2010-04-01
[17:20:00.0000] <TabAtkins_>
hober: I think your Change Proposal is good.

[17:33:00.0000] <hober>
thanks! any suggestions for improvements?

[19:05:00.0000] <TabAtkins>
/me doesn't know who people are at this talk, so is just hanging out until we have a panel discussion.

[19:07:00.0000] <annevk>
there's a talk first?

[19:07:01.0000] <TabAtkins>
Man, apparently?

[19:07:02.0000] <TabAtkins>
http://voyageofhtml5.eventbrite.com/

[19:07:03.0000] <Dashiva>
So where's our april's fools joke?

[19:08:00.0000] <TabAtkins>
There is no April Fools joke this year.

[19:08:01.0000] <Dashiva>
Is that the joke?

[19:08:02.0000] <TabAtkins>
You'll never know.

[19:08:03.0000] <Dashiva>
"Consider HTML5 is such a joke to begin with, WHATWG has decided to instead commit to a April's Fools serious activity"

[19:09:00.0000] <TabAtkins>
Heh, where's that from?

[19:09:01.0000] <Dashiva>
I made it up

[19:09:02.0000] <Dashiva>
April's fools!

[19:09:03.0000] <TabAtkins>
Ah, gotcha.

[19:09:04.0000] <annevk>
laughing out loud!

[19:55:00.0000] <boblet>
MikeSmithX: you there?

[19:55:01.0000] <MikeSmith>
boblet: yep

[19:56:00.0000] <MikeSmith>
sorry for not replying to e-mail

[19:56:01.0000] <MikeSmith>
I was traveling yesterday

[19:56:02.0000] <boblet>
np :) thought you’re busy so maybe IRC is faster

[20:01:00.0000] <MikeSmith>
boblet: let's ping annevk later when he's on

[20:01:01.0000] <annevk>
I'm here

[20:01:02.0000] <MikeSmith>
oh

[20:02:00.0000] <annevk>
waddup?

[20:02:01.0000] <boblet>
heh

[20:02:02.0000] <MikeSmith>
annevk: boblet's going to Fukuoka also on April 24 for the Web event I mentioned

[20:02:03.0000] <MikeSmith>
we need to decide how many days to stay there

[20:02:04.0000] <boblet>
annevk: hear you’re coming to Japan to partake in Mike’s … unique hospitality style :)

[20:03:00.0000] <boblet>
should be lots of fun

[20:03:01.0000] <annevk>
it was cool last time :)

[20:03:02.0000] <annevk>
my only constrain is that I fly out May 8

[20:04:00.0000] <boblet>
I think Oikawa-san is doing a same-day trip, but Shiraishi-san wants to stay Saturday night. dunno if Yakura-san has mentioned his plans yet

[20:05:00.0000] <boblet>
I’ve got work on 25th, so am trying to decide whether to leave 24th after event or crazy early 25th

[20:06:00.0000] <boblet>
It’ll be my first time to Kyuushuu so thinking to go 23rd, and wondering what your plans are…

[20:06:01.0000] <boblet>
what time constraints do you have Mike?

[20:07:00.0000] <MikeSmith>
none

[20:07:01.0000] <MikeSmith>
so I guess we can arrive on the 23rd

[20:08:00.0000] <MikeSmith>
and then stay over on the 23rd and 24th at least

[20:08:01.0000] <MikeSmith>
maybe head back on the 25th or even 26th

[20:10:00.0000] <boblet>
it sounds like they’re gonna organise a hotel, but I’m guessing that’d be for a night. ok I might stay 23rd then head back on 24th. I’ll miss the after-party but I wouldn’t be able to drink anyhow

[20:11:00.0000] <MikeSmith>
we can have a before-party on the 23rd

[20:11:01.0000] <boblet>
MikeSmith: you know we’re meant to send an indication of costs right? and check the profile page by this w/e?

[20:11:02.0000] <MikeSmith>
yeah, I saw that

[20:11:03.0000] <boblet>
before-party sounds great, esp. as it’s an afternoon event

[20:11:04.0000] <MikeSmith>
but I have not had time to check on flights or hotel

[20:12:00.0000] <MikeSmith>
did Shirakawa-san give any hotel info yet?

[20:16:00.0000] <boblet>
MikeSmith: apparently Shimokawa-san in Fukuoka has offered to book a hotel for us. I guess somewhere reasonable (they’re paying) and near the venue, which hopefully means convenient for you both outside the covered day(2?)

[20:16:01.0000] <boblet>
s/day(2)/day(s)/

[20:17:00.0000] <boblet>
annevk: they’re also wondering if you’d like to have your name associated with the event…

[20:18:00.0000] <boblet>
not sure if that means just saying that you’ll be there, or if they’re sounding you out as a speaker

[20:18:01.0000] <MikeSmith>
boblet: do we have a URL for this thing yet?

[20:19:00.0000] <MikeSmith>
/me shuold probably go back and read the e-mail

[20:43:00.0000] <annevk>
for continuity, the above was handled in /msg

[20:47:00.0000] <annevk>
so for the CSS value API we need some kind of hashmap and some kind of list

[20:47:01.0000] <annevk>
e.g. for 'margin' you need a hashmap and for 'cursor' you'd need a list

[20:48:00.0000] <annevk>
the hashmap can be immutable, but the list needs to be mutable

[20:49:00.0000] <annevk>
another constrain is that both the hashmap interface and list interface might need to be implemented on the same object in certain scenarios (for backwards compatibility; if something changes from being just a hashmap into a comma-separated list of hashmaps)

[20:49:01.0000] <annevk>
current sketch I have:

[20:49:02.0000] <annevk>
map { mapLength, mapKey(), getter getMapItem() }

[20:51:00.0000] <annevk>
list { listLength, insertListItem), deleteListItem(), getter getListItem() }

[20:52:00.0000] <annevk>
since getListItem() takes an int and getMapItem() takes a string the getter thing ought to work

[20:52:01.0000] <annevk>
thoughts?

[21:04:00.0000] <annevk>
http://twitter.com/Bess/status/11402770234 -- 'Tab Atkins says "sexiest" thing in HTML5 is CSS transformation. CSS camp is more peaceful than JS camp. Too much fight over JS camp'

[21:04:01.0000] <annevk>
o_O I wonder what was actually said

[21:25:00.0000] <djbpython>
anyone have any good examples? having trouble parsing some html into an lxml etree

[22:30:00.0000] <othermaciej>
Hixie: I am sad that the YouTube HTML5 beta no longer works :-(

[22:42:00.0000] <MikeSmith>
djbpython: using html5lib?

[22:54:00.0000] <shepazutoo>
hsivonen: ping

[23:03:00.0000] <MikeSmith>
othermaciej: my WebKit keeps crashing

[23:03:01.0000] <MikeSmith>
r56855

[23:04:00.0000] <othermaciej>
MikeSmith: hmmm

[23:04:01.0000] <othermaciej>
I am running my own build of r56882

[23:04:02.0000] <othermaciej>
it does not seem particularly crashy

[23:04:03.0000] <othermaciej>
is there any particular site that reproduces it?

[23:06:00.0000] <MikeSmith>
I got a bunch of tabs open

[23:06:01.0000] <MikeSmith>
and it seems to be crashing not on any particular site

[23:07:00.0000] <MikeSmith>
I'll mess around with it some more and see

[23:22:00.0000] <MikeSmith>
othermaciej: crashing on the "Thank you for testing WebKit" page

[23:22:01.0000] <othermaciej>
MikeSmith: consistently?

[23:22:02.0000] <othermaciej>
http://nightly.webkit.org/start/about is not crashing for me in my build

[23:24:00.0000] <MikeSmith>
so it opens http://nightly.webkit.org/start/trunk/56855 , then if I close that window (so there are no more windows open), then go to my History and select the "Thank you for testing WebKit" page, then let it open, then close it, then it crashes, consistently

[23:25:00.0000] <MikeSmith>
it gives me a "Do you want to resent application settings for WebKit?" dialog

[23:25:01.0000] <othermaciej>
MikeSmith: can't repro on my current build; will try it with an actual nightly

[23:26:00.0000] <MikeSmith>
OK

[23:26:01.0000] <MikeSmith>
I tried the "Reset and relaunch" option, but it seems to have made no difference

[23:28:00.0000] <hsivonen>
shepazutoo: pong

[23:29:00.0000] <hsivonen>
today is the day when you can't trust news :-(

[23:29:01.0000] <MikeSmith>
othermaciej: maybe it's an OS problem.. I've not restarted my machine for a long time

[23:30:00.0000] <othermaciej>
tried with the nightly - cannot repro

[23:31:00.0000] <MikeSmith>
ok

[23:31:01.0000] <MikeSmith>
I guess I'll try restarting my machine

[23:37:00.0000] <Lachy>
Has anyone come across any obvious or funny april fools day hoaxes yet?

[23:37:01.0000] <othermaciej>
I hope the YouTube HTML5 beta is an April Fool's joke

[23:37:02.0000] <othermaciej>
er the HTML5 beta being broken

[23:57:00.0000] <Lachy>
http://youtube-global.blogspot.com/2010/03/textp-saves-youtube-bandwidth-money.html

[23:58:00.0000] <Lachy>
I wish I could find a video that that new feature works on.

[00:00:00.0000] <othermaciej>
Lachy: works on this: http://www.youtube.com/watch?v=iDe4v318f64&feature=featured&textp=fool

[00:01:00.0000] <Lachy>
doesn't work for me

[00:02:00.0000] <othermaciej>
worked for me

[00:02:01.0000] <othermaciej>
in fact, works every time I visit that page

[00:04:00.0000] <Lachy>
damn geo-ip restrictions. Only works for me when I connect to a VPN in the US

[00:18:00.0000] <Lachy>
hmm, maybe it's not intentional geo-ip restrictions after all. It works from London too. It just doesn't work within Norway. It might just be something to do with their CDN caching, and not having the updated Flash video distributed everywhere yet.

[01:00:00.0000] <hsivonen>
Hixie: Did you make defer on VBScript invalid on purpose?

[01:00:01.0000] <hsivonen>
Hixie: everything that's not JS is a data block, right?

[01:55:00.0000] <MikeSmith>
othermaciej: restarted my machine but still crashing

[02:17:00.0000] <MikeSmith>
hmm, "lxml.etree.XMLSyntaxError: htmlParseCharRef: invalid xmlChar value 55349, line 59559, column 2233"

[02:17:01.0000] <MikeSmith>
when running the splitter

[02:19:00.0000] <Philip`>
That's a surrogate thingy?

[02:19:01.0000] <Philip`>
Sounds like the Anolis serialisation bug

[02:19:02.0000] <Philip`>
creating invalid HTML which lxml is therefore unable to load

[02:20:00.0000] <Hixie>
hsivonen: no, what makes you think that?

[02:20:01.0000] <Philip`>
(i.e. it's not my fault)

[02:20:02.0000] <Hixie>
hsivonen: executable code = script. non-executable = data.

[02:20:03.0000] <jgraham>
MikeSmith: You obviously don't read the logs :)

[02:21:00.0000] <MikeSmith>
jgraham: I read them, but I skip all the boring parts

[02:22:00.0000] <jgraham>
I stand by my original statement then

[02:23:00.0000] <jgraham>
Anyway it is a bug in the python stdlib that makes the html5lib serializer spit out the wrong character references for non BMP characters

[02:23:01.0000] <jgraham>
gsnedders promised to fix it yesterday, so complain to him

[02:24:00.0000] <jgraham>
(and then to me to update pms I guess)

[02:27:00.0000] <zcorpan>
Hixie: should we invent a mime type specifically for <script> data blocks?

[02:27:01.0000] <jgraham>
(and, then, for completenes, at Hixie to regen the spec)

[02:27:02.0000] <MikeSmith>
jgraham: Ok, thanks

[02:27:03.0000] <MikeSmith>
btw, I did just find it in the logs also

[02:28:00.0000] <Hixie>
zcorpan: wouldn't the mime type depend on what type the data block uses?

[02:28:01.0000] <zcorpan>
Hixie: so that people know what to use and newer scripting languages are less constrained by legacy

[02:28:02.0000] <MikeSmith>
/me wonders what to do in the mean time, because this is breaking the build of the multipage W3C copy of the spec

[02:29:00.0000] <zcorpan>
Hixie: a parameter or data-* could be used to distinguish different types of data blocks

[02:29:01.0000] <MikeSmith>
"some of the named character references are in fact combining characters".. great stuff

[02:30:00.0000] <jgraham>
I'm not sure that is insane from a MathML point of view

[02:30:01.0000] <Hixie>
zcorpan: i don't understand what problem that would solve, though it seems like it would make things more complicated in general

[02:31:00.0000] <zcorpan>
Hixie: i don't know what to put in type='' if i want to use a custom data block

[02:32:00.0000] <zcorpan>
Hixie: if i make something up, it might clash with some future scripting langauge that makes up the same mime type as i did

[02:32:01.0000] <MikeSmith>
jgraham: MathML should really be discouraging people from used named character references, instead of facilitating it

[02:32:02.0000] <Hixie>
the same type you would use for content-type if you were sending that block over http

[02:32:03.0000] <Hixie>
MikeSmith: i'm not sure if the characters being combining is intentional -- we should ask the mathml wg about that

[02:33:00.0000] <zcorpan>
hmm, i guess that would be text/plain

[02:33:01.0000] <MikeSmith>
Hixie: I see

[02:34:00.0000] <jgraham>
MikeSmith: If hand authoring is a goal there is no other sane way to do it

[02:34:01.0000] <jgraham>
If had authoring is not a goal we get into "the tools will save us" territory

[02:35:00.0000] <MikeSmith>
god save anybody who tries to do hand authoring of any serious mathml

[02:36:00.0000] <zcorpan>
indeed

[02:36:01.0000] <jgraham>
Well yeah it is painful

[02:36:02.0000] <jgraham>
But moving it from painful to effectively impossible is still not good

[02:40:00.0000] <hsivonen>
Hixie: so a validator can't check the requirement because the validator can't know which MIME types are executable by some UA

[02:40:01.0000] <hsivonen>
Hixie: I thought data blocks were defined to be things that aren't JS

[02:41:00.0000] <hsivonen>
Should I expect class selector matching to be sufficient proof of what encoding was used for a CSS file in IE6?

[02:42:00.0000] <Hixie>
hsivonen: well, you can hard-code that text/javascript and a few other script types are script types and assume most other types are not, but yeah, there's no way to reliably check that

[02:42:01.0000] <Hixie>
hsivonen: but then, you have the same problem with reliably checking the syntax of the scripts themselves

[02:42:02.0000] <Hixie>
hsivonen: you can only check the syntax of things you know about

[02:42:03.0000] <hsivonen>
right

[02:42:04.0000] <zcorpan>
Hixie: but it's easy to check async/defer

[02:42:05.0000] <zcorpan>
Hixie: and src

[02:43:00.0000] <hsivonen>
oh. right. you aren't supposed to use src with data blocks

[02:43:01.0000] <hsivonen>
Hixie: the newly-introduced requirement is redundant, then

[02:43:02.0000] <hsivonen>
which confused me

[02:44:00.0000] <Hixie>
how so?

[02:44:01.0000] <Hixie>
i made a bunch of changes today

[02:44:02.0000] <Hixie>
i'm not sure which you're talking about

[02:44:03.0000] <zcorpan>
async and defer must not be present if src isn't

[02:44:04.0000] <zcorpan>
src is not allowed  on data blocks

[02:44:05.0000] <hsivonen>
Hixie: the one that says async and defer are prohibited on data blocks

[02:45:00.0000] <hsivonen>
r4927

[02:45:01.0000] <Hixie>
oh, oops, i missed the requirement on async and defer needing src

[02:45:02.0000] <Hixie>
only saw it for charset

[02:45:03.0000] <boblet>
Dear cabal, for your amusement: http://twitter.com/boblet/status/11414105037 /cc TabAtkins

[02:48:00.0000] <hsivonen>
is the REPLACEMENT CHARACTER supposed to match in a selector just like any other character?

[02:48:01.0000] <Philip`>
boblet: That actually has been proposed on the WHATWG list, I believe

[02:50:00.0000] <Philip`>
boblet: http://www.w3.org/Bugs/Public/show_bug.cgi?id=6999

[02:50:01.0000] <Philip`>
(I was wrong)

[02:52:00.0000] <boblet>
Philip`: omg, not even in April!

[02:52:01.0000] <boblet>
wow, reality rules

[02:53:00.0000] <hsivonen>
yeah, reality is always crazier than April Fools. 

[02:54:00.0000] <jgraham>
Also from WHATWG "Should we have attributes to indicate whether <video> content [...] is NSFW"

[02:54:01.0000] <hsivonen>
this XHTML 3.0 announcement is an old joke, but it's pretty much like XSL-FO: http://www.cs.tut.fi/~jkorpela/html/xhtml3.html

[02:55:00.0000] <Hixie>
hsivonen: looks like what a lot of people in public-html are suggesting should be valid

[02:56:00.0000] <hsivonen>
Hixie: I don't know if that's the criterion the chairs will apply.

[02:56:01.0000] <Hixie>
which criterion?

[02:57:00.0000] <asmodai>
heh, got to love xkcd's april's fool

[02:57:01.0000] <hsivonen>
Hixie: that things that a lot of people suggest to be valid should be

[02:58:00.0000] <jgraham>
hsivonen: One of the principal suggesetees is himself a chair

[02:59:00.0000] <Hixie>
hsivonen: oh i wasn't suggesting that it would pass or fail or that it was a good or bad idea, i was just pointing out that that april fools was another example of something that some people were suggesting seriously

[02:59:01.0000] <hsivonen>
Hixie: oh

[03:00:00.0000] <hsivonen>
jgraham: well, previously one complaint was that Hixie had opinions of his own. at least Maciej and Paul participate in gauging if Sam's opinion enjoys consensus

[03:01:00.0000] <MikeSmith>
/me puts a s/5[5-6][0-9][0-9][0-9]/32/g into his makefile

[03:02:00.0000] <Hixie>
we haven't actually seen what the chairs base their decisions on yet

[03:03:00.0000] <Hixie>
so far as far as i'm aware we haven't seen any non-editorial decisions from the chairs

[03:03:01.0000] <Hixie>
and their editorial decisions have been more or less random

[03:03:02.0000] <zcorpan>
MikeSmith: won't that affect other parts of the spec? e.g. the websocket handshake examples have some random numbers

[03:04:00.0000] <zcorpan>
maybe they don't match the pattern

[03:05:00.0000] <Hixie>
(e.g. microdata is in its own draft despite being pretty tightly integrated with html, but a random meta name keyword is in the same draft as <meta> despite being mostly useless and completely independent)

[03:05:01.0000] <hsivonen>
Hixie: even though the basis isn't overt, all the three chairs are supposed to participate in the decision

[03:05:02.0000] <Hixie>
(have there been other decisions?)

[03:06:00.0000] <Hixie>
hsivonen: they might be using dice for all i can tell :-)

[03:07:00.0000] <Hixie>
we don't really have enough data to be able to tell yet

[03:08:00.0000] <MikeSmith>
zcorpan: I meant it just as a workaround until it's fixed or worked around in html5lib

[03:09:00.0000] <hsivonen>
argh. the rejection of non-text/css sheets makes stuff harder to test

[03:09:01.0000] <hsivonen>
I need to configure my server to have a separate set of extensions for misconfigured versions of each MIME type

[03:10:00.0000] <hsivonen>
I already map .html to text/html; charset=utf-8 and .htm to text/html

[03:11:00.0000] <hsivonen>
and yes, apparently REPLACEMENT CHARACTER works like any other character for selector matching

[03:19:00.0000] <gsnedders>
Yay for 4933 giving us a load to update in html5lib

[03:19:01.0000] <hsivonen>
has Opera started rejecting non-text/css sheets even in the quirks mode?

[03:19:02.0000] <zcorpan>
hsivonen: no

[03:19:03.0000] <Hixie>
ok i should sleep now

[03:19:04.0000] <Hixie>
nn

[03:20:00.0000] <hsivonen>
zcorpan: oops. I forgot to change my default encoding in Opera back to Windows-1252 after setting it to a Cyrillic encoding

[03:24:00.0000] <hsivonen>
It would be very interesting to test this document.open() stuff in IE5

[03:24:01.0000] <hsivonen>
and Netscape 4

[03:26:00.0000] <hsivonen>
Hixie: do you happen to remember why you made the spec say that the encoding for document.open()ed docs is UTF-16?

[03:26:01.0000] <hsivonen>
oh. Hixie went to sleep 

[03:30:00.0000] <zcorpan>
"You can tell if a browser has so designated its content handler

[03:30:01.0000] <zcorpan>
by trying to display a JPEG in <embed>."

[03:30:02.0000] <zcorpan>
browsers support jpeg in embed

[03:50:00.0000] <asmodai>
Hixie sleeps?

[03:50:01.0000] <asmodai>
I thought he was a script.

[03:53:00.0000] <zcorpan>
scripts can sleep

[03:57:00.0000] <Dashiva>
They just lock up the application when they do

[03:58:00.0000] <zcorpan>
Hixie locks up HTML5 when he sleeps

[04:05:00.0000] <asmodai>
heh

[04:47:00.0000] <Lachy>
today's xkcd is brilliant :-)

[04:47:01.0000] <asmodai>
Lachy: tried sudo apt-get dist-upgrade yet?

[04:48:00.0000] <Lachy>
asmodai, no. Just tried (sudo) make me a sandwich

[04:50:00.0000] <asmodai>
interesting read about that privacy issue with :visited

[04:50:01.0000] <asmodai>
http://hacks.mozilla.org/2010/03/privacy-related-changes-coming-to-css-vistited/

[04:52:00.0000] <gsnedders>
dbaron posted something too

[04:52:01.0000] <gsnedders>
On his own site

[04:53:00.0000] <Lachy>
haha, try the cat command :-)

[04:54:00.0000] <Dashiva>
It doesn't jive well with having browser shortcuts on . and -

[04:54:01.0000] <Lachy>
yeah, it doesn't seem to work in opera well at all

[04:58:00.0000] <asmodai>
always Opera... ;)

[04:59:00.0000] <asmodai>
gsnedders: but it's linked to from that link I pasted

[04:59:01.0000] <gsnedders>
Oh, I didn't notice that when I read it yesterday

[05:01:00.0000] <Philip`>
It saves a lot of effort if you just read http://github.com/chromakode/xkcdfools/blob/master/xkcd_cli.js instead of typing in commands

[05:02:00.0000] <Dashiva>
I just put the page source into jslint

[05:04:00.0000] <asmodai>
rofl, even a small adventure clone built in

[05:04:01.0000] <Dashiva>
I didn't get this one for the asl: '6/M/Battle School'

[05:05:00.0000] <asmodai>
 oneliners['emacs'] = 'You should really use vim.';

[05:05:01.0000] <asmodai>
oneliners['vi'] = oneliners['vim'] = 'You should really use emacs.';

[05:05:02.0000] <Philip`>
Ender?

[05:06:00.0000] <Philip`>
http://en.wikipedia.org/wiki/List_of_Ender's_Game_series_organizations#Battle_School

[05:06:01.0000] <asmodai>
Never read it (yet).

[05:07:00.0000] <gsnedders>
asmodai: Uh, yeah. That comes from a xkcd comic.

[05:08:00.0000] <asmodai>
gsnedders: Read them all but can't remember that one

[05:20:00.0000] <gsnedders>
hsivonen: http://validator.nu/?doc=http%3A%2F%2Fjacobrask.net%2Fcontenteditable-style.htm

[05:21:00.0000] <gsnedders>
hsivonen: it doesn't seem to like <!doctype html><body> <style scoped></style>

[05:21:01.0000] <gsnedders>
(note the spac)

[05:21:02.0000] <Creap>
<- jacobrask

[05:21:03.0000] <gsnedders>
Oh, the space doesn't matter

[05:22:00.0000] <gsnedders>
<!doctype html><title></title><body><style scoped></style> is invalid if you believe the validator

[05:22:01.0000] <gsnedders>
Creap: Do you need different nicks on different servers!?

[05:22:02.0000] <brucel>
c'mon, who believes a validator? puh-lease!

[05:22:03.0000] <gsnedders>
I don't ;P

[05:23:00.0000] <jacobrask>
gsnedders: work and non-work ;)

[05:23:01.0000] <gsnedders>
/me just gets confused because his Opera and W3C usernames are "gsneddon"

[05:27:00.0000] <brucel>
nice glass dolphin, eh, gsnedderws

[05:30:00.0000] <nessy>
nice new currency ;)

[05:31:00.0000] <hsivonen>
gsnedders: thanks. I have no idea how that bug arose.

[05:36:00.0000] <asmodai>
hsivonen: Curious thought: when do you anticipate the mathml to stop throwing errors? :)

[05:38:00.0000] <hsivonen>
asmodai: I want to say "later this month", but because I don't want to make promises I can't keep, I'm not promising anything

[05:38:01.0000] <asmodai>
hsivonen: Ok, but it's on a fairly soonish schedule, which is cool

[06:03:00.0000] <hsivonen>
/me notes that the commenters on the Chromium blog and the commenters on the Flash Player blog have very different attitudes towards bundling Flash with Chrome

[06:13:00.0000] <shepazutoo>
hsivonen: yt?

[06:13:01.0000] <hsivonen>
shepazutoo: yes

[06:18:00.0000] <shepazutoo>
hsivonen: I'm starting to write up wording for allowing @href/@src instead of/in addition to @xlink:href, and I was wondering if you had any thoughts on how we can resolve the dual-attribute conflicts... 

[06:24:00.0000] <hsivonen>
shepazutoo: the most obvious thing that comes to mind would be to use xlink:href it it is present and the new thing otherwise

[06:31:00.0000] <jgraham>
Happy mailman reminder day everyone

[06:33:00.0000] <shepazutoo>
hsivonen: what about the DOM value reflection for say, @href and @src?

[06:45:00.0000] <hsivonen>
shepazutoo: is the DOM property now href?

[06:48:00.0000] <hsivonen>
hmm. it is

[06:49:00.0000] <hsivonen>
adding no-namespace href and src may be more trouble than they are worth

[06:50:00.0000] <hsivonen>
shepazutoo: I don't have a good idea for reflecting no-namespace href and src consistenly

[06:50:01.0000] <asmodai>
When did Firefox stop giving clipboard access by default? 3.6?

[06:51:00.0000] <hsivonen>
shepazutoo: since .href setter will have to continue to set xlink:href 

[06:51:01.0000] <hsivonen>
or I'd assume that has to happen for compat

[06:53:00.0000] <hsivonen>
shepazutoo: for setters to work sanely, you should probably have new DOM properties for no-name href and src

[06:53:01.0000] <hsivonen>
shepazutoo: but I can't think of a sane name for the href property

[06:54:00.0000] <hsivonen>
shepazutoo: why do you want to introduce new attributes with duplicate behavior?

[06:54:01.0000] <hsivonen>
setAttributeNS sucks, but if script authors use the href setter, they don't need the NS-aware attribute setter

[06:55:00.0000] <hsivonen>
and writing xlink:href is not *that* terrible when factoring in the backwards compat benefit

[06:57:00.0000] <shepazutoo>
hsivonen: it's been a pain point for several developers (though I didn't mind so much myself), and for it would be nice for SVG newbies to have more consistency between SVG and HTML

[06:57:01.0000] <hsivonen>
shepazutoo: if we're making incompatible changes to SVG, let's kill the load event

[06:57:02.0000] <hsivonen>
shepazutoo: I see

[06:58:00.0000] <shepazutoo>
hsivonen: maybe that (the load event) should be on the table, too)

[06:58:01.0000] <hsivonen>
shepazutoo: anyway, it seems to me that the hardest part is figuring out the DOM property setter story for the no-namespace href

[06:58:02.0000] <hsivonen>
shepazutoo: sorry I don't have a solution

[06:58:03.0000] <hsivonen>
shepazutoo: has the SVG WG already agreed to make the load event async?

[06:59:00.0000] <hsivonen>
shepazutoo: zcorpan said Opera implements it as async

[06:59:01.0000] <shepazutoo>
hsivonen: re: load, I don't recall... we did discuss it

[06:59:02.0000] <shepazutoo>
that will be resolved in SVG 2, I imagine

[07:00:00.0000] <hsivonen>
/me would love to have a preliminary resolution before we ship the HTML5 parser

[07:00:01.0000] <shepazutoo>
hsivonen: re: href setter, yeah, it's tricky

[07:00:02.0000] <shepazutoo>
hsivonen: what's your timeframe?

[07:00:03.0000] <hsivonen>
shepazutoo: no hard timeframe yet

[07:00:04.0000] <hsivonen>
turning it on by default was a Q1 goal but now it's a Q2 goal...

[07:01:00.0000] <hsivonen>
shepazutoo: anyway, resolving the bug is scheduled for Q2

[07:01:01.0000] <hsivonen>
the bug being that the HTML5 parser doesn't dispatch the SVG load event

[07:04:00.0000] <shepazutoo>
hsivonen: then let's focus on getting something in the SVG 2 spec in that tiemframe

[07:42:00.0000] <brucel>
Can I just grab some svg and paste it into my html5 code and have it render? I know I will be able to, but can I now?

[07:47:00.0000] <jgraham>
brucel: In Gecko with html5.enable, yes

[07:47:01.0000] <jgraham>
For most values of svg

[07:48:00.0000] <hsivonen>
FWIW, double hyphens in comments are truly interop-sensitive: https://bugzilla.mozilla.org/show_bug.cgi?id=556037

[07:49:00.0000] <Philip`>
brucel: And in IE9

[07:54:00.0000] <jgraham>
brucel: Obligatory example: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/441

[07:54:01.0000] <hsivonen>
jgraham: I get an Internal Error

[07:55:00.0000] <jgraham>
Oh

[07:55:01.0000] <jgraham>
Yeah

[07:55:02.0000] <jgraham>
Well nevermind, make up your own demo :)

[07:55:03.0000] <jgraham>
(it was probably too long)

[08:05:00.0000] <brucel>
thanks. So I need an html5 parser (or IE9!) to have SVG in HTML5 in practice, tho its allowed in the spec as is

[08:06:00.0000] <beowulf>
http://jonnycampbell.com/ # i think the sections should be articles…

[08:09:00.0000] <jgraham>
brucel: Yes, presumably most browsers are going to add SVG-in-text/html support as part of larger parser changes for HTML5 compliance

[08:10:00.0000] <brucel>
thx. 

[08:32:00.0000] <beowulf>
ok, that link i posted is a little weird, for me it showed code, for someone else it's showing a weblog. ah well.

[08:37:00.0000] <jgraham>
beowulf: Seems to be a design decision

[08:37:01.0000] <jgraham>
s/incorrect spellings/correct spellings/

[09:22:00.0000] <JonathanNeal>
Mornin!

[10:23:00.0000] <gsnedders>
So, apparently I should make some decision about university.

[10:23:01.0000] <gsnedders>
I don't like decisions.

[10:24:00.0000] <jgraham>
Well it seems like you don't have that many options so it should be really easy

[10:25:00.0000] <othermaciej>
gsnedders: what kind of decision? which one to attend?

[10:28:00.0000] <gsnedders>
othermaciej: So I have an offer from Glasgow for CS (almost everywhere else rejected me). Do I bother to go to Glasgow or not, will I gain enough from it for it to be worthwhile?

[10:29:00.0000] <othermaciej>
gsnedders: I don't have a good picture of what schools are good in the UK

[10:29:01.0000] <othermaciej>
gsnedders: this is undergrad, or graduate?

[10:29:02.0000] <gsnedders>
undergrad

[10:29:03.0000] <jgraham>
gsnedders: FWIW I think you are underestimating the amount of useful stuff in the degree course

[10:30:00.0000] <jgraham>
I can't imagine you already know all the useful theory + mathematics + so on

[10:30:01.0000] <othermaciej>
gsnedders: I think what you learn in an undergrad CS program is likely to be very useful, assuming you are interested in a career in programming or related professions

[10:30:02.0000] <gsnedders>
/me isn't sure about taht

[10:30:03.0000] <gsnedders>
*that

[10:31:00.0000] <othermaciej>
I came into college thinking I knew a lot, but I learned a lot, including important frameworks for how to think about things

[10:31:01.0000] <jgraham>
gsnedders: Which part? The career interest or the fact that the material is useful?

[10:31:02.0000] <gsnedders>
The career

[10:31:03.0000] <jgraham>
Well if you're not interested in it you are doing a very poor job of demonstrating it

[10:31:04.0000] <othermaciej>
in the US you don't have to pick your field of study before you start

[10:32:00.0000] <jgraham>
othermaciej: Typically here you do

[10:32:01.0000] <jgraham>
(although of course there is always more flexibility than they let on)

[10:32:02.0000] <othermaciej>
gsnedders: are there other things you might be interested in pursuing as a career?

[10:33:00.0000] <gsnedders>
othermaciej: Quite a few English related careers (writing, journalism, editing, publishing I all would quite like) as well aeronautical engineering

[10:35:00.0000] <othermaciej>
gsnedders: at MIT, the aero/astro program was famous for being brutally hard, and yet with not so great career prospects

[10:35:01.0000] <othermaciej>
it was widely believed to be the hardest major

[10:36:00.0000] <othermaciej>
not saying that should dissuade you, just reporting

[10:36:01.0000] <gsnedders>
/me didn't get the impression it had that bad career prospoects

[10:36:02.0000] <jgraham>
If you are mainly interested in career prospects (which are not the only, or necessarily even the main consideration) all the things you listed are pretty poor

[10:36:03.0000] <othermaciej>
I expect writing is something you could pursue as a minor or simply by taking the right classes while doing another degree

[10:37:00.0000] <othermaciej>
gsnedders: it's not that getting a job is impossible, it's just that it apparently doesn't pay very well compared to other engineering fields, in light of how hard it is

[10:39:00.0000] <jgraham>
(in particular all beginning English students want to be writers and/or journalists. They all end up being teachers)

[10:39:01.0000] <jgraham>
(note: there is some degree of generalisation there)

[10:40:00.0000] <beowulf>
i studied geology, it has something like 3% employment in the field rate for grads 

[10:40:01.0000] <gsnedders>
jgraham: I know, that's part of the reason why I'm not quite taht enthusastic

[10:40:02.0000] <beowulf>
you can always do another one :)

[10:42:00.0000] <gsnedders>
othermaciej: I guess I'll do Gaelic at university then...

[10:51:00.0000] <gsnedders>
So yeah, I have no idea.

[13:29:00.0000] <Lachy>
http://www.trainsignal.com/WindowsTraining.aspx

[13:41:00.0000] <gsnedders>
jgraham: Will you murder me if I stay in Sweden?

[13:47:00.0000] <gsnedders>
No RFCs published yet today. :(

[14:01:00.0000] <jgraham>
gsnedders: Possibly. And do what?

[14:47:00.0000] <AryehGregor>
Bwahaha, http://annevankesteren.nl/test/contenteditable-style.htm is great.

[14:50:00.0000] <AryehGregor>
Of course, the cool part is the execution, not the actual functionality.

[14:50:01.0000] <AryehGregor>
The latter would be pretty easy.

[15:34:00.0000] <gsnedders>
jgraham: Hide my body?

[15:48:00.0000] <jgraham>
gsnedders: I doubt that would work. At least that's what TV taught me

[15:49:00.0000] <jgraham>
(there could be an element of selection bias though)

[15:49:01.0000] <jgraham>
Anyway it clearly wasn't what I meant :)

[15:49:02.0000] <gsnedders>
jgraham: More seriously, stay at pera

[15:49:03.0000] <gsnedders>
*Opera

[15:49:04.0000] <gsnedders>
jgraham: Since when did I ever give you answers straight away which were what you meant when there was an obvious mis-interpretation?

[15:52:00.0000] <jgraham>
gsnedders: Since it took you 1.5 hours to come up with that response I would hardly call it straight away


2010-04-02
[17:10:00.0000] <dglazkov>
wicked

[17:12:00.0000] <dglazkov>
I met annevk, fantasai, hakon, jens, and arron in one day.

[17:19:00.0000] <Hixie>
hm, looks like something is checking in changes under me

[17:19:01.0000] <Hixie>
it's causing my w3c-html upload script to fail

[17:23:00.0000] <JonathanNeal>
hey all

[18:26:00.0000] <othermaciej>
Hixie: did you figure out the problem?

[18:27:00.0000] <othermaciej>
I see a bunch of commits from mike marked [updated by splitter]

[18:32:00.0000] <Hixie>
MikeSmith: your temporary workaround for the bug in anolis is affecting unrelated parts of the spec

[18:32:01.0000] <Hixie>
it's also affecting my checking in, but i've worked around that by just forcing the checkin each time

[18:32:02.0000] <Hixie>
in my script

[18:32:03.0000] <MikeSmith>
Hixie: OK, I'll turn it off

[18:33:00.0000] <Hixie>
well it's not really causing me any problems at this point

[18:34:00.0000] <Hixie>
just thought you should know :-)

[18:34:01.0000] <MikeSmith>
OK

[18:35:00.0000] <MikeSmith>
so I guess I'll leave it on for now, and try to see if I can refine it so that it doesn't affect any other part of the spec

[18:36:00.0000] <MikeSmith>
maybe I should instead have my makefile for the multipage build just drop the whole named-character-references table for now

[18:38:00.0000] <Hixie>
really we just need anolis fixed :-)

[18:40:00.0000] <MikeSmith>
strangely enough, I running the multipage build script with the workaround removed, but I'm not getting the errors I was getting before

[18:48:00.0000] <myakura>
Hixie: per 2.2.2 Extensibility it follows vendor-specific attributes start with an underscore, but the examples shown above all start with a dash...

[19:31:00.0000] <Hixie>
man, lots of feature requests in this batch of bugs

[20:46:00.0000] <MikeSmith>
Hixie: I tweaked my script to just have the multipage build remove the whole named-character-references table

[20:46:01.0000] <MikeSmith>
and also fixed it so that I think you won't need to force your commit

[20:47:00.0000] <MikeSmith>
you probably will need to force it the next time you commit

[20:47:01.0000] <MikeSmith>
but I think you won't need to after that

[21:29:00.0000] <Hixie>
MikeSmith: cool

[22:11:00.0000] <MikeSmith>
Hixie: copy-paste error in you r4941 commit

[22:11:01.0000] <MikeSmith>
http://html5.org/tools/web-apps-tracker?from=4940&to=4941

[22:11:02.0000] <MikeSmith>
+   <dt><dfn title="attr-body-margintop"><code>marginheight</code></dfn> on <code>body</code> elements</dt>

[22:11:03.0000] <Hixie>
i'll file a bug thanks

[22:12:00.0000] <MikeSmith>
hai

[22:15:00.0000] <rjett>
Is this the correct place to ask a question about an issue I've run in to when tinkering with Web Sockets?

[22:26:00.0000] <Hixie>
rjett: sure

[22:27:00.0000] <rjett>
I've been having trouble getting my client to connect to my server from my University, but it works everywhere else

[22:27:01.0000] <rjett>
Shouldn't Web Sockets get passed any firewall issues? 

[22:42:00.0000] <Hixie>
rjett: which port?

[22:44:00.0000] <rjett>
for the Web Socket? I'm using 12345... 

[22:44:01.0000] <Hixie>
can you telnet to that port from your university network?

[22:44:02.0000] <rjett>
I couldn't ping my server from the University, but I could access web pages from the server

[22:47:00.0000] <rjett>
I did not try to telnet to that port

[22:49:00.0000] <Hixie>
sounds like you have a proxy that's blocking outgoing connections

[22:49:01.0000] <Hixie>
i recommend either looking to see if you can use a SOCKS proxy, assuming your browser supports that, or using port 443 and TLS-encrypting the websocket connection

[22:50:00.0000] <Hixie>
shepazutoo: yt?

[22:53:00.0000] <rjett>
Thanks a lot. I really appreciate your help. :) 

[22:53:01.0000] <Hixie>
np

[22:53:02.0000] <Hixie>
i expect this to be a common problem

[22:54:00.0000] <rjett>
From what I had read I was under the impression that Web Sockets would magically handle these kinds of issues

[22:54:01.0000] <Hixie>
sadly not

[23:25:00.0000] <wirepair>
quick question, that validator site is a neat idea, but aren't you worried about it getting abused?

[23:25:01.0000] <wirepair>
the 'address' option in particular

[23:27:00.0000] <Hixie>
validator site?

[23:27:01.0000] <wirepair>
http://html5.validator.nu

[23:27:02.0000] <wirepair>
i was under the impression some one here ran it

[23:27:03.0000] <Hixie>
abused how?

[23:27:04.0000] <wirepair>
as in i can make it scan any site i want

[23:28:00.0000] <Hixie>
how is that different than, say, validator.w3.org?

[23:28:01.0000] <wirepair>
it's not

[23:28:02.0000] <wirepair>
i could use that too ;>

[23:30:00.0000] <wirepair>
but if you're not worried about people using it as an open proxy and attacking other sites, then don't worry about it ;>

[23:36:00.0000] <Hixie>
i don't really see how they could attack other sites

[23:41:00.0000] <wirepair>
well, i could easily script up a tool that would point it at a site

[23:41:01.0000] <wirepair>
then look for common vulnerabilities, such as sql injection

[23:42:00.0000] <wirepair>
and since i can see the source of the responses

[23:42:01.0000] <wirepair>
i can just parse out the response

[23:43:00.0000] <wirepair>
and all of the attacks would look like they are coming from the validator, not from the attacker

[23:52:00.0000] <Hixie>
i expect henri has set some rate-limiting on the validator

[23:55:00.0000] <wirepair>
yeah that or some form of a captcha would be good

[23:56:00.0000] <wirepair>
but yeah i was just curious if anyone thought about it before, so i brought it up.

[23:59:00.0000] <Hixie>
hsivonen is the guy who runs it, so he's the one to ask

[00:03:00.0000] <wirepair>
will do, thanks

[00:55:00.0000] <boblet>
Hixie: you there?

[00:55:01.0000] <Hixie>
yeah

[00:56:00.0000] <Hixie>
notwithstanding the weird <wbr> interaction with nowrap, is it the same as ZWSP?

[00:56:01.0000] <Hixie>
anyone know?

[00:56:02.0000] <boblet>
re: Ruby, I noticed the w3 ruby spec uses annotating WWW with World Wide Web as an example. I’m wondering how using ruby like that relates to the abbr element

[00:56:03.0000] <Hixie>
relates how?

[00:56:04.0000] <boblet>
http://www.w3.org/TR/ruby/

[00:57:00.0000] <Hixie>
i mean, they're similar... how do you mean?

[00:57:01.0000] <Hixie>
ruby renders the annotation by default and abbr doesn't, is the main difference

[00:57:02.0000] <boblet>
well, if you’re adding ruby to an abbreviation, then I’m guessing that you could also wrap the ruby word in abbr

[00:58:00.0000] <boblet>
but the abbr title text would duplicate the rt text, so maybe not necessary

[01:00:00.0000] <boblet>
abbr and ruby seem pretty similar, so I guess I’m wondering if there’s any difference apart from the way they display the annotation

[01:01:00.0000] <JonathanNeal>
hey all

[01:01:01.0000] <JonathanNeal>
hi Hixie

[01:02:00.0000] <boblet>
finally since complex ruby didn’t make it in, would using a title on abbr or span be a way to add the second term

[01:02:01.0000] <boblet>
hey JonathanNeal 

[01:02:02.0000] <Hixie>
boblet: well there's lots of differences, e.g. they use different element names, and i wouldn't recommend using <abbr> for annotating bopomofo readings... dunno what to tell you

[01:02:03.0000] <Hixie>
hey JonathanNeal 

[01:03:00.0000] <JonathanNeal>
hey boblet

[01:03:01.0000] <Hixie>
boblet: what is the complex ruby example you want to do?

[01:03:02.0000] <JonathanNeal>
For a while I thought no one could read my messages.  Yay, I exist.

[01:03:03.0000] <boblet>
Hixie: ok, I’ll mock up some code and ping you for feedback later

[01:04:00.0000] <Hixie>
the amount of work required to add a new element to the spec is inane

[01:04:01.0000] <boblet>
Hixie: writing an article on it, so referring to the W3 Ruby doc example of Japanese with furigana and romaji

[01:04:02.0000] <Hixie>
i really need to work on this

[01:04:03.0000] <Hixie>
boblet: ah

[01:04:04.0000] <JonathanNeal>
What element are you thinking of adding?

[01:04:05.0000] <boblet>
JonathanNeal: it’s the nsfw element

[01:05:00.0000] <boblet>
hixie is trying to sneak it in while no one is looking ;-)

[01:05:01.0000] <Hixie>
trying to turn <wbr> from an obsolete elements into a valid element

[01:05:02.0000] <Hixie>
it's proving quitethe tedious task

[01:06:00.0000] <Hixie>
k i think i have it.

[01:06:01.0000] <Hixie>
/me regens

[01:06:02.0000] <JonathanNeal>
Shouldn't that be left to a style?

[01:06:03.0000] <JonathanNeal>
How is <wbr> meaningful as information?

[01:07:00.0000] <Hixie>
you can't really do it in style

[01:08:00.0000] <boblet>
Hixie: btw, what was the reason complex ruby was dropped? too complex for not enough use cases?

[01:09:00.0000] <Hixie>
it wasn't dropped

[01:09:01.0000] <Hixie>
it was just never added

[01:09:02.0000] <boblet>
or just deferred til later?

[01:09:03.0000] <Hixie>
it was not added because IE didn't do it

[01:09:04.0000] <Hixie>
HTML5's ruby is a reverse engineering of IE's, I didn't pay much attention to the ruby spec

[01:09:05.0000] <boblet>
dropped meaning it was specced in the ruby spec

[01:09:06.0000] <boblet>
aah

[01:10:00.0000] <boblet>
any thoughts of adding it later? say post-HTML5?

[01:10:01.0000] <Hixie>
are there real use cases for it?

[01:11:00.0000] <Hixie>
and can those use cases not be handled by simply nesting two rubys in each other?

[01:11:01.0000] <Hixie>
(i don't know the answers to these questions)

[01:11:02.0000] <othermaciej>
Hixie: I did check what at least some sites use <nobr> for

[01:12:00.0000] <othermaciej>
Hixie: the only one I recall offhand is YouTube, which uses it for the username of the person uploading a video in some screens, presumably because it is atomic and must not be broken across lines, even if it has a space

[01:12:01.0000] <Hixie>
JonathanNeal: the main use case is the one shown by the example in the spec: http://www.whatwg.org/specs/web-apps/current-work/#the-wbr-element

[01:12:02.0000] <boblet>
Hixie: I’ve got some introductory Japanese textbooks that do the kanji with kana and English. will mock up some nested ones and see how I go

[01:12:03.0000] <Hixie>
othermaciej: sounds like a stylistic concern. If my screen is small, I'd want it to wrap.

[01:12:04.0000] <boblet>
prolly not enough use cases to justify complexity

[01:12:05.0000] <JonathanNeal>
Hixie, checking

[01:13:00.0000] <othermaciej>
Hixie: yeah, I will look around and see if I find any real uses that seem more semantic

[01:13:01.0000] <Hixie>
boblet: yeah it's a hard call, when there are valid use cases but they are rare... it might just be something to do in a future version, or it might be something we never bother with because it's not that common, or it might be something that there are acceptable workarounds for (like nesting rubys). I'd love to see mockups or photos of real use cases though.

[01:14:00.0000] <JonathanNeal>
Weeeeird

[01:14:01.0000] <JonathanNeal>
Niche

[01:14:02.0000] <Hixie>
it is indeed niche, much like <br>

[01:15:00.0000] <Hixie>
but since it's already implemented and widely included in tutorials already, it's almost free to support

[01:15:01.0000] <Hixie>
so since there's a use case, why not

[01:15:02.0000] <othermaciej>
Hixie: do you recall what's partial about the acceptance on <http://www.w3.org/Bugs/Public/show_bug.cgi?id=9350> (the <wbr> bug)?

[01:15:03.0000] <othermaciej>
Hixie: wondering if I should check anything before I mark it CLOSED

[01:15:04.0000] <Hixie>
othermaciej: just the parenthetical, i think

[01:15:05.0000] <boblet>
Hixie: k. also thinking that for non-bopomofo use ruby is almost like a phrase version of small with pre-set styling (“side comments”)

[01:16:00.0000] <Hixie>
boblet: yeah. it's mostly intended for the CJK typographical ruby thing, the other uses are somewhat esoteric.

[01:16:01.0000] <Hixie>
othermaciej: i.e. partial acceptance of the argument, not of the change.

[01:17:00.0000] <boblet>
I was somewhat surprised by the WWW example

[01:17:01.0000] <boblet>
(in W3 Ruby)

[01:17:02.0000] <Hixie>
yeah me too to be honest

[01:17:03.0000] <Hixie>
you'll notice the html5 spec doesn't show any such examples

[01:17:04.0000] <Hixie>
:-)

[01:17:05.0000] <othermaciej>
Hixie: I would consider that fully accepted, but fine by me either way

[01:17:06.0000] <boblet>
heh

[01:18:00.0000] <boblet>
that’s what made me wonder if such examples were legit

[01:18:01.0000] <Hixie>
othermaciej: to be perfectly honest i marked it partially accepted because i have two buttons, one that says "accepted" and prefills the rationale as concuring with the reporter, and one that says "partially accepted" and leaves the rationale blank, and so since i wanted to write a more detailed rationale, i clicked the second one to save me the selecting the rationale and replacing it :-P

[01:19:00.0000] <othermaciej>
Hixie: lol

[01:19:01.0000] <Hixie>
boblet: yeah, i dunno. might be worth asking richard ishida or other people more expert in that area what they think.

[01:20:00.0000] <boblet>
will do

[01:35:00.0000] <JonathanNeal>
Oh man ... PeeWee Herman makes me smile.

[01:45:00.0000] <JonathanNeal>
MikeSmith has quit

[01:45:01.0000] <JonathanNeal>
There goes HTML5.

[01:45:02.0000] <othermaciej>
umm, just from IRC

[01:48:00.0000] <JonathanNeal>
There goes IRC.

[01:48:01.0000] <kennyluck>
He is now sleeping :)

[01:49:00.0000] <JonathanNeal>
There goes being awake.

[01:49:01.0000] <Bolkonskij>
good morning

[01:49:02.0000] <JonathanNeal>
There goes it being night.

[01:49:03.0000] <Bolkonskij>
Heh

[01:50:00.0000] <JonathanNeal>
Next thing you know <wbr> will be back in the spec.

[01:51:00.0000] <Hixie>
ok time for bed

[01:51:01.0000] <Hixie>
nn

[01:52:00.0000] <JonathanNeal>
Night Hixie!

[05:45:00.0000] <gsnedders>
Hixie: Well, really, we just need Python to be fixed :

[08:51:00.0000] <annevk>
why is http://dev.w3.org/csswg/css3-2d-transforms/#transform-property not comma-separated?

[09:05:00.0000] <JonathanNeal>
mornin'

[10:03:00.0000] <JonathanNeal>
Hi all!

[10:15:00.0000] <JonathanNeal>
Anyone here versed in zen html coding?

[10:15:01.0000] <JonathanNeal>
Where you write HTML like this: a#skip-to-main-content[href=#main-content]

[10:15:02.0000] <boblet>
woah, crazy. ruby-position:inline mentioned in #5 Properties Index, but not in #4.1 Ruby Positioning. Wonder how inline differs from right? http://www.w3.org/TR/css3-ruby/

[10:16:00.0000] <boblet>
JonathanNeal: I saw a screencast once

[10:17:00.0000] <JonathanNeal>
I just don't get how you move out of a selector.

[10:18:00.0000] <JonathanNeal>
Say I want to make <div><span></span></div><div><p></p></div>a

[10:20:00.0000] <boblet>
JonathanNeal: I’m pretty sure they did multiple blocks in the screencast. I’d guess whitespace (extra returns?)

[10:21:00.0000] <JonathanNeal>
(parens)

[10:21:01.0000] <tabatkins>
What do you mean by "move out of a selector", JonathanNeal?

[10:21:02.0000] <JonathanNeal>
"JonathanNeal, This is #html, not #wackyhtmltemplateengines."

[10:25:00.0000] <JonathanNeal>
tabatkins, I meant I didn't know how to write css that didn't constantly move down the chain

[10:25:01.0000] <JonathanNeal>
But I realized you can control the scope with parens

[10:25:02.0000] <boblet>
nice

[10:26:00.0000] <tabatkins>
JonathanNeal: Um?  You can do what with the who now?

[10:26:01.0000] <tabatkins>
Selectors move down the chain [or forward, with sibling selectors].  That's how they're made.

[10:26:02.0000] <tabatkins>
Are you trying to abuse selectors to do something crazy?

[10:31:00.0000] <JonathanNeal>
Yes, and I just did.

[10:32:00.0000] <JonathanNeal>
You've never heard of zen coding?

[10:32:01.0000] <JonathanNeal>
html>(head>meta[charset=UTF-8]+title)+body>#main-document>(a#skip-to-main-content[href=#main-content])+(header#banner[role=banner]>(hgroup#heading>(h1.site-title[href=@site-url@]>span)+(h2.page-title[href=@page-url@]>span))+(nav#navigation>h1+ul>li*3))+(div#content>(nav#breadcrumbs>(h1>span)+p)+div#main-content)+footer#content-info[role=content-info]>p

[10:32:02.0000] <JonathanNeal>
That's a complete html5 page with heading, navigation, breadcrumb, content, footer

[10:35:00.0000] <tabatkins>
That looks like a combination of disgusting, unreadable, and ridiculous.

[10:35:01.0000] <AryehGregor>
Parentheses don't work like that in CSS.  That looks like a syntax error to me.

[10:35:02.0000] <AryehGregor>
If it's meant to be an actual selector.

[10:38:00.0000] <annevk>
o_O

[10:38:01.0000] <lazni1>
it works backward from a selector: serialize a tree matching that ext-CSS

[10:38:02.0000] <jgraham>
I assume it is just supposed to be some way of generating a document structure

[10:39:00.0000] <jgraham>
That happens to reuse selector synatx, mostly

[10:39:01.0000] <boblet>
/me is surprised y’all are not up with what the kids are playing with these days

[10:39:02.0000] <boblet>
it’s just generating html verbiage from quasi-CSS, to cut down on typing

[10:39:03.0000] <annevk>
boblet, like engineer barbie?

[10:40:00.0000] <AryehGregor>
Oh, I see.

[10:40:01.0000] <AryehGregor>
That seems kind of pointless.  It's not that much shorter.

[10:40:02.0000] <AryehGregor>
Although it's true that the redundancy of closing tags in SGML-ish languages is annoying.

[10:41:00.0000] <boblet>
annevk: Google is my friend http://www.ni.com/news/releases/april0601.htm ;-)

[10:41:01.0000] <jgraham>
Reusing selector syntas to make things more readable is an... odd approach

[10:42:00.0000] <JonathanNeal>
annevk, exactly like engineer barbie

[10:42:01.0000] <jgraham>
You could of course use sexprs and cut out all the end tags

[10:42:02.0000] <tabatkins>
I've certainly seen that stuff before.  It's just horrifying when extended beyond toy examples.

[10:42:03.0000] <boblet>
TextMate does most of it for me, so the only appeal of zen coding is prolly making tables

[10:42:04.0000] <annevk>
boblet, it's not a joke anymore I believe, but it doesn't really matter :)

[10:43:00.0000] <boblet>
annevk: There’s a computer engineer Barbie, but I can’t find the one with the wrench & grease on her face

[10:44:00.0000] <JonathanNeal>
that would be a strange computer engineer.

[10:44:01.0000] <jgraham>
Yeah, who keeps a wrench on their face?

[10:46:00.0000] <annevk>
that is not how it works?

[10:46:01.0000] <JonathanNeal>
No idea.

[10:47:00.0000] <AryehGregor>
I use screwdrivers sometimes, but not sure where you'd use a wrench in computing.

[10:47:01.0000] <jgraham>
In other news I think I have a fix for the spec generation bug

[10:47:02.0000] <boblet>
Computer Engineer Barbie; “I ran the measurements on my leg-to-torso ratio against the 95 percentile, and the results are … well…”

[10:48:00.0000] <jgraham>
Nah, computer engineers son't know statistics

[10:48:01.0000] <jgraham>
*don't

[10:49:00.0000] <JonathanNeal>
Computer Engineer Barbie: "I will  be on vacation tomorrow."

[11:01:00.0000] <annevk>
is anyone able to dig up that email on default styles for form controls on the WHATWG list?

[11:01:01.0000] <annevk>
someone from Google wrote it I believe

[11:01:02.0000] <annevk>
(could've been a different list maybe, but I doubt it)

[11:01:03.0000] <JonathanNeal>
What do we want?  Self Closing DIVS and SPANS!  When do we want them?  NOW!

[11:02:00.0000] <annevk>
I'm trying to write a few words about HTML5 forms and it'd be nice to include a pointer to that research

[11:03:00.0000] <JonathanNeal>
html5 forms seems to be the big thing lately, please link me to that when you're done annevk

[11:03:01.0000] <JonathanNeal>
I'd love to learn what it is you folks are doing to them.

[11:05:00.0000] <jgraham>
OK, just updated pms.net 

[11:06:00.0000] <annevk>
e.g. stuff like http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2009-April/019478.html

[11:07:00.0000] <seutje>
cool, specs have to list every browsers behaviour? I figured it was meant to set a standard and hope browsers will addapt

[11:08:00.0000] <AryehGregor>
seutje, some specs try that, but they aren't particularly useful to anyone.

[11:09:00.0000] <AryehGregor>
HTML5 doesn't try to list every browser's behavior, but it tries to match their behavior when it's consistent.

[11:09:01.0000] <AryehGregor>
Or, at any rate, it doesn't require things that it doesn't think implementers will refuse to do.

[11:09:02.0000] <jgraham>
It tries to specify a single behaviour that, due to legacy, msut be almost the same as existing deployed behaviour from market leaders

[11:13:00.0000] <AryehGregor>
At least in cases where sites rely on the behavior.

[11:13:01.0000] <seutje>
I see

[11:14:00.0000] <JonathanNeal>
It would be cool if Simon's html5-elements page could load up individual pages for those elements, instead of having to #get-there

[11:14:01.0000] <seutje>
u have a funny way of putting it though

[11:14:02.0000] <seutje>
"it doesn't think.."

[11:14:03.0000] <jgraham>
Which is more cases than you can imagine

[11:14:04.0000] <seutje>
the spec became self-aware!

[11:15:00.0000] <jgraham>
(well it is hopefully as many cases as we can collectively imagine or there will still be cases with required but undefined behaviour)

[11:15:01.0000] <jgraham>
(so what I really mean is that it is more cases then most people naively imagine)

[11:15:02.0000] <jgraham>
*than

[11:17:00.0000] <gsnedders>
/me thought jgraham was going away

[11:18:00.0000] <gsnedders>
Yet he fixed the bug

[11:18:01.0000] <jgraham>
gsnedders: Maybe I am using my telekinetic powers to bug fix by remotly typing on someone else's keyboard. Maybe it is your keyboard. Have you been watching for unexpected movements?

[11:19:00.0000] <gsnedders>
Well, no.

[11:19:01.0000] <jgraham>
Tell you what, I fixed the bug so you can fix the tests :)

[11:22:00.0000] <gsnedders>
However, I need to go and cook now ;P

[11:23:00.0000] <JonathanNeal>
gsnedders, master chef

[11:26:00.0000] <seutje>
hmm, that reminds me, shouldn't forget to eat

[11:31:00.0000] <dglazkov>
Hixie: is this the latest? http://dev.w3.org/html5/spec/forms.html#the-select-element

[11:32:00.0000] <annevk>
should be

[11:32:01.0000] <dglazkov>
Hixie: I want to add a corner case for handling default-selected item in select element

[11:32:02.0000] <dglazkov>
annevk: http://trac.webkit.org/export/57003/trunk/LayoutTests/fast/forms/select-change-popup-to-listbox.html

[11:33:00.0000] <dglazkov>
it's a bit weird, but all browsers seem to agree, so probably should be spec'd

[11:33:01.0000] <dglazkov>
for size == 1, first item is selected by default

[11:33:02.0000] <dglazkov>
if no other items are selected

[11:34:00.0000] <dglazkov>
but then if you change the size, the default selection is preserved.

[11:34:01.0000] <dglazkov>
annevk: sorry, I did't try on Opera!

[11:34:02.0000] <dglazkov>
:)

[11:34:03.0000] <annevk>
no worries

[11:35:00.0000] <dglazkov>
I wonder how mobile browsers deal with this?

[11:36:00.0000] <annevk>
http://dev.w3.org/html5/spec/the-xhtml-syntax.html#the-select-element-0

[11:36:01.0000] <annevk>
defines UI

[11:36:02.0000] <annevk>
i.e. what to render for various values of size

[11:38:00.0000] <annevk>
I think it is defined also that the first is selected

[11:38:01.0000] <annevk>
"If the multiple attribute is absent, whenever there are no option elements in the select element's list of options that have their selectedness set to true, the user agent must set the selectedness of the first option element in the list of options in tree order that is not disabled, if any, to true."

[11:38:02.0000] <annevk>
(that's in section 4.10 again)

[11:41:00.0000] <dglazkov>
annevk: great! do you think it's worth outlining the case of preserving the selectedness when the size changes?

[11:41:01.0000] <dglazkov>
(that's the one I am interested about)

[11:42:00.0000] <dglazkov>
hang on.. reading this one more time :)

[11:42:01.0000] <annevk>
I suppose, but that would have to be added to the rendering section; you could add a comment via http://whatwg.org/html#the-select-element-0

[11:43:00.0000] <annevk>
(if you fill the thing in at the bottom that will file a bug for you)

[11:44:00.0000] <annevk>
(not sure how many other such liveness things there are though and whether they all need to be pointed out; in general nothing changes unless stated otherwise)

[11:52:00.0000] <JonathanNeal>
So what's so bad about allowing xml style self-closing on all elements?

[11:52:01.0000] <JonathanNeal>
Or on divs and spans?

[11:53:00.0000] <Philip`>
Legacy compatibility

[11:54:00.0000] <AryehGregor>
JonathanNeal, no existing browser parses them that way in text/html.

[11:54:01.0000] <AryehGregor>
And none can change, because there are pages that use the self-closing syntax but actually rely on the fact that it's not self-closing.

[11:55:00.0000] <tabatkins>
I've never been a fan of the term "rely" in that context, because it implies they did it on purpose, when really they just did it, then hacked around it until it sort of worked.

[11:56:00.0000] <Philip`>
Why do you think reliance is purposeful?

[11:57:00.0000] <Philip`>
I rely on lots of things that I've never even considered

[11:57:01.0000] <dglazkov>
annevk: actually I think the spec is perfectly fine now that I read it

[11:57:02.0000] <dglazkov>
(read it vs. skim through it)

[11:57:03.0000] <Philip`>
and I can't think of a better word than "rely" for them

[12:00:00.0000] <tabatkins>
Hmm, not sure what a better term would be.

[12:07:00.0000] <annevk>
dglazkov, cool

[12:17:00.0000] <JonathanNeal>
AryehGregor, so you don't do it so it's more compatible with the old html version?

[12:18:00.0000] <AryehGregor>
JonathanNeal, there is no "old HTML version".  HTML5 is the one and only specification to fully specify the details of the text/html format that has been used for years.

[12:18:01.0000] <JonathanNeal>
ok

[12:18:02.0000] <AryehGregor>
All these implementation requirements (or variants there of) were a de facto standard for a long time, now they're becoming de jure.

[12:19:00.0000] <AryehGregor>
Previous HTML specs just didn't say anything about error handling by implementations.

[12:20:00.0000] <AryehGregor>
So <div /> was non-conforming in HTML prior to XHTML, and browsers consistently ignored the "/" (which was allowed by the HTML specs of the time).

[12:21:00.0000] <JonathanNeal>
So it's just a property of hypertext to not have self-closing elements?

[12:21:01.0000] <JonathanNeal>
or bi-closing elements?

[12:21:02.0000] <AryehGregor>
The closing "/" was not valid in any version of HTML prior to XHTML 1.0.

[12:22:00.0000] <AryehGregor>
It's an XML feature, not a text/html feature.

[12:22:01.0000] <AryehGregor>
It would raise validation errors if you used it, which no one did, because it did nothing anyway.

[12:23:00.0000] <AryehGregor>
XHTML allowed the self-closing feature, but browsers didn't implement it in text/html for compatibility reasons.

[12:23:01.0000] <AryehGregor>
And still don't, and presumably never will.

[12:24:00.0000] <Philip`>
Actually it wouldn't raise validation errors, because SGML-based HTML4 parsers (e.g. validator.w3.org) consider <br/> to be correct and equivalent to <br>&gt; or whatever it is

[12:25:00.0000] <JonathanNeal>
Right, and there is meta and link too

[12:25:01.0000] <JonathanNeal>
and img

[12:25:02.0000] <Philip`>
It's never meant "self-closing" in non-X HTML

[12:25:03.0000] <AryehGregor>
Oh, really?  Interesting.

[12:26:00.0000] <Philip`>
SHORTTAG and NET are relevant keywords

[12:27:00.0000] <Philip`>
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2004-August/001940.html

[12:27:01.0000] <Philip`>
(The W3C validator does warn about these things now, though)

[12:30:00.0000] <JonathanNeal>
So <br> <img> are bad?

[12:32:00.0000] <Philip`>
No

[12:32:01.0000] <Philip`>
Why would they be?

[12:32:02.0000] <JonathanNeal>
Ah, so it's because they're NOT self-closing, right?

[12:32:03.0000] <Philip`>
I have no idea what you're talking about :-)

[12:33:00.0000] <JonathanNeal>
<img> is okay because it's not an element that opens and closes, it just is.

[12:33:01.0000] <Philip`>
You've always been able to write "<br>" and "<img>" in HTML, because that's how the syntax is defined

[12:33:02.0000] <tabatkins>
JonathanNeal: In old-style SGML, the / was a way of closing a tag without explicitly using the end tag.

[12:33:03.0000] <JonathanNeal>
So ... <br /> is also valid.

[12:33:04.0000] <JonathanNeal>
So what would be bad about <div /> then?

[12:33:05.0000] <tabatkins>
So you could, frex, do "<div>stuff goes here/" and be equivalent to "<div>stuff goes here</div>"

[12:33:06.0000] <Hixie>
not quite

[12:34:00.0000] <Hixie>
<div/stuff goes here/

[12:34:01.0000] <Hixie>
or <br/ (since there's no end tag, no end / either)

[12:34:02.0000] <tabatkins>
Ah yeah, right.

[12:34:03.0000] <JonathanNeal>
No > ?

[12:34:04.0000] <tabatkins>
Nope.

[12:34:05.0000] <Hixie>
the first / takes places of the / and the next / takes place of the entire end tag

[12:34:06.0000] <Hixie>
er

[12:34:07.0000] <Hixie>
let me retype that

[12:34:08.0000] <tabatkins>
Which means that, if browsers were conforming SGML parsers, you'd see a linebreak followed by a > character in your page.

[12:35:00.0000] <Hixie>
the first / takes place of the > and the next / takes place of the entire end tag

[12:35:01.0000] <JonathanNeal>
Wow!  That's not implemented that way in any browser, is it?

[12:35:02.0000] <AryehGregor>
No, browsers don't actually implement SGML.

[12:36:00.0000] <Hixie>
there's a reason html5 gives up on the sgml fiction :-)

[12:36:01.0000] <JonathanNeal>
So, you can do <div /> because it wouldn't be properly conforming to something no browsers actually implement?

[12:36:02.0000] <JonathanNeal>
*can't

[12:36:03.0000] <tabatkins>
No, you can't do <div /> because it looks exactly like "<div>" to all browsers.

[12:36:04.0000] <JonathanNeal>
Well, in HTML4.

[12:36:05.0000] <tabatkins>
[That is, it looks like an open tag.]

[12:36:06.0000] <Hixie>
here, if you want to blow your mind: http://damowmow.com/playground/mental.html http://damowmow.com/playground/not-html-yet-valid.html 

[12:37:00.0000] <Hixie>
both of those validate in an html4 validator

[12:37:01.0000] <tabatkins>
Html nothing.  I'm talking about browsers here, not languages.  All existing browsers, current and legacy, treat it like that.

[12:37:02.0000] <tabatkins>
[Unless they are specifically in an XHTML mode, of course.[

[12:38:00.0000] <tabatkins>
So allowing <div /> as a short form of <div></div> wouldn't be backwards compatible in any relevant browser straight away, let alone all the pages that accidentally use <div /> and would be screwed up if it actually acted like <div></div>.

[12:39:00.0000] <annevk>
/me wonders when the SGML interest will die off

[12:40:00.0000] <tabatkins>
Dammit, JonathanNeal.  I didn't realize you'd dropped, and typed several lines at you.  >_<

[12:40:01.0000] <JonathanNeal>
tabatkins, not only did i drop

[12:40:02.0000] <JonathanNeal>
my whole computer restarted

[12:40:03.0000] <tabatkins>
That's fun.

[12:40:04.0000] <JonathanNeal>
hixie's links really blew my mind

[12:40:05.0000] <JonathanNeal>
and lappy's.

[12:40:06.0000] <tabatkins>
Anyway, check the logs.  I don't want to retype those.

[12:40:07.0000] <JonathanNeal>
I was like "This is a cell" and my computer was like "this fuuuuuuuuuuuuuuu"

[12:40:08.0000] <JonathanNeal>
Where do I find the logs?

[12:40:09.0000] <Philip`>
tabatkins: There's this fancy new concept known as "copy and paste" which you could use

[12:40:10.0000] <tabatkins>
Check the topic.

[12:41:00.0000] <tabatkins>
Philip`: Too fancy for me.

[12:41:01.0000] <JonathanNeal>
checking

[12:42:00.0000] <JonathanNeal>
tabatkins, aren't nav and video not backwards compatible?

[12:42:01.0000] <JonathanNeal>
I believe <section> causes fuglies in ff2 and ie6-7-8

[12:42:02.0000] <tabatkins>
They have fallback.

[12:43:00.0000] <JonathanNeal>
they don't have fallback ... not in those browsers.  What do you mean by "fallback"?

[12:43:01.0000] <tabatkins>
ff2 is luckily ignorable.

[12:43:02.0000] <tabatkins>
ie6-8 has the hack, which is nice.

[12:43:03.0000] <JonathanNeal>
ie6-7-8?

[12:44:00.0000] <JonathanNeal>
so it's okay because there's a hack? :-|

[12:44:01.0000] <tabatkins>
The hack makes it usable *now*, so... yes?  Things are a lot less appealing when there's no way to use them until all browsers update.

[12:45:00.0000] <tabatkins>
That doesn't mean we'll avoid *all* things without fallback, but they need a better reason to exist.

[12:45:01.0000] <tabatkins>
Self-closing syntax has no real reason to exist besides "It's what XML does."

[12:45:02.0000] <JonathanNeal>
ie supports it.

[12:45:03.0000] <Philip`>
No it doesn't

[12:45:04.0000] <Philip`>
except sometimes

[12:46:00.0000] <Hixie>
ie's parsing behaviour is seriously messed up

[12:46:01.0000] <Hixie>
and very internally inconsistent

[12:46:02.0000] <Philip`>
like in unknown elements that are recognised due to being declared with xmlns:foo or with the createElement hack, or due to being any normal unknown element in IE9

[12:46:03.0000] <Philip`>
(i.e. <section/> closes the element in IE9)

[12:47:00.0000] <Philip`>
(but <div/> doesn't)

[12:49:00.0000] <JonathanNeal>
Ah, I see.

[12:50:00.0000] <JonathanNeal>
I tried This is <span style="background: #0F0; padding: 10px;" /> self closing. and it failed in all browsers.

[12:51:00.0000] <JonathanNeal>
So it can't be implemented because it would break on all browsers.  You're lucky you got any new elements in :)

[12:51:01.0000] <JonathanNeal>
Considering how ff used to handle these things.

[12:51:02.0000] <tabatkins>
If FF still handles unknown elements as badly as it used to, I suspect the new elements would have had a much more uphill battle.

[12:52:00.0000] <JonathanNeal>
aye

[13:00:00.0000] <tabatkins>
/me goes to learn himself a python for great justice.

[13:10:00.0000] <mr_daniel>
is it really possible that Firefox still has no support for websockets? Just downloaded the newest version 3.6.3, but the included websocket chat in Jetty fails to run

[13:12:00.0000] <AryehGregor>
mr_daniel, https://bugzilla.mozilla.org/show_bug.cgi?id=472529

[13:13:00.0000] <AryehGregor>
Work ongoing.

[13:15:00.0000] <Hixie>
mr_daniel: websockets isn't really ready yet, the wg is still discussing the protocol

[13:20:00.0000] <annevk>
Philip`, apparently their plan was to conform to HTML5

[13:20:01.0000] <annevk>
Philip`, maybe you should file bugs?

[13:25:00.0000] <MikeSmith>
mr_daniel: if you are comfortable building from sources, you can download the patch from that bug and apply it and build and it should work

[13:26:00.0000] <MikeSmith>
Hixie: it seems ready enough for experimenting with at least

[13:26:01.0000] <MikeSmith>
and for playing Quake2

[13:44:00.0000] <Hixie>
MikeSmith: we're making non-backwards-compatible changes to the protocol, so everything will break

[13:46:00.0000] <MikeSmith>
Hixie: yeah, I realize

[13:46:01.0000] <MikeSmith>
but it's understandable that people are anxious to try it out in the mean time in spite of any instability of the protocol spec

[13:47:00.0000] <Hixie>
sure, i've done it myself :-)

[13:47:01.0000] <Hixie>
in my spare time i'm writing a web-based mud using websockets

[13:48:00.0000] <annevk>
mud?

[13:48:01.0000] <annevk>
multi-user d...?

[13:48:02.0000] <paul_irish>
daemon

[13:50:00.0000] <MikeSmith>
Hixie: you should make a mud where the character is a spec editor who needs to try to create spec for Web-platform features and get them standardized and implemented

[13:50:01.0000] <Hixie>
heh

[13:50:02.0000] <Hixie>
annevk: multi-user dungeon

[13:51:00.0000] <paul_irish>
oops. :x   http://en.wikipedia.org/wiki/MUD

[13:51:01.0000] <MikeSmith>
/me thought it was "domain"

[13:53:00.0000] <MikeSmith>
/me reads annevk blog and finds out about https://wiki.mozilla.org/User:Mounir.lamouri/HTML5_Forms

[13:53:01.0000] <MikeSmith>
would be interesting to know how soon active work might get started on all the items in the red rows

[13:54:00.0000] <MikeSmith>
oh, I see that some of them do have bugs associated with them already

[13:56:00.0000] <MikeSmith>
I wonder if anybody knows/remembers how long it took Opera engineering to finish their implementation of Web Forms 2 at the time

[13:56:01.0000] <MikeSmith>
or how many people-hours it took

[14:03:00.0000] <MikeSmith>
/me wonders if recent TC39 f2f spent any time discussing https://cvs.khronos.org/svn/repos/registry/trunk/public/webgl/doc/spec/TypedArray-spec.html or other related proposals

[14:03:01.0000] <annevk>
one person on and off for half a year or so?

[14:03:02.0000] <annevk>
/me doesn't remember the details

[14:03:03.0000] <MikeSmith>
annevk: ok

[14:04:00.0000] <MikeSmith>
it seems like Webkit devs have been working on it very actively for more than 6 months now

[14:04:01.0000] <annevk>
btw, http://en.wikipedia.org/wiki/MUD

[14:05:00.0000] <annevk>
it could be one person for a year on and off

[14:06:00.0000] <annevk>
but not more than a year and not more than one person developing; I created tests together with someone else

[14:06:01.0000] <MikeSmith>
ok

[14:06:02.0000] <annevk>
but then our UI is not great...

[14:06:03.0000] <annevk>
but then WebKit does not have UI yet I believe...

[14:06:04.0000] <annevk>
oh well

[14:07:00.0000] <MikeSmith>
I guess it's hard to decide what's the appropriate UI for some of these controls

[14:07:01.0000] <MikeSmith>
btw, W3C Bugzilla instance was upgraded today

[14:08:00.0000] <tabatkins>
A decent rule, of course, is "Do whatever the most popular equivalent jQuery plugin does."

[14:08:01.0000] <annevk>
well, you need a UI engineer, not someone who's typically deeply involved in DOM/parsing/etc. code

[14:08:02.0000] <MikeSmith>
right, exactly

[14:10:00.0000] <estellevw>
i've been doing some testing of the various form elements in recent browser releases. webkit seems to understand from a DOM perspective new input types, but does nothing with them

[14:10:01.0000] <othermaciej>
WebKit's UI for many of the controls is not so great

[14:10:02.0000] <estellevw>
in terms of UI

[14:11:00.0000] <annevk>
estellevw, yeah, that is somewhat troublesome

[14:12:00.0000] <annevk>
estellevw, but as long as nobody ships final releases I guess it could work

[14:13:00.0000] <estellevw>
no, i have been testing the actual releases, not the nightly builds... so it is what has been released

[14:13:01.0000] <MikeSmith>
annevk: or even not just a UI engineer, but a UX expert too

[14:14:00.0000] <annevk>
just a little insight might have done wonders, but we were short on time too

[14:14:01.0000] <annevk>
and with no competition and not much real world usage we didn't move it beyond experimental

[14:15:00.0000] <annevk>
estellevw, i.e. in Safari 4 and Chrome 4?

[14:15:01.0000] <estellevw>
for example, chrome supports color, date and email, but doesn't validate

[14:15:02.0000] <MikeSmith>
yeah, but all told, it's still a pretty good implementation

[14:15:03.0000] <estellevw>
yeah, exactly

[14:15:04.0000] <annevk>
ugh :/

[14:15:05.0000] <annevk>
MikeSmith, yeah, it's pretty cool

[14:15:06.0000] <MikeSmith>
Hixie: W3C bugzilla is down

[14:16:00.0000] <estellevw>
i've been compiling results here: http://www.standardista.com/html5/html5-web-forms (not finished ... a work in progress)

[14:16:01.0000] <MikeSmith>
Hixie: systems team is working on it

[14:16:02.0000] <miketaylr>
estellevw: this might be helpful for you: http://www.miketaylr.com/code/input-type-attr.html

[14:16:03.0000] <miketaylr>
also, http://www.miketaylr.com/code/html5-forms-ui-support.html

[14:16:04.0000] <estellevw>
MIke- i came across your site last week, and found it extremely useful

[14:17:00.0000] <miketaylr>
ahh, k

[14:17:01.0000] <estellevw>
and then started visually comparing

[14:17:02.0000] <miketaylr>
yeah i'm just doing feature testing

[14:17:03.0000] <estellevw>
so, assumed DOM support from your excellent resource, and then sandboxed each 

[14:18:00.0000] <estellevw>
nice to virtually meet you, by the way

[14:18:01.0000] <miketaylr>
:)

[14:18:02.0000] <miketaylr>
same

[14:19:00.0000] <MikeSmith>
are all 192.* IP addresses reserved (or whatever it's called), or just 192.168.* ones? 

[14:22:00.0000] <annevk>
oh hey, someone is picking up the ball I sort of dropped: http://mail.apps.ietf.org/ietf/charsets/msg01877.html

[14:24:00.0000] <annevk>
MikeSmith, do you know what happened to http://lists.w3.org/Archives/Public/ietf-charsets/ ?

[14:25:00.0000] <annevk>
it scores highest on Google for ietf-charsets but has dropped the ball seemingly after 2004

[14:25:01.0000] <MikeSmith>
annevk: wfm

[14:25:02.0000] <MikeSmith>
oh, I see

[14:25:03.0000] <MikeSmith>
no idea

[14:25:04.0000] <MikeSmith>
it's an IETF list, just hosted at W3C

[14:26:00.0000] <MikeSmith>
maybe Richard or Martin Duerst would know

[14:26:01.0000] <annevk>
the actual list is hosted by IANA

[14:26:02.0000] <annevk>
well, it uses an IANA email address

[14:27:00.0000] <MikeSmith>
other IANA lists are hosted somewhere else

[14:27:01.0000] <MikeSmith>
like the language-subtags list

[14:27:02.0000] <MikeSmith>
/me tries to find

[14:27:03.0000] <MikeSmith>
http://www.alvestrand.no/mailman/listinfo/ietf-languages

[14:28:00.0000] <MikeSmith>
but that seems to be the same list as ietf-languages⊙io

[14:29:00.0000] <annevk>
W3C lists are nice

[14:29:01.0000] <Philip`>
MikeSmith: http://en.wikipedia.org/wiki/Private_network#Private_IPv4_address_spaces

[14:29:02.0000] <MikeSmith>
Philip`: ah, thanks

[14:30:00.0000] <annevk>
WHATWG mailing list archive links have been broken at least once (and not fixed) and now several different entry points are used

[14:30:01.0000] <annevk>
it's a bit of a mess

[14:30:02.0000] <MikeSmith>
so it is only the 192.168.* range

[14:31:00.0000] <Philip`>
MikeSmith: Yep

[14:31:01.0000] <MikeSmith>
annevk: ietf-charsets⊙io 

[14:31:02.0000] <MikeSmith>
but that appears to lead nowhere

[14:33:00.0000] <annevk>
MikeSmith, the actual archive is now at  http://mail.apps.ietf.org/ietf/charsets/

[14:33:01.0000] <annevk>
MikeSmith, just wondering why the W3C one was not kept updated

[14:34:00.0000] <MikeSmith>
annevk: dunno, I'll ask RIchard

[14:41:00.0000] <jgraham>
Hixie: Did you try regenning the spec recently? Did the non-BMP entity problems go away?

[14:42:00.0000] <MikeSmith>
jgraham: fwiw, I can try it now

[14:43:00.0000] <Dashiva>
"Developers should know better than to use a GET request for teleportation."

[14:44:00.0000] <annevk>
for that td:last-child:hover effect a transform might look better

[14:45:00.0000] <annevk>
though might also work less well come to think of it

[14:45:01.0000] <annevk>
unless we get some hit testing control

[14:45:02.0000] <MikeSmith>
annevk: so I'm told the w3c ietf-charsets list was shut down at the beginning of 2004 due disuse and to being overwhelmed with spam

[14:46:00.0000] <annevk>
kk, too bad

[14:46:01.0000] <annevk>
with the ietf mailing list archive you get everything on a single page

[14:46:02.0000] <jgraham>
MikeSmith: OK

[14:46:03.0000] <annevk>
it 1) takes ages to load and 2) doesn't browse very well

[14:47:00.0000] <jgraham>
/me is futzing about with python on the server at the moment as well so it is not impossible that something will accidentially break

[14:47:01.0000] <jgraham>
(it shouldn't but...)

[14:48:00.0000] <gsnedders>
(I wouldn't trust jgraham.)

[14:49:00.0000] <MikeSmith>
annevk: if the list is no longer spam-inundated, I can ask for the mirroring to be turned back on

[14:49:01.0000] <annevk>
nah it's okay, thanks

[14:49:02.0000] <JonathanNeal>
hey again

[14:50:00.0000] <jgraham>
Dashiva: where is that from?

[14:50:01.0000] <Dashiva>
http://code.google.com/p/chromium/issues/detail?id=31482

[14:50:02.0000] <annevk>
going to bed, hopefully I can sleep ten or so hours and defeat da jetlag

[14:50:03.0000] <MikeSmith>
annevk: oyasumi

[14:50:04.0000] <annevk>
/me only slept like two hours on the plane...

[14:51:00.0000] <annevk>
MikeSmith, ta!

[14:56:00.0000] <MikeSmith>
jgraham: this only affects the serializer behavior, right?

[14:56:01.0000] <gsnedders>
MikeSmith: Yes

[14:56:02.0000] <MikeSmith>
ok

[14:59:00.0000] <Hixie>
jgraham: trying now

[15:00:00.0000] <jgraham>
MikeSmith: You will need to wait a bit I think

[15:00:01.0000] <MikeSmith>
jgraham: hai

[15:00:02.0000] <jgraham>
The python messing changed some paths, I am just fixing now

[15:00:03.0000] <jgraham>
(well actually I am taking the oppertunity to upgrade some of the installed versions of things at the same time)

[15:05:00.0000] <JonathanNeal>
I wish there was a way to tell the w3 validator to ignore certain things, like X-UA-Compatible

[15:05:01.0000] <JonathanNeal>
Just because it makes me sad when I see the red tape.

[15:05:02.0000] <Hixie>
jgraham: getting 500s

[15:10:00.0000] <gsnedders>
/me blames jgraham, looks innocent himself

[15:13:00.0000] <jgraham>
MikeSmith, Hixie: Should be back up now

[15:14:00.0000] <jgraham>
JonathanNeal: That is sort of the point

[15:14:01.0000] <MikeSmith>
jgraham: hg pull says no updates found

[15:14:02.0000] <jgraham>
MikeSmith: Pull of what?

[15:14:03.0000] <JonathanNeal>
jgraham, but poor IE!

[15:14:04.0000] <MikeSmith>
um, from html5lib repo

[15:15:00.0000] <MikeSmith>
I thought that's where you were making the changes

[15:16:00.0000] <gsnedders>
He did a lot earlier

[15:16:01.0000] <jgraham>
MikeSmith: I made a change there earlier (rev. 4113ad9d98) but I was just futzing with the server

[15:16:02.0000] <jgraham>
now

[15:16:03.0000] <jgraham>
I needed to upgrade python and that changed the default python path

[15:16:04.0000] <MikeSmith>
aha

[15:17:00.0000] <MikeSmith>
I was confused

[15:17:01.0000] <gsnedders>
MikeSmith: like normal

[15:17:02.0000] <MikeSmith>
heh

[15:17:03.0000] <MikeSmith>
I had forgotten that this was a change needed to gsnedders adonais tool

[15:18:00.0000] <jgraham>
(the need to upgrade python was entirely unrelated to this issue)

[15:19:00.0000] <MikeSmith>
I guess I will need to wait til Hixie checks in again

[15:20:00.0000] <MikeSmith>
because I'm just working with his post-adonais generated output

[15:20:01.0000] <jgraham>
OK, well I am going to sleep in a moment so if it is broken it will have to stay broken for a while

[15:23:00.0000] <JonathanNeal>
http://achecker.ca/checker/index.php <html lang="en"> fails ... is "en" not valid ISO 639 specification for the HTML lang attribute?

[15:24:00.0000] <Hixie>
regenning again

[15:26:00.0000] <Hixie>
hrm i overloaded pms

[15:26:01.0000] <Hixie>
let me add more delays and try again

[15:29:00.0000] <Hixie>
hm, wait, it's not my fault

[15:29:01.0000] <Hixie>
jgraham: i'm getting xml syntax errors

[15:29:02.0000] <Hixie>
jgraham: when doing the specs that have the annotations

[15:30:00.0000] <Hixie>
jgraham: http://www.whatwg.org/specs/web-apps/current-work/.w3c-html-core/Overview.html

[15:30:01.0000] <jgraham>
Oh. That sounds bad

[15:32:00.0000] <jgraham>
Hixie: Would it be a problem to turn off the annotations until I have a chance to fix?

[15:34:00.0000] <Hixie>
no

[15:34:01.0000] <Hixie>
not for me, at least

[15:34:02.0000] <Hixie>
can't speak for everyone else... there's always someone who'll fine a problem with something

[15:35:00.0000] <Hixie>
what parameter do i need to remove to stop it annotating?

[15:35:01.0000] <Hixie>
just annotation= ?

[15:35:02.0000] <Hixie>
or annotate_w3c_issues=?

[15:35:03.0000] <Hixie>
both?

[15:36:00.0000] <Hixie>
i prefixed both with xxx-, we'll see if that helps

[15:38:00.0000] <Hixie>
MikeSmith: checking in

[15:39:00.0000] <MikeSmith>
OK

[15:39:01.0000] <MikeSmith>
http://lists.w3.org/Archives/Public/www-validator/2010Apr/0011.html is entertaining

[15:39:02.0000] <MikeSmith>
Jukka Korpela

[15:39:03.0000] <MikeSmith>
'"HTML 5" is codename for a sketch of a draft of a working paper. "Validating" against something as vague as that is pointlss.'

[15:39:04.0000] <jgraham>
Hixie: It might well work with annotations now

[15:40:00.0000] <MikeSmith>
dude is such a sweetheart, you just gotta love him

[15:40:01.0000] <Hixie>
jgraham: k trying

[15:40:02.0000] <Hixie>
MikeSmith: i guess he hasn't looked at the spec recently

[15:41:00.0000] <Hixie>
MikeSmith: or he draws WAY bigger sketches than i do

[15:41:01.0000] <MikeSmith>
heh

[15:41:02.0000] <tabatkins>
I can't imagine the draft of a working paper that HTML5 is a sketch of, let alone the working paper itself.

[15:41:03.0000] <Hixie>
i mean, there are lots of ways to dismiss html5, but calling it a "sketch" is somewhat silly

[15:41:04.0000] <MikeSmith>
he's an expert, so he's not obligated to actually read specs

[15:42:00.0000] <Hixie>
/me lies to his script to get it to actually regen the specs even though he didn't change anything

[15:42:01.0000] <tabatkins>
Hixie: irssi actually isn't that bad.  I'll continue to use that on my linux boxes I guess.

[15:43:00.0000] <Hixie>
dude i've been telling you that for ages :-P

[15:43:01.0000] <tabatkins>
If "for ages" you mean "once, last week", then yes.

[15:44:00.0000] <Hixie>
jgraham: looks like it worked

[15:44:01.0000] <Hixie>
tabatkins: i live in the now, anything in the past happened ages ago :-P

[15:44:02.0000] <Hixie>
MikeSmith: ok, looks like the astral plane stuff is all fixed

[15:46:00.0000] <MikeSmith>
/me runs splitter on W3C copy

[15:47:00.0000] <MikeSmith>
yippee

[15:47:01.0000] <MikeSmith>
http://dev.w3.org/html5/spec/named-character-references.html

[15:47:02.0000] <MikeSmith>
back to normal

[15:47:03.0000] <Hixie>
now if only browsers supported absolute positioning correctly in columns

[15:47:04.0000] <Hixie>
so the hover effect worked in more than the first column

[15:48:00.0000] <JonathanNeal>
It's kinda weird how the letter disappears

[15:48:01.0000] <JonathanNeal>
when you hover

[15:49:00.0000] <JonathanNeal>
Oh ... I see ... it's just not working right in Chrome.

[15:49:01.0000] <MikeSmith>
people should just ignore those other columns

[15:49:02.0000] <tabatkins>
Wow, that's... quite a table.

[15:49:03.0000] <MikeSmith>
http://www.w3.org/Bugs/Public/ appears to back in action too

[15:50:00.0000] <tabatkins>
JonathanNeal: ?_?

[15:51:00.0000] <JonathanNeal>
tabatkins, what are you ?_? ing?

[15:51:01.0000] <tabatkins>
Your nick change.

[15:51:02.0000] <paul_irish>
tabatkins: http://paulirish.com/i/ec20.png 

[15:51:03.0000] <tabatkins>
Ah, I see.

[15:52:00.0000] <JonathanNeal>
And http://www.imdb.com/title/tt0087182/quotes#qt0328198

[15:52:01.0000] <JonathanNeal>
And it was Paul ... there was no resisting.

[15:53:00.0000] <MikeSmith>
/me wonders what a "mouse locking feature" is and why WebGL-based Quake2 implementation would highlight the need for one

[15:54:00.0000] <Philip`>
MikeSmith: You should be able to spin around by moving your mouse rightwards a lot

[15:54:01.0000] <Philip`>
for a longer distance than the width of your screen

[15:55:00.0000] <Philip`>
without the mouse leaving the web page window and stopping controlling the game

[15:55:01.0000] <MikeSmith>
I see

[15:55:02.0000] <MikeSmith>
isn't this what joysticks are for?

[15:56:00.0000] <AryehGregor>
Because joysticks are much better supported by the web platform, right?

[15:56:01.0000] <MikeSmith>
heh

[15:56:02.0000] <tabatkins>
Do you own a joystick that works with your computer, MikeSmith?

[15:56:03.0000] <AryehGregor>
(and no, no one uses joysticks for FPSes on the PC, only for flight sims or something)

[15:57:00.0000] <MikeSmith>
tabatkins: I don't own any joystick at all

[15:57:01.0000] <MikeSmith>
unless you meant the question figuratively

[15:57:02.0000] <Philip`>
I own two dead joysticks

[15:57:03.0000] <Philip`>
I think the modern term you want is "joypads" or "gamepads" or whatever

[15:58:00.0000] <Philip`>
which people are more likely to own

[15:58:01.0000] <Philip`>
but they don't exactly work on the web either

[15:58:02.0000] <divya>
A question, In CSS3 the display property refers to block/inline elements http://www.w3.org/TR/css3-layout/#declaring-templates-the-display-property but HTML5 does not define block/inline elements. Is there something I am missing? 

[15:58:03.0000] <Philip`>
and they're rubbish compared to mouse+keyboard for FPSes anyway

[15:59:00.0000] <tabatkins>
divya: In terms of CSS, a block/inline element is one which has display:block/inline set on it in the default UA stylesheet.

[15:59:01.0000] <AryehGregor>
divya, look in the rendering section.

[15:59:02.0000] <AryehGregor>
http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#rendering

[16:00:00.0000] <divya>
tabatkins: so does this mean display no longer alters the "content model" of an element? 

[16:00:01.0000] <Hixie>
a block/inline element in CSS is one which has display:block/inline as its computed value, not necessarily in the UA style sheet

[16:00:02.0000] <Hixie>
divya: it never did

[16:01:00.0000] <tabatkins>
Hixie is more correct than I was.  I was referring to what elements were by default.

[16:02:00.0000] <Hixie>
technically even its _used_ style, not computed style, though i think with 'display' they're always the same

[16:02:01.0000] <Hixie>
(or does 'position' not affect the computed style?)

[16:03:00.0000] <Hixie>
an element with display:inline;position:absolute is block-level

[16:03:01.0000] <Hixie>
ok yeah for such an element the computed value is 'block'

[16:03:02.0000] <Hixie>
so yes, computed value

[16:05:00.0000] <divya>
Hixie: I was under the illusion display can alter the content model because of this sentence in the HTML4.01 spec http://www.w3.org/TR/html401/struct/global.html#h-7.5.3 "Style sheets provide the means to specify the rendering of arbitrary elements, including whether an element is rendered as block or inline. "

[16:05:01.0000] <tabatkins>
Rendered as a block or inline, yes.  That doesn't change what kind of elements are legal to put inside of it, though.

[16:06:00.0000] <Hixie>
divya: yeah, that confusion is why i stopped using "block" and "inline" in html5

[16:06:01.0000] <Hixie>
it confused a lot of people

[16:06:02.0000] <Hixie>
you're not alone, by a long shot :-)

[16:07:00.0000] <divya>
Hixie: thanks, that is a comfort :) 

[16:07:01.0000] <Hixie>
:-)

[16:08:00.0000] <divya>
so in a sense display:inline/block are legacy CSS syntaxes with no real-world relationship with the elements?

[16:09:00.0000] <tabatkins>
I'm not sure what you mean by "legacy" or "real-world" there.  CSS has no interaction with what's legal or illegal in HTML.

[16:09:01.0000] <Hixie>
that's not how i would put it

[16:09:02.0000] <Hixie>
i'd just say display:inline/block conrtol the rendering

[16:09:03.0000] <Hixie>
and that's all :-)

[16:09:04.0000] <divya>
ok that makes sense :)

[16:10:00.0000] <divya>
(I think it is hard for me to not get confused between css inline/block and html4 inline/block)

[16:10:01.0000] <tabatkins>
Easiest is to probably just ignore the HTML4 definitions entirely.  ^_^

[16:11:00.0000] <divya>
tabatkins: :)

[16:11:01.0000] <AryehGregor>
I bet that's why HTML5 doesn't use those terms to begin with.

