00:05
<Hixie>
TabAtkins: checked in
00:08
<TabAtkins>
What section is it in?
00:09
<Hixie>
search for cssElementMap
00:09
<Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#dom-document-csselementmap
00:09
<TabAtkins>
Sigh. To do that I have to load the single-page version.
00:09
<TabAtkins>
Ah, that works.
00:09
<Hixie>
(it's in the DOM tree accessors section, for lack of a better place to put it)
00:10
<Hixie>
the single-page version works fine with chrome, fwiw
00:10
<TabAtkins>
You messed up a <dfn> tag in the first sentence.
00:10
<Hixie>
oops
00:10
<Hixie>
will fix
00:10
<Hixie>
thanks
00:12
<TabAtkins>
Looks good to me, Hixie. Thanks.
00:12
<Hixie>
np
00:12
<TabAtkins>
Hm. The "specified" seems redundant. Wanna drop that?
00:13
<Hixie>
sure
00:14
<TabAtkins>
Oh, wait, one thing.
00:15
<Hixie>
yes?
00:15
<TabAtkins>
Only img, canvas, and video should be assignable to the map while out-of-document.
00:15
<Hixie>
can't enforce that since you can add it to the map and then remove an element
00:15
<TabAtkins>
Any other element shouldn't be assignable while out-of-document, and should remove itself from the map if removed from teh doc.
00:16
<Hixie>
that would be rather confusing -- better to just make it that a reference to an unusable element gets ignored on the CSS side, imho
00:16
<Hixie>
or i can make those elements not define the mapping for you
00:16
<Hixie>
but still have them be in the API's map
00:17
<Hixie>
it'd be confusing because you don't want it to matter whether you create, then add to the list, then append to the doc, or create, then append, then add to hte list
00:17
<Hixie>
and you don't want things to break when all you're doing is moving an element around the dom (which involves being removed from the doc briefly)
00:17
<TabAtkins>
Yeah, makes sense. So yeah, make it so they don't map to the CSS identifier while not in the document.
00:17
<Hixie>
k
00:18
<Hixie>
img, video, canvas is a rather arbitrary list
00:18
<Hixie>
what about <input type=image>, <object>, <embed>...?
00:19
<TabAtkins>
<input type=image>, sure. object, embed, iframe I'm not sure of right now - I've already asked about them on the list.
00:19
<TabAtkins>
I think the answer should be "yes", though there are wrinkles, like iframes not loading until they're added to the document.
00:20
<Hixie>
do i have to have the HTML side also handle the case of the <img> element's bitmap not being available yet?
00:20
<TabAtkins>
It should just be any element whos rendering is unaffected by surrounding elements by default.
00:20
<Hixie>
or of a <canvas> with height=0?
00:20
<Hixie>
it really seems best to have this be all handled in the CSS side
00:20
<TabAtkins>
No, CSS should handle that with "with an appearance exactly equivalent to that of the referenced element"
00:20
<Hixie>
well then if you can handle <img alt="x" src="bogus"> out of doc why can't you handle <span>x</span>?
00:21
<Hixie>
battery's dying, gotta go. will check in once i'm near power again.
00:21
<TabAtkins>
The problem with doing this on the CSS side is that CSS has no notion of elements outside the document. So I'm not sure how to talk about them.
00:33
<karlcow>
http://www.barrypearson.co.uk/articles/layout_tables/history.htm
01:21
<karlcow>
about implementing current practices - http://ksi.cpsc.ucalgary.ca/archives/HTML-WG/html-wg-94q3.messages/0042.html
01:35
<karlcow>
Where ... might be any number of idioms, including nothing at all, i.e.
01:35
<karlcow>
<!DOCTYPE HTML>
01:35
<karlcow>
I would take this declaration to mean "gimme the current version of the
01:35
<karlcow>
HTML DTD."
01:35
<karlcow>
in http://ksi.cpsc.ucalgary.ca/archives/HTML-WG/html-wg-94q3.messages/0056.html
01:36
<TabAtkins>
Heh, that's exactly what it means now, except without the whole nasty DTD business. ^_^
01:46
<karlcow>
TabAtkins, I spend a bit of time in the archives on a regular basis. And the discussions are *always* the same.
