00:34
<Yuhong>
Personally I consider "HTML5" even as a buzzword a misnomer nowadays.
00:36
<Yuhong>
How many people using the buzzword know for example that canvas dates back to 2005?
00:45
<Yuhong>
Personally, I hope this can be finished: http://www.w3.org/wiki/Evolution
00:46
<Yuhong>
It will be helpful in proving the W3C process is flawed.
00:47
<Yuhong>
Hopefully, the TAG elections will get the right people in and help get this finished.
00:49
<Hixie>
if the people in question don't have time to update it now, why would they have time to update it after joining the mailing list and getting more mail than they do now?
00:55
<Yuhong>
Hixie: http://logbot.glob.com.au/?c=freenode%23whatwg&s=dec+31+2012&e=4+Jan+2013#c759102
01:01
<Yuhong>
Unrelated, but in case you didn't see it.
02:46
<rniwa>
any mozilla folks here?
02:47
<rniwa>
ideally someone working on css.
02:49
<heycam>
rniwa, I can try to help
02:49
<rniwa>
heycam: hi!
02:49
<heycam>
hey
02:49
<rniwa>
heycam: do you know if mozilla has any tests for CSS calc?
02:49
<heycam>
rniwa, we ought to
02:49
<rniwa>
heycam: if so, if you guys are planning to upstream those tests to W3C?
02:49
<rniwa>
heycam: i've checked http://test.csswg.org/source/contributors/mozilla/submitted/ but i don't see any tests there.
02:50
<rniwa>
heycam: mikelawther from google is implementing css calc in webkit
02:50
<heycam>
rniwa, people have been writing new tests in an upstreamable form, but I don't know if we have concrete plans to upstream existing tests
02:50
<heycam>
although it would be nice to
02:50
<rniwa>
heycam: and we would like to be abel to import your tests.
02:50
<rniwa>
heycam: ah, i see.
02:50
<heycam>
oh yes I met mikelawther at LCA last year
02:50
<rniwa>
heycam: cool :)
02:50
<heycam>
let me take a look at the existing tests
02:50
<rniwa>
heycam: could you keep him in the loop about your upstream effort?
02:51
<heycam>
to see how easy they would be to upstream
02:51
<heycam>
they are here btw: https://hg.mozilla.org/mozilla-central/file/tip/layout/reftests/css-calc/
02:54
<heycam>
rniwa, those tests seem reasonable self contained, so I think they shouldn't be hard to upstream
03:00
<rniwa>
heycam: great.
03:00
<rniwa>
heycam: is there any Bugzilla bug filed anywhere regards to this?
03:00
<heycam>
just filed one
03:00
<heycam>
and CCed you
03:00
<rniwa>
heycam: thanks.
03:00
<rniwa>
heycam: oh
03:00
<heycam>
couldn't find mike as a bugzilla user there though
03:01
<rniwa>
heycam: yeah.
03:01
<rniwa>
heycam: i don't think he has Mozilla bug account.
03:01
rniwa
waits for an email :)
03:01
<rniwa>
heycam: thanks a lot for the help.
03:01
<heycam>
no problem!
13:00
<annevk>
Hixie: ah yeah, I abuse @WHATWG now and then
13:02
<annevk>
Hixie: my excuse has been that the idea was that it could be used for whatever, but maybe we should keep it scoped nowadays
16:01
<annevk>
Man... I'm almost done with annotating parse errors...
16:01
<annevk>
No fun was had
17:59
<Hixie>
annevk: i don't have any problem with you abusing @whatwg, i trust your judgement
18:00
<Hixie>
annevk: i can't help but wonder how many of @whatwg's subscribers really care about the TAG though :-)
18:00
<Hixie>
then again, i'm baffled by why _you_ care, so... :-P
20:17
<Hixie>
hey bruce is asking me what my biggest mistake was for some article he's writing -- anyone got any suggestions? (i've already listed pushState, appcache, postMessage()'s security model...)
20:18
<Ms2ger>
Margin collapsing?
20:18
<Ms2ger>
The HTMLWG?
20:18
<WeirdAl>
Hixie - listening to developers for Acid 3 tests?
20:18
WeirdAl
being one of those idiot developers
20:18
<Hixie>
ooh, hadn't tought about margin collapsing and acid3, i was focused on the spec
20:19
<Ms2ger>
Acid2 and <p><table>
20:23
<Hixie>
I actually think <p><table> is fine, personally, but yeah, the acid tests in general
20:24
<Ms2ger>
You think the one quirk in the parser is fine?! :)
20:24
<annevk>
Hixie: localStorage?
20:24
<WeirdAl>
compared to the horrors that the acid tests required, Ms2ger? Yeah.
