2010-03-01
[18:58:00.0000] <Hixie>
huh

[18:59:00.0000] <Hixie>
ever since he retracted his formal objection, larry's been posting to the public-html list with a different (non-adobe) e-mail address

[18:59:01.0000] <mpilgrim>
AryehGregor: I can't find an authoritative statement on chrome.blogspot.com, blog.chromium.org, or dev.chromium.org about supported codecs

[18:59:02.0000] <mpilgrim>
which surprises me

[19:00:00.0000] <mpilgrim>
and since i'm on the chrome developer relations team, i will personally find out what each product supports and blog about it in an authoritative place

[19:02:00.0000] <mpilgrim>
for <video>, chromium only supports theora/vorbis in an ogg container

[19:02:01.0000] <boblet>
http://000000book.com/ GML; graffiti markup language (yep, there’s a markup language for that)

[19:02:02.0000] <mpilgrim>
for <audio>, chromium only supports vorbis

[19:02:03.0000] <mpilgrim>
UNLESS

[19:02:04.0000] <mpilgrim>
you install the non-free codecs

[19:03:00.0000] <mpilgrim>
this is packaged in the ubuntu PPA as chromium-codecs-ffmpeg-nonfree.deb

[19:03:01.0000] <mpilgrim>
then chromium will supports exactly what chrome supports

[19:03:02.0000] <mpilgrim>
but i'm not exactly sure what that is

[19:03:03.0000] <mpilgrim>
it includes H.264/AAC in an MP4 container

[19:03:04.0000] <mpilgrim>
for <video>

[19:04:00.0000] <mpilgrim>
it also includes MP3 for <audio>

[19:04:01.0000] <mpilgrim>
i believe chrome also supports MP3 tracks in a video

[19:04:02.0000] <mpilgrim>
but i will double-check that

[19:05:00.0000] <mpilgrim>
and since it's based on ffmpeg, i *think* it supports up to and including High Profile H.264, all profiles of AAC, etc.

[19:05:01.0000] <mpilgrim>
but i will double-check that too

[19:07:00.0000] <mpilgrim>
i have no information on if, when, or whether chrome will add support for VP8 or any of the other On2 codecs that we just acquired

[19:10:00.0000] <mpilgrim>
neither chromium nor chrome use gstreamer, directshow, quicktime, or any other media backends

[19:10:01.0000] <mpilgrim>
we built our own, some details are here: http://dev.chromium.org/developers/design-documents/video

[19:11:00.0000] <mpilgrim>
there is a bug asking us to rip that out and use gstreamer instead: http://code.google.com/p/chromium/issues/detail?id=32861

[19:13:00.0000] <mpilgrim>
it's marked "WONTFIX"

[23:14:00.0000] <JonathanNeal>
hey all

[00:51:00.0000] <jgraham>
http://blog.mozilla.com/webdev/2010/02/25/new-python-library-bleach/

[00:53:00.0000] <Philip`>
I thought for a moment that was a Python port of http://search.cpan.org/~dconway/Acme-Bleach-1.12/lib/Acme/Bleach.pm

[00:54:00.0000] <jgraham>
Philip`: People with that level of evil malice are all perl programmers

[00:57:00.0000] <othermaciej>
Philip`: that is evil

[01:01:00.0000] <roc>
oooff

[02:08:00.0000] <annevk>
some of these guys on the hybi list aah

[02:09:00.0000] <hsivonen>
I guess I should still be glad I don't read the hybi list

[02:14:00.0000] <Dashiva>
hsivonen: Careful or annevk might start quoting it :P

[02:15:00.0000] <annevk>
Dashiva, you'd like that, wouldn't you? :p

[02:16:00.0000] <Dashiva>
Pssh, I can read it for myself :P

[02:28:00.0000] <annevk>
unrelated to what I said above, Julian on hybi: "Hm, no. Break backwards compatibility if it brings clear benefits, and the damage is not too big."

[02:28:01.0000] <othermaciej>
oh no, there's more posts on hybi

[02:28:02.0000] <annevk>
jaja

[02:33:00.0000] <othermaciej>
I should really teach myself not to post to standards lists late at night

[02:33:01.0000] <othermaciej>
comes out snarky too often

[02:33:02.0000] <gsnedders>
othermaciej? Snarky? NEver!

[02:35:00.0000] <othermaciej>
there's also return of the son of the versioning permathread on public-html

[02:40:00.0000] <roc>
someone needs to write "Extending The Web Platform For Dummies"

[02:40:01.0000] <jgraham>
/me looks at public-html notices a thread called (roughly) html-versioning vs. longdesc, wonders who came up with the idea of permathread deathmatches as a way or resolving issues

[02:42:00.0000] <annevk>
also something about Web Platform protocol/API/markup language design for dummies

[02:42:01.0000] <othermaciej>
jgraham: that was Larry's idea

[02:44:00.0000] <othermaciej>
if we could really arrange to have two permathreads enter, and only only one will leave, I would be all for it

[02:44:01.0000] <othermaciej>
but I fear that two will enter, and three will leave

[02:44:02.0000] <Hixie>
rofl at othermaciej's last e-mail

[02:45:00.0000] <othermaciej>
Hixie: that was probably the one I shouldn't have sent

[02:52:00.0000] <Dashiva>
Amusing that some people consider atom 0.3 vs 1.0 a success

[02:54:00.0000] <gsnedders>
I'd say it was

[02:55:00.0000] <Dashiva>
A success in ensuring there are now two incompatible atom formats that both need to be supported

[02:55:01.0000] <Philip`>
A success in ensuring there are only two incompatible formats and not more

[04:49:00.0000] <gsnedders>
Wow, Hixie accepted my feedback.

[04:49:01.0000] <gsnedders>
:P

[05:07:00.0000] <jgraham>
What was it? Conforming html5 documents MUST contain a link to donate to your paypal account?

[05:11:00.0000] <gsnedders>
I'm under 18, I'm not allowed a paypal account

[05:11:01.0000] <jgraham>
There's lots of things you're not allowed but that doesn't stop you

[05:11:02.0000] <jgraham>
Anyway you will be 30 by the time HTML5 is a REC

[05:37:00.0000] <asmodai>
Mmm, interesting little project: http://code.google.com/p/html5media/

[05:37:01.0000] <asmodai>
GUess it was pasted to the channel before.

[05:44:00.0000] <hsivonen>
asmodai: I like how the script hard-codes a magic H.264 profile

[05:45:00.0000] <hsivonen>
in practice, mortals are not going to get the codecs parameter right for their particular H.264

[05:45:01.0000] <hsivonen>
profiles FAIL

[05:47:00.0000] <annevk>
that line is part of the unwritten book on Web format design

[05:48:00.0000] <asmodai>
hsivonen: Hadn't even seen that yet ;)

[05:54:00.0000] <Philip`>
Just need to make sure encoding programs give you a nice UI so you can choose some combination of "iPhone compatible" and "Maximum compression (not compatible with mobile devices)" etc, and then they encode all the versions and spit out the appropriate chunk of HTML to use

[05:54:01.0000] <Philip`>
Surely that can't be hard to do

[05:54:02.0000] <MikeSmith>
/me reads discussion about validation on public-html

[05:55:00.0000] <jgraham>
"Surely that can't be hard to do" is more or loss what got us the state of the web today

[05:55:01.0000] <jgraham>
*less

[05:55:02.0000] <Philip`>
Sure, but that's not going to stop me being sure about it

[05:55:03.0000] <hsivonen>
Philip`: those piece will be even harder to keep together than keeping files and MIME types without parameters together

[05:55:04.0000] <Philip`>
(It'll stop me being correct, of course)

[05:57:00.0000] <hsivonen>
I'm struggling with my old nemesis: the onload event

[05:57:01.0000] <hsivonen>
s/onload/load/

[06:02:00.0000] <annevk>
I just realized

[06:02:01.0000] <annevk>
reading public-html is like reading the newspaper, almost always bad news

[06:04:00.0000] <jgraham>
/me is finding the blizzard outside rather distracting

[06:04:01.0000] <gsnedders>
Me too.

[06:04:02.0000] <gsnedders>
The fact I know I have to walk home in that makes me sad.

[06:05:00.0000] <jgraham>
Camp at the office?

[06:06:00.0000] <gsnedders>
Seems rather dull.

[06:06:01.0000] <gsnedders>
But so does my flat.

[06:06:02.0000] <gsnedders>
Oh well.

[06:07:00.0000] <annevk>
you should come party in the Netherlands

[06:08:00.0000] <gsnedders>
Hmm… Flights to Amsterdam…

[06:09:00.0000] <asmodai>
annevk: Where you at in NL anyway?

[06:09:01.0000] <annevk>
Utrecht

[06:09:02.0000] <asmodai>
Ahh, Utreg

[06:10:00.0000] <asmodai>
Area of the clogged motorways ^^

[06:10:01.0000] <annevk>
no car, no problems :)

[06:10:02.0000] <asmodai>
True, true.

[06:11:00.0000] <annevk>
you're Dutch too right? /whois says so anyway

[06:11:01.0000] <asmodai>
Aye, work in R'dam, live in Barendrecht

[06:12:00.0000] <annevk>
might go to bar with the gang from #fronteers on Wednesday here in Utrecht

[06:12:01.0000] <annevk>
pretty sure everyone here would be welcome to come and otherwise I'll take a hit

[06:13:00.0000] <asmodai>
#fronteers?

[06:14:00.0000] <annevk>
IRC channel from http://fronteers.nl/

[06:15:00.0000] <asmodai>
Ah

[06:16:00.0000] <asmodai>
Wasn't even aware of that group.

[06:16:01.0000] <asmodai>
Useful, since we do a shitload of web dev here at the uni.

[06:17:00.0000] <MikeSmith>
I'm trying to figure out what, if any, new information there is in http://escholarship.org/uc/item/0rp834wf "Privacy Issues of the W3C Geolocation API"

[06:17:01.0000] <annevk>
krijnh throws a party for each new member

[06:19:00.0000] <Peter`>
He still owes me one in that case :p

[06:30:00.0000] <asmodai>
annevk: hahaha

[06:30:01.0000] <asmodai>
annevk: Sounds ideal. :)

[06:34:00.0000] <boblet>
are there any HTML5 elements that aren’t media-independent now?

[06:35:00.0000] <Dashiva>
audio?

[06:37:00.0000] <boblet>
Dashiva: I’m guessing that the media independence of img, video, canvas & audio is the mentioning of a11y alternatives, right?

[07:07:00.0000] <MikeSmith>
I notice http://wiki.commonjs.org/wiki/Binary/F was added recently.. wondering if the Ecmascript TC has been discussing anything about it or other JS binary-datatype proposals recently

[08:17:00.0000] <beilabs>
Quick question, is The Web Hypertext Application Technology Working Group the same group as WebApps WG?

[08:20:00.0000] <jgraham>
No

[08:52:00.0000] <zcorpan>
/me does s/autobuffer/preload/ in http://simon.html5.org/html5-elements

[09:40:00.0000] <othermaciej>
I need to update mine

[09:41:00.0000] <Dashiva>
Looks like someone tried to implement acid2: http://25.media.tumblr.com/tumblr_kycn8zaXYz1qzpwi0o1_500.jpg

[11:16:00.0000] <JonathanNeal>
Hey all!

[11:17:00.0000] <TabAtkins>
yo, JonathanNeal.

[11:27:00.0000] <JonathanNeal>
What's new in the htmizzle?

[11:30:00.0000] <Necrathex>
the <bling> element

[11:31:00.0000] <Philip`>
/me groans

[11:40:00.0000] <jwalden>
the data-pimped-out="data-pimped-out" attribute

[11:45:00.0000] <zcorpan__>
data-mining=""

[11:50:00.0000] <zcorpan__>
someone should set up a twitter account announcing the daily new html5 data-* attribute

[11:51:00.0000] <zcorpan__>
sometimes with a new attribute on <embed>

[11:53:00.0000] <jwalden>
<blinquee data-scrolling-speed="11" />

[11:59:00.0000] <Philip`>
Someone should make an HTML5 Applicable Specification generator

[11:59:01.0000] <Philip`>
You tell it what elements and attributes you want to be valid, and it writes a spec for you

[12:00:00.0000] <zcorpan__>
it should also output a pluggable schema and the appropriate validator.nu url

[12:05:00.0000] <zcorpan__>
mpilgrim: btw, i guess you should update diveintohtml5 wrt autobuffer

[12:08:00.0000] <zcorpan__>
mpilgrim: the js impl would probably look for a preload attribute with value empty string or ascii-case-insensitive "auto" to mean the same as yesterday's autobuffer

[13:18:00.0000] <annevk>
Google acquired a Flash-based app? that's funny giving all the recent commotion

[13:25:00.0000] <annevk>
that thing on Wednesday in Utrecht now has an associated blog post: http://fronteers.nl/blog/2010/03/spontaanborrel-woensdag

[13:26:00.0000] <annevk>
http://twitter.com/jreschke/status/9818873134 -- didn't masinter say he switched to using that address because of The Incident?

[13:26:01.0000] <annevk>
on his blog or something

[13:26:02.0000] <annevk>
I'm now wondering why I even care

[13:26:03.0000] <annevk>
/me goes back to reading QC

[13:27:00.0000] <othermaciej>
maybe the claimed piece of misinformation was that Larry "retracted his Formal Objection"

[13:27:01.0000] <othermaciej>
(since according to Larry, he never made a Formal Objection)

[13:29:00.0000] <annevk>
ooh, I guess you're right, lol

[13:32:00.0000] <JonathanNeal>
data-picard="great combo"

[13:39:00.0000] <othermaciej>
annevk: could you update HTML4 diffs for publication as a Working Draft this Thursday?

[13:39:01.0000] <othermaciej>
annevk: see email

[13:39:02.0000] <othermaciej>
Hixie: same for your drafts and we have a minor status change request

[13:42:00.0000] <jwalden>
krijnh: would it be possible to change the name/nick-styling cookie on the logs site to be effectively permanent?  month-long lifetime doesn't serve much purpose given the contents of the cookies in question

[13:47:00.0000] <Hixie>
othermaciej: tuesday march 2nd? sure

[13:47:01.0000] <Hixie>
othermaciej: what's the change?

[13:47:02.0000] <othermaciej>
Hixie: Thursday March 4 please

[13:47:03.0000] <othermaciej>
Hixie: see http://lists.w3.org/Archives/Public/public-html/2010Mar/0025.html

[13:47:04.0000] <othermaciej>
for the change

[13:47:05.0000] <Hixie>
oh thursday, sorry, misread

[13:47:06.0000] <othermaciej>
can explain further if needed but I have an interview in 6 minutes

[13:48:00.0000] <Hixie>
nothing more to explain, will do

[13:49:00.0000] <othermaciej>
thanks much

[13:52:00.0000] <Hixie>
changed s/specification is/contents of this specification are also/

[13:53:00.0000] <Hixie>
oh wait the requested change was more than that

[13:53:01.0000] <zcorpan__>
/me was just going to say...

[14:12:00.0000] <annevk>
I will tomorrow or so

[14:12:01.0000] <annevk>
I guess that would be a good time to integrate recent changes as well

[14:13:00.0000] <annevk>
data-driven="marcos"

[14:50:00.0000] <Hixie>
othermaciej: is the registry for issue 27 testable yet?

[14:51:00.0000] <Hixie>
othermaciej: i'm happy to be a volunteer to test the registry once it exists, though we should probably have several people do so to get a reasonable idea of the results

[14:54:00.0000] <othermaciej>
Hixie: I believe the essential part of the requested status change was s/this specification/the contents of this specification/

[14:54:01.0000] <Hixie>
sam and i figured it out on #html-wg

[14:54:02.0000] <othermaciej>
k

[14:55:00.0000] <othermaciej>
Hixie: in issue 27 - the registry is not testable, the ETA we were given for when it would be testable was "about 6 weeks", that being 2 weeks ago now

[14:55:01.0000] <Hixie>
k

[14:55:02.0000] <othermaciej>
Hixie: we would accept someone volunteering with a long deadline to account for that, and a plan to re-evaluate if the registry is not ready in time

[14:56:00.0000] <Hixie>
k, i'll volunteer to organise a test of the registry when it's available

[14:56:01.0000] <Hixie>
issue-31 is very confusing... any chance we can get that split into one issue per actual problem instead of just having a catch-all?

[14:57:00.0000] <Hixie>
-32 also

[14:57:01.0000] <othermaciej>
Hixie: it does seem to me that (a) the three Change Proposals submitted for 31 appear to be largely orthogonal; and (b) Laura's in particular bundles multiple changes

[14:58:00.0000] <othermaciej>
I am not sure we want to flood the issue tracker, but I can see how it may get confusing with counter-proposals or alternatives

[14:58:01.0000] <Hixie>
as far as i can tell, the confusing issues are always the ones that are about a topic and not about a specific problem

[14:58:02.0000] <othermaciej>
On -32, I want to see what the accessibility TF actually ends up submitting

[14:58:03.0000] <othermaciej>
indeed, and sadly many of our pre-new-process issues are about a topic

[14:59:00.0000] <Hixie>
maybe i should file a bug on the process to ask that we change the issue mechanism to require that issues specify a specific problem :-)

[14:59:01.0000] <othermaciej>
well, that's effectively true for new issues, since they need to be escalated from a bug

[14:59:02.0000] <othermaciej>
question is really what, if anything, to do for the finite number of pre-existing issues that are more open-ended than we'd like for new issues

[15:01:00.0000] <Hixie>
ok off to work to get food so i can resume editing websocket

[15:01:01.0000] <Hixie>
back in a bit

[15:05:00.0000] <annevk>
I replied to public-html on some email about schemas -- big mistake

[15:06:00.0000] <annevk>
Learning from past mistakes is something I should take more seriously...

[15:08:00.0000] <annevk>
It was a while ago though: http://annevankesteren.nl/2007/04/html-red-pill

[15:08:01.0000] <annevk>
/me still likes that post

[15:09:00.0000] <othermaciej>
annevk: interesting to look at how many of those we are still discussing

[15:09:01.0000] <othermaciej>
No one seems to care about "No SGML" any more

[15:19:00.0000] <Dashiva>
You can remove the SGML from HTML, but you can't remove the SGML from their hearts


2010-03-02
[18:28:00.0000] <othermaciej>
just waiting on html4diffs and h:tml to be updated now

[18:33:00.0000] <MikeSmith>
othermaciej: I just now updated the h:tml draft and waiting for the cvs commit to complete

[18:33:01.0000] <MikeSmith>
oK

[18:33:02.0000] <MikeSmith>
done

[18:33:03.0000] <othermaciej>
MikeSmith: hawt

[18:33:04.0000] <othermaciej>
MikeSmith: so now we're just waiting on Anne

[18:33:05.0000] <MikeSmith>
OK

[18:34:00.0000] <MikeSmith>
annevk still awake?

[18:34:01.0000] <othermaciej>
I asked him to update earlier today, so hopefully he'll do it whenever he gets up

[18:34:02.0000] <othermaciej>
I am assuming no

[18:34:03.0000] <othermaciej>
I asked him to update ~5 hours ago

[18:35:00.0000] <othermaciej>
even though it is an arbitrary milestone, I am for some reason very excited about publishing

[18:39:00.0000] <MikeSmith>
othermaciej: I think the biggest positive effect of WD publication is that it gets documents a lot wider attention than just editor's-draft publication does

[18:40:00.0000] <MikeSmith>
at least as far as W3C publication goes

[18:40:01.0000] <MikeSmith>
and for a certain period of time, at least

[18:40:02.0000] <othermaciej>
in the context of HTML5, I have not seen it result in a higher level of technical review

[18:40:03.0000] <MikeSmith>
yeah

[18:40:04.0000] <othermaciej>
though it does get some sort of attention in the tech press

[18:42:00.0000] <MikeSmith>
othermaciej: yeah, the downside of that is that one of the first questions all the media people seem to ask every time W3C publishes and HTML5 WD is that, "OK, so when will HTML5 be done?" question

[18:43:00.0000] <MikeSmith>
to which it seems to me the only proper response is to try to get them to see that they're asking the wrong question

[18:43:01.0000] <MikeSmith>
but that's not usually successfull

[18:43:02.0000] <othermaciej>
then we can just point to the charter and indicate that the next milestone is Last Call

[18:43:03.0000] <othermaciej>
which is on schedule, so long as we can invent time travel technology in the next few months

[18:43:04.0000] <MikeSmith>
heh

[18:45:00.0000] <Hixie>
the ietf guys argued that websocket would get wider review if we went through the ietf, but so far all the useful feedback i've received has been either off-list or on the whatwg list

[18:46:00.0000] <Hixie>
similar to how it was argued that html5 would get more review if we went through the w3c

[18:46:01.0000] <Hixie>
looks to me like the standards organisations are out of date in terms of what gets wider useful feedback :-)

[18:47:00.0000] <othermaciej>
Hixie: I sent useful feedback on the ietf list... I think

[18:48:00.0000] <Hixie>
i think everything you said on the list you first said on the adam/you/me thread we had

[18:48:01.0000] <Hixie>
and adam and you would have reviewed the spec anyway, regardless of whether we went w3c, whatwg, or ietf

[18:48:02.0000] <othermaciej>
nah, I pointed out the reverse cross-protocol problem on the list first, then discussed it with you on IRC, then on the email thread with Adam

[18:48:03.0000] <othermaciej>
I would like to say I would have given that feedback anyway but I only thought of it due to people questioning the handshake

[18:49:00.0000] <Hixie>
fair enough

[18:49:01.0000] <MikeSmith>
Hixie: standards organizations are out of date in a lot of ways.. I guess until we actually work on coming up with viable alternative, we are stuck with what we got, and hopefully at least we can (or have) managed to get some improvements made

[18:49:02.0000] <othermaciej>
I mean, I might have still had the thought

[18:50:00.0000] <Hixie>
MikeSmith: my most successful standard i think has been pingback, which didn't involve a standards organisation at all

[18:50:01.0000] <othermaciej>
it's not clear to me how to make a good standards org

[18:50:02.0000] <Hixie>
MikeSmith: so it's not clear to me that we _need_ an alternative

[18:50:03.0000] <MikeSmith>
I think the existence of the Web Sockets discussion at IETF has helped to changed the IETF culture for the better

[18:50:04.0000] <othermaciej>
all the existing ones either come down to pay-to-play voting or dictatorship by an elite oligarchy, if you push hard enough

[18:51:00.0000] <Hixie>
really? how so?

[18:51:01.0000] <Hixie>
er, that was to mike

[18:51:02.0000] <Hixie>
othermaciej: i think the oligarchy mechanism is the only one that truly works, but it only works so long as the oligarchy has the respect of the implementors

[18:52:00.0000] <Hixie>
othermaciej: and i think it pretty much stops having the respect as soon as there's a process in place, because a process forcibly distances the oligarchy from the implementors

[18:52:01.0000] <Hixie>
part of the problem the w3c and ietf both have is that they try to solve too many problems at once

[18:52:02.0000] <Hixie>
having multiple focus areas is a huge red flag for a std org imho

[18:52:03.0000] <MikeSmith>
Hixie: I think the implementors issue is something that the Web Sockets work has helped to raise awareness about in the IETF

[18:53:00.0000] <othermaciej>
oligarchy can work with the right oligarchs

[18:53:01.0000] <othermaciej>
and so long as it does not get overwhelmed with legitimacy disputes

[18:53:02.0000] <Hixie>
legitimacy disputes = lack of respect

[18:54:00.0000] <MikeSmith>
Hixie: the awareness being that it is risky to develop a spec without also working hard to get implementor buy-in during the spec-development process

[18:54:01.0000] <othermaciej>
IETF and W3C try to have consensus/voting as the front line decision-making tool, with escalation to a dictator/oligarchy

[18:54:02.0000] <othermaciej>
sometimes I wonder if the other way around would work better

[18:54:03.0000] <othermaciej>
of course, in the HTML WG we kinda have a sandwich

[18:54:04.0000] <Hixie>
not really

[18:55:00.0000] <Hixie>
we just have three tiers of <span>abort the

[18:55:01.0000] <Hixie>
er

[18:55:02.0000] <Hixie>
mispaste

[18:55:03.0000] <Hixie>
we just have three tiers of oligarchy

[18:55:04.0000] <othermaciej>
I guess layers at least limit the damage that can be done by any one layer going off the rails

[18:56:00.0000] <Hixie>
each tier being further removed from the issues, leading to more and more random resolutions as an issue is escalated ;-)

[18:57:00.0000] <othermaciej>
I find I need to have a lot more familiarity with the issues than I think I should have to for my role

[18:58:00.0000] <roc>
/me wonders what Apple's Webkit devs are doing

[18:58:01.0000] <othermaciej>
right now?

[18:58:02.0000] <othermaciej>
most of them are either at home, or having dinner

[18:59:00.0000] <Hixie>
roc: i think he means his role as htmlwg chair, not webkit manager :-P

[18:59:01.0000] <othermaciej>
oh

[18:59:02.0000] <othermaciej>
yeah

[18:59:03.0000] <roc>
no, my comment was random

[18:59:04.0000] <MikeSmith>
heh

[19:01:00.0000] <othermaciej>
trac would give you a good approximation if you can filter down to Apple committers

[19:01:01.0000] <roc>
indeed

[19:01:02.0000] <MikeSmith>
webkit devs seem to be doing a lot of bug fixing these days, as opposed to feature implementation

[19:02:00.0000] <othermaciej>
MikeSmith clearly reads the webkit commits twitter feed

[19:03:00.0000] <MikeSmith>
I do notice that Dirk Schulze seems to still be actively been doing some refinements to the svg filters stuff

[19:05:00.0000] <MikeSmith>
I wish trac had a better way to track commits per-developer

[19:06:00.0000] <MikeSmith>
per-committer feeds

[19:06:01.0000] <Hixie>
i wish trac had a lot of things

[19:12:00.0000] <MikeSmith>
Hixie: who's developed the http://html5.org/tools/web-apps-tracker UI, annevk ?

[19:12:01.0000] <Hixie>
yes

[19:12:02.0000] <Hixie>
iirc

[19:12:03.0000] <Hixie>
at last he hosts it

[19:12:04.0000] <MikeSmith>
OK

[19:12:05.0000] <Hixie>
it's in google code if you want to offer patches

[19:12:06.0000] <Hixie>
product html5, iirc

[19:12:07.0000] <MikeSmith>
OK

[19:13:00.0000] <MikeSmith>
I would like for it to have a feature that lets users supplement that indicators

[19:13:01.0000] <roc>
man, the per-platform test results make Webkit SVN enormous

[19:30:00.0000] <Hixie>
ok, bbl, food time.

[19:30:01.0000] <Hixie>
if anyone wants to review proposals for websocket, i just regenned the complete.html spec with the new proposed handshake

[21:32:00.0000] <MikeSmithX>
Hixie: (when you get back) - I'm wondering if you made any changes at all in response to Noah's message about explicitly defining the term "conforming document" - http://lists.w3.org/Archives/Public/public-html-comments/2010Feb/0011.html

[21:54:00.0000] <MikeSmith>
othermaciej: I think at some point we should publish a WD of the static "author view" of HTML5

[21:54:01.0000] <MikeSmith>
http://dev.w3.org/html5/spec-author-view/

[21:54:02.0000] <othermaciej>
MikeSmith: I agree

[21:55:00.0000] <othermaciej>
MikeSmith: should it still have all the trappings of normativity?

[21:55:01.0000] <MikeSmith>
that part I dunno

[21:55:02.0000] <othermaciej>
I feel weird having multiple normative references for the same thing even if they are mechanically generated from a single source

[21:55:03.0000] <othermaciej>
but other than that - we did suggest to the TAG that we'd like to do this

[21:55:04.0000] <MikeSmith>
and that TAG seems to actually like that doc OK

[21:56:00.0000] <MikeSmith>
Noah at least does, I know

[21:56:01.0000] <MikeSmith>
well, I guess I shouldn't say "I know", but I think he has publicly commented quite positively

[21:57:00.0000] <MikeSmith>
anyway, I suppose we should start discussion about publishing after we get the current round of WDs out

[21:57:01.0000] <MikeSmith>
*publishing the author view

[21:58:00.0000] <othermaciej>
yeah I'd like to get the current round done

[21:58:01.0000] <othermaciej>
and get more issues in the pipeline

[21:58:02.0000] <othermaciej>
we have 10 sitting on "Chairs" now

[22:04:00.0000] <MikeSmith>
othermaciej: I also have a couple of bugzilla bugs assigned to me that have not been escalated to issues yet and that I'm hoping won't have to be

[22:05:00.0000] <MikeSmith>
hmm, http://lists.w3.org/Archives/Public/public-html-bugzilla/2010Mar/0008.html

[22:05:01.0000] <MikeSmith>
"Can the script element please allow the scope attribute in the same semantic way as the style element? The dom would be limited to only elements under that node."

[22:06:00.0000] <othermaciej>
interesting idea

[22:06:01.0000] <othermaciej>
hard to implement and use

[22:06:02.0000] <othermaciej>
probably not viable as as a security feature, if that is what the commentor intends

[22:13:00.0000] <MikeSmith>
not sure what the commenter intends -- the embedded commenting feature is great for reporting outright bugs but doesn't encourage a lot of elaboration 

[22:36:00.0000] <hsivonen>
Hixie: so your most successful spec depends on xml-rpc!

[23:03:00.0000] <MikeSmith>
hsivonen: which one is that?

[23:06:00.0000] <Hixie>
MikeSmith: i don't think so... did he file a bug?

[23:07:00.0000] <MikeSmith>
Hixie: no, he didn't.. I can suggest to him that he should, or I can file one myself, I guess. He just posted only to the public-html-comments list about it, as far as I can see

[23:07:01.0000] <MikeSmith>
/me goes to raise new bug for it

[23:07:02.0000] <hsivonen>
MikeSmith: pingback according to the log

[23:08:00.0000] <MikeSmith>
ah

[23:08:01.0000] <Hixie>
hsivonen: yeah, that's embarassing as heck :-)

[23:13:00.0000] <annevk>
MikeSmith, I'll update nowish I guess

[23:13:01.0000] <annevk>
/me is trying to wake up

[23:13:02.0000] <MikeSmith>
annevk: great

[23:13:03.0000] <annevk>
/me has a headache for unknown reasons

[23:13:04.0000] <annevk>
/me wonders if it's because the heating is working again and his body is no longer used to the warmth

[23:14:00.0000] <MikeSmith>
Hixie: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9178 (new bug for defining "conforming document")

[23:14:01.0000] <MikeSmith>
btw, I notice that pubrules still doesn't recognize CSS3 color names

[23:15:00.0000] <MikeSmith>
I would say, "That seems like it would be relatively easy to fix" but if I did I would probably get asked to fix it

[23:26:00.0000] <MikeSmith>
othermaciej: I guess one advantage of publishing the author view of the spec as non-normative would be that we'd not need to do an LC round for it

[23:26:01.0000] <MikeSmith>
nor CR

[23:26:02.0000] <othermaciej>
MikeSmith: we'd probably want to anyway, to let it stay in sync with the main HTML5 draft

[23:26:03.0000] <MikeSmith>
yeah, true

[23:27:00.0000] <MikeSmith>
no, not true

[23:27:01.0000] <othermaciej>
MikeSmith: how hard do you think it would be to get pubrules changed to allow HTML5 as a publication format?

[23:27:02.0000] <othermaciej>
I periodically have the urge to try to pursue that, but I am not sure if it is worth my time

[23:27:03.0000] <MikeSmith>
pubrules would be relatively easy to change, actually

[23:28:00.0000] <MikeSmith>
it's not a technical problem that blocks that

[23:28:01.0000] <othermaciej>
I didn't mean "how technically hard"

[23:28:02.0000] <MikeSmith>
well, not hard, then

[23:28:03.0000] <othermaciej>
I mean, "If I suggested it, is there a chance I'd get anywhere?"

[23:28:04.0000] <MikeSmith>
I meant, "well, hard, then"

[23:28:05.0000] <othermaciej>
and if so, what would be the right venue

[23:28:06.0000] <othermaciej>
should I ask privately what the actual blocker is?

[23:29:00.0000] <MikeSmith>
no need to ask privately

[23:29:01.0000] <othermaciej>
what's the actual blocker?

[23:30:00.0000] <othermaciej>
I wondered if maybe there was some idea that a format spec had to be at REC before it could be used, but I notices that even Working Drafts of XHTML1 were published as XHTML1, not as HTML4

[23:30:01.0000] <MikeSmith>
the blocker is that the team doesn't believe it's appropriate yet to be publishing documents that use features from HTML5 that are not at least in a PR draft

[23:30:02.0000] <othermaciej>
so it doesn't seem like there is a hard and fast rule

[23:30:03.0000] <MikeSmith>
as far as XHTML1 and HTML4, that was then, this is now

[23:31:00.0000] <MikeSmith>
there were a whole lot of corners cut in publishing HTML4 and XHTML1

[23:31:01.0000] <MikeSmith>
I would suggest that we don't want to use their publication approach as a precedent

[23:31:02.0000] <othermaciej>
it just seems like an embarrassment to me that we can't publish HTML5 as HTML5

[23:31:03.0000] <othermaciej>
but from what you say, it sounds like it would be a waste of time to pursue it

[23:32:00.0000] <Hixie>
it's an embarassment to the w3c

[23:32:01.0000] <Hixie>
html5 has been published as html5 for years

[23:32:02.0000] <MikeSmith>
fwiw, I have argued that simply using <!doctype html> as the doctype on a document does not constitute publishing it as HTML5

[23:32:03.0000] <othermaciej>
it's an embarrassment to the HTML WG, even though it is not up to us

[23:34:00.0000] <othermaciej>
anyway I'll mentally file it away as "it's political and not worth pursuing" unless the other HTML WG chairs decide they care too at some point

[23:35:00.0000] <MikeSmith>
I am happy to pursue it if the chairs will agree to support pushing for it 

[23:35:01.0000] <othermaciej>
I'll ask the other co-chairs then

[23:35:02.0000] <othermaciej>
another data point: XHTML 1.1 Working Drafts were also published as XHTML 1.1

[23:35:03.0000] <othermaciej>
that's not quite as long ago as HTML4 or XHTML 1.0, but still pretty long ago

[23:37:00.0000] <MikeSmith>
XHTML 1.1 is another case of a spec that I don't think it'd be prudent to use as a precedent for anything

[23:37:01.0000] <MikeSmith>
anyway, I don't want to (re)raise it if somebody's going to end up pulling the rug out from underneath

[23:37:02.0000] <othermaciej>
I'm not asking you to pursue it right now

[23:37:03.0000] <othermaciej>
I just wanted to understand the issue

[23:37:04.0000] <MikeSmith>
OK

[23:38:00.0000] <MikeSmith>
I do want to say that a technical way we could address this is in part is by having <!doctype html> be defined in a separate spec that we could move through Rec track more quickly

[23:39:00.0000] <othermaciej>
this is the relevant rule, right: http://www.w3.org/2005/07/pubrules?uimode=filter&uri=#format

[23:40:00.0000] <MikeSmith>
othermaciej: yeah

[23:40:01.0000] <othermaciej>
interesting note: the XHTML 1.1 REC is in violation of that rule

[23:41:00.0000] <MikeSmith>
yep

[23:43:00.0000] <MikeSmith>
anyway, to be clear, I do think it is worth pursuing at least getting agreement to be allowed to publish the spec with the <!doctype html> doctype

[23:44:00.0000] <MikeSmith>
but I suggest that not be considered the same thing as "publishing 

[23:44:01.0000] <hsivonen>
othermaciej: occasionally, I wish people who raise process issues about the HTML WG went raise the same issues about the XHTML2 WG first

[23:44:02.0000] <othermaciej>
the most recent violation of pubrules format requirements I can find so far was January 16, 2009

[23:44:03.0000] <annevk>
MikeSmith, I doubt that would go to REC more quicly

[23:44:04.0000] <MikeSmith>
meant to write, I suggest that not be considered the same thing as publishing the spec "as HTML5"

[23:44:05.0000] <MikeSmith>
annevk: why?

[23:45:00.0000] <annevk>
MikeSmith, because there's a bunch of open debates around the DOCTYPE

[23:45:01.0000] <annevk>
(I also have no idea how you would separate it out, but aside from that...)

[23:46:00.0000] <othermaciej>
to be fair this is only a Note, perhaps those are not considered to have any normative versions at all

[23:46:01.0000] <othermaciej>
http://www.w3.org/TR/2009/NOTE-xhtml-media-types-20090116/

[23:46:02.0000] <MikeSmith>
hsivonen: it does seem that the HTML WG gets singled for process problems that are not unique to this group

[23:47:00.0000] <othermaciej>
but the pubrules for WG Notes still have the format requirements

[23:47:01.0000] <othermaciej>
this one seems to have slipped through the checker

[23:47:02.0000] <hsivonen>
if the doctype were a separate spec, versioning fans could object to it and say in public that they aren't objecting to HTML5--just the doctype

[23:48:00.0000] <MikeSmith>
othermaciej: I'm not sure whether the pubrules checker has always checked for the doctype constraint or not

[23:48:01.0000] <MikeSmith>
I think mi

[23:48:02.0000] <MikeSmith>
it might just pass it to the validator as-is

[23:48:03.0000] <MikeSmith>
hsivonen: I suppose

[23:49:00.0000] <othermaciej>
I do fear that splitting off the doctype could harm substantive progress

[23:49:01.0000] <JonathanNeal>
Is role="banner" gone?

[23:49:02.0000] <MikeSmith>
so the thing is, we would then need to define what we mean by "publishing as HTML5"

[23:49:03.0000] <othermaciej>
JonathanNeal: it still exists - just not default

[23:49:04.0000] <JonathanNeal>
I couldn't find it anywhere in the draft.

[23:50:00.0000] <MikeSmith>
for example, does "publishing as HTML5" mean we want it to be OK for a document to include features that don't yet have any implementation support?

[23:50:01.0000] <othermaciej>
JonathanNeal: it's specified in WAI-ARIA, which is a reference

[23:50:02.0000] <JonathanNeal>
or the WAI ARIA @ http://www.w3.org/TR/wai-aria/#banner (seems to have been removed?)

[23:51:00.0000] <hsivonen>
JonathanNeal: ARIA changed to publishing only the TOC at the main URL

[23:51:01.0000] <othermaciej>
JonathanNeal: http://www.w3.org/TR/wai-aria/roles#banner

[23:51:02.0000] <othermaciej>
it's a multipage spec

[23:51:03.0000] <JonathanNeal>
Thanks hsivonen and othermaciej, found it now.

[23:51:04.0000] <hsivonen>
/me doesn't like multipage specs

[23:51:05.0000] <JonathanNeal>
/me agrees with hsivonen on that one.

[23:51:06.0000] <othermaciej>
MikeSmith: I think we would want to exercise judgment and not use markup features that break in current browsers or that are expected to break (or change) in the future

[23:52:00.0000] <othermaciej>
MikeSmith: perhaps the WC3 Team would not trust the HTML WG to exercise that level of judgment

[23:52:01.0000] <othermaciej>
I like my find in page, so me three

[23:52:02.0000] <JonathanNeal>
A lot of these roles are now elements in HTMl5.

[23:53:00.0000] <annevk>
MikeSmith, what does publishing as HTML4 mean?

[23:53:01.0000] <annevk>
MikeSmith, we could publish HTML4 that validates perfectly fine but is not usable in any browser, is that more acceptable than HTML5?

[23:54:00.0000] <annevk>
I'd argue that with HTML5 such a problem is less likely to occur

[23:55:00.0000] <othermaciej>
annevk: you could write validating HTML5 that is not usable in any browser (for example if it makes assumptions about default rendering)

[23:55:01.0000] <MikeSmith>
annevk: you're preaching to the choir here -- I'm saying we need to make sure we be clear about what it is we want to get approval for

[23:56:00.0000] <othermaciej>
I will ask the other two co-chairs if they feel this is an important issue

[23:56:01.0000] <othermaciej>
to me it's lower on the priority list than getting to Last Call, but it does kinda bother me

[23:57:00.0000] <MikeSmith>
othermaciej: as far as trust, it would seem to me at least that the W3C Team does trust that chairs of the HTML WG

[23:57:01.0000] <MikeSmith>
otherwise they would not continue to be the chairs

[23:57:02.0000] <othermaciej>
all trust has limits, I don't think they would give me the W3C's bank account number

[23:57:03.0000] <othermaciej>
then again, nor do I want it

[23:59:00.0000] <sicking>
i wouldn't mind if someone gave it to me

[00:00:00.0000] <MikeSmith>
sicking: !

[00:00:01.0000] <sicking>
as long as I got permission to use it for whatever i wanted :)

[00:00:02.0000] <sicking>
hey Mike!

[00:00:03.0000] <MikeSmith>
long time no see

[00:01:00.0000] <sicking>
yeah, i think my irc client wasn't configured to autoconnect here for a long time

[00:01:01.0000] <sicking>
iirc as a result of my mac dying

[00:02:00.0000] <sicking>
/me looks at maciej

[00:02:01.0000] <sicking>
though really it was seagates fault, it was due to hard drive failure

[00:03:00.0000] <MikeSmith>
some might argue that hard drive failures are inevitable, so that any unforeseen consequences of a hard-drive failure are actually pilot error (that is, lack of preparing for the inevitable) :)

[00:06:00.0000] <othermaciej>
sicking: your Mac died?

[00:06:01.0000] <othermaciej>
sorry to hear

[00:07:00.0000] <sicking>
MikeSmith: it is true. Our desktop admin had asked me to get backup for quite some time

[00:07:01.0000] <Hixie>
why is the doctype an issue?

[00:07:02.0000] <Hixie>
that's not the interesting part of html5

[00:07:03.0000] <sicking>
othermaciej: work mac. HD failed. It ended up not being a huge deal, just lost a few days worth of work

[00:07:04.0000] <othermaciej>
I see

[00:08:00.0000] <sicking>
othermaciej: could have been *much* worse, but i got lucky

[00:08:01.0000] <othermaciej>
I concur with blaming yourself and/or the drive manufacturer

[00:08:02.0000] <sicking>
now i back up religiously. I have to say that time machine rocks

[00:08:03.0000] <othermaciej>
my most important things are checked into repositories or on mail or web servers

[00:08:04.0000] <othermaciej>
it would mostly be a personal annoyance, not a work one, if my drive totally failed

[00:09:00.0000] <roc>
for Mozilla development it's pretty easy to keep your work in Mercurial patch queues and save them to http://hg.mozilla.org/users/rocallahan_mozilla.com/ etc

[00:10:00.0000] <roc>
sicking: just don't lose XBL2!

[00:11:00.0000] <sicking>
roc: you push your mq repository there?

[00:11:01.0000] <annevk>
XBL2 is more than a pipe dream? :)

[00:11:02.0000] <roc>
I push some there

[00:11:03.0000] <roc>
now that I mention it, I really should push my other big queue there too

[00:13:00.0000] <othermaciej>
I guess for extra safety I could post more works-in-progress to bugs.webkit,.org

[00:13:01.0000] <othermaciej>
as a manager I don't have the time to code anything that complicated though

[00:13:02.0000] <roc>
/me anxiously scans his patch queue for anything embarrassing

[00:14:00.0000] <roc>
hg qdel video-h264

[00:15:00.0000] <annevk>
othermaciej, MikeSmith, http://dev.w3.org/html5/html4-differences/

[00:16:00.0000] <othermaciej>
annevk: thanks

[00:17:00.0000] <othermaciej>
if you're gonna do that then I guess I should delete my ogg patch from bugs.webkit.org

[00:18:00.0000] <MikeSmith>
heh

[00:18:01.0000] <MikeSmith>
annevk: ah, cool

[00:19:00.0000] <sicking>
roc: dude, you're sitting there watching youtube on <video> but holding out on the rest of us?

[00:20:00.0000] <sicking>
roc: i want lolcats in HTML too!!

[00:21:00.0000] <roc>
annevk: that's a great document

[00:24:00.0000] <annevk>
thanks!

[00:46:00.0000] <zcorpan>
annevk: "Web Sockets" should not be a link?

[00:47:00.0000] <zcorpan>
annevk: under 5.3. Changes from 12 February 2009 to 23 April 2009

[00:49:00.0000] <zcorpan>
annevk: "potential hostile content inline" - potentially

[00:52:00.0000] <MikeSmith>
ah, cool.. Opera "Check for Updates" works for me now in my 10.5 alpha install and lets me install beta 1

[00:53:00.0000] <annevk>
zcorpan, potentially, really?

[00:53:01.0000] <annevk>
zcorpan, there's two separate drafts, wasn't sure what to link to

[00:56:00.0000] <Hixie>
it's called "WebSocket" now btw (the specs are WebSocket API and WebSocket protocol)

[00:57:00.0000] <annevk>
i guess I can rename it

[00:58:00.0000] <annevk>
Hixie, maybe remove "The" before WebSocket API? e.g. it's also Selectors API and hardly any other spec has "The"

[01:03:00.0000] <zcorpan>
annevk: i'm no english expert but i'd say potentially there

[01:03:01.0000] <annevk>
zcorpan, same here, but then for potential

[01:04:00.0000] <annevk>
MikeSmith, opinions?

[01:05:00.0000] <MikeSmith>
annevk: sorry, wasn't paying attention.. what's the particular thing you guys have been talking about?

[01:06:00.0000] <annevk>
the phrase "potential hostile content inline" in my draft

[01:07:00.0000] <MikeSmith>
annevk: I think zcorpan is right that "potentially" would be better

[01:08:00.0000] <MikeSmith>
because "potentially" is an adverb modifying "hostile"

[01:08:01.0000] <MikeSmith>
whereas otherwise I suppose it might be ambiguous about meaning "potential content"

[01:08:02.0000] <annevk>
fixored

[01:10:00.0000] <MikeSmith>
so, I guess I can start getting the drafts staged into the dated TR URLs

[01:35:00.0000] <othermaciej>
/me wonders where to find 15-year-old HTML

[01:35:01.0000] <othermaciej>
/me also wonders where to find whatever dope dbaron has been smoking

[01:37:00.0000] <annevk>
img { border:0 } is one of the few things I always use when I add images

[01:38:00.0000] <othermaciej>
does Opera put borders on img in links by default?

[01:38:01.0000] <annevk>
nope

[01:38:02.0000] <annevk>
which is a reason I not always use that rule anymore I think

[01:38:03.0000] <annevk>
just forget about it sometimes

[01:39:00.0000] <Philip`>
I wish Opera did, so that I wouldn't accidentally make pages that look uglier than expected in Firefox

[01:39:01.0000] <annevk>
I believe zcorpan did some research at some point on this subject

[01:40:00.0000] <othermaciej>
we have never had it in WebKit

[01:40:01.0000] <othermaciej>
and I don't think we ever got a bug, internal or external, to add it

[01:41:00.0000] <hsivonen>
othermaciej: Mac IE 5 didn't have the border

[01:42:00.0000] <othermaciej>
hsivonen: yeah, but in the early days, Mozilla/Phoenix/Firebird/Firefox was most people's standard of "correct' rendering

[01:42:01.0000] <othermaciej>
for people who filed bugs anyway

[01:43:00.0000] <asmodai>
annevk: If you ever see TMS (again), make sure to bring him stroopwafels 

[01:46:00.0000] <othermaciej>
ok, I found a bug related to borders around images

[01:46:01.0000] <othermaciej>
it was that we still drew a border for <input type=image border=0> in Safari 0.6

[01:48:00.0000] <othermaciej>
hmm I take it back, I found a site that deliberately adds a blue border to image links:

[01:48:01.0000] <othermaciej>
http://news.google.com/

[01:49:00.0000] <hsivonen>
othermaciej: that's not the *default* blue border, though, in any browser

[01:49:01.0000] <othermaciej>
hsivonen: indeed

[01:54:00.0000] <annevk>
asmodai, TMS being?

[01:55:00.0000] <asmodai>
TMS!~Thomas⊙poc

[01:56:00.0000] <zcorpan>
annevk: i researched image borders?

[01:56:01.0000] <annevk>
zcorpan, maybe I misremembered

[01:56:02.0000] <zcorpan>
i might well have, don't remember either :)

[01:58:00.0000] <annevk>
asmodai, ah

[01:58:01.0000] <othermaciej>
I could understand arguing they are needed for compat, I was surprised at the argument that it's actually a good default

[02:17:00.0000] <Hixie>
there's a request that we report whether a websocket connection closed gracefully or not

[02:17:01.0000] <Hixie>
i see two ways to do this:

[02:17:02.0000] <Hixie>
1. add some state data to the 'close' event

[02:17:03.0000] <Hixie>
2. add a new event

[02:17:04.0000] <Hixie>
if we go with 2, does anybody have any suggestions for what the two events should be?

[02:17:05.0000] <Hixie>
onclose and onerrorclose?

[02:18:00.0000] <Hixie>
(onerror is going to be used for when an unexpected frame type is received)

[02:20:00.0000] <jgraham>
I think I prefer 1). Having to register two different event handlers in the case that you don't care about the error seems unweildy

[02:21:00.0000] <Philip`>
Are people usually going to want to perform the same processing in response to both types of closing?

[02:21:01.0000] <jgraham>
Plus the state seems needed anyway if there is more than one possible type of error

[02:21:02.0000] <Hixie>
k

[02:21:03.0000] <othermaciej>
Hixie: are there any non-fatal errors?

[02:21:04.0000] <othermaciej>
I would say "close" and "error" if there were two events

[02:21:05.0000] <Hixie>
othermaciej: yeah, receiving a frame of an unexpected type

[02:21:06.0000] <othermaciej>
I also think extra info in the "close" event is good

[02:21:07.0000] <Hixie>
k

[02:22:00.0000] <othermaciej>
it would be nice if the close event could tell you the last message known to be delivered, but that requires more than clean close I think

[02:24:00.0000] <Hixie>
v2.

[02:24:01.0000] <Hixie>
actually you couldn't do that in the close event anyway

[02:24:02.0000] <Hixie>
you need to reconnect to find that information

[02:25:00.0000] <annevk>
can't we use a different event for a frame of an unexpected type

[02:25:01.0000] <annevk>
the error event has always been used for network errors

[02:25:02.0000] <Hixie>
we can use whatever event you want

[02:26:00.0000] <Hixie>
what would you like

[02:28:00.0000] <annevk>
messageerror?

[02:28:01.0000] <Hixie>
done

[02:29:00.0000] <annevk>
http://isgeolocationpartofhtml5.com/ sweet

[02:29:01.0000] <othermaciej>
Hixie: last message *known* to be delivered - it would be a pessimistic estimate in the error case, and exact in the clean close case

[02:29:02.0000] <othermaciej>
Hixie: would not require reconnectin

[02:29:03.0000] <othermaciej>
Hixie: but it would require some form of per-message acks

[02:30:00.0000] <Hixie>
othermaciej: definitely v2.

[02:30:01.0000] <othermaciej>
not saying this is essential, just theorizing

[02:30:02.0000] <othermaciej>
annevk: a lot of people would disagree with that site

[02:30:03.0000] <othermaciej>
Hixie: I don't know if it has to be v-anything

[02:31:00.0000] <othermaciej>
there is a tension in designing this protocol

[02:31:01.0000] <othermaciej>
on the one hand, it would be hugely valuable to have practical experience before adding a lot of stuff

[02:31:02.0000] <othermaciej>
on the other hand, you really don't want to accidentally lock in a flawed design prematurely

[02:31:03.0000] <othermaciej>
that seems like a weakness in Roy's proposed "deploy first, then standardize" model

[02:33:00.0000] <Hixie>
specs have to be written while implementations grow

[02:33:01.0000] <Hixie>
there's no magical solution, you just have to be careful

[02:33:02.0000] <Hixie>
anyway, what should this event attribute be. event.closeError?

[02:34:00.0000] <Hixie>
(boolean)

[02:35:00.0000] <othermaciej>
I like booleans to start with something like "has" or "is" but that's hard to do in this case

[02:36:00.0000] <Hixie>
event.wasClean?

[02:37:00.0000] <Hixie>
what's the opposite of a clean close

[02:37:01.0000] <Hixie>
abrupt?

[02:37:02.0000] <Hixie>
terminated?

[02:37:03.0000] <annevk>
why do we need an event attribute if you have error/close?

[02:37:04.0000] <Hixie>
annevk: it was argued that we should have only one event

[02:37:05.0000] <annevk>
that's not really consistent with <img>, XHR, etc.

[02:38:00.0000] <Hixie>
since in most cases you won't care

[02:38:01.0000] <Hixie>
well on those this event is called "load"

[02:38:02.0000] <annevk>
you can just do socket.onerror = x; socket.onclose = x;

[02:38:03.0000] <Hixie>
(which isn't really especially meaningful here)

[02:38:04.0000] <Hixie>
annevk: yeah but that means the simple case is harder

[02:38:05.0000] <Hixie>
which is bad API design

[02:39:00.0000] <annevk>
i guess those will get img.onloadend or some such

[02:39:01.0000] <annevk>
at some point

[02:39:02.0000] <annevk>
hmm

[03:09:00.0000] <MikeSmith>
zcorpan: how about "A void element is an element whose content model never allows it to have contents under any circumstances." ?

[03:09:01.0000] <MikeSmith>
that would seem to exclude the <colgroup span> case

[03:10:00.0000] <Hixie>
an element being void or not has nothing to do with its content model, in theory

[03:10:01.0000] <MikeSmith>
Hixie: so what does it have to do with?

[03:10:02.0000] <Hixie>
(though of course in practice there's a relationship)

[03:10:03.0000] <Hixie>
it's just a syntax thing

[03:10:04.0000] <Hixie>
there's a list of elements that are void

[03:10:05.0000] <Hixie>
they are the ones that never have an end tag

[03:11:00.0000] <Hixie>
that's all there is to it

[03:11:01.0000] <MikeSmith>
so you are not defining them as "void" in the XML syntax?

[03:11:02.0000] <Hixie>
"void" is a text/html syntax feature, it has no equivalent in XML

[03:12:00.0000] <Hixie>
it's similar to optional tags

[03:12:01.0000] <Hixie>
or RCDATA elements

[03:13:00.0000] <asmodai>
hahaha

[03:13:01.0000] <asmodai>
http://i.imgur.com/Zdk4B.jpg

[03:13:02.0000] <asmodai>
Now that's nifty

[03:16:00.0000] <zcorpan>
Hixie: why messageerror and not just error?

[03:16:01.0000] <MikeSmith>
Hixie: I guess rather than getting hung up on the word "void", I am more interested in finding a way to describe what the common characteristic of this particular set of elements is in the abstract language, rather than in any particular syntax

[03:16:02.0000] <Hixie>
zcorpan: anne asked for it, see about an hour ago in the irc logs

[03:16:03.0000] <Hixie>
zcorpan: i'd rather have error, so if you convince him to change his mind, i'll change it :-)

[03:17:00.0000] <Hixie>
MikeSmith: the concept of "void" is a syntax thing, it has nothing to do with the abstract language

[03:17:01.0000] <zcorpan>
annevk: error isn't always about network errors

[03:17:02.0000] <zcorpan>
annevk: what's the benefit of messageerror over error?

[03:18:00.0000] <Hixie>
MikeSmith: you can probably come up with some characteristic of the abstract language that all the void elements share, but it'll be just a coincidence

[03:19:00.0000] <zcorpan>
MikeSmith: how do you describe elements that have optional tags?

[03:21:00.0000] <MikeSmith>
zcorpan: I don't label them with anything. Is your suggestion that it would be an improvement to not have any special label for "elements that are not allowed to have contents under any circumstances"?

[03:21:01.0000] <zcorpan>
MikeSmith: it's similar in concept; i don't suggest which approach is better

[03:21:02.0000] <annevk>
zcorpan, I thought we were going to have an event for network errors as well

[03:22:00.0000] <zcorpan>
MikeSmith: but if you're talking about void elements, i think the definition should be accurate

[03:22:01.0000] <annevk>
zcorpan, and since I thought that was the case I thought it should be error rather than closeerror

[03:22:02.0000] <MikeSmith>
OK

[03:23:00.0000] <annevk>
zcorpan, Hixie, so since we now expose this information on close I suppose we can use error after all

[03:23:01.0000] <annevk>
zcorpan, Hixie, though having said that, is there a context where error is dispatched more than once?

[03:23:02.0000] <annevk>
hmm, maybe applicationCache

[03:23:03.0000] <Hixie>
script onerror

[03:24:00.0000] <Hixie>
as in, window.onerror

[03:24:01.0000] <zcorpan>
but that's not an actual event :)

[03:24:02.0000] <zcorpan>
media elements can get error several times if you load() it several times

[03:24:03.0000] <Hixie>
workers too

[03:25:00.0000] <zcorpan>
or change src

[03:25:01.0000] <annevk>
zcorpan, well yeah, but goes for <img>, XMLHttpRequest etc. too

[03:25:02.0000] <annevk>
zcorpan, I meant one operation causing it to be dispatched multiple times

[03:25:03.0000] <annevk>
/me forgot how window.onerror worked

[04:09:00.0000] <hsivonen>
sigh. I broke sync XHR semantics

[04:09:01.0000] <hsivonen>
(locally only but still annoying)

[04:12:00.0000] <zcorpan>
should .ogg be video/ogg or audio/ogg ?

[04:12:01.0000] <Philip`>
No

[04:13:00.0000] <Lachy>
application/ogg

[04:13:01.0000] <Lachy>
I think

[04:13:02.0000] <Lachy>
.ogv is conventially video/ogg

[04:13:03.0000] <Philip`>
Seems like you need more information than the file extension to make a correct choice

[04:13:04.0000] <zcorpan>
/me leaves out .ogg from his article

[04:14:00.0000] <jgraham>
It seems like media is all too confusing and should be served with a mime type like media/theres-no-way-I-made-the-right-choice

[04:16:00.0000] <Dashiva>
Thanks to .ogg existing, you can't even tell whether it's audio or video...

[04:17:00.0000] <GarethAdams|Home>
zcorpan: if you could determine MIME types based solely on file extension, there wouldn't be a need for MIME types

[04:17:01.0000] <Lachy>
Dashiva, that's an inherent problem with container formats that can contain 1 or more streams of multiple formats

[04:18:00.0000] <annevk>
jgraham, just omit Content-Type!

[04:19:00.0000] <Dashiva>
media/unknown

[04:19:01.0000] <Philip`>
unknown/ogg

[04:20:00.0000] <Dashiva>
Lachy: Makes me wonder, what happens if you give a file containing multiple audio streams as input to <audio>?

[04:21:00.0000] <Lachy>
I believe it will select the first audio stream

[04:21:01.0000] <Lachy>
since there is no stream selection in the api

[04:21:02.0000] <hsivonen>
/me notes that Larry took credit on putting MIME into HTTP in his latest blog post

[04:21:03.0000] <hsivonen>
s/on/for/

[04:21:04.0000] <Lachy>
but UAs might provide some stream selection mechanism

[04:25:00.0000] <Dashiva>
"In a normal standards group, the group would have a discussion, and come to some conclusion, and the editor would follow along with the group consensus."

[04:25:01.0000] <Dashiva>
Kind of like how the group had a discussion and concluded canvas was in scope

[04:32:00.0000] <asmodai>
Well, that was funny, opened a linked Google Wave and the scripts it uses is making Firefox cry and hang. Good thing I got a stop script at one point.

[05:11:00.0000] <foolip>
zcorpan: .ogg should actually be audio/ogg according to some RFC, for legacy reasons mostly

[05:14:00.0000] <zcorpan>
foolip: yeah i've heard that, although i think some conversion tools output ogg videos with .ogg extensions

[05:15:00.0000] <zcorpan>
which kinda makes the rfc out of touch with reality, and we'll probably end up with videos labeled as audio/ogg

[05:16:00.0000] <zcorpan>
which is why browsers will assume <video> when loading audio/ogg content directly

[05:34:00.0000] <asmodai>
zcorpan / annevk: congratz on 10.50

[05:37:00.0000] <foolip>
zcorpan: doesn't matter, serving everything as any "maybe" or "yes" mime type (e.g. .m4v as audio/x-wav) would work

[05:38:00.0000] <foolip>
I'm not sure if browsers rejecting e.g. text/plain will actually lead to the right mime type being used

[05:41:00.0000] <zcorpan>
asmodai: thanks

[05:42:00.0000] <zcorpan>
foolip: gif/jpg/ico are in the same situation, but are mostly correctly labeled (i think)

[05:43:00.0000] <zcorpan>
and png

[05:48:00.0000] <Dashiva>
For some definition of mostly

[05:48:01.0000] <Dashiva>
I get warnings about mislabeled images from irfanview regularly

[06:36:00.0000] <gsnedders>
Hixie: yt?

[08:06:00.0000] <hsivonen>
ok, so now Opera has Theora support, too. now if xiph released a thusnelda version of xiphqt, I could write a tutorial without hand-waving about future software

[08:19:00.0000] <TabAtkins>
Argh, why would you use script and abspos to simulate fixpos?  Pretty sure everyone supports it.

[08:27:00.0000] <miketaylr>
TabAtkins: 'cept for ie6, yes

[08:35:00.0000] <TabAtkins>
miketaylr: Man, seriously?  I always forget what IE6 doesn't support, dammit.  >_<

[08:36:00.0000] <TabAtkins>
Lucky me I'm allowed to ignore it.

[08:36:01.0000] <miketaylr>
me too :D

[08:36:02.0000] <miketaylr>
this guy sometimes helps, http://a.deveria.com/caniuse/#feat=css-fixed

[08:37:00.0000] <TabAtkins>
Ah, right.  I like that page.

[08:54:00.0000] <annevk>
sad that Apple is not using patents just for defense

[08:55:00.0000] <jgraham>
sad that the BBC is proposing to close down 6Music

[08:55:01.0000] <jgraham>
OK, that might just be me

[08:55:02.0000] <jgraham>
But I just found out and am devestated

[08:55:03.0000] <Philip`>
It wouldn't be until the end of 2011, apparently

[08:56:00.0000] <jgraham>
That doesn't help much if they do in fact do it

[08:57:00.0000] <jgraham>
It only means there is a little time to try to stop them

[08:57:01.0000] <Philip`>
It also means you can spend the next two years listening to it

[08:57:02.0000] <jgraham>
/me is confused by the whole thing, like how they propose to make more money abroad whilst simultaneously cancelling their most popular shows abroad like Top Gear

[08:58:00.0000] <Philip`>
You could even save the two years' output to disk, and play it on loop for the rest of forever

[08:58:01.0000] <annevk>
they are cancelling Top Gear?

[08:58:02.0000] <annevk>
aaah

[08:58:03.0000] <Philip`>
since you'll have forgotten about the earlier shows later on

[08:58:04.0000] <annevk>
whenever I see it that show is fun

[08:58:05.0000] <Philip`>
/me hadn't heard that about Top Gear

[08:58:06.0000] <Philip`>
I'd heard they were planning to sell the Top Gear magazine, but that's quite different

[08:58:07.0000] <jgraham>
Philip`: It just means I will spend two years being upset at the stupidity of whoever thinks that UK Commerical radio is an acceptable alternative

[08:59:00.0000] <jgraham>
Oh maybe I got the wrong idea about Top Gear

[08:59:01.0000] <jgraham>
I was still in shock by that part of the article

[08:59:02.0000] <annevk>
/me listens to Norwegian radio now and then nowadays

[08:59:03.0000] <jgraham>
/me has not listened to much Swedish radio, but it has always been utterly dreadful

[09:00:00.0000] <annevk>
oh yes

[09:00:01.0000] <annevk>
my HD arrived

[09:01:00.0000] <annevk>
time to power down and start over I guess

[09:02:00.0000] <jgraham>
(I guess I don't have very Swedish taste in music)

[09:07:00.0000] <AryehGregor>
Could I file a bug with the W3C validator team asking them that if they find a page using a strict doctype is invalid, that they check it against HTML5 and report it as valid HTML5 if it is?

[09:07:01.0000] <AryehGregor>
That would make MediaWiki's switch to HTML5 significantly smoother.

[09:08:00.0000] <AryehGregor>
Does anyone know who I could talk to about it?

[09:19:00.0000] <JonathanNeal>
Is <nav role="navigation"> completely unnecessary or good accessibility practice?

[09:22:00.0000] <AryehGregor>
JonathanNeal, completely unnecessary.

[09:23:00.0000] <JonathanNeal>
Is the role attribute alltogether unnecessary, or can it be used for good accessibility practice?

[09:25:00.0000] <paul_irish>
JonathanNeal: peep these two sections: http://www.whatwg.org/specs/web-apps/current-work/#annotations-for-assistive-technology-products-(aria) and http://www.w3.org/WAI/PF/aria-implementation/

[09:25:01.0000] <AryehGregor>
If it were altogether unnecessary, would it have been added to the spec?

[09:26:00.0000] <workmad3>
in that case, I'd say completely unnecessary as it's redundant... you're in a navigation element, so repeating that the navigation element is used for navigation is redundant

[09:26:01.0000] <workmad3>
but you can put more useful roles in

[09:26:02.0000] <AryehGregor>
<nav> is defined to have a default role of "navigation".

[09:26:03.0000] <workmad3>
heh :) there we go

[09:26:04.0000] <workmad3>
/me should really look at aria a bit more)

[09:27:00.0000] <AryehGregor>
At least, I assume it is.

[09:27:01.0000] <JonathanNeal>
I hate how these pages hang in Firefox.

[09:27:02.0000] <AryehGregor>
Yep.

[09:27:03.0000] <JonathanNeal>
I have to wait until Firefox is done doing whatever, and then open them in Chrome.

[09:27:04.0000] <AryehGregor>
JonathanNeal, you can add "multipage/" before the "#" and it will load fine, even with a section anchor.

[09:27:05.0000] <AryehGregor>
I think it's fixed in Firefox 3.7, anyway.

[09:27:06.0000] <workmad3>
ho hum... waiting for FF to sort itself

[09:28:00.0000] <JonathanNeal>
Well, it works in Chrome just fine.

[09:29:00.0000] <JonathanNeal>
Oh my, so header is banner and hgroup is header?

[09:30:00.0000] <JonathanNeal>
header element -> No role, if specified, role must be banner ... and then ... hgroup element -> heading role

[09:41:00.0000] <TabAtkins>
JonathanNeal: banner is a page-unique role, so they can't just apply it to all <header>s by default (though we tried to at first).  That's why <header> has no default role.

[09:41:01.0000] <JonathanNeal>
I'm aware, I just figured that <header> would be "heading" and <hgroup> would be "banner"

[09:41:02.0000] <JonathanNeal>
Since usually the banner does not contain the navigation.

[09:42:00.0000] <JonathanNeal>
And usually the heading does.

[09:42:01.0000] <AryehGregor>
MikeSmith, do you know anyone I could contact on the W3C validator team to suggest that if a document has an obsolete but conforming doctype, and fails parsing under that doctype, the W3C validator should try parsing as HTML5 and declare it valid if it's valid HTML5?

[09:42:02.0000] <AryehGregor>
Otherwise Wikipedia (and all other MediaWikis) will look like invalid XHTML 1 Strict, which is kind of a pain for evangelism.

[09:43:00.0000] <AryehGregor>
(well, all other MediaWikis by default, unless they disable well-formed XML)

[10:00:00.0000] <TabAtkins>
JonathanNeal: Why did you think that?  Did you think that <h1-6> were "banner" instead of "heading"?

[10:02:00.0000] <TabAtkins>
That is to say, is it the aria names that were confusing you, or the HTML names?

[10:04:00.0000] <JonathanNeal>
It was the combination of <header> being "banner" and <hgroup> being "heading" but it makes sense with the understanding of H1

[10:06:00.0000] <TabAtkins>
Gotcha.

[10:09:00.0000] <JonathanNeal>
Also, I was suprised to have a <nav> inside a role="banner"

[10:11:00.0000] <TabAtkins>
So the name for that ARIA role seems unintuitive for you?

[10:14:00.0000] <JonathanNeal>
No, I think I can adjust my meaning of banner, which isn't specific.

[10:14:01.0000] <JonathanNeal>
A banner can include navigation.

[10:14:02.0000] <JonathanNeal>
I'm just bringing it forward as I processed it.

[10:15:00.0000] <TabAtkins>
Yeah, but your earlier understanding of the word didn't match up.  I'm just narrowing down where the confusion originated.  ^_^

[10:43:00.0000] <annevk>
installed

[10:43:01.0000] <annevk>
that took ages

[10:43:02.0000] <annevk>
for some reason the USB stick was broken after all so I had to create a new one which was kind of tricky

[10:45:00.0000] <htcn>
when you create a Pattern in a canvas context, can you offset it

[10:45:01.0000] <htcn>
or just center it

[10:46:00.0000] <htcn>
I used a pattern for drawing a hue/saturation/brightness colorwheel's hue circle

[10:47:00.0000] <JonathanNeal>
Would anyone be willing to look at the source of http://sandbox.thewikies.com/html5-layout/ and share their thoughts on the notes I've included?

[10:51:00.0000] <roc>
annevk: "sad that Apple is not using patents just for defense" --- what was that about?

[10:53:00.0000] <annevk>
roc, according to daringfireball they're attacking HTC

[10:53:01.0000] <TabAtkins>
JonathanNeal: What's the "Zen" business sprinkled throughout your notes?

[10:55:00.0000] <TabAtkins>
Oh, I see.  For one of the display modes.

[10:55:01.0000] <roc>
ta

[10:59:00.0000] <JonathanNeal>
Zen, yea it's not entirely useful, but I kept it so I could mark meaningful classnames used throughout the document versus (structural classnames).

[11:01:00.0000] <roc>
annevk: it looks like they're suing over mostly software patents :-(

[11:05:00.0000] <paul_irish>
JonathanNeal: looks good. nice to have the implied role's in there

[11:06:00.0000] <JonathanNeal>
paul_irish, thanks, I'll work on a better name than "Zen", but something that implies to us "this is meaningful markup"

[11:08:00.0000] <AryehGregor>
Yes, well, everyone sensible always knew Apple was evil.

[11:10:00.0000] <annevk>
according to gruber it's the first attack they've made

[11:10:01.0000] <annevk>
i kind of hoped they'd never do that

[11:11:00.0000] <annevk>
guess I'm going to consider switching away from apple hardware entirely

[11:11:01.0000] <AryehGregor>
/me has never owned an Apple product, and doesn't ever plan to.

[11:28:00.0000] <JonathanNeal>
I also think I could remove <nav class="portlet-toolbar"><ul ... /></nav> for <menu class="portlet-toolbar"><ul ... /></nav> what do you guys think?

[11:34:00.0000] <roc>
the problem is that you can only debug Mac bugs on Apple hardware, because you can't virtualize Mac OS (because Apple is evil)

[11:37:00.0000] <annevk>
according to markp soon everything will be iPhone OS'd so that shouldn't be an issue

[11:44:00.0000] <roc>
why do people on www-font care about better tools for creating EOT fonts

[11:46:00.0000] <Philip`>
Because they want to make sites that look the same in IE as in other browsers

[11:46:01.0000] <TabAtkins>
Because (1) if we want to use @font-face widely, EOT is still necessary, and (2) CWT, one of the deliverables for FontWG, is basically EOT.

[11:47:00.0000] <annevk>
we're having a Font WG after all? oh god

[11:47:01.0000] <othermaciej>
Is CWT still in the Font WG deliverables?

[11:47:02.0000] <TabAtkins>
Just to define WOFF and CWT, and get tests for font-face.

[11:47:03.0000] <annevk>
I'm very much opposed to all this

[11:48:00.0000] <TabAtkins>
annevk: You just hate all the non-TTF formats.  ^_^

[11:48:01.0000] <othermaciej>
Apple is not enthusiastic about implementing random new font formats, but with Mozilla backing the Fonts WG and WOFF there is not much point trying to oppose it

[11:51:00.0000] <roc>
we're not backing CWT

[11:51:01.0000] <TabAtkins>
Hrm, having trouble finding the recent email about the proposed charter.

[11:51:02.0000] <roc>
I'm not even sure we're backing the Fonts WG

[11:52:00.0000] <Philip`>
http://www.w3.org/2009/08/WebFonts/charter says just WOFF

[11:54:00.0000] <TabAtkins>
Ah damn, that did get taken out, didn't it.

[11:54:01.0000] <TabAtkins>
;_;

[11:54:02.0000] <roc>
you say that like it's a bad thing

[11:55:00.0000] <TabAtkins>
Everyone has a stupid kneejerk reaction to CWT just because it's an EOT version.  It's just a custom header on top of a TTF file!

[11:55:01.0000] <JonathanNeal>
http://downforeveryoneorjustme.com/whatwg.org

[11:55:02.0000] <JonathanNeal>
"It's not just you! http://whatwg.org  looks down from here."

[11:55:03.0000] <annevk>
TabAtkins, and for a reason

[11:56:00.0000] <TabAtkins>
annevk: Reason being?

[11:56:01.0000] <annevk>
TabAtkins, adding complexity and obfuscation to TTF just for poltical reasons is silly

[11:56:02.0000] <TabAtkins>
Well, no, it's for compat reasons.  CWT is usable in IE6+.  It's a useful variant of TTF.

[11:57:00.0000] <TabAtkins>
It happens to also fulfill the "light obfuscation" thing that some vendors want, but that's not its reason for existing.

[11:57:01.0000] <AryehGregor>
What ever happened to the same-origin thing?

[11:58:00.0000] <TabAtkins>
Fonts are supposed to be same-origin only, modulated by CORS.

[11:58:01.0000] <roc>
CWT is useless because to enforce the same-origin stuff that font vendors want, you have to use Referer checking

[11:58:02.0000] <AryehGregor>
When I left www-font, the prevailing objection was that either you made the root string blank and IE would serve it from any domain, or you didn't and other browsers would ignore root strings of existing files, which is also bad.

[11:58:03.0000] <annevk>
I was talking about WOFF

[11:58:04.0000] <annevk>
also, I'm opposed to abusing CORS as I previously explained

[11:58:05.0000] <AryehGregor>
Well, and also if you used the root string, people would have to actually maintain it to get it to work with IE, but I guess that's no worse than just using EOT.

[11:59:00.0000] <annevk>
CWT seems even more silly

[11:59:01.0000] <TabAtkins>
roc: It's exactly equivalent in restriction to the other formats; the same level of (non)restriction is present if you server TTF or WOFF.

[11:59:02.0000] <AryehGregor>
My position was always that if there was any solution that allowed one font file to be served to everyone, take it, however hacky.

[12:00:00.0000] <TabAtkins>
AryehGregor: Exactly, though CWT is at least minimally hacky.  You don't even have a rootstring. (Not even a rootstring hidden in padding, in the current proposed version.)

[12:00:01.0000] <roc>
TabAtkins: not so. we can add convenient same-origin restrictions to TTF and WOFF (and have, in Firefox). that is not an option when you serve CWT to IE.

[12:00:02.0000] <AryehGregor>
TabAtkins, then IE just accepts it from any domain?

[12:01:00.0000] <AryehGregor>
But we figure it's not such a big deal because it will fail in Firefox, so people won't be enthusiastic to do that?

[12:01:01.0000] <TabAtkins>
roc: That's only a problem if people are okay with their fonts *only* working in legacy IE.

[12:01:02.0000] <TabAtkins>
AryehGregor: Yeah.

[12:01:03.0000] <AryehGregor>
Seems reasonable enough to me.

[12:01:04.0000] <AryehGregor>
Although there are other problems with IE, IIRC, like not supporting italic/bold fonts easily?  I got out of this a long time ago.

[12:02:00.0000] <AryehGregor>
Glad to hear that a solution was reached that's acceptable to both Mozilla and MS.

[12:02:01.0000] <roc>
If a font license requires the author to protect the font from cross-origin access, and the author doesn't but "it's OK because only IE can access the font cross-origin", how many corporate legal departments would be OK with that?

[12:02:02.0000] <AryehGregor>
Now if only we could solve the "page doesn't render while font downloads" problem.

[12:02:03.0000] <TabAtkins>
Legacy IEs have buggy @font-face support, but you can work around it.

[12:02:04.0000] <zcorpan>
there's no event when 'buffered' changes because cached data has been thrown away

[12:03:00.0000] <zcorpan>
so it's hard to know when to update the UI

[12:03:01.0000] <AryehGregor>
roc, do any foundries state things exactly that way?  Or do they say exactly what technologies they permit?

[12:03:02.0000] <TabAtkins>
roc: Any website which leeches your font won't work in a large percentage of browsers.

[12:03:03.0000] <TabAtkins>
It's just not a good deal.

[12:03:04.0000] <AryehGregor>
Anyway, has anyone thought about progressive rendering for fonts?

[12:03:05.0000] <zcorpan>
should we fire 'progress' in that case?

[12:03:06.0000] <AryehGregor>
Like putting all the ASCII characters in the front and ensuring that the browser can render those right away?  Is this impossible with TTF?

[12:03:07.0000] <AryehGregor>
I guess the Chinese are out of luck regardless.  :)

[12:03:08.0000] <TabAtkins>
AryehGregor: You can get mild progressive rendering by ordering the tables correctly.  I think FF already does that somewhat?

[12:04:00.0000] <AryehGregor>
It still has FOUC-like effects.

[12:04:01.0000] <AryehGregor>
Unless you use heavy subsetting, maybe?

[12:04:02.0000] <TabAtkins>
WOFF organizes the font in such a way as to present some layout information immediately.

[12:04:03.0000] <AryehGregor>
How big is a font with only basic Latin, a few kilobytes?  If it fits in a single TCP window . . .

[12:04:04.0000] <roc>
AryehGregor: I believe that's how it works (or rather, is going to work). I poked and prodded the font vendors who supported CWT/EOT to try to get specific details, they were not cooperative

[12:05:00.0000] <AryehGregor>
roc, maybe if you were a customer they'd be more willing to explain.  :)

[12:05:01.0000] <TabAtkins>
They're still too tied up in their legal department wranglings.  >_<

[12:05:02.0000] <roc>
TabAtkins: I understand that, but ignoring font licensing requirements because "the font vendor isn't going to care in practice" isn't going to be acceptable to lawyers in general

[12:06:00.0000] <AryehGregor>
No, but if they don't mind in principle, and enough customers ask them about it, they'll tell their legal department to allow it explicitly.

[12:06:01.0000] <TabAtkins>
roc: We'd need details on exactly what they're trying to prevent, though.  If it's generic enough, then just allowing your font to be downloaded with wget may be enough to violate the license.

[12:06:02.0000] <roc>
in theory the font vendors could carve out an exception in their licenses, but my suggestive prodding failed to elicit such a plan

[12:06:03.0000] <AryehGregor>
One Mozilla developer asking them, however awesome he is, is probably not enough to get them to ask a lawyer to look at it.  :)

[12:07:00.0000] <AryehGregor>
But if a bunch of customers ask, or one large customer, that would be a different story.

[12:07:01.0000] <roc>
what if the answer to the question might affect Mozilla's support for their font format?

[12:07:02.0000] <TabAtkins>
I should prod them too.

[12:08:00.0000] <TabAtkins>
roc: Well, since their non-answer's effect seems to be "Mozilla wont' support it", they don't have much to lose.  ^_^

[12:08:01.0000] <roc>
sure

[12:08:02.0000] <roc>
good luck

[12:08:03.0000] <AryehGregor>
roc, then you're deadlocked.  This is what fora like a Fonts WG are supposed to prevent.  :)

[12:08:04.0000] <roc>
we're not deadlocked

[12:08:05.0000] <paul_irish>
what's the prodding for? i have a few good contacts.

[12:09:00.0000] <roc>
so have I

[12:09:01.0000] <AryehGregor>
paul_irish, would EOT with no root strings be okay, if non-IE browsers implemented it with cross-origin restrictions?

[12:09:02.0000] <roc>
they could change the landscape by just announcing that they will license CWT fonts in a way that lets you deploy them on IE without any cross-origin protection

[12:09:03.0000] <TabAtkins>
paul_irish: Seeing if it would be acceeptable to font foundries for a website to serve CWT, which will be same-origin protected on modern browsers but not on legacy IEs.

[12:09:04.0000] <roc>
I asked them to make such an announcement

[12:09:05.0000] <roc>
they didn't do so

[12:09:06.0000] <roc>
<shrug>

[12:10:00.0000] <paul_irish>
didnt FontFont just announce their licensing their work for CWT and woff only?

[12:10:01.0000] <zcorpan>
http://simon.html5.org/temp/2d0zbqtzv.html - not finished, but feedback welcome (i'll read the logs)

[12:11:00.0000] <paul_irish>
and ascender, of course, is behind CWT.. i havent seen much foundry-based opposition to it

[12:12:00.0000] <AryehGregor>
"Opera requires that your video file is served as video/ogg for it to play."  Why?

[12:12:01.0000] <zcorpan>
that's not entirely accurate; we also accept application/ogg and audio/ogg and audio/wav etc

[12:12:02.0000] <zcorpan>
but text/plain and text/html etc are rejected

[12:12:03.0000] <zcorpan>
because the spec says so

[12:13:00.0000] <AryehGregor>
Oh, feh.

[12:13:01.0000] <AryehGregor>
Why does the spec say so?  These aren't script, are there security problems?

[12:14:00.0000] <AryehGregor>
Apache seems to serve application/ogg by default for Ogg, or at least a rule in /etc/apache2/magic seems to say so.

[12:14:01.0000] <AryehGregor>
Still not sure why this is necessary.

[12:15:00.0000] <zcorpan>
it's just to avoid mislabeled content

[12:15:01.0000] <AryehGregor>
Which will fail anyway when you feed it to GStreamer, no?

[12:16:00.0000] <AryehGregor>
The server might mislabel, the video player will know for sure whether it can play the file.

[12:16:01.0000] <AryehGregor>
Server-set MIME types should be treated as hints of intent, not an actual description of what the content is, because how should the server know that?  But maybe there's a good reason here.

[12:17:00.0000] <zcorpan>
sure, but if we play text/plain videos, then we need to sniff for video for text/plain if we want to be able to play mislabeled videos when loaded directly

[12:17:01.0000] <zcorpan>
plus, we want to reject video/mp4 if we can't play mpeg-4

[12:17:02.0000] <AryehGregor>
Well, I assume you already do other types of sniffing there anyway, so why not?

[12:18:00.0000] <zcorpan>
so it's not much effort to also reject text/plain

[12:18:01.0000] <zcorpan>
because we don't know what gstreamer supports so we don't know what to sniff for

[12:18:02.0000] <AryehGregor>
/me doesn't get what the benefit is to users or authors that outweighs the annoyance of authors potentially having to configure their servers.

[12:18:03.0000] <AryehGregor>
So let GStreamer sniff.  It presumably returns an error graciously if it can't play it, right?

[12:19:00.0000] <AryehGregor>
GStreamer is the only part of the system that knows for sure what can be played.

[12:19:01.0000] <AryehGregor>
Once you've already received the HTTP headers, you've received the start of the content too, so just feed that to GStreamer and that will be a lot more reliable than guessing based on header.

[12:19:02.0000] <AryehGregor>
Of course, it makes sense to sniff based on MIME type in the HTML, because that way you can avoid unnecessary requests.

[12:24:00.0000] <AryehGregor>
zcorpan, you should mention that old browsers may return "no" from canPlayType().  Your examples should be updated to reflect that too.

[12:27:00.0000] <zcorpan>
AryehGregor: old video-supporting browsers return bogus results for canPlayType anyway, iirc

[12:31:00.0000] <zcorpan>
added a note

[12:31:01.0000] <roc>
Safari 3

[12:39:00.0000] <asmodai>
/me eyes Google Docs...

[12:39:01.0000] <asmodai>
Am I the only one for which their spreadsheet is acting weird on FF 3.6?

[12:55:00.0000] <zcorpan>
nn

[13:06:00.0000] <hober>
hsivonen: http://blogs.msdn.com/ie/archive/2010/03/02/how-ie8-determines-document-mode.aspx

[13:07:00.0000] <hober>
I think http://ieblog.members.winisp.net/images/MarcSil_IE8_Document_Mode_2.png looks even more complicated than http://hsivonen.iki.fi/doctype/ie8-mode.png

[13:18:00.0000] <mpilgrim>
/me catches up on the font format discussion

[13:19:00.0000] <mpilgrim>
yeah... this isn't doing much to change my opinion of the font foundries

[13:19:01.0000] <JonathanNeal>
I'm not sure the context for <menu type="toolbar"> and <menu type="context menu"> after reading the docs.  I have visible buttons in the upper-righthand area of an application, so for the visible buttons I used <menu type="toolbar">, and then for the drop down menus I used <menu type="context menu">

[13:19:02.0000] <JonathanNeal>
Does that sound right?

[13:19:03.0000] <asmodai>
mpilgrim: Obssessive people? :)

[13:20:00.0000] <TabAtkins>
JonathanNeal: I think context menu <menu>s are intended only for actual context menus; that is, right-click menus.

[13:21:00.0000] <JonathanNeal>
I wasn't sure if context could be triggered by left clicking an icon.

[13:23:00.0000] <TabAtkins>
Well, that's supposed to be handled by the UA.

[13:24:00.0000] <JonathanNeal>
Sure, well in that case, I will leave it blank for the "menu" role.

[13:24:01.0000] <TabAtkins>
Ideally, the UA exposes the commands in a UA-specific manner when the user asks for a context menu.

[13:24:02.0000] <TabAtkins>
Yeah.

[13:24:03.0000] <TabAtkins>
/me should put together a toy impl of that tonight.

[13:27:00.0000] <MikeSmith>
AryehGregor: about the validator question you asked, the place to discuss that would be on the public-qa-dev list or www-validator list

[13:28:00.0000] <MikeSmith>
there really is not currently much of a team working actively on maintaining the existing validator

[13:28:01.0000] <MikeSmith>
it mostly one guy, Ville Skyttä

[13:29:00.0000] <JonathanNeal>
Thanks TabAtkins, you can see how I implemented it @ http://sandbox.thewikies.com/html5-layout/ (read source code on line 247+ for documentation)

[13:29:01.0000] <JonathanNeal>
It's cross browser too.

[13:30:00.0000] <Philip`>
hober: Ooh, nice that they're documenting it in some actual detail now

[13:30:01.0000] <MikeSmith>
AryehGregor: what you suggest sounds like something that could really be a feature of validator.nu itself

[13:31:00.0000] <Philip`>
(Still seem to be entirely missing the details about how they determine the modes for doctypes, though)

[13:44:00.0000] <roc>
my opinion of the font vendors ("foundries" implies an unwarranted special status IMHO) is pretty low too. However I think it's worth making a small compromise to get interoperable Web fonts more widely accepted, faster. WOFF is so simple, it's a very small compromise indeed from my point of view.

[13:46:00.0000] <TabAtkins>
Sigh.  "I'd like light ranch dressing, please." "Ok, ranch dressing." "LIGHT ranch." "Ok, here you go, italian dressing."

[14:07:00.0000] <mpilgrim>
i'm just waiting for the first firefox extension that notices embedded WOFF fonts, automatically converts them to TTF, installs them in your local font directory, and puts up a toaster-style notification saying "Congratulations, your font library just got expanded!" Preferably with an icon of an "R" wearing a pirate patch.

[14:08:00.0000] <TabAtkins>
Why the R?

[14:09:00.0000] <roc>
sure

[14:09:01.0000] <roc>
perhaps Linux systems will get native support for WOFF too

[14:09:02.0000] <mpilgrim>
wouldn't have to be an "R". could be an "A". i guess the concept of "typography" is used iconified using an "A", isn't it?

[14:09:03.0000] <roc>
doesn't matter

[14:09:04.0000] <TabAtkins>
Ah, got it.

[14:09:05.0000] <mpilgrim>
but an "R" with a pirate patch would probably look cooler

[14:10:00.0000] <mpilgrim>
anyway, the entire thing is an exercise in making copying bits less convenient

[14:10:01.0000] <mpilgrim>
that never ends well, regardless of good intentions

[14:10:02.0000] <TabAtkins>
Well, WOFF actually comes with some nice benefits over TTF.

[14:10:03.0000] <TabAtkins>
CWT doesn't have any direct benefits over TTF, but it good simply because of increased compat.

[14:11:00.0000] <mpilgrim>
does one of them include "native support on every major computing platform on the planet"?

[14:11:01.0000] <TabAtkins>
No?

[14:11:02.0000] <roc>
once you unwrap and gunzip, yes

[14:11:03.0000] <TabAtkins>
Well, yes.

[14:11:04.0000] <mpilgrim>
i serve my embedded TTF fonts gzipped already

[14:13:00.0000] <roc>
WOFF reorders the tables and compresses them independently, which could be helpful if you want to analyze just the CMAP

[14:15:00.0000] <mpilgrim>
could i not do that with a TTF file directly? (non-rhetorical question)

[14:18:00.0000] <roc>
you could reorder the tables and gzip the whole thing. Then the browser could read the CMAP quicker, but when you wanted the other tables you'd have to re-uncompress the whole file from scratch, unless you saved the gzip state. It's considerably more complex.

[14:19:00.0000] <TabAtkins>
Sigh.  Why do tables have to act so weird?  Specifically with respect to floating and positioned descendants.

[14:21:00.0000] <mpilgrim>
thanks, roc

[14:22:00.0000] <mpilgrim>
and how does this all help the font vendors in their quest to make bits less copyable?

[14:22:01.0000] <mpilgrim>
i.e. why are they behind such a format?

[14:22:02.0000] <roc>
simply that you cannot download a WOFF font and drop it in your Fonts folder and have it work

[14:22:03.0000] <roc>
that's all

[14:22:04.0000] <TabAtkins>
It's a "garden fence", for now.

[14:22:05.0000] <roc>
well

[14:23:00.0000] <roc>
I guess there's also the fact that the only browser that implements WOFF today has a default same-origin restriction, so it's easy for authors to comply with font licenses that require them to protect fonts from cross-site linkage. But strictly speaking that's an author benefit.

[14:26:00.0000] <TabAtkins>
So, roc, am I restating your objection to CWT correctly if I say that it's *too* interoperable; eg, the problem is that it works in browsers that don't have same-origin restrictions?

[14:27:00.0000] <roc>
I wouldn't put it that way

[14:28:00.0000] <roc>
I'd say that IE has origin restrictions for fonts, but CWT forbids you from using them

[14:29:00.0000] <TabAtkins>
I'd say that's at least as biased a phrasing as what I provided.  ^_^

[14:29:01.0000] <roc>
definitely :-)

[14:30:00.0000] <TabAtkins>
It also makes it seem like everything would be better if only CWT allowed you to use IE's origin-restriction mechanism, but in fact allowing that mechanism was one of your objections to the earlier CWT draft, iirc.

[14:31:00.0000] <roc>
IIRC I have not objected to having CWT say that the header is opaque and hence may contain data that IE would interpret as a rootstring

[14:31:01.0000] <TabAtkins>
All right, I may be risremembering.  I know that several people *did* object to precisely that.

[14:31:02.0000] <roc>
yes

[14:32:00.0000] <roc>
I may be misremembering too

[14:32:01.0000] <TabAtkins>
/me doesn't want to comb through his archives to find the answer.

[14:32:02.0000] <roc>
such an approach has some problems, like the fact that different browsers using different access control policies would be suboptimal

[14:32:03.0000] <TabAtkins>
Indeed.

[14:33:00.0000] <Philip`>
/me wonders if anyone happened to notice that Microsoft removed the 5000-byte name string limit (which broke lots of fonts that embed the Open Font License) in a security update recently

[14:33:01.0000] <TabAtkins>
Ooh, I didn't.  Good to know.

[14:33:02.0000] <TabAtkins>
Augh, god, SHODAN keeps scaring me.

[14:34:00.0000] <TabAtkins>
I have her flashing for a fraction of a second every few minutes on the GLaDOS system at work.

[14:44:00.0000] <TabAtkins>
/me is pissed that he has to wrap the contents of a <td> in a <div height:100%> just to provide a positioning root.

[14:55:00.0000] <othermaciej>
TabAtkins: can't you just put position:relative on the TD?

[14:55:01.0000] <TabAtkins>
othermaciej: Nope.  Not working in FF3.6, at least.  It *should* work, but it's not.

[14:56:00.0000] <othermaciej>
roc: we're probably gonna support double-clicking WOFF fonts to install them, same as for OpenType

[14:56:01.0000] <othermaciej>
roc: if only because it would be extra work not to

[14:56:02.0000] <othermaciej>
so we see WOFF as pure waste

[14:57:00.0000] <othermaciej>
TabAtkins: I know Firefox has a problem with the table not being eligible to be a containing block for absolute positioned content, did not know there was an issue with cells

[14:58:00.0000] <roc>
cells don't necessarily support relative positioning in CSS 2.1

[14:58:01.0000] <roc>
"The effect of 'position:relative' on table-row-group, table-header-group, table-footer-group, table-row, table-column-group, table-column, table-cell, and table-caption elements is undefined."

[14:58:02.0000] <TabAtkins>
Argh, that is stupid and wrong.

[14:58:03.0000] <TabAtkins>
Presumably bugwards compatibility.

[14:58:04.0000] <paul_irish>
othermaciej: seriously? doubleclick to install woff fonts?

[14:59:00.0000] <othermaciej>
it's what we do for OpenType

[14:59:01.0000] <othermaciej>
and our easiest path to WOFF is to treat them same as any other font in the font system

[14:59:02.0000] <roc>
I'm surprised that that's the easiest thing to do, but OK

[15:00:00.0000] <othermaciej>
well, we could have a layer to translate from WOFF to TrueType in WebKit, if we were specifically motivated to prevent these fonts from working in apps that don't use WebKit for display

[15:00:01.0000] <othermaciej>
though increasingly more apps use WebKit for display, so that wouldn't even be very effective

[15:01:00.0000] <othermaciej>
TabAtkins: no, just the fact that tables are underdefined in CSS

[15:01:01.0000] <roc>
I don't see how your support work work if it doesn't do that. You'd add WOFF support to Quartz?

[15:01:02.0000] <TabAtkins>
othermaciej: Indeed, they are.  And that's stupid and wrong.  ^_^

[15:01:03.0000] <TabAtkins>
Long-term goal: overdefine CSS.

[15:02:00.0000] <roc>
it looks like Webkit doesn't actually support relative positioning on table cells

[15:03:00.0000] <roc>
but position:relative still makes the cell a container for abs-pos elements

[15:03:01.0000] <TabAtkins>
Argleasdjf;alf

[15:03:02.0000] <roc>
oh hang on

[15:03:03.0000] <roc>
Webkit behaves exactly like Firefox here

[15:04:00.0000] <roc>
totally ignores position:relative on cells

[15:04:01.0000] <othermaciej>
that's believable

[15:04:02.0000] <roc>
the cell doesn't become an abs-pos container

[15:04:03.0000] <roc>
ok, everyone move along

[15:04:04.0000] <othermaciej>
seems like it is useful to make a cell be an absolute positioned containing block

[15:04:05.0000] <TabAtkins>
It is very useful.

[15:04:06.0000] <TabAtkins>
I am making a calendar right now which could use it.

[15:04:07.0000] <othermaciej>
roc: same parts of the system that support OpenType/TrueType would support WOFF

[15:05:00.0000] <othermaciej>
on Mac

[15:05:01.0000] <othermaciej>
at least that is our current tentative plan

[15:05:02.0000] <roc>
ok

[15:05:03.0000] <TabAtkins>
<td><div height:100%; position:relative;>foo</div></td> works, but is obviously stupid.

[15:08:00.0000] <roc>
I presume, though, you must have some support in Webkit for font formats, since you need it for SVG fonts, so I presume you have some good reason to not add WOFF support there (since all ports would benefit from that)

[15:09:00.0000] <roc>
(now SVG fonts --- *that*'s a pure waste :-) )

[15:09:01.0000] <Rik`>
iirc, the iphone only supports svg fonts :(

[15:13:00.0000] <othermaciej>
roc: we might also add it there for ports where we can't change the font system, but on operating systems controlled by Apple the long-term goal would be to make it just another font format

[15:13:01.0000] <othermaciej>
the iPhone only supports SVG fonts as Web fonts, currently anyway

[15:14:00.0000] <Rik`>
othermaciej: do you know why ?

[15:14:01.0000] <roc>
that's unfortunate, since SVG fonts are pretty bad

[15:14:02.0000] <othermaciej>
Rik`: still evaluating security / bandwidth / perf impact of OpenType

[15:14:03.0000] <othermaciej>
it may change in the future, it may not, that is all I can say

[15:15:00.0000] <othermaciej>
interesting side note: some iTunes LPs and iTunes Extras use SVG fonts

[15:15:01.0000] <roc>
do you know why?

[15:15:02.0000] <othermaciej>
as a cheapass way to do subsetting

[15:15:03.0000] <roc>
weird

[15:18:00.0000] <Rik`>
why supporting OTF on the Mac if there is still a security evaluation on the iPhone ?

[15:22:00.0000] <TabAtkins>
At the upcoming CSS ftf, we're totally going to have to reintroduce display-role and display-model (though maybe as -outside and -inside, to be more intuitive).

[15:22:01.0000] <TabAtkins>
Otherwise, how will I ever make a table-cell also use Template Layout?

[15:23:00.0000] <roc>
I think I'd like that

[15:24:00.0000] <TabAtkins>
/me wants to use Template and Flexbox, or their spiritual successors, *so bad*.

[15:32:00.0000] <Philip`>
Rik`: Maybe security is stricter on the iPhone than on the Mac, because it needs to prevent users doing dastardly things such as choosing to run unauthorised programs

[15:32:01.0000] <Philip`>
(Or maybe there's more sensible reasons)


2010-03-03
[18:59:00.0000] <MikeSmithX>
Hixie: about shortnames on for the WDs to publish tomorrow, annevk mentioned "I believe we decided that md would be microdata", but I can't recall a decision about that. Can you?

[19:07:00.0000] <othermaciej>
MikeSmithX: hmm, the TransReq asked for approval of a new shortname but did not cite the shortname

[19:07:01.0000] <othermaciej>
I do recall discussing "microdata" as the shortname

[19:08:00.0000] <MikeSmithX>
othermaciej: well, I think it would be an improvement for clarity to use it as the shortname instead

[19:08:01.0000] <othermaciej>
I think so too

[19:08:02.0000] <MikeSmithX>
also, another issue is the URI "http://www.w3.org/TR/markup/"

[19:09:00.0000] <othermaciej>
our approval does not seem to have record of a shortname, nor was Sam specific about what shortname he was requesting

[19:09:01.0000] <othermaciej>
yeah H:TML probably should not use that as a shortname

[19:09:02.0000] <MikeSmith>
yeah, the lack of explicit mention of the shortname requests was my fault for not making it clear to Sam that they needed to be included

[19:10:00.0000] <MikeSmith>
about H:TML, I suggest we request http://www.w3.org/TR/html-markup/ instead

[19:10:01.0000] <MikeSmith>
about microdata, I suggest we consider qualifying that short name further as well

[19:11:00.0000] <MikeSmith>
I we want consistency with the existing RDFa shortname, it would be http://www.w3.org/TR/microdata-in-html/

[19:12:00.0000] <MikeSmith>
if we don't care too much about consistency, http://www.w3.org/TR/html-microdata/

[19:13:00.0000] <MikeSmith>
and perhaps http://www.w3.org/TR/canvas-2dcontext/

[19:13:01.0000] <MikeSmith>
or even http://www.w3.org/TR/canvas-2d/

[19:14:00.0000] <othermaciej>
html-microdata is getting kinda long for a short name

[19:14:01.0000] <othermaciej>
I would say just microdata - if there are other specs to disambiguate later, they can get a longer name

[19:14:02.0000] <MikeSmith>
OK

[19:14:03.0000] <MikeSmith>
how about "2dcontext"?

[19:14:04.0000] <othermaciej>
canvas-2d or canvas2d sounds better to me than 2dcontext, if not qualified otherwise

[19:14:05.0000] <MikeSmith>
OK

[19:15:00.0000] <MikeSmith>
I can send mail to the webmaster (reply to Sam's pub request) shortly

[19:15:01.0000] <MikeSmith>
othermaciej: are you OK with http://www.w3.org/TR/html-markup/ ?

[19:16:00.0000] <othermaciej>
MikeSmith: what shortname does the diffs document get published with?

[19:16:01.0000] <othermaciej>
just for comparison

[19:16:02.0000] <MikeSmith>
http://www.w3.org/TR/html5-diff/

[19:20:00.0000] <MikeSmith>
<twitter!squaredeye> .@w3c Are their any specs in future releases of CSS for blending modes with colors?

[19:20:01.0000] <MikeSmith>
...not sure what the even means

[19:20:02.0000] <MikeSmith>
TabAtkins: ↑

[19:21:00.0000] <othermaciej>
html-markup sounds ok to me

[19:21:01.0000] <othermaciej>
MikeSmith: maybe he means ability to do Porter-Duff compositing operations other than SrcOver?

[19:21:02.0000] <othermaciej>
or, like, photoshop blend modes, like dodge and burn

[19:21:03.0000] <MikeSmith>
ah

[19:24:00.0000] <TabAtkins>
Mikesmith: Not sure what the dude means, but othermaciej seems like he has a decent handle on it.

[19:24:01.0000] <MikeSmith>
OK

[19:24:02.0000] <TabAtkins>
Assuming that's approximately right, then no.

[19:25:00.0000] <TabAtkins>
At least, nothing in the pipeline about how colors blend.

[19:25:01.0000] <othermaciej>
blend modes are not really a color issue so he might just be confused

[19:25:02.0000] <TabAtkins>
If you can reply and ask for more details, I can give a better answer.

[19:26:00.0000] <MikeSmith>
maybe he meant more like "blending modes, with colors" .. though not really clear, either 

[19:26:01.0000] <MikeSmith>
TabAtkins: yeah, I will

[19:26:02.0000] <TabAtkins>
Back to Mass Effect and beer.  Basically the best night possible.

[19:27:00.0000] <MikeSmith>
ah, I now notice he also tweeted, ".@webkit are their any experiments with blending modes (ie: multiply) with colors in the browser?"

[19:27:01.0000] <MikeSmith>
TabAtkins: what's Mass Effect?

[19:27:02.0000] <othermaciej>
I assume he means blend modes then

[19:27:03.0000] <MikeSmith>
yeah

[19:27:04.0000] <MikeSmith>
http://twitter.com/squaredeye

[19:28:00.0000] <othermaciej>
and we do have -webkit-background-composite

[19:28:01.0000] <othermaciej>
to set the compositing mode

[19:28:02.0000] <othermaciej>
in addition to having different compositing modes in canvas

[19:35:00.0000] <paul_irish>
othermaciej: earlier on the WOFF topic.. i just wanted to clarify.. you said it'd be easier to just make woff fonts installable on mac os, right?

[19:36:00.0000] <othermaciej>
paul_irish: it is simpler for us to treat them like any other font than to try to make the font system only handle them for WebKit

[19:37:00.0000] <paul_irish>
roger that. i'm just kinda wtf'ing because i thought the intention of the woff format was to avoid a trivial-to-install scenario

[19:37:01.0000] <othermaciej>
I'm sure that is some people's intent, I don't think there is any such requirement in the spec though

[19:38:00.0000] <paul_irish>
hehe. this is true. :) 

[19:39:00.0000] <paul_irish>
othermaciej: is it okay if i mention this to some webfont friends of mine?

[19:48:00.0000] <othermaciej>
paul_irish: then it might sounds like I am making a commitment about future Apple products, which I definitely am not

[19:48:01.0000] <othermaciej>
we haven't even started on implementing WOFF yet

[19:49:00.0000] <paul_irish>
gotcha.

[19:49:01.0000] <paul_irish>
my expectation is that people wont be very pleased. :)

[19:49:02.0000] <othermaciej>
I can't even commit to implementing it at all ever

[19:49:03.0000] <paul_irish>
k. thanks much.

[19:50:00.0000] <othermaciej>
if we do, then I am not sure we have huge motive to make such fonts hard to install

[19:50:01.0000] <othermaciej>
if we end up participating in the WG then we may be able to discuss our plans

[19:50:02.0000] <othermaciej>
in more specifics

[19:50:03.0000] <othermaciej>
if we have plans by then

[19:50:04.0000] <paul_irish>
aye

[19:50:05.0000] <othermaciej>
I expect font people are most worried about piracy on Windows

[19:50:06.0000] <othermaciej>
Windows is the main platform where piracy happens, from what I hear

[19:51:00.0000] <othermaciej>
at least for other media

[19:51:01.0000] <paul_irish>
i pirate on mac all the time. :)

[19:51:02.0000] <miketaylr>
shhhh

[19:51:03.0000] <paul_irish>
oh fu-

[19:59:00.0000] <jcranmer>
most people I know don't like macs

[20:02:00.0000] <paul_irish>
especially not after today. sheesh

[20:08:00.0000] <othermaciej>
why, did a Mac go rogue and kill its trainer today?

[21:18:00.0000] <TabAtkins>
MikeSmith: http://en.wikipedia.org/wiki/Mass_Effect

[21:20:00.0000] <MikeSmith>
TabAtkins: I see.. glad it's not this: http://en.wikipedia.org/wiki/Mass_effect_(medicine)

[21:23:00.0000] <TabAtkins>
Heh, indeed.

[21:31:00.0000] <MikeSmith>
hmm, I don't think Jonas's assertion that "Validation only checks for a certain, very limited, types of errors." is entirely accurate

[21:32:00.0000] <othermaciej>
it only checks for syntax errors

[21:32:01.0000] <othermaciej>
not other kinds of errors

[21:34:00.0000] <MikeSmith>
I would not classify syntax errors as "very limited types of errors"

[21:34:01.0000] <othermaciej>
Jonas is probably speaking from the programmer's perspective

[21:34:02.0000] <othermaciej>
the errors we spend most of our time dealing with are not syntax errors

[21:35:00.0000] <MikeSmith>
right, I understand that

[21:35:01.0000] <MikeSmith>
but a markup validator is essentially a lint checker

[21:35:02.0000] <othermaciej>
right

[21:35:03.0000] <MikeSmith>
a static checker

[21:36:00.0000] <othermaciej>
I think static checkers, lint-like tools, and validators are all awesome things

[21:36:01.0000] <othermaciej>
yet I agree with Jonas that the errors they can find are but a subset of the possible interesting errors

[21:36:02.0000] <othermaciej>
though, for document-like content it's probably a higher fraction of errors you may care about

[21:37:00.0000] <othermaciej>
than for app-type content

[21:37:01.0000] <othermaciej>
we compile WebKit at a crazy high warning level

[21:37:02.0000] <MikeSmith>
yeah, agreed

[21:39:00.0000] <MikeSmith>
checking/validating the initial markup of some particular a Web app that changes state is of limited usefulness 

[21:53:00.0000] <othermaciej>
our warning flags: -Werror -Wmissing-prototypes -Wnon-virtual-dtor -Wnewline-eof -Wall -Wextra -Wcast-align -Wcast-qual -Wchar-subscripts -Wextra-tokens -Wformat=2 -Winit-self -Wmissing-format-attribute -Wmissing-noreturn -Wpacked -Wpointer-arith -Wredundant-decls -Wundef -Wwrite-strings -Wno-trigraphs

[21:53:01.0000] <othermaciej>
(-Werror makes the build break if someone violates one of these)

[21:57:00.0000] <JonathanNeal>
Hey all

[22:31:00.0000] <MikeSmith>
othermaciej: I realize that the TAG specifically suggested a RelaxNG schema and that it should be an appendix to the H:TML document

[22:31:01.0000] <MikeSmith>
and that may be what they think they want

[22:31:02.0000] <othermaciej>
MikeSmith: I just filed the bug based on what they said - if something else would make them happy, I have no complaint

[22:31:03.0000] <MikeSmith>
but I don't think that's what they really want

[22:32:00.0000] <othermaciej>
actually, I'll go farther - I don't even care if we do make them happy, as long as they get the same due process right to be heard as anyone else

[22:33:00.0000] <MikeSmith>
OK, well, if we keep that bug summary as specifically adding and appendix with and RelaxNG schema to the H:TML doc, then what I am inclined to do is change it to resolved/wontfix with a comment that I don't plan to add it

[22:33:01.0000] <MikeSmith>
I dunno, maybe that's the best way to do it

[22:34:00.0000] <othermaciej>
or you could make it "partially accepted" with a resolution of doing something different than what was asked

[22:34:01.0000] <othermaciej>
or you could get the TAG's agreement on repurposing the bug

[22:34:02.0000] <othermaciej>
many options

[22:34:03.0000] <MikeSmith>
OK, yeah, I suppose I should follow up with them about it directly on the TAG list

[22:35:00.0000] <MikeSmith>
I will suggest to them that we repurpose that bug

[22:38:00.0000] <othermaciej>
do you have a concrete alternate approach in mind?

[22:42:00.0000] <MikeSmith>
othermaciej: not really

[22:42:01.0000] <othermaciej>
hmm

[22:42:02.0000] <MikeSmith>
well, not short of just publishing the set of formalisms

[22:42:03.0000] <othermaciej>
I probably shouldn't be the first vote on these accessibility surveys

[22:43:00.0000] <othermaciej>
is there a way to delete my answer?

[22:43:01.0000] <MikeSmith>
othermaciej: yeah

[22:43:02.0000] <othermaciej>
ah, there is

[22:43:03.0000] <MikeSmith>
that brings to mind one other thing I wanted to ask about

[22:44:00.0000] <othermaciej>
I will register my opinion later if it seems appropriate

[22:44:01.0000] <othermaciej>
I do like both proposals

[22:44:02.0000] <Hixie>
MikeSmith: there are short names assumed in the various drafts, look at the "current version" urls

[22:44:03.0000] <MikeSmith>
othermaciej: I think it's likely we will get agreement to proceed with those

[22:45:00.0000] <MikeSmith>
Hixie: looking now

[22:45:01.0000] <othermaciej>
yeah, I am just looking to be supportive

[22:45:02.0000] <othermaciej>
but I do not want to inadvertently influence people for or against too much

[22:45:03.0000] <othermaciej>
I gotta go

[22:46:00.0000] <MikeSmith>
Hixie: ah, OK

[22:48:00.0000] <MikeSmith>
"microdata" seems fine, but I already sent mail suggesting "canvas-2d", so I guess I need to wait unless/until anybody replies to that

[22:49:00.0000] <MikeSmith>
if PLH or nobody else thinks it needs to be changed, then we can just leave it as-is

[22:50:00.0000] <Hixie>
canvas-2d is a bit misleading since there's no canvas element in the spec :-)

[22:53:00.0000] <MikeSmith>
Hixie: I don't think the shortname necessarily implies it's a spec for the canvas element

[22:54:00.0000] <MikeSmith>
and the abstract does read, "This specification defines the 2D Context for the HTML canvas element.

[22:54:01.0000] <MikeSmith>
and the abstract does read, "This specification defines the 2D Context for the HTML canvas element"

[22:54:02.0000] <MikeSmith>
oops

[22:54:03.0000] <MikeSmith>
anyway

[22:54:04.0000] <MikeSmith>
I don't feel strongly about it

[22:55:00.0000] <MikeSmith>
it's just that I let the genie out of the bottle already, as far as having proposed it, so I need to wait to see what the responses are

[22:55:01.0000] <MikeSmith>
I guess

[22:56:00.0000] <MikeSmith>
Hixie: anyway, I wanted to say something related to Maciej's comment that he probably should not be the first vote on the accessibility surveys

[22:59:00.0000] <MikeSmith>
Hixie: which is, I think if you genuinely want to help reduce the amount of unproductive discussion on the public-html list, one thing you can consider doing more of is waiting to post replies on threads that you know are going to be contentious

[23:01:00.0000] <nessy>
othermaciej: I made the mess larger by replying to your questions on the a11y list :-)

[23:18:00.0000] <Hixie>
MikeSmith: i almost always do

[23:20:00.0000] <nessy>
foolip: do you think the media track API is not ready yet?

[23:22:00.0000] <MikeSmith>
Hixie: OK. I think even it non-contentious it helps others to allow time for some people form their own opinions

[23:22:01.0000] <MikeSmith>
*non-contentious discussions

[23:46:00.0000] <Hixie>
MikeSmith: in general i only respond to threads weeks after they've finished, because i just stuff all incoming e-mail into my folders

[23:46:01.0000] <Hixie>
MikeSmith: actually for public-html mail now i tend to just not reply at all, and just wait for a bug to be filed

[23:46:02.0000] <Hixie>
occasionally there are exceptions, of course, like when someone asks me a question directly

[23:48:00.0000] <MikeSmith>
Hixie: that sounds reasonable to me

[23:49:00.0000] <MikeSmith>
we need to get the overall list traffic down on public-html

[23:50:00.0000] <MikeSmith>
I think moving more discussions to bugzilla is one good way to do that

[23:51:00.0000] <MikeSmith>
and encouraging people to comment in bugzilla instead

[04:09:00.0000] <asmodai>
mmm

[04:09:01.0000] <asmodai>
http://www.theregister.co.uk/2010/03/02/virgin_america_html_flash/

[04:09:02.0000] <asmodai>
Guess that made the rounds too?

[04:12:00.0000] <annevk>
not on this channel

[04:13:00.0000] <asmodai>
Hadn't seen it in my scrollback buffer, but hey, could've missed it. :)

[04:16:00.0000] <boblet>
anyone have an opinion on using <dl> for forms? (<dt> for labels, <dd> for <input> etc)

[04:18:00.0000] <hsivonen>
/me wishes Lufthansa dropped Flash, too

[04:18:01.0000] <Philip`>
boblet: Why not use <label> for labels and <input> for <input>?

[04:19:00.0000] <boblet>
Philip`: sorry, I meant putting <label> in <td> etc

[04:19:01.0000] <boblet>
sheesh

[04:19:02.0000] <Philip`>
Seems like anything else is just a presentation concern

[04:19:03.0000] <boblet>
in <dt>

[04:20:00.0000] <boblet>
Philip`: agreed, but replying to the latest HTML5Doctor Q&A, and wondering why <dl> feels wrong for me (I’d say a list, or maybe a table for a very complex form, eg multi-col)

[04:21:00.0000] <Philip`>
"Virgin anticipates moving this new site to HTML 5, once it's cleared standards ratification at the World Wide Web Consortium (W3C)" - in 2022?

[04:22:00.0000] <boblet>
I was impressed they were using it as a buzzword

[04:22:01.0000] <Philip`>
"the vast majority of web content [...] are built using [Flash]" - huh?

[04:22:02.0000] <boblet>
HTML5 = Web 2.1? or Web 3.0? ;-)

[04:25:00.0000] <jgraham>
Philip`: Yeah, the grammar in that quote sounds all wrong

[04:25:01.0000] <MikeSmith>
hey lazychannel, what's the hg command-line syntax for pulling a specific revision in mercurial?

[04:25:02.0000] <hsivonen>
hg pull -r 431432421 -u

[04:25:03.0000] <MikeSmith>
hsivonen: thanks

[05:03:00.0000] <zcorpan>
http://my.opera.com/core/blog/2010/03/03/everything-you-need-to-know-about-html5-video-and-audio-2

[05:03:01.0000] <zcorpan>
hmm, i managed to screw up the url. oh well

[05:03:02.0000] <gsnedders>
n00b

[05:08:00.0000] <Philip`>
Cool URIs don't change, but screwed up URIs should change so that they become cool

[05:09:00.0000] <mut>
hello

[05:09:01.0000] <boblet>
I dunno, I think you could add some more keywords in there yet

[05:09:02.0000] <boblet>
hey mut

[05:10:00.0000] <mut>
im trying to make a site with canvas

[05:10:01.0000] <mut>
well its all working pretty well for me so far

[05:10:02.0000] <boblet>
zcorpan: what about adding -ogg-vorbis-h264 at the end?

[05:10:03.0000] <mut>
but im using .translate(), i just wondered how do you reset the origin to 0,0 and clear the canvas? 

[05:14:00.0000] <mut>
like im drawing something based on user interaction based on a form, but when they select something, i wanna clear the canvas, and start again

[05:16:00.0000] <mut>
ah if i change/reset the width/height of the canvas it "resets" the canvas... :)

[05:18:00.0000] <Philip`>
mut: Usually you should use save()/restore() to reset the origin

[05:18:01.0000] <Philip`>
and clearRect to clear the bitmap

[05:18:02.0000] <Philip`>
though if you really want to reset the entire canvas then changing size will work too

[05:21:00.0000] <mut>
oh

[05:21:01.0000] <mut>
save() restore()

[05:21:02.0000] <mut>
hmm ill take a look

[05:22:00.0000] <mut>
so i save() when the canvas is clear, and then do restore()?

[05:24:00.0000] <Philip`>
They don't affect the canvas bitmap at all

[05:24:01.0000] <Philip`>
save() just remembers the current origin transformations, fill styles, etc

[05:24:02.0000] <Philip`>
and restore() resets them back to the last saved values

[05:24:03.0000] <Philip`>
Oh, wait

[05:24:04.0000] <Philip`>
That's what you said

[05:24:05.0000] <Dashiva>
Wow, Jobs actually called animating an object "original technology"?

[05:25:00.0000] <mut>
ok

[05:25:01.0000] <Philip`>
so, yes :-)

[05:25:02.0000] <mut>
errm but when i try that, the canvas dosent clear

[05:25:03.0000] <mut>
oh i nead to clearrect toto

[05:25:04.0000] <mut>
too

[05:25:05.0000] <mut>
heh

[05:25:06.0000] <Dashiva>
Is clearrect cheaper than .width = .width?

[05:26:00.0000] <Philip`>
Dashiva: Yes, in theory, because it has to do less work

[05:26:01.0000] <Philip`>
(although implementations might be suboptimal and do more work in the common case, in order to handle more general cases)

[05:27:00.0000] <mut>
philip, i have used transform to move things around, but i want to reset the origin to the canvas 0,0

[05:27:01.0000] <mut>
err can i give you a link?

[05:27:02.0000] <Philip`>
Sure

[05:28:00.0000] <mut>
cheers

[05:28:01.0000] <mut>
it just keeps stacking images

[05:28:02.0000] <mut>
if that makes sense

[05:31:00.0000] <Philip`>
It's probably best to not call ctx.save in init(), and instead make the drawStairs function do "ctx.clearRect(...); ctx.save(); ...do all the drawing...; ctx.restore()"

[05:31:01.0000] <mut>
ok

[05:33:00.0000] <Philip`>
That way the canvas would always be in a 'normal' state, except temporarily while it's inside the drawing function (where it'll mess around with the state but it'll restore everything at the end)

[05:34:00.0000] <mut>
ok thats almost there  :) cheers

[05:35:00.0000] <mut>
heh wierd, its just messing up on one small thing now

[05:35:01.0000] <mut>
thanks for that philip, *back to work

[05:35:02.0000] <mut>
:)

[05:47:00.0000] <Creap>
What would you suggest to use to accomplish something like <nobr>, for instance to keep the unit on the same row when writing "10 GB"

[05:47:01.0000] <Creap>
any particular element or just a span with some custom class

[05:48:00.0000] <Creap>
or 10&nbsp;GB?

[05:48:01.0000] <Philip`>
"&pound;10" :-)

[05:49:00.0000] <Philip`>
Oops

[05:49:01.0000] <Philip`>
I thought you said GBP

[05:49:02.0000] <Philip`>
/me isn't good at reading today

[05:50:00.0000] <Creap>
hehe

[05:50:01.0000] <Philip`>
nbsp seems to me the semantically appropriate thing when you really want a space that doesn't break

[05:51:00.0000] <Philip`>
though maybe it renders narrower than you'd want

[05:51:01.0000] <Creap>
true, I didn't even think of that, &nbsp; makes the eyes bleed because of the misuse

[05:51:02.0000] <Creap>
or well, I thought of that, but not until I've already asked ;)

[05:52:00.0000] <Philip`>
I wouldn't really recommend using &nbsp; for layout purposes, but it sounds like a good fit here :-)

[05:54:00.0000] <jgraham>
You can always insert a literal U+00A0 character. That would be less ugly, but possibly also harder to maintain

[06:00:00.0000] <asmodai>
zcorpan: Mmm, looks like my.opera.com has some issues now.

[06:01:00.0000] <zcorpan>
yeah, i noticed. maybe my post was too long

[06:01:01.0000] <asmodai>
zcorpan: Oh noes! You broke it

[06:02:00.0000] <asmodai>
Ah, there it is again

[06:02:01.0000] <karlcow>
http://www.theregister.co.uk/2010/03/02/virgin_america_html_flash/print.html

[06:02:02.0000] <mut>
hey philip any chance of 1 more quick piece of advice?

[06:03:00.0000] <mut>
same again, but it all works

[06:03:01.0000] <mut>
except the bottom part of the drawing dosent reset

[06:03:02.0000] <mut>
its bizzare

[06:03:03.0000] <asmodai>
zcorpan: Guess you guys moved away from Python/Genshi for it nowadays?

[06:04:00.0000] <zcorpan>
asmodai: i have no idea what's used for my.opera

[06:04:01.0000] <asmodai>
Shoot

[06:04:02.0000] <asmodai>
I know it was Python with Genshi at one point, since I had some contact with the dev guys on your side

[06:05:00.0000] <zcorpan>
ah

[06:05:01.0000] <gsnedders>
asmodai: my.opera uses Perl

[06:06:00.0000] <zcorpan>
/me gotta go catch a train

[06:12:00.0000] <asmodai>
ah, sorry

[06:12:01.0000] <asmodai>
was widgets.opera.com

[06:12:02.0000] <asmodai>
gsnedders: http://my.opera.com/WebApplications/blog/2008/02/07/widgets-opera-com-2-0

[06:12:03.0000] <gsnedders>
asmodai: Yeah, that uses Python

[06:13:00.0000] <asmodai>
Still does?

[06:13:01.0000] <gsnedders>
Yeah

[06:16:00.0000] <Philip`>
mut: Hmm, that might be because you're calling clearRect after having called scale, so the cleared area will be scaled too, so it might not be covering the whole canvas

[06:19:00.0000] <mut>
hmmm

[06:19:01.0000] <mut>
ok cheers

[06:19:02.0000] <mut>
ill take a look

[06:21:00.0000] <mut>
philip` I dont think its that, because, it seems to leave behind the last drawing of the bottom tread

[06:21:01.0000] <mut>
if that makes sense

[06:21:02.0000] <mut>
so whereever it was last, gets left behing

[06:21:03.0000] <mut>
sometimes the new one is further down than that

[06:22:00.0000] <mut>
(although im sure that the scale thing isnt helping :)

[06:28:00.0000] <Philip`>
mut: Not sure if this is the problem, but I think you need to call beginPath more

[06:29:00.0000] <Philip`>
Typically you should always do "beginPath(); moveTo and lineTo etc to set up the path; fill() or stroke();"

[06:29:01.0000] <mut>
oh ok

[06:29:02.0000] <Philip`>
because fill/stroke don't reset the current path, so it will continue to exist for your next path-drawing operation

[06:29:03.0000] <Philip`>
and so you'll keep building up a longer and longer path, and keep refilling/restroking the early parts lots of times

[06:30:00.0000] <Philip`>
e.g. in the drawStairRun loop

[06:30:01.0000] <Philip`>
and it looks like that'll also be drawing the bits of path left over from the end of the previous time you redrew the screen

[06:30:02.0000] <mut>
Ok, ill go and have a look (only found canvas this morning! so Im on the bottom of the learning curve) ill go and sort out the beginpath stuff out first and see if that helps

[06:30:03.0000] <Philip`>
so beginPath should help :-)

[06:30:04.0000] <mut>
ok  ta

[06:32:00.0000] <mut>
Im liking this canvas stuff, all the functions like .arc etc are very similar to CNC coding, so its coming quite quickly to me

[06:33:00.0000] <jgraham>
hsivonen: http://blogs.msdn.com/ie/archive/2010/03/02/how-ie8-determines-document-mode.aspx

[06:34:00.0000] <paul_irish>
i was kinda curious why they were replicating henri's flowchart two years later...

[06:35:00.0000] <paul_irish>
like they forgot to hit the publish button on this one?

[06:36:00.0000] <Rik|work>
I think they're trying to prove that "a picture is worth a thousand words" is a wrong adage

[06:36:01.0000] <hsivonen>
jgraham: interesting. I need to verify that my flowchart matches

[06:37:00.0000] <mut>
Philip` sorted, thanks

[06:40:00.0000] <jgraham>
hsivonen: I think they have a few things you don't and you have a few things that they don't

[06:41:00.0000] <Philip`>
mut: Good :-)

[06:48:00.0000] <hsivonen>
paul_irish: maybe they didn't forget to publish it earlier but it took this long to map the thing out :-)

[06:49:00.0000] <paul_irish>
hahaha

[06:52:00.0000] <Philip`>
Word and Excel really ought to detect when you're pressing printscreen and disable the spelling/grammar checker underlines

[06:53:00.0000] <lazni>
I think a lower level thingy catches the pressing

[06:57:00.0000] <mut>
i always hit print preview before i printscreen cus of that

[07:21:00.0000] <karlcow>
77% of French people do not understand the ballot screen for browsers http://www.silicon.fr/fr/news/2010/03/02/une_majorite_d_internautes_ne_connait_pas_le_ballot_screen_de_microsoft

[07:23:00.0000] <Philip`>
23% understand the concept of web browser choice? That's much higher than I'd expect

[07:26:00.0000] <mut>
:P

[07:26:01.0000] <jgraham>
Philip`: That's roughly the same percentaqge that already use something other than IE I expect

[07:29:00.0000] <Philip`>
jgraham: I thought most of those only used not-IE because a family member installed Firefox for them and set it as the default

[09:59:00.0000] <TabAtkins>
So, hey, does anyone know of any examples of video on the web that are *not* done through <video> or plugins?

[09:59:01.0000] <TabAtkins>
I don't know if this actually exists or not.

[10:00:00.0000] <TabAtkins>
And more specifically, if it *does* exist, does it act like <video> with respect to letter-boxing when the container isn't the right size?  Or does it try and fully scale itself like an image would?

[10:00:01.0000] <Philip`>
You mean native browser support for videos?

[10:00:02.0000] <TabAtkins>
I guess.  Not sure.  Hakon says there is some.  I doubt this.

[10:01:00.0000] <TabAtkins>
It's the point upon which our argument was hinging on today in the conf call.

[10:01:01.0000] <Philip`>
There's animated GIFs

[10:01:02.0000] <Philip`>
Maybe also http://msdn.microsoft.com/en-us/library/ms533606(VS.85).aspx

[10:02:00.0000] <TabAtkins>
They resize like images, right?

[10:02:01.0000] <TabAtkins>
/me has several animated gifs on hand, and can check.

[10:02:02.0000] <Philip`>
I don't know what else exists that isn't <embed>/<object>, though

[10:02:03.0000] <Philip`>
TabAtkins: Yes, they're just the same as any image

[10:02:04.0000] <Philip`>
(at least when you use them in <img>)

[10:03:00.0000] <TabAtkins>
Yeah.

[10:03:01.0000] <TabAtkins>
Ah, nobody cares about HTML+TIME.

[10:03:02.0000] <Philip`>
Seems strange to hinge an argument on the behaviour of clearly obscure technologies where you don't even know if they exist :-)

[10:04:00.0000] <TabAtkins>
Well, if they *do* exist, we have to add a new value that means "do whatever is appropriate for the media".  If they don't, then the existing ones are fine.

[10:05:00.0000] <Philip`>
Why do we have to do what they do?

[10:05:01.0000] <Philip`>
<video> will fail to support a load of things that people do already

[10:06:00.0000] <Philip`>
(When you say "plugins", do you just mean Flash, or do you also mean to exclude things like WMP/QuickTime/RealPlayer plugins that were used before Flash video came along?)

[10:07:00.0000] <TabAtkins>
Anything that would invoke the plugin architecture.

[10:07:01.0000] <TabAtkins>
Also, sorry, was referring to a CSS discussion, not HTML.

[10:08:00.0000] <TabAtkins>
Specifically, the image-fit property, which determines how replaced content scales in its content box.

[10:08:01.0000] <TabAtkins>
So, those other things you mentioned count as plugins.

[10:08:02.0000] <Philip`>
Oh

[10:13:00.0000] <TabAtkins>
Hm.  I need to eat a distinctly Texan lunch today, since it will be one of my last few before I move.  Wonder where I should go?

[10:13:01.0000] <Philip`>
I suggest going to Texas

[10:13:02.0000] <TabAtkins>
Step 1: Done.

[10:14:00.0000] <Philip`>
Sounds like you're well on the way to having a genuine Texan lunch, then

[10:15:00.0000] <TabAtkins>
I also need to finally go buy myself a cowboy hat tonight, since it is a conspicuous lack in my life.

[10:15:01.0000] <TabAtkins>
Gotta play up the "y'all" and "ma'am" when I move.

[12:08:00.0000] <AryehGregor>
MikeSmith, what exactly is the relationship between validator.nu and validator.w3.org?  Could my problem be fixed in validator.nu and it would also work on the W3C validator?  The latter is the one that most people use.

[12:13:00.0000] <MikeSmith>
AryehGregor: the current W3C validator is essentially a Perl script that wraps around James Clark's old C-based "sp" validation engine, which is an SGML tool that does DTD-based validation. if it either encounters a <!doctype html> doctype or is manually told to check a doc as HTM5, what that perl script does is, it uses the validator.nu REST API to validate the document, then processes the results of that and reports them

[12:14:00.0000] <AryehGregor>
Oh, so it's literally requesting results from validator.nu, not just running the same software?  Interesting.

[12:15:00.0000] <MikeSmith>
AryehGregor: no, actually, to be precise, it's usually a locally-installed headless instance of the validator.nu backend

[12:15:01.0000] <AryehGregor>
Ah.

[12:15:02.0000] <AryehGregor>
It seems like I'd want 1) validator.nu to say it's valid HTML5 if it is, even with a confusing doctype, and 2) the Perl script to try passing things to validator.nu if it has an obsolete but conforming doctype but doesn't validate according to that doctype.

[12:15:03.0000] <AryehGregor>
Does that sound right?

[12:16:00.0000] <MikeSmith>
AryehGregor: yeah

[12:16:01.0000] <AryehGregor>
Okay, thanks.

[12:17:00.0000] <MikeSmith>
but I would not recommend calling those doctypes "obsolete but conforming" in discussions on the qa-dev and www-validator lists

[12:18:00.0000] <MikeSmith>
you would likely just get people taking issue with that wording rather than listening to what you're proposing

[12:19:00.0000] <AryehGregor>
Er, okay.

[12:19:01.0000] <AryehGregor>
That's the term the HTML5 spec uses, right?

[12:19:02.0000] <MikeSmith>
right, but even the intent of that is the specific context of HTML4

[12:19:03.0000] <MikeSmith>
*HTML5

[12:20:00.0000] <MikeSmith>
there are lot of people who don't care about validating their documents as HTML5 at this point and who may never want to change or update their content to get it validate against HTML5, if it's already valid HTML4 or XHTML1

[12:20:01.0000] <MikeSmith>
which is fine

[12:21:00.0000] <AryehGregor>
Right.  I'd say only kick it to validator.nu if it fails validation under the given doctype.

[12:22:00.0000] <MikeSmith>
even in that case, I don't think you will get a warm reception if you propose that it does that automatically, or especially if it does it without reporting their errors against whatever doctype the document actually uses

[12:23:00.0000] <MikeSmith>
for one thing, it would mean validation of every invalid document would take longer, because it would get automatically checked twice

[12:25:00.0000] <AryehGregor>
Maybe we could use a secret code, like adding <!-- html5 --> right after the doctype to trigger HTML5 validation mode?

[12:25:01.0000] <AryehGregor>
Hacky, but it would be a lot simpler, I imagine.

[12:25:02.0000] <MikeSmith>
icky indeed

[12:26:00.0000] <MikeSmith>
I don't know what the target users would be who would take the time to do that

[12:26:01.0000] <AryehGregor>
Or could the HTML5 obsolete-but-conforming doctypes be modified so that they specify the same DTD/entities, but are not actually conforming HTML4 or whatever?

[12:26:02.0000] <AryehGregor>
Take the time to do what?

[12:26:03.0000] <AryehGregor>
I'd do it in MediaWiki, that's my concern here.  The current setup has MediaWiki not validating even if it's valid HTML5, if in well-formed XML mode.

[12:27:00.0000] <MikeSmith>
I meant, take the time to add a special-purpose comment like that

[12:27:01.0000] <MikeSmith>
but I think you'd get a warmer reception if the suggestion were for an opt-in mechanism that prompts users to choose whether they want to re-validate the doc at HTML5 or not

[12:27:02.0000] <AryehGregor>
I meant, I'll add the comment to MediaWiki if it's the only way to validate.

[12:27:03.0000] <MikeSmith>
I see

[12:27:04.0000] <AryehGregor>
I'm concerned about people sticking the page in the validator and getting a "This document is invalid!"

[12:28:00.0000] <AryehGregor>
Nobody's going to notice a revalidate button unless it's unreasonably prominent.

[12:28:01.0000] <AryehGregor>
(since most such documents will not be valid HTML5 either)

[12:28:02.0000] <AryehGregor>
I don't know much about doctypes -- is there some way to make the doctype invalid HTML4/XHTML1 while still keeping the same behavior in browsers and other XML processors?

[12:29:00.0000] <AryehGregor>
Or is every part that's specified essential?

[12:29:01.0000] <MikeSmith>
I don't think there's any way to do that, no

[12:29:02.0000] <AryehGregor>
/me thinks it all looks essential, although he doesn't understand all of it.

[12:30:00.0000] <MikeSmith>
only thing I can think of is that SGML does not require the system ID

[12:31:00.0000] <MikeSmith>
that is, the URI part  in the doctype

[12:31:01.0000] <MikeSmith>
the second quoted string

[12:31:02.0000] <MikeSmith>
it only needs to first

[12:31:03.0000] <AryehGregor>
That's necessary for XML validators to figure out the entities, though.

[12:31:04.0000] <MikeSmith>
leaving off the system ID makes it invalid XML

[12:31:05.0000] <AryehGregor>
Hmm.

[12:31:06.0000] <MikeSmith>
actually, non-well-formed XML, I gues

[12:31:07.0000] <AryehGregor>
Well, a comment would be hacky, but I had actually thought of it earlier anyway, because it will also serve to inform human readers that the document is actually HTML5.

[12:32:00.0000] <AryehGregor>
Lest they get confused.

[12:32:01.0000] <AryehGregor>
Do you think that would stand a chance of getting approved?

[12:32:02.0000] <MikeSmith>
I very much doubt it

[12:32:03.0000] <AryehGregor>
Bleh.

[12:32:04.0000] <MikeSmith>
I think most people consider specialized comments of that kind to be an anti-pattern

[12:33:00.0000] <MikeSmith>
that's the kind of thing that SGML/XML processing instructions are meant for

[12:33:01.0000] <MikeSmith>
but such processing instructions are not valid in HTML5, for good reasons

[12:33:02.0000] <AryehGregor>
Can you think of anything else that might possibly get approved?  Maybe I should just explain the problem and outline the pros and cons I see with each possible solution and hope someone finds something acceptable beyond "call it invalid"?

[12:34:00.0000] <MikeSmith>
AryehGregor: explaining the problem first would be good, yeah

[12:34:01.0000] <AryehGregor>
Well, I was planning on doing that anyway.  :)

[12:34:02.0000] <AryehGregor>
I'll give that a shot when I have the time.

[12:35:00.0000] <AryehGregor>
Thanks.

[12:35:01.0000] <MikeSmith>
OK. proposing specific solutions is fine, too, but the risk is always that people then get hung up on arguing about the merits or faults of the proposed solution rather than thinking about the problem

[12:36:00.0000] <AryehGregor>
Right.

[12:36:01.0000] <AryehGregor>
Got to go now, I'll post a message to one of those lists soon.

[12:51:00.0000] <MikeSmith>
<twitter!andymantell> @w3c Still don't get why there is no previous sibling combinator in CSS3, it'd be so so useful - any chance this could make it in one day?

[12:55:00.0000] <Dashiva>
That would lead to possible cycles

[12:57:00.0000] <TabAtkins>
MikeSmith: I can answer that.  Gimme a sec; busy right now, then I'll give you a details response.

[12:57:01.0000] <MikeSmith>
TabAtkins: thanks

[12:58:00.0000] <MikeSmith>
or feel free to reply directly to him on twitter if you want 

[12:58:01.0000] <MikeSmith>
Dashiva: performance issue?

[13:03:00.0000] <Dashiva>
I'm not sure if it's only performance or if it could lead to loops in the layout as well

[13:04:00.0000] <TabAtkins>
No, just within selectors you can't get cycles.

[13:04:01.0000] <TabAtkins>
It's performance reasons, though previous-sibling and parent combinators are okay enough that implementors are all right with doing them.

[13:04:02.0000] <Dashiva>
Oh, my bad then

[13:04:03.0000] <TabAtkins>
We just haven't started writing Selectors 4 yet.

[13:05:00.0000] <TabAtkins>
They're pretty much certain to be in there.

[13:06:00.0000] <TabAtkins>
Basically browsers have a somewhat unintuitive performance profile that makes all the reverse-versions of the current 4 combinators pretty inefficient, even though they can be done fairly easily in selectors engines like Sizzle.

[13:07:00.0000] <TabAtkins>
It's because nearly all selectors matching in browsers happens on an incomplete DOM as a page is being loaded, while js-based selectors engines operate on complete doms.

[13:09:00.0000] <TabAtkins>
And the existing combinators happen to work great with that, as they only need information that is guaranteed to already be in the DOM.

[13:09:01.0000] <TabAtkins>
If you are trying to match div > span, when a span appears you are guaranteed that, if it has a div parent, you'll have already seen it and know about it because of the way HTML's structure works.

[13:10:00.0000] <TabAtkins>
While "span < div" doesn't offer the same promise - when you see a div, you then have to wait until all its children have been loaded before you can be sure if it has a span child.

[13:13:00.0000] <TabAtkins>
MikeSmith: So there you go.  Try condensing that down to 140 characters.  ^_^

[13:14:00.0000] <TabAtkins>
I want :has-child() *so bad*, but I understand why it hasn't appeared so far.  It will appear in the future, though.

[13:14:01.0000] <TabAtkins>
/me mainly wants it so he can do "label:has-child(:checked)".

[13:18:00.0000] <MikeSmith>
TabAtkins: thanks

[13:23:00.0000] <MikeSmith>
TabAtkins: you mentioned that implementors are OK with doing previous-sibling and parent combinators

[13:30:00.0000] <MikeSmith>
Hixie: I need some help in getting the Microdata draft through pubrules

[13:31:00.0000] <MikeSmith>
TabAtkins: so you are saying that implementors are planning to add support for previous-sibling?

[14:38:00.0000] <Hixie>
MikeSmith: here

[14:40:00.0000] <MikeSmith>
Hixie: I am told we can't have URLs for images and other things in examples that don't resolve

[14:40:01.0000] <MikeSmith>
particularly to third-party sites

[14:40:02.0000] <MikeSmith>
e.g., the oreilly.com URLs in the examples

[14:41:00.0000] <MikeSmith>
so I think we should change those to example.com URLs

[14:43:00.0000] <Philip`>
example.com is a third-party site, and most URLs on it won't resolve

[14:43:01.0000] <Philip`>
so that seems an inconsistent policy

[14:43:02.0000] <MikeSmith>
yeah, but we know they are not meant to reolve

[14:45:00.0000] <MikeSmith>
I believe in part the logic of this is that we don't want somebody contacting O'Reilly to complain about a broken URL that we created

[14:46:00.0000] <MikeSmith>
Hixie: the bogus dev.w3.org house.jpeg and mailbox.jpeg URLs as well

[14:46:01.0000] <MikeSmith>
Philip`: example.com is not actually a third-party site

[14:46:02.0000] <MikeSmith>
it is non-site

[14:47:00.0000] <Philip`>
It looks like a site to me

[14:47:01.0000] <Philip`>
Not a very exciting one, admittedly

[14:47:02.0000] <Philip`>
It's got a web server and everything

[14:48:00.0000] <MikeSmith>
well, it's a site that tells you to go read an RFC to find out why there's actually no content at the site

[14:48:01.0000] <TabAtkins>
MikeSmith: Yeah, that's the indication so far.  Previous adjacent sibling and parent aren't *too* bad, and they're useful enough to justify the cost.  General previous sibling (reverse of ~) and ancestor, though, are much worse and can't justify themselves as well.

[14:49:00.0000] <MikeSmith>
TabAtkins: OK

[14:49:01.0000] <Hixie>
MikeSmith: looking...

[14:50:00.0000] <MikeSmith>
Philip`: and it's site that we can create arbitrary fully-qualified domain names for without somebody nagging me to fix them

[14:50:01.0000] <MikeSmith>
s/create/invent|make up/

[14:51:00.0000] <TabAtkins>
Hehe, I love people who don't know how mailing lists work, or how they got on them.

[14:51:01.0000] <TabAtkins>
Specifically, I love the line "I'M SO FATUP WITH THIS!!!".

[14:51:02.0000] <Hixie>
MikeSmith: the oreilly URLs were provided by an O'Reilly employee, so I presume that they are valid RDF URLs (which don't resolve by design)

[14:52:00.0000] <Hixie>
MikeSmith: i couldn't find any dev.w3.org house.jpeg URLs

[14:53:00.0000] <MikeSmith>
Hixie: the example in http://dev.w3.org/html5/md/#associating-names-with-items

[14:54:00.0000] <Hixie>
wtf

[14:54:01.0000] <Hixie>
that's not in the source

[14:54:02.0000] <Hixie>
what on earth

[14:55:00.0000] <Hixie>
c.f. http://www.whatwg.org/specs/web-apps/current-work/complete.html#associating-names-with-items

[14:56:00.0000] <MikeSmith>
weird

[14:56:01.0000] <Hixie>
very weird

[14:56:02.0000] <MikeSmith>
I hope I'm not to blame for this

[14:56:03.0000] <Hixie>
it's in the version i check in

[14:56:04.0000] <Hixie>
what on earth is going on here

[14:57:00.0000] <Hixie>
/me walks through this pipeline

[14:58:00.0000] <MikeSmith>
Hixie: about the O'Reilly URLs, if you can point me to an e-mail message or IRC log or whatever where those came from, I can pass it on

[14:59:00.0000] <Hixie>
http://gavin.carothers.name/2009/08/13/trying-to-understand-microdata-rdfa/

[15:00:00.0000] <MikeSmith>
thanks

[15:02:00.0000] <Hixie>
(this is his google profile http://www.google.com/profiles/gcarothers )

[15:04:00.0000] <Hixie>
apparently the URL gets screwed up by my filter-for-w3c script

[15:04:01.0000] <Hixie>
aha, i see the problem

[15:06:00.0000] <MikeSmith>
Hixie: it's still not clear whether those are real or just examples he's using for illustration purposes

[15:07:00.0000] <MikeSmith>
I would really rather we just replace them with example.com URLs instead of me needing to argue with the publication team to convince them we're not violating the policy

[15:08:00.0000] <Hixie>
k

[15:08:01.0000] <Dashiva>
<dd><span itemprop="http://purl.org/dc/terms/creator">Wil Wheaton</span></dd>

[15:08:02.0000] <Dashiva>
Is the span really necessary there?

[15:09:00.0000] <TabAtkins>
No.

[15:14:00.0000] <MikeSmith>
/me steps away for a bit

[15:48:00.0000] <Hixie>
MikeSmith: ok, checked in changes to fix those two problems


2010-03-04
[16:46:00.0000] <MikeSmith>
Hixie: thanks for the changes

[17:40:00.0000] <TabAtkins>
Ah, Whataburger, the most Texan of all fast-food joints.

[17:40:01.0000] <TabAtkins>
/me is filling up on stereotypical food.

[22:20:00.0000] <JonathanNeal>
Hey all

[22:24:00.0000] <Hixie>
hey

[22:46:00.0000] <JonathanNeal>
Hey Hixie, how's it goin' tonight?

[22:53:00.0000] <MikeSmith>
Hixie: if you can indulge me, there's another minor issue in the markup for the HTML5 spec that I'd hope you can fix upstream -- it's that there's no "type" attribute on the style element that contains the .domintro:before selector

[22:53:01.0000] <Hixie>
JonathanNeal: good, good, you?

[22:54:00.0000] <Hixie>
MikeSmith: updating...

[22:54:01.0000] <othermaciej>
MikeSmith: are you fixing the pubrules checker issues that the webmaster pointed out?

[22:54:02.0000] <JonathanNeal>
The battle of HTML5 seems all but over within the company, we're good to go.  I've been building out the next version of the theme that ships with the product.

[22:55:00.0000] <othermaciej>
MikeSmith: (sounds like you are, just wanted to confirm)

[22:55:01.0000] <othermaciej>
JonathanNeal: what is your company's product and/or service?

[22:56:00.0000] <JonathanNeal>
Liferay, our company site is all HTML5'd out now, but now I've been working to put it in the product. 

[22:57:00.0000] <JonathanNeal>
Along with moving off TABLE-based layouts, and so far it all looks good.  Today I built the new default theme that will ship with the next version of the software coming out later this month or sometime about then.

[22:57:01.0000] <MikeSmith>
othermaciej: yeah, plh fixed a bunch and I'm fixing more now

[22:58:00.0000] <MikeSmith>
othermaciej: scare quotes implied

[22:58:01.0000] <JonathanNeal>
I always forget who each of you guys work for, it seems to be a pretty well mixed bunch.

[22:59:00.0000] <MikeSmith>
othermaciej:  in some cases we're not really fixing anything, but instead just trying to get the pubrules checker to stfu

[23:00:00.0000] <Hixie>
MikeSmith: fixed

[23:01:00.0000] <othermaciej>
MikeSmith: that's unfortunate, glad things are getting fixed though

[23:01:01.0000] <MikeSmith>
Hixie: thanks again

[23:03:00.0000] <MikeSmith>
othermaciej: a number of the drafts are already in place at the 2010/WD-* URLs, so we just need to wait for the webmaster to add symlinks for those

[23:04:00.0000] <MikeSmith>
the rest I should have in place well before the html wg telcon tonight

[23:04:01.0000] <othermaciej>
MikeSmith: hawt

[23:06:00.0000] <JonathanNeal>
othermaciej, what do you do?

[23:07:00.0000] <othermaciej>
JonathanNeal: I'm on the WebKit team at Apple

[23:08:00.0000] <JonathanNeal>
Oh neato, I like you guys *except that whole gradient spec* :D

[23:08:01.0000] <othermaciej>
I blame TabAtkins for the gradient spec

[23:09:00.0000] <JonathanNeal>
Really?  TabAtkins did it?  Say it ain't so.  He helped move me to HTML5.

[23:09:01.0000] <JonathanNeal>
othermaciej, fine, can I blame you for your quicktime-ish HTML5 video player?

[23:10:00.0000] <othermaciej>
JonathanNeal: I think he's currently editing the gradient spec, though I think we were the first to have some form of gradients in CSS

[23:11:00.0000] <knowtheory>
othermaciej: i saw that this issue http://www.w3.org/html/wg/tracker/issues/10 was closed w/o prejudice

[23:11:01.0000] <knowtheory>
and i was just curious whether that means no consideration was taken regarding SMIL?

[23:11:02.0000] <knowtheory>
(i'm just trying to figure out what the status is atm, besides, you know, closed)

[23:11:03.0000] <othermaciej>
JonathanNeal: what aspect of our video player do you find too "quicktime-ish"?

[23:12:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-video-framing/ othermaciej

[23:12:01.0000] <othermaciej>
knowtheory: no one made a concrete proposal to make the <video> element more like SMIL

[23:13:00.0000] <othermaciej>
knowtheory: if someone were to come forward with a proposal, we could reopen it, though at this point I'd suggest any proposals along those lines start with a bug report

[23:13:01.0000] <knowtheory>
othermaciej: great, that's what i wanted to know.

[23:13:02.0000] <knowtheory>
i'm disappointed that Dick Bulterman never got back to you guys

[23:14:00.0000] <othermaciej>
JonathanNeal: that's not what the latest controls look like

[23:14:01.0000] <knowtheory>
i work w/ a startup who's doing SMIL based video delivery, and when i found this, i thought it was both interesting, and potentially useful

[23:14:02.0000] <knowtheory>
especially if there is interoperability between SMIL and (X)HTML5

[23:14:03.0000] <JonathanNeal>
I'll either update or delete the page when the next Safari is out then.

[23:14:04.0000] <othermaciej>
knowtheory: for what it's worth, I think a proposal to incompatibly change the core features of <video> and <audio> would have a huge uphill battle at this point

[23:14:05.0000] <JonathanNeal>
othermaciej, will there be keyboard controls, do you know?

[23:15:00.0000] <othermaciej>
JonathanNeal: in Safari 4.04 on SnowLeopard, the controls I get on the live video do not look anything like your screenshot

[23:15:01.0000] <JonathanNeal>
You don't need to answer that, I'll find out when I find out.

[23:15:02.0000] <knowtheory>
othermaciej: yeah i appreciate that (which is why i'm particularly disappointed that Bulterman never got back to you guys), i figure i should at least take a look

[23:15:03.0000] <JonathanNeal>
othermaciej, oh I guess it's a PC thing.

[23:16:00.0000] <knowtheory>
we're interested in adoption of SMIL, but if that can be accomplished by better coordiation w/ existing standards, and being able to do things like integrate SMIL timing semantics via XML modules, so much the better IMO

[23:16:01.0000] <othermaciej>
JonathanNeal: we might not have updated Safari on Windows to the new controls yet

[23:17:00.0000] <JonathanNeal>
You might be interested in seeing http://sandbox.thewikies.com/html5-layout/7cogs-website.html it's filled with all sorts of HTML5 / CSS3 -webkit goodness.

[23:17:01.0000] <JonathanNeal>
-moz goodness too, even some IE love.

[23:18:00.0000] <othermaciej>
JonathanNeal: the controls are not keyboardable by default if that is your concern

[23:18:01.0000] <othermaciej>
but the look and hover behavior are different

[23:18:02.0000] <knowtheory>
othermaciej: i apologize for being out of the loop on this, i've only really started paying attention to the HTML5 process in earnest recently, i wish i had tuned in sooner :\

[23:19:00.0000] <knowtheory>
(apologize for raising this issue at this late stage that is)

[23:20:00.0000] <othermaciej>
knowtheory: I am just giving you fair warning that you may not have much luck if you propose incompatible changes

[23:20:01.0000] <othermaciej>
unless you have a really really really compelling rationale

[23:20:02.0000] <othermaciej>
JonathanNeal: nice looking web page

[23:23:00.0000] <knowtheory>
othermaciej: Thanks, i appreciate that, and don't really expect anything, but again, squandering the opportunity seems just as foolish, if something amenable could be worked out.

[23:23:01.0000] <knowtheory>
othermaciej: btw, i see that last call for comments is out, when is that process going to draw to a close?

[23:24:00.0000] <othermaciej>
knowtheory: the WHATWG copy of the spec went into Last Call, but then converted to unversioned, so I am not sure what the plan is there

[23:24:01.0000] <othermaciej>
at the W3C, the spec is not yet in Last Call

[23:24:02.0000] <Hixie>
the stability is just marked section-by-section

[23:25:00.0000] <Hixie>
going to last call was really a shorthand for saying all the sections were in last call

[23:25:01.0000] <knowtheory>
Ah okay, thanks.

[23:25:02.0000] <Hixie>
(you can see the states in the margins of the whatwg spec)

[23:25:03.0000] <knowtheory>
I saw the Status there under http://www.w3.org/TR/html5/video.html#video

[23:26:00.0000] <knowtheory>
so i wasn't sure, exactly what the implications of Last Call were here

[23:26:01.0000] <hsivonen>
"Larry wrote 2 March 2010 member-only email with suggested path forward (for some reason, it's not in the tag⊙wo archives, hence not linkable from this agenda)

[23:26:02.0000] <hsivonen>
"

[23:26:03.0000] <Hixie>
knowtheory: oh you don't want to look at the version on the TR/ page, that's almost always out of date

[23:26:04.0000] <hsivonen>
does anyone know if the email in question is available elsewhere?

[23:26:05.0000] <knowtheory>
(again i apologize that i haven't had more time to poke around and find out what the W3C process is yet, i saw that the relevant ticket was closed 2 months ago, and wanted to pop in to find out if it was too late to do anything)

[23:27:00.0000] <hsivonen>
context: http://lists.w3.org/Archives/Public/www-tag/2010Mar/0011.html

[23:27:01.0000] <knowtheory>
Hixie: where should i be looking then?

[23:27:02.0000] <Hixie>
knowtheory: you want either http://whatwg.org/html or http://dev.w3.org/html5/spec/

[23:27:03.0000] <JonathanNeal>
othermaciej, thanks

[23:27:04.0000] <knowtheory>
ah fantastic thanks Hixie

[23:27:05.0000] <Hixie>
the former has a feedback form in the lower right which might be useful

[23:28:00.0000] <knowtheory>
Cool, much appreciated

[23:34:00.0000] <othermaciej>
hsivonen: interesting - haven't seen the email, but then I looked at the February tag archives and found much fascinating content

[23:44:00.0000] <hsivonen>
othermaciej: http://lists.w3.org/Archives/Member/tag/2010Feb/0065.html and http://lists.w3.org/Archives/Member/tag/2010Feb/0057.html in particular

[23:53:00.0000] <othermaciej>
hsivonen: I'm wary of linking specific emails lest there be complaints that discussion on secret lists isn't being kept secret enough

[23:55:00.0000] <hsivonen>
sorry. I thought in was common practice to point to the URLs.

[23:56:00.0000] <knowtheory>
Incidentally othermaciej & Hixie is the HTML5 group aware of XHTML+SMIL? http://www.w3.org/TR/XHTMLplusSMIL/

[23:57:00.0000] <othermaciej>
hsivonen: me too but apparently it gets some people's panties in a bunch

[23:57:01.0000] <othermaciej>
hsivonen: you shouldn't feel the need to stop doing so

[23:57:02.0000] <othermaciej>
I'm just explaining why I am not doing so at the moment

[23:57:03.0000] <hsivonen>
I see

[23:58:00.0000] <foolip>
knowtheory: SMIL has been raised from time to time, but no ones seems particularly interested and its not clear what problems it would solve or how to integrate it with HTML

[23:58:01.0000] <othermaciej>
hsivonen: the emails you linked are indeed particularly interesting

[23:58:02.0000] <knowtheory>
foolip: Funnily enough, those are all questions i'm seeking to answer ;)

[23:59:00.0000] <foolip>
knowtheory: there was some discussion in the a11y taskforce not too long ago, perhaps you can search the archives

[23:59:01.0000] <knowtheory>
a11y?  I will do some googling, anything you could recommend i look for specifically?

[00:00:00.0000] <knowtheory>
ah the accessibility group

[00:00:01.0000] <foolip>
knowtheory: for reference (i.e. if you want to hit me), I'm Philip Jägenstedt in those threads.

[00:00:02.0000] <knowtheory>
thanks :) much appreciated.

[00:00:03.0000] <MikeSmith>
knowtheory: http://lists.w3.org/Archives/Public/public-html-a11y/

[00:01:00.0000] <knowtheory>
thanks MikeSmith 

[00:01:01.0000] <othermaciej>
knowtheory: my impression is the same as Philip's; people are generally aware of SMIL, but there is not much interest in using it

[00:02:00.0000] <knowtheory>
Yep, i appreciate that.

[00:03:00.0000] <hsivonen>
SMIL is a classic case of a W3C WG doing its own thing without enough implementors of incumbent products participating

[00:03:01.0000] <knowtheory>
Yeah, i've spent the past 3 months researching wtf happened w/ the SMIL standard

[00:03:02.0000] <knowtheory>
what it can do, what it can't do, how it interacts w/ other standards

[00:04:00.0000] <danbri>
what can it do?

[00:04:01.0000] <danbri>
in practice i mean

[00:04:02.0000] <danbri>
are there .js implementations yet?

[00:04:03.0000] <knowtheory>
well, SMIL's general purpose is to provide general timing semantics in a declarative fashion

[00:04:04.0000] <knowtheory>
there are some components of SMIL that have been built by people here and there

[00:05:00.0000] <danbri>
the only bit i've historically been interested in was timed text, which wasn't even in the older standards, they deferred to whatever realaudio etc happened to consume

[00:05:01.0000] <danbri>
so i'm curious if http://www.w3.org/TR/2010/CR-ttaf1-dfxp-20100223/ will get any traction

[00:05:02.0000] <knowtheory>
yeah.  the SMIL3 standard has timed text as a module at this point

[00:05:03.0000] <danbri>
and happy it got renamed to something saner

[00:05:04.0000] <knowtheory>
i mean most people know that some of SMIL's timing semantics have been integrated into SVG

[00:05:05.0000] <danbri>
<- http://lists.w3.org/Archives/Public/public-tt/2010Feb/0004.html

[00:05:06.0000] <knowtheory>
in order to do very basic animation

[00:05:07.0000] <foolip>
DFXP/TTML is pretty weird too, at least the styling markup

[00:06:00.0000] <knowtheory>
yeah i'm aware of TTML though not the details of how weird it can get.

[00:06:01.0000] <othermaciej>
DFXP has the worst acronym I have ever seen for a technology

[00:07:00.0000] <knowtheory>
the thing w/ SMIL is that it can provide basic timing semantics, which is handy for a variety of tasks, and is particularly useful when coordinating multiple pieces of content

[00:07:01.0000] <knowtheory>
i can give you examples of what the company i work for has been doing, and is pushing to continue to do

[00:07:02.0000] <knowtheory>
if people are interested

[00:07:03.0000] <othermaciej>
my impression of SMIL is that it feels like something designed for the era of multimedia CD-ROMs

[00:07:04.0000] <knowtheory>
othermaciej: that's sort of true, and sort of not.  My biggest complaint about SMIL right now is that the layout tools feel really clunky

[00:07:05.0000] <othermaciej>
it doesn't seem to fit that well with what people want to do with animation or timed media today

[00:07:06.0000] <foolip>
The expansion of DFXP is not very helpful either: Distribution Format Exchange Profile

[00:08:00.0000] <knowtheory>
which is why i'm getting interested in the possibility of integrating SMIL timing semantics into something with a more mature layout model, like XHTML

[00:08:01.0000] <othermaciej>
I think what people want is things that integrate really well into HTML/CSS, not a new top-level technology

[00:08:02.0000] <knowtheory>
Well that's the thing, since SMIL's been modularized it's actually possible to pick and choose components to support

[00:08:03.0000] <knowtheory>
if it's possible to integrate the standards at a base level

[00:08:04.0000] <othermaciej>
I don't view that as a plus

[00:09:00.0000] <foolip>
but HTML has no namespaces, so we'd just be copying the element names

[00:09:01.0000] <othermaciej>
picking and choosing what components to support leads to fragmentation and lack of interoperability

[00:09:02.0000] <knowtheory>
Yes, i'm going to have to admit, that i am walking into this somewhat unprepared so i dn't know that i have all the answers or a firm grasp of where the points of pain would be for integration.

[00:10:00.0000] <knowtheory>
othermaciej: yeah, well w/ SMIL it's used on such a wide variety of platforms with such a wide variety of capabilities

[00:10:01.0000] <hsivonen>
the design of DFXP is really sad. unfortunately, the spec fills a current accessibility void, so there's a risk that the Web gets stuck with it

[00:10:02.0000] <knowtheory>
they basically had to modularize the standard

[00:10:03.0000] <knowtheory>
MMS functions through SMIL for instance

[00:10:04.0000] <othermaciej>
at least some uses of SMIL actually support such tiny subsets that it's almost silly to pretend they are SMIL

[00:10:05.0000] <othermaciej>
MMS is a good example

[00:10:06.0000] <hsivonen>
does anyone use MMS?

[00:11:00.0000] <knowtheory>
In this day?  Probably not

[00:11:01.0000] <othermaciej>
Apple initially considered putting some SMIL support into WebKit to do it but the subset actually supported and used on real phones is so trivial that it was easier to just handle the common special cases

[00:12:00.0000] <knowtheory>
I mean, at this point there's like 3 issues.  First is why is SMIL the way it is (given it's history), second is what does SMIL actually provide to users at large, and 3rd, is it worth supporting and/or does it have a future as a tech

[00:12:01.0000] <knowtheory>
othermaciej: sure, i think that there is something to be said for being able to declaritively define timed behavior

[00:12:02.0000] <knowtheory>
the other alternative being JS

[00:13:00.0000] <knowtheory>
Depending on what your environment is, and whom your accepting data from, being able to rely on just recieving XML is a good thing

[00:13:01.0000] <othermaciej>
there's also css animations and css transitions

[00:13:02.0000] <othermaciej>
although those do not have SMIL's more robust concept of timebases

[00:14:00.0000] <knowtheory>
yeah.  A lot of what we're trying to do specifically involves video, and being able to manipulate it in a variety of ways

[00:14:01.0000] <knowtheory>
which is why we're interested in what's happening w/ HTML5

[00:15:00.0000] <knowtheory>
and SMIL's semantics for doing stuff like that are really good, if you want to concatenate video, or play video along side each other in a single timed environment

[00:15:01.0000] <knowtheory>
it's pretty solid

[00:16:00.0000] <knowtheory>
again i need to look at the issue more regarding HTML5s behavior, but i'm not sure how easy something like that would be to do w/ the in browser players that exist now

[00:16:01.0000] <knowtheory>
there are a bunch of toy usecases that exist already (i've been thinking about this because i intend to talk to people about it at SxSWi)

[00:16:02.0000] <knowtheory>
for instance, a youtube doubler would be easy to do in SMIL

[00:17:00.0000] <knowtheory>
and offer better control over timing

[00:17:01.0000] <knowtheory>
than the existing youtube doubler, which relies on remoting in and out of the flash player w/ JS (although that'll probably change w/ the HTML5 player and direct JS access)

[00:18:00.0000] <danbri>
'doubler'?