[16:12:00.0000] <divya>
I wonder why not many people brought this up. :/

[16:12:01.0000] <divya>
(when they were writing their "Top Ten CSS Tricks" posts)

[16:13:00.0000] <AryehGregor>
Why would they bring it up?

[16:14:00.0000] <divya>
AryehGregor: It is confusing and deserves clarity. More than cool tricks that can be accomplished with display: inline/block 

[16:14:01.0000] <estellevw>
i found the distinction of block, inline and flow very helpful to understanding the semantics of elements

[16:14:02.0000] <Hixie>
/me has no idea what to do about character references that don't end with ;

[16:14:03.0000] <AryehGregor>
Hixie, are they used too much to just be dropped?

[16:15:00.0000] <Hixie>
yes

[16:15:01.0000] <Hixie>
there are bugs that ask for various changes to the way they're parsed

[16:15:02.0000] <Hixie>
9351, 9207

[16:15:03.0000] <Hixie>
or reported as errors

[16:15:04.0000] <Hixie>
9352

[16:16:00.0000] <Hixie>
we could make unescaped & not be an error but leave charrefs that don't have a ; be an error, i guess

[16:16:01.0000] <Hixie>
so "?a=b&c=d" would be fine

[16:16:02.0000] <GarethAdams|Home>
how do you tell the difference?

[16:17:00.0000] <Hixie>
but "?apple=banana&copy=ditto" would be an error (since the &copy gets parsed in an unexpected way)

[16:17:01.0000] <GarethAdams|Home>
oh, so from the fixed list of known entities

[16:17:02.0000] <Hixie>
or we could make &copy= not be parsed as a char ref in attributes

[16:17:03.0000] <Hixie>
just like &copypasta already isn't

[16:18:00.0000] <Hixie>
but that makes attributes even more different than content

[16:18:01.0000] <tabatkins>
Are we sure that char references without a trailing ; appear in attributes, especially @href and similar, often enough to cause a problem?

[16:18:02.0000] <Hixie>
tabatkins: cause problems in which direction?

[16:18:03.0000] <tabatkins>
In the direction of, if we make them not parse as entities, links will break.

[16:19:00.0000] <Hixie>
tabatkins: i expect there are plenty of pages that say "?foo=bar&copy=ditto" and are broken, and i expect there are plenty of pages that say "this file &copy 2010 by me" and expect it to be an entity.

[16:19:01.0000] <tabatkins>
Sure.

[16:19:02.0000] <Hixie>
the latter is even valid in ghtml4

[16:19:03.0000] <Hixie>
s/g//

[16:20:00.0000] <Hixie>
so yes, things will break if we change anything, and things are broken if we don't change anything

[16:20:01.0000] <tabatkins>
I think I'm fine with attributes being different from content here.

[16:20:02.0000] <Hixie>
i'm only talking about attributes here

[16:21:00.0000] <Hixie>
they're already different than content, so that ship has sailed -- teh question is whether we can keep that to a minimum, so that people moving stuff in and out of attributes don't get screwed

[16:21:01.0000] <Hixie>
i'd really like to make any valid content in one have the same meaning in the other

[16:21:02.0000] <Hixie>
invalid content is less of an issue

[16:22:00.0000] <tabatkins>
I'd like links to be easy to author, and not have to worry about the names of my query paramaters.

[16:22:01.0000] <AryehGregor>
It's too bad we can't just make &copy equivalent to &amp;copy in content as well.

[16:22:02.0000] <AryehGregor>
But the extra difference between content and attributes seems like the lesser evil here.

[16:22:03.0000] <AryehGregor>
It's not like it should come up very often in practice.

[16:23:00.0000] <AryehGregor>
You could make "&copy" and such raise a warning or error if used in content.

[16:23:01.0000] <tabatkins>
It would be interesting to try and see how much outcry there is (try making entities require a ;, and allow unescaped & otherwise).

[16:23:02.0000] <AryehGregor>
Yeah, I wonder if implementers would do that.  How many pages are we talking about?  It wouldn't be huge breakage in any case, if it's largely in copyright boilerplate then most users probably won't even notice.

[16:24:00.0000] <AryehGregor>
(even if they view an affected page)

[16:24:01.0000] <AryehGregor>
/me wonders why tabatkins has decapitalized himself

[16:24:02.0000] <tabatkins>
Indeed.

[16:24:03.0000] <Hixie>
we're talking billions of pages here

[16:24:04.0000] <tabatkins>
Eh, accidental.  irssi is taking my account name from linux, and I'm not bothering to correct it.

[16:24:05.0000] <Hixie>
AryehGregor: it would be a disaster in non-english locales

[16:25:00.0000] <AryehGregor>
Oh well.

[16:25:01.0000] <Hixie>
e.g. &egrave is one of these entities

[16:27:00.0000] <Hixie>
suggests the simplest option to improve compatibility on the web is to make = make these entities not get replaced

[16:27:01.0000] <Hixie>
er, http://www.w3.org/Bugs/Public/show_bug.cgi?id=9207#c5 suggests

[16:27:02.0000] <tabatkins>
= and _, apparently.

[16:27:03.0000] <AryehGregor>
That sounds like a clever solution.

[16:27:04.0000] <Hixie>
_ was not quite as clear

[16:29:00.0000] <Hixie>
ok let's add = and then reexamine the conformance problem.

[16:31:00.0000] <Hixie>
i suppose i could make &foo= be not a parse error in attributes, but otherwise leave these rules alone

[16:31:01.0000] <Hixie>
or i could just make all &foo that don't match an entity be not parse errors

[16:32:00.0000] <Hixie>
the latter seems simpler, though it would lead to authors never bothering to escape their &s, which could get them in trouble if they accidentally walk into &amp;s

[16:38:00.0000] <Hixie>
this would make &foo; (where "foo" isn't anything known) into _not_ an error...

[16:42:00.0000] <dglazkov>
Hixie: I demand a recount

[16:42:01.0000] <dglazkov>
:)

[16:43:00.0000] <dglazkov>
can you point me to the place where the spec says that if the size of select is >1, multiple is not set, and no options have selectedness set to true, we shouldn't mark any item as selected?

[16:45:00.0000] <Hixie>
no, because the spec doesn't say that :-)

[16:45:01.0000] <Hixie>
"If the multiple attribute is absent, whenever there are no option elements in the select element's list of options that have their selectedness set to true, the user agent must set the selectedness of the first option element in the list of options in tree order that is not disabled, if any, to true."

[16:45:02.0000] <dglazkov>
but all browser that I know of do it.

[16:45:03.0000] <Hixie>
do what?

[16:45:04.0000] <dglazkov>
not select if size > 1

[16:46:00.0000] <dglazkov>
annevk told me to do the comment thingy

[16:46:01.0000] <dglazkov>
should I go ahead and do it?

[16:47:00.0000] <Hixie>
how weird

[16:48:00.0000] <Hixie>
why would size="" have any effect on this?

[16:48:01.0000] <Hixie>
please do file a bug

[16:48:02.0000] <Hixie>
it's an easy fix

[16:48:03.0000] <Hixie>
i wonder why browsers act like this

[16:50:00.0000] <othermaciej>
Hixie, dglazkov: I suspect the browser rule is "if it renders as a pop-up menu select the first item, if it renders as a listbox, don't select anything"

[16:50:01.0000] <othermaciej>
because listboxes have a reasonable "nothing selected" state

[16:50:02.0000] <dglazkov>
othermaciej: yes, something like that

[16:50:03.0000] <othermaciej>
that would be why size= has an effect

[16:50:04.0000] <dglazkov>
but if you change the size from X to 1 and then back to X (X>1), you'll have a selected element.

[16:51:00.0000] <othermaciej>
sure, because size changes don't cause the selected element to be deselected

[16:51:01.0000] <dglazkov>
but they do cause the element to be selected :)

[16:51:02.0000] <othermaciej>
and browsers don't keep track of whether something got selected by default

[16:51:03.0000] <dglazkov>
yep

[16:51:04.0000] <dglazkov>
/me is writing a layout test for that

[16:51:05.0000] <othermaciej>
yes, if you enter the "must have a selection" state (no multiple or size > 1), then an item gets default-selected

[16:52:00.0000] <othermaciej>
but when you leave that state, nothing gets deselected

[16:52:01.0000] <othermaciej>
seems logical to me, at least to the extent one can expect logic of the Web platform

[16:54:00.0000] <Hixie>
i guess it makes sense given visual UAs using different widgets and how single-select list boxes can have an (normally unreachable) no-selection state

[16:54:01.0000] <Hixie>
doesn't make much sense from an abstract point of view though


2010-04-03
[17:00:00.0000] <othermaciej>
Hixie: it's not unreachable (at least on Mac)

[17:00:01.0000] <othermaciej>
you can always cmd-click the selected item to deselect it

[17:03:00.0000] <Hixie>
othermaciej: wow, so you can

[17:03:01.0000] <Hixie>
i had no idea

[17:07:00.0000] <othermaciej>
/me thought it was obvious from the way cmd-click normally works in lists...

[17:16:00.0000] <TabAtkins>
Does it break some notion of semantics somewhere for a document to not have a location?

[17:17:00.0000] <othermaciej>
that question is too abstract for me to understand it

[17:17:01.0000] <TabAtkins>
The context is an intended-to-be-ephemeral document created by a page, with a location that is only valid until the creating page is trashed.  I think you can navigate to that url fine (by the time the document is trashed and the url becomes meaningless, the navigation action is already moving).  But what does document.location then return in the new doc?

[17:19:00.0000] <Hixie>
TabAtkins: is this a doc created by document.open(), createDocument(), XHR, window.open(), navigating to a File API URL, a Stream URL, or something else?

[17:19:01.0000] <TabAtkins>
Created by a FileWriter.

[17:20:00.0000] <TabAtkins>
Note: This isn't defined.  We're asking what it *should* return, and if it's valid for it to have no location.

[17:20:01.0000] <Hixie>
wait, FileWriter returns a Document?

[17:20:02.0000] <Hixie>
you don't want to do that

[17:20:03.0000] <Hixie>
just have FileWriter use the same URL generation mechanism as the File API

[17:21:00.0000] <Hixie>
looks like the recent bugzilla update broke my automatic scripts, i wonder if the whatwg bug submission thing still works even

[17:21:01.0000] <Hixie>
oh christ it even broke my bookmarklets

[17:22:00.0000] <TabAtkins>
Hixie: That's already being used, but the url generated thusly is intended to be trashed/meaningless when the document that created it went away.

[17:22:01.0000] <Hixie>
bbiab, tgif

[17:22:02.0000] <TabAtkins>
So if you navigate the creating document to the url of the created document (assuming this makes sense), what's the location of the created document?

[17:25:00.0000] <othermaciej>
why would FileWriter return a Document if it's a writer, not a reader?

[17:54:00.0000] <JonathanNeal>
Is sitemap a valid role in html5?

[17:59:00.0000] <othermaciej>
JonathanNeal: is it a valid role in ARIA? if so, it's valid in HTML5

[18:01:00.0000] <JonathanNeal>
Guess not http://www.w3.org/TR/2007/WD-aria-role-20071019/#roles

[18:02:00.0000] <JonathanNeal>
Or maybe this is more accurate / recent  http://www.w3.org/TR/wai-aria/roles#role_definitions

[18:04:00.0000] <othermaciej>
JonathanNeal: I believe your second link is the most accurate

[18:04:01.0000] <othermaciej>
JonathanNeal: is the "navigation" role (or the HTML5 <nav> element) appropriate for what you are trying to do?

[18:05:00.0000] <JonathanNeal>
It is, but I get confused every now and again when I see an HTML5 blogpost use "breadcrumbs" or "sitemap"

[18:05:01.0000] <othermaciej>
as a role?

[18:06:00.0000] <JonathanNeal>
yes, usually next to "navigation" on the ul.

[18:07:00.0000] <JonathanNeal>
Which is weird anyway, because nav has "navigation" inherently. 

[18:11:00.0000] <webben>
JonathanNeal: Don't think sitemap has ever been an ARIA role.

[18:11:01.0000] <JonathanNeal>
poor confused web designers out there.

[18:12:00.0000] <webben>
JonathanNeal: "breadcrumbs" used to be in the draft: http://www.w3.org/TR/2006/WD-aria-role-20060926/

[18:14:00.0000] <JonathanNeal>
Yea ... I often try to convince folks that breadcrumbs shouldn't be a list.

[18:14:01.0000] <JonathanNeal>
Rather, a paragraph.

[18:15:00.0000] <JonathanNeal>
I think the spec has it that way.

[18:15:01.0000] <JonathanNeal>
And I saw it, thought it was odd, then thought about it, and thought it was neat.

[18:15:02.0000] <TabAtkins>
JonathanNeal: What's your reasoning?  I do breadcrumbs as an <ol>.

[18:35:00.0000] <othermaciej>
JonathanNeal: seems like it is semantically a list

[20:08:00.0000] <JonathanNeal>
hi all

[20:10:00.0000] <JonathanNeal>
We had a big company meeting and it was really engaging, sorry for neglecting to reply to you guys, TabAtkins and othermaciej.

[20:10:01.0000] <othermaciej>
JonathanNeal: no worries, I wasn't saying anything important

[20:11:00.0000] <JonathanNeal>
I think breadcrumbs can be either.  I've pushed for them to be paragraphs because I think they can work as paragraphs too.

[20:13:00.0000] <JonathanNeal>
I can see how they would be ol, because they're part of a hierarchical structure.

[20:13:01.0000] <JonathanNeal>
That's actually the reason I had us change from <ul>'s to <p>'s.

[20:14:00.0000] <JonathanNeal>
My initial understanding was tha tI didn't think of <ol> and <ul> as accurately portraying a breadcrumb in a heirarchical manner.

[20:14:01.0000] <othermaciej>
well

[20:14:02.0000] <othermaciej>
you can think of it two ways:

[20:14:03.0000] <othermaciej>
1) a nesting hierarchy

[20:15:00.0000] <othermaciej>
2) an ordered list of the user's path through the site

[20:15:01.0000] <othermaciej>
in a way they are logically equivalent, because a list is a degenerate tree

[20:16:00.0000] <JonathanNeal>
I might hear you describing two types of breadcrumbs, the hansel and gretel style where it targets your journey, or a branch style where you see are in relationship to the site hierchy (which is how I see it)

[20:17:00.0000] <JonathanNeal>
I like the later, as I've stated, and I felt it worked excellent with the idea and example @ http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#hierarchical-link-types

[20:44:00.0000] <TabAtkins_>
JonathanNeal: Yeah, no problem.  Actually I use the latter style of breadcrumbs, where it shows the current path from the site root, but still believe <ol> is an appropriate tag for that.

[20:45:00.0000] <Hixie>
i think things are back in working order after the bugzilla upgrade

[20:45:01.0000] <Hixie>
but we'll see tomorrow

[20:46:00.0000] <shepazutoo>
Hixie: I think you pinged me last night when I was away?

[20:47:00.0000] <Hixie>
yeah, was going to ask you if you were aware of the plans for killing DOM Views, and if so, if you had plans for dealing with it in UIEvent

[20:47:01.0000] <Hixie>
i ended up just going with changing viewArg to dummyArg in the drag and drop event stuff, but i'll change it to whatever you end up deciding on

[20:48:00.0000] <Hixie>
gotta go right now though

[20:48:01.0000] <Hixie>
should be online tomorrow

[20:48:02.0000] <shepazutoo>
Hixie: ok, let's talk tomorrow

[20:48:03.0000] <Hixie>
k

[21:58:00.0000] <JonathanNeal>
whos up!

[22:28:00.0000] <zcorpan>
JonathanNeal: i accept patches (re http://krijnhoetmer.nl/irc-logs/whatwg/20100402#l-471 )

[22:28:01.0000] <JonathanNeal>
reading

[22:28:02.0000] <JonathanNeal>
zcorpan, you're simon's html5 elements page?

[22:29:00.0000] <zcorpan>
yes

[22:29:01.0000] <JonathanNeal>
Then I will get to work immediately.

[22:29:02.0000] <zcorpan>
cool :)

[22:31:00.0000] <JonathanNeal>
Is there an even more multipage version of the html spec?

[22:31:01.0000] <JonathanNeal>
I want to make sure I'm not reinventing the wheel unnecessarily.

[22:31:02.0000] <zcorpan>
there's an index in the multipage spec

[22:32:00.0000] <zcorpan>
http://www.whatwg.org/specs/web-apps/current-work/multipage/section-index.html#index

[22:33:00.0000] <zcorpan>
i guess the table of contents also has all the elements listed

[22:34:00.0000] <zcorpan>
the toc is in the right order

[22:34:01.0000] <zcorpan>
i guess html5-elements' order is out of date

[22:34:02.0000] <JonathanNeal>
I just noticed that the html5 spec is written in the us language of hixie

[22:35:00.0000] <zcorpan>
it would be nice if i could generate html5-elements with a script that scrapes the spec

[22:36:00.0000] <JonathanNeal>
Well, it should be pretty easy with ajax.

[22:36:01.0000] <JonathanNeal>
It just means we'll need to host a local coy and fetch it via ajax.

[22:36:02.0000] <JonathanNeal>
Or curl it to proxy ajax.

[22:37:00.0000] <zcorpan>
generating it with javascript fails on the same-origin policy

[22:38:00.0000] <zcorpan>
could write a bookmarklet

[22:40:00.0000] <zcorpan>
actually a bookmarklet would not be a bad idea. would be easy to just appendChild()ing the right elements and getting innerHTML when you're done

[22:41:00.0000] <zcorpan>
the elements from the toc, and the attributes of each element from the index (loaded in an iframe)

[23:01:00.0000] <JonathanNeal>
zcorpan, i don't get, how does the bookmarklet work?

[23:01:01.0000] <JonathanNeal>
It runs locally so how can it get the content?

[23:02:00.0000] <zcorpan>
you run it when having the html5 spec loaded

[23:03:00.0000] <JonathanNeal>
oh ...

[23:03:01.0000] <JonathanNeal>
Maybe it would be better if I just copied the parts I wanted from the spec into separate files :P

[02:37:00.0000] <zcorpan>
/me adds wbr and device to html5-elements

[06:36:00.0000] <gsnedders>
/me wonders what a graph of number of songs owned per artist against number of plays per artist would look like

[06:37:00.0000] <gsnedders>
/me expects for him it'd be a line through the origin, more or less

[07:21:00.0000] <annevk>
zcorpan, can't you just prefix the current links with http://www.whatwg.org/html

[07:21:01.0000] <annevk>
zcorpan, the automatic redirects should take care of the rest

[08:29:00.0000] <JonathanNeal>
Goodmorning!

[08:43:00.0000] <annevk>
certainly feels like morning

[08:53:00.0000] <annevk>
shouldn't the spec explain how <wbr> is different from zero width whitespace?

[10:24:00.0000] <divya>
is it right to say that there is no stable release of a browser with HTML5 parser?

[10:25:00.0000] <lazni>
fx 3.6 has one, but disabled

[10:26:00.0000] <lazni>
by default

[10:26:01.0000] <divya>
lazni: true, but was wondering if there were anything else I did not know about. 

[10:57:00.0000] <Dashiva>
Six different types of "conforming", I guess it'll be easy to adhere to at least one of them

[11:02:00.0000] <jgraham>
I think I am losing my mind

[11:02:01.0000] <Philip`>
Present tense?

[11:03:00.0000] <jgraham>
What happened to the bit of the websockects protocol spec where it talks about constructing the random tokens?

[11:03:01.0000] <jgraham>
Philip`: Add the word "further" if that helps

[11:04:00.0000] <Dashiva>
Whoa, Shelley is riding the deletionist wave like a pro...

[11:04:01.0000] <jgraham>
Why does WAHTWG not publish a version of websockts in a sane format (i.e. HTML)

[11:05:00.0000] <jgraham>
and without a multitude of different versions

[11:28:00.0000] <Dashiva>
Eating broccoli sure is fun :D

[12:42:00.0000] <Hixie>
jgraham: complete.html has the websockets protocol in it

[12:44:00.0000] <gsnedders>
/me wants a tiling window manager that actually behaves as he wants one to :P

[12:54:00.0000] <Hixie>
gsnedders: ion?

[13:03:00.0000] <Philip`>
gsnedders: Write your own

[13:03:01.0000] <gsnedders>
Philip`: That's my general feeling

[13:03:02.0000] <gsnedders>
Gah.

[13:03:03.0000] <Philip`>
based on xmonad

[13:04:00.0000] <gsnedders>
That means learning HAskell

[13:04:01.0000] <gsnedders>
*Haskell

[13:04:02.0000] <Philip`>
That makes it a doubly good option

[13:39:00.0000] <jgraham>
Hixie: Yeah I realised that eventually

[13:39:01.0000] <jgraham>
We could do with a multipage complete.html though

[13:40:00.0000] <jgraham>
/me likes multipage when he knows exactly what he is looking for and the section is relatively self contained

[13:44:00.0000] <jgraham>
gsnedders: Define "as you want one to"

[13:45:00.0000] <jgraham>
/me tried xmonad for a while but found that it was very inconvenient when you wanted some unusual layout

[13:45:01.0000] <jgraham>
(so prefers the relatively simple solution of using compiz+the grid plugin)

[13:46:00.0000] <jgraham>
(but would like it if the grid plugin had more tiling-window-manager like features)

[15:39:00.0000] <Hixie>
jgraham: i'd be up for that if someone sets up the multipage splitter in such a way that i can use it from the script

[15:40:00.0000] <gsnedders>
/me is somewhat scared that he can go into an IRC channel he's never been in before, and have said to him from someone he's never heard of, "Hey Geoffrey! Nice to see you here!"

[15:41:00.0000] <Hixie>
i know the feeling

[15:41:01.0000] <gsnedders>
I mean, since when was I that well known on the intartubes?

[15:41:02.0000] <Hixie>
i usually reply "um, hi"

[15:41:03.0000] <gsnedders>
Yeah, my normal response too

[15:42:00.0000] <gsnedders>
Obviously your influence is wearing off on me :P

[15:47:00.0000] <jgraham>
Philip`: Can you do whatever needs to be done with the multipage splitter?

[16:21:00.0000] <gsnedders>
<meta http-equiv="text/html" value="text/html;charset=utf-8"> works if I'm testing this correctly


2010-04-04
[17:51:00.0000] <annevk>
jgraham, I discussed this a week ago or so

[17:51:01.0000] <annevk>
jgraham, Philip` suggested I patch his processor script

[17:51:02.0000] <annevk>
btw, the W3C multipage version has better chosen splits than the WHATWG one

[17:56:00.0000] <othermaciej>
annevk: the page boundaries are different?

[18:02:00.0000] <annevk>
yeah

[18:02:01.0000] <annevk>
see e.g. forms

[18:18:00.0000] <othermaciej>
odd

[18:20:00.0000] <karlcow>
http://blog.threepress.org/2010/04/02/designing-ebooks-for-epub-reading-engines/

[18:29:00.0000] <Kronuz>
hello

[18:31:00.0000] <Kronuz>
hey, I'm trying to make a html satinize function using html5lib satinizer in python, I'm trying this code: http://pastebin.org/133583

[18:32:00.0000] <Kronuz>
but then I tried to input this: sanitizeHtml('<div style="text-align:center;margin-top:20px;"></div>')

[18:33:00.0000] <Kronuz>
and I get: '<div style="margin-top: 20px;"></div>'

[18:33:01.0000] <Kronuz>
where did my text-align go? and why? :(

[18:35:00.0000] <Kronuz>
...also, is there a way I can tell html5lib to add / to the self closing elements, such as printing <img ... /> or <br /> if I input <img ...> or <br> ?

[18:36:00.0000] <Hixie>
why would you want to do the latter?

[18:36:01.0000] <Hixie>
(not sure why you lose text-align, maybe html5lib is doing CSS sanitising also and strips text-align?)

[18:37:00.0000] <Kronuz>
Hixie, yep, it is I suppose, but why is it stripping it?

[18:37:01.0000] <Hixie>
not sure

[18:37:02.0000] <Hixie>
jgraham, annevk: yt?

[18:37:03.0000] <Kronuz>
and about the <... /> it's 'cause I want to have valid xhtml

[18:37:04.0000] <Hixie>
html5lib outputs html, not xhtml

[18:37:05.0000] <Hixie>
i think you have to pick an xml mode output if you want valid xhtml output

[18:38:00.0000] <Kronuz>
I would have guessed text-align was a "default" css property

[18:38:01.0000] <Hixie>
(ending <img> with /> is the least of your problems if you want valid xml output)

[18:38:02.0000] <Kronuz>
yep, well, I need a valid chunk, not the whole thing

[18:39:00.0000] <Kronuz>
I mean I wouldn't want it to output the doctype and stuff

[18:39:01.0000] <Hixie>
why are you using xhtml, anyway?

[18:39:02.0000] <Kronuz>
I'd be inserting it myself

[18:39:03.0000] <Hixie>
there's a lot more to outputting valid xhtml than just the doctype, namespaces, and />

[18:39:04.0000] <Hixie>
there's all kinds of crazy complications

[18:40:00.0000] <Kronuz>
anyway, the main problem right now is to understand what's wrong with my satinizer that doesn't let text-align (and who knows what else) thru ... :P

[18:41:00.0000] <Hixie>
yeah dunno about that

[18:41:01.0000] <Hixie>
/me isn't an html5lib guy

[19:05:00.0000] <Kronuz>
is there any way to make html5lib sanitize the css inside <style> tags?

[19:06:00.0000] <Kronuz>
or is there any way I can sanitize those later on using the html5lib css sanitizer?

[20:10:00.0000] <shepazu>
Hixie: what's this about views?

[21:59:00.0000] <Hixie>
i love how leif cites html4 then claims it says the exact opposite of what it says

[22:03:00.0000] <Hixie>
(i'd argue the case on the mailing list, but what html4 says seems somewhat academic)

[01:22:00.0000] <gsnedders>
/me waits for his hands to thaw so that he can type

[01:22:01.0000] <gsnedders>
It's a useful thing to be able to do

[02:38:00.0000] <Hixie>
hm

[02:38:01.0000] <Hixie>
is there any reason we shouldn't make <form> transparent and allow it in phrasing content elements?

[02:38:02.0000] <Hixie>
oh i guess <form> closes <p>, huh

[02:42:00.0000] <othermaciej>
so after reading the names for the algorithms that are alternatives to adoption agency, I am really curious what they do

[02:42:01.0000] <othermaciej>
and which one is which browser

[02:42:02.0000] <othermaciej>
I am guessing "incest algorithm" is IE

[02:42:03.0000] <Hixie>
yes

[02:43:00.0000] <othermaciej>
I don't know what Heisenberg or secret affair are

[02:43:01.0000] <Hixie>
Heisenberg is mozilla's old algorithm, which depended on packet boundaries

[02:43:02.0000] <othermaciej>
but I assume one is Firefox and the other is Opera

[02:43:03.0000] <Hixie>
and "secret affair" is opera, because it stores additional information beyond what you can see in the dom

[02:44:00.0000] <othermaciej>
I can see how adoption agency turned out to be the lesser evil

[02:44:01.0000] <Hixie>
e.g. <span style="a"> aaaaa <b style="b"> bbbbb </span> bbbb </b> will result in a DOM like <span> aaaaa <b> bbbbb </b></span> bbbb

[02:44:02.0000] <Hixie>
but it'll use the styles from <b> when rendering the last text node

[02:46:00.0000] <othermaciej>
that reminds me vaguely of how WebKit handles blocks inside inlines

[02:46:01.0000] <othermaciej>
(though the evil magic there is solely in the render tree)

[02:46:02.0000] <othermaciej>
but that could be just because I am currently trying to fix a bug with inline continuations

[02:47:00.0000] <othermaciej>
specifically how they interact with the accessibility tree

[02:47:01.0000] <othermaciej>
I may have to give up and ask hyatt for help

[03:30:00.0000] <Hixie>
huh, neither mozilla, nor safari, nor google reader complained about my xml error in my test atom document

[03:30:01.0000] <Hixie>
it took opera to point it out to me

[03:30:02.0000] <Hixie>
good work atom implementors!

[03:32:00.0000] <annevk>
am now Hixie

[03:32:01.0000] <Hixie>
annevk: some guy was asking html5lib questions

[03:32:02.0000] <annevk>
oh, I wouldn't have known either; it's been too long since I played with that

[03:33:00.0000] <othermaciej>
Hixie: I'm pretty sure Atom gets parsed in a tag soupey way

[03:33:01.0000] <othermaciej>
in Safari

[03:33:02.0000] <othermaciej>
our RSS/Atom parsing is kind of bunk

[03:55:00.0000] <Hixie>
man the state of atom implementations is terrible

[03:59:00.0000] <Hixie>
ok bed time

[03:59:01.0000] <Hixie>
nn

[04:08:00.0000] <annevk>
is this issue-96 proposal thread worth a read?

[04:08:01.0000] <annevk>
it seems to start out with shelley not understanding why ARIA is not the final solution...

[04:09:00.0000] <othermaciej>
it was a pretty one-sided thread (Shelley vs. everyone else) but I'd say it's worth reading and perhaps you can add something useful

[05:34:00.0000] <gsnedders>
Hixie: not well-formed in what way?

[11:21:00.0000] <jgraham>
Kronuz left I guess

[11:22:00.0000] <jgraham>
Oh well for future referne you can make html5lib give trailing slashes but if you really want xml output you should use an xml serializer in some xml library

[11:23:00.0000] <jgraham>
And the CSS sanitizer is some horrible regexp that I want to delete

[11:23:01.0000] <jgraham>
(but can't really delete until it can be replaced by a proper CSS parser)

[11:24:00.0000] <jgraham>
But there is a whitelist of properties. I can't see why text-align wouldn't be on it though

[13:33:00.0000] <annevk>
the amount of things people try to infer from vague lines in old specs is amazing

[13:33:01.0000] <annevk>
pro tip: specs have bugs

[13:41:00.0000] <Dashiva>
Isn't that the point of vague specs, letting people interpret them freely? :)

[13:56:00.0000] <AryehGregor>
/me really wishes <input type=email> would catch on so sites actually accept e-mail addresses with + in them

[13:56:01.0000] <AryehGregor>
It's extremely sad how many sites reject such addresses.

[13:56:02.0000] <Dashiva>
No kidding

[13:57:00.0000] <Dashiva>
I've been told it's a "security risk"

[13:59:00.0000] <karlcow>
/me wonders which kind of security risk. Spam is another security risk ;) 

[13:59:01.0000] <Dashiva>
Maybe they don't escape it properly, so the + becomes a space or something

[14:00:00.0000] <Philip`>
Dashiva: I've seen that bug in one site

[14:00:01.0000] <Philip`>
(An Intel one, I think)

[14:00:02.0000] <Philip`>
where it constructed a URL like "register.cgi?user=foo+bar⊙bc"

[14:04:00.0000] <Dashiva>
Especially find is the examples like sun

[14:04:01.0000] <Dashiva>
Where they also reject + in the email address in the form to send feedback about registration

[14:05:00.0000] <Dashiva>
(I'm not sure what I meant instead of 'find')

[14:06:00.0000] <Philip`>
Maybe you meant "fun"?

[14:06:01.0000] <Dashiva>
I guess

[14:07:00.0000] <Philip`>
Or perhaps you meant something like "finickity"

[14:08:00.0000] <Philip`>
which admittedly wouldn't have made sense


2010-04-05
[17:23:00.0000] <MikeSmith>
anybody seen nessy around today?

[17:42:00.0000] <MikeSmith>
Hixie: about the broken links in IDLs in the the author view of the spec, for the static W3C copy of the author-view that I'm generating, I guess I can have the script for that rewrite the fragment IDs in the IDLs so that they point to the domintro stuff instead, but not clear to me if/how it's fixable in the dynamic/JS-switchable view of the spec

[17:44:00.0000] <MikeSmith>
and since not all attributes in the IDLs have corresponding domintro content, I guess on I can just have the build of the static author view un-link them and instead just display them as normal text

[17:45:00.0000] <MikeSmith>
It seems like that's mainly the case for DOM attributes that have same names as content attributes

[17:45:01.0000] <MikeSmith>
or whatever they're currently called in the spec to distinguish them from one another

[17:46:00.0000] <MikeSmith>
hsivonen: bugzilla.validator.nu seems to be responding very slowly

[17:54:00.0000] <KaOSoFt>
Hello.

[17:56:00.0000] <MikeSmith>
KaOSoFt: good morning

[18:03:00.0000] <KaOSoFt>
I've just recently been practicing  HTML5, and while thinking in semantics, I thought about formatting (giving it meaning) my data. I read some web-logs which mentioned Microformats, and while it looks like an interesting approach, some others barely wrote about RDFa.

[18:03:01.0000] <KaOSoFt>
I then checked some documents about RDFa, but I don't seem to find any examples using HTML5. I suppose it's because both need to get the recommendation status, but I wanted to check nevertheless. Given there is a DTD for RDFa with XHTML but since HTML5 doesn't use it, I wanted to know how should I start approaching RDFa in HTML5: is it okay to just start using RDFa properties, or do you rather know of/have an example?

[18:04:00.0000] <KaOSoFt>
I still need to know exactly how HTML5 is going to work with DTD, or without them.

[18:04:01.0000] <MikeSmith>
KaOSoFt: you know of the http://dev.w3.org/html5/rdfa/ ?

[18:04:02.0000] <MikeSmith>
KaOSoFt: HTML5 does not rely on DTDs

[18:05:00.0000] <MikeSmith>
nobody should be using DTDs at this point, for anything

[18:05:01.0000] <MikeSmith>
they have been obsolete for a long time

[18:05:02.0000] <KaOSoFt>
1. Didn't know about that, thanks. :)

[18:05:03.0000] <KaOSoFt>
For some reason I never got into that link. :s

[18:05:04.0000] <Hixie>
MikeSmith: yeah i don't know what to do in the dynamic case

[18:05:05.0000] <MikeSmith>
if you need a schema to use for validation, the right thing to use these days is either RelaxNG or W3C XML Schema

[18:05:06.0000] <KaOSoFt>
2. I see, I wanted to be sure anyways.

[18:06:00.0000] <Hixie>
MikeSmith: in the static case it might make more sense to start from the pre-anolis draft, change the domintro elements to dfn, and rerun anolis

[18:06:01.0000] <MikeSmith>
Hixie: hmm, yeah

[18:06:02.0000] <MikeSmith>
Hixie: for the dynamic case, maybe l e

[18:06:03.0000] <MikeSmith>
oops

[18:08:00.0000] <MikeSmith>
Hixie: for the dynamic case, maybe one of your scripts could cause an alert to pop up, saying "The target of this link is not in the author view, do you want to switch to the full (implementor) view instead?"

[18:08:01.0000] <MikeSmith>
or something like that

[18:08:02.0000] <MikeSmith>
I don't know how feasible that would be, though

[18:09:00.0000] <Hixie>
ooh that's a good idea

[18:11:00.0000] <KaOSoFt>
MikeSmith- Fortunately, that document just mentions the changes to make the XHTML+RDFa recommendation work with HTML5, which implies it is still pretty much XHTML+RDFa.

[18:11:01.0000] <KaOSoFt>
Thank you.

[18:11:02.0000] <KaOSoFt>
^^_

[18:11:03.0000] <KaOSoFt>
^_^

[18:11:04.0000] <Hixie>
MikeSmith: can you file a bug to that effect?

[18:12:00.0000] <MikeSmith>
Hixie: yeah, will do

[18:12:01.0000] <Hixie>
thanks

[18:13:00.0000] <KaOSoFt>
Hixie- This example is yous, isn't it? http://www.w3.org/TR/xhtml-rdfa-primer/alice-example.html

[18:13:01.0000] <Hixie>
lol, laura just implied that if the WAI isn't consulted on something, it can't be good for accessibility

[18:13:02.0000] <MikeSmith>
KaOSoFt: there are some unresolved issues with full integration of RDFa into text/html

[18:13:03.0000] <Hixie>
KaOSoFt: never seen it before

[18:14:00.0000] <KaOSoFt>
MikeSmith- Yeah, I thought so. I know they are still not Recommendations, but wanted to be sure how far they were. Thanks.

[18:16:00.0000] <MikeSmith>
KaOSoFt: have you looked at the HTML5 Microdata feature?

[18:17:00.0000] <KaOSoFt>
Hixie- Oh, sorry, I got my tabs mixed, and thought it was yours. :P

[18:19:00.0000] <KaOSoFt>
MikeSmith- Well, I was reading what some search engine crawlers understood, and that's why I knew of Microformats and RDFa.

[18:19:01.0000] <KaOSoFt>
Now that you mentioned HTML5 Microdata, I'm currently reading this: http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html

[18:22:00.0000] <KaOSoFt>
The current HTML5 specification is quite large. I have to check the multipage version, otherwise the browsers take a while to load it all.

[18:24:00.0000] <KaOSoFt>
Hmmm, looks interesting.

[18:29:00.0000] <MikeSmith>
Google's rich-snippets feature also supports HTML5 Microdata now

[18:29:01.0000] <MikeSmith>
if your goal is to have search-engine support

[18:30:00.0000] <KaOSoFt>
Yes, exactly. :)

[18:33:00.0000] <KaOSoFt>
By the way, is this the correct channel for these inquiries? I don't mean to disturb whatever you do here.

[18:33:01.0000] <KaOSoFt>
._.

[18:37:00.0000] <MikeSmith>
KaOSoFt: here is fine

[18:38:00.0000] <MikeSmith>
especially if you have questions about the Microdata feature

[18:38:01.0000] <MikeSmith>
foolip and others have used it some and can help a bit probably

[18:39:00.0000] <KaOSoFt>
Yeah, I'm currently searching using "HTML5 Microdata", and have found some articles. :D

[18:39:01.0000] <KaOSoFt>
The first step was knowing about it. :P

[18:59:00.0000] <KaOSoFt>
Wow, there are already a lot of tools for testing/trying/enjoying Microdata.

[18:59:01.0000] <KaOSoFt>
I'm currently checking foolip's Live Microdata. :D

[19:00:00.0000] <kennyluck>
KaOSoFt, any pointer?

[19:00:01.0000] <KaOSoFt>
Pointer?

[19:00:02.0000] <KaOSoFt>
http://gitorious.org/microdatajs

[19:00:03.0000] <kennyluck>
yeah, like this, thatnks.

[19:00:04.0000] <KaOSoFt>
http://james.html5.org/microdata/

[19:00:05.0000] <KaOSoFt>
http://philip.html5.org/demos/microdata/demo.html

[19:00:06.0000] <KaOSoFt>
:D

[19:01:00.0000] <kennyluck>
Thanks! :)

[19:10:00.0000] <MikeSmith>
kennyluck: hey, you at t45 already?

[19:10:01.0000] <kennyluck>
MikeSmith, yes I am! Why?

[19:11:00.0000] <MikeSmith>
just wondering

[19:11:01.0000] <MikeSmith>
some trains seem to be delayed

[19:11:02.0000] <kennyluck>
The rain wasn't that big when I came here.

[19:16:00.0000] <MikeSmith>
kennyluck: if you have time in working on some HTML5-related stuff, there's a number of things that might be interesting

[19:16:01.0000] <MikeSmith>
work on the validator.nu backend, for one

[19:16:02.0000] <MikeSmith>
also, the RDFa WG is working on and RDFa API

[19:17:00.0000] <kennyluck>
MikeSmith, cool.

[19:17:01.0000] <kennyluck>
Yep, I would like to join RDFa WG, I just got my W3C member account.

[19:17:02.0000] <MikeSmith>
great

[19:19:00.0000] <kennyluck>
MikeSmith: I am also now discussing with Fumi about holding an event for Semantic Web & Linked Data. I wonder whether I can invite you to talk about Microdata.

[19:20:00.0000] <kennyluck>
Actually not an event, but gatherings twice a month.

[19:27:00.0000] <MikeSmith>
if you do the event you talked about, make sure to invite Kanzaki-san

[19:30:00.0000] <kennyluck>
MikeSmith, yup, that's on the plan. It was very sad that I didn't meet him last year. I even bought 5 of his books, and I gave Tim one of them (should have got his signature)

[19:33:00.0000] <MikeSmith>
anyway, I may not be the best person to present about Microdata

[19:34:00.0000] <MikeSmith>
I'm pretty sure we could find someone who's already using it in pages

[19:34:01.0000] <kennyluck>
OK, but you are always welcome :) I just forwarded my plan to you.

[19:34:02.0000] <MikeSmith>
OK

[19:34:03.0000] <MikeSmith>
adding RDFa support to validator.nu would be another thing worth exploring

[19:34:04.0000] <MikeSmith>
i'm not sure that's completely possible right now without some changes being made to the HTML5 parsing algorithm

[19:35:00.0000] <MikeSmith>
but hsivonen or Hixie would know better

[19:35:01.0000] <MikeSmith>
hsivonen did an implementation a while back that didn't support CURIEs but the RDFa community did not seem to be too thrilled with that

[19:35:02.0000] <kennyluck>
We should be happy...don't know why.

[19:36:00.0000] <MikeSmith>
heh

[19:36:01.0000] <MikeSmith>
http://www.html5.jp/blog/2010/04/04/html5-api/ is interesting

[19:36:02.0000] <kennyluck>
But I should say, I have little knowledge about the difference of HTML+RDFa and XHTML+RDFa.

[19:36:03.0000] <kennyluck>
Does HTML+RDFa means RDFa 1.1 ?

[19:38:00.0000] <kennyluck>
By the way, CURIE like "[wiki:Test]" is a mess, we RDF community people can't live without "foaf:knows". Maybe that's what happened. 

[19:38:01.0000] <kennyluck>
s/, we/, but we/

[19:38:02.0000] <kennyluck>
Ughhhh..

[19:39:00.0000] <MikeSmith>
hsivonen could give you the details about what the validator implementations issues are with respect to supporting RDFa

[19:39:01.0000] <MikeSmith>
HTML+RDFa means supporting whatever version of RDFa in text/html MIME type

[19:40:00.0000] <MikeSmith>
as opposed to XML MIME type

[19:40:01.0000] <kennyluck>
Yes, so HTML+RDFa doesn't have to meet the "copy and paste" requirement. Am I right?

[19:40:02.0000] <MikeSmith>
yeah, arguably

[19:41:00.0000] <MikeSmith>
to put it in other terms, RDFa+HTML means supporting it RDFa documents that are parsed with HTML parsers instead of XML parsers

[19:42:00.0000] <MikeSmith>
or, I guess, in any tools that don't rely on the pseudo-namespace mechanism that RDFa uses

[19:43:00.0000] <kennyluck>
OK, I need to look it up. I am not familiar with the difference between HTML DOM and XML DOM..

[19:55:00.0000] <KaOSoFt>
Well, I've read enough for the day. Tomorrow I'll practice a little bit more, and try to implement Microdata in a project I'm working on. Thank you all.

[19:55:01.0000] <KaOSoFt>
Goodbye.

[19:55:02.0000] <kennyluck>
KaOSoFt, bye.

[20:20:00.0000] <othermaciej>
good evening everyone

[21:43:00.0000] <MikeSmith>
Hixie: I'm only now noticing that hostname on the From address of mail generated by W3C bugzilla has changed from wiggum.w3.org to jessica.w3.org

[21:43:01.0000] <MikeSmith>
I don't know if that makes any difference for any mail filtering you have set up

[21:43:02.0000] <MikeSmith>
it does for me

[21:44:00.0000] <MikeSmith>
I think it's also breaking some of the hacky commit-notification stuff I have set up

[21:44:01.0000] <MikeSmith>
I'll have to check and see

[21:45:00.0000] <Hixie>
that was one of the few changes that didn't actually break anything for me :-)

[21:45:01.0000] <MikeSmith>
OK

[21:46:00.0000] <MikeSmith>
the may be some other stuff that's borked due to the bugzilla upgrade

[21:46:01.0000] <MikeSmith>
anyway, I'm going to be away for the next 12+ hours or so

[21:46:02.0000] <MikeSmith>
traveling

[21:47:00.0000] <MikeSmith>
so if in the mean time you discover anything busted, please ping shepazu about it

[21:48:00.0000] <Hixie>
k

[21:50:00.0000] <MikeSmith>
Hixie: btw, as far as the general rationale for having document-conformance criteria in the spec, I think another reason that could possibly be cited is that non-browser tools that consume HTML also benefit from the existence of constraints that bound the possible markup cases they are designed to process

[21:50:01.0000] <Hixie>
are there any conformance criteria that we've added for that reason?

[21:51:00.0000] <MikeSmith>
I don't know if there are some that have been added for that specific reason

[21:51:01.0000] <MikeSmith>
but it is at least a beneficial side effect

[21:51:02.0000] <MikeSmith>
tools can't be expected to process every insane markup case that some author wants to use for whatever bizarre reasons

[21:52:00.0000] <MikeSmith>
e.g., whatever recent bug was arguing that div should be allowed as a child of span

[21:53:00.0000] <Hixie>
depends on the tool

[21:53:01.0000] <MikeSmith>
true

[21:53:02.0000] <Hixie>
not really sure what kind of tool would have problem with that kind of thing

[21:53:03.0000] <MikeSmith>
I guess cases like that actually break the element semantics 

[21:55:00.0000] <Hixie>
anyone know of a character that in a particular character encoding other than UTF-16 or -32 becomes a multibyte character with one of the bytes being 0x20?

[22:09:00.0000] <wirepair>
http://www.fileformat.info/info/unicode/char/2022/index.htm

[22:09:01.0000] <wirepair>
%u2022

[22:09:02.0000] <wirepair>
er sorry that is utf-165

[22:09:03.0000] <wirepair>
er 16 ;>

[22:12:00.0000] <wirepair>
looks like shift_jis and euc_jp don't have any

[22:29:00.0000] <Hixie>
well if there aren't any then that makes my life easier. :-)

[00:32:00.0000] <JonathanNeal>
ahoy!

[02:20:00.0000] <Hixie>
othermaciej: 94?

[02:20:01.0000] <othermaciej>
Hixie: pardon?

[02:20:02.0000] <Hixie>
how did you count 94 open bugs?

[02:20:03.0000] <Hixie>
i count 43

[02:20:04.0000] <othermaciej>
Hixie: did I say 94?

[02:20:05.0000] <othermaciej>
I meant 43

[02:21:00.0000] <Hixie>
that's quite the typo :-P

[02:21:01.0000] <othermaciej>
Hixie: d''oh, I forgot to edit it from last month's value

[02:21:02.0000] <Hixie>
you had me momentarily worried that my scripts were missing some huge number of bugs

[02:22:00.0000] <othermaciej>
Hixie: no, I distinctly remember doing a query and getting 43

[02:23:00.0000] <Hixie>
good, good

[02:23:01.0000] <othermaciej>
and indeed the ones I called out as oldest reopened and oldest non-reopened will be in that set

[02:23:02.0000] <othermaciej>
sorry for scaring you

[02:23:03.0000] <Hixie>
so long as i'm not blithely ignoring half the bugs out there, i'm happy :-)

[03:50:00.0000] <annevk>
/me wonders if it would be better to have File/LocalFile/Blob rather than File/Blob with Blob overloaded with type/URL

[03:56:00.0000] <annevk>
defining maps for properties is still hard considering forward extensibility

[03:57:00.0000] <annevk>
at least if we do things like .property.partOfMap

[03:57:01.0000] <annevk>
since partOfMap might conflict with a future value API

[03:57:02.0000] <annevk>
maybe maps and list should be on their own attributes

[03:57:03.0000] <annevk>
e.g. margin.m.left

[03:57:04.0000] <annevk>
or b.l[0].m.color

[03:58:00.0000] <annevk>
euh, I meant background.l[0].m.color

[03:59:00.0000] <annevk>
meh, complicated

[03:59:01.0000] <annevk>
also fun

[04:41:00.0000] <annevk>
more thinking time tomorrow

[09:20:00.0000] <ment>
is there a sheet somewhere listing supported <audio>, <video> codecs in various browsers?

[09:25:00.0000] <paul_irish>
ment: theres a page on html5doctor that tries to capture audio codecs.. video support.. that one should be easier to find. btw you can use Modernizr to detect these on the fly.

[09:34:00.0000] <ment>
paul_irish: modernizr it is! thanks

[09:56:00.0000] <annevk>
"I dont think there is so much difference between say :hover and mouseover, or :visited and click (except that click is more general), or state and event in general."

[09:56:01.0000] <annevk>
sounds like it is time to give up on this thread

[10:26:00.0000] <dglazkov>
ah what the heck

[10:26:01.0000] <dglazkov>
good morning, Whatwg!

[10:28:00.0000] <annevk>
morning dglazkov!

[10:28:01.0000] <annevk>
/me wonders if everyone has Easter Monday off

[10:28:02.0000] <dglazkov>
nope.

[10:29:00.0000] <dglazkov>
but some peeps do.

[10:30:00.0000] <annevk>
we do, though I've done some work anyway

[10:32:00.0000] <JonathanNeal>
You're a Unicorn!

[10:34:00.0000] <annevk>
orly?

[10:48:00.0000] <JonathanNeal>
yes, annevk, rly.

[11:09:00.0000] <JonathanNeal>
Is there no -moz-linear-gradient equiv in Opera, still?

[11:15:00.0000] <jwalden>
wasn't there a CSS Opera logo recently that I thought used one?  or did someone just make one for the heck of it?

[11:16:00.0000] <Philip`>
http://desandro.com/articles/opera-logo-css/

[11:16:01.0000] <Philip`>
It doesn't work in Opera

[11:16:02.0000] <paul_irish>
http://people.opera.com/howcome/2010/tests/operalogo.html

[11:17:00.0000] <Dashiva>
Absolut semantic markup

[11:19:00.0000] <annevk>
we don't do gradients yet

[11:19:01.0000] <JonathanNeal>
ah, got it.

[11:20:00.0000] <JonathanNeal>
iz cool, just making sure I was supporting Opera's syntax if it was there.

[11:21:00.0000] <Dashiva>
How would you make that accessible the best way?

[11:23:00.0000] <JonathanNeal>
Dashiva, was that question for me?  If so, I don't quite understand the question.

[11:23:01.0000] <Dashiva>
<div><div></div><div></div><div></div><div></div><div></div><div></div><div></div><div></div></div> doesn't explain very much what's going on

[11:26:00.0000] <paul_irish>
Yah where's the semantic O in the middle of it?

[11:26:01.0000] <paul_irish>
a buncha slackers.

[11:30:00.0000] <JonathanNeal>
They have classnames, Dashiva :)

[11:32:00.0000] <gsnedders>
/me notes we can render the Opera logo using SVG

[11:37:00.0000] <JonathanNeal>
You'll have gradient soon enough, fo sho.

[11:39:00.0000] <TabAtkins>
Moz just happened to be the first to implement the new gradient syntax.  Others will get on the ball soon enough.

[11:44:00.0000] <TabAtkins>
exit

[11:45:00.0000] <JonathanNeal>
entrance?

[11:46:00.0000] <TabAtkins>
Eh, forgot the /.  Wanted to reload my prefs.

[11:50:00.0000] <JonathanNeal>
Oh, you weren't just being dramatic :)

[12:04:00.0000] <TabAtkins>
othermaciej: So should I wait on writing counterproposals?  I was going to block out some time this week for them.

[12:04:01.0000] <othermaciej>
TabAtkins: it's never a problem to start early, or to submit proposals without a formal call

[12:05:00.0000] <TabAtkins>
Sure, but if something might get rejected before having counterproposals asked for, it's a waste of time for me to do anything.  Is that a possibility?

[12:05:01.0000] <TabAtkins>
Basically I am trying to minimize the amount of time I waste on countering ridiculousness.

[12:06:00.0000] <othermaciej>
TabAtkins: I don't know. I have suggested to the other co-chairs that if a proposal has no support besides the author, and significant opposition, we should consider just doing a call for consensus to reject it.

[12:06:01.0000] <othermaciej>
TabAtkins: (a) Not sure what they will think. (b) Not sure if Laura's comments can be counted as "support".

[12:06:02.0000] <TabAtkins>
Kk.

[12:06:03.0000] <othermaciej>
we will probably decide on some course of action for at least some of the proposals by early this week

[12:07:00.0000] <TabAtkins>
K.  I'll stay in the loop, but not plan on anything yet.

[12:12:00.0000] <othermaciej>
Hixie said he thought one counter-proposal might be sufficient to counter all the semantic element/attribute proposals

[12:12:01.0000] <othermaciej>
perhaps not literally identical text, but at least a template

[12:12:02.0000] <othermaciej>
you could maybe consider using a wiki, since others (e.g. John Foliot) have expressed interest in contributing

[12:13:00.0000] <TabAtkins>
Yeah, I'll put it up on the wiki if I start on them.

[12:22:00.0000] <hober>
I had started my ISSUE-95 zero-edit CP a while ago; I'd be willing to generalize it into a broader one for the several related issues

[12:23:00.0000] <hober>
here's the existing text; I have several recent emails flagged in gmail as source material for updating the CP this week: http://hober.jottit.com/issue-95_change_proposal

[13:05:00.0000] <annevk>
http://lists.w3.org/Archives/Public/www-html-editor/2010JanMar/0013.html o_O

[13:05:01.0000] <annevk>
http://lists.w3.org/Archives/Public/www-html-editor/2010JanMar/att-0011/00-part apparently a common thing

[13:05:02.0000] <annevk>
/me remembers someone from the W3C telling him about it

[13:06:00.0000] <Dashiva>
What is happening?

[13:06:01.0000] <annevk>
nothing much, it's a free day

[13:23:00.0000] <annevk>
whoa: http://www.rijksoverheid.nl/copyright

[13:23:01.0000] <annevk>
that's pretty cool

[13:40:00.0000] <annevk>
"PS Notice lack of smiley" lol

[13:41:00.0000] <annevk>
are all her emails like that? maybe I should read some more

[13:42:00.0000] <othermaciej>
annevk: what email is that from?

[13:46:00.0000] <annevk>
http://lists.w3.org/Archives/Public/www-archive/2010Apr/0007.html

[13:53:00.0000] <annevk>
http://twitter.com/shelleypowers/status/11657402980 -- "It begins to look like a month of my work will be rejected out of hand at the HTML5 WG http://bit.ly/cgsEb9 because of five lazy assholes"

[13:53:01.0000] <annevk>
which is followed with

[13:54:00.0000] <annevk>
http://twitter.com/shelleypowers/status/11657500563 -- "I wonder when the new boss at the W3C will finally so something about the atmosphere of hostility and bias at the HTML5 WG"

[13:54:01.0000] <annevk>
classic

[13:56:00.0000] <othermaciej>
I'm definitely sensing some hostility

[13:57:00.0000] <dglazkov>
TabAtkins: are you to blame for everything?

[13:58:00.0000] <annevk>
dglazkov, you guys must regret what you got yourselves into by hiring this TabAtkins figure

[13:58:01.0000] <annevk>
dglazkov, nasty Texans :p

[13:58:02.0000] <JonathanNeal>
I blame Tab when little things go wrong.

[13:58:03.0000] <dglazkov>
what? we hired him?! I thought it was just April Fool's

[13:58:04.0000] <JonathanNeal>
My girlfriend misunderstands me.

[13:58:05.0000] <JonathanNeal>
I hit a red light even though there are no cars coming the other way.

[14:00:00.0000] <JonathanNeal>
I say to myself, "Tabbit."

[14:03:00.0000] <TabAtkins>
I am, basically, to blame for all of your small ills.

[14:03:01.0000] <TabAtkins>
Re Shelley, sigh.

[14:21:00.0000] <JonathanNeal>
So whatwg has no rendering requirement or recommendation for :link { color }, right? 

[14:24:00.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#fonts-and-colors

[14:26:00.0000] <JonathanNeal>
Hixie, you know I always look at that guy ... it's just that sometimes I miss 'em.

[14:26:01.0000] <JonathanNeal>
Like body having a margin, it's not writen out css style.

[14:29:00.0000] <JonathanNeal>
Oh, there it is, blue and purple

[14:31:00.0000] <JonathanNeal>
brain fail

[14:31:01.0000] <dglazkov>
hey, standards brains, is there anything more modern and specific than this on appendChild()? http://www.w3.org/TR/DOM-Level-3-Core/core.html#ID-184E7107

[14:32:00.0000] <dglazkov>
specifically, the actual algorithm of appending document fragments.

[14:32:01.0000] <othermaciej>
dglazkov: DOM Level 3 Core is the most recent official standard

[14:33:00.0000] <othermaciej>
dglazkov: inserting DocumentFragments into the DOM is a particularly hairy area though, due to mutation events

[14:33:01.0000] <dglazkov>
othermaciej: ok. If I wanted to specify more, how would I go about it?

[14:33:02.0000] <dglazkov>
othermaciej: yes indeedy :)

[14:33:03.0000] <othermaciej>
dglazkov: volunteer to edit DOM Level 4 Core?

[14:33:04.0000] <dglazkov>
never!

[14:33:05.0000] <dglazkov>
:)

[14:34:00.0000] <dglazkov>
file a bug mebbe?

[14:34:01.0000] <gsnedders>
/me notes shepauz did

[14:34:02.0000] <gsnedders>
(or whatever his IRC nick is)

[14:39:00.0000] <JonathanNeal>
fieldset is display inline according to you folks, right?

[14:40:00.0000] <JonathanNeal>
I never see where it is recommended or required to be display block.

[14:40:01.0000] <JonathanNeal>
Though I want to infer that from the margin and padding.

[14:40:02.0000] <dglazkov>
othermaciej: FF has the sanest behavior on document fragments + mutation events, so I'll just use theirs

[14:41:00.0000] <jwalden>
dear lazyirc, anyone feel like reporting to Opera that they've mis-implemented calculation of the background positioning area in 10.51?  testcase is <http://whereswalden.com/files/mozilla/background-size/bg-position.html>, cowbell should scale over entire canvas, not over the area of the rounded dashed rectangle+margins

[14:41:01.0000] <othermaciej>
dglazkov: is there anything wrong with the way WebKit does it currently?

[14:41:02.0000] <othermaciej>
dglazkov: is there a compat issue caused by this?

[14:43:00.0000] <dglazkov>
othermaciej: in a word, yes

[14:43:01.0000] <gsnedders>
jwalden: Are you sure we actually support cover at all?

[14:43:02.0000] <jwalden>
gsnedders: yes

[14:44:00.0000] <jwalden>
(make the window tall and narrow to demonstrate the bug, to elaborate)

[14:44:01.0000] <gsnedders>
/me doesn't have 10.5x at hand

[14:47:00.0000] <TabAtkins>
JonathanNeal: fieldset is crazytimes, and not actually done with CSS in any browser

[14:47:01.0000] <JonathanNeal>
Sure, but you're recommending margins and padding, so why not display type?

[14:48:00.0000] <TabAtkins>
The hardest part is positioning <legend> (which, though it is required to be the first child of <fieldset>, for legacy reasons must be accepted as any child) and styling the border (since it breaks behind the <legend>).

[14:48:01.0000] <TabAtkins>
Shrug, dunno.

[14:53:00.0000] <othermaciej>
I think <fieldset> is essentially a block other than the special handling of <legend>

[14:54:00.0000] <TabAtkins>
It is.  If we could find a way to specify the border rendering, and the positioning of <legend> without making all <fieldset>s into positioning roots, we could do it all in CSS.

[14:56:00.0000] <othermaciej>
<fieldset> is kind of legacy cruft - it isn't very useful in modern UI design

[14:56:01.0000] <JonathanNeal>
It is essentially a block, so is it a bad idea just to write it in?

[14:56:02.0000] <othermaciej>
so not worth spending too much effort on it

[14:56:03.0000] <JonathanNeal>
It's not legacy if you're putting it in the new spec.

[14:56:04.0000] <TabAtkins>
(While I don't really like selectors appearing in values, I suspect that's what would be required to specify <legend>'s positioning.  position-to: "fieldset"; would grab the closest ancestor that matches the selector, and treat it as the positioning root (unless there was a closer ancestor who is actually a positioning root from normal means).

[14:57:00.0000] <TabAtkins>
That would be useful in more than just <fieldset>/<legend>.

[14:57:01.0000] <othermaciej>
in WebKit, fieldset in fact has a display: block rule in the UA stylesheet

[14:57:02.0000] <othermaciej>
TabAtkins: to really do fieldset/legend, you need the notion of an interrupted border with something positioned in it, and an item that acts as the border caption for the relevant container

[14:58:00.0000] <othermaciej>
TabAtkins: so I think it can be done without selectors in values

[14:58:01.0000] <KrocCamen>
Aren’t fieldsets needed to get radio buttons to know each other? (press one, and the others depress)

[14:58:02.0000] <othermaciej>
TabAtkins: same basic concept as display: table and display: table-caption

[14:58:03.0000] <KrocCamen>
(Or am I thinking of VB6, perhaps I am)

[14:58:04.0000] <TabAtkins>
border-caption...  That's an interesting way to capture that.

[14:58:05.0000] <TabAtkins>
KrocCamen: No, radios know each other by @name.

[14:58:06.0000] <othermaciej>
TabAtkins: the latter finds its containing table without the table having to act as a containing block and without specifying it explicitly by selector

[15:02:00.0000] <TabAtkins>
Aw man, I love Template.  So good!

[15:22:00.0000] <JonathanNeal>
I also noticed that HR has no recommendation for margin, so no margin is recommended.

[15:23:00.0000] <TabAtkins>
Hm, really?  Pretty sure <hr> *does* have margin.

[15:24:00.0000] <TabAtkins>
It's got ".5em auto" in Chrome, at least.

[15:26:00.0000] <TabAtkins>
<hr> has some odd legacy issues, but none that would actually require any changes to CSS.  Everybody's implementation of it is just an empty block box with an inset border and a small vertical margin.

[15:26:01.0000] <TabAtkins>
It's just that, for some reason, some browsers don't actually expose it as such through CSS.  It's possible that all the major ones work correctly now, though?  I'd have to test to be sure.

[15:26:02.0000] <TabAtkins>
And I can't test properly until I get my Windows box.

[15:38:00.0000] <JonathanNeal>
I was just wondering because I couldn't find a requirement or recommendation in the spec (maybe I'm blind again)

[15:38:01.0000] <JonathanNeal>
http://www.iecss.com/whatwg.html experimental comparison

[15:39:00.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#the-hr-element-0 is all there is about <hr> styles

[15:39:01.0000] <Hixie>
no vertical margins currently

[15:40:00.0000] <JonathanNeal>
Hixie, right .. that's what I was curious about.

[15:40:01.0000] <TabAtkins>
IE* and 9 seem to be setting up the properties correctly, they just set a few additional ones as well for some reason.  No clue why it has overflow or text-align on it.

[15:40:02.0000] <TabAtkins>
Hixie: I think modern browsers have converged on .5em vertical margins on <hr>.  Can we have that in the spec?

[15:41:00.0000] <Hixie>
file a bug

[15:41:01.0000] <Hixie>
but yes

[15:41:02.0000] <TabAtkins>
Will do.

[15:42:00.0000] <Hixie>
also mention that i don't set 'height' if size="" isn't present currently

[15:42:01.0000] <Hixie>
oh i guess i don't need to, since i don't set box-sizing

[15:42:02.0000] <Hixie>
so nevermind that

[15:43:00.0000] <TabAtkins>
Never minded.

[15:45:00.0000] <TabAtkins>
You don't need to talk about height *anyway*.  You never manipulate height or width on it - it just acts like a normal empty block box.

[15:49:00.0000] <JonathanNeal>
but vertical margin may still added?

[15:50:00.0000] <TabAtkins>
Yes.  <hr> should have "margin: .5em auto;" on it.

[15:51:00.0000] <JonathanNeal>
Since I'm so hit and miss, I'll ask another question I couldn't directly identify in the spec --- is there a recommended font-family (serif?), font-size (16px?), line-height (1.2)?

[15:52:00.0000] <Hixie>
no

[15:52:01.0000] <Hixie>
that's css's territory

[15:56:00.0000] <JonathanNeal>
Most browsers settle on something like 16px/1.2 and some kind of serif.

[15:56:01.0000] <JonathanNeal>
By most I mean all major browsers (though I'm not sure if they do it by px)

[15:58:00.0000] <TabAtkins>
I think we leave that explicitly up to the browser.  It relies a lot on user defaults and such.

[16:00:00.0000] <Hixie>
16px really should be made the only allowed value for font-size:medium, imho

[16:00:01.0000] <Hixie>
http://www.hixie.ch/specs/css/font-size-ui/font-size-ui

[16:06:00.0000] <JonathanNeal>
neato Hixie

[16:19:00.0000] <JonathanNeal>
All right, so these are what I gather would be the requirements and recommendations of whatwg @ http://www.iecss.com/whatwg.css

[16:29:00.0000] <TabAtkins>
JonathanNeal: I'll go through and vet it in a bit.

[16:29:01.0000] <JonathanNeal>
TabAtkins, that would be killer awesome, I would totally appreciate it.


2010-04-06
[17:43:00.0000] <JonathanNeal>
/me patiently awaits TabAtkins vetting.  The sensation must be similar to pregnancy results.

[17:52:00.0000] <othermaciej>
JonathanNeal: that is an odd way to organize the rules

[17:53:00.0000] <JonathanNeal>
Do you mean the entire sections, like /* Display types */?  I organized that per the spec.  If you mean the rules within those sections, you're right, I organized them alphabetically by the first property.

[17:54:00.0000] <JonathanNeal>
It could be organized better for readability in that sense, would you recommend I reorganize it so display:none is followed by display:block etc?

[17:56:00.0000] <othermaciej>
JonathanNeal: I'm not sure what the purpose of your document is, so I don't really have a preference

[17:56:01.0000] <JonathanNeal>
othermaciej, really?  Well, it's to write out a css stylesheet of all the presentational requirements and recommendations made by the WHATWG for HTML5.

[18:02:00.0000] <othermaciej>
JonathanNeal: we'll probably have most of these rules in our UA stylesheet eventually, but probably organized by element rather than by specific property value

[18:21:00.0000] <estellevw>
JonathanNeal: I would agree with the organization comments.  I would do a little more grouping by element type and a little less by property. So would group the heading margins and font sizes together for example. 

[20:10:00.0000] <JonathanNeal>
Hello

[20:14:00.0000] <TabAtkins_>
Yo, JonathanNeal, I don't see <details> in the display section.

[20:15:00.0000] <JonathanNeal>
Checking

[20:16:00.0000] <JonathanNeal>
Having some trouble reading the spec.

[20:17:00.0000] <JonathanNeal>
First time I've had to switch OUT of Chrome.

[20:18:00.0000] <JonathanNeal>
Ah, I see it's defined in a paragraph and not as CSS, adding.

[20:19:00.0000] <JonathanNeal>
Okay, added.

[21:29:00.0000] <JonathanNeal>
TabAtkins_, was the rest up to par?

[21:30:00.0000] <TabAtkins_>
Didn't finish looking at it before I left work.

[01:48:00.0000] <Hixie>
annevk: yt?

[01:52:00.0000] <annevk>
Hixie, yo

[01:52:01.0000] <Hixie>
annevk: FormData for a <form> -- should I add a method to do it, or are you creating a constructor for it?

[01:52:02.0000] <annevk>
btw, the form submission stuff remembered me of something, I haven't verified it yet, but I believe implementations do UTF-16 -> UTF-8 there too

[01:53:00.0000] <Hixie>
don't i have that covered already? file a bug if you want me to look at it

[01:53:01.0000] <Hixie>
and let me know what you mean exactly :-)

[01:53:02.0000] <annevk>
Hixie, I can add a constructor for it

[01:53:03.0000] <Hixie>
k

[01:54:00.0000] <Hixie>
thanks

[01:54:01.0000] <annevk>
though is there a generic name/value algorithm for associated form controls defined at this point?

[01:54:02.0000] <annevk>
iirc it's intertwined with submission

[01:54:03.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#constructing-form-data-set

[01:54:04.0000] <annevk>
cool

[01:55:00.0000] <Hixie>
i suppose i should make that its own separate section

[01:55:01.0000] <Hixie>
feel free to file a bug for that

[01:55:02.0000] <Hixie>
if you want it more cleanly separated out

[01:55:03.0000] <Hixie>
(i expect for your purposes you'll be ignoring the "type" entry in the tuple)

[01:56:00.0000] <annevk>
aah, it was already handled

[01:56:01.0000] <annevk>
"Otherwise, if the document's character encoding is an ASCII-compatible character encoding, then that is the selected character encoding."

[01:56:02.0000] <Hixie>
cool

[01:57:00.0000] <annevk>
I searched the document for UTF-16 before, as that was used for URLs

[02:10:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9427 -- with bugs like these we'll never get done

[02:13:00.0000] <Hixie>
closed

[02:13:01.0000] <Hixie>
well, resolved

[03:27:00.0000] <zcorpan>
ukai: hi

[04:43:00.0000] <remysharp>
quick question: I see how formnovalidate attrib works, but I don't see an example of how novalidate is used

[04:46:00.0000] <annevk>
<form novalidate> ?

[04:47:00.0000] <remysharp>
sure, but if I put novalidate on the form, why bother putting an attribute like required on the input field?

[04:47:01.0000] <annevk>
E.g. with forced previews you could put it on <form> though I suppose in most cases you might as well put it on the button

[04:47:02.0000] <remysharp>
not sure what a forced preview is - sorry!

[04:48:00.0000] <remysharp>
the example in the spec for formnovalidate makes absolute sense, but I can't see the use case for novalidate

[04:48:01.0000] <annevk>
remysharp, e.g. you have to roundtrip with the server at least once before you can actually submit a comment

[04:48:02.0000] <annevk>
remysharp, for that first roundtrip validation is not needed, but user feedback might still be useful

[04:49:00.0000] <remysharp>
I still don't see how this works in <form novalidate>

[04:49:01.0000] <remysharp>
sorry, I don't mean to sound like a pain/muppet

[04:49:02.0000] <remysharp>
trying to understand the difference :)

[04:50:00.0000] <remysharp>
one mo, let me mock up my understanding of formnovalidate, and perhaps you can tweak the source to show me how novalidate would work - if that's cool?

[04:52:00.0000] <annevk>
sure

[04:53:00.0000] <remysharp>
example: http://jsbin.com/inaxo3 - editable source: http://jsbin.com/inaxo3/edit

[04:53:01.0000] <remysharp>
so I've got a formnovalidate on the "save" button

[04:53:02.0000] <remysharp>
annevk: hit "new revision" to save any changes you make

[04:57:00.0000] <annevk>
my form would have only one submit button

[04:58:00.0000] <annevk>
see the form here: http://annevankesteren.nl/2010/03/css-wg-brainstorming#comments

[04:58:01.0000] <remysharp>
sure - but imagine the form is some big ebay style thing where there's lots of content being saved

[04:59:00.0000] <remysharp>
it's a contrived example I know :)

[04:59:01.0000] <remysharp>
it's more to show the formnovalidate attribute working

[04:59:02.0000] <remysharp>
but I'm struggling to see how novalidate comes in to play

[04:59:03.0000] <annevk>
sure, i was just saying that for preview you might not care about validation yet

[05:00:00.0000] <annevk>
and therefore add novalidate to <form>

[05:00:01.0000] <remysharp>
if you use it on the form element, then what's the point in having validation on the form at all?

[05:00:02.0000] <annevk>
user feedback while entering controls

[05:00:03.0000] <annevk>
you still have the form controls apply and all etc. it just doesn't prevent submission

[05:01:00.0000] <annevk>
sites might want to use this too if they don't care if the user got the email address correct or not

[05:01:01.0000] <annevk>
but still want the control to integrate with the address book and create better input methods

[05:01:02.0000] <remysharp>
okay, *that's* what I was looking for

[05:01:03.0000] <remysharp>
perfect, cheers.

[05:02:00.0000] <remysharp>
I can see it being used on the type="email" input, and as you said, still benefits from web forms integration

[05:02:01.0000] <remysharp>
cheers :)

[05:04:00.0000] <annevk>
to be clear, "novalidate" cannot be specified on <input>, just on <form>

[05:05:00.0000] <remysharp>
crap.

[05:05:01.0000] <annevk>
/me wonders if validate and formvalidate would have been better

[05:05:02.0000] <remysharp>
okay, so - say we didn't want to validate the email address field, and used type="email" to still integrate as you suggested

[05:06:00.0000] <annevk>
no, it's either all or nothing

[05:06:01.0000] <remysharp>
but I want to validate the required field of their name

[05:06:02.0000] <remysharp>
crap.

[05:06:03.0000] <annevk>
I just gave email as an example

[05:06:04.0000] <remysharp>
okay.

[05:06:05.0000] <annevk>
for <input required> it's the same

[05:06:06.0000] <annevk>
ideally you would people to fill in a name and give feedback to that effect, but if they don't it's no biggy

[05:07:00.0000] <annevk>
and you put (anonymous + md5(ip)) there or something :)

[05:07:01.0000] <remysharp>
formnovalidate is really useful, I'm not sure novalidate is /that/ useful... :)

[05:07:02.0000] <annevk>
it's mostly there for consistency

[05:08:00.0000] <annevk>
all attributes on <form> for which it makes sense are exposed on submit buttons

[05:11:00.0000] <asmodai>
annevk: managed to overcome the jetlag?

[05:11:01.0000] <annevk>
not sure it makes much sense for enctype come to think of it

[05:11:02.0000] <annevk>
asmodai, yeah, pretty much

[05:12:00.0000] <asmodai>
Good ^^

[05:13:00.0000] <annevk>
warming up for a new jetlag next week :)

[06:17:00.0000] <boblet>
annevk: you’ll have the sento to help you get over that tho, and late night izakaya drinking ;-)

[06:17:01.0000] <boblet>
easy peasy

[06:39:00.0000] <annevk>
boblet, hehe :)

[06:49:00.0000] <annevk>
wait, boblet, you are the same as danield right?

[06:49:01.0000] <annevk>
your whois info confuses me

[06:57:00.0000] <MikeSmith>
boblet is not danield

[07:02:00.0000] <annevk>
wow, my bad

[07:02:01.0000] <annevk>
/me hides

[07:23:00.0000] <boblet>
hey annevk, yep not Daniel. Oli of HTML5Docs IRL ;-)

[07:23:01.0000] <boblet>
hey Rem. ’sup

[07:23:02.0000] <remysharp>
fieldsets and validation.

[07:23:03.0000] <boblet>
annevk: what in whois was confusing?

[07:24:00.0000] <remysharp>
so - fieldset elements have the checkValidity method

[07:24:01.0000] <boblet>
heh, still on that huh :)

[07:24:02.0000] <boblet>
hats off

[07:24:03.0000] <remysharp>
I thought this would trigger the validation against the element collection against that fieldset

[07:24:04.0000] <remysharp>
but it doesn't... so -

[07:24:05.0000] <remysharp>
just wondering if I missed a trick or if someone can explain what is expected when you run fieldset.checkValidity()

[07:25:00.0000] <remysharp>
that was an open question btw :)

[07:31:00.0000] <remysharp>
damn

[07:31:01.0000] <boblet>
heh, poor Remy

[07:31:02.0000] <annevk>
boblet, that your name was not danield :p

[07:31:03.0000] <boblet>
huh

[07:31:04.0000] <annevk>
indeed, my apologies

[07:32:00.0000] <boblet>
annevk: who is this danield fellow, obviously using the good name of boblet for villany!? :)

[07:32:01.0000] <boblet>
np yo

[07:32:02.0000] <annevk>
ourmaninjapan

[07:32:03.0000] <annevk>
quite literally

[07:32:04.0000] <boblet>
aah

[07:32:05.0000] <boblet>
*click*

[07:32:06.0000] <boblet>
that DanielD

[07:33:00.0000] <boblet>
nope, he is far better at Ukulele and far more employed than I :|

[07:38:00.0000] <boblet>
anyone here had experience with Arabic, Hebrew or Korean text? have a ruby question for you if so…

[10:26:00.0000] <JonathanNeal>
hi all

[10:27:00.0000] <MikeSmith>
JonathanNeal: hey

[10:47:00.0000] <JonathanNeal>
What time are TabAtkins and Hixie usually about?  I'm uber excited to get my css vetted.

[10:48:00.0000] <TabAtkins>
Hixie's not usually around until the early afternoon.

[10:48:01.0000] <TabAtkins>
And I'm looking over the CSS *right this minute*.

[10:49:00.0000] <gsnedders>
Hixie just lives weird hours

[10:50:00.0000] <JonathanNeal>
Constantly flying in an airplane ... never knows when it's gonna be day or night.

[10:52:00.0000] <TabAtkins>
No, he just keeps weird hours normally.  I don't think he flies that much.

[10:52:01.0000] <TabAtkins>
He's certainly not a talk-circuit junkie.

[10:53:00.0000] <gsnedders>
He avoids flying much.

[10:53:01.0000] <TabAtkins>
Anyone know if there's an easy setting in Gnome to make focus-follow-mouse, but not autoraise-on-focus?

[10:53:02.0000] <JonathanNeal>
No, I was joking, rather saying that he is constantly in a fantastic flight around the world, rarely touching down, like some kind of airborne hughes.

[10:55:00.0000] <jgraham>
TabAtkins: I doubt it

[10:56:00.0000] <jgraham>
gnome is not known fro exposing loads of settings

[10:56:01.0000] <TabAtkins>
;_;

[10:56:02.0000] <jgraham>
(I love gnome to pieces btw)

[10:56:03.0000] <TabAtkins>
I got so used to focus-follows-mouse after using TweakUI in XP.  It kills me to not have it anymore.

[10:56:04.0000] <TabAtkins>
Having to raise a window just to scroll it is barbaric.

[10:57:00.0000] <jgraham>
Hmm? I just mouse over the window and twiddle the scroll wheel

[10:57:01.0000] <jgraham>
no focus needed

[10:58:00.0000] <TabAtkins>
Okay, true, scrolling works.  But not typing!

[10:58:01.0000] <jgraham>
TabAtkins: If you install CompizConfig settings manager then you can change the option

[10:58:02.0000] <jgraham>
under General Options -> Focus and Raise behaviour

[10:58:03.0000] <TabAtkins>
Excellent.

[10:59:00.0000] <jgraham>
Also, not being able to type in window that you have not focused is a feature not a bug

[10:59:01.0000] <jgraham>
(s/focused/explicitly focused/)

[10:59:02.0000] <TabAtkins>
I explicitly focus windows by putting my mouse in them.

[11:00:00.0000] <gsnedders>
/me explicitly reminds himself that devianART is not productive

[11:00:01.0000] <TabAtkins>
If I want to raise them, I'll click.  But raising should be unrelated to whether it's active.

[11:00:02.0000] <jgraham>
My experience of focus follows mouse is that I typically inexplicably focus windows by placing the mouse over them

[11:00:03.0000] <jgraham>
/me explicitly reminds gsnedders that he is supposed to be on holiday

[11:00:04.0000] <TabAtkins>
jgraham: That happened to me for like a week, then I got used to it and it was much easier.

[11:01:00.0000] <gsnedders>
jgraham: But I could be doing more productive things like reading a book

[11:01:01.0000] <gsnedders>
jgraham: Not staring at a computer screen

[11:01:02.0000] <TabAtkins>
Your notion of 'productive' seems oddly tweaked.

[11:01:03.0000] <jgraham>
TabAtkins: Also you should install the compizconfig thing anyway because otherwise you can't enable the grid plugin and without the grid plugin life just isn't good

[11:01:04.0000] <gsnedders>
Oh, and listening to metal, just to prove I'm not called "James", no matter what people say.

[11:02:00.0000] <gsnedders>
TabAtkins: And writing, that's the other thing I meant to do toady

[11:02:01.0000] <gsnedders>
(number of words written: 0)

[11:02:02.0000] <TabAtkins>
jgraham: that on apt?

[11:02:03.0000] <gsnedders>
jgraham: yeah

[11:02:04.0000] <gsnedders>
or rather, TabAtkins: yeah

[11:02:05.0000] <gsnedders>
gah

[11:02:06.0000] <gsnedders>
/me is too tired

[11:02:07.0000] <jgraham>
gsnedders: Please stop defining reading a book as "productive". It reminds me of the painful hacker news thread where people were trying to defend reading fiction on the basis that you might learn facts about the real world

[11:03:00.0000] <jgraham>
rather than because, say, it is fun

[11:03:01.0000] <TabAtkins>
Named?  apt-get install compizconfig is no good.

[11:03:02.0000] <gsnedders>
jgraham: You might! The Zombie Survival guide is useful!

[11:03:03.0000] <jgraham>
TabAtkins: no idea try synaptic + search

[11:03:04.0000] <gsnedders>
or aptitude if ou think a GUI is too much effort

[11:03:05.0000] <jgraham>
gsnedders: Also, I am worried in case you have taken to referring to me as your doppelganger

[11:04:00.0000] <TabAtkins>
They are!

[11:04:01.0000] <Philip`>
TabAtkins: You could install KDE, and then System Settings > Window Behaviour > Focus > Policy > Focus Follows Mouse

[11:04:02.0000] <gsnedders>
jgraham: In reference to my Facebook status?

[11:04:03.0000] <TabAtkins>
Then I'd have to go install kde, though.  (I do like KDE a bit better, though.)

[11:05:00.0000] <jgraham>
gsnedders: Yeah. Or twitter thing or whatever it was

[11:05:01.0000] <gsnedders>
jgraham: No

[11:05:02.0000] <gsnedders>
jgraham: Merely how I behaved at school, which was nothing like me

[11:05:03.0000] <jgraham>
Philip`: That would have the unfortunate side effect of forcing you to use KDE

[11:05:04.0000] <Philip`>
jgraham: s/un//

[11:05:05.0000] <jgraham>
s//un/

[11:06:00.0000] <TabAtkins>
I hope you didn't mean that last regexp, jgraham.

[11:06:01.0000] <Philip`>
m/oo/n

[11:06:02.0000] <jgraham>
magic irc regexps

[11:06:03.0000] <jgraham>
they have the effect that you indend as long as what you write can be interpreted as vaugely alluding to the right change

[11:07:00.0000] <TabAtkins>
I prefer real regexps, so they can be properly interpreted by our minuting script.

[11:07:01.0000] <jgraham>
It is fortunate then that no one is minuting in here

[11:07:02.0000] <TabAtkins>
Curses, foiled again!

[11:08:00.0000] <jgraham>
I would say it is fortunate that nothing important enough it happening in here for anyone to bother taking minutes but I have seen minuted meetings

[11:08:01.0000] <gsnedders>
/me heads off to do this whole "fun" thing

[11:09:00.0000] <TabAtkins>
I think it's more fortunate that we all minute ourselves voluntarily.

[11:10:00.0000] <jgraham>
Well yeah, the logs are a good thing, no question

[11:10:01.0000] <TabAtkins>
Now, about this compiz business.  Just how much do I have to install of this?  I see a bunch of packages here, and one is named "compiz-gnome".

[11:11:00.0000] <TabAtkins>
I suppose the config one is the compizconfig-settings-manager package?

[11:11:01.0000] <jgraham>
Yes

[11:12:00.0000] <TabAtkins>
/me got far too used to running as root most of the time.

[11:15:00.0000] <TabAtkins>
Yay!  All fixed.

[11:17:00.0000] <TabAtkins>
(There was indeed a setting, in System->Prefs->Windows

[11:19:00.0000] <jgraham>
But now you can use the grid plugin too :)

[11:19:01.0000] <jgraham>
(assuming you got the compiz settings thing installed)

[11:20:00.0000] <TabAtkins>
I installed it, but now I can't figure out how to run it.  Damn these command lines!

[11:21:00.0000] <TabAtkins>
nm, found it.

[11:22:00.0000] <TabAtkins>
Is it the "Window Rules" setting?

[11:50:00.0000] <estellevw>
jonathanNeal: are you going to add article, section, etc to your css?

[11:51:00.0000] <JonathanNeal>
Strange, they're missing, aren't they.

[11:52:00.0000] <JonathanNeal>
Okay, they're in --- strange how they wound up missing.

[11:59:00.0000] <jgraham>
TabAtkins: No it is under plugins -> Grid or something

[12:00:00.0000] <jgraham>
search for grid anyway

[12:00:01.0000] <jgraham>
/me can't check right now as he is on the wrong computer

[12:31:00.0000] <JonathanNeal>
Does the spec say anything about the required / recommended outline for :link ?

[12:45:00.0000] <annevk>
http://www.w3.org/TR/2010/WD-file-writer-api-20100406/

[12:46:00.0000] <annevk>
JonathanNeal, that should be on :link:focus, but the spec does not mention it

[12:46:01.0000] <annevk>
JonathanNeal, you should prolly file a bug

[12:46:02.0000] <othermaciej>
/me wishes File Writer could be a Web Apps spec so I could have some place sensible to send review comments

[12:46:03.0000] <JonathanNeal>
I've never filed a bug ... linkie?

[12:47:00.0000] <TabAtkins>
JonathanNeal: Just go to the spec, click in the section you want to file on, and type in the issue in the box at the bottom.

[12:49:00.0000] <JonathanNeal>
Okay

[12:49:01.0000] <annevk>
othermaciej, oh it's not?

[12:49:02.0000] <othermaciej>
annevk: I think it's DAP

[12:49:03.0000] <annevk>
othermaciej, meh, there's so much overlap

[12:50:00.0000] <JonathanNeal>
Done! :)

[12:52:00.0000] <JonathanNeal>
Firefox suggests the outline should be on anything focusable.

[12:53:00.0000] <JonathanNeal>
Is there a pseudo selector for that?  :selectable ?

[12:53:01.0000] <annevk>
:focus

[12:53:02.0000] <miketaylr>
:focus?

[12:53:03.0000] <JonathanNeal>
There you go.. better to place it on that then :link

[12:54:00.0000] <JonathanNeal>
Although I won't until it's gonna be in the spec.

[14:45:00.0000] <TabAtkins>
jgraham: D'oh, I'm on Ubuntu 8.04, Grid doesn't appear to be installed by default, and make errors out if I follow the install process on the wiki.  ;_;

[14:50:00.0000] <karlushi>
http://www.museumsassociation.org/about all the fonts are recreated with a canvas element and then we can't select the text. View source is working but copy/paste doesn't

[14:52:00.0000] <miketaylr>
yes, cufon has that limitation

[14:57:00.0000] <jgraham>
TabAtkins: Ah

[14:58:00.0000] <jgraham>
It installed for me in 8.10 and comes by default in (at-least) 9.10

[14:58:01.0000] <TabAtkins>
I blame Google.

[14:58:02.0000] <jgraham>
TabAtkins: So does everyone else

[14:58:03.0000] <jgraham>
For everyhing

[14:58:04.0000] <jgraham>
So it's ood to get ome practice in so you are used to it

[14:59:00.0000] <jgraham>
+ some random letters

[15:01:00.0000] <jgraham>
TabAtkins: I guess if you are sticking to LTS versions, 10.04 is not far away

[15:02:00.0000] <gsnedders>
10.10 was going to be LTS and not 10.04, I thought

[15:02:01.0000] <gsnedders>
No, 10.04

[15:02:02.0000] <gsnedders>
/me thought he saw something that said LTS was being pushed to 10.10

[15:03:00.0000] <jgraham>
Ah, well I was just going with the "every two years" thing

[16:00:00.0000] <Hixie>
where is philipj

[16:48:00.0000] <TabAtkins>
MikeSmith: Somehow I stopped receiving public-html email on April 1, and I no longer appear on the member's list.  This might be related to my affiliation change on Thursday.

[16:59:00.0000] <volkmar>
Hixie: ping

[16:59:01.0000] <othermaciej>
TabAtkins: did your AC rep renominate you?


2010-04-07
[17:01:00.0000] <Hixie>
volkmar: here

[17:02:00.0000] <volkmar>
i've posted a message about [PutForwards=value] for htmlFor attribute in output element

[17:02:01.0000] <volkmar>
i've been told you can explain this choice

[17:03:00.0000] <Hixie>
yeah, it's on my pile of feedback

[17:03:01.0000] <Hixie>
in short though the answer is that it makes the authoring experience better

[17:05:00.0000] <TabAtkins>
othermaciej: Yeah, I just talked to TV and he said he did.

[17:08:00.0000] <volkmar>
Hixie: does it make the authoring experince so much better ?

[17:09:00.0000] <volkmar>
Hixie: i think it's a bit odd to set a string to a DOM object to access directly to an attribute

[17:09:01.0000] <volkmar>
i'm not sure it worths it

[17:11:00.0000] <Hixie>
volkmar: it's how window.location works

[17:12:00.0000] <Hixie>
volkmar: i think it's a heck of a lot easier to say  foo.className = 'x y z'; foo.htmlFor = 'a b c';  than to have to make them different as in foo.className = 'x y z'; foo.htmlFor.value = 'a b c';

[17:23:00.0000] <MikeSmith>
TabAtkins: I'll check now

[17:24:00.0000] <MikeSmith>
TabAtkins: so I see two accounts for you in the database

[17:25:00.0000] <MikeSmith>
both affiliated correctly

[17:25:01.0000] <MikeSmith>
one of those accounts shows you as an active participant in the CSS WG

[17:26:00.0000] <MikeSmith>
but it doesn't show you as a participant in the HTML WG

[17:26:01.0000] <MikeSmith>
the other account doesn't have any working groups associated with it at all

[17:30:00.0000] <volkmar>
Hixie: ok, then let's go with that

[17:30:01.0000] <volkmar>
Hixie: thanks :)

[17:35:00.0000] <Hixie>
volkmar: np

[17:48:00.0000] <othermaciej>
TabAtkins: I guess poking MikeSmith is the right next step then

[17:48:01.0000] <othermaciej>
ah, I see MikeSmith is on it already

[17:50:00.0000] <MikeSmith>
yeah

[17:51:00.0000] <MikeSmith>
I can't find any record of Tab being re-added, so I've asked the systems team to look into it

[17:57:00.0000] <hober>
TabAtkins: re: your email, I'll go ahead and convert my ISSUE-95 zero-edit CP from Markdown to MediaWiki syntax & I'll get it up on one of the WGs' wikis. We can work on generalizing it from there

[18:07:00.0000] <TabAtkins>
hober:  kk

[18:07:01.0000] <TabAtkins>
MikeSmith: Thanks.

[18:51:00.0000] <Hixie>
maciej, you should configure your mail client to alert you when you include the string "file://" in it. :-P

[19:07:00.0000] <annevk>
blahdieblah

[19:07:01.0000] <annevk>
just because

[19:08:00.0000] <othermaciej>
Hixie: d'oh!

[19:08:01.0000] <othermaciej>
Hixie: maybe you should configure yours to automatically translate to http://dev.w3.org/ URLs :-)

[23:22:00.0000] <Hixie>
i like how shelley thinks styling controls is some new topic that's never come up before

[23:22:01.0000] <Hixie>
given that i've written entire specifications (e.g. xbl2) on the topic

[23:27:00.0000] <othermaciej>
Hixie: I didn't even mention XBL2 as part of the story for deep styling of custom controls, but of course it is a potential part of the toolset

[23:27:01.0000] <othermaciej>
maybe I am just embarrassed that we haven't implemented it yet

[23:27:02.0000] <Hixie>
xbl2 is even designed to be extended to have predefined control bindings, much like what html5 references in the rendering section

[23:28:00.0000] <Hixie>
and supports the pseudos that people are suggesting

[23:28:01.0000] <Hixie>
these are all essentially solved problems at the theoretical level... it's all coding from now on, if y'all agree with my approach. :-P

[23:30:00.0000] <othermaciej>
I don't think XBL2 solves the problem of minting brand new pseudos... or does it let you make up arbitrary new ones?

[23:31:00.0000] <othermaciej>
also doesn't define all the right appearance property values

[23:31:01.0000] <othermaciej>
I think XBL2 is a good tool, but shouldn't be the only way to style form controls

[23:31:02.0000] <Hixie>
we'd need some pseudos from the csswg, and yeah, appearance needs rewriting

[23:31:03.0000] <othermaciej>
nor should form controls have to be implemented with XBL2 to be stylable, although it may turn out to be a convenient implementation technique at some point

[23:31:04.0000] <Hixie>
well, depends what you mean by "only way"

[23:32:00.0000] <Hixie>
i think conceptually it makes sense to pretend the form control is a binding, for expository purposes if not necessarily implementation purposes

[23:32:01.0000] <Hixie>
but anyway

[23:32:02.0000] <Hixie>
i think all the concepts are in place, though some polishing may be needed in various places

[23:33:00.0000] <othermaciej>
well one thing I'd like to see defined is have clear conformance requirements for simply putting CSS properties on <input type=text> and <input type=submit>

[23:34:00.0000] <Hixie>
if we had a conceptual binding combined with a well-defined 'appearance' property, that would be one solution that would work pretty well

[23:34:01.0000] <Hixie>
and would extend well to a world with xbl2

[23:34:02.0000] <othermaciej>
Hixie: let me be clear about the basic level I am talking about

[23:34:03.0000] <othermaciej>
the styling that <http://www.google.com/> applies to the text field and two buttons...

[23:35:00.0000] <othermaciej>
I would like the result of that to be interoperable and defined by spec

[23:36:00.0000] <othermaciej>
per spec, the effect of the CSS properties put on those controls is completely implementation-defined

[23:36:01.0000] <Hixie>
anyone who tries to write a spec to do that will fail to prevent scope creep all the way to defining every property, unless they are _extremely_ resistant to feedback

[23:36:02.0000] <Hixie>
hence why i'd just go all the way and just define it in a way that uses conceptual bindings and true 'appearance' work

[23:36:03.0000] <othermaciej>
yes, you have to define how every property works

[23:37:00.0000] <othermaciej>
I don't see how appearance helps

[23:37:01.0000] <othermaciej>
that page doesn't set the appearance property

[23:37:02.0000] <Hixie>
it's set by the UA sheet

[23:37:03.0000] <othermaciej>
nor do I see how "conceptual bindings" help, because that page does not use them either

[23:37:04.0000] <Hixie>
they're used by the UA sheet

[23:38:00.0000] <Hixie>
if you define how setting a property overrides 'appearance', and how setting a property affects specific parts of a binding, you ground your definition in something useful

[23:38:01.0000] <Hixie>
it's not the only solution, certainly

[23:38:02.0000] <othermaciej>
overriding 'appearance', I see your point

[23:38:03.0000] <Hixie>
but i think other solutions are far less likely to be sanely scalable or workable in the future

[23:38:04.0000] <othermaciej>
'specific parts of a binding', I don't see how that is relevant to text fields or buttons

[23:38:05.0000] <othermaciej>
I can see how it is relevant to compound controls

[23:38:06.0000] <Hixie>
for text fields and buttons, maybe not

[23:38:07.0000] <othermaciej>
that need to have subparts individually styled

[23:39:00.0000] <Hixie>
but e.g. for <input type=file>, you need to say what the element's styles apply to

[23:39:01.0000] <Hixie>
anyway

[23:39:02.0000] <Hixie>
i'm not planning on doing this work

[23:39:03.0000] <Hixie>
and if others want to do it in a different way, i'm certainly not going to stop them

[23:39:04.0000] <boblet>
annevk: you there?

[23:42:00.0000] <othermaciej>
Hixie: yeah, for <input type=file> we have a separate pseudo for the button

[23:43:00.0000] <othermaciej>
Hixie: I hope at some point we can find a person with the skills and experience to specify this.

[23:43:01.0000] <Hixie>
add it to the list

[23:52:00.0000] <othermaciej>
it's a big list!

[23:53:00.0000] <othermaciej>
I'm thinking a bit less about the big list these days because I want to focus on getting HTML5 through the W3C meat grinder

[23:53:01.0000] <othermaciej>
cause that's gonna be some delicious sausage when it's done

[00:09:00.0000] <othermaciej>
TabAtkins: lol at "CSS T&A"

[00:45:00.0000] <JonathanNeal>
hello all

[00:45:01.0000] <estellevw>
hi

[00:47:00.0000] <JonathanNeal>
what's new?

