2012-07-01
[20:12:00.0000] <MikeSmith>
will be interesting to see where http://paulrouget.com/e/servo/  goes

[23:59:04.0000] <annevk>
gsnedders: it's less than half typically


2012-07-02
[19:24:33.0000] <MikeSmith>
https://groups.google.com/forum/?fromgroups#!topic/es-operating-system/8oWtRZnDK_w

[19:25:12.0000] <MikeSmith>
Shiki Okasaka  wrote his own new browser engine

[22:47:45.0000] <annevk>
MikeSmith: http://code.google.com/p/es-operating-system/wiki/Cplusplus0xBinding is also pretty interesting

[22:48:12.0000] <annevk>
MikeSmith: did he implement HTML/CSS in JavaScript or C/C++?

[22:55:58.0000] <MikeSmith>
annevk: the engine is all C++ I think

[22:56:18.0000] <MikeSmith>
the UI is HTML and CSS

[22:57:43.0000] <annevk>
cool cool cool

[22:58:18.0000] <MikeSmith>
we should go visit Shiki down in Kyoto next time you're here

[22:58:27.0000] <MikeSmith>
he owns a cafe there

[22:58:38.0000] <annevk>
yeah I know, would be nice :)

[23:19:53.0000] <zcorpan>
for <ol reversed>, isn't the simplest thing to compute the start value by counting child <li> elements, and let the result be bogus if you do something bogus?

[23:20:21.0000] <annevk>
how is computed now?

[23:20:27.0000] <annevk>
it /\

[23:22:20.0000] <Ms2ger>
That's the way it's counted

[23:22:40.0000] <Ms2ger>
Only if you do something bogus the spec isn't too clear about markers

[01:00:11.0000] <bdc>
hey! I was wondering if there's a way to completely remove the app cache from a website? Deleting the file and reference isn’t enough apparently :/

[01:01:53.0000] <shwetank>
you mean programatically? or just from the perspective of the user (i.e, by asking the user to go in the browser's settings and delete it)?

[01:02:07.0000] <bdc>
programatically :)

[01:06:18.0000] <shwetank>
well, it should be enough ... which browser you're trying in? trying reloading a couple of times I guess, deleting the file and/or reference to it should be enough (i've always done this myself and had no issues)

[01:08:29.0000] <bdc>
hm ok I'll give it another try then… testing on mobile safari / ios5

[01:09:28.0000] <zcorpan>
TabAtkins_: please add a new value to 'clear', clear:fix

[01:09:48.0000] <bdc>
I might be wrong but I think I've read some article saying it was actually not enough to just remove references

[01:09:49.0000] <zcorpan>
which will be the new new new way to do it

[01:10:24.0000] <zcorpan>
bdc: did you serve the manifest with no-cache headers?

[01:10:55.0000] <bdc>
zcorpan: good question :p

[01:11:45.0000] <zcorpan>
see gotcha #4 http://www.alistapart.com/articles/application-cache-is-a-douchebag/

[01:12:42.0000] <bdc>
that's exactly the issue I have

[02:27:54.0000] <annevk>
Velmont: zcorpan: fancy creating a proper SVG out of http://lists.w3.org/Archives/Public/public-webappsec/2012Jun/att-0045/Hodges-CORS-Diagram-SimpleRequest-00.pdf ?

[02:28:17.0000] <odinho>
annevk: hm, ok.

[02:28:36.0000] <annevk>
odinho: unless you think it's not useful of course, but it looks fairly useful

[02:31:08.0000] <odinho>
Yeah. I might've decoded it faster if it was more visual though. But it's helpful either way.

[02:33:03.0000] <annevk>
if it would've helped you, future generations will like you for making it that way :)

[02:39:48.0000] <zcorpan>
annevk: doesn't seem useful for *me*

[02:41:40.0000] <annevk>
:)

[02:42:14.0000] <odinho>
zcorpan: Yeah, but you need to put yourself in a poor webdev's shoes and then think about it like that.

[02:42:56.0000] <zcorpan>
yeah i'm not saying it's not useful for everyone but me

[02:43:31.0000] <jgraham>
It seems pretty confusing

[02:43:32.0000] <odinho>
zcorpan: It's not like I didn't know how these requests were supposed to be made :P

[02:43:55.0000] <annevk>
jgraham: are you on public-webappsec?

[02:44:00.0000] <annevk>
jgraham: might want to mention that

[02:44:01.0000] <jgraham>
No

[02:44:11.0000] <annevk>
email to reply to is http://lists.w3.org/Archives/Public/public-webappsec/2012Jun/0045.html

[02:44:18.0000] <jgraham>
I can mention it here

[02:44:34.0000] <jgraham>
It seems inconsistent in the use of speech bubbles

[02:45:03.0000] <jgraham>
The initial request doesn't have one. They don't all list the same things

[02:45:41.0000] <odinho>
Yeah, I'm kinda changing it a bit now.

[03:59:51.0000] <matjas>
`'_'.anchor('x')` is `'<a name="x">_</a>'` in all browsers, except in IE, where it’s `'<A NAME="x">_</A>'` instead

[04:00:49.0000] <matjas>
should I tweak http://mathias.html5.org/specs/javascript/#string to allow uppercase, or could I ignore IE in this case as the resulting string is equivalent anyway?

[04:06:30.0000] <Ms2ger>
IE is dumb, spec the lower case thing :)

[04:06:46.0000] <matjas>
yay, then no tweaks are needed!

[04:07:11.0000] <matjas>
I guess the Strings chapter is complete then. I wrote tests for it, too: http://mathias.html5.org/tests/javascript/strings/

[04:08:25.0000] <matjas>
I spent some time speccing the Object extensions too, in case anyone wants to review: http://mathias.html5.org/specs/javascript/#object

[04:09:15.0000] <zcorpan>
matjas: as a general rule, don't allow two things in a spec just because browsers disagree. specs are supposed to get rid of the "disagree" part, so pick the majority or the most sensible option

[04:09:33.0000] <matjas>
zcorpan: thanks; that makes sense

[04:13:00.0000] <zcorpan>
matjas: is your algorithm of substr different from the one in the ES spec?

[04:13:30.0000] <matjas>
it produces the same results, but the algo is a bit different

[04:13:58.0000] <matjas>
i had already written this when you told me it was already defined in the ES5.1 spec

[04:14:29.0000] <matjas>
i think this version is much more readable, but i guess that doesn’t matter

[04:14:31.0000] <Ms2ger>
matjas, do you want to check IsCallable() (http://ecma-international.org/ecma-262/5.1/#sec-9.11) for defineGetter/defineSetter?

[04:14:35.0000] <zcorpan>
i suggest you drop it and just reference the ES one

[04:14:38.0000] <matjas>
/me removes

[04:15:01.0000] <zcorpan>
having two algorithms means implementors need to spend time one evaluating whether they are equivalent

[04:16:19.0000] <zcorpan>
matjas: i guess all of the appendix in the ES spec is needed for compat, right? maybe say that all of it must be supported?

[04:18:34.0000] <matjas>
zcorpan: good call. would it be useful to have the separate features listed out in our spec, or would you just mention annex b once and never explicitly mention e.g. String#substr

[04:19:47.0000] <Ms2ger>
matjas, and what do Object.prototype.__defineGetter__ and friends do if you call them on non-objects?

[04:19:54.0000] <zcorpan>
either way works, your call :-)

[04:20:08.0000] <matjas>
Ms2ger: do you think that’s better / more clear than what it says now? “If function is not a Function instance (of type Function), throw a TypeError and return”

[04:20:33.0000] <Ms2ger>
Well, is it equivalent, and if not, what do browsers do?

[04:21:06.0000] <matjas>
seems like it is, as only Function instances have an internal [[Call]] property set to `true`

[04:21:10.0000] <matjas>
afaict

[04:21:40.0000] <matjas>
i guess it’s clearer to re-use the ES spec lingo though

[04:22:30.0000] <matjas>
thanks for the feedback! <3

[04:22:36.0000] <Ms2ger>
Np :)

[04:23:24.0000] <Ms2ger>
Hmm, http://mxr.mozilla.org/mozilla-central/source/js/src/tests/ecma_5/extensions/extension-methods-reject-null-undefined-this.js

[04:26:25.0000] <Ms2ger>
/me goes off for a bit

[04:28:25.0000] <matjas>
interesting! I didn’t know Mozilla had a `String#quote`

[04:36:55.0000] <hasather>
matjas: In http://mathias.html5.org/specs/javascript/#tohtmltag, step 2, there is an extra quotation mark in "\"">"

[04:38:02.0000] <matjas>
hasather: good catch; thanks!

[04:38:38.0000] <hasather>
matjas: I'm also skeptic about the name, ToHTMLElement instead?

[04:39:44.0000] <matjas>
hasather: i don’t know. IMHO an element is what’s in the DOM, tags are what devs write

[04:40:18.0000] <matjas>
it doesn’t actually create an element (like `document.createElement`) so I think that would be confusing

[04:41:02.0000] <matjas>
hmm, it’s not just one “tag” though

[04:41:18.0000] <matjas>
could it be just “ToHTML”?

[04:45:46.0000] <hasather>
matjas: that's better. I wouldn't say that element is only a DOM concept though (XML has a production for it, and so has SGML)

[04:47:42.0000] <zcorpan>
ToTagSoup

[04:48:00.0000] <hasather>
:D

[04:48:49.0000] <zcorpan>
i think that's about as accurate as it gets :-)

[04:51:04.0000] <matjas>
:')

[04:54:17.0000] <zcorpan>
matjas: s#<code>"&amp;quot;"</code>#"<code>&amp;quot;</code>"#

[04:56:06.0000] <matjas>
zcorpan: thanks. i assume the same goes for the strings in the http://mathias.html5.org/specs/javascript/#tohtml algo?

[04:56:21.0000] <zcorpan>
yeah

[04:57:10.0000] <zcorpan>
actually the backslash thing is a bit un-spec-y

[04:58:01.0000] <zcorpan>
i think i'd use http://simon.html5.org/tools/js/unicodeize/ notation

[04:59:21.0000] <matjas>
so, "<code>"></code>" (U+0022 U+003E)

[04:59:39.0000] <matjas>
nice tool btw!

[05:00:10.0000] <zcorpan>
yeah

[05:00:21.0000] <zcorpan>
maybe i should make it emit <code>s too

[05:00:28.0000] <matjas>
do it!

[05:01:27.0000] <hasather>
matjas: is there a reason to say "six-character string" instead of "string"?

[05:02:05.0000] <matjas>
hasather: I saw that in the HTML spec somewhere and assumed it was the right way of doing things

[05:02:40.0000] <matjas>
hasather: it also made it more clear that the quotes aren’t part of the string; but that’s not an issue anymore with zcorpan’s suggestions

[05:03:03.0000] <hasather>
matjas: seems redundant, and it's not used like that always like that in the HTML spec, but I don't know if there's a reason for it

[05:03:20.0000] <hasather>
matjas: anyway, just nit-picking :)

[05:03:43.0000] <matjas>
hasather: thanks for nit-picking! very useful feedback

[05:06:30.0000] <zcorpan>
matjas: done

[05:07:18.0000] <matjas>
zcorpan: thanks! could you add .replace(/</g, '&lt;').replace(/&/g, '&amp;')?

[05:07:46.0000] <zcorpan>
i did :-)

[05:07:57.0000] <matjas>
thanks!

[05:10:57.0000] <hasather>
matjas: "=\"" still has a backslash, and not all of them have code points, but maybe you're currently updating that

[05:11:10.0000] <matjas>
hasather: I am, but thanks anyway :)

[06:02:03.0000] <odinho>
hmmm. Is this clearer?  http://odin.s0.no/tmp/Hodges_odinho-CORS-Diagram-SimpleRequest.svg

[07:24:27.0000] <matjas>
odinho: minor typo: s/Do Access-Control-Allow-Origin/Does Access-Control-Allow-Origin/

[07:24:42.0000] <odinho>
matjas: Fack, and I "fixed" it from Does to Do :P

[07:24:51.0000] <odinho>
Was unsure, so.

[07:40:00.0000] <odinho>
matjas: But no other comments?

[07:41:17.0000] <odinho>
So, I never liked that IndexedDB used .continue() and .delete() as they are reserved in JS, but thought, ohwell. -- But it actually breaks "with".   with(os) { delete(); continue(); }  doesn't work at all...

[07:41:25.0000] <matjas>
odinho: looks very clear to me. maybe explain what happens if access is not allowed

[07:41:36.0000] <jgraham>
odinho: I call "feature"

[07:41:46.0000] <Ms2ger>
^

[07:42:21.0000] <odinho>
matjas: Yea, thought about that. ... Hmm. Maybe I should try. Don't want it to be too advanced. And I'm basically only beautifying and SVG-ing something Jeff Hodges made.

[07:42:30.0000] <matjas>
odinho: only for backwards compatibility with ES3. in ES5, `continue` and `delete` are valid property names: http://mothereff.in/js-properties#delete

[07:43:18.0000] <odinho>
matjas: You mean that with will work correctly then?

[07:43:32.0000] <matjas>
odinho: as a property, yes

[07:44:26.0000] <matjas>
odinho: in your `with` example, you’re using `delete` as an identifier, which is not allowed

[07:44:30.0000] <odinho>
Yea, but property.  continue() isn't a property? It has to hang onto somewhere?

[07:44:31.0000] <matjas>
even in ES5

[07:44:48.0000] <odinho>
matjas: Yep, so what I was saying :P  -- We have  object.delete() and object.continue() working fine.

[07:45:09.0000] <matjas>
yeah, because they’re property names

[07:45:12.0000] <matjas>
ES5 allows the use of unquoted reserved words as property names

[07:45:38.0000] <matjas>
odinho: reserved words are valid IdentifierNames, but not valid Identifiers, in ES5

[07:46:01.0000] <odinho>
Yeah, but it's a bit strange to use them in a web API like IDB...

[07:46:06.0000] <matjas>
see http://mathiasbynens.be/notes/javascript-identifiers and http://mathiasbynens.be/notes/javascript-properties if you’re interested

[07:46:08.0000] <odinho>
Highlight in vim is strange and all.

[07:46:45.0000] <odinho>
matjas: Yeah, I have somewhat control over that, keyPaths have to be valid identifiers, so I've read those parts of the ES spec.

[07:48:20.0000] <AryehGregor>
odinho, 'with' is evil anyway.

[07:48:43.0000] <odinho>
yeah, noone needs to tell me.

[07:49:09.0000] <jgraham>
odinho: It seems they do, because you keep talking as if breaking it is a problem

[07:49:15.0000] <jgraham>
;)

[07:49:16.0000] <odinho>
But it's also evil to make API's that break features.

[07:49:28.0000] <jgraham>
"with" ain't no feature

[07:49:34.0000] <odinho>
And that highlight strangely in vim. :P

[07:49:46.0000] <AryehGregor>
Aha, so now we see your real motive.

[07:49:48.0000] <odinho>
jgraham: misfeatures then.

[07:52:45.0000] <matjas>
Gecko just unprefixed backface-visibility, perspective, perspective-origin, transform, transform-origin, transform-style — nice!

[07:53:19.0000] <AryehGregor>
Yep!

[07:53:23.0000] <AryehGregor>
Hurrah!

[07:53:28.0000] <AryehGregor>
Thank Ms2ger.

[07:55:55.0000] <Ms2ger>
Yay me :)

[08:00:43.0000] <Ms2ger>
matjas, all the String.prototype functions in ES5 call CheckObjectCoercible, and it looks like Gecko does that as well

[08:20:06.0000] <matjas>
Ms2ger: ah, only now I understand explains the Gecko source link you posted earlier

[08:20:08.0000] <matjas>
Ms2ger: thanks!

[08:20:24.0000] <Ms2ger>
Np

[08:43:04.0000] <matjas>
Ms2ger: do you know what I’m doing wrong here? test(function() { assert_throws(null, String.prototype.substr.call(null)); }); — http://mathias.html5.org/tests/javascript/strings/

[08:43:22.0000] <matjas>
I tried `assert_throws('TypeError', fn)` first

[08:44:07.0000] <Ms2ger>
assert_throws(null, function() { String.prototype.substr.call(null); });

[08:44:22.0000] <Ms2ger>
Or assert_throws(new TypeError(), function() { String.prototype.substr.call(null); });

[08:44:31.0000] <matjas>
without wrapping it in test(), you mean?

[08:44:58.0000] <Ms2ger>
No, within the test

[08:45:05.0000] <Ms2ger>
But you need another function()

[08:45:15.0000] <matjas>
ah, duh!

[08:45:19.0000] <matjas>
thanks

[09:02:45.0000] <karlcow>
I wonder if the kobo touch JP (webkit) handles Furigana http://img.asiajin.com/blog/wp-content/uploads/2012/07/rakuten-kobo.png

[09:03:01.0000] <karlcow>
hmm yes it does.

[09:03:08.0000] <karlcow>
stupid me visible on the pic

[09:07:09.0000] <matjas>
interesting, WebKit & IE don’t perform the CheckObjectCoercible step for the String extensions

[09:17:13.0000] <Ms2ger>
Opera and Gecko do, though

[09:17:29.0000] <Ms2ger>
Actually, is this WebKit or V8?

[09:18:35.0000] <jgraham>
V8 I guess

[09:41:12.0000] <Ms2ger>
matjas, http://mathias.html5.org/specs/javascript/ is empty

[10:21:20.0000] <matjas>
Ms2ger: fixed

[10:21:29.0000] <Ms2ger>
Ta

[12:49:21.0000] <smokeyrat>
What is WhatWG again? Some kind of subset of W3C? And why did W3C rename to W3?

[12:50:46.0000] <TabAtkins_>
smokeyrat: WHATWG is an independent group that was formed after the W3C decided not to work on HTML.  The W3C later changed its mind, and now HTML is developed jointly by both groups.

[12:50:58.0000] <TabAtkins_>
The W3C didn't rename to W3.  It just uses "w3.org" as its url.

[12:58:56.0000] <smokeyrat>
You mean when they went all XHTML?

[12:59:16.0000] <TabAtkins_>
Yes.

[12:59:50.0000] <smokeyrat>
How come WhatWG has a say? Who made it? Usually, sensible groups that break out from a big organization have no say.

[12:59:59.0000] <smokeyrat>
That whole XHTML thing was bizarre.

[13:00:55.0000] <Ms2ger>
Because browsers vendors believe the WHATWG has a say

[13:01:20.0000] <othermaciej>
whatwg is also sort of a w3c community group but still in some ways independent

[13:01:51.0000] <smokeyrat>
I always found the W3C site to be a strange place. They used all kinds of bad practice in their own code, and even in examples, and at some point they used AdSense for Search on their site. They just gave a very lousy impression.

[13:02:46.0000] <othermaciej>
The w3c takes feedback on their site design

[13:06:08.0000] <TabAtkins_>
The WHATWG was created *by* browser vendors, who were annoyed at the W3C stopping work on HTML.  It was just an alternate location to keep developing HTML, since the W3C wasn't going to allow it anymore.

[13:06:39.0000] <smokeyrat>
W3C = George Lucas?

[13:06:54.0000] <smokeyrat>
"HTML doesn't exist. HTML is dead to us. XHTML is the future."

[13:07:12.0000] <smokeyrat>
But yeah, they seem to have gotten *some* sense back.

[13:07:24.0000] <smokeyrat>
Too bad they still didn't make a sensible doctype syntax for HTML 5.

[13:07:59.0000] <Hixie>
the doctype in HTML now was developed in the WHATWG actually :-)

[13:08:04.0000] <othermaciej>
<!DOCTYPE html> was invented by the WHATWG

[13:08:07.0000] <Hixie>
it's the shortest thing we could come up with that still worked

[13:08:17.0000] <Hixie>
(we tried pretty hard to find shorter ones)

[13:08:18.0000] <othermaciej>
w3c is not really pushing XHTML-everything any more

[13:08:27.0000] <smokeyrat>
Why in the freaking heck not <html version="5">?

[13:08:35.0000] <othermaciej>
though there is a new desire to try to shove anything and everything into HTML

[13:08:44.0000] <smokeyrat>
Why was it ever an incredibly long and impossible-to-memorize line?

[13:08:51.0000] <othermaciej>
<html version="5"> is not a doctype

[13:09:03.0000] <othermaciej>
the impossible-to-memorize line was due to SGML

[13:09:19.0000] <othermaciej>
"<!doctype html>" is the shortest string that triggers standards mode

[13:09:23.0000] <othermaciej>
that is the sole reason for it

[13:09:29.0000] <smokeyrat>
The only thing I know of that is in HTML 5 that I don't get (but I'm sure there is more) is some kind of client-side SQL database. Which makes no sense at all.

[13:09:32.0000] <othermaciej>
you really want to be in standards more, not quirks mode

[13:09:38.0000] <othermaciej>
otherwise html5 would have no doctype at all

[13:09:56.0000] <smokeyrat>
The entire concept of standards/quirks mode still exists in modern IEs?

[13:10:23.0000] <Hixie>
sadly yes

[13:10:29.0000] <Hixie>
it's even specced now

[13:10:44.0000] <Hixie>
we had no choice, there were pages depending on both modes

[13:10:46.0000] <TabAtkins_>
In all modern browsers, not just IE.

[13:10:55.0000] <Hixie>
(four modes, really)

[13:11:11.0000] <smokeyrat>
Dang ol' mess.

[13:11:16.0000] <Hixie>
welcome to html

[13:11:45.0000] <smokeyrat>
Not really new to it... I remember when IE 6 was bliss compared to Netscape Navigator 4.8.

[13:20:28.0000] <smokeyrat>
Some PNGs in IE are not properly transparent, it seems. Or maybe I should phrase it like this: the transparency of PNGs in today's current stable versions of IE, Opera, Firefox, Safari and Chrome seems to vary. Can somebody confirm this?

[13:24:22.0000] <c|oneman>
probably fine in latest version

[13:24:29.0000] <c|oneman>
I knoew early IE's hae trouble

[13:24:33.0000] <smokeyrat>
IE 9 here.

[13:24:35.0000] <smokeyrat>
Not fine.

[14:00:30.0000] <tsatse>
Hello I'd like to submit a proposal about the canvas API

[14:00:54.0000] <tsatse>
In my use case, I'd like to be able to manipulate canvases with lower bit-depth

[14:01:08.0000] <tsatse>
like say one colour channel per pixel

[14:01:50.0000] <tsatse>
that is useful if you need to store a lot of offscreen canvases but don't really need all that pixel information

[14:02:17.0000] <tsatse>
it would allow to stuff more offscreen canvases in the same memory chunk

[14:02:35.0000] <tsatse>
that could be useful in games, image manipulation apps

[14:02:39.0000] <tsatse>
etc.

[14:02:57.0000] <tsatse>
and it should improve the processing time too in some cases

[14:03:16.0000] <tsatse>
because it would mean less pixels to copy when blitting for example

[14:51:59.0000] <jamesr>
tsatse, interesting! take a look at http://wiki.whatwg.org/wiki/FAQ#Is_there_a_process_for_adding_new_features_to_a_specification.3F if you haven't seen it. the proposal sounds reasonable, but as an implementor i would wonder if there are enough use cases to justify adding the API surface area / implementation code

[14:52:24.0000] <jamesr>
tsatse, if you just do the naive thing and use a canvas with full bit-depth, how wasteful is it in practice? i.e. how many games / image manipulation apps / etc see significant overhead?

[14:52:37.0000] <jamesr>
and if there are a lot, is canvas 2d the right primitive for them?

[15:11:26.0000] <tsatse>
jamesr, thanks. someone suggested me to store lower-depth image data in a typed array and manually move the data back and forth between the canvas when needed

[15:11:39.0000] <tsatse>
which is reasonable in my use case

[15:12:28.0000] <tsatse>
jamesr, I was thinking this was lacking because it's so common in opengl for instance to manipulate different colour-depth textures

[15:13:22.0000] <tsatse>
jamesr, and thanks for the pointer to the faq. I was looking for that formal information

[15:13:38.0000] <jamesr>
it's more advice than formal information

[15:13:47.0000] <jamesr>
webgl does have support for different texture formats, etc

[15:14:04.0000] <tsatse>
but it's less supported than canvas-2d

[15:14:04.0000] <jamesr>
canvas 2d is a bit higher-level

[15:14:10.0000] <tsatse>
yes

[15:14:27.0000] <jamesr>
it didn't have any support for controlling filtering until very recently

[15:15:12.0000] <tsatse>
it's good to have those features coming to canvas 2d because it means a higher reach for users


2012-07-03
[00:29:19.0000] <matjas>
JS guys: \c escapes for control characters in regular expressions are non-standard too, right? I can’t find anything about it in ES5.1.

[00:33:10.0000] <zcorpan>
you mean like "\1" ?

[00:37:24.0000] <matjas>
zcorpan: no, like `/\cA/.test('\x01'); // true`

[00:37:35.0000] <matjas>
`\cA` till `\cZ`,

[00:37:39.0000] <matjas>
so to speak

[00:40:44.0000] <matjas>
e.g. \cA sends Ctrl+A, which is U+0001

[00:41:53.0000] <zcorpan>
i totally didn't know about this

[00:45:05.0000] <matjas>
if it’s so obscure even zcorpan had never heard of it, it must be non-standard

[00:48:09.0000] <zcorpan>
i don't see it in the syntax of regex literals in the spec and nothing in annex b

[00:48:35.0000] <matjas>
zcorpan: ok, thanks for confirming

[00:49:27.0000] <matjas>
not in ES3 either

[00:49:35.0000] <matjas>
/me specs

[01:17:31.0000] <matjas>
http://mathias.html5.org/specs/javascript/#control-character-escape-sequences

[01:21:26.0000] <zcorpan>
matjas: regular expression *literals* no?

[01:22:27.0000] <matjas>
zcorpan: fixed

[01:23:52.0000] <zcorpan>
matjas: saying "must be supported" doesn't really say what is supposed to happen. i think there needs to be something that extends the syntax or regular expression literals and then some prose that says what the semantic is of the added syntax

[01:25:56.0000] <zcorpan>
i guess add a line in RegularExpressionNonTerminator ::

[01:26:03.0000] <matjas>
zcorpan: right

[01:26:32.0000] <matjas>
zcorpan: i guess it should say more than just “literals” too, as RegExp('\\cA').test('\x01') should work too

[01:27:03.0000] <matjas>
/me ​’s TODO list explodes

[01:27:07.0000] <zcorpan>
oh, ok, i thought it was a quirk for just literals

[01:32:13.0000] <zcorpan>
hey, hold on

[01:32:25.0000] <zcorpan>
c ControlLetter

[01:32:33.0000] <zcorpan>
http://ecma-international.org/ecma-262/5.1/#sec-A.7

[01:33:09.0000] <matjas>
FFFFFUUUUU

[01:33:16.0000] <zcorpan>
LOL

[01:34:44.0000] <zcorpan>
http://ecma-international.org/ecma-262/5.1/#sec-15.10.2.10 has the semantic requirement

[01:35:48.0000] <hasather>
matjas: __lookupGetter__ says "If a getter for the prop property has been defined on the this object through the use of __defineGetter__, return the getter function; else, return undefined.". It should also return the getter if it was defined in the object literal. (Same for Setter).

[01:36:44.0000] <matjas>
hasather: defined in the object literal? how?

[01:37:18.0000] <hasather>
matjas: var o = {get p() { return 1; }}

[01:38:02.0000] <matjas>
woah, i thought that was Fx-only for some reason

[01:38:24.0000] <matjas>
this complicates things

[01:38:33.0000] <zcorpan>
odinho: the new graph looks better, but it's not really clear how the webapp makes a request to the UA. maybe the arrow should have an XHR code snippet to make it concrete (and more correct, since the app doesn't really make a HTTP request to the browser)

[01:38:37.0000] <hasather>
matjas: it's in http://es5.github.com/#x11.1.5

[01:39:42.0000] <odinho>
zcorpan: Hmm. Okay. That might work and still be short enough. Like     xhr.open('get', 'http://example.com/resource.txt')   xhr.send()

[01:39:49.0000] <odinho>
It's a bit longer.

[01:39:52.0000] <matjas>
hasather: thanks

[01:40:00.0000] <zcorpan>
odinho: yeah

[01:40:32.0000] <zcorpan>
odinho: dunno if the final arrow also should have a code snippet for symmetry

[01:41:02.0000] <zcorpan>
xhr.onload = function() { alert(xhr.response) }

[01:41:15.0000] <odinho>
zcorpan: Although it's conceptual the whole image, -- and CORS is really not tied to XHR.

[01:41:16.0000] <hasather>
matjas: isn't the name "Web ECMAScript" enough? This spec is not what people mean when they say "JavaScript". :) Plus, "JavaScript" is a registered trademark.

[01:41:48.0000] <matjas>
hasather: annevk suggested it, and i liked it

[01:42:03.0000] <matjas>
hasather: i’d argue this spec is what people mean when they say “javascript”

[01:42:32.0000] <zcorpan>
odinho: yeah but not everything matches the flow chart

[01:42:49.0000] <zcorpan>
odinho: e.g. <img> sans crossorigin doesn't use CORS

[01:43:10.0000] <matjas>
hasather: e.g. 'foo'.substr() is not ECMAScript, but it is JavaScript (who wouldn’t call it so?)

[01:43:13.0000] <hasather>
matjas: people say JavaScript and refer to ECMAScript usually, not the browser extensions

[01:43:37.0000] <odinho>
zcorpan:   get ( lyrics.com/hello.txt )   looks cleaner, though. -- But can't find a way to codify the return.

[01:43:54.0000] <odinho>
The return has a different color tho'. Because it's supposed to be a body etc.

[01:44:16.0000] <odinho>
I'm not sure if any code will make it clearer, I think it makes it harder to visualize at least :P

[01:44:27.0000] <hasather>
matjas: s/not the browser extensions/not *only* the browser extensions

[01:44:31.0000] <matjas>
hasather: #whatwg regulars people might refer to ECMAScript more often, but I’ll bet the average web developer means “anything that looks like ECMAScript and works in all browsers”

[01:44:40.0000] <odinho>
zcorpan: "request lyrics.com/hello.txt"  then? Instead of GET

[01:44:52.0000] <zcorpan>
odinho: wfm

[01:45:02.0000] <matjas>
hasather: this spec requires an ES implementation + the extras, so it’s the whole package

[01:45:05.0000] <odinho>
Okay :D It looks better than the old.

[01:45:06.0000] <hasather>
matjas: yea, but this only specs a subset

[01:45:25.0000] <hasather>
matjas: ok then :)

[01:45:29.0000] <odinho>
zcorpan: Updated.

[01:47:56.0000] <zcorpan>
odinho: it should really also be GET /hello.txt HTTP/1.1 :-)

[01:48:29.0000] <odinho>
zcorpan: No. I put my foot down to your correctness :P

[01:48:46.0000] <odinho>
shortness > correctness  for explanation picture :]

[01:48:51.0000] <zcorpan>
fine :-P

[01:49:03.0000] <zcorpan>
i'm still gonna grumble like an old fart

[01:49:05.0000] <odinho>
Well, technical non-relvevant detail correctness.

[01:49:18.0000] <odinho>
minus spelling errors.

[01:50:11.0000] <odinho>
Speaking of which, fixed matjas spelling fix also. Only waiting for anne now...

[01:58:54.0000] <matjas>
hasather: fixed — “If a getter for the prop property has been defined on the this object through the use of __defineGetter__ or the get syntax in the object initializer, return the getter function; else, return undefined.” + added examples

[03:34:34.0000] <jgraham>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1647

[03:34:50.0000] <jgraham>
Seems to be some disagreement about whether document.write causes load events

[03:35:15.0000] <jgraham>
In any case the spec is wrong (it's an infinite loop in the spec)

[03:35:39.0000] <jgraham>
(for some value of 'loop')

[03:39:02.0000] <jgraham>
matjas: __lookupGetter__ and friends probably want to be defined in terms of AccessorDescriptors as per ES5

[03:41:06.0000] <jgraham>
Look at e.g. http://es5.github.com/#x8.7.1 for clues

[03:51:33.0000] <volkmar>
 /win 18

[04:14:23.0000] <matjas>
jgraham: thanks, added to TODO. I was trying to keep things simple :'[

[04:19:57.0000] <annevk>
MikeSmith: can you add matjas' work to http://platform.html5.org/ please?

[04:20:09.0000] <MikeSmith>
sure

[04:20:16.0000] <MikeSmith>
what's the URL?

[04:20:35.0000] <matjas>
MikeSmith: http://mathias.html5.org/specs/javascript/

[04:20:52.0000] <matjas>
(WIP)

[04:21:24.0000] <annevk>
everything is WIP :)

[04:24:11.0000] <MikeSmith>
OK added

[04:24:16.0000] <niloy>
does this mean IE will have to implement __proto__?

[04:24:17.0000] <MikeSmith>
did like this: "JavaScript (based on ECMAScript)"

[04:24:57.0000] <gsnedders>
niloy: No. Like all web standards it doesn't force anyone to do anything.

[04:25:10.0000] <niloy>
but they should right?

[04:25:32.0000] <gsnedders>
niloy: IE will (sadly, IMO) be implementing __proto__ fairly soon, when ES6 support starts to come along.

[04:26:11.0000] <gsnedders>
(It's simply not tenable to not support it on mobile where so much content relies upon WebKit behaviour.)

[04:26:43.0000] <niloy>
great! I was just wondering if exposing __proto__ is a good thing

[04:26:59.0000] <gsnedders>
Not really. Immutable prototype chains would be far nicer.

[04:27:27.0000] <gsnedders>
However, the web relies on it, so it's just another inelegant thing that's part of the web platform.

[04:28:28.0000] <niloy>
oh... sad

[04:28:34.0000] <annevk>
MikeSmith: nice

[04:28:53.0000] <MikeSmith>
seemed like the simplest way to handle it

[04:29:14.0000] <gsnedders>
niloy: As currently spec'd in ES6 it's really horrible, a magic data property with magic in [[Get]] and [[Put]]. Would be far nicer to have an accessor property.

[04:31:00.0000] <niloy>
yeh

[04:32:36.0000] <matjas>
niloy: if they want to be ES6-compliant, yeah

[04:33:04.0000] <gsnedders>
matjas: They're going to do it regardless of whether it's in ES6.

[04:33:15.0000] <matjas>
gsnedders: oh, cool

[04:33:36.0000] <gsnedders>
matjas: It's simply not viable on mobile anymore with the ubquity of WebKit-expecting content.

[04:33:49.0000] <niloy>
I always felt taht not exposing __proto__ was one of the good things IE has done, apart from ajax

[04:34:56.0000] <gsnedders>
I'd still like to see first trying Object.defineProperty(Object.prototype, "__proto__", {get: function(o){return Object.getPropertyOf(o);},configurable:true});

[04:35:13.0000] <gsnedders>
(i.e., a read-only __proto__, which as such can be entirely described by ES5 semantics)

[04:35:34.0000] <gsnedders>
But apparently, in large part down to a couple of libraries, that isn't possible for the mobile web.

[04:41:36.0000] <benvie>
it seemed that part if it was because most __proto__ uses are in order to mutate the [[prototype]] of newly created objects, and a point of contention was the syntax that was to be introduced to cover that important use-case. At the end of the day, it was new, non-backward compatible syntax vs. __proto__ which works on mobile and works on now the majority of desktop environments by user percent.

[04:41:37.0000] <benvie>
Given the fact that using the new syntax is a syntax error in existing browsers, and that __proto__ is already in the wild and widely used, the discussion shifted toward formalizing what already exists

[04:43:26.0000] <gsnedders>
benvie: Certainly almost everyone would've rather have gotten rid of the setter if possible, even if that meant introducing new syntax.

[04:45:11.0000] <benvie>
that's where `<|` and `.{` came in and, I may have misinterpreted what I read, but those were what eventually fell by the wayside in favor of just using __proto__

[04:52:17.0000] <jgraham>
http://play-ttd.com/ <-- don't tell Hixie

[04:56:52.0000] <benvie>
I don't know what the status of this is, but it'd be ideal http://wiki.ecmascript.org/doku.php?id=harmony:proto_operator

[05:05:43.0000] <annevk>
jgraham: sweet

[05:05:53.0000] <annevk>
jgraham: now someone make it happen for Dungeon Keeper 2

[05:14:16.0000] <MikeSmith>
annevk: do we still want to publish the Notification LCWD this week?

[05:14:59.0000] <MikeSmith>
I see the message from Artur Ortega came in

[05:15:11.0000] <annevk>
MikeSmith: I guess we should address those comments first somehow

[05:15:15.0000] <MikeSmith>
OK

[05:15:32.0000] <annevk>
MikeSmith: need to find out what Apple/Google think

[05:15:37.0000] <MikeSmith>
annevk: I don't know how you plan to resolve the remaining i18n comments

[05:15:51.0000] <annevk>
MikeSmith: exposing direction seems sensible, language seems overkill

[05:16:12.0000] <MikeSmith>
annevk: OK, I'm planning to talk to Richard about that tonight

[05:16:41.0000] <jgraham>
annevk: Aren't you supposed to be depleting the crustacean population of western europe?

[05:16:44.0000] <annevk>
MikeSmith: at the end of the day, it's a lot of additional complexity to put into those systems and the benefit is small

[05:16:52.0000] <MikeSmith>
yeah

[05:16:55.0000] <annevk>
jgraham: we're about to go cycling :)

[05:17:16.0000] <jgraham>
Oh, fun :)

[05:42:19.0000] <MikeSmith>
jgraham: do you get notifications about testharness.js pull requests from github?

[05:43:31.0000] <jgraham>
MikeSmith: Yes

[05:43:36.0000] <MikeSmith>
k

[05:43:44.0000] <jgraham>
I got ones for the requests Lachy made anyway

[05:44:15.0000] <jgraham>
Kind of annoying that step 3 has to be "now commit using mercurial"

[05:44:23.0000] <MikeSmith>
yeah :(

[05:44:27.0000] <jgraham>
Well step 4 I guess

[05:44:40.0000] <MikeSmith>
the actually-important step

[05:44:57.0000] <jgraham>
1) Make patch 2) Upload to github 3) Get review 4) commit using mercurial to dvcs.w3.org

[05:45:07.0000] <guestlald>
can anyone tell me what "whatwg" is?

[05:45:36.0000] <odinho>
MikeSmith, jgraham: I want notification as well.

[05:45:44.0000] <tanepiper>
Anyone know if Simon Pieters is in here and awake?

[05:45:56.0000] <odinho>
tanepiper: --> zcorpan:

[05:45:58.0000] <guestlald>
I got this error on w3c validator: Bad value rpuPostId for attribute name on element meta: Keyword rpupostid is not registered.

[05:46:01.0000] <MikeSmith>
odinho: I can add you to the github w3c account

[05:46:07.0000] <odinho>
MikeSmith: Ohyea!

[05:46:12.0000] <tanepiper>
odinho: cool, thanks

[05:46:15.0000] <guestlald>
how can I fix that?

[05:46:20.0000] <MikeSmith>
odinho: what is your user name?

[05:46:24.0000] <odinho>
MikeSmith: Velmont

[05:46:28.0000] <MikeSmith>
hai

[05:46:36.0000] <tanepiper>
(p.s. if anyone else might be interested, http://webcomponents.github.com/ :)

[05:46:40.0000] <jgraham>
guestlald: a mailing list on which people, including several major browser vendors, discuss proposals for changes to the web; principally the HTML spec

[05:46:55.0000] <MikeSmith>
guestlald: you need to read the full text of that error message

[05:46:55.0000] <tanepiper>
(x-tags ftw)

[05:47:13.0000] <guestlald>
full text?

[05:47:36.0000] <guestlald>
<meta name='rpuPostId' content='1914'><meta name='rpuPlugin' content='v2.7'/>… Syntax of metadata name: A metadata name listed in the HTML specification or listed in the WHATWG wiki. You can register metadata names on the WHATWG wiki yourself.

[05:48:20.0000] <jgraham>
guestlald: rpuPlugin is not a valid value for meta @name

[05:48:43.0000] <guestlald>
its a plugin I use on my blog

[05:48:47.0000] <tanepiper>
zcorpan: when you are awake can you ping me please :)

[05:48:58.0000] <guestlald>
I didn't put that in meta

[05:48:59.0000] <MikeSmith>
guestlald: it's not registered anywhere

[05:49:04.0000] <jgraham>
/me expects zcorpan is awake

[05:49:17.0000] <jgraham>
Unless he is taking an afternoon nap

[05:49:18.0000] <zcorpan>
tanepiper: i'm awake

[05:49:19.0000] <odinho>
/me expects he is busy writing CORS inline tests

[05:49:23.0000] <guestlald>
should I just remove the plugin?

[05:49:24.0000] <jgraham>
Which would noy be a bad idea!

[05:49:43.0000] <MikeSmith>
guestlald: you can safely ignore that message if you want

[05:49:44.0000] <guestlald>
or should I not worry about those errors?

[05:49:48.0000] <guestlald>
okay

[05:49:49.0000] <MikeSmith>
yeah, that

[05:49:50.0000] <tanepiper>
zcorpan: cool - Patrick Lauke said you are probably the person I need to talk to re: web components in Opera

[05:49:53.0000] <odinho>
guestlald: If you don't want that message, you can fix the plugin, remove it, or ignore the message.

[05:50:08.0000] <guestlald>
doesn't slow down my site if I ignore message?

[05:50:20.0000] <odinho>
guestlald: No.

[05:50:29.0000] <jgraham>
tanepiper: zcorpan or I are probably good to talk to. Maybe odinho also

[05:50:33.0000] <guestlald>
thank you :)

[05:50:53.0000] <guestlald>
anyone know if a w3c validation irc chatroom exists?

[05:50:53.0000] <zcorpan>
tanepiper: what jgraham said :-)

[05:51:16.0000] <MikeSmith>
guestlald: no there is not one. but you can always ask here

[05:51:35.0000] <guestlald>
A charset attribute on a meta element found after the first 512 bytes. …134593'></script> <meta charset="UTF-8"/><title>You Are Stronger | The Optimis…

[05:51:56.0000] <guestlald>
http://www.anthonysabilities.com/blog/yas/

[05:51:58.0000] <odinho>
guestlald: You should put it at the top, right after <!doctype html>

[05:52:09.0000] <guestlald>
above head?

[05:52:24.0000] <odinho>
guestlald: You don't need head, but if you have <head> there do it right after <head> then.

[05:52:39.0000] <tanepiper>
jgraham: ah cool - yea, basically I;ve spoken to Daniel Buchner at Moz about this, and he's on board with getting the idea going - basically I'd like to do this 'not in a vacuum' but I've been tracking web components for a bit of time now, since Alex Russel talked about them - and after the GoogleIO talk and seeing the X-Tags implementation I've kicked off http://webcomponents.github.com/

[05:52:51.0000] <guestlald>
its already like this <head>   <meta charset="<?php bloginfo('charset'); ?>" />    <title><?php atom()->documentTitle(); ?></title>

[05:53:13.0000] <guestlald>
but my minify plugin adds more code before it

[05:53:17.0000] <odinho>
guestlald: So you have nasty plugins doing nasty things.

[05:53:59.0000] <guestlald>
<!DOCTYPE html><html dir="ltr" lang="en-US" xmlns:fb="http://www.facebook.com/2008/fbml"><head><link rel='stylesheet' href='http://www.anthonysabilities.com/blog/wp-content/plugins/wp-minify/min/?f=wp-content/plugins/custom-about-author/cab-style.css,wp-content/plugins/wp-single-post-navigation/css/single-post-navigation.css,wp-content/themes/mystique/css/core.css,wp-content/themes/mystique/css/style-blue.css,wp-content/themes/my

[05:54:14.0000] <tanepiper>
ideally i'd like to get a dev rel member on board from each of the browser vendors (sans Apple, because mehh) so we have people who can sanity check and make sure X-Tags, and moving on to proper web components, that are developer-proposed are sane :)

[05:54:19.0000] <guestlald>
ignore that code, didn't paste right lol

[05:54:40.0000] <MikeSmith>
odinho: added you https://github.com/w3c don't know if that means you will get the notifications or not (I don't understand how github handles the settings for that)

[05:54:50.0000] <zcorpan>
tanepiper: what's mehh about Apple?

[05:54:51.0000] <jgraham>
tanepiper: Oh well none of us are dev-rel

[05:54:53.0000] <guestlald>
but wpminify adds that link code

[05:55:00.0000] <guestlald>
before meta

[05:55:01.0000] <odinho>
MikeSmith: Cool, guess that should be enough.

[05:55:16.0000] <guestlald>
can I just put the meta before head?

[05:55:28.0000] <guestlald>
after <!DOCTYPE html>

[05:55:44.0000] <odinho>
tanepiper: ted is cool.

[05:55:44.0000] <tanepiper>
zcorpan: afaik they don't do much dev rel around Safari? I may be wrong, but that's the impression I get

[05:56:44.0000] <MikeSmith>
guestlald: I guess you could try omitting the head start tags and end tags completely. Though dunno that might make your plugin choke.

[05:57:04.0000] <guestlald>
<!DOCTYPE html> <html <?php language_attributes('html'); ?>>  <head>   <meta charset="<?php bloginfo('charset'); ?>" />    <title><?php atom()->documentTitle(); ?></title>

[05:57:07.0000] <zcorpan>
tanepiper: just so we understand each other, what do you mean with dev rel?

[05:57:09.0000] <guestlald>
change to

[05:57:21.0000] <guestlald>
<!DOCTYPE html> <meta charset="<?php bloginfo('charset'); ?>" /><html <?php language_attributes('html'); ?>>  <head>      <title><?php atom()->documentTitle(); ?></title>

[05:57:27.0000] <tanepiper>
The organisation exists around https://github.com/WebComponents - and i'm hoping to finaise the submission guidelines this week, get Moz's existing X-tags from http://mozilla.github.com/x-tag/ imported in as repos and get some discussion started on improving them

[05:57:29.0000] <guestlald>
?

[05:57:43.0000] <odinho>
tanepiper: Edward O'Connor is apple and very responsive and cool.

[05:57:54.0000] <guestlald>
I need the head tags

[05:57:55.0000] <MikeSmith>
guestlald: that's not going to validate but it will if you drop the head start tag and end tag

[05:58:01.0000] <MikeSmith>
guestlald: why?

[05:58:01.0000] <tanepiper>
zcorpan: good question i suppose - really we want to make sure that proposed web components are sane for the browser vendors

[05:58:13.0000] <guestlald>
wordpress put them there

[05:58:54.0000] <zcorpan>
tanepiper: apple people can have good feedback, i wouldn't exclude them up front :-)

[05:58:55.0000] <MikeSmith>
guestlald: then you're probably better off asking in a wordpress forum

[05:59:04.0000] <guestlald>
okay

[06:00:01.0000] <jgraham>
/me doesn't really know what x-tags is, but web components in general is quite vendor driven

[06:00:10.0000] <odinho>
tanepiper: We're spec and web tech people in Opera. We have something called devrel too, they are more working with developers, w/ some specs on the side.

[06:00:15.0000] <tanepiper>
ideally at least one member from each will have full admin rights on the organisation to help keep things in check, potentially do merges on repos from incoming patch requests (the plan is to initially start with X-tags so we can have the likes of http://mozilla.github.com/x-tag/demo.html#tabbox or http://mozilla.github.com/x-tag/demo.html#modal - get these speced)

[06:00:49.0000] <jgraham>
tanepiper: You have seen http://dvcs.w3.org/hg/webcomponents/raw-file/tip/explainer/index.html ?

[06:01:06.0000] <tanepiper>
jgraham: X-Tag is based on the Custom Elements portion of the W3 Web Components spec.

[06:01:26.0000] <tanepiper>
i.e. a polyfill developers can use now :)

[06:01:53.0000] <tanepiper>
it's less to do with the implementation and more to do with what developers will actually do with this stuff

[06:02:15.0000] <tanepiper>
e.g. one tab compoenent to rule them all vs 50,000,000 different tab components

[06:03:46.0000] <tanepiper>
odinho: that's maybe who I should be speaking to then, but I was directed here via a question on twitter :)

[06:04:22.0000] <odinho>
tanepiper: Yea, you spoke to the devrel people, they sent you to us, who have our hands a bit more dirty and deeper into the specs etc.

[06:06:31.0000] <zcorpan>
tanepiper: it's probably my fault you ended up here :-) but at least now we know a bit more about what you're after

[06:07:33.0000] <tanepiper>
zcorpan: :) Cool, yea this is a project very much about defining standard <element>'s - not how you guys are implementing what we'll be using as devs

[06:12:22.0000] <zcorpan>
tanepiper: yeah, our devrel people seems more suitable for that kind of thing

[06:12:41.0000] <zcorpan>
assuming one of them can commit to it

[06:14:22.0000] <zcorpan>
maybe they'll just go "LOL I dunno nothing about web components, buy me a slab of beer and i'll *consider* it"

[06:17:38.0000] <MikeSmith>
/me thought Patrick Lauke was in dev rel

[06:17:47.0000] <odinho>
MikeSmith: He is.

[06:18:07.0000] <MikeSmith>
so you dudes should pass the hot potato back to him

[06:18:12.0000] <MikeSmith>
or karlcow

[06:20:10.0000] <tanepiper>
zcorpan: i just as a MS developer evangelist 'trying to make html5 work for everyone' - he also asked what are web components :)

[06:20:21.0000] <tanepiper>
*le sigh*

[06:20:44.0000] <odinho>
But well, I don't know... For my part, I feel normal spec/webtech people is better suited than devrel. Although, I don't really know what different stuff devrel do, -- they do more Twitter and less tests than the rest of us at least.

[06:20:56.0000] <odinho>
Less arguing on mailing lists.

[06:21:40.0000] <tanepiper>
hmm well testing will be a big part of this - we don't want x-tags (custom elements) going out that will break things badly :D

[06:22:30.0000] <karlcow>
/me is reading the backlog for the hot potato

[06:24:54.0000] <karlcow>
MikeSmith: in fact there is one for validation or more exactly two ;) question of guestald #validator on freenode, #validator on w3c irc

[06:26:21.0000] <karlcow>
"Web Components Organisation — We don't need no stinkin' standards bodies" …

[06:30:01.0000] <AryehGregor>
Wait, so for computed styles, what does "As specified, but with relative lengths converted into absolute lengths" actually mean?

[06:30:11.0000] <AryehGregor>
What unit should be used for the absolute length, for instance?

[06:30:49.0000] <karlcow>
I have read the http://webcomponents.github.com/ a few days ago… and not sure what the goal is, specifically with the inflammatory tone at the start.

[06:31:21.0000] <zcorpan>
AryehGregor: i think it means converted to px, but that's not what it says is it...

[06:31:25.0000] <AryehGregor>
No.

[06:31:40.0000] <AryehGregor>
Hmm, you know, I guess it makes no difference except for in the CSSOM.

[06:32:20.0000] <AryehGregor>
Because CSSOM aside, computed value is only used for inheritance, so units are irrelevant.

[06:33:33.0000] <AryehGregor>
Seems like "convert to px" is how all UAs handle getComputedStyle() for line-height, at least.

[06:43:09.0000] <karlcow>
tanepiper: not sure what do you want exactly

[07:02:22.0000] <MikeSmith>
oh wow http://darobin.github.com/test-harness-tutorial/docs/using-testharness.html is nice

[07:02:27.0000] <MikeSmith>
bravo Berjon

[07:02:45.0000] <matjas>
gsnedders, Ms2ger et al. is there any way to predefine the slugs instead of letting Anolis create them?

[07:03:00.0000] <Ms2ger>
Slugs?

[07:03:08.0000] <matjas>
Ms2ger: e.g. I’d like http://mathias.html5.org/specs/javascript/#array.prototype.splice(start,-deletecount-,-item1-,-item2-,-…-)) to just say http://mathias.html5.org/specs/javascript/#array.prototype.splice

[07:03:53.0000] <tanepiper>
hasn't every good distruptive project started by having a go at the old way :)  And the goal is, browser vendors are finally giving developers the necessary tools to build our own tags and components, outwith defined specs, which means we'll start to see the likes of <x-modal>, <x-datepicker>, etc - but rather than have the current situation where you have 50 different date pickers to choose from - why not

[07:03:56.0000] <tanepiper>
instead have developers come together and try agree one to rule them all.  Then as browser vendors scower the web post-html5 final spec, looking for new tags to add nativly (as per the suggestion in the GoogleIO talk) - then all these should be speced out rather well

[07:04:05.0000] <Ms2ger>
<h3 id=array.prototype.splice>... should work, I think

[07:04:39.0000] <matjas>
Ms2ger: hah, just like that huh :)

[07:05:12.0000] <Ms2ger>
All thanks to gsnedders, that :)

[07:05:20.0000] <gsnedders>
DON'T BLAME ME.

[07:06:05.0000] <AryehGregor>
Hmm.  So why does the resolved value for transform-origin and perspective-origin resolve percentages to absolute lengths, but transform not?

[07:07:43.0000] <AryehGregor>
tanepiper, nobody has a problem with dev-written libraries, like jQuery or whatever.  Specs still need to exist to define things that are better implemented in browsers directly, such as for reasons of efficiency or because scripts lack some capabilities or because not everyone wants to pull in a 100k library to get basic functionality.

[07:08:20.0000] <Ms2ger>
I have more than enough problems with jQuery :)

[07:08:32.0000] <AryehGregor>
Yeah, but things like it are going to exist, and that's fine.

[07:10:56.0000] <tanepiper>
well why can't these libraries, instead of creating their own markup and functionality over and over for the same thing (and making it almost impossible to mix-and-match libraries too) why can't they use agreed upon custom components?  Isn't this what custom components are all about?

[07:12:08.0000] <jgraham>
Not really… I mean if you have two libraries that implement datepickers with different feature sets and expose different APIs, it will be quite hard to make them the same

[07:12:31.0000] <tanepiper>
By giving us the shadow DOM, etc, do you really want developers going out their way to write <mycustomvideo> tags? Or shouldn't you want them to use the pre-spec'ed <video> tag that's there - same with custom components.  Why does the web need to be filled with <tanes-datepicker>, <joes-datepicker>, etc (i used those tag names as descriptive only)

[07:13:00.0000] <jgraham>
Well I sort of agree with Hixie that I don't want them writing their own tags at all

[07:13:16.0000] <tanepiper>
and how will google, scraping the web, know that both these are really the same thing and lots of people are using them

[07:13:21.0000] <tanepiper>
anyway, meeting

[07:19:04.0000] <karlcow>
tanepiper: any point of extensions will be used… in a way which is not consensual. This is the goal of extensions. I understand your goals but it seems a case of let create yet another thing on the pile of all other things.

[07:19:49.0000] <zcorpan>
MikeSmith: the first code snippet is wrong though, should use root-relative url

[07:20:18.0000] <karlcow>
tanepiper: more exactly it is a case of XKCD 927 http://xkcd.com/927/

[07:20:52.0000] <MikeSmith>
zcorpan: oh. alert Berjon. But I think he's mostly away this week

[07:23:21.0000] <jgraham>
zcorpan: It is wrong for submissions to the W3C, but right for the text he gives there

[07:23:35.0000] <jgraham>
in that it is the easiest way to get started writing a test

[07:24:29.0000] <Ms2ger>
Oh yes, I should poke Lachy about using root-relative urls

[07:24:49.0000] <Lachy>
I did, didn't I?

[07:26:58.0000] <Lachy>
Ms2ger, which URLs did I use wrongly?

[07:27:16.0000] <Ms2ger>
Looks like you got it right

[07:27:26.0000] <Ms2ger>
Now I'm confused

[07:27:36.0000] <zcorpan>
jgraham: yeah, that's fair enough. i haven't finished reading it, but maybe it should have a note about submission to w3c since people who wanna write tests and submit to w3c might well read this as how to do it

[07:28:25.0000] <Ms2ger>
jgraham, also to get them imported into browser test suites, no?

[07:28:40.0000] <jgraham>
Ms2ger: Yes

[07:29:11.0000] <jgraham>
I am aware though that there is a tension between "how do I get going quickly" and "how do I write perfect tests"

[07:29:31.0000] <jgraham>
I think making step 1) "install a local web server" might be a bit off-putting

[07:29:42.0000] <zcorpan>
"assert_equals(actual, expected, description) checks that actual and expected have the same value (without necessarily being the same object)." seems wrong

[07:29:57.0000] <Lachy>
jgraham, depends. There are very easy ways to get set up with a local web server on some platforms

[07:30:21.0000] <jgraham>
Lachy: Sure. But as a prerequisite it increases the barrier to entry

[07:31:27.0000] <jgraham>
Although we should just make a node.js based server specifically for running the tests. People don't take projects seriously anymore unless step 1) is "install node.js and npm"

[07:32:05.0000] <Ms2ger>
Yeah, running V8, that sounds like a good idea :)

[07:32:10.0000] <jgraham>
(note: that was satire. If you think it is a serious suggestion, your doctor advises that you start monitoring your koolaid consumption)

[07:33:54.0000] <Ms2ger>
Btw, is Opera going to implement the Blob constructor at some point?

[07:35:25.0000] <jgraham>
Why wouldn';tt we?

[07:35:25.0000] <zcorpan>
Ms2ger: i think that's already done (might not be shipped yet)

[07:36:46.0000] <Ms2ger>
Doesn't look implemented in my opera-next :)

[07:37:05.0000] <Ms2ger>
Anyway, you may be interested in http://w3c-test.org/webapps/FileAPI/tests/submissions/Ms2ger/Blob-constructor.html

[07:38:10.0000] <zcorpan>
Ms2ger: nice, thanks

[07:38:19.0000] <jgraham>
Ooh cool

[07:39:09.0000] <Ms2ger>
(I wanted to write a test for passing null for dictionaries, but it got a bit out of hand)

[07:41:44.0000] <jgraham>
Ms2ger: (making a function called assert_blob that lives outside a test() is very weird. Rename it to test_blob?)

[07:44:18.0000] <Ms2ger>
Fair point, done

[07:45:29.0000] <jgraham>
Thanks

[07:54:34.0000] <tanepiper>
karlcow: you are the second person to link me to that (or maybe it was you on HN?) but the fact is, we're getting the tools to do this, it's happening - but lets not let idiots fuck it up for the rest of us - so what if there is 1 more competing spec, better than everyone coming up with their own individual snowflake spec :)

[07:54:52.0000] <tanepiper>
/me doesn't want the web to explode

[08:01:10.0000] <karlcow>
tanepiper: the Web will not explode. ☺ Not in that way at least.

[08:01:10.0000] <karlcow>
So you are targeting mainly the Web developers and encouraging a kind of best practices. Not sure how browsers devrel have any impacts or should have impacts on it. You were promoting Anarchy (let's say ala Proudhon and not Bakounine), it will then settle down by itself with the agreement of people participating to your project.

[08:15:27.0000] <tanepiper>
well a) devrel can help promote it beyond my efforts (as I mentioned the Moz guy I spoke to is keen) b) Despite the anarchy, there still needs to be final agreement on things that won't break browsers (and the anarchy is really more around the proposal stage, let devs go hog wild with suggestions, and through discussion the wheat and chaff can be seperated)

[08:22:38.0000] <karlcow>
ah not anarchy then

[08:23:53.0000] <Lachy>
Given a WebIDL dictionary, which defines a value like the following:

[08:23:56.0000] <Lachy>
dictionary X {

[08:23:56.0000] <Lachy>
  RegExp foo;

[08:23:58.0000] <Lachy>
}

[08:24:27.0000] <Lachy>
Is that handled by the ECMAScript conversion algorithm for object http://dev.w3.org/2006/webapi/WebIDL/#es-object ?

[08:24:53.0000] <Ms2ger>
Invalid, there's no such thing as "RegExp" in WebIDL

[08:25:01.0000] <Lachy>
hmm.

[08:25:09.0000] <Lachy>
That's what I was afraid of.

[08:27:24.0000] <Lachy>
Ms2ger, so, then to handle such a case in a spec, it would have to be defined as { object foo; } and then use prose to check if it's a regular expression?

[08:28:02.0000] <Ms2ger>
Either that or change WebIDL

[08:29:24.0000] <tanepiper>
karlcow: Not anarchy in the sense of 'lets throw stuff through the bank window' - that would be silly.  But also no situation where one person can say, decide a tag is no longer required despite the weight of developers disagreeing with that

[08:29:29.0000] <Lachy>
well, it's for an internal opera extensions feature that I'm doing QA for. So we could just accept it as being a willful violation too.

[08:29:34.0000] <tanepiper>
*cough*

[08:30:09.0000] <karlcow>
no no ☺ Anarchy has a precise meaning at least for me. Anarchy != mess

[08:30:31.0000] <karlcow>
https://en.wikipedia.org/wiki/Anarchism

[08:31:04.0000] <tanepiper>
Maybe Anarchy is wrong, but it's certainly not democratic (because Democracy is a suckers game) and no dictatorship because that never works

[08:32:50.0000] <tanepiper>
oh, and it's not against the 'state' here (despite my opening paragraph) - quite the opposite, it's working with the state :)

[08:33:13.0000] <tanepiper>
that'll get toned down once I actually sit down this week and finish some of the documentation around it

[08:45:42.0000] <AryehGregor>
/me can tell TabAtkins_ is the primary editor of a draft when it uses terms like "webapps" and "multicol" in spec prose

[08:46:02.0000] <AryehGregor>
You sure love your contractions.

[08:48:50.0000] <zcorpan>
Ms2ger: http://w3c-test.org/webapps/FileAPI/tests/submissions/Opera/

[08:50:27.0000] <jgraham>
Hmm, so is it well/correctly defined how the load event in an iframe interacts with the parent document

[08:50:52.0000] <jgraham>
The spec says "When there is an active parser in the iframe, and when anything in the iframe is delaying the load event of the iframe's browsing context's active document, the iframe must delay the load event of its document"

[08:51:36.0000] <jgraham>
And it has an active parser whilst the parser is not stopped or aborted

[08:51:45.0000] <jgraham>
But "stopped" links to "the end"

[08:52:22.0000] <jgraham>
So it isn't clear if "has an active parser" becomes untrue at the start or at the end of that set of steps

[08:55:18.0000] <jgraham>
The text makes it sound like the end of those steps, but that can't be true or you would get the load event on the parent document before the load event on the iframe

[08:55:53.0000] <tantek>
annevk - congrats on Fullscreen being published as a FPWD

[08:57:25.0000] <tantek>
http://www.w3.org/TR/2012/WD-fullscreen-20120703/

[08:58:56.0000] <zcorpan>
jgraham: maybe The end should have a step that says "now the document no longer *has an active parser*" that it can hook into, or so?

[09:00:07.0000] <jgraham>
zcorpan: Yeah, something like that

[09:02:03.0000] <tantek>
all links to terms like "browsing context" have been updated only in the static /TR/2012/ copy to link to http://www.w3.org/TR/2012/WD-html5-20120329/ instead of http://www.whatwg.org/specs/web-apps/current-work/multipage/browsers.html#browsing-context - no change in dvcs

[09:02:43.0000] <Ms2ger>
zcorpan, hmm, that wasn't there before :)

[09:02:44.0000] <tantek>
I wonder if that means that someone will do those (semi)automatically again next for the next TR snapshot, or if it doesn't matter that there's another TR snapshot.

[09:03:41.0000] <zcorpan>
Ms2ger: indeed

[09:04:28.0000] <tantek>
well at least the FPWD kicks off the call-for-exclusions

[09:04:48.0000] <TabAtkins_>
AryehGregor: Hehe.

[09:23:17.0000] <jwalden>
gsnedders: working on fixing one last issue with __proto__ as accessor, hopefully it'll be done today and I'll send mail

[09:23:43.0000] <jwalden>
thought I had it done yesterday, ran into a last-minute issue I didn't have time to debug before end of day

[09:32:30.0000] <gsnedders>
/me tries to find an empty workspace to try and finish his stuff off

[09:35:06.0000] <gsnedders>
jwalden: You going for origin-restricted? (i.e., the setter only works on same-global objects?)

[09:35:47.0000] <jwalden>
gsnedders: origin-restricted, or global-restricted?

[09:35:57.0000] <jwalden>
gsnedders: right now I'm going for no restrictions

[09:36:22.0000] <jwalden>
gsnedders: since there are no restrictions on __proto__ across globals right now, other than the ones enforced by the normal property lookup process

[09:37:53.0000] <jwalden>
I

[09:37:55.0000] <jwalden>
er

[09:44:14.0000] <gsnedders>
jwalden: I mean taking the __proto__ setter (through getOwnPropertyDescriptor) from another global. That's something you can't do today, because of the lack of getters.

[09:46:20.0000] <jwalden>
once the __proto__ setter escapes, it goes where it will

[09:50:05.0000] <gsnedders>
Right, so you're following the same route as JSC, against the wiki proposal and what's needed for SES.

[09:50:41.0000] <jwalden>
there's an actual wiki proposal for it right now?  wasn't aware of that

[09:53:36.0000] <gsnedders>
http://wiki.ecmascript.org/doku.php?id=strawman:magic_proto_property#normative-optional_proposal_currently_in_es5_spec_language

[09:53:51.0000] <gsnedders>
"If O is not an object from this context, throw a TypeError exception." notably.

[09:54:04.0000] <gsnedders>
(which JSC doesn't implement, and I really don't want to for reasons I've said before)

[09:57:04.0000] <gsnedders>
/me was going to poison Object.getOwnPropertyDescriptor(Object.prototype, "__proto__").set instead

[09:59:42.0000] <gsnedders>
(But that requirement, or something else making the setter useless or unavailable is needed for SES)

[10:31:42.0000] <gsnedders>
jwalden: I know Brendan was especially worried about letting the setter out in the wild, even with the wiki proposal. Poisioning .set seems the simple way to do this.

[10:32:19.0000] <jwalden>
brendan is always worried about backwards-compatibility, with some frequency to the point of excess, I think

[10:32:38.0000] <gsnedders>
It's not the BC concern here, but the extra security surface risk.

[10:32:40.0000] <jwalden>
witness why we still have E4X, although thankfully that's dying very very soon now

[10:33:11.0000] <gsnedders>
(Hey, if I could convince others, I'd be all for dropping the setter and trying to evangelize it through)

[10:33:18.0000] <gsnedders>
(But that looks less and less likely :()

[10:33:25.0000] <jwalden>
also for security, too, although with somewhat more justification there

[10:33:47.0000] <jwalden>
the more we simplify our algorithms and code, the easier it is to make changes with less security-conscious hesitance

[10:34:07.0000] <jwalden>
and if there's one thing we are doing all the time now, it's simplifying

[10:39:07.0000] <gsnedders>
jwalden: I've got .set poisoned now, hopefully get that in (along with moving it to an accessor) really soon.

[10:40:02.0000] <gsnedders>
jwalden: Basically, the SES people want to make it impossible to mutate prototypes within a global context, so other global's setter functions can't work on it.

[10:41:06.0000] <jwalden>
if we want to make __proto__ setting not work across contexts, that should be fairly simple to bolt on after the initial patch for us

[10:41:24.0000] <gsnedders>
 1 file changed, 14 insertions(+), 1 deletion(-)

[10:41:29.0000] <gsnedders>
I think that counts as simple. :)

[10:50:30.0000] <AryehGregor>
gsnedders, you made me curious.  Now I'm doing this on mozilla-central: hg log -I '**/*.cpp' -I '**/*.h' -p -u ayg | diffstat

[10:50:40.0000] <AryehGregor>
Have I made Gecko simpler or more complicated so far?  :)

[10:54:06.0000] <gsnedders>
Heh. Will, given there are commits from CVS-days which makes git believe I've written code I haven't…

[10:54:09.0000] <gsnedders>
*Well

[10:57:16.0000] <AryehGregor>
 155 files changed, 6085 insertions(+), 6713 deletions(-)

[10:57:17.0000] <AryehGregor>
Not bad.

[10:58:59.0000] <AryehGregor>
Ms2ger is:  2430 files changed, 52892 insertions(+), 116678 deletions(-)

[11:03:31.0000] <jwalden>
AryehGregor: "yes"

[13:31:20.0000] <jgraham>
annevk: I'm not sure what "a camping full of hippies" is. Campsite, perhaps?

[13:32:26.0000] <Hixie>
camping site?

[13:32:54.0000] <Hixie>
in french, and maybe other non-english languages, there's a single noun for "camping site" that sounds like "camping"

[13:33:20.0000] <jgraham>
Yeah, and in english there's "campsite" :)

[13:33:34.0000] <Hixie>
oops, i read that as "composite" -_-

[13:35:37.0000] <jgraham>
Hixie: Did you see the question about delaying the load event in iframes earlier

[13:35:43.0000] <Hixie>
i did not

[13:35:58.0000] <jgraham>
http://krijnhoetmer.nl/irc-logs/whatwg/20120703#l-774

[13:37:08.0000] <jgraham>
(except I think I might have said "end" and meant "start" but I'm not sure)

[13:37:59.0000] <jgraham>
(yes I did)

[13:39:46.0000] <Hixie>
jgraham: it doesn't link to "the end" it links to the definition of "stops parsing"

[13:39:54.0000] <Hixie>
jgraham: which is the condition that begins running the steps

[13:40:19.0000] <Hixie>
jgraham: the parser says when you "stop parsing" (typically in response to an EOF token)

[13:42:14.0000] <jgraham>
Hixie: But then the document would stop delaying the load event of the parent before those steps run, no?

[13:43:26.0000] <jgraham>
Whereas you have to have the load event on the child fire before the load event on the parent

[13:46:40.0000] <Hixie>
it's quite possible the spec has a bug :-)

[13:47:22.0000] <jgraham>
Well I can file a bug if you like :) I'm just checking first in case I am missing something subtle or indeed something blindingly obvious

[13:48:15.0000] <Hixie>
as specced, if the parser in a nested iframe is the only thing delaying the load event of the parent then the parent will likely fire 'load' before the child.

[13:48:20.0000] <Hixie>
as far as i can tell

[13:48:29.0000] <Hixie>
so if that's not compatible, yes, please send feedback

[13:48:30.0000] <Hixie>
(or file a bug)

[13:50:42.0000] <annevk>
jgraham: thanks will fix

[13:51:20.0000] <jgraham>
Hixie: OK. I'm pretty sure that frames always have to fire load events before their parents

[13:52:26.0000] <annevk>
jgraham: thought camping was pretty universal for some reason

[13:52:38.0000] <jgraham>
Filed bug 17680

[13:53:32.0000] <jgraham>
annevk: Also I feel that "stinged" should probably be "stang" or "stung"

[13:53:49.0000] <jgraham>
Perhaps the latter? My native-speaker intuition fails me

[13:55:25.0000] <annevk>
anything wrong with clean-shaven or shaven?

[13:55:32.0000] <annevk>
spellchecker is marking them as incorrect...

[13:55:48.0000] <annevk>
(it's also not happy with Nirvana)

[13:58:47.0000] <Hixie>
jgraham: yeah, i doubt i tested that when writing that part of the spec to be honest

[13:59:11.0000] <jgraham>
annevk: I think [clean-]shaven is fine

[13:59:38.0000] <annevk>
cool cool

[14:06:22.0000] <TabAtkins_>
clean-shaven is definitely fine.

[14:06:51.0000] <TabAtkins_>
shaven by itself seems less common to my ears, but there's nothing wrong with it.  I think it's usually conjugated as "shaved" when alone, though.

[14:14:48.0000] <guzzlefry>
Do many user-agents actually use the media attribute of the style tag?

[14:15:35.0000] <TabAtkins_>
Yes.

[14:15:56.0000] <guzzlefry>
How are they selected? Example:

[14:16:05.0000] <TabAtkins_>
Or wait, I don't know if <style media> is used.  <link rel=stylesheet media> definitely does.

[14:16:22.0000] <guzzlefry>
that applies also, since I'd be using it instead of <style>

[14:17:21.0000] <guzzlefry>
<link rel="stylesheet" type="text/css" media="all" href="yada.css" /> <link rel="stylesheet" type="text/css" media="screen" href="yada_screen.css" />

[14:17:27.0000] <TabAtkins_>
Media Queries just return a boolean - do they match or not?  Then all the ones that match are used.

[14:17:49.0000] <guzzlefry>
ok

[14:18:01.0000] <guzzlefry>
TabAtkins_: so the regular cascade order applies, nothing fancy?

[14:18:05.0000] <TabAtkins_>
Yup.

[14:18:09.0000] <guzzlefry>
cool, thank you

[16:37:23.0000] <smokeyrat>
Do you think that HTTP 2.0 will be released soon in a Windows 95 style commercial jippo with new versions of all major browsers coming out within a week with full support for it and where the protocol supports native and standard and super-safe payments totally free from transaction fees and where any webpage (such as our HTML 5 games) may ask if the user wants to insta-pay $0.1 to play the game,

[16:37:24.0000] <smokeyrat>
or any other amount, and when they click yes, it's transferred instantly to our bank account? Because that would be a very, very, very welcome addition to HTML 5 + CSS 3 + modern JavaScript.


2012-07-04
[17:03:55.0000] <TabAtkins_>
...what the hell?

[17:04:45.0000] <smokeyrat>
What the hell what?

[17:05:23.0000] <TabAtkins_>
Your bizarro train-of-thought spew.

[17:08:29.0000] <smokeyrat>
Bizarro?

[17:08:35.0000] <smokeyrat>
It is a very good idea.

[17:08:41.0000] <smokeyrat>
And something I've wanted for ages.

[17:09:16.0000] <tantek>
Is there a REQUEST-PONY header in HTTP 2.0?

[17:10:05.0000] <smokeyrat>
Why are you making fun of this?

[17:10:24.0000] <smokeyrat>
Do you have any idea how this would revolutionize the Internet?

[17:10:26.0000] <tantek>
is "Windows 95 style commercial jippo" not a humorous reference?

[17:10:34.0000] <smokeyrat>
I meant that it would be big-scale.

[17:10:39.0000] <smokeyrat>
Something that everyone hears about.

[17:10:41.0000] <tantek>
or "all major browsers coming out within a week with full support for"

[17:10:49.0000] <tantek>
(which has never happened on any spec/feature AFAIK)

[17:10:51.0000] <smokeyrat>
Yes, "very soon".

[17:11:06.0000] <tantek>
or "super-safe payments" (what does that mean?)

[17:11:09.0000] <smokeyrat>
Maybe they have had some kind of secret board meeting and agreed on this and worked behind the scenes.

[17:11:32.0000] <smokeyrat>
The point is that it would be truly global, instant, no hazzles and one-click.

[17:11:33.0000] <tantek>
or "totally free from transaction fees" is that a wish? or a requirement in a standard? or an expected market outcome?

[17:12:10.0000] <smokeyrat>
Or the transaction fee would have to be truly in %, so if the payment is $0.1, the fee truly is just a fraction of that. No minimum charges.

[17:12:19.0000] <smokeyrat>
This would open so many doors.

[17:12:39.0000] <tantek>
sounds like you're asking for some kind of micropayment standard

[17:12:43.0000] <tantek>
searching for that...

[17:12:47.0000] <smokeyrat>
I would sure as hell pay $0.1 to support a nice game or article, but I would not bother today to send that amount.

[17:13:05.0000] <smokeyrat>
No... I mean that it should be right into the HTTP protocol, thus in the native browser UI.

[17:13:41.0000] <tantek>
the HTTP protocol is a standard

[17:13:48.0000] <tantek>
and $0.1 would be a micropayment

[17:13:57.0000] <tantek>
so I repeat, sounds like you're asking for some kind of micropayment standard

[17:14:08.0000] <tantek>
looking up micropayment...

[17:14:10.0000] <tantek>
http://en.wikipedia.org/wiki/Micropayment#History

[17:14:25.0000] <tantek>
looks like there is bad news

[17:14:28.0000] <tantek>
"During the late 1990s, there was a movement to create microtransaction standards,[5] and the World Wide Web Consortium (W3C) worked on incorporating micropayments into HTML even going as far as to suggest the embedding of payment-request information in HTTP error codes.[4] The W3C has since stopped its efforts in this area,[4]"

[17:15:58.0000] <tantek>
If you'd like to propose a micropayment standard e.g. as an extension of HTTP, perhaps consider attempting to start a W3C Community Group on "HTTP micropayments" and see if you can get other people involved: http://www.w3.org/community/

[17:16:20.0000] <smokeyrat>
I can't propose anything. It will have to come from the big boys who people listen to.

[17:16:35.0000] <smokeyrat>
That they have not sat down in a conference room years ago is a mystery.

[17:16:51.0000] <smokeyrat>
There must be some kind of evil conspiracy requiring them to not make it easy for people to pay and charge.

[17:16:51.0000] <tantek>
most successful standards start from individuals with good ideas, not "big boys"

[17:17:23.0000] <tantek>
from the Wikipedia article, sounds like they did discuss it (whether in a conference room or not is unknown) and eventually stopped efforts.

[17:17:39.0000] <tantek>
no evil conspiracy is required for inaction. that's simply inertia.

[17:19:14.0000] <tantek>
anyway, if you want to make it happen, try starting a community group to see if there is interest from others. if no one else is interested, then it means it's not worth creating a standard for it. if there are others interested, then you have to do the hard work of figuring out how it would work.

[17:23:55.0000] <smokeyrat>
:|

[17:24:04.0000] <smokeyrat>
It's obvious that it's needed.

[17:27:31.0000] <TabAtkins_>
Not necessarily.  Micropayments have been attempted in lots of forms, both online and off.  They don't work very well.  Human psychology makes a huge deal of the difference between "free" and "not free", even if the "not free" is "so cheap it's practically free".

[17:28:04.0000] <TabAtkins_>
So, asking someone to pay 1 cent is psychologically equivalent to asking them to pay a few bucks, in terms of how willing the average person is willing to do it.

[17:28:41.0000] <TabAtkins_>
This might be changeable - it's not clear if this is something built into our psyche, or if it's a result of microtransactions never really being viable in the real world before the last few decades.

[17:28:56.0000] <TabAtkins_>
But resistance will be high regardless.

[17:30:00.0000] <TabAtkins_>
(There were simple experiments with real-world things, like giving people a choice between a 15c truffle and a 1c hershey kiss - 3/4 chose the truffle.  But if you drop the kiss to free, 2/3 choose it instead. Most people's brains break when they see "free".)

[17:32:20.0000] <tantek>
TabAtkins - the positive examples are music downloads (iTunes, Amazon) for < $1, and ringtones as well

[17:32:50.0000] <tantek>
as well as all those things people buy on Zynga games Farmville etc.

[17:32:59.0000] <tantek>
so there is some psychological precedent for it

[17:33:09.0000] <tantek>
especially in the use-case mentioned of (HTML5) games

[17:33:30.0000] <tantek>
why should Zynga make (nearly?) all the online/web game micropayment $?

[17:33:34.0000] <smokeyrat>
TabAtkins_: Well, the price example was just an example.

[17:34:05.0000] <TabAtkins_>
tantek: That's true.  We do have a lot more microtransactions in today's world.

[17:34:05.0000] <smokeyrat>
But even paying $1 is ridiculous today. It's eaten up by fees and is not anywhere near instant or hazzle-free.

[17:34:23.0000] <smokeyrat>
In fact, it's a nightmare of credit card info, codes, bank hardware dongles and fees.

[17:34:32.0000] <smokeyrat>
Or paying $10.

[17:36:54.0000] <tantek>
smokeyrat, Apple, Amazon, Zynga and others have figured out how to make < $1 payments work, perhaps you could research how they've done it and report back your results.

[17:37:35.0000] <tantek>
oh and please post your research on the web somewhere, preferably at a permalink with a creative commons / public domain license - that helps for work towards a standard

[17:39:56.0000] <smokeyrat>
For them there isn't anything to figure out. It's a closed system with lots of users.

[17:40:08.0000] <smokeyrat>
They enter cash and then they have some virtual credits to spend instantly.

[17:40:12.0000] <smokeyrat>
Couldn't be simpler.

[17:40:22.0000] <tantek>
smokeyrat - perhaps you could write up your analysis in a blog post

[17:40:37.0000] <TabAtkins_>
You're thinking of Zynga.  Tantek also pointed out Apple and Amazon, which do microtransactions directly off your CC.

[17:41:11.0000] <smokeyrat>
I was not aware that they do that. But that would require CC entering, etc.

[17:41:22.0000] <smokeyrat>
So basically they store your CC info. Is that novel?

[17:41:41.0000] <TabAtkins_>
Storing CC information is trivial - browsers do that already for autocomplete.

[17:41:45.0000] <smokeyrat>
That went well for Sony...

[17:41:52.0000] <TabAtkins_>
The difficulty appears to lay elsewhere.

[17:46:33.0000] <zewt>
TabAtkins_: well, browsers seem to attempt to *not* store CC info

[17:46:53.0000] <zewt>
(but that's a special case)

[17:53:42.0000] <TabAtkins_>
/me just realized that </figcaption> isn't an implied end tag. :/

[19:12:56.0000] <guzzlefry>
should the xml:lang attribute be used in HTML5?

[19:13:21.0000] <MikeSmith>
guzzlefry: no not in text/html documents

[19:13:31.0000] <zewt>
just use @lang

[19:14:04.0000] <guzzlefry>
k, thanks

[22:01:10.0000] <guzzlefry>
Does HTML5 have implicit end tags like older versions (pre-XHML)?

[22:02:26.0000] <guzzlefry>
s/XHML/XHTML

[22:23:21.0000] <daurnimator>
do you guys know of some way to test a http client? I wrote one yesterday and what to check it against strange corner cases. connection losses, timeouts, shutdowns; broken ssl handshakes and unexpected re-negotiations. etc

[22:40:38.0000] <MikeSmith>
daurnimator: I thought the HTTPbis WG had a test suite

[22:43:51.0000] <daurnimator>
MikeSmith: not much of one... and definiently not one that really tests the underlying tcp connection

[22:44:07.0000] <MikeSmith>
I see

[22:44:21.0000] <MikeSmith>
maybe could use some tests from browser projects

[00:40:21.0000] <zcorpan>
TabAtkins_: @import should maybe use the same rules as <link> when it comes to requiring text/css type and quirks mode

[00:40:43.0000] <zcorpan>
TabAtkins_: opera seems to use the same logic while chrome and firefox allow text/plain also in standards mode

[00:46:24.0000] <zcorpan>
TabAtkins_: hmm, chrome seems to allow text/html cross-origin @imports which i think is a security problem

[00:46:32.0000] <zcorpan>
in quirks mode

[02:00:32.0000] <jgraham>
The internet is funny. CERN live streams a seminar intended for research scientists at which Higgs boson results are announced. Internet complains that a) research level particle physics is hard to follow and b) they used comic sans in the presentation

[02:01:13.0000] <shwetank>
they wanted an apple keynote of the whole thing

[02:02:39.0000] <moo-_->
are we confident that Higgs is smaller than pixels in retina display?

[02:11:20.0000] <jgraham>
shwetank: I guess. Sadly the real world doesn't work like that

[02:11:56.0000] <jgraham>
(also I feel great now I got to describe academica as "the real world")

[02:18:24.0000] <Lachy>
moo-_-, that depends, how do you define the mass of a pixel in the retina display?

[02:22:05.0000] <Philip`>
Lachy: Mass of screen divided by number of pixels?

[02:24:00.0000] <Lachy>
do you include the mass of the glass covering it, or just the mass of the phosphorus material? Or you could define it as the energy emitted from the pixel when it's on.

[02:34:29.0000] <Ms2ger>
var scrollTo = window.scrollTo;

[02:34:29.0000] <Ms2ger>
w(scrollTo);

[02:34:29.0000] <Ms2ger>
What should be logged?

[02:37:05.0000] <jgraham>
Ohn interesting

[02:38:59.0000] <jgraham>
Presumably it depends on whether var scrollTo creates a new global given that one already exists

[02:39:05.0000] <jgraham>
on the window object

[02:39:19.0000] <Ms2ger>
I, too, would presume that :)

[02:39:29.0000] <Ms2ger>
/me looks for gsnedders 

[02:41:24.0000] <ashemedai>
ObligatoryVenting: not being able to discern at on-hover whether a link opens a JavaScript window on top of your existing window, loads content in your current window, or if it opens a new window/tab.

[02:41:33.0000] <ashemedai>
So frustrating and unclear at times.

[02:41:35.0000] <Ms2ger>
They used Comic Sans MS?!

[02:41:44.0000] <ashemedai>
Ms2ger: Yes, ultimate troll :)

[02:41:52.0000] <Ms2ger>
What, me?

[02:41:54.0000] <Ms2ger>
Thanks :)

[02:41:58.0000] <ashemedai>
The Comic Cans

[02:41:59.0000] <ashemedai>
Sans

[02:43:37.0000] <ashemedai>
Ultimate Science requires levity in the form of a funny font

[02:45:03.0000] <Philip`>
Surely real science should be conveyed with a marker pen and transparency and OHP

[02:45:28.0000] <Ms2ger>
Chalk, dammit

[02:45:37.0000] <jgraham>
Using comic sans for presentations is a tradition that dates back for as long as people have been making computer-based slideshows

[02:46:26.0000] <Ms2ger>
Though, the ultimate troll would be using Comic Sans MS in a LaTeX presentation

[02:46:44.0000] <jgraham>
In an ideal world each slide is a couple of words plus a graph anyway

[02:47:12.0000] <jgraham>
Nah, the ultimate troll was the guy that used 12pt Times New Roman

[03:00:34.0000] <Philip`>
Nothing beats a graph with red-squiggly-underlined words that's clearly been screenshotted from Excel

[03:03:01.0000] <odinho>
annevk: Did you see the svg?

[03:03:12.0000] <annevk>
odinho: no missed it

[03:03:25.0000] <Ms2ger>
Nothing beats a maintainer for those canvas tests...

[03:05:12.0000] <annevk>
odinho: pointer/

[03:05:13.0000] <annevk>
?

[03:05:48.0000] <odinho>
Computer strugling

[03:06:09.0000] <odinho>
omfg. some memory leak there!

[03:06:19.0000] <Ms2ger>
Opera had memory leaks first

[03:06:32.0000] <odinho>
Ms2ger: You bet'cha

[03:06:53.0000] <odinho>
Ms2ger: But then Fx came and owned us :-(

[03:06:57.0000] <Ms2ger>
:D

[03:07:55.0000] <odinho>
annevk: http://odin.s0.no/tmp/Hodges_odinho-CORS-Diagram-SimpleRequest.svg

[03:08:27.0000] <annevk>
that looks like the Core-Doxcs t-shirt :)

[03:08:35.0000] <odinho>
...

[03:08:38.0000] <odinho>
Did I save over it :P

[03:08:39.0000] <Ms2ger>
Zing

[03:09:16.0000] <annevk>
euh

[03:09:18.0000] <annevk>
Core-Docxs

[03:11:00.0000] <annevk>
minor announcement: http://annevankesteren.nl/2012/07/leaving-opera

[03:11:14.0000] <odinho>
annevk: http://odin.s0.no/tmp/Hodges_odinho-CORS-Diagram-SimpleRequest.svg Tada

[03:11:38.0000] <annevk>
my IRC client still displays the same picture funnily enough

[03:11:43.0000] <odinho>
Hm. Although I used Ubuntu font, so maybe I should use something neutral like arial if it messes up other places.

[03:11:55.0000] <odinho>
annevk: IRC client with pictures? Wat?

[03:12:06.0000] <odinho>
My lingogi shows the correct one now.

[03:12:22.0000] <annevk>
LimeChat shows images inline

[03:12:26.0000] <annevk>
including SVG

[03:13:20.0000] <annevk>
odinho: instead of .com it should prolly say .example or something

[03:13:49.0000] <annevk>
odinho: also GET lyrics.com/hello.txt is wrong

[03:13:58.0000] <annevk>
it's GET /hello.txt

[03:14:03.0000] <annevk>
Host: lyrics.com

[03:14:53.0000] <Philip`>
/me thinks using a non-conventional TLD like ".example" would make the diagram harder to understand

[03:15:34.0000] <odinho>
annevk: It's not supposed to be real. It also misses the HTTP/1.1 stuff

[03:15:40.0000] <odinho>
I agree with Philip`

[03:15:49.0000] <odinho>
I'm just trying to make it easy for them newbies :P

[03:15:53.0000] <odinho>
But, lunch out in the sun!

[03:16:00.0000] <Philip`>
(since "webapps.com" is clearly a domain name, but "webapps.example" isn't, so it'd have to be written like "http://webapps.example" to make it unambiguous but then it'd still look kind of weird)

[03:16:34.0000] <odinho>
won't anybody think of the children!!11

[03:16:38.0000] <annevk>
yeah but there are some unwritten rules against using actual domain names

[03:17:00.0000] <smaug____>
annevk: Leaving Opera? What will you do?

[03:17:19.0000] <annevk>
smaug____: nothing much for now :)

[03:17:34.0000] <smaug____>
oh

[03:17:48.0000] <smaug____>
that is possibly a good choice

[03:18:09.0000] <smaug____>
and after that, you'd be welcome to Mozilla ;)

[03:20:26.0000] <smaug____>
annevk: does this change affect to your role as a spec editor?

[03:26:52.0000] <annevk>
smaug____: dunno, hope not too much, I like writing specs

[03:27:25.0000] <MikeSmith>
annevk: http://www.w3.org/Consortium/Recruitment/#techeditor

[03:28:04.0000] <annevk>
MikeSmith: the link makes it sound interesting, but then there's "for HTML Working Group" on the actual page :)

[03:28:30.0000] <smaug____>
what... " lead editor for the HTML Working Group"

[03:30:21.0000] <annevk>
the HTML WG is the new old HTML WG

[03:31:53.0000] <MikeSmith>
annevk: trail of broken hearts behind you. First you leave the CSS WG, then the HTML WG, and now this

[03:32:55.0000] <annevk>
MikeSmith: quit while you're ahead?

[03:33:06.0000] <MikeSmith>
heh

[03:33:11.0000] <MikeSmith>
are you ahead?

[03:33:41.0000] <MikeSmith>
annevk: leave them wanting more

[03:33:51.0000] <annevk>
dunno really, but soon I'm an unemployed bum and I like that idea

[03:35:24.0000] <annevk>
CSS WG was too much abuse, HTML WG was too much process abuse and irrelevant anyway given the WHATWG

[03:36:46.0000] <annevk>
I had no conflicts with Opera though, just felt like trying something else

[03:36:47.0000] <Ms2ger>
WCSSTWG?

[03:39:35.0000] <annevk>
CSS WG was actually also mostly process nonsense, like how to divide WDs rather than how to solve problems

[03:39:56.0000] <annevk>
combined with a madman in charge that just gives problems

[03:47:30.0000] <MikeSmith>
maybe the CSS WG would work better if it were more like the WebApps WG

[03:48:20.0000] <MikeSmith>
WebApps WG seem to somehow be able to get stuff done without anybody trying to micro-manage it all and layer further process on it

[03:49:19.0000] <MikeSmith>
anyway, question: Is there really no HTTP test suite anywhere?

[03:49:56.0000] <annevk>
IETF and testing? mwaha

[03:50:07.0000] <zcorpan>
is there a HTTP spec that doesn't suck?

[03:50:10.0000] <annevk>
jreschke has a couple of tests for some features though

[03:50:21.0000] <annevk>
zcorpan: rhetorical?

[03:51:41.0000] <MikeSmith>
it seems whacky that IETF culture does not see any need for requiring test suites for RFCs

[03:52:02.0000] <Ms2ger>
Why would you need tests? They're only requesting comments

[03:52:08.0000] <MikeSmith>
heh

[03:52:17.0000] <MikeSmith>
but don't browser projects have HTTP tests?

[03:52:22.0000] <annevk>
for RFCs it makes sense I guess, but not STDs

[03:52:32.0000] <jgraham>
We should rename W3C specs RFTs

[03:52:39.0000] <MikeSmith>
I would think Mozilla would at least have some HTTP tests

[03:52:50.0000] <jgraham>
I expect everybody has some HTTP tests

[03:53:03.0000] <jgraham>
Possibly not packaged as a suite

[03:53:07.0000] <Ms2ger>
Mm

[03:53:08.0000] <jgraham>
Or easy to reuse

[03:53:22.0000] <Ms2ger>
And ours run JS on the server

[03:54:12.0000] <Ms2ger>
So unless everyone wants to run a Gecko server... ;)

[03:55:34.0000] <MikeSmith>
wonder if it's time to write an HTTP spec

[03:56:11.0000] <ashemedai>
annevk: Eh? Unemployed? How so?

[03:56:45.0000] <ashemedai>
MikeSmith: As an update to HTTP or merely a more concise up-to-date spec?

[03:57:05.0000] <jgraham>
ashemedai: Oslo ran out of bacon cheeseburgers

[03:57:25.0000] <MikeSmith>
ashemedai: as a real spec to document how it actually works

[03:58:03.0000] <ashemedai>
jgraham: Noooooes!

[03:58:04.0000] <annevk>
ashemedai: well you know, when you quit your job and there's no direct follow-up, that's what happens

[03:58:06.0000] <annevk>
:)

[03:58:22.0000] <ashemedai>
MikeSmith: I can tell you that I know a bunch of folks who would be Real Happy(tm) if you do that.

[03:58:54.0000] <annevk>
MikeSmith: gsnedders worked on a part of that a long time ago

[03:58:54.0000] <ashemedai>
annevk: Ahh ok, it was a choice on your end. :)

[03:59:09.0000] <annevk>
MikeSmith: but never finished it unfortunately

[03:59:22.0000] <MikeSmith>
interesting

[03:59:29.0000] <MikeSmith>
I kind of vaguely remember that now

[03:59:33.0000] <annevk>
ashemedai: ah yeah, it was, and I'm welcome back ;)

[03:59:46.0000] <ashemedai>
annevk: Hehe, what kind of work do you want to do though?

[04:01:08.0000] <jgraham>
Here's a thing

[04:01:42.0000] <jgraham>
If I have a document with two framesets, is there a way to set an onload listener and tell *which* frameset loaded

[04:02:21.0000] <annevk>
jgraham: you mean two <frame>s?

[04:02:22.0000] <jgraham>
Like <frameset rows="*" onload="something"><frameset></frameset>

[04:02:29.0000] <jgraham>
I mean two framesets

[04:02:39.0000] <jgraham>
You can nest framesets

[04:02:57.0000] <annevk>
sure, but they do not load anything

[04:03:18.0000] <zcorpan>
<frameset onload> reflects window.onload

[04:03:32.0000] <zcorpan>
right?

[04:03:46.0000] <jgraham>
Oh wait

[04:03:52.0000] <jgraham>
I am misunderstanding this test

[04:04:07.0000] <annevk>
zcorpan: right

[04:04:21.0000] <jgraham>
I thought it was giving a load event on the frameset when all the subframes had loaded

[04:04:25.0000] <jgraham>
But it isn't

[04:04:27.0000] <jgraham>
Ignore me

[04:06:04.0000] <jgraham>
Right, I now have a less confusing, non-racy test

[04:27:36.0000] <Ms2ger>
annevk, so did you leave because Facebook is going to buy Opera? :)

[04:28:45.0000] <jgraham>
Maybe facebook bought annevk

[04:31:10.0000] <zcorpan>
he left. surely that means something.

[04:31:32.0000] <icaaq>
zcorpan: hehe

[04:32:48.0000] <Ms2ger>
Is he not yet allowed to disclose that he's going to work for facebook?

[04:38:23.0000] <Ms2ger>
matjas, thanks

[04:39:12.0000] <matjas>
Ms2ger: I’ll leave a comment with some tests results in other browsers

[05:08:42.0000] <annevk>
Ms2ger: I think those rumors started after I quit

[05:12:37.0000] <Ms2ger>
Aha!

[05:17:00.0000] <annevk>
the plot thickens? lol

[05:26:57.0000] <matjas>
jgraham: “eval.apply({}, code) should throw EvalError” → no browser seems to do this; am I testing this incorrectly?

[05:29:39.0000] <zcorpan>
isn't EvalError unused?

[05:51:28.0000] <jgraham>
matjas: I don't remember. Possibly everyone changed that behaviour?

[05:51:53.0000] <matjas>
jgraham: that’s good news, then

[05:53:08.0000] <MikeSmith>
do any engines other than gecko have support for the opus audio codec yet?

[06:10:21.0000] <zcorpan>
hsivonen: MikeSmith: http://validator.keegan.st/

[06:11:14.0000] <MikeSmith>
zcorpan: wow

[06:11:20.0000] <MikeSmith>
that's pretty nice

[06:11:40.0000] <matjas>
woah i know this guy

[06:11:48.0000] <MikeSmith>
we should add this upstream

[06:14:03.0000] <Ms2ger>
assertThrows(function(){ worker.port.onmessage = 1 }, [], TypeError, {});

[06:14:14.0000] <Ms2ger>
zcorpan, sounds bogus to me ^

[06:14:49.0000] <zcorpan>
Ms2ger: it probably matched the spec when i wrote it

[06:15:04.0000] <zcorpan>
now it should set to null, right?

[06:15:11.0000] <Ms2ger>
Right

[06:26:02.0000] <matjas>
wait, does testharness.js not have an equals() method that compares objects?

[06:26:21.0000] <Ms2ger>
How would you compare them?

[06:26:59.0000] <matjas>
iterate over all keys/values in both objects and see if there’s a difference

[06:27:16.0000] <matjas>
would cover most cases

[06:28:34.0000] <matjas>
QUnit does it like this: https://github.com/jquery/qunit/blob/b0c3da8d09b1dcbe5dd0877e27386f4b9e8050a1/qunit/qunit.js#L1343-1525

[06:30:10.0000] <jgraham>
matjas: Because there are a huge number of weird edge cases

[06:30:25.0000] <jgraham>
e.g. loops in the object graph

[06:30:48.0000] <jgraham>
and wanting to compare particular keys with something other than assert_equaks

[06:30:54.0000] <jgraham>
*assert_equals

[06:31:16.0000] <matjas>
jgraham: why not add a simple method and document that it won’t work for every possible use case?

[06:31:38.0000] <jgraham>
Because people will be surprised when it doesn't work for their use case

[06:32:29.0000] <matjas>
fair enough

[06:32:33.0000] <jgraham>
And because it's hard to get right even in simple cases

[06:32:56.0000] <jgraham>
and doesn't promote good quality tests (a single assert can only give a single error message)

[06:33:18.0000] <jgraham>
(what was your use case?)

[06:34:28.0000] <matjas>
comparing fn.arguments with the arguments object in a function (it should be ===)

[06:34:47.0000] <matjas>
i just use assert_equals(argumentsProp === argumentsObject, true) now

[06:35:00.0000] <matjas>
but in other cases that wouldn’t be possible

[06:35:08.0000] <jgraham>
But if they are === then assert_equals should work

[06:35:24.0000] <jgraham>
assert_equals(argumentsProp, argumentsObject)

[06:35:41.0000] <matjas>
oh, right

[07:19:26.0000] <matjas>
do any of you know if this is a compatibility measure, or just a bug that all engines appear to have? https://mail.mozilla.org/pipermail/es-discuss/2012-July/023888.html

[07:22:38.0000] <Ms2ger>
Heh, people still care about acid2

[07:23:39.0000] <gsnedders>
Ms2ger: var scrollTo = window.scrollTo; doesn't create a new variable binding

[07:24:30.0000] <Ms2ger>
In the meantime, I found https://bugs.ecmascript.org/show_bug.cgi?id=78#c0

[07:26:06.0000] <gsnedders>
Ms2ger: I believe var self = 1; would mutate Window.prototype.self to equal 1.

[07:26:34.0000] <gsnedders>
(I could be wrong, I've never been that happy with this stuff)

[07:27:20.0000] <Ms2ger>
That sounds rather unlikely

[07:27:30.0000] <gsnedders>
That's what I think.

[07:27:54.0000] <jgraham>
Huh, how would that work?

[07:27:57.0000] <gsnedders>
But you get a reference for "scrollTo", which is a reference to Window.prototype.scrollTo, and then put a value on that reference.

[07:28:00.0000] <gsnedders>
Or am I being stupid?

[07:28:50.0000] <zcorpan>
that sounds rather likely :-P

[07:28:56.0000] <gsnedders>
Yes, it does. :P

[07:29:05.0000] <Ms2ger>
No comment :)

[07:37:41.0000] <matjas>
Ms2ger: heh, that’s the link from http://mathias.html5.org/specs/javascript/#top-level-var-statements

[07:38:19.0000] <Ms2ger>
\o/

[07:45:35.0000] <Ms2ger>
matjas, <jorendorff> Ms2ger: if you look at ES5.1 �10.5, arguments object is created in steps 6-7; variables aren't bound until step 8

[07:48:10.0000] <matjas>
Ms2ger: cool, filed it here: Thanks for confirming. I filed it here: https://bugs.ecmascript.org/show_bug.cgi?id=440.

[07:48:15.0000] <matjas>
whoops copypasta

[07:49:27.0000] <Ms2ger>
So the spec is correct, right?

[07:50:27.0000] <Ms2ger>
Also...

[07:50:29.0000] <Ms2ger>
A test!

[07:50:50.0000] <matjas>
Ms2ger: the spec text is wrong, but the steps are right

[07:51:33.0000] <Ms2ger>
Oh, �10.6 disagrees with �10.5?

[07:52:53.0000] <matjas>
unless i misread the text

[07:53:59.0000] <matjas>
“an `arguments` object is created unless […] the identifier `arguments` occurs as […] the Identifier of a VariableDeclaration […] contained in the function code.” seems to conflict with 10.5

[07:54:24.0000] <Ms2ger>
Looks like we all agree on that point :)

[07:59:40.0000] <matjas>
so in the case of function fn() { var arguments = undefined; return typeof arguments; }; fn(); // "undefined"

[08:00:48.0000] <matjas>
the `arguments` object is created in step 7

[08:01:03.0000] <matjas>
but then overwritten by the variable in step 8, iiuc

[08:02:17.0000] <matjas>
heh

[08:04:34.0000] <SimonSapin>
I’m now passing Acid2 :) http://weasyprint.org/#acid2

[08:05:21.0000] <matjas>
SimonSapin: nice work

[08:07:11.0000] <SimonSapin>
matjas: thanks

[08:07:56.0000] <SimonSapin>
We’re a bit late to the party but we only started 14 months ago

[08:15:14.0000] <Ms2ger>
matjas, no, it's not overwritten due to 8.c

[08:18:22.0000] <matjas>
Ms2ger: oic. but how is `var arguments;` different from `var arguments = undefined;`, then?

[08:18:37.0000] <matjas>
aren’t both VariableDeclarations?

[08:18:45.0000] <Ms2ger>
`var arguments = undefined;` is `var arguments; argument = undefined;`

[08:18:58.0000] <Ms2ger>
<jorendorff>       function fn() { alert(typeof arguments); var arguments = undefined;  alert(typeof arguments); }

[08:18:59.0000] <Ms2ger>
<jorendorff> alerts "object" then "undefined"

[08:18:59.0000] <Ms2ger>
<jorendorff> All the steps in 10.5 happen before the function body starts running.

[08:21:40.0000] <Ms2ger>
Is anyone else doing proxies, btw?

[08:22:59.0000] <zcorpan>
/me awaits an xkcd about higgs

[08:23:12.0000] <jgraham>
Ms2ger: Not yet afaik

[08:23:30.0000] <annevk>
zcorpan: me too, I wonder if it'll be about the font :)

[08:25:06.0000] <matjas>
Ms2ger: thanks! http://ecma-international.org/ecma-262/5.1/#sec-12.2 VariableDeclaration : Identifier vs. VariableDeclaration : Identifier Initialiser

[08:31:43.0000] <matjas>
i guess i was thrown off by this: function fn() { alert(typeof arguments); var arguments; alert(typeof arguments); }; fn(); // alerts "object" twice

[08:38:00.0000] <Ms2ger>
That's consistent with what I said, I think :)

[09:03:34.0000] <matjas>
sure, it all makes sense in YOUR head!

[09:05:39.0000] <Ms2ger>
:D

[09:08:34.0000] <matjas>
Ms2ger: http://ecma-international.org/ecma-262/5.1/#sec-12.2 says “Variables are initialised to `undefined` when created” — so I expected `var foo;` to set it to undefined upon creation

[09:09:12.0000] <Ms2ger>
But 8.c means there's no variable

[09:11:23.0000] <matjas>
…as varAlreadyDeclared is false, because the `arguments` object is already created at that point

[09:11:37.0000] <matjas>
so only when you do `[var] arguments = undefined` it overwrites that

[09:11:40.0000] <matjas>
FINALLY

[09:12:20.0000] <matjas>
I don’t know why I keep trying to read specs… not even all this Higgs boson talk makes me feel this stupid :)

[09:14:11.0000] <matjas>
/s/false/true/ a couple of messages back

[09:15:40.0000] <Ms2ger>
ES is horrible, I can only do this thanks to a smart guy in another channel :)

[09:16:54.0000] <jgraham>
ES is just quite complicated

[09:17:10.0000] <matjas>
the spec is a bit weirder than other specs, too

[09:17:18.0000] <jgraham>
I'm not sure that any other programming language spec would be simpler

[09:17:22.0000] <matjas>
e.g. “If experValue is null or undefined, return (normal, empty, empty).” — what the hell is (normal, empty, empty)?

[09:17:32.0000] <jgraham>
That's defined elsewhere

[09:18:09.0000] <matjas>
undoubtedly. but in WHATWG/W3C specs you’d get a clickable link to the explanation/definition

[09:18:35.0000] <matjas>
stuff like that helps make things a bit more n00b-friendly

[09:18:45.0000] <jgraham>
http://es5.github.com/#x8.9

[09:18:47.0000] <matjas>
although i can see how that is not the goal of any spec

[09:19:19.0000] <jgraham>
One major problem with ECMAScript is that it doesn't use HTML natively so obvious features like hypoerlinks of definitions are missing

[09:20:12.0000] <jgraham>
e.g. if that was phrased If experValue is null or undefined, return the [completion type] (normal, empty, empty).", with the bit in square brackets being a link it would be a lot easier to work with

[09:20:36.0000] <matjas>
jgraham: exactly

[09:20:45.0000] <matjas>
/me ♥ Anolis

[09:22:30.0000] <Ms2ger>
:)

[09:25:12.0000] <gsnedders>
matjas: We love you too.

[10:03:48.0000] <matjas>
Ms2ger: ask jorendorff to accept https://github.com/jorendorff/es-spec-html/pull/8 please

[10:03:55.0000] <matjas>
/me wants to delete fork

[10:06:27.0000] <Ms2ger>
Breaking news: CSSWG is renaming flexbox properties again

[10:10:43.0000] <Ms2ger>
... I guess that wasn't a surprise to anybody

[10:21:18.0000] <shepazu>
krijnh: hey, when did your IRC logs change? they look like they've lost flagging functionality, or is that just me?

[10:22:28.0000] <Ms2ger>
Just you

[10:23:09.0000] <zewt>
been missing for me for a while too

[10:24:45.0000] <shepazu>
it's weird… and all of the comments by Ms2ger are now rendered in white text on a black background, with little unicode skulls at the end of each line

[10:25:50.0000] <Ms2ger>
Oh really ? ☠

[10:26:20.0000] <shepazu>
took you a while to find that skull, Ms2ger, I'm disappointed :)

[10:26:44.0000] <Ms2ger>
/me patpats shepazu 

[10:27:33.0000] <shepazu>
srsly, do you have to be one of the in-crowd, or signed in, or has the functionality just be removed or broken?