[00:18:01.0000] <othermaciej>
knowtheory: concatenation - you'd have to use script

[00:18:02.0000] <foolip>
knowtheory: it seems the difficult part here is actually getting media to play in sync or concatented without gaps, not the syntax for expressing it

[00:18:03.0000] <othermaciej>
knowtheory: playing side-by-side - there is no good way to tell the media engine to keep the timebases in sync

[00:19:00.0000] <knowtheory>
othermaciej: SMIL sets up semantics for dealing with slippage incidentally

[00:19:01.0000] <hsivonen>
othermaciej: s/the media engine/QuickTime/ ? or more generally?

[00:19:02.0000] <danbri>
ah http://www.youtubedoubler.com/ :)

[00:19:03.0000] <foolip>
othermaciej: what is "the media engine"? it seems to be possible in e.g. GStreamer.

[00:19:04.0000] <knowtheory>
sorry danbri was looking for a particular one to link you to

[00:19:05.0000] <knowtheory>
but you beat me to the site

[00:20:00.0000] <knowtheory>
here danbri: http://www.youtubedoubler.com/?video1=http://www.youtube.com/v/HsF_2CNV9v4&start1=3&video2=http://www.youtube.com/v/06CvUjLgK5g&start2=0&h=1

[00:20:01.0000] <MikeSmith>
foolip: btw, thanks for the replies about the surveys. I have to say I'm a bit confused about whether the issues you brought up were discussed previously among the media subgroup members

[00:21:00.0000] <othermaciej>
hsivonen, foolip: I mean there is no good interface at the HTML level 

[00:21:01.0000] <knowtheory>
othermaciej: here's an example i found out today that a friend of mine hacked up of a presentation using HTML5's audio: http://labs.toolness.com/ff-herdict-preso/

[00:21:02.0000] <othermaciej>
I know media playback engines *can* keep playback in sync if told do so, there is just no way to tell them

[00:21:03.0000] <knowtheory>
i told him i'd put that example together a SMIL version of that as well

[00:21:04.0000] <foolip>
othermaciej: indeed, but that isn't the hard part IMO, the hard part is in the media framework

[00:22:00.0000] <othermaciej>
knowtheory: syncing with a slideshow type thing is doable - though we have also discussed better ways to do it than the best currently possible

[00:23:00.0000] <knowtheory>
oh yeah?

[00:23:01.0000] <knowtheory>
that'd be interesting to hear about

[00:23:02.0000] <foolip>
MikeSmith: I think it's mostly been me, Silvia and Eric discussing the technical details, so there are plenty of small issues. But I don't think it's meaningful to make it perfect since things will likely be remade after it hits the HTML WG anyway.

[00:24:00.0000] <othermaciej>
/me would like to see it hit the HTML WG ASAP

[00:24:01.0000] <foolip>
/me too

[00:25:00.0000] <annevk>
subtitle format debates scare me

[00:25:01.0000] <othermaciej>
knowtheory: there used to be a concept of "cue ranges"

[00:26:00.0000] <foolip>
annevk: which parts do you find scary?

[00:26:01.0000] <othermaciej>
annevk: I would lean towards HTML5 being agnostic about supported formats

[00:27:00.0000] <hsivonen>
othermaciej: that doesn't really address the problem that we need a format the all the relevant vendors agree to implement

[00:28:00.0000] <hsivonen>
othermaciej: especially if the format is external to the video file

[00:28:01.0000] <hsivonen>
since in that case there's no value in maintaining the MPEG vs. Xiph duality

[00:28:02.0000] <othermaciej>
hsivonen: I expect all vendors will support SRT

[00:28:03.0000] <foolip>
I think no browser vendor has found SRT objectionable, yet

[00:28:04.0000] <foolip>
it's just that it lacks a spec

[00:29:00.0000] <hsivonen>
othermaciej: original SRT or RSSesquely extended?

[00:29:01.0000] <othermaciej>
hsivonen: I expect over time we will also agree on either DXFP or perhaps a subset, but I am less sure of that

[00:29:02.0000] <othermaciej>
I think trying to encode either or both of those thoughts in the HTML5 spec is likely to spawn flamewars

[00:29:03.0000] <hsivonen>
RSS <title> all over again scares me

[00:29:04.0000] <foolip>
say no SRT extensions!

[00:29:05.0000] <hsivonen>
/me agrees with foolip

[00:30:00.0000] <othermaciej>
I do not really care, but if it lacks a spec, I suppose someone should make one

[00:30:01.0000] <annevk>
if we can cover all relevant use cases with some SRT extensions that seems favorable over DXFP

[00:30:02.0000] <foolip>
and I think we shouldn't support the pseudo-HTML typically found in SRT either

[00:30:03.0000] <annevk>
we'll need some kind of SRT spec to nail all those things down

[00:30:04.0000] <knowtheory>
SRT?

[00:31:00.0000] <hsivonen>
annevk: I'd be OK with having a position up vs. position down marker extension in SRT, but the pseudo-HTML is serious badness

[00:31:01.0000] <foolip>
I expect writing the parsing section for SRT would be fun given the weird stuff that existing in the wild

[00:31:02.0000] <foolip>
hsivonen: position could be done with CSS couldn't it?

[00:32:00.0000] <foolip>
hsivonen: I mean completely external to the SRT file

[00:32:01.0000] <annevk>
foolip, that'd suck

[00:32:02.0000] <hsivonen>
(I can believe that position up vs. down could be dressed into being a legal requirement, but I have really hard time believing italics is a legal requirement unless someone shows me the relevant chapter and verse in a statute)

[00:32:03.0000] <annevk>
foolip, you want the ability to move the subtitles around a little to not cover onscreen text

[00:32:04.0000] <hsivonen>
foolip: I meant positioning a particular string in a particular time window near the top of the video frame when the video frame itself has relevant text at the bottom

[00:33:00.0000] <MikeSmith>
knowtheory: http://en.wikipedia.org/wiki/SubRip

[00:33:01.0000] <foolip>
I don't really like the idea of pushing SRT towards the "complex" side of the scale, I'd much rather have one format on each extreme

[00:34:00.0000] <foolip>
i.e. SRT + something

[00:35:00.0000] <knowtheory>
thanks MikeSmith :)

[00:35:01.0000] <foolip>
these legal requirements that are cited from time to time, who do they apply to? TV stations? Set-top box makers? Browser vendors?

[00:36:00.0000] <knowtheory>
So there's an interest in subtitles/captioning on video?

[00:36:01.0000] <knowtheory>
foolip: government bodies definitely

[00:36:02.0000] <foolip>
knowtheory: yes, very much so

[00:36:03.0000] <knowtheory>
i believe that companies also have an obligation

[00:36:04.0000] <knowtheory>
but i'm not sure what it is.

[00:36:05.0000] <knowtheory>
(off the top of my head anyway)

[00:36:06.0000] <danski>
the laws change from country to country but by and large they do apply to private organisations as well

[00:37:00.0000] <foolip>
So it could be illegal to not support e.g. positioning subtitles in a software product? That doesn't sound very reasonable.

[00:38:00.0000] <danski>
foolip: i'm trying to catch up with the discussion in progress here :) there are accessibility requirements that vary from country to country, is what i'm saying

[00:40:00.0000] <knowtheory>
Okay so basically, I should go justify SMIL's existence ;) and then see whether there's any interest (or even a way) to propose some kind of interoperability w/ HTML5

[00:42:00.0000] <hsivonen>
knowtheory: even if you manage to justify SMIL's existence, it's not a given that it's the most appropriate syntax for text/html.

[00:43:00.0000] <knowtheory>
hsivonen: in what contexts?

[00:45:00.0000] <hsivonen>
knowtheory: it's not a given that mixing SMIL syntax into text/html is a good idea even if SMIL turned out to have some desirable features

[00:45:01.0000] <knowtheory>
Actually, what is the deal with XHTML5?  someone made mention of a flath namespace?

[00:46:00.0000] <hsivonen>
XHTML5 is the HTML5 language parsed through an XML parser

[00:46:01.0000] <knowtheory>
Yeah, i get that

[00:46:02.0000] <knowtheory>
supports nameespaces and all that fun stuff then right?

[00:46:03.0000] <knowtheory>
so XHTML5 should support imbedding SVG

[00:46:04.0000] <knowtheory>
or a SMIL document as a result?

[00:46:05.0000] <hsivonen>
syntactically, yes

[00:46:06.0000] <knowtheory>
(even if browsers wouldnt' know what to do with the content)

[00:46:07.0000] <knowtheory>
yes, fine

[00:47:00.0000] <foolip>
knowtheory: see the very end of http://hsivonen.iki.fi/xhtml2-html5-q-and-a/

[00:47:01.0000] <annevk>
why do we care about SMIL?

[00:47:02.0000] <knowtheory>
Because HTML5 is converging on the media space, which is what SMIL was made for

[00:47:03.0000] <knowtheory>
and because SMIL does things like provide timing semantics which don't really exist in other standards

[00:48:00.0000] <danski>
timing semantics. in this specific example, the synching of captions to video, but there are many other applications.

[00:48:01.0000] <knowtheory>
and because SMIL is already integrated into other standards, like SVG

[00:48:02.0000] <knowtheory>
I'm not saying that HTML5 has to support it, but i'd like to try to make a case

[00:48:03.0000] <knowtheory>
that it might be useful/cool if it did

[00:49:00.0000] <knowtheory>
And i'm not suggesting that HTML5 should contain SMIL :P

[00:49:01.0000] <knowtheory>
I was just curious whether there was a possibility for interoperability given this issue: http://www.w3.org/html/wg/tracker/issues/10

[00:49:02.0000] <zcorpan__>
SVG uses a forked subset of SMIL, not SMIL proper

[00:50:00.0000] <knowtheory>
Yes, there's a common functional base though

[00:59:00.0000] <annevk>
hmm, maybe somewhat

[01:00:00.0000] <knowtheory>
foolip: thanks for the link

[02:03:00.0000] <gsnedders>
jgraham: http://www.petapixel.com/2010/03/03/canon-l-lens-look-alike-coffee-mug/

[02:25:00.0000] <mut>
hey, is there any way to measure the extremities of a drawing on canvas

[02:25:01.0000] <mut>
like so i can scale it correctly?

[02:36:00.0000] <Philip`>
mut: No

[02:36:01.0000] <Philip`>
(other than keeping track of it manually in your scripts)

[02:42:00.0000] <othermaciej>
was XMLHttpRequest once specified in HTML5?

[02:43:00.0000] <zcorpan__>
othermaciej: http://www.whatwg.org/specs/web-apps/2005-09-01/#scripted-http

[02:43:01.0000] <mut>
Philip` ok cheers... :(

[02:43:02.0000] <othermaciej>
whoah, I didn't know there were historical copies of the spec on whatwg.org

[02:44:00.0000] <othermaciej>
is there any way to see what previous versions exist?

[02:45:00.0000] <zcorpan__>
try url-hacking

[02:45:01.0000] <zcorpan__>
the 2006 version also has xhr

[02:45:02.0000] <othermaciej>
were Workers ever in HTML5 or were they always separate (can't remember)

[02:46:00.0000] <othermaciej>
here is my list of specs spun off from HTML5 so far:

[02:46:01.0000] <annevk5>
othermaciej, http://www.whatwg.org/specs/ has copies

[02:46:02.0000] <annevk5>
othermaciej, Web Workers was always separate, though is part of Web Applications 1.0 now

[02:46:03.0000] <annevk5>
s/copies/pointers/

[02:46:04.0000] <othermaciej>
- 2dcontext, microdata, server-sent events, web messaging, web sockets api, web sockets protocol, web sql data base, web storage, xmlhttprequest

[02:46:05.0000] <othermaciej>
now that I typed that I realize I am missing the sniffing spec

[02:47:00.0000] <zcorpan__>
workers was initially a separate source doc but was later merged with the main source (but has never been part of the generated html5 spec)

[02:49:00.0000] <othermaciej>
can anyone else think of specs that got split out besides what I listed?

[02:50:00.0000] <othermaciej>
I guess there is also Web Addresses, though that is in a messed up state

[02:50:01.0000] <othermaciej>
also are any of what I listed wrong?

[02:51:00.0000] <zcorpan__>
window was split but merged back

[02:52:00.0000] <othermaciej>
I don't think it was ever actually removed from the spec, though I did try to write a separate spec for it at one point

[02:52:01.0000] <Hixie>
selectors api was originally an issue marker in the html5 spec, though there was never any normative prose written for it in the html5 spec

[02:52:02.0000] <Hixie>
the alternative stylesheet stuff in cssom

[02:52:03.0000] <Hixie>
the url stuff

[02:52:04.0000] <Hixie>
the origin stuff

[02:55:00.0000] <Hixie>
<device>

[02:59:00.0000] <annevk5>
the internet

[02:59:01.0000] <annevk5>
oh wait

[03:07:00.0000] <MikeSmith>
anybody know any Hadoop committers, or anybody involved with Hadoop that might be interested in doing a presentation about it?

[03:08:00.0000] <MikeSmith>
annevk5: btw, thank you for having the html5-diff doc pubrules-clean (unlike me, who got a serious beat-down yesterday for not having the h:tml doc in the proper state)

[03:12:00.0000] <annevk5>
Hixie, in http://www.whatwg.org/specs/web-apps/current-work/complete.html#protocol-overview the ABNF does not look nice

[03:14:00.0000] <annevk5>
hmm, maybe I don't have the correct fonts or something

[03:16:00.0000] <Hixie>
"doesn't look nice"?

[03:17:00.0000] <Hixie>
what doesn't look nice about it?

[03:20:00.0000] <annevk5>
for some reason no monospace font is selected on my machine

[03:20:01.0000] <MikeSmith>
the comment for the name-char production wraps in an unsightly way

[03:21:00.0000] <MikeSmith>
that comment would be less prone to unsightliness if it were to be split into two lines

[03:21:01.0000] <MikeSmith>
the any-char production risks potential unsightliness as well

[03:22:00.0000] <MikeSmith>
*the comment for the any-char production

[03:24:00.0000] <annevk5>
encoding as UTF-8 seems wrong btw for resource name

[03:25:00.0000] <annevk5>
it would have to be percent-escaped i'd think

[03:25:01.0000] <annevk5>
also, it seems the connection steps stop mentioning UTF-8 encoded from step 7 on

[03:26:00.0000] <annevk5>
oh, nm

[03:27:00.0000] <Hixie>
MikeSmith: i recommend getting a bigger monitor

[03:27:01.0000] <Hixie>
that way it won't wrap ata ll

[03:28:00.0000] <MikeSmith>
/me sets aside his iPhone and goes to check on his main work machine, a TTY

[03:28:01.0000] <MikeSmith>
no luck there

[03:28:02.0000] <Hixie>
i do not buy that you were looking at that page on your iphone

[03:28:03.0000] <Hixie>
it's a 5MB HTML page of DOOM

[03:28:04.0000] <MikeSmith>
<snort>

[03:29:00.0000] <Hixie>
your iphone would probably catch fire if it didn't hang first

[03:29:01.0000] <MikeSmith>
I don't even have an iPhone, actually

[03:29:02.0000] <Hixie>
aha!

[03:29:03.0000] <MikeSmith>
I would welcome the donation of one from a generous benefactor

[03:29:04.0000] <gsnedders>
Liar!

[03:29:05.0000] <MikeSmith>
/me puts out his pants fire

[03:30:00.0000] <MikeSmith>
but it does honestly look like crap on my TTY

[03:30:01.0000] <MikeSmith>
and my Apple IIe

[03:30:02.0000] <MikeSmith>
/me goes back to playing Wizardry

[03:47:00.0000] <mut>
:/ just realised that the save() and restore()  functions on canvas are stacked

[03:47:01.0000] <mut>
that could have saved me some time about 3 or 4 hours ago

[03:47:02.0000] <Hixie>
how did you think they worked before?

[03:47:03.0000] <Hixie>
one-off?

[03:47:04.0000] <mut>
yea

[03:48:00.0000] <mut>
dont laugh :P

[03:48:01.0000] <Hixie>
anything in particular make you think that? if the spec is misleading i can try to fix it

[03:48:02.0000] <mut>
nah im noob

[03:48:03.0000] <Hixie>
everyone's a noob when they first read the spec, anything i can do to help noobs is helpful to many :-)

[03:48:04.0000] <mut>
seems realy obvious now tbh

[03:48:05.0000] <Hixie>
k

[03:48:06.0000] <mut>
just didn't realise

[03:49:00.0000] <gsnedders>
Hixie: Make the spec shorter

[03:49:01.0000] <mut>
just spotted it in a tutorial and facepalmed

[03:49:02.0000] <Hixie>
gsnedders: press command+minus a few times

[03:50:00.0000] <mut>
i love how canvas is similar to cnc gcode :) saved my learning curve

[03:50:01.0000] <gsnedders>
/me doesn't like how he knows nothing about 2D graphics APIs

[03:56:00.0000] <jgraham>
Hixie: That's not really shorter by any useful metric

[03:57:00.0000] <Hixie>
jgraham: he didn't specify

[03:57:01.0000] <jgraham>
I reccommend doing s/[aeiou]//

[03:58:00.0000] <jgraham>
It is srprsng hw sy it is t rd stff wth n vwls

[03:58:01.0000] <gsnedders>
You included "i" there.

[03:58:02.0000] <jgraham>
Er, yes

[03:58:03.0000] <jgraham>
I failk

[03:58:04.0000] <gsnedders>
Indeed. Like normal.

[03:59:00.0000] <jgraham>
I isn't a real vowel anyway

[03:59:01.0000] <gsnedders>
No, you're jgraham.

[03:59:02.0000] <gsnedders>
Not much like a consunent either

[03:59:03.0000] <gsnedders>
And I still can't spell.

[03:59:04.0000] <jgraham>
Was that supposed to say constanant?

[03:59:05.0000] <gsnedders>
Yes

[04:00:00.0000] <jgraham>
Well at least the fail is spread around :)

[04:01:00.0000] <gsnedders>
jgraham made me take off my headphones :(

[04:01:01.0000] <gsnedders>
/me doesn't like that man in the corner

[04:02:00.0000] <jgraham>
Oh you were probably only listening to Christian noisecore or some such anyway

[04:02:01.0000] <jgraham>
I'm sure your ears thanked me

[04:02:02.0000] <gsnedders>
Within Temptation, actually

[04:02:03.0000] <gsnedders>
Though I guess symphoic metal is something you'll hate anyway

[04:02:04.0000] <gsnedders>
I mean, includes "metal" in the name of the genre

[04:03:00.0000] <jgraham>
Much like Midge Ure, that means nothing to me

[04:03:01.0000] <workmad3>
gsnedders: do you mean 'symphonic metal' by any chance?

[04:03:02.0000] <mut>
heh canvas is ace, couldnt think of many applications to use it at first, but applications for it keep popping into my head

[04:03:03.0000] <gsnedders>
workmad3: Have I not already proven I cannot spell?

[04:03:04.0000] <workmad3>
/me prefers operatic metal e.g. 'nightwish' personally

[04:03:05.0000] <workmad3>
gsnedders: very true :)

[04:04:00.0000] <gsnedders>
workmad3: jgraham prefers the Eels and Radiohead.

[04:04:01.0000] <gsnedders>
But I guess he could tell you that.

[04:04:02.0000] <jgraham>
Not exclusively

[04:05:00.0000] <gsnedders>
Well, I don't like sym[autocomplete this yourself]

[04:06:00.0000] <gsnedders>
metal

[04:06:01.0000] <gsnedders>
exclusively

[04:06:02.0000] <gsnedders>
I pressed return a bit too soon then

[04:06:03.0000] <gsnedders>
I mean, even the intersection of our music tastes manages to be a non-empty set.

[04:06:04.0000] <Hixie>
the new stuff in the websocket protocol added 40 lines of code to my simple server

[04:07:00.0000] <workmad3>
I don't like operatic metal exclusively... I'll listen to most rock or metal and enjoy it to some extent :)

[04:07:01.0000] <Philip`>
gsnedders: Symbiotic metal?

[04:07:02.0000] <workmad3>
I don't like RnB, hiphop, rap, or pop though

[04:08:00.0000] <Philip`>
/me prefers plastic to metal

[04:08:01.0000] <workmad3>
easier to stamp out the same thing time after time? :P

[04:09:00.0000] <jgraham>
/me notes that gsnedders listed two bands and compared then to a genre

[04:09:01.0000] <gsnedders>
jgraham: Okay, okay… in 48 days I'll start being sensible.

[04:09:02.0000] <gsnedders>
Whatever "ikag" means.

[04:11:00.0000] <jgraham>
Ijak

[04:11:01.0000] <jgraham>
Or rather !Ijak

[04:11:02.0000] <gsnedders>
Whatever that word we were discussing earlier on the whiteboard means.

[04:13:00.0000] <Hixie>
woo, my server returns the same handshake as the spec's example

[04:13:01.0000] <Hixie>
sweet

[04:13:02.0000] <gsnedders>
Hixie: Congrats, you managed to write Perl without killing yourself.

[04:17:00.0000] <Hixie>
ok time for bed

[04:18:00.0000] <MikeSmith>
workmad3: how about folk metal?

[04:18:01.0000] <MikeSmith>
I'm a hardcore fan of that folkmetal stuff

[04:19:00.0000] <MikeSmith>
this kind of thing:

[04:19:01.0000] <MikeSmith>
http://upload.wikimedia.org/wikipedia/commons/0/08/Skyforger_16127_Cernunnos_Fest_11.jpg

[04:21:00.0000] <daedb>
MikeSmith: You like Finntroll?

[04:21:01.0000] <jgraham>
That's a real genre? What does it consist of? Metal riffs played on the banjo?

[04:21:02.0000] <MikeSmith>
/me scrambles to look up Fintroll in Wikipedia

[04:22:00.0000] <jgraham>
Hmm the photo suggests normal metal + bagpipes for more aural pain

[04:22:01.0000] <MikeSmith>
jgraham: don't mock the gods of folktalic metal

[04:22:02.0000] <gsnedders>
Those bagpipes don't even have enough pipes to be very loud

[04:22:03.0000] <daedb>
Finntroll is damn near the ultimate folk metal band... combine metal with humppa and have finns singing in swedish, it's awesome :)

[04:23:00.0000] <MikeSmith>
daedb: I am mostly a fan of the physique of the Fintrolls

[04:23:01.0000] <MikeSmith>
http://upload.wikimedia.org/wikipedia/commons/9/96/Finntroll%2C_Metaltown_2008.JPG

[04:23:02.0000] <MikeSmith>
/me wonders what those guys eat, of if they do in fact eat

[04:23:03.0000] <daedb>
MikeSmith: I was at the concert where that photo was taken :)

[04:23:04.0000] <MikeSmith>
daedb: I am pretty sure those dudes are tweakers

[04:24:00.0000] <MikeSmith>
daedb: I don't actually know nothing about them.. I suppose I might seriously like them if I actually listened

[04:25:00.0000] <MikeSmith>
or especially if I went to a live show

[04:25:01.0000] <daedb>
I actually expected them to be more fun live than they were, but it wasn't bad.

[04:26:00.0000] <MikeSmith>
I've found there's a lot of music I don't really care about much if I listen to a recording, but can like it at lot of I see it live 

[04:26:01.0000] <mut>
metalheads

[04:27:00.0000] <MikeSmith>
I saw this band a while back and there music is a joke but they actually they actually tore up seriously when they played live -

[04:27:01.0000] <MikeSmith>
http://www.pinkypiglets.com/

[04:28:00.0000] <MikeSmith>
jgraham: I'm pretty sure http://upload.wikimedia.org/wikipedia/commons/0/08/Skyforger_16127_Cernunnos_Fest_11.jpg is actually a scene from the movie Spinal Tap

[04:28:01.0000] <MikeSmith>
and if it's not, it should be

[04:31:00.0000] <jgraham>
Sight. Metal. Can't live with it

[04:31:01.0000] <jgraham>
*Sigh

[04:31:02.0000] <jgraham>
Sigh

[04:31:03.0000] <gsnedders>
At least you've _listened_ to some, unlike too many people who slag it off endlessly.

[04:33:00.0000] <MikeSmith>
jgraham: if you think you've seen the worst of it, try some Christian metal

[04:33:01.0000] <MikeSmith>
that'll change your mind

[04:33:02.0000] <gsnedders>
Hey, there is some good Christian metal

[04:33:03.0000] <gsnedders>
/me hides

[04:34:00.0000] <jgraham>
I don't mean to slag it off endlessly or anything. It's just that it seems to be what *all* the kids here (and by kids, I don't just mean gsnedders, I eman anyone under the age of 40) are into

[04:34:01.0000] <MikeSmith>
"Christian metal" is kind of like "Satanic bake sale"

[04:34:02.0000] <gsnedders>
jgraham: Welcome to Sweden :P

[04:34:03.0000] <daedb>
Especially 80's christian metal, like Stryper :p

[04:35:00.0000] <MikeSmith>
not to use the name of Lord Satan in vain, by the way

[04:35:01.0000] <jgraham>
Nothing like some cookies of deatho or sponge cake of hell to brighten a rainy sunday morning

[04:35:02.0000] <MikeSmith>
daedb: yes, the Stryper guys did indeed have some serious armadillos on their trowsers

[04:36:00.0000] <MikeSmith>
Stryper did spandex better than anybody

[04:37:00.0000] <jgraham>
In other news, the lack of takers for the Ukelele Orchestra is most disappointing

[05:26:00.0000] <Lachy>
jgraham, how many people said they'd go?

[05:26:01.0000] <jgraham>
Lachy: None

[05:27:00.0000] <jgraham>
Unless gsnedders changes his mind I guess

[05:27:01.0000] <gsnedders>
I said I might.

[05:27:02.0000] <jgraham>
s/changes/makes up/

[05:32:00.0000] <Dashiva>
Amazing how big signatures some people on public-html have

[05:33:00.0000] <Dashiva>
And how the readability of quote-heavy threads seems inversely proportional to how much the thread deals with a11y

[05:33:01.0000] <Lachy>
jgraham, you just need to promote it better.

[05:35:00.0000] <jgraham>
Dashiva: My understanding is that is because the requirements for making email understandable to blind users are the exact opposite of the requirements to make the conversations understandable to everyone else

[05:35:01.0000] <jgraham>
This seems to be a deficiency in the email clients avaliable to the blind

[05:37:00.0000] <jgraham>
Lachy: I have used all my reserves of [whatever the antonym of shyness is] already. 

[05:38:00.0000] <Dashiva>
I wonder how a11y people in general feel about making content inaccessible to non-disabled users for the benefit of disabled users

[05:39:00.0000] <MikeSmith>
jgraham: what is Ukulele Orchestra?

[05:39:01.0000] <Lachy>
jgraham, assertiveness is an antonym of shyness

[05:40:00.0000] <jgraham>
MikeSmith: http://www.youtube.com/user/UkuleleOrchestra

[05:40:01.0000] <jgraham>
Lachy: Hmm. Maybe

[05:42:00.0000] <Lachy>
also, the requirements of making e-mail readable by blind people should be the same as for everyone else. But AIUI, the problem is that mail clients don't deal with quotes properly, which seems odd because converting a plain text e-mail to equivalent HTML markup (using <blockquote>, <ul>, etc. is trival

[05:43:00.0000] <Lachy>
and after the conversion, reading the mail is the same as reading an HTML page.

[05:43:01.0000] <jgraham>
Lachy: Yes the quoting thing is just what I meant

[05:43:02.0000] <jgraham>
Non blind people need it to follow the thread. Blind people allegedly get confused

[05:46:00.0000] <MikeSmith>
jgraham: going to a live performance?

[05:46:01.0000] <jgraham>
MikeSmith: That's the plan

[05:46:02.0000] <jgraham>
Although I haven't actually bought any tickets yet

[05:47:00.0000] <MikeSmith>
where are you guys these days? still Liknoping?

[05:47:01.0000] <MikeSmith>
*Linkoping

[05:47:02.0000] <jgraham>
Yes

[05:47:03.0000] <MikeSmith>
hmm, "Liknoping" sounds better

[05:47:04.0000] <MikeSmith>
I think I will call it that from now now

[05:47:05.0000] <MikeSmith>
*now on

[05:49:00.0000] <annevk>
I just call it Sweden

[05:51:00.0000] <jgraham>
I think that's quite unfair to the rest of Sweden

[05:52:00.0000] <MikeSmith>
good god almighty, I did a search just now and I find that there is no heavy metal band yet named Mjöllnir

[05:53:00.0000] <MikeSmith>
what the hell are these heavy-metal guys thinking?

[05:53:01.0000] <MikeSmith>
i mean, "Finntroll" vs. "Mjöllnir"?

[05:53:02.0000] <MikeSmith>
no brainer

[05:53:03.0000] <boblet>
maybe they’re all too busy playing Halo 3?

[05:54:00.0000] <MikeSmith>
those guys should might just as well called themselves MoominTroll instead

[05:54:01.0000] <boblet>
(Master Chief wears Mjolnir-brand underoos)

[05:54:02.0000] <MikeSmith>
dudes, we need to start a heavy metal band named Mjöllnir, right away

[05:55:00.0000] <davidhund__>
would this be a good time (and place) to ask the Swedish Heavy Metal Gods a question re: alt attribute on images in a <figure>?

[05:56:00.0000] <davidhund__>
I was just wondering, from a usability p.o.v., would it be correct to state that an image in a <figure> should always have alternative text (in the 'alt' attribute)?

[05:56:01.0000] <boblet>
if someone could programmatically turn emotions in email into noise, you wouldn’t even need to write any music; just point it at public-html and press rec

[06:00:00.0000] <Philip`>
davidhund__: The spec explicitly allows alt to be omitted if you're in a <figure> with a <figcaption>

[06:00:01.0000] <davidhund__>
But, this means that the figcaption could describe content that's 'not there', right?

[06:01:00.0000] <Philip`>
and indicates it's the appropriate thing to do when you don't have a real textual alternative to the image (e.g. you don't know what's in the image) but you do have some other text that's related to the image (e.g. a title)

[06:02:00.0000] <Philip`>
An <img> with no alt isn't 'not there'

[06:03:00.0000] <Philip`>
<img alt=""> is not there, but just <img> is unknown and dependent on context

[06:03:01.0000] <Philip`>
(as I understand it)

[06:03:02.0000] <davidhund__>
hrm

[06:05:00.0000] <Philip`>
I think the idea is that if you can replace the image with equivalent text, then that text goes in alt; otherwise you shouldn't specify alt at all (and shouldn't put in text that's descriptive but not equivalent - that should go in a <figcaption> instead)

[06:05:01.0000] <davidhund__>
How would <figure><img ... [some chart] alt=""><figcaption>The latest trends in Foo</figcaption></figure> be helpful to people with screenreaders?

[06:07:00.0000] <Philip`>
Empty alt="" should only be used for decorative images or similar things where the image can be omitted entirely; you shouldn't use it in that example

[06:08:00.0000] <davidhund__>
exactly.

[06:08:01.0000] <davidhund__>
Hence my question: "Would it be correct to say that alt text is required on images in a <figure>"

[06:11:00.0000] <Philip`>
No, because there are other examples where alt should be omitted (when it's impossible to do anything that would be helpful to people with screenreaders, and putting bogus alt text would be worse than omitting it and having associated non-alt text in a figcaption etc)

[06:12:00.0000] <Philip`>
like the spec's example of a photo sharing site saying "<figure> <img src="1100670787_6a7c664aef.jpg"> <figcaption>Bubbles traveled everywhere with us.</figcaption> </figure>"

[06:13:00.0000] <Philip`>
(where the site knows nothing about the image, other than a description of it (which is not equivalent to the image itself))

[06:15:00.0000] <davidhund__>
sure, just wondering about the usefulness of the figcaption in that case then…

[06:15:01.0000] <davidhund__>
thanks Philip'

[06:15:02.0000] <MikeSmith>
Lachy: I think you should be the part of Loki in our assemblage of rock gods known as Mjöllnir

[06:15:03.0000] <MikeSmith>
I shall be Thor, of course

[06:16:00.0000] <MikeSmith>
gsnedders: you will be Baldr, the emo-God

[06:16:01.0000] <Philip`>
davidhund__: The usefulness is that UAs can automatically determine the text that's associated with the image, and announce it to the user, so the user has better than no idea about what the image is

[06:17:00.0000] <daedb>
MikeSmith: http://www.vikingarock.se/banden/mjolner.php

[06:18:00.0000] <daedb>
Different spelling, but there is at least one Mjöllnir-band

[06:18:01.0000] <MikeSmith>
oh shit

[06:19:00.0000] <MikeSmith>
well, our first act as a band will have to be to find those guys and kick the crap out of them and steal their name

[06:19:01.0000] <Dashiva>
I thought your "band" was StandardsSuck

[06:19:02.0000] <Philip`>
MikeSmith: You should just apply for a trademark on their name, and then get your lawyers to force them to change

[06:20:00.0000] <MikeSmith>
Dashiva: Marcos is not of Northern European stock, so he can't be in the band

[06:20:01.0000] <boblet>
MikeSmith: who drives the @w3c twitter account?

[06:20:02.0000] <MikeSmith>
or Nordic-Germanic

[06:20:03.0000] <Dashiva>
MikeSmith: He could be the manager

[06:20:04.0000] <MikeSmith>
boblet: a bot

[06:20:05.0000] <daedb>
MikeSmith: Also, http://www.metal-archives.com/band.php?id=22679 (different band, same name, seems more serious)

[06:20:06.0000] <Dashiva>
http://www.bradcolbow.com/archive.php/?p=204

[06:21:00.0000] <boblet>
give the bot my thanks :)

[06:21:01.0000] <Philip`>
davidhund__: (By the way, I don't fully understand the spec's current alt requirements, and think they're crazy (because they're too hard to fully understand or even read), so don't trust anything I say :-) )

[06:21:02.0000] <MikeSmith>
Dashiva: we will let Marcos carry our guitar cases and select the fans who have the privilege to come backstage and bask in our glory

[06:22:00.0000] <MikeSmith>
boblet: the bot says, "give W3C more props"

[06:22:01.0000] <davidhund__>
Philip' yeah alt-attr seems so simple but I find it quite hard to 'get'.  Since a figure is supposed to be a "unit of self-contained content" it is referenced from somewhere, right? So now the subject referenced ('Bubbles') is not described at all. The figcaption is not describing the subject, simply giving more context (in your example)

[06:22:02.0000] <gsnedders>
w00t! ꦕꦫꦏꦤꦤ꧀

[06:23:00.0000] <boblet>
heh

[06:26:00.0000] <Lachy>
MikeSmith, I have no idea waht Mjöllnir is

[06:27:00.0000] <Lachy>
and what makes Thor and Loki rock gods? Thor is the god of thunder, I can't remember what Loki is

[06:27:01.0000] <daedb>
Mjöllnir is Thor's hammer.

[06:28:00.0000] <Philip`>
I remember some old shareware game where Loki was a snake guy who irritatingly could only be hit from behind

[06:29:00.0000] <MikeSmith>
jgraham: Lachy's comments make me think that you may be more worthy of donning the mantle of Loki

[06:29:01.0000] <Philip`>
but fortunately your hammer was like a boomerang so you could do that alright

[06:29:02.0000] <Philip`>
(I guess you were Thor in the game?)

[06:29:03.0000] <Philip`>
Also it had skunks

[06:29:04.0000] <MikeSmith>
jgraham: but I had been thinking you should be Tannhäuser

[06:29:05.0000] <MikeSmith>
which I realize Tannhäuser is not a god

[06:29:06.0000] <MikeSmith>
but he is better than a god

[06:30:00.0000] <MikeSmith>
he is the inspiration for Wagner's greatest music

[06:30:01.0000] <daedb>
/me wants to be Baldr

[06:30:02.0000] <MikeSmith>
and Wagner is the original inspiration behind all heavy metal

[06:30:03.0000] <gsnedders>
daedb: MikeSmith already assigned me to him

[06:31:00.0000] <daedb>
bah

[06:31:01.0000] <MikeSmith>
as heavy metal Gods, we bow before two ÜberLords, Lord Wagner and Lord Satan 

[06:32:00.0000] <Philip`>
http://www.mobygames.com/game/dos/god-of-thunder/screenshots/gameShotId,4870/ - ooh, that was it

[06:32:01.0000] <Philip`>
Thor was so cute

[06:34:00.0000] <MikeSmith>
that is not Thor

[06:34:01.0000] <MikeSmith>
there can be not cuteness in Thor

[06:35:00.0000] <MikeSmith>
only anger and righteous vengefulness

[06:35:01.0000] <MikeSmith>
the greatest incarnation of Thor of all time was this guy:

[06:35:02.0000] <MikeSmith>
http://en.wikipedia.org/wiki/Beta_Ray_Bill

[06:36:00.0000] <Dashiva>
Wait, there are HTML action figures? http://www.bradcolbow.com/archive.php/?p=106

[06:36:01.0000] <MikeSmith>
this is a guy that was such a baddass he fought the existing Thor and beat him and then became Thor in his place

[06:38:00.0000] <MikeSmith>
Dashiva: those action figures are missing their orthopedic chairs and laptops and coffee cups

[06:40:00.0000] <Dashiva>
That sounds more like inaction figures

[06:41:00.0000] <boblet>
lol

[06:43:00.0000] <MikeSmith>
Philip`: the spec's current discussion of alt text is baroque to the point that I think the only thing more it needs now is accompanying footnotes or commentary in the margins, like a Thomspson Study Bible

[06:44:00.0000] <MikeSmith>
I think a graduate course could be constructed around that section

[06:44:01.0000] <MikeSmith>
taught by Jacques Derrida

[06:45:00.0000] <MikeSmith>
and/or Jacques De Molay

[06:47:00.0000] <jgraham>
MikeSmith: In unrelated news, thanks for being roughly the only person on linkedin who isn't entirely po-faced

[06:47:01.0000] <MikeSmith>
jgraham: I don't understand. my linkedin profile is deadly serious

[06:48:00.0000] <davidhund__>
but MikeSmith I don't think Derrida would appreciate the spec metanarrative to start… :-P

[06:48:01.0000] <Dashiva>
I have problems determining whether linkedin is actually professionally useful or not

[06:49:00.0000] <MikeSmith>
Dashiva: one thing, it's useful for hooking up with sexy women

[06:50:00.0000] <MikeSmith>
in my experience at least

[06:50:01.0000] <MikeSmith>
Is it supposed to good for professional stuff of some kind?

[06:51:00.0000] <MikeSmith>
/me is wondering if he somehow has missed understood the "Linked" implications of the "LinkedIn" name

[06:51:01.0000] <Philip`>
I've only ever used it for stalking people

[06:52:00.0000] <MikeSmith>
davidhund__: overall, there is too much clarity in the HTML5 spec to interest Derrida

[06:52:01.0000] <MikeSmith>
he likes ambiguity and confusion

[06:53:00.0000] <knowtheory>
everything is open to interpretation

[06:53:01.0000] <boblet>
MikeSmith: Beta Ray Bill seems like an inspiration for the Halo series Brutes

[06:53:02.0000] <knowtheory>
;)

[06:53:03.0000] <davidhund__>
he would be opposed to the many truth claims for sure :-) It's just a powerplay by a small group… :-P

[06:53:04.0000] <MikeSmith>
davidhund__:  that's why I think he would be much more at home in the section that attempts to give guidance on alt text

[06:53:05.0000] <Dashiva>
MikeSmith: I've got a partial understanding that it's supposed to be useful for making connections, getting job opportunities, etc

[06:54:00.0000] <MikeSmith>
Dashiva: that doesn't sound like much fun

[06:54:01.0000] <davidhund__>
Mikesmith agreed re: alt

[06:55:00.0000] <jgraham>
Dashiva: I would like some evidence to back up that idea. I tend to suspect it is just clever hype to make people buy in

[06:56:00.0000] <MikeSmith>
boblet: I think Beta Ray Bill was an inspiration for many 

[06:56:01.0000] <MikeSmith>
Walt Simonson is one of the greats of all time

[06:57:00.0000] <Dashiva>
jgraham: Yes, that's how I feel too

[06:57:01.0000] <Philip`>
jgraham: Maybe it's just hype, but maybe it's real, and surely you don't want to take the risk of missing out on exciting career opportunities

[06:57:02.0000] <Dashiva>
If it isn't, what does linkedin offer over, say, facebook?

[06:58:00.0000] <jgraham>
/me has a perfectly fine job at the moment

[06:58:01.0000] <Dashiva>
jgraham: Sure, but by the time things change it's a bit late to start preparing :P

[07:00:00.0000] <Philip`>
Maybe tomorrow Microsoft will buy Opera so they can replace Trident and then what'll you do?

[07:00:01.0000] <jgraham>
Laugh

[07:01:00.0000] <gsnedders>
jgraham: "fine"? You call a job in a country where it's still -5°C in March _fine_?

[07:01:01.0000] <jgraham>
I wan't using "fine" as a synonym for "sunny"

[07:04:00.0000] <beowulf>
MikeSmith: o rly? o.O

[07:04:01.0000] <Dashiva>
gsnedders: It's a small price to pay for a liveable july

[07:04:02.0000] <beowulf>
/me has another look at LinkedIn

[07:05:00.0000] <gsnedders>
Dashiva: Well, July in (Eastern) Scotland is liveable… just depressingly gray.

[07:06:00.0000] <beowulf>
MikeSmith++ # man about town

[07:24:00.0000] <zcorpan__>
"Thus I don't think having normative dependencies on non-free specifications could be allowed."

[07:25:00.0000] <zcorpan__>
/me notes that iso 8879 isn't freely available

[07:27:00.0000] <Philip`>
zcorpan: That seems like a good example of why it's a terrible idea

[07:28:00.0000] <Philip`>
Someone should complain about the [BEZIER] reference, because that appears to almost entirely not exist

[07:29:00.0000] <Philip`>
It looks like it probably was published somewhere but I can't find anything except second-hand references to it

[07:29:01.0000] <Philip`>
(and it's a normative dependency)

[07:30:00.0000] <zcorpan>
Philip`: go ahead

[07:30:01.0000] <Philip`>
But the someone would have to come up with a better reference to use instead

[07:30:02.0000] <zcorpan>
no, someone could just complain

[07:30:03.0000] <Philip`>
and I couldn't find any that didn't require payment so I can't tell if they actually define what's needed

[07:31:00.0000] <Philip`>
Someone could, but it wouldn't be very useful

[07:31:01.0000] <jgraham>
References to things that require payment are almost useless

[07:32:00.0000] <Philip`>
Almost but not quite

[07:32:01.0000] <jgraham>
Right, but to the extent that technically less correct, but free, references, should be preferred where possible

[07:33:00.0000] <jgraham>
and references to non-free documents only included if there is literally no other possibility

[07:34:00.0000] <Philip`>
The problem with the BEZIER one is it's not non-free, it's entirely unfindable

[07:58:00.0000] <annevk>
didn't the SVG WG write up the BEZIER math?

[07:58:01.0000] <annevk>
/me wonders how hard it would be for someone to write down the relevant math so we can just include it directly

[08:00:00.0000] <Philip`>
The maths is trivial, it's just like http://upload.wikimedia.org/math/5/e/3/5e32b674a98a9f70e492851f9ad92b61.png

[08:00:01.0000] <Philip`>
but Hixie seemingly wanted something as close to the 'original' source as possible

[08:03:00.0000] <annevk>
seems easier to spec the math and save everyone some time

[08:04:00.0000] <Philip`>
In practice, it's easier to not include a reference or definition at all (I don't see one in SVG) because everyone knows what it means and/or can look it up on Wikipedia

[08:05:00.0000] <Lachy>
Philip`, what is that formula you linked to?

[08:06:00.0000] <Philip`>
Lachy: Cubic Bézier curves

[08:09:00.0000] <Philip`>
(...from http://en.wikipedia.org/wiki/Bézier_curve#Examination_of_cases)

[09:34:00.0000] <MikeSmith>
zcorpan, you around?

[10:28:00.0000] <annevk>
Microsoft just filed a bug on <canvas>: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9189

[10:32:00.0000] <Philip`>
/me told them to :-)

[10:36:00.0000] <AryehGregor>
Does anyone know if they'll be implementing it in IE9, or what?

[10:36:01.0000] <Philip`>
I assume they know, but nothing has been announced

[10:37:00.0000] <Philip`>
Given the detail that they're reviewing the spec in, I assume it's because they're either implementing it or planning to implement it at some point in the future

[10:37:01.0000] <AryehGregor>
I figured that much, but I wonder if it's for IE9 or 10.

[10:37:02.0000] <AryehGregor>
Actually, have they even announced when IE9 will be released?

[10:37:03.0000] <AryehGregor>
Even approximately?

[10:38:00.0000] <Philip`>
/me has no idea

[11:01:00.0000] <LoneStar99>
was told to come here for html5 canvas inquires is that correct?

[11:02:00.0000] <AryehGregor>
LoneStar99, yes, this is one place you can ask about that.

[11:08:00.0000] <LoneStar99>
i have a drawing program that works, fine but since it is a mobile, app it broke.  noticed line 7 of the following makes the code fail http://pastebin.com/fvRpPFeu 

[11:11:00.0000] <LoneStar99>
need to figure out how to not use "context.beginPath();"  but lines are still anti-aliased

[11:11:01.0000] <Philip`>
LoneStar99: It's not clear what that code is meant to be doing

[11:12:00.0000] <Philip`>
It sounds like you're calling it each time the user clicks somewhere?

[11:12:01.0000] <Philip`>
If you don't call beginPath, then every time you call lineTo you're making the current path get longer and longer

[11:13:00.0000] <Philip`>
and each time you call stroke() you're re-drawing every line segment that's ever been added to that path

[11:13:01.0000] <zcorpan>
MikeSmith: am now

[11:13:02.0000] <LoneStar99>
yes it is based o this code: http://devfiles.myopera.com/articles/649/example1.js

[11:13:03.0000] <Philip`>
which might be what makes it look non-antialiased (because it's overdrawing the lines)

[11:13:04.0000] <Philip`>
What you probably want is to do beginPath, and then moveTo with the *previous* mouse coordinates, and then lineTo with the current mouse coordinates

[11:14:00.0000] <Philip`>
so that each time you're just drawing a single line between the previous and current mouse positions

[11:15:00.0000] <Philip`>
Hmm, that example seems to do it wrong (it never resets so it'll stroke longer and longer paths)

[11:16:00.0000] <LoneStar99>
the weird thing is that drawing works, without "context.beginPath();" but once uncommented, it fails to display

[11:16:01.0000] <Philip`>
If you just call beginPath and then lineTo, it's not drawing a line from somewhere to somewhere

[11:16:02.0000] <Philip`>
because you've only given it a single point

[11:16:03.0000] <Philip`>
which is why you first need to moveTo the previous point

[11:18:00.0000] <LoneStar99>
how would I moveTo previous point?

[11:19:00.0000] <LoneStar99>
was chatting with another developer, and he said he does not even use "context.beginPath();"  to draw lines... how would I be able to do that?

[11:20:00.0000] <Philip`>
You need to add some global variables and store ev._x and ev._y in them each time you draw, and then use those values in a moveTo the next time you draw (after beginPath and before lineTo)

[11:21:00.0000] <Philip`>
If you don't use beginPath then you'll never reset to an empty path, so you'll have the problem of repeatedly drawing over the early parts of the path

[11:21:01.0000] <Philip`>
(which is an easy mistake because it's hard to notice, but it eventually makes things go really slow and look ugly)

[11:21:02.0000] <LoneStar99>
simply need to anti-alias anything drawn...  what is the best way?

[11:23:00.0000] <LoneStar99>
this is the sample code working, but without anti-aliasing http://devfiles.myopera.com/articles/649/example1.html 

[11:25:00.0000] <Philip`>
It is antialiasing, but it's got this overdraw problem because it never resets the path

[11:25:01.0000] <Philip`>
so if you move mouse a hundred times then the first part of the line will have been stroked a hundred times

[11:25:02.0000] <Philip`>
and the antialiased grey gets drawn a hundred times and turns black

[11:26:00.0000] <Philip`>
which is why it looks aliased

[11:29:00.0000] <LoneStar99>
kinda confused...

[11:31:00.0000] <Philip`>
If you say: ctx.strokeStyle = 'red'; ctx.moveTo(0, 0); ctx.lineTo(100, 0); ctx.stroke(); ctx.strokeStyle = 'green'; ctx.lineTo(100, 100); ctx.stroke();

[11:31:01.0000] <Philip`>
then it will stroke the first line segment in red, and then it will stroke *both* line segments in green

[11:32:00.0000] <Philip`>
because they're both still part of the current path

[11:32:01.0000] <Philip`>
and that's what that example is doing (except it's a lot more than two line segments, and it's stroking them all in black)

[11:35:00.0000] <LoneStar99>
i need lines to look like this: http://caimansys.com/painter/index.html

[11:35:01.0000] <LoneStar99>
not generic looking...

[11:36:00.0000] <Philip`>
/me goes away for a while

[12:19:00.0000] <Hixie>
annevk: if an implement thinks the maths would have helped them, send me the maths and i'll add it.

[12:50:00.0000] <Philip`>
/me thinks anybody realistically implemented Bézier curves is going to need a lot more information than the equation

[12:50:01.0000] <Philip`>
*implementing

[12:50:02.0000] <Dashiva>
Nah, edge cases never happen in the real world :P

[12:50:03.0000] <Philip`>
It'd only be useful if there was ambiguity about what "cubic Bézier curve" means

[12:50:04.0000] <Philip`>
but there isn't

[12:50:05.0000] <Philip`>
because there's only one definition and everybody uses it

[12:53:00.0000] <Hixie>
i agree, hence "if an implementor thinks" :-)

[13:07:00.0000] <othermaciej>
yay, drafts are published

[13:07:01.0000] <othermaciej>
/me is glad that is over with

[13:10:00.0000] <annevk>
oh wait, I guess I confused bezier with compositing

[13:11:00.0000] <Philip`>
Oh

[13:11:01.0000] <Philip`>
Not quite the same thing :-p

[13:12:00.0000] <Philip`>
The original paper on compositing is online for free so that's nice

[13:13:00.0000] <Hixie>
othermaciej: how long did the process take from start to finish?

[13:13:01.0000] <othermaciej>
Hixie: what do you count as the start?

[13:13:02.0000] <annevk>
yeah I know it's not the same, didn't realize bezier curves had a similar issue

[13:13:03.0000] <Hixie>
othermaciej: when you decided we should try to publish, i guess

[13:14:00.0000] <othermaciej>
start of Call for Consensus to publish: Feb 10

[13:14:01.0000] <othermaciej>
recorded WG decision to publish: Mar 24

[13:15:00.0000] <othermaciej>
er

[13:15:01.0000] <othermaciej>
Feb 24

[13:15:02.0000] <othermaciej>
so about a month, or 8 days, depending on how you care to count

[13:15:03.0000] <Hixie>
not too bad i guess

[13:15:04.0000] <othermaciej>
this signals that if we are serious about the heartbeat requirement, the next publication round should start in 2 months

[13:15:05.0000] <Hixie>
yeah

[13:17:00.0000] <Dashiva>
As discussed in #html-wg earlier, the whole process might get easier if we published more than heartbeat required

[13:25:00.0000] <LoneStar99>
Philip': my code is posted in the last thread of this link http://uoid.me/canvas2

[13:42:00.0000] <othermaciej>
Dashiva: I could get behind publishing monthly 

[13:43:00.0000] <Philip`>
LoneStar99: I think you need something like http://pastie.org/854584

[13:47:00.0000] <LoneStar99>
Philip': something random the original code was written by a dude that just left the chat... 

[13:47:01.0000] <LoneStar99>
sorry the channel "ROBOd "

[13:52:00.0000] <colinsullivan>
Can anyone help me with an HTML5 audio implementation?

[13:53:00.0000] <LoneStar99>
Philip': that works perfect!

[13:53:01.0000] <LoneStar99>
thank you so muc

[13:53:02.0000] <LoneStar99>
h

[14:13:00.0000] <Hixie>
well, we've not put JavaScript on the TR/ page

[14:14:00.0000] <Hixie>
now, even

[14:23:00.0000] <LoneStar99>
Philip': i noticed it works, but when I draw anything, it automatically connect to another line...

[14:28:00.0000] <othermaciej>
Hixie: how come "hide UA text" gets a colored border but "highlight UA text" does not?

[14:55:00.0000] <Hixie>
othermaciej: because it's obvious that text is highlighted, but not at all obvious that text is hidden

[14:56:00.0000] <Hixie>
othermaciej: and i needed a way to remind myself that text was hidden when i was in UA mode so i wouldn't spend hours looking for text that was hidden

[15:12:00.0000] <othermaciej>
Hixie: heh, I see


2010-03-05
[16:55:00.0000] <jarib>
z~

[17:25:00.0000] <MikeSmith>
are there any command-line utilities that, given a file, will report the character encoding of the file?

[17:27:00.0000] <jcranmer>
file, maybe?

[17:35:00.0000] <MikeSmith>
jcranmer: indeed

[17:35:01.0000] <MikeSmith>
thanks

[17:36:00.0000] <MikeSmith>
/me did not know about the --mime-encoding switch in file

[17:55:00.0000] <Hixie>
Philip`: http://canvex.lazyilluminati.com/survey/2007-07-17/analyse.cgi/pages/tagattr/img/longdesc 500

[18:54:00.0000] <MikeSmith>
file command doesn't seem to know the difference between a us-ascii file and a utf-8 one

[19:05:00.0000] <wirepair>
file -bi doesn't work?

[19:07:00.0000] <MikeSmith>
/me tries file -bi

[19:08:00.0000] <MikeSmith>
wirepair: nope

[19:08:01.0000] <wirepair>
and isn't us-ascii part of utf-8? heh

[19:08:02.0000] <jcranmer>
presumably it can't tell if there are no high-bit characters

[19:09:00.0000] <wirepair>
yeah or there's no BOM

[19:10:00.0000] <MikeSmith>
validator.nu seems to be smarter - http://validator.nu/?doc=http%3A%2F%2Fwww.w3.org%2FTR%2Fhtml-markup%2F

[19:10:01.0000] <MikeSmith>
too smart for its own damn good

[19:13:00.0000] <wirepair>
no it's not

[19:13:01.0000] <wirepair>
it's reading the servers charset tag

[19:13:02.0000] <wirepair>
;>

[19:14:00.0000] <wirepair>
Content-Type: text/html; charset=utf-8

[19:14:01.0000] <wirepair>
that's why it disagree's i imagine

[19:14:02.0000] <wirepair>
er disagrees

[19:14:03.0000] <wirepair>
i see no bom nor non-ascii codes in that html file

[19:14:04.0000] <wirepair>
only reference to utf-8 is the server's content-type header

[19:26:00.0000] <MikeSmith>
wirepair: ah

[19:36:00.0000] <wirepair>
just my guess though ;>

[19:37:00.0000] <MikeSmith>
wirepair: I'm not familiar with that part of the validator.nu

[19:38:00.0000] <wirepair>
oh duh it's right there

[19:38:01.0000] <wirepair>
under parser

[19:38:02.0000] <wirepair>
says parser: 'automatically from content-type'

[19:38:03.0000] <MikeSmith>
but the fact that hsivonen has is reporting, "disagrees with the actual encoding of the document", made me think that he must have it doing something to actually check the encoding, not just relying on the headers

[19:39:00.0000] <MikeSmith>
wirepair: I think that only is used for deciding whether to parse is as text/html or as XML

[19:39:01.0000] <wirepair>
gotcha

[19:39:02.0000] <MikeSmith>
/me makes note to ask hsivonen about it when he gets back

[19:40:00.0000] <wirepair>
that site isn't very helpful

[19:40:01.0000] <MikeSmith>
which site?

[19:40:02.0000] <wirepair>
validator

[19:40:03.0000] <MikeSmith>
what's not useful about it?

[19:40:04.0000] <wirepair>
the fact that it doesn't tell you where it decided it matches utf-8

[19:40:05.0000] <wirepair>
and not ascii

[19:41:00.0000] <wirepair>
because i'm looking at the raw bytes, and i see no multibytes just straight up ascii heh

[19:41:01.0000] <MikeSmith>
yeah, it probably could provide more details

[19:41:02.0000] <MikeSmith>
I guess I should just go ahead and look at the htmlparser code and see for myself

[19:43:00.0000] <MikeSmith>
hmm, there's not "ack" macports port, it seems

[19:44:00.0000] <MikeSmith>
/me falls back to find . | xargs grep

[19:49:00.0000] <MikeSmith>
wirepair: http://hg.mozilla.org/projects/htmlparser/file/546412142175/src/nu/validator/htmlparser/io/Driver.java

[19:49:01.0000] <MikeSmith>
in case you're curious

[19:50:00.0000] <MikeSmith>
it seems that at the point where it emits that error, it hasn't recorded how it actually determined the encoding

[19:50:01.0000] <MikeSmith>
/me looks at getActualHtmlEncoding()

[19:56:00.0000] <wirepair>
found something?

[19:59:00.0000] <MikeSmith>
it's using InputSource.getEncoding()

[19:59:01.0000] <MikeSmith>
from org.xml.sax

[20:00:00.0000] <MikeSmith>
http://java.sun.com/j2se/1.5.0/docs/api/org/xml/sax/InputSource.html#getEncoding()

[20:00:01.0000] <MikeSmith>
I have no clue what means getEncoding actually uses to determine the encoding

[20:01:00.0000] <wirepair>
yeh,, no idea heh

[20:07:00.0000] <MikeSmith>
actually, it's just a simple getter

[20:07:01.0000] <wirepair>
wtf

[20:07:02.0000] <wirepair>
heh

[20:08:00.0000] <wirepair>
so something is supplyin gthe stream with an encoding ?

[20:08:01.0000] <wirepair>
kinda makes the whole validation thing moot doesn't it?

[20:08:02.0000] <MikeSmith>
the magic is here: http://hg.mozilla.org/projects/htmlparser/file/546412142175/src/nu/validator/htmlparser/io/HtmlInputStreamReader.java

[20:09:00.0000] <wirepair>
i didn't see a BOM so...

[20:09:01.0000] <wirepair>
again, would be better if the validator says where it no longer matches the expected type

[20:10:00.0000] <MikeSmith>
It seems to be using jchardet and/or ICU4J

[20:10:01.0000] <wirepair>
so people wouldn't have to look through the validator code to figure out how it's determining hehe

[20:10:02.0000] <MikeSmith>
true

[20:10:03.0000] <wirepair>
i say just wait ;>

[20:12:00.0000] <MikeSmith>
well, at this point, I'm reasonably sure it is actually doing genuine character-encoding detection, not just relying on the Content-Type header

[20:12:01.0000] <MikeSmith>
actually, I already knew that, because there are times when it will report that the character-encoding in the Content-Type header doesn't match the actual encoding of the document

[20:13:00.0000] <MikeSmith>
anyway, I need to get some grub

[20:13:01.0000] <MikeSmith>
wirepair: btw, you have a lot of sento in Oota-ku

[20:14:00.0000] <MikeSmith>
a number of which are real onsen

[20:14:01.0000] <MikeSmith>
well, real onsen water

[20:14:02.0000] <wirepair>
ah not much of an onsen person

[20:14:03.0000] <MikeSmith>
http://www21.ocn.ne.jp/~spa-mich/todofuken/tokyo/tokyo.htm

[20:14:04.0000] <MikeSmith>
I am nuts for onsen

[20:15:00.0000] <wirepair>
ha

[20:15:01.0000] <wirepair>
i'll show the wife

[20:15:02.0000] <wirepair>
thanks ;>

[20:16:00.0000] <MikeSmith>
yeah, for some reason you all have more onsen-sento there than any other part of the city, by a wide margin

[20:17:00.0000] <wirepair>
weird..

[20:17:01.0000] <wirepair>
wonder why that is

[20:31:00.0000] <MikeSmith>
wirepair: I think one reason might be that Oota-ku is just really big

[20:31:01.0000] <MikeSmith>
geographically

[20:48:00.0000] <wirepair>
how long you been in japan mike?

[20:51:00.0000] <MikeSmith>
wirepair: 8 years

[20:51:01.0000] <wirepair>
wow, quite a bit of time ;>

[20:51:02.0000] <wirepair>
2x me hehe

[20:51:03.0000] <wirepair>
i imagine your japanese must be pretty damn good by now hehe

[20:54:00.0000] <MikeSmith>
wirepair: my Japanese is actually pretty bad

[20:54:01.0000] <wirepair>
didn't study or gave up studying? 

[20:54:02.0000] <wirepair>
hehe

[20:55:00.0000] <MikeSmith>
I studied some, but mostly learned on the job and from my daughter, who's 12 years old now

[20:55:01.0000] <MikeSmith>
she and I only speak Japanese to each other

[20:56:00.0000] <wirepair>
interesting

[20:56:01.0000] <MikeSmith>
she understands a lot of English, but doesn't speak English much

[20:56:02.0000] <wirepair>
yeah i've seen that, where the child understand what you say, but will just answer in japanese

[20:56:03.0000] <MikeSmith>
yeah

[20:57:00.0000] <MikeSmith>
she is starting to study English more, and will be studying it at school when she starts junior-high in April

[20:57:01.0000] <wirepair>
crazy

[20:57:02.0000] <MikeSmith>
she teaches me a lot of words, actually

[20:57:03.0000] <wirepair>
i imagine her pronounciation will be better than most

[20:57:04.0000] <wirepair>
hehe

[20:58:00.0000] <wirepair>
funny i was teaching a family in laws kid a little english (he's like 3) and his father was there, i said truck in english pronounciation, and his father said it in japanese

[20:58:01.0000] <wirepair>
the kid was like 'which is it?!'

[20:58:02.0000] <MikeSmith>
heh

[20:58:03.0000] <wirepair>
he thought they were completely different words

[20:59:00.0000] <MikeSmith>
yeah, it's like orange, or cake, or many others

[20:59:01.0000] <MikeSmith>
truck is a tough one, though

[20:59:02.0000] <wirepair>
yup

[21:00:00.0000] <MikeSmith>
"tr" is just not amenable to katakana-ized pronunciation

[21:00:01.0000] <wirepair>
chi-a-ru

[21:00:02.0000] <wirepair>
hehe

[21:01:00.0000] <MikeSmith>
but I think it's good for kids to learn that people pronounce the same words in different ways

[21:01:01.0000] <wirepair>
yeah definitely

[21:01:02.0000] <MikeSmith>
it's definitely confusing when they are young, though

[21:03:00.0000] <MikeSmith>
one example is the way some Japanese teachers of English in secondary schools pronounce the word "the"

[21:03:01.0000] <wirepair>
za-?

[21:03:02.0000] <MikeSmith>
which is like "za"

[21:03:03.0000] <wirepair>
yeah

[21:03:04.0000] <MikeSmith>
yeah

[21:04:00.0000] <MikeSmith>
some people think it's just wrong to teach kids to pronounce it that way

[21:04:01.0000] <wirepair>
i don't think too highly of the japanese educational system to be honest though

[21:04:02.0000] <wirepair>
not from the things i've heard

[21:04:03.0000] <MikeSmith>
yeah, the system has some problems, for sure

[21:04:04.0000] <MikeSmith>
but about particular case of "za", I don't think it's a big deal at all

[21:05:00.0000] <MikeSmith>
because for one thing, it's also the way some European non-native-English speakers say it soo

[21:05:01.0000] <wirepair>
yeah, i'm more concerned with my child sitting for 8hours and then being expected to regurgitate the information back

[21:05:02.0000] <wirepair>
;>

[21:05:03.0000] <wirepair>
not that i have any at the moment, but when i do heh

[21:06:00.0000] <MikeSmith>
well, my own secondary education wasn't exactly a thrill ride

[21:07:00.0000] <wirepair>
at the time i thought mine was terrible, looking back and hearing other peoples experiences i guess it was pretty damn good

[21:07:01.0000] <wirepair>
heh

[21:07:02.0000] <MikeSmith>
I had teachers who were supposedly better trained and more enlightened, but still most of them sucked at actually teaching

[21:07:03.0000] <wirepair>
i only had a few like that

[21:07:04.0000] <MikeSmith>
teaching well is a rare talent

[21:07:05.0000] <wirepair>
indeed.

[21:08:00.0000] <MikeSmith>
I taught high school on contract for 3 years or so

[21:08:01.0000] <wirepair>
ah so you got to see it firsthand ;>

[21:09:00.0000] <MikeSmith>
well, it was enough for me to realize I couldn't do it well, and needing to find some other way to earn money

[21:09:01.0000] <wirepair>
ahha

[21:09:02.0000] <wirepair>
gotcha

[21:09:03.0000] <MikeSmith>
anyway, in 6 years of primary school, my daughter has had some pretty good teachers, and some pretty bad ones

[21:09:04.0000] <MikeSmith>
the teachers do make a big difference

[21:10:00.0000] <wirepair>
yeah

[21:10:01.0000] <MikeSmith>
last year, her teacher was a young-guy, first-year teacher, and he didn't really know what he was doing, and her grades went down

[21:10:02.0000] <wirepair>
ahh that's a shame

[21:11:00.0000] <MikeSmith>
this year, she had a much older, more experienced, no-nonsense teacher and she learned a lot more and turned all her grades around

[21:12:00.0000] <MikeSmith>
I know there are some good teachers in secondary schools here, too

[21:12:01.0000] <MikeSmith>
along with mediocre and bad ones

[21:13:00.0000] <MikeSmith>
it's partly just the luck of the draw, I guess

[21:13:01.0000] <wirepair>
yeah i'd like to know what the curriculum is like, my understanding is there's little to know real teaching of the scientific process until college

[21:13:02.0000] <wirepair>
s/know/no

[21:14:00.0000] <wirepair>
but hey what do i know, i never even went to college ;>

[21:14:01.0000] <MikeSmith>
I think the system expects -- or even requires -- a lot more personal responsibility on the part of students and their parents

[21:15:00.0000] <wirepair>
yeah, unfortunately what i've seen here is an unbelievable amount of pressure on teachers to take care of kids

[21:15:01.0000] <MikeSmith>
the thing I value most about my college experience was the many opportunities it provided for relaxed dope smoking

[21:15:02.0000] <wirepair>
hahah

[21:16:00.0000] <wirepair>
see i did that in private school for 2 years, i got it out of my system

[21:16:01.0000] <wirepair>
then i got a job during the bubble, it bust, and all my friends got out of college and couldn't find jobs when i had 4 years experience ;>

[21:17:00.0000] <wirepair>
my timing was impeccable

[21:17:01.0000] <MikeSmith>
you planned that well

[21:18:00.0000] <wirepair>
just very lucky

[23:12:00.0000] <othermaciej>
MikeSmith: check out the status on this: http://www.w3.org/TR/2010/WD-html-bidi-20100304/#bidi-isolation

[23:12:01.0000] <othermaciej>
MikeSmith: "This document is an editors' copy that has no official standing."

[23:13:00.0000] <othermaciej>
that's probably not right

[23:13:01.0000] <MikeSmith>
oops

[23:13:02.0000] <othermaciej>
MikeSmith: also I surprisingly hadn't heard about that document

[23:13:03.0000] <MikeSmith>
I will fix it in-place now

[23:13:04.0000] <othermaciej>
/me wonders why the proposals in that document weren't proposed to the HTML WG and hopes they will be at some point

[23:14:00.0000] <MikeSmith>
othermaciej: I probably should not admit this, but the first I saw of that doc was this announcement

[23:15:00.0000] <MikeSmith>
I think Richard mentioned something about it to me, but it didn't register as being something closely related to HTML

[23:15:01.0000] <othermaciej>
glad it's not just me

[23:16:00.0000] <othermaciej>
Looking at the TransReq Philippe did say it should be be brought to the attention of the HTML WG

[23:17:00.0000] <MikeSmith>
well

[23:18:00.0000] <MikeSmith>
it's odd that they chose to publish it as WD first

[23:18:01.0000] <MikeSmith>
the "bring to attention of the HTML WG" step would seem like it should have better occurred first

[23:18:02.0000] <MikeSmith>
wait

[23:19:00.0000] <MikeSmith>
it's about bidi

[23:19:01.0000] <MikeSmith>
so they are doing everything backwards

[23:19:02.0000] <MikeSmith>
sorry, not backwards, just different-wards

[23:19:03.0000] <MikeSmith>
it's all relative

[23:23:00.0000] <MikeSmith>
hsivonen: btw, I wanted to ask you about the character-detection mechanism that validator.nu uses

[23:24:00.0000] <MikeSmith>
hsivonen: http://validator.nu/?doc=http%3A%2F%2Fwww.w3.org%2FTR%2Fhtml-markup%2F

[23:24:01.0000] <MikeSmith>
"Internal encoding declaration us-ascii disagrees with the actual encoding of the document (utf-8)"

[23:25:00.0000] <MikeSmith>
but that document actually has no non-ascii characters in it, and no BOM

[23:26:00.0000] <MikeSmith>
so I assume the character-detection code is actually also using the Content-type header

[23:27:00.0000] <MikeSmith>
not just checking the "actual encoding" of the document

[23:28:00.0000] <MikeSmith>
if so, I'm wondering if that error message might be refined to say, "disagrees with the actual encoding of the document and/or the Content-Type header the document is being served with (utf-8)"

[23:28:01.0000] <MikeSmith>
or something similar

[23:30:00.0000] <hsivonen>
I guess the actual encoding is the one named on the http level in this case

[23:30:01.0000] <zcorpan>
MikeSmith: pong

[23:31:00.0000] <MikeSmith>
hsivonen: OK

[23:31:01.0000] <hsivonen>
maybe 'actual' is a philosophical/onntological question and the message should say something else

[23:31:02.0000] <othermaciej>
heh, bidi

[23:32:00.0000] <MikeSmith>
hsivonen: yeah, the word "actual" is what threw me

[23:32:01.0000] <hsivonen>
in this case, actual means the encoding that's actually used for decoding

[23:32:02.0000] <MikeSmith>
ah

[23:32:03.0000] <MikeSmith>
that's more like maybe "effective"

[23:33:00.0000] <MikeSmith>
I can't think of any great suggestions other than that one

[23:33:01.0000] <MikeSmith>
which is not great

[23:33:02.0000] <MikeSmith>
anyway, I'm not sure it's worth changing

[23:33:03.0000] <MikeSmith>
I had just been wondering

[23:34:00.0000] <MikeSmith>
zcorpan: I wanted to tell you I made some more changes to the markup doc based on the errors you reported a while back

[23:35:00.0000] <MikeSmith>
zcorpan: http://www.w3.org/TR/html-markup/syntax.html#doctype-syntax

[23:36:00.0000] <MikeSmith>
my rationale for organizing the description of doctypes that way (instead of the way Hixie has it) is that I wanted to try to have the description of what a "normal" HTML doctype is to be as simple as possible

[23:37:00.0000] <MikeSmith>
I don't know if "normal doctype" is the best choice as far a terminology

[23:38:00.0000] <MikeSmith>
but anyway, I did correct the description of "deprecated doctype" to remove any reference to the XML definition of what a docytpe is

[23:38:01.0000] <MikeSmith>
http://www.w3.org/TR/html-markup/syntax.html#deprecated-doctype

[23:39:00.0000] <annevk>
the piece on W3C news about H:TML says more about what it is not than what it is :/

[23:39:01.0000] <MikeSmith>
heh

[23:39:02.0000] <zcorpan>
MikeSmith: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9145

[23:40:00.0000] <MikeSmith>
annevk: my fault, since it's word-for-word from the abstract

[23:40:01.0000] <MikeSmith>
zcorpan: looking now

[23:40:02.0000] <MikeSmith>
OK

[23:41:00.0000] <MikeSmith>
/me makes that same change now

[23:42:00.0000] <MikeSmith>
hsivonen: "computed encoding"?

[23:42:01.0000] <zcorpan>
MikeSmith: "Conditionally, depending on whether it is part of a permitted-public-ID-system-ID-combination, the following parts, in exactly the following order:" doesn't say whether the condition is true or false

[23:42:02.0000] <zcorpan>
MikeSmith: maybe "If it (what?) is part of a permitted-..."

[23:42:03.0000] <MikeSmith>
yeah

[23:43:00.0000] <MikeSmith>
problem is, I don't know how to describe that condition

[23:43:01.0000] <MikeSmith>
the condition is basically, if it ends up looking like one of the doctypes in the list below that, then true

[23:43:02.0000] <MikeSmith>
otherwise false

[23:44:00.0000] <zcorpan>
it's a bit confusing

[23:44:01.0000] <zcorpan>
i think Hixie's text is less confusing, at least to me :)

[23:45:00.0000] <MikeSmith>
I may end up changing it to that

[23:46:00.0000] <MikeSmith>
I wanted to first try to see if there were any way to make it any clearer

[23:46:01.0000] <zcorpan>
MikeSmith: maybe you could have two tables, one for html which has optional SI, and one for xhtml which doesn't

[23:46:02.0000] <MikeSmith>
hmm, yeah

[23:46:03.0000] <MikeSmith>
that would be slightly better at least

[23:47:00.0000] <MikeSmith>
zcorpan: I also changed the description of what a conformant XML document is (the other error you reported)

[23:48:00.0000] <hsivonen>
I wonder if a MacBook can be fooled to switch its VNC resolution to something higher than the res of the built-in screen without connecting anything to the mini-DVI port

[23:48:01.0000] <MikeSmith>
http://www.w3.org/TR/html-markup/documents.html#syntax-document-xml

[23:48:02.0000] <othermaciej>
annevk: looks like it's about 50/50 between what it is and what it isn't

[23:49:00.0000] <othermaciej>
it is funny that the one document not produced by the HTML WG (and something of a surprise to us) has by far the longest description

[23:49:01.0000] <MikeSmith>
zcorpan: I am planning to make a couple more tweaks based on the other message you sent, about "this specification does not define any additional syntax-level requirements"

[23:51:00.0000] <annevk>
othermaciej, the W3C clearly hates HTML5 much like it hates babies

[23:52:00.0000] <othermaciej>
annevk: I thought that was Adobe

[23:52:01.0000] <annevk>
oh, they're not the same?

[23:52:02.0000] <othermaciej>
heh

[23:52:03.0000] <othermaciej>
seriously though, HTML5 plus three other documents get less total space in the description than any other drafts

[23:52:04.0000] <zcorpan>
/me wonders what he has to do to see the browser choice screen

[23:54:00.0000] <zcorpan>
http://www.w3.org/TR/2010/WD-html-bidi-20100304/#br-as-separator - wonder if that matches what html5 says

[23:55:00.0000] <annevk>
zcorpan, visit http://www.browserchoice.eu/ ?

[23:56:00.0000] <othermaciej>
I dunno, it's funny that they used HTML4 as their baseline reference

[23:56:01.0000] <annevk>
zcorpan, I think dbaron mentioned that HTML5 should say something about that (iirc)

[23:56:02.0000] <annevk>
zcorpan, not sure if it was followed through

[23:56:03.0000] <zcorpan>
annevk: i thought i'd get something automatically by setting ie to default browser and doing a windows update

[23:57:00.0000] <annevk>
zcorpan, oh like that

[23:57:01.0000] <annevk>
zcorpan, dunno

[23:58:00.0000] <zcorpan>
how many of the non-top-5 browsers in the screen use trident?

[23:58:01.0000] <paul_irish>
maxthon does

[23:59:00.0000] <zcorpan>
yes, and flock doesn't, what about the others?

[23:59:01.0000] <zcorpan>
k-meleon also doesn't

[23:59:02.0000] <zcorpan>
avant is trident

[00:00:00.0000] <zcorpan>
looks like sleipnir is a trident browser

[00:01:00.0000] <zcorpan>
flashpeak too

[00:01:01.0000] <paul_irish>
and maybe AOL? 

[00:02:00.0000] <othermaciej>
there are several trident browsers, according to the news

[00:03:00.0000] <othermaciej>
did they fix the randomization on that page yet?

[00:04:00.0000] <zcorpan>
trident: 6, gecko: 3, webkit: 2, presto: 1

[00:05:00.0000] <othermaciej>
zcorpan: looks like it doesn't match HTML5 - "A br element does not separate paragraphs for the purposes of the Unicode bidirectional algorithm. [BIDI]" http://dev.w3.org/html5/spec/Overview.html#the-br-element

[00:08:00.0000] <annevk>
/me wonders what browsers implement now

[00:08:01.0000] <othermaciej>
according to the bidi requirements draft, WebKit and IE do not implement the HTML5/HTML4 rule but instead do treat <br> as a paragraph break for bidi algorithm purposes

[00:09:00.0000] <othermaciej>
the draft also says this behavior is better (I think)

[00:11:00.0000] <annevk>
I see

[00:12:00.0000] <annevk>
I wonder why they want through the trouble of publishing a separate document rather than sending feedback on HTML5...

[00:17:00.0000] <othermaciej>
I dunno

[00:18:00.0000] <othermaciej>
this could be, like, 10-20 bug reports

[00:18:01.0000] <othermaciej>
most of which would be uncontroversial except perhaps trivial syntax details

[00:19:00.0000] <othermaciej>
(for example the proposed bdi attribute should be a standard html-style boolean attribute; shouldn't use attributes with underscore; etc)

[00:20:00.0000] <annevk>
some of the guidelines should not be in HTML I think, e.g. the bit about <pre> in http://www.w3.org/TR/2010/WD-html-bidi-20100304/#newline-as-separator seems like it should be in CSS

[00:24:00.0000] <MikeSmith>
zcorpan: how's the following sound?

[00:24:01.0000] <MikeSmith>
[[

[00:25:00.0000] <MikeSmith>
A conformant document in the XML syntax must be a well-formed XML document, as defined in the XML specification [XML], and its root element, as defined in the XML specification [XML], must be an html element.

[00:25:01.0000] <MikeSmith>
]]

[00:25:02.0000] <zcorpan>
MikeSmith: the next paragraph seems to repeat the well-formedness requirement

[00:26:00.0000] <roc>
'bdi' is a terrible name

[00:26:01.0000] <zcorpan>
MikeSmith: how about saying it must be namespace-well-formed, once?

[00:26:02.0000] <roc>
some of these issues were discussed briefly on public-html ages ago

[00:26:03.0000] <MikeSmith>
zcorpan: OK

[00:27:00.0000] <MikeSmith>
roc: yeah

[00:27:01.0000] <MikeSmith>
I recall now Richard mentioning the <br> issue to me a while back

[00:35:00.0000] <MikeSmith>
zcorpan: 

[00:35:01.0000] <MikeSmith>
"A conformant document in the XML syntax must be a namespace-well-formed XML document, as defined in the XML specification [XML] and Namespaces in XML 1.0 specification [XMLNS], and its root element must be an html element."

[00:35:02.0000] <MikeSmith>
and I turned the "Documents in the XML syntax must not make use of any features of the HTML syntax that do not follow XML well-formedness constraints (for example, documents in the XML syntax must not use unquoted attribute value syntax and must not omit tags).

[00:35:03.0000] <MikeSmith>
" part into a note

[00:37:00.0000] <zcorpan>
s/as defined in the XML specification [XML] and //

[00:38:00.0000] <MikeSmith>
OK

[00:38:01.0000] <zcorpan>
isn't the root element required by the content models, btw?

[00:41:00.0000] <MikeSmith>
zcorpan: it's otherwise not explicitly stated anywhere for the XML/XHTML document case

[00:41:01.0000] <MikeSmith>
it is stated explicitly for the text/html case -

[00:41:02.0000] <MikeSmith>
file:///opt/workspace/html5/markup/documents.html#conformant-documents

[00:42:00.0000] <MikeSmith>
oops

[00:43:00.0000] <MikeSmith>
http://dev.w3.org/html5/markup/documents.html#syntax-document-html

[00:43:01.0000] <zcorpan>
/me gets a twitter account

[00:48:00.0000] <MikeSmith>
hsivonen: "operative encoding"

[00:51:00.0000] <Philip`>
Hixie: The analyse.cgi stuff is intentionally broken since I moved to a new server and don't really want to bother setting up all the databases and everything

[00:51:01.0000] <Philip`>
It's fixable but it depends on whether anybody cares enough

[00:53:00.0000] <annevk>
/me follows zcorpan

[00:57:00.0000] <Hixie>
Philip`: k

[02:43:00.0000] <mut>
#

[02:43:01.0000] <jgraham>
!

[02:43:02.0000] <annevk>
~

[02:44:00.0000] <mut>
$(':(').show();

[02:44:01.0000] <jgraham>
‽

[02:44:02.0000] <annevk>
you usually hide sad faces

[02:45:00.0000] <jgraham>
Only if you're a clown

[02:45:01.0000] <gsnedders>
jgraham: It's a Javascript library, duh. $ is awesome. :P

[02:45:02.0000] <gsnedders>
Normally my eyes are _above_ my mouth.

[02:46:00.0000] <jgraham>
/me is horribly confused

[02:46:01.0000] <gsnedders>
/me is good at making jgraham horribly confused

[02:46:02.0000] <jgraham>
You are quite the confusing person

[02:47:00.0000] <mut>
im about to get trout slappy

[02:47:01.0000] <beowulf>
i haven't been slapped by a wet trout in years

[02:47:02.0000] <mut>
"This program has stopped responding" <----- :(

[02:47:03.0000] <jgraham>
Fun game: read ꦕꦫꦏꦤ꧀ 

[02:48:00.0000] <beowulf>
does the internet still have such things?

[02:48:01.0000] <jgraham>
(gsnedders isn't allowed to play)

[02:48:02.0000] <mut>
would you like to lose all of your work

[02:48:03.0000] <gsnedders>
Spoil sport.

[02:48:04.0000] <mut>
/me troutslaps beowulf

[02:48:05.0000] <mut>
;)

[02:48:06.0000] <workmad3>
/me slaps mut around a bit with a wet trout

[02:48:07.0000] <mut>
haha

[02:48:08.0000] <gsnedders>
jgraham is such a mean guy.

[02:48:09.0000] <mut>
ewwww moist

[02:48:10.0000] <workmad3>
you need to get a troutslap right :P

[02:49:00.0000] <mut>
ah im on a java applet, no troutslaps

[02:49:01.0000] <workmad3>
well, I just type em in manuallp

[02:49:02.0000] <workmad3>
*manually

[02:49:03.0000] <jgraham>
/me feels his game may have been lost in YAMPITS

[02:49:04.0000] <workmad3>
but I think mIRC used to have a /slap command to do it for you :)

[02:49:05.0000] <mut>
show a man a troutslap button, he will troutslap for a day, teach a man /me, and he will troutslap for life

[02:50:00.0000] <beowulf>
lol

[02:50:01.0000] <gsnedders>
jgraham: That's what you get for posting arbitary Unicode codepoints

[02:51:00.0000] <jgraham>
They're not arbitay. That's the point :)

[02:51:01.0000] <gsnedders>
Well, seeming I'm forbidden from playing, you could probably guess I know what they are.

[02:52:00.0000] <annevk>
jgraham, for that I would have to be funny

[02:52:01.0000] <beowulf>
jgraham: ok, explain the game, i didn't get to kill or blow anyone up

[02:53:00.0000] <Philip`>
Google is worringly fast

[02:53:01.0000] <jgraham>
annevk: Now you confused me

[02:53:02.0000] <gsnedders>
Go annevk!

[02:53:03.0000] <Philip`>
I searched for jgraham's string and it found the IRC logs from here

[02:54:00.0000] <jgraham>
That's how Hixie reads irc

[02:56:00.0000] <annevk>
/me waits for jgraham to implode

[02:57:00.0000] <mut>
/me sidesteps away from jgraham

[02:58:00.0000] <annevk>
/me advices people with nicks starting with a "j" to invoke /nick

[02:59:00.0000] <zcorpan>
jgraham: is it carakan?

[02:59:01.0000] <gsnedders>
zcorpan: yes.

[02:59:02.0000] <zcorpan>
yay

[03:01:00.0000] <annevk>
MessagePort.onmessage is somewhat ugly

[03:01:01.0000] <annevk>
but I guess it works

[03:03:00.0000] <jgraham>
zcorpan: Specifically is is the word "carakan" spelt out in Carakan symbols

[03:05:00.0000] <jgraham>
Sadly I can't find any fonts with support for it

[03:09:00.0000] <Philip`>
jgraham: Could probably make one easily by renumbering the font in http://hanacaraka.fateback.com/dok&down.htm

[03:12:00.0000] <jgraham>
Intersting

[03:14:00.0000] <Philip`>
Not really, because there's no reason to ever read or write anything in that language

[03:16:00.0000] <jgraham>
Sure there is. Just 30 minutes ago it would have helped you understand this very channel

[03:30:00.0000] <gsnedders>
Philip`: That's what I thought yesterday.

[03:30:01.0000] <gsnedders>
Philip`: If you're in school in Java it's useful

[03:40:00.0000] <virtuelv>
Opera's IRC client has /slap which slaps you around a bit with a wet trout

[03:41:00.0000] <virtuelv>
Aren't google going to start realtime indexing of results, soon?

[03:41:01.0000] <virtuelv>
as in you can use PubSubHubbub

[03:42:00.0000] <jgraham>
Yes

[03:44:00.0000] <zcorpan>
/me slaps a large trout around a bit with virtuelv

[03:45:00.0000] <jgraham>
/me thinks he has accidentially discovered that there are two bands called Camera Obscura

[03:45:01.0000] <gsnedders>
By Spotify being confused

[03:45:02.0000] <gsnedders>
?

[03:45:03.0000] <zcorpan>
virtuelv: it's large, not wet

[03:45:04.0000] <jgraham>
gsnedders: Yes

[03:45:05.0000] <gsnedders>
I see that too often

[03:46:00.0000] <jgraham>
It's a little jarring when you go from melodic Scottish pop to... something

[03:46:01.0000] <jgraham>
s/pop/indie pop/ or something

[03:49:00.0000] <jgraham>
Ah "post hardcore" apparentley

[03:49:01.0000] <jgraham>
At least according to wikipedia

[03:49:02.0000] <gsnedders>
Oh dear.

[03:49:03.0000] <jgraham>
Oh dear isn't even close

[04:28:00.0000] <mut>
haha sweet, my canvas page works on my android phone! :)

[06:37:00.0000] <Lachy>
it just seems like the DE proponents won't be happy with anything less than XML-style namespaces, despite having so many extensibility points in HTML5 already.

[06:37:01.0000] <zcorpan>
http://chikuyonok.ru/ambilight/

[06:38:00.0000] <Lachy>
nice.

[06:40:00.0000] <gsnedders>
Lachy: There's a lot of specs already written that are designed for the current namespace structure within the DOM, and none of the HTML5 extensbility points allow me to put, e.g., dc into HTML5

[06:40:01.0000] <LoneStar99>
Morning all!   

[06:41:00.0000] <Lachy>
gsnedders, so? I consider that a feature, not a bug

[06:41:01.0000] <gsnedders>
Lachy: But what do I do when I want to add DC to my markup?

[06:41:02.0000] <LoneStar99>
Can anything be drawn on canvas which does not use " beginPath(); "?

[06:42:00.0000] <gsnedders>
Lachy: So that I can take my HTML document and pass the DOM into my existing toolchain?

[06:42:01.0000] <Lachy>
what do you want to add DC markup for in the first place, which isn't catered for by existing metadata elements in HTML, or which couldn't be addressed through microformats or microdata?

[06:44:00.0000] <gsnedders>
Lachy: The summary of a section cannot be fetched programmatically; µf and µdata both require forks in my codebase at a DOM level for handling HTML differently to generic XML.

[06:44:01.0000] <Philip`>
LoneStar99: Not sure what you mean

[06:45:00.0000] <Philip`>
LoneStar99: You ought to use beginPath every time you start drawing a new path, but things like drawImage and fillRect don't use paths

[06:45:01.0000] <Lachy>
gsnedders, are you giving a serious use case here and actually arguing that we should support namespaces in HTML, or are you just being hypothetical?

[06:45:02.0000] <LoneStar99>
in other words, can a drawing "sketching" be done without using "beginPath();" "moveTo"  and  "lineTo"

[06:45:03.0000] <Philip`>
LoneStar99: You could do it by just drawing images or drawing lots of little rectangles, but that'd be a silly way to draw lines compared to using lineTo :-)

[06:45:04.0000] <gsnedders>
Lachy: I'm trying to show you there are serious use cases here, and that people do have codepaths to deal with these things and supporting the HTML 5 way of doing things leads to little gain but a lot of additional complexity.

[06:47:00.0000] <Lachy>
sure, but the additional complexity of trying to support namespaces in HTML, compared with the little gained by supported those largely obscure and insignificant cases in practice, proves it's just not worth it

[06:47:01.0000] <gsnedders>
No, the additional complexity _depends_ upon the serialization chosen.

[06:48:00.0000] <Lachy>
then don't choose HTML if you depend on XML-specific features

[06:48:01.0000] <jgraham>
Requiring namespaces introduces a lot of complexity for authors

[06:48:02.0000] <jgraham>
By comparison I have rather little sympathy for implementors

[06:48:03.0000] <gsnedders>
Lachy: Feeds online are not an obscure or insignificant case, and currently require duplicated data

[06:49:00.0000] <gsnedders>
jgraham: Indeed, I think there are plenty of ways to cope with namespaces without additional complexity for authors in the pure-HTML case.

[06:49:01.0000] <Lachy>
gsnedders, wtf do feeds have to do with supporting namespaces in the HTML serialisation?

[06:49:02.0000] <jgraham>
e.g. not having namespaces?

[06:49:03.0000] <jgraham>
(and I also don't get the relevance of feeds)

[06:50:00.0000] <LoneStar99>
Philip`:  how would I be able to drawing a sketching "drawing" tons of rectangles? 

[06:50:01.0000] <LoneStar99>
with

[06:50:02.0000] <gsnedders>
Lachy: Why do I have to post both an HTML and Atom edition of my blog for feed readers? Why can feed readers not programmatically read content from the HTML?

[06:50:03.0000] <Philip`>
gsnedders: Because HTML parsers are rubbish and buggy

[06:51:00.0000] <Philip`>
or slow

[06:52:00.0000] <jgraham>
gsnedders: They can. There is no problem here

[06:52:01.0000] <gsnedders>
Philip`: All OS X feed readers have a decent HTML parser available at an OS level; I presume the same is true on Windows with mshtml.ddl

[06:52:02.0000] <gsnedders>
jgraham: But they cannot extract all the semantics they can get with Atom

[06:53:00.0000] <jgraham>
But DE wouldn't magically make that happen because in practice no one would use the extensions anyway

[06:53:01.0000] <Philip`>
gsnedders: Does it provide an actual standalone HTML parser than gives you back a DOM and is just as easy to use as XML parser libraries?

[06:53:02.0000] <jgraham>
An author that cares can already use the HTML->Atom conversion algorithm to get almost everything

[06:53:03.0000] <LoneStar99>
Philip': issue I have with "beginPath();" is that  my code used to work without a  "moveTo();"  the code now, needs a " moveTo();" and notice on drawings, line are automatically connected when they should not be connected

[06:54:00.0000] <gsnedders>
Philip`: My memory from when I looked at it before was that it wasn't very hard.

[07:05:00.0000] <Lachy>
gsnedders, there have been various attempts at doing that (hAtom, the "extract an Atom feed" algorithm in HTML5), but so far it seems the market isn't really interested in pushing those kinds of techniques. Besides having a CMS generate an Atom feed automatically is a solved problem that authors just don't have to worry about.

[07:09:00.0000] <gsnedders>
So duplicated data isn't a problem when we have tools?

[07:12:00.0000] <Lachy>
if it wasn't already a solved problem, I'd care more about the duplication

[07:21:00.0000] <Philip`>
gsnedders: Data isn't duplicated - there's a single copy of each blog post in the database

[07:21:01.0000] <Philip`>
and so you can avoid all the problems like duplicates getting out of sync with each other

[07:40:00.0000] <mut>
anyone know of a good canvas tutorial?

[07:45:00.0000] <LoneStar99>
mut: what u trying to do with canvas?

[07:46:00.0000] <miketaylr>
mut: http://dev.opera.com/articles/view/html-5-canvas-the-basics/ && https://developer.mozilla.org/en/Canvas_tutorial are pretty good

[07:46:01.0000] <mut>
im drawing staircases in canvas

[07:47:00.0000] <mut>
but just thinking of some other things to do, which are currently beyond me :P

[07:47:01.0000] <LoneStar99>
yeah, try the above links, those are the ones I have uded

[07:47:02.0000] <LoneStar99>
used

[07:47:03.0000] <mut>
thanks mike btw :)

[07:48:00.0000] <mut>
yea im sure I can do what i was planning, but I have had a few more ideas for usablility (zoom and drag functionality ala googlemaps)

[08:13:00.0000] <TabAtkins>
Oh man, only 5 days before my lightcone envelopes HR222.

[08:15:00.0000] <Philip`>
http://thomas.loc.gov/cgi-bin/bdquery/z?d109:h.r.00222: - "To prohibit the expenditure of Federal funds to conduct or support research on the cloning of humans, and to express the sense of the Congress that other countries should establish substantially equivalent restrictions."

[08:16:00.0000] <Philip`>
TabAtkins: ^ That HR222?

[08:16:01.0000] <TabAtkins>
No, look for the first one involving stars.

[08:17:00.0000] <Philip`>
Oh

[08:18:00.0000] <Philip`>
Happy almost about 24.3rd birthday, then

[08:20:00.0000] <Dashiva>
http://twitter.com/heycam/status/10003477514

[08:20:01.0000] <Dashiva>
Maybe I should tell heycam how buses in Norway don't play radio anymore because of inspectors from the local RIAA equivalent

[08:47:00.0000] <mut>
hmm opera dosent seem to handle save() and restore() when combined with scale()...well not the same as firefox or ie +excanvas

[08:48:00.0000] <Philip`>
The problem might be that it applies the transformation when you call stroke/fill, instead of applying it when you're adding points to the path

[08:48:01.0000] <Philip`>
(so if you change the transformation between constructing the path and stroking/filling then it'll break)

[09:11:00.0000] <AryehGregor>
Where are the W3C Editor's Drafts?

[09:11:01.0000] <AryehGregor>
They seem to be cunningly hidden.

[09:11:02.0000] <AryehGregor>
Even the link on the HTMLWG page to the latest editor's draft links to a WD.

[09:12:00.0000] <Philip`>
http://dev.w3.org/html5/ ?

[09:13:00.0000] <AryehGregor>
Oh, wait, I get it.

[09:13:01.0000] <Philip`>
I think the ones like http://dev.w3.org/html5/spec/ just haven't been changed back to the ED colours yet

[09:13:02.0000] <AryehGregor>
The latest editor's drafts are actually WDs right now?

[09:13:03.0000] <AryehGregor>
Or something.

[09:17:00.0000] <Philip`>
AryehGregor: Yes

[09:17:01.0000] <Philip`>
They were changed in e.g. http://dev.w3.org/cvsweb/html5/spec/index.html.diff?r1=1.807&r2=1.808

[09:18:00.0000] <Philip`>
("front matter updates for publication this week", 2.5 weeks ago)

[09:18:01.0000] <Philip`>
and will presumably be changed back soon

[09:26:00.0000] <LoneStar99>
Philip': you know of a tutorial, of how to draw a lines with tons of squares, like you mentioned before?

[09:27:00.0000] <Philip`>
LoneStar99: I expect there's lots of things talking about how to implement line rasterization algorithms, but it seems like a terrible idea :-)

[09:28:00.0000] <LoneStar99>
Philip: it does, but i know of an app, that does it and works really well...

[09:29:00.0000] <annevk>
omg http://www.w3.org/Bugs/Public/show_bug.cgi?id=9197

[09:29:01.0000] <annevk>
evil

[09:29:02.0000] <Philip`>
Using beginPath/moveTo/lineTo is the right way to draw lines, and it should be able to do everything you need, so if your code has bugs then you just need to fix those bugs

[09:30:00.0000] <LoneStar99>
Philip`: the thing it might not be my app, it might be the OS and need to find a work around...

[09:33:00.0000] <Philip`>
That seems unlikely

[09:33:01.0000] <LoneStar99>
tried everything... it just seems that my app randomly begins to connect lines and i think it has to do with "context.moveTo(this.previous_x, this.previous_y);"  but it need the line to work...

[09:35:00.0000] <Philip`>
It sounds like you just have a bug somewhere, but I don't know what your code is doing so I have no idea where it might be wrong

[09:38:00.0000] <LoneStar99>
Philip' http://pastebin.com/uNxvLdL4

[09:43:00.0000] <Philip`>
LoneStar99: You need to set previous_x and previous_y in mousedown, otherwise they'll be left as whatever they were in the last mousemove even if you released the mouse, I guess

[09:46:00.0000] <LoneStar99>
Philip': have tried that and it still begins to screw up after awhile

[09:47:00.0000] <TabAtkins>
Tried logging the events and the values they see for previous_x and previous_y?

[09:48:00.0000] <Philip`>
LoneStar99: By the way, why is the drawImage line there?

[09:48:01.0000] <Philip`>
Also, why is the stroke() not at the end of mousemove?

[09:50:00.0000] <tantek>
Hixie, does microdata itemref permit referencing an id of an ancestor?

[09:50:01.0000] <tantek>
e.g. <span itemscope id=a><span itemprop=turtle itemscope id=b itemref=a></span></span>

[09:51:00.0000] <TabAtkins>
It might run afoul of the cycle-breaking algo, but there's no explicit restriction against such a thing.

[09:51:01.0000] <TabAtkins>
I haven't kept up with the cycle algo, though.

[09:51:02.0000] <TabAtkins>
foolip would have more info, I think.

[09:51:03.0000] <tantek>
TabAtkins - what cycle breaking algo? none specified here: http://www.w3.org/TR/microdata/#attr-itemref

[09:52:00.0000] <tantek>
it just says "a list of additional elements to crawl to find the name-value pairs of the item."

[09:53:00.0000] <tantek>
which in the example above, gives you infinitely nested items, each with a turtle property whose value is the next nested item with a turtle property etc.

[09:53:01.0000] <tantek>
turtles all the way down ;)

[09:53:02.0000] <tantek>
when do you stop crawling?

[09:53:03.0000] <LoneStar99>
Philip': i makes things look anti-aliasing

[09:54:00.0000] <TabAtkins>
tantek: http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html#associating-names-with-items

[09:54:01.0000] <TabAtkins>
Dunno if this is the latest version of the cycle-breaker.  You'd have to ask foolip, or get Hixie.

[09:55:00.0000] <tantek>
ah - ok: If root is in memory, then the algorithm fails; abort these steps.

[09:55:01.0000] <tantek>
therefore microdata effectively makes referencing an id of an ancestor a NOP

[09:56:00.0000] <TabAtkins>
If the ancestor is already part of the item, yes.

[09:56:01.0000] <TabAtkins>
If it's an ancestor in the DOM, higher up than the start of the item, though, then should be okay.

[09:56:02.0000] <tantek>
sure

[09:56:03.0000] <tantek>
microformats include-pattern has the same restriction

[09:57:00.0000] <tantek>
but I didn't see it in the W3C microdata draft

[09:57:01.0000] <tantek>
so I wondered if Ian had somehow permitted itemref inclusion of ancestors

[09:57:02.0000] <tantek>
apparently not

[09:57:03.0000] <TabAtkins>
Is that algo in the w3c draft, or are the two versions differeing significantly?

[09:58:00.0000] <LoneStar99>
Philip': Maybe not, the drawImage stuff does nothing from i can see

[09:59:00.0000] <tantek>
TabAtkins - the w3c draft says the same: http://www.w3.org/TR/microdata/#associating-names-with-items

[09:59:01.0000] <TabAtkins>
kk.

[09:59:02.0000] <tantek>
I just didn't find it because it doesn't reference itemref explicitly

[09:59:03.0000] <tantek>
itemref on an ancestor being a NOP is a side effect you have to figure out from the algo

[09:59:04.0000] <tantek>
rather than being explicit

[09:59:05.0000] <TabAtkins>
Yes.

[09:59:06.0000] <tantek>
which is going to confuse authors

[10:00:00.0000] <tantek>
when they ask why doesn't this work

[10:00:01.0000] <TabAtkins>
The simple and obvious answer would be "because that would make an infinite loop, don't do that".

[10:00:02.0000] <TabAtkins>
You don't need to point them at the algo for that.

[10:02:00.0000] <tantek>
much better to describe it up front IMHO

[10:02:01.0000] <tantek>
e.g.

[10:02:02.0000] <tantek>
http://microformats.org/wiki/include-pattern#in_general

[10:02:03.0000] <tantek>
To prevent infinite loops, if a class="include" refers to itself or to an ancestor in the parse tree, then it is ignored and has no effect on the parser.

[10:02:04.0000] <LoneStar99>
are there any webOS developers here?

[10:03:00.0000] <TabAtkins>
tantek: But the situation is more subtle with microdata, because it permits reference to ancestors; it merely prevents such a reference from causing an infinite loop.

[10:03:01.0000] <tantek>
LoneStar99 try #webkit

[10:04:00.0000] <tantek>
TabAtkins - not true - the ancestor is already in memory from the parse in progress

[10:04:01.0000] <tantek>
thus the reference to ancestor immediately fails in step 1

[10:04:02.0000] <tantek>
assuming you're parsing microdata from the root of document downward

[10:05:00.0000] <tantek>
which all the "converting to ..." (insert format here) algorithms do

[10:05:01.0000] <LoneStar99>
tantek: there is a webos channel, but wanted to see if someone could see the issues i see with my code... but anyways time for lunch! thanks for all the help be back later

[10:05:02.0000] <tantek>
LoneStar99 - good to know!

[10:08:00.0000] <tantek>
btw - given that itemref causes an actual crawl of the targets of the itemref it is not really a "ref" (like href), it is more like a "src" (like <img src>)

[10:08:01.0000] <tantek>
itemsrc would more accurately convey what itemref does

[10:08:02.0000] <TabAtkins>
tantek: I believe that that algo starts by assuming the root is an item, that is, it has @itemscope.

[10:09:00.0000] <tantek>
TabAtkins - sure - but that makes no difference to the breaking of recursion

[10:09:01.0000] <tantek>
or ancestral reference

[10:09:02.0000] <tantek>
alternatively, iteminclude would also better convey what itemref does

[10:10:00.0000] <TabAtkins>
<div id=foo itemprop=foo>foo <div itemscope itemref=foo></div></div> would affect it, wouldn't it?

[10:13:00.0000] <tantek>
TabAtkins - yes - not a particularly useful case though

[10:13:01.0000] <tantek>
as typically you get a need in nested objects to refer to (include) a common parent object

[10:13:02.0000] <TabAtkins>
No, but that's the reason why Ian often forgoes prose descriptions of algos - they don't hit all cases.

[10:14:00.0000] <tantek>
sure, but then that makes it harder for authors

[10:14:01.0000] <tantek>
authors don't typically "get" the procedural descriptions of the algos

[10:14:02.0000] <TabAtkins>
Indeed.

[10:15:00.0000] <tantek>
and one of the claims of microdata is that it is "easier" for authors

[10:15:01.0000] <tantek>
thus I would say that claim is refuted by this

[10:15:02.0000] <tantek>
TabAtkins - if you want to see a discussion of a real world example of nested objects referring to a parent object, /join #microformats

[10:16:00.0000] <tantek>
we're discussing hProduct + hReview there now

[10:16:01.0000] <TabAtkins>
Oh, I don't doubt the use-case.  I use circularly-nested data structures all the time.

[10:16:02.0000] <tantek>
all the time?

[10:16:03.0000] <tantek>
or are you being facetious ;)

[10:16:04.0000] <TabAtkins>
No, I really do use them a lot.

[10:17:00.0000] <TabAtkins>
It's very useful to embed a reference to the parent in a child.

[10:18:00.0000] <tantek>
a reference yes. but an include no. and itemref is an include, not a reference. hence my point about terminology.

[10:18:01.0000] <TabAtkins>
That's valid.

[10:19:00.0000] <TabAtkins>
Then suggest some accurate prose, or suggest a change to the algo to reflect the simpler prose.

[10:19:01.0000] <TabAtkins>
Such as disallowing all ancestor references.

[10:19:02.0000] <TabAtkins>
/me goes to lunch.

[10:20:00.0000] <JonathanNeal_oww>
Hey all.

[10:20:01.0000] <JonathanNeal>
I'm trying to come up with practical naming conventions for the IDs and Classnames in my layout template.

[10:21:00.0000] <JonathanNeal>
I was trying to follow the ARIA roles list to name most of my elements.

[10:58:00.0000] <Philip`>
JonathanNeal: I suggest using boys' names for classes and girls' names for IDs

[10:58:01.0000] <JonathanNeal>
Ha.

[10:58:02.0000] <Philip`>
Helps to keep them easily distinguishable

[10:58:03.0000] <JonathanNeal>
I found an ARIA name.

[10:58:04.0000] <JonathanNeal>
role="main" so I named it main.

[11:07:00.0000] <Hixie>
tantek: please send microdata feedback to the list or use the review comments tool in the whatwg version to file a bug so i don't lose it -- thanks

[11:18:00.0000] <mpilgrim>
I don't really understand the purpose of the "fallback" section in a cache manifest file

[11:19:00.0000] <mpilgrim>
i see the use case for the "explicit" section (stuff you need offline)

[11:19:01.0000] <mpilgrim>
and i see the use case for the "online whitelist" section (stuff like tracking CGI scripts that don't make sense offline)

[11:20:00.0000] <mpilgrim>
there's only one example in the spec of using a "fallback" section, and i don't fully understand it

[11:20:01.0000] <mpilgrim>
scratch that, i don't understand it at all

[11:22:00.0000] <Hixie>
imagine a site like flickr

[11:22:01.0000] <Hixie>
which has one page per image

[11:23:00.0000] <Hixie>
now imagine the user, while offline, adds a bunch of images

[11:23:01.0000] <Hixie>
and follows links to those images

[11:23:02.0000] <Hixie>
if we didn't have the fallback section, those links would be dead

[11:24:00.0000] <mpilgrim>
...

[11:25:00.0000] <mpilgrim>
ok, so offline flickr

[11:26:00.0000] <mpilgrim>
suppose http://www.flickr.com/photos/f8dy/ links to a manifest file

[11:27:00.0000] <Hixie>
crap meeting gotta go

[11:28:00.0000] <mpilgrim>
http://www.flickr.com/photos/f8dy/327774195/ (and, say, 99 other recent photo pages) is listed in the explicit section

[11:28:01.0000] <mpilgrim>
along with http://farm1.static.flickr.com/142/327774195_2d2b67753e.jpg which is the image on that page

[11:31:00.0000] <tantek>
Hixie, will do. For reference, my feedback items are described here: http://krijnhoetmer.nl/irc-logs/whatwg/20100305#l-781 I'll clean up that text and send it to public-html.

[11:32:00.0000] <tantek>
/me was trying to figure out what about that photo of towels over mpilgrim's feed made it "explicit".

[11:32:01.0000] <mpilgrim>
haha

[11:33:00.0000] <mpilgrim>
pooh porn

[11:33:01.0000] <mpilgrim>
rule 34 predicts its existence

[11:36:00.0000] <mpilgrim>
a fun fact about the cache manifest: if there are no explicit section headers, all the listed resources are implicitly in the "explicit" section

[11:37:00.0000] <mpilgrim>
i tried to explain this with a straight face for my book, but i failed

[12:08:00.0000] <knowtheory>
Hey othermaciej :)

[12:08:01.0000] <othermaciej>
hello

[12:11:00.0000] <knowtheory>
are CSS animations incorporated into a component of HTML5 or into a new CSS standard?

[12:11:01.0000] <knowtheory>
I'm trying to figure out the degree of complexity and organizational capaobilities of CSS animations (basically, how one would work w/ 'em for complex examples)

[12:24:00.0000] <TabAtkins>
knowtheory: CSS Animations are (unsurprisingly ^_^) a CSS Module.

[12:24:01.0000] <knowtheory>
TabAtkins: in HTML5 then? :D

[12:24:02.0000] <TabAtkins>
Haha.

[12:25:00.0000] <knowtheory>
hunh, exciting, http://www.w3.org/TR/css3-ui/ just crashed a bunch of tabs in chrome

[12:26:00.0000] <knowtheory>
yeah freaky, consistent crash on that page.

[12:26:01.0000] <knowtheory>
safari renders it fine though

[12:26:02.0000] <TabAtkins>
Huh, no crash for me.  Windows, Chrome 4.

[12:27:00.0000] <knowtheory>
yeah osx 5.0.307.11 b

[12:27:01.0000] <miketaylr>
works fine in 5.0.345.0 (40651)/Mac

[12:27:02.0000] <miketaylr>
(well, chromium)

[12:27:03.0000] <knowtheory>
even more exciting! :)

[12:37:00.0000] <knowtheory>
ah there we go: http://dev.w3.org/csswg/css3-animations/

[12:42:00.0000] <TabAtkins>
knowtheory: Ah man, I'm sorry, I would have linked you to it immediately if I knew you were looking for it.

[12:43:00.0000] <knowtheory>
TabAtkins: no worries!

[12:59:00.0000] <tantek>
I did encode actual example uses of CSS3 UI properties in the spec itself - I wonder if that may be causing the crash - certainly was not intentional.

[13:05:00.0000] <knowtheory>
tantek: it's all your fault!

[13:05:01.0000] <knowtheory>
since someone mentioned that chromium wasn't seeing that behavior

[13:05:02.0000] <knowtheory>
i figure it's just a bug in this version of chrome

[13:07:00.0000] <miketaylr>
knowtheory: heh, i think you're right: http://miketaylr.com/post/bcbf19c8.png

[13:07:01.0000] <miketaylr>
i crash using the same version of chrome, but am fine in chromium

[13:08:00.0000] <knowtheory>
miketaylr: i filed a bug on the chromium tracker :)

[13:09:00.0000] <miketaylr>
cool

[13:11:00.0000] <knowtheory>
miketaylr: http://code.google.com/p/chromium/issues/detail?id=37546

[13:11:01.0000] <knowtheory>
(also for you tantek if you are interested)

[13:16:00.0000] <knowtheory>
man, that was fast, the bug i reported has already been triaged and assigned

[13:25:00.0000] <dglazkov>
knowtheory: speed is what we do :)

[13:25:01.0000] <dglazkov>
that didn't come out right

[13:26:00.0000] <knowtheory>
haha :D

[13:27:00.0000] <mpilgrim>
woohoo, i have a working offline web application

[13:28:00.0000] <mpilgrim>
whose sole purpose is to display information on whether or not it is a working offline web application

[13:30:00.0000] <mpilgrim>
it works in 2 browsers!

[13:31:00.0000] <mpilgrim>
my work here is done

[13:38:00.0000] <othermaciej>
hawt

[13:38:01.0000] <othermaciej>
annevk: does html5-diffs have a bug component?

[13:52:00.0000] <mpilgrim>
no, scratch that

[13:52:01.0000] <mpilgrim>
i can't get it to register updates in firefox

[13:53:00.0000] <mpilgrim>
http://diveintohtml5.org/examples/clock.html

[13:53:01.0000] <mpilgrim>
manifest is at http://diveintohtml5.org/examples/clock.html5cache

[13:53:02.0000] <Hixie>
why "html5"cache?

[13:53:03.0000] <mpilgrim>
firefox properly downloads it and can use it offline

[13:54:00.0000] <mpilgrim>
but when i update the JS file with new code and update the manifest file with a version-specific comment, it never picks up the updated resources

[13:54:01.0000] <Hixie>
does it try to fetch the manifest when you load the page?

[13:54:02.0000] <mpilgrim>
curl -I confirms that the etag and last-modified headers of the manifest file are changing

[13:55:00.0000] <mpilgrim>
oh bloody hell, my expires headers are wrong

[13:55:01.0000] <othermaciej>
what's the expiration on the manifest?

[13:55:02.0000] <mpilgrim>
yeah, just got there

[13:55:03.0000] <mpilgrim>
hang on

[14:07:00.0000] <temoto>
Hello. I want to try html5lib instead of BeautifulSoup to parse arbitrary HTML found on the web. What i'm [currently] interested is only about finding <a> elements and taking their href attribute value. (as opposed to keeping Soup interface) So which is the fastest way to iterate all <a> elements?

[14:14:00.0000] <mpilgrim>
woohoo, http://diveintohtml5.org/examples/offline/clock.html works in 2 browsers!

[14:14:01.0000] <mpilgrim>
hixie: i don't know why i called it ".html5cache"

[14:15:00.0000] <mpilgrim>
i needed a unique file extension to set the content-type

[14:15:01.0000] <Hixie>
and you didn't like .manifest like the spec suggests? :-P

[14:16:00.0000] <mpilgrim>
i missed that

[14:16:01.0000] <mpilgrim>
that doesn't conflict with anything else in MIME-world?

[14:16:02.0000] <Hixie>
not to my knowledge

[14:17:00.0000] <mpilgrim>
("MIME-world" reminds me of the underwater minus level in Super Mario Bros. -- hard to get into and impossible to win)

[14:24:00.0000] <mpilgrim>
while we're on the subject, do you have a suggestion for a MIME type for .ttf files?

[14:24:01.0000] <Hixie>
font/truetype?

[14:24:02.0000] <Hixie>
might need registering

[14:24:03.0000] <mpilgrim>
it's one of those MIME-world trick questions, i think

[14:25:00.0000] <mpilgrim>
AFAICT, there's nothing registered

[14:25:01.0000] <mpilgrim>
and all the experimental or vendor-specific alternatives are horrific

[14:25:02.0000] <Hixie>
ah. then you just volunteered to register it!

[14:25:03.0000] <mpilgrim>
apparently it doesn't matter because font-face-supporting browsers ignore the mime type

[14:26:00.0000] <Hixie>
yeah well

[14:26:01.0000] <mpilgrim>
but i got email complaints because my fonts were being served as text/plain, which is arguably the worst default ever

[14:26:02.0000] <Hixie>
/me wishes we could abandon MIME types for anything that can be unambiguously identified by the leading bytes

[14:26:03.0000] <mpilgrim>
and judging by the online discussions i've found on the topic, registering it is impossible

[14:26:04.0000] <Hixie>
why is registering it impossible?

[14:27:00.0000] <mpilgrim>
maybe not theoretically impossible

[14:27:01.0000] <mpilgrim>
but no one has managed it yet

[14:27:02.0000] <mpilgrim>
despite years of trying

[14:27:03.0000] <mpilgrim>
and no, i'm not volunteering to register it

[14:28:00.0000] <mpilgrim>
i don't need any more email from Those People

[14:30:00.0000] <Hixie>
<img alt="Stamp It" onerror="this.src='/icon.gif';" src="" align="left" border="0" height="32" width="32">

[14:30:01.0000] <Hixie>
wtf

[14:30:02.0000] <mpilgrim>
hooray for alt text!

[14:30:03.0000] <mpilgrim>
does that ever work?

[14:30:04.0000] <mpilgrim>
(the onerror part, i mean)

[14:33:00.0000] <Hixie>
apparently

[14:33:01.0000] <Hixie>
src="" triggers onerror, which sets the icon, or something

[14:33:02.0000] <mpilgrim>
awesome

[14:34:00.0000] <mpilgrim>
so you were explaining the fallback section of the manifest

[14:34:01.0000] <mpilgrim>
using flickr as an example

[14:34:02.0000] <Hixie>
yes

[14:34:03.0000] <mpilgrim>
i don't understand it yet

[14:35:00.0000] <Hixie>
it lets you cache a small part of a site and have graceful fallback for the rest of the site while offline

[14:37:00.0000] <mpilgrim>
and by "graceful fallback" you mean "a page that says this page is not available offline"

[14:37:01.0000] <mpilgrim>
in lieu of a browser-level error message

[14:37:02.0000] <mpilgrim>
?

[14:37:03.0000] <Hixie>
not necessarily; if you have the information locally to render the page, e.g. if bugzilla has a copy of the bug in its client-side database, you can just render the page locally

[14:38:00.0000] <temoto>
Hello. I want to try html5lib instead of BeautifulSoup to parse arbitrary HTML found on the web. What i'm [currently] interested is only about finding <a> elements and taking their href attribute value. (as opposed to keeping Soup interface) So which is the fastest way to iterate all <a> elements?

[14:39:00.0000] <mpilgrim>
that makes sense for bugzilla because each page is "show_bug.cgi?id=foo" and the only thing that changes is foo

[14:39:01.0000] <mpilgrim>
earlier you mentioned a hypothetical offline flickr

[14:39:02.0000] <mpilgrim>
flickr uses unique URLs for each photo page

[14:41:00.0000] <mpilgrim>
are you saying that you could set up a wildcard so that all unknown (uncached) URLs got handled by a master page like show_photo.cgi?id=foo

[14:41:01.0000] <Hixie>
i just meant an image site in general, dunno abotu flickr specifically

[14:41:02.0000] <mpilgrim>
?

[14:41:03.0000] <Hixie>
that's what the fallback stuff does, yes

[14:42:00.0000] <mpilgrim>
i see

[14:43:00.0000] <mpilgrim>
that was the missing link

[14:43:01.0000] <mpilgrim>
i didn't know that fallback URLs could be wildcards

[14:43:02.0000] <mpilgrim>
now i see the bit about "fallback namespace" and "URLs as prefix match patterns"

[14:43:03.0000] <Hixie>
they work the same as the online whitelist, but with a fallback url

[14:44:00.0000] <mpilgrim>
can i declare a specific URL in the explicit section that would also match a fallback namespace?

[14:44:01.0000] <mpilgrim>
like

[14:44:02.0000] <mpilgrim>
CACHE MANIFEST

[14:44:03.0000] <mpilgrim>
CACHE:

[14:44:04.0000] <mpilgrim>
images/cat.jpg

[14:45:00.0000] <mpilgrim>
FALLBACK:

[14:45:01.0000] <mpilgrim>
images/ missing.jpg

[14:45:02.0000] <Hixie>
what do you mean by "can i" exactly?

[14:46:00.0000] <mpilgrim>
will the browser cache images/cat.jpg and use it when i say <img src="images/cat.jpg"> ?

[14:46:01.0000] <mpilgrim>
the URL matches a resource listed in the explicit section, and also a fallback namespace listed in the fallback section

[14:46:02.0000] <mpilgrim>
which one wins?

[14:46:03.0000] <Hixie>
yes

[14:46:04.0000] <Hixie>
If a resource is listed in the explicit section or as a fallback entry in the fallback section, the resource will always be taken from the cache, regardless of any other matching entries in the fallback namespaces or online whitelist namespaces.

[14:46:05.0000] <mpilgrim>
ok

[14:48:00.0000] <mpilgrim>
now here's another question

[14:48:01.0000] <mpilgrim>
i've been trying to set up all of diveintohtml5.org as an offline web app

[14:48:02.0000] <mpilgrim>
but i don't want to do it by default

[14:49:00.0000] <mpilgrim>
i want you to click a link "go offline" and go to a page that gives a progress bar

[14:49:01.0000] <mpilgrim>
(in firefox this page would also display the prompt "do you want this site to store local data" or some such)

[14:49:02.0000] <mpilgrim>
(in safari there is apparently no prompt)

[14:49:03.0000] <mpilgrim>
IOW, i don't want every page to point to the manifest file

[14:50:00.0000] <Hixie>
if a page doesn't point to the manifest, it can't be cached for offline use

[14:50:01.0000] <mpilgrim>
but i want the go_offline.html page to point to ta manifest file that includes all the other HTML pages (and their scripts, stylesheets, and images)

[14:50:02.0000] <Hixie>
each page that is to work offline has to opt in to working offlien

[14:51:00.0000] <mpilgrim>
so page A *must* point to a manifest file in order to be used offline?

[14:51:01.0000] <mpilgrim>
even if page B points to a manifest file that includes page A?

[14:52:00.0000] <Hixie>
yes

[14:52:01.0000] <mpilgrim>
even after the user visits page B and consents to storing local data?

[14:52:02.0000] <Hixie>
yes

[14:52:03.0000] <mpilgrim>
that sucks

[14:53:00.0000] <mpilgrim>
i don't want every page of my site popping up a scary infobar about "storing local data"

[14:53:01.0000] <Hixie>
your site isn't an application

[14:53:02.0000] <mpilgrim>
sure it is

[14:53:03.0000] <mpilgrim>
it's an interactive ebook

[14:53:04.0000] <mpilgrim>
it has interactive elements

[14:54:00.0000] <mpilgrim>
and scripts that give live results based on the current browser environment

[14:54:01.0000] <LoneStar99>
is Philip` around?

[14:54:02.0000] <Philip`>
Maybe

[14:54:03.0000] <Hixie>
mpilgrim: then just stick a manifest on each page

[14:54:04.0000] <Hixie>
mpilgrim: and let the user decide whether to go offline or not

[14:55:00.0000] <LoneStar99>
Philip` have a video with the issue at hand http://video.yahoo.com/watch/7098996/18475648

[14:55:01.0000] <mpilgrim>
that user experience sucks

[14:55:02.0000] <Philip`>
/me can't watch videos currently

[14:55:03.0000] <Hixie>
the user experience is in the hands of the user agents, so get them to make it better

[14:56:00.0000] <LoneStar99>
Philip` ok will ping you later 

[14:56:01.0000] <mpilgrim>
that's a shitty cop-out of an answer

[14:56:02.0000] <mpilgrim>
this was possible in gears, was it not?

[14:56:03.0000] <mpilgrim>
because everything was scripted, including the downloading of the manifest file, etc.

[14:56:04.0000] <Hixie>
no idea

[14:57:00.0000] <Hixie>
letting the user agent be in charge of the user experience is not a cop-out answer, it's the way the web is designed to work

[14:58:00.0000] <LoneStar99>
if anyone has exprience with canvas can you look at the following video and tell me what is going on: http://video.yahoo.com/watch/7098996/18475648

[14:58:01.0000] <mpilgrim>
in html5, you can't script the downloading of the manifest file because you can't dynamically set the manifest attribute

[14:59:00.0000] <Hixie>
correct

[14:59:01.0000] <Hixie>
a page is either part of an application or it isn't

[14:59:02.0000] <mpilgrim>
i want a page to decide at runtime whether it is part of an application or it isn't

[15:00:00.0000] <Hixie>
can't be done

[15:00:01.0000] <Hixie>
well, you could have the server dynamically decide whether to set manifest="" or not, i guess

[15:00:02.0000] <Hixie>
but you'd have to reload the page or load it in an iframe or something to trigger teh load

[15:00:03.0000] <mpilgrim>
based on what? HTTP request headers?

[15:01:00.0000] <Hixie>
cookies, probably

[15:01:01.0000] <mpilgrim>
i suppose that would work

[15:01:02.0000] <Hixie>
it's not the intended use case though

[15:01:03.0000] <Hixie>
the intended use case is applications that are unambiguously applications

[15:02:00.0000] <mpilgrim>
MY SITE IS UNAMBIGUOUSLY AN APPLICATION

[15:02:01.0000] <mpilgrim>
it is ambiguously an offline application

[15:02:02.0000] <Hixie>
your site is apparently insecure about its application status, if it's not willing to always be an offline application :-)

[15:03:00.0000] <mpilgrim>
how are applications like gmail supposed to migrate to html5 if they can't offer an online application and an offline application at the same URL?

[15:03:01.0000] <Hixie>
what's the difference?

[15:04:00.0000] <Hixie>
between an "online" app and an "offline" app, i mean

[15:05:00.0000] <mpilgrim>
i navigate to http://mail.google.com/ in firefox

[15:05:01.0000] <mpilgrim>
i log in

[15:05:02.0000] <mpilgrim>
i click "settings"

[15:05:03.0000] <mpilgrim>
i click "offline"

[15:06:00.0000] <Hixie>
why?

[15:06:01.0000] <mpilgrim>
i click "enable offline access"

[15:06:02.0000] <Hixie>
here's how it should work:

[15:06:03.0000] <Hixie>
you navigate to http://mail.google.com/ in firefox

[15:06:04.0000] <Hixie>
you log in

[15:06:05.0000] <Hixie>
end of story

[15:06:06.0000] <mpilgrim>
it pops up a pseudo-dialog titled "install offline access for gmail"

[15:07:00.0000] <mpilgrim>
because not every user who logs into gmail wants, needs, or is operating in a safe browser environment to have offline access

[15:07:01.0000] <mpilgrim>
the pseudo-dialog i am reading right now says "This feature will download your email messages onto this computer. Please make sure it is not a public or shared computer."

[15:08:00.0000] <Hixie>
sure, actually downloading the mail might be something you want to do after checking a pref

[15:08:01.0000] <Hixie>
but that's just making the mail available offline, that's not making the _app_ available offline

[15:09:00.0000] <Hixie>
the data part has nothing to do with the cache manifest

[15:09:01.0000] <Hixie>
you can download that without making an "offline" app

[15:09:02.0000] <mpilgrim>
so your ideal user experience is signing up for gmail, logging in for the first time, and getting an infobar saying "this site is asking to store local data [allow] [deny] [learn more]"

[15:09:03.0000] <Hixie>
no

[15:09:04.0000] <Hixie>
there's no reason for the user agent to ask the user about caches

[15:09:05.0000] <Hixie>
it's basically no different than an HTTP cache

[15:10:00.0000] <mpilgrim>
but firefox does, demonstrably, ask the user for permission before storing any data in the applicationCache

[15:10:01.0000] <Hixie>
file a bug

[15:10:02.0000] <Hixie>
or use a different browser

[15:10:03.0000] <mpilgrim>
i'm using firefox 3.6 on windows

[15:10:04.0000] <mpilgrim>
though i assume firefox on other platforms offers an identically bad experience

[15:10:05.0000] <Hixie>
browsers can show whatever prompts they want if they want to make their UI less than ideal

[15:11:00.0000] <Hixie>
it's a tradeoff they have to make between convenience and giving the appearance of... something? not sure what the dialog helps with in this case, privacy? security? dunno.

[15:13:00.0000] <mpilgrim>
presumably the site could use it as a super-tracking cookie? i don't know how useful it would be since it wouldn't be accessible cross-domain

[15:13:01.0000] <Hixie>
it can set cookies without prompting, too

[15:13:02.0000] <mpilgrim>
indeed

[15:14:00.0000] <Hixie>
i agree that if you enable the cookie dialog, then it makes sense to prompt for appcache, web storage, web db, etc

[15:14:01.0000] <Hixie>
probably should also prompt with every link click, since the IP address, User-Agent string, and other parts of the HTTP message can more or less uniquely identify users even without cookies

[15:17:00.0000] <mpilgrim>
so, assuming i wanted to serve different static HTML files based on cookies

[15:17:01.0000] <mpilgrim>
i'd need a CGI script that set the cookie

[15:17:02.0000] <Hixie>
or javascript

[15:17:03.0000] <mpilgrim>
ok

[15:17:04.0000] <mpilgrim>
expires/caching headers set to 0 so HTML pages would always refresh from the server

[15:18:00.0000] <mpilgrim>
a Vary: Cookie header

[15:18:01.0000] <mpilgrim>
two copies of each .html file (one with the manifest attribute, one without)

[15:18:02.0000] <mpilgrim>
and some mod_rewrite magic that served the appropriate .html file based on the cookie in the HTTP request headers

[15:19:00.0000] <Hixie>
or just put the manifest on every file and make the manifest itself be the only one with magic

[15:19:01.0000] <mpilgrim>
/me is suddenly overwhelmed with the memory of the classic "and then a miracle occurs -->" tshirt

[15:19:02.0000] <mpilgrim>
no, i'm pretty sure the presence of the manifest attribute is what triggers firefox's infobar

[15:20:00.0000] <Hixie>
oh, right, if you want to work around firefox's ui issues you'd probably have to do that, yes

[15:20:01.0000] <mpilgrim>
EVERYONE WILL WANT TO WORK AROUND FIREFOX'S UI ISSUES

[15:21:00.0000] <Hixie>
if everyone hates firefox's ui, i imagine it'll change

[15:21:01.0000] <mpilgrim>
I AM JUST THE FIRST PERSON WHO HAS EVER TRIED

[15:21:02.0000] <Hixie>
i would recommend filing a bug with them explaining that the prompt doesn't really do anything helpful rather than yelling here :-)

[15:21:03.0000] <mpilgrim>
mobile gmail already does UA sniffing and only presents the offline-capable version to supported browsers (based on HTTP_USER_AGENT, i assume)

[15:22:00.0000] <mpilgrim>
and everyone else is still using gears

[15:22:01.0000] <mpilgrim>
(where "everyone else" includes zoho, rememberthemilk, and wordpress)

[15:23:00.0000] <mpilgrim>
(and the "supported browsers" that mobile gmail sniffs are all based on webkit and don't present any stupid UI prompts)

[15:23:01.0000] <mpilgrim>
(so it is quite likely that this problem has literally never come up)

[15:23:02.0000] <annevk>
othermaciej, I just use my inbox

[15:24:00.0000] <annevk>
othermaciej, so no

[15:26:00.0000] <mpilgrim>
i suppose my firefox bug report would be stronger if I blindly offline-enabled every page of my site and then included copies of the hate mail i expect to get from people who think my site is taking over their computer

[15:28:00.0000] <annevk>
this peer-to-peer stuff is interesting

[15:29:00.0000] <annevk>
are we taking on IM? :)

[15:29:01.0000] <Hixie>
annevk: dunno, i just threw IM in there because i noticed most if not all video conferencing systems also do peer-to-peer IM during the connection

[15:31:00.0000] <annevk>
so the way this API works is that you negotiate a unique identifier via a third-party server to establish a peer-to-peer connection?

[15:32:00.0000] <Hixie>
no third-party server in theory

[15:33:00.0000] <Hixie>
it depends on what protocol we end up building this on, but getAddress() e.g. could just return an IP address and port

[15:33:01.0000] <Hixie>
(of the user's machine)

[15:33:02.0000] <annevk>
but how do you give that to a friend?

[15:34:00.0000] <Hixie>
oh well i assume you're talking to them somehow already, e.g. you have a websocket connection to the server of the page the script is from

[15:34:01.0000] <Hixie>
and they do too

[15:34:02.0000] <Hixie>
e.g. for chatting in-game

[15:34:03.0000] <Hixie>
you would both be connected to the game server

[15:34:04.0000] <annevk>
right, so you have a third-party server

[15:34:05.0000] <Hixie>
well it's more a second-party server, but yeah

[15:35:00.0000] <Hixie>
(user, server, other player -- the other player is the third party)

[15:35:01.0000] <Hixie>
you could also just sent it by e-mail or something :-)

[15:35:02.0000] <annevk>
sure or you could phone the guy

[15:35:03.0000] <Hixie>
yeah

[15:36:00.0000] <Hixie>
but generally i don't think these address strings will be pretty

[15:36:01.0000] <annevk>
but that's not really going to happen I think :)

[15:36:02.0000] <annevk>
right

[15:36:03.0000] <Hixie>
you could imagine if we do NAT traversal that the string will include all kinds of stuff for how to poke holes in the firewalls

[15:36:04.0000] <Hixie>
but again, that depends on whatever we decide to build this on

[15:38:00.0000] <Hixie>
annevk: btw would be good to get opera's feedback on http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-February/025280.html

[15:38:01.0000] <annevk>
i didn't quite get it

[15:39:00.0000] <Hixie>
hsivonen: your feedback would likely also be useful (on http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-February/025280.html )

[15:39:01.0000] <Hixie>
annevk: that's probably good feedback to send :-)

[15:40:00.0000] <annevk>
btw, do I misunderstand the subprotocol idea or do the people on hybi misunderstand it?

[15:40:01.0000] <annevk>
maybe both :)

[15:40:02.0000] <Hixie>
there's certainly been a lot of confusion on that issue

[15:40:03.0000] <Hixie>
not sure why

[15:41:00.0000] <Hixie>
i don't remember seeing anything from you about it so no idea if you understand it or not :-)

[15:43:00.0000] <franksalim>
annevk, how do you understand it?

[15:45:00.0000] <annevk>
I don't

[15:45:01.0000] <annevk>
e.g. "I would like the toString method to be called with 'html '"

[15:46:00.0000] <annevk>
but toString() has no arguments in this context

[15:46:01.0000] <annevk>
afaik

[15:46:02.0000] <franksalim>
where is that?

[15:46:03.0000] <annevk>
oh, are you talking about the subprotocol thingie?

[15:47:00.0000] <franksalim>
oh yes, i meant subprotocol. i see you were talking about http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-February/025280.html

[15:49:00.0000] <annevk>
well, I think the idea is that the subprotocol indicates how the data transmitted in the frame types is supposed to be processed, but I don't think the idea is that you can simply use all the frame types if you have a custom subprotocol

[15:50:00.0000] <franksalim>
I think that is correct

[15:51:00.0000] <franksalim>
Ports usually indicate the protocol, but because ws:// does not use host/port addressing, 'subprotocol' indicates the protocol instead

[15:52:00.0000] <franksalim>
Either new protocols or existing TCP protocols bound to ws

[15:52:01.0000] <annevk>
i.e. I am pretty certain we end up having distinct frame types for ByteArray and Stream and that a subprotocol cannot change the meaning of that

[15:53:00.0000] <franksalim>
annevk, so you're pretty sure we need frame types and not just a bidirectional byte stream like TCP?

[15:54:00.0000] <annevk>
yeah, exposing TCP to JavaScript directly won't fly

[15:55:00.0000] <franksalim>
for technical reasons or other reasons?

[15:58:00.0000] <annevk>
security, mostly

[15:58:01.0000] <franksalim>
I would be perfectly happy with TCP in JavaScript after the WebSocket handshake

[15:58:02.0000] <annevk>
oh I see

[15:58:03.0000] <annevk>
interesting thought

[15:59:00.0000] <annevk>
the reason for that is that JavaScript prolly works better being event-based

[15:59:01.0000] <annevk>
but I'm on shaky grounds here

[15:59:02.0000] <annevk>
Hixie would know


2010-03-06
[16:02:00.0000] <Hixie>
yeah doing stream-based programming from JS is just exposing way too much complexity to most authors

[16:02:01.0000] <franksalim>
annevk, ok. fwiw, JavaScript events and TCP work fine together

[16:02:02.0000] <Hixie>
almost all protocols end up being framed anyway

[16:03:00.0000] <franksalim>
Hixie, it is true that most protocols end up being framed, but now each framed protocol needs to have a WebSocket binding defined

[16:03:01.0000] <Hixie>
annevk: btw i think the main source of argument on this topic is that many of the people on hybi for some reason want to use websockets for something other than browser-to-server communication

[16:03:02.0000] <Hixie>
franksalim: why?

[16:04:00.0000] <franksalim>
Hixie, how do you later XMPP over WebSocket?

[16:05:00.0000] <franksalim>
Hixie, align each XMPP stanza to a ws frame, or ignore ws frames and just send XMPP as strings?

[16:05:01.0000] <Hixie>
why would you layer XMPP over WebSocket?

[16:05:02.0000] <Hixie>
put your XMPP client on the server-side

[16:06:00.0000] <Hixie>
and then just speak an internal protocol to get the data the "last mile" from your server to the browser

[16:06:01.0000] <franksalim>
the internal protocol might as well be XMPP

[16:06:02.0000] <Hixie>
putting all the XMPP logic in the client seems like a lot of work for no reason

[16:07:00.0000] <franksalim>
it is the same amount of work as writing a new protocol that covers all of the XMPP semantics you want

[16:08:00.0000] <Hixie>
you hardly have to implement anything at all if you put the XMPP client on the WebSocket server

[16:08:01.0000] <franksalim>
and because XMPP is well known and widely interoperable, you have the option of connecting to other servers that are outside of your direct control

[16:08:02.0000] <Hixie>
wait you want your JS page to connect to arbitrary XMPP servers?!

[16:08:03.0000] <franksalim>
yes

[16:08:04.0000] <Hixie>
i suppose if you really wanted to you could do that

[16:08:05.0000] <Hixie>
dunno why you would though

[16:08:06.0000] <franksalim>
but for a moment let me pretend i am only concerned with first party servers

[16:08:07.0000] <franksalim>
and am willing to write my own protocol to go the last mile

[16:09:00.0000] <franksalim>
the complexity of the custom protocol is necessarily the same as xmpp

[16:09:01.0000] <Hixie>
no

[16:09:02.0000] <franksalim>
except, possibly, in the wire representation

[16:10:00.0000] <Hixie>
it's the same complexity as the protocol an XMPP client would have with the OS's graphics subsystem, maybe

[16:10:01.0000] <Hixie>
"show these strings as being online"

[16:10:02.0000] <Hixie>
"display this message"

[16:10:03.0000] <annevk>
Hixie, yeah, I get the idea from some emails they have started already

[16:12:00.0000] <franksalim>
Hixie, that is one approach

[16:13:00.0000] <Hixie>
do people speak XMPP over hanging GETs today?

[16:13:01.0000] <franksalim>
yes

[16:13:02.0000] <Hixie>
url?

[16:13:03.0000] <franksalim>
http://xmpp.org/extensions/xep-0124.html

[16:13:04.0000] <Hixie>
i'd love to see that

[16:14:00.0000] <Hixie>
what pages use that?

[16:14:01.0000] <franksalim>
chesspark, collecta

[16:14:02.0000] <franksalim>
i don't have a list. those are two i have used

[16:14:03.0000] <franksalim>
http://code.stanziq.com/strophe/

[16:15:00.0000] <franksalim>
also, Kaazing has an XMPP library that uses the WebSocket API. It implements the XMPP protocol over ws://

[16:15:01.0000] <Hixie>
man, and here i was thinking people would finally have an excuse to dump XMPP and use something saner

[16:15:02.0000] <Hixie>
oh well

[16:15:03.0000] <franksalim>
that might happen too :-)

[16:16:00.0000] <franksalim>
but nobody has rallied around a chat protocol with a JSON representation, for instance

[16:16:01.0000] <franksalim>
XMPP is the best we currently have

[16:20:00.0000] <Hixie>
anyway, to answer your earlier question, you would just send each stanza with its own WebSocket.send() call

[16:20:01.0000] <Hixie>
and would receive each stanza in a separate onmessage call

[16:20:02.0000] <franksalim>
that seems natural, to align each frame in the existing protocol to a websocket frame

[16:20:03.0000] <Hixie>
not sure how you'd deal with the wacky XML parsing rules

[16:20:04.0000] <franksalim>
but these bindings will need to be defined

[16:21:00.0000] <Hixie>
sure

[16:21:01.0000] <Hixie>
if people want to communicate across multiple servers

[16:21:02.0000] <Hixie>
just like HTTP and XMPP are defined above TCP

[16:21:03.0000] <Hixie>
you can define protocols on top of WebSocket

[16:21:04.0000] <Hixie>
i expect most to be proprietary though

[16:22:00.0000] <franksalim>
Hixie, I expect most (of the important ones) to be based on existing standard protocols

[16:22:01.0000] <Hixie>
i expect most of them not to be important :-)

[16:23:00.0000] <franksalim>
maybe that's too loaded a word

[16:23:01.0000] <franksalim>
interoperable clients and servers

[16:24:00.0000] <Hixie>
i think most uses of websocket will be for a dedicated server speaking to a dedicated page from that server

[16:24:01.0000] <franksalim>
Hixie, what do you think about TCP relays over WebSocket? A few of those have already popped up

[16:25:00.0000] <Hixie>
e.g. reporting analytics, state in multiplayer games, sending back and forth ad information

[16:25:01.0000] <Hixie>
s/from that server/from that server or using a script from that server/, i should say

[16:25:02.0000] <Hixie>
much like today XHR is used with XML and JSON files that are proprietary formats

[16:26:00.0000] <franksalim>
Hixie, since ws is cross-origin from day 1, it is too appealing for cross-site use to be relegated to that use case

[16:26:01.0000] <franksalim>
Hixie, so what about public AJAX services? "APIs"

[16:27:00.0000] <othermaciej>
annevk: we should probably make one for consistency with other HTML WG drafts, though there is probably not much need to file bugs

[16:28:00.0000] <annevk>
if you want it's not a problem

[16:29:00.0000] <othermaciej>
I'll ask MikeSmith to add one

[16:30:00.0000] <Hixie>
franksalim: yeah, those are almost all proprietary

[17:57:00.0000] <MikeSmith>
http://code.google.com/p/web-messaging-rpc/

[18:19:00.0000] <franksalim>
MikeSmith, I like the labels for that project: html5, messaging, rpc, worker, webworkers, xdm, websockets, server-sent-events

[18:20:00.0000] <MikeSmith>
yeah

[00:23:00.0000] <hsivonen>
my knee jerk reaction to making document.write more magic is negative, but I need to think about it

[01:24:00.0000] <Lachy>
can anyone in here make any sense out of these examples? http://www.w3.org/TR/2010/WD-html-bidi-20100304/#basedirection

[01:25:00.0000] <Lachy>
Especially the 2nd and 3rd examples, which seem to randomly have some words with their letters reversed, and seemingly random word order for no apparent reason

[01:25:01.0000] <Lachy>
well, the 3rd one has all the letters reveresed and word order reversed, expect for "html" which has its letters in the usual order.

[01:26:00.0000] <Lachy>
but I also don't get how the second one can claim to be displayed LTR, and not just use normal english word and letter ordering.

[01:27:00.0000] <annevk>
I think there is a convention that uppercase means characters with RTL hinting or some such

[01:29:00.0000] <annevk>
and if you just have hinting it stops at word boundaries or some such instead of the entire sentence

[01:29:01.0000] <Lachy>
it would make more sense if they just used examle letters from actual RTL languages, instead of trying to confuse people with latin letters

[01:29:02.0000] <annevk>
in the bidi world the convention is pretty well established I think and it is much easier to use characters that fall within the ASCII range

[01:30:00.0000] <annevk>
also, by using characters in the ASCII range you do not get the problem that it might be displayed incorrectly because the bidi implementation is buggy

[01:30:01.0000] <annevk>
which seems important

[01:31:00.0000] <Lachy>
they could just give the examples using real RTL letters, and then provide an image with the intended rendering

[01:32:00.0000] <annevk>
it explains this even: http://www.w3.org/TR/2010/WD-html-bidi-20100304/#notation

[01:32:01.0000] <annevk>
seems like a lot of additional work

[02:38:00.0000] <annevk>
all this crap about modular

[02:38:01.0000] <annevk>
all I see is more boilerplate

[02:46:00.0000] <Philip`>
/me doesn't see that, since he uses the WHATWG copy instead

[02:47:00.0000] <Dashiva>
I suppose it's an externality then :P

[02:52:00.0000] <annevk>
Philip`, fair point

[03:46:00.0000] <annevk>
fun, Ubuntu does not ship with either curl or cvs

[03:46:01.0000] <annevk>
guess that makes sense

[03:49:00.0000] <Dashiva>
What about svn?

[03:54:00.0000] <Philip`>
As far as I'm aware it doesn't even ship with GCC

[03:55:00.0000] <Philip`>
Software developers clearly aren't the primary audience of the default install

[03:58:00.0000] <annevk>
Dashiva, nope

[03:59:00.0000] <annevk>
still ships with apt-get though

[03:59:01.0000] <Dashiva>
Doesn't the update manager just frontend apt-get?

[04:00:00.0000] <annevk>
prolly

[04:44:00.0000] <mut>
arrg anyone around, could really do with a little help re .translate() on canvas

[04:49:00.0000] <Philip`>
/me guesses some people may be around

[04:49:01.0000] <mut>
heh

[04:49:02.0000] <mut>
hey Philip`  ;)

[04:50:00.0000] <mut>
im having an issue, I have some things im drawing, and moving around the canavas with translate

[04:50:01.0000] <mut>
I was kind of stacking the translates if you know what i mean

[04:51:00.0000] <mut>
and I decided it would be easier if i make a variable, and change that, and do 1 translate at the end of the function, so when i restore it, i have a reference as to where the origin is for that drawing

[04:51:01.0000] <mut>
but it just breaks when i do that, and i cannot figure out why

[04:51:02.0000] <mut>
:P

[04:52:00.0000] <Philip`>
Uh, don't quite understand what you're doing

[04:52:01.0000] <mut>
http://stairbox.com/tom/test/ << thats what im working on

[04:52:02.0000] <mut>
and it works at the mo

[04:53:00.0000] <mut>
if you select left turn for ex

[04:53:01.0000] <mut>
butif you look at the code, i commented out what i was trying to do 

[04:53:02.0000] <mut>
if that makes sense

[04:53:03.0000] <mut>
its in drawRun2 where the problem is

[04:53:04.0000] <mut>
but i used that method in drawRun1, and it seems to be fine

[04:54:00.0000] <mut>
as far as i can tell im using .translate correctly

[04:57:00.0000] <Philip`>
mut: Looks like the problem is you're mixing translate and rotate in there

[04:58:00.0000] <Philip`>
If you do e.g. "translate(10, 0); rotate(180 degrees); translate(10, 0)" then the second translation is relative to the rotated origin, so you end up back where you started

[04:58:01.0000] <mut>
oh

[04:58:02.0000] <Philip`>
and it's not equivalent to "rotate(180 degrees); translate(20, 0)" which is what you get if you just accumulate all the translations

[04:58:03.0000] <mut>
errm

[04:58:04.0000] <mut>
ok think im with you

[04:59:00.0000] <mut>
oh right, so if i shift the rotates beyond the last translate

[04:59:01.0000] <mut>
gotcha

[04:59:02.0000] <mut>
:)

[05:00:00.0000] <Philip`>
Hmm, maybe you're not actually doing translate-rotate-translate but you seem to be switching between translate-rotate (in the original code) and rotate-translate (in the run2Origin code) and they're not equivalent

[05:00:01.0000] <mut>
yea im with you

[05:00:02.0000] <Philip`>
Should be okay if you're consistent about the order of operations

[05:00:03.0000] <mut>
sorted

[05:00:04.0000] <mut>
:)

[05:01:00.0000] <mut>
yea sorted that, ill just expand and make sure it works across everything

[05:02:00.0000] <mut>
thanks a lot Philip` 

[05:02:01.0000] <mut>
heh

[05:02:02.0000] <mut>
saved my bacon

[05:05:00.0000] <Philip`>
Mmm, bacon

[05:10:00.0000] <mut>
english bacon, or american bacon?

[05:13:00.0000] <Dashiva>
Carrying a coconut?

[05:15:00.0000] <mut>
eh

[05:17:00.0000] <Magical1>
is there a particular tag used to mark up an area of a document containing dynamic content (AJAX) in HTML5, before I apply WAI-ARIA to it?

[05:35:00.0000] <MikeSmith>
Magical1: <div class="dynamic-content">... ?

[06:32:00.0000] <lyhana8>
hi,  how an old browser that doesn't know the <header> tag or other will behave with them ?

[06:34:00.0000] <Dashiva>
More or less like a <div>

[06:36:00.0000] <Philip`>
Less if the old browser is IE

[06:38:00.0000] <lyhana8>
but it can totally blow up a layout if the behaviour is unpredictable

[06:40:00.0000] <mut>
i dont know why they dont just use one of the open source rendering engines for ie

[06:40:01.0000] <mut>
started advertising ie8 on tv now :/

[06:42:00.0000] <lyhana8>
mut: which country  ?

[06:42:01.0000] <mut>
uk

[06:42:02.0000] <mut>
banging on about its saftey and private browsing etc

[06:42:03.0000] <mut>
pfffft

[06:44:00.0000] <lyhana8>
did they explain tabs browsing ?? Cause a lot of people doesn't use them u_u

[07:47:00.0000] <mut>
errm

[07:48:00.0000] <mut>
yea tabbed browsing too i  think

[07:57:00.0000] <Dashiva>
/me wonders how make firefox stop lying and actually open new windows as tabs

[09:59:00.0000] <LoneStar99>
Philip' you around?

[10:00:00.0000] <Philip`>
No

[10:01:00.0000] <LoneStar99>
k, figured out the following problem: http://urloid.com/error1

[10:01:01.0000] <LoneStar99>
it is a video 

[10:04:00.0000] <Philip`>
/me will look when he is around

[10:04:01.0000] <LoneStar99>
the fix is a global, which would maintain the the last positions, for some reason after leaving the drawing scene it would start happening

[10:04:02.0000] <LoneStar99>
so a global X and Y fixed it

[10:09:00.0000] <LoneStar99>
thanks for your help

[10:58:00.0000] <mut>
any way to flip something in canvas

[10:59:00.0000] <mut>
with scale(-1, 1) or something

[10:59:01.0000] <mut>
cant find any way

[11:03:00.0000] <mut>
nm

[11:03:01.0000] <mut>
it was me being daft

[11:03:02.0000] <mut>
heh


2010-03-07
[04:26:00.0000] <MikeSmith>
AryehGregor: if you around, I wanted to ask if/how template pages can be created in mediawiki

[04:27:00.0000] <MikeSmith>
oh, just Template*, I guess?

[08:53:00.0000] <webben>
Magical1: http://www.w3.org/WAI/PF/aria-practices/#docmgt

[09:38:00.0000] <AryehGregor>
MikeSmith, any page can be used as a template in MediaWiki.  If you say {{Foo:Bar}}, it will transclude the page "Foo:Bar".  The only way templates are different is that if you don't state a namespace, Template: is assumed, so {{Bar}} is the same as {{Template:Bar}}.  (You can get main-namespace pages with an explicit empty namespace, e.g., {{:Bar}}.)

[15:22:00.0000] <Bolkonskij>
hi


2010-03-08
[18:38:00.0000] <MikeSmith>
Hixie: about http://www.w3.org/Bugs/Public/show_bug.cgi?id=9198

[18:39:00.0000] <MikeSmith>
since hspace and vspace are already now allowed on embed, I assume that bug must be asking that they be explicitly listed in the obsolete-and-nonconforming section

[23:00:00.0000] <Hixie>
/me wonders what the chart on http://blogs.msdn.com/ie/archive/2010/03/05/ie8-smartscreen-filter-protecting-users-at-internet-scale.aspx shows

[23:05:00.0000] <MikeSmith>
Hixie: "mean block rate for socially engineered software"

[23:05:01.0000] <MikeSmith>
s/software/malware/

[23:06:00.0000] <MikeSmith>
(in the full PDF report that's linked to there) 

[23:10:00.0000] <MikeSmith>
It might help if they actually used something closer to commonly used terms

[23:11:00.0000] <MikeSmith>
since as far as I can see it appears to be comparing the anti-phishing mechanisms that various browsers provide

[23:15:00.0000] <nessy>
and what do the percentages mean?

[23:15:01.0000] <nessy>
number of attacks captured?

[23:15:02.0000] <MikeSmith>
nessy: yeah, seems so

[23:16:00.0000] <nessy>
they should compare the absolute number of attacks that succeeded, that would be more interesting :)

[23:16:01.0000] <othermaciej>
I've seen the similar study published before and it seemed questionable to me

[23:16:02.0000] <othermaciej>
1) they don't give enough information to reproduce their results (neither original data, nor enough detail of the methodology)

[23:18:00.0000] <othermaciej>
2) what they do say sounded like there was a lot of arbitrary fiddling with the original test sites (e.g. to remove ones that exploit security bugs in some way other than social engineering, which seems pretty point-missing)

[23:18:01.0000] <othermaciej>
3) it's not clear if their data source for what sites contain malware in the first place is independent from Microsoft's

[23:19:00.0000] <othermaciej>
I would also say it's very suspicious that Firefox, Safari and Chrome get such different results, considering they all use Google's malware/phishing data

[23:40:00.0000] <MikeSmith>
othermaciej: hmm, updated my Webkit and now I get "Nightly builds of WebKit are not supported on Mac OS X 10.5 at this time" 

[23:40:01.0000] <othermaciej>
MikeSmith: that's a surprise to me

[23:42:00.0000] <othermaciej>
MikeSmith: it's also a surprise to the guy who makes the nightlies, so it's probably a bug

[23:42:01.0000] <MikeSmith>
ah, OK

[23:42:02.0000] <othermaciej>
MikeSmith: does it work if you relaunch it?

[23:43:00.0000] <othermaciej>
I'm told that alert sometimes pops up incorrectly

[23:43:01.0000] <MikeSmith>
doesn't work even if I relaunch it

[23:43:02.0000] <MikeSmith>
I tried a few times now

[23:44:00.0000] <MikeSmith>
fwiw, Get Info shows r55610

[00:01:00.0000] <MikeSmith>
othermaciej: btw, can you advise me on how best to proceed with discussion about getting the author view of the HTML5 spec published as a WD?

[00:02:00.0000] <MikeSmith>
Hixie: btw, are you supportive of publishing the static author view of the spec as a WD?

[00:02:01.0000] <MikeSmith>
http://dev.w3.org/html5/spec-author-view/

[00:02:02.0000] <othermaciej>
MikeSmith: if the editor proposes it, then the chairs would likely put an FPWD resolution before the group

[00:02:03.0000] <othermaciej>
I believe it meets the threshold of being a bona fide collaborative product of the WG

[00:03:00.0000] <othermaciej>
and I believe it is reasonably in scope

[00:03:01.0000] <MikeSmith>
OK, thanks

[00:03:02.0000] <MikeSmith>
so I guess will propose to Hixie that he propose it

[00:03:03.0000] <othermaciej>
I did notice that the author view radio button is now available on the TR page

[00:03:04.0000] <MikeSmith>
yeah

[00:03:05.0000] <MikeSmith>
let's please not make a lot of noise about that

[00:04:00.0000] <othermaciej>
MikeSmith: btw, my preference would be to start the publication cycle about 2 months from now

[00:04:01.0000] <MikeSmith>
no one has ever explicitly told me that we can't have JS in docs in TR

[00:04:02.0000] <othermaciej>
it took about 1 month from the time the publication CfC was posted to the time we published

[00:04:03.0000] <MikeSmith>
othermaciej: works for me

[00:04:04.0000] <othermaciej>
so we have to start the CfC in 2 months to actually satisfy the heartbeat requirement

[00:04:05.0000] <MikeSmith>
yeah

[00:04:06.0000] <othermaciej>
if there were additional drafts ready by then, that would be cool

[00:05:00.0000] <othermaciej>
also if we publish WDs more often then hopefully it can be less dramatic

[00:05:01.0000] <MikeSmith>
yes

[00:05:02.0000] <MikeSmith>
so that means restarting WD discussion in early May, publishing in June

[00:06:00.0000] <MikeSmith>
it would good to get new drafts out before the July/August holiday period

[00:06:01.0000] <MikeSmith>
when a lot of people tend to be away for weeks at a time

[00:07:00.0000] <Hixie>
MikeSmith: i'm happy for it to be there, though i think if we publish it we should probably do it as a subcomponent of the HTML5 spec, not a separate /TR/ deliverable

[00:07:01.0000] <Hixie>
MikeSmith: probably should make it clear that it's non-normative, too

[00:07:02.0000] <othermaciej>
Hixie: how would you do it as a subcomponent? separate link in the table of contents?

[00:11:00.0000] <othermaciej>
MikeSmith: bdash suggests "download a fresh build from nightly.webkit.org, and to save the bad build and file a bug report with it"

[00:12:00.0000] <Hixie>
yeah, i guess

[00:12:01.0000] <Hixie>
well, no

[00:12:02.0000] <Hixie>
more like the same way we do the multipage version

[00:12:03.0000] <Hixie>
there's a w3c convention for this kind of thing

[00:13:00.0000] <Hixie>
links to pdf versions, multipage versions, etc

[00:14:00.0000] <MikeSmith>
ah yeah

[00:14:01.0000] <MikeSmith>
good point

[00:14:02.0000] <MikeSmith>
we can just treat it as a variant of the full spec

[00:14:03.0000] <othermaciej>
ah, alternate non-normative formats

[00:18:00.0000] <boblet>
MikeSmith: re: http://dev.w3.org/html5/spec-author-view/ what’s up with the double asterisk overlapping “call” in “Status: last call for comments”

[00:18:01.0000] <boblet>
eg http://dev.w3.org/html5/spec-author-view/sections.html#the-article-element

[00:19:00.0000] <boblet>
looks like generated content, but I can’t see from where

[00:20:00.0000] <MikeSmith>
boblet: no idea

[00:21:00.0000] <MikeSmith>
some borkedness in the CSS stylesheet I'll need to unwind

[00:21:01.0000] <boblet>
heh

[00:24:00.0000] <Hixie>
boblet: what browser?

[00:24:01.0000] <MikeSmith>
othermaciej: OK, grabbed a fresh build and it's working fine

[00:24:02.0000] <othermaciej>
MikeSmith: woot

[00:25:00.0000] <MikeSmith>
othermaciej: so what should I attache to the bug report?

[00:25:01.0000] <MikeSmith>
zip up the WebKit.app ?

[00:25:02.0000] <boblet>
Hixie: seeing those asterisks in Chrome and FF latest for Mac. Only me?

[00:25:03.0000] <Hixie>
boblet: does it disappear if you hit command++/command+- ?

[00:26:00.0000] <boblet>
nope. changing body font to serif changes position slightly though

[00:26:01.0000] <boblet>
(moves it to end of “call”)

[00:27:00.0000] <boblet>
holler if you want screenshots

[00:27:01.0000] <Hixie>
boblet: sounds like a browser bug to me -- misplaced ::before or ::after from a class="XXX" element.

[00:28:00.0000] <Hixie>
odd that it doesn't move when you resize though

[00:28:01.0000] <MikeSmith>
.XXX:before, .XXX:after { content: " ** "; position: absolute; left: 0; width: 8em; text-align: right; }

[00:29:00.0000] <boblet>
MikeSmith: there you go. Pity inspector & Firebug don’t show generated content (& make it hard to access eg hover state CSS)

[00:29:01.0000] <MikeSmith>
indeed

[00:29:02.0000] <MikeSmith>
anyway, I just checked in a fiew

[00:29:03.0000] <MikeSmith>
fix

[00:30:00.0000] <MikeSmith>
eagle eye

[00:30:01.0000] <boblet>
happy to help ;-)

[00:30:02.0000] <MikeSmith>
/me thinks boblet must read through specs with a jeweler's loupe

[00:31:00.0000] <boblet>
“hey, designers *are* good for something!”

[00:31:01.0000] <MikeSmith>
heh

[00:31:02.0000] <Hixie>
MikeSmith: what was the bug? (what did you change?)

[00:34:00.0000] <MikeSmith>
Hixie: I made a copy of the http://www.whatwg.org/style/specification stylesheet

[00:35:00.0000] <MikeSmith>
and have I that document referencing it

[00:35:01.0000] <MikeSmith>
I can't quite remember why, now

[00:35:02.0000] <MikeSmith>
but that's what it's doing

[00:35:03.0000] <MikeSmith>
and what I did was just to comment out that part

[00:35:04.0000] <MikeSmith>
.XXX:before, .XXX:after { content: " ** "; position: absolute; left: 0; width: 8em; text-align: right; }

[00:38:00.0000] <Hixie>
neither of those should be necessary :-)

[00:38:01.0000] <boblet>
maybe someone who really really liked Markdown wrote the CSS?

[00:39:00.0000] <MikeSmith>
I blame the CSS gremlins

[00:42:00.0000] <Hixie>
boblet: i wrote that css, but it shouldn't be doing what you describe

[00:42:01.0000] <Hixie>
boblet: hence why i think it's a browser bug

[00:44:00.0000] <Creap>
is there any input which can represent a time including seconds?

[00:46:00.0000] <boblet>
Hixie: what was the effect you wanted? It seems like it’s putting ** 8em from the left inside .XXX, and that happens to be about where class is

[00:47:00.0000] <zcorpan>
Creap: "A time consists of a specific time with no time-zone information, consisting of an hour, a minute, a second, and a fraction of a second."

[00:48:00.0000] <zcorpan>
Creap: from the spec on <input type=time>

[00:49:00.0000] <Creap>
ok, I was confused by http://www.w3.org/TR/html-markup/input.time.html#input.time + Opera's implementation lacking seconds

[00:51:00.0000] <Hixie>
boblet: oh, i thought you meant it was nowhere near anything that should have had the stars

[00:51:01.0000] <Hixie>
Creap: make sure to set step=1

[00:52:00.0000] <Hixie>
boblet: it's supposed to attract attention to the issues

[00:52:01.0000] <boblet>
Hixie: the ** appear inside a paragraph with .XXX, so it’s as expected

[00:52:02.0000] <Hixie>
boblet: ah ok

[00:53:00.0000] <Creap>
ah.

[00:54:00.0000] <zcorpan>
Creap: could you file a bug about opera's impl?

[00:55:00.0000] <Creap>
should it include seconds by default? using step=1 as hixie suggested makes it show seconds

[00:55:01.0000] <boblet>
Hixie: if you were wanting ** before and after, I think you might need to define :before and :after separately

[00:58:00.0000] <mut>
do i have to use .closePath? or not?

[01:09:00.0000] <Hixie>
boblet: why?

[01:10:00.0000] <Hixie>
Creap: step=60 is the default, so it hides seconds by default

[01:10:01.0000] <Creap>
ok

[01:11:00.0000] <Creap>
is datetime's step also specified in seconds?

[01:11:01.0000] <Hixie>
iirc, yes

[01:12:00.0000] <Hixie>
ok i'm outta here. g'night all.

[01:12:01.0000] <MikeSmith>
おつかれ

[01:16:00.0000] <zcorpan>
Creap: step="any" should make it show seconds

[01:18:00.0000] <MikeSmith>
/me wonders if there's a dev.opera.com article on forms features

[01:18:01.0000] <zcorpan>
http://dev.opera.com/articles/view/improve-your-forms-using-html5/

[01:19:00.0000] <MikeSmith>
hmm

[01:19:01.0000] <zcorpan>
it'd be nice with an article with more coverage

[01:19:02.0000] <MikeSmith>
yeah

[01:19:03.0000] <MikeSmith>
http://dev.opera.com/articles/tags/forms/ :(

[01:28:00.0000] <MikeSmith>
hmm, we are really going to need some how-to guides about the new input types and other forms features

[01:29:00.0000] <MikeSmith>
this stuff is not especially intuitive

[01:48:00.0000] <annevk>
http://www.w3.org/People/Jaffe/

[02:00:00.0000] <knowtheory>
yeh, new ceo eh annevk?

