bonsai.mozilla.org/cvsview2.cgi?diff_mode=context&whitespace_mode=show&root=/cvsroot&subdir=mozilla/editor/libeditor/text&command=DIFF_FRAMESET&file=nsPlaintextDataTransfer.cpp&rev2=1.35&rev1=1.34

[15:10:07.0000] <timeless>
2003 there

[15:11:00.0000] <timeless>
but that was probably just the rename from `explicitOriginalTarget` to `tmpRealOriginalTarget`

[15:12:00.0000] <timeless>
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/dom/public/idl/events/nsIDOMNSEvent.idl&rev=1.4.32.1#100

[15:12:01.0000] <annevk>
heh

[15:12:02.0000] <annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=674696

[15:12:03.0000] <timeless>
the change is from 2003

[15:12:04.0000] <timeless>
sorry, may 2003

[15:12:05.0000] <timeless>
so, it's only 8 years old

[15:12:06.0000] <annevk>
guess I'll file one on WebKit too

[15:13:00.0000] <timeless>
anyway, it looks like explicit could be made [noscript] trivially

[15:13:01.0000] <timeless>
which would be sufficient for your requirements

[15:14:00.0000] <annevk>
no the constants, originalTarget etc. are all proprietary

[15:15:00.0000] <jcranmer>
AryehGregor: there is an interface in Thunderbird which has "name", "prettyName", and "prettiestName" (with about 3 or 4 other "name"s)

[15:15:01.0000] <AryehGregor>
As the saying goes, there are only two hard problems in computer science: cache invalidation and naming things.

[15:16:00.0000] <Ms2ger>
And off-by-one errors

[15:16:01.0000] <annevk>
and off by one errors

[15:16:02.0000] <annevk>
:)

[15:16:03.0000] <Ms2ger>
:)

[15:18:00.0000] <annevk>
http://daringfireball.net/linked/2011/07/27/html5-fullscreen hmm

[15:32:00.0000] <annevk>
filed bugs ms2ger

[15:33:00.0000] <annevk>
and good night

[15:42:00.0000] <AryehGregor>
k, posted.

[15:42:01.0000] <AryehGregor>
timeless, subscribe to the WHATWG blog to get weekly-ish updates on interesting stuff that's happening in WHATWG-land.

[15:43:00.0000] <Hixie>
AryehGregor: cool, thanks

[15:46:00.0000] <annevk>
AryehGregor, could use some <code> love

[15:46:01.0000] <AryehGregor>
Yeah, maybe.

[15:46:02.0000] <AryehGregor>
Feel free if you like.  :)

[15:47:00.0000] <AryehGregor>
I was content with adding links.

[15:47:01.0000] <annevk>
meh

[15:53:00.0000] <AryehGregor>
Oh, so Google Plus lets you make a post that's officially public but you can secretly block arbitrary people from posting there?

[15:53:01.0000] <AryehGregor>
That seems sinister.

[15:53:02.0000] <annevk>
TabAtkins, yes it does

[15:54:00.0000] <annevk>
TabAtkins, but you get initial containing block and maybe that gets in the way

[15:54:01.0000] <annevk>
TabAtkins, like with a checkbox being the root

[15:58:00.0000] <annevk>
ok people

[15:58:01.0000] <annevk>
so here is the original email from Maciej

[15:58:02.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapi/2006Apr/0161.html

[15:58:03.0000] <annevk>
anyone who can find

[15:58:04.0000] <annevk>
Node item(in long index);

[15:58:05.0000] <annevk>
in WebKit and when/if it was removed gets a cookie

[15:59:00.0000] <jamesr>
can you do that?

[15:59:01.0000] <Hixie>
sure you just use Set-Cookie right?

[16:00:00.0000] <annevk>
AryehGregor, you just block people

[16:00:01.0000] <timeless>
will it be ;secure? :)

[16:00:02.0000] <AryehGregor>
annevk, yes, but it seems sinister that the discussion looks like it's public but some unknown list of people is blocked.

[16:01:00.0000] <AryehGregor>
Although I guess the discussion starter can also delete posts and things.

[16:01:01.0000] <AryehGregor>
So it's really controlled by them anyway.

[16:01:02.0000] <annevk>
AryehGregor, seems normal to me for discussions

[16:01:03.0000] <annevk>
pretty much any public forum has some amount of moderation

[16:02:00.0000] <annevk>
so they keep functioning

[16:02:01.0000] <AryehGregor>
Yeah, but it's usually some official moderator people or such, not the person who happened to start the thread.

[16:02:02.0000] <AryehGregor>
It means I might need to respond differently based on who started the thread, which seems weird to me.

[16:02:03.0000] <AryehGregor>
But my background is mailing lists and forums, not Facebook-style stuff.

[16:04:00.0000] <AryehGregor>
"To Areyh, I understand you are not a subject matter expert. Consequently, I put its importance back to major."

[16:04:01.0000] <AryehGregor>
Meh, see if I bother doing anything with a11y bugs but ignore them from now on.

[16:04:02.0000] <abarth>
Hixie: maybe ddhal would like to see the crypto-related feedback to whatwg?  he's driving.  i'm just helping him out

[16:05:00.0000] <timeless>
AryehGregor: which browser?

[16:05:01.0000] <jamesr>
so..when do i have to use adoptNode()?

[16:05:02.0000] <AryehGregor>
timeless, what?

[16:05:03.0000] <AryehGregor>
jamesr, AFAICT, you don't.

[16:05:04.0000] <timeless>
"To Areyh, I understand you are not a subject matter expert. Consequently, I put its importance back to major."

[16:06:00.0000] <jamesr>
AryehGregor: hm, ok

[16:06:01.0000] <jamesr>
i have a page where i create an iframe, put a canvas inside of it, then move the iframe to a popup window's document

[16:06:02.0000] <timeless>
oh, w3.org

[16:06:03.0000] <AryehGregor>
timeless, not a browser: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13176#c5

[16:06:04.0000] <timeless>
yeah, google :)

[16:06:05.0000] <jamesr>
the canvas itself doesn't survive the transition in firefox or opera

[16:06:06.0000] <jamesr>
why?

[16:07:00.0000] <AryehGregor>
I mean, I'm pretty sure severity is supposed to be set by the editor and/or editorial assistants, but frankly I don't care, I'll just ignore them from now on.

[16:07:01.0000] <jamesr>
hah, rich schwerdflkjerlkj

[16:07:02.0000] <jamesr>
AryehGregor: that guy is just a dick

[16:07:03.0000] <timeless>
AryehGregor: generally people don't understand how those fields work

[16:07:04.0000] <AryehGregor>
timeless, I object to the condescension.

[16:07:05.0000] <timeless>
and bugzilla at some day should be changed to explain it

[16:08:00.0000] <timeless>
ideally they're the reporter's realm until a qa comes

[16:08:01.0000] <timeless>
once the qa claims it, they're the qa's until an assignee touches the bug

[16:08:02.0000] <timeless>
at which point, they're the assignee's

[16:08:03.0000] <AryehGregor>
If they want to be jerks, their bugs can sit around forever without resolution so they can't escalate without chair intervention.  I don't care.

[16:08:04.0000] <timeless>
but bugzilla doesn't enforce that

[16:08:05.0000] <timeless>
heh

[16:08:06.0000] <AryehGregor>
jamesr, you're right, I was too hasty.  I'll just ignore his bug reports, not other a11y people's.

[16:09:00.0000] <AryehGregor>
timeless, we have guidelines for severity: http://wiki.whatwg.org/wiki/Bugzilla_conventions#Severities

[16:09:01.0000] <AryehGregor>
"major" is unused.

[16:09:02.0000] <AryehGregor>
But I don't know if severity is a useful field anyway.

[16:09:03.0000] <AryehGregor>
In my experience, it's almost always inaccurate enough that it's useless, and changing it starts fights.

[16:10:00.0000] <timeless>
the last bit is always true :)

[16:10:01.0000] <AryehGregor>
Often of the form "developer sets severity to enhancement, reporter gets offended because they interpret that as 'not important'".

[16:10:02.0000] <timeless>
be happy it isn't a private bug tracker

[16:10:03.0000] <timeless>
at nokia, we had a group who kept adding `escalate` everywhere

[16:11:00.0000] <timeless>
/me thinks that was the l10n testing group as opposed to the a11y group, but..

[16:12:00.0000] <timeless>
do they have a sample page demoing a real use case for this?

[16:12:01.0000] <timeless>
it's way too abstract

[16:13:00.0000] <jamesr>
i don't know if you want to start going down the 'can we get use cases' path

[16:13:01.0000] <timeless>
i don't mean use case as in `tell me that you need accessibility`

[16:13:02.0000] <timeless>
i mean `show me a site that's trying and has something so i can see how the problem really manifests`

[16:13:03.0000] <timeless>
problems you can't touch, grasp, or hold are much harder to address properly

[16:14:00.0000] <jamesr>
http://lists.w3.org/Archives/Public/public-canvas-api/ is full of people asking them (richard schwesdflkjsdlkfj and charles prichard) for exactly that and them writing novellas that don't quite give concrete use cases but sure do use a lot of words

[16:14:01.0000] <timeless>
heh

[16:15:00.0000] <timeless>
i'm not an expert, but i do actually *use* accessibility agents

[16:16:00.0000] <timeless>
and i file bugs against them (sometimes even with patches)

[16:17:00.0000] <timeless>
/me waves goodbye to irccloud.com

[16:17:01.0000] <clair>
Hi guys, I want to check before I do so, is it ok for non w3c members to sign up to get CC notifications of w3c bugs?

[16:18:00.0000] <timeless>
clair: i don't see why it'd be a problem

[16:18:01.0000] <timeless>
other than the fact that the bugs will probably be 99% noise

[16:18:02.0000] <clair>
It's just one particular bug I'm interested in, really, I assume you can subscribe to a single one?

[16:19:00.0000] <timeless>
clair: although, i'm not sure `cc notifications` means what you want it to mean

[16:19:01.0000] <timeless>
yes

[16:19:02.0000] <timeless>
you can get an account and cc yourself to the bug

[16:19:03.0000] <clair>
Yeah that's what I'm thinking - I assume I then get email replies to the ticket?

[16:20:00.0000] <timeless>
yes

[16:20:01.0000] <timeless>
you can control which types

[16:20:02.0000] <clair>
OK, thanks :)

[16:21:00.0000] <timeless>
http://www.bugzilla.org/docs/4.0/en/html/userpreferences.html#emailpreferences

[16:21:01.0000] <timeless>
is vaguely relevant

[16:21:02.0000] <timeless>
sadly, the doc doesn't actually have much to say directly about email

[16:21:03.0000] <timeless>
it should have a section about it

[16:21:04.0000] <timeless>
note to self: file bug

[16:39:00.0000] <Hixie>
abarth: k

[16:43:00.0000] <Hixie>
hm, if we're going to have other specs on whatwg.org maybe we should have a subdomain

[16:43:01.0000] <Hixie>
like .specs.whatwg.org

[16:43:02.0000] <Hixie>
so e.g. fullscreen.specs.whatwg.org

[16:43:03.0000] <Hixie>
or fullscreen.s.whatwg.org

[16:43:04.0000] <Hixie>
for tantek's

[16:43:05.0000] <Hixie>
and wa1.s.whatwg.org or html.s.whatwg.org or whatever we call the complete spec

[16:44:00.0000] <nessy>
leave out the specs part

[16:44:01.0000] <abarth>
would those domains point to whatwg servers, or how would we push specs there?

[16:44:02.0000] <Hixie>
i'd give the editor of each spec a separate account

[16:44:03.0000] <nessy>
fullscreen.whatwg.org , html.whatwg.org work for me

[16:44:04.0000] <Hixie>
or we could point them to whever is most appropriate

[16:44:05.0000] <Hixie>
nessy: the idea would be to keep them separate from blogs. forums. status. and so on

[16:45:00.0000] <nessy>
ah, hmm

[16:45:01.0000] <nessy>
.doc. ?

[16:45:02.0000] <abarth>
Hixie: ok, i'd like to try that for the sniffing spec.  Pete Resnick seems pretty opposed to it moving forward at IETF

[16:45:03.0000] <Hixie>
.s. :-)

[16:45:04.0000] <Hixie>
abarth: ah, k

[16:45:05.0000] <nessy>
.s. ist a bit meaningless that's all

[16:45:06.0000] <Hixie>
abarth: do you have a timeline on this?

[16:45:07.0000] <nessy>
looks like  a typo

[16:45:08.0000] <Hixie>
short for .spec.

[16:46:00.0000] <abarth>
Hixie: whenever is convenient for you.  there's no real rush

[16:46:01.0000] <nessy>
yeah, I know, but still looks like a typo ;-)

[16:46:02.0000] <Hixie>
html.spec.whatwg.org fullscreen.spec.whatwg.org

[16:46:03.0000] <Hixie>
maybe .spec. is short enough

[16:46:04.0000] <nessy>
WFM

[16:46:05.0000] <abarth>
html.TR.whatwg.org

[16:46:06.0000] <Hixie>
where are ms2ger and anne when you need them

[16:46:07.0000] <Hixie>
abarth: hahaha

[16:47:00.0000] <Hixie>
tempting

[16:50:00.0000] <Hixie>
jeez, whatwg.org has a lot of subdomains already

[16:52:00.0000] <Hixie>
blog, c, demos, developers, forums, help, history, images, n, status, svn, validator, wiki, xn--7ca. (history and help are dead, demos and status are just subdomains for other subsubdomains)

[16:54:00.0000] <jamesr>
what's the correct name for "all unicode characters whose code points are <= 255"?

[16:55:00.0000] <jamesr>
ascii plus some

[16:55:01.0000] <Hixie>
"Unicode codepoints in the range U+0000 to U+00FF"


2011-07-28
[17:02:00.0000] <annevk>
I sort of dislike subdomains but I guess it's the easiest

[17:04:00.0000] <Hixie>
subdirectories would be much more difficult to deal with

[17:04:01.0000] <jcranmer>
obviously, you should replace all subdirectories with subdomains

[17:05:00.0000] <Hixie>
also subdomains have the advantage of being great in autocomplete in browsers :-)

[17:05:01.0000] <Hixie>
jcranmer: i mean, when you have multiple people involved

[17:05:02.0000] <annevk>
whatwg.org/c versus html.spec.whatwg.org

[17:05:03.0000] <annevk>
i guess we could still have shortcuts

[17:05:04.0000] <Hixie>
whatwg.org/c would still work

[17:05:05.0000] <Hixie>
but remember, that's why we added c.whatwg.org

[17:06:00.0000] <Hixie>
if i do this btw i'm merging complete.html and HTML together and calling the whole thing HTML again

[17:06:01.0000] <annevk>
yes

[17:06:02.0000] <annevk>
want

[17:07:00.0000] <annevk>
so html.spec.whatwg.org dom.spec.whatwg.org

[17:07:01.0000] <annevk>
and spec.whatwg.org gives an overview?

[17:08:00.0000] <Hixie>
hadn't thought of spec.whatwg.org but sure

[17:08:01.0000] <Hixie>
we can have that show the same as platform.html5.org

[17:08:02.0000] <annevk>
yeah or a redirect to the wiki

[17:08:03.0000] <Hixie>
yeah

[17:09:00.0000] <Hixie>
btw in other news i think i've done the merging of all the dom core stuff

[17:09:01.0000] <Hixie>
annevk: do you have a dreamhost account?

[17:10:00.0000] <annevk>
yes

[17:11:00.0000] <Hixie>
do you want dom.spec.whatwg.org in your account?

[17:11:01.0000] <Hixie>
i wonder if we can do that

[17:14:00.0000] <Hixie>
hmm

[17:18:00.0000] <Hixie>
i'll ask them

[17:22:00.0000] <annevk>
you can by setting the nameservers accordingly

[17:22:01.0000] <annevk>
hmm actually, dunno

[17:23:00.0000] <annevk>
because you'd have the same nameservers :)

[17:27:00.0000] <Hixie>
yeah

[17:28:00.0000] <Hixie>
i sent a support req

[17:34:00.0000] <annevk>
cool about DOM Core integration

[17:34:01.0000] <annevk>
will review

[18:04:00.0000] <annevk>
so SVG wants on*

[18:04:01.0000] <annevk>
so infrastructure should probably be in DOM Core

[18:04:02.0000] <annevk>
problem: SVG needs "evt" exposed

[18:04:03.0000] <annevk>
also, it should be in DOM Core because browsers implement it on Element, not HTMLElement

[18:13:00.0000] <Hixie>
are we having all the on* handlers on every element?

[18:13:01.0000] <Hixie>
or only the SVG ones on SVGElement and HTML ones on HTMLElement?

[18:13:02.0000] <Hixie>
also which are going on Window?

[18:14:00.0000] <annevk>
all on Element

[18:14:01.0000] <annevk>
with the "evt" compat integrated

[18:14:02.0000] <annevk>
I guess on Document and Window too

[18:15:00.0000] <Hixie>
that's a _lot_ of redundant event handler attributes

[18:15:01.0000] <annevk>
I was planning on defining the infrastructure

[18:15:02.0000] <annevk>
isn't that the situation we have now?