[10:28:32.0000] <gsnedders>
WFM, and it doesn't have my nick saved or anything

[10:30:00.0000] <zewt>
definitely doesn't work for me

[10:30:08.0000] <gsnedders>
Weird.

[10:30:24.0000] <zewt>
in ff13 or chrome 19

[10:30:58.0000] <shepazu>
gsnedders: I'm on FF13

[10:31:54.0000] <shepazu>
gsnedders: you can flag something, and it persists when you reload? it's not just the highlighted line behavior?

[10:32:10.0000] <zewt>
it doesn't even give me the UI to do it

[10:32:23.0000] <gsnedders>
shepazu: I didn\t try reloading.

[10:32:30.0000] <gsnedders>
Why would I do something so radical?

[10:33:32.0000] <shepazu>
gsnedders: well, you are an iconoclast, I thought it would be second nature to doubt

[10:33:55.0000] <gsnedders>
Right, persisted here, at least.

[10:33:58.0000] <shepazu>
zewt: same here

[10:34:09.0000] <gsnedders>
You should be using Opera, obviously. :)

[10:34:22.0000] <shepazu>
I guess sgsnedders is one of the in-crowd :)

[10:34:39.0000] <shepazu>
zewt: … well, I'm sorry...

[10:34:44.0000] <shepazu>
you didn't make the cut

[10:35:02.0000] <gsnedders>
shepazu: I guess not, given krijnh has no way to tell me from anyone else. :)

[10:38:08.0000] <zewt>
looks like something broken on the php side, since it's just not outputting the span that the script is looking for

[10:54:43.0000] <odinho>
I thought we had to use usercss and stuff.

[12:01:46.0000] <jgraham>
shepazu: You aren't trying to flag something from not today, are you?

[12:02:11.0000] <jgraham>
krijnh changed it so that flagging gets disabled after the day is no longer today

[12:02:23.0000] <shepazu>
ahhhh

[12:02:26.0000] <shepazu>
why's that?

[12:02:48.0000] <Philip`>
For caching reasons, I thought

[12:03:14.0000] <Philip`>
(making all archived logs be static HTML instead of running slow PHP each time)

[12:03:28.0000] <Philip`>
though I may just be imagining remembering that

[12:03:37.0000] <jgraham>
Oh, I thought it was just so that people living in silicon valley would have to try double hard to get things in yellow

[12:03:53.0000] <jgraham>
Since the cutoff is halfway through the day there

[12:03:55.0000] <shepazu>
oh, that's too bad, usually flagging doesn't take place the same day, I'd have thought

[12:04:15.0000] <jgraham>
If it was me I would have let things be flagged for a few days

[12:04:35.0000] <jgraham>
But it's not, and I  don't have to pay the bandwidth so...

[12:06:14.0000] <shepazu>
I'm looking at improving the W3C IRC bot stuff

[12:06:40.0000] <shepazu>
including something like krijn's logs

[12:19:41.0000] <Philip`>
Is there really no decent off-the-shelf IRC logging software that could be used?

[12:25:40.0000] <jgraham>
It does sound like a solved problem, doesn't it

[13:36:49.0000] <annevk>
jgraham: it's not bandwidth, it's performance

[13:37:09.0000] <annevk>
jgraham: but maybe krijnh is willing to tweak it to make it possible to change it for the last three days or so

[13:37:15.0000] <annevk>
jgraham: dunno how flexible the setup is

[14:02:08.0000] <gsnedders>
Fairly agressive caching could help without making it immutable.

[14:17:54.0000] <annevk>
webkit-dev is amusing

[14:18:05.0000] <annevk>
Intel guy #1: Look, user timing API implementation!

[14:18:14.0000] <annevk>
Intel guy #2: Nice feature!

[14:18:19.0000] <annevk>
</thread>

[14:31:10.0000] <guzzlefry>
Is it preferable to set image width and height in CSS or the actual HTML?

[14:31:39.0000] <annevk>
guzzlefry: if you set them in CSS you might as well not set them

[14:31:56.0000] <guzzlefry>
Why is that?

[14:32:13.0000] <annevk>
guzzlefry: because by the time the style sheet is loaded the images are prolly too

[14:32:26.0000] <guzzlefry>
ah okay, thanks

[14:38:21.0000] <jgraham>
annevk: Could make the highlights a seperate JSON file and applied in client script so that you always serve static files and the only dynamic bits are in response to POSTs that update the highlighting JSON

[14:38:49.0000] <annevk>
I'm sure krijnh has plenty of free time :)

[14:39:21.0000] <annevk>
Which reminds me, how is updating pimpmyspec.net going?

[14:39:29.0000] <jgraham>
Sure, I realise it is work

[14:40:49.0000] <jgraham>
Not at all. The main missing feature is the xrefs which I don't really know how to implement

[14:44:28.0000] <zewt>
annevk: there's no way there's a performance problem with toggling highlights in an IRC log, heh

[14:44:45.0000] <zewt>
unless there's some catastrophically bad code involved

[14:46:42.0000] <annevk>
it's not about the highlighting itself

[14:46:55.0000] <annevk>
it's about generating the pages

[14:47:49.0000] <zewt>
it's not like there are thousands of highlights being toggled a second, heh


2012-07-05
[18:35:55.0000] <Hixie>
can anyone try and see if this weasyprint thing can handle the html spec?

[18:48:52.0000] <guzzlefry>
Hixie: It says it passes Acid2?

[18:49:09.0000] <Hixie>
yes?

[18:53:32.0000] <guzzlefry>
Hixie: Going to assume I need more knowledge about HTML than I currently have to help with this. :P

[18:53:40.0000] <Hixie>
why?

[18:54:01.0000] <Hixie>
i just mean can someone install weasyprint and try to print the html spec and see whether it does it or not :-)

[18:54:10.0000] <Hixie>
as opposed to crashing or taking four hours

[18:54:19.0000] <guzzlefry>
ah, I might be able to do that.

[18:55:08.0000] <Hixie>
we used to use prince to print the PDF version but it required so much RAM to create the PDF that my server would kill the process before it could finish

[18:56:15.0000] <guzzlefry>
Is the spec on whatwg.org or w3c?

[18:56:30.0000] <Hixie>
http://whatwg.org/c

[20:09:44.0000] <guzzlefry>
Well, weasyprint doesn't like Gentoo. :P

[21:26:22.0000] <abarth>
MikeSmith: you around?

[21:26:35.0000] <MikeSmith>
abarth: yup

[21:26:37.0000] <MikeSmith>
what's up?

[21:26:54.0000] <abarth>
I was wondering if we could give Mike West commit access to http://dvcs.w3.org/hg/content-security-policy/

[21:27:03.0000] <MikeSmith>
sure

[21:27:08.0000] <MikeSmith>
he has a W3C account already?

[21:27:08.0000] <abarth>
thanks

[21:27:28.0000] <abarth>
he says he had an old individual account from a while back

[21:27:36.0000] <abarth>
but he just requested a member account a few minutes ago

[21:27:41.0000] <abarth>
(he works for Google now)

[21:27:43.0000] <MikeSmith>
yeah found it

[21:29:13.0000] <MikeSmith>
OK he should have access now

[21:29:17.0000] <abarth>
many thanks

[21:29:21.0000] <abarth>
you're the best :)

[21:29:26.0000] <MikeSmith>
hah

[21:29:29.0000] <MikeSmith>
cheers man

[22:24:09.0000] <MikeSmith>
Hixie: I installed weasyprint but it fails when I try it with the HTML spec

[22:24:40.0000] <MikeSmith>
never gets past validating the stylesheet

[22:25:21.0000] <MikeSmith>
it appears to be first trying to validate the stylesheet before processing the document

[22:26:10.0000] <Hixie>
k

[22:26:11.0000] <Hixie>
oh well

[22:27:06.0000] <MikeSmith>
I did get it to work on some smaller docs

[22:27:52.0000] <MikeSmith>
it doesn't seem to use much memory but it's extremely slow and hogs up CPU

[22:29:39.0000] <MikeSmith>
Hixie: so yeah anyway I don't think it would work even if somebody took time to figure out how to workaround the thing that's causing it to fail with the HTML spec now

[01:51:50.0000] <zcorpan>
matjas: why doesn't http://mathias.html5.org/tests/javascript/string/ test fontsize?

[01:52:12.0000] <matjas>
zcorpan: because whoops. good catch

[01:54:19.0000] <zcorpan>
matjas: a wrapper function in tests is usually unnecessary

[01:54:44.0000] <matjas>
zcorpan: no test() calls? just the assert*()s?

[01:55:08.0000] <zcorpan>
no, i mean the outer function

[01:55:19.0000] <zcorpan>
(function() { .... })();

[01:55:51.0000] <odinho>
You can make them globals, no prob ;]

[01:55:57.0000] <matjas>
ah, yeah that’s just me. don’t want to leak the typeError var

[01:56:18.0000] <matjas>
must not pollute global scope… not even in tests </ocd>

[01:56:19.0000] <odinho>
I did that at first as well. Then I learned to relax and stop worrying.

[01:56:31.0000] <jgraham>
That's just weird :)

[01:56:50.0000] <zcorpan>
best practice for tests is not exactly the same as best practice in general :-)

[01:56:57.0000] <matjas>
jgraham: you don’t know me very well, do you

[01:57:01.0000] <zcorpan>
tests should usually cut the crap

[01:57:01.0000] <odinho>
Figured out, less code -> better. Nice looking tests is best.

[02:05:26.0000] <zcorpan>
matjas: what's the moz bug number for " in link()?

[02:05:36.0000] <matjas>
https://bugzilla.mozilla.org/show_bug.cgi?id=352437

[02:06:36.0000] <zcorpan>
thanks

[02:57:23.0000] <MikeSmith>
anybody know where's the place to report bugs in the mercurial sources?

[02:58:29.0000] <MikeSmith>
nm

[02:58:32.0000] <jgraham>
http://bz.selenic.com/

[02:58:33.0000] <MikeSmith>
found it

[02:59:10.0000] <smaug____>
oh, they have switched to a saner bug tracker

[03:06:34.0000] <zcorpan>
annevk: i think backslash to forward slash in URLs can be restricted to file, http and https

[03:08:28.0000] <zcorpan>
annevk: and should only happen before the first ? or #

[03:09:08.0000] <zcorpan>
annevk: (except # for file seems to not be treated as a fragment delimiter in chrome)

[03:09:33.0000] <MikeSmith>
Content-Type: text/html; charset="UTF-8" is not valid, right?

[03:09:37.0000] <MikeSmith>
(note the quotes)

[03:09:55.0000] <MikeSmith>
a media-type should not have quotes in there, right?

[03:10:04.0000] <MikeSmith>
or are they optional or something?

[03:10:37.0000] <zcorpan>
quotes are allowed iirc

[03:12:27.0000] <jgraham>
annevk: re: notifications, it does seem to make somse sense to me to add language metadata even if all notification mechanisms don't support it. But it doesn't seem like 1 language / field is enough in many cases, and that can't be solved without markup (which is a really bad idea)

[03:13:08.0000] <zcorpan>
MikeSmith: http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.7

[03:13:30.0000] <zcorpan>
3.6:       value                   = token | quoted-string

[03:13:41.0000] <MikeSmith>
yeah, I remember I had this discussion somewhere before

[03:13:57.0000] <MikeSmith>
so it does allow a quoted string

[03:15:24.0000] <zcorpan>
quoted-string  = ( <"> *(qdtext | quoted-pair ) <"> ) http://www.w3.org/Protocols/rfc2616/rfc2616-sec2.html

[03:15:33.0000] <zcorpan>
so only allows *double* quotes

[03:15:49.0000] <MikeSmith>
ah yeah I remember that now too

[03:15:50.0000] <MikeSmith>
nuts

[03:16:15.0000] <zcorpan>
xrefs in rfcs would be nice

[03:17:54.0000] <Ms2ger>
HTML in rfcs would be nice

[03:20:07.0000] <SimonSapin>
like this? http://tools.ietf.org/html/

[03:21:03.0000] <Ms2ger>
No, not like that

[03:21:37.0000] <MikeSmith>
anyway filed http://bz.selenic.com/show_bug.cgi?id=3531

[03:22:08.0000] <MikeSmith>
and I hacked the hgweb sources on dvcs.w3.org to drop the quotes around the charset value

[03:22:17.0000] <MikeSmith>
that was causing the validator to complain

[03:22:32.0000] <Ms2ger>
Fix the validator? ;)

[03:22:55.0000] <MikeSmith>
I remember having a discussion with Henri about that

[03:23:11.0000] <MikeSmith>
iirc it's because of third-party code

[03:23:19.0000] <MikeSmith>
maybe that chardet things

[03:23:52.0000] <MikeSmith>
but yeah it should be fixed, because it's wrong

[03:35:58.0000] <zcorpan>
Ms2ger: if you (or someone else) have a preference for what to do with https://bugzilla.mozilla.org/show_bug.cgi?id=352437 please say so in a comment

[03:36:55.0000] <Ms2ger>
I've got a patch to match matjas's spec, just need to clean it up a bit

[03:37:19.0000] <matjas>
FWIW, I’ve got a V8 patch ready to get rid of the unneeded escapes

[03:37:55.0000] <matjas>
let’s align some browser implementations \o/

[03:38:30.0000] <zcorpan>
Ms2ger: ok, nice

[03:50:43.0000] <Von_Davidicus>
I have a question about website coding--particularly the use of a database.

[03:51:35.0000] <Von_Davidicus>
If you had a field that required a massive amount of text, would you store it in the database, or in a seperate text file, and refer to it via database?

[03:51:50.0000] <Philip`>
What's "massive"?

[03:51:54.0000] <Philip`>
(Megabytes? Gigabytes?)

[03:52:59.0000] <Von_Davidicus>
Anywhere from 2 KB to 25 KB.  They're short stories.

[03:53:07.0000] <Philip`>
That sounds tiny :-)

[03:53:28.0000] <Von_Davidicus>
My question still stands.

[03:53:30.0000] <Philip`>
so I'd store it in a MySQL TEXT column (or equivalent in other databases)

[03:54:02.0000] <Philip`>
since storing it externally would be a lot of hassle for no real benefit

[03:54:32.0000] <Von_Davidicus>
Aside from editability, but I think I see your point.

[03:55:05.0000] <Philip`>
(Oh, I suppose not actually TEXT, since that's stupidly small; might as well use LONGTEXT)

[03:56:37.0000] <Von_Davidicus>
The reason I had them in seperate text files was so that I could edit them in a text editor.

[03:58:05.0000] <Philip`>
I imagine it'd be useful to do things like store historical revisions of every story, and store metadata like who last modified the data and when, so you'd always be required to edit them through the proper database-based interface instead of using a text editor directly

[03:58:34.0000] <Von_Davidicus>
Okay...

[03:59:02.0000] <Philip`>
unless you're absolutely certain your design will never need any data other than what the filesystem already provides

[03:59:27.0000] <zcorpan>
matjas: have you filed a bug with IE about this?

[03:59:42.0000] <matjas>
zcorpan: not yet, only about the fact that IE uppercases the tag names

[03:59:44.0000] <jgraham>
Philip`: Obviosuly you should put git into the mix to get revisions

[04:00:02.0000] <Von_Davidicus>
Meaning?

[04:00:33.0000] <matjas>
zcorpan: https://connect.microsoft.com/IE/feedback/details/752283 if you’re interested. I’ll file another one for the " escape thing

[04:00:47.0000] <Philip`>
Von_Davidicus: Knowing jgraham, that was probably sarcasm

[04:00:52.0000] <zcorpan>
matjas: thanks. i'm interested in the " thing :-)

[04:01:10.0000] <Von_Davidicus>
/me was directing the "meaning" to you, Philip`

[04:01:15.0000] <Von_Davidicus>
With the text files, all I do is use PHP's include() function.

[04:01:26.0000] <Philip`>
Ah :-)

[04:01:32.0000] <Von_Davidicus>
Heck, I could almost get away with using SSI.

[04:01:57.0000] <Philip`>
Sure, you could choose to use horrid technologies if you really wanted to :-p

[04:02:41.0000] <Von_Davidicus>
SSI's not bad.  I've used it on a page with a several -massive- tables.

[04:03:09.0000] <Von_Davidicus>
As I said, I could -almost- get away with it; I get the title and story rating info from the database.

[04:07:40.0000] <Von_Davidicus>
Besides, you want horrid?  Try using XML files as your database.

[04:08:02.0000] <Von_Davidicus>
BOY, that gets complicated.

[04:09:36.0000] <Philip`>
If you store the story text in the database too then you get simpler code, more flexibility, more nice ACID properties, easier backups, better performance, etc, and if you want to edit a story in a text editor then just write a script that downloads it from the database into a text file and then you edit it and then upload it back again

[04:10:01.0000] <zcorpan>
matjas: maybe you could put in a recommendation for authors saying something along the lines of "use of these methods is not recommended; using DOM APIs is recommended. however, if these methods are used nevertheless, the characters & and " should be escaped as &amp; and &quot;, respectively"

[04:10:06.0000] <Philip`>
so I don't see why you'd make life harder for yourself :-)

[04:10:06.0000] <Von_Davidicus>
ACID properties?

[04:10:32.0000] <Philip`>
Like in http://en.wikipedia.org/wiki/ACID

[04:10:43.0000] <matjas>
zcorpan: filed https://connect.microsoft.com/IE/feedback/details/752391

[04:11:22.0000] <matjas>
zcorpan: i have a note here: http://mathias.html5.org/specs/javascript/#escapeattributevalue but i guess it would be clearer to have one near the actual string method definitions

[04:11:41.0000] <Philip`>
(e.g. the database system will prevent the data and metadata getting out of sync with each other, but you get no such guarantees if you involve the filesystem unless you're extremely careful)

[04:12:38.0000] <Von_Davidicus>
Hmmm... never had a problem with it before.

[04:13:01.0000] <zcorpan>
matjas: hmm, i don't have permission to view it

[04:13:31.0000] <matjas>
zcorpan: http://i.imgur.com/G73ww.png

[04:13:43.0000] <Philip`>
Von_Davidicus: Rare bugs are the worst :-)

[04:14:22.0000] <Von_Davidicus>
I suppose I could throw this into the database... and have PHP insert the tab characters and line breaks in the code where they should be.

[04:14:40.0000] <Philip`>
Von_Davidicus: Everything will look like it's running fine for several years as you get more and more users, and then the server's PSU blows up and you're very unlucky and you find that half of your data is corrupted and irrecoverable

[04:15:53.0000] <Von_Davidicus>
That, sir, is why I have a backup site on my own computer.  And I back --that-- up, too.

[04:17:00.0000] <Philip`>
If your backup is a snapshot of the server while it's in a temporarily inconsistent state (e.g. because it's updated its metadata but not yet written the actual data) then that still won't solve the problem

[04:18:13.0000] <Von_Davidicus>
You have it backwards, sir.  What's on the -server- is a snapshot of what's on my computer.  I edit my site on my computer, then I upload whatever files I need to and run whatever SQL I have to.

[04:23:51.0000] <zcorpan>
annevk: hmm, the chrome file # thing seems is just for file://#/ or file:\\#\ - hash anywhere else is fragment. but that seems bogus and opera/firefox don't do that

[04:26:02.0000] <zcorpan>
http://# resolves to http:/# in chrome and http:///# in opera/firefox

[04:34:40.0000] <matjas>
zcorpan: “using DOM APIs is recommended” — should I add a reference there? to [DOM2HTML], of something else?

[04:34:46.0000] <matjas>
s/of/or/

[04:36:32.0000] <zcorpan>
matjas: dom2html is obsolete, so preferably not :-)

[04:36:55.0000] <zcorpan>
i guess dom4 and html if you want to reference something

[04:37:17.0000] <matjas>
i couldn’t find property attributes in DOM4 right away so i got confused

[04:41:08.0000] <zcorpan>
matjas: i get an empty doc now

[04:42:46.0000] <matjas>
zcorpan: fixed

[04:46:06.0000] <zcorpan>
matjas: thanks. i guess the dom api recommendation applies to all of those, but it seems annoying to repeat it over and over. maybe put the note at the top of the section, and say that the escaping recommendation only applies to those that take an argument, or so?

[04:46:22.0000] <matjas>
zcorpan: but there’s also substr

[04:46:25.0000] <AryehGregor>
matjas, I suggest explicitly saying in the note that browsers have historically not escaped " in the argument, so anyone using the function is likely to be bitten by XSS unless they escape.

[04:48:16.0000] <annevk>
zcorpan: not ftp?

[04:48:39.0000] <zcorpan>
matjas: i'd solve that by removing substr from that section and say in a separate section that annex b must be supported :-)

[04:48:44.0000] <annevk>
zcorpan: I figured if we have a whitelist of hierarchical schemes we can do all that magic there

[04:48:50.0000] <annevk>
zcorpan: including path normalization and such

[04:48:52.0000] <zcorpan>
annevk: no occurrences of backslashes in ftp in web200904

[04:49:26.0000] <zcorpan>
annevk: see CT-3084

[04:50:04.0000] <zcorpan>
annevk: isn't it good to keep the backslash rewriting to a minimum?

[04:50:37.0000] <annevk>
zcorpan: I rather keep the special scheme lists to a minimum

[04:50:55.0000] <zcorpan>
annevk: why?

[04:51:04.0000] <annevk>
less code paths

[04:51:13.0000] <annevk>
one code path for hierarchical schemes, one for the rest

[04:51:46.0000] <annevk>
not sure what the point is in not admitting defeat on \

[04:52:19.0000] <zcorpan>
well, i dunno, new schemes get to use \ as \

[04:53:24.0000] <zcorpan>
but i don't care about it strongly

[04:53:45.0000] <zcorpan>
it can be escaped

[04:54:14.0000] <annevk>
right, and it seems better if new hierarchical schemes behave consistently with existing hierarchical schemes

[04:57:08.0000] <annevk>
jgraham: so apparently there's a way to embed language metadata through Unicode

[04:57:26.0000] <annevk>
jgraham: but it's "deprecated" or some such

[04:57:30.0000] <jgraham>
:-o

[04:57:37.0000] <jgraham>
Really?

[04:57:39.0000] <jgraham>
Wow

[04:58:25.0000] <annevk>
http://en.wikipedia.org/wiki/Unicode_control_characters#Language_tags

[05:01:18.0000] <zcorpan>
matjas: assert: rfc2119 words in a note :-P

[05:01:41.0000] <matjas>
how do you check for that? is that some anolis setting?

[05:02:31.0000] <zcorpan>
it isn't. but that's not a bad idea, actually

[05:05:04.0000] <annevk>
can you nest dictionaries?

[05:05:50.0000] <annevk>
dictionary UIString { DOMString value; DOMString lang; Direction dir; }

[05:05:54.0000] <annevk>
or some such

[05:06:19.0000] <zcorpan>
var a = { dir: a }

[05:12:51.0000] <annevk>
boo

[05:19:35.0000] <Lachy>
zcorpan, that doesn't work, since a is undefined at time of assignment. This works: var a = {}; a.dir = a

[05:20:54.0000] <zcorpan>
Lachy: good point

[05:38:12.0000] <matjas>
V8 will remove the unneeded escapes in its HTML String methods as soon as other browsers add the &quot; escape: http://code.google.com/p/v8/issues/detail?id=2217#c8

[05:44:33.0000] <gsnedders>
matjas: They're not pointless if the web relies on them. :)

[05:44:59.0000] <matjas>
gsnedders: no other browser does it so i doubt the web relies on them

[05:45:20.0000] <annevk>
opposite is more likely

[05:45:41.0000] <gsnedders>
It seems better to escape too much than too little, IMO

[05:47:41.0000] <zcorpan>
gsnedders: let's not muddy the waters

[05:48:27.0000] <gsnedders>
I'd rather have the methods not lead to trivial XSS holes.

[05:48:37.0000] <zcorpan>
<>' does't

[05:48:41.0000] <zcorpan>
" does

[05:48:48.0000] <zcorpan>
the spec says to escape only "

[05:49:02.0000] <zcorpan>
let's stick to that

[05:49:02.0000] <gsnedders>
Oh, matjas's new spec?

[05:49:05.0000] <zcorpan>
yes

[05:49:33.0000] <zcorpan>
we can't escape & since that would lead to double-escaping for sites that "do the right thing"

[05:49:45.0000] <zcorpan>
so escaping <>' is pointless

[05:53:27.0000] <matjas>
zcorpan: if & was escaped into &amp;, would it be useful to escape <>'?

[05:53:56.0000] <annevk>
matjas: why would you escape & or < and >?

[05:54:20.0000] <matjas>
annevk: i wouldn’t, just wondering… i can’t think of any XSS vector using <>' even if &amp; gets escaped

[05:54:38.0000] <matjas>
in quoted attribute values with double quotes, that is

[05:54:45.0000] <zcorpan>
matjas: no

[05:54:56.0000] <zcorpan>
innerHTML doesn't escape those either

[05:55:41.0000] <zcorpan>
for innerHTML it was a compat problem to escape <>

[05:57:12.0000] <zcorpan>
(opera always did it, ie never did it, webkit and firefox usually escaped it but didn't for href="javascript:<>", and of course a sitecompat bug was filed on opera where it expected it to not be escaped for href="javascript:<>")

[05:57:18.0000] <zcorpan>
(iirc)

[05:57:37.0000] <gsnedders>
zcorpan: In Carakan or Futhark?

[05:57:57.0000] <zcorpan>
gsnedders: innerHTML

[05:58:29.0000] <zcorpan>
don't recall if it was pre- or post- carakan shipping

[05:58:48.0000] <gsnedders>
zcorpan: Ah, not String.prototype.*

[06:01:33.0000] <zcorpan>
matjas: so "Note: For security reasons, Web authors should not rely on any such escaping, as older, non-JavaScript-compliant implementations may not perform this operation correctly." shouldn't use "should" or "may"

[06:01:59.0000] <matjas>
zcorpan: yeah, havent gotten around to it yet :) which verb to pick…

[06:02:10.0000] <zcorpan>
k

[06:02:56.0000] <zcorpan>
matjas: it kinda repeats the other note, doesn't it?

[06:06:06.0000] <matjas>
the one near EscapeAttributeValue? yeah

[06:06:48.0000] <matjas>
but this way, Web authors who only read a specific part of the spec, get to see the warning anyway

[06:06:51.0000] <zcorpan>
that's the one i quoted

[06:07:07.0000] <matjas>
only having the warning once would assume everyone read the full spec

[06:07:22.0000] <annevk>
matjas: instead of class=note, use class=warning

[06:07:32.0000] <annevk>
matjas: and keep it normative

[06:07:34.0000] <matjas>
annevk: thanks

[06:07:43.0000] <matjas>
you’re all so helpful, i love it

[06:08:09.0000] <zcorpan>
that works, though "may not" should still be changed

[06:09:31.0000] <matjas>
s/may not/might not/

[06:10:09.0000] <zcorpan>
yep

[06:29:25.0000] <zcorpan>
i wrote an introduction, http://simon.html5.org/specs/quirks-mode#introduction

[06:33:01.0000] <annevk>
"standards mode but expected quirks" missing comma?

[06:33:41.0000] <zcorpan>
dunno, does it need a comma?

[06:33:42.0000] <annevk>
"apart from CSS" -> "outside of CSS"

[06:33:51.0000] <annevk>
I thought but would typically be preceded by a comma

[06:35:19.0000] <gsnedders>
annevk: Matter of style.

[06:36:39.0000] <zcorpan>
annevk: added a comma but a bit earlier. also changed the other thing. thanks

[06:37:06.0000] <matjas>
s/further/since/ but I suppose that’s style too

[06:38:53.0000] <zcorpan>
fixed

[06:39:20.0000] <matjas>
zcorpan: what’s your workflow for writing specs and hosting them on simon.html5.org? did you manage to get Python + Anolis running on the server? I just run everything locally and rsync

[06:39:49.0000] <zcorpan>
i run it locally

[06:40:54.0000] <zcorpan>
my plan is to move it to dvcs.w3.org

[07:32:15.0000] <smaug____>
jgraham: ping

[07:33:37.0000] <smaug____>
or any Opera-er

[07:33:40.0000] <AryehGregor>
matjas, I edit things like that on my server.  Works for me because I use vim anyway.

[07:34:23.0000] <matjas>
AryehGregor: cool. I was asking ’cause I had no rights to install Python / Anolis on the html5.org server

[07:35:06.0000] <zcorpan>
smaug____: pong

[07:35:16.0000] <matjas>
I use https://github.com/cowboy/grunt to “watch” the source files, so whenever I save, Anolis generates a new spec version

[07:35:31.0000] <smaug____>
zcorpan: do you happen to know about Presto's JS<->C++ handling

[07:35:49.0000] <smaug____>
especially, is it difficult to handle cases when C++ keeps JS alive

[07:35:56.0000] <smaug____>
like event listeners

[07:36:19.0000] <zcorpan>
smaug____: i don't really, sorry

[07:36:37.0000] <smaug____>
/me is amazed that it seems to be so hard in WebKit that webkit-guys want to change an API just because of that

[07:37:02.0000] <smaug____>
zcorpan: who might know?

[07:37:30.0000] <zcorpan>
gsnedders or jgraham maybe

[07:37:31.0000] <odinho>
I don't understand the question. But I've only scratched Opera's code. Done some very small DOM stuff.

[07:38:14.0000] <zcorpan>
else we can ask some relevant developer

[07:38:23.0000] <smaug____>
odinho: when you do .addEventListener("foo", somefunction), native code needs to keep that JS function object alive

[07:38:48.0000] <smaug____>
odinho: based on comments from abarth it is hard to not leak in that case in webkit

[07:39:01.0000] <smaug____>
eventlisteners are apparently very much special cased there

[07:39:10.0000] <jgraham>
smaug____: I will ask someone who knows

[07:39:30.0000] <jgraham>
But I think their proposed replacement looks horrible

[07:39:49.0000] <odinho>
Where is this, I want to see it :D

[07:39:53.0000] <smaug____>
yeah, the new DOMTranction() thing is very odd

[07:39:59.0000] <smaug____>
odinho: webapps wg

[07:40:00.0000] <odinho>
Oh, that.

[07:40:02.0000] <zcorpan>
i know we have special cases for event listeners to make sure we don't gc the object if it can still get events for those listeners (like an open websocket or xhr)

[07:40:04.0000] <smaug____>
about UndoManager

[07:40:16.0000] <gsnedders>
smaug____: The only real hard case for us is stuff like XHR event handlers when the XHR object itself is GC'd, I believe.

[07:40:23.0000] <odinho>
Yea, I saw undomanager and was like "oh, wait till the end of the day"

[07:40:41.0000] <gsnedders>
(Which I imagine is fairly inelegant for everyone)

[07:40:59.0000] <smaug____>
In Gecko XHR object can't be collected if the connection is still open

[07:41:14.0000] <odinho>
smaug____: Yea, -- someone said here that we have no such problem.

[07:41:34.0000] <jgraham>
odinho: Who?

[07:41:41.0000] <odinho>
jgraham: joaoe :-)

[07:41:54.0000] <smaug____>
odinho: was that about XHR, or JS callbacks in general?

[07:42:06.0000] <odinho>
smaug____: js callbacks in general.

[07:42:12.0000] <smaug____>
ok, good

[07:43:54.0000] <gsnedders>
/me is somewhat amused by how he avoids DOM code, despite dealing with both JS and CSS at times

[07:44:33.0000] <zcorpan>
and i try to avoid JS :-P

[07:49:48.0000] <AryehGregor>
We apparently have this magical thing called a cycle collector that's built out of macros and pixie dust, which mysteriously makes things not leak if you enlist one of the two people who know how it works to tell you the right incantation to add to your classes.

[07:53:42.0000] <MikeSmith>
hah

[07:54:17.0000] <MikeSmith>
I read that thread

[07:55:21.0000] <AryehGregor>
/me discovers that Element.style is readonly . . . why?

[07:55:36.0000] <AryehGregor>
Wouldn't you expect setting it to work like setAttribute(), same as practically all such attributes?

[07:57:59.0000] <hober>
you would

[07:58:00.0000] <zcorpan>
you do style.foo = 'bar' instead

[07:58:13.0000] <hober>
but el.style is not a string corresponding to <el style="">

[07:59:24.0000] <AryehGregor>
hober, sure, but it could still be settable.  .classList is settable, right?

[08:00:03.0000] <zcorpan>
nope

[08:10:00.0000] <AryehGregor>
Hmm, so it's not.

[08:10:02.0000] <AryehGregor>
Oh well.

[08:10:04.0000] <AryehGregor>
It confused me.

[08:15:38.0000] <zcorpan>
MikeSmith: how do i create a new repo on dvcs.w3.org?

[08:15:55.0000] <Ms2ger>
zcorpan, ask MikeSmith :)

[08:16:10.0000] <MikeSmith>
yah

[08:16:15.0000] <MikeSmith>
just lemme know the name

[08:16:41.0000] <zcorpan>
ok, how about 'quirks-mode'

[08:18:00.0000] <MikeSmith>
OK

[08:18:23.0000] <MikeSmith>
zcorpan: who should have write access?

[08:19:06.0000] <zcorpan>
um. me, to begin with, i guess we can add more later if needed

[08:21:28.0000] <zcorpan>
annevk: why http://dvcs.w3.org/hg/encoding/rev/55a9e6c1aa32 ?

[08:22:32.0000] <annevk>
zcorpan: lawyers?

[08:22:53.0000] <annevk>
zcorpan: tantek reverted that for Fullscreen

[08:23:04.0000] <annevk>
zcorpan: so maybe it's okay again

[08:23:34.0000] <annevk>
MikeSmith: spec-writers group?

[08:23:57.0000] <MikeSmith>
annevk: for quirks-mode draft?

[08:24:03.0000] <annevk>
zcorpan: in particular it's something chaals combined with W3C cared about and nobody else

[08:24:05.0000] <AryehGregor>
annevk, I was asked to get rid of it but politely said no, and no one sicced any lawyers on me yet.

[08:25:17.0000] <annevk>
AryehGregor: took them about six months to get back to me and then I lost some silly internal argument and gave up for now

[08:29:55.0000] <zcorpan>
annevk: ok :-/

[08:30:19.0000] <annevk>
should not be a problem now though

[08:34:23.0000] <matjas>
where does the web ecmascript identifier thing originate?

[08:34:30.0000] <matjas>
trying to understand if it can be dropped or not

[08:36:01.0000] <annevk>
if IE/Gecko dropped it chances are the others want to follow

[08:36:12.0000] <zcorpan>
matjas: probably at the time we implemented carakan, jgraham noticed that everyone else did it so assumed it had to be supported

[08:37:15.0000] <matjas>
yay \o/

[08:38:09.0000] <annevk>
matjas: filing bugs on the other engines with "remove this because IE/Gecko don't" is usually sufficient

[08:39:56.0000] <jgraham>
matjas: Right, what zcorpan said. I was generally pretty liberal with the definition of "everyone" since it is better to have too much compat than too little

[09:00:53.0000] <MikeSmith>
zcorpan: https://dvcs.w3.org/hg/quirks-mode created

[09:01:40.0000] <zcorpan>
MikeSmith: thanks!

[09:06:45.0000] <annevk>
zcorpan: I guess the best you can do is to just keep it under the current copyright and publish it through WHATCG once the W3C resolves some of the outstanding questions

[09:07:07.0000] <zcorpan>
http://dvcs.w3.org/hg/quirks-mode/raw-file/tip/Overview.html

[09:07:30.0000] <annevk>
or that :)

[09:07:38.0000] <annevk>
do you want me to make it official?

[09:07:45.0000] <zcorpan>
uh, the copyright shouldn't say fullscreen....

[09:08:13.0000] <annevk>
yeah

[09:09:18.0000] <annevk>
zcorpan: the only other thing that's semi-required is an email like this one: http://lists.w3.org/Archives/Public/public-whatwg-contrib/2012Apr/0008.html

[09:11:32.0000] <zcorpan>
hmm does MakeFile need a special chmod or something again?

[09:11:53.0000] <matjas>
chmod +x Makefile

[09:12:25.0000] <zcorpan>
still, "make: Nothing to be done for `all'."

[09:12:42.0000] <annevk>
zcorpan: "touch Overview.src.html"

[09:12:44.0000] <annevk>
in doubt

[09:13:06.0000] <zcorpan>
oh wait i edited the wrong file

[09:13:07.0000] <zcorpan>
heh

[09:17:19.0000] <zcorpan>
annevk: i don't need to subscribe first, right?

[09:17:47.0000] <zcorpan>
shit, need to run

[09:25:44.0000] <matjas>
what’s the advantage of having a spec be hosted on w3.org?

[09:28:43.0000] <smaug____>
matjas: comparing to what?

[09:28:59.0000] <matjas>
smaug____: compared to anywhere else, e.g. on your own site

[09:29:10.0000] <smaug____>
well, some credibility

[09:29:23.0000] <Lachy>
matjas, depends what you mean. The patent policy can help if you develop the spec in a WG, task force, community group or whatever.

[09:30:11.0000] <Lachy>
but there are specs that have been implemented despite being written and hosted on personal sites, at least initially.

[09:31:41.0000] <odinho>
And it kinda helps finding it...

[09:37:46.0000] <matjas>
I see. Use of the bug tracker is nice too, I imagine

[10:45:20.0000] <annevk>
zcorpan: hmm dunno

[10:45:34.0000] <annevk>
zcorpan: if you join http://www.w3.org/community/whatwg/ you are subscribed

[10:46:04.0000] <annevk>
zcorpan: I'll just publish it and you can sort that out later

[10:46:45.0000] <annevk>
zcorpan: oh also, ij prefers if instead of "Standard" it says "Specification"

[11:25:33.0000] <hober>
Lachy: are you still editing the about: spec?

[11:26:14.0000] <Lachy>
yes, sort of.

[11:26:19.0000] <Lachy>
though I haven't looked at it in a while.

[11:26:31.0000] <hober>
where should people send feedback? to the whatwg lists with [about] in the subject?

[11:26:50.0000] <smaug____>
what is about: spec about?

[11:26:54.0000] <smaug____>
about:blank ?

[11:26:55.0000] <Lachy>
probably better to send it to the IETF list.

[11:27:04.0000] <Lachy>
smaug____, yes, and otehrs

[11:27:07.0000] <Lachy>
*others

[11:27:15.0000] <smaug____>
what all others

[11:27:33.0000] <hober>
do you mean uri⊙wo?

[11:28:08.0000] <smaug____>
oh, it is about the protocol, not about the about:* pages

[11:29:10.0000] <Lachy>
hober, ietf⊙io

[11:29:23.0000] <hober>
ahh, uri-review⊙io the number of url mailing lists is too damn high.

[11:29:28.0000] <Lachy>
oh, yes

[11:29:29.0000] <Lachy>
that's the one

[11:29:50.0000] <Lachy>
discussion has shifted so many times for that draft, even I lose track of where it should be

[11:30:09.0000] <Lachy>
at one stage, they even started having discussions on a list that I wasn't subscribed to without telling me

[11:30:16.0000] <hober>
oh awesome

[11:30:19.0000] <Lachy>
that was fun catching up on everything I missed

[11:31:24.0000] <Lachy>
smaug____, http://www.ietf.org/id/draft-ietf-appsawg-about-uri-scheme-07.txt

[11:31:30.0000] <Lachy>
that's the latest draft

[11:35:38.0000] <Lachy>
WTF? someone has completely re-written that and taken over

[11:35:44.0000] <Lachy>
I'm not even listed as editor any more

[11:36:02.0000] <smaug____>
"nice"

[11:37:02.0000] <Lachy>
oh well, I guess I'm not the editor of that any more.

[11:42:34.0000] <hober>
Lachy: weird

[11:42:43.0000] <hober>
there's an about registry now? ugh.

[11:43:00.0000] <Lachy>
yeah, I know. They pushed that through in the previous draft, despite my objections.

[11:43:06.0000] <Lachy>
After that, I just sort of gave up

[11:43:57.0000] <hober>
it doesn't even say where comment should go anymore

[11:44:46.0000] <Lachy>
I think the latest location for discussion is here now. apps-discuss⊙io

[11:45:00.0000] <Lachy>
I'm pretty sure that's the group that took over the work

[11:53:59.0000] <annevk>
hober: Lachy: not sure why you still bother with discussing it there

[11:54:07.0000] <annevk>
hober: Lachy: did anything useful come out of it?

[11:54:24.0000] <annevk>
hober: Lachy: last I heard it was mostly bad (e.g. setting up another IANA registry...)

[11:54:34.0000] <Lachy>
annevk, I haven't done anything on it for about a year

[11:55:03.0000] <Lachy>
If there's enough interest, I could republish the draft elsewhere and just completely ignore the IETF

[11:55:55.0000] <annevk>
http://tools.ietf.org/html/draft-ietf-appsawg-about-uri-scheme is a better link

[11:56:55.0000] <annevk>
maybe we should fold it into the URL spec

[11:57:18.0000] <annevk>
it should really not be more than a couple of paragraphs anyway

[12:23:35.0000] <Lachy>
annevk, fine with me. If you like, I can write up a draft, dropping a whole bunch of unnecessary and overly complex stuff, and make that easier to fold in to the URL spec.

[12:24:01.0000] <Lachy>
where is the current URL spec though?

[12:41:01.0000] <pablof>
http://dvcs.w3.org/hg/url/raw-file/tip/Overview.html ?

[12:44:48.0000] <jgraham>
timeless: Thank you

[12:54:41.0000] <jgraham>
pablof: Was there some context for that "question"?

[12:55:24.0000] <pablof>
jgraham: "Lachy: where is the current URL spec though?"

[12:55:50.0000] <pablof>
the question mark was meant to imply that i'm somewhat, but not totally sure :-)

[12:56:00.0000] <jgraham>
I see

[12:56:26.0000] <Lachy>
thanks

[12:56:27.0000] <jgraham>
I might have got that if the line had started "Lachy: "

[12:56:46.0000] <pablof>
yeah, i should've done that

[12:56:55.0000] <jgraham>
Or if I read the scrollback more closely ;)

[13:01:54.0000] <jgraham>
Argh gmail rage

[13:03:21.0000] <jgraham>
It makes me so sad that they can't get basic formatting of email replies correct

[13:03:31.0000] <jgraham>
How hard can it possibly be?

[13:07:24.0000] <hober>
jgraham: srsly

[13:07:42.0000] <hober>
fortunately, you can point gnus at gmail's imap. :)

[13:32:49.0000] <Yuhong>
Hey, I have an idea: Why not use the versioned HTML specs produced by the HTML WG as *authoring* requirements only?

[13:33:26.0000] <Yuhong>
That validators can validate.

[13:34:46.0000] <timeless>
jgraham: what did i do?

[13:34:59.0000] <timeless>
oh

[13:35:05.0000] <timeless>
um, i don't think i sent it to enough groups

[13:35:14.0000] <timeless>
that was on my todo list for one group that hasn't started messing up yet

[13:35:19.0000] <timeless>
i need to send it out to a couple of others

[13:35:47.0000] <timeless>
i'm not even sure i sent it to the right offender group

[13:35:59.0000] <timeless>
iirc the primary offenders are in DAP and i sent it to coremob

[13:37:55.0000] <timeless>
i sent out a better welcome for the intents list, and they were well behaved

[13:39:38.0000] <hober>
timeless: ?

[13:46:46.0000] <timeless>
hober: i'm assuming jgraham thanked me for sending an eitquette email to some list

[13:47:08.0000] <timeless>
http://lists.w3.org/Archives/Public/public-coremob/2012Jul/0015.htmlhttp://lists.w3.org/Archives/Public/public-coremob/2012Jul/0015.html

[13:47:12.0000] <timeless>
grr

[13:47:12.0000] <timeless>
http://lists.w3.org/Archives/Public/public-coremob/2012Jul/0015.html

[13:47:20.0000] <timeless>
/me kicks chrome for being awful

[13:48:19.0000] <hober>
ahh. yes, lovely, thanks. :)

[13:48:37.0000] <timeless>
please please feel free to plagiarize that

[13:48:55.0000] <timeless>
or reference it, whichever, i'm not picky

[13:50:11.0000] <Velmont>
timeless: Cool. But do you think people understand it?

[13:50:37.0000] <timeless>
dunno, i got a thanks from an offender

[13:50:55.0000] <timeless>
but he used outlook style --- original message--- w/ top quoting as he madehis reply...

[13:51:31.0000] <othermaciej>
timeless: I have to admit, I hate the footnote style for links in w3c mailing list mail

[13:51:36.0000] <othermaciej>
timeless: where did that originate?

[13:51:48.0000] <timeless>
it predates me

[13:52:02.0000] <timeless>
it probably relates to wanting to have a clen message

[13:52:14.0000] <othermaciej>
(I hate it because my mail client, and probably most reasonably modern ones, know how to make the URL clickable, but not the little bracketed number, so it would work better to just put the URL inline)

[13:52:15.0000] <timeless>
and also avoiding stupid line wrapping breaking links

[13:52:21.0000] <timeless>
or quoting breaking links

[13:52:33.0000] <timeless>
and uoting w/ line wrapping dfefinitely breaks links

[13:52:49.0000] <timeless>
this chrome session is causing tremendous lag to my typing, so i can't see anything

[13:52:53.0000] <Velmont>
I normally write urls after a paragrah with its own line, 4 spaces to indent it.

[13:52:53.0000] <timeless>
/me blames chrome for sucking

[13:53:01.0000] <Velmont>
timeless: Try Opera ;-)

[13:53:10.0000] <timeless>
Velmont: hrm

[13:53:16.0000] <othermaciej>
yeah, I like them inline on their own line, unless there is an excessive number of links and/or they are very short

[13:53:17.0000] <Velmont>
We did it first!

[13:53:35.0000] <othermaciej>
doing the [1] [2] thing makes it much harder to follow links with context in mind, especially for a long email

[13:53:36.0000] <timeless>
Velmont: you did sucking first?

[13:54:12.0000] <timeless>
othermaciej: you can cheat by copying the bottom into a notepad and having it alongside

[13:55:03.0000] <Velmont>
timeless ;-)

[13:55:05.0000] <othermaciej>
timeless: I think you just convinced me to never do the [1] [2] thing again

[13:55:18.0000] <othermaciej>
(I do it sometimes in my emails, but no more)

[13:55:31.0000] <Hixie>
i generally do the [1] [2] thing if I don't expect people to follow the links but just want to include them for completeness

[13:55:36.0000] <Hixie>
e.g. to back up a point i'm making

[13:55:48.0000] <timeless>
i think Hixie 's explanation is better

[13:55:51.0000] <Hixie>
whereas if i want people to look at the link, i include it inline

[13:55:53.0000] <Velmont>
Hixie: Seems reasonable. Doing it for "real" footnotes.

[13:55:57.0000] <othermaciej>
I like to actually check people's references

[13:56:18.0000] <othermaciej>
if emails could be HTML then of course you could have clickable links to the footnotes, getting sort of the best of both worlds

[13:56:20.0000] <Hixie>
Velmont: yeah, sometimes they're not even URLs, they're just textual footnotes

[13:56:38.0000] <othermaciej>
by W3C email has the same modern text technology as IETF RFCs

[13:56:59.0000] <othermaciej>
s/by/but/

[13:57:13.0000] <Velmont>
othermaciej: Well helped by useless html implementation by email clients...

[13:57:18.0000] <othermaciej>
but yeah, I guess it is a good technique if you have references that you don't care to have checked

[13:57:59.0000] <othermaciej>
people send HTML email to each other all the time at Apple and no one ever even notices, but we have a mostly monolithic set of client software with known good html rendering and editing

[13:58:01.0000] <timeless>
Velmont: Sorry, we’re only supporting browsers with WebSockets enabled at the moment

[13:58:41.0000] <timeless>
othermaciej: and people don't use fancy watermarks / backgrounds? :)

[13:59:20.0000] <othermaciej>
mostly they use bold, italic, monospace fonts for code sections, and the occasional distinguished section header, as well as, of course, links

[13:59:38.0000] <othermaciej>
and including images inline when relevant

[13:59:47.0000] <timeless>
iow you have people who aren't stupid

[14:00:00.0000] <timeless>
silly company self-selecting for intelligence

[14:00:04.0000] <othermaciej>
well, the mail client doesn't have a ready affordance for setting an ugly background

[14:00:39.0000] <othermaciej>
I expect most html email with fugly backgrounds is generated by a script, not a conventional mail client

[14:00:44.0000] <timeless>
silly company trying to restrict users from the freedom to send mail that's unreadable!

[14:01:00.0000] <othermaciej>
but perhaps I am naiive

[14:01:21.0000] <timeless>
i should try to do a survey of that, but i'm not sure where a valid sample input would be

[14:01:29.0000] <timeless>
none of my mailboxes are valid

[14:01:41.0000] <timeless>
all but one of them have 99.99% bugzlila bugmail

[14:01:51.0000] <Hixie>
/me notes that he never sees the backgrounds, since his e-mail client is a black-and-white monospace text-mode client with one font size :-)

[14:02:16.0000] <timeless>
does it support black-on-black or white-on-white? :)

[14:02:26.0000] <timeless>
Velmont: anyway, opera was a total failure

[14:02:29.0000] <timeless>
maybe i should try ie9?

[14:02:30.0000] <Velmont>
timeless: It has WebSocket, but it's disabled by default :P The old protocol that was vulnerable.

[14:02:55.0000] <Velmont>
timeless: ie10 then. If Opera fails, I would be sad if IE9 worked.

[14:02:56.0000] <timeless>
is is there a version of opera that's useful?

[14:03:19.0000] <Hixie>
timeless: it essentially doesn't support colour at all, just text

[14:03:30.0000] <Velmont>
timeless: No need to be snarky, -- ofc it's useful. I use it every day.

[14:03:43.0000] <timeless>
Velmont: for my purposes

[14:03:51.0000] <timeless>
i'm ok w/ a beta or alpha as long as someone tells me how to get it

[14:03:58.0000] <timeless>
just as you suggested ie10

[14:04:07.0000] <Velmont>
timeless: I wouldn't know. Don't know what you do etc.

[14:04:15.0000] <timeless>
alpha.irccloud.com

[14:04:23.0000] <timeless>
i don't browse the web or read email or anything like that

[14:04:39.0000] <Yuhong>
Hixie: Why not use the versioned HTML specs produced by the HTML WG as *authoring* requirements only (for validators)?

[14:04:49.0000] <Hixie>
Yuhong: ?

[14:05:10.0000] <Yuhong>
I am referring to the W3C HTML WG.

[14:05:26.0000] <Hixie>
what is the problem you are trying to solve?

[14:05:27.0000] <timeless>
this is #whatwg, not irc.w3.org #html ...

[14:05:42.0000] <Hixie>
timeless: there's no topic restriction in this channel

[14:05:44.0000] <jgraham>
timeless: New websockets is coming

[14:05:58.0000] <timeless>
jgraham: nick highlight me when they're available

[14:06:18.0000] <jgraham>
Hah, like I will remember that

[14:06:22.0000] <timeless>
i'll gladly replace chrome canary, since righ now it paints random triangular wedges across my screen

[14:06:50.0000] <timeless>
ie9 didn't work either..

[14:07:00.0000] <timeless>
Velmont: just so you don't feel bad :)

[14:07:24.0000] <timeless>
/me considers replacing chromium w/ nightly for a bit

[14:07:47.0000] <Yuhong>
Browsers typically treat all content the same, but *validators* often lets you choose different HTML versions to validate against.

[14:08:07.0000] <Hixie>
Yuhong: a validator that lets you test against an old version seems kinda silly

[14:08:09.0000] <jgraham>
/me also uses alpine rather often fwiw

[14:08:21.0000] <Hixie>
Yuhong: why would you want to know if a page was valid against some know-buggy rules?

[14:08:23.0000] <timeless>
/me uses mail.google.com/mail/u/0/x/ often enough

[14:08:24.0000] <Velmont>
timeless: I couldn't test irccloud because it's not possible to make a user :P

[14:08:37.0000] <Yuhong>
But most validators do support it.

[14:08:41.0000] <timeless>
Velmont: i can probably get you an invite if you're interested

[14:08:56.0000] <timeless>
Hixie: your patience continues to impress me

[14:09:08.0000] <Hixie>
Yuhong: most validators don't support any version of HTML from the past decade

[14:09:23.0000] <Velmont>
timeless: I'll make an operatester account. I like irssi for IRC myself, and have dogfooded on the Opera IRC client a bit, -- but then I went back to irssi :P

[14:09:25.0000] <timeless>
(most developers don't use any validators)

[14:09:28.0000] <Hixie>
Yuhong: so i don't think argument from popularity is a good idea in this case :-)

[14:09:48.0000] <Yuhong>
Except that Validator.nu does support validating against HTML4.

[14:10:12.0000] <Hixie>
Yuhong: so?

[14:10:17.0000] <timeless>
i don't think that's from the last decade

[14:10:28.0000] <timeless>
/me wonders what the pub date on html 4 is anyway

[14:10:49.0000] <Hixie>
timeless: 1997-2001 depending on which minor revisionversion you mean

[14:10:54.0000] <Hixie>
revision/version even

[14:10:56.0000] <timeless>
 July 8, 1997 -- The World Wide Web Consortium (W3C) today announced the first public working draft of HyperText Markup Language (HTML) 4.0

[14:11:17.0000] <timeless>
Yuhong: note that the past decade doesn't go past 2002 :)

[14:12:27.0000] <Yuhong>
That is because W3C was screwing up with XHTML2.

[14:13:49.0000] <timeless>
/me shrugs

[14:13:53.0000] <MikeSmith>
Yuhong: I think the HTML4 support in validator.nu is likely to be dropped eventually

[14:14:03.0000] <timeless>
i think it's a transitional thing

[14:14:11.0000] <timeless>
to make it easier for validator.nu to replace the w3 validator

[14:14:15.0000] <MikeSmith>
yeah

[14:14:25.0000] <timeless>
when html5 is published, i'd expect them to formally favor html5

[14:14:30.0000] <timeless>
and then probably drop it a few months later

[14:14:36.0000] <timeless>
or at least hide it

[14:14:48.0000] <Hixie>
html5 was published in like 2007

[14:15:04.0000] <timeless>
/me shrugs

[14:15:08.0000] <Velmont>
timeless: irccloud seems a bit strange in their free plan, -- because you can only connect to irc.irccloud.com. I'm at loss as to how they'll get users hooked in the first place then.

[14:15:20.0000] <timeless>
/me doesn't want to get into html5 pub dates

[14:15:28.0000] <timeless>
Velmont: the beta doesn't enforce the limits

[14:15:34.0000] <timeless>
which makes addiction *much* easier :)

[14:16:01.0000] <timeless>
Velmont: everyone knows why drug dealers are successful

[14:16:11.0000] <timeless>
you give the product away for free :)

[14:16:18.0000] <timeless>
and i don't mean "you can only smoke it here"

[14:16:18.0000] <Yuhong>
Yea, my point is that validators do let you validate against different HTML versions.

[14:16:22.0000] <timeless>
i mean "free"

[14:16:31.0000] <timeless>
Yuhong: that's a transitional thing

[14:16:44.0000] <timeless>
but i don't see how that helps you make a request to hixie

[14:16:49.0000] <timeless>
nor do i understand what the request is

[14:17:00.0000] <timeless>
if you're actually doing maintenance on a web site

[14:17:01.0000] <Hixie>
Yuhong: my point is that such features are pointless

[14:17:06.0000] <timeless>
you want it to work w/ modern browsers

[14:17:07.0000] <Yuhong>
So why not use the versions produced by W3C HTML WG as targets?

[14:17:25.0000] <timeless>
and the current spec already works to ensure it doesn't break older browsers (in general)

[14:17:28.0000] <Hixie>
Yuhong: because there's no point using _any_ fixed reference as a target, as it is by definition going to be out of date

[14:17:31.0000] <timeless>
and future specs also do that

[14:17:51.0000] <timeless>
a validator won't tell you if a page/site will work w/ a given browser

[14:17:52.0000] <Hixie>
Yuhong: whether that's one of the w3c fixed versions or any of the thousands of whatwg fixed versions

[14:17:56.0000] <timeless>
you'd need to use that actual browser

[14:18:24.0000] <timeless>
if you have validation errors while you're actually fixing your site, you fix them.

[14:18:30.0000] <jgraham>
Validating against some old version of a spec doesn't ensure that the page will work either in browsers that were contemporary with the spec or in modern browsers

[14:19:50.0000] <Hixie>
Validating against some old version of a spec ensures only that you are using known-buggy rules for validatior

[14:19:53.0000] <Hixie>
validation

[14:20:04.0000] <Hixie>
it is an entirely futile task imho

[14:22:03.0000] <timeless>
Yuhong: why does validation interest you in the first place?

[14:22:20.0000] <timeless>
... back to Hixie 's original question: What actual problem are you trying to solve?

[14:22:40.0000] <Yuhong>
OK, I realized it now.

[14:25:13.0000] <Velmont>
:-)

[14:25:19.0000] <Yuhong>
And consolidating features into versions is good for buzzwords but don't fit reality.

[14:25:46.0000] <Velmont>
Yuhong: You're fast man.

[14:26:01.0000] <Velmont>
Normally people use a long time to realize stuff like that.

[14:26:28.0000] <Yuhong>
I wonder it is time to stop usage of the "HTML5" buzzword.

[14:27:43.0000] <Yuhong>
I once saw a comic in which

[14:27:57.0000] <Yuhong>
I once saw a comic in which "HTML5" was defined as what is not in IE8.

[14:28:17.0000] <Yuhong>
I think I posted a link to it before.

[14:28:19.0000] <karlcow>
comic sans

[14:28:29.0000] <Yuhong>
http://msdn.microsoft.com/en-us/library/ff405772(v=vs.85).aspx

[14:28:49.0000] <Yuhong>
Not the comic, but "IInternet Explorer HTML 4.01 Standards Support Document"

[14:30:19.0000] <timeless>
i'd like to +1 Velmont 's you're fast note

[14:35:18.0000] <Yuhong>
AFAIK the use of "HTML5" as a buzzword began before WHATWG declared it a living standard.

[14:35:58.0000] <Hixie>
it was one of the reasons we gave up on the name :-)

[14:38:58.0000] <Yuhong>
BTW, on the Quirks Mode "Standard", I think this has a good explaination: http://hsivonen.iki.fi/almost-precedent/

[14:43:33.0000] <Yuhong>
So what do you think the "HTML5" parser should be called?

[14:43:58.0000] <Hixie>
"HTML parser"? :-)

[14:44:44.0000] <Yuhong>
Except there was also the old one.

[14:44:51.0000] <Yuhong>
"spec-compliant parser"?

[14:44:57.0000] <Hixie>
old one?

[14:46:15.0000] <Yuhong>
I mean, browsers was touting the "HTML5" parser as a feature.

[14:46:45.0000] <Hixie>
there is in theory not supposed to be much difference between the old parser browsers had and the specced parser

[14:46:49.0000] <Hixie>
that's the whole point :-)

[14:47:02.0000] <Yuhong>
The most recent being IE10.

[14:52:15.0000] <Yuhong>
Most features are easy to refer to without using the "HTML5" term, but...

[14:53:18.0000] <Hixie>
there's not, in my mind, any reason to refer to "implementing the parser according to the spec"

[14:53:26.0000] <Hixie>
at least not in marketing documents

[14:57:44.0000] <Yuhong>
http://meta.slashdot.org/comments.pl?sid=2769161&cid=39610913

[14:57:53.0000] <Yuhong>
Notice no use of "HTML5" anywhere.

[15:31:35.0000] <zewt>
... can anyone explain why <title> is "required"? so far, nobody's ever been able to give any rationale, and I can't even contrive any reason, heh

[15:34:34.0000] <edwardbc_>
well, a document should have a title as part of its metadata, so why shouldn't it be there?

[15:34:42.0000] <zewt>
"should" != "must"

[15:35:30.0000] <zewt>
http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jun/0282.html suggests there's some more pressing reason than "titles are nice"

[16:43:40.0000] <annevk>
data:text/html,<title>test</title> vs data:text/html,<title></title> is interesting

[16:44:48.0000] <zewt>
afaik empty title usually behaves the same as not specifying any


2012-07-06
[19:43:14.0000] <kennyluck>
"Grand Unified HTML Parser"

[19:44:29.0000] <Hixie>
zewt: <title> is required because UIs break down if they don't have a way to refer to the page

[19:47:05.0000] <zewt>
"break down"?

[19:47:12.0000] <Hixie>
become crappy

[19:47:18.0000] <Hixie>
bad user experience

[19:47:22.0000] <zewt>
i've made plenty of test pages with no title and it works just fine

[19:47:49.0000] <Hixie>
our opinions on what is "just fine" are clearly quite different :-)

[19:50:04.0000] <zewt>
given that <title></title> appears to be legal, and that browsers treat that the same (from the user's perspective) as no title element at all, I'm not seeing the problem

[19:50:59.0000] <zewt>
(the URL as a default title works just fine)

[20:01:09.0000] <MikeSmith>
is "called getter WebElement findElement(in Locator locator)" valid WebIDL?

[20:08:33.0000] <Hixie>
yeah we should probably make that non-conforming

[20:08:45.0000] <heycam>
MikeSmith, should be "caller"

[20:08:54.0000] <heycam>
MikeSmith, also getters have to get either unsigned long or DOMString as their first arg

[20:09:17.0000] <MikeSmith>
OK

[20:09:19.0000] <MikeSmith>
thanks

[20:09:46.0000] <MikeSmith>
shouldn't be "legacycaller"?

[20:09:56.0000] <zewt>
(yeah, please create more bogus validator warnings)

[20:29:12.0000] <Hixie>
zewt: it's not bogus, having an empty <title> seems rather pointless and probably an authoring mistake

[20:31:48.0000] <zewt>
or "just use the URL and stop nagging me to add a pointless <title>"; the URL is just fine as a default title (and better than the "foo" I'd put in if I had to enter something to shut up the validator)

[20:33:39.0000] <Hixie>
URLs aren't really supposed to be user-facing, notwithstanding that they're plastered on busses and are a core part of the web's origin security model

[22:40:53.0000] <zcorpan>
matjas: the grunt thing sounds interesting, care to give some hints for how to set it up? (i tried reading the documentation but still didn't really know how to do it)

[23:12:15.0000] <MikeSmith>
heycam: "caller getter ..." thing I mentioned earlier, should be "legacycaller getter ..." instead?

[23:12:23.0000] <heycam>
MikeSmith, ah yep

[23:12:29.0000] <MikeSmith>
ok

[23:33:48.0000] <zcorpan>
MikeSmith: can i has a bugzilla component?

[23:33:57.0000] <MikeSmith>
sure

[23:34:04.0000] <MikeSmith>
for quirks-mode?

[23:34:28.0000] <zcorpan>
yep

[23:43:18.0000] <matjas>
zcorpan: install node, then run `npm install -g grunt`

[23:44:36.0000] <matjas>
zcorpan: then, in your spec dir, create a `grunt.js` file — here’s mine: http://pastebin.com/raw.php?i=hR6yAev7 (required running `npm install grunt-shell` in that dir first)

[23:45:17.0000] <matjas>
after that you can just cd into the spec dir, and enter `grunt` to watch the index.src.html file for changes and automatically regenerate index.html when needed

[23:46:10.0000] <zcorpan>
matjas: thanks, i'll give it a go later

[00:10:37.0000] <MikeSmith>
zcorpan: added https://www.w3.org/Bugs/Public/describecomponents.cgi?product=Web%20Platform%20%28other%29

[00:12:59.0000] <zcorpan>
thanks

[00:34:32.0000] <MikeSmith>
hey cool another ambiguously-worded unimplementable CP thanks to "amicable resolution"

[00:35:31.0000] <MikeSmith>
""Conformance checkers *may* use the information given on the WHATWG Wiki MetaExtensions page to establish if a value is allowed or not: values defined in this specification or marked as "proposed" or "ratified" must be accepted, whereas values marked as "discontinued" or not listed in either this specification or on the aforementioned page must be *reported* as invalid."

[00:37:38.0000] <MikeSmith>
so conformance checkers both must use the information on the wiki and also may use it

[00:38:00.0000] <MikeSmith>
oh and btw "Analogic edit has to be done in sections "4.2.5.4 Other pragma directives" and "4.12.4.14 Other link types"."

[00:38:21.0000] <MikeSmith>
real precision there

[01:28:48.0000] <matjas>
Ms2ger: how’s that patch coming along?

[01:31:25.0000] <matjas>
MS is currently looking into the two bug reports re: the HTML String.prototype methods \o/

[01:32:41.0000] <matjas>
Opera and V8 are waiting for other browsers to escape " into &quot; first

[01:32:49.0000] <matjas>
/me files JavaScriptCore bug

[01:54:12.0000] <opera>
timeless: Hmmm. Works rather well here ;_)

[02:24:34.0000] <zcorpan>
you can now select text in quirks mode and a link will appear to file a bug

[02:26:23.0000] <zcorpan>
(should be keyboard-accessible if you can select text with the keyboard; the buglink has accesskey 1)

[02:27:38.0000] <odinho>
zcorpan: Gotta love that, I likey :D

[02:28:38.0000] <zcorpan>
the script should be spec-agnostic so can be copied verbatim (but needs to be placed after the "file a bug" link to work)

[02:38:30.0000] <zcorpan>
(also two lines of css need to be copied)

[02:49:29.0000] <annevk>
MikeSmith: I sure hope Hixie stops paying attention to them

[02:51:09.0000] <MikeSmith>
annevk: I bet Hixie didn't even read that one

[02:51:17.0000] <MikeSmith>
I didn't myself until today

[02:52:09.0000] <MikeSmith>
that's failure point number in this scheme

[02:52:50.0000] <MikeSmith>
that it depends on people actually being motivated to invest time in reading totally half-baked change proposals

[02:52:55.0000] <annevk>
it's clear that the bureaucrats have taken over and everyone doing real work moved elsewhere

[02:53:28.0000] <MikeSmith>
which CPs nobody is responsible for vetting first to make sure they are sane before they are hoisted up to the whole WG

[02:54:44.0000] <MikeSmith>
annevk: I would at least like to get a heads-up about any CP that involves requirements for conformance checkers

[02:54:53.0000] <MikeSmith>
and/or Henri should

[02:55:19.0000] <MikeSmith>
from the title of that CP I had no idea what it was supposed to be about

[02:56:02.0000] <MikeSmith>
it sounded to me like another pet "I don't like the way you are handling citation style of the references section" thing

[03:00:46.0000] <annevk>
I wonder why the W3C is only looking for an editor for HTML

[03:15:29.0000] <Stevef>
putting all CP's in the crap basket and not worthy of consideration is not helpful or useful, some CPs are useful and result in changes that make HTML better me thinks...

[03:22:03.0000] <annevk>
I haven't seen any

[03:24:07.0000] <annevk>
CPs are just some layer of bureaucracy to please the insane crowd coming from WAI as far as I can tell; it's not actually helping produce a high quality specification

[03:24:40.0000] <Stevef>
well you and I have differing opinions on whats useful and makes HTML better, luckily neither of us has a stranglehold on what is good for the web

[03:25:37.0000] <annevk>
nobody does

[03:25:50.0000] <annevk>
but that seems besides the point

[03:26:12.0000] <Stevef>
annevK: yeah thanks for that mature characterization, I am one of those insane WAI people

[03:26:38.0000] <zcorpan>
annevk: wdyt about the file-bug script in quirks mode? something you want to adopt in your specs? if so, maybe we can rip it out and maintain it as an external script somewhere?

[03:30:19.0000] <annevk>
Stevef: WAI seems largely non-functional and has way too much bureaucracy; they somehow successfully ported that to the HTML WG

[03:31:21.0000] <annevk>
zcorpan: looks interesting yeah!

[03:31:29.0000] <annevk>
zcorpan: what's the status thing?

[03:32:32.0000] <zcorpan>
status? oh, that's just the boxes "Status: Awaiting implementation feedback. Has tests." that i maintain manually for now

[03:33:10.0000] <annevk>
okay

[03:33:25.0000] <zcorpan>
where should we put the script? should it inject the styles it needs also?

[03:34:35.0000] <annevk>
I think that would be nice

[03:35:16.0000] <annevk>
you could put it inside the quirks-mode repository I suppose

[03:35:36.0000] <annevk>
then I can use <script src=/hg/quirks-mode/raw-file/tip/file-bug.js> or some such

[03:36:14.0000] <annevk>
zcorpan: can we make it use the "file a bug" link to pick the appropriate component?

[03:36:37.0000] <annevk>
zcorpan: so we don't need to pass in any data specially

[03:38:03.0000] <zcorpan>
it uses the component from the "file a bug" link it expects to be there

[03:38:17.0000] <zcorpan>
good enough?

[03:38:33.0000] <Stevef>
annevk: the HTML WG process has nothing to do with WAI, its maciej's and sam rubys baby

[03:38:55.0000] <annevk>
zcorpan: ah yeah it already does

[03:39:40.0000] <annevk>
Stevef: only because WAI insisted on "a clear process" they could then abuse

[03:40:05.0000] <Stevef>
annevk: rather than the no process that whatwg could abuse

[03:40:16.0000] <annevk>
zcorpan: looks pretty great

[03:43:20.0000] <annevk>
zcorpan: could use cloneNode btw

[03:44:01.0000] <annevk>
Stevef: Given the number of differences between W3C HTML and WHATWG HTML I think the WHATWG worked pretty well and way faster too

[03:44:40.0000] <annevk>
s/worked/works/

[03:48:06.0000] <Stevef>
annevk: i do not disagree that the whatwg works well in some respects, but when it comes to disagreement it fails, it is not the perfect model, which is why I work within the html wg framework as stakeholders other than chosen have a voice.

[03:49:05.0000] <Stevef>
s/chosen/the chosen/

[03:50:04.0000] <annevk>
but as long as the HTML WG is ignored that hardly seems like an effective use of time

[03:50:28.0000] <annevk>
and the HTML WG seems even further away from "the perfect model"

[03:51:00.0000] <zcorpan>
annevk: http://dvcs.w3.org/hg/quirks-mode/raw-file/tip/file-bug.js

[03:51:02.0000] <annevk>
it's sort of working now because Hixie is still cooperating to some extent

[03:51:46.0000] <annevk>
zcorpan: maybe advice to just put the id on the <script>?

[03:53:10.0000] <Stevef>
its not always ignored, just by a set of people who you deam as important, my main issues have been with author conformance stuff which is not the domain of browser implementers, if hixie chooses not to chnage the whatwg spec in that regard it is a problem but not end of the world

[03:53:32.0000] <Stevef>
s/chnage/change/

[03:54:19.0000] <annevk>
zcorpan: added to fullscreen, cool

[03:54:47.0000] <annevk>
zcorpan: only problem seems to be source code newlines get copied over

[03:55:52.0000] <Ms2ger>
timeless, another etiquette point... No "This email may contain confidential information" signatures ;)

[03:56:42.0000] <Ms2ger>
MikeSmith, no, it should not have either "caller" or "legacycaller"

[03:57:32.0000] <zcorpan>
annevk: changed to dispatch on data-no-style=""

[03:58:13.0000] <zcorpan>
annevk: any suggestion about source newlines?

[03:58:33.0000] <Stevef>
annevk: without the whatwg i would not have had the impetus or freedom within the W3C to edit the various documents I have: alt guidance/using aria in html/html acc implementation, so cheers for that!

[03:58:37.0000] <Ms2ger>
annevk, it's looking for an HTML *WG* editor, not an HTML editor, no?

[03:58:54.0000] <annevk>
zcorpan: no sorry :/

[03:59:37.0000] <annevk>
Ms2ger: sure, editor for the HTML spec

[03:59:49.0000] <jgraham>
Ms2ger: You think they're hiring an assasin?

[04:00:25.0000] <Ms2ger>
jgraham, why would they, they already have MikeSmith :)

[04:00:46.0000] <annevk>
Stevef: without the WHATWG the W3C would still be a closed shop with lots of F2F meetings and teleconferences

[04:01:21.0000] <Stevef>
annevk: thats what I was saying

[04:01:46.0000] <jgraham>
annevk: Like the CSS WG, then?

[04:02:22.0000] <annevk>
jgraham: well technically CSS is chartered as a public group now, but somehow they kept their Member-only list around without crackdown from above

[04:03:10.0000] <annevk>
Stevef: ah sorry

[04:03:18.0000] <jgraham>
And they mainly work by F2F meetings and teleconfs

[04:04:27.0000] <jgraham>
http://dev.opera.com/articles/view/opera-spdy-build/

[04:05:10.0000] <annevk>
jgraham: uhuh :/

[04:05:14.0000] <Ms2ger>
jgraham, if by "work" you mean "bikeshed the names of flexbox properties" :)

[04:05:38.0000] <annevk>
flexbox should have been ready a decade ago

[04:05:49.0000] <annevk>
but you know, resources

[04:06:00.0000] <jgraham>
Ms2ger: I am still in denial about that. I am working on the hypothesis that you must have been joking

[04:06:30.0000] <Ms2ger>
Should I point out they did another rename this week?

[04:06:44.0000] <jgraham>
That's the one I'm in denail about

[04:06:47.0000] <jgraham>
*denial

[04:07:11.0000] <Stevef>
annevk: I can happily edit specs/docs with the minumum of beaurocracy that sometimes challenge ideas in other specs and working groups I don't have to attend telecons unless i want to, i get support and feedback...

[04:07:14.0000] <jgraham>
Because that would be so stupid it would be like writing a big sign saying "Incompetent" and sticking it to your face

[04:08:18.0000] <Stevef>
annevk: and yes that is because the HTML WG is a less confining WG than others thanks to the WHATWG influence

[04:12:53.0000] <Philip`>
jgraham: And then crossing out the "Incompetent" and writing "Foolish" over it

[04:13:06.0000] <Philip`>
and then crossing that out too

[04:28:38.0000] <Ms2ger>
Looks like unprefixed gradients in Gecko are coming soon

[04:30:07.0000] <MikeSmith>
Ms2ger: about caller thanks yeah I think Stephen Stewart already fixed that in the place I had been asking about it, the WebDriver spec

[04:31:28.0000] <jgraham>
*Simon

[04:31:58.0000] <jgraham>
(Unless there are multiple people with very similar names working on WebDriver)

[04:33:27.0000] <MikeSmith>
yeah I meant Simon

[04:36:20.0000] <MikeSmith>
oh cool no week is complete until I get at least one e-mail message from somebody explaining to me what my job is and how to do it

[04:36:40.0000] <Ms2ger>
MikeSmith, assassin? :)

[04:38:47.0000] <MikeSmith>
Superman and Mr. Mxyzptlk

[04:39:03.0000] <jgraham>
Oh, there are four s's in assassin

[04:39:38.0000] <MikeSmith>
I guess I really shouldn't read e-mail on Friday nights

[04:41:37.0000] <annevk>
MikeSmith: haters gonna hate

[04:41:44.0000] <MikeSmith>
yeah

[04:42:31.0000] <MikeSmith>
I do love to get lectured though

[04:43:05.0000] <Ms2ger>
Don't we all?

[04:43:33.0000] <MikeSmith>
especially when I'm doing something that's in contrast a mundane non-Alice-through-the-looking-glass kind of thing, like, say, trying to help get errors in the WebDriver spec corrected

[05:54:04.0000] <matjas>
does Opera allow volunteers to access (parts of) https://bugs.opera.com/?

[05:54:57.0000] <jgraham>
matjas: Yes, some people who have signed NDAs

[05:55:06.0000] <matjas>
jgraham: where do I sign?

[05:55:10.0000] <timeless>
Ms2ger: yeah, so

[05:55:10.0000] <jgraham>
I don't know how you become such a person

[05:55:20.0000] <timeless>
... that is added by my corporation outside my mailer

[05:55:27.0000] <timeless>
... i have *zero* control over it

[05:55:29.0000] <jgraham>
But I think maybe Philip` has access?

[05:55:46.0000] <timeless>
... i used to be upset about that bit, until i discovered how it was configured

[05:56:01.0000] <timeless>
... and learned that it really isn't something people sending it can control

[05:57:52.0000] <jgraham>
matjas: Alternatively, you get bug access with an employment contract ;)

[05:58:42.0000] <timeless>
that's the normal way :)

[06:00:37.0000] <Philip`>
matjas: There's a volunteer thing where you get access to non-public pre-release builds and non-secret parts of the bug tracker

[06:01:04.0000] <Philip`>
(and have to sign an NDA)

[06:01:32.0000] <Philip`>
though nowadays it seems a lot of the pre-release builds are public anyway

[06:01:55.0000] <matjas>
Philip`: do want, if only to check the status of bugs i reported. who do i have to talk to?

[06:02:33.0000] <Philip`>
I have no idea

[06:03:12.0000] <Philip`>
Some Opera person invited me, back when I used to do vaguely useful stuff instead of dossing around

[06:03:20.0000] <matjas>
hah

[06:03:42.0000] <Ms2ger>
timeless, I know, but still, silly corporations :)

[06:03:53.0000] <Philip`>
(mainly since it was useful for Opera if I could test their pre-release builds to report bugs, I guess)

[06:04:29.0000] <timeless>
Ms2ger: well

[06:04:37.0000] <timeless>
to be fair, i don't think we add it to all employees

[06:04:42.0000] <timeless>
i work in the Legal division

[06:04:53.0000] <timeless>
it actually makes sense for them to add that disclaimer to our correspondence

[06:05:02.0000] <timeless>
you really don't want to let the lawyers remove or modify it

[06:05:27.0000] <Ms2ger>
timeless, you're a lawyer now? :)

[06:05:33.0000] <timeless>
no

[06:06:09.0000] <matjas>
I love how anyone can get the CORE-ID of any Opera bug based on its DSK-ID, even though it’s useless information unless you have access to the bug tracker

[06:07:01.0000] <zcorpan>
matjas: i just moved the bug, wrote a tc and assigned it

[06:07:08.0000] <zcorpan>
matjas: thanks for filing

[06:07:25.0000] <zcorpan>
with that i call it a day

[06:07:28.0000] <matjas>
yay \o/

[06:07:55.0000] <matjas>
zcorpan: DSK-369398 → CORE-47441, right?

[06:08:35.0000] <matjas>
:P

[06:09:19.0000] <odinho>
matjas: Since I love following bugsl I'm watching that one :]

[06:09:49.0000] <Ms2ger>
Btw, if bug access is being handed out... I'm in line too :)

[06:09:51.0000] <matjas>
with every Opera employee in #whatwg tracking bugs for me, it’s almost like having real bug tracker access

[06:10:04.0000] <matjas>
Ms2ger: how’s your patch coming along?

[06:10:06.0000] <odinho>
^_^

[06:10:32.0000] <Ms2ger>
matjas, running tests

[06:10:56.0000] <matjas>
Ms2ger: \o/

[06:23:24.0000] <annevk>
matjas: talk to andreas, he can hopefully sort you out

[06:23:28.0000] <Stevef>
hixie: here http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#hit-regions  says "control (default null) An element (that is a descendant of the canvas) to which events are to be routed, and which accessibility tools are to use as a surrogate for describing and interacting with this region" The focusable elements are also interacted and can be activated  by a...

[06:23:29.0000] <Stevef>
...keyboard user who navigates to them so doesn't seem correct to say only for 'accessibility tools'

[06:24:37.0000] <matjas>
annevk: thanks

[06:31:14.0000] <Stevef>
hixie: test and video demos of current canvas sub DOM iimplementation and how it works with keyboard here: http://www.paciellogroup.com/blog/2012/06/html5-canvas-accessibility-in-firefox-13/

[06:44:33.0000] <gsnedders>
matjas: In JIRA if you move a bug from one project to another it gets a new number. The DSK bug number becomes an alias for the CORE bug number. If you move it back to DSK, it gets another new DSK number. :)

[06:46:48.0000] <matjas>
gsnedders: e.g. https://bugs.opera.com/browse/DSK-369398 is kind enough to redirect to https://bugs.opera.com/browse/CORE-47441 before telling me i have no access :)

[06:47:02.0000] <gsnedders>
Hah, that's lovely data leakage.

[08:09:33.0000] <Ms2ger>
Found 99 tests

[08:09:33.0000] <Ms2ger>
99 Pass

[08:10:31.0000] <Ms2ger>
matjas, and it doesn't look like I'm failing other tests...

[08:10:41.0000] <matjas>
Ms2ger: \o/

[08:33:37.0000] <odinho>
Ms2ger: What tests are you using?

[08:33:49.0000] <Ms2ger>
http://mathias.html5.org/tests/javascript/string/

[08:33:58.0000] <odinho>
matjas: I've made a patch for it in Opera :-) But I need to learn all that review stuff etc etc etc. It'll take some time.

[08:35:57.0000] <odinho>
Oh, that's not the ones I was looking for. I was looking for the  v\u061r  stuff.

[08:37:28.0000] <Ms2ger>
matjas, you should check those tests into hg, btw

[08:52:33.0000] <odinho>
matjas: You should update your v\u0061r test in http://mathias.html5.org/tests/javascript/identifiers/

[08:55:14.0000] <matjas>
odinho: done

[11:15:48.0000] <Yuhong>
http://news.ycombinator.com/item?id=3945953

[11:26:16.0000] <zcorpan>
Yuhong: was there something specific in quirks mode that should be changed/corrected?

[11:26:47.0000] <Yuhong>
No, just an editorial change to the spec.

[11:27:08.0000] <Yuhong>
To explain why there is a need for a quirks mode "standard".

[11:27:32.0000] <zcorpan>
what do you suggest?

[11:27:46.0000] <Yuhong>
http://hsivonen.iki.fi/almost-precedent/

[11:28:03.0000] <zcorpan>
do you suggest i put that in the spec verbatim?

[11:28:15.0000] <Yuhong>
Not the entire thing.

[11:28:25.0000] <Yuhong>
"Mode Switches are a Product of Their Time" is the section I am thinking of.

[11:29:04.0000] <zcorpan>
hmm, gotta go, but please elaborate in a bug report, thanks!

[11:30:23.0000] <Yuhong>
Perhaps with some changes.

[11:59:44.0000] <matjas>
I’ve wondered why hsivonen’s site loads 5 MB of fonts for ages now

[16:12:01.0000] <zewt>
apparently firefox used to have an option to not allow target=_blank, and now it's hidden; why do I have to jump through hoops to get things that should be the default :(

[16:12:52.0000] <zewt>
might not really be doing the old thing, either

[16:14:45.0000] <zewt>
ah, got it--browser.link.open_newwindow 1, browser.link.open_newwindow.override.external 3


2012-07-07
[03:23:22.0000] <smaug____>
annevk: thanks, that was quick

[03:23:36.0000] <annevk>
smaug____: was just checking email :)

[03:32:38.0000] <annevk>
http://w3cmemes.tumblr.com/post/26651456066/decider-george-does-not-understand-what-dbaron-is sounds kind of bad

[03:32:54.0000] <annevk>
I hope someone makes sure they're not going to introduce that for identifiers or something dumb like that

[03:36:10.0000] <annevk>
hober: ^^

[04:48:37.0000] <Velmont>
01:14 < zewt> apparently firefox used to have an option to not allow target=_blank, and now it's hidden; why do I have to jump through << yes, hate it.

[04:48:55.0000] <Velmont>
zewt: Found out big chinese sites like 163.com opens EVERY link in _blank. Fscking infuriating.

[08:38:45.0000] <Ms2ger>
smaug____, should exceptions thrown from event handlers be propagated to the dispatchEvent() caller?

[08:42:16.0000] <smaug____>
Ms2ger: no

[08:43:06.0000] <smaug____>
there can be several exceptions, so it would be unclear which exception to throw to dispatchEvent caller

[08:44:26.0000] <Ms2ger>
/me files bugs

[08:46:05.0000] <smaug____>
Ms2ger: do exceptions propagate in some implementations ?

[08:46:22.0000] <Ms2ger>
No, but the specs aren't clear

[09:42:43.0000] <zewt>
Velmont: it's probably the single most abused "feature" on the web, and bizarre that it's allowed

[11:59:00.0000] <Ms2ger>
Looks like unprefixed gradients will be in the next Nightly

[12:05:39.0000] <tantek>
Ms2ger, one bug report at a time, progress.


2012-07-08
[18:06:49.0000] <zewt>
heh, greasemonkeying one thing that IE has always done better than FF: implicit submit on enter when a SELECT is focused

[20:10:48.0000] <nonge>
zewt, sounds like bad behavior to me … when you navigate with keyboard and want tab through the select options, the form gets submitted each time (for each option)

[20:12:37.0000] <zewt>
no it doesn't; up and down changes options, enter submits

[20:13:09.0000] <zewt>
(it shouldn't submit if you press enter to select an option when the dropdown is open, which I can't emulate with GM, but it's good enough as a workaround--whenever I have the dropdown open I'm on the mouse anyway)

[20:54:21.0000] <Hixie>
why is chrome validating fields before the 'click' event runs

[20:54:23.0000] <Hixie>
jeez

[20:55:42.0000] <Hixie>
hm, it's not

[20:55:42.0000] <Hixie>
wtf

[03:36:44.0000] <annevk>
*summon* Ms2ger

[03:37:17.0000] <annevk>
in 4.7. of http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-event-listener-invoke we could write to not propagate exceptions

[03:37:18.0000] <annevk>
however

[03:37:34.0000] <annevk>
they should still reach window.onerror

[03:40:29.0000] <annevk>
added comments to https://www.w3.org/Bugs/Public/show_bug.cgi?id=17713 instead

[04:32:24.0000] <jarek>
do you know when browser vendors are going to kill the old flexbox?

[04:32:47.0000] <jarek>
it looks like Safari 6 will ship with old flexbox only

[04:32:54.0000] <jarek>
I'm not sure about IE10

[04:33:15.0000] <jarek>
if I recall correctly they have implemented only the old spec as well

[04:35:24.0000] <jarek>
and the new spec is still in WD state, it won't get rewritten from scratch again, right?

[05:15:47.0000] <annevk>
Hixie: maybe you should stop displaying "Issues" on the issues graph

[05:15:59.0000] <annevk>
Hixie: only Bugs and Emails are interesting

[05:23:25.0000] <annevk>
hmm, ended up reading http://grumet.net/weblog/archives/000118.html this morning

[06:46:24.0000] <Velmont>
jarek: Well, IMHO (and I guess vendors agree), it'd be stupid to only support the new version, given that if you have a new version, it's not that hard to also support the old version.

[06:47:47.0000] <Velmont>
zewt: I agree with you there. I want <enter> to submit the form most often.

[07:37:58.0000] <Ms2ger>
/me wanders in

[07:38:19.0000] <Ms2ger>
*summon* annevk

[07:46:31.0000] <Velmont>
Ms2ger: That summoning trick isn't really working all that well...

[07:46:53.0000] <Ms2ger>
He should be here by 8pm ;)

[07:47:21.0000] <Ms2ger>
By which time I may or may not still be around

[07:47:53.0000] <Velmont>
Oh, the suspense.

[08:11:57.0000] <Ms2ger>
Looks like we need to tickle window.onerror

[08:41:46.0000] <Hixie>
anne: i still need to track the XXX markers (what the graph calls issues) -- it's only because i'm tracking them that i've kept them so low :-)

[09:30:53.0000] <Workoh>
Ello folks

[10:07:21.0000] <AlexNRoss>
https://www.youtube.com/watch?v=tWOQrpfWRdQ

[10:11:17.0000] <annevk5>
The water here is too cold

[10:12:31.0000] <annevk5>
Hixie: ah k

[10:13:30.0000] <Ms2ger>
Velmont, what did I say? :)

[10:13:43.0000] <annevk5>
Ms2ger: later might work. Any ideas as to how to define it? Seems to generally apply to callbacks

[10:14:12.0000] <Ms2ger>
Haven't really thought about it

[10:16:25.0000] <annevk5>
Most callbacks are just async so you don't think you have to mention it

[10:17:14.0000] <annevk5>
I think it's something IDL/HTML need to define

[10:17:20.0000] <Ms2ger>
Makes sense

[10:17:34.0000] <annevk5>
Typing uppercase words on a phone sucks

[10:17:39.0000] <zewt>
s/uppercase//

[10:17:57.0000] <Ms2ger>
s/words//

[10:18:27.0000] <annevk5>
Not everything

[10:18:38.0000] <annevk5>
💩

[10:18:49.0000] <Ms2ger>
Is that a pile of poo?

[10:18:51.0000] <zewt>
nice default glyph you've got there

[10:19:06.0000] <annevk5>
Uhuh

[10:19:29.0000] <Workoh>
Hello folks. I've got a question as far as commenting goes, E.G: <!--[if gt IE 8]> <html class="no-js" lang="en"> <![endif]-->

[10:19:59.0000] <Workoh>
Those tags do not do anything beyond comment in a funky way, correct?

[10:21:11.0000] <AlexNRoss>
That is a horrible method of going about it and isn't proper either as the only thing before <html> should be a doctype declaration.

[10:21:26.0000] <zewt>
am I the only person in the world that has always seen that Firefox bug where the save-as dialog can't be submitted without pressing alt-s like there times? heh

[10:21:42.0000] <AlexNRoss>
Even though it is "comment code", it is still before the HTML if the user is IE7 or less.

[10:21:52.0000] <Ms2ger>
Comments there are fine

[10:21:54.0000] <AlexNRoss>
As well, other browsers wouldn't render that gt 8 either.

[10:22:02.0000] <zewt>
has happened to me since days lost to history

[10:22:03.0000] <Ms2ger>
The browser sniffing less so

[10:22:15.0000] <Ms2ger>
zewt, I don't remember hitting that

[10:22:20.0000] <AlexNRoss>
Ms2ger: Fine at a level of it will still be processed. But at a technical level, not fine.

[10:22:35.0000] <Ms2ger>
AlexNRoss, nonsense

[10:22:52.0000] <AlexNRoss>
That is the purpose of the <html> entity

[10:23:07.0000] <AlexNRoss>
Because the document is mime-type text/html so it is already implied that it is HTML.

[10:23:28.0000] <AlexNRoss>
So the <html> tag initiates the beginning of the HTML code.

[10:23:43.0000] <Ms2ger>
I'm not sure what you're trying to say

[10:23:57.0000] <Workoh>
<!--[if lt IE 7]> <html class="no-js lt-ie9 lt-ie8 lt-ie7" lang="en"> <![endif]-->

[10:23:59.0000] <Workoh>
<!--[if IE 7]>    <html class="no-js lt-ie9 lt-ie8" lang="en"> <![endif]-->

[10:24:00.0000] <Workoh>
<!--[if IE 8]>    <html class="no-js lt-ie9" lang="en"> <![endif]-->

[10:24:02.0000] <Workoh>
<!--[if gt IE 8]> <html class="no-js" lang="en"> <![endif]-->

[10:24:04.0000] <Workoh>
 <head>

[10:24:22.0000] <Workoh>
Sorry for the spam.   Give you a bigger picture perhaps.

[10:25:11.0000] <AlexNRoss>
What I'm saying is that comment code is declared in the HTML specification, therefore technically it should be inside the <html></html> as opposed to outside. Will it work outside? Yes. Is it proper? Not really.

[10:25:11.0000] <Workoh>
Simple question is: Are they hacks or will they effectively disable JavaScript?

[10:25:30.0000] <Ms2ger>
You are mistaken

[10:26:02.0000] <AlexNRoss>
Ms2ger: How so?

[10:26:22.0000] <Ms2ger>
There's nothing improper about comments outside the root element

[10:27:18.0000] <Workoh>
Being a bit of an idiot, I was just making sure I could remove those comments as that style of commenting is something I've not looked into.

[10:27:32.0000] <zewt>
(as long as they're after the doctype and charset decls)

[10:28:19.0000] <Ms2ger>
Workoh, nothing in there to disable scripting

[10:28:49.0000] <Ms2ger>
Search for Conditional Comments if you're interested in what they do

[10:28:59.0000] <Workoh>
Will-do. Thank you.

[10:29:25.0000] <Ms2ger>
(The one issue there is that non-IE browsers don't seem to get an html start tag)

[10:29:45.0000] <Workoh>
<!doctype html> above

[10:29:58.0000] <Ms2ger>
But no <html>

[10:30:05.0000] <Ms2ger>
Anyway, I'm off for a bit

[10:30:31.0000] <Workoh>
Hm. It seemed to validate fine but that's 5 for us.

[10:31:27.0000] <Workoh>
...using modernize.js, should of thrown that in.


2012-07-09
[18:25:29.0000] <MikeSmith>
what does "QoI testing" mean?

[18:25:44.0000] <MikeSmith>
http://www.w3.org/2012/06/26-coremob-minutes.html#item06

[18:25:50.0000] <MikeSmith>
quality of implementation?

[18:37:04.0000] <zewt>
(what does "core mobile web platform" even mean? sounds roughly opposite the design of the platform)

[19:35:10.0000] <Hixie>
has the CSS 'cursor' property gone out of favour?

[19:35:20.0000] <Hixie>
i can't get it to use a png as a cursor image on either chrome or firefox

[19:38:59.0000] <zewt>
seems to work with .ico and .png files, but not .jpg

[19:39:12.0000] <zewt>
(not sure why you'd want to use a format without alpha as a cursor, but it's pretty weird that it doesn't work at all)

[19:39:39.0000] <Hixie>
maybe it's the size that matters...

[19:40:04.0000] <Hixie>
aha, yes

[19:40:11.0000] <Hixie>
browsers just don't seem to resize cursors down

[19:40:17.0000] <zewt>
https://zewt.org/~glenn/temp.html

[19:40:35.0000] <zewt>
do they fall back if you give multiple size cursors and earlier ones are too big?

[19:40:54.0000] <gsnedders>
(Opera doesn't support cursor: url(…) at all)

[19:41:31.0000] <Hixie>
zewt: looks like yes

[19:41:46.0000] <zewt>
looks like no on my quick test

[19:41:51.0000] <zewt>
in ff

[19:42:03.0000] <Hixie>
yes in chrome

[19:42:18.0000] <zewt>
yeah

[19:42:31.0000] <Hixie>
also looks like both accept cursors up to 128x128 on mac

[19:42:41.0000] <gsnedders>
OS limit?

[19:42:50.0000] <zewt>
ff seems to be loading both cursors, then ignoring both when the first doesn't work and jumping to the fallback cursor

[19:43:00.0000] <zewt>
gsnedders: yeah, i think the cursor size is an OS thing (but the fallback issue is a browser one)

[19:43:35.0000] <zewt>
oh, chrome also falls back properly

[19:43:48.0000] <zewt>
so ff just doesn't handle them so well

[19:43:50.0000] <zewt>
afk a few

[01:20:41.0000] <zcorpan>
Yuhong: i've reworded the intro a bit, lemme know what you think

[02:17:09.0000] <matjas>
ES6 Annex B is normative for web browsers (yay!), but not for other ES implementations ಠ_ಠ

[02:26:33.0000] <odinho>
matjas: What update did you get on the Opera bug, if any (from JIRA)?

[02:27:01.0000] <matjas>
odinho: none at all :(

[02:27:45.0000] <matjas>
waiting for hallvors to contact me with the NDA details

[02:31:02.0000] <odinho>
matjas: Okay, it has a patch and is reviewed, but it has to bubble up to integration, hopefully it'll be accepted. Haven't run it through regression tests though.

[02:31:38.0000] <matjas>
I did get notified of your comment to the ECMAScript bug, though :)

[03:08:05.0000] <Ms2ger>
Animations and transitions looks like they're going to be unprefixed in tomorrow's Nightly

[03:22:41.0000] <annevk>
what happened here: https://www.w3.org/Bugs/Public/show_bug.cgi?id=17426 ?

[03:23:50.0000] <Ms2ger>
Something strange

[03:29:53.0000] <Ms2ger>
Given an interface declared without [Constructor], should its interface object have a 'length' property?

[03:37:13.0000] <annevk>
Ms2ger: in WebKit and Opera, yup

[03:39:11.0000] <annevk>
Ms2ger: IE/Gecko, no

[03:39:15.0000] <annevk>
Ms2ger: for Node

[03:52:49.0000] <Ms2ger>
annevk, and the value? zero?

[03:55:29.0000] <annevk>
Ms2ger: yeah

[03:55:58.0000] <Ms2ger>
That matches Gecko's new bindings, let's go with that

[03:56:25.0000] <annevk>
poor IE

[03:56:35.0000] <annevk>
does it match IDL?

[03:56:51.0000] <Ms2ger>
I'm not entirely sure

[03:57:14.0000] <Ms2ger>
IDL says that the interface object is a Function object

[03:57:30.0000] <Ms2ger>
And ES seems to suggest that Functions get a length property

[03:57:36.0000] <annevk>
okay

[03:58:02.0000] <annevk>
would be nice if IDL had some examples of how objects would be represented in ES

[03:58:12.0000] <Ms2ger>
OTOH, WebIDL says "Interface objects for interfaces declared with a [Constructor] extended attribute MUST have a property named �length��"

[03:59:44.0000] <Ms2ger>
And by WebKit you mean V8, I guess.

[03:59:50.0000] <Ms2ger>
s/./?/

[04:28:14.0000] <annevk>
Ms2ger: yes

[04:28:25.0000] <annevk>
Ms2ger: but I can test Safari

[04:28:46.0000] <annevk>
Safari says undefined

[04:28:46.0000] <Ms2ger>
So can heycam, I guess :)

[04:29:07.0000] <annevk>
(Version 5.1.7 (7534.57.2))

[05:06:23.0000] <zcorpan>
Ms2ger: is it intentional that using xspec xref doesn't add the referenced spec to the References section?

[05:07:36.0000] <Ms2ger>
Sorta

[05:08:17.0000] <zcorpan>
is it fine that it's not in References, or am i expected to place a reference somewhere?

[05:26:00.0000] <Ms2ger>
I've used something like Some of the terms used in this specification are defined in Web IDL, XML and Namespaces in XML. [WEBIDL] [XML] [XMLNS]

[05:41:57.0000] <zcorpan>
ok

[06:58:34.0000] <matjas>
should CORS apply when requesting a data URL from an http:// resource? I thought `data` was a scheme as per RFC 2397 (in which case it would be a cross-origin request)

[06:59:56.0000] <matjas>
i’m asking because browser implementations seem to differ

[09:16:35.0000] <hool>
hello whatwg

[09:18:05.0000] <Ms2ger>
Good evening

[09:19:08.0000] <zcorpan>
TabAtkins_: "[, ], (, ), [, ]." in css3-syntax, i guess you meant to have {, } in there

[09:36:31.0000] <TabAtkins_>
Yes, I'll fix.

[09:41:13.0000] <JonathanNeal>
Can I access the prototype or property descriptors of className?

[09:41:36.0000] <TabAtkins_>
className is a string, no?

[09:42:18.0000] <JonathanNeal>
yes, but I can't seem to find it on Element.prototype or HTMLElement.prototype, I can override the getter / setter on an element, but I can't even find it's property descriptor.

[09:43:04.0000] <TabAtkins_>
Oh, I see what you're saying.

[09:43:23.0000] <TabAtkins_>
I *think* it should be defined on HTMLElement.prototype, but it wouldn't surprise me if it's magic.

[09:45:15.0000] <hool>
has anyone here worked with frame accuracy in html5 video?

[09:45:26.0000] <hool>
specifically with 29.97fps?

[09:45:29.0000] <hool>
(drop frame)

[09:45:44.0000] <JonathanNeal>
Yea, defineProperty(HTMLElement.prototype, 'className') will not hijack className. I can hijack an individual element via defineProperty(myNode, 'className'), but I can't getOwnPropertyDescriptor for it.

[09:46:59.0000] <JonathanNeal>
Which defeats the purpose of using defineProperty to hijack a getter or setter without destroying the actual functionality. I have never seen this for any other property.

[09:47:30.0000] <TabAtkins_>
Nah, you've got similar magic with event listeners.

[09:47:41.0000] <TabAtkins_>
And it's a bug that we're trying to fix.

[09:53:56.0000] <zcorpan>
JonathanNeal: iirc webidl says that idl attributes are supposed to be visible on the prototype, but opera and webkit don't do that (yet)

[10:00:09.0000] <JonathanNeal>
Well, I found a way around it. I defineProperty the individual node, then on set I use delete to remove the setter, set the value, and then restore the alternate setter.

[10:12:07.0000] <annevk>
matjas: data URLs are considered same-origin under certain circumstances

[10:12:11.0000] <annevk>
matjas: see HTML

[10:13:40.0000] <annevk>
matjas: hmm maybe it needs some additional wording to make that happen

[10:13:47.0000] <annevk>
matjas: I kind of forgot how that works

[10:21:45.0000] <JonathanNeal>
Thanks for the heads up on magic, TabAtkins_. I would have kept trying to do something that didn't work.

[10:22:19.0000] <JonathanNeal>
But now I have a classList polyfill that handles errors, enumerables, className that works in IE8+ https://gist.github.com/2956234

[10:22:59.0000] <TabAtkins_>
Heh, cool.

[10:38:02.0000] <annevk>
hmm

[10:38:05.0000] <annevk>
CSSOM still not updated

[10:38:24.0000] <TabAtkins_>
Poke Glenn and Shane in w3c-css-wg?

[10:38:39.0000] <TabAtkins_>
Even if they're working on it, they should be updating live.

[10:39:29.0000] <annevk>
just observing what happens to stuff I've abandoned

[10:39:52.0000] <TabAtkins_>
Okay, so you're not actually interested in it. ^_^

[10:40:25.0000] <annevk>
well I'm kind of interested in having the IDL updated

[10:40:33.0000] <Ms2ger>
I know complaints to the chairs have been made

[10:40:43.0000] <Ms2ger>
They claimed the editors were "new"

[10:40:54.0000] <Ms2ger>
I prefer "incompetent"

[10:41:08.0000] <annevk>
like the chairs?

[10:41:21.0000] <Ms2ger>
No comment

[10:41:31.0000] <annevk>
oh my, I better go back vacationing

[10:41:58.0000] <Ms2ger>
Or, well, start editing CSSOM again outside the WG

[10:47:32.0000] <annevk>
there's some minor fixing that could be done, but the best long term approach is writing out the CSS model (i.e. rewriting CSS 2.1 a la "HTML5") and then having the API map on that

[10:48:14.0000] <annevk>
that's the main thing I learned writing the CSSOM and I have tried to convey that at various points but the CSS WG is just not wired to think about specifications that way

[10:48:52.0000] <annevk>
it always turned quickly into W3C Process discussions and splitting drafts, etc. etc.

[10:50:58.0000] <annevk>
prescient https://twitter.com/hober/status/5511711

[10:55:35.0000] <hober>
annevk: :)

[10:56:13.0000] <annevk>
hober: or just maybe this keeps following you around because you really are :p

[10:57:07.0000] <hober>
annevk: i think this is the first tweet mentioning 'html5', at least according to the kellan.io/oldtweets search: https://twitter.com/BenWard/status/882773/

[10:57:47.0000] <hober>
annevk: "hober: claiming to not be the bottleneck he really is since 2007."

[10:58:01.0000] <annevk>
nice find

[11:00:29.0000] <Lachy>
Given a WebIDL dictionary defined as: dictionary D { DOMString foo = "" }, I need some clarification about what happens when I pass the value as {foo: null}

[11:00:48.0000] <Lachy>
I initially thought it would stringify as "null", but reading WebIDL, just made me a little confused.

[11:01:33.0000] <Lachy>
oh, no. nevermind. I was misreading the algorithm.

[11:10:28.0000] <Ms2ger>
Yes, "null"

[11:39:04.0000] <JonathanNeal>
length, item, contains, add, remove, and toggle are enumerables on a DOMTokenList. Is this intentional?

[11:41:26.0000] <Ms2ger>
Yes, see http://dev.w3.org/2006/webapi/WebIDL/#es-attributes and http://dev.w3.org/2006/webapi/WebIDL/#es-operations

[11:44:20.0000] <JonathanNeal>
Ms2ger: I couldn't actually understand where the logic was being presented. I thought they would not be enumerable so that for each would only go over the indexes of the list.

[11:45:10.0000] <Ms2ger>
That's never been true in browsers

[11:45:28.0000] <Ms2ger>
for (var a of arr) will do that, though

[11:46:27.0000] <kennyluck>
Ms2ger, just curious who have made the complaints to the chairs (re. CSSOM)?

[11:46:57.0000] <Ms2ger>
I'm not sure I should tell you that

[11:47:28.0000] <kennyluck>
Okay.

[11:49:51.0000] <JonathanNeal>
Ms2ger: yea, I always liked how it did that on Array. To be clear, the spec does not prefer that? If you're able to chime in why, it would satisfy my curiosity, but I wouldn't be hurt if you didn't.

[11:50:25.0000] <Ms2ger>
Probable web-compat issues, like usual :)

[11:51:11.0000] <annevk>
there's no compat issues for DOMTokenList though

[11:52:49.0000] <JonathanNeal>
annevk: do you know why that's the preference then?

[11:53:23.0000] <annevk>
probably just not considered

[11:54:09.0000] <kennyluck>
It's indeed quite annoying to see Glenn post to almost every mailing list and join HTML WG calls but not edit the spec for which he is an editor.

[11:54:40.0000] <annevk>
JonathanNeal: could you file a bug on DOM?

[11:55:00.0000] <JonathanNeal>
if it's a bug, I will, yes.

[11:55:06.0000] <Ms2ger>
It would be rather silly to make DOMTokenList inconsistent with all the other lists, though

[11:55:22.0000] <annevk>
JonathanNeal: well it's something we can consider changing

[12:02:14.0000] <JonathanNeal>
Ms2ger: what other lists are there?

[12:02:19.0000] <JonathanNeal>
Besides Array, of course.

[12:03:29.0000] <annevk>
JonathanNeal: NodeList, StyleSheetList, HTMLCollection, etc.

[12:03:45.0000] <Ms2ger>
NodeList, PropertyNodeList, HTMLCollection, HTMLOptionsCollection, HTMLPropertiesCollection, DOMTokenList, DOMSettableTokenList, ClientRectList, PaintRequestList, TouchList, FileList, SVGLengthList, SVGNumberList, SVGPathSegList, SVGPointList, SVGTransformList,

[12:03:57.0000] <Ms2ger>
Are just a few

[12:04:09.0000] <JonathanNeal>
Right, duh. My bad.

[12:05:13.0000] <JonathanNeal>
length, and item seem to be enumerable on NodeList. What's the link to file a ticket?

[12:05:28.0000] <BenoitRen>
Hi there.

[12:05:40.0000] <annevk>
JonathanNeal: see http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html

[12:06:29.0000] <BenoitRen>
Does anyone know where I can find resources on how to mark up a message board (commonly called forums) semantically? I've been drawing a blank for a long time and googling isn't helping.

[12:11:58.0000] <annevk>
BenoitRen: what problem are you running into?

[12:14:28.0000] <annevk>
friend of mine once did that exercise but it seems it's no longer online

[12:15:19.0000] <BenoitRen>
annevk: It's not a specific problem, really. I don't even know where to start.