[02:00:01.0000] <knowtheory>
and his first post: http://www.w3.org/QA/2010/03/w3c_ceo_blog_first_posting_mar.html

[02:30:00.0000] <hsivonen>
does ie8 smartscreen differ from the phishing and malware blacklist features of other browsers in terms of features? or is the difference microsoft's server side instead of google's?

[02:35:00.0000] <othermaciej>
they have some purely client-side heuristics too, I believe

[02:40:00.0000] <knowtheory>
othermaciej: what time zone are you in? o_O

[02:40:01.0000] <othermaciej>
"other"

[02:41:00.0000] <knowtheory>
heh, ah yes, that one, i know it well.

[02:41:01.0000] <knowtheory>
(it's awkward living in EST, and working w/ people in GMT and PST)

[02:44:00.0000] <zcorpan>
http://www.w3.org/WAI/PF/src/role-attribute-src.html

[02:48:00.0000] <annevk>
the examples in that document seem somewhat misguided

[03:18:00.0000] <MikeSmith>
does Webkit already have some mechanism for exposing orientiation/acceleromerter events to Web applications?

[03:18:01.0000] <MikeSmith>
was reading the new Orientation Event spec http://dev.w3.org/geo/api/spec-source-orientation.html

[03:19:00.0000] <MikeSmith>
and noticed that it informatively references https://developer.mozilla.org/en/XPCOM_Interface_Reference/nsIDOMOrientationEvent

[03:19:01.0000] <MikeSmith>
and I remember recently seeing a guy do a File API demo in Mozilla that appeared to be reacting to changes in orientation of his Macbook

[03:28:00.0000] <knowtheory>
hah, that's cool :D

[03:33:00.0000] <MikeSmith>
cool also to see that work seems to have started on implemented CSS3 Paged Media in Webkit - https://bugs.webkit.org/show_bug.cgi?id=15548

[05:00:00.0000] <zcorpan>
hsivonen: "but it looks like once it reads a conditional comment, it's too late to change X-UA-Compatible." -- http://www.sitepoint.com/forums/showthread.php?t=663679

[05:12:00.0000] <hsivonen>
zcorpan: Yeah, I'm aware that conditional comments (like scripts) make later X-UA-Compatible ineffective. I just haven't gotten around to updating the flowchart.

[05:24:00.0000] <zcorpan>
ok

[05:40:00.0000] <asmodai>
mmm

[05:41:00.0000] <asmodai>
So why doesn't HTML 5 include ruby markup in the spec? :)  (per Ishida's email on www-int)

[05:41:01.0000] <annevk>
it doesn't?

[05:42:00.0000] <asmodai>
Hmm, apparently not

[05:43:00.0000] <myakura>
It does http://www.whatwg.org/html5#the-ruby

[05:43:01.0000] <asmodai>
hmmm

[05:43:02.0000] <asmodai>
yeah, just looking at that

[05:43:03.0000] <annevk>
http://lists.w3.org/Archives/Public/www-international/2010JanMar/0107.html seems Richard is asking about complex ruby

[05:43:04.0000] <annevk>
which was deliberately left out

[05:43:05.0000] <asmodai>
annevk: there's complex?

[05:44:00.0000] <asmodai>
/me thought there was only 1 Ruby spec

[05:44:01.0000] <annevk>
see http://www.w3.org/TR/2001/REC-ruby-20010531/#complex

[05:44:02.0000] <annevk>
asmodai, well yeah, but nobody implements /TR/ruby/

[05:45:00.0000] <asmodai>
Mmm, have never seen the complex case being used for Asian languages.

[05:45:01.0000] <asmodai>
annevk: Aside from the addon for Firefox?

[05:45:02.0000] <asmodai>
annevk: XHTML Ruby

[05:48:00.0000] <annevk>
asmodai, I doubt that plugin implements it correctly

[05:48:01.0000] <annevk>
but yeah

[05:48:02.0000] <asmodai>
annevk: Actually, it passes all tests IIRC

[05:49:00.0000] <asmodai>
annevk: http://www.w3.org/International/tests/results/results-ruby-markup-2

[05:50:00.0000] <asmodai>
bit older version of the addon

[06:02:00.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/www-international/2010JanMar/0107.html

[06:18:00.0000] <annevk>
asmodai, if it was done correctly it would've made it into Gecko by now

[06:28:00.0000] <hsivonen>
annevk, asmodai: addons don't go deep into layout and add CSS box types. proper Ruby support needs support from the CSS formatter, IIRC.

[06:42:00.0000] <annevk>
hsivonen, right

[07:07:00.0000] <asmodai>
hsivonen: I know nothing of that. :)

[07:07:01.0000] <boblet>
did anything come of using @datetime on <article>? I’m guessing better to not have as an invisible attribute, but…

[07:07:02.0000] <asmodai>
hsivonen: So I'll take your word for it.

[07:09:00.0000] <boblet>
huh. datetime isn’t mentioned under the-article-element and isn’t in global attribs, so I guess no

[07:12:00.0000] <MikeSmith>
boblet: I seem to remember it being there for a while, then being removed

[07:12:01.0000] <MikeSmith>
but I may have imagined it all

[07:12:02.0000] <boblet>
I definitely remember it being discussed for blog posts, but yeah not there now

[07:14:00.0000] <Philip`>
I thought it was called pubdate now

[07:14:01.0000] <mr_danie1>
is the following sentence in the book 'div into html5' (http://diveintohtml5.org/) book really true:

[07:14:02.0000] <mr_danie1>
...The HTTP header is the preferred method, and it overrides the <meta> tag if present....

[07:15:00.0000] <mr_danie1>
you can find it in this page: http://diveintohtml5.org/semantics.html#encoding

[07:15:01.0000] <mr_danie1>
I thought the reason why the <meta charset=...> tag is available is because most authores CANNOT set the http content-type header

[07:16:00.0000] <mr_danie1>
but when the http content-type header, which could be wrong', overwrites any authors attempt to alter the charset, the <meta charset=...> wouldn't make sense

[07:17:00.0000] <boblet>
Philip`: aah nice, thanks

[07:17:01.0000] <mr_danie1>
I think this must be a typo in the book. can someone confirm this?

[07:17:02.0000] <boblet>
mr_danie1: I was under the impression that http header > meta charset

[07:18:00.0000] <workmad3>
mr_danie1: I think the book is correct

[07:18:01.0000] <workmad3>
http header outranks the meta charset

[07:18:02.0000] <Philip`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/parsing.html#determining-the-character-encoding

[07:18:03.0000] <boblet>
mr_danie1: http://lists.w3.org/Archives/Public/www-validator/2008Mar/0024.html

[07:18:04.0000] <boblet>
heh

[07:18:05.0000] <Philip`>
"1. If the transport layer specifies an encoding, and it is supported, return that encoding with the confidence certain, and abort these steps."

[07:19:00.0000] <Philip`>
(where HTTP is a transport layer, and specifies encodings via Content-Type)

[07:23:00.0000] <workmad3>
mr_danie1: in the case where the server is serving data that it doesn't know the content type for, it should omit the content-type header completely I believe

[07:23:01.0000] <workmad3>
rather than send a false type that will be taken as exact

[07:24:00.0000] <mr_danie1>
I think  meta charset > http header would be better, for example: the server sets content-type:charset=utf-8, but the author sets <meta charset=iso-...

[07:24:01.0000] <mr_danie1>
and when the http header, the content will be displayed wrong, because the content is iso-...

[07:25:00.0000] <mr_danie1>
...http header WIN, the...

[07:25:01.0000] <Philip`>
mr_danie1: Can't do that because it would be incompatible with all current browsers

[07:25:02.0000] <workmad3>
mr_danie1: if the server is sending an incorrect content type, the server has a bug

[07:25:03.0000] <Philip`>
and probably incompatible with lots of current pages

[07:26:00.0000] <workmad3>
or is configured incorrectly... either way, it's not a problem with the standard

[07:26:01.0000] <Philip`>
You should always use UTF-8 anyway :-)

[07:26:02.0000] <mr_danie1>
ok, so html5 must use this policy because of its 'incremental html improvement' approach not to break the current web

[07:26:03.0000] <mr_danie1>
workmad3: true, then the server has a problem

[07:27:00.0000] <workmad3>
also, I believe in a lot of cases, it's the <meta charset= that is incorrect, because it was set on old content that has since been re-encoded

[07:27:01.0000] <mr_danie1>
Philip`: true, utf-8 is my favourite; it hate encoding issues, especially when I work in a team with lot of people

[07:27:02.0000] <Philip`>
mr_danie1: Yes, and also because it's probably better for the server to be authoritative

[07:28:00.0000] <workmad3>
and (checking the wiki page on http encodings) it looks like <meta> content types should be used by the server, not the browser

[07:28:01.0000] <Philip`>
You already have to trust the server to tell you the content is text/html rather than image/png or whatever, so it'd be odd if you trusted it to say the "content-type: text/html" part but not the "; charset=utf-8" part

[07:29:00.0000] <Philip`>
workmad3: ?

[07:29:01.0000] <Philip`>
Servers shouldn't be parsing HTML

[07:29:02.0000] <workmad3>
'A known misconception about <meta http-equiv="Content-Type">  is that meta element is intended to be interpreted directly by a browser, like an ordinary HTML tag. According to WWW Consortium, it helps HTTP server[1]  to generate some headers when it serves the document.'

[07:30:00.0000] <workmad3>
that said, it's a wiki page... it could easily be incorrect :)

[07:30:01.0000] <Philip`>
I assume that was the original intention, hence it being called "http-equiv" and being equivalent to HTTP headers, but that's crazy and not how it works today

[07:31:00.0000] <workmad3>
it does seem to gel with the description here as well though: http://www.w3.org/TR/html401/struct/global.html#edef-META

[07:31:01.0000] <Philip`>
Indeed, but HTML4 is crazy and doesn't reflect reality either

[07:32:00.0000] <workmad3>
heh :)

[07:37:00.0000] <Dashiva>
Imagine how fun it would be if every web server had its own HTML parser

[07:38:00.0000] <workmad3>
Dashiva: how about if web servers sniffed the browser types and rewrote the HTML they were serving to be compatible? :D

[07:39:00.0000] <Philip`>
Most web pages are generated dynamically and all the headers are output before any content, so it wouldn't even be possible to parse it and add headers

[07:41:00.0000] <asmodai>
annevk: Can you elaborate why it was a deliberate decision btw? (complex ruby)

[07:41:01.0000] <workmad3>
it's all just crazy anyway... encoding issues only exist because people were insane, and they turn sane people into insane people even now

[07:45:00.0000] <annevk>
asmodai, I believe that nobody really requested more than what IE already supported

[07:48:00.0000] <Philip`>
workmad3: I don't think it's because people were insane - it was logical and sensible to invent all sorts of 8-bit encodings when computers rarely communicated with each other (and especially not internationally), to suit their particular needs

[07:48:01.0000] <Philip`>
and then it was necessary to continue supporting them for legacy compatibility

[07:49:00.0000] <Philip`>
which results in all these encoding issues

[07:50:00.0000] <workmad3>
Philip`: maybe... I'll stick with my 'people are insane' view though, it justifies me cursing encodings more ;)

[07:50:01.0000] <Philip`>
I'm not saying people aren't insane, just saying that there's other reasons for this situation :-)

[07:51:00.0000] <asmodai>
annevk: ah ok

[07:51:01.0000] <Philip`>
and insanity seems a reasonable explanation for why people continue to expose themselves to encoding issues even if they can avoid it (like when inventing new formats and languages), instead of sticking with Unicode internally and UTF-8 externally and not having to worry about it

[07:51:02.0000] <asmodai>
annevk: I now feel like a proxy XD

[07:52:00.0000] <annevk>
pretty sure this was already discussed before with Richard

[07:57:00.0000] <asmodai>
annevk: could be, at least I pointed it out to him now, I'm sure he'll get in contact

[08:08:00.0000] <asmodai>
annevk: Thanks :)

[08:09:00.0000] <annevk>
np

[09:31:00.0000] <MikeSmith>
anybody remember what as the rationale for not making the <rb> element conforming?

[09:31:01.0000] <MikeSmith>
just for simplification?

[09:32:00.0000] <MikeSmith>
if there is existing content that uses <rb> and that works as expected in IE, it would seem counter-productive to make such content invalid in HTML5

[09:38:00.0000] <Philip`>
MikeSmith: I believe it's not supported in IE

[09:39:00.0000] <Philip`>
(or it's no more supported than the <thisdoesnotexist> tag)

[09:39:01.0000] <MikeSmith>
Philip`: OK

[09:39:02.0000] <MikeSmith>
I guess the issue is that IE does not object to it

[09:39:03.0000] <Philip`>
It's used sometimes but not always in existing content

[09:39:04.0000] <MikeSmith>
I see

[09:40:00.0000] <Philip`>
http://philip.html5.org/demos/html/ruby/wild-examples.html

[09:40:01.0000] <MikeSmith>
/me looks now

[09:42:00.0000] <Philip`>
I'd guess the reason for excluding is simplification - there's no point requiring people to do something that's unnecessary, and making it optional is adding needless complexity for authors

[09:42:01.0000] <MikeSmith>
Philip`: I see

[09:43:00.0000] <Philip`>
though I don't see any explicit rationale in the mailing lists, other than stating it's unnecessary and IE doesn't support it

[09:43:01.0000] <Philip`>
Also it doesn't seem a very widely used feature so it doesn't hurt too much to make current uses non-conforming

[09:44:00.0000] <MikeSmith>
that's always a judgement call, I guess

[09:44:01.0000] <Philip`>
(Apparently I saw it on 0.3% of .jp sites in dmoz.org)

[09:45:00.0000] <Philip`>
It's nowhere near as numerically significant as making e.g. <br/> conforming

[09:45:01.0000] <MikeSmith>
I see enough instances of <rb> in that sample at least that I am inclined to think that by prohibiting we are going to risk frustrating and confuseing a significant number sites that are already using it

[09:45:02.0000] <Philip`>
and people still argued for keeping the syntax simpler in that case

[09:47:00.0000] <Philip`>
MikeSmith: I expect confusion could be alleviated by validators printing a message saying "the <rb> tag is unnecessary and not permitted in HTML5 - remove the tags and your page will work just fine" rather than just "invalid element <rb>"

[09:52:00.0000] <MikeSmith>
Philip`: yeah, that we can do

[09:53:00.0000] <MikeSmith>
and our contract is with users is not that we are guaranteeing that everything that was valid in HTML4 or XHTML1 will continue to be valid in HTML5

[12:23:00.0000] <Hixie>
MikeSmithX: .3% iirc is almost as low as the number of pages that use <image>, and we're not making _that_ conforming :-P

[13:11:00.0000] <knowtheory>
Hixie: if i wanted to write up a proposal for something, is http://dev.w3.org/html5/status/issue-status.html a good place to look at examples of how people have structured other proposals?

[13:15:00.0000] <othermaciej>
knowtheory: you should look here: <http://dev.w3.org/html5/decision-policy/decision-policy.html>

[13:16:00.0000] <Hixie>
knowtheory: no, not really

[13:16:01.0000] <Hixie>
knowtheory: if you want to propose something, best thing to do is to e-mail whatwg with a description of the problem you want to solve

[13:16:02.0000] <Hixie>
imho

[13:17:00.0000] <knowtheory>
okay, thanks guys

[14:08:00.0000] <JonathanNeal>
Woohoo, today we're putting the new HTML5 stuff into the product.

[14:48:00.0000] <othermaciej>
knowtheory: the Change Proposal structure is only needed if something has been escalated, not for an initial informal proposal

[14:48:01.0000] <knowtheory>
othermaciej: okay :)

[14:50:00.0000] <knowtheory>
i've signed up to the whatwg list, so i'll post there if that's appropriate?

[14:57:00.0000] <Hixie>
knowtheory: that is one of the appropriate places you can send feedback, yes

[14:57:01.0000] <Hixie>
(it's one of the places i guarantee a reply eventually, too)

[14:57:02.0000] <Hixie>
(the other option is the bugzilla system in the html working group)

[14:58:00.0000] <Hixie>
s/option/place/

[14:58:01.0000] <Hixie>
AryehGregor: yt?

[14:58:02.0000] <Hixie>
or sicking

[14:59:00.0000] <Hixie>
i'm looking at the empty attribute thread

[14:59:01.0000] <sicking>
which one?

[14:59:02.0000] <Hixie>
the one talking about not fetching things for <Script src="">, etc

[14:59:03.0000] <sicking>
ah

[14:59:04.0000] <knowtheory>
thanks Hixie

[14:59:05.0000] <sicking>
Hixie: what about it?

[14:59:06.0000] <Hixie>
if we make <script src=""> not fetch anything, should we also make it non-conforming?

[14:59:07.0000] <Hixie>
and does that mean <script src=""> and <script src=" "> would be different from each other?

[15:00:00.0000] <sicking>
Hixie: i don't feel strongly on conforming vs. non-conforming

[15:00:01.0000] <Hixie>
(<script src=" "> is non-conforming currently)

[15:00:02.0000] <sicking>
Hixie: is whitespace generally trimmed from these attributes?

[15:00:03.0000] <Hixie>
(but it does result in a fetch, which would make it different)

[15:00:04.0000] <Hixie>
well when resolving the url it is

[15:00:05.0000] <sicking>
Hixie: i.e. what does " " do currently?

[15:00:06.0000] <Hixie>
not sure what you mean by "generally"

[15:01:00.0000] <Hixie>
" " resolves to the base URL

[15:01:01.0000] <knowtheory>
do CSS Animations actually have a formal relationship to the HTML5 spec?  Or is it a detail of browser implementation, that CSS Animations would be included w/ HTML5?

[15:01:02.0000] <sicking>
Hixie: why does it resolve to the baseurl?

[15:01:03.0000] <Hixie>
knowtheory: CSS animations are part of CSS, nothing to do with HTML

[15:01:04.0000] <sicking>
Hixie: wouldn't it convert to "%20" which is appended to the base?

[15:01:05.0000] <Hixie>
sicking: spaces get trimmed first

[15:01:06.0000] <knowtheory>
Hixie: cool, just wanted to make sure i wasn't like way off base.

[15:02:00.0000] <sicking>
Hixie: ah, for relative urls? Or for all values of the attribute?

[15:02:01.0000] <Hixie>
i'm loathe to make "" invalid everywhere given that it's a valid URL, but making it invalid only in some places is a bit weird...

[15:02:02.0000] <Hixie>
all values

[15:02:03.0000] <sicking>
/me ponders

[15:02:04.0000] <Hixie>
the "web addresses" stuff, assuming larry gets around to fixing it properly, trims spaces before resolving urls

[15:03:00.0000] <Hixie>
seems like making <a href="">foo</a> invalid is a bit weird

[15:03:01.0000] <sicking>
Hixie: ok, from an implementation point of view, i'd prefer to just tread "" special. I.e. "" != " ". The former does nothing, the latter fetches the base url

[15:03:02.0000] <Hixie>
but making <script src=""> valid but not do anything is also weird...

[15:03:03.0000] <Hixie>
oh i agree that "" should be special, not in any way suggesting we make " " be handled differently

[15:04:00.0000] <Hixie>
it's the conformance aspects of this i'm struggling with

[15:04:01.0000] <sicking>
Hixie: on the validity issue i don't feel strongly. I agree both options are weird in different ways

[15:05:00.0000] <sicking>
Hixie: one argument is that it might actually be convenient for authors to be able to use <script src="">

[15:05:01.0000] <Hixie>
if we make "" invalid for external-resource <link>s, <link rel="stylesheet index" href=""> would be simultaneously valid and invalid, which is confusing as all hell

[15:06:00.0000] <Hixie>
but if we make it valid, then <link rel=stylesheet href=""> would be valid but would not have the effect it should have given the url ""

[15:06:01.0000] <sicking>
Hixie: as to avoid having to do <%if ($bar) {%><script src="<%$bar%>"></script><%}%>

[15:07:00.0000] <sicking>
Hixie: yeah, i agree with your argument too

[15:08:00.0000] <sicking>
Hixie: politically it might be easier to make it invalid

[15:08:01.0000] <Hixie>
i'm not worried about the politics

[15:08:02.0000] <othermaciej>
really? document validity tends to be more political than implementation conformance

[15:08:03.0000] <Hixie>
it's the usability of the language i'm worried about

[15:09:00.0000] <sicking>
then i would say don't make it invalid

[15:09:01.0000] <sicking>
to avoid forcing authors to use the complex syntax above

[15:09:02.0000] <Hixie>
having <link rel="stylesheet index" href=""> create just one link but <link rel="stylesheet index" href="#"> create two is very confusing

[15:09:03.0000] <Hixie>
imhio

[15:09:04.0000] <Hixie>
imho, even

[15:09:05.0000] <sicking>
though, i guess you could also argue that <script src=""> is likely a bug, and thus we'd help people by making it invalid so that the validator highlights the bug

[15:10:00.0000] <Hixie>
that was the argument for making it not fetch anything in the first palce

[15:10:01.0000] <Hixie>
place

[15:10:02.0000] <Hixie>
that it was always a bug

[15:10:03.0000] <sicking>
true

[15:10:04.0000] <sicking>
ok, make it invalid then :)

[15:10:05.0000] <Hixie>
is <a href="">foo</a> always a bug? maybe i should just make all empty URLs invalid

[15:11:00.0000] <sicking>
good question

[15:11:01.0000] <Hixie>
having some be invalid and some not is very poor form

[15:11:02.0000] <sicking>
well.. i don't think there are any good solutions here

[15:11:03.0000] <sicking>
so we'll have to pick one bad solution

[15:12:00.0000] <Hixie>
if we're picking bad solutions, it seems like making "" actually fetch the local page is the least bad solution from a usability perspective

[15:12:01.0000] <sicking>
why?

[15:12:02.0000] <Hixie>
it makes the language self-consistent and predictable

[15:13:00.0000] <Hixie>
we could say that if you refer to the local page that the UA must use the cached copy always

[15:13:01.0000] <sicking>
that seems like picking language purity over author friendlyness

[15:13:02.0000] <Hixie>
"purity"?

[15:13:03.0000] <Hixie>
it's picking author friendliness over server load friendliness

[15:13:04.0000] <sicking>
authors are generally in charge of server load. so the distinction doesn't make sense IMHO

[15:13:05.0000] <Hixie>
if we say that if you refer to the local page that the UA must use the cached copy always, it doesn't even affect the load, which would be good

[15:14:00.0000] <sicking>
i.e. you'd be hurting the same guy as you're helping

[15:14:01.0000] <Hixie>
granted

[15:14:02.0000] <sicking>
and probably hurting him more than helping

[15:14:03.0000] <Hixie>
but if we use the cache we're not hurting at all

[15:14:04.0000] <Hixie>
while still being predictable

[15:14:05.0000] <sicking>
there isn't always a cache around

[15:14:06.0000] <Hixie>
well, in those cases we can hurt, i guess

[15:14:07.0000] <Hixie>
there usually is

[15:15:00.0000] <sicking>
that doesn't seem to be true for <img src="">

[15:15:01.0000] <sicking>
given that many browsers gave that special treatment

[15:15:02.0000] <Hixie>
what isn't true?

[15:15:03.0000] <sicking>
that there is a cached copy around

[15:15:04.0000] <Hixie>
i don't see why the browser behaviour indicates that one way or the other

[15:16:00.0000] <Hixie>
it's just easier to implement "do nothing" than "fetch from the cache and then do nothing because HTML isn't an image format"

[15:16:01.0000] <sicking>
browsers implemented an exception because there was a detectable difference

[15:16:02.0000] <Hixie>
by "browsers" you mean FF and Opera

[15:17:00.0000] <Hixie>
all the other browsers fetch the image, according to the data collected in this thread

[15:17:01.0000] <sicking>
ah, thought it was more

[15:17:02.0000] <Hixie>
in fact it's one of the few cases where IE does fetch the URL -- most of hte time it doesn't if it's ""

[15:18:00.0000] <sicking>
i still don't see how we're helping anyone by saying it should be fetched from cache

[15:18:01.0000] <sicking>
also, how is "should fetch from cache" different than not saying so? No browsers that i know of fetch from the network for the heck of it

[15:19:00.0000] <Hixie>
well if fetching from the cache doesn't reduce the load, it's not clear to me why there's a problem here

[15:19:01.0000] <sicking>
so staying silent on the issue would result in the same browser behavior as far as i can see

[15:19:02.0000] <sicking>
huh?

[15:19:03.0000] <sicking>
fetching from the cache would reduce the load

[15:19:04.0000] <Hixie>
right now, many browsers fetch something in these "" cases. I'm saying we should make it not fetch anything new but just use the available data instead. That seems like it would reduce the load, which is the stated problem.

[15:20:00.0000] <sicking>
but obviously at least firefox, opera and yahoo felt that the cache didn't reduce the load enough

[15:20:01.0000] <daedb_>
Why would any author want to use empty src/href? That does not look sane to me :)

[15:20:02.0000] <Hixie>
i dunno about "obviously"

[15:20:03.0000] <sicking>
firefox and opera obviously since they implemented it

[15:21:00.0000] <sicking>
yahoo in general i agree. This guy from yahoo obviously since he spent a lot of time figuring this out while developing the site

[15:21:01.0000] <Hixie>
i think you're drawing conclusions from the data that aren't warranted -- e.g. opera's behaviour could be a random bug.

[15:21:02.0000] <Hixie>
anyway, let's start from the beginning

[15:22:00.0000] <Hixie>
the problem is that authors say src="", data="", etc, but don't mean to

[15:22:01.0000] <Hixie>
yet they (presumably) say <a href=""> and _do_ mean to

[15:22:02.0000] <Hixie>
we ideally would want to catch the errors but not flag the intentional cases

[15:22:03.0000] <Hixie>
we ideally would want to not fetch data from the network for the error cases

[15:23:00.0000] <Hixie>
we have certain language features where a single href="" can be in both situations at the same time, namely <link rel="stylesheet index" href="">

[15:23:01.0000] <sicking>
practically speaking though, when would that markup ever make sense?

[15:24:00.0000] <sicking>
given that i can't think of a case when it would, I don't care very much what the spec says about it

[15:24:01.0000] <sicking>
(i only care enough that i don't want to write a bunch of code to handle it, as that is effectively dead code)

[15:24:02.0000] <Hixie>
<link rel="prefetch next" href=""> might, especially with scripting involved

[15:25:00.0000] <sicking>
neither "prefetch" or "next" is on the list of suggested changes though, is it?

[15:25:01.0000] <Hixie>
"prefetch" is

[15:25:02.0000] <Hixie>
it's an external resource link

[15:26:00.0000] <sicking>
true

[15:26:01.0000] <Hixie>
and who knows what future external resource links might be invented

[15:26:02.0000] <sicking>
ok, so what is your point?

[15:27:00.0000] <Hixie>
no point, i was just describing the problem

[15:27:01.0000] <Hixie>
for daedb_'s benefit

[15:28:00.0000] <sicking>
it seems to me that everyone in that thread agreed on what UA behavior should be

[15:28:01.0000] <sicking>
so IMHO we should spec that behavior

[15:28:02.0000] <Hixie>
there's also the mildly interesting case of (XML) <img xml:base="...image..." src="" alt=.../>

[15:28:03.0000] <ap>
Hixie: <https://bugs.webkit.org/show_bug.cgi?id=30303>, if you want more opinions on the topic

[15:28:04.0000] <sicking>
the question of conformance has not been treated though

[15:28:05.0000] <sicking>
i don't have a strong opinion on that

[15:29:00.0000] <sicking>
though I think that in general things that are extremely likely to be bugs should be non-conforming

[15:29:01.0000] <Hixie>
ap: reading...

[15:29:02.0000] <sicking>
as a rule of thumb

[15:30:00.0000] <Hixie>
i guess we could make <a href="" rel=next> valid but <link href="" rel=next> invalid

[15:31:00.0000] <Hixie>
and most authors wouldn't run into it

[15:34:00.0000] <sicking>
Hixie: sounds ok to me

[15:36:00.0000] <Hixie>
should <script src=""></script> fire onload or onerror or neither?

[15:36:01.0000] <Hixie>
and if an event is fired, should it be fired synchronously or asynchronously when parsing?

[15:37:00.0000] <sicking>
Hixie: i'd say fire onerror. To keep the consistency that onload or onerror is always fired

[15:37:01.0000] <sicking>
Hixie: and fire asynch, to keep that consistency that all events are async

[15:37:02.0000] <sicking>
IMHO

[15:38:00.0000] <Hixie>
events with <script> aren't always async

[15:39:00.0000] <Hixie>
e.g. <script>document.write('<script onload="alert(1)">alert(0)<\/script>');alert(2);</script> per spec

[15:40:00.0000] <Hixie>
mind you either firefox nor safari fire that event at all as far as i can tell

[15:41:00.0000] <Hixie>
actually no browser seems to

[15:41:01.0000] <Hixie>
huh

[15:41:02.0000] <othermaciej>
spec bug?

[15:42:00.0000] <Hixie>
unclear

[15:42:01.0000] <Hixie>
i think firing onload in that case was an intentional addition

[15:42:02.0000] <Hixie>
not sure about the sync vs async

[15:47:00.0000] <Hixie>
ok i made it async in the internal case

[15:53:00.0000] <Hixie>
should <embed type="application/plugin" src=""> instantiate the plugin based on the type="" attribute, or not at all?

[15:56:00.0000] <Hixie>
i'll follow <object> and make <embed src=""> do nothing

[15:56:01.0000] <Hixie>
er, <embed src="" type="..."> that is


2010-03-09
[16:02:00.0000] <annevk>
Hixie, how does the UI for peer-to-peer work?

[16:14:00.0000] <Hixie>
annevk: you mean the UA UI?

[16:14:01.0000] <Hixie>
there is no UA UI defined for peer-to-peer... it would be the script that sets that up

[16:16:00.0000] <annevk>
yeah

[16:16:01.0000] <annevk>
shouldn't I agree to connect my computer to some computer elsewhere?

[16:17:00.0000] <annevk>
maybe the implications are not too bad

[16:22:00.0000] <Hixie>
annevk: it doesn't seem to be any different than XHR or WebSocket

[16:51:00.0000] <othermaciej>
preventing exfiltration isn't really practical, so there's probably not much more reason to worry about peer-to-peer

[17:16:00.0000] <MikeSmithX>
the more interesting number for rb would be to measure how many sites that use ruby have rb as part of their ruby 

[17:52:00.0000] <MikeSmith>
Hixie: so it would seem we are probably going to need for somebody to write an actual spec for SRT

[18:09:00.0000] <Hixie>
MikeSmith: yeah i figured as much. i can do that at some point, it's a pretty simple format. I wrote a draft already, actually, it's in some e-mail somewhere

[18:11:00.0000] <MikeSmith>
OK, that's good news at least

[18:11:01.0000] <MikeSmith>
/me looks at Hixie's "valid non-empty URL" change

[18:13:00.0000] <MikeSmith>
ah hey, yeah, you already had valid non-empty URL as a datatype for other attribute values

[18:13:01.0000] <MikeSmith>
/me goes to look at v.nu schema

[18:15:00.0000] <Hixie>
really?

[18:16:00.0000] <MikeSmith>
oh, maybe not

[18:17:00.0000] <MikeSmith>
if not, this might require adding a new datatype to the v.nu HTML5 datatype library

[18:18:00.0000] <MikeSmith>
hmm, yeah, @manifest was previously allowed to be empty but must be non-empty now too

[18:18:01.0000] <MikeSmith>
etc.

[18:21:00.0000] <MikeSmith>
yeah, it seems this will require a new datatype

[18:21:01.0000] <MikeSmith>
/me files a v.nu bug

[18:22:00.0000] <MikeSmith>
I wish there were some other contributors working on v.nu

[18:22:01.0000] <MikeSmith>
this would be an easy bug with which to get familiar with that part of the code

[18:23:00.0000] <MikeSmith>
Hixie: btw, about http://www.w3.org/Bugs/Public/show_bug.cgi?id=9198

[18:23:01.0000] <MikeSmith>
since hspace and vspace are already now allowed on embed, I assume that bug must be asking that they be explicitly listed in the obsolete-and-nonconforming section

[18:23:02.0000] <MikeSmith>
*not allowed

[18:25:00.0000] <MikeSmith>
/me wonders if he can manage to get his spellchecker to forget the words "now" and "not" so that they will always get highlighted and I can get alerted to double-check that I'm using the right one

[22:01:00.0000] <MikeSmith>
Hixie: I notice there was never any response on the hybi list to John Fallows's "WebSocket API - proposal for binary frame support in JavaScript"

[22:02:00.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-html-comments/2010Jan/0009.html

[22:02:01.0000] <MikeSmith>
oh, sorry

[22:02:02.0000] <MikeSmith>
I see it was on the public-html-comments list

[22:02:03.0000] <MikeSmith>
(which makes sense because it's about the API..)

[22:03:00.0000] <MikeSmith>
Hixie: should I maybe open a bugzilla bug for it?

[22:03:01.0000] <MikeSmith>
/me really does need to get some food, will step out for a bit

[22:04:00.0000] <MikeSmith>
hmm, I see that you did actually respond.. I missed the reply because I was looking in the online archives

[22:05:00.0000] <MikeSmith>
and it got split across the month boundary

[22:05:01.0000] <wirepair>
mike, you have any more information on that proposal for binary frame support?

[22:05:02.0000] <wirepair>
that seems kinda iffy

[22:05:03.0000] <MikeSmith>
wirepair: I know no more than what John posted in the message above

[22:06:00.0000] <wirepair>
gotcha

[22:06:01.0000] <wirepair>
ps. don't catch norovirus

[22:06:02.0000] <MikeSmith>
hai

[22:06:03.0000] <wirepair>
i've been out for 3 days ;<

[22:06:04.0000] <wirepair>
things killing me

[22:06:05.0000] <MikeSmith>
I gargle podine iodine to ward off the bugs

[22:07:00.0000] <wirepair>
good deal. heh

[22:07:01.0000] <MikeSmith>
the brown stuff with the hippo character

[22:08:00.0000] <MikeSmith>
/me grabs his umbrella and heads out

[22:55:00.0000] <othermaciej>
MikeSmith: shouldn't it be on the webapi wg list?

[00:22:00.0000] <othermaciej>
there have been 666 bugs filed in the HTML5 spec components since TPAC

[00:25:00.0000] <annevk>
nice number

[00:28:00.0000] <annevk>
with the peer-to-peer thingie it seems that everyone who gets your address can just connect with you

[00:29:00.0000] <annevk>
depending on what chat system is running on the site you might get tricked into giving information away to third parties you did not mean to

[00:29:01.0000] <othermaciej>
I just calculated our total incoming and outgoing bugs and issues since TPAC

[00:29:02.0000] <othermaciej>
neither number is really trending to 0

[00:30:00.0000] <othermaciej>
where is the peer-to-peer thingie spec'd?

[00:30:01.0000] <annevk>
http://html5.org/tools/web-apps-tracker?from=4828&to=4829 is all there is

[00:30:02.0000] <annevk>
so there is some speculation from me is to how it works since it's not all defined

[00:34:00.0000] <othermaciej>
the hard part of peer-to-peer is discovering peers

[00:34:01.0000] <othermaciej>
most systems start with a well-known host that does the initial coordination

[00:35:00.0000] <othermaciej>
also not clear what protocol these proposed interfaces would use

[00:35:01.0000] <othermaciej>
also this probably won't work very well for users behind firewalls if it works in any kind of obvious way

[00:36:00.0000] <othermaciej>
to get to users on network segments using NAT or otherwise firewalled, you need to tunnel through a central server

[00:36:01.0000] <othermaciej>
like I gather Opera Unite does

[00:38:00.0000] <annevk>
the coordination goes through some kind of server in typical usage (non-typical would be to coordinate it over the phone)

[00:39:00.0000] <othermaciej>
telling someone your IP address over the phone is not an 80% use case

[00:42:00.0000] <annevk>
yeah

[00:46:00.0000] <othermaciej>
not even sure if peer-to-peer in general is an 80% use case

[00:46:01.0000] <othermaciej>
I'd like to see WebSocket implemented and deployed first

[00:46:02.0000] <othermaciej>
which reminds me that I need to review the new handshake and also get abarth to do so

[00:47:00.0000] <annevk>
IM is not 80%?

[00:50:00.0000] <othermaciej>
you don't need peer-to-peer for IM

[00:51:00.0000] <othermaciej>
Jabber goes through a server, so does IRC, to the best of my knowledge AIM/MSN/ICQ/etc also go through a central server

[00:52:00.0000] <othermaciej>
the only true peer-to-peer chat I get on my Mac, as far as I know, is Bonjour chat in iChat, which is only a neat trick because it autodiscovers people on the local network segment

[00:52:01.0000] <othermaciej>
but the interfaces you linked don't seem to have any provision for autodiscovery

[00:53:00.0000] <othermaciej>
autodiscovery is the hard part

[00:55:00.0000] <othermaciej>
(IRC does have the possibility of DCC which is directly peer-to-peer but it doesn't seem to be used that often)

[00:58:00.0000] <othermaciej>
I would guess audio or video chat probably does go peer-to-peer usually

[00:58:01.0000] <othermaciej>
but not always

[00:59:00.0000] <othermaciej>
there is group audio chat like Ventrillo which definitely uses a central server

[01:01:00.0000] <annevk>
mkay

[01:05:00.0000] <annevk>
autodiscovery would be nice

[01:08:00.0000] <virtuelv>
othermaciej: in terms of Opera unite it does one of two things

[01:08:01.0000] <virtuelv>
router configuration (uPnP) and/or a central proxy

[01:09:00.0000] <othermaciej>
oh, I think MobileMe also has name-based peer-to-peer WAN autodiscovery by name through a central server

[01:09:01.0000] <virtuelv>
there is also local network discovery

[01:09:02.0000] <othermaciej>
discovery is the hard part

[01:10:00.0000] <roc>
/me spams webkit-dev

[01:10:01.0000] <othermaciej>
and letting a web page advertise your host for autodiscovery without permission is the potentially scary security bit

[01:13:00.0000] <annevk>
ohuh

[01:13:01.0000] <Hixie>
othermaciej: iChat is peer-to-peer also once you've established the link, as i understand it. Certainly Skype is.

[01:13:02.0000] <Hixie>
annevk: there's no reason the address couldn't be unique per PeerToPeerServer -- indeed, that seems like a good idea

[01:13:03.0000] <othermaciej>
Hixie: iChat is an app not a protocol - which protocol do you mean?

[01:14:00.0000] <Hixie>
othermaciej: the one used for video chat, sorry

[01:14:01.0000] <othermaciej>
I believe its use of AIM and Jabber is not peer-to-peer, but I believe audio and video chat and Bonjour chat are peer-to-peer

[01:14:02.0000] <Hixie>
sounds right

[01:14:03.0000] <othermaciej>
(audio and video peer discovery being done via the central server initially)

[01:14:04.0000] <virtuelv>
skype uses STUN, afaict

[01:14:05.0000] <virtuelv>
http://www.ietf.org/rfc/rfc3489.txt

[01:14:06.0000] <Hixie>
video conferencing is the main use case i'm imagining this stuff being used for

[01:14:07.0000] <Hixie>
anyway this is all highly experimental so far

[01:15:00.0000] <virtuelv>
plus http://tools.ietf.org/html/draft-ietf-behave-turn-16

[01:15:01.0000] <Hixie>
i'm just poking around with proposals because it comes up every now and then

[01:15:02.0000] <virtuelv>
http://adamfisk.wordpress.com/2006/08/16/cringely-skype-open-infrastructure/

[01:16:00.0000] <annevk>
Hixie, yeah, I was thinking it had to be tied to the origin or something

[01:16:01.0000] <annevk>
but unique would be even better

[01:17:00.0000] <Hixie>
annevk: i hope that we don't have to define the network-level stuff but can instead just defer to a separate spec, the way the websocket api defers to a websocket protocol spec, except without me having to write the network side :-)

[01:18:00.0000] <annevk>
I wonder what the quality of such a spec would be

[01:18:01.0000] <Philip`>
/me thought the point of Skype was that it tried loads of connection methods, which is necessary in order to make something that works for all users

[01:19:00.0000] <Philip`>
e.g. if one user can accept external network connections, they get used as a proxy between users that can't

[01:19:01.0000] <othermaciej>
like I said - discovery is the hard part

[01:20:00.0000] <Philip`>
(so it'll silently use up your bandwidth)

[01:20:01.0000] <Philip`>
and it does various tricks like trying to use port 80 to get around firewalls

[01:20:02.0000] <Philip`>
(as well as all of the NAT traversal)

[01:22:00.0000] <Philip`>
This isn't the discovery part, it's the part where you figure out how to send traffic to someone after you've discovered them

[01:30:00.0000] <othermaciej>
when I say "discovery" I mean including the ability to actually establish a connection to the peer

[01:30:01.0000] <othermaciej>
I guess that's really discovery + routing

[01:34:00.0000] <Philip`>
Ah

[01:46:00.0000] <Hixie>
othermaciej: the change proposal link in ISSUE-66 in the issue-status list links to something other than it claims

[01:46:01.0000] <othermaciej>
Hixie: probably my falt, let me see

[01:46:02.0000] <othermaciej>
Hixie: ugh - it looks like the forwarding links from the old wiki are not working

[01:46:03.0000] <othermaciej>
MikeSmithX: are you around?

[01:47:00.0000] <othermaciej>
Hixie: probably all other links to the esw wiki are broken

[01:47:01.0000] <MikeSmithX>
othermaciej: I am now

[01:47:02.0000] <othermaciej>
MikeSmithX: esw wiki doesn't seem to be forwarding to the new wiki

[01:47:03.0000] <othermaciej>
or rather, all old links to it forward to the main page

[01:48:00.0000] <othermaciej>
is that expected?

[01:50:00.0000] <MikeSmithX>
it's expected but we are working on fixing

[01:50:01.0000] <MikeSmithX>
iy

[01:50:02.0000] <MikeSmithX>
it

[01:50:03.0000] <othermaciej>
ok

[01:50:04.0000] <othermaciej>
I should probably fix all the links to the old wiki on the issue status page

[01:52:00.0000] <othermaciej>
would it work to just global replace http://esw.w3.org/topic/HTML/ with http://www.w3.org/html/wg/wiki/

[01:52:01.0000] <MikeSmithX>
yeah, it should

[01:53:00.0000] <othermaciej>
Hixie: fix't

[01:56:00.0000] <Hixie>
thanks

[01:57:00.0000] <othermaciej>
MikeSmith: by the way, your "age of oldest unresolved bug" number doesn't look so hot (from my recent email) - I hope you get a chance to do a pass over H:TML bugs soon

[01:58:00.0000] <MikeSmith>
othermaciej: yeah, there are a couple of old ones I know I can close out

[01:58:01.0000] <MikeSmith>
because I filed them myself

[01:58:02.0000] <othermaciej>
heh

[01:59:00.0000] <othermaciej>
I will forgive you if you skip the standard boilerplate on self-filed bugs (assuming no one else is going to be looking to escalate them or anything)

[02:00:00.0000] <MikeSmith>
ok

[02:02:00.0000] <othermaciej>
this article sounds amazingly authoritative given that the author clearly has no idea what he's talking about: http://www.betanews.com/article/Latest-HTML5-working-draft-published-despite-claims-of-sabotage/1268084800

[02:09:00.0000] <MikeSmith>
aha, authoritative indeed

[02:09:01.0000] <MikeSmith>
'By "accessibility," Le Hégaret was using a keyword in the argument in favor of breaking Canvas out from HTML5'

[02:09:02.0000] <othermaciej>
I am just amazed how may simple declarative statements there are false

[02:10:00.0000] <MikeSmith>
sharp character, this Scott M. Fulton, III is

[02:10:01.0000] <MikeSmith>
he understands all the secret code words

[02:10:02.0000] <Hixie>
i couldn't read it all the way through, it was too painful

[02:17:00.0000] <nessy>
what a twisted world view

[02:17:01.0000] <nessy>
some people can make plots out of anything!

[02:26:00.0000] <annevk>
oh hey, my blog is linked

[02:26:01.0000] <annevk>
good times

[02:27:00.0000] <hsivonen>
http://twitter.com/jdowdell/status/10187314697

[02:28:00.0000] <hsivonen>
I guess the "range of expectation" is narrow

[02:29:00.0000] <Hixie>
i'm sure glad people are tweeting rather than filing bugs

[02:29:01.0000] <annevk>
the more important question is of course, do you care about SWF?

[02:29:02.0000] <Hixie>
makes my life a lot easier

[02:30:00.0000] <hsivonen>
annevk: at least on the desktop market, it would be considered a bug if CSS and SWF colors didn't look the same for a given RGB value

[02:31:00.0000] <hsivonen>
on a related note, neither swfdec nor gnash manages to show me any .swf files on Ubuntu (64-bit)

[02:32:00.0000] <hsivonen>
the context menus suggests the plug-ins do load

[02:43:00.0000] <othermaciej>
one of the top two reasons we can't colormatch images in HTML in Safari is because SWF doesn't colormatch vector drawing

[02:45:00.0000] <othermaciej>
I almost @replied to jdowdell but I regained my senses

[02:45:01.0000] <othermaciej>
because one of the top two reasons we can't colormatch everything in Safari is SWFs!

[02:59:00.0000] <MikeSmith>
hsivonen: thanks for the comment on the iri-ref v.nu bug I raise

[02:59:01.0000] <MikeSmith>
*raised

[03:00:00.0000] <MikeSmith>
hsivonen: the solution of changing the iri-ref datatype to never accept the empty string and changing the allowed value for attributes that do accept either the empty string or iri-ref does sound like the best way to handle it

[03:00:01.0000] <MikeSmith>
shall I go ahead and make that change?

[03:10:00.0000] <MikeSmith>
othermaciej, annevk - I created a new bugzilla HTML WG component for the HTML5 differences from HTML4 doc

[03:29:00.0000] <hsivonen>
MikeSmith: go ahead

[03:30:00.0000] <hsivonen>
I'm getting really close to point where the HTML5 parser in Gecko can be turned on by default

[03:30:01.0000] <hsivonen>
I hope to have some cycles for V.nu hacking when that's done

[03:32:00.0000] <MikeSmith>
hsivonen: cool

[06:24:00.0000] <Philip`>
"Exactly what is the purpose for this change proposal? Are you saying that you want to change the section in the specification?" - I thought it was fairly clear that the purpose of a change proposal is to propose a change to the specification

[06:27:00.0000] <Philip`>
(Maybe the confusion is that Shelley thinks it necessarily reflects the personal opinion of the person who is writing the proposal, rather than just being intended as an option for the WG to consider and potentially vote on)

[06:30:00.0000] <hsivonen>
Did Safari 4 change WebKit's doctype sniffing inspired by HTML5 as Leif claims?

[06:41:00.0000] <annevk>
i somewhat doubt it

[06:44:00.0000] <annevk>
doesn't look like it looking at trac

[06:44:01.0000] <gsnedders>
hsivonen: hyatt rewrote it around the Saf4 time, IIRC

[06:44:02.0000] <gsnedders>
(hyatt definitely rewrote it to match HTML 5 sometime)

[06:44:03.0000] <annevk>
really?

[06:45:00.0000] <annevk>
the doctype string file hasn't been touched for four years

[06:45:01.0000] <annevk>
and to match HTML5 that would need changing

[06:45:02.0000] <annevk>
furthermore the HTML parser code doesn't seem to be changed around doctypes either

[06:46:00.0000] <annevk>
meanwhile I noticed Safari implements <hgroup> among other new elements

[06:46:01.0000] <annevk>
well, WebKit

[06:48:00.0000] <annevk>
gsnedders, I can't find evidence for your claim

[06:48:01.0000] <annevk>
would be nice though if he did

[06:48:02.0000] <gsnedders>
annevk: He certainly changed it a fair bit to get it pretty close to HTML 5

[06:49:00.0000] <annevk>
gsnedders, not per SVN

[06:49:01.0000] <gsnedders>
https://bugs.webkit.org/show_bug.cgi?id=17639 is one bug

[06:50:00.0000] <annevk>
hmm, that bug is marked WFM

[06:50:01.0000] <annevk>
and the file it talks about hasn't been touched for four years

[06:50:02.0000] <gsnedders>
Indeed

[06:51:00.0000] <gsnedders>
Was the list in HTML 5 not originally based upon WebKit?

[06:51:01.0000] <annevk>
yes

[07:07:00.0000] <Philip`>
Shelley's view of change proposals seems to make compromise impossible, if everybody only proposes what they personally think is best (given that the result must be the acceptance of exactly one change proposal)

[07:09:00.0000] <annevk>
/me wonders if everyone has the experience outlined here: http://camendesign.com/blog/choice

[07:10:00.0000] <annevk>
balot screen kind of fails if that's true

[07:11:00.0000] <gsnedders>
annevk: Apparently happens when you upgrade to IE8 and ballot screen at the same time

[07:13:00.0000] <meledin>
Those pictures are pure win

[07:14:00.0000] <Philip`>
At least the IE8 upgrade won't happen on Win7 (since it ships with that version)

[07:15:00.0000] <meledin>
Even if it

[07:15:01.0000] <meledin>
is accidental, I'd have to give props to Microsoft for ingeniously circumventing the ballot

[07:18:00.0000] <Philip`>
Forcing users to respond to an unsolicited prompt that expects them to make an important decision is always going to be bad UI design

[07:18:01.0000] <Philip`>
and I don't see an obvious way that Microsoft could have made it much better

[07:21:00.0000] <Philip`>
(It seems they only do these prompts for IE8 (which is important for security and compatibility, and is a major user-visible change so it shouldn't be done silently, so it's worth annoying the user for it) and browser choice (which is legally required), and otherwise successfully avoid it)

[07:22:00.0000] <hsivonen>
considering that the ballot is legally required, I'm surprised they haven't done more code review and QA on it

[07:42:00.0000] <gsnedders>
Anyone know of any Python WebIDL parser?

[07:46:00.0000] <Philip`>
gsnedders: http://suika.fam.cx/www/markup/html/whatpm/readme has a Perl one, which is close

[07:46:01.0000] <gsnedders>
/me slaps Philip` 

[07:47:00.0000] <Philip`>
:-(

[07:47:01.0000] <Philip`>
/me often chooses what language to use based on available libraries rather than on properties of the language itself

[07:47:02.0000] <gsnedders>
So, should I go for one and use GHC, and use that as a Python extension? :P

[07:47:03.0000] <gsnedders>
*use GHC for it

[07:48:00.0000] <gsnedders>
Philip`: That decision has already been made

[07:48:01.0000] <Philip`>
Do you mean write your own?

[07:48:02.0000] <Philip`>
Also, do you mean GHC as in Haskell?

[07:48:03.0000] <gsnedders>
Yes

[07:49:00.0000] <gsnedders>
http://hackage.haskell.org/package/webidl

[07:50:00.0000] <Philip`>
That sounds more complex than using one written in a sensible language like Perl, and making it emit the parse tree as JSON so you can load it into Python

[07:50:01.0000] <gsnedders>
Maybe I should just rewrite Anolis in Haskell :P

[07:50:02.0000] <Philip`>
Alternatively, maybe you shouldn't :-p

[07:52:00.0000] <Philip`>
In any case, it's probably best to treat the IDL parser as a standalone tool that emits an easy-to-parse text stream, instead of tightly coupling your Python code to it

[07:56:00.0000] <boblet>
There’s a new doctor in the house — me ;-)

[07:56:01.0000] <boblet>
http://html5doctor.com/i-b-em-strong-element/

[07:56:02.0000] <boblet>
/me really hopes no one spots an obvious mistake in like 5 seconds (gulp)

[07:57:00.0000] <paul_irish_>
very nice post.

[07:57:01.0000] <gsnedders>
Philip`: That means trying to write Perl ;P

[07:58:00.0000] <boblet>
paul_irish_: thanks! your powers of speed reading are … impressive!

[07:58:01.0000] <Philip`>
"i-b-em" - that sounds familiar

[07:58:02.0000] <paul_irish_>
haha

[07:58:03.0000] <boblet>
hehe

[07:59:00.0000] <gsnedders>
boblet: Would the examples for the b element not normally be ::first-line?

[07:59:01.0000] <boblet>
it just worked out that way, honest guv’nor!

[07:59:02.0000] <Philip`>
gsnedders: Probably only a little bit of Perl

[08:01:00.0000] <MikeSmith>
boblet: great stuff

[08:01:01.0000] <boblet>
gsnedders: one would, and the other wouldn’t — the :first-line example is the ”use :first-line not b to do this” one

[08:01:02.0000] <boblet>
thanks for your help MikeSmith 

[08:01:03.0000] <MikeSmith>
boblet: I especially like that way you curl your capital letter Is

[08:01:04.0000] <MikeSmith>
or at least that one I

[08:02:00.0000] <boblet>
I am but your humble padawan ;-)

[08:02:01.0000] <MikeSmith>
do more of that

[08:02:02.0000] <MikeSmith>
oh my god

[08:02:03.0000] <MikeSmith>
that O outdoes the I

[08:02:04.0000] <MikeSmith>
that is audacious O

[08:02:05.0000] <MikeSmith>
that is like a "Story of O" O

[08:02:06.0000] <boblet>
versals in the house ;-) @font-face of course

[08:03:00.0000] <MikeSmith>
boblet: in future articles, you :first-letter examples should just successively use bigger and bigger letters, until the get up to 320px or so

[08:04:00.0000] <gsnedders>
boblet: But do you really often do such things with less than the opening line? My memory could alwasy be wrong… :P

[08:05:00.0000] <boblet>
MikeSmith: click through to the example page — it makes more sense there (the “O” is the first drop cap, the “I” is the size of subsequent drop caps ;-) )

[08:05:01.0000] <MikeSmith>
blb

[08:05:02.0000] <MikeSmith>
boblet: sorry, I got distracted

[08:05:03.0000] <boblet>
gsnedders: in traditional typography the first word or phrase is common

[08:06:00.0000] <MikeSmith>
reading "Story of O"

[08:06:01.0000] <boblet>
traditional = scribes in brown habits, early years of the printing press etc

[08:07:00.0000] <gsnedders>
Yet of course it remained primarily stylisitc with print

[08:08:00.0000] <gsnedders>
And what I was taught was one very specific style of one printer

[08:08:01.0000] <MikeSmith>
boblet: anyway, seriously, very nice article

[08:08:02.0000] <gsnedders>
Anyone know off-hand what Ada's full name in "Ada or Ardor" is?

[08:09:00.0000] <boblet>
MikeSmith: I raise my mug of green tea to your good health :)

[08:09:01.0000] <gsnedders>
(for some obscure memonic reason)

[08:09:02.0000] <MikeSmith>
boblet: you have set a higher bar for the other HTML doctors

[08:09:03.0000] <MikeSmith>
you have made them look like HTML interns

[08:10:00.0000] <MikeSmith>
or HTML candy-stripe nurses

[08:10:01.0000] <gsnedders>
/me cheats and takes advantage of the fact that the text is available online

[08:10:02.0000] <gsnedders>
"A girl was born on July 21, 1872, at Ardis, her putative father's seat in Ladore County, and for some obscure mnemonic reason was registered as Adelaida.

[08:10:03.0000] <MikeSmith>
http://html5candystripenurses.com/

[08:10:04.0000] <boblet>
gsnedders: I’ll have to ask John from ILT about it

[08:11:00.0000] <gsnedders>
ILT?

[08:11:01.0000] <boblet>
http://ilovetypography.com/

[08:12:00.0000] <boblet>
MikeSmith: you’re too kind. But yeah I worry I may have got a little carried away. Hopefully they’ll forgive an enthusiastic locum

[08:12:01.0000] <Philip`>
http://ihatetypography.com/ - someone needs to register that

[08:12:02.0000] <boblet>
/me is very disappointed that nurses site isn’t registered

[08:13:00.0000] <gsnedders>
/me hates the opening chapters of Ada

[08:20:00.0000] <lazni>
in windows 7, ie8 still pops up that modal dialog at first load

[08:22:00.0000] <Philip`>
The upgrade-from-IE7 one?

[08:24:00.0000] <lazni>
yes, "Welcome to IE8"

[09:10:00.0000] <AryehGregor>
Hixie, could you say that if a URL resolves to the current resource, and the user agent knows that a text/html response will cause an error, it can decline to fetch the resource even if it's not actually in cache, on the assumption that it will be text/html?  (Similar for XML MIME types.)

[09:12:00.0000] <AryehGregor>
As for validation, surely something somewhere already says that script/stylesheet/etc. links must point to documents of the correct type.  E.g., if <script src=""></script> points to a text/html file, then that violates the sentence "If the language is not that described by "text/javascript", then the type attribute must be present, as described below."

[09:12:01.0000] <AryehGregor>
So a validator could recognize that "" is not a text/javascript resource and raise an error.

[09:15:00.0000] <Philip`>
AryehGregor: Some people (like hsivonen, I think) don't like the validity of a document depending on external resources

[09:16:00.0000] <Philip`>
It makes it really hard to check that a document is valid and that it stays valid

[09:16:01.0000] <AryehGregor>
You can't check all validity requirements anyway.

[09:16:02.0000] <Philip`>
(and you'd get different results when copying-and-pasting into a text input box vs pointing at the URL)

[09:16:03.0000] <AryehGregor>
Hmm, true.

[09:16:04.0000] <Philip`>
That's not a reason to add more requirements that can't be checked

[09:17:00.0000] <Philip`>
particularly since it's a lot of complexity for something that nobody is going to do on purpose

[09:17:01.0000] <AryehGregor>
The requirement already exists, I'm not suggesting it be added.  :)

[09:17:02.0000] <Philip`>
s/particularly since/Also,/

[09:17:03.0000] <Philip`>
Oh, okay

[09:17:04.0000] <AryehGregor>
Validators that don't want to depend on whether you copy-paste or point at the URL can implement only special cases of this check -- like saying it's an error only if the URL resolves to the current document, but HTML is the wrong type for the link.

[09:18:00.0000] <AryehGregor>
Which is the current case.

[09:19:00.0000] <Philip`>
Actually, <script src=""> doesn't point to a text/html file

[09:20:00.0000] <Philip`>
because (per the spec) it doesn't resolve to anything

[09:20:01.0000] <AryehGregor>
It doesn't?

[09:20:02.0000] <Philip`>
It hasn't done for at least 16 hours

[09:21:00.0000] <Philip`>
So I think that's not a problematic case, because it points at nothing and is clearly an error

[09:23:00.0000] <Philip`>
The problematic case was like <link rel="stylesheet index" href="">

[09:23:01.0000] <Philip`>
where the stylesheet link doesn't point to anything, while the index link points to the base URL

[09:24:00.0000] <AryehGregor>
Well, I was making suggestions to Hixie based on the previous version of the spec.

[09:24:01.0000] <AryehGregor>
As an alternative fix.

[09:25:00.0000] <Philip`>
The previous version of the spec would download the current page again if you wrote <script src="">

[09:25:01.0000] <Philip`>
which was the problem

[09:25:02.0000] <Philip`>
so the fix would have to involve changing it to not download the current page again

[09:25:03.0000] <Philip`>
and so it couldn't then define validity in terms of the downloaded resource

[09:26:00.0000] <Philip`>
although I think I forgot to read what you actually said

[09:27:00.0000] <AryehGregor>
Right.  My suggestion was that the spec say that if a URL is known to resolve to the current resource, then if that URL is used for something that's not supposed to be HTML, the UA can raise an error without bothering to actually retrieve the resource.

[09:27:01.0000] <AryehGregor>
Or just ignore it, or something.

[09:28:00.0000] <AryehGregor>
And validators can say it's an error if they know it resolves to the current resource.

[09:28:01.0000] <AryehGregor>
I guess that's messier than just adding some special cases, but it seems less hackish.

[09:29:00.0000] <Philip`>
That seems like a lot of complexity for something that nobody is going to do on purpose

[09:31:00.0000] <Philip`>
It doesn't seem non-hackish to me, because it's still going to do weird things in hypothetical cases like a POST response page saying <script src=""> where the GET returns a different content-type

[09:32:00.0000] <Philip`>
and it'll be hard for authors to predict behaviour because it'll depend on whether the UA cached enough information about the page

[09:35:00.0000] <AryehGregor>
I guess.

[11:00:00.0000] <Dashiva>
It's so easy to forget that we are currently living in an era of low public-html traffic

[11:04:00.0000] <knowtheory>
what?

[11:13:00.0000] <Philip`>
Dashiva: We are?

[11:13:01.0000] <Philip`>
Monthly post counts: http://chart.apis.google.com/chart?cht=lc&chd=t:1773,1307,1099,1313,1258,576,398,491,324,307,441,370,802,767,462,455,965,515,301,592,263,694,818,734,288,657,857,954,1477,1220,1144,681,518,1459,908&chds=0,2000&chs=300x150

[11:15:00.0000] <othermaciej>
Philip`: what is that counting?

[11:15:01.0000] <Dashiva>
The graph is deceitful, it doesn't compensate for christmas

[11:16:00.0000] <Philip`>
othermaciej: Monthly posts :-)

[11:16:01.0000] <Philip`>
The numbers from http://lists.w3.org/Archives/Public/public-html/ in particular

[11:16:02.0000] <othermaciej>
monthly posts to what?

[11:16:03.0000] <Philip`>
(excluding the periods that aren't months)

[11:16:04.0000] <othermaciej>
public-html?

[11:16:05.0000] <Philip`>
public-html

[11:19:00.0000] <knowtheory>
Dashiva: how is it decietful?

[11:38:00.0000] <othermaciej>
since we're throwing stats around - I made some monthly bug stats: http://spreadsheets.google.com/ccc?key=0AoCAfo_LQ5_kdFFWWmpCMWxsLUN2TW9VYi1uNEJGenc&hl=en

[11:41:00.0000] <othermaciej>
it's interesting that find rate and fix rate seem to roughly track each other, yet the resulting net incoming rate seems to oscillate quite a bit

[14:00:00.0000] <Hixie>
AryehGregor: that would fail to handle this http://www.hixie.ch/tests/evil/html/external-resources/001.test

[14:00:01.0000] <Hixie>
AryehGregor: (note no content-type headers)

[14:01:00.0000] <Hixie>
AryehGregor: (also note that that test is now out of date)


2010-03-10
[17:18:00.0000] <Hixie>
apparently shelley's reading comprehension isn't very good. In response to the e-mail where I explain my opinion is irrelevant, she asks for my opinion again

[17:18:01.0000] <Hixie>
I wonder what her obsession with my opinion today is about, given that she usually thinks my opinion is worthless

[17:21:00.0000] <Dashiva>
Hixie: She needs to know your opinion to make sure she doesn't agree with it? :P

[17:52:00.0000] <Hixie>
Dashiva: that is sadly far too consistent with previous behaviour

[21:15:00.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-webapps/2010JanMar/0792.html is interesting

[21:16:00.0000] <MikeSmith>
Charles Pritchard, "Advanced Pointer Events: Use Cases"

[21:46:00.0000] <MikeSmith>
boblet: hey, I just realized something, which is that I'm not going to be in town on March 27

[21:46:01.0000] <MikeSmith>
off-by-one error

[21:46:02.0000] <MikeSmith>
the one being one month

[21:46:03.0000] <MikeSmith>
I had been thinking we'd talked about April 27, not March 27

[22:24:00.0000] <hsivonen>
have I understood correctly that geopriv would involve the user making privacy choices in the browser chrome but the browsers would be unable to enforce the settings and instead the user would have to trust the site to be benevolent and obedient?

[22:26:00.0000] <hsivonen>
/me has read www-tag again

[22:58:00.0000] <boblet>
MikeSmith: that sucks

[23:02:00.0000] <Hixie>
hsivonen: yes

[23:29:00.0000] <MikeSmith>
bb

[23:30:00.0000] <MikeSmith>
boblet: I got a copy of hatano-san's book in the mail today

[23:30:01.0000] <MikeSmith>
beefy

[23:30:02.0000] <MikeSmith>
550+ pages

[23:45:00.0000] <boblet>
MikeSmith: hey, great stuff! He must be relieved

[23:45:01.0000] <MikeSmith>
yeah

[23:46:00.0000] <boblet>
what’s it like? manual style? explanation? examples? tutorials? practical use cases?

[23:46:01.0000] <boblet>
I’m guessing more explaining the spec style with made-up examples

[23:46:02.0000] <MikeSmith>
boblet: reference-like

[23:47:00.0000] <boblet>
that fits with his site

[23:48:00.0000] <MikeSmith>
HTML5 マークアプ ガイドブク

[23:48:01.0000] <MikeSmith>
マークアップ, make that

[23:49:00.0000] <MikeSmith>
hmm, which sounds very unnatural as a transliteration

[23:49:01.0000] <boblet>
man 550 pages. what a machine

[23:49:02.0000] <MikeSmith>
seems like it would better just be マーカップ

[23:49:03.0000] <MikeSmith>
but oh well

[23:50:00.0000] <boblet>
c’mon you know Japanese transliteration ignores naturalness

[23:50:01.0000] <MikeSmith>
heh

[23:52:00.0000] <Hixie>
MikeSmith: i got the html5 books btw

[23:52:01.0000] <Hixie>
can't read anything but the acronyms, but... :-)

[23:52:02.0000] <MikeSmith>
Hixie: ah, cool

[23:53:00.0000] <boblet>
Hixie: both the Japanese ones? nice :)

[23:53:01.0000] <MikeSmith>
you may have a copy of Hatano-san's book on the way to you as well, not sure

[23:53:02.0000] <MikeSmith>
oh, "books"

[23:53:03.0000] <Hixie>
boblet: yeah

[23:53:04.0000] <MikeSmith>
great

[23:54:00.0000] <boblet>
yeah we were telling Shiraishi-san he should definitely send you a copy

[23:54:01.0000] <boblet>
good to see Japan showing up all the English authors

[23:54:02.0000] <MikeSmith>
both Shiraishi-san and Hatano-san are speaking at the Google DevFest event here tomorrow

[23:54:03.0000] <boblet>
(or sticking to publisher schedules regardless of the spec ;-) )

[23:56:00.0000] <Hixie>
:-)

[23:59:00.0000] <MikeSmith>
Futomi gots his portrait picture on the About the Author page of his book 

[00:00:00.0000] <MikeSmith>
he looks kind of pissed off, like he's getting ready to kick somebody's ass

[00:00:01.0000] <boblet>
was prolly meant to be his serious face (unless writing the book really did take a toll ;-)

[00:01:00.0000] <MikeSmith>
I will say that both dudes should have included an prominent URL for an Errata/Updates page

[00:01:01.0000] <boblet>
MikeSmith: why?

[00:01:02.0000] <boblet>
/me rofls

[00:02:00.0000] <boblet>
there won’t be changes will there?

[00:02:01.0000] <boblet>
heh

[00:03:00.0000] <MikeSmith>
hopefully the first printing will sell out quickly and they can add a URL to the next print run

[00:03:01.0000] <boblet>
definitely seems like it’s the hot topic, so here’s hoping

[00:06:00.0000] <Hixie>
if we do end up using a non-versioned model in the future, we'll have to find some solution to handle referring to the latesrt version for books and the like

[00:07:00.0000] <Hixie>
maybe the w3c will just publish annual RECs that include whatever the latest implemented stuff is :-)

[00:11:00.0000] <nessy>
wishful thinking ;)

[01:28:00.0000] <gsnedders>
Hixie: Still awake?

[01:30:00.0000] <Hixie>
gsnedders: here

[01:32:00.0000] <gsnedders>
Hixie: So for Anolis 2 I was planning on changing the way TOC/num worked; Anolis 1 just gets all headers of a certain depth, but I was planning on changing it to only get children of a certain section (specified by class on either the sectioning content or heading content element). That seem reasonable?

[01:33:00.0000] <Hixie>
come again?

[01:33:01.0000] <gsnedders>
Anolis 1 you just have min-depth and max-depth that controls what is in the TOC, right?

[01:37:00.0000] <Hixie>
i wasn't aware there was even that much

[01:37:01.0000] <Hixie>
so long as it grabs all the headers except the ones marked notoc, i'm good

[01:38:00.0000] <Hixie>
if it does that by looking at the notoc class, or some other magic like determining automatically which should have had notoc, i don't mind

[01:38:01.0000] <gsnedders>
But how does it know not to number the top no-toc, and start level one at its children?

[01:39:00.0000] <Hixie>
magic?

[01:39:01.0000] <gsnedders>
As normally that would end up as 0.1 as the first child

[01:39:02.0000] <Hixie>
i don't understand the problem you're trying to solve :-)

[01:40:00.0000] <gsnedders>
Oh, just trying not to remove features inherieted from the PP, and not wanting to add magic

[01:41:00.0000] <Hixie>
well so long as it doesn't change, i don't really mind either way

[01:41:01.0000] <Hixie>
the output, i mean

[01:42:00.0000] <gsnedders>
Hixie: But if we just go with that, which proibably works in any case for everyone who uses it, can you add no-num no-toc to the hgroup in the whatwg header?

[01:42:01.0000] <gsnedders>
Hixie: And to the h1 in the w3c header?

[01:54:00.0000] <Hixie>
gsnedders: sure

[02:02:00.0000] <MikeSmith>
Andrey Kosyakov. Web Inspector: display list of active workers & support debugging with fake workers. https://bugs.w… http://a.gd/wk55771

[02:02:01.0000] <MikeSmith>
that sounds pretty cool

[05:07:00.0000] <hsivonen>
/me wonders if IE does the isindex magic if there's a valueless <input type=image> before the <input name=isindex>

[05:15:00.0000] <Lachy>
hsivonen, I can test it for you if you have a test page

[05:15:01.0000] <Lachy>
if you don't have IE available yourself

[05:16:00.0000] <MikeSmith>
hsivonen must have IE now

[05:16:01.0000] <MikeSmith>
Lachy: btw, where you been man

[05:16:02.0000] <MikeSmith>
Lachy: seems like you've not been around on #whatwg so much lately

[05:17:00.0000] <hsivonen>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/398

[05:18:00.0000] <hsivonen>
looks like Firefox 3.5 submits x and y in that case

[05:19:00.0000] <hsivonen>
looks like I was worried about dead code

[05:21:00.0000] <Lachy>
MikeSmith, I've been really busy for the past few weeks doing Carakan QA, getting ready for shipping 10.50. And since then, I've mostly just been taking a break from HTMLWG stuff

[05:22:00.0000] <Lachy>
hsivonen, in IE, if I enter "test" into the input and press enter, it just submits "test". If I click the image, it also submits the x and y coords.

[05:22:01.0000] <hsivonen>
Lachy: thanks

[05:24:00.0000] <boblet>
interesting. Last week Google Search cached pages used <b><mark>term</mark></b>, but now use nested b

[05:25:00.0000] <hsivonen>
Lachy: apparently Firefox counts the return as clicking type=image at 0,0 but IE doesn't

[05:51:00.0000] <gsnedders>
MikeSmith: We kidnapped Lachy and moved him to Sweden for a while.

[05:52:00.0000] <gsnedders>
MikeSmith: What he just said was really what we ordered him to say while we had him hostage

[05:53:00.0000] <Philip`>
If that was the case, why would you admit it?

[05:54:00.0000] <Philip`>
Perhaps what's really happened is that he has kidnapped you, and is ordering you to say the opposite in order to allay suspicions

[06:14:00.0000] <MikeSmith>
gsnedders: I wonder if Lachy loves Sweden as much as he loves Norway

[06:15:00.0000] <Lachy>
No. Norway is better

[06:16:00.0000] <MikeSmith>
heh

[06:16:01.0000] <Lachy>
well, at least Oslo is better than Linköping.

[06:16:02.0000] <MikeSmith>
<snort>

[06:16:03.0000] <Philip`>
At least it's easier to spell

[06:16:04.0000] <MikeSmith>
well, nothing can compare to the wonders of Oslo

[06:16:05.0000] <Dashiva>
Which wonders are those?

[06:17:00.0000] <Lachy>
Linköping isn't hard to spell, once you learn how to enter the ö character

[06:17:01.0000] <Lachy>
it is hard to pronounce, though

[06:18:00.0000] <Lachy>
it's not spelled in any sensible way for how it is pronounced.

[06:18:01.0000] <workmad3>
Lachy: my pronunciation of it would suck - I'm thinking of it as 'link o ping'

[06:19:00.0000] <MikeSmith>
Dashiva: the Stargate, just to name one highlight

[06:19:01.0000] <Dashiva>
Did they build that recently? I don't recall seeing it

[06:19:02.0000] <Philip`>
<link ping="...">

[06:19:03.0000] <MikeSmith>
Dashiva: it comes and goes

[06:19:04.0000] <Philip`>
Hmm, not supported in HTML5 :-(

[06:19:05.0000] <Lachy>
it's more like Lin-chirping, but, the 'ch' is softer sound than usual

[06:20:00.0000] <Lachy>
gsnedders can probably tell you the IPA for it

[06:20:01.0000] <workmad3>
Lachy: more like a 'sh' than a 'ch' perhaps?

[06:20:02.0000] <MikeSmith>
"LPI" is the IPA for it, I'm pretty sure

[06:20:03.0000] <Lachy>
sort of a cross between them, I think

[06:20:04.0000] <Philip`>
MikeSmith: Do they have one of those cool giant spinny things like in Contact, or is it just a boring shimmering hole?

[06:21:00.0000] <MikeSmith>
Philip`:  "shimmering hole" is more like it

[06:21:01.0000] <MikeSmith>
shimmering like a puddle of something

[06:21:02.0000] <MikeSmith>
with an oil slick of some kind on the surface

[06:22:00.0000] <MikeSmith>
or at least you hope it's just oil and not something worse

[06:22:01.0000] <MikeSmith>
seriously, the parts of Oslo where there are actual trees are wonderful

[06:23:00.0000] <MikeSmith>
like the part the runs along the river that's next to the Opera office

[06:23:01.0000] <asmodai>
not sure if this was shared yet http://mrdoob.com/lab/javascript/harmony

[06:31:00.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/www-international/2010JanMar/0116.html

[06:31:01.0000] <MikeSmith>
bravo Murakami-san

[06:35:00.0000] <annevk>
IE doesn't even support <rp>

[06:35:01.0000] <asmodai>
Hah, fucking awesome. Hadn't seen his reply yet.

[06:35:02.0000] <annevk>
UAs ought to always support ruby, screw fallback

[06:37:00.0000] <Dashiva>
annevk: Make UAs ignore bracket-type characters inside <ruby>?

[06:42:00.0000] <annevk>
huh?

[06:43:00.0000] <annevk>
I mean that UAs should always support ruby rendering so authors do not have to care about <rp> or fallback for ruby

[06:48:00.0000] <Dashiva>
Old browsers exist, we can't do much about that

[06:48:01.0000] <annevk>
MikeSmith, you still there?

[06:48:02.0000] <annevk>
/me wonders whether the <cite> element really changed

[06:48:03.0000] <MikeSmith>
annevk: yeah, here for a bit

[06:49:00.0000] <annevk>
HTML4 does not say much about it, though HTML5 does contradict a non-normative example that one of the editors of HTML4 thought was wrong

[06:49:01.0000] <annevk>
not really sure what to say about it

[06:49:02.0000] <MikeSmith>
I think <cite> is changed at least as far as the fact that the example in the HTML4 spec is no longer conformant

[06:49:03.0000] <MikeSmith>
the example that uses <cite> for a name

[06:49:04.0000] <annevk>
yeah

[06:49:05.0000] <MikeSmith>
annevk: yeah, what you said

[06:49:06.0000] <annevk>
but there's no normative text that supports the example

[06:49:07.0000] <MikeSmith>
agreed

[06:50:00.0000] <MikeSmith>
but that's sort of not exceptional in the case of HTML4

[06:50:01.0000] <MikeSmith>
I guess

[06:50:02.0000] <MikeSmith>
I think there are other instances of the examples having resulted in bad practices that the spec did not normatively describe

[06:51:00.0000] <MikeSmith>
so I guess we could say it's a change to existing practices that the HTML4 spec implicitly encouraged

[06:51:01.0000] <MikeSmith>
or some such

[06:52:00.0000] <MikeSmith>
I do think Jeremy is right to say that users should be alerted to this as being a more explicit restriction in HTML5

[06:55:00.0000] <annevk>
mkay

[06:56:00.0000] <Philip`>
annevk: I thought IE did support <rp> when I tested it

[06:56:01.0000] <Philip`>
though IE8 beta in IE8-mode had regressions

[06:57:00.0000] <annevk>
ok, it's not listed on msdn...

[06:57:01.0000] <annevk>
guess I shouldn't rely on that

[06:59:00.0000] <Philip`>
Reading MSDN is never quite an adequate substitute for testing

[07:04:00.0000] <Philip`>
/me tests IE8

[07:04:01.0000] <Philip`>
annevk: It does indeed hide <rp>

[07:04:02.0000] <Philip`>
(in IE8 mode and in quirks mode)

[07:12:00.0000] <hsivonen>
/me wonders if IPA is expressive enough to capture the sv-SE and sv-FI differences in the pronunciation of the 'k' in Linköping

[07:18:00.0000] <gsnedders>
hsivonen: k in sv-FI is more dorsal, or am I misremembering?

[07:19:00.0000] <mut>
hey,can i change the letter spacing in canvas?

[07:20:00.0000] <Philip`>
No

[07:20:01.0000] <mut>
:/

[07:20:02.0000] <Philip`>
except by doing fillText("H e l l o") etc

[07:20:03.0000] <mut>
ok :)

[07:20:04.0000] <mut>
ok

[07:20:05.0000] <mut>
I wanted to group it closer

[07:20:06.0000] <mut>
no problem though

[07:21:00.0000] <Philip`>
(You can only change the CSS 'font' property for text, and letter-spacing isn't part of that)

[07:22:00.0000] <mut>
ok

[07:22:01.0000] <mut>
shame :s

[07:22:02.0000] <mut>
its ok though, i can just overlay an image

[07:23:00.0000] <mut>
the text isnt dynamic so no problem

[07:23:01.0000] <hsivonen>
gsnedders: I guess it counts as more dorsal compared to sv-SE.

[07:24:00.0000] <hsivonen>
gsnedders: it's more like 'ch' in English "check"

[07:25:00.0000] <Philip`>
mut: I think the idea is that there should eventually be a drawHTMLElement() function, rather than putting more and more of CSS into custom canvas drawing functions

[07:28:00.0000] <mut>
ok

[07:28:01.0000] <mut>
ok yea that makes sense

[07:29:00.0000] <mut>
hmm

[07:29:01.0000] <mut>
im drawing the same drawing twice on 2 different canvass(canvii!?) the second of wich is in a full screen div, so that i can show a preview,and have the full size one slide down

[07:30:00.0000] <mut>
one scaled up, one scaled down

[07:30:01.0000] <mut>
works fine in ff, for some reason it justwont work in ie

[07:30:02.0000] <mut>
only the smaller one shows

[07:30:03.0000] <mut>
any idea?

[07:30:04.0000] <Philip`>
Using excanvas in IE?

[07:30:05.0000] <mut>
yea

[07:31:00.0000] <Philip`>
I have no idea how that works

[07:31:01.0000] <Philip`>
other than having an idea that it often doesn't work

[07:31:02.0000] <Philip`>
so I wouldn't be surprised by it breaking :-)

[07:31:03.0000] <mut>
all im doing is ctx = smallcanvas; drawCanvas(); ctx = largeCanvas; drawCanvas

[07:31:04.0000] <mut>
so if it works on the small on, surely.....?

[07:31:05.0000] <mut>
heh

[07:32:00.0000] <Philip`>
/me shrugs

[07:32:01.0000] <mut>
mabe im putting too much faith into logic

[07:32:02.0000] <mut>
;)

[07:39:00.0000] <mut>
hmmm

[07:39:01.0000] <mut>
seems like excanvas dosentsupport multiple canvases 

[07:41:00.0000] <Philip`>
That seems odd

[07:52:00.0000] <mut>
yea

[07:52:01.0000] <mut>
im gonna do a quick test with something real simple just to prove it

[08:16:00.0000] <Philip`>
/me didn't realise there was http://lists.w3.org/Archives/Public/public-html-ig-jp/ explicitly about HTML5

[08:56:00.0000] <annevk>
http://dvcs.w3.org/hg/html/file/4d46b00c2a90/tests/submission/Microsoft/htmldom/HTMLAnchorElement01.html the Microsoft submitted testcases are actually DOM2HTML tests?

[08:56:01.0000] <annevk>
not even updated?

[08:56:02.0000] <annevk>
/me wonders what the value of that is

[08:59:00.0000] <Philip`>
annevk: http://lists.w3.org/Archives/Public/public-html-testsuite/2010Feb/0009.html etc

[09:00:00.0000] <Philip`>
I don't think they're expected to be correct, they're just believed to be better than starting from scratch

[09:01:00.0000] <Philip`>
because maybe some of them still reveal bugs

[09:03:00.0000] <Philip`>
/me thinks the effort would be better spent on a systematic comprehensive approach to testing the HTML5 IDL instead

[09:07:00.0000] <annevk>
yeah

[09:07:01.0000] <annevk>
me too

[09:07:02.0000] <annevk>
updating tests is tricky

[09:10:00.0000] <Philip`>
/me started an IDL test script but then got bored and gave up

[10:05:00.0000] <LoneStar99>
what is the best way to delete a canvas path, when the cursor is dragged on it?

[14:05:00.0000] <jgraham>
/me agrees that trying to reuse DOM2 tests is a bad idea

[14:05:01.0000] <jgraham>
Maybe I shoul say o on the list or someting

[14:06:00.0000] <jgraham>
/me wonders if having conference calls is making people decide stuff in calls rather than ask the list

[14:06:01.0000] <jgraham>
(or if I just haven't been paying enough attention)

[14:07:00.0000] <annevk>
I haven't paid much attention either to be honest...


2010-03-11
[16:14:00.0000] <othermaciej>
jgraham: you should comment on the list

[16:15:00.0000] <othermaciej>
jgraham: I think no one but Kris and PLH is attending the conference calls, but Kris may also feel no one else is really giving input

[23:09:00.0000] <Mrmil>
Hi, just a quick question: is there a way to simulate tree structure in html5?

[23:18:00.0000] <MikeSmithX>
Mrmil: datagrid?

[23:22:00.0000] <Mrmil>
MikeSmithX: Wasn't that dropped?

[23:35:00.0000] <Hixie>
Mrmil: you can style nested <ul>s... search for "would be turned into the following DOM tree" in the HTML5 spec for an example (in the intro)

[00:27:00.0000] <othermaciej>
MikeSmith: are you about?

[00:44:00.0000] <zcorpan>
hmm, i was pondering about a potential security issue with the legacy encoding decl

[00:45:00.0000] <zcorpan>
consider a service that filters input but allows some text in a <meta name=description value=...> (or keywords)

[00:46:00.0000] <zcorpan>
attacker inserts text/html; charset=something

[00:46:01.0000] <zcorpan>
if that's after 512 bytes or 1024 bytes, it can cause scripts to execute twice

[00:47:00.0000] <zcorpan>
it can also cause the page to be interpreted in the wrong encoding, causing further problems

[00:48:00.0000] <zcorpan>
i wonder if we can kill re-navigation, and require http-equiv=content-type for the encoding decl

[00:50:00.0000] <annevk>
seems somewhat unlikely

[00:51:00.0000] <annevk>
the vulnerability is of the same level of complexity we expect people using srcdoc="" to handle just fine

[00:52:00.0000] <JonathanNeal>
You guys hear about the html5 shiv problem with js?

[00:53:00.0000] <zcorpan>
annevk: i don't follow

[00:53:01.0000] <annevk>
zcorpan, I don't see why value would not be a quoted attribute and why the service would not quote-escape the input

[00:55:00.0000] <zcorpan>
annevk: the problem isn't with quoting

[00:56:00.0000] <zcorpan>
annevk: <zcorpan> attacker inserts text/html; charset=something

[00:57:00.0000] <zcorpan>
annevk: per spec, any meta is respected as an encoding decl, so it's easy to convert an innocent meta into an encoding decl without breaking out of the attribute value

[00:57:01.0000] <zcorpan>
hmm the attribute needs to be content='', but still

[00:58:00.0000] <annevk>
oh wait, you mean the content attribute

[00:58:01.0000] <zcorpan>
yes

[00:58:02.0000] <annevk>
I think some UAs do require it

[00:59:00.0000] <annevk>
it seems like a very minor issue though

[00:59:01.0000] <zcorpan>
the spec requires just look at content=''

[01:00:00.0000] <annevk>
also, the dangerous encodings have been banned

[01:01:00.0000] <zcorpan>
doesn't mean it's not problematic to change the encoding of a page

[01:01:01.0000] <zcorpan>
you change the meaning of urls and form submission

[01:02:00.0000] <annevk>
it seems far-fetched to me

[01:02:01.0000] <zcorpan>
sure

[01:02:02.0000] <zcorpan>
so is the problem with reparsing

[01:03:00.0000] <annevk>
also, sites can prevent this from happening easily

[01:03:01.0000] <zcorpan>
sure

[01:05:00.0000] <zcorpan>
JonathanNeal: what problem?

[01:11:00.0000] <zcorpan>
Philip`: http://canvex.lazyilluminati.com/survey/2007-07-17/analyse.cgi/index is broken :(

[01:15:00.0000] <gsnedders>
How do you get the textContent of a node in XSL?

[01:15:01.0000] <JonathanNeal>
zcorpan, well..

[01:16:00.0000] <JonathanNeal>
var div = document.createElement('div');

[01:16:01.0000] <JonathanNeal>
 div.innerHTML = '<section> From innerHTML </section>';

[01:16:02.0000] <JonathanNeal>
 alert(div.innerHTML);

[01:16:03.0000] <JonathanNeal>
 // in IE

[01:16:04.0000] <JonathanNeal>
ooooh poo, why doesn't ychat tell me when it's going to do that.

[01:22:00.0000] <JonathanNeal>
Anyway, it's a problem, but I found a solution.

[01:22:01.0000] <JonathanNeal>
I'm tired, I'll finish the blog later and share it with you guys.  It's kinda important went trying to work with the dom in JS in IE with HTML5.

[01:22:02.0000] <JonathanNeal>
*when

[01:23:00.0000] <zcorpan>
i don't have ie; what does it do?

[01:24:00.0000] <JonathanNeal>
it returns "From innerHTML </section>"

[01:24:01.0000] <JonathanNeal>
It kills all html set via innerHTML.

[01:24:02.0000] <JonathanNeal>
all html5.

[01:24:03.0000] <zcorpan>
weird

[01:24:04.0000] <JonathanNeal>
It is very weird.

[01:24:05.0000] <JonathanNeal>
But fixable.

[01:25:00.0000] <JonathanNeal>
It has to do with the shiv and how dynamic events aren't attached to a shiv'd body.

[01:25:01.0000] <JonathanNeal>
so i just create all my nodes on a sister fragment when i need to manip, then detach. 

[01:26:00.0000] <MikeSmith>
gsnedders: <xsl:value-of select="."/>

[01:26:01.0000] <MikeSmith>
when the node you want is the current context node

[01:26:02.0000] <JonathanNeal>
Adds about 0.1ms to the fn.

[01:50:00.0000] <hsivonen>
Bug 9225 doesn't suprise me, but it's sad that the platform has had that bug for so long.

[02:00:00.0000] <Philip`>
zcorpan: Moved it to a new server and haven't bothered fixing the database stuff (it was using PostgreSQL but I don't fancy running that on the new server just for this)

[02:01:00.0000] <zcorpan>
Philip`: ok

[02:03:00.0000] <zcorpan>
Philip`: have you researched <meta>s that get interpreted as legacy encoding decl but don't have http-equiv=content-type?

[02:04:00.0000] <zcorpan>
/me finds http://lists.w3.org/Archives/Public/public-html/2008Mar/0029.html

[02:04:01.0000] <Philip`>
http://philip.html5.org/data/charsets.html#http-equiv-values

[02:05:00.0000] <zcorpan>
hmm no name='' values

[02:06:00.0000] <Philip`>
/me probably didn't collect that data

[02:06:01.0000] <Philip`>
It might be in the "none" ones

[02:07:00.0000] <zcorpan>
yeah

[02:07:01.0000] <zcorpan>
would be interesting to know what they are if we're changing the spec

[02:07:02.0000] <zcorpan>
maybe some pages use httpequiv=content-type or other

[02:19:00.0000] <zcorpan>
hmm writing a grep correct for this seems seems non-trivial; maybe it's better to run an html parser and inspect attributes

[02:22:00.0000] <zcorpan>
how convenient that the algorithm is removed from both html5 and mimesniff just when i want to look at it

[02:23:00.0000] <Dashiva>
Check an old revision of html5?

[02:23:01.0000] <Philip`>
I've got an easily-processable list of meta http-equiv/name/content values

[02:23:02.0000] <Philip`>
but it doesn't include any other attributes

[02:25:00.0000] <zcorpan>
http://www.google.com/codesearch?hl=en&lr=&q=%3Cmeta%5Cs%5B%5E%3E%5D%2Ahttpequiv+lang%3Ahtml&sbtn=Search

[02:31:00.0000] <zcorpan>
/me would like a field in code search "exclude ones that match this regexp:"

[03:37:00.0000] <Lachy>
interesting, yet convoluted way to watch h.264 encoded videos embedded with <video> in Firefox http://lifehacker.com/5488607/can-i-play-html5-youtube-videos-in-firefox-right-now

[03:38:00.0000] <Lachy>
basically, use the IETab extension to force YouTube to always load IE instead of Gecko, and install Google Chrome Frame

[03:39:00.0000] <Lachy>
so Firefox which loads IE, which in turn loads Chrome, just to watch h.264 in a browser that looks like Firefox, but really isn't.

[03:39:01.0000] <zcorpan>
seems easier to install a firefox skin to chrome

[03:39:02.0000] <Lachy>
yep :-)

[03:40:00.0000] <annevk>
:/

[03:51:00.0000] <zcorpan>
hsivonen: why did you retract the isindex request?

[03:52:00.0000] <hsivonen>
zcorpan: Implementing it is already sunk cost now, so I no longer care.

[03:52:01.0000] <hsivonen>
zcorpan: also, I discussed it with sicking, who pointed out that the HTML5 way will allow us to clean up some CSS formatter stuff down the road

[03:53:00.0000] <hsivonen>
and it we do that, we end up with less crazy code as a whole

[03:53:01.0000] <zcorpan>
ok

[03:55:00.0000] <hsivonen>
how can get the whatwg list route messages where I'm in To or Cc also though the list?

[03:55:01.0000] <hsivonen>
/me wants the right List-Id header on all whatwg discussions

[03:58:00.0000] <zcorpan>
i wonder if ie checks type=text for submission of <isindex type=checkbox>

[04:02:00.0000] <Lachy>
hsivonen, if I'm correctly understanding what you're asking for, Mailman has subscriber preferences that you can set for yourself so that you always receive a copy of the e-mail via the list, even if you're addressed in To or CC fields.

[04:02:01.0000] <annevk>
hsivonen, by allowing for duplicates

[04:03:00.0000] <annevk>
hsivonen, if you have set the software to prevent duplicates it will not give you the list copy if you already got a copy in a different manner (it cannot prevent giving you that copy)

[04:04:00.0000] <hsivonen>
/me fails to see an entry point to mailman settings

[04:05:00.0000] <Lachy>
http://lists.whatwg.org/listinfo.cgi/whatwg-whatwg.org

[04:05:01.0000] <Lachy>
see the "Unsubscribe or edit options" button at the bottom

[04:06:00.0000] <hsivonen>
Lachy: thanks. that "edit options" part sure is well hidden

[04:08:00.0000] <hsivonen>
huh. I already have avoid duplicates set to No.

[04:08:01.0000] <hsivonen>
/me tries re-setting

[04:12:00.0000] <Lachy>
is it possible that the duplicate is being detected and deleted by something on your mail server?

[04:12:01.0000] <hsivonen>
Lachy: unlikely, since it doesn't happen for w3.org lists

[04:28:00.0000] <hsivonen>
hmm. I've put commas is nonsensical places in a lot of email lately :-(

[04:28:01.0000] <hsivonen>
I wonder why

[04:29:00.0000] <hsivonen>
the correctness of my written English has deteriorated since high school even though I write more :-(

[04:31:00.0000] <Dashiva>
hsivonen: You're being exposed to native speakers and their sloppy ways

[04:57:00.0000] <asmodai>
Dashiva: Heh, I always cringe whenever I see 'would of'.

[04:58:00.0000] <Dashiva>
Indeed

[05:00:00.0000] <Lachy>
what is it about 'would of' that makes you cringe?

[05:01:00.0000] <Lachy>
the fact that it's grammatically incorrect version of would've or would have?

[05:01:01.0000] <Dashiva>
The fact that it's so obviously wrong

[05:02:00.0000] <MikeSmith>
Zakim, drop MIke

[05:03:00.0000] <Dashiva>
#whatwg

[05:03:01.0000] <Lachy>
it's just one of those phrases that people hear spoken, mis-interpret and treat as an idiomatic phrase. Happens quite a bit with a few phrases.

[05:07:00.0000] <asmodai>
Lachy: I never can understand how people manage to get that done. If I say would have/would've it doesn't sound like would of. Maybe it's because I learnt Queen's English in school.

[05:08:00.0000] <Dashiva>
Lachy: That assumes learning English entirely through audio and never even covering basic verb conjugation

[05:09:00.0000] <gsnedders>
A large part of my English didn't come from audio

[05:09:01.0000] <gsnedders>
But nor it come from learning formal grammar rules

[05:10:00.0000] <Dashiva>
Eliminate vocabulary and the ratios change I expect

[05:11:00.0000] <gsnedders>
I'd say the majority of the grammar I learnt aurally was mainly the more simple rules

[05:12:00.0000] <gsnedders>
And the increasingly complex came more from writing

[05:13:00.0000] <TabAtkins>
asmodai: In standard American, "would've" and "would of" do sound identical.

[05:13:01.0000] <jgraham>
hmm? "would've" and "would of" sound quite similar in home counties british english (which I assume you mean by "Queeen's English" since I can't imagine you learnt to speak in an incredibly posh accent)

[05:13:02.0000] <Lachy>
asmodai, that depends on the speaker's accent and how much they stress the 'o' in 'of'

[05:13:03.0000] <jgraham>
s/eee/ee

[05:15:00.0000] <MikeSmith>
TabAtkins: both "wooduv" but some people say "woulda" instead

[05:15:01.0000] <MikeSmith>
hey

[05:15:02.0000] <MikeSmith>
http://dev.w3.org/geo/api/spec-source-orientation.html

[05:16:00.0000] <MikeSmith>
question:

[05:16:01.0000] <TabAtkins>
MikeSmith: Yeah, but I'd pronounce both as "wooda" too.  ^_^

[05:16:02.0000] <MikeSmith>
http://dev.w3.org/geo/api/spec-source-orientation.html#examples

[05:17:00.0000] <MikeSmith>
"A device lying flat on a horizontal surface with the top of the screen pointing West has the following orientation... theta: 90"

[05:17:01.0000] <MikeSmith>
why is that theta: -90 ?

[05:18:00.0000] <MikeSmith>
why is that not theta: -90 ?

[05:18:01.0000] <MikeSmith>
that is, why positive 90 and not negative 90?

[05:19:00.0000] <MikeSmith>
"C is perpendicular to the screen, positive out of the screen... rotate the device frame around its C axis by theta degrees, with theta in [0, 360)"

[05:20:00.0000] <TabAtkins>
MikeSmith: Because it's been rotated 90 degrees in the positive direction (CCW).

[05:20:01.0000] <annevk>
it follows from that it cannot be negative

[05:20:02.0000] <MikeSmith>
TabAtkins: CCW is positive?

[05:20:03.0000] <TabAtkins>
Yes.

[05:20:04.0000] <TabAtkins>
Always has been.

[05:20:05.0000] <MikeSmith>
oh.

[05:20:06.0000] <MikeSmith>
dint know that

[05:21:00.0000] <TabAtkins>
Standard rules of angles is that CCW is positive, and the 0 angle is to the east.

[05:21:01.0000] <TabAtkins>
(When directionality is important, as well as simple magnitude of change.)

[05:21:02.0000] <MikeSmith>
damn I'm ig'nant

[05:22:00.0000] <annevk>
btw, why do we need CSSRule.type or Node.type?

[05:22:01.0000] <annevk>
is that something inflicted upon us by Java?

[05:22:02.0000] <annevk>
Node.nodeType even

[05:23:00.0000] <annevk>
I guess it's somewhat useful when iterating through a bunch of them; nevermind

[05:24:00.0000] <Philip`>
TabAtkins: I'd be cautious in calling any geometry conventions a "standard" :-)

[05:24:01.0000] <Philip`>
People can't even decide which handedness to use for axes, or which way around to write matrices

[05:25:00.0000] <Philip`>
That orientation spec says "Positive rotation around an axis is defined as clockwise when viewed along the positive direction of the axis."

[05:26:00.0000] <MikeSmith>
oh, handedness... this is ringing a bell now, thumbs-up, positive is in direction of the way fingers curl

[05:26:01.0000] <MikeSmith>
so something like that

[05:28:00.0000] <Philip`>
MikeSmith: The spec says C is perpendicular to the screen and outwards, so if the screen is lying flat on its back then C is vertically upwards

[05:28:01.0000] <Philip`>
so you have to look along the upwards direction

[05:28:02.0000] <Philip`>
i.e. from beneath

[05:28:03.0000] <Philip`>
and then rotate clockwise (for positive angles)

[05:28:04.0000] <MikeSmith>
ah

[05:28:05.0000] <Philip`>
which is equivalent to anticlockwise when you're looking down on it

[05:29:00.0000] <MikeSmith>
I see

[05:31:00.0000] <Philip`>
(and you start with the screen lying flat on its back with the top pointing North)

[05:31:01.0000] <Philip`>
(and rotate around C first)

[05:31:02.0000] <Philip`>
(so that's the right way to turn it to make its top point West)

[05:31:03.0000] <MikeSmith>
了解

[05:49:00.0000] <MikeSmith>
annevk: btw, some chrome devs in Japan are starting work CSS Paged Media implementation

[05:52:00.0000] <annevk>
sad day for paper

[05:52:01.0000] <annevk>
:p

[05:52:02.0000] <annevk>
but euh, yeah, the CSSOM story is kind of weak for paged media

[05:53:00.0000] <MikeSmith>
annevk: is the CSSDOM a comedy or a tragedy?

[05:53:01.0000] <MikeSmith>
annevk: is the editor of the CSSOM spec a tragic hero?

[05:55:00.0000] <annevk>
he lacks time and superpowers

[05:55:01.0000] <gsnedders>
/me wants to go and study English.

[05:55:02.0000] <annevk>
but he's doing some work on it at the moment, mostly fixing minor things

[05:56:00.0000] <MikeSmith>
...and he will eventually reveal himself in his full glory

[05:56:01.0000] <MikeSmith>
like the Transfiguration

[05:57:00.0000] <gsnedders>
/me wonders when that comment comes from MikeSmith whether it means a Gnostic Transfiguration

[05:57:01.0000] <annevk>
MikeSmith, it's just me

[05:59:00.0000] <Dashiva>
Is it strange that I find MikeSmith more understandable when he speaks Japanese?

[05:59:01.0000] <MikeSmith>
heh

[06:00:00.0000] <MikeSmith>
gsnedders: "I believe he's coming back, like he said, ... in the twinkling of an eye, he'll split the Eastern sky"

[06:07:00.0000] <MikeSmith>
heh, "The Web leaves no edge case unexercised"

[06:21:00.0000] <annevk>
MikeSmith, but anything specific about the CSSOM?

[06:21:01.0000] <Creap>
has there been discussion on having something like an validate=false attribute on input elements?

[06:22:00.0000] <annevk>
there is a novalidate attribute on submit buttons

[06:22:01.0000] <Creap>
oh, I'll try that

[06:22:02.0000] <MikeSmith>
annevk: no, was just thinking that it'd be nice to see CSSOM spec see that light of day eventually

[06:23:00.0000] <annevk>
here it is: http://dev.w3.org/csswg/cssom/

[06:23:01.0000] <Creap>
I think it would be useful on an input level though

[06:24:00.0000] <Creap>
I have some URL fields which are to contain variables represented with %s

[06:24:01.0000] <Creap>
but not all fields in the form

[06:25:00.0000] <annevk>
just use type=text?

[06:25:01.0000] <Creap>
yes, I'm doing that now

[06:25:02.0000] <Creap>
but it will mean other benefits of the URL fields are lost, such as special keyboards on mobiles and URL completion

[06:26:00.0000] <annevk>
/me wonders if Leonard is a Larry-proxy :p

[06:36:00.0000] <hsivonen>
regarding http://www.useragentman.com/blog/2010/03/09/cross-browser-css-transforms-even-in-ie/ what font rasterization method do Chrome and Opera use on Windows?

[06:37:00.0000] <hsivonen>
/me wildly guesses Firefox and Safari let the Windows font rasterizer perform the rasterization with a matrix

[06:38:00.0000] <gsnedders>
AFAIK we have our own

[06:38:01.0000] <hsivonen>
do Chrome and Opera do the same bug force ClearType on even if the user has it off system-wide?

[06:38:02.0000] <hsivonen>
gsnedders: ok

[06:38:03.0000] <hsivonen>
gsnedders: can you say if it's FreeType or something yet different?

[06:38:04.0000] <gsnedders>
(But that's just a vague memory, and may be totally wrong)

[06:39:00.0000] <hsivonen>
Chrome on Linux in those screenshots is using subpixel AA but Windows and Mac aren't

[06:39:01.0000] <hsivonen>
which hints at OS font rasterizer

[06:41:00.0000] <hsivonen>
aside: I don't understand why systems offer subpixel AA. On my LCDs, the color channels are always so well aligned that grayscale AA looks nice but subpixel AA looks like colors are leaking at the edges of black glyphs

[06:42:00.0000] <hsivonen>
/me always configures Macs to use the "CRT" AA and Ubuntu to use grayscale AA

[06:42:01.0000] <hsivonen>
even with LCDs that is

[06:44:00.0000] <hsivonen>
am I using unusually good screens or is subpixel AA snakeoil?

[06:51:00.0000] <jgraham>
hsivonen: I get noticably nicer (to my eyes) results with subpixel smoothing

[06:51:01.0000] <jgraham>
But I think that it is more related to expectations than anything

[06:55:00.0000] <gsnedders>
/me has had varying results

[06:55:01.0000] <annevk>
MikeSmith, I'm not particularly interested to be honest in putting CSSOM on TR/

[06:56:00.0000] <annevk>
MikeSmith, I have personally not noticed a whole lot of positive things coming from TR/. Actually mostly emails from people being confused or looking at TR/ while they should be looking at the editor's draft...

[06:58:00.0000] <Philip`>
/me prefers it with sub-pixel rendering enabled on Linux

[06:58:01.0000] <Philip`>
on a cheap laptop display and on a better LCD

[06:59:00.0000] <Philip`>
It's subtle enough that I don't notice the colours, and its makes vertical lines a tiny bit thicker so they better match the non-axis-aligned lines, I think

[06:59:01.0000] <Philip`>
s/its/it/

[07:10:00.0000] <jgraham>
On my display subpixel rendering gives better contrast compared to greyscale smoothing

[07:13:00.0000] <boblet>
hsivonen: maybe you’ve got good eyesight, or maybe you’re sitting too close to the monitor. AA works when your brain does the smoothing

[07:17:00.0000] <hsivonen>
I also think ClearType is ugly

[07:18:00.0000] <hsivonen>
and one reason I didn't switch from Mac to Gnome earlier was that around 2003 FreeType sucked too much

[07:19:00.0000] <boblet>
typography on Windows is offensive :)

[07:20:00.0000] <hsivonen>
hinting is *so* nineties

[07:20:01.0000] <hsivonen>
shape-preserving AA FTW!

[07:43:00.0000] <TabAtkins>
Argh, I really wish Richard would send plain-text emails.  It's very apparent that there's a mismatch between the visual size of some font metric on his screen and mine.  (Likely pts.)

[07:49:00.0000] <annevk>
/me just reads it as text/plain

[07:59:00.0000] <JonathanNeal>
Hello

[08:00:00.0000] <TabAtkins>
annevk: I'd read it as text/plain too, except that some people use indent-based quoting which I need to read as HTML to follow.

[08:01:00.0000] <Dashiva>
If people only used semantic markup ;)

[08:03:00.0000] <annevk>
TabAtkins, either my email client picks up on that or I just don't read those emails... not sure which it is

[08:03:01.0000] <TabAtkins>
I want everyone to just use Markdown like normal people.

[08:09:00.0000] <Dashiva>
TabAtkins: I think you need a new definition of normal

[08:09:01.0000] <TabAtkins>
The point of Markdown is that it *is* what normal people type.  ^_^

[08:11:00.0000] <Dashiva>
No, it's what some dude thinks looks somewhat similar to what normal people type

[08:43:00.0000] <zcorpan>
Hixie: where's the text that says to fire 'error' for <img src="">?

[08:56:00.0000] <Dashiva>
This is rich"These actions make it more

[08:56:01.0000] <Dashiva>
difficult to determine exactly what the person wants

[08:57:00.0000] <Dashiva>
Coming from the person who suggested splitting out microdata to another spec, and then immediately turned around and opposed that split-out spec

[08:58:00.0000] <MikeSmith>
annevk: a positive effect of publishing in TR/ is that doing so obligates and WG members to either disclose any related patents they may hold, or if they don't disclose them, they are agreeing to license them freely to any implementors of the spec

[08:59:00.0000] <MikeSmith>
I don't know whether that is an important consideration for the CSSOM spec or not, though

[08:59:01.0000] <annevk>
most features are in a previous copy of the spec, but doesn't the patent policy only take effect for RECs?

[09:00:00.0000] <annevk>
if we published REC-snapshots regularly the incentive would be bigger I guess

[09:00:01.0000] <MikeSmith>
no, it takes effect on publication of the FPWD

[09:00:02.0000] <annevk>
but now that getting to REC takes a decade I wonder how valuable the patent policy is

[09:00:03.0000] <MikeSmith>
after FPWD, WG members have 5 months to either disclose or default to licensing freely if not 

[09:01:00.0000] <othermaciej>
FPWD and LCWD trigger patent review and disclosure obligations

[09:02:00.0000] <annevk>
until REC they can still sue afaik

[09:02:01.0000] <annevk>
http://www.w3.org/Consortium/Patent-Policy-20040205/#sec-Requirements

[09:02:02.0000] <othermaciej>
technically I think the obligation to actually license is not until REC

[09:02:03.0000] <annevk>
indeed

[09:02:04.0000] <annevk>
which makes the policy somewhat incompatible with how specs are made

[09:03:00.0000] <MikeSmith>
hmm, yeah

[09:03:01.0000] <MikeSmith>
or, rather, I should say, I can understand it being seen that way

[09:06:00.0000] <othermaciej>
the potentially problematic part is that we call for implementations at CR, but patents may not be licensed until REC

[09:07:00.0000] <Dashiva>
/me wonders if XHTML is by definition well-formed XML, or if non-well-formed XML is (some classification, e.g. invalid) XHTML

[09:13:00.0000] <MikeSmith>
Dashiva: one way to answer that question is to not use the term XHTML at all but instead always refer to it as the "XML syntax for HTML"

[09:14:00.0000] <Dashiva>
MikeSmith: That makes XHTML seem like a second-class citizen, subordinate to the glorious HTML

[09:16:00.0000] <workmad3>
XHTML is an XML language for HTML, like KML is an XML language for map data, and SVG is an XML language for vector graphics

[09:18:00.0000] <MikeSmith>
exHTML

[09:46:00.0000] <AryehGregor>
"Even though the Internet was invented in the U.S. over 20 years ago,"  http://googleblog.blogspot.com/2010/03/internet-in-america-youtube-interview.html

[09:46:01.0000] <gsnedders>
Well, over 40.

[09:46:02.0000] <AryehGregor>
I guess it's true that 35 years ago is over 20 years ago, but that seems like a funny way to put it.

[09:46:03.0000] <AryehGregor>
40 is a bit of a stretch.  TCP/IP isn't that old.

[09:52:00.0000] <AryehGregor>
Neat: http://2.bp.blogspot.com/_LJuAPqyUVas/S5kFX5CnVYI/AAAAAAAAACs/vnO3XvDQ7HQ/s1600-h/sputnik.png

[10:01:00.0000] <zcorpan>
http://www.w3.org/mid/63df84f1003110952i369edbbfueea857bbc64cda86⊙mgc - the xml core wg has the same reasoning as julian (while i agree with sicking)

[10:02:00.0000] <gsnedders>
Challenge: teach me about pointers in C++.

[10:02:01.0000] <jgraham>
gsnedders: I'm not sure they're as complex as you think

[10:11:00.0000] <Philip`>
gsnedders: They're memory addresses

[10:11:01.0000] <Philip`>
That's all you need to know, and you can deduce the rest

[10:14:00.0000] <zcorpan>
Philip`: you think you could find pages that have a <meta> that could be interpreted as an encoding decl but doesn't have http-equiv=content-type?

[10:17:00.0000] <jgraham>
Philip`: Presumably you also need to know that they are typed

[10:25:00.0000] <jgraham>
(unless you are using a weird language like IDL -- Interactive Data Language, not the other one -- of course. But the less said about that the better)

[10:25:01.0000] <TabAtkins>
gsnedders: What Philip` said, plus pointers are often used to gain pass-by-reference behavior when only pass-by-value is allowed.

[10:26:00.0000] <TabAtkins>
That is, passing a pointer as an argument to a function allows you to pass large data structures around without copying them, and allows you to modify them in functionally-impure ways.

[10:28:00.0000] <Philip`>
jgraham: True - you need to know they're memory addresses, and (intptr_t)((T*)p+1) == (intptr_t)p+sizeof(T), and you can deduce the rest

[10:29:00.0000] <Philip`>
zcorpan: Maybe, but it depends on exactly what that means and whether it's easy to code

[10:31:00.0000] <gsnedders>
http://pastebin.com/NhuD82La

[10:31:01.0000] <gsnedders>
How does that crash?

[10:32:00.0000] <Philip`>
You should run it in Valgrind

[10:32:01.0000] <Philip`>
which tells you useful things about pointer errors

[10:32:02.0000] <Philip`>
In this case, p1 etc are uninitialised values

[10:33:00.0000] <Philip`>
so they contain random garbage (whatever happened to be on the stack)

[10:33:01.0000] <Philip`>
which is interpreted as a memory address (because pointers are memory addresses)

[10:33:02.0000] <Philip`>
and when you try to write something to the content of that memory address (*p1 = ...) it writes to a garbage location

[10:52:00.0000] <AryehGregor>
Which most likely means the program tries writing to a page that's not in its page tables, so the CPU raises a fault, and the OS sees you're doing something crazy and decides to put you down for your own good.

[10:53:00.0000] <AryehGregor>
You need to ensure that p1 points to something first, like: p1 = malloc(sizeof p1); or something (my C isn't very good).

[10:53:01.0000] <Philip`>
That would allocate the size of the pointer, not the size of the pointee

[10:54:00.0000] <AryehGregor>
Rats.

[10:54:01.0000] <AryehGregor>
p1 = malloc(sizeof int); then.

[10:54:02.0000] <Philip`>
Better to do int* p1 = new int;

[10:54:03.0000] <AryehGregor>
malloc() will pick a memory address on the heap where there's enough free space, note internally that the memory is being used, and return the address at the start of the allocated block.

[10:54:04.0000] <Philip`>
since this is C++

[10:54:05.0000] <AryehGregor>
Well, if you're using C++.

[10:54:06.0000] <Philip`>
gsnedders said C++ :-)

[10:54:07.0000] <AryehGregor>
So he did.

[10:55:00.0000] <Philip`>
He could have been lying, though

[10:55:01.0000] <AryehGregor>
Anyway, so do int *p1 = new int; int *p2 = new int; int *p3 = new int; or whatever, then you can use them largely as you expect as long as you put * before them whenever you use them.

[10:56:00.0000] <AryehGregor>
Except that they're allocated on the heap, not the stack, so for instance, you can return the pointer from a function safely.

[10:56:01.0000] <AryehGregor>
And allocate very large things to them.  And if you don't delete them later, you've caused a memory leak.

[10:58:00.0000] <Philip`>
You can return ints from a function safely too :-p

[10:59:00.0000] <Philip`>
Oh, I guess you mean like "int* f() { int x = 25; int* y = &x; return y; }"

[10:59:01.0000] <TabAtkins>
Well, you can return auto-generated copies of ints safely.

[11:00:00.0000] <AryehGregor>
Well, yes.  I'm not entirely sure why you'd want a pointer to an int to begin with.  But they're essential for things like structs.

[11:01:00.0000] <AryehGregor>
TabAtkins, returning a copy of an int costs less than returning a pointer to it, so it's not really a useful distinction in the case of ints.

[11:01:01.0000] <annevk>
did sicking just offer me cookies?

[11:01:02.0000] <AryehGregor>
Well, I guess it is if you want to modify it.

[11:01:03.0000] <TabAtkins>
annevk: Yes.  Take him up on it and send some to me.

[11:01:04.0000] <AryehGregor>
annevk, where?

[11:02:00.0000] <TabAtkins>
AryehGregor: Reference vs value distinction is very important!

[11:02:01.0000] <annevk>
/me likes cookies

[11:02:02.0000] <zcorpan>
/me too

[11:02:03.0000] <AryehGregor>
Okay, but "returning x" doesn't necessarily mean "returning the value of x" or "returning a reference to x", it can mean either.

[11:02:04.0000] <AryehGregor>
So your correction was unnecessary, anyway.

[11:02:05.0000] <AryehGregor>
/me hasn't had cookies in a while, should get some today

[11:03:00.0000] <AryehGregor>
/me doesn't think it's worth it to send cookies intercontinentally, though

[11:03:01.0000] <Philip`>
I thought cookies were considered obsolete

[11:04:00.0000] <zcorpan>
/me goes to get a homemade cookie

[11:05:00.0000] <zcorpan>
mmmmmm

[11:05:01.0000] <annevk>
Philip`, that has been removed from the draft

[11:05:02.0000] <annevk>
zcorpan, aaah, i want!

[11:06:00.0000] <AryehGregor>
I haven't had a homemade cookie in *ages*.

[11:06:01.0000] <AryehGregor>
Why do they always taste so much better than store-bought ones?  :(

[11:06:02.0000] <AryehGregor>
Someone needs to invent packaged food that tastes good.

[11:07:00.0000] <TabAtkins>
AryehGregor: Sure, but often when talking about "returning *an* int" or whatnot you are implying that it's the same object throughout.  Understanding how value semantics break that implication is important.

[11:08:00.0000] <TabAtkins>
AryehGregor: There are good cookies in the store, but generally they're made by the bakery there.  There is no such thing as a factory-produced cookie that tastes good.

[11:08:01.0000] <zcorpan>
you should buy cookies from a pastry

[11:08:02.0000] <Philip`>
AryehGregor: How does the cost of the ingredients and time taken to homemake them compare to the price of the store-bought ones?

[11:09:00.0000] <TabAtkins>
I usually eat pastries, rather than bartering with them.

[11:09:01.0000] <zcorpan>
wait pastry might not be the word i'm looking for

[11:09:02.0000] <TabAtkins>
Pastry shop?

[11:09:03.0000] <Philip`>
Pâtisserie?

[11:09:04.0000] <AryehGregor>
Philip`, time is worth much more to me than tasty cookies, alas.

[11:10:00.0000] <AryehGregor>
If there were a good bakery around I could use that, I guess.

[11:10:01.0000] <AryehGregor>
But the only kosher bakery in my neighborhood makes cookies that taste about as good as packaged stuff.

[11:10:02.0000] <TabAtkins>
Baking cookies is fun by itself, so the time cost is negligible.

[11:10:03.0000] <TabAtkins>
AryehGregor: Simple solution: stop eating kosher.

[11:12:00.0000] <zcorpan>
annevk: wait you have a conforming badge?

[11:12:01.0000] <AryehGregor>
TabAtkins, or I could move to Israel, where at least half the stores are kosher even in the least religious parts.

[11:13:00.0000] <TabAtkins>
That's also a good idea.  The IDF girls are *hot*.

[11:13:01.0000] <AryehGregor>
My sister is in the IDF.

[11:13:02.0000] <AryehGregor>
Air force, specifically.  Non-combat, though.

[11:13:03.0000] <AryehGregor>
She works in a chemistry lab of some kind.

[11:13:04.0000] <annevk>
zcorpan, who doesn't?

[11:14:00.0000] <zcorpan>
annevk: i don't. but i have cookies

[11:14:01.0000] <AryehGregor>
Conforming badge?

[11:15:00.0000] <annevk>
there's something wrong here

[11:17:00.0000] <zcorpan>
Philip`: what i want is a list of occurrences (with the associated url) of <meta> elements that have a content='' attribute whose value matches /^[^;]*;\s*charset=/i and doesn't have an http-equiv attribute with the value 'content-type' (case insensitively)

[11:17:01.0000] <zcorpan>
Philip`: i guess the whole thing could be written as grep, but i think it'd be a bit complex and might be easier to do running an html parser

[11:28:00.0000] <hsivonen>
I though my mobile browsers ran out of memory and cut the irc logs short, but it seem krijnh's server is terminating the stream early

[11:30:00.0000] <annevk>
yeah, the timeout is fairly short

[11:32:00.0000] <zcorpan>
Philip`: (?i)<meta\s(([^h]|h(?!ttp-equiv=("|')?content-type))[^>]*content=("|')?[^>"';]+;\s*charset=)|(content=("|')?[^>"';]+;\s*charset=[^>]*[^h]|h(?!ttp-equiv=("|')?content-type)))

[11:32:01.0000] <AryehGregor>
Under what circumstances is rel="external" actually useful?

[11:33:00.0000] <zcorpan>
AryehGregor: i've only seen it as a way to hide target=_blank from validators

[11:34:00.0000] <zcorpan>
insert the word 'used' somewhere

[11:34:01.0000] <AryehGregor>
/me inserts it before "hide" and becomes confused

[11:39:00.0000] <TabAtkins>
Unrelated: http://poloassassin.typepad.com/blog/2010/03/awesome-dinosaur-comics-that-my-computer-wrote.html

[11:39:01.0000] <AryehGregor>
/me saw that

[11:40:00.0000] <AryehGregor>
It says something about Dinosaur comics that most of those are pretty plausible.

[11:54:00.0000] <hsivonen>
the TOC of the French translation of Dive into HTML5 isn't in the alphabetical order :-(

[11:56:00.0000] <AryehGregor>
hsivonen, neither is the English translation, as far as I can tell.  :)

[11:56:01.0000] <hsivonen>
also, the French version demonstrates why it's bad to subset fonts manually and fail to adjust the repertoire when the need for glyphs grows

[11:58:00.0000] <hsivonen>
oh. I got Dive into HTML5 and Dive into Python 3 mixed up

[11:58:01.0000] <hsivonen>
and it's the drop caps, not the headings

[13:16:00.0000] <LoneStar99>
have a canvas app, and need to draw a dynamic box which moves around as the cursor is moved... 

[13:17:00.0000] <LoneStar99>
app = element

[13:17:01.0000] <Dashiva>
Which part are you having problems with?

[13:18:00.0000] <LoneStar99>
drawing the box and allowing to drag along with the cursor

[13:19:00.0000] <LoneStar99>
http://devfiles.myopera.com/articles/649/example2.html

[13:20:00.0000] <LoneStar99>
want the box to drag along, so that a user knows where the cursor is....

[13:20:01.0000] <LoneStar99>
like a red box

[13:21:00.0000] <Dashiva>
Use mousemove to track the position of the cursor

[13:22:00.0000] <Dashiva>
Each frame, draw the content and then draw the square at the stored cursor position

[13:22:01.0000] <LoneStar99>
yeah got that, confused on if the "box" should be a drawing or "css" box...

[13:25:00.0000] <TabAtkins>
If you want it to be part of the <canvas>, should be a drawing.  If you're fine with it being a separate overlay, "css box".

[13:28:00.0000] <TabAtkins>
The "make an element overlay the <canvas>" approach has the advantage that you don't need to redraw the canvas just to make the box follow the mouse, but if you're redrawing the canvas anyway, it doesn't really matter.

[13:31:00.0000] <LoneStar99>
would I need two canvas elements one for the "drawing" as in http://devfiles.myopera.com/articles/649/example2.html  and the other for the "red box cursor?"  

[13:35:00.0000] <LoneStar99>
ok i think I need two canvas elements... one that draws the red box only once and the other that draws the dynamic drawing

[13:56:00.0000] <TabAtkins>
LoneStar99: You could do that, but it's not necessary.  The <canvas> is just a big box for you to draw everything in.  You have to handle ensuring that you draw the red box last, but you can certainly do so in the same canvas as your other drawing.

[13:57:00.0000] <TabAtkins>
Everyone else: I'm having an issue with some scripting that works in Firefox, but fails identically in Chrome and IE.

[13:58:00.0000] <TabAtkins>
At this page - http://74.54.176.50/~spinnerg/index.php?page=spin-schedule - FF makes the table cells grow when you hover them and on click brings up an overlay and a datepicker.

[13:58:01.0000] <LoneStar99>
TabAtkins: is it possible to have different layers in the same drawing?

[13:58:02.0000] <LoneStar99>
sorry canvas

[13:58:03.0000] <TabAtkins>
Both IE and Chrome, though, fail at some point earlier, and won't animate the cells properly.

[13:58:04.0000] <TabAtkins>
LoneStar99: Sure, though you have to handle that yourself.  <canvas> does not have any native idea of 'layers'.

[13:58:05.0000] <TabAtkins>
(Though you can fake it with multiple <canvas>es put on top of each other, as you suggested.)

[13:59:00.0000] <LoneStar99>
ok

[13:59:01.0000] <LoneStar99>
let me try that

[13:59:02.0000] <TabAtkins>
Within a single <canvas>, though, you just have to keep track of drawing order by yourself, and ensure that things on higher "layers" get drawn after things on lower ones.

[14:00:00.0000] <LoneStar99>
ok

[14:00:01.0000] <LoneStar99>
thanks, let me experiement

[14:08:00.0000] <TabAtkins>
Hmm, looks like my problem might lie in the color animation library.  Damn.


2010-03-12
[17:36:00.0000] <MikeSmith>
othermaciej: for some reason, on my system, no progress messages seem to get written to the console during the process of updating to a new nightly build

[17:36:01.0000] <MikeSmith>
or maybe I'm looking in the wrong place

[17:37:00.0000] <MikeSmith>
but I just updated to the latest nightly and when I go to check All Message in Console.app, I see just one Safari/WebKit-related message there: 10/03/12 10:37:32 Safari[43511] WebKit r55834 initialized.

[17:39:00.0000] <othermaciej>
MikeSmith: I know nothing about the nightly update mechanism since I always just build my own

[17:39:01.0000] <MikeSmith>
ah yeah

[17:41:00.0000] <MikeSmith>
well, I just mentioned it because bdash suggested checking the console messages as a way of troubleshooting any install problems.. I guess I'll ping him on IRC when he's around again and ask him if I need to switch on verbosity somewhere or something

[17:42:00.0000] <MikeSmith>
othermaciej: anyway, about doing the regular checks/reports on resolution status, etc., of HTML WG bugs, I definitely want to help out with that

[17:43:00.0000] <MikeSmith>
and with anything else we can think of that would help with getting issues resolved, and help with getting us closer to being ready for LC

[17:48:00.0000] <MikeSmith>
I'll start doing the 2nd thing you mentioned

[17:48:01.0000] <MikeSmith>
http://krijnhoetmer.nl/irc-logs/html-wg/20100311#l-88

[17:48:02.0000] <MikeSmith>
"someone should be checking bugs in RESOLVED state to make sure they contain the required info (the Editor's Response, a link to the spec diff if there was actually a spec change, a rationale that is actually some sort of meaningful rationale"

[17:50:00.0000] <othermaciej>
MikeSmith: great - I'm going to mass-move the bugs that predate the policy info VERIFIED tomorrow, and the rest can be done at any time, no set schedule

[17:50:01.0000] <MikeSmith>
OK

[17:51:00.0000] <MikeSmith>
It will help in general if we had some direct SQL query access to the bugzilla DB

[17:51:01.0000] <MikeSmith>
maybe I can get that

[17:52:00.0000] <MikeSmith>
but rather than checking it manually, I'm hoping there is maybe a way that I can write a script to generate report" is something that of course needs to be checked by a human

[17:53:00.0000] <MikeSmith>
but having the Editor's Decision comments in some kind of easy-to-scan report form would make that much easier

[17:54:00.0000] <othermaciej>
a script would be great

[17:54:01.0000] <othermaciej>
there's actually a couple of scripts we could use:

[17:54:02.0000] <othermaciej>
- Script to compile Editor's Resolution and spec diff link for bugs in RESOLVED (for verification)

[17:54:03.0000] <othermaciej>
(Hixie uses a consistent format so it shouldn't be too hard)

[17:56:00.0000] <othermaciej>
- Script to count bugs that are still not in a terminal state (anything but CLOSED or VERIFIED+NoReply)

[17:57:00.0000] <othermaciej>
- Script to generate a Disposition of Comments based on bugs that *are* in a terminal state (we'll need this for getting out of Last Call)

[18:01:00.0000] <MikeSmith>
the last one we are absolutely going to need eventually, one way or the other

[18:02:00.0000] <MikeSmith>
the other two I can definitely work on trying to get set up

[18:02:01.0000] <MikeSmith>
I vaguely remember asking the systems team some time ago about if there were a way I could get SQL access to the bugzilla DB

[18:03:00.0000] <MikeSmith>
but I can remember if the answer was a definition No, or just a "No, we'd rather you didn't do that, so we're not going to let you" No

[18:09:00.0000] <MikeSmith>
.me notices new Audits tab in WebKit

[18:09:01.0000] <MikeSmith>
including "Remove unused CSS rules"

[18:09:02.0000] <MikeSmith>
this is nice

[18:10:00.0000] <MikeSmith>
it would also be nice to have as an online checker somewhere

[18:10:01.0000] <MikeSmith>
or maybe it already is

[18:11:00.0000] <MikeSmith>
wow, I guess the Audits tab has been in there for a while.. Not sure how I missed it

[18:12:00.0000] <MikeSmith>
I think the name "Audits" scared me off

[18:12:01.0000] <MikeSmith>
if is said something more like "Optimize" that would be a lot more attractive

[18:18:00.0000] <othermaciej>
that's a good idea

[18:19:00.0000] <othermaciej>
"Audits" sounds like it will poke in your financial records

[18:19:01.0000] <MikeSmith>
yeah

[18:53:00.0000] <MikeSmith>
http://googlewebmastercentral.blogspot.com/2010/03/microdata-support-for-rich-snippets.html

[18:53:01.0000] <MikeSmith>
wow, the news was posted on March 9?

[18:54:00.0000] <MikeSmith>
anyway, so much for the argument about Microdata not having any implementation support

[18:56:00.0000] <othermaciej>
hey neat

[19:04:00.0000] <Hixie>
anyone know if the changes discussed for the File API are going to make it?

[19:05:00.0000] <Hixie>
in particular, moving .type to Blob and renaming .urn to .URL?

[19:11:00.0000] <Hixie>
othermaciej: how do you think i should address the use case of allowing a canvas to be uploaded to a server?

[19:13:00.0000] <othermaciej>
Hixie: you mean other than the possibilities of uploading a data: URL or uploading the ImageData in a way you serialize yourself?

[19:14:00.0000] <othermaciej>
Hixie: I am going to try to get binary data on the agenda for the TC-39 meeting at Apple later this month, if they can be brought to actually care about the issue, then I'd recommend that

[19:15:00.0000] <Hixie>
k, i'll delay a bit longer then

[19:15:01.0000] <Hixie>
failing that i'll just provide a toBlob() to match toDataURL()

[19:16:00.0000] <othermaciej>
Hixie: I'm not super enthusiastic about the idea of using a File or Blob because it seems broken to me to use an interface that's intended for async-only access to persistently stored data for this use case

[19:16:01.0000] <othermaciej>
Hixie: but it may be better than nothing

[19:16:02.0000] <Hixie>
yeah that's more or less my position

[21:20:00.0000] <paradisaeidae_>
I am attempting to factor in a number of html5 elements to a page. When I have a number of javascripts, gifs, svg and xhtml:body sections, then a page can reach 500k.

[21:20:01.0000] <paradisaeidae_>
This is the sort of size similar functionality is delivered by flash.

[21:20:02.0000] <paradisaeidae_>
Though there is a presentation hint possible with flash.

[21:20:03.0000] <paradisaeidae_>
The '...loading...' message.

[21:20:04.0000] <paradisaeidae_>
Is there a function in html5 spec which does anything similar?

[23:01:00.0000] <annevk>
Google supports Microdata per @diveintomark

[23:01:01.0000] <annevk>
http://googlewebmastercentral.blogspot.com/2010/03/microdata-support-for-rich-snippets.html

[23:44:00.0000] <annevk>
oh hey, the IETF is having discussion on whether text/plain is still a suitable format on http://www.ietf.org/mail-archive/web/ietf/current/maillist.html

[23:44:01.0000] <annevk>
funny

[23:48:00.0000] <hsivonen>
hah. framing the discussion as PS/PDF vs. plain text as opposed to HTML vs plain text

[23:49:00.0000] <annevk>
yeah, I've no idea how you can even start with suggesting PDF

[23:49:01.0000] <annevk>
something must be seriously wrong in the workflow of that guy

[23:51:00.0000] <hsivonen>
and then the thread turn to .epub

[23:51:01.0000] <hsivonen>
wow

[23:52:00.0000] <annevk>
oh, I didn't get that far

[23:52:01.0000] <annevk>
/me goes back reading

[23:52:02.0000] <annevk>
oh, that's actually early on, I skipped a few fun emails it seems

[23:53:00.0000] <hsivonen>
/me wonders ow many messages it takes for someone to remind the others that this Web thing happened while they weren't paying attention

[23:56:00.0000] <hsivonen>
the credit goes to Julian: http://www.ietf.org/mail-archive/web/ietf/current/msg60576.html

[23:58:00.0000] <zcorpan>
i wonder why people refer to plain text as "ASCII"

[23:59:00.0000] <hsivonen>
RFC plain text is ASCII, but yeah, it's interesting that moving to UTF-8 plain text wasn't proposed

[00:00:00.0000] <annevk>
Japanese guy freaked out over moving away from ASCII: http://www.ietf.org/mail-archive/web/ietf/current/msg60589.html

[00:00:01.0000] <annevk>
the interwebs is full of fascinating plot twists

[00:00:02.0000] <hsivonen>
oh, it's mentioned later in the thread

[00:01:00.0000] <annevk>
as far as I can tell Julian and Tim are the only sane voices in that discussion

[00:01:01.0000] <annevk>
way to go IETF...

[00:04:00.0000] <MikeSmith>
they are right that all the fancy graphics just detract from the essentials of the gaming experience

[00:05:00.0000] <MikeSmith>
the best games are just plain text, like the original Star Trek

[00:06:00.0000] <annevk>
yeah

[00:06:01.0000] <annevk>
http://xkcd.com/91/

[00:06:02.0000] <MikeSmith>
heh

[00:07:00.0000] <MikeSmith>
"you don't need any extra junkware to be able to read the RFCs"

[00:07:01.0000] <MikeSmith>
they should put that the RFC boilerplate

[00:08:00.0000] <MikeSmith>
in bold type

[00:08:01.0000] <MikeSmith>
and/or italiced

[00:08:02.0000] <MikeSmith>
and in red

[00:08:03.0000] <annevk>
but they can't

[00:08:04.0000] <annevk>
that's not ASCII o_O

[00:08:05.0000] <MikeSmith>
oh damn

[00:09:00.0000] <MikeSmith>
I was going to say that they could make it hyperlink to a detailed explanation

[00:09:01.0000] <hsivonen>
You can read PDFs using less if you are cool enough. I have.

[00:10:00.0000] <hsivonen>
(be sure not the have flate streams in there, though, unless you can run deflate mentally)

[00:10:01.0000] <zcorpan>
hsivonen is cool

[00:11:00.0000] <MikeSmith>
you can read a lot of stuff using strings(1)

[00:11:01.0000] <MikeSmith>
with implied quotes in there somewhere

[00:11:02.0000] <MikeSmith>
maybe around the word read

[00:12:00.0000] <zcorpan>
i read PNGs by reading the PNG spec and then opening in a hex editor

[00:12:01.0000] <MikeSmith>
zcorpan: I think they like the word ASCII because is more l33t

[00:13:00.0000] <MikeSmith>
welcome to our hackers club. we do everything in ASCII. if you dont like it, get out of the computer room and go eat lunch in the cafeteria or somewhere instead

[00:14:00.0000] <annevk>
that hackers club must have some kind of millennium bug...

[00:16:00.0000] <MikeSmith>
they accidentally creating a worm hole with their mad hacking skillz and they are stuck in it

[00:16:01.0000] <MikeSmith>
and happy to be stuck in it

[00:20:00.0000] <zcorpan>
is google's impl of microdata conforming?

[00:22:00.0000] <hsivonen>
zcorpan: dunno, but anyone wnting to use Google's impl for Microdata advocacy should find out.

[00:39:00.0000] <Lachy>
where is Google's implementation of Microdata? Have they released the source, or just incorported it into their search engine somehow?

[00:40:00.0000] <annevk>
hsivonen, that's not how advocacy for RDFa worked...

[00:40:01.0000] <annevk>
or works, haven't followed it recently

[01:06:00.0000] <annevk>
/me wonders how long we'll have to wait before the lawyers at Google have sorted out VP8

[01:26:00.0000] <hsivonen>
annevk: This is Microdata advocacy's chance to hold the moral high ground compared to RDFa advocacy.

[01:27:00.0000] <othermaciej>
so I just noticed that it's effectively impossible to google for the phrase, "5 > 2"

[01:27:01.0000] <othermaciej>
that's kinda sad

[01:28:00.0000] <annevk>
hopefully it's a high priority bug

[01:28:01.0000] <othermaciej>
it should be!

[01:28:02.0000] <othermaciej>
(it's unbingable as well)

[01:29:00.0000] <annevk>
the other day I saw some people theorizing over what it might mean

[01:29:01.0000] <annevk>
apparently it was in a Google I/O presentation but not clarified; they thought it had something to do with Web 2.0

[01:30:00.0000] <othermaciej>
heh

[01:30:01.0000] <othermaciej>
was it in there as an inside joke or did they expect people to get the reference?

[01:31:00.0000] <zcorpan>
http://www.google.com/codesearch?hl=en&lr=&q=%5Cb5%5C+%3E%5C+2%5Cb&sbtn=Search

[01:31:01.0000] <annevk>
probably an inside joke; it has been an inside joke from the moment you made that t-shirt I think

[01:31:02.0000] <MikeSmith>
http://www.w3.org/2010/02/rdfa/wiki/RDFa-DOM-API

[01:31:03.0000] <annevk>
we should make some new t-shirts

[01:31:04.0000] <othermaciej>
it's been an inside joke since before I made the t-shirt

[01:31:05.0000] <othermaciej>
5 > 2 is no longer an interesting statement

[01:32:00.0000] <othermaciej>
sadly 5 > f is not true (at least in hexadecimal)

[01:32:01.0000] <annevk>
<i> &hearts; <html> is one I suggested for #fronteers once

[01:32:02.0000] <Philip`>
/me would hypothesise that Google uses the same 'HTML' parser for Microdata as for RDFa, i.e. one that's more like an error-correcting XML parser, so it'll handle things like "<p whatever>...<p whatever>" as if they were nested

[01:32:03.0000] <annevk>
or s/<html>/HTML5/

[01:32:04.0000] <zcorpan>
1.0 > 5

[01:33:00.0000] <othermaciej>
so here's one thing I don't get about people making RDF APIs

[01:33:01.0000] <othermaciej>
why is the interface always a flat list of triples?

[01:33:02.0000] <othermaciej>
I thought the whole point of RDF was that it has a graph structure, and the interesting thing to do is to traverse or make queries against the graph

[01:34:00.0000] <othermaciej>
if you completely drop the graph structure, what's the point?

[01:34:01.0000] <othermaciej>
it would be like having a DOM API where the only way to traverse nodes was as a flat list of all the nodes in the document

[01:35:00.0000] <othermaciej>
that's cute

[01:35:01.0000] <othermaciej>
too bad HTML5 has no version indicator so you can't do it in the syntax

[01:35:02.0000] <othermaciej>
finally, a critical use case for a version indicator!

[01:35:03.0000] <annevk>
<i> &hearts; <!doctype html> says enough

[01:36:00.0000] <annevk>
now we need some designer to make it look pretty :)

[01:37:00.0000] <othermaciej>
tragically monospace fonts don't seem to have a decent-looking solid heart glyph (at least not the ones on my machine)

[01:37:01.0000] <othermaciej>
I am amused that Mark Pilgrim still uses 5>2 as his twitter icon

[01:38:00.0000] <othermaciej>
I wonder how many of his followers get the reference

[01:38:01.0000] <zcorpan>
http://creatr.cc/creatr/logo/i%20hearts%20doctype%20html.png?1268387003

[01:38:02.0000] <hsivonen>
Is http://www.amazon.com/HTML5-Up-Running-Mark-Pilgrim/dp/0596806027/ref=pd_sim_b_1 the same content as Dive into HTML5?

[01:38:03.0000] <zcorpan>
hsivonen: yes

[01:39:00.0000] <othermaciej>
whoah there's a surprising number of HTML5 books now

[01:40:00.0000] <MikeSmith>
othermaciej: really?

[01:40:01.0000] <othermaciej>
well, maybe I am naiive to be surprised

[01:41:00.0000] <MikeSmith>
I didn't know any printed HTML5 books had been published yet other than the ones that got published here in Japan

[01:41:01.0000] <annevk>
if you search on Amazon you can see there's quite a few coming up

[01:41:02.0000] <MikeSmith>
I see now

[01:42:00.0000] <annevk>
most by people I don't know, which is encouraging

[01:42:01.0000] <hsivonen>
http://www.amazon.com/dp/0321719913 by tantek

[01:42:02.0000] <MikeSmith>
the book equivalent of vaporware, except I guess for Mark's, which is online already

[01:42:03.0000] <MikeSmith>
*content of which is already online

[01:42:04.0000] <zcorpan>
/me notes that his video article is under BSD so someone could rip it off and use it in a book (assuming i understand the license correctly)

[01:42:05.0000] <hsivonen>
http://www.amazon.com/Introducing-HTML5-Voices-That-Matter/dp/0321687299/ref=pd_sim_sbs_b_2 brucel&remysharp

[01:44:00.0000] <zcorpan>
/me also notes that he is happy to review books on html5, for anyone writing such books

[01:44:01.0000] <MikeSmith>
me sees franksalim as co-author on another of those books

[01:44:02.0000] <hsivonen>
I've volunteered to make some simple cross-browser <video> boilerplate and I can't decide if I should have unscripted Cortado fallback with the user of the cookbook having to put the video size and URL in multiple places or if I should have scripted fallback that copies the URL and size from the video element

[01:45:00.0000] <MikeSmith>
I want a T-shirt that symbolizes the statement, "Using HTML or PDF for RFCs is about the same as moving from English language RFCs to mandarin language RFCs"

[01:45:01.0000] <MikeSmith>
I think maybe the symbol of caveman hitting himself in the head with a brick might symbolize that sentiment pretty well

[01:45:02.0000] <othermaciej>
not sure if it can be expressed as a t-shirt slogan

[01:46:00.0000] <zcorpan>
hsivonen: i think scripted easy-to-use are more successful

[01:48:00.0000] <othermaciej>
MikeSmith: the idea that HTML is seen as some mysterious foreign thing is kind of mind-boggling

[01:48:01.0000] <zcorpan>
hsivonen: does cortado support autoplay and different preload strategies?

[01:48:02.0000] <othermaciej>
is there really anyone with a computing environment that can't handle HTML at *least* as well as plain text?

[01:49:00.0000] <vininim>
hm, what's the best of way of inserting a <nav> block into the <body> tree of a page using external html?

[01:49:01.0000] <Philip`>
My mail clients can't

[01:49:02.0000] <MikeSmith>
othermaciej: I think the Martin Rex guy must be a sockpuppet attempt to make the RFCs-in-ASCII argument look as absurd as possible

[01:50:00.0000] <othermaciej>
my favorite part is when people linked the HTML versions of RFCs to demonstrate the awesomeness of plain text

[01:50:01.0000] <zcorpan>
vininim: external html?

[01:50:02.0000] <hsivonen>
MikeSmith: wow. I missed the "mandarin" argument when I first skimmed the thread

[01:50:03.0000] <Dashiva>
"Stone tablets were good enough for my ancestors, we don't need no stinking HTML specs"

[01:51:00.0000] <hsivonen>
zcorpan: I don't remember what tunable Cortado has

[01:51:01.0000] <zcorpan>
is there an iPad in stone?

[01:52:00.0000] <Philip`>
/me notes that http://tools.ietf.org/rfc/rfc303.txt is still non-ASCII

[01:52:01.0000] <hsivonen>
Dashiva: stone tables have a proven archival track record

[01:52:02.0000] <Dashiva>
iPads are the devil's attempt at stone tablets

[01:52:03.0000] <vininim>
zcorpan: suppose I have a nav.html, boilerplater html and a nav block, and I want to insert this nav block on other pages.

[01:53:00.0000] <othermaciej>
Philip`: what is non-ASCII about it?

[01:53:01.0000] <zcorpan>
vininim: use server-side include

[01:53:02.0000] <Philip`>
othermaciej: The é

[01:54:00.0000] <vininim>
was going to use that as last resort, javascripting seems more fun!

[01:54:01.0000] <webben>
vininim: Less robust.

[01:55:00.0000] <zcorpan>
vininim: google won't be able to navigate your site if you include the nav with js

[01:55:01.0000] <zcorpan>
vininim: not to mention users that don't have js enabled

[01:56:00.0000] <webben>
or anyone in an untested browser where your JS happens to fail.

[01:56:01.0000] <MikeSmith>
Henrik Levkowetz deserves some serious props for setting up and maintaining the tools.ietf.org HTML stuff

[01:57:00.0000] <annevk>
why CSSImportRule has .media and .href is beyond me

[01:57:01.0000] <annevk>
grmbl

[01:57:02.0000] <othermaciej>
annevk: maybe the next t-shirt should be like a che guevara t-shirt but with Hixie instead

[01:57:03.0000] <othermaciej>
or would that bee too much out of MLW's playbook?

[01:57:04.0000] <annevk>
heh

[01:57:05.0000] <hsivonen>
do I need to turn of JavaScript to read the page Roy linked to without registering with Day Software?

[01:58:00.0000] <annevk>
what link?

[01:58:01.0000] <hsivonen>
http://dev.day.com/docs/en/cq/current/administering/multi_site_manager.html

[01:59:00.0000] <othermaciej>
hsivonen: that seems to show up ok for me without registering

[01:59:01.0000] <hsivonen>
hmm. appartly, the box can be dismissed by logging in with empty credentials

[01:59:02.0000] <hsivonen>
*apparently

[01:59:03.0000] <zcorpan>
Philip`: was my suggested grep completely broken?

[02:00:00.0000] <othermaciej>
hsivonen: oh interesting - I only got the annoyance box the second time I visited the page

[02:00:01.0000] <Philip`>
zcorpan: Don't know - I haven't got around to looking at it

[02:00:02.0000] <Philip`>
though I haven't completely forgotten it yet

[02:00:03.0000] <zcorpan>
ok

[02:02:00.0000] <MikeSmith>
othermaciej: MLW is way ahead of you, as usual

[02:02:01.0000] <hsivonen>
(I would recommend avoiding putting people engaged in violent activities on t-shirts that are supposed to convey a positive marketing message)

[02:02:02.0000] <MikeSmith>
he's already got a great T-shirt design - http://2.bp.blogspot.com/_27WEu4KHCYU/S5oK11NyJQI/AAAAAAAAAT0/nxCrMZB8sjw/s1600-h/microdata.gif

[02:03:00.0000] <othermaciej>
MikeSmith: not as iconic as Che, but sure, Microdata is Communism, I think that's obvious to anyone

[02:03:01.0000] <annevk>
I get a register overlay on that day.com site

[02:03:02.0000] <MikeSmith>
maybe MLW is Dick Cheney's daughter Liz

[02:04:00.0000] <annevk>
ctrl+w

[02:04:01.0000] <othermaciej>
in that case we should remake the Obama "Hope" poster

[02:04:02.0000] <zcorpan>
othermaciej: Che?

[02:04:03.0000] <othermaciej>
to signal our solidarity with the worker

[02:05:00.0000] <othermaciej>
zcorpan: this gentleman, who is a popular staple of hipster t-shirts: http://www.radiofuturamusic.com/v2/images/categories/TshirtCheRed.jpg

[02:05:01.0000] <othermaciej>
(probably most people wearing him on a shirt have no idea that he killed huge numbers of people)

[02:06:00.0000] <zcorpan>
ah

[02:06:01.0000] <annevk>
the onion variant is sort of funny, though not as good as their other t-shirts: http://hideyourarms.com/wp-content/uploads/2009/07/437.jpg

[02:07:00.0000] <othermaciej>
heh, yeah

[02:07:01.0000] <Dashiva>
othermaciej: They were probably bad people anyway

[02:10:00.0000] <MikeSmith>
I think what we really need is some kind of image that is iconic but that won't offend anybody; for example, an image of Jesus on the cross, with text underneath saying, "Jesus died for your XHTML sins."

[02:12:00.0000] <MikeSmith>
http://www.flickr.com/photos/mrlastweek/3730602918/ would make a great actual T-shirt

[02:12:01.0000] <MikeSmith>
I say we just go with that

[02:12:02.0000] <othermaciej>
how about a picture of Muhammad with HTML tags in his turban? 

[02:12:03.0000] <MikeSmith>
now you're thinking

[02:12:04.0000] <MikeSmith>
some excellent brainstorming we go going here

[02:12:05.0000] <Philip`>
/me likes the http://myserendipityblog.files.wordpress.com/2009/04/in-the-loop-0051.jpg version of the Hope poster

[02:13:00.0000] <Dashiva>
MikeSmith: Replace his face with someone else?

[02:13:01.0000] <MikeSmith>
Dashiva: <img hidden>

[02:13:02.0000] <MikeSmith>
Hixie looks like Sabretooth in that flickr T-shirt

[02:14:00.0000] <MikeSmith>
the brothers Wolverine, Sabretooth, Hixie

[02:14:01.0000] <Dashiva>
"XHTML2 died for your well-formedness errors"

[02:14:02.0000] <MikeSmith>
Dashiva: sublime

[02:15:00.0000] <MikeSmith>
hey maybe #whatwg should be the "non-X-Men"

[02:15:01.0000] <MikeSmith>
put a Magneto helmet on Hixie

[02:16:00.0000] <Dashiva>
Fighting Professor XHTML? :P

[02:16:01.0000] <othermaciej>
the Brotherhood of Markup Mutants?

[02:17:00.0000] <Dashiva>
Mutant Markup is sort of appropriate

[02:17:01.0000] <Dashiva>
All kinds of random errors, but some thrive and others fail to parse properly

[02:17:02.0000] <Dashiva>
Together, they drive the browser conformance criteria forward

[02:21:00.0000] <zcorpan>
Markup Mutant Ninja Microdata

[02:33:00.0000] <asmodai>
Mmm, interesting post by Roland Steiner on the ruby subject @wwwint

[02:34:00.0000] <jgraham>
zcorpan: I now have the theme tune fron Teenage Mutant Hero Turtles in my head, damn you

[02:34:01.0000] <workmad3>
'heros in a half-shell, TURTLE POWER!'

[02:34:02.0000] <workmad3>
or was it hard shell?

[02:34:03.0000] <workmad3>
been too long :)

[02:35:00.0000] <workmad3>
'Thundercats are on the move, Thundercats are loose'

[02:35:01.0000] <zcorpan>
jgraham: now i know what to use as my ring tone

[02:35:02.0000] <annevk>
Lord of Markup

[02:35:03.0000] <asmodai>
workmad3: half shell I think

[02:35:04.0000] <asmodai>
jgraham: You must be British btw

[02:35:05.0000] <asmodai>
jgraham: Since only the UK did the Ninja -> Hero conversion for what I remember :)

[02:35:06.0000] <annevk>
http://lists.w3.org/Archives/Public/www-international/2010JanMar/0143.html

[02:36:00.0000] <jgraham>
zcorpan: Only if you want your phone brutally bludgeoned

[02:36:01.0000] <jgraham>
asmodai: Indeed

[02:36:02.0000] <workmad3>
asmodai: I think it was actually the cartoon / film difference

[02:36:03.0000] <asmodai>
annevk: Ah yes, apologies for not providing a link.

[02:36:04.0000] <workmad3>
cartoon = hero turtles, films = ninja turtles

[02:36:05.0000] <jgraham>
workmad3: I'm pretty sure the cartoon was Ninja in the US and Hero in the UK

[02:36:06.0000] <jgraham>
(I have no idea why)

[02:36:07.0000] <workmad3>
ah :)

[02:37:00.0000] <workmad3>
spotty recollections of my youth :)

[02:37:01.0000] <asmodai>
The Teenage Mutant Ninja Turtles (or simply Ninja Turtles, and previously known in the United Kingdom, Germany, Austria, Ireland  and Scandinavia as Teenage Mutant Hero Turtles)

[02:37:02.0000] <asmodai>
http://en.wikipedia.org/wiki/Teenage_Mutant_Ninja_Turtles#Teenage_Mutant_Hero_Turtles

[02:37:03.0000] <asmodai>
Anyway

[02:37:04.0000] <workmad3>
(and besides, I didn't care what it was called in the states when I was 10 :P )

[02:37:05.0000] <asmodai>
annevk: how far off is that vertical text implementation for most UAs?

[02:37:06.0000] <othermaciej>
apparently the word "Ninja" was too violent for the UK (which strikes me as hella weird)

[02:38:00.0000] <asmodai>
annevk: I am not even sure if the HTML5 spec touches that, does it?

[02:39:00.0000] <workmad3>
all this discussion leads to one important question though... who was your favourite hero/ninja turtle? :)

[02:39:01.0000] <annevk>
asmodai, far and not it doesn't (it shouldn't either)

[02:39:02.0000] <annevk>
s/not/no/

[02:39:03.0000] <annevk>
well, IE has support

[02:40:00.0000] <annevk>
it's clear to me people are actually waiting for vertical text support

[02:40:01.0000] <asmodai>
annevk: It would make some things easier, yes. :)

[02:40:02.0000] <jgraham>
workmad3: No it doesn't, it leads to the question if WHATWG/HTMLWG members were turtles, who would be April O'Neill

[02:40:03.0000] <annevk>
if we ever do it there would be a large amount of bugs and it would be vastly undertested :/

[02:40:04.0000] <annevk>
it's not clear to me, damn it

[02:40:05.0000] <workmad3>
jgraham: that's also an important one, along with who would be shredder? :)

[02:40:06.0000] <asmodai>
Especially if you consider the case of Mongolian and traditional writing of Japanese and/or Chinese and such languages.

[02:40:07.0000] <annevk>
/me blames the keyboard

[02:41:00.0000] <annevk>
asmodai, but Japanese Web designers have not really been asking for it

[02:41:01.0000] <annevk>
so far it's just the print industry

[02:41:02.0000] <asmodai>
annevk: True.

[02:41:03.0000] <annevk>
and I don't really care about print

[02:41:04.0000] <asmodai>
annevk: Awww, why not?

[02:42:00.0000] <annevk>
because printing of the Web is dying

[02:42:01.0000] <asmodai>
/me is a dying breed. :(

[02:43:00.0000] <annevk>
it's also wasteful imo

[02:43:01.0000] <annevk>
mind you, howcome strongly disagrees with me

[02:43:02.0000] <workmad3>
well, when you need a couple dozen rainforests to print out the entire thing, it's not hard to see why :P

[02:43:03.0000] <asmodai>
I hate reading on-screen.

[02:43:04.0000] <asmodai>
Especially since the DPI is still lagging far behind printed materials.

[02:43:05.0000] <othermaciej>
printing is obsolete

[02:44:00.0000] <daedb>
http://www.penny-arcade.com/comic/2002/7/1/

[03:25:00.0000] <Lachy>
http://www.youtube.com/watch?v=saqO_ZqX6uY

[03:53:00.0000] <zcorpan>
wonder why http://www.whatwg.org/issues/data.html doesn't work in opera

[03:53:01.0000] <zcorpan>
only shows the yellow bars and the text for me

[03:57:00.0000] <Philip`>
zcorpan: Looks like just the averages lines are missing for me

[03:58:00.0000] <Philip`>
which is probably http://philip.html5.org/tests/canvas/suite/tests/results.html#2d.path.lineTo.ensuresubpath.2

[03:59:00.0000] <zcorpan>
all lines are missing for me, but my build might be a bit old

[04:00:00.0000] <zcorpan>
Hixie: the graph looks a lot better if the months are abbreviated to three-letter labels

[04:00:01.0000] <Philip`>
or maybe it's not

[04:01:00.0000] <Philip`>
Oh, yes, it is that

[04:01:01.0000] <Philip`>
Opera foolishly implemented what the spec said, then the spec changed so now Opera is broken

[04:02:00.0000] <Philip`>
(Adding some moveTo()s at the beginning of the line-drawing functions fixes it in Opera)

[04:03:00.0000] <Philip`>
When I said "averages lines" I actually meant "averages lines and also the lines which they are averages of"

[05:44:00.0000] <zcorpan>
<[a-zA-Z][^>]*=\s*(("[^"]*)|('[^']*)|([^"']*))&(AElig|AMP|Aacute|Acirc|Agrave|Aring|Atilde|Auml|COPY|Ccedil|ETH|Eacute|Ecirc|Egrave|Euml|GT|Iacute|Icirc|Igrave|Iuml|LT|Ntilde|Oacute|Ocirc|Ograve|Oslash|Otilde|Ouml|QUOT|REG|THORN|Uacute|Ucirc|Ugrave|Uuml|Yacute|aacute|acirc|acute|aelig|agrave|amp|aring|atilde|auml|brvbar|ccedil|cedil|cent|copy|curren|deg|divide|eacute|ecirc|egrave|eth|euml|frac12|frac14|frac34|gt|iacute|i

[05:44:01.0000] <zcorpan>
uml|laquo|lt|macr|micro|middot|nbsp|not|ntilde|oacute|ocirc|ograve|ordf|ordm|oslash|otilde|ouml|para|plusmn|pound|quot|raquo|reg|sect|shy|sup1|sup2|sup3|szlig|thorn|times|uacute|ucirc|ugrave|uml|uuml|yacute|yen|yuml)[^;a-zA-Z0-9]

[05:45:00.0000] <zcorpan>
Philip`: the above should find no-semicolon entities in attribute value that's not followed by a-zA-Z0-9

[05:45:01.0000] <zcorpan>
Philip`: could you run a grep for that? :)

[06:57:00.0000] <Philip`>
zcorpan: Remind me in several hours :-)

[07:05:00.0000] <zcorpan>
Philip`: i might not be around then

[07:06:00.0000] <Philip`>
/me will try to remember it himself

[07:23:00.0000] <hsivonen>
has Hixie's Live Dom Viewer changed recently to push the content to the parser differently and less often?

[07:50:00.0000] <Dashiva>
ASCII not what the spec can do for you, ask what you can do for the spec

[07:57:00.0000] <annevk>
the Live DOM Viewer is buggy in Chrome

[07:57:01.0000] <annevk>
especially the w() function does not seem to execute instantly like I expect and like it does in other browsers

[07:57:02.0000] <annevk>
Hixie, ^^

[08:00:00.0000] <Dashiva>
So the main arguments for the RFC format seems to be "works for my 30-year old printer" and "it's eco-friendly"

[08:00:01.0000] <Dashiva>
Surely this is the perfect organization to lead us into the future

[08:12:00.0000] <jgraham>
Dashiva: You seem to be concentrating on the logical arguments. The _main_ argument seems to be "get off my lawn"

[08:14:00.0000] <Dashiva>
Good thing there's nobody around to blow up the ascii for names thing as racism and western imperialism

[08:16:00.0000] <jgraham>
I feel I should insert some sort of qualifier before "logical". Like "almost, but not"

[08:19:00.0000] <ment>
from my programmer POV, ASCII manuals and references are much more convenient than PDFs

[08:20:00.0000] <Dashiva>
That's a false dilemma

[08:21:00.0000] <Dashiva>
(ASCII is a charset, not a file format)

[08:21:01.0000] <ment>
erm, plain text

[08:22:00.0000] <ment>
i thought you were arguing about format rather than charset

[08:22:01.0000] <Dashiva>
Both, really

[08:23:00.0000] <ment>
well, with utf-8 i can get my cp437 box characters back

[08:23:01.0000] <jgraham>
And with HTML I can have hyperlinks

[08:24:00.0000] <jgraham>
Seriously I don't know why PDF is even being discussed

[08:24:01.0000] <jgraham>
The question should be phrased as "why aren't RFCs being published as (utf-8) HTML"

[08:25:00.0000] <Dashiva>
Because that won't print on my 30-years old printer using tools made before HTML came into being

[08:26:00.0000] <jgraham>
To which the correct response is "so?"

[08:27:00.0000] <Dashiva>
"Take your technical arguments elsewhere, whippersnapper"

[08:29:00.0000] <jgraham>
Anyway plain text RFCs are inconvenient too transcribe by hand-of-monk with illuminated letters since the number of characters per line is highly variable in that case, and the hardcoded headers and page breaks confuse the monks no end

[08:29:01.0000] <jgraham>
s/too/to/

[08:30:00.0000] <Dashiva>
hand-of-monk is obsolete since RFC 1217

[08:31:00.0000] <ment>
(replaced by hand-of-monkey)

[08:50:00.0000] <Philip`>
If you allow people to write specs in HTML, soon they'll be using SVG graphics and embedding JavaScript in every page

[08:51:00.0000] <Philip`>
and there would be no end to the madness

[08:54:00.0000] <Philip`>
Also, you can bet a lot of people would write their HTML-formatted specs in Microsoft Word

[08:55:00.0000] <annevk>
yeah, because that is exactly what happened at the W3C

[08:56:00.0000] <jgraham>
Gosh, if only technology had advanced to the level where we could automatically check a file against a set of rules before publication. Then one could enforce whatever rules one pleased and avoid that whole issue

[08:57:00.0000] <jgraham>
It's sad that no one has managed to develop such a system

[08:59:00.0000] <annevk>
governments would pay millions

[09:29:00.0000] <AryehGregor>
Ah, so Google spiders finally support microdata.

[09:33:00.0000] <Philip`>
Presumably with bugs, though

[09:33:01.0000] <boblet>
Philip`: maybe Microdata is so simple they don’t have any bugs?

[09:34:00.0000] <boblet>
;)

[09:39:00.0000] <Dashiva>
Philip`: I thought you would have a test case showing the bugs by now

[09:48:00.0000] <Philip`>
Dashiva: I didn't really care enough

[09:48:01.0000] <Philip`>
but things like http://www.google.com/webmasters/tools/richsnippets?url=http%3A%2F%2Fphilip.html5.org%2Fdemos%2Fmicrodata%2Fsibling-p.html are obvious bugs

[09:48:02.0000] <Dashiva>
So your bias is coming through :P

[09:48:03.0000] <Philip`>
(The same bugs affect the RDFa parsing)

[09:48:04.0000] <Philip`>
(because it's not using anything like an HTML parser when parsing the HTML)

[09:49:00.0000] <AryehGregor>
Interesting.

[09:49:01.0000] <Dashiva>
Well, that's equal opportunity at least

[09:49:02.0000] <AryehGregor>
Hixie had sort of vaguely hinted at some point that Google might possibly be using an HTML5 parser for some things internally.

[09:49:03.0000] <Philip`>
Clearly this isn't one of those things

[09:50:00.0000] <Dashiva>
This isn't a HTMLx parser for any value of x

[09:50:01.0000] <Philip`>
It seemed kind of like XML5 minus the complex bits (doctypes etc) when I last looked

[09:50:02.0000] <AryehGregor>
There's no such thing as an HTMLx parser for x < 5, at least if we're confining our discussion to real life.

[09:51:00.0000] <Dashiva>
Philip`: And minus the well-formedness?

[09:51:01.0000] <annevk>
I thought Ian once said they use some kind of graph representation

[09:51:02.0000] <Philip`>
AryehGregor: There are parsers which attempt to be compatible with the syntax defined by HTMLx for x < 5, and they'll all handle <p>foo<p>bar correctly since that's a basic part of HTML syntax

[09:52:00.0000] <Philip`>
Dashiva: I thought XML5 didn't do well-formedness

[09:52:01.0000] <Dashiva>
I thought it did, but with relaxed constraints

[09:52:02.0000] <AryehGregor>
Google doesn't handle <p>foo<p>bar properly?  o_O

[09:52:03.0000] <Dashiva>
Not in their snippets tool

[09:52:04.0000] <Philip`>
I assume they use a different parser for serious things like search indexing

[09:53:00.0000] <Philip`>
annevk: I vaguely remember him just saying they parse into some kind of internal representation that isn't RDF

[09:54:00.0000] <Philip`>
(Presumably they use the same one for RDFa and microdata and microformats, so just the parser is changed)

[09:55:00.0000] <annevk>
Philip`, I meant for HTML in general

[09:55:01.0000] <annevk>
Philip`, not this stuff

[09:55:02.0000] <annevk>
Philip`, because they want streaming and do not want to modify a tree

[09:56:00.0000] <annevk>
Philip`, but do want to recognize the effects of <i><b>test</i>test</b>

[09:58:00.0000] <Philip`>
Oh

[09:58:01.0000] <Philip`>
/me is unaware of that

[11:02:00.0000] <annevk>
MikeSmithXX, I forwarded someone in your direction regarding HTML5; she wanted to chat to some other people involved

[11:22:00.0000] <franksalim>
MikeSmithXX, I see that book, too :-)

[12:09:00.0000] <zcorpan>
Philip`: reminder :)

[12:39:00.0000] <hsivonen>
/me notes that Google's link to "Official microdata specification" points to whatwg.org--not w3.org

[12:49:00.0000] <hsivonen>
Hixie: why doesn't http://www.whatwg.org/specs/web-apps/current-work/#current-document-readiness have the uninitialized and interactive states?