01:56
<MikeSmith>
we really should just write a separate spec for <!doctype html>
01:57
<MikeSmith>
and take it to Rec
01:57
<MikeSmith>
unattached from HTML5
01:57
<MikeSmith>
which would enable anybody to use it on any doc they wanted for whatever purpose they wanted
02:00
<karlcow>
MikeSmith: ahahaha :D
02:00
<TabAtkins>
Man, this is a fascinating read: http://www.w3.org/TR/1998/NOTE-CSS-potential-19981210
02:01
<TabAtkins>
Most of it is stuff we're either just now doing, or plan to do in the very near future.
02:01
<TabAtkins>
They had the future of CSS mapped out over a decade ago. ^_^
02:01
<TabAtkins>
(Seriously, there's very little that we're doing now or plan to do soon that isn't already on the list.)
02:02
<MikeSmith_>
karlcow: really -- because as TabAtkins said, it really has nothing to do with any DTD at all
02:03
<MikeSmith_>
it's not asking for a gimme of anything
02:03
<MikeSmith>
and it's not asserting the document is an HTML5 document
02:04
<MikeSmith>
if it's asserting anything at all, it's simply that the document is an HTML document, period
02:05
<MikeSmith>
what the spec should say is that it's only needed to prevent UAs from processing the document in quirks mode
02:06
<TabAtkins>
Remember - all sequences of bytes are HTML documents. They may not be valid, but they can be processed as HTML.
02:17
<karlcow>
[21:02] <MikeSmith> which would enable anybody to use it on any doc they wanted for whatever purpose they wanted
02:17
<karlcow>
I thought you were proposing doctype html for word, wordperfect, multimate and wordstar documents
02:18
<MikeSmith>
ah, no
02:20
<karlcow>
hmm I haven't found a page which explains the history mess of html mailing lists such as w3c and ietf one running in parallel.
02:37
<karlcow>
MikeSmith: :)))) http://lists.w3.org/Archives/Public/www-html-editor/2010JulSep/0001.html
02:46
<karlcow>
ok that is a start :) not perfect but still gives a way to people to find their ways http://www.w3.org/html/wiki/HTML_Mailing_Lists
02:54
<Hixie>
TabAtkins: i don't understand what "with an appearance exactly equivalent to that of the referenced element" means. Consider an inline element, what is its appearance?
02:55
<Hixie>
give line breaking, e.g.
02:55
<Hixie>
or consider an element with a top-margin collapsed with a larger following margin -- what is its appearance? specifically, what is its origin?
02:55
<Hixie>
or consider an element with display:runin
02:58
<roc>
In Gecko, we draw the border-box of the first CSS box
02:59
<roc>
er wrong
02:59
<oojacoboo>
is it really recommended not to use quote with html attributes?
02:59
<oojacoboo>
quotes*
03:00
<roc>
what we do is
03:00
<roc>
we take all the CSS boxes for the element
03:00
<roc>
we compute the bounding box that contains all those CSS boxes
03:00
<boogyman>
oojacoboo: depends on the context, single word attribute values its acceptable, multi-word values must be encapsulated by quotes
03:00
<boogyman>
it's*
03:01
<roc>
er, the bounding box is the union of all the border-boxes of the CSS boxes
03:01
<oojacoboo>
I would think that just for semantics and good use, using them all the time, would be the most appropriate
03:01
<roc>
that defines the image bounds
03:01
<boogyman>
however in the spirit of consistent coding technique, I always use quotes and lower-case
03:01
<oojacoboo>
html5 spec is littered with attributes without quotes
03:01
<roc>
then we render all the CSS boxes into it in flow order
03:02
<othermaciej>
use of quotes or not is a matter of taste
03:02
<othermaciej>
html5 examples strive to use a variety of styles, to avoid expressing an opinion on what is best
03:02
<othermaciej>
particular quoting conventions are not more or less semantic
03:03
<oojacoboo>
othermaciej: then what is it?
03:03
<othermaciej>
what is what?
03:03
<oojacoboo>
the use or not use of quotes
03:03
<othermaciej>
it is a matter of taste
03:03
<othermaciej>
like I said
03:03
<oojacoboo>
it's not a matter of taste or semantics, what is it?