[18:15:03.0000] <annevk>
anyway, I was planning on defining the infrastructure and let the "expose this list" to other specs

[18:18:00.0000] <Hixie>
we don't have that many events currently

[18:18:01.0000] <Hixie>
i guess it depends on how many SVG would be adding

[18:18:02.0000] <Hixie>
i'm used to SVG adding six bazillion of everything

[18:32:00.0000] <annevk>
whoa http://www.w3.org/Bugs/Public/show_bug.cgi?id=12417

[18:32:01.0000] <annevk>
Hixie, SVG overlaps a lot with what we have

[18:32:02.0000] <annevk>
Hixie, in the meeting this morning they identified about 4/5 events

[18:33:00.0000] <annevk>
Hixie, that would need to be added

[18:33:01.0000] <annevk>
Hixie, I think sort of prefer this stays in HTML though given the dependency on script context, and returnValue and such

[18:34:00.0000] <annevk>
but I think bz wants them to move from HTMLElement to Element

[18:35:00.0000] <Hixie>
ok, glad it's not much

[18:36:00.0000] <Hixie>
i still think we should just merge dom core and html, personally

[18:36:01.0000] <Hixie>
but... :-)

[18:36:02.0000] <Hixie>
i mean, it's not like SVG can avoid depending on Window

[18:37:00.0000] <annevk>
hmm yeah

[18:37:01.0000] <annevk>
only Progress Events solely has a dependency on DOM Core

[18:37:02.0000] <annevk>
and not HTML

[18:38:00.0000] <annevk>
anyway, I rather not go there for now

[18:38:01.0000] <annevk>
slow change works better with the rest of the world

[22:21:00.0000] <annevk>
what is the use case of DocumentFragment? so you can have several element siblings instead of always a root?

[22:24:00.0000] <nimbu>
annevk: ut?

[22:26:00.0000] <annevk>
yeah

[22:26:01.0000] <nimbu>
k pmm

[22:53:00.0000] <annevk>
added a link to platform.html5.org on http://html5.org/

[23:04:00.0000] <annevk>
where is ms2ger?

[23:04:01.0000] <annevk>
hmm

[23:04:02.0000] <annevk>
http://html5.org/temp/dom-mutation-methods.txt

[23:04:03.0000] <annevk>
i think those steps are sufficient

[23:06:00.0000] <annevk>
but I'd like some review before I go change the spec

[23:34:00.0000] <boblet>
wassup with the blog roll nav example in #the-aside-element? I would have thought a blog roll would prolly not be major navigation…

[23:35:00.0000] <boblet>
just a case of “major nav for the author of this page”?

[00:07:00.0000] <hsivonen>
I wonder what possessed the Google News team to use dc.date.issued instead of dc.issued or dcterms.issued

[01:05:00.0000] <hsivonen>
"SVG would probably have remained a dead technology if Google hadn't started carrying its banner with Chrome"

[01:05:01.0000] <hsivonen>
eh?

[01:07:00.0000] <heycam>
s/Google/MS/ and s/Chrome/IE/? :)

[01:13:00.0000] <hsivonen>
it's remarkable how Chrome has managed to create a reality distortion field like this

[01:13:01.0000] <heycam>
who wrote that quote? just some random person?

[01:14:00.0000] <hsivonen>
heycam: Kurt Cagle

[01:14:01.0000] <heycam>
huh.

[01:14:02.0000] <hsivonen>
https://plus.google.com/114141433688365651943/posts/aGDpEJX7jj8

[01:14:03.0000] <heycam>
he used to be big into SVG, so I am surprised he would have that "reality-challenged" view

[01:14:04.0000] <heycam>
(being close to things a number of years ago, that is)

[03:27:00.0000] <boblet>
http://www.htmlvalidator.com/htmlval/whycseisbetter.html lol

[03:28:00.0000] <boblet>
also, ppl pay for validators? :o

[03:28:01.0000] <boblet>
hsivonen: ^^^

[03:32:00.0000] <espadrine`>
boblet: "All HTML documents should have titles and "Untitled" is not a good title"...

[03:32:01.0000] <espadrine`>
 They are checking for the word "Untitled" in <title>!

[03:32:02.0000] <boblet>
espadrine`: validator+ ? ;) that’s more a linting feature

[03:33:00.0000] <boblet>
(it is good advice tho)

[03:35:00.0000] <hsivonen>
boblet: I'm quite happy with my business model that doesn't involve worrying about collecting my income from end users

[03:36:00.0000] <boblet>
hsivonen: I was meaning more why would you pay for a validator when validator.nu is so awesum

[03:37:00.0000] <espadrine`>
it *does* find more errors than their CSE HTML Validator Lite v9.02

[03:39:00.0000] <hsivonen>
I wonder what accessibility and SEO checking involves in practice

[03:39:01.0000] <Ms2ger>
longdesc and meta keywords?

[03:40:00.0000] <jgraham>
hsivonen: "Missing meta description tag used by some search engines"

[03:40:01.0000] <hsivonen>
jgraham: is that an actual quote from that product?

[03:41:00.0000] <jgraham>
hsivonen: Yes

[03:42:00.0000] <jgraham>
See the page boblet linked

[03:43:00.0000] <boblet>
espadrine`: also that’s only validation errors, not other stuff (linting, SEO etc)

[03:43:01.0000] <hsivonen>
jgraham: oh

[03:43:02.0000] <boblet>
W3 validator finds a bunch using HTML5 mode too, unsurprisingly

[03:43:03.0000] <hsivonen>
boblet: the product seems to be a syntax-highlighting text editor and validators in one product

[03:44:00.0000] <hsivonen>
boblet: which is useful

[03:44:01.0000] <hsivonen>
boblet: Validator.nu (alone) doesn't give you an editor to fix the errors in

[03:44:02.0000] <boblet>
yeah I’d like validator.nu locally to operate on a project level

[03:44:03.0000] <hsivonen>
hendry has integrated Validator.nu into vi, though

[03:44:04.0000] <boblet>
hsivonen: but they’re putting the validation aspect front and center in their branding huh

[03:45:00.0000] <boblet>
but that would mean learning vi :p

[03:46:00.0000] <hsivonen>
boblet: well, the validation features seem to be the bulk of the product

[03:46:01.0000] <hsivonen>
boblet: so it would be silly to advertise it as an editor that, oh by the way, has a validator

[03:47:00.0000] <espadrine`>
hsivonen: Is vi integration offline?

[03:47:01.0000] <boblet>
hey adactio, nice article on IDs and ARIA roles from Jan that I somehow missed. linked it up

[03:47:02.0000] <espadrine`>
I assumed it used internet

[03:47:03.0000] <hsivonen>
espadrine`: if you run Validator.nu on localhost, yes

[03:47:04.0000] <adactio>
boblet: merci

[03:48:00.0000] <espadrine`>
That could be interesting!

[03:48:01.0000] <espadrine`>
One day, there will be a port dedicated to html validation.

[04:23:00.0000] <Ms2ger>
[07:12] <annevk> where is ms2ger?

[04:23:01.0000] <Ms2ger>
At 7 AM? What were you thinking? :)

[05:25:00.0000] <matjas>
any feedback on http://www.w3.org/Bugs/Public/show_bug.cgi?id=13118? (“Consider firing the `input` event for contenteditable areas”) /cc Hixie

[05:33:00.0000] <matjas>
seems like a no-brainer to me, I must be missing something

[05:47:00.0000] <Ms2ger>
All you're missing is that there's a couple of hundred bugs that need to be addressed before yours

[06:25:00.0000] <hsivonen>
foolip's Java looks Pythonic. it's shockingly compact as far as Java code goes

[06:26:00.0000] <foolip>
hsivonen, I'm not sure if that is a compliment or not :)

[06:27:00.0000] <hsivonen>
foolip: it's a compliment

[06:27:01.0000] <foolip>
found any horrible bugs yet?

[06:27:02.0000] <hsivonen>
foolip: not yet

[06:28:00.0000] <foolip>
great, hope it works out

[06:28:01.0000] <foolip>
I was hacking on collecting itemValue yesterday, with that vocabulary validation should be in reach

[06:28:02.0000] <foolip>
but I'm not sure how to structure

[06:28:03.0000] <foolip>
it

[06:29:00.0000] <foolip>
did you have ideas about transforming it to something and using existing schema languages?

[06:30:00.0000] <hsivonen>
foolip: I've considered mapping Microdata to XML in order to use RELAX NG on it

[06:30:01.0000] <foolip>
hsivonen, could you do that while still warning at the appropriate source location?

[06:30:02.0000] <jgraham>
Needs more AbstractSingletonProxyFactoryBean

[06:30:03.0000] <hsivonen>
foolip: so far, I'm unsure if that would help or hurt

[06:30:04.0000] <hsivonen>
foolip: the source location could be preserved

[06:31:00.0000] <hsivonen>
foolip: but I'm not sure what would happen with property names that aren't NCNames

[06:31:01.0000] <foolip>
I was thinking about just representing items and properties with the Element class, mirroring the DOM API

[06:31:02.0000] <hsivonen>
so far, I've learned that itemref doesn't mean what I thought it meant

[06:31:03.0000] <foolip>
ok, what did you think it did?

[06:32:00.0000] <foolip>
the main question is of course how to represent all of schema.org

[06:32:01.0000] <hsivonen>
I thought its referent had to be an item so that it gave a dislocated item value to an item-valued property

[06:32:02.0000] <foolip>
ah, the item* prefix is misleading sometimes

[06:33:00.0000] <foolip>
just like itemValue is not the value of an item, but of a property

[06:34:00.0000] <foolip>
there's bound to be things that need code to validate, something like http://schema.rdfs.org/all.json won't suffice for all vocabularies

[06:35:00.0000] <hsivonen>
as I read the spec more, itemref is totally different from what I thought it was

[06:35:01.0000] <hsivonen>
I wonder why I thought it was something other than what it is

[06:36:00.0000] <foolip>
without itemref, microdata would be totally trivial

[06:36:01.0000] <foolip>
I wonder if Google search will actually implement it properly

[06:37:00.0000] <foolip>
it seems rather unlikely, really

[06:38:00.0000] <hsivonen>
what use case lead to the current design of itemref?

[06:39:00.0000] <foolip>
let me dig up a real-world example

[06:39:01.0000] <hsivonen>
the spec itself mentions having items as columns of a table while the properties are in the cells

[06:39:02.0000] <hsivonen>
but that's not really a use case

[06:39:03.0000] <foolip>
http://www.2gc.co.uk/a2gc-people

[06:40:00.0000] <hsivonen>
foolip: hmm. ok

[06:40:01.0000] <foolip>
in a nutshell: the information is interleaved throughout the page

[06:41:00.0000] <foolip>
I'm not sure we're not going to regret itemref down the line, but so far it's not too bad

[06:42:00.0000] <cygri>
 at least itemref degrades somewhat gracefully

[06:42:01.0000] <foolip>
if leaving out half of the information is graceful, then yes :)

[06:43:00.0000] <foolip>
50% is infinitely better than 0%

[06:43:01.0000] <foolip>
(as if 1%)

[06:43:02.0000] <foolip>
s/if/is/

[07:07:00.0000] <hsivonen>
foolip: given how much complexity itemref adds to the language, it seems that the it wouldn't be that big of a deal to make properties forward-compatible with changing from string-valued to item-valued

[07:08:00.0000] <hsivonen>
although that's in the department of theoretical problems until we get far enough that someone has that problem

[07:08:01.0000] <hsivonen>
(and then it's too late)

[07:08:02.0000] <foolip>
hsivonen, I'm not sure what you mean

[07:08:03.0000] <foolip>
what's string-valued vs item-valued?

[07:08:04.0000] <hsivonen>
foolip: what TabAtkins sent email to the list about

[07:08:05.0000] <hsivonen>
foolip: properties have either strings or items as their values

[07:09:00.0000] <foolip>
right

[07:09:01.0000] <foolip>
is this an old email I've forgotten?

[07:09:02.0000] <foolip>
the one I replied to perhaps

[07:09:03.0000] <hsivonen>
foolip: so suppose you have a list of tracks of an album and each track is a string that is the tracks name

[07:09:04.0000] <hsivonen>
foolip: then someone who isn't the author of the page writes consumption code that expects string values

[07:10:00.0000] <hsivonen>
foolip: then publishers decide that tracks are now items

[07:10:01.0000] <hsivonen>
foolip: the consuming code breaks

[07:10:02.0000] <foolip>
yeah, I agree that it is a problem

[07:10:03.0000] <foolip>
but I'm not thrilled about having a dual interpretation of properties where some consumers end up having to consider *both* at once

[07:11:00.0000] <hsivonen>
it could be address by having the consumer say to his/her microdata tooling "get property foo as string"

[07:11:01.0000] <foolip>
for example, because people only tested with consumers that consider it a string, and didn't notice the item representation was broken

[07:11:02.0000] <hsivonen>
and the tooling to have a rule about what to do if property foo is an item

[07:11:03.0000] <hsivonen>
foolip: good point

[07:11:04.0000] <hsivonen>
maybe I'm just a Complicator here

[07:12:00.0000] <foolip>
and if you also have the reverse, you have to look at both and apply heuristics

[07:12:01.0000] <foolip>
but without seeing what the consuming code would look like, it's hard to make guesses at solutions

[07:13:00.0000] <hsivonen>
foolip: the solution could be an itemname that's a privileged property

[07:13:01.0000] <foolip>
itemNAME, a new attribute?

[07:14:00.0000] <hsivonen>
foolip: so all vocabularies would have to use <span itemname>Whatever</span> for their most nameish or titleish property

[07:14:01.0000] <hsivonen>
instead of <span itemprop=title>Whatever</span>

[07:15:00.0000] <foolip>
solving it by making authors cater to the problem sounds like it'll only work 1% of the type

[07:15:01.0000] <foolip>
it broke because the author was careless in upgrading from string to item without checking what existing consumers do...

[07:15:02.0000] <hsivonen>
well, the solution requires vocabulary designers to cooperate

[07:16:00.0000] <hsivonen>
not authors per se

[07:16:01.0000] <foolip>
the problem is bound to happen, but I've failed to come up with an opinion on the matter :)

[07:16:02.0000] <hsivonen>
/me gets some gloves http://thedailywtf.com/Articles/The_Complicator_0x27_s_Gloves.aspx

[07:17:00.0000] <hsivonen>
though I don't know what the gloves would be in this case

[07:17:01.0000] <foolip>
oh, is The Complicator a meme of some sort?

[07:18:00.0000] <foolip>
/me replies to bug 850

[07:18:01.0000] <hsivonen>
foolip: yes

[08:46:00.0000] <matjas>
Ms2ger: I can wait :) was just wondering if this is something that has a chance to be specced, since only WebKit has this behavior atm

[08:46:01.0000] <Ms2ger>
If webkit has it and Gecko is in favour, sounds like it does

[09:21:00.0000] <foolip>
hsivonen, do you want the changes based on your feedback as a new complete patch, or as a delta on top of the old one?

[09:53:00.0000] <dglazkov>
good morning, Whatwg

[09:53:01.0000] <jcranmer>
wake up and smell the ashes

[09:56:00.0000] <hober>
morning dglazkov

[09:56:01.0000] <hsivonen>
foolip: either way works

[09:56:02.0000] <foolip>
ok

[09:57:00.0000] <Ms2ger>
Evening

[09:57:01.0000] <Michael>
Hello :)

[09:58:00.0000] <dglazkov>
you are a friendly bunch I see

[11:08:00.0000] <AryehGregor>
. . . why does http://platform.html5.org/ link to the W3C version of HTML?

[11:08:01.0000] <Ms2ger>
Because MikeSmith works for the W3C?

[11:09:00.0000] <Ms2ger>
Not saying it's a conspiracy, just saying

[11:09:01.0000] <AryehGregor>
Phooey.

[11:16:00.0000] <annevk>
do we need to define that "an /x/ element" means "an element with local name /x/ and namespace ?!"?

[11:17:00.0000] <Ms2ger>
Where?

[11:20:00.0000] <annevk>
in DOM Core

[11:20:01.0000] <AryehGregor>
annevk, that would be nice, actually.

[11:20:02.0000] <annevk>
we talk about "html element" sometimes but that is never really mapped to local name

[11:20:03.0000] <annevk>
I guess you would still need to mention the namespace

[11:20:04.0000] <AryehGregor>
Currently I just make /x/ an xref to the relevant element in the HTML spec.

[11:20:05.0000] <Ms2ger>
Mm

[11:21:00.0000] <annevk>
an /x/ element in namespace /y/

[11:21:01.0000] <AryehGregor>
But that only works if I have a specific name, and it's not a variable.

[11:21:02.0000] <annevk>
also that does not work for DOM Core as long as we do not want to depend on HTML

[11:21:03.0000] <AryehGregor>
Otherwise I'm forced to say "an <span>HTML element</span> with <span ...>local name</span> <var>tag name</var>" or such.

[11:22:00.0000] <AryehGregor>
Where I've defined "HTML element" to mean "Element in the HTML namespace".  Dunno if that duplicates definitions elsewhere.

[11:33:00.0000] <AryehGregor>
Okay, my latest Google+ post is going to cause a massive drama-fest.

[11:33:01.0000] <AryehGregor>
But I hope Jeni et al. appreciate the points I make and it makes some kind of difference.

[11:33:02.0000] <AryehGregor>
If the peanut gallery wants to freak out, I'm happy to ignore them.