[00:47:01.0000] <estellevw>
In terms of the discussion from last hour, i think generally the appearance of input elements should be left out of the HTML5 spec, other than most general specifications. The appearance, other than box, checkbox, radio button, button and button with box is all that should be stated in the spec... 

[00:48:00.0000] <estellevw>
i would like to see the input type file enable styling of the button and box too

[00:50:00.0000] <JonathanNeal>
left out, okay.

[00:50:01.0000] <estellevw>
in terms of "Hixie: i'm not planning on doing this work", I am fairly new here, but willing to do the work... need guidance on what to do though

[00:51:00.0000] <estellevw>
sorry JonathanNeal... nothing new, other than dealing with a skunk spray this evening... but that has noting to do with HTML5

[00:55:00.0000] <Hixie>
estellevw: TabAtkins would be the guy to coordinate with -- he works with the css working group more than most of us here

[00:55:01.0000] <estellevw>
thanks. 

[00:57:00.0000] <estellevw>
But putting that the element should have two components does not seem to be in the spec. 

[00:58:00.0000] <estellevw>
There is more description of appearance in the spec for other imputs, like 'image' stating it's an image or a button

[00:58:01.0000] <estellevw>
but i'll check the css wg archives

[01:59:00.0000] <JonathanNeal>
what's the current status of microformatting in html5?  Can I set the profile in head or something?

[01:59:01.0000] <zcorpan>
you can use microformats, but you can't use the profile attribute

[02:00:00.0000] <JonathanNeal>
do i have to wrap everything in a class="vcard" still?

[02:03:00.0000] <JonathanNeal>
I'm interested in adopting the hcard h* microformats, but it seems unnecessary to always wrap everything in a vcard

[02:05:00.0000] <zcorpan>
why are you interested? what are you trying to do?

[02:08:00.0000] <JonathanNeal>
zcorpan, I'm interested because it pleases me, you know me by now?  Always trying to adopt and standardize. :)

[02:09:00.0000] <JonathanNeal>
I would have made a great pharisee. 

[02:17:00.0000] <zcorpan>
JonathanNeal: personally i wouldn't adopt anything without a clear reason, i.e. trying to solve a problem or enabling something cool

[02:17:01.0000] <JonathanNeal>
Well, it would be cool to have Google picking up my content mo' betta.

[02:18:00.0000] <JonathanNeal>
It would allow me to standardize some classnames now and again, making my development mo' readable to other developers.

[02:18:01.0000] <zcorpan>
i doubt that adding microformats will make it more readable

[02:21:00.0000] <JonathanNeal>
Well, if there is a popularized verbatim of classnames, then using said classnames would more-than-not make the content more readable to other readers, especially when those readers are fellow developers or search engines.

[02:22:00.0000] <JonathanNeal>
more readable for readers, yo

[02:26:00.0000] <annevk>
boblet, am now

[02:27:00.0000] <boblet>
hey. Just want to check how to pronounce your name for katakana-ization

[02:27:01.0000] <boblet>
anne or an-nay?

[02:28:00.0000] <boblet>
(or should that be ann or an-nay)

[02:28:01.0000] <boblet>
I guess van is as expected. Kesteren too?

[02:31:00.0000] <annevk>
an-nuh or some such

[02:32:00.0000] <annevk>
the rest is as expected, yay

[02:32:01.0000] <annevk>
an-nuh-san, even :p

[02:32:02.0000] <svl>
annevk: don't you think "as expected" with "van" would be the vehicle?

[02:32:03.0000] <zcorpan>
it's annö

[02:32:04.0000] <boblet>
aah, nice

[02:33:00.0000] <annevk>
svl, true, but I care less :)

[02:33:01.0000] <boblet>
thought I didn’t have it quite right

[02:34:00.0000] <boblet>
svl:  unless you were enquiring about pronunciation

[02:34:01.0000] <boblet>
:|

[02:34:02.0000] <zcorpan>
van is like 'fun'

[02:34:03.0000] <boblet>
heh

[02:34:04.0000] <svl>
like the "an" in the latin word "anno"

[02:34:05.0000] <boblet>
ohrly. svl, my apologies

[02:35:00.0000] <svl>
:)

[02:35:01.0000] <boblet>
hrm, I guess アンヌ・ヴン・ケステレン

[02:36:00.0000] <boblet>
かな？

[02:36:01.0000] <boblet>
MikeSmith: yt?

[02:36:02.0000] <MikeSmith>
yeah

[02:36:03.0000] <boblet>
that look right to you?

[02:36:04.0000] <MikeSmith>
yep

[02:36:05.0000] <boblet>
cool

[02:36:06.0000] <annevk>
アンヌ・ヴン・ケステレン is me in Japanese?

[02:37:00.0000] <boblet>
yep

[02:37:01.0000] <annevk>
sweet

[02:37:02.0000] <boblet>
badass huh ;-)

[02:37:03.0000] <annevk>
I need to get a tattoo now

[02:40:00.0000] <jgraham>
Maybe stick with a t-shirt?

[02:47:00.0000] <svl>
google translate back says Susan Anne Kesteren; I'm assuming susan is anne, and anne is van?

[02:48:00.0000] <MikeSmith>
svl: that's somewhat bizarre

[02:50:00.0000] <svl>
アン ファン ケステレン is its translation the other way round

[02:51:00.0000] <estellevw>
jonathanNeal are you the one who did http://www.iecss.com/whatwg.css

[02:52:00.0000] <MikeSmith>
svl: which is "Ann Fan Kesteren"

[02:53:00.0000] <payman>
annevk: http://dev.w3.org/csswg/cssom/#introduction : s/where/were ?

[02:57:00.0000] <annevk>
o_O thanks

[03:02:00.0000] <myakura>
I would spell like アナ・ヴァン・ケステレン but I'm not sure if this sound as expected.

[03:04:00.0000] <MikeSmith>
myakura: zcorpan said the "van" part is pronounced more like "fun"

[03:04:01.0000] <MikeSmith>
schwa sound, I guess

[03:05:00.0000] <myakura>
MikeSmith: Ah.

[03:05:01.0000] <MikeSmith>
but there is no schwa sound in Japanese

[03:05:02.0000] <myakura>
right.

[03:05:03.0000] <MikeSmith>
but I guess ヴン is more like "foon", not "fun"

[03:05:04.0000] <MikeSmith>
boblet: so maybe ヴァン is better

[03:06:00.0000] <svl>
The starting letter is definitely a "v", not an "f"; the a is the problem when trying to find an English equivalent

[03:06:01.0000] <svl>
maybe like "varsity"

[03:09:00.0000] <annevk>
it's somewhat like fun, but with a really weak f

[03:11:00.0000] <MikeSmith>
myakura's ヴァン suggestion is the closest approximation in Japanese, I think

[03:12:00.0000] <MikeSmith>
dammit

[03:12:01.0000] <MikeSmith>
my skype on Mac appears to be making a quacking noise

[03:12:02.0000] <MikeSmith>
I have no idea why

[03:13:00.0000] <MikeSmith>
help

[03:13:01.0000] <MikeSmith>
please help me stop the quacking

[03:16:00.0000] <myakura>
Ah. I remember that the "van" part in Pieter van den Hoogenband is spelled as ファン so the Japanese could live with that :)

[03:17:00.0000] <annevk>
ah, that would at least be consistent :)

[05:11:00.0000] <jgraham>
Am I blind or is window.stop missing from HTML5? Is that deliberate (does IE support it?)

[05:13:00.0000] <Lachy>
jgraham, it does appear to be missing

[05:13:01.0000] <Lachy>
do you know if it's supported in webkit and opera?

[05:14:00.0000] <jgraham>
Lachy: Well w(window.stop) in the ldv shows function() {/* native code*/} so I guess yes

[05:14:01.0000] <jgraham>
but I haven't actually compared behaviour

[05:17:00.0000] <Lachy>
IE8 doesn't support window.stop()

[07:34:00.0000] <annevk>
what is window.stop()?

[07:34:01.0000] <jgraham>
I think it interrupts parsing

[07:35:00.0000] <jgraham>
Beyond that I don't know; I hadn't heard of it untill this morning

[07:35:01.0000] <annevk>
funnest

[07:36:00.0000] <jgraham>
There are a few bugs in our bug tracker about it at least

[07:36:01.0000] <jgraham>
Although some of them are userjs and similar non-web things

[07:49:00.0000] <Lachy>
AIUI, window.stop() basically behaves like the browser's stop button

[07:50:00.0000] <Lachy>
I'm not really sure what the use case for the method is though

[07:56:00.0000] <annevk>
seems vital for implementing a browser in a browser

[07:56:01.0000] <gsnedders>
And we all know the browser OS is the future

[07:56:02.0000] <annevk>
and what's an OS without a browser?

[07:57:00.0000] <annevk>
it's going to be so recursive it hurts

[08:00:00.0000] <zcorpan>
we should implement a browser in flash, to implement flash

[08:01:00.0000] <JonathanNeal>
estellevw, yes

[08:02:00.0000] <JonathanNeal>
Hii! :)

[08:05:00.0000] <annevk>
ok, epic after-lunch-e-mail overload

[08:05:01.0000] <gsnedders>
/me has epic after-Easter e-mail overload

[08:05:02.0000] <gsnedders>
I don't dare start on it though, even if I am working a bit today :P

[08:28:00.0000] <JonathanNeal>
gsnedders, I just realized I use your outliner :)

[08:28:01.0000] <JonathanNeal>
/me gives him a paper and pen for the ol' autograph

[08:43:00.0000] <Dashiva>
So "resource" is short for "representation of resource" when convenient. How convenient.

[08:43:01.0000] <Philip`>
Abbreviate it to "re'source"

[08:44:00.0000] <Philip`>
That should avoid any potential ambiguity

[08:46:00.0000] <jgraham>
Is it too obvious to suggest namespaces?

[08:50:00.0000] <jgraham>
Actually I don't understand this sentence at all: """a URI can identify a

[08:50:01.0000] <jgraham>
resource that is an SVG document which has a representation

[08:50:02.0000] <jgraham>
that is also and SVG document.

[08:50:03.0000] <jgraham>
"""

[08:50:04.0000] <jgraham>
How (in the HTTP sense) can a resource be a concrete type?

[08:50:05.0000] <jgraham>
e.g. "a resource that is a SVG document"

[08:51:00.0000] <zcorpan>
/me wonders how jgraham pronounces SVG

[08:53:00.0000] <jgraham>
I think that was more like a typo than anything deep

[08:53:01.0000] <jgraham>
I mean I pronounce it ess-vee-gee. That's what everyone does, right?

[08:53:02.0000] <jgraham>
What else could you do? svug?

[08:54:00.0000] <zcorpan>
soovge

[08:55:00.0000] <zcorpan>
like doovde and ready for de hud

[08:55:01.0000] <jgraham>
I can't even say that with it written out

[08:56:00.0000] <annevk>
the transition and animation thread should take a break

[09:05:00.0000] <gsnedders>
JonathanNeal: *signs*

[09:05:01.0000] <JonathanNeal>
woooow

[09:05:02.0000] <JonathanNeal>
Are microformat and HTML5 microdata mutually exclusive?

[09:06:00.0000] <jgraham>
No

[09:06:01.0000] <annevk>
is the native elements vs scripted thread a rehash of older debates?

[09:08:00.0000] <JonathanNeal>
jgraham, it seems like it would be a lot of clutter to follow both --- is there a preference?  I really like that microformat has its own active community, but I'm just learning about this all.  I'm not sure how the two work together.

[09:09:00.0000] <jgraham>
JonathanNeal: Well you have sort of decided on clutter if you do either

[09:09:01.0000] <Dashiva>
The idea is that new (or updated old) microformats would be layered on top of microdata

[09:10:00.0000] <JonathanNeal>
jgraham, they're not TERRIBLY cluttered.  The documentation for microdata feels cluttered (especially after reading microformats), but Dashiva that makes sense --- oi but to learn microdata doesn't seem as friendly.

[09:11:00.0000] <Dashiva>
If you ignore the parts added to support RDF use cases, it's pretty simple

[09:11:01.0000] <jgraham>
JonathanNeal: I guess microdata hasn't been through all the blogs and round the conference circuit until the explainations are smoothly polished yet

[09:14:00.0000] <jgraham>
But it shouldn't be that hard. You should become an expert and get invited to do well paid speaking gigs or something

[09:16:00.0000] <JonathanNeal>
jgraham, har har :)

[09:18:00.0000] <JonathanNeal>
Except Google has posted about it- so I imagine someone else has.

[09:21:00.0000] <gsnedders>
/me doesn't like it when people try and bully him into speaking on stages

[09:22:00.0000] <JonathanNeal>
in that case we'll just bully you into including the multiple headings in the hgroup in the outliner, at least optionally :)

[09:22:01.0000] <gsnedders>
/me is amazed at how much use people get out of something he hacked together in 30 mins

[09:23:00.0000] <gsnedders>
And how much more time people want me to spend on it :P

[09:23:01.0000] <JonathanNeal>
Well, it's not as easy to write in php, so we love you for it.

[09:23:02.0000] <gsnedders>
It wouldn't be that hard to write in PHP

[09:24:00.0000] <JonathanNeal>
... I tried :\

[09:24:01.0000] <JonathanNeal>
/me hangs his head low. awww

[09:24:02.0000] <gsnedders>
You just need, to do something very basic, DOMDocument::loadHTML and some method of iterating over the DOM

[09:24:03.0000] <gsnedders>
Like http://github.com/gsnedders/complexpie/blob/master/src/domiterator.php

[09:26:00.0000] <gsnedders>
Then you just need to create some basic structure to store the classes in

[09:27:00.0000] <gsnedders>
/me thinks the challenge of PHP is as much knowing the language well enough to do stuff in reasonable ways as well as learning how to write clean PHP

[09:30:00.0000] <JonathanNeal>
Is <address /> really only meant for article and body, not just any sectioning element?

[09:30:01.0000] <JonathanNeal>
"The address element represents the contact information for its nearest article or body element ancestor"

[09:31:00.0000] <zcorpan>
JonathanNeal: it is, but you can put it in any sectioning element

[09:32:00.0000] <zcorpan>
or indeed pretty much anywhere

[09:34:00.0000] <JonathanNeal>
That seems like something that should be clarified or decided upon.

[09:34:01.0000] <miketaylr>
would you want a seperate <address> for a <nav>?

[09:35:00.0000] <miketaylr>
s/seperate/separate/

[09:35:01.0000] <JonathanNeal>
Well, would you distribute the nav as independant content?

[09:36:00.0000] <miketaylr>
of course not

[09:36:01.0000] <JonathanNeal>
s/independant/independent

[09:36:02.0000] <miketaylr>
:)

[09:37:00.0000] <JonathanNeal>
Well, I believe <nav> is sectioning content, which is why gsnedders goes as far as to "untitled section" it when its left out, so an optional <address /> allowed in any sectioning element means that yes, it would be allowed in <nav />

[09:37:01.0000] <miketaylr>
right

[09:37:02.0000] <JonathanNeal>
Maybe <nav> shouldn't be a sectioning element :)

[09:38:00.0000] <miketaylr>
but the way its defined now, that wouldn't imply that the <address> content denotes the author of the <nav>

[09:38:01.0000] <miketaylr>
but the body or article its found in

[09:38:02.0000] <JonathanNeal>
Sure, so what is the benefit of <nav /> being its own section?

[09:39:00.0000] <JonathanNeal>
Because if its sectioning content because you can include other copy or independent 

[09:39:01.0000] <JonathanNeal>
s/its/it's

[09:40:00.0000] <JonathanNeal>
... document-ish stuff, then <address /> actually works.

[09:41:00.0000] <miketaylr>
sure, you can stick an <address> in there

[09:42:00.0000] <miketaylr>
but i don't see how that affects its representation of "the contact information for its nearest article or body element ancestor"

[09:44:00.0000] <JonathanNeal>
it doesn't, so maybe it's a potential yet impractical bug (still find nav impractical as sectioning content) ... I should continue this discussion in my nice, cool office.

[09:45:00.0000] <JonathanNeal>
Be back in 20 or so, miketaylr thank you very, very much for the discussion.

[09:45:01.0000] <miketaylr>
/me waits in his stuffy office

[10:26:00.0000] <volkmar>
the current implementation of the keygen elements are all based on the Netscape keygen element ?

[10:27:00.0000] <jgraham>
volkmar: AFAIK it is that and a wing and a prayer

[10:28:00.0000] <volkmar>
:)

[10:28:01.0000] <JonathanNeal>
back!

[10:28:02.0000] <volkmar>
i was reading carefuly the keygen element specs this morning and it sounds weird to have an interactive element with the autofocus attribute which has no mandatory UI

[10:29:00.0000] <gsnedders>
JonathanNeal: Don't blame me, I just implemented the spec.

[10:29:01.0000] <jgraham>
JonathanNeal: fingers!

[10:31:00.0000] <JonathanNeal>
knees and toes, knees and toes

[10:33:00.0000] <annevk>
volkmar, is there no UI suggested in the rendering section?

[10:34:00.0000] <volkmar>
annevk: oups, i did not check

[10:34:01.0000] <annevk>
/me did not either

[10:38:00.0000] <volkmar>
annevk: "the element is expected to render as an 'inline-block' box containing a user interface to configure the key pair to be generated."

[10:39:00.0000] <volkmar>
that is quite similar to what we have in the keygen element spec

[10:43:00.0000] <annevk>
not much indeed

[10:45:00.0000] <volkmar>
at the moment, all UA implementing the keygen element have a list to select the key length and a text field

[10:45:01.0000] <volkmar>
but the text field looks to be read only on opera

[10:46:00.0000] <volkmar>
and I suppose it is used for the algorithm

[10:46:01.0000] <volkmar>
should that be added to the specs ?

[10:54:00.0000] <annevk>
well, it really depends on the UA how they expose it

[10:56:00.0000] <annevk>
could just be a selectbox, the field could be completely readonly i suppose

[10:56:01.0000] <annevk>
if the ua doesn't allow for picking sizes

[11:08:00.0000] <JonathanNeal>
Microdata, Microformats, RDFa --- a sordid tale

[11:36:00.0000] <TabAtkins_>
Anyone know if gradient support was removed from FF 3.5?  I know it was working before, but now my wife can't see any gradients in her 3.5.7.

[11:36:01.0000] <TabAtkins_>
And now the MDC page says it's a 3.6 feature.

[11:36:02.0000] <miketaylr>
hmm, i didn't even know it was in 3.5 at all

[11:38:00.0000] <TabAtkins_>
Yeah, as -moz-linear-gradient() and -moz-radial-gradient()

[11:38:01.0000] <miketaylr>
do you have a test page?

[11:39:00.0000] <TabAtkins_>
I know it was working a month ago when I added it to my old company's main site.

[11:39:01.0000] <miketaylr>
i suppose i could find one myself :P

[11:39:02.0000] <TabAtkins_>
miketaylr: www.xanthir.com/etc/gradient.html

[11:39:03.0000] <miketaylr>
i'll start with 3.5.0 : http://mac.oldapps.com/firefox.php

[11:39:04.0000] <miketaylr>
:)

[11:39:05.0000] <JonathanNeal>
hey speaking of gradients, opera :)

[11:40:00.0000] <TabAtkins_>
Does Opera do gradients now?  Prefixed, hopefully?

[11:40:01.0000] <estellevw>
I tested these just last months

[11:40:02.0000] <estellevw>
let me see if i can find my sandbox

[11:40:03.0000] <estellevw>
i mean, browser results

[11:40:04.0000] <annevk>
we don't

[11:41:00.0000] <TabAtkins_>
That's what I thought, annevk.

[11:41:01.0000] <annevk>
we were waiting for syntax to stabilize

[11:43:00.0000] <daedb>
Has it stabilized now?

[11:43:01.0000] <JonathanNeal>
annevk, by stabilize do you mean for moz / webkit to support a uniform pattern?

[11:43:02.0000] <miketaylr>
TabAtkins_: is it supposed to look like this? http://miketaylr.com/post/f30cd6ac.png

[11:43:03.0000] <TabAtkins_>
miketaylr: No, the background is supposed to fade from white at the top to gray at the bottom.

[11:44:00.0000] <estellevw>
FF and webkit use differnt syntax

[11:44:01.0000] <TabAtkins_>
daedb: There's only one place where I might still change the syntax, and I doubt I will there.

[11:44:02.0000] <TabAtkins_>
(The one place is precisely how to handle angles.)

[11:44:03.0000] <estellevw>
IE can do gradients too, but completely different syntax than the spec

[11:44:04.0000] <daedb>
TabAtkins_: ok, thanks

[11:45:00.0000] <annevk>
JonathanNeal, the spec

[11:45:01.0000] <miketaylr>
maybe i'm missing something then, TabAtkins_. it looks the way i just pasted in chromium/mac, safari, ff 3.6, ff3.5, etc

[11:45:02.0000] <miketaylr>
oh wait. scratch that.

[11:45:03.0000] <miketaylr>
:/

[11:45:04.0000] <JonathanNeal>
/me scratches that

[11:45:05.0000] <TabAtkins_>
Does it work in one of the FF versions?

[11:46:00.0000] <miketaylr>
not in 3.5.0 or 3.5.8

[11:46:01.0000] <TabAtkins_>
Wtf.  I *know* that it was working a month ago when I redid my company's site template to use gradients.

[11:46:02.0000] <estellevw>
background: -moz-linear-gradient(top, yellow, red);

[11:46:03.0000] <estellevw>
background: -webkit-gradient(linear, 0 top, 0 bottom, color-stop(0, yellow), color-stop(1, red));

[11:47:00.0000] <TabAtkins_>
I could *see* it working, before I slotted in fallback images for other browsers.

[11:47:01.0000] <estellevw>
produce the same thing. 

[11:47:02.0000] <TabAtkins_>
roc, dbaron, anyone else from Moz?

[11:47:03.0000] <miketaylr>
it's the drugs, man.

[11:47:04.0000] <dbaron>
TabAtkins_, eh?

[11:47:05.0000] <TabAtkins_>
Any clue what happened to gradient support in FF?

[11:47:06.0000] <dbaron>
it's in 3.6

[11:47:07.0000] <dbaron>
not 3.5

[11:48:00.0000] <TabAtkins_>
3.6 is public, right?

[11:48:01.0000] <dbaron>
yes

[11:48:02.0000] <estellevw>
yeah

[11:48:03.0000] <TabAtkins_>
Argh, I'm all sorts of crazy, then.

[11:48:04.0000] <estellevw>
Moz is at 3.6.3 right now

[11:48:05.0000] <JonathanNeal>
It didn't work until 3.6.

[11:49:00.0000] <JonathanNeal>
They wouldn't let me put in the 3.6 css until 3.6 was out, had to use images.

[11:49:01.0000] <estellevw>
https://developer.mozilla.org/en/CSS/-moz-linear-gradient

[11:49:02.0000] <estellevw>
Firefox 3.6+ supported it

[11:49:03.0000] <TabAtkins_>
I was confused because my wife's comp was on 3.5, but that's just because she uses Chrome as her primary, and so hadn't gotten an auto-update in a while.

[11:49:04.0000] <TabAtkins_>
I figured she was on the latest.  Shrug.

[11:49:05.0000] <TabAtkins_>
/me is on 3.0 on this comp, unfortunately.

[11:49:06.0000] <dbaron>
(and, for the record (in response to "anyone else from Moz", I see 9 Mozilla employees currently on the channel, not counting those who are on twice... though roc and I are the ones who'd be most likely to know about gradients)

[11:50:00.0000] <estellevw>
http://standardista.com/sandbox/linear_gradient.html if interested

[11:50:01.0000] <JonathanNeal>
annevk or anyone, how are consolidation efforts going between the camps in regards to gradients?

[11:50:02.0000] <TabAtkins_>
dbaron: Heh, I totally don't know who all works in Moz.  But yeah, I figured you and roc were the most knowledgeable to ask anyway.

[11:50:03.0000] <dbaron>
(and I'm probably missing a few... er, make that 10)

[11:50:04.0000] <JonathanNeal>
http://css3please.com/ also has gradients for ff, saf, chrome, ie6+

[11:51:00.0000] <dbaron>
(never mind other community members)

[11:52:00.0000] <TabAtkins_>
/me heads back to his desk anyway, so he can work on getting his Windows machine to let him log in.

[11:52:01.0000] <annevk>
JonathanNeal, I think it's pretty much done now

[11:52:02.0000] <JonathanNeal>
So, which format won?  bluray or h...... moz of webkit?

[11:53:00.0000] <JonathanNeal>
s/of/or

[11:53:01.0000] <TabAtkins_>
Moz implements the spec version.  I created that version in response to webkit's version.

[11:56:00.0000] <JonathanNeal>
/me points!

[11:56:01.0000] <JonathanNeal>
IT WAS YOOOOOUERRRR FAULT

[12:26:00.0000] <TabAtkins_>
JonathanNeal: Yes, it was my fault.  Webkit's original syntax sucked.  ^_^

[12:27:00.0000] <JonathanNeal>
/me agrees with you too.

[12:28:00.0000] <JonathanNeal>
So, sounds like annevk needs to go back and add gradient support to Opera today.

[12:28:01.0000] <JonathanNeal>
Get a new version out Friday.

[12:29:00.0000] <JonathanNeal>
Meanwhile I'm going to invent wings for flying.

[12:31:00.0000] <AryehGregor>
I'm fairly sure wings were already invented some billions of years ago.

[12:31:01.0000] <AryehGregor>
It's conventional to focus effort on inventing things that do not, in fact, already exist.

[12:32:00.0000] <annevk>
TabAtkins_, when you say alpha before winter, does that mean about three to five months?

[12:33:00.0000] <annevk>
I guess more like five, but that seems a pretty long wait :/

[12:36:00.0000] <hober>
Despite the 'keephidden' URL, http://www.w3.org/html/wg/wiki/User:Eoconnor/keephidden is evolving into a zero-edit CP for several of the "remove this new feature" CPs that have come in in the last few weeks. Any and all contributions welcome.

[12:38:00.0000] <hober>
In particular, it'd be great to get the <figure>, <aside>, <details>, <progress>, and <meter> rationales fleshed out.

[12:38:01.0000] <TabAtkins>
annevk: alpha is expected out in August.

[12:38:02.0000] <hober>
And, by "fleshed out," I mean "written at all," since those sections are currently empty. :)

[12:38:03.0000] <TabAtkins>
It's my main project at work besides spec work.

[12:38:04.0000] <TabAtkins>
But I have to learn a *lot* of new things to do it.

[12:43:00.0000] <KaOSoFt>
http://www.thewhitehouse.gov/ uses RDFa :o

[12:43:01.0000] <AryehGregor>
TabAtkins, alpha of what?

[12:44:00.0000] <JonathanNeal>
RDFa, arrrrr

[12:47:00.0000] <KaOSoFt>
._.

[12:50:00.0000] <JonathanNeal>
/me waves the microformats banner that he just picked up a few hours ago.

[12:55:00.0000] <JonathanNeal>
Looks like yelp isn't using the address element correctly.  Time to dispatch the httPolice

[12:56:00.0000] <TabAtkins>
AryehGregor: The Cross-Activity Testing System, for automated handling of test-suites for various important web technologies across all browsers.

[12:56:01.0000] <AryehGregor>
Interesting.

[12:57:00.0000] <TabAtkins>
The alpha will just handle manually running tests, and reporting success/failure.

[13:01:00.0000] <miketaylr>
sounds neat.

[13:08:00.0000] <TabAtkins>
Well, in addition to all the framework.  It'll take until August because I have to build the entire framework first, from collection to reviewing to running to reporting.

[13:09:00.0000] <TabAtkins>
And I really don't know how to do most of that yet.

[13:19:00.0000] <jgraham>
TabAtkins: Interesting

[13:20:00.0000] <jgraham>
How does that compare to the W3C browser test harness thing?

[13:21:00.0000] <jgraham>
And why do you care about manual tests

[13:27:00.0000] <jgraham>
/me has a strong interest in test automation and that sort of thing

[13:36:00.0000] <KaOSoFt>
JonathanNeal- Are you the author of cms.txt?

[13:44:00.0000] <jgraham>
gsnedders: Got/heard the new Rufus Wainwright album?

[13:46:00.0000] <gsnedders>
jgraham: no

[13:46:01.0000] <gsnedders>
jgraham: I totally forgot about it since, uh, the middle of lst month

[13:47:00.0000] <gsnedders>
Oh, it's out here but not there

[13:47:01.0000] <gsnedders>
I should get it.

[13:47:02.0000] <gsnedders>
/me is rambling now

[13:50:00.0000] <estellevw>
In understanding  <section> versus <article> I think of it like the SUnday Paper. The newspaper had sections & within the sections there are articles. Those articles can have sections too. Is that a correct understanding?

[13:58:00.0000] <gsnedders>
/me wonders about the origins of thinking of life as a hand dealt by Fate

[14:00:00.0000] <gsnedders>
Also, "Would?" is an awesome song

[14:01:00.0000] <KaOSoFt>
estellevw- That's exactly my view of it so far.

[14:02:00.0000] <KaOSoFt>
I'm currently doing a HTML5 (and CSS3) template, and that's how I'm making it.

[14:02:01.0000] <KaOSoFt>
If there is a text with an explanation of something, preferable with a heading line, to me, that's an article.

[14:03:00.0000] <KaOSoFt>
/me looks up for "article" definition on dictionary.com

[14:14:00.0000] <jgraham>
gsnedders: "Would?"?

[14:17:00.0000] <KaOSoFt>
estellevw- This is the layout I'm currently starting with: http://pastebin.com/CuX6Qs4z

[14:17:01.0000] <KaOSoFt>
._.

[14:18:00.0000] <KaOSoFt>
estellevw- This is the layout I'm currently starting with: http://pastebin.com/CuX6Qs4z

[14:23:00.0000] <estellevw>
KaOSoft: I think you could do the same with less: unsure of the need for small in the footer or if encapsulating your logo in a <figure> is needed. Did you want to add more semantics to those two elements?

[14:25:00.0000] <KaOSoFt>
Well, I recently read about <small>, and it kind of mentioned to be used for license stuff. I don't remember where I read that, though. I think it was in http://html5doctor.com

[14:25:01.0000] <KaOSoFt>
I think it shouldn't exist, but oh well.

[14:25:02.0000] <KaOSoFt>
At least not with that name.

[15:15:00.0000] <gsnedders>
jgraham: Alice in Chains

[15:16:00.0000] <Philip`>
Is that the missing book in Lewis Carroll's trilogy?

[15:17:00.0000] <daedb>
Alice in Chains is awesome.

[15:18:00.0000] <gsnedders>
daedb: Totally

[15:19:00.0000] <gsnedders>
http://twitter.com/gsnedders/status/11781964799

[15:23:00.0000] <TabAtkins>
jgraham: Who manages the w3c browser test harness?  Is that plh?

[15:25:00.0000] <TabAtkins>
If so, then it's not directly related, but I may be looking at it for some inspiration.  fantasai and arronei and I spent all last Thursday talking about the design, and about similar efforts that preceded this.

[15:26:00.0000] <annevk>
yeah, you should contact plh

[15:26:01.0000] <annevk>
there's a testing IRC channel on irc.w3.org too, or some such

[15:26:02.0000] <TabAtkins>
Already planning on it.

[15:29:00.0000] <annevk>
would be nice if it was done already :)

[15:29:01.0000] <TabAtkins>
Too bad!

[15:30:00.0000] <annevk>
sort of interested in working on tests again, but I'm not going to before something is deployed

[15:30:01.0000] <TabAtkins>
Yeah, don't worry about it for now.

[15:41:00.0000] <TabAtkins>
Hehe, it's fun using a substantially different default font-color than most people, because then you can sometimes see exactly where an email was edited before being forwarded to you.

[15:41:01.0000] <TabAtkins>
(Since some email programs naturally change the font-color of text types in a forward/reply, and the sender changes that to their own font-color, rather than to 'automatic'.)

[15:42:00.0000] <murz>
lol, that is funny

[15:43:00.0000] <annevk>
first advantage ever for HTML-based email

[15:43:01.0000] <annevk>
/me is impressed

[15:43:02.0000] <TabAtkins>
For realz.

[15:43:03.0000] <gsnedders>
/me sighs at his own IM conversations

[15:44:00.0000] <gsnedders>
Alternating between quoting Romeo and Juliet, making both characters himself...

[15:44:01.0000] <TabAtkins>
Wherefore art thou, gsnedders?

[15:45:00.0000] <gsnedders>
Shall I hear more, or shall I speak at this?

[15:45:01.0000] <TabAtkins>
I'm not going any further as Juliet.

[15:46:00.0000] <gsnedders>
You only had the first line of the three she says there

[15:46:01.0000] <gsnedders>
And you missed the start of the line

[15:46:02.0000] <gsnedders>
/me checks that's right

[15:46:03.0000] <gsnedders>
Oh, she says four lines there

[15:46:04.0000] <TabAtkins>
/me doesn't actually know R&J, especially not at the level of accurate quoting.

[15:46:05.0000] <TabAtkins>
Were you at least communicating between two different IM contacts of yourself, so they came from different speakers?

[15:47:00.0000] <gsnedders>
/me ended up having a dicussion about the foreshadowing in R&J earlier, and hence can remember a lot of the quotes

[15:47:01.0000] <gsnedders>
No, both from the infamous gsnedders 

[15:47:02.0000] <annevk>
watch Curb Your Enthusiasm

[15:47:03.0000] <annevk>
much better use of time imo

[15:48:00.0000] <gsnedders>
Nah, I have shit vampire books to read

[15:48:01.0000] <gsnedders>
/me = teenage girl

[15:48:02.0000] <annevk>
oh well, at least I tried

[15:49:00.0000] <gsnedders>
Speaking of TV shows, I should finally get around to the second series of Californication

[15:50:00.0000] <annevk>
oh wow, your priorities are backwards

[15:50:01.0000] <annevk>
:p

[16:36:00.0000] <abarth_>
Hixie: is there a reason not to have the following work:

[16:36:01.0000] <abarth_>
<frameset> 

[16:36:02.0000] <abarth_>
<frame src="sf.html" sandbox> 

[16:36:03.0000] <abarth_>
</frame> 

[16:36:04.0000] <TabAtkins>
Other than "why are you caring about frames?"?

[16:36:05.0000] <abarth_>
apparently someone inside google wants to use @sandbox

[16:36:06.0000] <abarth_>
but they're using framesets

[16:36:07.0000] <abarth_>
instead of iframes

[16:37:00.0000] <TabAtkins>
Interesting.  Could you point me to them so I can help them not use framesets?

[16:37:01.0000] <abarth_>
haha

[16:37:02.0000] <abarth_>
ok

[16:38:00.0000] <abarth_>
done

[16:40:00.0000] <TabAtkins>
Ah yes, hmm.  I'll play with this for a sec and see what we can do about it.

[16:41:00.0000] <TabAtkins>
In any case, I don't see any particular reason it shouldn't be supported, other than that <frameset> and <frame> are obsolete elements.

[16:42:00.0000] <abarth_>
we're getting close to shipping @sandbox

[16:43:00.0000] <abarth_>
but we can add it to other kinds of frames later


2010-04-08
[17:24:00.0000] <JonathanNeal>
TabAtkins, you may find this interesting http://sandbox.thewikies.com/html5-wai/ it attempts to hit many birds with one stone; html5, outline, wai, 508, microformat.

[17:28:00.0000] <TabAtkins>
JonathanNeal: Why are you using <div>s for the main content rather than <article>?

[17:28:01.0000] <TabAtkins>
Also, role=document is assumed automatically on <body>, so there's no reason to put a wrapper <div> just to get that.

[17:29:00.0000] <JonathanNeal>
It's there in the event I add an application aside of it.

[17:29:01.0000] <TabAtkins>
?_?

[17:29:02.0000] <JonathanNeal>
say a dockbar on the top, a chatbar on the side, or some other role="application" aside the main document but on the same page.

[17:30:00.0000] <TabAtkins>
Regardless, the <body> is still a role=document, and I don't think it needs to be repeated on the main page.

[17:30:01.0000] <TabAtkins>
I assume that anything which isn't explicitly wrapped is considered part of the body's document.

[17:32:00.0000] <JonathanNeal>
If that was the case, then there would never be a reason to use more than one role per page.

[17:32:01.0000] <JonathanNeal>
clarify --- role="document" || role="application"

[17:33:00.0000] <TabAtkins>
I'm not sure what you mean.  You just gave a potential reason to use more than one role.  You're embedding applications inside a document.

[17:33:01.0000] <JonathanNeal>
I can see removing it because it is inherited from body (since body is not overwritten)

[17:34:00.0000] <TabAtkins>
Btw, the lynx viewer link you have gets angry if it sees a referer from another page.

[17:34:01.0000] <JonathanNeal>
<body><div role="application>dock application</div><div role="document">main document</div><div role="application>chat application</div></body>

[17:34:02.0000] <JonathanNeal>
I know, I go there and then just re-highlight the addy and return.

[17:34:03.0000] <TabAtkins>
Kk, just saying.

[17:34:04.0000] <JonathanNeal>
So ... you're saying that it's unnecessary because in that case ...

[17:34:05.0000] <JonathanNeal>
<body><div role="application>dock application</div><div>main document</div><div role="application>chat application</div></body> ... is just the same

[17:35:00.0000] <TabAtkins>
Yeah.

[17:35:01.0000] <TabAtkins>
I think.

[17:35:02.0000] <TabAtkins>
/me isn't an expert on ARIA, but that seems to make sense.

[17:35:03.0000] <JonathanNeal>
Yeah, that makes sense to me too, I think you're right.

[17:44:00.0000] <KaOSoFt>
I'm just a novice but, why <div id="main-content"> and not <section id="main-content">? I'm coding an HTML5 template right now, and I use a <div id="wrap">, but I also (inside the "#wrap") use a <section id="main-content">.

[17:45:00.0000] <TabAtkins>
I don't see a particularly good reason for either of those <div>s.  The whole <div><div><section> structure should be swappable for a single <article> or <section>.

[17:47:00.0000] <KaOSoFt>
Well, I read somewhere (unfortunately don't remember where) that now with <section> and <article>, <div> had no semantic meaning, that's why I use it as a meaningless container.

[17:48:00.0000] <KaOSoFt>
Well, while <header> and <footer> should make things "easier", I also have a point of view where they are still <section>s, which means, that I'd do it: <section id="header">, <section id="main-content"> and <section id="footer">, for example.

[17:48:01.0000] <TabAtkins>
<div> Never had semantics meaning, actually.  It was always purely there for the purpose of grouping things to make it easier to throw CSS and JS at the page.

[17:48:02.0000] <TabAtkins>
Same with <span>.

[17:50:00.0000] <TabAtkins>
The idea of <section>, <article>, <aside>, etc is that some of those groupings *were* semantics, and commonly expressed by authors (often through #ids on the elements), so we can usefully capture those semantics and allow other technologies to use them.

[17:54:00.0000] <JonathanNeal>
For me, having a <section id="main-content"> would not be so easy to title.

[17:54:01.0000] <KaOSoFt>
Like I said, elements like <header> and <footer> are just short-cuts to me. They are indeed <section>s. Oh, well, I could just stay with my idea, and use <section id="header"> and the like, but future-proof machines wouldn't understand my web site as well.

[17:54:02.0000] <JonathanNeal>
and something like <section id="wrap"> would default the sectioning purpose of <body>

[17:55:00.0000] <JonathanNeal>
I think there is still room for divs to be organizational for the content, which you could say is semantic.

[17:55:01.0000] <TabAtkins>
<header> isn't a section, actually, and you don't want it to be - it would scope the headings you put in it to itself, and not let them refer to the outer section you put the <header> in!

[17:55:02.0000] <JonathanNeal>
Since the requirements and recommendations for sectioning are too strict on <section|article|aside|etc>

[17:56:00.0000] <KaOSoFt>
TabAtkins- Hmm, yeah, I'm checking the code on your test-site, and it indeed doesn't use <header> as the header of the document. It uses a <div>, and you use <header> to encompass the headings.

[17:56:01.0000] <TabAtkins>
That's JonathanNeal.  ^_^

[17:56:02.0000] <KaOSoFt>
Er, yeah, that.

[17:56:03.0000] <KaOSoFt>
:P

[17:56:04.0000] <JonathanNeal>
KaOSoFt, yea that's my test site.  TabAtkins may not want to be associated with that little nugget of joy.

[17:57:00.0000] <KaOSoFt>
Hahaha.

[17:57:01.0000] <KaOSoFt>
God, semantics sometimes can be deceitful.

[17:57:02.0000] <TabAtkins>
You can use <header> as the header of the site!  Nothing wrong with that; in fact, plenty right with it.  It's just not a "section", in the way that <section> and <article> and <aside> are.

[17:59:00.0000] <TabAtkins>
Precisely because "sections" scope the headings inside of them, and you *want* headings inside of <header> to apply to the content outside of <header>.

[17:59:01.0000] <KaOSoFt>
Where do you take the role from? Is it part of HTML5 spec?

[17:59:02.0000] <KaOSoFt>
"role="

[17:59:03.0000] <JonathanNeal>
<div id="main-document" /> extends body (the scope isn't changed by sectioning content), and there are instances where I will have other applications aside it, like I mentioned earlier.  The benefit is that, even in instances where there are NOT applications aside of it, I can keep the div because it doesn't have any meaning.

[17:59:04.0000] <TabAtkins>
Officially from the ARIA spec, but HTML5 defines how you can use @role on HTML elements.

[18:00:00.0000] <KaOSoFt>
I see.

[18:00:01.0000] <KaOSoFt>
/me goes to check the ARIA spec

[18:01:00.0000] <JonathanNeal>
So then it's useful for styling.  Giving it the role was to enforce its intended usage, but without the other DIVs it doesn't seem as necessary, and as TabAtkins is pointing out, the document role may already be inherited.

[18:06:00.0000] <TabAtkins>
KaOSoFt: Basically, ARIA is a way to add additional semantics to documents in ways that the document language doesn't yet allow, so that other types of UAs, like screen readers, can understand what something is.  For example, javascript libraries often let you define a slider control, and build it out of several <div>s.  You can tell what this is and how to use it by looking at it, but if you were having the page read to you you'd have n

[18:07:00.0000] <TabAtkins>
A lot of HTML5 elements now carry native semantics that render a lot of ARIA unnecessary, which is a good thing.  Frex, <input type=range> should replace many uses for sliders.

[18:28:00.0000] <KaOSoFt>
So many things to synchronize. :S

[18:29:00.0000] <KaOSoFt>
Everytime I read about a specification, it makes me wonder how long will it take to really release HTML5.

[18:31:00.0000] <KaOSoFt>
I mean:

[18:31:01.0000] <KaOSoFt>
It's like, there are so many things that HTML5 has currently addressed, that every time I see something new (to me), I wonder how long will it take to make it work together, all synchronized.

[18:34:00.0000] <Hixie>
it's probably a bad sign that having decided i wasn't sure if "punt" meant what i thought it meant, i looked it up on wikitionary, and the only example usage of the word in the sense i meant it is quoting two people i work with.

[18:35:00.0000] <KaOSoFt>
._.

[18:36:00.0000] <KaOSoFt>
Oh, by the way, I've had this inquiry for a while, and I think this channel would be the right place to ask:

[18:37:00.0000] <gsnedders>
Hixie: Which meaning?

[18:38:00.0000] <KaOSoFt>
I've always read that we should specify the image height and width in the <img> tag, so browsers render them faster, but is it really "okay"? I mean, if for some reason I change the image, which could possibly change its size, I'd have to also change the HTML file.

[18:38:01.0000] <TabAtkins>
There's nothing wrong with either way.

[18:48:00.0000] <Hixie>
gsnedders: not doing it straight away

[18:48:01.0000] <Hixie>
as in, i think we should punt on doing this for now

[18:49:00.0000] <KaOSoFt>
Oh, well, time to go home.

[18:49:01.0000] <KaOSoFt>
Goodnight.

[19:36:00.0000] <yael_>
hello, can I ask a ARIA related question here?

[19:37:00.0000] <davidb>
yael: sure, there is also #wai-aria

[19:38:00.0000] <davidb>
oh sorry that is on w3c

[19:41:00.0000] <yael_>
I am workin on connecting the progress element to accessibility technology. Since ARIA defines one role for both determinate and indeterminate progress bar, how does the browser convey the difference to the accessibility software?

[19:43:00.0000] <yael_>
On the mac, it says "9 percent" for indeterminate progress bar, and that is obviously wrong.

[19:43:01.0000] <Hixie>
if you're implementing the <progress> element in a browser, you wouldn't use ARIA, you'd use the platform-native accessibility APIs

[19:43:02.0000] <Hixie>
or am i misunderstanding

[19:45:00.0000] <yael_>
You are right, but I am still defining the role for the progress element , and pass that to the platform API.

[19:45:01.0000] <othermaciej>
yael_: if you're talking about WebKit, you would want to special case this in the accessibility code

[19:45:02.0000] <othermaciej>
yael_: we have the ability to represent roles that don't map to an ARIA role

[19:46:00.0000] <yael_>
othermaciej: Do you have an example?

[19:47:00.0000] <othermaciej>
yael_: look at the AccessibilityRole enum in AccessibilityObject.h

[19:47:01.0000] <othermaciej>
yael_: you can add roles to that if you need to

[19:47:02.0000] <othermaciej>
but it's also worth checking what the mac AX APIs expect for an indeterminate progress bar

[19:48:00.0000] <yael_>
sounds good, thanks

[19:48:01.0000] <othermaciej>
yael_: Mac has only one role for progress indicators; you can see the reference for all roles here: http://developer.apple.com/mac/library/documentation/Cocoa/Reference/ApplicationKit/Protocols/NSAccessibility_Protocol/Reference/Reference.html#//apple_ref/doc/c_ref/

[19:49:00.0000] <othermaciej>
yael_: what you probably need is to expose the right attributes

[19:50:00.0000] <othermaciej>
yael_: you can see what a native progress bar exposes for role and attributes with Accessibility Verifier.app

[19:50:01.0000] <othermaciej>
yael_: for both determinate and indeterminate state

[19:50:02.0000] <othermaciej>
yael_: your goal at least in the Mac accessibility interface should be to replicate that

[20:45:00.0000] <JonathanNeal>
Hi all!

[23:12:00.0000] <othermaciej>
hello everyone

[23:15:00.0000] <JonathanNeal>
hi othermaciej

[23:51:00.0000] <othermaciej>
Hixie: are you around?

[23:52:00.0000] <othermaciej>
Hixie: wanted to ask why you went with leading underscore instead of leading hyphen for vendor-specific attributes

[23:52:01.0000] <othermaciej>
Hixie: so I can appropriately advise people thinking of using vendor-specific attributes

[00:19:00.0000] <Hixie>
othermaciej: leading hyphen is illformed in xml

[00:20:00.0000] <othermaciej>
Hixie: stupid bondange&discipline markup language!

[00:21:00.0000] <Hixie>
*shrug*

[00:21:01.0000] <Hixie>
i just work with what we have

[00:23:00.0000] <hsivonen>
:-( crazy volume of email on public-html. Will I miss anything if I mark all as read?

[00:23:01.0000] <othermaciej>
what's weird is that NAmespaces in XML allows essentially anything

[00:23:02.0000] <othermaciej>
hsivonen: which thread?

[00:24:00.0000] <hsivonen>
othermaciej: everything since Friday

[00:25:00.0000] <othermaciej>
hsivonen: I can summarize for you, since I am crazy enough to read all of it

[00:26:00.0000] <hsivonen>
Shelley wanted to remove stuff and others disagreed?

[00:26:01.0000] <othermaciej>
Hixie: superficially, it looks like Namespaces in XML allows leading underscore unless I am reading wrong: 'In XML documents which conform to this specification, element and attribute names MUST match the production for QName and MUST satisfy the "Namespace Constraints".'

[00:26:02.0000] <othermaciej>
Hixie: or is the prevailing interpretation that you have to match both QName and Name?

[00:26:03.0000] <othermaciej>
hsivonen: that was a lot of it

[00:27:00.0000] <othermaciej>
hsivonen: with subthreads on accessibility and control styling

[00:27:01.0000] <Hixie>
xmlns doesn't change xml syntax

[00:27:02.0000] <Hixie>
it just adds more constraints

[00:27:03.0000] <othermaciej>
I thought the post about control styling that I tweeted was marginally interesting

[00:27:04.0000] <Hixie>
unless you've found a bug in xmlns

[00:28:00.0000] <hsivonen>
othermaciej: I read the email you tweeted. thanks

[00:28:01.0000] <Hixie>
QName doesn't match anything Name doesn't match

[00:28:02.0000] <othermaciej>
Hixie: I am pretty sure the QName production in Namespaces in XML 1.0 (Third Edition) is not a proper subset of the Name production in XML 1.0 (Fifth Edition

[00:29:00.0000] <Hixie>
sure it is

[00:29:01.0000] <Hixie>
it's defined in terms of Name in fact

[00:29:02.0000] <othermaciej>
oh I see

[00:29:03.0000] <othermaciej>
I failed to comprehend the '-'

[00:30:00.0000] <Hixie>
there's even a comment explaining it :-P

[00:30:01.0000] <othermaciej>
yeah, but who has time to read specs :-p

[00:30:02.0000] <othermaciej>
I don't know if I can bring myself to tell people to make attributes starting with _webkit-

[00:30:03.0000] <othermaciej>
if I tell them it's for XML compatibility they will just laugh at me

[00:31:00.0000] <Hixie>
well if it helps, mozilla's been doing it for about a decade now

[00:31:01.0000] <othermaciej>
hsivonen: other stuff of interest - the lengthy ISSUE-88 (content-language) continues

[00:31:02.0000] <Hixie>
and fwiw, if this discourages people from making extensions, that's fine too

[00:31:03.0000] <othermaciej>
hsivonen: we have a possible solution on referencing the IRI spec if Larry ever wakes up

[00:31:04.0000] <Hixie>
(not so fine if it encourages them to just trample over the language of course)

[00:31:05.0000] <othermaciej>
Hixie: there are _moz- attributes?

[00:32:00.0000] <Hixie>
yeah

[00:32:01.0000] <othermaciej>
example?

[00:32:02.0000] <othermaciej>
hsivonen: also DanC submitted a very mild proposal for the "resource" vs "resource representation" issue

[00:32:03.0000] <othermaciej>
and Julian said it would be fine with him to remove the description of @profile from the spec

[00:33:00.0000] <hsivonen>
In my opinion, _moz attributes shouldn't be exposes as attributes but should be stored on the DOM nodes or maybe CSS frames as private data

[00:33:01.0000] <othermaciej>
/me hopes we can actually resolve some issues maybe

[00:33:02.0000] <hsivonen>
othermaciej: thanks

[00:34:00.0000] <hsivonen>
othermaciej: try parsin <keygen> in text/html or some non-trivial MathML in application/xhtml+xml in Firefox and inspect the DOM to see _moz attributes

[00:35:00.0000] <othermaciej>
hsivonen: I see - so it's Mozilla internal stuff?

[00:35:01.0000] <othermaciej>
as opposed to extensions anyone is meant to see?

[00:36:00.0000] <hsivonen>
othermaciej: right (at least for the _moz attributes I have seen)

[00:36:01.0000] <othermaciej>
I am not sure we store anything like that as DOM attributes in WebKit

[00:36:02.0000] <othermaciej>
maybe in editing code, though I think that is mostly limited to funky class names

[00:37:00.0000] <othermaciej>
yeah, just funky class names afaict

[00:37:01.0000] <Hixie>
othermaciej: the only example i can think of is _moz-unknown which is set on elements the (old?) parser didn't recognise

[00:38:00.0000] <Hixie>
oh yeah as hsivonen says there's also some _moz stuff around <keygen>

[00:38:01.0000] <othermaciej>
the parser adding random attributes seems really bad, however they are prefixed :-(

[00:38:02.0000] <hsivonen>
yes

[00:38:03.0000] <Hixie>
indeed

[00:38:04.0000] <Hixie>
i didn't say it was a good idea :-)

[00:38:05.0000] <Hixie>
just that they have existed

[00:39:00.0000] <othermaciej>
I get it

[00:39:01.0000] <othermaciej>
I just cry with sad on the inside

[00:39:02.0000] <Hixie>
hehe

[01:57:00.0000] <jgraham>
TabAtkins: If there is some way that I can be in the loop on the testing thing, let me know what it is :)

[02:05:00.0000] <hsivonen>
wirepair: Yes, I have considered the threat of using validator.nu with show source enabled as an open proxy

[02:06:00.0000] <hsivonen>
wirepair: a mitigating factor is that the resource being retrieved has to go through the HTML parser or the XML parser without fatal errors

[02:07:00.0000] <Hixie>
hah

[02:07:01.0000] <hsivonen>
wirepair: so it's useless as a general purpose proxy for anonymous porn surfers

[02:07:02.0000] <zcorpan>
so if you don't want your site vulnerable, include start your html page with <head></head><link>

[02:08:00.0000] <hsivonen>
wirepair: so far, there haven't been problems with people using it as a text-only open proxy

[02:08:01.0000] <hsivonen>
wirepair: so I haven't tried to make that kind of use harder

[02:09:00.0000] <hsivonen>
wirepair: after all, the show source feature is pretty useful for its intended purpose

[02:10:00.0000] <hsivonen>
wirepair: as for 3rd party security, if someone has a vulnerability that can be exploited by cookieless GET from the public network, they have bigger problems than validators

[02:37:00.0000] <hsivonen>
what was the "subdued errors concept"

[02:37:01.0000] <hsivonen>
my memory is bad

[02:38:00.0000] <Hixie>
what we had before we moved to comforming-but-obsolete-and-with-warning

[02:39:00.0000] <Hixie>
something that's an error but which the validator is expected to collapse into a single statement that "there are also some other less important things to worry about"

[02:41:00.0000] <othermaciej>
it was "downplayed errors", wasn't it?

[02:42:00.0000] <hsivonen>
yeah, I thought that was "downplayed errors"

[02:45:00.0000] <othermaciej>
whatever one names that category, I think collapsing all the notices into a single statement, yet still labeling the document as nonconforming, is the opposite of what is useful for very minor issues

[02:51:00.0000] <Hixie>
aren't "downplayed" and "subdued" the same thing?

[02:51:01.0000] <jgraham>
Sigh. The amount of effort that has gone into naming and renaming that categoty of things is so sad

[02:51:02.0000] <Hixie>
if i said "subdued", i was referring to "downlplayed"; if someone else was, dunno

[02:51:03.0000] <jgraham>
*category

[03:21:00.0000] <hsivonen>
Hixie: I see

[04:36:00.0000] <annevk>
oh fun, maybe <basefont> can be nuked after all

[04:49:00.0000] <zcorpan>
nuked from dom and parser?

[04:49:01.0000] <annevk>
not parser

[04:50:00.0000] <annevk>
since the parser cost is extremely low I don't really care about that

[04:55:00.0000] <zcorpan>
what's the new information about basefont?

[04:58:00.0000] <Lachy>
Which browser's support basefont? Is it just IE?

[04:59:00.0000] <Lachy>
and do you have information about it being removed from IE9 or something?

[04:59:01.0000] <jgraham>
I guess Mozilla just removed it

[04:59:02.0000] <annevk>
zcorpan, sicking wrote a patch for Gecko

[05:00:00.0000] <Lachy>
ah, that explains why Minefield isn't supporting it now.

[05:02:00.0000] <annevk>
not really, it hasn't landed

[05:02:01.0000] <zcorpan>
i wonder what ie9 does with basefont

[05:02:02.0000] <zcorpan>
Lachy: mozilla hasn't supported rendering for basefont before

[05:03:00.0000] <zcorpan>
Lachy: but has had HTMLBaseFontElement or whatever

[06:05:00.0000] <hsivonen>
Lachy: IIRC, no one got around to implementing <basefont> in Gecko in the first place

[06:28:00.0000] <annevk>
would be fun to transition text-align

[07:00:00.0000] <gsnedders>
Is it bad I can't help but think of XML when othermaciej says "this change proposal is well-formed"

[07:01:00.0000] <Dashiva>
Yes

[07:01:01.0000] <Dashiva>
So is a non-well-formed change proposal still a change proposal, or is it just mislabeled text?

[07:01:02.0000] <jgraham>
Nah, it just doesn't exist, by definition

[07:02:00.0000] <jgraham>
A theoretical impossibility

[07:35:00.0000] <zcorpan>
where's the webaddress spec?

[07:35:01.0000] <zcorpan>
/me needs the 'resolve an address' algorithm

[07:36:00.0000] <gsnedders>
You don't get no stinkin' address.

[07:36:01.0000] <zcorpan>
:'(

[07:36:02.0000] <jgraham>
/me wonders if that algorithm has actually been written yet

[07:36:03.0000] <zcorpan>
i thought it was in the webaddress spec

[07:37:00.0000] <zcorpan>
http://www.w3.org/html/wg/href/draft

[07:37:01.0000] <jgraham>
I mean it *did* exist but then people decided it wasn't good enough just to have all the useful information, there also had to be fights about where the useful information lived

[07:37:02.0000] <jgraham>
and some information got lost in the fights

[07:37:03.0000] <jgraham>
But maybe not that

[09:15:00.0000] <TabAtkins>
jgraham: Join the public-test-infra list.  We're stealing it for our purposes.

[09:16:00.0000] <gsnedders>
TabAtkins: People and their mailing lists

[09:16:01.0000] <jgraham>
Yay, mailing lists!

[10:24:00.0000] <JonathanNeal>
hi all!

[10:24:01.0000] <TabAtkins>
yo

[10:25:00.0000] <TabAtkins>
Oh man, anyone in MV that wants brownies should come by my office.

[10:26:00.0000] <JonathanNeal>
Mission Viejo?

[10:27:00.0000] <TabAtkins>
Sure.  It'll be a drive, but if you're still welcome to come by my office.

[10:36:00.0000] <KaOSoFt>
Good morning.

[10:49:00.0000] <KaOSoFt>
:o

[10:49:01.0000] <KaOSoFt>
Google already supports Microdata.

[11:01:00.0000] <KaOSoFt>
In HTML5, can <img> be used outside of <figure>? Like: http://pastebin.com/BzcDZqct

[11:01:01.0000] <TabAtkins>
Yeah, you can still use <img> wherever you want.

[11:26:00.0000] <KaOSoFt>
While what I'm about to say might be a taboo, have you people made a poll with the users on this channel about what they're currently using, be it RDFa or Microdata?

[11:26:01.0000] <TabAtkins>
Is it important to do so?

[11:27:00.0000] <KaOSoFt>
Statistics are always important, whether you take them into account or not.

[11:27:01.0000] <TabAtkins>
Statistics of a random chat room probably aren't.  ^_^  Usage on the web is.

[11:28:00.0000] <KaOSoFt>
Well, the poll would be about people using it.

[11:28:01.0000] <KaOSoFt>
._.

[11:30:00.0000] <JonathanNeal>
KaOSoFt, you didn't even mention microformatting.

[11:30:01.0000] <JonathanNeal>
That's like not including Google :)

[11:30:02.0000] <TabAtkins>
Microformats are already part of HTML, though.  We're talking *dedicated* metadata embedding formats.

[11:30:03.0000] <KaOSoFt>
Well, to be "impartial", let's include Microformatting, though I don't like it that much.

[11:30:04.0000] <JonathanNeal>
TabAtkins, well maybe everyone is okay with that.

[11:31:00.0000] <KaOSoFt>
Besides, I guess usage shows a bigger RDFa user-base than Microdata, given their age.

[11:32:00.0000] <JonathanNeal>
Say 1000 people use one of those three; 300 use microdata, 200 use RDFa, and 500 use microformatting --- disregarding microformatting is strange --- because they're all trying to accomplish the same functionality.

[11:32:01.0000] <JonathanNeal>
It's not like comparing people that don't do anything to people that do something --- you should try to include all popular "something"s.

[11:33:00.0000] <KaOSoFt>
Hmm... though it's an example, you mention at least 300 people using Microdata. Has user-base really grown that much?

[11:33:01.0000] <KaOSoFt>
I mean, in general.

[11:33:02.0000] <JonathanNeal>
The internet is SERIOUS BUSINESS KaOSoFt!

[11:37:00.0000] <TabAtkins>
In any case, Microformats are not defined by their embedding.  They're a structure to apply to data.  They happened to use an embedding method built around existing HTML4 syntax, because that's what was available at the time, but writing an hcard in Microdata or RDFa is still using Microformats.

[11:37:01.0000] <KaOSoFt>
I guess in the end it's better to have multiple "supported" options, than to have one unsupported.

[11:38:00.0000] <KaOSoFt>
That's kind of why I don't like Microformats.

[11:42:00.0000] <JonathanNeal>
It's kinda what the class attribute was designed to do.

[11:43:00.0000] <JonathanNeal>
Not that I'm against Microdata or RDFa, even though they kill unicorns.

[11:43:01.0000] <TabAtkins>
Well, yes and no.  Class isn't necessarily supposed to be communicating anything meaningful to the outside world.  It's meant for private communication between the page author and their tools.

[11:44:00.0000] <JonathanNeal>
RDFa has nicer documention and adoption than Microdata IMHO.

[11:46:00.0000] <TabAtkins>
While I, on the other hand, find Microdata's overall syntax and data model to be immensely simpler and easier to write and read.

[11:48:00.0000] <JonathanNeal>
TabAtkins, I bet I would feel the same way if the documentation wasn't such a unrainbow.

[12:11:00.0000] <KaOSoFt>
Oh, damn, wait... if not even Internet Explorer recognizes HTML5 elements (<header>, <section>, etc.), how can I make print stylesheets work? :o

[12:12:00.0000] <TabAtkins>
The IE printing problem is a hard one.  If print stylesheets working in IE is important to you, don't use HTML5 yet.

[12:13:00.0000] <KaOSoFt>
Internet Explorer 8*

[12:14:00.0000] <KaOSoFt>
Yeah. It's not a problem for MY website, but if I were to ever code a website for someone else, I guess I'd have to switch back to <div>.

[12:14:01.0000] <TabAtkins>
I think that most people don't actually care how websites print, so you're probably okay.

[12:15:00.0000] <TabAtkins>
IE can handle the new elements just fine on the screen, with the createElement hack.

[12:16:00.0000] <KaOSoFt>
Yeah.

[12:16:01.0000] <KaOSoFt>
HTML5Shiv was a life-saver for a lot of people.

[12:16:02.0000] <JonathanNeal>
Shiv the printer!

[12:16:03.0000] <TabAtkins>
AFAIK, no one's figured out how to.  ;_;

[12:16:04.0000] <JonathanNeal>
http://sandbox.thewikies.com/html5-experiments/phase-01.html http://sandbox.thewikies.com/html5-experiments/phase-02.html

[12:18:00.0000] <JonathanNeal>
Any blogs giving an example of how the html5 ie printing problem works?

[12:18:01.0000] <JonathanNeal>
Maybe I can look into it.

[12:20:00.0000] <KaOSoFt>
JonathanNeal- None that I know of. :'(

[12:21:00.0000] <KaOSoFt>
JonathanNeal- I'm going to (try to) check all your SandBox examples. http://sandbox.thewikies.com/

[12:21:01.0000] <KaOSoFt>
:D

[12:21:02.0000] <JonathanNeal>
Don't judge me, brah.

[12:27:00.0000] <KaOSoFt>
JonathanNeal- Phase 02 images don't work.

[12:27:01.0000] <KaOSoFt>
In Experiments, I mean.

[12:29:00.0000] <KaOSoFt>
Is this alright?

[12:29:01.0000] <KaOSoFt>
article, aside, figure, footer, header,hgroup, menu, nav, section {display: block;}

[12:29:02.0000] <KaOSoFt>
Am I missing anything?

[12:30:00.0000] <TabAtkins>
Check JonathanNeal's HTML default stylesheet.

[12:33:00.0000] <KaOSoFt>
Hmm... where? I don't find it in the sandbox folder, not in the document's own CSS.

[12:35:00.0000] <KaOSoFt>
nor*

[12:35:01.0000] <TabAtkins>
I didn't bookmark it last time I saw it; ask JonathanNeal 

[12:36:00.0000] <JonathanNeal>
You mean http://www.iecss.com/whatwg.css ?

[12:36:01.0000] <TabAtkins>
Yeah

[12:37:00.0000] <KaOSoFt>
Hmm, it's a reset stylesheet. Good, let's check it out.

[12:37:01.0000] <KaOSoFt>
Thanks.

[12:40:00.0000] <KaOSoFt>
It specifies everything.

[12:40:01.0000] <KaOSoFt>
-_-

[12:41:00.0000] <TabAtkins>
Yes, it's meant to be an example of what an actual UA stylesheet for HTML5 could look like.

[12:41:01.0000] <KaOSoFt>
UA means a browser?

[12:42:00.0000] <TabAtkins>
Yeah.  User Agent.

[12:42:01.0000] <KaOSoFt>
User Agent, oh, yes.

[12:42:02.0000] <TabAtkins>
In this context, browser, but it can mean lots of things.

[12:42:03.0000] <KaOSoFt>
Yeah, I know.

[12:42:04.0000] <miketaylr>
should that include stuff like :valid, :invalid, :out-of-range, etc?

[12:43:00.0000] <miketaylr>
or would that be handled elsewhere?

[12:43:01.0000] <TabAtkins>
HTML5 doesn't specify any default renderings for those.

[12:43:02.0000] <miketaylr>
/me doesn't know

[12:43:03.0000] <miketaylr>
ah.

[12:43:04.0000] <KaOSoFt>
Hmm... and while this should provide a good example, is that CSS publicized so that UA developers know of it?

[12:43:05.0000] <miketaylr>
yeah, gotcha TabAtkins

[12:43:06.0000] <TabAtkins>
KaOSoFt: It's just a distillation of what HTML5 says, in a perhaps-more-understandable and centralized format.

[13:13:00.0000] <dglazkov>
TabAtkins has the best brownies

[13:28:00.0000] <JonathanNeal>
I found a way to work around the HTML5 printing bug in IE

[13:28:01.0000] <TabAtkins>
paul_irish: ^^^

[13:36:00.0000] <miketaylr>
JonathanNeal: how!?!??!

[13:36:01.0000] <TabAtkins>
Gah, I knew I wanted to tell someone else, but I forgot who.  Sorry, miketaylr.

[13:36:02.0000] <miketaylr>
haha no worries :)

[13:36:03.0000] <miketaylr>
all of my css expression hack attempts failed

[13:38:00.0000] <KaOSoFt>
JonathanNeal- Link, or it doesn't exist...!

[13:38:01.0000] <KaOSoFt>
._.

[13:45:00.0000] <JonathanNeal>
Okay.

[13:46:00.0000] <miketaylr>
JonathanNeal: i'm super curious :)

[13:48:00.0000] <KaOSoFt>
http://blog.whatwg.org/usability-testing-html5

[13:48:01.0000] <KaOSoFt>
Interesting conclusions.

[14:00:00.0000] <paul_irish>
JonathanNeal: is it using behavior: ?

[14:08:00.0000] <JonathanNeal>
no

[14:10:00.0000] <Hixie>
KaOSoFt: microdata is very new, i wouldn't expect many people to be using it today

[14:11:00.0000] <Hixie>
KaOSoFt: however, it has the support of some pretty big companies, e.g. google added support for it recently

[14:11:01.0000] <TabAtkins>
I just now wrote a quick tutorial on Microdata this is hopefully informative and accessible.  Review comments welcome: http://www.xanthir.com/document/document.php?id=a3c2c6bb66f089ef0cec212c7409c83b7f4ab3af217f1cc4bb664a135a7d172c364c625be73b7da5750fd01bed1ea739cb7fa61de8fc7f46d01334df1164d70e

[14:12:00.0000] <TabAtkins>
Or, less ridiculous, http://www.xanthir.com/:gg5

[14:20:00.0000] <KaOSoFt>
JonathanNeal- HTML5 + Internet Explorer = Print? :'(

[14:21:00.0000] <paul_irish>
he's got a solution. sounds really promising. it's coming... :)

[14:21:01.0000] <KaOSoFt>
I need to set up a blog. That solution will be my first entry on it.

[14:21:02.0000] <KaOSoFt>
^_^

[14:22:00.0000] <paul_irish>
do it!

[14:26:00.0000] <KaOSoFt>
Arggh, I guess WordPress is the fastest solution.

[14:28:00.0000] <JonathanNeal>
thanks paul_irish, I'll have something up today.

[14:29:00.0000] <KaOSoFt>
My first blog entry will make me rich...!

[14:29:01.0000] <KaOSoFt>
:D

[14:37:00.0000] <AryehGregor>
TabAtkins, is the author of document.php a big fan of GUIDs and thinks it's easier to use SHA512 than figure out MAC addresses or such?  Or maybe they just want really secure content-addressable storage?

[14:48:00.0000] <TabAtkins_>
AryehGregor: Whirlpool hash, actually.

[14:49:00.0000] <TabAtkins_>
And I'm the author.  Woo!

[14:52:00.0000] <AryehGregor>
TabAtkins_, so why are you using 512-bit hashes for id's?

[14:53:00.0000] <Hixie>
he's VERY prolific

[14:53:01.0000] <jgraham>
Y'know I was just about to ask that

[14:54:00.0000] <jgraham>
hint: if it's longer than the hash you get when you check it into your VCS, it's too long :)

[14:54:01.0000] <TabAtkins_>
Because shrug, that's why.

[14:54:02.0000] <Hixie>
to use up the entire address space over 100 years of blogging, you'd have to blog at a rate of 4.2 x 10^144 Hz

[14:55:00.0000] <AryehGregor>
jgraham, well, git uses SHA1, which isn't advisable if you want actual security . . .

[14:55:01.0000] <Hixie>
that is an INSANE frequency

[14:55:02.0000] <AryehGregor>
Hixie, but only the square root of that to have a high chance of collision, right?  That's much more reasonable.

[14:56:00.0000] <Hixie>
um, even to avoid a collision, if you were to have that frequency as EM waves, it'd still be 40 orders of magnitude narrower wavelength than the width of an atom

[14:57:00.0000] <AryehGregor>
Well, so you can just write a lot of blog posts in parallel.

[14:57:01.0000] <AryehGregor>
A million monkeys and all.  Or 10^60 monkeys, whatever it takes.

[14:57:02.0000] <TabAtkins_>
Dude, I coded my document writer in like 10 minutes.  I had just used whirlpool for something else, because it's fun, so it was on my mind.

[14:57:03.0000] <AryehGregor>
(I'm not sure there's enough mass in the observable universe to create 10^60 monkeys.)

[14:57:04.0000] <TabAtkins_>
They'd have to be very small monkey, but you could.

[14:58:00.0000] <Hixie>
very small

[14:58:01.0000] <Hixie>
there's only ~10^80 atoms in the visible universe

[14:58:02.0000] <Hixie>
and most of that is hydrogen

[14:58:03.0000] <Hixie>
not so useful for making monkeys

[14:59:00.0000] <Philip`>
http://www.guzer.com/pictures/very-small-monkey.jpg ?

[15:00:00.0000] <Philip`>
You could use the hydrogen to make inflatable monkeys

[15:00:01.0000] <Hixie>
there's about 10^27 atoms in that monkey assuming he's about 1kg

[15:01:00.0000] <AryehGregor>
That looks like a heck of a lot less than 1 kg.

[15:01:01.0000] <AryehGregor>
Hmm, maybe not so much less.

[15:01:02.0000] <Hixie>
so that's 10^53 monkeys max

[15:01:03.0000] <TabAtkins_>
Yes,  by "very small" I meant "microscopic".

[15:01:04.0000] <AryehGregor>
I dunno, I'm not a primatologist.

[15:01:05.0000] <AryehGregor>
TabAtkins_, it has to be big enough to type blog posts at random, otherwise it won't work.

[15:02:00.0000] <TabAtkins_>
By the time I'm converting the mass of the universe into monkeys, I think I can devise alternate typing methods.

[15:02:01.0000] <Philip`>
How much entropy does a typical monkey contain?

[15:02:02.0000] <Hixie>
also alternate hashing methods

[15:03:00.0000] <AryehGregor>
Yeah, it'd probably be relatively easy to switch to a 1024-bit hash, if you have the resources to make that many monkeys.

[15:07:00.0000] <jgraham>
TabAtkins_: You know when the TAG said URLs were opaque identifiers you weren't supposed to take it as a commandment to make them as opaque as possible

[15:08:00.0000] <TabAtkins_>
WELL MAYBE THEY SHOULDN'T HAVE SUGGESTED IT THEN.

[15:08:01.0000] <TabAtkins_>
/me sulks.

[15:08:02.0000] <Dashiva>
jgraham: How do you know we weren't supposed to?

[15:10:00.0000] <AryehGregor>
jgraham, that's not as opaque as possible.  It has "document.php" in it.

[15:12:00.0000] <TabAtkins_>
AryehGregor: That's why I provided the more opaque but less entropic link /:gg5

[15:12:01.0000] <AryehGregor>
"less entropic"?

[15:12:02.0000] <TabAtkins_>
lower entropy

[15:13:00.0000] <TabAtkins_>
i was constrained in my wording by parallelism.

[15:13:01.0000] <TabAtkins_>
more/less

[15:14:00.0000] <AryehGregor>
How is it lower-entropy?

[15:14:01.0000] <TabAtkins_>
Less possible states.

[15:20:00.0000] <AryehGregor>
What's your definition of microstates and macrostates here?

[15:20:01.0000] <JonathanNeal>
Back

[15:20:02.0000] <TabAtkins_>
/me waves hands wildly.

[15:20:03.0000] <AryehGregor>
I mean, you say a macrostate is low-entropy if there are few microstates corresponding to it.

[15:21:00.0000] <TabAtkins_>
I really don't have a sufficient background to properly defend my statement.

[15:24:00.0000] <JonathanNeal>
TabAtkins_ I'll try to have that html5 ie print thing shared soon enough.

[15:24:01.0000] <JonathanNeal>
I hope I don't come across like a dick, I just want to give something that has had some testing.

[15:24:02.0000] <TabAtkins_>
Man, whatever, dick.

[15:24:03.0000] <TabAtkins_>
/me pfft.

[15:27:00.0000] <Hixie>
ok, finally finished dealing with e-mail

[15:27:01.0000] <Hixie>
now to find some food

[15:27:02.0000] <Hixie>
bbiab

[15:28:00.0000] <JonathanNeal>
:|

[15:31:00.0000] <gsnedders>
TabAtkins_: Bitch.

[15:31:01.0000] <TabAtkins_>
/me takes it to the street.

[15:55:00.0000] <KaOSoFt>
JonathanNeal- Internet Explorer Collection could help you in testing: http://finalbuilds.edskes.net/iecollection.htm

[15:58:00.0000] <KaOSoFt>
In fact, I had not tried it on Wine. Perhaps I should...

[15:58:01.0000] <KaOSoFt>
/me is downloading the IE Collection to try it on Ubuntu

[16:00:00.0000] <paul_irish>
dunno about through wine, but the spoon.net browsers are the most solid (aside from the VPC vm's) alternate IE versions there are

[16:01:00.0000] <KaOSoFt>
I think I had visited that site before, but it didn't have all those programs.

[16:02:00.0000] <KaOSoFt>
Hmm... I don't see IE.

[16:04:00.0000] <paul_irish>
http://spoon.net/Browsers/

[16:06:00.0000] <KaOSoFt>
Ohh, I see, since I'm on Linux it says IE is currently unavailable on my device, which explains why I didn't see it in the Browsers section.

[16:07:00.0000] <KaOSoFt>
In fact, ALL of it says my device is not supported. :'(

[16:08:00.0000] <KaOSoFt>
The a element without an href attribute now represents a "placeholder link". It can also contain flow content rather than being restricted to phrase content.

[16:08:01.0000] <KaOSoFt>
What does "placeholder link" mean?

[16:11:00.0000] <JonathanNeal>
IT WORKS!!!!!!

[16:12:00.0000] <KaOSoFt>
Link, or it doesn't exist...!

[16:12:01.0000] <KaOSoFt>
._.

[16:13:00.0000] <KaOSoFt>
"dir has been obsoleted in favor of ul". Does this include dir as in 'dir="ltr"'?

[16:15:00.0000] <jgraham>
KaOSoFt: A placeholder link is typically something that a script will come and fill in with a href value later on

[16:17:00.0000] <KaOSoFt>
Hmm, I see. I didn't use those because it wasn't unobtrusive, but if they are to be allowed now, all hail HTML%

[16:17:01.0000] <KaOSoFt>
HTML5!*

[16:24:00.0000] <annevk>
wow TabAtkins; URL fail :p

[16:25:00.0000] <annevk>
Hixie's URLs are not that great, but you are pushing it

[16:26:00.0000] <TabAtkins_>
KaOSoFt: That passage is talking about <dir>, not @dir.  Completely different things.

[16:28:00.0000] <TabAtkins_>
annevk: I'll note that the urls are much better for me.  The one I share is specifically made to be unguessable.  I just go a little overboard with it.  ^_^

[16:29:00.0000] <jgraham>
TabAtkins_: Why do you want an unguessable URL?

[16:30:00.0000] <jgraham>
I mean isn't it easier just to set up authentication?

[16:30:01.0000] <TabAtkins_>
jgraham: No.

[16:31:00.0000] <jgraham>
It seems easier to leak a URL

[16:32:00.0000] <jgraham>
(becasue you give it to someone, they link to it, google finds it, and you have lost)

[16:32:01.0000] <jgraham>
(whereas if you give them a username / password they implicitly understand it is supposed to be private)

[16:34:00.0000] <TabAtkins_>
That's fine.  Once I've given it to someone, I obviously want it to be findable.

[16:35:00.0000] <jgraham>
I have occasionally given out links to people that I did not want to be public (or, sometimes, they did not want to be public)

[16:35:01.0000] <Philip`>
/me once saw a site that encrypted a password plus today's date rounded to 6 months, to generate a URL for a private page

[16:36:00.0000] <Philip`>
and if you got the wrong password then you'd get the wrong URL, and if you had the URL then it'd only be valid for 6 months

[16:36:01.0000] <Philip`>
but I could find a few of those 6-month URLs since they'd leaked out onto Google

[16:37:00.0000] <Philip`>
and (since the encryption algorithm wasn't very good) use them to work out the password

[16:37:01.0000] <Philip`>
which then allowed access to the most recent page

[16:37:02.0000] <jgraham>
So I don't really buy the idea that everything shared is implicitly public

[16:38:00.0000] <TabAtkins_>
jgraham: You don't have to.  I was expressing my own feelings on content that I share.  ^_^

[16:39:00.0000] <jgraham>
TabAtkins_: Just don't be surprised if this happens every time you share a link :)

[16:39:01.0000] <TabAtkins_>
Philip`: Was it actually encrypted, or just hashed?

[16:39:02.0000] <jgraham>
Where by "this" I mean an inquisition about your URL scheme

[16:39:03.0000] <TabAtkins_>
Also, that seems pretty bad, since you have to prevent password collisions, and thus leak information about other pages on the system.

[16:40:00.0000] <TabAtkins_>
jgraham: That's why I need to remember to share my shorturls instead.  ^_^

[16:40:01.0000] <TabAtkins_>
Also, why I need to finally sit down and build myself a proper blogging platform out of the tools I have lying around.

[16:40:02.0000] <Philip`>
TabAtkins_: It was some rubbish JS hash-like algorithm

[16:41:00.0000] <TabAtkins_>
Philip`: Ah, homerolled crypto.  Nice.

[16:41:01.0000] <jgraham>
It is long past bedtime

[16:41:02.0000] <jgraham>
Goodnight

[16:42:00.0000] <KaOSoFt>
Goodnight.

[16:45:00.0000] <annevk>
jgraham, good point

[16:46:00.0000] <annevk>
I'm failing with bedtimes

[16:47:00.0000] <annevk>
Well, I value fun over sleep, which is something that does not scale, but I've yet to come to terms with that

[16:49:00.0000] <TabAtkins>
annevk: Surefile solution: get married.

[16:50:00.0000] <annevk>
I don't think I'm ready for that yet :)

[16:52:00.0000] <Hixie>
TabAtkins: how does getting married help with that problem?

[16:52:01.0000] <TabAtkins>
Well, it makes sure I get to bed before midnight every day.

[16:53:00.0000] <TabAtkins>
If Rachel's gone for some reason, my bedtimes hover around 3 or 4am

[16:53:01.0000] <Hixie>
i just ended up making carey go to bed around 4am...

[16:54:00.0000] <JonathanNeal>
Ready?


2010-04-09
[17:25:00.0000] <KaOSoFt>
JonathanNeal- Yes, yes, yes!

[17:26:00.0000] <KaOSoFt>
Well, I used to go to sleep at 1 AM or 2 AM, but since a month ago I started a routine in which I went to sleep early (9 PM, 10 PM) and woke up early as well (5 AM, 6AM).

[17:27:00.0000] <KaOSoFt>
It's tough in the beginning, but after a while you get used to it.

[17:27:01.0000] <JonathanNeal>
Well ... there you go.

[17:27:02.0000] <KaOSoFt>
There is one flaw, though: you get sleepy earlier. I NEVER got sleepier with my old schedule.

[17:28:00.0000] <KaOSoFt>
Arggh, I don't have IE here. Let me check a server.

[17:28:01.0000] <KaOSoFt>
>_>

[17:30:00.0000] <KaOSoFt>
Shim? 0_o

[17:31:00.0000] <JonathanNeal>
Verified, IE6, 7, 8

[17:33:00.0000] <KaOSoFt>
Black borders, white filled, right?

[17:34:00.0000] <KaOSoFt>
Hmm... let me check on Firefox, because Chrome doesn't have Page preview.

[17:36:00.0000] <KaOSoFt>
It works on Internet Explorer 7.

[17:37:00.0000] <KaOSoFt>
Now I'm checking ieHTML5.js

[17:38:00.0000] <KaOSoFt>
It replaces block elements with <div>, and in-line with <span>, right?

[17:39:00.0000] <KaOSoFt>
I'm a novice for everything Web related, besides browsing.

[17:39:01.0000] <KaOSoFt>
._.

[17:39:02.0000] <JonathanNeal>
Yes, please do not blog this until it's had a chance

[17:40:00.0000] <JonathanNeal>
to get bug tested by TabAtkins, ben alman, paul irish, and some of the other guys I trust.

[17:40:01.0000] <KaOSoFt>
Yeah, I'm going to keep testing it at home anyways.

[17:40:02.0000] <JonathanNeal>
Okay :) thanks

[17:41:00.0000] <KaOSoFt>
I haven't fully set up my blogging environment anyways.

[17:41:01.0000] <KaOSoFt>
I'm still deciding which template to use.

[17:42:00.0000] <KaOSoFt>
At least until I've fully coded an HTML5 template.

[17:45:00.0000] <KaOSoFt>
JonathanNeal- You had done research on it before, or you just started when I brought the topic up?

[17:46:00.0000] <JonathanNeal>
Haha, well actually it was because you were talking about it.

[17:47:00.0000] <JonathanNeal>
Or whoever brought it up in here or the other room.

[17:47:01.0000] <KaOSoFt>
Well, I guess (looking at the code) it was simple logic, after all.

[17:49:00.0000] <KaOSoFt>
I mean, the implementation method was the "hard?" thing.

[17:49:01.0000] <JonathanNeal>
No, it was just a matter of someone doing it.

[17:49:02.0000] <KaOSoFt>
Yeah, well, that. :P

[17:50:00.0000] <JonathanNeal>
Until I heard the fuss today, I didn't think it was a big deal (maybe it isn't) but at least it's fixed now.

[17:50:01.0000] <KaOSoFt>
By the way, what does "/*@cc_on!@*/" mean? I've seen it on HTML5Shiv as well.

[17:50:02.0000] <JonathanNeal>
Google it!  It's an IE proprietary thing, and I'm using it to keep other browsers from doing anything unnecessary.

[17:51:00.0000] <KaOSoFt>
Ugh, IE, as always.

[17:53:00.0000] <KaOSoFt>
So, IE reads JavaScript even for Page previewing, or your code did something to the DOM?

[17:53:01.0000] <KaOSoFt>
This is me trying to look like I know "something".

[17:54:00.0000] <KaOSoFt>
Oh, you're improving the example.

[17:54:01.0000] <KaOSoFt>
You're TEH man.

[17:54:02.0000] <TabAtkins>
Ok, interesting.  Not so much a fix as a workaround, but sure.

[17:55:00.0000] <Hixie>
ok updated my -88 change proposal and sent a new one

[17:56:00.0000] <JonathanNeal>
It doesn't fix the bug, but it fixes the problem.

[17:56:01.0000] <gsnedders>
Damned books that end on lame cliffhangers that make me want to get the next book

[17:57:00.0000] <gsnedders>
(and are generally shit)

[17:57:01.0000] <TabAtkins>
Solution: stop reading shit books

[17:58:00.0000] <gsnedders>
But I want to know what happens!

[17:58:01.0000] <daedb_>
Which book?

[17:58:02.0000] <gsnedders>
daedb_: I'm sorry, but I don't want to embarrass myself that much

[17:59:00.0000] <daedb_>
It can't be that bad :)

[17:59:01.0000] <gsnedders>
Oh yes it can.

[18:00:00.0000] <gsnedders>
I mean, I may be willing to admit I think Twilight (at least the first book) is awesome, but to drop this low?

[18:01:00.0000] <gsnedders>
Also: there aren't enough books remaining to buy to be able to get them for three for two.

[18:01:01.0000] <gsnedders>
And I don't want to buy them full price.

[18:01:02.0000] <gsnedders>
They're too shit for that.

[18:06:00.0000] <daedb_>
/me was not aware that there's something more embarrasing on the internet than Twilight...

[18:06:01.0000] <gsnedders>
Oh, really, there is.

[18:06:02.0000] <gsnedders>
Imagine Twilight without the only good bit.

[18:07:00.0000] <gsnedders>
(I guess the fact I'm implying the very existance of a good bit of Twilight implies I'm a teenage girl)

[18:07:01.0000] <daedb_>
To be fair, I know just about nothing about Twilight except that it has vampires and everyone (on the internet at least) seems to hate it

[18:07:02.0000] <gsnedders>
(Oh well. All the teenage girls I hung around with the entire time at school concluded I was a teenage girl.)

[18:11:00.0000] <gsnedders>
(Those same girls may have something to do with the fact I read Twilight. It may have taken them around three years to bully me into it, though.)

[18:13:00.0000] <gsnedders>
/me heads off, after having embarrassed himself on the intarwebs

[18:30:00.0000] <estellevw>
"I just fixed the #html5 printing bug in #IE - woohoo!" - yay!

[18:31:00.0000] <estellevw>
oops, was quoting. I didn't solve it. jonathan neal did

[18:31:01.0000] <gsnedders>
https://lists.webkit.org/pipermail/webkit-dev/2010-April/012235.html

[18:31:02.0000] <estellevw>
anyone have an url?

[18:33:00.0000] <TabAtkins>
gsnedders: Yeah, that causes some grumbling over here.  "Hey, we like that multiprocess idea you guys have so much, we decided to do it completely differently!"  ^_^

[18:34:00.0000] <TabAtkins>
estellevw: JonathanNeal is still doing testing/tweaking, so he's not releasing it publicly quite yet.  We'll let you know.  ^_^

[18:34:01.0000] <estellevw>
i wanted to retweet, but figured it would be more interesting with a link.

[18:35:00.0000] <gsnedders>
TabAtkins: Having an API designed for it, and to be totally non-blocking, seems better than the current state of Chrom(e|ium)

[18:35:01.0000] <estellevw>
TabAtkins: congrats on the Google gig by the way. how was your first (second) week?

[18:35:02.0000] <TabAtkins>
I have no opinion on the technical merits of either approach.  Just commenting.

[18:36:00.0000] <gsnedders>
Yeah, I can't imagine is being brilliant ;P

[18:36:01.0000] <TabAtkins>
estellevw: Doing good!  It's nice to have the freedom to do just spec work without stealing time from my "real" job.

[18:36:02.0000] <gsnedders>
/me needs to work out whether he'll have a job in September

[18:37:00.0000] <gsnedders>
Just, like, minor decision in life

[18:37:01.0000] <Dashiva>
Are you becoming a redundancy?

[18:37:02.0000] <TabAtkins>
gsnedders: Come to America and we'll land you a job with one of us stateside browsers.

[18:37:03.0000] <gsnedders>
No, just a question of whether I go to uni or not.

[18:37:04.0000] <gsnedders>
TabAtkins: heh

[18:38:00.0000] <Dashiva>
I'm sure he has a non-compete clause

[18:39:00.0000] <gsnedders>
Some of my colleagues seem me to want to stay and not go to uni for fear of me doing not returning and going to a competitor

[18:39:01.0000] <gsnedders>
TabAtkins: Equally, none of them would hire me without a degree

[18:40:00.0000] <Dashiva>
/me might be to blame there

[18:40:01.0000] <TabAtkins>
gsnedders: You think my degree had a thing to do with me getting my job?

[18:40:02.0000] <TabAtkins>
Answer: no.

[18:41:00.0000] <gsnedders>
TabAtkins: In general it seems a requirement, but one that actually is just expected as a baseline and otherwise ignored

[18:41:01.0000] <TabAtkins>
In general, yes.  In specific, once you get past the non-engineers doing recruiting, you're golden.

[18:42:00.0000] <KaOSoFt>
I wish I'd have your chance.

[18:43:00.0000] <KaOSoFt>
I've been in the same spot for two years now, and I think I deserve better.

[18:43:01.0000] <gsnedders>
Like, I'm nowhere near even meeting the requirements for an internship at Google

[18:43:02.0000] <KaOSoFt>
</hater>

[18:44:00.0000] <KaOSoFt>
I went back to finish my degree only because I saw engineers earning a whole lot more than I do, and in the end (and by my boss orders), it was me who did the job. :'(

[18:45:00.0000] <KaOSoFt>
Once I get my degree, I'll start sending curriculums everywhere if I don't get promoted here.

[18:50:00.0000] <KaOSoFt>
Time to leave (work). Goodnight.

[19:05:00.0000] <JonathanNeal>
Back home!

[20:49:00.0000] <boblet>
hey all. anyone know of any spec repos that accept patches as a way of filing bugs? I found http://github.com/html5/spec but looks like Manu isn’t auto-updating it

[00:02:00.0000] <JonathanNeal>
TabAtkins, TabAtkins_,  you around?

[00:15:00.0000] <JonathanNeal>
Hey all.

[00:44:00.0000] <estellevw>
/me hears crickets chirping

[00:45:00.0000] <othermaciej>
hi!

[00:50:00.0000] <JonathanNeal>
haha

[00:50:01.0000] <JonathanNeal>
Boy, I was hoping we could coordinate testing for this html5 print thing.

[01:25:00.0000] <estellevw>
i can help you test

[01:25:01.0000] <estellevw>
i am on XP on VMWare with Real IE8 and faux IE6

[02:16:00.0000] <JonathanNeal>
estellevw, you still available?

[02:16:01.0000] <JonathanNeal>
I was working on a new version.

[02:28:00.0000] <zcorpan>
http://forums.whatwg.org/viewtopic.php?t=4271 - any cache manifest people here?

[02:33:00.0000] <annevk>
the reason is for debugging

[02:33:01.0000] <annevk>
you need to add the online whitelist flag

[02:33:02.0000] <annevk>
if you want it to work

[02:34:00.0000] <nessy>
Hixie: around?

[02:35:00.0000] <Hixie>
nessy: vaguely

[02:35:01.0000] <nessy>
I just submitted my first bug and wanted to know if that's the right form: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9452

[02:36:00.0000] <nessy>
if you don't have time to look, that's ok, too

[02:37:00.0000] <Hixie>
looks fine to me

[02:37:01.0000] <nessy>
ok, thanks

[02:37:02.0000] <nessy>
I'll do at least one more :)

[02:37:03.0000] <othermaciej>
hey nessy 

[02:37:04.0000] <nessy>
hey othermaciej :)

[02:37:05.0000] <Hixie>
nessy: thank _you_! :-)

[02:37:06.0000] <nessy>
finally getting to it!

[02:38:00.0000] <nessy>
pleasure :)

[02:38:01.0000] <othermaciej>
woot, multitrack API

[02:38:02.0000] <nessy>
the recent a11y TF F2F was most interesting - and actually somewhat productive

[02:39:00.0000] <othermaciej>
yeah?

[02:39:01.0000] <othermaciej>
MikeSmith sent me some notes, but I don't know the details

[02:39:02.0000] <nessy>
need to work over the caption proposal a bit before submitting, but I'm also planning to get that into the bug tracker today

[02:40:00.0000] <nessy>
well, the biggest issue that held back the two existing proposals were really the SMIL people's objections

[02:40:01.0000] <nessy>
but we managed to explain that their objections are not really with our proposals, but with the lack of SMIL functionality in HTML5

[02:40:02.0000] <othermaciej>
are you going to submit these as change proposals to the WG as well or just as bugs?

[02:40:03.0000] <othermaciej>
that's good

[02:40:04.0000] <nessy>
which for the state of media in HTML5 is a fair state IMO

[02:41:00.0000] <nessy>
well, Mike told me to submit bugs

[02:41:01.0000] <othermaciej>
Multitrack API seems particularly unrelated to SMIL

[02:41:02.0000] <nessy>
they are essentially change proposals, so now I don't know if I should also submit change proposals and how to do so

[02:41:03.0000] <nessy>
indeed!

[02:41:04.0000] <othermaciej>
to submit a Change Proposal you just need to send mail to public-html with the contents or a link to it

[02:42:00.0000] <nessy>
I think I will do that, too

[02:42:01.0000] <nessy>
probably even to WHATWG

[02:42:02.0000] <nessy>
the more eyes are aware of it and look at it, the better

[02:42:03.0000] <othermaciej>
definitely

[02:43:00.0000] <othermaciej>
I'm sure there are potential refinements in both cases

[02:43:01.0000] <othermaciej>
I am glad it is moving forward

[02:44:00.0000] <nessy>
so am I - it was really hard to do without stepping on people's toes

[02:45:00.0000] <boblet>
Validator.nu text field input + validate just reloads homepage. just me or known issue?

[02:45:01.0000] <nessy>
but I was very surprised about the lack of technical understanding of the HTML5 media elements with some folks - that was really the core learning of the F2F

[02:46:00.0000] <nessy>
it's probably natural not to be experts across everything

[02:46:01.0000] <othermaciej>
nessy: what were some of the misunderstandings?

[02:46:02.0000] <nessy>
still - that lack of understanding held back progress, which is what really surprised me

[02:47:00.0000] <nessy>
well, one expectation is that just by introducing audio and video into HTML5, HTML5 will magically do everything that DAISY was built for

[02:47:01.0000] <nessy>
all the table of contents stuff, direct access to content, sequencing of content etc

[02:48:00.0000] <nessy>
all of this is, of course, possible using javascript with HTML5 now, but the expectation was that it'll just be in the declarative parts

[02:48:01.0000] <othermaciej>
I don't really understand what DAISY is

[02:48:02.0000] <nessy>
think of talking books

[02:48:03.0000] <nessy>
seen any of those?

[02:48:04.0000] <othermaciej>
like audiobooks?

[02:48:05.0000] <othermaciej>
I'm familiar with the concept, though not a user myself

[02:48:06.0000] <nessy>
not just CDs that are read out books

[02:49:00.0000] <othermaciej>
I'm looking at the DAISY consortium page

[02:49:01.0000] <nessy>
more books and news etc that has been read in chapters and is directly accessible

[02:50:00.0000] <nessy>
http://www.readhowyouwant.com/formats/audio-format.aspx <- is quite instructive

[02:50:01.0000] <Hixie>
like audible.com books?

[02:50:02.0000] <asmodai>
oh nice - http://www.mozilla.com/en-US/firefox/lorentz/

[02:50:03.0000] <othermaciej>
do audible.com book have structure?