[14:03:00.0000] <asmodai>
heh, sweet http://desandro.com/articles/opera-logo-css/

[14:34:00.0000] <annevk>
now all that's left between me and baked cookies is the HTTP WG

[14:36:00.0000] <jgraham>
annevk: Presumably he might poison the cookies to reduce dissent

[15:13:00.0000] <zcorpan>
/me finds http://www.robodesign.ro/mihai/blog/html5-demo-video-and-canvas

[15:16:00.0000] <annevk>
"I could move only a really small part of the code into a worker, and the performance penalty of sending the pixels in a message from the main thread to the worker thread is too significant. In the end, it runs quite slower than the script which does not use any web worker." seems relevant

[15:17:00.0000] <ap>
annevk: worked better for http://nerget.com/rayjs-mt/rayjs.html


2010-03-13
[16:19:00.0000] <Philip`>
Hixie: You forgot a comma after adding "Adam Hepton"

[16:19:01.0000] <Hixie>
thanks

[16:31:00.0000] <JonathanNeal>
Phase #1 complete @ http://sandbox.thewikies.com/html5-experiments/phase-01.html woohoo

[17:01:00.0000] <karlcow>
http://gist.github.com/330963

[17:01:01.0000] <karlcow>
 Test for browser redirection handling WRT URI fragments, quick and dirty.  

[17:20:00.0000] <zcorpan>
Hixie: search for 'hidden' in http://my.opera.com/core/blog/2010/03/03/everything-you-need-to-know-about-html5-video-and-audio-2

[19:25:00.0000] <mpilgrim>
any mac/safari users around?

[19:25:01.0000] <mpilgrim>
i need to test something and i'm not near any macs

[19:26:00.0000] <miketaylr>
im on a mac

[19:27:00.0000] <miketaylr>
w/ safari 4.0.4

[19:27:01.0000] <mpilgrim>
can you install ClickToFlash and test something for me?

[19:27:02.0000] <miketaylr>
sure, one sec

[19:27:03.0000] <mpilgrim>
http://rentzsch.github.com/clicktoflash/

[19:29:00.0000] <miketaylr>
k, installed

[19:30:00.0000] <miketaylr>
mua ha ha http://miketaylr.com/post/b8950f24.png

[19:31:00.0000] <mpilgrim>
ha

[19:31:01.0000] <mpilgrim>
ok, go to http://diveintomark.org/tmp/flashblock-detect.html

[19:31:02.0000] <mpilgrim>
tell me if it says "a Flash blocker was detected"

[19:32:00.0000] <ttepasse>
Here: no flash blocker was detected.

[19:32:01.0000] <miketaylr>
it says: no Flash blocker was detected

[19:32:02.0000] <mpilgrim>
damn it

[19:32:03.0000] <miketaylr>
but, http://miketaylr.com/post/37e864ed.png

[19:32:04.0000] <ttepasse>
(Leopard, Safari 4.x, CtF 1.5.3)

[19:32:05.0000] <miketaylr>
looks blocked to me :p

[19:32:06.0000] <mpilgrim>
yeah

[19:34:00.0000] <mpilgrim>
force-refresh now (shift+refresh button i think)

[19:35:00.0000] <miketaylr>
a Flash blocker was detected

[19:35:01.0000] <miketaylr>
\o/

[19:35:02.0000] <mpilgrim>
aha

[19:39:00.0000] <mpilgrim>
ok, force-refresh again to make sure i didn't just break anything

[19:40:00.0000] <miketaylr>
nope, still good

[19:40:01.0000] <mpilgrim>
ok, final test

[19:40:02.0000] <mpilgrim>
go into the clicktoflash preferences (under your application menu IIRC) and whitelist diveintomark.org

[19:41:00.0000] <mpilgrim>
then force-refresh

[19:41:01.0000] <mpilgrim>
should now say "no flash blocker was detected"

[19:41:02.0000] <mpilgrim>
(after about 5 seconds)

[19:42:00.0000] <miketaylr>
hmmm, still a Flash blocker was detected

[19:42:01.0000] <mpilgrim>
oh damn

[19:47:00.0000] <mpilgrim>
oh bloody hell

[19:47:01.0000] <mpilgrim>
i know what's wrong

[19:47:02.0000] <mpilgrim>
but i don't know how to fix it

[19:49:00.0000] <mpilgrim>
BOY THIS SURE WOULD BE EASIER IF WEBKIT WOULD JUST GO AHEAD AND IMPLEMENT DOMATTRMODIFIED

[19:50:00.0000] <miketaylr>
>:|

[19:53:00.0000] <mpilgrim>
ok, last try

[19:53:01.0000] <mpilgrim>
force-refresh again

[19:53:02.0000] <mpilgrim>
with diveintomark.org still whitelisted

[19:55:00.0000] <mpilgrim>
/me throws a sheep at miketaylr

[19:56:00.0000] <miketaylr>
/me ducks

[19:56:01.0000] <miketaylr>
whitelisted, no Flash blocker was detected

[19:56:02.0000] <mpilgrim>
that's good

[19:56:03.0000] <mpilgrim>
now clear the whitelist

[19:56:04.0000] <miketaylr>
k

[19:56:05.0000] <mpilgrim>
and refresh

[19:57:00.0000] <miketaylr>
a Flash blocker was detected

[19:57:01.0000] <mpilgrim>
HOORAY

[19:57:02.0000] <miketaylr>
, for great victory

[19:57:03.0000] <mpilgrim>
AFAIK, this is the world's first flashblock detector that works in 3 browsers

[19:58:00.0000] <mpilgrim>
er, world's first PURE JAVASCRIPT flashblock detector

[19:58:01.0000] <miketaylr>
oh, sweet

[19:58:02.0000] <mpilgrim>
thanks for your help

[19:58:03.0000] <miketaylr>
np, sir

[20:46:00.0000] <mpilgrim>
http://code.google.com/p/flashblockdetector/

[20:46:01.0000] <mpilgrim>
going to sleep now

[02:00:00.0000] <annevk>
/me reads http://lists.whatwg.org/pipermail/help-whatwg.org/2010-March/000432.html by coincidence

[02:00:01.0000] <annevk>
/me is not subscribed to help⊙wo

[02:09:00.0000] <asmodai>
annevk: funny if you don't monitor it :)

[04:13:00.0000] <hsivonen>
what's the motivating use case behind http://code.google.com/p/flashblockdetector/ ? analytics?

[04:16:00.0000] <Lachy>
I don't know for sure, but one use case might be for advertisers to serve an alternate image instead of flash if they can detect the user won't see the flash.

[04:17:00.0000] <Philip`>
"One good use of this would be to inform users that they need to opt-in to Flash for unobtrusive Flash enhancements (such as invisible audio players) to work on that page."

[05:04:00.0000] <hsivonen>
I thought you were supposed to use <audio> for that in all the browsers that have supported Flash blockers

[05:06:00.0000] <Philip`>
Writing something once in <audio> and once in Flash seems like more work than writing it once in Flash and telling users to turn Flash on

[06:47:00.0000] <boblet>
what’s the difference between rel="license" and <small> if the license info is linked?

[06:49:00.0000] <boblet>
aah, from the example under link type license it seems they’re complementary; wrap rel=license link in small

[06:49:01.0000] <boblet>
thanks :)


2010-03-14
[23:23:00.0000] <Tripknotix>
hey everyone

[23:23:01.0000] <Tripknotix>
long time no see

[23:23:02.0000] <Tripknotix>
i had a question on whether or not webgl or ogl would be implemented in html5, or if it will continue to only be pushed by the browser companies themselves

[23:26:00.0000] <Hixie>
i don't understand the difference

[23:30:00.0000] <othermaciej>
Tripknotix: html5 is a spec, not an implementation

[23:30:01.0000] <othermaciej>
so nothing will be "implemented in html5"

[23:31:00.0000] <othermaciej>
if you're asking whether the WebGL spec will be incorporated into the HTML5 specification, rather than being published as a separate specification by a different standards body, I believe the answer is "no".

[23:31:01.0000] <Tripknotix>
haha

[23:31:02.0000] <Tripknotix>
are you on the whatwg team?

[23:31:03.0000] <Tripknotix>
because that is interesting news =0

[23:31:04.0000] <Tripknotix>
=)

[23:31:05.0000] <Tripknotix>
have not gotten a positive answer before

[23:32:00.0000] <othermaciej>
let me correct that and say the answer is "probably not"

[23:32:01.0000] <othermaciej>
in other words, WebGL will most likely continue to be a separate spec, published by Khronos

[23:33:00.0000] <othermaciej>
although I would not describe that as "only [] pushed by the browser companies themselves"

[23:33:01.0000] <Tripknotix>
well you know the difference

[23:33:02.0000] <Tripknotix>
=)

[23:38:00.0000] <Tripknotix>
thanks for the help

[04:53:00.0000] <MikeSmith>
Hixie: interesting discussion about ruby annotations continues on the HTML5 Japanese IG list -

[04:53:01.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-html-ig-jp/2010Mar/thread.html#msg7

[04:54:00.0000] <MikeSmith>
at some point we'll probably want to post a summary of that to the public-html list

[04:58:00.0000] <MikeSmith>
note that Shinyu Murakami is an implementor of an XSL-FO engine that also now has CSS print support (like Prince)

[04:58:01.0000] <MikeSmith>
hmm, I guess that most of the substantive discussion there is all in English

[04:59:00.0000] <MikeSmith>
except for the examples

[04:59:01.0000] <MikeSmith>
Murakami-san's latest message is more about meta issues

[05:01:00.0000] <MikeSmith>
othermaciej: on Nikunj asking about review of the IndexedDB spec, I guess you've seen that Jeremy Orlow seems to have started on an implementation

[05:02:00.0000] <othermaciej>
MikeSmith: I've seen that he has been talking about it

[05:02:01.0000] <MikeSmith>
othermaciej: he's made at least one or two related commits so far, I think

[05:03:00.0000] <MikeSmith>
though the commit messages described them as "baby steps" or something

[05:03:01.0000] <othermaciej>
I am not sure why Nikunj asked me and not Mozilla

[05:04:00.0000] <othermaciej>
I am also not sure Apple will have substantially different feedback than Google, since if we ship it we'll likely use the same implementation

[05:04:01.0000] <MikeSmith>
maybe he asked because he's already talked with Mozilla off-list

[05:05:00.0000] <othermaciej>
(though I don't believe Jeremy consulted with anyone at Apple about his spec feedback or his nascent implementation)

[05:06:00.0000] <MikeSmith>
othermaciej: I think he may have actually written a design doc

[05:06:01.0000] <MikeSmith>
/me goes to look

[05:06:02.0000] <MikeSmith>
http://sites.google.com/a/chromium.org/dev/developers/design-documents/indexeddb

[05:08:00.0000] <MikeSmith>
maybe he is just planning to wait to get that posted to webkit-dev after more discussion within the chromium team

[05:09:00.0000] <othermaciej>
he did post a comment on webkit-dev, but not a pointer to the design document

[05:09:01.0000] <othermaciej>
and it was more of an "I'm doing it" comment than an "I want your input" comment

[05:35:00.0000] <MikeSmith>
othermaciej: in the recent public-html thread about conformance, you mention both authoring conformance and document conformance. Do you mean to make any distinction between those two, or are you using them synonymously?

[05:36:00.0000] <othermaciej>
MikeSmith: I meant them to be synonyms, but I may have used the in confusing ways

[05:36:01.0000] <MikeSmith>
no, not really confusing

[05:36:02.0000] <othermaciej>
MikeSmith: the biggest surprise to me was how many issues are purely editorial

[05:36:03.0000] <MikeSmith>
I just wasn't sure if it was intentional

[05:36:04.0000] <MikeSmith>
othermaciej: yeah, I saw your note about that

[05:37:00.0000] <MikeSmith>
given that the are just editorial, it would see likely that we could get agreement about those more easily

[05:38:00.0000] <MikeSmith>
or maybe I'm just being optimistic

[05:38:01.0000] <MikeSmith>
overly optimistic

[05:40:00.0000] <othermaciej>
if it was easy to get agreement about those topics, they wouldn't be tracker issues...

[05:41:00.0000] <MikeSmith>
true

[05:41:01.0000] <MikeSmith>
but there are some cases of bugs getting escalated prematurely

[05:41:02.0000] <MikeSmith>
"premature escalation" is the technical process term, I think

[05:41:03.0000] <MikeSmith>
from Robert's Rules of Order

[05:42:00.0000] <othermaciej>
I haven't heard the term

[05:42:01.0000] <othermaciej>
what do you mean exactly?

[05:42:02.0000] <MikeSmith>
I'm joking about the term

[05:42:03.0000] <othermaciej>
I see

[05:43:00.0000] <MikeSmith>
but I mean that some bugs have gotten escalated without much discussion

[05:44:00.0000] <MikeSmith>
or to look at it from another perspective, they maybe should not have been been moved to resolved=wontfix without more discussion

[05:44:01.0000] <MikeSmith>
but then some of them probably should not have been raised to begin with

[05:46:00.0000] <othermaciej>
I do think that for many of those editorial issues, there wasn't a good faith effort to work out an acceptable compromise

[05:46:01.0000] <othermaciej>
or to determine if the issue actually matters to anyone besides the original before escalating

[05:46:02.0000] <othermaciej>
not that those are required, but it would certainly save me work if people did that before escalating

[05:47:00.0000] <MikeSmith>
yeah

[05:50:00.0000] <othermaciej>
oh, that reminds me, I promised to move old bugs to VERIFIED

[05:52:00.0000] <othermaciej>
MikeSmith: did I ever show you the graph I made of our bug in/out rate?

[05:53:00.0000] <othermaciej>
http://spreadsheets.google.com/ccc?key=0AoCAfo_LQ5_kdFFWWmpCMWxsLUN2TW9VYi1uNEJGenc&hl=en

[05:53:01.0000] <othermaciej>
I'm still not sure what happened in August that so massively increased use of the bug tracker

[05:53:02.0000] <MikeSmith>
othermaciej: I had not seen this yet

[05:55:00.0000] <othermaciej>
if you look at this chart, and also at issue in/out rates, you would conclude that we didn't start serious work until August 2009

[05:55:01.0000] <Philip`>
August was when the automatic bug submission form was added to the spec, which might explain it

[05:56:00.0000] <MikeSmith>
Philip`: well, that would be one factor

[05:56:01.0000] <othermaciej>
ok, I thought that might be it (part of it, anyway)

[05:56:02.0000] <MikeSmith>
the other factor is that's when some people started putting serious effort into bug resolution

[05:56:03.0000] <MikeSmith>
e.g., Maciej

[05:57:00.0000] <Philip`>
Looks like over 50% of reported bugs come from that tool

[05:57:01.0000] <MikeSmith>
but most of those are typically less controversial

[05:58:00.0000] <othermaciej>
of those 211 bugs, 183 were originated by contributor⊙wo

[05:59:00.0000] <othermaciej>
conclusion: if we stopped publishing the WHATWG copy of the spec, we'd get 90% less feedback in bugzilla

[05:59:01.0000] <MikeSmith>
heh

[05:59:02.0000] <othermaciej>
total filed since last august: 1463; total filed by contributor⊙wo: 1097

[06:00:00.0000] <othermaciej>
ok, so only 75%

[06:00:01.0000] <MikeSmith>
still, the biggest factor, for sure

[06:00:02.0000] <MikeSmith>
but not the most interesting factor

[06:01:00.0000] <MikeSmith>
I think the life cycle of most of those contributor⊙wo bugs is relatively short

[06:01:01.0000] <othermaciej>
right now the life cycle of nearly all bugs is short

[06:01:02.0000] <MikeSmith>
some of them are just background noise

[06:02:00.0000] <othermaciej>
I don't have a problem with lots of incoming bugs as long as the rate of resolving them keeps up

[06:02:01.0000] <MikeSmith>
also, a non-insignificant percentage of those contributor⊙wo bugs are spam

[06:02:02.0000] <othermaciej>
I think lots of incoming bugs is a very postive sign

[06:02:03.0000] <othermaciej>
very few bugs get closed as INVALID each month, so the spam proportion is pretty low

[06:02:04.0000] <MikeSmith>
yeah

[06:03:00.0000] <othermaciej>
I think there's people who wish the whatwg copy of the spec did not exist, and they may be seriously underestimating the degree to which it helps us get good technical feedback, even in the HTML WG

[06:04:00.0000] <othermaciej>
I did not expect before looking that so many of our bugs came from the comment form

[06:04:01.0000] <othermaciej>
I personally hate the comment form because it makes it too hard to track bugs I filed, but clearly it is a powerful tool for getting feedback

[06:05:00.0000] <MikeSmith>
as far as any increase of the non-contributor⊙wo bugs, I think a big part of the reason for that would be that we began to advertise bugzilla to the group as being an encouraged means for requesting spec changes

[06:05:01.0000] <Philip`>
If I couldn't use the comment form, I'd just send email instead (because that's easier for me than manually filing bugs)

[06:05:02.0000] <Philip`>
so it probably wouldn't have a significant effect on the amount of feedback I would send

[06:05:03.0000] <MikeSmith>
we should add a comment form to the W3C version

[06:06:00.0000] <MikeSmith>
too

[06:06:01.0000] <Philip`>
(I track bugs by appending "[pt]" to each one, so they're easy to find in a Bugzilla query)

[06:07:00.0000] <Philip`>
MikeSmith: Then you'd need to run a CGI script on the domains that the spec is hosted on (dev.w3.org and www.w3.org, I guess)

[06:07:01.0000] <Philip`>
Is that easy/possible?

[06:08:00.0000] <MikeSmith>
anything's possible

[06:08:01.0000] <MikeSmith>
but dev.w3.org is not set up for now, at least

[06:08:02.0000] <othermaciej>
let's see how bugzilla does by mass-marking 500 bugs

[06:08:03.0000] <MikeSmith>
Philip`: I think maybe www.w3.org is not either

[06:09:00.0000] <othermaciej>
Philip`: you'd probably send email to whatwg⊙wo, but in your case I imagine it would go to public-html instead of the whatwg list didn't exist

[06:14:00.0000] <othermaciej>
MikeSmith: non-comment-form bug rate did not go up until September, but it has been about 2.5 times the previous level

[06:20:00.0000] <MikeSmith>
I see

[06:23:00.0000] <MikeSmith>
othermaciej: did you intend to include issue 5758 in those you moved to resolved=verified?

[06:24:00.0000] <othermaciej>
MikeSmith: I just did a batch move on all bugs that were resolved before last October

[06:24:01.0000] <othermaciej>
it looks like that was one of them, so yes

[06:24:02.0000] <MikeSmith>
OK

[06:25:00.0000] <othermaciej>
looking at it now, it looks to me like it should me marked with TrackerIssue and associated with ISSUE-9

[06:25:01.0000] <MikeSmith>
yeah

[06:25:02.0000] <othermaciej>
(but still VERIFIED)

[06:26:00.0000] <othermaciej>
or reopened - I'm not really sure what to do with a LATER resolution

[06:26:01.0000] <othermaciej>
I wish our bugzilla didn't have LATER or REMIND because they don't make sense

[06:26:02.0000] <othermaciej>
I just count them as rejections when I make the monthly bug stats

[06:28:00.0000] <MikeSmith>
155 new bugmail messages in my inbox..

[06:34:00.0000] <nessy>
5758 should not be counted as a rejection, indeed

[06:35:00.0000] <nessy>
but alas, I don't think we really need an open bug to introduce media a11y - we all know it's required :)

[16:04:00.0000] <wycats>
who goes there?

[16:43:00.0000] <annevk>
geez, waddup with the bugmail spam?

[16:46:00.0000] <Dashiva>
Is it the sound of accessibility?

[16:46:01.0000] <Dashiva>
No, it's the sound of bugs being marked CLOSED, I guess

[16:52:00.0000] <annevk>
and they're gone from my inbox

[16:58:00.0000] <Dashiva>
Now we wait two weeks for 500 more emails from all those being marked NoReply


2010-03-15
[17:00:00.0000] <Philip`>
It's also the sound of people rediscovering bugs they'd forgotten about and deciding they shouldn't be allowed to die quietly

[17:03:00.0000] <Philip`>
/me wonders what CCN is

[17:03:01.0000] <Philip`>
(in relation to http://lists.w3.org/Archives/Public/public-html-bugzilla/2010Mar/0764.html)

[17:04:00.0000] <Philip`>
Oh, I guess it's http://www.ccnx.org/

[17:05:00.0000] <Dashiva>
It refers to http://lists.w3.org/Archives/Public/www-tag/2009Sep/0055.html at least

[17:05:01.0000] <Philip`>
Hmm, I think I saw Van Jacobson talking about that but I can't remember anything about it :-/

[17:05:02.0000] <Dashiva>
"content-centric networking" sounds like a buzzword without much real substance

[17:06:00.0000] <Philip`>
/me should have paid more attention, if it's going to replace IP

[17:07:00.0000] <Philip`>
Since he's one of the main developers of TCP, I think he knows a bit about the substance of networking

[17:09:00.0000] <Dashiva>
Lots of failed projects were done by people with excellent technical skills

[17:10:00.0000] <Dashiva>
These use cases seem to focus on money and control, with a little attribution on the side

[17:11:00.0000] <Philip`>
I believe it's "content-centric" because you care more about the content than about where it comes from - e.g. instead of sending packets to an IP address, you send requests to a content identifier and it can be satisfied by any router that has cached that content, or else it'll be forwarded towards a more authoritative source for that data

[17:12:00.0000] <Dashiva>
"We expect that most organizations will have a way to intake GPL code." :)

[17:12:01.0000] <Dashiva>
They've made a custom binary encoding of XML...

[17:14:00.0000] <Philip`>
(and you route based on longest-prefix matches of the content identifier, so it can efficiently handle hierarchical naming schemes but isn't strictly constrainted to a hierarchy, so it still gives freedom to separate content from location)

[17:16:00.0000] <Dashiva>
I wonder how they plan to translate from user desires to interest messages

[17:18:00.0000] <Philip`>
Users click desirable links on web pages which link to the content name, I'd guess

[17:18:01.0000] <Philip`>
or programs generate them automatically

[17:18:02.0000] <Philip`>
just like they do with URLs today

[17:19:00.0000] <Philip`>
only without today's underlying layers that collapse the whole rich URL structure into a 4-byte address of a host that contains that content

[17:20:00.0000] <Dashiva>
It seems like they're pulling the search engines into the transport layer

[17:21:00.0000] <Philip`>
Why does it seem like that?

[17:21:01.0000] <Philip`>
/me sees it as more like routing on URLs (plus caches and security)

[17:22:00.0000] <Dashiva>
Well, if the search engines remain a necessary part of the system, what is gained from losing the location-dependency?

[17:22:01.0000] <Dashiva>
Search results already span different locations

[17:22:02.0000] <Philip`>
(going by the CoNEXT paper at least)

[17:22:03.0000] <Dashiva>
P2P networks handle distributed downloads too

[17:24:00.0000] <Dashiva>
I'm watching a video presentation, he says they're NOT Google. That they do the inverse. Google is "content to names", they are "names to content"

[17:24:01.0000] <Dashiva>
Seems 'content' is rather overloaded as a term here

[17:25:00.0000] <Philip`>
If you use some extra service like a search engine to map identifiers onto locations, then higher layers don't have to worry about locations and you've got an emulation of location-independence and that should work okay

[17:25:01.0000] <Philip`>
but it'd be awfully inefficient since it's layered on top of an internet that's fundamentally designed around locations

[17:26:00.0000] <Philip`>
and you'd have to do a lot more communication than if the underlying layers were designed to support that properly

[17:27:00.0000] <Dashiva>
Doing a near-broadcast interest message, and then all the possible responders have to manage to alert each other so only one (ideally) responds

[17:27:01.0000] <Philip`>
P2P works for some things, but it doesn't help distribute the front page of a news site or videos on YouTube

[17:27:02.0000] <Dashiva>
Well, if you remove the user uploading, you get a CDN

[17:27:03.0000] <Philip`>
I suppose FreeNode supports distributing web pages but it has the problem of being rubbish

[17:27:04.0000] <Philip`>
Uh

[17:27:05.0000] <Philip`>
FreeNet

[17:27:06.0000] <Philip`>
or whatever it is

[17:28:00.0000] <Dashiva>
How would this system handle spam

[17:28:01.0000] <Dashiva>
You could reply to any interest message with Cialis :)

[17:30:00.0000] <Philip`>
By cryptographically signing content messages so that they're tied to the content name

[17:30:01.0000] <Dashiva>
But the name is included in the request, so you'd know the name

[17:31:00.0000] <Philip`>
You map the namespace hierarchy onto a tree of public/private keys

[17:31:01.0000] <Philip`>
so that only people with the key for a particular name can sign content purporting to be from that name

[17:32:00.0000] <Dashiva>
How would the keys be distributed and authenticated?

[17:35:00.0000] <Dashiva>
I suppose you could use a DNS-like system

[17:36:00.0000] <Philip`>
You trust somebody (perhaps a centralised authority like with DNS) who asserts the validity of other keys, recursively

[17:36:01.0000] <Philip`>
and keys are just content so they're distributed like any other content (under some kind of naming convention)

[17:38:00.0000] <Dashiva>
"I tell the server to store _my_ pictures." "I tell the camera the pictures it takes are _mine_"

[17:39:00.0000] <Dashiva>
Seems like you'd need to tell the server about the camera (or the inverse), or you'd be pinging every machine everywhere for pictures

[17:52:00.0000] <Dashiva>
"Most communication isn't point to point anymore"

[17:55:00.0000] <Philip`>
I think one of my doubts was that I expect things are going to become more point-to-point, not less

[17:56:00.0000] <Philip`>
because everyone is a fan of personalisation and customisation

[17:56:01.0000] <Philip`>
When you visit a web site you're never going to get precisely the same content as anybody else, because you'll have logged in and it'll be tailored for you

[17:56:02.0000] <Philip`>
and YouTube will do on-the-fly encoding of targeted adverts into the videos you watch

[17:57:00.0000] <Dashiva>
This video presentation seems to focus on personal data, all the examples are gadget-related

[17:57:01.0000] <Philip`>
and the idea of caching content for multiple users will be obsolete

[17:57:02.0000] <Philip`>
and the idea of requesting the same piece of data twice will also be obsolete

[17:58:00.0000] <Philip`>
and people will care about services, not about content

[17:59:00.0000] <Dashiva>
Although perhaps non-personal content (audio, video media) might have big enough in volume to still matter

[18:04:00.0000] <Philip`>
/me sleeps

[18:05:00.0000] <Dashiva>
They sign the content, yes

[18:06:00.0000] <Dashiva>
They'll need to include some kind of DRM, though...

[18:17:00.0000] <Dashiva>
One huge problem seems to be that people are expected to manage (sub)namespaces

[18:17:01.0000] <Dashiva>
I don't think most people will be able to make sensible hierarchies

[18:26:00.0000] <jwm>
lots of chatting tonight

[18:26:01.0000] <jwm>
hehe

[18:42:00.0000] <Dashiva>
"Security: each packet will be named; the naming will be registered and secure (in the same way that IP addresses are now). "

[18:43:00.0000] <Dashiva>
If IP addresses are registered and secure, and each packet contains an IP address...

[22:47:00.0000] <redder>
hi guys & girls

[22:49:00.0000] <MikeSmith>
redder: hey

[22:49:01.0000] <redder>
how are you today ?

[22:50:00.0000] <MikeSmith>
so, regarding http://www.w3.org/Bugs/Public/show_bug.cgi?id=7546#c10, "HTML 5 being superseded by CCN or a similar structural change", anybody know what "CCN" is?

[22:51:00.0000] <MikeSmith>
redder: I am doing swimingly

[22:51:01.0000] <redder>
lol

[22:51:02.0000] <redder>
cool

[22:52:00.0000] <redder>
i am looking for someone who can design a site for me :)

[22:52:01.0000] <MikeSmith>
content-centric networking

[22:53:00.0000] <MikeSmith>
redder: I think this would probably not be the best place to look

[22:53:01.0000] <MikeSmith>
there are not so many designers here

[22:53:02.0000] <redder>
any suggestions ?

[22:54:00.0000] <redder>
i do not know web building

[22:54:01.0000] <MikeSmith>
redder: how much would you be able to pay somebody to design the site?

[23:14:00.0000] <MikeSmith>
it seems that CCNx is the more common abbreviation for content-centric networking these days

[23:15:00.0000] <redder>
MikeSmith

[23:16:00.0000] <MikeSmith>
yeah

[23:16:01.0000] <redder>
depends on what they ask

[23:16:02.0000] <redder>
lol

[23:16:03.0000] <redder>
i am fair guy

[23:16:04.0000] <redder>
and what they can offer

[23:16:05.0000] <redder>
just NO Flash

[23:16:06.0000] <redder>
lol

[23:25:00.0000] <othermaciej>
what is CCN?

[23:25:01.0000] <othermaciej>
as in, what does "content centric networking" actually mean

[23:27:00.0000] <MikeSmith>
othermaciej: yeah, though not clear how it'd ever "supersede" HTML5

[23:27:01.0000] <othermaciej>
whether it would or not, I am now curious what it means

[23:27:02.0000] <othermaciej>
I seriously have no idea

[23:27:03.0000] <othermaciej>
it is not a term I have heard before

[23:28:00.0000] <zcorpan>
/me guesses it's the supposedly next term after DHTML, AJAX, HTML5

[23:30:00.0000] <othermaciej>
from googling, it sounds considerably lower level than HTML5

[23:30:01.0000] <othermaciej>
sounds like it is more a replacement for TCP/IP

[23:31:00.0000] <MikeSmith>
yeah

[23:31:01.0000] <othermaciej>
(though I am still not sure what it is exactly)

[23:33:00.0000] <MikeSmith>
Hixie: I would like to bring http://www.w3.org/Bugs/Public/show_bug.cgi?id=9178 to your attention

[23:33:01.0000] <MikeSmith>
request for Noah/TAG : add definition of "conforming document"

[23:34:00.0000] <MikeSmith>
othermaciej: http://www.w3.org/Bugs/Public/show_bug.cgi?id=7034 is one I raised but I'm not quite sure where we're at with it

[23:35:00.0000] <MikeSmith>
it really seems like a meta-issue, not a request for a specific change to the spec

[23:35:01.0000] <othermaciej>
MikeSmith: where I'd like to be at with it is one or more bugs that propose specific changes to conformance requirements

[23:35:02.0000] <othermaciej>
MikeSmith: I don't like having this open-ended bug where everyone commenting has a different idea of what it means

[23:36:00.0000] <MikeSmith>
yeah

[23:36:01.0000] <othermaciej>
"remove all document conformance requirememts

[23:36:02.0000] <othermaciej>
" would be a very clear bug

[23:36:03.0000] <othermaciej>
but if this bug ever represented that request, then Sam has apparently repurposed it for something different

[23:37:00.0000] <othermaciej>
Larry also made yet another different request

[23:38:00.0000] <othermaciej>
and there's all the old history

[23:38:01.0000] <MikeSmith>
what that bug started out as was 'change "conformance checker" to "ideology checker" or "loyalty checker"'

[23:38:02.0000] <othermaciej>
yeah, which was kind of a joke

[23:38:03.0000] <othermaciej>
then at least some people supported removing all document conformance requirements as a serious idea

[23:39:00.0000] <MikeSmith>
If I were to close it myself at this point, I guess somebody else would likely just re-open it

[23:39:01.0000] <othermaciej>
now Sam seems to be proposing some set of adding requirements, removing requirements, and making requirements optional

[23:40:00.0000] <othermaciej>
if you feel Sam's request differs from yours, you could work with him to put it in one or more other bugs, or you could go ahead and let him repurpose the bug

[23:40:01.0000] <othermaciej>
the biggest thing I'd worry about is if multiple people try to repurpose the bug, all for different purposes

[23:46:00.0000] <MikeSmith>
I would prefer to remove myself as the reporter of that bug, but I guess there's no way to do actually do that

[23:47:00.0000] <othermaciej>
not without hacking the database

[23:48:00.0000] <othermaciej>
btw it looks like moving all those bugs to VERIFIED has been successful at getting issues to move to either CLOSED or REOPENED instead of languishing at RESOLVED

[23:48:01.0000] <othermaciej>
which makes me happy

[23:49:00.0000] <othermaciej>
we're up to 325 CLOSED, where yesterday I think we were around 270

[23:50:00.0000] <MikeSmith>
I just closed all of the ones I had reported 

[23:51:00.0000] <MikeSmith>
except for two

[23:51:01.0000] <othermaciej>
I did all of mine except one

[23:52:00.0000] <othermaciej>
we still have a backlog of 1278 bugs in RESOLVED or VERIFIED

[23:53:00.0000] <othermaciej>
of which 34 are marked with TrackerIssue or TrackerRequest

[23:53:01.0000] <othermaciej>
so there's clearly a big backlog of bugs that haven't gotten a reply from the originator

[23:53:02.0000] <othermaciej>
thanks for doing your part to reduce that number

[00:01:00.0000] <MikeSmith>
I actually didn't do all mine.. the two I have not closed yet are 7034 (the one we were just discussing) and 9178 (add definition of "conforming document")

[00:08:00.0000] <JonathanNeal>
http://img710.imageshack.us/img710/2292/smurfperception.png

[00:12:00.0000] <JonathanNeal>
opps, meant that for the ot

[00:18:00.0000] <othermaciej>
9178 is still NEW, so that's not an issue

[02:00:00.0000] <hsivonen>
is window.location.href supposed to be the empty string while the window is loading?

[03:54:00.0000] <MikeSmith>
othermaciej: the TAG request for a schema makes we wonder if we should add a new HTML WG bugzilla component for "proposed deliverables" or some such

[03:55:00.0000] <othermaciej>
MikeSmith: so far our way of adding deliverables has been that someone writes it, and then proposes it as a publication

[03:55:01.0000] <MikeSmith>
true

[03:55:02.0000] <othermaciej>
MikeSmith: I'm not sure what we'd do with a proposal for a new deliverable that doesn't come with a commitment to produce it

[03:56:00.0000] <MikeSmith>
well, in this case, there is already a schema that's been produced

[03:58:00.0000] <othermaciej>
a commitment to edit it as a WG deliverable then

[03:59:00.0000] <MikeSmith>
yeah, that makes sense, then

[03:59:01.0000] <MikeSmith>
that way of describing it, I mean

[04:12:00.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-html-ig-jp/2010Mar/0037.html is a good example of a case where <rb>-less HTML5 Ruby markup makes for easier authoring and code readability than the equivalent way of marking it up in XHTML 1.1 complex-ruby 

[04:40:00.0000] <zcorpan>
MikeSmith: the </rt> tags can be omitted too

[04:40:01.0000] <MikeSmith>
http://news.cnet.com/8301-30685_3-20000423-264.html

[04:40:02.0000] <MikeSmith>
"Web guru Tim Bray takes Google Android job"

[04:42:00.0000] <MikeSmith>
zcorpan: in HTML5 ruby?

[04:42:01.0000] <zcorpan>
MikeSmith: yes

[04:43:00.0000] <MikeSmith>
I don't understand.. what would be the semantics of that?

[04:44:00.0000] <zcorpan>
the same; the rt element is closed at </ruby> in that example

[04:52:00.0000] <zcorpan>
hsivonen: no (re location.href)

[04:52:01.0000] <othermaciej>
I was surprised to learn that drawImage() with a "copy" composite operation is no faster than with the default "src-over" operation

[04:52:02.0000] <othermaciej>
(at least in Safari)

[04:54:00.0000] <othermaciej>
(drawImage is drawing a canvas that happens to contain all 100% opaque pixels in this case, but I would not expect the implementation to specifically check for that case)

[04:57:00.0000] <zcorpan>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Ciframe%3E%3C%2Fiframe%3E%3Cscript%3E%0Aw(window%5B0%5D.location.href)%0A%3C%2Fscript%3E vs http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Ciframe%3E%3C%2Fiframe%3E%3Cscript%3E%0Aw(window%5B0%5D.location.href)%3B%0A%3C%2Fscript%3E

[04:57:01.0000] <zcorpan>
...is interesting in chrome

[04:57:02.0000] <MikeSmith>
zcorpan: about the ruby thing, I misread what you wrote. had thought you were saying the rt element could be omitted

[04:58:00.0000] <zcorpan>
also if i replace the ; with a space

[04:58:01.0000] <zcorpan>
weird bug

[04:59:00.0000] <Philip`>
othermaciej: Drawing something that might contain alpha?

[04:59:01.0000] <Philip`>
(I know that at least Firefox has separate compositing paths for images that are known to be solid RGB)

[05:00:00.0000] <Philip`>
((or Cairo does, anyway))

[05:00:01.0000] <othermaciej>
I was testing in Safari, and I believe we don't ever assume a canvas backing store is all RGB

[05:00:02.0000] <Philip`>
Ah, so doing drawImage(canvas) rather than drawImage(image)?

[05:03:00.0000] <Philip`>
About ImageData: If the fastest possible unpremultiplying-image-copy code is still slow enough to be a problem, could browsers do higher-level optimisations like having getImageData do a direct copy of the image data and only perform the unpremultiplication if the user accesses .data?

[05:03:01.0000] <Philip`>
(and then putImageData could do a direct copy with no premultiplying if it knows the user didn't fiddle with .data at all)

[05:04:00.0000] <Philip`>
(so it can be used for blitting between regions on a canvas or between multiple canvases, doing nothing more computationally complex than memcpy)

[05:04:01.0000] <othermaciej>
I think that would slow down doing poking at the contents of ImageData

[05:04:02.0000] <othermaciej>
by at least the branch to do the check for lazy un-premult

[05:05:00.0000] <othermaciej>
even then, it would be overall slower for cases where you actually do modify the ImageData

[05:05:01.0000] <hsivonen>
zcorpan: I guess window.location in Gecko is broken, then

[05:05:02.0000] <Philip`>
I presume anyone who cares about performance would do "var d = imagedata.data; d[i]..." rather than doing the .data lookup every time

[05:05:03.0000] <Philip`>
even if it's doing nothing more complex than a normal property lookup

[05:05:04.0000] <zcorpan>
hsivonen: i get 'about:blank' in gecko

[05:06:00.0000] <othermaciej>
Philip`: so you're assuming the lazy un-pre-multiply would be on accessing the data property?

[05:07:00.0000] <zcorpan>
hsivonen: or the document's url if it's known when the script runs

[05:07:01.0000] <othermaciej>
Philip`: it would still mean you have to do the original copy and un-premultiply as a second pass

[05:07:02.0000] <Philip`>
othermaciej: Yes, so it's fast if you don't look at the data at all, but if you do then it'll have a one-time conversion cost and then pixel access will be as fast as usual

[05:07:03.0000] <hsivonen>
zcorpan: oh. I get "" for window.location.href and "about:blank" for window.document.documentURI

[05:08:00.0000] <othermaciej>
Philip`: that one-time conversion cost is going to be a lot for a large image - comparable to the cost of getImageData now, but you have the original cost of the raw copy on top of that

[05:08:01.0000] <othermaciej>
Philip`: as it is, if you just want to pass image data around and *not* directly poke at pixels, you are better off passing a HTMLCanvasElement than an Image Data

[05:08:02.0000] <Philip`>
Is the raw copy significant in comparison to that?

[05:09:00.0000] <Philip`>
Looks like Firefox actually already does a raw copy from the canvas into the ImageData, and then unpremultiplies as a separate stage within getImageData, rather than doing copy+unpremultiply simultaneously

[05:09:01.0000] <hsivonen>
boohoo. Live DOM saving is broken

[05:10:00.0000] <hsivonen>
zcorpan: http://bit.ly/9GBt2V

[05:10:01.0000] <Philip`>
Looks like Firefox also uses a 64KB lookup table for unpremultiplication, which sounds a bit weird to me

[05:10:02.0000] <othermaciej>
Philip`: that might partly explain why its getImageData is so slow

[05:11:00.0000] <Philip`>
(https://bugzilla.mozilla.org/show_bug.cgi?id=519400)

[05:11:01.0000] <Philip`>
/me goes away

[05:12:00.0000] <zcorpan>
hsivonen: seems it's different for data: urls

[05:13:00.0000] <zcorpan>
hsivonen: about:blank and data: gives empty string in gecko, it seems

[05:13:01.0000] <zcorpan>
hsivonen: but src=document gives 'about:blank'

[05:19:00.0000] <hsivonen>
zcorpan: wow. that's weird.

[06:32:00.0000] <Philip`>
"Roughly speaking 'sce' should be to 'bdi' what <em> is to <i>." - I imagine people have widely varying views on whether that is a positive or a negative

[06:33:00.0000] <Dashiva>
We need _more_ stuff messing up text directions?

[06:35:00.0000] <Dashiva>
(And is there a practical solution to the problem of users inserting magic characters that mess up text direction?)

[06:39:00.0000] <Philip`>
Do you mean magic characters like in http://www.google.com/search?q=%E2%80%AE ?

[06:39:01.0000] <erlehmann>
Dashiva, what is wrong with a) filtering them out ? or b) enforce the opposite markers after user input ?

[06:40:00.0000] <erlehmann>
(i can see the problem with option a)

[06:40:01.0000] <MikeSmith>
sce as a name is even more opaque than bdi

[06:40:02.0000] <MikeSmith>
what does sce stand for?

[06:40:03.0000] <jgraham>
enforce * after user input seems to fail for any *

[06:40:04.0000] <Dashiva>
The problem with b) is that it isn't direction-independent

[06:42:00.0000] <erlehmann>
indeed :<

[06:43:00.0000] <Dashiva>
I suppose it's not impossible to build an index of characters that change text direction, but even then you'd be forbidding characters that would be legal in one

[06:43:01.0000] <Dashiva>
of the directions (or at least harmless)

[06:45:00.0000] <Dashiva>
http://lists.w3.org/Archives/Public/public-html/2010Mar/0335.html

[06:45:01.0000] <zcorpan>
Dashiva: there's a POP BIDI OVERRIDE character or similar

[06:45:02.0000] <Dashiva>
Good thing it's text and not XHTML

[06:45:03.0000] <Dashiva>
zcorpan: Is that harmless if used in excess?

[06:46:00.0000] <zcorpan>
i think so

[06:47:00.0000] <Dashiva>
Although having an equal number of that character to the number of characters in user input seems bad

[06:47:01.0000] <zcorpan>
oh i guess it might mess things up if used in the middle of words

[06:49:00.0000] <Dashiva>
As long as you can contain the effects to the user input, the user can only mess up his own content, so that's okay

[06:50:00.0000] <Dashiva>
Maybe a single POP DIRECTIONAL FORMATTING is enough for any number of modifiers

[06:52:00.0000] <Philip`>
If the user can mess up his own content, he could use RTL markers in order to write naughty words that bypass profanity filters because they're written backwards

[06:54:00.0000] <Dashiva>
I don't really worry too much about censorship, but I suppose that's also an issue that needs to be adressed

[06:58:00.0000] <erlehmann>
Philip`, i routinely bypass wordfilters by using special unicode chars. like 𝚝𝚑𝚒𝚜

[06:58:01.0000] <erlehmann>
seems that nobody to filter unicode math characters that look like letters

[06:59:00.0000] <erlehmann>
nobody bothers. my sentence a verb

[06:59:01.0000] <karlushi>
http://www.tbray.org/ongoing/When/201x/2010/03/15/Joining-Google

[07:00:00.0000] <karlushi>
"I’m going to have to get savvier about HTML5-based applications, because a lot of smart people think the future’s there, that the “native app” notion will soon seem quaint." -- Tim Bray

[07:01:00.0000] <MikeSmith>
I hope he's not holding his breath on that one

[07:01:01.0000] <Dashiva>
When was the last time native apps were on death row?

[07:01:02.0000] <erlehmann>
but it will be *real soon* (holds breath)

[07:01:03.0000] <erlehmann>
*cough*

[07:01:04.0000] <MikeSmith>
every year for the last 5 years

[07:01:05.0000] <MikeSmith>
soon, very soon

[07:02:00.0000] <MikeSmith>
the Messiah will return and deliver us all from the evils of Native Apps

[07:02:01.0000] <Dashiva>
We could have a pool for what happens first: year of linux on the desktop, or death of native apps

[07:02:02.0000] <MikeSmith>
/me snorts green tea out of his nose

[07:03:00.0000] <MikeSmith>
Dashiva: the year of Linux on the desktop has already arrived

[07:03:01.0000] <MikeSmith>
it's just that you didn't notice

[07:03:02.0000] <jgraham>
Well since he is working on non-native apps on a popular linux based consumer device...

[07:03:03.0000] <MikeSmith>
you need special glasses to see it

[07:03:04.0000] <erlehmann>
because its a slow process ?

[07:03:05.0000] <Dashiva>
I think it's possible to predict an increase in non-native apps without the native app dying out

[07:04:00.0000] <karlushi>
there was a Web site for crystall ball mantra… trying to dig the url

[07:04:01.0000] <MikeSmith>
I think we are still in the Year of Web Services

[07:04:02.0000] <MikeSmith>
it's just a very long year

[07:06:00.0000] <Dashiva>
Well, it's still september, after all

[07:06:01.0000] <erlehmann>
maybe the WHATWG should get some lobbyists akin to the RIAA so that congress votes to extend the year by 20 years

[07:06:02.0000] <Dashiva>
What recovery rules would you apply to this? http://lists.w3.org/Archives/Public/public-html/2010Mar/0335.html

[07:07:00.0000] <erlehmann>
the Mike Smith web services term extension act ;)

[07:07:01.0000] <Dashiva>
I see two instances of <Rich> without a matching </Rich>

[07:08:00.0000] <Philip`>
I use the recovery rule that if it's too hard to work out what's a quote and what's new material then I don't bother reading it

[07:08:01.0000] <MikeSmith>
it's like the "we are currently in a stage of high alert" messages that play continuously in the airport for getting close to 10 years now or whatever

[07:08:02.0000] <Dashiva>
If the second one is an empty comment, saying "imply </Rich> on encountering an empty line" seems reasonable

[07:09:00.0000] <karlushi>
aaaah found

[07:09:01.0000] <karlushi>
http://www.longbets.org/

[07:09:02.0000] <erlehmann>
Dashiva, isnt there a parsing specification for that to get a proper DOM out of it ? ;)

[07:09:03.0000] <gsnedders>
/me does backflip

[07:10:00.0000] <Dashiva>
erlehmann: I'm not aware of any specs that include the <Rich> element

[07:10:01.0000] <karlushi>
/me is surprised there is no bet for 2022

[07:11:00.0000] <Philip`>
MikeSmith: Would you prefer them to stop being highly alert, and to allow more unimaginative terrorists onto their planes with bombs?

[07:12:00.0000] <MikeSmith>
“By 2060 the total population of humans on earth will be less than it is today.” .. "total population" instead of just "population".. that's clearly because it means "net population", and this guy knows about the arrival of the anti-humans who cancel out the population of us normal humans

[07:12:01.0000] <Dashiva>
Now we have to kill you, MikeSmith

[07:12:02.0000] <erlehmann>
then … let me see … data state, tag open state, tag name state, data state … wtf am i reading

[07:12:03.0000] <Dashiva>
I mean, someone else has to kill you

[07:12:04.0000] <Dashiva>
... I said nothing

[07:13:00.0000] <MikeSmith>
Dashiva: you'd have to catch me off-guard first, which will be hard for you guys to do, because I have my special glasses that I can detect what your true appearance

[07:13:01.0000] <Philip`>
Does "by 2060" mean "on at least point in time between today and 2060"?

[07:14:00.0000] <karlushi>
there is a missing view of the bets by dates. This month's bets results :) 

[07:14:01.0000] <MikeSmith>
Philip`: no, I would prefer that they be honest and just say, We will forever from now on be in a state of high alert because that works well for us

[07:14:02.0000] <Philip`>
If so, what about tomorrow? Maybe the random fluctuations in population are stronger than the gradual upward trends, so it'll go down a bit before it goes up again

[07:14:03.0000] <Dashiva>
MikeSmith: But high alert doesn't work well, by definition. It means more effort than ideal.

[07:15:00.0000] <Philip`>
By what definition?

[07:15:01.0000] <Philip`>
If a cake is highly tasty, is that bad by definition because it means more tasty than ideal?

[07:16:00.0000] <Dashiva>
It's bad for the cake baker

[07:16:01.0000] <Dashiva>
Since it means more effort to produce the cake

[07:16:02.0000] <Philip`>
Maybe it means 10% more effort to make it twice as tasty

[07:16:03.0000] <Philip`>
which would seem to be worthwhile

[07:17:00.0000] <erlehmann>
The WHATWG is committed to the well being of all participants. Cake and grief counseling will be available at the conclusion of the specification. Thank you for helping us help you help us all.

[07:17:01.0000] <Dashiva>
Cake... or DEATH?

[07:18:00.0000] <Dashiva>
http://www.longbets.org/382

[07:18:01.0000] <erlehmann>
Curse your sudden and inevitable discovery of the true reason behind all this !

[07:18:02.0000] <Dashiva>
Seems like a rather useless bet to make

[07:19:00.0000] <erlehmann>
„will go to National Rifle Association if Keane wins“

[07:19:01.0000] <erlehmann>
so they can shoot the remains ?

[07:19:02.0000] <erlehmann>
when will the bet be called ?

[07:20:00.0000] <erlehmann>
ah 2018

[07:21:00.0000] <Philip`>
"For the purpose of the bet, Earth should be considered "destroyed" if, at the end of the term of this bet in 2018, zero human beings reside on the surface of the planet."

[07:21:01.0000] <Dashiva>
http://www.longbets.org/244

[07:21:02.0000] <Philip`>
What if we all move into underwater cities?

[07:21:03.0000] <Dashiva>
Who would the VP be?

[07:22:00.0000] <MikeSmith>
the Submariner and Aquaman would have to fight for it

[07:23:00.0000] <Philip`>
/me would be sure to sign up as a Big Daddy

[07:27:00.0000] <erlehmann>
i wanna be mermaid man :>

[07:27:01.0000] <aednsi>
Do it up

[09:47:00.0000] <MikeSmithX>
reading http://ajaxian.com/archives/a-better-mobile-web-what-else

[09:47:01.0000] <MikeSmithX>
proposing -webkit-viewport-fixed

[09:49:00.0000] <annevk>
would be nice if fixed positioning just worked...

[09:50:00.0000] <annevk>
position:fixed-i-really-mean-it-this-time ...

[09:51:00.0000] <MikeSmithX>
position: fixed important!!!

[09:51:01.0000] <MikeSmithX>
position: fixed dammit!

[09:52:00.0000] <MikeSmithX>
position: fixed goddammit!

[10:03:00.0000] <annevk>
we could define a new kind of priority

[10:03:01.0000] <annevk>
position:fixed !omgiamsonotkiddingthistimearound

[10:05:00.0000] <Dashiva>
It isn't a priority issue, though, is it?

[10:05:01.0000] <Dashiva>
It seems like they put a viewport onto the viewport

[10:08:00.0000] <jwalden>
yo dawg

[10:48:00.0000] <caio_>
I have a small question. I will can send the <video> contents (from webcam <device>) to a server, through HTML 5 / JavaScript? 

[10:56:00.0000] <AryehGregor>
<device> isn't even fully specced last I checked, let alone implemented.

[11:02:00.0000] <caio_>
Yes, I Know. There is a way to send the contents of <video> to a server, maybe through Web Sockets?

[11:03:00.0000] <caio_>
for Web Conferences.

[11:07:00.0000] <AryehGregor>
If you have some data in JavaScript, you can send it via Web Sockets or AJAX or whatever.

[11:07:01.0000] <AryehGregor>
But there's no standards-based interface right now to get webcam data, AFAIK.

[11:32:00.0000] <Philip`>
zcorpan: Ping

[11:40:00.0000] <AryehGregor>
/me wonders why http://tstarling.com/blog/2009/12/response-about-climate-change-action/ is still cached by Google as having his name in the title despite having changed ages ago

[11:40:01.0000] <AryehGregor>
Second result for my name.  Although at least the first is my actual website now.

[11:44:00.0000] <Philip`>
Maybe because it indexes pages based partly on the text used in links to them, not just the text in the page itself

[11:44:01.0000] <Philip`>
Oh, but that wouldn't explain the cache

[11:45:00.0000] <AryehGregor>
Also, I doubt there are any links to that post from other sites at all, let alone ones containing my name.

[11:45:01.0000] <Philip`>
except, I don't see your name in a cached version of it

[11:45:02.0000] <Philip`>
so I don't understand the problem at all

[11:45:03.0000] <AryehGregor>
Search for "Aryeh Gregor".  Does that page show up in the top ten results?

[11:46:00.0000] <Dashiva>
I don't see your name in the title, just in the snippet from the text

[11:47:00.0000] <AryehGregor>
Fascinating.

[11:47:01.0000] <Philip`>
It's the 5th page in the list (3rd site in the list)

[11:47:02.0000] <AryehGregor>
Oh, I get different search results.

[11:47:03.0000] <AryehGregor>
Due to customization.

[11:47:04.0000] <AryehGregor>
The customized version has my name in the title, the uncustomized one doesn't.

[11:47:05.0000] <AryehGregor>
Puzzling.

[11:48:00.0000] <Philip`>
The first result I get is "Re: Inconsistency in HTML 4.01 regarding NBSP from Aryeh Gregor on ..."

[11:48:01.0000] <AryehGregor>
Oh, well, I guess I'll just have to wait for that post to fade into obscurity.  This is why I use "Simetrical" whenever I'm discussing politics or anything, but Tim Starling didn't know that.

[11:48:02.0000] <Dashiva>
I didn't know you were Simetrical...

[11:48:03.0000] <AryehGregor>
Did you see Simetrical elsewhere and not realize I was the same person, or you never heard of Simetrical?

[11:48:04.0000] <Dashiva>
I suppose that's the point :)

[11:49:00.0000] <AryehGregor>
The point of using Simetrical these days is mainly to avoid gaming sites and stuff coming up as top hits for my name.

[11:49:01.0000] <Dashiva>
I think I remember you from #wikia and related places

[11:50:00.0000] <AryehGregor>
Oh, didn't know you were involved in wiki stuff.

[11:50:01.0000] <Philip`>
I avoid that kind of problem by having the same name as lots of other people

[11:51:00.0000] <Philip`>
so nobody can easily tell if a particular person is really me

[11:51:01.0000] <AryehGregor>
I'm probably the only person in the world named Aryeh Gregor.

[11:51:02.0000] <Dashiva>
Not much. I co-ran two gaming wikis, and reported the template variables are unavailable to parser functions bug :)

[11:51:03.0000] <AryehGregor>
Since my first name is extremely Jewish and my last name is extremely non-Jewish.

[11:51:04.0000] <AryehGregor>
Dashiva, oh, that one.

[11:51:05.0000] <AryehGregor>
That's an old one.

[11:52:00.0000] <AryehGregor>
I seem to recall {{#tag:}} is the official workaround, and nobody ever bothered actually fixing it.

[11:52:01.0000] <Dashiva>
It was properly fixed last year, I believe

[11:52:02.0000] <AryehGregor>
I avoid templates.  :P

[11:53:00.0000] <Dashiva>
I love templates

[11:53:01.0000] <Philip`>
/me wonders if MediaWiki supports template metaprogramming

[11:53:02.0000] <AryehGregor>
Sadly, yes.

[11:53:03.0000] <Philip`>
Is it Turing complete?

[11:53:04.0000] <Dashiva>
Not quite

[11:53:05.0000] <Philip`>
Alas

[11:53:06.0000] <AryehGregor>
I've heard it rumored that it is, but the resource limits are such that it isn't in practice.

[11:53:07.0000] <Dashiva>
But we did manage to totally overload the job queue on tabularasa

[11:53:08.0000] <AryehGregor>
Pfft, you don't even have to try to do that.

[11:54:00.0000] <Dashiva>
Well, we did get some central wikia person posting asking us to do something about it :)

[11:54:01.0000] <Philip`>
If you care about resource limits, then a desktop PC isn't Turing complete either

[11:54:02.0000] <AryehGregor>
Wikimedia sysadmins tend to just truncate the job queue if it gets too long.  :)

[11:54:03.0000] <Philip`>
so it's probably best to ignore those limits

[11:54:04.0000] <AryehGregor>
Philip`, yeah, but there are theoretical limits and practical limits.

[11:55:00.0000] <Dashiva>
We also ran into the maximum include size

[11:55:01.0000] <AryehGregor>
Meant to prevent dastardly template metaprogramming from getting out of hand.

[11:55:02.0000] <Dashiva>
:)

[11:56:00.0000] <AryehGregor>
Although it doesn't work very well, you can still make pages that take half a minute or more to render on Wikipedia.


2010-03-16
[18:31:00.0000] <MikeSmithX>
http://www.w3.org/2010/02/rdfa/wiki/RDFa-DOM-API

[18:35:00.0000] <MikeSmith>
boblet: http://html5.org/tools/web-apps-tracker?from=4846&to=4847

[18:37:00.0000] <boblet>
MikeSmith: interesting

[18:49:00.0000] <boblet>
MikeSmith: was there a particular reason for pinging me on that btw?

[18:49:01.0000] <MikeSmith>
um, I thought you had asked me about something related to that

[18:50:00.0000] <MikeSmith>
but I guess it was somebody else

[18:51:00.0000] <boblet>
not that I recall… :)

[22:27:00.0000] <MikeSmith>
Hixie: I don't see what http://www.w3.org/html/wg/wiki/ChangeProposals/fixedprefixsimple would give us that we can't already do with the class attribute

[00:03:00.0000] <wycats>
e

[00:20:00.0000] <Hixie>
MikeSmith: i don't think we'd have wanted apple to invent a new class name for <canvas>

[00:20:01.0000] <Hixie>
class names are for authors

[00:24:00.0000] <MikeSmith>
Hixie: I guess I can't see that Rob's fixedprefixsimple proposal would be much value in practice for anything more author-private use

[00:25:00.0000] <MikeSmith>
Hixie: is Apple has used this approach for canvas, what would it have looked like? <img -apple-canvas> ?

[04:13:00.0000] <annevk>
no_mind, we've yet to figure out styling form controls unfortunately

[04:14:00.0000] <annevk>
though what WebKit is doing for controls in combination with appearance and all seems quite nice, but nobody has written down all the details in a specification yet...

[04:14:01.0000] <no_mind>
hmm so that spec is yet to be written...

[04:15:00.0000] <annevk>
pretty much

[04:15:01.0000] <no_mind>
how one start writing such spec

[04:15:02.0000] <no_mind>
where ever I give html5 demo, all designers ask this question

[04:15:03.0000] <annevk>
usually by figuring out the fundamentals

[04:16:00.0000] <annevk>
e.g. how does form styling currently work across various platforms, what is the effect of setting various properties, etc.

[04:16:01.0000] <annevk>
then figuring out how appearance affects that, what pseudo-classes are supported, how they all work together

[04:17:00.0000] <annevk>
and while testing all that writing down implementation criteria for someone that tries to emulate it

[04:17:01.0000] <annevk>
once we know the fundamentals we can start poking holes and enable new stuff

[04:17:02.0000] <annevk>
that's pretty much how we're doing most new features that are somewhat based on old APIs

[04:18:00.0000] <no_mind>
ok

[04:18:01.0000] <no_mind>
that will mean we have to add new css controls or attributes ?

[04:19:00.0000] <annevk>
maybe

[04:19:01.0000] <annevk>
another idea is that appearance:none would disable all styling and you would create the entire control yourself using XBL

[04:20:00.0000] <annevk>
though we'd need some hooks from XBL into the form API to make that completely feasible

[04:20:01.0000] <annevk>
e.g. so you can declare a form control matches :out-of-range and things like that

[04:21:00.0000] <annevk>
anyway, as things stand today CSS is not defined for form controls

[04:21:01.0000] <annevk>
the CSS WG declared it out of scope and nobody volunteered to fix it

[04:24:00.0000] <MikeSmith>
XBL2 will save us

[04:31:00.0000] <Dashiva>
But who will save us from XBL2?

[04:32:00.0000] <no_mind>
annevk, while leaving other form controls is fine, but styling of calender and time control is very important

[04:45:00.0000] <asmodai>
wtb: printing of MathML formulae in Mozilla

[04:53:00.0000] <MikeSmith>
fwiw, I concur with no_mind about calendar-widget styling being important for designers

[04:53:01.0000] <MikeSmith>
any time I've ever shown a demo of the WF2 input calendar, one of the very first questions is always about whether the style can be changed

[04:54:00.0000] <no_mind>
MikeSmith, same here.

[04:54:01.0000] <no_mind>
I think its time to write a specification for this. 

[04:54:02.0000] <MikeSmith>
no_mind: so write one :)

[04:55:00.0000] <no_mind>
I am confused should this be a separate spec or should we add to an existing HTML5 spec

[04:55:01.0000] <MikeSmith>
separate spec

[04:56:00.0000] <MikeSmith>
I'm not myself convinced it's so important to end users

[04:56:01.0000] <MikeSmith>
I do realize that some designers believe it's importan

[04:57:00.0000] <MikeSmith>
the alternative for designers who do is just to keep doing whatever they're doing now to provide calendar widgets in their content

[05:01:00.0000] <no_mind>
not to end users but to designers, it is. Like font-face in css

[05:05:00.0000] <MikeSmith>
no_mind: well, I think users do actually benefit a lot more from the effects of font-face than they would of calendar-widget styling

[05:05:01.0000] <klevison>
Whati is best browser to test html5 app?

[05:05:02.0000] <klevison>
*What

[05:06:00.0000] <MikeSmith>
no_mind: so getting good font-face support is something that to me at least seems actually pretty important for users

[05:07:00.0000] <MikeSmith>
klevison: is this a test?

[05:07:01.0000] <MikeSmith>
we can only pick one?

[05:08:00.0000] <MikeSmith>
what do I win if I pick the right one?

[05:08:01.0000] <klevison>
I'm new in this topic

[05:08:02.0000] <klevison>
I'm developer.

[05:09:00.0000] <klevison>
But I wanna know which browser I should user to dev. html5 app

[05:09:01.0000] <MikeSmith>
I guess most people test in multiple browsers

[05:09:02.0000] <MikeSmith>
including in development/pre-release builds

[05:10:00.0000] <no_mind>
MikeSmith, how font-support was important while calendar styling is not ? Both are used by designers

[05:10:01.0000] <klevison>
In your opinion which browser is the best?

[05:10:02.0000] <MikeSmith>
klevison: lynx

[05:10:03.0000] <MikeSmith>
no_mind: because it's more important to me, and I consider myself the ideal end user

[05:11:00.0000] <klevison>
MikeSmith: windows browsers

[05:11:01.0000] <no_mind>
MikeSmith, okies, from my perspective the calender widget needs to blend with rest of look and feel :)

[05:12:00.0000] <MikeSmith>
klevison: I've heard that a lot of Windows users tend to use Internet Explorer for some reason

[05:13:00.0000] <MikeSmith>
if that rumor is true, I'd guess that might be a good target to test your app in

[05:13:01.0000] <MikeSmith>
no_mind: fair enough

[05:40:00.0000] <asmodai>
MikeSmith: Given the existence of http://jqueryui.com/themeroller/ I'd say styling is quite important to designers at least.

[05:43:00.0000] <MikeSmith>
asmodai: I wasn't claiming datepicker styling isn't important to designers

[05:44:00.0000] <asmodai>
MikeSmith: Ah ok, my bad, must have misunderstood.

[05:44:01.0000] <Philip`>
zcorpan: Hello

[05:45:00.0000] <zcorpan>
Philip`: hello

[05:45:01.0000] <Philip`>
klevison: Browsers all have different levels of support for various features, and different sets of bugs, so none is unequivocally best, and you should test in as many as possible and focus on the ones that your users are most likely to use

[05:46:00.0000] <Philip`>
zcorpan: I recently remembered that I forgot to do http://philip.html5.org/data/entities-in-attribute-with-no-semicolon-or-alphanumeric.txt

[05:47:00.0000] <Philip`>
I changed your regexp a bit to exclude > from attribute 'values', because otherwise it seemed to pick up a lot of entities that were in content after an element with attributes

[05:47:01.0000] <klevison>
Philip`: Is there a ranking of most use browsers?

[05:47:02.0000] <zcorpan>
Philip`: sweet, thanks

[05:48:00.0000] <Philip`>
zcorpan: Also I think the file is probably truncated, because it seemed to get stuck at some point (I guess a page hit a worst-complexity case in the pattern) so I killed it after it had stopped for a few minutes

[05:49:00.0000] <Philip`>
klevison: There are lots of rankings, all slightly different

[05:49:01.0000] <Philip`>
and it depends a lot on target audience, and country, etc

[05:49:02.0000] <zcorpan>
Philip`: ok

[05:49:03.0000] <Philip`>
but http://en.wikipedia.org/wiki/Usage_share_of_web_browsers gives some indications

[05:50:00.0000] <Philip`>
klevison: The theory is that HTML5 applications ought to work the same in all browsers, so for main development work you should pick whichever one gives you the best development tools (debuggers and network monitors and all that stuff)

[05:51:00.0000] <klevison>
phil

[05:51:01.0000] <klevison>
Philip`: ok

[05:51:02.0000] <klevison>
thanks

[06:03:00.0000] <Philip`>
/me wonder what exciting IE9 features will be revealed today

[06:03:01.0000] <annevk>
oh, IE9 is announced later today?

[06:03:02.0000] <Dashiva>
60% on acid3!

[06:03:03.0000] <annevk>
/me wonders if it will beat Opera Mini 5 in standards compliance

[06:04:00.0000] <asmodai>
Dashiva: rofl

[06:04:01.0000] <jcranmer>
SVG and canvas support?

[06:05:00.0000] <jgraham>
ES5

[06:05:01.0000] <asmodai>
Anyone else who does/did MathML? I wonder if an accented variable is done with an <msup> or just as an <mi><mo> pair without <msup>.

[06:05:02.0000] <jcranmer>
the fastest JS engine?

[06:06:00.0000] <asmodai>
integrated silverlight? :P

[06:06:01.0000] <annevk>
we have some (imo) crippled supported for MathML

[06:06:02.0000] <annevk>
support*

[06:06:03.0000] <asmodai>
annevk: Opera 5 mini?

[06:06:04.0000] <asmodai>
err, opera mini 5

[06:06:05.0000] <Philip`>
annevk: The existence of IE9 has already been announced, but it'll apparently be discussed in more detail at MIX later today

[06:07:00.0000] <annevk>
asmodai, we released it earlier today

[06:07:01.0000] <jgraham>
asmodai: <mover> I think

[06:07:02.0000] <annevk>
Philip`, always so correct

[06:07:03.0000] <asmodai>
annevk: Sorry, the mathml support I mean. You meant it's in opera mini 5?

[06:08:00.0000] <asmodai>
jgraham: Let me try that

[06:08:01.0000] <annevk>
asmodai, prolly, but also in the normal version of Opera

[06:08:02.0000] <asmodai>
jgraham: No, not over.

[06:08:03.0000] <asmodai>
jgraham: That's placing it directly over the variable.

[06:08:04.0000] <asmodai>
jgraham: what you do with, say, vectors

[06:08:05.0000] <Philip`>
According to Twitter, "Word has it #IE9 will support #CSS3 and #HTML5"

[06:09:00.0000] <jgraham>
Philip`: What about Web 3.0?

[06:10:00.0000] <Philip`>
jgraham: I'd guess they'll only have about 80% support for Web 3.0

[06:10:01.0000] <jgraham>
asmodai: Oh you mean like <mi>r`</mi> or so?

[06:10:02.0000] <Dashiva>
Philip`: As in "someone made this statement up" or, "reliable sources tell us"

[06:10:03.0000] <asmodai>
jgraham: if you have Mozilla 3.6 with html5 enabled: http://www.in-nomine.org/~asmodai/3d.html

[06:10:04.0000] <asmodai>
jgraham: See Kijiya's and rendering equation formulae

[06:10:05.0000] <Rik`>
Philip`: do we know when they will talk about it ? today, tomorrow ?

[06:11:00.0000] <asmodai>
First is using <msup> and the second is without.

[06:11:01.0000] <Philip`>
Dashiva: Who knows - the origin is lost in an infinite chain of retweets, and Twitter doesn't encourage accurate attribution and evidence for all statements

[06:12:00.0000] <jgraham>
asmodai: I think it should be without

[06:12:01.0000] <asmodai>
jgraham: looks better huh?

[06:12:02.0000] <Philip`>
Rik`: I think it's the keynote today (and some later sessions), which http://live.visitmix.com/Schedule says is 9am in Las Vegas

[06:13:00.0000] <asmodai>
oh shoot, need to fix something in that bidirectional

[06:13:01.0000] <jgraham>
Arguably it should be a combining mark of some sort so it is all one effective character or something

[06:13:02.0000] <asmodai>
jgraham: I wonder who's authoratative on these kind of issues.

[06:13:03.0000] <jgraham>
asmodai: Have you looked at the spec?

[06:13:04.0000] <asmodai>
jgraham: From time to time. But it is not a good way to stay sane :)

[06:14:00.0000] <asmodai>
/me goes to check again

[06:14:01.0000] <zcorpan>
/me sees mathml entities in http://www.in-nomine.org/~asmodai/3d.html

[06:15:00.0000] <zcorpan>
was it hand-written or generated?

[06:15:01.0000] <zcorpan>
s/generated/converted from something else/

[06:15:02.0000] <jgraham>
asmodai: It looks nice btw

[06:16:00.0000] <asmodai>
jgraham: Thanks! I am not a designer though, but try to make it clear to read.

[06:16:01.0000] <asmodai>
Sure as hell goes faster now that we have the firemath addon for firefox. Saves me a lot of time designing.

[06:17:00.0000] <asmodai>
Designing the formulae that is.

[06:19:00.0000] <jgraham>
Holy crap, the Stix fonts *still* haven't been released

[06:19:01.0000] <asmodai>
jgraham: Should get a site update this week.

[06:19:02.0000] <asmodai>
/me stopped poking them whenever that date passed :)

[06:19:03.0000] <asmodai>
jgraham: http://groups.google.com/group/netscape.public.mozilla.mathml/msg/110fb7f0d1d11c52

[06:19:04.0000] <asmodai>
jgraham: Looks like it's raised by default when used as a postfix operator

[06:19:05.0000] <jgraham>
Well yeah but after hearing they will be ready "real soon now" for the best part of a decade

[06:20:00.0000] <asmodai>
/me grins

[06:20:01.0000] <asmodai>
Must say that standard Windows 7 with FF 3.6 does work decently for MathML

[06:21:00.0000] <Philip`>
I like how the Stix web site's front page has a quote from 2002

[06:23:00.0000] <jgraham>
I like the timeline best of all

[06:24:00.0000] <asmodai>
http://www.w3.org/Math/testsuite/mml2-testsuite/Topics/Primes/primes1.xml <-- I am not sure I understand why they switch between &prime; and ' or not even using &Prime; instead of ''

[06:24:01.0000] <asmodai>
But there they use msup

[06:26:00.0000] <Philip`>
Isn't a font generally a quite predictable thing to schedule?

[06:27:00.0000] <Philip`>
You know you need to create n glyphs, and you work out how long that takes based on the zillions of pretty much identical font development projects that have been done before

[06:28:00.0000] <Philip`>
It's not like they trying to land on the moon or collide hadrons or write a Duke Nukem game

[06:39:00.0000] <asmodai>
Philip`: Yeah, not sure how they managed to let that go on so long

[06:39:01.0000] <boblet>
re: Stix, I’d worry about a group making a font when they use a screenshot of an italic font with Windows-level aliasing. does not bode well

[06:44:00.0000] <Philip`>
boblet: I was going to say they'd probably been using that screenshot since 2002

[06:44:01.0000] <Philip`>
but actually it looks like it was added to the site in 2007

[06:44:02.0000] <boblet>
Philip`: zing!

[06:44:03.0000] <boblet>
given creating a typeface is almost 100% details, you’d think they’d be concerned about … well, details like that

[06:45:00.0000] <Philip`>
Maybe the people who are working on the font have been too busy to update the site, so they got someone else to do it

[06:58:00.0000] <asmodai>
ah

[06:58:01.0000] <asmodai>
jgraham: found something:

[06:58:02.0000] <asmodai>
Not all "mathematical identifiers" are represented by mi elements — for example, subscripted or primed variables should be represented using msub  or msup respectively. 

[06:59:00.0000] <asmodai>
Mmm, but that's a primed variable. *sigh*

[07:21:00.0000] <boblet>
another HTML5 Doctor article: http://html5doctor.com/small-hr-element/

[07:21:01.0000] <boblet>
sorry Mike, nothing to top the mother of all O’s from last time

[07:22:00.0000] <MikeSmith>
boblet: you need a pipe

[07:22:01.0000] <MikeSmith>
in your photo

[07:22:02.0000] <boblet>
lol

[07:23:00.0000] <MikeSmith>
and/or a lab coat

[07:23:01.0000] <MikeSmith>
all y'all htmldoctors should wear lab coats

[07:23:02.0000] <MikeSmith>
or whatever the white coats that doctors wear are called

[07:23:03.0000] <boblet>
yeah, that’s a good idea

[07:23:04.0000] <MikeSmith>
or scrubs

[07:23:05.0000] <boblet>
scrubs might be a bit hardcore

[07:24:00.0000] <boblet>
but doctor/boffin lab coats are a plan

[07:25:00.0000] <workmad3>
you can get lab coats pretty cheap on ebay :)

[07:25:01.0000] <workmad3>
/me has an extra large one at home that cost about £5 or £10

[07:25:02.0000] <MikeSmith>
or you can just photoshop your heads onto existing images of doctors in lab coats

[07:26:00.0000] <workmad3>
that would be good :D especially if you did it deliberately badly ;)

[07:26:01.0000] <workmad3>
or (as a final idea), get a lab coat, and then photoshop the head of a random scrubs castmember onto your own body?

[07:27:00.0000] <boblet>
hehe. these suggestions are gold!

[07:27:01.0000] <boblet>
/me rolls his eyes when everyone looks the other way

[07:27:02.0000] <boblet>
:P

[07:27:03.0000] <workmad3>
I saw that :P

[07:32:00.0000] <boblet>
curses! ;-)

[07:37:00.0000] <annevk>
really, someone needs to rewrite CSS

[07:37:01.0000] <annevk>
CSSOM would work so much better if there was an actual parsing specification that says to create CSSStyleRule objects and all and says to create CSSImportRule objects, that says for @import user agents have to "fetch" style sheets

[07:38:00.0000] <jgraham>
annevk: Are you volunteering?

[07:38:01.0000] <annevk>
and that when they are fetched they are to be parsed, and then CSSStyleSheet objects are to be created etc.

[07:38:02.0000] <annevk>
I have it all in my head, I just don't care about writing it all down because it's so much frigging work

[07:38:03.0000] <annevk>
just writing down the bits specific to CSSOM is already so much work

[07:38:04.0000] <annevk>
it annoys me

[07:40:00.0000] <annevk>
I would be happy to give directions if anyone is interested though

[07:49:00.0000] <MikeSmith>
annevk: what you are describing is what I had thought the CSSOM spec was

[07:49:01.0000] <MikeSmith>
before I actually read it

[07:50:00.0000] <MikeSmith>
I even told other people that's what CSSOM was

[07:50:01.0000] <MikeSmith>
I guess I should actually read specs before I tell other people what they are specs for

[07:51:00.0000] <zcorpan>
that'd take away all the fun

[07:52:00.0000] <Philip`>
MikeSmith: It only becomes a problem if those other people then read the spec, which is unlikely

[07:52:01.0000] <MikeSmith>
heh

[07:52:02.0000] <MikeSmith>
touché

[07:53:00.0000] <annevk>
you thought CSSOM was a CSS parser spec?

[07:54:00.0000] <gsnedders>
We all know CSSOM is the only spec that matters any more.

[07:54:01.0000] <annevk>
I have thought of including a section that actually says how CSSOM integrates with the vague notion of CSS parser in the normal CSS specs

[07:55:00.0000] <MikeSmith>
annevk: for some crazy reason I thought CSSOM probably meant "CSS Object Model"

[07:55:01.0000] <annevk>
which would say that once you parse an @import rule you create a CSSImportRule object, you fetch the indicated style sheet (unless you decide not to based on the media query provided), etc.

[07:55:02.0000] <annevk>
MikeSmith, well it does

[07:56:00.0000] <annevk>
MikeSmith, what I'm complaining about is that the link between object model, syntax, and processing, is missing

[07:56:01.0000] <annevk>
or links

[07:56:02.0000] <Philip`>
MikeSmith: You should think of it as analogous to HTML4 + DOM

[07:56:03.0000] <annevk>
currently we have some notion of syntax which is various CSS specs patched together

[07:56:04.0000] <Philip`>
which define documents, and a document object model, but don't bother with the chasm in the middle

[07:57:00.0000] <annevk>
some vague notion of processing (not at all the detail of e.g. HTML5 fetch)

[07:57:01.0000] <annevk>
and a not so good object model that I'm trying to improve, but I keep running into processing model and syntax to object model mapping issues

[07:57:02.0000] <MikeSmith>
yeah, the chasm is the part I had kind of assumed was provided instead as a mountain of information the spec

[07:59:00.0000] <annevk>
I sort of have the feeling that only a few people care about the level of detail

[07:59:01.0000] <MikeSmith>
this seems similar in lackingness to the HTML4 spec

[07:59:02.0000] <Philip`>
/me 's grammar parser has difficulties towards the end of MikeSmith's statement

[07:59:03.0000] <annevk>
but they are the people that matter if we want this to work in the same way...

[07:59:04.0000] <zcorpan>
or XML/DOM

[08:00:00.0000] <Dashiva>
Argh

[08:00:01.0000] <asmodai>
boblet: Any St Patrick's Day parades in Osaka? :)

[08:00:02.0000] <Dashiva>
Now it's <RSS> instead of <Rich>, I need a new parser

[08:00:03.0000] <MikeSmith>
annevk: if the only few people who are a implementors, then that would seem to make it still pretty important

[08:00:04.0000] <jgraham>
Dashiva: There are plenty of <RSS> parsers on the net

[08:00:05.0000] <jgraham>
Try one of them ;)

[08:01:00.0000] <boblet>
asmodai: nope, although I’m sure any Irish etc pub will be using it for sales

[08:01:01.0000] <MikeSmith>
but if implementors don't care, then I guess it doesn't matter so much who does

[08:01:02.0000] <asmodai>
boblet: heh, I'm sure they would

[08:01:03.0000] <asmodai>
/me misses the kansai region

[08:01:04.0000] <annevk>
MikeSmith, yeah, they're also the people that don't really have much time to give feedback on-the-go

[08:01:05.0000] <annevk>
meh

[08:01:06.0000] <boblet>
you used to be here?

[08:01:07.0000] <Philip`>
At some point there would be few enough people who care about a detail that it takes less effort to get them to converge via bug reports than by writing a spec

[08:01:08.0000] <boblet>
cool!

[08:02:00.0000] <asmodai>
boblet: 3 or 4 times by now

[08:02:01.0000] <annevk>
I'm slowly making progress though; a lot more than two years ago

[08:02:02.0000] <asmodai>
If all goes according to plan, back in May or so

[08:02:03.0000] <boblet>
what for?

[08:02:04.0000] <annevk>
Philip`, yeah, except without a spec it's bad for newcomers

[08:02:05.0000] <asmodai>
boblet: Making my own okonomiyakis is ok, but still nice to taste the real thing there.

[08:03:00.0000] <asmodai>
boblet: visiting friends and such. My ex-gf is a Japanese living up in tochigi-ken, near utsunomiya

[08:03:01.0000] <MikeSmith>
Philip`: I had meant in that sentence to refer to the chasmatic-lackingness-of-necessary-details quality perhaps being similar to the corresponding lackingness in the HTML4 spec

[08:03:02.0000] <boblet>
へー so you’ve lived here rather than visited huh?

[08:04:00.0000] <asmodai>
Visited. never lived.

[08:04:01.0000] <MikeSmith>
Philip`:  relative to the contrasting cornucopia of such detail in the HTML5

[08:04:02.0000] <boblet>
ping Mike or me when you have dates, as there’s a local HTML5 event in May I could take you to :)

[08:04:03.0000] <asmodai>
boblet: Although I guess I saw more of Japanese normal life than most people do when they visit Japan.

[08:05:00.0000] <asmodai>
boblet: I'll keep it in mind :)

[08:05:01.0000] <MikeSmith>
Philip`: or the my-cup-floweth-over level of detail in the HTML5 spec

[08:05:02.0000] <boblet>
having a Japanese gf does tend to give you a different perspective than a tourist

[08:06:00.0000] <asmodai>
boblet: She and her mother were also quite keen on teaching me the reasoning, as far as they knew, behind certain elements of Japanese culture and all that.

[08:06:01.0000] <boblet>
ping me on Twitter @boblet if I’m not here

[08:06:02.0000] <MikeSmith>
asmodai: I live in Tokyo but prefer to socialize exclusively with people from Kansai or Kyushu

[08:06:03.0000] <Philip`>
MikeSmith: I meant the statement before that one

[08:06:04.0000] <MikeSmith>
or anywhere except Tokyo or Kanagawa-ken

[08:06:05.0000] <jgraham>
MikeSmith: Do you have a fetish for people from places starting with K?

[08:06:06.0000] <boblet>
MikeSmith: heh

[08:07:00.0000] <asmodai>
Of course, I was also brought to the Kumon class of her mother's to help the children with their English pronunciation. And being 1,94m tall is always fun. eeeeeeeeeeeee ooki gaijin da na!

[08:07:01.0000] <boblet>
that’s pretty huge

[08:07:02.0000] <asmodai>
MikeSmith: I love kansaiben and the region, maybe it's because I am Dutch or something.

[08:07:03.0000] <asmodai>
boblet: Good thing my ex is something like 1,75m ;)

[08:07:04.0000] <MikeSmith>
jgraham: yes, but even more so for people from places that have K in the name but instead the K is pronounced like "sh" or something for some bizarre reason

[08:08:00.0000] <jgraham>
MikeSmith: That explains so much

[08:08:01.0000] <MikeSmith>
asmodai, boblet - I'm joking about that, anyway

[08:08:02.0000] <MikeSmith>
There are actually some nice people who are native Tokyoites

[08:08:03.0000] <MikeSmith>
or so I have heard, at least

[08:09:00.0000] <MikeSmith>
I trust that it's true

[08:09:01.0000] <boblet>
MikeSmith: no, that sounds a pretty accurate description of your ‘liking ppl’ habits to me :P

[08:13:00.0000] <MikeSmith>
anyway, speaking of sento, it's time for me to go to the sento before it closes

[08:14:00.0000] <asmodai>
I always find people in Tokyo to be a bit nervous.

[08:14:01.0000] <wirepair>
hu? it's 12:20... heh

[08:14:02.0000] <asmodai>
Stressy maybe is more the term.

[08:14:03.0000] <wirepair>
shouldn't it be closed by now?

[08:14:04.0000] <asmodai>
wirepair: I guess not in Tokyo ;)

[08:14:05.0000] <wirepair>
/me is in tokyo

[08:14:06.0000] <wirepair>
;>

[08:15:00.0000] <wirepair>
didn't realize they stayed open that late ;>

[08:15:01.0000] <boblet>
wirepair: Mike’s sento is extra-late

[08:15:02.0000] <wirepair>
guess so

[08:15:03.0000] <boblet>
normally last in 11 close and midnight, but that one is like open til 1am or something

[08:16:00.0000] <boblet>
maybe 1:30 or 2 if Mike’s going now

[08:16:01.0000] <gsnedders>
MikeSmith just lives weird hours.

[08:16:02.0000] <wirepair>
hehe

[08:16:03.0000] <boblet>
sry meant to type ‘last in at 11, close at midnight’

[08:16:04.0000] <Philip`>
gsnedders: It's not just his hours that are weird

[08:17:00.0000] <asmodai>
/me chuckles

[08:17:01.0000] <gsnedders>
Philip`: Have you even laid your eyes upon the weirdness of MikeSmith?

[08:17:02.0000] <gsnedders>
I went running last time I saw him.

[08:17:03.0000] <boblet>
lol

[08:18:00.0000] <boblet>
gsnedders: the ’stache?

[08:18:01.0000] <gsnedders>
boblet: He didn't have that when I saw him

[08:19:00.0000] <boblet>
gsnedders: well if you thought he was funny-lookin’ then…

[08:19:01.0000] <gsnedders>
:D

[08:20:00.0000] <boblet>
Philip`: how come you don’t have Deja Vu Sans Mono on your Web Font Optimizer?

[08:21:00.0000] <Philip`>
boblet: I guess I never (thought|bothered) to add it

[08:22:00.0000] <boblet>
Philip`: come to think of it, how come you aren’t listing them alphabetically *seeing as I like coded that for you and everything*??

[08:22:01.0000] <boblet>
[angry] smiley here

[08:22:02.0000] <Philip`>
Because I haven't updated that site at all since before your feedback

[08:22:03.0000] <boblet>
or maybe the sideways glance one would be appropriate

[08:22:04.0000] <boblet>
raspberries to you

[08:22:05.0000] <Philip`>
/me should probably get back to it some time

[08:23:00.0000] <Philip`>
but I'm hoping if I wait long enough then it'll be obsolete and I won't have to

[08:23:01.0000] <boblet>
i think i tried installing locally but it needed a hack to some library and i gave up

[08:23:02.0000] <boblet>
wonder what @mpilgrim is using…

[08:23:03.0000] <Philip`>
It ought to work with a modern default Perl installation with no fiddling, I believe

[08:23:04.0000] <boblet>
might give it another go

[08:24:00.0000] <Philip`>
The only slightly unusual module is something like Digest::SHA which I think is now standard in Perl 5.10

[08:40:00.0000] <erlehmann_>
boblet, of course, he is using python ;)

[08:41:00.0000] <boblet>
Python? aaw man no wonder. that’s soo 3 years ago. He should be using something newer like Go :P

[08:42:00.0000] <Philip`>
boblet: Did you mean font subsetting? I believe he was using my tool for that

[08:42:01.0000] <Philip`>
(The offline version)

[08:42:02.0000] <boblet>
Philip`: yep, and that’s what I guessed tttt

[08:42:03.0000] <boblet>
hold on, do ppl say tttt for to tell the truth?

[08:43:00.0000] <boblet>
i think i just made that one up

[08:43:01.0000] <boblet>
woops

[08:45:00.0000] <Philip`>
I assumed it was some kind of crazy smiley

[08:45:01.0000] <boblet>
yep, that sounds like me, crazy smiley man

[08:45:02.0000] <boblet>
:|

[08:50:00.0000] <Philip`>
http://ie.microsoft.com/testdrive/

[08:51:00.0000] <Philip`>
http://ie.microsoft.com/testdrive/info/ReleaseNotes/Default.html

[08:51:01.0000] <Philip`>
"XHTML Documents"

[08:52:00.0000] <Philip`>
"SVG"

[08:52:01.0000] <Philip`>
"DOM L2 Events"

[08:53:00.0000] <Philip`>
"The demo of HTML5 <video> at MIX10 was a preview of a future release of the Platform Preview. Stay tuned for more!"

[08:55:00.0000] <zcorpan>
crap, need vista or 7

[08:58:00.0000] <Philip`>
http://blogs.zdnet.com/microsoft/?p=5572 - "Microsoft’s new JavaScript engine is built to take advantage of two cores, with the second core compiling JavaScript down to native machine code to help speed up the browser."

[09:00:00.0000] <AryehGregor>
So now everyone's using JIT.  Awesome.

[09:01:00.0000] <Philip`>
The preview build license doesn't allow you to publish benchmark results

[09:02:00.0000] <gsnedders>
Same was true of IE8, no?

[09:02:01.0000] <annevk>
yeah, that's going to go down well

[09:02:02.0000] <AryehGregor>
So download it, agree to the license, and get someone else to publish benchmark results who didn't agree to the license.

[09:02:03.0000] <AryehGregor>
Easy.

[09:03:00.0000] <Philip`>
That would presumably count as publishing results to the someone else

[09:03:01.0000] <AryehGregor>
Why?  You're not running the benchmark.

[09:03:02.0000] <jgraham>
Not if they run the benchmarks

[09:03:03.0000] <AryehGregor>
It doesn't say you can't let anyone use the browser who didn't agree to the license, right?

[09:03:04.0000] <Philip`>
Feel free to argue that with Microsoft's lawyers :-p

[09:03:05.0000] <Philip`>
or alternatively disregard the license entirely

[09:04:00.0000] <jgraham>
My uninformed guess is that they are more interested in discouraging people by threat then in having anything enforcable

[09:04:01.0000] <jgraham>
*than

[09:04:02.0000] <AryehGregor>
It's so rare that I'm actually excited about a Microsoft product release.

[09:05:00.0000] <zcorpan>
is there a vista vm for xp?

[09:05:01.0000] <erlehmann>
Philip`, what video format are they using ?

[09:05:02.0000] <AryehGregor>
jgraham, my guess is that the provisions are thrown in just in case they'd be handy, not because they even intend to discourage people.

[09:05:03.0000] <erlehmann>
zcorpan, drop XP, it's older than my little sister ;)

[09:05:04.0000] <AryehGregor>
So in the event they do have a lawsuit over something or other, it's one extra claim they can try to get to stick.

[09:06:00.0000] <AryehGregor>
erlehmann, TeX is older than my big brother, but it's still awesome.

[09:06:01.0000] <Philip`>
/me contributes to Hixie's server load

[09:06:02.0000] <Philip`>
(55/100)

[09:06:03.0000] <erlehmann>
:D

[09:08:00.0000] <Philip`>
"GPU-Powered HTML5"

[09:09:00.0000] <AryehGregor>
erlehmann, also, there's no <video> support in this version, but they've said there will be in a future version.

[09:09:01.0000] <AryehGregor>
I assume they'll just use Windows' built-in media stuff, the way Apple uses QuickTime, and they'll support all system codecs.

[09:09:02.0000] <Dashiva>
Oh ho

[09:10:00.0000] <AryehGregor>
Which would mean H.264 will be supported by default, but not Theora, unless they add Theora support.

[09:10:01.0000] <AryehGregor>
AFAICT.

[09:10:02.0000] <AryehGregor>
But who knows, we'll have to see.

[09:10:03.0000] <Dashiva>
If IE9 supports XHTML, maybe we can bypass the whole polyglot documents mess

[09:10:04.0000] <zcorpan>
"The SVG in HTML5 embedding method without using foreign object (i.e. just create an <svg> tag in your HTML file)"

[09:12:00.0000] <svl_>
"We did see one cool HTML5 feature we weren't expecting: IE9 working using YouTube's implementation of the HTML5 video element." <- means h.264 :( http://arstechnica.com/microsoft/news/2010/03/platform-preview-gives-web-developers-first-taste-of-ie9.ars

[09:13:00.0000] <AryehGregor>
I think it was obvious that IE9 would support H.264.  MS is part of MPEG-LA.  The question is whether it would support Theora too.

[09:13:01.0000] <AryehGregor>
I'd guess not, but you never know what MS will do.

[09:14:00.0000] <gsnedders>
They already support Vorbis is some things

[09:15:00.0000] <AryehGregor>
Why is the first picture in the "HTML5" section an Acid3 screenshot, when Acid3 doesn't test any specs that have anything to do with HTML5?

[09:15:01.0000] <gsnedders>
It does, it has the same author

[09:16:00.0000] <AryehGregor>
Also, one thing that Acid3 tests is support for downloadable TTFs, which IE has said they'll never support, so they'll never get 100 anyway.

[09:16:01.0000] <zcorpan>
the links to specs go to dev.w3.org

[09:16:02.0000] <boblet>
so does "We love HTML5 so much that we want to do it right" mean “2022 suckaz!”?

[09:17:00.0000] <zcorpan>
i hope they will send spec feedback; implementing a spec usually involves finding bugs in the spec, but i haven't seen much feedback from ms

[09:17:01.0000] <AryehGregor>
They sent some useful feedback.

[09:17:02.0000] <gsnedders>
But that'd give away they were implementing it

[09:17:03.0000] <gsnedders>
And management wouldn't like that

[09:18:00.0000] <gsnedders>
(oh yay for MS)

[09:18:01.0000] <AryehGregor>
The IE9 preview actually edges out Fx3.7a2 on SunSpider.  Awesome.

[09:19:00.0000] <AryehGregor>
We are finally arriving at an era where the most recent version of every browser is non-crippled.

[09:19:01.0000] <boblet>
AryehGregor: and somewhat embarassing for Moz :)

[09:19:02.0000] <workmad3>
AryehGregor: now we just need to enter the era where all users have the most recent version of a browser

[09:20:00.0000] <AryehGregor>
boblet, they're already doing a major rework of their JS engine, Jager-whatever it is.

[09:20:01.0000] <Philip`>
JaegerMonkey

[09:20:02.0000] <zcorpan>
AryehGregor: unless you get bitten by http://hsivonen.iki.fi/doctype/ie8-mode.png (which will get even more complicated in ie9)

[09:20:03.0000] <jgraham>
Jägermonkey

[09:20:04.0000] <Philip`>
The preview build says it has IE5, IE7, IE8 and IE9 rendering modes

[09:21:00.0000] <boblet>
ya gotta admit, there would have been some happy engineers at Redmond when they finally got a better SunSpider score than a modern browser

[09:21:01.0000] <zcorpan>
wonder if ie10 will replace the ie9 mode or add another

[09:21:02.0000] <workmad3>
http://www.taranfx.com/firefox-hardware-acceleration

[09:21:03.0000] <AryehGregor>
The discrepancy might be because Firefox is falling back to an interpreter in some places.  If the blog posts about Jägermonkey are right, it will beat everyone when that's done.

[09:22:00.0000] <AryehGregor>
Of course, one possible result here is that IE will beat the pants off everyone else by IE10 because Microsoft is pouring billions into it, and then we'll see an IE6-like stagnation.  But that seems a lot less likely now, since Google and Apple have a heck of a lot more money than Netscape ever did.

[09:22:01.0000] <workmad3>
I expect JS performance is going to be a leapfrog race for a year or two until all browsers have superb JS performance :)

[09:22:02.0000] <AryehGregor>
So, I'm extremely optimistic.

[09:22:03.0000] <AryehGregor>
Yay!

[09:22:04.0000] <workmad3>
hell, it's already started

[09:23:00.0000] <gsnedders>
So their JS engine is called "Chakra"

[09:24:00.0000] <workmad3>
so it seems :)

[09:24:01.0000] <AryehGregor>
Yeah, just chart performance over the last couple of years.  Two years ago, no one was even using JITs, right?

[09:24:02.0000] <workmad3>
AryehGregor: don't think so... I think that started with chrome 1 or 2 or something

[09:24:03.0000] <AryehGregor>
Chrome was released less than two years ago, no?

[09:24:04.0000] <workmad3>
good point :)

[09:24:05.0000] <jgraham>
AryehGregor: (without having any particular clue what I'm talking about) I would be pretty surprised if there was another factor of 10 improvement in js peformance

[09:25:00.0000] <AryehGregor>
September 2008.

[09:25:01.0000] <workmad3>
for some reason I thought it was back in my uni days, but it wasn't :)

[09:26:00.0000] <AryehGregor>
jgraham, I imagine that JS in the fastest browsers right now is less than ten times slower than C, so I'd agree.

[09:26:01.0000] <workmad3>
who knows... maybe comps will start using 'javascript cards' that essentially run JS on chip? :)

[09:27:00.0000] <workmad3>
but yeah... I doubt a 10-fold improvement over a year or 2 is likely

[09:28:00.0000] <gsnedders>
I expect significant improvements, still

[09:28:01.0000] <AryehGregor>
Yes, I agree.

[09:28:02.0000] <gsnedders>
But I think nothing so radical, and nothing so sudden now

[09:28:03.0000] <workmad3>
yeah

[09:29:00.0000] <workmad3>
the only reason such a large improvement was really possible is because JS was pretty much ignored and there was a lot of interpreted language improvements that could be applied very quickly to the interpreters

[09:29:01.0000] <workmad3>
such as the JIT compilation :)

[09:29:02.0000] <boblet>
shepazu: so did you have your Mix10 preso already? :)

[09:30:00.0000] <workmad3>
it was almost along the lines of jumping straight from a java JVM from '95 to one from '05 in only a year :)

[09:31:00.0000] <JonathanNeal>
they'll support video that's nice of them.l

[09:32:00.0000] <AryehGregor>
If they support Theora, there's still a chance of it winning the format war.  If not . . . oh well.

[09:34:00.0000] <jgraham>
I would speculate that the best-case realistic possibility is that they support whatever codecs you have installed

[09:34:01.0000] <AryehGregor>
Yes, but they also control what codecs are installed by default.

[09:34:02.0000] <AryehGregor>
Same as Apple.

[09:34:03.0000] <AryehGregor>
So they could choose to add Theora support by default to the OS, or not.

[09:34:04.0000] <AryehGregor>
Up to them.

[09:35:00.0000] <Philip`>
http://blogs.msdn.com/ie/archive/2010/03/16/html5-hardware-accelerated-first-ie9-platform-preview-available-for-developers.aspx

[09:36:00.0000] <gsnedders>
http://ie.microsoft.com/testdrive/benchmarks/Acid3/Default.html

[09:37:00.0000] <Philip`>
"we’re looking for developer feedback on our implementation of HTML5’s parsing rules, Selection APIs, XHTML support, and inline SVG"

[09:37:01.0000] <Philip`>
Parsing rules?

[09:38:00.0000] <gsnedders>
Hmmm...

[09:39:00.0000] <AryehGregor>
Philip`, so that question means "I'm going to go write up a bunch of test cases to see if IE9 actually implements HTML5 parsing", right?

[09:39:01.0000] <gsnedders>
What about the html5lib runner?

[09:39:02.0000] <jgraham>
gsnedders: You read my mind

[09:39:03.0000] <gsnedders>
http://gsnedders.html5.org/html5lib-tests/runner.html is the most up to date version

[09:39:04.0000] <gsnedders>
So if anyone has Vista/Win7 at hand :P

[09:42:00.0000] <boblet>
oyasumi

[09:42:01.0000] <AryehGregor>
I have one, but it's like two rooms away, so I'm too lazy to test it presently.

[09:46:00.0000] <Philip`>
gsnedders: http://philip.html5.org/misc/html5lib-runner-ie9.txt

[09:47:00.0000] <gsnedders>
/me wonders how up to date that copy is

[09:48:00.0000] <jgraham>
/me wonders what's with the </B> elements

[09:49:00.0000] <Philip`>
It seems to *fewer* of those tests when it's in IE8 mode

[09:49:01.0000] <Philip`>
but only a few fewer

[09:51:00.0000] <gsnedders>
Philip`: Try running it again, I've just updated it

[09:51:01.0000] <gsnedders>
(Sorry)

[09:51:02.0000] <zcorpan>
jgraham: presumably unexpected end tags still create "/B" elements

[09:51:03.0000] <zcorpan>
so not a full html5 parser

[09:54:00.0000] <zcorpan>
Test 1 of 24 in data/tests10.dat

[09:54:01.0000] <zcorpan>
|     <svg svg>

[09:54:02.0000] <zcorpan>
and mathml

[09:54:03.0000] <zcorpan>
no wait

[09:54:04.0000] <zcorpan>
they don't parse mathml

[09:55:00.0000] <annevk>
they support SVG in text/html per HTML5 rules?

[09:55:01.0000] <Philip`>
gsnedders: http://philip.html5.org/misc/html5lib-runner-ie9.txt

[09:55:02.0000] <Philip`>
(Updated)

[09:55:03.0000] <gsnedders>
Philip`: Thanks

[09:55:04.0000] <gsnedders>
annevk: yes

[09:56:00.0000] <Philip`>
http://twitter.com/IE/status/10577927335 - "we're going to support the H.264 video codec, it's the de facto standard for video across the broadcasting industry"

[09:56:01.0000] <annevk>
10 years from now we'll finally find out whether the IE-was-the-problem advocates of XHTML were right

[09:56:02.0000] <annevk>
I can't wait

[09:57:00.0000] <erlehmann>
Philip`, all is lost

[09:57:01.0000] <annevk>
Philip`, it's also evil

[09:57:02.0000] <zcorpan>
Test 17 of 24 in data/tests10.dat - they don't break out of foreign lands :-/

[09:57:03.0000] <erlehmann>
maybe the EU should do a default codec ballot screen :P

[10:01:00.0000] <Philip`>
Going by the Zombie DOM Viewer, <foo><bar/><baz> is parsed with XMLish nesting

[10:01:01.0000] <AryehGregor>
Can someone tweet @IE asking if they'll support Theora in addition to H.264, or if they're committed to not supporting Theora?

[10:01:02.0000] <Philip`>
i.e. unknown tags are no longer empty elements

[10:01:03.0000] <Philip`>
and trailing slashes make them closed

[10:01:04.0000] <annevk>
o_O

[10:01:05.0000] <erlehmann>
AryehGregor, as if they would.

[10:01:06.0000] <AryehGregor>
erlehmann, we could ask.

[10:03:00.0000] <Philip`>
Unmatched closing tags of known elements ("</a>" etc) turn into empty elements named "/A" etc in the DOM

[10:03:01.0000] <zcorpan>
Test 21 of 24 in data/tests10.dat - they don't parse xlink:href into a namespaced attribute :-/

[10:03:02.0000] <Philip`>
(</div> and </p> turn into <div> and <p>, though)

[10:03:03.0000] <Philip`>
(and </unknown> is dropped)

[10:04:00.0000] <erlehmann>
AryehGregor, do it, then. i'm not on twitter.

[10:04:01.0000] <AryehGregor>
I'm not either.

[10:04:02.0000] <AryehGregor>
Someone here must be.

[10:04:03.0000] <JonathanNeal>
http://twitter.com/jon_neal

[10:04:04.0000] <JonathanNeal>
there you go AryehGregor.

[10:05:00.0000] <AryehGregor>
I wouldn't have phrased it that way, but okay.

[10:05:01.0000] <JonathanNeal>
What's wrong with the way I phrased it?  I can always tweet again.

[10:05:02.0000] <annevk>
then get your own twitter account AryehGregor

[10:05:03.0000] <annevk>
:)

[10:05:04.0000] <JonathanNeal>
I don't take twitter seriously enough to worry how I look.

[10:06:00.0000] <AryehGregor>
It sounds more confrontational than I'd have said it.

[10:06:01.0000] <Philip`>
http://hsivonen.iki.fi/test/svg-and-mathml-in-html.html displays the nice tiger

[10:06:02.0000] <zcorpan>
if they are changing the parser in radical ways, one wonders why they don't just implement it per spec in one go

[10:06:03.0000] <AryehGregor>
I'd have said something like "Might IE9 support Theora in addition to H.264, or is that out of the question?"

[10:08:00.0000] <erlehmann>
AryehGregor, I would assume that professional PR team behind that twitter account will pack their „no theora 4 u“ message into a nice, fluffy package regardless.

[10:08:01.0000] <AryehGregor>
What a nicely cynical prediction.

[10:08:02.0000] <AryehGregor>
They might just ignore it, though.

[10:17:00.0000] <JonathanNeal>
erlehmann is right, which is why I really didn't care :)

[10:17:01.0000] <JonathanNeal>
They don't care about me, I don't care about them, it's just a call for information

[10:19:00.0000] <annevk>
http://www.zeldman.com/2010/03/16/ie9-preview/ is arguing for convergence on WebKit

[10:20:00.0000] <annevk>
I don't see why Zeldman does not realize that one rendering engine would be quite bad as well...

[10:22:00.0000] <AryehGregor>
It's always nice to have competition.

[10:22:01.0000] <JonathanNeal>
hey paul_irish!

[10:22:02.0000] <jgraham>
Fortunately there is no realistic possibility of Microsoft switching to WebKit

[10:22:03.0000] <erlehmann>
actually, as a non-corporate person, i wonder why.

[10:23:00.0000] <erlehmann>
shouldn't „instant better browser“ be a management metric ?

[10:23:01.0000] <AryehGregor>
Not if it means you can never beat some of your key competitors in certain areas, because they get all your improvements.

[10:23:02.0000] <AryehGregor>
WebKit is largely LGPL.

[10:24:00.0000] <AryehGregor>
Using code from Chrome, now, that might make sense, but it would look bad for Microsoft.

[10:24:01.0000] <AryehGregor>
Probably they figure the extra time and money is worth the benefits to MS's image.  It's already widely viewed as a company that just steals good ideas from others.

[10:24:02.0000] <AryehGregor>
Without having to actually take *code* from others.

[10:25:00.0000] <Philip`>
Microsoft would have its own ideas about how the browser architecture should be developed in order to provide the best experience for their users (e.g. scrap all the cross-platform layers and make it heavily optimised for Direct2D etc), which are likely to conflict with the goals of Apple's WebKit developers

[10:25:01.0000] <Philip`>
so they'd have to fork it instead of contributing upstream

[10:25:02.0000] <AryehGregor>
That's true too.  They're integrating heavily with Windows-specific stuff.

[10:26:00.0000] <AryehGregor>
I think it probably makes sense for them to work on their own.

[10:26:01.0000] <Philip`>
which would mean they lose the benefits of sharing future development effort from other people, and they have to fix all bugs themselves in a codebase which they have no experience with

[10:52:00.0000] <roc_>
"scrap all cross-platform layers" doesn't really buy you that much

[10:53:00.0000] <AryehGregor>
Surely it does if you can ask the Windows team to add any new APIs you'd like.

[10:54:00.0000] <AryehGregor>
(not that I'd know, I don't ever do platform-specific code)

[10:54:01.0000] <asmodai>
mm

[10:55:00.0000] <Philip`>
Asking the Windows team to add APIs doesn't seem a great strategy, because (a) they'll probably say no, and (b) you'll have to wait five years for it to be released and adopted before you can start relying on it

[10:55:01.0000] <asmodai>
http://ie.microsoft.com/testdrive/

[10:56:00.0000] <AryehGregor>
Philip`, I'm talking about the IE team asking the Windows team to add APIs.

[10:56:01.0000] <Philip`>
AryehGregor: So am I

[10:56:02.0000] <AryehGregor>
And backport to Vista if necessary.

[10:56:03.0000] <jgraham>
Philip`: Well IE9 needs things relased since IE8

[10:56:04.0000] <jgraham>
*released

[10:56:05.0000] <AryehGregor>
Why do you think either (a) or (b) is the case?  Since MS views IE as part of its core platform, you'd think it would tell the Windows and IE teams to cooperate.

[10:57:00.0000] <erlehmann>
seems they politely ignored the theora question.

[10:57:01.0000] <AryehGregor>
Also, hasn't MS historically used undocumented APIs in some of their programs?

[10:57:02.0000] <AryehGregor>
That's the same sort of idea.

[10:57:03.0000] <jgraham>
s/things/windows versions/

[10:58:00.0000] <AryehGregor>
I'm pretty sure the various MS teams coordinate with each other to a considerable degree.  Various things Raymond Chen has said tell me that the compiler team does things at request of the Windows team, for example.

[10:58:01.0000] <AryehGregor>
Or at least in coordination with them.

[11:00:00.0000] <asmodai>
Wasn't Opera Mini 5 released out of beta?

[11:00:01.0000] <gsnedders>
Yes

[11:00:02.0000] <gsnedders>
Today

[11:01:00.0000] <asmodai>
Mmm, guess the android market place is not updated yet.

[11:01:01.0000] <gsnedders>
Sometime this morning before I got to the office :P

[11:01:02.0000] <asmodai>
Heh

[11:02:00.0000] <asmodai>
If I install the beta from the market place, will it get the release?

[11:02:01.0000] <gsnedders>
/me claims no knowledge of Mini

[11:02:02.0000] <asmodai>
Heh

[11:02:03.0000] <gsnedders>
Provided the rendering engine works…

[11:02:04.0000] <asmodai>
I assume they'll replace the beta with the release version on the market place. Lets just install it

[11:07:00.0000] <Philip`>
AryehGregor: I don't know how closely teams cooperate, but it seems there's a fair amount of independence

[11:07:01.0000] <AryehGregor>
That kind of statement is hard to disagree with.

[11:07:02.0000] <Philip`>
e.g. "The OS team decides what OS features to update in patches. The hardware team decides what hardware to release. The two teams do not necessarily have the same priorities, especially since the OS team views the hardware team as "just another hardware vendor."" (http://blogs.msdn.com/oldnewthing/archive/2008/08/06/8835316.aspx)

[11:07:03.0000] <AryehGregor>
Hardware probably isn't as important to MS as IE is.

[11:07:04.0000] <MikeSmith>
hats off to the IE product-dev team

[11:08:00.0000] <Philip`>
and I'd expect that to some extent the IE team is seem as just another software vendor (though a very important one, but not necessarily more important than all third-party vendors)

[11:09:00.0000] <MikeSmith>
this is real progress and always good to leave the audience wanting more

[11:09:01.0000] <annevk>
given that Trident is used by other Microsoft software and Windows itself wouldn't it be more important?

[11:09:02.0000] <AryehGregor>
MS views IE as a key part of Windows, so I'd say it's probably viewed as a lot more important than most vendors.

[11:09:03.0000] <Philip`>
The IE-favouritism and undocumented API things got Microsoft into a little bit of legal bother, so I'm not sure they want to do much more of that

[11:09:04.0000] <AryehGregor>
They can still release documented APIs.  Probably no one will sue them too much for that.

[11:10:00.0000] <roc_>
it's not good to leave the audience wanting <canvas>, though :-)

[11:10:01.0000] <MikeSmith>
annevk: as far as the webkit-convergence think, I would think that most people would be smart enough to never take advice from a guy with a mustache, or a hat, or worse yet a mustache and a hat.. because that guy is clearly trying to hide something

[11:11:00.0000] <AryehGregor>
/me has both a mustache and a hat.

[11:11:01.0000] <AryehGregor>
Although I have a beard too, so I don't know if it counts.

[11:11:02.0000] <MikeSmith>
beard is a sign of wisdom

[11:11:03.0000] <MikeSmith>
prophets must have beards

[11:11:04.0000] <MikeSmith>
nobody takes them seriously otherwise

[11:12:00.0000] <MikeSmith>
but how many prophets can you think of who just had a mustache?

[11:12:01.0000] <annevk>
IE9 does not have <canvas>?

[11:12:02.0000] <annevk>
weird

[11:12:03.0000] <Philip`>
No(t yet)

[11:13:00.0000] <AryehGregor>
Why is it weird?

[11:13:01.0000] <jgraham>
What's the point of a fast js engine if you can't play js games?

[11:13:02.0000] <AryehGregor>
Most of the stuff they're adding seems like it's higher-priority than <canvas>.

[11:13:03.0000] <AryehGregor>
Performance, <video>, SVG.

[11:13:04.0000] <Philip`>
They'll apparently update the platform preview every 8 weeks or so

[11:13:05.0000] <AryehGregor>
Heh.

[11:13:06.0000] <AryehGregor>
Point.

[11:13:07.0000] <MikeSmith>
hopefully it's just the first beta and canvas will be added shortly

[11:13:08.0000] <AryehGregor>
It's not a beta, it's pre-beta.

[11:13:09.0000] <Philip`>
and maybe they'll add new features then, when they're more confident they'll definitely be able to ship them

[11:13:10.0000] <AryehGregor>
But that would be awesome, yes.

[11:14:00.0000] <Philip`>
jgraham: It might be useful for, like, useful things instead

[11:14:01.0000] <MikeSmith>
AryehGregor: is your beard a serious manly beard, or more like a billy-goat beard?

[11:14:02.0000] <Philip`>
like email and office applications

[11:15:00.0000] <AryehGregor>
/me wonders why the heck he gets pictures of naked women when Googling "beard"

[11:15:01.0000] <Philip`>
Personalised search?

[11:15:02.0000] <AryehGregor>
. . .

[11:15:03.0000] <othermaciej>
I would have thought <canvas> is higher priority than SVG given author demand

[11:15:04.0000] <AryehGregor>
My beard is kind of like this, just a half-inch trim overall: http://en.wikipedia.org/wiki/File:JoeyHerrington.png

[11:15:05.0000] <AryehGregor>
Except darker-colored.

[11:16:00.0000] <AryehGregor>
SVG is older than <canvas>, I guess, and better known, and part of Acid3 . . . dunno.

[11:16:01.0000] <AryehGregor>
Personally, I skimmed the canvas section of the spec, because I don't understand all this crazy graphics stuff.

[11:17:00.0000] <jgraham>
Philip`: Typically email and office applications aren't really limited by raw execution performance afaict

[11:17:01.0000] <MikeSmith>
that beard is too tame.. beards should be wild and unruly and awe-inducing

[11:17:02.0000] <jgraham>
(obviously some applications could be but the obvious examples would also want <canvas>)

[11:18:00.0000] <AryehGregor>
C'mon, benchmarks are the only reason to improve JS performance.  Doesn't everyone know that?

[11:18:01.0000] <Philip`>
jgraham: I assume they occasionally are limited by JS performance, for a short period of time that is long enough to cause visible latency, which is a good enough reason to optimise JS execution

[11:19:00.0000] <Philip`>
(even if most of the time they're idle or network-limited)

[11:20:00.0000] <MikeSmith>
here's a dude with substantial beard:

[11:20:01.0000] <MikeSmith>
http://offthemeatrack.files.wordpress.com/2007/11/mel4.jpg

[11:21:00.0000] <MikeSmith>
he looks a bit sad and confused, though

[11:21:01.0000] <MikeSmith>
I guess sometimes prophets gets sad and confused too

[11:21:02.0000] <MikeSmith>
anyway, we do need canvas

[11:22:00.0000] <jgraham>
Philip`: There was a microsoft research paper suggesting that they were more likely to be limited by DOM + Events

[11:22:01.0000] <MikeSmith>
hopefully some soonly upcoming non-pre beta will add that

[11:22:02.0000] <jgraham>
Although I didn't think the paper was that great

[11:23:00.0000] <jgraham>
(I mean it was fine, but the conclusions were not that surprising)

[11:23:01.0000] <AryehGregor>
Someone can try tweeting them to ask if they're planning on <canvas>.

[11:23:02.0000] <gsnedders>
You'll get no answer

[11:23:03.0000] <gsnedders>
It's futile

[11:23:04.0000] <AryehGregor>
Also, if more people harass them on if they'll support Theora in addition to H.264, maybe they'll answer that.

[11:23:05.0000] <AryehGregor>
gsnedders, they answered a video codec question.

[11:23:06.0000] <AryehGregor>
You never know.

[11:24:00.0000] <Philip`>
That question was already answered in published material

[11:24:01.0000] <Philip`>
I don't think they're going to release any new information via Twitter

[11:24:02.0000] <AryehGregor>
Hmm, where?

[11:24:03.0000] <AryehGregor>
I didn't see it.

[11:25:00.0000] <Philip`>
http://blogs.msdn.com/ie/archive/2010/03/16/html5-hardware-accelerated-first-ie9-platform-preview-available-for-developers.aspx - "we showed HTML5 video support at the MIX conference, as well as how HTML5 video (specifically industry standard HD-encoded, H.264 720p) [...]"

[11:25:01.0000] <miketaylr>
here, i believe, AryehGregor: http://twitter.com/IE/status/10577927335

[11:25:02.0000] <miketaylr>
oops

[11:25:03.0000] <AryehGregor>
Philip`, point.

[11:25:04.0000] <AryehGregor>
Okay, then, Twitter is probably pointless.

[11:25:05.0000] <AryehGregor>
(as usual)

[11:27:00.0000] <MikeSmith>
or if we want it to become true, we could just do I want things to become true, which is, I just add them to Wikipedia with a citation to some site I control where I emphatically affirm that it's true, quoting "sources that asked to remain unnamed in order to protect their confidentiality"

[11:27:01.0000] <AryehGregor>
Revert, personal website, does not meet [[WP:RS]].

[11:28:00.0000] <Philip`>
Then you wait until a news site cites the wiki article, and then you make the wiki article cite the site

[11:28:01.0000] <MikeSmith>
it worked for the Judith "Weapons of Mass Destruction" Miller and the New York Times

[11:28:02.0000] <AryehGregor>
I wish I had a good response to that, but I don't, because it's actually happened.

[11:28:03.0000] <AryehGregor>
The moral for me is that Wikipedia shouldn't cite journalists on subjects where you could conceivably cite anything more authoritative.

[11:28:04.0000] <Philip`>
You can verify the consistency of statements, but you can't verify the truth

[11:28:05.0000] <MikeSmith>
it's trivial to do, despite the editards

[11:29:00.0000] <Philip`>
and it's easy to set up consistent loops of untruths

[11:29:01.0000] <MikeSmith>
yes

[11:29:02.0000] <MikeSmith>
and easy to base a whole career on them

[11:29:03.0000] <AryehGregor>
Philip`, I'm fairly sure that I can verify the truth of the statement "Manhattan is not currently a desolate wasteland after a surprise nuclear strike by China this morning."

[11:30:00.0000] <MikeSmith>
witness Charles Pellegrino and Henry Holt and the Last Train to Hiroshima saga

[11:30:01.0000] <Philip`>
AryehGregor: You can only verify it by determining that it is inconsistent with your observations

[11:30:02.0000] <MikeSmith>
I hope China does that just to rid us forever of a setting for Sex and the City

[11:31:00.0000] <erlehmann>
MikeSmith, then it would continue like CSI … SatC: Miami

[11:31:01.0000] <AryehGregor>
Philip`, however, I can verify the truth of the observation its negation would be inconsistent with, insofar as I'm in Manhattan right now and can directly observe that it's not a desolate wasteland.

[11:31:02.0000] <AryehGregor>
(at least, not the whole thing; haven't checked downtown lately)

[11:32:00.0000] <miketaylr>
as long as they keep the fallout from my neighborhood in southern brooklyn, i'm okay with that scenario as well

[11:32:01.0000] <MikeSmith>
Miami is high on the list too.. when I was in Texas and watched Armageddon when it first came out, everybody cheered with the aliens destroyed Houston

[11:32:02.0000] <roc_>
personally I would have prioritiized canvas higher than SVG

[11:32:03.0000] <roc_>
it's a lot easier, and probably more used on the Web

[11:33:00.0000] <Philip`>
Is it used by any of the top 7000 sites?

[11:34:00.0000] <MikeSmith>
the waste toxicity of the east river provides a barrier that would protect Brooklyn for a Manhattan attack

[11:34:01.0000] <miketaylr>
truth.

[11:34:02.0000] <MikeSmith>
roc_: I would rather they'd done both

[11:34:03.0000] <roc_>
I kind of expect they will

[11:35:00.0000] <MikeSmith>
yeah

[11:35:01.0000] <roc_>
Philip`: I don't know, but I bet SVG isn't

[11:35:02.0000] <Philip`>
I doubt they'd ask so many questions about the spec if they didn't have some intention of implementing it at some point

[11:36:00.0000] <Philip`>
They've probably asked more questions about canvas than about all the parts of HTML5 they've already implemented and shipped or announced, combined

[11:36:01.0000] <Philip`>
(Admittedly that's still only about five questions)

[11:36:02.0000] <MikeSmith>
maybe their canvas needs more work and they judged that having a partial SVG implementation was OK but that waiting to have a non-partial canvas would be better

[11:36:03.0000] <othermaciej>
Philip`: I believe Google Maps can optionally use <canvas>

[11:37:00.0000] <roc_>
Google Maps can use SVG too

[11:37:01.0000] <gsnedders>
That is true

[11:37:02.0000] <roc_>
implementing canvas on top of Direct 2D is really quite easy

[11:37:03.0000] <Philip`>
Depends if you care about all the edge cases

[11:37:04.0000] <Philip`>
(I'd guess)

[11:38:00.0000] <roc_>
maybe, but existing browsers don't get all the edge cases right either

[11:38:01.0000] <hsivonen>
If google maps can use SVG, why doesn't it?

[11:38:02.0000] <roc_>
it would certainly be easy to get something going for a "tech preview"

[11:39:00.0000] <roc_>
hsivonen: it does, it uses SVG for the routes in Gecko, last I checked

[11:39:01.0000] <hsivonen>
roc_: ah. not for the map, though

[11:39:02.0000] <roc_>
for the actual map tiles, I expect SVG is a giant performance lose

[11:39:03.0000] <othermaciej>
I think it is capable of using CSS transforms and transitions for the map tiles now

[11:41:00.0000] <hsivonen>
does the IE9 preview run on XP?

[11:41:01.0000] <Philip`>
No

[11:41:02.0000] <hsivonen>
are they doing video using Vista/7-only APIs?

[11:41:03.0000] <Philip`>
Yes

[11:42:00.0000] <Philip`>
They're seemingly doing all rendering using Vista/7-only APIs

[11:42:01.0000] <hsivonen>
that's an interesting decision

[11:42:02.0000] <Philip`>
(Direct2D)

[11:42:03.0000] <hsivonen>
I guess I need to purchase another Windows license in order to poke at IE

[11:43:00.0000] <Dashiva>
I suppose it's time to try out DX 10+ too...

[11:43:01.0000] <Philip`>
Unsurprisingly, it doesn't appear to work in Wine

[11:43:02.0000] <miketaylr>
hsivonen: i found out the hard way: http://miketaylr.com/post/dbf12ce8.png

[11:44:00.0000] <roc_>
that's probably a good decision for them

[11:44:01.0000] <Philip`>
(http://ie.microsoft.com/testdrive/info/FrequentlyAskedQuestions/Default.html mentions XP)

[11:44:02.0000] <miketaylr>
ah yes, FAQS. i've heard of those.

[11:44:03.0000] <no_mind>
why doesnt microsft puts IE to rest

[11:45:00.0000] <Philip`>
"Will Internet Explorer 9 run on Windows XP? It’s too early to talk about features of the Internet Explorer 9 Beta."

[11:45:01.0000] <Philip`>
They're not saying it *won't* run

[11:45:02.0000] <Philip`>
so they could add a non-hardware-accelerated rendering mode, perhaps

[11:46:00.0000] <Dashiva>
(I think people sticking to IE6 are a bigger problem than XP users on IE8, overall)

[11:47:00.0000] <hsivonen>
does D2D work on VirtualBox?

[11:49:00.0000] <Philip`>
IE9 works for me in VMware on Linux

[11:49:01.0000] <roc_>
I don't know

[11:49:02.0000] <roc_>
VMWare does support D2D, accelerated even

[11:49:03.0000] <Philip`>
so it's not getting real hardware access there

[11:50:00.0000] <roc_>
VMWare does support some level of hardware acceleration

[11:50:01.0000] <Philip`>
/me has no idea what layers perform emulation in this case

[11:51:00.0000] <Philip`>
My hardware doesn't really support much acceleration, and I'd be shocked if the OS/driver/X11/VMware/etc compatibility was good enough for it to reach all the way through to a VM :-)

[11:52:00.0000] <hsivonen>
well, I guess VMWare has the record-and-replay bonus, although it's not relly productized on Linux yet

[11:53:00.0000] <hsivonen>
/me just decided against purchasing Windows 7 plus VMWare two weeks ago. 

[11:53:01.0000] <Philip`>
/me is using the free VMware Player so it doesn't do anything fancy like that

[11:54:00.0000] <Philip`>
For some reason it doesn't even like transmitting mouse-up events if the cursor is not in the top-left few hundred pixels of the screen, either

[11:54:01.0000] <Philip`>
which makes resizing windows tricky

[12:08:00.0000] <Philip`>
http://msdn.microsoft.com/en-us/ie/ff468705.aspx

[12:08:01.0000] <Philip`>
"Added HTML5 support 

[12:08:02.0000] <Philip`>
* Improved parsing of generic elements

[12:08:03.0000] <Philip`>
* Improved parsing of overlapping tags

[12:08:04.0000] <Philip`>
* script and style block parsing changes

[12:08:05.0000] <Philip`>
* Text selection APIs"

[12:14:00.0000] <Philip`>
"Internet Explorer Platform Preview follows the HTML5 draft spec and resolves overlapping tags at parse time."

[12:25:00.0000] <zcorpan>
"* script and style block parsing changes" - wonder what they changed. looking at html5lib tests they haven't implemented the double escaped dash dash stuff

[12:25:01.0000] <gsnedders>
No reparsing?

[12:25:02.0000] <gsnedders>
But if they have the old stuff from the spec…

[12:25:03.0000] <gsnedders>
Well, I guess they have their compat modes

[12:27:00.0000] <zcorpan>
i think they still reparse if i interpret the output correctly

[12:27:01.0000] <Philip`>
zcorpan: It says later on

[12:28:00.0000] <Philip`>
"Internet Explorer 8 and earlier versions do not pass text, in either a script or style block, to text nodes in the DOM. Internet Explorer Platform Preview persists text in script or style blocks to the DOM as text nodes, as expected. This enables you to display source code on your page using script."

[12:28:01.0000] <zcorpan>
Philip`: ok. i don't see that in the html5lib test output though

[12:29:00.0000] <zcorpan>
oh wait

[12:29:01.0000] <zcorpan>
i was looking at tests without a doctype

[12:30:00.0000] <zcorpan>
that explains it

[12:49:00.0000] <Dashiva>
"Bray is the co-inventor of the XML Web standard"

[12:49:01.0000] <Dashiva>
Does "invent" make sense for something like XML?

[12:51:00.0000] <erlehmann>
Dashiva, what would you say ? That he discovered the fundamentals of XML ? ;)

[12:51:01.0000] <gsnedders>
I invented a subset of SGML!

[12:51:02.0000] <erlehmann>
That's nothing ! I invented a subset of ASCII containing only the letter x !

[12:52:00.0000] <erlehmann>
xxxxxxxxxxxxxx … see ?

[12:52:01.0000] <gsnedders>
Wow.

[12:52:02.0000] <gsnedders>
Impresssive.

[12:53:00.0000] <Dashiva>
erlehmann: I would say designer, editor, or even architect

[12:53:01.0000] <Dashiva>
Inventor sounds wrong to me

[12:53:02.0000] <zcorpan>
what about Namespaces?

[12:55:00.0000] <erlehmann>
Dashiva, fine. I hereby proclaim myself the architect of lowercase-x-notation !

[12:56:00.0000] <erlehmann>
to collect royalties, x-LA will be set up soon ;)

[12:57:00.0000] <erlehmann>
i think inventor is fundamentally right, even if it may ring the wrong bells here or there

[12:57:01.0000] <Dashiva>
Namespaces are evil enough that I can imagine a mad scientist inventing them

[12:58:00.0000] <erlehmann>
you mean a mad engineer

[12:58:01.0000] <erlehmann>
unless there's a hypothesis hidden in the spec

[12:58:02.0000] <erlehmann>
tag soup parsers are the control group !

[13:00:00.0000] <Dashiva>
The hypothesis was that people would actually use it

[13:01:00.0000] <erlehmann>
mad science indeed

[13:15:00.0000] <annevk>
Hixie, your response on the progress element question does not really answer the question

[13:35:00.0000] <Philip`>
/me sees a Q&A video with Dean Hachamovitch

[13:36:00.0000] <Philip`>
in which the second question the host asks is about canvas support

[13:37:00.0000] <Philip`>
and he doesn't say anything specific but notes that all of IE9's graphics are hardware-accelerated, and that there are more things to come than have been announced

[15:14:00.0000] <zcorpan>
so ie9 won't need the html5 shiv

[15:18:00.0000] <othermaciej>
nice

[15:18:01.0000] <othermaciej>
they have correct handling of unknown elements?

[15:18:02.0000] <zcorpan>
no, <foo/> gets self-closed

[15:19:00.0000] <jgraham>
http://research.microsoft.com/en-us/um/people/cormac/papers/2009/SoLongAndNoThanks.pdf seems like it might be interesting

[15:21:00.0000] <jgraham>
/me notices the text of the license on that paper forbids "redistribution to lists", wonders what that means and whether posting a URL counts

[15:23:00.0000] <Philip`>
Summary: "Microsoft recommends leaving users ignorant to simple security vulnerabilities"

[15:24:00.0000] <Philip`>
although, in the interest of not entirely blatantly misrepresenting it, I suppose I should note that it ends with "this paper is not to be read as an encouragement to end-users to ignore security policies or advice. The opinions expressed are those of the author."

[15:25:00.0000] <Philip`>
(Also that was my summary, not a direct quote)

[15:25:01.0000] <jgraham>
Philip`: I haven't read it all yet but it doesn't seem to say that, really

[15:26:00.0000] <jgraham>
It notes that it is somewhat rational to ignore security advice if the cost of understanding and following that advice is greater than the expected loss from not following that advice

[15:26:01.0000] <Philip`>
Sure, but worrying about what it really says is not the way to write headlines on tech news sites

[15:27:00.0000] <jgraham>
Which seems like a major over simplification since you need to look not just at the expected loss but at the distribution of possible losses

[15:27:01.0000] <Philip`>
and for most people the cost of reading the actual paper is less than the benefit they gain from doing so

[15:27:02.0000] <jgraham>
e.g. if you can have all your money stolen that is really bad even if there is only a 1 in a million chance it happens to you and so the expected loss is < $1

[15:28:00.0000] <Philip`>
Why is that really bad?

[15:28:01.0000] <Philip`>
Or, (why) is it more bad than two million people each guaranteed to lose < $1?

[15:30:00.0000] <jgraham>
I guess the total possible effect on your life scales more than linearly with the amount of money lost (beyond a certian point)

[15:35:00.0000] <Philip`>
jgraham: Only up to a certain point

[15:35:01.0000] <Philip`>
with that point being when you are bankrupt

[15:35:02.0000] <Philip`>
and then you don't care any more

[15:35:03.0000] <jgraham>
(the other problem with this typ of calculation is that the asumption of rational behaviour is totally flawed. In particular people are probably totally incapable of estimating the probability of a given attack, so if they happen to make "rational" decisions it is probably happenstance)

[15:37:00.0000] <Philip`>
(Incidentally, would you suggest it is rational to play the lottery, because super-linear scaling means a 10^-7 chance of winning £1M outweighs a 10^0 chance of losing £1?)

[15:38:00.0000] <Dashiva>
Philip`: No, the value of the million is subject to decreasing value since it's so big

[15:38:01.0000] <Philip`>
The abstract sounds ambiguous about the suggested behaviour

[15:38:02.0000] <roc>
you can argue that due to various kinds of poverty traps, playing the lottery is rational in some cases

[15:39:00.0000] <Philip`>
i.e. whether it means user behaviour (driven by ignorance and laziness) happens to coincide with rational behaviour, or whether it's actually suggesting users are making rational evaluations themselves

[15:39:01.0000] <Dashiva>
I think the paper should also be read as implementor-facing: "These behaviors are bad for users, so don't make your security depend on them"

[15:40:00.0000] <jgraham>
Dashiva: That was the interesting point of view for me

[15:40:01.0000] <othermaciej>
is risk-aversion natural?

[15:40:02.0000] <othermaciej>
er rational?

[15:40:03.0000] <othermaciej>
most people dislike losing what they have more than they like gaining something

[15:40:04.0000] <Dashiva>
only if they understand the risk, though

[15:40:05.0000] <roc>
it depends

[15:41:00.0000] <roc>
I think the main point of the paper is that we tend to not rationally consider the costs of security measures

[15:41:01.0000] <othermaciej>
it's a well-known result in behavioral economics: http://en.wikipedia.org/wiki/Endowment_effect

[15:41:02.0000] <roc>
we == vendors and researchers

[15:42:00.0000] <roc>
asking everyone in the world to do something that takes only a few seconds is actually a tremendous cost

[15:42:01.0000] <roc>
that's a humbling thought for browser developers :-)

[15:42:02.0000] <othermaciej>
otoh I don't think endowment effect applies in the case of a very low probability of a large loss

[15:42:03.0000] <othermaciej>
this is why I hate conformation dialogs as a security measure

[15:43:00.0000] <othermaciej>
even if everyone read it and carefully considered the implications, the cost would likely still exceed the benefit

[15:43:01.0000] <roc>
people are irrationally fearful of certain kinds of losses, like shark attacks

[15:43:02.0000] <Philip`>
When there was a 2 minute silence in America for some people who had died, I wondered how many lifetime's worth of productivity was being lost because of that

[15:43:03.0000] <Philip`>
But it was only about 20 years of productivity, so it wasn't even a whole person

[15:43:04.0000] <othermaciej>
people are scared of low-probability risk of death or serious injury, if the scenario is particularly vivid

[15:44:00.0000] <othermaciej>
that's another well-known cognitive bias

[15:44:01.0000] <roc>
so there's the answer

[15:44:02.0000] <roc>
"if you don't look at this site's certificate, you will be eaten by a shark"

[15:44:03.0000] <Dashiva>
Philip`: Sort of like that rant about emails sent to large lists

[15:44:04.0000] <othermaciej>
however it doesn't tend to apply to risks where the consequences seem commonplace

[15:44:05.0000] <othermaciej>
especially where costs of avoiding the risk would require ongoing active effort

[15:45:00.0000] <Dashiva>
(For certificates, pretending it's http seems like a great solution to me)

[15:47:00.0000] <othermaciej>
I think the paper ignores the fact that certain security measures have a very high false positive rate only because they exist

[15:47:01.0000] <othermaciej>
for example, if there was no error for invalid certs, a lot more phishing sites would use them

[15:47:02.0000] <jgraham>
I was going to point out the same thing

[15:47:03.0000] <othermaciej>
they only avoid using them because there is a scary warning

[15:48:00.0000] <Dashiva>
But as long as the sites weren't flagged as valid certs, it would be no different from now in effeciency

[15:48:01.0000] <roc>
othermaciej: is that actually true

[15:48:02.0000] <roc>
?

[15:49:00.0000] <othermaciej>
roc: I haven't seen the full counter-factual world of high degree of phishing/pharming but no warning whatsoever on invalid certs

[15:49:01.0000] <othermaciej>
but it may mean the level of scariness of the warning is nearly irrelevant

[15:49:02.0000] <roc>
right

[15:50:00.0000] <othermaciej>
roc: I do know that at least some malware modifies your local DNS settings and system root certs

[15:50:01.0000] <Philip`>
http://news.netcraft.com/archives/2010/03/05/new_browser_reports_over_half_of_ssl_sites_may_be_unsafe.html

[15:50:02.0000] <Philip`>
"A new internet browser, Comodo Dragon, reports that more than half of the world's valid SSL certificates are unsafe."

[15:50:03.0000] <othermaciej>
whence I conclude that it is at least somewhat valuable to attackers to be able to appear 100% like the real site, if they were able to do so

[15:50:04.0000] <Philip`>
Also: "Although Comodo states that many websites connected to cyber-crimes use domain-validated certificates, Netcraft's phishing site feed shows that only 0.3% of reported phishing sites use HTTPS, including those running on compromised servers with SSL certificates already in place."

[15:51:00.0000] <othermaciej>
phishing from a non-SSL server is extremely effective

[15:52:00.0000] <othermaciej>
it's hard to tell how much marginal impact there is on the success rate of phishing from measures that make it harder to look exactly like the real site in all ways

[15:52:01.0000] <roc>
right

[15:53:00.0000] <othermaciej>
the people aware enough to benefit from the protection were already at much lower risk of phishing in the first place

[15:54:00.0000] <roc>
you might be right that phishers would be more likely to use invalid certs if there was no warning, but I haven't seen any evidence that that would actually make a difference to users

[15:58:00.0000] <Dashiva>
<a href="www.evil.com">www.PayPal.com</a>

[15:58:01.0000] <Dashiva>
Nice example :)

[16:00:00.0000] <Philip`>
It'd be a nicer example if they hadn't forgotten the http://

[16:05:00.0000] <Dashiva>
You could draw a connection from cost-vs-benefit of security to the otherwise normal practice of insurance

[16:05:01.0000] <Dashiva>
It's not irrational to pay more than expected average risk if it lets you avoid large worst case risks

[16:17:00.0000] <othermaciej>
indeed, but it depends on the magnitude and likelihood of the risk

[16:20:00.0000] <Dashiva>
The paper also ignores non-monetary costs borne by users

[16:21:00.0000] <Dashiva>
If you get phished, the bank doesn't just give you the money right away, there's probably many hours of calling and talking and writing to recover it.

[16:27:00.0000] <AryehGregor>
Philip`, for an example of the Windows team doing stuff for the IE team, look at Vista's Low Integrity mode.  I've seen a lot of stuff that implies that was intended mainly for IE -- official MS sources tend to mention low-integrity mode and protected IE in the same breath.

[16:28:00.0000] <AryehGregor>
jgraham, the usual assumption of economists is that rational agents should attempt to maximize expected utility.  Of course, utility doesn't map linearly to money, but the actual distributions don't matter, only the expectation (of utility).  There are pretty good theoretical reasons for this; if you don't do this, you can be money-pumped (at least sometimes).

[16:30:00.0000] <AryehGregor>
Also, people often do tend to act rationally, and not just by chance.  Your average person hasn't been hit by online fraud, no one they knows has, and they notice that no one else they know seems to be very concerned, so they (correctly) conclude that it's not a big problem for them.

[16:32:00.0000] <AryehGregor>
Any way you slice it, it's clear that things like raising a big scary warning for cert errors are just stupid.

[16:32:01.0000] <AryehGregor>
It's security charade.

[16:40:00.0000] <Dashiva>
Dang, only 55 on Acid3

[16:40:01.0000] <Dashiva>
My prediction was 60

[16:43:00.0000] <AryehGregor>
It will probably exceed 60 by the final release.

[16:44:00.0000] <AryehGregor>
Possibly even by the beta.

[16:44:01.0000] <Dashiva>
My prediction was for MIX, though

[16:45:00.0000] <AryehGregor>
Oh.  Then you fail.  Pretty close, though, good job anyway.

[16:49:00.0000] <zcorpan>
"if you don't look at this site's certificate, you will be eaten by a shark" - LOL

[16:51:00.0000] <Dashiva>
It's a trick, the shark is sneaking up behind you while you read it


2010-03-17
[17:38:00.0000] <wycats>
Hixie: you around?

[18:52:00.0000] <TabAtkins>
Yay, finally in California!

[18:52:01.0000] <TabAtkins>
/me 's roadtrip has gone on far too long.

[18:58:00.0000] <othermaciej>
heya TabAtkins 

[18:58:01.0000] <TabAtkins>
Yo maciej.

[20:09:00.0000] <GPHemsley>
Hixie: Thanks for combining all your replies based on their general topic. :)

[20:09:01.0000] <GPHemsley>
Hixie: It would have been even better if you'd included the original subject before each quote, though. ;)

[21:05:00.0000] <Hixie>
GPHemsley: unfortunately pine doesn't make that very easy :-(

[23:19:00.0000] <asmodai>
So is that platform preview of IE9 installable next to IE8 or will it overwrite?

[23:21:00.0000] <asmodai>
ah, looks like it can be installed side by side

[23:21:01.0000] <asmodai>
We designed the Platform Preview to be installed, side by side with IE8. The Platform Preview is not a replacement for your daily browser; [...]

[23:21:02.0000] <hsivonen>
personal sites aren't WP:RS? Mine seems to be.

[23:21:03.0000] <asmodai>
WP:RS?

[23:27:00.0000] <hsivonen>
asmodai: reputavble source for wikipedia

[23:28:00.0000] <othermaciej>
their notion of what is a reputable source always strikes me as confusing

[23:32:00.0000] <asmodai>
hsivonen: Aye.

[23:32:01.0000] <asmodai>
othermaciej: It's based on whatever whim some editor now has.

[23:33:00.0000] <othermaciej>
there's pages where a direct edit from me would be original research, my email to someone else would not be a reliable source, but me posting the same info on the webkit.org does count as a reliable source

[01:43:00.0000] <asmodai>
http://desandro.com/resources/curtis-css-typeface/

[02:42:00.0000] <asmodai>
Mmm, IE9 breaks down on the DOM attributes (Prototype) of dromaeo

[02:46:00.0000] <hsivonen>
can a window or window.document object be queried for its intent to navigate?

[02:48:00.0000] <hsivonen>
problem: I have a library method that takes a window object and tries to decide of the window object is done navigating and loading

[02:48:01.0000] <hsivonen>
if the window object came from window.open immediately before, it'll have about:blank in readyState "complete"

[02:48:02.0000] <hsivonen>
so it's indistinguishable from a window that intends to stay as about:blank

[03:28:00.0000] <hsivonen>
XML is still awesome: http://hsivonen.iki.fi/screen/visitmix-ysod.png

[03:29:00.0000] <asmodai>
hsivonen: hehehe

[03:38:00.0000] <asmodai>
Kuruma: I love the nickname :) 車ですね。。。

[03:42:00.0000] <othermaciej>
hsivonen: seems to be fixed now - I guess someone noticed?

[03:44:00.0000] <jgraham>
Just another benefit of XML: when you have a minor irrelevant problem someone has to scramble to fix it urgently

[03:46:00.0000] <othermaciej>
XML probably made more sense when most content was static

[03:46:01.0000] <othermaciej>
did IE9 get a negative response in some quarters?

[03:47:00.0000] <othermaciej>
(I'm reading mollydotcom's twitter feed and it sounds defensive)

[03:52:00.0000] <Philip`>
hsivonen: I get text/html on that page - how come it was XML?

[03:53:00.0000] <asmodai>
So far I am impressed with IE9, there might be hope for MS on the standards front.

[03:53:01.0000] <lazni>
I also get a/h+x

[03:53:02.0000] <lazni>
*a/x+x

[03:54:00.0000] <lazni>
only in fx

[03:54:01.0000] <Philip`>
/me was looking in Firefox 3.6

[03:55:00.0000] <Philip`>
Do things like http://live.visitmix.com/search?term=cheese%01 break too?

[03:56:00.0000] <Philip`>
(The cheese is probably unnecessary)

[03:57:00.0000] <zcorpan>
Philip`: yes: value="cheese&#1;"

[03:58:00.0000] <lazni>
3.7a4pre

[03:58:01.0000] <lazni>
renders a bit, then ysod with no text

[03:58:02.0000] <Philip`>
I get text/html there too

[03:59:00.0000] <Philip`>
Hmm, now http://live.visitmix.com/ gives me a YSOD

[03:59:01.0000] <Philip`>
so it's not fixed yet

[03:59:02.0000] <Philip`>
when using some version of Minefield/3.7a2pre

[04:00:00.0000] <zcorpan>
i guess now when ie will support xhtml, we'll get to see a lot more YSODs

[04:00:01.0000] <Philip`>
Seems to be working reliably in my Firefox and failing reliably in my Minefield

[04:00:02.0000] <Philip`>
(where "working" means "using text/html")

[04:01:00.0000] <Philip`>
Someone should start a blog that collects YSODs on the sites of people who ought to know better

[04:05:00.0000] <jgraham>
WFM in Opera + Chrome FFM in Minefield

[04:06:00.0000] <lazni>
/me installs UA switcher

[04:12:00.0000] <Philip`>
Viewing the page in IE9, I get a script error dialog box

[04:13:00.0000] <Philip`>
It gets sent text/html

[04:16:00.0000] <Philip`>
(based on User-Agent sniffing)

[04:17:00.0000] <annevk>
whatwg cabal -- always breaking your interwebs

[04:18:00.0000] <asmodai>
hah

[04:23:00.0000] <asmodai>
annevk: Do you guys also fix intarwebs?

[04:25:00.0000] <annevk>
we try to fix part of the "foundation"

[04:27:00.0000] <Philip`>
But mostly we just point and laugh

[04:27:01.0000] <Philip`>
(or cry)

[04:28:00.0000] <annevk>
or shrug etc.

[04:30:00.0000] <asmodai>
Philip`: Sometimes I think cry is the more appropriate reaction.

[04:31:00.0000] <asmodai>
Then again, if I had to shed a tear for every abuse of the web standards I find I'd be a dried out husk by now.

[04:46:00.0000] <asmodai>
annevk: 10.10 vs 10.50 -- 10.50 had a completely overhauled js engine or was it incremental.

[04:46:01.0000] <gsnedders>
asmodai: Completely new

[04:47:00.0000] <asmodai>
Thankee

[04:48:00.0000] <asmodai>
someone @ tweakers asking about the perf numbers

[04:48:01.0000] <annevk>
asmodai, version numbers are a game marketing plays that nobody understands, not even those involved

[04:50:00.0000] <asmodai>
annevk: Hahaha, oh I know. Just wanted to enlighten somenoe.

[04:50:01.0000] <asmodai>
annevk: http://tweakers.net/nieuws/66227/microsoft-zet-previewversie-internet-explorer-9-online.html

[04:50:02.0000] <gsnedders>
/me pretends to be annevk and understand that

[04:51:00.0000] <asmodai>
When I commented there were no replies yet.

[04:52:00.0000] <annevk>
asmodai, ah, I guess the confusion in that thread is that our version numbers are like decimal numbers nowadays, confusingly different from what everyone else is doing

[04:52:01.0000] <annevk>
asmodai, though not everyone internally agrees they are like decimal numbers, making matters even more weird

[04:53:00.0000] <annevk>
fortunately we offer the latest version on opera.com and you don't have to pick one :)

[04:53:01.0000] <Dashiva>
Do like ATI and release Opera X1 next

[04:53:02.0000] <mpilgrim>
so as far as i can tell, julian seems determined to trap hixie into stating that there will always be a group maintaining html within the organization that has been trying to kill and/or replace html since the day it was founded

[04:53:03.0000] <mpilgrim>
which, when you say it out loud, does seem kind of silly

[04:54:00.0000] <Dashiva>
More like as long as anyone cares about HTML, there will be someone mainining HTML

[04:54:01.0000] <Dashiva>
And when nobody cares anymore, who cares? (Nobody)

[04:54:02.0000] <asmodai>
Anyone, anyone, Bueller?

[04:55:00.0000] <mpilgrim>
Maybe he's saying that, once hixie gets hit by a bus, the Serious Standards People promise to stop working on HTML. Again.

[04:55:01.0000] <Dashiva>
WHATWG is the prime example of this

[04:55:02.0000] <mpilgrim>
which is certainly not outside the realm of possibility

[04:55:03.0000] <Dashiva>
w3c dropped the ball completely, but people still cared about HTML, so WHATWG was formed to continue the work

[04:56:00.0000] <mpilgrim>
/me is pretty sure he knew that already :O

[04:56:01.0000] <mpilgrim>
anyway, kind of a weird argument to be having

[04:57:00.0000] <Dashiva>
You know, maybe some of our viewer don't? :)

[04:57:01.0000] <Dashiva>
*viewers

[04:57:02.0000] <mpilgrim>
but i suppose it's not nearly as weird as arguing over the normative reference to ASCII

[04:58:00.0000] <mpilgrim>
or any of the myriad of other trivial things that julian chooses to spend his time arguing about

[04:58:01.0000] <annevk>
clearly he has a lot of spare time

[04:58:02.0000] <jgraham>
But imagine the pain if a Serious Standard got the wrong normative reference to ASCII!

[04:58:03.0000] <asmodai>
Can we clone Hixie?

[04:59:00.0000] <Dashiva>
We could breed him

[04:59:01.0000] <jgraham>
Are you suggesting using Hixie like a stallion?

[04:59:02.0000] <Dashiva>
mpilgrim: I was quite surprised to see him arguing in favor of non-text non-ascii RFCs

[05:00:00.0000] <mpilgrim>
oh, that rfc format discussion? yeah, it's scary to think that, in some circles, julian is the voice of reason

[05:00:01.0000] <mpilgrim>
i see why he feels so frustrated in the HTMLWG

[05:00:02.0000] <asmodai>
jgraham: I'm sure he'd like being a studmuffin.

[05:01:00.0000] <mpilgrim>
UNICODE IS A CONSPIRACY TO BREAK OUR TOOLS AND STEAL OUR WOMEN

[05:01:01.0000] <Philip`>
Dashiva: Doesn't seem that surprising given that he's written a tool to generate HTML versions of RFCs

[05:02:00.0000] <mpilgrim>
AND MEN, AND NON-GENDER-SPECIFIC DOMESTIC PARTNERS

[05:02:01.0000] <asmodai>
annevk: oh btw, the opera mini 5 beta (latest on the android market  place) doesn't seem to grok any MathML. So I doubt the released version has rudimentary support as you said yesterday

[05:02:02.0000] <asmodai>
annevk: I'll be glad to be proven wrong though :D

[05:03:00.0000] <Philip`>
I think they should have stuck with hand-written faxed RFCs

[05:03:01.0000] <annevk>
asmodai, we implement it via some CSS rules; I guess those are not included on the mini server

[05:03:02.0000] <jgraham>
Do they still accept RFCs in that format?

[05:03:03.0000] <Philip`>
That's the only technology that is sure to last for centuries, long after all our ASCII text tools have died out

[05:04:00.0000] <Dashiva>
What if English dies out?

[05:04:01.0000] <annevk>
asmodai, did you try Opera 10.50?

[05:04:02.0000] <asmodai>
annevk: moment

[05:04:03.0000] <asmodai>
/me fires up 10.50

[05:05:00.0000] <Philip`>
Dashiva: Archeological linguists can rediscover it and then translate the RFCs

[05:05:01.0000] <asmodai>
I get something like: S p = C p cos &ApplyFunction; i 1 - d + d + W &ApplyFunction; i cos&ApplyFunction; s n

[05:05:02.0000] <mpilgrim>
just like anything else, the only reason the IETF will ever change is if someone stands up to them

[05:05:03.0000] <mpilgrim>
someone who has something they want

[05:05:04.0000] <asmodai>
Which is the raw data in between the matml tags

[05:05:05.0000] <mpilgrim>
like some emerging standard

[05:05:06.0000] <asmodai>
annevk: So doesn't look like it does.

[05:05:07.0000] <mpilgrim>
and says "no, fuck you and your 30-year-old formatting rules"

[05:05:08.0000] <annevk>
asmodai, got a link?

[05:05:09.0000] <annevk>
asmodai, chaals is interested

[05:06:00.0000] <annevk>
:)

[05:06:01.0000] <asmodai>
annevk: sure

[05:06:02.0000] <asmodai>
annevk: http://www.in-nomine.org/~asmodai/3d.html

[05:06:03.0000] <mpilgrim>
tim bray complained about this many times, because the IETF's ASCII-only formatting rules mandated misspelling the name of one of the atompub editors

[05:06:04.0000] <mpilgrim>
and that can't be the first time that's happened

[05:07:00.0000] <virtuelv>
mpilgrim: whose name?

[05:07:01.0000] <annevk>
asmodai, ah yeah, we don't do the entities, but the rest doesn't look that great either :/

[05:07:02.0000] <asmodai>
annevk: for all I know that should be html 5 compliant. hsivonen's validator only complains about the mathml elements, but that's because he's still working on adding that.

[05:08:00.0000] <mpilgrim>
Bill de hÓra

[05:08:01.0000] <virtuelv>
ah

[05:08:02.0000] <annevk>
asmodai, aah, we only enable it for XML

[05:08:03.0000] <asmodai>
annevk: Ach zo

[05:08:04.0000] <asmodai>
annevk: Would that include the XHTML + MathML profile?

[05:09:00.0000] <annevk>
asmodai, hopefully

[05:09:01.0000] <annevk>
as you can tell I have not done a whole lot with this myself :)

[05:09:02.0000] <virtuelv>
actually, also asbjørn ulsberg's name is misspelled in rfc 4287

[05:09:03.0000] <virtuelv>
as in the transliteration of his name is wrong

[05:10:00.0000] <virtuelv>
(the proper substitute for ø (&oslash;) is oe

[05:10:01.0000] <virtuelv>
)

[05:10:02.0000] <Dashiva>
But that's his own fault, apparently

[05:12:00.0000] <asmodai>
annevk: yeah, looks like XHTML + MathML gets a better treatment

[05:12:01.0000] <asmodai>
annevk: checked out my pre-conversion XHTML version: http://www.in-nomine.org/~asmodai/3d.xhtml

[05:13:00.0000] <asmodai>
/me needs to double check these old formulae for consistency.

[05:14:00.0000] <asmodai>
annevk: How well versed is chaals in MathML btw?

[05:16:00.0000] <annevk>
not sure

[05:18:00.0000] <annevk>
mpilgrim, btw, why the positive stance for H264? (it seems somewhat contradictory with the thing on Firefox icon licensing, though I realize it is not at all the same)

[05:25:00.0000] <zcorpan>
asmodai: it seems mathml entities don't work in o10.50 for some reason. it works in internal builds

[05:28:00.0000] <zcorpan>
asmodai: a workaround is to use straight characters or NCRs and remove the doctype (which will give better perf in gecko and opera)

[05:29:00.0000] <jgraham>
Although I guess that may be impractical if your editor doens't support it

[05:29:01.0000] <jgraham>
(it should, really)

[05:29:02.0000] <asmodai>
zcorpan: In my case I switched to firemath for creating the first versions of my formulae.

[05:30:00.0000] <asmodai>
zcorpan: also, that xhtml file is now reworked to html 5. It hardly has any doctype ;)

[05:30:01.0000] <zcorpan>
asmodai: yeah, i meant for the xhtml case

[05:37:00.0000] <asmodai>
zcorpan: I dropped xhtml in favour of html 5. The latter seems to be much wider implemented, even now. :)

[05:40:00.0000] <asmodai>
Someone posted http://underdark.nl/dump/ie9-debunk/ in response to the results @ http://samples.msdn.microsoft.com/ietestcenter/

[05:41:00.0000] <zcorpan>
asmodai: mathml in xhtml is implemented in 2 shipping browsers, mathml in html is implemented in 0 shipping browsers, and 1 non-shipping

[05:42:00.0000] <asmodai>
zcorpan: mathml in html 5 works pretty well in FF 3.6 (when you enable html5 in the config).

[05:42:01.0000] <zcorpan>
asmodai: i count that as non-shipping, but fair enough, 1 shipping browser if you flip a pref

[05:43:00.0000] <asmodai>
zcorpan: And since it's not public yet, I can limit the exposure. But point taken. But the xhtml route is also wrought with the typical xhtml problems/considered dangerous.

[05:43:01.0000] <asmodai>
zcorpan: So I guess, by the time it's really in shape to be shown publically html 5 ought to be a bit more in position.

[05:43:02.0000] <asmodai>
I am still wondering how to do a decent fallback though.

[05:46:00.0000] <zcorpan>
i guess you could detect lack of support and implement a renderer in javascript or something

[05:47:00.0000] <zcorpan>
or parse the html using the js port of the v.nu html parser

[05:49:00.0000] <asmodai>
zcorpan: Yeah, will think about it some more.

[05:49:01.0000] <asmodai>
I can so many areas in which MathML could help the research institutes here at uni.

[05:50:00.0000] <asmodai>
/me wonders if he can bribe Opera devs with Dutch cheese to get html 5 mathml working

[05:51:00.0000] <zcorpan>
asmodai: try it

[05:51:01.0000] <zcorpan>
/me likes Dutch cheese

[05:51:02.0000] <asmodai>
I'll even throw in some french as needed :)

[05:52:00.0000] <asmodai>
zcorpan: I wonder how easily mathml in html 5 could be made to work with userjs. I remember this whitewolf or what's his name guy who wrote a mathml userjs back in the day

[05:57:00.0000] <zcorpan>
asmodai: it would be easy, but you wouldn't want to apply it for all pages since it'd kill perf

[05:58:00.0000] <zcorpan>
asmodai: http://krijnhoetmer.nl/irc-logs/whatwg/20080814#l-242

[06:00:00.0000] <annevk>
http://underdark.nl/dump/ie9-debunk/ is incorrect on some things though

[06:00:01.0000] <annevk>
e.g. border and backgrounds is in CR, which means vendors are encouraged to support properties without prefix

[06:01:00.0000] <annevk>
but they don't have to

[06:01:01.0000] <Philip`>
The presentation of the test results seems intentionally misleading

[06:01:02.0000] <Philip`>
since it makes it look like IE supports all the standards perfectly

[06:01:03.0000] <annevk>
oh yeah, no doubt about that

[06:02:00.0000] <annevk>
extremely poor form

[06:02:01.0000] <annevk>
only publishing the tests they pass

[06:03:00.0000] <Philip`>
It's good to have tests and to know what browsers have bugs, but it's silly to compare colour-coded percentages when it's a tiny subset of tests

[06:03:01.0000] <zcorpan>
maybe they didn't make any more tests, but made the impl pass all their tests

[06:04:00.0000] <Philip`>
Compare http://codedread.com/svg-support.php

[06:04:01.0000] <Philip`>
where IE gets far less than anything else, on a much wider range of tests

[06:06:00.0000] <hsivonen>
more XMl fun: http://live.visitmix.com/MIX10/Sessions/KEY01 unescaped ampersand

[06:07:00.0000] <annevk>
that one loads fine in Opera

[06:07:01.0000] <zcorpan>
so opera and batik 1.7 have "A+"

[06:08:00.0000] <annevk>
oh, it's text/html

[06:08:01.0000] <hsivonen>
do you get an A+ for supporting the SVG 1.2 Tiny features that are Bad for the Web?

[06:09:00.0000] <hsivonen>
(XML Events and textArea)

[06:09:01.0000] <zcorpan>
dunno

[06:09:02.0000] <zcorpan>
i think we might have broken xml events while fixing the 'load' event recently

[06:10:00.0000] <zcorpan>
we should probably just remove xml events

[06:10:01.0000] <hsivonen>
zcorpan: how did you fix the load event?

[06:10:02.0000] <zcorpan>
hsivonen: we fired a bubbling event on document before

[06:11:00.0000] <zcorpan>
hsivonen: now we do per spec

[06:11:01.0000] <hsivonen>
I haven't implemeted teh SVG load event at all in text/html, because the SVG WG still hasn't responded to my feedback

[06:11:02.0000] <hsivonen>
zcorpan: sync or async?

[06:11:03.0000] <zcorpan>
hsivonen: how to test?

[06:12:00.0000] <hsivonen>
zcorpan: trying to think of a reliable way...

[06:13:00.0000] <jcranmer>
that graph is bleh

[06:13:01.0000] <jcranmer>
it looks like you could get at best ~70% for not supporting animation

[06:14:00.0000] <hsivonen>
unreliable would be <script>var loadRan =false;</script><svg onload="loadRan=true'></svg><script>alert(loadRan)</script>

[06:14:01.0000] <zcorpan>
the svg testsuite should probably have an order or magnitude more tests anyway to have the coverage we aim for for html5

[06:18:00.0000] <zcorpan>
hsivonen: i get 'false' for that in xhtml

[06:18:01.0000] <Philip`>
"The following picture indicates the results of running all 280 tests in many popular SVG implementations in the wild."

[06:18:02.0000] <Philip`>
How much does each test test?

[06:18:03.0000] <zcorpan>
hsivonen: but i don't know if we fire load on svg sync or async

[06:18:04.0000] <Philip`>
/me has half an order of magnitude more tests for canvas than that, though a lot of them are testing very similar or very boring things

[06:19:00.0000] <zcorpan>
the interesting number is not how many tests are in a testsuite, but how many bugs are found

[06:22:00.0000] <Philip`>
But that's a hard number to measure, and it's generally correlated to the number of tests and the extensiveness of each test, which are easier to measure and compare

[06:22:01.0000] <hsivonen>
zcorpan: false means it can't be sync, right? (assuming the onload attribute works)

[06:27:00.0000] <asmodai>
http://www.readwriteweb.com/archives/npr_and_wsj_building_ipad-only_websites.php

[06:27:01.0000] <asmodai>
The Wall Street Journal, a News Corp. property, is also building an iPad-only version of their site - well, actually just an iPad-only front page. Unlike NPR's iPad site, which will be 100% Flash-free, WSJ visitors who follow links deeper into the website will soon discover that not all its page have been converted. 

[06:32:00.0000] <zcorpan>
hsivonen: yeah, svg load seems to be async

[06:33:00.0000] <zcorpan>
hsivonen: i was talking about the normal window.onload though

[06:39:00.0000] <hsivonen>
zcorpan: ok. I hope the WG agrees with async. I don't want to make it sync.

[06:40:00.0000] <zcorpan>
hsivonen: maybe send an email to ed

[06:42:00.0000] <hsivonen>
zcorpan: I've already sent email to the WG and ed has seen the email

[06:43:00.0000] <gsnedders>
othermaciej: you awake?

[06:44:00.0000] <zcorpan>
hsivonen: ok

[06:45:00.0000] <zcorpan>
hsivonen: have you pointed out that opera fires async?

[06:48:00.0000] <hsivonen>
zcorpan: nope

[06:51:00.0000] <annevk>
can't we kill the silly svg load events?

[06:52:00.0000] <annevk>
(assuming that is what you're talking about)

[06:54:00.0000] <zcorpan>
http://canvex.lazyilluminati.com/misc/dom-viewer/x.html?%3Chtml%20xmlns%3D'http%3A%2F%2Fwww.w3.org%2F1999%2Fxhtml'%3E%3Cscript%3EloadRan%3Dfalse%3C%2Fscript%3E%3Csvg%20xmlns%3D'http%3A%2F%2Fwww.w3.org%2F2000%2Fsvg'%20onload%3D'loadRan%3Dtrue'%2F%3E%3Cscript%3Ealert(loadRan%3F'sync'%3A'async')%3C%2Fscript%3E%3C%2Fhtml%3E

[06:54:01.0000] <zcorpan>
seems sync in gecko and chrome

[07:51:00.0000] <annevk>
ok, I dumped the abstract style sheet concept from CSSOM

[07:51:01.0000] <annevk>
style sheets are CSS

[07:51:02.0000] <annevk>
as they said in one episode of Better off Ted, deal with it

[07:59:00.0000] <hsivonen>
ah. UA sniffing. a site has one code branch for "Firefox" and another for everything else

[07:59:01.0000] <hsivonen>
and the Firefox branch doesn't work in Gecko+HTML5 parser but the everything else branch works

[08:05:00.0000] <Dashiva>
The web UI for the laundromat here has branches for IE and Netscape

[08:05:01.0000] <Dashiva>
That is, it assumes everything that doesn't have appName Netscape is IE...