[12:16:15.0000] <annevk>
BenoitRen: seems like it would just be a bunch of <article>'s with some content inside them really

[12:16:40.0000] <annevk>
BenoitRen: with topic overview lists as lists, etc.

[12:16:48.0000] <annevk>
BenoitRen: kind of depends on the style you're going for too

[12:17:59.0000] <BenoitRen>
annevk: But the format is always: "poster X with this avatar and this location and amount of posts says <article> and here's some a bunch of buttons to quote it oh and here's a PM button".

[12:19:11.0000] <BenoitRen>
annevk: That's the main problem I'm stuck on. How do I mark something like that up semantically?

[12:19:26.0000] <annevk>
BenoitRen: that's just some <p>s with classes

[12:19:50.0000] <annevk>
BenoitRen: maybe <aside> if it's not too relevant

[12:19:58.0000] <annevk>
BenoitRen: and I would put that inside the <article>

[12:21:30.0000] <TabAtkins_>
<article class=thread><article class=post><address>Urist McDwarf <img src></address> <p>Blah blah blah</article>...</article>

[12:21:44.0000] <BenoitRen>
annevk: It's that simple? I find that hard to accept... How would that read?

[12:22:12.0000] <annevk>
BenoitRen: simple is good :)

[12:22:51.0000] <annevk>
Ms2ger: https://www.w3.org/Bugs/Public/show_bug.cgi?id=17318 I'm not sure I follow entirely

[12:23:07.0000] <BenoitRen>
annevk: Sure, but a HTML document like that seems weird. One paragraph is just a poster name, then the avatar, then a paragraph with just the post count, and then suddenly actual text?

[12:23:26.0000] <annevk>
Ms2ger: oh I do now

[12:23:49.0000] <Ms2ger>
Order :)

[12:23:52.0000] <TabAtkins_>
BenoitRen: Yeah, that's about right.

[12:24:05.0000] <annevk>
BenoitRen: I guess you would do <article> <aside> poster info </aside> post </article>

[12:24:40.0000] <annevk>
BenoitRen: maybe <article> <header> poster info </header> post </article>

[12:25:31.0000] <JonathanNeal>
annevk: I went to overview but i couldn't find a link to anything that looked like it was the right place to file a ticket, is there another place?

[12:25:55.0000] <annevk>
JonathanNeal: :(

[12:26:01.0000] <annevk>
JonathanNeal: it's right under "Participate"

[12:26:16.0000] <annevk>
JonathanNeal: but it's kind of bad you didn't find that, I wonder how we can make it more obvious

[12:26:41.0000] <annevk>
Ms2ger: do you have a fix in mind?

[12:28:01.0000] <BenoitRen>
How about... <article><h2>MyAwesomeUserName</h2><aside>My Cool Title</aside><aside>Posts: 9001</aside><p>My awesome post with knowledge.</p></article>

[12:28:33.0000] <annevk>
the username is not the header

[12:28:39.0000] <annevk>
heading*

[12:28:55.0000] <zcorpan>
JonathanNeal: which spec?

[12:29:02.0000] <BenoitRen>
Why not?

[12:29:14.0000] <annevk>
zcorpan: DOM

[12:30:03.0000] <zcorpan>
annevk: maybe bug should be a line on its own?

[12:30:49.0000] <zcorpan>
JonathanNeal: where did you look? :-)

[12:30:51.0000] <Ms2ger>
annevk, for?

[12:31:06.0000] <annevk>
Ms2ger: the replaceChild thing

[12:31:22.0000] <Ms2ger>
Reordering the first few steps of the algorithm, I think

[12:31:44.0000] <annevk>
so sometimes Gecko does throw NotFoundError?

[12:33:25.0000] <annevk>
hmm confusing

[12:34:36.0000] <Ms2ger>
Yeah

[12:34:52.0000] <BenoitRen>
TabAtkins_: I just read the spec on <address>. How you propose using it is valid, but I don't understand why it wouldn't be a header.

[12:35:16.0000] <annevk>
Ms2ger: I was hoping it would be an easy fix, postponed now :)

[12:35:23.0000] <Ms2ger>
Aww :)

[12:35:48.0000] <annevk>
Ms2ger: those algorithms are starting to become scary :/

[12:36:00.0000] <Ms2ger>
The code is worse :/

[12:36:13.0000] <odinho>
Opening for cleanup!!11

[12:36:15.0000] <odinho>
:D

[12:36:36.0000] <Ms2ger>
Why are you using that name this time of the day?

[12:36:57.0000] <odinho>
Yeah, I really need to go home... :-/

[12:37:00.0000] <odinho>
*leaving

[12:37:11.0000] <smaug____>
oh, odinho is odinho only in the office

[12:37:34.0000] <odinho>
smaug____: There's a system to it ;-)

[12:39:55.0000] <Ms2ger>
A method, if you like

[12:45:59.0000] <Hixie>
(updated my e4h proposal)

[12:46:16.0000] <Ms2ger>
Now, your h spec ;)

[12:51:28.0000] <BenoitRen>
/me feels ignored

[12:51:41.0000] <TabAtkins_>
BenoitRen: Sorry, was doing other things.

[12:52:01.0000] <TabAtkins_>
BenoitRen: The poster's name isn't a *heading*.  Headings are title for their section.

[12:52:08.0000] <TabAtkins_>
Someone's name isn't a title for their post.

[12:52:27.0000] <TabAtkins_>
<header> is a different beast, it's structual markup.  It's fine to put it the poster's name in a <header> if you want.

[12:53:15.0000] <BenoitRen>
I thought <header> was a way to cram two headings in one semantically. Like Title: Subtitle.

[12:53:29.0000] <TabAtkins_>
No, that's <hgroup>.

[12:53:33.0000] <TabAtkins_>
English is confusing.

[12:53:34.0000] <BenoitRen>
Oh.

[12:54:49.0000] <BenoitRen>
So I put the poster's name in an address along with their avatar (with empty alt attribute), then the rest of their info in an aside?

[12:55:20.0000] <TabAtkins_>
Sure, that sounds reasonable.

[12:55:56.0000] <Hixie>
BenoitRen: <header> was renamed <hgroup> a few years ago when we added what is now <header>, fwiw

[12:55:58.0000] <zcorpan>
Hixie: it needs xlink: support as well, right? i'm not sure the prefix story really works for writing svg and mathml

[12:56:30.0000] <BenoitRen>
Hixie: Ah, I see. Thanks. By the way, don't forget my W3 bug. ;)

[12:56:48.0000] <Hixie>
zcorpan: why xlink? i thought the svg guys were going away from xlink

[12:56:55.0000] <TabAtkins_>
We are.  Fuck xlink.

[12:57:21.0000] <Hixie>
BenoitRen: i have 506 open bugs on the pile, but am prioritising the 1245 open e-mails first, sorry :-)

[12:57:47.0000] <BenoitRen>
Hixie: B-but I've been waiting for over 6 months! :'(

[12:58:20.0000] <TabAtkins_>
Everyone else has been waiting longer. ^_^

[12:58:28.0000] <Hixie>
BenoitRen: oldest e-mail on my pile is from June 2009. :-)

[12:59:01.0000] <Hixie>
of the things i haven't intentionally deferred, i'm currently up to January 2012 or so

[12:59:12.0000] <Hixie>
so if your bug is older than that, let me know

[12:59:15.0000] <Hixie>
i might have missed it

[12:59:31.0000] <zcorpan>
TabAtkins_: it wasn't clear to me whether that was going to be reality or just fiction

[12:59:35.0000] <Hixie>
(and if it's truly urgent, e.g. a browser vendor is blocked on it, let me know and i'll prioritise it)

[12:59:51.0000] <TabAtkins_>
zcorpan: It's one of the less controversial breaks, so it's reality afaik.

[13:00:05.0000] <Hixie>
zcorpan: if it turns out xlink is important, i don't mind adding it later, but that seems like a minor issue overall

[13:00:21.0000] <zcorpan>
yeah sure

[13:07:43.0000] <BenoitRen>
TabAtkins_: What about the buttons? Things like the PM buttons is simple: just put them in a <p> below the message. But the quote button is often found at the top right of the message... :S

[13:08:58.0000] <Hixie>
TabAtkins_: btw, re <style scoped> (since you're the csswg member who posted on the relevant thread) -- my intent is to leave the HTML spec as is until someone can take over on the CSS side and define the cascade, @global, et al properly, and then I'll coordinate with them to have the rights hooks in the specs

[13:12:33.0000] <zcorpan>
BenoitRen: it's position visually doesn't mean it can't go in a <p> below the message

[13:13:00.0000] <BenoitRen>
zcorpan: It makes it very hard to style, though. :(

[13:13:56.0000] <TabAtkins_>
Hixie: fantasai and I plan to do that later this year.

[13:14:12.0000] <Hixie>
cool

[13:14:27.0000] <TabAtkins_>
BenoitRen: Yeah, that's the rub right now.  It'll get fixed as Flexbox and Grid propagate out.  Do what you need to do for now, and styling will catch up with the proper semantics later.

[13:22:14.0000] <JonathanNeal>
Woot, got it in, thanks annevk, I don't know how I missed it, but I really did. https://www.w3.org/Bugs/Public/show_bug.cgi?id=17726

[13:31:38.0000] <zcorpan>
JonathanNeal: a link should also appear when you select some text in the spec, btw

[13:37:53.0000] <annevk>
Hixie: what's new?

[13:38:06.0000] <Hixie>
new where?

[13:38:07.0000] <Hixie>
what?

[13:38:17.0000] <Hixie>
/me blinks like a deer in headlights

[13:38:36.0000] <annevk>
a Google Plus

[13:38:39.0000] <annevk>
ah*

[13:38:45.0000] <Hixie>
oh on e4h?

[13:39:32.0000] <Hixie>
attributes="can have inline {substitutions}" and the new checked?={booleanexpression} or title?={title} syntax for optional attributes

[14:12:04.0000] <annevk>
neat

[14:12:37.0000] <annevk>
and there's @<test>{substitutions}</test> too right?

[14:14:01.0000] <annevk>
confirmed

[14:14:13.0000] <Hixie>
yeah

[14:14:39.0000] <TabAtkins_>
What does @ do, repeat as a list?

[14:15:03.0000] <Hixie>
nothing really, it just prevents XSS attacks by being invalid XML

[14:15:52.0000] <TabAtkins_>
Explain?

[14:17:22.0000] <Hixie>
you at your cube?

[14:18:01.0000] <TabAtkins_>
yeah

[14:18:31.0000] <Hixie>
it's hard to explain over irc. let me just come over.

[14:18:35.0000] <Velmont>
oh man

[14:18:58.0000] <Velmont>
/me was following along from home ;]

[14:19:04.0000] <Velmont>
But go do that high bandwidth thing.

[14:22:24.0000] <annevk>
Velmont: http://code.google.com/p/doctype-mirror/wiki/ArticleE4XSecurity

[14:22:47.0000] <gsnedders>
jwalden: I got __proto__ into Carakan again! :) I got __proto__ reverted in Carakan again! :(

[14:23:01.0000] <jwalden>
\o/ /o\

[14:23:16.0000] <jwalden>
(I read that left to right and reacted exactly that way :-D )

[14:23:35.0000] <gsnedders>
Well, at least we know how the bidi algorithm applies here!

[14:23:49.0000] <jwalden>
I am still busy shaving some yaks, but I can almost see something pretty underneath it, so I'm making good progress

[14:24:13.0000] <jwalden>
I leave for a month-long vacation probably next Wednesday, so I need to push here to get this done now

[14:24:21.0000] <gsnedders>
:(

[14:25:21.0000] <annevk>
https://twitter.com/eeppa/status/221661734824443904 :)

[14:26:23.0000] <Velmont>
annevk: Trying to get us to start using RDF? :P

[14:26:46.0000] <annevk>
depends, are you a hipster?

[14:27:23.0000] <Velmont>
Are we not all? :]

[14:29:00.0000] <gsnedders>
jwalden: I'm just hoping I still beat you. (Though jl seems to be doing most of the follow-up work now…) :P

[14:29:05.0000] <annevk>
relevant for whatwg⊙wo: http://i.qkme.me/3q13ei.jpg

[14:29:59.0000] <hober>
that should be revise *an* rfc :)

[14:30:02.0000] <jwalden>
/me wasn't a hipster before it was cool not to be a hipster

[14:30:14.0000] <jwalden>
hober++

[14:31:01.0000] <hober>
annevk: also, should you post that one or should i? :)

[14:31:02.0000] <annevk>
hober: jwalden: have it your way: http://i.qkme.me/3q13fu.jpg

[14:31:05.0000] <TabAtkins_>
Depends on your pronunciation, I suppose.

[14:31:17.0000] <annevk>
hober: you are at work, so you can do it :)

[14:31:20.0000] <jwalden>
\o/

[14:31:21.0000] <TabAtkins_>
"reff-seh"

[14:31:22.0000] <Hixie>
Velmont: basically, look up the e4x security problems; the @ thing just stops those being viable in e4h

[14:31:28.0000] <hober>
annevk: hahahaha. :)

[14:32:12.0000] <TabAtkins_>
Velmont: The confusing bit for me was that I didn't realize the @ came at the *start* of the document.  That's why an e4h literal is invalid XML - it has character data before the root element.

[14:32:16.0000] <gsnedders>
TabAtkins_: No, it shouldn't, as it'll be the same phonemes. But, well, whether you treat it as an acronym or an initialism would change it.

[14:32:45.0000] <TabAtkins_>
gsnedders: That's what I'm saying, yeah.

[14:33:17.0000] <BenoitRen>
Well, I'm off. Thanks for the help, all!

[14:33:36.0000] <TabAtkins_>
hober: Got some questions about image-set() impl if you have a few.

[14:33:43.0000] <hober>
annevk: posted.

[14:33:50.0000] <hober>
TabAtkins_: ok, shoot

[14:34:27.0000] <TabAtkins_>
Does the Nx dimension in its syntax work the same as the Nx component in @srcset, where it adjusts the used resolution accordingly?

[14:35:08.0000] <hober>
it does the same intrinsic dimension calculation, if that's what you mean

[14:35:18.0000] <TabAtkins_>
Yeah, that's what I mean. kk.

[14:35:23.0000] <hober>
word.

[14:36:05.0000] <TabAtkins_>
To addess the "Indicating it's optional..." thread, I'm thinking of allowing someone to specify a fallback color as well, for use if the UA can't download any of them, or just doesn't want to download anything.  Sound okay?

[14:36:29.0000] <annevk>
hober: haha, nice line to go with it :p

[14:36:35.0000] <TabAtkins_>
Well, really for the latter, since the former can be done by nesting it in an image().

[14:36:36.0000] <hober>
TabAtkins_: isn't that what image() is for?

[14:37:08.0000] <TabAtkins_>
Basically, "use this color if you're so constrained you'd prefer not to download the images at all".

[14:38:11.0000] <jwalden>
gsnedders: yeah, might or might not happen; these yaks in the way need the work regardless, and it would be painful to hack around them

[14:38:56.0000] <hober>
TabAtkins_: hmm. i doubt there's much of a use case there. also, that breaks the homogeneity of the comma-separated bits

[14:39:27.0000] <TabAtkins_>
Eh, not really.  It's like the other arguments, except with a resolution indicator of 0x. ^_^

[14:39:39.0000] <hober>
well, wait a sec. couldn't you do that with image-set(url(foo.png) 1x, url(bar.ping) 2x, color(black)) or some such?

[14:39:55.0000] <hober>
yeah, also the arguments aren't ordered

[14:40:03.0000] <hober>
i assume the fallback color would come last

[14:40:11.0000] <TabAtkins_>
Yeah, the ordering is irrelevant here - you can distinguish between a color and an image.

[14:40:15.0000] <Velmont>
TabAtkins_: Yeah, I started by looking at the parser. -- I needed the opposite thing, knowing why it could be vulnerable at all :P

[14:40:16.0000] <hober>
which breaks the unorderedness

[14:40:30.0000] <TabAtkins_>
No need to put it last.

[14:41:00.0000] <hober>
I think ultimately YAGNI

[14:42:18.0000] <TabAtkins_>
I disagree.  ^^; Avoiding the roundtrip can be even better than reducing the transfer size.

[14:43:13.0000] <Velmont>
or even put a small dataurl there :P

[14:43:24.0000] <TabAtkins_>
But, hm.  Using image(black) to generate an image, and setting it with an extremely low x multiplier, might be sufficient.

[14:43:44.0000] <TabAtkins_>
Perhaps special-case 0x, or require 1x for normal images and special-case a lack of an x multiplier.

[14:44:07.0000] <hober>
lack-of-x-multiplier should be equivalent to 1x

[14:44:36.0000] <TabAtkins_>
That's what I thought.  Then we'd need either 0x or some other indicator in place of an x multiplier.

[14:44:46.0000] <hober>
i think even if you really want such a fallback, doing it with image(black) 0.1x is better: it's weird looking, and what you're doing is weird. also, it keeps the feature straightforward

[14:45:22.0000] <hober>
the problem with 0x or <1x with an image that has no intrinsic size is, well.

[14:45:28.0000] <Hixie>
ojan: did you ever decide (when implementing this for webkit) what the intrinsic width of a seamless iframe should be when it's floating?

[14:46:01.0000] <TabAtkins_>
hober: The problem is identical for vector images without intrinsic sizes.

[14:46:41.0000] <TabAtkins_>
hober: In that case the value of the x doesn't actually do anything, and the intrinsic size is determined normally for dimensionless images, as specified in Image V"

[14:46:45.0000] <TabAtkins_>
Values.

[14:47:41.0000] <TabAtkins_>
hober: Is image-set() exposed to the web anywhere yet?

[14:48:17.0000] <hober>
-webkit-image-set() is enabled in ToT WebKit

[14:48:34.0000] <TabAtkins_>
Okay, it wasn't in m21 so I was wondering.

[14:48:52.0000] <hober>
m21?

[14:48:57.0000] <TabAtkins_>
Chrome 21.

[14:49:00.0000] <TabAtkins_>
Current dev channel.

[14:49:22.0000] <hober>
ahh. i dunno if chrome enables it

[14:50:42.0000] <TabAtkins_>
It would be nice if we didn't do so yet. ^_^

[14:53:00.0000] <Hixie>
anyone know if any non-chrome browsers implement seamless="" yet?

[14:53:28.0000] <Hixie>
TabAtkins_: any suggestions on what i should make the intrinsic width of shrinkwrapping seamless iframes?

[14:53:54.0000] <TabAtkins_>
I assume you dont' want to expose the min-content width of the contained document.

[14:53:58.0000] <TabAtkins_>
So... probably 300px.

[14:54:50.0000] <Hixie>
i think the ideal width to expose would be the canvas shrink-wrap width, but that seems impossible to define in a way that isn't sometimes impossible to calculate

[14:55:40.0000] <TabAtkins_>
The canvas doesn't have a shrink-wrap width, unless you mean the document's shrink-wrap width.  That's always possible to calculate, but it will expose data about the document.

[14:56:05.0000] <Hixie>
exposing data is fine, these are always same-origin

[14:56:09.0000] <Hixie>
we're already exposing the height

[14:56:48.0000] <TabAtkins_>
Oh, okay.  Then yeah, just get the document's own intrinsic width.  That's always well-defined.

[14:57:02.0000] <Hixie>
fair enough

[14:57:36.0000] <Hixie>
(what should i reference?)

[14:57:47.0000] <TabAtkins_>
CSS 2.1.

[14:59:02.0000] <Hixie>
i don't see anything in css2.1 about documents having intrinsic dimensions, unless i'm missing something.

[14:59:13.0000] <Hixie>
there's stuff about the root element, but that depends on the viewport

[14:59:17.0000] <Hixie>
and it's the viewport i'm trying to size

[14:59:51.0000] <TabAtkins_>
You want the fit-content size of the root element.

[15:00:22.0000] <TabAtkins_>
fit-content depends on min-content, max-content, and the available space, none of which are dependent on the viewport size.

[15:00:54.0000] <TabAtkins_>
Though, hm, actually the available space would depend on the viewport size naively.

[15:01:16.0000] <TabAtkins_>
So instead, just use the definition directly - max(min-content, min(max-content, available space for the iframe))

[15:02:06.0000] <Hixie>
"available space for the iframe" is undefined in this case, i think (we're talking about out-of-flow or inline-blocks iframes)

[15:02:23.0000] <TabAtkins_>
No, it's well-defined.  It's the width of the iframe's containing block.

[15:02:32.0000] <Hixie>
(and fit-content doesn't seem to be in css2.1)

[15:02:34.0000] <TabAtkins_>
Which, in the worst case, is the width of the iframe's viewport.

[15:03:01.0000] <TabAtkins_>
No, it's called "shrinkwrap width" or something in 2.1.  fit-content is defined in Writing Modes currently.

[15:03:21.0000] <Hixie>
oh. css2.1 doesn't define the shrink-wrap width. :-)

[15:03:29.0000] <Hixie>
that's why i was asking in the first place :-)

[15:03:48.0000] <TabAtkins_>
(By "iframe's viewport" I mean the viewport of the document the iframe is in, not the viewport it defines for its contained document.)

[15:04:15.0000] <Hixie>
(sure)

[15:04:41.0000] <TabAtkins_>
Okay, then refer to Writing Modes, say that it's the fit-content width of the document's root element, except with the available space of the iframe used in place of the available space in the algorithm.

[15:04:46.0000] <Hixie>
even if i do use the css2.1 underdefined "shrink-to-fit" width, that still won't really work, i think.

[15:04:56.0000] <Hixie>
because if the document just consists of positioned elements, the shrink-to-fit width will be 0

[15:05:01.0000] <TabAtkins_>
Yes.

[15:05:10.0000] <TabAtkins_>
That's a problem?

[15:05:52.0000] <Hixie>
well for the height i avoid the problem by fixing the width and using the canvas boundary extents or some such

[15:06:06.0000] <TabAtkins_>
Why are you going to that much trouble?

[15:06:20.0000] <TabAtkins_>
Just let it fall down to 0x0.

[15:06:38.0000] <Hixie>
that seems... less than ideal

[15:06:41.0000] <TabAtkins_>
They can set width/height on either the iframe, the document's root, or something else in there if they want.

[15:06:52.0000] <TabAtkins_>
You're trying to be smarter than necessary. ^_^

[15:07:36.0000] <Hixie>
story of my life :-)

[15:08:02.0000] <Hixie>
i'm thinking of just punting and making it 300x150 if they're floating/positioning the iframe

[15:08:10.0000] <Hixie>
or 300xresulting-height

[15:08:14.0000] <TabAtkins_>
<iframe seamless> is supposed to act like you just transplanted the document's root into the outer document.

[15:08:20.0000] <TabAtkins_>
So just do that.

[15:08:42.0000] <Hixie>
yeah, that's fair enough

[15:08:45.0000] <TabAtkins_>
If the inner document is all positioned, shrug, its root will be 0x0, just like a container element in the same circumstances.

[15:08:45.0000] <Hixie>
ok

[15:10:04.0000] <Velmont>
Hehe. Seems easy enough to relate to.

[15:15:32.0000] <ojan>
Hixie: i agree with TabAtkins_. i'm pretty sure that's roughly what webkit does

[15:16:02.0000] <ojan>
Hixie: although i just reviewed the code. eseidel (on #webkit) wrote the code...he might know better.

[15:16:49.0000] <ojan>
making float do the wrong thing because of the edge case of a document only containing positioned elements would be a shame

[15:17:43.0000] <Hixie>
webkit seems to do weird things

[15:18:29.0000] <Hixie>
e.g. http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1655

[15:18:34.0000] <Hixie>
why scrollbars?

[15:18:36.0000] <Velmont>
film at 11?

[15:19:15.0000] <Hixie>
(with overflow:hidden: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1656 )

[15:19:46.0000] <Hixie>
/me joins a CG that has published some FSA specs to see what the UI looks like when you have published FSA specs

[15:19:57.0000] <Hixie>
(prepare for the linked data community to think i've gone over to their side...)

[15:20:01.0000] <ojan>
Hixie: lol...it's a bug...there's some crazy hackery w scrollbars and seamless iframes

[15:20:20.0000] <Hixie>
ojan: k

[15:20:33.0000] <Hixie>
aw, man, i can't join the damn group without going through my AC rep

[15:20:36.0000] <ojan>
Hixie: i'll file these. hopefully eseidel will fix them. :)

[15:21:03.0000] <TabAtkins_>
Hixie: I know, it's so annoying. >_<

[15:21:10.0000] <ojan>
Hixie: anyways...the 0x0 behavior is what a regular element would do...so i don't see why it's problematic for a seamless iframe to do that

[15:21:26.0000] <Hixie>
ojan: k

[15:21:32.0000] <hober>
Hixie: I'm sure tvr would be thrilled that you've developed an interest in linked data :)

[15:21:44.0000] <Hixie>
yeah but not if i wanted to leave the group 5 minutes later :-)

[15:21:53.0000] <hober>
heh

[15:22:13.0000] <Hixie>
is anyone here either a member of a CG that has published an FSA or an individual member who isn't a member company employee who could do me a favour?

[15:22:38.0000] <JonathanNeal>
Hixie: do I qualify?

[15:22:59.0000] <Hixie>
JonathanNeal: possibly :-)

[15:23:18.0000] <Hixie>
JonathanNeal: i'm trying to get a screenshot of the page that lets you sign on to the FSA deed

[15:23:26.0000] <Hixie>
JonathanNeal: once you're a member of teh group, it should be either on this page: http://www.w3.org/community/json-ld/

[15:23:33.0000] <Hixie>
or this page: http://www.w3.org/community/json-ld/spec/22/commitments

[15:24:13.0000] <Hixie>
oh i guess i could just create a bogus account for testing purposes

[15:25:11.0000] <Hixie>
except i bet w3c has a human in that loop

[15:25:17.0000] <Hixie>
so that wouldn't work either

[15:25:32.0000] <JonathanNeal>
Sorry.

[15:25:51.0000] <JonathanNeal>
I saw the "Join this group" button, but if I can't see the agreement without a human approving something, meh.

[15:26:17.0000] <Hixie>
oh i don't think you need a human to approve joining

[15:26:25.0000] <Hixie>
unless you don't have an account

[15:26:28.0000] <Hixie>
in which case don't worry

[15:27:39.0000] <JonathanNeal>
I definitely have some kind of account with this site, I used it to join a css group.

[16:21:26.0000] <TabAtkins_>
Hixie: For <iframe seamless>, I wonder if it the inner document should use the outer document's information for viewport-related media queries.

[16:22:21.0000] <Velmont>
Opera has something in the speeddails there. Never looked at it though.

[16:25:50.0000] <Hixie>
TabAtkins_: possibly, yeah. file a bug?

[16:26:11.0000] <TabAtkins_>
Related: Does inline SVG create a sub-document with its own viewport size?

[16:26:32.0000] <TabAtkins_>
/me doesn't want to hunt for the relevant details to try and answer the question himself.

[16:38:03.0000] <Hixie>
TabAtkins_: svg always kinda confuses me on this front

[16:38:07.0000] <Hixie>
it used to be very poorly defined

[16:38:13.0000] <Hixie>
now it's defined, but i'm not 100% sure where

[16:51:17.0000] <TabAtkins_>
Ah, cool, dbaron already filed the bug about seamless and MQ.


2012-07-10
[17:10:02.0000] <dbaron>
TabAtkins_, actually, Hixie filed it for me (as I noticed after posting to www-style)

[17:12:23.0000] <TabAtkins_>
Ah, heh.

[17:12:30.0000] <TabAtkins_>
Well, as long as it's filed, I'm happy.

[21:48:30.0000] <zewt>
heh youtube seems to be out to give <video> a boost by switching to hopelessly broken buffering behavior

[21:48:46.0000] <zewt>
now it won't buffer more than 30 seconds ahead, which means i ... can't watch things

[21:48:55.0000] <zewt>
(unless i feel like watching it in 30 second chunks)

[22:00:39.0000] <Hixie>
you can't watch online?

[22:02:09.0000] <zewt>
it won't stream fast enough to watch it

[22:02:34.0000] <zewt>
normally i'd just pause it and let it buffer to full in the background ... but now it stops buffering after the buffer is 30s ahead of the play position, which broke that

[22:04:26.0000] <Hixie>
it doesn't buffer fast enough even if it's playing?

[22:06:51.0000] <zewt>
wow, the html5 player is doing it too (capping the buffer); wonder if that's something they're doing in script (can you even do that?) or if FF is doing the same thing

[22:07:02.0000] <Hixie>
i would assume it's the server

[22:08:08.0000] <zewt>
oh wait, the html5 one kept buffering (eventually)

[22:08:23.0000] <zewt>
(the flash one stops after 30s read-ahead and stays stopped indefinitely)

[22:10:52.0000] <zewt>
(the html5 one working is what I expected--i suppose you could probably jump the hoops to cripple prebuffering like that with <video> if you wanted to badly enough, though)

[23:47:31.0000] <MikeSmith>
has anybody written tests for CSS animations?

[23:48:16.0000] <MikeSmith>
/me finds http://test.csswg.org/source/contributors/microsoft/incoming/CSS3/animations/

[00:48:05.0000] <matjas>
lol @ XHTML 1.0 Strict

[00:49:07.0000] <odinho>
Yes, could do that once every day.

[03:28:04.0000] <SimonSapin>
border-collapse with colspan/rowspan is "funny" ... http://dabblet.com/gist/3082417 (compare webkit and gecko)

[03:29:12.0000] <zcorpan>
firefox 16 will be the unprefixing release, huh?

[03:30:09.0000] <Ms2ger>
Yeah

[03:30:10.0000] <zcorpan>
SimonSapin: opera has a third rendering :-)

[03:30:19.0000] <Ms2ger>
/me looks for other things to unprefix

[03:30:24.0000] <SimonSapin>
zcorpan: yes, I just checked

[03:30:45.0000] <zcorpan>
SimonSapin: what's correct per spec? does it say?

[03:31:01.0000] <SimonSapin>
the spec doesn’t say anything …

[03:31:03.0000] <Ms2ger>
zcorpan, Ha. Ha. Ha.

[03:31:40.0000] <SimonSapin>
I’ll just do whatever is more convenient in WeasyPrint

[03:32:11.0000] <zcorpan>
anyone have ie to check?

[03:32:25.0000] <zcorpan>
i bet it has its own take on it

[03:32:58.0000] <SimonSapin>
I don’t even what to bother reporting this to www-style, nothing’s gonna happen until someone makes css3-tables in a few decades

[03:33:54.0000] <zcorpan>
would be nice to have issues documented somewhere though when that someone wants to take it

[03:34:49.0000] <zcorpan>
maybe a wiki page is better than emails to www-style

[03:36:44.0000] <SimonSapin>
I think that CSSWG has both a wiki and a bugzilla for tracking issues

[03:37:29.0000] <zcorpan>
heycam|away: shouldn't webidl drop the "must have a length" if ES already requires length to be present?

[03:46:27.0000] <Ms2ger>
SimonSapin, and Tracker

[03:47:29.0000] <SimonSapin>
Ms2ger: indeed

[03:48:31.0000] <Ms2ger>
Might be useful to ask MikeSmith to create a bugzilla component for tabkes

[03:48:32.0000] <Ms2ger>
tables*

[03:51:35.0000] <SimonSapin>
Ms2ger: wrote about this to www-style, I’ll leave the rest to the WG

[04:51:49.0000] <MikeSmith>
Ms2ger: CSS tables?

[04:54:11.0000] <MikeSmith>
btw, I added a bunch of CSS features to http://platform.html5.org/

[04:54:24.0000] <MikeSmith>
lots more green flags there now

[05:01:44.0000] <zcorpan>
i love how http://html5please.com/#WebSockets recommends disabling websocket for safari but at the same time using Flash Sockets

[05:02:11.0000] <AryehGregor>
Is there really no Array.remove() or something in JS?

[05:02:45.0000] <AryehGregor>
arr.splice(i, 1) seems a bit abstruse compared to arr.remove(i) or such.

[06:18:47.0000] <kennyluck>
CSS Undefined Module

[06:19:19.0000] <Ms2ger>
Do you mean CSS 2.1?

[06:31:50.0000] <kennyluck>
yeah :)

[07:32:43.0000] <zcorpan>
what's the procedure of publishing a Final Specification?

[07:35:21.0000] <Ms2ger>
Sacrifice a goat

[07:36:07.0000] <zcorpan>
:-(

[07:36:23.0000] <zewt>
declare the spec dead?

[07:37:00.0000] <zcorpan>
i thought it was possible to publish Final Specifications several times

[07:37:31.0000] <zewt>
i thought that was just Final Fantasy

[07:37:38.0000] <Ms2ger>
I think that's the idea

[07:37:53.0000] <Ms2ger>
Maybe AryehGregor knows

[07:38:07.0000] <AryehGregor>
Yes, that's my understanding.

[07:38:15.0000] <AryehGregor>
A Final Specification is a frozen snapshot.

[07:38:22.0000] <AryehGregor>
But a CG can publish however many it wants.

[07:38:34.0000] <AryehGregor>
And/or continue work on the specs after forking them.

[07:38:51.0000] <AryehGregor>
I imagine the procedure for publishing a Final Specification is "Talk to Ian Jacobs, he'll be very excited because no one's done it yet".

[07:39:02.0000] <zcorpan>
should whatcg publish FSs for all the drafts (and one for html)?

[07:39:35.0000] <zcorpan>
i see whatcg doesn't list html under current drafts

[07:39:51.0000] <zcorpan>
but i thought the purpose was to publish html as FS at some point

[07:40:31.0000] <Ms2ger>
The purpose was the patent policy

[07:40:46.0000] <Ms2ger>
And I thought that happened earlier than FS

[07:44:12.0000] <zcorpan>
well we haven't published a draft either so i guess the CLA doesn't apply for html yet

[07:44:48.0000] <zcorpan>
i haven't read the CLA and FSA in detail yet

[07:50:31.0000] <zcorpan>
http://www.w3.org/community/reports/reqs/ copyright boilerplate for FS says to link to https://www.w3.org/community/about/agreements/fsa/ but that redirects to fsa-deed, i guess final/ was intended

[07:52:28.0000] <zcorpan>
also uses https on some links but they seem to redirect to http

[07:54:44.0000] <zcorpan>
/me sends an email

[08:20:38.0000] <zcorpan>
fast ij is fast

[08:33:43.0000] <AryehGregor>
Ms2ger, the patent policy before FS only applies to contributors, I think.  At FS, everyone is invited to agree to license their patents on the subject matter, although it's optional.  (I think.)

[08:33:50.0000] <AryehGregor>
I guess I should publish a FS of the editing spec.

[08:34:05.0000] <AryehGregor>
But if the WHATCG wants to test the waters first, please be my guest.  :)

[10:44:30.0000] <matjas>
annevk: re: HTTP & data URLs & same-origin: http://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html#data:-urls-and-http seems to allow it

[10:44:50.0000] <matjas>
annevk: couldn’t find it in the HTML spec

[10:45:51.0000] <matjas>
annevk: or did I misunderstand that section ^?

[10:50:34.0000] <quinnirill>
hey everyone, I have a question about the XHR2

[10:51:07.0000] <quinnirill>
is Anne here?

[10:51:17.0000] <zewt>
that's an easy question, nope

[10:51:23.0000] <quinnirill>
hehe

[10:51:44.0000] <quinnirill>
ok, I'll try again later then, thanks

[10:51:49.0000] <TabAtkins_>
Anne's on vacation, except when he checks in.

[10:51:52.0000] <quinnirill>
ahh

[10:52:21.0000] <quinnirill>
hmm, I'll just drop this in to boil now then

[10:53:01.0000] <quinnirill>
it looks to me like the current spec doesn't have any support for streaming, am I correct? has this been discussed?

[10:57:19.0000] <quinnirill>
the thing is, we're building these JS audio decoders and *not* streaming would be horrible UX as potential use cases can include 1h+ songs and even for 3m+ songs downloading the whole song before starting playback is usually unacceptable

[10:57:58.0000] <quinnirill>
so for now, we're using partial requests, but that's hogging bandwidth like a... well, hog.

[10:58:26.0000] <quinnirill>
I was wondering if there's a better solution cooking and whether we could maybe contribute ideas

[10:59:10.0000] <zewt>
iirc you can access arraybuffer while loading to get the data so far (never tried it/don't know if it's implemented/might be wrong)

[10:59:25.0000] <zewt>
don't think that lets you actually stream (eg. discard data as you read to read an unbounded stream)

[11:00:40.0000] <quinnirill>
zewt: ok, I'll try that and see if it would be good enough, thanks!

[11:17:45.0000] <Hixie>
i don't suppose we can just drop xslt support from the platform?

[11:18:00.0000] <TabAtkins_>
I'm cool with this.

[11:18:00.0000] <Hixie>
/me stares at hsivonen's scary-looking feedback regarding how readyState should work with xslt

[11:20:44.0000] <Hixie>
oh hey, i already have a section that basically says "i give up on xslt, but in the absence of xslt defining this stuff, here's some handwaving suggestions"

[11:20:49.0000] <Hixie>
/me adds more to it

[12:27:49.0000] <zewt>
gar does google really not index w3 tickets

[12:27:59.0000] <zewt>
site:https://www.w3.org/Bugs/Public/ 1 result

[12:32:07.0000] <TabAtkins_>
view-source:https://www.w3.org/Bugs/Public/robots.txt

[12:32:26.0000] <TabAtkins_>
So, no.

[12:32:46.0000] <zewt>
(obviously, the real question was "why does w3's tracker have a completely BS robots.txt")

[12:33:32.0000] <zewt>
that's beyond absurd ... even for the w3c

[12:38:51.0000] <gsnedders>
jwalden: __proto__ is in Carakan!

[12:44:28.0000] <jwalden>
yay!

[12:46:21.0000] <Hixie>
zewt: bugzilla in general disables robots because of load, iirc

[12:46:49.0000] <Hixie>
also the real robots.txt that matters is https://www.w3.org/robots.txt

[12:46:55.0000] <Velmont>
They should learn it some caching :-)

[12:47:00.0000] <Hixie>
(a robots.txt in a subdirectory doesn't work)

[12:47:06.0000] <Velmont>
Do some varnish magic.

[12:54:37.0000] <jarek>
which spec should I be looking at to get the list of all possible color names?

[12:54:53.0000] <gsnedders>
css3-color

[12:54:55.0000] <jarek>
SVG? CSS2.1? CSS3 Color module?

[12:55:11.0000] <jarek>
http://www.w3.org/TR/css3-iccprof

[12:55:18.0000] <jarek>
but it's from 2001 :/

[12:55:29.0000] <gsnedders>
http://www.w3.org/TR/css3-color/ is a REC from last year

[12:55:38.0000] <jarek>
cool, thanks

[12:55:54.0000] <TabAtkins_>
jarek: If you want to use the list for something, I have a few versions of the list in JSON.

[12:56:49.0000] <jarek>
TabAtkins_: I need a list of all color names that could be used with "fill" SVG property

[12:57:18.0000] <jarek>
TabAtkins_: how many colors are on your list? 147?

[12:57:39.0000] <TabAtkins_>
In browsers, 'fill' takes every single CSS color, including ones that SVG doesn't officially define.

[12:57:55.0000] <TabAtkins_>
All of which are defined in css3-color, that gsnedders linked to.

[13:00:17.0000] <jarek>
TabAtkins_: css3-color seems to contains exactly the same colors as SVG 1.1

[13:00:22.0000] <TabAtkins_>
Yes.

[13:00:34.0000] <TabAtkins_>
The colors haven't changed for a decade.

[13:02:50.0000] <jarek>
there is gray and gray which are basically the same color :/

[13:03:37.0000] <jarek>
I mean gray and grey

[13:05:04.0000] <TabAtkins_>
Yes, they're identical.  The named colors list is *retarded*.

[13:05:56.0000] <jarek>
I assume that decision to include both gray and grey on the list was purely political?

[13:06:07.0000] <Hixie>
probably dates back to X11

[13:09:55.0000] <jarek>
I didn't know that X11 was designed by a committee

[13:10:09.0000] <jarek>
it looks more like addition from W3C

[13:10:14.0000] <gsnedders>
Totally backwards compatible, most of it never used.

[13:10:28.0000] <TabAtkins_>
Not political. The X11 list was copied from an old rgb.txt file that someone had.

[13:10:45.0000] <TabAtkins_>
Someone on that Linux distro decided that that particular list was worthwhile.

[13:10:58.0000] <TabAtkins_>
I respectfully submit that that distro maintainer was stupid.

[13:11:42.0000] <Hixie>
http://en.wikipedia.org/wiki/X11_color_names suggests it's even more complicated than i feared

[13:12:02.0000] <Hixie>
looks like netscape was on the path of blame

[13:15:09.0000] <jarek>
"Some browsers like Netscape navigator insisted on an ‘a’ in any ‘Gray’ except for ‘Light Grey’."

[13:16:13.0000] <gavinc>
I this mess starts in the 19th century :P

[13:16:55.0000] <gavinc>
http://grammarist.com/spelling/gray-grey/

[13:18:04.0000] <gsnedders>
There's 50 Shades of Grey, but only 1 Shade of Gray.

[15:28:36.0000] <hober>
It's that time of year again when I post in some WBS surveys and see if anyone cares enough to fill them out.

[15:28:51.0000] <hober>
Specifically, https://www.w3.org/2002/09/wbs/40318/issue-200-objection-poll/ and https://www.w3.org/2002/09/wbs/40318/issue-189-objection-poll/

[15:37:51.0000] <zewt>
TabAtkins_: WS post might have mixed up quotees a little

[15:40:52.0000] <zewt>
were there two threads about that same topic? could have sworn I replied to this already

[15:43:13.0000] <zewt>
must have been a discard


2012-07-11
[19:01:45.0000] <Hixie>
zewt: actually scripts _can_ detect the conditions you mention for when the network comes back -- they just need to listen for the 'online' event

[19:26:01.0000] <zewt>
Hixie: assuming transient issues (eg. ip changes) ever make it consider itself "offline"

[19:26:55.0000] <zewt>
i really don't understand the notion that pages should have to deal with this individually, when so many other web APIs go to efforts to eliminate uncommon error states and make things just work

[19:33:18.0000] <Hixie>
zewt: it's possible we need to change the design. so far people have just been saying "we should slightly change this api for this one case", not "we need to rethink the basic design of the api"

[19:38:14.0000] <zewt>
simply saying that reconnection behavior, at least for most users, will be done much better if the UA handles it, and that we'll have a lot of buggy and badly-behaved apps if not

[19:38:43.0000] <zewt>
certainly doesn't seem like a huge change, no

[19:39:50.0000] <Hixie>
i don't know how big a change it would be since i haven't examined the implications of a fundamental design change

[19:40:06.0000] <Hixie>
sometimes one can change the basic idea of the feature without actually changing the API at all

[19:40:24.0000] <Hixie>
other times, it simply makes no sense to even have a similar API with a change of design philosophy

[19:43:19.0000] <zewt>
i havn't looked at it in depth either, but roughly i'd expect it to look equivalent to if it still only retried once, but it just took a long time (eg. it'd sit in CONNECTING for a long time)

[19:43:49.0000] <zewt>
(don't really know if it'd need to expose more than that)

[19:46:13.0000] <zewt>
(i should have said I don't expect it to be a big script-visible change)

[20:00:13.0000] <Hixie>
zewt: i don't know, but e.g. maybe the protocol needs other features, maybe there needs to be attributes exposing or controlling it, i dunno

[20:02:20.0000] <zewt>
yep

[02:09:24.0000] <annevk>
matjas: the intention is to allow it

[02:09:41.0000] <annevk>
matjas: not sure anything defines it as being same-origin though

[02:09:51.0000] <annevk>
matjas: but UAs are starting to implement it that way

[03:12:24.0000] <annevk>
matjas: hmm yeah so we need to make that more explicit somehow

[03:13:25.0000] <matjas>
yes please. if it wasn’t for the non-normative “note” i wouldn’t have even mentioned it

[03:14:30.0000] <matjas>
annevk: want me to file a CORS spec bug?

[03:14:44.0000] <annevk>
it's not a bug in CORS

[03:14:58.0000] <annevk>
it's an XHR bug

[03:15:08.0000] <matjas>
it doesn’t just apply to XHR though

[03:15:19.0000] <matjas>
data URL images should be same-origin too

[03:15:33.0000] <matjas>
or is that case already covered elsewhere?

[03:15:36.0000] <annevk>
"If the XMLHttpRequest origin and the request URL are same origin" needs to expand to also include request URLs whose scheme is data:

[03:15:59.0000] <annevk>
matjas: that is covered elsewhere at the moment

[03:16:56.0000] <annevk>
matjas: maybe we can flip things around and only special case data URLs where allowing them to be same-origin would be harmful, but that would require working through all the same-origin checks in the platform

[03:17:14.0000] <annevk>
matjas: it's not okay for instance for shared workers

[03:18:44.0000] <annevk>
somehow blob URLs are defined to be same origin with the script that created them

[03:18:56.0000] <annevk>
that prolly works

[03:19:45.0000] <annevk>
but that cannot work for data URLs

[03:22:56.0000] <annevk>
matjas: I can fix the XHR bug easily enough sometime later, but maybe we should look at data URLs again in various contexts to see if updating the same-origin definition and special casing shared workers would be enough

[03:30:09.0000] <annevk>
hmm so XHR -> cross-origin -> data URL would still need to be cross-origin

[03:31:58.0000] <annevk>
so yeah, it would kind of depend on where we special case it how such scenarios would work

[03:32:10.0000] <annevk>
e.g. with my proposal XHR -> same-origin -> data URL would fail

[03:43:32.0000] <cheron>
What is the intention of the "challenge" attribute of the HTML 5 keygen element? Are there any use cases/examples?

[03:46:33.0000] <annevk>
cheron: it influences the generated signature

[03:46:47.0000] <annevk>
cheron: the use case is mainly supporting some legacy sites that use <keygen>

[03:47:15.0000] <annevk>
cheron: so maybe we should move the element to the "obsolete" section

[03:50:35.0000] <cheron>
annevk: ty, this helps me much with my bachelor's thesis

[03:51:01.0000] <annevk>
what's the subject of the thesis?

[03:51:31.0000] <cheron>
Security in HTML5

[03:51:52.0000] <smaug____>
uh, we should definitely move keygen to obsolete

[03:52:13.0000] <annevk>
cheron: cool

[03:52:19.0000] <annevk>
smaug____: file a bug ;)

[03:53:35.0000] <smaug____>
whaat, keygen has got some new properties in HTML spec

[03:59:20.0000] <cheron>
smaug__: if you file a bug can you send me the link, pls

[04:01:17.0000] <cheron>
annevk: yeah, pretty interesting

[04:48:28.0000] <zcorpan>
/me sends an email to site-comments about robots.txt

[04:51:44.0000] <MikeSmith>
zcorpan: what's up with robots.txt?

[04:52:15.0000] <zcorpan>
MikeSmith: http://lists.w3.org/Archives/Public/site-comments/2012Jul/0007.html

[04:52:58.0000] <MikeSmith>
I can fix the dvcs.w3.org one myself

[04:53:04.0000] <MikeSmith>
what should it have?

[04:53:20.0000] <MikeSmith>
the www.w3.org one I'll leave up to others

[04:54:25.0000] <zcorpan>
MikeSmith: is don't see any reason to not use `User-agent: *\nAllow: *` but maybe there is

[04:55:24.0000] <MikeSmith>
zcorpan: I know they have had serious problems in the past with bots crawling dev.w3.org

[04:55:35.0000] <MikeSmith>
creating massive load

[04:56:33.0000] <zcorpan>
bratell just had issues finding the HTML Editing APIs spec because of this, and i found out yesterday that the quirks mode spec isn't visible in google

[04:56:48.0000] <MikeSmith>
yeah that's not good

[04:56:57.0000] <MikeSmith>
I will talk to the systems team about it

[04:57:01.0000] <zcorpan>
thanks!

[04:57:54.0000] <zcorpan>
the /Bugs/ thing was someone complaining here in #whatwg the other day

[04:59:09.0000] <MikeSmith>
OK

[08:59:00.0000] <zcorpan_>
http://simon.html5.org/sandbox/bookmarklets/w3c-cg-spec-checker.js

[09:17:24.0000] <ashemedai>
Yay, I get to annoy the Opera guys again

[09:19:12.0000] <zcorpan>
matjas: i have filed a spec bug about CORS-enabled fetch to consider data: to be same-origin

[09:20:03.0000] <zcorpan>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17461

[11:29:45.0000] <smaug____>
rniwa: values from dictionaries are read when the object is passed to some native method

[11:29:55.0000] <smaug____>
so I don't quite understand your email about the problem

[11:30:13.0000] <rniwa>
smaug____: well, you can then modify those properties

[11:30:20.0000] <smaug____>
so?

[11:30:30.0000] <rniwa>
smaug____: so you can do undoManager.transact({execute: function () { this.undo = ~; }})

[11:30:50.0000] <rniwa>
smaug____: so my point is that it's not a problem specific to having DOMTransaction interface

[11:30:57.0000] <smaug____>
this.undo = wouldn't do anything sane

[11:31:06.0000] <rniwa>
smaug____: what do you mean?

[11:31:15.0000] <rniwa>
smaug____: it'll do something extremely sane

[11:31:17.0000] <smaug____>
I mean, it would put .undo property to 'this'

[11:31:28.0000] <smaug____>
this wouldn't be the dictionary object

[11:31:32.0000] <rniwa>
smaug____: yeah but "this" is the dictionary object!

[11:32:08.0000] <rniwa>
smaug____: since execute is supposed to be called on the dictionary object passed into.

[11:32:22.0000] <smaug____>
no

[11:32:24.0000] <rniwa>
transact

[11:32:31.0000] <rniwa>
smaug____: why no?

[11:32:32.0000] <smaug____>
execute is a property of the dictionary

[11:32:41.0000] <smaug____>
and stored separately

[11:32:42.0000] <rniwa>
smaug____: yes.

[11:32:44.0000] <rniwa>
smaug____: no.

[11:32:45.0000] <smaug____>
when passed to native

[11:32:50.0000] <rniwa>
smaug____: that's not how it works

[11:33:05.0000] <rniwa>
maybe my spec isn't clear.

[11:33:13.0000] <rniwa>
smaug____: but the intent is for the dictionary to be kept alive

[11:33:25.0000] <rniwa>
smaug____: and execute, undo, redo, & executeAutomatic to be called upon on the dictionary object

[11:33:39.0000] <smaug____>
then you should use some nointerfacecobject interface

[11:33:49.0000] <smaug____>
not dictionary

[11:33:50.0000] <rniwa>
smaug____: I'm using that.

[11:33:55.0000] <rniwa>
smaug____: it's DOMTransaction

[11:34:00.0000] <rniwa>
smaug____: with [NoInterface]

[11:34:07.0000] <smaug____>
ok, let's not talk about dictionaries then :)

[11:34:33.0000] <Ms2ger>
Eh, NoInterfaceObject, you shouldn't use that

[11:34:37.0000] <rniwa>
smaug____: yeah, that's the confusing part.

[11:34:38.0000] <smaug____>
ojan_away was talking about dictionaries in his email

[11:35:06.0000] <smaug____>
so you're probably talking about different things... and I'm not sure what exactly :)

[11:35:23.0000] <rniwa>
smaug____, Ms2ger: http://dvcs.w3.org/hg/undomanager/raw-file/tip/undomanager.html#dom-domtransaction

[11:35:42.0000] <rniwa>
well... it works like Dictionary to some extent

[11:35:51.0000] <rniwa>
at least in ES5, you can create one like Dictionary

[11:35:59.0000] <rniwa>
but yeah... it's not really the same thing

[11:36:10.0000] <rniwa>
the fact we're talking about this in ES5 is making everything super confusing

[11:36:20.0000] <smaug____>
Dictionary is pretty well-defined in WebIDL

[11:36:36.0000] <rniwa>
smaug____: yes, but i don't think ojan_away was talking about that.

[11:36:52.0000] <smaug____>
I thought he was talking about dictionaries

[11:36:59.0000] <rniwa>
smaug____: he and everyone else are referring to dictionary was in ordinary programming language term.

[11:37:08.0000] <rniwa>
s/ was/ as/

[11:37:18.0000] <smaug____>
huh

[11:37:30.0000] <Ms2ger>
Interfaces aren't implemented by content scripts

[11:37:38.0000] <rniwa>
Ms2ger: they could be.

[11:37:43.0000] <Ms2ger>
No

[11:37:51.0000] <rniwa>
Ms2ger: but this is EXACTLY why I want to change the interface

[11:37:55.0000] <rniwa>
because it's so damn foreign to the Web

[11:38:01.0000] <smaug____>
when talking about WebIDL interfaces 'dictionary' certainly refers to the concept in  WebIDL

[11:38:42.0000] <Ms2ger>
So, what is this DOMTransaction thing?

[11:38:43.0000] <rniwa>
Ms2ger: http://www.w3.org/TR/WebIDL/#es-user-objects

[11:39:05.0000] <rniwa>
Ms2ger: it's a user object implementing a callback interface

[11:39:25.0000] <rniwa>
but we had never used this semantics in any other API

[11:39:26.0000] <Ms2ger>
What do you use it for?

[11:39:38.0000] <rniwa>
Ms2ger: you use it to represent an undo "step"

[11:39:49.0000] <rniwa>
Ms2ger: a unit of work that can be done, undone, or redone via undo manager

[11:40:07.0000] <rniwa>
Ms2ger: it represents a "command" in the command design pattern

[11:42:17.0000] <smaug____>
I don't buy the "foreign to the Web". { handleEvent: function(evt) {} } isn't foreign  to the Web

[11:42:36.0000] <smaug____>
the fact the the object may have some more properties does change anything

[11:42:50.0000] <smaug____>
s/does/doesn't/

[11:44:29.0000] <rniwa>
smaug____: i don't think most authors are aware of the fact an event handler is an object though

[11:44:47.0000] <Ms2ger>
So, you pass it to transact()?

[11:44:49.0000] <rniwa>
smaug____: i literally had never seen anyone using that explicit object form

[11:44:56.0000] <rniwa>
Ms2ger: right.

[11:45:05.0000] <Ms2ger>
Why doesn't the IDL for transact say that?

[11:46:35.0000] <Ms2ger>
Looks like a dictionary would work

[11:46:48.0000] <rniwa>
Ms2ger: yeah... i should update the spec :\

[11:47:21.0000] <smaug____>
dictionary doesn't work if rniwa wants the 'this' handling he mentioned above

[11:47:29.0000] <Ms2ger>
Well, no

[11:48:06.0000] <rniwa>
smaug____, Ms2ger: that "this" handling is pretty important.

[11:48:21.0000] <rniwa>
otherwise the benefit of using dictionary degrades significantly

[11:50:00.0000] <smaug____>
so don't use dictionary but callback interface :)

[11:50:06.0000] <rniwa>
smaug____, Ms2ger: so are you guys opposed to yuval's proposal?

[11:50:55.0000] <Ms2ger>
Link?

[11:51:23.0000] <smaug____>
I'm not sure I understand what is the "final" proposal

[11:51:51.0000] <matjas>
zcorpan: thanks for the link

[11:52:16.0000] <Ms2ger>
Anyway, write it up, and we'll complain then :)

[11:52:21.0000] <rniwa>
http://lists.w3.org/Archives/Public/public-webapps/2012JulSep/0058.html

[11:52:53.0000] <smaug____>
that hybrid form is really odd

[11:54:55.0000] <smaug____>
rniwa: just curious, did this all start from the memory management problems, or were you thinking about changing the API even before someone started to implement the API ?

[11:55:30.0000] <rniwa>
smaug____: i had never liked the API because it seemed very different from other APIs

[11:55:38.0000] <rniwa>
smaug____: and my intern had been wokring on this patch

[11:56:06.0000] <rniwa>
and memory management thing motivated me enough to start the thread

[11:56:48.0000] <smaug____>
/me can't wait to see cycle collector to be implemented for webkit ;)

[11:58:05.0000] <Ms2ger>
Let's try workers first

[11:58:40.0000] <rniwa>
smaug____: that oddness solves one important weridness

[11:58:50.0000] <rniwa>
smaug____: which comes with "native" DOMTransactions

[11:58:54.0000] <rniwa>
smaug____: editing commands add

[11:59:21.0000] <Ms2ger>
(Hah, XSLT3)

[12:00:39.0000] <rniwa>
Ms2ger: how about some sane NodeList & HTMLCollection

[12:00:44.0000] <smaug____>
rniwa: well, it is odd to have some sorts of callbacks and event listeners

[12:00:53.0000] <Ms2ger>
rniwa, what about NodeList?

[12:00:54.0000] <rniwa>
smaug____: no... there will be no callbacks

[12:01:06.0000] <rniwa>
Ms2ger: WebKit has namedItem or whatever on NodeList

[12:01:11.0000] <smaug____>
new DOMTransaction(function () { ~ })

[12:01:17.0000] <Ms2ger>
Kill it

[12:01:24.0000] <smaug____>
what is that function if not a callback?

[12:01:25.0000] <rniwa>
smaug____: on but that won't be added as a property

[12:01:35.0000] <rniwa>
smaug____: it'll be hidden within the object.

[12:01:44.0000] <smaug____>
so?

[12:01:49.0000] <smaug____>
it is still a callback

[12:01:54.0000] <rniwa>
smaug____: yeah... :\

[12:02:07.0000] <rniwa>
smaug____: so should we just do

[12:02:14.0000] <rniwa>
t = new DOMTransaction

[12:02:22.0000] <rniwa>
t.onexecute = function () {~}

[12:02:29.0000] <rniwa>
t.onundo = function () {~}

[12:02:34.0000] <rniwa>
t.onredo = function () {~}

[12:02:35.0000] <rniwa>
?

[12:02:53.0000] <smaug____>
that one looks better

[12:06:31.0000] <smaug____>
(but I'm sure the jQuery fans will complain ;) )

[12:08:19.0000] <rniwa>
smaug____: well, people who like jQuery can wait for jQuery to support it

[12:08:22.0000] <Ms2ger>
t.onexecute = function () { ~; return this; }

[14:00:50.0000] <Ms2ger>
https://twitter.com/dastels/status/223059799699501056

[14:05:57.0000] <jsbell>
Not sure if praise or damning praise.

[15:24:24.0000] <ojan>
a bit late to the party...but, for the record, I was talking about WebIDL dictionaries

[15:29:04.0000] <Hixie>
so what's the browser interest on the http+aes stuff?

[15:33:01.0000] <espadrine_>
isn't that what https is supposed to do? What's http+aes?

[15:34:14.0000] <Yuhong>
<Hixie> looks like netscape was on the path of blame

[15:34:29.0000] <Yuhong>
I think IE1 was the first browser to support named colors.

[15:35:13.0000] <Hixie>
espadrine_: a way to have content served to a client from an untrusted CDN

[15:37:37.0000] <espadrine_>
some kind of CORS with a brand new protocol then?

[15:38:12.0000] <Hixie>
no, new scheme that includes a key to decrypt the payload of an HTTP resource

[15:38:15.0000] <Hixie>
it's in the spec currently

[15:38:20.0000] <Hixie>
but i'm wondering if we should drop it

[15:38:25.0000] <Hixie>
since it doesn't seem anyone really cares

[15:38:51.0000] <Hixie>
we added it because people said there was a use case for it, but i'm starting to think the use case was a fiction intended to help push a DRM scheme

[15:41:01.0000] <espadrine_>
oh, I see

[15:41:53.0000] <smaug____>
rniwa: it still sounds like ojan is talking about WebIDL dictionaries and you're talking about callback objects :)

[15:42:50.0000] <ojan>
what are callback objects?

[15:43:03.0000] <smaug____>
{ handleEvent: function() {} }

[15:43:12.0000] <smaug____>
and you store reference to that object

[15:43:18.0000] <rniwa>
smaug____: yeah so it seems.

[15:43:49.0000] <ojan>
callback objects seem like a bad idea

[15:44:07.0000] <smaug____>
callback objects are a nice idea :)

[15:44:35.0000] <smaug____>
but in this particular case dictionary should work well too

[15:44:53.0000] <ojan>
it simplifies many things in this case

[15:45:06.0000] <ojan>
e.g. what happens when you reuse the object for two transact calls

[15:46:01.0000] <ojan>
i'm about at the end of my standards arguing rope on this discussion though...and i don't think there's much more to say...so, unless some new information comes to light...i don't have more to add to that thread

[15:47:00.0000] <smaug____>
I don't care too much, except that the proposal to use callbacks in some cases and event listeners in some other cases looked really really odd.

[15:47:14.0000] <smaug____>
either event listeners, or dictionary or callback object

[15:47:51.0000] <zewt>
smaug____: that seems like a pointless idea, heh

[15:48:13.0000] <zewt>
very Java-ish (or more generally, any language that doesn't have first-class function objects-ish)

[15:48:25.0000] <smaug____>
from implementation point of view callback object is easiest in Gecko, but other options are easy too

[15:48:56.0000] <zewt>
it's just completely alien to JS as a callback paradigm

[15:49:31.0000] <zewt>
i mean, i don't know why a *user* would ever want to use that method

[15:50:07.0000] <smaug____>
(yet { handleEvent: function() {}} is commonly used, but let's not start this discussion again. It is getting late here :) )

[15:50:26.0000] <zewt>
never once seen it in real-world code :)

[15:51:17.0000] <smaug____>
I'm fine with dictionary or event listener based approaches

[15:51:34.0000] <smaug____>
zewt: you haven't looked at Firefox UI code ;)

[15:51:58.0000] <zewt>
i have a bit; i try to avoid it at any cost :)

[15:51:59.0000] <smaug____>
or, IIRC, OSX Dashboard

[15:52:37.0000] <rniwa>
ojan: i replied on the thread but there were use cases for being able to keep the original object

[15:52:42.0000] <rniwa>
ojan: in particular properties on the original object.

[15:53:39.0000] <smaug____>
( that is the use case for { handleEvent: function() {}} too. One can store state in the object easily. )

[15:54:06.0000] <zewt>
(not really a use case for that, since you can easily pass in a bound function or a closure)

[15:54:11.0000] <zewt>
(not specific to that, I mean)

[15:54:53.0000] <smaug____>
"you can easily" doesn't mean the other approach is bad

[15:55:21.0000] <zewt>
it is if the object is meant to be a simple callback (as opposed to, say, a *set* of callbacks, in which case it's fine)

[15:55:50.0000] <smaug____>
zewt: in this case, UndoManager, we're talking about a set of callbacks

[15:55:52.0000] <rniwa>
zewt: in undo manager's case it'll be 1-3 callbacks so it'll be odd.

[15:56:04.0000] <rniwa>
zewt: in fact, you can't just pass in an object

[15:56:12.0000] <rniwa>
ugh.. i mean pass in a function

[15:56:17.0000] <zewt>
then OK (that's very different from the weird handleEvent thing on event listeners)

[15:56:24.0000] <rniwa>
you need to explicitly pass in an object

[15:56:31.0000] <rniwa>
zewt: yeah.

[15:56:51.0000] <rniwa>
I DO like the syntax of pure Dictionary or user object implementing DOMTransaction

[15:57:06.0000] <rniwa>
the fact we can't have any readonly property, etc...

[15:57:14.0000] <rniwa>
make things too hary

[15:57:17.0000] <rniwa>
hairly*

[15:57:19.0000] <smaug____>
zewt: it is not at all different from event listeners. you can pass { handleEvent: function() {}, undo: function() {}, redo: function() {}} to UndoManager and to addEventListener

[15:58:11.0000] <zewt>
it's very different from event listeners, because event listeners are a single simple callback which can (just for compatibility reasons) also be passed in as an object

[15:58:50.0000] <zewt>
event listeners aren't a set of functions, they're just a function

[15:59:37.0000] <smaug____>
eh, but you can pass the *same* object implementing several callbacks

[15:59:42.0000] <smaug____>
that is the point

[15:59:55.0000] <smaug____>
no need to play with bind() or anything

[16:01:00.0000] <zewt>
i don't know what you're saying; I can think of nothing you can do with addEventListener by passing in an object that you can't do easily with a function

[16:01:40.0000] <zewt>
(not sure what you mean by "several callbacks"; event handlers have only one callback, handleEvent)

[16:01:43.0000] <smaug____>
{ handleEvent: function() {}, undo: function() {}, redo: function() {}}

[16:02:10.0000] <smaug____>
you pass the same object to addEventListener, and also to UndoManager

[16:02:19.0000] <zewt>
that's pointless and unnecessary

[16:02:22.0000] <smaug____>
you can keep some state in the object

[16:02:35.0000] <smaug____>
this isn't useful discussion :)

[16:03:06.0000] <zewt>
if you say so :)

[16:03:43.0000] <smaug____>
if the argument is all the time something like "that's pointless and unnecessary" :p

[16:04:16.0000] <zewt>
it lets you do nothing that you can't do easily and cleanly without it; that's how I describe that

[16:05:23.0000] <rniwa>
anyways, i'm inclined to take the flame and use the event listener approach

[16:05:27.0000] <zewt>
anyway, we're stuck with it for event handlers, but I hope I won't be the only one to object if someone suggests supporting it for other simple callbacks :)

[16:05:43.0000] <rniwa>
i'm sad about the verbosity but it seems better than the user object approach.

[16:06:07.0000] <rniwa>
zewt: object to what?

[16:07:08.0000] <zewt>
rniwa: that a function taking a single, simple callback should accept either a function or an object with some method, like event handlers

[16:07:29.0000] <zewt>
(not what you're doing)

[16:07:48.0000] <rniwa>
zewt: oh yeah... that's a very odd design indeed.

[16:08:37.0000] <zewt>
rniwa: as long as you expose event attributes, it's not *too* much more verbose (compared to having to addEventListener each of them)

[16:09:01.0000] <rniwa>
zewt: oh yeah, definitely not addEventListener

[16:09:27.0000] <smaug____>
well, there will be also addEventListener

[16:09:48.0000] <rniwa>
zewt, smaug____: or maybe I should just give up and stick with the original design.

[16:09:54.0000] <zewt>
yeah, if you're an EventTarget then you get addEventListener automatically

[16:10:39.0000] <smaug____>
rniwa: original was callback object, right?

[16:10:58.0000] <smaug____>
I'm fine with that, and I'm fine with dictionary, and with event listeners

[16:11:04.0000] <smaug____>
but just don't mix  them :)

[16:11:18.0000] <smaug____>
I kind of like the callback object approach

[16:11:28.0000] <rniwa>
smaug____: yeah.

[16:11:49.0000] <rniwa>
smaug____: sicking really liked the callback object approach.

[16:12:14.0000] <zewt>
hmm

[16:13:27.0000] <zewt>
we don't have any way to feature detect dictionary-based APIs, do we

[16:13:49.0000] <zewt>
akin to func.length for simple functions

[16:14:06.0000] <zewt>
(in cases where you can't just call it and see what happens, which isn't always workable)

[16:14:45.0000] <zewt>
thinking of createObjectURL(blob, {autoRevoke...})

[16:16:03.0000] <zewt>
aside from https://www.w3.org/Bugs/Public/show_bug.cgi?id=16953#c4, which is a bit insane

[16:16:52.0000] <rniwa>
zewt: no :(

[16:17:08.0000] <zewt>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17746 heh people copying and pasting a mail from me into a bug is ... confusing

[16:17:11.0000] <rniwa>
zewt: another thing about user object is that you can't do expandos

[16:17:12.0000] <zewt>
"havn't I read this before?"

[16:17:31.0000] <zewt>
rniwa: i think the dictionary-feature-detection thing should be solved generically (don't know how though)

[16:17:48.0000] <rniwa>
zewt: how do you do that?

[16:17:58.0000] <zewt>
i said i don't know :P

[16:17:59.0000] <rniwa>
zewt: also, what is there to feature-detect?

[16:18:11.0000] <zewt>
whether an argument supports a particular options argument

[16:18:22.0000] <zewt>
eg. whether createObjectURL supports autoRevoke

[16:18:32.0000] <zewt>
rniwa: which case can't you do expandos?

[16:18:52.0000] <zewt>
if it's a dictionary then the API doesn't really store the object itself anyway, as I understand it

[16:19:44.0000] <zewt>
(WebIDL basically makes a shallow copy)

[16:20:57.0000] <rniwa>
zewt: dictionary.

[16:21:21.0000] <rniwa>
zewt: i guess i meant the user object case then.

[16:21:36.0000] <rniwa>
zewt: in either case, you can't really add new functionality to dom transactions.

[16:21:41.0000] <rniwa>
unless you manually modify each object.

[16:22:26.0000] <zewt>
yeah, you don't get a reference back to the original object (since the API never even sees it)

[16:22:45.0000] <zewt>
it's not quite like passing around handler objects like JS libraries like to do

[16:23:10.0000] <zewt>
it doesn't really lose you anything compared to event handlers, either, though

[16:24:33.0000] <rniwa>
zewt: well in the event handler case, the object uses DOMTransaction interface

[16:24:42.0000] <zewt>
i don't have any strong opinion on "return an EventTarget and the user attaches listeners to it" vs. "pass in an object with a set of callbacks"; there's really very little difference

[16:24:43.0000] <rniwa>
zewt: so you can feature detect on DOMTransaction interface if any

[16:25:14.0000] <rniwa>
zewt: one big difference is that you can add event listeners to the former after the fact

[16:25:25.0000] <rniwa>
zewt: and observe undo/redo per dom transaction

[16:25:40.0000] <rniwa>
zewt: wheres for the latter, we need to do that at element/document level


2012-07-12
[23:51:54.0000] <AryehGregor>
"It almost seems to me that nobody cares over here what the W3C document actually says, as there is that other "more helpful" version. In which case I wonder why it's published at all?"

[23:52:02.0000] <AryehGregor>
. . . I guess it's probably best not to respond to that.

[23:53:14.0000] <AryehGregor>
/me cries when he reads stuff like http://lists.w3.org/Archives/Public/www-style/2012Jul/0267.html that adds ever more declarative special cases instead of writing imperatively

[23:54:11.0000] <annevk>
you mean the calc() feature itself?

[23:54:29.0000] <AryehGregor>
I mean basically all of CSS.

[23:55:09.0000] <AryehGregor>
Which currently consists of lots of declarative requirements scattered across a zillion specs with all interactions undefined unless someone specifically points them out.

[23:55:13.0000] <annevk>
I saw a somewhat plausible answer to why adding imperative stuff to CSS would be problematic

[23:55:22.0000] <annevk>
ooh

[23:55:27.0000] <annevk>
you mean spec-wise

[23:55:33.0000] <AryehGregor>
Right.

[23:55:48.0000] <annevk>
yeah dunno about that

[23:55:50.0000] <AryehGregor>
For that matter, a lot of the terminology is undefined too -- specs will refer to things like "the element's border box" when elements can have any number of border boxes from zero and up.

[23:56:03.0000] <annevk>
uhuh

[23:56:07.0000] <annevk>
it's kind of crazy really

[23:56:11.0000] <AryehGregor>
Clearly it's harder to write imperatively than DOM stuff, because JavaScript is imperative and CSS is declarative.

[23:56:21.0000] <AryehGregor>
But still.

[23:56:25.0000] <annevk>
it's hard to write because CSS is hard

[23:56:41.0000] <annevk>
so instead of defining CSS they describe it which is a little less hard

[23:56:48.0000] <annevk>
but then you run into problems soon enough

[23:57:20.0000] <AryehGregor>
I don't think CSS is inherently any harder than, say, the HTML parser, or editing.

[23:57:31.0000] <AryehGregor>
It's a lot harder than DOM Core, yes, but that's not saying much.

[23:58:04.0000] <annevk>
dunno, CSS is big

[23:58:20.0000] <AryehGregor>
Bigger than HTML?

[23:58:29.0000] <AryehGregor>
One person did that.  :)

[23:58:43.0000] <annevk>
I think CSS is bigger and definitely more complex

[23:59:01.0000] <annevk>
but I have hopes of someone fixing it still :)

[23:59:03.0000] <AryehGregor>
Really?  The WHATWG spec is a lot bigger than all CSS specs combined.

[23:59:12.0000] <AryehGregor>
Although that's partly because it actually specifies all details, obviously.

[23:59:37.0000] <annevk>
all CSS specs?

[23:59:50.0000] <annevk>
including the object model and such?

[00:00:01.0000] <AryehGregor>
I'm pretty sure.

[00:00:11.0000] <AryehGregor>
I think I once counted using wget and wc -l.

[00:00:32.0000] <AryehGregor>
Retrieving all the CSS specs from the Current Work page.

[00:00:51.0000] <AryehGregor>
CSS2.1 is the only one of appreciable size.  (Guess what: co-edited by Hixie.)

[00:00:57.0000] <annevk>
CSS 2.1 alone is half of HTML

[00:01:13.0000] <AryehGregor>
Really?

[00:01:14.0000] <annevk>
I reckon if you rewrite that proper style it will be bigger

[00:01:26.0000] <annevk>
the PDF is 487 pages

[00:02:01.0000] <zcorpan>
AryehGregor: annevk: did you see http://krijnhoetmer.nl/irc-logs/whatwg/20120711#l-533 ?

[00:02:05.0000] <annevk>
and CSS 2.1 does not include the object model and such

[00:02:38.0000] <AryehGregor>
zcorpan, no, what does it do?

[00:03:56.0000] <zcorpan>
AryehGregor: checks http://www.w3.org/community/reports/reqs/

[00:04:02.0000] <annevk>
zcorpan: that's kind of neat, maybe get ij to do something with it?

[00:07:30.0000] <zcorpan>
i'm not sure a bookmarklet is the best way to implement it, maybe something that can be run from the command line is better so editors can run it at the same time as anolis

[00:08:44.0000] <zcorpan>
i checked yesterday and it finds some errors in Fullscreen and HTML Editing APIs

[00:08:51.0000] <annevk>
ideally it's some kind of service that just emails you the violations

[00:09:11.0000] <zcorpan>
or files a bug maybe?

[00:09:40.0000] <annevk>
yeah

[00:09:49.0000] <annevk>
would be cool for the markup validator too

[00:10:04.0000] <zcorpan>
hsivonen: ^

[00:10:25.0000] <MikeSmith>
what would be cool for the validator?

[00:10:45.0000] <zcorpan>
to email or file bug when your page doesn't validate anymore

[00:11:53.0000] <MikeSmith>
oh

[00:13:05.0000] <zcorpan>
but i guess that's possible to implement with v.nu's existing api

[00:13:09.0000] <zcorpan>
with a simple script

[00:13:51.0000] <zcorpan>
just file-upload to v.nu after running anolis and if it gives errors, file a bug

[00:35:54.0000] <hsivonen>
zcorpan: you want the validator to file bugs in the W3C bugzilla?

[00:36:30.0000] <zcorpan>
hsivonen: yeah. except i realized that's possible already without more code in v.nu

[00:37:43.0000] <hsivonen>
I see

[01:07:33.0000] <annevk>
MikeSmith: zcorpan: if you want to allow indexing of W3C Bugzilla https://www.w3.org/robots.txt will need to be modified (note https)

[01:08:09.0000] <MikeSmith>
ah yeah

[01:38:21.0000] <hsivonen>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17693 someone missed the memo on <ins> and <del> (which became synonyms for <u> and <s> since it's really hard to do what Word does)

[02:04:48.0000] <smaug____>
Uh, http://dev.w3.org/2006/webapi/FileAPI/#dfn-autoRevoke :/

[02:05:15.0000] <smaug____>
no arun here

[02:32:52.0000] <zcorpan>
smaug____: file a bug?

[02:33:22.0000] <smaug____>
perhaps I should discuss with arun first

[02:33:52.0000] <smaug____>
though, I could file a bug to remove "stable state" concept from HTML :)

[02:35:54.0000] <zcorpan>
what's wrong with stable state?

[02:39:02.0000] <smaug____>
zcorpan: it is a weird concept

[02:39:12.0000] <smaug____>
it is not really detectable from JS

[02:39:47.0000] <zcorpan>
what do you suggest instead for algorithms that currently use it?

[02:39:55.0000] <zcorpan>
queue a task?

[02:40:42.0000] <smaug____>
that is often a possible choice

[02:40:55.0000] <smaug____>
There is also microtask

[02:41:21.0000] <smaug____>
microtask was added because stable state and end-of-task are so vague

[02:41:35.0000] <zcorpan>
have a link to microtask?

[02:42:17.0000] <smaug____>
http://www.whatwg.org/specs/web-apps/current-work/#perform-a-microtask-checkpoint

[02:42:37.0000] <smaug____>
for some odd reason the spec talks only about mutation observers

[02:44:24.0000] <zcorpan>
thanks

[02:55:34.0000] <annevk>
is that revoking language even correct?

[02:56:12.0000] <annevk>
you only want it to be revoked after you have used the URL somehow

[02:56:56.0000] <annevk>
that file API spec is turning into a mess

[02:59:18.0000] <smaug____>
why should it say anything about using the URL ?

[03:03:00.0000] <annevk>
if you just create it, then wait a while before using it (waiting on some event) it would already be revoked

[03:03:05.0000] <annevk>
that seems kind of pointless

[03:05:16.0000] <zcorpan>
yeah the autorevoke thing only allows the url to be referenced in the same script. and many things that fetch urls are themselves async, like e.g. <img src>

[03:15:00.0000] <zcorpan>
should websocket binary messages allow scripts to read the message in a streaming fashion while it's being received, like with xhr?

[03:15:24.0000] <zcorpan>
right now the message event is only fired once the whole message has been received, afaict

[03:17:00.0000] <annevk>
pretty much nothing in XHR is streaming fwiw

[03:17:06.0000] <annevk>
only responseText at the moment

[03:17:16.0000] <annevk>
maybe also responseType = "text"

[03:35:58.0000] <zcorpan>
annevk: can't you get the partial data as blob or arraybuffer?

[03:36:41.0000] <annevk>
zcorpan: I don't really see how

[03:36:57.0000] <annevk>
we thought of allowing that at some point

[03:37:13.0000] <annevk>
with "I don't really see how" I mean that the current spec doesn't do it

[03:37:27.0000] <annevk>
see http://dvcs.w3.org/hg/xhr/raw-file/tip/Overview.html#the-response-attribute

[03:38:37.0000] <zcorpan>
ah. i thought it wasn't null while LOADING

[03:40:34.0000] <zcorpan>
wouldn't it work to return a blob early and have it act like a file that loads slowly?

[03:41:27.0000] <zcorpan>
or maybe something different makes more sense for streaming use cases

[03:49:52.0000] <annevk>
yeah that would work

[03:49:59.0000] <annevk>
I think implementors had some concerns

[03:50:12.0000] <annevk>
but we can still make that work in the future

[03:52:06.0000] <annevk>
time to see the outside world

[04:06:20.0000] <Ms2ger>
Pff

[04:06:26.0000] <Ms2ger>
The outside world is overrated

[04:23:40.0000] <hsivonen>
hmm. does the new GWT ingest IDL from Gecko and WebKit or really C++?

[04:23:56.0000] <hsivonen>
if it really eats C++, that would be a failure for IDL

[04:25:31.0000] <smaug____>
hsivonen: link?

[04:27:23.0000] <hsivonen>
smaug____: http://www.youtube.com/watch?v=VOf27ez_Hvg&feature=plcp

[04:27:44.0000] <hsivonen>
(not really a time-efficient reference)

[06:52:52.0000] <zcorpan_>
isn't it awesome when people talk in XML tags on mailing lists and suggest to replace HTML+CSS with XML+XSL?

[08:02:09.0000] <matjas>
hsivonen: re: 404 page on <audio> testbed — http://twitter.com/yvg/status/223346356088094721

[08:11:16.0000] <miketaylr>
matjas: pfffffffffft

[08:15:02.0000] <smaug____>
zewt: an event listener can't know if something else runs before the next stable state

[08:15:14.0000] <smaug____>
(just an example)

[08:15:28.0000] <smaug____>
also, the same task may for example dispatch several events

[09:33:27.0000] <hsivonen>
matjas: thanks

[09:58:59.0000] <gsnedders>
/me contemplates starting a new cross-browser JS testsuite, again

[10:02:15.0000] <Ms2ger>
gsnedders, at least join us at the W3C :)

[10:25:08.0000] <sedovsek>
Hey, Is there any alternative for document.webkitHidden on iOs and Android devices?

[10:25:20.0000] <Hixie>
what does it do?

[10:30:50.0000] <sedovsek>
https://developers.google.com/chrome/whitepapers/pagevisibility

[10:30:55.0000] <sedovsek>
Here are some examples.

[11:13:27.0000] <matjas>
sedovsek (if you read the logs): http://mths.be/visibility demonstrates how you can fall back to focus/blur

[11:16:09.0000] <bencc>
are svg animations standard in html5?

[11:17:41.0000] <shepazu>
bencc: technically, because SVG animations (SMIL) are part of SVG 1.1, which is the spec that HTML5 references, they are "standard"… but IE doesn't support them, and HTML does strictly require full support for SVG 1.1

[11:19:27.0000] <bencc>
shepazu: it did support it in the past I think

[11:19:40.0000] <shepazu>
bencc: nope

[11:19:40.0000] <bencc>
shepazu: do you know if there are plans to support SMIL in IE10 or later?

[11:19:49.0000] <shepazu>
no plans for IE10

[11:20:13.0000] <bencc>
shepazu: that's sad :(

[11:20:30.0000] <shepazu>
I agree

[11:20:43.0000] <shepazu>
they are focusing on CSS animations, which you should be able to use with SVG

[11:20:59.0000] <shepazu>
I think that's all still being worked out, though

[11:21:09.0000] <bencc>
shepazu: ok

[11:21:23.0000] <bencc>
shepazu: can I do feature detection for beginElement or animate?

[11:21:42.0000] <shepazu>
bencc: you should think about using FakeSMIL

[11:22:06.0000] <shepazu>
it's a script lib shim that handles most of SMIL in SVG for IE

[11:22:22.0000] <bencc>
cool. this? http://fakesmil.blogspot.co.il/

[11:22:27.0000] <Hixie>
wow, not counting the responsive images threads and not counting e-mails about adding new features, i'm down to just about 250 e-mails!

[11:22:47.0000] <shepazu>
yeah

[11:22:52.0000] <Hixie>
(including those two categories, 1200 e-mails.)

[11:23:09.0000] <Hixie>
(also including bugs: 1700 e-mails.)

[11:23:51.0000] <shepazu>
bencc: it's a bit old, but it still works

[11:24:12.0000] <shepazu>
it's not perfect, but it does the job for most things

[11:26:20.0000] <bencc>
 shepazu: I've just tried it and I still get the error

[11:26:38.0000] <shepazu>
bencc: what error?

[11:26:43.0000] <bencc>
 shepazu: I'm adding the animate element dynamically so I'm not sure if the script cover this use case

[11:26:50.0000] <shepazu>
it should

[11:26:52.0000] <bencc>
Object doesn't support property or method 'beginElement'

[11:28:05.0000] <shepazu>
bencc: yeah, it's not going to add that support if you're doing feature detection

[11:28:33.0000] <shepazu>
bencc: you have a jsfiddle?

[11:29:10.0000] <bencc>
shepazu: not yet http://dpaste.com/769928/

[11:29:30.0000] <bencc>
shepazu: what do you mean? I'm not doing feature detection

[11:29:35.0000] <shepazu>
ok

[11:29:37.0000] <bencc>
shepazu:  I just included the script

[11:30:00.0000] <shepazu>
bencc: that snippet doesn't include fakesmil...

[11:30:35.0000] <bencc>
shepazu: I thought it should just work

[11:31:05.0000] <shepazu>
bencc: I can't debug if I don't have an example that should work...

[11:31:30.0000] <bencc>
shepazu: sure, I'll make one

[11:38:06.0000] <bencc>
shepazu:  this is without smil http://jsfiddle.net/DgMDV/1/

[11:38:14.0000] <bencc>
shepazu: works in FF but not in IE9

[11:38:25.0000] <bencc>
shepazu: now trying to add smil but I'm not sure how to use it

[11:38:54.0000] <bencc>
I mean fakesmil

[11:39:32.0000] <shepazu>
http://leunen.me/fakesmile/faq.html#dev

[11:43:12.0000] <bencc>
shepazu: I've added the script to the jsfiddle but there must be something more I need to do

[11:43:51.0000] <shepazu>
bencc: let me make a simple example

[11:43:57.0000] <shepazu>
it might take me a few minutes

[11:44:14.0000] <bencc>
thanks!

[11:48:41.0000] <Hixie>
hmm, i wonder if there's a way to make bugzilla search results more computer-readable

[11:53:51.0000] <bencc>
shepazu: looks like there is a more recent version here: http://bazaar.launchpad.net/~smilteam/smil/MAIN/view/head:/smil.user.js

[12:00:05.0000] <shepazu>
bencc: I just realized I can't test it… I'm on a mac, and I don't have a browser that doesn't support SMIL :(

[12:00:37.0000] <bencc>
shepazu: I can

[12:00:48.0000] <bencc>
the example will be useful in any case

[12:07:24.0000] <bencc>
shepazu: I'm trying to add it with <svg ...><script type="text/ecmascript" xlink:href="smil.user.js"/>...</svg> but I'm not sure if it is loaded

[12:09:01.0000] <MikeSmith>
bencc: you serving it with an XML mime type?

[12:10:43.0000] <bencc>
MikeSmith: no

[12:11:02.0000] <bencc>
MikeSmith: I don't know how to use it

[12:12:36.0000] <MikeSmith>
bencc: I think what you're trying to do won't work unless you serve it with an XML mime type

[12:12:45.0000] <MikeSmith>
but I dunno really

[12:12:57.0000] <MikeSmith>
shepazu knows better than me

[12:13:02.0000] <bencc>
ok

[12:17:19.0000] <shepazu>
bencc: my suggestion is to google uses of fakesmil, and find an example that works for you

[12:18:49.0000] <bencc>
shepazu: I tried but can't find how to dynamically include it

[13:21:29.0000] <MikeSmith>
hmm JSON-LD

[15:18:42.0000] <loincloth>
what what

[16:06:59.0000] <zewt>
cool, "new" paypal page broke the login form in some weird way

[16:08:08.0000] <zewt>
haha and they use a position: absolute block hovering on top of the input form instead of @placeholder, and don't bother turning off pointer-events, so when i click the input, it randomly selects the placeholder instead of the input

[16:08:22.0000] <zewt>
it's always amazing when the highest-profile pages on the internet screw up so badly

[16:09:44.0000] <TabAtkins_>
yay!

[16:10:40.0000] <hober>
s/amazing/sadly unsurprising/

[16:10:55.0000] <zewt>
i was trying to be a *little* generous

[16:11:06.0000] <zewt>
okay I guess I wan't :)

[16:11:08.0000] <zewt>
also wasn't

[16:13:13.0000] <zewt>
(somehow having a simple login form break autocompletion twice and then select the word "address" three times makes me feel ungenerous)


2012-07-13
[21:28:05.0000] <kodkod>
irc is so oldschool

[21:28:27.0000] <zewt>
yeah, it's way too functional for today's internet

[21:29:58.0000] <kodkod>
heh there were times when i use to be an ircop on some servers :)

[22:15:20.0000] <zcorpan>
Hixie: turns out "soon" didn't match reality eh? i really don't like this. can fixed bugs be marked in some other way that is possible to search/filter for?

[22:42:29.0000] <zcorpan>
role attribute?

[22:42:34.0000] <zcorpan>
/me blinks

[22:43:29.0000] <zcorpan>
i thought the PFWG had taken the attribute and defined it completely in the ARIA spec *years* ago, since the role spec was useless

[01:05:45.0000] <annevk>
zcorpan: it's that time when everything old is new again

[01:06:11.0000] <annevk>
see also: HTML WG

[01:21:47.0000] <annevk>
yay zewt for following up on the File API mess

[01:22:52.0000] <annevk>
zewt: maybe it can be some kind of post-URL-parser steps that do this

[01:23:10.0000] <annevk>
zewt: that way not all specs have to special case Blob URLs

[01:43:40.0000] <sedovsek>
Anyone knows whether Page Visibility works on iOS/Androids (hint: it does not :/)? Are there any workarounds?

[01:43:50.0000] <sedovsek>
http://www.w3.org/TR/2011/WD-page-visibility-20110602/

[01:53:03.0000] <annevk>
okay, I'm no longer stalking people in Mozilla or WebKit Bugzilla

[01:53:19.0000] <annevk>
this means that if you cc me there's a somewhat bigger chance I'll take a look

[01:58:15.0000] <Ms2ger>
Did I spam you too much? :)

[02:23:55.0000] <annevk>
I had 80k unread Gecko threads and 30k unread WebKit

[02:24:04.0000] <annevk>
time to give up the experiment :)

[02:24:23.0000] <annevk>
now marking everything as read and see how it goes

[02:28:01.0000] <annevk>
When is Gmail going to remove the "Invite a friend" box?

[02:38:30.0000] <sedovsek>
@annevk : )

[02:38:49.0000] <sedovsek>
I've seen you're giving a talk at the fronteers conf.

[02:39:08.0000] <sedovsek>
Looking forward already.

[02:40:56.0000] <annevk>
ah yeah, Fronteers will be fun :)

[02:41:18.0000] <sedovsek>
Yea… go to Amsterdam they said.

[02:41:25.0000] <sedovsek>
It will be fun they said.

[02:55:54.0000] <annevk>
no worries, our http://en.wikipedia.org/wiki/Levee have been safe since the fifties

[02:56:20.0000] <annevk>
(too many words for one thing in English)

[03:02:34.0000] <SimonSapin>
funny how English uses a French word where French has a different word (digue)

[03:02:35.0000] <annevk>
English has that too

[03:02:36.0000] <annevk>
dike/dyke

[03:02:36.0000] <annevk>
dijk in Dutch

[03:11:50.0000] <Lachy>
be careful of using the word dike in English. It has an unfortunate slang meaning.

[03:13:47.0000] <beverloo>
Don't you mean dyke?

[03:34:42.0000] <hsivonen>
almost every English word seems to have slang meanings that make you uneasy about using the word if you look it up on urbandictionary :-(

[03:34:57.0000] <hsivonen>
(not to suggest disagreement with Lachy's warning)

[03:37:47.0000] <annevk>
http://blogs.msdn.com/b/ie/archive/2012/07/12/ie10-user-agent-string-update.aspx cross-platform :/

[03:38:40.0000] <hsivonen>
looks like they are bold enough to break the Desktop/Phone/Tablet taxonomy and instead go with a touch / no touch

[03:39:00.0000] <hsivonen>
unless, of course, they later ship IE10 for phones with yet another UA string anyway

[03:39:23.0000] <hsivonen>
in which case they could have said Tablet instead of touch for compat with Firefox and Opera

[03:39:31.0000] <annevk>
it doesn't really seem compatible with Microsoft's platform vision of being able to interact with everything in whatever way you want either

[03:40:20.0000] <annevk>
if the UA advertises touch and I have a keyboard hooked up there's a fair chance I'm going to be in trouble, which seems kind of contrary to recent Microsoft vision announcements

[03:40:34.0000] <hsivonen>
fwiw, I expect the strategy of putting touch screens on desktops to flop for ergonomic and aesthetic reasons

[03:40:48.0000] <Ms2ger>
annevk, maybe you should get yourself hired by MS :)

[03:41:00.0000] <annevk>
hsivonen: well yeah, me too

[03:41:12.0000] <annevk>
hsivonen: just pointing out some inconsistency here :)

[03:42:21.0000] <hsivonen>
annevk: well, Mozilla's platform engineers supposedly prefer capability sniffing, but we still cater to non-top-tier Web authors who want to assume a Desktop/Tablet/Phone taxonomy and UA sniff for it

[03:43:05.0000] <annevk>
seems bad for fingerprinting, but I guess it's exposed in other ways already

[03:43:28.0000] <annevk>
(though it does make it easier)

[03:43:56.0000] <hsivonen>
I should note that Opera's local-Presto products do it, too

[03:44:26.0000] <hsivonen>
In which light I find it odd that Opera Mini doesn't expose tabletness in the UA string

[03:44:57.0000] <hsivonen>
even though Mini spews out fingerprinting surface without any concern by exposing the exact device model

[03:44:59.0000] <annevk>
I have attempted to fix various bits of our UA string, but somehow missed out when we started screwing it up again :/

[03:46:00.0000] <hsivonen>
I was trying to advocate that Mozilla do it exactly like Opera (with "Mobi" and "Tablet"), but we ended up doing it with "Mobile" and "Tablet"

[03:46:32.0000] <hsivonen>
which still allows clueful people handle Firefox and Opera form-factor sniffing using the same code

[03:47:27.0000] <hsivonen>
not sure how much of a lost cause it is to design UA strings to be nicely sniffable by clueful sniffers

[03:48:47.0000] <annevk>
for a moment there I was hopeful the IE blog used <mark>

[03:48:50.0000] <annevk>
turns out it's <span style="background-color: yellow;">Touch</span>

[03:49:15.0000] <hsivonen>
<mark> would have made it so much more Semantic

[03:49:43.0000] <annevk>
:)

[03:52:26.0000] <annevk>
the web needs a new Mark Pilgrim for some level-headed commentary

[03:52:47.0000] <annevk>
there's @mattur but he's limited to 140 characters

[03:53:52.0000] <niloy>
the IE blog is a sad piece of software, no rounded corners, no gradients, looks like 2006

[03:54:08.0000] <hsivonen>
the AC should be outraged that the W3C's resources have gone into publishing an XHTML2 module, but chances are no one even gets a slap on the wrist for it

[03:55:30.0000] <annevk>
W3C doesn't know what it's doing; film at eleven

[03:58:42.0000] <hsivonen>
it'll be interesting to see what will be put forward as the two interoperable implementations

[03:59:07.0000] <hsivonen>
my guess is: two XML editors that don't barf on a DTD that declares an attribute called role

[03:59:42.0000] <Ms2ger>
My guess is that they won't bother and nobody will complain

[04:00:34.0000] <hsivonen>
Björn might

[04:00:48.0000] <hsivonen>
/me looks forward to another spiderman episode

[04:20:44.0000] <hsivonen>
hmm. looks like the email where I proposed Mozilla stop prefixing APIs got Warnocked

[04:21:04.0000] <hsivonen>
well, on the positive side, while I was away, various CSS things were unprefixed in Gecko

[04:22:46.0000] <smaug____>
it is not clear to me when to prefix an API an when not

[04:23:11.0000] <smaug____>
btw, IDB got unprefix

[04:23:13.0000] <smaug____>
ed

[04:23:16.0000] <hsivonen>
I think we should not prefix

[04:23:22.0000] <hsivonen>
yay for IDB unprefixing

[04:23:39.0000] <smaug____>
hsivonen: even when implementing highly unstable APIs?

[04:23:46.0000] <hsivonen>
and I think we shouldn't ship to release channel stuff that we don't expect to be able to live with forever

[04:23:57.0000] <hsivonen>
smaug____: I think we shouldn't ship those

[04:24:04.0000] <smaug____>
that is possibly true

[04:24:08.0000] <hsivonen>
smaug____: or if we ship, we shouldn't pretend they are unstable

[04:24:41.0000] <hsivonen>
smaug____: I'd expect people who write apps for the first release of Firefox OS to be unhappy if we treated their apps as having been written on top of unstable APIs

[04:24:59.0000] <smaug____>
but the problem is that we get even less feedback from web devs, if APIs aren't in a release

[04:25:33.0000] <smaug____>
getting feedback from web devs is always difficult

[04:25:36.0000] <hsivonen>
smaug____: we get so little feedback even now that I think the delta isn't worth the trouble prefixing causes

[04:25:54.0000] <smaug____>
could be

[04:26:09.0000] <hsivonen>
I guess I should go write a couple of patches now

[04:28:55.0000] <hsivonen>
speaking of feedback from Web devs, it would be interesting to count the bugzilla items we've got from various Web devs who position themselves as standards advocates

[04:39:15.0000] <hsivonen>
/me finds out about http://annevankesteren.nl/2012/07/leaving-opera whoa

[04:39:32.0000] <Ms2ger>
Did you hear about chaals? :)

[04:42:51.0000] <hsivonen>
Ms2ger: I heard a claim that he left Opera

[04:43:09.0000] <hsivonen>
Ms2ger: I haven't heard if it was true or where he went yet

[04:43:19.0000] <Ms2ger>
It's true, and Yandex

[04:45:10.0000] <Ms2ger>
/me is fascinated by the addendum at http://annevankesteren.nl/2012/07/interview being older than the post otself

[04:45:12.0000] <Ms2ger>
*itself

[05:11:41.0000] <karlcow>
hsivonen: Chaals left Opera for Yandex.

[05:14:19.0000] <hsivonen>
interesting destination

[05:19:43.0000] <karlcow>
Ms2ger: the fascinating thing is that the home page says now, and the article page says July 13th

[05:20:03.0000] <karlcow>
hsivonen: interesting in which way?

[05:21:41.0000] <hsivonen>
karlcow: in the way that Yandex is not known as a browser vendor and (AFAICT) not as a particularly active participant at the W3C

[05:23:46.0000] <karlcow>
hsivonen: there is a life (and businesses) outside of browser vendors ;). For Yandex and W3C, I guess they have indirect influence by their huge market share in Russia.

[05:26:30.0000] <karlcow>
indeed participation is subpar for now, at least for people using yandex emails

[05:26:31.0000] <karlcow>
http://www.w3.org/Search/Mail/Public/search?keywords=&hdr-1-name=from&hdr-1-query=yandex-team.ru&index-grp=Public_FULL&index-type=g&type-index=

[05:27:35.0000] <karlcow>
ah there is also

[05:27:36.0000] <karlcow>
http://www.w3.org/Search/Mail/Public/search?keywords=&hdr-1-name=from&hdr-1-query=yandex.ru&index-grp=Public_FULL&index-type=g&type-index=

[05:28:01.0000] <karlcow>
maybe generic emails yandex.ru

[05:29:41.0000] <zcorpan>
chaals⊙oc Results : 2117

[05:32:08.0000] <karlcow>
Sidar → RMIT → W3C → Opera → Yandex

[05:59:30.0000] <zcorpan>
TabAtkins_: does http://lists.w3.org/Archives/Public/www-style/2012Apr/0321.html "a00000" still apply? afaict it's an identifier and would be converted. please file a bug if the spec's wrong

[06:00:10.0000] <zcorpan>
/me goes offline now for a few weeks

[06:11:06.0000] <kodkod>
hello

[06:16:35.0000] <hsivonen>
sigh. Source Maps have a format version identifier

[06:27:59.0000] <hsivonen>
Do we have a canned reference for explaining why versioning is an anti-pattern for Web formats?

[06:46:13.0000] <kennyluck>
Well Marat Tanalin is pretty active in every mailing list.

[06:48:03.0000] <Ms2ger>
Is that Marat Tanalin | Advertising?

[06:48:51.0000] <kennyluck>
huh

[06:49:10.0000] <hsivonen>
kennyluck: does he/she work for Yandex or just use their webmail?

[06:49:31.0000] <hsivonen>
as in: people with @yahoo.com addresses don't necessarily work for Y!

[06:49:37.0000] <hsivonen>
either

[06:49:43.0000] <kennyluck>
hsivonen, good question. I don't know.

[06:50:32.0000] <hsivonen>
/me wonders how long until )]} prepended to JSON becomes an official part of JSON

[06:56:54.0000] <SimonSapin>
hsivonen: prepended how?

[06:57:28.0000] <hsivonen>
SimonSapin: Characters ")]}" at the start of the file before normal JSON content

[06:58:00.0000] <hsivonen>
SimonSapin: in order to make sure the file is *not* a valid JS program

[06:58:38.0000] <Lachy>
hsivonen, isn't the whole point of JSON supposed to be that it is syntactically valid JavaScript?

[06:59:06.0000] <Lachy>
which sites prepend such characters to their JSON?

[06:59:51.0000] <hsivonen>
Lachy: well, JSON works independent of eval nowadays (and also in non-JS languages)

[07:00:34.0000] <Lachy>
yeah, I know. But JSON.parse() doesn't support prepending )]}, does it?

[07:00:48.0000] <hsivonen>
Lachy: the point of )]} is to prevent cross-side data theft by overriding JS Array and then loading the file as a script

[07:00:54.0000] <SimonSapin>
hsivonen: is this for security reasons as described in http://flask.pocoo.org/docs/security/#json-security ?

[07:01:06.0000] <hsivonen>
Lachy: it doesn't support it to my knowledge, but I haven't tested

[07:01:16.0000] <hsivonen>
Lachy: presumably Google uses this for source maps

[07:01:28.0000] <hsivonen>
Lachy: the source maps spec explicitly extends JSON like this

[07:01:49.0000] <hsivonen>
SimonSapin: yes

[07:02:43.0000] <hsivonen>
the source maps spec could use some Hixie-style spec writing. I will have to send some feedback.

[07:02:53.0000] <Lachy>
I don't know what source maps is.

[07:04:23.0000] <Lachy>
found this. Reading. http://www.html5rocks.com/en/tutorials/developertools/sourcemaps/

[07:04:57.0000] <hsivonen>
yeah, that

[07:20:30.0000] <Lachy>
so I'm now looking into rewriting the about: URI spec properly (since the IETF has now taken it in the wrong direction).

[07:21:20.0000] <Lachy>
I need to figure out a term to use, instead of "reserved about URI", which means any URI defined by a specification for use in a defined context, but where it doesn't suggest any kind of exclusivity of that URI for that spec.

[07:26:15.0000] <Lachy>
I'm also defining it such that all about: URIs other than about:blank are non-dereferencable, outside of application-specific uses, which avoids any possible conflict of two specs using the same about: URI, and thus removing any need for a normative registry..

[08:02:21.0000] <matjas>
TIL about the “octal escapes in regular expression literals” compatibility requirement for JavaScript: http://mathias.html5.org/specs/javascript/#octal-escapes-in-regular-expression-literals

[08:15:55.0000] <Hixie>
Lachy: how on earth did the ietf manage to screw up something as simple as about: ???

[08:16:19.0000] <Hixie>
zcorpan: i'm literally working on a fix to the bug issue today (spent most of yesterday on it too)

[08:16:38.0000] <Hixie>
zcorpan: (writing a script that will clone the bugs)

[08:17:20.0000] <Lachy>
well, they define things that just don't match reality. percent-encoding is not supported, for example, and yet is defined in the spec. I would have changed it, but the feedback was handled around the time someone else took over editing.

[08:18:21.0000] <Lachy>
they also introduced a silly registry that gives exclusivity of a given about: URI to the first spec that defines it, rather than realising that independent uses of the same token in different specs doesn't conflict in any case.

[08:18:31.0000] <Lachy>
they also now fail to define how about:blank is handled.

[08:19:57.0000] <Ms2ger>
In their defense, nobody knows how about:blank needs to be handled

[08:20:03.0000] <Lachy>
so I think about: URIs, at least for use in specifications, are to be handled as opaque identifiers, rather than real URLs.

[08:20:50.0000] <Lachy>
Ms2ger, it needs to be handled as an opaque identifier not subject to URL normalisation, and behaves as though it returns an empty UTF-8 encoded HTML document.

[08:22:13.0000] <Lachy>
<iframe src="about:blan%6B"> is not the same as <iframe src="about:blank"> in any implementation.

[08:22:39.0000] <Ms2ger>
But clearly it *should* be!

[08:24:29.0000] <Lachy>
that depends on the reasons for browsers not applying any normalisation. Boris sent feedback before to the IETF list about this, and the potential for security implications for changing it.

[08:26:10.0000] <Lachy>
Ms2ger, also, despite the about: spec being available and saying that for a long time now, not one browser has changed. So I'd rather have a spec that documents reality.

[08:26:11.0000] <hsivonen>
when in doubt, let bz take precedence over the IETF

[08:26:48.0000] <Ms2ger>
s/the IETF/everyone else/, pretty much

[09:06:07.0000] <Hixie>
Lachy: good times

[09:33:49.0000] <hober>
Lachy: TabAtkins_ apparently managed to register about:invalid with the iana folks

[09:46:45.0000] <TabAtkins_>
Yup, it was pretty easy.

[10:05:03.0000] <Lachy>
hober, yes, I know. I saw that.

[10:23:35.0000] <TabAtkins_>
zcorpan: Assuming you read logs when you return for a few weeks, looks like I was wrong.  a00000 will indeed be handled correctly by your quirk.

[10:29:10.0000] <annevk>
hober: TabAtkins_: fwiw, CSS should use about:blank, just like every other place where we lack a base URL

[10:30:08.0000] <annevk>
Lachy: Ms2ger: %-normalization does not appear to happen in the browser, so supporting just "about:blank" should be pretty trivial

[10:30:47.0000] <annevk>
Lachy: Ms2ger: especially once the URL specification is somewhat more in shape, though not sure how much of that I'll do during vacation, other than a bit of planning

[10:31:03.0000] <TabAtkins_>
annevk: No can do. That's a valid HTML page, which won't do what we want if we, say, de-magicify iframes into CSS.

[10:31:35.0000] <annevk>
it's not valid and I'm not sure what you mean

[10:31:49.0000] <TabAtkins_>
Hm? about:blank is a perfectly valid page.

[10:32:04.0000] <annevk>
not validator valid

[10:32:15.0000] <TabAtkins_>
Who cares about that?

[10:32:39.0000] <annevk>
dunno, but I'm not sure what you mean

[10:32:44.0000] <scott_gonzalez>
Is there a way to prevent a browser from displaying options for <input list> or would you need to remove the list attribute?

[10:32:49.0000] <TabAtkins_>
I mean that it's a resolveable URL that returns an HTML page.

[10:33:03.0000] <annevk>
scott_gonzalez: you'd need to remove the attribute

[10:33:05.0000] <Lachy>
annevk, right, which is why I think percent normalisation should be abandoned. It might be easier to just say that when used by a specification in a specific context, they are non-dereferencable magic tokens that look like URLs, but which aren't really. Except for about:blank, no other about: URI actually needs to refer to any specific resource.

[10:33:20.0000] <annevk>
TabAtkins_: what you mean with de-magicify iframes

[10:33:32.0000] <TabAtkins_>
The point of the default value for attr(foo url) is that it shouldn't every be dereferenceable, no matter what context it appears in.

[10:33:49.0000] <TabAtkins_>
annevk: Like, "content: frame(http://foo);"

[10:34:06.0000] <scott_gonzalez>
Is there anything documenting how to search against the options or is it up to the UA?

[10:34:11.0000] <annevk>
TabAtkins_: why not?

[10:34:32.0000] <annevk>
scott_gonzalez: up to the UA

[10:34:32.0000] <TabAtkins_>
Because then it's not a very good default value?  Also: symmetry.

[10:34:48.0000] <annevk>
TabAtkins_: symmetry would be following the rest of the platform

[10:34:55.0000] <scott_gonzalez>
annevk: Do you think in makes sense for @list to be ignored for @autocomplete=off?

[10:35:09.0000] <annevk>
scott_gonzalez: no, autocomplete is orthogonal

[10:35:09.0000] <Hixie>
what's about:invalid?