03:03
<oojacoboo>
you said it's NOT
03:03
<othermaciej>
no, I said:
03:03
<othermaciej>
"use of quotes or not is a matter of taste"
03:04
<oojacoboo>
:P
03:04
<boogyman>
oojacoboo: <element class=one> and <element class="one"> are both acceptable, however <element class=one two> and <element class="one two"> have separate meanings
03:04
<othermaciej>
you sometimes need quotes for certain values
03:04
<oojacoboo>
othermaciej: don't write confusing shit like that at this hour ;)
03:04
<othermaciej>
it's just like the Unix shell
03:04
<oojacoboo>
boogyman: right
03:04
<othermaciej>
sometimes you need quotes around command parameters
03:04
<oojacoboo>
yea
03:04
<othermaciej>
much of the time you don't
03:04
<othermaciej>
you can still use them when you don't need them
03:04
<oojacoboo>
I'll stick with my anal quoting routine :)
03:05
<othermaciej>
most people choose not to
03:05
<othermaciej>
in HTML, some people feel you should quote even when you don't need to
03:05
<oojacoboo>
yea, like me
03:05
<othermaciej>
others feel you should quote only when needed
03:05
<othermaciej>
HTML5 lets you do it either way
03:05
<oojacoboo>
anyone here familiar with the <figcaption> element?
03:06
<oojacoboo>
I was trying to get some specifics on it's intent for block level use
03:07
<oojacoboo>
the spec is extremely vague on this element, so far, I've just gone ahead and done it the way I think it should be, but if someone has some intimate knowledge into this element, I'd love to ehar it
03:07
<oojacoboo>
hear*
03:07
<oojacoboo>
http://dev.rentpost.com/managers (you'll see where I have used it in the <figure> at near the bottom on this page
03:07
<oojacoboo>
+)
03:07
<Hixie>
<figcaption> is allowed in one place only, and that's as a child of <figure>
03:08
<oojacoboo>
ummm... duh ;)
03:18
<oojacoboo>
curious on all the new input type values. How are the current UAs interpreting these?
03:18
<oojacoboo>
are there any backward compatibility techniques
03:31
<Hixie>
oojacoboo: they're doing some stuff on them, not great so far
03:32
<Hixie>
oojacoboo: the back compat story is that older browsers treat them type type=text
03:32
<oojacoboo>
mmm, ok
03:32
<oojacoboo>
I was considering trying to code up some forms with html5, but I think that's probably going to be a waste at this time
03:33
<oojacoboo>
maybe type would be ok to use if they fall back to "text"
03:33
<oojacoboo>
at least for email on mobile you might not get the uppercase default onFocus
03:36
<oojacoboo>
Hixie: do you have any resources on backward compatibility at least for safe use, even if you have to replicate functionality in certain instances
03:48
<Hixie>
not off-hand
03:48
<Hixie>
diveintohtml5.org might have some useful stuff
03:48
<oojacoboo>
are people using things like modernizr and just using multiple form elements?
03:49
<oojacoboo>
not really wanting to write the html twice, doubling the markup
03:49
<boogyman>
oojacoboo: how about you do a couple of simple test cases before making templates for your "production" site
03:50
<oojacoboo>
boogyman: what do you mean?
03:50
<oojacoboo>
I don't plan on coding this straight to production, we do have a dev and stage server...
03:51
<boogyman>
create a simple html example and test in some user-agents
03:51
<oojacoboo>
I'd prefer not to waste the time coding up a bunch of things that someone already has the answer on
04:19
<Hixie>
roc: except for <img>, <video>, and <canvas>, right?
04:20
<Hixie>
roc: my point was just that the css spec would have to define all this, so having the css spec define what happens when the referenced element is an <img> with broken image data, or a <video> that's still spooling, or a <canvas> with height=0, etc, is something that would have to be defined in css anyway
04:20
<Hixie>
roc: so it's not something the html spec should define
04:20
<roc>
sure
04:20
<roc>
we actually treat <video>/<canvas>/<img> differently depending on whether they have CSS boxes or not
04:20
<Hixie>
makes sense
05:07
<oojacoboo>
Hixie: you write the lang type for the spec?