[08:10:00.0000] <zcorpan>
isn't that true these days?

[08:10:01.0000] <zcorpan>
or is opera still opera?

[08:10:02.0000] <zcorpan>
oh it is

[08:11:00.0000] <zcorpan>
/me thought it was changed

[08:21:00.0000] <jgraham>
/me is particularly annoyed by people who do if (window.attachEvent) {/*IE + Opera*/} else {/*Everyone else*/}

[08:22:00.0000] <annevk>
you could also be annoyed at Opera for implementing support for attachEvent

[08:23:00.0000] <lazni>
http://intertwingly.net/blog/2010/03/17/Non-Draconian-XHTML

[08:25:00.0000] <zcorpan>
the xml spec doesn't really require showing an error message

[08:25:01.0000] <Philip`>
You need to show an error message for debuggability

[08:25:02.0000] <Philip`>
because otherwise I guess nobody will even realise there's a problem

[08:25:03.0000] <annevk>
yeah, I guess you need some text after the error to see what happens

[08:26:00.0000] <Philip`>
If I look at the MIX site in IE with a faked UA string, then it just stops midway through a news summary with no indication why

[08:26:01.0000] <Philip`>
*IE9

[08:26:02.0000] <zcorpan>
i think experimental features should be user-opt-in (like webgl and html5 parser in firefox) instead of being ugly syntax

[08:27:00.0000] <Philip`>
WebGL is opt-in because it's insecure, not because it's experimental

[08:27:01.0000] <Philip`>
(It has a prefix on the context name because it's experimental)

[08:29:00.0000] <zcorpan>
still, making experimental features being user-opt-in solves the problem of people depending on it

[08:33:00.0000] <Dashiva>
Any sufficiently non-stable release branch does the same, though

[08:34:00.0000] <zcorpan>
yes, not shipping experimental features also works

[08:34:01.0000] <Philip`>
/me remembers problems where Facebook(?) depended on experimental features that had only shipped in Firefox betas, and broke when it changed to follow an updated spec

[08:34:02.0000] <Philip`>
so I guess Firefox betas are insufficiently non-stable

[08:35:00.0000] <Dashiva>
How about this: experimental features only work on intranets :)

[08:37:00.0000] <Philip`>
That'd make testing hard, since many people don't have intranets

[08:37:01.0000] <jgraham>
Maybe on experimental.example.org subdomains

[08:37:02.0000] <Philip`>
Better to do something like force the page's background colour to magenta if it uses experimental features

[08:37:03.0000] <Philip`>
That wouldn't prevent technical testing of the feature, but would discourage its use in production sites

[08:38:00.0000] <Dashiva>
Oh, oh

[08:38:01.0000] <jgraham>
It would if the feature was supposed to affect background colour

[08:38:02.0000] <Dashiva>
Make it play a song

[08:38:03.0000] <Dashiva>
Something by crash test dummies, maybe

[08:39:00.0000] <zcorpan>
make the browser crash after a few minutes

[08:40:00.0000] <Philip`>
jgraham: Those features can change the text colour instead

[08:40:01.0000] <asmodai>
rofl, on reddit: In honor of St. Patrick's Day, I will drink a beer for every upvote I get!

[08:40:02.0000] <asmodai>
He now has 1573 upvotes XD

[08:41:00.0000] <jgraham>
Dashiva: If you want audio, surely it should be a clip of Mr T. saying " I pity the fool who uses experimental features" played at random intervals

[08:41:01.0000] <Dashiva>
No, it has to be sufficiently bad to be discouraging

[08:41:02.0000] <Dashiva>
Mr. T. could easily be seen as a feature

[08:43:00.0000] <Philip`>
I can imagine a web developer may use a desktop computer with no speakers attached

[08:44:00.0000] <Philip`>
so they wouldn't discover Mr. T's contributions to their site until they'd launched it and received comments from bemused users

[08:44:01.0000] <jgraham>
Well then that would teach them to be the fool that used experimental features

[09:28:00.0000] <zcorpan>
hsivonen: i talked with ed and we concluded that it seems most reasonable to remove SVGLoad from most elements but keep it for <svg>, <image> etc, and to make it async

[09:31:00.0000] <annevk>
and for <svg> it is delayed like it is delayed for <body>?

[09:31:01.0000] <annevk>
at some point someone should work out how that interacts

[09:32:00.0000] <zcorpan>
apparently it is in opera currently, though i don't care if it's async when end tag is parsed or when everything has been parsed

[09:39:00.0000] <zcorpan>
hsivonen: it seems opera doesn't send SVGLoad on elements if there's no <svg> ancestor

[09:39:01.0000] <zcorpan>
hsivonen: another alternative is to kill SVGLoad and make <svg onload> work like <body onload> when it's the root element

[09:44:00.0000] <hsivonen>
zcorpan: killing svg onload would be my preference

[09:46:00.0000] <hsivonen>
if feasible considering compat

[09:49:00.0000] <zcorpan>
hsivonen: i'd prefer killing it as well

[09:49:01.0000] <annevk>
/me thirds

[09:52:00.0000] <boblet>
hey all, anyone know how Google Rich Snippets is detecting uF?

[09:53:00.0000] <boblet>
if they have code that supports uF and Microdata, that could be helpful for Tantek and other uF tool maintainers

[10:02:00.0000] <boblet>
will check logs if anyone can answer that later

[10:08:00.0000] <AryehGregor>
boblet, what do you mean "how"?  Presumably Google wrote some software to do it for them, which they haven't released, no?

[10:10:00.0000] <boblet>
AryehGregor: how as in a modified version of html5lib, a custom library they wrote, something they’re planning to release, something they’re not planning to release

[10:11:00.0000] <AryehGregor>
Why do you assume they made any information public about it?

[10:11:01.0000] <AryehGregor>
Their entire search engine and pretty much everything associated with it is closed-source and quite secret, as far as I've seen.

[10:12:00.0000] <boblet>
detecting uF and Microdata isn’t exactly secret sauce-level secrets

[10:12:01.0000] <boblet>
they have the rich snippet testing tool for testing after all

[10:13:00.0000] <boblet>
of course what they do with the data is, but I’m wondering if anyone here has heard about the detecting part

[10:15:00.0000] <Philip`>
I've heard nothing about the details of any of their rich snippets processing stuff

[10:16:00.0000] <Philip`>
so the only information I have comes from testing, or from making it up

[10:17:00.0000] <Philip`>
e.g. it's not using any kind of proper HTML parser (judging by some testing)

[10:17:01.0000] <boblet>
Philip`: aah interesting

[10:18:00.0000] <boblet>
that prolly means nothing to do with html5lib, and something in-house

[10:18:01.0000] <Philip`>
and in fact it is powered by snakes that use iPhones (I made that up)

[10:19:00.0000] <boblet>
snakes, eh

[10:19:01.0000] <boblet>
good to know

[10:20:00.0000] <Philip`>
/me wonders how well cold-blooded animals can use touch-sensitive screens

[10:20:01.0000] <jgraham>
/me imagines Samuel L. Jackson wandering around the Google server room muttering about motherfucking snakes on motherfucking iPhones

[10:21:00.0000] <jgraham>
Not that Samuel L. Jackson mutters you understand

[10:22:00.0000] <asmodai>
Given how Google uses Python

[10:22:01.0000] <asmodai>
...

[10:37:00.0000] <boblet>
asmodai: so it’s the *snakes* that are programming in Python! of course!! it all becomes clear :)

[10:43:00.0000] <asmodai>
boblet: :)

[10:43:01.0000] <boblet>
asmodai: you truly have deep insight into the workings of things. thank you for enlightening me ;-)

[11:21:00.0000] <mpilgrim>
annevk: where did you get the impression that i was cheerleading for h.264?

[11:27:00.0000] <mpilgrim>
i have to talk about it in diveintohtml5.org because it's (unfortunately) required to reach all platforms that people would reasonably care about

[11:27:01.0000] <mpilgrim>
i think i re-tweeted a mention of IE supporting it because it seemed like big news

[11:28:00.0000] <mpilgrim>
that doesn't mean i *like* the fact that web video is standardizing on a patent-encumbered format

[11:29:00.0000] <mpilgrim>
(de facto standardizing, not formally standardizing)

[11:29:01.0000] <mpilgrim>
then there's the whole "lesser of two evils" thing with naked-H.264-in-a-video-element vs. H.264-played-by-a-Flash-client

[11:30:00.0000] <erlehmann>
indeed. it suddenly becomes easier to scrape websites for content.

[11:30:01.0000] <mpilgrim>
seeing as how Flash ships with H.264 plus a bunch of other proprietary stuff (that is not only patent-encumbered but completely vendor-specific and unlicensable under any terms)

[11:30:02.0000] <erlehmann>
and no flash cookies etc.

[11:31:00.0000] <mpilgrim>
it seems like a huge win if we can convert all the video sites away from flash

[11:31:01.0000] <erlehmann>
hehe, wake me up when porn sites begin using <video>

[11:31:02.0000] <mpilgrim>
or even some of them, some of the time, for some percentage of users

[11:33:00.0000] <mpilgrim>
during the most recent html5-vs-flash flamewar, i read that many porn sites have mobile editions which offer pure-H.264-in-MP4 downloads for iPhone and other flash-free platforms

[11:33:01.0000] <mpilgrim>
dunno if that's true, but it certainly seems plausible

[11:34:00.0000] <mpilgrim>
they're probably doing H.264 encoding anyway to serve up through flash

[11:35:00.0000] <mpilgrim>
anyway, video has always been a mess, and it's unrealistic to expect the entire world to shift on a dime and suddenly be non-messy just because we defined a bunch of angle brackets

[11:39:00.0000] <crash\>
there was once a post on the Whatwg mailing list about the acces restriction of <video> for external ressources and that those restriction should had applied to <img>, <object> etc. when the Web was created

[11:39:01.0000] <crash\>
Has anybody got the link?

[11:40:00.0000] <AryehGregor>
Once?

[11:40:01.0000] <AryehGregor>
Pretty sure it's come up several times.

[11:42:00.0000] <TabAtkins>
I know I supported that assertion.  Let me look...

[11:43:00.0000] <crash\>
it was more written like an article :)

[11:45:00.0000] <TabAtkins>
Darn, my gmail-fu is weak.

[11:46:00.0000] <crash\>
http://lists.xiph.org/pipermail/theora/2008-November/001958.html that one :)

[12:40:00.0000] <hsivonen>
 I tought svg load was only on the svg element

[12:41:00.0000] <hsivonen>
thought

[15:04:00.0000] <MikeSmith>
gsnedders: you still around?

[15:25:00.0000] <jgraham>
MikeSmith: As of about 20 minutes ago he was out

[15:25:01.0000] <jgraham>
Erm

[15:26:00.0000] <jgraham>
Semantics fail

[15:26:01.0000] <MikeSmith>
ok

[15:26:02.0000] <jgraham>
I mean "20 minutes ago, when my information was last known to be correct, he was out"

[15:29:00.0000] <MikeSmith>
jgraham: understood, thanks

[15:30:00.0000] <jgraham>
mpilgrim: Arguably, for a new entrant to the browser market, H.264 is worse than Flash because the barrier to entry of millions of dollars a year for licenses is a greater impediment than the need to implement the Netscape plugin API and let adobe worry about licenses

[15:31:00.0000] <jgraham>
(for end users it is better as long as they don't mind browsers requiring large companies backing them and don't care about open source or small players)

[15:39:00.0000] <Dashiva>
jgraham: Decoder distribution licenses, you mean?

[15:41:00.0000] <jgraham>
Yes

[15:42:00.0000] <Dashiva>
That's pretty much an OS feature by now, though

[15:42:01.0000] <jgraham>
If you don't care about small players or new entrants

[15:43:00.0000] <jgraham>
And if the media backend on the OS has the APIs you need for the particular requirments of web video

[15:43:01.0000] <Dashiva>
That seems to be the case for all major OSes at the moment

[15:44:00.0000] <jgraham>
My understanding is that it is non-trivial to get good support in some cases

[15:45:00.0000] <jgraham>
using the native APIs

[15:45:01.0000] <jgraham>
(especially if you are not the OS vendor and can't change / fix the APIs / codecs as needed)

[15:46:00.0000] <jgraham>
(and for Linux it is only the case that you can play H.264 if you ignore the laws that prevent you doing so)

[15:47:00.0000] <jgraham>
+ other OSes that don't license the codecs correctly

[15:48:00.0000] <Dashiva>
This is all very vague and ominous

[15:56:00.0000] <jgraham>
Dashiva: Not really

[15:58:00.0000] <jgraham>
I guess "it's possible to just use the platform-supplied codecs and push the licensing concerns onto someone else" is a valid counter argument

[15:59:00.0000] <jgraham>
But there are real disadvantages

[16:02:00.0000] <Dashiva>
jgraham: sounds similar to "It's possible to just use the flash-supplied codecs and push the licensing concerns onto someone else"

[16:04:00.0000] <jgraham>
Dashiva: You probably have to implement NPAPI anyway and it is probably mostly shared between platforms whilst video would need N implementations to ship on N platforms

[16:04:01.0000] <Dashiva>
Oh, sure, it's more work. But it's just another feature to support, there's nothing inherently harder about it.

[16:31:00.0000] <Hixie>
i love feedback that just says "Section 1.2.3, step 4 should be changed to do X because Y" where X and Y assume you know what section 1.2.3 step 4 is.

[16:31:01.0000] <Dashiva>
/me tries to parse

[16:31:02.0000] <Dashiva>
Oh

[16:32:00.0000] <Hixie>
given that there are over a dozen specs each with different section numbers, several of which are the same text, and all of which change regularly, and given that i only look at feedback with a latency of a few months...

[16:32:01.0000] <AryehGregor>
Dashiva, Windows doesn't have H.264 codecs installed by default before Windows 7, as far as I've heard.

[16:34:00.0000] <AryehGregor>
And it's true that Linux only does on the theory of "downloading my distribution copy of ffmpeg-nonfree isn't pirating, honest".

[16:38:00.0000] <AryehGregor>
Okay, who has any idea what might be the issue here?  http://www.mediawiki.org/wiki/Special:Code/MediaWiki/63842

[16:38:01.0000] <AryehGregor>
I'm not sure what would cause Mac and Windows Firefox to treat px differently.  Different DPI settings or something?

[16:41:00.0000] <Dashiva>
Maybe they use different baselines to offset from

[16:52:00.0000] <Hixie>
othermaciej: would be good to have some webkit feedback on http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2010-March/025403.html

[16:52:01.0000] <Hixie>
annevk: opera feedback too ^

[16:53:00.0000] <othermaciej>
Hixie: it looks to me like that message is itself WebKit feedback

[16:55:00.0000] <Hixie>
true

[16:55:01.0000] <Hixie>
annevk! opera's feedback would be most welcome :-)

[16:58:00.0000] <roc>
AryehGregor: we've heard that Microsoft is going to ship an H.264 codec in Vista SP2

[16:59:00.0000] <roc>
a lot of XP users will upgrade to Windows 7 relatively soon, but there will probably be a very long tail of XP users too


2010-03-18
[17:00:00.0000] <roc>
AryehGregor: I suspect that person is probably just confused ... it's more likely font differences that are causing a problem

[17:05:00.0000] <daedb>
roc: Vista SP2 has been out for quite awhile, with no h.264 codec.

[17:05:01.0000] <roc>
ok, some future service pack then?

[17:05:02.0000] <daedb>
SP3 maybe :)

[17:05:03.0000] <roc>
they're going to need an H.264 codec for IE9, certainly

[17:08:00.0000] <JoePeck>
Reading up on <input type=range> at http://www.whatwg.org/specs/web-apps/current-work/multipage/number-state.html#range-state

[17:08:01.0000] <JoePeck>
[[ When the element is http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#suffering-from-an-underflow, the user agent must set the element's http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#concept-fe-value to a http://www.whatwg.org/specs/web-apps/current-work/multipage/common-microsyntaxes.html#valid-floating-point-number that represents 

[17:08:02.0000] <JoePeck>
http://www.whatwg.org/specs/web-apps/current-work/multipage/common-input-element-attributes.html#concept-input-min. ]]

[17:08:03.0000] <JoePeck>
eek, sorry!

[17:08:04.0000] <JoePeck>
[[ When the element is suffering from an underflow, the user agent must set the element's value to a valid floating point number that represents the minimum. ]]

[17:08:05.0000] <JoePeck>
I read that and I see, "if the element is suffering from an under, set its value so its no longer underflows"

[17:09:00.0000] <JoePeck>
Am I reading that correctly? Can a range input ever underflow?

[17:09:01.0000] <JoePeck>
/me goes to test what browsers do

[17:11:00.0000] <mpilgrim_>
jgraham: i have little sympathy for companies that get all faint-y about H.264 licensing fees but continue to ship browsers that support NPAPI

[17:12:00.0000] <mpilgrim_>
we have over 10 years of experience with NPAPI, and it's pretty clear that its only purpose is to serve as a backdoor for all manner of proprietary crap to sneak into browsers, shit all over the web, and then complain about the smell

[17:12:01.0000] <Hixie>
JoePeck: it can momentarily suffer from an underflow if you use script to set it to a value below the minimum

[17:12:02.0000] <Hixie>
JoePeck: but that sentence means the steady state can't suffer from an underflow

[17:13:00.0000] <Hixie>
JoePeck: if that makes sense

[17:13:01.0000] <JoePeck>
Hixie: okay, but correct behavior should then immediately set the value to be the minimum, and its overflow validity state would then be false?

[17:13:02.0000] <mpilgrim_>
if you really want to support an open web, drop support for NPAPI plugins

[17:13:03.0000] <mpilgrim_>
otherwise, STFU

[17:17:00.0000] <Philip`>
Seems a bit late for that to have any effect other than driving away users to other browsers

[17:17:01.0000] <Philip`>
whereas it's not necessarily too late yet to influence how people use <video>

[17:18:00.0000] <Hixie>
JoePeck: yeah

[17:18:01.0000] <JoePeck>
Hixie: ahh, I just changed my mind

[17:18:02.0000] <JoePeck>
Hixie: The "value sanitization algorithm" which modifies the value is only run [[ When an input element's type attribute changes state, and when the element is first created ]].

[17:19:00.0000] <JoePeck>
Hixie: so if the input element is already a range, and a script sets its value to an underflowing value, "what then"

[17:19:01.0000] <Hixie>
setting .value also invokes the value sanitization algorithm

[17:20:00.0000] <JoePeck>
Hixie: ahh I see thanks.

[17:20:01.0000] <JoePeck>
a lot of "ahh" moments, I'll need to read this more completely after I work out this bug

[17:20:02.0000] <JoePeck>
Hixie: thanks!

[17:20:03.0000] <Hixie>
my pelasure

[17:21:00.0000] <Hixie>
pleasure also

[17:21:01.0000] <Hixie>
i wish it was easier to read

[17:21:02.0000] <Hixie>
but i don't know how to make it easier to read

[17:21:03.0000] <Hixie>
fundamentally what we're speccing here is a mess

[17:21:04.0000] <Hixie>
so the spec kinda has to be a mess too

[17:21:05.0000] <JoePeck>
the links help, but in this case

[17:22:00.0000] <Hixie>
a key thing to know about the links is that when you reach the bold <dfn> of a term, you can click that to get a list of references to it

[17:22:01.0000] <JoePeck>
the "value sanitization algorithm" link only went to the first case, and not the .value case

[17:22:02.0000] <Hixie>
(only works in the one-page version though)

[17:22:03.0000] <JoePeck>
thanks again =)

[18:09:00.0000] <mpilgrim>
Philip`: yeah, people are really fleeing the iPhone in droves

[18:12:00.0000] <Dashiva>
Is this tweet a case for a standardized RT end tag? http://twitter.com/diveintomark/status/10645203889

[18:14:00.0000] <erlehmann_>
Dashiva, you have <q> and </q> for that already. stupid twitter not supporting markup (also, counting it towards the character limit gave rise to short url services that break the web)

[18:14:01.0000] <othermaciej>
most people put "via" at the end if they want a retweet with commentary, in this case it wouldn't have worked for the humor purpose of the post

[18:15:00.0000] <Dashiva>
I'm pretty sure tinyurl was around before twitter

[18:15:01.0000] <erlehmann_>
HTML 2.0 WTF

[18:16:00.0000] <erlehmann_>
isn't „gave rise“ something akin to „made it popular“ ?

[18:16:01.0000] <erlehmann_>
also, ars technicas site is XHTML 1.0 Transitional. where did they get that 2.0 thingy from ?

[18:17:00.0000] <Dashiva>
Their CMS is probably XHTML, but their web server error pages are not part of the CMS

[18:18:00.0000] <erlehmann_>
well, i guess they would fail on a 500 too in most cases if they were ;)

[18:21:00.0000] <sicking>
hum, does HTML not support hex numeric entities? Like &#x40; ?

[18:22:00.0000] <sicking>
never mind, i'm just doing it wrong

[18:22:01.0000] <mpilgrim>
OH NOES I'M USING TWITTER INCORRECTLY

[18:22:02.0000] <JoePeck>
Hixie: can I run a very simple example past you to see if you think its correct or not?

[18:25:00.0000] <Hixie>
sure

[18:25:01.0000] <JoePeck>
var x = document.createElement('input'); x.type='range'; // range input

[18:25:02.0000] <JoePeck>
x.max = 2; x.value = 3; // this causes overFlow, value gets changed to 

[18:25:03.0000] <JoePeck>
x.validationMessage; // shouldn't be any, because value is now stable

[18:26:00.0000] <Hixie>
sounds right, but i could be mistaken

[18:26:01.0000] <JoePeck>
okay. I feel like that removes the usefulness of element.validity.rangeOverflow and rangeUnderflow for type=range

[18:27:00.0000] <JoePeck>
but I do feel like this should be correct behavior

[18:30:00.0000] <Hixie>
well there is no usefulness of element.validity.rangeOverflow and rangeUnderflow for type=range, so i don't think removing it is a problem :-)

[18:31:00.0000] <JoePeck>
I was just going to try to get some implementor feedback, maybe send a question to the mailing list

[18:31:01.0000] <JoePeck>
I haven't checked the other input types affected rangeUnderflows/rangeOverflows

[18:31:02.0000] <Dashiva>
Sleep... Norwegian letters in the Safari 4 installer become chinese symbols

[18:31:03.0000] <Dashiva>
*Sleek

[18:32:00.0000] <sicking>
that's internationalization taken to new levels

[18:32:01.0000] <sicking>
"install safari, become a world traveler"

[18:33:00.0000] <Dashiva>
Actually, it might be Japanese

[18:34:00.0000] <Hixie>
JoePeck: many of the validity flags are only useful for particular controls, i wouldn't worry about that

[18:35:00.0000] <JoePeck>
Hixie: okay. Its just WebKit had a bunch of tests for overflow / underflow which, after reading the spec, didn't seem to make sense

[18:36:00.0000] <Hixie>
ah ok

[18:36:01.0000] <Hixie>
that's quite possible :-)

[18:53:00.0000] <roc>
mpilgrim: for a long time there was (and to some extent still is) nothing comparable to the iPhone

[18:53:01.0000] <roc>
the desktop browser space is a lot more competitive

[18:54:00.0000] <roc>
if we (Mozilla) thought we could ditch NPAPI support and remain competitive, we probably would

[18:54:01.0000] <roc>
it would be a lot less work, that's for sure

[18:54:02.0000] <Dashiva>
Really? Isn't the API pretty much static?

[18:55:00.0000] <roc>
it evolves

[18:55:01.0000] <roc>
but all browsers are doing lots of work to isolate plugins into their own processes etc

[18:57:00.0000] <roc>
mpilgrim: in fact, even today there is nothing really comparable to the iPhone that supports Flash yet

[18:58:00.0000] <Dashiva>
mpilgrim: I saw a claim on hackernews that youporn has a <video> mobile edition, but I can't verify ;)

[19:02:00.0000] <othermaciej>
Dashiva: once we have the porn sites its game over

[19:20:00.0000] <mpilgrim>
dashiva: that's probably where i read it too

[19:23:00.0000] <mpilgrim>
roc: i understand your present position, but outsourcing all the patent-encumbered stuff to a third-party plugin (via an API where third-party vendors can add unlimited amounts of new patent-encumbered stuff at any time) is hardly noble

[19:23:01.0000] <roc>
that is not our goal

[19:23:02.0000] <mpilgrim>
that is the reality of NPAPI

[19:23:03.0000] <roc>
that is not our goal for Web video

[19:23:04.0000] <roc>
our goal is to not depend on patent-encumbered stuff at all

[19:23:05.0000] <mpilgrim>
...except for all the stuff you've outsourced to adobe

[19:24:00.0000] <roc>
tthat is temporary

[19:24:01.0000] <mpilgrim>
everything is temporary if you live long enough

[19:24:02.0000] <mpilgrim>
i don't see "removing NPAPI" on any mozilla roadmap

[19:25:00.0000] <mpilgrim>
did i miss a memo?

[19:25:01.0000] <roc>
I don't believe we can remove NPAPI and remain relevant

[19:26:00.0000] <roc>
hopefully at some point in the future this will no longer be true

[19:26:01.0000] <mpilgrim>
that's a pretty loose definition of "temporary"

[19:26:02.0000] <roc>
yes it is

[19:26:03.0000] <roc>
I can't predict what will happen on the Web, sorry

[19:26:04.0000] <jcranmer>
/me points to the temporary Mork database in mozilla code

[19:26:05.0000] <roc>
it's gone

[19:26:06.0000] <roc>
anyway

[19:27:00.0000] <jcranmer>
about 10 years later

[19:27:01.0000] <mpilgrim>
well, i hope we all live long enough to see it

[19:27:02.0000] <roc>
so do I

[19:27:03.0000] <roc>
we'll need NPAPI support as long as there's a mass of Flash content on the Web that people want to use

[19:27:04.0000] <roc>
I can't control that

[19:28:00.0000] <othermaciej>
jcranmer: Mork?

[19:28:01.0000] <jcranmer>
I suspect the answer is there always will be

[19:28:02.0000] <roc>
we *can* try to rescue video on the Web from a pay-to-play patent-encumbered prison

[19:28:03.0000] <jcranmer>
othermaciej: something with which I have much more experience than I would like to have

[19:28:04.0000] <jcranmer>
think about all those little flash games

[19:28:05.0000] <mpilgrim>
not with just a client, you can't

[19:29:00.0000] <othermaciej>
if Wikipedia (or some other site) gets a critical mass of Theora-only content, that seems more likely to drive Ogg 

[19:29:01.0000] <roc>
I agree

[19:30:00.0000] <roc>
we might lose

[19:30:01.0000] <roc>
but mpilgrim said "I have little sympathy for companies that get all faint-y about H.264 licensing fees but continue to ship browsers that support NPAPI" and I have a problem with that

[19:31:00.0000] <othermaciej>
I can understand Mozilla's desire not to ship H.264 out of the box, but I am puzzled that a third party is allowed to plug it in to <object> but there is by design no mechanism for a third party to plug it into <video>

[19:31:01.0000] <mpilgrim>
i did say that, and i stand by it

[19:31:02.0000] <mpilgrim>
you refuse to defer to OS-level codecs

[19:31:03.0000] <mpilgrim>
so people are going to continue to use flash

[19:31:04.0000] <othermaciej>
that seems to help plugins at the expense of <video> more than it helps Ogg at the expense of H.264

[19:31:05.0000] <mpilgrim>
you don't even let firefox extensions add codec support

[19:32:00.0000] <roc>
if we had to do H.264 we probably would defer to OS codecs, which is why "H.264 licensing fees" is a total non-issue

[19:32:01.0000] <mpilgrim>
but you've stated publicly that you won't do that

[19:32:02.0000] <jcranmer>
I suspect FF would refuse to support external <object> handlers were that a tenable option

[19:32:03.0000] <mpilgrim>
(defer to OS codecs)

[19:33:00.0000] <mpilgrim>
yet you continue to allow NPAPI plugins, which you acknowledge contain multiple patent-encumbered codecs, including H.264 in particular

[19:33:01.0000] <mpilgrim>
so yeah, people are gonna keep using flash

[19:33:02.0000] <mpilgrim>
congratulations, that's what you've achieved

[19:33:03.0000] <mpilgrim>
i don't really care what your original goals were

[19:33:04.0000] <roc>
supporting H.264 is not going to make Flash go away

[19:34:00.0000] <roc>
it really isn't

[19:34:01.0000] <mpilgrim>
yes, i know, all those awesome flash games

[19:34:02.0000] <roc>
if dropping H.264 would let us drop support for NPAPI, that would be worth serious consideration

[19:34:03.0000] <roc>
sorry

[19:34:04.0000] <roc>
if *supporting* H.264...

[19:34:05.0000] <mpilgrim>
not sure why kongregate.com can't just release a standalone client for playing them anyway

[19:35:00.0000] <mpilgrim>
why does that have to happen in a browser?

[19:35:01.0000] <roc>
the other thing is the content producer side

[19:36:00.0000] <mpilgrim>
yes, i am aware i work for the single largest producer of H.264 web video

[19:36:01.0000] <deltab>
mpilgrim: there already is a standalone player, though not everything will work in it (due to the lack of surrounding page)

[19:37:00.0000] <mpilgrim>
deltab: my point is that "flash games" is a common retort for why browsers have to keep supporting flash, despite the fact that the entire flash game industry has consolidated around a few key players who may as well release their own client (and may already be working on that, for all i know)

[19:38:00.0000] <mpilgrim>
my friend's son programs flash games, makes a decent living at it (for a 15 year old)

[19:38:01.0000] <mpilgrim>
the browser is just a weird quirky delivery device to him

[19:38:02.0000] <roc>
it's not just Flash games

[19:38:03.0000] <roc>
in China tons of Web sites are Flash-based

[19:38:04.0000] <roc>
http://www.smashingmagazine.com/2010/03/15/showcase-of-web-design-in-china-from-imitation-to-innovation-and-user-centered-design/

[19:39:00.0000] <roc>
if you think it's so easy to drop Flash support, start at home with Chrome

[19:39:01.0000] <mpilgrim>
don't you think i've made that argument?

[19:41:00.0000] <mpilgrim>
not with chrome specifically (which has all the same "market relevance" problems on the windows PC desktop), but with chrome os in particular (which is more like an iPhone in being a specialized almost-but-not-quite general-purpose-computing device)

[19:42:00.0000] <roc>
I would also like to mention that support for plugging codecs into <video> is not omitted "by design"

[19:42:01.0000] <roc>
codec extensibility would be a ton of extra work

[19:42:02.0000] <roc>
we have simply not chosen to invest in that

[19:43:00.0000] <roc>
In Windows 7 Microsoft has effectively dropped support for codec extensibility because it's such a pain

[19:44:00.0000] <roc>
third-party codec extensibility, I mean

[20:57:00.0000] <mpilgrim>
hixie, you around?

[21:09:00.0000] <mpilgrim>
/me adds a disclaimer to the latest chapter in "dive into html5" to the effect that he is probably talking out of his ass, while he waits for hixie to return and explain the online whitelist wildcard flag to him

[21:32:00.0000] <Traveler>
Hi

[21:32:01.0000] <Traveler>
Should I use aside for a contact form or a div?

[21:32:02.0000] <MikeSmith>
Hixie: did you manage to get any good suggestions for short hidden='' examples? 

[21:34:00.0000] <MikeSmith>
Traveler: either would be fine probably

[21:34:01.0000] <Traveler>
I see..

[21:34:02.0000] <MikeSmith>
aside would not be inappropriate

[21:35:00.0000] <Traveler>
Great.

[21:36:00.0000] <Traveler>
What about nested asides?

[21:36:01.0000] <MikeSmith>
what is the use case?

[21:37:00.0000] <MikeSmith>
why would you want to put inside the aside

[21:37:01.0000] <Traveler>
Well I have two aside elements but I need to wrap them in something so I could float them.

[21:37:02.0000] <Traveler>
Div is better then?

[21:38:00.0000] <MikeSmith>
why not an aside with two child divs? 

[21:38:01.0000] <Traveler>
Hmm

[21:38:02.0000] <Traveler>
I did it the other way around

[21:39:00.0000] <Traveler>
Your idea sounds better though.

[23:33:00.0000] <MikeSmith>
does the iPhone have both an accelerometer and a compass?

[23:39:00.0000] <othermaciej>
yeds

[23:39:01.0000] <othermaciej>
*yes

[23:47:00.0000] <MikeSmith>
othermaciej: thanks

[23:47:01.0000] <MikeSmith>
I was looking at Andrei's Orientation Event spec

[23:48:00.0000] <MikeSmith>
http://dev.w3.org/geo/api/spec-source-orientation.html

[23:49:00.0000] <MikeSmith>
I assume the OrientationEvent in that is meant to rely on a compass sensor

[23:49:01.0000] <MikeSmith>
while the AccelerometerEvent event uses the accelerometer

[23:50:00.0000] <othermaciej>
I have no idea

[23:50:01.0000] <othermaciej>
I wish OrientationEvent didn't clash with the iPhone's "orientationchange" event

[23:50:02.0000] <othermaciej>
which btw we'd like to standardize

[23:51:00.0000] <MikeSmith>
yeah, the name of that in Andrei's spec should maybe be changed

[23:51:01.0000] <othermaciej>
is his event intended to be absolute orientation or relative orientation?

[23:51:02.0000] <MikeSmith>
relative

[23:51:03.0000] <MikeSmith>
I believe

[23:51:04.0000] <othermaciej>
can a digital compass even tell you your pitch and yaw?

[23:52:00.0000] <othermaciej>
I thought that data came from the accelerometer

[23:52:01.0000] <MikeSmith>
yeah, I guess I'm not clear about what the relationship is

[23:53:00.0000] <MikeSmith>
or why the spec has two interfaces instead of just one

[23:53:01.0000] <othermaciej>
I'm also not clear why this spec is being done in the geolocation wg

[23:53:02.0000] <othermaciej>
yeah I'm not clear on why there are two separate events instead of one

[23:55:00.0000] <MikeSmith>
as far as why it's in the geolocation wg, I guess it may not be the ideal place, but I'm not sure it'd be worthwhile to bring it up -- I think Andrei and the chairs of that group did a pretty good job of keeping the geolocation API in scope and getting it published and seeing it get implemented

[23:57:00.0000] <othermaciej>
I have no quarrel with them, just not clear to me how it fits in their charter more than Web Apps WG or DAP

[23:58:00.0000] <othermaciej>
I guess I should count myself lucky it did not end up in DAP

[23:58:01.0000] <MikeSmith>
no comment

[00:12:00.0000] <MikeSmith>
/me wonders what kind of response dude expects to his "Why WebGL? Why now and not 10 years ago?" message on the webgl list

[00:18:00.0000] <othermaciej>
that's a funny question

[00:32:00.0000] <boblet>
jeez. promo in blog of leading Japanese web events group for event on HTML5 + CSS3 (logo, title, speaker images and names, date and venue) … as a jpg

[00:34:00.0000] <boblet>
don’t know whether to take the piss on Twitter, take the time to show them how to use, y’know, HTML5 & CSS3, or just go outside and scream

[00:34:01.0000] <boblet>
not even any alt text ffs

[00:34:02.0000] <MikeSmith>
CSSNite?

[00:34:03.0000] <boblet>
MikeSmith: :X

[00:34:04.0000] <boblet>
heh

[00:35:00.0000] <boblet>
that was meant to be the “I’m unable to talk” emoticon

[00:35:01.0000] <boblet>
seeing as I no doubt know all the organisers… gah

[00:36:00.0000] <MikeSmith>
Takano-san is a great guy, but I guess he's not as tuned into standards-based Web development as some others are

[00:37:00.0000] <boblet>
nope. He’s tuned into Adobe-based web development :|

[00:37:01.0000] <MikeSmith>
yeah, CSSNite is sometimes more like DreamWeaverNite

[00:37:02.0000] <MikeSmith>
but a lot of the people who attend seem to like that about it

[00:37:03.0000] <boblet>
lolsigh. guess what’s on in Kyoto tonight

[00:38:00.0000] <MikeSmith>
having a common set of tools/authoring environment as a baseline keeps the how-to discussion pretty tightly focused

[00:38:01.0000] <MikeSmith>
boblet: what's on?

[00:39:00.0000] <boblet>
“DreamWeaver Town Meeting in Kyoto” :|

[00:39:01.0000] <MikeSmith>
good for them

[00:39:02.0000] <boblet>
oh hold on, 21st April

[00:39:03.0000] <boblet>
saw it in the Twitters today, guess they announced it

[00:40:00.0000] <MikeSmith>
certainly sounds like a lot better place to hook up than a Vim Town Meeting or an Emacs Town Meeting

[00:40:01.0000] <boblet>
heh

[00:41:00.0000] <MikeSmith>
I hope the high interest in HTML5 among Japanese developers will result in demand to have some really good integration for HTML5 features in Dreamweaver

[00:44:00.0000] <MikeSmith>
looking at reports about Google's TV venture with Sony and Intel reminds me that it seem like broadcast TV and video-on-demand industry has by far the most to gain financially from having royalty-free video codecs

[00:44:01.0000] <MikeSmith>
for IPTV

[00:45:00.0000] <othermaciej>
MikeSmith: I think there's room in the market for a tool to make really sweet rich interactive sites using HTML5+CSS3, but I am not sure Dreamweaver is a good starting point for it

[00:45:01.0000] <MikeSmith>
othermaciej: not sure what other contenders there are

[00:45:02.0000] <MikeSmith>
there is definitely a real business opportunity there

[00:46:00.0000] <boblet>
sooner or later someone is gonna do it. probably when browsers y’know actually implement specs and stuff

[00:47:00.0000] <MikeSmith>
anybody who wants to get market share and developer mindshare away from Dreamweaver has a big challenge on their hands

[00:48:00.0000] <MikeSmith>
the figures I have heard on Dreamweaver in Japan at least are that 90%+ of working Web designers in Japan use Dreamweaver as their main authoring tool

[00:49:00.0000] <boblet>
sad ;(

[00:49:01.0000] <MikeSmith>
why sad?

[00:49:02.0000] <MikeSmith>
from what I've seen the output from it is pretty good

[00:49:03.0000] <othermaciej>
MikeSmith: I think the best opportunity is to create a new tool focused on creating the UI for front end of an interactive AJAX-based web app, that integrates with one or more JS libraries and server-side frameworks

[00:49:04.0000] <boblet>
because it’s suffering Adobe-itis

[00:50:00.0000] <othermaciej>
but maybe I am just unaware of Dreamweaver's goodness

[00:50:01.0000] <MikeSmith>
othermaciej: that indeed would be something

[00:50:02.0000] <boblet>
where Adobe builds a great product, then tries to kill it little by little

[00:50:03.0000] <othermaciej>
I would love to see a tool like the Cappuccino UI builder, but that generates markup as the output and integrates with jQuery or Dojo or something

[00:50:04.0000] <MikeSmith>
well, Adobe didn't originally build Dreamweaver

[00:50:05.0000] <MikeSmith>
Macromedia did

[00:51:00.0000] <MikeSmith>
boblet: but I get your point

[00:51:01.0000] <boblet>
I guess it’s better than that bs HomeSite Builder program

[00:51:02.0000] <othermaciej>
I would probably want to work on such a thing if I didn't love my current job

[00:51:03.0000] <boblet>
which is also quite popular for corporates

[00:51:04.0000] <boblet>
(in Japan)

[00:51:05.0000] <MikeSmith>
yeah, Cappuccino is inspiring

[00:53:00.0000] <MikeSmith>
boblet: I worked as a tech writer at one time, and did all my work in FrameMaker, which was a really great product already at the time when Adobe acquired the company that made it

[00:53:01.0000] <MikeSmith>
but after the acquired it, the made very few real improvements to it

[00:55:00.0000] <boblet>
MikeSmith: I *loved* Freehand back in the day — used it to do a crazy 100 page booklet as full printer’s spreads when I was in University, doing things that would be near impossible in Illustrator and damn hard in InDesign

[00:55:01.0000] <boblet>
yeah Framemaker was almost bought to prevent competition with the inferior PageMaker huh

[00:55:02.0000] <MikeSmith>
roc, doublec - have either of you guys ever met Rob Glidden? I'm wondering what he's up to these days. He still seems to be blogging now and then

[00:57:00.0000] <roc>
never heard of him

[00:57:01.0000] <roc>
I wrote my PhD thesis in Framemaker

[00:57:02.0000] <roc>
it was pretty cool

[00:58:00.0000] <MikeSmith>
yeah, there are features in Frame that I have still never seen in other products

[00:58:01.0000] <MikeSmith>
like the mechanism that it provides for creating and managing cross-references

[00:58:02.0000] <MikeSmith>
and the one that it has for creating indexes

[00:59:00.0000] <roc>
one way to attack the tools market would be to make the tools Web-based

[00:59:01.0000] <MikeSmith>
that would certainly fit with othermaciej Cappuccino idea

[01:00:00.0000] <MikeSmith>
roc: Rob Glidden is http://www.robglidden.com/2009/12/royalty-free-codec-standards-dont-settle-for-less/

[01:00:01.0000] <MikeSmith>
http://www.robglidden.com/

[01:00:02.0000] <roc>
Microsoft's Expression stuff seems good but I'm told they've had a really hard time making inroads with designers who are all hooked on Adobe

[01:00:03.0000] <othermaciej>
Making a web-based tool is an interesting idea

[01:00:04.0000] <roc>
if Microsoft can't make progress there, it'd be even harder for players without Microsoft's tools leverage

[01:01:00.0000] <MikeSmith>
roc: yeah, I would expect that's going to be the case with anybody else trying to get a piece of that market

[01:01:01.0000] <MikeSmith>
roc: Rob used to work at Sun and was involved, I think, with the Sun OMS Video work

[01:01:02.0000] <roc>
Microsoft's tools integrate with Visual Studio, ASP.NET, IIS, etc ... you'd think that'd be useful for someone :-)

[01:04:00.0000] <roc>
that guy looks interesting

[01:04:01.0000] <roc>
unfortunately I believe OMS is dead

[01:04:02.0000] <MikeSmith>
roc: yeah, have not heard anything about it for a long time now

[01:07:00.0000] <MikeSmith>
"Bob Lord is now Manager, Network Security and Infrastructure, at Twitter"

[01:09:00.0000] <MikeSmith>
"Christian Sejersen has an updated current title: Acting Head of Web Runtime at Nokia"

[01:09:01.0000] <MikeSmith>
I guess I should read my LinkedIn updates more often

[01:21:00.0000] <asmodai>
Any of you have any idea how I can make Firefox stop asking a proxy user/password combination? Even opted to save it, which works, but it still pops up every now and then.

[01:23:00.0000] <asmodai>
https://support.mozilla.com/en-US/forum/1/596899

[01:23:01.0000] <asmodai>
Guess that's relevant

[01:24:00.0000] <Philip`>
You could reconfigure the proxy to not require authentication

[01:27:00.0000] <asmodai>
Philip`: Fat chance, not in my power.

[01:27:01.0000] <asmodai>
That toggle did work though.

[01:28:00.0000] <asmodai>
Now I don't have to hit that password entry box all the timne

[01:46:00.0000] <asmodai>
Christ, that still doesn't fix it. Tried some other settings and it is still not giving me a rest.

[02:04:00.0000] <annevk>
/me had to think way too long about "Sure" (see WHATWG list)

[02:05:00.0000] <annevk>
/me still isn't sure it was the right answer

[02:06:00.0000] <asmodai>
annevk: Maybe maybe was better ;)

[02:16:00.0000] <annevk>
"sure" seemed better for process and given the data the most sane option

[02:16:01.0000] <annevk>
but that doesn't mean I like it :)

[02:19:00.0000] <annevk>
aah crap

[02:19:01.0000] <annevk>
GoW3 does not come before April 2011

[02:19:02.0000] <asmodai>
/me pats annevk 

[02:19:03.0000] <asmodai>
I can let you play some SC2 on my computer though ;)

[03:39:00.0000] <gsnedders>
MikeSmith: I am around now

[03:40:00.0000] <MikeSmith>
I can't remember what I wanted to ask you about now.. :(

[03:45:00.0000] <gsnedders>
MikeSmith: Forgetful old man.

[03:54:00.0000] <hendry>
we have some spam on the wiki http://wiki.whatwg.org/wiki/MBT_Shoes

[03:55:00.0000] <Philip`>
Who buys shoes over the internet anyway?

[03:55:01.0000] <Philip`>
/me can't imagine not going to a shop and trying them on first

[03:58:00.0000] <jgraham>
Philip`: Apparently it is rather coomon. YOu just order a whole bunch and send back the ones that don't fit

[03:58:01.0000] <jgraham>
*common

[03:58:02.0000] <jgraham>
It seems kinda silly to me

[04:04:00.0000] <hendry>
jgraham: some people don't have the time or courage to go to a shopping mall

[04:04:01.0000] <hendry>
added a note about vim+beautifier at http://wiki.whatwg.org/wiki/IDE

[04:06:00.0000] <jgraham>
hendry: It seems like sending back all the shoes that dodn't fit would be at least as time consuming as going to a physical shop, especially given that people typically have to visit physical shops occasionally anyway

[04:07:00.0000] <jgraham>
As for "courage"... How much courage does it take to buy shoes? Even I can manage that and I can be shy to a fault

[04:08:00.0000] <hendry>
jgraham: i take delivery at work and give stuff to reception for returns, so i don't need to go to a post office

[04:08:01.0000] <hendry>
jgraham: i find shopping malls host to traumatic experiences

[04:56:00.0000] <Dashiva>
http://cowbirdsinlove.com/comics/107/strawman.png

[04:56:01.0000] <Dashiva>
Not related to anything in particular

[05:03:00.0000] <jgraham>
Is that how these debating club/competition things work? I guess that explains why people love to argue by naming fallacies that the other party is supposedly making

[05:42:00.0000] <lazni>
http://www.iphone.vinaphone.com.vn/home/ In minefield 3.7a4pre, after the loading bar filled up, I can see the menu but the main content is blankly white

[05:42:01.0000] <lazni>
I mean the rectangle between the menu and "Tin tức"

[05:45:00.0000] <lazni>
html5 parser not enabled

[05:49:00.0000] <MikeSmith>
lazni: what happens if you do html5.enabled

[05:50:00.0000] <lazni>
I don't need to restart right? reloading gives the same result

[05:52:00.0000] <MikeSmith>
I don't think you should need to restart

[05:53:00.0000] <MikeSmith>
/me tries that page in Safari

[05:53:01.0000] <lazni>
nvm, it works correctly in safemode

[05:53:02.0000] <boblet>
seems like people are reading the logs these days. krijnhoetmer.nl is getting really unresponsive

[05:53:03.0000] <lazni>
a possible cause is ad block plus

[05:54:00.0000] <MikeSmith>
lazni: hmm, yeah, that might do it

[05:54:01.0000] <MikeSmith>
boblet: it's been that way for a while.. krijnh runs it from a Web server installed on his sewing machine

[05:55:00.0000] <boblet>
I know, but it seems to getting worse

[05:55:01.0000] <boblet>
Maybe he migrated it to the overlocker

[05:56:00.0000] <boblet>
try three to load the full day

[05:56:01.0000] <lazni>
I can see in the ABP blocked items list that a filter is triggered there

[05:56:02.0000] <boblet>
…made it 7 lines in ;-(

[05:57:00.0000] <MikeSmith>
boblet: there do seem to be one or more people with too much time on their hands, who go through the logs and highlight all kinds of pointless and uninteresting stuff

[05:57:01.0000] <boblet>
“pointless and uninteresting”!? but the highlights are where I get all my info from! :O

[05:58:00.0000] <boblet>
or are you just commenting on the mustache conversation we were having while you were at the sento? :P

[05:58:01.0000] <Dashiva>
I move for a reexamination of the proposal of multiple types of 'interesting' flagging

[05:59:00.0000] <boblet>
Dashiva: interesting facial hair conversations, interesting tangentially related links

[06:00:00.0000] <Dashiva>
More along the lines of 'interesting', 'informative', 'funny', 'conspiratory', 'controversial', etc

[06:08:00.0000] <annevk>
ooh look, more HTML5 fan sites: http://www.html5video.org/

[06:09:00.0000] <annevk>
mpilgrim, your explanation on how you view NPAPI vs H264 makes it clear, thanks

[06:09:01.0000] <annevk>
mpilgrim, while I'm not quite in agreement, I see your point

[06:09:02.0000] <hsivonen>
lazni: works for me with html5.enable=true and ABP

[06:10:00.0000] <lazni>
hsivonen: I subscribe to the Vietnamese list+Easy list filters

[06:10:01.0000] <hsivonen>
mpilgrim: so are Chrome and the Android browser getting rid of NPAPI? to me it seems NaCl&Pepper is even extending it

[06:11:00.0000] <lazni>
the rule triggered is something about id=banner

[06:11:01.0000] <hsivonen>
lazni: I see

[06:11:02.0000] <lazni>
VN list + Easy list is no longer on the ABP rule page I think

[06:11:03.0000] <hsivonen>
mpilgrim: as I understand it, NaCl&Pepper allows proprietary codecs to be downloaded dynamically, but the permanently-installed browser can be Free Software

[06:12:00.0000] <hsivonen>
(cue Stallman and non-Free JavaScript programs)

[06:12:01.0000] <annevk>
what is Pepper?

[06:12:02.0000] <annevk>
is that the bytecode part of NaCl?

[06:12:03.0000] <hsivonen>
annevk: Pepper is a cross-platform NPAPI layer

[06:13:00.0000] <hsivonen>
annevk: instead of an operating system-specific drawing surface, etc.

[06:14:00.0000] <hsivonen>
annevk: https://wiki.mozilla.org/Plugins:PlatformIndependentNPAPI

[06:15:00.0000] <hsivonen>
unless I've completely missed the point, Pepper is about delivering NaCl-based NPAPI plug-ins

[06:17:00.0000] <lazni>
/me writes a blog as an atom feed, and use client-side XSLT for pretty html 

[06:17:01.0000] <hsivonen>
as for actually discontinuing support for NPAPI, having a common codec for HTML5 video and having swfdec integrated into browsers natively (without Flash video support) would go a long way

[06:18:00.0000] <lazni>
fx and opera insist on applying their feed subscription interface, regardless of the mimetype

[06:18:01.0000] <hsivonen>
I was able to browse cluelessly authored restaurant menus just fine with swfdec

[06:18:02.0000] <Dashiva>
"valid, extended HTML"

[06:18:03.0000] <hsivonen>
but YouTube blocked swfdec...

[06:18:04.0000] <Dashiva>
What is valid supposed to mean again?

[06:18:05.0000] <lazni>
where can I find more about this sniffing?

[06:19:00.0000] <lazni>
return code of v.nu is 0?

[06:19:01.0000] <hsivonen>
Dashiva: badge and proprietary extensions at the same time

[06:19:02.0000] <hsivonen>
lazni: return code?

[06:20:00.0000] <lazni>
ISTR v.nu has a gnu error format output, which may have something equivalent

[06:20:01.0000] <lazni>
but then, you're the author

[06:20:02.0000] <Dashiva>
Having non-standardized extensions be valid seems sort of like allowing a contradiction in your logic system

[06:20:03.0000] <Dashiva>
Suddenly everything is valid

[06:20:04.0000] <lazni>
isn't that the purpose of data-?

[06:21:00.0000] <Dashiva>
data is just data, not behavior

[06:21:01.0000] <hsivonen>
ah. with the GNU format, the return value is an empty string when the content is valid

[06:21:02.0000] <Philip`>
Dashiva: It means intentional things are valid and errors are still invalid

[06:22:00.0000] <hsivonen>
Dashiva: it's a contradiction if you think validation in terms of de jure conformance. It may make sense if you think validation as spell check with a user dictionary.

[06:22:01.0000] <Philip`>
rather than overloading "invalid" to mean both "errors" and "features that the standardisation group have not incorporated"

[06:23:00.0000] <MikeSmith>
kudos to you for using client-side XSLT.. some developers appear to be religious objections against it

[06:23:01.0000] <Dashiva>
hsivonen: So I'll extend my dictionary with Japanese and release a book written in "extended English"

[06:23:02.0000] <jgraham>
mpilgrim: I was surprised you gave such a poor argument that browser vendors should put up and shut up about H.264. An argument of "If you do unpleasant thing X you lose all right to object to being required to do unplesant thing Y" is a total non-sequiter

[06:23:03.0000] <hsivonen>
I don't like using FBML as an example in DE discussions when (AFAICT) Facebook reps haven't asked for spec changes and some DE proposals would break Facebook in Firefox if implemented

[06:24:00.0000] <lazni>
ISTR the starcraft2.com website use client-side XSLT

[06:24:01.0000] <hsivonen>
Dashiva: for example. Though it's more conventional to extend English with French or Spanish

[06:24:02.0000] <Dashiva>
But that's no longer English

[06:25:00.0000] <MikeSmith>
boblet: sorry to have missed out on the mustache discussion.. 

[06:25:01.0000] <hsivonen>
Dashiva: is http://www.flickr.com/photos/hsivonen/190450883/ English?

[06:25:02.0000] <hsivonen>
alt="Home of the Kosher Style Burrito"

[06:25:03.0000] <boblet>
MikeSmith: don’t worry, you too can reload the relevant log page 4 times to read it

[06:26:00.0000] <Dashiva>
hsivonen: Webster says yes

[06:26:01.0000] <hsivonen>
Dashiva: I guess the extensions have been adopted to the spec already.

[06:27:00.0000] <Dashiva>
But that's a red herring, since it's only not soon-to-be-formalized extensions

[06:27:01.0000] <Dashiva>
There's also the "I want to make up stuff without anyone telling me it's stupid, and it should still be valid"

[06:27:02.0000] <Philip`>
boblet: Feel free to set up your own version of the logs page :-)

[06:27:03.0000] <Philip`>
/me can probably supply raw log files if needed

[06:28:00.0000] <hsivonen>
Dashiva: yes, there's the use case of BigCos selling stuff to governments and not wanting to be told their stuff is non-standard

[06:28:01.0000] <boblet>
Philip`: c’mon! the peanut gallery is for ppl to moan about free stuff they couldn’t possibly recreate themselves

[06:28:02.0000] <Dashiva>
And that's introducing the contradiction

[06:28:03.0000] <hsivonen>
Dashiva: indeed

[06:28:04.0000] <Philip`>
It shouldn't be hard to recreate a version without all the features

[06:28:05.0000] <Dashiva>
Philip`: sounds useful :P

[06:28:06.0000] <Philip`>
and the features are presumably what make it slow

[06:29:00.0000] <hsivonen>
Dashiva: good luck getting anyone whose actual use case that is to admit it

[06:29:01.0000] <lazni>
yeah, http://starcraft2.com/faq.xml is xhtml, then modified by XSLT for some unfathomable purpose

[06:29:02.0000] <boblet>
Philip`: I thought the running it on a sewing machine part was what made it slow ;-)

[06:29:03.0000] <Philip`>
I guess most people would be happy with a static slightly-formatted HTML version of the logs split by day and updated every few minutes, and nothing fancier

[06:29:04.0000] <Dashiva>
hsivonen: Indeed, no hopes there. But we _can_ make them try to state an alternative acceptable use case

[06:30:00.0000] <lazni>
but the XSLT wouldn't be used because the page is text/html

[06:30:01.0000] <Dashiva>
Philip`: I suppose the personalization (nick color etc) makes it impossible to cache

[06:30:02.0000] <boblet>
Dashiva: surely that’s JS-able, no?

[06:31:00.0000] <Dashiva>
Sure

[06:31:01.0000] <Dashiva>
If you don't mind being a script-requiring fascist

[06:31:02.0000] <lazni>
/me sees the effect of opera turbo clearly on starcraft2.com, all pixelly

[06:32:00.0000] <boblet>
heh

[06:32:01.0000] <Dashiva>
I could make a log page for fun, but I wouldn't want to have to support it

[06:32:02.0000] <Dashiva>
So making one would just make one party sad

[06:33:00.0000] <Philip`>
Could the personalisation be done with just CSS, in case people are silly and object to JS?

[06:34:00.0000] <Dashiva>
I suppose you could add class=nickname to every <span>nickname</span>, but that would bloat the page a lot for little (or no) gain

[06:35:00.0000] <Philip`>
The gain would be cachability

[06:35:01.0000] <Philip`>
and the avoidance of extreme slowness

[06:37:00.0000] <Philip`>
although doing it in JS would have those benefits and would also be more efficient and more powerful and easier

[06:37:01.0000] <Philip`>
Things like the current logs' index page with personalised line counts couldn't be done client-side, though

[06:38:00.0000] <Philip`>
and I'd expect that's pretty hard on performance

[06:38:01.0000] <Dashiva>
I don't think it's a log viewer's job to tell you how popular you are :)

[06:39:00.0000] <Philip`>
I guess it's useful to see if somebody addressed queries to you while you were offline

[06:39:01.0000] <Philip`>
(if you were offline for the entire day)

[06:39:02.0000] <Philip`>
(and if you're not just going to read all the logs anyway)

[06:39:03.0000] <boblet>
surely the server just needs to be upgraded from a sewing machine to say a networked TV

[06:40:00.0000] <boblet>
or maybe one of those really fancy high-end clock radios

[06:40:01.0000] <Dashiva>
If you got the swine flu vaccine, it could run on the microchips in your bloodstream

[06:40:02.0000] <boblet>
mmm bacon and chips

[06:41:00.0000] <lazni>
ugh, firegestures doén't work with 3.7a4pre

[07:02:00.0000] <Dashiva>
hsivonen: What do you mean when you called Stroustrup's attitude Hixie-ish, by the way?

[07:08:00.0000] <no_mind>
is canvas element a form of image ?

[07:09:00.0000] <no_mind>
type of image*

[07:09:01.0000] <Philip`>
It has no relation to the <img> element, but it can be thought of as a kind of programmable dynamic <img>

[07:25:00.0000] <hsivonen>
Dashiva: backwards compat, trying to push little new things to make things better in the future nonetheless, focus on the stuff people actually use to do stuff even if it looks ugly to the purists

[07:25:01.0000] <hsivonen>
Dashiva: also avoidance of vendor lock-in 

[07:46:00.0000] <no_mind>
Philip`, then why does firefox gives you a "View image" and "Save image" options on right clicking a canvas element ? This will make user think that canvas element is same as image

[09:04:00.0000] <AryehGregor>
Philip`, I recently bought new boots online and regretted it.  Had to return two pairs before getting the right size, and still not quite sure they're the right size.  I buy shoes online if they're the exact same type of shoe that I already have but my current pair has just worn on.

[09:06:00.0000] <jgraham>
AryehGregor: What type of boots?

[09:06:01.0000] <AryehGregor>
I dunno, what types are there?

[09:08:00.0000] <jgraham>
Well I mean like walking boots or ordinary shoe-like boots or...

[09:09:00.0000] <AryehGregor>
Well, I do plan to put them on my feet, and hopefully walk in them.  They're rather taller than shoes.

[09:10:00.0000] <jgraham>
/me is familiar with the basic distinguishing features of boots

[09:11:00.0000] <jgraham>
I was just curious because a lot of outdoors0type shops that otherwise have web stores refuse to sell walking boots online due to the liklihood of a poor fit

[09:12:00.0000] <AryehGregor>
It was from Muck Boots.

[09:27:00.0000] <Philip`>
no_mind: Because those are useful features for users when they see something on the screen that they want to save

[09:35:00.0000] <no_mind>
hmm

[09:35:01.0000] <no_mind>
I thought canvas cannot be saved

[09:35:02.0000] <Dashiva>
Then what would the 'save image' do?

[09:35:03.0000] <Philip`>
Firefox demonstrates it can :-)

[09:36:00.0000] <no_mind>
what if I want to present some stuff that I dont want to be saved

[09:36:01.0000] <no_mind>
specially with copyrighted photographs

[09:36:02.0000] <Philip`>
You can't

[09:36:03.0000] <Philip`>
The user could just take a screenshot anyway

[09:37:00.0000] <Dashiva>
Or the user could use toDataUrl() with a bookmarklet and save that

[09:37:01.0000] <Philip`>
You could prevent that by drawing a different-origin image onto the canvas

[09:39:00.0000] <Dashiva>
The user could use a firefox extension to override that? :)

[09:40:00.0000] <Philip`>
Sure

[09:41:00.0000] <Philip`>
They could modify the browser source code to save all images, too

[09:41:01.0000] <no_mind>
well the average users is not taht much talented

[09:41:02.0000] <Philip`>
or they could just grab the photo when you were sending it over the network to the page's scripts

[09:42:00.0000] <Philip`>
If you just want to stop average users right-clicking on the canvas, you could stick a <div style=opacity:0> in front of it or something

[09:42:01.0000] <no_mind>
good idea

[09:43:00.0000] <othermaciej>
MikeSmith: (a) if you reopen NEEDSINFO bugs, you should probably be adding some info

[09:43:01.0000] <Philip`>
/me has encountered one irritating site that circumvents Opera's ad blocking feature by putting a transparent.gif in front of the ad, so clicking on it just stops the transparent image from rendering

[09:43:02.0000] <othermaciej>
MikeSmith: (b) it's not good to reopen bugs that are tagged TrackerISsue

[09:44:00.0000] <Philip`>
and the ad itself is a CSS background-image, so if you disable CSS in Opera (in an attempt to stop the overlaying of elements) then it disappears and you can't click it

[09:46:00.0000] <jgraham>
no_mind: Not really a good idea

[09:46:01.0000] <no_mind>
why

[09:47:00.0000] <jgraham>
Because it seems like a lame, ineffective, technical solution to something that is mostly a non-problem

[09:47:01.0000] <jgraham>
Anyone sufficiently determined can download your content anyway

[09:47:02.0000] <no_mind>
ok let me explain my problem

[09:48:00.0000] <no_mind>
I have a client who uses flash to create slideshows and they want to move away from flash

[09:48:01.0000] <no_mind>
the single point stopping them is the fact that users cannot download the images inside the flash 

[09:48:02.0000] <AryehGregor>
So just use a transparent overlay.

[09:48:03.0000] <no_mind>
with html plus js based slideshow, a user can download individual images

[09:49:00.0000] <AryehGregor>
That's what YouTube uses in its HTML5 video, I think.

[09:49:01.0000] <AryehGregor>
It will stop basic right-clicking.

[09:49:02.0000] <Philip`>
Won't help with advanced right-clicking, but that takes years of practice

[09:50:00.0000] <jgraham>
no_mind: There are at least a variety of tools on the internet that claim to allow images to be extracted from flash

[09:50:01.0000] <AryehGregor>
jgraham, his client doesn't know that.

[09:50:02.0000] <AryehGregor>
:)

[09:50:03.0000] <Philip`>
The problem is how to provide the illusion of protection, not how to actually provide protection that works

[09:50:04.0000] <no_mind>
neither my client nor his clients know this :)

[09:51:00.0000] <jgraham>
Well by all means do the transparent overlay thing if it is the only way to get paid

[09:51:01.0000] <no_mind>
I was thinking of streaming the image and receiving it in a JS variable to draw the image on a canvas element

[09:52:00.0000] <AryehGregor>
Overly complicated.

[09:52:01.0000] <Philip`>
And slow

[09:52:02.0000] <jgraham>
And less effective even than the transparent overlay

[09:52:03.0000] <no_mind>
that works in most browsers except FF, which gives a "Save image" option

[09:52:04.0000] <Philip`>
If you want to lie to your clients about protection, you might as well not make it significantly technically inferior too :-)

[09:52:05.0000] <AryehGregor>
Relatively position the image's container, stick in a <div style="z-index: 1; left: 0; top: 0; right: 0; bottom: 0">.

[09:52:06.0000] <AryehGregor>
Er, and add position: absolute.

[09:53:00.0000] <AryehGregor>
To the div's style.

[09:53:01.0000] <AryehGregor>
Or whatever, something like that.

[09:53:02.0000] <AryehGregor>
Maybe you need opacity: 0 too.

[09:53:03.0000] <no_mind>
AryehGregor, that I know how to do :)

[09:53:04.0000] <jgraham>
And hope that no one who wants to download the pictures has hear of firebug/web inspector/dragonfly

[09:53:05.0000] <jgraham>
*heard

[09:53:06.0000] <AryehGregor>
jgraham, no, he doesn't care about that.  All that matters is his client hasn't heard of that.

[09:53:07.0000] <AryehGregor>
If he has, then show him a tool that extracts the images from Flash just as easily.

[09:54:00.0000] <jgraham>
Well sure

[09:54:01.0000] <AryehGregor>
This is about the "garden fence" kind of thing, anyway.  If most people don't know how to do it, that's probably good enough for the client.

[09:55:00.0000] <Philip`>
Depends on what you're trying to prevent

[09:55:01.0000] <Philip`>
If your goal is to stop people breaking into your house, a garden fence is useless even if most people will respect it

[09:55:02.0000] <jgraham>
Possibly most people aren't actually interested in downloading the content and using it in some bad way anyway

[09:55:03.0000] <AryehGregor>
Yes, so really this is all just a game of pleasing the client.

[09:55:04.0000] <AryehGregor>
Not much to discuss.

[09:55:05.0000] <annevk>
"I'll have a relationship of some kind with the magazine until I'm an old web coot telling young people about how we edited our HTML by hand rather than having our digital sex pony avatars do it for us in our Farmbooks." -- http://www.theawl.com/2010/03/a-conversation-with-paul-ford-the-now-former-web-editor-of-harpers-magazine

[09:55:06.0000] <annevk>
via markp

[09:56:00.0000] <annevk>
or mpilgrim

[09:56:01.0000] <annevk>
or some such

[09:56:02.0000] <no_mind>
well in case of my client he had 3 incidents in last one year where people have just right-click and saved images to use them in other places and not give him credits

[09:56:03.0000] <jgraham>
Wht type of conetent is it?

[09:56:04.0000] <jgraham>
*content

[09:56:05.0000] <no_mind>
mostly images and some text with images

[09:57:00.0000] <jgraham>
What type of images?

[09:57:01.0000] <no_mind>
Monuments, people, some interesting photography tricks

[09:57:02.0000] <jgraham>
Like, is it more effective just to put a watermark in the image itself?

[09:57:03.0000] <jgraham>
For photography it probably is

[09:58:00.0000] <AryehGregor>
Surely watermarks are more obnoxious than a transparent div overaly.

[09:58:01.0000] <AryehGregor>
overlay.

[09:58:02.0000] <no_mind>
putting watermark spoils the whole feel

[09:58:03.0000] <no_mind>
these people are were specific about the overall feel of the work.

[09:58:04.0000] <no_mind>
that is why he switched to flash

[09:59:00.0000] <jgraham>
Depends how you do it really, but they are typically more effective against people who are determined enough to actually download an image and use it on their own site

[09:59:01.0000] <jgraham>
(I expect)

[09:59:02.0000] <no_mind>
then I showed him step by step replacement of all his requirements in HTML5, JS and SVG

[09:59:03.0000] <no_mind>
to which he agrees but is stuck on this one point

[09:59:04.0000] <no_mind>
those who are determined cant be stopped :)

[10:00:00.0000] <no_mind>
but the probability is fairly low

[10:01:00.0000] <AryehGregor>
jgraham, if there are three cases in the last year, a transparent overlay is probably enough to stop them all with fairly good probability.

[10:01:01.0000] <AryehGregor>
It will stop most people.

[10:01:02.0000] <jgraham>
Well the probability of someone taking clearly-copyrighted work and assuming they can use it on their own site without permission or reprecussions is quite low to start with

[10:02:00.0000] <jgraham>
I would have thought that the sort of people who were prepared to do this at all were also the sort of people that would be prepared to work around a trivial technical barrier

[10:02:01.0000] <jgraham>
But I might be wrong

[10:03:00.0000] <AryehGregor>
To most people, trivial technical barriers are not trivial.

[10:04:00.0000] <AryehGregor>
My mother is an art historian.  She downloads out-of-copyright images all the time from websites to use in her lectures.

[10:04:01.0000] <jgraham>
If these people are putting things on the web there is a higher than usual possibility they have some understanding of how the web works

[10:04:02.0000] <AryehGregor>
But if right-clicking doesn't work, she has to call me in.  Or, more likely, give up, since I'm not going to do some complicated thing with Firebug for dozens of images in a row.

[10:04:03.0000] <AryehGregor>
It's amazingly easy to create a website these days.

[10:04:04.0000] <no_mind>
AryehGregor, exactly

[10:05:00.0000] <AryehGregor>
My mother puts things up on her website by using folders that my brother and I have carefully set up to automatically log in via FTP but behave just like a normal directory.

[10:05:01.0000] <no_mind>
and people do not care about copyright as much as we expect them to be. In many cases people are not shy to hotlink images from other websites as their own

[10:05:02.0000] <AryehGregor>
She can't create HTML pages, but she puts up images and slideshows and things like that.

[10:08:00.0000] <gsnedders>
But does circumventing it fall under the circumventing DRM clause in the DCMA?

[10:08:01.0000] <AryehGregor>
Who knows?  Ask a lawyer.

[10:08:02.0000] <gsnedders>
If it does, then we would be adding DRM to HTML, which seems bad.

[10:08:03.0000] <AryehGregor>
What?

[10:08:04.0000] <AryehGregor>
We're not adding anything to HTML.

[10:08:05.0000] <gsnedders>
AryehGregor: Most Swedish lawyers won't know.

[10:08:06.0000] <gsnedders>
/me has just seen fragments of this

[10:08:07.0000] <AryehGregor>
Nor will most non-lawyers, which is what this channel is made up of.

[10:09:00.0000] <gsnedders>
I don't know :P

[10:09:01.0000] <no_mind>
is DCMA accepted in Sweeden ?

[10:09:02.0000] <AryehGregor>
Nor will most American lawyers, unless they're assigned to research it and come up with an answer, and that answer might end up being "depends" or "maybe".

[10:09:03.0000] <AryehGregor>
no_mind, no, the DMCA is an American law.

[10:09:04.0000] <AryehGregor>
The EUCD is similar, but I can't be bothered figuring out if Sweden is in the EU and/or has implemented it.

[10:10:00.0000] <ment>
maybe we all get DCMA through ACTA soon

[10:10:01.0000] <AryehGregor>
gsnedders, we were just talking about some guy who used Flash to do a slideshow so people couldn't take his images, and we were telling him to use a transparent div overlay to stop it in HTML/CSS.

[10:10:02.0000] <AryehGregor>
I've seen a lot of FUD about ACTA, but nothing that I'm actually worried about.

[10:10:03.0000] <AryehGregor>
These things need to be passed by legislatures before they do anything, you know.

[10:10:04.0000] <no_mind>
Seeden is in EU, I know it sitting in India :p

[10:10:05.0000] <AryehGregor>
Lots of treaties that executives have negotiated but legislatures didn't ratify.

[10:11:00.0000] <AryehGregor>
The League of Nations and the Kyoto Protocol are two US examples.

[10:11:01.0000] <AryehGregor>
And before it's passed by legislatures, it needs to be made public, so everyone will have a chance to review it and object at some point.  No need to fret now when we don't know anything about it.

[10:11:02.0000] <no_mind>
The problem is that such treaties can be used in proxy even if legislature did not ratify them

[10:12:00.0000] <AryehGregor>
What do you mean?

[10:12:01.0000] <no_mind>
We have seen the case of fairplay in India. Where Apple and RIAA managed to take the host to court, even though US patents do not apply in India

[10:13:00.0000] <AryehGregor>
Well, you can always take someone to court.  Did they win?

[10:13:01.0000] <no_mind>
though the host backed out later and hosting was moved to some other place

[10:13:02.0000] <AryehGregor>
You can try to get judgments in one country enforced in another, but I'm not clear on what the procedure is.

[10:14:00.0000] <no_mind>
Well in India they wont have managed to force the judgement. Anti software patents lobby is pretty strong here

[10:14:01.0000] <no_mind>
but it was enough to scare other hosts

[10:14:02.0000] <AryehGregor>
For obvious reasons.  :)

[10:15:00.0000] <AryehGregor>
Poorer countries are going to be in favor of lax IP rights, because Western IP is usually priced for Westerners.  Whereas rich countries will be in favor of strong international IP for the same reason.

[10:16:00.0000] <Dashiva>
Poorer countries also often have the advantage of not having any legacy laws that snuck in before anyone realized what a stupid idea software patents was

[10:16:01.0000] <no_mind>
well India favors IP but not software patents

[10:17:00.0000] <no_mind>
copyright laws and trademark laws are pretty strong here, in terms of enforcement

[10:17:01.0000] <no_mind>
even patents

[10:17:02.0000] <no_mind>
but software patents and business patents are a strict no-no

[10:18:00.0000] <AryehGregor>
So people buy legitimate copies of Windows from Microsoft, for how much money?

[10:19:00.0000] <no_mind>
starts from about INR 7000/- roughly (USD 150)

[10:19:01.0000] <no_mind>
though piracy of windows is as rampant as any Asian country

[10:20:00.0000] <AryehGregor>
And people pay that rather than pirating?  I've heard of other countries (Russia, etc.) that you can get pirated Windows all over the place on the street.

[10:20:01.0000] <no_mind>
yes you do get pirated windows here, easily

[10:20:02.0000] <AryehGregor>
I guess "strongly enforced" is relative.  Richer countries can inevitably pay for more enforcement.

[10:20:03.0000] <no_mind>
but you are caught easily too

[10:21:00.0000] <no_mind>
AryehGregor, Enforcement depends on comapny too. While it is difficult to find and survive with pirated Adobe products, it is easy to do this with Microsoft

[10:22:00.0000] <no_mind>
Adobe follows stricter policies

[10:22:01.0000] <no_mind>
while Microsoft is very laxed here. They wait for 2-4 years before filing a case

[10:22:02.0000] <no_mind>
when the person is totally dependent on windows

[10:23:00.0000] <no_mind>
then they come threatening, pay us the license fee or we will file a complaint. Piracy happens to be a non-bailable offence with minimum of 3 years imprisonment

[10:24:00.0000] <AryehGregor>
Interesting.  In the US you can only get fined, I think, not imprisoned, unless you're distributing.

[10:25:00.0000] <no_mind>
here if a complaint is filed, it will be minimum of 2 months before you get a bail. And if out of court settlement happens fine else you get fine plus imprisonment

[10:25:01.0000] <AryehGregor>
Fascinating.

[10:25:02.0000] <no_mind>
there seems to be a nexus here because what I have noticed that distributors always get away while user suffers

[10:25:03.0000] <AryehGregor>
Probably because it's good business for MS to hook people on Windows when it looks like it's free.

[10:26:00.0000] <AryehGregor>
Glad I use Linux.

[10:26:01.0000] <no_mind>
and since M$ sales in India have started to fall, they are imposing anti piracy measures even in smaller cities

[10:26:02.0000] <gsnedders>
/me wonders about the HTC Legend

[10:38:00.0000] <MikeSmith>
othermaciej: apologies about bug 7539

[10:38:01.0000] <MikeSmith>
that change was a mistake

[10:38:02.0000] <MikeSmith>
about the others, I will try to go back in today and add more info for those

[10:39:00.0000] <othermaciej>
MikeSmith: I put the ones that had TrackerIssue back in VERIFIED

[10:39:01.0000] <MikeSmith>
oh.. I didn't know I'd done that for others than 7539

[10:39:02.0000] <othermaciej>
MikeSmith: if there are any that the TF intended to "de-escalate" (or whatever) then please re-reopen and remove TrackerIssue so they're not in an inconsistent stte

[10:39:03.0000] <MikeSmith>
OK

[10:40:00.0000] <MikeSmith>
/me looks back again at bugmail

[10:40:01.0000] <othermaciej>
I got bugzilla mail about a bunch of them and moved them back

[10:44:00.0000] <MikeSmith>
othermaciej: yeah, I see those now too

[10:44:01.0000] <MikeSmith>
sorry for muffing that up

[10:44:02.0000] <MikeSmith>
I will avoid doing that again

[10:45:00.0000] <othermaciej>
thanks, it seems to be a common mistake, so it's good to raise awareness

[10:46:00.0000] <MikeSmith>
well, at least I should know better.. problem was I was changing those during the last 10 minutes of the a11y telcon and wasn't really looking at them carefully

[11:26:00.0000] <Philip`>
What's the appropriate <source type> for MP3s?

[11:35:00.0000] <othermaciej>
audio/mp3 ?

[11:39:00.0000] <mpilgrim>
AFAICT, "distributed extensibility" is defined as "whatever HTML5 doesn't have yet"

[11:39:01.0000] <Dashiva>
And "whatever HTML5 doesn't have anymore" :)

[11:39:02.0000] <AryehGregor>
Well, that's why it's distributed.

[11:39:03.0000] <AryehGregor>
If HTML5 had it, it would be centralized, right?

[11:39:04.0000] <mpilgrim>
the contrapositive is also true -- if it is determined that a use case can be solved with something already in HTML5 (like "data-*" attributes), then the use case is automatically NOT "distributed extensibility"

[11:40:00.0000] <mpilgrim>
once i realized this, the ongoing discussions about ISSUE-41 made much more sense to me

[11:41:00.0000] <Philip`>
othermaciej: Without any crazy codecs= or profiles or anything? That seems far too easy

[11:41:01.0000] <Philip`>
Maybe I should switch to AAC so that it satisfies my expected level of complexity

[11:42:00.0000] <Philip`>
Any guesses as to whether IE9 will support MP3 and/or AAC?

[11:43:00.0000] <Philip`>
/me assumes at least the latter, since it goes along with H.264

[11:48:00.0000] <mpilgrim>
Philip`: MS claimed to have the Youtube HTML5 beta working, so whatever audio codec is used in those videos is certainly supported

[11:48:01.0000] <mpilgrim>
(or maybe codecs, plural)

[11:49:00.0000] <mpilgrim>
i don't actually know which codecs those are, but presumably you could violate the youtube terms of service, download a video, and check

[11:50:00.0000] <mpilgrim>
as a google employee, i can not actually recommend any tools for this purpose, but i'm told we still have a search engine that might help you

[11:50:01.0000] <Philip`>
http://blogs.msdn.com/ie/archive/2010/03/16/html5-hardware-accelerated-first-ie9-platform-preview-available-for-developers.aspx - "Not WMV. IE9 supports MP4 h.264 video, MP3 or AAC audio"

[11:51:00.0000] <Philip`>
I guess that answers it without me having to risk being sued by Google

[11:52:00.0000] <Philip`>
I wouldn't dare search for YouTube downloading tools, the Google central intelligence will surely detect it and scan my emails and consider me a subversive

[11:54:00.0000] <daedb_>
No wmv? I figured they'd at least allow that as an option alongside h264...

[12:10:00.0000] <Philip`>
daedb_: I certainly won't complain if they promote non-proprietary standards instead of WMV :-)

[12:11:00.0000] <Philip`>
It's almost as if they wanted to encourage interoperability

[13:06:00.0000] <AryehGregor>
Who was that Mozilla person who wrote a blog post about how H.264 is patented outside of the US?

[13:06:01.0000] <AryehGregor>
Ah, bzbarsky.

[13:06:02.0000] <AryehGregor>
/me should Google before asking

[16:35:00.0000] <paul_irish>
data-* attributes are gone from all current specs, right?

[16:37:00.0000] <Dashiva>
No?

[16:37:01.0000] <paul_irish>
oh its in html5.. embedding-custom-non-visible-data

[16:37:02.0000] <Dashiva>
Yup

[16:37:03.0000] <paul_irish>
i thought it was thrown out with the microdata bathwater. phew.

[16:38:00.0000] <AryehGregor>
/me begins reading through a two-week backlog of spec mail

[16:38:01.0000] <Dashiva>
Have fun

[16:58:00.0000] <AryehGregor>
Why does <http://wiki.whatwg.org/favicon.ico> not have a transparent background?


2010-03-19
[17:25:00.0000] <JonathanNeal>
AryehGregor, yea what gives

[17:30:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/wiki.whatwg.org.favicon.ico There you go, AryehGregor, you go ahead and enjoy a favicon with a transparent background 

[17:31:00.0000] <AryehGregor>
/me remembers that he has shell access to wiki.whatwg.org

[17:32:00.0000] <AryehGregor>
Yay, works.

[17:33:00.0000] <AryehGregor>
JonathanNeal, thanks.  :)

[17:33:01.0000] <JonathanNeal>
/me can finally tell his kids one day, "Kids, your dad took the white background off the whatwg's wiki favicon once."

[17:33:02.0000] <JonathanNeal>
They'll say "Wow, it looks so much better in Chrome, Father dearest."

[17:34:00.0000] <JonathanNeal>
Haha, I didn't even realize you already have a png!

[17:35:00.0000] <AryehGregor>
Yeah, it's mainly in Chrome that it looks ugly.  Blue background and all.

[17:37:00.0000] <JonathanNeal>
I had no idea you would actually use that ... okay I took it a little more seriously.  Refresh my favicon and use that one, it's actually a 16x16 ico of your png.

[17:37:01.0000] <JonathanNeal>
Otherwise, you'll have these little tiny white edges on your favicon and I'll just kill myself.

[17:38:00.0000] <JonathanNeal>
Then Marty will never be born, and there'll be no one to go back and warn Doc Brown.

[17:38:01.0000] <AryehGregor>
Done.

[17:38:02.0000] <AryehGregor>
Looks the same to me.

[17:39:00.0000] <AryehGregor>
But that's why I'm a programmer, not a designer.  :P

[17:39:01.0000] <JonathanNeal>
np, lemme know if you ever need stuff like that, it takes but 20 seconds.

[17:40:00.0000] <AryehGregor>
I could probably have done it myself, actually, although not in 20 seconds.

[17:40:01.0000] <AryehGregor>
In fact, it was more of a random gripe, but then when you gave the improved version, I remembered I have shell access, so . . .

[17:41:00.0000] <JonathanNeal>
Now I can die in peace, ty AryehGregor. :P

[17:45:00.0000] <AryehGregor>
/me heartily endorses http://www.w3.org/TR/2010/WD-html-bidi-20100304/

[17:46:00.0000] <JonathanNeal>
just supporting dir in all elements?

[17:46:01.0000] <AryehGregor>
No, it has a detailed list of bidi problems that should be fixed.

[17:46:02.0000] <AryehGregor>
I've run into a lot of them.

[17:47:00.0000] <AryehGregor>
(since I know Hebrew, and occasionally write it on computers, although not often)

[17:48:00.0000] <AryehGregor>
/me got through six days' worth of spec mail in the last hour or so, ten days to go . . . but bedtime now

[17:54:00.0000] <JonathanNeal>
Nite nite!

[01:29:00.0000] <annevk>
MikeSmith, I think our process for going out of Last Call is not having outstanding issue/discussion threads for a while

[01:30:00.0000] <annevk>
though ultimately, just like with the W3C, it'll be a judgment call

[02:06:00.0000] <hsivonen>
it's really sad that I felt compelled to file http://www.w3.org/Bugs/Public/show_bug.cgi?id=9279

[02:06:01.0000] <hsivonen>
the effects of putting UI strings in the DOM and the html5lib test suite...

[02:08:00.0000] <annevk>
that UI strings can be generated by the HTML parser is insane to begin with...

[02:09:00.0000] <asmodai>
http://www.in-nomine.org/~asmodai/mathml/msup.html <-- needs html 5 + mathml

[02:10:00.0000] <hsivonen>
I guess privacy features that mask the UI language should also emit the isindex prompt in en-US

[02:10:01.0000] <hsivonen>
mask the UI language from servers that is

[02:10:02.0000] <asmodai>
bottom two look horrendous, the third from the bottom looks the best but is probably not the way the mathml is supposed to be, and the fourth, well, definitely not how it should

[02:11:00.0000] <hsivonen>
IIRC, some folks in the IETF argued that you can ethnically profile people from their UI language and then ethnically discriminate

[02:11:01.0000] <hsivonen>
(no, I don't have a link to email archives at hand)

[02:12:00.0000] <annevk>
hsivonen, oh yeah, hadn't even considered that the HTML parser leaked privacy information therefore

[02:12:01.0000] <annevk>
only slightly insane o_O

[02:13:00.0000] <hsivonen>
I should make a note for this for Mozilla's security review of the HTML5 parser

[02:14:00.0000] <asmodai>
hsivonen: yeah, using 'here' is such a bad usability string.

[02:14:01.0000] <jgraham>
Why do people care about leaking UA language information?

[02:15:00.0000] <hsivonen>
jgraham: see above: ethnic profiling and discrimination

[02:16:00.0000] <asmodai>
Is it scary that I never even considered discrimination based on UI language? I wonder if that makes me naïve, idealistic, or just plain dumb.

[02:17:00.0000] <hsivonen>
it's not a given that everything perceived to be a threat at the IETF actually is

[02:18:00.0000] <jgraham>
That sounds farfetched, illegal, and easilly worked arounf

[02:19:00.0000] <jgraham>
So not really something I would be that concerned about

[02:19:01.0000] <annevk>
but some people are

[02:20:00.0000] <jgraham>
Right, but some people are concerned with lots of things

[02:20:01.0000] <jgraham>
Doesn't make them all equally worrysome

[02:28:00.0000] <MikeSmith>
annevk: thanks (for the clarification about whatwg last call)

[02:29:00.0000] <MikeSmith>
annevk: but I think in the community there remains some unproductive confusion about what it means for the whatwg HTML5 spec to be in last call

[02:29:01.0000] <hsivonen>
MikeSmith: what's the "community" in this case?

[02:30:00.0000] <MikeSmith>
the community represented by the several people so far who have asked me personally about this

[02:30:01.0000] <hsivonen>
ok

[02:30:02.0000] <annevk>
MikeSmith, maybe you should suggest them to add a question to the FAQ

[02:31:00.0000] <MikeSmith>
I guess

[02:31:01.0000] <MikeSmith>
or maybe somebody who understands what it does mean can just go ahead and add the question and the answer

[02:32:00.0000] <MikeSmith>
because I have to say I personally don't know what the answer should be

[02:32:01.0000] <annevk>
you can just add a question without giving an answer

[02:32:02.0000] <annevk>
i.e. by putting a ? there

[02:34:00.0000] <hsivonen>
list of Frequently Asked Questions. no one promised answers.

[02:34:01.0000] <MikeSmith>
true

[02:34:02.0000] <MikeSmith>
anyway, the simple and candid answer that I give to people who ask me in person is that they can safely ignore the fact that the whatwg version is in last call

[02:34:03.0000] <MikeSmith>
but I don't think that's a suitable answer to put into the FAQ

[02:37:00.0000] <MikeSmith>
also, it's not even clear to me if it is actually still in last call

[02:37:01.0000] <MikeSmith>
e.g., the section on the device element clearly is not

[02:38:00.0000] <hsivonen>
MikeSmith: for a candid answer, see also http://krijnhoetmer.nl/irc-logs/whatwg/20091029#l-273

[02:38:01.0000] <MikeSmith>
/me looks now

[02:39:00.0000] <MikeSmith>
/me waits

[02:39:01.0000] <MikeSmith>
heh

[02:40:00.0000] <MikeSmith>
yeah, well, that's also probably not the answer we'd want to put in the FAQ

[02:40:01.0000] <jgraham>
Isn't the answer just that all the subsections that are considered "HTML5" are individually in LC status?

[02:53:00.0000] <jgraham>
I hate javascript, it has made me reflexively write === everywhere

[02:59:00.0000] <virtuelv>
jgraham: then I'm wondering what you think of Perl

[02:59:01.0000] <virtuelv>
cause it makes me reflexiely write @_#!¤#&%! everywhere

[02:59:02.0000] <annevk>
jgraham, you're in the ideal position to fix it!

[02:59:03.0000] <annevk>
virtuelv, lol

[03:02:00.0000] <asmodai>
/me shoots Perl.

[03:06:00.0000] <jgraham>
virtuelv: I try not to think of Perl

[03:15:00.0000] <hsivonen>
Will Perl6 be released before 2022?

[03:21:00.0000] <Philip`>
/me predicts yes

[03:45:00.0000] <no_mind>
well HTML5 will be released before Perl6 :)

[03:49:00.0000] <Philip`>
Depends what "released" means, I guess

[03:50:00.0000] <Philip`>
http://use.perl.org/~pmichaud/journal/39411

[04:13:00.0000] <no_mind>
So is Perl like Osho ? Never born never dead, visiting the planet earth. ?

[04:21:00.0000] <gsnedders>
http://pegjs.majda.cz/ — a parser generator for JS

[05:06:00.0000] <lazni>
how can I tell opera not to apply its feed view for http://hdh.dyn-o-saur.com/plos/plos.xhtml?

[05:08:00.0000] <lazni>
http://hdh.dyn-o-saur.com/plos/plos.html would show the text, but the CSS PI isn't applied in text/html 

[05:08:01.0000] <asmodai>
hahaha, nice bugtracking category @ Opera: ALIEN? :)

[05:11:00.0000] <karlcow>
lazni: plos.xhtml is viewed as an xml document as well in Firefox

[05:14:00.0000] <lazni>
but firefox doesn't apply its feed view

[05:15:00.0000] <lazni>
(because .xhtml doesn't contain "<feed"

[05:16:00.0000] <lazni>
hm, maybe I should try writting it polyglottically instead of XSLT

[05:23:00.0000] <zcorpan>
lazni: http://simon.html5.org/dump/html+js+css+atom.html

[05:25:00.0000] <lazni>
great, thanks

[05:55:00.0000] <annevk>
asmodai, oooh, that's secret

[05:55:01.0000] <annevk>
at least so we were told by the Norwegian and American government

[05:57:00.0000] <Philip`>
annevk: I think you meant "whopping", not "whooping"

[06:03:00.0000] <annevk>
that is too bad, since "whooping" sounds a lot nicer

[06:03:01.0000] <annevk>
thanks though

[06:27:00.0000] <asmodai>
annevk: hahaha

[06:27:01.0000] <asmodai>
annevk: http://my.opera.com/desktopteam/blog/2010/03/13/opera-10-51-rc1-for-windows

[06:27:02.0000] <asmodai>
annevk: Not so secret eh? ;)

[06:40:00.0000] <annevk>
asmodai, they must have missed the memo

[06:58:00.0000] <hsivonen>
this is exciting. I've tested three browser engines for w.document.readyState immediately after var w = window.open(...

[06:58:01.0000] <hsivonen>
so far three different results

[06:58:02.0000] <hsivonen>
will the fourth engine have a fourth value

[06:58:03.0000] <hsivonen>
readyState has four possible values...

[07:00:00.0000] <jgraham>
(theoretically)

[07:01:00.0000] <hsivonen>
boohoo. IE had the same value as Safari

[07:01:01.0000] <hsivonen>
aside: the value given by IE and Safari is the least useful

[07:02:00.0000] <hsivonen>
Minefield: uninitialized; Safari 4.0.5 and IE8: complete; Opera 10.50: interactive

[07:02:01.0000] <hsivonen>
w00t. got a fourth value out of IE8!

[07:02:02.0000] <hsivonen>
depends on caching!

[07:03:00.0000] <hsivonen>
interop FTW!

[07:04:00.0000] <annevk>
/me wonders if the rules of the channel have been violated

[07:04:01.0000] <annevk>
/me can't believe we all went ahead and implemented the crime that is document.readState in the first place

[07:05:00.0000] <hsivonen>
/me can't see a way to empty the cache in IE8's UI

[07:06:00.0000] <hsivonen>
found it now.

[07:07:00.0000] <hsivonen>
maybe IE doesn't depend on caching but on something else more random

[07:07:01.0000] <garbeam_>
I'm reading this http://dev.w3.org/html5/eventsource/ and stumbled across section 8: "Authors are also cautioned that HTTP chunking can have unexpected negative effects on the reliability of this protocol[..]" - any evidence for that?

[07:07:02.0000] <hsivonen>
like network speed or something

[07:08:00.0000] <hsivonen>
does IE have an equivalent of document.documentURI?

[07:08:01.0000] <annevk>
garbeam_, prolly more effective to mail public-webapps⊙wo with that question

[07:18:00.0000] <jgraham>
Hmm, I live in the wrong country

[07:19:00.0000] <gsnedders>
jgraham: Because there's too much ice and snow here?

[07:19:01.0000] <garbeam_>
annevk: ok done, thx

[07:19:02.0000] <gsnedders>
It's kinda sad that's apparently WONTFIX BYDESIGN

[07:20:00.0000] <jgraham>
gsnedders: Because I just saw the lineup for the latitude festival, and now I am glum

[07:20:01.0000] <jgraham>
Moreso than usual I mean

[07:21:00.0000] <gsnedders>
Yeah, I was going to say. When are you not glum?

[07:21:01.0000] <annevk>
garbeam_, you posted to the WHATWG list, but that works too

[07:21:02.0000] <gsnedders>
But oh dear, I have to put up with an even more glum jgraham…

[07:21:03.0000] <garbeam_>
annevk: yes, I try to avoid W3C

[07:23:00.0000] <asmodai>
gsnedders: Nothing a tranq dart can't cure.

[07:23:01.0000] <MikeSmith>
garbeam_: me too, but I'm not always so successful at it

[07:23:02.0000] <garbeam_>
MikeSmith: it's a matter of self-discipline ;)

[07:24:00.0000] <jgraham>
I'm not sure tranquiliser darts are a cure for glumness

[07:24:01.0000] <MikeSmith>
garbeam_: in my case, I guess it's more a matter of masochism

[07:24:02.0000] <jgraham>
At the best they are a temporary respite

[07:27:00.0000] <asmodai>
jgraham: It will keep you from going totally over to the dark side and starting to cut yourself and all that.

[07:30:00.0000] <jgraham>
jgraham: No, gsnedders is the emo kid

[07:30:01.0000] <gsnedders>
jgraham: I AM NOT EMO.

[07:30:02.0000] <gsnedders>
/me wonders how he can prove to jgraham he is not emo.

[07:30:03.0000] <jgraham>
No, your're not *emu*, you are emo

[07:31:00.0000] <asmodai>
gsnedders paints his nails black and all that?

[07:32:00.0000] <gsnedders>
No, I don't.

[07:32:01.0000] <gsnedders>
Oh, jgraham's saying something. Maybe I ought to take my headphones off. On second thought, he's probably just calling me emo, so there's little point.

[07:34:00.0000] <asmodai>
rofl

[07:34:01.0000] <TabAtkins>
/me stole a girl fomr someone as a result of painting his nails black + black lipstick.  That was goth, though, not emo.

[07:34:02.0000] <gsnedders>
fomr?

[07:35:00.0000] <Philip`>
"for"

[07:35:01.0000] <TabAtkins>
"from".

[07:35:02.0000] <gsnedders>
See, if jgraham just called me goth all the time, I wouldn't complain anywhere near so much. I'm a lot closer to being a goth than being emo…

[07:35:03.0000] <gsnedders>
jgraham would never accept that.

[07:35:04.0000] <Philip`>
He runs a girl thievery business for clients

[07:35:05.0000] <TabAtkins>
That would be lucrative!

[07:35:06.0000] <gsnedders>
He has to call me emo just to annoy me

[07:36:00.0000] <gsnedders>
I guess the fact that I'm wearing black entirely and have pale skin is cause for jgraham to think I'm emo…

[07:37:00.0000] <gsnedders>
There again, I have a ponytail, so I can't be emo.

[07:38:00.0000] <asmodai>
goth++

[07:39:00.0000] <asmodai>
And that said, time for some Trail of Tears

[07:39:01.0000] <jgraham>
gsnedders: You're no more goth than I am

[07:40:00.0000] <gsnedders>
Yet I'm more emo? Okay…

[07:41:00.0000] <jgraham>
I you were goth, I would expected, died, straightened, black hair, at the very least

[07:41:01.0000] <asmodai>
Aside from the goatee and the moustache and semi-long hair I doubt I'd pass as remotely metal ;)

[07:41:02.0000] <jgraham>
*If

[07:41:03.0000] <asmodai>
s/died/dyed I hope

[07:41:04.0000] <jgraham>
Oh

[07:42:00.0000] <jgraham>
Well yes

[07:42:01.0000] <jgraham>
Although dead hair would be even more goth

[07:42:02.0000] <TabAtkins>
Goths don't require black hair.  Clothing and makeup will do.

[07:42:03.0000] <TabAtkins>
Your hair is presumably already dead?

[07:42:04.0000] <jgraham>
gsnedders has clothing, I'm happy to say

[07:42:05.0000] <asmodai>
/me always has to laugh at those folks who put on face paint

[07:42:06.0000] <jgraham>
Not very goth, but enough to cover himself

[07:43:00.0000] <jgraham>
makeup not so much

[07:43:01.0000] <gsnedders>
/me has eyeliner at home, as well as black lipstick…

[07:43:02.0000] <TabAtkins>
See?  Goth it is.

[07:43:03.0000] <jgraham>
Not *at home*

[07:43:04.0000] <jgraham>
*Note

[07:43:05.0000] <gsnedders>
And black clothes, plenty of black.

[07:43:06.0000] <TabAtkins>
Or punk metal, I suppose.

[07:43:07.0000] <jgraham>
He probably means "in Scotland"

[07:43:08.0000] <gsnedders>
TabAtkins: Oh, I am no way punk metal

[07:43:09.0000] <jgraham>
i.e. not even important enough to bring with him

[07:43:10.0000] <jgraham>
Not goth

[07:44:00.0000] <gsnedders>
In Linköping, actually

[07:44:01.0000] <jgraham>
Wear it or I don't believe you

[07:45:00.0000] <TabAtkins>
Tits or gtfo.

[07:45:01.0000] <gsnedders>
TabAtkins: Sorry, I'm not getting a sex change just for the sake of that

[07:52:00.0000] <gsnedders>
Is it bad I'm now arguing against the existance of rationale?

[07:53:00.0000] <gsnedders>
Or does the belief in emotion over rationale make me emo, somehow?

[07:55:00.0000] <jgraham>
Yes, yes it does

[07:57:00.0000] <gsnedders>
_How?_

[08:23:00.0000] <jgraham>
It's not self evident?

[08:24:00.0000] <gsnedders>
No. I obviously don't self-harm…

[08:25:00.0000] <JonathanNeal>
http://html5gallery.com/ woo Liferay!

[08:25:01.0000] <jgraham>
I don't think that's the only criterion

[08:37:00.0000] <erlehmann>
how would you name „emotion over reason“, gsnedders ?

[08:38:00.0000] <Dashiva>
"Shrill"

[08:48:00.0000] <gsnedders>
erlehmann: I wouldn't, I'd just use such a phrase :P

[09:09:00.0000] <boblet>
JonathanNeal: congrats :)

[09:09:01.0000] <JonathanNeal>
thanks boblet

[09:10:00.0000] <boblet>
that reminds me I have some sites to submit too

[09:12:00.0000] <lazni>
does mime a/x+x always force standard mode?

[09:13:00.0000] <gsnedders>
lazni: Yes, quirks mode exists only for text/html content

[09:13:01.0000] <JonathanNeal>
This brought up an interesting question though, how do you organize preview content.

[09:14:00.0000] <JonathanNeal>
What if you show part of an article that takes you to its own page, is that preview "intended to be independently distributable or reusable"?

[09:14:01.0000] <JonathanNeal>
If not, what is it, is it part of the outline?

[09:16:00.0000] <jgraham>
I would use <section> YMMV

[09:16:01.0000] <JonathanNeal>
You Make Me Vomit?

[09:16:02.0000] <jgraham>
Well that too

[09:16:03.0000] <JonathanNeal>
Your Milage May Vary?

[09:17:00.0000] <jgraham>
I guess we should go with the latter on the basis of public hygine

[09:18:00.0000] <gsnedders>
We need namespaces for acronyms.

[09:18:01.0000] <gsnedders>
s/acronyms/abbrevations/

[09:18:02.0000] <JonathanNeal>
:) yea I could see how they could belong to a section.

[09:18:03.0000] <JonathanNeal>
It's a section containing links to other sections and articles.

[09:18:04.0000] <JonathanNeal>
I s'pose.

[09:24:00.0000] <Philip`>
gsnedders: I suggest a namespacing syntax in which you put disambiguating lowercase letters after each of the acronym's letters

[09:26:00.0000] <jgraham>
Philip`: That doesn't appear to involve URIs and therefore doesn't allow for Distributed Extensibility

[09:39:00.0000] <adactio>
Can I just quickly check with you guys that I'm understanding the script element correctly: if you're linking off to a file using the src attribute, you don't need to use the type attribute but if you are putting code between the opening and closing script tags insetad, you *must* use the type attribute?

[09:40:00.0000] <gsnedders>
No, that's wrong

[09:40:01.0000] <gsnedders>
You never need the type attribute

[09:40:02.0000] <gsnedders>
Except if the script isn't JS

[09:41:00.0000] <adactio>
This bit of the spec: "When used to include data blocks, the data must be embedded inline, the format of the data must be given using the type attribute, and the src attribute must not be specified." makes it sound like type is mandatory when not using src.

[09:42:00.0000] <adactio>
http://www.whatwg.org/specs/web-apps/current-work/multipage/scripting-1.html#script

[09:42:01.0000] <Dashiva>
The sentence before it explicitly says it isn't

[09:42:02.0000] <AryehGregor>
"When used to include data blocks"

[09:42:03.0000] <AryehGregor>
JavaScript isn't data blocks.

[09:43:00.0000] <AryehGregor>
I was never sure what that section of the spec was meant to refer to.

[09:43:01.0000] <adactio>
The sentence before it refers only to scripts referenced with the src attribute, not inline scripts.

[09:43:02.0000] <AryehGregor>
Some hypothetical usage, or do people actually do that?

[09:43:03.0000] <AryehGregor>
"When used to include dynamic scripts, the scripts may either be embedded inline or may be imported from an external file using the src attribute. If the language is not that described by "text/javascript", then the type attribute must be present, as described below."

[09:43:04.0000] <AryehGregor>
That covers all usage of JavaScript, since you're including dynamic scripts, not data blocks.

[09:43:05.0000] <Dashiva>
adactio: "either embedded inline or..."

[09:44:00.0000] <adactio>
Sounds like the sentence that's confusing me could be clarified by adding the words "non-JavaScript" i.e.: "When used to include non-JavaScript data blocks..."

[09:45:00.0000] <AryehGregor>
I think the intent is that all data blocks are non-JavaScript, but it's sort of vague about what "data blocks" means.

[09:45:01.0000] <AryehGregor>
I don't know why the distinction is drawn.

[09:45:02.0000] <AryehGregor>
I suggest you file a bug or post to whatwg.

[09:45:03.0000] <Dashiva>
Data is not code

[09:45:04.0000] <adactio>
I'll post a comment.

[09:45:05.0000] <jgraham>
Sum Ruby might disagree

[09:46:00.0000] <jgraham>
*Sam

[09:46:01.0000] <jgraham>
argh

[09:46:02.0000] <AryehGregor>
adactio, that's the same as filing a bug.

[09:46:03.0000] <gsnedders>
The sum of all rubies?

[09:46:04.0000] <adactio>
comment/bug posted/filed

[09:47:00.0000] <Philip`>
Data blocks are things like <script type="text/glsl-vs">here's some text that other scripts can extract, and it can say x < y etc safely</script>

[09:48:00.0000] <Philip`>
You could do something like put JSON data in there too

[09:48:01.0000] <Dashiva>
Or XML...

[09:48:02.0000] <jgraham>
Or HTML

[09:48:03.0000] <Philip`>
or even put JS code which should be accessible to the rest of the page and should not be executed

[09:48:04.0000] <Philip`>
though then you'd have to come up with a new content-type for it

[09:49:00.0000] <Dashiva>
text/not-javascript-I-promise

[09:49:01.0000] <Philip`>
Dashiva: XML is not great because it might include the string "</script>" and you can't escape that at all

[09:50:00.0000] <Philip`>
but I think some people use it for that anyway

[09:50:01.0000] <AryehGregor>
JSON might also include the string "</script>".

[09:51:00.0000] <Philip`>
You can write "<\/script>" then

[09:52:00.0000] <jgraham>
Philip`: Presumably you can escape the XML

[09:52:01.0000] <Philip`>
Then it wouldn't be XML

[09:52:02.0000] <Philip`>
It'd be XML in some wrapper format

[09:53:00.0000] <jgraham>
Yes, if you're being pedantic

[09:54:00.0000] <jgraham>
(although strictly I don't know if you can embed XML in some other format and call it XML anyway._

[09:54:01.0000] <jgraham>
)

[09:55:00.0000] <AryehGregor>
Random thought: how long until browser JITs implement speculative execution and branch prediction so they can use multiple cores for single-threaded JS?  :)

[09:58:00.0000] <Philip`>
I imagine the synchronisation overhead would make that kind of fine-grained concurrency impossible

[09:59:00.0000] <AryehGregor>
What if you did it while the script was blocked on something, like a network read?

[10:00:00.0000] <Philip`>
The script is almost certainly going to use the read data almost immediately

[10:00:01.0000] <Philip`>
so you couldn't speculate ahead very far

[10:00:02.0000] <AryehGregor>
What if it's a long-running app like Gmail and you can guess the read data?

[10:01:00.0000] <gsnedders>
And there are relatively few cases where you are blocked on network

[10:01:01.0000] <Philip`>
If you're blocked on the network then your performance is terribly anyway

[10:01:02.0000] <Philip`>
and nobody's going to notice a few extra milliseconds

[10:01:03.0000] <AryehGregor>
/me just realized that once IE9 is released, he should update his website to use inline SVG-in-HTML someplace, so that his site is once again so standard that it no longer works interoperably.

[10:02:00.0000] <AryehGregor>
It will be awesomely ironic if my site is so standards-compliant that it only works properly in IE.

[10:02:01.0000] <AryehGregor>
Actually, it wouldn't work properly in any browser, since IE won't support CSS gradients (probably).  Even better.

[10:02:02.0000] <Philip`>
I suppose they could do more things as background optimisations, like guess parameters for functions that haven't been called yet and call them in a separate thread in order to get traces and statistics for the JIT

[10:03:00.0000] <Philip`>
though the actual execution of code would still be single-threaded

[10:03:01.0000] <AryehGregor>
/me realizes he doesn't really know anything about SVG, and should read the spec.

[10:03:02.0000] <AryehGregor>
What spec should I start with?

[10:03:03.0000] <Philip`>
I suggest the SVG spec

[10:03:04.0000] <AryehGregor>
There are multiple SVG specs, as far as my understanding goes.

[10:03:05.0000] <AryehGregor>
Like multiple versions?  Profiles?  That kind of thing?

[10:04:00.0000] <Philip`>
I believe http://www.w3.org/TR/SVG/ is what most browsers target

[10:04:01.0000] <Philip`>
i.e. 1.1

[10:04:02.0000] <JonathanNeal>
A Question about HTML5 outlines for those interested in particulars: I have a collection of blog previews, like 10 titles each with the first paragraph and a "click to read more" link.  Are each of these previews an article, or do they all belong to one article/ section and have no individual outlining weight of their own?

[10:04:03.0000] <Philip`>
(not 1.2 Tiny)

[10:12:00.0000] <AryehGregor>
An IE person responded here, but ignored questions asking about <canvas>: http://blogs.msdn.com/ie/archive/2010/03/18/svg-in-ie9-roadmap.aspx

[10:12:01.0000] <AryehGregor>
I'm guessing they haven't fully decided whether they'll do <canvas> yet or not.

[10:16:00.0000] <paul_irish>
you saw the amd press release that hinted at it, yes?

[10:16:01.0000] <paul_irish>
http://blogs.amd.com/developer/2010/03/16/ie9-takes-advantage-of-the-gpu/ " The <canvas> element will be accelerated on the GPU via Direct2D and...."

[10:17:00.0000] <AryehGregor>
Interesting.

[10:17:01.0000] <AryehGregor>
I wonder if AMD knows what it's talking about.

[10:17:02.0000] <AryehGregor>
That looks like a blog post, not a press release.

[10:18:00.0000] <AryehGregor>
"The information presented in this document is for informational purposes only and may contain technical inaccuracies, omissions and typographical errors. Links to third party sites are for convenience only, and no endorsement is implied."

[10:18:01.0000] <paul_irish>
hah. true.

[10:18:02.0000] <AryehGregor>
Google isn't clear on whether Robin Maffeo works for Microsoft or AMD.

[10:24:00.0000] <JonathanNeal>
Any responses / opinions about article previews?

[10:25:00.0000] <JonathanNeal>
I'm hoping this question clears up future decisions I make about how to organize my content.

[10:39:00.0000] <Philip`>
AryehGregor: Or they've decided but don't want to announce it quite yet

[10:40:00.0000] <AryehGregor>
I guess.

[10:40:01.0000] <AryehGregor>
Maybe they don't want to announce plans for anything they haven't implemented, in case they have to scrap it.

[10:40:02.0000] <Philip`>
Need some extra features to keep people interested in future releases

[10:40:03.0000] <AryehGregor>
I expect that if they knew for sure they wouldn't implement it, they'd probably say so.

[10:40:04.0000] <AryehGregor>
They've said that for other things, like Acid3.

[10:41:00.0000] <Philip`>
They've announced <video> but not shipped it in the preview build

[10:41:01.0000] <Philip`>
so the current version is definitely not the complete set of features

[10:43:00.0000] <Philip`>
They'd probably never say no to anything, but they'd probably do what they did before IE8 and avoid the quesion by just saying they're going to implement things based on customer demand

[10:44:00.0000] <Philip`>
instead of the current approach of avoiding the question by not responding to it or (if they can't avoid it that way) saying to wait and see

[10:45:00.0000] <AryehGregor>
They demoed <video>.

[10:46:00.0000] <Philip`>
(It's fun trying to reach conclusions based on non-information, even if it's not a very accurate method)

[10:55:00.0000] <adactio>
I believe the plan is for Microsoft to release updates to the developer preview of IE9 at regular intervals. I suspect each release will include more updates.

[10:57:00.0000] <AryehGregor>
I'd assume that they wouldn't make a new release with no updates.

[10:57:01.0000] <AryehGregor>
It's every eight weeks, by the way.

[10:59:00.0000] <Philip`>
Roughly every eight

[11:03:00.0000] <Philip`>
(... i.e. they're not committing to a fixed schedule in the style of other browsers' nightly builds)

[11:04:00.0000] <Philip`>
(Not that that makes much difference, as long as you're not marking the dates on your calendar in advance)

[11:14:00.0000] <AryehGregor>
You mean you're not?  :(

[11:15:00.0000] <Philip`>
I don't have a calendar

[11:18:00.0000] <AryehGregor>
"For years, Viacom continuously and secretly uploaded its content to YouTube, even while publicly complaining about its presence there. It hired no fewer than 18 different marketing agencies to upload its content to the site. It deliberately "roughed up" the videos to make them look stolen or leaked. It opened YouTube accounts using phony email addresses. It even sent employees to Kinko's to upload clips from computers that couldn't be traced

[11:18:01.0000] <AryehGregor>
 to Viacom. And in an effort to promote its own shows, as a matter of company policy Viacom routinely left up clips from shows that had been uploaded to YouTube by ordinary users. Executives as high up as the president of Comedy Central and the head of MTV Networks felt "very strongly" that clips from shows like The Daily Show and The Colbert Report should remain on YouTube." http://youtube-global.blogspot.com/2010/03/broadcast-yourself.html

[11:18:02.0000] <AryehGregor>
Wow.

[11:18:03.0000] <AryehGregor>
Of course, it was posted by a lawyer and relates to an ongoing legal case, so possibly it's all lies, but still.

[11:39:00.0000] <AryehGregor>
Is there a standards-compliant way to tell whether a page has already finished loading from JavaScript, if the script might be executed after the load event has fired (e.g., from a bookmarklet)?

[11:49:00.0000] <AryehGregor>
I mean a standards-compliant way that's not completely hackish, needless to say.

[11:49:01.0000] <paul_irish>
document.readyState though it doesnt have great support.

[11:50:00.0000] <AryehGregor>
(apropos: http://tech.slashdot.org/comments.pl?sid=1588514&cid=31536576)

[11:57:00.0000] <AryehGregor>
Oh well, I guess he was right then.

[12:27:00.0000] <AryehGregor>
/me finds that when working through his spec mail backlog, he only gets to the really long threads later on

[13:48:00.0000] <zcorpan>
http://flashisdead.tumblr.com/

[13:51:00.0000] <Dashiva>
Google finance doesn't look like it uses any magical aspects of flash, but maybe I'm missing advanced features

[13:52:00.0000] <AryehGregor>
Analytics seems similar.  It looks like it would work fine with canvas and/or SVG.

[13:53:00.0000] <AryehGregor>
Chatroulette is fair until <device> works.

[13:53:01.0000] <AryehGregor>
Dunno about the others.

[13:53:02.0000] <Dashiva>
Homestar runner is fair enough, flash is made for animation. But it's not that it's impossible, there just aren't any tools to do it the HTML5 way

[13:53:03.0000] <MikeSmith>
"let's see your html5 do this" is a nice slogan

[13:53:04.0000] <Dashiva>
Sound mixing has been requested several times, so that's fair too

[13:54:00.0000] <Dashiva>
Image editors... seems like canvas could handle that fine

[13:56:00.0000] <zcorpan>
/me wants to see html5 do those things

[13:56:01.0000] <Dashiva>
Important difference between html5 can do x, and someone has actually done the work to create a html5 tool doing x

[13:56:02.0000] <AryehGregor>
Wait until IE < 9 has, let's say, under 5% market share.

[13:56:03.0000] <AryehGregor>
So, I don't know, 2026?

[13:56:04.0000] <zcorpan>
Dashiva: indeed

[13:57:00.0000] <Dashiva>
As long as flash has 98% or whatever, most people will be happy using existing tools

[13:57:01.0000] <Dashiva>
Or at least grudgingly accepting :P

[13:58:00.0000] <AryehGregor>
Not if they know HTML+CSS but not Flash.

[13:58:01.0000] <AryehGregor>
They have to learn HTML+CSS+JS anyway, so if they no longer have any reason to use Flash, they'll stick with a unified setup.

[13:58:02.0000] <AryehGregor>
That's our major advantage.

[13:58:03.0000] <Dashiva>
I mean the users

[13:58:04.0000] <Dashiva>
As long as there exist tools that solve their problems, there's reduced demand for creating equivalent html5 tools

[13:59:00.0000] <JonathanNeal>
If I want to say my page is in english in HTML5, where should I place the attributes that dictate this?

[13:59:01.0000] <AryehGregor>
It's the authors who will control the switch.  Users aren't going to know or care whether you use HTML5 or Flash, on the whole.

[13:59:02.0000] <AryehGregor>
Existing tools have to be maintained, maybe rewritten occasionally.

[13:59:03.0000] <Dashiva>
Authors respond to demand, though

[13:59:04.0000] <JonathanNeal>
<html lang="en-US"> ?  <meta http-equiv="content-language" content="en">

[13:59:05.0000] <AryehGregor>
They'll fade over time.

[13:59:06.0000] <JonathanNeal>
Both?

[13:59:07.0000] <AryehGregor>
Authors will adopt HTML5 over Flash because it makes their lives easier or improves the quality of their product, not because their users explicitly demand it.

[13:59:08.0000] <zcorpan>
JonathanNeal: just <html lang>

[13:59:09.0000] <AryehGregor>
JonathanNeal, <html lang> is the better bet.

[14:00:00.0000] <JonathanNeal>
So <html en-US></html> ?

[14:00:01.0000] <Dashiva>
Again, not direct demand

[14:00:02.0000] <JonathanNeal>
Oh!

[14:00:03.0000] <JonathanNeal>
<html lang="en-US">, got it.

[14:01:00.0000] <zcorpan>
yep

[14:03:00.0000] <JonathanNeal>
So <meta http-equiv="content-type" content="text/html; charset=utf-8"> became <meta charset="UTF-8">?  If so, does this only affect http-equiv="content-type", or any other http-equiv?

[14:04:00.0000] <Dashiva>
Only content-type

[14:05:00.0000] <Dashiva>
It's the result of supporting <meta http-equiv=content-type content=text/html; charset=utf-8>

[14:06:00.0000] <JonathanNeal>
So meta http-equiv="content-type" content="text/html; charset=utf-8" became meta charset="UTF-8" and meta http-equiv="content-language" content="en-US" became html lang="en-US" ... cool.

[14:06:01.0000] <Dashiva>
No, lang appeared on its own

[14:06:02.0000] <Dashiva>
The header content-language and @lang are different in subtle and confusing ways

[14:07:00.0000] <Dashiva>
(Otherwise we might've been saved the thread about it...)

[14:22:00.0000] <AryehGregor>
I've spent more than two hours reading all my spec mail.  :(

[14:22:01.0000] <AryehGregor>
But I'm finally done.

[14:37:00.0000] <JonathanNeal>
Question about writing out Element Attributes; should they always be arranged alphabetically?  href="@page-url@" title="@go-to@ @page-title@",  content="@site-author@" name="author", or do they have an arrangement written somewhere?

[14:39:00.0000] <AryehGregor>
What?

[14:39:01.0000] <JonathanNeal>
AryehGregor, what to me?

[14:39:02.0000] <AryehGregor>
Element attributes are logically unordered.  Use whatever order you like.

[14:40:00.0000] <AryehGregor>
I'm not sure what your question is, exactly.

[14:40:01.0000] <JonathanNeal>
Okay, you answered my  question, I didn't know if I'm supposed to write id="myId" class="myclassname" or if class="myclassname" id="myId" was just as correct, sorting alphabetically.

[14:43:00.0000] <JonathanNeal>
Thanks, I was wondering if there was a logical order.

[14:43:01.0000] <JonathanNeal>
Because if element attributes are logically unordered, we could order them alphabetically, like we do with CSS properties.

[14:43:02.0000] <Philip`>
That doesn't sound logical

[14:43:03.0000] <Philip`>
You'd end up writing <a class="foo" href="bar"> etc

[14:44:00.0000] <Philip`>
when clearly it should be <a href...>

[14:44:01.0000] <AryehGregor>
Any order is fine.

[14:45:00.0000] <miketaylr>
i never alphabetize my css

[14:46:00.0000] <Dashiva>
The only CSS that's alphabetical is the indexes :P

[14:46:01.0000] <miketaylr>
:)

[14:48:00.0000] <Philip`>
I line up my DVDs in order of increasing height of the age rating icon on the spine, but even I don't alphabetise CSS

[14:49:00.0000] <JonathanNeal>
What's up with meta content="IE=8;chrome=1" http-equiv="X-UA-Compatible" not validating?

[14:49:01.0000] <AryehGregor>
X-UA-Compatible is a non-standard extension.

[14:49:02.0000] <JonathanNeal>
You don't alphabetize your css properties?  Huh.

[14:49:03.0000] <AryehGregor>
So it's not supposed to validate as standard.

[14:49:04.0000] <AryehGregor>
Also, will that value actually work?

[14:50:00.0000] <AryehGregor>
It seems like you could just leave out the IE part, unless you specifically don't want your site to work in IE9 for some reason.

[14:51:00.0000] <JonathanNeal>
Well, no the issue is that IE runs in IE7 mode on intranets, so I just learned from paul_irish

[14:51:01.0000] <AryehGregor>
Oh, right.

[14:51:02.0000] <AryehGregor>
That's true.

[14:51:03.0000] <AryehGregor>
If it's an intranet, you could just reconfigure them all, couldn't you?

[14:51:04.0000] <AryehGregor>
I'm told that's easy on Windows intranets.

[14:52:00.0000] <JonathanNeal>
Sure.

[14:53:00.0000] <JonathanNeal>
So, writing <body><div id="document" role="document"></div></body> is the worst idea ever and standards-breaking, or valid and at my descretion?

[14:57:00.0000] <AryehGregor>
It seems to be technically valid as far as HTML5 is concerned, unless I missed the part that bans it.

[14:57:01.0000] <AryehGregor>
It seems pointless, though.

[14:57:02.0000] <AryehGregor>
<body> has implicit document semantics.

[14:59:00.0000] <JonathanNeal>
Right, but I think this could be useful when you're in a modx, Facebook, or another portal based page where you have a dock or a chat application that sits aside the div#document.

[14:59:01.0000] <Philip`>
JonathanNeal: If you use x-ua-compatible, you should use "IE=edge" so that it always uses the latest version

[15:00:00.0000] <AryehGregor>
Are you testing the effect that role is having here in an actual UA that supports role, or just speculatively adding markup in case maybe for someone it would be better than not having the markup?

[15:00:01.0000] <Philip`>
(so that you benefit from improved performance and standardiness in future versions)

[15:01:00.0000] <paul_irish>
Philip`: yah i was thinking the same.

[15:02:00.0000] <JonathanNeal>
Philip`, good call.

[15:02:01.0000] <Dashiva>
Or so you contribute to IE=edge becoming frozen as IE8 because too many people use it and depend on quirks~

[15:02:02.0000] <JonathanNeal>
AryehGregor, both.

[15:03:00.0000] <AryehGregor>
JonathanNeal, don't do the latter, ever.  Do not add features you can't test.  They'll probably make things worse.  UAs can deal with you ignoring their existence, they can't deal with you giving them explicit instructions that turn out to make no sense because you didn't test them.

[15:03:01.0000] <AryehGregor>
I've learned this the hard way.

[15:03:02.0000] <JonathanNeal>
Dashiva, I have corrected this.

[15:03:03.0000] <JonathanNeal>
AryehGregor, what tools do you recommend I use to test these?

[15:05:00.0000] <AryehGregor>
JonathanNeal, I don't know.  But if you aren't testing, don't use the feature, period.  You *will* do it wrong and your site *will* break when someone finally does implement the feature.  Guaranteed.

[15:05:01.0000] <Philip`>
Dashiva: He should use IE=edge and not depend on quirks

[15:05:02.0000] <JonathanNeal>
I take the red pill, AryehGregor.

[15:05:03.0000] <Philip`>
and should test in various browsers and in IE betas

[15:06:00.0000] <AryehGregor>
JonathanNeal, what?

[15:06:01.0000] <JonathanNeal>
AryehGregor, http://izanbardprince.files.wordpress.com/2008/08/red-pill-or-blue-pill.jpg

[15:06:02.0000] <Dashiva>
Philip`: Yeah. The problem is that if enough other people don't, his site will still be affected.

[15:07:00.0000] <Philip`>
Dashiva: He'll be affected less than if he uses IE=8

[15:07:01.0000] <Dashiva>
That's true

[15:08:00.0000] <AryehGregor>
JonathanNeal, I've seen The Matrix.  I have no idea what you mean in context.

[15:09:00.0000] <JonathanNeal>
AryehGregor, I mean that I'm risky and I wanna see what happens --- but really if no tool exists to evaluate this then I can theorize.

[15:10:00.0000] <JonathanNeal>
The spec says I can, and it makes sense to me this way, so the only reason I could provide against it would be that someone implements it differently.

[15:10:01.0000] <AryehGregor>
No, that is an absolutely terrible idea.  Drop it immediately.  You're doing a disservice to yourself and your users and the web as a whole.  Deploying a feature that hasn't yet been implemented is almost guaranteed to break, and that either 1) hurts your users if their site breaks, or (worse) 2) hurts browsers because they get complaints from users that the site breaks in the new version, since the new version implements the feature you're u

[15:10:02.0000] <AryehGregor>
sing brokenly.

[15:11:00.0000] <AryehGregor>
What will happen is that either you will misunderstand the spec somehow, or else the implementation will have some quirk you weren't expecting.  I have had this happen to me *multiple* times.

[15:11:01.0000] <AryehGregor>
When I learned that Opera supposed datalist, I added support to MediaWiki and tested in Opera -- only to find out that actually there was a slight quirk I never would have predicted that made it behave worse than the non-datalist functionality.

[15:12:00.0000] <AryehGregor>
When I learned that Opera supported attributes like pattern on inputs, I added pattern to lots of inputs, and tested.  It turned out that if a password field fails a constraint, it will print the value of the password field in plaintext.

[15:13:00.0000] <AryehGregor>
In both cases, if I had blindly added the feature without testing it in any actual user-agent, it would have caused significant *harm* to users when Opera added the feature.

[15:13:01.0000] <JonathanNeal>
You could test, what do I have to test?  JAWS?

[15:14:00.0000] <AryehGregor>
In another case, when I was young and stupid and didn't yet realize the absolute necessity of testing, I added a handheld stylesheet to MediaWiki without having any mobile UA handy, figuring something was better than nothing.  This actually broke Opera Mobile (or Mini or something), to the point that they went to the effort to dig up wikitech-l and ask there for us to remove our handheld stylesheet because it was broken.

[15:14:01.0000] <AryehGregor>
If you care about blind users, then yes, absolutely test in JAWS.

[15:14:02.0000] <JonathanNeal>
Just because I'm pushing back doesn't mean I'm disregarding what you're saying, just developing with you.

[15:14:03.0000] <AryehGregor>
Add whatever features demonstrably make your users' experience better.

[15:15:00.0000] <AryehGregor>
Do not add features that don't do anything and might or might not do something in the future just because you can.

[15:16:00.0000] <JonathanNeal>
Right, change for change sake.  I'm trying to take advantage of meaning to enforce existing standards.  For instance, most folks need a <div id="wrapper" /> or something like that on their site.  Something that wraps all the pages content and separates it from another application on the page.

[15:16:01.0000] <JonathanNeal>
And that lead me to look into the role document and application, which seemed to fit perfectly in this paradigm, and describes itself as such.

[15:16:02.0000] <JonathanNeal>
And goes as far as to say this scope doesn't apply to the whole page, but part of the page.

[15:17:00.0000] <JonathanNeal>
So I thought, let's take this great thing and mix it in with the existing need for a wrapper, consolidating the two.

[15:17:01.0000] <AryehGregor>
Semantics for semantics' sake is just as bad.

[15:18:00.0000] <AryehGregor>
Keep this in mind: if you use role nowhere at all, your page will be like 99.9% of pages out there.  Screen readers for the blind will be able to handle that.

[15:18:01.0000] <AryehGregor>
So it will work fine.

[15:18:02.0000] <JonathanNeal>
What about consolidation by means of semantics?

[15:18:03.0000] <AryehGregor>
If you use role *incorrectly*, your site will work *worse*.  It might become unusable, even.

[15:19:00.0000] <AryehGregor>
The gains are not worth the risk from any kind of practical standpoint.

[15:19:01.0000] <AryehGregor>
JonathanNeal, that's a string of buzzwords, not an actual reason.

[15:19:02.0000] <AryehGregor>
An example of a reason to use role="something or other" would be "so JAWS does X instead of Y".

[15:19:03.0000] <AryehGregor>
Where X is better than Y.

[15:19:04.0000] <JonathanNeal>
As we get to know each other more, you'll learn I'm not a buzzword guy.

[15:22:00.0000] <JonathanNeal>
I think there's a balance, and sometimes you have to edge past the 99.9% paradigm, get the test results, learn, and develop.  Then progress is made.  And it looks like I have good counsel in this room to remind me how dangerous it can be to force users to adopt my suggestions :)

[15:25:00.0000] <AryehGregor>
I've added a whole lot of HTML5 features to MediaWiki (thus to eventually be deployed on Wikipedia), and I have always followed this maxim: only add something if I can clearly spell out a user-visible benefit.

[15:27:00.0000] <AryehGregor>
/me remembers he needs to work out something about validation of HTML5 with non-HTML5 doctype, ugh

[15:30:00.0000] <othermaciej>
Hixie: are you around?

[15:31:00.0000] <JonathanNeal>
I like that AryehGregor, and thanks for taking the time to chat about it with me.

[15:31:01.0000] <AryehGregor>
You can look at <http://www.mediawiki.org/wiki/HTML5> for a list of features we (I) have implemented.

[15:31:02.0000] <AryehGregor>
I have to go now, though.

[15:32:00.0000] <othermaciej>
AryehGregor: neat

[15:32:01.0000] <AryehGregor>
othermaciej, it might be enabled on Wikipedia soon, but I've been saying that for like six months.  :P

[15:33:00.0000] <zcorpan>
AryehGregor: just use http://html5.validator.nu/

[15:33:01.0000] <AryehGregor>
Code review has been very slow lately, and since I'm not a sysadmin I can't sneak in my changes without a full deployment, but 1.16rc1 has been released, and a deployment of 1.16 to Wikimedia is supposed to follow before 1.16 final.

[15:33:02.0000] <AryehGregor>
zcorpan, I can, but other people use validator.w3.org.

[15:35:00.0000] <zcorpan>
if (ua == 'validator') echo '<!doctype html><title></title>'; else { ...

[15:35:01.0000] <JonathanNeal>
AryehGregor, yea the long term elements will be an interesting piece of tackle, since many yesterday mobile browsers won't work with html5 elements, even with the "shiv" javascript hack.

[15:35:02.0000] <JonathanNeal>
I look forward to seeing how that goes, have fun.

[15:36:00.0000] <AryehGregor>
zcorpan, . . . that is so evil that I never even thought of it, but it just might work.

[15:36:01.0000] <AryehGregor>
/me will keep it in mind.

[15:37:00.0000] <AryehGregor>
Won't work on Wikipedia, though, Squid caching.

[15:55:00.0000] <othermaciej>
is removing <head> conforming in XHTML5?

[15:56:00.0000] <othermaciej>
hmm I see, it's not trying to produce XHTML5


2010-03-20
[17:30:00.0000] <othermaciej>
wow: http://intertwingly.net/blog/2010/03/17/Non-Draconian-XHTML

[17:31:00.0000] <miketaylr>
wow indeed.

[17:37:00.0000] <Dashiva>
Well, as mentioned in the comments, it's not actually non-draconian, it's non-error-message-displaying

[17:39:00.0000] <othermaciej>
I guess that is a lesser failing

[17:59:00.0000] <Dashiva>
http://techdirt.com/articles/20100318/1240568623.shtml

[17:59:01.0000] <Dashiva>
I guess software patents aren't the only problem anymore...

[18:07:00.0000] <TabAtkins>
Sigh.  Nobody understands patents.

[18:07:01.0000] <TabAtkins>
I mean, okay, I can understand the jury's ruling.

[18:08:00.0000] <TabAtkins>
But a government-required technology *can not* be patent-encumbered.

[18:08:01.0000] <TabAtkins>
ARgha;sd;asldkjf;ads

[18:10:00.0000] <TabAtkins>
Though, wait, further details are hitting my brain.

[18:10:01.0000] <TabAtkins>
No, the jury is a group of unwashed illiterates.

[18:10:02.0000] <TabAtkins>
You can't sue a company for not putting in a non-required safety feature.

[18:15:00.0000] <othermaciej>
you can sue someone for negligence if you can make a case that they did not take adequate safety precautions

[18:15:01.0000] <othermaciej>
it doesn't matter if the safety precautions were legally mandated

[18:16:00.0000] <othermaciej>
I can see how this may be an odd result, but it's not because the safety feature was non-mandatory

[21:48:00.0000] <othermaciej>
so, only 4 of the Alexa worldwide top 20 validate as their declared doctype

[21:59:00.0000] <JonathanNeal>
Heyyo!

[22:04:00.0000] <othermaciej>
hmm

[22:05:00.0000] <othermaciej>
two of the Alexa top 50 are unreachable sites

[22:05:01.0000] <othermaciej>
I am starting to doube the accuracy of these rankings

[00:38:00.0000] <othermaciej>
is type="text/css" necessary or useful in HTML5?

[00:38:01.0000] <othermaciej>
(on <link rel="stylesheet">)

[00:45:00.0000] <JonathanNeal>
Heyo

[03:03:00.0000] <Philip`>
othermaciej: No

[03:03:01.0000] <Philip`>
"If the attribute is omitted, but the external resource link type has a default type defined, then the user agent must assume that the resource is of that type."

[03:03:02.0000] <Philip`>
"The default type for resources given by the stylesheet keyword is text/css."

[03:04:00.0000] <Philip`>
so it should make no difference

[03:29:00.0000] <scotfl>
   \

[06:32:00.0000] <boblet>
hey all. regarding the old acronym/abbr confusion, I thought they were to indicate an acronym (abbreviation read as a word like NATO) and an initialism (abbreviation read as individual letters like HTML) to screen readers

[06:32:01.0000] <boblet>
does anyone know if screen readers actually did that, or if they used some other method, like a built-in dictionary of mappings?

[07:04:00.0000] <Dashiva>
boblet: They were just different tags for different browsers

[07:04:01.0000] <Dashiva>
There was no semantic difference

[07:05:00.0000] <boblet>
Dashiva: aah they came out of NN vs IE then?

[07:05:01.0000] <boblet>
makes sense

[07:05:02.0000] <Dashiva>
Yeah

[07:06:00.0000] <lazni>
google translates "diveintomark" as if it was "dive into bookmark"

[07:25:00.0000] <Dashiva>
Did anyone seriously believe microsoft would not use h264 for IE9?

[07:29:00.0000] <Dashiva>
"We, the users, didn't decide to download Flash - we were forced to"

[07:29:01.0000] <Dashiva>
Because Flash obviously had 98% market penetration from day one

[07:32:00.0000] <boblet>
Dashiva: I think there were some hopes that they’d offer an alternative codec too, but those hopes were … hmm

[07:39:00.0000] <Dashiva>
boblet: naive? foolish? deluded?

[07:42:00.0000] <boblet>
I was thinking something less blunt… optimistic?

[07:43:00.0000] <boblet>
how do you say naive but with a positive meaning?

[07:59:00.0000] <Philip`>
Dashiva: Some people may have expected they'd use WMV instead

[08:00:00.0000] <lazni>
/me would pick directshow

[08:00:01.0000] <Philip`>
boblet: Innocent?

[08:01:00.0000] <Philip`>
/me can't think of a better word

[08:01:01.0000] <boblet>
Philip`: yeah but that has connotations too

[08:01:02.0000] <boblet>
hopeful/optimistic is as good as it gets huh

[08:01:03.0000] <Philip`>
All words have connotations :-)

[08:01:04.0000] <boblet>
does put Moz in a bit of a spot

[08:02:00.0000] <Dashiva>
I suppose you could say they 'have faith'

[08:02:01.0000] <boblet>
“I want to believe!”

[08:02:02.0000] <lazni>
http://www.onelook.com/?w=naive+without+connotation&ls=a

[08:57:00.0000] <JonathanNeal>
Mornin'!

[09:00:00.0000] <paul_irish>
'afternoon.

[09:01:00.0000] <JonathanNeal>
o'h

[09:06:00.0000] <JonathanNeal>
I've been trying to think of the label for the main document content on a page that sits after the dock bar and before the chat bar

[09:09:00.0000] <JonathanNeal>
Microsoft puts their entire site in a form field.

[09:10:00.0000] <JonathanNeal>
Yahoo doesn't wrap all their content.

[09:21:00.0000] <JonathanNeal>
Analyzing some more... No wrapping divs for Drupal, ModX, or Wordpress.  Joomla uses "page-holder", BBC uses "blq-main".  Foxnews and msn use "wrapper", latimes, cssZen, Digg, and twitter use "container"

[09:21:01.0000] <JonathanNeal>
YUI uses "yui-t7" (oh very descriptive, YUI), Facebook uses "globalContainer", Alexa uses "pageContainer", MySpace uses "wrap", Google uses "main", YouTube kinda uses "baseDiv", Blogger kinda uses "body", Bing uses a TABLE "layout", Wikipedia uses "globalWrapper", 

[09:30:00.0000] <JonathanNeal>
456 Berea Street uses wrap, Blogger uses "home", ebay and cnn have a <div align="center">, Go has a <table align="center"> and an annoying ad, Disney has "flashcontent" and a more annoying ad, Apple has "home", Photobucket has "centeredPage", Mozilla kinda has "wrapper", Adobe kinda has "home-fma", Amazon kinda but not really has "page-wrap", PirateBay uses "browseContainer".  Wow, no standard.

[09:37:00.0000] <JonathanNeal>
WhatWG nothin', w3c "w3c_container", jQuery "jq-siteContain", Party Poker "mainbg", Download "rb_bodyWrap", Zedo (a table), Netflix "headlessshadowwrapper", Huffington "wrapper", HP "centerContent", metacafe "PageWrapper", Vimeo "outer_wrap" (and then "inner_wrap" (and then "everything")), PaulIrish "wrap" --- and that covers all the major sites on the internet :)

[10:45:00.0000] <JonathanNeal>
A long, boring list of the first DIV's @ http://sandbox.thewikies.com/first-div/

[11:36:00.0000] <paul_irish>
AryehGregor: based on yesterdays discussion of implementing role=document and such.... would you say that   <meta http-equiv="X-UA-Compatible" content="IE=edge;chrome=1" > would be a bad idea because things could break in ie9 vs locking in the ie8 rendering mode?

[12:43:00.0000] <krijnh>
MikeSmith: a sewing machine? ;)

[12:43:01.0000] <lazni>
http://www.openclipart.org/image/800px/svg_to_png/GMcGlinn_cloud_Rainbow.png does this logo work for "free spectrum"?

[12:57:00.0000] <Philip`>
Is there a known problem with sending content-encoding:gzip .oga files to Chrome?

[12:57:01.0000] <Philip`>
/me appears to get no playback unless he disables compression

[12:57:02.0000] <krijn>
MikeSmith: I think my hdd is about to die or something

[13:24:00.0000] <krijn>
/me starts implementing some sort of cache for /irc-logs/

[13:42:00.0000] <roc>
sending gzipped .oga is a bad idea in general

[13:42:01.0000] <roc>
I expect you won't get useful extra compression, everyone will do extra work, and you won't be able to seek

[13:44:00.0000] <Philip`>
I'm just surprised that it seems to make it stop working entirely

[13:44:01.0000] <Philip`>
whereas gzipped MP3s work fine

[13:44:02.0000] <Philip`>
in Chrome

[13:44:03.0000] <roc>
sure, that's probably a bug

[13:46:00.0000] <roc>
can't say I blame them; if we have such a bug in Firefox, I wouldn't prioritize it

[13:52:00.0000] <krijn>
/me has a thereifixedit.com moment with the irc logs

[13:54:00.0000] <krijn>
For those who want to know, URLs from 2006-2009 are now cached once they are loaded, which speeds up requests afterwards

[13:54:01.0000] <krijn>
http://krijnhoetmer.nl/irc-logs/whatwg/20090331 for example

[13:54:02.0000] <krijn>
Flagging of lines is disabled for cached files

[13:56:00.0000] <annevk>
also highlighting?

[13:56:01.0000] <krijn>
Flagging is highlighting

[13:56:02.0000] <krijn>
Flagged lines are still flagged

[13:56:03.0000] <annevk>
cookie-based highlighting

[13:57:00.0000] <krijn>
You just can't flag any new lines

[13:57:01.0000] <krijn>
Ow, crap

[13:57:02.0000] <krijn>
Good point

[13:57:03.0000] <annevk>
might make sense to implement that client-side regardless

[13:57:04.0000] <krijn>
Indeed

[13:57:05.0000] <krijn>
Now the cached files include who requested it, nice

[13:57:06.0000] <krijn>
<-- stupid

[13:58:00.0000] <krijn>
Is the nick highlighting thing even useful in older files?

[13:59:00.0000] <annevk>
why wouldn't it be?

[13:59:01.0000] <krijn>
I don't know, I never use it :)

[14:00:00.0000] <lazni>
/me suggests scrapping it altogether; people with big egos can use greasemonkey

[14:00:01.0000] <krijn>
Or just a user stylesheet

[14:01:00.0000] <krijn>
The only thing the cookie thing does is add some extra css

[14:01:01.0000] <lazni>
how would USS work without bloating the page with classes?

[14:01:02.0000] <lazni>
nvm, offtopic from caching

[14:02:00.0000] <krijn>
Ah, it's not only CSS

[14:05:00.0000] <annevk>
krijn, it highlights lines of people addressing you; it's useful when you weren't available or connected

[14:05:01.0000] <krijn>
Ow, wow, the nicks database contains 16000 records

[14:05:02.0000] <krijn>
Fu spammers!

[14:06:00.0000] <annevk>
nog een rede voor client-side :)

[14:06:01.0000] <annevk>
je zou met localStorage kunnen gaan spelen

[14:06:02.0000] <annevk>
kan ik ook voor html5.org gaan doen

[14:07:00.0000] <krijn>
Of het probleem verplaatsen

[14:07:01.0000] <krijn>
Waardoor het een heel stuk simpeler wordt

[14:07:02.0000] <krijn>
nicks db droppen

[14:07:03.0000] <krijn>
<li id="l-29" _a="Hixie" _r="christ"><a href="#l-29">#</a> [00:21] &lt;Hixie&gt; christ, i have a two-line testcase for mutation events and three browsers and they ALL DO SOMETHING DIFFERENT <span> </span></li>

[14:07:04.0000] <krijn>
lol

[14:08:00.0000] <krijn>
<li id="l-54" _a="annevk3" _r="true"><a href="#l-54">#</a> [00:57] &lt;annevk3&gt; true, I rather not bug him about ES gossip though :) <span> </span></li>

[14:08:01.0000] <krijn>
Okay, not that great

[14:08:02.0000] <krijn>
But great enough

[14:08:03.0000] <annevk>
dat kan ook nog, maar is het niet simpeler om zoiets met JS te doen?

[14:10:00.0000] <krijn>
Dat kan nu dus

[14:10:01.0000] <krijn>
Of ja, kan sowieso

[14:10:02.0000] <krijn>
Nu kan een user stylesheet iig ook

[14:10:03.0000] <lazni>
what's _r?

[14:11:00.0000] <krijn>
Reply to

[14:11:01.0000] <krijn>
Or something

[14:11:02.0000] <krijn>
_a is Author

[14:16:00.0000] <krijn>
Zo dan

[14:16:01.0000] <krijn>
I mean, There, I fixed it

[14:19:00.0000] <lazni>
does the nick highlight instruction have to be so urgent-looking?

[14:19:01.0000] <krijn>
Nope, removed

[14:22:00.0000] <lazni>
there's no CSS like text-transform: sentence-case, eh?

[14:28:00.0000] <krijn>
Hmm, only one problem

[14:28:01.0000] <krijn>
If people load an old log, it is immediately cached

[14:29:00.0000] <krijn>
So the flagging of lines doesn't work anyway..

[14:29:01.0000] <lazni>
krijn: s/Options/Highlight/

[14:29:02.0000] <lazni>
I thought you don't want to change old line flagging

[14:32:00.0000] <krijn>
Fixed

[14:32:01.0000] <krijn>
Nog flagging possibilities in 2009 and earlier

[14:32:02.0000] <krijn>
What a shame

[14:33:00.0000] <krijn>
*No

[14:33:01.0000] <Dashiva>
If each class had an id, the flagging could be done purely css without having to change the cached html

[14:33:02.0000] <Dashiva>
*each line

[14:33:03.0000] <krijn>
And then constantly include a new CSS file?

[14:33:04.0000] <krijn>
Hmm

[14:34:00.0000] <krijn>
There are 15000 flagged lines.. You guys have to start typing less interesting stuff!

[14:36:00.0000] <Dashiva>
Is it still running on your sewing machine?

[14:36:01.0000] <krijn>
Yes

[14:36:02.0000] <krijn>
It is in a clothing shop, so that does make sense :)

[14:40:00.0000] <Dashiva>
Maybe we could upgrade to a 286 or something

[14:40:01.0000] <JonathanNeal>
Please stand by for a few moments. Our database system is currently scaling up to meet the demand of a recent surge in traffic. The site should be operational again in a few minutes.

[14:40:02.0000] <JonathanNeal>
"No, we're not down, we're just .... we're SCALING YES THAT'S IT

[15:26:00.0000] <MikeSmith>
krijn: well, I figured it either had to be a sewing machine or Nintendo DS

[15:26:01.0000] <krijn>
Heh

[15:27:00.0000] <krijn>
I hope it's a bit better now

[15:39:00.0000] <MikeSmith>
krijn: yeah, it does definitely seem better

[15:39:01.0000] <krijn>
The index is still pretty slow

[15:41:00.0000] <lazni>
http://tsung.erlang-projects.org/ you can stress test it to get numbers

[15:41:01.0000] <MikeSmith>
I don't usually use the index

[16:26:00.0000] <lazni>
the  Internet will be, err, transparent.  Which, I like to think, is kind of  fitting for the  information network that interprets censorship as damage and routes  around it… http://www.daniweb.com/news/post972308.html

[16:36:00.0000] <Dashiva>
I wonder if what the correlation (postive, negative, insignificant) is between people who support D.E. and people who think the no-standards wild-west internet of the past was a good idea

[16:36:01.0000] <Dashiva>
-if

[16:38:00.0000] <lazni>
DE?

[16:38:01.0000] <Dashiva>
Distributed extensibility

[16:55:00.0000] <othermaciej>
Dashiva: you'd have to consider the control group for that experiment

[16:55:01.0000] <othermaciej>
is any person who is reasonably aware of standards likely to hold the view that "the no-standards wild-west internet of the past was a good idea"?

[16:56:00.0000] <Dashiva>
I would think the answer is 'no'

[16:57:00.0000] <Dashiva>
But that was also a perfect example of distributed extensibility


2010-03-21
[17:09:00.0000] <TabAtkins>
The "what's the language of the document" thread is the most boring thread in existence.  Even worse than the process threads.

[17:27:00.0000] <Dashiva>
I wonder, the systems that use http-equiv content-language to assign metadata for use by the http server, how do they handle the language for non-html documents?

[17:34:00.0000] <AryehGregor>
paul_irish, if your page works in all major browsers including IE8 without browser-specific hacks, I think IE=edge is likely to be the most useful, because the next version of IE will probably implement more good features than bad.  Moreover, the implications of X-UA-Compatible are rather evil, so I'd prefer to avoid it on principle if possible.

[17:37:00.0000] <AryehGregor>
/me ORLYs at http://www.mediawiki.org/w/index.php?title=HTML5&diff=0&oldid=311452

[17:37:01.0000] <AryehGregor>
/me tests

[17:40:00.0000] <Dashiva>
I can see how it'd make the compression less efficient, but I can't see it leading to an increase in size compared to uncompressed

[17:42:00.0000] <Dashiva>
"Also note Manual:$wgValidateAllHtml, but is tidy ready for HTML5?" -- no

[17:47:00.0000] <AryehGregor>
It means that it would end up being larger than the version that uses quotes consistently.

[17:50:00.0000] <AryehGregor>
/me found not, although the difference is too trivial to matter by itself

[17:51:00.0000] <AryehGregor>
/me is wondering if he should move weird quoting to its own config option for people who don't like the aesthetics.

[18:08:00.0000] <Dashiva>
AryehGregor: If you through some magical incantation managed to start using a model to construct the page instead of string concatenation, it would almost solve itself

[18:09:00.0000] <AryehGregor>
What would almost solve itself?

[18:09:01.0000] <Dashiva>
The quoting

[18:09:02.0000] <Dashiva>
It would just be a parameter to the serialization

[18:10:00.0000] <AryehGregor>
It already is a parameter to the serialization, it's just currently rolled into $wgWellFormedXml.

[18:10:01.0000] <AryehGregor>
Search for "$quote": http://svn.wikimedia.org/viewvc/mediawiki/trunk/phase3/includes/Html.php?view=markup

[18:11:00.0000] <AryehGregor>
Of course, that method isn't called everywhere we output an element.

[18:11:01.0000] <AryehGregor>
Partly because that would be too slow.  Some places are pretty performance-sensitive for this sort of thing, I'm told.

[18:12:00.0000] <Dashiva>
Are there examples of such places?

[18:12:01.0000] <AryehGregor>
I've been told that if I make Sanitizer.php use Html::element(), I'll have my commit reverted, because extra function calls make a difference there.

[18:13:00.0000] <AryehGregor>
That's what ends up outputting stuff for the actual wikitext.

[18:13:01.0000] <Dashiva>
Isn't that... the majority of stuff?

[18:14:00.0000] <AryehGregor>
Yep!

[18:14:01.0000] <AryehGregor>
/me notices only 60 bytes saving after gzip on a large article for disabling $wgWellFormedXml, which is 0.6%.

[18:15:00.0000] <AryehGregor>
Even pre-gzip it's only 121 bytes.

[18:15:01.0000] <AryehGregor>
Er, no.

[18:15:02.0000] <AryehGregor>
It's 210 bytes pre-gzip.

[18:15:03.0000] <AryehGregor>
0.5%.

[18:16:00.0000] <othermaciej>
TabAtkins: at lest it's relevant to the content of the spec

[18:16:01.0000] <AryehGregor>
/me wonders whether there's any actual point in making $wgWellFormedXml = false do anything other than change the doctype to <!doctype html>.

[18:18:00.0000] <Dashiva>
AryehGregor: Well, does it affect the skin?

[18:18:01.0000] <AryehGregor>
Parts of it.

[18:18:02.0000] <TabAtkins>
othermaciej: Glad to hear it.  I try to start reading the thread and my brain shuts down out of self-defense.

[18:18:03.0000] <AryehGregor>
wiki.whatwg.org is using $wgWellFormedXml = false, so you can look there to see what's covered.

[18:18:04.0000] <AryehGregor>
It's not much.

[18:19:00.0000] <AryehGregor>
Most of the <head>.

[18:19:01.0000] <othermaciej>
well, the impact on the spec is off in the long tail where I don't care much about the specific outcome

[18:19:02.0000] <AryehGregor>
Basically nothing in the <body>

[18:20:00.0000] <AryehGregor>
.

[18:38:00.0000] <AryehGregor>
Is there a way in WebKit's Web Inspector to see styles that apply to pseudo-elements like ::-webkit-placeholder?

[18:40:00.0000] <AryehGregor>
Doesn't matter, I guess.

[18:40:01.0000] <AryehGregor>
/me struggles to find a color that light gray looks bad against.

[18:41:00.0000] <AryehGregor>
/me concludes "an almost exactly matching color of light gray" works

[18:41:01.0000] <othermaciej>
I don't think so, but that would be a good improvement

[18:49:00.0000] <JoePeck>
AryehGregor: write up a bug report at http://webkit.org/new-inspector-bug

[18:49:01.0000] <JoePeck>
AryehGregor: pfeldman (in #webkit-inspector) has been doing a lot of work recently with CSS Styles

[18:52:00.0000] <AryehGregor>
JoePeck, https://bugs.webkit.org/show_bug.cgi?id=36421

[18:52:01.0000] <JoePeck>
AryehGregor: thanks!

[18:54:00.0000] <JoePeck>
AryehGregor: for what its worth, the user agent style is input::-webkit-input-placeholder { color: darkGray; }

[18:54:01.0000] <JoePeck>
not light gray?

[18:54:02.0000] <AryehGregor>
It's something like #bbb, so that seems right.

[18:54:03.0000] <AryehGregor>
CSS3 says darkgray is #a9a9a9.

[19:11:00.0000] <othermaciej>
How do you do nested lists in MediaWiki syntax>

[19:12:00.0000] <othermaciej>
I see, multiple *s

[21:33:00.0000] <AryehGregor>
Does anyone have stats on what doctypes are used?  In particular, I'd expect XHTML 1.1 to be less common than XHTML 1.0 Strict, right?

[22:29:00.0000] <AryehGregor>
List of top 100 Alexa websites that validate according to the W3C: live.com, wikipedia.org, msn.com, wordpress.com, bbc.co.uk, mozilla.com, files.wordpress.com, thepiratebay.org.

[22:29:01.0000] <AryehGregor>
8/98 checkable, it seems.

[22:29:02.0000] <AryehGregor>
I'm only checking the index page, of course.

[22:29:03.0000] <AryehGregor>
That's somewhat higher than I expected.

[22:30:00.0000] <AryehGregor>
/me notes two Microsoft sites validate, while all Google sites fail miserably.

[22:31:00.0000] <AryehGregor>
(The two uncheckable are 1e100.net and bp.blogspot.com, which don't actually resolve.)

[22:32:00.0000] <AryehGregor>
The next two fail, so I'll just say 8% of the top 100 Alexa domain names that actually resolve.

[22:34:00.0000] <othermaciej>
AryehGregor: that's what I found two

[22:34:01.0000] <AryehGregor>
s/two/too/?

[22:34:02.0000] <othermaciej>
yes

[22:35:00.0000] <othermaciej>
"too"

[22:35:01.0000] <AryehGregor>
And "that" meaning everything I said, or some particular part of it?

[22:35:02.0000] <AryehGregor>
Validation is much more common among the top 10 sites than among the top 100.

[22:35:03.0000] <AryehGregor>
I imagine it gets less common as you go down, really.

[22:35:04.0000] <othermaciej>
I started this page: http://www.w3.org/html/wg/wiki/ConformanceErrorStudy

[22:35:05.0000] <AryehGregor>
Ah, neat.

[22:36:00.0000] <othermaciej>
any data you can add would be appreciated

[22:36:01.0000] <AryehGregor>
I think Google doesn't use an HTML5 doctype, so much as independently choosing the shortest possible doctype that will avoid quirks mode (without regard to standards).

[22:36:02.0000] <AryehGregor>
How did you test?

[22:36:03.0000] <othermaciej>
I think classifying the errors and filing bugs on any that seem bogus is more productive than what Sam is doing

[22:36:04.0000] <AryehGregor>
What's Sam doing?

[22:36:05.0000] <othermaciej>
so instead of arguing with him I have started ding it myself

[22:36:06.0000] <othermaciej>
help would be appreicated

[22:36:07.0000] <othermaciej>
his weird "remove or replace author conformance requirements" bug

[22:36:08.0000] <AryehGregor>
And where are you filing bugs?  On what basis?

[22:37:00.0000] <othermaciej>
where he refuses to get specific on what's wrong

[22:37:01.0000] <othermaciej>
I have not yet filed any bugs

[22:37:02.0000] <othermaciej>
my plan is to file some bugs against the spec in W3C bugzilla

[22:37:03.0000] <AryehGregor>
E.g., do you think we should propose that some common presentational attributes be made conforming in the spec?

[22:37:04.0000] <othermaciej>
that ask for removal or change of specific conformance requirements, if there are any that seem to be more noisy than beneficial

[22:38:00.0000] <othermaciej>
I haven't yet concluded whether to file a bug on presentational elements and attributes

[22:38:01.0000] <othermaciej>
my personal feeling is that there is not much benefit to making them nonconforming

[22:38:02.0000] <othermaciej>
AryehGregor: by the way, I do believe that google.com very deliberately chose <!DOCTYPE html> to match HTML5

[22:38:03.0000] <othermaciej>
it used to have no doctype at all

[22:38:04.0000] <AryehGregor>
I have to say, it would be nice if like all Wikipedia articles didn't suddenly fail validation when we switch to HTML5.  They tend to use presentational stuff all over the place.

[22:39:00.0000] <AryehGregor>
Oh, really?  From looking at the HTML source, it looks to me like they're just trying to get the absolute minimum number of bytes possible.

[22:39:01.0000] <AryehGregor>
Oh, odd.

[22:39:02.0000] <AryehGregor>
They use <head>, and <meta http-equiv="content-type" content=etc.> instead of <meta charset>.

[22:39:03.0000] <AryehGregor>
So not sure my hypothesis works.

[22:40:00.0000] <othermaciej>
I suspect some of their use of presentational elements could be replaced with class names and CSS that are a smaller total number of source bytes

[22:40:01.0000] <othermaciej>
I also suspect that after gzip there will be no difference

[22:40:02.0000] <AryehGregor>
I've found that there's usually some difference after gzip.

[22:40:03.0000] <othermaciej>
I think part of the reason for the way google.com is written may be to work with pre-CSS UAs

[22:40:04.0000] <AryehGregor>
With this sort of thing.

[22:40:05.0000] <AryehGregor>
What pre-CSS UAs?

[22:41:00.0000] <othermaciej>
truly ancient browsers

[22:41:01.0000] <othermaciej>
I think google.com aims to target even the long tail of very rarely used browsers

[22:41:02.0000] <othermaciej>
but it's hard to say why they do what they do

[22:41:03.0000] <AryehGregor>
Yeah.

[22:41:04.0000] <AryehGregor>
"Unescaped & in URL attribute - 87"  Is there any actual reason to prohibit this?

[22:41:05.0000] <othermaciej>
on most other pages that use presentational attributes, it seems to be an oversight, not a purposeful choice

[22:41:06.0000] <othermaciej>
I don't think unescaped & in URL attributes should be prohibited

[22:42:00.0000] <othermaciej>
btw if you click on the yes/no links you can see my methodology

[22:42:01.0000] <othermaciej>
which I guess I should document

[22:42:02.0000] <othermaciej>
I am using validator.nu to check for HTML5 and XHTML5 conformance

[22:42:03.0000] <othermaciej>
I am using validator.w3.org to check for conformance to declared doctype, if it is something other than HTML5

[22:42:04.0000] <othermaciej>
I am counting only errors, not warnings, in both cases

[22:42:05.0000] <AryehGregor>
My methodology for bulk testing was: for domain in `cat top-domains-201003`; do echo -en "$domain:\t"; curl -Ss "http://validator.w3.org/check?uri=http://$domain" | grep -o 'This document was successfully checked\|Errors found while checking this document\|This document can not be checked'; done

[22:43:00.0000] <othermaciej>
My goal here is detailed review of a fairly small number of sites, with the errors classified by hand

[22:43:01.0000] <AryehGregor>
Right.  Makes sense for your purpose.

[22:43:02.0000] <othermaciej>
since in some cases, the validator reports errors that are not errors per spec, or misclassifies the nature of the error

[22:43:03.0000] <AryehGregor>
I was looking for a statistic, for an unrelated purpose.

[22:44:00.0000] <othermaciej>
hmmm

[22:44:01.0000] <othermaciej>
how do I add text that goes before the table of contents in MediaWiki?

[22:45:00.0000] <AryehGregor>
Just add it before the first section, the TOC is put before the first section by default.  If you don't like the automatic TOC positioning, you can position it manually with __TOC__.

[22:45:01.0000] <othermaciej>
I see

[22:45:02.0000] <AryehGregor>
/me tries to figure out what his login is on the W3C wiki

[22:46:00.0000] <AryehGregor>
Hey, it does username autocompletion.  I wonder who wrote that.

[22:47:00.0000] <AryehGregor>
/me figures out his username

[22:47:01.0000] <AryehGregor>
1.12.0, yay.

[22:47:02.0000] <AryehGregor>
Any particular reason you used CamelCase for the page name?

[22:48:00.0000] <othermaciej>
it's the only thing I knew how to do

[22:49:00.0000] <othermaciej>
if I had my choice, I would call it "HTML5 Authoring Conformance Study"

[22:49:01.0000] <othermaciej>
or something like that

[22:49:02.0000] <othermaciej>
if you know how to fix it, that would be sweet

[22:49:03.0000] <othermaciej>
I just moved the TOC to where I want it

[22:49:04.0000] <AryehGregor>
Your choice is granted.

[22:50:00.0000] <AryehGregor>
MediaWiki doesn't require camel-case names.  I'm told that's one of its major historical innovations, although it was well before my time.

[22:50:01.0000] <AryehGregor>
othermaciej, is it a bug in the validator that's flagging autocomplete in the validator, or the fact that it's on <input type=hidden>?

[22:51:00.0000] <othermaciej>
oh

[22:51:01.0000] <othermaciej>
is it always on input type=hidden?

[22:51:02.0000] <othermaciej>
is that a conformance error?

[22:52:00.0000] <othermaciej>
it does look like <input type=hidden autocomplete=off> is nonconforming

[22:53:00.0000] <AryehGregor>
It's on <select> in some cases.

[22:53:01.0000] <othermaciej>
is it disallowed on <select> too?

[22:53:02.0000] <AryehGregor>
Seems so.

[22:53:03.0000] <othermaciej>
looks like it is

[22:53:04.0000] <othermaciej>
mea culpa

[22:53:05.0000] <othermaciej>
not sure offhand if the spec should allow those

[22:54:00.0000] <othermaciej>
in the case of <select>, it seems like it would be useful, and in the case of <input type=hidden>, it seems harmless but not useful either

[22:55:00.0000] <othermaciej>
fixing it

[22:55:01.0000] <AryehGregor>
Do browsers autocomplete <select>?

[22:55:02.0000] <othermaciej>
I believe Safari may, in at least some cases

[22:57:00.0000] <othermaciej>
right now I'm classifying the rest of the errors on yahoo.com

[22:57:01.0000] <othermaciej>
if you want to do any of the others, feel free

[22:57:02.0000] <othermaciej>
it's interesting how few of the Alexa top 100 validate

[22:57:03.0000] <othermaciej>
I think wikipedia.org may be the only one that is HTML5 and validates

[22:57:04.0000] <dbgi>
if anyone needs a free shell on a really nice webserver for  a website or whatever, send me a msg

[22:58:00.0000] <AryehGregor>
Well, that's because I wrote the current front page.  :P

[22:58:01.0000] <othermaciej>
:-)

[22:58:02.0000] <AryehGregor>
Or, rather, fixed it up.

[22:58:03.0000] <AryehGregor>
Also because I poked the right people about getting validator bugs fixed.

[22:58:04.0000] <AryehGregor>
So, not really surprising.  :P

[23:02:00.0000] <AryehGregor>
I did all the easy (short) ones.

[23:02:01.0000] <AryehGregor>
Let me try one or two of the big ones, then I'll go to bed.