[11:33:03.0000] <AryehGregor>
https://plus.google.com/u/0/105458233028934590147/posts/h7nsT7wuNmX

[11:33:04.0000] <AryehGregor>
/me can't figure out if you can link directly to posts

[11:34:00.0000] <AryehGregor>
/me finds an id for the post of z135hbfqbruhcpvbj04chplqevjesd4xigk#1311874590176000, but somehow doubts that's meant to be used for anchors

[11:36:00.0000] <Ms2ger>
"I know you're all perfectly reasonable people who are committed to improving the web"

[11:36:01.0000] <Ms2ger>
My sarcasm detector is failing me

[11:36:02.0000] <AryehGregor>
I wasn't being sarcastic at all.

[11:37:00.0000] <AryehGregor>
I think they're doing a bad job at improving the web, in certain key respects, but they're trying.

[11:37:01.0000] <Ms2ger>
Maybe it's just my cynicism that made me doubt that :)

[11:39:00.0000] <Philip`>
AryehGregor: Given that Google thinks that strings like 105458233028934590147 and h7nsT7wuNmX and aapbdbdomjkkjkaonfhkkikfgjllcleb etc are good things to use in URLs in place of human-readable identifiers, I don't see why they wouldn't think z135hbfqbruhcpvbj04chplqevjesd4xigk#1311874590176000 was perfectly acceptable too

[11:39:01.0000] <AryehGregor>
Philip`, to start with, it has a # in it, so it wouldn't register as a valid URL in lots of contexts.

[11:41:00.0000] <Ms2ger>
FWIW, I have vague plans to move my specs to the W3C, but it's low enough priority that it may never happen

[11:41:01.0000] <Philip`>
Wouldn't that just get replaced with a %34 or whatever?

[11:42:00.0000] <Philip`>
Oh, %23

[11:42:01.0000] <Philip`>
/me was only off by one (twice)

[11:49:00.0000] <brucel>
Who's up for anwering a stoopid microdata question (prompted by TabAtkins' blogpost)?

[11:50:00.0000] <annevk>
AdobeGuest network is so slow

[11:50:01.0000] <annevk>
just ask brucel

[11:50:02.0000] <Ms2ger>
Maybe if you pushed an Adobe employee into a lake...

[11:50:03.0000] <annevk>
more chance than with asking to ask

[11:51:00.0000] <brucel>
OK - itemid: according to spec, "The itemid attribute must not be specified on elements that do not have both an itemscope attribute and an itemtype attribute specified, and must not be specified on elements with an itemscope attribute whose itemtype attribute specifies a vocabulary that does not support global identifiers for items, as defined by that vocabulary's specification.?

[11:52:00.0000] <brucel>
how do I know if a vocab "supports" global indetifiers?

[11:53:00.0000] <brucel>
Tabh's blogpost http://www.xanthir.com/blog/b4570 leads me to assume that it means: if a vocab has something that's primary-key like, eg ISBN or social security  number, then you can use that

[11:53:01.0000] <cygri>
brucel: the documentation of the vocabulary should explain whether it supports itemid

[11:54:00.0000] <cygri>
if the documentation doesn't say anything about itemid, then don't use it

[11:54:01.0000] <brucel>
but "supports" could mean that it conforms to some technical schema full of OWLs and SPARQLs and things that upset my hippocampus

[11:55:00.0000] <brucel>
microdata spec tells me that certain vocabs support global identifiers, but then don't use itemid in the examples, so am not any wiser

[11:56:00.0000] <cygri>
brucel, look at the vevent, vcard and so on vocabs that are part of the microdata spec

[11:56:01.0000] <cygri>
they explain where you can use itemid

[11:56:02.0000] <cygri>
OWL and SPARQL has nothing to do with this really

[11:57:00.0000] <annevk>
euh

[11:58:00.0000] <annevk>
wtf is going on here: http://lists.w3.org/Archives/Public/www-archive/2011Jul/0025.html

[11:58:01.0000] <annevk>
hint: dates

[11:59:00.0000] <brucel>
cygri have been reading here http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html#mdvocabs but none of the examples employ itemid

[12:00:00.0000] <cygri>
brucel, hm, how to explain this

[12:00:01.0000] <cygri>
itemid usually wouldn't be something like the social security number

[12:01:00.0000] <cygri>
it would just be a URL that, according to the publisher of the HTML page, is a globally unique identifier for that item

[12:02:00.0000] <cygri>
so, if i were to add microdata to my own homepage, i'd use itemid="http://richard.cyganiak.de/#cygri" or something like that, and not my social security number

[12:03:00.0000] <brucel>
cygri thanks - and what's the benefit of your doing that

[12:04:00.0000] <Ms2ger>
lolurls

[12:05:00.0000] <Ms2ger>
Having taken ten years to confirm that we implemented your feedback...

[12:05:01.0000] <cygri>
well, then this url can be used in an itemprop value to refer to me

[12:05:02.0000] <Ms2ger>
That's worse than Hixie

[12:05:03.0000] <annevk>
Ms2ger, did you look at http://html5.org/temp/dom-mutation-methods.txt ?

[12:05:04.0000] <Ms2ger>
" This draft is expected to be updated or made obsolete within three months of its publication (3 October 2002)."

[12:06:00.0000] <Ms2ger>
Looked at it, yes

[12:06:01.0000] <Ms2ger>
Read, no

[12:07:00.0000] <cygri>
brucel, a better example might be an event. you can give it a unique identifier with itemid, and then someone could use that id to state that they're attending the event or whatever, assuming you have a vocabulary that makes use of itemids in that way

[12:07:01.0000] <annevk>
SVG WG is going to discuss HTML in SVG soonish btw

[12:07:02.0000] <annevk>
if people have thoughts

[12:07:03.0000] <cygri>
brucel, you *could* still use things like urn:isbn:123456 for itemids

[12:09:00.0000] <brucel>
ok, thanks cygri

[12:09:01.0000] <cygri>
yw brucel

[12:10:00.0000] <Philip`>
Ms2ger: At least they implemented the feedback promptly, whereas Hixie can take 5 years to read a message and then replies asking for clarification on some of the points

[12:10:01.0000] <brucel>
feels a bit ... edgecasey.. but I think I understand enough to know that I don't need to use it if I'm marking up people, isbns, events

[12:11:00.0000] <Ms2ger>
Philip`, please elaborate.

[12:12:00.0000] <Philip`>
Ms2ger: I'll elaborate in 2016 if that's alright

[12:12:01.0000] <AryehGregor>
Philip`, be fair, he just said that his oldest piece of feedback was from 2007.

[12:12:02.0000] <Ms2ger>
Sure

[12:12:03.0000] <AryehGregor>
That's only four years.

[12:13:00.0000] <AryehGregor>
(BTW, Hixie, it is a real problem that you don't respond to feedback for months or years, it massively discourages feedback)

[12:13:01.0000] <AryehGregor>
(although you edit so much that maybe you don't have any option)

[12:13:02.0000] <AryehGregor>
/me plan to never edit so many specs that he can't respond to feedback promptly

[12:14:00.0000] <dglazkov>
Hixie has become a one-person bureaucracy!!! :P

[12:14:01.0000] <Philip`>
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2011-June/032026.html was 6.5 years after some comments, or 5 if you're feeling charitable

[12:16:00.0000] <Philip`>
(...and asks non-rhetorical questions in a few cases)

[12:16:01.0000] <Ms2ger>
annevk, looks good

[12:17:00.0000] <annevk>
cool, now I need to figure out how to rewrite it nicely

[12:26:00.0000] <dglazkov>
I am genuinely surprised by negative reaction to AryehGregor's spec announcement.

[12:27:00.0000] <AryehGregor>
dglazkov, then you aren't familiar with W3C politics.

[12:27:01.0000] <jamesr>
negative reaction where?

[12:27:02.0000] <dglazkov>
I do avoid them at all costs.

[12:27:03.0000] <Ms2ger>
All over the W3C, I guess

[12:34:00.0000] <annevk>
inner substeps yay

[12:35:00.0000] <annevk>
at least these either continue or fail

[12:35:01.0000] <AryehGregor>
/me pokes hober with a pointy stick in the direction of innerText

[12:35:02.0000] <AryehGregor>
Pretty soon I'm going to give up on getting a response from WebKit here.

[12:42:00.0000] <gsnedders>
Ms2ger: All those specs are WebApps WG not HTML WG, so nothing for the HTML testsuite

[12:42:01.0000] <Ms2ger>
appcache?

[12:42:02.0000] <gsnedders>
Oh, that isn't

[12:42:03.0000] <gsnedders>
/me can't read

[12:43:00.0000] <Ms2ger>
And webapps doesn't have a test list afaik

[12:47:00.0000] <jgraham>
WebSQL is dead isn't it?

[12:56:00.0000] <MacTed>
jgraham - spec appears so, judging by  http://www.w3.org/TR/webdatabase/ ...

[12:56:01.0000] <MacTed>
unfortunate, and surprising that they say all implementors used same backend (SQLite), as we implemented bridges to ODBC amd XMLA -- http://wikis.openlinksw.com/dataspace/owiki/wiki/UdaWikiWeb/InstallConfigHTML5SQLBridges -- and thus backends may be Oracle, DB2, Informix, MSSQL, MySQL, PostgreSQL, etc.

[12:58:00.0000] <Ms2ger>
Very fortunate

[12:58:01.0000] <jgraham>
Tests should be ported to testharness.js too

[12:58:02.0000] <Ms2ger>
Yeah

[12:58:03.0000] <MacTed>
Ms2ger - why fortunate?

[12:58:04.0000] <Ms2ger>
Because that would have led to a sqlite monoculture

[12:59:00.0000] <Ms2ger>
And as nice as it might be, it would have had all the problems of a monoculture

[12:59:01.0000] <MacTed>
Ms2ger - did you miss the part where we implemented ODBC and XMLA binding?  as in, monoculture no more?

[12:59:02.0000] <annevk>
will we avoid the monoculture with IndexedDB or is everyone going to use the open source work from Google for Chrome?

[13:01:00.0000] <Ms2ger>
We have an independent implementation AFAIK

[13:01:01.0000] <annevk>
on top of SQLite

[13:01:02.0000] <jamesr>
currently there are sqlite backends for indexdb and leveldb

[13:01:03.0000] <annevk>
teehee

[13:01:04.0000] <jamesr>
so that's 2

[13:01:05.0000] <annevk>
what is leveldb?

[13:01:06.0000] <jamesr>
http://code.google.com/p/leveldb/

[13:01:07.0000] <jamesr>
storage engine

[13:02:00.0000] <annevk>
but it uses sqlite?

[13:02:01.0000] <jamesr>
no

[13:02:02.0000] <jamesr>
it's a storage engine

[13:02:03.0000] <annevk>
i got thrown of by sqlite backend above

[13:03:00.0000] <MacTed>
I don't think you mean "backend" the same as I do...

[13:03:01.0000] <jamesr>
i'm just talking about indexdb

[13:03:02.0000] <MacTed>
...

[13:04:00.0000] <MacTed>
"there are sqlite backends for indexdb and leveldb"

[13:04:01.0000] <MacTed>
please rephrase?

[13:05:00.0000] <MacTed>
I would understand that to mean "sqlite is (or can be) the storage engine used by indexdb and leveldb"

[13:05:01.0000] <jamesr>
that is true

[13:07:00.0000] <MacTed>
so... indexdb and leveldb are 2 masks/wrappers over 1 body/engine (sqlite)

[13:07:01.0000] <MacTed>
how does that avoid monoculture?

[13:08:00.0000] <MacTed>
reading further in http://www.w3.org/TR/webdatabase/, this line seems like it should have been in the read box above it -- "The Web Applications Working Group continues work on two other storage-related specifications: Web Storage and Indexed Database API. "

[13:08:01.0000] <jamesr>
wait

[13:08:02.0000] <jamesr>
sorry i misread

[13:08:03.0000] <jamesr>
sqlite is _not_ the storage engine used by leveldb

[13:09:00.0000] <annevk>
i think what you meant is that indexeddb has two potential backends: sqlite and leveldb

[13:09:01.0000] <jamesr>
right

[13:09:02.0000] <Philip`>
I think the main problem is the API being a non-standardised monoculture (which WebSQL's is, since it incorporates SQLite's SQL syntax and semantics which aren't defined clearly enough for independent implementation)

[13:10:00.0000] <jamesr>
websql had only one possible backend, sqlite

[13:10:01.0000] <annevk>
if we had kept the SQL parsing separate it would have been better

[13:10:02.0000] <annevk>
but nobody ever defined that

[13:11:00.0000] <Philip`>
It's okay if everyone uses the same implementation, as long as the API is standardised and lets them easily use a different implementation if they ever had any reasons to do so

[13:11:01.0000] <dglazkov>
https://plus.google.com/103035368214666982008/posts/43eHRWqsMEP

[13:11:02.0000] <Ms2ger>
dglazkov, will I regret following that link?

[13:11:03.0000] <dglazkov>
define regret

[13:11:04.0000] <Ms2ger>
I.e., is it goatse or W3C process discussion?

[13:11:05.0000] <jamesr>
Ms2ger: is there a difference?

[13:12:00.0000] <MacTed>
I think I get you...

[13:12:01.0000] <MacTed>
OK, so we'll have to go back to the labs and implement IndexedDB-to-ODBC.  I think that's a can-do.

[13:12:02.0000] <jamesr>
Philip`: it is a little dangerous if everyone uses the same impl since in that situation it's very difficult to avoid creating strange dependencies

[13:12:03.0000] <MacTed>
(now looking at http://www.w3.org/TR/IndexedDB/ and http://www.w3.org/TR/webstorage/ )

[13:13:00.0000] <Philip`>
MacTed: Does your ODBC/etc bridge support precisely the same SQL syntax/semantics as SQLite?

[13:15:00.0000] <Philip`>
If it's just the same WebSQL method calls but different SQL then it's not really compatible with WebSQL (people will write web pages depending on the quirks of SQLite so they won't work if you try to swap it out for a different SQL implementation)

[13:18:00.0000] <gsnedders>
Does importScripts work cross-origin?

[13:20:00.0000] <jgraham>
Anybody got any idea what to do when you get ImportError: /home/jgraham/lib/python2.6/lib-dynload/unicodedata.so: undefined symbol: _PyUnicodeUCS4_T\

[13:20:01.0000] <jgraham>
oNumeric

[13:20:02.0000] <gsnedders>
"Attempt to fetch each resource identified by the resulting absolute URLs, from the entry script's origin, with the synchronous flag set." — what about scripts not from the same origin? Silently ignore them.

[13:20:03.0000] <jgraham>
trying to run something under mod_wsgi

[13:20:04.0000] <gsnedders>
jgraham: Looks like you have something compiled for UCS4 and something compiled for UCS2

[13:20:05.0000] <jgraham>
That seems to work fine under normal python

[13:21:00.0000] <jgraham>
With the same executable and the same python path

[13:21:01.0000] <jgraham>
gsnedders: Well I worked that much out

[13:21:02.0000] <jgraham>
I just can't work out *what* it might be

[13:22:00.0000] <gsnedders>
jgraham: mod_wsgi?

[13:22:01.0000] <jgraham>
Well that is the only thing I haven't compiled by hand

[13:22:02.0000] <jgraham>
But it doesn't *say* this should be necessary

[13:22:03.0000] <gsnedders>
Oh, the "from the entry script's origin" is an argument to "fetch". Well that's certainly unclear.

[13:26:00.0000] <AryehGregor>
Can I disable section numbering in anolis?

[13:26:01.0000] <AryehGregor>
I don't want people reporting things by section numbers, since they don't appear in the source.

[13:27:00.0000] <Ms2ger>
.secno { display: none; }

[13:27:01.0000] <AryehGregor>
Blech.

[13:27:02.0000] <Philip`>
Shouldn't you try to make things as easy as possible for reviewers, to maximise the amount of feedback you get?

[13:27:03.0000] <AryehGregor>
They can give the section name.

[13:27:04.0000] <gsnedders>
AryehGregor: Anolis 2 made it easier. :P

[13:27:05.0000] <Philip`>
(even if that means you have to cross-check an old HTML version to match up numbers)

[13:27:06.0000] <Ms2ger>
It probably wouldn't be hard to hack that in

[13:27:07.0000] <AryehGregor>
Giving the section name is about as easy for them and saves me effort.

[13:28:00.0000] <AryehGregor>
I should also change all my ol's to ul's.

[13:33:00.0000] <MacTed>
Philip` - partially true.  anyone who writes to the quirks of SQLite wouldn't be able to move to our bridge, but anyone who didn't want SQLite could start (and stay) with our bridge, whether they then used ODBC standard (recommended) or DBMS-specific dialect

[13:33:01.0000] <Hixie>
AryehGregor: given how much feedback we get, discouraging feedback isn't a real problem :-P

[13:34:00.0000] <MacTed>
Philip` - and if they wrote to ODBC spec, and then wanted to use SQLite ... the bridge still works, with their chosen backend.

[13:34:01.0000] <AryehGregor>
Hixie, I'd love to get more feedback myself.

[13:35:00.0000] <Ms2ger>
Hixie, how much feedback *you* get, you mean ;)

[13:41:00.0000] <jamesr>
lawl

[13:41:01.0000] <jamesr>
now charles pritchard is quoting IRC chatlogs on email on public-canvas-api

[13:41:02.0000] <jamesr>
the circle has completed

[13:46:00.0000] <othermaciej>
did his email get quoted on IRC first?

[13:46:01.0000] <jamesr>
not directly

[13:46:02.0000] <annevk>
Ms2ger, how does this read: "If node is a DocumentFragment node, inserts its children (preserving tree order), before child or at the end of parent if child is null."