[10:35:24.0000] <Lachy>
about:invalid in CSS Values spec

[10:35:26.0000] <annevk>
Hixie: same as about:blank but without the content-type

[10:35:31.0000] <TabAtkins_>
Hixie: I just got it registered.  Guaranteed non-dereferencable url.

[10:35:31.0000] <Hixie>
ah

[10:35:41.0000] <scott_gonzalez>
annevk: Really? @list seems to me like a way to build a specific set of autocomplete options.

[10:35:59.0000] <Hixie>
in other news, i am about to start testing my script for cloning htmlwg bugs into whatwg bugs (the script to do the opposite is next, but the needs are different so i'm doing them in two different steps)

[10:36:04.0000] <annevk>
scott_gonzalez: autocomplete relates to UA-autocomplete, not developer-autocomplete

[10:36:33.0000] <annevk>
Hixie: is the opposite needed?

[10:36:48.0000] <Hixie>
annevk: well i assume the htmlwg doesn't want to lose all the bugs, i mean, the bugs apply to their spec too

[10:37:04.0000] <Hixie>
MikeSmith: i'll let you know when i'm ready to just do a batch job, but for now since i'm doing these one bug at a time i think it's ok to leave the bugmail turned on

[10:37:25.0000] <annevk>
yeah I guess, don't really see the point in having both specs if they're not edited by the same person

[10:37:25.0000] <Hixie>
incidentally, i am calling this "operation convergence".

[10:37:38.0000] <Hixie>
annevk: no argument from me there, but that's up to them

[10:38:08.0000] <annevk>
Hixie: sounds like a name straight from Arrested Development

[10:38:12.0000] <Lachy>
TabAtkins_, just curious what you think about the %-encoding issue. Should about:invali%64 work in CSS, or would you rather it be just an opaque identifier that must match "about:invalid" to be valid?

[10:38:23.0000] <TabAtkins_>
Lachy: I have absolutely no opinion on this.

[10:38:26.0000] <Lachy>
ok

[10:38:44.0000] <annevk>
Lachy: I think everything after about: should just be treated as a literal

[10:38:56.0000] <annevk>
Lachy: apart from normalization that is done by the URL parser

[10:39:07.0000] <annevk>
Lachy: e.g. turning " " into "%20" and such

[10:39:10.0000] <Hixie>
ok.

[10:39:15.0000] <Hixie>
/me pushes the button to run the first bug through

[10:40:01.0000] <annevk>
Tabatkins: it's still not really clear to me why about:blank is not okay

[10:40:21.0000] <annevk>
Tabatkins: it's used e.g. for Document objects created by createDocument or createHTMLDocument

[10:40:31.0000] <annevk>
it's used by <iframe>s

[10:40:33.0000] <annevk>
etc.

[10:40:34.0000] <Tabatkins>
Those are trying to create HTML documents, yes.

[10:40:39.0000] <Hixie>
ok. this is what my script does for bugs that don't need cloning but don't apply to whatwg bugs: https://www.w3.org/Bugs/Public/show_bug.cgi?id=14028

[10:40:40.0000] <annevk>
no they're not

[10:40:42.0000] <Hixie>
that looks ok

[10:40:45.0000] <Hixie>
/me pushes the button again

[10:42:00.0000] <Hixie>
this is what bugs look like when they are cloned: https://www.w3.org/Bugs/Public/show_bug.cgi?id=17769

[10:43:01.0000] <Hixie>
/me makes the lines between comments longer for aesthetic reasons

[10:44:10.0000] <annevk>
does not look too bad

[10:44:32.0000] <Tabatkins>
annevk: The problem that I alluded to with my "symmetry" comment is that we have mechanisms in CSS now that have special behavior for "invalid" URLs, that don't dereference to a resource of the expected type (the image() function).

[10:44:37.0000] <annevk>
but it's going to be confusing when people add comments all over

[10:45:07.0000] <Hixie>
annevk: i'm very open to suggestions for making it better

[10:45:21.0000] <Tabatkins>
annevk: It's reasonable to assume that, in the future, if we have properties that expect to take a URL pointing to an HTML page, that we might have similar functionality.

[10:45:45.0000] <Tabatkins>
annevk: So, for symmetry, might as well claim a URL up-front that will work for all media types.

[10:45:50.0000] <annevk>
Tabatkins: why would it ever point to about:blank directly?

[10:46:06.0000] <annevk>
Tabatkins: could you give an example?

[10:46:13.0000] <Tabatkins>
Because of attr(foo url), when the element doesn't have a "foo" attribute.

[10:46:47.0000] <annevk>
Tabatkins: ooh, why wouldn't you fallback to nothing?

[10:46:56.0000] <Tabatkins>
What is "nothing"?

[10:47:28.0000] <annevk>
Tabatkins: is this mainly about the computed value?

[10:47:36.0000] <Tabatkins>
Yeah.

[10:48:12.0000] <annevk>
Tabatkins: wouldn't the computed value be "foo"?

[10:48:14.0000] <Hixie>
hm, bug in my script; it didn't properly update the previous bug

[10:48:16.0000] <Hixie>
why not

[10:49:00.0000] <Tabatkins>
annevk: No, that would be weird.  It would mean that attr(foo url) computes to either a string or a url, depending on whether the attribute exists.

[10:49:05.0000] <Tabatkins>
We can't do useful type-checking with that.

[10:49:41.0000] <Tabatkins>
And computing to url(foo) is no good either, because that's potentially a valid resource.

[10:51:15.0000] <annevk>
thanks for explaining

[10:51:26.0000] <Tabatkins>
It makes sense now?

[10:51:54.0000] <annevk>
not sure whether it was worth a new URL, but the explanation makes sense

[10:52:11.0000] <Tabatkins>
kk, good enough.

[10:58:20.0000] <Hixie>
/me fixes bug and pushes button again

[10:59:55.0000] <Hixie>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=10824 became https://www.w3.org/Bugs/Public/show_bug.cgi?id=17770

[11:00:03.0000] <Hixie>
looks like that worked, reassignment of the original and everything

[11:05:42.0000] <Hixie>
ok last test: https://www.w3.org/Bugs/Public/show_bug.cgi?id=11004 cloned to https://www.w3.org/Bugs/Public/show_bug.cgi?id=17771

[11:05:50.0000] <Hixie>
that one shows some of my attempts to be clever in the cloning

[11:05:53.0000] <Hixie>
(compare the comments)

[11:06:07.0000] <Hixie>
ok. this script seems to work. let's work on the script for the other direction now.

[11:07:08.0000] <Ms2ger>
Killing the "EDITOR'S RESPONSE" boilerplate and the "mass-moved component" comments?

[11:07:36.0000] <Hixie>
and stripping quotations from comments for the form ">lots of quotations from immediately prior comment\ncomment"

[11:08:49.0000] <Ms2ger>
What's the other direction for?

[11:11:08.0000] <Hixie>
the bugs that are in the 'other hixie drafts' component because they were filed from the whatwg side, but that apply to the htmlwg deliverables

[11:14:25.0000] <Ms2ger>
Is there anyone who's going to look at them?

[11:16:12.0000] <Hixie>
how do you mean?

[11:16:48.0000] <Ms2ger>
At the bugs in the htmlwg components

[11:17:23.0000] <Hixie>
the chairs have said that they are going to find an editor who is going to do what i was doing but with an eye to going to REC, so i presume so

[11:17:39.0000] <Hixie>
it's what their process requires, no?

[11:18:07.0000] <Hixie>
it would be pretty disastrous to the htmlwg spec if they just stopped fixing the problems with it, not to mention being a violation of both the wg's process and the w3c's process

[11:20:15.0000] <Ms2ger>
Well...

[11:20:16.0000] <Ms2ger>
Yes

[11:20:48.0000] <Hixie>
anyway, that's their problem

[11:20:54.0000] <Ms2ger>
Otoh, the W3C just published an XHTML2 module

[11:21:25.0000] <Hixie>
seriously?

[11:22:03.0000] <Ms2ger>
http://www.w3.org/TR/2012/CR-role-attribute-20120712/

[11:22:32.0000] <Hixie>
...

[11:23:42.0000] <annevk>
well also the HTML Media Capture draft

[11:23:49.0000] <annevk>
which is really just a new attribute for <input>

[11:24:52.0000] <annevk>
Tabatkins: I think in the specific case where you want to represent a network error via a URL and not have the URL for base URL purposes something like about:invalid is probably okay

[11:25:09.0000] <annevk>
Tabatkins: it seems somewhat likely there's a precedent for that somewhere, but maybe not

[11:25:12.0000] <Tabatkins>
Yeah, that's basically the intent here.

[11:25:22.0000] <Ms2ger>
Oh, and I hadn't even seen http://www.w3.org/MarkUp/Forms/2012/WD-xforms20-20120628/ yet

[11:25:52.0000] <annevk>
Ms2ger: there was also an XForms module I think

[11:26:30.0000] <Ms2ger>
Anyway, I don't really care what the W3C does anymore

[11:26:55.0000] <Hixie>
annevk: btw is anyone implementing capture="" yet? if they are we should just spec it

[11:27:15.0000] <annevk>
I thought Android had it, but I'm not sure if Chrome Android has it too

[11:27:22.0000] <annevk>
beverloo?

[11:29:02.0000] <Hixie>
if it's just one vendor then it's not so interesting

[11:31:35.0000] <Hixie>
annevk: btw is this your bug? https://www.w3.org/Bugs/Public/show_bug.cgi?id=10213

[11:32:54.0000] <annevk>
it's from simon

[11:33:07.0000] <annevk>
you can reassign to the URL spec

[11:33:26.0000] <annevk>
it's going to be solved by the whitelist of hierarchical schemes

[11:33:52.0000] <annevk>
I can reassign too

[11:35:45.0000] <annevk>
Hixie: maybe instead of "other Hixie drafts" we should ask for a "WHATWG" or "WHATCG" product?

[11:36:10.0000] <annevk>
Hixie: and then have HTML / Encoding / Fullscreen / Quirks there

[11:37:53.0000] <Hixie>
i'd rather not have to fix all my scripts to use the new product/component

[11:38:01.0000] <Hixie>
but i can certainly see the logic

[11:38:07.0000] <Hixie>
especially for the other drafts

[11:38:37.0000] <Hixie>
i guess it wouldn't be so bad, not that many scripts to update really

[11:38:38.0000] <Hixie>
sure

[11:38:40.0000] <Hixie>
let's do that

[11:38:44.0000] <Hixie>
let's wait til after this migration though

[11:39:07.0000] <annevk>
sure, I think it'll also help people filing bugs against WHATWG work

[11:39:23.0000] <Hixie>
yeah

[11:39:45.0000] <hober>
word

[11:41:01.0000] <annevk>
zewt: are you around?

[11:41:40.0000] <Hixie>
(incidentally, i'm not changing the qa contact field during this, but i guess after the migration we should change that too)

[11:43:07.0000] <Hixie>
ok looks like part 2 is ready too

[11:43:31.0000] <Hixie>
MikeSmith: ok i'm ready whenever you are

[11:45:36.0000] <Hixie>
back in bit, lunch

[11:55:46.0000] <annevk>
zewt: added a comment to the bug instead

[11:59:32.0000] <jtcranmer>
is there any sort of spec for txt-to-html conversion or vice versa?

[12:01:09.0000] <SimonSapin>
jtcranmer: there are many text-based lightweight markup languages that can convert to HTML: markdown, restructuredText, …

[12:01:45.0000] <jtcranmer>
that doesn't answer my question :-)

[12:03:02.0000] <SimonSapin>
then, what do you mean by txt-to-html?

[12:03:25.0000] <annevk>
jtcranmer: there's no standardized way other than lots of de facto standards

[12:03:59.0000] <jtcranmer>
there are de facto standards at least for things like *bold* /italics/ _underline_ |code|

[12:04:52.0000] <SimonSapin>
yes, and markdown formalizes them somewhat

[12:04:52.0000] <jtcranmer>
and I've noticed that things like copying HTML and pasting as plain text tend to do things like turn <ol> into # and <ul> into * or similar ventures

[12:05:26.0000] <SimonSapin>
there are many variants, like `code` instead of |code|

[12:05:43.0000] <SimonSapin>
if you pick on precisely, you get something like markdown

[12:05:54.0000] <SimonSapin>
http://daringfireball.net/projects/markdown/syntax

[12:07:57.0000] <Tabatkins>
Yes, Markdown is generally accepted as the "standard" txt-markup language now.  There are several incompatible extensions, but the core is stable.

[12:08:59.0000] <jtcranmer>
how about the other direction?

[12:09:43.0000] <SimonSapin>
it has been done, but it is always lossy

[12:09:58.0000] <jtcranmer>
I don't doubt that it is

[12:24:45.0000] <Tabatkins>
It's actually a good bit easier.  Just look up what markdown can encode, then reverse it in the obvious way.

[12:25:00.0000] <Tabatkins>
<h1>foo</h1> becomes "foo\n===", etc.

[12:25:45.0000] <SimonSapin>
Tabatkins: easier because we already have HTML parsers?

[12:25:55.0000] <Tabatkins>
Yes, it's a simple tree-walk then.

[12:26:13.0000] <Tabatkins>
I'm assuming this is done in the browser.

[12:26:24.0000] <Tabatkins>
Outside of it, the HTML parser is harder than the Markdown parser. ^_^

[12:27:03.0000] <SimonSapin>
okay. I guess md-to-html is alsa just a tree-walk once the Markdown is parsed

[12:27:09.0000] <SimonSapin>
also*

[12:27:43.0000] <Tabatkins>
Well, sure.

[12:28:44.0000] <annevk>
http://softwaremaniacs.org/playground/showdown-highlight/

[12:30:41.0000] <Hixie>
the official way to convert text/plain to HTML is <pre>... :-)

[12:31:00.0000] <Tabatkins>
<plaintext>, you mean.

[12:31:22.0000] <Tabatkins>
Also, <xmp> is better for actual text/plain.

[12:31:23.0000] <Hixie>
that's the unofficial way

[12:31:33.0000] <Tabatkins>
;_;

[12:32:10.0000] <jtcranmer>
Hixie: that's how you should display a text/plain as a DOM

[12:32:34.0000] <Tabatkins>
annevk: Hrm, the default styling there doesn't do anything special for blockquotes, which looks confusing.

[12:32:35.0000] <MacTed>
<pre> fails at text/plain conversion to HTML, as there's no line-wrapping within <pre> elements

[12:32:37.0000] <jtcranmer>
Hixie: for email clients receiving a text/plain, requirements are slightly different

[12:32:54.0000] <Tabatkins>
MacTed: <pre style="whitespace: pre-wrap;">

[12:34:11.0000] <MacTed>
Tabatkins - I believe that will work in many (most) browsers. but it's a different way than <pre>.

[12:34:12.0000] <Hixie>
jtcranmer: an e-mail client receiving text/plain e-mail doesn't have to convert to HTML, it has to convert to pixels.

[12:34:15.0000] <Hixie>
(typically)

[12:34:34.0000] <Hixie>
there's no line wrapping in text/plain either :-)

[12:35:51.0000] <MacTed>
Hixie - what the rendering engine does with text/plain is outside of scope.  <pre> mandates no wrapping during render.  text/plain does no such thing.

[12:36:25.0000] <Hixie>
<pre> doesn't mandate anything on rendering

[12:36:37.0000] <Hixie>
you could render <pre> contents to a 3d orange and it wouldn't be non-compliant

[12:36:48.0000] <Hixie>
anyway

[12:37:17.0000] <Ms2ger>
/me throws a 3d orange at Hixie 

[12:37:42.0000] <Hixie>
any browser vendors other than chrome looking at this autocompletetype thing?

[12:37:51.0000] <Ms2ger>
autowhat?

[12:38:30.0000] <Hixie>
chrome proposal from a while back about a way to do better autofill

[12:47:09.0000] <Hixie>
anyone from mozilla or opera have an opinion on it?

[12:47:46.0000] <Hixie>
http://lists.w3.org/Archives/Public/public-whatwg-archive/2011Dec/0194.html was the original e-mail

[12:48:20.0000] <Ms2ger>
volkmar, ^

[12:48:50.0000] <Ms2ger>
I don't care much for it

[12:51:28.0000] <rniwa>
sigh... HTMLCollection is such a mess.

[12:51:41.0000] <Ms2ger>
s/HTMLCollection/the web/

[12:51:46.0000] <rniwa>
why does HTMLAllCollection::namedItem returns another HTMLAllCollection?

[12:51:55.0000] <rniwa>
why can't it be HTMLCollection

[12:52:02.0000] <rniwa>
or NodeList

[12:52:08.0000] <Ms2ger>
I dunno

[12:52:14.0000] <Ms2ger>
/me looks what Gecko does

[12:52:27.0000] <rniwa>
HTMLOptionsCollection::namedItem returns live NodeList

[12:52:35.0000] <rniwa>
and HTMLFormCollection::namedItem returns RadioNodeList

[12:52:54.0000] <rniwa>
HTMLPropertiesCollection::namedItem returns PropertyNodeList

[12:55:21.0000] <hsivonen>
Hixie: reinventing the RFC WF2 used to reference, eh?

[12:55:43.0000] <hsivonen>
(although in fairness, decoupling the autofill semantic from the field name is slightly different)

[12:56:41.0000] <Hixie>
hsivonen: yeah

[12:56:59.0000] <Ms2ger>
rniwa, afaict, Gecko just goes with the usual NodeList-HTMLCollection hybrid

[12:57:11.0000] <rniwa>
Ms2ger: we just return static node list LOL

[12:57:23.0000] <Hixie>
hsivonen: they had some good reasons for thinking the rfc failed due to design problems and not due to lack of need, can't remember if they told me them in person or talked about it on the list

[12:57:38.0000] <hsivonen>
Hixie: ok

[12:57:41.0000] <Tabatkins>
Should we be prefixing the new path apis on canvas?

[12:57:46.0000] <Hixie>
tano

[12:57:48.0000] <Hixie>
er

[12:57:49.0000] <Hixie>
Tabatkins: no

[12:57:52.0000] <Tabatkins>
kk

[12:58:00.0000] <hsivonen>
Tabatkins: prefixing in what way?

[12:58:24.0000] <Hixie>
i guarantee that any spec i write will not conflict with implementations, either by renaming the things in the spec or aligning with implementations directly

[12:58:35.0000] <Hixie>
so there's never any need to prefix stuff i spec

[12:58:39.0000] <Tabatkins>
hsivonen: webkitPath(), etc.

[12:58:48.0000] <Ms2ger>
Let's rename some more flexbox stuff

[12:58:54.0000] <Tabatkins>
VIOLENCE

[12:59:01.0000] <Hixie>
lol

[12:59:23.0000] <hsivonen>
Tabatkins: I wish by now it was clear that the answer to "should we vendor prefix?" is "NOOOO!" :-(

[12:59:30.0000] <Ms2ger>
But order is such a bad name...

[13:00:08.0000] <Ms2ger>
Hmm

[13:00:16.0000] <Ms2ger>
document.all.item("item")

[13:00:28.0000] <Ms2ger>
I bet Gecko doesn't handle that right

[13:00:32.0000] <Ms2ger>
Otoh, I don't care

[13:00:32.0000] <annevk>
Tabatkins: for some reason that reminds me of http://ln.hixie.ch/?start=1063663989&count=1

[13:01:02.0000] <Ms2ger>
"Margin collapsing is now well defined."

[13:01:19.0000] <Tabatkins>
Oh, the naivete of youth...

[13:01:24.0000] <hsivonen>
lol

[13:01:37.0000] <annevk>
Ms2ger: where is that from?

[13:01:49.0000] <Ms2ger>
annevk, ... guess?

[13:01:50.0000] <Tabatkins>
That post you linked.

[13:01:53.0000] <annevk>
oh doh

[13:02:08.0000] <annevk>
now you reminded me of the one where Hixie is a fan of XHTML 2.0

[13:02:15.0000] <hsivonen>
how many Last Calls did CSS 2.1 have?

[13:02:37.0000] <Tabatkins>
I consider it just one long 8-year LC.

[13:03:08.0000] <Hixie>
hey, margin collapsing at that point _was_ well defined compared to before

[13:03:21.0000] <Hixie>
it was defined in like 4 hand-wavy lines before that point

[13:03:42.0000] <Ms2ger>
"compared to before" is pretty essential

[13:04:06.0000] <Hixie>
it's like html parsing before, and after, the month i spent first speccing it

[13:04:21.0000] <Hixie>
sure, it wasn't perfectly "well defined" after either, but compared to before, hell

[13:05:31.0000] <Hixie>
ok let's see. i have a bunch of features that are only requested by one vendor but which nobody seems opposed to. if i do one such feature per vendor, that should be considered fair, right?

[13:05:36.0000] <Hixie>
apple want getImageDataHD

[13:05:46.0000] <Hixie>
chrome want like a zillion things, but let's start with autocompletetype

[13:06:04.0000] <Hixie>
mozilla want... data: in web workers was a mozilla request right?

[13:06:10.0000] <Ms2ger>
No, Opera

[13:06:21.0000] <espadrine>
The autocompletetype effort is very good

[13:06:24.0000] <Ms2ger>
At least, I'm going to put it on Opera's bill

[13:06:27.0000] <espadrine>
do they want it in the html spec?

[13:06:37.0000] <espadrine>
it feels like it doesn't belong

[13:06:41.0000] <Hixie>
ok, so what does mozilla want that nobody else wants

[13:06:57.0000] <Ms2ger>
The entire output of the WebAPI WG

[13:07:01.0000] <Hixie>
espadrine: yeah, it'd be a small subsection in the web forms section

[13:07:09.0000] <Hixie>
Ms2ger: i mean, that goes in the whatwg spec

[13:07:29.0000] <Hixie>
oh the plugin click-to-play thing

[13:07:31.0000] <Hixie>
that's a mozilla thing

[13:07:35.0000] <Hixie>
ok

[13:07:42.0000] <Ms2ger>
That's a thing by some Mozilla guy ;)

[13:07:48.0000] <Ms2ger>
And I think he retracted it

[13:08:11.0000] <Hixie>
and what does Microsoft want

[13:08:17.0000] <Hixie>
removeHitRegion?

[13:08:27.0000] <espadrine>
it looks sort of like scheme.org… but having a subsection in the spec makes it visible

[13:08:51.0000] <Hixie>
espadrine: i'd probably put scheme.org in the html spec if they asked :-)

[13:09:00.0000] <Hixie>
schema.org?

[13:09:03.0000] <Hixie>
whatever it's called

[13:09:16.0000] <Hixie>
wait, you mean the microdata thing for search engines right?

[13:09:26.0000] <Ms2ger>
Meh

[13:09:27.0000] <espadrine>
yep, misspelled ;)

[13:09:31.0000] <Hixie>
k

[13:09:37.0000] <Ms2ger>
We already implemented your microdata API

[13:09:48.0000] <Hixie>
anyway, i'd probably put that in if they wanted it in the html spec, there are other vocabs in there

[13:09:50.0000] <Ms2ger>
About which I note there are some outstanding bugs ;)

[13:10:12.0000] <Hixie>
i'm stuck on bugs until MikeSmith gets online

[13:10:17.0000] <Hixie>
anyway bugs aren't controversial

[13:10:24.0000] <Hixie>
everyone wants them fixed, generally

[13:11:08.0000] <annevk>
MikeSmith is prolly ~5h away from being online

[13:11:28.0000] <annevk>
espadrine: why should it not be in HTML?

[13:11:47.0000] <rniwa>
Ms2ger: i've sent an angry email about htmlcollection now.

[13:11:56.0000] <Ms2ger>
I noticed :)

[13:12:21.0000] <Ms2ger>
Oh, we want window.scrollMax*

[13:12:21.0000] <espadrine>
annevk: maybe it should, I'm not sure

[13:12:49.0000] <annevk>
espadrine: that and a wiki page for extensions that get later folded in if successful seems like a rather good model

[13:13:22.0000] <Hixie>
scrollMax => CSSOM

[13:14:01.0000] <Ms2ger>
Might as well say /dev/null

[13:14:22.0000] <espadrine>
annevk: as long as those extensions are not vendor-prefixed

[13:14:36.0000] <annevk>
Ms2ger: didn't quite realize it would be this bad :(

[13:14:53.0000] <annevk>
espadrine: uhuh :)

[13:14:58.0000] <Tabatkins>
I'm annoyed, but I'm giving them til the timeline that Shane gave me expires.

[13:15:01.0000] <Tabatkins>
Which is in August.

[13:16:49.0000] <Ms2ger>
When did they claim they would edit?

[13:16:54.0000] <Ms2ger>
January?

[13:17:10.0000] <Tabatkins>
Don't recall.

[13:17:45.0000] <Ms2ger>
/me doesn't believe anything useful will happen on that spec

[13:18:07.0000] <Hixie>
ok i can't get hold of josh and the thread on click-to-play does give a plausible alternative that involves no spec changes

[13:18:12.0000] <Tabatkins>
If nothing useful happens, I'll take it.  But like I said, I'm still waiting for my coworker to exceed his timeline. ^_^

[13:18:14.0000] <Hixie>
so... what else do mozilla people want

[13:20:02.0000] <annevk>
Ms2ger: it was very shortly after http://lists.w3.org/Archives/Public/www-style/2011Nov/0782.html

[13:20:22.0000] <annevk>
Ms2ger: glazou was so happy with the new editors he tweeted enthusiastically about them before anything was done

[13:20:50.0000] <annevk>
Ms2ger: but you know, it's only over half a year later, give them some time

[13:22:14.0000] <Tabatkins>
"all the email addresses in the registries have had

[13:22:15.0000] <Tabatkins>
their "@" symbols automatically converted into ampersands in order to

[13:22:16.0000] <Tabatkins>
thwart harvesting.

[13:22:22.0000] <Tabatkins>
wut

[13:23:18.0000] <Ms2ger>
Hixie, we'd like you to define table layout? :)

[13:23:21.0000] <annevk>
if url contains ietf.org look for "&"

[13:23:26.0000] <annevk>
or iana.org

[13:23:36.0000] <annevk>
ooh

[13:23:40.0000] <annevk>
is that how it works

[13:23:45.0000] <annevk>
please define CSS Hixie

[13:24:11.0000] <annevk>
I'm soon unemployed, but the people will appreciate it

[13:24:33.0000] <Tabatkins>
annevk: What actual job did you end up taking? The page you linked to isn't in English. ^_^

[13:25:01.0000] <Ms2ger>
Learn Dutch :)

[13:25:16.0000] <annevk>
Tabatkins: http://robbertbroersma.nl/ looks English to me

[13:25:56.0000] <Tabatkins>
...huh. I didn't see the blog posts before.

[13:26:13.0000] <annevk>
Tabatkins: but I'm going to be a co-founder of sorts

[13:26:49.0000] <Ms2ger>
Working on XSLT?

[13:33:33.0000] <annevk>
Ms2ger: not planning on it, but maybe, we'll see

[13:35:13.0000] <annevk>
Ms2ger: planning on having fun, writing specs, making our idea successful, and enjoying the freedom of not representing anyone

[13:35:41.0000] <Ms2ger>
Join Mozilla, you don't need to represent us either ;)

[13:36:58.0000] <annevk>
would've been the same for Google

[13:37:14.0000] <Ms2ger>
Yeah, but we're good, and they're evil :)

[13:37:20.0000] <Tabatkins>
Hey, fuck you. ^_^

[13:37:28.0000] <Ms2ger>
<3 Tabatkins

[13:38:32.0000] <Hixie>
Ms2ger: i think you're missing the part where i edit the HTML spec...

[13:38:48.0000] <Ms2ger>
Hixie, can't I troll in here? :(

[13:39:34.0000] <Hixie>
sure you can troll

[13:39:40.0000] <Hixie>
and tab can say fuck you too :-)

[13:39:50.0000] <Ms2ger>
I noticed! ;)

[13:40:05.0000] <espadrine>
annevk: which idea is that?

[13:40:19.0000] <Hixie>
(oh, you meant telling me to spec css was trolling. i thought you were referring to saying google was evil.)

[13:40:27.0000] <Hixie>
(my bad)

[13:40:31.0000] <Ms2ger>
No, mine :)

[13:40:35.0000] <Hixie>
ok well i can't find anything mozilla wants

[13:40:42.0000] <Hixie>
so i guess you guys get nothing

[13:40:45.0000] <Ms2ger>
I didn't realize you were talking about CSS

[13:40:55.0000] <annevk>
espadrine: dunno really how much we're telling, I'm on vacation and we're prolly not properly founded until October or so

[13:41:20.0000] <espadrine>
annevk: I see. Good luck in your venture then! :)

[13:41:22.0000] <Hixie>
Ms2ger: so basically we have no idea what either of us are saying. sounds like this conversation is going swimmingly. :-P

[13:41:28.0000] <Ms2ger>
:D

[13:41:30.0000] <hober>
:)

[13:41:48.0000] <Hixie>
anne: wait hold on, hold on

[13:42:03.0000] <Hixie>
annevk: that site says XSLT on it

[13:42:04.0000] <Hixie>
annevk: twice

[13:42:07.0000] <Hixie>
annevk: in big letters

[13:42:13.0000] <hober>
Hixie: re: autocompletetype="", we like the overall idea of (long-term) reducing the pain of autofill code

[13:42:21.0000] <annevk>
Hixie: :) Robbert is a fan

[13:42:47.0000] <Hixie>
...and you want to cofound something with him? o_O

[13:42:52.0000] <hober>
hahahha

[13:42:53.0000] <Ms2ger>
Hixie, I didn't miss that you edit the HTML spec, but if you're ever looking for more work to do... CSS is an option :)

[13:43:25.0000] <Hixie>
Ms2ger: i'm looking for an HTML feature that mozilla wants because i'm doing the rounds of feature requests and granting one to each vendor that no other vendor is super-eager about

[13:43:37.0000] <espadrine>
I wondered how long it would take before some XSLT trolling…

[13:43:39.0000] <Hixie>
Ms2ger: (not those that other vendors are against)

[13:44:12.0000] <Ms2ger>
Yeah

[13:44:24.0000] <annevk>
Hixie: looking forward to it even :)

[13:44:49.0000] <Ms2ger>
Hixie, can we keep that for a future feature request? :)

[13:45:05.0000] <Hixie>
Ms2ger: nah :-P

[13:45:21.0000] <Ms2ger>
XBL2? :)

[13:45:27.0000] <Hixie>
already specced that

[13:45:36.0000] <Ms2ger>
And I guess everyone else hates it

[13:45:49.0000] <Hixie>
nah only dglazkov hates it :-P

[13:46:51.0000] <Hixie>
man, of the four features here, _two_ are canvas-related

[13:46:56.0000] <Hixie>
ain't that always the way

[13:47:20.0000] <hsivonen>
in case anyone has opinions about http vs. internal declaration precedence for source maps, now is your chance to comment on https://bugzilla.mozilla.org/show_bug.cgi?id=765993

[13:47:25.0000] <Ms2ger>
Weren't you going to wait with more canvas stuff until we implemented any of the new stuff? :)

[13:47:37.0000] <Hixie>
Ms2ger: i was

[13:48:02.0000] <Hixie>
hsivonen: what's a source map?

[13:48:38.0000] <espadrine>
source map: file that contains data to go from original minified JS to readable JS

[13:48:46.0000] <espadrine>
or from coffeescript to JS

[13:48:58.0000] <espadrine>
it's useful for devtools, mainly

[13:49:12.0000] <Hixie>
oh like symbol maps for compiled binary code?

[13:49:21.0000] <espadrine>
exactly

[13:49:27.0000] <hsivonen>
Hixie: https://docs.google.com/document/d/1U1RGAehQwRypUTovF1KRlpiOFze0b-_2gc6fAH0KY0k/edit?_escaped_fragment_=

[13:50:03.0000] <Hixie>
i had no idea some people used the term "source map" instead of "symbol map", weird

[13:50:11.0000] <Hixie>
sounds like a useful feature though!

[13:51:30.0000] <Ms2ger>
Want to spec it? :)

[13:52:38.0000] <Hixie>
hsivonen: is anyone gonna use the http header? seems like a pointless feature

[13:53:06.0000] <Hixie>
hsivonen: internal makes a lot of sense (p.s. i'd still love to see a filename/lineno pragma for validator.nu)

[13:53:07.0000] <espadrine>
hsivonen: having the X in X-SourceMap is strange

[13:53:19.0000] <Ms2ger>
espadrine, that's what hsivonen said :)

[13:53:34.0000] <Hixie>
hober: i assume you want backingStorePixelRatio to return the ratio of device pixels to coordinate space units, not to CSS pixels, right?

[13:53:35.0000] <espadrine>
Ms2ger: Oops. Didn't see :)

[13:53:47.0000] <Hixie>
er, s/device pixels/backing store pixels/

[13:53:50.0000] <hsivonen>
Hixie: good point. dunno

[13:54:27.0000] <annevk>
hsivonen: why does the header have an X- prefix?

[13:54:40.0000] <annevk>
hsivonen: http://tools.ietf.org/html/rfc6648

[13:54:50.0000] <hsivonen>
annevk: I don't know, but the usual guess applies

[13:55:19.0000] <zewt>
annevk: i'm here in case you don't want to spam the bug btw

[13:55:37.0000] <hsivonen>
annevk: cool. I didn't realize that one had gotten out of the draft stage

[13:55:47.0000] <annevk>
zewt: my idea is that you have a global list of identifiers

[13:56:12.0000] <annevk>
zewt: the identifiers have some kind of revoke flag and origin tied to them, and optionally the payload data

[13:56:49.0000] <annevk>
zewt: that way blob URLs always map to them and APIs don't need special knowledge about the state of blob URLs

[13:57:04.0000] <zewt>
not quite following ... what do you mean by "revoke flag"?

[13:57:07.0000] <annevk>
zewt: but if the revoke flag is set for a particular identifier parsing the URL should not work

[13:57:09.0000] <zewt>
if a blob URL is revoked, it simply no longer exists

[13:57:21.0000] <annevk>
yeah I think we should change that concept

[13:57:26.0000] <Hixie>
hober: also, do you want it on canvas of the 2d context? i'm guessing 2d context, since the webgl guys have their own issues on this front

[13:57:27.0000] <annevk>
I think it should simply fail to parse

[13:57:28.0000] <Tabatkins>
Hixie: Yes, definitely.

[13:57:35.0000] <espadrine>
hsivonen: since Chrome's implementation of source maps is not in stable builds yet, maybe we can change this before it's too late?

[13:57:38.0000] <annevk>
because otherwise the blob URL can no longer be fetched

[13:57:40.0000] <Tabatkins>
Hixie: Sorry, regarding the backingStorePixelRatio.

[13:57:54.0000] <annevk>
you'd need a whole bunch of special casing everywhere

[13:58:05.0000] <hober>
Hixie: yes to your first question

[13:58:07.0000] <zewt>
but you have to free up the URL at some point, or else a loop creating blob URLs would leak forever

[13:58:29.0000] <annevk>
zewt: it always uses the same identifier?

[13:58:34.0000] <hsivonen>
espadrine: oh. cool

[13:58:40.0000] <zewt>
no, it's different each time you create it

[13:58:54.0000] <annevk>
zewt: right, so you need to free up the identifier at some point

[13:58:55.0000] <hober>
Hixie: re: where it should hang off of, it would be nice if it were available to script before creating a canvas

[13:59:12.0000] <zewt>
i don't follow what all the special casing is about

[13:59:14.0000] <annevk>
zewt: and then the URL can succeed to parse, but simply references no existing identifier and therefore still fails

[13:59:43.0000] <zewt>
the only possible per-spec thing I think might be needed is saying when the reference to the underlying blob data is released (and even that might not be needed--not sure)

[13:59:55.0000] <Hixie>
hober: how can you know it before knowing which canvas you're asking about? or are you saying there should not be a way for UAs to do it on a per-canvas basis?

[13:59:57.0000] <hober>
Hixie: [also, while we're on this, i dislike the *HD names and would be happy with better ones. :)]

[14:00:05.0000] <zewt>
see the pseudocode in my last post (to see if we're thinking of what resolve would do in the same way)?

[14:00:11.0000] <annevk>
zewt: <img>.src = url; you're saying 1) that if that's a blob URL <img> somehow needs to store its payload data and 2) that when fetching it cannot fetch the blob URL but must fetch that payload data instead

[14:00:12.0000] <Hixie>
/me kinds likes the HD names :-P

[14:00:27.0000] <Hixie>
kinda, even

[14:00:28.0000] <hober>
Hixie: yeah, we suggested them because we couldn't come up with better, but "least bad" != good

[14:00:29.0000] <annevk>
zewt: so all APIs need 1) special storage and 2) special fetching

[14:00:46.0000] <zewt>
annevk: no, resolve would handle 1) and fetch would handle 2)

[14:00:48.0000] <zewt>
img doesn't care

[14:00:58.0000] <annevk>
zewt: how does resolve handle it?

[14:01:05.0000] <zewt>
see my last post on the bug

[14:01:33.0000] <annevk>
zewt: so you're storing the data on the URL object?

[14:01:57.0000] <zewt>
yeah, though it's just a string from what I understand

[14:02:09.0000] <annevk>
zewt: what is just a string?

[14:02:11.0000] <zewt>
(doesn't matter for these purposes)

[14:02:16.0000] <zewt>
the result of "resolve a URL"

[14:02:32.0000] <zewt>
not intimately familiar with that algorithm

[14:02:35.0000] <annevk>
it's going to be an object of some kind

[14:02:50.0000] <annevk>
and it's going to be called something with parsing

[14:02:54.0000] <hober>
Hixie: re: exposing backingStorePixelRatio in a way that doesn't require you to create a canvas first, see dino's mail here: https://www.khronos.org/webgl/public-mailing-list/archives/1206/msg00193.html

[14:03:03.0000] <zewt>
that's fine either way, the point is the blob data is attached to that result (which the caller API would mostly ignore, except that when it hands that URL to fetch, fetch now has access to it)

[14:03:18.0000] <zewt>
which encapsulates most of this into those two algorithms

[14:03:31.0000] <hober>
Hixie: if you do this naively, you create a huge canvas before you realize you shouldn't :(

[14:03:47.0000] <annevk>
zewt: I guess that works, and supposedly it's just a pointer on URL anyway so that can be shared

[14:04:06.0000] <annevk>
zewt: it's kind of ugly, but not more or less than what I came up with

[14:04:19.0000] <zewt>
yeah it's a bit of a hack but the best I've thought of

[14:04:20.0000] <annevk>
zewt: maybe a little more since it affects both parsing and fetching

[14:04:34.0000] <annevk>
but then mine changes URLs in weird ways

[14:04:38.0000] <zewt>
i don't fully understand your idea

[14:04:50.0000] <Hixie>
hober: i don't see how that contradicts the spec

[14:04:57.0000] <Tabatkins>
hober: Ah, so it exposes the implicit multiplier, not the final multiplier.  Cool.  So it's 1 on iOS devices, right?

[14:05:01.0000] <annevk>
global list of blob identifiers that can have a revoke flag set

[14:05:09.0000] <zewt>
when is that list purged?

[14:05:13.0000] <hober>
Tabatkins: right

[14:05:21.0000] <annevk>
once a blob URL is parsed that references an identifier whose revoke flag is set, it's turned into about:invalid

[14:05:22.0000] <Hixie>
hober: my problem is that if you are creating a canvas every few minutes, and the user is zooming every few minutes, you're going to want a different scale factor each time.

[14:05:56.0000] <Hixie>
hober: i guess we could say that each <canvas> gets a scale factor assigned at birth and that there's a global attribute that returns the scale factor the UA will use if you create a canvas immediately

[14:05:59.0000] <annevk>
zewt: same as when identifiers cease to exist now

[14:06:09.0000] <zewt>
"identifiers"?

[14:06:15.0000] <annevk>
blob:identifier

[14:06:21.0000] <zewt>
you mean blob URLs?

[14:06:24.0000] <zewt>
(sorry if I'm being thick, I'm just missing a piece somewhere)

[14:06:36.0000] <annevk>
a blob URL is blob: followed by the identifier

[14:06:47.0000] <zewt>
but that wouldn't work

[14:06:53.0000] <annevk>
but the identifier holds other concepts that are not part of the URL

[14:06:54.0000] <annevk>
such as origin

[14:07:01.0000] <zewt>
the URL->blob data mapping needs to outlive the blob URL itself

[14:07:28.0000] <annevk>
right

[14:07:43.0000] <hober>
Hixie: that would work for me

[14:07:48.0000] <annevk>
identifier->data mapping outlives the time blob URL can be successfully parsed

[14:08:13.0000] <zewt>
but didn't you say that the mappings would be purged when the blob URL (or "identifier"--not sure if you mean some distinction) is revoked

[14:09:04.0000] <annevk>
once the revocation happens the identifier's revoke flag is set which causes further parsing of blob URLs using that identifier to fail (turn into about:invalid)

[14:09:20.0000] <zewt>
but when are items removed from the mapping entirely?

[14:09:21.0000] <annevk>
but if you already have a URL object with that specific blob URL it could still be fetched

[14:09:41.0000] <annevk>
zewt: that's the same as your quality of implementation thingie

[14:10:06.0000] <zewt>
feels fuzzier, but maybe just because i don't have as much of a mental handle on it

[14:10:15.0000] <annevk>
afaict it's the same model

[14:11:07.0000] <annevk>
but I'm okay with special casing parsing, the URL object, and fetching

[14:11:31.0000] <zewt>
i guess i think of a list/map of stuff that you insert into as needing to remove it explicitly, where attaching data to an object just feels GC-y

[14:11:34.0000] <annevk>
need to find a better term for URL object, since there's also window.URL which is somewhat different

[14:11:57.0000] <zewt>
guess if you call it a weakmap it makes more sense

[14:12:09.0000] <zewt>
well anyway

[14:13:04.0000] <zewt>
with the URL-property approach, do you think leaving releasing to QoI is good enough?

[14:13:20.0000] <zewt>
one thing is that specs still need to pay attention to it, to make sure it results in the right conclusion

[14:15:01.0000] <zewt>
eg. that they don't accidentally require keeping blob references around longer than expected

[14:15:20.0000] <annevk>
I guess reloading should fail afterwards? e.g. <iframe>.contentWindow.location.reload()

[14:15:21.0000] <zewt>
(such as if xhr allowed calling send() multiple times per open(), that would have happened)

[14:15:47.0000] <zewt>
hmm

[14:16:25.0000] <zewt>
any time you're able to reload the data later it'd force keeping the blob around ... so it's a judgement call

[14:17:42.0000] <annevk>
can't fetch just free it up?

[14:17:56.0000] <annevk>
after fetch completes, clear it from the URL object?

[14:18:09.0000] <annevk>
so if the URL object is reused by the API after that, it fails

[14:18:27.0000] <annevk>
and if the API reuses it before the fetch completes, it works

[14:18:30.0000] <zewt>
i guess it could, with two caveats: 1: it'd prevent ever using it multiple times (maybe there's never any reason to) and 2: it'd still depend on QoI for freeing (because you might eg. never call xhr.send() at all)

[14:19:11.0000] <annevk>
well yeah, unused objects that might be used will use memory

[14:19:15.0000] <annevk>
not sure if there's a way around that

[14:19:24.0000] <zewt>
i mean, after discarding the xhr object

[14:19:49.0000] <zewt>
(if you call open() and then keep the xhr around forever and never send() it, yeah, keeping the blob around forever is what the user is asking for)

[14:20:18.0000] <annevk>
yeah the implementation would need to make sure that the objects stored on the XHR object are freed

[14:20:21.0000] <annevk>
including the URL object

[14:20:23.0000] <annevk>
and everything it has

[14:20:30.0000] <annevk>
that seems pretty straightforward though

[14:21:08.0000] <annevk>
XHR.collect() -> will collect the associated URL object which will collect itself

[14:21:15.0000] <zewt>
yeah

[14:21:42.0000] <annevk>
not sure when there would be a multiple times scenario

[14:21:45.0000] <zewt>
though you'd also want to release the blob ref when entering DONE, without waiting for GC (but that's also just QoI)

[14:22:06.0000] <annevk>
if you enter done fetch has finished and you'd be required to

[14:22:22.0000] <Hixie>
hm, i guess we'll need a createImageDataHD() version too

[14:22:53.0000] <zewt>
Hixie: webgl decided to ignore how canvas handles pixel ratios, and make everyone do it by hand :(

[14:23:00.0000] <zewt>
consistency smash

[14:23:46.0000] <annevk>
main problem: fixing the URL spec will take a while

[14:23:53.0000] <annevk>
but then File API has been taking a long time

[14:24:15.0000] <zewt>
seems like this is doable with the current resolve algorithm as-is

[14:24:38.0000] <zewt>
maybe not as cleanly as if it was returning a concrete object

[14:25:03.0000] <annevk>
oh sure, some temporary hack could be fabricated

[14:25:22.0000] <annevk>
not going to invest much time in that though

[14:25:24.0000] <zewt>
i'm just afraid that if this gets held up, the terrible "oneTimeOnly" hack in IE will get a foothold

[14:26:26.0000] <annevk>
I think if you add this model as a separate comment to the bug Arun should be able to define how it works in a hacky way

[14:26:41.0000] <annevk>
then we'll clean it up once the URL spec is further along

[14:27:17.0000] <zewt>
shouldn't this be defined in HTML anyway (since that's where fetch and resolve live)

[14:28:00.0000] <Hixie>
zewt: they're screwed, their API has assumptions in it that means they have no choice

[14:28:18.0000] <zewt>
Hixie: i gave them a way to do it consistently; they ignored me

[14:28:30.0000] <zewt>
(or at least much more consistently than what they landed on)

[14:50:44.0000] <annevk>
https://twitter.com/rybesh/status/223823857449041921 about teaching Microdata and RDFa

[14:54:58.0000] <Hixie>
annevk: he must be biased

[14:55:00.0000] <Hixie>
:-P

[14:56:26.0000] <zewt>
hmm, i don't think microtasks are right for autoRevoke, but there are some quirks with stable states too

[14:59:17.0000] <Tabatkins>
Ugh, the JS i18n API uses RangeError to mean "doesn't match one of the values in this enum". >_<

[15:01:32.0000] <Hixie>
uh

[15:29:04.0000] <tantek>
annevk - I've heard similar comparisons of microdata and RDFa, of course even more so I hear that they're both still too much of a bother (e.g. compared to microformats).

[15:30:03.0000] <tantek>
I was originally more optimistic about microdata, but it turns out it (and RDFa) suffer from the "bolt-on" problem (ala longdesc) - they're just not part of web designers' normal workflow, so they get ignored.

[15:30:41.0000] <tantek>
in contrast, all web designers spend time in class attributes, so to put in a few microformats classes while they're in there is anywhere from trivial to easy.

[15:32:29.0000] <annevk>
I find all of them too much of a bother typically :)

[15:32:37.0000] <Hixie>
aha, tantek!

[15:32:39.0000] <annevk>
haven't seen much useful stuff done with them

[15:33:28.0000] <Hixie>
/me agrees with anne :-)

[15:33:38.0000] <zewt>
as a web developer i have no idea what either of them are, and the drama around both of them makes me not very interested in finding out :)

[15:34:25.0000] <tantek>
indeed, if it doesn't seem useful to you, then don't bother is a reasonable short-term prioritization approach

[15:35:24.0000] <tantek>
annevk - "too much of a bother" - often true unfortunately, no matter what markup we're talking about (from <section> to <abbr> to RDFa)

[15:36:01.0000] <annevk>
yeah, I stopped using <abbr> too

[15:36:08.0000] <annevk>
that's something dictionaries can solve

[15:36:26.0000] <tantek>
it's pretty clear from years of teach web designers, that one must demonstrate that some semantic markup provides benefits to them that outweighs the "bother" in order for them to do it and keep doing it.

[15:37:08.0000] <annevk>
<section> I might use if I had more complex sites

[15:37:09.0000] <tantek>
this can be improved by increasing the benefits received, or lowering the amount of bother, or both.

[15:37:21.0000] <annevk>
currently using <h2> works for me

[15:37:45.0000] <Hixie>
i wish i had <section> available when i started the html spec

[15:37:51.0000] <Hixie>
that would have made moving stuff around way easier

[15:38:06.0000] <Hixie>
as it is now i have a horrible tangle of pragmas to bump the hx numbers up and down on output

[15:38:07.0000] <tantek>
with microformats 2, we're focusing first on lowering the "bother" (since that worked for microformats in the first place, as compared to things like XML data islands etc.).

[15:38:08.0000] <annevk>
ah yeah, for that it's super useful

[15:38:25.0000] <annevk>
that would be useful on the WHATWG blog too

[15:38:33.0000] <tantek>
does anything actually support <section> scoped Hn tags?

[15:38:43.0000] <annevk>
tantek: microformats though has a lot of nonsense

[15:38:49.0000] <annevk>
e.g. class=url on <a>

[15:38:54.0000] <annevk>
although I think that's being removed now

[15:39:54.0000] <tantek>
annevk - yeah, we tried just saying let all <a> links mean something, and that resulted in too many false positives. however, the class=url has been removed on simple hyperlinked things.

[15:40:37.0000] <tantek>
and I'd say your instinct is in general a good one, if some markup feels like nonsense, then avoid it.

[15:41:08.0000] <tantek>
annevk - what's being removed? the section scoped Hn thing?

[15:41:09.0000] <annevk>
learned it the hard way

[15:41:29.0000] <annevk>
came from the "it's in the spec so it makes sense" camp

[15:41:40.0000] <annevk>
tantek: making class=url optional

[15:44:01.0000] <annevk>
tantek: dunno about <section> scoped <h1>-<h6> apart from that not having to renumber headings would be useful

[15:44:22.0000] <tantek>
agreed that it would be useful. I just keep wondering if anyone is going to implement it.

[15:44:43.0000] <tantek>
sorry, not "useful" but rather, "less of a bother"

[15:44:46.0000] <annevk>
btw zewt, please don't quote in Bugzilla if it's directly in reply to the previous comment

[15:45:52.0000] <annevk>
tantek: dunno about that, cost is quite high, benefit not that big

[15:46:37.0000] <annevk>
controls are where the real benefit is at I think

[15:46:43.0000] <annevk>
such as <dialog>

[15:47:13.0000] <annevk>
some way to do a dropdown without lots of scripting would be cool too

[15:48:35.0000] <annevk>
nn

[16:06:32.0000] <Hixie>
should createImageDataHD() also have the imagedata argument variant? It'd be identical to the non-HD version... (/cc hober)

[16:07:21.0000] <Tabatkins>
For usability, I'd say yes.

[16:08:53.0000] <Hixie>
that's what i was thinking, but i wonder if it'll confuse people, thinking it's a way to convert a low-res imageData to a high res one or something...

[16:30:14.0000] <hober>
/me doesn't have a strong opinion either way


2012-07-14
[21:10:51.0000] <Yuhong>
http://lists.w3.org/Archives/Public/www-html/2001Mar/0023.html

[21:11:17.0000] <Yuhong>
about document.all and document.getElementById

[21:11:34.0000] <Yuhong>
from 2001 before IE6 was released by Hixie.

[21:12:05.0000] <Yuhong>
from 2001 before IE6 was released, by Hixie.

[21:37:49.0000] <jackjack>
Hello, everyone!

[21:38:25.0000] <jackjack>
is there anyone who is familiar with HTML tokenization states?

[21:39:02.0000] <jackjack>
I need a little help

[21:39:10.0000] <jackjack>
Thanks!

[21:44:17.0000] <espadrine`>
jackjack: do you have a question?

[21:45:42.0000] <jackjack>
Yes, as mentioned by HTML standards: "	In some cases, the tokenizer state is also changed by the tree construction stage.

[21:45:42.0000] <jackjack>
"

[21:46:12.0000] <jackjack>
I am wondering what are the "some cases"?

[21:48:58.0000] <jackjack>
 "The exact behavior of certain states depends on the insertion mode and the stack of open elements."

[21:50:04.0000] <jackjack>
What are those "certain states" ?

[21:50:32.0000] <jackjack>
Anyone can help me ?

[21:52:14.0000] <espadrine`>
I would assume that this has to do with document.write, but I cannot assert that for sure

[21:54:49.0000] <espadrine`>
although it's bad to use document.write

[21:55:17.0000] <espadrine`>
cats die when you do that

[22:58:40.0000] <SimonSapin>
Tabatkins: Looking at your new CSS parser now. Do you exepect to have JS tests for it? How would these relate to a test suite for the spec?

[23:41:37.0000] <jwalden>
gsnedders: the yak-shaving and the __proto__ patch test out and are posted for review now!

[23:41:42.0000] <jwalden>
\o/ \o/ \o/ \o/

[23:42:13.0000] <jwalden>
https://bugzilla.mozilla.org/show_bug.cgi?id=773850 and https://bugzilla.mozilla.org/show_bug.cgi?id=770344 if you care

[23:44:28.0000] <espadrine`>
yay for __proto__!

[23:44:39.0000] <espadrine`>
not sure what yak-shaving does though

[00:18:22.0000] <jwalden>
boo for __proto__

[00:18:36.0000] <jwalden>
but yay for its being implemented sanely, if it must be implemented at all

[00:19:02.0000] <jwalden>
the yak-shaving was some API cleanup/improvement over what we had before, which was thoroughly bizarre and unreadable

[00:39:28.0000] <annevk>
jackjack: <plaintext>, <script>, and such cause the tokenizer to change from the parser

[00:40:22.0000] <annevk>
jackjack: espadrine: document.write() just injects into the character stream, it does not actually change state (although the characters it injects could of course)

[01:08:10.0000] <rniwa>
annevk: not sure if you've read my rant about htmlcollection but can we somehow name HTML*Collection::namedItem saner?

[01:08:18.0000] <rniwa>
annevk: as is, it's implementation/API nightmare.

[01:08:31.0000] <rniwa>
somehow make*

[01:10:40.0000] <Ms2ger>
"<Yuhong> from 2001 before IE6 was released by Hixie."

[01:10:45.0000] <Ms2ger>
/me didn't read that as intended

[01:11:29.0000] <annevk>
how was that intended?

[01:12:24.0000] <Ms2ger>
A comma after "released"

[01:12:24.0000] <annevk>
rniwa: I recommend filing a bug on Gecko and Opera to see if they agree to implement saner behavior

[01:12:52.0000] <rniwa>
annevk: i don't think neither Gecko nor Opera implements what the spec says.

[01:12:54.0000] <Ms2ger>
rniwa, and test if we don't do that already :)

[01:12:57.0000] <annevk>
rniwa: if Hixie disagrees the simplest approach is to convince the other vendors of your right

[01:13:19.0000] <rniwa>
annevk: hixie mention that there are some comat. issues :\

[01:13:23.0000] <rniwa>
mentions*

[01:13:39.0000] <annevk>
rniwa: no he presumes there are

[01:14:20.0000] <rniwa>
annevk: hm... maybe i should try that then.

[01:14:54.0000] <annevk>
there might be compat issues and the people that can tell you are the other vendors

[01:15:23.0000] <annevk>
but sometimes e.g. Gecko is willing to take a small short term hit for long term API improvements

[01:25:31.0000] <rniwa>
annevk: yeah, Mozilla is good at that.

[01:25:38.0000] <rniwa>
annevk: wish we could do the same in WebKit :\

[01:25:45.0000] <rniwa>
annevk: but we have some constraints :(

[01:25:49.0000] <annevk>
you can in Chromium ;)

[01:26:04.0000] <rniwa>
annevk: oh well, we try not to diverse too much between ports

[01:26:37.0000] <rniwa>
annevk: we don't want to end up with 4-5 different flavors of webkit :)

[01:27:16.0000] <rniwa>
annevk: arguably, we're more liberal about changing web-facing APIs than other ports but still.

[01:29:20.0000] <Ms2ger>
And here I thought you already had ended up with 4-5 different flavours of webkit :)

[01:30:21.0000] <rniwa>
Ms2ger: noooo. don't point out the painful fact.

[01:30:28.0000] <Ms2ger>
:D

[01:31:16.0000] <rniwa>
Ms2ger: ironically, i use firefox on my android tablet because that's the only browser available on my ICS tablet that reliably lets me review/comment on bugs.webkit.org

[01:32:20.0000] <Ms2ger>
Yw ;)

[01:32:34.0000] <Ms2ger>
Now, what do you think about Chrome-on-Gecko? :)

[01:33:49.0000] <rniwa>
Ms2ger: you mean chrome inside BtG?

[01:33:52.0000] <rniwa>
BtoG*

[01:34:00.0000] <rniwa>
or was it B2G?

[01:34:01.0000] <espadrine>
rniwa: Firefox OS now

[01:34:09.0000] <rniwa>
espadrine: oh, that's right.

[01:34:18.0000] <rniwa>
espadrine, Ms2ger: so you want Chrome on Firefox OS?

[01:34:43.0000] <Ms2ger>
No, I want Chrome to get rid of webkit and use Gecko instead :)

[01:34:55.0000] <rniwa>
Ms2ger: oh... not sure about that.

[01:35:05.0000] <rniwa>
Ms2ger: you guys don't support multi process model still :\

[01:35:14.0000] <rniwa>
Ms2ger: and don't support all those webkit extensions.

[01:35:33.0000] <rniwa>
Ms2ger: anyway, i think it's nice to have competition between different browser engines

[01:35:36.0000] <Ms2ger>
That's coming soon enough

[01:35:56.0000] <rniwa>
Ms2ger: maybe we'll consider once you've removed all the comments in your code :)

[01:36:05.0000] <Ms2ger>
Hah

[01:36:16.0000] <Ms2ger>
To make it look more like webkit? :)

[01:36:49.0000] <rniwa>
Ms2ger: http://lists.webkit.org/pipermail/webkit-dev/2012-July/thread.html

[01:37:32.0000] <Ms2ger>
/me snorts

[01:37:53.0000] <Ms2ger>
Hey, if you join us, you don't need to write ChangeLog entries ;)

[01:38:40.0000] <espadrine>
although you will need to struggle with mercurial queues

[01:39:13.0000] <rniwa>
Ms2ger: no, i love change logs :)

[01:39:22.0000] <rniwa>
espadrine: must use svn ;)

[01:39:28.0000] <Ms2ger>
Shrug, MQ is nice

[01:40:02.0000] <espadrine>
ouch. svn. I thought webkit people all used git as an svn front-end

[01:40:25.0000] <rniwa>
espadrine: no.

[01:40:36.0000] <rniwa>
espadrine: i actually prefer svn over git

[01:40:42.0000] <rniwa>
espadrine: git is way too complicated for my work flow

[01:41:18.0000] <rniwa>
espadrine, Ms2ger: if i do join mozilla, you'll find out how "unconventional" and how unhip i am :)

[01:41:26.0000] <Ms2ger>
rniwa, we hate git too ;)

[01:41:48.0000] <Ms2ger>
Except the a-team and webdev

[01:42:14.0000] <rniwa>
Ms2ger: http://lists.webkit.org/pipermail/webkit-dev/2012-March/019808.html for your enjoyment

[01:42:25.0000] <rniwa>
or http://lists.webkit.org/pipermail/webkit-dev/2012-March/thread.html

[01:42:29.0000] <Ms2ger>
WebKittens?

[01:42:36.0000] <rniwa>
yeah

[01:43:28.0000] <espadrine>
one day we'll all have to move to git anyway

[01:43:46.0000] <espadrine>
Servo is handled by git :)

[01:44:13.0000] <rniwa>
espadrine: i'm not convinced.

[01:44:38.0000] <espadrine>
rniwa: don't worry, I won't make the elevator pitch.

[01:45:07.0000] <rniwa>
anyway, i have a serious problem with people who advocate for one-fits-all solution.

[01:47:15.0000] <rniwa>
they irritate me to no end

[01:47:35.0000] <rniwa>
just like people who jokingly talk about everything would have been easy if we all spoke English, etc...

[01:48:21.0000] <Ms2ger>
Imagine how many would speak en-AU then

[01:52:59.0000] <rniwa>
Ms2ger: I had a Scottish English teacher who lied in Ireland for years.

[01:53:05.0000] <rniwa>
lived*

[01:53:21.0000] <Ms2ger>
The original was more fascinating

[01:53:38.0000] <Ms2ger>
/me imagines an obsessive liar in a kilt

[01:53:59.0000] <Ms2ger>
... walking through the Irish countryside

[01:54:04.0000] <rniwa>
hehehe

[01:55:11.0000] <rniwa>
Ms2ger: actually, Dublin is the only city in Europe i've been to.

[01:55:27.0000] <rniwa>
/me is excited about TPAC in Lyon :)

[01:55:28.0000] <Ms2ger>
For the beer? :)

[01:55:36.0000] <rniwa>
Ms2ger: i was 17 at the time

[01:55:39.0000] <rniwa>
Ms2ger: so no beer for me.

[01:55:54.0000] <Ms2ger>
I see

[01:56:01.0000] <Ms2ger>
/me is never excited about TPAC

[01:58:16.0000] <volkmar>
Hixie: for the autocompletetype, my opinion (and bz when we talked) is that it is a very bad idea and we can solve that we current input types

[01:58:32.0000] <volkmar>
I've never found time to reply though, still in my todo list

[02:00:18.0000] <espadrine>
volkmar: you mean, by having a better "name=" de-facto standard?

[05:45:13.0000] <gsnedders>
Exciting times in ES land, trying to subvert what some people want…

[06:32:39.0000] <annevk>
gsnedders: isn't that what usually happens? :)

[06:37:24.0000] <gsnedders>
annevk: No, not in ES land.

[06:38:00.0000] <annevk>
ES4?

[06:39:11.0000] <gsnedders>
Well, that wasn't so much subvertion as just outright refusal to implement something they'd never been involved in. But I was really meaning Harmony onwards.

[06:45:48.0000] <annevk>
gsnedders: so what's going on?

[06:46:14.0000] <gsnedders>
__proto__, and whether it should be an accessor property or not.

[06:54:51.0000] <annevk>
I should invest time in learning JavaScript again at some point

[06:59:16.0000] <annevk>
Ms2ger: re the addendum stuff, writing time != publication time

[07:00:02.0000] <annevk>
e.g. Linköping story was well over a year old

[07:00:06.0000] <annevk>
the /\

[11:25:11.0000] <zewt>
grr no onerror for <link rel=stylesheet> on ios

[11:27:07.0000] <zewt>
at least <script defer> works

[11:30:36.0000] <zewt>
no onerror, yet onload works

[11:30:46.0000] <zewt>
guess i might be able to fudge it with that

[11:31:09.0000] <zewt>
wait, it doesn't work; oh well

[12:50:57.0000] <Hixie>
volkmar: please make sure to reply on the list soon, because right now all the feedback is positive and i'm ready to spec it

[13:00:22.0000] <Yuhong>
BTW, it is unfortunate that it took until just after Gecko introduced Almost Standards Mode before the first draft of CSS 2.1 was introduced.

[13:02:53.0000] <Yuhong>
*published.

[13:12:41.0000] <jackjack>
annevk: espadrine': thanks for answering my question.

[15:04:59.0000] <divya>
annevk: YOU MADE IT TI W3C MEMESS

[15:53:48.0000] <annevk>
divya: hahaha


2012-07-15
[22:17:09.0000] <divya>
annevk: how cool iszat

[22:17:16.0000] <divya>
u gotta thank tantek for thememe

[02:14:06.0000] <annevk>
I wonder why http://www.belshe.com/2012/06/04/the-web-only-works-thanks-to-reload-and-why-the-mobile-web-fails/ has not been discussed

[02:14:22.0000] <annevk>
Although I have to say I don't encounter those problems often myself

[04:43:13.0000] <Ms2ger>
And looks like Nightly got unprefixed calc() in time for 16 too

[12:52:12.0000] <GPHemsley>
Ms2ger: Could you point me to the section that defines the header rendering?

[12:52:50.0000] <GPHemsley>
oh wait

[12:53:03.0000] <GPHemsley>
Ms2ger: Is it this? http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#sections-and-headings

[16:05:43.0000] <Hixie>
how do i work out the size of the viewport in JS?

[16:06:13.0000] <Hixie>
document.documentElement.clientHeight and clientWindth?

[16:18:37.0000] <espadrine`>
Hixie: innerHeight and innerWidth, I think

[16:18:53.0000] <Hixie>
innerheight and width give the element's content area width or some such

[16:18:58.0000] <Hixie>
there's no element that's the size of the viewport

[16:19:34.0000] <espadrine`>
cssom-view defined innerWidth as "the viewport width including the size of a rendered scroll bar (if any)"

[16:19:51.0000] <Hixie>
oh. on what object then?

[16:19:58.0000] <espadrine`>
on window

[16:19:59.0000] <Hixie>
not on an Element i assume

[16:20:00.0000] <Hixie>
ohhhh

[16:20:02.0000] <Hixie>
ok

[16:20:03.0000] <Hixie>
that might work

[16:58:59.0000] <zewt>
what the

[16:59:04.0000] <zewt>
chrome doesn't have modal alert()?

[16:59:09.0000] <zewt>
wow


2012-07-16
[01:09:10.0000] <hsivonen>
Hixie: the other day, you were trying to find spec changes "Mozilla" wants. Have you noticed that Mozilla is reviving the inputmode attribute (with different values if I understand correctly)?

[02:06:09.0000] <hsivonen>
MikeSmith: this thing uses the W3C's servers as its back end, right? http://w3clove.com/plans-and-pricing

[02:06:41.0000] <MikeSmith>
hsivonen: yeah I believe it does

[02:07:59.0000] <hsivonen>
MikeSmith: mildly interesting that the front end talks about priority levels then

[02:08:26.0000] <hsivonen>
but I guess there are priorities in the front end queue

[04:35:41.0000] <Ms2ger>
(CVS $Revision: 87e938c14c3f $)

[04:54:07.0000] <hsivonen>
so now I have to figure out what the navigation timing spec says about restoration from the BF cache...

[04:55:40.0000] <hsivonen>
restore from BF cache isn't supposed to fire a readystatechange, is it?

[04:57:57.0000] <hsivonen>
nice. the navigation timing spec doesn't cover BF cache

[04:58:00.0000] <hsivonen>
sigh.

[04:58:09.0000] <hsivonen>
let's see if HTML5 does

[04:58:14.0000] <hsivonen>
or HTML LS rather

[05:01:01.0000] <hsivonen>
hooray. HTML LS actually covers BF cache and it's pretty clear my assertions found a Gecko bug

[05:01:12.0000] <hsivonen>
I hope no one is depending on the bug yet...

[05:11:31.0000] <Ms2ger>
\o/

[05:33:17.0000] <MikeSmith>
BF cache?

[05:33:39.0000] <hsivonen>
MikeSmith: back/forward cache for Document objects

[05:33:50.0000] <MikeSmith>
OK

[05:34:52.0000] <hsivonen>
the next time someone gives a gift like readyState, we should be more careful about making sure the gift comes with unit tests

[05:45:36.0000] <hsivonen>
  } else {

[05:45:36.0000] <hsivonen>
    // XXX: Should fire error event to the document...

[05:45:36.0000] <hsivonen>
  }

[05:45:51.0000] <hsivonen>
doesn't look exactly compliant there

[05:48:35.0000] <hsivonen>
Hixie: sorry about the tone of my latest email in the readyState thread. I'm rather frustrated about this not being a bug fix in the past yet.

[07:21:30.0000] <hsivonen>
how fun. removing-editable-xslt.html

[07:21:41.0000] <hsivonen>
because everyone wants to have editable XSLT

[07:26:22.0000] <[tm]>
hahaha

[07:27:18.0000] <hsivonen>
the generated documents for image, video, audio and plug-in URLs (when loaded alone in a browsing context) and XSLT documents are my least favority kind of documents

[07:27:23.0000] <hsivonen>
after about:blank of course

[07:30:10.0000] <Ms2ger>
hsivonen, well, fortunately nobody scripts the those :)

[07:34:33.0000] <[tm]>
if browser vendors would just implement xslt 3.0 I'm sure it would solve all those problems

[07:37:47.0000] <[tm]>
seriously though xpath and xslt are still a better API than the standard DOM

[09:00:07.0000] <dglazkov>
good morning, Whatwg!

[09:24:41.0000] <kennyluck>
good night, Whatwg!

[09:25:36.0000] <espadrine`>
kennyluck: are you doing the opposite of what dglazkov is doing every morning?

[09:26:27.0000] <kennyluck>
espadrine`, sort of, but I am about to go to bed too.

[09:27:07.0000] <espadrine`>
good night then!

[10:16:06.0000] <annevk>
dglazkov: kennyluck: it's WHATWG \o/

[10:16:27.0000] <annevk>
also happy e-day

[10:17:41.0000] <espadrine`>
annevk: what's an e-day?

[10:26:34.0000] <annevk>
not sure, but I meant it as a timezone neutral greeting

[10:34:02.0000] <Hixie>
hsivonen: ah, excellent, just the kind of thing i was looking for. do you have any links to that? there's some bugs asking for inputmode but they're all stuck on people doing some research so we can work out what it actually needs to be

[10:35:31.0000] <Hixie>
annevk: try "hi" :-P

[10:37:12.0000] <Hixie>
hsivonen: "Firefox, IE10, and Opera" is less than 50% of the market, what are you talking about "rocking the interop boat"?

[10:38:01.0000] <Hixie>
especially since in the text you quote you point out firefox isn't doing the same thing either

[10:39:32.0000] <gsnedders>
jwalden: Yay!

[10:44:39.0000] <Hixie>
MikeSmith: yt?

[10:56:29.0000] <bencc>
shepazu: you were the one recommending fakesmile?

[11:48:44.0000] <hsivonen>
Hixie: Firefox currently fires the event you say shouldn't fire

[11:49:29.0000] <hsivonen>
the way current Firefox differs from what I want is its omission to transition to "interactive"

[11:50:49.0000] <Hixie>
firing DOMContentLoaded doesn't make much sense if you're not firing 'load'

[11:50:59.0000] <Hixie>
and firing DOMContentLoaded when you haven't loaded the DOM content is silly IMHO

[11:51:10.0000] <Hixie>
i don't see what possible advantage it would be to authors

[11:51:55.0000] <hsivonen>
Hixie: about inputmode: https://bugzilla.mozilla.org/show_bug.cgi?id=746142

[11:52:16.0000] <hsivonen>
Hixie: I see no advantage in not firing it

[11:53:53.0000] <Hixie>
do you plan to fire 'load'?

[11:54:45.0000] <hsivonen>
no. not because I think it makes sense not to but because everyone refrains from firing it

[11:58:48.0000] <Hixie>
hsivonen: refrain from firing DOMContentLoaded and then most people will refrain from firing that, too

[11:58:58.0000] <Hixie>
s/people/browsers/

[11:59:16.0000] <Hixie>
hsivonen: DOMContentLoaded is basically the same event as 'load', just before scripts

[11:59:29.0000] <Hixie>
hsivonen: it would be silly to fire one and not the other, IMHO

[11:59:32.0000] <Hixie>
hsivonen: and very confusing to authors

[12:58:29.0000] <annevk>
not sure anyone cares, but @WHATWG is nearing 6666

[12:58:55.0000] <Hixie>
6666 what?

[12:59:02.0000] <Tabatkins>
tweets, presumably

[12:59:42.0000] <annevk>
followers

[13:00:14.0000] <annevk>
tweets is 3752

[13:02:54.0000] <Hixie>
are we still tweeting all the non-editorial changes?

[13:03:37.0000] <annevk>
yes

[13:03:49.0000] <Hixie>
man, that's 6666 crazy people

[13:03:49.0000] <annevk>
we omit things that say "typo" and "xref"

[13:04:37.0000] <annevk>
the editorial stuff is quite popular sometimes

[13:04:46.0000] <Hixie>
heh

[13:05:00.0000] <annevk>
people retweet things like "fix obsolete non-normative text"

[13:05:06.0000] <Hixie>
o_O

[13:05:09.0000] <annevk>
and pay no attention to normative changes

[13:05:14.0000] <annevk>
I'm not sure how this works

[13:07:18.0000] <Hixie>
maybe instead of createImageDataHD i should just have a third argument that takes the scale factor

[13:07:49.0000] <Hixie>
nah

[13:07:52.0000] <Hixie>
it's tied to a canvas

[13:07:57.0000] <Hixie>
ok

[13:08:01.0000] <Hixie>
createImageDataHD it is

[13:08:05.0000] <hober>
it's hard to feature detect optional args v. new methodas

[13:08:09.0000] <hober>
err, methods

[13:09:29.0000] <annevk>
so there will be a new object ImageDataHD too/

[13:09:30.0000] <annevk>
?

[13:10:37.0000] <Tabatkins>
hober: One reason I'm excited about @supports - makes it easier to feature-detect new values in existing properties.

[13:59:32.0000] <Hixie>
hober: doesn't .length return the length of the longer one?

[13:59:34.0000] <Hixie>
annevk: no

[13:59:57.0000] <Hixie>
hober: (and yeah, i went back to the HD idea anyway)

[14:00:16.0000] <Hixie>
hober: what was the attribute you wanted, again, btw?

[14:13:32.0000] <Tabatkins>
hsivonen: Some discussion about <template> led to an interesting insight.

[14:14:35.0000] <Tabatkins>
hsivonen: Regarding <iframe srcdoc>, it was designed as an attribute to make escaping easier, but ignoring that, it would obviously have been easier to do a <sandbox> element, where you just put in user content directly.

[14:15:07.0000] <Tabatkins>
hsivonen: In that case, you'd have elements parsed as part of the outer document, but bundled up into a separate document hanging off of the main doc instead.

[14:15:23.0000] <Tabatkins>
hsivonen: Which is, of course, essentially identical to what <template> does.

[14:15:47.0000] <Tabatkins>
hsivonen: Do your arguments against <template> apply equally to <sandbox>?  Or are there significant differences that I'm not seeing?

[14:17:51.0000] <othermaciej>
Tabatkins: the reason a <sandbox> element wouldn't work is that it's too hard to prevent hostile content prematurely closing the <sandbox> tag

[14:18:13.0000] <Tabatkins>
Yes, I referred to that above.  That's why we used an attribute instead of an element.

[14:18:36.0000] <Tabatkins>
But disregarding that consideration, @srcdoc is effectively embedding a second document into the content of the first, inline.

[14:19:08.0000] <Tabatkins>
Basically I'm just probing at hsivonen's internal consistency here, to see if there are good reasons to be against one and not the other.

[14:19:10.0000] <Hixie>
it's much like what <style> and <script> already do

[14:19:18.0000] <Tabatkins>
Yeah.

[14:19:22.0000] <Hixie>
i suppose we could have <template> wrk more like those

[14:19:25.0000] <Hixie>
have the text be in the outer doc

[14:19:32.0000] <Hixie>
and then have a parsed dom hang off the side

[14:19:49.0000] <othermaciej>
since <sandbox> is hypothetical, I'm not sure it's a useful query of hsivonen's views

[14:19:52.0000] <Hixie>
much like how <style> has a textContent node unparsed, and a CSSOM off the side, post-parse

[14:19:56.0000] <othermaciej>
since being for or against it has no consequence

[14:20:41.0000] <Tabatkins>
othermaciej: Just checking. If hsivonen would be okay with <sandbox> (ignoring the security consideration), then I'd be curious why, and how that applies to his arguments against <template>.

[14:20:44.0000] <othermaciej>
if you made <template> work like that, then presumably in xhtml, you would have to escape its contents as a CDATA section

[14:20:48.0000] <Tabatkins>
Hixie: No problem with that from me.

[14:20:57.0000] <Tabatkins>
othermaciej: I guess, yeah.

[14:21:13.0000] <Hixie>
<template> as specced just doesn't work in xhtml

[14:21:21.0000] <Hixie>
as proposed, i should say

[14:21:38.0000] <othermaciej>
if <sandbox> was inline, then it might not want to create a subframe and a separate document

[14:21:44.0000] <Tabatkins>
If that was the objection, and saying "just escape as CDATA" gets around it, that's not a problem. ^_^

[14:21:46.0000] <Hixie>
or maybe "as specced" is right, the proposal is pretty detailed :-)

[14:22:05.0000] <Tabatkins>
othermaciej: I think most of the sandboxing stuff would be very annoying / error-prone to implement if you didn't use a separate document.

[14:22:11.0000] <Hixie>
othermaciej: we used a subframe for its security properties

[14:22:11.0000] <othermaciej>
I have to admit I don't really understand the reasons why <template> can't be in the normal DOM

[14:22:35.0000] <Hixie>
othermaciej: it's similar to sandboxing, actually. you need a way to neuter all the features. forms, images. etc.

[14:22:58.0000] <annevk>
also selectors

[14:23:30.0000] <othermaciej>
I don't understand why selectors have to be neutered (that argument somewhat confuses me) and I don't understand how a separate document helps w/ those other aspects

[14:23:31.0000] <annevk>
you can theoretically do the others by adding conditionals

[14:23:38.0000] <Tabatkins>
Yeah, just like sandboxing, you *can* do it in the same document, but it's way easier conceptually (and you're less likely to have errors) if you do it as a separate doc.

[14:24:00.0000] <Tabatkins>
othermaciej: You get them for free!  Disconnected documents don't load anything, don't contribute to forms, don't match against selectors from the outer document...

[14:24:10.0000] <othermaciej>
preventing script execution and subdocument loading and the like is already a conditional in WebKit's DOM, you'd merely need to change the condition

[14:24:26.0000] <othermaciej>
er, subresource loading

[14:25:16.0000] <othermaciej>
like, there's nothing about how XHR-loaded documents do those things that intrinsically needs to be document-scoped

[14:25:45.0000] <othermaciej>
selector matching is the only anomaly, and I am not sure why it needs to be one (maybe I would if I read more about it)

[14:25:54.0000] <Hixie>
othermaciej: you want to neuter everything. appearing in document.images, appearing in an ancestor's form.elements, appearing in an ancestor select.options, form submission, creating of nested browsing contexts, running scripts, etc etc etc

[14:26:13.0000] <Hixie>
othermaciej: there are so many places we'd have to add checks it would be unmanageable

[14:26:14.0000] <othermaciej>
I think <template> has a hard problem that it won't be able to offer any of its features as currently designed unless you require some sort of escaping of the content

[14:26:38.0000] <othermaciej>
er, won't be able to offer them in non-supporting browsers

[14:26:49.0000] <Hixie>
i think the idea is to make <template> work exactly like XHR-loaded documents

[14:26:52.0000] <othermaciej>
so the more differences it has from normal content, yet without escaping, the harder it is to use

[14:27:07.0000] <Hixie>
(except inline)

[14:27:16.0000] <othermaciej>
unless you wait until the hypothetical day when all browsers have it

[14:27:16.0000] <Tabatkins>
Hixie: I think that's right, yeah.

[14:27:25.0000] <othermaciej>
or to put it another way, it doesn't degrade gracefully

[14:27:31.0000] <Tabatkins>
othermaciej: Those hypothetical days come sooner all the time. ^_^

[14:27:41.0000] <Hixie>
yeah, it definitely doesn't degrade that well as proposed

[14:27:57.0000] <Hixie>
we've tried a graceful degradation approach before (see web forms 2's repetition model), it wasn't that great

[14:28:01.0000] <Tabatkins>
We're okay with pushing on a good idea even if it'll be a few years before it's usable in practice.  Degrading is great, but you can't always degrade gracefully.

[14:28:09.0000] <othermaciej>
Tabatkins: I wonder if a version of the feature that has the templates out-of-line or escaped would be viable

[14:28:16.0000] <Hixie>
certainly if someone can come up with a way to make it degrade, i'd be all for it

[14:28:26.0000] <othermaciej>
are those options so bad that it's worth the greater transition cost?

[14:28:39.0000] <Tabatkins>
othermaciej: We're down with out-of-band templates too.  If that's the way you ahve to use them until support evolves (parsing on your own in legacy browsers), that's also fine.

[14:28:54.0000] <othermaciej>
<template src="foo.html#id"></template>, the way to degrade is use XHR

[14:29:06.0000] <Hixie>
Tabatkins: have they (you? we?) considered doing something like srcdoc="" for <template>?

[14:29:14.0000] <othermaciej>
that would have an excellent graceful degradation story

[14:29:22.0000] <Tabatkins>
Hixie: I don't think we have.

[14:29:26.0000] <othermaciej>
<template srcdoc=""> would also have a good story

[14:29:36.0000] <Hixie>
might be worth looking at

[14:29:42.0000] <othermaciej>
if you give up on "inline template as normal markup", all the issues go away

[14:30:23.0000] <othermaciej>
where "give up" might just mean "don't focus on initially"

[14:30:49.0000] <Tabatkins>
It's kinda nasty, but if it's a second inline option alongside proper children, that sounds like it would be palatable.

[14:38:00.0000] <hober>
Hixie: (sorry for the delay) a static attr on HTMLCanvasElement, window, or something like that which is "the backingStorePixelRatio that you'll get if you create a new canvas right now."

[14:38:19.0000] <Hixie>
no preferred name or location though?

[14:41:16.0000] <Hixie>
/me discovers that toDataURL and toBlob aren't actually required to return the canvas' image at all

[14:41:19.0000] <Hixie>
oops

[14:43:39.0000] <hober>
I don't have a particular name or location preference, no, so long as you don't need to instantiate a canvas to get at it

[14:44:08.0000] <Hixie>
do we have a spec that specs window.screen? maybe that'd be a good place for it

[14:45:51.0000] <hober>
that's a pretty reasonable place, yeah. AFAIK there's no spec for window.screen though. :(

[14:46:43.0000] <Hixie>
http://dev.w3.org/csswg/cssom-view/#screen !

[14:46:45.0000] <Hixie>
we do have it!

[14:46:46.0000] <Hixie>
awesome

[14:47:06.0000] <Hixie>
man i wish anne was still editing those specs

[14:47:37.0000] <Hixie>
anyone know who i have to speak to to get things added to that?

[14:47:40.0000] <hober>
"Editor's Draft 13 March 2012" sadface. they were supposed to update the ED last week...

[14:47:47.0000] <Tabatkins>
I'm gonna poke the thread again.

[14:47:57.0000] <Hixie>
oh that's those specs that are in limbo isn't it

[14:48:02.0000] <hober>
yup

[14:48:17.0000] <Tabatkins>
Yup.  I'll pick them up in September if they stay in limbo.

[14:48:45.0000] <Hixie>
aren't you, like, already under a pile of specs three miles high

[14:48:52.0000] <Tabatkins>
Your point?

[14:49:00.0000] <Hixie>
yeah i guess glass houses and all...

[14:51:30.0000] <Tabatkins>
Hixie: Feel free to add something to it from the HTML spec for now, but shoot an email to www-style about it so we can track it.

[14:51:37.0000] <Hixie>
k

[14:55:54.0000] <Hixie>
hober: i've also added toDataURLHD and toBlobHD

[14:56:05.0000] <Hixie>
hober: we probably need to do something to drawImage, too

[14:56:32.0000] <Hixie>
hober: though maybe people can deal with the latter by giving the expected dimensions, so nevermind

[14:57:31.0000] <hober>
nice, thanks

[15:31:18.0000] <Tabatkins>
Ah, CSSOM started updating today.

[16:07:45.0000] <linelevel>
 From the W3 spec on HTML5 data-* attributes: "It would be inappropriate, however, for the user to use generic software not associated with that music site to search for tracks of a certain length by looking at this data. This is because these attributes are intended for use by the site's own scripts, and are not a generic extension mechanism for publicly-usable metadata."

[16:07:50.0000] <linelevel>
I'm wondering what the rationale for that restriction is. It seems like a great way to embed public metadata in HTML.

[16:08:12.0000] <Tabatkins>
Two reasons.

[16:08:38.0000] <Tabatkins>
1) Eliminate any possibility of accidental collisions, like a search engine assigning meaning to certain data-* attributes that you happen to be using for something else.

[16:08:55.0000] <Tabatkins>
There's no need for a registry or anything if we explicitly disallow this kind of thing, so that all usage is private.

[16:09:24.0000] <Tabatkins>
2) That's what Microdata is for - it's better designed for decorating pages with metadata.

[16:11:21.0000] <linelevel>
Tabatkins: Thanks, that makes sense.


2012-07-17
[22:35:47.0000] <hsivonen>
Tabatkins: your hypothetical <sandbox> would get similar objections from me as <template>

[23:31:28.0000] <Hixie>
hsivonen: how do you feel about the hypothetical <template srcdoc="">?

[23:38:13.0000] <hsivonen>
Hixie: I think it would be mostly inconvenient compared to letting the nodes be in the tree but it wouldn't have the DOM Consistency problems if it parsed consistently regardless of the XMLness of the host doc

[23:39:04.0000] <hsivonen>
Hixie: I'm not actually advocating it as a solution, because it would be syntactically inconvenient to author

[23:39:05.0000] <Hixie>
do you really think it's plausible to put the dom nodes in the dom inline and then change the zillions of things that need to be neutered? it seems really intractable to me

[23:39:31.0000] <Hixie>
(i'm not a big fan of the attribute solution either)

[23:39:41.0000] <Hixie>
(but i don't understand your objection to the inline thing)

[23:41:46.0000] <hsivonen>
Hixie: one possibility is that <template> is the wrong thing to want in the first place and we should instead focus on making it easy to write document fragments inline inside .js files

[23:42:20.0000] <Hixie>
those seem like orthogonal concerns

[23:42:29.0000] <hsivonen>
Hixie: I do think it's tractable to make inline nodes as static as nodes in an XSLT program

[23:42:30.0000] <Hixie>
but i'm definitely 100% behind solving the latter problem

[23:42:50.0000] <Hixie>
i have no idea if nodes in an XSLT program are static

[23:43:03.0000] <Hixie>
i wish we could just drop XSLT

[23:43:09.0000] <hsivonen>
Hixie: they aren't quite orthogonal to the extent <template> doesn't do anything in declaratively on its own

[23:43:10.0000] <Hixie>
it's so ill-defined

[23:43:33.0000] <Hixie>
(and used almost but not quite rarely enough to be dropped, grr)

[23:43:37.0000] <hsivonen>
Hixie: if <template> exists only to provide scripts a fragment to clone, the fragment might as well be put in the script

[23:43:58.0000] <Hixie>
my understanding is that <template>'s main purpose is for web components, is that wrong?

[23:44:18.0000] <hsivonen>
Hixie: if that's the main purpose, I've missed some big-picture memo

[23:44:22.0000] <Hixie>
i wasn't aware scripts would do much with them, i kinda assumed it would be mostly declarative like the repetition model

[23:44:42.0000] <Hixie>
anyway i agree that html-in-script is something more interesting to solve

[23:45:02.0000] <Hixie>
can you get anyone at mozilla to implement http://hixie.ch/specs/e4h/strawman ? :-)

[23:45:14.0000] <hsivonen>
Hixie: unlikely

[23:45:32.0000] <Hixie>
:-(

[23:45:36.0000] <hsivonen>
Hixie: Conway's Law: it would belong to a different team

[23:47:38.0000] <hsivonen>
Hixie: anyway, my DOM Consistency concern would be less defensible if the proponents of <template>-as-a-wormhole-to-another-world were ready to introduce the same model to XML parsing and face the wrath of XML people

[23:48:05.0000] <hsivonen>
Hixie: if there are reasons not to do that with XML parsing, consider those same reasons raised as reasons not to do it in HTML parsing

[23:48:25.0000] <hsivonen>
(I'm behind in my W3C mailing list reading, so I'm not sure if the position on XML parsing has been reversed by now)

[23:49:01.0000] <hsivonen>
Hixie: that said, I'm uneasy about the wormhole model even if it applied to both XML and HTML

[23:50:27.0000] <Hixie>
the reason not to do it with xml parsing is "the xml spec doesn't say to do it"

[23:50:38.0000] <Hixie>
and... conway's law, we don't control the xml spec :-)

[23:50:54.0000] <Hixie>
(i don't think they really care about xml)

[23:51:15.0000] <Hixie>
(can't say i blame them)

[23:51:22.0000] <hsivonen>
well, XML is in the same department as XSLT: not a priority but can't be dropped

[23:51:28.0000] <Hixie>
yeah

[23:51:32.0000] <Hixie>
but doesn't have to be supported well either

[23:51:54.0000] <hsivonen>
rewriting Gecko's XML loading code paths is on my list of things to do

[23:52:15.0000] <hsivonen>
so XML is still planned to get engineering time

[23:52:37.0000] <hsivonen>
I even feel surprisingly motivated to do it

[23:52:46.0000] <Hixie>
mozilla uses xml more than other browsers

[23:53:17.0000] <hsivonen>
Hixie: oh the Mozilla-special parts are not planned to get overhauled

[23:53:36.0000] <Hixie>
loading xml has an impact on firefox load time

[23:53:41.0000] <Hixie>
it doesn't have an impact on other browsers

[23:54:01.0000] <hsivonen>
it would be so much nicer not to have to worry about RDF and XUL

[23:54:08.0000] <hsivonen>
or XSLT

[23:54:17.0000] <hsivonen>
those are the 3 special cases

[23:54:32.0000] <hsivonen>
within the XML universe that is

[00:04:42.0000] <abarth>
hsivonen: i'm not sure why XML should be allowed to constrain the development of HTML

[00:05:29.0000] <annevk>
hsivonen: sounds more like the Gecko XML universe :)

[00:05:38.0000] <abarth>
if you want to use feature A that isn't supported in XML, then don't use XML.

[00:07:00.0000] <hsivonen>
abarth: well, at some point someone will use <template> in XML and then we'll be in a place where everyone loses if one browser vendor tries to be nice and support that case

[00:08:23.0000] <annevk>
wouldn't that be prevented by a test case?

[00:08:36.0000] <annevk>
and the note in the spec "Does not work in XML"

[00:08:42.0000] <hsivonen>
we have been in a situation where we had DOM inconsistency

[00:08:49.0000] <hsivonen>
the situation was bad

[00:08:59.0000] <hsivonen>
and we worked hard to get DOM Consistency

[00:09:22.0000] <annevk>
I always thought of that whole concept more as sharing the same DOM model

[00:09:29.0000] <annevk>
not as the same mapping-to-the-DOM model

[00:09:56.0000] <hsivonen>
annevk: well, they aren't going to share the same DOM model if one supports wormholes and one doesn't.

[00:10:04.0000] <annevk>
like using the same namespace, having the DOM behave the same in both environments (more or less)

[00:10:30.0000] <abarth>
hsivonen: one supports entity references and the other doesn't :)

[00:11:01.0000] <hsivonen>
abarth: different level of abstraction

[00:11:09.0000] <annevk>
hsivonen: I don't really see DOM consistency as a requirement that the effects the markup one of them can create in the DOM, the other has to too

[00:11:23.0000] <abarth>
anyway, there's no reason a DOM generated from XML can't support wormholes

[00:11:32.0000] <annevk>
hsivonen: e.g. XML can create namespaced nodes

[00:11:35.0000] <abarth>
isn't that just a property on the HTMLTempalteElement?

[00:11:48.0000] <annevk>
hsivonen: and now HTML would be able to create template trees

[00:11:57.0000] <annevk>
hsivonen: doesn't seem like a huge deal to me

[00:12:45.0000] <annevk>
abarth: no it's also how the contents are parsed and put on the fragment that attribute returns

[00:12:49.0000] <annevk>
put in*

[00:12:53.0000] <hsivonen>
annevk: I think making the DOM not be a syntax tree and turning it into an application-specific tree is a huge deal

[00:13:17.0000] <abarth>
/me is confused

[00:13:22.0000] <abarth>
doesn't shadow DOM do that?

[00:13:32.0000] <abarth>
I'm clearly missing something here

[00:13:55.0000] <annevk>
abarth: the idea is that <template>something</template> turns into <template/> in the DOM with an associated fragment that contains something

[00:14:03.0000] <hsivonen>
abarth: does the shadow DOM have a serialization?

[00:14:23.0000] <abarth>
I feel like you're conflating two things

[00:14:52.0000] <abarth>
1) there's a DOM, independent of serialization

[00:14:59.0000] <abarth>
that's shared between HTML and XML

[00:15:12.0000] <abarth>
2) there are HTML and XML serializations

[00:15:21.0000] <abarth>
<template> is not supported in the XML serilaization

[00:15:23.0000] <abarth>
that's it

[00:15:37.0000] <hsivonen>
abarth: but the DOM isn't just any tree. It's a tree designed to represent the syntax of tags-and-text markup

[00:15:47.0000] <abarth>
none of that has changed

[00:15:57.0000] <annevk>
well it kind of has

[00:16:05.0000] <annevk>
because the contents of <template> are elsewhere

[00:16:11.0000] <abarth>
not any more that <style><div>hi</div></style> changed it

[00:16:24.0000] <hsivonen>
abarth: so saying that the correspondence between tags-and-text and the DOM nodes is now different is a huge departure from what the DOM was designed to be

[00:16:26.0000] <abarth>
s/that/than/

[00:16:39.0000] <abarth>
so?

[00:16:41.0000] <annevk>
abarth: I think it's more; that's still in the DOM

[00:16:48.0000] <abarth>
having script in web page was a huge departure too

[00:16:53.0000] <abarth>
and it has turned out to be a good thing

[00:17:06.0000] <abarth>
well, don't put it in the DOM

[00:17:18.0000] <abarth>
put it in an HTMLTemplateElement propertly

[00:17:22.0000] <abarth>
like a shadow DOM

[00:18:31.0000] <abarth>
your argument has now shifted to not make reference to XML

[00:18:44.0000] <abarth>
which, IMHO, is a more interested conversation

[00:18:53.0000] <abarth>
s/interested/interesting/

[00:18:58.0000] <abarth>
/me can't type

[00:19:13.0000] <hsivonen>
abarth: no, XML is a useful aid for making my argument, because the correspondence between tags-and-text and DOM nodes is even more straight-forward in XML

[00:19:32.0000] <abarth>
the XML issue is off-point

[00:19:50.0000] <abarth>
it's more interesting to discuss whether this is a good or bad thing for HTML

[00:19:55.0000] <abarth>
irrespective of the existence of XML

[00:20:03.0000] <abarth>
for most intents and purposes, XML doesn't exist

[00:20:53.0000] <annevk>
it sounds like hsivonen's point is that you can serialize the DOM without element-specific knowledge, but that is already not true for HTML

[00:21:10.0000] <Hixie>
hober: yt?

[00:22:14.0000] <hsivonen>
at the most basic, my argument is that 1) DOM isn't an app-specific API first; it's a direct representation of tags-and-text syntax. 2) That essential nature of the DOM shouldn't be changed lightly, because changing fundamentals has unforeseen consequences. 3) Stuff that has unforeseen consequences shouldn't be done in gung-ho "let's just implement this feature [in Chrome]" manner

[00:22:47.0000] <Hixie>
1) is clearly false IMHO, since all the HTML elements have dedicated interfaces into the DOM

[00:22:50.0000] <abarth>
(3) seems prematurely argumentative

[00:23:06.0000] <abarth>
I'm not aware of that happening with <template>

[00:23:10.0000] <Hixie>
in fact, it was first an application-specific API (new Image(), new Option(), etc) onto which a generic model was then layered

[00:23:15.0000] <Hixie>
retroactively

[00:23:16.0000] <hsivonen>
so I'm not trying to argue what specific bad effects this might have, because my fear is that changing fundamentals will have consequences I don't see now

[00:23:45.0000] <annevk>
but you also don't see the potential good effects

[00:23:51.0000] <Hixie>
and clearly nobody is doing 3) since if this was gung-ho, it'd be in the spec already instead of people trying to convince you :-)

[00:24:45.0000] <hsivonen>
annevk: true

[00:25:55.0000] <hsivonen>
Hixie: well, just brushing off XML to the side instead of having a design that integrates will all existing parts of the platform seems gung-ho to me

[00:26:25.0000] <hsivonen>
(it's so weird to be the person who seems to care the most about XML)

[00:26:38.0000] <annevk>
might be worth studying http://msdn.microsoft.com/en-us/library/windows/desktop/ms766512(v=vs.85).aspx

[00:26:45.0000] <Hixie>
(on an unrelated note: i think i'm going to put little "fingerprint" icons in the spec in the places that leak fingerprintable data. If anyone has an icon to contribute, or a list of places that should get the icon, please let me know, e.g. by filing a bug or sending me an e-mail.)

[00:26:48.0000] <annevk>
but I cannot find much information about success/failure

[00:27:45.0000] <Hixie>
hsivonen: i'm all for solutions that work in XML too. I just haven't seen one. I don't think we should sacrifice the web's progress on the altar of XML -- if we can't find a solution, we have to move on, imho.

[00:27:57.0000] <hsivonen>
annevk: yeah, data islands are precedent. and failed precedent

[00:29:06.0000] <hsivonen>
Hixie: about #1, even though elements have convenience properties and methods, the tree structure corresponds to tags-and-text rather than application-specific considerations

[00:29:57.0000] <abarth>
it doesn't though

[00:30:03.0000] <annevk>
hsivonen: failed because unappealing or because lack of adoption by other vendors?

[00:30:11.0000] <abarth>
or rather it only does if you know some special rules for some of the tags

[00:30:21.0000] <annevk>
hsivonen: successful enough for https://developer.mozilla.org/en/Using_XML_Data_Islands_in_Mozilla it seems

[00:30:28.0000] <hsivonen>
annevk: unappealing to other vendors ;-)

[00:30:50.0000] <Hixie>
hsivonen: the tree structure was not first

[00:31:36.0000] <hsivonen>
Hixie: the current crop of implementations is no longer based on the model that existed before the tree structure

[00:31:38.0000] <Hixie>
hsivonen: the tree structure was layered on top of the app-specific API years later.

[00:31:56.0000] <Hixie>
hsivonen: that doesn't seem relevant to the question of what it is

[00:31:58.0000] <hsivonen>
Hixie: rather, it was layered under

[00:32:50.0000] <Hixie>
i'm not arguing that the tree structure isn't important, just that it's not some core concept that should be considered holy

[00:33:26.0000] <Hixie>
the dom _is_ an application specific api, was from the start, and we haven't made it any less so (q.v. srcdoc="")

[00:33:37.0000] <abarth>
hsivonen: how would you feel if we allowed XL in <script> ?

[00:34:02.0000] <hsivonen>
XL?

[00:34:03.0000] <abarth>
it just so happens that the syntax of JavaScript doesn't look like tags-and-text

[00:34:07.0000] <abarth>
http://en.wikipedia.org/wiki/XL_(XML_programming_language)

[00:34:25.0000] <abarth>
but there isn't any reason why we shouldn't be able to use a scripting language

[00:34:30.0000] <abarth>
whose syntax is tags-and-text

[00:34:50.0000] <abarth>
then the <script> tag would do the same thing that's bothering you about the <template> tag

[00:34:57.0000] <annevk>
the HTML spec already caters to that possibility

[00:34:58.0000] <Hixie>
abarth: xforms actions is the language i usually reference in this context

[00:35:22.0000] <abarth>
ok, there are lots of them

[00:35:28.0000] <abarth>
XL was the first one I found :)

[00:35:32.0000] <Hixie>
:-)

[00:35:49.0000] <Hixie>
CSS even _has_ a tree structure _in the DOM_

[00:36:05.0000] <hsivonen>
abarth: I think supporting that sort of language in <script> would be a can of worms

[00:36:38.0000] <hsivonen>
abarth: I'm aware that WHATWG HTML allowes (allowed?) it, but fortunately, no one really supports a languages like like

[00:36:39.0000] <abarth>
what if we replaced the < brackets with {

[00:36:57.0000] <hsivonen>
abarth: it would make it less of a can of worms from the DOM perspective

[00:37:01.0000] <abarth>
it's unclear to me why < has a special status in the syntax of scripting languages

[00:37:24.0000] <hsivonen>
abarth: because < is special in the wrapper

[00:37:39.0000] <abarth>
i mean, JavaScript uses <

[00:37:55.0000] <hsivonen>
abarth: but this is theoretical. I disapprove of non-JS <script> language in general, including Dart and VBScript

[00:37:56.0000] <abarth>
it just doesn't look like tags

[00:38:12.0000] <abarth>
IMHO, that's short sighted

[00:38:15.0000] <hsivonen>
abarth: < in inline JS is a Polyglotness problem

[00:38:25.0000] <Hixie>
abarth: i think for this line of argumentation CSS is the better path, because it is already parsed into an exposed OM

[00:38:27.0000] <abarth>
yes, hence not supporting <template> in XML

[00:38:32.0000] <abarth>
so, not allowed in polyglot

[00:38:40.0000] <abarth>
ok, consider CSS

[00:38:46.0000] <abarth>
but using < rather than {

[00:39:19.0000] <hsivonen>
abarth: you are basically trying to argue that since we already have cans of worms, it's OK to open some more

[00:39:39.0000] <abarth>
my argument is that your argument from purity ignores that we're already impure

[00:39:39.0000] <Hixie>
we don't just have a can of worms, dude. We're swimming in a swamp of worms.

[00:40:01.0000] <hsivonen>
Hixie: I had them in plural already :-)

[00:40:05.0000] <abarth>
the correspondence between DOM and tags-and-text isn't as tight as you make it sound

[00:40:23.0000] <hsivonen>
abarth: from impl point of view, <script> is the most special of elements

[00:40:43.0000] <hsivonen>
abarth: I accept it as legacy, but I don't view it as a role model for introducing anything else

[00:40:44.0000] <Hixie>
i dunno about _that_

[00:40:53.0000] <Hixie>
<script> is bad, granted.

[00:40:55.0000] <abarth>
DOCTYPE has got to be more special

[00:40:57.0000] <abarth>
:)

[00:40:58.0000] <Hixie>
but <table> parsing is crazy

[00:41:03.0000] <Hixie>
<iframe> is buts

[00:41:06.0000] <Hixie>
nuts even

[00:41:11.0000] <Hixie>
<object> is a land of its own

[00:41:14.0000] <Hixie>
<frameset>...

[00:41:32.0000] <Hixie>
not to mention the magic that is nested <a> parsing!

[00:41:40.0000] <Hixie>
ok i'll stop now

[00:42:20.0000] <hsivonen>
Hixie: none of those can do the kind of things to the parser state as <script>

[00:42:32.0000] <Hixie>
<script> is only crazy because of document.write()

[00:42:48.0000] <Hixie>
which by the way is why i neutered that for svg, which you keep arguing we should bring back! :-P

[00:43:12.0000] <hsivonen>
Hixie: on the XML side, it can suspend the parser as a nod to document.write even though document.write doesn't work

[00:43:30.0000] <abarth>
/me needs to go to sleep soon

[00:43:40.0000] <Hixie>
i don't recall what we did in xml mode

[00:43:46.0000] <Hixie>
i don't really care about xml

[00:51:40.0000] <Hixie>
hsivonen: i don't see anything in the spec that does the document.write() thing in xml, am i missing something?

[00:54:17.0000] <hsivonen>
non-async/defer script block the xml parser to show a predictable DOM state to scripts

[00:54:32.0000] <hsivonen>
maybe it's not specced anywhere

[00:54:45.0000] <Hixie>
<Script src=""> blocks the parser's modifications of the DOM, sure

[00:54:50.0000] <Hixie>
that's got nothing to do with document.write()

[00:55:01.0000] <hsivonen>
nothing?

[00:55:04.0000] <Hixie>
that's just the scripts being synchronously loaded

[00:55:16.0000] <Hixie>
the same would occur if document.write() didn't exist at all

[00:55:39.0000] <hsivonen>
I don't believe it would

[00:56:04.0000] <Hixie>
ok

[00:56:06.0000] <Hixie>
either way

[00:56:12.0000] <Hixie>
i did not have document.write() in mind when speccing that

[00:56:20.0000] <Hixie>
and it has nothing to do with document.write() in the spec

[00:56:35.0000] <Hixie>
it's just sync external script loading

[00:56:49.0000] <Hixie>
and it's hardly "crazy"

[00:56:59.0000] <Hixie>
compared to things like <object>'s processing model

[00:57:45.0000] <hsivonen>
object is very light in craziness from the parser impl POV

[00:58:24.0000] <Hixie>
oh for parsing, sure

[00:58:43.0000] <Hixie>
parsing in general is very light in craziness from the big picture point of view :-)

[00:58:48.0000] <Hixie>
XML parsing

[00:58:54.0000] <hsivonen>
script has the worst parsing complications

[00:59:07.0000] <Hixie>
in XML parsing, it has the only complications

[00:59:31.0000] <Hixie>
well, i guess </video> and </object> technically have some minor complications

[00:59:39.0000] <Hixie>
but that's about it, in xml

[01:00:14.0000] <hsivonen>
and <svg>

[01:00:34.0000] <Hixie>
oh because of their event nonsense?

[01:00:41.0000] <hsivonen>
right

[01:00:55.0000] <Hixie>
yeah i haven't been affected by that so it doesn't come to mind :-)

[01:02:24.0000] <Hixie>
anyway. i think the original point stands: if there's a solution that works for XML, I'm sure people would be happy to hear about it, but so far none have been put forward, and it doesn't seem sensible to block progress on that. And when you forget about XML, it really doesn't seem like the proposal is unreasonable; it reuses existing logic.

[01:06:39.0000] <Hixie>
ok, HD canvas is done, modulo all the bugs i just introduced that hober and zcorpan will point out

[01:06:52.0000] <Hixie>
next up is removeHitRegion(), i guess

[01:06:53.0000] <Hixie>
tomorrow

[01:13:14.0000] <annevk>
how do URLs in <template> descendant elements resolve?

[01:13:38.0000] <annevk>
I guess base URLs would have to work

[01:14:17.0000] <annevk>
we could just modify XML really

[01:14:46.0000] <annevk>
introduce <xml:template> or something odd like that

[01:22:44.0000] <annevk>
http://what-if.xkcd.com/3/ is awesome

[03:32:39.0000] <annevk>
Is there any circumstance where you do not have a base URL?

[03:33:01.0000] <annevk>
(other than the address bar)

[03:36:41.0000] <hsivonen>
<xml:foo> is never the right solution

[03:37:00.0000] <hsivonen>
for evidence, see xml:lang, xml:id and <xml:include>

[03:38:40.0000] <annevk>
hsivonen: yeah after I mentioned that I wondered whether I should correct it to html:template

[03:38:49.0000] <annevk>
hsivonen: or just let it be warnocked

[03:45:39.0000] <MikeSmith>
http://www.w3.org/community/xmlhypermedia/ created

[03:45:49.0000] <MikeSmith>
by the xml:href guy

[03:48:21.0000] <MikeSmith>
interesting list of technologies on this "Anatomy of a Web Developer" T-shirt http://rlv.zcache.com/anatomy_of_a_web_developer_brain_driven_by_mysql_tshirt-p235273746211126463b7wq7_400.jpg

[03:55:08.0000] <hsivonen>
/me didn't know there was *the* xml:href guy

[04:06:24.0000] <hsivonen>
so not only have I semi-accidentally come into contact with browsing context code, I have even more accidentally came into contact with B/F cache code :-( :-(

[04:07:31.0000] <hsivonen>
let's see if Hixie has conceptualized the B/F cache better than the Gecko code that introduced it...

[04:11:13.0000] <hsivonen>
I wonder if Gecko ever puts non-"complete" docs in the B/F cache in the first place...

[04:11:23.0000] <hsivonen>
smaug____: do you happen to know?