[02:51:00.0000] <Hixie>
well they have chapters and stuff

[02:51:01.0000] <Hixie>
like a regular MPEG file

[02:51:02.0000] <Hixie>
not sure what else you would want from an audio book

[02:51:03.0000] <nessy>
e.g. http://www.youtube.com/watch?v=bgWOoQ6FgA0

[02:51:04.0000] <annevk>
blargh, Tyler's rhetoric is starting to annoy me

[02:52:00.0000] <othermaciej>
well at least I got him to comment on what "subset" means

[02:53:00.0000] <annevk>
the continued insistence on there being a problem with CORS while there's not at all agreement on that; saying UMP exists to solve the non-definition of credentials in CORS while I pointed out several months ago that was a fixed issue, etc.

[02:54:00.0000] <annevk>
he doesn't want to work together, he just wants to push his thing through

[02:55:00.0000] <othermaciej>
well, I expect Apple at least will ignore it either way

[02:55:01.0000] <annevk>
in almost similar tactic to the XHTML2 WG; by defining something that cannot be implemented and therefore avoiding all the critical steps in the WD-process

[02:55:02.0000] <othermaciej>
I wonder if any vendor will announce plan to implement UMP

[02:55:03.0000] <annevk>
(the XHTML2 WG used modularization as an excuse to say it would not directly be implemented but rather other things would build on top of it and those would be implemented and tested)

[02:55:04.0000] <Hixie>
if there's no vendor willing to implement, we should reguse to publish it

[02:56:00.0000] <Hixie>
there's no point working on things without vendor support

[02:56:01.0000] <othermaciej>
there's the highly dubious "it's already sort of supported if you support CORS" argument

[02:57:00.0000] <othermaciej>
so far no vendors but Apple have stated their intent on the list

[02:57:01.0000] <othermaciej>
though privately it seem like Google folks aren't really interested either

[02:57:02.0000] <annevk>
pretty much the same as when Steven claimed Opera supported XHTML2

[02:57:03.0000] <Hixie>
if it's a strict subset of CORS and all the vendors are implementing CORS then UMP doesn't seem to add anything

[02:58:00.0000] <othermaciej>
it adds purity of essence

[02:58:01.0000] <nessy>
(man, there are no short videos to explain DAISY! this is one is a bit better http://www.youtube.com/watch?v=H9gKj2hLFoo )

[02:58:02.0000] <othermaciej>
otherwise the Origin header will introduce a foreign substance into our precious bodily fluid

[03:00:00.0000] <othermaciej>
it sound like DAISY is something you could implement given JavaScript and HTML5, but it's not just built-in HTML5 functionality

[03:02:00.0000] <nessy>
indeed

[03:02:01.0000] <nessy>
that's what is hard to explain

[03:02:02.0000] <nessy>
and many in the a11y TF want to drive HTML5 to support it declaratively

[03:03:00.0000] <annevk>
more crazy markup languages?

[03:04:00.0000] <annevk>
o_O

[03:04:01.0000] <othermaciej>
that would be a pretty complicated and specialized addition

[03:05:00.0000] <annevk>
yeah that's not going to happen

[03:05:01.0000] <othermaciej>
maybe worth adding someday, but not as important as critical things like captioning and the like

[03:06:00.0000] <nessy>
check out this video at about 2min in http://www.youtube.com/watch?v=mxI8zhLqs3M

[03:07:00.0000] <othermaciej>
so is the interesting bit the fact that it syncs with the text?

[03:07:01.0000] <nessy>
I think over time we might get those features - but right now is too early

[03:07:02.0000] <nessy>
it has all sorts of interesting features

[03:07:03.0000] <othermaciej>
IBooks btw will be accessible via audio without the need to make a special version with audio files in it

[03:08:00.0000] <nessy>
you can search through long audio recordings by jumping from section to section, search by text, also get screenreader functionality in visible stuff like the Web etc

[03:08:01.0000] <nessy>
what standard to IBooks follow?

[03:09:00.0000] <othermaciej>
they are basically ePub

[03:09:01.0000] <othermaciej>
except that we don't restrict them to the epub subset of HTML

[03:09:02.0000] <othermaciej>
arbitrary XHTML5 can be used

[03:10:00.0000] <jgraham>
othermaciej: Man, now I feel conflicted. You embraced and extended a standrd, but did it by making it use another open standard and made it better

[03:10:01.0000] <othermaciej>
or at least, anything that WebKit can support

[03:10:02.0000] <jgraham>
Does it restrict to XML?

[03:11:00.0000] <nessy>
ePub is also a DAISY standard

[03:12:00.0000] <jgraham>
(I think ePub would be a lot better if it were HTML+an optional manifest and some conventions about how to extract data from the HTML)

[03:12:01.0000] <jgraham>
(and honestly it would be nice if you could avoid the manifest)

[03:13:00.0000] <nessy>
interesting - maybe Apple has already managed to marry the two worlds ...

[03:16:00.0000] <othermaciej>
jgraham: it's actually typically multiple XHTML files - one per chapter

[03:17:00.0000] <othermaciej>
jgraham: I actually don't know offhand if you have to use XML or can use the HTML syntax - most existing epub books are XHTML though

[03:17:01.0000] <zcorpan>
"though ~5% of another sample were using the pragma incorrectly anyway [2]."

[03:18:00.0000] <zcorpan>
Hixie: i don't see how to draw that conclusion from [2] http://hsivonen.iki.fi/test/moz/analysis.txt ?

[03:19:00.0000] <Hixie>
oh i misread it

[03:19:01.0000] <Hixie>
it's cases of content-language, not cases of bad values of content-language

[03:20:00.0000] <Hixie>
so i guess it's just repeating the previous datum

[03:20:01.0000] <zcorpan>
indeed

[03:21:00.0000] <jgraham>
othermaciej: Yeah, I didn't mean "a single HTML file" so much as "an arbitary number of HTML files"

[03:21:01.0000] <jgraham>
I believe ePub theoretically requires well formed XML, which would explain most books having that format

[03:21:02.0000] <othermaciej>
jgraham: it basically is XHTML plus a manifest - I'm actually not sure if text/html would work

[03:22:00.0000] <othermaciej>
once I get an iPad I'll try sticking handmade epub books on it to see

[03:22:01.0000] <zcorpan>
/me wonders whether ePub will go down the legacy mobile path and treat application/xhtml+xml as text/html

[03:22:02.0000] <jgraham>
othermaciej: As far as I can tell both the manifest and the chapter files are supposed to be well-formed XML

[03:23:00.0000] <jgraham>
With namespaces and all sorts of crap

[03:23:01.0000] <jgraham>
It seems altogether unnecessary

[03:25:00.0000] <othermaciej>
jgraham: definitely "supposed to" according to epub - I don't know if our implementation enforces that restriction

[03:54:00.0000] <hsivonen>
It's kinda sad that the ePub folks are still tracking the XHTML2 WG's stuff instead of XHTML5

[03:55:00.0000] <hsivonen>
is the book reader on iPad using WebKit?

[03:55:01.0000] <annevk>
othermaciej just said so

[03:55:02.0000] <hsivonen>
oh

[03:55:03.0000] <annevk>
it supports whatever WebKit supports

[03:55:04.0000] <hsivonen>
/me reads scrollback

[03:56:00.0000] <othermaciej>
it's basically epub + xhtml5

[03:56:01.0000] <hsivonen>
so what's the deal with reports saying you can't use <video> in ePub on iPad?

[03:56:02.0000] <othermaciej>
dunno, could be a bug

[03:56:03.0000] <annevk>
ooh, WebKit is doing <progress> already to some extent

[03:56:04.0000] <hsivonen>
and is CSS restricted somehow to make all the books look the same?

[03:56:05.0000] <othermaciej>
maybe there are artificial restrictions I don't know about

[03:56:06.0000] <othermaciej>
books can include their own CSS

[03:59:00.0000] <annevk>
ah, Nokia is adding it

[04:37:00.0000] <hsivonen>
Is there an up-to-date report of PICS support in browsers?

[04:37:01.0000] <hsivonen>
hmm. PICS in ungooglable

[04:38:00.0000] <hsivonen>
/me guesses that Google treats "PICS" as a synonym for "pictures"

[04:38:01.0000] <hsivonen>
/me files a search quality complaint

[04:39:00.0000] <Philip`>
Searching for "+pics" finds the right PICS, about a dozen places down

[04:41:00.0000] <hsivonen>
Philip`: thanks

[04:43:00.0000] <othermaciej>
Safari has no PICS support

[04:43:01.0000] <othermaciej>
I believe IE does have it

[04:45:00.0000] <hsivonen>
so PICS label is basically an IE only feature

[04:45:01.0000] <hsivonen>
and in the wishful W3C land PICS has been superseded by an RDF vocabulary

[04:45:02.0000] <hsivonen>
/me wonders if IE support the RDF thingy

[04:46:00.0000] <hsivonen>
did FCC and FTC drop interest shortly after IE implemented PICS?

[04:47:00.0000] <hsivonen>
or why don't other browsers support PICS?

[04:47:01.0000] <hsivonen>
if IE dropped PICS support today, would FTC and FCC wake up again?

[04:48:00.0000] <othermaciej>
was PICS ever driven by legal requirements?

[04:49:00.0000] <othermaciej>
PICS basically seems like a way for sites to claim they won't do bad things with your personal information, in machine-readable form

[04:49:01.0000] <othermaciej>
I am not sure what "support" even means really

[04:49:02.0000] <hsivonen>
othermaciej: I thought the whole point of PICS was that the Web folks could say to the FTC and the FCC that the Web is a self-regulating industry

[04:49:03.0000] <hsivonen>
othermaciej: PICS is a mechanism for saying "we don't have nudity here"

[04:49:04.0000] <hsivonen>
othermaciej: it's not about personal info

[04:50:00.0000] <othermaciej>
oh

[04:50:01.0000] <hsivonen>
like the MPAA self-rates almost every movie as R and says they are a self-regulating industry so the Congress can stay out

[04:50:02.0000] <othermaciej>
I guess porn sites are not usually motivated to claim not to be porn sites

[04:51:00.0000] <othermaciej>
I guess in theory then Parental Controls mode in Safari could do something with PICS, but I am not sure enough sites use it

[04:51:01.0000] <hsivonen>
www.apple.com claims to have no nudity or sexual material, no violence, no potentially offensive language and no "other topics"

[04:51:02.0000] <hsivonen>
gotta love the lack of other topics

[04:52:00.0000] <othermaciej>
I believe "other topics" would be against our developer agreement

[04:53:00.0000] <jgraham>
/me wonders if one should take "other" literally and conclude that there is no material at all on apple.com

[04:53:01.0000] <hsivonen>
maybe "other topics" means the category labeled "potentially harmful activities" elsewhere

[04:54:00.0000] <hsivonen>
i.e. tobacco, alcohol, drugs, weapons, gambling, setting bad example to children, horror and distrimination based on gender, sexual orientation, ethnic, religious or national identity

[04:56:00.0000] <hsivonen>
*discrimination

[04:57:00.0000] <annevk>
I wonder how often IRI issues have to be explained again

[04:58:00.0000] <annevk>
So far this IRI WG does not seem very effective...

[05:00:00.0000] <hsivonen>
othermaciej: P3P was the privacy thing

[05:01:00.0000] <othermaciej>
hsivonen: ah, I always get those confused apparently

[05:01:01.0000] <hsivonen>
IIRC that was about telling the FTC and the EU that the Web is a self-regulating industry and doesn't need governmental regulation

[05:01:02.0000] <annevk>
o_O http://lists.w3.org/Archives/Public/public-iri/2010Apr/0015.html email client fail

[05:02:00.0000] <annevk>
fortunately I wrote the character out too

[05:14:00.0000] <hsivonen>
hmm. I wonder if I should bother to escalate the <nobr> but to the tracker...

[05:17:00.0000] <hsivonen>
Microsoft has published documentation about PICS support on 2010-03-25

[05:17:01.0000] <othermaciej>
hsivonen: you disagree with Hixie's stance on nobr?

[05:17:02.0000] <hsivonen>
othermaciej: yes

[05:18:00.0000] <othermaciej>
I thought about reopening if I could find a case that is as arguably semantic as <pre>

[05:18:01.0000] <hsivonen>
I think <nobr> isn't used without thinking, so banning it doesn't serve an educational purpose

[05:18:02.0000] <othermaciej>
it seems that in principle <nobr> is no less semantic than <pre>

[05:18:03.0000] <hsivonen>
and having to use span + style is less convenient than the interoperable dedicated element

[05:18:04.0000] <annevk>
Isn't PICS that P3P thing or so? That they use to determine whether third party cookies can be set or something silly like that?

[05:19:00.0000] <othermaciej>
apparently I made a better case for <wbr>

[05:19:01.0000] <hsivonen>
annevk: PICS is about declaring "no nudity". P3P is about declaring "we don't do bad things with cookies"

[05:20:00.0000] <annevk>
ah

[05:20:01.0000] <annevk>
all this having faith in the supplier seems kind of misguided

[05:22:00.0000] <hsivonen>
annevk: they are smokescreens to repel government authorities

[05:26:00.0000] <annevk>
some people genuinely believe in them

[05:26:01.0000] <annevk>
did you follow the geolocation mess?

[05:26:02.0000] <annevk>
they wanted something like PICS/P3P for geolocation

[05:26:03.0000] <annevk>
I'm somewhat glad that failed to materialize

[05:28:00.0000] <zcorpan>
PICS for geolocation? to prevent zooming in on nude people on google maps?

[05:29:00.0000] <othermaciej>
geopriv

[05:29:01.0000] <othermaciej>
they think that whenever you transmit a piece of geolocation data, you should label it with the user's preferences for what they'd like to allow the receiver to do with it

[05:29:02.0000] <zcorpan>
or don't give away my location while i'm at the strip club

[05:29:03.0000] <annevk>
zcorpan, hehe, did you see the Google Maps / Chatroulette mashup? it's scary

[05:32:00.0000] <zcorpan>
hadn't

[07:42:00.0000] <remysharp>
Regarding media elements - if I want to visualise the download progress of the playable timerange, should I be listening for the durationChanged event or the progress event? 

[07:43:00.0000] <zcorpan>
remysharp: progress

[07:44:00.0000] <remysharp>
I was using progress in a test, but (I think Firefox) wasn't firing all the way to the end of the media downloading

[07:44:01.0000] <zcorpan>
remysharp: durationchange is only fired when the duration changes, which will likely be just once very early

[07:44:02.0000] <remysharp>
i.e. I'd read the video.seekable.end() when progress fired, and it wouldn't be the completed length

[07:45:00.0000] <zcorpan>
remysharp: i think that's a bug then

[07:45:01.0000] <remysharp>
or should I have been using the video.duration prop on the progress event firing?

[07:46:00.0000] <zcorpan>
duration doesn't say how much has been downloaded

[07:46:01.0000] <remysharp>
no - sorry I should have been clear - it was to update a playhead scrubber

[07:46:02.0000] <zcorpan>
remysharp: "Once the entire media resource has been fetched (but potentially before any of it has been decoded)

[07:46:03.0000] <zcorpan>
Queue a task to fire a simple event named progress at the media element."

[07:46:04.0000] <remysharp>
i.e. what amount of the video can be played

[07:47:00.0000] <remysharp>
ah - so the video is decoded after that event

[07:47:01.0000] <zcorpan>
might be

[07:47:02.0000] <remysharp>
therefore the duration wouldn't potentially be calculated on the last progress event firing?

[07:48:00.0000] <remysharp>
but - I guess that said, if the meta data comes down first, I'd have the duration before the video can be played, right?

[07:48:01.0000] <zcorpan>
the duration can be known before having decoded any video, e.g. mozilla supports an http header to indicate duration

[07:48:02.0000] <remysharp>
okay

[07:48:03.0000] <zcorpan>
if you want to know which parts you can seek to, you probably want .seekable

[07:49:00.0000] <remysharp>
I've just checked my code example, and I'm listening for the durationChange *and* progress, but updating the scrubber based on the seekable.end() value

[07:49:01.0000] <zcorpan>
(it's durationchange btw with lowercase c)

[07:49:02.0000] <remysharp>
aye, but I want it to show how much has loaded as it's loading

[07:49:03.0000] <remysharp>
(lower c - ta)

[07:55:00.0000] <remysharp>
zcorpan: cheers for your help.

[08:11:00.0000] <hsivonen>
regarding http://html5doctor.com/your-questions-answered-6/ : the doctor says he doesn't know why Mozilla have taken that stance (re: not using OS MP3 decoder)

[08:11:01.0000] <hsivonen>
hasn't this been communicated pretty clearly?

[08:11:02.0000] <asmodai>
mmm, webkit 2? https://lists.webkit.org/pipermail/webkit-dev/2010-April/012235.html

[08:20:00.0000] <zcorpan>
hsivonen: i recall reading several statements about h.264, but i don't recall anything about mp3 or using os decoders in general

[08:25:00.0000] <hsivonen>
zcorpan: http://shaver.off.net/diary/2010/01/23/html5-video-and-codecs/ and http://weblogs.mozillazine.org/roc/archives/2009/06/directshow_and.html don't limit the discussion to H.264

[08:38:00.0000] <zcorpan>
hsivonen: i don't recall having read those two :-)

[09:00:00.0000] <JonathanNeal>
Hi all

[09:04:00.0000] <hsivonen>
hi

[09:13:00.0000] <hsivonen>
/me wonders if <option> has been in the "special" category in the past

[09:45:00.0000] <remysharp>
another media question -

[09:45:01.0000] <remysharp>
I've been using the canplay event to say "this is the time I should enable my custom controls"

[09:45:02.0000] <remysharp>
but sometimes the video download will suspend, so canplay doesn't fire

[09:46:00.0000] <remysharp>
I'm thinking that I should enable my controls at the loadeddata event - would that be a better approach?

[09:46:01.0000] <remysharp>
the user can then start the playback, which then eventually fires canplay (I'm seeing this in Firefox btw)

[09:50:00.0000] <TabAtkins>
I think that's the right thing anyway; iirc for some types of video (live streams?) FF will *never* fire canplay,

[09:50:01.0000] <remysharp>
cheers - exactly what I was after

[09:51:00.0000] <remysharp>
ta

[09:53:00.0000] <remysharp>
ooh - one question tho -

[09:53:01.0000] <remysharp>
the preload="metadata" attribute

[09:53:02.0000] <remysharp>
do you know (I'm trying to work out from the spec now) whether that will only fire the loadedmeta data

[09:53:03.0000] <remysharp>
and not the loadeddata

[09:54:00.0000] <remysharp>
because then I'd be stuck with no ui

[09:54:01.0000] <remysharp>
so perhaps I could hook at the loadedmetadata instead?

[09:54:02.0000] <TabAtkins>
Not sure at that point.

[09:54:03.0000] <remysharp>
Darn :-\

[09:54:04.0000] <TabAtkins>
I think roc is working on video in FF.

[09:55:00.0000] <remysharp>
roc? Is that someone in here?

[09:55:01.0000] <TabAtkins>
robert o'callahan.  Don't know if he's in the room right now.

[09:56:00.0000] <TabAtkins>
/me still hasnt' figured out all of irssi's commands.

[09:56:01.0000] <remysharp>
okay, cheers for that - I'll keep my eyes peeled :)

[10:07:00.0000] <Philip`>
TabAtkins: /names to see who's here, or just type their name and press tab and see if it autocompletes

[10:08:00.0000] <TabAtkins>
Ah, thanks for /names.  roc has a short enough name that I thought perhaps robod was blocking him or something.

[10:10:00.0000] <Philip`>
Type in the whole name and tab-complete and it'll add a ": " if the person is here

[10:11:00.0000] <TabAtkins>
Yeah, I guess that's true.

[10:11:01.0000] <Philip`>
or just the first few letters and tab will cycle through all the options

[10:11:02.0000] <TabAtkins>
Ah, didn't know about the cycling.

[10:11:03.0000] <TabAtkins>
I just delete and type more if it doesn't hit the right person first time.

[10:12:00.0000] <gsnedders>
/me just learnt to use irssi by hitting random keys

[10:19:00.0000] <hsivonen>
hmm. WebKit closes stuff more aggressively on </form> than old Gecko or IE do

[10:19:01.0000] <hsivonen>
I wonder if the spec should change to be like WebKit on <form><foo>bar</form>baz

[10:19:02.0000] <hsivonen>
anyway, I believe <button> shouldn't be scoping

[10:20:00.0000] <JonathanNeal>
Hi everybody!

[10:20:01.0000] <hsivonen>
bug that made me look at <button> and </form>: https://bugzilla.mozilla.org/show_bug.cgi?id=558302

[10:21:00.0000] <TabAtkins>
hsivonen: Damn you XML!

[10:24:00.0000] <ROBOd>
TabAtkins: me? blocking? who? :)

[10:25:00.0000] <TabAtkins>
If roc was in the room, your name would probably come directly before him.  So I wasn't sure if he was actually not here or what.  ^_^

[10:30:00.0000] <JonathanNeal>
How would you describe the HTML5 ie problem?

[10:30:01.0000] <TabAtkins>
The specific treats-unknown-element-start-and-end-tags-as-void-elements problem?

[10:31:00.0000] <JonathanNeal>
Yea.

[10:32:00.0000] <JonathanNeal>
Hmm, looking to summarize what my script will do.

[10:32:01.0000] <TabAtkins>
I would describe it as "IE treats unknown element start and end tags as void elements".

[10:33:00.0000] <JonathanNeal>
Yea, yea.

[10:33:01.0000] <JonathanNeal>
Hmm, so I'm describing it as "IE Print Protector is a script that allows you to print HTML5 pages in Internet Explorer" and I'll use what you wrote to describe the issue.

[10:33:02.0000] <paul_irish>
Thus making them impossible to style and they instantly collapse, spewing their contents into the DOM

[10:33:03.0000] <TabAtkins>
Probably describe wtf that actually means to the page author, which is "the unknown elements no longer wrap their contents".

[10:35:00.0000] <miketaylr>
Print Protector™

[10:43:00.0000] <annevk>
oh, looks like I hit a nerve

[10:49:00.0000] <JonathanNeal>
No TM, miketaylr :-P I don't want anyone to think I somehow am a genius.  It will only give them reason to prove I am not. :)

[10:49:01.0000] <JonathanNeal>
"Oh, I packaged this whole thing up for you, I wrote this script, look at me, I'm so cool"

[10:50:00.0000] <miketaylr>
hahaha

[10:50:01.0000] <JonathanNeal>
We test it, make it look sexy so those who need it find it easily and merrily.

[10:51:00.0000] <JonathanNeal>
Like charitable prostitution

[11:24:00.0000] <JonathanNeal>
TabAtkins, wanna help me switch to microdata for the ie print page I'll be building?  I need a link to your tutorial again :)

[11:31:00.0000] <TabAtkins>
Sure!

[11:31:01.0000] <TabAtkins>
www.xanthir.com/:gg5

[11:47:00.0000] <KaOSoFt>
Good afternoon.

[11:48:00.0000] <JonathanNeal>
Thanks!

[11:51:00.0000] <KaOSoFt>
?

[11:52:00.0000] <KaOSoFt>
JonathanNeal- I tested it at home (IE8), and it worked marvelously.

[11:58:00.0000] <JonathanNeal>
KaOSoFt, awesome.

[11:59:00.0000] <KaOSoFt>
JonathanNeal- Whenever you fell it's ready, let me know. My blog site is not quite ready yet, but I'll publish whenever is possible/viable.

[12:09:00.0000] <JonathanNeal>
Cool.

[12:55:00.0000] <KaOSoFt>
feel*

[13:00:00.0000] <AryehGregor>
It annoys me when page speed analyzers complain that you aren't using a CDN.

[13:01:00.0000] <AryehGregor>
. . . Maybe it shouldn't.  Are there cheap/easy-to-use CDNs around?

[13:01:01.0000] <AryehGregor>
/me gets the impression that at least the high-profile ones are "call for a quote"-type things

[13:07:00.0000] <TabAtkins>
AryehGregor: afaik, you're right.

[13:07:01.0000] <Dashiva>
Coral is free :)

[13:07:02.0000] <Dashiva>
But I have trouble imagining coral would be faster than not using a CDN

[13:08:00.0000] <AryehGregor>
Yeah, I'm also wondering about whether you'd get real performance improvements from using some smaller CDNs.

[13:08:01.0000] <AryehGregor>
In principle, a CDN with a lot of connectivity might be able to charge *less* for the same bandwidth than the data center where your site is hosted.

[13:09:00.0000] <AryehGregor>
Although the storage probably costs them more, due to the duplication.

[13:09:01.0000] <TabAtkins>
I just ignore that recommendation when I'm using YSlow, but it does annoy me that I can't get a perfect report card.  ;_;

[13:09:02.0000] <AryehGregor>
/me notes that Akamai's website appears to be written entirely in Flash

[13:09:03.0000] <AryehGregor>
I was just using webpagetest.org.

[13:10:00.0000] <AryehGregor>
/me didn't realize his images were so badly compressed

[13:10:01.0000] <KaOSoFt>
AryehGregor- Just make a subdomain of the like: media.yoursite.com

[13:10:02.0000] <TabAtkins>
YSlow has a "profile" you can set, though - the "small site" setting doesn't complain about CDNs.

[13:10:03.0000] <AryehGregor>
KaOSoFt, what's the point in that?

[13:10:04.0000] <KaOSoFt>
That'll trick the web browser into making another thread for downloading those other files.

[13:10:05.0000] <AryehGregor>
Other than a few more parallel requests, I guess.

[13:10:06.0000] <KaOSoFt>
That.

[13:10:07.0000] <AryehGregor>
Well, there's not much point in putting *all* included content there.

[13:10:08.0000] <AryehGregor>
If you want to use that hack, you have to split them among several subdomains.

[13:10:09.0000] <TabAtkins>
AryehGregor: Doing subdomains also lets you more easily tweak compressions/caching/etc.

[13:11:00.0000] <AryehGregor>
Compression/caching works fine for me right now.

[13:11:01.0000] <TabAtkins>
Kk.

[13:11:02.0000] <AryehGregor>
After I fixed some JS not being compressed because of application/x-javascript vs. application/javascript vs. text/javascript . . .

[13:11:03.0000] <KaOSoFt>
If your site needs more than, let's say, 8 simultaneous requests (images, styles, scripts, content, etc.), tricking it that way should work.

[13:12:00.0000] <Philip`>
AryehGregor: The Amazon CDN sounds pretty easy to use

[13:12:01.0000] <AryehGregor>
KaOSoFt, but if all the simultaneous requests are on the same domain, it doesn't help you if it's separate from your main site.

[13:12:02.0000] <AryehGregor>
Philip`, yeah, I'd heard about that.

[13:12:03.0000] <AryehGregor>
I tried SimpleCDN, but it didn't work.

[13:12:04.0000] <TabAtkins>
Amazon is sorta expensive.

[13:12:05.0000] <AryehGregor>
Amazon sounds plausible, but I'm wondering about cost.

[13:12:06.0000] <Philip`>
http://aws.amazon.com/cloudfront/

[13:12:07.0000] <TabAtkins>
My old company used it to serve our actual software, but that's it.

[13:12:08.0000] <Philip`>
"$0.150 per GB – first 10 TB / month data transfer out"

[13:13:00.0000] <TabAtkins>
Oh, that's something different.  I was using s3.

[13:13:01.0000] <KaOSoFt>
Philip`- That's cheap.

[13:13:02.0000] <KaOSoFt>
._.

[13:14:00.0000] <Philip`>
CloudFront is basically a faster geographically-distributed frontend to S3

[13:15:00.0000] <Philip`>
for approximately the same cost per GB

[13:15:01.0000] <AryehGregor>
/me tries to work out if $0.150 per GB is more or less expensive than $50/Mbps at the 95th percentile

[13:15:02.0000] <AryehGregor>
/me gives up, needs a traffic distribution

[13:15:03.0000] <KaOSoFt>
Haha.

[13:15:04.0000] <AryehGregor>
I think this stuff is actually a pretty small part of my bandwidth, though.

[13:15:05.0000] <Philip`>
That seems impossible to calculate, since you might go 0Mbps for 95% of the time and then ~infinity Mbps for 5% of the time

[13:16:00.0000] <AryehGregor>
/me discovers that some site admin or another stuck zillions of ImageShack-hosted images on his front page, sigh

[13:16:01.0000] <TabAtkins>
AryehGregor: link?

[13:16:02.0000] <AryehGregor>
TabAtkins, twcenter.net

[13:17:00.0000] <KaOSoFt>
Are you... are you the webmaster of that site?

[13:17:01.0000] <TabAtkins>
The same site admin doesn't realize that jpegs don't play nicely with text in image.

[13:17:02.0000] <AryehGregor>
KaOSoFt, yes.

[13:17:03.0000] <KaOSoFt>
AryehGregor- I got Empire - Total War because of the mods there.

[13:17:04.0000] <AryehGregor>
I didn't make the front page, though, don't blame me.

[13:17:05.0000] <KaOSoFt>
:D

[13:17:06.0000] <AryehGregor>
KaOSoFt, :)

[13:17:07.0000] <AryehGregor>
I'm the primary webmaster, anyway.

[13:18:00.0000] <AryehGregor>
There are like eight admins, but I do most of the techy stuff.

[13:18:01.0000] <AryehGregor>
Our primary bandwidth drain is, in fact, downloads.

[13:19:00.0000] <KaOSoFt>
TabAtkins- Why are there like, two more nicknames like yours?

[13:19:01.0000] <KaOSoFt>
I mean, here, in the channel.

[13:19:02.0000] <TabAtkins>
KaOSoFt: The one I'm talking as right now is my work desktop.  The other one or two are my laptops.

[13:19:03.0000] <KaOSoFt>
Oh, I see.

[13:20:00.0000] <TabAtkins>
Much easier to just keep irc up at all times than manually launch it whenever I use a laptop.

[13:20:01.0000] <gsnedders>
You should just log in to one server over ssh and run irssi on it

[13:20:02.0000] <TabAtkins>
gsnedders: My current solution requires less work than that.

[13:21:00.0000] <TabAtkins>
And all the nicks are set to hilight for me, so shrug.

[13:23:00.0000] <AryehGregor>
. . . @ http://img23.imageshack.us/img23/2130/50633715.png being a PNG

[13:27:00.0000] <TabAtkins>
What's the closest thing to public domain that has a good chance of being recognized internationally?  BSD? MIT?

[13:28:00.0000] <AryehGregor>
In some jurisdictions, no open-source license will be recognized internationally.

[13:28:01.0000] <KaOSoFt>
MIT is as free as it could be.

[13:28:02.0000] <AryehGregor>
You can use the CC public domain dedication.

[13:32:00.0000] <TabAtkins>
How about the cc0 license?

[13:32:01.0000] <TabAtkins>
http://creativecommons.org/publicdomain/zero/1.0/

[13:33:00.0000] <AryehGregor>
Ugh, my front page is a complete trainwreck.  4.4 MB of content.

[13:33:01.0000] <AryehGregor>
/me gives up on it

[13:33:02.0000] <TabAtkins>
Ooh, http://farm5.static.flickr.com/4024/4499455073_a729acd621_o.png

[13:34:00.0000] <TabAtkins>
^^^ This is why a fast, simple auto-update is ideal.

[13:34:01.0000] <AryehGregor>
You bet.

[13:35:00.0000] <AryehGregor>
Chrome is by far the best at auto-updating.  Firefox is a pain.

[13:35:01.0000] <AryehGregor>
I don't know why people are so paranoid about silent auto-update.

[13:35:02.0000] <Dashiva>
Because when stuff breaks, users get mad

[13:36:00.0000] <AryehGregor>
/me regularly gets nagged on his Linux machine at school to update Firefox . . . although probably can't, since you know, he's not root?

[13:36:01.0000] <AryehGregor>
Dashiva, easy solution: don't break stuff on upgrade.

[13:36:02.0000] <AryehGregor>
(Chrome breaks too often for me, but that's probably because I'm on the dev channel and too lazy to switch to beta)

[13:36:03.0000] <Dashiva>
Firefox in particular has that entire forest of extensions

[13:37:00.0000] <AryehGregor>
Yes, and then it prompts you *synchronously* every time you start whether you want to check for updates.

[13:38:00.0000] <AryehGregor>
STFU, I don't care.

[13:38:01.0000] <AryehGregor>
Oh, but you're saying they might be broken by upgrade.

[13:38:02.0000] <TabAtkins>
That synchronous checking is really quite annoying.

[13:38:03.0000] <AryehGregor>
Well, that's because Firefox extensions are much more powerful than Chrome extensions.

[13:38:04.0000] <AryehGregor>
Comes at a cost.

[13:38:05.0000] <AryehGregor>
I guess.

[13:38:06.0000] <TabAtkins>
I've often started FF, gone to do something else (because with all my extensions, starting FF takes 3 or 4 minutes), only to come back to it saying an extension has been updated.

[13:39:00.0000] <AryehGregor>
My impression is that Firefox gives too much power to extensions, so they wind up being must-haves, but slowing the browser down, breaking on upgrade . . .

[13:39:01.0000] <AryehGregor>
Chrome extensions seem to be just web pages, which have a few magic powers that use narrow and well-defined APIs that aren't likely to break across upgrade.

[13:39:02.0000] <AryehGregor>
I think Firefox is going that way with Jetpacks or whatever they're called?  Dunno, I don't actually have any idea what I'm talking about.  :D

[13:40:00.0000] <AryehGregor>
But I found that installing lots of extensions on Firefox is a bad idea, in the end.

[13:40:01.0000] <AryehGregor>
Too much pain.

[13:41:00.0000] <TabAtkins>
That's approximately right.

[13:43:00.0000] <Dashiva>
In particular ones that update frequently...

[13:44:00.0000] <AryehGregor>
I'm also suspicious of the performance impact.

[13:47:00.0000] <KaOSoFt>
Chrome is quite fast, and it's got this eerie smoothness that I like.

[13:47:01.0000] <KaOSoFt>
But I think it lacks some functionality.

[13:47:02.0000] <AryehGregor>
Works well enough for me.

[13:47:03.0000] <KaOSoFt>
I'm using it right now.

[13:47:04.0000] <KaOSoFt>
This extra functionality doesn't do me any good at work.

[13:48:00.0000] <KaOSoFt>
So I don't need it.

[13:49:00.0000] <AryehGregor>
Doesn't do me any good at home either.

[13:50:00.0000] <AryehGregor>
Sadly, work (i.e., school) runs RHEL5, so it's hard to install Chrome.

[13:50:01.0000] <AryehGregor>
I haven't bothered trying.

[13:50:02.0000] <AryehGregor>
Libraries too outdated . . .

[13:50:03.0000] <TabAtkins>
Only two things I've found that I miss in Chrome: the ability to actually view the source that a page came in with ("View Source" makes a new request, which screws up some things) and the ability to get at image information.

[13:51:00.0000] <TabAtkins>
/me supposes he should file bugs about those.

[13:51:01.0000] <AryehGregor>
Yeah, those two are mildly annoying.

[13:51:02.0000] <gsnedders>
How does that work with POST?

[13:51:03.0000] <TabAtkins>
Badly.

[13:51:04.0000] <AryehGregor>
gsnedders, it doesn't, that's the problem.

[13:51:05.0000] <miketaylr>
it doesn't

[13:51:06.0000] <gsnedders>
Does it still make another request?

[13:51:07.0000] <gsnedders>
Or does it just not give you view sourcE?

[13:51:08.0000] <TabAtkins>
Yes, but as GET, I think.

[13:51:09.0000] <AryehGregor>
You can at least get image resolution if you open it in a new tab, I noticed.

[13:51:10.0000] <gsnedders>
Oh dear

[13:51:11.0000] <AryehGregor>
It displays in the title.

[13:51:12.0000] <AryehGregor>
If you hover over the tab title.

[13:51:13.0000] <TabAtkins>
I know that it screws up the CSS module post-processor, so I have to use FF when I'm editting a spec.

[13:51:14.0000] <AryehGregor>
Discoverability FTW.

[14:17:00.0000] <KaOSoFt>
Today I leave early, though there is a storm out there... Oh, well, goodbye anyways.

[14:45:00.0000] <JonathanNeal>
All I need is a style for the page.

[15:23:00.0000] <JonathanNeal>
Anyone care to review http://www.iecss.com/print-protector/ let me know if I missed anything?

[15:25:00.0000] <Philip`>
JonathanNeal: The ri kerning in the title image is ugly :-(

[15:26:00.0000] <JonathanNeal>
Noted, what font should I use?

[15:27:00.0000] <Philip`>
Comic Sans?

[15:27:01.0000] <Philip`>
Can you stop calling it a shiv, and call it something like a shim instead, since that would make much more sense?

[15:30:00.0000] <JonathanNeal>
Philip`, is it really that important?  http://ejohn.org/blog/html5-shiv/ http://remysharp.com/2009/01/07/html5-enabling-script/

[15:30:01.0000] <Hixie>
a shiv is a weapon

[15:30:02.0000] <Hixie>
a shim is a thin strip of material to make things align

[15:31:00.0000] <JonathanNeal>
The term shiv originates from John Resig, the creator of jQuery, who used the word as a metaphor to describe shiving support for HTML5 elements into Internet Explorer via javascript, where shiving  is a slang term for a sharp object used as a knife-like weapon. There's also a chance that John had a moment of false etymology and meant to use the word shim, which in computing means an application compatibility workaround.

[15:31:01.0000] <TabAtkins>
I'm going with the second possibility, which makes much more sense.

[15:32:00.0000] <Hixie>
you could use the word ship, claiming that you are "providing a way to get there from here"

[15:32:01.0000] <JonathanNeal>
Well, I don't think I'll be offending any SHIV people by changing the wording to SHIM --- done!

[15:32:02.0000] <jgraham>
You could use the word "hack" instead and then spend lots of time explaining to your grandparents that NSA aren't after you

[15:33:00.0000] <Hixie>
or or "shin", and say that you are giving IE legs to stand on

[15:33:01.0000] <Philip`>
Replacing things like <video> and <canvas> with a <div> on printing seems like a terrible idea, because it won't do anything useful in IE8 and it'll completely break printing of the elements in IE9

[15:33:02.0000] <Hixie>
or "shia" and claim it to be a metaphor for forking standards...

[15:34:00.0000] <Philip`>
(Same for all the other elements that have special not-purely-CSS renderings, too)

[15:34:01.0000] <TabAtkins>
Ooh, good point.  You need a cc targeting IE8 and earlier only.

[15:36:00.0000] <Philip`>
I think IE9 parses unknown elements in an XMLish way so you don't actually need to do anything there

[15:36:01.0000] <JonathanNeal>
TabAtkins, I will do that.

[15:56:00.0000] <JonathanNeal>
Okay, it's all updated.

[16:26:00.0000] <JonathanNeal>
s'all good?

[16:31:00.0000] <othermaciej>
MikeSmith: I'm around if you want to chat at all

[16:32:00.0000] <TabAtkins>
Very interesting with the newer updates, JonathanNeal.

[16:35:00.0000] <TabAtkins>
JonathanNeal: Does IE ignore media=print on stylesheets or something?

[16:36:00.0000] <JonathanNeal>
What do you mean?

[16:37:00.0000] <nessy>
othermaciej: now sent the change proposals email for media a11y - let's see what happens

[16:37:01.0000] <TabAtkins>
You have the _exposePrint and _concealPrint functions, which just turn <style media=print> into <style media=screen old_media=print>, etc.

[16:44:00.0000] <JonathanNeal>
Well, that's because I replace html5 elements with div/span

[16:44:01.0000] <JonathanNeal>
so exposePrint and consealPrint temporarily change the stylesheets so I can grab the correct styling.

[16:44:02.0000] <JonathanNeal>
Then it creates a custom stylesheet that styles the new div / span elements.

[16:45:00.0000] <TabAtkins>
Oh, now I get it.  You're switching them to screen just so you can analyze the styles *while* it's in screen mode.

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

[16:45:02.0000] <TabAtkins>
Can you use CSSOM to do that more cleanly?  (Dunno what the level of support is in IE6+.

[16:45:03.0000] <JonathanNeal>
I've never heard of that.

[16:46:00.0000] <TabAtkins>
Search for CSS Object Model.

[16:46:01.0000] <JonathanNeal>
Yea, I pulled up some spec from the w3 but I don't ... know how I would use it.

[16:47:00.0000] <TabAtkins>
Grab the appropriate stylesheet, iterate through and read out the styles, producing a new stylesheet as you go along with your substitutions.

[16:47:01.0000] <TabAtkins>
Rather than reading off the styles of individual elements, you'd change selectors that reference the current element names to use, I dunno, a unique class or something.

[16:48:00.0000] <JonathanNeal>
Ah!  That's a good idea.

[16:48:01.0000] <JonathanNeal>
http://www.quirksmode.org/dom/w3c_cssom.html will this page tell you whether IE6 supports it?

[16:48:02.0000] <JonathanNeal>
While I understand what you're saying ... I haven't seen the syntax yet to do what you described.

[16:49:00.0000] <TabAtkins>
No, that's just for CSSOM View.

[16:52:00.0000] <JonathanNeal>
I am just failing at the right google result to see how to do this.

[16:53:00.0000] <JonathanNeal>
http://www.quirksmode.org/dom/w3c_css.html#access ?

[16:54:00.0000] <TabAtkins>
Yes, that's it.


2010-04-10
[17:05:00.0000] <othermaciej>
nessy: we shall see! I think things will be fine

[17:06:00.0000] <nessy>
I look forward to the further discussion!

[17:10:00.0000] <JonathanNeal>
TabAtkins, all right, I'm halfway through the new parser.

[17:38:00.0000] <TabAtkins>
Woo, first pass of new table-layout algorithm, custom-designed for Template Layout. www.xanthir.com/:tm7

[17:41:00.0000] <Hixie>
JonathanNeal: btw it's spelt conceal, not conseal :-)

[17:41:01.0000] <JonathanNeal>
Shame on me.

[17:41:02.0000] <JonathanNeal>
:)

[17:42:00.0000] <estellevw>
JonathanNeal: did you post yet?

[17:43:00.0000] <JonathanNeal>
I posted the first version @ http://www.iecss.com/print-protector/

[17:43:01.0000] <estellevw>
cool

[17:43:02.0000] <estellevw>
now i can tweet it

[17:43:03.0000] <JonathanNeal>
but the next version uses an idea TabAtkins pitched where I parse through the stylesheets and generate a secondary HTML5 css file that works with classnames.

[17:43:04.0000] <JonathanNeal>
Wellllll I dunno if we wanna tweet it yet, do we?

[17:59:00.0000] <Philip`>
JonathanNeal: Since you've mentioned it on IRC, surely it'll spread onto and around the Twitsphere by itself, so it's too late to try to stop it now

[18:00:00.0000] <JonathanNeal>
Yeap.

[18:00:01.0000] <JonathanNeal>
Can't I make your glass half full ever, Philip`? :)

[18:01:00.0000] <Philip`>
I don't expect you can fill that precisely - it'll always be slightly more or slightly less than half full

[18:02:00.0000] <paul_irish>
:) JonathanNeal i think its a good time to start leaking out

[18:02:01.0000] <JonathanNeal>
Yea go for it!

[18:14:00.0000] <JonathanNeal>
I finished the new version.

[18:17:00.0000] <JonathanNeal>
http://www.iecss.com/print-protector/example_print_iepp_1-2-0.html

[18:18:00.0000] <TabAtkins>
/me will try to print it when he gets home.

[18:19:00.0000] <JonathanNeal>
TabAtkins, you can see the new code @ http://www.iecss.com/print-protector/javascript/iepp.1-2-0.js

[18:48:00.0000] <boblet>
JonathanNeal: nice work on iepp!

[18:50:00.0000] <boblet>
“is piece of javascript” → “is a piece of javascript”

[18:51:00.0000] <JonathanNeal>
Good catch, updated.

[18:51:01.0000] <JonathanNeal>
boblet, you have a copy of ie?

[18:52:00.0000] <boblet>
yes, but in a VM that is broken :/

[18:52:01.0000] <boblet>
so I guess that’s kinda no huh

[18:53:00.0000] <boblet>
/me fails at replying

[19:02:00.0000] <JonathanNeal>
:)

[19:13:00.0000] <MikeSmith>
Hixie: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9472 (deal with broken links in dynamic author view of spec by prompting users to choose to switch to full view)

[19:13:01.0000] <MikeSmith>
from discussion last monday

[19:14:00.0000] <MikeSmith>
I didn't manage to get around to raising it til now

[19:16:00.0000] <MikeSmith>
btw, I notice something else while looking at a validator.nu bug about option/@type=selected descendants of select/@multiple elements

[19:16:01.0000] <MikeSmith>
http://bugzilla.validator.nu/show_bug.cgi?id=734

[19:17:00.0000] <MikeSmith>
er, not type=selected of course, but just selected

[19:18:00.0000] <MikeSmith>
anyway, I can't find anywhere in the spec that states the conformance constraint that an option/@selected can only occur as a descendant of select/@multiple

[19:20:00.0000] <MikeSmith>
in contrast, the section on the file-upload state for the input element has a paragraph explicitly marked with a "Constraint validation" subhead that says, "Unless the multiple attribute is set, there must be no more than one file in the list of selected files"

[19:20:01.0000] <MikeSmith>
http://dev.w3.org/html5/spec/forms.html#file-upload-state

[19:33:00.0000] <MikeSmith>
er, I guess the "Constraint validation" part is not relevant here, but the paragraph I cited that follows it is

[19:35:00.0000] <MikeSmith>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9473

[19:38:00.0000] <JonathanNeal>
Okay! http://www.iecss.com/print-protector/ --- updated!  Happy weekend, everyone!

[19:44:00.0000] <gsnedders>
/me waves

[19:44:01.0000] <MikeSmith>
gsnedders: hey

[19:44:02.0000] <gsnedders>
MikeSmith: I bought The Third Policeman, BTW

[19:45:00.0000] <MikeSmith>
are you waving goodbye, hello, or waving for help?

[19:45:01.0000] <MikeSmith>
gsnedders: ah, great book

[19:45:02.0000] <gsnedders>
waving hello

[19:45:03.0000] <MikeSmith>
well, I envy you for being able to read The Third Policeman for the first time

[19:47:00.0000] <gsnedders>
/me is going to end up writing a terrible blog post that he'll scrap, again...

[19:50:00.0000] <gsnedders>
/me rambles on endlessly

[19:50:01.0000] <MikeSmith>
publish it under an assumed name

[19:50:02.0000] <gsnedders>
That won't help. It's still nonsense.

[19:50:03.0000] <MikeSmith>
e.g. Geoffrey Snedders O'Brien

[19:50:04.0000] <MikeSmith>
that'll throw people off

[19:51:00.0000] <gsnedders>
/me still wants to finally start the art project he's had in his mind for a while

[19:52:00.0000] <gsnedders>
But everytime I think about it, I think of another way I have to conseal my identity

[19:52:01.0000] <gsnedders>
Like, to properly have an anon blog is a load of work

[19:57:00.0000] <MikeSmith>
gsnedders: here's another idea - just start a "The Fake Geoffrey Snedders Blog" and write stuff that would fit the stereotype that others might have of your private thoughts

[19:58:00.0000] <MikeSmith>
people would read it and chuckle about it and say, Man, whoever this guy is, he's got Geoffrey pegged

[19:59:00.0000] <gsnedders>
:P

[19:59:01.0000] <gsnedders>
So, basically, act emo?

[20:02:00.0000] <MikeSmith>
act like you're somebody who's not emo but who's trying to act emo 

[20:02:01.0000] <gsnedders>
Man, that's a lot of acting

[20:04:00.0000] <MikeSmith>
well, it's only acting like you're not yourself but actually just being yourself

[20:05:00.0000] <MikeSmith>
a celebration of yourself

[20:05:01.0000] <gsnedders>
But that wouldn't be matching the stereotype

[20:05:02.0000] <MikeSmith>
well, that's why it's the Fake Geoffrey Snedders Blog

[20:06:00.0000] <MikeSmith>
it's a chance for you to explore another side of yourself

[20:06:01.0000] <MikeSmith>
the self-parody side

[20:06:02.0000] <gsnedders>
The wreckless, fearless, uncompromising one?

[20:06:03.0000] <MikeSmith>
The Third Policeman will give you some inspiration

[20:06:04.0000] <MikeSmith>
gsnedders: the laughing-through-the-tears side

[20:07:00.0000] <gsnedders>
Like I don't do that already.

[20:07:01.0000] <gsnedders>
I mean, all the world's a stage.

[20:07:02.0000] <MikeSmith>
yep

[20:07:03.0000] <MikeSmith>
and you are the star

[20:07:04.0000] <MikeSmith>
the tragic hero

[20:08:00.0000] <MikeSmith>
tragic comedy

[20:08:01.0000] <MikeSmith>
celebrate yourself in the manner of Walt Whitman

[20:08:02.0000] <MikeSmith>
except with humor like Flann O'Brien

[20:08:03.0000] <JonathanNeal>
heyo, hey MikeSmith :)

[20:09:00.0000] <MikeSmith>
JonathanNeal: word up

[20:09:01.0000] <gsnedders>
Someday I'll just have to scream out in here that I'm married.

[20:09:02.0000] <gsnedders>
And that I'll start being euphoric all the time

[20:09:03.0000] <JonathanNeal>
...

[20:09:04.0000] <gsnedders>
And get rid of this entire persona people have of me.

[20:09:05.0000] <JonathanNeal>
gsnedders, you okay?

[20:10:00.0000] <gsnedders>
JonathanNeal: Yeah, this is what happens when you enter mid-conversation

[20:11:00.0000] <gsnedders>
I mean, if I posted that on the Fake gsnedders Blog it'd blatantly be fake.

[20:49:00.0000] <wirepair>
are there any plans to reserve a port for websockets? or is it up to the developer to choose?

[20:50:00.0000] <wirepair>
or only on the usual http/https ports?

[20:51:00.0000] <wirepair>
nm, it's in the ietf draft ;>

[04:59:00.0000] <annevk>
http://google-opensource.blogspot.com/2010/04/interesting-times-for-video-on-web.html

[05:07:00.0000] <Philip`>
Summary for lazy people: Google is helping fund an ARM code optimised version of the Theora video codec

[05:13:00.0000] <annevk>
ah great, saves me some reading

[05:13:01.0000] <annevk>
:p

[05:21:00.0000] <KaOSoFt>
Good morning.

[08:45:00.0000] <mikelifeguard>
What doctype(s) must I use for a page including <video> to validate?

[08:49:00.0000] <annevk>
<!doctype html>

[08:55:00.0000] <mikelifeguard>
What'll happen to old or ancient browsers that see the html5 doctype?

[08:56:00.0000] <annevk>
nothing

[09:13:00.0000] <divya>
mikelifeguard: http://www.w3.org/TR/html5-diff/#doctype

[11:13:00.0000] <estellevw>
good morning everyone

[11:15:00.0000] <KaOSoFt>
Good morning.


2010-04-11
[00:40:00.0000] <estellevw>
There were five global attributes related to microdata including itemid, itemprop, itemref, itemscope and itemtype that i remember seeing, but don't see in the spec right now. Am I missing something?

[00:46:00.0000] <myakura>
estellevw: they got split off from the main HTML5 spec about a month (or two) ago.

[00:46:01.0000] <estellevw>
thanks

[00:47:00.0000] <estellevw>
is it likely to come back in?

[00:48:00.0000] <myakura>
I guess the WHATWG version of HTML5 still incorporates it.

[00:48:01.0000] <myakura>
for W3C version, see http://www.w3.org/TR/microdata/ . 

[00:49:00.0000] <estellevw>
yeah, looking at it

[00:49:01.0000] <estellevw>
i think it needs some editing. The attributes are in the body but not in the table of contents

[00:51:00.0000] <estellevw>
http://www.w3.org/TR/microdata/#attr-itemref and the like have lost their original stature.

[00:52:00.0000] <estellevw>
it states The following attributes are added as global attributes to HTML elements:

[00:52:01.0000] <estellevw>
    * itemid

[00:52:02.0000] <estellevw>
    * itemprop

[00:52:03.0000] <estellevw>
    * itemref

[00:52:04.0000] <estellevw>
    * itemscope

[00:52:05.0000] <estellevw>
    * itemtype

[00:52:06.0000] <estellevw>
oops, sorry

[00:52:07.0000] <estellevw>
but if i recall correctly, those aren't listed in teh global attributes anymore

[00:53:00.0000] <estellevw>
the global attributes being here: http://www.w3.org/TR/html5/dom.html#global-attributes

[00:56:00.0000] <myakura>
because they are two separete specs; Microdata is build on top of HTML5 so those item* attributes cannot be defined in HTML5.

[00:59:00.0000] <estellevw>
ok, makes sense as to why role, and the aria-* attributes aren't listed as global even though they are too

[00:59:01.0000] <estellevw>
thanks

[00:59:02.0000] <myakura>
yeah..

[01:00:00.0000] <myakura>
just refer to the WHATWG version and you won't get confused :)

[01:00:01.0000] <myakura>
MikeSmith: aloha

[01:01:00.0000] <estellevw>
i can still get confused, but i'll have to blame myself ;)

[04:28:00.0000] <MikeSmith>
myakura: here now

[04:29:00.0000] <MikeSmith>
hsivonen: if/when you are around and have time to chat, please ping me

[04:29:01.0000] <MikeSmith>
in regard to http://dev.w3.org/html5/spec/text-level-semantics.html#guidance-for-conformance-checkers

[04:42:00.0000] <MikeSmith>
writing code for a conformance checker to check "The img element is part of the only paragraph directly in its section" or even "only non-whitespace content in the only paragraph directly in its section" is not easy

[04:45:00.0000] <annevk>
prolly also depends on how you implement things

[04:46:00.0000] <MikeSmith>
well, it's not practical at all to implement using a grammar-based schema, so we forget about that completely

[04:48:00.0000] <MikeSmith>
and it's quite complicated to implement in the Java code that validator.nu currently uses for things that can't be checked practically using grammar-based checking

[04:50:00.0000] <MikeSmith>
as far as I can see, it will require adding code that is very unlike existing code for any other checking that is being done by validator.nu

[04:58:00.0000] <othermaciej>
MikeSmith: I think that particular exception is not very well justified in the first place

[04:58:01.0000] <othermaciej>
MikeSmith: but yeah, to check an "only paragraph in its section" condition would require running the HTML5 outline algorithm I think

[04:59:00.0000] <MikeSmith>
It would be help to have the spec provide the intended rationale for that exception

[05:00:00.0000] <othermaciej>
MikeSmith: Hixie explained it a bit in the specific bug Laura filed about it

[05:00:01.0000] <MikeSmith>
ok

[05:00:02.0000] <othermaciej>
let me see if I can find it

[05:01:00.0000] <othermaciej>
MikeSmith: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9217

[05:02:00.0000] <MikeSmith>
I see

[05:02:01.0000] <othermaciej>
MikeSmith: I would personally think it was fine to say "just use figure"

[05:02:02.0000] <MikeSmith>
well, I think we could all live without that particular exception

[05:02:03.0000] <MikeSmith>
yeah, what you said

[05:03:00.0000] <othermaciej>
because associating the relevant section header with the image is much more complicated

[05:03:01.0000] <othermaciej>
as much as you don't want to implement it in the validator, I *really* don't want to implement the outline algorithm solely for our accessibility code to handle this case

[05:04:00.0000] <MikeSmith>
yep

[05:06:00.0000] <othermaciej>
I'm adding a comment to the bug

[05:06:01.0000] <MikeSmith>
OK

[05:10:00.0000] <othermaciej>
done

[05:11:00.0000] <othermaciej>
I mentioned the validator issue but feel free to comment as well

[05:12:00.0000] <othermaciej>
the rate of incoming bugs is just ridiculous

[05:12:01.0000] <othermaciej>
can't believe we are up to 90 already, there were about 42 just a few days ago

[05:13:00.0000] <MikeSmith>
well, in some ways, incoming bugs is a sign of progress

[05:13:01.0000] <othermaciej>
it's a sign of people reviewing the spec

[05:13:02.0000] <MikeSmith>
e.g., with the media-accessibility bugs that Silvia raised

[05:13:03.0000] <othermaciej>
which is good

[05:13:04.0000] <MikeSmith>
true that

[05:14:00.0000] <othermaciej>
I don't feel bad about incoming bugs, as long as the rate of outgoing bugs is also high

[05:47:00.0000] <myakura>
MikeSmith: I talked with Shiraishi-san on Friday about what we're gonna talk in Fukuoka.

[05:48:00.0000] <MikeSmith>
OK

[05:50:00.0000] <myakura>
MikeSmith: and their plan is to: Takuya will do the keynote-like talk (.5hr), and You and I do about Web Standards in general (like what you did at DevFest)

[05:50:01.0000] <MikeSmith>
sounds good so far

[05:51:00.0000] <myakura>
MikeSmith: Oli for markup stuff, Shiraishi-san will talk about APIs, and Anne to do about "CSS latest status"

[05:51:01.0000] <myakura>
and I'm not sure you guys heard about that

[05:51:02.0000] <myakura>
and whether you guys are okay with that.

[05:52:00.0000] <boblet>
oh, good timing

[05:52:01.0000] <boblet>
hey all

[05:54:00.0000] <myakura>
boblet: heya.

[05:54:01.0000] <boblet>
myakura: by markup stuff do you mean sectioning elements etc?

[05:54:02.0000] <boblet>
or also CSS?

[05:55:00.0000] <myakura>
MikeSmith: btw we have an hour for each session but for yours and Anne's I need to interpret so you guys will have shorter time (40min?)

[05:56:00.0000] <MikeSmith>
myakura: that sounds fine

[05:56:01.0000] <myakura>
boblet: the table only shows "HTML5 Mark up" ...

[06:00:00.0000] <boblet>
myakura: will check it out. thanks for the info. Guess I better read Mike’s slides

[09:34:00.0000] <jgraham>
It seems like AT would want access to the document outline in any case

[09:36:00.0000] <jgraham>
And presumably webkit will have to implment it anyway once we get :section(n)

[09:37:00.0000] <annevk>
if we ever decide to do that

[09:37:01.0000] <jgraham>
Well sure

[09:37:02.0000] <jgraham>
If we don't decide to do that the whole outline algorithm is likely a waste of time

[10:32:00.0000] <micheil>
hmm.. would this be the place to ask a quick question about one of your spec proposals?

[10:32:01.0000] <jgraham>
Yes

[10:33:00.0000] <micheil>
okay, with the websocket protocol, would I be right in assuming that one server would be able to server multiple different sockets / data sets to clients based on the paths at which they connect?

[10:34:00.0000] <jgraham>
With the proviso I am not an expert on that spec, yes

[10:34:01.0000] <micheil>
okay

[10:34:02.0000] <jgraham>
The server can do whatever it wants based on anything the client sends including the path part

[10:35:00.0000] <jgraham>
(subject to it meeting the requirements for a successful connection of course)

[10:36:00.0000] <micheil>
yeah, I'm just trying to work out how to direct data about by using that, all the reference implementations I can find don't indicate anything on that

[10:38:00.0000] <jgraham>
Well when the server recieves the client handshake it parses out the path so you can arrange for it to be passed to the app

[10:38:01.0000] <jgraham>
Not sure what existing implementations do though

[10:39:00.0000] <micheil>
yeah, just trying to think if it's possible to send data to a specific connection or not

[10:39:01.0000] <micheil>
(ie, rather then just being a broadcast type service)

[10:40:00.0000] <micheil>
I suppose because the writes to the network stream would be done from within a connection, you'd handle it there maybe..

[10:41:00.0000] <jgraham>
If I follow you, that sounds right

[10:42:00.0000] <jgraham>
From the point of view of the server each client is a seperate connection so you can read and write to each client independently

[10:43:00.0000] <micheil>
hopefully

[10:43:01.0000] <jgraham>
I don't really see how it could work otherwise

[10:44:00.0000] <micheil>
I suppose I should just work out the raw tcp send switching first, then add on the websocket protocol

[10:46:00.0000] <MikeSmith>
jgraham: conformance checkers should not require access to the document outline

[10:47:00.0000] <jgraham>
MikeSmith: I didn't mean to have any opinion on conformance checkers

[10:47:01.0000] <jgraham>
Just other types of UA

[10:47:02.0000] <MikeSmith>
I see

[10:47:03.0000] <MikeSmith>
I can see that AT having access to the document outline would be good

[10:48:00.0000] <MikeSmith>
but then there are many things that AT should be doing that they are not currently

[10:49:00.0000] <jgraham>
Right, but I disagree with othermaciej's assertion that implementing the outline algorithm in the accessibility code would soley be for this case

[10:53:00.0000] <MikeSmith>
yeah, it'd certainly seem it could end j useful for

[10:54:00.0000] <MikeSmith>
.. end up being useful for lth

[10:54:01.0000] <MikeSmith>
oops

[10:54:02.0000] <MikeSmith>
...useful for other things

[10:54:03.0000] <MikeSmith>
(my fingers are cold)

[10:55:00.0000] <jgraham>
ah, I was imagining that you had insane keyboard macros that could expand a few characters into whole sentences

[10:55:01.0000] <jgraham>
and they had broken

[10:55:02.0000] <MikeSmith>
heh

[10:59:00.0000] <MikeSmith>
I just came back from sento, in 水風呂 water that's 17.1 degrees

[11:04:00.0000] <jgraham>
/me assumes that is the kanji for "fucking cold"

[11:05:00.0000] <micheil>
it'd have to be less then 17 degrees outside here

[11:06:00.0000] <jgraham>
Well sure it is less than 17 degrees C outside here too

[11:06:01.0000] <jgraham>
But I wouldn't go in water that temperature

[11:06:02.0000] <jgraham>
The thermal conduction is a killer

[11:07:00.0000] <micheil>
actually. 20ºC here is cold.

[11:08:00.0000] <AryehGregor>
It's barely above 20°C here now, and it's starting to be summer already.

[11:09:00.0000] <micheil>
jgraham: I think the key to my websocket problem is in the http parsers (I'm working with node.js btw)

[11:10:00.0000] <micheil>
it's not even winter yet and it's 7.6ºC outside.

[11:10:01.0000] <micheil>
then again, I'm usually getting 30-45º in summer.

[11:17:00.0000] <MikeSmith>
micheil: I think maybe websocket isn't meant to be used with existing http parsers

[11:17:01.0000] <micheil>
MikeSmith: it isn't. however, the key to building a robust node.js websocket server implementation lies in how http servers work

[11:18:00.0000] <MikeSmith>
I'm not familiar with node.js or what its main use cases are

[11:19:00.0000] <micheil>
async / evented javascript with network and file i/o

[11:19:01.0000] <micheil>
built on the same javascript engine that powers chrome / chromium

[11:38:00.0000] <jgraham>
micheil: Do you get raw socket access with node.js? 

[11:38:01.0000] <jgraham>
If you do I would start from there rather than with the existing http parser

[11:38:02.0000] <micheil>
yeah, raw tcp / sockets

[11:39:00.0000] <micheil>
although, node's http parser is built on it's net sockets, so I can use that as reference

[11:45:00.0000] <jgraham>
micheil: You might be just as well working from the simple echo server example in the documentation

[11:46:00.0000] <micheil>
not realy

[11:46:01.0000] <micheil>
not everything is documented, so there are some different ways to do things under the hood

[11:46:02.0000] <jgraham>
Or you sould take the existin websocket-in-node.js implementation and check that it supports the new handshake

[11:46:03.0000] <jgraham>
*could

[11:46:04.0000] <micheil>
(I wish the current net implementation was around when I started working on node-smtp-client)

[11:47:00.0000] <micheil>
jgraham: nawh, that takes away half the fun :P

[11:47:01.0000] <jgraham>
/me should finish his websocket-server-in-python-diesel implementation

[11:48:00.0000] <micheil>
heh heh, what I'm wanting to get from this is pretty specific

[11:55:00.0000] <AryehGregor>
Hmm, so Google is funding efficient ARM decoding of Theora.  I still have hope that they'll switch YouTube from H.264 someday.

[11:59:00.0000] <jgraham>
AryehGregor: Well I don't doubt they will. They question is will it be to theora and soon or to H.264.next and not-so-soon

[11:59:01.0000] <AryehGregor>
Heh.

[11:59:02.0000] <JonathanNeal>
updated ie print protector; added summary element, wrapped entire script in ie conditional (instead of individual fns), and compressed window/document variables (all based on remy sharp suggestions)

[12:10:00.0000] <micheil>
hmm.. due to the way the websocket protocol is, I'm not likely to need much of a really heavy parser for the handshake (initial GET request), am I?

[12:11:00.0000] <jgraham>
micheil: You need to parse out the headers to get the websocket-sec-key (or whatever they are called) values

[12:12:00.0000] <jgraham>
And you need the random bytes

[12:12:01.0000] <micheil>
yeah, but I want need the same level of parser as what a standard http server would

[12:12:02.0000] <jgraham>
and there are some requirements about when you must drop the connection

[12:12:03.0000] <micheil>
hmm.. random bytes..

[12:12:04.0000] <micheil>
/me checks it

[12:12:05.0000] <jgraham>
But no, a full HTTP stack isn't necessary

[12:13:00.0000] <micheil>
hmm..

[12:13:01.0000] <micheil>
random bytes as in the ^n:ds[4u from the spec?

[12:14:00.0000] <jgraham>
The 8 bytes after the end of the part that looks like HTTP headers

[12:14:01.0000] <micheil>
which, according to http spec should probably be: ...headers..\r\n\r\nDATA\r\n\r\n

[12:14:02.0000] <micheil>
I think

[12:15:00.0000] <jgraham>
The HTTP spec isn't really relevant

[12:15:01.0000] <jgraham>
From the point of view of the HTTP spec they are the first 8 bytes of the body or so, I think

[12:19:00.0000] <micheil>
yeah, although, I'm just working out what I'll need to do to parser it

[12:36:00.0000] <Hixie>
simplest answer to that is to follow the rules in the parser section that tell you how to write the parser :-)

[12:36:01.0000] <Hixie>
section 5.1

[12:36:02.0000] <micheil>
Hixie: heh, I suppose you'd be the man to ask :P

[12:37:00.0000] <micheil>
Hixie: is there any sane way that you read those documents?

[12:40:00.0000] <Hixie>
how do you mean?

[12:47:00.0000] <micheil>
well, there seems to be a lot of extra formatting on the IEFT / RFC type documents, is there a cleaner way to view them?

[13:40:00.0000] <jgraham>
micheil: If you can use the complete.html spec on the WHATWG site. It is a bitch to load but once it is loaded it is rather good to read

[14:23:00.0000] <annevk>
http://adactio.com/journal/1654/ -- I still think we should drop <article> and get <content>

[14:24:00.0000] <annevk>
though maybe just dropping <article> for now and see what patterns emerge

[14:40:00.0000] <Hixie>
annevk: i don't understand why people are so confused by them... they're completely differnet

[14:40:01.0000] <Hixie>
one is for chapters and the other is for syndicatable content

[14:40:02.0000] <Hixie>
they're different use cases with almost no overlap

[14:42:00.0000] <Dashiva>
Hixie: The confusion seems to be with the descriptions, maybe it's just a matter of condensing it down to a better explanation

[14:47:00.0000] <Hixie>
the descriptions are completely different

[14:47:01.0000] <Hixie>
i think the problem is that the element names aren't intuitive based on the descriptions

[14:48:00.0000] <Hixie>
but that's not a big problem, people will learn the difference in due course

[14:49:00.0000] <Hixie>
it's like <dl>, <ul>, and <ol>. Given just those element names and then descriptions of what they're for, how would you guess which was which?

[14:50:00.0000] <annevk>
hmm, would've been good to test that out

[14:51:00.0000] <othermaciej>
<article> and <section> as names sound very distinct

[14:51:01.0000] <othermaciej>
maybe the problem is with the descriptions

[14:54:00.0000] <othermaciej>
"The section element represents a generic document or application section." --> ""The section element represents a generic section of a document or application"

[14:54:01.0000] <othermaciej>
I bet that by itself would reduce confusion

[14:54:02.0000] <othermaciej>
if you skim fast, the current text looks like "The ____ element represents a document"

[14:55:00.0000] <othermaciej>
"The article element represents a component of a page that consists of a self-contained composition in a document, page, application, or site and that is intended to be independently distributable or reusable, e.g. in syndication."

[14:55:01.0000] <othermaciej>
and that one reads as "The ____ element represents a component" if you skim

[14:56:00.0000] <othermaciej>
could be something like "The article element represents a self-contained composition in a document, page, application, or site that is intended to be independently distributable or reusable, e.g. in syndication."

[14:56:01.0000] <othermaciej>
Hixie: ^

[14:56:02.0000] <othermaciej>
(annevk also)


2010-04-12
[17:13:00.0000] <Hixie>
othermaciej: please file bugs, i'd be happy to move words around if it helps

[17:13:01.0000] <othermaciej>
Hixie: will do

[17:13:02.0000] <Hixie>
thanks

[17:27:00.0000] <othermaciej>
Hixie: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9491

[17:59:00.0000] <Hixie>
othermaciej: ta

[18:31:00.0000] <jonnybarnes>
Hey guys

[19:00:00.0000] <boblet>
MikeSmith: the wording of the footer element’s example about links in footer has changed in WHATWG spec, so you might want to update nav details box

[19:03:00.0000] <boblet>
s/element’s/element/ (or more precisely the example about links in the footer)

[19:04:00.0000] <boblet>
Hixie: given this change to nav, would the defining characteristic of a group of links that warrants nav still be that they’re important enough to have a title? something else?

[19:05:00.0000] <boblet>
eg, it seems the Exampland wiki footer links could also now be nav, so wondering if this is stylistic or if there’s still a reason they’re not

[19:39:00.0000] <JonathanNeal>
hey all

[19:41:00.0000] <MikeSmith>
JonathanNeal: what it be like

[19:41:01.0000] <JonathanNeal>
it be like this.... yo IE8 is scor'n 18/160 on html5test.com

[19:43:00.0000] <JonathanNeal>
Actually ... more like "hey guys, let's grill hotdogs tonight" "fo sho Jon" and everyone is over.

[19:44:00.0000] <MikeSmith>
who maintains html5test.com ?

[19:44:01.0000] <miketaylr>
it just showed up on twitter today, afaik

[19:45:00.0000] <JonathanNeal>
Niels Niels Leenheer

[19:45:01.0000] <miketaylr>
magically

[19:45:02.0000] <MikeSmith>
sounds like a fake name

[19:46:00.0000] <JonathanNeal>
Well, I accidentally wrote his first name twice, the ills of not copying and pasting.

[19:46:01.0000] <miketaylr>
too bad, Neils Neils is kinda catchy

[19:47:00.0000] <JonathanNeal>
I agree.

[19:48:00.0000] <MikeSmith>
it's an anagram for "See Hell Nine Re"

[19:48:01.0000] <MikeSmith>
which sounds much more like a real name

[19:48:02.0000] <MikeSmith>
HTML5 will unleash the Nine Rays of Hell in 2011

[19:49:00.0000] <miketaylr>
the good news is that 2011 is just around the corner

[19:53:00.0000] <MikeSmith>
one side product of public-html discussions is the innovative quoting non-conventions that people seem to come up with 

[19:53:01.0000] <JonathanNeal>
Like?

[19:54:00.0000] <MikeSmith>
e.g., http://lists.w3.org/Archives/Public/public-html/2010Apr/0348.html

[19:54:01.0000] <MikeSmith>
ME] Yes

[19:54:02.0000] <JonathanNeal>
YOU] I can make you say things now too

[19:57:00.0000] <MikeSmith>
heh

[20:00:00.0000] <JonathanNeal>
YOU] We're releasing HTML5 this summer, so I can spend more time buying new iPhones.

[20:00:01.0000] <JonathanNeal>
ME] Wow!  Great idea, Mike.

[20:02:00.0000] <paul_irish>
MikeSmith: http://twitter.com/rakaz is the html5test guy.  already pinged him about some weak tests.

[20:02:01.0000] <paul_irish>
apparently the site's been out for a month. 

[20:02:02.0000] <paul_irish>
funny how that happens

[20:04:00.0000] <MikeSmith>
it does seem like a useful site I guess

[20:05:00.0000] <MikeSmith>
even though its secret purpose is to quietly prepare for the unleashing of the Nine Rays of Hell 

[20:06:00.0000] <JonathanNeal>
That is the ULTIMATE purpose!

[20:06:01.0000] <MikeSmith>
paul_irish: is dude using modernizer at all?

[20:06:02.0000] <paul_irish>
nah

[20:06:03.0000] <MikeSmith>
modernizr make that

[20:06:04.0000] <MikeSmith>
paul_irish: I wonder why he's not

[20:06:05.0000] <paul_irish>
that findmebyip.com site that's gotten a lot of play does, though.

[20:06:06.0000] <MikeSmith>
ok

[20:07:00.0000] <paul_irish>
theres a few things he's got i'm gonna pick up.. like output/progress/meter elements.. and undoManager

[20:07:01.0000] <miketaylr>
yeah, saw those

[20:08:00.0000] <MikeSmith>
"A “Browser” is a piece of software used to access & display the Internet on your computer. ".. thanks, sensei

[20:08:01.0000] <paul_irish>
miketaylr: im sure you saw him fall into the chrome webforms false positive trap

[20:08:02.0000] <miketaylr>
ITS A TRAAAP

[20:08:03.0000] <miketaylr>
paul_irish: yeah

[20:08:04.0000] <JonathanNeal>
A "Browser" is a giant turtle with spikes who wants to kill Mario ... wait.

[20:09:00.0000] <MikeSmith>
paul_irish: the Forms 2.0 detection on that site seems wacky

[20:10:00.0000] <MikeSmith>
/me compares it with results at miketaylr site

[20:11:00.0000] <paul_irish>
miketaylr has his UI tests that are a bit better

[20:11:01.0000] <JonathanNeal>
paul_irish, what do you guys think of the ie7-js scripts Dean Edwards maintained that would add hover and type="" support to IE?  Do you think you'd ever implement stuff like that into Modernizr?

[20:11:02.0000] <paul_irish>
but they were necessitated because his earlier tests (same as these) did the same falsepositive shit in webkit

[20:12:00.0000] <paul_irish>
JonathanNeal: hover no. i'm not into parsing stylesheets. authors have been working around div:hover in ie6 forever

[20:12:01.0000] <paul_irish>
whats type="" support?

[20:12:02.0000] <miketaylr>
JonathanNeal: whatchu mean by type=""

[20:12:03.0000] <miketaylr>
o. what paul_irish just said

[20:13:00.0000] <JonathanNeal>
Oh, I was just talkin' about the new input types.

[20:13:01.0000] <JonathanNeal>
Oh he parsed stylesheets to add that support?  I could build something like that.

[20:14:00.0000] <miketaylr>
oh you mean his web forms script

[20:14:01.0000] <miketaylr>
i wonder if he'll ever finish that

[20:14:02.0000] <Hixie>
boblet: one of the main advantages of <nav> is it lets speech readers skip navigation sections, so in footers it's not especially useful unless you want to style it

[20:14:03.0000] <JonathanNeal>
Maybe ... you guys are the experts ... I just talk a lot :)

[20:15:00.0000] <MikeSmith>
nav++

[20:15:01.0000] <MikeSmith>
I wonder why nobody has written a change proposal for removing nav

[20:16:00.0000] <paul_irish>
miketaylr: oh shittttt whoa.. http://code.google.com/p/webforms2/ our buddies weston and zoltan have been hacking on that too

[20:17:00.0000] <MikeSmith>
wow

[20:17:01.0000] <miketaylr>
ahh cool. i showed that to weston a while back

[20:17:02.0000] <miketaylr>
back when dean told me "it's beta, please don't blog about it" :/

[20:17:03.0000] <MikeSmith>
did they actually implement the repetition model too?

[20:17:04.0000] <miketaylr>
oh wait, no sorry. this is weston's old one

[20:18:00.0000] <boblet>
Hixie: unless you wanted to access those links I guess

[20:18:01.0000] <paul_irish>
MikeSmith: http://sourceforge.net/projects/wf2/forums/forum/466583/topic/1604767?message=3994714 looks like it?

[20:18:02.0000] <Hixie>
boblet: yeah, but who ever uses footer navigation links?

[20:19:00.0000] <MikeSmith>
paul_irish: hmm, yeah, 2006

[20:19:01.0000] <Hixie>
i mean, if they were meant for common use, they'd be up front

[20:19:02.0000] <boblet>
well, the Emperor of Exampland, for one

[20:19:03.0000] <boblet>
? ;-)

[20:19:04.0000] <miketaylr>
afaik, weston hasn't touched this in years

[20:20:00.0000] <boblet>
Hixie: I’m still fuzzy on what is and isn’t a major (née primary) navigation block though. the previous “footer links wouldn’t qualify b/c they’re random” was informative, but that’s gone now

[20:20:01.0000] <Hixie>
only the author can know what's major and what isn't

[20:21:00.0000] <boblet>
MikeSmith: the wording of the example about links in footer in the nav element has changed in WHATWG spec, so you might want to update nav details box

[20:21:01.0000] <boblet>
Hixie: sounds koan-y. me likes

[20:22:00.0000] <MikeSmith>
I still think it would be useful to have a standard way to distinguish primary navigation from secondary navigation

[20:22:01.0000] <Hixie>
primary navigation is first? :-)

[20:23:00.0000] <MikeSmith>
in document order? it could be last if the author wants to, and use CSS to position it

[20:24:00.0000] <MikeSmith>
seriously, the main distinction that most designers seem to make is between primary and secondary nav

[20:24:01.0000] <MikeSmith>
if you look at class values and such in more granularity

[20:24:02.0000] <MikeSmith>
but wtfu do I know anyway

[20:24:03.0000] <MikeSmith>
I'm not a designer

[20:24:04.0000] <MikeSmith>
and don't want to be

[20:25:00.0000] <MikeSmith>
CSS is just to frustrating

[20:25:01.0000] <MikeSmith>
*too

[20:25:02.0000] <boblet>
traditionally the hardcore standardista is content then site nav, but with the state of CSS positioning (…implementations) it pretty much always ends up being nav then content

[20:26:00.0000] <boblet>
s/standardista/standardista approach/

[20:26:01.0000] <boblet>
MikeSmith: CSS is great, it’s just the implementations that are … frustrating

[20:26:02.0000] <MikeSmith>
I guess we will know more after nav gets used -- I'd expect to see a lot of <nav class=primary> and <nav class=secondary> and such

[20:26:03.0000] <MikeSmith>
boblet: same could be said about many features of the web platform

[20:27:00.0000] <MikeSmith>
it's all great until you actually try to use it

[20:27:01.0000] <boblet>
given ppl who care about such things generally add skip links, that could be used as a designator of primary nav

[20:28:00.0000] <boblet>
generally only one “skip to nav” link, even if there are multiple things that could be marked up with nav

[20:28:01.0000] <boblet>
(although I’ve seen “skip to search”)

[20:28:02.0000] <MikeSmith>
Hixie: trying to implement http://dev.w3.org/html5/spec/text-level-semantics.html#guidance-for-conformance-checkers is giving me heartburn, man

[20:29:00.0000] <MikeSmith>
I really need to chat with hsivonen 

[20:29:01.0000] <KaOSoFt>
boblet- Why should content be first than navigation? Sorry, I'm kind of new to this Web development topic.

[20:30:00.0000] <boblet>
KaOSoFt: the idea is that for content pages (ie not the homepage or the archives etc) the user generally wants to get to the content. For text-to-speech reading through the navigation and other stuff in the header takes time (visual equivalent is Flash loading animation)

[20:31:00.0000] <boblet>
KaOSoFt: hence skip to content links and putting nav after content in HTML (which as mentioned is generally a massive pita)

[20:32:00.0000] <KaOSoFt>
boblet- Hmmm, "content pages". I get it. Thanks. :)

[20:32:01.0000] <boblet>
KaOSoFt: happy to help :)

[20:53:00.0000] <othermaciej>
MikeSmith: have you found any of the conditions hard other than the paragraph/section one?

[20:54:00.0000] <MikeSmith>
othermaciej: the figcaption exception is PITA, though not really hard I suppose

[20:55:00.0000] <MikeSmith>
the fact that that you the figcaption can occur before or after the img in document order

[20:55:01.0000] <MikeSmith>
so have to hold off on that check until endelement event for the figure element

[20:55:02.0000] <othermaciej>
MikeSmith: I was gonna say - you could track that you are in a figure

[20:56:00.0000] <MikeSmith>
yeah

[20:56:01.0000] <MikeSmith>
we just don't have a similar case like this yet to steal from

[20:56:02.0000] <MikeSmith>
in the existing code

[20:57:00.0000] <MikeSmith>
we will after this, though, I guess

[20:57:01.0000] <MikeSmith>
though I hope we won't be needing any more of this kind of stuff

[21:00:00.0000] <othermaciej>
yeah I don't know what other requirements might be like this

[21:01:00.0000] <MikeSmith>
hopefully none

[21:01:01.0000] <MikeSmith>
because from where I sit, it's pretty much an anti-pattern for conformance criteria

[21:02:00.0000] <MikeSmith>
*for document-conformance criteria

[21:03:00.0000] <MikeSmith>
it's also a challenge to try to figure out how to provide a concise error message for this set of cases

[21:03:01.0000] <MikeSmith>
or impossible

[21:03:02.0000] <MikeSmith>
I think the only practical thing to do is to provide a link to some document that provides detailed guidance

[21:04:00.0000] <MikeSmith>
e.g., as with the "Use CSS instead" link we have now

[21:05:00.0000] <othermaciej>
MikeSmith: I would describe the error as just missing alt

[21:05:01.0000] <MikeSmith>
yeah

[21:06:00.0000] <othermaciej>
MikeSmith: then briefly mention there are some other alternatives for describing an image and link to the list

[21:06:01.0000] <MikeSmith>
I think "element img is missing required attribute alt" is just about all the error message itself can say

[21:07:00.0000] <othermaciej>
really, there's no ability to give anything more descriptive for a missing attribute?

[21:07:01.0000] <MikeSmith>
yeah, there is

[21:07:02.0000] <MikeSmith>
it can add what you just described

[21:08:00.0000] <othermaciej>
ok, good

[21:08:01.0000] <othermaciej>
I found a lot of suboptimal error messages during the conformance study, at some point I should be a good citizen and file more validator bugs

[21:08:02.0000] <MikeSmith>
what I meant is, just the description of what the actual error is, I don't know what else to have it say

[21:08:03.0000] <MikeSmith>
some of those I have fixed already

[21:08:04.0000] <othermaciej>
ah, right, as a short description, I think that is fine

[21:08:05.0000] <othermaciej>
yeah, I've seen some of your fixes going by

[21:10:00.0000] <othermaciej>
anyway - that's a mostly accurate description of the error, even though there are a few exemptions to the alt requirement

[21:10:01.0000] <MikeSmith>
right

[21:25:00.0000] <JonathanNeal>
heyo

[21:35:00.0000] <Hixie>
MikeSmith: file bugs

[23:15:00.0000] <Hixie>
othermaciej: any suggestions for a shorter term than "valid URL potentially surrounded by spaces"?

[23:16:00.0000] <othermaciej>
Hixie: "valid URL attribute value" is a little shorter (assuming that is equally accurate)

[23:16:01.0000] <othermaciej>
or even just "URL attribute value" if there is no need for the non-valid version

[23:17:00.0000] <Hixie>
i'd rather not call the syntax by where it's used

[23:18:00.0000] <Hixie>
that always ends up giving me trouble later when i reuse the syntaxes

[23:21:00.0000] <othermaciej>
"space-delimited URL"

[23:24:00.0000] <Hixie>
i'll just go with potentially surrounded by spaces, it's less ambiguous

[23:34:00.0000] <othermaciej>
it does seem like a pretty unambiguous choice

[23:45:00.0000] <MikeSmith>
me tries to figure out which way the wind blowing on adding presence of role=presentation as an exception to the alt-required constraint

[23:45:01.0000] <micheil>
morning good chaps & ladies.

[23:46:00.0000] <MikeSmith>
if it is favorable I could add it speculatively/experimentally now to the checker in the interest of giving something broader to test against

[23:46:01.0000] <MikeSmith>
or I could hold off and just do it for now as close to the existing spec as I can get

[23:47:00.0000] <MikeSmith>
micheil: hey

[23:47:01.0000] <micheil>
MikeSmith: if it weren't for the variations in spellings, we'd have the same first/last names.

[23:48:00.0000] <MikeSmith>
your first name has a (tm) in it also?

[23:56:00.0000] <Hixie>
MikeSmith: the role=presentation exception is pointless... it's longer than the alternative (just writing alt="", which implied role=presentation automatically)

[23:56:01.0000] <Hixie>
s/implied/implies/

[23:57:00.0000] <micheil>
MikeSmith: ah, no. 

[23:58:00.0000] <micheil>
MikeSmith: besides, mine's more trademarkable

[23:59:00.0000] <MikeSmith>
Hixie: OK

[23:59:01.0000] <micheil>
hey, Hixie thanks for the slight tip last night on where to look for the protocol details on websockets, so far I've got the server accepting clients and reading the headers

[23:59:02.0000] <Hixie>
cool

[23:59:03.0000] <MikeSmith>
micheil: your parents clearly have more imagination than mine

[00:00:00.0000] <MikeSmith>
Hixie: how about aria-labelledby ?

[00:01:00.0000] <micheil>
Hixie: I've been working from a slightly different use case, so far all the server's I've seen don't make use of the path in the initial GET request

[00:01:01.0000] <micheil>
is it intended to be used?

[00:01:02.0000] <Hixie>
MikeSmith: http://www.w3.org/Bugs/Public/show_bug.cgi?id=6496

[00:01:03.0000] <Hixie>
MikeSmith: yes

[00:01:04.0000] <Hixie>
er

[00:01:05.0000] <Hixie>
micheil: yes

[00:02:00.0000] <micheil>
Hixie: cool, do you know of any other servers they make use of it?

[00:03:00.0000] <Hixie>
micheil: i wrote one that does, but it's a hack :-) (i used it to smuggle in the username/password of the user, instead of sending it in the first frame)

[00:03:01.0000] <micheil>
so far most I've seen don't even parse the headers correctly

[00:03:02.0000] <Hixie>
yeah

[00:03:03.0000] <micheil>
because, theoretically there aren't any headers not allowed

[00:04:00.0000] <micheil>
plus, they use a fixed order for them, which isn't correct

[00:04:01.0000] <Hixie>
that's why the latest draft requires the servers to parse two headers, because otherwise the servers wouldn't parse any and would be vulnerable to a cross-protocol attack from HTTP

[00:04:02.0000] <MikeSmith>
Hixie: I see, thanks (about aria-labelledby)

[00:05:00.0000] <micheil>
two headers? I've missed that one.. I'll have to check up on it

[00:05:01.0000] <Hixie>
the two Sec-WebSocket-Key headers

[00:05:02.0000] <Hixie>
MikeSmith: as a general rule, ARIA is a layer above HTML, so it should be possible to take any HTML file with ARIA, remove all the ARIA, and not make the page less conforming

[00:05:03.0000] <micheil>
oh, I thought they were part of the first GET request

[00:06:00.0000] <Hixie>
micheil: they are

[00:06:01.0000] <Hixie>
micheil: by "header" i meant "field" in websocket terminology, my bad

[00:07:00.0000] <micheil>
okay

[00:09:00.0000] <MikeSmith>
Hixie: yeah, that does seems rational, thanks

[02:33:00.0000] <hsivonen>
ouch. Tony Ross misdescribed the fallback mechanism of <video> at MIX.

[02:33:01.0000] <hsivonen>
claiming that the fallback content is rendered if the codec isn't available

[02:33:02.0000] <hsivonen>
I wonder if they've implemented it that way in IE9

[02:34:00.0000] <hsivonen>
and I wonder how many MIX10 visitors will now assume they can rely on the described behavior

[02:36:00.0000] <MikeSmithX>
hsivonen: are you going to around still in a couple hours from now?

[02:36:01.0000] <MikeSmithX>
I really want to talk with you when you have time

[02:36:02.0000] <MikeSmithX>
but I need to head out for a bit first

[02:44:00.0000] <micheil>
Hixie: good lordy is that spec on WebSockets very complete with all basis covered..

[02:58:00.0000] <zcorpan>
Hixie++

[02:58:01.0000] <micheil>
Hixie: also, with the usage of read 8 bytes from the server/client, I've found an ambiguity: if it's in UTF8, 8 bytes would be variable

[02:59:00.0000] <micheil>
it may be 8 characters or it could be four

[02:59:01.0000] <zcorpan>
micheil: the random bytes only use characters in the ascii range

[02:59:02.0000] <micheil>
ah

[02:59:03.0000] <jgraham>
micheil: it need not be characters at all

[03:00:00.0000] <micheil>
so, in which case the characters would have a string length of 8

[03:00:01.0000] <jgraham>
+utf8

[03:00:02.0000] <jgraham>
micheil: It is better to think of it as a bytearray of length 8

[03:00:03.0000] <micheil>
:/

[03:00:04.0000] <jgraham>
Dunno is node.js supports that

[03:00:05.0000] <micheil>
not sure I have access to bytearray types

[03:00:06.0000] <jgraham>
concept

[03:00:07.0000] <micheil>
considering it's js

[03:00:08.0000] <jgraham>
Right

[03:01:00.0000] <jgraham>
Hopefully soon

[03:01:01.0000] <gsnedders>
js--

[03:01:02.0000] <zcorpan>
tc39--

[03:01:03.0000] <jgraham>
If the TC39 people pull their finger out a little

[03:01:04.0000] <micheil>
gsnedders: js++*Infinity.

[03:01:05.0000] <jgraham>
(it looks like WebGL will force the issue though)

[03:02:00.0000] <gsnedders>
I'm sorry, GWT has made me hate JS forever.

[03:02:01.0000] <jgraham>
Anyway, the point is that in a sensible languge (note: very few such languages exist) that makes a clear distinction between a string and a sequence of bytes

[03:02:02.0000] <micheil>
gsnedders: GWT's mainly java though

[03:02:03.0000] <gsnedders>
micheil: The JS it generates is horrid.

[03:02:04.0000] <jgraham>
you would treat the random bytes as bytes, not as a string

[03:03:00.0000] <micheil>
yeah, I'm not generating JS, I'm writing it.

[03:03:01.0000] <jgraham>
gsnedders: Presumably the anyone feels the same way debugging generated bytecode

[03:03:02.0000] <micheil>
well, I do have a byte buffer that I could use (interfaces with C), but that may be a little heavy

[03:03:03.0000] <hsivonen>
/me hopes the byte array issue gets resolved before any browser ships a release version with WebGL enabled

[03:04:00.0000] <jgraham>
/me seriosuly thinks it would help to replace all the variable names in GWT code with random english words

[03:05:00.0000] <jgraham>
But feels that eval might screw things up

[03:07:00.0000] <jgraham>
(the idea being it is easier to figure out what is going on if you can say that "sheep calls confuse which calls ancilliary" rather than "xzy calls dfg which calls ijk"

[03:07:01.0000] <jgraham>
)

[03:13:00.0000] <hsivonen>
it's sure nicer to figure out what GWT is doing if you possess the original Java source and can recompile without minification

[03:14:00.0000] <hsivonen>
what bits of Google Wave are open source?

[03:14:01.0000] <hsivonen>
is any of the front end GWT stuff Open Source?

[03:18:00.0000] <zcorpan>
annevk: i think it would be awesome if the top five hundred or so websites in each country were accessible

[03:18:01.0000] <jgraham>
/me wonders what metric you are using for top 500

[03:19:00.0000] <zcorpan>
/me is quoting annevk's blog

[03:19:01.0000] <jgraham>
I mean quite a few of those are probably porn. DO you get accessible porn?

[03:19:02.0000] <zcorpan>
it would be awesome

[03:21:00.0000] <gsnedders>
So, the alt text for images is just erotic writing instead of the pictures?

[03:22:00.0000] <zcorpan>
what about video?

[03:23:00.0000] <jgraham>
Yeah, I guess a transcript wouldn't help much

[03:23:01.0000] <annevk>
zcorpan, yeah, maybe I should have lowered the number, wasn't sure

[03:23:02.0000] <hsivonen>
http://diveintomark.org/archives/2006/07/27/imagination

[03:25:00.0000] <jgraham>
hsivonen: I didn't dount that there sould be a demand for accessible porn

[03:25:01.0000] <jgraham>
Well I did obviously

[03:25:02.0000] <jgraham>
But I didn't mean to

[03:25:03.0000] <jgraham>
I meant to doubt that it would be an alternative presentation of other porn

[03:26:00.0000] <jgraham>
rather than a totally different resource

[03:26:01.0000] <jgraham>
s/sould/would/

[03:27:00.0000] <jgraham>
s/dount/doubt/

[03:27:01.0000] <jgraham>
You know I read that several times and didn't notice that I had totally misspelt the key word in the sentence

[03:30:00.0000] <annevk>
now http://lists.w3.org/Archives/Member/w3c-archive/2010Apr/0039.html (Member-only) is interesting

[03:30:01.0000] <annevk>
finally some hidden agenda gossip

[03:37:00.0000] <micheil>
hmm.. what would be an example of the /resource name/ for a websocket on the server?

[03:37:01.0000] <micheil>
morning there jeremy (adactio)

[03:37:02.0000] <adactio>
Morning.

[03:42:00.0000] <jgraham>
micheil: Presumably it can be anything you want

[03:42:01.0000] <micheil>
hmm..

[03:42:02.0000] <micheil>
okay

[03:42:03.0000] <jgraham>
It is just an opaque string that gets used by the server however it choses

[03:42:04.0000] <micheil>
it's definition seems rather ambiguous, I don't know if it should be clarified (/cc Hixie )

[03:43:00.0000] <micheil>
under the definition, it states that it should be the resource given in the clients handshake, then there's two blank lines and something about True / False if encrypted

[03:44:00.0000] <micheil>
True if the connection is encrypted or if the server expected it to be encrypted; false otherwise.

[03:44:01.0000] <jgraham>
The two blank lines sound like they should be talking about the secure flag

[03:44:02.0000] <jgraham>
What document are you reading?

[03:47:00.0000] <zcorpan>
ok now someone needs to test for canvas bugs in flash cs5

[03:48:00.0000] <annevk>
hmm, I meant W3C Member-only above...

[03:48:01.0000] <jgraham>
annevk: What other type of Member could you have meant?

[03:49:00.0000] <jgraham>
whatwg Member I guess

[03:57:00.0000] <micheil>
http://www.whatwg.org/specs/web-socket-protocol/

[04:00:00.0000] <micheil>
same is present in the latest draft: http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-75#section-5.1

[04:01:00.0000] <annevk>
-75 and and the whatwg.org version are different

[04:01:01.0000] <micheil>
on may be more up-to-date

[04:02:00.0000] <annevk>
whatwg.org is

[04:02:01.0000] <micheil>
but the problem is present from about -67 onwards

[04:02:02.0000] <jgraham>
If you have a trunk gecko or webkit based browser to hand (or maybe even lynx or something) I seriously recommend using http://www.whatwg.org/specs/web-apps/current-work/complete.html?slow-browser=1#network

[04:03:00.0000] <jgraham>
It is always the latest version and you can link to any part you are having trouble with

[04:03:01.0000] <micheil>
that is hella slow, but anyway :)

[04:03:02.0000] <jgraham>
I think, but without complete certianty, that the slow-browser thing stops some scripts from running

[04:03:03.0000] <jgraham>
which makes it faster

[04:04:00.0000] <jgraham>
(but that might just be the main spec or I might have forgotten the right parameter name)

[04:06:00.0000] <zcorpan>
jgraham: ?slow-browser is right

[04:06:01.0000] <micheil>
hmm.. I suppose rather then setting my key_1 & key_2 to be the values of the two Sec-WebSocket_Key[1|2] fields, I should just set them to be the numbers, as that's all that's actually used, isn't it?

[04:07:00.0000] <micheil>
no, actually that's wrong.

[04:07:01.0000] <zcorpan>
the numbers and the number of spaces are used

[04:07:02.0000] <micheil>
yeah

[04:07:03.0000] <micheil>
I might have to make that a sub-method.. 