[13:46:03.0000] <jamesr>
but the IRC conversation was about public-canvas-api email threads

[13:46:04.0000] <jamesr>
and how many of them were kind of useless

[13:46:05.0000] <annevk>
Ms2ger, child is the new refChild

[13:47:00.0000] <Ms2ger>
s/inserts/insert/?

[13:47:01.0000] <annevk>
if that's it, great

[13:47:02.0000] <Ms2ger>
I think it's fine

[13:48:00.0000] <AryehGregor>
Okay, is there any git hosting site that will serve text/html from my repo without $$$?  I'd prefer to use something like github or gitorious instead of gitweb.cgi here.

[13:49:00.0000] <annevk>
alright

[13:49:01.0000] <annevk>
completed insert node

[13:49:02.0000] <annevk>
now replace node

[13:49:03.0000] <annevk>
then we are pretty close to being able to spec any kind of mutation events

[13:50:00.0000] <jgraham>
AryehGregor: Will github not do that?

[13:50:01.0000] <Ms2ger>
Then, LC

[13:50:02.0000] <The_8472>
AryehGregor... get a virtual server and do it yourself?

[13:50:03.0000] <AryehGregor>
jgraham, AFAICT you need to buy some kind of better account.

[13:50:04.0000] <The_8472>
cloud is overrated

[13:50:05.0000] <AryehGregor>
Maybe I should.

[13:50:06.0000] <AryehGregor>
The_8472, I have a dedicated server that I'm already using.  The problem is, gitweb stinks compared to sites like github.

[13:50:07.0000] <AryehGregor>
Has way fewer features, is uglier, etc.

[13:51:00.0000] <The_8472>
mhm... and no better software available?

[13:51:01.0000] <AryehGregor>
It has syntax highlighting too.

[13:51:02.0000] <AryehGregor>
Nothing nearly as slick as the professional stuff that I'm aware of.

[13:51:03.0000] <jgraham>
Oh, I assumed all those people with github hosted pages were using the free service

[13:52:00.0000] <AryehGregor>
Hmm, does it have syntax highlighting?

[13:52:01.0000] <AryehGregor>
You'd assume so.

[13:52:02.0000] <annevk>
Ms2ger, apparently we're going to spec on*

[13:52:03.0000] <AryehGregor>
But I don't see it.

[13:53:00.0000] <annevk>
Ms2ger, and DOM Range should prolly move in once AryehGregor fixed all its bugs :)

[13:53:01.0000] <Ms2ger>
Sounds good

[13:53:02.0000] <annevk>
DOM Range does a lot of mutation stuff so having that tightly coupled would be good

[13:53:03.0000] <jgraham>
AryehGregor: "All plans come with [...] Wikis, Issues, Downloads, Pages & more"

[13:54:00.0000] <jgraham>
Which sounds like text/html

[13:54:01.0000] <Philip`>
AryehGregor: Use GitHub plus cron on your own server to clone and publish the HTML?

[13:54:02.0000] <AryehGregor>
Philip`, hmm.  Interesting thought.

[13:54:03.0000] <Ms2ger>
Ugh, 22-step algorithms

[13:54:04.0000] <AryehGregor>
jgraham, I mean serving raw text/html of files in the repo.

[13:54:05.0000] <AryehGregor>
Ms2ger, only 22?

[13:54:06.0000] <Ms2ger>
That's the first I noticed

[13:55:00.0000] <AryehGregor>
"Delete the contents" is 32 steps.

[13:55:01.0000] <AryehGregor>
Plus lots of substeps, naturally.

[13:55:02.0000] <AryehGregor>
insertParagraph is also 32.

[13:55:03.0000] <AryehGregor>
Did I ever mention execCommand is complicated?

[13:55:04.0000] <Ms2ger>
Those aren't in DOM Range :)

[13:57:00.0000] <annevk>
back in 30-45min

[13:57:01.0000] <AryehGregor>
Yeah, but the 22-step algorithm in DOM Range was written by me too.  :)

[13:58:00.0000] <Ms2ger>
AryehGregor, is commonAncestorContainer hard or did you just not get around to it?

[13:58:01.0000] <AryehGregor>
Ms2ger, I never got around to it.

[13:59:00.0000] <AryehGregor>
I'm pretty sure it's trivial.

[13:59:01.0000] <Ms2ger>
OK

[13:59:02.0000] <AryehGregor>
var ret = range.startContainer; while (ret != range.endContainer && !isAncestor(ret, range.endContainer)) ret = ret.parentNode; return ret;

[13:59:03.0000] <AryehGregor>
Something like that.

[14:14:00.0000] <jgraham>
Victory!

[14:14:01.0000] <jgraham>
http://test-review.hoppipolla.co.uk/changeset/6e56634565c9/

[14:14:02.0000] <jgraham>
Or in general http://test-review.hoppipolla.co.uk/

[14:14:03.0000] <jgraham>
The UI kinda sucks and there are a few non-trivial problems with the software

[14:14:04.0000] <jgraham>
But it is something

[14:15:00.0000] <jgraham>
I don't think itt's setup so that people can comment at the moment

[14:18:00.0000] <Ms2ger>
502 Bad Gateway

[14:18:01.0000] <Ms2ger>
It rocks!

[14:19:00.0000] <jgraham>
Try again?

[14:19:01.0000] <AryehGregor>
Hixie, there's no longer any plan to merge the editing spec into HTML, right?

[14:19:02.0000] <Ms2ger>
So, how about you test it by reviewing http://test-review.hoppipolla.co.uk/changeset/abb5f54243bf/?

[14:20:00.0000] <jgraham>
Ms2ger: I see what you did there

[14:20:01.0000] <Ms2ger>
:)

[14:20:02.0000] <Ms2ger>
Want me to link you to the spec? :)

[14:21:00.0000] <jgraham>
If you like

[14:22:00.0000] <Ms2ger>
http://www.whatwg.org/specs/web-apps/current-work/multipage/webappapis.html#handler-window-onload

[14:24:00.0000] <AryehGregor>
What's the Mac equivalent of Shift-Enter?

[14:27:00.0000] <jgraham>
Evidently the allow_anon_comments setting doesn't do much :)

[14:28:00.0000] <dglazkov>
option-enter

[14:28:01.0000] <AryehGregor>
dglazkov, thanks.

[14:29:00.0000] <dglazkov>
it actually varies from editor to editor :(

[14:29:01.0000] <AryehGregor>
:(

[14:30:00.0000] <dglazkov>
for key in ['shift','option','command']:

[14:31:00.0000] <AryehGregor>
Hmm, should backspacing immediately after a link unlink?

[14:31:01.0000] <AryehGregor>
I think yes.

[14:31:02.0000] <AryehGregor>
IE9 and Word 2007 do it, but other browsers and OO don't.

[14:32:00.0000] <AryehGregor>
I'll change the spec to match IE and Word, it seems more useful.

[14:33:00.0000] <Ms2ger>
/me touches DOM Range for once

[14:33:01.0000] <dglazkov>
BOOM

[14:34:00.0000] <david_carlisle>
Ms2ger:  on the xml serialialisation bug, might be interested in helping spec that , but no time  until end of august, just wanted to get comments in before LC deadline

[14:34:01.0000] <Ms2ger>
/me whacks dglazkov 

[14:34:02.0000] <dglazkov>
:D

[14:35:00.0000] <Ms2ger>
david_carlisle, Shelly might throw a tantrum over it, but innerHTML probably will be moved out of HTML soon

[14:35:01.0000] <Ms2ger>
And there's no LC in sight for my spec :)

[14:35:02.0000] <AryehGregor>
Into where, DOMPS?

[14:35:03.0000] <david_carlisle>
Ms2ger:  well 'twas me put in a buzilla request asking for that...

[14:35:04.0000] <Ms2ger>
Yeah

[14:36:00.0000] <david_carlisle>
so it would apply to element not htmlelement and mathml would benefit

[14:37:00.0000] <foolip_>
what does innerHTML do outside of HTML?

[14:38:00.0000] <Ms2ger>
Same as it does inside

[14:38:01.0000] <david_carlisle>
foolip if you have a html span with mathml  in it then innerhtml on the span has to, by spec and current implementations serialise the math

[14:38:02.0000] <foolip_>
parse it as HTML and insert?

[14:38:03.0000] <david_carlisle>
but you can't access the seialisation by doing innerhtml on math

[14:38:04.0000] <foolip_>
ah, getting innerHTML, not setting it

[14:38:05.0000] <david_carlisle>
well both

[14:38:06.0000] <foolip_>
setting doesn't magically parse as XML or anything, right?

[14:39:00.0000] <Ms2ger>
No

[14:39:01.0000] <foolip_>
maybe I should just read the spec when I'm bored :)

[14:39:02.0000] <foolip_>
ignore me

[14:39:03.0000] <david_carlisle>
parsing of mathml is fully specified in html(5) parse algorithm

[14:39:04.0000] <gsnedders>
It parses it as XML in XHTML, at lesat.

[14:39:05.0000] <Ms2ger>
Like using <svg> in HTML doesn't magically parse as XML ;)

[14:39:06.0000] <gsnedders>
So it'd make sense to parse as XML in XML.

[14:39:07.0000] <foolip_>
right, silly me

[14:39:08.0000] <david_carlisle>
gsnedders: in application/xml yes not in text/html

[14:40:00.0000] <gsnedders>
david_carlisle: Well, that's not XHTML or XML.

[14:40:01.0000] <david_carlisle>
gsnedders: sorry I'm lost, back up:-)

[14:41:00.0000] <Ms2ger>
text/html \not\in \{XHTML, XML\}

[14:41:01.0000] <david_carlisle>
gsnedders: worry about xml mime types later, immediate concern is mathml in html served as text/html

[14:41:02.0000] <Ms2ger>
And apologies for the LaTeX

[14:41:03.0000] <gsnedders>
david_carlisle: innerHTML parses as XML in XHTML. I made no comment about HTML.

[14:41:04.0000] <gsnedders>
Ms2ger: Can't you hand-write MathML?

[14:41:05.0000] <Ms2ger>
Only with innerHTML

[14:42:00.0000] <david_carlisle>
Ms2ger:http://www.w3.org/Bugs/Public/show_bug.cgi?id=11204

[14:42:01.0000] <david_carlisle>
LaTeX, what's that?

[14:42:02.0000] <Ms2ger>
"Firefox doesn't know how to open this address, because the protocol (ms2ger) isn't associated with any program."

[14:43:00.0000] <david_carlisle>
sorry firefox (well chatzilla) thought it would be clever and stick your name at front, probably i caught the tab key or something:-)

[14:43:01.0000] <Ms2ger>
:)

[14:44:00.0000] <annevk>
sounds like a bug in Firefox

[14:45:00.0000] <Ms2ger>
annevk, of course not, there are no bugs in Firefox

[14:54:00.0000] <Michael>
^^

[15:34:00.0000] <AryehGregor>
/me seriously does not understand why stuff like http://www.w3.org/Bugs/Public/show_bug.cgi?id=13431 has to mention anything about accessibility at all

[15:40:00.0000] <TabAtkins_>
AryehGregor: That's explained.  People in the a11y TF came up with it, so they decided to phrase it in terms of a11y.

[15:41:00.0000] <AryehGregor>
Yes, but I don't understand how the explanation makes sense given that it really manifestly has nothing at all to do with a11y.

[15:41:01.0000] <AryehGregor>
But it's good feedback in this case anyway.

[15:46:00.0000] <annevk>
it's just missing "e.g. "

[15:48:00.0000] <annevk>
it's good feedback, but to report a small typo they could have done with a somewhat simpler bug report

[15:49:00.0000] <AryehGregor>
Yeah, but they're used to getting their bug reports closed NEEDSINFO with demands for explanation and use cases, so I don't blame them.

[16:01:00.0000] <Hixie>
AryehGregor: define "plan"

[16:01:01.0000] <Hixie>
AryehGregor: i don't think long-term that we should have execCommand specced in a different spec than HTMLDocument

[16:02:00.0000] <Hixie>
AryehGregor: then again on the long term I don't think we should have HTMLDocument specced in a different spec than Document either

[16:02:01.0000] <AryehGregor>
Hixie, you don't think we should have more than one web spec in the long term.  :)

[16:02:02.0000] <AryehGregor>
(except maybe orthogonal stuff like HTML vs. SVG)

[16:02:03.0000] <Hixie>
i expect i'll merge the editing stuff into HTML in the next year or two

[16:02:04.0000] <AryehGregor>
Even if I'm still actively editing it?

[16:03:00.0000] <Hixie>
if you're actively editing it, i'm happy to delay the merge

[16:03:01.0000] <annevk>
AryehGregor, that is not completely true; Hixie wants multiple specs, just not that small ;)

[16:04:00.0000] <AryehGregor>
Given that we're going to have multiple people editing significant things indefinitely, don't you think it makes more sense to have separate specs that reference each other?  You're going to have somewhat different stylistic conventions and so on, I don't think it makes sense to actually merge everything.

[16:04:01.0000] <AryehGregor>
The editing stuff is pretty orthogonal to the rest of HTML.

[16:05:00.0000] <annevk>
I think we should try to converge on stylistic behavior when possible

[16:05:01.0000] <AryehGregor>
I don't think we should bother, unless it's likely to confuse anyone.

[16:05:02.0000] <annevk>
Makes it easier for people to read specs

[16:05:03.0000] <AryehGregor>
Otherwise we're reinventing pubrules.

[16:05:04.0000] <annevk>
style guide is way different from pubrules

[16:05:05.0000] <AryehGregor>
Different people will have different styles, it's not something that we should waste time negotiating about if different editors disagree.

[16:06:00.0000] <gsnedders>
annevk: At what level?

[16:06:01.0000] <annevk>
pubrules is about boilerplate

[16:06:02.0000] <gsnedders>
Someone using en-gb and someone using en-us is mostly irrelevant, for example.

[16:06:03.0000] <annevk>
sure

[16:06:04.0000] <Hixie>
AryehGregor: well as an example, i intend to work with ms2ger so that the common text in dom core and html actually is merged in from one common source

[16:07:00.0000] <Hixie>
annevk: ok dreamhost said they can do this co-hosting thing

[16:07:01.0000] <AryehGregor>
I mean yeah, let's all use the same stylesheets and same basic approach and not use conflicting terminology or anything, but if I want to xref everything as direct links and not have a references section, or want to xref more or fewer common terms than other people, or want to use <ul> instead of <ol> because otherwise people will report bugs using ever-changing step numbers that aren't in the source, I don't think we should worry about it.

[16:07:02.0000] <annevk>
sweet, so I just claim a subdomain and it works?

[16:07:03.0000] <annevk>
i guess that would be pretty bad

[16:07:04.0000] <AryehGregor>
Hixie, why isn't it just in DOM Core and referenced from HTML?

[16:08:00.0000] <annevk>
not everything is needed for DOM Core

[16:08:01.0000] <Hixie>
annevk: we have to both agree

[16:08:02.0000] <Hixie>
annevk: and then tell them

[16:08:03.0000] <Hixie>
annevk: so what do you want, dom.spec.whatwg.org?

[16:08:04.0000] <Hixie>
AryehGregor: because that makes HTML a huge pain to read

[16:08:05.0000] <AryehGregor>
Hixie, how so?

[16:09:00.0000] <Hixie>
AryehGregor: i don't have the details at hand, but basically html right now says "a, b, c, d" and dom core says "b, d" and each of these is like one sentence long

[16:09:01.0000] <annevk>
Hixie, that makes the most sense I think

[16:09:02.0000] <Hixie>
AryehGregor: it takes just as much to point the reader to the other spec as to just say it

[16:10:00.0000] <Hixie>
AryehGregor: plus it gets confusing having a and c in one spec yet have to read another for b and d when they are all so similar concepts

[16:10:01.0000] <AryehGregor>
Yeah, for minor things like that duplication makes sense.

[16:10:02.0000] <Hixie>
none of this is anything but definitions and so on

[16:10:03.0000] <AryehGregor>
But if there's a huge chunk that's orthogonal to the rest of the spec, I don't think it makes sense to try having it be part of the same spec.

[16:11:00.0000] <Hixie>
i took out all those bits yesterday

[16:11:01.0000] <Hixie>
(execCommand() isn't orthogonal though)

[16:11:02.0000] <AryehGregor>
Unless the same person winds up editing it.  Obviously if you wind up editing the editing stuff, then you'll want to merge it into your spec.

[16:12:00.0000] <AryehGregor>
How is it not orthogonal?

[16:12:01.0000] <Hixie>
it's HTML's editing model

[16:13:00.0000] <Hixie>
it's less orthogonal than, say, PeerConnection, which i also think belongs in HTML

[16:14:00.0000] <gsnedders>
/me thinks anything that has only one-way references should be a separate spec

[16:15:00.0000] <Hixie>
annevk: what's your dreamhost id?

[16:15:01.0000] <Hixie>
annevk: or dreamhost e-mail address

[16:15:02.0000] <Hixie>
annevk: i'll cc you

[16:23:00.0000] <annevk>
annevankesteren⊙gc

[16:24:00.0000] <annevk>
yay for dglazkov

[16:24:01.0000] <dglazkov>
I win!

[16:24:02.0000] <dglazkov>
what did I win?