[04:12:11.0000] <smaug____>
it shouldn't

[04:12:22.0000] <hsivonen>
smaug____: ok. thanks

[04:21:52.0000] <MikeSmith>
hsivonen: btw it would be nice to try to fix the the parser bug that causes double-quoted charset parameter values to be reported as errors in the validator

[04:22:21.0000] <MikeSmith>
I can't remember where the source of that is but I vaguely recall being in the chardet code maybe

[04:23:04.0000] <hsivonen>
MikeSmith: does the spec not treat those as errors?

[04:23:21.0000] <MikeSmith>
no, not as far as I know

[04:23:33.0000] <hsivonen>
interesting

[04:24:15.0000] <MikeSmith>
I don't know that the HTML spec says anything about the syntax for the header case

[04:24:37.0000] <MikeSmith>
anyway, the mime-type RFCs allow it

[04:24:52.0000] <hsivonen>
oh this is about header parsing rather than processing the value after header parsing?

[04:24:54.0000] <MikeSmith>
though only for double quotes, not for single once

[04:25:02.0000] <MikeSmith>
yeah

[04:25:50.0000] <MikeSmith>
one common place where it gets reported is for the built-in Web server that ships with mercurial

[04:25:52.0000] <MikeSmith>
hgweb

[04:26:07.0000] <hsivonen>
ok

[04:36:04.0000] <MikeSmith>
hsivonen: related mercurial bug I filed: http://bz.selenic.com/show_bug.cgi?id=3531

[06:18:04.0000] <odinho>
Ms2ger: Problem is ofc that they're really so damn verbose. Should be a way to say "I want to see the important stuff". So then it's possible to fix that first...

[06:18:10.0000] <odinho>
Ms2ger: (the idlharness that is)

[06:18:51.0000] <Ms2ger>
Well, all of idlharness is obscure edge cases, really :)

[06:20:16.0000] <odinho>
Nope, checking that properties exist, is readonly, returning expected objects back that behave as described. That stuff is useful for doing a quick skeleton of code to further refine and build on. :]

[06:20:49.0000] <odinho>
Well, depends on obsucrity level ofc, but not that obscure :P

[06:23:47.0000] <Ms2ger>
Easy; you just import the IDL file and let the bindings code do its wo... Oh

[06:24:24.0000] <odinho>
...

[06:24:58.0000] <Ms2ger>
/me curses Gecko's manual bindings

[06:26:10.0000] <odinho>
Hmm? Thought Gecko had a binding system?

[06:26:37.0000] <Ms2ger>
About 5

[06:35:29.0000] <Ms2ger>
And the hand-written parts don't look nice...

[06:52:29.0000] <annevk>
why does http://trac.webkit.org/browser/trunk/Source/WebCore/platform/KURL.cpp#L74 not list ports for ws: and wss:?

[06:52:35.0000] <annevk>
they're the same as for http?

[06:57:17.0000] <annevk>
indeed they are

[08:03:13.0000] <MikeSmith>
hsivonen: so I find the cause of the parser not handling double-quotes in the Content-Type header charset parameter is not in chardet but in the validator util code

[08:03:23.0000] <MikeSmith>
https://bitbucket.org/validator/util/src/de960161c5f1/src/nu/validator/xml/ContentTypeParser.java#cl-37

[08:03:37.0000] <MikeSmith>
private static final Pattern CHARSET = Pattern.compile("^\\s*charset\\s*=\\s*(\\S+)\\s*$");

[08:04:29.0000] <MikeSmith>
should be:

[08:04:32.0000] <MikeSmith>
 private static final Pattern CHARSET = Pattern.compile("^\\s*charset\\s*=\\s*\"(\\S+)\"\\s*$");

[08:04:37.0000] <MikeSmith>
instead I guess

[08:04:59.0000] <MikeSmith>
hmm are the spaces allowed there?

[08:05:11.0000] <MikeSmith>
/me goes back to read whatever RFC it is

[08:05:51.0000] <gsnedders>
Yes, they are.

[08:05:59.0000] <zewt>
heh, nothing is quite as horrid as double-escaped regexes in languages without a regex syntax

[08:06:31.0000] <Ms2ger>
perl comes close

[08:06:37.0000] <zewt>
it always does

[08:08:15.0000] <MikeSmith>
gsnedders: oh yeah?

[08:08:50.0000] <zewt>
who was it that objected to just "img.src = blob", again? was it just MS?

[08:09:02.0000] <zewt>
(because they seemed to object to anything that wasn't their own broken oneTimeOnly thing)

[08:10:42.0000] <Ms2ger>
I think that img.src = non-string is silly

[08:11:11.0000] <MikeSmith>
from productions in RFC 2616 at least I can't see that the spaces are allowed

[08:11:30.0000] <MikeSmith>
gsnedders: ↑

[08:11:38.0000] <MikeSmith>
I guess I must be missing something

[08:14:16.0000] <annevk>
zewt: it was sicking

[08:14:29.0000] <annevk>
zewt: primarily because of the CSS thing, which I don't think people will use

[08:14:55.0000] <zewt>
i'm personally on the fence (for whatever that's worth, not being a sicking myself)

[08:15:13.0000] <zewt>
having to add new entry points for every single api seems a lot bigger than changing the existing ones

[08:15:19.0000] <zewt>
but it is nicer

[08:17:06.0000] <zewt>
one thing that makes me nervous about it is that it makes it "optional" for APIs

[08:17:24.0000] <zewt>
some apis will inevitably go "why would you want to do that?" and not add the entry point, instead of it being a universal feature

[08:20:49.0000] <annevk>
that just requires a bit of hammering

[08:21:00.0000] <annevk>
but we can go with what we have now

[08:21:50.0000] <MikeSmith>
hsivonen: and would be better to be using a real parser there instead of regexp :) I'll file a TODO bug for it

[08:22:53.0000] <zewt>
annevk: well, modulo https://www.w3.org/Bugs/Public/show_bug.cgi?id=16790#c13

[08:26:00.0000] <zewt>
calling a black-box script that does a document.write("<script...") and having blob URLs in the caller be revoked would be bad

[08:47:43.0000] <dglazkov>
good morning, Whatwg!

[08:51:56.0000] <jwalden>
morning, dglazkov

[09:38:44.0000] <gsnedders>
MikeSmith: Implicit white space is what you're missing.

[09:39:08.0000] <gsnedders>
jwalden: So, we now outnumber these accessor property people. Mwhaha! :)

[09:39:27.0000] <jwalden>
:-)

[09:39:45.0000] <jwalden>
I landed the __proto__ patch about an hour ago, we'll see how it goes

[11:14:09.0000] <annevk>
hmm authority is annoying to parse

[11:14:53.0000] <annevk>
[bits[:bits]@]bits[:bits] is the grammar or some such

[11:15:01.0000] <annevk>
but then wildly more complicated

[11:50:55.0000] <Philip`>
Hmm, I got a "DMCA Removal Notice - Link Removal Request" about "an unauthorised link to our clients web site" on http://philip.html5.org/data/doctypes.html

[11:51:33.0000] <Philip`>
annevk: Since you're hosting it and would presumably have to deal with any problems if they're serious: Should I care?

[11:53:08.0000] <zewt>
Philip`: it's it a felony to send bogus DMCA requests or something? heh

[11:53:56.0000] <Philip`>
zewt: I guess that only applies to actual take-down notices, not to emails which happen to mention "DMCA" in their subject line

[11:54:20.0000] <zewt>
"something something DMCA - do whatever I tell you to"

[11:54:28.0000] <zewt>
Philip`: DMCA Removal Notice - buy me a pizza

[11:54:37.0000] <Philip`>
This one says "If the links are not removed as soon as possible then we may have to take further action such as requesting a DMCA take-down notice or reporting the links to Google"

[11:55:03.0000] <Philip`>
(I'd have thought Google was already aware of all my links, since that's what their business is based on...)

[11:55:52.0000] <zewt>
"you're going to tell a search engine about my webpage? please do!"

[12:01:00.0000] <Ms2ger>
Oh, look

[12:01:22.0000] <Ms2ger>
First action as a CSSOM editor: "Hey, do you *really* want me to *define* things?"

[12:20:10.0000] <MikeSmith>
gsnedders: I don't think I'm missing implicit white space as far the the RFCs go

[12:20:46.0000] <MikeSmith>
there's nothing implicit about it in RFC 2616

[12:21:10.0000] <MikeSmith>
rather it's just simply not allowed

[12:21:44.0000] <MikeSmith>
if browser engines allow it, then that should be specified somewhere

[12:29:50.0000] <annevk>
Ms2ger: oh, something is fixed?

[12:30:42.0000] <Ms2ger>
Dunno

[12:30:45.0000] <annevk>
Philip`: sounds like you can just ignore it

[12:33:00.0000] <karlcow>
Philip`: just ignore.

[12:57:01.0000] <Hixie>
hober: your quick review of the HD stuff would be nice, to make sure i'm not barking up the wrong tree

[12:57:04.0000] <Hixie>
MikeSmith: ping

[13:02:35.0000] <annevk>
am I reading it correctly that IRIs don't allow non-BMP characters?

[13:03:31.0000] <annevk>
http://tools.ietf.org/html/rfc3987#page-8

[13:03:41.0000] <annevk>
oh well, not that it matters what that spec says

[13:04:26.0000] <zewt>
an established tradition with rfcs

[13:15:54.0000] <annevk>
for URL parsing it's actually easier if the internal URL object stores path segments and the filename

[13:16:25.0000] <annevk>
so when you resolve "../" against something else, you just copy the path segments from the base URL (along with everything else, but not the filename and such) and just pop the topmost one

[13:16:46.0000] <annevk>
(after you tokenize "../" per path tokenization rules of course)

[13:17:30.0000] <annevk>
(and only if these are hierarchical schemes)

[13:18:25.0000] <annevk>
and then .pathname is just the concat of path segments with / followed by / followed by filename

[13:19:55.0000] <Tabatkins>
Philip`: There's no such thing as an unauthorized link.  You can ignore that.

[13:28:52.0000] <karlcow>
annevk: what is the filename?

[13:29:16.0000] <annevk>
karlcow: remainder except for query/fragment

[13:29:45.0000] <karlcow>
http://example.org/path/foo/morefoo?blah

[13:29:51.0000] <annevk>
morefoo

[13:30:11.0000] <annevk>
path segments would be an array of path and foo

[13:30:19.0000] <karlcow>
In which context it is a filename?

[13:30:23.0000] <annevk>
query would be blah

[13:30:31.0000] <annevk>
it's just a term

[13:31:16.0000] <karlcow>
/me scratches his head

[13:32:51.0000] <annevk>
yeah, the URL related terms are not too great

[13:33:05.0000] <annevk>
but we are kind of stuck with them given the API

[13:33:09.0000] <karlcow>
the path is "/path/foo/morefoo"

[13:33:32.0000] <annevk>
though I'm somewhat hesitant to call fragment hash instead

[13:33:37.0000] <annevk>
and query search...

[13:36:37.0000] <Hixie>
it's often tempting in writing these specs to just call everything "opaque term #391" and so forth

[13:36:50.0000] <Tabatkins>
But then you need a registry...

[13:36:53.0000] <Hixie>
just to shortcircuit the confusion and bikeshed bickering that inevitably results

[13:37:07.0000] <Hixie>
Tabatkins: "html spec opaque term #391"

[13:37:15.0000] <Tabatkins>
Ah, distributed extensibility.

[13:38:53.0000] <annevk>
Hixie: shortcircuit the confusion with terms like that? interesting universe

[13:39:07.0000] <Hixie>
yeah the problem is it introduces a whole different confusion

[13:39:13.0000] <Hixie>
where everyone has no idea what anything means

[13:39:25.0000] <Hixie>
but seriously, people sometimes get so hung up on the terms and ignore their definitions, it drives me crazy

[13:39:41.0000] <annevk>
uhuh

[13:39:48.0000] <Hixie>
e.g. the "cookie-free document" thing recently, which was very clearly defined as meaning "document that cannot get cookies", not "document that has no cookies"

[13:39:58.0000] <Hixie>
yet people wrote code assuming it meant the latter!

[13:40:08.0000] <Hixie>
(i changed that to "cookie-averse documents")

[13:40:09.0000] <Tabatkins>
omg

[13:40:37.0000] <Hixie>
and the definition was like two sentences from where the offending requirement was

[13:40:38.0000] <Tabatkins>
I... want to refuse to believe that people would write code based solely on the name of a feature, rather than reading any of the actual spec.

[13:40:48.0000] <Hixie>
i wish

[13:41:21.0000] <Hixie>
anyway, that kind of thing is what makes me want to use really opaque terms so that there's no way to implement it without truly understanding it

[13:41:29.0000] <Hixie>
obviously a non-starter in practice

[13:42:31.0000] <Hixie>
ok, i've added removeHitRegion() for microsoft

[13:42:42.0000] <Hixie>
that's apple and microsoft dealt with

[13:43:09.0000] <Hixie>
opera wants data: in web works, mozilla wants... what was it, someone had a suggestion earlier

[13:43:53.0000] <Hixie>
inputmode, i think

[13:43:58.0000] <Hixie>
that's gonna require research

[13:44:31.0000] <karlcow>
/me now wonders what are the APIs in different std libs for programming languages.

[13:48:58.0000] <karlcow>
* for URIs

[14:54:54.0000] <Hixie>
ok... data: workers done

[14:55:03.0000] <Hixie>
that leaves autocompletetype for chrome and inputmode for mozilla

[14:56:47.0000] <annevk>
Hixie: how did you handle same-origin that redirects to a data: URL?

[14:57:35.0000] <smaug____>
really really odd, pages for mobile browsers seem to use event.srcElement reasonable often

[14:57:40.0000] <smaug____>
I wonder why

[14:57:48.0000] <smaug____>
srcElement is IEism

[14:57:53.0000] <smaug____>
not webkitism

[14:58:20.0000] <Hixie>
annevk: bummer. redirects.

[14:58:22.0000] <Hixie>
/me reopens the editor

[14:58:57.0000] <annevk>
I thought it was an easy fix for XHR too and then decided to pun on it when I realized redirects

[14:59:42.0000] <Hixie>
annevk: oh it's already dealt with. As soon as you have a redirect to a different origin, the fetch aborts.

[15:00:03.0000] <Hixie>
annevk: the way i specced it, it's only data: URLs given in the actual argument that are special-cases

[15:00:08.0000] <Hixie>
special-cased

[15:00:14.0000] <annevk>
I guess that's okay

[15:00:34.0000] <annevk>
the redirect case is kind of weird

[15:13:28.0000] <hober>
If you're an implementor and object to the changes described in http://www.w3.org/wiki/User:Mtanalin/legend-placement#Details it would be nice if you said so in the wbs poll on the matter before friday: https://www.w3.org/2002/09/wbs/40318/issue-200-objection-poll/

[15:15:58.0000] <Tabatkins>
Oh jesus.

[15:18:18.0000] <hober>
Tabatkins: exactly.

[15:32:07.0000] <Hixie>
adding some screenshots to http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control

[15:32:22.0000] <Hixie>
hober: why does it matter? surely if the wg just makes the spec say something dumb, it'll just be ignored

[15:35:42.0000] <hober>
Hixie: all things being equal, it'd be nice if the wg took its charter requirement to actively pursue convergence.

[15:35:49.0000] <hober>
s/took/took seriously/

[15:36:56.0000] <Hixie>
hober: that's up to the chairs

[15:37:10.0000] <Hixie>
hober: and the editors

[15:37:43.0000] <hober>
indeed, and they base their decisions on, among other things, the wbs survey results. having voices of reason in the survey results helps improve the outcome

[15:41:51.0000] <Hixie>
hober: if they took their charter requirement to actively pursue convergence seriously, they wouldn't base their decisions on surveys

[15:43:41.0000] <tantek>
/me scrolls up

[15:43:52.0000] <tantek>
nicely written objection TabAtkins

[15:44:00.0000] <tantek>
I'm not sure I could really add anything substantial to that

[15:44:28.0000] <tantek>
my big objection is that there is insufficient benefit shown to offset the cost of adding a new <ilegend> element - in fact, quite the opposite, it looks like language pollution.

[15:45:04.0000] <Tabatkins>
tantek: Simply adding that Mozilla objects for the same reason would be useful.

[15:45:08.0000] <tantek>
however, that's not really a new objection compared to what you wrote, so I'm not going to bother adding a "me too" answer since the chairs say they don't look at # of votes etc. but rather just the strength of arguments.

[15:45:16.0000] <Tabatkins>
While votes themselves aren't counted, weight of impl opinion is.

[15:45:23.0000] <tantek>
it shouldn't be an (un)popularity contest

[15:45:36.0000] <Tabatkins>
It should absolutely include popularity among implementors.

[15:45:47.0000] <Tabatkins>
Given that they're the ones implementing the features. ^_^

[15:45:54.0000] <Hixie>
it shouldn't be a contest at all

[15:45:57.0000] <tantek>
I'd rather see the chairs make a good rational decision based on your objection alone. I think that will show that strength of argument matters more than specific parties objecting.

[15:46:14.0000] <karlcow>
plus I would rather prefer having an element called <iamlegend> or even better <thelastmanonearth>

[15:46:24.0000] <Hixie>
the objections have long been discussed, if the chairs wanted to make rational decisions they could have done so months ago (when i did)

[15:46:30.0000] <Tabatkins>
Past experience suggests the chairs don't make good rational decisions on these things.

[15:46:35.0000] <Hixie>
(not to mention that in this case the problems are self-evident)

[15:46:51.0000] <tantek>
Tabatkins - my experience has been contrary - e.g. decisions on Issue 183, Issue 184. Have been quite rational.

[15:47:13.0000] <Hixie>
/me gave up hope on the chairs when they decided on a proposal that was literally self-contradictory

[15:47:24.0000] <Tabatkins>
I'm willing to believe they're not consistently irrational. ^_^

[15:47:31.0000] <tantek>
Hixie, baby/bathwater. No one is perfect.

[15:48:04.0000] <Hixie>
this isn't about lacking perfection

[15:48:19.0000] <tantek>
Anyway, I think the chairs will also be increasingly biased *against* new elements in an effort to get closure on the HTML5 feature set.

[15:48:34.0000] <Hixie>
it was clear in their decision that they didn't even know what they were deciding on -- the thing they pointed to had two different proposals, both internally inconsistent, and they hadn't even noticed

[15:48:49.0000] <Hixie>
and they continued to insist for something like 6 months, even after every single problem was discussed to death, that their decision was right

[15:48:59.0000] <tantek>
Hixie, not everyone is as good at spotting deep internal inconsistencies as you are :P

[15:49:09.0000] <Hixie>
this isn't "deep internal inconsistencies"

[15:49:23.0000] <Hixie>
this proposal wasn't even consistent at a shallow level

[15:49:33.0000] <tantek>
/me isn't even sure which issue/proposal(s) we're talking about, but suspects it has something to do with accessibility.

[15:49:42.0000] <Hixie>
some canvas thing last year

[15:56:09.0000] <tantek>
quick question for those of you that went to TPAC2012 and are considering going to TPAC2013:

[15:57:12.0000] <gavinc>
... isn't TPAC2012 in October?

[15:57:26.0000] <Tabatkins>
Very end of Oct/beginning of Nov.

[15:57:31.0000] <tantek>
make that went to TPAC2011 and are considering going to TPAC2012 :)

[15:57:39.0000] <Tabatkins>
Oh, got it.

[15:57:50.0000] <Tabatkins>
/me missed the implication of the question. ^_^

[15:57:59.0000] <tantek>
Did you get value out of the TPAC2011 pre-scheduled morning sessions (before the BarCamp-style sessions), or not? (e.g. lots, some value, none, negative, walked out) And if you did get value, any suggestions for "global" topics for TPAC2012?

[15:58:16.0000] <tantek>
(that's the question, before was just setup)

[15:58:30.0000] <tantek>
(totally unscientific poll)

[15:58:39.0000] <Tabatkins>
I don't remember what they were, so... apparently not.

[15:58:52.0000] <Tabatkins>
I think fantasai and I skipped them to work on the slides for out later session.

[15:58:54.0000] <othermaciej>
do you mean on the tech plenary day?

[15:59:03.0000] <tantek>
othermaciej yes

[15:59:12.0000] <tantek>
Tabatkins I'

[15:59:22.0000] <tantek>
I'll take that as "none" for you and fantasai then

[15:59:53.0000] <gavinc>
/me got no value what so ever as the WGs he's part of didn't go to TPAC last year, and are going this year but now the TPAC is in Europe and can't get funding and there is no remote on the TPAC site this year :( I am a sad panda

[16:00:00.0000] <Tabatkins>
More of an "abstain", but close enough to be "none".

[16:00:14.0000] <Tabatkins>
gavinc: I'll take you in my baggage.

[16:01:52.0000] <tantek>
more of a "decided to not go" than "abstain" which if anything is a vote against based on pre-evaluation.

[16:02:10.0000] <smaug____>
uh, almost forgot tpac... more traveling this year

[16:02:15.0000] <hober>
IIRC the unconference part of the day was worthwhile

[16:02:24.0000] <hober>
but I don't remember the non-unconference part.

[16:02:31.0000] <Tabatkins>
tantek: Yeah, but we had things to do, and also the weather was *gorgeous* outside.

[16:02:31.0000] <Hixie>
i got a lot of OpenTTD done that week

[16:02:33.0000] <tantek>
hober - I'll take "don't remember" as a "none"

[16:03:16.0000] <Hixie>
other than that, pretty much every meeting i was in that i thought was productive (not many to start with) ended up being a no-op because the decisions made there were reversed or abandoned within a month or two

[16:03:29.0000] <Hixie>
so i'd have to go with "negative"

[16:03:42.0000] <Tabatkins>
Hixie: He's not polling about WG meetings.

[16:03:49.0000] <Hixie>
i didn't go to any wg meetings

[16:03:54.0000] <Tabatkins>
Oh, ok then.

[16:04:07.0000] <Hixie>
those i _knew_ are a waste of time :-)

[16:05:16.0000] <tantek>
Hixie - I thought I saw you in the HTMLWG meeting - or was I hallucinating that?

[16:05:25.0000] <Hixie>
oh yeah there was that tv-related thing

[16:05:31.0000] <Hixie>
forgot that that was a wg meeting

[16:05:37.0000] <tantek>
I have a vague memory of time element discussions that were fairly productive

[16:06:10.0000] <tantek>
but some of that was also in informal lobby conversations as well

[16:06:11.0000] <Hixie>
the time element discussions were productive in that you and i spent 5 minutes at lunch time and that resulted in the whatwg spec being updated to the right thing, and the htmlwg still hasn't figured out the resolution to that topic.

[16:06:24.0000] <tantek>
either way - it was good to see you at TPAC even if for only a day or two.

[16:06:45.0000] <Hixie>
last year i was thinking of doing a "whatwg team building day" on the wednesday

[16:06:55.0000] <Hixie>
but the w3c and hsivonen convinced me not to

[16:06:56.0000] <tantek>
Hixie - htmlwg has resolved on both issue 183 and 184 AFAIK, and consistent with the whatwg spec changes/resolutions.

[16:07:03.0000] <Tabatkins>
Dammit, I wanted to do that.

[16:07:04.0000] <tantek>
(those are the time/data issues)

[16:07:21.0000] <Hixie>
tempted to try again in 2013 (assuming 2012 is in europe)

[16:07:36.0000] <tantek>
what would be a whatwg team building day? like go rock climbing somewhere together?

[16:07:41.0000] <Hixie>
tantek: i thought there were still some oustanding, but ok, i can't claim to have looked closely

[16:07:56.0000] <Hixie>
my idea was to rent a bunch of buses and take everyone down to the exploratorium

[16:08:00.0000] <tantek>
Hixie, there were, but there's been progress today on both issues.

[16:08:02.0000] <Hixie>
and do some sort of puzzle hunt there

[16:08:08.0000] <Hixie>
"progress"

[16:08:15.0000] <Hixie>
c.f. the whatwg spec resolved this last year

[16:08:29.0000] <tantek>
"progress" meaning actual HTMLWG decisions and issue closures.

[16:08:49.0000] <Hixie>
pardon me for having stopped caring entirely :-)

[16:08:54.0000] <Hixie>
(about the htmlwg)

[16:09:05.0000] <tantek>
Hixie, I realize the delay in changes/decisions in comparison. I decided to in good faith attempt to pursue the issue resolution path in HTMLWG.

[16:10:38.0000] <Hixie>
anyway, yeah, if the tpac is back in the bay area some time i'll look into doing some sort of proper team-building day again

[16:10:46.0000] <tantek>
I expect so in 2013

[16:10:49.0000] <Hixie>
right

[16:11:07.0000] <hober>
sounds fun

[16:11:25.0000] <tantek>
BTW I think they're constructing a new Exploratorium on the bay (forgot which Pier #) and moving everything.

[16:11:31.0000] <Hixie>
oh neat!

[16:11:35.0000] <tantek>
not sure if that will be done or not by 2013 TPAC

[16:12:00.0000] <tantek>
http://www.exploratorium.edu/piers/

[16:12:07.0000] <Hixie>
they say "in 2013"

[16:12:09.0000] <tantek>
Pier 15 apparently

[16:12:16.0000] <Hixie>
spring 2013!

[16:12:19.0000] <Hixie>
dude that's awesome

[16:12:25.0000] <Hixie>
we should definitely do that

[16:12:49.0000] <Hixie>
anyway, that's like 18 months away

[16:13:07.0000] <Hixie>
afk, bbiab

[16:14:37.0000] <tantek>
couldn't find a timeline, but yeah spring 2013 sounds good. also this is cool: http://www.exploratorium.edu/piers/creatingcampus.php?show=inside


2012-07-18
[17:10:51.0000] <Tabatkins>
Hixie: Regarding your suggestion about hanging the source of the template off of a string property, do you want the actual original text, or is a reserialization sufficient?

[18:12:42.0000] <Hixie>
Tabatkins: can you elaborate on your question?

[18:40:19.0000] <volkmar>
Hixie: ping

[19:47:17.0000] <Hixie_>
volkmar: here briefly

[19:48:25.0000] <volkmar>
Hixie: about that autocompletetype

[19:48:39.0000] <volkmar>
could you wait until the end of the month for my feedback?

[19:48:43.0000] <volkmar>
I will not have the time this week

[19:48:48.0000] <volkmar>
maybe not next week

[19:48:57.0000] <volkmar>
but after, I will be on vacation so time I should find ;)

[19:49:14.0000] <Hixie>
can you give me a synopsis of your feedback?

[19:49:22.0000] <Hixie>
like, a one line "elevator pitch"?

[19:52:16.0000] <volkmar>
Hixie: basically, my feeling is that we should be able to use cleverly @type instead of @autocompletetype

[19:52:35.0000] <volkmar>
a few autocompletetype values are already @type values

[19:52:43.0000] <volkmar>
and we could extend @type instead of having that new thing

[19:52:53.0000] <Hixie>
i thought type="" was for the data type and autocompletetype="" was for the field name?

[19:53:26.0000] <Hixie>
like, type would be something like "integer" or "string", and autocomplete type something like "x" and "y", or "name" and "description"

[19:56:20.0000] <Hixie>
dglazkov: in xbl, there were different events with event objects retargetting too (there had to be, since we sometimes crossed security boundaries)

[19:58:51.0000] <zewt>
i'd sooner not have @type expanded for things that use the same widget; I wish type=password had just been <input type=text datatype=password> or something like that, since it's less special casing when I just want to hook "all text-input-box <input>s"

[19:59:35.0000] <zewt>
having type=name or something like that means every case where I want to do that, I have to add "|| e.target.type == 'name'", and so on (if I'm following right, which I may not be since I'm preoccupied)

[19:59:55.0000] <Hixie>
zewt: search and text are more similar than text and password, at least text and password are slightly different controls. search is barely more than the same control with a stylistic change.

[20:02:53.0000] <zewt>
yeah, often a completely script-invisible one too (like changing onscreen keyboard layouts)

[20:08:30.0000] <zewt>
sigh, another TR link hits the archives, lurking in wait for somebody to trip up

[21:08:18.0000] <volkmar>
Hixie: IIRC, autocompletetype wasn't what you describe but I should have another look

[21:09:13.0000] <Hixie>
the thing i'm talking about is for the autofill feature; credit card numbers, addresses, etc

[21:09:24.0000] <volkmar>
yes

[21:10:32.0000] <Hixie>
well e.g. for "credit card expiry date", you'd want type=month, but autocompletetype="credit card expiry date"; but for "anniversary" you'd want type=month but autocompletetype="wedding day" or some such

[21:10:35.0000] <volkmar>
Hixie: fwiw, the example of the official blog post is something like "<input type=”text” name=”field1” x-autocompletetype=”email” />"

[21:10:46.0000] <Hixie>
"official"?

[21:10:55.0000] <Hixie>
i dunno about any blog post :-)

[21:10:58.0000] <volkmar>
from google blog post

[21:11:02.0000] <Hixie>
ah

[21:11:15.0000] <volkmar>
but i'm not sure about thta autocompletetype="credit card expiry date"

[21:11:26.0000] <volkmar>
i thought the idea was to have a defined list

[21:11:29.0000] <Hixie>
i don't know what the value would actually be

[21:11:42.0000] <volkmar>
http://wiki.whatwg.org/wiki/Autocompletetype

[21:12:57.0000] <volkmar>
oh

[21:13:03.0000] <volkmar>
i actually replied to the whatwg thread

[21:14:25.0000] <Hixie>
cc-exp, right

[21:14:42.0000] <volkmar>
Hixie: http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jan/0193.html

[21:14:51.0000] <Hixie>
note 4 pretty much summarises what i was trying to say above

[21:15:48.0000] <Hixie>
i don't understand how it would work

[21:15:55.0000] <Hixie>
type=tel just says "this is a kind of phone number"

[21:16:08.0000] <Hixie>
it doesn't say whether it's a home phone, work phone, fax number, voicemail number, ...

[21:17:17.0000] <volkmar>
yes, indeed

[21:18:13.0000] <volkmar>
for those information we might want to use stuff like contacts api

[21:19:23.0000] <Hixie>
the advantage of the proposed solution is that it paves a very well-travelled cowpath

[21:19:29.0000] <volkmar>
also, do we have any idea of web dev willingness to use that attribute?

[21:20:24.0000] <Hixie>
chrome guys tell me they have some uptake, but i don't know of any first-hand

[21:22:19.0000] <volkmar>
I think <input type='address'> and <input type='contact'> could solve quite a lot of what they do but the chance to see that happening soon are null

[21:22:30.0000] <volkmar>
however, I don't feel that this feature is very urgent

[21:23:51.0000] <Hixie>
i'm currently specifically going through features that vendors don't think are urgent except for one vendor, this is just the one (of many chrome asked for) that chrome is getting :-)

[21:24:00.0000] <Hixie>
mozilla is getting inputmode

[21:25:02.0000] <volkmar>
inputmode use cases are quite clear

[21:25:10.0000] <Hixie>
so are autofill use cases :-)

[21:25:23.0000] <Hixie>
i mean, this is stuff browsers have done since the dawn of time

[21:25:24.0000] <volkmar>
there might be other ways to solve this but I'm sure web devs want a feature like that

[21:27:05.0000] <volkmar>
Hixie: sure but browsers are already kind of solving the problem

[21:27:27.0000] <volkmar>
seems like the main use case of that proposal is for website selling stuff

[21:27:42.0000] <Hixie>
i'm not sure what you're arguing for or against

[21:27:48.0000] <Hixie>
but i encourage you to post on the thread

[21:28:05.0000] <volkmar>
my first reply has been ignored

[21:28:10.0000] <volkmar>
i can write another one

[21:28:19.0000] <volkmar>
but not before 10 days

[21:28:27.0000] <Hixie>
i won't ignore your previous mail, i'm sure it's still on my pile

[21:28:44.0000] <Hixie>
i don't think it makes sense though :-)

[21:30:13.0000] <volkmar>
that would be a good reason to ignore it actually

[21:30:34.0000] <Hixie>
"ignore" would mean that i don't even bother reading it or considering it

[21:30:50.0000] <volkmar>
actually, I wasn't speaking about you ignoring it

[21:31:04.0000] <volkmar>
but the person who sent the proposal

[21:31:14.0000] <volkmar>
but maybe he/she didn't reply because it was just non-sense

[01:06:52.0000] <annevk>
Hixie: do you think it's better to add an entry for data URLs and XMLHttpRequest with origins to your origin table or should I add an equivalent entry to XMLHttpRequest?

[01:09:35.0000] <Hixie>
dunno right now. send mail or file bug. probably best to have it all in one place though (and we sohuld remove the duplicate table from the origin spec)

[01:10:51.0000] <annevk>
that prolly means removing that spec from the IETF as updating RFCs is hard

[01:11:14.0000] <annevk>
I tried filing errata against it and they just wontfixed it without even taking it up in the WG...

[01:25:11.0000] <annevk>
also holy batman, 67 bugmails

[01:26:10.0000] <djdingo14>
Will HTTP 2.0 come out tomorrow with support in all browsers and support directly attaching your bank account to some kind of global ID which you can use to transfer and receive any amount of money anonymously for $0.01 per transaction regardless of the fee instantly without any hassle which will open up fantastic new abilities to finally monetize good content without relying on awkward scams such as PayPal?

[01:26:13.0000] <djdingo14>
Because that would be some news that would actually get me excited, for the first time in about a decade.

[01:31:53.0000] <annevk>
go outside, life a little ;)

[01:31:56.0000] <annevk>
live*

[01:36:08.0000] <annevk>
MikeSmith: you around?

[01:36:23.0000] <MikeSmith>
yeah

[01:36:37.0000] <annevk>
MikeSmith: we had the idea to make a new product, named WHATWG or WHATCG

[01:37:06.0000] <annevk>
MikeSmith: it could be "Web platform (other)" renamed

[01:37:48.0000] <annevk>
MikeSmith: and then move Fullscreen/Encoding/HTML there

[01:38:08.0000] <annevk>
MikeSmith: the HTML that is currently "other Hixie drafts (editor: Ian Hickson)"

[01:38:30.0000] <MikeSmith>
sounds fine to me

[01:39:07.0000] <MikeSmith>
I can just change the name of the "Web Platform (other)" one

[01:39:44.0000] <djdingo14>
annevk:  I am outside quite often. What the hell's your point?

[01:40:35.0000] <MikeSmith>
annevk: what should be put for the description? "Deliverables of the WHATCG"?

[01:40:42.0000] <annevk>
djdingo14: don't try to get excitement from things you don't influence

[01:40:58.0000] <annevk>
MikeSmith: WHATWG standards?

[01:41:02.0000] <annevk>
MikeSmith: or WHATWG specifications

[01:44:28.0000] <MikeSmith>
annevk: ok

[01:44:35.0000] <MikeSmith>
I made the component name WHATCG

[01:45:00.0000] <MikeSmith>
https://www.w3.org/Bugs/Public/describecomponents.cgi?product=WHATCG

[01:46:02.0000] <annevk>
MikeSmith: hmm Fullscreen is also under WebAppsWG?

[01:46:20.0000] <MikeSmith>
yeah

[01:48:51.0000] <annevk>
MikeSmith: is it easy to move a component?

[01:49:04.0000] <annevk>
MikeSmith: otherwise just keep it under WebApps and remove the one under WHATCG

[01:50:44.0000] <MikeSmith>
annevk: OK I'll just remove the one under WHATCG for now, since there are no bugs there anyway

[01:57:02.0000] <annevk>
okay

[01:57:50.0000] <annevk>
MikeSmith: if you can, coordinate with Hixie when to move "other Hixie drafts (editor: Ian Hickson)" over as "HTML" so he can update his scripts afterwards

[01:58:14.0000] <MikeSmith>
annevk: yeah will do

[01:58:30.0000] <MikeSmith>
have to wait til tomorrow though

[01:59:39.0000] <annevk>
there's no rush

[02:26:54.0000] <hsivonen>
argh. I made a mistake in my poll answer

[02:27:14.0000] <hsivonen>
has the ability to revise been removed or is this the poll-systems bogus cache control again?

[02:29:48.0000] <hsivonen>
it's the cacheability

[02:30:06.0000] <hsivonen>
and when I read my comment again, it seems I already said what I meant

[02:32:41.0000] <annevk>
still caring about the HTML WG?

[02:50:46.0000] <hsivonen>
annevk: yeah. it turns out that forming an informed opinion for responding to a poll is always a time sink

[02:50:58.0000] <hsivonen>
I tend to regret paying attention to the polls

[02:51:49.0000] <annevk>
I stopped caring after the last meeting and got chaals to remove me from the group

[02:51:58.0000] <annevk>
Seems Hixie is pretty much out now too

[03:27:42.0000] <MikeSmith>
hsivonen: about the validator not handling double quotes in charset param, is it OK with you if i check in the change described at http://krijnhoetmer.nl/irc-logs/whatwg/20120717#l-783 ?

[03:27:58.0000] <MikeSmith>
ah whoops

[03:28:05.0000] <MikeSmith>
need to make the quotes optional of course

[03:34:56.0000] <MikeSmith>
hsivonen: Pattern CHARSET = Pattern.compile("^\\s*charset\\s*=\\s*\"?(\\S+)\"?\\s*$");

[03:38:50.0000] <MikeSmith>
damn that ain't going to work

[03:40:45.0000] <hsivonen>
MikeSmith: did you intentionally not enforce the string having both start and end quote or neither?

[03:41:02.0000] <MikeSmith>
no not intentionally

[03:41:20.0000] <MikeSmith>
could deal with it by using two patterns

[03:41:45.0000] <MikeSmith>
one for quoted charset param and one for unquoted

[03:41:52.0000] <MikeSmith>
and check for the quoted case first

[03:42:07.0000] <MikeSmith>
does that sound reasonable?

[03:43:21.0000] <hsivonen>
MikeSmith: isn't it possible to bake both into one pattern?

[03:43:38.0000] <hsivonen>
might be slightly unpleasant with numbering of the captures, but should work otherwise

[03:44:24.0000] <hsivonen>
do we have a spec for parsing this stuff in non-compliant cases, yet?

[03:44:38.0000] <MikeSmith>
none that I know of

[03:44:40.0000] <hsivonen>
i.e. do we really want to insist on both quotes or no quotes?

[03:45:25.0000] <hsivonen>
anyway, baking both cases into one pattern should be marginally more performant than having two patterns

[03:45:27.0000] <MikeSmith>
yeah I had another question about that, which is, those spaces around the "=" are not conforming as far as RFC 2616 says

[03:45:32.0000] <MikeSmith>
OK

[03:45:50.0000] <MikeSmith>
so but how do I have a regex that says \S+ but not \" ?

[03:46:35.0000] <hsivonen>
MikeSmith: [^\"\s]+ IIRC

[03:47:12.0000] <hsivonen>
if \s inside [...] works

[03:47:21.0000] <MikeSmith>
yeah I think it does

[03:55:30.0000] <annevk>
MikeSmith: spaces are conforming

[03:55:42.0000] <annevk>
MikeSmith: RFC 2616 has a notion of "implicit whitespace"

[03:55:46.0000] <annevk>
MikeSmith: it's fucked up

[03:55:53.0000] <MikeSmith>
crazy train

[03:56:09.0000] <MikeSmith>
magic whitespace

[03:56:23.0000] <annevk>
yeah, between all tokens unless noted otherwise

[03:56:35.0000] <annevk>
was too hard to write their crazy BNF otherwise

[04:01:45.0000] <MikeSmith>
BNF is a really great way to obscure clarity

[04:02:45.0000] <karlcow>
/me notes that BNF is Bibliothèque Nationale de France… and that there is indeed some truth into that statement :p

[04:08:25.0000] <Ms2ger>
Like the CSSWG-BNF, then

[04:09:00.0000] <MikeSmith>
hsivonen: does https://gist.github.com/3135601 look sane?

[04:09:20.0000] <hsivonen>
what's "operation convergence"?

[04:10:09.0000] <MikeSmith>
string to make the bugs identifiable so people can filter bugmail

[04:10:16.0000] <hsivonen>
MikeSmith: yes, except the parentheses around the quoted variant should be non-capturing

[04:10:19.0000] <MikeSmith>
because of the mailbomb we sent

[04:10:31.0000] <Ms2ger>
hsivonen, Hixie was requested to stop updating the HTMLWG spec, so he's duplicating the bugs to get work done on HTML

[04:10:46.0000] <hsivonen>
i.e. the parentheses inside the quotes need to be capturing but the ones around don't

[04:10:55.0000] <hsivonen>
Ms2ger: I see

[04:11:30.0000] <MikeSmith>
hsivonen: ah yeah

[04:12:50.0000] <hsivonen>
oh and the unquoted variant should probably use [^\"\\s]+ as well

[04:12:57.0000] <hsivonen>
otherwise it will capture lone quotes

[04:13:22.0000] <MikeSmith>
ok

[04:13:28.0000] <MikeSmith>
/me updates it

[04:13:40.0000] <hsivonen>
thanks

[04:15:48.0000] <MikeSmith>
hsivonen: updated patch at https://gist.github.com/3135601

[04:17:16.0000] <hsivonen>
MikeSmith: r+

[04:17:27.0000] <MikeSmith>
thanks

[04:34:46.0000] <annevk>
validator.nu moved to github?

[04:35:15.0000] <MikeSmith>
no there's just a mirror there

[07:26:00.0000] <zewt>
https://www.w3.org/Bugs/Public/show_bug.cgi?id=16790 https://www.w3.org/Bugs/Public/show_bug.cgi?id=17988 ... what the hell?

[07:26:49.0000] <Ms2ger>
zewt, the former is to be fixed by the hypothetical HTML5 editor, the latter is actually going to be fixed by Hixie

[07:27:02.0000] <zewt>
uh

[07:27:24.0000] <zewt>
it's nothing but facking up the bug tracker, heh

[07:27:44.0000] <Ms2ger>
Thank the chairs for that :)

[07:30:59.0000] <zewt>
also the bug isn't CC'd on any lists now?

[07:33:52.0000] <zewt>
also all kinds of metadata wasn't copied over (blocks, priority)

[07:34:05.0000] <zewt>
okay whatever i'm going to work to do *my* job now :)

[07:42:10.0000] <jarek>
what was the reason for allowing units inside SVG documents?

[07:42:23.0000] <jarek>
I mean something like:

[07:42:25.0000] <jarek>
<rect x="0.5cm" y="0.5cm" width="2cm" height="1cm"/>

[07:43:15.0000] <jarek>
it really doesn't make sense considering the fact that any of the ancestor elements could have transform applied

[07:46:25.0000] <cheron>
Is the CORS standard not part of the whatwg HTML specification?

[07:53:55.0000] <hsivonen>
cheron: it's a separate spec

[07:56:13.0000] <cheron>
hsivonen: yes, from W3C, but i can't find it on WHATWG. Does that mean there is none?

[07:58:52.0000] <Ms2ger>
Indeed

[08:02:34.0000] <cheron>
ty

[08:05:48.0000] <MikeSmith>
zewt: doing what to the bug tracker?

[08:09:24.0000] <Hixie>
MikeSmith: we should call the new product "WHATWG", not "WHATCG", imho

[08:10:21.0000] <MikeSmith>
Hixie: that carries a higher risk of somebody kneejerk flaming about it

[08:10:33.0000] <Hixie>
the cg's shortname is "whatwg"

[08:11:09.0000] <Hixie>
i'm happy to take any heat from flames, redirect them to me :-) it just seems needlessly confusing to people filing bugs

[08:12:26.0000] <MikeSmith>
OK, changed the product name to "WHATWG (flame Hixie instead of MikeSmith)"

[08:12:32.0000] <Hixie>
hah

[08:12:37.0000] <MikeSmith>
:)

[08:12:52.0000] <MikeSmith>
OK, it's just plain WHATWG now

[08:13:45.0000] <annevk>
okay, I will update the one character in Encoding

[08:14:02.0000] <annevk>
maybe I should checkout quirks mode and fix it for simon too

[08:19:37.0000] <annevk>
done

[08:29:09.0000] <Hixie>
MikeSmith: thanks

[08:31:39.0000] <MikeSmith>
Hixie: btw I have multiple IMEs for Japanese input installed in my Android mobile

[08:31:50.0000] <MikeSmith>
will try to make time to upload some screenshots

[08:35:59.0000] <Hixie>
sweet

[08:39:37.0000] <Hixie>
MikeSmith: do you have access to logs for https://www.w3.org?

[08:40:20.0000] <MikeSmith>
Hixie: web-server logs you mean?

[08:40:24.0000] <Hixie>
yeah

[08:40:28.0000] <MikeSmith>
nope

[08:40:32.0000] <Hixie>
i can't work out for the life of me what some of these errors are

[08:40:33.0000] <MikeSmith>
at least I don't think I do

[08:41:01.0000] <MikeSmith>
it's several machines I guess you know

[08:41:06.0000] <Hixie>
yeah

[08:41:33.0000] <MikeSmith>
but I don't have shell access on any of those

[08:41:53.0000] <MikeSmith>
I can ask somebody who does

[08:42:26.0000] <Hixie>
nah, sok

[08:46:36.0000] <Hixie>
MikeSmith: ok, can we try to turn off the list bugmail again for a bit?

[08:49:51.0000] <MikeSmith>
yup

[08:49:54.0000] <MikeSmith>
gimme a minute

[08:51:02.0000] <MikeSmith>
ah man I still had it disabled it seems

[08:51:10.0000] <MikeSmith>
forgotted to re-enable it

[08:51:10.0000] <Hixie>
hah ok, np

[08:51:21.0000] <MikeSmith>
I blame it on too much breakfast beer

[08:54:40.0000] <MikeSmith>
8 years of fun so far to reach harmonic convergence

[08:55:16.0000] <MikeSmith>
or closing on 9 years now I guess

[08:56:01.0000] <MikeSmith>
/me puts Whatever Happened to the Teenage Dream on the turntable

[09:00:48.0000] <Hixie>
ok. all htmlwg bugs are now cloned to whatwg bugs. now trying again on the reverse operation.

[09:05:45.0000] <annevk>
apart from Opera no browser (maybe Internet Explorer?) seems to terminate authority on ; http://dvcs.w3.org/hg/url/raw-file/tip/Overview.html#authority-terminating-character

[09:05:53.0000] <annevk>
I wonder why that was there

[09:06:45.0000] <MikeSmith>
wha

[09:06:59.0000] <MikeSmith>
what is "terminate authority"?

[09:07:36.0000] <MikeSmith>
would be fun to work a "... with extreme prejudice in there"

[09:07:59.0000] <Hixie>
i've gotten "prejudice" into the specs i edit a few times :-)

[09:08:08.0000] <MikeSmith>
sweet

[09:08:13.0000] <MikeSmith>
let's have more of that please

[09:08:40.0000] <MikeSmith>
annevk: so btw I might be in Amsterdam in September and asked to do 3.5 hours of presentation on Web platform

[09:08:55.0000] <MikeSmith>
if you could help me out with presenting that would be be appreciated

[09:09:05.0000] <MikeSmith>
and/or anybody else you'd recommend

[09:09:40.0000] <MikeSmith>
I will definitely be elsewhere in Europe in September for other things

[09:11:32.0000] <annevk>
I'm around

[09:11:43.0000] <annevk>
sounds like fun

[09:15:04.0000] <MikeSmith>
annevk: I volunteered to visit some EU W3C offices for 10 days or so but it appears to be morphing into I'll end up spending the entire foreseeable future there once I arrive

[09:15:43.0000] <MikeSmith>
which is OK because it's not like I have anything else productive to do with my time

[09:16:13.0000] <Hixie>
heh

[09:16:56.0000] <MikeSmith>
"An authority terminating character is either a slash character, U+003F ("?"), U+0023 ("#"), or U+003B (";")."

[09:16:58.0000] <annevk>
MikeSmith: we'll make it fun ;)

[09:17:18.0000] <MikeSmith>
we always do :-)

[09:17:22.0000] <MikeSmith>
so why ";" ?

[09:17:31.0000] <annevk>
uhuh

[09:30:36.0000] <Ms2ger>
<glenn> can someone file a bug against cssom to record this?

[09:30:46.0000] <Ms2ger>
Yeah, filing bugs on your own spec is soooo hard...

[09:35:36.0000] <Hixie>
oh crap

[09:35:44.0000] <Hixie>
silly silent errors

[09:36:10.0000] <Hixie>
/me cloned a bunch of bugs without noticing the old bugs weren't getting annotated with the new bug id

[09:36:41.0000] <Hixie>
causing a bunch of dupes

[09:37:13.0000] <Hixie>
oh not so many dupes it seems

[09:37:15.0000] <Hixie>
that's lucky

[09:39:08.0000] <Hixie>
ok through some lucky fluke, no dupes

[09:39:17.0000] <Hixie>
still, need to add the annotations now

[09:42:49.0000] <Hixie>
oh, this problem is tractable. only five bugs.

[09:42:52.0000] <Hixie>
ok.

[09:43:41.0000] <Ms2ger>
Now, whether 'tis nobler to write a script to fix those five bugs or to fix them manually

[09:44:55.0000] <Hixie>
script

[09:50:08.0000] <Hixie>
wtf

[09:50:14.0000] <Hixie>
bugzilla is refusing my tokens

[09:50:43.0000] <MikeSmith>
Hixie: hitting rate limiting maybe?

[09:50:59.0000] <Hixie>
that seems to usually be 502s and 504s

[09:51:17.0000] <Hixie>
this is a 200, saying i should hit back because the token is bad

[09:51:29.0000] <Ms2ger>
/me wonders if the CSSWG is going to manage not to rename flexbox again

[09:52:04.0000] <Ms2ger>
I guess it's about 50/50

[09:52:10.0000] <MikeSmith>
nothing the CSSWG does will ever surprise me

[09:52:27.0000] <MikeSmith>
they have already achieved

[09:53:56.0000] <MikeSmith>
anyway I thought flexbox was dead in the water in the face of grid

[09:54:54.0000] <MikeSmith>
not that I know jack about either

[09:56:28.0000] <MikeSmith>
hmm "Unknowing Information Practices"

[09:56:34.0000] <MikeSmith>
interesting phrase

[10:06:14.0000] <dglazkov>
good morning, Whatwg!

[10:06:32.0000] <Ms2ger>
Good day

[10:07:49.0000] <Ms2ger>
Turns out the CSSWG, as usual, can't decide what to do and what reason to invent for the decision

[10:09:58.0000] <Hixie>
oh dear lord

[10:09:59.0000] <Hixie>
<-- idiot

[10:10:29.0000] <Ms2ger>
News at 11 :)

[10:58:07.0000] <Hixie>
almost done

[11:00:25.0000] <Hixie>
i wonder why there was a bug bump yesterday around the time i was doing the cloning

[11:00:43.0000] <MikeSmith>
?

[11:01:01.0000] <MikeSmith>
coincidence?

[11:01:05.0000] <Hixie>
i hope so

[11:01:28.0000] <Hixie>
in theory this bug cloning thing shouldn't have affected the number of bugs i have assigned to me, because each bug cloned was originally assigned to me, and as each is cloned, one of the two ends up not assigned to me

[11:01:36.0000] <Hixie>
but there's a bump on http://www.whatwg.org/issues/data.html?period=1

[11:01:41.0000] <MikeSmith>
ah

[11:01:46.0000] <Hixie>
not a big one

[11:02:15.0000] <MikeSmith>
that's got to be some artifact for sure

[11:02:27.0000] <Hixie>
went from 522 to 531 between 07:00 and 08:00

[11:02:34.0000] <Hixie>
which is exactly when we were doing it yesterday

[11:02:38.0000] <Hixie>
i think

[11:02:57.0000] <MikeSmith>
yeah

[11:04:16.0000] <MikeSmith>
consequence of doing violence against bugzilla and bugmail

[11:04:33.0000] <Hixie>
i don't really see how it could go up rather than down :-)

[11:04:44.0000] <Hixie>
but i guess that's at least not the destructive direction...

[11:04:51.0000] <MikeSmith>
yeah

[11:04:55.0000] <Hixie>
ok looks like it's done

[11:05:05.0000] <Hixie>
let me kust rerun the scripts to catch any stragglers

[11:05:08.0000] <MikeSmith>
sweet

[11:05:20.0000] <Hixie>
no stragglers

[11:05:21.0000] <Hixie>
sweet!

[11:05:24.0000] <Hixie>
ok cool

[11:05:39.0000] <Hixie>
MikeSmith: can you also change the default assignee for the htmlwg bugs to dave.null rather than me?

[11:05:51.0000] <MikeSmith>
I'll re-enable bugmail for the lists

[11:05:57.0000] <Hixie>
cool, thanks

[11:06:00.0000] <MikeSmith>
Hixie: yeah

[11:06:58.0000] <MikeSmith>
btw the secret plan is that I become the new HTML WG editor

[11:07:05.0000] <MikeSmith>
so that I get paid double

[11:07:19.0000] <Hixie>
i doubt the money is worth it :-)

[11:07:29.0000] <MikeSmith>
hah

[11:07:48.0000] <MikeSmith>
I'm waiting to see how that goes

[11:08:29.0000] <MikeSmith>
in unrelated news btw anybody raise your hand if you think basing an operating system is the DOM is a great idea

[11:08:57.0000] <Ms2ger>
volkmar, he's waiting for you

[11:10:52.0000] <MikeSmith>
"I hope that b2g is as grand a success as ChromsOS and WebOS have been so far"

[11:11:03.0000] <Ms2ger>
Ouch

[11:11:44.0000] <Hixie>
i can't quite get the times and numbers to line up but there does seem to have been a bunch of bugs filed in the last day

[11:11:56.0000] <Hixie>
so i'm going to assume timezones account for most of this

[11:12:37.0000] <MikeSmith>
Hixie: weird I don't think so many bugs were filed in the last day

[11:12:55.0000] <Hixie>
there were at least nine "Everything!!! Someone else in into my computer and controls it" bugs filed

[11:12:56.0000] <MikeSmith>
and I get Cc'ed on all of them

[11:13:06.0000] <MikeSmith>
ah yeah

[11:13:43.0000] <MikeSmith>
I dunno i have so much bugmail even normally that it's hard for me to tell I guess

[11:13:48.0000] <Hixie>
which, if the times on the three "you njust saved" bugs were not so close as to make the maths bogus, would account for the 522 to 531 increase

[11:14:04.0000] <MikeSmith>
geez

[11:14:15.0000] <MikeSmith>
that's nothing man

[11:14:31.0000] <Hixie>
the bug filed 12 seconds before called "i m searching action" also belies my explanation

[11:14:34.0000] <Hixie>
but anyway

[11:18:19.0000] <Hixie>
ok

[11:18:33.0000] <Hixie>
now i need to change the default QA on these bugs and reassign them all to the WHATWG product

[11:18:45.0000] <MikeSmith>
OK

[11:18:53.0000] <MikeSmith>
can do that

[11:19:07.0000] <Hixie>
oh i can do it, it's easy

[11:19:23.0000] <MikeSmith>
ok

[11:22:35.0000] <Hixie>
1075 bugs, hm

[11:22:43.0000] <Hixie>
this ought to be fun

[11:23:33.0000] <Hixie>
ah, no component yet

[11:23:44.0000] <Ms2ger>
Now they're all touched in the last few days... Maybe you could prioritize by bz being cc'd? :)

[11:23:50.0000] <Hixie>
hehe

[11:24:05.0000] <Hixie>
now that i don't have to worry about the escalation proces, i expect the closure rate will be much higher

[11:29:32.0000] <Hixie>
MikeSmith: can you create an "HTML" component in the WHATWG product, default assignee ian⊙hc, qa contributor⊙wo? pretty please :-)

[11:31:05.0000] <MikeSmith>
yah of course

[11:33:12.0000] <Hixie>
i love the name of the "Unwelcome" component

[11:33:19.0000] <MikeSmith>
hah

[11:33:43.0000] <MikeSmith>
yeah couldn't think of another good name for that

[11:34:00.0000] <Hixie>
"Behind the seat cushion"

[11:34:22.0000] <Hixie>
or "The crack", as in "it fell through the cracks", though people would interpret that wrongly :-P

[11:34:59.0000] <MikeSmith>
hahaha

[11:35:32.0000] <MikeSmith>
anyway for now "HTML" component created

[11:36:23.0000] <Hixie>
sweet

[11:36:34.0000] <Hixie>
ok, moving all the bugs over from "other Hixie drafts" now

[11:36:50.0000] <MikeSmith>
k

[11:37:00.0000] <Hixie>
...there's no way to do this without sending (me in particular) several thousand e-mails, is there

[11:37:39.0000] <MikeSmith>
me too

[11:37:54.0000] <Hixie>
indeed

[11:38:09.0000] <Hixie>
ok here goes

[11:39:03.0000] <MikeSmith>
/me waits to receive yet another mail bomb

[11:39:28.0000] <Hixie>
zewt already said i should have my decision making privileges revoked for all the spam he got

[11:39:55.0000] <MikeSmith>
hah

[11:40:40.0000] <MikeSmith>
so I guess he figures I should be stripped of decision-making privileges too

[11:40:49.0000] <MikeSmith>
which is fine by me me

[11:41:00.0000] <Hixie>
well he said whoever made the decision to do it should be

[11:41:04.0000] <Hixie>
so i think you may be safe

[11:41:14.0000] <MikeSmith>
ah OK

[11:41:16.0000] <MikeSmith>
damn

[11:41:38.0000] <MikeSmith>
I was going to suggest he contact my boss with a concrete proposal

[11:41:47.0000] <Hixie>
ok i updated my bug filing script and the links in the boilerplate for the whatwg specs

[11:41:52.0000] <Hixie>
wonder where else needs updating

[11:42:12.0000] <smaug____>
Hixie: so HTML bugs should go now to product:WHATWG?

[11:42:24.0000] <MikeSmith>
Hixie: dunno really other than that

[11:42:27.0000] <Hixie>
smaug____: yeah

[11:42:31.0000] <smaug____>
ok, good.

[11:42:31.0000] <Hixie>
smaug____: product WHATWG, component HTML

[11:43:05.0000] <Hixie>
smaug____: i'll do some PSAs in IRC channels soon, but feel free to start broadcasting the news :-)

[11:43:19.0000] <Hixie>
smaug____: e-mail is still fine, too, of course

[11:43:53.0000] <Ms2ger>
Whoa, why did my inbox just jump to 259?

[11:43:55.0000] <Ms2ger>
Oh, Hixie

[11:43:57.0000] <Hixie>
MikeSmith: it's about 30% done with the move, but the page itself has timed out so i only know this from doing queries :-)

[11:44:21.0000] <MikeSmith>
hmm  OK

[11:44:33.0000] <Hixie>
uh oh, looks like it stalled

[11:45:01.0000] <Hixie>
uh

[11:45:25.0000] <Hixie>
393 left in old component + 315 in new component != 1075 which we had before...

[11:45:47.0000] <Hixie>
that doesn't bode well

[11:46:55.0000] <Hixie>
oh, i see

[11:47:00.0000] <Hixie>
i was excluding some by mistake

[11:48:28.0000] <Hixie>
134 to go...

[11:48:34.0000] <Hixie>
88...

[11:48:40.0000] <Hixie>
oh no, it overwhelmed ms2ger!

[11:48:45.0000] <Hixie>
59...

[11:49:00.0000] <Hixie>
21...

[11:49:07.0000] <Hixie>
tis done!

[11:49:21.0000] <Hixie>
MikeSmith: can you delete the "other hixie drafts" component? and then i think we're done entirely with this

[11:49:33.0000] <Hixie>
/me informs the chairs

[14:18:26.0000] <Yuhong>
Hixie: I wonder whether the "HTML5 in 2022" talk inspired you to make HTML a living standard.

[14:23:01.0000] <Yuhong>
Wrote a blog article about the buzzword: http://yuhongbao.blogspot.com/2012/07/why-html5-buzzword-is-misnomer.html

[14:23:09.0000] <Yuhong>
And submitted this to Reddit and HN.

[14:33:13.0000] <Hixie>
Yuhong: it did not; it was the realisation that to finish in 2022 would mean not adding features for about a decade, which is simply a non-starter

[14:34:18.0000] <Yuhong>
Of course. Even HTML4 did not take that long.

[14:34:49.0000] <Hixie>
HTML4 took about 6 months, but they didn't do any of what we're doing really

[14:35:02.0000] <Hixie>
compared to the modern spec, HTML4 is more like a position paper

[14:36:24.0000] <Yuhong>
CSS 2.0 is even worse. It is unfortunate it took until after almost standards mode was created before the first draft of CSS 2.1 was published.

[14:37:43.0000] <Hixie>
same people, more or less

[14:42:01.0000] <Yuhong>
Which was after IE6 was released.

[14:44:47.0000] <Yuhong>
HN submission:

[14:44:48.0000] <Yuhong>
http://news.ycombinator.com/item?id=4238188

[14:45:17.0000] <Yuhong>
Reddit: http://www.reddit.com/r/web_design/comments/whdid/why_the_html5_buzzword_is_a_misnomer/

[14:48:23.0000] <Yuhong>
From that era: http://www.mozillazine.org/talkback.html?article=1524

[14:48:31.0000] <Yuhong>
"Is it just me, or is WaSP entirely composed of people who've never written anything other than HTML (and related standards)? Do they not realize that it takes a while to rewrite a browser from the ground up, when a community has 1/1000th the manpower of the competitor?"

[14:50:36.0000] <Yuhong>
Remember Netscape "Mariner"?

[14:55:52.0000] <annevk_>
https://www.w3.org/Bugs/Public/buglist.cgi?product=WHATWG&component=HTML&resolution=--- sweet!

[14:56:06.0000] <annevk_>
nice work Mike and Hixie!

[14:56:11.0000] <Hixie>
volkmar: any help on http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control would be awesome, i'm having great trouble learning anything about non-English IMEs and input modes

[14:56:17.0000] <Hixie>
annevk_: :-)

[14:56:40.0000] <Yuhong>
WaSP refers to Web Standards Project.

[14:56:43.0000] <Yuhong>
WaSP refers to Web Standards Project, of course.

[14:58:52.0000] <Yuhong>
Also see: http://www.google.com/url?sa=t&rct=j&q=netscape%20cannot%20win&source=web&cd=1&ved=0CFAQFjAA&url=http%3A%2F%2Fwww.mozillazine.org%2Ftalkback.html%3Farticle%3D1708&ei=mzEHUKWZAcelrQHX1r3bCA&usg=AFQjCNGQU0fG-3MvSFWRNwIMn7O4cEkSkQ&cad=rja

[14:58:58.0000] <Yuhong>
http://www.mozillazine.org/talkback.html?article=1708

[14:59:01.0000] <Yuhong>
Sorry.

[15:01:32.0000] <volkmar>
Hixie: do you want me to point to the list we thought about adding?

[15:04:52.0000] <annevk_>
volkmar: the use cases for the list

[15:08:20.0000] <volkmar>
sure

[15:11:16.0000] <volkmar>
annevk, Hixie : can one of you creates me an account on the wiki?

[15:11:28.0000] <annevk>
sure

[15:11:29.0000] <annevk>
email?

[15:11:37.0000] <volkmar>
mounir⊙lf

[15:12:21.0000] <annevk>
done

[15:13:30.0000] <Velmont>
Bah, I alwas write w3.org/Public/bugs when it' sthe other way around.

[15:13:55.0000] <annevk>
W3C URLs are confusing

[15:14:04.0000] <annevk>
but at least they don't change

[15:14:35.0000] <Velmont>
They're some cool URI's in that way ;-)

[15:21:49.0000] <volkmar>
annevk, Hixie: i added something to the wiki

[15:22:26.0000] <volkmar>
all of those are implementable on Androd

[15:22:29.0000] <volkmar>
*Android

[15:22:33.0000] <volkmar>
actually, I should push the patches...

[15:24:15.0000] <annevk>
Hixie: closed some bogus bugs

[16:33:20.0000] <Velmont>
Was is a good "dagger stop" to use is javascript? someone is using "z", but that's hardly the last letter in the alphabet. I guess I don't want \uffff but rather something not *too* strange (so it doesn't uncover stupid sleeping browser bugs :P)

[16:36:03.0000] <Velmont>
Hmmm, \uffff seems to work in Dragonfly at least. Will propose than then. :P Just hoping it isn't best practice ;]

[16:36:56.0000] <pablof>
Am i misunderstanding something or are the descriptions of setRangeText's selectionModes "select" and "preserve" switched? http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#textFieldSelection

[16:47:05.0000] <pablof>
Hixie: ^


2012-07-19
[18:02:18.0000] <gsnedders>
Hmm, is there any good way to mark up verse with using caesura and represent it (using CSS) as a long space?

[18:04:07.0000] <gsnedders>
*caesurae if I actually decline correctly

[18:10:22.0000] <gsnedders>
Most existing things seem to use &nbsp;, which isn't great.

[18:14:08.0000] <gsnedders>
(esp if you want to line up the half-lines, assuming that's how the caesura is used))

[19:00:34.0000] <Hixie>
volkmar: awesome, thanks. interesting that you are not looking at doing script-related inputmode control; is that not needed?

[19:02:05.0000] <Hixie>
pablof: they seem right, am i missing something?

[19:06:21.0000] <Hixie>
gsnedders: <span class="caesura">&#x1D113;</span> and hide the content with CSS? :-)

[19:06:34.0000] <Hixie>
but that's more a musical caesura

[19:13:39.0000] <gavinc>
Wrong place to ask, but does anyone know the best way to contact someone at the W3C about membership issue? (Email to membership unanswered for a while)

[19:50:17.0000] <Hixie>
gavinc: your AC rep, if you have one

[19:50:25.0000] <Hixie>
gavinc: failing that, try ij@ maybe?

[20:17:11.0000] <gavinc>
Hixie: Yeah, am the AC rep ;) trying to figure out why our membership didn't start on the first

[20:18:07.0000] <Hixie>
ah :-)

[20:18:24.0000] <Hixie>
try ian jacobs (ij@)

[20:18:40.0000] <Hixie>
when MikeSmith is around he can help too, probably by telling you to contact ij@ :-)

[20:19:07.0000] <MikeSmith>
gavinc: what Hixie said :)

[20:19:19.0000] <MikeSmith>
gavinc: I can check the DB at least

[20:37:19.0000] <MikeSmith>
Hixie: the script is done right?

[20:37:27.0000] <Hixie>
yup

[20:37:32.0000] <MikeSmith>
k

[20:37:50.0000] <Hixie>
there remains the issues i sent the chairs in the e-mail earlier (you were cc'ed)

[20:37:59.0000] <Hixie>
but i don't plan on doing anything on those unless they ask me to

[20:38:37.0000] <MikeSmith>
OK

[20:39:22.0000] <MikeSmith>
Hixie: you don't want any list Cc'ed on that component?

[21:09:34.0000] <Hixie>
not particularly

[21:20:42.0000] <volkmar>
Hixie: script-related input mode control?

[21:20:56.0000] <Hixie>
like, chinese simple vs traditional

[00:45:30.0000] <hsivonen>
so much "wrong tomorrow" with smart card Web logins

[00:46:25.0000] <hsivonen>
printed pad of passcodes each of which is used once will always be more compatible than any hardware-based solution

[00:46:55.0000] <hsivonen>
but even as far as hardware goes, USB connectors are more common now than smartcard slots

[01:27:45.0000] <annevk>
Is there a way to combine inputmode and autocompletetype somehow?

[01:28:06.0000] <annevk>
From the autocompletetype anyway the inputmode is often implied

[01:39:43.0000] <MikeSmith>
annevk: so it turns out the HTTP spec does specifically disallow whitespace in media-type parameters

[01:39:54.0000] <MikeSmith>
though it does allow it in other parameters

[01:39:55.0000] <MikeSmith>
http://tools.ietf.org/html/rfc2616#section-3.7

[01:40:17.0000] <AryehGregor>
annevk, I had a comment in that vein at some point, saying that there should be some higher-level attribute along the lines of "what type of input is expected" without specifically tying it to autocomplete or whatever.

[01:40:22.0000] <MikeSmith>
"Linear white space (LWS) MUST NOT be used between the type and subtype, nor between an attribute and its value."

[01:40:48.0000] <MikeSmith>
I see now that PhilipJ pointed this out in a list discussion last year

[01:40:52.0000] <annevk>
what is an attribute in that context?

[01:41:02.0000] <annevk>
they mean parameter?

[01:41:16.0000] <MikeSmith>
the first half of the parameter before the equal sign

[01:41:33.0000] <annevk>
ah okay

[01:41:36.0000] <MikeSmith>
the define the parameter as an attribute+value pair

[01:41:48.0000] <MikeSmith>
"Parameters MAY follow the type/subtype in the form of attribute/value pairs"

[01:41:51.0000] <annevk>
so charset = utf-8 would not work?

[01:42:02.0000] <MikeSmith>
yeah and it doesn't work actually

[01:42:06.0000] <MikeSmith>
in most browsers

[01:42:11.0000] <MikeSmith>
I think it works only in IE

[01:42:16.0000] <annevk>
really?

[01:42:20.0000] <MikeSmith>
other browsers ignore the parameter

[01:42:23.0000] <MikeSmith>
yeah try it

[01:42:47.0000] <MikeSmith>
you have to make sure you have charset-sniffing turned off in the browser

[01:43:00.0000] <MikeSmith>
and make sure it's not set to default to UTF-8

[01:43:05.0000] <MikeSmith>
Julian has a test case

[01:43:09.0000] <MikeSmith>
lemme get the URL

[01:43:37.0000] <MikeSmith>
http://greenbytes.de/tech/tc/httpcontenttype/textplainutf8ws.asis

[01:43:58.0000] <MikeSmith>
summary of results for various browsers is at http://greenbytes.de/tech/tc/httpcontenttype/#textplainutf8ws

[01:44:49.0000] <MikeSmith>
though when I try it in Chrome it does seem to use UTF-8 no matter what

[01:45:07.0000] <MikeSmith>
I can't tell whether it's getting it from the header or not

[01:45:29.0000] <MikeSmith>
but Opera and Firefox both ignore it at least

[01:47:32.0000] <MikeSmith>
hsivonen: do you think it's worth updating the charset part of the ContentTypeParser code to make it proper error-reporting parser instead of that existing regexp-based one?

[01:47:45.0000] <MikeSmith>
if so I think I'll go ahead and do that

[01:49:38.0000] <MikeSmith>
it seems good to alert a user with a useful error message for the "text/html; charset = utf-8" case if that's not going to actually work as expected in browsers

[01:56:12.0000] <hsivonen>
MikeSmith: yeah, that would be nice

[01:56:23.0000] <MikeSmith>
ok

[01:57:01.0000] <hsivonen>
the JS i18n API will suck for non-top-tier locales that won't be supported in all browser from day 1

[01:58:44.0000] <hsivonen>
the WHATWG-style notion of interop is rather incompatible with the politics of never closing the door on the set of supported locales

[01:59:12.0000] <hsivonen>
hmm. that statement wasn't quite right

[01:59:44.0000] <hsivonen>
the WHATWG-style notion of interop is rather incompatible with the politics of not having a normative list of which locales are in the supported set today

[02:01:03.0000] <hsivonen>
eventually, the standard will be "whatever ICU supports" and everyone has to ship ICU as the back end impl

[02:01:24.0000] <hsivonen>
except Microsoft probably won't an will roll their own

[02:09:54.0000] <annevk>
ICU is kind of terrible

[02:10:08.0000] <annevk>
at least for encodings the architecture is way too complex

[02:25:48.0000] <annevk>
MikeSmith: I wonder if other headers with similar parameter-style are parsed in the same way

[02:26:01.0000] <MikeSmith_>
annevk: I think not

[02:26:11.0000] <annevk>
MikeSmith: this does not sound like the way e.g. <meta http-equiv=content-type> is parsed

[02:26:15.0000] <MikeSmith_>
I think Content-Disposition is not, for one

[02:26:27.0000] <annevk>
and they call HTML a mess

[02:26:33.0000] <MikeSmith>
heh

[02:26:42.0000] <MikeSmith>
yeah

[03:24:17.0000] <david_carlisle>
MikeSmith: (or anyone) If I want to comment on one of these bugs that's been cloned into HTML WG and WhatWG components, am I supposed to comment on the old one, the new one or both?

[03:24:51.0000] <MikeSmith>
the new one

[03:24:55.0000] <MikeSmith>
I suggest

[03:25:48.0000] <Ms2ger>
Then again, we would say that in this channel

[03:33:57.0000] <david_carlisle>
MikeSmith:  OK thanks, it might be worth posting to html and or wg lists saying what just happened, I got several mails I expect some got a lot more:-)

[03:34:21.0000] <MikeSmith>
yeah sorry about the bugmail bomb

[03:34:37.0000] <MikeSmith>
I got a lot more :-)

[03:34:52.0000] <david_carlisle>
MikeSmith:  good:-)

[04:09:12.0000] <annevk>
david_carlisle: yeah, the clarifying email will come

[04:09:42.0000] <annevk>
david_carlisle: prolly end of this week or early next, whenever Hixie gets around to it

[04:10:20.0000] <annevk>
david_carlisle: basically, the HTML WG forked development on HTML

[04:11:10.0000] <david_carlisle>
annevk: yes but I hoped they ony _say_ they forked not actually have different bugs with different resolutions, which would be ... unfortunate

[04:11:21.0000] <Ms2ger>
Well

[04:11:22.0000] <david_carlisle>
only

[04:11:35.0000] <Ms2ger>
More likely they'll have a lot of bugs without resolutions instead

[04:12:09.0000] <david_carlisle>
Ms2ger:  good plan

[04:25:32.0000] <hsivonen>
does the HTML WG have an editor for its fork, yet?

[04:26:54.0000] <hsivonen>
I wonder if someone has scientifically figured out why the WebApps WG kinda works but the HTML WG doesn't

[04:35:21.0000] <odinho>
Culture.

[04:37:24.0000] <Ms2ger>
Are you a scientist? :)

[04:43:37.0000] <odinho>
Òó

[05:03:26.0000] <annevk>
hsivonen: chairs that tolerate bullshit?

[05:03:55.0000] <annevk>
hsivonen: maybe also because WebApps is more difficult so you get less bullshit to start with

[05:04:41.0000] <annevk>
hsivonen: well, HTML is deceptively simple anyway, not actually

[05:09:42.0000] <hsivonen>
annevk: yeah, I'd prefer the chairs to be more like "can't behave like that here" or "nope, not gonna do that" bouncers as opposed to putting everything to a poll out of fairness

[05:15:34.0000] <odinho>
Ms2ger: Absolutely not ;]

[05:32:12.0000] <hsivonen>
it's unfortunate that the Security Considerations section of CORS doesn't cover the concern related to authentication by IP address or authentication by being on the right side of a firewall

[05:35:37.0000] <odinho>
That sections has been a bit back and forth IIUC.

[05:37:05.0000] <hsivonen>
time and again there's always someone new who doesn't realize that CORS protects network resources whose authentication relies on the network topology

[05:48:52.0000] <odinho>
true dat

[05:50:43.0000] <hsivonen>
another repeating thing: Ars writes that new release of browser Foo now implements API Bar without explicitly spelling out that care was taken in making API Bar secure. Reader in the comments freaks out assuming that the security issues related to the subject matter of API Bar haven't been addressed or thought about.

[06:01:28.0000] <hsivonen>
I wonder if hober is on track to become the new HTML WG editor considering all the CPs...

[06:01:58.0000] <annevk>
the W3C is trying to hire someone for it, in case you missed that

[06:02:02.0000] <annevk>
as a fulltime job

[06:02:13.0000] <annevk>
I think it's the first time the W3C offers to pay an editor

[06:04:24.0000] <hsivonen>
annevk: interesting. I missed that.

[06:04:40.0000] <hsivonen>
annevk: have you applied for the job now that you are no longer at Opera? :-)

[06:06:01.0000] <Ms2ger>
hsivonen, nah, he's going to do XSLT :)

[06:06:20.0000] <annevk>
hsivonen: not sure why I would put time in what Hixie is already doing quite well

[06:07:14.0000] <hsivonen>
annevk: speaking of editorships, are you keeping the editorship of the specs you were editing as an Opera employee?

[06:08:11.0000] <annevk>
hsivonen: yeah, I plan to apply for invited expert status

[06:09:33.0000] <hsivonen>
annevk: cool. I hope whatever your new business model is allows you to keep editing.

[06:10:21.0000] <annevk>
hsivonen: I hope so too :)

[06:22:19.0000] <gsnedders>
Hixie: U+1D113 is the musical symbol, which is different to what you want in verse.

[06:33:55.0000] <hsivonen>
/me finds http://www.w3.org/Consortium/Recruitment/#techeditor

[07:19:24.0000] <annevk>
" I believe the WG already has consensus on this point." (on public-webappsec)

[07:19:41.0000] <annevk>
not sure why we need to have consensus on facts, but I guess...

[07:29:21.0000] <hsivonen>
hooray. not declaring the character encoding is now invalid.

[07:29:34.0000] <hsivonen>
so I get to remove some complexity from the HTML parser's Java version

[07:52:34.0000] <odinho>
hsivonen: Oh the joys in life!

[07:58:43.0000] <MikeSmith>
hsivonen: I just finished rewriting the Content-Type parsing

[07:58:45.0000] <MikeSmith>
testing it now

[07:58:57.0000] <MikeSmith>
will send you a patch when I think it's ready

[09:41:58.0000] <Ms2ger>
Philip`, any chance you could put a warning on http://philip.html5.org/tests/canvas/ that those tests aren't updated anymore?

[09:46:31.0000] <dglazkov>
good morning, Whatwg!

[09:58:30.0000] <Hixie>
annevk: where would you suggest i e-mail, and what would you suggest i say in said e-mail?

[10:55:52.0000] <MikeSmith>
so I made a patch that adds an error-reporting Content-Type header parser to the validator

[10:55:56.0000] <MikeSmith>
and pushed it to http://qa-dev.w3.org:8888/

[10:56:11.0000] <MikeSmith>
please test

[10:56:33.0000] <MikeSmith>
or if anybody has any Content-Type test cases please point me to them

[11:51:00.0000] <Hixie>
there's some content-type tests on hixie.ch/tests/adhoc/http i think, dunno if they'd help

[12:20:59.0000] <sunshinehappy>
te//o

[12:25:15.0000] <sunshinehappy>
hello

[12:25:18.0000] <sunshinehappy>
what is it

[12:26:35.0000] <Hixie>
what is what

[12:27:07.0000] <Philip`>
Ms2ger: Yes, if I still remember by the time I have a proper computer+internet set up, and am not too grossly lazy

[12:29:08.0000] <sunshinehappy>
whatwg

[12:31:20.0000] <Hixie>
sunshinehappy: it's where we write the html standard and some other specs

[12:31:41.0000] <sunshinehappy>
why do you guys write HTML standard? I thought that's what W3C do

[12:32:02.0000] <Hixie>
we work with the w3c

[12:32:13.0000] <Hixie>
they're currently more focused on releasing a snapshot

[12:32:15.0000] <sunshinehappy>
cool!

[12:32:18.0000] <Hixie>
we're doing the work on future stuff

[12:32:43.0000] <Hixie>
(thought right now they are not making any changes, so they're also missing the fixes we're making to the old stuff)

[12:32:52.0000] <Hixie>
(but they plan on fixing that)

[12:33:43.0000] <sunshinehappy>
what's are the ideas for future?

[12:33:45.0000] <Hixie>
MikeSmith: do you know if r12a is on irc anywhere?

[12:34:10.0000] <Hixie>
sunshinehappy: right now i'm doing research for an inputmode="" attribute on <input> to help people write web pages for mobile phones

[12:34:25.0000] <Hixie>
sunshinehappy: so you can say e.g. "this input field should have automatic capitalisation" or "this should default to all lowercase"

[12:34:34.0000] <sunshinehappy>
cool :)

[12:34:36.0000] <Hixie>
sunshinehappy: trying to fill in this page: http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control

[12:34:54.0000] <Hixie>
unfortunately i'm finding it hard to find any information on how input modalities work in asian markets

[12:35:00.0000] <Hixie>
in particular, chinese, japanese, and korean

[12:35:18.0000] <Hixie>
also vietnamese i hear needs something special but i can't find any data on it

[13:05:17.0000] <loincloth>
Hixie: that example and a number of HTML5 form features strike me as blurring the line between markup and behaviors.. do i miss the point or is that a legitimate observation?

[13:05:54.0000] <loincloth>
i'm sure HTML5 wasn't the first to dance around that line, either.. i sometimes think JS and CSS should be one language, too, so i'm just thinking out loud.. curious for reactions

[13:35:36.0000] <Tabatkins>
loincloth: We've been blurring that line for forever.  It's not a sacrosanct barrier.

[13:36:58.0000] <Tabatkins>
HTML blends content and behavior in a lot of form stuff, and <details>, etc.  It blend content and style in its half-presentational elements.  CSS blends style and content with pseudo-elements.  It blends style and behavior with :hover/:focus/:active, and Animations.  JS blends everything, because it's a fully-powered programming language.

[13:44:37.0000] <loincloth>
Tabatkins: indeed.. as i said, i know this musing isn't new.. but a new example was given that sparked further discussion

[14:19:50.0000] <Hixie>
loincloth: not sure what the difference is

[14:20:36.0000] <loincloth>
Hixie: between?

[14:34:54.0000] <Hixie>
loincloth: markup and behaviours

[14:35:00.0000] <Hixie>
loincloth: or rather, they seem to be orthogonal

[14:35:18.0000] <Hixie>
loincloth: you can use markup to describe behaviours... it's like the difference between "english" and "poetry"

[14:44:46.0000] <annevk>
Hixie: whatwg⊙wo and maybe public-html⊙wo

[14:45:43.0000] <annevk>
Hixie: just state that from the perspective of the WHATWG we'll keep updating HTML and do so by paying attention to the WHATWG/HTML Bugzilla component and that for W3C HTML people can contact the W3C

[14:46:08.0000] <annevk>
Hixie: and maybe elaborate on the scripts you wrote to keep the two parties in sync as far as incoming information goes

[14:49:31.0000] <Hixie>
well my goal for keeping stuff in sync in the future is just that i'll keep an eye on the w3c side

[14:50:08.0000] <Hixie>
and the chairs haven't asked me to do anything for their side

[14:50:15.0000] <Hixie>
so e.g. bugs aren't being automatically filed on both sides

[14:51:26.0000] <Hixie>
not sure i should post anything to public-html, i'd probably get flamed to death

[14:51:36.0000] <annevk>
I think we should at least inform whatwg⊙wo about why some people received an email about a duplicated bug

[14:51:44.0000] <Hixie>
fair enough

[14:51:53.0000] <annevk>
and how W3C HTML and WHATWG HTML are no longer in sync

[14:52:29.0000] <Hixie>
might be worth you doing a blog post about it too

[14:52:39.0000] <Hixie>
so feel free to crib from the e-mail i'm going to write

[14:53:18.0000] <annevk>
yeah was planning on doing that

[14:53:32.0000] <annevk>
maybe I'll include one of the memes people made

[14:53:41.0000] <Hixie>
there are memes?

[14:53:42.0000] <annevk>
http://w3cmemes.tumblr.com/post/27561016967 is fun

[14:53:59.0000] <annevk>
http://w3cmemes.tumblr.com/post/27558722901/stereotypical-html-wg-poll-respondent about the polls is kind of funny too

[14:54:14.0000] <Hixie>
heh

[14:54:31.0000] <Hixie>
let's not insult w3c too prominently... :-P

[15:07:57.0000] <othermaciej>
I wonder what this one's about? http://w3cmemes.tumblr.com/post/27215424834/the-most-interesting-standards-guy-in-the-world

[15:10:39.0000] <smaug____>
that is a good one :)

[15:12:47.0000] <annevk>
othermaciej: ask hober ;)

[15:12:52.0000] <annevk>
nn

[15:16:53.0000] <pablof_>
Hixie: re: http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#dom-textarea/input-setrangetext

[15:17:09.0000] <pablof_>
Hixie: in step 6, shouldn't "selection start" be "start" and "selection end" be "end"?

[15:17:42.0000] <pablof_>
Hixie: also, in step 8, for "select", shouldn't "selection end" be set to "start + replacement.length()"? similar for mode "end"

[15:17:56.0000] <Hixie>
yikes, that does seem plausible

[15:18:03.0000] <Hixie>
can you file a bug using the widget at the bottom right?

[15:18:14.0000] <pablof_>
sure

[15:20:12.0000] <Hixie>
thanks

[15:49:40.0000] <Hixie>
ok, long rambling administrivia e-mail sent

[16:08:49.0000] <heycam>
the name "operation convergance" is presumably ironic then?

[16:09:22.0000] <zewt>
operation fragment bug discussion

[16:09:45.0000] <Hixie>
heycam: not really, the worry was that if we didn't clone the bugs that each bug would only be dealt with by one bug, so it'd be even worse

[16:10:09.0000] <Hixie>
heycam: at least this way all the bugs are handled by both groups, so whoever gets to a bug second can first check what the other group did and try to keep in step

[16:10:17.0000] <heycam>
Hixie, I see

[16:10:21.0000] <Hixie>
heycam: it's pretty inevitable that the specs will diverge more, though, imho

[16:10:25.0000] <heycam>
Hixie, yeah

[16:10:43.0000] <heycam>
Hixie, hopefully when the W3C find an editor on their side that they can keep up

[16:10:58.0000] <Hixie>
yeah well that's another thing

[16:11:09.0000] <Hixie>
i don't think it'll be physically possible for the wg to keep up, given their process

[16:11:15.0000] <Hixie>
whoever they have as editor

[16:11:19.0000] <heycam>
interesting times, then

[16:11:28.0000] <Hixie>
i mean, one of the reasons i was in favour of this split is that i was being slowed down by the process

[16:11:35.0000] <Hixie>
and i didn't have to keep up with anyone...

[16:11:41.0000] <heycam>
heh yeah

[16:14:51.0000] <Hixie>
interesting times indeed, i'm really quite honestly curious to see where this goes

[16:34:46.0000] <MikeSmith>
Hixie: thanks will take a look at those content-type tests

[16:35:05.0000] <MikeSmith>
and no Richard's not around on irc right now

[16:35:29.0000] <MikeSmith>
he's not always on so many hours

[16:35:50.0000] <MikeSmith>
I think he may come on when he has meetings and drop off sometimes when he's working

[16:36:22.0000] <MikeSmith>
but if it's about the input-mode stuff, I can help with modalities for Japanese


2012-07-20
[17:35:52.0000] <Hixie>
MikeSmith: see my google+ post :-)

[17:37:29.0000] <MikeSmith>
/me reads

[17:39:07.0000] <MikeSmith>
so one quick thing I can tell you is that on almost all Japanese IMEs I have used on mobile devices, there are no separate hiragana and katakana modes

[17:40:11.0000] <MikeSmith>
there's only general kana mode, alphabet mode, and number mode at the base

[17:41:23.0000] <MikeSmith>
and usually a separate symbol mode for typing in emoticons and special characters

[17:43:03.0000] <MikeSmith>
I'll try to make some screenshots today and annotate them

[17:43:52.0000] <MikeSmith>
I have three different Japanese IMEs installed on my Android mobile, and that's the common case for all of them

[17:44:03.0000] <MikeSmith>
I guess I should buy ATOK and take a look at that

[17:45:10.0000] <espadrine`>
gsnedders: do you know anything about DOM properties that are enumerable, but are not on the prototype chain? Is that considered a bug?

[17:45:15.0000] <MikeSmith>
ah cool they have a trial version

[17:47:04.0000] <hober>
people still use google+?

[17:47:21.0000] <espadrine`>
gsnedders: I noticed that Opera doesn't have any, but Firefox and Chrome have a bunch.

[17:48:21.0000] <divya>
shhh hober

[17:49:07.0000] <hober>
/me forgot to use his inner monologue

[17:50:14.0000] <divya>
/me offers to be hober's conscience

[17:51:42.0000] <MikeSmith>
hey I need a conscience too

[17:51:54.0000] <MikeSmith>
at least that's what some people tell me

[17:53:52.0000] <hober>
yeah, i think MikeSmith is ahead of me in the getting-a-conscience queue, divya

[17:53:53.0000] <zewt>
MikeSmith: that's pretty odd (no hiragana/katakana toggle); does it depend on heuristics to convert or are people just OK with that?

[17:54:14.0000] <divya>
but you are in sf hober ! I can pretty much do that job IRL too.

[17:54:32.0000] <divya>
MikeSmith: i shall be your conscience digitally

[17:55:05.0000] <MikeSmith>
divya: :)

[17:55:38.0000] <MikeSmith>
zewt: it includes both the hiragana and katakana in the candidate lists it gives back

[17:55:57.0000] <zewt>
some of those jp ime's are weirdo

[17:56:07.0000] <zewt>
i guess plenty of western inputs are weirdo too (swype, etc)

[17:57:03.0000] <MikeSmith>
I really like swype

[17:57:18.0000] <MikeSmith>
swype has a Japanese IME too

[17:57:32.0000] <MikeSmith>
but it doesn't work nearly as well as the English one

[18:35:31.0000] <MikeSmith>
Hixie: http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control#Android_Japanese_IMEs

[18:38:17.0000] <espadrine`>
gsnedders: related to the prototype-invisible properties issue, here's the data I have, and some code to fiddle with: http://3.jsbin.com/ebukuv/2/edit

[18:42:45.0000] <Hixie>
MikeSmith: thanks man

[18:43:07.0000] <Hixie>
hober: keeps growing, see our i/o announcements :-D

[18:44:32.0000] <MikeSmith>
Hixie: if you want screenshots of submodes lemme know but really those are the same as as with other IMEs. Well except that there are a bunch of different subcategories of emoji. But I don't think anybody wants to have separate input modes for those subcategories.

[18:46:01.0000] <Hixie>
MikeSmith: you describe four modes but have three pictures -- can you clarify which relates to which?

[18:47:03.0000] <MikeSmith>
Hixie: those are three different IMEs: the first is one called ATOK that's very commonly used, the second is Google Japanese, and the third is the IME that's built-in on Samsung mobiles

[18:47:13.0000] <Hixie>
aah, ok

[18:47:20.0000] <Hixie>
they all correspond to "normal" mode?

[18:47:21.0000] <MikeSmith>
they're all shown in kana mode

[18:47:26.0000] <Hixie>
ah cool

[18:47:26.0000] <Hixie>
thanks

[18:48:25.0000] <Hixie>
MikeSmith: so there's no separate "kanji" mode or whatnot?

[18:48:37.0000] <MikeSmith>
you can see they're all basically the same: you have one button for switching among あ-A-1, and one other button with a smiley face for getting emoticons and symbols

[18:48:47.0000] <MikeSmith>
no there is no separate kanji mode

[18:49:03.0000] <MikeSmith>
making the kanji is a two-step process that starts by typing in kana

[18:49:26.0000] <MikeSmith>
so there is no need for a kanji mode in addition to kana mode

[18:49:33.0000] <MikeSmith>
they're the same thing

[18:50:00.0000] <MikeSmith>
and same goes getting katakana actually

[18:50:06.0000] <MikeSmith>
you have to first time in hiragana

[18:50:12.0000] <MikeSmith>
on mobile IMEs that is

[18:50:14.0000] <Hixie>
MikeSmith: so just the same ones as western IMEs (normal, numeric, e-mail, url, telephone), with submodes like symbols/emoji, and then the additional mode for specifically entering in "latin" input?

[18:50:24.0000] <MikeSmith>
on desktop IMEs there are usually a separate katakana mode

[18:50:36.0000] <MikeSmith>
Hixie: yeah

[18:50:40.0000] <Hixie>
k

[18:50:58.0000] <Hixie>
that's surprisingly mundane and restrained for the japanese :-P

[18:51:02.0000] <MikeSmith>
heh

[18:51:12.0000] <Hixie>
i guess they let loose in the emoji submodes...

[18:51:17.0000] <MikeSmith>
yeah

[18:51:41.0000] <Hixie>
thanks a ton for this

[18:51:54.0000] <Hixie>
you are so far basically the only person to give me any information :-)

[18:51:55.0000] <MikeSmith>
those have separate subcategories for emotions for laughing, for crying, for being angry, etc.

[18:52:11.0000] <Hixie>
/me has seen the unicode emoji pages, good lord

[18:52:11.0000] <MikeSmith>
Hixie: somebody else put up iOS screenshots there

[18:52:32.0000] <MikeSmith>
yeah actually before when I said emoji I meant emotions

[18:52:38.0000] <MikeSmith>
*emoticons

[18:52:44.0000] <Hixie>
i put the ios ones up :-(

[18:52:48.0000] <MikeSmith>
ah

[18:53:39.0000] <MikeSmith>
on Android smartphones at least there is not way to type in all those unicode emoji, or the emoji that we have on legacy keitai mobiles

[18:53:43.0000] <Hixie>
(found them on some stackoverflow question)

[18:53:54.0000] <Hixie>
oh really?

[18:53:57.0000] <Hixie>
interesting

[18:53:57.0000] <MikeSmith>
yeah

[18:54:13.0000] <MikeSmith>
it's one of the fun Japanese things that's lost in going to smartphones

[18:54:14.0000] <Hixie>
does android have much penetration?

[18:54:20.0000] <MikeSmith>
yeah very much

[18:54:50.0000] <MikeSmith>
Docomo has tons of different Android devices, and the number two carrier Au does too

[18:56:19.0000] <Hixie>
weird, i thought not having emoji was considered a blocker to deployment in .jp and that was used as the argument for adding the six zillion pages to unicode

[18:57:03.0000] <zewt>
are you sure?

[18:57:14.0000] <zewt>
i'd think it's more round-trip compat with sjis or something like that

[18:58:13.0000] <MikeSmith>
it does suck to not have them. I used to use them all the time in mobile e-mail messages

[18:58:24.0000] <MikeSmith>
beer glass, hypodermic needle, etc.

[19:00:04.0000] <MikeSmith>
a wait they are still there, but only in the e-mail client

[19:00:16.0000] <Hixie>
they're the only part of unicode that specifies actual _colours_

[19:00:48.0000] <MikeSmith>
being only available in mobile e-mail makes sense because that's the only place you could reliably use them anyway

[19:01:21.0000] <MikeSmith>
I just don't notice because I avoid using Docomo's mobile e-mail client because it's so bad and painful to use

[19:04:09.0000] <MikeSmith>
anyway for all those emoji characters I don't think anybody would want a separate way to specify an emoji inputmode for form fields in Web applications. Because you never want to start out typing in emoji. You use the emoji to decorate your prose.

[19:04:48.0000] <Hixie>
oh yeah i wasn't suggesting anyone would want that

[19:06:28.0000] <Hixie>
MikeSmith: is there a concept of a difference between a text field with capitalisation or not, the same way there's a concept of a difference between message text fields and username text fields in the west?

[19:06:42.0000] <Hixie>
for kana input, i mean

[19:07:21.0000] <MikeSmith>
anyway to be clear again what I said before about emoji not being available on smartphones as wrong. It's still available in the same one place it's always been, in the carrier-specific mobile e-mail clients. But it's not available in Gmail and such so I think since many more people here are just using Gmail they don't use emoji in their messages.

[19:08:10.0000] <MikeSmith>
Hixie: no concept of a difference in capitalization for Japanese/kana mode

[19:08:19.0000] <Hixie>
how about a way to turn off the ability to upconvert from hiragana to kanji or katakana?

[19:08:27.0000] <Hixie>
any fields that prevent that?

[19:08:37.0000] <MikeSmith>
no not that I know of

[19:08:42.0000] <Hixie>
very interesting

[19:08:45.0000] <MikeSmith>
what use case relates to that?

[19:08:52.0000] <Hixie>
no idea

[19:08:59.0000] <Hixie>
i'm mostly stumbling in the dark here

[19:09:29.0000] <MikeSmith>
at some point you'll probably want to try to get more details from a Japanese native speaker/user

[19:10:01.0000] <Hixie>
that's what my g+ post was attempting to get :-)

[19:10:26.0000] <Hixie>
i should record another video for those guys who sent me a book

[19:10:37.0000] <Hixie>
"hello hero worshippers! i have a mission for you now!"

[19:10:59.0000] <MikeSmith>
ahaha

[19:11:32.0000] <MikeSmith>
hey a good source to get info from at Google Japan is Hiroyuki Komatsu

[19:11:59.0000] <MikeSmith>
he leads the product-dev team for Google Japanese IME

[19:11:59.0000] <Hixie>
i asked internally at google too (on our internal g+), will see what response i get overnight and poke people if i don't

[19:12:22.0000] <MikeSmith>
talk to Takuya Oikawa if you know him

[19:12:42.0000] <MikeSmith>
he can put you in touch with Komatsu-san or whoever

[19:12:54.0000] <MikeSmith>
or ask Fette because he knows all these guys too

[19:13:03.0000] <MikeSmith>
Or Dominic Clooney

[19:13:15.0000] <Hixie>
my office is literally four meters from the i18n team

[19:13:16.0000] <Hixie>
so i'm good

[19:13:25.0000] <Hixie>
(not that i ever go to that building)

[19:13:26.0000] <MikeSmith>
ok

[19:13:28.0000] <Hixie>
(but that's another story)

[19:20:23.0000] <Hixie>
so i guess the "normal" mode should just be "user default", and japanese just adds one extra mode

[19:20:55.0000] <Hixie>
and "capitalise" mode should be "latin captitalise" and "prediction off" should be "latin without prediction" and so on

[19:21:06.0000] <Hixie>
i guess it depends a lot on what other languages want

[19:21:17.0000] <Hixie>
but my point basically is that xforms' model is apparently wrong

[19:21:32.0000] <Hixie>
and rather than script * mode, it's just a list of modes

[19:33:18.0000] <zewt>
android is really annoying, it tends to merge "autocapitalization" and "prediction" into a single setting

[19:33:33.0000] <zewt>
so I can't get IRC with prediction/autocorrection but without autocap

[20:11:47.0000] <MikeSmith>
Hixie: yeah the Xforms model doesn't match anything except some twinkle in the eye of an architecture astronaut

[20:12:16.0000] <Hixie>
yeah, seems that way

[20:12:58.0000] <MikeSmith>
and yeah to be clear there is no way to turn off prediction in Japanese mode. The prediction in that case is really more like completion and is fundamental to the way the whole thing works.

[20:13:55.0000] <MikeSmith>
that said, sometimes the IME prediction conflicts with stuff like Google Suggest so it would be good to have a way to disable the system IME sometimes

[20:14:09.0000] <MikeSmith>
but that's what the IME API spec is attempting to do (in part)

[20:14:55.0000] <Hixie>
not sure what disabling the IME altogether would mean, without script, but with script i'd presume the API would turn off the system-level stuff, yeah

[20:15:32.0000] <MikeSmith>
zewt: about no auto-caps for me at least I get more sane behavior for that in Swype. Or it seems like at least

[20:15:50.0000] <zewt>
yeah just talking about the default latin input

[20:16:38.0000] <zewt>
MikeSmith: if google suggest is broken by "people inputting japanese", isn't effectively disabling japanese input the wrong solution? :)

[20:17:46.0000] <MikeSmith>
lemme get a screenshot to show what I mean

[20:18:10.0000] <MikeSmith>
it's a problem of the system-IME candidate list obscuring the Google Suggest candidate list

[20:18:18.0000] <zewt>
google search is much less polished than it used to be; for example, pressing arrows while its suggestion dropdown is displayed operates on keyup (instead of down) and doesn't key repeat--very rough stuff

[20:18:46.0000] <zewt>
MikeSmith: wouldn't the solution be a hint of where you want IME overlays to go? i think most mature ones have something like that

[20:18:56.0000] <zewt>
"i need the space below the input box, so try to put stuff on top"

[20:23:31.0000] <MikeSmith>
this: http://people.w3.org/mike/mobile-ime/ime-conflict.png

[20:23:58.0000] <Hixie>
heh

[20:24:00.0000] <Hixie>
oops

[20:24:03.0000] <MikeSmith>
there are Google Suggest candidates that are obscured by the system IME

[20:24:24.0000] <zewt>
MikeSmith: but given the two, isn't the IME the more important thing to see?

[20:24:33.0000] <MikeSmith>
so to get to them I have to either drag the system IME window out of the way, or scroll the page

[20:24:48.0000] <zewt>
of course, the IME should go away once you pick something...

[20:24:49.0000] <MikeSmith>
zewt: no actually it's often not

[20:24:57.0000] <MikeSmith>
yeah it does do that at least

[20:25:14.0000] <zewt>
how can you even use the IME if there's no completion?

[20:25:30.0000] <zewt>
unless you happen to get what you want on the first guess

[20:25:45.0000] <MikeSmith>
you do usually, with Google Suggest at least

[20:26:16.0000] <MikeSmith>
it is much more often the case that you want to see the smart Google Suggest candidates rather than the generic system IME ones

[20:30:00.0000] <MikeSmith>
zewt: I'm not sure what the best behavior would be, but I think maybe having the system IME on when you compose the first clause and then turning it off until you start to type another clause

[20:30:06.0000] <MikeSmith>
if you see what I mean

[20:30:53.0000] <zewt>
are there any use cases other than gsuggest (out of curiosity)?

[20:31:04.0000] <MikeSmith>
i need the IME on to do the initial composition but once I've stopped typing I don't need that candidate window from the system IME

[20:31:08.0000] <MikeSmith>
zewt: yeah Bing

[20:31:10.0000] <MikeSmith>
same problem

[20:31:23.0000] <MikeSmith>
or any Web app that has as similar suggest feature

[20:32:02.0000] <MikeSmith>
it's a general problem and was one of the primary cases if not the primary case for coming up with the IME API

[20:37:48.0000] <zewt>
well, most autocomplete dropdowns in web apps aren't smart enough to work for japanese without the native IME's help for kanji, but okay

[20:38:06.0000] <zewt>
the high-profile ones (search engines, amazon) are

[20:41:59.0000] <MikeSmith>
zewt: yeah that's the use case we had in mind anyway

[20:42:25.0000] <MikeSmith>
those are the ones that cause the widest user frustration

[21:09:24.0000] <MikeSmith>
hey it's abarth in a video http://www.youtube.com/watch?v=EDtiWN42lHs

[21:16:10.0000] <MikeSmith>
wtf <browser src="http://news.google.com/" width="640" height="480"></browser>

[21:16:16.0000] <MikeSmith>
http://code.google.com/chrome/extensions/trunk/apps/app_external.html#H3-4

[21:16:40.0000] <MikeSmith>
and <a name="H3-4"></a> in the source

[21:17:26.0000] <MikeSmith>
"Unlike iframes, the browser tag runs in a separate process."

[21:18:32.0000] <zewt>
why would authors care?

[21:18:56.0000] <MikeSmith>
security for installed apps

[21:19:23.0000] <MikeSmith>
it's not for the Web I guess

[21:19:30.0000] <MikeSmith>
but for the un-Web

[23:31:00.0000] <hsivonen>
what's the typical connotation of the meme picture at http://w3cmemes.tumblr.com/post/27558722901/stereotypical-html-wg-poll-respondent ?

[23:32:42.0000] <Hixie>
it's the background of advice animal memes

[23:32:56.0000] <Hixie>
thus the implication is "nothing" or "nobody"

[23:34:09.0000] <hsivonen>
Hixie: I see

[23:36:46.0000] <hsivonen>
sigh. Web and TV IG is confidential

[23:37:12.0000] <hsivonen>
and they are supposed to request features for HTML video, right?

[23:37:27.0000] <Hixie>
haven't they already requested all the features they want?

[23:37:40.0000] <Hixie>
one presumes they have been requesting them in order of priority

[23:37:49.0000] <hsivonen>
Hixie: maybe

[23:37:51.0000] <Hixie>
and some of there requests are pretty far down what i would consider the priority list

[23:38:23.0000] <hsivonen>
What's Browser Testing and Tools? Should I go observe their meeting at TPAC?

[23:38:51.0000] <Hixie>
/me is going to assume the question about whether someone should do something at TPAC is not aimed at him :-P

[23:41:29.0000] <annevk>
is there some deadline for TPAC coming up?

[23:41:42.0000] <hsivonen>
the TPAC registration form could use some "is the logged-in person an AC rep" logic

[23:41:46.0000] <annevk>
I kind of want to go, but I also kind of not want to pay a whole lot for doing so

[23:41:54.0000] <hsivonen>
annevk: DL coming up on Oct 16

[23:42:03.0000] <annevk>
thanks

[23:42:10.0000] <annevk>
sounds like I have plenty of time then

[23:42:54.0000] <MikeSmith>
hsivonen: Browser Testing and Tools is really just for the WebDriver API at this point

[23:43:11.0000] <MikeSmith>
I wrote the charter in a way that will let us add more to that group if we want

[23:43:28.0000] <MikeSmith>
but right now it's exclusively about WebDriver

[23:43:59.0000] <MikeSmith>
if you know David Burns he can tell you more and help you decide if it's worth your time

[23:44:16.0000] <Hixie>
charters should have more distributed extensibility

[23:44:23.0000] <MikeSmith>
heh

[23:44:26.0000] <Hixie>
and with that, i bid you all adieu for the evening!

[23:44:46.0000] <MikeSmith>
wow Hixie quitting before midnight

[23:44:50.0000] <MikeSmith>
that's a rarity

[23:44:53.0000] <hsivonen>
MikeSmith: thanks. So I won't observe that.

[23:44:56.0000] <MikeSmith>
hsivonen: nn

[23:44:59.0000] <MikeSmith>
oops

[23:45:01.0000] <MikeSmith>
Hixie: nn

[23:45:04.0000] <MikeSmith>
hsivonen: OK

[23:45:54.0000] <MikeSmith>
hsivonen: I'm not sure we'll actually be having a f2f at TPAC anyway

[23:45:58.0000] <MikeSmith>
for that group

[23:45:58.0000] <hsivonen>
huh. TAG is confidential even though www-tag isn't

[23:46:15.0000] <MikeSmith>
TAG meetings have always been confidential I thought

[23:46:31.0000] <hsivonen>
ok

[23:46:46.0000] <annevk>
there's tag⊙wo too

[23:46:59.0000] <MikeSmith>
Browser Testing and Tools WG is essentially chair-less right now

[23:47:16.0000] <MikeSmith>
if anybody would like to volunteer

[23:47:46.0000] <MikeSmith>
in the mean time I guess I have to be the one to respond to the TPAC survey for the group

[23:47:57.0000] <MikeSmith>
hsivonen: URL for the survey?

[23:48:15.0000] <hsivonen>
MikeSmith: https://www.w3.org/2002/09/wbs/35125/TPAC2012/?login

[23:48:18.0000] <MikeSmith>
thanks

[23:53:18.0000] <MikeSmith>
hmm I guess the fact the group is already listed in the survey means somebody must have already requested space for a f2f

[01:03:39.0000] <annevk>
http://blog.whatwg.org/html-and-html5

[01:03:49.0000] <annevk>
Hixie: ^^

[01:04:37.0000] <annevk>
oh Hixie is asleep, can someone else do a quick review?

[01:06:21.0000] <hsivonen>
annevk: looks ok.