05:07
<oojacoboo>
html lang="en-US-x-hixie"
05:45
<jamesr_>
Hixie: you there? i'm trying to figure out how to upload something to your live dom viewer
05:46
<jamesr_>
i want to look at a test page i've created on an ipad
05:46
<jamesr_>
i tried clicking 'upload' and have no idea what happened
05:47
<jamesr_>
oh i probably want 'save'
05:55
<jamesr_>
figured it out!
06:28
<Hixie>
oojacoboo: i write the spec
06:29
<jacobolus>
oojacoboo: it's sort of a joke I think :)
06:44
<hsivonen>
http://www.w3.org/News/2010.html#entry-8967
06:44
<hsivonen>
I'm very confused by that announcement. When did the election start?
06:45
<hsivonen>
I thought just a couple of days ago, JeniT was contemplating whether to run for the election
07:59
<othermaciej>
hsivonen: there was a nomination period, it just ended
08:00
<othermaciej>
hsivonen: apparently, the number of nominees was the same as the number of seats up for election, thus making it a bit of a banana republic election
08:00
<jamesr_>
everyone wore a neutral-colored fleece and a scarf?
08:01
<othermaciej>
shows how much you know
08:02
<othermaciej>
banana republic is all about semi-desaturated non-primary tones
08:07
<hsivonen>
othermaciej: I see. It would have been helpful for the announcement to say that it was an election without votes.
08:07
<hsivonen>
I guess it also means that Jeni Tennison chose not to run.
08:08
<othermaciej>
I guess so
08:08
<othermaciej>
there would have been more nominees, but I couldn't convince Dave Singer to nominate people who explicitly said they'd refuse
08:37
<hsivonen>
good to see some chairing on bugs that attempted to circumvent the WG Decision on ISSUE-30
08:54
asmodai
eyes the latest chrome dev build
08:56
<asmodai>
It's not even loading the html5test.com test anymore :|
08:58
<hsivonen>
Chrome no longer supports HTML5!11!11!11111!!!1
09:00
<asmodai>
hsivonen: yea :)
09:01
<asmodai>
anyway
09:01
<asmodai>
so me and my colleagues are wondering wtf WebSQL stopped
09:01
<asmodai>
I mean, isn't it a good thing if all vendors settled on an implementation?
09:01
<asmodai>
So what's so much a problem about them all using SQLite?
09:05
<asmodai>
Mmm, funny, same version of Chrome on Mac works
09:10
<asmodai>
aha, chrome --disable-accelerated-compositing makes it work on Windows at least
09:12
<hsivonen>
asmodai: Staying compatible with the version of SQLite that Safari used for the JS-exposed SQL database isn't a goal of the SQLite project
09:12
<hsivonen>
asmodai: thus, keeping compatibility will potentially involve forking SQLite and being on the hook for security patches independently
09:14
<asmodai>
hsivonen: ok, I can understand that not being a goal of SQLite, but it shouldn't be an issue if some fork for webbrowsers was available? Would only ease maintenance for all vendors if they cooperated on that. Maybe I just see things too rose-coloured :)
09:14
<hsivonen>
asmodai: also, it's considered healthy that a feature practically requires all vendors to ship the same code
09:14
<Ms2ger>
unhealthy? :)
09:14
<hsivonen>
Ms2ger: oops. s/healthy/UNhealthy/
09:15
<hsivonen>
I think gzip might be an area where everyone ships the same code, but it's not clear if everyone *has* to
09:15
<hsivonen>
for PNG, everyone probably ships the same code, but probably wouldn't have to
09:16
<hsivonen>
investigating how realistic it would be to ship a JFIF impl. other than the Independent JPEG Group's would be an interesting exercise
09:17
<asmodai>
s/gzip/zlib/
09:17
<asmodai>
At least, I see more acknowledgements for zlib rather than gzip ;)
09:17
<hsivonen>
but at least the zlib project itself aims to keep compatibility
09:17
<asmodai>
fair enough
10:30
jgraham
notes that if you have <div class="wrap1">…<div class=wrap8> all nested inside each other, you are probably doing it wrong
10:40
Workshiva
endorses the idea of w3c member archives on wikileaks
10:47
<jgraham>
Seems kinda dull for them. This week on wikileaks: New secret emails reveal that Sir Tim Berners-Lee prefers Fudge Ice Cream to Vanilla"
10:55
<zcorpan>
that's what they discuss in TAG
11:21
<asmodai>
jgraham: sort of like: http://www.theatlanticwire.com/features/view/feature/Wikileaks-Cables-Reveal-That-Canada-Is-Boring-2668
12:36
<annevk>
ooh, the TAG thing was not even an election? hahaha
12:38
<jgraham>
annevk: You should "stand" next time
12:38
<jgraham>
cat amongst the pigeons
12:38
<annevk>
Maciej threatened to get me nominated during TPAC
12:38
<annevk>
I have since been wondering what I have done to upset him
12:43
<hsivonen>
does anyone happen to remember if removing a script from the DOM before it has been executed resets 'already started' in IE and Opera?