[04:07:04.0000] <micheil>
bbl. dinner

[04:08:00.0000] <micheil>
thanks for the help folks :)

[04:10:00.0000] <zcorpan>
hmm wonder if it's problematic if the client inserts the spaces at the start or at the end of the string

[04:11:00.0000] <annevk>
not if you have a WebSocket server

[04:11:01.0000] <annevk>
but if you do both HTTP and WebSocket you need some kind of switching before continuing parsing

[04:12:00.0000] <annevk>
or raw buffer access or something I suppose

[05:18:00.0000] <micheil>
annevk: the solution is to make proper use of the upgrade / connection headers

[05:21:00.0000] <annevk>
so if you are an HTTP server you parse the first bit as HTTP until you hit the upgrade header at which point you continue parsing using a WebSocket parser?

[05:27:00.0000] <micheil>
yeah

[05:27:01.0000] <jgraham>
FWIW I think assuming HTTP servers have a particular architecture is probably a bad idea

[05:27:02.0000] <jgraham>
including an architecture required to conform to the spec

[05:27:03.0000] <jgraham>
(HTTP spec that is)

[05:28:00.0000] <micheil>
jgraham: true, hence the reason I'm implementing my node websocket server on top of raw network sockets

[05:28:01.0000] <jgraham>
Especially for little-used features

[05:29:00.0000] <micheil>
I know that the headers in a websocket client are going to be in the format: GET {path} HTTP/1.1\r\n{headers}\r\n\r\n{body}\r\n\r\n

[05:29:01.0000] <micheil>
or something like that

[05:29:02.0000] <jgraham>
micheil: Right but from the point of view of websockets it means we should be wary of security features that might fail due to likely bugs in servers

[05:29:03.0000] <jgraham>
(i.e. I agree with zcorpan about the location of the spaces)

[05:29:04.0000] <micheil>
and then the headers are going to be separated as {key}: {value}\r\n

[05:30:00.0000] <micheil>
and according to the spec for websockets, you should be fine, however, I wouldn't think of sticking websockets and a httpserver on the same network process

[05:30:01.0000] <micheil>
(ie, reading from the same stream of data)

[05:31:00.0000] <jgraham>
Me neither but some people seem to want that

[05:31:01.0000] <jgraham>
Or something

[05:31:02.0000] <micheil>
you do two processes

[05:31:03.0000] <micheil>
different ports

[05:32:00.0000] <micheil>
eg, your webserver on *.host.tld:80, and then your websockets on say, ws.host.tld:80

[05:32:01.0000] <micheil>
or just host:80 for http, host:8080 for websockets

[05:33:00.0000] <micheil>
nothing actually stops the two network sockets from talking to each other, depending on how you architect your app, take PusherApp.com for example, they have a rails app somehow pushing data onto a websocket server

[05:33:01.0000] <annevk>
jgraham, requiring a WebSocket server to understand HTTP semantics is also bad

[05:33:02.0000] <annevk>
jgraham, same for a WebSocket client

[05:34:00.0000] <micheil>
annevk: there needs to be some form of handshaking, and if that's done over upgrade, so be it.

[05:34:01.0000] <jgraham>
annevk: It doesn't have to at the moment, really

[05:35:00.0000] <jgraham>
It just has to follow the spec

[05:38:00.0000] <micheil>
as far as I'm concerned, I'll leave the spec writing to the guys who are the computer scientists or know what they're doing. I'm just interested in implementing and using things to the spec.

[05:39:00.0000] <zcorpan>
micheil: the spec likely has a number of bugs. if you're implementing it, you might catch bugs as you go, and it'd be really useful if you sent comments about your findings

[05:39:01.0000] <micheil>
yeah

[05:40:00.0000] <micheil>
I will do, so far I've only had issues with wording on it

[05:40:01.0000] <zcorpan>
cool

[05:41:00.0000] <jgraham>
/me wonders how many people are actually Computer Scientists by training. I mean I know Hixie isn't (I know a few people are but most people I just have no idea)

[05:42:00.0000] <micheil>
jgraham: Hixie isn't? I always had the impression that a lot of the guys on the w3c board / the spec writers were computer scientists and things like that

[05:42:01.0000] <micheil>
and I'm sure if you asked most web developers, they'd think the same

[05:42:02.0000] <zcorpan>
raise hands, everyone, who are computer scientists?

[05:43:00.0000] <asmodai>
What's a computer scientist? Someone who theorizes about computer stuff?

[05:43:01.0000] <annevk>
jgraham, indeed, that's a good thing

[05:43:02.0000] <asmodai>
/me did 1 year of something akin to CS and then just quit uni altogether

[05:43:03.0000] <workmad3>
asmodai: that's one thing

[05:43:04.0000] <annevk>
jgraham, but it does assume HTTP server architecture that also want to implement WebSocket can deal with that

[05:43:05.0000] <hsivonen>
/me has a CS degree

[05:43:06.0000] <jcranmer>
although I'm still at uni, so....

[05:43:07.0000] <workmad3>
/me has a CS masters

[05:43:08.0000] <asmodai>
/me is one of those autodidact scary types

[05:43:09.0000] <workmad3>
I wouldn't say I'm a computer scientist though

[05:44:00.0000] <Lachy>
/me has an IT degree, which AIUI, is basically the same as a CS degree, but had a different name at my university.

[05:44:01.0000] <workmad3>
CS is more about research into areas of computing more than actual software development

[05:44:02.0000] <micheil>
I've so far done.. 1 subject in a computer science degree.. but that's only because I can't do studies fulltime with highschool still on

[05:45:00.0000] <workmad3>
and I'd say the majority of CS courses and CS professors couldn't program for toffee

[05:45:01.0000] <jgraham>
/me wonders how a course would program

[05:45:02.0000] <workmad3>
jgraham: bad phrasing there

[05:45:03.0000] <workmad3>
CS courses don't tend to teach people how to program

[05:46:00.0000] <workmad3>
they teach students some programming languages, and then it's pretty much ignored

[05:46:01.0000] <Lachy>
workmad3, do they teach more about hardware aspects of computing?

[05:46:02.0000] <jcranmer>
pretty much

[05:46:03.0000] <Lachy>
like processor architectures, etc?

[05:46:04.0000] <jcranmer>
although in the course I TA, I do make it a point to beat in good coding practices

[05:47:00.0000] <workmad3>
Lachy: some do, or they'll teach things things like language design principles, or automated program and theorem proving, or lots of advanced db theory

[05:47:01.0000] <jcranmer>
/me will be taking Advanced Algorithm theory, compiler theory, and ethics as CS courses next semester

[05:47:02.0000] <workmad3>
but (mine at least) stopped on the programming front at the point of 'you can churn out some basic code now, you're a programmer'... no use of VCS, no code design or architecture, basically nothing to make a decent software engineer

[05:48:00.0000] <jgraham>
/me notes that the theoretical CS courses seem the most interesting / useful

[05:48:01.0000] <jgraham>
(but that could be my bias)

[05:48:02.0000] <Lachy>
my IT course covered a range of subjects including Java, C++, HTML/CSS/JS, databases and SQL, with little bits about hardware

[05:48:03.0000] <jcranmer>
well, given the scope of a project in a class, using VCS is pretty much overkill

[05:48:04.0000] <Lachy>
also covered things like project development and management

[05:49:00.0000] <jgraham>
(because you can pretty easilly lean to program by programming but learning, say, analysis of complex algorithsm on your own is pretty hard)

[05:49:01.0000] <micheil>
that's rather confusing.. in that complete work webapp thing, the numbering of sections is totally different to in the IEFT drafts

[05:49:02.0000] <jcranmer>
I think the largest one I managed to hit was 1K lines in magnitude

[05:49:03.0000] <workmad3>
jcranmer: if it's over about 100 LoC, I'd want it in a VCS now

[05:49:04.0000] <workmad3>
although I'd probably use git and just keep it local admittedly :)

[05:50:00.0000] <jcranmer>
lemme rephrase... it's one or two files, essentially

[05:50:01.0000] <micheil>
workmad3: what, you mean you don't work like a designer? myclass.final.12.final.proposal.png

[05:50:02.0000] <micheil>
:P

[05:50:03.0000] <jcranmer>
VCS comes into play when either a) I want to publish it or b) it's too complex that changing a file or three at a time stops working

[05:50:04.0000] <workmad3>
micheil: heh :)

[05:51:00.0000] <micheil>
/me wonders if the whatwg specs are in a git repo..

[05:51:01.0000] <workmad3>
jcranmer: for me, VCS comes into play when a) I want to do some code

[05:51:02.0000] <jcranmer>
micheil: svn I believe

[05:51:03.0000] <jcranmer>
workmad3: wuss :-)

[05:51:04.0000] <workmad3>
it's just a standard thing to do for me now... especially with how easy it is to just create a git repo for it

[05:51:05.0000] <micheil>
hmm.. public?

[05:52:00.0000] <jgraham>
VCS and so on seem like practical skills that are really easy to pick up

[05:52:01.0000] <jgraham>
Kinda like a physics degree teaching you how to solder

[05:52:02.0000] <micheil>
what's physics got to do with soldering? 

[05:52:03.0000] <jgraham>
(although I admit I am hopeless at soldering and when I had to do that in my physics degree I made a total mess of it)

[05:52:04.0000] <workmad3>
jgraham: there were people on electrical engineering course at my uni in the third year who couldn't solder

[05:53:00.0000] <jcranmer>
micheil: http://html5.org/tools/web-apps-tracker

[05:53:01.0000] <jgraham>
workmad3: I didn't say engineering on purpose

[05:53:02.0000] <micheil>
jcranmer: neat

[05:54:00.0000] <workmad3>
jgraham: my view personally is that if a uni course is claiming to teach programming, then they should teach the good habits that go with it

[05:55:00.0000] <workmad3>
anyway, I need to go get lunch

[05:55:01.0000] <jcranmer>
-20 points: used #include "mylib.c"

[05:56:00.0000] <jgraham>
(the rather strained analogy I was trying to make is that soldering is often a required skill for practising physicists. But the point of the degree is to make sure that you know about quantum machanics and relativity and ways of thinking about solving problems like a physicist, not about how to solder)

[05:56:01.0000] <jcranmer>
-100 points: code doesn't compile due to cyclic include structure (WTF?)

[05:56:02.0000] <micheil>
hmm.. with the ASCII case-insensitive comparing, would an implementor be expected to convert from UTF8 to ASCII to compare, or would change the variable value to lowercase and checking it against another value that is lowercase be sufficient?

[05:57:00.0000] <jgraham>
micheil: It is defined somewhere, but it basically means change ACSII uppercase to ASCII lowercase and leave all other characters untouched

[05:57:01.0000] <micheil>
yeah

[05:58:00.0000] <micheil>
so equivilant to javascripts string.toLowerCase

[05:58:01.0000] <Philip`>
That does Unicode lowercasing, I thought

[05:58:02.0000] <micheil>
uhh.. I'm not sure

[05:58:03.0000] <jgraham>
Yes, not equivalent, for the reasons Philip gave

[05:59:00.0000] <Philip`>
You just want tr/A-Z/a-z/ (or whatever the syntax is)

[05:59:01.0000] <micheil>
hmm..

[05:59:02.0000] <micheil>
yeah, so, I'd probably have to use regex, and then use a callback on it

[06:00:00.0000] <annevk>
or you could map A-Z to a-z and then compare

[06:01:00.0000] <micheil>
huh?

[06:01:01.0000] <annevk>
but if you implement in JS I guess regexp might be faster

[06:01:02.0000] <micheil>
yeah, I think so

[06:47:00.0000] <micheil>
how does one report a bug?

[06:47:01.0000] <micheil>
oh, wait, not a bug

[06:58:00.0000] <erlehmann>
how does one report a feature ? hehehe

[06:58:01.0000] <annevk>
local authorities

[07:00:00.0000] <micheil>
erlehmann: no, it was just the way I read it

[07:00:01.0000] <micheil>
"append ws:// or wss:// to the url"

[07:01:00.0000] <micheil>
it sounds like you would get the url being: example.comws://

[07:01:01.0000] <micheil>
until you read that it's actually url begins empty

[08:08:00.0000] <annevk>
nessy, it's not reserved, just pointing out how it might affect existing features

[08:09:00.0000] <nessy>
I think that would be a nice side-effect actually!

[08:11:00.0000] <nessy>
I don't actually think the attribute name makes much of a difference - but sure, we gotta pay attention to details ;)

[08:59:00.0000] <annevk>
hsivonen, it should be srclang then arguably

[08:59:01.0000] <annevk>
hsivonen, or s/src/href/ but that seems worse

[09:00:00.0000] <hsivonen>
annevk: yeah, I noticed but I went with hreflang still

[09:00:01.0000] <hsivonen>
annevk: do you have an opinion on my concrete captioning mechanism suggestion at the end of the email?

[09:04:00.0000] <annevk>
not really sure what to think about pseudo-HTML for captioning to be honest

[09:04:01.0000] <annevk>
it sure beats TTML

[09:05:00.0000] <hsivonen>
it's not pseudo! innerHTML is the real thing!

[09:05:01.0000] <annevk>
heh

[09:06:00.0000] <annevk>
it does seem relatively straightforward

[09:06:01.0000] <annevk>
and much more Web-like

[09:07:00.0000] <TabAtkins>
/me boggles at 7 XML namespaces.

[09:28:00.0000] <TabAtkins>
hsivonen: How are you proposing that the time of the text string be encoded in the HTML version?

[09:28:01.0000] <annevk>
it's just SRT with HTML instead of text

[09:28:02.0000] <TabAtkins>
So you still write a number, followed by a time range, then followed by some HTML and finally a linebreak?

[09:29:00.0000] <TabAtkins>
s/linebreak/empty line/

[09:29:01.0000] <annevk>
yeah

[09:30:00.0000] <TabAtkins>
ok

[09:30:01.0000] <jgraham>
I swaer I saw a caption the other day where they blurred a swearword

[09:30:02.0000] <TabAtkins>
Um.

[09:30:03.0000] <jgraham>
But I was very unsure afterwards if I had just imagined it

[09:30:04.0000] <TabAtkins>
/me will be brb, rebotting.

[09:30:05.0000] <jgraham>
I always knew TabAtkins had to be a bot

[09:33:00.0000] <micheil>
question: how does one work out in the websockets if key-number1 is not an integral multiple of spaces1?

[09:34:00.0000] <micheil>
is it simple a Mod style expression?

[09:34:01.0000] <jgraham>
Yes

[09:34:02.0000] <micheil>
right

[09:35:00.0000] <micheil>
so what, key_number1 % spaces1 ?

[09:35:01.0000] <jgraham>
== 0

[09:36:00.0000] <jgraham>
or != 0 I guess

[09:36:01.0000] <micheil>
yeah

[09:36:02.0000] <jgraham>
or !== in javascript :)

[09:36:03.0000] <micheil>
good point there too

[09:36:04.0000] <micheil>
bitwise vs standard

[09:37:00.0000] <jgraham>
Well type converting vs non-type converting

[09:37:01.0000] <jgraham>
("non-strict" vs "strict" in ES-spec parlance)

[09:38:00.0000] <micheil>
yeah

[09:40:00.0000] <gsnedders>
(of course unrelated to strict-mode)

[09:40:01.0000] <gsnedders>
(I <3 the ES spec)

[09:40:02.0000] <micheil>
it's surprising how many people are implementing websocket servers wrongly

[09:40:03.0000] <jgraham>
micheil: The spec changed recently

[09:41:00.0000] <jgraham>
I expect they are implementing the old spec

[09:41:01.0000] <micheil>
not that greatly though

[09:41:02.0000] <jgraham>
or rather implemented

[09:41:03.0000] <micheil>
in the sense that they are using fixed header fields and very fixed input

[09:42:00.0000] <jgraham>
Well the whole handshake thing is new

[09:42:01.0000] <micheil>
ah, fair enough

[09:42:02.0000] <jgraham>
And there were lots of people pressing to make it easier to implment on top of an existing HTTP server

[09:45:00.0000] <gsnedders>
/me wonders what made Hixie change his mind

[09:45:01.0000] <gsnedders>
/me was arguing that within a few months of him spec'ing out the original WS draft

[09:45:02.0000] <micheil>
I wouldn't want websockets on top of http.

[09:46:00.0000] <micheil>
http has a fair bit of extra data baggage, compared with a raw tcp socket

[10:00:00.0000] <jgraham>
/me proposes ircroulette where you get dumped into a random irc channel until you leave or someone kicks you

[10:00:01.0000] <jgraham>
and then you get dumped into the next channel

[10:01:00.0000] <jgraham>
and then finally you get dumped by your SO for spending too much time on the internet, become depressed and die

[10:02:00.0000] <gsnedders>
I guess this is the advantage of not having a SO, you just skip right to the become depreessed and die part, hence wasting less time.

[10:02:01.0000] <gsnedders>
Optimization ftw.

[10:02:02.0000] <jgraham>
Nah, you just have to pick one up in one of the irc chennels along the way

[10:03:00.0000] <jgraham>
*channels

[10:03:01.0000] <gsnedders>
But I thought all dating IRC channel were 18+!

[10:03:02.0000] <jgraham>
Do you now understand *random*?

[10:04:00.0000] <jgraham>
*not

[10:04:01.0000] <jgraham>
sigh

[10:04:02.0000] <gsnedders>
jgraham's a terrible typist.

[10:04:03.0000] <jgraham>
Yes

[10:04:04.0000] <jgraham>
Sorry

[10:04:05.0000] <gsnedders>
You better be!

[10:05:00.0000] <gsnedders>
I've got a bottole of Irn-Bru and I'm not afraid to use it!

[10:05:01.0000] <gsnedders>
*bottle

[10:05:02.0000] <jgraham>
I'm a bee?

[10:05:03.0000] <jgraham>
Are you planning to use the Irn-Bru like jam at a picnic?

[10:06:00.0000] <jgraham>
(that is a waste of good Irn-Bru)

[10:06:01.0000] <gsnedders>
Well, no. More just to scare you with it's bright orange colour.

[10:06:02.0000] <gsnedders>
*its

[10:55:00.0000] <AryehGregor>
Why is Ian listed as the only editor in the W3C Editor's Draft, but David Hyatt listed as a second editor in the WDs?

[10:56:00.0000] <gsnedders>
AryehGregor: hyatt no longer an editor

[10:56:01.0000] <AryehGregor>
So the next WD won't list him as an editor?

[10:56:02.0000] <AryehGregor>
/me needs a source to cite to revert http://en.wikipedia.org/w/index.php?title=HTML5&diff=355515922&oldid=355096550

[10:56:03.0000] <AryehGregor>
/me suspects linking to the SVN logs would be considered "original research"

[10:57:00.0000] <gsnedders>
Indeed not

[10:57:01.0000] <AryehGregor>
Do you know of anywhere this is written down that I can cite?

[10:57:02.0000] <gsnedders>
Linking to the editor's draft would be to, but what else can you cite?

[10:58:00.0000] <gsnedders>
What does it currently cite?

[10:58:01.0000] <AryehGregor>
I mean, like, was a change agreed upon somewhere to remove David Hyatt's name as an editor?

[10:58:02.0000] <AryehGregor>
It currently cites the latest WD.

[10:58:03.0000] <gsnedders>
Then change it to cite the latest ED

[10:58:04.0000] <AryehGregor>
That's what I'm doing.

[10:58:05.0000] <AryehGregor>
I was hoping there'd be a clearer source, though.

[10:59:00.0000] <gsnedders>
othermaciej sent something to public-html

[10:59:01.0000] <miketaylr>
AryehGregor: http://lists.w3.org/Archives/Public/public-html/2010Apr/0096.html

[10:59:02.0000] <miketaylr>
/me wasn't sure if he imagined that

[11:00:00.0000] <AryehGregor>
Oh, thanks.

[11:00:01.0000] <AryehGregor>
Serves me right for not keeping up.

[11:01:00.0000] <AryehGregor>
It's fun to come across old posts by your n00b self: https://bugzilla.wikimedia.org/show_bug.cgi?id=986#c12

[11:01:01.0000] <AryehGregor>
"my HTML skills are fairly modest"

[11:01:02.0000] <gsnedders>
AryehGregor: You're probably more up to date than the guy who sent the first ever message to whatwg.

[11:01:03.0000] <AryehGregor>
/me doesn't get what that means

[11:02:00.0000] <gsnedders>
He's get a tens of thousands of unread emails :)

[11:02:01.0000] <micheil>
jgraham: hmm.. the WebSockets security stuff is probably going to take a while to implement..

[11:03:00.0000] <TabAtkins_>
gsnedders: The one about the XML submission format?

[11:03:01.0000] <gsnedders>
/me can't remember what it's about

[11:03:02.0000] <TabAtkins_>
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2004-April/000000.html

[11:03:03.0000] <gsnedders>
Yeah

[11:03:04.0000] <inimino>
when was all this Sec-WebSocket-Key1 and Key2 stuff added to Web Sockets?

[11:04:00.0000] <micheil>
oh. hey inimino 

[11:04:01.0000] <micheil>
:P

[11:04:02.0000] <inimino>
oh hey micheil ;)

[11:06:00.0000] <inimino>
I remember the ideal of having Web Socket servers be written in a few lines of Perl, did that die in the IETF?

[11:08:00.0000] <micheil>
I suppose until I get the Security stuff working, I can leave it out

[11:08:01.0000] <jgraham>
micheil: To stop cross-protocl attacks

[11:08:02.0000] <jgraham>
*protocol

[11:12:00.0000] <gsnedders>
inimino: It still can be done in a few lines of Perl. It just depends how readable you want your Perl to be.

[11:13:00.0000] <inimino>
gsnedders: indeed :)

[11:16:00.0000] <TabAtkins_>
<keygen> is generally not something we like, right?  It's only specced because it's common on the web and a few browsers have at least parsing support for it?

[11:17:00.0000] <jgraham>
TabAtkins_: Right

[11:17:01.0000] <jgraham>
(alos because we don't have a good other solution for the use cases)

[11:17:02.0000] <TabAtkins_>
k, was just making sure before I responded to the one email about it.

[11:18:00.0000] <TabAtkins_>
Well, hrm.  I still don't know enough to respond to the email correctly.  :/

[11:20:00.0000] <jgraham>
/me discovers that magic from-the-intertubes package delivery+installation isn't so good when the tubes get diconnected

[12:06:00.0000] <mbelshe>
annevk: yt?

[14:05:00.0000] <cardona507>
College4

[14:11:00.0000] <AryehGregor>
Neat, now Firefox has resizable text areas in 3.7.  That's one of the features I actually notice missing when I use Firefox instead of Chrome.

[14:12:00.0000] <annevk>
mbelshe, am now

[14:20:00.0000] <TabAtkins>
Yay!

[14:20:01.0000] <TabAtkins>
/me loves resizable text areas.

[14:21:00.0000] <annevk>
mbelshe, ah, you mailed

[14:23:00.0000] <cardona507>
resizable text areas are pretty sweet 

[14:44:00.0000] <jgraham>
/me wonders if setting pupils on a per-subject basis according to ability is legal in Sweden

[14:54:00.0000] <AryehGregor>
What do you mean?

[15:00:00.0000] <jgraham>
AryehGregor: Me?

[15:02:00.0000] <AryehGregor>
Yes, you.

[15:03:00.0000] <jgraham>
Having a setup where, for each subject, pupils are grouped according to their ability in that subject, and each ability group is taught seperately (typically for secondary eduction i.e. 11-16 but potentially for younger or older pupils too)

[15:04:00.0000] <AryehGregor>
You mean would it be legal for a private school to do that?  I can't see why not.  Or a public one?

[15:05:00.0000] <Dashiva>
In Norway the problem is bypassed with elective courses

[15:06:00.0000] <AryehGregor>
In America it's pretty typical to tier students by ability to some extent.

[15:07:00.0000] <AryehGregor>
Not that I really know, I went to a yeshiva.

[15:07:01.0000] <Dashiva>
(But we also have an acknowledged problem of gifted students being neglected in many places)

[15:08:00.0000] <jgraham>
AryehGregor: It is pretty typical in England too, especially in Maths, English and Science

[15:09:00.0000] <jgraham>
I caan't tell if it happens at all here but I at least know of schools that don't set in maths, which is very uncommon in the UK

[15:10:00.0000] <jgraham>
(because of the remarkable difficulty of teaching to the whole range of abilities without making half the class suicidal through either boredom or frustration)

[15:10:01.0000] <jgraham>
(that is, it is very uncommon not to set for maths in the UK)

[15:11:00.0000] <jgraham>
(or removing the double negative, it is very common to set)

[15:12:00.0000] <jgraham>
AryehGregor: I mean any school. The system of public/private schools is not quite the same as the US I think

[15:13:00.0000] <jgraham>
In particular there seem to be laws that restrict teaching practices in both

[15:45:00.0000] <Clubbed>
hi

[15:49:00.0000] <Clubbed>
why label="" attribute instead of tag? an attribute can contain only text, no markup... i mean that title="" attribute should be <tooltip>... attributes should contains technical informations like urls or ids classes, not visible text

[15:50:00.0000] <Hixie>
label=""?

[15:50:01.0000] <Clubbed>
yes

[15:50:02.0000] <Clubbed>
<menu type="context-menu" label="File">

[15:50:03.0000] <Hixie>
oh on <menu>

[15:51:00.0000] <Hixie>
because operating systems typically just ask for a single string for the menu labels

[15:53:00.0000] <Clubbed>
uhm

[15:53:01.0000] <Clubbed>
<label>ddd<b>d</b></label>    label.textContent

[15:55:00.0000] <Clubbed>
limit ui designers means that no one will use label attribute

[15:56:00.0000] <Clubbed>
this is just my opinion!

[15:56:01.0000] <Clubbed>
that do you think about it hixie

[15:57:00.0000] <Clubbed>
*what

[15:57:01.0000] <Clubbed>
what do you think about

[15:58:00.0000] <Hixie>
i think if people want a context menu they'll use it :-)

[15:59:00.0000] <Clubbed>
no they will use a random old dhtml javascript context menu :D

[16:00:00.0000] <TabAtkins>
Why?  So they can use <b> in their context menu?

[16:00:01.0000] <TabAtkins>
/me wants everyone to start being able to use <menu> yesterday.

[16:00:02.0000] <Clubbed>
yes, it is just an example

[16:00:03.0000] <Clubbed>
just like icon="" attribute

[16:01:00.0000] <Clubbed>
if i want to use two icons 

[16:01:01.0000] <Clubbed>
?

[16:01:02.0000] <Clubbed>
and

[16:01:03.0000] <TabAtkins>
Your context menu is, fundamentally, a limited thing.  When you right-click on things, you get a series of plain text labels that execute some command when you click on it.

[16:01:04.0000] <Hixie>
context menus on windows and mac don't generally have bold text and so forth, why would web authors want something fundamentally different?

[16:03:00.0000] <TabAtkins>
On any distro of Linux that I've personally experienced, too.

[16:03:01.0000] <Clubbed>
tabatkins 

[16:03:02.0000] <annevk>
Opera context menus at least have a concept of bold

[16:03:03.0000] <annevk>
select some text, right click, see under "search with"

[16:03:04.0000] <Clubbed>
webapp should be independent (but integrated in the os)

[16:04:00.0000] <Clubbed>
so if the need is a text-only label

[16:04:01.0000] <Clubbed>
the syntax should be

[16:04:02.0000] <annevk>
i could imagine that if we ever "replace" the OS people would want to experiment beyond somewhat old-fashioned limitations too

[16:04:03.0000] <Clubbed>
<label value="String" />

[16:04:04.0000] <annevk>
but presumably you would not use <menu> then

[16:04:05.0000] <Clubbed>
so i can put icons around it

[16:05:00.0000] <TabAtkins>
Again, though, you don't see icons in right-click menus.  They're just plain text strings.

[16:05:01.0000] <annevk>
Opera has icons in context menus

[16:05:02.0000] <TabAtkins>
Dammit, Opera!

[16:05:03.0000] <TabAtkins>
Why you gotta be complicating things.

[16:06:00.0000] <annevk>
we're full of win

[16:06:01.0000] <Clubbed>
they are not complicating

[16:06:02.0000] <Clubbed>
*complicated

[16:06:03.0000] <Clubbed>
fkg language difficulties xD

[16:07:00.0000] <Clubbed>
anyway tabatkins 

[16:07:01.0000] <Hixie>
annevk: i wouldn't look at Opera for UI advice. :-P

[16:07:02.0000] <Hixie>
Clubbed: <menu> supports icons in context menus

[16:07:03.0000] <Clubbed>
yes but just only one

[16:07:04.0000] <Hixie>
um yes

[16:07:05.0000] <Hixie>
one per command

[16:07:06.0000] <Clubbed>
it's just an example of a limitation

[16:07:07.0000] <Hixie>
what's the use case for more than one per command?

[16:08:00.0000] <Clubbed>
i dont need more icons, but i think visible content should be placed inside tags, not attributes

[16:09:00.0000] <Clubbed>
for a lot of reasons

[16:09:01.0000] <Clubbed>
if i want to style only the label=""

[16:09:02.0000] <Clubbed>
how i should do?

[16:09:03.0000] <Hixie>
you can't style a context menu

[16:09:04.0000] <Hixie>
it's an operating system widget

[16:09:05.0000] <Hixie>
it's styled by the operating system

[16:10:00.0000] <Clubbed>
but is what designers want?

[16:10:01.0000] <Clubbed>
for example

[16:10:02.0000] <Clubbed>
facebook dropdown menus

[16:10:03.0000] <Hixie>
users are more important than designers, and users want their menus to be the same across all their applications

[16:10:04.0000] <Clubbed>
will use <menu> tag?

[16:11:00.0000] <Clubbed>
i think not

[16:11:01.0000] <Clubbed>
they will continue to use lists

[16:11:02.0000] <Clubbed>
or nav

[16:11:03.0000] <Hixie>
hey does anyone recall whether <time> is supposed to get replaced by its contents? I can't figure out where we last discussed this.

[16:11:04.0000] <TabAtkins>
I won't.  I'd shoot people, right in the face, if that's what it took to get an OS-native right-click menu in my webapps.

[16:12:00.0000] <Clubbed>
tabatkins

[16:12:01.0000] <TabAtkins>
Hixie: ?_?  You mean, like, converted automagically into the user's time settings?

[16:12:02.0000] <Hixie>
man, TabAtkins must have missed the "Life of Don't Be Evil" class

[16:13:00.0000] <TabAtkins>
I did! o_O

[16:13:01.0000] <Hixie>
TabAtkins: i mean <time datetime="foo"></time> being turned into locale-specific rendering of foo

[16:13:02.0000] <Clubbed>
my os controls must appear exactly like os skin defined they

[16:14:00.0000] <Clubbed>
i'm talking about context menus on gmail, on facebook, on wave, and toolbars too

[16:14:01.0000] <TabAtkins>
Hixie: I don't recall discussions about this, but that would be great.

[16:14:02.0000] <TabAtkins>
Clubbed: I'm also talking about those.

[16:14:03.0000] <TabAtkins>
And I want them to look like a right-click menu.

[16:14:04.0000] <TabAtkins>
Just like every single other application on my entire machine.

[16:16:00.0000] <TabAtkins>
I mean, I *actively desire* there to be no significant styling control of those things.  Just like I don't want styling control of scrollbars.

[16:16:01.0000] <Clubbed>
ok, but is our "nerd" opinion, if you want to build an os-like application we can use xul, but "commercial" applications needs more styling and scripting features

[16:17:00.0000] <JonathanNeal>
BUT I WANNA MAK MA SCROLLBAHS PUPLE!

[16:17:01.0000] <Clubbed>
lolsdjfgsdf

[16:17:02.0000] <TabAtkins>
Apparently they don't.  Commercial applications use ordinary right-click menus.  And ordinary scrollbars.  The *very* few that go out of their way to create custom versions of those are extremely annoying, though luckily quite rare.

[16:18:00.0000] <Hixie>
annevk: do you recall the current state of the <time> thing by any chance? whether we decided to make it not localise or something?

[16:18:01.0000] <Hixie>
the spec right now is inconsistent on the matter

[16:18:02.0000] <Hixie>
and i can't work out if it's just that i missed something while adding it or while removing it

[16:19:00.0000] <Clubbed>
tabatkins i see a lot of custom widget that they can not be reproduced with html5 markup

[16:19:01.0000] <Clubbed>
for example on wpf

[16:19:02.0000] <Clubbed>
i see textboxes inside the contextmenu

[16:20:00.0000] <TabAtkins>
I've never seen such a thing in any application I've ever used.  :shrug:

[16:21:00.0000] <Clubbed>
me too, but i use only notepad++ and eclipse lol so maybe my idea of an application is "limited"

[16:21:01.0000] <TabAtkins>
I use plenty, and speak as a web designer who likes making pretty things.

[16:23:00.0000] <Clubbed>
anyway, you are right, people can't change color of our scrollbars but they do, so they will do the same with context menus, they will never use <menu>

[16:24:00.0000] <Clubbed>
xhtml teachs to us how it works

[16:24:01.0000] <TabAtkins>
I can't parse what you just said.

[16:24:02.0000] <Clubbed>
i'm sorry

[16:24:03.0000] <TabAtkins>
People do use scrollbars.  They're a basic part of UI.

[16:25:00.0000] <TabAtkins>
And people do use context menus.  They're all over the place.  The one place they're missing is in webapps.

[16:25:01.0000] <Clubbed>
people cannot change the color of my browser scrollbar, but they inevitably do, with scripts

[16:25:02.0000] <Clubbed>
or flash

[16:26:00.0000] <Clubbed>
and they will use scripts instead of <menu>

[16:26:01.0000] <TabAtkins>
Very few people use scripts for context menus now.  I doubt that will change.

[16:26:02.0000] <TabAtkins>
Adding <menu> will make context menus super easy to make, so I think you'll see them a lot more.

[16:26:03.0000] <Hixie>
very few people restyle their scrollbars. some do, but not many.

[16:26:04.0000] <Hixie>
the same will apply to <menu>.

[16:26:05.0000] <Clubbed>
... :| gmail does wave does everything on google does, social networks too

[16:27:00.0000] <Hixie>
some will go out of their way to do special menus, but most will do the right thing and use the OS default

[16:27:01.0000] <TabAtkins>
Gmail doesn't.

[16:27:02.0000] <TabAtkins>
Some of the Google apps do, but they're not doing anything that can't be handled by <menu>.

[16:28:00.0000] <Clubbed>
exactly what i'm saying, developers will continue to use absolutely positioned divs 

[16:28:01.0000] <Clubbed>
for emulate context menus

[16:28:02.0000] <TabAtkins>
That's the opposite of what I just said.  I said that the Google apps that *do* use custom right-click menus can migrate to using <menu> without any problem.

[16:29:00.0000] <Clubbed>
lol

[16:30:00.0000] <Hixie>
Clubbed: if your hypothesis was correct -- that is, if authors liked styling widgets so much that <menu> will fail -- then one could predict that authors would uniformly restyle scrollbars instead of using the default rendering for them

[16:30:01.0000] <Hixie>
Clubbed: however, that prediction does not match reality

[16:30:02.0000] <Clubbed>
i'm sorry but you do not understand... google wave is new, it uses custom scrollbars

[16:30:03.0000] <Clubbed>
so

[16:30:04.0000] <Hixie>
Clubbed: therefore the hypothesis is wrong

[16:30:05.0000] <TabAtkins>
Frex, take a look at Google Maps.  Custom right-click menu.  And yet, it has nothing but plain-text labels.

[16:30:06.0000] <Clubbed>
why did they not use os scrollbars?

[16:30:07.0000] <Hixie>
Clubbed: now certainly, SOME pages will make custom scrollbars and custom menus and so forth, but that's fine

[16:30:08.0000] <Clubbed>
it's a commercial thing

[16:30:09.0000] <TabAtkins>
They could swap to <menu> immediately once it was supported widely enough.

[16:30:10.0000] <Hixie>
Clubbed: the point is that MOST will not

[16:30:11.0000] <Clubbed>
exactly!

[16:31:00.0000] <Clubbed>
they could but they will not swap to <menu>!

[16:31:01.0000] <Hixie>
Clubbed: MOST pages will just use the default menu, just like they use the default scrollbar

[16:31:02.0000] <Clubbed>
never! they want custom design!

[16:31:03.0000] <Hixie>
who, the wave team?

[16:32:00.0000] <Clubbed>
everyone, included me

[16:32:01.0000] <Hixie>
so why doesn't everyone do their own scrollbars?

[16:32:02.0000] <Hixie>
(on a sidenote, i just hit r5000!)

[16:34:00.0000] <Clubbed>
just because script is hard to make

[16:34:01.0000] <TabAtkins>
Wave's custom scrollbars are *really weird* and very hard to use if there is a lot of overflow.  I need to go complain at them.

[16:35:00.0000] <TabAtkins>
Clubbed: You don't need scripting to change the color of your scrollbars.  IE supports some proprietary scrollbar CSS properties.

[16:35:01.0000] <Hixie>
yeah it's actually pretty easy to change scrollbar colorus

[16:36:00.0000] <Clubbed>
the point is not this, i dont want to change my context menu style or scrollbar style, never

[16:36:01.0000] <Clubbed>
but

[16:37:00.0000] <Clubbed>
i think developers dont want to be limited

[16:37:01.0000] <Clubbed>
come on

[16:37:02.0000] <Clubbed>
the web is commercial

[16:37:03.0000] <Clubbed>
is not a desktop app

[16:37:04.0000] <Clubbed>
it must be controlled but not limited

[16:38:00.0000] <TabAtkins>
Developers not wanting to be limited is unimportant.  Does it help or harm users?

[16:38:01.0000] <TabAtkins>
If it harms them, then it's out.

[16:39:00.0000] <Hixie>
developers aren't limited here anyway, if they don't want their menus to look like native menus, they're welcome to write their own menus with script, css, and aria

[16:39:01.0000] <Hixie>
just like they do today

[16:39:02.0000] <Hixie>
the point of <menu> is specifically to let them use the default OS styles

[16:39:03.0000] <Hixie>
to fit in better

[16:39:04.0000] <TabAtkins>
And it's probably not a good idea to defend things that you don't like anyway.  You can come up with much better justifications if you personally want the change you're arguing for.

[16:40:00.0000] <Clubbed>
hixie

[16:40:01.0000] <Hixie>
clubbed

[16:40:02.0000] <Clubbed>
if i want to use the default os style

[16:40:03.0000] <Clubbed>
i can use xul

[16:40:04.0000] <Clubbed>
embedded

[16:40:05.0000] <Clubbed>
with a namespace

[16:40:06.0000] <TabAtkins>
No you can't.

[16:40:07.0000] <Clubbed>
for example

[16:40:08.0000] <Hixie>
you can also use <menu>

[16:41:00.0000] <TabAtkins>
Not in HTML, at least.  Maybe some browsers let you embed XUL in XHTML, I dunno.

[16:41:01.0000] <Clubbed>
tabatkins im using html5 as xml with xul on firefox and it works good

[16:42:00.0000] <Clubbed>
with xhtml dtd obviously

[16:42:01.0000] <TabAtkins>
I don't want to use XML.  And I want my websites to work outside of Firefox.

[16:43:00.0000] <Clubbed>
thats not the point! xD

[16:43:01.0000] <Clubbed>
anyway closed discussion

[16:43:02.0000] <TabAtkins>
Sure it is, since you just said that the way to have OS-native context menus is to use embedded XUL.

[16:45:00.0000] <Clubbed>
we will talk again in... five years?

[16:45:01.0000] <TabAtkins>
If you're trying to change the design of <menu>, 5 years from now will be too late.  ^_^

[16:45:02.0000] <Clubbed>
the prevision is: developers will continue to try to override default os controls

[16:46:00.0000] <Clubbed>
with custom divs and scripts and random shit

[16:46:01.0000] <TabAtkins>
I agree that some will.  I think that the vast majority won't, because an OS-native display is either exactly what they want (to blend in with every other context menu the user sees on their machine) or is sufficient (because they just want the functionality, and don't care about the appearance as long as it's "good enough").

[16:47:00.0000] <Clubbed>
but default functionality, predefined graphics and scripting should be override-able in some cases

[16:48:00.0000] <Clubbed>
i dont want to customize the style of the contextmenu

[16:48:01.0000] <TabAtkins>
The important question is, why should they?  Sometimes there are good reasons to allow such, sometimes not.

[16:48:02.0000] <Clubbed>
i want to put two icons for each item!

[16:49:00.0000] <Clubbed>
i want to put a textfield inside my toolbar

[16:49:01.0000] <TabAtkins>
Create a single image with both icons in it.

[16:49:02.0000] <Clubbed>
etc

[16:49:03.0000] <TabAtkins>
Or, you know, don't.

[16:49:04.0000] <Clubbed>
lol, just examples

[16:49:05.0000] <Clubbed>
title="" is limited

[16:49:06.0000] <Clubbed>
developers uses hidden div

[16:50:00.0000] <Clubbed>
to markup tooltips

[16:50:01.0000] <TabAtkins>
I know you can come up with examples of what someone could *possibly* do with more freedom.  That's unimportant.  We need to know if it's *important* to be able to do something that is currently restricted.

[16:54:00.0000] <Hixie>
i thought you said you didn't? <Clubbed> the point is not this, i dont want to change my context menu style or scrollbar style, never

[16:54:01.0000] <Hixie>
i'm confused

[16:54:02.0000] <Hixie>
do you want to change the style of <menu> ornot?

[16:54:03.0000] <Hixie>
and if you do, what is the use case

[16:55:00.0000] <Clubbed>
hixie, i can use html5 with no problems without styling at all and without custom labels and icons

[16:56:00.0000] <Clubbed>
but a lot of people will

[16:56:01.0000] <Clubbed>
xhtml fails for this

[16:56:02.0000] <Clubbed>
so

[16:56:03.0000] <Clubbed>
html5 should be semantic

[16:56:04.0000] <Clubbed>
with no limitations

[16:56:05.0000] <Clubbed>
for example

[16:57:00.0000] <Clubbed>
if html5 do <scrollbar>

[16:57:01.0000] <Clubbed>
developers can use <scrollbar> to customize the look

[16:57:02.0000] <Hixie>
the main problem i have with all that is the "a lot of people" part. What evidence do you have that it is a lot of people?

[16:58:00.0000] <Clubbed>
and UA can restrict by settings their styling

[16:58:01.0000] <Hixie>
I contend that it is not al ot of people, and i put forward as evidence the fact that not many people style their scrollbars, even though doing so is easy.

[16:58:02.0000] <TabAtkins>
Thus the reason why I said you shouldn't argue for things that you don't, personally, want.  Arguing for other people is always fraught with peril.


2010-04-13
[17:00:00.0000] <Hixie>
(if it _is_ "a lot of people" then you're right, we should investigate this further.)

[17:01:00.0000] <Clubbed>
tab, i'm building a webos,  personally i want, but is not absolutely necessary the styling, but if i want to put a special control like a text input in my toolbar, or a checkbox in my contextmenu?

[17:01:01.0000] <Clubbed>
how i do it

[17:01:02.0000] <Clubbed>
?

[17:01:03.0000] <Hixie>
a "webos"?

[17:01:04.0000] <Clubbed>
i will use an absolutely positioned div to emulate context menu

[17:01:05.0000] <Clubbed>
and a <ul>

[17:02:00.0000] <Clubbed>
for the toolbar

[17:02:01.0000] <Clubbed>
just old html4 

[17:02:02.0000] <Hixie>
sounds good

[17:02:03.0000] <Hixie>
what's the problem

[17:03:00.0000] <Clubbed>
the problem is that like me other people never will accept <menu>

[17:03:01.0000] <Clubbed>
they will continue to use divs

[17:03:02.0000] <Hixie>
what is your evidence that you are in the majority here?

[17:03:03.0000] <Hixie>
i've provided what i consider evidence that this is likely a minority position

[17:03:04.0000] <Hixie>
what is your counter-evidence?

[17:04:00.0000] <Clubbed>
uhm sorry i cannot understand this

[17:04:01.0000] <Clubbed>
my english is poor

[17:04:02.0000] <Clubbed>
try to write simply sorry

[17:04:03.0000] <Clubbed>
*simpler

[17:04:04.0000] <Hixie>
why do you think you more people want styling than not want styling

[17:05:00.0000] <Hixie>
more people want their scrollbars to be not styled than want them to be styled, if we look at the number of people styling their scrollbars

[17:05:01.0000] <Hixie>
so why do you think the opposite will be true for menus

[17:06:00.0000] <Clubbed>
uhm actually i think the same for the two cases

[17:07:00.0000] <Clubbed>
but styling is not adding an icon

[17:07:01.0000] <Clubbed>
my style can not be overriden, for example on <select>

[17:07:02.0000] <Clubbed>
but

[17:07:03.0000] <Clubbed>
in <option></option> of selects i want to put buttons

[17:07:04.0000] <Clubbed>
textareas

[17:07:05.0000] <Clubbed>
i want to put multiple icons

[17:08:00.0000] <Clubbed>
thats the point

[17:09:00.0000] <Hixie>
why?

[17:10:00.0000] <Clubbed>
why i can need it...

[17:11:00.0000] <Clubbed>
if the standard is the standard desktop toolkit

[17:11:01.0000] <Clubbed>
UIs will never go on

[17:11:02.0000] <Clubbed>
always the same

[17:12:00.0000] <TabAtkins>
The tools necessary to innovate in UI development already exist - use javascript and CSS, plus some accessibility help like ARIA.

[17:13:00.0000] <TabAtkins>
Then, *after* we've established that something new is wanted by a lot of people, we can add it to the language.

[17:13:01.0000] <TabAtkins>
Trying to guess what requirements people will have for new innovations is doomed to failure, and will just result in a lot of useless bloat in the spec that will end up getting in people's way more than it helps them.

[17:14:00.0000] <Clubbed>
ok but html5 is new, html4 is the experiment

[17:15:00.0000] <TabAtkins>
Sure, for some things it was.  Now HTML5 is the experiment for a bunch of new things.

[17:16:00.0000] <Clubbed>
anyway ok... i will continue to use divs xD

[17:17:00.0000] <TabAtkins>
You'll have to for a while anyway, until browsers support <menu>.

[17:19:00.0000] <Clubbed>
ive written implementation for menu using css bindings but specs, as you see lol, does not match my needs of a really good tool kit

[17:20:00.0000] <Clubbed>
anyway i will match aria's things if possible

[17:22:00.0000] <Clubbed>
ah another question... what WG will do about clipboard

[17:23:00.0000] <Clubbed>
something like drag&drop?

[17:24:00.0000] <Clubbed>
i know security issues about clipboard

[17:28:00.0000] <KaOSoFt>
http://tech.slashdot.org/story/10/04/12/1816205/Google-Rebuilds-Docs-Platform

[17:34:00.0000] <Clubbed>
tabatkins -> http://img199.imageshack.us/img199/3540/googleep.png

[17:35:00.0000] <Clubbed>
going to bed

[17:35:01.0000] <Clubbed>
good night

[17:36:00.0000] <Hixie>
i suppose we could have told him that tab and i work for google

[17:36:01.0000] <TabAtkins>
Eh, no reason.

[17:37:00.0000] <Hixie>
in case he comes back, the image he pasted is a good use case, but one we can easily support in the future by adding more features to <menu>

[17:37:01.0000] <Hixie>
it doesn't argue for arbitrary styling or anything like that

[17:37:02.0000] <TabAtkins>
Man, I'm tired of writing this counter-proposal for Issue 100.  It's annoying have to parse out actual objections from a wall of text, and further to separate objections to @srcdoc from the (much more numerous) objections to @sandbox itself.

[17:40:00.0000] <Hixie>
it's hard to write change proposals in general at the moment since we don't have much to go on in terms of figuring out how they'll be handled by the chairs

[17:41:00.0000] <Hixie>
in fact iirc all we have is the microdata one

[17:42:00.0000] <Hixie>
where a long list of good reasons didn't convince the chairs

[17:42:01.0000] <Hixie>
it's hard to know if a shorter list or a longer list would have had a different result

[17:48:00.0000] <JonathanNeal>
Was there a dialog element?

[17:48:01.0000] <JonathanNeal>
And was it removed?

[17:49:00.0000] <Hixie>
yes and yes

[17:51:00.0000] <JonathanNeal>
should embed be considered a new element?

[17:54:00.0000] <TabAtkins>
No, it's existed for a long time.

[17:56:00.0000] <JonathanNeal>
Hmm, it was listed @ http://www.w3.org/TR/html5-diff/#new-elements

[17:56:01.0000] <JonathanNeal>
Anyway, I was just trying to update the wikipedia page for new elements.

[17:57:00.0000] <TabAtkins>
It's new in the sense that it didn't exist in HTML4.  But it was certainly used back then.

[17:58:00.0000] <JonathanNeal>
I ended up just re-ordering two of them to make them alphabetical (you can't have all but two in order!) and removed dialog.

[18:00:00.0000] <JonathanNeal>
I think the deprecated list could be larger too.

[18:06:00.0000] <JonathanNeal>
okay, updated both... so productive :-P

[18:09:00.0000] <TabAtkins>
Sent Issue 100 counter proposal.  Now to look at Issue 92 and see if it's objectionable or not...

[18:17:00.0000] <TabAtkins>
/me finds no real reason to object to Issue 92.

[18:22:00.0000] <Hixie>
i wish i understood 92

[18:22:01.0000] <TabAtkins>
Basically, the example table you've used is unrealstic and contrived.  She's proposing a better example table.

[18:23:00.0000] <Hixie>
but that's not what the bug she escalated it from is asking for

[18:23:01.0000] <TabAtkins>
Complex enough to justify explanatory text about the headers, but still relatively simple to understand.

[18:23:02.0000] <TabAtkins>
Indeed, the bug is something different entirely.

[18:23:03.0000] <Hixie>
o_O

[18:23:04.0000] <TabAtkins>
Yeah, dunno.

[18:24:00.0000] <Hixie>
any idea which section she's trying to replace?

[18:24:01.0000] <Hixie>
i don't know what she means by "table element description section"

[18:25:00.0000] <TabAtkins>
In the description of the <table> element.

[18:25:01.0000] <Hixie>
oh

[18:25:02.0000] <Hixie>
what the whole thing?

[18:25:03.0000] <TabAtkins>
I'd have to look again.  I *think* it's just the example table and related verbage?

[18:25:04.0000] <Hixie>
oh i see, she's trying to readd the summary="" attribute without telling anyone about it

[18:26:00.0000] <TabAtkins>
Well, yeah.

[18:26:01.0000] <TabAtkins>
Oh man, skipped over that entire paragraph.

[18:27:00.0000] <Hixie>
wow her example is terrible

[18:27:01.0000] <TabAtkins>
I think it's a pretty good table to use, at least.

[18:27:02.0000] <Hixie>
except it doesn't need a tabel description

[18:27:03.0000] <Hixie>
since it can be entirely derived from the table structure

[18:28:00.0000] <Hixie>
and she put a footer in a table cell

[18:28:01.0000] <Hixie>
ok well we'll see what the chairs do with this

[18:30:00.0000] <Hixie>
i like how she managed to file a bug to remove an example on the grounds that it is extraneous, and turned it into a change proposal that instead adds a new attribute and a bigger example

[18:33:00.0000] <TabAtkins>
I don't find it fundamentally wrong to have a CP arguing for a *better* example as a suggested fix for an issue saying the example is bad and should be removed.

[18:33:01.0000] <TabAtkins>
"should be removed" is a possibe solution, sure, but so is making the example less bad.

[18:37:00.0000] <Hixie>
if all she was doing is suggesting a better example, i wouldn't have rejected the bug

[18:37:01.0000] <Hixie>
she didn't ask for a better example, nor provide one

[18:38:00.0000] <Hixie>
and then she escalated this editorial bug and turned it into a way to smuggle the summary="" attribute back in

[18:39:00.0000] <Hixie>
(also, the proposedw table isn't better -- it doesn't need a table description at all, since it can be entirely derived from the table structure, and it has a footer in a cell)

[18:39:01.0000] <Hixie>
i wish we did have a better table though

[18:42:00.0000] <Hixie>
(another reason her table is a poor one is that even if it was one that needed a table description, it is far too long to repeat several times showing different ways of exposing the table description)

[19:46:00.0000] <MikeSmith>
boblet: I removed the Details section from the doc for nav 

[19:46:01.0000] <MikeSmith>
I can re-add something there again later

[21:00:00.0000] <othermaciej>
hi everyone

[21:29:00.0000] <MikeSmith>
othermaciej: hey

[21:30:00.0000] <othermaciej>
hi MikeSmith!

[21:31:00.0000] <MikeSmith>
catching up with some unread mail and just reading about new webkit multiprocess stuff now

[21:42:00.0000] <othermaciej>
oh yeah that is neat stuff :-)

[21:46:00.0000] <wirepair>
whats this now?

[21:46:01.0000] <wirepair>
ah like chrome

[21:47:00.0000] <wirepair>
wonder if they will sandbox, or just keep split

[21:48:00.0000] <othermaciej>
split first

[21:49:00.0000] <wirepair>
yeah looks like they make hints at sandboxing

[21:49:01.0000] <wirepair>
"This can deliver benefits in responsiveness, robustness, security (through the potential to sandbox the web process) and better use of multicore CPUs"

[21:49:02.0000] <othermaciej>
/me is one of them

[21:49:03.0000] <wirepair>
oh ;>

[21:50:00.0000] <wirepair>
sorry i just kind of lurk here, not really up to speed with who everyone is and where they are at ;>

[21:51:00.0000] <othermaciej>
no worries

[21:51:01.0000] <othermaciej>
but you can ask questions if you want, here or on #webkit :-)

[21:51:02.0000] <wirepair>
good to know thanks ;>

[22:30:00.0000] <MikeSmithX>
the cafeteria at work is serving whale meat this semester

[22:35:00.0000] <othermaciej>
MikeSmithX: :-

[22:35:01.0000] <othermaciej>
:-(

[22:36:00.0000] <MikeSmithX>
yeah

[22:37:00.0000] <wirepair>
make them all watch that movie ;>

[22:37:01.0000] <wirepair>
or ask them if they are eating the meat for research

[01:30:00.0000] <Hixie>
MikeSmithX: your input on bug 9467 would be very welcome

[01:32:00.0000] <MikeSmith>
me looks around for Yudai

[01:32:01.0000] <MikeSmith>
Yudai, http://www.w3.org/Bugs/Public/show_bug.cgi?id=9467

[01:33:00.0000] <MikeSmith>
Hixie: fwiw, "on" and "kun" are different phonetic readings for kanji

[01:33:01.0000] <MikeSmith>
I still don't understand the question there either, though

[01:33:02.0000] <MikeSmith>
/me looks more

[01:33:03.0000] <hsivonen>
argh. html5test.com tests SQL database and encumbered codecs as part of its "HTML5" score

[01:34:00.0000] <Hixie>
MikeSmith: well if you don't understand it, i'm certainly going to be way out of my depth :-)

[01:34:01.0000] <hsivonen>
so stuff like this emerges: http://geektechnica.com/2010/04/chrome-passes-html5-test-with-flying-colors/

[01:34:02.0000] <Hixie>
MikeSmith: hopefully they'll comment further...

[01:34:03.0000] <Hixie>
yeah that test is a disaster

[01:34:04.0000] <othermaciej>
hsivonen: it also appears to have totally buggy tests

[01:35:00.0000] <Hixie>
i commented on reddit to that effect, but not sure what else can be done about it

[01:35:01.0000] <MikeSmith>
Hixie: I think you could just change it from "kanji reading" to "phonetic reading"

[01:35:02.0000] <othermaciej>
for example, it tells me my Safari supports Ogg Vorbis but not MP3 or AAC

[01:35:03.0000] <Hixie>
MikeSmith: k

[01:36:00.0000] <othermaciej>
also seems to think we support the <mark> element

[01:36:01.0000] <MikeSmith>
hsivonen: btw, I think I found a bug in the v.nu assertions code that is causing processChildContent() from ever being called at all

[01:36:02.0000] <hsivonen>
MikeSmith: oops...

[01:37:00.0000] <MikeSmith>
but it also seems like maybe the processChildContent stuff is dead code, anyway

[01:37:01.0000] <hsivonen>
which reminds me, that I was planning on redeploying V.nu yesterday

[01:37:02.0000] <hsivonen>
so I guess I should do it now

[01:37:03.0000] <Yudai>
I'm still stacking on rebuilding :)

[01:38:00.0000] <MikeSmith>
hsivonen: eclipse warns me: "The method isChildren() from the type Assertions.StackNode is never used locally ... The local variable parentName is never read ... The local variable add is never read"

[01:39:00.0000] <MikeSmith>
hsivonen: if something breaks when you rebuild, it may be because I forgot to commit something from my workspace

[01:39:01.0000] <MikeSmith>
Yudai is try to build from current source and says his build is failing 

[01:40:00.0000] <Yudai>
MikeSmith: wait a moment

[01:40:01.0000] <hsivonen>
I'm building now on Ubuntu using OpenJDK

[01:41:00.0000] <hsivonen>
the build works for me

[01:41:01.0000] <MikeSmith>
cool

[01:42:00.0000] <hsivonen>
hmm. I think I'll land my option/datalist fix in the parser repo even though it lacks review for Gecko

[01:42:01.0000] <MikeSmith>
hsivonen: btw, I don't know if you saw, but Sean Fraser was trying to put together a ruby gem of v.nu and raised a bug to report that the build initially failed for him

[01:43:00.0000] <MikeSmith>
because of the known issue of needing to run the build twice after a initial/fresh checkout

[01:43:01.0000] <hsivonen>
yeah, the build script needs polishing

[01:43:02.0000] <hsivonen>
I'm personally annoyed at having to set JAVA_HOME

[01:43:03.0000] <MikeSmith>
yeah

[01:44:00.0000] <wirepair>
mike,you manage the html5.validator.nu?

[01:44:01.0000] <MikeSmith>
no

[01:44:02.0000] <hsivonen>
wirepair: I do

[01:44:03.0000] <wirepair>
ah, are you takign steps to block repeated requests/requests to localhost?

[01:45:00.0000] <wirepair>
just out of curiousity ;>, i think i spoke with hixie about this

[01:45:01.0000] <hsivonen>
whoa. what requests to localhost?

[01:45:02.0000] <wirepair>
so if anattacker does

[01:45:03.0000] <MikeSmith>
Hixie: somebody may likely complain about "phonetic reading", but I don't know any other way to explain it easily (without going into further confusing/esoteric/useless-in-this-context details about various possible readings of kanji in Japanese)

[01:45:04.0000] <wirepair>
http://localhost:8080 ....

[01:45:05.0000] <Hixie>
MikeSmith: k

[01:45:06.0000] <wirepair>
are you blocking things like that

[01:46:00.0000] <wirepair>
i was using it to validate google, and i notice it gave me google frances' site, so i realized the validator could basically be used as a proxy

[01:48:00.0000] <MikeSmith>
Hixie: http://dev.w3.org/html5/spec/dom.html#inter-element-whitespace means that <figcaption>   </figcaption> does not contain content other than inter-element whitespace, right?

[01:48:01.0000] <MikeSmith>
because "Empty text nodes and text nodes consisting of just sequences of those characters are considered inter-element whitespace"

[01:49:00.0000] <Hixie>
yes, unless i'm missign something

[01:49:01.0000] <MikeSmith>
OK

[01:52:00.0000] <MikeSmith>
hsivonen: btw, I'm curious if you might remember why you have ignorableWhitespace implementation in the Checker code calling the characters method

[01:53:00.0000] <hsivonen>
MikeSmith: it's always legal per SAX to delegate ignorableWhitespace to characters if you aren't ignoring anything

[01:53:01.0000] <MikeSmith>
oh, OK

[01:53:02.0000] <MikeSmith>
didn't know that

[01:54:00.0000] <hsivonen>
wirepair: no, localhost isn't blocked

[01:54:01.0000] <wirepair>
hsivonen may want to do that ;>

[01:54:02.0000] <MikeSmith>
wirepair is a security maven

[01:54:03.0000] <hsivonen>
wirepair: localhost isn't supposed to have any services whose protection is IP-based

[01:54:04.0000] <wirepair>
so no jetty manager 

[01:54:05.0000] <wirepair>
jmx-console...

[01:54:06.0000] <wirepair>
etc/

[01:54:07.0000] <wirepair>
?

[01:55:00.0000] <wirepair>
MikeSmith nah just a concerned citizen in this case ;>

[01:55:01.0000] <hsivonen>
wirepair: there's not supposed to be any of those things

[01:55:02.0000] <wirepair>
hsivonen cool, as long as you aren't worried

[01:55:03.0000] <hsivonen>
wirepair: validator.nu uses Jetty as a library

[01:55:04.0000] <hsivonen>
wirepair: validator.nu starts a Jetty HTTP listener

[01:56:00.0000] <hsivonen>
wirepair: Jetty doesn't start validator.nu

[01:56:01.0000] <wirepair>
gotcha

[01:57:00.0000] <hsivonen>
but I agree it would be proper defense in depth to block localhost

[01:57:01.0000] <hsivonen>
I have a vague recollection that I had some kind of hand-wavy localhost blocking

[01:57:02.0000] <hsivonen>
but I removed it

[01:57:03.0000] <wirepair>
hehe, yeah also be on the look out for people abusing it

[01:58:00.0000] <wirepair>
not fun to have people requesting www.nsa.gov/somescript.jsp?blah=1' drop tables terrorists;-- ;D

[01:58:01.0000] <hsivonen>
if I readd it, I want to make sure the blocking actually works for IPv6 localhost and octal IPv4 localhost

[02:01:00.0000] <wirepair>
actually, if it were me i'd probably add an X-Forwarded-For: <origin ip>

[02:01:01.0000] <wirepair>
not sure if you are doing that already

[02:31:00.0000] <annevk>
gears of war 3 not until April 2011

[02:31:01.0000] <annevk>
sad day

[02:35:00.0000] <annevk>
did people see this: http://newteevee.com/2010/04/12/google-to-open-source-vp8-for-html5-video/

[02:35:01.0000] <zcorpan>
i wonder what the multiple sources they refer to

[02:36:00.0000] <Hixie>
annevk: you should play Bad Company 2 if you have a PS3

[02:37:00.0000] <MikeSmith>
annevk: didn't see that yet

[02:38:00.0000] <MikeSmith>
maybe the "multiple sources" are the same ones that said the Macbook refresh would happen today

[02:38:01.0000] <MikeSmith>
and they same ones who have been saying that for 2 months or whatever

[02:39:00.0000] <hsivonen>
let's see if the V.nu redeployment worked...

[02:39:01.0000] <hsivonen>
yay. it worked

[02:41:00.0000] <zcorpan>
/me wonders when xml-stylesheet checking will be deployed

[02:41:01.0000] <hsivonen>
zcorpan: has MikeSmith landed the code?

[02:41:02.0000] <zcorpan>
dunno

[02:41:03.0000] <MikeSmith>
no, I have not yet

[02:41:04.0000] <MikeSmith>
sorry

[02:41:05.0000] <MikeSmith>
hsivonen: I still need to make some refinements based on your review

[02:42:00.0000] <hsivonen>
MikeSmith: ok

[02:42:01.0000] <MikeSmith>
I will try to get back to that asap

[02:42:02.0000] <hsivonen>
anyway, the live v.nu should now match what's in the repos

[02:42:03.0000] <zcorpan>
cool

[02:44:00.0000] <hsivonen>
offtopic: has anyone purchased support from Canonical?

[02:45:00.0000] <annevk>
oh hey, Apple approved Opera Mini

[02:45:01.0000] <hsivonen>
their sales dept apparently isn't planning on telling me if a particular kind of problem is in scope of their support contract

[02:45:02.0000] <annevk>
hsivonen, ask @mpt on twitter?

[02:45:03.0000] <annevk>
/me is using Ubuntu without support

[02:45:04.0000] <hsivonen>
so I'm wondering if their support is has been any good for anyone else

[02:46:00.0000] <annevk>
well, if I have a real problem I can look up some phone numbers of people at the office that can help me out in case the interwebs has no answer

[02:47:00.0000] <zcorpan>
annevk: time for a blog post?

[02:49:00.0000] <annevk>
regarding Opera Mini you mean?

[02:49:01.0000] <zcorpan>
yeah

[02:49:02.0000] <annevk>
would make an easy filler for my one post a workday challange

[02:54:00.0000] <hsivonen>
validator.nu goes down for a kernel update

[02:55:00.0000] <hsivonen>
...and back

[02:56:00.0000] <MikeSmith>
annevk: btw, do you know of any way to get free wifi at Schiphol?

[02:56:01.0000] <MikeSmith>
the prices that they wifi provider there charges are crazy

[02:57:00.0000] <MikeSmith>
it's a relatively great airport otherwise

[03:01:00.0000] <annevk>
in the lounge...

[03:02:00.0000] <annevk>
you could sit outside the lounge and pick up the single, the password is klm100413 (for today) iirc

[03:02:01.0000] <annevk>
or maybe just kl

[03:02:02.0000] <MikeSmith>
ah, cool

[03:02:03.0000] <annevk>
s/single/signal/ doh

[03:03:00.0000] <MikeSmith>
yeah, I did actually try that when I was there -- the signal part

[03:03:01.0000] <MikeSmith>
just didn't know what password to try

[03:06:00.0000] <MikeSmith>
role values are case sensitive?

[03:07:00.0000] <hsivonen>
html5.validator.nu going down for kernel update

[03:08:00.0000] <hsivonen>
MikeSmith: IIRC, yes. They were born in the XML world.

[03:08:01.0000] <jgraham>
You make them sound like orcs

[03:11:00.0000] <hsivonen>
...and html5.validator.nu is back

[03:12:00.0000] <zcorpan>
hmm. r5032 is scary. i thought button was scoping at least in ie and opera?

[03:13:00.0000] <hsivonen>
zcorpan: button is crazy in IE

[03:13:01.0000] <hsivonen>
zcorpan: and magic in Opera

[03:13:02.0000] <hsivonen>
zcorpan: or if it's not magic, Opera allows links inside a <button> to be clicked

[03:14:00.0000] <hsivonen>
is there .deb for Opera 10.50 beta?

[03:15:00.0000] <annevk>
yeah

[03:15:01.0000] <annevk>
http://snapshot.opera.com/unix/snapshot-6306/ is the latest public build

[03:16:00.0000] <annevk>
well, it's not beta, but it's .deb and the latest build publicly available (10.51 actually)

[03:16:01.0000] <hsivonen>
annevk: thanks

[03:18:00.0000] <Hixie>
what should window.stop() do?

[03:19:00.0000] <Hixie>
cancel the fetching algorithm for the browsing context?

[03:19:01.0000] <hsivonen>
the menus in Opera 10.5x on Gnome are *much* nicer-looking than the old Windows 95-style menus

[03:20:00.0000] <mpt>
annevk, hsivonen, I don't work in the support department, sorry

[03:21:00.0000] <hsivonen>
mpt: ok. I guess I'll have to buy a 160-euro lottery ticket to make at least some progress

[03:21:01.0000] <micheil>
hey, which browsers support websockets?

[03:21:02.0000] <micheil>
 / do websockets like working from localhost?

[03:22:00.0000] <jgraham>
micheil: I don't know of any browser that implements the new handshake

[03:22:01.0000] <jgraham>
But I'm not sure

[03:22:02.0000] <jgraham>
No reason they shouldn't work on localhost

[03:22:03.0000] <annevk>
Chrome has support for -75 I think

[03:22:04.0000] <jgraham>
Ah, cool

[03:22:05.0000] <micheil>
I'm having trouble getting chrome to even send or receive data, none of the events get called

[03:22:06.0000] <micheil>
I know it connects.. but then all is silent

[03:23:00.0000] <jgraham>
Does the handshake fail?

[03:23:01.0000] <jgraham>
Does anyone here know how you test if subversion has compiled in support for gnome-keyring?

[03:24:00.0000] <micheil>
no, I've removed the handshake from my code atm

[03:24:01.0000] <micheil>
although, I have an oddity

[03:24:02.0000] <hsivonen>
hmm. Opera 10.5x on Ubuntu doesn't use my anti-aliasing settings

[03:24:03.0000] <micheil>
I connected to a websocket I have access to on pusherapp.com (ws://ws.pusherapp.com:8080)

[03:25:00.0000] <micheil>
and it worked fine

[03:25:01.0000] <jgraham>
Well if chrome really does support the handshake and you don't have it then it would fail

[03:25:02.0000] <micheil>
eg, onopen was fired, although, try doing that from localhost:7000 for the websocket address and it fails to fire onopen

[03:25:03.0000] <hsivonen>
zcorpan: hmm. http://software.hixie.ch/utilities/js/live-dom-viewer/saved/454 doesn't make the link clickable in Opera

[03:27:00.0000] <zcorpan>
hsivonen: indeed. doesn't seem to be clickable in firefox either

[03:27:01.0000] <hsivonen>
zcorpan: why are the links on http://www.gf-hunters.de/ clickable in Opera?

[03:28:00.0000] <hsivonen>
Inspect Element tells me that the links are nested inside the <button>

[03:28:01.0000] <micheil>
actually, probably the source of the error: http://code.google.com/p/chromium/issues/detail?id=36652

[03:28:02.0000] <hsivonen>
zcorpan: at least making <button> non-scoping wouldn't be randomly novel, since it isn't scoping in WebKit

[03:29:00.0000] <hsivonen>
(aside: Firefox really needs to bring FireBug into the main product. Having Inspect Element work out-of-the-box is very convenient.)

[03:30:00.0000] <annevk>
btw, you can subscribe to http://my.opera.com/desktopteam/blog/ for Opera build updates; automatic updates is not enabled for non-stable releases unfortunately

[03:30:01.0000] <hsivonen>
annevk: thanks.

[03:30:02.0000] <hsivonen>
/me notes that Chrome beta channel integrates very nicely into the Ubuntu Update Manager

[03:30:03.0000] <mpt>
hsivonen, http://www.ubuntu.com/support/services/desktopcomparison

[03:31:00.0000] <annevk>
hsivonen, yeah indeed, we should do that too, but we only do it for stable releases :/

[03:34:00.0000] <zcorpan>
hsivonen: i think that has to do with our hack where layout doesn't match the dom

[03:35:00.0000] <hsivonen>
zcorpan: well, then, that's not a suitable solution for the spec

[03:36:00.0000] <zcorpan>
hsivonen: indeed

[03:37:00.0000] <zcorpan>
hsivonen: i also noticed that opera closes button (but not object) for <ul><li><button><li>

[04:09:00.0000] <annevk>
zcorpan, done

[04:15:00.0000] <annevk>
funny: http://twitter.com/sgalineau/status/12070054472

[04:18:00.0000] <micheil>
jgraham: I've finally got chrome connecting to my server..

[04:19:00.0000] <asmodai>
Mmm, so next month VP8 will be made available under an OS-license. Interestink.

[04:24:00.0000] <annevk>
asmodai, indeed

[04:25:00.0000] <hsivonen>
annevk: when I follow the link from the tweet in Minefield, Bing warns me that my browser might not work well with Bing Maps and encourages me to install IE or Firefox

[04:26:00.0000] <annevk>
ugh, me too

[04:26:01.0000] <annevk>
/me didn't follow the link

[04:27:00.0000] <asmodai>
annevk: But what I understand Google wants to push Theora as lowest common denominator and then also make VP8 available

[04:28:00.0000] <hsivonen>
bah. Bing Maps 3D requires IE and .NET

[04:28:01.0000] <hsivonen>
still the old Microsoft

[05:02:00.0000] <Philip`>
Would it be possible to use parts of VP8 to improve Theora?

[05:03:00.0000] <Philip`>
(It's too late to change the file format but maybe there's new patented encoder techniques it could use, or something)

[05:28:00.0000] <micheil>
Hixie: hey, um, how should servers respond if the key3 is empty? eg, the client isn't using the new websocket authentication stuff

[05:41:00.0000] <jgraham>
micheil: Presumably you should drop the connection

[05:41:01.0000] <micheil>
which means no client could connect

[05:41:02.0000] <jgraham>
Well all clients have to send the 8 bytes

[05:41:03.0000] <micheil>
chrome doesn't.

[05:41:04.0000] <jgraham>
Yeah, because they need to change their implementation

[05:42:00.0000] <jgraham>
It is a breaking change

[05:42:01.0000] <micheil>
and so far it's only one of the major browsers supporting any form of things..

[05:42:02.0000] <jgraham>
Right, which is why we can still make breaking changes

[05:43:00.0000] <jgraham>
Anyway if you were to go against spec and support clients that don't do the handshake, you would be exposes to the security issues the handshake is supposed to prevent

[05:44:00.0000] <jgraham>
*exposed

[05:44:01.0000] <micheil>
yeah, true

[05:44:02.0000] <micheil>
another thing is to try and get the flash fallbacks up to speed

[05:46:00.0000] <jgraham>
Has anyone developed a test client yet?

[05:47:00.0000] <jgraham>
Like telnet with a handshake

[05:49:00.0000] <jgraham>
Oh it looks like the go langauge has a ws client in the stdlib or something

[05:49:01.0000] <beowulf>
Hixie: ps3 username?

[05:49:02.0000] <jgraham>
http://golang.org/src/pkg/websocket/client.go

[05:49:03.0000] <jgraham>
(with the handshake stuff)

[06:59:00.0000] <zcorpan>
hmm, i guess script shouldn't be a raw text element in #writing

[07:29:00.0000] <annevk>
heh, just as I say what the fuck apple they release new macbooks

[07:33:00.0000] <zcorpan>
they were waiting for you

[07:34:00.0000] <annevk>
makes sense

[07:36:00.0000] <MikeSmith>
o

[07:36:01.0000] <MikeSmith>
cool

[07:37:00.0000] <MikeSmith>
I've not had the opportunity for a long time to purchase a first-generation piece of hardware that is likely to be highly unstable

[07:37:01.0000] <MikeSmith>
and maybe even catch on fire

[07:38:00.0000] <myakura>
MikeSmith: They didn't update 13-inch models.

[07:39:00.0000] <MikeSmith>
hmm, ueah, I see that now

[07:39:01.0000] <MikeSmith>
so no i5 for 13-inch

[07:39:02.0000] <MikeSmith>
that sucks

[07:40:00.0000] <micheil>
MikeSmith: it's a pitty that the prices seem to have gone up a bit. (even on the educational store)

[07:40:01.0000] <MikeSmith>
I don't want to lug around a 15-inch machine around all my drinking spots in Kagurazaka

[07:41:00.0000] <micheil>
/me wouldn't go for anything smaller then 15"

[07:42:00.0000] <MikeSmith>
micheil: we are at the mercy of our benevolent hardware providers in Cupertino

[07:42:01.0000] <MikeSmith>
we should just be happy that they continue to allow us to buy their products at exorbitant prices 

[07:43:00.0000] <myakura>
annevk: I talked with guys organizing the HTML5 event in Fukuoka the other day and heard that they would like you to talk about the current status of CSS. Can you do that?

[07:43:01.0000] <micheil>
heh he

[07:43:02.0000] <annevk>
myakura, I saw that on wave, seems doable

[07:43:03.0000] <micheil>
MikeSmith: at any rate, you can't buy the i7 speced machine on the educational stores atm, it's a 404

[07:44:00.0000] <annevk>
myakura, I thought they did update the 13'' models, don't they have way more battery life?

[07:44:01.0000] <myakura>
annevk: cool.

[07:44:02.0000] <MikeSmith>
micheil: that's just a sign that you have been judged unworthy

[07:44:03.0000] <micheil>
:P

[07:45:00.0000] <annevk>
myakura, see http://www.apple.com/pr/library/2010/04/13mbp.html

[07:45:01.0000] <myakura>
annevk: whoa.

[07:46:00.0000] <Dashiva>
/me looks at 700 emails in whatwg and w3c folders

[07:49:00.0000] <zcorpan>
Hixie: i thought philipj and hsivonen argued for removing auto-localizing <time>

[07:50:00.0000] <hsivonen>
How Apple-like to say i5 and i7 without saying which ones in those families

[07:50:01.0000] <Dashiva>
"CSS pixels are by definition square (well, technically they're by definition circular)"

[07:51:00.0000] <hsivonen>
I'm glad I already bought an "i7" desktop while Apple wasn't delivering, so now I don't need to think which MacBook Pro model to get

[07:51:01.0000] <annevk>
zcorpan, Hixie, yeah, I thought the idea was to kill that and leave it to CSS

[07:52:00.0000] <Lachy>
It's a shame that the 17" MacBook Pro still has the useless express card slot instead of an SD card reader like the other models.

[07:52:01.0000] <annevk>
funny that the 17'' is not the most advanced

[07:52:02.0000] <annevk>
oh, you can get that via options

[07:54:00.0000] <hsivonen>
looks like I'd have to pay over 1000 euros more to get less CPU in the portable Mac case than I paid in the desktop non-Mac case

[07:54:01.0000] <Philip`>
/me prefers defining pixels as samples of a band-limited continuous function over a square grid

[07:54:02.0000] <hsivonen>
but to Apple's credit, "anti-glare" is back as an option

[07:56:00.0000] <Philip`>
Incidentally, hexagonal pixels (i.e. sample points in a hexagonal grid) are a better idea than square pixels, since (I think) they maximise the representable frequency for a given number of pixels

[07:57:00.0000] <Dashiva>
I've always wondered how three subpixels per pixel fit into a square grid

[07:58:00.0000] <Philip`>
That's just a square grid with 3x the horizontal resolution

[07:58:01.0000] <annevk>
hsivonen, but you no longer have a laptop?

[07:58:02.0000] <Philip`>
(where "square" obviously means "rectangular")

[07:59:00.0000] <Dashiva>
But the subpixels aren't rectangular in their light-emitting part, are they?

[07:59:01.0000] <hsivonen>
annevk: I still have my old MacBook, too.

[08:00:00.0000] <hsivonen>
annevk: but I need the portability a couple of times a year

[08:00:01.0000] <hsivonen>
annevk: and I need fast Firefox compiles every day

[08:00:02.0000] <annevk>
fair enough

[08:01:00.0000] <hsivonen>
the saddest part of my Ubuntu migration is that I need to keep the Mac running in order to make email work

[08:01:01.0000] <hsivonen>
I really, really need to get a better overall email solution

[08:02:00.0000] <hsivonen>
also, I haven't found a GraphicConverter replacement for Ubuntu

[08:03:00.0000] <Philip`>
Dashiva: The ones on my monitor are rectangular if I look through a magnifying glass

[08:04:00.0000] <Dashiva>
Oh really? Interesting

[08:04:01.0000] <Philip`>
What other shape would they be?

[08:04:02.0000] <annevk>
hsivonen, Opera? ;)

[08:04:03.0000] <Dashiva>
I figured they'd be rounded square-ish

[08:05:00.0000] <Philip`>
I don't think the shape should matter anyway - what's important is that there's a regular offset between the grids of (sub)pixel samples

[08:05:01.0000] <hsivonen>
annevk: what I want is roughly:

[08:05:02.0000] <hsivonen>
1) IMAP server

[08:05:03.0000] <hsivonen>
2) SMTP server that records the outgoing addresses in my email

[08:06:00.0000] <hsivonen>
3) server side filtering for the IMAP server so that the recorded previous outgoing addresses are whitelisted

[08:06:01.0000] <Philip`>
/me notes that his magnifying glass is one that came from a Christmas cracker many years ago, so he can't see precisely whether the coloured subpixels have sharp corners or not

[08:07:00.0000] <hsivonen>
4) an as Ajaxy/HTML5-y as it gets webmail client that uses the IMAP and SMTP servers but still works in Fennec, too

[08:07:01.0000] <hsivonen>
these requirements seem like no-brainers to me

[08:07:02.0000] <hsivonen>
but it seems I want a solution that isn't available

[08:14:00.0000] <annevk>
sounds nice

[08:21:00.0000] <annevk>
whoa, only eleven open bugs

[08:21:01.0000] <annevk>
way to go Hixie

[08:23:00.0000] <Philip`>
Why has http://lists.w3.org/Archives/Public/public-html-bugzilla/2010Apr/ had no updates since April 2?

[08:24:00.0000] <annevk>
no idea

[08:24:01.0000] <annevk>
MikeSmith, ^^

[08:28:00.0000] <MikeSmith>
Philip`: no idea

[08:29:00.0000] <MikeSmith>
bugzilla borkedness, I suppose

[08:29:01.0000] <MikeSmith>
I will check

[08:41:00.0000] <MikeSmith>
ah, they changed the sender name of the bugzilla user

[08:41:01.0000] <MikeSmith>
used to be bugzilla⊙wwo

[08:41:02.0000] <MikeSmith>
and now it's bugzilla⊙jwo

[08:42:00.0000] <MikeSmith>
why it changed I have no idea

[08:45:00.0000] <MikeSmith>
I think it should be fixed now

[08:56:00.0000] <TabAtkins>
Anyone know what plh's timezone is?

[09:00:00.0000] <Philip`>
TabAtkins: plh knows; you could ask him once he's awake

[09:01:00.0000] <TabAtkins>
I'll just tell him my timezone in this email and let him work out when the best time to call is, then.

[09:35:00.0000] <TabAtkins>
Babbage quote ftw!

[09:44:00.0000] <zcorpan>
TabAtkins: re unbalanced </iframe>s, that's not really the case

[09:44:01.0000] <zcorpan>
TabAtkins: since iframe is a rawtext element, it can't contain "<iframe></iframe>"

[09:48:00.0000] <gsnedders>
What's a sane terminal emulator for X11 with useful features like, uh, "find"?

[09:48:01.0000] <zcorpan>
TabAtkins: but having it as text content of the iframe would involve knowing the legacy situation with <!--

[09:49:00.0000] <Philip`>
gsnedders: Konsole?

[09:49:01.0000] <zcorpan>
wonder if we should add type="" to <pre> so that it can be exposed better to ATs

[09:53:00.0000] <gsnedders>
Philip`: That don't mean installing almost all of KDE?

[10:03:00.0000] <AryehGregor>
gsnedders, doesn't it only mean you need Qt or whatever?

[10:03:01.0000] <AryehGregor>
You don't need KDE's window manager or anything.

[10:03:02.0000] <AryehGregor>
(I hope)

[10:04:00.0000] <gsnedders>
kdebase4, kdelibs4…

[10:05:00.0000] <gsnedders>
Almost 100MB of stuff

[10:05:01.0000] <jgraham>
Anything from KDE needs all of KDE, always

[10:06:00.0000] <AryehGregor>
Well, if you're using Linux for a while, you'll probably have at least one KDE app installed.

[10:06:01.0000] <AryehGregor>
At least, I do.

[10:06:02.0000] <AryehGregor>
Random things like KCacheGrind.

[10:06:03.0000] <jgraham>
And a good goal in life is to have zero KDE stuff anywhere

[10:06:04.0000] <AryehGregor>
:P

[10:06:05.0000] <jgraham>
Because it is uniformly ugly

[10:06:06.0000] <TabAtkins>
That's an interseting definition of "good".

[10:06:07.0000] <AryehGregor>
It seems like GNOME Terminal really doesn't have "Find" anywhere that I can see.

[10:06:08.0000] <AryehGregor>
Interesting.

[10:07:00.0000] <jgraham>
No, it doesn't

[10:07:01.0000] <jgraham>
/me never noticed before

[10:07:02.0000] <AryehGregor>
jgraham, yeah, KDE apps look awful, at least if you're using GNOME.

[10:07:03.0000] <Philip`>
I don't understand why people want to avoid installing useful applications just because it'll use up a whole 0.01% of your disk space for dependencies

[10:07:04.0000] <AryehGregor>
Maybe they look better if the whole desktop is KDE.

[10:07:05.0000] <jgraham>
(but it would be useful I guess)

[10:07:06.0000] <gsnedders>
AryehGregor: Amazingly few terminal emulators do

[10:07:07.0000] <AryehGregor>
jgraham, you'd think so, but I don't think I've ever needed it in my whole time using terminals.

[10:07:08.0000] <gsnedders>
Sadly Konsole is the only good terminal emulator for X, AFAIK

[10:07:09.0000] <TabAtkins>
KDE apps do indeed look fine in KDE.  I used Kubuntu for a while.

[10:07:10.0000] <jgraham>
AryehGregor: My experience is that it looks equally bad if the whole desktop is KDE. Worse even because there is necessarily more of it

[10:08:00.0000] <AryehGregor>
Interesting.

[10:08:01.0000] <gsnedders>
/me restrains jgraham 

[10:08:02.0000] <gsnedders>
No being a GNOME fanboi.

[10:08:03.0000] <AryehGregor>
I prefer GNOME's minimalist philosophy, except of course when the default behavior is stupid *and* unchangeable.

[10:08:04.0000] <AryehGregor>
metacity is evil.

[10:08:05.0000] <AryehGregor>
When you have multiple monitors, it positions new windows on some random monitor of its choosing, not the current monitor.

[10:08:06.0000] <gsnedders>
But it's open source, what does it matter!!!11!!!!!!1!!eleventy!!

[10:08:07.0000] <jgraham>
(the problems I have are things like the fact that every app has about a million buttons)

[10:09:00.0000] <jgraham>
(to cover every usecase that every user ever had)

[10:09:01.0000] <AryehGregor>
Which gives you no control over where it goes when it's, e.g., fullscreen, or otherwise can't be moved between windows.  Unless you want to break out wmctrl on the command line.

[10:09:02.0000] <AryehGregor>
jgraham, wait, you think this is bad but you work for Opera?

[10:09:03.0000] <jgraham>
(so it is impossible to actually do any of the things)

[10:09:04.0000] <jgraham>
AryehGregor: No comment :p

[10:10:00.0000] <gsnedders>
10.50 doesn't have many buttons by default!

[10:10:01.0000] <jgraham>
AryehGregor: Anyway everyone uses compiz these days, no?

[10:10:02.0000] <AryehGregor>
Compiz doesn't work with nv, does it?  So I don't.

[10:10:03.0000] <jgraham>
and compiz has a shed load of options if you install the control panel thing

[10:10:04.0000] <jgraham>
Far too many

[10:10:05.0000] <AryehGregor>
If it works with nouveau, then I guess I will in a couple of weeks when I go to 10.04.

[10:10:06.0000] <TabAtkins>
I spent an entire day playing with the options before shutting off almost all of them.

[10:11:00.0000] <AryehGregor>
Yeah, I remember that.  I used to use it.

[10:11:01.0000] <jgraham>
But you need it to get the grid plugin working

[10:11:02.0000] <AryehGregor>
But Compiz interferes with Wine, and Wine is the only reason I'd bother installing a proprietary video driver.

[10:11:03.0000] <TabAtkins>
/me knows how to productively spend his time.

[10:11:04.0000] <jgraham>
And everything that is good about the grid plugin is (almost) everything that is bad about window mangement on OSX

[10:11:05.0000] <jgraham>
Actually that may be unfair

[10:12:00.0000] <jgraham>
There is a lot more wrong with window mangement on OSX than just the inability to place winwos nicely

[10:12:01.0000] <jgraham>
But that is really the main thing

[10:14:00.0000] <AryehGregor>
Window management on Windows always seemed to Just Work for me.

[10:15:00.0000] <TabAtkins>
It was never great, but it worked.  Not nearly enough support for things nicely snapping to edges of other things.

[10:15:01.0000] <AryehGregor>
I finally got it to work okay on Linux, but the dual-monitor window placement of metacity is really annoying.  I went to Xfce for a while and it worked much better.

[10:15:02.0000] <AryehGregor>
Yeah, that's a nice feature in Linux.

[10:15:03.0000] <AryehGregor>
Xfce panels are also way better than GNOME ones.

[10:15:04.0000] <AryehGregor>
Like, you can actually place them on non-primary monitors.

[10:15:05.0000] <AryehGregor>
And the icons don't randomly jump around every time you restart.

[10:15:06.0000] <AryehGregor>
/me tries to remember why he switched back to GNOME

[10:18:00.0000] <jgraham>
gsnedders: A gnome-terminal developer claims to be working on a patch for find btw

[10:18:01.0000] <gsnedders>
jgraham: Dude, it's already 2010. I don't care.

[10:18:02.0000] <gsnedders>
If you can't have find in 2010...

[10:19:00.0000] <jgraham>
You might get it by the end of 2010

[10:21:00.0000] <jgraham>
You could install xsel, do select all and on a second terminal do xsel | less

[10:22:00.0000] <jgraham>
Although it crahsed when I tried that...

[10:25:00.0000] <micheil>
anyone know if the body in the WebSocket initialization should be in utf8 or ascii encoding?

[10:26:00.0000] <micheil>
the spec doesn't specify: "After the first 0x0D 0x0A 0x0D 0x0A byte sequence, indicating the end of the fields, the client sends eight random bytes. These are used in constructing the server handshake."

[10:27:00.0000] <TabAtkins>
You mean the "8 random bytes" part?

[10:27:01.0000] <micheil>
yeah

[10:27:02.0000] <TabAtkins>
Bytes dont' have an encoding.  They're bytes.  8 bits.

[10:27:03.0000] <micheil>
how should the server interpret them?

[10:28:00.0000] <TabAtkins>
As bytes.

[10:28:01.0000] <micheil>
uhh..

[10:28:02.0000] <TabAtkins>
They're not characters, and shouldn't be interpreted as such.

[10:28:03.0000] <micheil>
righteo..

[10:28:04.0000] <jgraham>
micheil: They are 8 random numbers in the range 0 to 255

[10:29:00.0000] <jgraham>
s/numbers/integers/

[10:29:01.0000] <micheil>
jgraham: okay. I just have a few ways of implementing things

[10:29:02.0000] <micheil>
jgraham: I ended up removing the http parser I wrote and using the one from node's core

[11:00:00.0000] <Dashiva>
Good reason not to let code examples use smart quotes: client.open(’GET’, ‘demo.cgi’);

[11:02:00.0000] <TabAtkins>
Nothing should ever use smart quotes.  >_<

[11:02:01.0000] <micheil>
TabAtkins: apart from typographers.

[11:02:02.0000] <micheil>
and writers.

[11:03:00.0000] <micheil>
programmers, probably not.

[11:03:01.0000] <TabAtkins>
Let me clarify.  Nothing should ever use smartquotes in content.

[11:03:02.0000] <TabAtkins>
You can intelligently transform quotes into smartquotes at display time.

[11:07:00.0000] <Dashiva>
TabAtkins: Can you really, if there's code involved?

[11:07:01.0000] <TabAtkins>
Clearly, anything in <code> shouldn't have that applied.

[12:25:00.0000] <AryehGregor>
/me speculates about the rumored upcoming VP8 announcement

[12:26:00.0000] <TabAtkins>
Though I wouldn't leak anything if I *did* know anything, I still wish I knew something about it too.

[12:32:00.0000] <AryehGregor>
The question is, will it be like "Here's a cool demo, it might be practically more useful than some other codecs in some cases in the medium term maybe"?  Or more like "We just released VP8 support in the Chrome dev channel, and YouTube is migrating to it over the next few years"?

[12:32:01.0000] <AryehGregor>
It can't be too awesome, or they wouldn't have seen fit to fund Theora on ARM.

[12:32:02.0000] <AryehGregor>
But it can't be too useless either, or they wouldn't have bought it.

[12:32:03.0000] <TabAtkins>
That's not necessarily true.  We're a big org.  Sometimes the left and right arm are both doing awesome things.

[12:32:04.0000] <AryehGregor>
Like Android vs. Chrome OS.

[12:32:05.0000] <AryehGregor>
Yeah, true.

[12:32:06.0000] <TabAtkins>
Yes.

[12:33:00.0000] <AryehGregor>
Some of the Theora people I know say that it's really an issue of having a mature, high-quality decoder, more than the format itself, that H.264 is better than Theora only because it has a more mature encoder, and that therefore VP8 isn't going to be a huge deal, but Theora can be made better with some more resources.

[12:33:01.0000] <AryehGregor>
At least, I saw that once.

[12:33:02.0000] <TabAtkins>
I've heard chatter to that effect too.

[12:34:00.0000] <TabAtkins>
Apparently Moz funding Theora dev made a big difference recently.

[12:36:00.0000] <AryehGregor>
But maybe VP8 uses some brilliant new technique that leapfrogs the existing stuff?  I don't know why Google would buy it if it didn't think it could be better than Theora in the fairly near term.

[12:36:01.0000] <TabAtkins>
Yeah, honestly have no idea.

[12:36:02.0000] <TabAtkins>
Our <video> people aren't in Mountain View, anyway.

[12:36:03.0000] <othermaciej>
I have heard that Theora has some technical limitations which make it hard to be efficient at large (HD) video sizes

[12:37:00.0000] <othermaciej>
like motion vectors are limited to 8 pixels or something like that

[12:37:01.0000] <othermaciej>
which is fine at 320x200 but not so much at 1024x768

[12:37:02.0000] <othermaciej>
note: not a codec expert

[12:37:03.0000] <AryehGregor>
Interesting.

[12:38:00.0000] <AryehGregor>
I've also heard that some codecs have very particular advantages and disadvantages.

[12:38:01.0000] <AryehGregor>
Like Theora can evidently be decoded a lot faster in software than H.264 can.

[12:38:02.0000] <AryehGregor>
And so on.

[12:38:03.0000] <AryehGregor>
Oh well, we'll see.

[12:38:04.0000] <AryehGregor>
Speculation is only fun up to a point.

[12:38:05.0000] <othermaciej>
I believe that is true, about sw decoding complexity

[13:07:00.0000] <micheil>
Hixie: are you about?

[13:11:00.0000] <jgraham>
It kinda makes sense that you would fund theora on mobile devices and something new for non-mobile if all of the above were true

[13:11:01.0000] <jgraham>
Since mobile devices typically don't need to do HD

[13:12:00.0000] <jgraham>
micheil: (I suggest you just ask and Hixie will reply when he is around)

[13:13:00.0000] <micheil>
okay, well, with websockets, and sending the body on the request without a Content-Length or Transfer-Encoding seems to go against RFC2068

[13:13:01.0000] <micheil>
"The presence of a message-body in a request is signaled by the inclusion of a Content-Length or Transfer-Encoding header field in the request's message-headers"

[13:14:00.0000] <jgraham>
So the hypothetical architecture is one where the HTTP server handles the entire request including the random bytes and then passes it off to a websockets handler

[13:15:00.0000] <micheil>
yeah, sorta

[13:15:01.0000] <jgraham>
Couldn't you just treat it like a request with no body and then have the WS server read 8 bytes from the network to get the random bytes?

[13:15:02.0000] <jgraham>
I mean at some point you have to hand over to WS-specific code

[13:16:00.0000] <micheil>
no. because the body is separated from the headers with \r\n\r\n

[13:21:00.0000] <jgraham>
So you only have a problem if the HTTP layer reads the headers, notices it is a websocket client, hands over to the websocket server, but also reads the body (even though the headers don't say there is a body) and doesn't make that avaliable to the WS server

[13:22:00.0000] <jgraham>
It's not clear to me why that would happen

[13:22:01.0000] <jgraham>
(it seems you could get into trouble with either design)

[13:24:00.0000] <micheil>
hmm..

[13:24:01.0000] <micheil>
well, I've found a really hacked fix, not sure if it'll work though

[13:32:00.0000] <micheil>
I'm beginning to think it was easier to write my own header handler

[13:32:01.0000] <jgraham>
If you have the choice I expect it is

[13:33:00.0000] <jgraham>
I *think* parsing the initial headers is not that hard

[13:36:00.0000] <micheil>
I just hope every browser / client that implements websockets get's it write.

[13:36:01.0000] <micheil>
*right

[13:36:02.0000] <micheil>
jgraham: do you think it'd be possible for the body to be separated from the request at any stage?