20:24
<Hixie>
i think the quirk sucks, but i think not implying </p> at <table> is what sucks, not the fact that we have a quirk so much :-)
20:25
<Ms2ger>
Oh, SVG fonts
20:25
<Hixie>
annevk: oh yeah, the storage mutex, good call
20:25
<annevk>
Hixie did not do SVG fonts :-)
20:25
<Hixie>
i did in acid3
20:25
<annevk>
ooh
20:25
<annevk>
Hixie: SGML comments ;-)
20:25
<Hixie>
was that acid2 or 3?
20:25
<Ms2ger>
Oh man
20:25
<Hixie>
2, rigt?
20:25
<annevk>
yeah
20:25
<Ms2ger>
2, I think
20:27
<annevk>
ah, Ms2ger had margin collapsing :)
20:27
<annevk>
That's a long list, why do we trust this guy again?
20:27
<Hixie>
nobody else is willing to do it?
20:27
<Ms2ger>
Why, prefer the CSSWG?
20:28
<annevk>
Ms2ger: uhuh
20:31
<annevk>
looking forward to reading that
20:31
<annevk>
Hixie: I guess that's part of it; would be so great to have actual competition rather than a copy & paste lobby
21:47
<annevk>
MikeSmith: what does https://twitter.com/katoken0215/status/287224151620587520 say?
21:52
<zewt>
that you're a woman, for starters...
21:53
<annevk>
whoa
21:54
<annevk>
TR/domcore looks ugly
21:54
<annevk>
dglazkov: please link to dom.spec.whatwg.org instead from the Shadow DOM spec
21:56
<annevk>
dglazkov: when does redistribution happen? how is this represented in implementations? is the DOM sprinkled with special nodes?
21:57
<annevk>
zewt: if that's all it says it's not very interesting
22:12
<zewt>
annevk: describing what you said in your last post, i think
22:13
<annevk>
New Zealand though?
22:20
<zewt>
annevk: apparently he thought you were a european woman
22:21
<zewt>
https://twitter.com/katoken0215/status/287320663876833280
22:21
<zewt>
(he deleted the last one, heh)
22:23
<yroc>
Hi, wondering why the leading newline following an opening <pre> tag is stripped...
22:23
<yroc>
Is it to prevent the opening <pre> tag from displacing the initial line of preformatted text.
22:25
<annevk>
yroc: legacy thing from SGML iirc
22:26
<annevk>
yroc: http://www.w3.org/TR/html401/appendix/notes.html#notes-line-breaks
22:26
<annevk>
wycats_: you around?
22:30
<annevk>
Anyone else know if it gets any better than http://api.jquery.com/on/ ?
22:31
<annevk>
whoa
22:31
<annevk>
http://zeptojs.com/ document.write is still recommended anno 2012
22:32
<yroc>
annevk: Right. And now, with the white-space processing model, *all* newlines are ignored when white-space: normal;
22:32
<annevk>
yroc: right, so the quirk is only kept for <pre> and <textarea>, where it's observable
22:32
<annevk>
well, more directly observable
22:32
<yroc>
Exactly, where white-space: pre;
22:33
<yroc>
annevk: But *why* is the quirk kept?
22:33
<annevk>
yroc: because otherwise sites will render differently
22:33
<zewt>
the opposite of "on()" is "off()"? really? heh
22:34
<annevk>
zewt: I wonder if that's actually used
22:34
<yroc>
annevk: if the initial newline following <pre> were not stripped, sites would render differently?
22:35
<annevk>
yroc: right
22:35
<yroc>
Uh, can you give an example?
22:35
<annevk>
yroc: you would get a blank line
22:36
<yroc>
Oh, I see what you mean. I thought you meant different browsers would render differently (an interop problem). Yes, exactly, you'd get a blank line
22:36
<annevk>
yroc: http://www.w3.org/TR/CSS21/text.html seems to rely on it for instance
22:37
<annevk>
yroc: well we don't want to remove it because then you'd get an interop problem
22:37
<annevk>
yroc: and we'd break sites
22:38
<zewt>
annevk: doesn't look like it's actually a very nice interface (off(), i mean)
22:38
<yroc>
yroc: but isn't it a useful thing to keep? Doesn't it allow you to start the first line of preformatted text on a "fresh" line without the starting <pre> tag displacing the content?
22:38
<annevk>
zewt: I agree, I'm open to suggestions
22:38
<zewt>
since you still have to have a reference to the listener function, so it has the same problems as addEventListener(x, func.bind(something))
22:38
<yroc>
Oops, I meant annevk!
22:39
<annevk>
yroc: it's being kept
22:39
<zewt>
nicer to return an object representing the observation, eg. x = event.addEventListener(...); x.remove();
22:39
<zewt>
i seem to recall prototype has something like that (been a while though)
22:39
<annevk>
with s/event/element/?