12:45
<hsivonen>
now that I'm implementing execution prevention for scripts that have changed docs between "run" and "execute", it would kinda make sense to clone the IE/Opera behavior of canceling a script and unblocking the parser if a pending script is removed from the doc
12:48
<hsivonen>
awesome. the m-c was closed because a test tried to load .swf from zombo.com
12:48
<jgraham>
tests + external servers === evil
12:52
<gsnedders>
m-c?
12:52
<jgraham>
mozilla-central
12:52
<jgraham>
I assume
12:52
<annevk>
hsivonen, not using http://html5zombo.com/ ?!
12:58
<hsivonen>
html5zombo.com is superior to zombo.com except as a plug-in test case
13:14
<karlcow>
annevk: there is still a chance. The last seat is to be nominated by tim ;)
13:38
<hsivonen>
hmm. I wonder what MikeSmith's plan regarding build.py updating was
13:45
<hsivonen>
MikeSmith: I started hacking on build.py
13:47
<annevk>
I hate resolving URLs across multiple windows
13:47
<annevk>
before XMLHttpRequest I never even contemplated stuff like self[0].EventSource
13:49
<annevk>
kind of funny that I suspected the "first script" to be incorrect and the first time I run an actual test it does not appear to match implementations
13:49
<annevk>
I am disappoint
14:31
<hsivonen>
MikeSmith: I've migrated build.py over.
14:32
<hsivonen>
next step is to detect if the user has MQ enabled and use hg pull --rebase instead of --update if so
14:32
<hsivonen>
but not today
14:56
<annevk>
'When Microsoft says "HTML6," I hear "X-UA-Compatible-We-Fucking-Mean-It-This-Time."'
14:56
<annevk>
-- @diveintomark
15:19
<annevk>
hsivonen, btw, we have our own image code
15:19
<annevk>
hsivonen, re discussion earlier today
17:31
jgraham
curses DOM3 XPath (again)
19:03
<annevk>
Used <s> in a blog post
19:03
<annevk>
had to hack my blog software
19:07
<Ms2ger>
jgraham, s/DOM3 //
19:18
<annevk>
anyone with ideas on structuring weblog archives?
19:18
<annevk>
especially for blogs without tags
19:18
<hober>
annevk: http://diveintomark.org/archives comes to mind
19:18
<Philip`>
annevk: Add a link to your favourite search engine
19:19
<Philip`>
Nobody's going to bother reading the archives any other way
19:19
<annevk>
search engines are
19:19
<annevk>
and I am too
19:19
<Ms2ger>
http://ln.hixie.ch/?command=index comes too mind
19:19
<annevk>
hober, yeah, thinking about that
19:19
<annevk>
oh, Hixie's one should be a lot easier
19:20
<annevk>
but it'll be a long... very long
19:21
<annevk>
there's 2500 entries
19:21
<annevk>
(this includes links)
19:21
<annevk>
(called HREF at one point)
19:26
<zcorpan>
have one page for each year
19:26
<annevk>
I guess I should start with the yearly and monthly pages indeed
19:27
<annevk>
yearly already exists, but currently excludes link posts
19:30
<zcorpan>
oh. maybe http://annevankesteren.nl/archives could list the years instead of the months then?