[16:24:03.0000] <annevk>
writing the spec

[16:25:00.0000] <dglazkov>
that doesn't seem like a win

[16:26:00.0000] <dglazkov>
spec-writing for me is like squeezing water out of a stone.

[16:28:00.0000] <hober>
AryehGregor: yes, sorry, will get back to you today or tomorrow on that

[16:29:00.0000] <AryehGregor>
hober, thanks.

[16:33:00.0000] <Hixie>
foolip_: if things are marked with class=impl incorrectly, you can file that as a regular bug in bugzilla

[16:33:01.0000] <foolip_>
Hixie, ok, so class=impl is what it is

[16:33:02.0000] <foolip_>
will do

[16:33:03.0000] <annevk>
dglazkov, haha

[16:34:00.0000] <annevk>
dglazkov, without a spec it's kind of hard to see how Shadow DOM differs from XBL

[16:35:00.0000] <annevk>
and how it works :)

[16:35:01.0000] <hober>
AryehGregor: which list did that email go to?

[16:35:02.0000] <dglazkov>
annevk: writing this up now, btw

[16:35:03.0000] <AryehGregor>
hober, public-html, since I wanted feedback from Microsoft.

[16:36:00.0000] <annevk>
I have a vague idea and wonder how closely it needs to be defined together with DOM Core given how it affects that, but I'd like to know a little more

[16:39:00.0000] <annevk>
oh yes

[16:39:01.0000] <annevk>
done: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-node-insert

[16:40:00.0000] <annevk>
Ms2ger is sleeping already of course, but he can look tomorrow

[16:42:00.0000] <Hixie>
foolip_: so this algorithm was left in because it's used to define the validator requirements

[16:42:01.0000] <Hixie>
foolip_: any suggestions on getting around that?

[16:43:00.0000] <foolip_>
foolip_, well, I just implemented the validator bits and I wasn't looking at the developer version, so I don't think it's a practical problem

[16:43:01.0000] <foolip_>
you could try to express the validity constraints in prose instead, as it's rather obscure right now

[16:44:00.0000] <foolip_>
oops, talking to myself :) Hixie ^

[16:44:01.0000] <Hixie>
yeah, dunno

[16:45:00.0000] <foolip_>
the best I could come up with for an error message for "microdata error" was "The itemref attribute contained redundant references", which is neither completely true nor clear

[16:46:00.0000] <Hixie>
yeah coming up with good prose was hard, which is why the spec says what it does

[16:46:01.0000] <Hixie>
i can just hide these requirements from the developers version i guess

[16:46:02.0000] <foolip_>
that would solve the problem at hand :)

[16:58:00.0000] <benschwarz>
oh hai you two

[16:58:01.0000] <Hixie>
foolip_, benschwarz: fixed

[16:59:00.0000] <foolip_>
benschwarz, quick, make it work :P

[16:59:01.0000] <benschwarz>
I just saw.


2011-07-29
[17:03:00.0000] <benschwarz>
foolip_: check my last two commits - https://github.com/benschwarz/developers.whatwg.org/commits/master/

[17:04:00.0000] <foolip_>
that's more like it :)

[17:04:01.0000] <benschwarz>
also, its been deployed ;)

[17:04:02.0000] <foolip_>
hmm, I can't see any difference

[17:05:00.0000] <foolip_>
neither on the splitting or on the class=impl, whichever it was you fixed

[17:05:01.0000] <benschwarz>
double refresh, app cache is playing with you

[17:05:02.0000] <benschwarz>
(i think)

[17:06:00.0000] <benschwarz>
Hixie, foolip_, the spec splitter has split microdata to its own page, but for some reason its getting an extra wrapper div that the others are not.

[17:06:01.0000] <benschwarz>
compare the "5 microdata" header on http://developers.whatwg.org/microdata.html#microdata

[17:07:00.0000] <foolip_>
benschwarz, I think that's just what the source looks like

[17:07:01.0000] <benschwarz>
vs http://developers.whatwg.org/browsers.html#browsers

[17:07:02.0000] <foolip_>
to organize some stuff for W3C purposes I guess

[17:07:03.0000] <foolip_>
I had to fix a bug in the spec splitter related to it, I presume you got that with the pull as well

[17:10:00.0000] <benschwarz>
yeah… but its inconsistant. Anything we can do here @Hixie ?

[17:10:01.0000] <benschwarz>
http://cl.ly/330I1x3m0d2e1u0A3P3J

[17:10:02.0000] <foolip_>
benschwarz, does the extra nesting confuse your CSS?

[17:11:00.0000] <foolip_>
looks like it

[17:14:00.0000] <benschwarz>
yeah. sadly I had to use a > selector for those headers

[17:15:00.0000] <Hixie>
what's the problem?

[17:16:00.0000] <benschwarz>
Hixie: the microdata header has a extra wrapper div that other major sections don't have

[17:16:01.0000] <dglazkov>
hsivonen: are we friends? I can't decide which circle to add you to.

[17:17:00.0000] <Hixie>
odd

[17:17:01.0000] <Hixie>
oh the <div data-component> div?

[17:17:02.0000] <benschwarz>
yeah

[17:17:03.0000] <Hixie>
yeah not much i can do about that

[17:17:04.0000] <Hixie>
can you have your script strip it out?

[17:17:05.0000] <benschwarz>
if we need to

[17:19:00.0000] <annevk>
Ms2ger, https://bitbucket.org/ms2ger/dom-range/changeset/d3f907fe0dd7 looks kind of bogus (the text does not match the title values and e.g. previous sibling is different from preceding siblings)

[17:28:00.0000] <hober>
AryehGregor: email is forthcoming, but basically we much prefer option 3 to option 2, and aren't interested in option 1.

[17:31:00.0000] <Hixie>
benschwarz: these divs appear all over hte spec, i think it's just luck that this is the only one you've run into

[17:31:01.0000] <Hixie>
benschwarz: most of the others only appear in parts of the spec not included in developers.whatwg.org source material

[17:31:02.0000] <Hixie>
benschwarz: like websockets, webstorage, etc

[17:37:00.0000] <benschwarz>
Hixie: bad luck I guess :)

[18:12:00.0000] <Hixie>
annevk: looks like it's all up to your now. :-)

[18:12:01.0000] <annevk>
yeah

[18:13:00.0000] <annevk>
hopefully I can work with Ms2ger on it next week

[18:15:00.0000] <annevk>
kind of funny how it puts it under another domain I own that is completely unrelated

[18:15:01.0000] <Hixie>
heh

[18:18:00.0000] <Hixie>
anyone got a better suggestion for a name for the "Scripted-submit" flag?

[18:18:01.0000] <Hixie>
it's set when you use the submit() method

[18:18:02.0000] <TabAtkins_>
Hm, I thought Chrome had added [].foreach already. :/

[18:19:00.0000] <TabAtkins_>
RAGE (╯°□°)╯︵ ┻━┻

[18:19:01.0000] <Hixie>
maybe "submitted from method"

[18:19:02.0000] <hober>
script-triggered-submit? submitted-from-script? no, those kind of suck

[18:19:03.0000] <Hixie>
or "submitted from script() method"

[18:19:04.0000] <Hixie>
i guess it doesn't have to be short

[18:20:00.0000] <annevk>
"submit() flag"

[18:20:01.0000] <annevk>
XHR has a "send() flag"

[18:20:02.0000] <annevk>
though kind of different purpose

[18:20:03.0000] <Hixie>
i'll got with "submitted from script() method" for now

[18:20:04.0000] <annevk>
script() is not a method

[18:20:05.0000] <Hixie>
er

[18:20:06.0000] <Hixie>
submit()

[18:20:07.0000] <Hixie>
i got it right in the spec text :-P

[19:17:00.0000] <Hixie>
man, opera's support for <datalist> is... not per spec

[22:11:00.0000] <MikeSmith>
"CSE HTML Validator generates a message about missing meta description tags"

[22:14:00.0000] <MikeSmith>
“CSE enforces better style by letting you know when "td" and "tr" tags are not closed. Even though omitting these end tags is technically allowable in HTML 4.01, it is not recommended because it may cause rendering problems in some browsers”

[22:14:01.0000] <MikeSmith>
solid gold man

[22:38:00.0000] <erlehmann>
>Download Trial

[22:38:01.0000] <erlehmann>
>Buy Now

[22:38:02.0000] <erlehmann>
:3

[22:38:03.0000] <erlehmann>
Oh, and …

[22:39:00.0000] <erlehmann>
>Try CSE HTML Validator Pro free for 200 validations or 30 days, whichever comes first.

[22:39:01.0000] <erlehmann>
COMEDY GOLD

[22:39:02.0000] <erlehmann>
let me unpack the strategic popcorn reserves.

[04:16:00.0000] <jgraham>
foolip: I read machineval and "machin eval" and wonder what a machin is…

[04:16:01.0000] <jgraham>
I think this suggests industrial injury caused by over-exposure to js

[04:36:00.0000] <hsivonen>
Does this mean that TimBL doesn't approve of the idea of not having Microdata spec say how Microdata maps to RDF? http://lists.w3.org/Archives/Public/www-tag/2011Jul/0120.html

[04:52:00.0000] <erlehmann>
hsivonen, is that a rhetorical question?

[04:59:00.0000] <hsivonen>
erlehmann: no. I was semi-expecting Tim to be OK with Microdata not defining a mapping to RDF

[04:59:01.0000] <hsivonen>
erlehmann: since then Microdata would step on RDF's toes less than by defining a mapping that sucks

[05:00:00.0000] <jgraham>
Well he seems to want it to define a mapping that doesn't suck

[05:00:01.0000] <erlehmann>
i should probably learn about the reasons why a mapping would have to suck

[05:00:02.0000] <jgraham>
I wonder if a concrete proposal for such a mapping exists

[05:02:00.0000] <hsivonen>
I think it would be happier for everyone if a group of people who want to use such a mapping drove the definition of the mapping

[05:02:01.0000] <hsivonen>
instead of Hixie slapping together something when Hixie obviously isn't going to use the mapping

[05:07:00.0000] <erlehmann>
sounds reasonable.

[05:17:00.0000] <hsivonen>
TabAtkins: in a presentation of yours you said that new PDF readers don't support old PDF features

[05:17:01.0000] <hsivonen>
TabAtkins: any concrete examples?

[05:18:00.0000] <hsivonen>
TabAtkins: (I'm assuming that archivists aren't using the insecure parts of PDF that allow applicaiton invocation)

[05:19:00.0000] <hsivonen>
TabAtkins: IIRC, the only feature that PDF has ever dropped was LZW compression, but LZW *de*compression is still supported

[05:20:00.0000] <hsivonen>
hmm. Actually, I recall once seeing a batch of legacy PDFs whose fonts only worked in Adobe Reader--not in Evince

[06:48:00.0000] <hsivonen>
Hixie: is http://www.jenitennison.com/blog/node/161 on your feedback radar?

[07:07:00.0000] <erlehmann>
hsivonen, that is a perfect trolling opportunity to demand namespaces, again! :D

[07:08:00.0000] <Ms2ger>
Don't we have enough trolling opportunities already?

[08:48:00.0000] <karlcow>
/me proposed text for the itemid http://www.w3.org/Bugs/Public/show_bug.cgi?id=13452#c2

[08:57:00.0000] <smaug____>
is there a coding style for webidl interfaces?

[09:00:00.0000] <Ms2ger>
Yes :)

[09:00:01.0000] <Ms2ger>
</unhelpful>

[09:01:00.0000] <smaug____>
bah

[09:01:01.0000] <Ms2ger>
Two-space indentation for everything except for read-write attributes, those are on "  readonly ".length

[09:01:02.0000] <smaug____>
how to handle {}

[09:02:00.0000] <Ms2ger>
Not on a new line

[09:02:01.0000] <smaug____>
looks like HTML spec uses different coding style than gecko

[09:03:00.0000] <Ms2ger>
Much less indentation for methods, yes

[09:03:01.0000] <Ms2ger>
s/indentation/spaces in the middle of the line/

[09:16:00.0000] <hsivonen>
/me subscribes to www-tag

[09:17:00.0000] <jgraham>
Poor you

[09:21:00.0000] <jgraham>
(I suppose the less cynical respose is "I hope this means that you are considering standing for election; the general usefulness of the tag might be increased if it wasn't almost-entirely composed of linked data theoreticians rather than people with an interst in improving the actual web)

[09:21:01.0000] <jgraham>
s/)/")/

[09:24:00.0000] <Ms2ger>
But Linked Data is the actual web!

[09:24:01.0000] <Ms2ger>
/me goes back to his cave

[10:10:00.0000] <Hixie>
hsivonen: it was, but i didn't see anything actionable in a brief scan. Anything specifically I should be looking at?

[10:29:00.0000] <hsivonen>
Hixie: addressing the use case of supporting both SEO and extraction to calendar for event data

[10:30:00.0000] <Hixie>
you mean in a situation where you're targetting two UAs who use different vocabularies for the same thing?

[10:31:00.0000] <Hixie>
i don't see any good way to do that

[10:31:01.0000] <Hixie>
does anyone actually do the extraction to calendar using microdata?

[10:31:02.0000] <Hixie>
if they just use iCal then that might be the better way to go

[10:32:00.0000] <hsivonen>
Hixie: I'm not aware of anyone actually doing calendar extraction, but you wrote a spec for it, no?

[10:32:01.0000] <Hixie>
i wrote a spec for it assuming anyone would implement it

[10:32:02.0000] <hsivonen>
Hixie: yeah, targeting two UAs that use different vocabularies for the "same" thing on the high level but different in details

[10:32:03.0000] <Hixie>
if tehy don't implement it, it's kind of academic

[10:33:00.0000] <hsivonen>
Hixie: isn't it too early to tell?

[10:33:01.0000] <hsivonen>
Hixie: but it's not too early to tell that Google like their own thing more than iCalendar for the SEO use case

[10:35:00.0000] <Philip`>
Could support multiple independent microdata layers, like

[10:35:01.0000] <Hixie>
realistically i'd be very surprised if anyone ever did calendar extraction with microdata

[10:35:02.0000] <Philip`>
<li itemscope[0] itemtype[0]="http://schema.org/Event" itemid[0]="/2011/oscon/" itemscope[1] itemtype[1]="http://microformats.org/profile/hcalendar#vevent" itemid[1]="/2011/oscon/">

[10:35:03.0000] <Hixie>
but if they do we can fix the problem then

[10:36:00.0000] <hsivonen>
Hixie: why don't you zap the spec text that suggests it, then?

[10:37:00.0000] <Hixie>
it provides a guide to people who want to write their own vocabularies

[10:38:00.0000] <Hixie>
plus, it does solve the problem if anyone ever does want to solve it

[10:38:01.0000] <hsivonen>
ok

[10:39:00.0000] <Hixie>
looks like sicking wants us to drop the <datagrid>-fallback-to-<select> feature

[10:39:01.0000] <Hixie>
anyone have any opinions on that?

[10:40:00.0000] <Michael>
If I knew who "us" was maybe

[10:40:01.0000] <Michael>
missing context :D

[10:40:02.0000] <hsivonen>
I thought the feature made sense

[10:40:03.0000] <Hixie>
us = the web

[10:40:04.0000] <Ms2ger>
The spec

[10:40:05.0000] <Ms2ger>
Or that

[10:40:06.0000] <hsivonen>
but I don't know if it is actually used by authors

[10:40:07.0000] <hsivonen>
or if it's too soon to analyze that

[10:41:00.0000] <Michael>
Is the fallback in a library like Modernizr or?

[10:41:01.0000] <Michael>
oh in the spec?

[10:41:02.0000] <Michael>
seems like a weird thing to have in a spec

[10:42:00.0000] <Ms2ger>
What, features?

[10:42:01.0000] <Michael>
A fallback

[10:42:02.0000] <Michael>
I thought fallbacks were the jobs of 3rd party developers

[10:42:03.0000] <hsivonen>
Michael: how do you mean?

[10:43:00.0000] <Michael>
I think I'm still missing context.

[10:43:01.0000] <Michael>
I'm going to hush

[10:44:00.0000] <Michael>
hsivonen, I just saw "fallback from data-grid to select"

[10:44:01.0000] <Michael>
and thought "Shouldn't that be the job of a 3rd party tool like modernizr?"

[10:44:02.0000] <Michael>
But I don't know what spec you're talking about or really even who you guys are other than being related to whatwg.org

[10:45:00.0000] <hsivonen>
Michael: s/grid/list/

[10:45:01.0000] <Michael>
sorry

[10:46:00.0000] <hsivonen>
Michael: the fallback is easy to make work without any JS at all

[10:46:01.0000] <Michael>
Are you talking about the html5 spec?

[10:46:02.0000] <hsivonen>
Michael: yes

[10:46:03.0000] <hsivonen>
Michael: we also have scriptless fallback from <video> to Flash

[10:46:04.0000] <Michael>
interesting

[10:47:00.0000] <Philip`>
Michael: The second example in http://whatwg.org/html#the-list-attribute where it uses <select> for non-scripted fallback

[10:47:01.0000] <hsivonen>
though it sucks somewhat due to the lack of a baseline codec for video

[10:47:02.0000] <hsivonen>
"Unless this bug is substantially addressed, then I will continue to reopen it

[10:47:03.0000] <hsivonen>
until it is elevated to an Issue recorded against HTML5 for resolution by the

[10:47:04.0000] <hsivonen>
WG as a whole."