[01:06:25.0000] <annevk>
thanks

[01:24:06.0000] <hsivonen>
I foresee great implementor interest and success: http://tools.ietf.org/html/draft-karcz-uuas-00

[01:41:16.0000] <MikeSmith>
hsivonen: oh boy

[01:46:33.0000] <odinho>
http://drupal.org/node/1170478 <-- So drupal is meditating wether to put in a <picture> polyfill into their core. That would really be disastrous.

[01:47:23.0000] <MikeSmith>
odinho: yeah

[01:48:06.0000] <odinho>
I like more and more having srcset WITH the viewport declarations, and then having   <img src="balbla" defer>  for more advanced usage.

[01:48:50.0000] <MikeSmith>
I like whatever ends up getting shipped natively

[01:49:12.0000] <odinho>
Yes, I like that most :P

[01:49:40.0000] <odinho>
But seeing that drupal page, -- people really do like flexibility. And I like simplicity.

[01:50:01.0000] <odinho>
Having something like @defer would allow the flexibility whilst also being very very simple.

[01:51:15.0000] <MikeSmith>
advocating indignantly for certain syntax is almost as good a way to spend time as arguing for particular/more semantic elements

[01:53:16.0000] <odinho>
MikeSmith: Well, the behaviour is starkly different.

[01:53:37.0000] <odinho>
It is not about syntax. It's about who gets the power to decide what is best for the user. If it is the web page author, or the user agent.

[01:56:27.0000] <MikeSmith>
yeah I understand that part

[01:57:53.0000] <MikeSmith>
but if somebody believes that unilaterally implementing it in drupal is going to somehow push browser projects to implement it the way they want, I guess they are going to be in for further disappointment

[01:59:05.0000] <odinho>
I'm afraid it's doing more hurt, -- vendors have to think about being compliant with how it is used in Drupal. And so it might cripple the possible usage and flexibility for the spec down the road.

[02:00:54.0000] <abarth_>
browser vendors will just pick a different name

[02:00:58.0000] <abarth_>
to avoid the baggage

[02:01:27.0000] <odinho>
Yea, most possibly, if going down that route of having a new image element.

[02:01:34.0000] <odinho>
But seems like drupal doesn't understand that.

[02:01:47.0000] <odinho>
They think they'll get browser support "for free" in the future. :-)

[02:01:59.0000] <abarth_>
glhf

[02:02:07.0000] <abarth_>
it worked for json.js

[02:02:14.0000] <abarth_>
so, it's not impossible

[02:02:18.0000] <abarth_>
just difficult

[02:02:28.0000] <odinho>
Yes. :-) But they had a spec from the start.

[02:02:37.0000] <odinho>
Not just some strawman proposals.

[02:02:59.0000] <odinho>
"it could look like this", with lots of open holes. :] But yes, it might work, if if if .

[02:03:22.0000] <abarth_>
i'm actually shocked the json.js think worked out ok

[02:03:37.0000] <abarth_>
thought we'd have tons of pain with folks using loose JSON syntax

[02:03:43.0000] <abarth_>
but it seems to have worked

[02:04:32.0000] <hsivonen>
abarth_: IIRC, adding JSON.parse to Firefox broke Facebook

[02:04:32.0000] <annevk>
odinho: so Drupal no longer really cares about validator conformance?

[02:04:41.0000] <MikeSmith>
yeah but this is about markup so there's a broader set of people with opinions about it

[02:04:50.0000] <hsivonen>
abarth_: fortunately, Facebook is exceptionally responsive to problems like that

[02:05:13.0000] <hsivonen>
abarth_: I don't expect the same responsiveness from Drupal

[02:05:15.0000] <annevk>
yeah Facebook is awesome

[02:05:19.0000] <abarth_>
hsivonen: oh, I didn't know that

[02:05:33.0000] <odinho>
annevk: "It will be valid in the future, so it's okay!"

[02:05:37.0000] <abarth_>
yeah, Facebook has a good deployment process

[02:05:37.0000] <hsivonen>
even if the Drupal core team was responsive, it would be a pain to get the update deployed everywhere

[02:05:42.0000] <odinho>
annevk: They said that in thet thread :]

[02:05:56.0000] <annevk>
odinho: brilliant

[02:06:08.0000] <MikeSmith>
abarth: speaking of new elements, you all's <browser> element looks intriguing

[02:06:08.0000] <annevk>
odinho: but then Drupal added RDFa too I think

[02:06:14.0000] <MikeSmith>
abarth: nice video btw

[02:06:15.0000] <annevk>
odinho: full of silly

[02:06:56.0000] <hsivonen>
yeah, don't trust Drupal on the future of markup

[02:06:58.0000] <MikeSmith>
annevk:  <menuitem>

[02:08:27.0000] <abarth>
MikeSmith: thanks

[02:08:35.0000] <abarth>
yeah, the browser tag is something Firefox has had for a while

[02:08:43.0000] <MikeSmith>
oh

[02:08:45.0000] <MikeSmith>
didn't know

[02:08:55.0000] <MikeSmith>
I thought you all had minted it

[02:08:56.0000] <abarth>
well, its only exposed to "chrome" content in Firefox

[02:09:10.0000] <MikeSmith>
yeah I know

[02:09:15.0000] <abarth>
the trick with this sort of thing is to get the API right

[02:09:16.0000] <hsivonen>
abarth: we have <browser> for HTML??? I thought only XUL

[02:09:23.0000] <abarth>
yeah, XUL

[02:10:14.0000] <hsivonen>
If I understand correctly, the Gaia browser will use a special iframe that looks like a top-level browsing context from within

[02:10:30.0000] <abarth>
the problem is that iframe has the wrong API for a bunch of uses cases

[02:10:36.0000] <abarth>
e.g., you can't really build a browser out of it

[02:11:12.0000] <abarth>
I think the apps folks are going for something that has an API more like UIWebView

[02:11:17.0000] <abarth>
or the Android WebView

[02:11:40.0000] <hsivonen>
abarth: I believe there's going to be a magic attribute that morphs iframe into being more suitable

[02:12:05.0000] <abarth>
ok, then it might be a similar concept

[02:12:11.0000] <abarth>
why not use a new tag rather than morphing an iframe?

[02:12:42.0000] <hsivonen>
I'm not sure. I didn't participate in the design

[02:12:43.0000] <abarth>
in Chrome, the implementation is very different than iframe

[02:12:56.0000] <abarth>
it's basically a web component that hides a plugin

[02:13:12.0000] <abarth>
that embeds WebKit similarly to how a WebView embeds webkit

[02:13:18.0000] <hsivonen>
Chrome as a plug-in in Chrome?

[02:13:23.0000] <abarth>
yes

[02:13:27.0000] <hsivonen>
yo dawg

[02:13:29.0000] <MikeSmith>
wow

[02:13:34.0000] <MikeSmith>
hahaha

[02:13:37.0000] <abarth>
well, it even more awesome in Chrome Frame :)

[02:13:52.0000] <MikeSmith>
inception

[02:14:02.0000] <abarth>
the benefit of this technical design is that the content runs in a separate process due to out-of-process plugins

[02:14:14.0000] <abarth>
that means we get a strong security boundary between the app and the web content

[02:14:37.0000] <hsivonen>
I believe B2G runs the Gaia browser UI and the Web content in separate processes, too

[02:15:01.0000] <abarth>
using the morphing iframe?

[02:15:26.0000] <hsivonen>
so I've understood, but I'm really basing what I say on hearsay. I don't work on that stuff.

[02:16:42.0000] <hsivonen>
abarth: see https://wiki.mozilla.org/Security/Reviews/B2GBrowser

[02:17:37.0000] <abarth>
the goals are the same

[02:17:43.0000] <abarth>
sounds like the syntax is different

[02:17:54.0000] <abarth>
i'm hopeful that the SysApps WG can sort this sort of stuff out

[02:19:25.0000] <hsivonen>
http://mozillamemes.tumblr.com/post/27554392258/have-you-checked-out-the-source-for-the-gaia

[06:25:03.0000] <hsivonen>
"I couldn't find any of the original sites still responding, so it's hard to know if there is still a compat. problem here. If there isn't, the greater conceptual simplicity of the WebKit model is quite appealing."

[06:25:10.0000] <hsivonen>
not the first time :-(

[06:35:46.0000] <AryehGregor>
?

[06:38:36.0000] <hsivonen>
AryehGregor: it seems to be a recurring theme that 1) Gecko does something simple. 2) Some site breaks. 3) Gecko changes to do something complicated. 4) Years later WebKit does the simpler thing. 5) The site that broke in Gecko is gone.

[06:38:59.0000] <AryehGregor>
Well, nothing to complain about there.  We get to simplify our code, yay.

[06:40:08.0000] <hsivonen>
I'm slightly unhappy that Canonical introduces a system-specific notification API instead of using the multivendor API: http://developer.ubuntu.com/api/ubuntu-12.04/javascript/index.html

[06:41:13.0000] <odinho>
lol. Think I found a super stupid bug in testharness.js that I've been irritated by.

[06:44:55.0000] <Namtnad>
hsivonen: Me too. They even reinvented the notification wheel when we already have a notification spec that's already implemented.

[07:00:47.0000] <odinho>
https://github.com/w3c/testharness.js/pull/6 <- anyone want to r+ it quickly? :P

[07:58:40.0000] <gsnedders>
espadrine`: That's most certainly a bug.

[09:52:00.0000] <Hixie>
odinho: do you know why they don't like srcset=""? looks like it already handles their case, but maybe i'm missing something

[09:52:51.0000] <odinho>
Hixie: They want even more control. It's the "artistic use case" (or whatever we should call it) that is not 100% covered by that one.

[09:53:17.0000] <Hixie>
what more control can one have than with srcset=""?

[09:53:29.0000] <odinho>
Although I think that is something that will get *much* smaller usage, because it's much more specialized.

[09:53:49.0000] <Hixie>
(btw, on another note, http://tools.ietf.org/html/draft-karcz-uuas-00 claims to have no security implications, which is kinda funny given that this is a big part of the bits used by fingerprinting)

[09:54:17.0000] <Hixie>
(quite aside from the issue of speccing something without implementors, which appears to be going on here)

[09:54:54.0000] <odinho>
Hixie: Well, one "problem" is that it favours desktop first, over mobile first. Because the algorithm has to work (taking away pictures) from one direction.

[09:55:37.0000] <odinho>
Although I must confess I have mostly forgotten why they found srcset lacking at all. I was always rooted in the browser-vendor camp.

[09:55:42.0000] <Hixie>
not sure what you mean by "favour"

[09:56:08.0000] <Hixie>
the problem is intrinsically assymetric, since it's working on number lines that go from 0 to positive infinity

[09:56:20.0000] <Hixie>
but i don't see how that favours one side or the other

[09:57:32.0000] <Hixie>
i guess i should be going through this feedback, anyway

[09:57:36.0000] <Hixie>
got like 300 e-mails on the subject

[09:58:03.0000] <Hixie>
my todo list currently looks like: inputmode, autocompletetype, handlers/intents, responsive images

[09:58:05.0000] <odinho>
Yeah, I have written some of those :P But got a bit in need for a vacation after a bit. There's quite some repetition going on there.

[09:59:07.0000] <dglazkov>
good morning, Whatwg!

[10:00:22.0000] <Hixie>
inputmode, autocompletetype, handlers/intents, responsive images, then bugs, starting with WebVTT stuff, looks like

[10:01:07.0000] <Hixie>
abarth|zZz++

[10:01:22.0000] <abarth>
what did I do?

[10:01:29.0000] <Hixie>
your e-mail just now re legacy systems

[10:01:33.0000] <abarth>
ah :)

[10:01:35.0000] <abarth>
thanks

[10:02:41.0000] <Hixie>
MikeSmith: so for inputmode, i'm considering having the spec define a lot of modes, but for each one say which one to fall back to if you don't support it

[10:02:54.0000] <Hixie>
MikeSmith: basically, a tree of modes, all but the root being optional, and the root being "user default"

[10:03:22.0000] <odinho>
Hixie: Well, it's mostly about things like orientation:portrait, and stuff like being able to choose between min-width  and  max-width.

[10:03:52.0000] <odinho>
Hixie: That said, towards the end of the thread tsunami, a few people started understanding the proposal better and were not so hostile any more.

[10:04:54.0000] <odinho>
Hixie: So I guess much of the initial reactions and emails are coming more from prejudice than actual understanding.

[10:05:43.0000] <[tm]>
Hixie: interesting

[10:06:40.0000] <Tabatkins>
odinho: Yes, nearly all of the reaction in the first 24 hours can be thrown out, as it was almost completely based on misunderstanding.

[10:07:05.0000] <[tm]>
different approach but seems worth exploring i guess

[10:07:07.0000] <Hixie>
[tm]: an alternative is to have authors list tokens, and the UA picks the first one it implements

[10:07:39.0000] <Hixie>
[tm]: but that seems like it would be giving authors the job that we can just do, since it's not clear that you'd ever want a _different_ fallback

[10:07:58.0000] <Hixie>
odinho: can you elaborate on the orientation and min vs max thing?

[10:10:03.0000] <[tm]>
Hixie: btw maybe you saw i added a screen capture of desktop Japanese keyboard switcher

[10:10:06.0000] <odinho>
Hixie: Lots of small things to tweak. I don't think anyone actually came with a really compelling case of what they wanted to do, and where the mediaqueries were a much better fit. I had an email saying how you could do min-max with srcset - response was that it was ugly :-) Which is kinda true, but well.

[10:15:58.0000] <odinho>
Hixie: Ah, before I thought that the max-width mediaqueries etc would work on the *content size* around it, so a picture in a small column would actually load something that only fit in there.

[10:16:18.0000] <Hixie>
[tm]: yeah that's what made me think of this

[10:16:27.0000] <Tabatkins>
It's unfortunate, but that doesn't work very well. :/

[10:17:05.0000] <odinho>
But I realized that max-width in mediaqueries refer to the viewport as well, so that was my misunderstanding in the beginning about mediaqueries being more flexible and it actually having some useful flexibility.

[10:17:12.0000] <Hixie>
yeah we can't really do anything but viewport sizing, sadly

[10:17:20.0000] <Hixie>
my original idea was also based on content size but that's a nonstarer

[10:17:27.0000] <Hixie>
since you need to do the calculations before you have the content

[10:17:36.0000] <odinho>
Hixie: Unless you do it specifically for your site in script.

[10:17:47.0000] <Hixie>
well sure, but then you don't need any of this

[10:17:49.0000] <odinho>
(and to do that, you'd need to defer image load).

[10:18:02.0000] <Hixie>
deferring image load is easy :-)

[10:18:05.0000] <odinho>
Hixie: Yes, -- you want to hinder the browser from prefetching before doing javascript.

[10:18:32.0000] <odinho>
Hixie: well,   <img data-src=img.jpg><noscript><img src=img.jpg></noscript>  <-- it's very ugly.

[10:18:55.0000] <Hixie>
just <noscrpt><img src="..." alt="..."></noscript>

[10:19:24.0000] <Hixie>
script can figure out the rest

[10:20:41.0000] <odinho>
Yes, true :-) Hmm.

[10:21:49.0000] <Hixie>
or if you don't care about script-free cases and if the image is purely decorative, just something like <div data-src="..."></div>

[10:22:07.0000] <odinho>
Problem is that lazy-loading on scroll still poses a problem for Opera Mini. We cut execution after some seconds of js run, -- and we've had problems with many sites optimizing by lazy-loading images when you scroll down to them, and at that point, you just get white boxes, so mini breaks there.

[10:22:41.0000] <odinho>
Although I think Mini team has hacked around several of those things, maybe even done a real fix, I don't follow Mini too closely, just noticed breakage on my own phone some time ago.

[10:22:48.0000] <Hixie>
mini imho is not a conforming approach to a web browser :-)

[10:22:57.0000] <odinho>
WAT! :-O :P

[10:24:04.0000] <odinho>
But somewhat unrelated to that, - optimizing picture loading should be the user agent's job. Sadly that won't work because many sites expect pictures to be loaded before onload is fired etc.

[10:24:18.0000] <odinho>
So mobile browsers always load everything.

[10:24:54.0000] <odinho>
Having a @defer would allow to actually not to wait on those images. Even not loading them until you've scrolled down.

[10:25:00.0000] <Hixie>
afk bbiab to work on inputmode

[10:41:46.0000] <Yuhong>
After the Last Call, CSS 2.1 became CR in early 2004:

[10:41:47.0000] <Yuhong>
http://www.w3.org/TR/2004/CR-CSS21-20040225/

[10:42:07.0000] <Yuhong>
But by the time IE7 beta 1 was released, it has been pulled back to WD:

[10:42:12.0000] <Yuhong>
http://www.w3.org/TR/2005/WD-CSS21-20050613/

[10:42:31.0000] <Tabatkins>
Yes, CR back then was still insufficiently rigorous.

[10:42:32.0000] <Yuhong>
And did not become CR until after IE7 was released.

[10:42:39.0000] <Yuhong>
And did not become CR again until after IE7 was released.

[10:42:57.0000] <Yuhong>
And IE7 support of CSS 2.1 features was pretty limited.

[10:43:33.0000] <Yuhong>
*still pretty limited.

[10:43:44.0000] <Yuhong>
http://www.w3.org/TR/2007/CR-CSS21-20070719/

[10:44:09.0000] <Wilto>
odinho: The current proposal for `picture` covers the min/max-width use case while using `srcset` for resolution switching only, for what that’s worth. http://www.w3.org/community/respimg/2012/06/18/florians-compromise/

[10:45:25.0000] <Yuhong>
MS was working on IE8 by that time, and rewrote the CSS layout engine likely based on this CR.

[10:45:52.0000] <Yuhong>
Soon after IE8 was released, another CR was published: http://www.w3.org/TR/2009/CR-CSS2-20090423/

[10:46:40.0000] <Yuhong>
It ended up being pulled back to WD: 20101207/

[10:46:55.0000] <Yuhong>
It ended up being pulled back to WD again: http://www.w3.org/TR/2010/WD-CSS2-20101207

[10:47:02.0000] <Yuhong>
But that didn't last long.

[10:47:31.0000] <Yuhong>
It became PR soon after IE9 was released: http://www.w3.org/TR/2011/PR-CSS2-20110412/

[10:47:38.0000] <Ms2ger>
Are you trying to make a point?

[10:47:40.0000] <Ms2ger>
If so, make it

[10:48:10.0000] <Yuhong>
How IE development relates to CSS 2.1 spec development.

[10:50:45.0000] <odinho>
Wilto: Yes, but can you tell me what that is (with an example)? I believe you can do what you want with the current srcset proposal as well.

[10:52:07.0000] <Wilto>
You could for certain, but it would require a completely redundant syntax pattern. There’s no benefit to the `srcset` markup, for that purpose.

[10:52:32.0000] <Wilto>
For resolution switching there’s a great deal of benefit to `srcset`, both implementor and author side.

[10:54:05.0000] <odinho>
Wilto: No, but there's more to weigh for and against than just that case. And it's a very high price to pay for an extra convenience.

[10:54:49.0000] <Wilto>
What’s the benefit to using `srcset` for min/max width and such?

[10:55:18.0000] <Wilto>
I mean, assuming it’ll be expanded to cover both cases and units other than pixels.

[10:57:43.0000] <Wilto>
Just seems like a lot of reinvention for one. For another, once the appropriate _size_ image has been selected, the resolution seems like it’s better handled as an isolated concern.

[10:57:43.0000] <odinho>
Wilto: Plus for srcset; don't have to make a new element. Don't have to pull in the entirety of mediaqueries for picture loading (which is more work). It'll be shorter to write. With mediaqueries you *have* to follow them, with srcset the implementation can choose to optimize if it wants to (although it's not recommended to go outside the author's wishes in the common case, but uncommon cases exist).

[10:58:09.0000] <odinho>
Wilto: It won't be expanded for other things than pixels. Only pixels matter for screen size.

[10:58:20.0000] <Wilto>
I use almost strictly em-based media queries.

[10:58:28.0000] <Wilto>
That’s a very common practice.

[10:58:32.0000] <Tabatkins>
odinho: That's silly.  Ems are useful for plenty of things.

[10:58:34.0000] <odinho>
Wilto: By the way, it can't work in media queries either when doing preloading.

[10:58:43.0000] <Tabatkins>
Huh?

[10:59:03.0000] <Wilto>
I was told several times on the mailing list that viewport information is available well before preload.

[10:59:11.0000] <odinho>
Tabatkins: You have to know what size the em is.

[10:59:15.0000] <Tabatkins>
That's well-defined.

[10:59:21.0000] <Tabatkins>
It's the user's default font size.

[10:59:36.0000] <odinho>
But it's just a constant?

[10:59:48.0000] <Tabatkins>
(Which you shouldn't be toying with for body text anyway, and body text em size is what you are likely breaking against if you use ems.)

[10:59:48.0000] <odinho>
So it won't be different from px anyway.

[10:59:56.0000] <Tabatkins>
It's a constant *per user*.

[10:59:59.0000] <Wilto>
Unless the user zooms.

[11:00:00.0000] <Tabatkins>
Not something the author can predict.

[11:00:20.0000] <odinho>
Okay, didn't know that. CSS is not my field :-)

[11:00:31.0000] <hober>
essentially, em=rem in mqs

[11:00:46.0000] <odinho>
So how would that work?

[11:00:46.0000] <Tabatkins>
hober: Still not right.  rem can be changed by the author.

[11:01:17.0000] <Tabatkins>
odinho: How would what work?

[11:01:39.0000] <odinho>
Tabatkins: em's in srcset or in media max-width/min-width.

[11:01:52.0000] <odinho>
(for images, that is)

[11:02:43.0000] <odinho>
I've seen the behavior for mediaqueries on sites, which is kinda cool effect. :P I can't seem to apply that sensibly to the image case.

[11:03:15.0000] <Tabatkins>
Same deal.  You may want to use different images for different breakpoints on your site.

[11:03:22.0000] <Tabatkins>
And if your breakpoints are em-based, which is very reasonable...

[11:04:41.0000] <Tabatkins>
If you're switching from a two-column to single-column layout, for example, even though you don't know exactly what size an 'em' is, you at least know that your layout is, say, 2/3rds the size that it was, and so you supply an image that's narrower.

[11:06:23.0000] <Wilto>
odinho: I 100% agree that the UA should have final say in the assets that get requested, though, in the case of resolution. I think `srcset` gives us a great opportunity to introduce syntax that works as… a suggestion, overrideable by the UA/users’ settings. I wouldn’t want to introduce that kind of variance to media queries either.

[11:07:39.0000] <Wilto>
First the correct size is selected, in absolutes. Then `srcset` determines and delivers the optimal resolution, unless overridden by the browser. “Always download high-res,” “always download low-res,” “download high-res as bandwidth permits,” etc.

[11:07:43.0000] <Tabatkins>
I agree that in the "art direction" case (the one that cares about min/max size), you should exactly honor what the author wants.  In the "quality case" (the one that cares about resolution), you should let the UA decide what's best.

[11:07:53.0000] <Wilto>
Jinx, basically.

[11:07:57.0000] <Tabatkins>
haha, yes.

[11:08:32.0000] <odinho>
Designing a web site, I say   <img src=myimg-2000px.jpg srcset="myimg-4000px.jpg 2x">  <-- I have a big picture there. But my phone is also 2x, and so it'll download the 4000px one, although that's grossly too much for it.

[11:09:13.0000] <odinho>
Pulling in the entire new tag with all its requirements feels like a very heavy handed thing to do for such a common case.

[11:09:50.0000] <Wilto>
I mean, buzzwords aside, there’s no denying that a single code base approach across a huge range of devices is becoming increasingly common.

[11:09:56.0000] <odinho>
Whilst much of the art direction and advanced usage I'm sure will see much smaller use.

[11:10:48.0000] <odinho>
And so that's part of the reason why I'd really like to have something short and low on requirements for that case.

[11:11:17.0000] <Wilto>
Well, here’s the thing, there:

[11:11:22.0000] <odinho>
And rather make it possible for script (or maybe even CSS) to dictate much more advanced "art direction" uses later.

[11:11:31.0000] <Wilto>
Your example makes perfect sense. `<img src=myimg-2000px.jpg srcset="myimg-4000px.jpg 2x">`

[11:11:45.0000] <Wilto>
I mean, really, I’d prefer to have that. `srcset` for resolution switching only.

[11:11:49.0000] <odinho>
So that it's possible to do advanced stuff again, and even in a nice way.

[11:12:05.0000] <Wilto>
And later on, should we need to expand to a new element for the media query use case, we’d have `srcset` already available.

[11:12:18.0000] <Wilto>
There’s no reason it couldn’t be used on `img` to handle the resolution-switching case alone.

[11:12:58.0000] <Wilto>
I would just want to avoid inventing an ever-expanding syntax for `srcset` to cover the use case better served by MQ.

[11:13:03.0000] <odinho>
But for mobile phones I'd really like size switching straight away.

[11:13:06.0000] <Wilto>
`srcset` is the better solution to resolution.

[11:13:11.0000] <Wilto>
Oh, trust me: so would I.

[11:13:28.0000] <Wilto>
I just worry that we’d paint ourselves into a corner. What if it had been implemented with pixels only, y’know?

[11:13:42.0000] <Wilto>
Authors would eventually just have to find ways to work around the limitations we’d stand to introduce.

[11:14:08.0000] <Wilto>
We’d be duplicating the uses of media queries, over time. Two syntaxes developed in parallel to accomplish the same end goal.

[11:14:39.0000] <Wilto>
`srcset` is a great solution for what it set out to solve: 1x, 2x, and so on.

[11:14:43.0000] <odinho>
Well, it would work much much better than today. Where there is *nothing*. And I'm really not convinced that pixels for images are so bad, they are made of pixels themselves.

[11:15:15.0000] <odinho>
Wilto: "set" set out to solve 1x 2x. srcset was to solve both resolution, and also sizes of the same image :-)

[11:15:22.0000] <Wilto>
If the entire page is resized in ems, though, that image is no longer layout-appropriate.

[11:15:37.0000] <odinho>
hober: I think you called it 'set'?

[11:16:09.0000] <Tabatkins>
The CSS image-set() function, you mean?

[11:16:21.0000] <odinho>
Tabatkins: Yea, but he sent a version for HTML as well.

[11:16:26.0000] <Wilto>
The original proposal that became `srcset` was `set` I think, yeah.

[11:16:29.0000] <Tabatkins>
Oh, yeah, right.

[11:16:30.0000] <Wilto>
And just solved resolution.

[11:17:01.0000] <Tabatkins>
odinho: You're not understanding the pixels issue.  It's not very much about the image itself, it's about aligning with the breakpoints you're already defining in MQ.

[11:17:41.0000] <odinho>
Wilto: Not quite right either, they started out differently, but the unnamed proposal took in 'set' proposal and built on that when it came.

[11:17:46.0000] <Tabatkins>
And those can be defined in units other than px, and it's completely reasonable and useful to do so (not just something that's *possible* but a bad idea).

[11:18:03.0000] <odinho>
Wilto: hober saw Hixie was working on it, and so hober sent out his email that had been waiting as a draft.

[11:18:12.0000] <odinho>
Wilto: Ohwell, techincalities.

[11:20:57.0000] <odinho>
Tabatkins: Yea, I see that, but I'm not convinced of its weight. Some part about making the 90% of what you want to do easy, and leaving out the 10% that would steeply raise the complexity level.

[11:21:17.0000] <Wilto>
odinho: Raise the complexity level for authors, you mean?

[11:21:30.0000] <Tabatkins>
Sure, that's a valid argument in general, but the *whole point* of the art-direction side of srcset is to match your breakpoints.

[11:21:35.0000] <odinho>
And an easy fix like @defer would in fact be even more flexible than mediaqueries.

[11:21:55.0000] <odinho>
*And* it would actually be able to load images based on the *content size*. Which even mediaqueries is not able to do.

[11:22:11.0000] <Tabatkins>
What does <img defer> do?

[11:22:13.0000] <odinho>
And you could build it into your CMS in a very nice way. And do lots of nice stuff :-)

[11:22:16.0000] <Wilto>
Content size won’t be available at the time of prefetching, though.

[11:22:19.0000] <odinho>
It's the magic shampoo! :D

[11:22:29.0000] <Wilto>
You can’t know the available container size before the layout has been rendered.

[11:22:37.0000] <Tabatkins>
It's a floor wax and a dessert topping!

[11:22:39.0000] <Wilto>
Plus, with @defer, we lose prefetching altogether.

[11:22:52.0000] <odinho>
Wilto: Yes, that's what defer does, - it doesn't prefetch the image, but lets early javascript handle it in any way it wants.

[11:23:02.0000] <odinho>
Wilto: We didn't have prefetching until not so long ago.

[11:23:15.0000] <odinho>
Wilto: And @defer would be an opt-in.

[11:23:45.0000] <odinho>
Also, many sites do @defer on their own now, -- by using <div>'s and exchanging them with images when you scroll to them (and fades them in).

[11:24:13.0000] <Wilto>
Oh, yeah, we did something similar with Picturefill.

[11:24:33.0000] <Wilto>
https://github.com/scottjehl/picturefill

[11:24:43.0000] <Wilto>
I mean, that’s no long term solution, for obvious reasons.

[11:24:54.0000] <odinho>
Wilto: That is something I've looked at and thought about when discussing defer.

[11:25:26.0000] <Wilto>
Don’t get me wrong: I like the idea of having defer available.

[11:25:40.0000] <Wilto>
But I think you’re just gonna end up with a bunch of authors writing their own Picturefills.

[11:25:49.0000] <odinho>
Many mobile sites defer images already, -- it should be up to the browser to do that. Sadly it's not possible to do that (Opera tried) because the web expects every image to be loaded and ready when you fire onload.

[11:26:13.0000] <odinho>
Wilto: Defer would work something like picturefill on its own.

[11:26:52.0000] <odinho>
Ok, my idea with it was, -- when onload *has* run, and if not some script has explicitly said that it will do the picture loading, -- the browser may load them as it pleases.

[11:27:09.0000] <Wilto>
That’s fair, and I think it would be a great feature.

[11:27:25.0000] <Wilto>
I just don’t consider it a solution for this issue in and of itself, y'know?

[11:28:11.0000] <odinho>
So if you do (made up)    img:unloaded { opacity: 0.0 }   img:loaded { opacity: 1.0; transition: opacity 1s; }   then they might fade in on scroll-down just like picturefills do now. But without any javascript to delay the image load.

[11:28:53.0000] <odinho>
Obviously mostly mobile browsers would /maybe/ like to defer the images for loading like that.

[11:29:57.0000] <Wilto>
Er, Picturefill doesn’t do that. I think we might be talking about two different things.

[11:30:12.0000] <odinho>
Wilto: I'm thinking it would be a solution to the missing 10%, and it could possibly also enable other new features.

[11:30:19.0000] <Wilto>
I mean, I agree: that would be useful, but completely independent of the `picture`/`srcset` stuff.

[11:30:28.0000] <odinho>
Wilto: Oh, no, I was actually talking about something else now. :]

[11:30:32.0000] <Wilto>
Ah, okay.

[11:30:52.0000] <Wilto>
Yeah, I mean, I would not mind having @defer one bit, as an author.

[11:31:07.0000] <odinho>
I mean, having @defer to be able to do the last mile of the art directed respimg things, -- and *in the process of adding that, also getting some other new stuff for "free" (ohwell :P)*

[11:31:08.0000] <Wilto>
It's Not That Simple™ I'm sure, but I wouldn't be mad about it.

[11:33:23.0000] <odinho>
It would open up the posibility of actually doing content size dependent loads. So although those last 10% would loose prefetch, they would actually gain some extremely powerful tools.

[11:33:52.0000] <odinho>
s/last 10%/last 10% use cases/

[11:34:02.0000] <odinho>
Not people :-)

[11:37:49.0000] <Wilto>
I just don’t think that’s the end-all-be-all solution people are after, nice though it would be in a lot of ways.

[11:38:37.0000] <hober>
odinho: my original html proposal named the attribute srcset=""

[11:38:56.0000] <odinho>
hober: Oh, okay. Might be the other way around then :S

[11:43:07.0000] <hober>
I think searching for an end-all-be-all solution is a mistake in the first place

[11:43:37.0000] <tantek>
`hober, I kind of stopped paying attention when people kept arguing without referring to specific use-cases.

[11:43:54.0000] <tantek>
re: end-all-be-all - they're not even trying to do *that*

[11:44:10.0000] <tantek>
if they were, they'd actually reference the "alls"

[11:44:22.0000] <hober>
heh

[11:44:39.0000] <tantek>
/me is kind of tired of feature design/invention by "intuition"

[11:45:34.0000] <tantek>
such made up features (without citing documented use-cases) are no better than a Volcano with a fax number (no matter how cool you think that might be)

[11:45:54.0000] <Wilto>
tantek: This `defer` thing, you mean?

[11:46:14.0000] <tantek>
Wilto - can you cite the page documenting use cases of the 'defer' thing?

[11:46:21.0000] <odinho>
Wat!

[11:46:32.0000] <tantek>
if your answer is anything other than a URL, then you're doing it wrong.

[11:46:38.0000] <Wilto>
Woah, tantek, that’s not my deal.

[11:46:42.0000] <tantek>
;)

[11:46:51.0000] <Wilto>
odinho brought it up, and I didn’t say much more than “that could be cool, sure.”

[11:47:01.0000] <tantek>
so that's part of the problem IMHO

[11:47:10.0000] <tantek>
we need to not be so accepting of proposals without use-caess

[11:47:13.0000] <tantek>
cases even

[11:47:20.0000] <odinho>
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2012-February/034840.html

[11:47:26.0000] <tantek>
it's why feature bloat happens :(

[11:47:36.0000] <odinho>
tantek: I'm against bloat.

[11:47:59.0000] <tantek>
much better

[11:48:30.0000] <odinho>
tantek: Actually we (as in Opera) wanted to defer image loading on Mobile.

[11:49:04.0000] <tantek>
odinho - the scenarios described make sense and do sound related to responsive image scenarios

[11:56:27.0000] <tantek>
odinho, consider incorporating that use-case description into the broader set of responsive image use-cases described here: http://www.w3.org/wiki/Images

[11:56:33.0000] <tantek>
hopefully it will make them easier to find

[11:56:55.0000] <Ms2ger>
Fun, hitting ctrl+a in opera-next crashes

[12:00:18.0000] <tantek>
sigh, apparently /Images was too hard to find on the w3c wiki, so someone started http://www.w3.org/wiki/Adaptive_Images

[12:00:37.0000] <tantek>
or maybe I'm missing something? Wilto, odinho, is there a difference in framing between "responsive" images and "adaptive" images?

[12:01:25.0000] <Wilto>
Oh, no, both terms are used interchangeably. “Determining the most appropriate image source for the client’s capabilities,” if I had to cram it onto a business card.

[12:01:35.0000] <Wilto>
Which I pray I never have to do.

[12:02:39.0000] <Wilto>
Eesh, never seen http://www.w3.org/wiki/Adaptive_Images before. Maybe that’s an older one.

[12:04:22.0000] <tantek>
anyway - I leave it to those of you more passionate about the subject matter to reconcile/merge/cleanup the pages accordingly for better referencing / citability in the future (to encourage more rational discussion)

[12:07:47.0000] <Wilto>
Is the CG’s proposal worth posting on the WHATWG mailing list, despite “going to the HTML WG?”

[12:08:07.0000] <Wilto>
I’m not sure how clear that divide is supposed to be; I’m just a pixel-pusher in a strange land.

[12:08:27.0000] <Wilto>
I figure it’s better to have everyone on the same page rather than not, me.

[12:08:28.0000] <tantek>
Wilto - I'm not sure how productive the different email paths are either at this point.

[12:09:26.0000] <tantek>
I find put everything substantial (even disagreements) on a wiki page the best way get everyone on the same page.

[12:09:27.0000] <tantek>
(so to speak)

[12:09:40.0000] <Wilto>
(Ba-dum tss.)

[12:09:44.0000] <tantek>
:)

[12:10:12.0000] <tantek>
and then sending tweet-length emails with URL to said page is a reasonable notification mechanism to all the lists of people that care

[12:10:44.0000] <Hixie>
afk bbiab to work on inputmode

[12:10:46.0000] <Hixie>
er

[12:10:47.0000] <Wilto>
I mean, for the RICG’s stance, http://www.w3.org/community/respimg/2012/06/18/florians-compromise and the proposal itself at http://www.w3.org/community/respimg/wiki/Picture_Element_Proposal are kinda the final word. I guess a quick post couldn’t hurt, there.

[12:10:57.0000] <Hixie>
mistype

[12:13:03.0000] <tantek>
great, yet another W3C wiki-silo rather than using w3.org/wiki

[12:13:41.0000] <Hixie>
Wilto: anything you want to have impact the whatwg spec should be discussed on the whatwg list

[12:13:42.0000] <tantek>
(each wiki silo requires its own login etc. - I have no idea why W3C CGs made the creation of such silos the default)

[12:14:03.0000] <Hixie>
Wilto: i do try to read stuff outside the whatwg list, but i only guarantee to see and respond to stuff on the list

[12:14:18.0000] <tantek>
Wilto - maybe add links here: http://www.w3.org/wiki/HTML/next#Multimedia

[12:14:20.0000] <Hixie>
(and in practice, bugs in the WHATWG product HTML component assigned to ian⊙hc in the W3C Bugzilla)

[12:14:42.0000] <Hixie>
(but i sometimes just close those without detailed response, unlike e-mails to the list, where i try to always give a considered response)

[12:14:47.0000] <Hixie>
(so if you want a response, use the list)

[12:15:32.0000] <Wilto>
Hixie: Yeah, understandable. I’m a little wind-swept when it comes to what goes where -- lots of advice coming from lots of people, when one just wades into this particular pool.

[12:16:09.0000] <Hixie>
well basically there are two specs now, the WHATWG HTML spec and the W3C HTML5 spec, and they are independent

[12:16:21.0000] <Wilto>
( Guessing it’s best to not resurrect the ancient thread on the topic, too. )

[12:16:31.0000] <Hixie>
you have to make your own determination abotu which spec you care about :-)

[12:16:50.0000] <Hixie>
if you want to affect the W3C list, you want to file a bug on the W3C spec (W3C Bugzilla, HTML WG product, HTML5 spec component)

[12:16:54.0000] <Wilto>
Hah—which is the one where I can use these things on a website someday?

[12:17:10.0000] <Hixie>
well i guarantee that on the long term the WHATWG spec will match browsers

[12:17:20.0000] <Hixie>
but which browsers are following, i can't say really

[12:17:23.0000] <Hixie>
i hope the WHATWG one

[12:17:25.0000] <Hixie>
but we'll see

[12:17:28.0000] <tantek>
Wilto, even if it isn't official, HTML5 is fairly feature frozen these days.

[12:17:57.0000] <tantek>
Best bet for more rapid consideration / publication of new HTML features (via a spec) is to the WHATWG.

[12:18:05.0000] <Wilto>
Yeah. The HTML WG was discussing this as “HTML Next,” assuring me that didn’t necessarily mean a delay in implementation.

[12:18:13.0000] <Hixie>
tantek: my understanding is that the W3C is expecting to do a parallel play of HTML.next sooner than they finish HTML5

[12:18:13.0000] <Wilto>
But... one wonders.

[12:18:25.0000] <Hixie>
Wilto: well, implementations are independent of specs

[12:18:38.0000] <tantek>
Hixie, I expect to see the HTML.next effort adopt features developed from/in WHATWG

[12:18:39.0000] <Hixie>
Wilto: delay in the spec just means browsers make up their own stuff :-)

[12:18:52.0000] <tantek>
yeah

[12:19:14.0000] <Hixie>
tantek: from your lips to the working group's actions

[12:19:31.0000] <Hixie>
/me fails to find a good rephrasing of "to god's ears"

[12:19:33.0000] <tantek>
Hixie - it's certainly what I'll be contributing to the WG

[12:19:44.0000] <Wilto>
Yeah. There are a couple of implementors ready to haul off and do this; I just assumed it made sense to push for a spec first.

[12:19:56.0000] <tantek>
better to discuss more considered features rather than less developed ones

[12:20:26.0000] <Hixie>
Wilto: well on the whatwg side i watch the (browser) implementors closely and if they go off in one direction, the spec will follow shortly.

[12:20:42.0000] <Hixie>
Wilto: i always like it if they go off in the same direction as i already did with the spec, since that makes my life easier :-)

[12:20:48.0000] <tantek>
Wilto - your speediest chances for specification are likely to contribute to WHATWG, by whatever means the editor requests (modulo, he's been known to also accept emails with wiki page citations :) )

[12:21:17.0000] <Hixie>
like i said above, if you want a response to an argument, make sure to include the argument in the e-mail :-)

[12:21:27.0000] <tantek>
^^^

[12:21:38.0000] <Hixie>
citing a wiki page is fine but my response will likely just be "thanks, i read the wiki page and did what i thought was best considering it" :-)

[12:21:50.0000] <tantek>
good to know

[12:22:11.0000] <Hixie>
so if you care about results rather than knowing why the results came about, that's good enough :-)

[12:22:11.0000] <Wilto>
Naturally—I’ll post the wiki page and I’ll be sure to include the reasoning detailed in the accompanying post.

[12:22:22.0000] <Hixie>
(the reason being it's hard for me to quote wiki content in e-mail)

[12:22:25.0000] <tantek>
providing a summary of arguments in email is probably helpful then

[12:22:32.0000] <Hixie>
(or rather, harder than i'm willing to bother with)

[12:22:54.0000] <Wilto>
http://www.w3.org/community/respimg/2012/06/18/florians-compromise/ is probably a little long and prose-y for an email. I’ll pare it down.

[12:22:56.0000] <Hixie>
Wilto: btw if the points have already been made in the 300+ e-mail thread, then there's no need to post them again

[12:23:07.0000] <tantek>
Hixie, I've come across that in other discussions too - where the barrier to quotation/citation is what appears to be limiting the discussion.

[12:23:10.0000] <Hixie>
Wilto: i do consider each point even if it's deeply embedded in a thread

[12:23:20.0000] <Hixie>
tantek: yeah

[12:23:30.0000] <Wilto>
Some of it was covered in Florian’s posts; not all.

[12:24:05.0000] <Hixie>
anyway, you have a few days, as my current priorities are inputmode, autocompletetype, handlers/intents, then responsive images, then bugs, starting with WebVTT stuff

[12:24:31.0000] <tantek>
so basically, if you're going to post something on a wiki page (or a blog post even), make it easily quotable/citable. this applies to both the content of the page, and perhaps a/the UI for it, which should permit easy quoting/citing on at least a section if not paragraph level.

[12:24:38.0000] <tantek>
/me has been working on this for his own blog/site.

[12:24:45.0000] <tantek>
citation UIs that is

[12:25:30.0000] <Tabatkins>
Yeah, once I finish my own markdown parser, I'll have easy citation-links-on-hover for paragraphs, and citation-links-all-the-time on headers.

[12:25:50.0000] <tantek>
Tabatkins, what does the parser have to do with UI?

[12:25:59.0000] <tantek>
do you have sketches of your UI you could share?

[12:26:22.0000] <tantek>
(I've gone through numerous iterations, it seems like a pretty challenging problem)

[12:26:31.0000] <tantek>
the UI part, not the parser part

[12:26:32.0000] <Tabatkins>
tantek: I can't easily add markup to my posts unless I have control of th eparser.  Right now I just author in pure Markdown, and output the generated results.

[12:26:53.0000] <tantek>
are you talking about manually adding citation markup/UI on each post?

[12:26:54.0000] <Tabatkins>
I haven't given much thought to UI yet, since I don't have the necessary technical plumbing in place yet.

[12:27:05.0000] <tantek>
that sounds backwards

[12:27:10.0000] <Tabatkins>
Yes, for some definition of "manual".

[12:27:21.0000] <tantek>
the technical plumbing should be driven by what the UI needs

[12:27:38.0000] <Tabatkins>
The technical plumbing I'm talking about is "actually be possible to have citation links at all".

[12:28:04.0000] <tantek>
I actually think the lack of good citation UIs is what's driven the "share all the things" button explosion.

[12:28:10.0000] <Tabatkins>
Until I get that far, no reason to think too much about the UI.  I've seen enough permalinks on blogs to know that reasonable UI is possible.

[12:28:49.0000] <tantek>
one of the reasons I don't think much of the current "intent" work is worth much (if anything) - it's solving higher level problems with fundamentally flawed approaches.

[12:29:14.0000] <tantek>
Tabatkins - hopefully by the time you get to that point, I'll have updated my blog and you can just steal my UI :)

[12:29:26.0000] <tantek>
I might even blog about it

[12:31:49.0000] <Hixie>
volkmar: yt?

[12:32:56.0000] <Hixie>
volkmar: what are the use cases for the default-all-uppercase and auto-titlecase-all-words input modes?

[12:33:50.0000] <volkmar>
all uppercase: some forms want that

[12:34:02.0000] <volkmar>
title case has no strong use cases though

[12:34:18.0000] <volkmar>
i don't think it will be bad to have it still

[12:34:36.0000] <tantek>
isn't it bad to have anything without strong use-cases?

[12:37:36.0000] <Hixie>
can you elaborate on the uppercase thing?

[12:37:38.0000] <Hixie>
what forms?

[12:37:43.0000] <Hixie>
do they just automatically uppercase stuff?

[12:37:49.0000] <Hixie>
or do they actually need to default the user to uppercase?

[12:38:03.0000] <Hixie>
seems like just having the form uppercase it automatically would be best

[12:41:06.0000] <Hixie>
my current proposal is this: http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control#Ian_Hickson

[12:41:29.0000] <Hixie>
it only supports english (and languages with the same needs as english, so most of europe) and japanese, because nobody else has given me any info

[14:02:43.0000] <Hixie>
should i use 'kana' or 'hiragana' for the keyword that corresponds to the default inputmode for japanese input, given that 'katakana' will fall back to this mode and there is no separate mode for hiragana?

[14:27:05.0000] <hober>
Hixie: kana

[14:27:58.0000] <Hixie>
hober: yeah that was my conclusion too

[14:32:16.0000] <hober>
i knew minoring in japanese would come in handy one day

[14:32:53.0000] <Hixie>
hehe

[14:33:11.0000] <Hixie>
give me 3 minutes and you can proof-read my inputmode text

[14:33:16.0000] <Hixie>
since i have absolutely no idea what i'm doing here :-)

[14:33:58.0000] <hober>
heh

[14:34:24.0000] <hober>
have you thought about how well the feature maps onto, say, how the keyboard works on the ipad?

[14:34:25.0000] <hober>
:)

[14:34:52.0000] <Hixie>
i have definitely thought about it

[14:34:59.0000] <Hixie>
whether that produced any useful results remains to be seen!

[14:35:07.0000] <hober>
fair enough :)

[14:35:18.0000] <Hixie>
http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control is the research

[14:35:26.0000] <Hixie>
which includes screenshots from iOS, MacOS, and Android

[14:35:36.0000] <Hixie>
regenning spec now...

[14:35:58.0000] <Hixie>
fixing markup errors now...

[14:39:37.0000] <Hixie>
hober: http://www.whatwg.org/specs/web-apps/current-work/#attr-inputmode

[14:45:36.0000] <Hixie>
hober: any opinions?

[14:46:17.0000] <hober>
looking

[14:48:10.0000] <hober>
i'd rather this not be used when the feature could be implicitly handled because of the type="" value

[14:48:18.0000] <hober>
<input type=numeric inputmode=email>

[14:48:28.0000] <hober>
err, type=number

[14:48:57.0000] <hober>
the UA knows it's a number; why present the less useful keyboard?

[14:49:28.0000] <Hixie>
you'll notice it only applies to type=search and type=text

[14:49:30.0000] <Hixie>
:-)

[14:49:41.0000] <zewt>
Hixie: how can you say "turn off autocapitalization, but enable spelling autocorrection" (eg. irc mode)?

[14:49:55.0000] <zewt>
(a lack on Android that's always annoyed me)

[14:50:11.0000] <Hixie>
zewt: that's "latin" (and it's present on android, it's e.g. what search uses)

[14:50:28.0000] <hober>
oh nice. what about <textarea>?

[14:50:53.0000] <Hixie>
applies to that too, yup

[14:51:06.0000] <hober>
ok

[14:51:16.0000] <hober>
cursory/initial take: looks good

[14:51:19.0000] <zewt>
meh i need to revert from android 4.1, it's a little faster but the ui is a bunch of little steps back (adding to a big step back)

[14:51:39.0000] <Hixie>
hober: awesome, thanks

[14:51:40.0000] <zewt>
like I hit the search box on the home page and instead of letting me type right away, there's a big delay while it loads a search app

[14:51:47.0000] <Hixie>
volkmar: inputmode="" is in the spec

[14:51:55.0000] <Hixie>
zewt: odd

[14:52:00.0000] <zewt>
wtf? i hit the google button and it's making me go through a WELCOME TO GOOGLE bit of nonsense

[14:52:06.0000] <Hixie>
what hardware?

[14:52:11.0000] <zewt>
nexus s

[14:52:25.0000] <zewt>
the google search box is now a narrow thing at the top instead of the input box

[14:52:29.0000] <Hixie>
ah, yeah, that might just be too old to handle jellybean fast

[14:53:28.0000] <zewt>
that input box, at least, doesn't use autocorrection at all (probably depends on google's suggestions to handle it)

[14:53:41.0000] <zewt>
i'll try in the browser

[14:54:08.0000] <Hixie>
hm, yeah, i guess it doesn't

[14:54:11.0000] <Hixie>
never noticed that

[14:54:23.0000] <zewt>
dunno, amazon search seems verbatim too

[14:54:40.0000] <zewt>
i think the input API only has "helpful stuff on" and "helpful stuff off", at least last I poked through it

[14:54:45.0000] <Hixie>
you must be right, android just has equivalents of verbatim and latin-prose

[14:54:50.0000] <Hixie>
interesting

[14:55:15.0000] <Hixie>
well, the spec has an in between mode, anyway, so hopefully you are happy :-)

[14:55:29.0000] <zewt>
yeah, one more little nudge for android to support it some day :)

[14:57:28.0000] <zewt>
are there use cases for actually hinting half-width katakana?

[14:58:27.0000] <zewt>
(possibly a question for MikeSmith et al)

[14:59:32.0000] <Hixie>
zewt: it's apparently common for OSes to have half-width IMEs, dunno if there's a need to hint for it though

[14:59:39.0000] <zewt>
yeah that's what i mean

[14:59:49.0000] <Hixie>
ok, bbiab.

[15:00:00.0000] <zewt>
same question for full-width-latin (not sure how it would be used by authors)

[15:01:05.0000] <gsnedders>
This strikes me as a mess, and rather exemplifies why I want a decent solution for caesura: <div class="line"><a name="li2"></a>&thorn;&#x113;od-cyninga&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&thorn;rym gefr&#x16B;non,</div>

[15:01:29.0000] <gsnedders>
That is: þēod-cyninga      þrym gefrūnon,

[15:02:19.0000] <gsnedders>
I should just stop caring about markup. It would simplify many things.

[15:48:16.0000] <annevk>
what is the case for inputmode=url?

[15:53:05.0000] <zewt>
if you're entering a URL on a webpage then you want to be in a keyboard mode that puts eg. : and / on the front keyboard (as for how often people enter URLs except into the browser directly, dunno)

[15:54:06.0000] <annevk>
why would you not use type=url?

[15:54:47.0000] <zewt>
dunno

[15:55:49.0000] <annevk>
because that's really the question

[15:56:23.0000] <zewt>
the legacy conflation of type for both "data type" and "data contents" makes it all a little mixed up

[16:29:43.0000] <zewt>
gah

[16:30:00.0000] <zewt>
apparently there are some errors that are only shown on the old "error console" in FF and not on the web console

[16:30:14.0000] <zewt>
wish they'd merge those and stop having two consoles

[16:41:39.0000] <Hixie>
strikes me that we should probably look into how to make inputmode work for contenteditable

[16:41:42.0000] <Hixie>
bummer

[16:51:59.0000] <smaug____>
zewt: error console is very useful

[16:52:15.0000] <smaug____>
zewt: but sure, web page related errors should go to web console

[16:52:28.0000] <smaug____>
zewt: file a bug if something goes to the wrong console

[16:56:18.0000] <tantek>
so how about that Opus huh? http://hacks.mozilla.org/2012/07/firefox-beta-15-supports-the-new-opus-audio-format/


2012-07-21
[20:04:59.0000] <Hixie>
wow, the htmlwg is still arguing over r6890 while we're up at r7200. i sure hope they find an editor soon or it's going to be a huge amount of work just coordinating upstreaming of patches.

[00:32:03.0000] <annevk>
Hixie: will you consistently use "Fix some typos or copypasta." from now on?

[00:32:16.0000] <annevk>
Hixie: could filter those from the tweets too maybe, although nobody has complained much

[00:42:37.0000] <annevk>
tantek: is there a catch somewhere? sounds too good to be true

[01:10:45.0000] <tantek>
annevk - which part?

[01:11:06.0000] <annevk>
tantek: opus

[01:17:27.0000] <tantek>
well we're not out of the woods yet. we still need to see Webkit/Chrome/Safari updates.

[01:40:28.0000] <annevk>
do other audio codecs have faster decoders/encoders by any chance?

[01:40:34.0000] <annevk>
article didn't mention that

[01:40:37.0000] <annevk>
or hardware support

[01:41:20.0000] <tantek>
well you can try the Firefox 15 beta on your Android and see how well it does: firefox.com/beta

[01:44:00.0000] <annevk>
heh

[04:12:18.0000] <ruby_on_tails>
data in localStorage is never deleted ? what happens if 500 sites use 5mb of it on my machine ?

[04:12:27.0000] <ruby_on_tails>
how does an average user clear it ?

[04:21:04.0000] <smaug____>
that is up to the UA

[04:24:54.0000] <Philip`>
An average user will have way more than 2.5GB of free disk space, so it won't be a problem

[04:30:59.0000] <ruby_on_tails>
but that doesnt sound good

[04:36:37.0000] <smaug____>
[OT] I hope we could get rid of localStorage. It is a horrible API.

[06:18:46.0000] <Yaffle>
Hello!!!

[06:19:32.0000] <Yaffle>
@annevk, XMLHttpRequest has 50ms delay for progress events, WHY???

[06:34:16.0000] <annevk>
it's not a delay

[06:37:44.0000] <Yaffle>
delay or not, it is

[06:37:46.0000] <Yaffle>
why???

[06:40:00.0000] <Yaffle>
what do "50ms" in the XMLHttpRequest2 spec??? it is a bug

[06:40:11.0000] <annevk>
it's just an interval

[06:40:19.0000] <Yaffle>
for what?

[06:40:29.0000] <annevk>
so that they don't fire too often

[06:40:44.0000] <Yaffle>
it is also a delay,

[06:40:47.0000] <Yaffle>
because

[06:40:59.0000] <Yaffle>
"50ms or for every byte transmitted, whichever is least frequent."

[06:41:25.0000] <Yaffle>
so, progress will not fire for every byte(s), but there are will be a delay

[06:41:50.0000] <annevk>
not sure what you're trying to say

[06:41:51.0000] <Yaffle>
why this delay is needed?

[06:42:08.0000] <Yaffle>
i need latency = 1ms, not 50 ms

[06:42:52.0000] <annevk>
you'll have to explain a bit more and preferably in form of an emali

[06:42:54.0000] <annevk>
email*

[06:43:04.0000] <Yaffle>
Opera, Chrome, Firefox updates "responseText" every 10-12 ms

[06:43:22.0000] <Yaffle>
Chrome, Firefox fires "readystatechange" for every responseText update

[06:43:35.0000] <Yaffle>
so, it is fired more often, then "progress"

[06:43:58.0000] <annevk>
well that's a bug in Chrome/Firefox

[06:44:35.0000] <annevk>
but we can fire progress more often I guess if people want that, you haven't really elaborated on why it's needed, please include that in the email

[06:44:45.0000] <Yaffle>
don't think so

[06:45:46.0000] <Yaffle>
i want "progress" events for every byte(s) received; for 50 ms interval, i can use setInterval(callback, 50)

[06:46:53.0000] <annevk>
well, want is not a use case

[06:47:00.0000] <annevk>
I'm going back to my weekend now, this is getting boring

[06:48:21.0000] <Yaffle>
i can fix spec myself

[06:48:28.0000] <Yaffle>
just give access

[06:48:49.0000] <Yaffle>
whatwg is LOL

[07:35:17.0000] <Ms2ger>
/me wonders if anybody has tried to implement a WebIDL array

[08:18:51.0000] <zewt>
annevk: try not to get trolled heh

[08:35:25.0000] <Hixie>
annevk: i thought i was still sending you "typo" and "xref"... my script is just expanding those for e-mail purposes, i think... did i screw up?

[08:37:30.0000] <zewt>
uh, why is a Chair hopping on webapps and saying "take this discussion elsewhere"

[08:37:44.0000] <zewt>
since when does that ever happen--don't recall ever seeing it before since I subscribed

[08:39:01.0000] <zewt>
ah he's probably the webappsec chair

[08:39:05.0000] <zewt>
and didn't notice that it's cc: webapps

[08:54:50.0000] <[tm]>
zewt: good example of why cross posting is bad

[08:59:30.0000] <zewt>
eh

[08:59:56.0000] <zewt>
sometimes multiple audiences are relevant to a discussion and expecting everyone on one list to subscribe to the other for one thread is unreasonable

[12:36:50.0000] <annevk>
Hixie: ah, I didn't realize those were different

[12:37:11.0000] <annevk>
zewt: have to give people a chance, but that was going nowhere

[12:38:29.0000] <zewt>
yeah, just the "stop talking" thing never really happens on webapps

[16:13:52.0000] <norbert_>
hey all

[16:13:59.0000] <norbert_>
I just read this: http://developers.slashdot.org/story/12/07/21/2040257/html5-splits-into-two-standards

[16:14:08.0000] <norbert_>
and I'm curious, is this something that recently happened?

[16:14:59.0000] <norbert_>
or is /. reporting old stuff, and this is basically what you wrote about in Jan. 2011 http://blog.whatwg.org/html-is-the-new-html5

[16:19:49.0000] <Philip`>
norbert_: http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jul/0119.html is probably the most relevant discussion of the new news

[16:20:03.0000] <norbert_>
thanks

[16:20:03.0000] <Philip`>
(I haven't actually read that, but I presume it's relevant...)

[16:21:25.0000] <Philip`>
Previously the W3C version of the spec was mostly a subset of the WHATWG version, generated from the same source and edited by the same editor, but now they're effectively two independent forks (the WHATWG one with the original editor, the W3C one looking for a new editor), as far as I'm aware

[16:44:34.0000] <norbert_>
maybe W3C and WHATWG can at least agree on some kind of canvas standard?


2012-07-22
[17:52:24.0000] <Hixie>
norbert_: the WHATWG spec will track whatever implementors do

[03:09:50.0000] <annevk>
some articles on the fork

[03:09:53.0000] <annevk>
http://www.theverge.com/2012/7/22/3175248/html5-fork-w3c-whatwg

[03:09:58.0000] <annevk>
http://www.i-programmer.info/news/191-htmlcss/4531-which-html5-whatwg-and-w3c-split.html

[03:10:02.0000] <annevk>
http://techcrunch.com/2012/07/22/html5-work-splits-into-living-and-snapshot-standards-developers-need-not-standard/

[03:10:25.0000] <annevk>
nothing much interesting though

[07:49:25.0000] <BenoitRen>
Hi all.

[07:50:44.0000] <BenoitRen>
There's an old website I maintain (webmistress vanished from the interwebs more than a decade ago), and I've been looking through it again recently to improve the HTML further. One thing I'm having trouble with is images that get referenced explicitly in the text.

[07:51:23.0000] <BenoitRen>
Like, you have <p><img src="..."></p><p>This picture shows blahblahblah.</p>

[07:51:47.0000] <BenoitRen>
As this is supposed to be a standards compliant mirror of the original site, I don't have much leeway in changing the text. :/

[07:53:32.0000] <BenoitRen>
Example: http://users.skynet.be/fa258499/beyondalgo/story/you_guide.html

[08:58:21.0000] <AryehGregor>
BenoitRen, what's wrong with images that get referenced explicitly in the text?

[09:07:47.0000] <BenoitRen>
AryehGregor: The image might not be showing.

[09:08:21.0000] <AryehGregor>
BenoitRen, . . . and?  You can't really do much about that.  Provide alt text if that helps.

[09:08:42.0000] <AryehGregor>
But I mean, if the point of something is to talk about a picture, and the picture doesn't display, there's a limited amount you can do about it.

[09:08:49.0000] <BenoitRen>
AryehGregor: But what would I put in the alt text that logically precedes text like "This image shows...".

[09:08:56.0000] <BenoitRen>
?

[09:09:02.0000] <AryehGregor>
I dunno.  "Image of ..."

[09:09:15.0000] <BenoitRen>
That's usually bad alt text, though. :/

[09:09:23.0000] <BenoitRen>
alt text should be a replacement for the image.

[09:11:43.0000] <AryehGregor>
If possible.

[09:11:49.0000] <AryehGregor>
Or, rather, to the extent possible.

[09:12:16.0000] <AryehGregor>
A good way to think about it is: "If I were reading this page to someone over the phone, how would I describe it?"

[09:12:46.0000] <AryehGregor>
The answer here would probably be something like "Then there's a picture of X, and after that it says . . ."

[09:12:59.0000] <BenoitRen>
Yeah.

[09:13:27.0000] <AryehGregor>
Sometimes you can't do any better than that.

[09:13:46.0000] <AryehGregor>
In other cases, you might read what the picture says (if it's really text in picture format), or just skip it (if it's decorative), and that's fine too.

[09:14:03.0000] <AryehGregor>
Although I think I read a survey of blind users who say they'd prefer to have decorative images mentioned.

[09:14:41.0000] <AryehGregor>
I guess because that just gives them a better idea of what the page contains.

[09:16:41.0000] <boogyman>
AryehGregor: can you give me an example of a semantically appropriate decorative image

[09:16:50.0000] <AryehGregor>
I don't know what that means.

[09:17:08.0000] <AryehGregor>
So no.  :)

[09:18:07.0000] <boogyman>
An example of a decorative image that has meaning in the context of its placement, instead of just there for visual "decorative purposes".

[09:18:42.0000] <boogyman>
CSS and image replacement is around of a reason :| for the latter

[09:25:23.0000] <BenoitRen>
But a decorative image by definition isn't semantic...

[15:03:33.0000] <BenoitRen>
I'm going now. Bye! And thanks. :)

[15:25:36.0000] <david_carlisle>
MikeSmith: If someone opened a bug and it got cloned to WhatWG product, shouldn't they still be the reporter (or at least on the CC list) so that if (to take a random example) Hixie closes it as wontfix, the original reporter gets notified?

[15:32:04.0000] <annevk>
the clone script cannot set the reporter

[15:32:11.0000] <annevk>
does it not cc the original reporter?

[15:32:26.0000] <annevk>
also, you want to ask Hixie, he wrote the scripts

[15:38:12.0000] <david_carlisle>
annevk: apparently not:-)

[15:40:53.0000] <david_carlisle>
OK I thought I saw somewhere  they were Mike's scripts, I may push back on the wontfix (of 17798) in which case I'd be contacting Hixie anyway (or I may not bother, too late to decide tonight)


2012-07-23
[19:45:09.0000] <Hixie>
ah, crap, did i forget to cc the reporter?

[19:45:10.0000] <Hixie>
bummer

[20:23:37.0000] <nonge>
problem solved - that was fast :D thanks

[20:39:37.0000] <Hixie>
ok, i'm running a script that ccs the reporters on all the bugs i cloned to the whatwg product for those bugs where the original reporter isn't cc'ed

[23:45:20.0000] <MikeSmith>
Hixie: oh thanks for running

[23:45:27.0000] <MikeSmith>
thanks for running that

[23:45:50.0000] <MikeSmith>
I had gone through and manually added a few earlier today

[23:46:36.0000] <MikeSmith>
Hixie: hmm is that script done running?

[23:47:01.0000] <MikeSmith>
because if it had run on all of those bugs I think I would be seeing a lot more bugmail in my inbox right now

[23:47:47.0000] <MikeSmith>
because I'm also already Cc'ed on every one of them I think

[00:08:14.0000] <Hixie>
MikeSmith: most bugs weren't affected

[00:08:37.0000] <Hixie>
MikeSmith: (most bugs are filed with my script)

[00:11:02.0000] <MikeSmith>
Hixie: hmm OK

[00:11:17.0000] <MikeSmith>
but I still think I should have gotten more than I did

[00:11:23.0000] <MikeSmith>
/me checks a few bugs

[00:12:25.0000] <Hixie>
if you find any i missed let me know

[01:26:09.0000] <annevk>
matjas: it should be "Platform Design Principles" prolly

[01:26:23.0000] <annevk>
matjas: or just "Design Principles"

[01:26:38.0000] <matjas>
+∞

[01:27:37.0000] <annevk>
matjas: I've been thinking about putting it on the WHATWG Wiki instead and shortening some of the topics and making others more generic, but I'm not sure it's worth the effort

[01:28:22.0000] <annevk>
(publishing an update through the W3C HTML WG is impossible at this point)

[01:29:52.0000] <matjas>
more divergence, yay!

[01:32:38.0000] <annevk>
heh

[01:34:42.0000] <annevk>
we've yet to see what that actually means

[01:39:29.0000] <hsivonen>
sigh. still someone in the comments in The Verge longing for XHTML2

[01:42:17.0000] <annevk>
http://xkcd.com/386/

[01:55:56.0000] <david_carlisle>
Hixie: thanks for addressing the bugzilla  CC issue (not so much thanks for "won't fixing" my bug, but I'll think about that:-)

[02:18:23.0000] <hsivonen>
the HTML WG could use chairing like this: http://lists.w3.org/Archives/Public/public-webapps/2012JulSep/0212.html

[02:21:42.0000] <tantek>
annevk - it's likely worth the effort (putting it on the WHATWG Wiki instead and shortening some of the topics and making others more generic)

[02:21:55.0000] <tantek>
happy to contribute to it as well

[02:27:32.0000] <Ms2ger>
"Thank you for helping the HTML5 effort!"

[02:27:47.0000] <Ms2ger>
Hey, I wanted to help the HTML effort instead

[02:28:00.0000] <odinho>
:(

[02:28:09.0000] <odinho>
Ms2ger: Dang! You backed the wrong horse.

[02:28:17.0000] <Ms2ger>
/me shoots the horse

[02:28:20.0000] <annevk>
Ms2ger: where does it say that?

[02:28:57.0000] <tantek>
Ms2ger - what's the difference to you?

[02:29:26.0000] <Ms2ger>
annevk, http://www.whatwg.org/specs/web-apps/current-work/reviewer.js

[02:29:46.0000] <Ms2ger>
tantek, HTML is a useful specification, HTML5 is the W3C's waste of time

[02:30:12.0000] <tantek>
Ms2ger - what's wrong with a feature frozen stabilization fork?

[02:30:19.0000] <Ms2ger>
The HTMLWG

[02:30:31.0000] <annevk>
tantek: it's not maintained?

[02:30:59.0000] <tantek>
annevk - yes, that's certainly an issue until they empower new editors to take bug fixes from HTML.

[02:31:55.0000] <annevk>
you can only maintain a fork so long, because at some point features will come to depend on new features, or will be significantly rewritten, etc.

[02:32:04.0000] <tantek>
Ms2ger that sounds roughly like ad hominem - HTMLWG (or the people of)  being bad or not does not equate to HTML5 (or the details/ideas contained therein) being bad or not.

[02:32:08.0000] <annevk>
and maintaining a fork is a waste of resources I think

[02:32:27.0000] <tantek>
annevk - I doubt features will come to depend on new features, as that would likely break backward compat.

[02:32:31.0000] <annevk>
publishing a snapshot for lawyers, okay, but doing anything more is a waste of time

[02:32:38.0000] <tantek>
I expect most new features in HTML to be reasonably modularizable.

[02:32:43.0000] <annevk>
tantek: they do all the time actually

[02:32:52.0000] <tantek>
In many ways, HTML5 could be the CSS 2.1 of HTML.

[02:32:52.0000] <Ms2ger>
tantek, you miss the point

[02:33:02.0000] <annevk>
tantek: e.g. when CORS was added existing features were defined in a different way

[02:33:12.0000] <Ms2ger>
tantek, the HTMLWG is the reason the HTML5 spec is bad

[02:33:37.0000] <tantek>
Ms2ger - I'm willing to give the yet-to-be-named HTML5 editors a chance first.

[02:33:56.0000] <tantek>
annevk - "defined in a different way" != break backward compat

[02:34:05.0000] <annevk>
exactly

[02:34:07.0000] <Ms2ger>
tantek, I'm not talking about hypothetical editors, I'm talking about trolls and the chairs

[02:34:12.0000] <annevk>
but you cannot easily take in fixes at that point

[02:34:15.0000] <tantek>
in practice I expect any new post-HTML5 features to be reasonably modularizable.

[02:34:26.0000] <tantek>
it's not clear that there's a need for further bloating a monolithic spec

[02:34:38.0000] <annevk>
or when we merge CORS and fetching

[02:34:46.0000] <annevk>
you cannot easily take such fixes in

[02:34:47.0000] <tantek>
I expect to see a bunch of module specs that extend HTML5.

[02:34:50.0000] <annevk>
going to be a mess

[02:35:20.0000] <Ms2ger>
tantek, who is going to write those specs?

[02:35:31.0000] <tantek>
Ms2ger whoever wants the new features

[02:35:47.0000] <tantek>
annevk - perhaps merging CORS and rewriting fetching would be a good HTML5.1 update.

[02:36:00.0000] <Ms2ger>
I suspect it's more likely that Hixie is going to write them

[02:36:17.0000] <Ms2ger>
There aren't many competent editors around here

[02:36:29.0000] <Ms2ger>
Anyway, I could be doing something useful

[02:37:09.0000] <annevk>
tantek: it's that kind of talk why I left the CSS WG; planning about drafts instead of working on tangible things

[02:37:48.0000] <tantek>
annevk - it's easy enough to ignore such threads/conversations if they don't interest you.

[02:38:00.0000] <annevk>
well yes, I quit the HTML WG

[02:38:14.0000] <annevk>
I have no hope they're going to be doing anything useful

[02:38:17.0000] <tantek>
oh? did I miss that? since the May f2f?

[02:38:22.0000] <annevk>
the last meeting made that pretty clear

[02:38:32.0000] <annevk>
since a month ago or so

[02:38:36.0000] <tantek>
I remember you threatening as much - or at least seeming to give up

[02:38:43.0000] <tantek>
I didn't realize you'd formally quit HTMLWG

[02:38:58.0000] <annevk>
wrote a little bit about it here: http://annevankesteren.nl/2012/07/passion

[02:39:05.0000] <tantek>
/me reads

[02:39:07.0000] <annevk>
I got chaals to remove me from the group

[02:39:58.0000] <tantek>
not mikesmith?

[02:40:03.0000] <tantek>
or did you quit webapps too?

[02:40:12.0000] <annevk>
your AC representative needs to remove you

[02:40:28.0000] <annevk>
which was chaals at the time

[02:40:57.0000] <tantek>
re: "Until that point HTML was by and large developed by a committee in private meetings." - the sad thing is, originally HTML was developed on an open mailing list - I'd like to know when it got moved to private lists/meetings.

[02:41:16.0000] <annevk>
I have not quit WebApps, WebApps is reasonably well managed, though the meetings make me cringe every now and then

[02:41:20.0000] <tantek>
www-talk if I recall (where HTML was developed originally)

[02:42:32.0000] <Ms2ger>
Heh, meetings

[02:42:39.0000] <tantek>
yeah, the May f2f meetings were oddly mixed.

[02:43:35.0000] <tantek>
Ms2ger - do you ever go to W3C f2f meetings?

[02:43:53.0000] <Ms2ger>
No

[02:44:28.0000] <Ms2ger>
https://lists.w3.org/Archives/Member/w3c-html-wg/ maybe?

[02:44:30.0000] <annevk>
I go to the meetings because they're fun outside of the actual meetings, the meetings themselves are mostly useless

[02:44:49.0000] <tantek>
annevk - re: "I quit the HTML WG politics game as what happened the other day was inevitable. " - what specifically happened the other day?

[02:46:01.0000] <tantek>
Ms2ger - from that I found this: https://lists.w3.org/Archives/Member/html-erb/

[02:46:35.0000] <annevk>
tantek: editor split

[02:46:58.0000] <annevk>
I guess I can add one link

[02:46:59.0000] <tantek>
first message to html-erb was on 1995-12-07 https://lists.w3.org/Archives/Member/html-erb/1995OctDec/0000.html

[02:47:01.0000] <annevk>
to http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jul/0119.html

[02:50:54.0000] <tantek>
annevk - the response from Steve Faulkner is classic

[02:51:26.0000] <Ms2ger>
He's been on a troll lately

[02:51:27.0000] <Ms2ger>
*roll

[02:51:28.0000] <tantek>
who will be the keeper of canonical author level conformance requirements and advice?

[02:51:41.0000] <tantek>
Ms2ger - Freudian LOL

[02:51:54.0000] <Ms2ger>
tantek, we :)

[03:54:59.0000] <MikeSmith>
heh so the Register has chimed in

[03:55:06.0000] <MikeSmith>
"Judean People's Front and People's Front of Judea go their own way"

[03:55:11.0000] <MikeSmith>
http://www.theregister.co.uk/2012/07/23/html_5_fork/

[03:56:28.0000] <MikeSmith>
hmm sadly that title is the only funny part

[03:57:21.0000] <MikeSmith>
and that title was probably written by an editor and not by the author of the article

[03:57:29.0000] <odinho>
Hehe.

[04:21:57.0000] <tantek>
I can't imagine that many people would care on the level the Register is trying to explain - it must look like so much inside-baseball to the typical reader, even web developer.

[04:22:42.0000] <MikeSmith>
tantek: indeed

[04:23:02.0000] <MikeSmith>
and people don't read the Register to get information anyway

[04:23:05.0000] <tantek>
MikeSmith - I suspect we'll manage to make things work somehow regardless.

[04:23:41.0000] <MikeSmith>
tantek: yeah I'm sure we will

[04:24:34.0000] <MikeSmith>
and no matter what anybody thinks of the HTML WG for better or worse it's going to remain relevant

[04:25:05.0000] <Ms2ger>
*irrelevant

[04:25:17.0000] <MikeSmith>
cute

[04:25:37.0000] <MikeSmith>
but specifically in the case of document-conformance the W3C HTML5 spec going to be authoritative

[04:25:43.0000] <tantek>
MikeSmith - sure, I just hope those of us contributing to it can continue to do so productively.

[04:25:57.0000] <tantek>
(re: HTMLWG)

[04:26:08.0000] <Stevef_>
Ms2ger: using the troll word is a lazy way of trying to discredit a person

[04:26:41.0000] <Ms2ger>
Like your tweets, then :)

[04:27:03.0000] <MikeSmith>
tantek: yeah I do do but there are some people who have proven they often can't participate productively and who should have been permanently banned from public-html discussions long ago

[04:27:29.0000] <tantek>
/me gets a drink, sits back, props up his feet to watch the Msger / Stevef_ show.

[04:27:35.0000] <tantek>
*Ms2ger

[04:28:09.0000] <Ms2ger>
Have fun on your own, I've got work to do

[04:28:52.0000] <tantek>
*except when snarkily correcting MikeSmith ;)

[04:30:39.0000] <hsivonen>
it remains to be seen if Web devs end up caring about authoritative HTML5 conformance, authoritative HTML LS conformance, both or neither

[04:31:10.0000] <Stevef_>
I have no interest in a protracted discussion about it, just gotta call out BS when i see it

[04:31:34.0000] <hsivonen>
at some point, I will need to figure out a way to transition the functionality of html5.validator.nu to validator.nu and validator.nu to custom.validator.nu or complex.validator.nu

[04:31:47.0000] <hsivonen>
dunno how to do that without breaking the Web Service end points

[04:32:18.0000] <MikeSmith>
hsivonen: btw I'm wondering if maybe it's time to remove the HTML4 and XHTML1 presets from the validator UI

[04:32:31.0000] <MikeSmith>
I think very few people are actually using those

[04:32:44.0000] <hsivonen>
MikeSmith: only on Validator.nu or also on the W3C validator?

[04:32:52.0000] <MikeSmith>
both

[04:33:42.0000] <hsivonen>
MikeSmith: what about the legacy DTD stuff on the W3C validator?

[04:33:46.0000] <MikeSmith>
I think Jukka is the only one who has ever commented on those, and the only reason he has was to complain that the HTML4 checker doesn't follow SGML rules

[04:34:36.0000] <MikeSmith>
hsivonen: what about it? you mean the HTML4 facet of the W3C validator, or the 3.2 and older?

[04:34:40.0000] <hsivonen>
it does seem that it happens that people who think validator.nu (as opposed to html5.validator.nu) only does HTML5 validation accidentally end up doing HTML4 validation with the defaults

[04:35:13.0000] <MikeSmith>
ah

[04:35:13.0000] <hsivonen>
MikeSmith: do you mean keeping DTD-based anything around in the W3C Validator?

[04:36:24.0000] <MikeSmith>
hsivonen: no I don't want to keep anything DTD based around anywhere except in the legacy W3C validator

[04:36:45.0000] <hsivonen>
MikeSmith: but the legacy validator is staying around at the W3C?

[04:36:52.0000] <MikeSmith>
yeah

[04:36:52.0000] <MikeSmith>
b

[04:37:20.0000] <MikeSmith>
but eventually the Nu Markup Validator is going to be at http://validator.w3.org/

[04:37:35.0000] <hsivonen>
I see

[04:37:38.0000] <MikeSmith>
and the legacy validator will be moved to http://validator.w3.org/html4 or somewhere

[04:38:25.0000] <hsivonen>
but yeah, I'm not particularly enthusiastic about keeping the HTML4 schema around

[04:38:48.0000] <MikeSmith>
hsivonen: so I think that along with removing the HTML4 and XHTML1 presets, if we did that, we should change the doctype checking so that it doesn't ever try to use those

[04:39:01.0000] <MikeSmith>
basically, make the behavior like html5.validator.nu

[04:39:20.0000] <hsivonen>
MikeSmith: yeah

[04:39:53.0000] <MikeSmith>
and maybe include a message to say that if you want to do HTML4 or XHTML1 message, use the existing W3C DTD-based validator

[04:39:55.0000] <hsivonen>
on the Web Service side, there's the issue that validator.nu now accepts custom schemas but html5.validator.nu doesn't

[04:40:08.0000] <MikeSmith>
*HTML4 or XHTML1 checking

[04:40:20.0000] <MikeSmith>
hsivonen: I see

[04:40:27.0000] <hsivonen>
MikeSmith: I don't really want to encourage anyone to use DTD-based validation

[04:40:36.0000] <MikeSmith>
OK

[04:41:10.0000] <MikeSmith>
I don't want to encourage people to use XHTML1 or HTML4 doctypes

[04:41:14.0000] <annevk>
hsivonen: redirect html5.validator.nu to validator.nu, only expose the complex stuff as a service, and be done with it?

[04:41:58.0000] <hsivonen>
annevk: the thing is that the complex stuff has worse perf characteristics

[04:42:25.0000] <hsivonen>
annevk: so it's not nice to put the custom schema-enabled Web Service and the common-case GUI on the same host

[04:42:54.0000] <annevk>
does it have noticeable impact?

[04:43:13.0000] <annevk>
how many API consumers are there?

[04:43:31.0000] <hsivonen>
though maybe that could be addressed by 1) discontinuing Schematron support or 2) moving the whole thing to Heroku and hoping that superior infrastructure takes care of the user-visible issues

[04:43:48.0000] <hsivonen>
annevk: occasionally, validator.nu gets DoS'ed

[04:43:57.0000] <hsivonen>
annevk: non-malicuously AFAICT

[04:44:04.0000] <hsivonen>
annevk: I don't have proper metrics

[04:44:07.0000] <annevk>
you could ask the consumers to move to a different API during some time period

[04:44:10.0000] <hsivonen>
(I should probably get some)

[04:44:20.0000] <annevk>
and then disable it

[04:44:33.0000] <hsivonen>
annevk: yeah, except I have no communication channel with them

[04:44:59.0000] <annevk>
hsivonen: sounds like you should set one up :)

[04:45:56.0000] <hsivonen>
moving the validator to Heroku would be technically sensible

[04:46:05.0000] <hsivonen>
but it scares me to host anything in the United States

[04:46:20.0000] <hsivonen>
hosting it in France is legally reckless enough

[04:47:35.0000] <annevk>
are you scared the host will turn over data of validator users?

[04:48:42.0000] <hsivonen>
annevk: no, I'm scared about DMCA compliance and crazy liabilities I don't even know about

[04:49:30.0000] <hsivonen>
annevk: as for the EUCD, Finland and France are in the same boat anyway

[04:49:49.0000] <MikeSmith>
maybe switzerland

[04:50:13.0000] <hsivonen>
the most scary thing about France is their anti-Nazi laws

[04:51:07.0000] <hsivonen>
I don't intend to be a Nazi, but I don't know how crazy French enforcement is when it comes to echoing 3rd-party content (via Show Source or quotations in error messages)

[04:51:36.0000] <hsivonen>
I'm hoping it's clear enough that validator.nu doesn't really host any 3rd-party content it might echo

[04:54:12.0000] <hsivonen>
annevk: as for handing over use data, the good thing is that the validator isn't in the business of storing users' content

[04:55:33.0000] <hsivonen>
annevk: but it's still annoying that pretty much every competent hosting provider in Europe eventually wants to serve U.S. customers or have a data center in the U.S. to enable their European customers serve U.S. end users faster

[04:56:01.0000] <annevk>
makes sense though

[04:56:50.0000] <hsivonen>
so all your cloud data are belong to U.S. TLA agencies even if your provider is nominally European

[04:57:49.0000] <annevk>
I thought some of them offer options on only storing data in Europe

[04:58:05.0000] <annevk>
in particular for EU governments who do not want to store data in the US

[04:58:57.0000] <annevk>
this global internet, local laws thing sucks though

[04:59:05.0000] <annevk>
I think that's annoying

[04:59:33.0000] <hsivonen>
annevk: I doubt that helps much with staff to incarcerate in the U.S.

[05:00:15.0000] <hsivonen>
at least that's the general sentiment of the accidentally honest Microsoft rep

[05:01:40.0000] <annevk>
ooh

[05:08:17.0000] <hsivonen>
annevk: for [citation needed]: http://www.zdnet.com/blog/igeneration/microsoft-admits-patriot-act-can-access-eu-based-cloud-data/11225

[05:08:46.0000] <ruby_on_tails>
i am prepending a span in the new LIs created in a contenteditable OL using jquery, this causes issues in FF.  on hitting the return key in FF, a new LI is created but the caret does not move there, you have to hit return twice to get down to the new LI created, any idea how to solve this ?

[05:09:42.0000] <matjas>
lolol: http://www.theregister.co.uk/2012/07/23/html_5_fork/ — “Judean People's Front and People's Front of Judea go their own way”

[05:10:31.0000] <ruby_on_tails>
sorry got dc

[05:11:10.0000] <ruby_on_tails>
basically the caret doesnt move down due to the span

[05:16:12.0000] <AryehGregor>
hsivonen, FWIW, I believe Wikipedia stores all its data in the U.S. specifically because U.S. laws are most favorable to its interests.  The U.S. has very defendant-friendly defamation law, and strong free speech guarantees in general, relative to a typical European country.  German courts have ordered wikipedia.de taken down at least once, and there have been credible complaints of infringement of British copyright law, but in both cases Ame

[05:16:12.0000] <AryehGregor>
rican law was clearly on Wikipedia's side.

[05:16:53.0000] <AryehGregor>
America is a bad place to host U.S. classified documents, gambling sites, and maybe a couple other things (some types of pornography?), but overall it tends to be safer than most European countries.

[05:16:59.0000] <AryehGregor>
I think the Netherlands is good too.

[05:17:35.0000] <AryehGregor>
That's with respect to prosecution, though.  If you care about the government snooping on your data, the story might be different.

[05:18:53.0000] <AryehGregor>
Although I suspect most of Europe is relatively cooperative with U.S. investigations.

[05:19:43.0000] <hsivonen>
AryehGregor: indeed, hosting in the U.S. makes sense for Wikipedia

[05:20:45.0000] <hsivonen>
AryehGregor: and if you start with the assumption that U.S. TLA agencies can get your non-public data anyway, then hosting non-public data in the U.S. makes sense, because at least you are defending against snooping by *other* governments

[05:21:08.0000] <AryehGregor>
Right.

[05:21:47.0000] <AryehGregor>
For a typical site provider, the U.S. seems to be among the most legally friendly places to host your site, because it has very strong free speech guarantees in practice relative to most other countries.

[05:21:56.0000] <AryehGregor>
However much some people might complain.

[05:22:00.0000] <hsivonen>
AryehGregor: but that the U.S. makes sense for an org the size of Wikimedia Foundation doesn't mean it makes sense for someone smaller

[05:22:19.0000] <AryehGregor>
Well, it's true that legal protections on paper aren't very useful if you can't afford to be sued.

[05:22:20.0000] <hsivonen>
AryehGregor: because even if you have the law on your side, legal stuff in the U.S. is expensive

[05:22:27.0000] <AryehGregor>
Is it a lot cheaper other places?

[05:22:34.0000] <hsivonen>
AryehGregor: yes

[05:22:38.0000] <AryehGregor>
How so?

[05:23:17.0000] <hsivonen>
AryehGregor: well, insurance companies would know, and it's easier&cheaper to get legal insurance for any territory except the U.S. and Canada

[05:24:01.0000] <AryehGregor>
But realistically, you still strongly want to avoid a court case anyway, right?  You don't want to have to pay for insurance to start with, and insurance doesn't prevent countless hours of lost time.

[05:25:18.0000] <AryehGregor>
(I think part of what makes some countries much better for lawsuit insurance is the loser-pays rule some countries have for civil cases.)

[05:25:34.0000] <hsivonen>
AryehGregor: that's a big deal

[05:25:51.0000] <AryehGregor>
That doesn't really help a typical individual much, because they have to pay up-front in case they do lose, but if you're insured it probably makes a huge difference.

[05:25:58.0000] <hsivonen>
AryehGregor: another rule that makes civil cases saner is a ban on punitive damages

[05:26:17.0000] <hsivonen>
AryehGregor: i.e. a principle that the damages awarded to the plaintiff can never exceed the suffered harm

[05:26:37.0000] <AryehGregor>
Punitive damages aren't usually a huge deal in American tort cases, I don't think.  I think unreasonably high pay-outs for intangible harm like emotional damage are more of a problem, at least from anecdotal observation.

[05:26:46.0000] <hsivonen>
(in if you want to punish the defendant more, make it a criminal case and pay the fines to the state)

[05:28:22.0000] <hsivonen>
AryehGregor: anyway, hosting in the U.S. makes a great deal of sense if you have a U.S.-based legal department anyway

[05:28:29.0000] <AryehGregor>
Yes, that's certainly true.

[05:28:32.0000] <hsivonen>
AryehGregor: otherwise, not so much

[05:28:42.0000] <AryehGregor>
The best way to minimize legal liability in any event is to be in as few countries as possible.

[05:31:47.0000] <hsivonen>
(aside: I've occasionally wondered how Google dares to have a data center in Finland--a country that doesn't have proper judicial review of search warrants.)

[05:32:45.0000] <hsivonen>
(maybe they count on the officers who could perform a search being sufficiently afraid of Google's ability to pursue after-the-fact remedies)

[05:34:19.0000] <hsivonen>
(if the police in the U.S. had the kind of search powers the Finnish police has on paper, there'd be no end to abuse of police power in the U.S.)

[05:45:33.0000] <AryehGregor>
Fortunately, U.S. police search powers are very effectively curtailed by the exclusionary rule, since that means that it's generally pointless for them to perform illegal searches.

[06:38:13.0000] <AryehGregor>
Ms2ger, why do we even want contains()'s param to be "Node?" rather than just "Node"?

[06:38:46.0000] <Ms2ger>
Because smaug: https://www.w3.org/Bugs/Public/show_bug.cgi?id=14017

[06:39:39.0000] <Ms2ger>
Hmm, Opera doesn't support Node.contains?

[06:52:49.0000] <ruby_on_tails>
so, document.execCommand("insertBrOnReturn", false, "true"); throws error on opera, what to do ?

[06:53:53.0000] <odinho>
Ms2ger:  >>> document.querySelector("div").contains(document.querySelector('span'))

[06:53:56.0000] <odinho>
true

[06:54:03.0000] <odinho>
Ms2ger: Seems to here.

[06:55:18.0000] <Ms2ger>
Huh

[06:55:27.0000] <Ms2ger>
Aha

[06:55:36.0000] <Ms2ger>
Not on Node, it doesn't

[06:55:42.0000] <Ms2ger>
w(document.contains(document.querySelector('span')))

[06:55:51.0000] <Ms2ger>
error: Uncaught exception: TypeError: 'document.contains' is not a function on line 3

[07:04:07.0000] <annevk>
IE only had it on Element I think?

[07:04:14.0000] <Ms2ger>
Possible

[07:05:44.0000] <annevk>
https://plus.google.com/111991826926222544385/posts/QdGfrgtP6Eg MikeSmith is blogging again, teehee

[07:07:07.0000] <hsivonen>
annevk: I guess being called for trolling like that means it's time for introspection. (Though see the comment I left there.)

[07:08:34.0000] <annevk>
I'm not sure why you were trolling

[07:09:46.0000] <hsivonen>
annevk: well, I wasn't in the sense that I didn't intend to troll; I just think that the main value of snapshots is the PP, which admittedly was probably too much inside baseball for a tweet

[07:10:17.0000] <annevk>
I have not seen any other value demonstrated

[07:10:22.0000] <annevk>
or explained

[07:10:37.0000] <annevk>
so unless I missed a memo somewhere...

[07:11:05.0000] <Ms2ger>
"Microsoft"

[07:11:29.0000] <hsivonen>
interesting that Steve's FUD spree on Twitter didn't get called out, though

[07:11:48.0000] <hsivonen>
Ms2ger: isn't that part of the "PP" package?

[07:12:48.0000] <Ms2ger>
Well, maybe if you incorrectly assume they send useful feedback...

[07:13:35.0000] <hsivonen>
Ms2ger: I meant in the sense that having stuff under the PP makes it more likely that Microsoft implements

[07:13:48.0000] <Ms2ger>
Shrug

[07:14:15.0000] <Ms2ger>
I tend to assume that they'll implement eventually

[07:14:29.0000] <odinho>
When sitecompat problems grow?

[07:14:31.0000] <hsivonen>
annevk: your "I'm not sure..." sentence above is a bit ambiguous. Did it seem to you like I was trolling?

[07:14:35.0000] <Ms2ger>
Yeah

[07:14:58.0000] <annevk>
hsivonen: sorry, why you were considered to be trolling

[07:15:12.0000] <annevk>
odinho: or when developer complaints grow

[07:15:32.0000] <odinho>
annevk: Ofc, that's a good place too.

[07:15:37.0000] <karlcow>
annevk> I have not seen any other value demonstrated

[07:15:37.0000] <karlcow>
I guess we all have a values system which makes it very difficult to be able to see. It's why cultures exist.

[07:17:35.0000] <annevk>
fortunately we have you

[07:18:18.0000] <Ms2ger>
Heh

[07:19:40.0000] <karlcow>
annevk: unfortunately I have me

[07:20:43.0000] <Ms2ger>
I wonder by which definition part of community moving off and creating their own version without apparent coordination is a "branch" rather than a "fork"

[07:34:00.0000] <Stevef_>
hsivonen: if there is any of what you call FUD here: http://www.paciellogroup.com/blog/2012/07/w3c-html-and-whatwg-html-split-affect-on-html-accessibility/ please feel free to disabuse me

[07:36:55.0000] <annevk>
Stevef_: e.g. the bit about AT APIs is nonsense, we did show interest in that and discussed at some F2F

[07:37:12.0000] <annevk>
Stevef_: it was then decided it was best to work on that separately for now

[07:38:12.0000] <bencc1>
is it possible that a websocket will fail after successful connection?

[07:38:31.0000] <bencc1>
or if a connection is established I can assume that websocket works for that user?

[07:38:35.0000] <annevk>
bencc1: sure, if the network drops

[07:38:49.0000] <Stevef_>
annevk: not nonsense at all I was told by hixie that the document was not needed and not worthwhile, that browsers had been implementing it fine already (demonstratably nonsense)

[07:38:50.0000] <bencc1>
annevk: I mean stuff like proxy issues

[07:39:31.0000] <bencc1>
annevk: if websocket fails I'm using long-polling. my question is if ws.onconnect is being called, I can assume that long-polling is not needed

[07:39:31.0000] <annevk>
Stevef_: does not compute with "never shown interest" though

[07:40:16.0000] <Stevef_>
annevk: mincing words me thinks

[07:40:32.0000] <annevk>
Stevef_: "never shown interest" is FUD

[07:40:41.0000] <annevk>
Stevef_: you can call it mincing words, whatever

[07:43:15.0000] <Stevef_>
annevk: apart from that?

[07:43:29.0000] <Stevef_>
will mark it as FUD

[07:44:55.0000] <annevk>
dunno

[07:45:16.0000] <annevk>
that you managed to get some of your pet issues through the HTML WG and not through the WHATWG does not mean the WHATWG does not care about accessibility

[07:45:28.0000] <annevk>
but you obviously feel otherwise about that

[07:45:52.0000] <annevk>
and I don't really care

[07:46:37.0000] <annevk>
mostly curious what this looks like a year or five years from now

[07:46:53.0000] <Stevef_>
annvk: i have the same antipathy towards your set of beliefs

[07:47:44.0000] <karlcow>
Q.E.D.

[07:48:28.0000] <Stevef_>
annevk: but I have decided to engage with the WHATWG never the less

[07:48:56.0000] <annevk>
my beliefs?

[07:50:35.0000] <karlcow>
☮ ♥ ✌ please

[07:51:33.0000] <Stevef_>
karlcow: what's so funny about  peace love and understanding :-)

[07:52:25.0000] <karlcow>
Stevef_: nothing funny. Just need more of it.

[07:52:54.0000] <Stevef_>
karlcow: its a song reference if you didn't know

[07:53:08.0000] <karlcow>
I didn't know

[07:54:58.0000] <karlcow>
/me DDGed it and found Elvis Costello

[07:56:16.0000] <Stevef_>
Karlocw: originally by nick lowe http://en.wikipedia.org/wiki/%28What%27s_So_Funny_%27Bout%29_Peace,_Love,_and_Understanding

[08:00:12.0000] <Stevef_>
annevk: if you can provide anything concrete on it not being FUD, then I am happy to retract and apologise for FUD

[08:05:20.0000] <Ms2ger>
odinho, could you file contains-on-node or should I?

[08:07:27.0000] <AryehGregor>
Stevef_, I think the discussions between you and various WHATWG participants are often characterized by substantive points made on both sides that are obscured by inflammatory language, which encourages each side to dismiss the other.  So for instance, in this discussion just now, both you and annevk referred to things the other one said as "nonsense".  I would comment on the actual blog post you linked to, but it's not loading for me.

[08:12:16.0000] <Stevef_>
ArehGregor: i have attempted to be civil for the short while i have been joining the whatwg IRC, but i find that annevk in particular appears to only thinly veil or not veil at all his lack of respect for me, in such cases I find it difficult not to respond in kind, but will endeavour not to jump at the bait in future.

[08:15:22.0000] <odinho>
Ms2ger: Are you asking me to do it? If so, I can do it for you, but I was doing something else right now :P

[08:15:47.0000] <Ms2ger>
I can do it, just don't want to duplicate work

[08:16:11.0000] <Ms2ger>
And I don't have access to your BTS to check myself (wink nudge)

[08:18:31.0000] <odinho>
Ms2ger: Hmm. Yea, can't seem to find a bug like that. I'll make it. Do you have an URL handy by chance?

[08:18:41.0000] <odinho>
(to the place in the spec that is)

[08:19:17.0000] <Ms2ger>
http://w3c-test.org/webapps/DOMCore/tests/submissions/Ms2ger/Node-contains.xml / http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-node-contains

[08:19:20.0000] <Ms2ger>
Thanks

[08:19:38.0000] <odinho>
Ms2ger: Cool, love it when I can just put a TC directly in without doing it :P

[08:19:49.0000] <Ms2ger>
So do I :)

[08:21:07.0000] <Ms2ger>
/me should work some more on those workers tests

[08:25:32.0000] <annevk>
Stevef_: All I am saying is that spreading the message that WHATWG does not care about accessibility is spreading a lie. If we did not care about it, there would be no discussion about it, longdesc would not have been removed, etc.

[08:26:15.0000] <annevk>
Stevef_: There is disagreement about what the right way forward is, but no disagreement about whether accessibility is a field worthy of attention

[08:32:48.0000] <Stevef_>
annevk: i don't believe I have said that WHATWG does not care about accessibility, what I have said is that in regards to accessibility API work, it is being done at the W3C and I or my fellow editors never got anythting positive on the work or interest shown in the work from the WHATWG side. In regards  to the other aspects in the example i cited I did a put a lot of effort testing ,...

[08:32:49.0000] <Stevef_>
...researching, speaking with users on the effects of  the accessibility of title attribute as implemented in the real world and based my arguments on that, with little or no interest in arguing the issue on its merits, the work was largely ignored or dismissed by the WHATWG. What I do belive is that accessibility issues get more consideration at the W3C, but that is something we disagree upon

[08:57:32.0000] <annevk>
Stevef_: I think Hixie's main point about html-api-map is that it can use some cleanup

[08:58:52.0000] <annevk>
Stevef_: e.g. HTML in its section WAI-ARIA already defines the default roles

[08:59:03.0000] <annevk>
Stevef_: so that information should not be normative in html-api-map because it's redundant

[08:59:24.0000] <annevk>
Stevef_: but how e.g. "link role" maps to system APIs is new information and should be normative

[09:00:14.0000] <annevk>
Stevef_: that's probably why Hixie largely dismissed the document

[09:00:21.0000] <AryehGregor>
Stevef_, I haven't followed these discussions closely, but I really think the mutual attitude of dismissiveness comes more from the way the discussion is phrased than its substantive content.  The WHATWG people don't understand or agree with some of the things you're saying, this makes you get frustrated because you feel you can't make them listen to you, you lash out, they get annoyed at you and stop listening.  I don't think the problem is

[09:00:21.0000] <AryehGregor>
that anyone doesn't care about your work; it's some combination of a) you aren't phrasing it in a way that makes sense to them, and b) they have different assumptions or priorities.

[09:01:09.0000] <Stevef_>
annevk: its a work in progress and if anybody bothered to look at it lately its got a lot better, none of the formation is normative currently, the ARIA stuff is there for informative purposes, some of the ARIA mappings are differnet from what is currently in HTML as they are implemnted differently

[09:01:40.0000] <annevk>
Stevef_: I was looking at it just now

[09:01:50.0000] <annevk>
maybe I got the wrong version though?

[09:01:53.0000] <AryehGregor>
And then both sides find it easier to just ignore the others and do their own thing, which mostly means the WHATWG people win because implementers mostly listen to them.

[09:01:54.0000] <AryehGregor>
/me shrugs

[09:04:04.0000] <Stevef_>
areygregor: not true of acc implementation stuff as there is little or no  guidance from the WHATWG side, I work directly with the implementers from mozilla, webkit/chrome and we have cahnnesl into IE as one of the editors is from microsoft

[09:04:21.0000] <AryehGregor>
Okay.

[09:04:32.0000] <AryehGregor>
As I say, I haven't followed the discussions closely.

[09:04:51.0000] <Stevef_>
i spend a reasonable amount of my time filing bugs on the various browsers

[09:05:57.0000] <Stevef_>
annevk: as you know the editors draft is the only worthwhile copy http://dvcs.w3.org/hg/html-api-map/raw-file/tip/Overview.html

[09:06:09.0000] <annevk>
Stevef_: right so I was looking at that

[09:06:35.0000] <annevk>
got to go now

[09:07:34.0000] <Stevef_>
areyhgregor: thats because llittle of the dicsussion occurs on whatwg irc

[09:07:43.0000] <dglazkov>
good morning, Whatwg!

[09:09:11.0000] <AryehGregor>
Stevef_, I've seen the threads, I just don't read them because they don't relate to my work.  I mostly do editor-related stuff for Mozilla right now, and presumably that has a11y implications somehow, but no one told me about them, so I guess someone else must be handling them.

[09:10:55.0000] <AryehGregor>
Also, Chatzilla should support nick autocomplete, namely typing the first few letters of a nickname and then hitting Tab to complete it.  That way, you'll spell it correctly, and my IRC client will notify me that someone said my name.

[09:11:21.0000] <Stevef_>
aryehgregor: irc://irc.mozilla.org/accessibility is where the mozilla acc implementation work/dicussion occurs

[09:11:45.0000] <AryehGregor>
I'm not surprised.  I also think the accessible/ dir in the source tree has something to do with it.

[09:12:08.0000] <Ms2ger>
I can verify that :)

[09:12:41.0000] <Stevef_>
aryehgregor: yes that is a mine of info as to how acc is implemented

[09:19:53.0000] <beowulf>
/me wanders into whatwg and finds the same basic discussion happening a year or so later

[09:20:04.0000] <AryehGregor>
beowulf, only one year?

[09:20:50.0000] <beowulf>
AryehGregor: i'm not sure, it might be two

[09:38:40.0000] <tabatkins_>
hober: You still planning to do something about masks in CSS?

[10:03:19.0000] <Hixie>
i guess today I get to look into speccing this autofill feature

[10:13:40.0000] <Hixie>
tantek: is the list of tokens here ok with you? i remember that you had concerns about it not matching vCard or something: http://wiki.whatwg.org/wiki/Autocompletetype#4.10.7.3.1.1_Specifying_field_data_type_hints

[10:17:31.0000] <tantek>
Hixie - will take a look in a bit

[10:20:43.0000] <annevk>
Hixie: fwiw, that list is mostly for marking up existing forms on the web including those that are not taking advantage of new features

[10:21:20.0000] <Hixie>
if they use this feature, they're taking advantage of new features :-P

[10:22:06.0000] <annevk>
Hixie: sure, what I mean is that splitting a birthday in three separate fields is not necessarily something we want to support

[10:22:45.0000] <Hixie>
how woudl you support the existing forms if you don't?

[10:23:46.0000] <Hixie>
tantek: based on the thread it looks like the list is intended to match hCard, though some people asked if it could instead match schema.org

[10:24:11.0000] <annevk>
Hixie: I guess the question is whether we should cater to pre-WF2-forms

[10:24:44.0000] <Hixie>
ah right

[10:24:52.0000] <Hixie>
if it's easy to do so, it doesn't seem harmful

[10:24:57.0000] <Hixie>
but i don't have a strong opinion one way or the other

[10:25:05.0000] <Hixie>
does IE do the new types yet?

[10:25:23.0000] <tantek>
annevk - some (many?) avoid WF2 forms controls because of the crappy UI, e.g. input type=date

[10:25:53.0000] <tantek>
so it's probably better to not depend on them, but do take advantage of them if they're in use

[10:25:55.0000] <annevk>
not sure, and I guess it makes sense to support legacy forms too, it just seems kind of odd from a conformance perspective

[10:26:24.0000] <tantek>
depends on if you want the feature more or less adopted, more or less practical for authors vs. purity (per design principles)

[12:10:45.0000] <david_carlisle>
MikeSmith:  (and hsivonen)  I'd agree with losing the doctype based schemas in the v.nu interfaces (at least the web facing ones, local installations might be different)  although "<MikeSmith> I don't want to encourage people to use XHTML1 or HTML4 doctypes " I'd agree so it's unfortunate that the bug asking to have  a doctype for xhtml+xml that wasn't one of those (and works) just got...

[12:10:47.0000] <david_carlisle>
...wontfixed :(

[12:27:42.0000] <Hixie>
david_carlisle: if there are doctypes that work and aren't in the spec, please reopen the bug listing them. I said in the bug already that if there are cases that already work it's a no-brainer to add them. It's the ones that don't already work that I wontfixed.

[12:44:08.0000] <Yuhong>
Personally, I use the HTML4 Transitional option mostly when I want to suppress the deprecation warnings/errors.

[13:00:45.0000] <Hixie>
Yuhong: why would you want to use deprecated stuff?

[13:01:01.0000] <Yuhong>
To validate other pages, not my own.

[13:01:08.0000] <Hixie>
why do you validate other pages?

[13:03:18.0000] <Yuhong>
For fun, mostly, but sometimes there are reasons to do so.

[13:03:46.0000] <Yuhong>
What is wrong with validating other pages?

[13:07:28.0000] <tantek>
Hixie it is good to see that the autocomplete stuff has seen some iteration to better match hCard.

[13:07:42.0000] <tantek>
There's actually been a lot of prior art/research on the subject here: http://microformats.org/wiki/hcard-input

[13:08:05.0000] <tantek>
(nothing conclusive, but the research may be useful to help inform decisions for the autocomplete stuff)

[13:09:53.0000] <tantek>
(aside: the suggestion to use schema instead of vCard had to either be a joke/troll, or naïvete - surprised they didn't also try using Google Base, Google Data, FOAF, or if you really want to go back, MS Passport contact schemas for that matter)

[13:16:34.0000] <tantek>
Hixie - shall I go ahead and edit the wiki page directly with suggested fixes and simplifications?

[13:18:09.0000] <TabAtkins>
Hixie: What do you think about adding a toString() to Path that produces an SVG path string?

[13:18:30.0000] <tantek>
ok I'm going to edit first, ask (more) questions later

[13:18:47.0000] <TabAtkins>
Hixie: SVGWG would like to integrate with Path as possible, and it would be cool to be able to assign one directly to a <path> or whatever.

[13:20:22.0000] <Yuhong>
tantek: On modularization, remember XHTML 1.1

[13:20:24.0000] <Yuhong>
tantek: On modularization, remember XHTML 1.1?

[13:21:14.0000] <tantek>
Yuhong - yes, I was in the HTML working group at the time and very much remember the problems.

[13:21:27.0000] <tantek>
Hence why I'm suggesting the CSS model of modularization rather than the XHTML model.

[13:21:40.0000] <tantek>
XHTML modularization was both a) a disaster, and b) nearly useless

[13:21:42.0000] <Yuhong>
Which reminds me that if I was there in 2004, I would suggest proposing a new version of XHTML that IS backward compatible with XHTML1.

[13:22:25.0000] <Yuhong>
Before resorting to starting the WHATWG.

[13:22:31.0000] <tantek>
Yuhong - see my blog posts about it from that era, e.g. http://tantek.com/log/2003/01.html#L20030114t1345

[13:23:09.0000] <tantek>
I was pushing for backward compatible work on XHTML1/HTML4 back in 2003 - *in* the HTML Working Group itself. Unfortunately I was the only one advocating this focus (in the WG).

[13:23:40.0000] <tantek>
so yes, I've witnessed firsthand both a modularization failure (XHTML) and a modularization success (CSS2.1/3)

[13:23:57.0000] <Yuhong>
And unfortunately what got submitted by Mozilla/Opera was not that.