[13:37:00.0000] <micheil>
ie, rather then it being: "GET ... \r\n .... \r\n\r\nBODY\r\n\r\n"

[13:38:00.0000] <micheil>
it gets received as "GET ... \r\n ... \r\n\r\n" and then later "BODY\rn\r\n"

[13:45:00.0000] <franksalim>
micheil, that is entirely possible

[13:45:01.0000] <micheil>
damn

[13:45:02.0000] <micheil>
not the answer I was hoping for

[13:46:00.0000] <KaOSoFt>
Good morning.

[13:47:00.0000] <micheil>
In which case I have no idea how I'm going to handle the http headers / meta for websocket connections..

[13:49:00.0000] <franksalim>
micheil,  i don't understand why you can't hand the stream over to the websocket server

[13:49:01.0000] <micheil>
uh. I'm writing the server

[13:50:00.0000] <micheil>
so, how I go about parsing the headers from the clients is fairly important

[13:50:01.0000] <franksalim>
i mean your server

[13:50:02.0000] <franksalim>
to clarify: hand the stream from the http code that parsed the request headers to the websocket code

[13:50:03.0000] <micheil>
oh, wait, my WebSocket server is reading the stream, it's a matter of parsing the headers

[13:51:00.0000] <micheil>
umm.. because, I'm implementing my server on top of network sockets, not a http server

[13:51:01.0000] <franksalim>
so you are parsing everything from byte 1 in your own websocket code

[13:51:02.0000] <micheil>
yes

[13:52:00.0000] <franksalim>
does your server understand websocket and http or just websocket

[13:52:01.0000] <Hixie>
beowulf: HixieThePixie http://bfbc2.statsverse.com/stats/ps3/HixieThePixie

[13:52:02.0000] <micheil>
it understands the websocket, however, needs to handle the http request sent for the websocket client initialisation

[13:53:00.0000] <franksalim>
and you are using an http library for that or writing your own header parsing code?

[13:53:01.0000] <Hixie>
micheil: anything i can help with re websockets? (i haven't read the backlog)

[13:53:02.0000] <micheil>
it'd be much better if the websocket spec said that the Content-Length field needed to be set

[13:53:03.0000] <micheil>
Hixie: that ^^

[13:54:00.0000] <Hixie>
Content-Length: 0?

[13:54:01.0000] <Hixie>
i thought that was implied by the GET

[13:54:02.0000] <micheil>
it'd have to be Content-Length: 8

[13:54:03.0000] <micheil>
for the 8 bits

[13:54:04.0000] <Hixie>
the 8 bytes aren't part of the GET request

[13:54:05.0000] <micheil>
and no, it's not implied by the GET

[13:54:06.0000] <Hixie>
they're the first 8 bytes after the upgrade

[13:55:00.0000] <micheil>
0x0D 0x0A == \r\n, yeah?

[13:55:01.0000] <Hixie>
yep

[13:55:02.0000] <Hixie>
(unless you're using perl on windows)

[13:55:03.0000] <Hixie>
(or have changed the \n magic variable)

[13:55:04.0000] <Dashiva>
... you're kidding

[13:55:05.0000] <Hixie>
(iirc $/)

[13:55:06.0000] <micheil>
in which case, the key3 value would be the body of the HTTP request

[13:56:00.0000] <Hixie>
GETs don't have bodies if they don't have a Content-Length

[13:56:01.0000] <Hixie>
just handle it in the same code you're going to use to handle the websocket frames

[13:56:02.0000] <micheil>
um.. although, the request will probably look like..

[13:58:00.0000] <micheil>
http://gist.github.com/365024

[13:59:00.0000] <micheil>
Hixie: or am I wrong in thinking that?

[14:00:00.0000] <Hixie>
there's no \r\n after the 8 bytes

[14:00:01.0000] <micheil>
okay, well, any rate, leave off one and it would work?

[14:00:02.0000] <Hixie>
the first byte after the 8th byte will be a 0x00 byte from the first frame

[14:02:00.0000] <micheil>
okay

[14:02:01.0000] <micheil>
so, removing the two \r\n\r\n at the end of that data, and it's right?

[14:02:02.0000] <micheil>
eg, ...le.com\r\n\r\n^n:ds[4U

[14:04:00.0000] <micheil>
hmm..

[14:04:01.0000] <Hixie>
looks right

[14:04:02.0000] <Hixie>
but i haven't checked the details

[14:04:03.0000] <Hixie>
there's some examples in the spec if that helps

[14:05:00.0000] <micheil>
yeah, I'm constantly cross-referencing stuff

[14:05:01.0000] <Hixie>
cool

[14:06:00.0000] <micheil>
I'm wanting to make sure this client is spec-compliant

[14:06:01.0000] <micheil>
I'm still not sure if I want it to drop connections not using Sec-* (ie, all current webbrowser clients)

[14:11:00.0000] <micheil>
Hixie: any opinions on that?

[14:14:00.0000] <micheil>
Hixie: at any rate, thanks for the clarification and help :)

[14:14:01.0000] <Hixie>
well right now the spec isn't done

[14:14:02.0000] <Hixie>
so anything you do is going to need to change again before long

[14:14:03.0000] <micheil>
true

[14:15:00.0000] <Hixie>
whether you temporarily support the other work-in-progress implementations is up to you

[14:15:01.0000] <micheil>
hmm.. I could add a useStrict  flag in the options for the server as to whether to reject or allow clients not sending Sec-*

[14:16:00.0000] <micheil>
hmm.. to bed: it's 6:22am..

[14:18:00.0000] <franksalim>
micheil, goodnight

[15:40:00.0000] <TabAtkins>
echo?

[15:46:00.0000] <Dashiva>
Is Shelley still going on about SQL injection as a problem with @srcdoc?!

[15:46:01.0000] <TabAtkins>
Yes.  Yes she is.

[15:48:00.0000] <Dashiva>
I'm having trouble maintaining a belief in good faith intentions

[15:49:00.0000] <TabAtkins>
Not quite sure how an author can simultaneously be absolutely brain-dead stupid enough to think that using @srcdoc somehow protects against SQL injection from form-submitted data, but simultaneously be intelligent enough not to eat his own keyboard.

[15:49:01.0000] <TabAtkins>
Yeah, no, good-faith has long been abandoned in my mind.

[16:02:00.0000] <Philip`>
In a process where there's no penalty for making bad arguments, and you believe a proposal should be accepted, it seems only rational to provide as many arguments as you can think of regardless of their quality because maybe somebody will accidentally believe some of them

[16:03:00.0000] <Hixie>
yeah, that's pretty much the problem here

[16:05:00.0000] <Dashiva>
 <Philip`> In a process where there's no penalty for making bad arguments, and you believe a proposal should be accepted, it seems only rational to provide as many arguments as you can think of regardless of their quality because maybe somebody will accidentally believe some of them

[16:05:01.0000] <Dashiva>
For you, othermaciej :)

[16:05:02.0000] <TabAtkins>
Which unfortunately creates more work for everyone, as there's more crap to wade through, and then you still have to debunk the crap arguments.

[16:05:03.0000] <othermaciej>
Dashiva: pardon?

[16:05:04.0000] <Philip`>
(Of course a process that penalises good arguments because of nearby bad arguments would be irrational, because things should be decided based on the merits of good arguments)

[16:05:05.0000] <Philip`>
(so there's no good solution)

[16:08:00.0000] <Dashiva>
othermaciej: Talking about how there's no reason to not make bogus arguments to bog down "the others"

[16:09:00.0000] <othermaciej>
the only penalty for bad arguments is that readers may pay less heed to your good arguments

[16:10:00.0000] <Dashiva>
They don't have a choice with change proposals involved

[16:10:01.0000] <TabAtkins>
othermaciej: That's not a real penalty, as your bad arguments will still generate work for others.

[16:10:02.0000] <Dashiva>
You must address everything to be sure bad stuff doesn't happen

[16:11:00.0000] <othermaciej>
TabAtkins: the chairs have rejected Change Proposals before for lacking rationale

[16:11:01.0000] <othermaciej>
TabAtkins: but we're reluctant to do a really deep analysis

[16:12:00.0000] <Dashiva>
If you create a proposal with one real argument and ten bogus arguments, that's a lot of extra work for counter-proposals

[16:12:01.0000] <othermaciej>
we'll only reject if the rationale is completely missing, or just transparently inadequate (e.g. tautological or does not actually cover all changes in the proposal)

[16:12:02.0000] <TabAtkins>
othermaciej: That's not stopping bad arguments from dominating a change proposal.  A single good argument in rationale is enough to carry a bunch of bad arguments, all of which have to be addressed.  ;_;

[16:12:03.0000] <othermaciej>
TabAtkins: well, a counter-proposal doesn't necessarly have to address every single argument from any Change Proposal it is opposed to

[16:13:00.0000] <TabAtkins>
And since the CP accept/reject mechanism is completely opaque, we have no idea how important it is to address any particular argument.

[16:13:01.0000] <othermaciej>
it just has to make the case that the status quo is better than the change

[16:13:02.0000] <othermaciej>
the chairs would likely mention if it seems like there are points worth addressing that haven't been

[16:14:00.0000] <othermaciej>
anyway, I get a sense that you are somewhat dissatisfied, but I'm not sure if you have a specific proposal for what the chairs might do differently

[16:14:01.0000] <TabAtkins>
Not right now, no.

[16:14:02.0000] <Dashiva>
Perhaps allow striking of obviously invalid arguments

[16:16:00.0000] <TabAtkins>
Something along that line might work.  The essential problem is that, since the accept/reject process is a single opaque step at the end, there's no way to know *where* to apply energy.

[16:17:00.0000] <TabAtkins>
Feedback as things move along would at least allow CP authors to know which arguments to strengthen, and counter authors to know which ones don't need to be addressed.

[16:18:00.0000] <TabAtkins>
Not sure precisely how to address or obtain "more feedback", but I feel it's the basic weakness in the Issue process.

[16:18:01.0000] <murz>
You can use HTML5 elements on a page that uses an HTML4 or XHTML1.1 declaration, right? All that really matters it that the browser supports the element, the DTD is mostly just to trigger standards mode and to tell validators which rules to use?

[16:19:00.0000] <TabAtkins>
There's no practical difference, no.

[16:19:01.0000] <TabAtkins>
DTDs are *entirely* to trigger standards mode.

[16:20:00.0000] <paul_irish>
that's a great question. people also are commonly curious if javascript or RTL things change with different doctypes.

[16:23:00.0000] <TabAtkins>
level all

[16:27:00.0000] <annevk>
actually, DOCTYPEs are, DTDs are irrelevant

[16:47:00.0000] <MikeSmith>
TabAtkins: is "level all" a secret masonic code phrase or something?

[16:47:01.0000] <Dashiva>
It's the password to the secret whatwg treehouse

[16:47:02.0000] <TabAtkins>
Damn, was hoping that wasn't actually posted to everything.

[16:48:00.0000] <MikeSmith>
in full effect

[16:48:01.0000] <TabAtkins>
I've been trying to configure irssi to not show joins and quits.

[16:48:02.0000] <MikeSmith>
ah

[16:48:03.0000] <othermaciej>
TabAtkins: the chairs give feedback on form, the WG participants themselves hopefully give technical feedback on the actual validity / convincingness of the arguments

[16:49:00.0000] <TabAtkins>
othermaciej: That doesn't address my concern, which is that the chairs make the final decision on validity/convincingness of arguments, but we have no way of knowing how valid/convincing a given argument is going to be, and thus we must devote energy to *every* argument, even the stupid ones.

[16:50:00.0000] <othermaciej>
TabAtkins: on the Microdata issue, I believe the Chairs gave specific feedback before we went to a poll on which arguments did not seem sufficiently addressed by the other side

[16:50:01.0000] <Hixie>
and tab didn't convince you, so clearly he didn't have enough information on how to convince you :-)

[16:51:00.0000] <othermaciej>
that's like saying if you lose a case in court, you must not have known enough about the law

[16:51:01.0000] <TabAtkins>
I'm not talking about the Microdata issue.  At the moment, I'm talking about Issue 100, where there are a number of completely irrelevant or even nonsensical arguments, that I nonetheless had to address.

[16:51:02.0000] <TabAtkins>
I actually wrote more for several points before deciding to just be terse and say they were irrelevant.

[16:51:03.0000] <othermaciej>
your first audience for your Change Proposal should be the Working Group

[16:51:04.0000] <othermaciej>
so write whatever you think is useful to persuade WG members in general

[16:52:00.0000] <TabAtkins>
I will.  But my final audience is the chairs.

[16:52:01.0000] <othermaciej>
we want polls / written decisions to be the last resort

[16:52:02.0000] <othermaciej>
though I don't think we'll be so fortunate as to have 0 more of them

[16:52:03.0000] <Dashiva>
So what about all the "kill element x" proposals, that got massive pushback from the wg?

[16:53:00.0000] <othermaciej>
clearly the author of those did not do a good job of persuading the WG, at least so far

[16:53:01.0000] <Dashiva>
They aren't very likely to end in amicable resolution

[16:53:02.0000] <Dashiva>
Doesn't that ensure a chair decision or poll?

[16:53:03.0000] <TabAtkins>
Where "very likely" is defined as "0".  ^_^

[16:53:04.0000] <TabAtkins>
s/very likely/not very likely/

[16:53:05.0000] <othermaciej>
depends on whether the one supporter of those proposals agrees to back down after seeing the counter

[16:53:06.0000] <othermaciej>
if not, we'll decide what to do next


2010-04-14
[17:02:00.0000] <Hixie>
what do people think of making <s> valid again, with a semantic something like "indicates that the contents are no longer accurate or no longer relevant" or some such?

[17:04:00.0000] <Dashiva>
To avoid diluting the semantics of <del>?

[17:04:01.0000] <Hixie>
yeah

[17:04:02.0000] <Hixie>
(see http://www.w3.org/Bugs/Public/show_bug.cgi?id=9429 if you have any especially strong opinions one way or the other)

[17:06:00.0000] <annevk>
not opposed

[17:06:01.0000] <daedb>
Sounds fine to me, but I don't care strongly about it :p

[17:08:00.0000] <Dashiva>
I'm not opposed, but it sounds like a semantic fig leaf

[17:16:00.0000] <othermaciej>
Hixie: <s> is strikethrough? 

[17:16:01.0000] <othermaciej>
what's the difference between <s> and <strike> if any?

[17:16:02.0000] <othermaciej>
(change sounds good to me btw)

[17:17:00.0000] <TabAtkins>
othermaciej: No difference.  They appear to be the same element?

[17:17:01.0000] <othermaciej>
I wonder why both exist

[17:18:00.0000] <othermaciej>
HTML4 had both

[17:18:01.0000] <TabAtkins>
Presumably because HTML doesn't make any sense.

[17:21:00.0000] <annevk>
I guess Netscape had <s> and Microsoft had <strike> or vice versa

[17:21:01.0000] <TabAtkins>
Dashiva: Why do you think so?  There is a consistent semantic implied by the use of strike-through in text.

[17:22:00.0000] <TabAtkins>
It's sort of like <b> taking on the role of "everything that is bold for some important reason, but isn't otherwise covered by another element".

[17:22:01.0000] <othermaciej>
the <b> definition is also a semantic fig leaf

[17:22:02.0000] <Dashiva>
Well, what are the major semantic use cases for <s> that aren't covered by <del>?

[17:23:00.0000] <othermaciej>
<del> implies an edit or edit instruction

[17:23:01.0000] <othermaciej>
strikethrough in blogs is typically for uses of either "I made a mistake but am being transparent about correcting it" or an ironic fake version of that

[17:24:00.0000] <othermaciej>
it seems a little different to me than marking up document edits with <ins>/<del>

[17:24:01.0000] <othermaciej>
like the point of using strikethrough in that context is to *not* delete the stricken text, but rather let it stand and signal that it is no longer operative

[17:25:00.0000] <TabAtkins>
Actually, I think the first use-case *is* covered well by <del>.  They're editting the document.  If they were feeling really fancy, they could cite the comment from the pedant that pointed it out.

[17:25:01.0000] <TabAtkins>
But the ironic fake version probably is abusing <del>.

[17:25:02.0000] <Dashiva>
So if I'm making ironic emphasis in a quote, I shouldn't use <strong> or <em> either?

[17:25:03.0000] <othermaciej>
But they are not saying that text has been deleted or should be - rather they are flagging it as false

[17:25:04.0000] <othermaciej>
or mistaken or unintended

[17:26:00.0000] <othermaciej>
but explicitly letting it stand for the record

[17:27:00.0000] <TabAtkins>
Basically, the ironically-stricken text shouldn't be interpreted as ordinary text (that is, it shouldn't be just a <span style>), but it shouldn't be skipped over as a deletion either (and so shouldn't be a <del>).

[17:27:01.0000] <Dashiva>
(Like Tab says, transparent correction is very much ins/del, otherwise you'd just do the edit directly)

[17:28:00.0000] <othermaciej>
maybe I do not grok the deep semantics of <del>

[17:28:01.0000] <othermaciej>
I am not entirely sure what it is supposed to be for

[17:28:02.0000] <othermaciej>
I imagine if I marked up someone else's document to propose changes, I would use <ins> and <del>

[17:28:03.0000] <TabAtkins>
Yeah, that's right.

[17:28:04.0000] <othermaciej>
to me that seems different from transparent correction

[17:28:05.0000] <othermaciej>
where I am consciously *not* editing the text

[17:28:06.0000] <othermaciej>
and using strikethrough as an alternative

[17:29:00.0000] <Dashiva>
What is a correction if not editing the text?

[17:29:01.0000] <othermaciej>
well, newspaper corrections are generally an external annotation

[17:29:02.0000] <TabAtkins>
But you are!  In most cases I've seen strikethrough used as a transparent correction, the text could be read while completely ignoring the stricken text, and be correct.

[17:29:03.0000] <othermaciej>
the Web lets you do more of it inline

[17:30:00.0000] <othermaciej>
TabAtkins: I often see a strikethrough followed by a parenthetical remark explaining the correction, which would not read right if you skipped it entirely

[17:30:01.0000] <Dashiva>
The only problem with the ironic deletion seems to be that it shouldn't be collapsed

[17:30:02.0000] <othermaciej>
anyway, I am not the right person for this discussion

[17:30:03.0000] <othermaciej>
because I don't really care if elements are semantic or presentational, as long as they are useful

[17:30:04.0000] <Dashiva>
<del permanent-marker>

[17:31:00.0000] <TabAtkins>
Is that somehow better than <s>?

[17:32:00.0000] <Dashiva>
Personally I'd be fine with <del> meaning "something removed from the text" without also implying "something that can be removed from the markup"

[17:32:01.0000] <TabAtkins>
That might be the most pragmatic solution.

[17:33:00.0000] <TabAtkins>
Hixie, presumably you had a reason why you think that that wouldn't be sufficient?

[17:33:01.0000] <Dashiva>
If we're adding <s> as ironic <del>, we must also add <u> as ironic <ins>

[17:34:00.0000] <TabAtkins>
Haha.

[17:34:01.0000] <othermaciej>
underline seems at best very loosely tied to the semantic if "inserted text"

[17:34:02.0000] <othermaciej>
not the way strikethrough implies some form of deletion or removal

[17:35:00.0000] <othermaciej>
also people rarely feel the need to mark up inline correction additions in any way

[17:35:01.0000] <Dashiva>
Okay, then, <u> as ironic <a>

[17:36:00.0000] <othermaciej>
I think Hixie's proposed definition of <s> for strikethrough of falsity/irrelevance/inoperativeness captures the distinction between transparent corrections and <del> as text that is/should be deleted

[17:36:01.0000] <Hixie>
I think "indicates that the contents are no longer accurate or no longer relevant" is quite different than <del>'s edit semantic, personally

[17:36:02.0000] <othermaciej>
I think <a> works fine for ironic <a> :-)

[17:36:03.0000] <TabAtkins>
We need use-cases for ironic <a>s, damnit!

[17:36:04.0000] <Dashiva>
But it's not supposed to be a real link, just look like one

[17:36:05.0000] <othermaciej>
just put the </sarcasm> close tag at the end of your link

[17:36:06.0000] <Hixie>
anyway i didn't mean to start a huge conversation, there are far bigger problems :-P

[17:36:07.0000] <othermaciej>
ironic != fake

[17:36:08.0000] <TabAtkins>
Hixie, really?  What is the edit semantic if *not* "indicating that the contents are no longer accurate"?

[17:36:09.0000] <Dashiva>
<a></sarcasm>

[17:37:00.0000] <othermaciej>
TabAtkins: there are lots of reasons to delete content besides "no longer accurate"

[17:37:01.0000] <othermaciej>
most edits are for style, grammar or conciseness

[17:37:02.0000] <othermaciej>
and there are reasons to indicate content is "no longer accurate" without deleting it

[17:37:03.0000] <Dashiva>
I guess my personal interpretation of edit is any change to the document, not just an editioral review

[17:38:00.0000] <Hixie>
TabAtkins: consider a Fox News script... <del> there could be used for quite the opposite of "no longer accurate"

[17:38:01.0000] <TabAtkins>
No getting into arguments about absolute truth here.

[17:40:00.0000] <Dashiva>
Well, it won't be me having to defend it, so go ahead :)

[17:41:00.0000] <TabAtkins>
JonathanNeal: Your Print Protector showed up in Smashing Magazine's Buzz stream.  ^_^

[17:41:01.0000] <JonathanNeal>
:)

[17:43:00.0000] <JonathanNeal>
Remy Sharp is helping me shrink the code down even more.

[18:00:00.0000] <Hixie>
sweet jesus christ i am so tired of dealing with leif's massive essays

[18:00:01.0000] <Hixie>
every bug he files or reopens has pages and pages of text

[18:00:02.0000] <Hixie>
most of which seems to make no sense at all

[18:09:00.0000] <Hixie>
can someone please translate http://www.w3.org/Bugs/Public/show_bug.cgi?id=9417 for me?

[18:11:00.0000] <TabAtkins>
I like that the response to "Could you succinctly describe what the problem is?" is 4 paragraphs of text.

[18:14:00.0000] <TabAtkins>
Nope, sorry, can't parse it either.  I'd have to test what he's talking about in order to tell what the problem might be.  He hints at things that might be problems, but doesn't actually state precisely *what* the problem is.

[18:18:00.0000] <Dashiva>
I just want content-language to burn on the pyre of obsoletion

[18:23:00.0000] <Dashiva>
Hixie: NEEDSINFO, succinct description still missing

[18:25:00.0000] <Hixie>
Dashiva: he added even more text actually, i ended up just rejecting it based on the new text (which claimed a problem existed which, as far as i can tell, only exists if you go out of your way to make your life hard in the first place)

[19:59:00.0000] <jlebar>
Hixie, Can I find the old [MIMESNIFF] info somewhere?  I'm trying to figure out how to specify sniffing for resource packages (http://limi.net/articles/resource-packages/), but all the HTML5 sniffing language I might look at appears to be out of order.

[20:00:00.0000] <Hixie>
old MIMESNIFF?

[20:00:01.0000] <Hixie>
as opposed to the new MIMESNIFF?

[20:00:02.0000] <jlebar>
"The above is out of date now that the relevant section has been removed from MIMESNIFF. Stay tuned; I'll bring it back here soon."

[20:01:00.0000] <Hixie>
oh, that stuff

[20:01:01.0000] <Hixie>
yeah let me just add it back now, hold on

[20:01:02.0000] <jlebar>
Thanks.  :)

[20:13:00.0000] <Hixie>
jlebar: reload

[20:13:01.0000] <Hixie>
i'll check it in after lunch when i've reviewed it

[20:13:02.0000] <Hixie>
afk

[20:20:00.0000] <jlebar>
Thanks, Hixie.

[20:20:01.0000] <MikeSmithX>
HTML WG bugmail still does not seem to be getting posted to public-html-bugzilla⊙wo at all

[20:20:02.0000] <MikeSmithX>
..and I'm not sure why

[20:21:00.0000] <MikeSmith>
I think I need to kick something

[20:23:00.0000] <MikeSmith>
/me tries one more thing

[23:44:00.0000] <JonathanNeal>
hi all

[23:47:00.0000] <boblet>
hey JonathanNeal 

[23:47:01.0000] <JonathanNeal>
hey boblet

[23:55:00.0000] <boblet>
can any CSS3 people tell me what’s happening to writing-mode? Seems like it was in the text module but is not now. I’m wondering if that’s temporary or if direction & block-flow are obsoleting it

[00:10:00.0000] <JonathanNeal>
I suppose you can't link to local files using audio?

[00:29:00.0000] <boblet>
JonathanNeal: doesn’t file:///  or something relative to source html file work? would expect it to huh

[00:30:00.0000] <JonathanNeal>
no it doesn't work

[00:55:00.0000] <hsivonen>
Hixie: I support making <s> conforming

[00:56:00.0000] <zcorpan>
/me expects masses of people outside #whatwg will object to making <s> conforming

[00:58:00.0000] <hsivonen>
as I see things, <s> means striken but not removable from view and <del> means diff-annotation of an actual delete that could be safely completely deleted

[00:59:00.0000] <MikeSmith>
hsivonen: what about <strike> ?

[01:00:00.0000] <hsivonen>
MikeSmith: I'd be OK with having both <strike> and <s> and saying that they are exact synonyms

[01:00:01.0000] <hsivonen>
in which case it would be logical to make <acronym> conforming, too

[01:00:02.0000] <MikeSmith>
hmm, yeah

[01:00:03.0000] <MikeSmith>
so a bit of slipperly slope there maybe

[01:01:00.0000] <MikeSmith>
though I short slope, I guess

[01:01:01.0000] <hsivonen>
but if <acronym> were made conforming, it would need very, very obvious spec text saying that <acronym> and <abbr> are normatively synonymous and permathreads are hereby pre-empted

[01:02:00.0000] <MikeSmith>
/me looks back at http://dev.opera.com/articles/view/mama-phrase-block-list/#phrase

[01:02:01.0000] <MikeSmith>
we still have the unfortunate fact to deal with that <u> is much more commonly used than even <abbr> or <code>

[01:03:00.0000] <hsivonen>
I'd be OK with making <u> conforming, too

[01:03:01.0000] <zcorpan>
hsivonen: "Authors must not discuss which of abbr and acronym should be used."

[01:03:02.0000] <MikeSmith>
if we were going to pick another something to give new semantics to, that should be at the top of the list

[01:03:03.0000] <MikeSmith>
zcorpan: :)

[01:03:04.0000] <MikeSmith>
"Authors must STFU."

[01:03:05.0000] <hsivonen>
even though underlining on the Web generally means links, mapping word processing concepts from the 1990s to the Web benefits from having <u>

[01:04:00.0000] <hsivonen>
one interesting area of use cases for presentational markup is legal stuff

[01:04:01.0000] <MikeSmith>
true that

[01:04:02.0000] <hsivonen>
where there are ancient and typographically crummy conventions

[01:04:03.0000] <hsivonen>
but the styling is not optional

[01:04:04.0000] <MikeSmith>
right

[01:05:00.0000] <MikeSmith>
same with the actual language used in legal documents

[01:05:01.0000] <hsivonen>
zcorpan: we should have that requirement

[01:05:02.0000] <hsivonen>
zcorpan: we only need two authors to bite their tongues to make past CR

[01:06:00.0000] <zcorpan>
hsivonen: they also need to conform to all other criteria that apply to authors

[01:06:01.0000] <hsivonen>
oh right. I forgot we want two complete interoperable implementations

[01:06:02.0000] <zcorpan>
we're not svg

[01:06:03.0000] <hsivonen>
while feature-wise comparison is ok in the CSS WG

[01:13:00.0000] <myakura>
boblet: I don't know the detail, but it's been split out and moved into a separate module. see http://dev.w3.org/csswg/css3-text-layout/

[01:14:00.0000] <boblet>
myakura: doh — I was reading that and didn’t go down far enough. thanks!

[01:16:00.0000] <myakura>
hm http://lists.w3.org/Archives/Public/www-style/2010Apr/0278.html

[01:19:00.0000] <boblet>
myakura: do you know if there’s a display property for <rp>, like display:ruby-text?

[01:20:00.0000] <boblet>
(ie a custom one that isn’t display:none)

[01:28:00.0000] <boblet>
aah don’t worry. not possible to do that anyway (was trying to fake <rp> via generated content, but can’t then turn it off in browsers that support <rt>)

[01:29:00.0000] <myakura>
boblet: i dunno. i guess they prolly drop rps from the rendering tree (or something like such)

[01:39:00.0000] <boblet>
Hixie: re: complex ruby, the advantage of having rbc and rtc elements is the base text and the ruby text don’t need to be intermingled if you want to add ruby per character. Having the ruby all together after the base text is definitely better for non-supporting browsers and for copy&paste. current spec means you loose per-kanji coupling to do that

[01:40:00.0000] <boblet>
Hixie: PM me if you’d like to see the sample files I’m working on

[01:42:00.0000] <Hixie>
seems like copy-paste and legacy browsers are just a temporary problem

[01:42:01.0000] <Hixie>
i'd hate to make the language that much more complicated just for a temporary issue

[01:43:00.0000] <hsivonen>
I don't like the pattern of UA sniffing Firefox and feeding it some crazy code that it previously accedintally handled but feeding sane code to IE

[01:44:00.0000] <hsivonen>
I wonder if some guru has taught people that it's OK and even a good idea to document.write a script element to load a "module" in Firefox

[01:44:01.0000] <Hixie>
the opposite pattern is worse, and more common

[01:44:02.0000] <jgraham>
hsivonen: everything after and including the word Firefox is superflouous there

[01:44:03.0000] <hsivonen>
document.write a script from a timeout, that is

[01:45:00.0000] <hsivonen>
timeout is the key here

[01:45:01.0000] <jgraham>
UA snifing never seems to end well

[01:45:02.0000] <jgraham>
</qa_bias>

[01:45:03.0000] <hsivonen>
oh document.writing from timeout doesn't work in IE, let's do it in Firefox only

[01:45:04.0000] <hsivonen>
do people think what it means to document.write from a timeout?!11!!1eleventy

[01:46:00.0000] <jgraham>
What did it mean?

[01:46:01.0000] <jgraham>
It seems like it could have all sorts of exciting consequences

[01:46:02.0000] <hsivonen>
jgraham: content gets injected into a timing-sensitive point in the input stream, but only if the parser isn't done yet

[01:47:00.0000] <zcorpan>
shouldn't it blow away the document?

[01:47:01.0000] <Hixie>
it inserted it wherever the parser was, more or less (but only at certain points, iirc, e.g. only on token boundaries. i forget the details.)

[01:47:02.0000] <jgraham>
"please insert this script at a random point in my document, maybe"?

[01:47:03.0000] <hsivonen>
zcorpan: per spec, yes. in IE, yes. in Gecko with html5.enable=true, yes

[01:47:04.0000] <zcorpan>
iirc it does in opera also

[01:47:05.0000] <Hixie>
i remember testing and finding i could insert text into attributes

[01:47:06.0000] <hsivonen>
zcorpan: in Gecko with the old parser, no if the parser isn't done

[01:48:00.0000] <hsivonen>
/me is dissecting an exciting ad script that exhibits this behavior in Minefield on Windows but not on Mac or Linux

[01:49:00.0000] <boblet>
Hixie: don’t see how copy-paste issue is temporary

[01:50:00.0000] <Hixie>
boblet: well presumably it's not how browsers _want_ it to work

[01:52:00.0000] <boblet>
Hixie: I guess browsers could reorder the text when copying from ruby, but that seems somewhat unlikely. currently even del text shows up in copy-paste

[01:52:01.0000] <boblet>
adding nested ruby to the mix though…

[01:52:02.0000] <Hixie>
well copy/paste should do what the user expects

[01:53:00.0000] <jgraham>
At risk of jumping back to the <s> discussion, with the proviso that <ins> and <del> would be the first elements against the wall if my revolution came, I think that allowing <s> or <strike> or whatver is fine and the fact that there has been discussion about what it means indicates that people won't always use it for what it means

[01:53:01.0000] <Hixie>
boblet: and i'm pretty sure hte user isn't expecting it to do as you described

[01:53:02.0000] <boblet>
I definitely didn’t expect del’ed text to show up when c&p’ing a code snippet :O

[01:53:03.0000] <jgraham>
(or what it would be specced as meaning)

[01:54:00.0000] <jgraham>
So it should be given a definition that basicaly means "render this as struck through text where possible or in a manner that conveys struck-throughness where not possible"

[01:54:01.0000] <hsivonen>
/me wonders where GeckoActiveXObject comes from

[01:55:00.0000] <boblet>
i18n and a11y make my head hurt generally. combining them both into one element seems especially cruel :)

[01:57:00.0000] <boblet>
Does anyone have a clever idea about how to style ruby text in unsupporting browsers? I can change the font-size easily enough, but anything else will also affect supporting UAs…

[01:59:00.0000] <hsivonen>
haha. the same script assumes that if it find AppleWebKit in the UA string, it'll also find "Version"

[01:59:01.0000] <hsivonen>
which isn't a safe assumption in e.g. Chrome

[01:59:02.0000] <hsivonen>
or any non-Safari WebKit-based browser

[01:59:03.0000] <MikeSmith>
hmm, I just noticed a weird rendering problem in Webkit nightly

[02:00:00.0000] <hsivonen>
Why does Safari even bother putting the app version in the UA string?

[02:00:01.0000] <Hixie>
hsivonen: probably for easier stats analysis

[02:00:02.0000] <MikeSmith>
can somebody look at http://dev.w3.org/html5/markup/figcaption.html and check the "P" in "Phrasing content" and teh "f" in "figure" under "Permitted parent elements"?

[02:01:00.0000] <Hixie>
hsivonen: makes it more likely people will report "safari" as being used

[02:02:00.0000] <MikeSmith>
about above -- I mean, first click the link so it becomes a visited link, then check the letters

[02:03:00.0000] <MikeSmith>
in Webkit nightly they show up in the default (non-visited link) color, whereas the rest of the link is in visited-link color as expected

[02:04:00.0000] <MikeSmith>
hsivonen: btw, yesterday you mentioned about checking for a nested-figcaption case -- but figcaption can't nest, right? it's only allowed as a child of figure

[02:04:01.0000] <hsivonen>
MikeSmith: can't !== isn't allowed to

[02:05:00.0000] <MikeSmith>
ah

[02:05:01.0000] <hsivonen>
MikeSmith: that is, does the internal state of the checker go crazy, if the end tag of an illegally nested figcaption flips the bit early

[02:05:02.0000] <MikeSmith>
I see

[02:06:00.0000] <MikeSmith>
well, for the one test case I tried, it just first gives me an error about the figcaption not being allowed, and the message saying it stopped tehre and did not check the rest of the subtree

[02:11:00.0000] <boblet>
MikeSmith: re: webkit link coloring, me too on r56990

[02:15:00.0000] <MikeSmith>
boblet: thanks

[02:16:00.0000] <MikeSmith>
seems to be a new bug

[02:16:01.0000] <micheil>
hmm.. Hixie are websocket servers expected to reply with Set-Cookie headers if Set-Cookie headers were sent?

[02:17:00.0000] <Hixie>
servers can do whatever they feel is right with cookie headers, basically

[02:18:00.0000] <micheil>
okay, so, if a server were to drop the Set-Cookie headers, then that wouldn't be against spec

[02:18:01.0000] <micheil>
right?

[02:30:00.0000] <boblet>
when copying ruby with per-kanji rt in Chrome, each character and rt text is in source order with a line break between

[02:30:01.0000] <jgraham>
micheil: That's what I plan on doing initially at least

[02:31:00.0000] <micheil>
jgraham: do you have your implementation up anywhere?

[02:31:01.0000] <jgraham>
micheil: It doesn't quite work yet

[02:31:02.0000] <micheil>
jgraham: so far the only spec compliant server/client I've found is in Go

[02:31:03.0000] <Hixie>
micheil: right

[02:31:04.0000] <boblet>
Chrome also doesn’t highlight the second ruby rt correctly — can show highlighted but not copy, and copy when not highlighted

[02:31:05.0000] <boblet>
fun fun

[02:32:00.0000] <Hixie>
micheil: personally my web socket toys have not used cookies at all

[02:32:01.0000] <micheil>
Hixie: any idea who wrote the one for Go? They deserve a pat on the back or something :P

[02:32:02.0000] <Hixie>
no idea

[02:32:03.0000] <Hixie>
does it do the latest handshake?

[02:33:00.0000] <Hixie>
or is it just -75 still

[02:33:01.0000] <micheil>
yeah, I think so

[02:33:02.0000] <Hixie>
cool

[02:33:03.0000] <Hixie>
i wonder if chrome has updated yet

[02:33:04.0000] <micheil>
http://golang.org/src/pkg/websocket/server.go

[02:33:05.0000] <micheil>
does the challenge / fingerprint stuff

[02:33:06.0000] <Hixie>
yeah must be

[02:33:07.0000] <micheil>
also has a module for draftHixie75

[02:34:00.0000] <Hixie>
oh looks like ukai wrote it, cool

[02:34:01.0000] <micheil>
oh neat :)

[02:34:02.0000] <micheil>
all I could find on it was "The Go Authors"

[02:35:00.0000] <zcorpan>
micheil: you should look at the TODO markers ;)

[02:36:00.0000] <micheil>
oh, good point

[02:36:01.0000] <Hixie>
looks like he doesn't do the 0xFF trailing handshake yet

[02:36:02.0000] <Hixie>
but otherwise seems solid

[02:37:00.0000] <micheil>
trailing 0xFF handshake?

[02:37:01.0000] <micheil>
/me will have to check up on that

[02:37:02.0000] <zcorpan>
0xFF0x00 is the closing handshake

[02:38:00.0000] <micheil>
ah, okay

[02:38:01.0000] <micheil>
which is the inverse of a empty message packet: 0x000xFF

[02:39:00.0000] <Hixie>
yes, though that's a coincidence more than anything

[02:39:01.0000] <micheil>
true

[02:43:00.0000] <micheil>
looks like the current release of chrome doesn't do the closing handshake

[02:45:00.0000] <Hixie>
does it do the key stuff?

[02:45:01.0000] <Hixie>
the closing handshake was added at the same time

[02:46:00.0000] <beowulf>
Hixie: http://bfbc2.statsverse.com/stats/ps3/carisenda # i have more free time than you :)

[02:47:00.0000] <micheil>
I'm not sure, I haven't tested yet

[02:47:01.0000] <Hixie>
beowulf: yet my squad score is higher :-P

[02:47:02.0000] <micheil>
/me is just beginning to write a header parser..

[02:49:00.0000] <beowulf>
*mutter*

[02:49:01.0000] <Hixie>
:-P

[02:50:00.0000] <hsivonen>
so this time the script didn't call document.write from a timeout

[02:51:00.0000] <hsivonen>
it called it whenever the HTTP load for another network resource happened to finish

[02:52:00.0000] <hsivonen>
I still wish ad script authors had a more coherent mental model of how asynchronous processing and document.write interact

[02:53:00.0000] <annevk>
oh, it's also available for the xbox

[02:53:01.0000] <Hixie>
and pc

[02:54:00.0000] <hsivonen>
FWIW, WebKit devs might want to watch out for this same class of evangelism bugs when implementing HTML5-compliant document.write

[02:55:00.0000] <hsivonen>
this particular script tried to sniff for Safari version in order to be able to do the dangerous thing in new Safari releases that "support" it

[02:57:00.0000] <hsivonen>
IE and Opera should be fine, because they haven't previously "supported" crazy asynchronous document.wrtie

[02:58:00.0000] <hsivonen>
I have seen another case, though, that means trouble for WebKit and Presto, because the script assumed that only IE and Firefox support @defer and no one else does

[02:58:01.0000] <Hixie>
ain't the web grand

[02:59:00.0000] <hsivonen>
ain't document.write grand

[02:59:01.0000] <Hixie>
i wish validator.nu warned about empty paragraphs

[02:59:02.0000] <Hixie>
i keep finding places where i have <dt> instead of </dt> and it would have been caught by that warning

[03:00:00.0000] <Hixie>
oh actually it wouldn't be caught

[03:00:01.0000] <Hixie>
maybe i should make that warning also cover phrasing content elements

[03:01:00.0000] <Hixie>
instead of just flow content ones

[03:03:00.0000] <zcorpan>
Hixie: maybe you want a warning for omitted optional tags

[03:04:00.0000] <Hixie>
no, because i use those a lot

[03:04:01.0000] <Hixie>
and indeed things would have been fine here if i'd omitted them

[03:04:02.0000] <zcorpan>
true

[03:05:00.0000] <zcorpan>
maybe you want a warning for included optional tags :P

[03:05:01.0000] <Hixie>
wouldn't have fired in this case

[03:06:00.0000] <micheil>
okay.. turns out with a little convincing ryan dahl will be accepting a websocket protocol implementation into the core of node.js

[03:06:01.0000] <Hixie>
micheil: hopefully not before the protocol is done :-)

[03:06:02.0000] <micheil>
are many more changes expected?

[03:07:00.0000] <micheil>
(as in, will the security stuff stay similar?)

[03:07:01.0000] <Hixie>
who knows

[03:07:02.0000] <Hixie>
the last set of changes weren't "expected" either

[03:07:03.0000] <micheil>
oky

[03:07:04.0000] <Hixie>
ok bed time

[03:07:05.0000] <Hixie>
nn

[03:27:00.0000] <MikeSmithX>
so apparently a brief video clip of me being interviewed will be on a Japanese TV program tomorrow night

[03:27:01.0000] <MikeSmithX>
http://www.nhk.or.jp/itwb/

[03:28:00.0000] <MikeSmithX>
they tell me that during the interview, I said "this is the most rapid pace of progress that we have ever had happening around web technologies in the entire history of the development of the web"

[03:28:01.0000] <MikeSmithX>
and that's the part they're going to use

[03:29:00.0000] <MikeSmithX>
DJs, feel free to sample that and use it in your mix tapes

[03:31:00.0000] <Philip`>
More rapid than at the times in the past when people said "hey, let's add images to the web!" and "that sounds like a great idea! let's call it <img>" and two weeks later it was implemented and deployed in all major web browsers?

[03:32:00.0000] <zcorpan>
Philip`: was it two weeks? i thought it was more like two days

[03:33:00.0000] <Philip`>
Now it takes us five years just to add something like <figure> which doesn't even do anything

[03:33:01.0000] <MikeSmithX>
Philip`: yeah, it ain't the most accurate statement in the world

[03:33:02.0000] <MikeSmithX>
but the *spirit* of the message is what counts

[03:34:00.0000] <MikeSmithX>
I told them they should have used the part where I looked straight into the camera and said, "We are kicking ass all over the rest y'all and there's nothing you can do about it, so suck it up"

[03:34:01.0000] <MikeSmithX>
but for some reason they didn't like that part so much

[03:35:00.0000] <Philip`>
You should have just repeated that phrase for the entirety of the interview, so they'd have no choice but to pick it

[03:35:01.0000] <zcorpan>
MikeSmithX: did you also say "BTW Adobe eats babies"?

[03:35:02.0000] <MikeSmithX>
heh

[03:36:00.0000] <zcorpan>
Philip`: then they'd probably show the entire interview and MikeSmithX would look like a dork on tv

[03:37:00.0000] <nessy>
Adobe eats my macbook right now

[03:38:00.0000] <nessy>
if you compare the pace of innovation of html5 to the introduction of <canvas>, then it's light speed

[03:40:00.0000] <MikeSmith>
zcorpan: I told them, "My message to them is very simple: We drink your milkshake!"

[03:41:00.0000] <MikeSmith>
to paraphrase Daniel Day Lewis

[03:42:00.0000] <zcorpan>
/me doesn't know what MikeSmith is referring to

[03:43:00.0000] <Philip`>
MikeSmith: Did you then kill the interviewer?

[03:43:01.0000] <MikeSmith>
I did

[03:43:02.0000] <MikeSmith>
and then I ate the pieces of him that were left over

[03:44:00.0000] <zcorpan>
/me now knows not to interview MikeSmith

[03:44:01.0000] <annevk>
o_O

[03:54:00.0000] <micheil>
Hixie: latest google chrome beta doens't support the latest spec

[03:54:01.0000] <micheil>
(based on testing from go's websocket.Handler vs websocket.draft75Handler)

[03:55:00.0000] <micheil>
now I just need to work out how to use it's client version to test against my own implementations

[03:58:00.0000] <jgraham>
micheil: If you need really simple testing you can of course just copy the example from the spec and send it over a raw client socket 

[03:59:00.0000] <jgraham>
Won't test all edge cases or anything but should determine if your code works at all

[05:36:00.0000] <micheil>
jgraham: yeah, I've been doing telnet's to test so far

[06:31:00.0000] <Matjas>
I recently launched a new site in HTML5, is anyone willing to look at the source and tell me if I'm doing it right? :) http://mathiasbynens.be/

[06:32:00.0000] <Matjas>
I just changed section.note into div.note, to improve my document outline — but is there a better alternative?

[06:34:00.0000] <Matjas>
Also see http://gsnedders.html5.org/outliner/process.py?url=http://mathiasbynens.be/

[06:35:00.0000] <annevk>
shortcut is not a valid link relation

[06:36:00.0000] <annevk>
/me is not really inspired right now; should do some more packing for Tokyo

[06:36:01.0000] <adactio>
Matjas: I think maybe your div.note could be a header element. What do you think?

[06:38:00.0000] <Matjas_>
annevk: you mean 'shortcut icon'? what should i use instead?

[06:40:00.0000] <Matjas>
adactio: Hrm, I really don't know. It only appears on the index page

[06:40:01.0000] <Matjas>
If that even matters

[06:41:00.0000] <adactio>
Matjas: Well, I think the contents of div.note would make good header content for the section it's in (I'm thinking on the document level rather than a site level).

[06:43:00.0000] <Matjas>
adactio: I see. And it probably wouldn't be a problem that this header element only occurs on one page, right?

[06:43:01.0000] <adactio>
Matjas: I don't think so, no

[06:45:00.0000] <Matjas>
adactio: Thanks. Awesome, now I can even omit the class on the element

[06:46:00.0000] <Matjas>
annevk: Switched to rel="icon", thanks

[06:48:00.0000] <Matjas>
annevk: Does that work cross-browser?

[06:52:00.0000] <svl>
Matjas: no; but the browser in which it doesn't work just looks for favicon.ico in / regardless of what you tell it, so this way it'll work "as intended" everywhere.

[06:53:00.0000] <Matjas>
svl: Thanks. Which IE versions are we talking about? :)

[06:54:00.0000] <svl>
I think 6 and 7. not certain about 8

[06:56:00.0000] <Matjas>
So why is everyone using shortcut as well? Was is valid in HTML 4.01?

[06:57:00.0000] <Matjas>
I found this post on the WHATWG blog which contains an example with shortcut: http://blog.whatwg.org/the-road-to-html-5-link-relations#rel-icon

[06:58:00.0000] <Matjas>
/me is confused

[06:59:00.0000] <svl>
IE needs shortcut if you don't name your icon favicon.ico

[07:00:00.0000] <Matjas>
So shortcut is not in any spec, but is proprietary? :o

[07:00:01.0000] <svl>
Yes. IIRC it even needs it as the first value, rather than properly treating it as a space-separated value list.

[07:00:02.0000] <Matjas>
Mind = blown.

[07:03:00.0000] <Matjas>
Thanks svl!

[07:28:00.0000] <zcorpan>
http://www.w3.org/XML/2010/04/xml-stylesheet/ - PER in a few days

[09:10:00.0000] <JonathanNeal>
Heyo

[09:13:00.0000] <Matjas>
I'm tired of using technology

[09:13:01.0000] <Matjas>
Why don't you sit down on top of me

[09:13:02.0000] <Matjas>
OH SHI—

[09:15:00.0000] <JonathanNeal>
... good exit

[09:17:00.0000] <miketaylr>
indeed

[11:59:00.0000] <AryehGregor>
Why do Firefox and Chrome both render this with the <th> not centered?  data:text/html,<!doctype html><table style="text-align:left"><tr><th>Foo</th></tr><td>Barzabel</td></table>

[12:00:00.0000] <AryehGregor>
There should be a UA rule like th { text-align: center }, but there's not.  Instead, <th> contents seem to be centered by magic.

[12:00:01.0000] <zcorpan>
hmm, who was it that took my placeholder.js script and improved it a few days ago?

[12:01:00.0000] <zcorpan>
oh here it is http://fsfe.org/scripts/placeholder.js

[12:01:01.0000] <AryehGregor>
Well, I guess I mean: data:text/html,<!doctype html><table style="text-align:left"><tr><th>Foo<tr><td>Barzabel</table>

[12:01:02.0000] <AryehGregor>
Should produce the same DOM, anyway.

[12:01:03.0000] <AryehGregor>
Opera too.

[12:01:04.0000] <AryehGregor>
Only IE8 seems to work as expected.

[12:02:00.0000] <TabAtkins>
HTML5 specifies that <th> should have text-align:center, not magic.

[12:02:01.0000] <TabAtkins>
So it'll presumably move over in time.

[12:03:00.0000] <AryehGregor>
So does CSS 2.1.

[12:18:00.0000] <othermaciej>
WebKit has magic

[12:18:01.0000] <othermaciej>
it converts text-align: auto to text-align: center at style resolution time

[12:19:00.0000] <othermaciej>
for th elements

[12:19:01.0000] <AryehGregor>
https://bugs.webkit.org/show_bug.cgi?id=37590

[12:19:02.0000] <AryehGregor>
Is this necessary for compat?

[12:19:03.0000] <othermaciej>
I am not sure if it's possible to do all the weird legacy compat stuff with an explicit text-align: center rule in place

[12:20:00.0000] <AryehGregor>
What weird legacy compat stuff needs to be done?  Are th's sometimes magically not centered?

[12:20:01.0000] <AryehGregor>
Does this need to be specced somewhere?

[12:21:00.0000] <TabAtkins>
I'm curious too.  If you're referring to some of the aligning attributes or whatnot, those are on the author level, and thus override UA-level rules.

[12:51:00.0000] <othermaciej>
AryehGregor: I don't know if our behavior is necessary

[12:51:01.0000] <othermaciej>
AryehGregor: is the behavior cited in the bug above divergent from other browsers?

[12:51:02.0000] <othermaciej>
very specifically what we do is that for th, text-align: auto is interpreted as text-align: center at style resolution time

[12:51:03.0000] <AryehGregor>
othermaciej, I covered that in the bug.  Other browsers do the same, except IE8.

[12:52:00.0000] <othermaciej>
(instead of being equivalent to left as for most elements)

[12:52:01.0000] <othermaciej>
I suspect the bug could be fixed by making th have an explicit text-align: auto property in the UA stylesheet

[12:52:02.0000] <othermaciej>
but I don't know if this would break anything else

[12:52:03.0000] <zcorpan>
/me updates http://simon.html5.org/sandbox/js/placeholder-demo.htm

[12:53:00.0000] <AryehGregor>
Hmm, weird.

[13:39:00.0000] <smaug>
Hixie: thanks

[13:45:00.0000] <Hixie>
np

[14:09:00.0000] <estellevw>
hi. There's a small discussion going on about section versus article in my blog, and i would love to get some feedback (either here or within the comments section) as to what is right.  My understanding of the spec is that sections can be nested in articles, and articles can be nested in sections.  Am i wrong? http://www.standardista.com/html5-section-v-article/

[14:09:01.0000] <Dashiva>
That's right

[14:11:00.0000] <Dashiva>
It might possibly be easier to think "articles (which have sections) can contain articles" instead of "sections can contain articles"

[14:11:01.0000] <JonathanNeal>
Hey, everybody!

[14:12:00.0000] <paul_irish>
so in general, Dashiva, you expect sections will be contained by an article?

[14:14:00.0000] <JonathanNeal>
Whoa whoa whoa, are we saying that articles and sections are no longer mutually exclusive?

[14:15:00.0000] <paul_irish>
(i'm not asking for rules, just what we expect common use will be)

[14:16:00.0000] <Dashiva>
paul_irish: No

[14:17:00.0000] <Dashiva>
But you can consider the a sufficiently important child of the <body> element a pseudo-article

[14:21:00.0000] <estellevw>
i do think of it as.... if its datable, it's an article

[14:21:01.0000] <estellevw>
so, you can have articles.... dated items ... a group of them, in a section

[14:21:02.0000] <estellevw>
and you can have a bunch of sections in a very long article

[14:22:00.0000] <estellevw>
i don't see off the top of my head a use case for article within article, but i do section in section

[14:22:01.0000] <estellevw>
doesn't mean there isn't a use case for articles within articles

[14:22:02.0000] <estellevw>
i just can't come up with one

[14:23:00.0000] <Dashiva>
Newspaper letters to the editor

[14:23:01.0000] <Dashiva>
The page itself is one article, but each letter can also be an article

[14:24:00.0000] <estellevw>
ah, good one

[14:24:01.0000] <estellevw>
thank you

[14:30:00.0000] <JonathanNeal>
Asides can be in articles, right?

[14:35:00.0000] <Hixie>
sure

[14:37:00.0000] <JonathanNeal>
Goodie.

[14:39:00.0000] <othermaciej>
estellevw: the spec definitions of article and section just got reworded recently

[14:39:01.0000] <othermaciej>
estellevw: don't know if that addresses any of your comment debate but it might be worth looking at

[14:39:02.0000] <JonathanNeal>
oh neato

[14:40:00.0000] <othermaciej>
estellevw: I like your explanation of how a section can be either bigger or smaller than an article

[14:40:01.0000] <othermaciej>
that makes sense to me

[14:40:02.0000] <estellevw>
thanks

[14:40:03.0000] <estellevw>
re-reading the spec :(

[14:41:00.0000] <JonathanNeal>
/me looking for the change

[14:41:01.0000] <othermaciej>
reading the spec is a :( instead of a :)?

[14:41:02.0000] <othermaciej>
it's mainly in the first sentence of each definition I think

[14:41:03.0000] <estellevw>
lol. 

[14:41:04.0000] <JonathanNeal>
reading the spec is #)

[14:41:05.0000] <estellevw>
yeah, the way i was thinking of it still makes sense

[14:41:06.0000] <othermaciej>
sometimes reading the spec is O_o

[14:42:00.0000] <estellevw>
i hadn't read the spec since Friday night (because i have a life ;) )

[14:42:01.0000] <JonathanNeal>
i hadn't read the spec because I'm a zombie with his tongue sticking out :PF

[14:42:02.0000] <estellevw>
i was afraid it had changed this then... now just realized, i reread that on monday. 

[14:43:00.0000] <Hixie>
when reading the spec is O_o, please file a bug :-)

[14:44:00.0000] <estellevw>
Hixie: The spec is so wonderful, that I haven't read any thing else is a few months :D

[14:44:01.0000] <othermaciej>
Hixie: when I have that reaction, it's usually due to some crazy thing that's needed for legacy compat

[14:45:00.0000] <estellevw>
is there a way to see ... other than the top of the page, when something has changed

[14:45:01.0000] <estellevw>
like an SVN diff per line?

[14:45:02.0000] <othermaciej>
estellevw: there is a twitter feed of spec changes

[14:45:03.0000] <estellevw>
ah. twitterName please

[14:45:04.0000] <othermaciej>
http://twitter.com/html5

[14:45:05.0000] <estellevw>
thanks

[14:45:06.0000] <othermaciej>
that links to the diffs for each change

[14:45:07.0000] <Hixie>
if you just want non-editorial changes, twitter.com/whatwg might be more useful

[14:46:00.0000] <KaOSoFt>
Hello.

[14:46:01.0000] <Hixie>
it also includes diffs for other specs than html5 that are part of complete.html, like websockets

[14:46:02.0000] <Hixie>
also http://html5.org/tools/web-apps-tracker might be helpful

[14:47:00.0000] <estellevw>
ah, that is perfect. thank you

[14:48:00.0000] <othermaciej>
I like the examples from your post of a newspaper section vs. a section of an article

[14:48:01.0000] <othermaciej>
maybe the spec could cite examples like that

[14:48:02.0000] <Hixie>
happy to add more section and article examples, though there's already quite a few

[14:48:03.0000] <Hixie>
file a bug

[14:49:00.0000] <Hixie>
maybe i should have a summary of usage section for the sectioning elements

[14:49:01.0000] <Hixie>
like we do for phrasing

[14:49:02.0000] <Hixie>
s/have/add/

[14:49:03.0000] <KaOSoFt>
It isn't mandatory to have a form if I want an <input> element, right? I just need a text-box and a submit button for searching.

[14:49:04.0000] <othermaciej>
that sounds like a good idea

[14:49:05.0000] <othermaciej>
it looks like you don't have an example of a section containing articles

[14:50:00.0000] <Hixie>
KaOSoFt: correc

[14:50:01.0000] <Hixie>
t

[14:50:02.0000] <othermaciej>
oh, I guess you do, for the blog post with comments

[14:50:03.0000] <KaOSoFt>
Hixie- Thanks. ^_^

[14:53:00.0000] <KaOSoFt>
"<othermaciej> I like the examples from your post of a newspaper section vs. a section of an article"

[14:53:01.0000] <KaOSoFt>
Where is this?

[14:53:02.0000] <othermaciej>
KaOSoFt: http://www.standardista.com/html5-section-v-article/

[14:57:00.0000] <JonathanNeal>
estellevw's explanation is probably one of the best.

[14:59:00.0000] <JonathanNeal>
Newspaper letters to the editor, I was wondering if those are sections vs articles, much like a comment section might be a section.  However, in the comment section one could say that is definitely datable.

[15:01:00.0000] <JonathanNeal>
If I knew how to submit stuff to gsnedders's tool, I would suggest the outline allows marking which type of element is in use (it might make people feel better about those untitled navs and asides)

[15:01:01.0000] <JonathanNeal>
I would suggest == I would write and submit for review

[15:03:00.0000] <JonathanNeal>
What is the recommended element for comments on a blog post?

[15:03:01.0000] <estellevw>
the comments is a section, and each comment is dates do article

[15:04:00.0000] <estellevw>
do = so

[15:04:01.0000] <JonathanNeal>
And would that section belong to the article that is the blog post?

[15:04:02.0000] <estellevw>
yes

[15:04:03.0000] <estellevw>
at least, that's my take on it

[15:04:04.0000] <TabAtkins__>
/me came in late, but agrees with what estellevw is saying.

[15:05:00.0000] <JonathanNeal>
<article>... blog post heading and content <section>... blog post comment heading and <article ... various comments /></section></article>

[15:05:01.0000] <JonathanNeal>
where various comments represents multiple articles.

[15:06:00.0000] <JonathanNeal>
I agree too, just mulling the concept over.

[15:08:00.0000] <JonathanNeal>
In a traditional newspaper, each article references the date of the entire paper (usually at the top).  Wonder if there would be a geeky semantic way to write in that reference.  I would totally make a fake newspaper as HTML5 sample.

[15:24:00.0000] <TabAtkins__>
JonathanNeal: Nope, you'd have to include the <time pubdate> in every <article>.

[15:24:01.0000] <JonathanNeal>
No referencing?  Not even with microdata?

[15:31:00.0000] <TabAtkins>
If you have a metadata vocabulary appropriate to the task, then of course you can reference it.  But that has nothing to do with the implied connection between an <article> and a <time pubdate> that HTML5 defines.

[15:32:00.0000] <estellevw>
hhmmm, so a nested article doesn't inherit the pubdate of the parent article.

[15:33:00.0000] <estellevw>
makes sesne

[15:33:01.0000] <estellevw>
if instead of thinking of a daily newspaper, think of a monthly mag maybe

[15:33:02.0000] <TabAtkins>
Or just think of blog posts - the pubdate on the post has nothing to do with the pubdate on the comments.

[15:33:03.0000] <estellevw>
the magazine may come out on april 1, but the letters to the editor each have their own dates

[15:34:00.0000] <estellevw>
exactly

[15:34:01.0000] <JonathanNeal>
All of that would work even if it inherited the date.

[15:34:02.0000] <JonathanNeal>
If you do not specify the date, it could be implied from an ancestor.

[15:36:00.0000] <JonathanNeal>
Well, right now you can't, but in theory what issue would you take with that?

[15:38:00.0000] <TabAtkins>
The fact that it would be inaccurate?

[15:38:01.0000] <TabAtkins>
Especially in the common case, that of blog comments?

[15:44:00.0000] <JonathanNeal>
wouldn't the blog comments specify the date?

[15:44:01.0000] <JonathanNeal>
Otherwise, they're just as omitted in either case.

[15:44:02.0000] <TabAtkins>
Possibly.  If they do, they're covered.  But if they don't, your suggestion would imply that they have the same pubdate as the article they're attached to.

[15:46:00.0000] <JonathanNeal>
You'd rather they had nothing than possibly something.  That's fair, you can have them work strict in that sense.

[15:48:00.0000] <JonathanNeal>
It would be my preference that inheritance is an option.  I may be alone.

[15:49:00.0000] <TabAtkins>
What's wrong with just specifying <time pubdate> in the descendant articles?

[15:51:00.0000] <JonathanNeal>
It's easier to not do something than to do something.  I think it could be less strict.  I think there's wiggle room.  To compare, we've been talking about comments as articles, then comments aren't exactly independantly distributable content.

[15:52:00.0000] <JonathanNeal>
\s\

[15:52:01.0000] <JonathanNeal>
then\but

[15:52:02.0000] <TabAtkins>
They're independent chunks of content that can receive their own title and publication date, and are reasonable to link to and look at individually.

[15:52:03.0000] <TabAtkins>
That's close enough to make <article> a great match, I think.

[15:53:00.0000] <JonathanNeal>
Do individual comments commonly receive titles?

[15:54:00.0000] <TabAtkins>
I don't see it universally, but it's reasonable common.  Techdirt, for example, has titles on their comments.

[16:00:00.0000] <estellevw>
It does happen, but it's not the norm (yet).

[16:01:00.0000] <estellevw>
would most table of contents, though, list each comment individually, or just have the header "comments" in the TOC?

[16:01:01.0000] <estellevw>
it really depends on the blog i would think

[16:01:02.0000] <TabAtkins>
Probably the latter.

[16:02:00.0000] <estellevw>
the "great post" and chatter is not deserving

[16:02:01.0000] <estellevw>
but some blogs have well thought out responses

[16:02:02.0000] <estellevw>
even then, one would think the main author would want to determine their TOC

[16:02:03.0000] <estellevw>
not have their TOC partially filled in my commenters

[16:02:04.0000] <estellevw>
so, the latter... yup

[16:03:00.0000] <estellevw>
thinking with my fingers... sorry

[16:03:01.0000] <TabAtkins>
np

[16:16:00.0000] <JonathanNeal>
estellevw, thanks for the link to print protector :) just saw

[16:23:00.0000] <estellevw>
np


2010-04-15
[18:30:00.0000] <TabAtkins>
othermaciej: FWIW, that *was* me trying to be as civil as possible.

[18:32:00.0000] <TabAtkins>
Of course, now we'll probably be treated to another round of thinly-veiled accusations of sexism on alternate content streams.

[18:32:01.0000] <TabAtkins>
Sigh.  You just can't win when your opponent is certain that you're evil and wrong.

[18:32:02.0000] <othermaciej>
TabAtkins: just try not to pour fuel on the fire

[18:33:00.0000] <othermaciej>
I recognize that Shelley's tone went somewhat hostile, but a statement like "Also, seriously, chill" seems more likely to escalate the conflict than to cool it off, given the history

[18:34:00.0000] <TabAtkins>
Heh, that was meant as an amusing callback to our previous exchange.

[18:34:01.0000] <TabAtkins>
I decided to forgo the smilie this time, as that seems to have the opposite effect of what is intended.

[18:38:00.0000] <othermaciej>
I expect Shelley was not amused, and I also think you could have predicted that she would not be amused

[18:38:01.0000] <othermaciej>
I don't mean to be a hardass here, just giving the usual "no fighting in the war room" admonition

[18:42:00.0000] <estellevw>
inquiring minds want to know: where was the drama?

[18:42:01.0000] <Hixie>
public-html as always

[18:45:00.0000] <TabAtkins>
othermaciej: I am *seriously* not trying to fan any flames with my latest email.  Apologies if anyone does still get mad.  ;_;

[18:45:01.0000] <othermaciej>
now with up to 80% less drama!

[18:46:00.0000] <othermaciej>
just don't make me come back there

[18:47:00.0000] <TabAtkins>
You shouldn't have to, unless Shelley gets offended again.  I'm really trying to be as nice as possible here (as opposed to before, when I was being as nice as reasonable).

[18:49:00.0000] <Hixie>
what exactly are you proposing? I must have skipped past the part of the thread where you suggested something

[18:49:01.0000] <TabAtkins>
I asked her to remove the points from her rationale that are solely arguments against @sandbox, rather than against @srcdoc.

[18:50:00.0000] <Hixie>
oh

[18:51:00.0000] <Hixie>
well hell, we wouldn't want her to make her proposals more readable, they'd be more likely to pass!

[18:51:01.0000] <Hixie>
why are you suggesting such a thing

[18:51:02.0000] <Hixie>
oh wait, i get it, it's reverse psychology! very clever

[18:51:03.0000] <TabAtkins>
Because I have an irrational urge to make the process actually work as intended.

[18:52:00.0000] <Hixie>
you must be new here

[18:52:01.0000] <Hixie>
:-P

[18:52:02.0000] <Hixie>
ah well. we'll have you turned into a cynic soon enough. no need to rush it i guess.

[18:52:03.0000] <TabAtkins>
I'm a cynical idealist.  Why isn't this good enough?

[18:52:04.0000] <TabAtkins>
Anyway, out for the night.

[18:53:00.0000] <Hixie>
later

[20:13:00.0000] <TabAtkins___>
Hixie, brainstorming about timed text markup formats.  What were the extensions you were thinking of for timestamps and stage directions in <dialog>?

[21:13:00.0000] <MikeSmith>
Hixie: about http://www.w3.org/Bugs/Public/show_bug.cgi?id=9472 (deal with broken links in dynamic author view of spec by prompting users to choose to switch to full view)

[21:13:01.0000] <MikeSmith>
did you deploy a change for that already?

[21:48:00.0000] <boblet>
anyone know if Arabic Taškīl, Hebrew Niqqud, or other non-CJK languages would use ruby for marking up phonetic annotations or diacritics?

[21:59:00.0000] <MikeSmith>
boblet: Richard Ishida would know

[22:00:00.0000] <MikeSmith>
I seem to remember some examples in ruby supplementary docs of non-Chinese/Japanese languages that had ruby annotations

[22:00:01.0000] <MikeSmith>
but I dunno remember where

[22:00:02.0000] <boblet>
MikeSmith: you’ll be thinking of the dodgy abbreviation WWW and expiry date in Ruby Annotation I bet

[22:01:00.0000] <boblet>
yeah I need to ping Richard, he’d definitely know. will email him

[22:01:01.0000] <MikeSmith>
?

[22:02:00.0000] <boblet>
abbr as ruby: http://www.w3.org/TR/ruby/#simple-ruby1

[22:02:01.0000] <boblet>
ruby as formatting hack: http://www.w3.org/TR/ruby/#complex

[22:03:00.0000] <boblet>
MikeSmith: apart from that I’ve only found Chinese & Japanese examples. I can’t even determine if Korean use hangul ruby with hanzi or not

[22:04:00.0000] <MikeSmith>
boblet: I have never seen any examples of hanzi with ruby

[22:04:01.0000] <MikeSmith>
but I don't encounter a lot of Korean text anyway

[22:05:00.0000] <boblet>
MikeSmith: me neither, so wondering if they just don’t do it. South Korea is basically moving to all hangul, but I think they still use about 1000 hanzi. North Korea supposedly completely stopped teaching them in like the 60s but then reintroduced them due to Chinese

[22:06:00.0000] <boblet>
don’t know any Koreans I can ask :(

[22:06:01.0000] <MikeSmith>
they still use a lot of hanzi in Korean, from what I have seen while visiting there

[22:06:02.0000] <MikeSmith>
e.g., many buildings have names written in hanzi and such

[22:07:00.0000] <MikeSmith>
and 20-years-olds or so still know many kanji

[22:07:01.0000] <MikeSmith>
I think they still teach them in high school

[22:08:00.0000] <MikeSmith>
but I suspect in Korean the logograms all have only a single reading

[22:09:00.0000] <MikeSmith>
so the reading is unambiguous, and they wouldn't need ruby for showing phonetic readings

[22:09:01.0000] <MikeSmith>
I think

[22:10:00.0000] <boblet>
well, assuming you know how to read them ;-) wonder if they have hanzi/hangul/romanisation on eg train signs. don’t remember

[22:11:00.0000] <paul_irish>
boblet: what are you writing that article on?

[22:11:01.0000] <boblet>
paul_irish: HTML5 ruby, for HTML5Doctors

[22:11:02.0000] <paul_irish>
ruby! sweeet

[22:11:03.0000] <boblet>
it’s killing me >_<

[22:12:00.0000] <paul_irish>
ruby and rt are so sexy

[22:12:01.0000] <boblet>
emoticon fail that should be the Japanese ＞＿＜ emoji

[22:12:02.0000] <boblet>
paul_irish: like I said, i18n and a11y are bad enough on their own. combining them

[22:12:03.0000] <boblet>
ouch

[22:43:00.0000] <MikeSmith>
http://webkit.org/blog/1091/more-web-inspector-updates/

[22:43:01.0000] <MikeSmith>
cool stuff

[22:56:00.0000] <MikeSmith>
boblet: can you think of a Japanese equivalent of "Eeny, meeny, miny, moe"?

[22:56:01.0000] <MikeSmith>
oh

[22:56:02.0000] <MikeSmith>
found it

[22:57:00.0000] <MikeSmith>
eigiro to the rescue

[22:57:01.0000] <MikeSmith>
http://eow.alc.co.jp/Eeny/UTF-8/?ref=sa

[22:57:02.0000] <MikeSmith>
hmm, it's only an explanation

[22:58:00.0000] <MikeSmith>
doesn't say what equivalent if any there is

[23:10:00.0000] <wirepair>
doesn't everyone just do junken?

[23:10:01.0000] <wirepair>
oh nevermind, you mean for selecting somethign ;>

[23:18:00.0000] <MikeSmith>
wirepair: yeah, for picking somebody to be "it"

[23:18:01.0000] <wirepair>
mike, just asked a friend they do the same thing 'do re ni shi yo u ka na'

[23:18:02.0000] <wirepair>
ten no ka mi sa ma

[23:18:03.0000] <wirepair>
...

[23:18:04.0000] <wirepair>
etc

[23:19:00.0000] <wirepair>
so what is in that alc description is right hehe

[23:19:01.0000] <MikeSmith>
intersting

[23:19:02.0000] <MikeSmith>
I guess I can ask my daughter too

[23:19:03.0000] <MikeSmith>
for an authoritative 

[23:19:04.0000] <MikeSmith>
viewpoint

[23:19:05.0000] <MikeSmith>
since she must use something when playing kakurembo etc.

[23:20:00.0000] <wirepair>
i just found out they do the same thing with flower petals

[23:21:00.0000] <wirepair>
he seemed really surprisd we do it too ;>

[23:21:01.0000] <wirepair>
surprised rather

[23:56:00.0000] <Hixie>
MikeSmith: yes

[23:56:01.0000] <Hixie>
MikeSmith: (re the links in author mode thing)

[00:00:00.0000] <MikeSmith>
Hixie: what's the expected behavior?

[00:00:01.0000] <MikeSmith>
does it pop up a prompt?

[00:00:02.0000] <Hixie>
yeah

[00:00:03.0000] <MikeSmith>
Ok

[00:00:04.0000] <MikeSmith>
I'll have to try again

[00:01:00.0000] <MikeSmith>
I tried in Webkit and Minefield and it didn't seem to work

[00:01:01.0000] <Hixie>
hm

[00:01:02.0000] <Hixie>
what were you trying?

[00:02:00.0000] <MikeSmith>
I'm trying clicking on the links in the IDLs

[00:10:00.0000] <Hixie>
on what page?

[00:11:00.0000] <zcorpan>
it only seems to work when loading the page with a broken fragment, not loading then clicking a broken link

[00:14:00.0000] <Hixie>
make sure the browser you are using supports the 'hashchange' event

[00:22:00.0000] <zcorpan>
ok, works fine for me in chrome

[00:44:00.0000] <myakura>
Hixie: I got a comment in the public-html-ig-jp which asks whether <figcaption> can have flow content inside: http://lists.w3.org/Archives/Public/public-html-ig-jp/2010Apr/0010.html

[00:44:01.0000] <Hixie>
Content model:

[00:44:02.0000] <Hixie>
Phrasing content.

[00:44:03.0000] <Hixie>
so no.

[00:45:00.0000] <myakura>
I guess so. Then the example in http://www.whatwg.org/html5#table-descriptions (labbeled "Next to the table, in a figure's figcaption") is wrong :)

[00:46:00.0000] <Hixie>
oh, good point

[00:46:01.0000] <Hixie>
i'll file a bug, thanks

[00:46:02.0000] <zcorpan>
wait what, didn't we change the content model to flow before?

[00:46:03.0000] <zcorpan>
when it was called <legend>

[00:46:04.0000] <Hixie>
it probably changed back when i made it <dt>

[00:47:00.0000] <zcorpan>
it should allow flow, since there are use cases for it (same as <caption>)

[00:48:00.0000] <Hixie>
already filed the bug

[00:50:00.0000] <zcorpan>
ok, first i thought you wanted to change the example

[00:51:00.0000] <Hixie>
ah, no, sorry for the confusion

[00:52:00.0000] <myakura>
/me was about to reply that the example was wrong. phew.

[00:55:00.0000] <zcorpan>
Hixie: why does send() return a boolean? it seems the browser could return true but the sending will still fail (e.g. because the server has closed the connection before the data is actually sent)

[00:55:01.0000] <zcorpan>
Hixie: a more reliable check would involve looking at wasClean and bufferedAmount, and then the return value of send() seems useless

[00:56:00.0000] <Hixie>
the return value is just to help you if you're sending data on a timeout, i think

[00:56:01.0000] <Hixie>
it's a coarse and blunt instrument

[00:57:00.0000] <zcorpan>
what does help with?

[00:59:00.0000] <Hixie>
come again?

[00:59:01.0000] <zcorpan>
i mean, why do we have a boolean return value at all? what problem does it solve?

[01:00:00.0000] <Hixie>
someone asked for it at some point

[01:00:01.0000] <Hixie>
it seemed harmless to provide

[01:00:02.0000] <Hixie>
i forget the details

[01:06:00.0000] <Hixie>
fixed figcaption to be flow

[01:06:01.0000] <Hixie>
i am so not looking forward to working out how Selection.modify() works

[01:08:00.0000] <MikeSmith>
Hixie: OK, I see (about the link pop-up feature)

[01:08:01.0000] <MikeSmith>
thanks

[01:08:02.0000] <Hixie>
np

[01:08:03.0000] <Hixie>
if you know how to make it work in other browsers let me know

[01:09:00.0000] <Hixie>
should work at least in chrome and ie

[01:09:01.0000] <MikeSmith>
what browsers suppor the hashchange event? IE and CHrome?

[01:09:02.0000] <MikeSmith>
kk

[01:12:00.0000] <zcorpan>
Hixie: you could listen for click events

[01:13:00.0000] <zcorpan>
Hixie: although personally i'd just wait for browsers to implement hashchange :)

[01:14:00.0000] <MikeSmith>
yeah

[01:15:00.0000] <MikeSmith>
this gives other browser projects some good incentive to implement it

[01:21:00.0000] <myakura>
MikeSmith: so how was the tv show? did you watch it?

[01:23:00.0000] <MikeSmith>
myakura: it's actually on tonight

[01:23:01.0000] <MikeSmith>
at 11:30

[01:26:00.0000] <myakura>
MikeSmith: d'oh. I thought it was 14th.

[01:39:00.0000] <Hixie>
i actually thought everyone did hashchange when i implemented it, i woulda checked otherwise :-)

[01:52:00.0000] <JonathanNeal>
ben_alman wrote a pretty good program for detecting hash changes.

[01:53:00.0000] <JonathanNeal>
May not be at all what you're talking about or looking for, but just in case http://benalman.com/projects/jquery-bbq-plugin/

[01:53:01.0000] <zcorpan>
http://yfrog.com/06h39aj

[01:54:00.0000] <zcorpan>
w00t http://www.apple.com/euro/itunes/charts/apps/top10appstorefree.html

[01:55:00.0000] <JonathanNeal>
yea that's pretty awesome for Opera.

[01:55:01.0000] <hsivonen>
zcorpan: something wrong with Australia :-)

[01:56:00.0000] <zcorpan>
hsivonen: yeah, what's up there

[01:57:00.0000] <JonathanNeal>
I dunno, they were Opera earlier today.

[01:57:01.0000] <JonathanNeal>
They love their birds, I guess.

[01:57:02.0000] <annevk>
tweeted

[01:58:00.0000] <JonathanNeal>
I saw Opera on there the other day, it was so freaking fast and snappy.

[01:58:01.0000] <annevk>
or twitted or ...

[01:58:02.0000] <othermaciej>
neat!

[01:58:03.0000] <othermaciej>
though it makes me want to download Bird Strike

[01:58:04.0000] <JonathanNeal>
Don't do it!  It's #2 in the US.

[02:17:00.0000] <MikeSmithX>
Hixie: I recommend mixing around your browser usage now and then :p

[02:17:01.0000] <MikeSmithX>
your hashchange assumption is a clear sign that you are secretly a dedicated IE user/fanboy

[02:18:00.0000] <JonathanNeal>
isn't everybody an ie fanboy

[02:18:01.0000] <zcorpan>
i wonder if ie copes with complete.html

[02:29:00.0000] <MikeSmith>
I just got a great error message from my IMAP server via mutt : "Too many consecutive protocol violations."

[02:30:00.0000] <MikeSmith>
Hixie, othermaciej - I'm thinking of setting up a mailbot that automatically raises bugzilla bugs for all new messages posted to public-html-comments

[02:31:00.0000] <othermaciej>
MikeSmith: hmmm

[02:31:01.0000] <MikeSmith>
where new = does not have an In-reply-to or References header

[02:31:02.0000] <MikeSmith>
othermaciej: bad idea?

[02:31:03.0000] <Hixie>
you're gonna have to hire someone else to help me with dealing with the noise if you do that

[02:31:04.0000] <othermaciej>
I'm thinking about whether typical emails to that list make for a good bug

[02:31:05.0000] <MikeSmith>
hmm, OK

[02:31:06.0000] <othermaciej>
I think a lot of the emails there have one of the following properties:

[02:32:00.0000] <othermaciej>
- really an idea for discussion, not an actionable problem report

[02:32:01.0000] <othermaciej>
- include large numbers of potentially separate issues

[02:32:02.0000] <othermaciej>
there have also been comments against different specs

[02:32:03.0000] <MikeSmith>
ok

[02:32:04.0000] <MikeSmith>
point taken

[02:32:05.0000] <othermaciej>
I don't know if I have seen a single one that would meet all our criteria for good info to put in a bug report

[02:33:00.0000] <MikeSmith>
clearly some human needs to be in the decision pipe

[02:33:01.0000] <othermaciej>
maybe an on-demand version of this tool would be useful

[02:33:02.0000] <othermaciej>
where it's easy for some trusted set of people to convert a public-html-comments email into a bug

[02:33:03.0000] <MikeSmith>
yeah

[02:33:04.0000] <MikeSmith>
that's a better approah

[02:33:05.0000] <MikeSmith>
*way

[02:34:00.0000] <MikeSmith>
just thinking about how we handle comments on that list after we start LC

[02:35:00.0000] <othermaciej>
lots of actual discussion there too, at least in the past month :-/

[02:35:01.0000] <othermaciej>
I think our Last Call statement needs to indicate that the way to give a Last Call comment and get a response is to file a bug

[02:35:02.0000] <MikeSmith>
I would be very happy with that

[02:35:03.0000] <MikeSmith>
would make all our lives a lot easier

[02:35:04.0000] <othermaciej>
Paul Cotton seemed to indicate that this was seen as acceptable in the past for other WGs

[02:36:00.0000] <MikeSmith>
yeah

[02:36:01.0000] <MikeSmith>
so let's do that

[02:36:02.0000] <MikeSmith>
make it the responsibility of the commenter to actually raise the bug themselves

[02:36:03.0000] <MikeSmith>
now that I say that, that's the only way the really makes sense anyway

[02:36:04.0000] <MikeSmith>
raising bugs on behalf of somebody else is not a good idea

[02:38:00.0000] <hsivonen>
will public-html-comments be closed down with all comments directed to bugzilla?

[02:46:00.0000] <annevk>
what does tl;dr mean in email context?

[02:46:01.0000] <annevk>
is it like ignored and deleted or some such?

[02:47:00.0000] <annevk>
(see hybi for an example)

[02:47:01.0000] <othermaciej>
in general it means "too long; didn't read", but responding to an email on a standards list that way is kind of ridiculous

[02:47:02.0000] <annevk>
aah, what an asshole

[02:49:00.0000] <MikeSmith>
hmm, http://www.w3.org/XML/2010/04/xml-model/ .. "This document allows schemas using any schema definition language to be associated with an XML document by including one or more processing instructions with a target of xml-model in the document's prolog"

[02:51:00.0000] <MikeSmith>
annevk: careful, that dude was 'nominated one of the "50 most influential people in IP" by Managing Intellectual Property magazine'

[02:51:01.0000] <MikeSmith>
so you better give him some due respect

[02:52:00.0000] <MikeSmith>
I wonder if that means he was nominated, but lost

[02:53:00.0000] <annevk>
heh

[02:55:00.0000] <myakura>
/me waves to annevk. welcome back :)

[02:55:01.0000] <MikeSmith>
"This specification is complementary technology which can be used when it is necessary to store ad-hoc schema associations directly inside XML document."

[02:56:00.0000] <annevk>
thanks myakura! nice to be back :)

[02:56:01.0000] <zcorpan>
is it a good idea to subscribe to hybi?

[02:57:00.0000] <zcorpan>
/me subscribes

[02:58:00.0000] <MikeSmith>
zcorpan: you still in the XML Core WG

[02:58:01.0000] <MikeSmith>
?

[02:58:02.0000] <MikeSmith>
I think I'll file an editorial bug saying that they wrote "necessary" in the sentence above, when what they actually mean is "convenient"

[02:59:00.0000] <zcorpan>
MikeSmith: yes

[03:01:00.0000] <MikeSmith>
zcorpan: please beat some sense into those guys

[03:04:00.0000] <annevk>
zcorpan, no, but I guess you should

[03:05:00.0000] <hsivonen>
zcorpan: it may be a good idea only in the sense that subscribing to public-html is a good idea

[03:07:00.0000] <hsivonen>
I wonder what rock the hybi folks have been living under not to have seen code written very, very badly by Web authors

[03:08:00.0000] <othermaciej>
hsivonen: you don't have to look at content to serve it I guess...

[03:10:00.0000] <annevk>
http://twitter.com/benschwarz/statuses/12193163314 -- must have been inevitable that tweets like this would happen

[03:13:00.0000] <othermaciej>
annevk: I don't understand what he is responding to there

[03:14:00.0000] <othermaciej>
or what his question means...

[03:16:00.0000] <anarchos>
hi

[03:19:00.0000] <annevk>
othermaciej, yeah, not really sure either, but the implication seems to be that "those people" were also responsible for HTML4

[03:20:00.0000] <othermaciej>
I think he's claiming that the people on the WHATWG list prevented progress in HTML after  HTML4

[03:20:01.0000] <othermaciej>
which is a very interesting reading of history

[03:21:00.0000] <annevk>
right

[03:23:00.0000] <Philip`>
othermaciej: "maybe an on-demand version of this tool would be useful" "where it's easy for some trusted set of people to convert a public-html-comments email into a bug" - how would the tool be more useful than simply copying-and-pasting into a bug report?

[03:24:00.0000] <othermaciej>
Philip`: fewer mouse movements?

[10:10:00.0000] <zcorpan>
/me files bugs on html5test

[10:10:01.0000] <zcorpan>
were there other things wrong other than the codecs?

[10:13:00.0000] <paul_irish>
he's been really active in development of the new version.. lots of fixes and changes have already went in

[10:13:01.0000] <paul_irish>
even moved some of the sections so people wouldnt miscontrue that they were part of html5..

[10:20:00.0000] <zcorpan>
hmm i should have looked at trunk source instead of deployed source

[10:24:00.0000] <paul_irish>
zcorpan: and here's his thoughts on including codecs.. http://rakaz.nl/2010/03/the-html5-test.html#comment-1603

[10:47:00.0000] <zcorpan>
"IE9 video tag should follow standard object fallback principals if this video is any indication:

[10:47:01.0000] <zcorpan>
http://live.visitmix.com/MIX10/Sessions/CL27" -- http://www.cssquirrel.com/2010/03/22/comic-update-ie-nine-means-business/#comment-31899

[10:48:00.0000] <zcorpan>
srsly? if that's true, microsoft are clearly incapable of implementing the spec

[11:12:00.0000] <zcorpan>
"if the video element fails for any reason, it'll fallback to teh contents inside" he says in the video

[11:12:01.0000] <zcorpan>
hrmm

[11:13:00.0000] <TabAtkins>
/me refuses to watch videos of talks, but will ping a friend over at MS for confirmation about what's happening there.

[11:27:00.0000] <AryehGregor>
I also hate videos of talks.

[11:27:01.0000] <AryehGregor>
Way too hard to skim.

[11:27:02.0000] <AryehGregor>
And audio.

[11:28:00.0000] <TabAtkins>
Yus.  Both are my primary difficulties with the format.  Gimme a transcript, much easier to use.

[11:28:01.0000] <TabAtkins>
Score another point for a11y's unintended positive effects!

[11:29:00.0000] <AryehGregor>
/me reflects on Chrome's decision to strip "http://" from the beginning of URLs in UI

[11:29:01.0000] <AryehGregor>
Seems to make sense.

[11:29:02.0000] <AryehGregor>
I don't really notice it.

[11:30:00.0000] <TabAtkins>
That hasn't made it to stable yet.  I suppose I should follow dev channel.  What's the easiest way to put myself on such?

[11:30:01.0000] <AryehGregor>
Google "Chrome dev channel" and follow the links?

[11:30:02.0000] <AryehGregor>
I've actually found it kind of annoying.

[11:31:00.0000] <AryehGregor>
It's a little too crashy/unstable.

[11:31:01.0000] <AryehGregor>
I'd switch to the beta channel if I weren't too lazy.

[11:31:02.0000] <TabAtkins>
kk

[11:31:03.0000] <AryehGregor>
You can switch to it easily, but switching back is trickier, I think.

[11:31:04.0000] <AryehGregor>
Not sure how well it works.

[11:32:00.0000] <TabAtkins>
Ah, nm, doesn't look like they have beta releases for linux?

[11:32:01.0000] <TabAtkins>
I guess I can run it on my W7 box.

[11:32:02.0000] <AryehGregor>
They only have beta and dev for Linux, last I checked.

[11:32:03.0000] <AryehGregor>
No stable.

[11:33:00.0000] <AryehGregor>
(beta != dev, of course)

[11:33:01.0000] <TabAtkins>
Yeah, dev is weekly pushes, beta is monthly, stable is quarterly.

[11:34:00.0000] <AryehGregor>
Something like that.

[11:34:01.0000] <AryehGregor>
I don't think that's precise, though.

[11:34:02.0000] <TabAtkins>
That's what I heard yesterday when one of our engineers asked what the update schedule was, at least. 

[11:35:00.0000] <AryehGregor>
Why do Google people call programmers "engineers"?

[11:35:01.0000] <TabAtkins>
Dunno.

[11:35:02.0000] <AryehGregor>
Everyone at Google does it consistently, but like no one else does.

[11:36:00.0000] <TabAtkins>
That's just our term for some reason.  We're all hired as "software engineers".

[11:36:01.0000] <AryehGregor>
Well, that's a very common term in bureaucrat-speak, but in most places normal people don't use it.

[11:36:02.0000] <TabAtkins>
Shrug.

[11:36:03.0000] <AryehGregor>
It sounds vaguely grandiose for some reason.  Not sure why, since it's not like actual engineering is particularly harder or more prestigious or better-paying than software development.

[11:37:00.0000] <AryehGregor>
Maybe because it implies a kind of exactness and certainty that doesn't really exist in software.

[11:37:01.0000] <TabAtkins>
I suppose it was considered the most appropriate term that still covered all the different things that we hire people for, not all of which is programming.

[11:37:02.0000] <TabAtkins>
(Excluding non-technical things like advertising or sales or HR, obviously.)

[11:38:00.0000] <AryehGregor>
Maybe that's a flaw in the entire concept of job titles, then.  I once heard (never checked) that the credits for Valve games mostly don't have titles.  Just a long undifferentiated list of people.

[11:39:00.0000] <AryehGregor>
Because apparently they each do whatever they're interested in and good at doing.

[11:39:01.0000] <AryehGregor>
Or something.

[11:39:02.0000] <TabAtkins>
We don't really *have* job titles, except as a very generic slotting mechanism, like I said.  Every single technical hire at Google is a "software engineer".  Nothing more, nothing less.

[11:39:03.0000] <AryehGregor>
Interesting.

[11:40:00.0000] <AryehGregor>
http://www.mobygames.com/game/windows/half-life-2/credits

[11:40:01.0000] <hsivonen>
AryehGregor: I thhink it's standard practice elsewhere, too, not just at Google to treat "programmer" as a dirty word and call people software developers or software engineers

[11:40:02.0000] <TabAtkins>
Then we just get picked up by a particular team, and identify ourselves that way.  We're encourage to move around between teams, though.

[11:40:03.0000] <AryehGregor>
Undifferentiated except for voices, face usage, and legal.

[11:40:04.0000] <AryehGregor>
hsivonen, "developers" I'm used to, "engineers" strikes me as a bit weird.

[11:41:00.0000] <AryehGregor>
Anyway, whatever.

[11:41:01.0000] <TabAtkins>
Developers is somewhat inaccurate for some people, though!  I'm not a developer right now, frex.

[11:41:02.0000] <TabAtkins>
Whereas we're all engineering software to some degree.

[11:41:03.0000] <AryehGregor>
. . . kind of.

[11:42:00.0000] <AryehGregor>
/me votes that everyone's job title should be "Employee".

[11:42:01.0000] <TabAtkins>
Can I change my name to my SSN, too?

[11:42:02.0000] <mpt>
"SharedWorker"

[11:42:03.0000] <miketaylr>
i wanna be Señor Employee

[11:43:00.0000] <TabAtkins>
As we live in Amurrica, I will make it a point to call you *Mr.* Employee, miketaylr.

[11:43:01.0000] <AryehGregor>
TabAtkins, I'm pretty sure a judge wouldn't approve that.

[11:43:02.0000] <AryehGregor>
(the name change)

[11:44:00.0000] <TabAtkins>
/me has always lived in primarily Hispanic areas anyway.

[11:44:01.0000] <miketaylr>
same.

[11:45:00.0000] <AryehGregor>
Me too, I guess.

[11:45:01.0000] <AryehGregor>
Although I don't think of it that way, since among neighborhood people I practically only interact with the Jews.

[11:45:02.0000] <AryehGregor>
Although that's because I don't leave the house except when I have to go pray, to be fair.

[11:45:03.0000] <miketaylr>
hmm, now i'm in an predominantly arab, irish, and norwegian 'hood tho

[11:45:04.0000] <AryehGregor>
If it weren't for that, I'd spend weeks at a time without leaving my house, at least when I don't have school.

[11:46:00.0000] <TabAtkins>
Mountain View has a lot more asian varieties mixed in, but I'm still in a Hispanic area of town.  Which a huge plus, because good hole-in-the-wall taquerias OMNOMNOMNOM

[11:47:00.0000] <miketaylr>
mmmm

[11:47:01.0000] <miketaylr>
that's #1 on my list when i head out for the SF jQuery conf.

[11:47:02.0000] <miketaylr>
burros

[11:51:00.0000] <TabAtkins>
When's the jQuery conf?

[11:51:01.0000] <miketaylr>
next weekend the 24th 25th

[11:52:00.0000] <miketaylr>
feel free to insert between 1 and 3 commas in that

[11:52:01.0000] <TabAtkins>
I choose a comma and "and"

[11:52:02.0000] <TabAtkins>
Any of you guys doing a dinner some night?

[11:53:00.0000] <miketaylr>
paul_irish would be the guy to ask, being the insider

[11:57:00.0000] <miketaylr>
they might be doing dinner onsite at the conf? dunno

[11:58:00.0000] <TabAtkins>
Yeah, I'll wait for him to come around and ask then.

[11:59:00.0000] <hober>
/me will be at the SFBA jQuery conf too

[12:00:00.0000] <miketaylr>
word up

[12:00:01.0000] <miketaylr>
i should like, write my talk

[12:01:00.0000] <paul_irish>
TabAtkins: sunday night we'll probably pull some folks together. i'll keep you in the loop (& hober too)

[12:01:01.0000] <TabAtkins>
Awesome.  I'd like to meet you guys ftf.

[12:02:00.0000] <AryehGregor>
Whoa: http://to./

[12:02:01.0000] <miketaylr>
totally

[12:02:02.0000] <AryehGregor>
A TLD URL shortener.

[12:02:03.0000] <AryehGregor>
You can only beat that if someone convinces ICANN to approve a one-letter TLD.  :P

[12:02:04.0000] <AryehGregor>
Good PR on Tonga's part?

[12:03:00.0000] <TabAtkins>
I'm not sure how that works.  It looks like the tld is the empty string.

[12:03:01.0000] <TabAtkins>
And the server name is just "to"

[12:03:02.0000] <AryehGregor>
Trailing dots in domain names are usually ignored.

[12:03:03.0000] <AryehGregor>
I'm not sure why it's required here.

[12:04:00.0000] <AryehGregor>
Observe: http://google.com./

[12:04:01.0000] <AryehGregor>
The trailing dot is always present in the actual DNS request, I think.  If you omit the dot, then the domain name is relative or something.

[12:04:02.0000] <AryehGregor>
I dunno, don't remember the details.

[12:04:03.0000] <AryehGregor>
Anyway, that's the .to TLD.

[12:04:04.0000] <TabAtkins>
Ok.

[12:06:00.0000] <hsivonen>
running a TLD is expensive. is there money in shorteners?

[12:06:01.0000] <jlebar>
Does HTML5 specify that we should ignore stylesheet <link>s which don't appear in the document's <head>?  I can't find language one way or another.

[12:06:02.0000] <paul_irish>
twitter doesnt autolink the to./ shortened url's sadly. otherwise its the shortest shortener (though very close to  tinyarro.ws)

[12:06:03.0000] <TabAtkins>
Depends.  If you think people will abide by you showing a page while redirecting, you can do ads.  Otherwise, you can sell usage info.

[12:07:00.0000] <othermaciej>
jlebar: I don't believe it requires ignoring them - I think it requires implementations to respect them

[12:07:01.0000] <othermaciej>
jlebar: however, it's nonconforming for content to have a stylesheet <link> outside the <head> section

[12:08:00.0000] <jlebar>
othermaciej, Thanks.

[12:18:00.0000] <hober>
paul_irish, miketaylr, TabAtkins: sounds good to me.

[12:42:00.0000] <zcorpan>
TabAtkins: let me know if MS says anything about video

[13:01:00.0000] <Hixie>
hober: yt?

[13:25:00.0000] <Hixie>
i find it amusing that some of the people who are most adament that we shouldn't violate the Atom spec's authoring requirements are the same people who argued that one option for HTML was to drop all authoring requirements

[13:34:00.0000] <othermaciej>
isn't that basically just Sam? (I don't think Julian ever suggested "drop all authoring requirements")

[14:52:00.0000] <TabAtkins>
/me keeps getting secret invites to beta tests for font things for some reason.

[14:55:00.0000] <TabAtkins>
Slightly related: FF's 3.6's font rendering on linux is occasionally extremely ugly.  It's like, every once in a while, for no reason, they lose the ability to do sub-pixel stuff.

[14:56:00.0000] <TabAtkins>
So letters either gain odd extra-pixel spaces between them, or randomly switch between having 1 or 2px wide verticals, etc.