[10:48:00.0000] <hsivonen>
should I point out that that's not how the Process works?

[10:48:01.0000] <Michael>
Philip`, It makes sense. I suppose I'm just not used to seeing fallbacks in specifications

[10:48:02.0000] <Michael>
What is Sicking's reason for wanting it removed?

[10:48:03.0000] <Ms2ger>
hsivonen, no

[10:49:00.0000] <Ms2ger>
People who want to process-troll should first read the process

[10:49:01.0000] <Hixie>
Ms2ger++

[10:50:00.0000] <Philip`>
Michael: HTML has always been designed to allow pages to use new features with relatively graceful degradation in old browsers that don't support them, like using <script><!-- ... --></script> to hide the script text from pre-<script> browsers

[10:50:01.0000] <Michael>
good point!

[10:50:02.0000] <Michael>
I didn't know that was in the spec. I thought that was a 'hack' someone found

[10:50:03.0000] <Michael>
/me sits

[10:50:04.0000] <Ms2ger>
Philip`, I hope you're not trying to call that one a success :)

[10:51:00.0000] <Hixie>
hey it did work

[10:51:01.0000] <Hixie>
even if it caused us undue pain over the yeras

[10:51:02.0000] <Philip`>
(http://dev.w3.org/html5/html-design-principles/#degrade-gracefully)

[10:51:03.0000] <adactio>
I woud add to that: the design of the datalist element (and the way that it can contain a select element for older UAs) is one of the best examples of designing for backwards compatibility that I've seen.

[10:51:04.0000] <Ms2ger>
I guess the </script> handling was the real problem

[10:52:00.0000] <adactio>
(I wrote a bit about the design of datalist here: http://adactio.com/journal/4272/ )

[10:54:00.0000] <Philip`>
Ms2ger: It was presumably successful at getting people to use scripts without worrying about how it looked all ugly in Netscape 1, given how many people started using scripts :-)

[11:06:00.0000] <dglazkov>
good morning, Whatwg!

[11:09:00.0000] <Ms2ger>
Evening dglazkov

[11:09:01.0000] <Ms2ger>
Is it me or are you always getting on later? :)

[11:10:00.0000] <dglazkov>
early meeting, just now got to the keyboard

[11:11:00.0000] <dglazkov>
<insert excuse>

[11:18:00.0000] <webben>
Hixie: I'm confused ... I thought you'd given up on speccing datagrid for the moment anyway?

[11:19:00.0000] <Hixie>
datalist, not datagrid

[11:19:01.0000] <Hixie>
i misspoke

[11:20:00.0000] <Michael>
^^ my fault

[11:21:00.0000] <Ms2ger>
^^ Hixie's fault for inventing those names :)

[11:21:01.0000] <Michael>
lol

[11:21:02.0000] <Michael>
btw - thank you guys for writing specs :)

[11:21:03.0000] <Michael>
Disney has embraced HTML5 and we're all loving it.

[11:22:00.0000] <Hixie>
i've kind of given up on datagrid for now

[11:22:01.0000] <Hixie>
waiting for dglazkov to finish fixing xbl

[11:25:00.0000] <dglazkov>
/me is procrastinating writing Python code

[11:26:00.0000] <Ms2ger>
Hey, python is fun!

[11:26:01.0000] <Michael>
*cough*ruby

[11:26:02.0000] <Michael>
jk python is also great

[11:28:00.0000] <erlehmann>
Michael, has disney also embraced open standards?

[11:28:01.0000] <Michael>
For our software? No

[11:29:00.0000] <Michael>
We have envious competitors that we have to protect ourselves from unfortunately

[11:30:00.0000] <Michael>
We did release TEA as open source years back but closed it. Now TEA is deprecated and being replaced.

[11:31:00.0000] <erlehmann>
what is TEA?

[11:31:01.0000] <Michael>
A terrible view template language based on Java

[11:32:00.0000] <Michael>
The idea was to limit developers from low level functions like accessing the filesystem etc

[11:32:01.0000] <Michael>
http://www.pdmfc.com/tea-site/info/index.html

[11:32:02.0000] <erlehmann>
Michael, i meant with regards to A/V codecs in HTML5. last i checked, i lived in a country where firefox is the majority browser.

[11:32:03.0000] <erlehmann>
>best features from Scheme, Tcl and Java

[11:32:04.0000] <erlehmann>
heh :D

[11:32:05.0000] <Michael>
haha

[11:33:00.0000] <remysharp>
if I want to make use of setCustomValidity should I be setting on form.onsubmit or is there an event before the submit?

[11:33:01.0000] <remysharp>
(I'm getting mixed results between chrome and opera)

[11:33:02.0000] <Michael>
erlehmann, I'm honestly not sure. That's a good question

[11:33:03.0000] <annevk>
remysharp, onblur maybe?

[11:34:00.0000] <annevk>
kind of depends on the UI you want I guess

[11:34:01.0000] <remysharp>
annevk: say I just want to set the email validation fail message to "that's wrong"

[11:34:02.0000] <remysharp>
annevk: I'm doing that onblur? that seems...weird...

[11:36:00.0000] <remysharp>
annevk: in fact, it needs to be oninput instead of onblur, because user can hit enter

[11:36:01.0000] <remysharp>
annevk: but then I'm setting the custom validity message on every key stroke

[11:36:02.0000] <remysharp>
that feels really wrong.

[11:37:00.0000] <erlehmann>
remysharp, why not set it if it is not set?

[11:37:01.0000] <erlehmann>
hehe ;)

[11:37:02.0000] <erlehmann>
(i know, i know.)

[11:37:03.0000] <remysharp>
erlehmann: yeah.... you're going to have to clarify "it if it is not" bit :-\

[11:37:04.0000] <annevk>
remysharp, it could be submit and blur

[11:38:00.0000] <Rik`>
remysharp: onformsubmit should be ok, what's the problem?

[11:38:01.0000] <remysharp>
annevk: so chrome doesn't get to the submit event if the fields are invalid - so I can't set at that point

[11:38:02.0000] <remysharp>
so what I'm trying to figure out is if chrome is wrong, or if I'm wrong

[11:38:03.0000] <remysharp>
(and normally - the browsers are 'righter' than me!)

[11:39:00.0000] <annevk>
remysharp, oh yeah

[11:39:01.0000] <remysharp>
working example: http://jsbin.com/icejow/3/edit

[11:39:02.0000] <annevk>
remysharp, Chrome is correct

[11:39:03.0000] <remysharp>
annevk: okay, I suspected so

[11:40:00.0000] <remysharp>
which means, really, I need to hook a pre-submit event

[11:40:01.0000] <remysharp>
otherwise I'm listening for every keystroke

[11:40:02.0000] <remysharp>
every keystroke example: http://jsbin.com/icejow/4/edit#html,live - obviously works - but feels utterly wrong

[11:40:03.0000] <Rik`>
remysharp: onformsubmit and oninvalid ?

[11:41:00.0000] <remysharp>
Rik`: there's an oninvalid event?

[11:41:01.0000] <remysharp>
that's what I'm looking for then - one mo /me tests...

[11:41:02.0000] <annevk>
oh yeah there is

[11:41:03.0000] <remysharp>
hmm

[11:41:04.0000] <annevk>
and the event is called invalid

[11:42:00.0000] <annevk>
one day I will write that terminology blog post

[11:42:01.0000] <remysharp>
yeah, that's not flying

[11:42:02.0000] <remysharp>
might be my test - but I doubt it..: http://jsbin.com/icejow/5/edit#html,live

[11:43:00.0000] <Ms2ger>
Which?

[11:43:01.0000] <remysharp>
which what?

[11:43:02.0000] <Ms2ger>
(@annevk)

[11:43:03.0000] <Rik`>
remysharp: addEventListener('invalid')

[11:44:00.0000] <remysharp>
Rik`: ah - no form.oninvalid love.

[11:44:01.0000] <Rik`>
I don't know, I'm guessing :)

[11:44:02.0000] <remysharp>
Rik`: ah - I thought you were stating

[11:44:03.0000] <remysharp>
then no - that didn't fly either

[11:46:00.0000] <Rik`>
remysharp: works for me

[11:46:01.0000] <remysharp>
Rik`: in chrome?

[11:46:02.0000] <Rik`>
yeah

[11:46:03.0000] <Rik`>
but chrome dev

[11:46:04.0000] <remysharp>
Rik`: which build?

[11:46:05.0000] <remysharp>
frackin' dev builds :(

[11:46:06.0000] <Rik`>
14.0.835.8

[11:47:00.0000] <remysharp>
what's annoying is chrome has the oninvalid event listener - it just doesn't look like it's firing on 12.

[11:47:01.0000] <remysharp>
okay, if that's a bug - oninvalid sounds like the right place to be listening

[11:48:00.0000] <remysharp>
Rik`: you get the pirate speak error message?

[11:48:01.0000] <Rik`>
yes

[11:48:02.0000] <annevk>
Ms2ger, like tag vs element, but then event vs event handler, event type vs event object

[11:48:03.0000] <Rik`>
remysharp: oh no, right

[11:48:04.0000] <Ms2ger>
Please do

[11:49:00.0000] <Rik`>
remysharp: but does chrome support setCustomValidity ?

[11:49:01.0000] <remysharp>
Rik`: yeah, it does

[11:50:00.0000] <remysharp>
okay, I just think oninvalid is broken in Chrome - doesn't ever seem to fire

[11:50:01.0000] <Rik`>
that might be true

[11:50:02.0000] <Rik`>
I just hate Chrome's implementation of webforms…

[11:51:00.0000] <remysharp>
well, Opera has them licked, Firefox seems not to care, webkit gets what Chrome has, and Safari's release cycle is so fucking slow I stopped caring. It's a bit of a ball ache really.

[11:52:00.0000] <Ms2ger>
remysharp, look better at Fx

[11:52:01.0000] <Rik`>
yeah Firefox has the best non-broken implementation in my mind

[11:52:02.0000] <Rik`>
but there's not a lot of APIs

[11:52:03.0000] <Rik`>
(and I might be biased)

[11:52:04.0000] <hober>
/me points out that Safari just released...

[11:52:05.0000] <remysharp>
which Firefox? I'm on 5

[11:52:06.0000] <Rik`>
remysharp: 5

[11:53:00.0000] <Rik`>
well 4 also

[11:54:00.0000] <remysharp>
really? maybe I've been overlooking Fx - well - at least they have the same bug chrome have regarding this oninvalid

[11:55:00.0000] <Rik`>
volkmar: ping volkmar for forms discussion ;)

[11:56:00.0000] <remysharp>
anyway, cheers. gotta file some bugs then I guess.

[11:56:01.0000] <volkmar>
remysharp: what bug do we have with the invalid event?

[11:57:00.0000] <Rik`>
remysharp: volkmar implemented webforms in Firefox

[11:57:01.0000] <volkmar>
Rik`: webforms 2

[11:57:02.0000] <volkmar>
i don't want to take any credits for the old ones :D

[11:57:03.0000] <remysharp>
volkmar: I'm right in saying that Fx doesn't support: date, datetime, time, month, etc, range, color?

[11:58:00.0000] <volkmar>
remysharp: yes

[11:58:01.0000] <remysharp>
but it does have datalist?

[11:58:02.0000] <volkmar>
remysharp: though, you will have a hard time finding a good implementation of those...

[11:58:03.0000] <volkmar>
remysharp: datalist/list, yes

[11:58:04.0000] <remysharp>
and no number support

[11:58:05.0000] <remysharp>
I guess I'm not really asking

[11:59:00.0000] <remysharp>
can I assume they're making their way in?

[11:59:01.0000] <volkmar>
remysharp: number is nearly ready

[11:59:02.0000] <remysharp>
cool

[12:00:00.0000] <remysharp>
and oninvalid doesn't work?: http://jsbin.com/icejow/5/edit

[12:02:00.0000] <volkmar>
remysharp: could you explain me why you are doing that:

[12:03:00.0000] <volkmar>
http://jsbin.com/icejow/5/edit

[12:03:01.0000] <volkmar>
oups

[12:03:02.0000] <volkmar>
email.setCustomValidity('');

[12:03:03.0000] <remysharp>
to reset the error flag that gets set on setCustomValidity('foo')

[12:03:04.0000] <volkmar>
if (!email.validity.valid) {

[12:04:00.0000] <volkmar>
but setCustomValidity makes the element valid

[12:04:01.0000] <volkmar>
and then you check for it being invalid

[12:04:02.0000] <remysharp>
yep

[12:04:03.0000] <volkmar>
anyway

[12:04:04.0000] <volkmar>
actually, i'm in a herry

[12:04:05.0000] <volkmar>
brb

[12:13:00.0000] <remysharp>
volkmar: ignore that if you like - a new version that just listens for invalid - that doesn't fire: http://jsbin.com/icejow/7/edit#html,live

[12:15:00.0000] <volkmar>
remysharp: change 'email.form.oninvalid' to 'email.oninvalid' and it will work

[12:16:00.0000] <volkmar>
the form element isn't expected to receive the invalid event

[12:20:00.0000] <remysharp>
volkmar: but shouldn't the form fire the invalid event too?

[12:22:00.0000] <volkmar>
remysharp: http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#statically-validate-the-constraints

[12:22:01.0000] <volkmar>
remysharp: the invalid event is fired at the invalid fields

[12:22:02.0000] <volkmar>
the form never gets an invalid event according to the specs

[12:22:03.0000] <remysharp>
awesome - so shouldn't the event be removed from the form element then?

[12:23:00.0000] <remysharp>
but in fact my whole test gets borked with oninvalid (though it's a useful event)

[12:23:01.0000] <Ms2ger>
The event *handler* is on all elements for simplicity

[12:24:00.0000] <remysharp>
that's a pretty poor reason to have it there.

[12:25:00.0000] <volkmar>
remysharp: that's the case for nearly all events handler

[12:26:00.0000] <volkmar>
i think the expceptions are some events for body and a couple of other elements

[12:27:00.0000] <AryehGregor>
Hmm.  So changing your e-mail on Bugzilla is impossible, I guess, because it uses it as an ID?

[12:27:01.0000] <AryehGregor>
Annoying.

[12:28:00.0000] <remysharp>
I'm struggling to understand the use of setCustomValidity then

[12:28:01.0000] <remysharp>
Here's my example - now using email.oninvalid - http://jsbin.com/icejow/9/edit

[12:28:02.0000] <remysharp>
which works - but you have to hit submit *twice* after a corrected validation error

[12:29:00.0000] <remysharp>
before I was binding to the onsubmit - but that, as annevk, et al said - is wrong, onsubmit shouldn't submit before the form has validated

[12:29:01.0000] <remysharp>
setting it on the input event (as per the spec example) is the wrong way

[12:30:00.0000] <remysharp>
so does anyone have another example of how it should be used? (I'm sure you do!)

[12:32:00.0000] <annevk>
why is the spec example wrong?

[12:33:00.0000] <remysharp>
using oninput to set or remove the custom message?

[12:34:00.0000] <remysharp>
every keystroke it's updating the validation message - it shouldn't be tested on every keystroke

[12:34:01.0000] <remysharp>
that's one use case, but not a common use case

[12:34:02.0000] <remysharp>
if you had a input[type=email] and as you typed it told you the field was invalid, it would annoy you (or it would me)

[12:35:00.0000] <remysharp>
so we wait until they hit submit - i.e. the validation feedback should be set at the "trying to submit" phase

[12:35:01.0000] <AryehGregor>
annevk, I actually once tried to write down what cases cause HIERARCHY_REQUEST_ERR for a DOM Range algorithm, and it was ridiculously hard.

[12:36:00.0000] <AryehGregor>
(so I gave up and did it differently)

[12:36:01.0000] <AryehGregor>
The DOM is way too complicated.  :(

[12:36:02.0000] <Ms2ger>
Heh

[12:36:03.0000] <Ms2ger>
Let's kill it and start over

[12:36:04.0000] <annevk>
remysharp, it actually should for :valid and :invalid

[12:36:05.0000] <annevk>
remysharp, arguably, anyway

[12:37:00.0000] <annevk>
remysharp, ah yeah, that's why Mozilla has :ui-invalid or some such

[12:37:01.0000] <AryehGregor>
Ms2ger, I wish.  :(

[12:37:02.0000] <annevk>
hmm

[12:37:03.0000] <annevk>
AryehGregor, how is it done now?

[12:37:04.0000] <volkmar>
remysharp: at Mozilla, we are trying to push somithng that might fit you needs: https://developer.mozilla.org/en/HTML/Element/input#attr-x-moz-errormessage

[12:37:05.0000] <Ms2ger>
Hey, we edit the DOM specs, we can do this, right?

[12:37:06.0000] <Ms2ger>
Right?

[12:38:00.0000] <AryehGregor>
annevk, I don't remember, but nothing that would work for you.

[12:38:01.0000] <annevk>
Ms2ger, kill it?

[12:38:02.0000] <AryehGregor>
I think I just said "call insertBefore()" or something.

[12:38:03.0000] <remysharp>
volkmar: yeah, that's something that's specifically talked about in 2nd edition of Introducing HTML5 - which is really good - but I'm trying to get the JS side of things running

[12:38:04.0000] <annevk>
sure man

[12:38:05.0000] <annevk>
just when everyone starts to rely on DOM Core we nuke that spec from orbit

[12:38:06.0000] <remysharp>
volkmar: unless you can man handle all the others browsers to add this property for me :)

[12:39:00.0000] <annevk>
or replace it with "HAHAHAHA FOOLED YOU KTHXBAI"

[12:39:01.0000] <Ms2ger>
"Only available at the WHATWG"

[12:43:00.0000] <annevk>
Ms2ger, "gor" in DOM Range ;)

[12:43:01.0000] <remysharp>
right, my flight is landing - I'm sure you've had enough of me anyway. Thanks for the help - sorry for the hassling.