[13:24:24.0000] <tantek>
thus given that data, I'm advocating CSS-style modularization, where we agree to modularize on top of a monolithic spec, rather than to replace it.

[13:24:24.0000] <Yuhong>
Back in 2004, of course.

[13:24:39.0000] <tantek>
thus as we have CSS 2.1, we now have HTML5

[13:24:59.0000] <tantek>
and we have CSS3 modules that build on top of CSS2.1, we can have HTML modules that build on top of HTML5.

[13:27:03.0000] <Yuhong>
Sorry, did not realize the difference between CSS and XHTML.

[13:27:43.0000] <Yuhong>
modularization.

[13:27:43.0000] <tantek>
there's more than one way to modularize

[13:27:52.0000] <Yuhong>
Agreed.

[13:28:19.0000] <tantek>
so it's easy enough to pick one that is mostly working, and improve on it, rather than the one that was a nearly a completely failure

[13:29:05.0000] <Yuhong>
Of course.

[13:29:07.0000] <tantek>
Yuhong - you'll notice that WHATWG does tend to leave CSS stuff to the W3C CSS WG - because it tends to do a good enough practical job that it's not worth reinventing.

[13:29:30.0000] <tantek>
it's not perfect, but no standards org/community is.

[13:35:01.0000] <Hixie>
yutak: nothing wrong, just curious :-)

[13:35:04.0000] <Hixie>
tantek: cool, thanks for the link

[13:35:23.0000] <tantek>
regarding the autocomplete stuff, it's funny that it fails to mention the prior art by Microsoft - the VCARD_NAME field: http://msdn.microsoft.com/en-us/library/ms533032%28VS.85%29.aspx

[13:35:26.0000] <Hixie>
tantek: ideally post them to the list with rationale

[13:35:46.0000] <tantek>
(research that's in the microformats wiki on this subject)

[13:35:58.0000] <Hixie>
tantek: i'm probably adding them to the spec today or tomorrow, but they can always change after

[13:36:01.0000] <tantek>
judging by the screenshots, I'd estimate MS has supported that since IE6

[13:36:18.0000] <tantek>
Hixie - do you care at all about IE's implementation?

[13:36:35.0000] <tantek>
or should we design a brand new feature here?

[13:36:39.0000] <Hixie>
tabatkins: seems not unreasonable, if svg paths can describe everything canvas paths can do... i guess someone would have to come up with an algorithm

[13:36:46.0000] <tantek>
(I'm not saying there's an obvious right answer - I'm just looking for your general thinking on the matter)

[13:37:28.0000] <Hixie>
tantek: i care about all relevant data

[13:37:40.0000] <Hixie>
there's been a bunch of attempts at addressing this

[13:37:52.0000] <Hixie>
post mortems for those are what i'd really love to see

[13:37:59.0000] <tantek>
"shipping in IE for many years" is a bit stronger than "attempt" ;)

[13:38:20.0000] <tantek>
as I think you said when this topic first came up, it's not clear how many web developers care about this at all

[13:38:51.0000] <Hixie>
netscape shipped support for ECM for years too

[13:39:01.0000] <Hixie>
there's lots of shipped failures in this field :-(

[13:39:09.0000] <tantek>
I think IE might still support this

[13:39:35.0000] <tantek>
longest continuous (still) shipping implementation likely counts for more than abandoned attempts

[13:40:14.0000] <tantek>
maybe I'll ask on Twitter if anyone has heard of or uses it, or if IE even supports it still

[13:40:36.0000] <tantek>
do you have a link to the "netscape shipped support for ECM" documentation so I can add it to my research?

[13:40:50.0000] <Hixie>
i do not

[13:42:36.0000] <tantek>
ok I'll mention it as anecdotal then - let me know if you do find something

[13:42:50.0000] <tantek>
do you have a rough year range of when netscape shipped support for ECM?

[13:43:05.0000] <Hixie>
6.x range

[13:43:10.0000] <Hixie>
iirc

[13:43:40.0000] <Hixie>
man, searching the net for vcard_name is unfulfilling

[13:44:56.0000] <Velmont>
Yeah, if only someone here worked at a company knowing lots about search or something... :/

[13:45:04.0000] <Velmont>
:P

[13:46:22.0000] <Hixie>
?

[13:46:26.0000] <Hixie>
i don't mean the results are bad

[13:46:36.0000] <Hixie>
i mean the results show that people don't really know what it is

[13:47:59.0000] <tantek>
Hixie, are you searching HTML source for it?

[13:48:20.0000] <Hixie>
i was looking for tutorials and examples of people using it

[13:48:20.0000] <tantek>
for usage stats?

[13:48:32.0000] <tantek>
the only tutorials I know of are on MSDN

[13:48:38.0000] <tantek>
for VCARD_NAME

[13:48:45.0000] <tantek>
like most microsoft-only extensions

[13:48:47.0000] <Hixie>
hence my using a search engine to look for others :-P

[13:49:40.0000] <tantek>
can you do a search over crawl data for VCARD_NAME in the source? I'm curious about adoption in the wild.

[13:49:54.0000] <tantek>
(I don't know of any actual sites that use it - would be useful to have a few examples to look at)

[14:05:37.0000] <tantek>
I'm going to ask folks in #microformats to take a look at our research and iterate on it as well

[14:06:23.0000] <tantek>
anyone interested in autocomplete / vCard auto-fill type stuff should feel free to jump into #microformats and contribute - we're a fairly friendly bunch

[14:06:25.0000] <tantek>
(though we do ask for the usual, use-cases, data, real world examples, and sometimes tend to summarily dismiss theoretical examples and wishful thinking)

[14:20:08.0000] <annevk>
tantek: fwiw, the CSS-stye modularization is starting to become a mess for at least implementors and specification precision

[14:20:54.0000] <annevk>
not sure I'd classify it as a success

[14:22:31.0000] <tantek>
annevk - if you mean "total success" no

[14:22:41.0000] <tantek>
however *a* success, certainly yes, based on two very easy metrics

[14:22:49.0000] <tantek>
1. for spec-types - modules have made it to REC

[14:23:00.0000] <tantek>
2. for devs - browsers have interop on several modules

[14:23:01.0000] <Hixie>
making to REC is hardly a measure of success :-)

[14:23:09.0000] <Hixie>
2 is valid though

[14:23:22.0000] <tantek>
Hixie - I knew you were going to say that :P

[14:23:47.0000] <tantek>
perhaps I should have said "for process-type" rather than "for spec-types" for 1.

[14:24:16.0000] <tantek>
annevk - also, I consider overall CSS modularization a success especially when *some* modules fail.

[14:24:50.0000] <Hixie>
the biggest problem with CSS is the vagueness of the specs

[14:24:51.0000] <tantek>
that means that we successfully isolated features that the market was not as interested in, and stopped them from holding back features that the market wanted

[14:25:09.0000] <Hixie>
but that's more or less inevitable for the modules because they all rely on a foundation that is too vague

[14:25:11.0000] <tantek>
Hixie - s/CSS/specs in general

[14:25:18.0000] <Hixie>
no, not any more

[14:25:20.0000] <annevk>
euhm, features that the market wanted? you mean like a better layout model?

[14:25:33.0000] <annevk>
I mean sure, some tiny stuff was successful, such as color

[14:25:51.0000] <Hixie>
tantek: XHR, HTML, Origin, CORS, lots of the new APIs -- lots of new specs these days are written very tightly and not vaguely

[14:25:53.0000] <annevk>
but that's not really what most people are waiting for

[14:25:58.0000] <tantek>
Hixie - I think "too vague" is a bit harsh - based on the 2.1 test suite and impl reports.

[14:26:12.0000] <Hixie>
tantek: don't get me wrong, CSS is way better than what came before the 2.1 era

[14:26:26.0000] <Hixie>
tantek: but it hasn't kept improving, and is now no longer close to cutting edge in terms of precision

[14:26:52.0000] <tantek>
and as I said, CSS modularization is not perfect at all - it's just shown a degree of success (in contrast to other modularization attempts which have shown nearly total faliure)

[14:27:07.0000] <Hixie>
it's aroudn the same level of precision as SVG specs, which is to say, way better than the status quo around 2000, but not at the level we expect in 2010

[14:27:14.0000] <tantek>
Hixie - we just need to get TabAtkins to stop renaming Flexbox properties ;)

[14:27:37.0000] <tantek>
Hixie - fair enough, the bar for writing specs has been raised.

[14:27:38.0000] <Hixie>
what things are named isn't really a matter of precision :-)

[14:28:21.0000] <tantek>
even though we know how to write better specs, I would offer that lack of spec precision isn't stopping implementers of the CSS modules

[14:28:44.0000] <tantek>
so yes, the specs *can* be better written, however, they are doing a 'good enough' job to help the platform move forward

[14:29:10.0000] <tantek>
could we do an even better job? of course. do we have enough editors to do so? no. so we settle for incremental improvements.

[14:30:58.0000] <Hixie>
clearly lack of precision has never stopped implemntors

[14:31:03.0000] <Hixie>
witness HTML in the 90s

[14:31:08.0000] <Hixie>
or CSS in the 90s

[14:31:17.0000] <Hixie>
or HTTP

[14:31:18.0000] <Hixie>
or...

[14:31:58.0000] <Hixie>
it just results in poorer interop

[14:32:44.0000] <tantek>
right, and as precision has improved, so has interop

[14:32:56.0000] <Hixie>
precision and test suites, yeah

[14:32:58.0000] <tantek>
thus measured by interop, CSS 2.1 (incl test suite) has done a decent job

[14:33:30.0000] <Hixie>
for what's tested

[14:33:38.0000] <tantek>
/me looking forward to seeing what gets dropped from HTML5 during CR due to lack of impl/interop per test suites.

[14:33:39.0000] <Hixie>
i mean, entire sections of that spec are intentionally undefined

[14:33:41.0000] <Hixie>
e.g. the table model

[14:38:18.0000] <tantek>
yeah

[14:39:03.0000] <tantek>
Hixie, re: autocomplete, would it be reasonable to ask you to shuffle your task items to postpone spec'ing it for a week to see what the microformats community can come up with in terms of updating the research documented at http://microformats.org/wiki/hcard-input and sub-pages?

[14:39:10.0000] <tantek>
have just today/tomorrow is a bit short notice.

[14:45:00.0000] <Hixie>
it's hardly short notice, you participated in the threads back in january :-)

[14:47:49.0000] <tantek>
is it reasonable to ask for a week of updating research etc.?

[14:48:15.0000] <tantek>
back in January (or was it December?) it seemed like a Google-only thing

[14:48:23.0000] <tantek>
that didn't do any previous research etc.

[14:48:34.0000] <tantek>
now that it's gotten some basis in previous work, it's more interesting

[14:50:16.0000] <Hixie>
well the spec can always change :-)

[14:50:29.0000] <Hixie>
i haven't done anything yet, but it's what i'm workong on now

[14:51:39.0000] <tantek>
sure, the spec can always change, however you've shown a pattern of less change rather than more after a first iteration "lands" in the spec

[14:52:46.0000] <tantek>
also - implementers seem to take your first landing of a feature as a go signal to implement - so it has a lot of weight (even if unintentionally so) - gets press articles written, etc.

[14:52:48.0000] <Hixie>
i do less change after something exists to change than before something exists to change? o_O

[14:52:51.0000] <tantek>
sometimes even flamewars ;)

[14:53:11.0000] <Hixie>
well chrome already implements something on these lines, so that's not a concern here :-)

[14:53:14.0000] <tantek>
btw - overall comment - too many feature ;)

[14:53:16.0000] <tantek>
features

[14:53:20.0000] <tantek>
in autocomplete

[14:53:31.0000] <Hixie>
anyway, if you want to send feedback, send feedback :-)

[14:54:03.0000] <Hixie>
i assure you i take it as much into account if it's sent before i've looked at a thread as if it's sent after

[14:54:47.0000] <tantek>
will do

[14:55:08.0000] <tantek>
I'd just rather take the time to send you feedback with citations rather than feedback off the top of my head.

[14:56:51.0000] <tantek>
also - has anyone seen isherman around? would love to be able to do some of this interactively with him (since he wrote up and iterated http://wiki.whatwg.org/wiki/Autocomplete_Types )

[15:07:58.0000] <Hixie>
hmm, https://src.chromium.org/viewvc/chrome/trunk/src/chrome/test/data/autofill/heuristics/input/ is apparently a list of forms synthesised from real world forms which apparently the chrome team did when they originally came up with this

[15:18:11.0000] <tantek>
Hixie - thanks - will add that to the hcard-input-examples page for further analysis

[15:23:28.0000] <tabatkins>
tantek: Slanderer. ^_^ I stopped renaming things once we entered LC, and successfully blocked an attempted rename recently.

[15:28:35.0000] <tantek>
TabAtkins - I get all my news from W3Cmemes.

[15:29:07.0000] <tabatkins>
There's your problem.

[15:52:04.0000] <Hixie>
tantek: why is the country in hCard country-name rather than country?

[15:52:31.0000] <tantek>
because that's the phrasing the vCard spec used and in the interest of minimal change I chose their literal phrasing

[15:52:42.0000] <tantek>
same with "postal-code"

[15:52:55.0000] <Hixie>
o_O

[15:53:06.0000] <tantek>
(modulo singularization of plural names)

[15:53:26.0000] <Hixie>
seems a bit silly to derive a formal term from inconsistent prose rather than making it consistent :-)

[15:53:38.0000] <tantek>
nah, the prose was fairly consistent across properties

[15:53:47.0000] <tantek>
as what happens when people copy/paste property descriptions

[15:53:47.0000] <Hixie>
vcard 4 uses just "country" in the BNF

[15:53:55.0000] <tantek>
oh does it

[15:53:57.0000] <tantek>
/me looks

[15:55:06.0000] <tantek>
alrighty then, first google result for "vcard4" is a wiki page I created. thanks google.

[15:55:13.0000] <Hixie>
hehe

[15:55:17.0000] <Hixie>
i get that kind of stuff all the time

[15:55:29.0000] <Hixie>
search for some obscure thing and the first hit is some e-mail i wrote asking for help on the very topic

[15:56:20.0000] <tantek>
Hixie - in 6.3.1 it uses "country name" http://tools.ietf.org/html/rfc6350#section-6.3.1

[15:56:33.0000] <Hixie>
yes but that's the prose

[15:56:41.0000] <Hixie>
the formal bnf uses just "country"

[15:56:43.0000] <tantek>
historically the prose tends to be more accurate

[15:56:50.0000] <tantek>
since that's what human read/review

[15:56:51.0000] <Hixie>
the prose here is inconsistent

[15:56:58.0000] <tantek>
tons of problems in the grammar have been found over time

[15:56:59.0000] <Hixie>
uses "country name" but not "city name", e.g.

[15:57:22.0000] <Hixie>
anyway, my point is that it's dumb to use "-name" on the country field and not the others, it adds nothing :-)

[15:57:43.0000] <tantek>
so you're arguing from a shortening/simplicity perspective

[15:57:46.0000] <tantek>
to make an exception

[15:57:47.0000] <tantek>
b

[15:57:48.0000] <tantek>
e

[15:57:49.0000] <tantek>
ca

[15:57:50.0000] <tantek>
us

[15:57:51.0000] <tantek>
e

[15:57:52.0000] <tantek>
(darnit)

[15:58:02.0000] <tantek>
because the prose is fairly consistent in 6.3.1

[15:58:06.0000] <tantek>
         the post office box;

[15:58:06.0000] <tantek>
         the extended address (e.g., apartment or suite number);

[15:58:07.0000] <tantek>
         the street address;

[15:58:09.0000] <tantek>
         the locality (e.g., city);

[15:58:10.0000] <tantek>
         the region (e.g., state or province);

[15:58:11.0000] <tantek>
         the postal code;

[15:58:12.0000] <tantek>
         the country name (full name in the language specified in

[15:58:21.0000] <Hixie>
i'm arguing for consistency within the grammar itself, rather than with prose that is internally inconsistent

[15:58:26.0000] <tantek>
the {propertyname} {(optional parenthetical remark} ;

[15:58:27.0000] <Hixie>
s/grammar/vocabulary/

[15:58:36.0000] <tantek>
where is the prose internally inconsistent?

[15:58:51.0000] <Hixie>
it says "locality" not "locality name", and "country name" not "country"

[15:59:12.0000] <Hixie>
yet gramatically those (and region) should be equivalent

[15:59:24.0000] <tantek>
well if it used "name" as a suffix on all of them, then it would be consistent to drop it on all of them

[15:59:33.0000] <tantek>
I only kept -name on country because that was the only one that had it

[15:59:50.0000] <Hixie>
i understand what you did, i'm just saying it's silly :-)

[15:59:58.0000] <tantek>
it's not silly, it's just literal :)

[16:00:08.0000] <tantek>
so you think it's worth 5 characters to break from that?

[16:00:32.0000] <tantek>
or apply a new property naming rule of drop "-name" from properties from other formats?

[16:00:41.0000] <Hixie>
i think basing names on prose that isn't giving normative vocabulary terms is something that must be done with great care and not blindly

[16:01:26.0000] <tantek>
Hixie - I think a methodology that's more easily reproduceable (in the scientific sense) is better than a picky/careful one that may result in different results depending on who applies it and their opinions

[16:01:34.0000] <tantek>
hence a simpler methodology tends to be better

[16:01:56.0000] <tantek>
anyway - that was done nearly 8 years ago, so now you want to drop the 5 characters?

[16:02:11.0000] <Hixie>
not in hcard, no, of course not :-)

[16:02:14.0000] <tantek>
are you going to want to rename "org" to "organisation" [sic] like the RDF folks did?

[16:02:29.0000] <tantek>
or worse, "orgName

[16:02:31.0000] <tantek>
"

[16:02:33.0000] <tantek>
?

[16:02:35.0000] <Hixie>
hcard is shipped, there's not much one can do

[16:02:44.0000] <Hixie>
i'm just being an armchair critic :-)

[16:03:06.0000] <tantek>
I'm on the fence about it

[16:03:22.0000] <tantek>
over the years it has bugged me, but I haven't had sufficient reason to change it

[16:03:53.0000] <Hixie>
but i strongly disagree with your description of how to design a vocab. language design isn't something you should do by methodically transferring terms from a description of another vocab to the new vocab, it's more of an art form, imho

[16:03:54.0000] <tantek>
and IMHO that's the right way with standards, don't change something unless there is a good reason to do so

[16:04:21.0000] <tantek>
nah - the folks that claim "artform" are the ones that create new vocabs out of laziness

[16:04:28.0000] <tantek>
they don't do their home work

[16:04:37.0000] <tantek>
or they want to reinvent the world

[16:04:42.0000] <tantek>
(arrogance, ego)

[16:04:44.0000] <tantek>
like most of schema.org

[16:04:49.0000] <Hixie>
i'm not saying it should be done in the dark

[16:04:54.0000] <gavinc>
tantek: +1

[16:04:58.0000] <tantek>
but that's what happens in practice

[16:05:06.0000] <Hixie>
you obviously want to derive new vocabs from existing vocabs, and even more, from what is learnt from those vocabs having shipped

[16:05:06.0000] <gavinc>
/me goes back to creating vocabs as an art form

[16:05:17.0000] <tantek>
folks that claim vocab is an artform are usually doing so because they're too lazy to be scientific about it

[16:05:42.0000] <Hixie>
i'm just saying that you have to _apply_ those learnings when creating a new vocab, not just copy the mistakes on down, or introduce new ones without a need

[16:06:02.0000] <tantek>
there is some art to vocab design yes, but there is a heck of a lot more principled / constraint based design that goes into it

[16:06:11.0000] <gavinc>
better to copy the old mistake then mostly copy the old mistake and make a bit worse

[16:06:34.0000] <tantek>
Hixie - most folks that try to "fix" old "mistakes" make *worse* mistakes

[16:06:37.0000] <tantek>
that's the problem

[16:06:45.0000] <tantek>
and typically no two people fix the same mistake in the same way

[16:06:46.0000] <Hixie>
you're not most folks

[16:06:49.0000] <tantek>
so you get divergence

[16:06:49.0000] <Hixie>
i expect better of you :-)

[16:06:51.0000] <tantek>
and babble

[16:06:52.0000] <tantek>
hah!

[16:07:06.0000] <tantek>
scientists aren't supposed to assume that they're special (as an individual)

[16:07:33.0000] <Hixie>
language design isn't just a science

[16:07:38.0000] <tantek>
there are exceptions to the "fix" old "mistake" divergence pattern

[16:07:47.0000] <tantek>
the biggest exception I know of

[16:07:52.0000] <Hixie>
it's, to possibly coin a new phrase, evidence-based art

[16:07:57.0000] <tantek>
is that multiple independent parties have renamed "fn" to "name"

[16:08:09.0000] <tantek>
so given that experience of seeing that happen

[16:08:14.0000] <tantek>
we've done that as well in microformats2

[16:08:27.0000] <danbri>
how is microformats2 coming along?

[16:08:34.0000] <tantek>
evidence-based art - I like that

[16:08:40.0000] <tantek>
danbri - quite well

[16:08:51.0000] <danbri>
i dropped a mail to the list a while back, didn't see a reply... but i liked the general direction

[16:09:20.0000] <tantek>
danbri - I replied!

[16:09:31.0000] <tantek>
I think you were even explicitly cc'd on it

[16:09:34.0000] <danbri>
hmm, mail snafu then, investigating!

[16:09:41.0000] <tantek>
could be on either end

[16:09:47.0000] <tantek>
or perhaps email IS the snafu ;)

[16:10:07.0000] <tantek>
Hixie - have you seen other vocabularies "correct" 'country-name' to just 'country" ?

[16:10:18.0000] <Hixie>
tantek: vcard4 :-)

[16:10:29.0000] <tantek>
if you can cite examples of independently agreeing corrections, it's worthy of consideration

[16:10:35.0000] <danbri>
/me remembers going looking to double check it was all quiet :( searching "microformats 2" now I see an unread Tantek reply

[16:10:43.0000] <tantek>
hah - that's just the grammer production being shortened to fit in 70 char widths

[16:10:46.0000] <tantek>
grammar even

[16:10:57.0000] <tantek>
a-ha!

[16:11:09.0000] <tantek>
so you did get the reply. well at least there's that.

[16:11:15.0000] <Hixie>
tantek: adding "-name" is not a length problem, it would still fit

[16:11:21.0000] <danbri>
ah ok it was 20 days later, likely I gave up after a fortnight assuming a billion mails would have swamped it in your inbox subsequently

[16:11:24.0000] <tantek>
danbri - microformats2 is getting a good reception among web designers

[16:11:26.0000] <danbri>
thanks for replying, sorry I missed it!

[16:11:30.0000] <tantek>
the simpler design in particular is greatly applying

[16:11:34.0000] <Hixie>
tantek: i think they just did it for consistency in the grammar (though i'm just guessing) and thought the prose scanned better with some having "name" and some not

[16:11:37.0000] <tantek>
appealing even

[16:11:38.0000] <tantek>
s

[16:11:39.0000] <tantek>
h

[16:11:40.0000] <tantek>
ees

[16:11:40.0000] <tantek>
h

[16:11:42.0000] <tantek>
it may be coffee time

[16:11:55.0000] <tantek>
Hixie - it's been "country-name" since...

[16:12:45.0000] <danbri>
sleep o'clock here

[16:12:50.0000] <danbri>
i'll check in w/  Ben Ward  re parsing

[16:13:04.0000] <Hixie>
on another note, is there a vcard/hcard extension for IMs?

[16:13:07.0000] <tantek>
danbri - feel free to ping on #microformats any time

[16:13:11.0000] <Hixie>
do people just do a new field per IM type?

[16:13:16.0000] <tantek>
Hixie - yes, impp

[16:13:20.0000] <danbri>
will do, cheers

[16:13:26.0000] <Hixie>
or is there a set of terms for "im-type" and "im-whateveryoucallthevalue"?

[16:13:27.0000] <tantek>
it's been incorporated into vCard4

[16:13:32.0000] <tantek>
nah

[16:13:36.0000] <tantek>
it's just a variant of 'url'

[16:13:46.0000] <gavinc>
nwalsh's last version of the vcard owl lines up with hCards naming too, not that too many people care

[16:13:48.0000] <tantek>
the "type" is still URL

[16:13:56.0000] <tantek>
gavinc - URL?

[16:14:07.0000] <Hixie>
what url do you use for, say, aim? or icq?

[16:14:24.0000] <tantek>
Hixie - see tantek.com for an example aim: URL :D

[16:14:33.0000] <danbri>
/me reminded of PortableContacts threads about OpenSocial I meant to read up on

[16:14:36.0000] <gavinc>
tantek: http://nwalsh.com/rdf/vCard

[16:14:50.0000] <Hixie>
is that a scheme that anyone actually consumes?

[16:15:32.0000] <tantek>
Hixie - yes - if you click it, and have e.g. Adium installed, or iChat, it will work

[16:15:41.0000] <Hixie>
wikipedia says people use x-aim, x-icq, x-jabber, x-msn, x-yahoo, x-twitter, x-skype-username...

[16:15:46.0000] <Hixie>
in vcard

[16:15:52.0000] <tantek>
Hixie, re: aim:, Yahoo, MSN, Skype, all documented here: http://microformats.org/wiki/hcard-examples#New_Types_of_Contact_Info

[16:15:56.0000] <gavinc>
It's a scheme that has been used to store hCard extracted data in SPARQL stores

[16:16:10.0000] <gavinc>
but rdf data tends towards foaf over vcard

[16:17:27.0000] <Hixie>
tantek: thanks, perfect

[16:18:19.0000] <gavinc>
the most intesting part being "This ontology attempts to model a subset of vCards in RDF using modern (circa 2005) RDF best practices. The subset selected is the same subset that the microformats community has adopted for use in hCard."

[16:19:37.0000] <tantek>
ICQ… hmm...

[16:19:45.0000] <tantek>
anyone use ICQ and know how to construct icq: URLs?

[16:20:23.0000] <gavinc>
/me learns that ICQ still exists

[16:20:47.0000] <zewt>
every IM ends up still existing forever, for the isolated pockets of people who use it among each other, heh

[16:21:00.0000] <zewt>
(was: re: irc)

[16:21:07.0000] <tantek>
Hixie - which wikipedia article did you find about x-aim, x-icq, x-jabber, x-msn, x-yahoo, x-twitter, x-skype-username ?

[16:21:13.0000] <gavinc>
Yeah, was about to say ... he says on IRC ;)

[16:21:17.0000] <Hixie>
tantek: http://en.wikipedia.org/wiki/VCard

[16:21:49.0000] <tantek>
"a number of which are in common usage." yeah right.

[16:21:53.0000] <tantek>
CITATION NEEDED

[16:21:54.0000] <tantek>
;)

[16:22:00.0000] <zewt>
heh

[16:22:03.0000] <Hixie>
my citation was wikipedia :-)

[16:22:21.0000] <zewt>
also it's 2012 and every messaging system still has its own isolated client; the future is pathetic

[16:22:40.0000] <zewt>
all I want is to be able to contact people without caring what protocol they use or what protocol I'm using, let me pick a person and send a message and have it work

[16:23:28.0000] <zewt>
bluh i hate bad defaults, because inevitably it means that when you change it to the right setting, it ends up being an obscure buggy code path

[16:24:04.0000] <tantek>
zewt, Adium works reasonably well

[16:24:17.0000] <tantek>
indeed, we've regressed since mailto:

[16:24:22.0000] <zewt>
re: changing firefox to not allow pages to open new windows/tabs, which works--except it looks like if a page opens what would normally have been a new window (which navigates into the same window under these settings), it means the page can close the tab even though it wasn't actually opened by script

[16:24:26.0000] <tantek>
and btw, updated: https://en.wikipedia.org/wiki/VCard#vCard_extensions

[16:24:27.0000] <tantek>
:P

[16:24:28.0000] <zewt>
/me grump gripe

[16:26:00.0000] <tantek>
Hixie - please tell me you're not doing format/vocab design based on wikipedia documentation of asserted vCard extensions - most of those are crap / obsolete.

[16:26:46.0000] <tantek>
as they say in school (journalism?), Wikipedia is not a primary source, does not itself make for a good citation. It may help you *find* primary sources / good citations, but that's about it.

[16:28:12.0000] <tantek>
btw just to point out how badly out of date wikipedia can be - that vCard article doesn't list vCard4 properties here: https://en.wikipedia.org/wiki/VCard#Properties

[16:29:19.0000] <gavinc>
tantek: http://hg.adium.im/adium/file/9f195727e81d/Source/AIURLHandlerPlugin.m#l139 is what your referring to for Adium>

[16:29:20.0000] <gavinc>
?

[16:29:45.0000] <tantek>
http://www.adium.im/

[16:29:50.0000] <tantek>
is what I've got installed

[16:29:55.0000] <tantek>
and am reasonably happy with

[16:30:05.0000] <gavinc>
I meant the list of schemes :P

[16:30:07.0000] <tantek>
it's certainly better than the default iChat

[16:30:25.0000] <tantek>
oh look at that

[16:30:27.0000] <Hixie>
tantek: i use all kinds of sources, but yeah, i do use wikipedia

[16:30:39.0000] <tantek>
just as long as you keep in mind how obviously flawed it can be

[16:30:49.0000] <tantek>
all sources are not equal

[16:30:53.0000] <Hixie>
oh don't worry, i've read the HTML-related pages

[16:30:57.0000] <tantek>
hahaha

[16:32:01.0000] <tantek>
gavinc that's an interesting list of schemes, but not enough documentation to actually construct working URIs with each scheme

[16:32:37.0000] <tantek>
which is what the hCard examples provide: http://microformats.org/wiki/hcard-examples#New_Types_of_Contact_Info

[16:33:19.0000] <tantek>
e.g. how do you construct an irc: URI to open a private message window to particular nickname?

[16:33:31.0000] <tantek>
I know this works for a channel: irc://irc.freenode.net/microformats

[16:33:46.0000] <tantek>
but how do you message a nickname on a network?

[16:36:22.0000] <tantek>
nm re: icq scheme - there isn't one. apparently we documented that too: http://microformats.org/wiki/hcard-examples#ICQ

[16:41:58.0000] <Hixie>
tantek: i sent an e-mail rather than updating the spec, with a proposal. so please reply to it soonish if you want to reply before i edit the spec.

[16:43:18.0000] <tantek>
Hixie, I'm surprised to see you suggest "x-" prefixes

[16:43:23.0000] <tantek>
I thought for sure you would be against those

[16:43:35.0000] <tantek>
even the IETF recently obsoleted the "x-" prefix practice

[16:43:43.0000] <Hixie>
for things that are purely experimental and don't ship?

[16:43:53.0000] <tantek>
the problem is things leak

[16:44:08.0000] <Hixie>
there's a place for x- prefixes

[16:44:21.0000] <Hixie>
the problem is when people use them for extensions, rather than experiments

[16:58:36.0000] <gavinc>
<name of person doing experiment>-blah

[16:58:45.0000] <gavinc>
That's unlikely to leak ;)


2012-07-24
[17:01:57.0000] <tantek>
gavinc, oh like <name of company doing experiment> - blah

[17:02:19.0000] <tantek>
or perhaps short abbreviation instead of name, like ms- ;)

[17:22:03.0000] <gavinc>
hehe

[17:42:19.0000] <zewt>
my favorite HTML nonsense of the day: my wifi router uses type=password for keys, even though it fills out the current key, which just means I have to jump to the inspector to find out what the current key is

[17:44:38.0000] <zewt>
not sure if i should blame the router's HTML for that, or browsers for still not having any "show this password" context menu option on password inputs

[17:47:53.0000] <tantek>
Does the spec provide UI/UX guidance in that regard?

[17:48:02.0000] <tantek>
(for <input type=password> )

[17:48:46.0000] <tantek>
zewt - feel free to file a bug in Bugzilla on it for the UI you'd like to see in the browser and we'll see what our security people have to say about it: bugzilla.mozilla.org

[17:49:25.0000] <zewt>
haven't had much good experience filing bugs on firefox so i'm inclined only to spend time for security issues, heh

[01:12:07.0000] <hsivonen>
surprising preliminary result: http://lawrencemandel.com/2012/07/23/mobile-web-compatibility-july-20-2012/

[01:15:59.0000] <tantek>
hsivonen wonderful news

[01:16:33.0000] <tantek>
so perhaps we don't have to sacrifice the platform for the sake of webkit compat

[01:16:44.0000] <rniwa>
nice!

[01:18:45.0000] <hsivonen>
tantek: whether it's wonderful depends on the point of view. It's not wonderful that we can't improve compat unilaterally in our code without evangelism.

[01:19:03.0000] <hsivonen>
(also, the post doesn't say if the aliasing happened in the CSS parser only or also in CSSOM)

[01:20:01.0000] <tantek>
hsivonen - john's crawling showed evidence of sites using prefix properties in CSS, not CSSOM

[01:20:12.0000] <tantek>
if you have evidence of CSSOM dependencies, perhaps you can provide them

[01:20:13.0000] <rniwa>
hsivonen: i guess a good question to ask is whether the experience was worse on Firefox than on WebKit.

[01:20:41.0000] <rniwa>
it might be that those websites don't depend on webkit prefixed properties

[01:21:51.0000] <tantek>
rniwa - the websites were found to have stylesheets that had webkit prefixed properties in their CSS

[01:22:13.0000] <tantek>
whether or not the properties actually matter to the UX of the site - that's a different question

[01:22:22.0000] <rniwa>
tantek: sure but that doesn't necessarily mean that they "depend" on them

[01:22:24.0000] <tantek>
like does anyone notice a 1px difference in rounded borders?

[01:22:25.0000] <rniwa>
tantek: exactly.

[01:22:53.0000] <rniwa>
tantek: i may use webkit prefixed properties for some minor cosmetic reasons but it may not affect UX much.

[01:23:38.0000] <rniwa>
in fact, graceful digression is one of the strengths of css.

[01:23:50.0000] <tantek>
rniwa - precisely - and one of the reasons why the chicken-little "implement all the webkit prefixes!" crowd was talking irrationally

[01:24:21.0000] <rniwa>
tantek: yeah… that was quite radical after all.

[01:24:38.0000] <tantek>
which is why when I introduced the problem in the CSS WG in February, I pointed out we'd only change things justified by data, not the ranting emails/blog posts calling for massive changes in a reactionary manner.

[01:24:44.0000] <rniwa>
tantek: i'm glad people have realized that prefixing DOM APIs is a really bad idea now

[01:25:20.0000] <tantek>
rniwa - really? what's the evidence for that?

[01:25:31.0000] <tantek>
DOM APIs may have a different dynamic than CSS properties - that I'll grant

[01:25:38.0000] <tantek>
but I haven't seen the negative evidence that you speak of

[01:26:11.0000] <rniwa>
tantek: well, imagine we had document.webkitActiveElement, document.mozActiveElement, document.mzActiveElement, etc...

[01:26:18.0000] <rniwa>
ms*

[01:26:33.0000] <hsivonen>
tantek: IIRC, we had a bug where a site used Modernizr to detect support for 3D transforms and then used WebKit-only 3D transforms if Modernizr said yes

[01:26:46.0000] <rniwa>
hsivonen: :(

[01:27:03.0000] <hsivonen>
tantek: it's of course possible that it was an isolated incident and not a pattern

[01:27:15.0000] <annevk>
tantek: instead you rant about other browsers that did feel justified to support -webkit-border-radius?

[01:28:00.0000] <tantek>
annevk - data for different browsers was different

[01:28:21.0000] <tantek>
or so I'm assuming - only Mozilla published data on this AFAIK

[01:28:45.0000] <tantek>
annevk - if you have sources of data for other browsers, please provide and I'll happily link to it

[01:29:49.0000] <hsivonen>
rniwa: how is the realization that prefixing DOM APIs is bad showing in practice?

[01:30:10.0000] <hsivonen>
(I agree that prefixing DOM APIs is bad--I wish Mozilla didn't do it.)

[01:30:29.0000] <rniwa>
hsivonen: i think people are prefixing new DOM APIs less.

[01:30:42.0000] <tantek>
would prefixing the Localstorage APIs have helped with the incompatibilities that we've seen there?

[01:31:01.0000] <hsivonen>
tantek: no, AFAICT

[01:31:02.0000] <tantek>
since that was an instance of *not* prefixing, and then being stuck with implementation-specific behaviors

[01:31:26.0000] <annevk>
localStorage was mostly fast adaption by everyone and then realizing it might not be an ideal API

[01:31:33.0000] <annevk>
adoption*

[01:31:34.0000] <hsivonen>
tantek: did we get stuck with implementation-specific behaviors? I thought we got stuck with ill-considered specced behavior

[01:31:37.0000] <tantek>
specificly, IE8 localstorage, threading

[01:31:51.0000] <tantek>
hsivonen - not sure

[01:33:06.0000] <tantek>
though I suppose implementing WD spec'd behavior is another use-case for prefixes

[01:33:08.0000] <hsivonen>
tantek: Hixie tried to fix the multiprocess race conditions after the fact, but no one seriously wanted to implement the fix

[01:33:31.0000] <tantek>
no one wanted us to implement the fix because they didn't want their code to break

[01:33:38.0000] <tantek>
though if their code was dependent on a prefixed API

[01:33:45.0000] <tantek>
then we could have fixed it in the unprefixed API

[01:33:51.0000] <tantek>
or go from mozFoo to moz2Foo etc.

[01:34:13.0000] <annevk>
no that was not the reason tantek

[01:34:21.0000] <annevk>
the fix was simply not worth the effort

[01:34:21.0000] <tantek>
annevk - happy to be corrected

[01:34:24.0000] <hsivonen>
tantek: a truly proper fix is a completely different API, and we can mint a new API even if the first attempt isn't prefixed

[01:34:32.0000] <hsivonen>
tantek: the new API is IndexedDB

[01:34:43.0000] <hsivonen>
tantek: which for many localStorage uses is an overkill

[01:35:10.0000] <hsivonen>
tantek: so we really don't have an API with callbacks but otherwise localStorage-level simplicity

[01:35:20.0000] <tantek>
and is underkill on browsers that don't support IndexedDB

[01:35:32.0000] <rniwa>
wow! youtube.com just updated their UI.

[01:35:55.0000] <hsivonen>
tantek: the fix that would have retained the API structure of localStorage was too complex to be implemented

[01:36:15.0000] <hsivonen>
tantek: besides, localStorage already was the second iteration. after globalStorage

[01:36:23.0000] <hsivonen>
and globalStorage was unprefixed

[01:36:28.0000] <hsivonen>
and we still removed it

[01:36:35.0000] <tantek>
hsivonen - what I find odd is that cookie API has existed for years, and a simple attempt to expand it was apparently much harder than expected.

[01:36:47.0000] <tantek>
twice

[01:37:08.0000] <hsivonen>
sure, Hotmail broke, but if it had been mozGlobalStorage, Hotmail would have used it and gotten broken upon removal anyway

[01:37:26.0000] <hsivonen>
tantek: the cookie API has race conditions, too

[01:37:43.0000] <hsivonen>
tantek: but it turns out they are race conditions we can live with without the sky falling

[01:38:18.0000] <odinho>
storeGet = function(key, cb) { db.transaction('store').objectStore('store').get(key).onsuccess = function(e) { cb(e.target.result) }; }

[01:38:27.0000] <tantek>
the storage API debacle actually provides a strong case for versioned APIs (if not prefixed)

[01:38:33.0000] <hsivonen>
If I understand correctly, the cookie API has always had race conditions even in single-threaded browsers

[01:38:41.0000] <odinho>
Yeah, kinda verbose, -- but it's not many lines to shim a very basic easier API :P

[01:39:10.0000] <odinho>
(you would also need a similar Set function, which would be the same only using add instead of get, and you'd need an init function opening the db)

[01:39:17.0000] <hsivonen>
tantek: why? globalStorage, localStorage, Web SQL Database and IndexedDB don't have namespace collisions even without prefixes

[01:39:26.0000] <tantek>
local storage is particularly critical on mobile, and due to lack of IndexedDB on Webkit - telling webdevs to "just use IndexedDB" is impractical

[01:39:57.0000] <tantek>
so instead we wait for frameworks to abstract some *other* storage API on top of whatever each engine happens to support

[01:40:09.0000] <hsivonen>
tantek: why would it be more practical to tell them to use a bunch of different localStorages?

[01:40:26.0000] <tantek>
e.g. is there a IndexedDB polyfill that uses WebSQL underneath on Webkit?

[01:40:48.0000] <hsivonen>
IIRC there is

[01:40:52.0000] <hsivonen>
/me searches the Web

[01:41:10.0000] <tantek>
hsivonen - use bunch of different localStorages = can actually run on multiple browsers, rather than use just indexedDB = does not run on webkit = impractical

[01:41:15.0000] <tantek>
that seems like pretty obvious logic

[01:41:17.0000] <odinho>
It should do the same in Opera then hopefully.

[01:41:52.0000] <odinho>
Well, WebKit should fix their IDB. They have an implementation, although it's old and buggy.

[01:42:00.0000] <hsivonen>
tantek: depends on how different the localStorages are

[01:42:17.0000] <tantek>
indeed

[01:42:29.0000] <hsivonen>
tantek: fixing the problem localStorage has in a proper way would change it to a different callback-based API

[01:42:32.0000] <tantek>
would be hard to answer without actually implementing the aforementioned strategy

[01:42:39.0000] <hsivonen>
tantek: Google search says http://blog.nparashuram.com/2012/05/introducing-indexeddb-shim-over-websql.html

[01:42:46.0000] <hsivonen>
(I have no idea if it works)

[01:43:25.0000] <tantek>
ooh shims all the way down

[01:43:26.0000] <tantek>
"WebSql iteself has shims implemented and this should work well over those shims too."

[01:43:28.0000] <tantek>
from that post

[01:43:47.0000] <tantek>
nice find hsivonen

[01:43:53.0000] <tantek>
perhaps there is hope for IndexedDB after all

[01:43:57.0000] <hsivonen>
tantek: if different localStorages diverge enough, we might as well call one version Web SQL Database and another IndexedDB

[01:44:13.0000] <tantek>
right - hence "indeed" above

[01:44:35.0000] <hsivonen>
tantek: anyway, intuitively, it should be easier to emulate IndexedDB over Web SQL than vice versa

[01:44:46.0000] <tantek>
agreed

[01:45:07.0000] <tantek>
I think IndexedDB has a subset of functionality

[01:45:16.0000] <tantek>
plus WebSQL is a dead end draft

[01:45:28.0000] <tantek>
so we shouldn't be encouraging anyone to code to that API

[01:50:21.0000] <hsivonen>
well, if we were implementing localStorage.new and WebKit was implementing localStorage.old, your practicality argument would seem to involve encouraging people to write for localStorage.old for WebKit

[01:50:31.0000] <hsivonen>
seems like the same situation with Web SQL

[01:51:06.0000] <hsivonen>
except the two APIs are separate designs instead of iterations of one design family

[01:57:04.0000] <tantek>
right, presumably with iterations of the same API your business logic can still remain relatively similar. two completely different APIs usually means very different business logic / maintenance - except of course if you can polyfill one with the other as already discussed.

[02:00:10.0000] <hsivonen>
tantek: yes, except with localStorage the problem is so fundamental that an API fix would go from sync API to async API which virtually always involves significant code structure changes to the caller

[02:00:35.0000] <hsivonen>
anyway, I still think prefixing would not have made the localStorage situation better

[02:01:11.0000] <hsivonen>
so as far as historical datapoints go, I count localStorage as a Real Problem case against which prefixing would have been useless

[02:01:33.0000] <tantek>
sure, we could have always introduced asyncLocalStorage instead

[02:01:42.0000] <hsivonen>
indeed

[02:01:43.0000] <tantek>
and left the existing one in place

[02:02:04.0000] <tantek>
the so-called "new names are easy" argument per the IETF RFC that obsoleted x- prefixes

[02:02:14.0000] <tantek>
I'm sure that's made the rounds here right?

[02:02:36.0000] <hsivonen>
I've noticed the RFC

[02:02:43.0000] <tantek>
worth reading the whole thing

[02:03:22.0000] <hsivonen>
kudos to Peter St.Andre for getting it published at the IETF which is generally more change-averse than the W3C

[02:03:36.0000] <tantek>
indeed - it came as quite a surprise

[02:03:37.0000] <annevk>
hsivonen: did anything came out of https://twitter.com/BrendanEich/status/223802905596133376 ?

[02:03:41.0000] <tantek>
impressively so

[02:03:56.0000] <hsivonen>
annevk: I don't know.

[02:04:23.0000] <hsivonen>
annevk: a couple of B2G APIs have been unprefixed lately, IIRC, but I don't know if that should be credited to Brendan's influence

[02:04:38.0000] <tantek>
annevk - do you mean aliasing -moz-opacity, or unprefixing future JS APIs?

[02:05:57.0000] <annevk>
I was looking for some kind of statement

[02:07:27.0000] <hsivonen>
was there a statement about CSS unprefixing at Mozilla, either? I saw dbaron's policy proposal and then code action--but not a statement of the policy getting adopted

[02:07:49.0000] <hsivonen>
(hooray for code action!)

[02:08:07.0000] <tantek>
hsivonen - I believe dbaron's policy proposal has been brought to CSSWG where we can further argue it there

[02:08:34.0000] <tantek>
it's a bigger win if we can alter CSSWG policy than just Mozilla policy

[02:08:51.0000] <tantek>
that's our current path of "statement" / "policy" change

[02:08:58.0000] <annevk>
I wonder why file URLs have a different code path from other hierarchical URLs in Chrome/Gecko

[02:09:05.0000] <hsivonen>
tantek: yes, but I thought it was supposed to get adopted by each vendor (routing around the CSS WG)

[02:09:43.0000] <hsivonen>
"yes" as in bigger win *if* the CSSWG adopts it

[02:10:12.0000] <hsivonen>
which I, admittedly as a CSSWG outsider, see as a big "if"

[02:10:14.0000] <tantek>
hsivonen - well, we got adoption in the specific cases of TAA

[02:10:17.0000] <tantek>
in the CSSWG

[02:10:30.0000] <tantek>
hsivonen - that's fine (to see it as a big if)

[02:10:41.0000] <tantek>
that's dbaron and my job to make the big if actually succeed

[02:11:17.0000] <hsivonen>
ok

[02:11:41.0000] <tantek>
San Diego should be interesting

[02:11:59.0000] <hsivonen>
CSSWG meeting in San Diego before TPAC?

[02:12:04.0000] <tantek>
august

[02:12:12.0000] <hsivonen>
ok

[02:13:59.0000] <annevk>
contrast e.g. ws:test and file:test

[02:14:11.0000] <annevk>
and file://test/ and file:///test/ vs ws:///test/

[02:15:16.0000] <hsivonen>
who is responsible for cloning https://www.w3.org/Bugs/Public/show_bug.cgi?id=18367 and the like into the WHATWG component?

[02:15:44.0000] <annevk>
hsivonen: a script Hixie wrote

[02:15:59.0000] <hsivonen>
(I think cloning bugs is bad compared to having flags for tracking snapshots)

[02:16:31.0000] <hsivonen>
(Mozilla doesn't close bugs for ESR but tracks ESR status using flags that are secondary to the main bug FIXED status)

[02:18:12.0000] <hsivonen>
s/close/clone/

[02:18:34.0000] <annevk>
ESR?

[02:19:23.0000] <hsivonen>
annevk: Extended Support Release

[02:19:39.0000] <hsivonen>
annevk: the enterprise edition of Firefox

[02:20:29.0000] <odinho>
So like LTS only a NIH name? :D

[02:20:48.0000] <hsivonen>
odinho: yeah

[02:21:33.0000] <odinho>
Cool. Wonder what we should call it if we ever do something like that. Need to make some more acronyms!

[02:21:55.0000] <hsivonen>
Mozilla's "Extended" is less extended than Canonical's "Long" is long

[02:23:06.0000] <hsivonen>
"Extended" is roughly one fifth of "Long"

[02:23:10.0000] <odinho>
Yeah, 12.04 LTS will be there a looong time, -- I'm in the process of upgrading all servers to it :]

[02:23:27.0000] <odinho>
12.04 LTS is ELTS - Extra long term support :P

[02:32:04.0000] <AryehGregor>
hsivonen, in the HTML5/HTML-LS case, the specs are edited by different people who might conceivably decide on entirely different resolutions, so I don't think the Mozilla way would work well.

[02:33:39.0000] <tantek>
AryehGregor - we don't know who the HTML5 editor(s) will be so it's pointless to make such suppositions.

[02:33:49.0000] <tantek>
They might just take all upstream bugfixes for example

[02:34:05.0000] <tantek>
I'd prefer to give the new HTML5 editor(s) a chance to prove themselves.

[02:34:08.0000] <AryehGregor>
tantek, in that case, there's no point in splitting the bugs, no.

[02:34:19.0000] <AryehGregor>
But I don't think the Mozilla-style solution is *necessarily* workable.

[02:34:28.0000] <tantek>
there's always risks

[02:34:46.0000] <tantek>
if it weren't for people in general doing the right thing, open standards would fail, and open source would fail

[02:35:07.0000] <tantek>
so it's pointless to worry about worst cases of human behavior like that, or rather to do so and still try to participate.

[02:35:32.0000] <tantek>
(in open standards and open source)

[02:41:05.0000] <AryehGregor>
All I'm saying is that since the HTML5 editor will be perfectly entitled to resolve bugs differently from the HTML editor, we can't give them the same bugs in the same fashion as is done in Mozilla, where entirely different resolutions are not accepted for older versions.

[02:41:19.0000] <AryehGregor>
The workflow doesn't work.

[02:41:46.0000] <AryehGregor>
If the HTML5 editor commits to only backport HTML changes, then no, there's no reason to have separate bugs, but nothing guarantees that will be the case (unlike at Mozilla).

[02:44:07.0000] <Bonob0h>
salut, y a t il des francophones dans le coin ?

[02:46:08.0000] <hsivonen>
For expectations, it would help if the HTML WG chairs characterized the WHATCG spec as the trunk instead of characterizing it as one possible proposal for HTML.next.

[02:46:48.0000] <hsivonen>
AryehGregor: I wish the HTML5 editor ends up backporting fixes in which case bug cloning will have been unnecessary.

[02:46:59.0000] <AryehGregor>
hsivonen, yeah, that would be nice.

[02:47:30.0000] <hsivonen>
making completely different resolutions would be like the XHTML2 WG all over again

[02:48:01.0000] <tantek>
hsivonen - that's optimistic

[02:48:11.0000] <tantek>
xhtml2 wg was much worse than that

[02:48:27.0000] <Stevef_>
i also hope that implementation bugs are fixed as per HTML, and think that will occur. I don't think that all bugs should be accepted though

[02:49:07.0000] <AryehGregor>
Yeah, there should be some things that aren't backported to HTML5, just as with any kind of snapshot-thing.

[02:49:52.0000] <tantek>
I actually expect to see features *dropped* from HTML5 as it makes its way through CR and test show no one implements things, or not interoperably so.

[02:50:11.0000] <AryehGregor>
Yes, that's likely given W3C Process.

[02:50:29.0000] <AryehGregor>
In particular, if large bugs are found in a particular feature, it seems likely that the feature might be dropped instead of fixed.

[02:50:36.0000] <AryehGregor>
That's fine, it's a subset spec anyway.

[02:50:46.0000] <tantek>
indeed

[02:50:55.0000] <Stevef_>
the only time there would be disagreement on implementation stuff is impementers don't agree right?

[02:51:03.0000] <AryehGregor>
Well, you'd hope so.

[02:51:03.0000] <tantek>
it will be interesting to see how the whatwg spec evolves to incorporate (or not) such feature cutting

[02:51:14.0000] <AryehGregor>
There have been and are W3C specs that just ignore implementations entirely.

[02:51:19.0000] <tantek>
feature dumping that is

[02:51:34.0000] <tantek>
aryehgregor - unfortunately so

[02:51:34.0000] <AryehGregor>
I.e., they spec something that implementers have said they don't want to implement.

[02:51:38.0000] <AryehGregor>
Hopefully that won't happen again here.

[02:52:03.0000] <tantek>
AryehGregor - we could use more folks with that perspective in HTMLWG

[02:52:07.0000] <Stevef_>
tantek: if there is stuff that is not being implemented they shoudn't hang around in any specs should they?

[02:52:11.0000] <tantek>
and every other HTMLWG

[02:52:23.0000] <hsivonen>
dropping features is to be expected if the HTML WG adheres to the interop reqs of the Process like the CSSWG and doesn't get sucked into a political vortex where interop matters less

[02:52:26.0000] <AryehGregor>
Although it's true that W3C Process these days makes it impossible to get to REC if you don't have two implementations.

[02:52:31.0000] <tantek>
Stevef_ different specs may have different "experimentation" tolerances

[02:52:40.0000] <hsivonen>
(I sure hope it doesn't get sucked into that political vortex)

[02:52:45.0000] <tantek>
CR, especially upon exit has a very low tolerance for experimental features

[02:53:08.0000] <Stevef_>
tantek: right, but exit won't happen for years and years no?

[02:53:38.0000] <hsivonen>
(the sort of place where RDFa seems to be with headed or already is)

[02:53:53.0000] <Ms2ger>
AryehGregor, that might be true for the Process as written; I doubt it's true for the Process as actually enforces

[02:53:55.0000] <Ms2ger>
*enforced

[02:54:02.0000] <tantek>
stevef_ depends on how long the WG sets for the CR period

[02:54:11.0000] <tantek>
and whether the WG wants to bounce between LC and CR

[02:54:16.0000] <AryehGregor>
Ms2ger, if there are people in the WG who will call people on it . . .

[02:54:17.0000] <annevk>
Ms2ger: AryehGregor: I don't think that's actually in the Process or enforced

[02:54:20.0000] <hsivonen>
Ms2ger: so far the HTML WG Chairs have almost enforced the process too much...

[02:54:22.0000] <Stevef_>
and if no interop exists on features after years and years and no indication that it will from implementers then...

[02:54:23.0000] <annevk>
Ms2ger: AryehGregor: it's just good practice

[02:54:25.0000] <AryehGregor>
annevk, IIRC, it is in the process.

[02:54:26.0000] <AryehGregor>
/me looks

[02:54:43.0000] <Ms2ger>
(In general, not particularly for the HTMLWG)

[02:54:58.0000] <AryehGregor>
Oh, it's a "should", I guess.

[02:54:59.0000] <AryehGregor>
http://www.w3.org/2005/10/Process-20051014/tr.html#cfr

[02:55:06.0000] <AryehGregor>
"each feature of the technical report has been implemented. Preferably, the Working Group should be able to demonstrate two interoperable implementations of each feature. If the Director believes that immediate Advisory Committee review is critical to the success of a technical report, the Director may accept to Call for Review of a Proposed Recommendation even without adequate implementation experience"

[02:55:15.0000] <AryehGregor>
So it's not a hard rule.

[02:56:22.0000] <annevk>
and many on the W3C Team will tell you that two implementations of a feature means two implementations passing a testcase

[02:56:24.0000] <tantek>
no it's something that requires WG members to push for it

[02:56:55.0000] <tantek>
(regarding not a hard rule)

[02:57:00.0000] <annevk>
fortunately some groups have pursued higher standards

[02:57:27.0000] <tantek>
and other specs are fiction

[03:01:26.0000] <hsivonen>
Stevef_: do you happen to recall how (according to Hixie) the API mapping doc contradicts HTML-LS?

[03:02:38.0000] <hsivonen>
Stevef_: does your alt guidance doc have any WAI consensus behind it? (My vague recollection is that I've seen some misgiving from the WAI about publishing it; maybe I should search the archives.)

[03:03:15.0000] <Stevef_>
hsivonen: no that information was never provided, to my knowledge bugs filed against the spec would help resolve differences or at least provide clarification on why there are differnces

[03:03:51.0000] <Stevef_>
hsivonen: FYI the API guide is currently non normative

[03:04:22.0000] <hsivonen>
Stevef_: anyway, to your earlier question, http://www.paciellogroup.com/blog/2012/07/w3c-html-and-whatwg-html-split-affect-on-html-accessibility/ looks pretty accurate on the basis of facts but some of the facts taking without additional info are misleading on their own

[03:05:04.0000] <Stevef_>
hsivonen: the WAI have provided feedback on the alt doc and have no major issues with what it contains, but they are pushing to have it moved out of HTML WG into WAI

[03:05:27.0000] <hsivonen>
Stevef_: specifically, even though the API mapping guide is nominally published through the HTML WG, it is really developed by a separate small group of interested people and isn't something that the HTML WG really pays attention to

[03:05:46.0000] <Stevef_>
hsivonen: OK am happy to clarify any particular points on the post

[03:06:24.0000] <hsivonen>
Stevef_: so the fact that the API doc work is happening under the W3C umbrella doesn't really say much about W3C HTML5 vs. WHATWG HTML-LS.

[03:06:58.0000] <hsivonen>
Stevef_: any more than some other related doc work happening at the W3C

[03:07:21.0000] <Stevef_>
hsivonen: its in the HTML WG because that I think is the right place for it as its subject is HTML and its an open (in w3c terms) working group

[03:07:22.0000] <hsivonen>
Stevef_: the refusal to link to it from the WHATWG spec says *something* about *something* though

[03:07:38.0000] <hsivonen>
Stevef_: I'm not suggesting it's in a wrong place

[03:08:40.0000] <hsivonen>
Stevef_: I'm just saying that the API doc nominally being in the HTML WG is neither a merit nor a dismerit for W3C HTML5 and is neither proof nor non-proof of the general interest in the topic at the HTML WG

[03:09:19.0000] <Stevef_>
hsivonen: as i said to anne yesterday, there has gernerally been no input on the API guide from the WHATWG side, and only negative non actionable feedback from hixie in particular

[03:09:47.0000] <hsivonen>
Stevef_: FWIW, I think I've said occasional positive things about the API doc and haven't been negative about it, but then it would be correct not to count me as representing the WHATWG

[03:10:24.0000] <Stevef_>
hsivonen: meanwhile we have had direct input from browser acc engineers

[03:11:09.0000] <hsivonen>
Stevef_: the lack of WHATWG input is probably explained by the lack of detailed knowledge of browser accessibility API mapping implementations among the people you consider to be "the WHATWG side"

[03:11:36.0000] <Stevef_>
hsivonen: i am happy to get constructive feedback and bugs on the doc from anyone

[03:12:56.0000] <Bonob0h>
personne ne parle français ?

[03:13:18.0000] <Stevef_>
hsivonen: sure, but what little has been said of the doc has not been useful or constructive

[03:13:24.0000] <hsivonen>
Bonob0h: typiquement, on parle anglais ici

[03:14:23.0000] <Bonob0h>
donc si on ne parle pas anglais on a pas le droit a la parole ! tout le web doit parler anglais ! bravo !

[03:14:37.0000] <Bonob0h>
et bien sur aucun channel dans différentes langues !!!

[03:15:36.0000] <annevk>
 zou behoorlijk lastig worden als we allemaal in onze eigen taal gaan chatten

[03:15:47.0000] <hsivonen>
Bonob0h: personne n'a dit que vous n'avez pas le droit a la parole

[03:15:59.0000] <Bonob0h>
j'espère que la chine va dominer le monde ! et que bientot tous les anglophone devront parler et coder en mandarin

[03:16:01.0000] <AryehGregor>
צריכים לדבר בעברית!

[03:16:28.0000] <Stevef_>
hsivonen: and surprisingly (to me) the information in it is being used to guide implementations in a number of browsers especially for new features

[03:16:47.0000] <Ms2ger>
AryehGregor, hey, other scripts aren't fair ;)

[03:17:01.0000] <hsivonen>
Stevef_: as I said on Twitter (and, I belive, at TPAC two years ago), I think it's useful work

[03:17:25.0000] <AryehGregor>
Ms2ger, fine.

[03:17:32.0000] <AryehGregor>
Tzrichim ledaber be'ivrit!

[03:17:35.0000] <annevk>
Stevef_: if you got the impression from me that I did not think API mapping was useful, sorry about that; I do think it's useful

[03:17:50.0000] <annevk>
Stevef_: just tried to explain why Hixie might have problems with some of the content

[03:18:36.0000] <hsivonen>
Stevef_: anyway, getting feedback and implementations from browser accessibility engineers is a better way to proceed that fighting about the link with Hixie. :-/

[03:18:48.0000] <hsivonen>
s/that/than/

[03:18:56.0000] <Ms2ger>
<tabatkins> I [�] successfully blocked an attempted rename recently.

[03:19:12.0000] <Ms2ger>
Jury / WG was still out on that last time I checked

[03:19:22.0000] <annevk>
Stevef_: but the simplest way forward here seems to be to continue demonstrating you do a better job and then when Hixie gets to looking at that particular part of HTML again he'll adjust the spec accordingly

[03:19:50.0000] <hsivonen>
/me thinks angle direction changes are more confusing that renames

[03:19:55.0000] <hsivonen>
s/that/than/

[03:19:55.0000] <annevk>
Stevef_: not much more needed than demonstrating implementors are actually following a different specification

[03:20:06.0000] <hsivonen>
/me has trouble with "than" today

[03:20:30.0000] <AryehGregor>
Yeah, one thing you can say for the WHATWG -- if you have implementers on your side, you win no matter what.

[03:20:40.0000] <Stevef_>
hsivonen: annevk: we know its incomplete, there is a lot of stuff that still needs to be added, but for the quite numerous features that are not covered by the ARIA imeplementation guide it is the ONLY document that provides info on how browsers do or should implement HTML acc across paltforms and APIs

[03:21:38.0000] <AryehGregor>
Well, I didn't ever convince Hixie to align with ES for numeric parsing, I don't think.

[03:22:01.0000] <AryehGregor>
But we don't have real interop there right now anyway.

[03:22:11.0000] <Stevef_>
hsivonen: annevk:it is also helping to work out what new stuff needs to be aded to ARIA (1.1), such as roles for new HTML features that are currently not in any API

[03:22:59.0000] <Stevef_>
hsivoen:annevk: on link to it, sure I am like a dog with a bone sometimes, probably not my best feature

[03:23:19.0000] <hsivonen>
Stevef_: my point was that that's all good for the API mapping doc, but it's not a merit for W3C HTML5 relative to WHATWG HTML-LS (except for the link maybe)

[03:23:57.0000] <hsivonen>
Stevef_: so insinuating (on Twitter) accessibility merit of W3C HTML5 because of the API mapping doc is off the mark

[03:24:53.0000] <Stevef_>
hsivonen: my point was that there is more that goes on in the HTML WG than HTML5, so calls to shut it down are inflammatory and insulting to the other work going on

[03:25:28.0000] <hsivonen>
Stevef_: fair enough.

[03:26:24.0000] <hsivonen>
this may be insulting, too, but I didn't remember that the API mapping was an HTML WG deliverable

[03:26:51.0000] <hsivonen>
I do remember that RDFa in HTML nominally is, but it seems in practice a different group works on it

[03:27:40.0000] <hsivonen>
most of the ISSUE Process seems to be focused on parts branched/forked from WHATWG HTML-LS

[03:27:44.0000] <Stevef_>
hsivonen: and i am attempting to engage with WHATWG on the stuff I do, for example the recent draft using ARIA doc which I intend to be published as a WG deliverable (note) i sent an email to whatwg about

[03:28:19.0000] <Stevef_>
hsivonen: no offense taken, it is a deliverable, but not in the HTML5 timeframe

[03:29:49.0000] <hsivonen>
Stevef_: FWIW, I didn't retweet fantasai in agreement but in the sense "see this", since I think it's noteworthy how the sentiment is turning against trusting the W3C HTML WG to do a good job with HTML

[03:30:18.0000] <hsivonen>
after all, participating is unpleasant enough that people (me included) have been tuning out

[03:30:23.0000] <Ms2ger>
hsivonen, I thought the E in ESR was for Enterprise, not Extended?

[03:31:01.0000] <Stevef_>
hsivonen: all I can do is publicise work and call for feedback http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jun/0269.html

[03:31:11.0000] <hsivonen>
Ms2ger: it's "Extended" https://www.mozilla.org/en-US/firefox/organizations/all.html

[03:31:37.0000] <Ms2ger>
I stand corrected, thanks

[03:34:20.0000] <Stevef_>
hsivonen: re retweet OK. participation in WHATWG is not exactly seen as pleasant for many of my peers either, thus the lack of participation, despite that I have tried to bite the bullet and do it as I am interested in improving HTML as I see improvement ;-), and realize that working only in the HTML WG is not enough

[03:37:00.0000] <Ms2ger>
Stevef_, I guess most people on the WHATWG list don't have much experience with accessibility APIs and prefer to remain silent rather than to make uninformed comments

[03:37:08.0000] <hsivonen>
while I'd prefer the PP to change such that RECs wouldn't be special from the PP perspective, if we take the PP as a given, I think publishing a snapshot is a good thing from PP perspective

[03:37:17.0000] <hsivonen>
so I hope the HTML WG succeeds there

[03:37:44.0000] <Stevef_>
Ms2ger: OK

[03:37:56.0000] <hsivonen>
but I also hope various people who have a beef with Hixie don't take it as an opportunity to stick their stuff into what was supposed to be a snapshot

[03:38:55.0000] <hsivonen>
I also am quite skeptical about there being enough participation to get a snapshot done, because it's always cooler to work on new stuff than to polish old stuff for the PP's sake

[03:39:55.0000] <hsivonen>
(CSS 2.1 got done, but the CSS WG is exceptional at the W3C)

[03:39:57.0000] <Ms2ger>
hsivonen, or for any sake, really

[03:40:10.0000] <Ms2ger>
Also in implementations

[03:40:26.0000] <hsivonen>
Ms2ger: hey, I'm polishing readyState and about:blank!

[03:40:55.0000] <Ms2ger>
hsivonen, and I sure am glad you are :)

[03:41:33.0000] <Ms2ger>
Though I guess you may not be as glad about undertaking it :)

[03:41:34.0000] <hsivonen>
everlasting fame awaits among people who care about the readyState of XSLT output!

[03:45:03.0000] <Stevef_>
hsivonen: "but I also hope various people who have a beef with Hixie" I only have a persoanl beef in so much as he decides what goes into the WHATWG spec, so naturally I am going to work on the spec that I feel I have a better chance of affecting change, i don't always succeed, but i put a lot of effort, research and data collection into my efforts and think that when they are accepted they...

[03:45:05.0000] <Stevef_>
...have merit

[03:47:22.0000] <Stevef_>
hsivonen: in particular authro conformance advice and requirments that are not dependent on browser implementers, and also aligning such with implementation (or lack of) realities that have a negative effect upon users

[03:50:54.0000] <Stevef_>
hsivonenL for example with the title attribute stuff, IF browsers provided input device independent access to the content I would have no problem with whats in the HTML spec, but they don't and its been that way for a long time and there is no indication that there will be a shift in current browser behaviours in a timeframe that will make the conformance requirments and advice align with reality

[03:53:06.0000] <Stevef_>
hsivonen: having said that IE 10 displays title attribute as a tooltip on focus, which is a start, so if other browsers follow suit at some point the HTML spec may actually provide useful advice for developers that does not lead to inaccessible content

[03:53:30.0000] <hsivonen>
Stevef_: displaying on keyboard focus makes sense

[03:53:58.0000] <hsivonen>
Stevef_: IIRC, Firefox for Android shows it for links on tap&hold. (though I'd have to test to be sure)

[03:54:37.0000] <Stevef_>
hsivonen: yes, which is what I have argued and written about for almost a deacade :-)

[03:55:48.0000] <Stevef_>
hsivoenen: but it still leaves issues for any non focusable content, such as the advice in the HTML spec to use title for short footnotes...

[03:56:38.0000] <hsivonen>
I think I disagree with the short footnote advice

[03:57:42.0000] <Stevef_>
hsivonen: I will rejoice when the day comes that browsers provide access to title attribute content so I don't constantly have to tell clients that what they are doing doesn't work for users, but that day is a way off

[03:59:51.0000] <Stevef_>
hsivonen: which is why I have pushed for the HTML5 spec not to include such advice as for better or worse it is a source for HTML authoring rules

[04:01:04.0000] <Stevef_>
anyway sorry for getting on my horse and gabbling on, as annevk siad yesterday, it one of my 'pet' issues

[04:01:16.0000] <hsivonen>
FWIW, I want a Gecko feature that'd force-draw a focus outline when the page has made an element not change appearance on focus but would let page-provided focus indication work if the page has customized :focus in a visible way

[04:01:47.0000] <hsivonen>
boilerplate * { outline: 0; } is very annoying

[04:02:13.0000] <Stevef_>
hsivonen: sounds good

[04:05:17.0000] <hsivonen>
speaking of WG dysfunctions: https://twitter.com/tabatkins/status/227683159637106689

[04:10:10.0000] <Ms2ger>
I believe Bert thinks it's incompatible with pubrules

[04:33:57.0000] <annevk>
Ms2ger: do we want http://tools.ietf.org/html/bcp47 as BCP47 or something like LANGUAGE?

[04:34:05.0000] <annevk>
or maybe LANG

[04:34:41.0000] <Ms2ger>
Readable names are nice

[04:43:53.0000] <annevk>
done

[04:57:28.0000] <hsivonen>
hmm. the spec says meta Refresh navigates with replacement enabled.

[04:57:52.0000] <hsivonen>
Gecko source suggests it happens with replacement enabled only if the delay was <= 15 seconds

[04:57:55.0000] <hsivonen>
fun

[05:04:29.0000] <annevk>
spec should prolly allow for that

[05:08:35.0000] <annevk>
MikeSmith: I don't think we should have URL.username and URL.password

[05:08:52.0000] <MikeSmith>
annevk: eh?

[05:08:58.0000] <annevk>
MikeSmith: the whole userinfo production is considered dubious, not sure why we would expose them to APIs, abarth, arv?

[05:09:03.0000] <MikeSmith>
ah

[05:13:25.0000] <annevk>
MikeSmith: I don't expect to do much on URLs before September

[05:13:36.0000] <MikeSmith>
OK

[05:13:43.0000] <MikeSmith>
annevk: how about Notifications?

[05:13:59.0000] <annevk>
MikeSmith: I have a start of a JavaScript implementation of how I think the parser should be written, but it's not quite complete yet

[05:14:14.0000] <MikeSmith>
OK

[05:14:23.0000] <annevk>
maybe if I make some more time for it this week I'll try put it up on github

[05:14:39.0000] <annevk>
MikeSmith: Notifications I worked on today

[05:14:53.0000] <annevk>
MikeSmith: just posted some updates to the list

[05:15:01.0000] <MikeSmith>
ah OK

[05:15:39.0000] <MikeSmith>
/me finds exactly 500 unread messages in his mailbox

[05:16:30.0000] <MikeSmith>
ok the language attribute

[05:16:49.0000] <annevk>
yeah now they just have lang/dir

[05:16:54.0000] <MikeSmith>
and responded to Jonas

[05:17:14.0000] <MikeSmith>
so we can publish an LC draft soonish?

[05:17:44.0000] <annevk>
I think so

[05:17:52.0000] <annevk>
lets do next tuesday unless we hear anything?

[05:18:15.0000] <annevk>
after that I'm going to Romania

[05:52:36.0000] <MikeSmith>
OK let's aim for Tuesday

[05:52:48.0000] <MikeSmith>
annevk: ↑

[05:53:24.0000] <MikeSmith>
annevk: btw we have to change your affiliation

[05:53:35.0000] <MikeSmith>
can we do it this week?

[06:04:04.0000] <hsivonen>
hmm. I predict adding <picture> to a snapshot spec after the feature freeze of the snapshow won't go well

[06:10:04.0000] <annevk>
MikeSmith: August 1

[06:10:13.0000] <MikeSmith>
annevk: ah oK

[06:10:14.0000] <annevk>
MikeSmith: I'm still employed

[06:10:19.0000] <MikeSmith>
oh right

[06:10:20.0000] <MikeSmith>
OK

[06:38:28.0000] <hsivonen>
Does Jellybean still come with the old WebKit/Android port for apps that embed WebKit using the system APIs?

[06:38:38.0000] <hsivonen>
has that WebKit port improved since ICS?

[06:39:17.0000] <hsivonen>
that is, Dolphin on JB still uses the old Android WebKit--not the same WebKit and Chrome, right?

[06:54:25.0000] <odinho>
annevk: Still paying for that small ad next to your name! ;D

[06:55:13.0000] <annevk>
odinho: heh

[07:53:36.0000] <AryehGregor>
/me wonders if it's feasible to redefine NULL to be nullptr too . . . probably not

[07:53:50.0000] <odinho>
Hmm.

[07:53:56.0000] <AryehGregor>
Oh, wrong channel.

[07:53:56.0000] <AryehGregor>
Feh.

[07:54:01.0000] <odinho>
Thought so ;D

[07:54:03.0000] <AryehGregor>
/me always makes that mistake

[07:54:10.0000] <odinho>
You do, you do.

[07:54:29.0000] <odinho>
Just mozillaians though.

[07:54:37.0000] <odinho>
Don't see much webkit or opera people doing that...

[07:55:01.0000] <Ms2ger>
/me would be interested to see Opera people do that

[07:59:16.0000] <odinho>
Ms2ger: Yeah, all that secretsauce infoz! :P

[08:00:02.0000] <Ms2ger>
I should get myself hired by Opera for a while, just to stick my nose in all your secrets :)

[08:00:54.0000] <odinho>
Ms2ger: Desk next to me is free, since annevk was last visiting Oslo ;-)

[08:02:06.0000] <Ms2ger>
I guess I'd need to hurry if I'd wanted to get in impersonating annevk

[08:05:44.0000] <odinho>
Ms2ger: I'll help you :D

[08:07:10.0000] <annevk>
mahahaha, you want to borrow http://annevankesteren.nl/dump/opera-keycard.jpg ? although I guess that won't get you any far in the new building

[08:32:05.0000] <odinho>
annevk: Oh man that picture

[09:44:49.0000] <Hixie>
hober: fwiw, the html spec already says that if there's no control, the region represents something that isn't interactive

[09:56:42.0000] <hober>
Hixie: i'll make the issue-201 proposal doesn't add redundant normative text then :)

[09:57:50.0000] <Hixie>
as usual if there's anything you think needs adding, please post about it on the whatwg list

[09:58:01.0000] <Hixie>
and if it's not something you think needs adding, why are you adding it :-P

[09:58:26.0000] <hsivonen>
what's the deal with asking the responsive imagea cg to bring a proposal to the html wg after the feature freeze?

[09:58:49.0000] <Ms2ger>
Maybe they'll be asked to create a HTML6 module?

[10:00:31.0000] <hober>
Hixie: indeed. :) thanks for removeHitRegion; I get to make this proposal smaller as a result.

[10:00:58.0000] <hober>
hsivonen: my understanding is that they intend to work on "html.next" in parallel to the html5 stabilization effort.

[10:01:22.0000] <hsivonen>
who is "they"?

[10:05:50.0000] <hober>
hsivonen: the html wg

[10:06:33.0000] <Ms2ger>
/me wonders what will be left in the spec after the stabilization effort

[10:06:49.0000] <Ms2ger>
table.rows?

[10:07:27.0000] <Ms2ger>
/me doesn't think so, after writing http://w3c-test.org/html/tests/submission/Ms2ger/the-elements-of-html/tabular-data/the-table-element/table-rows.html

[10:39:07.0000] <Hixie>
ok i guess today's task will be to look at the web intents stuff, register*Handler stuff, and the proposal from mozilla

[10:39:43.0000] <Hixie>
and see what I can do that merges the three together in a backwards-compatible simple API and/or declarative markup that satisfies all the relevant valid use cases

[10:58:08.0000] <divya>
anybody knows what is the short url for whatwg multipage html spec?

[10:58:12.0000] <divya>
annevk: ^

[10:58:27.0000] <divya>
(i know anne tweeted it or mentioned it somewhere)

[10:59:21.0000] <beverloo>
http://whatwg.org/html ?

[10:59:32.0000] <divya>
omg thnx beverloo !!!!

[10:59:39.0000] <beverloo>
http://whatwg.org/C works too

[10:59:42.0000] <beverloo>
but I prefer "html" :p

[10:59:45.0000] <divya>
ahaha :)

[10:59:48.0000] <divya>
yeah

[10:59:51.0000] <divya>
html is less cryptic

[10:59:55.0000] <divya>
what is the c for anyway?

[11:00:40.0000] <beverloo>
no idea

[11:01:49.0000] <Ms2ger>
Complete

[11:03:00.0000] <jgraham>
Though as a memonic seCret is better :p

[11:03:02.0000] <zewt>
crash your browser

[11:03:29.0000] <Ms2ger>
Hey, one of them is multipage

[11:03:46.0000] <annevk>
divya: lowercase c is single-page

[11:06:42.0000] <tantek>
Hixie, btw I've been checking to see what other vCard/hCard derivatives have done with "country" vs. "country-name" and it's so far been a fairly even split, with a slight leaning toward country-name.

[11:06:58.0000] <tantek>
e.g. PoCo used "country", but OGP used "country-name"

[11:08:30.0000] <tantek>
it does seem to be a common simplification to drop the "-name" part though. It's unlikely to result in an incompatibility. So I'm on the fence about it.

[11:09:29.0000] <tantek>
from a vocabulary design standpoint, if we adopt the principle of dropping redundant suffixes from terms similar to dropping plural suffixes, then we can proceed consistently in the instance of "country" vs. "country-name" and any future such occurences.

[11:12:22.0000] <TabAtkins>
Should events that fire at the document bubble or cancel?

[11:13:12.0000] <TabAtkins>
That's a yay/nay for each option, not a choice between two options.

[11:19:25.0000] <Ms2ger>
smaug____, ^

[11:19:49.0000] <smaug____>
/me doesn't understand the question

[11:19:59.0000] <smaug____>
it depends on the case

[11:20:07.0000] <smaug____>
some event bubble to the window

[11:20:09.0000] <smaug____>
some don't

[11:28:09.0000] <odinho>
smaug____: partypooper breaking the rules :P

[11:28:18.0000] <odinho>
Ohwell, I should really go home.

[11:28:32.0000] <Ms2ger>
Naah

[11:28:39.0000] <Ms2ger>
How do I get into the office then?

[11:28:46.0000] <odinho>
Ms2ger: Come tomorrow.

[11:28:50.0000] <Ms2ger>
Alright

[11:31:13.0000] <tantek>
ok I've tried to capture both the potential vocabulary design principle of dropping redundant/no-value suffixes and the specific instance of how that applies to country vs. country-name here: http://microformats.org/wiki/naming-principles#Under_Consideration

[11:31:26.0000] <tantek>
feel free to add more data/examples if you know of any

[11:36:40.0000] <annevk>
divya: you can add to your CSS spec stuff that these are the relatively simple specs ;)

[11:36:53.0000] <divya>
annevk: wat

[11:37:06.0000] <divya>
annevk: i mean why use /c or /C what is 'c'

[11:37:16.0000] <divya>
annevk: what is this about css spec stuff :/

[11:37:20.0000] <hober>
short for "complete"

[11:37:26.0000] <divya>
hober: AHA.

[11:38:06.0000] <annevk>
divya: the ones we completed in a decade

[11:38:11.0000] <annevk>
or a little over a decade

[11:38:13.0000] <divya>
annevk: hahahah :|||

[11:38:16.0000] <divya>
yes

[11:38:20.0000] <divya>
i know right?!

[11:38:23.0000] <divya>
it is hilarious.

[11:38:50.0000] <divya>
i think it shows how irrelevant the time taken is i suppose.

[11:41:55.0000] <tantek>
ok, it's been pointed out on #microformats that 'country' often implies ISO country code, especially in forms (which is the broader context of this discussion - the autocomplete stuff).

[11:42:08.0000] <tantek>
thus I'm now leaning toward leaving 'country-name' as-is and not messing with it.

[11:42:30.0000] <tantek>
clarifies '-name' vs. a potential ISO country '-code' property

[11:42:37.0000] <tantek>
who knows, in forms, we could have both instances

[11:50:58.0000] <Hixie>
divya: it used to be that /html went to the WHATWG "HTML5" spec and /c went to the WHATWG "Web Applications 1.0" spec, which was called complete.html and contained everything including all the HTML stuff

[11:51:13.0000] <Hixie>
divya: but at some point we dropped them all and just went back to one spec, so now they all point to the same spec

[11:51:20.0000] <divya>
Hixie: thanks!

[12:00:57.0000] <TabAtkins>
smaug____: Is there a consistent guideline for whether it should bubble to window or not?

[12:01:13.0000] <TabAtkins>
Context is a FontFaceLoaded event, which fires when a @font-face finishes loading.

[12:01:45.0000] <TabAtkins>
(And gets fired at the document, since you can't access the @font-face rules from cross-origin stylesheets.)

[12:03:21.0000] <smaug____>
there is no guideline

[12:04:04.0000] <TabAtkins>
Awesome.

[12:37:34.0000] <Hixie>
TabAtkins: generally speaking, don't bubble things targetted at a node that isn't an Element, and make things cancelable only if they have a default action you want to be able to cancel

[12:41:49.0000] <Hixie>
/me mumbles something about mozilla inventing apis in their mailing lists just like everyone else

[12:42:19.0000] <Hixie>
(https://groups.google.com/forum/?fromgroups#!topic/mozilla.dev.webapi/o8bkwx0EtmM which i found linked to from https://wiki.mozilla.org/WebAPI/WebActivities which was linked to from a post to public-device-apis and public-web-intents)

[12:42:32.0000] <Ms2ger>
Yeah, the WebAPI guys are jerks like that

[12:45:17.0000] <tantek>
I just try to keep putting links to relevant stuff on findable (via web search) wiki pages - re: WebAPIs

[12:46:00.0000] <tantek>
Hixie - many folks (outside of Google) have given up on google web intents for a number of reasons (not going to bother going into details which we could spend hours on)

[12:46:03.0000] <Hixie>
that's a start, but i don't think we should be proposing apis by hoping people will stumble into them ;-)

[12:46:32.0000] <Hixie>
yeah i'm trying to work out how to merge all the various proposals here into one coherent whole

[12:46:44.0000] <tantek>
Hixie, what do you do when the WGs where you "should" propose them are greatly outnumbered by folks with very different priorities and timelines?

[12:46:48.0000] <tantek>
DAP etc.

[12:46:54.0000] <Hixie>
(hopefully the problems you reference are documented somewhere)

[12:47:04.0000] <tantek>
Good luck - the intents stuff is an every growing blob of mess.

[12:47:05.0000] <Hixie>
tantek: well, DAP imho is never a place to propose anything

[12:47:28.0000] <tantek>
well DAP is where WebAPIs were supposed to be proposed (in W3C)

[12:47:39.0000] <Hixie>
nah, public-webapps is the right place for web apis

[12:47:44.0000] <tantek>
so instead, you can find all the real-world ship-driven work here: https://wiki.mozilla.org/WebAPI/

[12:47:51.0000] <Ms2ger>
And the HTML WG was where HTML was supposed to be worked on

[12:48:21.0000] <tantek>
maybe I should actually officially join public-webapps instead of merely being an observer then

[12:48:25.0000] <Hixie>
public-webapps for apis, whatwg for apis and markup, www-svg and www-math for svg and mathml, www-style for rendering, and that should be it, imho

[12:48:44.0000] <Hixie>
audio, webrtc, dap, intents, perf, etc, all shouldn't exist imho

[12:48:46.0000] <tantek>
that's a decent summary Hixie, have you put that on the WHATWG wiki somewhere?

[12:48:54.0000] <tantek>
e.g. in the "where do I propose something" FAQ

[12:48:54.0000] <Hixie>
i have not

[12:49:02.0000] <tantek>
seems to come up often enough

[12:49:06.0000] <Hixie>
the whatwg wiki says "whatwg" as "where do i propose something"

[12:49:13.0000] <Ms2ger>
"Create a new CG"

[12:49:16.0000] <Hixie>
which seems like a good first approximation to me :-)

[12:49:25.0000] <tantek>
so should I send ContactsAPI to WHATWG?

[12:49:42.0000] <tantek>
https://wiki.mozilla.org/WebAPI/ContactsAPI is another thing I edit.

[12:50:15.0000] <Hixie>
wtf is contacts api

[12:50:38.0000] <Hixie>
that should just be a js library, why would the browser support that

[12:53:54.0000] <gsnedders>
Hixie: Because many apps on phones want access to contacts, and having apps work cross-platform is desirable?

[12:55:12.0000] <Ms2ger>
Sounds like a library would work much better than an API per vendor, then

[13:01:00.0000] <tantek>
because it requires access to device specific storage of contacts

[13:01:15.0000] <tantek>
so it has to go in the platform since a library can't access that

[13:01:24.0000] <tantek>
you could write a library on top of ContactsAPI if you want

[13:02:25.0000] <tantek>
sorry Hixie - didn't mean to distract you - was just providing another example of a WebAPI - there's plenty more.

[13:02:31.0000] <tantek>
/me realizes previous link went to a blank page

[13:03:04.0000] <tantek>
ok, fixed with a redirect for posterity

[13:03:12.0000] <tantek>
plenty more: https://wiki.mozilla.org/WebAPI#APIs

[13:41:49.0000] <annevk>
TabAtkins: FontFaceLoaded event sounds wrong, events should be lowercase

[13:42:10.0000] <annevk>
TabAtkins: fontfaceload seems more in line with what we have thus far

[13:42:31.0000] <TabAtkins>
That's the name of the interface.  That should be uppercase, right?

[13:42:41.0000] <annevk>
TabAtkins: name of the interface should end in Event

[13:43:32.0000] <TabAtkins>
Sorry, it does end in Event in my actual proposal.

[13:44:06.0000] <annevk>
but that should prolly be FontFaceEvent and the name of the event fontfaceload or some such

[13:44:19.0000] <annevk>
depending on what we want

[13:44:35.0000] <annevk>
and some API to add an @font-face to some style sheet

[13:58:20.0000] <annevk>
tantek: does proposing mean dropping a link or working with everyone to turn it into something?

[13:58:53.0000] <tantek>
annevk - I suppose it's a spectrum.

[13:58:59.0000] <tantek>
the more open the better

[13:59:01.0000] <annevk>
tantek: either way is fine as far as emailing to whatwg⊙wo goes, just curious

[13:59:51.0000] <tantek>
I just find email iteration tedious/inefficient, so prefer to iterate with wiki + irc links/chats until things seem fairly solid, then I'm ok with dropping a link to email just for old school's sake.

[13:59:53.0000] <annevk>
tantek: but it seems those APIs are primarily driven by a group at Mozilla that is not too communicative so setting expectations in such an email might be nice

[14:00:26.0000] <tantek>
Jonas has been pretty communicative in WebApps

[14:00:35.0000] <annevk>
tantek: well until this IRC channel has 1500 members you're not going to reach everyone involved in the WHATWG

[14:00:37.0000] <tantek>
and he's driving most of WebAPI

[14:00:42.0000] <Ms2ger>
Oh?

[14:00:57.0000] <Ms2ger>
I haven't seen much about this webapi stuff in webapps

[14:01:42.0000] <tantek>
annevk - in practice 1500 people don't contribute to specs.

[14:02:40.0000] <annevk>
tantek: that does not mean they should be ignored until things are "solid"

[14:03:03.0000] <tantek>
or maybe they shouldn't be spammed until things are more mature

[14:03:24.0000] <Ms2ger>
They subscribed for a reason

[14:03:36.0000] <annevk>
tantek: you mean when they can no longer influence direction?

[14:03:41.0000] <tantek>
anyway, see above, I find email iteration tedious/inefficient thus it's a notification system of last resort, in order to stay more productive.

[14:03:53.0000] <tantek>
ms2ger - who knows why people subscribe to mailing lists

[14:04:18.0000] <Ms2ger>
Presumably because they want to receive emails

[14:04:23.0000] <Ms2ger>
That's why I do it, at least

[14:04:25.0000] <tantek>
hah!

[14:06:13.0000] <tantek>
annevk - if folks want to only be on email, I'm fine treating that as "support forum" level discussion. plenty of ways to influence direction, including writing blog posts, editing wiki pages, contributing to irc discussions etc.

[14:06:51.0000] <tantek>
email lists generally devolve into support forum level thrashing - some may wish to receive such emails, good for them. I find them unproductive.

[14:07:36.0000] <tantek>
the lists/emails that is - not the people

[14:08:44.0000] <zewt>
no, mailing lists are the most effective communications medium we have (irc is better for quick back and forth, but less inclusive)

[14:10:21.0000] <Ms2ger>
I find it unfortunate that things are not discussed in a multi-vendor forum because you find that unproductive

[14:10:21.0000] <annevk>
pretty sure mailing list email is the one way some people keep track of standards work

[14:10:21.0000] <annevk>
not keeping them in the loop because you prefer some other medium seems silly

[14:10:22.0000] <annevk>
I did not suggest debating anything via email, just giving people a heads up and a way to participate

[14:10:22.0000] <annevk>
if you think that should be IRC/Wiki for now, that's fine

[14:10:22.0000] <annevk>
but if you don't even clue them in on where to start, they'll never get there

[14:10:26.0000] <tantek>
I'm aware of your opinion zewt, and we've already chosen to disagree. 

[14:10:40.0000] <zewt>
(you're being rude; this discussion is not only for you)

[14:11:25.0000] <tantek>
No I'm being terse. There's a difference. And you've expressed these opinions before in IRC and we've disagreed before. No need to repeat ourselves.

[14:11:52.0000] <zewt>
nope; if you're going to repeat your position, then I'm going to repeat mine

[14:12:50.0000] <tantek>
that's an odd approach to productivity. ;)

[14:13:41.0000] <zewt>
hardly; you're presenting your position (presumably) because you believe there are people reading who are unaware of it, and if you're presenting your rather odd view to other people, then clearly other positions should be restated as well

[14:13:42.0000] <tantek>
annevk - since Jonas is leading up the WebAPI effort and in WebApps, I'll leave it to him to decide how/when to notify which groups of what.

[14:14:10.0000] <zewt>
(on the other hand, spending this much of your--and my--time telling me not to state my opinion is a rather odd thing to do)

[14:14:25.0000] <tantek>
zewt, nah, annevk and I were simply having a discussion. I don't think your meta-arguments apply. sorry.

[14:14:45.0000] <annevk>
tantek: he was not directing it to you in particular though and I had not seen zewt's opinion before

[14:14:45.0000] <annevk>
(or more likely, I forgot, I try to keep up with the logs :))

[14:15:20.0000] <zewt>
(I'm not discussing whether I should state my opinion on IRC any more because it's a waste of my time; I'll continue to state my opinion when I wish, thanks)

[14:15:23.0000] <tantek>
feel free state your opinion as much as you like. I'm just choosing to simply disagree and not both rehashing the arguments.

[14:15:33.0000] <annevk>
tantek: anyway, WHATWG and W3C's main medium of communication is the mailing lists, if you avoid them prepare for some trouble

[14:15:43.0000] <TabAtkins>
annevk: I have no idea how to design events well.  Ms2ger just suggested some improvements to the design as well.

[14:15:58.0000] <tantek>
annevk - we're slowly getting W3C to use wikis more

[14:16:03.0000] <tantek>
it's a transition, it'll happen.

[14:16:21.0000] <annevk>
still waiting for RDF

[14:16:33.0000] <tantek>
is that like waiting to exhale?

[14:17:00.0000] <annevk>
heh

[14:17:23.0000] <tantek>
annevk - yes, you can find a bunch in the irc archives by searching for "email" and "support forum"

[14:17:24.0000] <Hixie>
gsnedders, tantek: seems pretty clear to me that "contacts" as a concept tied to a device is yesterday's tech

[14:17:46.0000] <tantek>
Hixie, yesterday's tech is today's/tomorrow's tech in parts of the world

[14:18:13.0000] <tantek>
plus, as long as all the contact sync services keep screwing things up (whether iCloud, or Google contacts), contacts tied to a device will remain relevant

[14:18:22.0000] <tantek>
since no one has made anything better actually work reliably

[14:18:33.0000] <annevk>
the WHATWG Wiki is frequently used too, but not without mailing list discussion to make the rest of the group aware of it

[14:18:37.0000] <Hixie>
beg to differ but ok :-)

[14:19:01.0000] <tantek>
Hixie, which perfect contact syncing service do you use?

[14:19:02.0000] <annevk>
contacts tied to a device?

[14:19:05.0000] <annevk>
who would want that?

[14:19:14.0000] <tantek>
people that buy the phones apparently

[14:19:18.0000] <tantek>
and Macs

[14:19:20.0000] <Hixie>
tantek: i wouldn't use syncing. i would just use cloud contacts only.

[14:19:21.0000] <tantek>
and PCs

[14:19:42.0000] <tantek>
Hixie - which "cloud" contacts then? and do you not do anything with contacts offline?

[14:19:44.0000] <Hixie>
tantek: my approach to phones is the same as my approach to laptops: the hardware is just a dumb terminal

[14:19:57.0000] <annevk>
tantek: not sure if that means they really want it

[14:19:59.0000] <tantek>
right, the use-case is offline.

[14:20:10.0000] <tantek>
we have whole APIs for that last time I checked

[14:20:16.0000] <annevk>
tantek: cause I'm pretty sure that if the hardware burns in a fire they'd still like their contacts on the next phone they buy

[14:20:17.0000] <Hixie>
tantek: my contacts live in g+, personally, but google alone has at least three different cloud contacts systems all of which i find work fine :-)

[14:20:20.0000] <tantek>
nevermind contacts in particular

[14:20:29.0000] <Hixie>
offline is for when i sleep

[14:20:42.0000] <tantek>
Hixie - I know highly technical google employees who have had problems with Google's various cloud contacts systems losing data.

[14:20:48.0000] <annevk>
Hixie: no sleep monitor? :p

[14:20:52.0000] <tantek>
Same with Apple employees and iCloud etc.

[14:21:00.0000] <tantek>
they all suck

[14:21:02.0000] <Hixie>
i know highly technical people who have had problems with their devices losing data, what's your point? :-P

[14:21:25.0000] <tantek>
less so

[14:21:31.0000] <annevk>
or whose phone gets water damage :p

[14:21:47.0000] <tantek>
the people I'm talking about found their contacts became much more randomly lossy when using syncing/cloud services

[14:21:59.0000] <Hixie>
this is all anecdotal and as such pretty worthless

[14:22:22.0000] <annevk>
tantek: if you don't configure your phone correctly that can happen

[14:22:40.0000] <annevk>
tantek: like if creating new contacts still creates them on the device, as is the silly default on iOS

[14:22:52.0000] <tantek>
well it's anecdotal enough for myself and many others to believe that we're not quite at that bright and shiny cloud future yet, and who knows if we will ever be

[14:23:05.0000] <Ms2ger>
annevk, sometimes water damage caused by a competitor pushing them into a lake? :)

[14:23:06.0000] <Hixie>
ditto but vice versa :-)

[14:23:24.0000] <tantek>
right, blame the user instead of the defaults. that always works. :P

[14:23:54.0000] <annevk>
I blamed the defaults?

[14:24:07.0000] <annevk>
oh well, all short term issues

[14:25:07.0000] <annevk>
you don't want the situation where you create a contact on your phone and it's not available on your laptop the same minute

[14:25:27.0000] <Hixie>
hear hear

[14:25:28.0000] <annevk>
or where you drop your phone in the water and you lose 200 contacts

[14:25:51.0000] <annevk>
that's last decade problems

[14:25:53.0000] <Hixie>
or where you add a contact to your tablet and it's not on your phone an hour later

[14:26:07.0000] <tantek>
people still have these problems yes

[14:26:21.0000] <Hixie>
i haven't had those problems in years

[14:26:22.0000] <tantek>
especially people with lots of (frequently changing) contacts

[14:26:31.0000] <tantek>
and who add/delete them from different devices etc.

[14:26:36.0000] <tantek>
Hixie - you just said you don't use sync

[14:26:37.0000] <tantek>
s

[14:26:38.0000] <tantek>
o

[14:26:40.0000] <Hixie>
right

[14:26:42.0000] <Hixie>
i use cloud

[14:26:44.0000] <Hixie>
sync is dumb

[14:26:45.0000] <tantek>
naturally, you wouldn't

[14:26:53.0000] <tantek>
unless you want offline

[14:26:54.0000] <Hixie>
sync is a bandaid

[14:27:08.0000] <Hixie>
offline works fine too, just cache it

[14:27:48.0000] <tantek>
things seem to be moving more in the sync direction, e.g. dvcs, hg, github are more of a "sync" approach than central cloud server approach (over cvs, svn etc.)

[14:27:58.0000] <tantek>
offline doesn't work when you add/delete contacts while offline

[14:28:02.0000] <tantek>
because then you need to...

[14:28:03.0000] <tantek>
s

[14:28:04.0000] <tantek>
y

[14:28:04.0000] <tantek>
n

[14:28:05.0000] <tantek>
c

[14:28:34.0000] <Hixie>
i'm going to go back to working out how to add intents to the spec, as this conversation is going nowhere :-)

[14:29:08.0000] <tantek>
no, we're not going to refigure out syncing, that's for sure. but at least it's a much more developed problem space than intents. good luck.

[14:35:30.0000] <TabAtkins>
I love how dramatic tantek's weird letter-per-line problem is sometimes.

[14:35:59.0000] <dglazkov>
d

[14:36:00.0000] <dglazkov>
r

[14:36:01.0000] <dglazkov>
a

[14:36:02.0000] <dglazkov>
m

[14:36:03.0000] <dglazkov>
a

[14:36:53.0000] <annevk>
TabAtkins: what's wrong with using null?

[14:37:31.0000] <annevk>
TabAtkins: also, the dictionary needs to inherit from EventInit

[14:37:34.0000] <TabAtkins>
annevk: I dunno.  It's weird?

[14:37:47.0000] <annevk>
TabAtkins: it would be pretty consistent with most other APIs

[14:38:08.0000] <annevk>
TabAtkins: although maybe just using the empty string here is enough

[14:38:21.0000] <tantek>
TabAtkins LOL

[14:38:23.0000] <Ms2ger>
annevk, I said that :)

[14:38:51.0000] <TabAtkins>
Do DOM APIs just use null because IDL didn't allow them to use undefined?

[14:39:02.0000] <TabAtkins>
JS APIs, with few exceptions, use undefined and null properly for their separate things.

[14:39:30.0000] <Ms2ger>
What's undefined for here?

[14:40:08.0000] <TabAtkins>
A property that doesn't exist.

[14:40:31.0000] <TabAtkins>
Unfortunately, you can't *actually* make attributes exist only conditionally, but having them return undefined is nearly functionally identical.

[14:41:21.0000] <Ms2ger>
Oh, if the property isn't present in the @font-face rule?

[14:41:27.0000] <TabAtkins>
Yeah, exactly.

[14:41:35.0000] <Ms2ger>
null sounds good

[14:42:34.0000] <annevk>
nn

[14:42:37.0000] <TabAtkins>
Why do you think so?

[14:43:02.0000] <TabAtkins>
Ms2ger: annevk: Can you put this advice and requirements for designing new Event subtypes into the dom core spec?

[14:43:54.0000] <hober>
which one is better / worth posting? http://www.quickmeme.com/meme/3q7quj/ http://www.quickmeme.com/meme/3q7qvo/

[14:44:10.0000] <Ms2ger>
http://wiki.whatwg.org/wiki/Howto_spec#Events_implementing_their_own_interface ?

[14:44:46.0000] <TabAtkins>
hober: Second, but s/it's/i call it/

[14:46:20.0000] <TabAtkins>
Ms2ger: Ah, didn't realize that existed.  That doesn't seem to contain all of the advice you've given.  anne said that my init dict should inherit from EventInit as well.

[14:46:39.0000] <Ms2ger>
I said that too :)

[14:46:53.0000] <Ms2ger>
But it's a wiki, feel free to add that :)

[14:49:52.0000] <TabAtkins>
I'd have to log into the whatwg wiki, then.

[14:50:49.0000] <TabAtkins>
So, the policy is to define the interface for a group of events, and then define specific events by different 'type' values?

[14:51:21.0000] <Hixie>
define an interface (or preferred, use an existing interface if there is one) for each event that needs a particular set of data sent with the event

[14:51:27.0000] <Hixie>
most events don't need any data and can just use Event

[14:51:41.0000] <Hixie>
(the HTML spec has many examples if you want to crib from that)

[14:52:04.0000] <TabAtkins>
The html spec no longer has the event index.

[14:52:35.0000] <Hixie>
search for [Init : Event]

[14:52:43.0000] <Hixie>
or [eventInit]

[14:52:45.0000] <Ms2ger>
Interface / type are orthogonal, really

[14:53:02.0000] <Hixie>
btw you can make a property not existing in WebIDL, just define it as a named property

[14:53:24.0000] <Hixie>
and then say when it's in the list of supported property names or not

[14:53:28.0000] <Hixie>
pretty unusual and ugly though

[14:54:00.0000] <TabAtkins>
heycam just now said that's okay. ^_^

[14:54:27.0000] <Hixie>
generally speaking you want the property to be present so you can set it

[14:54:28.0000] <Ms2ger>
Matches how those are usually represented in CSSOM

[14:54:42.0000] <Hixie>
a property being present doesn't mean the underlying field is present

[14:55:08.0000] <Hixie>
it's pretty standard in apis to expose all the possible fields and have a value that means "no value", namely, null

[14:55:25.0000] <Hixie>
"undefined" is more of a JS thing meaning "this property doesn't exist"

[14:55:40.0000] <Hixie>
really only applicable where other languages would refuse to compile :-)

[14:56:55.0000] <zewt>
well, i like how Lua does it; there's no distinction between a nil item in a dictionary and not existing

[14:57:15.0000] <zewt>
setting a value to nil is how you remove it from a dictionary

[15:02:55.0000] <TabAtkins>
Ms2ger: Why, in the how2spec example, does "type" show up both as the first argument to the constructor, and in the init dictionary?

[15:03:18.0000] <Ms2ger>
Wait, type in the dictionary?

[15:03:26.0000] <Ms2ger>
Sounds like a bug

[15:05:19.0000] <TabAtkins>
Okay, cool.

[15:06:29.0000] <TabAtkins>
Is the "type" in the attributes a bug too, since it should be inheriting that from Event already?

[15:08:59.0000] <Hixie>
what's the url to this page so i can go clean it up

[15:12:57.0000] <TabAtkins>
http://wiki.whatwg.org/wiki/Howto_spec#Events_implementing_their_own_interface

[15:13:12.0000] <TabAtkins>
(Damn adobe network temporarily drops about every 20 minutes.)

[15:14:57.0000] <Hixie>
ah, i think whoever wrote that just got confused and used "type" because they forgot about Event.type

[15:15:03.0000] <Hixie>
it's not meant to be related

[15:15:35.0000] <Hixie>
reload now

[15:16:22.0000] <Hixie>
the event handler stuff is out of date

[15:16:26.0000] <Hixie>
let me fix that too

[15:18:50.0000] <gsnedders>
Hixie: Tied to a device isn't the relevant point. It's about having a centralized store for contacts — it doesn't matter whether the contacts API works by delegating to some web service or whether it stores it on the device directly

[15:19:12.0000] <Hixie>
should just be implemented as a web service, imho

[15:19:17.0000] <Hixie>
possibly delegated using something like intents

[15:19:24.0000] <gsnedders>
But what web service?

[15:19:49.0000] <gsnedders>
The point is to have prefs set in one place, across the device, that can control where the prefs are stored.

[15:19:52.0000] <TabAtkins>
Hixie: Thanks for the update.  Much less confusing now. ^_^

[15:20:01.0000] <Hixie>
TabAtkins: updated the next bit too

[15:20:59.0000] <TabAtkins>
Okay, so even if I plan to have only a single event in this category, I still need to have a type attribute?

[15:21:17.0000] <Hixie>
there's no type attribute

[15:21:19.0000] <Hixie>
what do you mean

[15:21:21.0000] <Hixie>
i'm confused

[15:21:21.0000] <Hixie>
what?

[15:21:31.0000] <TabAtkins>
So, like "new FontFaceEvent('load')" rather than "new FontFaceLoadEvent()"

[15:21:46.0000] <Hixie>
oh, yes, because anyone can create an event that uses this interface

[15:21:56.0000] <Hixie>
the Event-inheriting interface and the event type are entirely orthogonal

[15:22:20.0000] <Hixie>
gsnedders: imho a Contacts API should define a common API mediated over something like MessagePorts and initiated via something like intents, which both the OS and any Web sites can offer to implement and provide

[15:22:31.0000] <TabAtkins>
Okay, so we're back to what I was trying to say, that the interface defines a group of events, with type used to distinguish specifically which event from the group is being sent.

[15:22:35.0000] <Hixie>
gsnedders: and the web browser doesn't need to know anything about it

[15:23:19.0000] <Hixie>
TabAtkins: more or less. pedantically, the interface defines a set of fields that events can use, there's no formal grouping of events into event types.

[15:23:24.0000] <gsnedders>
/me knows little of web intents

[15:23:29.0000] <TabAtkins>
Right.  Okay, that makes sense then.

[15:23:36.0000] <Hixie>
TabAtkins: and when you fire an event, you pick the interface it should use.

[15:23:41.0000] <Hixie>
TabAtkins: and the event type name


2012-07-25
[20:31:29.0000] <zewt>
i'm declaring the most useless often-used feature of HTML: copying/pasting HTML

[20:32:00.0000] <zewt>
every single time I copy text from one webpage and paste it into something like gmail or evernote I have to paste it into vim and copy it to get rid fo styling, or it pastes in some gigantic font or something else equally useless

[23:45:31.0000] <benschwarz>
Hixie: !ping

[00:18:26.0000] <annevk>
the new memes cracked me up

[00:22:12.0000] <annevk>
http://golem.ph.utexas.edu/~distler/blog/archives/002545.html <video poster> implementations are kind of bad when it comes to small videos

[01:00:46.0000] <hsivonen>
annevk: I observe that Jacques Distler's blog is not using MathML-in-text/html yet :-/

[01:01:52.0000] <annevk>
what's the benefit for him?

[01:02:48.0000] <hsivonen>
no benefit, I suppose, at the point where all XML bugs have been fixed and everything works

[01:03:30.0000] <hsivonen>
well, at least we carefully engineered the parsing algorithm to cater to the use cases seen on his blog

[01:04:44.0000] <annevk>
:)

[01:05:29.0000] <annevk>
the main benefit of HTML I guess is that you can use math entities and they will always work

[01:05:36.0000] <annevk>
whereas in XML it depends on whether the DTD is loaded

[01:06:19.0000] <hsivonen>
also, with current implementations, the page might load a tiny bit faster as HTML thanks to speculative loading on the HTML code path

[01:07:05.0000] <annevk>
but I think he normalizes them away

[01:09:26.0000] <hsivonen>
http://98.245.80.27/tcpc/OSCON2011/gbu.html is interesting

[01:11:29.0000] <annevk>
pretty positive over Go

[01:19:27.0000] <hsivonen>
http://html5.org/tools/web-apps-tracker?from=7201&to=7202 isn't a particularly useful response to the email hober sent. Basically, it kicks the problem the validator developers' court

[01:19:48.0000] <MikeSmith>
hsivonen: indeed

[01:20:40.0000] <annevk>
hsivonen: only if you decide it's worth the additional UI

[01:20:49.0000] <MikeSmith>
hsivonen: have you seen http://validator.keegan.st/ yet?