[23:02:02.0000] <othermaciej>
cool, thans

[23:02:03.0000] <othermaciej>
*thanks

[23:03:00.0000] <othermaciej>
note that I'm only counting errors, not warnings

[23:03:01.0000] <othermaciej>
or info notes

[23:03:02.0000] <AryehGregor>
Right.

[23:04:00.0000] <othermaciej>
I am too lazy to find out which of the warnings are mandated by the spec

[23:04:01.0000] <othermaciej>
and which are just because the validator author thought they were a good idea

[23:09:00.0000] <othermaciej>
ok, finished yahoo.com

[23:09:01.0000] <AryehGregor>
I'm doing qq.com.

[23:10:00.0000] <AryehGregor>
/me is puzzled by "Required attributes missing on element style": <style type="text/css" media="screen">↩#ppIn

[23:10:01.0000] <AryehGregor>
What could that possibly mean?

[23:10:02.0000] <othermaciej>
I think that is the error you get when you have <style> outside <head> and it is not <style scoped>

[23:11:00.0000] <AryehGregor>
Aha.

[23:11:01.0000] <othermaciej>
I'm doing youtube

[23:11:02.0000] <othermaciej>
this exercise is less work than I initially expected, I may expand it to the top 20 plus the "honorable mention" and "HTML5 notable" sites that Sam mentioned on his blog

[23:12:00.0000] <othermaciej>
I am amused that not a single one of Google's Web properties in the top 100 validates

[23:12:01.0000] <othermaciej>
and many of them have huge numbers of errors

[23:12:02.0000] <othermaciej>
too bad the Web Standards Project no longer cares about that sort of thing

[23:12:03.0000] <AryehGregor>
They obviously don't care about validation at all, yeah.

[23:12:04.0000] <AryehGregor>
They care about interoperability in a practical sense only.

[23:12:05.0000] <AryehGregor>
I guess.

[23:13:00.0000] <othermaciej>
since Google folks are presumbly pretty smart about the Web, does this give evidence that validation is pointless?

[23:14:00.0000] <othermaciej>
or that validators are useful for a particular audience but others people/organizations may find it more productive not to validate

[23:14:01.0000] <AryehGregor>
You mean, you're thinking of validation as being a way to catch authoring errors from a pragmatic standpoint, and not a way to impress people with how standards-compliant you are?

[23:14:02.0000] <AryehGregor>
To be honest, that's a strange point of view to me.

[23:14:03.0000] <AryehGregor>
Validators are sometimes useful, it's true.

[23:15:00.0000] <AryehGregor>
But realistically, they're definitely much more annoying than useful, flagging all sorts of trivial things that work perfectly fine and rarely helping you with debugging.

[23:15:01.0000] <AryehGregor>
Sometimes they do, but not that often.

[23:15:02.0000] <AryehGregor>
At least in my experience.  YMMV.

[23:18:00.0000] <othermaciej>
Should validators (and notions of document conformance) be changed so they are helpful more often than they are annoying?

[23:18:01.0000] <othermaciej>
C compiler warnings are often annoying but if you just blindly fix them all you will fix some real bugs

[23:18:02.0000] <othermaciej>
probably more than if you tried to review them one at a time

[23:18:03.0000] <AryehGregor>
If that's the goal, presentational attributes definitely need to be allowed, for starters.

[23:18:04.0000] <AryehGregor>
Many of the errors I'm seeing here are reasonable, though.

[23:19:00.0000] <AryehGregor>
But if that's the only goal of a validator, then there's no need for the spec to say what's valid.

[23:19:01.0000] <AryehGregor>
You can have competition between different error-checkers.

[23:19:02.0000] <othermaciej>
people have indeed raised that proposition

[23:20:00.0000] <othermaciej>
hsivonen says that it's useful for there to be at least a reasonable common base of errors reported, so you can switch validators without breaking your whole workflow

[23:20:01.0000] <AryehGregor>
It would also still be useful to retain a way of checking whether you're only using standard, interoperable functionality.

[23:20:02.0000] <othermaciej>
the one error that seems clearly bogus to me based on info so far is unescaped & in an attribute

[23:21:00.0000] <AryehGregor>
/me agrees with that

[23:21:01.0000] <othermaciej>
it seems like a huge waste of time to escape all those

[23:21:02.0000] <othermaciej>
I am also dubious about the value of banning presentational elements/attributes

[23:21:03.0000] <othermaciej>
but that seems less clearcut

[23:21:04.0000] <othermaciej>
most of these sites show only a tiny handful of uses which may well be an oversight, not a deliberate strategy

[23:21:05.0000] <othermaciej>
the main exception being google.com, which is clearly using presentational markup very deliberately

[23:22:00.0000] <AryehGregor>
If you leave the top 100 sites, you'll find a lot more deliberate presentational markup.

[23:22:01.0000] <AryehGregor>
Or just look at a random Wikipedia page's content (not the skin).

[23:23:00.0000] <othermaciej>
it does seem to me that presentational markup is a matter of taste, not unambiguously an "error"

[23:23:01.0000] <othermaciej>
I am ok with saying <i>foo <b> bar </i> baz </b> is an error

[23:24:00.0000] <othermaciej>
because it's almost certainly not going to give quite the result you intended, so it is probably a mistake

[23:24:01.0000] <othermaciej>
likewise with duplicate ids

[23:24:02.0000] <othermaciej>
but I don't see how use of the <u> element is in that same category

[23:24:03.0000] <AryehGregor>
It's in the same category as omitting alt from <img>.

[23:25:00.0000] <AryehGregor>
I.e., the spec bans it because of political pressure, not because it's not a rational thing for some authors to do.

[23:25:01.0000] <AryehGregor>
Anyone who thinks it's usually in authors' interest to give good alt text is smoking something.  Good for blind users, yes; but most authors have a negligible number of blind users relative to the effort it would take to use alt text properly.

[23:25:02.0000] <AryehGregor>
But this is politically unacceptable to acknowledge in a spec, it seems to me.

[23:26:00.0000] <othermaciej>
some authors have legal or policy reasons to care about blind users disproportionate to their share of the audience

[23:27:00.0000] <AryehGregor>
Not most, though.

[23:27:01.0000] <othermaciej>
others (e.g. random using uploading Flickr photo or writing a GMail message to grandma) do not

[23:27:02.0000] <AryehGregor>
The requirement is unreasonable for the large majority of authors, from a rational/economic standpoint.

[23:27:03.0000] <othermaciej>
it's plausible as a legal requirement for governments and large corporations or charities

[23:28:00.0000] <AryehGregor>
Yes, but that's not most authors.

[23:28:01.0000] <AryehGregor>
It's a small minority.

[23:28:02.0000] <AryehGregor>
But the requirement in the spec is absolute.

[23:28:03.0000] <AryehGregor>
Banning presentational markup is similar.

[23:28:04.0000] <othermaciej>
by visit count it probably affects most content, by page count, no

[23:28:05.0000] <othermaciej>
see also: power law distribution

[23:29:00.0000] <othermaciej>
I'm not sure that banning presentational markup even has an accessibility benefit

[23:29:01.0000] <AryehGregor>
Your hypothesis suggests that major sites would reliably have good alt text.

[23:29:02.0000] <AryehGregor>
Or at least major competently-run sites would.

[23:29:03.0000] <AryehGregor>
Is this the case?  I doubt it.

[23:30:00.0000] <othermaciej>
my hypothesis is that they have a legal/policy requirement, not that they are always attentive to it

[23:30:01.0000] <othermaciej>
for better or for worse, it's probably something they see as a cost of doing business, not an opportunity

[23:32:00.0000] <AryehGregor>
Nevertheless, it's not useful advice to the large majority of authors.  Maybe if you weight by page views, but I doubt authors use validators proportional to their page views.

[23:32:01.0000] <AryehGregor>
It's not useful advice to the large majority of authors who are trying to check their work in a validator.

[23:32:02.0000] <AryehGregor>
So it's not useful if your goal is a lint-like tool.

[23:33:00.0000] <AryehGregor>
I see two major purposes in a validator: 1) Informing authors of possible mistakes.  2) Informing concerned users about how standards-compliant a site is (e.g., for clients who are interested in having a standards-based site for policy reasons).

[23:34:00.0000] <othermaciej>
are duplicate IDs a useful error to report?

[23:34:01.0000] <othermaciej>
it seems like (2) is circular as a reason to have more conformance requirements

[23:34:02.0000] <othermaciej>
if the only reason to have a standards conformance requirement is to report to concerned users that a site fails that requirement, then that in itself is not a very good reason

[23:35:00.0000] <AryehGregor>
(2) would suggest reporting things that suggest the page is using non-standard functionality, such as unrecognized elements or attributes, or <object>s containing Flash, etc.

[23:35:01.0000] <othermaciej>
37 errors left to classify on YouTube...

[23:35:02.0000] <othermaciej>
<object>s containing Flash is something that no current validator will complain about

[23:35:03.0000] <AryehGregor>
By "standards-compliant" I meant "following de facto or de jure standards so as to behave interoperably in different clients".

[23:35:04.0000] <othermaciej>
I suspect making it a conformance error would mae lots of people mad

[23:35:05.0000] <AryehGregor>
I think they should.

[23:36:00.0000] <AryehGregor>
(complain, not be mad :) )

[23:36:01.0000] <othermaciej>
as for unrecognized elements, it still seems circular to me to say that they should be nonstandard so that the validator can report that they are nonstandard

[23:36:02.0000] <othermaciej>
if the spec allowed arbitrary custom elements and attributes, then they wouldn't be nonstandard

[23:36:03.0000] <AryehGregor>
Then the spec would not be doing its job of specifying behavior that can be implemented interoperably.

[23:37:00.0000] <AryehGregor>
I don't care about standards for standards' sake, I care about ensuring interoperability.

[23:37:01.0000] <othermaciej>
saying that they should be nonstandard so a valid document doesn't accidentally rely on uninteroperable behavior seems reasonable

[23:37:02.0000] <othermaciej>
though that is an aid to the author too, not just vigilant users

[23:37:03.0000] <AryehGregor>
Yes.

[23:39:00.0000] <AryehGregor>
Finished qq.com.

[23:40:00.0000] <AryehGregor>
Lots of miscellaneous errors.  It seems closer to what a typical site looks like.

[23:41:00.0000] <AryehGregor>
X-UA-compatible is another notable occurrence.

[23:42:00.0000] <AryehGregor>
Well, I guess only three sites have it

[23:42:01.0000] <AryehGregor>
.

[23:44:00.0000] <AryehGregor>
Okay, time for sleep.  I'm going to be a wreck tomorrow.

[23:44:01.0000] <othermaciej>
ok done with youtube

[23:44:02.0000] <othermaciej>
thanks for the help

[00:42:00.0000] <othermaciej>
these results are really thrown off by the sheer number of Google Web properties that rank high in Alexa

[01:10:00.0000] <othermaciej>
wow, chromium.org is a disaster markup-wise: http://validator.nu/?doc=http%3A%2F%2Fwww.chromium.org%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full-aria.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F&parser=html5

[03:19:00.0000] <jgraham>
othermaciej, AryehGregor: You know that hsivonen has code to collect and categorise validator errors, right?

[03:20:00.0000] <othermaciej>
jgraham: Sam Ruby made a script that I think parses the XML output of the validator, but the current validator misclassifies some of its own errors

[03:20:01.0000] <othermaciej>
jgraham: so I'm finding I can't get good quality results without at least a bit of human judgment

[03:20:02.0000] <jgraham>
OK

[03:21:00.0000] <jgraham>
Duplicsate IDs seem like a real and obvious error btw

[03:22:00.0000] <jgraham>
(because they are easy to add by accident and can have unexpected consequences if you don't realise that you have made the error)

[03:27:00.0000] <othermaciej>
I do think duplicate IDs are a real error that should be reported, yes

[03:27:01.0000] <othermaciej>
I am not sure that Sam thinks otherwise, despite his nominal stance that all document conformance requirements should be removed

[03:28:00.0000] <othermaciej>
I think he actually believes that some should be removed, but does not want to do the work to describe which ones

[03:38:00.0000] <othermaciej>
I am really curious how two sites ended up in the alexa top 100 that are unreachable

[03:39:00.0000] <gavin>
heh

[03:43:00.0000] <othermaciej>
kinda implies that there is a major flaw in how they gather their traffic stats

[03:51:00.0000] <othermaciej>
Philip`: that's neat data, how fresh is it?

[03:52:00.0000] <Philip`>
"The source data is 131,072 randomly-selected HTTP URIs from the Open Directory Project, downloaded on 2008-02-26."

[03:52:01.0000] <othermaciej>
/me would guess the HTML5 doctype is now more popular than the HTML2 doctype

[03:52:02.0000] <othermaciej>
though that might be overly optimistic

[03:52:03.0000] <othermaciej>
might not be in HTML 3.2 territory yet

[03:53:00.0000] <othermaciej>
wow, nearly half of pages have no doctype? (or at least did at time of study)

[03:54:00.0000] <Philip`>
http://www.pcmech.com/article/the-mysterious-1e100-net/

[03:54:01.0000] <Philip`>
Sounds like that one's a Google CDN of some kind

[03:54:02.0000] <Philip`>
hence getting lots of traffic, but not being an actual web site

[03:55:00.0000] <othermaciej>
oh, hahaha

[03:56:00.0000] <othermaciej>
I should have spotted the pun

[03:58:00.0000] <othermaciej>
I wonder what bp.blogspot.com is then

[03:59:00.0000] <othermaciej>
ah, another subdomain-only google domain

[03:59:01.0000] <othermaciej>
http://1.bp.blogspot.com/

[04:03:00.0000] <Philip`>
I wouldn't expect the number of no-doctype pages to change much

[04:03:01.0000] <Philip`>
since the number of people who are lazy and/or ignorant of the problems of quirks mode is unlikely to change much

[04:04:00.0000] <Philip`>
so they'll be creating the same proportion of no-doctype new pages

[04:05:00.0000] <Philip`>
(When people hear about the HTML5 doctype that might overcome laziness, but that'll take years)

[04:06:00.0000] <othermaciej>
yeah I wouldn't expect no-doctype to change

[04:06:01.0000] <othermaciej>
and I would expect attrition rate of truly ancient doctypes to be quite slow

[04:06:02.0000] <othermaciej>
what I'm expecting is use of the HTML5 doctype may have gone up

[04:07:00.0000] <othermaciej>
I think none of the alexa top 100 sites that use it were doing so 2 years ago

[04:09:00.0000] <othermaciej>
most ironic non-use of HTML5 doctype: http://www.youtube.com/html5

[04:10:00.0000] <othermaciej>
no wait, I lie, the truly most ironic non-use is http://www.w3.org/TR/html5/

[04:15:00.0000] <MikeSmith>
in a sane world, authors wouldn't have to worry about including a doctype and would not be seen as lazy if they just omitted it

[04:15:01.0000] <othermaciej>
it would be good if no special talisman is needed

[04:15:02.0000] <MikeSmith>
and the issue of what doctype the HTML5 TR should have would be a non-issue

[04:16:00.0000] <othermaciej>
I do feel like triggering standards mode is a benefit that is worth the cost of "<!DOCTYPE html>"

[04:16:01.0000] <MikeSmith>
yeah, absolutely

[04:16:02.0000] <othermaciej>
it would be better if we could get rid of even the vestigial remainder

[04:16:03.0000] <MikeSmith>
indeed

[04:17:00.0000] <othermaciej>
but it's probably not possible, due to the mass of content in quirks mode (and therefore presumably authored to quirks mode)

[04:56:00.0000] <annevk>
we could invent a mime type that takes the pain away

[04:56:01.0000] <annevk>
but that's prolly not worth it

[04:56:02.0000] <annevk>
text/htmls

[05:20:00.0000] <othermaciej>
that just sounds like more pain

[05:20:01.0000] <annevk>
text/pain

[05:21:00.0000] <annevk>
gonna be a hit

[05:42:00.0000] <annevk>
/me is having fun on twitter

[05:43:00.0000] <annevk>
/me should really do something useful; like writing the presentation he has to give tomorrow

[05:48:00.0000] <MikeSmith>
annevk: where are you presenting?

[05:51:00.0000] <annevk>
somewhere I can cycle towards

[05:51:01.0000] <annevk>
I think it's a first

[05:52:00.0000] <annevk>
well, apart from school and university

[05:52:01.0000] <annevk>
ironically, and perhaps unsurprisingly, this is at university

[05:52:02.0000] <annevk>
but I'm the teacher now, or something :p

[06:08:00.0000] <othermaciej>
Philip`: I wish there were a count of the number of pages using xmlns attributes instead of just count of attributes

[06:08:01.0000] <Philip`>
othermaciej: Which file are you looking at?

[06:09:00.0000] <othermaciej>
oops, my paste failed

[06:09:01.0000] <othermaciej>
http://philip.html5.org/data/xmlns-bindings.txt

[06:09:02.0000] <othermaciej>
actually, I'm not even sure what those counts are

[06:09:03.0000] <othermaciej>
are they counts of the number of times that prefix declaration appeared, or the number of pages on which it appeared?

[06:09:04.0000] <Philip`>
"Counts are the number of pages with that attribute name+value."

[06:10:00.0000] <Philip`>
so number of pages

[06:10:01.0000] <Philip`>
http://philip.html5.org/data/xmlns-attributes.txt is similar but newer

[06:11:00.0000] <Philip`>
(though it includes non-HTML pages too, I think)

[06:12:00.0000] <othermaciej>
what I'm curious about that I don't think I can get from this output is the number of pages that contain at least one xmlns attribute

[06:12:01.0000] <othermaciej>
in your older study, 79% of the occurrences of namespace declarations had "microsoft" in the namespace URI

[06:12:02.0000] <Philip`>
Do you mean xmlns:* (i.e. not counting xmlns="http://www.w3.org/1999/xhtml")?

[06:12:03.0000] <othermaciej>
that seems less likely with the new data

[06:13:00.0000] <othermaciej>
yeah xmlns:*

[06:14:00.0000] <othermaciej>
(also of potential interest would be a separate count of xmlns= that are neither the xhtml namespace URI nor empty, if there are nay such)

[06:14:01.0000] <Philip`>
I could upload my list of all (url, element-name, attribute-name-which-begins-with-"xmlns", attribute-value) from the dotbot data so you can process it yourself, if you want

[06:16:00.0000] <othermaciej>
that would be neat

[06:16:01.0000] <Philip`>
http://philip.html5.org/data/xmlns-attributes-raw.txt.bz2 (~5MB)

[06:17:00.0000] <Philip`>
(with all the usual caveats about sample selections and bias and bugs and so on)

[06:25:00.0000] <othermaciej>
$ grep 'xmlns:' xmlns-attributes-raw.txt | sed -e 's/    *//g;' | sort | uniq | xwc -l

[06:25:01.0000] <othermaciej>
   18718

[06:26:00.0000] <Philip`>
Your sed looks wrong

[06:26:01.0000] <Philip`>
They're tab-separated fields

[06:26:02.0000] <othermaciej>
I used a tab

[06:26:03.0000] <othermaciej>
(use ^V to enter a literal tab)

[06:27:00.0000] <othermaciej>
by that count, 4.4% of pages contain an XML namespace prefix declaration

[06:27:01.0000] <Philip`>
Oh, it looks like spaces in IRC :-)

[06:27:02.0000] <Philip`>
(I'd use "cut -f1" to extract the first field instead)

[06:27:03.0000] <othermaciej>
I checked that it stripped all fields but the first

[06:28:00.0000] <othermaciej>
noted!

[06:28:01.0000] <othermaciej>
cut -f1 gives a different result so I may have made a mistake

[06:28:02.0000] <Philip`>
I get 18718 if I *don't* do any sed/cut

[06:29:00.0000] <othermaciej>
my sed line is wrong

[06:29:01.0000] <othermaciej>
(it doesn't do nothing, but it does do the stripping wrong

[06:29:02.0000] <Philip`>
Oh, good point

[06:29:03.0000] <othermaciej>
what I wanted was sed -e 's/    .*//g;', except that cut is better

[06:29:04.0000] <Philip`>
It's just stripping out the tabs, I guess

[06:30:00.0000] <othermaciej>
so 7601

[06:30:01.0000] <othermaciej>
so 1.8% of pages include at least one XML Namespace prefix declaration

[06:31:00.0000] <othermaciej>
Is it ok if I put a link to the raw data in the HTML WG wiki?

[06:31:01.0000] <Philip`>
Sure

[06:32:00.0000] <Philip`>
Note that quite a few of these pages appear to be RSS feeds

[06:32:01.0000] <Philip`>
which annoyingly happen to be included in the data set

[06:33:00.0000] <Philip`>
/me doesn't claim that it is high-quality data

[06:33:01.0000] <othermaciej>
2476 pages have an xml namespace prefix declaration that has "microsoft" in the namespace URI

[06:35:00.0000] <othermaciej>
$ grep 'xmlns:' xmlns-attributes-raw.txt | egrep -v 'rss|feed' | cut -f1 | sort | uniq | xwc -l

[06:35:01.0000] <othermaciej>
    6853

[06:35:02.0000] <othermaciej>
$ grep 'xmlns:.*microsoft' xmlns-attributes-raw.txt | egrep -v 'rss|feed' | cut -f1 | sort | uniq | xwc -l

[06:35:03.0000] <othermaciej>
    2458

[06:36:00.0000] <othermaciej>
so more like 1.6%, and 36% of those are Microsoft namespace URIs

[06:58:00.0000] <MikeSmith>
othermaciej: the autocomplete=off instances on the facebook homepage are not conformant; autocomplete is explicitly not allowed on input/@type=hidden

[06:58:01.0000] <MikeSmith>
http://dev.w3.org/html5/spec/forms.html#hidden-state

[06:59:00.0000] <othermaciej>
MikeSmith: I know - I forgot to close the validator bug when I found that out

[06:59:01.0000] <MikeSmith>
ok, I'll close it myself now

[07:18:00.0000] <othermaciej>
MikeSmith: question: do you understand the error in the doctype reported here? http://validator.nu/?doc=http%3A%2F%2Ffacebook.com%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F

[07:18:01.0000] <othermaciej>
MikeSmith: it seems to be complaining about the newline - I think

[07:18:02.0000] <othermaciej>
is that actually noncomforming?

[07:20:00.0000] <othermaciej>
ok, looks like newlines in the doctype are indeed disallowed in HTML5

[07:20:01.0000] <othermaciej>
is that also true in XML, or SGML?

[07:23:00.0000] <TabAtkins>
AryehGregor: When talking about blind users, don't forget the ones that *are* measurably important to all authors, the search engines.

[07:24:00.0000] <TabAtkins>
AryehGregor: (That is, ironically, why I have to keep *removing* alt text from my company's site, because the idiot SEO people just keep cramming keywords into them.)

[07:24:01.0000] <jgraham>
othermaciej: Newlinwa in doctype appear to be OK per XML

[07:24:02.0000] <jgraham>
wtf

[07:24:03.0000] <jgraham>
Newlines

[07:25:00.0000] <othermaciej>
I'm not sure why HTML5 only allows spaces as the whitespace in doctypes

[07:25:01.0000] <othermaciej>
(at least that's what it appears to do)

[07:36:00.0000] <MikeSmith>
othermaciej: as far as the code that's emitting that error message, it's in the htmlparser code, so I have less insight into to that

[07:36:01.0000] <MikeSmith>
I'm also have no idea why it's disallowed

[07:36:02.0000] <MikeSmith>
I think it may just be a mistake

[07:37:00.0000] <Philip`>
othermaciej: Why do you think the newline has any effect?

[07:37:01.0000] <Philip`>
All whitespace should be handled the same, and I don't see any different behaviour here

[07:38:00.0000] <othermaciej>
Philip`: this is an error: http://validator.nu/?doc=http%3A%2F%2Ffacebook.com%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F

[07:38:01.0000] <othermaciej>
Philip`: but this is not: http://validator.nu/?doc=http%3A%2F%2Fwww.msn.com%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F&parser=html5&laxtype=yes

[07:39:00.0000] <othermaciej>
hmm, the error goes away if I change the parser to HTML5

[07:39:01.0000] <othermaciej>
mea culpa

[07:40:00.0000] <Philip`>
I think the validator is consistent across sites, it's just got occasionally confusing modes

[07:41:00.0000] <MikeSmith>
so that error was coming from it being parsed by the XML parser?

[08:21:00.0000] <othermaciej>
MikeSmith: I think using the HTML4 parser instead of the HTML5 parser makes XHTML doctypes an error

[08:29:00.0000] <othermaciej>
what's the third error here about: http://validator.nu/?doc=http%3A%2F%2Fwordpress.com%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full-aria.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F&parser=html5

[08:29:01.0000] <othermaciej>
(it's a <script> without src, but it's also non-empty)

[08:30:00.0000] <othermaciej>
I wish it would say what attribute it thinks is missing!

[08:31:00.0000] <annevk>
if you specify charset you must specify src I think

[08:31:01.0000] <Philip`>
"The charset attribute gives the character encoding of the external script resource. The attribute must not be specified if the src attribute is not present."

[08:32:00.0000] <othermaciej>
ah, I see

[08:35:00.0000] <MikeSmith>
othermaciej: In my workspace, I have a fix for the problem of v.nu not reporting the names of missing attributes

[08:36:00.0000] <MikeSmith>
but I need to make some changes to it before committing it

[08:36:01.0000] <othermaciej>
I guess it's pointless to specify charset without src

[08:36:02.0000] <Philip`>
It should have no effect, and is confusing because people reading the markup might expect it to have an effect

[08:37:00.0000] <MikeSmith>
hmm, that may be a bug, though

[08:37:01.0000] <MikeSmith>
http://qa-dev.w3.org:8888/?doc=http%3A%2F%2Fwordpress.com%2F does not report it as an error

[08:38:00.0000] <othermaciej>
Is it trying to apply HTML5 validation?

[08:38:01.0000] <othermaciej>
that error only shows up if you force HTML5 mode

[08:38:02.0000] <Philip`>
(and preventing confusion is a common part of the rationale for conformance requirements)

[08:39:00.0000] <MikeSmith>
the fact that it's not reported as an error by the qa-dev v.nu instance appears to be that I have a bug in my workspace code 

[08:39:01.0000] <Philip`>
MikeSmith: "XHTML 1.0 Transitional doctype seen. [...] Using the schema for HTML 4.01 Transitional + IRI / XHTML 1.0 Transitional + IRI."

[08:39:02.0000] <MikeSmith>
oh, OK

[08:40:00.0000] <MikeSmith>
http://qa-dev.w3.org:8888/?doc=http%3A%2F%2Fwordpress.com%2F&schema=http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fhtml5full-aria.rnc+http%3A%2F%2Fs.validator.nu%2Fhtml5%2Fassertions.sch+http%3A%2F%2Fc.validator.nu%2Fall%2F&parser=html5

[08:40:01.0000] <MikeSmith>
Error: Element script is missing required attribute src.

[08:40:02.0000] <MikeSmith>
I really need to get that fix committed

[08:41:00.0000] <Philip`>
That's still a far less helpful error than saying it's required because of the charset

[08:41:01.0000] <Philip`>
People will think they have to add a src, they won't think they have to remove the charset

[08:42:00.0000] <Dashiva>
I agree

[08:42:01.0000] <Philip`>
I'm only saying that because I don't have to implement it, of course

[08:42:02.0000] <Dashiva>
It should say something like "charset attribute only applies to external scripts" or something.

[08:43:00.0000] <MikeSmith>
well, this is fundamental problem with grammar-based constraint checking

[08:43:01.0000] <MikeSmith>
we could move the constraint check for this case to the assertions-checking code, where we have precise control over the error message

[08:43:02.0000] <Dashiva>
Can't you... invert the constraint or something?

[08:43:03.0000] <MikeSmith>
but if we do this check using the RelaxNG schema, we have no real control over the error that's reported

[08:43:04.0000] <annevk>
it should prolly say "remove the charset attribute"

[08:43:05.0000] <Dashiva>
Instead of saying charset requires src, say !src requires !charset

[08:44:00.0000] <Dashiva>
(Disclaimer: I don't know relaxng)

[08:44:01.0000] <MikeSmith>
Dashiva: if we move this out of the schema, then anybody using the schema only (without also at least using the associated Schematron assertions) is not going to see any error message at all for this

[08:45:00.0000] <Dashiva>
So the schema doesn't allow negative assertions like that one?

[08:45:01.0000] <MikeSmith>
Dashiva: no, we can't really invert the constraint in the scheam

[08:45:02.0000] <MikeSmith>
Dashiva: no, it doesn't

[08:46:00.0000] <Dashiva>
I don't think optimizing for solutions that are known to be fundamentally lacking is a good idea, though

[08:46:01.0000] <MikeSmith>
this is not just a case for charset, it's the same case for defer and async

[08:46:02.0000] <MikeSmith>
Dashiva: you are preaching to the choir on that

[08:47:00.0000] <Philip`>
Who is the anybody using the schema only?

[08:47:01.0000] <MikeSmith>
see my recent messages to www-tag, if you care to

[08:47:02.0000] <MikeSmith>
Philip`: the availability of the schema risks people actually using it alone

[08:47:03.0000] <Philip`>
i.e. are they hypothetical persons, or real practical implementations?

[08:47:04.0000] <MikeSmith>
people will use it alone

[08:47:05.0000] <MikeSmith>
this is not about implementations

[08:48:00.0000] <MikeSmith>
people can grab the schema and use it with Jing or whatever other RelaxNG validation tool they want

[08:48:01.0000] <Dashiva>
Well, aren't these errors here mainly harmless?

[08:48:02.0000] <MikeSmith>
use it from the command line

[08:48:03.0000] <Dashiva>
So it doesn't matter if someone aiming for mediocrity doesn't get the warnings

[08:49:00.0000] <MikeSmith>
Dashiva: no a particularly helpful way to describe it

[08:49:01.0000] <MikeSmith>
the schema is useful on its own as a way to guide context-sensitive editing

[08:49:02.0000] <MikeSmith>
that is, integrated into editing apps

[08:50:00.0000] <MikeSmith>
someone can right now take that schema and use it to edit HTML5 documents in Emacs, for example

[08:50:01.0000] <MikeSmith>
in nxml-mode

[08:50:02.0000] <MikeSmith>
that is not a case of anybody aiming for mediocrity

[08:50:03.0000] <MikeSmith>
ask hober 

[08:50:04.0000] <Dashiva>
Why wouldn't they include the schematron portions as well?

[08:51:00.0000] <MikeSmith>
Dashiva: because nxml-mode does not do Schematron checking

[08:52:00.0000] <MikeSmith>
that does not mean it's useless 

[08:52:01.0000] <Dashiva>
I didn't say it was useless, I said it was mediocre

[08:52:02.0000] <MikeSmith>
whatever

[08:52:03.0000] <Philip`>
I imagine RNG makes it possible to determine what is the next permitted token given an incomplete document, whereas Schematron doesn't

[08:52:04.0000] <Philip`>
so the latter isn't much good for auto-complete

[08:53:00.0000] <MikeSmith>
right

[08:53:01.0000] <MikeSmith>
all that Schematron does is test XPath expressions

[08:53:02.0000] <Philip`>
Can they be Turing-complete expressions?

[08:53:03.0000] <othermaciej>
so there's really no way to say an attribute is conditionally allowed based on presence of another?

[08:54:00.0000] <MikeSmith>
othermaciej: that is in fact what the schema currently says

[08:54:01.0000] <Dashiva>
I also don't see how _absense_ of src is going to lead to a benefit in autocomplete

[08:54:02.0000] <MikeSmith>
it says, charset is allowed if the src attribute is specific

[08:54:03.0000] <MikeSmith>
Dashiva: true

[08:54:04.0000] <othermaciej>
I see, so it's just the error message that is bad

[08:54:05.0000] <MikeSmith>
s/specific/specified/

[08:54:06.0000] <othermaciej>
it should say that charset is disallowed rather than src is missing

[08:55:00.0000] <othermaciej>
it seems like that would be the more natural way to express the constraint

[08:55:01.0000] <Philip`>
/me wonders if there's a subset of Schematron that allows the permitted next characters to be computable, while still being more expressive than RNG grammars

[08:55:02.0000] <MikeSmith>
othermaciej: as I said, I don't think we can have it report that with just the schema. we need to have it reported in the assertions code

[08:56:00.0000] <MikeSmith>
and if we keep it in the schema, then we get two errors reported

[08:56:01.0000] <MikeSmith>
Dashiva: this may indeed by a case where we can remove it from the schema with causing harm

[08:56:02.0000] <Dashiva>
Philip`: Surely it should be possible to expand RNG to handle absense of attributes too

[08:56:03.0000] <othermaciej>
I guess it is a problem of the schema engine sucking

[08:56:04.0000] <MikeSmith>
I hope we can remove it, personally

[08:57:00.0000] <MikeSmith>
this is a general deficiency in all grammar-based validation mechanisms that I am aware of

[08:57:01.0000] <othermaciej>
if it doesn't get fixed, then it's probably better to prioritize quality of error messages over usability of the schema for alternate purposes

[08:57:02.0000] <MikeSmith>
othermaciej: agreed

[08:58:00.0000] <MikeSmith>
othermaciej: again, see my comments about this that I posted in replies to DanC on the TAG mailing list

[08:58:01.0000] <MikeSmith>
this is one of the main reasons I don't want to have the HTML WG publish that schema

[08:58:02.0000] <MikeSmith>
me and others

[08:59:00.0000] <Philip`>
Dashiva: I expect negation is fundamentally incompatible with the theoretical model used by RNG

[08:59:01.0000] <Philip`>
(I don't know that's true, but it would explain why they don't have negation)

[09:01:00.0000] <Philip`>
In any case, if the grammar allows <script> and <script src charset> and the input is <script src> then it's easy to tell the input doesn't match the grammar, but there's no general correct way to decide which specific part of the grammar it's not matching

[09:02:00.0000] <Philip`>
so implementations have to make some arbitrary choice of which error to report, and in some cases that won't be ideal

[09:02:01.0000] <othermaciej>
Microsoft's homepage seems to return different content to http://validator.w3.org/ and http://validator.nu/ , even when forcing HTML5 mode in both

[09:03:00.0000] <othermaciej>
hmm, it seems to have stopped doing so

[09:03:01.0000] <MikeSmith>
Philip`: the choice that Jing makes (and likely any other RelaxNG implementation makes) is to report anything that is required but missing

[09:05:00.0000] <Philip`>
MikeSmith: If the grammar allows <foo bar> and <foo baz> and you write <foo>, then what does it report?

[09:06:00.0000] <MikeSmith>
it depends on the implementation

[09:06:01.0000] <MikeSmith>
the trunk of jing reports "required attributes missing"

[09:06:02.0000] <Philip`>
Ah

[09:06:03.0000] <MikeSmith>
you see why now, I think

[09:06:04.0000] <MikeSmith>
but the patched jing in my workspace reports something different

[09:07:00.0000] <MikeSmith>
hang on, I will get a URL

[09:07:01.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/www-archive/2010Jan/0077.html

[09:08:00.0000] <MikeSmith>
"Element object is missing one or more of the following attributes:  data type"

[09:08:01.0000] <Philip`>
Ah, right

[09:13:00.0000] <Philip`>
Validator.nu ought to save the error messages reported for each page, and the diffs between versions of a URL that are validated at least twice in quick succession

[09:14:00.0000] <Philip`>
and feed that into some magic AI algorithm that can determine the most frequent markup fix associated with each validation error

[09:14:01.0000] <Philip`>
and use that to suggest the correct fix to future users

[09:14:02.0000] <MikeSmith>
seriously, we should eventually have something like that

[09:15:00.0000] <Philip`>
If the magic AI algorithm was replaced by a person, then it could perhaps work

[09:15:01.0000] <MikeSmith>
heh

[09:15:02.0000] <MikeSmith>
yeah

[09:15:03.0000] <Philip`>
and if the saving of messages and pages was replaced by something that doesn't violate the privacy policy

[09:16:00.0000] <MikeSmith>
well, anybody who wanted to experiment with trying it could implement their own checker on top of the v.nu API

[09:16:01.0000] <MikeSmith>
with their own privacy policy

[09:16:02.0000] <Philip`>
Maybe it wouldn't be too hard for someone to validate a load of existing pages and check the results, and write a suite of test cases that include the markup error and the best guess at a fix and a human-readable description of the issue

[09:17:00.0000] <Philip`>
and then validators can try to find heuristics that get a reasonable match to the expected descriptions/fixes for those cases

[09:17:01.0000] <Philip`>
s/validators/validator developers/

[09:17:02.0000] <AryehGregor>
othermaciej, I assume that 1e100.net and the other don't have anything at the domain name itself, only on subdomains.

[09:17:03.0000] <othermaciej>
AryehGregor: that appears to be the case, yes

[09:18:00.0000] <Dashiva>
http://lists.w3.org/Archives/Public/public-html/2010Mar/0468.html

[09:18:01.0000] <Dashiva>
Does this message actually say anything?

[09:23:00.0000] <Philip`>
I think it says that we shouldn't be using Bugzilla for everything, and should have a different mechanism for higher-level issues

[09:24:00.0000] <Philip`>
but it says it in too polite a way, which waters down any meaning in the message

[09:24:01.0000] <Dashiva>
I guess my formulation of the question was a bit harsh.

[09:26:00.0000] <Dashiva>
But it seems to say things we already know and suggest things we already do.

[09:26:01.0000] <Philip`>
It's just a different tone to the usual focused directness on technical mailing lists

[09:27:00.0000] <Philip`>
The thing we already do is expect everything to be put into a Bugzilla bug

[09:28:00.0000] <Dashiva>
And use an issue tracker for larger issues

[09:29:00.0000] <Philip`>
and it seems to say we perhaps shouldn't do that, we should just use it for breaking down big problems into smaller ones and we need a different way of managing big issues

[09:29:01.0000] <Philip`>
The process requires use of the bug tracker before the issue tracker

[09:31:00.0000] <Dashiva>
That's a formality required to establish whether it's a large issue or not

[09:32:00.0000] <AryehGregor>
So, I have two spec-related goals today: ask the W3C validator team if we could work out some way for HTML5 to validate with an obsolete-but-conforming doctype; and post to whatwg arguing that presentational elements/attributes (at least some) should be obsolete but conforming instead of non-conforming.

[09:32:01.0000] <AryehGregor>
Dashiva, she's using "large" in the sense of "encompassing a broad topic", not "important".

[09:32:02.0000] <Philip`>
Dashiva: It might be a formality that makes it hard to address certain problems

[09:33:00.0000] <AryehGregor>
She makes a perfectly valid point: there is no way to argue about high-level, broad issues in the HTMLWG.

[09:33:01.0000] <AryehGregor>
This is by design, as far as I can tell.

[09:33:02.0000] <Dashiva>
Apart from the mailing list itself

[09:33:03.0000] <Philip`>
(since it might not be possible to express the problem as "a clear statement of a single problem with the spec")

[09:33:04.0000] <Dashiva>
You only need the tracking when you want to force a decision

[09:33:05.0000] <AryehGregor>
The mailing list lets you argue, fine, but it doesn't give you an avenue to get the spec changed, unless you can convince the editor.

[09:34:00.0000] <Dashiva>
But you can't change the spec in a high-level, broad way

[09:34:01.0000] <Dashiva>
You have to break it down into actual changes

[09:34:02.0000] <AryehGregor>
The only thing you can do is target one small issue and hope that the chairs' decision has larger-scale implications that the editor will choose to follow, as happened with splitting up the spec after the microdata decision.

[09:34:03.0000] <Dashiva>
It's fine to _argue_ high-level, but the text has to be written for specifics in the end

[09:35:00.0000] <AryehGregor>
In principle, we could permit Change Proposals that ask the editor to do something on a very high level, like "Try harder to align with other established specs" or something.  We don't allow that (I think for good reason), but we could.

[09:35:01.0000] <othermaciej>
all right, I have everything but the detailed error breakdown added for all the sites to http://www.w3.org/html/wg/wiki/HTML5_Authoring_Conformance_Study

[09:36:00.0000] <othermaciej>
many sites still need the detailed breakdown though

[09:36:01.0000] <Dashiva>
High-level changes like that are unlikely to succeed, since nobody agrees 100% with the editor

[09:36:02.0000] <Dashiva>
So we'll just end up going back a second time and fixing all the little pieces...

[09:38:00.0000] <AryehGregor>
Dashiva, right.  That's why we shouldn't allow such changes.  But we could allow them in principle anyway, and have the editor ask the chairs for clarification on specifics, say.

[09:38:01.0000] <othermaciej>
There's two ways in which an issue can be "large"

[09:38:02.0000] <othermaciej>
one is requiring a large diff to the spec, even though it can be stated relatively succinctly

[09:39:00.0000] <othermaciej>
that seems fine as a bug, though if it gets rejected, writing the Change Proposal will be potentially a lot of work

[09:39:01.0000] <othermaciej>
another is that it has many subproblems

[09:39:02.0000] <othermaciej>
for that, I would say the only good tools we have for breaking such issues down are informal ones (mailing list discussion, wikis, etc)

[09:41:00.0000] <othermaciej>
ideally we shouldn't have issues at this point that they are so complicated that they need extensive work to break them down before they can be processed, but yet are so critical that they must be addressed before Last Call

[09:41:01.0000] <othermaciej>
because we're supposed to be trying to converge right now

[09:41:02.0000] <Dashiva>
supposed to :)

[09:43:00.0000] <MikeSmith>
AryehGregor: about "ask the W3C validator team if we could work out some way for HTML5 to validate with an obsolete-but-conforming doctype"

[09:43:01.0000] <AryehGregor>
/me waits with bated breath

[09:44:00.0000] <Dashiva>
/me admires AryehGregor's spelling

[09:44:01.0000] <othermaciej>
AryehGregor: can't you do that using the drop-down on the validator page?

[09:44:02.0000] <MikeSmith>
part the way is towards that is for users to first explicitly choose HTML5 validation

[09:44:03.0000] <Philip`>
<meta http-equiv="X-UA-Compatible" content="w3validator=5">

[09:44:04.0000] <othermaciej>
under "More Options"

[09:44:05.0000] <Philip`>
/me can't remember whether the syntax as implemented in IE requires you to use a number or not

[09:44:06.0000] <Philip`>
(and of course it's undocumented)

[09:45:00.0000] <Philip`>
(and of course Chrome Frame implements it completely differently)

[09:45:01.0000] <AryehGregor>
othermaciej, sure, but I would like it to be autodetected correctly.  People will check if it passes and say "Oh, Wikipedia uses invalid markup", not "Oh, it must be using HTML5 with an obsolete but conforming doctype, let me manually choose HTML5 and see if it validates when I do that."

[09:45:02.0000] <MikeSmith>
if a document has an HTML4 doctype and it's validate HTML4 and you don't ask the W3C validator to validate it as HTML5, then it's going to validate it as HTML4

[09:45:03.0000] <AryehGregor>
s/Wikipedia/MediaWiki/ too.

[09:46:00.0000] <othermaciej>
AryehGregor: do you want it to try HTML4 validation first and then HTML5 if there's an error?

[09:46:01.0000] <MikeSmith>
AryehGregor: they odds of the W3C validator automatically switching to checking invalid HTML4 documents as HTML5 are extremely low

[09:46:02.0000] <othermaciej>
(assuming it's an HTML4 doctype)

[09:46:03.0000] <AryehGregor>
othermaciej, that was my first thought, but MikeSmith pointed out that that would mean it would check HTML5 an awful lot pointlessly.

[09:47:00.0000] <othermaciej>
AryehGregor: not only that, but it would not be a good change for people who actually do want to validate HTML4

[09:47:01.0000] <MikeSmith>
we might be able to get to the validator to prompt users and ask them if they want to re-validate a doc as HTML5

[09:47:02.0000] <AryehGregor>
MikeSmith, I'm currently thinking that perhaps it could be validated as HTML5 if 1) it has an XHTML 1.1 doctype, and 2) it has no xmlns attribute on the <html> element.  Kind of a hack, but unlikely to be disruptive.  Such documents cannot be valid XHTML 1.1.  In case it *was* meant to be an XHTML 1.1 document, a note can be output explaining the situation and saying to add a correct xmlns if you really want XHTML 1.1.

[09:47:03.0000] <AryehGregor>
(I assume XHTML 1.1 still requires xmlns, like XHTML 1.0.)

[09:47:04.0000] <MikeSmith>
othermaciej: yeah, it would mean that every invalid HTML4 doc would get checked twice as many times as it is now

[09:48:00.0000] <MikeSmith>
which would not seem like a particularly good idea

[09:48:01.0000] <MikeSmith>
AryehGregor: doing some kind of heuristic checks like that might be acceptable

[09:48:02.0000] <Dashiva>
AryehGregor: Who would create an XHTML 1.1 document without xmlns?

[09:49:00.0000] <AryehGregor>
Dashiva, they wouldn't, so it must be HTML5 with a conforming but obsolete doctype.

[09:49:01.0000] <AryehGregor>
That's the point.  :)

[09:49:02.0000] <Dashiva>
But why would the page exist in the first place?

[09:49:03.0000] <AryehGregor>
Currently MediaWiki uses XHTML 1.0 Strict, but XHTML 1.1 is less common, so would be a better choice.

[09:49:04.0000] <AryehGregor>
Because it can't use <!DOCTYPE html> because it wants to be well-formed XML and needs to have entity declarations.

[09:50:00.0000] <AryehGregor>
It's either that or try to get rid of every single &nbsp; and whatever, which is horribly error-prone.

[09:50:01.0000] <othermaciej>
do browsers give html entities under the XHTM 1.1 doctype?

[09:50:02.0000] <AryehGregor>
I'd have to test that.  If not, I could stick with XHTML 1.0 Strict.

[09:50:03.0000] <Dashiva>
So it's actually non-XHTML XML with an entity hack

[09:51:00.0000] <AryehGregor>
It's HTML5 that's also well-formed XML, is the idea, yes.

[09:52:00.0000] <Philip`>
Given that people are going to process it with XML tools, it seems bad to not have the XHTML xmlns

[09:52:01.0000] <Dashiva>
They'll process it with generic XML tools that don't care about namespaces, I bet :)

[09:52:02.0000] <MikeSmith>
I am personally not interesting at all in putting any more time and energy and money into further integrating HTML5 checking into the existing W3C validator.. good money after bad, and all that .. what we need to do instead is get a production instance of the HTML5 running on www.w3.org, and start to do promotion of that, and awareness-raiwing about what it's meant to be used for

[09:52:03.0000] <AryehGregor>
Correct.

[09:52:04.0000] <AryehGregor>
Just basic parsers.

[09:52:05.0000] <Philip`>
It'll also be harder for them to switch their tools to a proper HTML5 parser (which does give namespaces)

[09:53:00.0000] <AryehGregor>
MikeSmith, I can see that makes sense from your point of view, but from my point of view, it's important that MediaWiki and Wikipedia be perceived as standards-compliant.  Web standards are part of Wikimedia's mission and something I care about, and they need what PR they can get.

[09:53:01.0000] <MikeSmith>
AryehGregor: well, they are not standards-compliant

[09:53:02.0000] <AryehGregor>
In what way are they not?

[09:53:03.0000] <MikeSmith>
so you have a conflicting goal

[09:54:00.0000] <MikeSmith>
HTML5 is not a standard

[09:54:01.0000] <AryehGregor>
Draft-standard-compliant.

[09:54:02.0000] <AryehGregor>
I don't buy the W3C's insistence on glacial progression of standards along a formal track.

[09:54:03.0000] <MikeSmith>
geez

[09:54:04.0000] <MikeSmith>
the W3C is me, man

[09:54:05.0000] <MikeSmith>
me

[09:55:00.0000] <othermaciej>
if you want to get really technical, nothing the w3c publishes is a "standard"

[09:55:01.0000] <MikeSmith>
and other people

[09:55:02.0000] <AryehGregor>
:P

[09:55:03.0000] <MikeSmith>
AryehGregor: it is not a faceless blob of idiocy

[09:55:04.0000] <AryehGregor>
Draft standards that are interoperably implemented have pretty much all the benefits of actual standards, and it's not reasonable to avoid using them just because some totally unrelated features that happen to be in the same spec are not finalized yet.

[09:56:00.0000] <AryehGregor>
othermaciej, I think the W3C publishes standards in the dictionary sense of that term.

[09:56:01.0000] <AryehGregor>
Actually, HTML5 is a de facto standard at least, no matter what (at least the parts Wikipedia would bother using).

[09:56:02.0000] <Philip`>
othermaciej: plh said they're standards, in http://www.w3.org/QA/2009/05/_watching_the_google_io.html

[09:56:03.0000] <Philip`>
(comments)

[09:56:04.0000] <othermaciej>
I guess it's fair to say HTML5 is a conspicuous object (as a banner) formerly carried at the top of a pole and used to mark a rallying point especially in battle or to serve as an emblem

[09:58:00.0000] <AryehGregor>
It's that, too.

[09:59:00.0000] <AryehGregor>
Anyway, the WHATWG calls HTML5 a Draft Standard.

[09:59:01.0000] <AryehGregor>
Good enough for me.

[09:59:02.0000] <Philip`>
It started calling it a Standard after the aforementioned blog post

[10:01:00.0000] <AryehGregor>
Yeah, I saw in the comments.

[10:01:01.0000] <othermaciej>
that blog post has a giant flamewar in the comments, yay

[10:02:00.0000] <AryehGregor>
"An essential part of the Wikimedia Foundation's mission is encouraging the development of free-content educational resources that may be created, used, and reused by the entire human community.  We believe that this mission requires thriving open formats and open standards on the web to allow the creation of content not subject to restrictions on creation, use, and reuse."

[10:02:01.0000] <AryehGregor>
Maybe HTML5 is an "open format" if it's not an "open standard"?

[10:02:02.0000] <AryehGregor>
/me doesn't care, is calling it a standard regardless

[10:03:00.0000] <othermaciej>
if you want to market your compliance with HTML5, you probably need to do that actively by telling people how to check

[10:03:01.0000] <othermaciej>
I don't think it's the W3C's job to make it as easy as possible for you to impress people

[10:05:00.0000] <AryehGregor>
It's not about impressing people, it's about evangelism.  It's kind of hard for an organization to advocate web standards if it's perceived as not complying with them.

[10:06:00.0000] <AryehGregor>
Anyway, regardless of us, it's misleading to validator users to say that the document is non-standard XHTML 1.0 Strict when it's actually valid HTML5.  You could also say it's non-standard HTML 3.2, but that would be silly.

[10:06:01.0000] <othermaciej>
is there a place where I can find information on the standards Wikipedia (and/or MediaWiki) claims to conform to?

[10:07:00.0000] <AryehGregor>
Nowhere official.  Everything I've said here is my own opinion as a volunteer.  Historically, MediaWiki has attempted to conform to XHTML 1.0 Transitional, and now I'm trying to get it to conform as HTML5.

[10:08:00.0000] <AryehGregor>
It also tries to conform to CSS and so on, where practical.

[10:09:00.0000] <othermaciej>
I guess what I'm saying is, if it's important for Wikipedia to know about your standards compliance, it seems like one obvious first step would be for Wikipedia to document its own standards compliance efforts in a publicly visible way

[10:09:01.0000] <othermaciej>
but anyway I guess I have said enough on this topic

[10:09:02.0000] <AryehGregor>
Validators are what people use to judge standards compliance.

[10:10:00.0000] <Philip`>
Maybe MediaWiki should try to evangelise better ways of judging standards compliance

[10:10:01.0000] <AryehGregor>
That's the problem.  Few people understand anything about the standards, they rely on validators to tell them whether something is conforming.

[10:12:00.0000] <AryehGregor>
Wikimedia seems to mostly be focusing its evangelism efforts right now on Theora, although by now it's unfortunately turned out to be a lost cause (IMO).

[10:13:00.0000] <AryehGregor>
Although standards evangelism is far from a primary goal of Wikimedia, anyway.

[10:13:01.0000] <othermaciej>
I don't think wikipedia is the proper organization to evangelize standards

[10:14:00.0000] <othermaciej>
I'm thinking of a page that says things like "Wikipedia strives to be compliant to web standards including HTTP 1.1, HTML5 and CSS 2.1. Here are some ways in which we don't always meet that goal. Here is how to check for yourself." etc

[10:15:00.0000] <othermaciej>
writing is hard though and if the list grows to be more than a shortlist it can spawn excess debate

[10:15:01.0000] <othermaciej>
I guess some form of SVG should also be in that list

[10:16:00.0000] <AryehGregor>
That seems superfluous.  Isn't just saying "we try to be standards-compliant" enough?  It's pretty clear what HTML standard we use (currently XHTML 1.0 Transitional), and there are really no major conflicts between the versions of other standards (CSS, HTTP, etc.).  You can just assume the latest.

[10:16:01.0000] <AryehGregor>
(not that we find it necessary to officially say we try to be standards-compliant, we just do it, where it's reasonable)

[10:16:02.0000] <AryehGregor>
Does WebKit have an official list of standards that it tries to implement correctly?

[10:22:00.0000] <othermaciej>
we have tried to make such a list for WebKit but it started getting pretty long and hard to maintain

[10:22:01.0000] <othermaciej>
http://trac.webkit.org/wiki/SpecSupport

[10:22:02.0000] <AryehGregor>
Opera and IE have such lists, I think.

[10:22:03.0000] <othermaciej>
extremely incomplete list

[10:22:04.0000] <AryehGregor>
It doesn't seem like a worthwhile use of time to me.

[10:23:00.0000] <othermaciej>
and out of date

[10:23:01.0000] <othermaciej>
there's probably at least 50 things that could be added to that list

[10:24:00.0000] <othermaciej>
I think we need to have a more complete and up-to-date list though, if only for our own purposes

[10:24:01.0000] <AryehGregor>
Well, implementers tend to refer to actual specs a lot more than authors.  If anything, authors mostly just get it to validate and leave it at that.

[10:25:00.0000] <AryehGregor>
Maybe it's more useful to implementers to keep tabs on exactly what specs they need to refer to.

[10:25:01.0000] <AryehGregor>
It doesn't seem useful to me as an author.

[10:26:00.0000] <othermaciej>
it's useful for us so that we can keep track of whether patches are part of a standard we approve of, or whether they need closer examination

[10:26:01.0000] <othermaciej>
also to document the state of things for content authors, because they seem interested in what standards we support

[11:16:00.0000] <MikeSmith>
[me finally has a short meeting break]

[11:18:00.0000] <MikeSmith>
AryehGregor: fwiw, I will be working actively on getting a production instance of validator.nu set up in W3C webspace

[11:18:01.0000] <AryehGregor>
Great.

[11:19:00.0000] <MikeSmith>
but I will not be working actively on any further changes to the existing markup validator

[11:19:01.0000] <MikeSmith>
for that your best bet is to talk with Ville Skyttä

[11:19:02.0000] <AryehGregor>
Sure.  Thanks.

[11:20:00.0000] <MikeSmith>
there is not really any team of people working on the existing validator

[11:20:01.0000] <MikeSmith>
it is basically just Ville Skyttä

[11:20:02.0000] <MikeSmith>
/me didn't know about http://trac.webkit.org/wiki/SpecSupport

[11:21:00.0000] <othermaciej>
MikeSmith: it's very out of date so don't use it as a reference

[11:22:00.0000] <othermaciej>
but maybe it will be one of the next things I decide to fix up

[11:22:01.0000] <MikeSmith>
othermaciej: I just tweeted it

[11:23:00.0000] <othermaciej>
great, now I have to fix it

[11:23:01.0000] <MikeSmith>
It seems worth trying to keep up-to-date

[11:23:02.0000] <MikeSmith>
othermaciej: I would be willing to help keep it up to date

[11:25:00.0000] <MikeSmith>
othermaciej: or I also can delete that tweet

[11:26:00.0000] <MikeSmith>
/me deletes it

[11:26:01.0000] <othermaciej>
it's not actually that big a deal, sorry for making it sound like one

[11:27:00.0000] <othermaciej>
/me is loopy from lack of sleep and from stress over we will ever get to Last Call

[11:28:00.0000] <MikeSmith>
/me has spent quite a bit of time the previous several days discussing progress of HTML5 to Last Call

[11:28:01.0000] <MikeSmith>
othermaciej: I've been here in Cambridge since the 17th

[11:28:02.0000] <MikeSmith>
and will be here through the 31st

[11:29:00.0000] <othermaciej>
neat

[11:29:01.0000] <othermaciej>
I wonder what Hixie has been up to

[11:30:00.0000] <othermaciej>
there has not been much bug closing for a while

[11:30:01.0000] <othermaciej>
though we really need to get more issues closed out too

[11:32:00.0000] <MikeSmith>
Hixie seems to have had his head down doing especially concentrated work on something for a while now

[11:34:00.0000] <othermaciej>
seems like a pretty low volume of spec commits this month

[12:17:00.0000] <AryehGregor>
/me notes that Chrome and Gmail both offer to translate his e-mails now, except that Gmail is smart enough not to keep offering after he's gone on to the next e-mail

[12:18:00.0000] <MikeSmith>
AryehGregor: do you get a lot of e-mail in different languages?

[12:18:01.0000] <AryehGregor>
Yes, I get localization update e-mails every day from the MediaWiki commit list.

[12:19:00.0000] <AryehGregor>
Like: http://lists.wikimedia.org/pipermail/mediawiki-cvs/2010-March/085047.html

[12:19:01.0000] <AryehGregor>
Or: http://lists.wikimedia.org/pipermail/mediawiki-cvs/2010-March/085041.html

[12:19:02.0000] <AryehGregor>
I don't actually care to read them, obviously.

[12:41:00.0000] <MikeSmith>
I didn't know Chrome itself actually had a means for offering to translate a page

[12:42:00.0000] <MikeSmith>
I guess I need to start using it more

[12:42:01.0000] <AryehGregor>
They just added it to the dev channel.

[12:42:02.0000] <MikeSmith>
ah, OK

[12:43:00.0000] <AryehGregor>
/me has found the dev channel a bit too unstable for his liking, but is too lazy to switch to beta

[12:43:01.0000] <MikeSmith>
I try to mix up my browsing across as many major browsers/engines as I can

[12:43:02.0000] <MikeSmith>
but I find I end up using some a lot more than others

[12:45:00.0000] <AryehGregor>
I just use Chrome where possible now.  At least on Linux, I find it feels much more responsive than Firefox.

[12:45:01.0000] <AryehGregor>
And I never was an extension junkie, extensions are unreliable and slow.

[12:46:00.0000] <AryehGregor>
Firefox is second-place for me now.

[12:52:00.0000] <nessy>
can't live without firebug!

[12:53:00.0000] <Dashiva>
Does Chrome have live source editing within the browser?

[12:54:00.0000] <AryehGregor>
It has WebKit's Web Inspector.

[12:54:01.0000] <AryehGregor>
I've found it works about as well as Firebug.

[12:54:02.0000] <AryehGregor>
Maybe I missed out on some of Firebug's hard-core features.

[12:54:03.0000] <AryehGregor>
On the plus side, Web Inspector doesn't break all the time when you upgrade, and you can be sure it's not sucking up resources because everyone has it installed.

[12:56:00.0000] <AryehGregor>
I think Firebug is better on that than it used to be.  I remember a few years ago getting a warning from Gmail along the lines of "You're using Firebug, we suggest you don't because it tends to slow down Gmail horribly."

[12:57:00.0000] <AryehGregor>
That was before they made some things disabled by default, like network monitoring.

[13:04:00.0000] <Dashiva>
A minor diversion: http://www.foxtrot.com/comics/2010-03-21-75f89edd.gif

[13:18:00.0000] <MikeSmith>
nessy: Web Inspector works pretty well and is definitely worth trying

[13:19:00.0000] <MikeSmith>
AryehGregor: I guess Web Inspector is one of the features that leads me to end up using a particular browser/engine more than others

[13:20:00.0000] <nessy>
yeah, it's good that most browsers now have something like firebug

[13:20:01.0000] <nessy>
didn't used to be the case

[13:20:02.0000] <MikeSmith>
yep

[13:20:03.0000] <MikeSmith>
Firebug was really a big part of the inspiration for all the others

[13:21:00.0000] <MikeSmith>
I just wish that all browsers had development/debug tools built-in

[13:22:00.0000] <MikeSmith>
among other reasons, it just makes upgrading a lot easier

[13:22:01.0000] <MikeSmith>
if the debugging tools get tested a part of the normal build/release cycle

[13:23:00.0000] <MikeSmith>
including with the development builds

[13:25:00.0000] <Philip`>
Which browsers don't have them built-in?

[13:26:00.0000] <Philip`>
Opera has Dragonfly, IE has its developer tools, Safari/Chrome have Web Inspector

[13:26:01.0000] <Philip`>
Firefox is the only one that doesn't

[13:26:02.0000] <Philip`>
and installing Firebug is trivial

[13:29:00.0000] <Dashiva>
I'm looking for the equivalent of Opera's view source -> edit -> apply changes

[13:30:00.0000] <Philip`>
View source, copy, http://software.hixie.ch/utilities/js/live-dom-viewer/, paste

[13:32:00.0000] <Dashiva>
That doesn't work when doing script + stylesheet interaction :)

[13:33:00.0000] <Dashiva>
E.g. just now there was this script that failed in Opera, so I added a line to show the error condition and got the answer right away

[14:07:00.0000] <Philip`>
Hmph, I beat annevk by two minutes but used the wrong email address so my message got held up :-(

[14:08:00.0000] <annevk>
not here

[14:09:00.0000] <annevk>
teaches me to write email after 10PM

[14:31:00.0000] <jgraham>
FWIW I think the block-inside-<a> thing is probably a bad idea

[14:32:00.0000] <jgraham>
Because it doesn't work in a lot of cases to trying to use it has inconsistent results

[14:32:01.0000] <jgraham>
OTOH I don't feel like it is worth some huge war

[16:11:00.0000] <Dashiva>
View source and find to do page validation... right

[16:16:00.0000] <TabAtkins>
I don't want to violate my policy of "ignore Shelley", but yeah, that one *almost* got a reply out of me.

[16:17:00.0000] <Dashiva>
I have a better policy: Let's just write conforming documents to begin with

[16:30:00.0000] <Philip`>
Dashiva: HTML WG experts can often write mostly-conforming HTML by hand, so it seems best to design the spec around the expectation that everyone is like that


2010-03-22
[20:23:00.0000] <othermaciej>
TabAtkins: yeah I couldn't resist replying to that one, even though I should have

[23:37:00.0000] <lazni>
https://gist.github.com/f7d3e10233d436eb16dc comment on HTML6, I can't distill it down in one IRC line

[00:52:00.0000] <boblet>
krijnh: thanks for upgrading your sewing machine ;-) logs fully loading again

[00:57:00.0000] <JonathanNeal>
Hello all

[00:58:00.0000] <hsivonen>
hi

[00:59:00.0000] <JonathanNeal>
what's the haps, how goes html5?

[01:06:00.0000] <virtuelv>
huh?

[01:06:01.0000] <virtuelv>
http://www.w3.org/html/wg/wiki/HTML5_Authoring_Conformance_Study - why are there fewer, and different conformance errors on google.de compared to other googles?

[01:06:02.0000] <virtuelv>
s/, and different//

[01:22:00.0000] <hsivonen>
I need to switch to an OpenID provider whose authentication page doesn't break in Minefield.

[01:24:00.0000] <JonathanNeal>
So Wikipedia, Mozilla, and WHATWG get it right, virtuelv.

[01:29:00.0000] <virtuelv>
no, many sites gets it right

[01:29:01.0000] <virtuelv>
with regards to their declared doctypes and intended goals

[01:32:00.0000] <JonathanNeal>
well yes :)

[01:34:00.0000] <virtuelv>
the thing I find some comfort in, is that for the sites with many errors, the error count goes down when validated against HTML5

[01:39:00.0000] <boblet>
virtuelv: the very same reason will no doubt make many people agitated ;-)

[01:41:00.0000] <boblet>
hsivonen: are you planning to add an outliner to validator.nu?

[01:46:00.0000] <JonathanNeal>
virtuelv, how is that a good thing?

[01:46:01.0000] <JonathanNeal>
You think they'll be encouraged to fix their bugs if they have 10 instead of 20?

[01:46:02.0000] <JonathanNeal>
The most common validation error I see on corporate sites is improperly escaped / unescaped text.

[02:08:00.0000] <hsivonen>
boblet: yes, in some far future after a Firefox release has been shipped with the HTML5 parser

[02:08:01.0000] <boblet>
hehe

[02:09:00.0000] <boblet>
hsivonen: here’s hoping that far future isn’t too far, eg for block-level links etc

[02:13:00.0000] <hsivonen>
boblet: I thought we were going to enable the HTML5 parser on the trunk by the end of this month. However, now it seems that that's not going to happen. 

[02:14:00.0000] <hsivonen>
boblet: I don't really know what the time scales are

[02:14:01.0000] <boblet>
hsivonen: that sucks ;( did something big come up to stop it from happening?

[02:15:00.0000] <hsivonen>
boblet: we don't have continuous integration test coverage for speculative parsing

[02:15:01.0000] <boblet>
hsivonen: well … that_sounds_ big… :? hehe

[02:16:00.0000] <boblet>
or at least impressive

[05:26:00.0000] <Dashiva>
This seems to be somehwat overstating the abilities of CSS transforms, or am I wrong? "Actually, because CSS3 implements the same functionality as the canvas element (which makes me believe that the canvas element becomes redundant then)"

[05:42:00.0000] <annevk>
presumably you can implement <canvas> with <div> and CSS3

[05:42:01.0000] <annevk>
but then you can implement CSS3 with <font> and DOM

[05:42:02.0000] <annevk>
and sufficiently small font sizes

[05:47:00.0000] <jgraham>
So we conclude that you can implement <canvas> with <font> and DOM and sufficiently small font sizes

[05:47:01.0000] <annevk>
indeed, same for SVG btw

[05:47:02.0000] <annevk>
I wonder why we bothered with all this stuff in the first place

[05:48:00.0000] <annevk>
anyway, time to go

[05:48:01.0000] <annevk>
new passport + presentation coming up

[06:03:00.0000] <hsivonen>
with <font> going through DirectWrite, you even get GPU acceleration if you implement stuff on top of tiny fonts!

[06:17:00.0000] <virtuelv>
/me goes back to backlog JonathanNeal / boblet : my point is that the HTML5 spec is closer to specifying real authoring practices

[06:17:01.0000] <virtuelv>
if everyone ceased from creating new documents on the web, HTML5 UAs should eventually become more compatible with that content

[06:18:00.0000] <virtuelv>
*interoperable

[06:19:00.0000] <boblet>
virtuelv: is this re: the wiki authoring conformance study?

[06:19:01.0000] <virtuelv>
boblet: yes

[06:20:00.0000] <virtuelv>
of course, people will debate whether a specification that actually takes common legacy implementations into account is jumping the shark or not

[06:21:00.0000] <boblet>
my take was that those who e.g. advocate draconian error handling will see this as a watering down, a move in the wrong direction

[06:22:00.0000] <virtuelv>
I don't see how anyone can advocate draconian error handling

[06:22:01.0000] <Philip`>
Dashiva: I expect doing e.g. dynamic edge filtering of videos would be quite tricky with CSS transforms

[06:22:02.0000] <virtuelv>
(but that's a different matter)

[06:23:00.0000] <virtuelv>
draconian error handling is only possible on the syntactic level when you abstract away the syntax and let tools serialize your documents

[06:24:00.0000] <boblet>
or when you start from scratch and decree it

[06:25:00.0000] <Philip`>
boblet: Like with RSS?

[06:26:00.0000] <virtuelv>
/me has his name in RFC 4287

[06:26:01.0000] <virtuelv>
rss is still not gone :/

[06:26:02.0000] <boblet>
had TBL made draconian error handling a basic tenet of HTML from the outset it’s possible that the web would be using it

[06:26:03.0000] <virtuelv>
boblet: 

[06:26:04.0000] <virtuelv>
the short answer is "absolutely not"

[06:27:00.0000] <Philip`>
boblet: I expect the browser developers would have just ignored him

[06:27:01.0000] <boblet>
but I completely certain the web would be a shadow of its current state

[06:27:02.0000] <virtuelv>
my thesis is that had TBL made error handling draconian, the web wouldn't have taken off

[06:27:03.0000] <Philip`>
since they could make their products better for authors and users by violating the spec

[06:27:04.0000] <boblet>
virtuelv: yep, that’s what I think

[06:28:00.0000] <boblet>
Philip`: I don’t think there would have been a need, as browsers would only have been made/bought by academia and hobbyists

[06:28:01.0000] <Philip`>
http://www.walterzorn.com/jsgraphics/jsgraphics_e.htm - <canvas>-style graphics using <div>

[06:28:02.0000] <Philip`>
so it's not merely possible, it's already implemented

[06:30:00.0000] <virtuelv>
Philip`: I seem to recall digg at some point using that library

[07:34:00.0000] <Philip`>
/me continues to think that HTML5 shouldn't defer to RFC2119 for definitions of "must" etc, because RFC2119 doesn't make sense in the context of authoring requirements

[07:35:00.0000] <MikeSmith>
hmm, yeah, true

[07:36:00.0000] <Philip`>
(It's written about vendors implementing behaviours)

[07:37:00.0000] <MikeSmith>
Philip`: is it really, though?

[07:37:01.0000] <MikeSmith>
I'm reading it now and I don't immediately see anything that makes it bound so closely to implementations

[07:38:00.0000] <MikeSmith>
to being about requirements for implementors

[07:38:01.0000] <Philip`>
"the full implications should be understood and the case carefully weighed before implementing any behavior described with this label"

[07:38:02.0000] <MikeSmith>
ohg

[07:38:03.0000] <Philip`>
That's about implementing behaviours

[07:38:04.0000] <MikeSmith>
yeah, and "enhances the product while another vendor may omit the same item"

[07:39:00.0000] <Philip`>
The "MAY" definition is all about vendors and implementations

[07:39:01.0000] <MikeSmith>
right

[07:39:02.0000] <MikeSmith>
I see that now

[07:39:03.0000] <MikeSmith>
the Security Considerations section also

[07:40:00.0000] <Philip`>
Some of the definitions are more generic, but if the goal was for RFC2119 to apply to non-implementations requirements too then presumably the whole thing would have been written in a more generic way

[07:40:01.0000] <MikeSmith>
yes

[07:40:02.0000] <MikeSmith>
though perhaps it doesn't really matter what the original intent was

[07:41:00.0000] <Philip`>
It matters if people start trying to selectively interpret the intent of the original phrasing

[07:42:00.0000] <jgraham>
FWIW I agree with Philip` to the extent that RFC 2119 is clearly not intended to deal with authoring requirments when the client is liberal

[07:42:01.0000] <jgraham>
(or perhaps in general)

[07:42:02.0000] <MikeSmith>
Philip`: it is a mistake for them to do that

[07:42:03.0000] <jgraham>
I don't think this is a big problem except insofaras people try to spac lawyer as if it was

[07:42:04.0000] <Philip`>
I think the use of must/should/etc for authoring requirements is useful and sensible and good, but specs should be clear that it's a wilful violation of RFC2119

[07:43:00.0000] <Philip`>
so that people can't validly claim that the spec is silently inconsistent with the other specs it normatively references

[07:45:00.0000] <jgraham>
/me would prefer people just stopped that because it is obviously a trivial inconsistency, hasn't been a problem for other specs, and only matters to spec lawyers anyway

[08:04:00.0000] <Philip`>
/me wonders if 'it should be feasible to explain to authors exactly how to write valid HTML' is one of the criteria that Sam is missing

[08:06:00.0000] <Philip`>
(where 'to write an attribute value containing a '&', you can write it escaped or you can write '&' unless it's followed by one of these 106 strings of characters in unescaped form' is not considered a feasible explanation)

[08:07:00.0000] <Philip`>
(...'unless those strings are themselves followed by an alphanumeric character' or whatever the rules are now)

[08:28:00.0000] <MikeSmith>
interesting that an engineer from Microsoft Research is implementing a conformant HTML5 parser

[08:28:01.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/public-html-comments/2010Mar/0014.html

[08:28:02.0000] <MikeSmith>
I wonder what language he's coding it in

[08:31:00.0000] <Philip`>
http://langnetsymposium.com/2009/speakers.aspx - "Prior to joining MSR, Herman worked in the Developer Division, initially on JScript .NET. After completing the JScript .NET project Herman worked in collaboration with MSR on Comega and Spec#. Since joining MSR on a full time basis, Herman has worked on Spec#, the Verifying C compiler, CCI version 2 and lately on JScript again (initially in the context of the Gazelle browser project)."

[08:32:00.0000] <Philip`>
Sounds like he largely works on stuff around C#, so maybe that

[08:33:00.0000] <MikeSmith>
that would make sense, I guess

[08:33:01.0000] <Philip`>
Gazelle uses C# too

[08:34:00.0000] <Philip`>
(but reusing parts of Trident, like the parser, in the version in the published paper)

[08:38:00.0000] <MikeSmith>
I seem to vaguely remember previously hearing about someone already working on a C# HTML5 parser

[08:38:01.0000] <MikeSmith>
but maybe I imagined it

[08:38:02.0000] <MikeSmith>
anyway, it would seem to be nice to have one

[08:39:00.0000] <Philip`>
http://code.google.com/p/twintsam/

[08:41:00.0000] <MikeSmith>
ah yeah

[08:41:01.0000] <MikeSmith>
I rememeber that name now

[08:42:00.0000] <MikeSmith>
Thomas Broyer

[08:42:01.0000] <MikeSmith>
me tries to remember what "twinstam" means

[08:45:00.0000] <Philip`>
The Web Is Not Tag Soup Any More

[08:46:00.0000] <Dashiva>
Soup Tags?

[08:50:00.0000] <lazni>
<water><rice/><meat state=minced/></water>

[08:51:00.0000] <Dashiva>
I'd use a generic <minced> wrapper element to allow mincing arbitrary ingredients

[09:06:00.0000] <MikeSmith>
heh

[09:06:01.0000] <MikeSmith>
I want a <mike> tag

[09:07:00.0000] <MikeSmith>
we have that tag named after Mark Pilgrim already

[09:07:01.0000] <MikeSmith>
so it would only seem fair

[09:08:00.0000] <beowulf>
if you write enough documents with a <mike> tag doesn't it have to be in html6?

[09:08:01.0000] <Dashiva>
Not if it doesn't have special behavior

[09:08:02.0000] <beowulf>
i could wave my hands around, would that get it in?

[09:09:00.0000] <MikeSmith>
/me ponders what the behavior of the <mike> tag should be

[09:10:00.0000] <lazni>
<device type=microphone>

[09:22:00.0000] <Wolfman2000>
Morning/afternoon. As a reminder, is this the primary HTML5 channel on IRC?

[09:23:00.0000] <lazni>
/me looks at the channels on the log site in /topic

[09:23:01.0000] <Philip`>
It's where most discussion about the spec takes place

[09:24:00.0000] <Dashiva>
http://masinter.blogspot.com/2010/03/browsers-are-rails-web-sites-are-trains.html

[09:24:01.0000] <Wolfman2000>
...then perhaps you guys can be of assistance. I'm still a bit confused by the purpose of the <canvas> element. Is it supposed to allow for drawing whatever you want on it?

[09:25:00.0000] <Dashiva>
It seems rail/train doesn't really match how HTML works. It seems more like ocean/anything-that-floats

[09:25:01.0000] <Dashiva>
(It seems I like saying "It seems")

[09:25:02.0000] <Philip`>
Wolfman2000: Yes

[09:25:03.0000] <lazni>
programmatically

[09:25:04.0000] <Philip`>
Yes

[09:26:00.0000] <lazni>
no SVG importing?

[09:26:01.0000] <Wolfman2000>
Can this canvas be given a width and height, but with a height "cap" of sorts to force scrolling in one direction or another?

[09:26:02.0000] <Dashiva>
lazni: You can drawImage svg onto a canvas

[09:26:03.0000] <Philip`>
/me was interpreting "you" as meaning a page developer, not a user

[09:26:04.0000] <Philip`>
Dashiva: Only in Opera

[09:26:05.0000] <Philip`>
(I think)

[09:26:06.0000] <Dashiva>
Probably

[09:26:07.0000] <Philip`>
and not in the spec

[09:26:08.0000] <Philip`>
(except via <img src=foo.svg>)

[09:26:09.0000] <Wolfman2000>
/me is more a web developer than a user these days.

[09:26:10.0000] <Dashiva>
But I expect it will spread

[09:27:00.0000] <Philip`>
Wolfman2000: It has a width and a height, just like an <img>

[09:27:01.0000] <lazni>
Wolfman2000: you want something like http://yuiblog.com/assets/slicing/vr_test.html?

[09:27:02.0000] <lazni>
blog post: http://www.yuiblog.com/blog/2008/06/23/slicing/

[09:27:03.0000] <Wolfman2000>
hmm...unsure the best way to phrase this.

[09:28:00.0000] <Philip`>
Wolfman2000: It's just a rectangle on the page (of the given width/height), and scrolls like any other static object on a page

[09:29:00.0000] <Wolfman2000>
I want to turn a java application that I have and make it use the power of HTML5.  http://beta.pumpproedits.com/java/PPEdits.zip  <-- that's the app and the libraries. It's meant for creating machine dance step data for an arcade game, and the charts scroll downward.

[09:30:00.0000] <Philip`>
/me can't easily run Java applications

[09:30:01.0000] <Wolfman2000>
I'll get you a screenshot then if I can.

[09:31:00.0000] <Philip`>
If you want some kind of scrolling within the content of the <canvas> (rather than having it be like a large <img> where the whole page can scroll to reveal parts of it), then you'll just have to write code to detect user input and to offset all your drawing functions

[09:31:01.0000] <Wolfman2000>
http://tinypic.com/3ia31x0i Not just scrolling. Adding elements.

[09:32:00.0000] <Wolfman2000>
Either way, I can "possibly" do this with <div>s and lots of <img>s. I just want to try to take advantage of what HTML5 has to offer.

[09:33:00.0000] <Philip`>
Might be easier to do it with SVG than with canvas

[09:33:01.0000] <JonathanNeal>
heyo

[09:34:00.0000] <Wolfman2000>
Philip`: javascript and SVG? Well, the charts that I generate are SVG already.

[09:35:00.0000] <Wolfman2000>
http://beta.pumpproedits.com/chart/quick/143/rhythm <-- one of the recent ones, showing off a new game mode. (Firefox views it best)

[09:35:01.0000] <Wolfman2000>
Still...I could have thought plain <svg> files couldn't have a <script> or <title> tag.

[09:35:02.0000] <Wolfman2000>
How exactly can I combine SVG and Javascript?

[09:38:00.0000] <Philip`>
Wolfman2000: You can use <script>s in SVG files, and also you can embed SVG inside an XHTML page (sent as application/xhtml+xml, not text/html yet) in order to use normal HTML form inputs and scripts and so on

[09:39:00.0000] <AryehGregor>
You can embed SVG in text/html in Firefox with html5.enable and in IE9, and probably in other browsers at some point.

[09:39:01.0000] <Wolfman2000>
html5.enable is at true right now

[09:40:00.0000] <Philip`>
Wolfman2000: Not by default

[09:40:01.0000] <Wolfman2000>
...wonder how I turned it on then

[09:40:02.0000] <AryehGregor>
Then you can stick <svg><circle r=100 full=red></circle></svg> in your HTML documents and get SVG as part of the DOM.  (Except I don't know SVG, so that might not work.)

[09:40:03.0000] <AryehGregor>
(Do void elements auto-close in SVG-in-text/html?)

[09:41:00.0000] <Philip`>
Restricting yourself to IE9 + Firefox-with-non-standard-options is probably a bit excessively anti-user

[09:41:01.0000] <AryehGregor>
Maybe.

[09:41:02.0000] <Wolfman2000>
Then please suggest a good alternative

[09:42:00.0000] <Philip`>
Doing it in application/xhtml+xml will work in every browser except IE8 (and older IEs)

[09:42:01.0000] <Philip`>
AryehGregor: You have to say <circle></circle> or <circle/>

[09:42:02.0000] <Philip`>
The parser doesn't know the details of all SVG elements

[09:43:00.0000] <AryehGregor>
Ah, so <circle /> does work.

[09:43:01.0000] <AryehGregor>
This works in Firefox with html5.enable: data:text/html,<!doctype html><svg width=200 height=200><circle cx=100 cy=100 r=100 fill=red /></svg>

[09:43:02.0000] <AryehGregor>
Hurrah.

[09:43:03.0000] <Wolfman2000>
Philip`: since I'm using HTML5 on my website primarily, what requires changing for it to work as application/xhtml+xml?

[09:43:04.0000] <AryehGregor>
Wolfman2000, IE < 9 doesn't work with application/xhtml+xml.

[09:43:05.0000] <AryehGregor>
Nor does it support SVG at all, in fact.

[09:43:06.0000] <Wolfman2000>
AryehGregor: besides IE

[09:43:07.0000] <Philip`>
AryehGregor: and circle isn't an empty element in SVG anyway

[09:43:08.0000] <AryehGregor>
Oh, then you just have to make sure it's well-formed, or else it will die with a fatal error.

[09:44:00.0000] <AryehGregor>
So it's kind of inconvenient.

[09:44:01.0000] <AryehGregor>
Probably easier to use <object> for now.  You can still have <script> inside the SVG file itself, as I understand it.

[09:45:00.0000] <Wolfman2000>
so HTML5 and XHTML use the same thing?

[09:46:00.0000] <Wolfman2000>
...I'll rephrase better. All I have to do is make sure my HTML5 is well formed and pass the right mime type, and it magically works?

[09:47:00.0000] <Philip`>
Yes, except it's not magic because you're following all the steps needed to make it work :-)

[09:48:00.0000] <Wolfman2000>
I wasn't expecting magic

[09:48:01.0000] <Wolfman2000>
I think most of my code is well formed as is at least.

[09:48:02.0000] <Philip`>
It's generally very difficult to guarantee well-formedness, so you should probably avoid it as much as possible

[09:48:03.0000] <Philip`>
and only use XHTML for pages that really require it

[09:49:00.0000] <Philip`>
(There's also quite a few subtle behavioural differences, which can break things)

[09:49:01.0000] <Wolfman2000>
Philip`: the only pages that require XML are the generated charts...and hopefully the new javascript chart editor I want to get working.

[09:49:02.0000] <Wolfman2000>
so...what's the best validator to use in this case? W3 doesn't offer XHTML5

[09:50:00.0000] <Philip`>
AryehGregor might be sensible in suggesting using <object> instead, so most of your page is plain old HTML and it just embeds a minimal amount of SVG XML in the <object>

[09:50:01.0000] <Philip`>
and you can put scripts in the outer HTML page, and have them manipulate the inner SVG page's DOM

[09:51:00.0000] <Philip`>
http://html5.validator.nu/ should work

[09:54:00.0000] <lazni>
Wolfman2000: also be carefull with user input

[09:54:01.0000] <lazni>
unicode encoding and such

[09:55:00.0000] <Wolfman2000>
...no more summary attribute? huh

[09:56:00.0000] <lazni>
data-summary and provide your own behaviour

[09:56:01.0000] <Wolfman2000>
I meant for <table>

[09:56:02.0000] <Wolfman2000>
(yes, I have a valid use for table)

[09:57:00.0000] <lazni>
I meant as a workaround

[09:57:01.0000] <Philip`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/tabular-data.html#attr-table-summary

[09:58:00.0000] <AryehGregor>
summary is allowed, although it's controversial.

[09:58:01.0000] <Philip`>
The bit above that has lots of suggested better ways of expressing summary-style text

[10:02:00.0000] <Wolfman2000>
Lots to consider...if nothing else, thanks for the updates

[12:34:00.0000] <JonathanNeal>
Hello World

[12:37:00.0000] <lazni>
XML parsing error: missing XML prolog

[12:38:00.0000] <JonathanNeal>
Sounds like my Hello World test worked better than yours, lazni :P

[13:17:00.0000] <gsnedders>
Hello world!

[13:27:00.0000] <annevk>
when is Ubuntu finally going to work with projectors normally?

[13:27:01.0000] <lazni>
maybe with KMS

[13:28:00.0000] <lazni>
I think Lucid will use KMS with nouveau and intel

[13:28:01.0000] <annevk>
I had to actually remove xorg.conf, log out, and log in again to fix issues after letting Ubuntu write to it to make it recognize my normal resolution again

[13:28:02.0000] <AryehGregor>
/me guesses it will be a nonissue in less than five years

[13:28:03.0000] <annevk>
also, several icons are now in the wrong place

[13:28:04.0000] <AryehGregor>
GNOME panels always randomly reorganize themselves when you log out and back in.

[13:28:05.0000] <AryehGregor>
It's infuriating.

[13:28:06.0000] <AryehGregor>
Xfce panels are far superior in all respects.

[13:29:00.0000] <annevk>
it's mostly weird they haven't fixed this

[13:29:01.0000] <annevk>
you'd think fundamental problems like this would have been fixed ages ago

[13:29:02.0000] <AryehGregor>
That's not the kind of remark I'd expect from a software developer.

[13:29:03.0000] <annevk>
using a projector is sort of an important use case

[13:30:00.0000] <AryehGregor>
I'd have thought you'd realize that things that seem obviously important to you might be less important or more difficult to fix than you'd think.

[13:30:01.0000] <jgraham>
Argh, now I've missed the chance to make the "more functional hello world" joke

[13:30:02.0000] <AryehGregor>
Most people don't use projectors ever.  You give talks all the time, so you use projectors regularly.

[13:30:03.0000] <AryehGregor>
What use does a normal person have for a projector?

[13:30:04.0000] <AryehGregor>
I don't think I've ever used one in my life.

[13:31:00.0000] <annevk>
lots of people give talks

[13:31:01.0000] <jgraham>
Which would of course be a fibonacci sequence generator...

[13:31:02.0000] <jgraham>
AryehGregor: Lucky you

[13:31:03.0000] <AryehGregor>
A tiny minority of people give talks.

[13:31:04.0000] <annevk>
www-tag email is now classified as spam

[13:31:05.0000] <annevk>
way to go Opera Mail

[13:31:06.0000] <jgraham>
AryehGregor: I'm not sure that is true

[13:32:00.0000] <AryehGregor>
annevk, wait, so is that a false positive or not?

[13:32:01.0000] <gsnedders>
/me has got rejeced by Edinburgh uni

[13:32:02.0000] <jgraham>
Presentations are common in business and in academia

[13:32:03.0000] <AryehGregor>
jgraham, I don't give talks, therefore nobody does.  Very simple.

[13:32:04.0000] <AryehGregor>
Yes, yes, I know.

[13:32:05.0000] <gsnedders>
*rejected

[13:32:06.0000] <annevk>
meetings and presentations are all the rage in corporate culture

[13:32:07.0000] <annevk>
also in university

[13:32:08.0000] <AryehGregor>
Most people are not in business or academia, or if they're in business they're too low on the pecking order to give talks.

[13:32:09.0000] <jgraham>
gsnedders: Oh. Their loss really

[13:32:10.0000] <annevk>
also what jgraham said

[13:32:11.0000] <AryehGregor>
gsnedders, :(

[13:32:12.0000] <AryehGregor>
Anyway, obviously Ubuntu hasn't received too many complaints about projectors not working, so that favors my hypothesis.

[13:32:13.0000] <annevk>
projectors work

[13:33:00.0000] <gsnedders>
jgraham: I told you I should've applied for English ;P

[13:33:01.0000] <AryehGregor>
Not working well.  Whatever.

[13:33:02.0000] <jgraham>
gsnedders: Don't make me come over there

[13:34:00.0000] <annevk>
it's just that a) it requires Ubuntu writing to xorg.conf b) because of that it requires logging out and logging in c) because of the xorg.conf changes my original resolution can no longer be selected (which is somewhat weird admittedly, 16:10) d) it therefore requires me to sudo rm xorg.conf and log out and log in again...

[13:34:01.0000] <annevk>
none of this should be needed, of course

[13:35:00.0000] <gsnedders>
jgraham: I guess I'll have to follow the example of others on the Carakan team and go to Glasgow :P

[13:35:01.0000] <AryehGregor>
Why does it need to write to xorg.conf?

[13:35:02.0000] <AryehGregor>
Doesn't X autoconfigure these days?

[13:37:00.0000] <AryehGregor>
/me has written a 566-word post in response to the authoring conformance thread: trim down or post?

[13:37:01.0000] <AryehGregor>
566 words isn't *that* long.

[13:38:00.0000] <gsnedders>
Shorter than what I end up writing without meaning too...

[13:38:01.0000] <AryehGregor>
/me trims it down some

[13:38:02.0000] <Dashiva>
AryehGregor: Do you have a blog?

[13:38:03.0000] <AryehGregor>
Dashiva, no.

[13:38:04.0000] <AryehGregor>
I've been meaning to start posting blog-like stuff on aryeh.name, but never got around to setting it up.

[13:38:05.0000] <AryehGregor>
I'm too elitist to use Wordpress or something, you see.

[13:38:06.0000] <AryehGregor>
So I'd have to write it.

[13:38:07.0000] <Dashiva>
You could write a summary and paste the full entry below it

[13:39:00.0000] <jgraham>
post === email?

[13:39:01.0000] <lazni>
/me writes an atom feed, doesn't need html

[13:40:00.0000] <annevk>
so based on the IP address google.com.hk is either in simplified or traditional Chinese?

[13:40:01.0000] <annevk>
/me didn't quite understand that part of the announcement

[13:40:02.0000] <Dashiva>
The IP address...

[13:41:00.0000] <annevk>
and why the difference?

[13:42:00.0000] <AryehGregor>
/me gets it down to 528 words and gives up

[13:43:00.0000] <jgraham>
/me wonders if he should reply to Sam

[13:48:00.0000] <AryehGregor>
Which post?

[13:49:00.0000] <AryehGregor>
"You seem to suggest that the policy is: scream loudly enough and you too can get your favorite pet peeve disallowed"

[13:49:01.0000] <AryehGregor>
Wait, that isn't official policy?

[13:49:02.0000] <Dashiva>
Sure seems to me

[13:53:00.0000] <JonathanNeal>
I like that policy.

[13:53:01.0000] <JonathanNeal>
no ...

[13:53:02.0000] <JonathanNeal>
I LOVE THAT POLICY!

[13:55:00.0000] <Dashiva>
You need more exclamation points

[13:57:00.0000] <gsnedders>
Maybe I should just stay in Sweden...

[14:03:00.0000] <JonathanNeal>
exclamation points work like <h1-6> but opposite

[14:03:01.0000] <annevk>
hmm, per broadband.gov the US has more ambitious goals than what I have now

[14:03:02.0000] <JonathanNeal>
! (is of most importance) ! and !! (! is of secondardy importance, !! is of most importance)

[14:04:00.0000] <TabAtkins>
Theoretically, or practically?  I find that !!!!!!!!! is generally less important than a *period*.

[14:04:01.0000] <JonathanNeal>
WOW THAT'S REALLY IMPORTANT!!!!!!!!!!!!!!!!!!!!!

[14:04:02.0000] <annevk>
100/50 mbits is pretty good

[14:04:03.0000] <MikeSmith>
AryehGregor: well stated

[14:04:04.0000] <MikeSmith>
(your message on the authoring-conformance thread)

[14:05:00.0000] <Dashiva>
Little-endian exclamation points

[14:05:01.0000] <MikeSmith>
gsnedders: who went to Glasgow?

[14:05:02.0000] <gsnedders>
MikeSmith: Sigbjorn Finne

[14:05:03.0000] <MikeSmith>
ah

[14:05:04.0000] <MikeSmith>
I don't think I know him

[14:06:00.0000] <MikeSmith>
is beer and other alcohol as expensive in Sweden as it is in Norway

[14:06:01.0000] <gsnedders>
Less expensive

[14:06:02.0000] <MikeSmith>
do they have a beer in Sweden that's as bad as Rignes?

[14:10:00.0000] <Dashiva>
I like your message, AryehGregor

[14:10:01.0000] <Dashiva>
It didn't even seem that long

[14:10:02.0000] <jgraham>
MikeSmith: I think he is relatively new

[14:11:00.0000] <jgraham>
gsnedders: Presumably if you were planning on following grand Opera traditions you would go to Linköping univeristy

[14:11:01.0000] <gsnedders>
jgraham: And fall in love with LysKOM?

[14:12:00.0000] <jgraham>
AryehGregor: You mean which post from Sam? The one in the "bug 7034" thread

[14:12:01.0000] <jgraham>
gsnedders: I doubt it ould brain damage you that much

[14:12:02.0000] <MikeSmith>
ah, LysKOM

[14:12:03.0000] <jgraham>
*could

[14:12:04.0000] <gsnedders>
jgraham: What are you trying to imply!?

[14:13:00.0000] <MikeSmith>
gsnedders, gsnedders - have you come to realize the brilliance of LysKOM yet?

[14:13:01.0000] <gsnedders>
MikeSmith: Don't tell me you're a kommie!

[14:13:02.0000] <MikeSmith>
it's the greatest

[14:13:03.0000] <MikeSmith>
what do we need IRC for, when we have LysKOM?

[14:13:04.0000] <gsnedders>
I thought you were fucked up before, but now...

[14:13:05.0000] <jgraham>
What could be btter than a messanging system that allows for threads to form arbitary directed graphs?

[14:14:00.0000] <jgraham>
It makes designing good UI so easy

[14:14:01.0000] <MikeSmith>
or even, what do we need e-mail for? or the Web?

[14:14:02.0000] <gsnedders>
jgraham: Email does that to

[14:14:03.0000] <gsnedders>
jgraham: You just have a In-Reply-To header with an arbitrary list of message ids

[14:14:04.0000] <jgraham>
gsnedders: Curiously all email clients just pretend like that is impossible

[14:15:00.0000] <Dashiva>
_You_ try designing a good graph-based UI

[14:15:01.0000] <jgraham>
Dashiva: Well I might design the protocol in such a way that it didn't exclude good UI in the first place

[14:16:00.0000] <gsnedders>
jgraham: So do semi-sane KOM clients

[14:16:01.0000] <Dashiva>
jgraham: How would you design a graph-based protocol so as to not require graph-based UI?

[14:17:00.0000] <jgraham>
Dashiva: I would not design a graph-based protocol

[14:17:01.0000] <jgraham>
Or rather I would limit it to trees

[14:19:00.0000] <jgraham>
(I certianly wouldn't tout the graphiness as a desirable feature)

[14:20:00.0000] <jgraham>
gsnedders: There are semi-sane KOM clients?

[14:20:01.0000] <gsnedders>
jgraham: Maybe

[14:23:00.0000] <Dashiva>
jgraham: In that protocol that is email + ignoring references that would create cycles, that is already the case

[14:23:01.0000] <Dashiva>
So it seems like a solved problem

[14:46:00.0000] <gsnedders>
jgraham: http://www.computerra.ru/terralab/softerra/516485/ --- interview with jl

[14:46:01.0000] <gsnedders>
(apparently)

[14:54:00.0000] <jgraham>
google translate ftw

[14:56:00.0000] <jgraham>
"On the singularities of the new interpreter Javascript "Computerra-line" said Jens Lindström (Jens Lindström), a leading developer Carakan."

[14:56:01.0000] <jgraham>
/me would like to make it clear that contrary to popular belief we are not actually harnessing the power of black holes to make Carakan faster

[14:59:00.0000] <Dashiva>
If you did, I imagine the technology race would get a lot more exciting

[15:11:00.0000] <jgraham>
Well I guess it would at least explain the ALIEN- bugs in the changelogs :)

[15:14:00.0000] <Dashiva>
"Safari reclaims the throne on sunspider using actual sun to power their javascript engine"

[15:28:00.0000] <Hixie>
/me returns

[15:29:00.0000] <jgraham>
Oh well, there goes the neighborhood

[15:29:01.0000] <JonathanNeal>
/me is swept away.

[15:29:02.0000] <JonathanNeal>
Here crawls back the neighborhood.

[15:29:03.0000] <JonathanNeal>
/me crawls back.

[15:30:00.0000] <jgraham>
By which I mean of course "welcome back. Have you been doing anything fun?"

[15:33:00.0000] <gsnedders>
jgraham: shhh. you're not meant to comment on impl detail like our black-hole magic!

[15:33:01.0000] <gsnedders>
Hixie: Anolis 2 is close ;)

[15:34:00.0000] <Hixie>
jgraham: spent thursday-saturday reffing this year's silicon valley regional of the FIRST robotics competition

[15:34:01.0000] <Hixie>
and sunday sleeping

[15:34:02.0000] <Hixie>
i fear my inbox

[15:34:03.0000] <gsnedders>
Inboxes are overrated

[15:35:00.0000] <Philip`>
Select all, mark as read

[15:35:01.0000] <Philip`>
Anything important should be in Bugzilla anyway

[15:36:00.0000] <boblet>
I’d fear Hixie’s inbox every day

[15:36:01.0000] <gsnedders>
s/'s inbox//

[15:37:00.0000] <othermaciej>
hello Hixie

[15:37:01.0000] <gsnedders>
/me remembers he was going to ask othermaciej something, but forgets what.

[15:37:02.0000] <gsnedders>
Why can't people just be around when I want them? :P

[15:38:00.0000] <othermaciej>
Hixie: there are some bugzilla bugs where I'd like to request expedited handling if possible, but I will give you a day or two to deal with the inbox

[15:38:01.0000] <othermaciej>
gsnedders: sorry!

[15:38:02.0000] <Hixie>
othermaciej: can you mark them P1?

[15:39:00.0000] <othermaciej>
Hixie: sure, I was going to ask you if that was a good way to do it

[15:39:01.0000] <Hixie>
yes please

[15:39:02.0000] <jgraham>
gsnedders: About webkit bugs?

[15:41:00.0000] <jgraham>
/me doubts it but it is the only conversation he had with gsnedders today where othermaciej would be an obvious source of information

[15:42:00.0000] <gsnedders>
jgraham: It wasn't today I wanted him.

[15:45:00.0000] <jgraham>
gsnedders: Oh well in that case it doesn't matter that you can't remember today :p

[15:47:00.0000] <gsnedders>
/me is somewhat disturbed by the fact he thinks he knows almost everything in the first year CS course at Glasgow

[15:50:00.0000] <annevk>
hmm, talk about bad timing: http://html5.org/tools/web-apps-tracker?from=4857&to=4858

[15:51:00.0000] <Hixie>
bad timing?

[15:52:00.0000] <jgraham>
gsnedders: So try going to some more challenging classes like extra Maths or whatever

[15:52:01.0000] <Hixie>
annevk: ?

[15:53:00.0000] <annevk>
I posted something on my blog just before

[15:53:01.0000] <annevk>
/me updates now before sleeping

[15:53:02.0000] <Hixie>
/me looks

[15:53:03.0000] <Hixie>
oh heh

[15:53:04.0000] <gsnedders>
jgraham: Can't with a C at AH

[15:53:05.0000] <Hixie>
you had me worried that i'd posted some edit after someone was complaining that the worst thing i could do would be to work on this or something :-)

[15:54:00.0000] <gsnedders>
jgraham: My AH grades prohibit me from doing anything like that, although I know the vast majority of the first year material. It sucks.

[15:54:01.0000] <annevk>
Hixie, ah, no worries :)

[15:54:02.0000] <jgraham>
gsnedders: You can't take the exam. I guess you can go to the lectures or something

[15:54:03.0000] <jgraham>
(or, er, ask nicely. It may be that the rules aren't as rigid as you suppose)

[15:55:00.0000] <gsnedders>
jgraham: I got the impression before that they really didn't like breaking them

[15:56:00.0000] <jgraham>
gsnedders: Even so

[15:56:01.0000] <jgraham>
Where is the actual first year content listed?

[15:57:00.0000] <gsnedders>
In random PDFs

[15:57:01.0000] <gsnedders>
/me won't be able to find the URLs again

[15:57:02.0000] <gsnedders>
Ooooh! Awesome! I'd get to learn Javascript!

[15:57:03.0000] <gsnedders>
At last!

[15:58:00.0000] <gsnedders>
I might finally get my head around that language!

[15:59:00.0000] <annevk>
sounds like it's bedtime gsnedders

[15:59:01.0000] <annevk>
though come to think of it you always act like this :p

[15:59:02.0000] <jgraham>
gsnedders: Seriously, urls

[15:59:03.0000] <jgraham>
Also: why "computing science"?

[15:59:04.0000] <annevk>
/me goes to bed himself

[15:59:05.0000] <annevk>
nn

[16:00:00.0000] <gsnedders>
/me can't sleep

[16:00:01.0000] <gsnedders>
jgraham: For no real reason, as far as I can tell

[16:01:00.0000] <gsnedders>
/me wishes he had any idea about what he wanted to do

[16:01:01.0000] <jgraham>
I don't think it even makes grammatical sense

[16:01:02.0000] <jgraham>
(it sounds like an attempt to calculate science)

[16:01:03.0000] <jgraham>
Anyway it was bedtime a while ago

[16:02:00.0000] <gsnedders>
Also, some of my information comes from what was said and given out at the open day

[16:02:01.0000] <gsnedders>
I can't find all of it online

[16:04:00.0000] <gsnedders>
/me should blatantly have applied for Scandavian Studies at Edinburgh

[16:07:00.0000] <gsnedders>
Or Gaelic.

[16:10:00.0000] <Hixie>
wow, help@ got a lot of traffic

[16:33:00.0000] <TabAtkins>
 Hixie: help@?

[16:37:00.0000] <gsnedders>
help⊙wo


2010-03-23
[17:04:00.0000] <Hixie>
50% done on mail

[17:13:00.0000] <Wolfman2000>
hmm...I'm doing something wrong. I thought XHTML worked fine with HTML5. I'm being given the "This XML file doesn't appear to have style" warning. What am I missing here again?

[17:13:01.0000] <gsnedders>
Wolfman2000: XHTML namespace?

[17:14:00.0000] <Wolfman2000>
gsnedders: ...that's supposed to go in html, right?

[17:14:01.0000] <gsnedders>
(i.e., xmlns="http://www.w3.org/1999/xhtml" on the root element)

[17:14:02.0000] <gsnedders>
Wolfman2000: yes

[17:14:03.0000] <Wolfman2000>
HTML5 won't complain if xmlns is left even in text mode, right?

[17:14:04.0000] <gsnedders>
Yes

[17:15:00.0000] <Wolfman2000>
While it's on my mind...this particular document will be using svg. Do I define the xmlns:svg (or the reverse) in <html> as well, or just <svg>?

[17:16:00.0000] <gsnedders>
just do <svg xmlns="[whatever the SVG namespace is]">

[17:16:01.0000] <Wolfman2000>
along with the xlink stuff...right

[00:28:00.0000] <zcorpan>
hmm, <nav> doesn't seem like a good fit for a <video>'s scripted controls - http://www.multimediapedia.com/2010/03/building-html5-video-player-tutorial.html slide 16

[00:29:00.0000] <othermaciej>
zcorpan: agree

[00:39:00.0000] <jwalden>
o_O

[01:02:00.0000] <annevk>
in a timeframe of less than two hours Shelley manages to email public-html seven times on the same subject...

[01:03:00.0000] <abarth>
annevk: must be an important issue

[01:04:00.0000] <annevk>
right

[01:33:00.0000] <hsivonen>
This looks useful: http://www.ankoder.com/

[01:38:00.0000] <virtuelv>
/me is looking for a thunderbird bug about UTF-8 being an offensive encoding in some locations

[01:38:01.0000] <virtuelv>
(japan, namely)

[01:45:00.0000] <virtuelv>
ah

[01:45:01.0000] <virtuelv>
https://bugzilla.mozilla.org/show_bug.cgi?id=448842#c14

[01:48:00.0000] <othermaciej>
annevk: was that the figure thread?

[01:48:01.0000] <othermaciej>
annevk: I am impressed that even after being shown her rationale for something was completely false, she didn't seem to even consider changing her mind

[01:50:00.0000] <hsivonen>
virtuelv: so sad that the email client installed base in Japan still doesn't fully support UTF-8

[01:50:01.0000] <hsivonen>
(according to comments on the bug)

[01:56:00.0000] <virtuelv>
it might be sad, but it at least shows that one shouldn't make assumptions about the world as a whole by looking at a portion of it

[01:57:00.0000] <zcorpan>
http://www.themaninblue.com/writing/perspective/2010/03/22/ - i wonder what the graph looks like for opera

[02:01:00.0000] <hsivonen>
IE9 platform preview and Minefield with D2D would be interesting, too.

[02:02:00.0000] <virtuelv>
zcorpan: he offers the tests for download

[02:02:01.0000] <zcorpan>
it seems flash is a bit faster than canvas in opera, but not a lot faster (tried 1000, 2000 and 3000 particles)

[02:03:00.0000] <zcorpan>
html and svg are slow when there are many particles

[02:04:00.0000] <virtuelv>
btw, http://www.codedread.com/blog/archives/2010/03/22/know-the-beast/#more-757

[02:05:00.0000] <zcorpan>
virtuelv: i think that shouldn't make any difference, at least in opera

[02:07:00.0000] <zcorpan>
hsivonen: a comment says svg in ie9 is 45fps

[02:08:00.0000] <zcorpan>
hsivonen: compared to 3-4fps in firefox 3.6

[02:09:00.0000] <annevk>
Opera is pretty slow :/

[02:09:01.0000] <annevk>
othermaciej, yeah, sad

[02:10:00.0000] <zcorpan>
annevk: canvas seems comparable to flash for me

[02:12:00.0000] <annevk>
yeah, just meant the SVG results compared with Chrome

[02:13:00.0000] <zcorpan>
ok

[02:13:01.0000] <zcorpan>
file a bug :)

[02:15:00.0000] <annevk>
pretty sure ed will do that

[02:24:00.0000] <annevk>
is http://twitter.com/tj/status/10850669048 for real?

[02:24:01.0000] <annevk>
that is somewhat hilarious

[02:25:00.0000] <othermaciej>
if by "Republicans" he means "Republican members of Congress" then yeah, they have their own government-administered health plan

[02:25:01.0000] <othermaciej>
which is considerably beyond what normal people would get

[02:26:00.0000] <othermaciej>
on the other hand, the HCR legislation specifically exempts members of Congress, so Democrats in Congress have all opted out of the particular health care plan they passed

[02:28:00.0000] <annevk>
weird

[02:29:00.0000] <othermaciej>
actually, on closer review that's not true, but certain congressional staffers (mainly the ones who are staff for the leadership) will be exempted

[02:29:01.0000] <othermaciej>
that's even more weird

[02:33:00.0000] <annevk>
the latest extensibility proposal is somewhat interesting

[02:33:01.0000] <smaug>
hmm, what to do with websocket protocol...

[02:33:02.0000] <smaug>
the latest one is just ugly

[02:33:03.0000] <annevk>
the change proposal doesn't work

[02:33:04.0000] <Hixie>
smaug: if anyone has any proposals that satisfy the security constraints as well, i'm all ears

[02:33:05.0000] <annevk>
but the overall idea is somewhat sound I suppose, though you still get namespace clutter :/

[02:34:00.0000] <smaug>
Hixie: yeah, I'm hoping that someone, especially someone with lots of experience in protocols, could come up with something better

[02:34:01.0000] <smaug>
"hoping"

[02:35:00.0000] <Hixie>
unfortunately most of the people with lots of experience with protocols have lots of experience with protocols that are to be implemented by experts :-(

[02:36:00.0000] <smaug>
that is true

[02:48:00.0000] <Hixie>
othermaciej: good point about the image stuff

[02:48:01.0000] <Hixie>
i'll make that change at some point

[02:49:00.0000] <Hixie>
i must go to bed now though

[02:49:01.0000] <othermaciej>
sure

[02:49:02.0000] <othermaciej>
I should to actually

[02:49:03.0000] <othermaciej>
g'night

[02:52:00.0000] <zcorpan>
http://labs.toolness.com/ff-herdict-preso/ would benefit from preloading the next few images

[03:00:00.0000] <zcorpan>
i wonder if ie9 behaves the same as ie8 for things in http://wiki.whatwg.org/wiki/Web_ECMAScript

[03:09:00.0000] <zcorpan>
/me updates web ecmascript wrt --> and carakan

[03:41:00.0000] <MikeSmith>
/me is having trouble trying to machine-translate that interview with jl

[03:41:01.0000] <MikeSmith>
Google Translate says, "Sorry, we are unable to access the page you requested"

[03:43:00.0000] <annevk>
which interview?

[03:43:01.0000] <annevk>
nm

[03:44:00.0000] <annevk>
wfm MikeSmith

[03:44:01.0000] <annevk>
http://translate.google.com/translate?u=http%3A%2F%2Fwww.computerra.ru%2Fterralab%2Fsofterra%2F516485%2F&sl=auto&tl=en

[03:45:00.0000] <MikeSmith>
annevk: thanks

[03:45:01.0000] <MikeSmith>
that works for me too

[03:45:02.0000] <MikeSmith>
dunno what I was doing wrong

[03:46:00.0000] <annevk>
not a whole lot of new info

[03:46:01.0000] <MikeSmith>
" The team consisted of three to five developers who worked on the project all the time, and about the same number of engineers."

[03:47:00.0000] <MikeSmith>
developers=programmers and engineers=QA engineers ?

[03:49:00.0000] <MikeSmith>
zcorpan: I saw announcement that the xml-stylesheet spec has been updated and looked at the diff

[03:49:01.0000] <jgraham>
MikeSmith: I guess

[03:49:02.0000] <MikeSmith>
ok

[03:49:03.0000] <MikeSmith>
zcorpan: it seems like there's nothing that's changed as far as the syntax, right?

[03:50:00.0000] <jgraham>
MikeSmith: (although the numbers seem a little off for QA, but it depends on how you count)

[03:50:01.0000] <MikeSmith>
I see

[03:51:00.0000] <MikeSmith>
btw, it would of course be great to have it working on ARM along with x86

[03:51:01.0000] <MikeSmith>
but I assume that's already in the works

[04:03:00.0000] <zcorpan>
MikeSmith: only minor fixes

[04:04:00.0000] <MikeSmith>
ok

[04:04:01.0000] <zcorpan>
MikeSmith: if you had implemented the previous draft to the letter, then you'd have to implement the removing of quotes in PseudoAttValue

[04:05:00.0000] <MikeSmith>
I'll have to go back and check

[04:06:00.0000] <zcorpan>
i'm pretty sure you implemented it as it says in the new draft

[04:08:00.0000] <zcorpan>
the new draft is http://www.w3.org/XML/2010/03/xml-stylesheet/ for anyone interested

[05:06:00.0000] <annevk>
sigh, the WebSocket API / protocol split causes us issues too

[05:06:01.0000] <annevk>
might be because Ian has not made it explicit what is defined where, but still

[05:13:00.0000] <annevk>
/me wonders how http://www.opera.com/press/releases/2010/03/23_3/ will go down

[05:18:00.0000] <Dashiva>
With a bang or a whimper

[05:34:00.0000] <Philip`>
Do people do these "submitted to App Store" press releases (instead of just "now available on App Store") mainly as a way to pressure Apple into accepting their app, under the assumption that Apple doesn't want to disappoint users by rejecting it for breaking some rules?

[05:35:00.0000] <Dashiva>
More like prevent apple for stealth-rejecting it without good reason

[05:35:01.0000] <Dashiva>
*from

[06:05:00.0000] <hsivonen>
in there any public info on how Opera Mini for iPhone was developed? Does it use the same code as the usual Mini but with GCJ?

[06:06:00.0000] <hsivonen>
/me wonders if Opera is already busy porting Mini to Windows Phone 7

[06:11:00.0000] <gsnedders>
hsivonen: I'm pretty certain there is no public info.

[06:34:00.0000] <Philip`>
People seem to put a lot of significance on the priority values of their bugs

[06:34:01.0000] <Philip`>
Aren't they kind of meaningless?

[06:40:00.0000] <hsivonen>
gsnedders: ok.

[06:57:00.0000] <annevk>
Philip`, in the end all bugs need to be resolved...

[07:00:00.0000] <Philip`>
Priorities should be unbounded integers so that you never have to upset people by decreasing the priority of their bugs, you just increase the priority of all other bugs

[07:04:00.0000] <annevk>
people should whine less and do more maybe

[07:04:01.0000] <Philip`>
That too

[07:06:00.0000] <Dashiva>
Maybe we could just adjust the scale

[07:06:01.0000] <Dashiva>
So P1, P2, P3 becomes P1, P1.5, P2

[07:07:00.0000] <asmodai>
Heard about the EDPS suggestion for webbrowser builders?

[07:08:00.0000] <Dashiva>
No, what is it?

[07:08:01.0000] <asmodai>
EDPS is the European privacy watchdog

[07:08:02.0000] <asmodai>
they want that browser vendors by default refuse cookies from advertising networks

[07:08:03.0000] <asmodai>
http://www.edps.europa.eu/EDPSWEB/webdav/site/mySite/shared/Documents/Consultation/Opinions/2010/10-03-19_Trust_Information_Society_EN.pdf

[07:09:00.0000] <asmodai>
See page 17 onward

[07:09:01.0000] <MikeSmith>
"Validation can then be used as a quick check to determine whether the code is the clean work of a seasoned HTML author, or quickly hacked-together tag soup."

[07:10:00.0000] <asmodai>
The steps that users must take to opt-out seem not only complicated but also excessive (first he must set his browser to accept cookies, then exercise the opt-out option).

[07:11:00.0000] <asmodai>
Therefore, while legally speaking, Article 5(3) of the ePrivacy Directive provides for effective legal protection, in practice, Internet users are deemed to consent to be monitored for the purposes of sending behavioural advertisement when in fact, in many, if not most cases, they are fully unaware that the monitoring takes place.

[07:11:01.0000] <asmodai>
As described above, web browsers commonly allow a level of control over certain kinds of cookies. Currently, the default settings of most web browsers are accepting all cookies. In other words, by default, the browsers are set to accept all cookies, independently of the purpose of the cookie.

[07:12:00.0000] <asmodai>
[...] Furthermore, there is no privacy wizard on the first install or update of browser applications.

[07:12:01.0000] <asmodai>
Looks like they're urging on legislation as well.

[07:13:00.0000] <asmodai>
Which would mean that browsers in the EU would have to refuse 3rd party cookies by default

[07:13:01.0000] <asmodai>
and implement a privacy wizard

[07:13:02.0000] <Dashiva>
I wonder if they know about SharedObjects

[07:14:00.0000] <Dashiva>
"Currently, the default settings of most web browsers are accepting all cookies."

[07:15:00.0000] <Dashiva>
Is this actually true? I thought most browsers only allowed first-party and redirect cookies by default?

[07:16:00.0000] <Philip`>
My Opera says "Accept cookies" (not "only from the site I visit"), and I don't think I've ever changed it away from the default

[07:16:01.0000] <asmodai>
My firefox 3.6.2 under privacy has no specific cookie setting option, not unless you select use custom settings for history

[07:16:02.0000] <asmodai>
and that shows accept third-party

[07:16:03.0000] <gsnedders>
The default was changed to "Accept cookies only from the site I visit" in 10.50 and reverted in 10.51

[07:17:00.0000] <Dashiva>
The only vanilla browser I have to check is Safari, and it says "only sites I visit"

[07:18:00.0000] <Dashiva>
"There is a need for more granularity and clear information on the types of cookies and the usefulness of some of them."

[07:18:01.0000] <asmodai>
Fair point.

[07:18:02.0000] <Dashiva>
The user eduction tools will save us

[07:20:00.0000] <asmodai>
the LART?

[07:21:00.0000] <Dashiva>
I don't think many people would answer "Why yes, I do want my browsing habits to be tracked and recorded for other people's profit"

[07:22:00.0000] <Dashiva>
And I seem to recall from bug reports that some sites break without 3rd party cookies, so that's an incentive to enable it anyhow

[07:23:00.0000] <MikeSmith>
wow, presence of xmlns as talisman of best-practices opt-in

[07:24:00.0000] <Dashiva>
<html xmlns="I support best practices">

[07:25:00.0000] <MikeSmith>
we time-travels 10 years into the future and gets killed by people there who end up getting stuck with the long-term consequences of that one 

[07:25:01.0000] <asmodai>
Aren't those sites maybe broken by design then? :)

[07:25:02.0000] <Philip`>
If Opera changed back in 10.51, was that because of bug reports of sites breaking, or was it the result of a successful attempt to extort protection money for ad networks by showing them Opera's power to turn cookies off in 10.50?

[07:25:03.0000] <MikeSmith>
.me looks around the room for Sam

[07:26:00.0000] <Philip`>
Overloading syntax for multiple unrelated purposes is always fun

[07:26:01.0000] <Philip`>
That's why everyone loves doctypes and quirks mode

[07:27:00.0000] <MikeSmith>
yeah, we have a great precedent there

[07:27:01.0000] <Dashiva>
Don't forget UA sniffing

[07:28:00.0000] <MikeSmith>
Dashiva: indeed, another case where things have just worked out swimmingly

[07:28:01.0000] <MikeSmith>
thank you, geniuses of the past

[07:28:02.0000] <MikeSmith>
we are truly standing on the shoulders of giants

[07:28:03.0000] <asmodai>
as long as you're not standing on my left shoulder

[07:29:00.0000] <asmodai>
/me managed to land his full 100 kg on it earlier today playing squash >_<

[07:29:01.0000] <Dashiva>
I think you're going to overload the sarcasm meters

[07:31:00.0000] <Philip`>
The problem with being a giant is that it's much more painful when you fall over, so your mistakes will have serious consequences long into the future

[07:31:01.0000] <MikeSmith>
slagiatt

[07:31:02.0000] <asmodai>
Philip`: yeah, I blame my 1,94m :(

[07:31:03.0000] <Philip`>
/me wonders how tall a human would have to be before tripping over would be fatal

[07:31:04.0000] <Dashiva>
To the mathmobile!

[07:32:00.0000] <Philip`>
I guess the trick is to make sure you only stand on the shoulders of extremely careful giants

[07:32:01.0000] <Dashiva>
Or extremely soft giants

[07:32:02.0000] <asmodai>
Although, according to another colleague of mine, the save was spectacular to watch, as I fell, slid over the floor and then at the end, after bruising the shoulder, burning knee and hand, managed to bump my head in the corner of 2 walls.

[07:32:03.0000] <MikeSmith>
Eloi, Eloi, slagiatt

[07:36:00.0000] <asmodai>
MikeSmith: Why does that ring a bell. It's said as elo-i, right?

[07:38:00.0000] <Dashiva>
Seemed Like A Good Idea At The Time

[07:38:01.0000] <Dashiva>
=

[07:38:02.0000] <MikeSmith>
asmodai: dunno, it's something I heard in a bar or strip club or something

[07:38:03.0000] <Dashiva>
?

[07:38:04.0000] <MikeSmith>
I have no idea what it means

[07:39:00.0000] <asmodai>
Eloi is something I remember, I think, from hebrew being used in the bible.

[07:40:00.0000] <asmodai>
Unless I am misremembering, which is always possible.

[07:41:00.0000] <Philip`>
/me doesn't know much about bibles but they were in The Time Machine

[07:43:00.0000] <annevk>
didn't hsivonen post conformance criteria a long time ago on rubys blog?

[07:50:00.0000] <hsivonen>
annevk: possibly. though I'm not sure Hixie has written the spec according to criteria I've expressed on Sam's blog at one time or another

[07:51:00.0000] <gsnedders>
asmodai: eloi means father, and is part of what Jesus cried out before death upon the cross

[07:51:01.0000] <gsnedders>
("Father, father, why have you forsaken me?" is the normal English translation, which in Hebrew begins with "Eloi, Eloi..."

[07:52:00.0000] <asmodai>
gsnedders: Ahh, see, remembered something from my christian upbringing.

[07:53:00.0000] <asmodai>
Not that helps much being buddhist. *grin*

[07:53:01.0000] <hsivonen>
annevk: did you mean http://intertwingly.net/blog/2007/05/08/Dont-Break-The-Web#c1178698369 ?

[07:53:02.0000] <annevk>
hixie cited that once

[07:53:03.0000] <annevk>
and also added some other things I believe

[07:54:00.0000] <annevk>
it's all in an email on public-html somewhere

[07:54:01.0000] <annevk>
seems to come up whenever someone decides it's worth whining about again

[07:54:02.0000] <hsivonen>
annevk: the w3c list search doesn't find http://intertwingly.net/blog/2007/05/08/Dont-Break-The-Web on public-html

[08:03:00.0000] <annevk>
maybe it was separate

[08:12:00.0000] <annevk>
grmbl I cannot find it

[08:12:01.0000] <annevk>
the archives are vast and complex to search through

[08:14:00.0000] <asmodai>
Does it require sacrifices?

[08:15:00.0000] <gsnedders>
Philip`: sites breaking

[08:32:00.0000] <Philip`>
gsnedders: Oh

[09:07:00.0000] <AryehGregor>
asmodai, "Eloi" isn't Hebrew for "father", or Aramaic.  Maybe it's Greek?

[09:08:00.0000] <AryehGregor>
Greek likes ending words in "oi".

[09:08:01.0000] <AryehGregor>
Although usually as a masculine plural, I thought.

[09:10:00.0000] <AryehGregor>
"Father, Father, why have you forsaken me?" would probably be something like "Avi, avi, lama azavtani?" in Biblical Hebrew.

[09:11:00.0000] <Philip`>
Maybe Jesus was screaming profanities, but the English translators thought it would be better to pretend he said "Father" instead

[09:11:01.0000] <MikeSmith>
heh

[09:12:00.0000] <AryehGregor>
There are no profanities in Biblical Hebrew, it's all very euphemistic.  There aren't even words for things like "urine", as far as I know.  Presumably such words existed, but nobody saw fit to write them down in holy texts.

[09:38:00.0000] <gsnedders>
AryehGregor: And the New Testement was mainly written in Greek, no?

[09:39:00.0000] <AryehGregor>
Well, I'm hardly an expert on the New Testament, but that's my impression, yes.

[09:45:00.0000] <gsnedders>
Wikipedia agrees

[09:47:00.0000] <AryehGregor>
Wait, so now Google is serving Chinese people out of Hong Kong?  That's controlled by the PRC since 1997, I thought?  Is it not inside the Great Firewall?

[09:47:01.0000] <gsnedders>
"E'li, E'li, la'ma sa-bach-tha'ni?" is what Wikipedia gives for what Matthew and Mark give as the final, Aramaic, words

[09:47:02.0000] <gsnedders>
AryehGregor: Dunno, yes, yes

[09:48:00.0000] <AryehGregor>
"Eli" means "my God", not "my Father".  Don't recognize "sabachtani", not sure if it's cognate to a Hebrew word I recognize.

[09:48:01.0000] <gsnedders>
"My God, My God, why have you forsaken me?"

[09:49:00.0000] <AryehGregor>
What article?

[09:49:01.0000] <gsnedders>
But apparently He is quoted in Matthew and Mark in Aramaic, which fits with my memory of the quotation being Aramaic and of the text as a whole being Greek

[09:50:00.0000] <gsnedders>
AryehGregor: http://en.wikipedia.org/wiki/Sayings_of_Jesus_on_the_cross

[09:50:01.0000] <AryehGregor>
Ah, good, it has the Hebrew/Aramaic lettering.

[09:51:00.0000] <AryehGregor>
/me notes that "ēlî ēlî lamâ azavtanî" is identical to what he predicted, except for "eli" "my God" instead of "avi" "my Father"

[09:51:01.0000] <AryehGregor>
שבק is an Aramaic root I don't recognize.

[09:51:02.0000] <gsnedders>
Gah, in Jesus's case, what difference does it make? :)

[09:52:00.0000] <AryehGregor>
What difference does it make in anyone's case?  Doesn't Moses say in Deuteronomy, "You are all sons to the Lord your God"?

[09:52:01.0000] <AryehGregor>
Christian and Jewish prayers call God "Father" all over the place.

[09:52:02.0000] <Philip`>
"omg! omg! why have you forsaken me?"

[09:52:03.0000] <gsnedders>
/me isn't sure that's Deuteronomy, though

[09:52:04.0000] <AryehGregor>
/me is pretty sure it's Deuteronomy.

[09:52:05.0000] <gsnedders>
AryehGregor: Also: there was sarcasm in that last comment.

[09:53:00.0000] <AryehGregor>
14:1ץ

[09:53:01.0000] <AryehGregor>
Eרץ

[09:53:02.0000] <AryehGregor>
Er.

[09:53:03.0000] <AryehGregor>
14:1.

[09:53:04.0000] <AryehGregor>
/me doesn't see the sarcasm

[09:54:00.0000] <gsnedders>
Oh, my finger slipped, it was meant to be :P at the end

[09:54:01.0000] <gsnedders>
(But yeah, I was meaning it jokingly from the root of Father being used to refer to Gof all the time)

[09:54:02.0000] <gsnedders>
*God

[10:32:00.0000] <asmodai>
AryehGregor: Could very well be

[10:49:00.0000] <Philip`>
/me discovers a program that uses SGML for documentation, which uses the syntax like <em/this/

[10:49:01.0000] <Philip`>
I didn't know anybody seriously used that feature

[10:51:00.0000] <AryehGregor>
asmodai, it seems like it's a Greek bastardization of the Aramaic for "my God", "eli".

[10:59:00.0000] <hober>
http://www.w3.org/2010/Talks/doclicense-20100323/#(7) license proliferation fail

[11:08:00.0000] <krijnh>
boblet: Np! :)

[11:11:00.0000] <Philip`>
"Survey results on question "should W3C publish HTML 5 under a license that allows forking?". Of those who responded, approximately 79% answered "no" and 14% answered "yes" (and the rest abstained)."

[11:11:01.0000] <AryehGregor>
Does it matter?  It's all published by the WHATWG too.

[11:12:00.0000] <Philip`>
I wonder why they want to attempt to prohibit forking, when the W3C version is already a fork itself

[11:13:00.0000] <Dashiva>
It's only bad when people you don't like do it

[11:13:01.0000] <AryehGregor>
It's only bad when people other than you do it.

[11:14:00.0000] <Dashiva>
Fork is a loaded word, after all

[11:14:01.0000] <Dashiva>
It implies there still being an original

[11:15:00.0000] <Philip`>
Does it?

[11:15:01.0000] <AryehGregor>
/me just saw this smilie: ಠ_ಠ

[11:15:02.0000] <AryehGregor>
/me <3 Unicode

[11:15:03.0000] <AryehGregor>
/me ♥ Unicode, too

[11:15:04.0000] <Dashiva>
Otherwise it's just "resuming the work" or "picking up the project"

[11:15:05.0000] <Philip`>
When there's a fork in a road, I'd expect that to be two equally divergent paths

[11:16:00.0000] <AryehGregor>
"Fork" implies there's still an original, at the moment of forking.

[11:16:01.0000] <AryehGregor>
One or the other may subsequently die, of course.

[11:16:02.0000] <Philip`>
and in a utensil-style fork I wouldn't expect there to be a primary tine

[11:16:03.0000] <Dashiva>
Stick to spoons, Philip`

[11:16:04.0000] <Philip`>
so it's odd that in software it has a different connotation

[11:17:00.0000] <Dashiva>
It's not that odd. They simply focused only on the splitting aspect

[11:17:01.0000] <Dashiva>
Hum

[11:18:00.0000] <AryehGregor>
I could look up "fork" in the OED to see whether the name for the utensil or the meaning of bifurcation came first.

[11:18:01.0000] <Dashiva>
bifurcation doesn't imply a original either, does it?

[11:19:00.0000] <AryehGregor>
It implies one thing becoming two.

[11:19:01.0000] <AryehGregor>
Not with either one necessarily different from the other.

[11:19:02.0000] <Philip`>
"Conclusion: For HTML 5, create a new license that addresses all of the use cases cited by the HTML Working Group except forking." - maybe the best way to address the uses cases would be to not address them, to avoid wasting effort given the WHATWG copy already addresses the use cases in a far simpler and less proliferatively way

[11:20:00.0000] <Dashiva>
Clearly we need a new word

[11:21:00.0000] <Dashiva>
Cloning doesn't work, because a fork isn't identical except at the very beginning

[11:22:00.0000] <Philip`>
AryehGregor: It suggests the "pronged instrument" version was used two hundred years before any other use

[11:22:01.0000] <AryehGregor>
I guessed that from the ordering of the definitions in the AHD.

[11:22:02.0000] <AryehGregor>
I think that usually orders chronologically.

[11:22:03.0000] <Philip`>
"c1000 ÆLFRIC  Hom. (Th.) I. 430 {Edh}a  cwelleras..wi{edh}-ufan  mid heora forcum hine {edh}ydon."

[11:23:00.0000] <Philip`>
Dashiva: Clones aren't identical except at the beginning, either

[11:23:01.0000] <Philip`>
though they may follow somewhat similar paths of development

[11:24:00.0000] <MikeSmith>
spawn

[11:24:01.0000] <hsivonen>
it'll be extremely interesting to see how they are going to disallow forking while permitting the inclusion in open-source code so that the code stays as open source

[11:24:02.0000] <Philip`>
but e.g. Republic clone troopers didn't all end up as bounty hunters

[11:24:03.0000] <MikeSmith>
"a soft, jelly-like substance containing the eggs of fish, or of animals such as frogs"

[11:24:04.0000] <AryehGregor>
hsivonen, why does open-source code need to include a copy of HTML5?

[11:25:00.0000] <hsivonen>
AryehGregor: see the list of use cases

[11:25:01.0000] <AryehGregor>
/me has given up on the W3C anyway, honestly

[11:27:00.0000] <Dashiva>
Philip`: That's more like using a different compiler, the source code is still identical ;)

[11:27:01.0000] <Philip`>
Was http://www.w3.org/2010/Talks/0323-html-plh/ meant to have more than just a title page?

[11:27:02.0000] <Philip`>
Oh

[11:27:03.0000] <Philip`>
I need to press arrow keys, apparently

[11:27:04.0000] <Philip`>
/me was attempting to just disable stylesheets in order to get a readable version of it

[11:33:00.0000] <JonathanNeal>
Say I have a collection of 10 blog previews, each with a title and part of the first paragraph of the blog; would all of these pieces belong to a section or an article?

[11:33:01.0000] <hober>
each piece in <article>

[11:35:00.0000] <JonathanNeal>
Okay, so now say I have a rotating banner of various slogans pointing to articles, similar to the previews, but now in <img /> form, would you still recommend several <article>s?  I feel like the entire group is either an article or a section.

[11:36:00.0000] <JonathanNeal>
I may be changing our rotating banner @ http://www.liferay.com/ in response to http://html5gallery.com/2010/03/liferay/ since I'm on the fence about whether or not I agree with their assertion.

[11:37:00.0000] <hober>
see http://edward.oconnor.cx/2009/09/using-the-html5-sectioning-elements#archive-pages for more

[11:38:00.0000] <hober>
I would still recommend several <article>s, yeah

[11:38:01.0000] <hober>
<div class="fancy-banner-thingy"><article>...</article><article>...</article>...</div>

[11:39:00.0000] <hober>
(<article>s with <a rel="bookmark" ...><img ...></a> inside)

[11:39:01.0000] <JonathanNeal>
Well, in that case, wouldn't they belong to a banner section?

[11:40:00.0000] <hober>
Only use <section> if it a) has its own heading and b) should affect the document outline

[11:40:01.0000] <hober>
I don't see a heading that refers to the collection of those blurbs

[11:41:00.0000] <hober>
<section> doesn't just mean "<div> 2: <div> harder"

[11:41:01.0000] <JonathanNeal>
Well, in this case, since they are part of the larger page and not on a page of their own, they would have a heading.

[11:42:00.0000] <hober>
yes, each <article> has an obvious heading. I'm talking about the parent element.

[11:42:01.0000] <hober>
the parent element doesn't have its own heading, so far as I can see

[11:43:00.0000] <JonathanNeal>
Yes, I'm refering to the parent element which wraps all of the articles.  At Liferay.com, we do have a heading to label the banners.

[11:43:01.0000] <hober>
I don't see one

[11:43:02.0000] <JonathanNeal>
View the page without styling.

[11:43:03.0000] <JonathanNeal>
Our presentational style is to remove the heading.

[11:44:00.0000] <hober>
OK

[11:44:01.0000] <JonathanNeal>
As a group of previews of articles or an archive of articles, I think a heading is fitting, especially when there are other areas on the site which are not.

[11:44:02.0000] <hober>
in that case, yeah, I'd use <section><h1>banners</h1><article>...</article>...</section>

[11:44:03.0000] <JonathanNeal>
on the site = on the page

[11:45:00.0000] <asmodai>
mmm, that sputnik site is interesting

[11:46:00.0000] <JonathanNeal>
hober, I'm compelled to feel that way too, except that http://www.whatwg.org/specs/web-apps/current-work/multipage/sections.html#the-article-element does note <article> content is "intended to be independently distributable or reusable"

[11:47:00.0000] <JonathanNeal>
Whereas an archive may be such, previews of blog posts or a rotating banner may not.

[11:47:01.0000] <JonathanNeal>
They are pieces of independantly distributable or reusable content.

[11:47:02.0000] <asmodai>
In case people didn't know, http://sputnik.googlelabs.com/

[11:52:00.0000] <TabAtkins>
JonathanNeal: If it's just a snippet, I would *not* do <article>.  I don't want to see a snippet by itself, fullscreened or on its own page, so it's not appropriate for <article>.

[11:53:00.0000] <TabAtkins>
That's my rule-of-thumb.

[11:53:01.0000] <JonathanNeal>
TabAtkins, now what about a collection of snippits?

[12:20:00.0000] <gsnedders>
Damn food is expensive in Oslo :\

[12:30:00.0000] <JonathanNeal>
Any notes before I move this blog over to Liferay?  http://sandbox.thewikies.com/liferay-featured-at-html5-gallery.html

[12:46:00.0000] <JonathanNeal>
Word http://www.liferay.com/web/jonathan.neal/blog/-/blogs/liferay-featured-html5-gallery

[12:53:00.0000] <jgraham>
gsnedders: How did it take you so long to notice?

[12:54:00.0000] <gsnedders>
jgraham: I haven't bought food before on this trip, and despite having been here before, it still amazes me.

[12:57:00.0000] <gsnedders>
What's browser support for datalist like?

[12:57:01.0000] <gsnedders>
Only Presto?

[12:57:02.0000] <miketaylr>
think so

[12:59:00.0000] <gsnedders>
And there's an incomplete compile time option for WebKit, as far as I can tell

[14:04:00.0000] <paul_irish>
are localStorage caches retained during browser version upgrades?

[14:04:01.0000] <neotyk>
Hi *, I can't find a word in RFC2616 about multiple Content-Type headers, and what client should do with those, is there guideline for it?

[14:23:00.0000] <JonathanNeal>
Is there a list of all elements like this http://www.w3.org/TR/html401/index/elements.html but the html5 version?

[14:24:00.0000] <hober>
JonathanNeal: http://simon.html5.org/html5-elements

[14:25:00.0000] <JonathanNeal>
Thank you hober.

[14:25:01.0000] <Philip`>
JonathanNeal: http://www.whatwg.org/specs/web-apps/current-work/multipage/section-index.html#elements-1

[14:26:00.0000] <JonathanNeal>
thank you Philip`

[14:35:00.0000] <jgraham>
paul_irish: That;s up to the UA but I would expect them to be typically

[14:36:00.0000] <paul_irish>
jgraham: cool. thx

[14:41:00.0000] <Dashiva>
neotyk: HTTP doesn't say much about abnormal behavior

[14:41:01.0000] <Dashiva>
neotyk: It's specified for HTML5, though

[14:42:00.0000] <JonathanNeal>
<command> goes in <head> ?

[14:44:00.0000] <Dashiva>
They can appear as phrasing content too

[14:45:00.0000] <Dashiva>
E.g. as children of a <menu>

[14:57:00.0000] <neotyk>
Dashiva: but I'm asking in context of HTTP 1.1 client

[15:04:00.0000] <JonathanNeal>
How would they work in the head?

[15:11:00.0000] <Dashiva>
JonathanNeal: Different types of commands, I suppose?

[15:14:00.0000] <Dashiva>
JonathanNeal: I think the main type are accesskey-activated ones

[15:15:00.0000] <JonathanNeal>
I see, for web applications

[15:16:00.0000] <daedb>
ugh, the public-html figure thread is giving me a headache...

[15:16:01.0000] <zcorpan>
/me notes that <command> was allowed in head before accesskey was allowed

[15:17:00.0000] <Dashiva>
Maybe for use via script?

[15:17:01.0000] <zcorpan>
daedb: weird, it's not giving me a headache. hmm maybe it's because i'm not reading it

[15:18:00.0000] <daedb>
zcorpan: I'm considering reading the DE thread too, but I fear for my sanity if I read both tonight

[15:19:00.0000] <Dashiva>
It's somewhat amusing that thread subjects are getting shorter, it's a rare event :)

[15:23:00.0000] <daedb>
hmmm... I'm gonna go watch vampires on tv instead, less scary than public-html

[15:32:00.0000] <JonathanNeal>
area's parent is listed as phasing 

[15:32:01.0000] <JonathanNeal>
but isn't it map?

[15:33:00.0000] <JonathanNeal>
I do not believe it will even validate without being the child of a map

[15:33:01.0000] <zcorpan>
it needs to be a descendant of map, but not necessarily child

[15:42:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/

[15:45:00.0000] <JonathanNeal>
http://html5.validator.nu/?doc=http%3A%2F%2Fsandbox.thewikies.com%2Fhtml5-pangram%2F

[15:46:00.0000] <JonathanNeal>
<command> swallows.

[15:59:00.0000] <JonathanNeal>
Keygen disappears entirely.

[16:27:00.0000] <JonathanNeal>
TabAtkins, you never told me what would you wrap a collection of snippits with "If it's just a snippet, I would *not* do <article>.  I don't want to see a snippet by itself, fullscreened or on its own page, so it's not appropriate for <article>."

[16:28:00.0000] <hober>
My mental model is "Use <article> wherever I use class='hentry'"

[16:30:00.0000] <annevk>
http://twitter.com/Duckstadkrant/status/10945262710 want

[16:34:00.0000] <TabAtkins>
JonathanNeal: I did too!  ^_^  It was a while later, but I addressed it to you.

[16:36:00.0000] <JonathanNeal>
Did not :( I just scrolled up

[16:37:00.0000] <TabAtkins>
LIES.

[16:37:01.0000] <TabAtkins>
Unless I sent it while Pidgin was offline or something, and it decided not to warn me.

[16:39:00.0000] <TabAtkins>
I'll look for it.  One sec, moving desks.

[16:46:00.0000] <JonathanNeal>
/me imagines TabAtkins is doing things akin to the Adventures of Lolo 


2010-03-24
[17:58:00.0000] <boblet>
hey all, I’m under the impression that the hr element shouldn’t be used to indicate a break between sections, but the description and paragraph content model explanation don’t really say this explicitly

[17:59:00.0000] <boblet>
am I mistaken?

[18:01:00.0000] <boblet>
will check logs for reply… TIA

[18:51:00.0000] <werdnativ>
Is HTML5 practically set in stone now, or is there still some possibility to consider new elements?

[18:55:00.0000] <erlehmann>
werdnativ, there is a process for that. http://wiki.whatwg.org/wiki/Feature_Proposals#Guidelines

[18:56:00.0000] <werdnativ>
cool, thanks. so it's still open for consideration?

[19:20:00.0000] <wolf2k_ubuntu>
Evening. So far my XHTML5 SVG web app is coming along. I've run into an issue, however. Due to the nature of the app, I have to be able to insert SVG elements dynamically. Without some sort of z-index, all new elements are overlayed on top: that's not what I'm after. Based on my own testing, I don't think SVG elements respect HTML's z-index CSS attribute. Is there an alternative?

[19:24:00.0000] <wolf2k_ubuntu>
...nevermind. I got the answer. Painters model.

[19:52:00.0000] <othermaciej>
wolf2k_ubuntu: I am surprised that SVG elements don't respect z-index - I believe they should

[19:53:00.0000] <wolf2k_ubuntu>
othermaciej: Try talking to them for version 1.2, whenever that comes

[19:53:01.0000] <wolf2k_ubuntu>
Either way, you've got a new FAQ now. Even if it's not so F

[19:54:00.0000] <othermaciej>
by "should" I mean I expect that to work now - do you have a test case to show that it doesn't work?

[19:54:01.0000] <wolf2k_ubuntu>
othermaciej: Unless you plan on visiting my house and seeing my development area this instant, don't think I can easily show it

[19:54:02.0000] <wolf2k_ubuntu>
It's a combination of XHTML, SVG, and JS

[19:56:00.0000] <wolf2k_ubuntu>
This app will be adding and removing elements dynamically. Now we've hit the point where I basically have to keep track of the nodes.

[19:56:01.0000] <roc>
according to the spec, z-index should not work on SVG elements (except possibly a positioned <svg> element with non-SVG parent)

[19:56:02.0000] <roc>
everyone agrees we need to change SVG to support some form of z-index

[19:57:00.0000] <wolf2k_ubuntu>
I can work around it. jQuery will be a big help.

[19:57:01.0000] <roc>
jwatt and I have argued pretty hard that we should support CSS z-index directly, with stacking contexts etc

[19:57:02.0000] <wolf2k_ubuntu>
But if you need someone else to sign your petition, you've got me

[19:58:00.0000] <wolf2k_ubuntu>
I had a ball of a time getting a mouseover/mousemove shadow rectangle to display over the SVG due to lack of z-index

[19:58:01.0000] <roc>
other people have argued that CSS z-index stacking contexts are not necessary for SVG

[19:58:02.0000] <wolf2k_ubuntu>
I personally think it should be available.

[20:01:00.0000] <othermaciej>
it looks like WebKit ignores z-index on SVG elements

[20:01:01.0000] <othermaciej>
seems bogus to me

[20:01:02.0000] <wolf2k_ubuntu>
othermaciej: Firefox does it too

[20:01:03.0000] <othermaciej>
roc: you could just say that nothing besides the SVG root element creates a stacking context

[20:02:00.0000] <othermaciej>
roc: that's assuming CSS position does not apply to SVG elements

[20:02:01.0000] <roc>
Then SVG elements with z-index not 'auto' would be positioned but not create a stacking context for their descendants

[20:02:02.0000] <roc>
which would be weird

[20:02:03.0000] <roc>
I mean positioned in z-order

[20:03:00.0000] <roc>
Webkit and Firefox are just following the current spec

[20:03:01.0000] <othermaciej>
is there any way to escape your ancestor's stacking context in CSS?

[20:03:02.0000] <roc>
one reason to want stacking contexts for SVG is that you want 'opacity' and 'filter' and other effects to induce a stacking context

[20:03:03.0000] <roc>
no

[20:04:00.0000] <roc>
otherwise you get weirdness like a blur filter applied to an element subtree, but the element's children are not all part of the same rendering group because there's another non-blurred element in between

[20:05:00.0000] <roc>
and then it becomes really unclear what the rendered results should be, or how to produce them

[20:06:00.0000] <othermaciej>
I see

[20:06:01.0000] <othermaciej>
I think most SVG elements either have their own rendering, or contain other elements that may have rendering, but not both

[20:08:00.0000] <roc>
that's true, but doesn't help us here, I think

[20:45:00.0000] <othermaciej>
roc: the reason I mention it is that this may seem like a reason to not support stacking contexts, although you could argue the opposite on the same basis I think

[20:46:00.0000] <roc>
it may *seem* like a reason but I don't think it actually is :-)

[21:10:00.0000] <JonathanNeal>
I made an HTML5 Pangram http://sandbox.thewikies.com/html5-pangram/

[21:23:00.0000] <wolf2k_ubuntu>
JonathanNeal: I'm guessing pangrams are supposed to show off everything about a protocol?

[21:23:01.0000] <JonathanNeal>
wolf2k_ubuntu, yes

[21:24:00.0000] <wolf2k_ubuntu>
...maybe it was just me, but yours felt a little disorganized

[21:24:01.0000] <JonathanNeal>
With as few extra properties on elements as possible to validate.

[21:26:00.0000] <wolf2k_ubuntu>
...you know, it feel funny. The closer I get done with this HTML5 SVG app, the more I'm thinking that it's a good thing I didn't try to use GWT or whatever Google calls their Java -> Javascript thing.

[00:22:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/ I noticed that http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html doesn't have any rendering for video / audio / canvas

[00:27:00.0000] <hsivonen>
does anyone have data on what maximum values different browser engines place on attribute value length?

[00:29:00.0000] <othermaciej>
it looks like WebKit will accept arbitrarily long attribute names (until you run out of RAM or characters in your stream)

[00:29:01.0000] <othermaciej>
checking values now

[00:30:00.0000] <othermaciej>
wait, I'm wrong

[00:30:01.0000] <othermaciej>
1024 is the limit for attribute names

[00:30:02.0000] <othermaciej>
now checking vlues

[00:31:00.0000] <othermaciej>
attribute values can be as long as you can fit in the source file

[00:31:01.0000] <othermaciej>
at least as far as parsing goes

[00:31:02.0000] <hsivonen>
ouch

[00:32:00.0000] <othermaciej>
if added via the DOM, either name or value can be as long as you can fit in a String

[00:32:01.0000] <hsivonen>
so you can OOM with <a a="aaaaaa.....

[00:32:02.0000] <othermaciej>
it won't crash, but it can use up a lot of memory before an allocation fails

[00:32:03.0000] <othermaciej>
that's assuming your source was long enough for that in the first place

[00:33:00.0000] <hsivonen>
well, the HTML5 specs is probably larger than any page Schwab's banking UI

[00:33:01.0000] <othermaciej>
and did not itself get too big to fit in RAM as UTF-16

[00:33:02.0000] <hsivonen>
so if I made the DoS limit the size of the HTML5 spec, banking sites probably wouldn't break...

[00:34:00.0000] <hsivonen>
Gecko is getting an infallible malloc, so it might be a problem to have the parser thread request so much memory that infallible malloc has to scramble for more memory on the main thread

[00:35:00.0000] <hsivonen>
not sure how exactly these things are supposed to work

[00:35:01.0000] <othermaciej>
in fallible?

[00:35:02.0000] <hsivonen>
never returns null

[00:35:03.0000] <othermaciej>
what happens if you exhaust the address space?

[00:35:04.0000] <Hixie>
does it buy more ram for you?

[00:35:05.0000] <hsivonen>
I believe it kills the app

[00:36:00.0000] <othermaciej>
we have a version of malloc that will crash if it runs out of memory but we have a few places that are safe against allocation failure

[00:36:01.0000] <othermaciej>
and thus do not use that variant

[00:36:02.0000] <hsivonen>
though IIRC, some design doc says that it's going to ask things to stop when there's little RAM left

[00:36:03.0000] <othermaciej>
(theory being that clean crash beats a buffer overrun)

[00:37:00.0000] <hsivonen>
I wonder if IE has a hard limit on attribute value length.

[00:37:01.0000] <hsivonen>
maybe I should just bite the bullet and test

[00:37:02.0000] <othermaciej>
we did find that for attribute *names*, 1024 is a sufficient practical limit

[00:37:03.0000] <othermaciej>
not sure it would fly for values

[00:38:00.0000] <Philip`>
Does it say in the IE interoperability reports that were released recently?

[00:38:01.0000] <Philip`>
/me vaguely remembers them mentioning limits, compared to the limits HTML4 mentions, or something

[00:38:02.0000] <othermaciej>
tag names also have a 1024 limit

[00:38:03.0000] <othermaciej>
(in the HTML parser only)

[00:39:00.0000] <hsivonen>
my plan is to change the tokenizer to use the same buffer for both names and values

[00:39:01.0000] <Philip`>
I've seen pages with tens of kilobytes of text in an <img alt> attribute

[00:39:02.0000] <hsivonen>
because you can't be tokenizing an attribute name and value at the same time

[00:39:03.0000] <Hixie>
meta content (with name=keywords) also often has huge values

[00:39:04.0000] <Hixie>
sometimes it is missing quotes and you end up with bazillions of attributes

[00:39:05.0000] <hsivonen>
the only case where the tokenizer needs two accumulation buffers is when a character reference is being tokenized inside an attribute value

[00:39:06.0000] <othermaciej>
we use a variable-size buffer with a small inline capacity for attribute values

[00:40:00.0000] <othermaciej>
i.e. in the common case it doesn't malloc, just uses a fixed-size buffer, but will malloc on overflow

[00:40:01.0000] <hsivonen>
but in that case, the other buffer can't grow longer that the longest named character name

[00:40:02.0000] <hsivonen>
Hixie: truncating keywords isn't fatal

[00:41:00.0000] <hsivonen>
Hixie: truncating huge attributes on Schwab's netbank apparently is

[00:41:01.0000] <Hixie>
<input type=hidden value="">?

[00:42:00.0000] <hsivonen>
Hixie: not sure. I don't have login credentials.

[00:43:00.0000] <othermaciej>
Hixie: you're the only one who writes Change Proposal summaries that I can use as-is on the status page

[00:44:00.0000] <Hixie>
i do my best :-)

[00:45:00.0000] <hsivonen>
I think http://www.w3.org/Bugs/Public/show_bug.cgi?id=7670#c43 is a mockery of RDFa in HTML being a deliverable of the HTML WG

[00:45:01.0000] <hsivonen>
why bother pretending that it's contents aren't delivered by the RDFa WG?

[00:46:00.0000] <hsivonen>
s/it's/its/

[00:47:00.0000] <hsivonen>
scary. I do the "it's" thing now.

[01:02:00.0000] <hsivonen>
click and hold Dock icon for Exposé has to be the worst UI feature of Snow Leopard

[01:02:01.0000] <hsivonen>
I have never wanted to active it, but I accidentally activate it on a daily basis

[01:07:00.0000] <othermaciej>
hsivonen: I don't think accidental click-and-hold is a common user behavior

[01:19:00.0000] <jgraham>
I have done that a few times too

[01:20:00.0000] <jgraham>
Although plural of anecdotes and all that

[01:20:01.0000] <hsivonen>
othermaciej: I use a Wacom tablet--not a mouse

[01:20:02.0000] <othermaciej>
does that make it more likely?

[01:20:03.0000] <hsivonen>
dunno

[01:21:00.0000] <Philip`>
/me sometimes half-clicks on something before he's quite decided whether he meant to click on it or not, and then holds the mouse down while he's thinking

[01:27:00.0000] <asmodai>
http://www.smashingmagazine.com/2010/03/15/showcase-of-web-design-in-china-from-imitation-to-innovation-and-user-centered-design/  <-- I knew before how click intense Chinese websites are, but this really confirms it.

[01:29:00.0000] <othermaciej>
what parts of the article show the click-heaviness?

[01:29:01.0000] <othermaciej>
/me is just skimming

[01:31:00.0000] <annevk>
bah, got disconnected

[01:40:00.0000] <annevk>
/me wonders in what situation you would use hidden= on <td>

[01:44:00.0000] <lazni>
empty cell?

[01:45:00.0000] <Philip`>
It'd be more useful to support <col hidden>

[01:46:00.0000] <annevk>
again, would that usage match the semantics of hidden?

[01:50:00.0000] <Philip`>
<table><col><col id=rc hidden><tr><th>Test case input<th>Result<tr><td>1+1<td id=r0></table> <script>r0.innerHTML = '11'; rc.hidden = false;</script>

[01:51:00.0000] <zcorpan>
hmm, regarding allow-top-navigation, don't ads usually open links as target=_blank anyway?

[01:51:01.0000] <Philip`>
(Don't want to make the results visible until they're all filled in)

[01:51:02.0000] <zcorpan>
Philip`: does visibility:collapse work on colgroup and col?

[01:52:00.0000] <Philip`>
zcorpan: Don't ask me

[01:52:01.0000] <Philip`>
If it doesn't, HTML shouldn't be restricted by current bugs in CSS anyway

[01:53:00.0000] <zcorpan>
it seems to work

[02:04:00.0000] <annevk>
any chance btw we can get a multipage version of complete.html?

[02:04:01.0000] <annevk>
would be especially useful for WebSocket

[02:04:02.0000] <Hixie>
just get a better browser :-P

[02:05:00.0000] <Hixie>
seriously though, complete.html works fine in chrome on mac

[02:05:01.0000] <zcorpan>
still annoying for page search

[02:05:02.0000] <annevk>
on Ubuntu too, but loading a smaller file is still nice

[02:06:00.0000] <JonathanNeal>
I tried to translate the x h1 example from http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#margins-and-padding into actual css

[02:07:00.0000] <zcorpan>
maybe we should ditch the "html5" version and go back to only web apps 1.0

[02:07:01.0000] <JonathanNeal>
in LESS it is 20 lines, in actual css it is 1383 lines.

[02:08:00.0000] <roc>
the comment about "better browser" is slightly annoying given http://weblogs.mozillazine.org/bz/archives/020267.html

[02:10:00.0000] <JonathanNeal>
http://pastie.org/884348 written in LESS css, http://pastie.org/884350 written in normal css

[02:12:00.0000] <zcorpan>
JonathanNeal: it seems that's not correct

[02:13:00.0000] <JonathanNeal>
zcorpan, like I'm wrong, or it's just suprising?

[02:13:01.0000] <zcorpan>
JonathanNeal: you want article article article article article h1, not article h1 article h1 article h1 article h1 article h1,

[02:14:00.0000] <JonathanNeal>
Sure, I'll update the examples, they'll still be about 70 fold different.

[02:15:00.0000] <zcorpan>
/me doesn't follow

[02:16:00.0000] <zcorpan>
JonathanNeal: the LESS version should probably be something more like article, aside, etc { h1 { ... } article, aside, etc { h1 { ... if i understand the LESS syntax correctly

[02:18:00.0000] <JonathanNeal>
zcorpan, here you go http://pastie.org/884363 and http://pastie.org/884364

[02:19:00.0000] <zcorpan>
JonathanNeal: yep, that looks better

[02:20:00.0000] <JonathanNeal>
So, I'll have to use some kind of class on all sectioning content.

[02:23:00.0000] <annevk>
roc, maybe we should all start cheating

[02:24:00.0000] <roc>
'cheating' is a stronger word than I would use

[02:24:01.0000] <roc>
browsers make different performance vs correctness tradeoffs :-)

[02:26:00.0000] <zcorpan>
i really don't understand why anyone would want to use <img role="presentation" alt="" src="img" > or <img role="presentation" src="img" > instead of <img alt="" src="img" >

[02:27:00.0000] <lazni>
http://3273372964/en/weblog?weblogid=208188044

[02:27:01.0000] <Philip`>
annevk: Feel free to update http://code.google.com/p/html5/source/browse/trunk/spec-splitter/spec-splitter.py to work on it :-)

[02:27:02.0000] <Philip`>
(Maybe the only difference is it should have a different list of sections to split?)

[02:28:00.0000] <roc>
I shouldn't say this, but "figuring out how to fix selectors without regressing performance" reminds me of OJ searching for the real killers

[02:28:01.0000] <Philip`>
and then either run it yourself, or ask Hixie to change his server stuff (and me to change mine a bit)

[02:28:02.0000] <annevk>
Philip`, I suppose, how do you deal with W3C HTML5 vs WHATWG HTML5?

[02:30:00.0000] <Philip`>
annevk: I don't

[02:30:01.0000] <Philip`>
except to the extent of adding a --w3c option that twiddles a few things

[02:30:02.0000] <Philip`>
MikeSmith set it up to run on the W3C copy, I believe

[02:31:00.0000] <annevk>
I wonder if I can just add the new section names without having the HTML5 copy left broken

[02:33:00.0000] <annevk>
oh, it already splits on all <h2> elements

[02:33:01.0000] <annevk>
that sounds perfect

[02:33:02.0000] <Philip`>
The split_exceptions list is just a set of (approximately) childs of <body> that it'll split on, when the <h2>s are too big

[02:33:03.0000] <Philip`>
and it'll ignore anything that's in that set and doesn't exist in the document

[02:34:00.0000] <annevk>
oh, maybe not

[02:34:01.0000] <annevk>
cool, so we could add sections there for eventsource, websocket

[02:36:00.0000] <Philip`>
That should work

[02:37:00.0000] <Philip`>
/me tried to choose splits so each page was 100-200KB

[02:37:01.0000] <Philip`>
Oh, the source code already says that

[03:05:00.0000] <Philip`>
Does anyone want a free 1-year domain registration at gandi.net?

[03:05:01.0000] <Philip`>
/me has a couple of promo codes that expire at the end of the month

[03:09:00.0000] <hsivonen>
Philip`: for new domains only, I assume?

[03:09:01.0000] <Philip`>
hsivonen: Yes

[03:09:02.0000] <Philip`>
(https://ten.gandi.net/faq)

[03:12:00.0000] <hsivonen>
Philip`: then, I guess I don't have a use for it

[03:13:00.0000] <annevk>
registering at a new host is not worth the hassle either for something that costs EUR 5-10

[03:15:00.0000] <Dashiva>
"Currently <p /> is forbidden - validator.nu doesn't even permit it in polyglot mode, it seems. Yet, it has been permitted for 10 years. Or more: It is even permitted in HTML4 documents."

[03:15:01.0000] <Dashiva>
What black magic is this?

[03:16:00.0000] <Philip`>
Isn't that just <p>&gt; ?

[03:17:00.0000] <Dashiva>
That's what I thought

[03:17:01.0000] <hsivonen>
Dashiva: where's the quete from. It's wrong.

[03:17:02.0000] <Dashiva>
http://lists.w3.org/Archives/Public/public-html/2010Mar/0601.html

[03:17:03.0000] <Dashiva>
Leif seems to claim <p /> is identical to <p></p>

[03:17:04.0000] <Philip`>
validator.w3.org even gives a warning ("NET-enabling start-tag requires SHORTTAG YES")

[03:18:00.0000] <hsivonen>
Dashiva: incorrect

[03:18:01.0000] <hsivonen>
I need to remove the "polyglot" mode

[03:18:02.0000] <hsivonen>
since it hasn't gone anywhere after the first features

[03:18:03.0000] <Philip`>
validator.w3.org "show outline" of <h1><p/>foo shows how it results in a literal ">"

[03:19:00.0000] <hsivonen>
I don't know who dropped the ball. Me regarding repo access? Sam regarding impl? tantek and zeldman regarding requirements?

[03:19:01.0000] <jgraham>
hsivonen: Did you move to hg yet?

[03:20:00.0000] <jgraham>
If so it seems like repo access shouldn't be a big issue

[03:21:00.0000] <hsivonen>
jgraham: the parser is in hg, but Sam doesn't have access, because the central repo is under Mozilla committer rules

[03:21:01.0000] <hsivonen>
the rest is still in svn and Sam has access

[03:21:02.0000] <hsivonen>
s/access/write access/

[03:33:00.0000] <othermaciej>
hsivonen: you could ask Sam if he has any intent to continue the work

[03:33:01.0000] <annevk>
http://www.w3.org/QA/2010/03/update_on_html_5_document_lice.html -- hmm, it almost sounds like they want to invent yet another license

[03:34:00.0000] <othermaciej>
annevk: but see Ian Jacobs' comment at the end

[03:36:00.0000] <annevk>
I wonder how that will work out

[03:36:01.0000] <annevk>
It's annoying to have see the confidential information as well...

[03:41:00.0000] <othermaciej>
you mean the AC poll? or is there other confidential informaiton?

[03:44:00.0000] <annevk>
yeah

[03:44:01.0000] <annevk>
there's some presentation as well

[03:44:02.0000] <othermaciej>
I was surprised at Opera's vote, but then again, I was also surprised at Apple's vote

[03:45:00.0000] <annevk>
tell me about it

[03:45:01.0000] <Philip`>
"Having the specification available in two places is suboptimal, and we'll see whether we can improve the situation." - so they want to restrict the licensing of the W3C copy of the spec, and also get rid of the freely licensed copy?

[03:45:02.0000] <annevk>
oh, the slides are public

[03:46:00.0000] <annevk>
Philip`, the W3C copy is already restricted

[03:47:00.0000] <annevk>
a colleague of mine argued the WHATWG license was not adequate btw, but I've no idea whether that is actually true

[03:47:01.0000] <jgraham>
I note that "a license that is familiar to the free software community" would include, for example "all rights reserved"

[03:47:02.0000] <annevk>
though it would have been nicer if it just said MIT

[03:47:03.0000] <annevk>
jgraham, :p

[03:49:00.0000] <Philip`>
s/to restrict/to continue restricting/

[03:50:00.0000] <jgraham>
annevk: Well it is kind of important; you can't have a license that is actually a free software license and prevent forking. But "familiar to" doesn't mean "compatible with"

[03:50:01.0000] <annevk>
jgraham, if it's important you should add a comment

[03:51:00.0000] <annevk>
/me added a comment asking about non-HTML5 specifications

[03:51:01.0000] <annevk>
/me wonders if he should publish CORS/CSSOM/etc. under some different venue with another license

[03:51:02.0000] <othermaciej>
or XHR for that matter

[03:51:03.0000] <annevk>
do I just give up my copyright if I commit to dev.w3.org?

[03:52:00.0000] <annevk>
/me wonders how that works

[03:53:00.0000] <Dashiva>
No

[03:54:00.0000] <Philip`>
/me has a vague recollection that you remain the copyright owner of anything you produce (except work-for-hire etc) unless you explicitly sign an agreement to transfer ownership

[03:55:00.0000] <annevk>
cool, I haven't done that

[03:55:01.0000] <annevk>
how does it work if HTML WG agreed upon text gets included in the WHATWG copy?

[03:56:00.0000] <annevk>
do change proposal people give away their copyright?

[03:57:00.0000] <Philip`>
I think it works on the basis that the HTML WG only agrees on text that is low quality and needs to be rewritten by Hixie anyway

[03:57:01.0000] <Dashiva>
Indeed

[03:57:02.0000] <Philip`>
and also on the basis that nobody cares that much about the precise correct application of copyright laws

[03:58:00.0000] <othermaciej>
I think W3C participation agreement might include something about copyright

[03:59:00.0000] <annevk>
yeah, was afraid of that

[03:59:01.0000] <Dashiva>
What if a non-wg member posts to public-html?