[12:44:00.0000] <AryehGregor>
Oh, I just realized why we have such a ridiculous flood of bugs from TFs now: we're almost at the deadline for LC1.

[12:44:01.0000] <AryehGregor>
Sigh, bureaucracy.

[12:44:02.0000] <annevk>
seems to get people moving somewhat

[12:45:00.0000] <annevk>
I wonder though when they would comment if the W3C did not have snapshots

[12:45:01.0000] <annevk>
Ms2ger++ for CDATA patches

[12:45:02.0000] <Ms2ger>
And if the answer was "not", whether that would be a good or a bad thing

[12:46:00.0000] <Ms2ger>
I should get back to those...

[12:46:01.0000] <annevk>
Yeah, for a lot of these comments it is somewhat hard to see the long term value

[12:46:02.0000] <annevk>
Wording always changes over time, concepts are what you really want to comment on

[12:47:00.0000] <Ms2ger>
So don't comment on my "gor" :)

[12:50:00.0000] <annevk>
I was gonna add that grammar and such is of course important, but it seems that a TF would have a somewhat different focus

[12:50:01.0000] <annevk>
well, a TF on accessibility

[12:53:00.0000] <Ms2ger>
We need a Grammar TF

[12:53:01.0000] <Ms2ger>
timeless could chair it

[12:53:02.0000] <annevk>
:)

[12:54:00.0000] <annevk>
I guess where DOM Range uses insertBefore it can be changed to use http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-node-insert instead

[12:55:00.0000] <AryehGregor>
annevk, yes, I've been waiting for hooks like that to use.

[12:55:01.0000] <AryehGregor>
It will make the mutation stuff nicer too if everyone uses them.

[12:55:02.0000] <AryehGregor>
You should have the replace algorithm invoke the insert algorithm, surely?

[12:55:03.0000] <AryehGregor>
And define a remove algorithm too.

[12:56:00.0000] <AryehGregor>
Even if it's only one step right now, it would be a useful hook.

[12:57:00.0000] <annevk>
I'm not sure it should invoke the insert algorithm.

[12:57:01.0000] <annevk>
That kind of depends on how we handle mutation events

[12:58:00.0000] <AryehGregor>
How so?  Replace is remove followed by insert, no?

[12:59:00.0000] <annevk>
would you get two mutation events?

[12:59:01.0000] <AryehGregor>
How would you not get two mutation events?  You have to get both a remove event and an insert event, no?

[13:00:00.0000] <AryehGregor>
I assume there's no "replace" event.

[13:00:01.0000] <annevk>
i guess, I have not really thought about it much

[13:00:02.0000] <annevk>
but once mutation events are there I suspect this will be reorganized somewhat

[13:01:00.0000] <AryehGregor>
The only catch is you have to do all the exception checks prior to actually removing the existing node, but you can't do the insert followed by the remove because then you won't be able to replace children of Documents that you can only have one of.

[13:01:01.0000] <AryehGregor>
But you can definitely have a lot less copy-paste than you have now.

[13:01:02.0000] <AryehGregor>
Copy-paste of a large chunk of text with a few subtle differences is the worst.

[13:02:00.0000] <annevk>
not sure how to properly merge them

[13:02:01.0000] <AryehGregor>
Eh, it'll do for now.

[13:29:00.0000] <annevk>
Ms2ger, any ideas on how to make the boilerplate part of DOM Core easily changeable?

[13:29:01.0000] <Ms2ger>
Hmm?

[13:35:00.0000] <annevk>
Ms2ger, so we can publish one copy under a more permissive copyright

[13:36:00.0000] <Ms2ger>
--filter, I guess

[13:36:01.0000] <Ms2ger>
I don't want to end up with a Hixiespec-style trainwreck :)

[13:38:00.0000] <annevk>
agreed

[13:38:01.0000] <annevk>
I guess it is mostly things that need to be omitted and a few small things added

[13:38:02.0000] <annevk>
and a change of style sheet

[13:39:00.0000] <Ms2ger>
/me has a look

[13:40:00.0000] <Ms2ger>
Copyright, This Version, logo, SotD

[13:40:01.0000] <Ms2ger>
Anything else?

[13:42:00.0000] <Ms2ger>
And CSS

[13:43:00.0000] <Ms2ger>
Doctype/meta charset? I wouldn't bother

[13:45:00.0000] <AryehGregor>
Is there some short way to grab the current function as an object from within the function itself?

[13:47:00.0000] <annevk>
Ms2ger, add dfn.js?

[13:47:01.0000] <Ms2ger>
Yes

[13:47:02.0000] <Philip`>
AryehGregor: arguments.callee ?

[13:47:03.0000] <annevk>
Ms2ger, could maybe do that for the W3C Editor's draft as well

[13:47:04.0000] <AryehGregor>
Philip`, which throws an exception in strict mode?

[13:48:00.0000] <AryehGregor>
(why does it do that?)

[13:49:00.0000] <Ms2ger>
function foo() { ...foo... }

[13:49:01.0000] <annevk>
AryehGregor, messes up JIT

[13:50:00.0000] <AryehGregor>
Hmm, why?

[13:50:01.0000] <annevk>
because you can modify the function or some such? forgot the details

[13:50:02.0000] <AryehGregor>
Ms2ger, yeah, but that's a little annoying if the function is called areLooselyEquivalentValues or something.

[13:50:03.0000] <AryehGregor>
Wondering if there was a shorter way.

[13:50:04.0000] <AryehGregor>
annevk, can't you do the same by modifying the function's name itself?

[13:50:05.0000] <Ms2ger>
var callee = areLooselyEquivalentValues;

[13:50:06.0000] <AryehGregor>
. . .

[13:50:07.0000] <annevk>
http://stackoverflow.com/questions/103598/why-was-the-arguments-callee-caller-property-deprecated-in-javascript

[13:50:08.0000] <AryehGregor>
Clever.

[13:51:00.0000] <AryehGregor>
And that keeps a reference, right, so if I modify callee it will modify the actual function?

[13:51:01.0000] <AryehGregor>
/me is always confused by how assignment works in languages more sophisticated than C

[13:51:02.0000] <Philip`>
(In C terms, everything in JS is a double or a pointer)

[13:51:03.0000] <annevk>
heh, answered by olliej

[13:52:00.0000] <annevk>
can't get any better than that

[13:52:01.0000] <annevk>
Ms2ger, hahaha typo fix is great

[13:52:02.0000] <Ms2ger>
:)

[13:53:00.0000] <AryehGregor>
Philip`, that helps.

[13:53:01.0000] <Ms2ger>
I live to serve

[14:01:00.0000] <annevk>
How do you iterate over nodes in a Range?

[14:02:00.0000] <AryehGregor>
annevk, by writing a custom function.  :)

[14:02:01.0000] <AryehGregor>
Which is fairly nontrivial, in fact.

[14:02:02.0000] <annevk>
So I heard

[14:02:03.0000] <AryehGregor>
I wrote a getAllContainedNodes() function in my implementation.js that seems to work.

[14:02:04.0000] <annevk>
gotta go, talk to you later, might extend NodeIterator at some point

[14:22:00.0000] <timeless>
Ms2ger: sure..

[14:22:01.0000] <timeless>
but someone else has to take minutes :)

[14:23:00.0000] <Ms2ger>
Minutes? We're all pretty fond of confcall-less groups, here ;)

[14:23:01.0000] <jgraham>
/me would like to say that AFAIK the Testing TF has not provided any significant LC feedback but is doing work that will actually improve both the spec and the web

[14:23:02.0000] <timeless>
ok

[14:24:00.0000] <timeless>
/me needs to look up `TF` in a table

[14:25:00.0000] <Ms2ger>
jgraham, I should file my next bug as a response from the Testing TF :)

[14:29:00.0000] <AryehGregor>
I really need a reliable way in Chrome to get a backtrace when something goes into an infinite loop.

[14:29:01.0000] <AryehGregor>
I've tried leaving the debugger open and hitting pause, but it generally just crashes the tab.

[14:29:02.0000] <AryehGregor>
Also, the "would you like to kill this tab" dialog really needs to be non-modal.

[14:30:00.0000] <jamesr>
there's some level of v8 gdb support, iirc

[14:30:01.0000] <jamesr>
or you could just look at the native stack

[14:30:02.0000] <AryehGregor>
I'd kind of prefer using Web Inspector instead of gdb here.  :)

[14:31:00.0000] <AryehGregor>
Now the infinite loops disappeared somehow, yay.

[14:35:00.0000] <AryehGregor>
Okay, so why don't we define how many pixels xx-small/x-small/.../xx-large are?

[14:35:01.0000] <AryehGregor>
I guess because if the user changes the default font size, it doesn't change explicitly-specified font sizes?

[14:36:00.0000] <AryehGregor>
But that doesn't make a lot of sense.  Why not just give the user generic options to set a minimum font size and/or scale all text up by a certain factor?

[14:36:01.0000] <AryehGregor>
Or better yet, scale up everything on the page.

[15:22:00.0000] <AryehGregor>
Argh, fontSize is a total and absolute mess.

[15:22:01.0000] <AryehGregor>
/me stab stab stab

[15:25:00.0000] <AryehGregor>
The computed value is only related to the specified value in an impossible-to-handle way.

[15:25:01.0000] <AryehGregor>
The conversion is lossy and inherently varies between UAs.

[15:54:00.0000] <AryehGregor>
w

[15:54:01.0000] <AryehGregor>
a

[15:54:02.0000] <AryehGregor>
Ack.

[15:54:03.0000] <AryehGregor>
Awesome, Last Call feedback in the form of a Word document attached to a bug.  http://www.w3.org/Bugs/Public/show_bug.cgi?id=13461

[16:22:00.0000] <jgraham>
AryehGregor: I assume when the members of the Association of American Publishers publish dictionaries they will use this as an example of irony

[16:25:00.0000] <jgraham>
(unrelated fact: I am told that my friend's sister worked in Malaysia with Alanis Morrisette's English teacher. Rumours that she had fled Canada after being unable to face the shme are, however, almost certainly false)

[16:25:01.0000] <annevk>
Wi-Fi on the plane

[16:27:00.0000] <jgraham>
Or as Samuel L. Jackson would put it, motherfuckin' Wi-Fi on the motherfuckin' plane

[16:28:00.0000] <annevk>
SEO of specs

[16:29:00.0000] <annevk>
lol

[16:29:01.0000] <Hixie>
solid technical grounds, those

[16:30:00.0000] <Hixie>
which spec matches implementations?

[16:31:00.0000] <annevk>
I think theirs might match implementations, but if we want to get rid of EventException DOM Core is the way to go

[16:31:01.0000] <Hixie>
do we want to get rid of EventException?

[16:32:00.0000] <Hixie>
i thought we'd decided we were gonna have lots of new exception interfaces and they'd each have different names or something

[16:32:01.0000] <Hixie>
or was the decision something else

[16:32:02.0000] <annevk>
I'm not sure what the decision there was

[16:32:03.0000] <Hixie>
i forget what that discussion concluded

[16:32:04.0000] <annevk>
I don't think it has yet

[16:33:00.0000] <annevk>
anyway, having a new exception for one method where INVALID_STATE_ERR already matches the semantics seems dumb

[16:33:01.0000] <Hixie>
can't disagree there

[16:33:02.0000] <Hixie>
but if it's already implemented, that ship may have sailed

[16:33:03.0000] <Hixie>
especially if it's interoperable

[16:36:00.0000] <annevk>
i'm not sure it's interoperable

[16:36:01.0000] <annevk>
I think WebKit didn't fire it

[16:36:02.0000] <Hixie>
oh i thought you said the other spec matched implementations

[16:36:03.0000] <annevk>
well some of them

[16:51:00.0000] <hober>
AryehGregor: ok, replied on list

[16:51:01.0000] <hober>
AryehGregor: (finally)

[16:54:00.0000] <hober>
AryehGregor: haahahaha, I replied on the wrong thread, sorry. stop writing so many emails that I flag in my mailer as needing attention! :)


2011-07-30
[17:28:00.0000] <Hixie>
wtf, sort(1) just decided my locale had changed

[17:45:00.0000] <annevk>
that sounds kind of weird

[18:32:00.0000] <heycam>
hopefully not mid-sort...

[18:39:00.0000] <annevk5>
heycam, btw, I found one more problem with putting event handlers into DOM Core

[18:40:00.0000] <heycam>
oh yes?

[18:40:01.0000] <annevk5>
heycam, if you look at the bits where it sets scope for content attributes it references the HTML form element

[18:40:02.0000] <annevk5>
heycam, as well as defaulting to the Window object or some such

[18:40:03.0000] <heycam>
bah

[18:41:00.0000] <annevk5>
yeah, maybe just the IDL attributes ought to be moved

[18:41:01.0000] <annevk5>
and languages define content attributes themselves and how they map for now

[18:41:02.0000] <heycam>
ok, it's not too much work

[18:41:03.0000] <heycam>
you can have a hook for that in DOM Core for simple ones

[21:05:00.0000] <Hixie>
i do believe hober volunteered to spec innerText

[21:05:01.0000] <Hixie>
:-P

[07:57:00.0000] <annevk>
Did we ever consider standardizing the contains() method?

[07:57:01.0000] <annevk>
I cannot find it in hg log

[07:57:02.0000] <annevk>
Gecko does not have it, but Presto/WebKit/Trident do

[07:58:00.0000] <annevk>
boolean = Element.contains(Element)

[08:02:00.0000] <Ms2ger>
Never heard of

[08:33:00.0000] <The_8472>
annevk, is that about layout or dom traversal?

[08:34:00.0000] <annevk>
traversal

[08:34:01.0000] <annevk>
I wonder where the actual code for that method is located in WebKit

[08:34:02.0000] <The_8472>
then you can just iterate over the parents... it's just a few lines of code.

[08:35:00.0000] <annevk>
sure or you can use compareDocumentPosition

[08:38:00.0000] <bga_>
annevk you can standardtize _commonMinParent(node1, node2) because via Range API its so slow and look like hack

[08:39:00.0000] <annevk>
I'm not sure what you mean

[08:41:00.0000] <bga_>
get 2 nodes and find min parent node that contains given 2 nodes

[08:41:01.0000] <bga_>
min parent, not documentElement

[08:41:02.0000] <annevk>
and why do you want this?

[08:45:00.0000] <bga_>
annevk i can move parent <ul> by 2 given <li> even if one <li> is is deep inside

[08:47:00.0000] <bga_>
btw this operation can be generalized to many input nodes

[08:51:00.0000] <The_8472>
getNearestCommonAncestor would be a better name i think

[08:52:00.0000] <bga_>
yeah

[11:44:00.0000] <gsnedders>
/me wishes there was an IRC channel where JS people lurked

[13:29:00.0000] <mhausenblas>
gsnedders: re 'wish there was an IRC channel where JS people lurked' - I'd give #Node.js a try ;)

[13:30:00.0000] <gsnedders>
mhausenblas: Probably rather bias towards V8. :)

[13:30:01.0000] <mhausenblas>
oh well :P

[13:30:02.0000] <mhausenblas>
seems rather busy there tbh

[13:31:00.0000] <mhausenblas>
anyways ... /me back to https://github.com/mhausenblas/web.instata

[13:31:01.0000] <gsnedders>
(Node.js is based on V8, so, well, yeah.)

[13:31:02.0000] <mhausenblas>
sure

[13:31:03.0000] <mhausenblas>
but my impression was that quite some JS people hang out there

[13:32:00.0000] <mhausenblas>
laters

[15:19:00.0000] <jgraham>
Surely HTML5 doesn't really use "child" to mean "descendant"

[15:31:00.0000] <foolip_>
jgraham, [citation needed]

[16:04:00.0000] <charlvn>
jgraham: example?

[16:05:00.0000] <Dashiva>
hsivonen: A few days late, but my guess is a lot of people still use firefox 3.6 because of ubuntu

[16:06:00.0000] <charlvn>
Dashiva: why because of ubuntu?

[16:10:00.0000] <Dashiva>
Because the current LTS is lucid with 3.6

[16:11:00.0000] <charlvn>
not sure if this is typical but most ubuntu users i know upgrade rapidly to newer versions

[16:11:01.0000] <charlvn>
people only seem to care about lts for servers

[16:12:00.0000] <Dashiva>
Well, there's probably a similar thing going on debian too

[16:12:01.0000] <charlvn>
well debian uses iceweazel

[16:20:00.0000] <jgraham>
foolip_: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13480

[16:21:00.0000] <foolip_>
I wonder if we're supposed to be reading/filing spec bugs on Saturday nights...