[01:23:53.0000] <hsivonen>
MikeSmith: I haven't

[01:23:55.0000] <hsivonen>
/me looks

[01:26:53.0000] <hsivonen>
MikeSmith: supporting CORS without a specific use case seems to have paid off

[01:27:21.0000] <MikeSmith>
yeah

[01:27:49.0000] <MikeSmith>
think it would be worthwhile to upstream that filter feature?

[01:31:26.0000] <hsivonen>
potentially, yes. need to play with it some more to have a proper opinion

[02:55:48.0000] <annevk>
heh http://www.netmagazine.com/interviews/ian-hickson-html5-vs-html

[02:55:59.0000] <annevk>
they still use the same old picture

[02:58:45.0000] <annevk>
http://my.opera.com/hallvors/blog/2012/07/25/sometimes-it-helps-to-speak-norwegian-in-new-zealand sounds fucked up

[02:59:16.0000] <annevk>
an <input type=submit> always has a value?

[03:25:37.0000] <jgraham>
Sweet jesus someone has to find a better picture of Hixie

[03:35:21.0000] <tantek>
the netmagazine article was pretty decent

[03:35:31.0000] <tantek>
obligatory troll commentary aside

[03:35:43.0000] <tantek>
s/troll/trollish

[03:36:23.0000] <tantek>
jgraham - what's wrong with the picture of Hixie - is it too old?

[03:37:50.0000] <tantek>
annevk - that *is* an odd bug re: submit.

[03:39:44.0000] <Stevef_>
tantek: what is trollish about the comment?

[03:40:14.0000] <tantek>
Stevef_ not comment, just some trollish commentary in the article itself.

[03:40:28.0000] <tantek>
I didn't notice anything trollish about your comment.

[03:40:40.0000] <Stevef_>
tantek: OK just checking

[03:43:29.0000] <tantek>
Stevef_ if you're curious, the "quicksand" comparison was what I was referring to.

[03:44:02.0000] <Stevef_>
tantek: ah OK

[03:44:58.0000] <Stevef_>
tantek: missed mention of shelley previoulsy...

[03:54:21.0000] <odinho>
;-)

[03:56:48.0000] <Ms2ger>
/me thinks it better not to comment on Shelley

[04:19:19.0000] <hsivonen>
What's new in Opera Mobile for Android: 'Improved handling of meta-tag "format detection"'

[04:19:23.0000] <hsivonen>
what does that really mean?

[04:28:22.0000] <tantek>
hsivonen - plenty of "formats" have been proposed recently (the past year + ) that use meta-tags (e.g. Facebook OGP, Twitter Cards), perhaps they're referring to some of those.

[04:32:26.0000] <MikeSmith>
so never took time to count up myself how many specs for Web APIs are in development in various working groups, but Erik Wilde did: http://dret.typepad.com/dretblog/html5-api-overview.html

[04:32:36.0000] <MikeSmith>
so we have on the order of 50 different specs

[04:43:11.0000] <hsivonen>
It would be interesting to do a analysis 'Who ISSUEs HTML5?' and tabulate stats about whose grievances have been raised as ISSUEs, who writes CPs, whose CPs get accepted and whose accepted CPs also got accepted in WHATWG HTML and whose not

[04:57:25.0000] <odinho>
hsivonen: No idea.

[04:58:40.0000] <hsivonen>
odinho: ok. not the clearest release note line item

[04:59:28.0000] <odinho>
hsivonen: You could say that. :] I asked if anyone knew.

[04:59:39.0000] <odinho>
I could search around but I'm not that interested :P

[05:15:25.0000] <MikeSmith>
hmm I guess it would be useful to make an alphabetical list list of specs like http://dret.typepad.com/dretblog/html5-api-overview.html but also include the CSS3 specs and the non-W3C ones that Erik omits from that

[05:15:52.0000] <MikeSmith>
with those included it's more on the order of 60 specs

[05:22:53.0000] <hsivonen>
dealing with some of my least favorite parts of the platform: session history and window.open

[05:31:37.0000] <odinho>
hsivonen: Found this in the bugtracker (for that changelog entry):  The other problem worth to mention is Google Play limiting number of characters in changelogs, so we always need to be as short as possible ;)

[05:32:10.0000] <odinho>
hsivonen: But it's not linking to the real bug, so I still don't know what it means, but now know why the log entry is so criptic :P

[05:32:22.0000] <hsivonen>
odinho: thanks

[05:34:02.0000] <odinho>
hsivonen: FOUND IT!

[05:34:21.0000] <odinho>
http://developer.apple.com/library/safari/#featuredarticles/iPhoneURLScheme_Reference/Articles/PhoneLinks.html#//apple_ref/doc/uid/TP40007893-SW1 --->   <meta name = "format-detection" content = "telephone=no">

[05:34:25.0000] <odinho>
this here thing

[05:37:41.0000] <hsivonen>
odinho: thanks. totally non-obvious from the relnotes

[05:38:46.0000] <odinho>
hsivonen: yes, agreed. But probably because Play store had a length limit on changelogs. Should've been a link to something bigger though.

[05:51:35.0000] <hsivonen>
hmm. I don't see an invitation to bring <picture> to the HTML WG on the Responsive Images GC's mailing list

[05:51:50.0000] <hsivonen>
I wonder who did the invitation and why

[05:53:11.0000] <hsivonen>
I suspect the Responsive Images folks are going to face more of http://w3cmemes.tumblr.com/post/23122022271

[05:53:38.0000] <hsivonen>
"And then we told them ... to submit a proposal to the HTML WG"

[05:56:10.0000] <odinho>
lol, the tantek one is pure gold :D

[06:01:30.0000] <Ms2ger>
/me looks around http://www.cs.helsinki.fi/u/pettay/moztests/xmlevents/

[06:07:10.0000] <hsivonen>
at times, fixing about:blank feels hopeless

[06:32:03.0000] <hsivonen>
I wish the testing staff funded by Adobe, Microsoft and Google writes some good session history tests

[06:53:32.0000] <hsivonen>
am I the only one who read Maciej's advice about how to proceed with the era attribute as "Don't bother, but if you want a formal rejection, do this."?

[07:37:20.0000] <AryehGregor>
http://www.w3.org/TR/SVG11/svgdom.html#__svg__SVG_MATRIX_NOT_INVERTABLE

[07:37:22.0000] <AryehGregor>
Really?

[07:37:24.0000] <AryehGregor>
INVERTABLE?

[07:37:26.0000] <AryehGregor>
That made it into a spec?

[07:37:48.0000] <AryehGregor>
"Referer" is one thing, I'd like to think we're more professional now . . .

[07:40:54.0000] <Ms2ger>
Well

[07:40:57.0000] <Ms2ger>
SVGDOM

[07:42:46.0000] <odinho>
Referer <3

[08:03:04.0000] <heycam>
AryehGregor, sorry about that, I added a note to SVG 1.1 Second Edition pointing out the unconventional spelling :)

[08:03:18.0000] <heycam>
AryehGregor, (we're moving to Web IDL exception types soon anyway)

[08:03:22.0000] <AryehGregor>
Yay!

[08:03:29.0000] <AryehGregor>
Wait, WebIDL and not DOM?

[08:03:40.0000] <heycam>
AryehGregor, oh using the DOM ones

[08:03:44.0000] <AryehGregor>
Ah.

[08:03:44.0000] <heycam>
I don't think we need any specific SVG ones

[08:04:48.0000] <Ms2ger>
Oh

[08:04:55.0000] <Ms2ger>
I thought you guys did that ages ago

[08:08:37.0000] <jarek>
SVG DOM is awful

[08:10:01.0000] <jarek>
anyone here thinks that SVGPathSegCurvetoQuadraticSmoothRel is a good name?

[08:10:27.0000] <jarek>
looks a bit over-descriptive :/

[08:38:39.0000] <odinho>
jarek: Looks like Objective-C/Apple API's.

[08:55:19.0000] <gsnedders>
jarek: It's an awesome name!

[09:27:45.0000] <Hixie>
MikeSmith: to be fair, seven of those are fragments of the HTML spec

[09:28:49.0000] <Hixie>
and several of them are "native api" specs that really don't belong on the web in the first place imho

[09:29:03.0000] <Hixie>
or should be done as intent-driven web apis not implemented by the browser

[09:29:19.0000] <Hixie>
and he lists xhr and xhr2 as separate specs

[09:29:39.0000] <Hixie>
and media capture twice

[09:29:58.0000] <Hixie>
would be nice to have platform.html5.org updated to include any that were missed though

[09:44:34.0000] <annevk>
Hixie: maybe we need meta⊙wo? http://www.codinghorror.com/blog/2012/02/listen-to-your-community-but-dont-let-them-tell-you-what-to-do.html

[09:45:19.0000] <annevk>
on the other hand, we've been mostly good for eight years

[09:56:17.0000] <karlcow>
/me weird

[09:56:44.0000] <karlcow>
/me wonders what is the correct term instead of INVERTABLE?

[09:57:14.0000] <annevk>
INVERTIBLE

[09:57:23.0000] <karlcow>
annevk: thanks

[10:00:04.0000] <jzaefferer>
The spec is pretty ambigous about popstate event being fired on pageload: http://www.whatwg.org/specs/web-apps/current-work/multipage/history.html - now we have Chrome doing that, Firefox not. I'd like to see that fixed before we get stuck with that mess forever. Looking for some guidance how.

[10:01:58.0000] <jzaefferer>
from the MDN article https://developer.mozilla.org/en/DOM/window.onpopstate - "Browsers tend to handle the popstate event differently on page load. Chrome and Safari always emit a popstate event on page load, but Firefox doesn't." - that sucks for users

[10:03:25.0000] <jzaefferer>
MikeSmith: as my goto-whatwg person, any comment on the above?

[10:04:59.0000] <Ms2ger>
jzaefferer, file a bug to get it defined

[10:05:31.0000] <jzaefferer>
Ms2ger: okay, which tracker would be the right one?

[10:06:13.0000] <Ms2ger>
https://www.w3.org/Bugs/Public/enter_bug.cgi?product=WHATWG&component=HTML

[10:07:50.0000] <jzaefferer>
okay, thanks

[10:09:15.0000] <Ms2ger>
Np

[10:15:44.0000] <tantek>
annevk - the meta topic is interesting from a community management perspective.

[10:16:26.0000] <tantek>
however, it tends to eventually be dominated by people that don't love a community (as Atwood suggests), but by people that love meta for meta's sake.

[10:16:57.0000] <tantek>
prime example: the amount of time spent discussing, arguing, and trying to meet the steps for "process" at W3C.

[10:17:19.0000] <tantek>
so no, I don't think we need meta⊙wo

[10:18:55.0000] <tantek>
I mean unless you want to create a support group ;) for people who love meta for meta's sake *and* somehow found themselves involved with whatwg. It could be an interesting honeypot, to keep meta talk out of other fora.

[10:20:28.0000] <hober>
it could be useful as a honeypot

[10:24:29.0000] <karlcow>
/me is not subscribing to meta ☺ ever

[10:24:39.0000] <karlcow>
already enough of nonsense everywhere

[10:27:54.0000] <annevk>
tantek: fair enough

[10:28:22.0000] <tantek>
karlcow - but you *love* discussing process :)

[10:28:34.0000] <Ms2ger>
tantek, well, the point is to keep them away from the productive people ;)

[10:29:07.0000] <annevk>
tantek: I wasn't quite sure what we'd use it for either, just thought I'd suggest it so it could be considered

[10:30:44.0000] <tantek>
I feel like #whatwg provides a sufficient meta-forum. Perhaps consider meta@ lazily, that is, avoid it until someone makes a strong case+demand for it, and then give it to them.

[10:31:08.0000] <karlcow>
tantek: not really ☺ I love understanding/observing communities. I do not think there is one solution for all. I have the gut feeling that all Process often fails at having an annotated version with links to historical references.

[10:31:37.0000] <tantek>
+1 on Process often fails at having an annotated version with links to historical references

[10:33:04.0000] <tantek>
karlcow - in addition, I think the bigger failure of most process documentation is that it fails to document what's specifically been avoided/rejected (and why).

[10:41:34.0000] <karlcow>
/me is still thinking about that last sentence from Tantek

[10:41:38.0000] <Hixie>
tantek++ re not neeting a meta mailing list

[10:41:43.0000] <Hixie>
and re #whatwg being it

[10:41:55.0000] <Hixie>
needing, even

[10:42:07.0000] <Hixie>
that typo bodes poorly for the day's activities

[10:43:07.0000] <karlcow>
hmmm lightly to heavily documented https://en.wikipedia.org/wiki/Stare_decisis#Civil_law_systems differences depending on countries

[10:43:24.0000] <Ms2ger>
Hixie, might give you a nice set of easy typo-bugs tomorrow :)

[10:44:34.0000] <Hixie>
hah

[10:44:47.0000] <Hixie>
got plenty of those already

[10:48:26.0000] <Hixie>
volkmar: yt?

[10:52:51.0000] <volkmar>
Hixie: yes

[10:54:45.0000] <Hixie>
volkmar: do you know much abotu setMessageHandler()? as in https://groups.google.com/forum/?fromgroups#!topic/mozilla.dev.webapi/o8bkwx0EtmM

[11:05:59.0000] <volkmar>
damn, I was going to send you that link :)

[11:06:05.0000] <volkmar>
what do you want to know?

[11:48:36.0000] <Hixie>
tantek: do you have an ETA for the autocomplete feedback you were going to send btw? (for planning purposes)

[11:53:49.0000] <Hixie>
for anyone who is interested, my current (incomplete and in parts wrong) strawman proposal for merging all the various content handlers, web intents, web activities, system messages, etc, proposals is: http://www.whatwg.org/specs/web-apps/current-work/scratch

[12:31:40.0000] <Wilto>
hsivonen: Ian Jacobs reached out to me on behalf of the HTML WG chairs, to the best of my knowledge.

[12:32:21.0000] <Wilto>
hsivonen: That is, simply to post the proposal to the list.

[12:38:32.0000] <annevk>
http://html5.org/r/7203 lol

[12:38:56.0000] <Hixie>
:-P

[12:39:13.0000] <annevk>
haven't seen requests for XMLHttpRequest yet

[12:39:20.0000] <Hixie>
weird

[12:42:23.0000] <hober>
Hixie: what's the use case for having "new" AND "reuse"?

[12:42:53.0000] <Hixie>
most intents won't be able to handle the same page being used for two intents

[12:42:58.0000] <Hixie>
think, e-mail compose window intent handler

[12:43:09.0000] <Hixie>
so for those, you'd want "new"

[12:43:10.0000] <Ms2ger>
annevk, count this as a request :)

[12:43:27.0000] <Hixie>
"reuse" is for the cases where the app can in fact handle multiple things at once

[12:43:34.0000] <Hixie>
hober: does that make sense?

[12:43:39.0000] <annevk>
Ms2ger: you don't count :p

[12:44:00.0000] <Hixie>
hober: (btw, your input on this in general is very welcome, i'm very interested in safari and opera input on this since it has mostly been a mozilla and chrome thing so far)

[12:44:15.0000] <hober>
I suppose. Being stupid or lossy under the "reuse" behavior seems like a bad way to code an intent handler.

[12:44:19.0000] <Hixie>
(also interested in IE feedback, should any Microsoft people be lurking here or reading the logs)

[12:44:38.0000] <Hixie>
hober: right, hence why you'd say you're a "new" handler, then you don't have to worry about it

[12:44:52.0000] <Hixie>
hober: you wouldn't want all your e-mail compose intents to end up in the same tab :-)

[12:45:12.0000] <Hixie>
or all your "open PDF in new window" intents :-)

[12:45:48.0000] <hober>
why would "open pdf in new window" have anything to do with this mechanism? :)

[12:46:14.0000] <Hixie>
PDF viewing sites might well call registerContentHandler() for application/pdf

[12:46:43.0000] <hober>
"PDF viewing sites" *shudder*

[12:46:55.0000] <Hixie>
ok, 3D Model Viewing Sites

[12:46:57.0000] <Hixie>
whatever :-)

[12:47:23.0000] <Hixie>
think of Pages as a web page

[12:47:30.0000] <Hixie>
it handles the "open Pages app" intent

[12:47:33.0000] <Hixie>
er

[12:47:38.0000] <Hixie>
"open Pages doc" intent

[12:47:44.0000] <Hixie>
but each Pages doc ends up in a new Window

[12:47:47.0000] <Hixie>
so it's like "new"

[12:48:11.0000] <Hixie>
but then consider the "open MP3 doc" intent that iTunes supports

[12:48:18.0000] <Hixie>
it handles all of them in the same window

[12:48:27.0000] <Hixie>
so it's like "reuse"

[12:52:32.0000] <hober>
*nod*

[12:54:08.0000] <annevk>
oh hober

[12:54:10.0000] <annevk>
poor you

[12:54:29.0000] <hober>
On the plus side, I don't have to think much about what my Halloween costume will be at TPAC this year

[12:54:30.0000] <annevk>
I mean, have the funnest time

[12:55:08.0000] <annevk>
haha

[12:55:17.0000] <Hixie>
because you'll have given up and become an alpaca farmer by then? :-)

[12:55:23.0000] <hober>
(Zoidberg, obviously)

[12:56:00.0000] <hober>
hahahaha

[12:57:09.0000] <Hixie>
do you know when the four of you will begin editing?

[12:57:11.0000] <annevk>
I considered alpaca farming, but then the tent froze over during the night and I decided not to

[12:57:28.0000] <Hixie>
hober: (i'm mainly curious about when i should bother approaching y'all to see how you want to handle getting new bugs to both teams)

[12:58:24.0000] <Ms2ger>
I guess I could see that coming

[13:02:16.0000] <hober>
Hixie: the initial meeting hasn't even happened, so "not this week" is probably a safe answer.

[13:03:19.0000] <Hixie>
you're going to have meetings?

[13:03:25.0000] <Hixie>
oh poor poor hober

[13:03:38.0000] <Hixie>
lunch, bbiab

[14:17:11.0000] <tantek>
Hixie - re: autocomplete feedback completion ETA, good estimate would be next Monday, though I plan to work on it a bunch today.

[14:17:38.0000] <tantek>
I'll be basing my feedback on editing in real-time the hcard-input page in case anyone is interested in stalking/following along (edit notifications also go to #microformats) http://microformats.org/wiki/hcard-input#next_actions

[14:47:15.0000] <Hixie>
Any quirks mode spec people around?

[16:55:14.0000] <tantek>
Hixie, I've purged quirks mode details from my memory since leaving Microsoft. Hoping that many (most?) of them no longer matter.


2012-07-26
[17:03:21.0000] <Hixie>
well we have a quirks spec now so they're no longer really quirks :-)

[17:06:04.0000] <tantek>
I'm sure it'll evolve into the "no doctype" mode over time for ultraminimalists.

[18:08:10.0000] <hober>
is there a w3c bugzilla component for the fullscreen spec?

[22:04:41.0000] <hsivonen>
Wilto: interesting. not at all who I expected. thanks.

[22:05:08.0000] <hsivonen>
looks like my guess that hober was on track to editing HTML5 was right but not exclusively right

[22:06:30.0000] <hsivonen>
Wilto: anyway, to get adaptive images to implementation, I expect working through the WHATWG Community Group to be more productive than working through the HTML WG

[22:07:06.0000] <hsivonen>
Wilto: if you try to get it into HTML5, it will go into the Process bureaucracy and get rejected

[22:07:30.0000] <hsivonen>
Wilto: if you try to get it into HTML.next via the HTML WG, it will get delayed due to the focus on HTML5

[22:07:51.0000] <hsivonen>
Wilto: while the actual ongoing work on HTML happens at the WHATWG

[22:09:27.0000] <hsivonen>
congrats to the new editors

[22:55:28.0000] <MikeSmith>
hober: https://www.w3.org/Bugs/Public/buglist.cgi?product=WebAppsWG&component=Fullscreen&resolution=---

[22:58:44.0000] <Hixie>
how is that in WebApps

[22:58:46.0000] <Hixie>
isn't that a CSS thing?

[23:00:38.0000] <roc_>
arguably, but it's more DOM API than anything else

[23:00:56.0000] <Hixie>
i meant CSSWG thing

[23:01:00.0000] <Hixie>
agreed entirely that it's an API

[23:02:07.0000] <roc_>
I guess

[23:02:38.0000] <Hixie>
i shouldn't try to understand w3c stuff

[23:04:16.0000] <Hixie>
hey so i'm marking all the things in the spec that lead to fingerprinting

[23:04:23.0000] <Hixie>
anyone think of any i might have missed?

[23:16:33.0000] <benvie>
in js you can identify the browser and often js engine revision by identifying the MAX_STACK count though that's not a whatwg thing

[23:17:18.0000] <Hixie>
as in function foo() { count+=1; foo() } var count = 0; try { foo() } catch (e) { alert(count) }  ?

[23:17:26.0000] <benvie>
yeah

[23:17:31.0000] <benvie>
in v8 you can pin it to a specific revision

[23:18:02.0000] <benvie>
since for some reason it varies each release, but is static for that release

[23:18:24.0000] <Hixie>
huh

[23:18:56.0000] <Hixie>
heh, i guess firefox implements tail recursion optimisations because it never stopped for me :-)

[23:19:05.0000] <benvie>
yeah it's weird

[23:19:11.0000] <benvie>
it worse in spidermonkey shell, but not firefox

[23:19:16.0000] <benvie>
it works in any webkit and it works in IE

[23:19:23.0000] <AryehGregor>
Hixie, why are Latin and CJK the only scripts singled out in inputmode?  Is <input inputtype=latin> really meant to mean "whatever the usual script is for the current language"?  If so, why is it called "Latin"?

[23:19:23.0000] <benvie>
so jsc and v8 are similar

[23:20:00.0000] <benvie>
but in spidermonkey shell it always returned 4000 so it was useless except to identify generally that it was firefox

[23:20:08.0000] <benvie>
er spidermonkey

[23:23:14.0000] <benvie>
http://code.google.com/p/chromium/issues/detail?id=117531 is the issue I filed with Chromium since it's so specific as to almost be a security issue

[23:53:48.0000] <Hixie>
ok, i've added notes in the spec for fingerprinting risk

[23:54:24.0000] <Hixie>
AryehGregor: because nobody from any cultures other than the japanese responded to me six zillion requests for information, and i only knew about western (latin) input mode requirements.

[23:54:32.0000] <Hixie>
s/me/my/

[23:55:25.0000] <Hixie>
AryehGregor: so i finally said screw it, i'll give what the japaneses and the latin-script-using-west need, and if anyone else wants anything, they can get it when they provide me with the info i need

[00:00:01.0000] <hsivonen>
Hixie: yay for "latin" instead of "en", etc.

[00:00:46.0000] <hsivonen>
It's super-annoying when software is designed with the assumption that people who write in two languages write in two scripts

[00:01:41.0000] <hsivonen>
maybe once Spanish usage goes high enough in the U.S., more software designers realize that people input multiple languages with one Latin layout

[00:02:59.0000] <hsivonen>
/me has to turn off input prediction on Android, because it doesn't support predicting both English and Finnish without being explicitly told which one to predict at a given time

[00:04:02.0000] <hsivonen>
(on the bright side, Android support putting the U.S. English speech regornition button on a Finnish keyboard)

[00:04:08.0000] <hsivonen>
*supports

[00:05:25.0000] <hsivonen>
Hixie: btw, any reason not to reuse 4-letter script codes from the language tag RFC?

[00:11:56.0000] <Hixie>
hsivonen: on android you can make the space bar support switching languages, which is what i do for english vs french

[00:12:11.0000] <annevk>
Hixie: Fullscreen is in both WebApps and CSS :(

[00:12:15.0000] <Hixie>
hsivonen: but yeah, same keyboard, different language, that's a separate axis than inputmode

[00:12:28.0000] <Hixie>
hsivonen: reuse the script codes for what?

[00:12:47.0000] <annevk>
Hixie: And the CSS part is problematic with their sync decision making and mailing list I'm not subscribed to

[00:12:51.0000] <hsivonen>
Hixie: keyboard category

[00:13:02.0000] <Hixie>
annevk: why not just do it like your other specs?

[00:13:13.0000] <hsivonen>
Hixie: I don't want to tell Android whether I'm writing Finnish or English

[00:13:26.0000] <Hixie>
hsivonen: oh because i don't expect the script labels to match well to what keyboards inputmode="" needs

[00:13:28.0000] <annevk>
Hixie: The editor's draft is the WHATWG draft

[00:13:39.0000] <hsivonen>
Hixie: surely Google's mighty algorithms should be able to figure out which language of the two I'm writing after a word or so

[00:13:46.0000] <hsivonen>
Hixie: ok

[00:14:00.0000] <Hixie>
annevk: fun

[00:14:08.0000] <annevk>
so much

[00:15:16.0000] <Hixie>
hsivonen: it's pretty hard to distinguish poor typing from another language, probably. still, long-tap space bar, tap language, it's not exactly a burden to switch back and forth

[00:16:02.0000] <Hixie>
annevk: just FSA it and then you don't have to worry every again about the webapps/css versions :-)

[00:23:52.0000] <Hixie>
nn

[00:24:56.0000] <annevk>
Hixie: did you mark the Navigator object's friends?

[00:25:43.0000] <annevk>
looks like

[00:25:59.0000] <annevk>
that is one ugly icon btw

[00:26:02.0000] <annevk>
:)

[00:46:59.0000] <AryehGregor>
Hixie, what info do you need?

[00:47:36.0000] <AryehGregor>
I'd think most scripts would work the same as Latin, basically.

[00:48:13.0000] <AryehGregor>
What's the difference meant to be between "verbatim", "latin", and "latin-prose"?  What are the use-cases for the distinction?

[01:18:38.0000] <AryehGregor>
I can fill you in on any Hebrew-specific issues, if there are any.

[01:22:57.0000] <jgraham>
So, what did I miss?

[01:31:27.0000] <liebrand>
hi there… i've been trying to unsubscribe from the whatwg mailing list, but although it claims it has sent me a confirmation email to unsubscribe, I never receive that email.. i've tried a few times now and checked my spam folder as well.. is there any other way to unsubscribe?

[01:32:29.0000] <AryehGregor>
liebrand, are you sure you're unsubscribing the right e-mail address?

[01:32:49.0000] <liebrand>
yeah, 99% sure

[01:34:14.0000] <liebrand>
is there someone one here that can verify my email address?

[01:35:23.0000] <AryehGregor>
Well, Hixie can.  Dunno if anyone else has admin privileges to the list.

[01:37:46.0000] <liebrand>
ok thanks

[01:59:30.0000] <annevk>
liebrand: email ian⊙hc if all else fails

[02:09:47.0000] <Smylers>
Apologies for the duplicate mails to the list.

[02:10:21.0000] <Smylers>
Apparently my mailer was picking up both the CC: whatwg⊙wo and the List-Post: <mailto:whatwg⊙lwo> headers of the mail I was replying to.

[02:10:45.0000] <Smylers>
I'll try to watch out for that in future.

[02:17:12.0000] <jgraham>
So, uh, why did all my bugs get cloned?

[02:18:29.0000] <odinho>
jgraham: Because whatwg got its own HTML bugzilla place :)

[02:18:39.0000] <odinho>
jgraham: Read Administrativa email to the WHATWG-list.

[02:19:47.0000] <liebrand>
annevk: thanks!

[02:21:56.0000] <annevk>
Smylers: Gmail copes

[02:22:00.0000] <jgraham>
odinho: That email told me to stop reading after the first paragraph :p

[02:22:21.0000] <odinho>
jgraham: lol, that doesn't mean you.

[02:22:56.0000] <odinho>
"If you've been happily ignoring the W3C's involvement with HTML these past

[02:22:56.0000] <odinho>
few years, you can stop reading now. If you got a bunch of bugmail

[02:22:57.0000] <odinho>
recently and want to know why, the explanation is below."

[02:23:05.0000] <odinho>
Second part of that is absolutely you.

[02:23:14.0000] <Smylers>
annevk: In that case, apologies for bothering you here with an unnecessary apology!

[02:23:17.0000] <jgraham>
odinho: Then Hixie should reorder his if clauses

[02:23:40.0000] <jgraham>
Is there somewhere I can report a bug on the email?

[02:23:44.0000] <Smylers>
jgraham: You've been happily ignoring the W3C, while using their Bugzilla to report spec bugs?

[02:24:02.0000] <odinho>
jgraham: I think this channel is it.

[02:26:13.0000] <jgraham>
Smylers: It wouldn't make much difference to me if the bug tracker were run by the US president or Tibetan monks

[02:27:00.0000] <Smylers>
jgraham: OK, that covers the ‘ignoring’. Can you prove the ‘happily’?

[02:29:41.0000] <jgraham>
Hmm, so if the spec is Just Wrong, I now have to file two bugs rather than one?

[02:30:09.0000] <Smylers>
jgraham: Only if you care about the W3C's edition of the spec.

[02:30:38.0000] <Smylers>
jgraham: If you continue ignoring it — happily or otherwise — then you don't have to report it there.

[02:30:43.0000] <jgraham>
Or is the assumption that being wrong ("stable") is a desired property of the W3C spec?

[02:31:22.0000] <Smylers>
jgraham: Also, Hixie said he'll keep an eye on bugs reported on the W3C spec to see if they apply to the WhatWG spec too. Possibly the new W3C editors will do likewise in reverse.

[02:32:59.0000] <jgraham>
Smylers: It seems to me that the new system requires me to pay more attention to the W3C than the old system. Previously Hixie ensured that there were few differences in areas that are actually interoperability concerns

[02:33:12.0000] <jgraham>
Now everyone has to do this for themselves

[02:36:07.0000] <odinho>
jgraham: What with the pessimism? Lets try it out :P

[02:36:15.0000] <jgraham>
This wouldn't be a problem but people continue to percieve the W3C version of the spec as the "definitive" one that they should refer to from other specs

[02:36:40.0000] <jgraham>
odinho: I don't understand your negative attitude toward pessimism

[02:36:55.0000] <odinho>
:P

[02:37:41.0000] <Stevef_>
jgraham: suppose the assumption is that implementation bugs in HTMl5  will be fixed by the editorial team up until time when it gets frozen, then bugs will get fixed in htmlx

[02:39:56.0000] <jgraham>
/me retreats to the relative safety of public-webapps

[02:40:03.0000] <Stevef_>
jgraham: there are a number of people on the team that are switched on to the importance of keeping the specs aligned on the brwoser implementation side so the hope is...

[02:41:23.0000] <jgraham>
Stevef_: If people are going to put in the effort to keep the specs in alignment as far as UA conformance criteria go, that is great, but I fear the task will be Herculean

[02:42:15.0000] <Smylers>
jgraham: If the W3C spec suffers from bugs, because they've only been fixed in the WhatWG spec, that may lead to a change in the perception of which one is seen as definitive.

[02:42:48.0000] <jgraham>
Smylers: Ah, is that what optimisim sounds like?

[02:43:59.0000] <Stevef_>
jgraham: understand the concern, but there are 4 editors + their will be a full time technical editor, so they may have some chnace in keeping up with hixie. Also the canvas stuff will have a seperate team

[02:46:48.0000] <jgraham>
I suppose that Hixie should be flattered that he is being replaced by > 5 people

[03:06:08.0000] <AryehGregor>
It's not really a fair comparison, because Hixie is like the only full-time web spec editor in the universe.

[03:08:24.0000] <Ms2ger>
Presumably whoever the W3C ends up hiring will be the second, then

[03:09:10.0000] <AryehGregor>
Oh, I didn't realize there was a full-time job.

[03:09:12.0000] <AryehGregor>
Interesting.

[03:11:23.0000] <Stevef_>
aryehgregor: http://www.w3.org/Consortium/Recruitment/#techeditor

[03:12:20.0000] <Stevef_>
its still listed, but get the feeling they alreaedy have a candidate

[03:13:25.0000] <AryehGregor>
Interesting.

[03:14:03.0000] <AryehGregor>
I wonder if they'll get someone good.  The obvious candidates are all already employed.

[03:14:19.0000] <Ms2ger>
Will it be hober?

[03:14:25.0000] <Ms2ger>
Maybe Marcosc

[03:39:36.0000] <annevk>
Marcos is too radical for them

[03:39:48.0000] <annevk>
they want someone they can control

[03:40:08.0000] <annevk>
and hober is already happily employed

[04:12:13.0000] <Stevef_>
well hober is one of the HTML5 editors now

[04:13:12.0000] <Stevef_>
along with sylvia pfeiffer

[05:01:56.0000] <sedovsek>
There's still no support for break-before, break-after, etc. for Multiple Columns in any modern browser? :/

[06:41:38.0000] <annevk>
MikeSmith: hsivonen: I just approved this comment: http://blog.whatwg.org/vnu-parser-1-4#comment-43879

[06:43:31.0000] <annevk>
anyone know who Erika Doyle Navara is btw?

[06:43:46.0000] <Ms2ger>
Oh, that's me

[06:45:28.0000] <annevk>
of course you are

[06:47:06.0000] <odinho>
hehe

[07:32:01.0000] <david_carlisle>
annevk: don't know,  but linkedin has an Erika Doyle at "Senior Writer at Microsoft Internet Explorer Developer Content Team"  ?

[07:33:49.0000] <david_carlisle>
annevk:  oh and an Erika Navara at the same place:-)

[07:34:51.0000] <david_carlisle>
http://www.linkedin.com/pub/erika-navara/1/524/385

[07:35:11.0000] <annevk>
just says "Greater Seattle Area  Computer Software" for me

[07:37:07.0000] <david_carlisle>
annevk:  I think you get more text if you log in.

[07:37:45.0000] <annevk>
ah okay, don't have an account

[07:37:56.0000] <david_carlisle>
annevk: very wise:-)

[07:39:34.0000] <annevk>
maybe we should start spelling W3C W3 C or some such

[07:39:42.0000] <annevk>
http://www.w3.org/QA/2012/07/html5_and_htmlnext.html how hard can it be to get a name right

[07:41:24.0000] <annevk>
(personal pet peeve, misspellings of names annoys me)

[07:41:48.0000] <david_carlisle>
It's as easy to get a name right as it is to guess gender fro the name, isn't it ?

[07:41:54.0000] <david_carlisle>
from

[07:42:24.0000] <annevk>
getting the name right is one Google search

[07:42:39.0000] <annevk>
getting the gender right is one Google image search typically

[07:42:48.0000] <david_carlisle>
annevk:  well you show up in the odd google search as well:-)

[09:20:03.0000] <dglazkov>
good morning, Whatwg!

[11:06:04.0000] <jarek>
Hi

[11:06:36.0000] <jarek>
it looks like window.getComputedStyle completely ignores styles set with SVG presentational attributes

[11:07:03.0000] <jarek>
is this by design? If so, it doesn't make sense because presentational attributes are part of the cascade

[11:25:47.0000] <jarek>
nvm, it actually works as expected

[11:49:07.0000] <hober>
this is *screaming* to be meme-ified, but i fear the super cool ski instructor has made too many appearances lately. http://hueniverse.com/2012/07/oauth-2-0-and-the-road-to-hell/

[11:57:03.0000] <Ms2ger>
The question is, is dbaron's desk at Harvard OAuth 2.0 compliant?

[12:00:27.0000] <Hixie>
man, he's come a long way

[12:00:33.0000] <Hixie>
he was instrumental in bringing websocket to the ietf

[13:34:21.0000] <annevk>
hober: could use the hyatt meme

[13:34:27.0000] <annevk>
hober: where everything blows up

[13:35:02.0000] <annevk>
hober: caption: "Just use OAuth"

[14:42:41.0000] <hober>
hmm.

[14:42:53.0000] <Hixie>
hmm?

[14:43:04.0000] <hober>
contemplating annevk's meme suggestion

[14:43:09.0000] <matjas>
astral Unicode symbols in <input type=password> show up as two asterisks: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/1684

[14:43:29.0000] <Ms2ger>
Neat

[14:43:43.0000] <matjas>
in all browsers I tested so far… so i guess it’s standard

[14:44:00.0000] <matjas>
(OS X 10.8 seems to do the same for all password fields)

[16:26:33.0000] <zewt>
matjas: heh yeah, the hope of web developers ever getting non-BMP characters right is pretty remote when browsers don't even get it right

[16:27:33.0000] <zewt>
matjas: ie9 shows a single *, fyi

[16:30:27.0000] <zewt>
opera shows two dots but it edits like a single character (better than ff/chrome, anyway)

[16:38:15.0000] <Hixie>
i swear, i'm _this_ close to dropping the authorship information in the references section

[16:52:08.0000] <WeirdAl>
Hixie - tired of staring at a thousand names?

[16:58:01.0000] <gsnedders>
zewt: Opera on what platform? This is probably different between platforms.

[16:58:14.0000] <gsnedders>
Well, the impl is. The behaviour might be.

[16:58:16.0000] <zewt>
win7

[16:59:27.0000] <zewt>
haha context menu in opera on password inputs has "check spelling"

[16:59:35.0000] <zewt>
that might possibly be among the most useless context menu items ever


2012-07-27
[17:00:05.0000] <gsnedders>
/me wonders if it does anything

[17:00:11.0000] <zewt>
it should warn if you *don't* have typos

[17:02:01.0000] <gsnedders>
Okay, it doesn't do anything.

[17:02:46.0000] <zewt>
seems strange that browsers still don't allow copying out of password fields

[17:02:57.0000] <zewt>
even though every browser lets you get at it in inspector tools anyway

[17:10:31.0000] <Hixie>
WeirdAl: they keep fricking changing and keeping them up to date is of no value to me whatsoever

[17:11:43.0000] <WeirdAl>
I'll admit, it probably looks good on a thousand resumes.

[17:11:51.0000] <WeirdAl>
but other than that...

[17:13:01.0000] <Hixie>
not the acknowledgements, i mean the references section

[17:13:08.0000] <WeirdAl>
oh

[17:13:36.0000] <Hixie>
certainly not going to remove the names from the acknowledgements section, listing those names is the least i can do given how much help everyone has been

[17:16:52.0000] <Hixie>
annevk: btw i checked and i'm still sending your script the "typo"/"xref" thing, even when i'm sending the svn repo and e-mails a longer description

[23:03:52.0000] <hsivonen>
hober: The OAuth story really calls for "if you take your spec to the IETF, you're going to have a bad time", despite the frequency of the appearance of the super-cool ski instructor.

[23:04:48.0000] <hsivonen>
hober: I guess that the reason why the super-cool ski instructor appears so frequently is that it's the only advice animal that actually gives advice, and advice is needed when making standards.

[23:14:29.0000] <roc>
I believe the Opus IETF group wasn't terrible

[23:20:23.0000] <tantek>
that's good to hear

[23:24:11.0000] <annevk>
Hixie: cool

[23:24:23.0000] <annevk>
hober: my meme suggestion was a bad one

[23:45:44.0000] <annevk>
http://www.karlgroves.com/2012/07/26/selling-accessibility-negative-factors/ this is pretty how much how accessibility topics are handled in a F2F

[23:48:17.0000] <tantek>
annevk - wow

[23:48:30.0000] <tantek>
"The fact that accessibility is so hard and time consuming is partly because accessibility folks spend a lot of time chasing unicorns – seeking to pile on new requirements, check-points, and validation plans in hopes to make everything perfect. They separate instead of integrate and only show up when it is time to tell people they’ve messed everything up."

[23:48:53.0000] <annevk>
also the bit about hostility and FUD

[23:49:25.0000] <annevk>
every time I meet John Foliot he mentions lawsuits, rules of large corporations, new US legislation

[23:50:25.0000] <annevk>
not very compelling

[00:16:58.0000] <matjas>
zewt: re: astral symbols in <input type=password>, is it a bug to show them as two symbols or not? I’m not sure, with DOM strings being UTF-16 and all

[00:30:40.0000] <annevk>
matjas: depends on the UI requirements

[00:30:56.0000] <annevk>
matjas: it seems like suboptimal UI, but not necessarily a bug

[00:31:03.0000] <matjas>
annevk: is this stuff in a spec somewhere?

[00:31:11.0000] <annevk>
matjas: we don't spec UI

[00:31:13.0000] <hsivonen>
it seems like a bug but not one that's important to fix

[00:31:43.0000] <annevk>
matjas: you could take a look if the Rendering section says something, but I doubt it

[00:32:27.0000] <annevk>
"When the input-password binding applies, to an input element whose type attribute is in the Password state, the element is expected to render as an 'inline-block' box rendered as a text field whose contents are obscured."

[00:32:39.0000] <matjas>
that’s all I could find too

[00:33:25.0000] <kennyluck>
Hixie, you mentioned you were asking people about requirements on input mode control. Do you have a mail that summarizes your questions or something?

[00:33:26.0000] <hsivonen>
rendering two bullets for one character sure counts as obscured!

[00:33:43.0000] <annevk>
what I want fixed more than this is a "reveal password" thingie

[00:34:40.0000] <annevk>
kennyluck: we're look for use cases here: http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control

[00:35:22.0000] <kennyluck>
annevk, yeah, I know that page, though Hixie indicated that he sent some mails out, not sure if those were internal.

[00:35:53.0000] <kennyluck>
or I might have misinterpreted what he said.

[00:51:07.0000] <hsivonen>
MikeSmith: I think I should gather some stats about how validator.nu is used. Stuff like HTML5 versus HTML 4 usage numbers.

[00:51:21.0000] <MikeSmith>
yeah would be useful

[00:51:22.0000] <hsivonen>
MikeSmith: Any ideas what in particular I should measure?

[00:51:41.0000] <MikeSmith>
no particular ideas at the moment

[00:51:46.0000] <MikeSmith>
will try to think of some

[00:51:58.0000] <hsivonen>
MikeSmith: I'm not planning any fancy back end: just an object in RAM that can dump its data to a HTTP, so no disk I/O is needed.

[00:52:07.0000] <MikeSmith>
OK

[00:53:03.0000] <hsivonen>
My connectivity to Mozilla's important infrastructure is broken today, so I better take advantage of the validator hacking opportunity.

[00:53:23.0000] <MikeSmith>
hah

[00:53:27.0000] <MikeSmith>
yeah please do

[00:55:17.0000] <payman>
kennyluck: https://plus.google.com/107429617152575897589/posts/WwVmKNjcZ4p

[00:55:35.0000] <kennyluck>
payman, thanks!

[00:57:57.0000] <hsivonen>
MikeSmith: I think I should try to find out if people actually use custom Schematron

[01:01:05.0000] <annevk>
hsivonen: just measure all the settings?

[01:03:01.0000] <hsivonen>
annevk: it would be inappropriate to give a public dump of that

[01:03:10.0000] <hsivonen>
and I don't want to develop access-control

[01:03:46.0000] <annevk>
which would be wrong?

[01:03:57.0000] <hsivonen>
annevk: dumping the URLs of custom schemas

[01:05:07.0000] <annevk>
maybe you should hash them so you can at least count frequency?

[01:05:47.0000] <annevk>
but measuring first whether they're popular at all seems fine

[01:06:58.0000] <tantek>
/me just read that OAuth 2.0 and road to hell blog post and came upon Hixie's comment as well: http://hueniverse.com/2012/07/oauth-2-0-and-the-road-to-hell/#comment-42017

[01:09:03.0000] <annevk>
uhuh

[01:09:06.0000] <annevk>
WebSocket :(

[01:09:35.0000] <annevk>
although it didn't turn out all bad

[01:12:07.0000] <tantek>
I knew things about IETF were quirky and odd (e.g. the humming), but I didn't realize it was that broken (and susceptible to enterprise hijacking).

[01:16:11.0000] <ashemedai>
So with MathML in HTML5, am I supposed to use &Integral; or just use Unicode and put ∫ ?

[01:24:04.0000] <annevk>
up to you

[01:35:17.0000] <ashemedai>
annevk: Shouldn't cause any problems that you know of at least? I mean, semantically &Integral; might be passing along some additional info, not sure how UAs deal with this.

[01:39:24.0000] <hsivonen>
ashemedai: &Integral; doesn't pass any additional semantics

[01:41:37.0000] <ashemedai>
hsivonen: Brilliant, cheers.

[01:44:02.0000] <ashemedai>
Wonder if I should try compiling https://github.com/w3c/tidy-html5 under Windows one of these days.

[01:50:51.0000] <annevk>
ashemedai: the way it works is that &Integral; ends up as ∫ in the tree

[01:51:05.0000] <annevk>
ashemedai: so it's purely a source text affair and the source text is not exposed

[01:56:33.0000] <Stevef_>
hsivonen: your focus outline proposal is a crowd pleaser - thanks for filing and ccing me

[01:56:55.0000] <ashemedai>
annevk: gotcha

[01:59:05.0000] <ashemedai>
Microsoft's Math Input Panel loves to add a lot of extra <mrow>s in places it need not to.

[02:06:59.0000] <ashemedai>
Heh, even in a function specification like f(u,v) it made , a <mo>

[02:08:28.0000] <ashemedai>
Quite interesting how MathML can show the same rendering with quite varied MathML constructs.

[02:21:25.0000] <odinho>
I like that accessibility blog post.

[02:23:07.0000] <jgraham>
I would like it more if my experience wasn't that certain people it praises do all the things it criticises

[02:24:02.0000] <odinho>
jgraham: lol, you may replace them with "foo" and "bar" whilst reading.

[02:28:23.0000] <ashemedai>
Oh hey

[02:28:40.0000] <ashemedai>
The Wikipedia code has been updated. That MathJax change is now live on the main servers.

[02:28:50.0000] <ashemedai>
When logged in, under Appearance in Preferences

[02:50:42.0000] <Stevef_>
jgraham: I think you mean a certain person and I think it is important to draw a distinction between how people act in particular contexts (work) or when ineteracting with particular communities

[02:57:12.0000] <jgraham>
Stevef_: Yes, one person.

[02:58:49.0000] <Stevef_>
jgraham: John foliot can be a very positive force in web standards and bringing together the different communities  open web camp for example http://openwebcamp.org/

[02:59:58.0000] <Stevef_>
jgraham: and like myself and many other people in the web standards community he can be disruptive at times

[03:01:55.0000] <hsivonen>
Stevef_: I filed the focus thing out of my personal need--not to please a crowd. I CCed you, because I mentioned it to you earlier.

[03:02:41.0000] <Stevef_>
hsivonen: fine, what i meant was it is a good idea and was welcomed

[03:02:52.0000] <hsivonen>
Stevef_: ok

[03:03:15.0000] <Stevef_>
hsivonen: my choice of words may not have been the best

[03:03:36.0000] <jgraham>
Stevef_: He must have done something worthwhile because a number of people seem to respect him. That doesn't change the fact that the attitudes and behaviours that he is quoted as advocatin in the article are not those that I have experienced him employing himself

[03:07:53.0000] <Stevef_>
jgraham: yes I understand that, which is why i mentioned the work distinction as I believe he does exhibit the qualities he is quoted as advocating in his professional accessibility work.

[03:13:20.0000] <hsivonen>
Stevef_:         PARSER_XML_EXTERNAL ("Parser set to XML with external entities");

[03:13:23.0000] <hsivonen>
doh

[03:13:41.0000] <hsivonen>
Stevef_: is HTML working group work nonprofessional?

[03:20:12.0000] <Stevef_>
hsivonen: involvement is on a voluntary basis for many of us I think not affiliated with work. is what I meant. Also in my job (professional capacity) the clients I work with are on the whole not  from the standards communities, many have no idea it exists, adobe and google being the exceptions

[03:20:58.0000] <Stevef_>
maybe microsoft, but who knows what they know

[03:21:44.0000] <hsivonen>
ok

[03:25:13.0000] <jgraham>
(I think it seems very strange to say that one behaviour is needed to get positive action for accessibility in a "professional" context and then display quite different behaviour when trying to achieve similar things in a standards context)

[03:25:29.0000] <Stevef_>
hsivonen: i generally work with web app developers who build interfaces with tables and buttons using links and have little clue as to how to provide a label for a form control

[03:25:56.0000] <jgraham>
But what I actually wanted to ask is whether there is a good reason to fire beforescriptexecute after getting the source of the script rather than before

[03:26:12.0000] <jgraham>
In particular for inline scripts where beforescriptexecute changes the script text

[03:27:06.0000] <jgraham>
(Gecko gets thins "wrong")

[03:27:12.0000] <jgraham>
hsivonen: ^

[03:28:36.0000] <Stevef_>
jgraham: its a 2 way street the different behaviour is often exhibited due to different responses/reactions recieved

[03:30:53.0000] <Stevef_>
also note that it is not binary standards/professional , its interaction with particular people or particular subgroups that makes a difference for many people

[03:33:23.0000] <jgraham>
Stevef_: I still don't understand why you would believe that acting in a certian way will be best at effecting change in one context and acting in a very different way will be the best way to effect the same kinds of change in a very similar context

[03:34:37.0000] <hsivonen>
jgraham: sicking's design. He would be a better person to answer.

[03:35:10.0000] <hsivonen>
jgraham: If Gecko and the specs disagree, then Hixie disagreed with Gecko.

[03:36:50.0000] <Stevef_>
jgraham: I don't, i am not talking about what is the best strategy for reaching one's goals I am talking about how other peoples behaviour can effects how one behaves, not in a good way

[03:37:41.0000] <jgraham>
hsivonen: It seems that Gecko complies with the spec in the src case

[03:37:56.0000] <jgraham>
Which makes sense, sort of

[04:20:41.0000] <Stevef_>
on another note, anyone interested in an MSAA/IA2/UIA accessibility object inspection tool for windows, developed specifically for testing web pages, can try out http://www.paciellogroup.com/blog/2012/07/latest-aviewer-alpha-2012/ feedback welcome

[04:35:24.0000] <smaug____>
Stevef_: make sure to file browser bugs about the issues

[04:36:23.0000] <Stevef_>
smaug____: yeah am in contact with various browser acc engineers on the bugs

[04:38:02.0000] <Stevef_>
smaug____: main issue is that chrome does not currently expose API interface to enable the tool to get at innerhtml

[04:38:12.0000] <smaug____>
/me knows nothing about chrome

[04:39:05.0000] <smaug____>
but you mentioned few Firefox issues

[04:39:12.0000] <smaug____>
and would be good to have bugs filed

[04:39:41.0000] <Stevef_>
smaug____: i think the issues are on our/NVDAs side

[04:39:44.0000] <smaug____>
Stevef_: just curious, how does Opera behave ?

[04:39:52.0000] <smaug____>
Stevef_: ah

[04:39:56.0000] <odinho>
Prolly misbehaves :P

[04:40:32.0000] <Stevef_>
smaug____:last i checked, does not expose anything useful, unfortunately

[04:41:27.0000] <smaug____>
odinho was right then :)

[04:43:14.0000] <Stevef_>
bits of the browser chrome, none of the HTML

[04:51:25.0000] <odinho>
Stevef_: So, if Opera should be better, -- is there a progressive way to get there? I mean, like small feature A making it lots better, and then many more small fixes to make it progressively better. -- Or is it all requiring a big up-front investment?

[04:51:44.0000] <odinho>
I have no idea over Opera's accesibility, or even what tools expect to extract.

[04:52:35.0000] <Stevef_>
odinho: i wrote on the subject here: http://www.paciellogroup.com/blog/2012/06/open-letter-to-opera-re-accessibility-support/

[04:53:43.0000] <Stevef_>
odinho: to be honest, Opera should look at implementing accessibility in its mobile browsers first (where they have marketshare)

[04:55:14.0000] <Stevef_>
odinho: there was an aborted attempt 5 or so years ago: http://my.opera.com/desktopteam/blog/2007/07/27/screen-reader-support

[04:55:34.0000] <odinho>
Stevef_: Yes, I did see that post actually. But it seems like real tasks to do stuff like that. I am doing spec stuff, tests etc, but I like doing small bug fixes on the side, so I'm always looking for really easy stuff to fix :P

[04:55:47.0000] <Stevef_>
adinhO; and belive there was some work on mac accessibility

[04:57:39.0000] <Stevef_>
odinho: right, so if you want something small scale (relaitively) that would improve access for keyboard users see discussion with hallvord: http://www.paciellogroup.com/blog/2012/07/w3c-html-and-whatwg-html-split-affect-on-html-accessibility/comment-page-1/#comment-19628

[05:07:02.0000] <odinho>
Stevef_: Yeah, that one seems a bit chewable. At least after a while :-) Found the bug about it, seems like it's a very old bug (created 2001).

[05:07:21.0000] <odinho>
Stevef_: chaals made an extension for it as a stop-gap, https://addons.opera.com/en/extensions/details/key-title/ -- has been downloaded 905 times.

[05:08:44.0000] <Stevef_>
odinho: yeah I saw chaals extension, its been an issue since forever in all browsers, i have been banging on about it for at least the mid 2000's...

[05:09:57.0000] <Stevef_>
/me hope to see issues fixed before retirement...

[05:10:12.0000] <odinho>
Stevef_: Hehe, when is that?

[05:10:38.0000] <Stevef_>
odinho: 20 years or so :-)

[05:12:28.0000] <odinho>
Stevef_: Hehe, if we haven't fixed at least the tooltip thing in that time, I will sit down and do it.

[05:12:55.0000] <Stevef_>
odinho: lol will hold you to that

[05:13:02.0000] <odinho>
^_^

[05:19:49.0000] <Stevef_>
odinho: also Opera could look at improving support for screen magnifiers on the desktop and also look at providing further features for low vision users who don't use AT, there are many of them, opera could corner the market

[05:50:06.0000] <sedovsek>
I wrote about CSS Flexbox Layout model, if anyone is interested > http://galjot.si/css-flexbox

[05:58:49.0000] <jgraham>
sedovsek: *dynamically

[05:59:55.0000] <sedovsek>
jgraham: Thanks. I fixed it.

[06:13:18.0000] <david_carlisle>
ashemedai:   The comma should be in an <mo> (if it is explicit at all) or you could use mfenced and make the comma implict.  All distinct character tokens in MathML are in some element or other

[06:13:59.0000] <jgraham>
In other news MathML is rather verbose :p

[06:18:27.0000] <david_carlisle>
jgraham: <word><letter>w</letter><letter>e</letter></word><space> </space><word><letter>l</letter><letter>i</letter><letter>k</letter><letter>e</letter></word><space> </space><word><letter>o</letter><letter>u</letter><letter>r</letter></word><space> </space><word><letter>t</letter><letter>a</letter><letter>g</letter><letter>s</letter></word>

[06:19:33.0000] <jgraham>
:)

[06:46:34.0000] <payman>
sedovsek: "back in in"

[06:46:50.0000] <sedovsek>
:P

[06:53:27.0000] <payman>
sedovsek: s/Another two/Other two

[06:55:01.0000] <sedovsek>
payman: Thanks! Fixed it.

[06:55:27.0000] <payman>
didn't fix the first issue though :)

[06:56:27.0000] <sedovsek>
Oh, yea.

[06:56:35.0000] <sedovsek>
Thanks again.

[06:56:47.0000] <payman>
np

[07:18:38.0000] <zewt>
matjas/annevk: definitely a bug--it doesn't matter that the character is represented as a surrogate pair; it's still one character/one codepoint

[07:19:06.0000] <matjas>
implementation details exposed to the user etc etc

[07:19:19.0000] <matjas>
either way, not really a very important issue indeed

[07:19:41.0000] <zewt>
and being able to backspace over half of the surrogate pair (which you can in chrome and FF) is definitely a bug--you should never be able to create a broken surrogate pair (at least not with a simple UI like that--not sure what should happen in more obscure cases, like pasting one in)

[07:20:22.0000] <matjas>
so who’s gonna file the bugs? /me ducks

[07:20:35.0000] <zewt>
the fact that non-BMP characters are so rarely used is the only reason the utf-16 fustercluck is even tolerable :)

[07:20:42.0000] <matjas>
haha

[07:21:10.0000] <matjas>
i bet no one actually uses astral symbols in passwords

[07:21:10.0000] <zewt>
personally I wish we'd just change JS to ucs-4, under the logic that although we'd break some sites ... it'd probably fix far more :)

[07:21:25.0000] <matjas>
heh

[07:21:26.0000] <zewt>
(unfortunately, the sites that it would break would be the few that it actually matters for, since after all they spent the time making it work)

[07:25:47.0000] <smus>
hey, would like to polyfill a CSS function, but it appears that there's no consistent CSS OM API so the only way this can be done is by parsing CSS via JS (yuck!). can someone point to a relevant spec?

[07:26:36.0000] <beverloo>
a spec for what?

[07:26:41.0000] <paul_irish>
(because CSSOM wont reveal anything it cant succesfully parse..)

[07:26:52.0000] <beverloo>
the CSS OM is specified here: http://dev.w3.org/csswg/cssom/

[07:26:53.0000] <paul_irish>
well.. ideally CSSOM had a UnknownRule definition

[07:27:10.0000] <smus>
right. thx paul

[07:27:12.0000] <paul_irish>
so when it failed we could at least hack up that part with regex

[07:27:16.0000] <smus>
missing piece!

[07:27:51.0000] <paul_irish>
aside from that, the browser could expose essentially the stylesheet AST

[07:27:52.0000] <SimonSapin>
parsing CSS from JS? https://github.com/tabatkins/css-parser

[07:28:16.0000] <smus>
so if ur like background: -webkit-image("foo.webp", "foo.jpeg"); the parser just ignores that rule entirely. would be nice if it parsed as unknown

[07:28:21.0000] <beverloo>
the css syntax specifically defines that unknown properties/at rules need to be ignored

[07:28:26.0000] <paul_irish>
SimonSapin: the goal is a polyfill you'd use in production... and you can't ask folks to parse css in js clientside for real.

[07:29:15.0000] <paul_irish>
beverloo: sure, but you can see the problem. we can't polyfill to extend css when unknowns are discarded.

[07:29:23.0000] <beverloo>
the css syntax (including the rules for ignoring it) is specced here: http://www.w3.org/TR/css3-syntax/

[07:29:50.0000] <beverloo>
well, you'd have to polyfill the hypothetical CSSUnknownRule too :p

[07:30:00.0000] <paul_irish>
 hey now

[07:30:20.0000] <paul_irish>
i dont need to polyfill try/catch

[07:30:33.0000] <smus>
http://www.highdots.com/forums/javascript/retrieving-unknown-css-rules-48513.html

[07:30:50.0000] <smus>
specifically, http://lists.w3.org/Archives/Public/www-style/2003Oct/0347.html

[07:31:09.0000] <smus>
why drop unknowns?

[07:31:18.0000] <paul_irish>
omg. bert.

[07:31:42.0000] <beverloo>
it sounds like you should take this to www-style

[07:32:33.0000] <smus>
k, any known reason to drop unknowns?

[07:32:40.0000] <smus>
(so i can present a balanced case)

[07:33:03.0000] <paul_irish>
smus: its probably somewhere in this 2003 era archives of www style

[07:33:17.0000] <paul_irish>
beverloo:  yeah certainly. it'd be fun to get UnknownRule back

[08:30:01.0000] <smus>
cool, so <img srcset> doesn't seem to work in any browser. is there an existing reference implementation? http://smus.com/m/srcset/

[08:31:40.0000] <jreading>
iOS 6 mobile safari

[08:33:32.0000] <dglazkov>
good morning, Whatwg!

[08:34:00.0000] <SimonSapin>
good afternoon, dglazkov

[08:43:04.0000] <Wilto>
jreading: Any chance you have a link to the commit/some details handy? Curious as to whether it just covers resolution or the whole extended microsyntax.

[09:14:50.0000] <Velmont>
smus: ... You have an error in your syntax as well, it's with comma.

[09:15:48.0000] <smus>
Velmont: thx!

[09:30:39.0000] <jreading>
Wilto: http://trac.webkit.org/changeset/111637

[09:31:46.0000] <jreading>
and a polyfill: https://github.com/SunboX/-webkit-image-set-polyfill

[09:31:47.0000] <jreading>
:/

[09:37:19.0000] <miketaylr>
sweet README

[09:37:37.0000] <miketaylr>
accidentally the code

[09:38:17.0000] <jreading>
polyfills are the new domain squatting

[09:45:33.0000] <jreading>
i misspoke about srcset, it's "image-set", sorry for the confusion. *ducks*

[09:48:35.0000] <Hixie>
hsivonen: hey sometimes gecko and the spec disagree just because i made a mistake, or because mozilla people told me they disagree with gecko :-)

[11:43:22.0000] <Hixie>
hmmm

[11:43:48.0000] <Hixie>
i wonder how to spec the autocomplete="" field's grammar and parsing

[11:44:28.0000] <Hixie>
<empty> | on | off | [section] [subsection] ( generic-field | [contact-type] contact-field )

[11:44:47.0000] <Hixie>
i guess to parse it i split on spaces then go backwards through the tokens

[11:47:12.0000] <Hixie>
and to define the conformance criteria for authors, maybe an ordered list of tokens, and i just say what each token has to be or something

[11:47:15.0000] <Hixie>
ok

[12:06:18.0000] <Wilto>
jreading: Man, that’s gonna be a tough one to polyfill.

[12:38:15.0000] <tantek>
Hixie, sounds horribly complicated

[12:39:48.0000] <Hixie>
i'm certainly open to alternatives :-)

[13:02:52.0000] <zewt>
okay, the chrome thing where no form of refresh will make it reload css background: url() images is pretty much horrible

[13:05:49.0000] <TabAtkins>
I think Inspector is gaining some Refresh options soonish, so you can do hard refreshes or even full cache flushes.

[13:06:30.0000] <zewt>
that's what control-f5 is supposed to do, it just appears to not propagate to css resources

[13:07:48.0000] <TabAtkins>
That actually sounds like a bug.  File on us?

[13:26:08.0000] <annevk>
paul_irish: TabAtkins and I at some point came up with data-* CSS properties that would be exposed in the CSSOM

[13:26:31.0000] <annevk>
paul_irish: as opposed to exposing every property which would kinda ruin forward compatibility

[13:35:48.0000] <TabAtkins>
annevk: That's the CSS Variables spec.

[13:36:08.0000] <annevk>
ah yeah, it morphed into that

[13:36:20.0000] <TabAtkins>
Which we'll finalize for WD and soonafter LC next month at the meeting.

[13:36:33.0000] <TabAtkins>
But yeah, exposing every property is a non-starter.

[13:36:46.0000] <TabAtkins>
Like custom elements without an x- prefix or something.

[14:08:17.0000] <jgraham>
TimBL!

[15:41:33.0000] <Hixie>
TabAtkins: thanks for answering that intern's question

[15:41:40.0000] <Hixie>
TabAtkins: i was like "wtf are you talking about"

[15:44:15.0000] <TabAtkins>
Hixie: Haha, np

[15:44:22.0000] <TabAtkins>
I want to see the feature implemented!

[15:44:31.0000] <TabAtkins>
And now I have some questions to post to the list about it.

[15:50:42.0000] <Hixie>
i like your "just implement the spec" thing

[15:50:55.0000] <Hixie>
i always feel weird saying that :-)

[15:54:20.0000] <TabAtkins>
Why?


2012-07-28
[19:28:41.0000] <Hixie>
TabAtkins: because it undermines my ideal, namely that the standard should be good enough and be a close enough match to the implementors' desires that they would want to implement it without prompting

[01:04:07.0000] <annevk>
:(

[01:04:15.0000] <annevk>
DreamHost is still at MySQL 5.1

[01:04:20.0000] <annevk>
meaning http://golem.ph.utexas.edu/~distler/blog/archives/002539.html is still broken

[01:27:02.0000] <annevk>
so yeah, I tried putting 💩 in a blog entry and it failed miserably

[02:40:46.0000] <matjas>
annevk: surely they must be planning to upgrade to MySQL 5.3.3+ sometime?

[02:41:09.0000] <matjas>
annevk: back when my server had an outdated mysql, I just took care of it with some PHP string replacements

[02:41:23.0000] <matjas>
`{tetragram}` became `𝌆`, etc… annoying but at least it works

[02:42:25.0000] <matjas>
whenever someone left a comment using any other astral symbol, i would have to go in and add another dictionary entry – but luckily that never happened

[02:44:00.0000] <matjas>
also, converting your dbs and tables to utf8mb4 isn’t as simple as that blog post makes it seem

[02:44:13.0000] <matjas>
VARCHAR(191) is the new VARCHAR(255), for example

[02:46:18.0000] <matjas>
(for HTML, you could of course use character references)

[02:47:59.0000] <matjas>
(didn’t we talk about this before in #whatwg?)

[02:57:06.0000] <annevk>
matjas: we did, but I started looking into updating my stuff today

[02:59:38.0000] <annevk>
matjas: the comment in the MySQL documentation makes it seem as though the argument to CHAR is counting characters

[03:20:48.0000] <matjas>
i’d suggest using utf8mb4_unicode_ci instead of utf8mb4_bin, though

[03:20:59.0000] <matjas>
why would anyone want to use utf8mb4_bin?

[03:21:20.0000] <matjas>
i might do a write-up on how i converted my stuff

[03:21:57.0000] <matjas>
in case it helps anyone

[03:46:26.0000] <annevk>
matjas: I'm not actually using MySQL for anything other than dumb storage so it doesn't matter much for me I guess

[03:46:29.0000] <annevk>
at least not at the moment

[03:47:08.0000] <annevk>
matjas: but how to convert existing stuff and how to set up new stuff is always good to have given the rotten defaults :/

[04:23:54.0000] <Velmont>
matjas: So it's on mainline, I guess you got my email(?)

[07:06:49.0000] <matjas>
Velmont: huh?

[08:21:04.0000] <Velmont>
matjas: The var var, for, while; fix :-)

[08:45:45.0000] <zewt>
i wish all of the "reflected the content attribute of the same name" bits would link to said bits, since that's where I always land

[08:48:36.0000] <zewt>
http://www.jwz.org/blog/2012/07/non-bouncy-layouts/ i'd have expected setting img @width and @height to do this, but it doesn't :(

[08:49:47.0000] <zewt>
eg. saying <img width=800 height=200 style="width: 100%; height: auto;"> to always layout as if it had an 800x200 image loaded

[08:51:37.0000] <zewt>
(probably throw in display: inline-block for completeness, not that that helps)

[14:33:12.0000] <matjas>
Velmont: ah, got confused by your IRC nick, sorry — yeah I got the mail, thanks!


2012-07-29
[22:28:35.0000] <heycam>
anyone else having trouble pushing to dvcs.w3.org? I get "HTP Error 500: No usable temporary directory found in ['/tmp', '/var/tmp', '/usr/tmp', '/home/www-data']"

[22:28:59.0000] <heycam>
MikeSmith ^

[22:39:05.0000] <heycam>
(I've mailed sysreq)

[03:19:06.0000] <Velmont>
matjas: Ah, yes, the big confusion. I use odinho at work, and Velmont from home. :P

[03:45:32.0000] <hoolter1>
is there a simple way to take an image dropped onto the  page from the filesystem by html5 dnd and then just load it into an image tag?

[04:51:26.0000] <MikeSmith>
heycam|away: sorry was offline all day. Will take a look now

[08:08:27.0000] <Velmont>
hoolter: Well, DND api is not simple. It was not designed, it was built on top of IE's DND...

[08:08:47.0000] <Velmont>
hoolter: So not a "simple" way, but there is a very obvious way when just using the API.

[08:11:05.0000] <hoolter>
Velmont: i figured it out from remy's demo.  thanks though.  and did not know that - interesting.

[08:19:58.0000] <hoolter>
is <figure> block or inline?

[15:34:37.0000] <heycam>
MikeSmith, thanks. I just tried again now and it seems to be pushing fine.

[15:34:53.0000] <heycam>
MikeSmith, also no apologies needed for being offline all day on the weekend. :)

[15:43:45.0000] <bencc>
is it reasonable for a websocket client and server to remember the last message and resend it on reconnect?

[15:44:30.0000] <bencc>
or are websockets reliable enough that in case of failure something really bad happened and I need to recreate everything

[15:47:00.0000] <roc_>
the latter

[15:48:04.0000] <bencc>
roc: I did some testing. jsonp and xhr-polling fails sometimes and even RTMP flash connection

[15:48:11.0000] <bencc>
but Websocket doesn't

[15:48:13.0000] <bencc>
how come?

[15:49:03.0000] <roc>
I don't know

[15:49:37.0000] <bencc>
thanks


2012-07-30
[17:17:05.0000] <MikeSmith>
heycam: so I couldn't figure out specifically what might have caused that problem

[17:17:27.0000] <MikeSmith>
other than the file system just running out of space completely

[17:17:36.0000] <heycam>
MikeSmith, ok np. I'l l let you know if it comes up again.

[17:20:01.0000] <MikeSmith>
I think we just don't have enough space on that machine. There isn't a separate mount for /tmp on it. It's on the same mount as the main filesystem, which is only 4.3G and 93% full. So there less than 300MB free on it right now.

[17:20:11.0000] <MikeSmith>
anyway I'll talk to the systems team about it

[17:20:53.0000] <MikeSmith>
heycam: I don't mind getting pinged on the weekend, btw. But if you need to ping me and I don't seem to be around on IRC, feel free to e-mail at mike⊙wo

[17:21:37.0000] <MikeSmith>
If you do I'll get a notification on my mobile right away, and I can actually ssh into that machine from there

[17:22:11.0000] <heycam>
MikeSmith, righto, thanks

[17:53:16.0000] <bk>
hi

[17:55:38.0000] <bk_1>
is anyone here

[17:55:47.0000] <bk_1>
hmm

[18:01:19.0000] <|bk|>
anyone knows a good way to start with xhtml

[18:01:24.0000] <|bk|>
not sure where else to ask

[18:09:09.0000] <zewt>
"don't"

[18:11:46.0000] <GPHemsley>
<audio>/<video> should be able to contain <embed> as a fallback, right?

[18:34:06.0000] <GPHemsley>
I'm told that fallback content is only used by browsers that don't actually support <audio>/<video>

[18:34:49.0000] <GPHemsley>
But if the browser doesn't support the media type of the actual file, shouldn't it be able to treat it as it would <embed>?

[18:57:19.0000] <zewt>
GPHemsley: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2010-February/025028.html

[18:58:08.0000] <zewt>
nonsense, in my opinion (codec compatibility will *never* become a non-issue, so long as new codecs are being deployed that not everyone has)

[19:00:15.0000] <GPHemsley>
zewt: I didn't find the argument particularly convincing, particularly from a user experience point of view

[19:18:21.0000] <Hixie>
zewt: new video codecs come out even more rarely than new image formats

[19:18:43.0000] <Hixie>
(which have changed like, once, in the history of the web)

[19:59:57.0000] <zewt>
Hixie: what? they come out far, far more frequently (i'm watching stuff on my htpc right now that I had to buy an htpc to watch, because it's so new none of my set-top stuff can play it)

[20:00:42.0000] <zewt>
(namely, 10-bit h264, which both requires a lot of CPU and is not decode-compatible with regular h264)

[20:02:43.0000] <zewt>
(or whatever lower profile of h264 is widely used; not familiar with the particulars)

[20:23:30.0000] <Hixie>
zewt: browser-supported codecs?

[20:23:49.0000] <Hixie>
zewt: sure, new image formats and video codecs come out all the time, but we're talking specifically about browser-supported ones

[20:24:07.0000] <Hixie>
(and so far we don't even have one that's uniformly supported)

[23:09:50.0000] <MikeSmith>
anybody familiar with geolocation-api implementations, I'm wondering if it's really possible in practice for a Web app to get old position data for a user from the browser position-object cache

[23:11:05.0000] <MikeSmith>
I thought I saw something before that suggested browser implementations only allow a Web app to access position information which the user has already authorized that app to access

[23:12:58.0000] <MikeSmith>
I also found http://www.w3.org/2011/09/07-geolocation-minutes.html#item06

[23:13:07.0000] <MikeSmith>
where the Geolocation WG discussed this

[23:13:28.0000] <MikeSmith>
"The wg consensus is that no API change is required to address this, but that UAs should perhaps consider protecting against sites receiving old position data shared with other sites"

[23:13:57.0000] <MikeSmith>
but I don't understand why they decided the spec should not have some explicit requirement with regard to this

[01:03:03.0000] <hsivonen>
MikeSmith: looks like validator.nu defaults to HTML5 + SVG 1.1 + MathML 3.0 + RDFa Lite 1.1 for text/html with <!DOCTYPE html> but to XHTML5 + SVG 1.1 + MathML 3.0 for application/xhtml+xml with XHTML root namespace

[01:03:11.0000] <hsivonen>
MikeSmith: that can't be intentional, can it?

[01:03:48.0000] <hsivonen>
looks like there's a bug in the preset file

[01:07:56.0000] <hsivonen>
looks like the CML validator has moved and no longer builds on Validator.nu

[01:35:10.0000] <MikeSmith>
hsivonen: yeah that bug in the presets is almost certainly my fault

[01:35:11.0000] <MikeSmith>
sorry

[01:41:39.0000] <hsivonen>
MikeSmith: I can fix it later today

[01:41:46.0000] <MikeSmith>
ok

[01:41:56.0000] <hsivonen>
MikeSmith: I'm almost done with the statistics feature

[01:42:01.0000] <MikeSmith>
excellent

[01:42:11.0000] <hsivonen>
I'll put it behind a JVM system property

[01:42:44.0000] <MikeSmith>
once that lands I can enable it in the W3C backend as well, if you think that would be useful

[01:43:10.0000] <hsivonen>
MikeSmith: ok. right now, the schema stats don't support the W3C list of presets

[01:43:18.0000] <hsivonen>
but that could be arranged

[01:43:57.0000] <MikeSmith>
yeah I could tweak it locally afterward for now if it came to that

[01:44:09.0000] <MikeSmith>
I mean locally on those server instances

[01:44:35.0000] <MikeSmith>
I think we currently have 4 hosts behind validator.w3.org/nu

[02:27:40.0000] <matjas>
http://whatwg.org/html/webappapis.html#dom-windowbase64-btoa why was btoa specced to not support code points above U+00FF in the input string? was that how UAs had implemented it?

[02:31:04.0000] <gsnedders>
Yes.

[02:31:36.0000] <matjas>
gsnedders: thanks

[02:31:55.0000] <gsnedders>
Not sure if all did, but certainly most.

[02:32:02.0000] <AryehGregor>
All, IIRC.

[02:32:06.0000] <AryehGregor>
/me wrote the spec for that

[02:32:12.0000] <matjas>
Too bad. I’m used to stuff above U+FFFF breaking, but not > U+00FF.

[02:33:04.0000] <matjas>
what’s the reason UAs had implemented it that way? is there a good one?

[02:33:10.0000] <AryehGregor>
matjas, basically, btoa() treats its input as a sequence of bytes, not a string.

[02:33:24.0000] <AryehGregor>
The base64 algorithm acts on bytes, after all, not characters.

[02:33:41.0000] <AryehGregor>
How would you have said it should work?

[02:33:47.0000] <AryehGregor>
What would you do with codepoints above U+FF?

[02:34:43.0000] <matjas>
rebuild http://software.hixie.ch/utilities/cgi/data/data in HTML + JS, for example

[02:34:59.0000] <matjas>
it’s possible only if you don’t use `btoa` but instead write your own method

[02:35:52.0000] <matjas>
dev expectation (at least for me, without checking the spec) was that you can just feed any string to btoa() and get the base64-encoded result back, but that’s not true

[02:36:10.0000] <AryehGregor>
matjas, I mean, how would you modify the btoa() algorithm so that it does something sensible with chars above U+00FF?  What should it do with them?

[02:36:22.0000] <AryehGregor>
I'm not sure there's anything to do with them that makes sense.

[02:37:23.0000] <gsnedders>
Encode as UTF-8, like Encode() does.

[02:37:26.0000] <matjas>
encode strings per UTF-8, then base64 encode that?

[02:37:36.0000] <matjas>
but maybe that’s just me

[02:38:01.0000] <AryehGregor>
I guess that might have been plausible if it was done from the beginning, but it's not backward-compatible now.

[02:38:18.0000] <AryehGregor>
At the time the algorithm was designed, UTF-8 by default would have been a lot weirder than now, I think.

[02:38:44.0000] <matjas>
so instead of e.g. btoa('foo♥bar') you have to use btoa(unescape(encodeURIComponent('foo♥bar')))

[02:39:29.0000] <matjas>
AryehGregor: Yeah. I was just wondering why browsers never implemented it that way, as it seems to make much more sense IMHO.

[02:39:48.0000] <AryehGregor>
matjas, does encodeURIComponent actually always use UTF-8?

[02:40:31.0000] <matjas>
AryehGregor: it does as per ES 5.1; not sure about implementations though

[02:40:34.0000] <matjas>
http://ecma-international.org/ecma-262/5.1/#sec-15.1.3.4

[02:40:49.0000] <gsnedders>
All impls do.

[02:52:02.0000] <matjas>
function base64encode(string) { return btoa(unescape(encodeURIComponent(string))); } function base64decode(string) { return decodeURIComponent(escape(atob(string))); }

[02:52:12.0000] <matjas>
there we go.

[03:10:53.0000] <jgraham>
Why do you need [un]escape?

[03:12:39.0000] <annevk>
what's the difference between encodeURI and encodeURIComponent?

[03:13:30.0000] <annevk>
ah

[03:13:34.0000] <annevk>
encodeURI does not escape #

[03:14:15.0000] <annevk>
oh and whatever is uriReserved

[03:49:58.0000] <annevk>
welcome to the WHATCG odinho

[03:50:50.0000] <hsivonen>
MikeSmith: http://validator.nu/stats.html http://html5.validator.nu/stats.html

[03:51:09.0000] <hsivonen>
in the latter case, the schema is always automatically chosen but is never reported as such

[03:51:48.0000] <[tm]>
hsivonen: ok

[03:52:49.0000] <hsivonen>
there's no server-side persintence

[03:53:11.0000] <hsivonen>
the stats could be persisted by running wget on stats.html via cron or something

[03:53:50.0000] <hsivonen>
from my point of view, the time window between validator redeployments should be enough to get interesting numbers

[03:54:13.0000] <hsivonen>
already two things surprise me: validations per second < 1

[03:54:31.0000] <hsivonen>
and the Web service APIs seem more often used than the Web UI

[03:58:02.0000] <hsivonen>
[tm]: OK if I fix the bug in w3c-presets.txt, too, even though the bug is latent there (has no effect, AFAICT)?

[03:58:19.0000] <hsivonen>
(the bug about the default schema for XHTML, that is)

[04:04:49.0000] <matjas>
jgraham: [un]escape is not really needed here (it is only needed if you want to do proper UTF-8 {en,de}coding), but using it results in much smaller base64-encoded strings, and makes it match the output of other base64-encoders

[04:07:17.0000] <jgraham>
matjas: Right, that makes sense

[05:31:23.0000] <[tm]>
hsivonen: ok by me

[05:32:17.0000] <[tm]>
interesting that the Web-services API is used more than the Web UI

[05:39:35.0000] <hsivonen>
when going to TPAC, is it a good or a bad idea to take a TGV from CDG to Lyon?

[05:41:23.0000] <AryehGregor>
<http://html5.org/specs/dom-parsing.html> is loading slowly.

[05:42:38.0000] <AryehGregor>
Oh, it seems whatwg.org isn't loading.

[05:42:45.0000] <AryehGregor>
That's why.

[05:43:38.0000] <hsivonen>
[tm]: I pushed a fix for w3c-presets.txt

[05:44:31.0000] <AryehGregor>
This doesn't seem to say that any escaping happens?  http://html5.org/specs/dom-parsing.html#concept-serialize-xml

[05:45:05.0000] <AryehGregor>
(for text nodes)

[05:45:19.0000] <AryehGregor>
That's surely a bug.

[05:49:05.0000] <AryehGregor>
/me files

[06:00:46.0000] <MikeSmith>
hsivonen: thanks

[06:05:44.0000] <MikeSmith>
hsivonen: the fact that 75% of the requests are from the Web-services client makes me wonder if many of those are coming from the same small set of clients

[06:06:37.0000] <MikeSmith>
in other words, if there are is a small set of users or third-party services that account for most of the requests

[06:07:20.0000] <MikeSmith>
with users of the "normal" Web-based direct use case being the other 25%

[06:13:26.0000] <hsivonen>
MikeSmith: it would be interesting further research to try to figure out how varied the IP addresses for the Web service requests are

[06:15:06.0000] <MikeSmith>
hsivonen: yeah

[06:19:02.0000] <MikeSmith>
hsivonen: incidentally the W3C systems team has deployed a mechanism that when it detects excessive traffic from a particular IP, shows a page that says,

[06:19:03.0000] <MikeSmith>
Excessive traffic pattern blocked

[06:19:04.0000] <MikeSmith>
Due to excessive traffic we are temporarily blocking your IP address, please rate limit your use of W3C's Validator.

[06:19:05.0000] <MikeSmith>
Your IP address will clear automatically.

[06:19:30.0000] <hsivonen>
MikeSmith: is this in a load balancer independent of the validator code?

[06:19:38.0000] <MikeSmith>
yes

[06:20:02.0000] <MikeSmith>
I don't know the details of the rate-blocking part of that, though

[06:20:16.0000] <MikeSmith>
nor do I know how long they plan to set the blocking period in production

[06:20:56.0000] <MikeSmith>
nor how many requests per minute you have to make before it will block you and show that message

[06:21:29.0000] <MikeSmith>
previously it did not show that message

[06:21:55.0000] <MikeSmith>
it would just return a 403

[06:23:00.0000] <hsivonen>
It's annoying that SNCF won't sell train tickets more than three months in advance and the end of TPAC is more than three months away.

[07:08:28.0000] <hsivonen>
also very annoying that SNCF sells out some trains that are almost 3 months away

[07:09:00.0000] <hsivonen>
so I booked the Paris Lyon train now and will book Lyon Paris once available

[07:09:25.0000] <hsivonen>
at this rate, the Paris-Lyon tickets would be gone by the time the Lyon-Paris tickets go on sale

[07:09:29.0000] <hsivonen>
crazy

[07:26:06.0000] <icaaq>
Hi, if I have a sentence that someone has to say to make an accomplishment, would it be appropriate to use the q-element then?

[07:29:30.0000] <annevk>
<q> is annoying, better use ""

[07:30:03.0000] <icaaq>
why is it annoying?

[07:31:50.0000] <karlcow>
icaaq: It is not annoying.

[07:32:57.0000] <annevk>
icaaq: because it sometimes and sometimes not renders quotes

[07:33:21.0000] <annevk>
I guess most browsers do render quotes nowadays, but not always those you want

[07:33:41.0000] <annevk>
and since you never want to style it you might as well use ""

[07:34:47.0000] <icaaq>
annevk: ok, but that I dont care about :)

[07:35:28.0000] <icaaq>
I'm more interested in the right semantics

[07:35:48.0000] <karlcow>
or 「」(lang:ja) or « quote » (lang:fr) depending on the language.

[07:35:48.0000] <karlcow>
The question was about semantics. So icaaq is asking if it is appropriate to markup the expression of someone with q.

[07:36:54.0000] <icaaq>
karlcow: a future quote, so to speak, that not has been said yet

[07:36:56.0000] <karlcow>
icaaq: http://dev.w3.org/html5/spec-author-view/the-q-element.html#the-q-element

[07:36:58.0000] <jgraham>
karlcow: The author is pretty much the only person who can tell what the right quote character is, so that seems like a disadvantage

[07:37:31.0000] <karlcow>
icaaq: is more about contexts than content

[07:37:49.0000] <karlcow>
does it come from elsewhere with regards to the text you are writing.

[07:39:43.0000] <karlcow>
Basically, you are saying: I'm putting this (text) in here (<q></q>) and it's coming from there (cite="<uri>")

[07:40:14.0000] <icaaq>
Yes, but it has not been said yet ;)

[07:40:45.0000] <karlcow>
icaaq: it's a provenance context, not a type of discourse. :)

[07:41:31.0000] <karlcow>
basically it is not about narrative

[07:43:05.0000] <icaaq>
it's this pull request I'm wondering about. https://github.com/mozilla/bedrock/pull/296

[07:44:14.0000] <karlcow>
he is right, no need of q there

[07:45:02.0000] <hsivonen>
[tm]: whoa! somebody uses the legacy XHTML+SVG+MathML schema! No one validates SVG, though. http://validator.nu/stats.html

[08:04:06.0000] <david_carlisle>
hsivonen:  some of them might have been me (but you can discount those:-)

[08:04:46.0000] <icaaq>
karlcow: ok, thanks!

[10:07:30.0000] <dglazkov>
good morning, Whatwg!

[13:10:37.0000] <tantek>
wow - in an example of the validator actually *not helping* looks like a lot of folks running wordpress were very confused by rel="category tag" being marked as invalid

[13:11:44.0000] <tantek>
e.g. http://wordpress.org/support/topic/wordpress-abuses-rel-tag

[13:15:35.0000] <tantek>
(only discovered today when someone created a placeholder page for rel-category_tag due to someone adding "category tag" as a *single* value to the rel-registry - since corrected)

[13:16:29.0000] <TabAtkins>
It's a category *and* a tag, right?  Not a "category tag"?

[13:17:33.0000] <Hixie>
hah, that thread is awesome

[13:17:45.0000] <Hixie>
i love it when people say "it's not finalised, so it can change at any time"

[13:18:18.0000] <Hixie>
as if everything in HTML4 is still in HTML today unchanged, and as if HTML4 not changing itself was a good thing :-)

[13:28:41.0000] <tantek>
for more: https://www.google.com/search?q=rel-category+tag

[13:35:56.0000] <tantek>
TabAtkins, exactly. And as a result I had to write some text here: http://microformats.org/wiki/rel-category_tag

[14:43:32.0000] <Hixie>
bah, html sucks.

[14:43:48.0000] <Hixie>
/me finds himself wanting a non-rectangular cell in a table

[14:45:17.0000] <annevk>
sounds like a CSS problem

[14:45:26.0000] <edwardbc>
tables suck

[14:45:27.0000] <TabAtkins>
No, it's HTML.

[14:45:44.0000] <TabAtkins>
He means one that spans rows and columns without spanning the full cross of them.

[14:46:03.0000] <TabAtkins>
Not a cell that is, by itself, non-rectangular a la CSS Exclusions.

[14:46:36.0000] <annevk>
still a CSS problem

[14:46:46.0000] <TabAtkins>
I... don't understand.  It's content.

[14:46:49.0000] <annevk>
HTML does not dictate how to render a cell

[14:46:56.0000] <TabAtkins>
It's no more a CSS problem than @rowspan is.

[14:47:20.0000] <annevk>
I guess I misunderstood what Hixie meant by non-rectangular

[14:47:21.0000] <TabAtkins>
This is just a more complicated form of combining @rowspan and @colspan.

[14:47:40.0000] <annevk>
sleepy time

[14:47:43.0000] <TabAtkins>
That's why I tried to explain it. ^_^

[15:03:57.0000] <Hixie>
wow, i accidentally started a whole conversation of confusion there

[15:03:59.0000] <Hixie>
sorry about that

[15:06:33.0000] <Hixie>
hober: you editors had your meeting yet?

[16:07:57.0000] <UltraParadigm>
Will my webpage look right in windows xp ie if I use <header> <footer> instead of <div>

[16:13:58.0000] <zewt>
if you care about how your page looks in a browser, you're going to need to test in that browser :)

[16:15:48.0000] <UltraParadigm>
Yeah that's the best way but i think it's a pain in the butt get old version of IE.

[16:16:15.0000] <gsnedders>
UltraParadigm: Do you mean IE6/7/8?

[16:17:40.0000] <UltraParadigm>
I don't even use windows anymore except at work. We still have some xp at work u guess I can use to test.

[16:18:07.0000] <zewt>
if you aren't going to test in it yourself, you may as well give up, heh

[16:19:16.0000] <UltraParadigm>
yeah, alot of the world still uses xp unfortunatly, probably a good 30% at my work

[16:19:38.0000] <UltraParadigm>
Yeah gs

[16:20:33.0000] <UltraParadigm>
Gsnedders: ms wont let xp users upgrade to a html5 version if ie.

[16:21:08.0000] <gsnedders>
UltraParadigm: No, I meant, which IE versions do you care about?

[16:24:40.0000] <UltraParadigm>
7 and 8.

[16:25:24.0000] <gsnedders>
Okay, IE8 is fine. IE7 you can fool into working by creating such elements through scripts before the actual elements in the page.

[16:26:05.0000] <UltraParadigm>
Hopefully everyone has upgraded past 6 by now.  Although I know at my last job (cable vision tech support) they still had it. Too bad for them

[16:26:07.0000] <gsnedders>
<script>document.createElement("header");</script><header>foo</header> parses differently to just <header>foo</header>

[16:26:52.0000] <UltraParadigm>
Gsnedders: neat!

[16:53:00.0000] <Hixie>
aim:setstatus?message=wtf

[16:53:04.0000] <Hixie>
who designed this protocol

[16:53:45.0000] <abarth>
for real?

[16:54:29.0000] <Hixie>
http://webcache.googleusercontent.com/search?q=cache:http://dev.estage.aol.com/aimclient/OpenAIM182/technotes/AIMURL.html

[16:55:36.0000] <abarth>
AddBuddy?

[16:55:37.0000] <abarth>
:)

[16:55:39.0000] <Hixie>
also, jabber uses "jabber:" schemes for their namespace

[16:55:42.0000] <abarth>
is this actually used for something?

[16:55:52.0000] <Hixie>
abarth: apparently it's what the aim client supports

[16:56:23.0000] <Hixie>
i wonder how many people have yelled at the jabber folk for not buying into the distributed extensibility nonsense

[16:58:00.0000] <abarth>
I think we've got aim blacklisted in the external protocol handler\

[16:58:12.0000] <Hixie>
seems wise

[16:58:55.0000] <gavinc>
Hixie: errr, they did buy into it


2012-07-31
[17:03:42.0000] <zewt>
i've yelled at the jabber folk for their specs being a huge mass of "should" and "may"

[17:03:48.0000] <zewt>
total nightmare to do anything with

[17:05:31.0000] <zewt>
found their responsiveness to spec issues to be near-zero so i stopped bothering

[17:05:48.0000] <Hixie>
gavinc: "jabber:" isn't "distributed extensibility" (e.g. using dns), it's "landgrab namespacing" (what i advocate)

[17:06:04.0000] <zewt>
(seemed like a dead project, really, which is probably pretty much what xmpp is)

[17:08:45.0000] <gavinc>
Hixie: are we talking about the same thing? Jabber uses XML Namespaces

[17:09:05.0000] <TabAtkins>
Hixie's talking about the url protocol.

[17:09:29.0000] <gavinc>
aaaaahhh

[17:09:33.0000] <gavinc>
thank you :D

[17:10:21.0000] <Hixie>
actually you're both wrong :-)

[17:10:26.0000] <Hixie>
i'm talking about the url they use in their namespaces

[17:10:31.0000] <TabAtkins>
Oh!

[17:10:35.0000] <gavinc>
Oh!

[17:10:35.0000] <Hixie>
e.g. xmlns="jabber:client"

[17:10:44.0000] <TabAtkins>
JYNX

[17:11:10.0000] <Hixie>
which is pretty much exactly what i'd advocate, but nonetheless completely misses the point of distributed extensibility

[17:11:12.0000] <TabAtkins>
Or "jinx", rather.

[17:11:34.0000] <TabAtkins>
They're still going too far, imo.  Their ns should just be "jabber".

[17:12:01.0000] <Hixie>
well they don't really need namespaces at all, imho

[17:12:04.0000] <Hixie>
but that's another story

[17:12:11.0000] <TabAtkins>
Sure.

[17:12:18.0000] <Hixie>
(using xml for a stream-orientated protocol is pretty nutty)

[17:12:19.0000] <zewt>
there are a ton of namespaces in xmpp, though, not just those

[17:19:05.0000] <Hixie>
in other news, this table has had me as its bitch all day and i have only finally defeated it. http://www.whatwg.org/specs/web-apps/current-work/#autofill-field

[17:21:22.0000] <Hixie>
btw if there are any chrome qa people around who want something to do, the "M" in the "Mlle" example in that table is vertically aligned incorrectly

[17:21:36.0000] <Hixie>
i've been unable to determine why, but haven't looked closely

[18:03:02.0000] <jamesr_>
Hixie, what platform are you seeing bad alignment on?

[18:03:19.0000] <jamesr_>
you mean the "M<sup>lle</sup?" entry, right?

[18:11:15.0000] <Hixie>
jamesr_: mac and yes

[18:11:26.0000] <Hixie>
jamesr_: dev branch chrome

[18:13:30.0000] <jamesr_>
ah i see it now - it's a bit lower than it should be, right?

[18:55:08.0000] <gavinc>
ugh, balancing change/new features and doing what already works is very annoying.

[23:35:37.0000] <MikeSmith>
hsivonen: I wonder if the < 1 validations per second data you're seeing is due to the server hardware

[23:36:02.0000] <MikeSmith>
I'm seeing more like 1.9 validations per second

[23:41:37.0000] <hsivonen>
MikeSmith: unlikely. e.g. validating the front page of my site takes 44 milliseconds

[23:41:52.0000] <hsivonen>
MikeSmith: and the server is configured for concurrency

[23:42:07.0000] <MikeSmith>
OK

[23:42:37.0000] <hsivonen>
also, the CPU load is very low

[23:43:22.0000] <MikeSmith>
hsivonen: well we have some fairly beefy hardware behind the W3C service, and the data is pretty consistent across the majority of the hosts

[23:43:28.0000] <MikeSmith>
hsivonen: OK

[23:43:57.0000] <MikeSmith>
anyway, the W3C stats show those hosts consistently doing about 1.9 validations per second

[23:44:27.0000] <MikeSmith>
based on roughly 115,000 validations so for

[23:44:39.0000] <MikeSmith>
0.71 days

[23:45:31.0000] <hsivonen>
ok

[23:46:16.0000] <hsivonen>
the *completely* unused features so far: XHTML output, SVG schema, custom Schematron

[23:48:06.0000] <MikeSmith>
yeah that's as expected I guess

[23:48:29.0000] <hsivonen>
I expected someone to validate an occasional SVG file

[23:48:53.0000] <hsivonen>
I expected the legacy XHTML+SVG+MathML schema to be used less than the SVG schema

[23:49:03.0000] <MikeSmith>
OK

[23:49:15.0000] <shepazu>
I'm surprised that nobody's tried to validate SVG… is that what you're saying?

[23:49:16.0000] <hsivonen>
as for HTML vs. XML input, HTML is more than 99%

[23:49:20.0000] <hsivonen>
(Occupy XML)

[23:49:27.0000] <hsivonen>
shepazu: yes

[23:49:32.0000] <shepazu>
huh!

[23:49:44.0000] <MikeSmith>
shepazu: raw data is here: http://validator.nu/stats.html

[23:50:02.0000] <hsivonen>
and http://html5.validator.nu/stats.html

[23:50:29.0000] <shepazu>
oh, I believe you… I just went for years without an SVG validator, I never got in the habit myself (I just used XML validation when needed)

[23:51:57.0000] <shepazu>
9K validations doesn't seem like much… what time period, out of curiosity?

[23:52:36.0000] <hsivonen>
shepazu: see "Uptime in days" (which is < 1)

[23:53:08.0000] <shepazu>
oh

[00:24:33.0000] <hsivonen>
Underspecified feature request is underspecified: https://www.w3.org/Bugs/Public/show_bug.cgi?id=18439

[00:32:48.0000] <rniwa>
hsivonen: i can't even comprehend that sentence :(

[00:32:59.0000] <rniwa>
that is… the bug title.

[00:57:34.0000] <MikeSmith>
hsivonen: I don't know if you noticed yet but the settings for the filter feature of http://validator.keegan.st/ are persistent

[00:59:19.0000] <MikeSmith>
as in, once you filter out a certain class of errors, it's going to filter those out for you automatically from then on

[01:04:04.0000] <hsivonen>
I should pay more attention to the validator diagnostics. V.nu has been crashing due to an NPE that MikeSmith fixed over a month ago, because I had failed to redeploy after the fix

[01:04:20.0000] <MikeSmith>
damn

[01:04:22.0000] <MikeSmith>
sorry man

[03:04:56.0000] <hoolter1>
are opening and closing <html> tags required?

[03:05:01.0000] <hoolter1>
or are they implicit?

[03:06:12.0000] <jgraham>
Impolicit

[03:06:18.0000] <jgraham>
*Implicit

[03:06:35.0000] <hoolter1>
jgraham: cool.

[03:07:04.0000] <hoolter1>
jgraham: but then how can one specify doc language?

[03:07:10.0000] <hoolter1>
jgraham: any way?

[03:09:40.0000] <hoolter1>
iis it possible to read a file that your user DnD'd, then have the user open said file in, say, notepad, modify + save the file, and then for you to read the new changes without the user having to re-DnD the file again?

[03:30:30.0000] <hsivonen>
MikeSmith: it was my fault I didn't pay attention to diagnostics. the HTTP client had spammed me enough that it had taught me not to take the log4j emails too seriously :-(

[03:30:44.0000] <hsivonen>
MikeSmith: I guess I should upgrade to HttpClient 4.1

[03:30:53.0000] <hsivonen>
turns out the API has changed

[03:31:03.0000] <hsivonen>
not just a matter of changing package names

[03:48:39.0000] <AryehGregor>
hoolter, "doc language"?

[03:49:22.0000] <AryehGregor>
hoolter, for the latter question -- probably not.  That sounds like it would be a security vulnerability.

[03:49:39.0000] <AryehGregor>
Although maybe you can, I don't know.

[03:49:47.0000] <AryehGregor>
I take that back, yeah.

[03:49:52.0000] <AryehGregor>
I doubt browsers copy the file.

[03:49:56.0000] <AryehGregor>
Dunno how it works.

[03:50:01.0000] <AryehGregor>
What API are you using to read the file?

[04:19:10.0000] <hoolter>
AryehGregor: i've heard probably yes and probably no but i'm looking for someone who could tellme for sure.

[04:19:19.0000] <hoolter>
and i meant <html lang=en>

[04:19:43.0000] <AryehGregor>
hoolter, well, you can't omit the root tag if you want to specify any attributes on it.  You can still omit the end tag.

[04:20:00.0000] <AryehGregor>
If you want lang=en on the root element, you can't omit <html>, no.

[04:20:09.0000] <AryehGregor>
You could put it on <body>, although then you can't omit <body>.

[04:20:19.0000] <AryehGregor>
hoolter, with respect to the file question: what API are you using to read the file?

[04:32:54.0000] <hoolter>
AryehGregor: so I can have the following as my entire file and it's all totally fine then? |  <!doctype html><html lang=en><meta charset=utf-8><title>t</title><p>hello, world!</p>

[04:33:57.0000] <hoolter>
AryehGregor: IDK. I'm just wondering if it's possible at all.  See this video... just watch the first 25s of it or so, it's incredible.  but i can't reproduce it or wrap my mind around how it could work.  you know? www.youtube.com/watch?v=mY56fNmn2cE&feature=player_embedded

[04:34:37.0000] <AryehGregor>
hoolter, yes, that's valid.  Enter it into http://html5.validator.nu/ and see for yourself.  The shortest valid HTML document is: <!doctype html><title></title>

[04:38:45.0000] <hoolter>
AryehGregor: sweet.  do u mind checking the video?  It is fairly cool regardless of my question.

[04:39:06.0000] <AryehGregor>
hoolter, I have other things I'm doing right now.  If you have specific questions beyond what you've already asked, feel free to ask.

[04:39:07.0000] <hsivonen>
Now I remember why I went upgraded to HttpClient 4.1 yet. It's even more complicated to use than 3.1.

[04:39:23.0000] <hsivonen>
s/went/haven't/

[04:50:13.0000] <hoolter1>
AryehGregor: well i mean, is what i described possible using any api whatsoever that's thus yet been implemented by a version of chromium?

[04:50:42.0000] <AryehGregor>
hoolter1, 1) I don't know offhand, 2) what does Chromium have to do with it?  This is a standards channel.

[04:57:35.0000] <hoolter>
AryehGregor: well that's the one that was used to demo the feature in the video.  presumably it's at least a proposed standard of some sort.

[04:57:42.0000] <AryehGregor>
Probably.

[04:57:50.0000] <AryehGregor>
There are various file-related APIs floating around, I haven't been keeping track.

[05:09:01.0000] <hoolter>
AryehGregor: ah, ok.

[05:12:25.0000] <smaug____>
hoolter: yes, chrome implements some file system APIs

[05:12:43.0000] <smaug____>
I don't think anyone else is going to implement those APIs

[05:13:05.0000] <smaug____>
(but there will be some other file handling related APIs )

[07:55:45.0000] <zewt>
heh alt/title thread going on right when i hit a curious bit of nonsense

[07:56:20.0000] <zewt>
http://valgrind.org/docs/manual/dist.news.html has title="2. NEWS" on the whole thing, which apparently means that no matter where I idle the mouse, a "2. NEWS" tooltip pops up

[07:56:43.0000] <zewt>
(in Chrome, anyway)

[08:02:03.0000] <zewt>
now, if @title had simply been named @tooltip, nobody would be getting confused and thinking @title means something like "title for this section", heh

[08:02:35.0000] <zewt>
("but it might be displayed as something other than a tooltip!" ... that doesn't matter! heh)

[08:02:42.0000] <odinho>
heh. People.

[08:03:08.0000] <jgraham>
Maybe it shoudl have been named @banana

[08:04:00.0000] <zewt>
(because that would have been informative)

[08:06:13.0000] <jgraham>
Well it would stop people reading things into the name

[08:06:40.0000] <zewt>
reading things into the name is fine, if it's at least the right ballpark

[08:12:38.0000] <karlcow>
@hint ?

[08:13:15.0000] <zewt>
not like we can change it :)

[08:13:57.0000] <karlcow>
we can… it depends on the choices of life ;) being compatible or not with the deployed Web.

[08:14:14.0000] <karlcow>
being irrelevant as an art of life.

[08:16:29.0000] <zewt>
well, we already know the answer to that, heh

[08:45:04.0000] <Stevef>
some examples of title duplicating text content http://www.html5accessibility.com/tests/title-usage.html

[09:02:39.0000] <jgraham>
Hixie: Was there some change recently to the spec regarding pushState and base URLs? I (think I) remember you mentioning it, but I can't find it right now

[09:06:53.0000] <jgraham>
Ah http://krijnhoetmer.nl/irc-logs/whatwg/20120629#l-110

[09:08:23.0000] <Hixie>
jgraham: for some definition of "recently", yes

[09:08:46.0000] <Hixie>
jgraham: basically the concept of "current document's address" is entirely removed

[09:08:50.0000] <Hixie>
jgraham: it's a pretty big simplification

[09:09:03.0000] <Hixie>
jgraham: (unfortunately seems to screw opera for following the spec, sorry about that)

[09:10:25.0000] <jgraham>
Hixie: Well a month ago is pretty recently

[09:10:30.0000] <Hixie>
k :-)

[09:11:20.0000] <jgraham>
Well I was away for 2.5 weeks of that time as well :p

[09:11:32.0000] <jgraham>
So what's an observable change here?

[09:12:12.0000] <Hixie>
let me get you a test case, one sec

[09:12:44.0000] <Hixie>
http://hixie.ch/tests/adhoc/dom/level0/history/pushState/

[09:12:56.0000] <Hixie>
warning, these tests seriously confuse the back button

[09:13:06.0000] <jgraham>
Great, thanks

[09:13:12.0000] <jgraham>
Heh

[09:50:27.0000] <dglazkov>
good morning, Whatwg!

[09:50:41.0000] <edwardbc>
morning

[09:51:59.0000] <Hixie>
ok, i've added examples to the autocomplete field table. yay for timbl whose name has all the needed fields. :-)

[10:12:25.0000] <zewt>
a new (gross) one to me: someone sending mail with code in it ... with syntax highlighting

[11:13:07.0000] <hober>
Hixie: nope

[11:24:16.0000] <Hixie>
hober: k

[11:49:47.0000] <Hixie>
zewt: yt?

[11:49:58.0000] <Hixie>
zewt: re http://www.w3.org/mid/CABirCh-3Fa6xH=JA_j1b=d-7fGfrhAiG8WFUr=zbx0HDBGqEkQ⊙mgc -- what do you mean by "header line"?

[12:10:36.0000] <zewt>
Hixie: a line eaten by the "Header" section of the parser

[12:11:10.0000] <Hixie>
do you mean that this:

[12:11:13.0000] <Hixie>
WEBVTT

[12:11:16.0000] <Hixie>
language:fr

[12:11:22.0000] <Hixie>
timestamp --> timestamp

[12:11:24.0000] <Hixie>
foo

[12:11:34.0000] <Hixie>
...would treat "language:fr" as a "header" line?

[12:11:43.0000] <Hixie>
rather than as an "id" line?

[12:11:50.0000] <zewt>
yes, though it looks correct in that instance

[12:11:54.0000] <Hixie>
ok

[12:12:00.0000] <Hixie>
RESOLVED Working As Intended

[12:12:06.0000] <Hixie>
:-)

[12:12:52.0000] <zewt>
i mean if the "language:fr" line was actually something meant to be a "text track cue identifier", eg. step 31

[12:13:13.0000] <Hixie>
well the file is non-conforming either way

[12:13:21.0000] <Hixie>
this is just error handling

[12:13:42.0000] <Hixie>
for all we know, the line in question could be the cue text from some previous cue whose timestamp got croped

[12:13:45.0000] <Hixie>
cropped

[12:13:59.0000] <zewt>
huh? webvtt cue identifiers are in the syntax

[12:14:00.0000] <Hixie>
or it could be that the --> line is actually meant to be a comment

[12:14:15.0000] <Hixie>
the point is the file is invalid, so there's no way to know what the author meant

[12:14:32.0000] <Hixie>
we're just doing our best to salvage what we can from the mess

[12:14:46.0000] <zewt>
okay, looking at the syntax, the correct answer is "there should be a blank line before the first cue"

[12:18:06.0000] <Hixie>
right

[12:19:32.0000] <zewt>
(followed up on the thread to close that out)

[12:21:32.0000] <zewt>
(guessing that'll be a common mistake, but probably a harmless one since cue ids are probably uncommon)

[13:04:42.0000] <Areks|2>
Hello all.

[13:06:36.0000] <Areks|2>
Who made markup for tablet with android?

[13:08:17.0000] <Areks|2>
I use "Times New Roman" font in design, and all devices display it correctly except android.

[13:09:44.0000] <Areks|2>
Does anybody had same problem?

[13:11:28.0000] <Wilto>
Areks|2: Some older Androids will default to Droid Serif for all serifs and Droid Sans for all sans-serifs. It may not have Times New Roman installed.

[13:12:06.0000] <Wilto>
The #CSS channel is a more appropriate place for these sorts of questions -- hope that helps, though.

[13:14:48.0000] <Areks|2>
Wilto:  The problem is really interesting, because cyrillic symbols display correctly

[13:15:05.0000] <Areks|2>
Wilto: i'll try to ask on css, thanks

[13:15:27.0000] <Wilto>
No worries, man.