[04:00:00.0000] <othermaciej>
I don't know if it implies an exclusive copyright assignment or non-exclusive

[04:01:00.0000] <othermaciej>
http://www.w3.org/2009/12/Member-Agreement

[04:01:01.0000] <Dashiva>
The slides from yesterday said that contributors have the right to publish their own contributions, but not others', which seems to imply non-exclusive

[04:01:02.0000] <othermaciej>
(not sure if that is the latest version and not sure whether it also applies to standards documents

[04:01:03.0000] <workmad3>
it's probably something along the lines of contributors grant a non-exclusive, perpetual licence to w3c/whatwg

[04:01:04.0000] <Philip`>
http://www.w3.org/Consortium/Legal/2007/06-invited-expert.html - "The Invited Expert hereby grants to the W3C a perpetual, nonexclusive, royalty-free, world-wide right and license under any Invited Expert's copyrights on his or her contributions, to copy, publish and distribute the contribution under the W3C document licenses."

[04:01:05.0000] <othermaciej>
so you could publish under Opera's copyright

[04:02:00.0000] <othermaciej>
per the member agreement

[04:02:01.0000] <Philip`>
"The Invited Expert agrees to refrain from creating derivative works that include the Invited Expert's contributions when those derivative works are likely to cause confusion about the status of the W3C work or create risks of non-interoperability with a W3C Recommendation. «Branching» is one example of a non-permissible derivative work."

[04:02:02.0000] <othermaciej>
that doesn't seem to be in the Member agreement

[04:03:00.0000] <annevk>
I wonder if "c. The Member shall have access to software and documentation produced by the Consortium prior to release of such software and documentation to non-members, as set forth in Appendix 1." is still true with all the public drafts

[04:04:00.0000] <othermaciej>
I wonder if it is in the Process Document

[04:05:00.0000] <othermaciej>
the member agreement I linked is marked *** DRAFT *** but also linked as the latest

[04:05:01.0000] <Philip`>
Is something like http://philip.html5.org/tests/canvas/suite/tests/spec.html okay? It's a derivative work (of the WHATWG spec) that contains some of my contributions that were posted to public-html and were integrated into the W3C spec

[04:06:00.0000] <othermaciej>
you participate in public-html as an invited expert?

[04:06:01.0000] <Philip`>
and it's based on a 'branch' of the W3C specs (the HTML5 and 2d-context specs joined together with some extra text)

[04:06:02.0000] <othermaciej>
er in the HTML WG

[04:07:00.0000] <Philip`>
and it might confuse people into thinking it's related to the W3C spec

[04:07:01.0000] <Philip`>
othermaciej: Public invited expert, which I think involves the same agreement

[04:09:00.0000] <othermaciej>
I think it would take a lawyer to answer

[04:09:01.0000] <Philip`>
I think I won't bother asking one :-)

[04:09:02.0000] <othermaciej>
I don't think you made any contributions that would require a copyright grant

[04:09:03.0000] <othermaciej>
but I don't know if that affects the second clause

[04:10:00.0000] <othermaciej>
clause 2.4 of the Invited Expert agreement is interesting

[04:10:01.0000] <Philip`>
It just seems strange that it imposes more restrictions than the W3C document license itself

[04:11:00.0000] <othermaciej>
does the document license allow derivative works?

[04:12:00.0000] <Philip`>
The document license doesn't (can't) prevent works that are derivative of non-W3C works (e.g. the WHATWG spec, in my case)

[04:13:00.0000] <Philip`>
(because the license can only restrict the copying of documents which are published under that license)

[04:16:00.0000] <othermaciej>
ah, right

[04:16:01.0000] <othermaciej>
copyright can't prevent you from making a derivative work of something else, and most particularly of something on which you hold copyright

[04:19:00.0000] <Philip`>
The IE agreement sounds like it does try to prevent that, though

[04:19:01.0000] <othermaciej>
yes it does

[04:20:00.0000] <othermaciej>
I wonder why they make the assignment non-exclusive if they then try to restrict your use of your contributed material

[04:27:00.0000] <Philip`>
I'd guess because exclusive assignment is much more complex to do

[04:27:01.0000] <Philip`>
http://www.copyright.gov/title17/92chap2.html#204 - "A transfer of copyright ownership, other than by operation of law, is not valid unless an instrument of conveyance, or a note or memorandum of the transfer, is in writing and signed by the owner of the rights conveyed or such owner's duly authorized agent." etc

[04:28:00.0000] <Philip`>
("A “transfer of copyright ownership” is an assignment, mortgage, exclusive license, [...] but not including a nonexclusive license.")

[04:29:00.0000] <Philip`>
and I never had to write or sign anything to become a PIE

[04:30:00.0000] <othermaciej>
I see

[04:30:01.0000] <Philip`>
and I guess other sorts of contracts and agreements don't necessarily have to be in writing so it's okay to just do those over the internet and try to restrict people that way

[04:30:02.0000] <othermaciej>
FSF makes people sign things even for non-exclusive copyright assignment

[04:30:03.0000] <othermaciej>
(or maybe they do exclusive assignment and license your contribution back to you, I can't remember)

[04:31:00.0000] <hsivonen>
IIRC, FSF wants assignment and licenses stuff back

[04:31:01.0000] <hsivonen>
OpenOffice.org wants join copyright holdership

[04:31:02.0000] <hsivonen>
*joint

[04:32:00.0000] <hsivonen>
as I understand it, in the U.S. you need to have at least joint ownership to have standing to enforce a copyright

[04:36:00.0000] <othermaciej>
does Mozilla require any copyright assignment or joint ownership?

[04:36:01.0000] <othermaciej>
(neither Apple nor the WebKit project do, for WebKit)

[04:36:02.0000] <hsivonen>
IIRC, Mozilla Foundation requires neither.

[04:37:00.0000] <othermaciej>
is work done by MoCo employees copyright MoCo or copyright MoFo?

[04:37:01.0000] <hsivonen>
othermaciej: MoFo

[04:37:02.0000] <gsnedders>
othermaciej: Who should be CC'd on a JS parser bug in JSC?

[04:37:03.0000] <othermaciej>
gsnedders: oliver⊙ac, mjs⊙ac, ggaren⊙ac

[04:37:04.0000] <othermaciej>
(I'm not strictly necessary but I am curious)

[04:38:00.0000] <gsnedders>
/me thought olliej did more of the JIT stuff and not so much parser stuff

[04:39:00.0000] <othermaciej>
there are also some uni-szeged people who know the parser code but I can't recall names offhand

[04:39:01.0000] <othermaciej>
perhaps also some nokia people

[04:39:02.0000] <othermaciej>
olliej is fully qualified to hack any part of JSC

[04:41:00.0000] <gsnedders>
Shows my memory :)

[04:41:01.0000] <othermaciej>
now I wonder who has their svn username in the most distinct JSC source files

[04:41:02.0000] <othermaciej>
but I think I am too lazy to make a script to find out

[04:53:00.0000] <hsivonen>
aargh. looks like USB on my Linux box died and I don't have sshd up

[04:53:01.0000] <hsivonen>
one should always have sshd up

[05:09:00.0000] <hsivonen>
well, that wasn't nice. I did a hardware shutdown and reboot and grub wouldn't find my boot volume.

[05:09:01.0000] <hsivonen>
I was rather worried until I decided to reboot again

[05:09:02.0000] <hsivonen>
and then grub found the boot volume

[05:10:00.0000] <hsivonen>
things not going great today with computers

[05:11:00.0000] <hsivonen>
earlier today, my Mac was unusable for 3 hours, because Parallels decided to update Parallels Tools on a Windows VM and I was crazy enough to put other load on the machine at the same time (as opposed to just waiting for Parallels rendering the computer inoperable for half an hour or so)

[05:20:00.0000] <gsnedders>
Anyone with Minefield handy?

[05:20:01.0000] <jgraham>
Sure

[05:20:02.0000] <hsivonen>
/me has

[05:20:03.0000] <gsnedders>
Does <http://stuff.gsnedders.com/es/getterparsing.html> fail for the last four tests?

[05:21:00.0000] <hsivonen>
yes

[05:21:01.0000] <jgraham>
yes

[05:21:02.0000] <jgraham>
Also in Chrome

[05:21:03.0000] <gsnedders>
OK

[05:21:04.0000] <gsnedders>
I know about other browsers

[05:21:05.0000] <jgraham>
Yay for interoperability

[05:22:00.0000] <gsnedders>
/me wonders why he doesn't have up-to-date Minefield :\

[05:22:01.0000] <gsnedders>
(Like, the latest Fx nightly I have is 3.6 alpha2-pre)

[05:22:02.0000] <gsnedders>
(Which isn't even Minefield)

[05:22:03.0000] <annevk>
can you bikeshed just by yourself?

[05:22:04.0000] <jgraham>
You don't have the Mozilla PPA?

[05:23:00.0000] <gsnedders>
Uh, yeah. I could've just done it over remote desktop.

[05:23:01.0000] <gsnedders>
That's true

[05:23:02.0000] <jgraham>
annevk: Isn't that just called being indecisive?

[05:23:03.0000] <jgraham>
In which case: ask gsnedders :)

[05:23:04.0000] <gsnedders>
/me slaps jgraham 

[05:24:00.0000] <annevk>
hehe

[05:24:01.0000] <jgraham>
Sadly I'm out of reach, so I imagine that without a face to absorb the momentum, your hand just continued round and you actually slapped yourself on the other side of your head

[06:26:00.0000] <AryehGregor>
<roc> the comment about "better browser" is slightly annoying given http://weblogs.mozillazine.org/bz/archives/020267.html <-- Incorrect but fast can legitimately be called "better" than correct but slow.  Browsers are meant to create a good user experience, not conform to standards at all costs.  Realistically, IMO, refraining from fully implementing features until you can do so with no performance regression does make for a better browser f

[06:26:01.0000] <AryehGregor>
rom a user standpoint.  As Dave Hyatt points out on that post, it's just not fully implemented -- should they have not implemented it at all until they could get all corner cases right?

[06:29:00.0000] <annevk>
AryehGregor, I think the point is that it is is probably not possible to get it right without regressing performance

[06:29:01.0000] <annevk>
arguably we should nuke some selectors

[06:30:00.0000] <AryehGregor>
Interesting.

[06:30:01.0000] <AryehGregor>
Well, UAs often just refuse to implement features that would cost performance, right?

[06:31:00.0000] <AryehGregor>
That's not unique to WebKit.

[06:32:00.0000] <annevk>
yes, but that's not what happened here

[06:32:01.0000] <hsivonen>
if the selector features in question have been permanently vendor vetoed, the selector spec should change, IMO

[06:32:02.0000] <AryehGregor>
Looks like what happened here to me.

[06:32:03.0000] <AryehGregor>
"Boris, I'm just saying that I never looked at benchmarks or cared about benchmarks when adding this code. I held off on generalizing the code because I needed to think about a way to do it that would still be performant."

[06:33:00.0000] <AryehGregor>
= "won't implement it unless it can perform well"

[06:33:01.0000] <AryehGregor>
+ "probably not possible to get it right without regressing performance" => "won't implement", simple syllogism.

[06:34:00.0000] <hsivonen>
I suggest Hixie stops using the feature that has been vendor vetoed, then

[06:34:01.0000] <annevk>
Selectors is also in CR...

[06:35:00.0000] <annevk>
there'd be quite some outcry if we killed nth-child and friends

[06:36:00.0000] <AryehGregor>
Is Hixie actually using a CSS feature that kills performance in every browser that implements it?  That seems pointless.  I thought Firefox's painful handling of the HTML5 spec was mainly due to some JS thing and was fixed in 3.7.

[06:40:00.0000] <ment>
AryehGregor: the problem with css and js is, that you have to maintain correct styling during execution of the script

[06:42:00.0000] <AryehGregor>
What do you mean?  You don't have to rerender styles synchronously during script execution, right?  If I do foo.width = 100; foo.height = 100; then foo is never actually rendered with width 100 but not height 100, AFAIK.

[06:42:01.0000] <gsnedders>
AryehGregor: No, but if you reflow during script execution perf can be bad (as is the case with HTML5 and Opera, which reflows on a timeout during script execution)

[06:43:00.0000] <AryehGregor>
Mm.

[06:43:01.0000] <ment>
AryehGregor: yes, you could postpone rematching. but in some cases you just have to rematch the whole tree after insertion of one node in the right place

[06:44:00.0000] <AryehGregor>
Interesting, although this is somewhat over my head.

[06:44:01.0000] <ment>
during the time i've been writing css matcher i've wondered why the selector language wasn't designed with some theory backing it up

[06:46:00.0000] <gsnedders>
AryehGregor: Otherwise, only trying to access the styles through CSSOM causes a reflow during script execution

[06:54:00.0000] <hsivonen>
annevk: in that case, I think Hixie should make the spec obviously misrender if the selectors aren't implemented right :-)

[06:55:00.0000] <hsivonen>
misrender in an obvious way, that is

[06:56:00.0000] <hsivonen>
applying the strategy of making stuff look ugly in WebKit to pressure WebKit devs to fix WebKit isn't working on my personal site, but doing it with the spec might work

[06:57:00.0000] <othermaciej>
hsivonen: I didn't know your site looked any different in WebKit-based browsers than it was supposed to

[06:57:01.0000] <othermaciej>
what's the bug?

[06:58:00.0000] <othermaciej>
other than the Firefox ad

[06:58:01.0000] <othermaciej>
but I assume that is by design

[06:58:02.0000] <hsivonen>
othermaciej: bug 1: text doesn't use a local font while fonts are downloading

[06:58:03.0000] <othermaciej>
ah, that I've noticed

[06:58:04.0000] <hsivonen>
othermaciej: bug 2: the double-struck rounded corners are ugly

[06:59:00.0000] <othermaciej>
that I didn't notice, but I see your point

[07:00:00.0000] <hsivonen>
which reminds me that I should add support for -moz- transitions

[07:00:01.0000] <hsivonen>
/me doesn't like vendor prefixes

[07:00:02.0000] <othermaciej>
your site tells me to upgrade to Firefox 3.5 when I visit with Firefox 3.6

[07:01:00.0000] <hsivonen>
othermaciej: yeah, the back end for that ad in FAIL

[07:01:01.0000] <hsivonen>
I guess I should remove the ad

[07:01:02.0000] <hsivonen>
s/in/is/

[07:02:00.0000] <othermaciej>
hmm

[07:02:01.0000] <othermaciej>
I wonder if we are even using the right font for your site at all

[07:03:00.0000] <othermaciej>
the font it ends up with in Safari seems to be the one that flashes temporarily in Firefox

[07:03:01.0000] <AryehGregor>
I also noticed on IE6 that it shows the "You're settling for good" message for a second before switching to the "You don't use VHS" one.

[07:05:00.0000] <hsivonen>
othermaciej: because Apple doesn't allow Mac OS X to be virtualized and because Safari detects uninstalled local fonts(!), it's too hard for me to test if the page looks right in Safari on a Mac that doesn't already have the fonts installed locally

[07:06:00.0000] <hsivonen>
maybe I should test on my mother's Mac when I visit my parents

[07:08:00.0000] <jgraham>
ment: Any theory you hand in mind?

[07:09:00.0000] <hsivonen>
I also added -o- and -ms- transitions completely defeating the point of vendor prefixes

[07:10:00.0000] <MikeSmith>
http://twitter.com/futomi/status/10978894529

[07:10:01.0000] <MikeSmith>
"#html5 spec says video.width must be DOMString while canvas.width must be unsigned long. Why inconsequence?"

[07:10:02.0000] <MikeSmith>
of course he means, "Why the inconsistency?"

[07:10:03.0000] <MikeSmith>
but anybody know the answer to that?

[07:15:00.0000] <hsivonen>
can someone who understands CSS Transitions explain to me why my link styling transitions work on unvisited links but not on visited links?

[07:16:00.0000] <hsivonen>
also, what's the deal with Opera supporting border-radius but not outline-radius?

[07:17:00.0000] <annevk>
outline-radius is not an official property

[07:17:01.0000] <hsivonen>
furthermore, Opera mistransitions the background color of my links

[07:17:02.0000] <annevk>
bugs.opera.com/wizard

[07:18:00.0000] <hsivonen>
hmm. the visited links do transition in WebKit

[07:18:01.0000] <hsivonen>
I guess I got one Gecko bug report and one Presto bug report out of this

[07:18:02.0000] <karlcow>
MikeSmith: it seems there are two attributes on video

[07:19:00.0000] <karlcow>
  readonly attribute unsigned long videoWidth;

[07:19:01.0000] <karlcow>
attribute DOMString width;

[07:20:00.0000] <karlcow>
oh and img

[07:20:01.0000] <karlcow>
attribute unsigned long width;

[07:20:02.0000] <karlcow>
readonly attribute unsigned long naturalWidth;

[07:21:00.0000] <MikeSmith>
/me nods

[07:21:01.0000] <karlcow>
a videoWidth is not a naturalWidth ?

[07:21:02.0000] <MikeSmith>
dunno

[07:21:03.0000] <MikeSmith>
I guess naturalWidth must be the intrinsic width

[07:21:04.0000] <MikeSmith>
and the other one is the displayed width

[07:22:00.0000] <annevk>
nope, both are intrinsic

[07:22:01.0000] <MikeSmith>
oh, ok

[07:22:02.0000] <karlcow>
nativeWidth have been exterminated I guess

[07:22:03.0000] <karlcow>
:p

[07:22:04.0000] <annevk>
i complained about this once, but didn't fight it through

[07:22:05.0000] <MikeSmith>
anyway, it seems futomi's question is, why is one width value a string datatype and another is an unsigned long

[07:23:00.0000] <annevk>
naturalWidth was already there, and videoWidth was implemented when we found that out

[07:23:01.0000] <annevk>
maybe someone should complain again

[07:23:02.0000] <annevk>
MikeSmith, dunno really

[07:24:00.0000] <karlcow>
object has no intrinsic width, just   attribute DOMString width;

[07:24:01.0000] <annevk>
i guess to be the same as <iframe> and <object>

[07:24:02.0000] <annevk>
<img>.width is quite special

[07:30:00.0000] <karlcow>
% grep -i width html5.txt | grep attribute | grep long

[07:30:01.0000] <karlcow>
               attribute unsigned long width <#dom-img-width>;

[07:30:02.0000] <karlcow>
      readonly attribute unsigned long naturalWidth <#dom-img-naturalwidth>;

[07:30:03.0000] <karlcow>
      readonly attribute unsigned long videoWidth <#dom-video-videowidth>;

[07:30:04.0000] <karlcow>
               attribute unsigned long width <#dom-canvas-width>;

[07:30:05.0000] <karlcow>
           attribute unsigned long width <#dom-pre-width>;

[07:31:00.0000] <karlcow>
% grep -i width html5.txt | grep attribute | grep DOMString

[07:31:01.0000] <karlcow>
               attribute DOMString width <#dom-dim-width>;

[07:31:02.0000] <karlcow>
           attribute DOMString width <#dom-applet-width>;

[07:31:03.0000] <karlcow>
           attribute DOMString width <#dom-marquee-width>;

[07:31:04.0000] <karlcow>
           attribute DOMString marginWidth <#dom-frame-marginwidth>;

[07:31:05.0000] <karlcow>
           attribute DOMString width <#dom-col-width>;

[07:31:06.0000] <karlcow>
           attribute DOMString width <#dom-hr-width>;

[07:31:07.0000] <karlcow>
           attribute DOMString marginWidth <#dom-iframe-marginwidth>;

[07:31:08.0000] <karlcow>
           attribute DOMString width <#dom-table-width>;

[07:31:09.0000] <karlcow>
           attribute DOMString width <#dom-tdth-width>;

[07:32:00.0000] <annevk>
oh, and <canvas> is different because width/height are just different there, they define the grid, not the dimensions

[07:36:00.0000] <Philip`>
<iframe width="100%">

[07:36:01.0000] <Philip`>
(I think that works?)

[07:36:02.0000] <Philip`>
so it's not just an integer

[07:37:00.0000] <Philip`>
but canvas dimensions are the number of pixels in the coordinate space, so they must be a fixed number

[07:37:01.0000] <karlcow>
can video be 100% too

[07:37:02.0000] <Philip`>
/me shrugs

[07:37:03.0000] <annevk>
nope

[07:38:00.0000] <annevk>
oh yes

[07:39:00.0000] <annevk>
it works for both, but is also invalid

[07:39:01.0000] <hsivonen>
HTMLDocument has 3 charset-related properties. Seriously?

[07:40:00.0000] <MikeSmith>
hsivonen: btw, http://html5.org/tools/web-apps-tracker?from=4861&to=4862 adds a new "allow-top-navigation" enumerated value to the allowed values for iframe/@sandbox

[07:41:00.0000] <MikeSmith>
which I think means instead of 3! combinations of the allowed values for that in the schema

[07:41:01.0000] <MikeSmith>
we now have 4!

[07:41:02.0000] <MikeSmith>
4 factorial

[07:41:03.0000] <hsivonen>
combinations FTW!

[07:41:04.0000] <MikeSmith>
heh

[07:42:00.0000] <MikeSmith>
is there any alternative to using an RNG list structure for this? 

[07:42:01.0000] <MikeSmith>
I guess there's probably not

[07:42:02.0000] <hsivonen>
MikeSmith: a datatype in Java

[07:42:03.0000] <Philip`>
Isn't it 4!+3!+2!+1!+0!?

[07:42:04.0000] <MikeSmith>
Philip`: is it?

[07:42:05.0000] <Philip`>
Any permutation of 4 values + any permutation of 3 values etc

[07:43:00.0000] <MikeSmith>
Philip`: what we have in the schema now is just 6 combinations

[07:43:01.0000] <hsivonen>
fortunately, you don't need to list all possibilities, because you can use the ? quantifier inside a list {}

[07:43:02.0000] <hsivonen>
/me wishes list {} permitted interleave

[07:45:00.0000] <MikeSmith>
hsivonen: do you think this merits creating a new dataype? just adding all the combinations in the schema is easy enough to do, just ugly

[07:45:01.0000] <Philip`>
hsivonen: Ah, right

[07:47:00.0000] <hsivonen>
MikeSmith: it doesn't merit creating a datatype at this point of spec development if adding the combos is easy enough

[07:47:01.0000] <hsivonen>
Hixie: it would be nice to have the new Gecko textarea resizer in Live DOM

[07:51:00.0000] <jgraham>
hsivonen: (why) Doesn't it work automatically?

[08:01:00.0000] <hsivonen>
jgraham: it seems it doesn't work automatically. dunno why

[08:01:01.0000] <hsivonen>
twitter seems to use it already

[08:02:00.0000] <hsivonen>
/me discovers that Opera doesn't implement document.charset and document.characterSet

[08:03:00.0000] <hsivonen>
and WebKit doesn't appear to support reloading document.open()ed docs from location.reload()

[08:08:00.0000] <hsivonen>
whoa! is WebKit the only browser that supports document.charset?

[08:08:01.0000] <hsivonen>
why do we have it in the spec if Gecko, IE and Opera don't have it?

[08:10:00.0000] <hsivonen>
I'm confused. Live DOM w("Foo") not working in IE8

[08:11:00.0000] <jgraham>
hsivonen: It never has

[08:11:01.0000] <annevk>
hsivonen, maybe Hixie didn't check enough of the other impls?

[08:11:02.0000] <hsivonen>
jgraham: ooh! suprising

[08:11:03.0000] <hsivonen>
jgraham: it works occasionally

[08:12:00.0000] <hsivonen>
oh. IE has .charset after all

[08:12:01.0000] <jgraham>
Yeah, I guess I should have inserted the word "reliably" in there somewhere

[08:12:02.0000] <hsivonen>
and for document.open()ed docs it's "unicode"

[08:12:03.0000] <jgraham>
(it is also differently broken in Chrome)

[08:13:00.0000] <hsivonen>
browsers shouldn't be so broken that you can't black box poke at them

[08:16:00.0000] <hsivonen>
anyway, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/420 is what I have so far

[08:16:01.0000] <hsivonen>
it didn't work the way I expected in the old parser in Gecko

[08:16:02.0000] <hsivonen>
I wonder why

[08:17:00.0000] <hsivonen>
(I expected Gecko to reparse the written content upon .close())

[08:17:01.0000] <hsivonen>
(so that the written script would run thrice)

[08:21:00.0000] <TabAtkins>
JonathanNeal: You win.  My response was *not* sent to the room, so I can't find it in the logs.  ;_;

[08:22:00.0000] <TabAtkins>
The gist of it was that, sure, a collection of article snippets might be an <article>.  No reason why it can't be.  That seems like something that could legitimate be an independent chunk of content.

[08:30:00.0000] <JonathanNeal>
TabAtkins, and rotating banners then could be as well, s'pose?

[08:31:00.0000] <TabAtkins>
Apply the rule of thumb.  Would you want to look at it fullscreen, or as an independent page?

[08:31:01.0000] <JonathanNeal>
What would the difference be?

[08:32:00.0000] <JonathanNeal>
When I look at it full screen, is it still interactive (this would be necessary in the article snippits as a collective article viewed fullscreen)

[08:32:01.0000] <TabAtkins>
Sorry, that was an or choice.  Just two variants on the question.

[08:32:02.0000] <TabAtkins>
s/was/wasn't/

[08:32:03.0000] <TabAtkins>
argh

[08:32:04.0000] <JonathanNeal>
Got it.

[08:33:00.0000] <zcorpan>
/me wonders where that rule of thumb comes from

[08:33:01.0000] <TabAtkins>
It comes from my own reading of the spec, and reasoning about what the difference should be between an <article> and <section>

[08:33:02.0000] <JonathanNeal>
So section almost acts like a more generic version of article, in that it doesn't specify that the content is wholey complete on its own, but that it is an independent entity in some form?

[08:33:03.0000] <TabAtkins>
Also what can be easily and quickly applied.

[08:33:04.0000] <TabAtkins>
JonathanNeal: Yeah, in my head at least <section> is the most generic sectioning element, and all the others (<article>, <aside>, <nav>, etc.) are specializations on it.

[08:34:00.0000] <JonathanNeal>
Did you catch my attempt at realizing http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#margins-and-padding last night?

[08:34:01.0000] <TabAtkins>
Nope.

[08:35:00.0000] <zcorpan>
i'd want to look at <video>s fullscreen, but wouldn't in general wrap <video>s in <article>s

[08:35:01.0000] <TabAtkins>
zcorpan: Sure, but <video> has its fullscreen-ness built in.

[08:35:02.0000] <JonathanNeal>
The section of x h1, x x h1 etc.

[08:35:03.0000] <TabAtkins>
JonathanNeal: Oh gods.

[08:35:04.0000] <JonathanNeal>
I tried to realize what that css would look like.

[08:36:00.0000] <JonathanNeal>
http://pastie.org/884363 in less css it is about 30 lines of code.

[08:36:01.0000] <TabAtkins>
Ideally, it would look like :section(1) h1, :section(2) h1, or similar.

[08:36:02.0000] <JonathanNeal>
http://pastie.org/884364 in actual css it is about 1383 lines of code

[08:37:00.0000] <TabAtkins>
Failing that, :any(article, aside, nav, section) h1, :any(article, aside, nav, section) :any(article, aside, nav, section) h1, etc.

[08:38:00.0000] <JonathanNeal>
I'm guessing :section(1) and :any(section) aren't supported in any browser yet?

[08:40:00.0000] <TabAtkins>
They're not even close to specced yet.

[08:40:01.0000] <TabAtkins>
:any() at least has wide support, though we haven't started Selectors 4 yet.

[08:42:00.0000] <annevk>
You need both :section() and :heading() I guess...

[08:42:01.0000] <annevk>
And maybe ::section if you want to have boxes for implied sections...

[08:42:02.0000] <annevk>
But getting that implemented and working dynamically might be quite hard

[08:42:03.0000] <JonathanNeal>
I tried any() and it did not work

[08:43:00.0000] <TabAtkins>
JonathanNeal: No, it doesn't exist yet.

[08:43:01.0000] <jgraham>
"wide support" meaning in the CSSWG not in extant UAs 

[08:43:02.0000] <JonathanNeal>
Ha

[08:43:03.0000] <TabAtkins>
Nod to jgraham.

[08:43:04.0000] <JonathanNeal>
Be right back.

[08:44:00.0000] <TabAtkins>
annevk: Yeah, that's the issue.  I'd like to create some sort of entity in selectors that can be *either* a pseudoelement *or* an alias to an existing element, so I can transparently use ::section(1) and refer to either the implicit section or an explicit <section>.

[08:45:00.0000] <annevk>
:::section

[08:45:01.0000] <annevk>
more colons is better

[08:45:02.0000] <jgraham>
Selectors gets more and more like Perl every day

[08:45:03.0000] <annevk>
only in #whatwg

[08:46:00.0000] <annevk>
well, and TabAtkins' dreams

[08:46:01.0000] <annevk>
:p

[08:46:02.0000] <jgraham>
Well it is pretty close already

[08:46:03.0000] <jgraham>
You get to play "guess what this random bit of punctuation does"

[08:47:00.0000] <annevk>
especially the attribute selectors are tricky

[08:52:00.0000] <zcorpan>
i wonder why some people think it's suboptimal to have a spec in two places

[08:52:01.0000] <annevk>
it defies their sense of logic

[08:53:00.0000] <annevk>
prolly roughly corresponds to the people that hate this channel

[08:53:01.0000] <zcorpan>
the relaxng spec is also in two places (with editorial differences)

[08:53:02.0000] <zcorpan>
i think it's good because sometimes when i need to access the spec, whatwg.org is down or i can't reach it, or w3.org is down

[08:56:00.0000] <TabAtkins>
annevk: Answer my email I sent you 12 hours ago.

[08:56:01.0000] <annevk>
I don't have that email? :/

[08:57:00.0000] <annevk>
only email I have from you is about serializing Selectors

[08:57:01.0000] <TabAtkins>
Hrm.  I sent it to the address listed on your blog.

[08:57:02.0000] <TabAtkins>
(Easiest place to find it on my phone.)

[08:57:03.0000] <TabAtkins>
(Since I don't have internet at the house until tomorrow.)

[08:58:00.0000] <annevk>
oh, no email there either?

[08:58:01.0000] <annevk>
also not in spam

[08:59:00.0000] <annevk>
did you make a typo?

[08:59:01.0000] <TabAtkins>
Man, I dunno.  I"ll just msg you instead.

[09:00:00.0000] <zcorpan>
http://twitter.com/futomi/statuses/10978894529

[09:00:01.0000] <annevk>
we discussed that a little earlier

[09:03:00.0000] <jgraham>
zcorpan: I assumed it was all about control

[09:20:00.0000] <JonathanNeal>
back TabAtkins

[09:28:00.0000] <JonathanNeal>
How come the rendering section of the html5 spec doesn't mention video, audio, canvas being display: block?  Are they meant to be something else?

[09:29:00.0000] <gsnedders>
Wow. I managed to crash my laptop an hour ago.

[09:30:00.0000] <JonathanNeal>
You needed an hour to discover this?

[09:31:00.0000] <Dashiva>
That's how responsive it is, I guess

[09:32:00.0000] <JonathanNeal>
I'm looking over http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html trying to see where I might assume it is display block or inline-block, any guesses?

[09:32:01.0000] <JonathanNeal>
Facts will work too :)

[09:32:02.0000] <jgraham>
YOu clearly should be working harder if it took an hour to notice

[09:32:03.0000] <gsnedders>
jgraham: Hey, it's not my fault I was talking to James!

[09:34:00.0000] <jgraham>
You weren't talking to me!

[09:36:00.0000] <gsnedders>
Did you miss me!?

[09:37:00.0000] <JonathanNeal>
I noticed there are some vendor styles consistant across Opera, Chrome, Safari, and Firefox --- is there a list they are gathering these from?

[09:37:01.0000] <JonathanNeal>
For instance, canvas, by default is always 300x150

[09:41:00.0000] <gsnedders>
JonathanNeal: Reverse engineering more than anything else

[09:41:01.0000] <JonathanNeal>
Haha

[09:41:02.0000] <JonathanNeal>
So I'll create a second css sheet which is what the vendors have said.

[09:41:03.0000] <gsnedders>
Life of a browser QA involves knowing all your competitor's behaviour as well as your own

[09:41:04.0000] <gsnedders>
:P

[09:41:05.0000] <gsnedders>
JonathanNeal: See HTML 5 rendering appendix

[09:42:00.0000] <JonathanNeal>
gsnedders, ^^ that's what I was referencing.

[09:43:00.0000] <JonathanNeal>
What's weird is that the rendering spec mentions elements that are no in the html5 spec.

[09:43:01.0000] <JonathanNeal>
afaik

[09:43:02.0000] <gsnedders>
Indeed it does

[09:44:00.0000] <JonathanNeal>
That's kinda weird, huh?

[09:44:01.0000] <zcorpan>
why is it weird?

[09:44:02.0000] <gsnedders>
Some things you need to render in a certain way for compat even if they have no semantic meaning :P

[09:44:03.0000] <zcorpan>
(which elements are those, btw?)

[09:48:00.0000] <zcorpan>
looking at the rendering section, i wonder if we should make <iframe topmargin> same-origin-only, and whether doing so would break any pages

[09:48:01.0000] <zcorpan>
sorry, marginheight

[09:52:00.0000] <zcorpan>
/me can't find any elements in rendering that's not mentioned elsewhere in the spec

[09:53:00.0000] <annevk>
zcorpan, e.g. <center>

[09:54:00.0000] <zcorpan>
annevk: the obsolete section defines center

[09:55:00.0000] <annevk>
I'm assuming those elements were excluded

[09:55:01.0000] <beowulf>
ah <center>, old friend

[09:55:02.0000] <TabAtkins>
zcorpan: Why would it be necessary to make marginheight same-origin-only?  You fearing some negative margins used to position the page to a particular spot?

[09:56:00.0000] <zcorpan>
TabAtkins: ooh, i didn't even think of negative margins

[09:57:00.0000] <TabAtkins>
I don't see how positive margins could be maliciously used.

[09:58:00.0000] <zcorpan>
seems negative margins don't work

[09:59:00.0000] <zcorpan>
the spec says

[09:59:01.0000] <zcorpan>
"⚠Warning! The above requirements imply that a page can change the margins of another page (including one from another origin) using, for example, an iframe. This is potentially a security risk, as it might in some cases allow an attack to contrive a situation in which a page is rendered not as the author intended, possibly for the purposes of phishing or otherwise misleading the user."

[10:10:00.0000] <JonathanNeal>
I'm trying to write the rendering spec out as an actual css file.

[10:11:00.0000] <JonathanNeal>
It's weird how they confirm that form is display: block, but not fieldset

[10:12:00.0000] <zcorpan>
the rendering section has some requirements that can't be expressed in css

[10:12:01.0000] <JonathanNeal>
and then legend is confirmed, which is a child of fieldset

[10:12:02.0000] <JonathanNeal>
So it's best left implied by the vendors?

[10:13:00.0000] <zcorpan>
"The fieldset element is expected to establish a new block formatting context."

[10:14:00.0000] <zcorpan>
what would be useful is a testsuite for the rendering section

[10:14:01.0000] <annevk>
and a spec for form controls

[10:22:00.0000] <erlehmann>
some guy just realized that you can do <!dOcTyPe HtMl>

[10:23:00.0000] <erlehmann>
WHAT HAVE YOU DONE IAN HICKSON

[10:23:01.0000] <othermaciej>
clearly we need to support w4r3z d00d syntax for all tags and attributes

[10:24:00.0000] <JonathanNeal>
Here you go http://sandbox.thewikies.com/html5-pangram/

[10:25:00.0000] <annevk>
erlehmann, in HTML4 you can write <hTmL> too...

[10:25:01.0000] <Philip`>
I like writing my pages like <!Doctype Html><Html Lang="En"><Head><Title>... because then it looks more like proper English

[10:26:00.0000] <othermaciej>
I write <!DOCTYPE html>, I am not sure what convinced me that is right

[10:26:01.0000] <Philip`>
Alternatively, I like writing <body><div><span>things</SPAN> <span>like</SPAN> this</DIV></BODY> because it helps to visually distinguish the different types of tag

[10:26:02.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/rendering.whatwg.css --- that's a collection of many of the styles I could actually see recommended by the whatwg

[10:26:03.0000] <annevk>
othermaciej, that works in XML too

[10:26:04.0000] <Philip`>
You can make good use of capitalisation

[10:26:05.0000] <jgraham>
Philip`: You should do <DIV><span></span></DIV> to distinguish block and inline elements

[10:26:06.0000] <othermaciej>
annevk: maybe I should add xmlns="http://long/thing/i/cant/remember" to all my documents after all

[10:27:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/rendering.vendor.css --- that's a collection of many of the styles I could see the vendors were also following

[10:27:01.0000] <annevk>
othermaciej, might wanna check with Paul if it's worth brownie points :)

[10:28:00.0000] <annevk>
/me appreciated the beer at TPAC

[10:28:01.0000] <erlehmann>
JonathanNeal, a style for headings using sections would be nice to

[10:28:02.0000] <erlehmann>
too

[10:29:00.0000] <zcorpan>
/me remembers that li[type=I] is still broken

[10:30:00.0000] <JonathanNeal>
erlehmann, does that have to do with decendants?  I tried not to include anything I didn't have to that had to do with decendants.

[10:30:01.0000] <JonathanNeal>
Just to get it done.  Are you talking about h1's affected by being in sectioning elements?

[10:30:02.0000] <JonathanNeal>
"sectioned content" as it is referred to.

[10:31:00.0000] <erlehmann>
JonathanNeal, exactly. it would be a huge mess.

[10:31:01.0000] <erlehmann>
maybe i should autogenerate it

[10:31:02.0000] <erlehmann>
currently i just use h1 to h3, but it grows exponentially

[10:32:00.0000] <JonathanNeal>
erlehmann, heh I did, it's 1383 lines of code.

[10:33:00.0000] <JonathanNeal>
Just for four or five levels of h1's.

[10:33:01.0000] <erlehmann>
wait, wat

[10:33:02.0000] <erlehmann>
why so much ?

[10:33:03.0000] <erlehmann>
argl

[10:33:04.0000] <erlehmann>
you mean 1383 lines of stylesheet ?

[10:33:05.0000] <JonathanNeal>
Because they are exponetial decendants.

[10:34:00.0000] <JonathanNeal>
To test every scenario multiples by itself each time you add a new layer.

[10:34:01.0000] <JonathanNeal>
So article, aside, nav, or section, that's 4, then 16, then 256, then 65536

[10:35:00.0000] <erlehmann>
i am shocked and appalled

[10:35:01.0000] <JonathanNeal>
Don't take me to the bank on that, but when I actually rendered them out they ended up being 1383 lines of css

[10:35:02.0000] <JonathanNeal>
I compared this to an example of less css, where it was 30 lines.

[10:36:00.0000] <zcorpan>
be happy there aren't 5 sectioning elements

[10:37:00.0000] <JonathanNeal>
So, the best way around this would be to give classnames to all the sectioning elements, and for the exact classname I would borrow from the spec ".sectioning-content"

[10:37:01.0000] <zcorpan>
or 7 heading levels

[10:37:02.0000] <JonathanNeal>
Attach those to every article, aside, nav, and section element and you'll only need 20 lines of css.

[10:38:00.0000] <zcorpan>
or use h1-h6 appropriately and you need 0 lines of css

[10:39:00.0000] <JonathanNeal>
I would label them ".sectioning-content" vs ".sectioning-element" to describe what it contains rather than what it is (like how roles work, example being main the container isn't itself the main content but contains the main content) ... which (tangent) is also why I do not like naming things "container".

[10:45:00.0000] <MikeSmith>
paul_irish - you're in Boston, right?

[10:49:00.0000] <mpilgrim>
jgraham: is http://james.html5.org/microdata/ maintained?

[10:49:01.0000] <jgraham>
mpilgrim: Not really

[10:50:00.0000] <mpilgrim>
ok

[10:50:01.0000] <mpilgrim>
are there similar tools elsewhere, besides google's rich snippets testing tool?

[10:50:02.0000] <jgraham>
foolip has something

[10:50:03.0000] <Philip`>
http://philip.html5.org/demos/microdata/demo.html is similar, in that it's also unmaintained

[10:51:00.0000] <Philip`>
http://foolip.org/microdatajs/live/

[10:51:01.0000] <Philip`>
That one's better

[10:52:00.0000] <Philip`>
(Google's rich snippets testing tool is probably a bad way to attempt to determine correct microdata parsing)

[10:53:00.0000] <mpilgrim>
i agree

[10:53:01.0000] <mpilgrim>
but i'm going to talk about microdata in "dive into html5" and the hook to get people to care is "google parses it"

[10:53:02.0000] <jgraham>
Philip`: Did you have a good way to find that or did you have a reference to it stored somewhere?

[10:55:00.0000] <Philip`>
mpilgrim: I'm not aware of anybody having tested Google's support (other than some trivial things like it parsing <p><p> wrong), so if you try it out then it'd be interesting to know how well it matches the spec

[10:55:01.0000] <Philip`>
jgraham: I used a search engine

[10:55:02.0000] <Philip`>
They're quite good at finding things

[10:55:03.0000] <jgraham>
Not for me, in this case

[10:55:04.0000] <mpilgrim>
thanks for the links

[10:56:00.0000] <mpilgrim>
back to bed

[10:56:01.0000] <mpilgrim>
(sick today)

[10:56:02.0000] <Philip`>
I searched for "philip microdata" to find my one, and foolip's one came immediately after it

[10:57:00.0000] <Philip`>
so it found his one even though I *wasn't* searching for it, which is a step up from old-fashioned search engines that only find what you're searching for

[11:12:00.0000] <JonathanNeal>
href is not a required attribute of base?

[11:13:00.0000] <JonathanNeal>
It is ... but html5.validator.nu and validator.w3.org let it slide.

[11:14:00.0000] <JonathanNeal>
It took IE6 for me to find that out.

[11:14:01.0000] <annevk>
is it?

[11:14:02.0000] <MikeSmith>
I don't think href is required on base

[11:14:03.0000] <annevk>
it's not required

[11:14:04.0000] <JonathanNeal>
"A base element must have either an href attribute, a target attribute, or both."

[11:14:05.0000] <JonathanNeal>
http://www.whatwg.org/specs/web-apps/current-work/#the-base-element

[11:14:06.0000] <annevk>
right, so href is not required if target is there

[11:15:00.0000] <JonathanNeal>
Sure, I wasn't aware --- I was validating with <base />

[11:15:01.0000] <MikeSmith>
hmm, this is probably a bug in the validator

[11:15:02.0000] <MikeSmith>
in the schema

[11:16:00.0000] <MikeSmith>
yeah

[11:16:01.0000] <MikeSmith>
/me wonders when that spec change was made

[11:18:00.0000] <MikeSmith>
JonathanNeal: fwiw, after I make the fix for this, for this case, the validator will eventually report, "Element base is missing one or more of the following attributes: href target"

[11:18:01.0000] <paul_irishhhh>
MikeSmith: yessir i am.

[11:19:00.0000] <JonathanNeal>
Coolio Mike!  I found this out after IE6 crapped out on me, which had nothing to do with that requirement, but that's how I found it (IE6 just doesn't support the self-closing <base />)

[11:20:00.0000] <MikeSmith>
JonathanNeal: thanks for catching it

[11:34:00.0000] <annevk>
ooh, voicexml

[11:34:01.0000] <annevk>
it returns

[11:35:00.0000] <annevk>
/me reads rest of the message

[11:35:01.0000] <annevk>
ok, maybe not

[11:36:00.0000] <zcorpan>
/me figured out why dailymotion doesn't work in opera

[11:38:00.0000] <annevk>
/me is curious

[11:39:00.0000] <JonathanNeal>
http://localhost/html5-pangram/ you know those whatwg styles really do conform most of the browsers to look about the same.

[11:39:01.0000] <JonathanNeal>
You still need to set some styles not mentioned that the vendors are following on video, canvas, etc, but for the most part they're all starting to look the same.

[11:39:02.0000] <JonathanNeal>
localhost!  For shame ...

[11:39:03.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/

[11:40:00.0000] <zcorpan>
annevk: we fire canplaythrough before the script registers the listener

[11:40:01.0000] <zcorpan>
annevk: dispatching a canplaythrough event from the address bar makes it work

[11:42:00.0000] <annevk>
oh argh, event loop bugs

[11:43:00.0000] <zcorpan>
i'm not sure it's an event loop bug, just a bad assumption about when the event will be fired

[11:44:00.0000] <annevk>
but it always fires in other browsers at the right time?

[11:44:01.0000] <annevk>
or does it sometimes not work in Firefox/Chrome?

[11:45:00.0000] <zcorpan>
i guess it could sometimes not work in other browsers if the video is cached or if the script file is slow to load

[11:45:01.0000] <annevk>
not saying it's an event loop bug per se btw, more like an event loop issue

[11:46:00.0000] <annevk>
maybe other browsers do the event loop thingie for this in a certain way that guarantees this scenario works

[11:48:00.0000] <zcorpan>
i think it has more to do with other browsers only firing canplay and canplaythrough when the video actually can play (through)

[11:49:00.0000] <zcorpan>
while opera lies and fires them early because we haven't implemented it properly yet

[11:51:00.0000] <zcorpan>
still, it's a network race problem and the script should be smarter and check the state of the video when the script runs instead of blindly waiting for an event that might already have been fired

[11:53:00.0000] <MikeSmith>
/me reads http://www.ietf.org/mail-archive/web/http-state/current/msg00718.html (rough notes from http-state WG face-to-face meeting at IETF 77)

[12:08:00.0000] <zcorpan>
/me wants to be able to do elm.dispatchEvent(new Event('canplaythrough')) instead of var e = document.createEvent('Event'); e.initEvent('canplaythrough', false, false); elm.dispatchEvent(e);

[12:16:00.0000] <annevk>
there was some discussion on better DOM Event APIs

[12:16:01.0000] <annevk>
not sure what happened to it :/

[12:22:00.0000] <Philip`>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Cdiv%20style%3Dlist-style-image%3Aurl%28image%29%3E%3Caudio%20controls%3E

[12:22:01.0000] <Philip`>
Interesting styling in Firefox

[12:49:00.0000] <Hixie>
anyone know what the url to the hybi meeting audio is going to be?

[12:49:01.0000] <Hixie>
or what irc channel they're on?

[12:49:02.0000] <Hixie>
i guess they'd use a jabber channel

[12:49:03.0000] <othermaciej>
it's jabber

[12:51:00.0000] <Hixie>
any idea what the url to the jabber thing is?

[12:53:00.0000] <othermaciej>
I'm afraid not, but I think it may have been sent to the hybi list

[12:53:01.0000] <othermaciej>
I have to escape the meeting for a while

[12:55:00.0000] <annevk>
http://www.ietf.org/mail-archive/web/hybi/current/msg00822.html

[12:55:01.0000] <annevk>
Hixie, ^^

[12:56:00.0000] <Hixie>
cool thanks

[12:59:00.0000] <annevk>
what's the program that handles audio streams for Ubuntu?

[12:59:01.0000] <annevk>
I think I nuked it the other day when trying to fix something

[12:59:02.0000] <Hixie>
dunno

[12:59:03.0000] <annevk>
and now I forgot the name and don't have sound :)

[12:59:04.0000] <Philip`>
Do you mean like PulseAudio or something else?

[12:59:05.0000] <Hixie>
quicktime failed for me, but itunes worked

[13:00:00.0000] <Hixie>
http://videolab.uoregon.edu/events/ietf/ietf778.m3u is the audio url

[13:00:01.0000] <annevk>
Philip`, cheers!

[13:00:02.0000] <Philip`>
"mplayer -playlist http://videolab.uoregon.edu/events/ietf/ietf778.m3u" works for that kind of stream

[13:01:00.0000] <JonathanNeal>
Hola

[13:01:01.0000] <zcorpan>
[hidden]:not(colgroup):not(col):not(thead):not(tbody):not(tfoot):not(tr):not(td):not(th) { display:none }

[13:01:02.0000] <zcorpan>
colgroup[hidden], col[hidden], thead[hidden], tbody[hidden], tfoot[hidden], tr[hidden], td[hidden], th[hidden] { visibility:collapse }

[13:01:03.0000] <annevk>
cool

[13:02:00.0000] <zcorpan>
wonder if browsers would be ok with that in the ua style sheet

[13:02:01.0000] <annevk>
installing pulseaudio doesn't require a restart

[13:02:02.0000] <annevk>
i have to say, curb your enthusiasm was more fun

[13:02:03.0000] <JonathanNeal>
zcorpan, did you want me to change my css to reflect that?

[13:02:04.0000] <TabAtkins>
zcorpan: That's what I'm hoping for.

[13:02:05.0000] <zcorpan>
JonathanNeal: no, unrelated

[13:02:06.0000] <JonathanNeal>
(minus the fact that :not() and those are unsupported.

[13:02:07.0000] <JonathanNeal>
:)

[13:03:00.0000] <TabAtkins>
:not() is supported!  Just not across all browsers yet.

[13:05:00.0000] <zcorpan>
does ie9 support :not()?

[13:06:00.0000] <zcorpan>
/me guesses so since it supposedly passes all selectors tests

[13:06:01.0000] <TabAtkins>
Dunno, I don't have a Vista or W7 machine to test it on.

[13:07:00.0000] <annevk>
minutes for hybi are being kept here: http://etherpad.com/FWABRLTLpK

[13:07:01.0000] <annevk>
not sure if everyone can view it or not, there seem to be some limitations

[13:08:00.0000] <annevk>
/me wonders whether to suggest that Opera prefers -5

[13:10:00.0000] <Dashiva>
annevk: Is that a typo?

[13:10:01.0000] <TabAtkins_>
Dashiva: No, just a non-sequitur; anne is talking about the hybi minutes.

[13:11:00.0000] <TabAtkins_>
Or, wait, maybe a typo?

[13:11:01.0000] <Dashiva>
-5 seems malformed, I'd expect -75 or -05

[13:11:02.0000] <TabAtkins_>
(If he means -75).

[13:11:03.0000] <annevk>
Dashiva, they are zero padded?

[13:11:04.0000] <MikeSmith>
JonathanNeal: http://www.w3.org/html/check?doc=http%3A%2F%2Fdev.w3.org%2Fhtml5%2Ftests%2Fvalidation%2Ffull%2Finvalid%2Fmissing-attributes%2Fbase-missing-href-target.html

[13:11:05.0000] <annevk>
yup they are

[13:12:00.0000] <JonathanNeal>
:)

[13:19:00.0000] <annevk>
discussion so far is mostly procedural

[13:19:01.0000] <annevk>
guess that makes sense for a new WG

[13:19:02.0000] <annevk>
but it's boring

[13:33:00.0000] <annevk>
Hixie, WebSocket.send() should throw if DOMString cannot be converted to a Unicode stream, not just for surrogate characters

[13:34:00.0000] <Hixie>
what other inputs would fail that way?

[13:35:00.0000] <franksalim>
would a string with \0s in the middle fail like that?

[13:35:01.0000] <annevk>
you can get other invalid code points otherwise, no?

[13:35:02.0000] <gsnedders>
How do you do string concat in Perl?

[13:35:03.0000] <Hixie>
gsnedders: . operator

[13:36:00.0000] <Hixie>
franksalim: i don't think so

[13:36:01.0000] <Hixie>
annevk: i can't think of any others... which ones?

[13:37:00.0000] <franksalim>
Hixie, but you could get a bad websocket frame as currently speced, right?

[13:37:01.0000] <gsnedders>
Hixie: thx

[13:37:02.0000] <Hixie>
franksalim: why?

[13:37:03.0000] <Hixie>
or how?

[13:37:04.0000] <Hixie>
0x00 0x00 0x00 0xFF is fine

[13:38:00.0000] <Hixie>
it's a 2-character 0x00 frame with data U+0000 U+0000

[13:38:01.0000] <annevk>
Hixie, oh, maybe you're right, but maybe it should use http://dev.w3.org/2006/webapi/WebIDL/#idl-DOMString instead?

[13:38:02.0000] <annevk>
that is http://dev.w3.org/2006/webapi/WebIDL/#dfn-obtain-unicode

[13:39:00.0000] <annevk>
would make sense if we use either that consistently or consistently throw

[13:39:01.0000] <Hixie>
we could do that too

[13:39:02.0000] <Hixie>
i don't really mind either way

[13:39:03.0000] <franksalim>
Hixie, i see

[13:42:00.0000] <annevk>
/me can't really think of all the parts of the platform that use it

[13:42:01.0000] <MikeSmith>
JonathanNeal: fwiw, http://dev.w3.org/html5/markup/base.html updated as well

[13:42:02.0000] <annevk>
I guess this is one reason why it'd be nice to have an API for specs

[13:42:03.0000] <JonathanNeal>
That looks great, thanks MikeSmith.

[13:43:00.0000] <MikeSmith>
thanks again for catching it

[13:46:00.0000] <JonathanNeal>
Oh boy, now I can finally get that tee shit "All your <base /> are belong to us"

[13:46:01.0000] <JonathanNeal>
I mean ... shirt

[13:55:00.0000] <annevk>
update on hybi: at this point people are reiterating mailing list discussion

[13:56:00.0000] <franksalim>
annevk, probably because nothing was ever decided conclusively on list

[13:56:01.0000] <annevk>
i doubt we'll get there now :)

[13:56:02.0000] <franksalim>
i'm ever optimistic

[13:57:00.0000] <annevk>
nice

[13:57:01.0000] <annevk>
/me wants that back

[13:59:00.0000] <gsnedders>
/me lost a fair bit of unsaved data today from his computer crashing. gah.

[14:02:00.0000] <MikeSmith>
annevk, franksalim - who's scribing?

[14:03:00.0000] <franksalim>
MikeSmith, http://etherpad.com/FWABRLTLpK

[14:04:00.0000] <JonathanNeal>
Anahiem, why that's where I am, more or less.

[14:04:01.0000] <MikeSmith>
franksalim: I meant, who's the main person writing there?

[14:05:00.0000] <MikeSmith>
I realize it's world-writable

[14:05:01.0000] <franksalim>
i do not know

[14:05:02.0000] <MikeSmith>
k

[14:05:03.0000] <franksalim>
i can see him, but i do not know his name

[14:05:04.0000] <annevk>
this meeting just got a bit weirder

[14:09:00.0000] <JonathanNeal>
I wondered that, when I saw you had an etherpad link, I thought "somebody's gonna be like UNICORNS"

[14:11:00.0000] <annevk>
oh versioning

[14:16:00.0000] <annevk>
seems we went right past that discussion

[14:17:00.0000] <annevk>
though it was mentioned it's best practice to have versioning o_O

[14:17:01.0000] <jgraham>
(just as a personal opinion re: websockets, I don't think counting bytes in a string is at all hard compared to implementing the sockets stuff. So I don't think using simplicity as an argument for having terminators rather than explicit lengths is really a good one)

[14:19:00.0000] <jgraham>
Also, etherpad for live minutes seems good; better than irc

[14:19:01.0000] <annevk>
but it's one of the bits you'd likely implement yourself

[14:19:02.0000] <annevk>
whereas the sockets bit is likely copied from a tutorial on sockets

[14:19:03.0000] <franksalim>
as opposed to a tutorial on websockets? using a websocket library?

[14:20:00.0000] <jgraham>
annevk: To do anything non-trivial with sockets seems harder than just copying a tutorial

[14:20:01.0000] <jgraham>
at least all the tutorials I have read are like "this is a simple echo server"

[14:20:02.0000] <franksalim>
my opinion is that cut and paste code on the server is not nearly as much a problem as it is with markup

[14:21:00.0000] <annevk>
you should talk to one of the security guys from Yahoo!

[14:21:01.0000] <jgraham>
Which is quite different to having to actually do non-trivial stuff with multiple simultaneous clients

[14:22:00.0000] <franksalim>
i am not denying the existence of bugs

[14:23:00.0000] <jgraham>
The main issue is that it is reasonably likely that typical western users would only test with ascii. However having a testsuite avaliable would help with that a great deal

[14:25:00.0000] <TabAtkins>
Anyone have suggestions for a good irc program on linux?  Preferably easily skinnable?

[14:26:00.0000] <Hixie>
irssi

[14:27:00.0000] <TabAtkins>
Argh, these repos are useless.  >_<

[14:27:01.0000] <jgraham>
There is also supposed to be some thing that you can run as a daemon on one machine and connect to via text or graphical clients

[14:27:02.0000] <jgraham>
But I forget the name

[14:28:00.0000] <jgraham>
(if the curses nature of irssi puts you off)

[14:28:01.0000] <annevk>
Opera!

[14:28:02.0000] <tabatkin1>
Hrm, probably puts me off.

[14:28:03.0000] <tabatkin1>
Can I just run chatzilla on linux?

[14:28:04.0000] <tabatkin1>
/me doens't know.

[14:29:00.0000] <Dashiva>
Run mIRC in wine? :P

[14:29:01.0000] <daedb>
Quassel?

[14:29:02.0000] <franksalim>
tabatkin1, don't like xchat?

[14:29:03.0000] <MikeSmith>
TabAtkins: XChat

[14:29:04.0000] <TabAtkins>
franksalim: Haven't tried it yet.  At the moment, connecting with pidgin, which sucks.

[14:29:05.0000] <jgraham>
Seriously, being able to keep the client running all the time is teh awesome

[14:32:00.0000] <TabAtkins>
Argh, now what's the server name of the w3c?

[14:32:01.0000] <annevk>
irc:w3.org:80

[14:32:02.0000] <annevk>
irc.w3.org:80

[14:32:03.0000] <annevk>
i mean

[14:32:04.0000] <TabAtkins>
Hrm, xchat won't recognize it.

[14:33:00.0000] <jgraham>
Maybe I mean smuxi

[14:33:01.0000] <TabAtkins>
Looks like it's not successfully parsing out the port?

[14:34:00.0000] <jgraham>
iirc the way you enter ports in xchat is odd

[14:34:01.0000] <franksalim>
TabAtkins, works for me irc.w3.org/80

[14:34:02.0000] <franksalim>
slash instead of :

[14:34:03.0000] <TabAtkins>
"/server irc.w3.org/80"?

[14:34:04.0000] <annevk>
o_O

[14:35:00.0000] <TabAtkins>
?_?

[14:36:00.0000] <gavin>
/me would not expect irc.w3.org/80 to work anywhere

[14:36:01.0000] <TabAtkins>
Man, xchat sucks too.  I DECLARE IT SO.

[14:36:02.0000] <gavin>
unless you wanted to join channel #80 on irc.w3.org:6667

[14:36:03.0000] <TabAtkins>
/me goes to try and get chatzilla working, since he knows how that works.

[14:37:00.0000] <hober>
what ever happened to port 194?

[14:40:00.0000] <tabatkins__>
There, chatzilla is up.  Woo!

[14:40:01.0000] <tabatkins__>
Finally a sensible client.

[14:40:02.0000] <Dashiva>
I'm not going to take the obvious opening there...

[14:40:03.0000] <tabatkins__>
Now I just need to go grab my skin from my other laptop and I'll be set.

[14:49:00.0000] <Philip`>
"Bug 7034"

[14:49:01.0000] <Philip`>
Hooray for descriptive mailing list thread titles

[14:50:00.0000] <othermaciej>
if you post on that thread I'd suggest subject change

[14:53:00.0000] <Dashiva>
/me ponders making the list archives autolink bugs and issues

[15:15:00.0000] <annevk>
fixed the title

[15:16:00.0000] <annevk>
oh hey

[15:16:01.0000] <annevk>
shelley is leaving us again

[15:16:02.0000] <annevk>
http://twitter.com/shelleypowers/status/10993810486

[15:16:03.0000] <annevk>
http://twitter.com/shelleypowers/status/10995234211

[15:17:00.0000] <annevk>
"The spec is absolute crap"

[15:18:00.0000] <annevk>
if it was, surely authors would not like it and browser vendors would not implement it?

[15:18:01.0000] <TabAtkins_>
She uses a peculiar definition of "crap".

[15:18:02.0000] <Hixie>
dunno, never stopped them before :-P

[15:18:03.0000] <annevk>
hehe good point

[15:22:00.0000] <othermaciej>
when it comes to browser engineering, we spell quality with a *capital* K

[15:23:00.0000] <Dashiva>
Kwality? Or Kuality?

[15:24:00.0000] <Hixie>
Kaqlity!

[15:24:01.0000] <Hixie>
er

[15:24:02.0000] <Hixie>
i can't even spell my misspellings

[15:24:03.0000] <Hixie>
nevermind

[15:24:04.0000] <Hixie>
on another note: the hybi meeting was somewhat uneventful, fwiw, though i was a little concerned that the response to someone mentioning versioning was "well versioning is easy, we'll just copy some boilerplate from another draft to add version support" or some such

[15:25:00.0000] <annevk>
yeah... or maybe we don't

[15:25:01.0000] <othermaciej>
Dashiva: we feel that the exact spelling is an opportunity for differentiation and should be left up to the implementation

[15:34:00.0000] <TabAtkins_>
hsivonen: Is Leif making stuff up when he suggests that you are saying that @hidden should, generally, have a default CSS of something other than display:none?

[15:47:00.0000] <othermaciej>
TabAtkins_: the spec certainly says the default rendering of @hidden should be display: none

[15:47:01.0000] <TabAtkins_>
I know.  Leif is suggesting that hsivonen said something different, which I suspect is him being crazy.

[15:47:02.0000] <Dashiva>
Hum, Leif repeated the <p /> thing...

[15:49:00.0000] <Philip`>
/me finds the effort/reward ratio means he chooses never to read anything Leif ever says

[15:49:01.0000] <Philip`>
(which is a shame since sometimes he says things that seem useful and correct)

[15:49:02.0000] <TabAtkins_>
Philip`: Yeah, generally true.  This thread I'm in was just so batshit that I had to correct it.

[15:50:00.0000] <Hixie>
someone was wrong on the internet?

[15:51:00.0000] <Dashiva>
Is being wrong a violation of any conformance criteria?

[15:51:01.0000] <TabAtkins_>
Only in my dreams.

[15:52:00.0000] <Philip`>
You dream about conformance criteria?

[15:53:00.0000] <Hixie>
don't you?

[15:53:01.0000] <TabAtkins_>
...yes?

[15:53:02.0000] <TabAtkins_>
That's why you're not a spec editor, Philip`

[15:57:00.0000] <jgraham>
Philip` never sleeps so he can't dream

[15:59:00.0000] <Dashiva>
There is an installed base of RDFa-in-HTML that we can't risk destabilizing, how nice

[15:59:01.0000] <dbgi>
anyone in here who knows freebsd well and enjoys setting up and hosting web sites?

[16:00:00.0000] <Hixie>
does anyone implemented SharedWorker? I want to test zcorpan's suggested new examples

[16:02:00.0000] <annevk>
we have but it's not out in public yet I think

[16:02:01.0000] <annevk>
doesn't Chrome have it too?

[16:02:02.0000] <annevk>
anyways, bed time; nn

[16:02:03.0000] <Hixie>
nn

[16:03:00.0000] <Dashiva>
Sounds like a plan...


2010-03-25
[17:30:00.0000] <Hixie>
does webidl define how callback return values get converted to idl values?

[17:30:01.0000] <Hixie>
i can't find it anywhere

[19:25:00.0000] <hober>
with 3/31 approaching, I suppose I should work on http://hober.jottit.com/ISSUE-95_Change_Proposal more

[19:26:00.0000] <hober>
any and all feedback greatly appreciated.

[19:42:00.0000] <boblet>
hixie: you there?

[21:24:00.0000] <Hixie>
borey: here

[21:24:01.0000] <Hixie>
er

[21:25:00.0000] <Hixie>
nevermind, boblet left

[21:25:01.0000] <Hixie>
sorry borey, you were next on the nick completion list :-)

[00:00:00.0000] <zcorpan>
Hixie: was EXAMPLE workers/shared/001/test.html supposed to be replaced in the generated spec?

[00:27:00.0000] <zcorpan>
"our html5 involvement goes back years and years and years" - http://ecn.channel9.msdn.com/o9/ch9/8/2/8/8/3/5/ch9livemix10dean_ch9.mp4

[00:28:00.0000] <Hixie>
zcorpan: was it not replaced? hm

[00:28:01.0000] <Hixie>
wonder what's going on with wikileaks

[00:29:00.0000] <JonathanNeal>
Hey zcorpan, Hixie

[00:29:01.0000] <Hixie>
oh, too many /s for the EXAMPLE

[00:29:02.0000] <Hixie>
/me fixes

[00:31:00.0000] <zcorpan>
"will ie9 support canvas?" "everything in ie9 is hardware accelerated!"

[01:02:00.0000] <JonathanNeal>
did you guys see the pangram?

[01:05:00.0000] <hsivonen>
jgraham, Hixie: I see the textarea resizer on Live DOM now. I guess the resizer appears by default but I was using the wrong build yesterday

[01:06:00.0000] <zcorpan>
Hixie: the second example seems to be 001/

[01:11:00.0000] <Hixie>
i suck

[01:11:01.0000] <Hixie>
fixing

[01:11:02.0000] <Hixie>
thanks

[01:17:00.0000] <nessy>
it's bad hair day

[01:20:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-pangram/ and http://sandbox.thewikies.com/html5-pangram/short.html

[01:20:01.0000] <annevk>
/me always has a bad hair morning; unless it's extremely short

[01:22:00.0000] <zcorpan>
JonathanNeal: you should put some ads on that page

[01:23:00.0000] <JonathanNeal>
So it can be even uglier?

[01:23:01.0000] <JonathanNeal>
I'm not done, the short will be how I write an html5 normalize stylesheet

[01:23:02.0000] <zcorpan>
no, to match expectations; i think you've posted the link about 20 times now :P

[01:24:00.0000] <JonathanNeal>
It first follows the whatwg's rendering recommendation, followed then by popular vendor styling.

[01:24:01.0000] <annevk>
JonathanNeal, most people in this channel read the backlog

[01:24:02.0000] <JonathanNeal>
Note taken.

[01:24:03.0000] <JonathanNeal>
:)

[01:25:00.0000] <JonathanNeal>
I didn't even know there was a backlog to be read.  I thought the best strategy was to always stir up trouble.

[01:25:01.0000] <Hixie>
zcorpan: who implements the three-argument postMessage on Window?

[01:26:00.0000] <Hixie>
zcorpan: if nobody does, or if everyone is willing to change their implementations, i don't mind changing it

[01:26:01.0000] <zcorpan>
Hixie: opera does (not shipped yet, but is implemented), not sure about the others

[01:26:02.0000] <Hixie>
zcorpan: the reason postMessage()'s ports argument was added in the middle was to keep it closer to the message argument, where it makes more sense

[01:26:03.0000] <Hixie>
i.e. having "message, message, target" rather than "message, target, message"

[01:27:00.0000] <zcorpan>
Hixie: i think the common reaction to having it in the middle is "WTF?"

[01:27:01.0000] <Hixie>
if everyone is willing to change their implementations, i don't mind changing it

[01:27:02.0000] <Hixie>
no need to convince me on logical grounds :-)

[01:28:00.0000] <zcorpan>
/me can only speak for opera

[01:33:00.0000] <Hixie>
well if you can find out who else implements it i can go ask them

[01:34:00.0000] <annevk>
pretty sure WebKit must have it

[01:34:01.0000] <zcorpan>
Hixie: it seems chrome implements it

[01:35:00.0000] <Hixie>
k, i'll ask them about it tomorrow

[01:35:01.0000] <zcorpan>
but not firefox

[01:36:00.0000] <Hixie>
ooh, new freeman's mind

[01:36:01.0000] <Hixie>
bbiab!

[01:43:00.0000] <Hixie>
ok where was i

[01:43:01.0000] <zcorpan>
you were about to fix all my bugs

[01:48:00.0000] <Hixie>
oh yes

[02:11:00.0000] <zcorpan>
it's weird that people's reaction to "feature x and y seem like they should do what i want, but are buggy in browsers A and B" is "how about introducing feature z?" instead of "how about fixing x and y in A and B?"

[02:14:00.0000] <Hixie>
yeah

[02:14:01.0000] <Hixie>
i see that all the time also

[02:20:00.0000] <Hixie>
anyone else having difficulty using the bug filing thing in the spec?

[02:27:00.0000] <Hixie>
ok bed time

[02:27:01.0000] <Hixie>
nn

[02:31:00.0000] <zcorpan>
Hixie: nn

[02:34:00.0000] <gsnedders>
Anyone with IE9 around?

[02:35:00.0000] <Philip`>
Yes

[02:35:01.0000] <gsnedders>
Philip`: http://stuff.gsnedders.com/es/getterparsing.html

[02:35:02.0000] <gsnedders>
Philip`: What does that give?

[02:38:00.0000] <Philip`>
FAIL (threwEvalError: Expected ':')

[02:38:01.0000] <Philip`>
6 times

[02:39:00.0000] <gsnedders>
Huh

[02:39:01.0000] <gsnedders>
I thought IE supported it...

[02:40:00.0000] <lazni>
is opera's treatment of page-break-* in @media projection okay in standards?

[02:41:00.0000] <virtuelv>
lazni: to answer that question, you would have to first explain what it is you think Opera does with it

[02:41:01.0000] <annevk>
lazni, yes it is

[02:41:02.0000] <lazni>
it split advancement by pagedown by the pages

[02:41:03.0000] <annevk>
lazni, projection is paged-based media

[02:42:00.0000] <lazni>
okay

[02:42:01.0000] <lazni>
/me is writing a CSS+fragment slideshow

[03:17:00.0000] <hsivonen>
sigh. I just had a "Someone is wrong on the Internet" moment about http://robertnyman.com/2010/03/23/what-will-happen-to-open-video-on-the-web/

[03:18:00.0000] <hsivonen>
(I left a comment)

[03:24:00.0000] <annevk>
hmm, from that post (via #fronteers): "My friend spoke to a Mozilla guy at Libre Planet this last weekend who said Mozilla has a “big announcement” regarding video coming this week or next."

[03:35:00.0000] <jgraham>
Hmm, "my friend said someone said something was going to happen" doesn't exactly sound concrete

[03:35:01.0000] <Dashiva>
hsivonen: It's disappointing to see people still think using Vorbis has any weight at all on how safe Theora is

[03:37:00.0000] <hsivonen>
Dashiva: I didn't even read that far carefully before I went into the "Someone is wrong on the Internet" mode.

[03:39:00.0000] <zcorpan>
Dashiva: leave a comment?

[03:40:00.0000] <Dashiva>
People will still be wrong

[03:41:00.0000] <zcorpan>
but you know you get a warm fuzzy feeling when pointing it out

[03:43:00.0000] <Philip`>
If the Ogg people said Vorbis was perfectly safe and patent-free and encouraged everyone to use it, and then a few years later everybody got sued for infringing patents, then that would cast significant doubt over similar claims about Theora

[03:43:01.0000] <Philip`>
but everyone uses Vorbis and nobody has been sued yet, so that particular source of doubt has not been cast

[03:43:02.0000] <Philip`>
so it does seem to have some weight on the safety of Theora

[03:48:00.0000] <Dashiva>
In the same way having built a shed does give some weight to the guy who wants to build your skyscraper

[03:51:00.0000] <Philip`>
Sure

[03:51:01.0000] <jgraham>
For large values of "shed"

[05:58:00.0000] <lazni>
is there any CSS way I can make the nav#toc xx-large when there's no fragment id, but small when the fragment id is not "toc" in http://hdh.dyn-o-saur.com/public-domain/doc/sd1/goal-setting.html ?

[06:02:00.0000] <Rik`>
:target ? 

[06:12:00.0000] <lazni>
:not(:target) makes it large when another id is in focus too

[06:13:00.0000] <AryehGregor>
No, pretty sure there's no way to do that.  Use JS.

[06:18:00.0000] <hsivonen>
aaargh. I've broken validator.nu buildability with javac

[06:19:00.0000] <hsivonen>
ecj still works

[06:52:00.0000] <zcorpan>
hmm, maybe i should look at the entity data

[06:54:00.0000] <MikeSmith>
hsivonen: how'd you manage to break buildability with javac?

[06:58:00.0000] <zcorpan>
Philip`: have you tried http://krijnhoetmer.nl/irc-logs/whatwg/20100311#l-738 ?

[07:01:00.0000] <Philip`>
zcorpan: I think I forgot

[07:07:00.0000] <hsivonen>
MikeSmith: I changed NamedCharacters.java to use byte[][] for names

[07:07:01.0000] <hsivonen>
MikeSmith: ecj optimized the static initializer to under 2^16 bytes

[07:07:02.0000] <hsivonen>
MikeSmith: javac didn't

[07:07:03.0000] <MikeSmith>
ah

[07:08:00.0000] <hsivonen>
so javac failed to produce a legal .class file

[07:14:00.0000] <othermaciej>
howdy everyone

[07:15:00.0000] <annevk>
public-html :/

[07:15:01.0000] <annevk>
besides that pretty good

[07:15:02.0000] <annevk>
just had my first beer in the open, weather is changing for the better :)

[07:16:00.0000] <Philip`>
zcorpan: That regexp looks scary so maybe http://philip.html5.org/data/meta-content-charset-no-http-equiv-content-type.txt is better

[07:16:01.0000] <Philip`>
zcorpan: That's from egrep -i 'content="[^;"]*;\s*charset='|grep -iv 'http-equiv="content-type"'

[07:16:02.0000] <jgraham>
/me wonders who annevk is talking to

[07:17:00.0000] <Philip`>
on a file containing <meta uri http-equiv name content> extracted from parsed pages

[07:17:01.0000] <annevk>
jgraham, othermaciej?

[07:17:02.0000] <hsivonen>
/me wonders if WebKit and Opera have set themselves up for Microsoft moments like https://bugzilla.mozilla.org/show_bug.cgi?id=546857#c22

[07:18:00.0000] <hsivonen>
/me has no idea what kind of maintenance burdens Dashboard and Opera Widgets are

[07:19:00.0000] <jgraham>
annevk: Oh. I'm not really sure that follows from "Howdy everyone"

[07:19:01.0000] <annevk>
jgraham, it just did...

[07:20:00.0000] <othermaciej>
heh :-)

[07:20:01.0000] <annevk>
hsivonen, I suppose they're somewhat closer to HTML/CSS than XUL, other than that I have no idea

[07:21:00.0000] <zcorpan>
Philip`: nice, thanks

[07:22:00.0000] <AryehGregor>
"CSSStyleDeclaration" hurts my eyes.  Too many S's.

[07:23:00.0000] <annevk>
should have been CSSDeclarationBlock but it's too late for that now

[07:23:01.0000] <AryehGregor>
Acronyms in names should be capitalized like normal words.  That's the only sane way to do it.

[07:23:02.0000] <annevk>
much stuff in the CSSOM has weird names

[07:23:03.0000] <AryehGregor>
The convention that gives you CSSStyleDeclaration would also give you XMLHTTPRequest, which is just evil.

[07:23:04.0000] <AryehGregor>
CssStyleDeclaration would be so much more sensible.

[07:23:05.0000] <annevk>
euhm, now it is XMLHttpRequest

[07:24:00.0000] <annevk>
which is also weird

[07:24:01.0000] <AryehGregor>
Well, that's even more insane.

[07:24:02.0000] <zcorpan>
Philip`: there were no other attributes, like httpequiv? or http-equiv=contenttype?

[07:24:03.0000] <annevk>
yeah, not sure what you're saying

[07:24:04.0000] <hsivonen>
AryehGregor: more of a .NET than Java naming fan, eh?

[07:24:05.0000] <AryehGregor>
I don't touch either.

[07:25:00.0000] <AryehGregor>
annevk, I would do things like XmlHttpRequest.

[07:25:01.0000] <AryehGregor>
That way you can tell where words begin and end.

[07:26:00.0000] <annevk>
not really interesting to argue about such conventions now

[07:26:01.0000] <AryehGregor>
No, not at all, I was just randomly spamming.

[07:26:02.0000] <hsivonen>
I've used the .NET style in Java to improve readability

[07:26:03.0000] <hsivonen>
one of the main benefits .NET has over Java :-)

[07:26:04.0000] <AryehGregor>
Well, I guess it could be remembered for the future, if anyone gets to set new conventions for new languages or whatnot.

[07:26:05.0000] <AryehGregor>
hsivonen, so what *are* the .NET and Java styles?

[07:27:00.0000] <hsivonen>
CssStyleDeclaration would be .NET and CSSStyleDeclaration Java

[07:27:01.0000] <Philip`>
zcorpan: This data is just <meta uri http-equiv name content>

[07:27:02.0000] <Philip`>
i.e. it ignores any other attributes that may exist

[07:27:03.0000] <zcorpan>
Philip`: ok

[07:28:00.0000] <AryehGregor>
Then I'm a .NET convention fan.  One thing Microsoft did right.  :P

[07:28:01.0000] <zcorpan>
Philip`: i was kinda also interested in any other attributes that may exist :)

[07:29:00.0000] <Philip`>
zcorpan: I think I'd have to modify the Java code to do that properly

[07:29:01.0000] <Philip`>
which is possible

[07:29:02.0000] <Philip`>
(later, if I don't forget)

[07:30:00.0000] <Philip`>
The regexp wouldn't match anything across multiple lines so that's probably not so good

[07:43:00.0000] <hsivonen>
grr. maybe I should just ship a .class file compiled using Eclipse and stop supporting compilation using javac

[07:45:00.0000] <hsivonen>
alternatively, I could split NamedCharacter.java so that each member is in a class alone

[07:46:00.0000] <othermaciej>
is the file too big to compile with javac?

[07:46:01.0000] <hsivonen>
othermaciej: the output from javac is too big

[07:46:02.0000] <hsivonen>
to fit within the spec for .class files

[07:47:00.0000] <hsivonen>
the corresponding C++ file has been a problem with GCC and MSVC, too

[07:48:00.0000] <annevk>
hmm, CSS Value API is somewhat hard

[07:49:00.0000] <Philip`>
Making it harder to edit and recompile the code with standard tools doesn't sound like a good idea from a third-party contributor perspective

[07:51:00.0000] <hsivonen>
Philip`: Eclipse is the standard tool, isn't it. :-)

[07:51:01.0000] <hsivonen>
but yeah, I guess I need to split this file

[07:51:02.0000] <hsivonen>
and do the same split on the C++ side

[07:53:00.0000] <hsivonen>
this isn't the first time that javac produces more verbose bytecode than ecj

[07:54:00.0000] <Philip`>
Does javac attempt any kind of optimisation at all?

[07:54:01.0000] <hsivonen>
dunno

[07:54:02.0000] <Philip`>
/me briefly looked at its output once and it didn't seem to do any of the obvious things that would save space and time

[07:55:00.0000] <Philip`>
but I may have been mistaken

[07:56:00.0000] <JonathanNeal>
word

[07:57:00.0000] <Philip`>
JonathanNeal: excel

[07:57:01.0000] <JonathanNeal>
:)

[07:57:02.0000] <Philip`>
Wasn't that the pattern?

[07:58:00.0000] <JonathanNeal>
After one entry?  Very presumptuous Philip`

[08:00:00.0000] <Philip`>
You get more points for spotting the pattern sooner

[08:00:01.0000] <Dashiva>
More like assuming

[08:00:02.0000] <Philip`>
I was hoping "sharepoint workspace" would come next

[08:04:00.0000] <JonathanNeal>
Unfortunately, the next entry was to be "sentence"

[08:04:01.0000] <zcorpan>
hmm, opera doesn't do incremental rendering of giant text nodes in xml

[08:05:00.0000] <jgraham>
Microsoft Sentence? 

[08:05:01.0000] <annevk>
/me asks for thoughts on www-style

[08:15:00.0000] <MikeSmith>
hsivonen: fwiw, I don't think adding a dependency on ecj would be too unreasonable

[08:15:01.0000] <MikeSmith>
if that's what you were to decide to do

[08:15:02.0000] <MikeSmith>
it's cross-platform and easy to install

[08:17:00.0000] <zcorpan>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9207 - anyone feeling like analyzing the individual characters to see if they'd benefit from changing the parser?

[09:16:00.0000] <JonathanNeal>
Excell

[09:20:00.0000] <Philip`>
Where did the ent go?

[09:20:01.0000] <jgraham>
That sounds like the start of a Tolkien joke

[09:48:00.0000] <JonathanNeal>
Philip`, I don't know.

[10:18:00.0000] <CrazyFoam>
Hello, I have a question regarding XML Namespaces

[10:18:01.0000] <CrazyFoam>
say I have <a xmlns="urn:a" xmlns:b="urn:b"> <b:b> <c/> </b:b> </a>

[10:18:02.0000] <CrazyFoam>
element <c/> is assumed to be in namespace urn:a

[10:19:00.0000] <CrazyFoam>
but if I have <a:a xmlns:a="urn:a" xmlns:b="urn:b"> <b:b> <c/> </b:b> </a:a>

[10:19:01.0000] <CrazyFoam>
is element <c/> invalid or assumed to be in namespace urn:b?

[10:20:00.0000] <Philip`>
In the second case, the <c/> is in no namespace

[10:20:01.0000] <Philip`>
(which is a perfectly legitimate thing to do)

[10:20:02.0000] <CrazyFoam>
ah that's what I thought, but how is it legitimate

[10:20:03.0000] <CrazyFoam>
say the doc has no default namespace

[10:20:04.0000] <Philip`>
XML doesn't require namespaces

[10:20:05.0000] <CrazyFoam>
whatever processing the doc may not be able to understand <c/>

[10:21:00.0000] <Philip`>
You could write <a><b><c/></b></a> and the <c/> would be in no namespace, just as in your example

[10:21:01.0000] <Philip`>
Namespaces don't cause understanding

[10:22:00.0000] <annevk>
they just cause confusion

[10:22:01.0000] <CrazyFoam>
haha

[10:22:02.0000] <Philip`>
I could write <a:axghrlktxjhlkj xmlns:a="urn:a"/> and whatever's processing the document would have just as much chance of understanding it as of understanding <c/>

[10:22:03.0000] <Dashiva>
You could use the namespace URI as input to a RDF-based knowledge network that would provide all the data you need!

[10:23:00.0000] <Philip`>
You'd want some kind of schema if you want to restrict the document to known elements

[10:23:01.0000] <CrazyFoam>
knowledge network haha

[10:24:00.0000] <CrazyFoam>
okay well thanks for the info, really what i was looking for is that <c/> in the last case defeintiely shouldn't get interptreted as belonging to either urn:a or urn:b

[10:25:00.0000] <Philip`>
Yep - if it has no prefix then it's in the default namespace, and the default namespace is only affected by xmlns="..."

[10:27:00.0000] <CrazyFoam>
on that note, what do you call it when a doc changes default namespace all over the place, e.g. <doc xmlns="urn:doc"> <a xmlns="urn:a"/> <b xmlns="urn:b"/> </doc>, still valid right? can't think of a term to search for specifications - "default namespace switching"?

[10:36:00.0000] <Philip`>
CrazyFoam: You mean when it changes the default namespace in non-root elements?

[10:37:00.0000] <Philip`>
I guess it's neurotic by the definitions in http://lists.xml.org/archives/xml-dev/200204/msg00170.html

[10:40:00.0000] <CrazyFoam>
hmmm... is that post discussing maintaining prefix names through serialization?

[10:41:00.0000] <Philip`>
That's part of it

[10:41:01.0000] <CrazyFoam>
ah got to the neruotic part

[10:41:02.0000] <CrazyFoam>
what is a GI?

[10:43:00.0000] <Philip`>
Tag name

[10:43:01.0000] <Philip`>
('Generic identifier')

[10:47:00.0000] <lazni>
http://www.mikematas.com/ is rather empty in opera

[10:49:00.0000] <brucel>
Is there a webkit guru in the house?

[10:49:01.0000] <Philip`>
There might be more in #webkit

[10:49:02.0000] <jgraham>
brucel: I suggest you just ask the question and someone will probably answer

[10:50:00.0000] <jgraham>
lazni: Which version of OPera, and what exactly do you see?

[10:51:00.0000] <lazni>
10.51, I see only the 3 menu items on the left

[10:52:00.0000] <jgraham>
lazni: Curious. It works for me, although it is a bit slow to load (and has an obnoxious alert telling me to get a modern browser "such as Apple Safari")

[10:52:01.0000] <brucel>
Am trying to build up video demo that, if Ogg video isn;t supported (but video *is*, eg Safari) it will fall back to the contents of the video element. But can't get it working - onerror event seems not to fire.  http://people.opera.com/brucel/dev/html5-video-fallback.html Have a brain full of cold so am trying to understand where I'm going wrong

[10:53:00.0000] <lazni>
I restarted opera and the page loads now

[10:53:01.0000] <jgraham>
lazni: Could be cache bug or something I guess

[10:53:02.0000] <lazni>
but it doesn't pop up about safari

[10:54:00.0000] <jgraham>
If you work out how to reproduce then file a bug

[10:54:01.0000] <annevk>
brucel, did you check that error is dispatched at all?

[10:54:02.0000] <lazni>
okay

[10:55:00.0000] <brucel>
annevk, seems never to fire

[10:55:01.0000] <jgraham>
lazni: You're not missing much. There's nothing that says "retard" quite like being a supposed UI designer who uses a modal alert to tell people that a site might not work in a browser that it actually works fine in and that they should buy a Mac instead

[10:56:00.0000] <lazni>
and a totallly JS-driven site

[10:57:00.0000] <annevk>
brucel, fwiw, it should be dispatched on the video element

[10:57:01.0000] <annevk>
brucel, also, <video> does not have a type attribute

[10:58:00.0000] <brucel>
annevk whoops, will whisk that away and see how it goes

[10:59:00.0000] <jgraham>
lazni: Well yeah it would be nice if the site fell back to something sensible. I have to admit that it looks pretty cool though

[10:59:01.0000] <annevk>
(if it then works that would also be a bug, but I don't think it'll help)

[10:59:02.0000] <annevk>
brucel, if you add autoplay does it then fire maybe?

[11:01:00.0000] <annevk>
brucel, you can maybe hack around it by checking the associated error object on some kind of timeout loop

[11:01:01.0000] <annevk>
maybe they do properly update that

[11:03:00.0000] <brucel>
annevk didn't work when I removed type from video element and added autoplay

[11:03:01.0000] <brucel>
annevk hack, thanks, will try. But this *should* work, right?

[11:08:00.0000] <annevk>
yeah

[11:08:01.0000] <annevk>
you might wanna file a bug: https://bugs.webkit.org/

[11:14:00.0000] <brucel>
annevk thx

[12:41:00.0000] <mr_danie1>
everyon is so excited about html5, its great new features; also many times html5 is seen as a flash-killer, making hopefully anothe propritery technology obsolete, and in many aspect I think html5 will make the race with widely accepted standards

[12:42:00.0000] <mr_danie1>
but what is about p2p capabilites? for example chatroulette.com is completely build on flash, because it uses the p2p technology from flash to make direct end-user communication possible

[12:43:00.0000] <annevk>
it's work in progress

[12:43:01.0000] <mr_danie1>
so I am asking myself: is there a way to build p2p webapplications without flash? 

[12:43:02.0000] <mr_danie1>
is there a working draft document available annevk ?

[12:44:00.0000] <annevk>
http://www.whatwg.org/specs/web-apps/current-work/multipage/commands.html#peer-to-peer-connections

[13:19:00.0000] <zcorpan>
hmm, why doesn't the first video autoplay in opera? http://people.opera.com/brucel/dev/html5-video-fallback.html

[13:23:00.0000] <zcorpan>
annevk: "<annevk> brucel, fwiw, it should be dispatched on the video element" - no, it should fire on <source>

[13:25:00.0000] <zcorpan>
(at least when using <source>)

[13:32:00.0000] <zcorpan>
/me sees a your momma joke about html5 on twitter

[13:39:00.0000] <erlehmann>
this onerror thing is nice

[13:39:01.0000] <erlehmann>
zcorpan, link ?

[13:39:02.0000] <zcorpan>
http://twitter.com/iamkory/statuses/11051962698

[13:44:00.0000] <annevk>
zcorpan, ah yeah, I see now, but it'll bubble anyway so that shouldn't matter

[13:45:00.0000] <gsnedders>
/me opens up laptop and finds a listing of x86 jump instructions

[13:45:01.0000] <annevk>
the event summary makes this somewhat confusing

[13:45:02.0000] <gsnedders>
Oh dear.

[13:46:00.0000] <zcorpan>
annevk: error doesn't bubble

[13:47:00.0000] <annevk>
meh

[13:47:01.0000] <annevk>
seems cumbersome to put onerror on each source element

[13:47:02.0000] <annevk>
oh well

[13:47:03.0000] <zcorpan>
why would you put it on each source element?

[13:48:00.0000] <annevk>
debugging :)

[13:48:01.0000] <zcorpan>
you could use a capturing listener for debugging purposes :)

[13:50:00.0000] <annevk>
if i put the script in the right place, yeah...

[13:51:00.0000] <Hixie>
zcorpan: i think a "your momma so fat" joke would work better with the html spec :-)

[13:51:01.0000] <Hixie>
"your momma so fat, she makes the html spec look small"

[13:52:00.0000] <Dashiva>
More like "she got the HTML5 spec as a tattoo"

[13:52:01.0000] <zcorpan>
lol

[13:53:00.0000] <Hixie>
lol

[13:54:00.0000] <annevk>
hehe

[14:09:00.0000] <mr_danie1>
back, thanks for the link annevk 

[14:20:00.0000] <zcorpan>
content-aware fill looks nice

[14:25:00.0000] <paul_irish>
apparently its been available in GIMP for years as a plugin. looks almost better, actually: http://o3.tumblr.com/post/470608946/photoshops-caf-content-aware-fill-unbelievable

[14:26:00.0000] <Hixie>
"almost better" so "not quite as good"? :-)

[14:27:00.0000] <paul_irish>
heh. i think it's a little nicer, in fact. I perfered the resynthesizer's desert landscape job vs adobe's.

[14:29:00.0000] <zcorpan>
/me wasn't aware of GIMP-resynthesize

[14:29:01.0000] <zcorpan>
paul_irish: yes, the desert looks better

[14:30:00.0000] <zcorpan>
the sky in the first picture looks worse, but might be easily fixable

[15:21:00.0000] <Dashiva>
So RDFa 1.1 keeps xmlns unchanged, but also adds two additional ways to declare prefixes

[15:22:00.0000] <Dashiva>
And this is supposed to make it less complicated?

[15:24:00.0000] <Philip`>
Perhaps you could then define a profile of RDFa 1.1 that only includes one or two of the prefix declaration methods, to keep things simple

[15:26:00.0000] <gsnedders>
Anyone have any clue about ARM based computers? Like, what is decent and not overly expensive?

[15:32:00.0000] <jgraham>
gsnedders: Huh? Are there many ARM-based general purpose computers for sale?

[15:32:01.0000] <jgraham>
(or: what is your use case?)

[15:33:00.0000] <gsnedders>
jgraham: There are a few

[15:33:01.0000] <Philip`>
Does an iPhone count?

[16:08:00.0000] <Bolkonskij>
hi

[16:52:00.0000] <zcorpan>
reading the opera desktop team comments, it seems people don't understand what RC means

[16:57:00.0000] <gsnedders>
zcorpan: or "pre-pre-alpha" for that matter.


2010-03-26
[17:15:00.0000] <zcorpan>
/me wonders why http://hacks.mozilla.org/2010/02/firefox-46-features/ has stray </p>s in there

[18:55:00.0000] <gsnedders>
http://www.python.org/dev/peps/pep-0372/

[18:55:01.0000] <gsnedders>
html5lib mentioned in a PEP

[18:55:02.0000] <gsnedders>
(it's also mentioned in the Unladen Swallow PEP)

[18:55:03.0000] <gsnedders>
http://www.python.org/dev/peps/pep-3146/

[23:16:00.0000] <JonathanNeal>
It's really weird when Opera, Safari, and Firefox need a style, but Internet Explorer and Chrome do not.

[23:27:00.0000] <JonathanNeal>
Any folks around?

[23:28:00.0000] <Peter`>
that's just because of UA stylesheets

[23:28:01.0000] <Peter`>
also the reason reset-css files exist

[23:33:00.0000] <JonathanNeal>
Peter`, yes I don't want to spam this board, but that's why I'm labeling all the entries in my html5 reset.

[23:33:01.0000] <JonathanNeal>
I should also label when I'm setting the style to match ua and when i'm setting the style to match spec.

[00:10:00.0000] <JonathanNeal>
I'm sure it's a slippery slope, but there are some things I wish the spec didn't leave out and up to the ua.

[00:15:00.0000] <Peter`>
I concur, but I doubt it's going to change anytime soon

[00:20:00.0000] <JonathanNeal>
I didn't realize that type="a" and type="A" didn't work.

[00:22:00.0000] <JonathanNeal>
They work until you specify them in css ol[type="a"] {} ol[type="A"] {} both get eval'd.

[00:22:01.0000] <JonathanNeal>
for either a or A

[00:23:00.0000] <JonathanNeal>
Wow, in Chrome it even says the UA style is for ol[type=a] but when you style ol[type=A] it overwrites it anyway.

[00:24:00.0000] <JonathanNeal>
So ol[type=a] and ol[type=A] are different selectors within the UA's CSS, but the same outside.

[00:48:00.0000] <hsivonen>
JonathanNeal: I believe the dual selectors in the UA style sheet are there because the UA style sheet gets parsed in the case-sensitive mode to make it work right with XHTML

[02:42:00.0000] <annevk>
hsivonen, do you see validators as conversion help?

[02:42:01.0000] <annevk>
hsivonen, 'cause personally I would only start using the validator after I have made some attempt at following the rules

[02:46:00.0000] <hsivonen>
annevk: some people have already made fairly reasonable attempts to follow the previous spec

[02:47:00.0000] <hsivonen>
correction

[02:47:01.0000] <hsivonen>
the validator for the previous spec

[02:47:02.0000] <hsivonen>
which is Transitional

[02:47:03.0000] <hsivonen>
maybe I should have clarified that I'd like to loosen things based on usage data and that I wouldn't want to permit <font face> or <font size>

[02:48:00.0000] <hsivonen>
basically, I think the individual pre-existing features should be measured against usage on near-valid Transitional sites

[02:48:01.0000] <hsivonen>
and against Hixie's stated criteria

[02:48:02.0000] <annevk>
Can't you just group them all into one?

[02:49:00.0000] <hsivonen>
I think not

[02:49:01.0000] <annevk>
I think Hixie suggested that already

[02:49:02.0000] <annevk>
I guess schema-based validation does not give you the right messages for that...

[03:07:00.0000] <jgraham>
It never faiuls to amaze me how much people will pay to listen to some guy from the internet talk about stuff

[03:08:00.0000] <jgraham>
400 pounds for a one day workshop to learn about part of HTML 5 (or 200 if you're a student) is pretty insane

[03:09:00.0000] <jgraham>
/me hopes there are no students who can actually afford that

[03:09:01.0000] <Philip`>
Depends on the value of your time and how long it would take you to learn the same amount without that guidance

[03:09:02.0000] <MikeSmith>
/me wonders who the particular speaker is

[03:10:00.0000] <annevk>
Jeremy Keith I guess

[03:10:01.0000] <MikeSmith>
ah

[03:10:02.0000] <MikeSmith>
well, Jeremy's a great speaker

[03:10:03.0000] <annevk>
to his credit they organize a fairly cheap conference themselves in England

[03:11:00.0000] <jgraham>
I'm not disputing that he's a good speaker

[03:11:01.0000] <jgraham>
I'm just amazed that there are people who can afford it

[03:11:02.0000] <annevk>
I'd guess companies pay the bill

[03:11:03.0000] <jgraham>
Well sure. 

[04:09:00.0000] <zcorpan>
JonathanNeal: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9102

[04:27:00.0000] <annevk>
we should really remove those silly Selector rules

[04:27:01.0000] <annevk>
but I guess it is too late or something :/

[04:30:00.0000] <zcorpan>
hmm, i wonder if the spec is broken for <input type="COLOR"/> etc in xhtml

[04:31:00.0000] <annevk>
broken how?

[04:32:00.0000] <zcorpan>
the attribute value is case-insensitive in xhtml

[04:32:01.0000] <zcorpan>
but the selector matches case-sensitively

[04:33:00.0000] <annevk>
doesn't seem like a problem to me

[04:34:00.0000] <annevk>
well, maybe a problem with Selectors

[04:34:01.0000] <zcorpan>
it's a problem because writing valid xhtml would result in broken rendering

[04:34:02.0000] <zcorpan>
if the spec is implemented literally

[04:34:03.0000] <annevk>
oh, the spec

[04:35:00.0000] <annevk>
sorry, I see what you mean now

[04:35:01.0000] <annevk>
yeah, that seems broken

[04:35:02.0000] <zcorpan>
/me files a bug

[04:38:00.0000] <annevk>
i initially thought you meant for authors trying to style such inputs

[05:17:00.0000] <jgraham>
hsivonen: IO think it may be worth explaining why you use OpenOffice to create web pages if it involves a significant amount of manual post-processing work. At the moment it sounds a bit like the start of a "Doctor Doctor, it hurts when I do this" joke

[05:18:00.0000] <hsivonen>
jgraham: it lets me do paragraphs, headings and tables

[05:18:01.0000] <hsivonen>
jgraham: and Nvu wasn't better at that

[05:19:00.0000] <hsivonen>
jgraham: and getting a Dreamweaver license would be an overkill

[05:19:01.0000] <jgraham>
Tables I can somewhat understand, but it seems like for paragraphs + headigns you would be better off with a text editor

[05:19:02.0000] <hsivonen>
jgraham: I disagree

[05:19:03.0000] <hsivonen>
press return vs. typing all those tags

[05:20:00.0000] <jgraham>
Including the effort of later post-processing the output?

[05:20:01.0000] <hsivonen>
yes

[05:20:02.0000] <jgraham>
I'm pretty surprised

[05:20:03.0000] <jgraham>
(I guess tables also need post processing since they will not have proper headers or anything)

[05:21:00.0000] <annevk>
/me usually edits by hand

[05:22:00.0000] <hsivonen>
jgraham: tables from OOo do have <th>s 

[05:22:01.0000] <hsivonen>
the problem with tables in that OOo wraps the cell contents in <p>

[05:22:02.0000] <hsivonen>
and puts style and align all over the place

[05:22:03.0000] <jgraham>
In the right places (and in how many scenarios)? 

[05:23:00.0000] <hsivonen>
jgraham: I'm using OO.o Writer/Web. Are you talking about taking an OO.o Writer doc and exporting to HTML?

[05:23:01.0000] <hsivonen>
the "/Web" part is significant

[05:23:02.0000] <hsivonen>
uses HTML as the native format

[05:24:00.0000] <hsivonen>
and exposes semantic elements is the style palette

[05:24:01.0000] <annevk>
ooh great, I'm losing an hour sleep this Sunday?

[05:24:02.0000] <jgraham>
hsivonen: Ah. My OpenOffice install doesn't include anything that obviously looks like that

[05:24:03.0000] <MikeSmith>
hsivonen: so I pulled latest htmlparser source and updated my workspace and now trying to build... does ecj have the same command-line options as javac? can I just alias javac to ecj?

[05:24:04.0000] <annevk>
this trip to SFO is going to be sweet :/

[05:24:05.0000] <hsivonen>
MikeSmith: no idea. I'm working on fixing that right now (as soon as I stop typing on IRC)

[05:25:00.0000] <jgraham>
Oh, curious. There is no top level menu for it, but you can create a new HTML document which changes the application title

[05:25:01.0000] <jgraham>
s/top level menu/application menu item/

[05:26:00.0000] <jgraham>
The badness of the OO.org UI never fails to amaze me

[05:26:01.0000] <annevk>
creating a good editor is apparently a pretty hard problem

[05:26:02.0000] <hsivonen>
also, never ever open an existing HTML doc into OO.o

[05:27:00.0000] <annevk>
/me would love a good robust web-based editor

[05:27:01.0000] <hsivonen>
OO.o is write-only

[05:27:02.0000] <jgraham>
Totally. That doesn't excuse OO.org for having a lame UI in every respect, not just for creating clean HTML

[05:30:00.0000] <hsivonen>
Mozilla has made much better recovery progress from having a COM clone, an app-specific crossplatform UI toolkit, an app-specific cross-platform 2D graphics layer and a horrible GUI (Netscape 6 Preview Release) than OO.o

[05:32:00.0000] <jgraham>
OO.o has never recovered from the policy of making a pixel-by-pixel clone of Office 95

[05:32:01.0000] <jgraham>
(at least I assume it was a policy. It would be remarkable if it happened by accident)

[05:39:00.0000] <hsivonen>
MikeSmith: fix pushed to hg

[05:46:00.0000] <MikeSmith>
hsivonen: thanks

[05:47:00.0000] <hsivonen>
/me goes back to making the C++ version of that file compile nicely with GCC and MSVC

[05:47:01.0000] <hsivonen>
apperently, this area is too tough for 3 compilers

[06:15:00.0000] <MikeSmith>
hsivonen: btw, when I run jing from the command line in Terminal on OSX, I get "?" characters instead of curly quotes

[06:15:01.0000] <hsivonen>
MikeSmith: you need to set the terminal to UTF-8

[06:15:02.0000] <MikeSmith>
e.g., rror: Element ?base? from namespace ?http://www.w3.org/1999/xhtml? is missing one or more of the following attributes: [href, target]

[06:15:03.0000] <hsivonen>
MikeSmith: or if your terminal is in UTF-8, you need to set the System.out to UTF-8

[06:16:00.0000] <MikeSmith>
ah, OK

[06:16:01.0000] <hsivonen>
there's a magic java system property for that

[06:16:02.0000] <hsivonen>
it's very sad if Apple still keeps Java defaulting to MacRoman

[06:17:00.0000] <hsivonen>
the system property is file.encoding

[06:17:01.0000] <hsivonen>
-Dfile.encoding=UTF-8 should do the trick

[06:17:02.0000] <MikeSmith>
hsivonen: OK, trying that now

[06:18:00.0000] <MikeSmith>
indeed, yeah, that works

[06:18:01.0000] <MikeSmith>
thanks

[06:27:00.0000] <jgraham>
/me wonders if Sam actually read the IRC logs

[07:11:00.0000] <MikeSmith>
hsivonen: I'm trying to figure out where the w:string datatype is defined

[07:12:00.0000] <hsivonen>
MikeSmith: syntax/relaxng/datatype/java/src/org/whattf/datatype/AsciiCaseInsensitiveString.java

[07:13:00.0000] <MikeSmith>
ah

[07:13:01.0000] <MikeSmith>
thanks

[08:29:00.0000] <Philip`>
"I see value in being able to express notions such as "this page intentionally violates rule number 34"."

[08:29:01.0000] <Philip`>
I don't see how it's possible to violate http://tvtropes.org/pmwiki/pmwiki.php/Main/RuleThirtyFour

[08:32:00.0000] <Dashiva>
Perhaps by somehow providing content that cannot be made pornographic

[08:34:00.0000] <Philip`>
I don't see how it's possible

[08:37:00.0000] <Dashiva>
That's only because you lack faith

[09:20:00.0000] <JonathanNeal>
Thanks for the heads up on case insensative selectors, zcorpan.

[09:27:00.0000] <MikeSmith>
/me wonders what client do people use for jabber.ietf.org meetings

[09:27:01.0000] <annevk>
Empathy

[09:27:02.0000] <annevk>
it ships with Ubuntu

[09:28:00.0000] <annevk>
and does pretty much all messaging

[09:29:00.0000] <jgraham>
/me doesn't go to ietf meetings but uses bitlbee+irssi for jabber (donno if that has all the needed features for that usecase though)

[09:29:01.0000] <annevk>
somewhat tempted to install the new Ubuntu beta

[09:29:02.0000] <MikeSmith>
annevk: thanks

[09:29:03.0000] <MikeSmith>
wonder if there's anything that works on OSX

[09:29:04.0000] <MikeSmith>
jgraham: it uses some extension, XEP-0045

[09:30:00.0000] <annevk>
MikeSmith, http://www.pidgin.im/download/mac/ prolly works

[09:30:01.0000] <MikeSmith>
thanks

[09:30:02.0000] <annevk>
/me was using pidgin before

[09:40:00.0000] <lazni>
adiumx uses the pidgin lib

[09:42:00.0000] <annevk>
this is quite boring

[09:42:01.0000] <franksalim>
?

[09:42:02.0000] <annevk>
iri meeting

[09:43:00.0000] <franksalim>
i believe that

[09:43:01.0000] <jgraham>
Not just anne dissing the channel

[09:43:02.0000] <annevk>
mumbling on Unicode, UTF-8, octet sequences

[09:44:00.0000] <annevk>
just some meta discussion on the presentation of a IRI vs a IRI

[09:52:00.0000] <Dashiva>
IRI vs IRI?

[09:53:00.0000] <annevk>
dunno, we're now at LEIRI and HTML5 URL...

[09:53:01.0000] <annevk>
but it's going very slow

[10:01:00.0000] <Philip`>
The concept of competition between validators would be more convincing if more than one validator existed

[10:02:00.0000] <AryehGregor>
/me concurs.

[10:02:01.0000] <AryehGregor>
/me also isn't very inspired by how few web developers seem to use non-officially-endorsed tools like jslint, compared to Standards-Based Validators

[10:03:00.0000] <Philip`>
Also it would be more convincing if anybody used the HTML4 validators that compete with validator.w3.org

[10:03:01.0000] <Philip`>
(but as far as I can tell, everyone uses the latter)

[10:04:00.0000] <annevk>
i've seen people use validator.nu

[10:04:01.0000] <Philip`>
I tried using jslint once but then I realised it was a waste of time

[10:04:02.0000] <Dashiva>
^

[10:05:00.0000] <Philip`>
and I was making my code uglier and less readable just to satisfy its rule for semicolons and for not using the same name with var twice in one function

[10:05:01.0000] <Philip`>
s/rule/rules/

[10:07:00.0000] <AryehGregor>
Also, for compiled languages, where it's common to use warnings, everyone just uses their compiler's built-in warnings, so there's little to no competition on that score per se.

[10:08:00.0000] <AryehGregor>
(although I guess quality of warnings might be a contributing factor in the decision on which compiler to use, if there are multiple realistic choices for your platform)

[10:15:00.0000] <JonathanNeal>
Are there any other vendor stylesheets out there?  Like, has someone compiled a stylesheet of how one browser does everything?

[10:16:00.0000] <AryehGregor>
Every browser has its own UA stylesheets.

[10:17:00.0000] <Philip`>
I thought MikeSmith collected some for his guide thingy

[10:17:01.0000] <JonathanNeal>
I would love to get ahold of those.  I realized it might be a much more effective way to write mine.

[10:17:02.0000] <JonathanNeal>
I've been doing it through a process of laying out a pangram of every element in html5, then applying the styles from the html5 rendering guide.

[10:18:00.0000] <MikeSmith>
the h:tml doc uses the Webkit default UA stylesheet

[10:18:01.0000] <JonathanNeal>
Then applying additional styles in a vendor css file to catch whatever else was needed to normalize them.

[10:18:02.0000] <MikeSmith>
http://svn.webkit.org/repository/webkit/trunk/WebCore/css/html.css

[10:23:00.0000] <erlehmann>
the pecurialities are interesting. epiphany-webkit for example, does not have a default styling for <i>

[10:24:00.0000] <JonathanNeal>
Thanks MikeSmith, 0__qem sure is a strange value.

[10:26:00.0000] <MikeSmith>
JonathanNeal: I think that just gets interpreted as 0em

[10:27:00.0000] <MikeSmith>
I think for the build of the h:tml doc, I have it stripping out the __q part

[10:27:01.0000] <MikeSmith>
I don't remembering ever hearing what purpose the __q part serves

[10:28:00.0000] <MikeSmith>
yeah

[10:29:00.0000] <MikeSmith>
http://dev.w3.org/html5/markup/tools/css2xml

[10:29:01.0000] <MikeSmith>
(if you care to see all the ugliness)

[10:30:00.0000] <JonathanNeal>
I do care, I'm very interested in seeing how uas style and how people are using this info

[10:34:00.0000] <TabAtkins>
JonathanNeal: Poke around in your Firefox install to find it's stylesheet.  I've modified mine for a few things.  Relatively easy to get to.

[10:34:01.0000] <JonathanNeal>
Awesome!

[10:34:02.0000] <JonathanNeal>
Now just to get IE's

[10:38:00.0000] <MikeSmith>
annevk: is that Yngve speaking?

[10:46:00.0000] <annevk>
MikeSmith, I dropped out

[10:46:01.0000] <annevk>
MikeSmith, figured I should get some food

[10:48:00.0000] <AryehGregor>
http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/latest-trunk/ "This page is in Irish.  Would you like to translate it? [Translate] [Nope]"

[10:48:01.0000] <annevk>
I got bored with it too. All the high-level theoretical thought trains. How about starting with where contemporary implementations are?

[10:50:00.0000] <JonathanNeal>
Interesting how Firefox tightens up their css file (e.g. "margin: 1em 0") but then does "0px" and "margin: 0.5em auto 0.5em auto;" in other places.  It's nice to know these stylesheets are written by humans.

[10:52:00.0000] <Philip`>
JonathanNeal: Or it's worrying to know these stylesheets were written by robots that can pass the Turing test

[10:53:00.0000] <JonathanNeal>
Philip`, true dat, I am C.

[10:54:00.0000] <JonathanNeal>
If we find IE's, my assumption is that it would pass with flying colors.

[10:55:00.0000] <MikeSmith>
annevk: Larry is mentioning use of the term "URL" now

[10:56:00.0000] <annevk>
oh, maybe I should dial back in

[11:00:00.0000] <Philip`>
Does IE actually use a default stylesheet that's a real CSS file, rather than something more ad-hoc and legacyish?

[11:04:00.0000] <MikeSmith>
I think even for Mozilla and Webkit, the default styling behavior (logic?) is not totally captured in their UA stylesheets

[11:04:01.0000] <AryehGregor>
Philip`, I would guess so for IE >= 8.

[11:05:00.0000] <AryehGregor>
But yeah, some things aren't specifiable in CSS.

[11:05:01.0000] <JonathanNeal>
AryehGregor, we could write one if they don't already have one.

[11:05:02.0000] <AryehGregor>
They also all use vendor-specific extensions in their UA stylesheets, I assume.

[11:05:03.0000] <AryehGregor>
For anyone who hasn't seen it yet, this is interesting: http://themaninblue.com/writing/perspective/2010/03/22/

[11:05:04.0000] <MikeSmith>
AryehGregor: some things are, but they don't specify them in the UA stylesheet for some reason

[11:06:00.0000] <MikeSmith>
e.g., neither Gecko and Webkit UA stylesheets set bold on the th element

[11:06:01.0000] <MikeSmith>
they do it in program code instead

[11:06:02.0000] <AryehGregor>
Really?  Why?

[11:06:03.0000] <MikeSmith>
AryehGregor: dunno

[11:07:00.0000] <MikeSmith>
there may not be many (or any) more instances of that then the th case

[11:07:01.0000] <MikeSmith>
but I know of the th case, at least, so I suspect there might be others

[11:07:02.0000] <annevk>
how long does this meeting go on for?

[11:08:00.0000] <MikeSmith>
I think it was supposed to be 2.5 hours

[11:08:01.0000] <JonathanNeal>
I'm cleaning up the stylesheets and placing them @ http://sandbox.thewikies.com/html5-browser-stylesheets/

[11:09:00.0000] <annevk>
MikeSmith, kk

[11:09:01.0000] <JonathanNeal>
I wasn't sure what the version was on that webkit stylesheet, it was from their trunk, so I was going to add the version numbers once I know.

[11:09:02.0000] <annevk>
now they're talking about vests vs t-shirts

[11:09:03.0000] <annevk>
clearly ninjas win

[11:10:00.0000] <MikeSmith>
JonathanNeal: note that those stylesheets get updated/changed somewhat regularly

[11:10:01.0000] <MikeSmith>
so if you make copies of them, you are going to need to track the upstream changes

[11:12:00.0000] <JonathanNeal>
MikeSmith, definitely the reason why I want to grab the versions.  Yea, I can always tag it 56629

[11:19:00.0000] <JonathanNeal>
Tagged them as firefox-v3.6.2.css and webkit-r56629.css

[11:29:00.0000] <erlehmann>
JonathanNeal, better tag it gecko-1.9.xyz

[11:29:01.0000] <erlehmann>
after all its the rendering engines that change this

[11:29:02.0000] <JonathanNeal>
how would i verify whcih gecko it was?

[11:30:00.0000] <erlehmann>
doesn't ff report this ?

[11:30:01.0000] <JonathanNeal>
btw, from what I understand MS IE saves their stylesheet in an xml file.

[11:30:02.0000] <erlehmann>
check the about dialog.

[11:31:00.0000] <JonathanNeal>
1.9.2.2 for gecko erlehmann?

[11:31:01.0000] <erlehmann>
XML stylesheets, what a funny idea :D

[11:31:02.0000] <erlehmann>
JonathanNeal, dunno. but i have iceweasel 3.5.8, which has 1.9.1.6, so it seems plausible

[11:34:00.0000] <JonathanNeal>
I found IE6's stylesheet!

[11:36:00.0000] <Philip`>
Where?

[11:36:01.0000] <JonathanNeal>
Or I'm really dang close.

[11:36:02.0000] <JonathanNeal>
res://msxml.dll/defaultss.xsl

[11:38:00.0000] <Philip`>
That sounds like it's a stylesheet for XML files, not for HTML

[11:38:01.0000] <Philip`>
given that it's in msxml.dll

[11:42:00.0000] <JonathanNeal>
Yea, just need to figure out what the dll is for html

[11:42:01.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/defaultss.xls.txt

[11:42:02.0000] <JonathanNeal>
that's the one for xml, at least.

[11:47:00.0000] <Philip`>
HTML is lots of DLLs

[11:47:01.0000] <Philip`>
like mshtml.dll

[11:51:00.0000] <JonathanNeal>
Yea ... it sure feels close though.

[11:54:00.0000] <JonathanNeal>
According to http://msdn.microsoft.com/en-us/library/aa741312(VS.85).aspx it might be in mshtml.dll

[11:55:00.0000] <Philip`>
There are no strings in that file that look like CSS to me

[11:56:00.0000] <AryehGregor>
Maybe they're ROT13ed.

[11:56:01.0000] <AryehGregor>
Or base-64.

[11:56:02.0000] <jgraham>
Philip`: I'm pretty sure that most people that use jsLint turn of many of the more opinionated warnings

[11:56:03.0000] <JonathanNeal>
Maybe they're in the registry?

[11:56:04.0000] <jgraham>
and I don't think it's really comparable to HTML since javascript doesn't have any conformance criteria

[11:57:00.0000] <AryehGregor>
Some people want that for HTML too, apparently.  Or at least are open to considering it.

[11:58:00.0000] <jgraham>
(I guess theoretically one might like a porgram that told you if you were going to hit any features that change behaviour in ES5 strict mode. But that probably isn't statically determinable)

[11:58:01.0000] <jgraham>
AryehGregor: Yes some people want it. That doesn't mean it's a good idea

[12:11:00.0000] <AryehGregor>
Do all browsers support oninput?

[12:12:00.0000] <AryehGregor>
Meh, never mind.

[12:12:01.0000] <AryehGregor>
Doesn't matter.

[12:14:00.0000] <JonathanNeal>
I found a hacky way to grab 'em all.

[12:47:00.0000] <annevk>
i cannot access google.com

[12:47:01.0000] <annevk>
weird

[12:51:00.0000] <Dashiva>
Can you access this? http://www.schneier.com/blog/archives/2010/03/schneier_bloggi.html

[12:52:00.0000] <annevk>
sure

[12:53:00.0000] <annevk>
or google.nl etc.

[12:53:01.0000] <annevk>
it's very weird, my mother called me this morning and she had exactly this problem

[12:54:00.0000] <annevk>
maybe restarting the browser will fix it...

[12:54:01.0000] <annevk>
yup...

[13:21:00.0000] <jgraham>
Cookies?

[13:22:00.0000] <annevk>
i accept cookies

[13:22:01.0000] <annevk>
visit my site for address details

[14:24:00.0000] <annevk>
hmm, I wonder if the way Chrome (does not) implement(s) form validation will cause us problems going forward

[14:24:01.0000] <annevk>
i.e. required="" is "supported" but does not actually prevent form submission unless some JavaScript is used

[14:25:00.0000] <annevk>
not cool

[14:27:00.0000] <Hixie>
annevk: um yeah that's very not cool

[14:27:01.0000] <Hixie>
annevk: can you file a bug pointing out the very likely intrerop implications of this?

[14:28:00.0000] <TabAtkins>
Damn, anyone know where the hell image-fit is specified? (In CSS)

[14:28:01.0000] <Philip`>
TabAtkins: http://dev.w3.org/csswg/css3-page/#propdef-image-fit ?

[14:29:00.0000] <TabAtkins>
Danke, Philip`.

[14:29:01.0000] <Philip`>
Thank Google

[14:29:02.0000] <TabAtkins>
...I should have thought of that.

[14:31:00.0000] <annevk>
Hixie, would this be a bug on WebKit or Chrome?

[14:31:01.0000] <Hixie>
webkit i guess

[14:36:00.0000] <annevk>
https://bugs.webkit.org/show_bug.cgi?id=36677

[14:36:01.0000] <Hixie>
thanks

[16:17:00.0000] <Traveler>
hi

[16:49:00.0000] <JonathanNeal>
I set up a script to capture IE8's default styles.

[16:49:01.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/ie8.css this is what it returned me.

[16:56:00.0000] <erlehmann>
JonathanNeal, next step: use the script to automate the capturing job for all browsers ? maybe ab-using browsershots.org and AJAX ?

[16:56:01.0000] <TabAtkins>
JonathanNeal: Very cool.  I like it.

[16:57:00.0000] <erlehmann>
font-family: Courier New; — i suppose thats monospace ?

[16:58:00.0000] <TabAtkins>
Yeah, likely.  It's a monospace font, at least, and I guess it's JonathanNeal's default.

[16:58:01.0000] <erlehmann>
i wonder if the „zoom: 1;“ often seen is a IE hack :D

[16:58:02.0000] <TabAtkins>
It triggers hasLayout!

[16:59:00.0000] <JonathanNeal>
yes, it triggers hasLayout.


2010-03-27
[17:00:00.0000] <erlehmann>
even IE devs use IE hacks !

[17:00:01.0000] <erlehmann>
what a wonderfu^W awful world :(

[17:00:02.0000] <JonathanNeal>
Well, it may not be a direct translation --- I can't find how they may be triggering it.

[17:01:00.0000] <JonathanNeal>
This is a first draft.

[17:01:01.0000] <JonathanNeal>
A lot of weird detection going on to get these.

[17:01:02.0000] <TabAtkins>
JonathanNeal: What is the script doing to gather this info?

[17:02:00.0000] <JonathanNeal>
One I wrote.  See why I don't trust it.

[17:02:01.0000] <TabAtkins>
Care to share?

[17:02:02.0000] <JonathanNeal>
http://msdn.microsoft.com/en-us/library/ms535231(VS.85).aspx --- translated this into a checker.

[17:03:00.0000] <TabAtkins>
How did you extract information out of that?  Compare the values with some suitable "default" values?

[17:03:01.0000] <JonathanNeal>
Exactly, if they claim that everything ever should retain these current values, then I don't need to assign them.

[17:04:00.0000] <TabAtkins>
Gotcha.

[17:04:01.0000] <JonathanNeal>
And other browsers follow this convention, for instance webkit or firefox almost never assign display: inline

[17:05:00.0000] <JonathanNeal>
One could assume there is a vendor * {} but when I individually checked all of the default values, they checked out.

[17:06:00.0000] <TabAtkins>
Kk.  So you could automate this by just doing some statistical analysis on the properties, and ignoring values that appear too often.

[17:06:01.0000] <erlehmann>
haha

[17:06:02.0000] <JonathanNeal>
By either appearing too often or being universally accepted in multiple browsers as the default.

[17:07:00.0000] <JonathanNeal>
like how all xml elements are generally display: inline.

[17:07:01.0000] <TabAtkins>
Yeah, too often across all elements in a given browser, or across all browsers for a given element.

[17:07:02.0000] <TabAtkins>
We should make a table for this.

[17:08:00.0000] <JonathanNeal>
Sure, I'll brb in about 10 - 15 minutes, cool?

[17:10:00.0000] <TabAtkins>
Yeah.

[17:11:00.0000] <TabAtkins>
/me could work on it this weekend.

[17:11:01.0000] <TabAtkins>
/me is already tipsy now at the TGIF meeting.

[17:56:00.0000] <TabAtkins>
Hm.  What *is* the behavior of a CSS background image with only a defined width, or only a defined height?

[17:56:01.0000] <TabAtkins>
(Such an image could be produced through SVG, frex.)

[17:57:00.0000] <TabAtkins>
/me doesn't know enough SVG to actually define such an image and test it.

[19:51:00.0000] <othermaciej>
hola

[19:51:01.0000] <othermaciej>
TabAtkins: I think CSS now defines what the intrinsic size  is in "weird" cases like that

[20:47:00.0000] <Traveler1>
hi

[20:49:00.0000] <Traveler1>
anyone around?

[20:49:01.0000] <Traveler1>
hi TabAtkins

[20:52:00.0000] <TabAtkins>
yo

[20:52:01.0000] <TabAtkins>
othermaciej: Any idea where it might do so?  I was pretty sure it did too, but I can't find it at the moment.

[20:53:00.0000] <Traveler1>
What are your thoughts on multiple H1 tags on a page?

[20:54:00.0000] <TabAtkins>
Multiple top-level headings, or just using multiple <h1>s and letting them take implicit levels based on sectioning elements?

[20:54:01.0000] <TabAtkins>
(That is, doing <h1>foo</h1><section><h1>bar</h1></section> rather than <h1>foo</h1><h2>bar</h2>.)

[20:54:02.0000] <Traveler1>
Well if you're site calls for it

[20:55:00.0000] <Traveler1>
A lot of people say that one should use only once the h1 tag. 

[20:55:01.0000] <Traveler1>
Hm no

[20:56:00.0000] <Traveler1>
For example on the homepage of a blog. You will find a few posts(3,4) 

[20:56:01.0000] <TabAtkins>
Most of the time you'll have a single definite "top level" heading, and so a single <h1> is appropriate to mark that up, with <h2> or lower for other stuff.

[20:56:02.0000] <Traveler1>
The titles are h1's.

[20:56:03.0000] <Traveler1>
The subtitles(if any) are h2. and so on so forth.

[20:56:04.0000] <Traveler1>
Would you say that's a bad practice?

[20:57:00.0000] <TabAtkins>
No, I think that's just fine.  However, those <h1>s likely aren't actually the top-level heading of the page (probably the blog title is), so make sure that they are wrapped in something that properly scopes them, like <article> or <section>.

[20:57:01.0000] <TabAtkins>
That way they won't be registered as being the same level of heading as the blog title is.

[20:58:00.0000] <Traveler1>
so <article><h1>title</h1></article> ?

[20:59:00.0000] <Traveler1>
The blog title is just the title. I don't get why people wrap it in h1. 

[20:59:01.0000] <Traveler1>
It has it's freaggin tag <title> 

[20:59:02.0000] <TabAtkins>
Well, like <body><h1>My blog!</h1><article><h1>Post about my cat</h1><p>My cat rocks.</p></article><article><h1>More cat info!</h1><p>Really.</p></article></body>

[20:59:03.0000] <Traveler1>
not like that.

[21:00:00.0000] <Traveler1>
what's that lonely h1 doing after the body tag?

[21:00:01.0000] <TabAtkins>
...? People wrap the title in an <h1> because it's the top-level heading, and because it's difficult to display the <title>.

[21:00:02.0000] <TabAtkins>
That's the blog title.

[21:00:03.0000] <TabAtkins>
(Thus the words inside of it.)

[21:02:00.0000] <Traveler1>
A heading element briefly describes the topic of the section it introduces. 

[21:02:01.0000] <Traveler1>
Wouldn't it be logically correct to use it as I described it above? 

[21:03:00.0000] <Traveler1>
and just skip the h1 tag for the title

[21:04:00.0000] <TabAtkins>
You could, sure, but then the page as a whole doesn't have a heading.

[21:04:01.0000] <JonathanNeal>
Wow

[21:04:02.0000] <JonathanNeal>
That was a long 15 minutes.

[21:04:03.0000] <TabAtkins>
Because the <h1> is the heading for the article.

[21:04:04.0000] <TabAtkins>
Haha, JonathanNeal.

[21:04:05.0000] <JonathanNeal>
Turns out it was one of my great friends birthdays

[21:04:06.0000] <Traveler1>
The page as a whole doesn't need a heading. It has a title :P

[21:04:07.0000] <Traveler1>
The articles need headings

[21:04:08.0000] <JonathanNeal>
and I couldn't not drop code for my friend.

[21:05:00.0000] <TabAtkins>
Traveler1: Need?  No, you're right, it doesn't *need* one.  But most people will put one, since they'd like to have a header on their site.  ^_^  And that's perfectly normal.

[21:05:01.0000] <JonathanNeal>
TabAtkins, I'm rewriting my tests that generate those ie stylesheets.

[21:05:02.0000] <TabAtkins>
JonathanNeal: Congrats to your friend!

[21:06:00.0000] <Traveler1>
That doesn't bother me at all. I think it bothers me a little when people use the word must. 

[21:07:00.0000] <Traveler1>
Wouldn't you be bothered? 

[21:08:00.0000] <miketaylr>
Traveler1: i appreciate a h1 in the body since i use chromium which doesn't really show the <title> so well

[21:08:01.0000] <miketaylr>
e.g., http://miketaylr.com/post/806aae69.png

[21:09:00.0000] <Traveler1>
you use chromium as your main os?

[21:09:01.0000] <Traveler1>
(I had to ask)

[21:10:00.0000] <miketaylr>
heh. no. but i use chromium as my main browser.

[21:11:00.0000] <JonathanNeal>
But now I'm rewriting those tests

[22:45:00.0000] <Bolkonskij>
Hm

[22:46:00.0000] <Bolkonskij>
I can't believe this. I have to uninstall SP3 in order to remove IE8.  

[22:47:00.0000] <Bolkonskij>
Give me a freakin break. 

[23:33:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/ie8.css --- woo woo I think these are the default styles for ie8

[23:51:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/ie7.css --- and ie7

[00:05:00.0000] <othermaciej>
zoinks, Hixie pinged out

[00:24:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/ie6.css --- and ie6

[02:02:00.0000] <JonathanNeal>
heyo

[05:23:00.0000] <annevk>
oh wow

[05:23:01.0000] <annevk>
http://www.amuz.be/noflash.html (via krijnh)

[05:24:00.0000] <annevk>
roughly translated "we also have a site for blind people. click here for the site for blind people" (and then some mumbling for when you do not have Flash)

[06:24:00.0000] <hsivonen>
the rationale for conformance criteria is good to have

[06:27:00.0000] <hsivonen>
I think file size is the weakest of the rationales, because type=text/css is allowed

[06:27:01.0000] <annevk>
via your reply to gruber I found http://twitter.com/gruber/status/11119748935

[06:27:02.0000] <annevk>
funny

[06:28:00.0000] <annevk>
hsivonen, type=text/css is marginal compared to a 80% reduction overall

[06:54:00.0000] <karlcow>
annevk: indeed it is not the right questions, but I'm not sure what was the intent of Sam asking the questions. Both given examples may create accessibility challenge, they are both presentational, the only difference is size.

[08:10:00.0000] <Philip`>
Is it normal to get multiple requests per second, for many minutes, for an .ogg file served by <audio> to a Firefox 3.6 user?

[08:46:00.0000] <JonathanNeal>
Mornin!

[08:55:00.0000] <TabAtkins>
Yo, JonathanNeal.

[08:56:00.0000] <JonathanNeal>
Did you see the 5 css?  ie6, ie7, ie8, ie6 (ietester), ie7 (ietester) ?

[08:56:01.0000] <JonathanNeal>
I don't know why, but ietester would produce slightly different results.

[08:59:00.0000] <TabAtkins>
No, I left the room before you posted that, I guess.

[09:00:00.0000] <TabAtkins>
I found that ie emulators do an imperfect job of reproducing ie's quirks.

[09:00:01.0000] <TabAtkins>
I had some real problems when a site that looked fine in some ie emulation was actually still very unusable in the real ie6.

[09:02:00.0000] <JonathanNeal>
They are all @ http://sandbox.thewikies.com/html5-browser-stylesheets/

[09:08:00.0000] <TabAtkins>
Cool.

[09:09:00.0000] <TabAtkins>
/me goes to turn this into a huge table.

[09:10:00.0000] <JonathanNeal>
You've mentioned that, how are you planning to display this data in a table?

[09:10:01.0000] <TabAtkins>
Complicatedly.

[09:11:00.0000] <TabAtkins>
It's 3-dimensional, unfortunately, which means some trickery to display it in a <table>.

[09:11:01.0000] <TabAtkins>
I'm going with properties for the rows, with column groups corresponding to each element, and browsers in each column group.

[09:15:00.0000] <Philip`>
If it's 3D, you should use WebGL

[09:16:00.0000] <TabAtkins>
I doubt that would make the table easier to understand.

[09:17:00.0000] <Philip`>
No, but it would look cooler

[09:18:00.0000] <TabAtkins>
But I couldn't program it in ten minutes.

[09:21:00.0000] <JonathanNeal>
TabAtkins, I could always program my test to output the results differently.

[09:21:01.0000] <JonathanNeal>
Give me a structure on pastie or something that you want it to follow.

[09:22:00.0000] <TabAtkins>
Okay.  Can you output it for all browsers simultaneously?

[09:22:01.0000] <JonathanNeal>
No, but I can easily concat all 3 or 6 tests.

[09:22:02.0000] <JonathanNeal>
And each test can be written to be a piece of that table.

[09:23:00.0000] <TabAtkins>
Can you gimme a json structure containing nested objects keyed to element name, then property name?

[09:24:00.0000] <TabAtkins>
Ie {"a":{"border-color":"#000",...}...}

[09:24:01.0000] <JonathanNeal>
Sure

[09:26:00.0000] <TabAtkins>
Damn, wrong use of ie.  Should have been eg.

[09:28:00.0000] <Bolkonskij>
Do you still support Ie 6?  

[09:31:00.0000] <TabAtkins>
I "support" it in the sense that I will make sure most of my site is at least visible in ie6.  I don't care beyond that.

[09:37:00.0000] <JonathanNeal>
TabAtkins, http://sandbox.thewikies.com/html5-browser-stylesheets/ie7.json.txt and http://sandbox.thewikies.com/html5-browser-stylesheets/ie8.json.txt ?

[09:37:01.0000] <TabAtkins>
Yes, that is perfect.

[09:37:02.0000] <TabAtkins>
Now I go to eat breakfast, but I'll finish this later today.

[09:40:00.0000] <JonathanNeal>
ie6.json.txt is in there too.

[10:24:00.0000] <Bolkonskij>
http://cnettv.cnet.com/apple-vs-google/9742-1_53-50085458.html 

[10:38:00.0000] <lazni>
idea: federated feed syncing: several ppl in different timezones (and online habit) share the workload of subscribing to feeds

[10:38:01.0000] <lazni>
may reach the same sync reliability of an always-on feed reader

[10:38:02.0000] <lazni>
bonus point if implemented as an HTML offline-cached app

[11:31:00.0000] <GPHemsley>
annevk: (Where's Hixie?) "Errors that are intended to save the author time" in Section "1.10.3 Restrictions on the content model and on attribute values" has a typo

[11:31:01.0000] <GPHemsley>
"inlien" should be "inline"

[11:33:00.0000] <Philip`>
GPHemsley: Already in Bugzilla

[11:33:01.0000] <GPHemsley>
ah, OK, good :)

[11:34:00.0000] <GPHemsley>
glad I didn't make it official then :)

[12:02:00.0000] <TabAtkins>
JonathanNeal: http://www.xanthir.com/etc/default-stylesheets.php

[12:03:00.0000] <JonathanNeal>
Interesting

[12:03:01.0000] <JonathanNeal>
It would be neat if you could consolidate ones that repeat

[12:04:00.0000] <JonathanNeal>
so instead of italic, italic, italic, it was italic colspan="3"

[12:04:01.0000] <TabAtkins>
Sure.

[12:04:02.0000] <Bolkonskij>
Heh

[12:04:03.0000] <JonathanNeal>
I like the rollover and stuff too, that's nice.

[12:05:00.0000] <Bolkonskij>
This router is weird. I leave the desk for 2 mins and it disconnects me. It always happens when I'm away. 

[12:07:00.0000] <Bolkonskij>
/me is sick of the support guys telling him that the problem will be fixed in the next couple of days

[12:16:00.0000] <TabAtkins>
JonathanNeal: Done.

[12:16:01.0000] <JonathanNeal>
Soweet

[12:18:00.0000] <JonathanNeal>
Neat chart

[12:18:01.0000] <JonathanNeal>
How easy is it to update if we improve the json?

[12:19:00.0000] <TabAtkins>
Just hand me new json, I paste it in.

[12:24:00.0000] <TabAtkins>
Adding new browsers is a 1-line change too.

[12:24:01.0000] <JonathanNeal>
cool.

[12:26:00.0000] <lazni>
is there any websocket implementation? (not neccessarily in browsers)

[12:26:01.0000] <TabAtkins>
Doesn't Chrome have some early implementation of it?

[12:26:02.0000] <TabAtkins>
/me isn't certain.

[12:27:00.0000] <lazni>
/me would like to write apps wholly in web languages, to enforce agpl

[12:28:00.0000] <lazni>
http://blog.chromium.org/2009/12/web-sockets-now-available-in-google.html great

[12:28:01.0000] <TabAtkins>
agpl?

[12:28:02.0000] <TabAtkins>
Oh, wait, nm.

[12:29:00.0000] <lazni>
those languages are inherently distributed with the corresponding source code

[12:29:01.0000] <TabAtkins>
Yeah.

[12:29:02.0000] <TabAtkins>
Though with web sockets, you'll be talking back to a server.  That source isn't automatically distributed.

[12:30:00.0000] <TabAtkins>
Ooh, apparently ie8 implements some of the page-break properties?

[12:30:01.0000] <lazni>
can't WS be used to connect to other browsers? oops

[12:32:00.0000] <TabAtkins>
JonathanNeal: ie7 caption padding.  Really?

[12:38:00.0000] <lazni>
http://code.google.com/p/mochiweb/issues/detail?id=37 WS in minefield too

[12:40:00.0000] <MikeSmith>
lazni: there's a patch for adding WS support to Gecko, but I don't think the code has landed yet

[12:41:00.0000] <lazni>
you would know better than I do, as the above bug report isn't backed up

[12:42:00.0000] <MikeSmith>
https://bugzilla.mozilla.org/show_bug.cgi?id=472529

[12:43:00.0000] <MikeSmith>
the patch is at https://bug472529.bugzilla.mozilla.org/attachment.cgi?id=429716

[12:51:00.0000] <lazni>
each node sends its host:port to a tracker server, which distributes them to other nodes, and nodes run node.js and WS between themselves

[12:52:00.0000] <lazni>
kinda like bittorrent, which would suggest DHT as the next step

[12:52:01.0000] <lazni>
then the tracker is AGPL, and doesn't log anything

[13:16:00.0000] <Philip`>
lazni: WebSockets can't connect to browsers

[13:17:00.0000] <Philip`>
You'd need some kind of P2P protocol maybe like http://www.whatwg.org/specs/web-apps/current-work/multipage/commands.html#peer-to-peer-connections

[13:17:01.0000] <Philip`>
though you'll probably need to proxy a lot of the traffic through a server just to get around NATs and firewalls

[13:20:00.0000] <daedb>
TabAtkins: That's a pretty cool table :)

[13:21:00.0000] <TabAtkins>
Thank JonathanNeal equally.

[13:21:01.0000] <daedb>
Yes, JonathanNeal too ^^

[13:22:00.0000] <daedb>
I think it would be easier to read if the elements were on the left side instead on top though, so it wouldn't require so much horizontal scrolling.

[13:23:00.0000] <JonathanNeal>
TabAtkins, about 300zzz, you have no idea how many times I reran that trying to figure out what was going wrong.

[13:23:01.0000] <JonathanNeal>
It's honestly what the browser reports.

[13:24:00.0000] <TabAtkins>
JonathanNeal: How crazy.

[13:25:00.0000] <TabAtkins>
daedb: Eh, you either have lots of horizontal or lots of vertical.

[13:25:01.0000] <JonathanNeal>
So if it is a bug, it's with what the browser reports not the script.

[13:25:02.0000] <daedb>
TabAtkins: Horizontal is less annoying imo

[13:26:00.0000] <TabAtkins>
Good!

[13:26:01.0000] <TabAtkins>
(since that's what I'm doing now)

[13:27:00.0000] <daedb>
err... vertical, I mean... typo :p

[13:27:01.0000] <TabAtkins>
^_^

[13:27:02.0000] <TabAtkins>
The table would, in general, be a lot bigger if I swapped the axes, because of the width of various bits of text.

[14:16:00.0000] <jgraham>
Anyone got any ideas what could cause lcov to fail with messages like geninfo: WARNING: gcov did not create any files for /path/to/.gcda/file?

[14:16:01.0000] <jgraham>
/me realises there is likely a better channel for this somewhere


2010-03-28
[17:43:00.0000] <othermaciej>
/me would appreciate input on his recent bugs from people who know more about document conformance

[17:51:00.0000] <AryehGregor>
othermaciej, they all look good to me.

[18:52:00.0000] <othermaciej>
like I don't know the deep reason <nobr> and <wbr> are nonconforming

[18:52:01.0000] <othermaciej>
maybe just because they never were conforming before?

[18:52:02.0000] <AryehGregor>
I would guess it's because, unlike <pre> and <br>, they were not in any previous spec.

[18:52:03.0000] <AryehGregor>
Yeah.

[18:53:00.0000] <Dashiva>
They don't feel entirely semantic to me. Line breaking (or not) is fairly deep into CSS territory.

[18:53:01.0000] <AryehGregor>
Then why doesn't that apply to <pre>?

[18:53:02.0000] <AryehGregor>
Or <br>?

[18:53:03.0000] <Dashiva>
Of course, <wbr> can't really be emulated

[18:53:04.0000] <AryehGregor>
Also, as Maciej points out, <wbr> can't be emulated -- yeah.

[18:54:00.0000] <AryehGregor>
I'm also planning to file a bug requesting that most or all presentational attributes/elements be made conforming (possibly obsolete but conforming so they throw an error).  Something like <table cellpadding=5> is difficult to reproduce exactly in inline CSS, and even more difficult to convert automatically.  And some, like cellspacing, have no direct CSS equivalent that works in (for example) IE6.

[18:54:01.0000] <Dashiva>
<pre> is different, because it says the whitespace is significant

[18:54:02.0000] <AryehGregor>
s/so they throw an error/so they raise a warning/

[18:54:03.0000] <AryehGregor>
Dashiva, <nobr> says line breaks are significant.

[18:55:00.0000] <AryehGregor>
Er.

[18:55:01.0000] <AryehGregor>
Or, no.

[18:55:02.0000] <AryehGregor>
What does it do?

[18:55:03.0000] <AryehGregor>
/me has never used it

[18:55:04.0000] <Dashiva>
It prevents line breaking

[18:55:05.0000] <Dashiva>
<pre> gives meaning to the whitespace within it. <nobr> is just a formatting guideline.

[18:55:06.0000] <AryehGregor>
Then what does Maciej mean by "<nobr> should be considered just as semantic as <pre> or <br>, with a semantic of 'a block of text with preformatted line breaks.'"?

[18:55:07.0000] <AryehGregor>
It's not preformatted line breaks, it just doesn't have any line breaks.

[18:56:00.0000] <AryehGregor>
That does seem like CSS territory to me.

[18:56:01.0000] <AryehGregor>
<wbr> less so, because it's hard to see how you'd even implement such a feature in CSS -- it inherently requires a particular point to be flagged in the markup itself.

[18:57:00.0000] <Dashiva>
It's surely angels and pins territory, but to me <pre> is a clear step away from <br>, <wbr>, <nobr>

[18:58:00.0000] <AryehGregor>
Really?  <br> and <wbr> seem like they have to be in HTML to me.  <nobr> I agree with you on, more or less.

[18:58:01.0000] <Dashiva>
They have to be, but only because we can't express it any other way

[18:58:02.0000] <AryehGregor>
Well, eh.

[18:58:03.0000] <othermaciej>
AryehGregor: Sam Ruby has a bug open specifically about some particular examples of presentational table markup

[18:59:00.0000] <AryehGregor>
<nobr> really means "there are no line breaks here".

[18:59:01.0000] <AryehGregor>
othermaciej, oh, really?

[18:59:02.0000] <Dashiva>
<wbr> is sort of like &shy;

[18:59:03.0000] <othermaciej>
let me see if I can find it

[18:59:04.0000] <AryehGregor>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=7468

[18:59:05.0000] <Dashiva>
(only there's no hyphen involved, obviously)

[19:00:00.0000] <othermaciej>
I guess I described <nobr> incorrectly in my bug

[19:00:01.0000] <AryehGregor>
"This use case is handled by <style scoped> in HTML5."

[19:00:02.0000] <AryehGregor>
That's actually a good point.

[19:00:03.0000] <othermaciej>
AryehGregor: that's the one

[19:00:04.0000] <AryehGregor>
Once that's supported, the table markup could be easily replaced.

[19:00:05.0000] <Dashiva>
Well, zzz time, *poof*

[19:01:00.0000] <othermaciej>
I'm not sure I buy that argument

[19:01:01.0000] <othermaciej>
about <style scoped>

[19:01:02.0000] <othermaciej>
I guess it is more compact than putting a style= attribute on every cell, but it's probably still less compact than the presentational attributes

[19:01:03.0000] <othermaciej>
and it's not clear to me that the accessibility / media independence argument applies in this particular case

[19:03:00.0000] <mpilgrim>
ooh, if we added <nobr> and <wbr> to html5, ppk's compatibility tables could finally validate

[19:03:01.0000] <AryehGregor>
I think the real argument in favor of pushing for the use of CSS is consistency.  There's never any actual difference between using presentational markup and equivalent inline CSS, but CSS is a much more powerful and general mechanism, and is the only styling mechanism to be supported going forward.  A language that didn't include presentational markup beyond inline style would be much simpler.

[19:03:02.0000] <othermaciej>
mpilgrim: he uses those?

[19:03:03.0000] <AryehGregor>
I don't know whether that's a good enough reason to make them invalid, though, rather than obsolete but conforming.

[19:04:00.0000] <daedb>
http://www.quirksmode.org/oddsandends/wbr.html

[19:04:01.0000] <othermaciej>
I think whitespace treatment is something that is on the fine edge between semantics and presentation

[19:05:00.0000] <othermaciej>
sometimes, a particular whitespace treatment is really part of the information in the document, not just a matter of how you present it

[19:11:00.0000] <AryehGregor>
Poetry and computer code are standard examples of that.

[19:11:01.0000] <AryehGregor>
Although <pre> usually suffices for both, if you suppress the monotype font for poetry.

[19:32:00.0000] <Traveler7>
hi

[19:33:00.0000] <Traveler7>
What's correct <article><section><h1>Title</h1><p>Some content here</p></section> | <section><h1>Another title</h1></section></article> or <section><article><h1>Title</h1><p>Content</p></article></section> and so on.

[19:38:00.0000] <JonathanNeal>
hola

[19:40:00.0000] <AryehGregor>
/me commented on bug 7468

[19:43:00.0000] <othermaciej>
why does the validator dislike this docype:

[19:43:01.0000] <othermaciej>
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/1999/REC-html401-19991224/strict.dtd"

[19:43:02.0000] <othermaciej>
>

[19:43:03.0000] <othermaciej>
is that the wrong URL for the DTD?

[19:51:00.0000] <AryehGregor>
Are you using html5.validator.nu or just validator.nu?

[19:51:01.0000] <AryehGregor>
I think only the former accepts obsolete-but-conforming doctypes, for some reason.

[20:02:00.0000] <Traveler1>
hi

[20:08:00.0000] <AryehGregor>
othermaciej et al.: I just filed a bug "Many, most, or all obsolete features should be conforming".  http://www.w3.org/Bugs/Public/show_bug.cgi?id=9355

[20:10:00.0000] <othermaciej>
AryehGregor: I would recommend filing bugs that are narrower in scope than that

[20:10:01.0000] <AryehGregor>
Really?  Why?

[20:10:02.0000] <AryehGregor>
I give clear reasoning that applies to *all* obsolete features.

[20:10:03.0000] <AryehGregor>
I.e., I argue that features should not be non-conforming merely because they're obsolete.  The same logic applies to nearly all features that are non-conforming because they're obsolete.

[20:10:04.0000] <AryehGregor>
I don't see how I would be able to narrow the bug.

[20:11:00.0000] <othermaciej>
None of the features that are non-conforming "merely because they're obsolete"

[20:11:01.0000] <AryehGregor>
Really?

[20:11:02.0000] <AryehGregor>
It looks like they are to me.

[20:11:03.0000] <othermaciej>
obsolete is a specific subcategory of "non-conforming"

[20:11:04.0000] <othermaciej>
it is not a justification, it is a label

[20:11:05.0000] <othermaciej>
the rationale for things being noncomforming is in the new section Hixie just added

[20:12:00.0000] <AryehGregor>
Where?

[20:12:01.0000] <AryehGregor>
Maybe I missed it.

[20:12:02.0000] <AryehGregor>
The various features are listed as being obsolete, and the only hint as to why is that alternative markup is suggested.

[20:12:03.0000] <AryehGregor>
I mean, as non-conforming.

[20:12:04.0000] <othermaciej>
hang on, waiting for spec to load

[20:13:00.0000] <othermaciej>
it used to be there was no in-sepc rationale that said why those things are nonconforming

[20:13:01.0000] <othermaciej>
*in-spec

[20:13:02.0000] <AryehGregor>
The implication (to me) is that the only reason they're non-conforming is because these other features exist, which are in most cases exactly equivalent but are considered better for various reasons.

[20:13:03.0000] <AryehGregor>
They can't actually be very much better if they're functionally identical, of course.

[20:30:00.0000] <AryehGregor>
Okay, well, I'm going to bed.  If you have specific suggestions for how I should split up the bug or any further points that I should address, othermaciej, feel free to ping me to tell me of them.

[20:30:01.0000] <othermaciej>
hang on

[20:30:02.0000] <othermaciej>
my computer was hosed

[20:30:03.0000] <othermaciej>
read this: http://dev.w3.org/html5/spec/Overview.html#conformance-requirements-for-authors

[20:30:04.0000] <othermaciej>
AryehGregor: ^

[20:31:00.0000] <AryehGregor>
Aha.

[20:31:01.0000] <othermaciej>
that states many individual reasons for particular author conformance requirements

[20:31:02.0000] <othermaciej>
none of the justifications amounts to "is obsolete"

[20:31:03.0000] <othermaciej>
you may disagree with some of them, or you may feel some things marked obsolete and nonconforming when none of those criteria apply

[20:32:00.0000] <AryehGregor>
I didn't know about that.

[20:32:01.0000] <othermaciej>
yeah, it's new

[20:32:02.0000] <AryehGregor>
I'll look it over and amend the bug report accordingly, when I have time.

[20:32:03.0000] <othermaciej>
Hixie added that in response to bug 7034

[20:32:04.0000] <othermaciej>
I think your bug is arguably just a duplicate of bug 7034

[20:32:05.0000] <othermaciej>
(which is titled "remove or replace author conformance requirements")

[20:33:00.0000] <othermaciej>
I guess it is slightly less narrow in that it's limited to the obsolete section

[20:33:01.0000] <AryehGregor>
No, I only argue for removing a subset of them.

[20:33:02.0000] <othermaciej>
but that's kind of arbitrary

[20:33:03.0000] <AryehGregor>
There are lots of author conformance requirements that I think are a good idea.

[20:33:04.0000] <AryehGregor>
For instance, they can help to catch authoring mistakes.

[20:33:05.0000] <othermaciej>
note that not everything in the obsolete section is even presentational, though much of your argument is about presentational markups pecifically

[20:33:06.0000] <AryehGregor>
Or inform authors of areas where they might be running into interoperability problems, or pressure authors to do something that we have good reason to want them to do (e.g., alt="").

[20:34:00.0000] <AryehGregor>
Thus the many/most/all in the bug summary.

[20:34:01.0000] <AryehGregor>
Much of it applies to all of them, I think.

[20:35:00.0000] <AryehGregor>
It looks like the justification for presentational markup offers no arguments that couldn't equally apply to style="".

[20:35:01.0000] <othermaciej>
that's why it gives a specific reason for exempting style=""

[20:35:02.0000] <othermaciej>
which you may or may not agree with

[20:35:03.0000] <othermaciej>
anyway

[20:35:04.0000] <othermaciej>
I think a bug about just presentational markup might make sense

[20:35:05.0000] <AryehGregor>
Maybe I'll repurpose it.

[20:35:06.0000] <othermaciej>
I don't see how your argument applies to, say, <bgsound>

[20:35:07.0000] <othermaciej>
or <applet>

[20:36:00.0000] <othermaciej>
or <frameset>

[20:36:01.0000] <AryehGregor>
I don't know what the first two do, really.

[20:36:02.0000] <othermaciej>
but if you make the bug be about that, then I suggest you should address Ian's argument in the rationale

[20:36:03.0000] <othermaciej>
well if you don't know what they do, you probably don't have enough info to suggest they should be conforming :-)

[20:36:04.0000] <AryehGregor>
<frameset> is something that's banned because it's considered to be something authors should almost never do, because it has serious problems, although they might be tempted to do it.

[20:36:05.0000] <AryehGregor>
Which is why I didn't say all obsolete features should be conforming.

[20:37:00.0000] <othermaciej>
if you say many/most/all, then you are asking someone else to guess what you mean

[20:38:00.0000] <othermaciej>
all I'm saying is, make it clear enough that no one has to guess which ones you have in mind

[20:38:01.0000] <AryehGregor>
I guess I can present a narrow argument, and he can extend it to everything else it applies to if he likes.

[20:39:00.0000] <othermaciej>
well, even "presentational" is a fairly clear category, though there may be edge cases that are hard to judge

[20:40:00.0000] <othermaciej>
most obsolete nonconforming features are that way for reason of being presentational it seems, but there are quite a few exceptions which generally have an independent and plausible reason to be banned

[20:41:00.0000] <othermaciej>
anyway, this is mainly my advice on how to make a good bug report - if you choose to do nothing, then the project won't blow up or anything, but 

[20:41:01.0000] <othermaciej>
I think some revision would be wise

[20:41:02.0000] <othermaciej>
anyway, I won't keep you up any longer

[20:51:00.0000] <AryehGregor>
othermaciej, okay, how's this?  <http://www.w3.org/Bugs/Public/show_bug.cgi?id=9355>

[20:51:01.0000] <AryehGregor>
(now I'm really going to bed, won't see your response till tomorrow)

[21:35:00.0000] <JonathanNeal>
Hello

[22:24:00.0000] <lazni>
has anyone diffed DOMs using html5lib?

[22:31:00.0000] <lazni>
there's a firefox extension, which would use html5.enabled

[22:45:00.0000] <lazni>
ah, SAX output, that would reduce the load of comparing

[00:07:00.0000] <othermaciej>
AryehGregor: looks more focused to me

[01:36:00.0000] <JonathanNeal>
TabAtkins, if you read this message when you come back, I have new IE json stylesheets for you.

[03:55:00.0000] <jgraham>
AFAICT there is no difference between "obsolete but conforming" and "transitional"

[03:59:00.0000] <othermaciej>
jgraham: in plain english meaning - perhaps not

[03:59:01.0000] <othermaciej>
jgraham: operationally - seems like a fair bit of difference between "mandatory non-error warning" and "you can choose to make all of them errors or all of them silent with a flag inside the document"

[04:00:00.0000] <othermaciej>
from the developer perspective, I like the warning better, if validators have the UI in non-default mode to selectively silence them or force them into errors based on category

[04:03:00.0000] <jgraham>
The difference between "Your document is valid but uses some obsolete features" and "Tour document is validates as HTML transitional" seems rather subtle to me. Presumably the majority of authors using transitional doctypes know that they are using obsolete features

[04:03:01.0000] <jgraham>
*Your

[04:08:00.0000] <othermaciej>
jgraham: the difference in overall wording is subtle, but getting the specific list of uses of obsolete features is a big practical difference

[04:09:00.0000] <othermaciej>
(whether for good or ill, I don't know, though I tend to think for good)

[04:11:00.0000] <jgraham>
Yes, having the list presented is, I guess better

[04:13:00.0000] <othermaciej>
but hey, maybe calling them "transitional features" instead of "obsolete but conforming features" might be better branding

[04:14:00.0000] <jgraham>
If we really must have a list of "features you are allowed to use but validators must warn about", could be get rid of the oxymornic-sounding name, present them with a special obsolete marking, and just say "authors should not use obsolete features. Validators must warn when obsolete features are used"

[04:15:00.0000] <jgraham>
Although I think my preferred solution is still not to have that list

[04:15:01.0000] <othermaciej>
well, there's also obsolete features that are nonconforming and must result in an error

[04:15:02.0000] <othermaciej>
so you can't just use the unqualified term "obsolete"

[04:16:00.0000] <jgraham>
Yeah true

[04:16:01.0000] <jgraham>
They should be called this-is-a-bad-idea-and-you-don't-want-to-do-this-but-hey-experience-suggests-that-you-will-do-anyway features

[04:16:02.0000] <othermaciej>
I think that list and its status is on some level a compromise between Henri, who did not want to spam people with useless errors for markup that is in fact harmless, and Ian, who wanted to alert authors of new documents that there are certain effectively useless constructs which they should not bother with

[04:17:00.0000] <jgraham>
I thought it was mainly a result of the great accessibility wars

[04:17:01.0000] <jgraham>
But I could be misremembering

[04:17:02.0000] <othermaciej>
there is only one item in there on the basis of accessibility

[04:18:00.0000] <othermaciej>
I guess before that it was a list of "downplayed errors"

[04:18:01.0000] <jgraham>
Yes, that's what I recall

[04:18:02.0000] <othermaciej>
but the concept of "downplayed error" is silly

[04:18:03.0000] <jgraham>
Well so is "obsolete but conforming"

[04:18:04.0000] <othermaciej>
that too

[04:19:00.0000] <othermaciej>
my own suggestion was to just describe them in terms of the effect ('"validators must issue a warning when X") instead of as a conceptual category

[04:21:00.0000] <othermaciej>
yeah, other than summary, all the others seem to be more in the "waste of electrons" category

[04:22:00.0000] <othermaciej>
but there's only 6 things total in the "obsolete but conforming" category

[04:23:00.0000] <othermaciej>
which makes it seem like a waste to name it, especially when the name is confusing to so many peple

[04:23:01.0000] <othermaciej>
*people

[04:23:02.0000] <jgraham>
It seems like one possible future is that a lot more things end up as obsolete but conforming

[04:24:00.0000] <othermaciej>
could be

[04:24:01.0000] <othermaciej>
I think so far only Aryeh has outright advocated that path, but his particular proposal would indeed turn the category into "transitional"

[04:25:00.0000] <othermaciej>
it's kind of a lie to call it "transitional" if it's been over 10 years and the transition still hasn't happened...

[04:25:01.0000] <jgraham>
Indeed

[04:25:02.0000] <othermaciej>
anyway I am almost finished breaking down all the errors in the conformance error study

[04:26:00.0000] <othermaciej>
http://www.w3.org/html/wg/wiki/index.php?title=HTML5_Authoring_Conformance_Study

[04:26:01.0000] <othermaciej>
only ebay and nytimes remain to be typed in

[04:27:00.0000] <othermaciej>
it seems like unescaped & is the most common single error, by far

[04:28:00.0000] <othermaciej>
and after that I guess presentational markup

[04:28:01.0000] <othermaciej>
X-UA-Compatible is surprisingly common, but since it is anti-interoperability by design, it seems silly to make it conforming

[04:29:00.0000] <othermaciej>
there are surprisingly many sites with a stray close tag, also

[04:36:00.0000] <jgraham>
Not that surprising; it's almost always totally harmless, an easy error to make, and hard to spot

[04:39:00.0000] <othermaciej>
I think the main reason for stray close tag to be an error is that there are a lot of authoring mistakes that would show up that way (forgetting to close a tag at all near some that may implicitly close, wrong close tag, misnested tags, etc)

[05:16:00.0000] <Philip`>
Why is x-ua-compatible anti-interoperability, when including it makes pages work in *more* browsers than if it was not included?

[05:22:00.0000] <Philip`>
(It would be a problem if people used x-ua-compatible as a way to avoid having to write code that works in browsers that follow standards better, but in practice it seems more commonly used on sites that already have a standards-browser codepath plus an IE7 codepath and don't want to add an extra codepath for IE8 since it's too different from the existing ones)

[05:23:00.0000] <Philip`>
(so it's not causing any disadvantage to browsers that implement standards better than IE does)

[08:23:00.0000] <Dashiva>
"The absence of Theora (or Dirac) as baseline codec means that a good portion of user agents will likely omit Theora support"

[08:23:01.0000] <Dashiva>
If you accept that people will be wrong on the internet, is it preferale that they be wrong the same way even after being corrected, or that they come up with new ways to be wrong every time?

[09:17:00.0000] <JonathanNeal>
Goodmorning

[10:09:00.0000] <divya>
I have a question regarding the flow content as defined in HTML5 spec: http://www.w3.org/TR/html5/dom.html#flow-content

[10:11:00.0000] <divya>
the spec says "Most elements that are used in the body of documents and applications are categorized as flow content." but the pictorial representation seems to indicate the only way an element wont be a flow content is if it is meta data content (and somehow not be flow content). http://www.w3.org/TR/html5/images/content-venn.svg Is this right? 

[10:12:00.0000] <divya>
or are there hard rules that can be used to define if an element is flow content or not (spec does not define one)? 

[10:14:00.0000] <Dashiva>
Each element specifies which categories it is in

[10:16:00.0000] <divya>
Dashiva: can you elaborate? How does it specify? Is it because it belongs to the list of elements specified in the doc?

[10:16:01.0000] <Dashiva>
http://www.w3.org/TR/html5/semantics.html#the-hgroup-element

[10:16:02.0000] <Dashiva>
Categories: Flow content. Heading content. formatBlock candidate.

[10:17:00.0000] <divya>
Dashiva: got it. so there is no rule, except to refer to the doc to find out if an element is a flow element or not. 

[10:18:00.0000] <paul_irish>
<source> probably isnt flow content

[10:19:00.0000] <divya>
neither is <legend>

[12:41:00.0000] <AryehGregor>
The irony of a page entitled "HTML5: Cross-Browser Best practices" being written in Silverlight is . . . a lot.  http://live.visitmix.com/MIX10/Sessions/CL27

[12:59:00.0000] <gsnedders>
jgraham: FYI, I was in Sweden yesterday and am today (just now, actually) back in St Andrews

[13:02:00.0000] <othermaciej>
AryehGregor: lol

[13:03:00.0000] <AryehGregor>
They do offer the same content on the right-hand side in three other proprietary formats, though!

[13:03:01.0000] <AryehGregor>
And one open patent-encumbered format.

[13:27:00.0000] <othermaciej>
the width and height attributes on span don't do anything, right?

[13:27:01.0000] <workmad3>
othermaciej: not unless you change it's display to block

[13:28:00.0000] <othermaciej>
workmad3: I'm asking about attributes, not CSS properties - I am thinking the attributes don't ever do anything (unlike <div>, where they map to css propertieS)

[13:28:01.0000] <workmad3>
ah right

[13:28:02.0000] <workmad3>
wasn't even aware that width and height were valid attributes on a <span> :)

[13:29:00.0000] <othermaciej>
they are not

[13:44:00.0000] <AryehGregor>
/me is mildly fascinated by which of the language links on the left side of http://en.wikipedia.org/wiki/HTML5 have a space in the article name and which don't

[13:44:01.0000] <AryehGregor>
/me wonders if there are any patterns

[13:50:00.0000] <Dashiva>
I doubt it

[15:11:00.0000] <zcorpan>
http://urbanlistening.wordpress.com/2010/03/28/projekktor-zwei-html5-video-player/

[15:23:00.0000] <Dashiva>
The meta @keywords bug is quite verbose...

[15:24:00.0000] <Dashiva>
I don't really agree with the argument "content-producing parts of google use @keywords, this is proof google search uses @keywords"

[15:26:00.0000] <webben>
Yeah, that's bogus.

[15:27:00.0000] <webben>
"content-producing parts of google use @keywords, this suggests some Googlers think some search engine might use keywords" would work

[15:27:01.0000] <webben>
(where "might" can include future engines, of course)

[15:27:02.0000] <Dashiva>
Or just "content-producing parts of google use tools that think @keywords is useful"

[15:28:00.0000] <webben>
Mmm maybe.

[15:28:01.0000] <webben>
At Yahoo! we add keywords and that's definitely deliberate not an artefact of tools.

[15:31:00.0000] <Dashiva>
Besides, Google would have no trouble using @keywords for its own sites, since they would be known to be reliable

[15:33:00.0000] <AryehGregor>
Dashiva, I sure hope Google Search doesn't treat Google-hosted sites differently from any othrs.

[15:35:00.0000] <Dashiva>
AryehGregor: Not necessarily used for Google search

[15:37:00.0000] <TabAtkins>
Google has said explicitly that it does not pay attention to keywords.

[15:38:00.0000] <Philip`>
It could be mistaken, or lying

[15:40:00.0000] <deltab>
Google Web Search is not the only crawling Google does

[15:40:01.0000] <Hixie>
i'm not aware of anything we do that uses keywords

[15:41:00.0000] <AryehGregor>
Saying that doesn't violate your NDA or anything?