[15:01:00.0000] <paul_irish>
TabAtkins: ah you got that too? i would talk to you about it, but a checkbox i click seems to indicate that would be a faux pas..

[15:01:01.0000] <paul_irish>
i'd dying for directwrite support to land in FF

[15:06:00.0000] <Hixie>
i give up on this atom thread. Neither Julian nor Sam are listening to a word I'm saying, both just keep repeating the same argument over and over despite my having explained why both their arguments are inapplicable.

[15:08:00.0000] <TabAtkins>
I'm not quite sure about Julian (he posts too much in that thread, so I've just sort of skimmed his arguments) but I think Sam is listening to you, but disagreeing in a different direction.

[15:09:00.0000] <TabAtkins>
I think you can interpret his words as agreeing with you on "there's a bug in the Atom spec".  He'd probably say differently, but I believe the net effect of what he's saying is the same.

[15:17:00.0000] <hober>
Hixie: here

[15:24:00.0000] <Hixie>
hober: i replied to your mail

[15:24:01.0000] <Hixie>
TabAtkins: he keeps asking me to "test to see what happens when the IDs change" despite the fact that I'm not disagreeing with him on what happens if hte IDs change

[15:28:00.0000] <hober>
I think all Sam's trying to say with that is "feeds with shitty entry IDs and/or updated datetimes suck." Which, obviously, everyone knows so I don't know why he's banging on about that bit

[15:37:00.0000] <TabAtkins>
He's banging on about it because the HTML5 Atome extraction algo will, I think, generate shitty entry IDs (or no entry IDs?  Similar problems exist.).

[15:40:00.0000] <othermaciej>
HTML5 doesn't require you to preserve the entry ID if you generate the feed multiple times

[15:41:00.0000] <othermaciej>
(or rather, it is a SHOULD rather than a MUST)

[15:41:01.0000] <othermaciej>
but it's not clear how much of that requirement is needed to avoid aggregators breaking

[15:41:02.0000] <Hixie>
we could change it to "MUST if possible" i guess

[15:41:03.0000] <TabAtkins>
Yeah, I see it now.

[15:41:04.0000] <othermaciej>
I suspect just requiring that the identical document sent through the same converter to generate identical atom:ids is both feasible and would mostly solve that particular problem

[15:41:05.0000] <TabAtkins>
Nod to othermaciej.

[15:42:00.0000] <othermaciej>
"MUST if possible" means "SHOULD"!

[15:42:01.0000] <Hixie>
i know

[15:42:02.0000] <Hixie>
that's why i wrote SHOULD

[15:42:03.0000] <Hixie>
but apparently that's not good enough for julian and sam

[15:42:04.0000] <hober>
Hixie: use "OUGHT TO" ( ref: http://edward.oconnor.cx/2006/02/more-rfc-2119-humor ) 

[15:42:05.0000] <othermaciej>
I don't really understand the scope of the Atom spec's MUST

[15:42:06.0000] <TabAtkins>
It seems pretty easy to figure out how to generate a guurl from the article address and content.

[15:42:07.0000] <othermaciej>
when Julian has explained what he thinks it does and doesn't require, it sounds like he's making stuff up to suit his preferences rather than interpreting the Atom spec

[15:42:08.0000] <TabAtkins>
It is somewhat more difficult to generate one if the article changes between runs.

[15:43:00.0000] <TabAtkins>
Which I think they raised as a problem with it.

[15:43:01.0000] <othermaciej>
yeah, the only truly sound way to deal with "article changes between runs" is to require the HTML document to have unique IDs embedded

[15:43:02.0000] <hober>
othermaciej: FWIW, I thought your point re: "why the double standard between text->atom and html->atom" was spot-on

[15:44:00.0000] <othermaciej>
hober: I still don't feel like he gave a straight answer

[15:44:01.0000] <TabAtkins>
othermaciej: Right.  So I think I'm fine with that.  Changing the SHOULD to a MUST would still allow the ids to be different if the content changes between runs (it wont' be the "same input").

[15:44:02.0000] <othermaciej>
but I also don't understand the double standard between "two runs of same converter" and "runs of different converters"

[15:45:00.0000] <othermaciej>
TabAtkins: it's not clear to me if Julian or Sam would be fine with that, or whether they would see that as violating the Atom spec (a possibly separate question)

[15:45:01.0000] <TabAtkins>
othermaciej: That one, at least, is similar to the distinction between "run on the same input" versus "run on similar input".

[15:45:02.0000] <othermaciej>
the problem is that the Atom requirement just doesn't make sense when starting from non-Atom input

[15:45:03.0000] <TabAtkins>
Right.

[15:45:04.0000] <othermaciej>
it's totally plausible to apply to systems that are already operating in Atom

[15:45:05.0000] <hober>
exactly

[15:45:06.0000] <TabAtkins>
Unless the content happens to also embed an appropriate guid.

[15:45:07.0000] <othermaciej>
but it's unimplementable if your input doesn't already have globally unique IDs embedded

[15:46:00.0000] <Hixie>
yup

[15:46:01.0000] <Hixie>
that's exactly the line of thought i went through when i wrote this stuff

[15:46:02.0000] <Hixie>
hence the should

[15:46:03.0000] <TabAtkins>
So, yeah, bug in the Atom spec.  Change our spec to a MUST, and possibly explicitly call out the failing.

[15:46:04.0000] <othermaciej>
it *is* implementable for the special case of "same converter and identical content"

[15:47:00.0000] <TabAtkins>
It won't be a violation of our MUST if UAs generate new ids for altered content.

[15:47:01.0000] <Hixie>
one thing that's worth considering is whether we are interested in having a solution that outputs almost-valid atom that can be hand fixed to valid atom, or if we think the output must be valid atom

[15:48:00.0000] <TabAtkins>
The latter.  Not much point in an automatic extraction algorithm if you still have to hand-edit the result.

[15:48:01.0000] <TabAtkins>
Since the producer won't be the one doing the extracting, the consumer will.

[15:50:00.0000] <hober>
Hixie: my first choice is that the algorithm do what it's doing now. html -> atom tools can (and should) help their users improve their HTML so that the results are valid

[15:50:01.0000] <hober>
my second choice, if the output must be valid Atom, is to drop entries that we can't come up with IDs from. that way the entries that do end up in the feed don't annoy the crap out of subscribers

[15:51:00.0000] <TabAtkins>
hober: For your first choice, does that resolve to the former or latter option in hixie's question?

[15:51:01.0000] <TabAtkins>
It sounds like the latter, unless you're seriously suggesting that feed readers should ask their users to correct the HTML of feed producers.

[15:51:02.0000] <hober>
TabAtkins: the former, though I wouldn't suggest that people hand-edit the atom

[15:52:00.0000] <TabAtkins>
Who is supposed to edit the Atom, then?  At what point in the chain does it go from almost-valid to valid?

[15:52:01.0000] <hober>
I don't expect anybody to edit the Atom.

[15:53:00.0000] <hober>
I've got a site, and I decide to use an html2atom tool to make a feed from it

[15:53:01.0000] <hober>
I go to the tool, paste in my URL, and the result page is something like

[15:53:02.0000] <TabAtkins>
Then it never becomes valid, and you *can't* be agreeing with Hixie's former option.

[15:53:03.0000] <TabAtkins>
(Because that option, be definition, requires someone to edit the Atom at some point to make it valid.)

[15:53:04.0000] <hober>
"here's a url to an atom feed for your page. it doesn't look like it's valid; here are some ways you could improve your site so that your feed is more useful. have a nice day"

[15:54:00.0000] <hober>
or "here's a url... it's valid, congrats! here's a pat on the head"

[15:54:01.0000] <TabAtkins>
That's not the same use-case as "consume an HTML page as Atom".  That's an HTML->Atom validator.

[15:55:00.0000] <hober>
no, it's a converter, that's providing lint-tool-esque advice when used

[15:55:01.0000] <hober>
the feed consumers just get the url to the (possibly invalid) atom feed

[15:55:02.0000] <TabAtkins>
It's allowed to flag things that aren't right.  An actual conversion algo, though, either needs to output something valid, or we need to define exactly who turns it valid.

[15:55:03.0000] <hober>
it's the person feeding the site itno the converter that sees the lint/validation advice

[15:56:00.0000] <TabAtkins>
Okay, such a tool could work equally by only producing a valid feed, but also flagging things that it's skipped for invalidity.

[15:56:01.0000] <hober>
sure

[15:56:02.0000] <hober>
hence my compromise proposal to julian

[15:56:03.0000] <TabAtkins>
Then let's make sure that the algo just produces a valid feed.

[15:57:00.0000] <TabAtkins>
And let the market help people make their feeds better.

[15:57:01.0000] <hober>
sounds good to me

[16:12:00.0000] <JonathanNeal>
For any folks working on Microdata in here, it would be nice if I could put itemprop="fn org logo" on an <img /> with an alt.

[16:13:00.0000] <TabAtkins>
Hixie: ^^^ There really should be a way to get @alt in the property value.  Automatically, preferably.

[16:14:00.0000] <TabAtkins>
JonathanNeal: You wouldn't be able to put it on the <img> directly, anyway - that just grabs the @src instead.  But you *should* be able to put in on a <span> or <h1> or whatnot wrapping the <img>.

[16:14:01.0000] <JonathanNeal>
Yea that works too :)

[16:15:00.0000] <othermaciej>
TabAtkins: seems lame to add a span just to emit another property

[16:15:01.0000] <JonathanNeal>
I think in microformats you can place it on the <img /> itself.

[16:15:02.0000] <TabAtkins>
othermaciej: Shrug.  Grabbing the @src of an <img> is useful.

[16:15:03.0000] <JonathanNeal>
/me pokes you for keeping up with the Jones'.

[16:16:00.0000] <othermaciej>
TabAtkins: I agree - and as just mentioned, grabbing the alt is also useful

[16:16:01.0000] <TabAtkins>
And so it should continue to be easy to do so.  Using a wrapper is the obvious way to grab the alt without interfering with that.

[16:16:02.0000] <othermaciej>
I think it would be a design flaw if you had to add a superfluous DOM element just to emit both

[16:16:03.0000] <othermaciej>
although I do not have another design to propose

[16:16:04.0000] <TabAtkins>
You have to add wrappers for a lot of things in Microdata, if your data is part of prose.

[16:17:00.0000] <TabAtkins>
Though, with JonathanNeal's use-case, the <img> will commonly already have a wrapper <h1> to use.

[16:18:00.0000] <TabAtkins>
You either have to add wrapper, or add another level of indirection to specify where the property value comes from.  The former seems a lot better to me, in this context.\

[16:18:01.0000] <TabAtkins>
(If you asked me the same question in a CSS context, I'd say the opposite, because we love indirection.)

[16:23:00.0000] <othermaciej>
CSS WG must have a bunch of computer scientists in it

[16:23:01.0000] <TabAtkins>
Yes.

[16:25:00.0000] <Hixie>
when designing microdata i thought long and hard about how to get <img> to output one property for src="" and one for alt="" and i ended up deciding that was the mistake RDFa had made

[16:25:01.0000] <Hixie>
well, one of several, i guess

[16:27:00.0000] <TabAtkins>
Yeah, I don't think it should do it magically.  But I think that <span itemprop=foo><img itemprop=bar src=baz alt=qux></span> should output foo:"qux", bar:"baz".

[16:27:01.0000] <TabAtkins>
Agreed that trying to stack the two together is a mistake in RDFa.

[16:28:00.0000] <TabAtkins>
Essentially just, whenever a property is using the text content of an element for its value, <img> should substitute in its @alt value.  That's the whole point of @alt, after all - to be a textual equivalent of the image.

[16:31:00.0000] <othermaciej>
TabAtkins: in current Microdata, what would <span itemprop=foo><img src=baz alt=qux></span> emit?

[16:31:01.0000] <othermaciej>
(note lack of itemprop on the img itself)

[16:32:00.0000] <TabAtkins>
foo:""

[16:33:00.0000] <TabAtkins>
Because @alt isn't part of the element's textContent.

[16:34:00.0000] <othermaciej>
would you propose changing what the markup I cited emits, or only in the case where the img element has an itemprop?

[16:34:01.0000] <othermaciej>
and what if there is more than just the <img> element inside the span?

[16:35:00.0000] <othermaciej>
additional side note: is there a way to get the contents of a "title" attribute into a Microdata property?

[16:35:01.0000] <TabAtkins>
I would propose changing what the markup you cited emits.  It should emit foo:"qux".

[16:35:02.0000] <othermaciej>
those are the two cases I can think of where useful text is likely to be an attribute value instead of text content

[16:36:00.0000] <TabAtkins>
In the case of <span itemprop=foo>foo <img alt=bar> baz</span>, it should emit foo:"foo bar baz".

[16:36:01.0000] <othermaciej>
TabAtkins: that is a self-consistent model

[16:36:02.0000] <TabAtkins>
There is no way to get @title into a Microdata value.

[16:36:03.0000] <othermaciej>
although that solution doesn't scale to @title

[16:37:00.0000] <Hixie>
TabAtkins: yeah i've long been a fan of inventing some property that serialises HTML to text "correctly", e.g. turning <bdo> into unicode bidi formatting characters, inserting quotes for <q>, inserting newlines for <br>, etc. People haven't really been enthusiastic about it.

[16:37:01.0000] <TabAtkins>
@title isn't a textual equivalent of the element, so it shouldn't substitute itself normally.

[16:37:02.0000] <othermaciej>
Hixie: WebKit tries to do that for innerText

[16:37:03.0000] <Hixie>
TabAtkins: in practice btw there are actually few cases where you want the alt="" -- mostly you're grabbing the URL for purposes other than showing it in a context where you need the alt, e.g. you're grabbing it to label that image with metadata like a license

[16:37:04.0000] <Hixie>
othermaciej: really?

[16:38:00.0000] <TabAtkins>
And, unfortunately, it doesn't share the distinction of @src, @href, etc of being a special, use-restricted attribute so that you can usefully say "whenever @itemprop is <foo>, use @bar for the value".

[16:38:01.0000] <othermaciej>
TabAtkins: right - it's clearly ancillary information that shouldn't inject itself into the text content

[16:38:02.0000] <othermaciej>
Hixie: well at the very least we respect <br> and omit the text for some elements that are always unrendered

[16:38:03.0000] <othermaciej>
can't remember if we insert generated content

[16:38:04.0000] <othermaciej>
but innerText at least at one point did the same thing as Copy --> Past as Plain Text

[16:38:05.0000] <othermaciej>
*Paste

[16:39:00.0000] <Hixie>
interesting

[16:39:01.0000] <TabAtkins>
Hixie: I agree that the case of wanting *only* the @alt, as in <img itemprop=foo alt=bar>, are rare.  I don't care about that.  I would like that case to continue to emit the @src.

[16:39:02.0000] <Hixie>
TabAtkins: i'm not talking about wanting only the alt, i mean wanting both

[16:39:03.0000] <TabAtkins>
But the cases on pages that I write where I have <h1><img alt="My Company Name"></h1> are much more common.

[16:40:00.0000] <JonathanNeal>
Yes, that's common.

[16:40:01.0000] <TabAtkins>
I'm not sure what you're objecting to changing, Hixie?

[16:40:02.0000] <Hixie>
the algorithm to grab text for a property

[16:40:03.0000] <TabAtkins>
If you want both, you specify an @itemprop on both the <img> (for the @src) and a wrapper (for the @alt).

[16:41:00.0000] <Hixie>
just use <meta>

[16:41:01.0000] <TabAtkins>
That's ridiculous when the information is *right there*, and it's completely unambiguous which data you want.

[16:41:02.0000] <Hixie>
*shrug*

[16:42:00.0000] <Hixie>
it's a pretty rare case and would make the mechanism less intuitive

[16:42:01.0000] <TabAtkins>
It's silly that <span itemprop=foo>foo <img alt=bar> baz</span> doesn't emit foo:"foo bar baz", and the only way to make it do so is to duplicate the whole text value in the <meta>

[16:42:02.0000] <TabAtkins>
...no, it's much more intuitive.  @alt is the textual replacement for the image.  If you want the text of an element that contains an image, you'll get the textual replacement of it.

[16:43:00.0000] <TabAtkins>
This is basic a11y stuff, and is the whole reason that @alt exists.

[16:43:01.0000] <Hixie>
it's also "silly" that <span itemprop=foo> <q>test</q> <br> <bdo dir=rtl>abc</bdo> </span> doesn't emit quotes, a newline, and bidi formatting characters

[16:43:02.0000] <TabAtkins>
Agreed, to be honest.  Let's make all those changes.

[16:44:00.0000] <Hixie>
doing that leads to the implementation of microdata doubling or tripling in size

[16:44:01.0000] <Hixie>
it's not a good tradeoff

[16:45:00.0000] <Hixie>
and frankly it isn't necessary for most uses of microdata

[16:45:01.0000] <TabAtkins>
Let's make browsers include a property for emitting it.

[16:45:02.0000] <Hixie>
browsers aren't the main consumers

[16:47:00.0000] <TabAtkins>
Microdata currently defines that case as being the value of textContent.  This is a property exposed by whatever DOM implementation you're using.  We'd define a new property for DOM implementations to support that does "smart" text content.

[16:48:00.0000] <TabAtkins>
If you're not relying on an external parser to create and handle the DOM, then adding "smart" text content handling is *definitely* not a "doubling or tripling" of your code size.

[16:49:00.0000] <Hixie>
you'd be surprised

[16:49:01.0000] <Hixie>
textContent is implementable in about 3 lines

[16:49:02.0000] <Hixie>
what we're talking about here basically means implementing an HTML renderer for plain text

[16:49:03.0000] <Hixie>
that's 100+ lines

[16:49:04.0000] <Hixie>
maybe far more

[16:50:00.0000] <TabAtkins>
A "renderer" that handles, what, four cases total?  Plain text, <q> text, bidi text, and @alt.

[16:50:01.0000] <JonathanNeal>
All I want is alt :)

[16:50:02.0000] <TabAtkins>
If you're starting from a SAX parser or something, it's definitely not 100+ lines unless you really, really suck somehow.

[16:51:00.0000] <Hixie>
there's far more to HTML than just the cases I mentioned

[16:53:00.0000] <ment>
how to best benchmark how much time browser spends rematching css on dom tree when i change the dom tree using js?

[16:54:00.0000] <TabAtkins>
Let's see... I can see the possibility that you might want to do <sub> and <sup>, but I don't think there's an accept way to render them in plaintext.  You could *maybe* argue that <em> and <strong> deserve special formatting, but I can reasonably argue against it, I think.  I believe that's about it.

[16:55:00.0000] <TabAtkins>
Everything else is sufficient as plain text for any purpose that Microdata might need.

[16:55:01.0000] <Hixie>
i am highly skeptical

[16:55:02.0000] <Hixie>
and again, i really don't think the use cases need this

[16:55:03.0000] <TabAtkins>
I think that @alt is the highest-priority one.

[16:56:00.0000] <TabAtkins>
If you're rolling your own DOM, then just where you loop through the nodes and check if they're text nodes, also check if they're element nodes with the name "img".

[16:56:01.0000] <TabAtkins>
That is 2 lines (possibly more depending on your code formatting guidelines).

[16:57:00.0000] <TabAtkins>
You're already having to check if a node is an element node, to know if you have to recurse into it to search for more text nodes.

[16:58:00.0000] <JonathanNeal>
All I want is alt :)

[16:58:01.0000] <TabAtkins>
So I challenge your statement that this would greatly increase the implementation cost.  The increase is so minor as to be trivial.

[16:58:02.0000] <TabAtkins>
JonathanNeal: I'm fine with just @alt too.  ^_^

[16:59:00.0000] <TabAtkins>
At least with <q> and <bdo> the contents are *available*, even if formatted badly.  <img alt> is simply *omitted* as it stands.


2010-04-16
[17:07:00.0000] <TabAtkins>
/me will just go file a bug about ti.

[17:08:00.0000] <TabAtkins>
Do these bugs auto-cc the submitter, or do I need to manually put my email in the CC list?

[17:18:00.0000] <Hixie>
TabAtkins: the other elements that would need handling are all the flow ones, all the controls, dir="", lang="" (turning into unicode lang tags), things like <script> and <style> being hidden, <pre> not collapsing spaces but other things collapsing them, lists, <time>, it just gets more and more intractable

[17:19:00.0000] <TabAtkins>
Okay, then we skip that.  All of those still return a useful value for their textContent, even if it's somewhat misformatted.  <img> returns nothing, though.

[17:19:01.0000] <Hixie>
so don't use it

[17:19:02.0000] <Hixie>
it's really not a big problem

[17:20:00.0000] <Hixie>
i really don't see why <img> is important but, say, <br> is not

[17:21:00.0000] <TabAtkins>
For the reason I just stated.  Losing a linebreak makes the text somewhat misformatted, but the text is still *there* and can be read.

[17:23:00.0000] <Hixie>
<bdo> then

[17:23:01.0000] <TabAtkins>
Again, misformatted, but still present.

[17:24:00.0000] <TabAtkins>
A human can parse it, even if it's somewhat confusing.

[17:24:01.0000] <Hixie>
O_o neht ,dettamrof yldab tsuj si siht kniht uoy fi

[17:24:02.0000] <TabAtkins>
I can read it!

[17:25:00.0000] <Hixie>
the model works fine as is, uing textContent -- it's trivial to understand, doesn't need any special cases, and is easy to predict

[17:25:01.0000] <TabAtkins>
And it would only come out like that if a parent element's text contained a bidi override character.

[17:25:02.0000] <hober>
M-x reverse-region is your friend :)

[17:25:03.0000] <Hixie>
as soon as we start adding magic, things start going down the path to RDFa

[17:25:04.0000] <JonathanNeal>
): tla si tnaw I llA

[17:25:05.0000] <Hixie>
what's the use case for this anyway? what are you using microdata for for which you need alt?

[17:26:00.0000] <TabAtkins>
Dude, there's no magic.  You want text, you get text, including the text equivalent of things that have text equivalents.

[17:26:01.0000] <TabAtkins>
Sigh.  <h1 itemprop="fn org"><img src=foo alt="Example Corp"></h1>

[17:27:00.0000] <Hixie>
having to check each element's namespace and localname to see if it's an <img> from which to grab alt="" is magic.

[17:27:01.0000] <Hixie>
why would you use an image there?

[17:27:02.0000] <Hixie>
just use <h1>Example Corp</h1>

[17:27:03.0000] <TabAtkins>
Because that's extremely common.

[17:28:00.0000] <TabAtkins>
That way you can have a fancy image logo, while still having proper text for Google and screen readers.

[17:28:01.0000] <Hixie>
and you're putting a vcard on it? o_O

[17:28:02.0000] <TabAtkins>
Is there something wrong with that?

[17:28:03.0000] <Hixie>
i fail to see any point

[17:28:04.0000] <TabAtkins>
A Contact Us page is a fine place to put a vcard.

[17:29:00.0000] <TabAtkins>
Or, perhaps you put the business's contact information in a sidebar on each page (my old company does precisely this).

[17:29:01.0000] <Hixie>
sure, just stick it in a data: URL so I can actually use it, rather than requiring that I extract it from microdata :-P

[17:29:02.0000] <TabAtkins>
Might as well mark it up properly.

[17:29:03.0000] <JonathanNeal>
<section><h1>Companies using us</h1><article><h1 itemprop="fn org"><img src="google.png" alt="Google, Inc." /></h1></article></section>

[17:29:04.0000] <Hixie>
btw since you're going to have to have the "n" property elsewhere anyway, why can't you just do "fn" and "org" there too?

[17:30:00.0000] <JonathanNeal>
And in that <article>, you could be describing Google in some way.

[17:30:01.0000] <Hixie>
this really seems like far too overenthusastic use of microdata

[17:30:02.0000] <TabAtkins>
Hixie: ?_?

[17:30:03.0000] <TabAtkins>
I don't understand what you mean by your previous statement.

[17:30:04.0000] <Hixie>
for "n" you have to use itemscope

[17:30:05.0000] <Hixie>
rigt?

[17:30:06.0000] <Hixie>
so why is the duplication there ok but meanwhile you're trying really hard to avoid it here?

[17:31:00.0000] <TabAtkins>
Why do you need to use "n"?  This is a company name.  The correct idiom is to apply both "fn" and "org" to it.

[17:31:01.0000] <TabAtkins>
No "n" required.

[17:31:02.0000] <Hixie>
vcard requires "n" for all vcards

[17:31:03.0000] <TabAtkins>
Yes, and the microformat vocabulary allows it to be omitted and extracted from the fn, if present.

[17:31:04.0000] <Hixie>
the microdata vcard vocab has none of that magic

[17:31:05.0000] <TabAtkins>
If the Microdata embedding of such doesn't allow that, it's a bug in Microdata.

[17:33:00.0000] <Hixie>
speak to foolip, he argued for the magic to be removed since it basically doesn't work and isn't specified anywhere

[17:33:01.0000] <TabAtkins>
It's specified on the Microformats wiki!

[17:34:00.0000] <Hixie>
"specified" isn't the word i would use

[17:34:01.0000] <TabAtkins>
If I can believe the wiki, you don't need any of the implied fn magic.  If you're doing an org name, the org name should be specified as the value of both "fn" and "org", and *no* value shoudl be specified for the "n" properties (they should all be blank).

[17:34:02.0000] <Hixie>
for example, if n="TAMURA Kent", what is the vcard output?

[17:35:00.0000] <Hixie>
yeah for org="" the magic isn't quite so bad

[17:35:01.0000] <TabAtkins>
If n equals that?  That's wrong.  We're talking fn.

[17:35:02.0000] <Hixie>
er

[17:35:03.0000] <Hixie>
if fn="TAMURA Kent"

[17:35:04.0000] <Hixie>
what is the vcard output for "n"

[17:35:05.0000] <Hixie>
is "Kent" the given name or the surname?

[17:36:00.0000] <TabAtkins>
Then it gets split into given-name="TAMURA" and family-name="Kent".

[17:36:01.0000] <Hixie>
which is wrong

[17:37:00.0000] <TabAtkins>
Sure.  But it's right for most of the world.  And if your particular name-formatting practices dont' match that, just specify a full n value.

[17:37:01.0000] <Hixie>
uh, it's wrong for the majority of the population of earth

[17:37:02.0000] <Hixie>
it's right for the US and some parts of europe

[17:37:03.0000] <TabAtkins>
This makes the majority of hcards in the world much less verbose.

[17:37:04.0000] <TabAtkins>
It's wrong for East Asia.  Where else?

[17:38:00.0000] <Hixie>
a number of parts of europe

[17:38:01.0000] <TabAtkins>
Details?  I've never heard of such a thing.

[17:38:02.0000] <TabAtkins>
/me may simply be ignorant.

[17:39:00.0000] <Hixie>
i don't know the details, unfortunately

[17:39:01.0000] <Hixie>
foolip is the man to ask

[17:39:02.0000] <Hixie>
/me is not a fan of any of these data systems

[17:39:03.0000] <TabAtkins>
Will do.  In the meantime, we're talking about "fn org", which is a completely different matter.

[17:40:00.0000] <Hixie>
i really don't understand why anyone would ever care to get a vcard from a web page, personally

[17:40:01.0000] <TabAtkins>
It's not important that you do.  It's just important for you to understand that hCard exists, and is popular, for precisely that reason.

[17:40:02.0000] <Hixie>
hcard is hardly that popular

[17:40:03.0000] <TabAtkins>
Go ask Tantek for usage statistics.

[17:41:00.0000] <TabAtkins>
If we're talking about marking up metadata, it and licensing information are, I think, the most popular use-cases.

[17:41:01.0000] <Hixie>
you have been drinking too much kool-aid :-)

[17:41:02.0000] <Hixie>
licensing information i can sort-of buy

[17:42:00.0000] <Hixie>
that one definitely doesn't need alt="" though

[17:42:01.0000] <TabAtkins>
Sure.

[17:55:00.0000] <TabAtkins>
Anyway, submitted bug 9533 about it.

[17:57:00.0000] <Hixie>
did you mean to put it in "future versions"?

[17:58:00.0000] <TabAtkins>
Microdata is no longer part of HTML5, right?  Or did I just skip seeing the Microdata entry?

[17:58:01.0000] <Hixie>
oh i just treat it as part of html5

[17:58:02.0000] <Hixie>
"future versions" doesn't appear on my radar

[17:58:03.0000] <TabAtkins>
All right, that's cool.

[17:58:04.0000] <TabAtkins>
So, change it to "spec bugs"?

[17:59:00.0000] <Hixie>
i can do it

[17:59:01.0000] <TabAtkins>
kk

[18:20:00.0000] <MikeSmith>
so Chrome supports the File API now, I gues?

[18:20:01.0000] <TabAtkins>
If you were in the office, you could ask.

[18:20:02.0000] <TabAtkins>
/me doesn't know.

[18:21:00.0000] <MikeSmith>
reading http://gmailblog.blogspot.com/2010/04/drag-and-drop-attachments-onto-messages.html

[18:21:01.0000] <MikeSmith>
TabAtkins: If I were in the office?

[18:21:02.0000] <MikeSmith>
you mean if I was the other Mike Smith?

[18:21:03.0000] <TabAtkins>
... Sorry, I didn't look too closely and thought you were Hixie.

[18:22:00.0000] <MikeSmith>
me and Hixie are like twins

[18:22:01.0000] <MikeSmith>
or like Wolverine and Sabretooth

[18:22:02.0000] <TabAtkins>
It's true.  That's why you have to wear a hat at standards meetings.

[18:22:03.0000] <Hixie>
hah

[18:22:04.0000] <MikeSmith>
heh

[18:23:00.0000] <MikeSmith>
TabAtkins: one of the Chrome product managers is named Mike Smith

[18:23:01.0000] <MikeSmith>
in case you didn't know

[18:23:02.0000] <TabAtkins>
I'd given half-decent odds that there's a Mike Smith working for every browser company.

[18:23:03.0000] <TabAtkins>
Maybe not Opera, but who knows?

[18:23:04.0000] <MikeSmith>
yeh

[18:31:00.0000] <MikeSmith>
uh, I just checked in Gmail with Minefield and Chrome dev-channel on OSX, and the drag-and-drop file attachment thing don't seem to work

[18:42:00.0000] <Hixie>
MikeSmith did work for opera for a bit iirc, so maybe that counts for them

[18:44:00.0000] <TabAtkins>
I'll take it.

[18:54:00.0000] <Hixie>
hm, the hardware limitations clause is an interesting point

[18:54:01.0000] <Hixie>
it doesn't quite handle this case

[18:55:00.0000] <Hixie>
but we could make it handle this case and then turn it into a MUST

[20:05:00.0000] <kennyluck>
Morning, MikeSmith.

[20:11:00.0000] <MikeSmith>
kennyluck: hey

[20:11:01.0000] <MikeSmith>
annevk: i'm in the breakroom area

[20:11:02.0000] <annevk>
kk

[20:26:00.0000] <MikeSmith>
does html5-based drag-and-drop in Gmail only work on Windows maybe?

[20:27:00.0000] <MikeSmith>
still not working for me at least in Chrome and Minefield on OSX

[20:31:00.0000] <TabAtkins__>
Man, what computer do I still have on that is using TabAtkins_?

[20:31:01.0000] <TabAtkins__>
/me is puzzled.

[20:32:00.0000] <TabAtkins__>
/me has shut down the one that was claiming TabAtkins__, allowing me to change from ___ to __.

[22:38:00.0000] <MikeSmith>
I really like this page design: http://www.kuroge-wagyu.com/bc/delivery.html

[22:38:01.0000] <MikeSmith>
all sites should be like that

[22:39:00.0000] <kennyluck>
orz

[22:39:01.0000] <kennyluck>
a picture, huh...

[22:45:00.0000] <MikeSmith>
the table use is great as well

[22:47:00.0000] <MikeSmith>
I would like to be the guy who got paid to make that site

[22:52:00.0000] <MikeSmith>
kennyluck: click in the part where it TOP DELIVERY TAKEOUT etc.

[22:54:00.0000] <othermaciej>
hey everybody

[22:55:00.0000] <MikeSmith>
greetings

[22:56:00.0000] <MikeSmith>
othermaciej: dunno if you saw this yet: http://gmailblog.blogspot.com/2010/04/drag-and-drop-attachments-onto-messages.html

[22:57:00.0000] <othermaciej>
I have

[22:57:01.0000] <othermaciej>
not sure why they didn't make it work in Safari (we support the relevant features afaik)

[23:00:00.0000] <MikeSmith>
as far as I can tell so far, it doesn't work on OSX at all

[23:00:01.0000] <MikeSmith>
unless I am missing something

[23:01:00.0000] <MikeSmith>
maybe the platform file-system interaction part is only implemented for Windows so far?

[23:28:00.0000] <annevk>
TabAtkins, e.g. how would "Anne van Kesteren" work with this splitting system?

[23:31:00.0000] <othermaciej>
MikeSmith: is it a plugin or something?

[23:31:01.0000] <othermaciej>
I thought it was using standard file drag & drop functionality

[23:32:00.0000] <MikeSmith>
yeah, I thought so too.. but the interaction between the browser and the OS file system is platform code, right?

[23:36:00.0000] <othermaciej>
yeah but file DnD in WebKit was implemented on Mac first

[23:57:00.0000] <annevk>
TabAtkins, also, DOM insertion is not random, it is defined by the HTML parser

[23:58:00.0000] <JonathanNeal>
heyo

[23:59:00.0000] <TabAtkins__>
annevk: It woulnd't.  It only tries to do any optimizations if you have a single value after splitting spaces (in which case it assumes it is a nickname) or if you have two values (in which case it assumes given, then family name).

[00:00:00.0000] <JonathanNeal>
TabAtkins__, did we get alt?

[00:01:00.0000] <TabAtkins__>
I have a bug.  It's rejected right now, but I'm trying to gather some support from one of our Google teams

[00:01:01.0000] <othermaciej>
which bug?

[00:04:00.0000] <TabAtkins__>
9355?  something like that.

[00:06:00.0000] <annevk>
TabAtkins__, ah okay, but that indeed fails

[00:07:00.0000] <TabAtkins__>
annevk: You have any idea of non-east asian areas where it's common to write a name with family-name first, given-name last?

[00:07:01.0000] <annevk>
Hungary according to Wikipedia

[00:08:00.0000] <TabAtkins__>
Interesting.

[00:08:01.0000] <othermaciej>
this is true

[00:08:02.0000] <annevk>
optimizing for common usage failed to work out with e.g. <meter> and <progress> so I doubt it will here

[00:12:00.0000] <annevk>
I should add perceived* to that

[00:14:00.0000] <TabAtkins__>
othermaciej: 9533 is my bug, actually.

[00:16:00.0000] <MikeSmith>
TabAtkins, I thought I had heard before that family-name first is used in Finland in at least some cases

[00:18:00.0000] <hsivonen>
MikeSmith: it's rare

[00:19:00.0000] <hsivonen>
MikeSmith: doesn't happen in speech

[00:19:01.0000] <MikeSmith>
ok

[00:19:02.0000] <hsivonen>
MikeSmith: I guess some people think it's more official or something

[00:19:03.0000] <hsivonen>
MikeSmith: I've seen it in email From fields

[00:19:04.0000] <hsivonen>
MikeSmith: but very, very rarely

[00:19:05.0000] <hsivonen>
MikeSmith: definitely not common enough to cry i18n for failure to support that oddity

[00:20:00.0000] <MikeSmith>
I see

[00:20:01.0000] <MikeSmith>
hsivonen: in mail From fields to people put a comma after the family-name part?

[00:20:02.0000] <MikeSmith>
because I do see that in business mail in English too sometimes

[00:21:00.0000] <hsivonen>
MikeSmith: I think I've seen it without the comma

[00:23:00.0000] <MikeSmith>
has Twitter said what format their Annotations mechanism uses?

[00:24:00.0000] <MikeSmith>
are their docs yet for the API?

[00:24:01.0000] <micheil>
MikeSmith: i think so

[00:25:00.0000] <micheil>
MikeSmith: actually not yet, but I may be able to find out off the guys who went to chirp.

[00:25:01.0000] <micheil>
MikeSmith: I'm pretty sure it'd just be a json type blob

[00:26:00.0000] <micheil>
as it is, to update a status, you have to POST, so to add arbitrary json data to the body of that request would be easiest

[00:26:01.0000] <MikeSmith>
ok

[00:27:00.0000] <JonathanNeal>
So ... we did yet alt?

[00:38:00.0000] <MikeSmith>
JonathanNeal: ?

[00:38:01.0000] <JonathanNeal>
Earlier there was a large discussion going on about microdata.

[00:39:00.0000] <JonathanNeal>
In particular, whether or not folks would modify the spec to allow for the alt="text" to be read, example @ http://sandbox.thewikies.com/html5-layout/

[00:43:00.0000] <MikeSmith>
JonathanNeal: is the alt value in that case a functional equivalent of the image or is it a label or something?

[00:44:00.0000] <JonathanNeal>
it is the textual representation of the image.

[00:51:00.0000] <JonathanNeal>
That answer was bunk?

[00:54:00.0000] <JonathanNeal>
TabAtkins had a better explanation than I

[00:55:00.0000] <hsivonen>
what's the recommended way of building a Theora movie in Ogg container without soundtrack from a sequence of PNG images?

[00:56:00.0000] <hsivonen>
googling shows results from 2006, so I expect the state of the art may have advanced

[00:57:00.0000] <hsivonen>
hmm. http://wiki.povray.org/content/HowTo:Encode_animations_as_Ogg_Theora_Video might still be the state of the art

[01:02:00.0000] <hsivonen>
There's still a couple of Cinepak videos on my site that I made when I was in high school. I should re-encode those as Theora, even though those videos are a lot less impressive today than they were back then

[01:03:00.0000] <zcorpan>
hmm, should we make <figcaption> imply </p>?

[01:04:00.0000] <annevk>
<legend> wheee

[01:06:00.0000] <othermaciej>
it's probably unwise to make new things imply </p>

[01:06:01.0000] <othermaciej>
though I guess the spec probably already does that?

[01:06:02.0000] <annevk>
not that <legend> implies </p>

[01:06:03.0000] <annevk>
othermaciej, yes

[01:06:04.0000] <annevk>
othermaciej, e.g. <figure> does

[01:06:05.0000] <annevk>
othermaciej, and <section> and all

[01:06:06.0000] <othermaciej>
<p> inside <figure> seems too obscure a case for <figcaption> to imply </p>

[01:07:00.0000] <annevk>
a poem is not that obscure

[01:07:01.0000] <annevk>
it's in the spec even :)

[01:07:02.0000] <annevk>
(but with a </p>)

[01:08:00.0000] <othermaciej>
using a non-preformatted poem in a figure is, I think a somewhat obscure use of figure

[01:16:00.0000] <zcorpan>
"Please ensure that your comment is descriptive enough that the editor can understand it."

[01:17:00.0000] <zcorpan>
wtf?

[01:17:01.0000] <zcorpan>
what's the magic phrase i need to use?

[01:17:02.0000] <zcorpan>
Hixie: ^

[01:21:00.0000] <annevk>
more than one word

[01:21:01.0000] <annevk>
maybe?

[01:21:02.0000] <zcorpan>
my comment is "Consider the poem example of figure. It should be possible to omit the </p> tag there. Maybe we should make <figcaption> imply </p> in the parser."

[01:21:03.0000] <annevk>
oh dunno then

[01:24:00.0000] <zcorpan>
  } else if ((text.length <= 5) || (text.indexOf(' ') != text.lastIndexOf(' '))) {

[01:24:01.0000] <zcorpan>
    showAlert('Please ensure that your comment is descriptive enough that the editor can understand it.');

[01:26:00.0000] <zcorpan>
javascript:(function(){text='Consider the poem example of figure. It should be possible to omit the </p> tag there. Maybe we should make <figcaption> imply </p> in the parser.';alert(text.indexOf(' ') != text.lastIndexOf(' '))})()

[01:26:01.0000] <zcorpan>
true

[01:26:02.0000] <zcorpan>
Hixie: maybe you meant to use == instead of != there?

[01:29:00.0000] <annevk>
weird that nobody else noticed that

[01:34:00.0000] <zcorpan>
Hixie: i also didn't get any notification when the bug was successfully submitted

[01:39:00.0000] <MikeSmith>
http://wiki.whatwg.org/wiki/PragmaExtensions -> "Error: 1205 Lock wait timeout exceeded; try restarting transaction (database.wiki.whatwg.org)"

[01:39:01.0000] <annevk>
hybi list o_O moments

[01:40:00.0000] <annevk>
those people just wanna go in circles

[01:40:01.0000] <annevk>
creative bug filing zcorpan :)

[01:42:00.0000] <othermaciej>
I need to catch up on all that mail this weekend

[01:48:00.0000] <zcorpan>
annevk: what was i supposed to do? :)

[01:49:00.0000] <hsivonen>
nessy: W3C standards don't, in general, have the requirement not to break the Web

[01:49:01.0000] <nessy>
that's plain stupid - who made up that rule? ;)

[01:50:00.0000] <nessy>
joking aside: I still owe you an answer to your email

[01:50:01.0000] <nessy>
had a rather busy week, but will get to it

[01:50:02.0000] <nessy>
sorry about the delay

[01:53:00.0000] <annevk>
what would break the Web?

[01:54:00.0000] <hsivonen>
annevk: the money behind TTML or the money behind XSL-FO deciding to take either in a direction where the formatting behavior no longer matches the CSS formatting behavior for the same-named concepts

[01:54:01.0000] <nessy>
XSL-FO is really unfortunate

[01:54:02.0000] <annevk>
not sure if it's really bad by the way if we have some breakage with the SRT format; in the end deployment to Web browsers is somewhat new and require a few changes anyway

[01:55:00.0000] <annevk>
even without XSL-FO TTML is a terrible format imo

[01:55:01.0000] <nessy>
I don't disagree

[01:55:02.0000] <annevk>
as you showed hand authoring is way too hard

[01:55:03.0000] <nessy>
but it is being implemented all over the place

[01:55:04.0000] <annevk>
that's what people say about XForms too

[01:55:05.0000] <hsivonen>
nessy: interoperably?

[01:55:06.0000] <nessy>
I am not defending the best solution (which I think would be a new format)

[01:55:07.0000] <hsivonen>
nessy: what annevk said

[01:55:08.0000] <nessy>
I am defending a compromise

[01:56:00.0000] <hsivonen>
sXBL was a compromise

[01:56:01.0000] <nessy>
/me goes looking up sXBL

[01:57:00.0000] <annevk>
anything that involves re-using TTML will result in a fork or enormous bloat in the browser

[01:57:01.0000] <nessy>
yeah, I think the implementations are interoperable

[01:57:02.0000] <nessy>
most only implement the baseline anyway

[01:58:00.0000] <annevk>
though likely it will be both a fork and enormous bloat

[01:58:01.0000] <nessy>
I still don't understand what the problem with defining a mapping is

[01:58:02.0000] <nessy>
for SRT we define a mapping

[01:58:03.0000] <nessy>
nobody has a problem with that

[01:58:04.0000] <annevk>
for SRT we define the format I think

[01:58:05.0000] <nessy>
for TTML we define a mapping and it's a big problem

[01:58:06.0000] <nessy>
yeah, and then implement mappings

[01:58:07.0000] <annevk>
what do you mean mapping?

[01:59:00.0000] <annevk>
like text/plain is a mapping to an HTML file with a <pre> element?

[01:59:01.0000] <nessy>
what the start time means, what the end time means, how the text is to be parsed

[01:59:02.0000] <nessy>
yeah, somewhat like that

[01:59:03.0000] <annevk>
SRT is a simple format

[02:00:00.0000] <annevk>
a simple text-based format of which the Web has a couple already

[02:00:01.0000] <nessy>
not when we throw all the baggage at it that it needs

[02:00:02.0000] <annevk>
it makes sense

[02:00:03.0000] <annevk>
SRT-HTML would be a new format

[02:00:04.0000] <annevk>
it's just an idea of how we could approach the richer captioning format

[02:01:00.0000] <nessy>
that still has to be parsed

[02:01:01.0000] <annevk>
sure

[02:01:02.0000] <nessy>
most of the mapping of elements will easily map directly on a HTML element

[02:01:03.0000] <annevk>
as hsivonen defined it could reuse a lot of what already needs to be defined for SRT

[02:01:04.0000] <nessy>
that makes it easy

[02:02:00.0000] <nessy>
yeah, haven't looked at hsivonen's work yet, sorry

[02:02:01.0000] <nessy>
/me should really work on this client project

[02:02:02.0000] <annevk>
TTML is not easy

[02:02:03.0000] <nessy>
creating it isn't

[02:02:04.0000] <nessy>
but parsing it isn't hard

[02:03:00.0000] <annevk>
e.g. <color> does not match any concept of what browsers already have

[02:03:01.0000] <nessy>
?

[02:03:02.0000] <annevk>
http://www.w3.org/TR/ttaf1-dfxp/#style-value-color

[02:04:00.0000] <annevk>
i.e. compared to CSS a-value is different, there's no #rrggbbaa format, CSS has #rgb, system colors, etc.

[02:05:00.0000] <nessy>
that is not an element - it's describing the value that an attribute can take, IIUC

[02:05:01.0000] <annevk>
that's still parsing

[02:06:00.0000] <nessy>
sure - but it can be easily mapped onto an existing CSS value

[02:06:01.0000] <annevk>
that parsing XML is not hard is somewhat self-evident if you already have an XML parser

[02:06:02.0000] <annevk>
nessy, we'd need a new TTML-<color>-value parser first

[02:06:03.0000] <annevk>
it's insane

[02:06:04.0000] <annevk>
there's also bugs it seems

[02:06:05.0000] <zcorpan>
i think the point is not how hard it is to implement, the point is reuse of existing code

[02:07:00.0000] <annevk>
e.g. familyName can be any string except for quotation characters, but that gives parsing problems when used in http://www.w3.org/TR/ttaf1-dfxp/#style-attribute-fontFamily

[02:07:01.0000] <nessy>
I regard it like supporting different image formats

[02:08:00.0000] <nessy>
every image format brought in another decoding library

[02:08:01.0000] <nessy>
I regard formats of external timed text as exactly the same

[02:08:02.0000] <nessy>
it's just decoding a file

[02:08:03.0000] <hsivonen>
I'm not so concerned about parsing as I am about how exactly the stuff interacts with the CSS frame constructor, with the event loop, etc.

[02:09:00.0000] <jgraham>
A web browser is just decoding a file

[02:09:01.0000] <nessy>
I'm sure that in the long run we have to deal with much more than SRT and TTML, btw

[02:09:02.0000] <jgraham>
For large values of just

[02:10:00.0000] <annevk>
if it's all so easy I'd love to see a proper spec

[02:10:01.0000] <annevk>
not really convinced that TTML is one

[02:10:02.0000] <nessy>
I am just really concerned that if we do not specify how TTML is to be parsed, we end up with e.g. IE implementing a parsing that they like, then other browser vendors following and we have diverging behaviour

[02:11:00.0000] <nessy>
we will get huge flak from the a11y community if we ignore TTML

[02:11:01.0000] <nessy>
but yeah - all this is kinda akademic unless we specify the parsing

[02:11:02.0000] <zcorpan>
/me expects a flak from the a11y community regardless of what we do

[02:11:03.0000] <hsivonen>
nessy: we aren't ignoring it. more like rejecting it after taking a good look

[02:12:00.0000] <annevk>
the a11y community that designs this stuff could really do with some lessons from the people that made HTML/CSS

[02:13:00.0000] <nessy>
ok, in this case I will just sit back and watch the battle ;)

[02:13:01.0000] <hsivonen>
nessy: about mapping: if you map TTML to XSL-FO to CSS frame tree, you need a new frame constructor

[02:13:02.0000] <hsivonen>
FAIL

[02:13:03.0000] <hsivonen>
nessy: if you map TTML to style sheet plus DOM, you need to do it on the main thread

[02:13:04.0000] <nessy>
I'd avoid going via XSL-FO

[02:13:05.0000] <annevk>
all this namespace loving complexity (also with WAI-ARIA until we got involved) is just a major fail

[02:14:00.0000] <nessy>
can't style sheet plus DOM be thrown into an iframe?

[02:15:00.0000] <nessy>
there are two things that we are mixing up right now: the parsing of the file (basically the "decoding") and how to display it

[02:15:01.0000] <nessy>
I think we need to keep them separate

[02:15:02.0000] <annevk>
my comments are highly related though

[02:15:03.0000] <nessy>
what you do when you have some html-snipped for a certain time frame is independent of the file fromat

[02:16:00.0000] <annevk>
if they had taken lessons from the HTML/CSS people we would not be facing a lot of these questions

[02:16:01.0000] <nessy>
annevk, I don't like namespaces either

[02:16:02.0000] <nessy>
annevk, I think you're right - I wrote about DFXP last year and criticised that, too

[02:16:03.0000] <annevk>
with "namespace loving complexity" i mean more than just the complexity; i.e. designing things on top of XHTML2, XSL-FO, etc.

[02:16:04.0000] <nessy>
but DFXP/TTML was not developed as a web format, unfortunately

[02:17:00.0000] <nessy>
it was the time when things were done that way

[02:17:01.0000] <nessy>
a lot of specs came out of the W3C in that manner during that time

[02:17:02.0000] <nessy>
I blame history

[02:17:03.0000] <nessy>
it's unfortunate - just like XHTML2 is unfortunate

[02:17:04.0000] <jgraham>
TTML has the air of being designed for a situation where a bunch of people in suits turn up to help you do accessibility. SRT has the air of something designed so that some guy in his pants in his basement can subtitle anime. Historically the successful parts of the web have been much closer to the guy-in-his-basement design than the people-with-suits design

[02:18:00.0000] <jgraham>
It would be surprising if this case were different

[02:18:01.0000] <nessy>
that's why I fought for SRT in the a11y group

[02:18:02.0000] <hsivonen>
nessy: throwing style sheet plus DOM in an an anonymous iframe is what I suggested

[02:18:03.0000] <nessy>
yes, indeed, and that's a good idea

[02:19:00.0000] <hsivonen>
nessy: the point is that I had a concrete suggestion of what goes into the DOM without the complexity of a "mapping"

[02:19:01.0000] <nessy>
it keeps it separate from the rest

[02:19:02.0000] <nessy>
/me really has to make time to read the proposal

[02:19:03.0000] <nessy>
/me really should be having this discussion at a different time

[02:20:00.0000] <nessy>
sorry - I have to go, but I will get back to you, promised

[02:22:00.0000] <annevk>
the thing I liked about hsivonen's SRT-HTML was that SRT itself could also be mapped this way and that therefore it would also allow for styling SRT-based subtitles

[02:28:00.0000] <zcorpan>
it'd be nice to support the SRT legacy so authors can just use their existing subtitles (if they're SRT)

[02:31:00.0000] <zcorpan>
/me looks at http://wiki.xiph.org/index.php/OggKate

[02:32:00.0000] <MikeSmithX>
Cloud Print is not a "new approach to printing", not by a long shot .. it's actually quite an established approach that some companies have based entire businesses on

[02:32:01.0000] <MikeSmithX>
e.g., going back the Palladium project at MIT

[02:33:00.0000] <MikeSmith>
and Dazel in Austin

[02:33:01.0000] <MikeSmith>
where I worked for a while and where Dan Connolly worked for a time too

[02:34:00.0000] <MikeSmith>
but NIH and all applies as usual, I guess

[02:35:00.0000] <annevk>
i bet your approach didn't have Google account tie-in

[02:35:01.0000] <MikeSmith>
nope

[02:36:00.0000] <MikeSmith>
so I guess that's what the "new approach" amounts to

[02:36:01.0000] <annevk>
it also seems somewhat novel that you can address the printer settings from a page

[02:38:00.0000] <hsivonen>
/me replied to the whatwg list about captioning

[02:39:00.0000] <hsivonen>
zcorpan: Kate also has the problem of importing a whole new text layout engine to a browser

[02:40:00.0000] <zcorpan>
so it seems

[02:42:00.0000] <MikeSmith>
annevk: yeah, but the main part that really makes that novel is the whole idea of a page being an application, not really anything specific about being able to print

[02:42:01.0000] <MikeSmith>
but anyway

[02:44:00.0000] <hsivonen>
it's unclear to me what problem the cloud print system solves

[02:45:00.0000] <hsivonen>
is the problem being solved that printers today don't ingest a PDF via the IPP?

[02:45:01.0000] <hsivonen>
but in Google's diagram, legacy printers needed a local intermediate box anyway

[02:45:02.0000] <annevk>
i think the major thing it solves is that apps can skip the OS when talking to printers

[02:45:03.0000] <hsivonen>
so why not put the PDF to printer converter on that box?

[02:46:00.0000] <annevk>
and that the OS no longer needs to have printer support

[02:46:01.0000] <hsivonen>
and make the Chrome OS machine send a PDF via IPP to that box?

[02:46:02.0000] <hsivonen>
annevk: will they have some kind of generic HTTP to USB bridge on the intermediate box?

[02:47:00.0000] <hsivonen>
so that Google's cloud can talk what needs to travel over the last USB hop?

[02:47:01.0000] <annevk>
it prolly runs some exe file I suppose that gets info from the cloud software about what needs to happen

[02:47:02.0000] <annevk>
i.e. the same kind of software "cloud printers" would have

[02:48:00.0000] <hsivonen>
I though the IPP was supposed to solve this already

[02:48:01.0000] <hsivonen>
though I'm not sure if it really does

[02:50:00.0000] <annevk>
dunno, would love to know more

[04:47:00.0000] <zcorpan>
are subtitles ever streamed?

[04:48:00.0000] <hsivonen>
zcorpan: presumably they could if a TimBL presentation at TPAC were published as video and the captioner manages to keep up

[04:49:00.0000] <hsivonen>
zcorpan: the W3C has previously published live audio and live captioning of the audio as separate resources

[04:49:01.0000] <virtuelv>
/me has seen real-time caqtioning once

[04:49:02.0000] <virtuelv>
it's disturbing

[04:50:00.0000] <hsivonen>
in the U.S. CNN has live captioning

[04:50:01.0000] <virtuelv>
mostly because I consider live-captioning to be a superhuman feat

[04:50:02.0000] <hsivonen>
I wonder how live captioners and court reporters deal with RSI

[04:51:00.0000] <zcorpan>
i guess you could make real-time captioning less disturbing by delaying the video and audio the same amount of time it takes for the captioner to type a line

[05:32:00.0000] <nessy>
real-time captioners deal with it the same way that real-time signers do or real-time translaters

[05:32:01.0000] <nessy>
always have a second person as backup that can take over from you

[05:35:00.0000] <zcorpan>
that doesn't solve the latency problem, only the gotta-pee problem

[07:40:00.0000] <zcorpan>
Hixie: hmm, shouldn't we have a frame type that has no length for keep-alive purposes? currently a keep-alive frame would need two bytes (unless you keep a binary frame open and end it when you want to send a real message)

[07:41:00.0000] <zcorpan>
(like 0x80 .... 0x80 .... 0x80 .... 0x00 0x00 message 0xFF)

[07:45:00.0000] <zcorpan>
though maybe it doesn't matter if keep-alive frames are 1 byte or 2 bytes or 10 bytes

[07:45:01.0000] <Dashiva>
I would agree

[07:50:00.0000] <zcorpan>
although i quite like the idea of abusing the length bytes of a binary frame :)

[07:50:01.0000] <Philip`>
Keepalives have got to do a whole TCP packet or two anyway, so I don't think one byte is really worth worrying about

[07:51:00.0000] <zcorpan>
yeah

[08:06:00.0000] <boblet>
anyone have a decent (chunky) dictionary on hand? I’d like to know how it formats example sentences/phrases

[08:28:00.0000] <JonathanNeal>
morning all

[08:30:00.0000] <boblet>
hey Jon

[09:31:00.0000] <JonathanNeal>
datalist has been dropped from the spec?

[09:43:00.0000] <zcorpan>
no

[11:54:00.0000] <micheil>
Anyone got a forward-to for MikeSmith?

[11:54:01.0000] <micheil>
(the (tm) variant)

[12:16:00.0000] <estellevw>
is there an explanation in the spec on the attribute values of the href attribute of <a>. For example, mailto: links?

[12:17:00.0000] <TabAtkins>
That just depends on the scheme of the url.  The schemes are all defined in various other specs.

[12:24:00.0000] <estellevw>
do you know where I might find it?

[12:24:01.0000] <estellevw>
and would it make sense to include it within the HTML5 spec?

[12:24:02.0000] <TabAtkins>
I'd search for "mailto scheme rfc".  No idea which of the results are canonical.

[12:25:00.0000] <TabAtkins>
And, frankly, probably yes, but we've so far tried not to define too much what a url means (just how to decompose it, etc.) for political reasons.

[12:26:00.0000] <TabAtkins>
We've got some Issues active that are trying to remove more of our definitions of urls, and instead defer to other specs.

[12:27:00.0000] <estellevw>
ok, so i won't file a bug for it ;) I do think mailto:, tel: and others should be defined in the spec, but would rather see the spec approved :D

[12:28:00.0000] <estellevw>
defering to wikiipedia over the spec just seems odd

[12:45:00.0000] <AryehGregor>
Who's deferring to Wikipedia over the spec?

[12:45:01.0000] <AryehGregor>
There are specs for all these, somewhere.

[12:45:02.0000] <AryehGregor>
Probably.

[13:21:00.0000] <Hixie>
the definitions of schemes is in an iana registry

[14:09:00.0000] <TabAtkins>
Hixie: I changed a single word in the first paragraph of that clause, from SHOULD to MUST, and then added a non-normative note.  How can that possibly weaken anything?

[14:10:00.0000] <Hixie>
huh. the way you added the note made me interpret "same input" differently.

[14:11:00.0000] <TabAtkins>
That just means you have to define "same input" more precisely. ^^_

[14:11:01.0000] <Hixie>
indeed

[14:11:02.0000] <Hixie>
i hadn't realised that might have been a source of the confusion

[14:11:03.0000] <TabAtkins>
I'll cancel the email I was about to send and let you correct yourself.

[14:11:04.0000] <Hixie>
heh

[14:12:00.0000] <JonathanNeal>
Did we settle on @alt?

[14:13:00.0000] <TabAtkins>
No, I'm still talking with one of our internal teams.

[14:13:01.0000] <TabAtkins>
Dude, it's been less than 24 hours.  ^_^

[14:30:00.0000] <AryehGregor>
"I do not think a tautology machine would be a meaningful fulfillment of this success criterion."

[14:30:01.0000] <TabAtkins>
We should ask the tautology machine for confirmation.

[14:38:00.0000] <othermaciej>
AryehGregor: I said that weeks ago!

[14:38:01.0000] <othermaciej>
did it become relevant again all of a sudden?

[14:38:02.0000] <AryehGregor>
othermaciej, yes, I'm catching up.

[14:38:03.0000] <AryehGregor>
No, I just started reading the backlog.

[14:39:00.0000] <AryehGregor>
Got totally messed up by Passover.  250ish spec mails and >500 MediaWiki commit mails.

[14:42:00.0000] <Hixie>
what happened to the wiki

[14:44:00.0000] <AryehGregor>
Hmm.

[14:44:01.0000] <Hixie>
hey, there's a mediawiki person here! how convenient :-P

[14:44:02.0000] <Hixie>
any idea what i should do?

[14:45:00.0000] <AryehGregor>
I have shell access as the wiki user, so, I'm looking.  :P

[14:45:01.0000] <AryehGregor>
(someone gave it to me, can't remember who)

[14:45:02.0000] <AryehGregor>
Looks like a DB problem to me, not a MW problem.

[14:45:03.0000] <AryehGregor>
But let me see.

[14:45:04.0000] <Hixie>
well you have access to the db too if you have access to the shell :-P

[14:45:05.0000] <AryehGregor>
Yep.

[14:45:06.0000] <Hixie>
time for me to start changing passwords i guess :-P

[14:46:00.0000] <Hixie>
(i'm fine with you having access, but, um, people, please let me know who you give access to :-P )

[14:46:01.0000] <AryehGregor>
Interesting.

[14:47:00.0000] <AryehGregor>
Maybe it is a MediaWiki problem.

[14:47:01.0000] <AryehGregor>
Let me see if setting $wgCacheDirectory fixes it.

[14:48:00.0000] <Hixie>
i hope it was Lachy_ who gave you access

[14:49:00.0000] <AryehGregor>
There we go, fixed.

[14:49:01.0000] <AryehGregor>
Apparently the new localization cache system isn't too reliable if you use a database backend.

[14:49:02.0000] <Hixie>
good to know

[14:49:03.0000] <AryehGregor>
It's meant to use a filesystem-based cache, I think CDB or something.

[14:50:00.0000] <AryehGregor>
You can chown /home/whatwikiuser/wikicache to be webserver-writable but not world-writable, if you like (I just set 777).

[14:50:01.0000] <AryehGregor>
(since, obviously, I can't do anything else)

[14:52:00.0000] <Hixie>
the server runs as the user

[14:52:01.0000] <AryehGregor>
Oh, mod_cgi?

[14:52:02.0000] <AryehGregor>
Okay.

[14:53:00.0000] <Hixie>
yeah

[14:53:01.0000] <Hixie>
should do anyway

[14:53:02.0000] <AryehGregor>
Yeah, I see that now.

[14:57:00.0000] <Hixie>
thanks for fixing it btw! much appreciated

[14:57:01.0000] <Hixie>
dunno what i would have done :-P

[14:57:02.0000] <AryehGregor>
You could have asked for help in #mediawiki!

[14:57:03.0000] <AryehGregor>
(although you probably wouldn't be running unreleased versions, so you probably wouldn't have hit the problem)

[16:02:00.0000] <AryehGregor>
Is there a site that lets you sketch something in a <canvas> (not Flash) and then saves it so you can share the URL, pastebin-style?

[16:04:00.0000] <Dashiva>
Are you looking for the existence of such a site, or do you just want to share the results of a canvas drawing?

[16:04:01.0000] <othermaciej>
http://mugtug.com/sketchpad/

[16:04:02.0000] <othermaciej>
?

[16:04:03.0000] <Steve^>
If this works, then yes:  http://tinyurl.com/y5ybzlm

[16:04:04.0000] <AryehGregor>
Oh, cosketch.com actually seems to use canvas.  Interesting.

[16:04:05.0000] <Steve^>
translates to some data: thing

[16:04:06.0000] <AryehGregor>
I wanted to share the results of a quick sketch easily.

[16:04:07.0000] <othermaciej>
that data: URL didn't load for me

[16:04:08.0000] <AryehGregor>
I want it to use canvas because I have Flash disabled in my primary browser.

[16:05:00.0000] <Steve^>
I did that with the same link othermaciej mentioned

[16:05:01.0000] <Dashiva>
That URL looks broken, yes

[16:05:02.0000] <AryehGregor>
Didn't work for me either.

[16:05:03.0000] <AryehGregor>
Maybe it's truncated?

[16:05:04.0000] <Steve^>
opens in Opera, at least

[16:05:05.0000] <Steve^>
ooo, now it doesn't

[16:05:06.0000] <Steve^>
must have been a cached version

[16:05:07.0000] <Dashiva>
It opens, but the image is messed up

[16:05:08.0000] <othermaciej>
Steve^: odd, the data: URL works right from the web app, but not via bit.ly

[16:05:09.0000] <othermaciej>
er, via tinyurl

[16:06:00.0000] <othermaciej>
I think tinyurl may be cutting off the URL

[16:06:01.0000] <Steve^>
yea, maybe another shortener will be better

[16:06:02.0000] <othermaciej>
you can also draw with Sketchpad, copy the image local, and them upload to http://imgur.com/

[16:06:03.0000] <Steve^>
how would you save it locally?

[16:07:00.0000] <othermaciej>
drag & drop after clicking the save button

[16:07:01.0000] <othermaciej>
or use Save As... from the browser File menu after clicking save

[16:08:00.0000] <Dashiva>
Ideally the browser would let you do save as... directly on a canvas, but it doesn't seem to be high priority

[16:09:00.0000] <AryehGregor>
Copying locally is too annoying.  I could do it in the GIMP if I wanted to do that.

[16:09:01.0000] <Steve^>
indeed

[16:10:00.0000] <othermaciej>
sketchpad with a built in image hosting service would be really sweet

[16:10:01.0000] <AryehGregor>
Although a web page would probably load a lot faster than the GIMP.

[16:10:02.0000] <othermaciej>
looks like they may be vaguely planning that

[16:23:00.0000] <AryehGregor>
Haha: http://arewefastyet.com/

[16:24:00.0000] <AryehGregor>
Why are V8 and Nitro so close there?  Because it's not mostly raw JS performance, or what?

[16:24:01.0000] <AryehGregor>
/me also wonders if they're testing against Chromium/WebKit trunk, or a release

[16:24:02.0000] <AryehGregor>
/me likes the "NO"

[16:33:00.0000] <othermaciej>
AryehGregor: V8 and Nitro are fairly similar designs, though I am surprised at how close they are on that diagram

[16:43:00.0000] <dglazkov>
hey, brain people. especiallly TabAtkins

[16:43:01.0000] <dglazkov>
:P

[16:44:00.0000] <dglazkov>
can someone point me to a place that specifies submit-on-enter behavior for forms? I've been looking at http://www.whatwg.org/specs/web-apps/current-work/multipage/forms.html#forms and can't spot it

[16:45:00.0000] <dglazkov>
in other words, in which cases does hitting an enter in a textfield could submit a form?

[16:45:01.0000] <dglazkov>
IE and FF seem to agree that only if there's one text area and no submit button in the form.

[16:45:02.0000] <dglazkov>
WebKit just does it if there's no submit button.

[16:45:03.0000] <dglazkov>
either behavior seems arbitrary and weird.

[16:45:04.0000] <Hixie>
search for "default button" or some such

[16:46:00.0000] <Hixie>
or "default submit button"

[16:47:00.0000] <othermaciej>
dglazkov: seems like "enter" in a form should always activate the default button for the form (or submit if there isn't one)

[16:47:01.0000] <othermaciej>
dglazkov: "enter" in a text field in a form, rather

[16:47:02.0000] <othermaciej>
and perhaps also other controls when focused

[16:48:00.0000] <othermaciej>
dglazkov: seems like this is also more a UI issue than an interop issue - it might be irrelevant on platforms with a virtual keyboard that doesn't have "enter" when editing a text field for instance

[16:48:01.0000] <dglazkov>
othermaciej: try https://bug-21716-attachments.webkit.org/attachment.cgi?id=24462 in FF or WebKit.

[16:49:00.0000] <othermaciej>
dglazkov: in Safari/Mac + ToT WebKit pressing enter submits the form

[16:49:01.0000] <TabAtkins>
dglazkov: You want section 4.10.21.2

[16:49:02.0000] <dglazkov>
othermaciej: yep.

[16:49:03.0000] <Hixie>
right, what TabAtkins said

[16:49:04.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#implicit-submission

[16:49:05.0000] <othermaciej>
is that a bad thing?

[16:50:00.0000] <Hixie>
dglazkov: the spec covers that test case with no ambiguity

[16:50:01.0000] <TabAtkins>
No ambiguity, but with some leeway for UAs to do what they want.

[16:50:02.0000] <Hixie>
and says firefox is buggy

[16:50:03.0000] <Hixie>
no, not that test case

[16:51:00.0000] <TabAtkins>
Ah, right, true.

[16:51:01.0000] <Hixie>
there's no button, so it's pretty strictly defined

[16:51:02.0000] <TabAtkins>
Yes.

[16:51:03.0000] <TabAtkins>
Since FF does have an implicit submission method, there are some MUSTs it has to follow there.

[16:52:00.0000] <dglazkov>
so IE and FF are both wrong then?

[16:52:01.0000] <dglazkov>
and Opera too

[16:54:00.0000] <dglazkov>
when only one browser gets the spec right, is it: a) awesome b) horrifying or c) ain't right?

[16:55:00.0000] <TabAtkins>
Depends on the feature.  ^_^

[16:57:00.0000] <TabAtkins>
dglazkov: What behavior are you seeing in FF that makes you think its buggy?  FF submits a form with or without a submit button when you hit enter in a text input.

[16:57:01.0000] <TabAtkins>
That appears to be exactly what the spec says.

[16:58:00.0000] <jcranmer>
hmm, judging from roc's latest post, it seems like the worries about hardware decoding for theora are overblown

[16:59:00.0000] <othermaciej>
dglazkov: it seems like hitting Enter in a text input should pretty much always submit the form, unless some key event handler would prevent that

[16:59:01.0000] <othermaciej>
dglazkov: do you think that is wrong for some reason? are there any sites that rely on sometimes not submitting in that case?


2010-04-17
[17:00:00.0000] <TabAtkins>
Fwiw, the issues is that FF won't implicitly submit a form that has *two* text inputs and no submit button.

[17:01:00.0000] <dglazkov>
othermaciej: I don't know if it's wrong, it's just different. The spec and WebKit do one thing, the rest of the gang does the other.

[17:01:01.0000] <dglazkov>
https://bug-21716-attachments.webkit.org/attachment.cgi?id=24462 <-- try hitting enter in the text field in non-WebKit browsers.

[17:02:00.0000] <KaOSoFt>
Good evening.

[17:02:01.0000] <othermaciej>
dglazkov: Firefox doesn't submit, but that strikes me as unhelpful behavior

[17:03:00.0000] <othermaciej>
I would say it's a bug in Firefox unless there is some compat reason for it

[17:05:00.0000] <dglazkov>
othermaciej: I've been digging for gold in b.w.o, but only found one living site that still shows a compat issue: http://www.flo-joe.co.uk/cae/students/wordbank/wform.htm

[17:07:00.0000] <othermaciej>
dglazkov: the compat issue being that pressing enter does something useless?

[17:07:01.0000] <dglazkov>
othermaciej: yeah

[17:07:02.0000] <dglazkov>
othermaciej: https://bugs.webkit.org/show_bug.cgi?id=9756

[17:07:03.0000] <othermaciej>
I see your point but that seems pretty minor

[17:08:00.0000] <othermaciej>
dglazkov: only thing I could imagine changing is that if there is a button but no submit button, maybe one of the buttons should be the default action for pressing enter

[17:09:00.0000] <dglazkov>
othermaciej: I wonder if the current behavior of non-WebKit browsers is simply a leftover from isindex era

[17:14:00.0000] <annevk2>
hmm, I don't think HTML5 defines the crazy stuff in http://my.opera.com/hallvors/blog/2010/04/16/how-to-break-youtube does it?

[17:15:00.0000] <annevk2>
/me wonders why hsivonen hasn't complained about it

[17:46:00.0000] <annevk>
for http://wiki.whatwg.org/wiki/Timed_tracks I know that e.g. Ghost in the Shell 2 has things like that going on

[17:46:01.0000] <annevk>
but I don't have any screenshot

[17:47:00.0000] <Hixie>
i just meant that as an example of extremes

[17:48:00.0000] <Hixie>
hopefully most examples won't be that crazy :-)

[17:49:00.0000] <annevk>
those kind are actually hard to follow

[17:49:01.0000] <Hixie>
yeah

[17:50:00.0000] <annevk>
you listen to the Japanese, the subtitles are long and switch fast, while at the top the references in the text are explained

[17:51:00.0000] <Hixie>
the jokes are obviously only useful if you're pausing

[17:51:01.0000] <Hixie>
joke explanations i mean

[17:51:02.0000] <Hixie>
"footnotes"

[18:40:00.0000] <Hixie>
annevk: yt?

[18:41:00.0000] <Hixie>
annevk: is xhr's description of sync xhr complete, or is it pending on something?

[18:45:00.0000] <annevk>
i think it's complete

[18:45:01.0000] <annevk>
though maybe it should be updated to use the redirect flag of fetch

[18:45:02.0000] <annevk>
as that's new

[18:48:00.0000] <Hixie>
k

[18:49:00.0000] <Hixie>
so no tasks run while sync xhr is happening, right?

[18:49:01.0000] <Hixie>
i'm trying to work out how it interacts with workers

[18:49:02.0000] <Hixie>
seems fine currently

[18:49:03.0000] <annevk>
yeah, that's the idea

[18:49:04.0000] <annevk>
no timeouts, no events

[18:49:05.0000] <othermaciej>
you mean workings running a sync XHR, or running during a sync XHR?

[18:50:00.0000] <Hixie>
the former

[18:50:01.0000] <Hixie>
annevk: thanks

[18:51:00.0000] <Hixie>
annevk: should that change, you'll likely need to use the "spin the event loop" hook, in which case let me know because i'll have to define how that works with workers

[18:51:01.0000] <Hixie>
annevk: currnetly workers don't support 'spin the event loop' or 'await a stable state'/'synchronous sections', because nothing that runs in a worker can trigger those

[18:51:02.0000] <Hixie>
(i just noticed the storage mutex isn't handled right in workers and i'm fixing that)

[18:52:00.0000] <annevk>
okay

[21:36:00.0000] <piyushmishra>
hi

[21:37:00.0000] <piyushmishra>
I am trying to make a whiteboard app for multiple users in html5 canvas + JS and php

[21:37:01.0000] <piyushmishra>
can anyone please recommend good html5 canvas tutes/documentation etc

[21:37:02.0000] <piyushmishra>
?

[00:40:00.0000] <Hixie>
is http://www.w3.org/Bugs/Public/show_bug.cgi?id=9545 really true?

[00:41:00.0000] <Hixie>
you can't turn off the borders on <iframes> in IE with CSS?

[00:47:00.0000] <estellevw>
you have to include frameborder=0 or something like that in IE

[00:48:00.0000] <estellevw>
frameborder="0"

[00:52:00.0000] <nessy>
incidentally, can iframes be transparent

[01:00:00.0000] <nessy>
seems IE needs a allowTransparency  as well as CSS background-color:transparent 

[08:10:00.0000] <Tila>
hello

[08:11:00.0000] <Tila>
is this a place to ask questions about html5?

[08:12:00.0000] <miketaylr>
yes

[08:15:00.0000] <Tila>
i heard about js able to use storage on user disk, what kind of storage is it? files? a key/value db?

[08:45:00.0000] <miketaylr>
Tila: take a peek here:http://dev.w3.org/html5/webstorage/


2010-04-18
[18:55:00.0000] <boblet>
MikeSmithX: yt?

[18:57:00.0000] <MikeSmithX>
yeah

[18:57:01.0000] <MikeSmithX>
boblet: ↑

[19:53:00.0000] <cardona507>
I am trying to enable high accuracy on geolocation and don't totally understand the syntax. If I am using "navigator.geolocation.getCurrentPosition(function(position, error){//code});" - what is the correct way to set the PositionOptions object's enableHighAccuracy property to true?

[20:21:00.0000] <GPHemsley>
Is the W3C Validator down?

[20:21:01.0000] <GPHemsley>
well, just the HTML5 part of it

[20:22:00.0000] <GPHemsley>
"The error encountered was: 500 Can't connect to localhost:8888 (connect: Connection refused)"

[07:14:00.0000] <boblet>
JonathanNeal: yt?

[07:17:00.0000] <paul_irish>
still 6.23am there. he'll probably be up in a few

[07:22:00.0000] <boblet>
paul_irish: thanks

[07:22:01.0000] <boblet>
damn everytimezone.com for being so popular

[09:19:00.0000] <paul_irish>
TabAtkins: it's for the lulz :)

[09:20:00.0000] <miketaylr>
i'm terrible at sarcasm

[10:59:00.0000] <JonathanNeal>
hi boblet!

[10:59:01.0000] <JonathanNeal>
what's up

[11:58:00.0000] <miketaylr>
looking at <datalist>, "The contents of the element represents fallback content for legacy user agents, intermixed with option elements that represent the predefined options."

[11:58:01.0000] <miketaylr>
this means i can just do <datalist><select><option>...</select></datalist>, right?

[11:58:02.0000] <miketaylr>
s/can just do/get away with/

[16:44:00.0000] <Dashiva>
How does a feed reader that depends on atom:id for identification handle RSS feeds?

[16:48:00.0000] <Philip`>
RSS is RDF so surely it works using the magic of RDF

[16:50:00.0000] <Dashiva>
Non 1.0 RSS then :P


2010-04-19
[17:02:00.0000] <Dashiva>
I don't think Ian's cat would be a good change controller

[17:09:00.0000] <AryehGregor>
Why not?

[17:10:00.0000] <AryehGregor>
Cats are known to be very intelligent.

[17:10:01.0000] <AryehGregor>
Well, my cats are complete idiots, but maybe I just have losers.

[17:10:02.0000] <Dashiva>
Cats have short lives, it might die before HTML6 becomes PR

[17:10:03.0000] <AryehGregor>
Maybe we can cryogenically freeze it.

[17:11:00.0000] <Philip`>
Write an AI simulation of the cat

[17:11:01.0000] <AryehGregor>
Should be easy.

[17:12:00.0000] <Dashiva>
That's a typically "the tools will save us" fallacy, Philip`

[17:12:01.0000] <Dashiva>
I expected better from you

[17:12:02.0000] <AryehGregor>
"Stare plaintively at human until petted or fed.  Meow if necessary.  Run away from thrown objects."

[17:13:00.0000] <AryehGregor>
We can cryogenically freeze cats right now!

[17:13:01.0000] <AryehGregor>
Also, it could be hereditary.

[17:13:02.0000] <Dashiva>
What if it has multiple kittens, and they disagree?

[17:14:00.0000] <AryehGregor>
This was solved millennia ago by the doctrine of primogeniture.

[17:14:01.0000] <AryehGregor>
Don't reinvent the wheel.

[17:14:02.0000] <Dashiva>
Primogeniture is a rathole you don't want to go down

[17:15:00.0000] <Dashiva>
You have to resolve issues like whether male descendents have priority, and who inherits when the current controller has no children, and how to handle inlaws, etc etc

[17:15:01.0000] <Dashiva>
Not to mention the potential for abuse by guardians of underage heirs

[17:16:00.0000] <Philip`>
Put all the kittens into a last-man-standing fight to the death

[17:16:01.0000] <Philip`>
and then the victor can make uncontested decisions

[17:16:02.0000] <Dashiva>
That has a high chance of leading to the winner dying of wounds

[17:17:00.0000] <Philip`>
Wait for all the kittens to have kittens of their own, and then put the parent cats into the deathmatch

[17:17:01.0000] <Philip`>
and if they all die then repeat with the younger generation

[17:17:02.0000] <Philip`>
until there is one survivor

[17:18:00.0000] <Dashiva>
But by the time the younger generation is old enough, you'll have been without a change controller for a long time

[17:18:01.0000] <Philip`>
Force-feed them with rapid aging drugs

[17:19:00.0000] <Dashiva>
Also, I don't think fights to the death conforms to local laws in most jurisdictions

[17:19:01.0000] <Philip`>
Do it on the moon, so it's not in anybody's jurisdiction

[17:19:02.0000] <Dashiva>
I don't think W3C can afford that

[17:20:00.0000] <Philip`>
I'm sure TV companies would pay you a lot for live broadcast rights to lunar kitten deathmatches

[17:21:00.0000] <Dashiva>
If that's the case there should already exist such ventures

[17:21:01.0000] <Dashiva>
And the W3C could utilize existing infrastructure

[17:21:02.0000] <Philip`>
I'm sure such ventures do exist - Unreal Tournament demonstrates one of them

[17:22:00.0000] <Dashiva>
I'm sorry to say, but Unreal Tournament is not a documentary

[17:22:01.0000] <Philip`>
Oh

[17:23:00.0000] <Philip`>
/me will have to make new business plans, then :-(

[18:10:00.0000] <AryehGregor>
One annoying thing about reading old spec mail is you aren't sure whether to respond to old messages.

[18:10:01.0000] <AryehGregor>
"It will probably mean that Mediawiki some time in the future are going to disallow it, as well as Facebook, since using the built in functionality sooner or later will trickle into their code base."

[18:11:00.0000] <AryehGregor>
Does that suppose that MediaWiki or Facebook actually uses FILTER_VALIDATE_EMAIL?

[18:11:01.0000] <AryehGregor>
Because in MediaWiki it's been strpos( $email, '@' ) !== false for a long time.

[18:11:02.0000] <AryehGregor>
Too much effort to validate e-mail addresses, especially when mail agents tend to ignore the RFCs in practice and accept bad addresses.

[18:12:00.0000] <othermaciej>
probably depends on the current state of the issue

[18:12:01.0000] <othermaciej>
there were some bugs filed on this topic

[18:12:02.0000] <AryehGregor>
/me responds on a bug instead

[18:13:00.0000] <othermaciej>
I don't think any of them have been escalated to tracker issues but I am not sure what the current state is

[18:15:00.0000] <AryehGregor>
www.w3.org/Bugs/Public/show_bug.cgi?id=9392 is WONTFIX.

[18:19:00.0000] <Dashiva>
I know facebook rejects + in emails

[18:19:01.0000] <Hixie>
AryehGregor: unless it gets escalated i wouldn't worry about it

[18:19:02.0000] <AryehGregor>
Guess not.

[18:19:03.0000] <Hixie>
("it" being any general issue, not any one in particular)

[18:21:00.0000] <estellevw>
If the client has in memory maxed out browser cache, is there a spec as to how the browser / user agent should handle having no more local storage?

[18:22:00.0000] <Hixie>
estellevw: for appcache, http cache, local storage, cookies, something else?

[18:22:01.0000] <estellevw>
app cache

[18:23:00.0000] <Hixie>
no, nothing really defines that

[18:23:01.0000] <Hixie>
it's basically up to the ua

[18:23:02.0000] <estellevw>
i don't really see a way of controlling it, other than from the developer using dates... does the user have a way to set preferences.

[18:23:03.0000] <Hixie>
it falls under the "hardware limitations" clause

[18:23:04.0000] <estellevw>
ok

[18:24:00.0000] <estellevw>
should we be defining a path... oldest out first... or warning to user that their storage is full?

[18:24:01.0000] <estellevw>
or other?

[20:07:00.0000] <TabAtkins__>
AryehGregor: As far as I know, basically no one uses PHP's email validation filter, because it's buggy as all hell from what I understand.

[20:09:00.0000] <boblet>
re: email validation, I’d wager almost none of the scripts for that took multibyte domains into account

[20:09:01.0000] <boblet>
(random aside since I have one ;-)

[00:15:00.0000] <annevk>
hmm, Gecko has computed style for elements not in the DOM

[00:15:01.0000] <annevk>
how does that even work?

[00:17:00.0000] <annevk>
see e.g. http://dump.testsuite.org/2009/getcomputedstyle/liveness/002.htm

[00:17:01.0000] <annevk>
it's very weird

[00:17:02.0000] <annevk>
/me goes to download a nightly

[00:20:00.0000] <annevk>
same deal

[00:20:01.0000] <annevk>
CSS is matched against elements not even in the DOM, isn't that somewhat weird?

[00:47:00.0000] <annevk>
in other news, Tyler seems to continue suggesting forking CORS

[00:47:01.0000] <annevk>
http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0162.html

[00:47:02.0000] <annevk>
*sigh*

[00:49:00.0000] <Hixie>
i wish the hybi chairs would clear up once and for all whether amateur programmers are part of the target audience or not

[00:49:01.0000] <Hixie>
all we're doing is arguing around in circles because there are two completely different opinions about who the protocol should be aimed for

[00:50:00.0000] <annevk>
it seems Mike is sort of arguing for both sides, assuming SSL is too complex for amateur programmers

[00:50:01.0000] <Hixie>
i think he's saying what i said, namely that we should assume experts deploy using tls, but still allow a non-tls variant

[00:51:00.0000] <annevk>
doing everything over SSL and saving a roundtrip is the first novel suggestion in months on that list I think

[00:51:01.0000] <Hixie>
yeah

[00:51:02.0000] <MikeSmith>
a header named "U" could certainly not be accused of being verbose.. nor misleading, I guess.. since whatever meaning it's supposed to have can't be discerned from the name at all

[00:51:03.0000] <Hixie>
wish i understood enough about TLS to know if it made any sense :-)

[00:51:04.0000] <MikeSmith>
annevk: I think I'll post to suggest that the header should be named "X"

[00:52:00.0000] <annevk>
Hixie, ah yeah, I guess you can have two separate handshakes, one for non-tls usage and for tls usage

[00:52:01.0000] <annevk>
MikeSmith, it's more that he wants to replace current access control headers with something new

[00:53:00.0000] <MikeSmith>
yeah

[00:53:01.0000] <annevk>
MikeSmith, I think he's missing that this only solves the response side, the request side also has a whitelist

[00:53:02.0000] <MikeSmith>
I just particularly liked the suggestion for the alternativ eheader name

[00:53:03.0000] <annevk>
maybe I should point that out

[00:56:00.0000] <MikeSmith>
It seems like a lot of time might be saved if somebody from the Chrome team were to publicly say whether they'd have any plans at all to actually implement UMP

[00:56:01.0000] <MikeSmith>
or even not publicly

[00:59:00.0000] <annevk>
TabAtkins, maybe you can figure out what the Google Chromium team is interested in?

[00:59:01.0000] <annevk>
TabAtkins_, i.e. CORS or UMP

[00:59:02.0000] <annevk>
or Hixie, or anyone

[00:59:03.0000] <annevk>
it doesn't even matter so much I suppose

[01:00:00.0000] <annevk>
if CORS is not going to be removed it makes no sense to add UMP

[01:17:00.0000] <othermaciej>
ah, CORS vs UMP, too bad it's only about #10 on the list of total messes the I feel obligated to clean up

[01:18:00.0000] <othermaciej>
annevk: the Chrome security people I've talked to don't seem interested in removing CORS or in adding a separate UMP implementation, but no one has given an authoritative statement

[01:18:01.0000] <othermaciej>
MikeSmith: have any other browser vendors besides Apple made a public statement on the matter?

[01:20:00.0000] <annevk>
I could make a statement, but we haven't implemented yet so I'm not sure whether it's worth it

[01:22:00.0000] <othermaciej>
it seems relevant to me, since people specifically asked for statements of interest in implementing UMP or CORS

[01:22:01.0000] <othermaciej>
I'm not sure if there was any clear position taken by Mozilla

[01:25:00.0000] <MikeSmith>
othermaciej: none that I remember

[01:25:01.0000] <othermaciej>
or Microsoft for that matter

[01:26:00.0000] <othermaciej>
why were you specifically interested in the Chrome team's opinion?

[01:26:01.0000] <othermaciej>
fwiw neither Mark Miller nor Tyler Close are on the Chrome team, and Adam Barth is (as a contractor)

[01:27:00.0000] <annevk>
well, they implement it

[01:27:01.0000] <annevk>
and would need to implement UMP

[01:28:00.0000] <MikeSmith>
othermaciej: because they work at the same company

[01:28:01.0000] <othermaciej>
I don't know of them having any special inside track with the Chrome team, but I'm not the expert

[01:28:02.0000] <othermaciej>
certainly no one has proposed adding UMP to WebKit

[01:37:00.0000] <annevk>
emailed asking for implementor interest in a separate thread

[01:38:00.0000] <annevk>
I pointed out your reply in that email othermaciej and added Opera's position

[01:38:01.0000] <othermaciej>
gah, the hybi list generates stupid amounts of mail

[01:56:00.0000] <annevk>
othermaciej, if it was new information it might be more acceptable...

[01:56:01.0000] <MikeSmith>
hsivonen: any clue on what might be the problem here:

[01:56:02.0000] <MikeSmith>
http://validator.w3.org/check?uri=http%3A%2F%2Fwww.trigofacile.com%2Fjardins%2Flucullus%2Fgrec%2Fdemocrite%2Ffragments.htm&charset=%28detect+automatically%29&doctype=Inline&ss=1&group=0&user-agent=W3C_Validator%2F1.767#line-123

[01:57:00.0000] <MikeSmith>
HTML5 facet of W3C validator is reporting "End of file seen and there were open elements", but validator.nu says its valid

[01:57:01.0000] <othermaciej>
annevk: I'm supposed to be a co-editor for the requirements document but it's almost impossible to keep up with email enough to know what has consensus

[01:57:02.0000] <annevk>
MikeSmith, some kind of zero byte issue?

[01:58:00.0000] <othermaciej>
annevk: I think the hybi chairs need to raise some trac issues on clear permathreads so we can stop discussing those points endlessly

[01:58:01.0000] <annevk>
some kind of chair guidance would be good yes

[01:58:02.0000] <MikeSmith>
annevk: zero byte? it seems to actually be reading in the whole file OK

[01:59:00.0000] <annevk>
maybe the W3C has smaller buffers or some such?

[01:59:01.0000] <hsivonen>
MikeSmith: that sure is odd

[02:00:00.0000] <hsivonen>
MikeSmith: does it work with W3C-hosted v.nu with the same parser revision?

[02:00:01.0000] <MikeSmith>
annevk: if so, that's not the error I'd expect to see, I guess

[02:00:02.0000] <hsivonen>
MikeSmith: i.e. without the perl front-end but otherwise the same

[02:01:00.0000] <MikeSmith>
hsivonen: yeah - http://www.w3.org/html/check?doc=http%3A%2F%2Fwww.trigofacile.com%2Fjardins%2Flucullus%2Fgrec%2Fdemocrite%2Ffragments.htm

[02:01:01.0000] <MikeSmith>
though that's a different instance and a different machine even

[02:02:00.0000] <MikeSmith>
must be due to some monkeying that the perl layer is doing

[02:03:00.0000] <hsivonen>
MikeSmith: which part is responsible for showing the source?

[02:03:01.0000] <hsivonen>
MikeSmith: that is, does the full source showing prove anything about the full source reaching the back end?

[02:03:02.0000] <MikeSmith>
hsivonen: hmm, yeah, it may not prove that

[02:04:00.0000] <annevk>
http://www.w3.org/2009/12/vnu/style.css uses padding:none which should be padding:0 or some such

[02:04:01.0000] <MikeSmith>
I don't know much at all about how the W3C valdiator works

[02:04:02.0000] <hsivonen>
ok. I'm going to claim the front end cuts the content

[02:04:03.0000] <MikeSmith>
annevk: that one is my fault, thanks

[02:04:04.0000] <annevk>
it also has various -pre-wrap stuff that can prolly become just pre-wrap

[02:04:05.0000] <JonathanNeal>
and 0px is te same as 0

[02:04:06.0000] <JonathanNeal>
Those jerks!

[02:04:07.0000] <MikeSmith>
heh

[02:04:08.0000] <MikeSmith>
"jerk", singular

[02:04:09.0000] <MikeSmith>
=me

[02:05:00.0000] <MikeSmith>
I'm not a CSS Ninja

[02:05:01.0000] <JonathanNeal>
MikeSmith, doing something right in a way that could be done another way that others might prefer?

[02:05:02.0000] <MikeSmith>
I'm more of a CSS cro-magnon

[02:05:03.0000] <annevk>
CSS maximalist

[02:06:00.0000] <JonathanNeal>
I'm not wearing my "I Less-Than-Three MikeSmith" shirt tomorrow.

[02:06:01.0000] <MikeSmith>
JonathanNeal: I have perfected this method that I call "brute force trial-and-error"

[02:08:00.0000] <Hixie>
i just read an article that suggested the following as a strategy "Find a problem, find the simplest solution, punt the rest"

[02:08:01.0000] <Hixie>
i think that pretty much summarises a lot of how we've been approaching problems in html5 and related specs

[02:10:00.0000] <JonathanNeal>
I just want abbr!

[02:10:01.0000] <JonathanNeal>
Dangit ... I mean

[02:10:02.0000] <JonathanNeal>
alt!

[02:12:00.0000] <JonathanNeal>
the<wbr>rapist, what an infinitely useful element.

[02:12:01.0000] <JonathanNeal>
And by alt, I mean for microdata, Mr. Hixie.

[02:14:00.0000] <MikeSmith>
Hixie: no, that's all wrong. The correct strategy is to start with a vague "framework" that can be used to solve all kinds of problems -- including problems we didn't even know we had -- and then to introduce a magical "policy" mechanism to control it all, then people show up and ask to pile on all kinds of additional features and we do that to make them happy

[02:14:01.0000] <MikeSmith>
that's the way it's supposed to work

[02:17:00.0000] <wirepair>
wow you guys are bitter

[02:17:01.0000] <wirepair>
;>

[02:22:00.0000] <Hixie>
wirepair might have us figured out :-P

[02:22:01.0000] <wirepair>
hehe

[02:39:00.0000] <MikeSmith>
hsivonen: I tested with localhost 8888 on the W3C validator host and can't reproduce that error

[02:39:01.0000] <MikeSmith>
so it must be some problem in the perl layer

[02:40:00.0000] <hsivonen>
MikeSmith: ok. Good. :-)

[02:46:00.0000] <ment>
hmm this is maybe not the best channel to ask this question but it's somehow vaguely releated - are there any guidelines for writing css3 rule matcher for html*/xhtml* or any kind of previous work in this field?

[03:25:00.0000] <othermaciej>
I think I just killed w3c bugzilla

[03:26:00.0000] <annevk>
finally made IRIs work on my blog

[03:26:01.0000] <Yudai>


[03:26:02.0000] <Yudai>
oops

[03:26:03.0000] <Yudai>
MikeSmith: do you have any idea about "validator/entity-map" ?

[03:27:00.0000] <Yudai>
build.py all fails on loading the file

[03:29:00.0000] <othermaciej>
MikeSmith: I just discovered a downside to Cc-ing public-html on all bug keyword changes....

[03:29:01.0000] <othermaciej>
MikeSmith: any chance we could turn that off temporarily?

[03:30:00.0000] <annevk>
he's in some kind of meeting currently I believe

[03:30:01.0000] <othermaciej>
I just spammed public-html with hundreds of emails accidentally :-(

[03:31:00.0000] <annevk>
oh, so those twenty-odd were just the beginning

[03:31:01.0000] <othermaciej>
yeah

[03:31:02.0000] <othermaciej>
there's gonna be, like, 250

[03:31:03.0000] <othermaciej>
only noticed what was happening partway

[03:31:04.0000] <annevk>
I wonder what our company mail server will do

[03:36:00.0000] <roc>
I have nightmares about triggering something like this

[03:36:01.0000] <roc>
one time one of my kids found an open email window and invited a couple of dozen of my colleagues to GMail

[03:36:02.0000] <roc>
that was embarrassing

[03:37:00.0000] <othermaciej>
well - people will get to learn a lot about some of the old bugs we had hanging around :-/

[03:37:01.0000] <othermaciej>
it could be worse, I could have accidentally dropped prototype hardware in a seedy bar

[03:39:00.0000] <Dashiva>
You're more anonymous that way, though

[03:39:01.0000] <othermaciej>
looking at these bugs in a mail folder, some of them seem really ridiculous

[03:41:00.0000] <annevk>
othermaciej, ah, so it wasn't you? :p

[03:42:00.0000] <othermaciej>
annevk: honestly - I have no idea if it's even real

[03:42:01.0000] <othermaciej>
that kind of stuff is on a need-to-know basis

[03:42:02.0000] <roc>
what are you talking about?

[03:42:03.0000] <othermaciej>
god damn did I ever mailbomb the list

[03:43:00.0000] <annevk>
engadget claimed to have found an next-gen iphone

[03:43:01.0000] <annevk>
s/an/a doh

[03:45:00.0000] <roc>
hehe

[03:46:00.0000] <Dashiva>
othermaciej: lists, even. public-html-bugzilla as well :)

[03:47:00.0000] <othermaciej>
Dashiva: anyone on public-html-bugzilla has volunteered to be spammed by bugmail, I think

[03:49:00.0000] <Dashiva>
Well, there's bugmail and then there's 250 messages adding NoReply

[03:50:00.0000] <Lachy>
Hixie, re this about mediawiki from a few days ago, I thought you knew I'd given AryehGregor access and basically made him responsible for upgrading the wiki. http://krijnhoetmer.nl/irc-logs/whatwg/20100416#l-1070

[03:50:01.0000] <Dashiva>
It's too bad bugzilla can't generate one mail for each multichange

[03:52:00.0000] <Lachy>
othermaciej, what does the NoReply tag mean? 