19:37
<annevk>
zcorpan, yeah prolly
19:38
<karlcow>
annevk: what do you mean by structuring archives
19:38
<karlcow>
access to old entries? I do this http://www.la-grange.net/map
19:39
<karlcow>
http://www.la-grange.net/2009/ http://www.la-grange.net/2009/01/
19:39
<annevk>
yeah, I suspect I will do something like that
19:39
<annevk>
I like how your archives match the style of the site at the time
19:40
<annevk>
I can probably not do that
19:41
<karlcow>
annevk: yes I always kept the old style. There are part of the history of the content.
19:42
<karlcow>
I have a few broken links to fix here and there.
19:43
<annevk>
such a rebel
19:43
<annevk>
content != style karlcow
19:43
<annevk>
after so many years at the W3C you should know
19:43
<karlcow>
hehe
19:43
<karlcow>
but style is part of the memories ;)
19:44
<annevk>
no, you're right
19:44
<annevk>
my content is stored in a dumb database, rather than something smart like HTML
19:44
<karlcow>
hehe
20:14
<karlcow>
" I see my job as making the Mozilla developer outreach channels the Switzerland of HTML5, explaining and showing how to get all the cool new things to work across browsers, platforms and markets." -- http://www.readwriteweb.com/archives/html5_mozilla_and_the_future_of_the_web.php
20:14
<karlcow>
does it involve secret funds?
20:16
<annevk>
ok, yearly archives are now better
20:16
<annevk>
including a special style sheet for 2004 and 2005
20:16
<annevk>
(because of HREF)
20:18
<miketaylr>
seems like a standards body would be a better switzerland
20:24
<annevk>
oh yes http://annevankesteren.nl/archives
20:26
<karlcow>
annevk: cool
20:26
<Ms2ger>
http://annevankesteren.nl/2004/01/multicol-spacer
21:16
<oojacoboo>
anyone familiar with the pseudo class that targets the red outline/border for required input elements with html5
21:17
<oojacoboo>
particularly seems to be an incredible eye sore with gecko/-moz
21:18
<oojacoboo>
it's way too loud/in-your-face, whomever made the decision that that'd be the best method of indicated a required field either did it b/c they already had the code for :focus or just doesn't know what they are doing
21:18
<Ms2ger>
-moz-box-shadow: none?
21:19
<oojacoboo>
is that the style element?
21:19
<oojacoboo>
what's the pseudo class?
21:19
<oojacoboo>
:required ?
21:19
<Ms2ger>
:invalid, I guess
21:19
<oojacoboo>
I don't mind having this after a submit, but sitting before is insane
21:20
<oojacoboo>
I actually want it after a submit
21:20
<oojacoboo>
but not before!
21:20
<Ms2ger>
volkmar would know, but we might be fixing that
21:20
<oojacoboo>
something not so loud, PLEASE!
21:21
<oojacoboo>
it ruins the on page eye flow and funneling
21:23
<oojacoboo>
Ms2ger: that did the trick though, thanks
22:32
<oojacoboo>
anyone know what the style element is for the annoying 2px border around <button> on webkit?
22:32
<oojacoboo>
a border: 0 !important won't even nix it
22:42
<JoePeck>
oojacoboo: I don't see a border on => data:text/html,<button style="border:0">NoBorder
22:43
<JoePeck>
do you have a sample page / picture?