[15:43:00.0000] <Dashiva>
Everyone in #whatwg works for Google anyhow

[15:45:00.0000] <Hixie>
AryehGregor: it's part of my job to represent google on committees; part of that is determining what information i can say.

[15:45:01.0000] <AryehGregor>
Reasonable.

[15:46:00.0000] <AryehGregor>
Dashiva, no, they mostly seem to work for Opera as far as I can tell.

[15:46:01.0000] <AryehGregor>
Do we have chat statistics hanging around, of biggest talkers and such?  It would be fun to do breakdown by employer.

[15:46:02.0000] <Philip`>
Opera is just a front for Google

[15:47:00.0000] <gsnedders>
AryehGregor: Heh. To such an extreme that in 2008 MikeSmith predicted that within five years I would be working for Opera :)

[15:47:01.0000] <Dashiva>
Mozilla too, obviously

[15:47:02.0000] <Philip`>
It helps them monopolise the web without everybody noticing

[15:47:03.0000] <Philip`>
AryehGregor: See /topic

[15:47:04.0000] <jgraham>
Actually Google is just a front for Opera

[15:48:00.0000] <gsnedders>
jgraham: Go to bed.

[15:48:01.0000] <gsnedders>
Oh, wait, you're not an hour ahead of the time here.

[15:48:02.0000] <gsnedders>
It's just the time is still Europe/Stockholm on my laptop :P

[15:48:03.0000] <jgraham>
gsnedders: I thought I was? Did the clocks not change there?

[15:48:04.0000] <Dashiva>
The clocks changed last night

[15:48:05.0000] <gsnedders>
jgraham: They did

[15:48:06.0000] <gsnedders>
jgraham: Just not an hour ahead of the time here on my laptop

[15:48:07.0000] <Dashiva>
Philip`: Apparently you're unpopular

[15:49:00.0000] <gsnedders>
Shit, Lachy's almost caught up with me in number of lines

[15:49:01.0000] <jgraham>
Anyway, like I was saying it's always the little harmless looking one that turns out to weild ultimate power

[15:49:02.0000] <Dashiva>
"Hixie's faithful follower, Philip`, didn't speak so much: 534519 words."

[15:49:03.0000] <Dashiva>
Clearly Philip` is also a Google shill

[15:49:04.0000] <AryehGregor>
/me realizes he has no idea who a bunch of the people here work for, but is pretty sure at least four of the top ten there are Opera

[15:50:00.0000] <gsnedders>
AryehGregor: Of the top ten, five work for Opera

[15:50:01.0000] <AryehGregor>
And does anyone else employ more than one?

[15:50:02.0000] <gsnedders>
No

[15:51:00.0000] <zcorpan>
opera hires all #whatwg regulars

[15:51:01.0000] <jgraham>
Google, Opera, PhD, Mozilla (contractor), Apple, Opera, Opera, Opera, Opera, W3C

[15:51:02.0000] <othermaciej>
Dashiva: we already added back meta keywords, not sure why we had a second bug about it

[15:51:03.0000] <Dashiva>
The mozillas appear further down

[15:51:04.0000] <gsnedders>
zcorpan: Were you here before or after you joined Opera?

[15:52:00.0000] <zcorpan>
gsnedders: before

[15:52:01.0000] <jgraham>
Dashiva: What do you do these days?

[15:52:02.0000] <gsnedders>
zcorpan: When did you join here?

[15:52:03.0000] <othermaciej>
btw http://www.w3.org/html/wg/wiki/index.php?title=HTML5_Authoring_Conformance_Study is pretty much finished

[15:52:04.0000] <zcorpan>
gsnedders: 2005 iirc

[15:52:05.0000] <gsnedders>
zcorpan: Oh, my memory is really diabolical.

[15:52:06.0000] <othermaciej>
there are lots of interesting patterns in the data left to be teased out though

[15:52:07.0000] <othermaciej>
I may try to do a count of which errors are the most common across sites

[15:53:00.0000] <Dashiva>
jgraham: I'm finishing my master's degree (for real this time). I start at Google this summer.

[15:53:01.0000] <othermaciej>
or more specifically which presentational markup gets used the most

[15:53:02.0000] <othermaciej>
bus I filed so far are here: http://www.w3.org/html/wg/wiki/index.php?title=HTML5_Authoring_Conformance_Study#Bugs_Filed_Against_Spec

[15:53:03.0000] <jgraham>
Dashiva: Oh, cool. Know what you're doing at Google?

[15:53:04.0000] <othermaciej>
(well, and bugs other people filed)

[15:53:05.0000] <Dashiva>
Front end stuff

[15:54:00.0000] <Philip`>
Dashiva: Only 534519? That's not even as many as War and Peace :-(

[15:55:00.0000] <Dashiva>
Don't worry, it's only been a few years. You still have time.

[15:55:01.0000] <Dashiva>
2022, remember

[15:56:00.0000] <jgraham>
/me is worried that zcorpan has said almost as many more lines than him as the total number of lines from roc

[15:56:01.0000] <Dashiva>
Be worried about your own word count before worrying about others :P

[15:57:00.0000] <jgraham>
You misunderstand. I'm not worried for zcorpan's benefit

[15:57:01.0000] <roc>
I don't know what jgraham is worried about exactly, but there's nothing wrong with being terse

[15:58:00.0000] <gsnedders>
Ah, new snapshot with 2.5.24

[15:59:00.0000] <jgraham>
"Worried" is maybe the wrong word. It just seems like an incomprehensibly large gap

[16:00:00.0000] <Dashiva>
Tell his manager, grab his spot on the rankings :P

[16:00:01.0000] <gsnedders>
Hmm, http://developers.slashdot.org/story/10/03/28/208210/Best-Way-To-Land-Entry-Level-Job

[16:00:02.0000] <gsnedders>
I guess I circumvented this whole get-a-job-after-graduating problem by not going to uni

[16:02:00.0000] <Dashiva>
I've been lucky enough to never need to

[16:02:01.0000] <gsnedders>
Never need to what?

[16:02:02.0000] <Dashiva>
Apply for a job from the outside

[16:03:00.0000] <gsnedders>
Ah. I technically did for Opera, with an application thrown together very quickly on the final day

[16:03:01.0000] <gsnedders>
(including whole CV and everything)

[16:04:00.0000] <gsnedders>
My interview, however, was more just like a random conversation. :D

[16:04:01.0000] <Hixie>
it is incredibly hard to get the a11y people to actually suggest alt text

[16:04:02.0000] <Dashiva>
I got my interview with Opera from hanging out and giving help on #javascript :)

[16:05:00.0000] <Hixie>
they're eager to tell you what advice is _wrong_, but try to get them to give you positive advice, and sheesh

[16:05:01.0000] <gsnedders>
Dashiva: I got mine from hanging out here :)

[16:06:00.0000] <AryehGregor>
I can probably get a job at Wikimedia because I was a volunteer developer for them for the last 3.5 years, so I also hopefully won't have to apply for a job anytime soon.

[16:06:01.0000] <AryehGregor>
(currently working on a Ph.D., though)

[16:06:02.0000] <Dashiva>
Hixie: If there's no perfect solution, any near-optimal solution will still have flaws, and nobody wants to suggest flawed advice.

[16:06:03.0000] <AryehGregor>
Not having to actually be part of the job market is great.  \o/

[16:07:00.0000] <Hixie>
Dashiva: well then why complain about the advice in the spec that admits that any solution will be flawed?

[16:07:01.0000] <Dashiva>
Because it isn't perfect

[16:09:00.0000] <Hixie>
so it's better to advise people to do something impossible, and thus guarantee that everyone will do it wrong, than advise people to do the best that is possible, thus at least making it possible for some people to get it as right as possible?

[16:10:00.0000] <Dashiva>
Protecting the image of perfection can be important to some people

[16:10:01.0000] <Dashiva>
Or less pointedly, the moment you admit perfection is impossible, it can seem like you're accepting mediocrity

[16:11:00.0000] <AryehGregor>
Hixie, surely most feed readers strip <style>?

[16:11:01.0000] <AryehGregor>
/me supposes he can test

[16:12:00.0000] <Dashiva>
Wasn't syndication one of the main use cases for <style scoped?

[16:12:01.0000] <AryehGregor>
I have no idea.

[16:13:00.0000] <jgraham>
My conclusion about feed readers is that I almost always want them to remove document styles and I don't trust authors to use the right markup in cases where the style is essential compared to cases where the style is ephemeral

[16:13:01.0000] <AryehGregor>
jgraham, so if the author includes a table with borders, should it remove the borders?

[16:14:00.0000] <zcorpan>
i'd like the same table style on all tables in feeds

[16:14:01.0000] <zcorpan>
but i almost never see any tables in feeds anyway

[16:14:02.0000] <jgraham>
I think typically it would be fine to display in the standard style for the feedreader

[16:14:03.0000] <zcorpan>
so can't be bothered to write a user style sheet

[16:14:04.0000] <AryehGregor>
Even when one table is used presentationally (borders wreck the layout) and one is used semantically (borders necessary to look good)?

[16:15:00.0000] <jgraham>
Do items in feeds typically have tables for layout? I don't recall ever seeing that

[16:15:01.0000] <AryehGregor>
I have to admit, I agree that it's nice they all look consistent.

[16:15:02.0000] <AryehGregor>
No, I can't think of any time I've seen that.

[16:15:03.0000] <AryehGregor>
Hmm.

[16:15:04.0000] <AryehGregor>
In practice, do feed readers usually strip styles . . . ?

[16:15:05.0000] <jgraham>
I guess

[16:16:00.0000] <AryehGregor>
Inline styles, I mean, obviously.

[16:16:01.0000] <AryehGregor>
/me tries to find a likely example

[16:17:00.0000] <AryehGregor>
Google Reader doesn't strip inline style.

[16:17:01.0000] <daedb_>
cia.vc has styles in their feeds

[16:17:02.0000] <AryehGregor>
It's so obviously necessaryt \

[16:18:00.0000] <AryehGregor>
It's so obviously necessary for it to strip <style> that I'm not even going to test that, though.  (Don't have a feed handy that I really control . . .)

[16:18:01.0000] <AryehGregor>
Yeah, the styles on cia.vc feeds are really noticeable.  Quite annoying, honestly.

[16:19:00.0000] <webben>
Hixie: Seems to me that bug resolves around the question of text /equivalence/ versus identification and that Laura is asking for caption ("descriptive identification") in alt.

[16:22:00.0000] <webben>
Hixie: In particular, the bit quoted from the spec treats "alternative text" as equivalent, whereas the bit quoted from WCAG2 distinguishes "text alternative[s] that serve.. the equivalent purpose" from some superset of text alternatives.

[16:23:00.0000] <webben>
In WCAG2 "text alternative" pretty much just means "programmatically associated text": http://www.w3.org/TR/WCAG/#text-altdef

[16:24:00.0000] <webben>
I think the question needs to be whether inserting "descriptive identification" in @alt versus the other methods the spec mentions is useful or not.

[16:27:00.0000] <webben>
the bug's confused by the fact the webcam image actually contains textual information.

[16:31:00.0000] <Hixie>
webben: yeah i certainly agree that we should provide captions for these images, and that's what the spec suggests -- but it seems obvious that mixing captions and replacements is suboptimal

[16:32:00.0000] <Hixie>
webben: currently i'm just trying to get an accessibility expert to tell me what the alt="" text should be so that i can determine if it is a caption or not

[16:33:00.0000] <Dashiva>
othermaciej: The bugs you changed with "moved back to verified" comment seem to be only at resolved.

[16:33:01.0000] <webben>
Yep I get that. I think there's agreement in that ticket that providing programmatically associated captions is good, and the disagreement is about how to do so.

[16:35:00.0000] <webben>
From a WCAG1/2 Techniques perspective (and common practice I guess), @alt is already overloaded to mix caption and replacement.

[16:37:00.0000] <Hixie>
well wcag1/2 were written in a world where alt="" was required, so it makes sense they'd use that hammer for all their problems

[16:37:01.0000] <Hixie>
but we have the opportunity to think this out better now

[16:46:00.0000] <zcorpan>
wait why doesn't embed fire 'load'?


2010-03-29
[18:12:00.0000] <AryehGregor>
/me realizes that there was tons of public-html discussion about presentational markup that he missed before filing his bug -- oh well

[18:17:00.0000] <TabAtkins>
Hehe.

[18:18:00.0000] <AryehGregor>
I like how when people are using a Bugzilla that's not meant to track bugs in client software, it still auto-fills the platform field, so you can figure out whether everyone uses Windows/Mac/Linux.

[18:18:01.0000] <AryehGregor>
You'd think they'd make those fields disabled by default, since most Bugzilla instances surely don't need them . . .

[18:52:00.0000] <othermaciej>
Dashiva: if I missed one can you give me a URL for it?

[18:53:00.0000] <othermaciej>
Hixie: what's the benefit of being able to distinguish a text repacement from a non-replacement text description for an image?

[18:53:01.0000] <othermaciej>
Hixie: maybe the pragmatic thing to do is to loosen the definition of alt, rather than have complicated rules about what goes where

[18:53:02.0000] <TabAtkins>
A text replacement can be simply read as normal text, while a description can be denoted as such (rather than a seemingly-random bit of description in the middle of content).

[18:55:00.0000] <othermaciej>
I think screen readers typically distinguish both (there's no way typically to get it not to identify an image as such, even if it has alt text)

[18:55:01.0000] <othermaciej>
I suppose text-only UAs might treat them differently, though I am not sure they actually do anything with, for example, the title attribute

[18:55:02.0000] <othermaciej>
so it's hard to predict what they would do if they did

[18:57:00.0000] <othermaciej>
yowza, there's 178 open bugs

[18:58:00.0000] <Dashiva>
othermaciej: Maybe I just was confused, it seems the confirmed -> verified transition didn't generate bugmail

[18:58:01.0000] <Dashiva>
*resolved -> verified

[18:58:02.0000] <annevk>
/me is sleepy

[18:58:03.0000] <othermaciej>
Dashiva: yeah, it has to be done in two steps

[18:59:00.0000] <annevk>
btw othermaciej, will you be at the CSS get-together later today?

[18:59:01.0000] <othermaciej>
annevk: oh, I didn't know there was a thing today - I was considering coming for some of the meeting tomorrow and/or joining people for dinner during the week if there was a scheduled one

[19:00:00.0000] <annevk>
some people are meeting at 7pm at the hotel close to inf loop

[19:00:01.0000] <annevk>
not sure when the Apple organized dinner is

[19:01:00.0000] <othermaciej>
ah, doubt I can make it there in time (I'm in San Francisco)

[19:28:00.0000] <annevk>
it's on Tuesday per TabAtkins

[19:33:00.0000] <othermaciej>
cool

[19:50:00.0000] <JonathanNeal>
Ahoy!

[19:50:01.0000] <JonathanNeal>
TabAtkins, hhheyo

[20:49:00.0000] <Hixie>
othermaciej: the benefit is that for replacement text, you can forget that it was an image, and don't need to taunt the user with "image!" each time

[20:50:00.0000] <othermaciej>
Hixie: do any screen reader vendors do that, or plan to?

[20:50:01.0000] <othermaciej>
(VoiceOver does say "image" all the time; not aware of any plans to change)

[20:51:00.0000] <Hixie>
well i doubt any of them would plan to change unless we make the spec support that case

[20:51:01.0000] <Hixie>
currently it's a self-reinforcing race to the bottom

[21:03:00.0000] <Dashiva>
Not much to do about it as long as supporting Jaws 4 remains a requirement

[21:45:00.0000] <othermaciej>
does anyone have IE handy?

[21:45:01.0000] <othermaciej>
I am trying to determine whether any browser does what the spec says for <wbr> (rather than something essentially functionally identical to a zero-width space)

[21:49:00.0000] <othermaciej>
if anyone could save the text here into an HTML file and load it in IE, I would be much obliged: http://paste.lisp.org/display/97008

[21:50:00.0000] <othermaciej>
in Safari and Firefox, wbr just seems to create a breaking opportunity if line breaks are allowed at all, rather than doing so regardless of the whitespace mode

[21:51:00.0000] <Hixie>
check for open bugs on <wbr>, i remember doign it then someone saying what the spec said was wrong but i don't recall if i already fixed it

[21:51:01.0000] <othermaciej>
ok, I'll look

[21:51:02.0000] <othermaciej>
Opera does not grok <wbr> (still)

[21:52:00.0000] <Hixie>
yeah that's why i added it, zcorpan (iirc) told me they needed it for compat

[21:53:00.0000] <othermaciej>
I really should get my Windows partition working, and VMWare or something installed so I can test IE myself on my laptop

[21:54:00.0000] <othermaciej>
I see, there's http://www.w3.org/Bugs/Public/show_bug.cgi?id=9097

[21:54:01.0000] <Hixie>
i had windows 7 rc in a vm for a while but it expired

[21:54:02.0000] <Hixie>
i should get a real w7 license

[21:54:03.0000] <othermaciej>
I think Simon is correct that <wbr> does something essentially equivalent to a ZWSP, but I am not sure if it is equivalent in all details or all browsers

[21:55:00.0000] <othermaciej>
however, what the spec says now is definitely wrong:

[21:55:01.0000] <othermaciej>
"The wbr element is expected to override the 'white-space' property and always provide a line-breaking opportunity."

[21:55:02.0000] <othermaciej>
it provides a line-breaking opportunity but it does *not* (as far as I can tell) override the 'white-space' property

[21:55:03.0000] <othermaciej>
at least not in WebKit or Gecko

[21:55:04.0000] <Hixie>
well it definitely overrides <nobr> iirc

[21:55:05.0000] <Hixie>
no?

[21:56:00.0000] <othermaciej>
it doesn't override white-space: pre

[21:56:01.0000] <othermaciej>
let me try nowrap

[21:56:02.0000] <Hixie>
<wbr> and <nobr> are such a huge pain in the neck, i wish we could just get rid of them altogether (like <spacer>), rather than just obsolete them

[21:56:03.0000] <othermaciej>
ah, it does override nowrap

[21:56:04.0000] <othermaciej>
but not pre

[21:56:05.0000] <othermaciej>
so it's not the same as ZWSP

[21:57:00.0000] <othermaciej>
ZWSP doesn't override white-space nowrap

[21:57:01.0000] <othermaciej>
I proposed going the other way, since popular sites actually use them, and at least in the case of <wbr> there is apparently no other way to get the same effect

[21:57:02.0000] <Hixie>
popular sites use all kinds of crap

[21:57:03.0000] <Hixie>
doesn't make it a good idea

[21:58:00.0000] <Hixie>
there are plenty of ways of getting line breaking opportunities into text

[21:58:01.0000] <othermaciej>
"line-breaking opportunity" seems like a valid use case

[21:59:00.0000] <othermaciej>
I dunno, you can judge when you read my bugs about this

[21:59:01.0000] <Hixie>
it's what ZWSP is for

[21:59:02.0000] <othermaciej>
yeah, except ZWSP doesn't seem to quite match <wbr> in that it won't override white-space: nowrap

[21:59:03.0000] <miketaylr>
othermaciej: here's IE7/vista: http://miketaylr.com/post/865d1bfa.png

[22:00:00.0000] <miketaylr>
(alls i have on hand ATM)

[22:00:01.0000] <Hixie>
othermaciej: it's easy to turn off white-space:nowrap, so that doesn't seem like a problem

[22:00:02.0000] <othermaciej>
miketaylr: interesting - looks like in IE, both <wbr> and ZWSP override all white-space: pre

[22:00:03.0000] <miketaylr>
yeah

[22:01:00.0000] <othermaciej>
Hixie: sure, you could insert <span style="white-space: normal">&#8203;</span> instead of <wbr>

[22:01:01.0000] <Hixie>
well more likely you wouldn't do anything like that, you'd just do it all in css

[22:01:02.0000] <Hixie>
since where you want line-breaking depends on your styles

[22:02:00.0000] <othermaciej>
I don't see how you could do it all in CSS - line breaking opportunities, like hard line breaks, have to be places in specific places in the content

[22:02:01.0000] <othermaciej>
they depend on the content, not the style

[22:03:00.0000] <Hixie>
well it depends on what you're doing

[22:03:01.0000] <othermaciej>
miketaylr: would you mind trying a version of that with white-space: nowrap instead of white-space: pre?

[22:03:02.0000] <miketaylr>
sure one sec

[22:04:00.0000] <Hixie>
if you're strictly defining line-breaking opportunities without a particular style in mind, then sure, ZWSP is fine

[22:04:01.0000] <Hixie>
if you're actually trying to make headlines flow in various ways depending on the width of the page, or whatnot, then it's CSS

[22:04:02.0000] <Hixie>
sometimes it might be a mixture

[22:04:03.0000] <othermaciej>
at least one use case for <wbr> seems to be a very long run of text containing no whitespace; in many cases ZWSP will work just as well, is just more painful to type

[22:05:00.0000] <Hixie>
often though, <wbr> is used where you would have to move it if you had an alternative style sheet, and that indicates to me that it's not semantic

[22:05:01.0000] <Hixie>
and should be done in CSS instead

[22:05:02.0000] <othermaciej>
in the case of a run of text where you need to inhibit normal line breaking but want explicit break opportunities, then you probably need <wbr> (though I suppose you could replace all your normal whitespace with non-breaking space)

[22:06:00.0000] <othermaciej>
(an example I can imagine like that is pairs of words where you don't want to break in the middle of a pair, only between pairs)

[22:07:00.0000] <Hixie>
&nbsp; and &zwsp; seem like the right way to handle that. It'd even survive transferring to other mediums that use text but don't use HTML.

[22:08:00.0000] <Hixie>
(&zwsp; isn't valid though)

[22:08:01.0000] <Hixie>
(so you'd have to use the entity if you didn't have the unicode character inline)

[22:08:02.0000] <miketaylr>
...and white-space:nowrap: http://miketaylr.com/post/5d157284.png

[22:08:03.0000] <karlcow>
what about " "?

[22:08:04.0000] <othermaciej>
does &zwsp; do anything?

[22:09:00.0000] <othermaciej>
like is it working but invalid, or just nonexistent?

[22:09:01.0000] <Hixie>
non-existent

[22:09:02.0000] <karlcow>
" " = 00A0

[22:11:00.0000] <othermaciej>
so yes, &#8203; is a plausible alternative for <wbr>, though it doesn't do all the same things

[22:11:01.0000] <Hixie>
othermaciej: though you can use &zwnj; which is more or less the same as i understand it

[22:11:02.0000] <othermaciej>
as well as being way less fun to input

[22:11:03.0000] <Hixie>
i wouldn't be as opposed to <wbr> if it didn't come with a lot of legacy baggage -- it's more or less the same as <br>

[22:12:00.0000] <karlcow>
http://www.quirksmode.org/oddsandends/wbr.html

[22:13:00.0000] <othermaciej>
what's the legacy baggage?

[22:13:01.0000] <othermaciej>
just the special interaction with white-space: nowrap?

[22:13:02.0000] <Hixie>
othermaciej: all the weird rules of how it works

[22:14:00.0000] <othermaciej>
maybe I need to reverse-engineer it more thoroughly to understand the rules and figure out if they are truly weird

[22:15:00.0000] <othermaciej>
noninteroperable legacy behavior would definitely be a justifiable reason to make it nonconforming, since there is a close (if not identical) substitute

[22:22:00.0000] <karlcow>
/me is exploring the unicode table on macosx and discovering cool spaces

[22:25:00.0000] <othermaciej>
heh

[22:25:01.0000] <othermaciej>
yeah

[22:25:02.0000] <othermaciej>
see

[22:25:03.0000] <othermaciej>
the best thing about <wbr> is you don't have to be a master of the ludicrous number of whitespace characters to use it

[22:25:04.0000] <othermaciej>
anyway I guess it's best to let it set

[22:29:00.0000] <othermaciej>
Hixie: zwnj doesn't seem to create a breaking opportunity (in case you were curious)

[22:29:01.0000] <Hixie>
ah ok

[22:33:00.0000] <othermaciej>
I wonder if it would be compatible enough to just make <wbr> render as a zero-width space

[22:44:00.0000] <ap>
othermaciej: the main thing to test would probably be interaction with RTL text

[22:44:01.0000] <othermaciej>
ap: would you expect it to be different in that regard than zero-width space?

[22:45:00.0000] <othermaciej>
the main oddity seems to be interaction with different CSS white-space modes

[22:45:01.0000] <othermaciej>
(haven't found a browser yet that treats it any different under <nobr> than it would in any other white-space: nowrap context)

[22:46:00.0000] <ap>
othermaciej: I don't know. But all these spaces are subtly different, so it would be a coincidence if wbr matched it

[22:47:00.0000] <othermaciej>
ok, I'll try RTL

[22:49:00.0000] <othermaciej>
ap: seems like the only difference under RTL is same as LTR - it breaks even in white-space: nowrap

[22:49:01.0000] <ap>
othermaciej: how does it affect directionality of neutral text?

[22:50:00.0000] <othermaciej>
ap: how can I test that?

[22:50:01.0000] <othermaciej>
what I'm using as a test case is this:

[22:50:02.0000] <othermaciej>
"TheQuickBrownFoxJumpedOverTheLazyDog.<wbr>TheQuickBrownFoxJumpedOverTheLazyDog.<wbr>"

[22:50:03.0000] <othermaciej>
(repeated many times)

[22:50:04.0000] <othermaciej>
it seems to affect the period the same way ZWSP does, as far as I can observe

[22:50:05.0000] <ap>
othermaciej: that

[22:51:00.0000] <ap>
's how I would test

[22:51:01.0000] <ap>
othermaciej: actually, I'd also try <p>.<wbr><RTL text></p>

[22:51:02.0000] <ap>
othermaciej: with no strong directionality characters on one side

[22:52:00.0000] <ap>
othermaciej: but again, I don't know RTL

[22:52:01.0000] <othermaciej>
I think that might be beyond my level of understanding

[23:41:00.0000] <JonathanNeal>
aloha

[23:53:00.0000] <hsivonen>
Hixie: do the fancy unicode tricks get out of the way in clipboard export in shipped browsers?

[23:53:01.0000] <Hixie>
nbsp does, at least

[23:53:02.0000] <hsivonen>
oh. I thought it didn't

[23:55:00.0000] <hsivonen>
for Farsi use cases, it would be weird not to export zwnj to the clipboard

[23:56:00.0000] <hsivonen>
I think we should allow wbr and nobr. we don't delegate br to the unicode layer, either

[23:56:01.0000] <Hixie>
i could buy <wbr> if it didn't have the weird behaviours

[23:56:02.0000] <Hixie>
but <nobr>? really?

[23:57:00.0000] <Hixie>
what's the use case?

[23:57:01.0000] <Hixie>
people use it far more inappropriately than appropriately

[23:57:02.0000] <Hixie>
i'm fine with having the few rare appropriate use cases use &nbsp; if only to discourage the use of <nobr>

[23:59:00.0000] <hsivonen>
Hixie: the use case is to avoid breaking at hyphens

[00:00:00.0000] <Hixie>
white-space: nowrap

[00:00:01.0000] <hsivonen>
i'm on a browsers without copy and paste, but google for jukka korpela nobr

[00:01:00.0000] <hsivonen>
Hixie: how is that plus a span with class more maintainable, compact or accessible

[00:01:01.0000] <Hixie>
why would you use a span and a class?

[00:01:02.0000] <Hixie>
just use whatever the current parent is

[00:03:00.0000] <Hixie>
anyway, bed time

[00:03:01.0000] <Hixie>
nn

[00:04:00.0000] <hsivonen>
Hixie: I guess one would need a study of the real-world usage of white-space: nowrap; to see what it gets applied to

[00:04:01.0000] <hsivonen>
nn

[01:55:00.0000] <othermaciej>
filing these bugs from the conformance study is making me learn more than I wanted to know about all sorts of stuff (now the tokenizer)

[01:55:01.0000] <othermaciej>
hsivonen: is tokenization inside attributes and inside text content similar enough to be useful right now?

[02:00:00.0000] <JonathanNeal>
After mapping out the default stylesheets for ie, I tried running one on Google Chrome.

[02:00:01.0000] <JonathanNeal>
Not only does it crash Chrome, it crashes all of Chrome's tabs.

[02:01:00.0000] <othermaciej>
JonathanNeal: tried running one of what?

[02:01:01.0000] <othermaciej>
/me wonders if that is a Chrome bug or a WebKit bug and whether it has been reported

[02:03:00.0000] <JonathanNeal>
Stranger, it only crashes when I run it on my localhost

[02:04:00.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/style.php?b=ie6

[02:06:00.0000] <JonathanNeal>
Here's a pic of the crash @ http://sandbox.thewikies.com/html5-browser-stylesheets/crash.png

[02:07:00.0000] <peol>
JonathanNeal: What version of Chrome/Chromium?

[02:07:01.0000] <JonathanNeal>
peol, 4.1.249.1042 (42199)

[02:09:00.0000] <peol>
JonathanNeal: Hmm, and it only crashes on localhost, is it consistent?

[02:09:01.0000] <JonathanNeal>
yeappers

[02:09:02.0000] <JonathanNeal>
only when I run the php file too.

[02:09:03.0000] <peol>
JonathanNeal: Heh, I wonder what could cause it to crash locally but not remotely =p

[02:10:00.0000] <JonathanNeal>
You can use http://sandbox.thewikies.com/html5-browser-stylesheets/style.php.txt to save the php file locally.

[02:13:00.0000] <peol>
JonathanNeal: It seems to do all right over here, both locally and remotely on 41671

[02:14:00.0000] <JonathanNeal>
Yea, it's very mysterious.

[02:14:01.0000] <JonathanNeal>
Nevertheless, I have some nifty ie6, ie7, ie8 default stylesheets in there now.

[02:15:00.0000] <peol>
JonathanNeal: http://sandbox.thewikies.com/html5-browser-stylesheets/elements.ie.js seems to be broken too, doubt it's relevant though :)

[02:15:01.0000] <peol>
JonathanNeal: Yeah, did you extract them manually?

[02:15:02.0000] <othermaciej>
JonathanNeal: does the whole browser go down?

[02:16:00.0000] <othermaciej>
JonathanNeal: you might be crashing the app process, rather than the WebKit process

[02:16:01.0000] <JonathanNeal>
peol, no, only partially.  IE will report the default styling of elements in some instances.  The rest took many, many tests.

[02:16:02.0000] <othermaciej>
JonathanNeal: yeah, you did, from that screenshot

[02:16:03.0000] <JonathanNeal>
Discovering when it was using pt vs px vs em.

[02:16:04.0000] <othermaciej>
fwiw that page opens fine in Safari, but I dunno what was up with your local copy

[02:17:00.0000] <peol>
othermaciej, JonathanNeal: There might be something to it then, maybe in a newer build (or win32 only), I'm running under linux here without crashes

[02:17:01.0000] <JonathanNeal>
othermaciej, i can run it locally in safari all day without a crash, in chrome it takes but one refresh

[02:17:02.0000] <othermaciej>
definitely a Chrome bug, not a WebKit bug, then

[02:20:00.0000] <peol>
JonathanNeal: I bet it's the empty <object type="application/x-shockwave-flash"> that crashes it :)

[02:20:01.0000] <JonathanNeal>
Lemme check.

[02:21:00.0000] <JonathanNeal>
Okay, crashed it, now removing the object

[02:21:01.0000] <JonathanNeal>
oof, still crashes

[02:21:02.0000] <peol>
JonathanNeal: Or perhaps the empty source element in <video>, stuff like that could likely cause a crash in a daily

[02:22:00.0000] <othermaciej>
JonathanNeal: I would guess it's the filename crashing it

[02:22:01.0000] <othermaciej>
JonathanNeal: try renaming the file or moving it to a different path

[02:22:02.0000] <JonathanNeal>
Yessir, that was it, the <video> source element.

[02:22:03.0000] <hsivonen>
othermaciej: named character references currently tokenize the same way in attribute values and in element content

[02:22:04.0000] <othermaciej>
(just trying to guess things that would take down the Chrome UI process)

[02:22:05.0000] <othermaciej>
hah, wow

[02:23:00.0000] <JonathanNeal>
Yes, and I add it back in and it crashes.

[02:23:01.0000] <othermaciej>
hsivonen: whoah, are they officially called "named character references"? cause that would be hella confusing

[02:23:02.0000] <othermaciej>
(what with the "NCR" abbreviation)

[02:24:00.0000] <hsivonen>
othermaciej: the thingies formerly known as entities are now named character references

[02:24:01.0000] <othermaciej>
hsivonen: entity tokenization is the one thing that is very slightly different in quoted attribute values: http://dev.w3.org/html5/spec/Overview.html#attribute-value-double-quoted-state

[02:24:02.0000] <othermaciej>
(vs http://dev.w3.org/html5/spec/Overview.html#data-state )

[02:24:03.0000] <peol>
JonathanNeal: Sweet, now we know how to crash chromium dailies :)

[02:25:00.0000] <JonathanNeal>
:) glad I could help in a way entirely unrelated to what I was trying to accomplish.

[02:26:00.0000] <hsivonen>
othermaciej: oops. sorry. I misread my code.

[02:26:01.0000] <othermaciej>
hsivonen: in fact, just making "=" be treated as an extra "additional allowed character" would do the trick

[02:26:02.0000] <othermaciej>
hsivonen: that being said, I amended the bug because I think my reasons to limit the request were misguided

[02:28:00.0000] <othermaciej>
hello Lachy 

[02:28:01.0000] <Lachy>
hello othermaciej 

[02:31:00.0000] <hsivonen>
othermaciej: the "additional allowed character" takes effect immediately after the ampersand

[02:32:00.0000] <hsivonen>
othermaciej: so it's too interesting

[02:32:01.0000] <othermaciej>
hsivonen: s/too interesting/not too interesting/ ?

[02:32:02.0000] <hsivonen>
othermaciej: oops. right

[02:32:03.0000] <hsivonen>
othermaciej: the actually interesting bit is this: "If the character reference is being consumed as part of an attribute, and the last character matched is not a U+003B SEMICOLON character (;), and the next character is in the range U+0030 DIGIT ZERO (0) to U+0039 DIGIT NINE (9), U+0041 LATIN CAPITAL LETTER A to U+005A LATIN CAPITAL LETTER Z, or U+0061 LATIN SMALL LETTER A to U+007A LATIN SMALL LETTER Z, then, for historical reasons, all the characte

[02:33:00.0000] <hsivonen>
which I was trying to locate by reading code, but failed to see even though it was in plain view in my code

[02:33:01.0000] <othermaciej>
hsivonen: got cut off

[02:33:02.0000] <hsivonen>
If the character reference is being consumed as part of an attribute, and the last character matched is not a U+003B SEMICOLON character (;), and the next character is in the range U+0030 DIGIT ZERO (0) to U+0039 DIGIT NINE (9), U+0041 LATIN CAPITAL LETTER A to U+005A LATIN CAPITAL LETTER Z

[02:34:00.0000] <hsivonen>
, or U+0061 LATIN SMALL LETTER A to U+007A LATIN SMALL LETTER Z, then, for historical reasons, all the characters that were matched after the U+0026 AMPERSAND character (&) must be unconsumed, and nothing is returned.

[02:34:01.0000] <othermaciej>
hsivonen: I guess it's still true that there is already a difference, I was just wrong that the additional allowed character isthe key one

[02:34:02.0000] <othermaciej>
I found it in the spec, no worries

[02:34:03.0000] <hsivonen>
othermaciej: yeah

[02:34:04.0000] <othermaciej>
I guess that is the list to add = to, as previously identified by zcorpan on the old thread on this I believe

[02:35:00.0000] <hsivonen>
putting = on that list would certainly remove a sharp edge from the language

[02:36:00.0000] <hsivonen>
do we have data gathered by intrumenting that particular point in the tokenizer?

[02:36:01.0000] <hsivonen>
instrumenting

[02:36:02.0000] <othermaciej>
not as far as I know

[02:37:00.0000] <othermaciej>
however, one of the emails I linked in the bug cites raw data

[02:45:00.0000] <hsivonen>
looks like the only real danger is indeed &amp=

[02:45:01.0000] <hsivonen>
it would be possible to special-case &amp=

[02:45:02.0000] <hsivonen>
even though I'm generally scared of heavy DWIM

[02:46:00.0000] <hsivonen>
ooh. crazy. I misread the data

[02:47:00.0000] <othermaciej>
I guess my bug is a dupe of http://www.w3.org/Bugs/Public/show_bug.cgi?id=9207

[02:47:01.0000] <hsivonen>
indeed they seem to want &amp= to mean &amp;amp=

[02:47:02.0000] <othermaciej>
Hixie did find at least one case where it seemed "&amp=" was meant to be "&amp;" [sic]

[02:49:00.0000] <zcorpan>
othermaciej: i've added a new comment to that bug

[02:49:01.0000] <othermaciej>
zcorpan: yeah, that's how I noticed

[02:49:02.0000] <othermaciej>
trying to decide if I should dupe mine to that

[02:49:03.0000] <hsivonen>
I was such a bad idea to use an SGML special character as a special character in URLs, too

[02:49:04.0000] <othermaciej>
my bug makes a narrower request and has potentially useful comments, but I could always copy the comments

[02:50:00.0000] <othermaciej>
or I could cross-reference them and let Hixie decide whether it is a dupe

[02:50:01.0000] <hsivonen>
btw, is what the spec now says only for IE-compat and there are other browsers that don't expand &copy= to the copyright sign?

[02:51:00.0000] <hsivonen>
old Gecko expands :-(

[02:52:00.0000] <hsivonen>
WebKit, too

[02:56:00.0000] <zcorpan>
going by train is great, because then i have time to analyze things like this

[02:58:00.0000] <othermaciej>
hsivonen: what the spec says now matches browsers better than the proposed changes, the motive for changing parsing is that it may fix more content than it breaks, in addition to removing a sharp edge

[03:03:00.0000] <zcorpan>
&reg_nr= will still have the sharp edge

[03:06:00.0000] <othermaciej>
unless we add _ to the list too

[03:09:00.0000] <hsivonen>
/me wonders how & was added to URLs

[03:09:01.0000] <zcorpan>
yes but that would break some pages

[03:09:02.0000] <hsivonen>
presumably it wasn't part of TimBL's original design, because isindex doesn't need it

[03:09:03.0000] <othermaciej>
probably added whenever forms were added

[03:10:00.0000] <othermaciej>
it does seem like it would have been wise to pick a different separator

[03:10:01.0000] <hsivonen>
who was the first to add forms? Lynx?

[03:11:00.0000] <hsivonen>
http://people.cc.ku.edu/~grobe/early-lynx.html pointsto Lou Montulli

[03:12:00.0000] <othermaciej>
forms existed in HTML 2

[03:12:01.0000] <othermaciej>
so at the latest by 1995

[03:12:02.0000] <othermaciej>
Lou Montulli is personally responsible for a surprising number of completely broken things

[03:12:03.0000] <hsivonen>
http://www.montulli.org/lou

[03:12:04.0000] <othermaciej>
(not necessarily saying forms is one of them, but that's the context where I usually hear his name)

[03:13:00.0000] <hsivonen>
forms go back to 1992: http://lists.w3.org/Archives/Public/www-talk/1992NovDec/0212.html

[03:13:01.0000] <othermaciej>
cookies, server push, the blink tag, and animated GIFs

[03:13:02.0000] <othermaciej>
we get to enjoy all those things on the web thanks to him

[03:14:00.0000] <othermaciej>
I guess someone else gave us the wonder of framesets

[03:15:00.0000] <othermaciej>
so is there a record there of who decided & should be the field separator in URL queries?

[03:16:00.0000] <hsivonen>
w3.org doesn't have archives from 1993 and 1994, so the thread goes into unarchived territory

[03:17:00.0000] <othermaciej>
I wonder if any of the folks who were around back then remember

[03:18:00.0000] <hsivonen>
hmm. looks like served admins have accidentally knocked my personal site off the Web over the weekend

[03:21:00.0000] <othermaciej>
at least in 1993, forms were not in the Internet Draft for HTML: http://www.w3.org/MarkUp/draft-ietf-iiir-html-01.txt

[03:22:00.0000] <othermaciej>
"HTML 3.0 did not succeed for several reasons. The draft was considered very large at 150 pages...."

[05:04:00.0000] <karlcow>
hsivonen: look at the bottom of http://lists.w3.org/Archives/Public/www-talk/ for archives in 1991 and 1992

[05:04:01.0000] <hsivonen>
karlcow: yeah, but no 1993 or 1994 there

[05:05:00.0000] <karlcow>
ah yeah just realized you were talking about 93 94

[05:05:01.0000] <hsivonen>
karlcow: the interesting thread started in Dec 1992

[05:05:02.0000] <karlcow>
check here

[05:06:00.0000] <karlcow>
http://www.w3.org/Style/History/www.eit.com/www.lists/

[05:07:00.0000] <zcorpan>
/me gets not found errors when trying to view emails

[05:07:01.0000] <karlcow>
there are also accessible there http://ksi.cpsc.ucalgary.ca/archives/WWW-TALK/

[05:08:00.0000] <zcorpan>
/me finds http://1997.webhistory.org/www.lists/www-talk.1993q2/

[05:08:01.0000] <karlcow>
and http://1997.webhistory.org/www.lists/

[05:09:00.0000] <zcorpan>
http://1997.webhistory.org/www.lists/www-talk.1993q2/0557.html

[05:13:00.0000] <zcorpan>
http://1997.webhistory.org/www.lists/www-talk.1993q3/0810.html

[05:13:01.0000] <zcorpan>
http://1997.webhistory.org/www.lists/www-talk.1993q3/0812.html aha!

[05:13:02.0000] <zcorpan>
so Marc Andreessen is to blame

[05:18:00.0000] <zcorpan>
/me tests applying padding and border on <svg> and gets different results in all browsers

[06:16:00.0000] <zcorpan>
does <header> imply <body> in ie9?

[06:27:00.0000] <Philip`>
zcorpan: I'm fairly sure it's always treated the same as any other unrecognised element in IE9, and I vaguely remember that unrecognised elements don't imply <body>

[06:29:00.0000] <zcorpan>
ok

[06:30:00.0000] <Philip`>
(If there's a test in html5lib then the output is like in http://philip.html5.org/misc/html5lib-runner-ie9.txt)

[06:35:00.0000] <zcorpan>
ie9's result for Test 105 of 112 in data/tests1.dat looks weird

[06:35:01.0000] <zcorpan>
two head elements, a frameset and a body?

[06:38:00.0000] <Philip`>
Maybe it's one head element in two locations

[06:39:00.0000] <Philip`>
(Not that that'd be any less weird)

[06:39:01.0000] <zcorpan>
no, the output looks different in such a case

[06:39:02.0000] <zcorpan>
i think

[06:40:00.0000] <zcorpan>
or maybe not

[06:40:01.0000] <zcorpan>
the script only keeps track of the parent

[07:21:00.0000] <Philip`>
I dislike how Launchpad returns obscure error messages whenever you try to interact with it, if you have Referer headers disabled

[08:02:00.0000] <zcorpan>
foo.svg#svgView(viewBox(...)) seems awesome for image sprites

[08:03:00.0000] <zcorpan>
http://www.w3.org/TR/SVG11/linking.html#IntroFragmentsViews

[08:03:01.0000] <zcorpan>
seems to work in opera and chrome but not firefox

[08:04:00.0000] <zcorpan>
e.g. http://dev.w3.org/SVG/profiles/1.1F2/test/images/linkingCircle-f.svg#svgView(viewBox(64,227,72,72))

[09:20:00.0000] <JonathanNeal>
Fun thing to do; make a <caption> in ie7, then alert(document.getElementsByTagName('caption')[0].currentStyle.padding);

[09:20:01.0000] <JonathanNeal>
"padding: 1px 300zzz 500zzz;"

[09:21:00.0000] <MikeSmith>
JonathanNeal: what effect does that have?

[09:21:01.0000] <miketaylr>
only IE7?

[09:21:02.0000] <JonathanNeal>
Well, they're sleepy pixels, of course!

[09:23:00.0000] <paul_irish>
JonathanNeal: did you find what from ie6's default css was crashing chrome?

[09:23:01.0000] <TabAtkins>
miketaylr: Yup, only ie7.

[09:23:02.0000] <miketaylr>
interesting.

[09:24:00.0000] <miketaylr>
s/interesting/wtf/

[09:24:01.0000] <JonathanNeal>
paul_irish, we did ... but it was being triggered by a bizarre combination that was ultimately the <video> element's empty <source>

[09:25:00.0000] <JonathanNeal>
I never noticed till I applied ie stylesheets, but that on a php page in localhost would cause Chrome to consistantly crash until you removed the <source> in <video>

[09:25:01.0000] <TabAtkins>
miketaylr: Check out http://www.xanthir.com/etc/default-stylesheets.php, created from JonathanNeal's data.

[09:25:02.0000] <JonathanNeal>
TabAtkins, I've been meaning to tell you there are updated js files to reference for the browsers.

[09:25:03.0000] <TabAtkins>
Excellent.  Gimme!

[09:25:04.0000] <miketaylr>
oh that looks great, TabAtkins

[09:25:05.0000] <miketaylr>
/me looks

[09:25:06.0000] <JonathanNeal>
Same place http://sandbox.thewikies.com/html5-browser-stylesheets/

[09:26:00.0000] <JonathanNeal>
/me is not spamming, just re-referencing.

[09:28:00.0000] <JonathanNeal>
Actually TabAtkins I'll cue you in a minute when they're the latest latest.

[09:39:00.0000] <zcorpan>
blockquote[type=cite] { ...? what's this (from gecko's style sheet)?

[09:47:00.0000] <JonathanNeal>
TabAtkins quit :(

[09:51:00.0000] <Philip`>
He pung out, to be precise

[09:53:00.0000] <Philip`>
And so did you :-(

[09:55:00.0000] <smaug___>
zcorpan: seems like that has been there for ages http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/layout/style/html.css&rev=3.216#90

[10:07:00.0000] <Philip`>
http://homepages.laas.fr/felix/sky2c-data/mail/msg00044.html

[10:07:01.0000] <Philip`>
http://grouper.ieee.org/groups/1619/email/msg02384.html

[10:07:02.0000] <Philip`>
Is it the mailing list software generating the <blockquote type=cite>, or the mail clients?

[10:08:00.0000] <Philip`>
http://emperor.tidbits.com/webx?14@@.3cbfbb6d/45 too

[10:09:00.0000] <zcorpan>
i guess thunderbird uses that

[10:10:00.0000] <AryehGregor>
Is it just me, or does IE8 in standards mode drop support for things that aren't specced, but are interoperably supported among all browsers including IE7?

[10:11:00.0000] <AryehGregor>
Like <wbr>, or id's starting with a number . . . (both now specced in HTML5, of course.)

[10:11:01.0000] <AryehGregor>
That seems like it's a bad idea for interop.

[10:13:00.0000] <Xanthir>
http://www.xanthir.com/etc/default-stylesheets.php  Newest version, with JonathanNeal's latest data.

[11:00:00.0000] <JonathanNeal>
Yay for Chrome.

[11:31:00.0000] <gsnedders>
/me wishes he had any clue about how TC-39 works

[11:33:00.0000] <othermaciej>
gsnedders: what clues do  you lack?

[11:33:01.0000] <othermaciej>
gsnedders: I know some medium amount of stuff about it

[11:33:02.0000] <gsnedders>
othermaciej: Through what seemingly arbitrary process do things get into the spec?

[11:34:00.0000] <gsnedders>
As far as I can tell, it's easier to get things in that have already shipped (the point of the JS extensions in Gecko?), yet somethings like accessors end up spec'd in a way that is different to what is already implemented

[11:35:00.0000] <othermaciej>
gsnedders: I think you are supposed to make a "strawman" on the wiki, and it can eventually be upgraded to a "proposal"

[11:35:01.0000] <othermaciej>
the formal decision-making part is the part I least understand

[11:35:02.0000] <othermaciej>
not sure if it is consensus or voting or something else

[11:37:00.0000] <gsnedders>
othermaciej: Oh, and do you have any idea how it was decided what from ES4 got into ES5?

[11:37:01.0000] <othermaciej>
gsnedders: almost nothing from ES4 got into ES5, as far as I know...

[11:37:02.0000] <gsnedders>
Because that seems again tangential to what was already shipped by browsers

[11:38:00.0000] <othermaciej>
ES5 is largely based on what used to be in ES3.1

[11:38:01.0000] <JonathanNeal>
Do you guys think the non-html5 elements listed in Rendering should be removed from the css presentational hints?

[11:38:02.0000] <othermaciej>
the only ES4 stuff that got in was stuff that was in bth

[11:38:03.0000] <gsnedders>
ES3.1 existed before ES4 was abandoned?

[11:40:00.0000] <JonathanNeal>
Also, instead of /* LTR-specific: use 'margin-right' for rtl elements */, could you [dir=rtl] ?

[11:42:00.0000] <AryehGregor>
"JonathanNeal> Do you guys think the non-html5 elements listed in Rendering should be removed from the css presentational hints?"  What do you mean?

[11:43:00.0000] <AryehGregor>
[dir=rtl] is not the same as RTL, direction is inherited.

[11:43:01.0000] <annevk>
yeah, you'd need :rtl

[11:43:02.0000] <annevk>
which we might introduce in Selectors 4

[11:44:00.0000] <JonathanNeal>
Regarding the first comment, @ http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#the-css-user-agent-style-sheet-and-presentational-hints elements like basefont, center, plaintext are mentioned.

[11:45:00.0000] <JonathanNeal>
I know it doesn't break anything, but it seems unnecessary to include them in this section.  What do you think?

[11:47:00.0000] <JonathanNeal>
Also, when it says "The article, aside, nav, and section elements are expected to affect the margins of h1 elements", would this also apply to h2-h6?

[11:49:00.0000] <Xanthir>
JonathanNeal: default-css is updated.

[11:49:01.0000] <JonathanNeal>
Because I noticed that it seems to shift the margin and font-size for h1 down the scale of the h1-h6's.

[11:49:02.0000] <JonathanNeal>
It would seem fitting that the others would follow suit, or am I mistaken?

[11:50:00.0000] <annevk>
since user agents need to support the elements as per the rendering section it makes sense to include them

[11:50:01.0000] <JonathanNeal>
TabAtkins, groovy.  I'll check it out now.

[11:51:00.0000] <JonathanNeal>
TabAtkins, did you use the latest js?  border-color needn't show up on anchors

[11:52:00.0000] <TabAtkins>
JonathanNeal: Yes.

[11:52:01.0000] <TabAtkins>
Cache issues?  Do you see the [dir=rtl] stuff at the beginning?

[11:52:02.0000] <JonathanNeal>
I discovered that a few of these colors are set by color: myColor;, so I check to see if they match and negate to mention them if so.

[11:53:00.0000] <JonathanNeal>
I did see that stuff, that was my interpretation of how the direction stuff would work, since I was testing by changing <html> to <html dir="rtl"> to see the affects in an unstyled ie.

[11:53:01.0000] <othermaciej>
gsnedders: yes

[11:53:02.0000] <TabAtkins>
(Sorry, my latency is high right now because I'm scribing for CSSWG)

[11:54:00.0000] <othermaciej>
gsnedders: before TC-39 settled on what is now ES5, there were two competing efforts - ES4 and ES3.1, driven by different subsets of the group

[11:55:00.0000] <JonathanNeal>
AryehGregor, annevk, should h2-h6 be affected by sectioning content elements like article, aside, nav, and section too?

[11:55:01.0000] <AryehGregor>
I'd assume so.  Why not?

[11:56:00.0000] <JonathanNeal>
Well, may I recommend the spec be changed to reflect this?

[11:59:00.0000] <AryehGregor>
Post to whatwg or Bugzilla if you have a spec change suggestion.

[11:59:01.0000] <AryehGregor>
I haven't looked at the sectioning stuff, no idea how it works or why.

[12:07:00.0000] <TabAtkins>
JonathanNeal: I can explain the sectioning stuff in a bit.

[12:31:00.0000] <MikeSmith>
I just accidentally typed the word "equacally" in a bugzilla comment

[12:31:01.0000] <MikeSmith>
I kind of like it

[12:31:02.0000] <MikeSmith>
/me googles

[12:32:00.0000] <MikeSmith>
almost nothing

[12:32:01.0000] <MikeSmith>
I should lay claim to this word

[12:32:02.0000] <MikeSmith>
I'm just wondering what is should mean

[12:33:00.0000] <MikeSmith>
suggestions for the meaning of "equacally" welcom

[12:33:01.0000] <MikeSmith>
*welcome

[12:35:00.0000] <Dashiva>
"In a manner indistinguishable from a duck"

[12:37:00.0000] <MikeSmith>
heh

[12:37:01.0000] <MikeSmith>
that would be "equackally" I guess

[12:42:00.0000] <MikeSmith>
adequacally

[12:44:00.0000] <othermaciej>
heh

[12:44:01.0000] <MikeSmith>
in other news and fwiw, I added direct per-element links from the h:tml doc back to the HTML5 spec

[12:44:02.0000] <MikeSmith>
http://dev.w3.org/html5/markup/elements.html#elements

[12:45:00.0000] <othermaciej>
sweet

[12:45:01.0000] <MikeSmith>
the ⓘ symbol

[12:45:02.0000] <annevk>
btw, why is it in alphabetical order?

[12:45:03.0000] <zcorpan>
/me didn't know there was an ⓘ symbol

[12:45:04.0000] <MikeSmith>
also, on each element page; e.g., http://dev.w3.org/html5/markup/dt.html

[12:46:00.0000] <MikeSmith>
annevk: in alphabetical order because that's the least arbitrary way of ordering them

[12:47:00.0000] <MikeSmith>
is there an order that would be more useful?

[12:47:01.0000] <annevk>
html5 spec-order?

[12:47:02.0000] <annevk>
i.e. grouped by topic

[12:48:00.0000] <annevk>
then you could also have an introductory page per topic i guess... i.e. for forms or tables

[12:48:01.0000] <zcorpan>
http://simon.html5.org/html5-elements uses spec order but has a button to change the order

[12:48:02.0000] <MikeSmith>
that's pretty cool

[12:49:00.0000] <JonathanNeal>
equacally: describing something identical or equivalent to a well balanced compound water

[12:50:00.0000] <MikeSmith>
/me considers how to provide switchable order in the TOC

[12:50:01.0000] <gsnedders>
zcorpan: There's everything in a circle symbol with the COMBINING CIRCLE AROUND character

[12:51:00.0000] <MikeSmith>
JonathanNeal: "compound water"?

[12:51:01.0000] <JonathanNeal>
TabAtkins, still interested in hearing about the "sectioning stuff", which I think refers to my request to change the language @ "affect the font size of h1 elements", right?

[12:51:02.0000] <MikeSmith>
I don't know whether the circled "i" is the best choice here but I couldn't figure out what would work better

[12:51:03.0000] <MikeSmith>
there unfortunately is no circled-question-mark character

[12:52:00.0000] <JonathanNeal>
MikeSmith, in inorganic chemistry, aqua is a compound water.

[12:52:01.0000] <zcorpan>
gsnedders: aha, cool

[12:52:02.0000] <gsnedders>
(I think that's the Unicode name, at least, but such a combing character certainly exists)

[12:52:03.0000] <TabAtkins>
JonathanNeal: Yes, still minuting, give me a bit for a lunch break.

[12:53:00.0000] <JonathanNeal>
:D of course of course

[12:53:01.0000] <zcorpan>
MikeSmith: preload is not a boolean attribute

[12:53:02.0000] <MikeSmith>
oh

[12:53:03.0000] <MikeSmith>
zcorpan: then I think we must have it wrong in the schema

[12:53:04.0000] <MikeSmith>
TabAtkins: thanks

[12:54:00.0000] <MikeSmith>
oops

[12:54:01.0000] <MikeSmith>
make that: JonathanNeal, thanks (about aqua)

[12:56:00.0000] <JonathanNeal>
TabAtkins, I'd like to use the scripty you built for the ie default css tables to make something for w3c-html5.css too :)

[12:57:00.0000] <TabAtkins>
JonathanNeal: I'll send it to you.  Simple php script.

[12:57:01.0000] <JonathanNeal>
Grooviness

[13:03:00.0000] <JonathanNeal>
Also, miketaylr may or may not steal my jokes...

[13:03:01.0000] <miketaylr>
WAT

[13:03:02.0000] <TabAtkins>
dsinger: We have agreement on *not* 2.

[13:03:03.0000] <TabAtkins>
argh, dammit

[13:04:00.0000] <miketaylr>
sorry, took the 10.6.3 update plunge...so your jokes are safe for now JonathanNeal

[13:05:00.0000] <JonathanNeal>
/me sends affirmation to miketaylr.

[13:10:00.0000] <AryehGregor>
othermaciej, for what it's worth, extending "don't cook a kid in its mother's milk" to "don't mix milk and meat" wouldn't be considered "building a fence around the Torah".  In that particular case, the rabbinic interpretation is traditionally assumed to have been actually given to Moses orally on Sinai, not actually of rabbinic origin.  An example would be the fact that I'm not supposed to own leavened goods right now, even though the Torah'

[13:10:01.0000] <AryehGregor>
s prohibition only begins tonight -- this way, I won't put it off till the last minute and accidentally violate the Torah prohibition.

[13:10:02.0000] <AryehGregor>
:)

[13:11:00.0000] <othermaciej>
AryehGregor: that's not what wikipedia said, but I admit that is not a reliable reference on religious tradition

[13:11:01.0000] <AryehGregor>
What article?

[13:12:00.0000] <othermaciej>
I believe some aspects of that prohibition, such as how many hours you have to wait after eating hard cheese, or the exclusion on chicken, are in the "building a wall" category

[13:12:01.0000] <AryehGregor>
Yes, that's correct.

[13:12:02.0000] <othermaciej>
(in meeting right now)

[13:12:03.0000] <AryehGregor>
There are rabbinic prohibitions surrounding nearly every Torah prohibition.

[13:12:04.0000] <AryehGregor>
/me wonders how othermaciej came to know of this

[13:12:05.0000] <Philip`>
I hope the Torah doesn't have any prohibitions related to chocolate biscuits

[13:12:06.0000] <othermaciej>
why would a culturally literate person not know these kinds of things?

[13:12:07.0000] <othermaciej>
I think what I read was http://en.wikipedia.org/wiki/Milk_and_meat_in_Jewish_law

[13:12:08.0000] <AryehGregor>
othermaciej, very few people know things in this depth.

[13:13:00.0000] <AryehGregor>
Philip`, not any in the Torah itself that I can think of.  Of course, non-Jews don't have to follow the Torah anyway, so I'm pretty sure you're safe.

[13:14:00.0000] <othermaciej>
I am technically Jewish by matrilineal descent, but not observant or even brought up in the tradition

[13:14:01.0000] <AryehGregor>
Oh, really?  Interesting.

[13:14:02.0000] <othermaciej>
however, this topic is one I have been at various times curious about

[13:14:03.0000] <AryehGregor>
I like this quote: HTML 5 intends to make all markup interoperable. One can argue that every markup is willfully used. Following that path, validators/conformance checkers are completely unnecessary. HTML 5 validator could respond to every document with "Congratulations! Your document is... meh, whatever!"

[13:14:04.0000] <othermaciej>
I have also studied catholic dogmas to some extent

[13:15:00.0000] <qFox>
hi

[13:16:00.0000] <qFox>
video has no getImageData, does it?

[13:16:01.0000] <qFox>
like canvas does...

[13:16:02.0000] <gsnedders>
qFox: No, but you can write a video to a canvas

[13:16:03.0000] <Philip`>
qFox: No, but you can do canvas.drawImage(video) and then canvas.getImageData

[13:16:04.0000] <qFox>
yeah

[13:16:05.0000] <qFox>
my question was just, why doesn't video have or get it too?

[13:16:06.0000] <Philip`>
/me curses gsnedders

[13:16:07.0000] <qFox>
because now you have to paint it to a canvas first

[13:16:08.0000] <gsnedders>
/me beat Philip`!

[13:17:00.0000] <qFox>
unneccessary overhead for an action that'll probably done more than one cares for

[13:17:01.0000] <qFox>
+be

[13:17:02.0000] <qFox>
for any kind of frame processing, this overhead can add up. especially when ticks count

[13:18:00.0000] <qFox>
so i'm just wondering whether this has been brought up at all, and if so, what hte reasoning was behind it

[13:19:00.0000] <Philip`>
qFox: It seems likely that any processing you perform on the image data would be much more expensive than an extra blit onto a canvas, so it's probably not worth the specification/implementation complexity of adding API directly to <video> to optimise that step

[13:19:01.0000] <qFox>
ah well, but is it? i imagine that a blit is costly

[13:20:00.0000] <qFox>
and if the browser can easily blit video, i imagine that a simple getImageData on the video element shouldnt be much of a problem, at that stage

[13:20:01.0000] <qFox>
(or for any element, for that matter ;)

[13:21:00.0000] <qFox>
it just seems silly to have to create bogus canvas elements just to copy the video to get its pixels

[13:21:01.0000] <Philip`>
If you want to convince people that it's worthwhile on the basis of performance, you'd probably have to provide some profiling data that shows it's a real problem (and that it can't be fixed simply by optimising the implementation of drawImage(video))

[13:22:00.0000] <qFox>
i guess this is where i lack the knowledge of implementation details of browsers. but is it really such a giant step to implement?

[13:22:01.0000] <Philip`>
(Maybe the data will indicate that it is worthwhile - I don't really have any idea)

[13:22:02.0000] <Philip`>
It's probably just a little step, but there's a million equally little steps that could be implemented too

[13:22:03.0000] <qFox>
(or rather, my lack of that knowledge comes into play..)

[13:22:04.0000] <gsnedders>
It's not the implementation difficulty that's the concern

[13:23:00.0000] <Philip`>
e.g. someone might want getImageData on <img>, and on <svg>, and might want toDataURL on all those things, and maybe they'll want to be able to draw a bit onto it too

[13:23:01.0000] <qFox>
ah, well, if you're offering.. :p

[13:23:02.0000] <qFox>
yeah i guess i understand

[13:24:00.0000] <qFox>
i'll have to create some tests then

[13:24:01.0000] <Philip`>
The current approach seems to be to try to minimise the API surface by having all the pixel-related operations exposed by <canvas>, which keeps everything relatively straightforward

[13:24:02.0000] <qFox>
luckily it was an area i wanted to focus on anyways (benchmarking)

[13:24:03.0000] <qFox>
okay

[13:24:04.0000] <qFox>
that's the explanation i was kind of expecting

[13:27:00.0000] <Philip`>
Maybe the best way to optimise video processing is to use WebGL fragment shaders

[13:27:01.0000] <Philip`>
rather than trying to get the data into JS faster

[13:27:02.0000] <qFox>
how well does firefox implement those in 3.6? not at all, right?

[13:28:00.0000] <qFox>
well, from a complexity perspective you're correct. the paints aren't the main concern. but it is one obvious step that most likely to be expensive

[13:28:01.0000] <Philip`>
It's unstable and only for experimental usage at the moment

[13:28:02.0000] <Philip`>
(Not sure which browsers (if any) have shipped experimental support yet)

[13:29:00.0000] <qFox>
i believe, not sure, that 3.7a has some support? or was it webkit...

[13:29:01.0000] <Philip`>
but it'll be available sooner than <video>.getImageData would be :-)

[13:29:02.0000] <qFox>
hehe

[13:29:03.0000] <qFox>
ahwell. no matter

[14:00:00.0000] <AryehGregor>
You know, threaded e-mail clients can get really confusing if people are talking about related things in different threads and you're reading through several days later.

[14:01:00.0000] <Dashiva>
Like talking about h264 in a thread about window.print()?

[14:04:00.0000] <jgraham>
gsnedders: I'm not sure many things made the ES5 spec had previous implementations

[14:10:00.0000] <JonathanNeal>
TabAtkins, how we doin'?

[14:12:00.0000] <eighty4>
isn't there any background-opacity setting for image based backgrounds?

[14:12:01.0000] <eighty4>
surely there must be?

[14:12:02.0000] <eighty4>
><

[14:14:00.0000] <Hixie>
someone just asked me if i could reply to some questions about html5

[14:14:01.0000] <Hixie>
they want to phone me (i have no phone)

[14:14:02.0000] <Hixie>
they kindly provided the questions... in msword format

[14:14:03.0000] <Hixie>
and they want the reply by march 15th

[14:15:00.0000] <Hixie>
there are so many things wrong with this i don't even know where to begin

[14:16:00.0000] <MikeSmith>
Hixie: send your replies as a TeX file

[14:16:01.0000] <zcorpan>
Hixie: which year?

[14:16:02.0000] <Hixie>
no year specified

[14:17:00.0000] <Hixie>
in fact it just said "3/15" (mail was dated today)

[14:17:01.0000] <zcorpan>
well then send the replies 2022

[14:17:02.0000] <AryehGregor>
Hixie, I guess that's what really happens when journalists say "Mr. Smith did not respond to our request for comment by press time."

[14:17:03.0000] <Hixie>
the date honestly was the least of the problems in that e-mail

[14:17:04.0000] <Hixie>
sending me questions in some microsoft format was far worse

[14:18:00.0000] <zcorpan>
Hixie: i think there's some Google product that can grok microsoft formats

[14:18:01.0000] <AryehGregor>
Normal people probably don't even realize that .doc is a proprietary format.

[14:18:02.0000] <AryehGregor>
Why would they?

[14:18:03.0000] <Philip`>
'strings' can read .doc files

[14:18:04.0000] <Philip`>
.docx is much harder :-(

[14:18:05.0000] <Hixie>
zcorpan: i can grok plain text

[14:18:06.0000] <MikeSmith>
I think normal people don't even have much awareness of what "proprietary format" even means

[14:19:00.0000] <AryehGregor>
Yes, that too.

[14:19:01.0000] <franksalim>
I think normal people don't have any awareness of what "format" means

[14:19:02.0000] <AryehGregor>
. . . possibly, yeah.

[14:19:03.0000] <zcorpan>
i think normal people don't have any awareness

[14:20:00.0000] <MikeSmith>
Hixie: just reply with a message saying, "Here are some clarifying questions I have based on your request.." and then just write up append some semi-random questions

[14:20:01.0000] <zcorpan>
quote a random selection from the faq

[14:21:00.0000] <MikeSmith>
yeah

[14:21:01.0000] <MikeSmith>
you could write an e-mail bot to do it for you

[14:21:02.0000] <AryehGregor>
Philip`, strings '/home/aryeh/Documents/school/ccny/ENGL 10100 Essay 3.2a.doc' | less gives me output like: hZID?5=XwaWY8.@DJ]DCA96/+*/32;@DC

[14:21:03.0000] <Hixie>
/me often copies and pastes from replies he sent to other journalists :-P

[14:22:00.0000] <Philip`>
AryehGregor: You have to skip that bit and find the part where it stores chunks of text

[14:22:01.0000] <MikeSmith>
or just reply with, "Please direct your questions to <name>, who I prefer to vet all requests of this type", and give them somebody else's name and e-mail address

[14:22:02.0000] <Philip`>
If you edited the file a lot then I suppose they might be a bit muddled up

[14:23:00.0000] <AryehGregor>
Philip`, there are 1456 lines and I looked at all of them.

[14:24:00.0000] <Philip`>
Maybe you need 'strings -el'?

[14:25:00.0000] <Philip`>
Or maybe they changed the file format since I last looked at it, which was probably last millennium

[14:26:00.0000] <AryehGregor>
-el works, nice.

[14:27:00.0000] <AryehGregor>
I should use strings more.

[14:30:00.0000] <zcorpan>
MikeSmith: should i file a bug about preload?

[14:31:00.0000] <MikeSmith>
zcorpan: nope, I filed one already

[14:31:01.0000] <MikeSmith>
and fixed it

[14:31:02.0000] <MikeSmith>
in the schema

[14:31:03.0000] <zcorpan>
k

[14:31:04.0000] <MikeSmith>
I'm fixing it in the markup doc now

[14:31:05.0000] <MikeSmith>
thanks for catching it

[14:43:00.0000] <TabAtkins>
JonathanNeal: Back!  And not scribing!

[14:43:01.0000] <JonathanNeal>
Yay!

[14:43:02.0000] <TabAtkins>
So I'm still listening to the meeting, but I can actually concentrate on something else if necessary.

[14:43:03.0000] <TabAtkins>
So, what's the question?

[14:43:04.0000] <annevk>
Hixie, oh I got that email too...

[14:43:05.0000] <annevk>
Hixie, mine missed the attachment

[14:44:00.0000] <annevk>
(or maybe my mail client strips formats I dislike automatically, I wouldn't know)

[14:44:01.0000] <Hixie>
heh

[14:49:00.0000] <zcorpan>
/me looks at http://s.validator.nu/html5/media.rnc and gets confused

[14:50:00.0000] <zcorpan>
MikeSmith: that's the url the UI claims to use but doesn't match blackbox testing

[14:50:01.0000] <MikeSmith>
zcorpan: yeah, that's due to some build hackery I need to revert

[14:51:00.0000] <zcorpan>
MikeSmith: where can i see the schema that's actually used?

[14:51:01.0000] <MikeSmith>
I had thought http://s.validator.nu/html5/media-vnu.rnc would work

[14:51:02.0000] <MikeSmith>
but it appears not to

[14:52:00.0000] <MikeSmith>
it may be that it's not accessible except by v.nu itself

[14:53:00.0000] <MikeSmith>
fwiw, the reason I did this was in an attempt to improve reporting for required-but-missing attributes

[14:53:01.0000] <MikeSmith>
but it was a misguided approach

[14:54:00.0000] <MikeSmith>
and what we need to do instead is to patch jing

[14:54:01.0000] <zcorpan>
the schema on s.v.nu seems to be from 2008

[14:54:02.0000] <MikeSmith>
I guess the files at those URLs are not relevant to actual v.nu behavior

[14:55:00.0000] <MikeSmith>
v.nu uses an internal resolver to de-reference them to the schema it uses internally

[14:56:00.0000] <JonathanNeal_>
TabAtkins, you were going to tell me about selectors.

[14:57:00.0000] <TabAtkins>
I was?  I just saw a little bit of what you were talking about, and noted that no one was able to answer you properly.

[14:57:01.0000] <TabAtkins>
Could you reiterate what the issue was?

[14:57:02.0000] <JonathanNeal>
In the rendering section of the HTML5 spec, you don't mention that h2-h6 should also have their font-size and margin adjusted, you were going to tell me why.

[14:57:03.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-browser-stylesheets/table.html --- also, how do you feel about this formatting for the big browser css chart?

[14:59:00.0000] <TabAtkins>
That formatting seems cool.

[14:59:01.0000] <TabAtkins>
Re: h2-h6, they *should*, ideally, have their font-size and margin adjusted.  Problem is that doing it right is damn near impossible to express in CSS.

[14:59:02.0000] <Nick_G>
we need a CL wit MC and SC, Category List with Main Categorys and Sub Categories \o/

[15:00:00.0000] <jgraham>
AryehGregor: What's wrong with <strong style="color:red">. That is clearer and easier than using <strong> together with <font> for the style

[15:01:00.0000] <JonathanNeal>
TabAtkins, yes, but shouldn't it at least mention in the spec that the other elements will be affected?

[15:01:01.0000] <jgraham>
Which suggests that @style is less bad than <font> (because it combines more naturally with also picking the right markup)

[15:01:02.0000] <TabAtkins>
JonathanNeal: Yeah, it probably should.

[15:02:00.0000] <jgraham>
And a validator could chose to warn about <div style> or <span style>

[15:02:01.0000] <TabAtkins>
Ideally, what it would be able to say is just ":heading(1) { foo }, :heading(2) { foo }, etc.", but we can't do that.

[15:02:02.0000] <zcorpan>
that'd lead to people use other elements instead while still not caring about semantics or correctness

[15:04:00.0000] <jgraham>
True

[15:05:00.0000] <annevk>
TabAtkins, we can once Selectors 4 introduces them and HTML5 defines their semantics

[15:06:00.0000] <annevk>
(and then uses them to define their initial style rules)

[15:06:01.0000] <TabAtkins>
annevk: Yup.

[15:09:00.0000] <zcorpan>
/me wonders how to access complete.html when whatwg.org is out of reach for him

[15:11:00.0000] <JonathanNeal>
And the second question I had before was, why show the rendering guidelines for elements not in the html5 spec?

[15:14:00.0000] <jgraham>
Because interoperable behaviour from UAs is still needed

[15:18:00.0000] <JonathanNeal>
TabAtkins, you were gonna share that stylesheet php?

[15:18:01.0000] <TabAtkins>
JonathanNeal: What's your email?

[15:20:00.0000] <JonathanNeal>
jonathan t neal at gmail dot com

[15:29:00.0000] <TabAtkins>
JonathanNeal: Sent.

[15:29:01.0000] <TabAtkins>
I presume those spaces were to be removed?

[15:31:00.0000] <JonathanNeal>
yes and the at becomes an @

[15:32:00.0000] <JonathanNeal>
and the dot becomes .... the declaration of independence.

[15:33:00.0000] <zcorpan>
what does com become?

[15:35:00.0000] <JonathanNeal>
zcorpan, you don't even wanna know.

[15:39:00.0000] <TabAtkins>
zcorpan: Council Of Magisters.

[15:43:00.0000] <krijnh>
Did somebody clever in here already come up with a solution for the new elements in IE wrt print stylesheets?

[15:45:00.0000] <TabAtkins>
Does the js hack not work for print stylesheets?

[15:45:01.0000] <paul_irish>
miketaylr was working on one earlier but was defeated.

[15:46:00.0000] <krijnh>
TabAtkins: nope

[15:46:01.0000] <TabAtkins>
hrm.

[15:46:02.0000] <krijnh>
Hacking around with expression: doesn't seem to work either

[15:48:00.0000] <paul_irish>
i can tell you JonathanNeal found that documentfragments that contain the new elements need the hack applied to them as well.

[15:48:01.0000] <JonathanNeal>
it's true

[15:50:00.0000] <TabAtkins>
Yeah, and I know that the hack doesn't work on innerHTML'd content.

[15:51:00.0000] <TabAtkins>
(or maybe it needs to be reapplied?)

[15:51:01.0000] <JonathanNeal>
TabAtkins, I tried swapping the json with a file_get_contents inside the json_decode, but it returns null.  I read it could be the line returns, so I removed them, but it still returns null

[15:51:02.0000] <TabAtkins>
Your latest files have a ; at their end.  You need to remove that.

[15:51:03.0000] <JonathanNeal>
TabAtkins, it does, as long as it is writing innerHTML to a pre-hacked document.

[15:52:00.0000] <JonathanNeal>
Soweet, that worked.

[15:52:01.0000] <TabAtkins>
I removed the linebreaks just to make it easier to read.  It's not necessary to remove them.

[15:52:02.0000] <JonathanNeal>
Yes, but $values->ie6 = json_decode(rawurldecode(preg_replace('/\n|\r|\t/', '', file_get_contents('ie-6.js')))); reads really easily for me.

[15:52:03.0000] <TabAtkins>
Well, you can drop the preg_replace.

[16:11:00.0000] <AryehGregor>
jgraham, <strong style="color: red"> is preferable to <strong><font color=red>.  However, it is only aesthetically and pedagogically preferable, and I don't think that's good enough reason to say the latter is non-conforming.  A warning that there's a better way to do it would be appropriate, as I say.

[16:15:00.0000] <AryehGregor>
Of course, you can always debate what should be author conformance requirements and what should be up to validator authors.  But that doesn't excuse inconsistency in the spec -- and IMO, the treatment of style="" vs. other inline presentational markup is grossly inconsistent.

[16:16:00.0000] <AryehGregor>
style="" is clearly not *that* superior to other presentational markup -- empirical evidence strongly suggests that authors believe some other elements/attributes are better (or at least as good) in some cases.

[16:16:01.0000] <Hixie>
style="" is only in the spec because of the ridiculous amount of complains we got when we removed it

[16:17:00.0000] <Hixie>
complaints, even

[16:17:01.0000] <Hixie>
and <font> is out because people complained when it was in

[16:17:02.0000] <Hixie>
originally i wanted to have just <font style=""> and only allow it for wysiwyg editors

[16:18:00.0000] <AryehGregor>
Well, now I'm complaining that it's out!  Maybe if you make it obsolete but conforming you'll generate the fewest complaints.  :)

[16:18:01.0000] <AryehGregor>
Realistically, inline style is useful sometimes.  Sometimes you only want to use a style once.

[16:18:02.0000] <Hixie>
right, that's why we have style=""

[16:18:03.0000] <paul_irish>
or animation.

[16:18:04.0000] <AryehGregor>
Also, depending on your workflow it might just make more sense -- e.g., if you use a template-based HTML processor, then it might be just as easy to maintain, and more convenient in other ways.

[16:18:05.0000] <AryehGregor>
So I don't get all the hate.

[16:19:00.0000] <AryehGregor>
At worst, it hurts the author who uses it.  The thing to object to is non-semantic markup, not inline style.

[16:19:01.0000] <AryehGregor>
The two issues seem to be conflated.  You can use semantic and presentational markup together.  External CSS is no less presentational than inline CSS.

[16:19:02.0000] <Hixie>
the hate is because while there are a few authors who are competent enough to use it, the overwhelming majority take it as a sign that html is a presentation language

[16:20:00.0000] <Hixie>
same reason that we disallow <b><i></b></i> -- sure, some experts can use it correctly

[16:20:01.0000] <Hixie>
most can't

[16:20:02.0000] <AryehGregor>
Any HTML author these days knows that CSS is necessary for presentation.  You can't do even basic layouts without it, unless you use tables, in which case you are freaking insane.

[16:21:00.0000] <AryehGregor>
I'm pretty sure that almost no sites made in the last ten years don't use CSS.

[16:21:01.0000] <Hixie>
"any" dramatically overstates your case

[16:21:02.0000] <Hixie>
imho

[16:22:00.0000] <AryehGregor>
I usually try to be very careful not to overestimate the savviness of normal authors, but I'm pretty sure that even authors with extremely basic HTML knowledge understand that most styling is only possible with CSS.

[16:22:01.0000] <AryehGregor>
HTML styling is extremely limited, because no new features are being added to it.

[16:22:02.0000] <Hixie>
well if it's limited, why bother keeping any of it?

[16:22:03.0000] <Hixie>
i don't understand the use case for <font>, <big>, bgcolor="", etc

[16:23:00.0000] <AryehGregor>
Because existing pages use them, and it's annoying to say that to be valid the author has to exert significant effort that has no user-visible effect.

[16:23:01.0000] <AryehGregor>
Like xmlns="" or <img /> or whatever.

[16:24:00.0000] <AryehGregor>
Doesn't hurt anything much, compared to equivalent inline style.  Except pedagogically, so a warning is appropriate for those who care, but an error will just make people not want to validate.

[16:27:00.0000] <TabAtkins>
This can be solved equally well by just having validators separate errors into importance groups.

[16:27:01.0000] <Hixie>
it's exactly those existing uses whose authors are clearly not already using css

[16:28:00.0000] <Hixie>
and those pages are typically an accessibility nightmare

[16:28:01.0000] <Hixie>
i really see no value in supporting those pages as is

[16:28:02.0000] <Hixie>
afk bbiab

[16:28:03.0000] <AryehGregor>
TabAtkins, that's what warning vs. error *is*.

[16:28:04.0000] <TabAtkins>
Sure.  So what change is needed?

[16:29:00.0000] <AryehGregor>
TabAtkins, making legacy presentational stuff a warning instead of an error.

[16:29:01.0000] <AryehGregor>
Hixie, you seem to be arguing that the markup itself isn't bad, it's just correlated strongly with bad stuff, so we should ban the detectable bad stuff so that the undetectable bad stuff is probably flagged too.  But that seems like a weak argument.  Correlations like this tend to disappear when you push at them; the best we could hope for from banning presentational HTML is pages switching to CSS but still being inaccessible.

[16:30:00.0000] <AryehGregor>
There's no reason to think people will make their pages accessible if they switch to CSS, they'll just resort to divitis.

[16:30:01.0000] <TabAtkins>
Do you think that, were an inaccessible presentational-using site to upgrade, they'll just use only <div style> and <span style>?

[16:30:02.0000] <roc>
that's probably the cheapest option, so yes

[16:31:00.0000] <AryehGregor>
Either that, or they'll switch to divs and spans with classes.

[16:31:01.0000] <AryehGregor>
Which still doesn't help.

[16:31:02.0000] <AryehGregor>
Or they'll use tags like <strong> but misuse them.

[16:31:03.0000] <TabAtkins>
That last one is standard wysiwyg practice, anyway.

[16:31:04.0000] <AryehGregor>
Like they'll just replace <strike> by <del> everywhere.

[16:31:05.0000] <TabAtkins>
Also misusing <strong>.

[16:31:06.0000] <AryehGregor>
Yep.

[16:32:00.0000] <AryehGregor>
You're not going to fix accessibility problems by mandating a logically unrelated thing, even if presently that happens to be correlated.

[16:32:01.0000] <AryehGregor>
Most authors will do accessibility when either it's effortless, or there's immediate benefit to them.

[16:32:02.0000] <AryehGregor>
The latter is unlikely, and the former is very hard to engineer.

[16:33:00.0000] <TabAtkins>
I think that @style has an intrinsic benefit over the equivalent presentational markup, which I brought up on the list.

[16:33:01.0000] <TabAtkins>
Namely, that it uses CSS, which at least *exposed* that level of authors to CSS.

[16:33:02.0000] <TabAtkins>
s/exposed/exposes/

[16:33:03.0000] <AryehGregor>
Yes, that's absolutely a benefit.  A warning is enough to give that benefit, provided authors actually validate at all.

[16:34:00.0000] <TabAtkins>
Do you think that that level of author validates?

[16:34:01.0000] <AryehGregor>
No.  So it doesn't matter what the spec says at all.

[16:34:02.0000] <AryehGregor>
They'll ignore all author conformance requirements.

[16:34:03.0000] <TabAtkins>
Sure, so let's do what we want and strike them for fun!

[16:34:04.0000] <TabAtkins>
(And profit, if possible.)

[16:34:05.0000] <AryehGregor>
. . .

[16:35:00.0000] <AryehGregor>
What about authors who know about validation and might validate if it were useful, but not if it's spammed with noise about things that they know aren't actually a problem?

[16:35:01.0000] <TabAtkins>
I challenge you to find a significant set of authors that match what you just described.

[16:35:02.0000] <AryehGregor>
What about authors who care about validation, but not enough to rewrite entire bodies of legacy content?

[16:36:00.0000] <TabAtkins>
Wait, so for that last one, are you recommending we let their legacy content validate just to make them happy?

[16:36:01.0000] <AryehGregor>
That's what we do in lots of cases, isn't it?

[16:36:02.0000] <AryehGregor>
xmlns="", <img />, etc.

[16:36:03.0000] <AryehGregor>
It's basically harmless.

[16:36:04.0000] <annevk>
it's different from misuse though

[16:36:05.0000] <TabAtkins>
That really is harmless, though.  <img /> doesn't point to likely accessibility problems as well.

[16:37:00.0000] <Philip`>
It seems more useful to focus on improving the creation of new content, not the upgrading of legacy content, because there's a finite amount of legacy content and an infinite amount of future content

[16:37:01.0000] <AryehGregor>
TabAtkins, so are you really arguing we should ban markup because it's *correlated* with bad practices?

[16:37:02.0000] <Philip`>
(and most legacy content never gets upgraded anyway)

[16:37:03.0000] <TabAtkins>
Not by itself, no.  But that's a contributing reason, yes.

[16:37:04.0000] <AryehGregor>
annevk, <font> isn't misuse, it was allowed by a decade of HTML standards.

[16:37:05.0000] <AryehGregor>
Philip`, that argues for validators not complaining too loudly about legacy markup, or what?

[16:38:00.0000] <AryehGregor>
TabAtkins, why is it even a contributing reason?

[16:38:01.0000] <TabAtkins>
AryehGregor: The latter.  We should gear complaints toward encouraging better markup for new content.

[16:38:02.0000] <Philip`>
AryehGregor: It argues for designing validators to help authors who are writing new content, not to help authors who are checking or upgrading old content

[16:38:03.0000] <JonathanNeal>
TabAtkins, http://sandbox.thewikies.com/html5-browser-stylesheets/table.php

[16:38:04.0000] <TabAtkins>
TabAtkins: Because it correlates with bad practices?  I'm not sure how I can explain that further.

[16:39:00.0000] <AryehGregor>
Philip`, it's not always so easy to separate.  Content is often partly new and partly old.

[16:39:01.0000] <AryehGregor>
In MediaWiki's case, for instance, we can't ban presentational stuff in new pages.  That breaks copy-paste, etc.

[16:39:02.0000] <Philip`>
TabAtkins: Talking to yourself?

[16:39:03.0000] <TabAtkins>
JonathanNeal: darker table-cell borders, please.  I can't see the bordrs currently.

[16:39:04.0000] <JonathanNeal>
ok

[16:39:05.0000] <TabAtkins>
TabAtkins: Yes.

[16:42:00.0000] <annevk>
AryehGregor, a decade of wrong

[16:43:00.0000] <annevk>
AryehGregor, also, transitional meant it was going out

[16:52:00.0000] <JonathanNeal>
It's darker now, TabAtkins 

[16:52:01.0000] <TabAtkins>
JonathanNeal: Is it intended that <head> and <ruby> don't have cells for ie7 and ie8?

[16:53:00.0000] <JonathanNeal>
That's actually a bug

[16:53:01.0000] <JonathanNeal>
only ie8 should have the styles.

[16:53:02.0000] <TabAtkins>
Also: something's wrong with your coalescing code.  IE8 is never coalescing with ie6 and 7, even when it has the same styles.  See <strong>, frex.

[16:54:00.0000] <TabAtkins>
Well, sometimes ie8 coalesces with ie7.  But never with ie6 for some reason.

[16:56:00.0000] <JonathanNeal>
TabAtkins, fixed.

[16:56:01.0000] <TabAtkins>
head/ruby still broken, but coalescing good.


2010-03-30
[17:11:00.0000] <JonathanNeal>
TabAtkins, fixed.

[17:11:01.0000] <TabAtkins>
Cool!

[17:20:00.0000] <JonathanNeal>
Other suggestions?

[17:23:00.0000] <TabAtkins>
/me is minuting again.

[18:49:00.0000] <Hixie>
AryehGregor: i wanted to ban <div>, too.

[19:51:00.0000] <Hixie>
TabAtkins: jamesr was looking for you in #webkit. You might want to consider joining #chromium too.

[19:52:00.0000] <til>
hixie -  i just got your email re http://www.w3.org/Bugs/Public/show_bug.cgi?id=9160

[19:52:01.0000] <TabAtkins>
Hixie: Thanks for the heads up.  Joined now.

[19:52:02.0000] <til>
Hixie: what about a multiple choice survey question (check all that apply)

[19:52:03.0000] <til>
that's the situation i'm interested in

[19:53:00.0000] <Hixie>
what if none apply?

[19:53:01.0000] <Hixie>
seems very dodgy to me to use checkboxes in that way

[19:55:00.0000] <til>
hixie: i'm a supplier of survey systems for restaurants

[19:55:01.0000] <til>
there are times when there's a multiple choice non-exclusive question that is required

[19:56:00.0000] <Hixie>
like what?

[19:56:01.0000] <til>
i.e. what keeps you coming back? after someone has already said they dine frequently

[19:56:02.0000] <Hixie>
that's exactly the kind of question where i'd hate for hte browser to force me to pick at least one

[19:56:03.0000] <til>
if the client insists that a question be required, i have to make it required

[19:56:04.0000] <til>
i can tell them why it's not worth requiring

[19:56:05.0000] <til>
but they pay for me

[19:57:00.0000] <Hixie>
so much so that for similar questions in the past i've gone in with a DOM inspector and killed the JS that forces me to do it, and submitted it that way

[19:57:01.0000] <til>
i can force you to select one if it's mutually-exclusive

[19:57:02.0000] <til>
i should be able to do the same for multiple-select

[19:57:03.0000] <til>
Hixie: this isn't about form UI - it's about the API

[19:58:00.0000] <til>
i'm not saying that i like required survey questions

[19:58:01.0000] <Hixie>
the API is about form UI

[19:58:02.0000] <Hixie>
i don't think we should support this kind of UI

[19:58:03.0000] <til>
but there's a situation where you can require an input and it needs to be defined

[19:58:04.0000] <til>
e.g. what is a required checkbox

[19:58:05.0000] <Hixie>
right now it is defined as "the checkbox has to be checked"

[19:58:06.0000] <til>
i don't understand how my proposal endangers yours

[19:58:07.0000] <Hixie>
it doesn't

[19:58:08.0000] <til>
why would there be an input group if they weren't related?

[19:59:00.0000] <til>
so why should checkValidity only apply to that control

[19:59:01.0000] <til>
not the group, like with other grouped controls

[19:59:02.0000] <Hixie>
because what you're proposing is bad ui

[19:59:03.0000] <til>
applying it to the group seems more consistent - easier to grok

[19:59:04.0000] <til>
it's no less bad than requiring a radio group to be checked

[19:59:05.0000] <Hixie>
it would encourage people to make such interfaces

[20:00:00.0000] <Hixie>
i agree, with radio buttons the mistake is allowing any to not be checked in the first place

[20:00:01.0000] <Hixie>
unfortunately that boat has sailed

[20:00:02.0000] <til>
but required is essentially meaningless without applying it to the group

[20:00:03.0000] <Hixie>
no?

[20:00:04.0000] <Hixie>
it has a very well-defined meaning

[20:00:05.0000] <Hixie>
thereis no "group" with check boxes

[20:01:00.0000] <Hixie>
just checkboxes that happen to share a name

[20:01:01.0000] <Hixie>
just like you can have many text input fields with the same name -- doesn't make them into a group

[20:01:02.0000] <Hixie>
if you have <input type=text name=a required> <input type=text name=a required> you aren't required to type something into one or the other

[20:02:00.0000] <til>
in my mind, a checkbox is a type of radio that is mutually inexclusive

[20:02:01.0000] <Hixie>
you're required to type something into both

[20:02:02.0000] <til>
which is why this is distinct from a text input

[20:02:03.0000] <Hixie>
checkboxes and radio buttons are separate just like checkboxes and text fields

[20:02:04.0000] <til>
checkboxes are closer to radio than to text fields

[20:04:00.0000] <Hixie>
i think thinking of them as similar is misleading and leads one to designing poor UIs

[20:19:00.0000] <Hixie>
i really don't think that extending the wacky radio button required="" model to text fields makes any sense at all

[20:19:01.0000] <til>
anyway, is there even a concept of order in query parameters?

[20:20:00.0000] <til>
couldn't you have either Apt#B 123 Fake St and 123 Fake St Apt#B

[20:20:01.0000] <til>
depending on how the vendor parses it?

[20:20:02.0000] <Hixie>
it's always sent in form order

[20:20:03.0000] <til>
or does A before B in the HTML mean A is before B in the query?

[20:20:04.0000] <til>
ok

[20:21:00.0000] <til>
so your counterpoint is that someone could have address as the name for both fields, and only want to require the first line?

[20:21:01.0000] <til>
and want the second line filled to not fulfill the first line's requirement?

[20:22:00.0000] <til>
since there's no way to know on the server side which was filled, i don't see why the requirement should affect text fields any differently than radios or checkboxes

[20:25:00.0000] <til>
if you have text boxes A B and C called address, and get two values for address on the server side

[20:25:01.0000] <til>
you don't know which two you got

[20:25:02.0000] <Hixie>
you always know which one was filled in since you always get both, in the order they were in the form

[20:25:03.0000] <til>
but if B is empty, you don't get a key with no value

[20:25:04.0000] <Hixie>
whether or not they were filled in

[20:25:05.0000] <Hixie>
yes you do

[20:25:06.0000] <til>
really?

[20:25:07.0000] <til>
ok

[20:26:00.0000] <til>
my mistake

[20:32:00.0000] <til>
i'm less bothered by your opinion on the matter than that there is no discussion about it

[20:32:01.0000] <til>
i feel steamrolled

[20:32:02.0000] <til>
i'd really like to hear other people's POV on this, because I feel my original point remains valid

[20:39:00.0000] <Hixie>
til: you can ask for an escalation -- see the boilerplate i posted in the bug for details on how to do that

[20:40:00.0000] <til>
i misunderstood.  i'm not used to being able to reopen my own bugs

[20:43:00.0000] <Hixie>
well reopening just means you're asking me to reconsider -- if you want to escalate it to a working group issue that the chairs then decide on rather than me, you have to set some keyword or something

[20:43:01.0000] <Hixie>
and then some elves come along and automagically raise the issue for you

[20:43:02.0000] <Hixie>
(honestly not quite sure how that happens)

[20:44:00.0000] <Hixie>
there's a link in the boilerplate i pasted that explains it iirc

[20:44:01.0000] <til>
i'm working on that now

[20:44:02.0000] <Hixie>
k

[20:44:03.0000] <til>
i'm slow to respond because i'm typing in bugzilla :)

[20:44:04.0000] <Hixie>
:-)

[20:53:00.0000] <til>
Hixie: thanks for the discussion

[20:53:01.0000] <til>
i've escalated the bug

[20:53:02.0000] <til>
hopefully we can get more input

[21:46:00.0000] <JonathanNeal>
Added IE9 IE9 http://sandbox.thewikies.com/html5-browser-stylesheets/table.php

[00:00:00.0000] <JonathanNeal>
How can I get tweet feeds on the page.

[00:01:00.0000] <JonathanNeal>
like a kind-of twitter widget that will show anything related to #iecss

[00:48:00.0000] <hsivonen>
is this test case considered to reflect the current spec? http://simon.html5.org/test/opera/datalist.htm

[01:17:00.0000] <JonathanNeal>
http://www.iecss.com/ woot

[01:20:00.0000] <hsivonen>
JonathanNeal: interesting. I was unaware that IE had undergone unit category changes in the UA style sheet

[01:36:00.0000] <JonathanNeal>
hsivonen, between versions?  yes.

[01:53:00.0000] <zcorpan>
hsivonen: yes

[01:54:00.0000] <hsivonen>
zcorpan: I guess I'm looking at a parser bug, then

[01:54:01.0000] <zcorpan>
hsivonen: it's filed in v.nu bugzilla

[01:58:00.0000] <hsivonen>
zcorpan: thanks. I just filed it as https://bugzilla.mozilla.org/show_bug.cgi?id=555906 while I wasn't looking at IRC

[02:09:00.0000] <hsivonen>
/me notes that the Vodafone rep to the TAG still puts EXI as a bullet point on "mobile" slides

[02:18:00.0000] <othermaciej>
what is EXI?

[02:19:00.0000] <zcorpan>
the class='' recommendation on b and i doesn't make much sense

[02:19:01.0000] <hsivonen>
othermaciej: "efficient" binary compression representation for XML infosets

[02:20:00.0000] <othermaciej>
oh, that

[02:20:01.0000] <zcorpan>
/me bets OBML is more efficient than EXI

[02:21:00.0000] <zcorpan>
but maybe not a fair comparison since obml doesn't represent an infoset

[02:25:00.0000] <hsivonen>
zcorpan: is OBML what Opera Mini uses?

[02:25:01.0000] <hsivonen>
zcorpan: does it transfer the CSS box tree?

[02:26:00.0000] <zcorpan>
hsivonen: it's what opera mini uses

[02:27:00.0000] <zcorpan>
hsivonen: i don't know what it transfers exactly

[02:28:00.0000] <hsivonen>
zcorpan: ok

[03:24:00.0000] <zcorpan>
feature testing for structured clone in webworkers is similarly annoying

[03:25:00.0000] <zcorpan>
it would be easier if non-strings throwed from the start

[03:54:00.0000] <gsnedders>
/me gives up on trying to go to Glasgow

[03:55:00.0000] <gsnedders>
hsivonen: The answer for what OBML transmits is AFAIK "no comment" :)

[03:57:00.0000] <hsivonen>
gsnedders: I see. my css box tree guess remains

[05:40:00.0000] <zcorpan>
does anyone know which server is used for http://trac.webkit.org/browser/trunk/LayoutTests/websocket/tests

[05:47:00.0000] <zcorpan>
https://trac.webkit.org/browser/trunk/WebKitTools/Scripts/webkitpy/layout_tests/port/websocket_server.py ?

[06:05:00.0000] <MikeSmith>
zcorpan: tyoshino or ukai will know

[06:05:01.0000] <MikeSmith>
though they may be away

[06:06:00.0000] <MikeSmith>
(21:11 in Japan)

[06:07:00.0000] <tyoshino>
hi

[06:07:01.0000] <tyoshino>
ukai should know well but he's not here now.

[06:08:00.0000] <zcorpan>
ok, thanks

[06:09:00.0000] <tyoshino>
basically you're right. we're using pywebsocket and some helper scripts

[06:10:00.0000] <zcorpan>
ok

[06:22:00.0000] <annevk>
I now got that attachment (indeed looks like a Word document), but I cannot open it...

[06:23:00.0000] <jgraham>
Post it on the public web as a hacking challenge

[06:24:00.0000] <jgraham>
Give extra credit for inventing a time machine to crack it before March 15th 2010

[06:25:00.0000] <annevk>
someone with Word can prolly win that contest within minutes

[06:29:00.0000] <Philip`>
annevk: Delete a few bytes at random from the file, so Word will just report it as corrupted and fail

[06:32:00.0000] <jgraham>
annevk: Still you get to find out what the file says

[06:33:00.0000] <annevk>
just got a reply with the questions inline

[06:34:00.0000] <zcorpan>
but no time machine :(

[06:38:00.0000] <annevk>
you guys watch too much tv

[06:39:00.0000] <annevk>
and mostly science-fiction at that it seems :p

[06:41:00.0000] <Philip`>
It's strange how rarely science fiction focuses on technology problems like incompatible file formats

[06:41:01.0000] <Philip`>
The problems are always stuff like engines exploding

[06:41:02.0000] <Dashiva>
Cowboy Bebop had an excellent take on that

[06:41:03.0000] <Dashiva>
They found a really old video tape they wanted to play, but it was betamax

[06:42:00.0000] <Philip`>
Even with modern cars, it's pretty rare for the engine to explode and pretty common for the on-board computer to develop some kind of software fault, so presumably the same kind of thing will continue into the future

[06:46:00.0000] <Philip`>
The only exception I can think of (with my not-very-good memory) is in Halting State, where some guy stays up all night writing Perl scripts to translate data between two slightly different databases

[06:46:01.0000] <Philip`>
but then it's spoiled by having a computer that explodes later on

[06:46:02.0000] <Philip`>
(if I remember correctly)

[07:40:00.0000] <hsivonen>
I wish iki.fi had shorter delays on DNS updates...

[07:41:00.0000] <hsivonen>
the server that hosts hsivonen.iki.fi changed its IP address without prior notice

[08:05:00.0000] <erlehmann>
hsivonen, did you change that 404'd PNG link ?

[08:33:00.0000] <hsivonen>
erlehmann: not yet. I intend to fix it once my personal site becomes usable again.

[09:18:00.0000] <MikeSmith>
annevk: what's rumored to be in this Word file you've obtained?

[09:20:00.0000] <JonathanNeal>
Goodmorning :)

[09:22:00.0000] <jgraham>
Timezoneist

[09:25:00.0000] <MikeSmith>
in Japan, for people who work at bars and such at night, it's common to say おはよ (good morning) no matter what time of day it actually is

[09:25:01.0000] <MikeSmith>
when first seeing somebody for the day

[09:26:00.0000] <MikeSmith>
even if that happens to be 23:00 in the evening or whatever

[09:26:01.0000] <MikeSmith>
what is Viognier?

[09:26:02.0000] <MikeSmith>
hmm, Rhone valley

[09:27:00.0000] <MikeSmith>
this could be good

[09:27:01.0000] <MikeSmith>
/me looks around for the wine steward

[09:28:00.0000] <jgraham>
MikeSmith: In Japan it it common for shop assistants to compete for who can say "Irashaimasen!" in the most whiny voice. Doesn't mean it's a good idea :)

[09:28:01.0000] <MikeSmith>
heh

[09:29:00.0000] <MikeSmith>
jgraham: that's why I wear earblugs

[09:29:01.0000] <MikeSmith>
plugs

[09:30:00.0000] <MikeSmith>
one unfortunate thing in Japan is that being enthusiastic sometimes (often) means being noisy

[09:31:00.0000] <MikeSmith>
e.g., if you go to a baseball game, there is always this huge cheering section of people who do nothing but make noise for the entire game

[09:31:01.0000] <MikeSmith>
I think most of them are not actually paying any attention at all to what is going on in the game

[09:33:00.0000] <MikeSmith>
wow, this Viognier tastes somewhat like one of those chocolate oranges

[09:34:00.0000] <MikeSmith>
Terry's Chocolate Orange

[09:38:00.0000] <paul_irish>
what's the viognier?

[09:38:01.0000] <MikeSmith>
paul_irish: wine

[09:38:02.0000] <MikeSmith>
white wine

[09:38:03.0000] <paul_irish>
hahah no i mean who is the vineyard. :)

[09:38:04.0000] <MikeSmith>
oh

[09:39:00.0000] <MikeSmith>
Gadino Cellars, Greene County, Virginia

[09:39:01.0000] <MikeSmith>
apparently

[09:39:02.0000] <MikeSmith>
I don't recommend it

[09:39:03.0000] <MikeSmith>
unless you really like chocolate oranges

[09:40:00.0000] <MikeSmith>
paul_irish: btw, jgraham may know about the Opera UA stylesheet

[09:40:01.0000] <paul_irish>
Oh tight. I'd love to grab it to incorporate into JonathanNeal's table

[09:40:02.0000] <gsnedders>
/me notes jgraham doesn't

[09:40:03.0000] <MikeSmith>
s/jgraham/gsnedders/

[09:41:00.0000] <gsnedders>
/me knows a bit

[09:42:00.0000] <jgraham>
gsnedders: How do you know what I know?

[09:42:01.0000] <gsnedders>
jgraham: Because when it came up before I remember you not knowing much.

[09:42:02.0000] <gsnedders>
Of course, you may have learnt since

[09:42:03.0000] <jgraham>
Not really

[09:43:00.0000] <jgraham>
Also, I have no idea what the contex is for the email you just sent

[09:43:01.0000] <jgraham>
*context

[09:43:02.0000] <gsnedders>
See, you're cluelesss.

[09:43:03.0000] <gsnedders>
*clueless

[09:44:00.0000] <jgraham>
The first case was more serpentine than clueless

[09:44:01.0000] <gsnedders>
paul_irish: The basic answer is there is no CSS file available for the UA stylesheet

[09:44:02.0000] <gsnedders>
(But you probably know taht already)

[09:49:00.0000] <MikeSmith>
gsnedders: is there any way to get a dump of the CSS rules?

[09:49:01.0000] <paul_irish>
Hm, this was the case for IE too, but jon was able to extract the values from this nasty XML mess. regardless, if there is any opera resource for the default styling available.. i'd be great to document :)

[09:50:00.0000] <gsnedders>
MikeSmith: I can't think of any trivial way

[09:51:00.0000] <MikeSmith>
I would think there'd be some userjs way to do it

[10:11:00.0000] <gsnedders>
/me realizes he can't actually spell oliver

[10:15:00.0000] <gsnedders>
/me guesses we have at least another four hours until olliej wakes up

[10:16:00.0000] <gsnedders>
(Call me cynical, but...)

[11:22:00.0000] <JonathanNeal>
hello again

[11:25:00.0000] <JonathanNeal>
Did we ever find the Opera UA stylesheet?

[11:26:00.0000] <gsnedders>
It's not there to be found.

[11:26:01.0000] <JonathanNeal>
:(

[11:27:00.0000] <JonathanNeal>
Does Gecko have a separate icon?

[11:37:00.0000] <gsnedders>
http://en.wikipedia.org/wiki/File:Heringsschwarm.gif

[11:37:01.0000] <gsnedders>
w00t for animated gifs

[11:38:00.0000] <Philip`>
I like that the thumbnail is animated

[11:38:01.0000] <Philip`>
You don't get that with <video>s

[11:49:00.0000] <JonathanNeal>
More than halfway through capturing opera styles

[11:50:00.0000] <gsnedders>
Anyone able to summarize the Bug 7034 thread?

[11:55:00.0000] <Philip`>
gsnedders: I think the summary is that nobody knows of a consistent set of rules for precisely determining authoring conformance criteria, and some people think there should be a consistent set of rules, and nobody can agree on what any of the rules should be

[11:55:01.0000] <gsnedders>
OK, that's the impression I got from the start of the thread

[11:55:02.0000] <gsnedders>
/me marks as read

[11:56:00.0000] <Philip`>
and after about 120 emails, still nobody agrees on anything

[11:57:00.0000] <Philip`>
(though the spec now lists some rough principles)

[12:18:00.0000] <JonathanNeal>
I'm vertifying the styles.

[12:18:01.0000] <JonathanNeal>
I have a few tests where if the reported styles are false, then changing a few things like colors and font sizes will produce new results.

[12:42:00.0000] <JonathanNeal>
I would love to add a stylesheet of w3c recommendations @ iecss

[12:47:00.0000] <ttepasse>
Huh? Why is HTMLAnchorElement.hash readonly in Safari?

[14:00:00.0000] <JonathanNeal>
http://www.googlefight.com/index.php?lang=en_GB&word1=Ian+Hickson&word2=Mike+Smith har

[14:02:00.0000] <Dashiva>
Your googlefight priveleges are revoked for failing to use quotes

[14:04:00.0000] <Dashiva>
I'm a bit confused what the RDFa+HTML activity is supposed to be, beyond rubberstamping RDFa 1.1 Core

[14:07:00.0000] <othermaciej>
as far as I'm concerned, the HTML+RDFa editor still has to substantively respond to bugs, if he wants it to be an HTML WG deliverable

[14:17:00.0000] <Dashiva>
othermaciej: Sure, but with "incompatible with deployed content" in the list of available answers, I don't have high hopes

[14:28:00.0000] <JonathanNeal>
Dashiva, ah come on!

[14:29:00.0000] <JonathanNeal>
Mike Smith still wins.

[14:34:00.0000] <othermaciej>
Dashiva: if the editor gives answers that a commenter objects to, then of course they are free to raise it as an issue, whatever the nature of the reason given

[16:25:00.0000] <Hixie>
man, why are people suddenly all up in arms about changes that HTML_4_ made? (like getting rid of frames, font, etc)

[16:25:01.0000] <Hixie>
did they run out of things in HTML5 to complain about?

[16:26:00.0000] <JonathanNeal>
Who's complaining?

[16:26:01.0000] <gsnedders>
JonathanNeal: How far did you get with trying to get a stylesheet for Opera?

[16:26:02.0000] <Hixie>
people in www-style (about frames) and public-html (about everything under the sun, but more recently, not being allowed to use media-specific formatting features mixed in with their semantic markup)

[16:27:00.0000] <JonathanNeal>
gsnedders, http://www.iecss.com/ at the bottom is a link to the opera stylesheet.

[16:27:01.0000] <JonathanNeal>
along with firefox and webkit.

[16:45:00.0000] <JonathanNeal>
Hi TabAtkins!

[16:45:01.0000] <JonathanNeal>
Do you have a twitter?

[16:45:02.0000] <TabAtkins>
Yeah, @tabatkins

[16:46:00.0000] <JonathanNeal>
Wow ... your site

[16:46:01.0000] <JonathanNeal>
It has mario clouds

[16:49:00.0000] <JonathanNeal>
gsnedders, will that work for you?

[16:50:00.0000] <gsnedders>
JonathanNeal: I was just looking to see if I saw anything that totally jumps out at me as being wrong

[16:51:00.0000] <JonathanNeal>
Don't we have folks who work on Opera, Webkit, IE in this chan?

[16:51:01.0000] <gsnedders>
/me works at Opera

[16:51:02.0000] <JonathanNeal>
They should be able to say "no no no, you got that wrong, oooh you forgot this, ooooh good call didn't realize we had that in there"

[16:51:03.0000] <gsnedders>
We're mainly markup people here, though

[16:52:00.0000] <gsnedders>
What do we know about layout? :P

[16:52:01.0000] <JonathanNeal>
gsnedders, oh okay, I was going to ask "are you permitted to share your ua stylesheet?"

[16:52:02.0000] <gsnedders>
JonathanNeal: The answer is there is no CSS stylesheet.

[16:52:03.0000] <JonathanNeal>
Well, there is now :)

[16:57:00.0000] <TabAtkins>
JonathanNeal: My homepage is still *entirely* useless.  But yes, they are Mario clouds.

[16:57:01.0000] <TabAtkins>
Eventually, when I care about it, those clouds will be links.

[16:58:00.0000] <JonathanNeal>
I know I need to add something like audio[controls] { display: inline; }

[16:58:01.0000] <JonathanNeal>
for opera

[16:58:02.0000] <Philip`>
TabAtkins: You could make links rain out of the clouds

[16:58:03.0000] <TabAtkins>
Philip`: That sounds... ridiculous.

[16:59:00.0000] <JonathanNeal>
...ly awesome

[16:59:01.0000] <Philip`>
Someone should make a cloud-computing admin interface that consists of Mario jumping across clouds, and headbutting them to reboot them


2010-03-31
[17:02:00.0000] <TabAtkins>
The Wii download screen is mario jumping into a block for every % the download progresses.

[17:03:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9224 -- it seems highly unlikely implementations will follow through

[17:03:01.0000] <JonathanNeal>
I can't predict how Opera will handle video, the closest approximation would be video { height: 150px; width: 300px; } video[src] { height: auto; width: auto; }

[17:05:00.0000] <Hixie>
video { height: auto; width: auto; } would be accurate, since the spec says <video>'s intrinsic width defaults to 300x150 iirc

[17:07:00.0000] <JonathanNeal>
oh, does the spec also say <audio> should default to 300x25?

[17:07:01.0000] <gsnedders>
http://auriethepixie.deviantart.com/art/The-Secret-Life-of-Trees-158431899

[17:07:02.0000] <gsnedders>
The lighting of that is awesome.

[17:08:00.0000] <JonathanNeal>
Opera reports inline for audio, but shouldn't it be inline-block?

[17:10:00.0000] <annevk>
I guess we just make it a replaced element

[17:10:01.0000] <nessy>
spec says nothing about audio default width/height

[17:10:02.0000] <annevk>
seems to be it should do the same as <img> / <video> ...

[17:11:00.0000] <annevk>
pretty sure it is inline for <img>

[17:11:01.0000] <JonathanNeal>
Yea.

[17:11:02.0000] <JonathanNeal>
I always thought that was a grandfathered inline.

[17:12:00.0000] <JonathanNeal>
I'll follow your guys advice on this one audio { display: none; } audio[controls] { display: inline; } video { display: inline; }

[17:13:00.0000] <annevk>
"The audio element, when it is exposing a user interface, is expected to be treated as a replaced element about one line high, as wide as is necessary to expose the user agent's user interface features. When an audio element is not exposing a user interface, the user agent is expected to hide it, irrespective of CSS rules."

[17:14:00.0000] <annevk>
so in effect you should not be able to display it even if you use display:block

[17:14:01.0000] <annevk>
that seems kind of contrary to <head> and <title> and all...

[17:16:00.0000] <JonathanNeal>
yes, it does.

[17:16:01.0000] <JonathanNeal>
and contrary to <img>

[17:17:00.0000] <JonathanNeal>
If you expect to express itself visually, it should be much more like <img> and <video>

[17:18:00.0000] <JonathanNeal>
but thanks for the heads up, annevk :)

[17:25:00.0000] <Hixie>
iirc people demanded that it be like that

[17:25:01.0000] <Hixie>
i forget why

[17:26:00.0000] <nessy>
maybe because it is assumed that most use of audio is as background music to the site?

[17:26:01.0000] <JonathanNeal>
Interactive Illinois Report Card?

[17:26:02.0000] <roc>
/me wonders what Mark Pilgrim has to say about http://blog.chromium.org/2010/03/bringing-improved-support-for-adobe.html

[17:26:03.0000] <JonathanNeal>
:-P

[17:27:00.0000] <JonathanNeal>
nessy, that makes sense for a control-less <audio> but once controls are introduced it seems very off.

[17:28:00.0000] <nessy>
I'm not saying it's the right decision :)

[17:28:01.0000] <nessy>
might be worth another discussion on the list, actually

[17:29:00.0000] <JonathanNeal>
"the list"

[17:29:01.0000] <Hixie>
JonathanNeal: wait, what? why does it make no sense when you have controls?

[17:30:00.0000] <Hixie>
i thought that was the one case that DID make sense

[17:30:01.0000] <annevk>
roc, uhuh

[17:31:00.0000] <annevk>
roc, in general putting much effort into a better plugin API sort of seems like a waste of time

[17:31:01.0000] <roc>
yeah

[17:31:02.0000] <roc>
honestly I don't think anyone at Mozilla is excited about Pepper

[17:31:03.0000] <roc>
we're not blocking it

[17:31:04.0000] <roc>
in fact this is the first sign I've seen that Adobe is interested in it

[17:32:00.0000] <Hixie>
JonathanNeal: btw if you want to look into why that rule is there, it was added in r4083 on 2009-10-05

[17:32:01.0000] <JonathanNeal>
Okay, I'm very ready to be wrong about everything all the time, but it just seemed off.

[17:32:02.0000] <annevk>
oh, I have seen some other signs of people working for Adobe pushing for a better plugin API

[17:33:00.0000] <JonathanNeal>
I guess you don't want to be able to hide <audio controls> right?

[17:33:01.0000] <JonathanNeal>
Irrespective of CSS rules.

[17:33:02.0000] <annevk>
I managed to convince my manager I had better things to do then work on that

[17:34:00.0000] <JonathanNeal>
I think my issue was that I had never thought about "irrespective of CSS rules".  If that is the case, why would head get display: none in the rendering guide section?

[17:34:01.0000] <annevk>
I phrased that poorly, we discussed it briefly; it didn't require much convincing either way...

[17:34:02.0000] <JonathanNeal>
Does that make sense what I just asked, Hixie?

[17:34:03.0000] <JonathanNeal>
(also, I do not know how to access the log for r4083 on 2009-10-05

[17:35:00.0000] <annevk>
sicking, what you're suggesting will give us much of the same problems we faced with deploying HTML5 forms in the future whenever we do something like it again...

[17:35:01.0000] <annevk>
sicking, just allowing everything is not a good strategy; especially for extensibility it is harmful

[17:36:00.0000] <annevk>
and from all things conformance requirements help with that is probably the one I care about the strongest

[17:38:00.0000] <JonathanNeal>
I would love to compile a list of "irrespective" and "suggested" css rules.

[17:39:00.0000] <annevk>
JonathanNeal, you can display the head element using CSS

[17:39:01.0000] <annevk>
JonathanNeal, try <style>head, style { display:block } </style> in some cool browser

[17:40:00.0000] <annevk>
if you then put contenteditable on the <style> element you can even edit the CSS rules on the fly

[17:40:01.0000] <JonathanNeal>
I hope not.  If you can, it is an "irrespective" css rule, which would be contrary to #display-types

[17:40:02.0000] <annevk>
no it's not

[17:41:00.0000] <othermaciej>
roc: I think on some level the motivation for better plugin API is Chrome OS

[17:41:01.0000] <othermaciej>
roc: not that I have any specific information from Google about that

[17:41:02.0000] <annevk>
i really love data:text/html, <style contenteditable>head, style { display:block } </style>

[17:41:03.0000] <annevk>
it's deliciously weird and useless, but still a lot of fun

[17:41:04.0000] <othermaciej>
annevk: you just made my head explode

[17:42:00.0000] <annevk>
maybe I should blog about that tomorrow for mass head explosion effects around the globe

[17:43:00.0000] <annevk>
(or at least the tiny part of the population that tracks it :) )

[17:43:01.0000] <JonathanNeal>
hahaha

[17:44:00.0000] <JonathanNeal>
Just did <style contenteditable style="display: block; white-space: pre;" type="text/css"> in Chrome

[17:44:01.0000] <annevk>
via sicking: http://search.twitter.com/search?q=%23shaverfacts

[17:44:02.0000] <JonathanNeal>
You need the white-space to prevent the <p> getting in there from Opera.

[17:50:00.0000] <JonathanNeal>
In IE <head> is irrespectively display: none;

[17:50:01.0000] <JonathanNeal>
In Chrome and Opera it is malable.

[17:50:02.0000] <JonathanNeal>
(and Firefox)

[17:51:00.0000] <JonathanNeal>
Is it meant to be irrespectively hidden, or is it recommended to be hidden?

[17:59:00.0000] <Hixie>
JonathanNeal: not sure i understand the questions you were asking earlier (Sorry was afk)

[18:44:00.0000] <boblet>
Hixie: I have a question about the hr element and 3.2.5.2 paragraphs. Apart from the p element what other elements can hr be between? flow minus heading and sectioning elements?

[18:45:00.0000] <Hixie>
?

[18:49:00.0000] <boblet>
“The hr element represents a paragraph-level thematic break”, with paragraph linking to 3.2.5.2

[18:50:00.0000] <boblet>
but I’d like to confirm if that means hr only between p elements or eg between lists too

[18:53:00.0000] <boblet>
hr is used a lot as a section break in Japan, mainly because it translated well on mobile browsers. Am I right in thinking this would be incorrect now?

[18:55:00.0000] <Hixie>
3.2.5.2 in which version of the spec? :-)

[18:56:00.0000] <Hixie>
oh 3.2.5.3 Paragraphs in the complete.html spec, ok

[18:56:01.0000] <Hixie>
it's just paragraph-level

[18:56:02.0000] <boblet>
Hixie: curses! transcribing error; 3.2.5.3 (http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#paragraphs)

[18:56:03.0000] <Hixie>
lists are paragraph-level

[18:56:04.0000] <Hixie>
pretty much everything is paragraph-level :-)

[18:57:00.0000] <Hixie>
the spec is pretty explicit when something isn't allowed or if there are restrictions

[18:57:01.0000] <boblet>
Hixie: ohrly? hmm… sectioning and heading content too?

[18:57:02.0000] <Hixie>
what it means is that <hr> can be used anywhere <p> can, basically

[18:58:00.0000] <Hixie>
though the real way to work out where you can put <hr> is to look at the content models

[18:58:01.0000] <Hixie>
not try to decode mostly informative text :-)

[18:58:02.0000] <boblet>
aha! well, I completely misread #paragraphs then

[18:58:03.0000] <Hixie>
the word "represents" is the key word that means "this element means..." e.g. for rendering or analysis purposes

[18:58:04.0000] <Hixie>
so "paragraph-level" here just means that <hr> can be treated like a paragraph for the purposes of rendering

[18:59:00.0000] <Hixie>
/me tries to work out why his browser isn't showing him what the server is returning

[18:59:01.0000] <boblet>
so hr between section elements or after h1 would be fine then?

[19:02:00.0000] <Hixie>
after <h1> would be pretty silly but sure

[19:02:01.0000] <Hixie>
between section elements is pretty much redundant

[19:03:00.0000] <boblet>
…because sections convey the semantics & style can be applied via borders or background images right?

[19:04:00.0000] <Hixie>
yeah

[19:04:01.0000] <Hixie>
<hr> basically says "end of one section, start of another"

[19:04:02.0000] <Hixie>
which is the same semantic as </section> <section>

[19:05:00.0000] <Hixie>
(though in the case of <hr> the sections are generally more thematic rather than explicit -- e.g. in a novel think <section> being like a chapter and <hr> being like a bigger space between paragraphs)

[19:08:00.0000] <boblet>
hixie: I misread this as basically hr only between paragraphs of text (eg theme change in novel), which is a lot more restrictive than you’ve explained

[19:09:00.0000] <Hixie>
ah, ok

[19:09:01.0000] <boblet>
I’ve submitted a comment about this (#9371). Thanks for your explanation!

[19:09:02.0000] <Hixie>
np, thanks for the bug!

[19:09:03.0000] <boblet>
I better go update my html5doctor article :)

[19:16:00.0000] <gsnedders>
WTH, the KLM website is certain that to fly from Montreal to Amsterdam, where there are directly flights, I should go by Detroit

[19:19:00.0000] <Hixie>
gsnedders: looks like anolis has a bug with astral plane characters

[19:19:01.0000] <gsnedders>
Hixie: How so?

[19:19:02.0000] <gsnedders>
/me notes it's probably not Anolis itself, but rather a parser issue

[19:19:03.0000] <Hixie>
<span>&#x1D4CF;</span> in the input turns into <span>&#55349;&#56527;</span> in the output

[19:20:00.0000] <Hixie>
/me discovers some of the named character references are in fact combining characters

[19:21:00.0000] <gsnedders>
Gah.

[19:21:01.0000] <Hixie>
i'd really rather the output was just ascii personally

[19:21:02.0000] <Hixie>
i thought that's what i'd set it to

[19:21:03.0000] <Hixie>
haven't checked if my config is wrong yet

[19:22:00.0000] <gsnedders>
It should be, if you set it to be

[19:23:00.0000] <Hixie>
oh it is ascii

[19:23:01.0000] <Hixie>
i was briefly confused because &#x000A; turned into a real newline, but of course that's still ascii...

[19:24:00.0000] <Hixie>
so yeah, looks like there's an astral plane problem

[19:25:00.0000] <gsnedders>
Hixie: html5lib serializer bug

[19:26:00.0000] <gsnedders>
jgraham: ping

[19:27:00.0000] <Hixie>
i was about to say there was also a validator bug, since the validator didn't catch it, but of course i validate my post-processed pre-anolis source file, not the post-anolis file

[19:28:00.0000] <gsnedders>
Hixie: This on PMS I presume?

[19:28:01.0000] <Hixie>
yeah

[19:30:00.0000] <gsnedders>
https://code.google.com/p/html5lib/issues/detail?id=143

[19:31:00.0000] <gsnedders>
I guess if the spec depends on this it's somewhat urgent? :)

[19:32:00.0000] <gsnedders>
/me is too tired to want to look at this now

[19:33:00.0000] <Hixie>
i'm adding the glyphs to the big table at the bottom

[19:33:01.0000] <Hixie>
it'll make the output invalid, but i don't mind it being invalid for a while

[19:33:02.0000] <Hixie>
i doubt anyone will notice the broken glyphs for a few days

[19:33:03.0000] <Hixie>
(other than people who are reading this -- hi logs readers!)

[19:34:00.0000] <gsnedders>
/me thinks for a minute, and concludes that combining chars are always legal

[19:34:01.0000] <gsnedders>
Even at the start of a string, I think

[19:35:00.0000] <Hixie>
the combining chars thing is a separate problem that i'll need to talk to the mathml guys about

[19:36:00.0000] <gsnedders>
I'll try and fix the html5lib bug tomorrow

[19:36:01.0000] <gsnedders>
Well, later today

[19:36:02.0000] <Hixie>
does python have a "string S is one of strings A,B,C" construct?

[19:36:03.0000] <Hixie>
as in "s in ['aaa','bbb','ccc']" or some such?

[19:36:04.0000] <gsnedders>
Yeah

[19:36:05.0000] <gsnedders>
Exactly that syntax :)

[19:36:06.0000] <Hixie>
cool

[19:37:00.0000] <gsnedders>
Except if it's a fixed set, you probably want to use a frozenset

[19:37:01.0000] <Hixie>
and does it have ternary-? syntax? as in "foo ? true : false" ?

[19:37:02.0000] <gsnedders>
(as then it's just a hashtable lookup)

[19:37:03.0000] <gsnedders>
Hixie: no

[19:37:04.0000] <Hixie>
k

[19:37:05.0000] <Hixie>
frozenset?

[19:37:06.0000] <gsnedders>
a set which is immutable

[19:37:07.0000] <Hixie>
i just want a set literal

[19:37:08.0000] <gsnedders>
set([1,2,3])

[19:37:09.0000] <Hixie>
k

[19:37:10.0000] <gsnedders>
that gives a mutable set

[19:38:00.0000] <gsnedders>
s/set/frozenset/ gives an immutable one

[19:38:01.0000] <Hixie>
what does "[1,2,3]" give?

[19:38:02.0000] <gsnedders>
A list

[19:38:03.0000] <Hixie>
is frozenset() a method call, ora unary operator that gets optimised away?

[19:39:00.0000] <gsnedders>
It's a method call

[19:39:01.0000] <Hixie>
so probably not what i want here, since i'm in a loop

[19:39:02.0000] <Hixie>
there's no set literal syntax?

[19:39:03.0000] <gsnedders>
If it's a constant set, just have it declared outside the loop

[19:40:00.0000] <gsnedders>
No

[19:40:01.0000] <Hixie>
pah

[19:40:02.0000] <Hixie>
silly language

[19:40:03.0000] <Hixie>
i'll just use a list, it's short

[19:40:04.0000] <gsnedders>
/me notes there are plenty of things that are silly in Python

[19:40:05.0000] <Dashiva>
If you want dictionary lookup, can't you just use a dict instead of an array?

[19:40:06.0000] <gsnedders>
Dashiva: Then they need a value

[19:41:00.0000] <Hixie>
here's what i want:

[19:41:01.0000] <gsnedders>
Dashiva: If you want O(1) checking something is in a set, you don't need a value for every key

[19:41:02.0000] <franksalim>
a ternary operator was added to python in 2.5, btw

[19:41:03.0000] <Dashiva>
gsnedders: But you can't make a set literal

[19:41:04.0000] <Hixie>
(in pseudocode):  variable = foo in ['aaa', 'bbb', 'ccc'] ? 'xxx' : 'yyy'

[19:41:05.0000] <Hixie>
and i'm in a loop

[19:42:00.0000] <gsnedders>
franksalim: Oh, yeah, duh

[19:42:01.0000] <Hixie>
so i'd rather not do anything expensive like constructing something or allocating memory in every loop

[19:42:02.0000] <Dashiva>
And you can't declare the lookup list outside the loop?

[19:42:03.0000] <Hixie>
well i'd rather not since then to read the code you have to look in two places to work out what this does

[19:43:00.0000] <gsnedders>
variable = 'xxx' if foo in ['aaa', 'bbb', 'ccc] else 'yyy'

[19:43:01.0000] <Hixie>
sweet jesus that's worse than perl

[19:43:02.0000] <Hixie>
i'll just use a simple if/else :-P

[19:43:03.0000] <franksalim>
it's two lines

[19:43:04.0000] <gsnedders>
Hey, at least it's not PHP.

[19:44:00.0000] <Dashiva>
You can optimize for the common case by using assignment + only if!

[19:44:01.0000] <Hixie>
i really miss pascal. it was so clear compared to all these modern languages

[19:45:00.0000] <gsnedders>
/me notes there are still pascal compilers in existance

[19:45:01.0000] <Dashiva>
Javascript isn't so different from pascal, is it?

[19:45:02.0000] <Hixie>
Dashiva: javascript doesn't have a sane class system (sorry smalltalk people :-P)

[19:45:03.0000] <Hixie>
and it's not strongly typed

[19:45:04.0000] <Hixie>
gsnedders: i know, i'm writing my game in it

[19:46:00.0000] <Dashiva>
I will never understand how people can like dynamic strong typing

[19:46:01.0000] <Dashiva>
Worst of both worlds!

[19:47:00.0000] <franksalim>
sigh

[19:48:00.0000] <Hixie>
dynamic strong typing?

[19:48:01.0000] <Hixie>
pascal uses static strong typing

[19:48:02.0000] <Hixie>
as in, checked by the compiler

[19:48:03.0000] <Hixie>
well, provided you don't use Variants, but those are wacko and i recommend staying well clear unless you're doing weird COM stuff

[19:50:00.0000] <roc>
Pascal doesn't get used because important stuff like variable-length arrays didn't make it into the language

[19:51:00.0000] <Hixie>
any modern pascal has variable-length arrays

[19:51:01.0000] <Hixie>
but yeah, the original pascal was a pretty limited language

[19:51:02.0000] <roc>
if you want static type joy in a good language, use ML or Haskell

[19:51:03.0000] <Dashiva>
My bad, it's been too long since I used pascal, I misremembered it being staticly typed

[19:51:04.0000] <Hixie>
(pascal actually had quite a wide following on dos and windows, since borland made possibly the best RAD environments for those platforms)

[19:52:00.0000] <roc>
I was a big Turbo Pascal fan back in the day

[19:52:01.0000] <roc>
but C won

[19:52:02.0000] <Hixie>
(unfortunately borland's stuff was never truly cross-platform and they died when microsoft hired when of their people to do .NET and when microsoft made Visual Studio all one needed to work on windows)

[19:52:03.0000] <Hixie>
yeah

[19:53:00.0000] <ment>
Go is also static typed

[19:54:00.0000] <Hixie>
go is pretty nice in many ways, though they made some syntax decisions that are a bit odd

[19:54:01.0000] <Hixie>
still a very new language though

[19:55:00.0000] <Dashiva>
I always wanted to try D, but I can never get anything done unless I have a project to work on

[19:56:00.0000] <gsnedders>
/me feels the same with C/C++

[19:56:01.0000] <Hixie>
sanity check

[19:56:02.0000] <Hixie>
combining characters

[19:56:03.0000] <Hixie>
they come after teh character they combine with right?

[19:56:04.0000] <gsnedders>
Yes

[19:56:05.0000] <Dashiva>
Yes

[19:56:06.0000] <Hixie>
k

[19:56:07.0000] <Hixie>
i'm not going insane

[19:56:08.0000] <Hixie>
just checking

[19:56:09.0000] <Hixie>
or at least, that's not a symptom of my going insane

[19:58:00.0000] <Dashiva>
Going insane would require you to be sane to begin with

[19:58:01.0000] <gsnedders>
Hixie: I think you went insane a long time ago

[20:00:00.0000] <Dashiva>
Speaking of insane, was the last last week post the last?

[20:11:00.0000] <Hixie>
re my last checkin: if anyone can work out why (a) the columns no longer work in firefox or (b) why the :hover effect doesn't seem to work anywhere, please let me know

[20:11:01.0000] <Hixie>
(afk food)

[20:13:00.0000] <Dashiva>
Let's try validating

[20:14:00.0000] <Dashiva>
Just complaints about NFC...

[20:15:00.0000] <Dashiva>
Script error, though

[20:27:00.0000] <Dashiva>
The problem seems to be that the position: relative on the td is ignored. Probably because of the column stuff, since Opera doesn't support that and gets the positioning right.

[20:36:00.0000] <Dashiva>
Maybe this? "The effect of 'position:relative' on table-row-group, table-header-group, table-footer-group, table-row, table-column-group, table-column, table-cell, and table-caption elements is undefined."

[22:40:00.0000] <Hixie>
Dashiva: hm, thanks

[22:41:00.0000] <Hixie>
Dashiva: i figured they'd have defined it by now :-)

[22:52:00.0000] <JonathanNeal>
hello

[23:00:00.0000] <JonathanNeal>
Hixie, you around?

[23:10:00.0000] <Hixie>
JonathanNeal: vaguely

[23:22:00.0000] <JonathanNeal>
heh

[23:22:01.0000] <JonathanNeal>
So Hixie, why is it that <head> can be display: block; but <audio> shouldn't be overwritable?

[23:47:00.0000] <boblet>
JonathanNeal: thanks for iecss.com — great work!

[23:47:01.0000] <JonathanNeal>
boblet, thanks, and thank the guys on here for encouragement and advice.

[23:48:00.0000] <JonathanNeal>
Especially TabAtkins who wrote a php file I used to figure out how to write php files (which is how that table is rendered)

[01:25:00.0000] <hsivonen>
Well, this is interesting. And it's not the April Fools Day, yet. http://blog.chromium.org/2010/03/bringing-improved-support-for-adobe.html

[01:38:00.0000] <othermaciej>
hsivonen: I wonder if Mark Pilgrim will decry this move

[01:42:00.0000] <Hixie>
JonathanNeal: i don't recall, but the rule was added in r4083 on 2009-10-05, if you want to do some archeology to find out

[01:42:01.0000] <Hixie>
othermaciej: that's what roc said!

[01:44:00.0000] <othermaciej>
Hixie: yeah, I'm stealing his line

[01:44:01.0000] <Hixie>
ah :-)

[01:45:00.0000] <hsivonen>
othermaciej: that was my first thought, too

[01:45:01.0000] <Dashiva>
Man, what are we going to do when othermaciej goes to jail for theft

[01:46:00.0000] <othermaciej>
Dashiva: there will still be two co-chairs left

[01:46:01.0000] <othermaciej>
good thing we have double redundancy

[01:47:00.0000] <othermaciej>
hsivonen: I wonder if the bundle will come with the "Evil and Lazy" t-shirt

[01:47:01.0000] <Dashiva>
I'm not convinced you're all the same :P

[01:47:02.0000] <Dashiva>
It's more like a balance of power

[01:47:03.0000] <hsivonen>
othermaciej: :-)

[01:48:00.0000] <JonathanNeal>
where do i do said archeology?

[01:48:01.0000] <JonathanNeal>
Hixie

[01:49:00.0000] <roc>
my lines are available under an open source licence

[01:51:00.0000] <Dashiva>
JonathanNeal: Searching whatwg and public-html mail archives, and the irc channel logs if they go that far back

[01:51:01.0000] <othermaciej>
roc: viral or permissive?

[01:52:00.0000] <hsivonen>
/me wonders if Pepper surfaces can be GPU-hosted textures

[01:53:00.0000] <JonathanNeal>
Searched the mail archives but didn't see anything offering a detailed explanation.

[01:53:01.0000] <roc>
viral

[01:54:00.0000] <JonathanNeal>
Just instructional making-it-work-this-way talk

[01:54:01.0000] <roc>
in case anyone plans to make heaps of money out of books based on #whatwg logs

[01:54:02.0000] <roc>
hsivonen: I assume so, otherwise it'd be worse than what we already have (or will soon have)

[01:55:00.0000] <hsivonen>
roc: but GPL allows the making of heaps of money

[01:55:01.0000] <roc>
sure, in theory it does

[01:56:00.0000] <roc>
it destroys all the obvious business models and then says "good luck!"

[01:56:01.0000] <Dashiva>
It doesn't destroy the business model of relying on people thinking it's reasonable to pay for stuff

[01:56:02.0000] <roc>
note: I am a fan of the GPL for many things

[01:56:03.0000] <hsivonen>
or the business model of putting a Google search field on the product

[01:57:00.0000] <JonathanNeal>
Maybe there never was a reason to make <head> display blockable, but <audio> none or inline iirrespective of css

[01:57:01.0000] <JonathanNeal>
it should be consistant.

[01:57:02.0000] <Dashiva>
Making head display:block has been possible a long time, so I wouldn't be surprised if at least one site depends on it :)

[01:58:00.0000] <roc>
hsivonen: yeah, we're lucky

[02:00:00.0000] <roc>
the GPL makes it hard to make monopoly profits, which is what I think of as "heaps of money" :-). That's a good thing.

[02:00:01.0000] <JonathanNeal>
I understand, no controls on audio, then no audio, but having no control over the audio control's width or height or display seems very non <video / image / canvas / object> -ish

[02:06:00.0000] <JonathanNeal>
*then no audio = then no display

[02:25:00.0000] <Philip`>
/me suggests "s in ('aaa', 'bbb', 'ccc')"

[02:25:01.0000] <Philip`>
since rounded brackets are prettier, and tuples are theoretically faster than lists

[02:26:00.0000] <Philip`>
(If you've only got n=3 then I don't think it's really worth worrying about a O(n) list/tuple search vs O(1) set search)

[02:30:00.0000] <workmad3>
Philip`: are you sure that went into the right chan? :)

[02:30:01.0000] <Hixie>
yeah he was responding to something i said a while bacm

[02:30:02.0000] <Hixie>
back

[02:30:03.0000] <Hixie>
thanks Philip` 

[02:31:00.0000] <workmad3>
ah, I thought I'd accidentally joined #python again :)

[02:41:00.0000] <jgraham>
gsnedders: Unless I am misreading, it is actually a bug in the python stdlib

[02:49:00.0000] <jgraham>
(and I tend to agree with Philip`; for n == 3 you probably don't care about O(1) vs O(N) but about constant factors 

[02:50:00.0000] <jgraham>
)

[02:50:01.0000] <Philip`>
(If you're using Python you probably don't care about constant factors either)

[02:51:00.0000] <jgraham>
(Well you might if you were constrained to use Python and this was a particularly slow part of the code. But that seems unlikely in the extreme)

[02:51:01.0000] <Dashiva>
If it was a critical section of the code, I don't think moving the list out of the loop would be unacceptable

[02:52:00.0000] <jgraham>
Moving the list out of the loop seems like it should be acceptable in either case. If the loop body is so long that that seriously impacts readbility, you probably have two problems

[02:52:01.0000] <Dashiva>
He didn't say he was using regular expressions

[02:53:00.0000] <Philip`>
If it was a critical section of the code, rewriting that whole part of the program so it doesn't have to deal with strings in a tight loop may not be unacceptable

[03:04:00.0000] <zcorpan>
ukai: hi

[03:05:00.0000] <ukai>
zcorpan: hi

[03:06:00.0000] <zcorpan>
ukai: do you know if the webkit websocket tests need pywebsocket as an apache module or standalone server, or doesn't matter?

[03:06:01.0000] <ukai>
standalone

[03:06:02.0000] <zcorpan>
ok, thanks

[03:07:00.0000] <ukai>
WebKitTools/Scripts/run-webkit-tests launches pywebsocket's standalone.py in openWebSocketServerIfNeeded()

[03:08:00.0000] <zcorpan>
ukai: i'm thinking of running them with a remote server instead of a local server

[03:08:01.0000] <zcorpan>
ukai: the remote server might well be running all the time

[03:09:00.0000] <ukai>
zcorpan: what do you mean the remote server in this context?

[03:10:00.0000] <zcorpan>
ukai: on a different machine than the one running the test

[03:10:01.0000] <zcorpan>
ukai: the webkit tests seem to use 127.0.0.1

[03:11:00.0000] <ukai>
zcorpan: yes, but you need to rewrite all *.js in layout test code

[03:11:01.0000] <zcorpan>
ukai: yeah

[03:12:00.0000] <ukai>
zcorpan: so, what is the motivation?

[03:12:01.0000] <zcorpan>
ukai: trying to get it running in opera's regression testing system

[03:13:00.0000] <ukai>
ukai: ah. is there any problem with running on 127.0.0.1?

[03:14:00.0000] <ukai>
zcropan: ah. is there any problem with running on 127.0.0.1?

[03:15:00.0000] <zcorpan>
ukai: not sure, but it seemed like it would be simpler to install a websocket server on one machine than installing it on all test slaves

[03:18:00.0000] <ukai>
zcorpan: hmm, i'm not sure, but i believe we need to start pywebsocket server each time to run layout tests. some *_wsh.py need to be initialized before running test.

[03:22:00.0000] <zcorpan>
ukai: so different tests start the server with different configurations?

[03:26:00.0000] <ukai>
zcropan: no, it runs with same configurations.

[03:26:01.0000] <zcorpan>
/me -> lunch, bbiab

[03:27:00.0000] <ukai>
zcorpan: so, it might work with remote server that keep running.

[03:34:00.0000] <hsivonen>
is any browser knows to sniff charset-unlabeled JavaScript files for encoding patterns using a detector?

[03:34:01.0000] <hsivonen>
s/knows/known/

[03:47:00.0000] <jgraham>
hsivonen: I have a feeling that Opera did/does (and that it is considered a bug; don't recall if it got fixed for 10.50)

[05:09:00.0000] <zcorpan>
ukai: ok, thanks

[05:25:00.0000] <hsivonen>
/me concludes that most style sheets and scripts have to be ASCII if Opera has been get away with its external resource behavior with document.open()ed docs

[05:26:00.0000] <hsivonen>
when the behavior differs from IE, Gecko and WebKit

[05:36:00.0000] <zcorpan>
hsivonen: what's different?

[05:38:00.0000] <hsivonen>
zcorpan: the other's decode the external resources as UTF-8. Opera decodes as Windows-1252.

[05:38:01.0000] <hsivonen>
*others

[05:39:00.0000] <hsivonen>
scary how I've started to put ' randomly in front of trailing s

[06:00:00.0000] <zcorpan>
hsivonen: maybe there aren't that many document.open()ed docs that use external style sheets and scripts

[06:00:01.0000] <zcorpan>
hsivonen: could you file a bug about that?

[07:55:00.0000] <gsnedders>
jgraham: That was actually my blind guess

[08:43:00.0000] <divya>
@TabAtkins you mentioned (w.r.t to my query on legend elm's content model) < "Content model" describes what the element is allowed to *contain*.

[08:43:01.0000] <divya>
"Categories" describe what the element *is*.  (And then "Context"...describes where it can be used.) > Am I right in assuming HTML4 did not explicitly declare category/context and only HTML5 does so? 

[08:44:00.0000] <TabAtkins>
divya: I'd have to check HTML4 to be sure.  Been a while since I've read it.

[08:44:01.0000] <Dashiva>
It declared context, surely

[08:44:02.0000] <Dashiva>
Isn't that what the DTD is all about?

[08:45:00.0000] <divya>
@Dashiva  e.g. http://www.w3.org/TR/REC-html40/struct/global.html#h-7.5.1 for Body element there is explicit content model associated (%block; | SCRIPT)+ +(INS | DEL)

[08:46:00.0000] <divya>
@Dashiva Nothing explicit about context. 

[08:46:01.0000] <Philip`>
The "%block" definitions are sort of about context

[08:47:00.0000] <Philip`>
but in general the child->parent relationships aren't explicit, you have to derive them from te parent->child data in the DTD

[08:47:01.0000] <divya>
@Philip` %block defines what can go WITHIN body element and not what content model the element ITSELF is. 

[08:48:00.0000] <Philip`>
divya: But some other part of the DTD says that <div> etc *is* a %block, i.e. saying it can be used in all contexts that accept blocks

[08:49:00.0000] <divya>
@Philip` is there a ref here: http://www.w3.org/TR/REC-html40/sgml/dtd.html

[08:50:00.0000] <Philip`>
<!ENTITY % block "P | %heading; | %list; | %preformatted; | DL | DIV | NOSCRIPT | BLOCKQUOTE | FORM | HR | TABLE | FIELDSET | ADDRESS">

[08:51:00.0000] <Philip`>
That's kind of the explicit declaration of category

[08:52:00.0000] <divya>
@Philip` so any element in that list has a category of Block? 

[08:53:00.0000] <Philip`>
Yes, and can be used in any element whose content model includes "%block;"

[08:54:00.0000] <divya>
@Philip` ah k. but contexts are not specified though I think. They are assumed. E.g. ins and del. 

[08:55:00.0000] <jgraham>
divya: (this is not twitter, no need for the @ sign)

[08:56:00.0000] <divya>
jgraham: sorry :) too used to it.

[08:56:01.0000] <jgraham>
np :)

[08:59:00.0000] <gsnedders>
/me apparently just bought some "Nytol 0-A-Night" if you believe the recepit

[09:00:00.0000] <Philip`>
That doesn't seem very good product design, if you're expected to take none a night

[09:00:01.0000] <Philip`>
They'll never get any repeat customers

[09:00:02.0000] <jgraham>
Maybe you take it during the day

[09:00:03.0000] <Dashiva>
That doesn't stop homeopathy

[09:01:00.0000] <Philip`>
Dashiva: That's not how homeopathy works - you have to keep buying their expensive water regularly

[09:01:01.0000] <Philip`>
so they've got the right idea

[09:02:00.0000] <Dashiva>
You could just use half, and then refill yourself

[09:02:01.0000] <gsnedders>
Hopefully I can finally get off a 3am-10am sleep pattern :\

[09:02:02.0000] <gsnedders>
Being ill and ending up with that sort of sleep pattern scuks

[09:02:03.0000] <gsnedders>
*sucks

[09:02:04.0000] <Dashiva>
At least you have the excuse of being ill

[09:03:00.0000] <gsnedders>
It was so sad, I missed going to Oslo.

[09:03:01.0000] <Philip`>
gsnedders: You could move to a timezone several hours westwards, and then your pattern would be fine

[09:04:00.0000] <gsnedders>
I tried to do that, but the clocks changed too.

[09:04:01.0000] <gsnedders>
So I'm still in the same timezone, despite moving west.

[09:04:02.0000] <Philip`>
Go west more

[09:04:03.0000] <gsnedders>
Philip`: Get me a plane ticket.

[09:04:04.0000] <Dashiva>
Take up residence on a hot-air balloon

[09:05:00.0000] <Philip`>
Build an underground city

[09:07:00.0000] <Dashiva>
I don't have a problem being awake at night, it's just that other people insist on only being available during working hours

[09:10:00.0000] <Philip`>
It's strange that everyone has roughly the same working hours

[09:10:01.0000] <Philip`>
It means you're busy working while everyone else is available, and as soon as you stop working and have some free time everybody else stops working and becomes unavailable

[09:11:00.0000] <Philip`>
It'd be more efficient if everybody was assigned a random number between 0 and 24 at birth, which is the offset applied to all their working hours

[09:12:00.0000] <Philip`>
because then a large fraction of people would be available at all times, and you'd never have to worry that all the shops are about to shut

[09:16:00.0000] <Dashiva>
Well, it's sort of annoying if your business hours don't overlap with people you're supposed to interact with

[09:16:01.0000] <Dashiva>
(E.g. anyone who's ever been on a team with more than six hours time zone difference)

[09:16:02.0000] <jgraham>
In Sweden the shops almost all shut for much of the weekend

[09:17:00.0000] <jgraham>
typical opening hours might be 11-15 on Sat. only

[09:18:00.0000] <jgraham>
There are literally would-be shoppers listlessly wandering the steets looking for places to spend money

[09:18:01.0000] <jgraham>
This pretty much disproves economics

[09:19:00.0000] <Dashiva>
There are social arguments for having non-business hours

[09:19:01.0000] <Dashiva>
So the economists might just be repressed and unable to optimize

[09:19:02.0000] <daedb>
All the swedish weekend money gets spent in bars and clubs instead ^^

[09:59:00.0000] <TabAtkins>
othermaciej: I volunteer to write counter-proposals to all of Shelley's recent issues, except maybe the one for Issue 92 (that sounds like it might actually be reasonable; I need to review it more closely).

[09:59:01.0000] <TabAtkins>
I can do it by next Friday.

[09:59:02.0000] <gsnedders>
Is it just me or is Spotify really bad at the moment?

[10:05:00.0000] <JonathanNeal>
ahoy!

[10:06:00.0000] <JonathanNeal>
I've checked the public html mailing list, and I still haven't  found an explanation as to why <audio> and <audio controls> are to be styled the way they are irrespective of CSS. 

[10:07:00.0000] <JonathanNeal>
Would anyone currently on have enlightenment regarding this?

[10:10:00.0000] <jgraham>
http://www.publications.parliament.uk/pa/ld200910/ldhansrd/text/100303-0001.htm

[10:10:01.0000] <TabAtkins>
I suppose that <audio> is hidden no matter what because otherwise it's really annoying.  I've had problems caused by me adding a border to all body>*, and getting <style> and <script> showing up.

[10:11:00.0000] <jgraham>
(that link is unlikely to be enlightening on almost any topic)

[10:11:01.0000] <TabAtkins>
<audio controls>, not as sure, but I suspect it's so that users can depend on their browser always exposing the same-looking controls.

[10:11:02.0000] <gsnedders>
jgraham: Lovely.

[10:12:00.0000] <TabAtkins>
Essentially, I think <audio controls> is supposed to act like a replaced element.  Styling the contents of replaced elements is officially out-of-scope for CSS.

[10:12:01.0000] <jgraham>
gsnedders: The punchline is a thing of beauty

[10:15:00.0000] <JonathanNeal>
I'm looking for documentation on "replaced elements"

[10:16:00.0000] <JonathanNeal>
Hmmm, reading now.

[10:20:00.0000] <JonathanNeal>
I see why I never noticed this before, "4 The elements of HTML" seems not to be the section that would mention these things.

[10:21:00.0000] <JonathanNeal>
Where would I learn that <audio> is supposed to act like a replaced element?  Where would I learn the intrinsic height/width/ etc values it would have?

[11:15:00.0000] <JonathanNeal>
Found some @ http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#replaced-elements

[11:15:01.0000] <hober>
TabAtkins: you can skip ISSUE-95; I've got that one.

[11:52:00.0000] <JonathanNeal>
How, in CSS, might you represent that an img had successfully loaded a src?

[11:56:00.0000] <JonathanNeal>
though it doesn't exist, I feel like img[complete] would be correct. "image.complete Returns true if the image has been downloaded, decoded, and found to be valid; otherwise, returns false."

[12:11:00.0000] <JonathanNeal>
Actually, in theory a pseudo class would better represent this, like :complete

[12:22:00.0000] <annevk>
JonathanNeal, yeah

[12:23:00.0000] <annevk>
some browsers have something like that pseudo class I believe

[12:23:01.0000] <annevk>
maybe at some point we'll add something like that

[12:23:02.0000] <JonathanNeal>
I'm in la la hypothetical land, I know.  I just find it interesting and possibly useful as a reference guide.

[12:28:00.0000] <JonathanNeal>
how do you think sectioning-content would be defined?  :section ?

[12:31:00.0000] <gsnedders>
Wow. Transport is Scotland is really diabolical today.

[12:33:00.0000] <JonathanNeal>
annevk, http://pastie.org/897474

[12:34:00.0000] <annevk>
JonathanNeal, replaced elements don't need inline-block

[12:34:01.0000] <annevk>
also, controls is a content attribute so you can just do audio[controls]

[12:35:00.0000] <annevk>
also, audio is by default not displayed

[12:35:01.0000] <JonathanNeal>
They don't need it, no.  This is a reference guide to what you may be seeing.  Like, the presentational requirements and recommendations as represented in css.

[12:35:02.0000] <JonathanNeal>
annevk, by default it is not displayed?  But, in browsers where it is not supported I thought it works like a container?

[12:35:03.0000] <annevk>
and for :section I assume we'd use :section(<section-level>)

[12:35:04.0000] <annevk>
if it's not supported it's display:inline

[12:35:05.0000] <JonathanNeal>
Got it!

[12:36:00.0000] <annevk>
because then all properties are just their initial value

[12:36:01.0000] <JonathanNeal>
So, if it had to be represented, there would be audio {} and audio:supported {} ?

[12:37:00.0000] <annevk>
no

[12:37:01.0000] <annevk>
if it's not supported there's no representation

[12:37:02.0000] <annevk>
isn't that obvious?

[12:38:00.0000] <JonathanNeal>
It is obvious.  It wasn't obvious to me because I was realising that unsupported doesn't equate unstyled in some browsers, including IE with the createElement hack.

[12:40:00.0000] <JonathanNeal>
So, I was just being hypocrtical, willing to talk in theory, while then over-factoring in reality.

[12:41:00.0000] <JonathanNeal>
annevk, you're saying that :section() would possibly work like this?  :section h1 == :section(1) h1, :section :section h1 == :section(2) h1 ?

[12:43:00.0000] <annevk>
maybe

[12:43:01.0000] <annevk>
I think we probably need :heading(<heading-level>) too

[12:44:00.0000] <annevk>
or maybe we do not need either and authors can just style them based on knowing their structure

[12:44:01.0000] <annevk>
would probably be more efficient since :section/:heading would need some fairly complicated data structures if everything is to remain performant

[12:48:00.0000] <JonathanNeal>
annevk, funny you say that, I just changed my examples to have :heading

[12:49:00.0000] <JonathanNeal>
I didn't use :heading in the way you described though, since hgroup is also a heading.

[13:02:00.0000] <Hixie>
so of the change proposals shelley wrote that i've looked at so far, all the bugs for which she asked for changes that i rejected, she's now proposed removing the whole section instead, and for all the bugs where she asked me to remove the whole section, she's proposed changes instead

[13:03:00.0000] <gsnedders>
So, basically, it appears I got rejected for Edinburgh on grounds that they have a lot of good applicants, and having slightly worse grades than some on grounds of illness is sufficant reason to get rejected.

[13:04:00.0000] <Hixie>
d'oh

[13:06:00.0000] <Hixie>
(oh ok that was only the first few. the rest are more in line with what she originally proposed)

[14:05:00.0000] <othermaciej>
good afternoon

[14:05:01.0000] <TabAtkins>
Yo, Maciej.

[14:06:00.0000] <othermaciej>
TabAtkins: looks like you may get your wish to write some counter-proposals

[14:06:01.0000] <TabAtkins>
Indeed.

[14:07:00.0000] <TabAtkins>
I volunteer to write one for all of them except maybe 92 (I need to check it out more first) and maybe 95 (hober wrote one, I haven't reviewed it yet).

[14:08:00.0000] <othermaciej>
well we need to review the incoming proposals and have a discussion period first

[14:08:01.0000] <othermaciej>
but your general willingness is noted

[14:16:00.0000] <Hixie>
anyone remember why we droped <dialog>?

[14:16:01.0000] <daedb>
Because people bitched about it?

[14:16:02.0000] <Hixie>
people are bitching about wanting it back too :-)

[14:16:03.0000] <daedb>
/me still likes <dialog>

[14:16:04.0000] <Hixie>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9270

[14:16:05.0000] <Hixie>
me too

[14:16:06.0000] <Hixie>
but we can't just keep flipflopping

[14:17:00.0000] <JonathanNeal>
So object's as plugins (eg flash) are to be treated as replaced elements, right?

[14:18:00.0000] <Hixie>
dialog was dropped on 2009-09-15

[14:18:01.0000] <annevk>
Hixie, confusing with ARIA, contradicts HTML4 suggested markup

[14:18:02.0000] <annevk>
Hixie, doesn't hit 80/20

[14:20:00.0000] <TabAtkins>
Hixie: Dropped <dialog> because it was insufficient for the common use-cases.

[14:21:00.0000] <TabAtkins>
It worked for the simplest type of conversations, but not if there were any other information, like timestamps (common use-case for marking chat logs), stage directions (common use-case for marking scripts).

[14:21:01.0000] <Hixie>
based on my archeology, it looks like daedb has the real reason

[14:21:02.0000] <TabAtkins>
Well, yes, people bitched because of that.

[14:21:03.0000] <TabAtkins>
There may have been unjustified bitching as well.

[14:22:00.0000] <daedb>
Couldn't timestamps be handled by just allowing <time> as a child of <dialog>? Stage directions could be <p> children or whatever :p

[14:23:00.0000] <TabAtkins>
Sure, of course we could have.

[14:23:01.0000] <Hixie>
timestamps actually were handled

[14:24:00.0000] <Hixie>
and i had added an element to the parser to do stage directions eventually

[14:24:01.0000] <Hixie>
(had to get it parsing right first)

[14:28:00.0000] <jgraham>
I recall dialog being dropped because some people were unhappy with the design and no one could really make a convincing case for the benefits

[14:30:00.0000] <annevk>
pretty sure my reasons were part of it too

[14:33:00.0000] <othermaciej>
I think the bottom line is that people had their own preferred solutions, and many were not convinced that <dialog> was overall an improvement, so they didn't want it recommended as *the* way to do things

[14:45:00.0000] <annevk>
btw, according to w3.markmail.org Ian still has the most messages in public-html

[14:46:00.0000] <JonathanNeal>
TabAtkins, watcha think of this css? http://pastie.org/897474

[14:47:00.0000] <annevk>
and RB is out of the top 10, but still high

[14:47:01.0000] <annevk>
/me is almost out of the top 10

[14:48:00.0000] <annevk>
W3C-wide it is Ian, Julian, Bjoern, Anne, Dan...

[14:48:01.0000] <annevk>
with Ian having a solid lead

[14:48:02.0000] <annevk>
(not counting the bugzilla bot)

[14:49:00.0000] <annevk>
and public-html is on route of overtaking www-style

[14:49:01.0000] <annevk>
o_O

[14:49:02.0000] <annevk>
www-style has 35000 and public-html 28000

[14:49:03.0000] <TabAtkins>
JonathanNeal: This an attempt at a default UA stylesheet?

[14:50:00.0000] <JonathanNeal>
Yes, a "HTML5 Presentational Requirements and Recommendations Guide" following what I could gather from the spec, and making up a few hypothetical css selectors in the process.

[14:53:00.0000] <TabAtkins>
JonathanNeal: Why are you assuming a :section(n), but not a :heading(n)?

[14:53:01.0000] <JonathanNeal>
There's never an instance I could find where :heading(n) was more appropriate (where heading = h1 h2 h3 h4 h5 h6 hgroup)

[14:54:00.0000] <JonathanNeal>
Unless I misunderstand how :heading would work.

[14:55:00.0000] <TabAtkins>
Well, it would make a lot of your rules more simple.  Rather than ":section(1) h2, :section(2) h1" you can just have :heading(2).

[14:55:01.0000] <JonathanNeal>
eg: <section><hgroup><h1>foo</h1></hgroup></section> would be what :heading(n)?

[14:58:00.0000] <JonathanNeal>
My concern is that I wouldn't want hgroup to multiply the effect of :heading

[15:11:00.0000] <jgraham>
JonathanNeal: Presumably it would match all headings at level n, which iirc is defined in the spec

[15:12:00.0000] <jgraham>
(well the spec defines the outline and the heading for each section)

[15:13:00.0000] <jgraham>
(but I don't recall what it says about <hgroup> with multiple heading descendants; I think it might only talk about the first since that is the one that forms the actual heading)

[15:21:00.0000] <JonathanNeal>
jgraham, I understand -- i just wouldn't want 3em to become 9em

[15:23:00.0000] <jgraham>
JonathanNeal: I don't understand why you think that might happen

[15:23:01.0000] <JonathanNeal>
say <hgroup> and <h1> are both :heading's

[15:24:00.0000] <JonathanNeal>
:heading(1) { font-size: 3em; } <hgroup><h1> becomes 3em 3em

[15:24:01.0000] <jgraham>
I wouldn't expect <hgroup> to match :heading

[15:26:00.0000] <annevk>
it would make sense

[15:27:00.0000] <jgraham>
for it to match :heading?

[15:27:01.0000] <jgraham>
Not really, it's not a heading

[15:27:02.0000] <othermaciej>
annevk: yeah, I see that Ian is barely beating me on public-html still

[15:27:03.0000] <JonathanNeal>
it's heading content

[15:27:04.0000] <jgraham>
othermaciej: The winners on that list are the ones at the bottom :)

[15:28:00.0000] <othermaciej>
jgraham: the real winners are the ones not on the list

[15:28:01.0000] <jgraham>
That's like being at the bottom but turned up to 11

[15:31:00.0000] <JonathanNeal>
"The hgroup element represents the heading of a section."

[15:35:00.0000] <jgraham>
JonathanNeal: Unless I am misreading it (which is most likely) the draft is buggy

[15:36:00.0000] <jgraham>
OK, it is not quite buggy jsut confusing

[15:39:00.0000] <JonathanNeal>
Do you think it could make sense that :heading matches <hgroup>

[15:39:01.0000] <hober>
no

[15:40:00.0000] <jgraham>
JonathanNeal: Yes. I retract my earlier position. It should match hgroup but not a <hx> element in <hgroup>

[15:41:00.0000] <JonathanNeal>
jgraham, really you think it shouldn't match the hn's "These elements represent headings for their sections."

[15:43:00.0000] <jgraham>
JonathanNeal: Only if they are in <hgroup>

[15:43:01.0000] <jgraham>
Otherwise you have the problem you stated before

[15:45:00.0000] <JonathanNeal>
Eep, I dunno that sounds more confusing to me because then your single :heading pseudo-selector changes randomly.

[15:46:00.0000] <JonathanNeal>
randomly == perceivably unexpectedly

[15:48:00.0000] <jgraham>
Well you can't have it both ways :) What makes most sense?

[15:48:01.0000] <jgraham>
The outline algorithm actually says (contrary to what I previously stated) that it is the hgroup, not the <hx> inside, that forms the heading

[15:50:00.0000] <jgraham>
Which is why you might expect it to match :heading

[15:51:00.0000] <JonathanNeal>
Interesting, so you interpret the spec to say that h1-6 loses :heading status when inside <hgroup>

[15:53:00.0000] <jgraham>
JonathanNeal: See http://www.whatwg.org/specs/web-apps/current-work/multipage/sections.html#outlines

[15:54:00.0000] <JonathanNeal>
I just realized the spec doesn't say if body should have margin or how much it should have.

[15:54:01.0000] <jgraham>
In particular "When entering a heading content element"

[15:54:02.0000] <jgraham>
"If the current section has no heading..."

[15:54:03.0000] <JonathanNeal>
/me is reading that now

[15:55:00.0000] <jgraham>
But if you are in <hgroup> the current section already has a heading (the hgroup) so the <hx> elements are not seperately assigned as headings

[15:56:00.0000] <jgraham>
One could argue they should match :heading nonetheless

[15:56:01.0000] <jgraham>
But it doesn't work if you consider :heading to be a simple result of the elements assigned as headings by that algorithm

[15:57:00.0000] <JonathanNeal>
I don't suppose anyone in this chan has power to say one way or another?

[15:58:00.0000] <jgraham>
Well Hixie could confirm that we have read HTML5 correctly. No one can really say about :heading yet though since it doesn't actually exist afaik

[15:59:00.0000] <JonathanNeal>
Right, if he can confirm which way to read the spec, then I'll update immediately.

[15:59:01.0000] <JonathanNeal>
In the meantime, :section seemed safe (and matched the way they had presented it where :section was x)

[16:00:00.0000] <JonathanNeal>
"When a Document is in quirks mode, margins on HTML elements at the top or bottom of body, td, or th elements are expected to be collapsed to zero." but I don't see a margin being defined for body.

[16:01:00.0000] <JonathanNeal>
"then a default value of 8px is expected" oh there it is.

[16:07:00.0000] <JonathanNeal>
Now to figure out what the font size is supposed to be.