[16:22:00.0000] <charlvn>
bug with no references in spec = fail


2011-07-31
[02:15:00.0000] <annevk>
jetlag sucks

[02:23:00.0000] <othermaciej>
yep

[02:44:00.0000] <dydx>
hi othermaciej

[02:44:01.0000] <othermaciej>
hello

[03:02:00.0000] <annevk>
AryehGregor, so I agree it would be nice to share the "insert" stuff

[03:03:00.0000] <annevk>
AryehGregor, however that means the insert algorithm needs to be split into the steps you need to run before actually inserting and actually inserting the node

[03:03:01.0000] <annevk>
AryehGregor, calling one "insert" and one "actually insert" seems kind of weird

[03:04:00.0000] <annevk>
could have "pre-insert" and "insert" I suppose, but it seems kind of weird that appendChild invokes "pre-insert" which in turn invokes "insert"

[03:06:00.0000] <annevk>
pre-insert would be used by insertBefore/appendChild

[03:06:01.0000] <annevk>
insert would be used by pre-insert/replaceChild

[03:07:00.0000] <annevk>
pre-delete would be used by removeChild

[03:07:01.0000] <annevk>
delete would be used by pre-delete/replaceChild

[03:07:02.0000] <annevk>
I guess that could work

[03:08:00.0000] <annevk>
To pre-delete /node/ ...

[03:08:01.0000] <annevk>
s/delete/remove/

[03:41:00.0000] <annevk>
I made the above changes

[03:42:00.0000] <annevk>
s/above/aforementioned/

[07:58:00.0000] <annevk>
So contains() in WebKit is implemented on Node but only exposed on Element?

[07:59:00.0000] <annevk>
Another difference with Opera is that in Opera you can check for non-Element descendants. In WebKit that simply returns false.

[08:02:00.0000] <foolip_>
hsivonen, are you around?

[08:10:00.0000] <AryehGregor>
annevk, I am 147% in favor of a generic contains() on Node that can be used on any Node.

[08:10:01.0000] <AryehGregor>
That would be awesome.

[08:11:00.0000] <AryehGregor>
compareDocumentPosition is a nightmare.

[08:11:01.0000] <AryehGregor>
I wrote my own isAncestor() and isDescendant() helper methods.

[08:11:02.0000] <gsnedders>
+1 for contains working for any Node

[08:11:03.0000] <AryehGregor>
Plus, contains() must be cheaper to implement than compareDocumentPosition().

[08:12:00.0000] <AryehGregor>
Usually you don't care about the full position, you just want to know either ancestor/descendant or before/after.

[08:27:00.0000] <annevk>
AryehGregor, hmm maybe

[08:27:01.0000] <annevk>
AryehGregor, nobody has that though

[08:27:02.0000] <AryehGregor>
Standards are allowed to add new features sometimes, you know.  :)

[08:28:00.0000] <annevk>
True enough, but I rather be conservative here.

[08:28:01.0000] <annevk>
Adding a bunch of features is exactly what went wrong with the DOM.

[08:29:00.0000] <AryehGregor>
Are you talking about expanding contains() or adding the other stuff I mentioned?

[08:29:01.0000] <annevk>
Both :)

[08:30:00.0000] <annevk>
Even standardizing contains()

[08:30:01.0000] <AryehGregor>
Well, ask implementers.

[08:30:02.0000] <AryehGregor>
I think we could use a few more convenience features like that.

[08:32:00.0000] <hsivonen>
foolip_: I'm somewhat around

[08:41:00.0000] <foolip_>
hsivonen, I was adding integration with validator.nu on my live microdata too, see http://worf.foolip.org/microdatajs/live/

[08:41:01.0000] <foolip_>
but I can't figure out how to make the source show up on your side

[08:41:02.0000] <foolip_>
<input type="hidden" name="showsource" value="yes"> didn't do anything, it seems

[08:42:00.0000] <foolip_>
perhaps copySourceIntoTextArea is where the magic is, but I not sure what it's doing

[08:42:01.0000] <foolip_>
s/too/tool/

[08:43:00.0000] <foolip_>
meh, I should get an IRC client with a grammar checker :)

[08:52:00.0000] <hsivonen>
foolip_: I have no idea what happens

[08:52:01.0000] <hsivonen>
foolip_: the explanation might be buried in https://bitbucket.org/validator/validator/src/0a49bcad4ce5/src/nu/validator/servlet/MultipartFormDataFilter.java

[08:52:02.0000] <hsivonen>
but it's been a while since I wrote that, so I don't remember what oddities it has

[08:53:00.0000] <foolip_>
ok, I'll just hack away until I find the answer

[08:53:01.0000] <foolip_>
thanks

[08:54:00.0000] <hsivonen>
foolip_: IIRC, there's JS that copies the context from the source display to the textarea

[08:54:01.0000] <hsivonen>
s/context/content/

[08:54:02.0000] <foolip_>
hsivonen, but what is the source display?

[08:55:00.0000] <foolip_>
ok, I see how it works from the site itself

[08:55:01.0000] <hsivonen>
foolip_: It's an <ol>

[08:55:02.0000] <foolip_>
just need to figure out why my form doesn't have the same effect

[08:56:00.0000] <hsivonen>
hmm. maybe I put showsource=yes in the URL even though the content is POSTed

[08:56:01.0000] <hsivonen>
or something weird like that

[09:45:00.0000] <foolip_>
I sense frustration: https://plus.google.com/u/0/109022289777044064966/posts/j43r9owxqry

[09:47:00.0000] <erlehmann>
foolip_, a classic “U MAD” ;)

[09:57:00.0000] <foolip_>
hsivonen, odd, it looks like the order of the <input> elements matters, when switching name="content" and name="showsource" it works

[10:04:00.0000] <foolip_>
I'll just use the order that works without further investigation

[10:16:00.0000] <annevk>
foolip_, would be nice if G+ had a real subscribe function

[10:17:00.0000] <foolip_>
indeed

[10:17:01.0000] <foolip_>
I assume it's coming

[10:17:02.0000] <foolip_>
together with unsubscribe

[10:17:03.0000] <foolip_>
for all those derailed threads :)

[10:25:00.0000] <annevk>
mute is unsubscribe

[10:25:01.0000] <annevk>
although maybe that does not stop notifications? hmm

[10:26:00.0000] <foolip_>
hmm, that also hides in from the stream, though

[10:26:01.0000] <annevk>
what a bullshit thread that is by the way, on use cases

[10:26:02.0000] <foolip_>
better not respond further

[10:26:03.0000] <annevk>
we know what we want so we can skip the use cases test? mwaha

[10:27:00.0000] <foolip_>
go ahead and dive in if you want to be punched ;)

[10:27:01.0000] <annevk>
I never really bothered on mailing lists either so I am not sure I should start on Google+

[10:27:02.0000] <annevk>
:)

[10:27:03.0000] <foolip_>
I hear you're quite used to fountain-wrestling and such, it should suit you

[10:27:04.0000] <foolip_>
how did that actually happen?

[10:28:00.0000] <annevk>
well we were at Ann's house which is at Lake Union

[10:28:01.0000] <annevk>
and I asked Alex about going for a swim

[10:29:00.0000] <annevk>
I'm not sure why, but he walked over to the water and went on his knees to get a feel for how warm it was

[10:29:01.0000] <annevk>
that was my cue

[10:29:02.0000] <foolip_>
it was deliberate?!

[10:29:03.0000] <annevk>
well yes :)

[10:30:00.0000] <foolip_>
you're crazy

[10:30:01.0000] <annevk>
I'll take that as a compliment

[10:30:02.0000] <foolip_>
was he amused?

[10:30:03.0000] <annevk>
I jumped in after him, but was wearing swimming shorts

[10:30:04.0000] <annevk>
he was cool with it, provided I got him a new phone

[10:31:00.0000] <foolip_>
oh, I thought it was a laptop

[10:31:01.0000] <annevk>
these days pushing someone in the water is expensive

[10:31:02.0000] <foolip_>
you wrote "apple hardware" or some such

[10:31:03.0000] <annevk>
yeah, iPhone

[10:31:04.0000] <foolip_>
ah well, then you're not quite as crazy

[10:31:05.0000] <foolip_>
sorry to retract a compliment

[10:31:06.0000] <annevk>
aaah

[10:32:00.0000] <annevk>
back to Curb your Enthusiasm

[10:32:01.0000] <annevk>
LD is awesome

[10:45:00.0000] <erlehmann>
foolip_, i do not assume google+ gets better. after all, they see that the majority does not care. and feeds et al are not *that* difficult.

[10:45:01.0000] <foolip_>
there are feeds?

[10:55:00.0000] <erlehmann>
foolip_, no. but they could have had feeds and chose not to.

[10:56:00.0000] <foolip_>
do feeds that require authentication actually work in things like Google Reader?

[10:57:00.0000] <foolip_>
sounds unlikely

[11:36:00.0000] <hsivonen>
can folks who have a Mac and have never had Perian installed, please, test http://webm.html5.org/ in Safari?

[11:36:01.0000] <hsivonen>
it seems that I'm unable to revert to a Perianless state for retesting

[11:45:00.0000] <annevk>
"Your browser supports HTML5 video but does not support WebM.

[11:45:01.0000] <annevk>
You need to install Perian to view WebM content in your current browser."

[11:49:00.0000] <hsivonen>
annevk: ok. works. thank you

[11:49:01.0000] <hsivonen>
Hixie: is https://plus.google.com/u/0/109022289777044064966/posts/PRJjCL75KPu on your radar?

[11:51:00.0000] <gsnedders>
foolip_: WebKit/OSX has mappings that do that? I thought it all came from querying the QT APIs.

[11:54:00.0000] <annevk>
So after the big schema.org announcement people are going to try to change everything yet again?

[11:54:01.0000] <annevk>
Authors are going to be so thrilled

[11:55:00.0000] <hsivonen>
annevk: http://www.xkcd.com/927/

[12:09:00.0000] <foolip_>
annevk, not the people who announced schema.org ...

[12:10:00.0000] <foolip_>
gsnedders, I'd be surprised if the plugins exposed Content-Type-compatible names for containers and codecs, neither GStreamer, DirectShow or MediaFoundation do

[12:11:00.0000] <foolip_>
so it looks like Google+ is the new venue for creating web standards :)

[12:12:00.0000] <foolip_>
for some reason it seems easier to actually discuss things, is it because people have a picture and the font isn't monospaced?

[12:12:01.0000] <foolip_>
mysterious

[12:13:00.0000] <hsivonen>
foolip_: too bad the venue excludes pseudonymous people like Ms2ger

[12:14:00.0000] <foolip_>
hsivonen, it is

[12:14:01.0000] <foolip_>
I don't think they designed it primarily for people who only know each other by IRC nicks

[12:15:00.0000] <foolip_>
http://mercurial.selenic.com/wiki/GitConcepts#Git.27s_staging_area was unhelpful

[12:16:00.0000] <foolip_>
how am I supposed to write code with random debug print messages and not commit those using mercurial?

[12:16:01.0000] <foolip_>
without git add -p I am lost

[12:16:02.0000] <hsivonen>
foolip_: merqurial queues FTW

[12:16:03.0000] <Philip`>
Maybe since it's web-based there's more a feeling that it's public and you shouldn't say anything particularly rude and unhelpful and you should explain yourself with more context, whereas email feels more private

[12:17:00.0000] <foolip_>
Philip`, maybe, email archives certainly suck enough that almost no one will read them

[12:18:00.0000] <foolip_>
hsivonen, http://mercurial.selenic.com/wiki/MqExtension ?

[12:18:01.0000] <foolip_>
it seems like everything is an extension in mercurial :)

[12:20:00.0000] <Philip`>
Isn't modular software meant to be a good idea? :-)

[12:20:01.0000] <foolip_>
I guess, but the difference here is which modules are available by default

[12:20:02.0000] <foolip_>
perhaps hg tries to not scare newcomers

[12:20:03.0000] <foolip_>
git certainly makes no attempt to be approachable

[12:21:00.0000] <hsivonen>
foolip_: yes, that one. and yes, hg needs extensions to work

[12:21:01.0000] <foolip_>
my first few encounters with git were funny, always getting me into a state I didn't know how to get out of

[12:21:02.0000] <Philip`>
I think it's nice that you can treat default hg like a slightly fancier svn, and once you understand it and get annoyed by limitations you can find an extension that removes those limitations

[12:22:00.0000] <foolip_>
funny in a "haha I need to re-clone the repo and start over" kind of way

[12:22:01.0000] <Philip`>
instead of being thrown in at the deep end with a million features that you have no idea why you'd ever use

[12:23:00.0000] <foolip_>
certainly, but I wish there were good tools that allowed me to work with a hg or bzr repo using the git tools that are now hardcoded in my brain

[12:23:01.0000] <foolip_>
it's like switching between qwerty and dvorak all the time

[12:24:00.0000] <Philip`>
I wish either hg or git was rubbish, so everyone would standardise on one or the other

[12:25:00.0000] <foolip_>
yes, that would be great!

[12:25:01.0000] <foolip_>
doesn't look like the world works like that, though

[12:26:00.0000] <foolip_>
looks like http://stackoverflow.com/questions/883452/git-interoperability-with-a-mercurial-repository may be what I'm looking for

[12:26:01.0000] <foolip_>
I already found that n times, but always thought it was the reverse of what it is

[12:26:02.0000] <foolip_>
odd

[12:33:00.0000] <erlehmann>
Philip`, hg is rubbish. it is slow.

[12:34:00.0000] <erlehmann>
foolip_, i use hg-fast-export to merge upstream changes for a game, i maintain the dev branch. so that works, unless you have multiple heads

[12:35:00.0000] <erlehmann>
(i use git. upstream uses hg. everyone else uses git.)

[12:35:01.0000] <foolip_>
erlehmann, don't you need to re-export every time upstream changed and rebase --onto your changes?

[12:35:02.0000] <foolip_>
or is it clever enough to handle that?

[12:36:00.0000] <erlehmann>
foolip_, it is clever enough. however, upstream does not seem to be clever enough, so i now have duplicate commits.

[12:36:01.0000] <erlehmann>
as upstream evolves, the old commits have the same IDs

[12:36:02.0000] <erlehmann>
is that what you want?

[12:38:00.0000] <foolip_>
erlehmann, sounds like it, yes

[12:38:01.0000] <erlehmann>
also, if upstream does something that requires you to rebase, upstream is bad-bad!

[12:38:02.0000] <doublec>
I use hg-git for git/hg conversions

[12:39:00.0000] <erlehmann>
i do

[12:39:01.0000] <erlehmann>
hg-fast-export -r ~/share/src/games/minetest -o upstream

[12:39:02.0000] <erlehmann>
 git push origin upstream/master:upstream -f

[12:39:03.0000] <erlehmann>
(the -f is there because it is just intended as a mirror)

[12:39:04.0000] <foolip_>
I'd like the two repos to live in the same working tree

[12:40:00.0000] <foolip_>
doublec, do you do something like in the linked stackoverflow answer?

[12:40:01.0000] <erlehmann>
everyone telling me that git and hg are equivalent is a mercurial user.

[12:40:02.0000] <doublec>
foolip_: yes

[12:41:00.0000] <doublec>
foolip_: it's how I work on the mozilla hg repository - I maintain a git mirror and use that

[12:41:01.0000] <foolip_>
doublec, I'll try to follow that answer then

[12:41:02.0000] <doublec>
foolip_: http://www.bluishcoder.co.nz/2011/02/10/git-conversion-of-mozilla-central.html

[12:42:00.0000] <doublec>
foolip_: and http://www.bluishcoder.co.nz/2011/04/16/my-git-workflow-for-mozilla-development.html

[12:42:01.0000] <erlehmann>
reminds me of this one microsoft PR (?) guy i once met, who really wanted me to know that windows can do everything linux can. kinda annoying.

[13:39:00.0000] <twisted`>
yo, for some reason a piece of CSS get's totally ignored in Chrome/Firefox, got the html + css here: http://jsfiddle.net/pPpbx/

[13:39:01.0000] <twisted`>
the offending bit is: http://bastardoperatorfromhell.org/~twisted/jing/index.php?f=2011-07-31_2136.png&h=146&w=229

[13:40:00.0000] <twisted`>
the css class is .category_img {}

[13:40:01.0000] <twisted`>
and somehow doesn't get loaded :s

[13:42:00.0000] <The_8472>
there seems to be a special character in the css selector

[13:43:00.0000] <Philip`>
Looks like a U+0010

[13:43:01.0000] <Philip`>
(after ".category_img " before "{")

[13:49:00.0000] <twisted`>
really? let me check

[13:50:00.0000] <twisted`>
just deleted whatever was there and placed a space, didn't seem to change

[13:51:00.0000] <twisted`>
ah!

[13:51:01.0000] <twisted`>
The_8472 and Philip` THANKS!!!

[13:52:00.0000] <twisted`>
I'll now leave invisible chars on in Coda

[13:52:01.0000] <twisted`>
:)

[13:57:00.0000] <Philip`>
If you turn display of invisible characters on, they're no longer invisible characters, so they shouldn't be displayed by that option

[13:59:00.0000] <twisted`>
Philip`: it showed as a red questionmark in a rotated square

[13:59:01.0000] <twisted`>
the 'illegal' char, when using invisible chars

[13:59:02.0000] <twisted`>
but I get your logic ;)

[14:17:00.0000] <annevk>
http://technologizer.com/2011/07/29/a-brief-history-of-apple-not-buying-things/ is nice

[14:17:01.0000] <annevk>
analysts...

[16:02:00.0000] <Yuhong>
FYI, just tried http://www.hixie.ch/tests/adhoc/html/parsing/compat/004.html under Netscape 4.8 for fun.

[16:05:00.0000] <Yuhong>
FYI, it seems to be equivalent to <em><address></address></em>.