22:43
<oojacoboo>
it's not a "border" really, I think it's the accessibility border
22:43
<oojacoboo>
although, it kicks down the element 2px which I do not want it doing
22:44
<oojacoboo>
inspector says the ua stylesheet is styling border though
22:44
<oojacoboo>
which can't be overridden
22:44
<JoePeck>
hmm, "-webkit-box-sizing: border-box;" help at all? Just a thought. I am not familiar with accessibility items
22:44
<oojacoboo>
tried that, nope
22:44
<oojacoboo>
that's the default for it anyway
22:44
<oojacoboo>
one second, I can get an example
22:48
<oojacoboo>
http://dev.rentpost.com/managers @ JoePeck
22:48
<oojacoboo>
JoePeck: see the button at the bottom with the contact form, notice how in webkit it's positioned lower
22:50
<JoePeck>
oojacoboo: it has a margin of 2px
22:51
<JoePeck>
oojacoboo: although I don't know where it comes from
22:51
<oojacoboo>
it's not a margin, I don't think so
22:51
<oojacoboo>
almost 100% certain
22:51
<JoePeck>
oojacoboo: setting margin:0 made it look fine in my Safari
22:51
<oojacoboo>
:/
22:51
<oojacoboo>
sec
22:51
<JoePeck>
oojacoboo: check the "Metrics" section of the inspector
22:51
<oojacoboo>
yea
22:53
<JoePeck>
oojacoboo: I have no idea where that margin came from. Might be worth making a reduction. If its webkit only maybe webkit has a bug
22:53
<oojacoboo>
it's not in firefox
22:54
<oojacoboo>
I don't know why the browsers have such a hard time with <button>
22:54
<oojacoboo>
it's one of the most frustrating elements in all of html
22:54
<JoePeck>
oojacoboo: it could be the built-in browser theming. If you're familiar with WebKit code, RenderThemeMac, etc
22:55
<JoePeck>
oojacoboo: maybe try: -webkit-appearance: none
22:55
<oojacoboo>
it's likely there for accessibility reasons
22:55
<oojacoboo>
or so I hope
22:55
<oojacoboo>
but I'd think outline: would be beng used for that
22:55
<oojacoboo>
I'll try that
22:55
<JoePeck>
hmm, the -webkit-appearance made no difference. Well, please file a bug on this
22:56
<oojacoboo>
no go JoePeck
22:56
<oojacoboo>
link?
22:56
<JoePeck>
https://bugs.webkit.org/enter_bug.cgi
22:56
<JoePeck>
preferably with a reduction
22:57
<oojacoboo>
yea, figured as much
23:12
<hober>
annevk: reading the new reply to your aria feedback
23:12
<hober>
"To reduce some of the confusion caused by our "boolean " value type[...] we have renamed the name of the type to "true/false "."
23:12
<hober>
http://lists.w3.org/Archives/Public/public-pfwg-comments/2010OctDec/0020.html
23:26
<annevk>
hober, "and believe your proposal to allow host language remapping of ARIA data types would unnecessarily break implementations"
23:26
<annevk>
they are so full of shit
23:27
<annevk>
that was that they were saying
23:27
<annevk>
I was asking them to drop that "host language" stuff
23:27
<annevk>
glad we have gone full circle
23:29
<annevk>
"Response to the concerns raised in your acknowledgement" even says that
23:29
<annevk>
I guess this is what you get when you talk with committees rather than a single person
23:34
<annevk>
ooh, Hixie killed his style sheet
23:34
<annevk>
I was wondering what was going on
23:35
<annevk>
and his new style sheet follows w3.org conventions with meaningless dates?!
23:35
<annevk>
Hixie trying to get appointed by Tim? ;p
23:36
<Hixie>
it's not a meaningless date :-)
23:36
<Hixie>
it's the name of my stylesheet
23:36
<Hixie>
old one was "spaced out"
23:36
<Hixie>
then there was the never-completed "orange"
23:36
<Hixie>
and now "2010"
23:37
<Hixie>
it might get renamed once i've any idea what it looks like :-)
23:37
<Hixie>
also, wow, you noticed fast
23:38
<annevk>
I happened to be reading the archives
23:38
<Hixie>
aah
23:38
<annevk>
http://ln.hixie.ch/?start=1033035916&count=1 -- quite amusing
23:39
<Hixie>
wow, i was right
23:39
<Hixie>
check out the number of replies
23:40
<annevk>
heh, never really looked beyond the contents :)
23:58
<AryehGregor>
Wow, Tantek actually got the name "t" on Twitter? Was he involved in founding it, or did he pay a large sum of money, or what?
23:58
<gsnedders>
AryehGregor: Early enough.
23:58
<hober>
he originally had 'tantek'
23:58
<gsnedders>
(IIRC)
23:58
<hober>
but switched to 't'
23:59
<gsnedders>
(My memory could be wrong)
23:59
<hober>
while several of the one-letter usernames were still available