22:40
<zewt>
yeah
22:40
<annevk>
we also have the setInterval API we could mimic...
22:40
<zewt>
http://prototypejs.org/doc/latest/dom/Event/on/index.html
22:40
<yroc>
annevk: Yes, I know, but I'm just saying there's a good reason, which I'm trying to nail down (for example, "without this behavior, you'd get an undesirable blank line at the beginning of your preformatted content."
22:41
<yroc>
annevk: Would you say that's the true reason?
22:41
<annevk>
zewt: could you add that to https://www.w3.org/Bugs/Public/show_bug.cgi?id=16491 perhaps? I guess I'm gonna try to make some progress on that bug. It's been well over a year now...
22:41
<zewt>
these days i just pre-bind my event handlers (this.click_handler = this.click_handler.bind(this)), which is a little wordy but works fine in practice
22:42
<zewt>
not at home, i'll add it if i remember when i get back
22:44
<zewt>
lack of a capture argument to all of those apis is also annoying
22:51
<annevk>
well if nobody is using that...
22:52
<annevk>
we can do all kinds of things, e.g. make all events bubble for the purposes of this new API
22:52
<annevk>
I actually thought jQuery did that, but that page does not describe that behavior
22:54
<annevk>
http://prototypejs.org/doc/latest/dom/Event/Handler/new/index.html is an interesting API too
22:54
<annevk>
although it should implicitly call start
22:55
<Hixie>
yroc: generally if changing something would break web pages (as in this case), we don't even bother to look for other reasons :-)
22:57
<zewt>
yeah i'd much prefer an api that always bubbles, i can filter out the bubble phase easily enough if i really want that
22:58
<zewt>
personally the most annoying subtle weirdness of dom events for me is how capturing listeners aren't run before non-capturing listeners during the "on target" phase (or whatever it's called)
22:59
<annevk>
listeners are in adding order
23:00
<zewt>
but intuitively capturing listeners are before non-capturing listeners, but that's not actually the case when dispatching to the target itself
23:01
<yroc>
Hixie: I see. But incidentally, I think it's actually a good thing that the initial newline is stripped for the reason I outlined above.
23:01
<zewt>
anyway, fixing that even in a different api is probably not worth the added complexity of the subtle difference (and probably other weirdness as a result)
23:01
<Hixie>
yroc: probably, yeah
23:02
<annevk>
zewt: well we can do all kinds of stuff, e.g. let the developer completely control independently whether capture/bubble/target are included or not
23:03
<annevk>
zewt: but whether that's needed... it seems most cases are about target+bubbling, preferably with event delegation via selector
23:04
<zewt>
maybe i just do weird things, i dunno
23:04
<yroc>
Hixie: In fact, I think it would be good to strip the newline after a child <code> starting tag...
23:04
<yroc>
, so the author can start the first line flush with the rest of the code (just a suggestion!)
23:05
<yroc>
Hixie: Wondering why do we an element to represent a word wrap opportunity?
23:05
<zewt>
sometimes i'll put a capturing listener on a tree to grab an event, do something funky with it, then re-dispatch it (perhaps at a later time), eg. grabbing onclick and delaying the actual onclick by 0.5s to allow an animation to happen
23:05
<yroc>
Hixie: Why isn't zero width space sufficient?
23:05
<annevk>
zewt: you can do the same with preventDefault()
23:06
<zewt>
no, because i don't want any non-capturing event listeners to run (since often the actual behavior of the button or whatever is a scripted event, not the default)
23:07
<yroc>
Hixie: U+200B
23:07
<zewt>
anyway, it can be dealt with, it just means guaranteeing that the element where the capturing event listener lives is *above* any nodes that might receive the event later, to guarantee the capturing listener happens first--not a huge deal, just annoying when i forget
23:09
<annevk>
given that none of the libraries have it and pretty much no developer I have talked to really gets the final argument of addEventListener, I'm sceptical about the new API catering for it, but who knows
23:10
<tantek>
why not leave it out until someone needs it in a real world situation?
23:10
<jamesr>
the one that you just say "false" for because firefox complains if you don't?
23:11
<annevk>
tantek: well zewt had such a situation apparently ;)
23:12
<annevk>
jamesr: does it still? :/
23:12
<annevk>
hopefully once those new IDL bindings land that will no longer happen
23:12
<jamesr>
oh! guess it doesn't any more
23:13
<annevk>
since I started writing the core bit of Events it has been optional per spec
23:14
<annevk>
seems optional in Nightly 20something at least
23:18
<zewt>
annevk: i think a new events API should absolutely support capturing, but it'd be fine if it was an options dictionary item instead
23:19
<annevk>
dunno
23:20
<zewt>
capturing isn't used a lot but when you need it, it's really helpful
23:20
<zewt>
afk, heading home