[03:52:01.0000] <othermaciej>
Dashiva: that would have been much better in this case

[03:52:02.0000] <othermaciej>
Lachy: per our decision policy, it means we're going to assume we are not getting a reply to this bug, and it is now in a terminal state

[03:52:03.0000] <othermaciej>
Lachy: these are the bugs I moved to VERIFIED a month ago

[03:53:00.0000] <Dashiva>
Timeout while waiting for CLOSED

[03:53:01.0000] <othermaciej>
Lachy: that didn't get reopened, closed, or moved to the tracker

[03:53:02.0000] <Lachy>
ah, ok. So they have to move from VERIFIED FIXED to CLOSED?

[03:55:00.0000] <MikeSmith>
othermaciej: sorry, was afk

[03:55:01.0000] <MikeSmith>
I can try to turn it off now

[03:56:00.0000] <MikeSmith>
but I think it might be too late

[03:56:01.0000] <othermaciej>
MikeSmith: I stopped bugzilla at half the bugs

[03:56:02.0000] <MikeSmith>
ok

[03:57:00.0000] <othermaciej>
Lachy: technically, the policy says they stay in VERIFIED with the NoReply keyword, but in the next rev of the decision policy I'm proposing CLOSED with the NoReply keyword

[03:57:01.0000] <Dashiva>
Luckily it looks like nobody posted anything during the bug storm

[03:57:02.0000] <Lachy>
but for the 3 bugs I filed, I can now move them to closed myself because they are fixed?

[03:57:03.0000] <Dashiva>
That's what you were supposed to do these preceding weeks :)

[03:58:00.0000] <Lachy>
yeah, but for the last few weeks, I haven't been paying too much attention to HTMLWG stuff

[03:58:01.0000] <Dashiva>
But yes, go ahead. You can remove the NoReply keyword when you set it to CLOSED.

[04:00:00.0000] <MikeSmith>
othermaciej: you still want me to disable e-mail notifications for now?

[04:00:01.0000] <othermaciej>
MikeSmith: that would be nice, yes

[04:00:02.0000] <MikeSmith>
ok

[04:01:00.0000] <othermaciej>
Lachy: yes, you can move stuff to CLOSED and remove NoReply (and optionally add Disagree)

[04:01:01.0000] <othermaciej>
MikeSmith: I still have 235 more bugs to mass-modify

[04:01:02.0000] <MikeSmith>
hai

[04:01:03.0000] <Lachy>
does Disagree mean that I don't agree with the fix?

[04:03:00.0000] <othermaciej>
Lachy: it means you disagree with the resolution, but you don't care to escalate, reopen, or raise a formal objection

[04:04:00.0000] <Lachy>
ok. I don't disagree with them anyway, but good to know.

[04:05:00.0000] <MikeSmith>
othermaciej: OK, I just disabled all bugzilla mail for the public-html⊙wo user

[04:06:00.0000] <othermaciej>
why do people keep bringing up the idea of tunneling SPDY over WebSocket?

[04:07:00.0000] <othermaciej>
o_O

[04:16:00.0000] <Lachy>
MikeSmith, thank you! I've always hated having public-html spammed with bug mail

[04:19:00.0000] <othermaciej>
MikeSmith: let me know when the coast is clear please

[04:26:00.0000] <Lachy>
/me realises how even more seriously annoying it is to be spammed with 500 duplicate bug mails each on both public-html and public-html-bugzilla, on the same day my mail server seems to be responding rather slowly :-(

[05:25:00.0000] <zcorpan>
Hixie: should the spec say to fail the connection if the server sends a binary frame with a length that would cause a buffer overflow (say we store the length in a 32bit int)

[09:47:00.0000] <TabAtkins>
annevk: I was going to ask around today.  I'll give you a headsup on the thread later.

[10:43:00.0000] <franksalim>
othermaciej, re:SPDY over WebSocket, is that a rhetorical question?

[10:54:00.0000] <AryehGregor>
The real problem with the bug spam isn't the volume, it's that now I don't know which bugmail is actually meaningful, what with Gmail's conversation grouping . . .

[11:06:00.0000] <Philip`>
AryehGregor: Open your mail folder in an IMAP client, sort messages by date, select the ones in the offending time period, mark as read

[11:06:01.0000] <AryehGregor>
That sounds like more effort than just hitting ] a few hundred times.

[11:06:02.0000] <Dashiva>
Does it really

[11:06:03.0000] <AryehGregor>
(assuming Gmail doesn't decide that that should take ten seconds on a regular basis, as it is sometimes wont to do)

[11:07:00.0000] <AryehGregor>
Well, I've never used an IMAP client in my life.

[11:07:01.0000] <AryehGregor>
I used AOL up to 2005, and Gmail since then.

[11:07:02.0000] <jgraham>
Umm

[11:08:00.0000] <jgraham>
Wow

[11:08:01.0000] <gsnedders>
AryehGregor: Are you a masochist or something?

[11:08:02.0000] <Philip`>
It's fairly trivial to configure Thunderbird to access your Gmail account

[11:08:03.0000] <AryehGregor>
gsnedders, nope.  Webmail is convenient, why should I use a real mail client?

[11:08:04.0000] <AryehGregor>
I'm not clear on the benefits.

[11:08:05.0000] <Philip`>
Apparently webmail is inconvenient in this particular situation

[11:09:00.0000] <AryehGregor>
Well, one particular design choice of Gmail (which I mostly like a lot) is inconvenient in this particular situation.

[11:09:01.0000] <AryehGregor>
(I mean, I mostly like this design choice a lot, threaded viewing)

[11:10:00.0000] <AryehGregor>
Things like this seem to come up rarely.

[11:10:01.0000] <Philip`>
It's good to have an alternative for the rare inconvenient situations

[11:10:02.0000] <Philip`>
and continue to use webmail when it's good enough

[11:11:00.0000] <Dashiva>
The nice thing about IMAP is that you can have both

[11:11:01.0000] <AryehGregor>
Well, I suppose I technically do.  There's just never been a situation so inconvenient that I felt bothered to set it up.

[11:11:02.0000] <AryehGregor>
Maybe I should set up a mail client to at least download all my mail, though.

[11:11:03.0000] <jgraham>
AryehGregor: You know you can use a non-web client and a web client together on the smae mailbox, right?

[11:11:04.0000] <AryehGregor>
Yes, yes, I know.

[11:11:05.0000] <AryehGregor>
I've never seen the need.

[11:58:00.0000] <Hixie>
AryehGregor: do you know how to embed an external image on a wikimedia wiki?

[11:59:00.0000] <AryehGregor>
Hixie, you don't.  You can't include any external content for privacy reasons.

[11:59:01.0000] <AryehGregor>
IP sniffing and such.

[11:59:02.0000] <Hixie>
Lachy_: ah, indeed, nevermind then, just my poor memory :-)

[12:00:00.0000] <AryehGregor>
MediaWiki supports it (although kludgily), it's just disabled by default and on Wikimedia.

[12:00:01.0000] <Hixie>
AryehGregor: hm

[12:00:02.0000] <AryehGregor>
Did you mean Wikimedia or MediaWiki?

[12:00:03.0000] <Hixie>
oh i meant mediawiki

[12:00:04.0000] <AryehGregor>
Oh, right.

[12:00:05.0000] <Hixie>
didn't know there was a difference

[12:01:00.0000] <AryehGregor>
Different story then.

[12:01:01.0000] <Hixie>
ah, gotta go.brb

[12:01:02.0000] <Hixie>
wanted to add them to our subtitles page

[12:01:03.0000] <AryehGregor>
The names were chosen before Wikimedia had enough money to consider thinking seriously about stuff like branding.

[12:01:04.0000] <AryehGregor>
So, well, they're terrible.

[12:01:05.0000] <AryehGregor>
Wikimedia is the charitable foundation that runs Wikipedia, MediaWiki is the name of the software.

[12:02:00.0000] <AryehGregor>
Anyway, the way MW supports external images last I checked is that if you enable the feature, then bare links that look like image URLs are turned into <img> instead of <a>, basically.

[12:02:01.0000] <Philip`>
Is there a Mikiwedia too?

[12:02:02.0000] <AryehGregor>
Which doesn't allow specifying any attributes or anything.

[12:02:03.0000] <AryehGregor>
Philip`, apparently.  http://www.o--o.info/

[12:03:00.0000] <daedb>
Can't you just whitelist img elements?

[12:03:01.0000] <AryehGregor>
Well, yes, if you hack the code.  The element whitelist isn't configurable as it stands.

[12:03:02.0000] <AryehGregor>
(I'm not sure why not.)

[12:04:00.0000] <daedb>
oh, I thought I had read something a while back about that, but I must be misremembering then :)

[12:05:00.0000] <AryehGregor>
I could make it configurable, I suppose.

[12:05:01.0000] <AryehGregor>
The WHATWG wiki is already running off a development version, I could upgrade to a more recent one.

[12:06:00.0000] <AryehGregor>
(hopefully header processing hasn't changed again, or hsivonen will get annoyed again)

[12:09:00.0000] <othermaciej>
franksalim: it just seems like the opposite of the layering that makes sense

[12:12:00.0000] <Hixie>
back briefly, sorry for running off

[12:13:00.0000] <Hixie>
AryehGregor: is there something we can do to make it possibel for me to embed images from around the web on http://wiki.whatwg.org/wiki/Timed_tracks ?

[12:13:01.0000] <AryehGregor>
Hixie, I'm writing a patch right now to just allow whitelisting <img>.  That seems simplest.

[12:13:02.0000] <AryehGregor>
When I've committed that, I can update to latest trunk, fix any fallout, and then whitelist <img>.

[12:13:03.0000] <AryehGregor>
If someone reverts my commit, well, I'll deal with that.  :)

[12:14:00.0000] <Hixie>
sweet

[12:14:01.0000] <Hixie>
so in the meantime i just stick <img src="" alt=""> in the markup?

[12:17:00.0000] <AryehGregor>
I guess so.

[12:17:01.0000] <Hixie>
cool, thanks man, you rock

[12:19:00.0000] <AryehGregor>
This is a pretty horrendous interface.  I can imagine there might be objections to exposing it as an API we have to maintain.

[12:19:01.0000] <AryehGregor>
Hmm.

[12:20:00.0000] <AryehGregor>
Maybe a more targeted patch would be a better idea.

[12:38:00.0000] <AryehGregor>
Okay, feature added: http://www.mediawiki.org/wiki/Special:Code/MediaWiki/65286

[12:38:01.0000] <AryehGregor>
Now let's see if updating breaks anything exciting.

[12:42:00.0000] <AryehGregor>
Hixie, seems to be working.

[12:43:00.0000] <AryehGregor>
(I'm not sure why you don't upload those, though.  They look like they're likely to bitrot as-is.)

[12:57:00.0000] <AryehGregor>
/me ponders configurably whitelisting some new-to-HTML5 tags too

[13:39:00.0000] <Hixie>
AryehGregor: thanks

[13:39:01.0000] <Hixie>
AryehGregor: not uploading them because the lifetime of this page is going to be about 3 days

[14:03:00.0000] <Hixie>
well

[14:03:01.0000] <Hixie>
so far i haven't found any subtitles that do anything fancy at _all_

[14:03:02.0000] <Hixie>
not even italics!

[14:03:03.0000] <Hixie>
i must look harder

[14:07:00.0000] <AryehGregor>
I've noticed that in computer games, subtitles also rarely do anything fancy.  Most games don't even use italics or bold.

[14:07:01.0000] <AryehGregor>
Some actually even resort to *asterisks* for emphasis.

[14:08:00.0000] <Hixie>
heh

[14:08:01.0000] <Hixie>
amusingly irssi made your asterisks into bold :-P

[14:09:00.0000] <Philip`>
http://philip.html5.org/misc/portal-caption.jpg has colours and italics

[14:09:01.0000] <Philip`>
but only per-line styles, I think

[14:10:00.0000] <AryehGregor>
/me almost said "If subtitles are intended for blind people, it's not surprising", but remembered just in time that that makes absolutely no sense

[14:11:00.0000] <AryehGregor>
Philip`, it styles them according to speaker, I think, no?  Or something like that.  The pink italics is GLaDOS.

[14:11:01.0000] <AryehGregor>
Interesting approach.

[14:11:02.0000] <AryehGregor>
That way different speakers have distinct appearances, without having to tell you their name.

[14:12:00.0000] <Philip`>
AryehGregor: Well, there's only one speaking character in the game

[14:12:01.0000] <Philip`>
so it doesn't make much difference

[14:12:02.0000] <AryehGregor>
Not true!

[14:12:03.0000] <AryehGregor>
The robots speak.

[14:12:04.0000] <AryehGregor>
"I'm sorry."

[14:12:05.0000] <Philip`>
Oh, I suppose that's true

[14:12:06.0000] <AryehGregor>
. . . I guess that's all, though.

[14:13:00.0000] <Philip`>
They're more like sound effects than real speech

[14:13:01.0000] <AryehGregor>
How so?

[14:14:00.0000] <Philip`>
You could replace them with something like a whirring effect and it would make no real difference

[14:14:01.0000] <daedb>
/me has at least one file with subtitles that is in two languages (simultaneously), use different colours, and has at least one part with a different font...

[14:15:00.0000] <Philip`>
http://philip.html5.org/misc/subtitles-3.xml uses colours to differentiate speakers within a single line occasionally

[14:23:00.0000] <KaOSoFt>
Good morning.

[14:25:00.0000] <KaOSoFt>
Did you people know of this? http://apirocks.com/html5/html5.html

[14:28:00.0000] <AryehGregor>
KaOSoFt, to begin with, slide 3 is grossly wrong.

[14:28:01.0000] <AryehGregor>
HTML5 started in 2004, for practical purpose.s

[14:28:02.0000] <AryehGregor>
s.

[14:29:00.0000] <AryehGregor>
"Tableless Web Design" and "AJAX" also don't belong on the same list as actual standards.

[14:29:01.0000] <AryehGregor>
Some form of HTML predated 1991, too, AFAIK.

[14:30:00.0000] <AryehGregor>
I'm pretty sure not all the relations listed in slide 18 are new to HTML5, although I could be wrong.

[14:30:01.0000] <AryehGregor>
Slide 21 lists some things as unimplemented that are implemented, at least by Opera.

[14:31:00.0000] <AryehGregor>
The CSS stuff has nothing to do with HTML5, even if you construe it pretty broadly, unless by "HTML5" you mean "any vaguely recent development in web technology".

[14:31:01.0000] <AryehGregor>
That's just at a glance.

[14:35:00.0000] <jgraham>
(if this is the page I think it is, the fineprint somewhere says that it is only really about Chrome. Which is a pretty crappy thing to bury in the fineprint of a presentation nominally about a cross-vendor standard with multiple implementations)

[14:39:00.0000] <jgraham>
(where "only really about" means "the demos - which are basically the whole point of the presentation - have only been written to work in Chrome even when trivial modifications would allow them o work in more browsers)

[14:39:01.0000] <jgraham>
s/)/")/

[14:44:00.0000] <TabAtkins>
Slide 19 uses @item - this was changed to @itemscope after the usability testing.

[14:57:00.0000] <AryehGregor>
Thought that looked vaguely wrong.

[15:06:00.0000] <Hixie>
AryehGregor: I tried to put "<style scoped>img { max-width: 350px; }</style>" into a wiki page, but that didn't work... is there some way to set styles on a per-page basis?

[15:06:01.0000] <Hixie>
Philip`: thanks for that image, that's great

[15:12:00.0000] <daedb>
Hixie: Use the class that exists on the body element for that page?

[15:12:01.0000] <Hixie>
how do i do that?

[15:13:00.0000] <daedb>
In the global stylesheet, put in .page-pagename img { etc }

[15:14:00.0000] <Hixie>
sorry i'm a complete wikimedia n00b

[15:14:01.0000] <Hixie>
oh i can edit the global css?

[15:14:02.0000] <daedb>
http://wiki.whatwg.org/wiki/MediaWiki:Monobook.css

[15:14:03.0000] <daedb>
You can add styles there ^^

[15:14:04.0000] <Hixie>
nice

[15:14:05.0000] <daedb>
And every page has the pagename in a class on the body ;)

[15:15:00.0000] <Hixie>
yeah, just found it

[15:15:01.0000] <Hixie>
you rock, thanks

[15:15:02.0000] <Hixie>
sweet, it worked

[15:16:00.0000] <daedb>
Every user account has its own css file in Mediawiki too, which is really fun to play with :)

[15:17:00.0000] <Hixie>
cool

[15:19:00.0000] <TabAtkins>
+1 for <video>!  Switching youtube to <video> makes the audio actually work, instead of sounding like Max Headroom using the flash player.

[15:19:01.0000] <TabAtkins>
(on my goobuntu box)

[15:19:02.0000] <daedb>
http://wiki.whatwg.org/wiki/MediaWiki:Common.css is the really really global one btw. Monobook.css is for the monobook skin only.

[15:21:00.0000] <Hixie>
daedb: k

[15:55:00.0000] <Hixie>
i've put a bunch of examples in http://wiki.whatwg.org/wiki/Timed_tracks

[15:57:00.0000] <Hixie>
i've yet to find a single example that needs anything more than plain text and line feeds, the ability to have multiple distinct voices, some vague positioning, karaoke, and ruby

[16:00:00.0000] <TabAtkins>
It appears that removing most of the styling decisions from people would be a great improvement, if these examples are at all representative (and I'm betting they are...).

[16:00:01.0000] <Philip`>
/me notes that the Portal one isn't really timed tracks, since it's interactive rather than playing back pre-prepared sequences

[16:08:00.0000] <Philip`>
Hixie: http://philip.html5.org/misc/eva-captions.jpg is one with seemingly more important positioning requirements

[16:09:00.0000] <Hixie>
Philip`: wow, that's quite the trainwreck :-)

[16:11:00.0000] <Philip`>
Hixie: Why so? :-)

[16:12:00.0000] <Philip`>
It's just translating the text that's on the video, in a layout that roughly matches the video, plus the audio subtitles

[16:12:01.0000] <Hixie>
multiple fonts, three different kinds of text alignment, including one block of text with two different alignments... i'd have to see the video to tell how bad it really was in context

[16:13:00.0000] <Hixie>
but i've added it to the pile and changed the requirements section to match :-)

[16:14:00.0000] <TabAtkins>
Hixie: All but the bottom text is meant to more-or-less copy the alignment of the hand-written text on the picture, since it's a translation of that.

[16:14:01.0000] <Hixie>
oh i see what it's doing, but that doesn't make it any less of a trainwreck :-)

[16:15:00.0000] <Hixie>
a lot of these are pretty bad

[16:15:01.0000] <Hixie>
like the green subtitles in one of the examples

[16:15:02.0000] <Hixie>
ew

[16:15:03.0000] <TabAtkins>
Yeah, that's horrifying.

[16:15:04.0000] <TabAtkins>
I think it's outlined in red, too?

[16:15:05.0000] <Hixie>
or the second one in the "ruby" examples, though that one might have been ok before compression artefacts had their go at it

[16:15:06.0000] <Philip`>
Hixie: The video is http://www.youtube.com/watch?v=xG9KluukpJI at 2:43

[16:15:07.0000] <Philip`>
but that version has totally different text

[16:16:00.0000] <Hixie>
heh

[16:16:01.0000] <Hixie>
ok the start of that video is pretty insane too

[16:17:00.0000] <Philip`>
(The version I've got has the English translation of on-screen text baked into the video, plus an English subtitle track which translates a Japanese audio track, plus an English audio track which uses a different translation)

[16:17:01.0000] <Hixie>
between the subtitles, the subtitle credit, the main credit, and the youtube ad, i feel assaulted

[16:18:00.0000] <Hixie>
ooh, that video's 2:43 is interesting for another reason

[16:18:01.0000] <Hixie>
it adds another subtitle above the first one

[16:18:02.0000] <Hixie>
i wonder if that's an example of multiple voices or of positioning or something else

[16:18:03.0000] <Philip`>
(...using DivX video and Vorbis audio in an OGM container)

[16:20:00.0000] <Hixie>
actually i guess that should happen automatically just by having two subtitles with overlapping segments


2010-04-20
[17:13:00.0000] <Hixie>
nessy: yt?

[17:13:01.0000] <nessy>
starting to wake up...

[17:14:00.0000] <Hixie>
you wouldn't happen to be in the bay area in the near future by any chance would you?

[17:14:01.0000] <nessy>
so, yes :)

[17:14:02.0000] <nessy>
nothing planned, no

[17:14:03.0000] <nessy>
why?

[17:14:04.0000] <Hixie>
k

[17:14:05.0000] <Hixie>
i'm starting to look at this subtitles stuff

[17:14:06.0000] <Hixie>
if you were around i would have suggested lunch so i could get a brain dump

[17:14:07.0000] <Hixie>
but no worries

[17:15:00.0000] <Hixie>
i can read also :-)

[17:15:01.0000] <nessy>
I've got  to reply to your request and to henry's proposal etc

[17:15:02.0000] <nessy>
got lots to do - was just swamped with other stuff this week

[17:15:03.0000] <nessy>
(last week, really)

[17:15:04.0000] <nessy>
but shoot any questions you have

[17:15:05.0000] <Hixie>
i've 192 e-mails to read through on the subject (mostly to whatwg; most of the public-html-a11y ones i read as they passed and didn't save)

[17:16:00.0000] <Hixie>
so i'm sure i'll have questions :-)

[17:16:01.0000] <nessy>
best intro is probably http://blog.gingertech.net/2010/04/11/introducing-media-accessibilit-into-html5-media/

[17:16:02.0000] <nessy>
gives a background

[17:16:03.0000] <nessy>
most discussions on whatwg were about the caption format - ppl don't like DFXP/TTML

[17:17:00.0000] <nessy>
so mostly about developing a new format (microformats story all over?)

[17:17:01.0000] <Hixie>
cool, will read

[17:17:02.0000] <nessy>
won't go well with the a11y ppl, but I don't really care what format we use - I think it will be many over time

[17:18:00.0000] <Hixie>
yeah if we can avoid inventing yet another format that would certainly be good

[17:18:01.0000] <nessy>
are you starting to dig into it now?

[17:18:02.0000] <Hixie>
yeah

[17:18:03.0000] <nessy>
bummer - I wanted to give you a list of samples and use cases

[17:18:04.0000] <nessy>
will try and get to that today then

[17:18:05.0000] <Hixie>
oh it'll take me a while i'm sure

[17:18:06.0000] <Hixie>
no rush

[17:19:00.0000] <Hixie>
at this point i'm mostly just trying to ground my sense of what use cases are real and what are hypothetical

[17:19:01.0000] <Hixie>
by looking at what people actually do with subtitles

[17:19:02.0000] <Hixie>
for example i thought that there was a need to handle individual words being emphasised, but i've yet to find a single example of that in the real world

[17:19:03.0000] <Hixie>
though i have found two examples of ruby!

[17:20:00.0000] <Hixie>
i can't say i was ultra-pleased by that, given how ruby is like 10 times harder to deal with than italics :-P

[17:21:00.0000] <nessy>
most styled captions are commercial stuff

[17:21:01.0000] <nessy>
I'll see if I can find you some examples

[17:21:02.0000] <Hixie>
that would be awesome

[17:34:00.0000] <nessy>
with the use cases - are you ok if I take some screenshots of commercial stuff that may not be generally accessible, but a good example?

[17:35:00.0000] <Hixie>
yeah totally

[18:01:00.0000] <daedb>
Hixie: Dragon Age: Origins has individual words emphasized (in bold), if video games count as examples :)

[18:08:00.0000] <Hixie>
daedb: sure, take screenshots and upload them to the wiki page :-)

[18:10:00.0000] <daedb>
I'll try to remember that next time I see it (or see if Google can find a relevant screenshot)... it has italics too, btw, for character actions (like sighing) :p

[18:59:00.0000] <MikeSmith>
about http://www.w3.org/Bugs/Public/show_bug.cgi?id=7688

[18:59:01.0000] <MikeSmith>
what's POLA?

[19:00:00.0000] <Hixie>
something of least something?

[19:00:01.0000] <Hixie>
i've been playing Bad Company 2 too much. I just got spam about some sort of digital camera and the subject line mentioned 5x optical zoom and my brain instantly junked the e-mail on the basis that anything lower than 12x wasn't worth looking at

[19:01:00.0000] <doublec>
principle of least astonishment

[19:02:00.0000] <nessy>
Hixie: can I also add examples of other time-aligned text to that wiki page?

[19:02:01.0000] <Hixie>
sure

[19:02:02.0000] <nessy>
e.g. chapter markers, linguistic annotation 

[19:02:03.0000] <nessy>
ok, cool

[19:02:04.0000] <nessy>
and: yes, textual audio descriptions are just text with time markers

[19:03:00.0000] <MikeSmith>
doublec: thanks

[19:03:01.0000] <MikeSmith>
and an honorable mention to Hixie for "something of least something"

[19:35:00.0000] <nessy>
Hixie: are you sure the whatwg wiki lets me upload images?

[19:36:00.0000] <nessy>
I can't find a means and get the feeling that it's disabled in the mediawiki setup

[19:36:01.0000] <nessy>
also, http://wiki.whatwg.org/wiki/Special:ListFiles is empty

[19:41:00.0000] <Hixie>
nessy: oh you'll have to upload them elsewhere first

[19:41:01.0000] <nessy>
ah I see ...

[19:41:02.0000] <Hixie>
nessy: if you need somewhere to upload them use http://software.hixie.ch/utilities/cgi/uploader/uploader

[19:42:00.0000] <Hixie>
password is w643kJ6Gv43q3

[19:42:01.0000] <nessy>
ah, cool, thanks!

[19:55:00.0000] <nessy>
Hixie: good reading - http://main.wgbh.org/wgbh/pages/mag/services/captioning/faq/sugg-styles-conv-faq.html ;-)

[20:14:00.0000] <nessy>
I think I'm finding it hard to provide an example of musical notes in captions, since all online systems failed to render musical notes

[20:15:00.0000] <nessy>
I keep coming across ppl complaining about it ;-)

[20:15:01.0000] <nessy>
it's kinda ironic!

[20:23:00.0000] <nessy>
yay, found one!

[21:19:00.0000] <nessy>
Hixie: you'd enjoy watching http://svg-wow.org/audio/animated-lyrics.html

[21:23:00.0000] <MikeSmith>
annevk: is Daniel around?

[21:24:00.0000] <annevk>
sitting next to me

[21:24:01.0000] <annevk>
he's also on skype

[21:24:02.0000] <annevk>
I just told him about it, seems he's ok

[21:24:03.0000] <MikeSmith>
great

[21:24:04.0000] <annevk>
could discuss it more over lunch

[21:24:05.0000] <MikeSmith>
yep

[22:51:00.0000] <boblet>
Hixie: yt?

[23:53:00.0000] <annevk>
so TTML does not have ruby support it seems

[23:53:01.0000] <annevk>
already falls short of catering to in the wild use of captioning

[23:53:02.0000] <annevk>
while being fricking complex

[23:54:00.0000] <annevk>
genius

[23:56:00.0000] <hsivonen>
is ruby actually used for captioning in the wild?

[23:57:00.0000] <annevk>
http://wiki.whatwg.org/wiki/Timed_tracks#Ruby

[23:57:01.0000] <hsivonen>
which side of 80/20 inside locales that use ruby for anything at all?

[23:58:00.0000] <annevk>
it seems that in Japanese it would be required as otherwise meaning would be ambigious

[23:58:01.0000] <othermaciej>
that is some crazy exmples

[23:58:02.0000] <othermaciej>
*examples

[23:59:00.0000] <annevk>
the page is fascinating

[00:01:00.0000] <hsivonen>
well, at least the most complex cases aren't for accessibility

[00:01:01.0000] <hsivonen>
more like gratuitous "just because we can" for translation or karaoke

[00:01:02.0000] <hsivonen>
(I don't mean ruby here. I mean stuff on the page in general)

[00:04:00.0000] <hsivonen>
a lot of the plain text translations look very ugly cmpared to what we get here on tv. and I thought what we have wasn't pretty

[00:12:00.0000] <MikeSmith>
ruby annotations are used quite commonly in Japanese subtitles for films and TV programs and such (not just for karaoke or whatever other edge cases)

[00:14:00.0000] <hsivonen>
ok. seems like a FAIL for TTML then

[00:14:01.0000] <MikeSmith>
yeah

[00:16:00.0000] <MikeSmith>
a lot of the other Japanese caption examples there are not things I at least typically see in normal use

[00:16:01.0000] <hsivonen>
what we get is pretty readable comapred to many of the Latin examples: http://hsivonen.iki.fi/screen/subtitles.jpg

[00:17:00.0000] <MikeSmith>
interesting that that uses stroking

[00:21:00.0000] <annevk>
is CNN not an example of somewhat complex inline formatting btw?

[00:21:01.0000] <nessy>
what is required to support ruby?

[00:21:02.0000] <annevk>
ruby markup

[00:22:00.0000] <nessy>
/me goes google

[00:22:01.0000] <annevk>
and some CSS styling if you want to place it elsewhere prolly

[00:22:02.0000] <nessy>
i18n is so fascinating!

[00:22:03.0000] <annevk>
kind of weird the i18n group hasn't catched that with TTML

[00:23:00.0000] <annevk>
maybe they don't consider it a requirement

[00:23:01.0000] <nessy>
I'm not sure TTML is even properly finalised yet

[00:24:00.0000] <annevk>
it's in CR

[00:24:01.0000] <annevk>
which means it passed Last Call review

[00:24:02.0000] <nessy>
so there is a <ruby> tag in HTML?

[00:24:03.0000] <annevk>
yeah

[00:25:00.0000] <nessy>
cool - I didn't follow those discussions

[00:25:01.0000] <annevk>
http://www.whatwg.org/html#the-ruby-element

[00:31:00.0000] <othermaciej>
MikeSmith: btw did you change the bugzilla mail setup for public-html?

[00:32:00.0000] <hsivonen>
MikeSmith: subtitles ha used white text with black outline for decades over here.

[00:32:01.0000] <Hixie>
nessy: i'm removing the cnn example because it doesn't seem to be synchronised with the video -- let me know if it is in fact synchronised

[00:33:00.0000] <hsivonen>
s/ha/have/

[00:33:01.0000] <nessy>
let me check

[00:33:02.0000] <Hixie>
do you have any more info on the other ticker example?

[00:33:03.0000] <Hixie>
is it just reproducing the subtitle below the video?

[00:33:04.0000] <Hixie>
or is it something else?

[00:34:00.0000] <nessy>
oh, that was the one under ticker text?

[00:34:01.0000] <hsivonen>
isn't ticker by definition something else?

[00:34:02.0000] <nessy>
it is

[00:35:00.0000] <nessy>
the TV example (CNN) I wanted to use as an example how it shows on TV

[00:35:01.0000] <annevk>
but it comes with the video and should supposedly be exposed in some way...

[00:35:02.0000] <nessy>
but if you only want current web examples, that's fine

[00:35:03.0000] <MikeSmith>
othermaciej: yeah, I set the user prefs on the public-html⊙wo bugzilla account to disable all e-mail for the account

[00:35:04.0000] <nessy>
so the one that is currently under "ticker text" has both, captions AND ticker text

[00:35:05.0000] <hsivonen>
annevk: <marquee>

[00:35:06.0000] <Hixie>
i'm just looking for timed tracks, doesn't have to be web-only -- in fact, non-web is more interesting

[00:36:00.0000] <Hixie>
but the cnn tickers seem to just be a continuous unsynchronised stream of information that happens to be overlaid, you wouldn't want to use the same mechanism for those as for subtitles

[00:36:01.0000] <annevk>
hsivonen, that means you need to deploy different video streams

[00:36:02.0000] <nessy>
you will notice underneath the video a line of "NOW: Roubini: new regulations...."

[00:36:03.0000] <othermaciej>
MikeSmith: thanks!

[00:36:04.0000] <nessy>
that is the ticker text

[00:36:05.0000] <Hixie>
nessy: right but that's the same as the subtitle

[00:36:06.0000] <nessy>
just by chance

[00:36:07.0000] <Hixie>
nessy: is there somewhere i can experiment with that example in the wild?

[00:36:08.0000] <Hixie>
i don't really understand what it is showing

[00:36:09.0000] <nessy>
also, I am not sure if that is actually a subtitle or "burnt-in" text

[00:37:00.0000] <nessy>
there are multiple javascript libraries for doing ticker text

[00:37:01.0000] <hsivonen>
I wouldn't count newscast title of the story as a subtitle

[00:37:02.0000] <nessy>
let me see if I can track that one down

[00:37:03.0000] <hsivonen>
just like I wouldn't count mivie titles

[00:37:04.0000] <nessy>
hsivonen I don't think that's what that is - it's just intermediate

[00:38:00.0000] <othermaciej>
hmm

[00:38:01.0000] <othermaciej>
it looks like someone finished the mass move

[00:38:02.0000] <othermaciej>
that or bugzilla lied to me about the 235 bugs left

[00:38:03.0000] <Hixie>
the chapter stuff is interesting... wonder what to conclude from it

[00:39:00.0000] <Hixie>
there's such a wide variety of ways it is exposed that it might make more sense to just support it in the sense of exposing the data in the API but not doing anything with it otherwise

[00:40:00.0000] <hsivonen>
nessy: ok. I wouldn't count the cable news summary of the last couple of minutes, either

[00:40:01.0000] <nessy>
yeah, I was wondering about the chapters

[00:40:02.0000] <nessy>
if they are given inside the file, you will only get them successively normally

[00:40:03.0000] <nessy>
but for external tracks it is easy to grab them all in one go and display e.g. as a bar at the bottom to click directly to

[00:41:00.0000] <Hixie>
well certainly we shouldn't disallow the UA from showing UI for them, yeah

[00:42:00.0000] <Hixie>
annotations also seem like something we should probably leave to sites to figure out, giving them just an API so they don't have to do the timing themselves

[00:43:00.0000] <Hixie>
basically let them give us a file with the annotation data then have them hook into notifications for when to render each one

[00:44:00.0000] <annevk>
wouldn't they handle the file as well then?

[00:45:00.0000] <annevk>
otherwise you might as well add time data to the file format...

[00:46:00.0000] <Hixie>
?

[00:46:01.0000] <hsivonen>
fwiw, the reason why I wouldn't count titles and summary of the past couple of minutes is that they are customarily considered to be under the rule of the art department of the show

[00:46:02.0000] <nessy>
Hixie, for ticker text I am thinking of something like at http://www.scriptocean.com/dticker.html with time markers synchronised to the video

[00:47:00.0000] <Hixie>
does anyone do that?

[00:47:01.0000] <hsivonen>
so even if we had an elaborate mechanism for them, chances are that the art dept would still want to have full control

[00:47:02.0000] <hsivonen>
and treat them as part of the image

[00:47:03.0000] <Hixie>
i mean, as hsivonen says, that kind of thing would just be a title in most cases and so burnt in

[00:47:04.0000] <zcorpan>
why is http://dev.w3.org/html5/core/ not a 410 Gone?

[00:47:05.0000] <hsivonen>
and they are data on their own right as opposed to being a synchronized alternative of other data

[00:47:06.0000] <nessy>
Hixie: are you referring to me?

[00:47:07.0000] <Hixie>
nessy: you and hsivonen yeah

[00:48:00.0000] <zcorpan>
http://www.theopensourcery.com/keepopen/?p=2205 points to it

[00:48:01.0000] <nessy>
no, the text is not normally burnt in - only on TV

[00:48:02.0000] <hsivonen>
Hixie: to me, that kind of thing seems like something you'd put on the surrounding HTML page

[00:48:03.0000] <nessy>
it is an example of text that comes in at a certain time and leaves at a different

[00:48:04.0000] <hsivonen>
Hixie: not logically part of the video

[00:49:00.0000] <annevk>
zcorpan, you could commit a Redirect 410 /html5/core/ .htaccess line I suppose

[00:49:01.0000] <nessy>
possibly - but then there are definitely times when ticker text is synchronised to the video

[00:49:02.0000] <Hixie>
hsivonen: it's something i'd expect to see if i uploaded the video to my ipod or played it to my tv, so i'd expect it as part of the video data.

[00:49:03.0000] <annevk>
zcorpan, not entirely sure if Redirect was part of the syntax, but I thought it was

[00:49:04.0000] <zcorpan>
annevk: oh right i have cvs access

[00:50:00.0000] <Hixie>
nessy: would you agree that movie credits are text that comes in at a certain time and leaves at a different time but should be burnt in?

[00:50:01.0000] <hsivonen>
Hixie: in that case, I'd treat it as nice to have but not as a core requirement for the Web

[00:50:02.0000] <nessy>
it has very similar properties to captions and subtitles, which is why I think we can have a generic interface to deal with all of them

[00:50:03.0000] <hsivonen>
Hixie: uploading to iPod is not part of the Web mindset

[00:50:04.0000] <zcorpan>
annevk: actually i don't have access to dev.w3.org

[00:50:05.0000] <nessy>
movie credits are part of the video production process, therefore burnt in

[00:50:06.0000] <hsivonen>
Hixie: you are supposed to navigate to the original page using Mobile Safari on iPod Touch

[00:50:07.0000] <Hixie>
hsivonen: it's open content, i should be able to do whatever i like with it

[00:51:00.0000] <Hixie>
hsivonen: screw that :-P

[00:51:01.0000] <Hixie>
hsivonen: it's video, i want to treat it as video.

[00:51:02.0000] <hsivonen>
Hixie: depends on what kind of rathole of requirements come with your wishes

[00:51:03.0000] <Hixie>
nessy: i would expect the kind of thing i see at 

[00:51:04.0000] <nessy>
well, all external but linked text is either closer or less closely related to the video

[00:51:05.0000] <Hixie>
nessy: ... http://www.scriptocean.com/dticker.html to be part of the video production process also

[00:52:00.0000] <nessy>
but they are all synchronised to it, so they can all be treated the same way

[00:52:01.0000] <nessy>
no, ticker text is not part of the video production process - it is created by the publisher normally, not by the producer

[00:53:00.0000] <Hixie>
well the only time i've ever seen such ticker text is in news shows, where it's definitely part of the production

[00:53:01.0000] <Hixie>
but if there are other examples i'm certainly eager to have my eyes opened :-)

[00:53:02.0000] <nessy>
just because the producer and publisher are the same ;)

[00:53:03.0000] <nessy>
I've seen movies on TV with ticker text that is definitely not created by the production of the movie

[00:54:00.0000] <Hixie>
but it's synchronised with the movie?

[00:54:01.0000] <nessy>
e.g. when 9-11 happened, there was plenty of ticker text running over the bottom of a movie

[00:54:02.0000] <Hixie>
sure but that isn't timed text

[00:54:03.0000] <Hixie>
it's just a ticker overlay

[00:54:04.0000] <Hixie>
has nothing to do with the video

[00:54:05.0000] <nessy>
roughly synchronised in that case only - synchronised between specific larger sections

[00:55:00.0000] <annevk>
zcorpan, done

[00:55:01.0000] <nessy>
as I said: text that runs in parallel with video or audio can be more or less tightly related to the video content

[00:55:02.0000] <nessy>
that changes nothing in the fact that it is running in parallel to the video

[00:55:03.0000] <annevk>
zcorpan, lets see if we get complaints o_O

[00:56:00.0000] <zcorpan>
nice

[00:57:00.0000] <nessy>
I do see your point about the difference between a newscast and move examples though, with the newscast the semantic link is stronger 

[00:57:01.0000] <Hixie>
nessy: semantically, it is the difference between something that runs as part of a <video> element, and something that runs as part of wholy unrelated javascript with wholy unrelated <div>s or <canvas> or ARIA live regions or whatnot, which just happens, because of some CSS, to be rendered over the video

[00:57:02.0000] <Hixie>
nessy: that's a huge difference

[00:57:03.0000] <Hixie>
nessy: since we're talking about what <video> should provide

[00:57:04.0000] <othermaciej>
is "synchronized with the video but over the top of it" distinct from "synchronized with the video but not on top of it?"

[00:58:00.0000] <Hixie>
othermaciej: not especially, no

[00:58:01.0000] <Hixie>
othermaciej: i mean, the former might be handled natively while the latter would likely need JS support, but that's all

[00:58:02.0000] <othermaciej>
is time-based relationship to what the video is currently showing the key semantic relation?

[00:58:03.0000] <nessy>
I guess the question is: will it be going along its own timeline or along the timeline of the video

[00:59:00.0000] <Hixie>
othermaciej: they both would need to be expressed in some timed track format

[00:59:01.0000] <Hixie>
nessy: right

[00:59:02.0000] <nessy>
even if it goes along its own timeline for most of it, there is definitely a link at some point

[00:59:03.0000] <othermaciej>
I'm thinking about the ever-popular "video of presentation with synchronized HTML slideshow below" example

[00:59:04.0000] <nessy>
you won't have a ticker text on a video that starts ticking along without the video playing

[01:00:00.0000] <Hixie>
nessy: not for the 9/11 subtitles. they would have shown those subtitles irrespective of what movie was showing.

[01:00:01.0000] <nessy>
not if the video was paused

[01:00:02.0000] <Hixie>
well there's no way to pause it

[01:00:03.0000] <Hixie>
since it's all burnt in in that case

[01:00:04.0000] <nessy>
I'm talking about having a video like a youtube video on a site

[01:00:05.0000] <nessy>
TV is different

[01:01:00.0000] <Hixie>
if youtube had breaking-news tickers, they wouldn't pause when you pause the video.

[01:01:01.0000] <nessy>
if it's burnt-in it's none of our business anyway

[01:01:02.0000] <Hixie>
burnt-in subtitles tell us what people need on the web

[01:01:03.0000] <Hixie>
it's entirely our business

[01:01:04.0000] <Hixie>
it's what i'm using to evaluate the proposals :-)

[01:01:05.0000] <nessy>
if I go to a youtube page and the video sits there paused, I don't think I would expect the ticker text to be running

[01:02:00.0000] <nessy>
yes, TV gets you some way there, but we have a lot more flexibility on the Web ;)

[01:02:01.0000] <othermaciej>
YouTube has time-synchronized popover ads

[01:02:02.0000] <othermaciej>
but those don't change or go away, they just appear at a certain time

[01:02:03.0000] <boblet>
Hixie: I found out something interesting about ruby in Korean

[01:03:00.0000] <nessy>
I think we need to offer all that TV can do, but more :)

[01:03:01.0000] <boblet>
They’re kind of phasing kanji/hanzi out, and current usage is to have the hangul first, followed inline by the kanji in parentheses, so basically the opposite of the default display…

[01:04:00.0000] <Hixie>
nessy: i don't, fwiw :-) i think we should offer a small set of functionality in v1, and see what people feel they are truly lacking. baby steps and all.

[01:04:01.0000] <nessy>
that's fair enough

[01:04:02.0000] <boblet>
ruby associates rt with preceding non-rt ruby base text, and CSS3 Ruby doesn’t have this ruby-position, so currently impossible

[01:04:03.0000] <nessy>
we don't have to do ticker text in the first version

[01:04:04.0000] <Hixie>
i'm trying to find a web page that has tickers and video

[01:04:05.0000] <nessy>
I can, however, see a need

[01:04:06.0000] <Hixie>
but my network is acting poorly

[01:05:00.0000] <annevk>
boblet, CSS3 Ruby is somewhat obsolete as it is though and will be updated with HTML5 in mind

[01:05:01.0000] <nessy>
I found it difficult to find a video on the Web with ticker text actually, which is why that example came up

[01:05:02.0000] <nessy>
but let me look a bit more

[01:05:03.0000] <Hixie>
boblet: fwiw, "kanji/hanzi", "hangul", and "kanji" are all greek to me. :-)

[01:05:04.0000] <Hixie>
boblet: what does it mean in terms of html?

[01:06:00.0000] <boblet>
annevk: yeah, I read some of www-style on it. Interesting that the Korean usage is outside of CSS3 Ruby tho

[01:06:01.0000] <nessy>
for the craziest use of text with video ever got to http://www.masternewmedia.org/video_internet_television/video-annotation/online-video-annotation-takes-a-giant-leap-20070405.htm

[01:06:02.0000] <boblet>
Hixie: :)

[01:06:03.0000] <nessy>
I can find an example of almost anything you can think of on that page

[01:06:04.0000] <annevk>
boblet, if it's inline you don't need ruby markup though

[01:07:00.0000] <annevk>
boblet, you can just use normal typography :)

[01:07:01.0000] <annevk>
"normal" to be i18n-correct

[01:07:02.0000] <boblet>
Hixie: normally <ruby>base text <rt>ruby text</rt></ruby> to give ruby text above in supporting browsers or after in non-supporting browsers (default)

[01:07:03.0000] <boblet>
annevk: but that wouldn’t be *semantic* :P hehe

[01:07:04.0000] <othermaciej>
kanji and hanzi are the japanese and chinese names respecitvely for chinese ideographs

[01:08:00.0000] <othermaciej>
hangul is the korean alphabet

[01:08:01.0000] <annevk>
boblet, semantics are overrated

[01:08:02.0000] <boblet>
Hixie: current Korean usage is ruby text inline *before* base text, but old CSS3 ruby-position has every side but before

[01:08:03.0000] <boblet>
annevk: not to pedants

[01:09:00.0000] <annevk>
boblet, sure, but they're wrong

[01:10:00.0000] <annevk>
boblet, most of HTML is focused on pragmatism, not pure semantics

[01:10:01.0000] <boblet>
Hixie: also because no <rb> anymore source order is fixed it’s not possible to rearrange in source

[01:10:02.0000] <nessy>
ok, I found the Roubini example - I think what I thought was ticker text is actually an unrelated title div

[01:11:00.0000] <boblet>
(note that I’m not suggesting this is important enough to warrant <rb> as Korean is moving away from using anything that needs ruby)

[01:11:01.0000] <nessy>
Hixie: I don't mind about dropping ticker text from time-aligned text types and waiting till the market catches up

[01:11:02.0000] <boblet>
annevk: heh

[01:11:03.0000] <Hixie>
boblet: wait so koreans write <ruby><rt>ruby text</rt>base text</ruby> and it works in IE?

[01:11:04.0000] <Hixie>
nessy: k

[01:12:00.0000] <Hixie>
nessy: i'll remove that section for now

[01:12:01.0000] <nessy>
Hixie: after all, I waited for 10 years for the Web to catch up with my ideas of how hyperlinked video and captions should be done ;)

[01:12:02.0000] <nessy>
ok

[01:12:03.0000] <zcorpan>
http://dvcs.w3.org/hg/html/raw-file/tip/tests/submission/Microsoft/interface/anchor_href.htm - the framework seems to need unnecessary boilerplate and has the arguments backwards in assertEquals (compared to how i'd order them)

[01:12:04.0000] <boblet>
Hixie: I have yet to see a code example or make a test case

[01:13:00.0000] <nessy>
/me thinks there is a business idea in video ticker text offerings… ;)

[01:13:01.0000] <Hixie>
boblet: then i don't understand what you mean by "ruby text inline *before* base text"

[01:13:02.0000] <boblet>
Hixie: based on display of an image I was sent demonstrating current usage

[01:14:00.0000] <othermaciej>
zcorpan: I like the WebKit framework for script tests a *lot* better

[01:14:01.0000] <Hixie>
boblet: can you upload it so i can see?

[01:14:02.0000] <othermaciej>
both the output and the code there are icky

[01:14:03.0000] <boblet>
wip ;-)

[01:14:04.0000] <annevk>
I think boblet is saying they write "blah (blah)"

[01:14:05.0000] <annevk>
and therefore don't need ruby at all, but since it's called ruby pedants want to use ruby markup?

[01:14:06.0000] <boblet>
but the (blah) is what the blah would be in Japanese

[01:15:00.0000] <nessy>
haha, I just found a better example for ticker text, but also from Yahoo, see http://media.rgemonitor.com/images/blogs/10082008tt.jpg

[01:15:01.0000] <boblet>
annevk: there are some benefits in having reading semantically marked up you know. surely you’ve seen a few kanji by now? :P

[01:16:00.0000] <nessy>
and this: http://www.freewaregenius.com/wp-content/uploads/2009/11/Free-Stocks-Ticker-Screenshot.jpg

[01:16:01.0000] <nessy>
anyway - can all be done with javascript of course, so I don't really mind

[01:16:02.0000] <boblet>
Hixie: http://oli-studio.com/temp/ruby-korean.jpg

[01:16:03.0000] <Hixie>
nessy: the second one doesn't even have a video!

[01:17:00.0000] <nessy>
ah, you're right!

[01:17:01.0000] <Hixie>
boblet: ah well i'm happy to say that you can do that in text/plain

[01:17:02.0000] <Hixie>
nessy: i think that pretty much guarantees that i'm right that it's not linked to the video :-P

[01:17:03.0000] <boblet>
top line shows Japanese-style ruby, with ruby text above base text

[01:17:04.0000] <nessy>
oh, there are plenty of tickers that go without a video, that's for sure

[01:18:00.0000] <Philip`>
othermaciej: I think I got something slightly over 400 bug emails - was that the total number of bugs that you were changing (including the 235)?

[01:18:01.0000] <othermaciej>
Philip`: 493

[01:18:02.0000] <nessy>
but if no callbacks are required and all that is required is to start the thing on video playback and to stop it on video pause/end, then indeed the time-alignment is minimal

[01:19:00.0000] <boblet>
Hixie: true, but as mentioned there are linguistic benefits of indicating the phonetics of something. also if all we’re worried about with ruby is placement above it’s far more a presentational element than I thought :)

[01:19:01.0000] <Hixie>
nessy: i'm not even convinced there's that -- i

[01:19:02.0000] <Hixie>
nessy: ...i've yet to see a video with a ticker that isn't burnt in but that pauses when you pause the video

[01:19:03.0000] <nessy>
yeah, I saw :)

[01:19:04.0000] <annevk>
boblet, ruby markup makes sense for the top line in http://oli-studio.com/temp/ruby-korean.jpg but is not needed at all for the other lines

[01:19:05.0000] <nessy>
on the Web the tickers aren't bunt in, but they also may not stop, as you are saying, so same effect

[01:20:00.0000] <annevk>
boblet, I fully support ruby markup for examples similar to the top line (as used in Japan)

[01:20:01.0000] <MikeSmith>
boblet: that Korean markup case you mentioned is interesting (the logogram in parens after the phonetic reading, instead of the other way around)

[01:20:02.0000] <boblet>
hey Mike

[01:20:03.0000] <Hixie>
boblet: if all you want to do is render <ruby>bbb<rt>aaa</rt></ruby> as aaabbb then the answer is to use CSS (not ruby CSS, just regular CSS)

[01:21:00.0000] <Hixie>
boblet: though CSS doesn't yet support rendering things out of order like that, it's something the wg intends to provide eventually

[01:21:01.0000] <Hixie>
nessy: well if they don't stop with the video, they're definitely not timed tracks

[01:21:02.0000] <Hixie>
anyway

[01:21:03.0000] <nessy>
yeah, fair enough

[01:23:00.0000] <boblet>
Hixie: based on CSS3 Ruby module it seemed like that presentation slipped through the cracks. I’ll try to email about it tho, so hopefully it can be included in CSS3 Ruby mark 2

[01:24:00.0000] <Hixie>
boblet: it shouldn't be in the CSS ruby module

[01:24:01.0000] <Hixie>
boblet: it's not ruby presentation

[01:24:02.0000] <Hixie>
boblet: it's just regular inline rendering but with out of order flow

[01:25:00.0000] <boblet>
Hixie: then ruby-position: after also shouldn’t be there huh. ok that makes sense

[01:25:01.0000] <boblet>
which module would that end up in?

[01:25:02.0000] <Hixie>
what does 'ruby-position: after' do?

[01:25:03.0000] <Hixie>
no idea what module it should be in... TabAtkins or annevk might know

[01:25:04.0000] <boblet>
Hixie: according to that spec same rendering as non-supporting browser

[01:26:00.0000] <Hixie>
that seems like a waste of time then, they could have just used display:inline :-)

[01:26:01.0000] <boblet>
Hixie: so you read my article after all ;)

[01:26:02.0000] <boblet>
j/k, although I mentioned it in there

[01:43:00.0000] <boblet>
MikeSmith: thanks for the feedback

[02:05:00.0000] <jgraham>
Random, probably rather obvious, point; for sufficiently simple captions it's not clear to me that I want them overlayed on the video in the same way that they are on the TV. For fullscreen it makes sense but if I am watching something that only takes up 50% of my screen height I might want the subtitles to sit under the video

[02:06:00.0000] <jgraham>
zcorpan: It is not clear to me that there is a "natural" order for the arguments of assertEquals

[02:08:00.0000] <Philip`>
assertEquals(1+1, 2);

[02:08:01.0000] <Philip`>
assertEquals(2, 1+1);

[02:08:02.0000] <Philip`>
The first sounds more natural to me

[02:08:03.0000] <zcorpan>
jgraham: i often omit message, so from that point of view at least message should be last

[02:09:00.0000] <jgraham>
zcorpan: I tend to agree but I guess the idea here is that the message is mandatory

[02:09:01.0000] <othermaciej>
Philip`: agree, (test, expectation) seems like the natural order

[02:09:02.0000] <jgraham>
I'm not sure I agree

[02:10:00.0000] <jgraham>
If you wrote it like assert(1+1 == 2) I would agree

[02:11:00.0000] <jgraham>
But assert equals 2, 1+1 also makes sense

[02:11:01.0000] <jgraham>
Not typical english word order but logical

[02:12:00.0000] <gsnedders>
Balloons exploading are kinda loud.

[02:12:01.0000] <zcorpan>
gsnedders: happy birthday

[02:12:02.0000] <gsnedders>
zcorpan: thanks.

[02:14:00.0000] <jgraham>
zcorpan: In any case it's not clear to me that having lots of debates about the precise best format for testsuites is a good use of time. If microsoft are happy to write tests in this format, that is fine

[02:14:01.0000] <Philip`>
They should make vacuum-filled balloons

[02:14:02.0000] <Philip`>
Vacuum is much lighter than helium or even hydrogen, and the balloons could never explode

[02:15:00.0000] <jgraham>
Although I can't say that I would be inclined to use it myself

[02:15:01.0000] <jgraham>
(the only other option seems to be adopting some preferred framework in which case it is worthwhile to debate all the specifics)

[02:16:00.0000] <jgraham>
Which I guess is a point worth rasing on the list

[02:16:01.0000] <zcorpan>
jgraham: sure

[02:19:00.0000] <othermaciej>
jgraham: imagine if you also had assertFalse(), assertTrue(), assertThrowsException()

[02:19:01.0000] <othermaciej>
in each case you would want the test expression as the first parameter I think

[02:19:02.0000] <othermaciej>
so for consistency, assertEquals should also have it first, and the expectation second

[02:19:03.0000] <othermaciej>
my 2c anyway

[02:21:00.0000] <jgraham>
othermaciej: assertThrows should take some object to indicate the expected type and properties of the exception

[02:21:01.0000] <jgraham>
So that isn't an obvious counter-example

[02:30:00.0000] <othermaciej>
jgraham: it also needs to take a string to eval instead of an expression to do anything useful

[02:32:00.0000] <Philip`>
Isn't it better to pass a function(){} to call than a string to eval?

[02:32:01.0000] <othermaciej>
this is what WebKit's script tests look like: http://trac.webkit.org/export/57879/trunk/LayoutTests/fast/dom/HTMLAnchorElement/set-href-attribute-hash.html

[02:32:02.0000] <jgraham>
othermaciej: I would pass a function

[02:33:00.0000] <othermaciej>
and view the linked script-tests/set-href-attribute-hash.js to see what the actual test scripts look like

[02:33:01.0000] <jgraham>
/me has written an assertThrows with the signature assertThrows(expectedException, callable, message)

[02:33:02.0000] <othermaciej>
I think both the output and the source are more readable than Microsoft's format

[02:34:00.0000] <othermaciej>
(or whoever came up with that)

[02:34:01.0000] <othermaciej>
we use strings for both the test expression and the expectation

[02:34:02.0000] <othermaciej>
that way shouldBe (our version of assertEquals) can catch exceptions

[02:34:03.0000] <othermaciej>
so it can report unexpected throw instead of making the rest of the test bomb out

[02:35:00.0000] <othermaciej>
otherwise you have to either surround every test assertion with try/catch, or pass everything as closures

[02:35:01.0000] <jgraham>
I don't like using strings; making everything rely on eval can have surprising results

[02:36:00.0000] <othermaciej>
in practice it hasn't caused a problem

[02:36:01.0000] <othermaciej>
other than people getting confused about the need to double-quote actual strings

[02:37:00.0000] <othermaciej>
anyway, strings are not the main point

[02:37:01.0000] <othermaciej>
my point was (a) readable output; (b) readable test script

[02:37:02.0000] <jgraham>
I think quoting hell plus the lack of syntax highlighting for the test code is enough to put me off

[02:37:03.0000] <othermaciej>
could easily convert the system to use all closures

[02:39:00.0000] <othermaciej>
here is one with examples of more of the types of assertions: http://trac.webkit.org/export/57879/trunk/LayoutTests/fast/js/regexp-compile.html

[02:39:01.0000] <jgraham>
We have a whole bunch of javascript tests that use assertEquals(expected, actual, [message]) type syntax, use try/catch on the backend to handle failures and uses closures for cases where the test is expected to throw

[02:39:02.0000] <jgraham>
It has worked pretty well

[02:39:03.0000] <othermaciej>
and the corresponding JS source: http://trac.webkit.org/export/57879/trunk/LayoutTests/fast/js/script-tests/regexp-compile.js

[02:40:00.0000] <jgraham>
(nothing that I can link to sadly)

[02:41:00.0000] <othermaciej>
anyway, say what you will about the details of the assert function parameters, I think our output is the most useful I have seen for script tests

[02:41:01.0000] <othermaciej>
the color makes it easy to spot failures at a glance, it copies to plain text well, it doesn't have pointless tables, etc

[02:41:02.0000] <jgraham>
However, I think the main difficulty with DOM tests (rather than pure ECMAScript tests) is dealing with async results\

[02:42:00.0000] <jgraham>
e.g. if you want to test that some event is fired when it ought to be

[02:42:01.0000] <othermaciej>
another reason to use strings is to include the test expression in the test output

[02:42:02.0000] <othermaciej>
so you don't have to say what ou tested a second time

[02:42:03.0000] <othermaciej>
that's definitely a problem with *some* DOM tests

[02:43:00.0000] <othermaciej>
not with the attributes of the <a> element though

[02:43:01.0000] <jgraham>
Well no

[02:43:02.0000] <jgraham>
But in general, if you want a framework that deals well with testing DOM, it has to test async stuff well

[02:43:03.0000] <jgraham>
That is the problem that I don't feel I have ever found a good solution to

[02:44:00.0000] <othermaciej>
we use ad hoc methods

[02:44:01.0000] <jgraham>
(A good general solution rather than just making it up for each specific case)

[02:44:02.0000] <othermaciej>
the most annoying is if you want to test that some async event fires eventually, and one possible bug is that it never fires

[02:44:03.0000] <othermaciej>
then you kinda have to add an arbitrary timeout, unless you know of a later event that will likely fire even in the bug case

[02:44:04.0000] <jgraham>
Yeah, but that's a pretty typical case

[02:45:00.0000] <gsnedders>
othermaciej: But the problem with outputting the JS source is that doing it as the WebKit tests do relying upon eval seems bad

[02:45:01.0000] <othermaciej>
we end up just relying on the fact that our test framework times out eventually

[02:45:02.0000] <jgraham>
So it sucks if you have to reinvent the wheel every single time

[02:46:00.0000] <othermaciej>
gsnedders: I hate eval as much as the next guy, but the string thing works really well in this case - too bad JS doesn't have macros

[02:47:00.0000] <jgraham>
othermaciej: If you write tests with one test per function, you can toString the function to get the text of the test

[02:47:01.0000] <zcorpan>
i now have a default timeout of 2s (many tests depend on network so need quite a long timeout) and require an explicit end() when the test is done in my framework

[02:47:02.0000] <jgraham>
it's not quite as simple

[02:47:03.0000] <zcorpan>
it works pretty well for async tests

[02:48:00.0000] <gsnedders>
othermaciej: Until you come across cases where you have bugs and eval has bugs the normal codepath doesn't (or vice-versa)

[02:48:01.0000] <Philip`>
/me does assertEqual(1+1, 2, "1+1", "2"); in order to be able to output the JS source without relying on eval

[02:48:02.0000] <othermaciej>
jgraham: true, and then you are just trading off the oddities of eval and the awkwardness of quoting for the oddities of code in a function and the awkwardness of saying function() { } all the time

[02:48:03.0000] <othermaciej>
Philip`: I think repeating yourself is a worse programming sin than eval

[02:48:04.0000] <Philip`>
...with an ugly custom preprocessor language that converts "@assert 1+1 == 2;" into that syntax automatically

[02:48:05.0000] <Philip`>
so that I don't have to repeat anything :-)

[02:49:00.0000] <othermaciej>
using a preprocessor is cheating :-)

[02:49:01.0000] <gsnedders>
othermaciej: But functions work better in the case of JITing compilers that compile whole functions at once, as then you can just loop the function

[02:49:02.0000] <jgraham>
It's true that tests-in-a-function can have problems if you want to test something that specifically depends on global scope

[02:49:03.0000] <zcorpan>
/me does assertEquals(1+1, 2, "1+1") or just assertEquals(1+1, 2)

[02:49:04.0000] <othermaciej>
/me still wonders what y'all think of the output format, as opposed to the use of eval

[02:50:00.0000] <jgraham>
These problems are perhaps more pronounced if you are trying to test the ECMAScript engine itself rather than the DOM

[02:50:01.0000] <gsnedders>
The output format is I don't think an issue, and is really just a bikeshed :)

[02:51:00.0000] <othermaciej>
http://dvcs.w3.org/hg/html/raw-file/tip/tests/submission/Microsoft/interface/anchor_href.htm is really ugly

[02:51:01.0000] <jgraham>
othermaciej: It's pretty nice. Not quite to die for though ;)

[02:51:02.0000] <othermaciej>
and looks like it would be hard to spot failures

[02:51:03.0000] <othermaciej>
tables with double borders on everything, srsly

[02:51:04.0000] <jgraham>
Yes, I would turn off the table borders and use some colour there

[02:51:05.0000] <othermaciej>
so 1998

[02:52:00.0000] <othermaciej>
I don't get why it's a table at all

[02:52:01.0000] <othermaciej>
I also can't tell what it's actually testing from reading

[02:52:02.0000] <zcorpan>
/me 's output is a <pre> which says "PASS" if it passed, or "FAIL" followed by the failing messages if it failed

[02:52:03.0000] <othermaciej>
since the descriptions are arbitrary strings and don't show the code it ran

[02:53:00.0000] <othermaciej>
readable test output makes it much easier for implementors to look at the test and fix their implementations accordingly

[02:53:01.0000] <othermaciej>
which is why I think it's important, more so than the details of how test assertions are coded

[03:02:00.0000] <jgraham>
othermaciej: I need to read the source code in either case to be confident I know what is going on

[03:03:00.0000] <othermaciej>
which is why it helps for the source to be readable too, which is why I dislike the giant try/catch blocks, and the fact that it's hard to map the table output to the test assertions in the source

[03:03:01.0000] <jgraham>
Giant try/catch blocks are unnecessary

[03:04:00.0000] <jgraham>
You should be able to do function testAddition() {assertEquals(2, 1+1)}

[03:05:00.0000] <jgraham>
Then you only need to dive in deeper if you broke the harness itself not the test

[03:05:01.0000] <hsivonen>
Does IE9 mode have a default border on linked images?

[03:05:02.0000] <jgraham>
(and you should make the harness simple to avoid that)

[03:06:00.0000] <jgraham>
and you can use e.g. the <details> element in the output to show the full test source

[03:06:01.0000] <zcorpan>
jgraham: hmm, <details> seems to mean the framework is getting too complex

[03:07:00.0000] <jgraham>
zcorpan: In the output? 

[03:07:01.0000] <jgraham>
I don't see why

[03:07:02.0000] <jgraham>
Not all tests can be a single line of code

[03:08:00.0000] <othermaciej>
test assertions should generally be one line, though I could imagine multiple additional lines of setup code

[03:08:01.0000] <jgraham>
Yes, that's what I mean

[03:09:00.0000] <zcorpan>
well, would you want the <details> to work in existing browsers?

[03:09:01.0000] <jgraham>
Since the setup code is just as important as the assertion it makes sense to display it

[03:09:02.0000] <zcorpan>
you can view source

[03:10:00.0000] <jgraham>
zcorpan: Of course you can view source. othermaciej seems to be arguing that there is value in not having to view source for each fail

[03:10:01.0000] <jgraham>
Making details work somewhat in existing browsers is not the world's greatest challenge

[03:12:00.0000] <othermaciej>
jgraham: I don't mind if view source is required sometimes, what I find nice is if the fail message itself (a) often gives useful info in itself about what went wrong and (b) is easy to match up with the test assertion code in the source

[03:12:01.0000] <othermaciej>
I guess having those messages satisfies (b) but not (a), assuming the descriptions are unique

[03:55:00.0000] <Lachy_>
Does anyone in here know what should happen in the following window.stop() test? There's not much interop in this case between browsers

[03:55:01.0000] <Lachy_>
http://html5.lachy.id.au/output?data=%3C%21DOCTYPE+html%3E%0D%0A%3Cscript%3E%0D%0Adocument.write%28%22A%22%29%3B%0D%0Awindow.stop%28%29%3B%0D%0Adocument.write%28%22B%22%29%0D%0A%3C%2Fscript%3E&type=text%2Fhtml%3B+charset%3DUTF-8

[04:11:00.0000] <MikeSmith>
boblet: about your ruby article, one suggestion to maybe consider is adding some images of ruby annotations in non-Web contexts

[04:12:00.0000] <MikeSmith>
I was pointing out some to anne this morning on the train and I think he took some pictures

[04:12:01.0000] <MikeSmith>
but anyway, I think for people outside of Japan or Asia, it might help to show them some examples of real-world use of ruby

[04:15:00.0000] <gsnedders>
Apparently I'm meant to be responsible now. That seems bad.

[04:17:00.0000] <Lachy>
gsnedders, you're allowed to be completely irresponsible on your birthday.

[04:17:01.0000] <Lachy>
as long as there is alcohol involved

[04:19:00.0000] <hsivonen>
gsnedders: did you reach the legal drinking age in Sweden?

[04:19:01.0000] <jgraham>
This could be messy...

[04:20:00.0000] <Lachy>
if Sweden is anything like Norway, then I think at 18, you're allowed to drink low percentage alcohol, like beer

[04:20:01.0000] <Lachy>
and I guess wine too

[04:20:02.0000] <jgraham>
I don't think Sweden is like Norway

[04:21:00.0000] <jgraham>
He can't *buy* strong alcohol

[04:21:01.0000] <Lachy>
oh, then that's good. He'll be allowed to drink any alcohol. Anyone got a bottle of Austrian Rum for him?

[04:21:02.0000] <gsnedders>
I think, from what people were saying at lunch, I'm allowed to drink alcohol up to 3.5%

[04:22:00.0000] <gsnedders>
Lachy: If they got it with the intention of giving it to me, they've broken the law.

[04:22:01.0000] <jgraham>
gsnedders: Pointer? Wikipedia at least doesn't obviously agree with that

[04:23:00.0000] <MikeSmith>
gsnedders: the good part is that if we engage in felony crimes and get caught, you can go to the same prison with me .. instead of you being charged as a minor and going to junior-criminal jail 

[04:23:01.0000] <MikeSmith>
you get to go to the real prison

[04:23:02.0000] <gsnedders>
MikeSmith: I could from 15

[04:24:00.0000] <MikeSmith>
well, yippee-aye-kai-yai-yay

[04:24:01.0000] <gsnedders>
Oh, so I can drink alcohol legally. I can get the light beer sold in supermarkets.

[04:24:02.0000] <Philip`>
gsnedders: If you hypnotise somebody into buying it and giving it to you, does that count as intent?

[04:25:00.0000] <Lachy>
gsnedders, has there been a party planned for you that you know about yet?

[04:25:01.0000] <gsnedders>
But I can't go to the state-owned shop and buy anything.

[04:25:02.0000] <gsnedders>
Philip`: IANAL

[04:25:03.0000] <jgraham>
gsnedders: Yes, that is my understanding

[04:25:04.0000] <gsnedders>
Lachy: yes

[04:25:05.0000] <MikeSmith>
gsnedders: if we you were already fully prepared for serious criminal work at 15, I don't know what we've been waiting for, then. Let's get out and do some damage

[04:25:06.0000] <gsnedders>
MikeSmith: fuck yeah.

[04:27:00.0000] <jgraham>
gsnedders: Also I would like to reconfirm that Fleetwood Mac are indeed dire

[04:27:01.0000] <MikeSmith>
gsnedders: you got the right "can do" spirit

[04:29:00.0000] <gsnedders>
jgraham: You're wrong.

[04:32:00.0000] <jgraham>
On the other hand, the new Jónsi album is, 1.5 tracks in, pure awesome

[04:32:01.0000] <boblet>
MikeSmith: good idea — will add a couple

[04:32:02.0000] <MikeSmith>
friends, can anybody tell me where the latest Notifications API draft spec is?

[04:33:00.0000] <MikeSmith>
this? -> http://www.chromium.org/developers/design-documents/desktop-notifications/api-specification

[04:33:01.0000] <MikeSmith>
and is Mozilla also implementing it? or is it just Chromium project that plans to at this point?

[04:34:00.0000] <MikeSmith>
boblet: you might be able to even find some examples of add with vertical text and ruby

[04:35:00.0000] <boblet>
I’d like to do a haiku like that, but need to get IE working so I can actually get a screenshot

[04:35:01.0000] <zcorpan>
iirc, it's legal to give alcohol to children, but not to sell

[04:35:02.0000] <boblet>
broke the VM image :/

[04:35:03.0000] <MikeSmith>
I think you can find plenty of books and magazine or newspaper articles, but I think an ad with color and different font sizes and such would be more example-worthy

[04:35:04.0000] <zcorpan>
you're allowed to buy strong drinks in pubs but not in systembolaget

[04:35:05.0000] <zcorpan>
when you're 18

[04:35:06.0000] <boblet>
MikeSmith: aah, you mean images? yeah good idea

[04:35:07.0000] <boblet>
bbl

[04:36:00.0000] <MikeSmith>
boblet, yeah, actually images from train ads or whatever

[04:36:01.0000] <MikeSmith>
zcorpan: what is systembolaget? like vinmonopolet?

[04:37:00.0000] <Philip`>
boblet: Could use something like browsershots.org when you want IE screenshots

[04:37:01.0000] <zcorpan>
MikeSmith: yeah

[04:38:00.0000] <jgraham>
MikeSmith: Yeah, it's the government run "if-we-don't=sell-all-the-alcohol-and-not-on-a-Sunday-or-after-3-on-a-Saturday-then-you-will-all-become-alchoholics-obviously store"

[04:38:01.0000] <MikeSmith>
heh

[04:39:00.0000] <zcorpan>
a few years ago it was all of saturday also

[04:40:00.0000] <MikeSmith>
liquor stores are surprisingly harder to break into than pharmacies

[04:41:00.0000] <jgraham>
Until a few months ago there was also a government monopoly on selling pharmacuticals

[04:41:01.0000] <jgraham>
Then someone noticed that the only other countries with the same restiction were Cuba and North Korea

[04:41:02.0000] <jgraham>
At that point I guess it became a bit embarassing

[04:42:00.0000] <MikeSmith>
the Axis of Evil: Cuba, North Korea, and Sweden

[04:43:00.0000] <hsivonen>
gsnedders: well, congratulations on aging to a point where statutory conditions flip

[04:53:00.0000] <MikeSmith>
me finds http://dev.w3.org/2006/webapi/WebNotifications/publish/

[04:53:01.0000] <MikeSmith>
hsivonen: do you recall anybody from Mozilla posting about intentions to implement that?

[04:54:00.0000] <zcorpan>
/me does s/loadedfirstframe/loadeddata/ in https://developer.mozilla.org/En/Using_audio_and_video_in_Firefox but other things in there are probably also outdated

[04:55:00.0000] <zcorpan>
e.g. i don't get any canshowcurrentframe event

[04:55:01.0000] <MikeSmith>
I seem to remember recent discussion of the NOtifications api but can't recall where.. maybe the whatwg list?

[05:00:00.0000] <gsnedders>
hsivonen: thanks

[05:05:00.0000] <MikeSmith>
maybe I'm thinking about the notifications feature that gavin has been working on...

[05:05:01.0000] <MikeSmith>
which is unrelated, I think

[05:06:00.0000] <othermaciej>
MikeSmith: there has been discussion on webapps

[05:06:01.0000] <othermaciej>
dunno if it is of the same or a related API

[05:06:02.0000] <othermaciej>
I think it is related

[05:06:03.0000] <MikeSmith>
ok

[05:06:04.0000] <MikeSmith>
I'll go back and look there again

[05:07:00.0000] <MikeSmith>
ah, OK, I see now - http://lists.w3.org/Archives/Public/public-webapps/2010JanMar/thread.html#msg893

[05:07:01.0000] <MikeSmith>
last month, not this month

[05:07:02.0000] <MikeSmith>
having so much fun I lose track of the time

[05:08:00.0000] <MikeSmith>
oh

[05:10:00.0000] <othermaciej>
/me takes a peek at http://damowmow.com/playground/htmlwg/chart.html and decides to cry himself to sleep

[05:10:01.0000] <gsnedders>
/me hugs othermaciej 

[05:10:02.0000] <gsnedders>
You'll get over this tough time…

[05:11:00.0000] <othermaciej>
look at that trendline!

[05:14:00.0000] <jgraham>
othermaciej: I would be more worried about the fact that we are clearly due for a change of chair

[05:14:01.0000] <Lachy>
othermaciej, it's showing that the number of closed issues is steadily increasing, which is a good thing

[05:14:02.0000] <othermaciej>
jgraham: don't tempt me!

[05:14:03.0000] <jgraham>
That graph is predicting that you have a 1 in 3 chance of a nasty accident!

[05:15:00.0000] <othermaciej>
Lachy: that is a good thing - just need to get the number of closed issues decreasing

[05:15:01.0000] <othermaciej>
er

[05:15:02.0000] <othermaciej>
get the number of open issues decreasing

[05:16:00.0000] <othermaciej>
I think the closed issues have a much steeper slope under the current set of chairs than ever before, but it's kinda hard to see on this graph

[05:16:01.0000] <daedb>
So, what's the best place to get a free website (without ads) these days? Wordpress?

[05:17:00.0000] <Lachy>
at 2 out of the 3 chair changes so far, the graph did see a sudden rapid decline in open issues, so maybe that could be a good thing.

[05:38:00.0000] <nessy>
hsivonen: finally got time to read your proposal properly

[05:38:01.0000] <nessy>
I mostly like it ;)

[05:38:02.0000] <hsivonen>
nessy: nice

[05:38:03.0000] <hsivonen>
nessy: what's "mostly" about?

[05:38:04.0000] <nessy>
you'll see my reply

[05:39:00.0000] <hsivonen>
ok

[05:39:01.0000] <nessy>
I think we might need to deal with time-overlapping segments

[05:39:02.0000] <nessy>
but there may be a way to fit that onto the proposal

[05:39:03.0000] <nessy>
I'm not too worried about it for now

[05:40:00.0000] <nessy>
your proposal generally feels right

[05:40:01.0000] <nessy>
and I like the idea of calling events

[05:40:02.0000] <nessy>
anyway - gotta go - sorry

[05:41:00.0000] <Lachy>
which proposal are you referring to?

[05:43:00.0000] <hsivonen>
Lachy: http://lists.w3.org/Archives/Public/public-html/2010Apr/0365.html

[05:44:00.0000] <jgraham>
hsivonen: Your proposal explicitly prevents having the subtitles below the video (rather than overlaid)

[05:46:00.0000] <jgraham>
I know this is not common in non-web situations, but it unclear to me whether this is by design or just a limitattion of the typical medium

[05:47:00.0000] <MikeSmith>
http://www.w3.org/Submission/2010/SUBM-WOFF-20100408/ WOFF spec has been published as a W3C Member Submission

[05:48:00.0000] <hsivonen>
jgraham: to have them outside the video's CSS box, I think it's necessary to spec how to target the timed text to a non-anonymous iframe

[05:48:01.0000] <hsivonen>
jgraham: because establishing the rendering space for the captions outsize the <video> box using merely pseudos would suck

[05:49:00.0000] <hsivonen>
jgraham: so in that case, you'd need a concrete element to establish the box sanely

[05:49:01.0000] <jgraham>
Works for me as a suggestion

[06:11:00.0000] <nessy>
hsivonen: are you still there?

[06:14:00.0000] <nessy>
you mentioned SVG 1.2 Tiny textArea in the email to Sean - what is the status of that in HTML5?

[06:35:00.0000] <hsivonen>
nessy: http://lists.w3.org/Archives/Public/public-html/2009Sep/0005.html

[06:36:00.0000] <nessy>
thanks

[06:37:00.0000] <nessy>
Hixie: why did you remove all the other examples of time-aligned text from http://wiki.whatwg.org/wiki/Timed_tracks ?

[06:41:00.0000] <nessy>
ah I just found http://wiki.whatwg.org/wiki/Use_cases_for_API-level_access_to_timed_tracks

[06:42:00.0000] <nessy>
but I think the lyrics part should go back into the other page

[06:47:00.0000] <hsivonen>
I guess I should find an example of italics in otherwise plaintext subtitling

[06:48:00.0000] <hsivonen>
too bad the method for finding such an example is watching a lot of tv

[07:00:00.0000] <nessy>
hsivonen: yeah, italics, bold and colors

[07:02:00.0000] <hsivonen>
I don't remember seeing bold or colors on tv

[07:03:00.0000] <jcranmer>
karoake subbing for foreign theme songs ?

[07:03:01.0000] <jcranmer>
that uses multiple colors...

[07:04:00.0000] <zcorpan>
i see colors occasionally (e.g. indicating different speakers)

[07:04:01.0000] <zcorpan>
but that's a whole line, not just a word

[07:06:00.0000] <jcranmer>
hmm, umineko used multple colors for some lines because the color of text made a difference

[07:09:00.0000] <Philip`>
zcorpan: http://philip.html5.org/misc/subtitles-3.xml has differently coloured speakers on a single line

[07:09:01.0000] <Philip`>
<p begin="00:15:38.48" id="p205" end="00:15:41.76"><span tts:color="yellow">That's not very good looking. </span>You're<br />right. <span tts:color="cyan">It isn't, is it?</span></p>

[07:09:02.0000] <Philip`>
(Default text colour is white)

[09:31:00.0000] <TabAtkins>
Who's nessy?  Silvia?

[09:42:00.0000] <zcorpan>
nessy said she would reply to hsivonen's email, and i see a reply to his email from Silvia

[09:42:01.0000] <TabAtkins>
That's what I'm basing my guess on.

[09:43:00.0000] <TabAtkins>
/me doesn't like it when people have email and irc names that are significantly different.

[09:43:01.0000] <zcorpan>
/me takes note but won't change his nick :)

[09:44:00.0000] <TabAtkins>
That's fine.  Just change your email instead.

[09:44:01.0000] <zcorpan>
i had zcorpan in my email address before

[09:44:02.0000] <zcorpan>
but now i have to use simonp⊙oc

[09:44:03.0000] <TabAtkins>
You can adjust your display name to show zcorpan.

[09:45:00.0000] <zcorpan>
naw

[09:45:01.0000] <zcorpan>
whois or google will reveal my name

[09:46:00.0000] <TabAtkins>
Now you're just being difficult.

[09:54:00.0000] <gsnedders>
TabAtkins: pingy?

[09:55:00.0000] <TabAtkins>
gsnedders: No ping.  Going to breakfast.  Will reping when I'm back.

[09:55:01.0000] <gsnedders>
TabAtkins: I'm going out to supper so no ;P

[09:55:02.0000] <TabAtkins>
Dammit, ok.

[09:55:03.0000] <TabAtkins>
What's up.

[09:56:00.0000] <gsnedders>
No time now ;P

[09:56:01.0000] <TabAtkins>
Nuuuuu!

[09:57:00.0000] <gsnedders>
No, ni!

[09:57:01.0000] <zcorpan>
jgraham: i haven't had much need for deep equality checking

[09:57:02.0000] <TabAtkins>
http://www.harkavagrant.com/index.php?id=253 <-- Most adorable squid ever.

[09:58:00.0000] <jgraham>
zcorpan: I have needed it a few times

[09:58:01.0000] <jgraham>
One possible implementation strategy (that I think jQuery/qUnit uses) is just to implement assertObjectEquals

[09:59:00.0000] <jgraham>
that checks deeply, and use it for lots of different types of object 

[09:59:01.0000] <jgraham>
(instead of also having e.g. assertArrayEquals, etc.)

[09:59:02.0000] <TabAtkins>
So, gsnedders, for real should I go to breakfast or talk?

[09:59:03.0000] <gsnedders>
But you basically can't do anything with unknown properties or check things that aren't enumerable

[09:59:04.0000] <gsnedders>
TabAtkins: Breakfast

[10:00:00.0000] <jgraham>
TabAtkins: Talk

[10:00:01.0000] <Philip`>
function deeplyEqual() { return uneval(a) === uneval(b); }

[10:00:02.0000] <jgraham>
function isEvil(user) {return user.irc_nick === "Philip`"}

[10:01:00.0000] <gsnedders>
or if it is Philip on other servers

[10:01:01.0000] <gsnedders>
Philip`: And what is the defined behaviour of uneval?

[10:01:02.0000] <Philip`>
gsnedders: I define it to be what happens when I do it in Firefox

[10:02:00.0000] <gsnedders>
Philip`: What version of Firefox, what platform?

[10:02:01.0000] <Philip`>
(I do actually use uneval for equality checking tests, but only for non-browser-based SpiderMonkey-using code)

[10:02:02.0000] <zcorpan>
window unevals to {}

[10:02:03.0000] <Philip`>
gsnedders: 1.5, BeOS

[10:03:00.0000] <zcorpan>
sorry

[10:03:01.0000] <zcorpan>
({})

[10:03:02.0000] <jgraham>
I don't think Firefox 1.5 ever ran on BeOS

[10:03:03.0000] <jgraham>
did it?

[10:03:04.0000] <Philip`>
http://www.bebits.com/appver/3143

[10:04:00.0000] <Philip`>
Looks like it did

[10:04:01.0000] <jgraham>
Yeah, I was just looking at that

[10:05:00.0000] <jgraham>
/me would like to play with Haiku sometime

[10:22:00.0000] <AryehGregor>
Hixie, is there some particular reason that uploads are disabled on the wiki, or should I go ahead and enable them?

[10:22:01.0000] <AryehGregor>
(We should really support uploads-in-database, if only so they can be enabled by default . . .)

[10:24:00.0000] <AryehGregor>
I hypothesize that video games have more markup in subtitles because they're really targeted at everyone, not just the hard-of-hearing.  In many games, not all lines are even voice-acted, so subtitles have to be on.  In other cases the speaker might be far away or drowned out by loud noises, since dialog isn't always fully scripted.

[10:26:00.0000] <AryehGregor>
Also, they're always from scripts, so you can copy italics from the scripts if applicable.  In unscripted videos, or videos that follow a loose script, there's no transcript to tell you definitively whether something should be italicized or bolded or such.

[10:27:00.0000] <AryehGregor>
I don't think 3D video games are good examples to base HTML5 video stuff on, except if it's an actual video cutscene rather than an in-game dialog (and actual video cutscenes are rare these days in games).

[10:27:01.0000] <Philip`>
/me notes that Portal (and presumably HL2 too?) has both subtitles (for everyone) and closed captions (for hard-of-hearing)

[10:28:00.0000] <Philip`>
(where the latter includes textual alternatives to sound effects)

[10:28:01.0000] <AryehGregor>
Interesting, I didn't know that.

[10:28:02.0000] <AryehGregor>
Most games don't bother.

[10:28:03.0000] <Philip`>
I don't remember seeing that option on any others

[10:28:04.0000] <Philip`>
though I've never really looked for that option on any others

[11:11:00.0000] <TabAtkins>
Argh I hate bitly and url shorteners in general!

[11:15:00.0000] <Philip`>
http://bit.ly/9Cm8kU

[11:15:01.0000] <TabAtkins>
Thanks for that, Philip`.  Big help.

[11:16:00.0000] <Philip`>
Glad to be of assistance

[11:18:00.0000] <Philip`>
Aren't they doing the world a service by preventing web developers from having to waste time coming up with concise readable URL structures, given that their URLs are all going to get obscured anyhow?

[11:18:01.0000] <TabAtkins>
Concise readable url structures are all we have!

[11:19:00.0000] <TabAtkins>
On an unrelated note, are table cells automatically BFCs? 

[11:19:01.0000] <TabAtkins>
/me supposes he could test this himself.

[11:20:00.0000] <TabAtkins>
Yes, they are.

[13:10:00.0000] <Hixie>
AryehGregor: uploads were disabled so that we wouldn't have to manage getting rid of abusive uploads

[13:32:00.0000] <KaOSoFt>
Good morning.

[13:32:01.0000] <TabAtkins>
Holy crap. http://craftymind.com/factory/html5video/CanvasVideo.html

[13:33:00.0000] <KaOSoFt>
TabAtkins- :o

[13:34:00.0000] <KaOSoFt>
That looks totally cool.

[13:38:00.0000] <TabAtkins>
You can do so many awesome/ridiculous things when you aren't having to muck around in plugin land.

[13:44:00.0000] <paul_irish>
TabAtkins: in his notes he says ... "One feature I didn’t have time to figure out was adding depth to the explosion, so pieces that are closest to ground zero fly up into the air as they sail outward. With full shadow effects this could look really cool."

[13:44:01.0000] <paul_irish>
hopefully he adds that as well

[13:44:02.0000] <TabAtkins>
Yus.

[14:25:00.0000] <danbri>
browser geeks, may I ask for some help w/ rdfa idioms?

[14:25:01.0000] <danbri>
in rdfa/xhtml it is ok to put <object><span ...rdfablahblah></span></object> in the <head> of a doc...

[14:26:00.0000] <danbri>
...if such a block of markup found its way out of the xhtml universe and into html5 or pre-html5 html doc header, ... does anyone know of ugly side-effects of putting object/span there?

[14:27:00.0000] <Philip`>
danbri: The <object> would force an implicit </head><body> in most (all?) browsers

[14:31:00.0000] <Philip`>
(It seems pretty bogus that HTML4 (hence XHTML) allows <object> in <head>, since it doesn't work like that)

[14:34:00.0000] <danbri>
so if this is done, it should be the last section of <head> at least?

[14:35:00.0000] <jgraham>
It can't end up in head at all in HTML5/real text/html browsers

[14:35:01.0000] <danbri>
well this design is for the xhtml flavour, but it would be nice if it was non-destructive if it leaked elsewhere

[14:35:02.0000] <danbri>
what would the practical consequences be?

[14:36:00.0000] <danbri>
(hopefully the html/rdfa wg will come up with some microdata/rdfa hybrid in future and i won't have to be asking such things next year...)

[14:37:00.0000] <jgraham>
danbri: Play around with http://livedom.validator.nu/?%3C!DOCTYPE%20html%3E%0A%3Chead%3E%0A%3Cobject%3E%3C%2Fobject%3E

[14:38:00.0000] <jgraham>
Basically the practical consequences wouldbe that anything that expected the object to end up in head would break and there would be a high chance of collateral damage from things moving out of head that shouldn't

[14:39:00.0000] <danbri>
so only rdfa parsers would be looking for it in head...

[14:39:01.0000] <Philip`>
The practical consequences would be that HTML5 parser developers would curse you for exploiting misfeatures in the HTML4 DTD

[14:39:02.0000] <danbri>
...and if the <object> block was the last thing in </head> then nothing else would be bumped into body...

[14:40:00.0000] <Philip`>
and it would abort in pure streaming HTML5 parsers

[14:40:01.0000] <Philip`>
when it hits the explicit <body> later

[14:42:00.0000] <Dashiva>
A non-legacy timed text format that doesn't support ruby... heh

[14:44:00.0000] <Philip`>
danbri: <head><link etc><object></object><body foo> would (in non-streaming parsers) give the same DOM as <head><link etc></head><body foo><object></object> so that should indeed work

[14:48:00.0000] <danbri>
and spans inside the object would survive? page display and function generally ok in currently deployed browser?

[14:51:00.0000] <Philip`>
danbri: As far as I'm aware, yes

[14:51:01.0000] <Philip`>
although it looks like IE6 doesn't put the <object> element in the DOM at all

[14:52:00.0000] <Philip`>
and puts the <span> element as a child of the <head>

[14:52:01.0000] <Philip`>
Oh, and then it crashes

[14:52:02.0000] <Hixie>
note that putting <object> in <head> is non-conforming in XHTML5 and HTML5

[14:56:00.0000] <danbri>
thanks guys

[14:56:01.0000] <danbri>
so what's this about 'would abort'? I thought html5 parsers were softer, gentler more accepting creatures...

[14:57:00.0000] <Philip`>
In slightly rare cases, like if you want SAX output without potentially-unlimited internal buffering in the parser, there are some errors you can't correctly recover from

[14:58:00.0000] <Philip`>
e.g. <body foo><body bar> (which should be equivalent to <body foo bar> but it's too late if you emit the body element after reading the first tag)

[14:59:00.0000] <Philip`>
which (unless I'm mistaken) is what happens when you put non-head content inside a <head> (so it emits an implicit body element) and then have a body tag later

[15:01:00.0000] <Hixie>
yeah, if you have a streaming parser (unusual for web browsers, but probably not unusual for rdfa processors) then that would likely cause an abort

[15:01:01.0000] <Hixie>
unless the streaming processor basically buffers everything

[15:01:02.0000] <Hixie>
but then it's not really streaming

[15:01:03.0000] <danbri>
ok well it seems the folk i'm helping are gonna stick with 'meta tags' for now

[15:02:00.0000] <danbri>
which gives some breathing time at least

[15:08:00.0000] <erlehmann>
the examples are pure awesome :3 http://wiki.whatwg.org/wiki/Timed_tracks

[15:12:00.0000] <TabAtkins_>
erlehmann: I just love the one with a nun yelling that someone should get an afro.

[15:13:00.0000] <erlehmann>
TabAtkins_, i love all of it because it provides a rare insight into what kind of cultural background the people who create the spec have.

[15:13:01.0000] <erlehmann>
something you cannot easily discern from other material

[15:16:00.0000] <Philip`>
Hmm, the BBC iPlayer has subtitles using TTML with colours, but the Flash player still only displays them all in white

[15:16:01.0000] <Philip`>
which is particularly confusing when there's multiple speakers on a single line, differentiated by colour

[15:18:00.0000] <Hixie>
erlehmann: most of them are the result of me searching for "subtitle" or "soustitre" on google images

[15:18:01.0000] <Hixie>
erlehmann: with various other keywords

[15:18:02.0000] <Hixie>
like "anime"

[15:20:00.0000] <erlehmann>
owww

[15:20:01.0000] <erlehmann>
and i thought

[15:21:00.0000] <erlehmann>
but meh, it may give away the origin of subtitling to the overlords using timed text 4357 A.D.

[15:53:00.0000] <TabAtkins_>
Looks like a few people on www-style have been hit by a virus.  Interesting.

[16:00:00.0000] <daedb>
Hixie: Do you still want an example screenshot of subtitle with individual words emphasized?

[16:02:00.0000] <Dashiva>
http://dashiva.net/misc/1271498287684.jpg - Like this? :)

[16:04:00.0000] <daedb>
Yeah, sorta.

[16:05:00.0000] <daedb>
I took a screenshot from a game though, so not the same style as that one ^^

[16:56:00.0000] <Traveler9>
hi


2010-04-21
[17:17:00.0000] <Hixie>
daedb: if it's not too hard to get, sure

[17:21:00.0000] <daedb>
Hixie: I have one. It's not gonna be readable if it's scaled down to 400px width though (it's a 1680*1050 image). It can be scaled down somewhat and still be readable, but not down to 400px.

[17:22:00.0000] <Hixie>
that's ok, the shrunk down images are just to remind me of what they are

[17:25:00.0000] <daedb>
http://daed.site24.se/Screenshot20100420222809422.jpg is it.

[17:31:00.0000] <TabAtkins>
daedb: You could probably alter the image to work, by taking the full-size text and just transporting it into a scaled-down image.

[17:34:00.0000] <Hixie>
daedb: thanks

[17:34:01.0000] <daedb>
TabAtkins: http://daed.site24.se/dao_sub_cropped.png <-- like this? :)

[17:35:00.0000] <TabAtkins>
daedb: Yes.

[18:07:00.0000] <Hixie>
nessy: i've split it again, now there's three pages :-)

[18:10:00.0000] <Hixie>
nessy: is there a word to refer to a single segment of a timed track? whatever unit of text has a start and end time?

[18:10:01.0000] <nessy>
I've started calling them text cues

[18:11:00.0000] <Hixie>
cue, that works

[18:11:01.0000] <Hixie>
thanks

[18:16:00.0000] <nessy>
sorry, was on the phone

[18:17:00.0000] <nessy>
what's the third page now?

[18:17:01.0000] <Hixie>
one is use cases for timed tracks rendered over video by the UA, the other is use cases for API-level access to timed tracks.

[18:17:02.0000] <Hixie>
the main page has no images on it anymore

[18:18:00.0000] <nessy>
/me goes checking it out

[18:19:00.0000] <nessy>
ah I see

[18:19:01.0000] <nessy>
the links to the other pages are there, too, which got me confused yesterday

[18:19:02.0000] <Hixie>
links to the other pages?

[18:28:00.0000] <Hixie>
nessy: should a UA always grab all the tracks from the network, or should it only grab tracks when they're enabled?

[18:29:00.0000] <nessy>
external files: only when they are enabled

[18:29:01.0000] <Hixie>
interesting

[18:29:02.0000] <nessy>
save bandwidth where you can, I'd say

[18:29:03.0000] <nessy>
why?

[18:30:00.0000] <Hixie>
so we really need three modes when we have the API... on and displayed, on but only for hte API, and off

[18:30:01.0000] <nessy>
what do you mean by "on but only for the API"?

[18:30:02.0000] <Hixie>
suppose one of the timed tracks is a list of IDs of slides and times for when the slides turn on and off

[18:31:00.0000] <Hixie>
you wouldn't want the IDs actually shown on the video (or even listed in the UI)

[18:31:01.0000] <Hixie>
but you'd still want then "enabled" and triggering events for the script

[18:31:02.0000] <Hixie>
s/then/them/

[18:32:00.0000] <nessy>
are you saying that external tracks should trigger events even if they are not active?

[18:32:01.0000] <Hixie>
what do you mean by active?

[18:32:02.0000] <Hixie>
(if they're triggering events, presumably they are active)

[18:32:03.0000] <nessy>
hmm… I guess there is a difference between being loaded and being displayed

[18:34:00.0000] <nessy>
maybe this has to be done with two different attributes just like on the media elements

[18:34:01.0000] <nessy>
something akin to @controls for being displayed and something akin to @preload for being "available"

[18:35:00.0000] <nessy>
though those names of course don't match

[18:35:01.0000] <Hixie>
i can't see any good way to do it other than something in attributes, yes

[18:35:02.0000] <Hixie>
but for now i'm just trying to work out requirements

[18:36:00.0000] <Hixie>
so that i can go through your proposals

[18:36:01.0000] <nessy>
sure - my brain jumped ahead

[18:36:02.0000] <nessy>
there is some need for turning on and off UA display

[18:37:00.0000] <nessy>
and there is some need for enabling / disabling a track and its events

[18:38:00.0000] <nessy>
it may even make sense to have that fourth state: displaying, but no events triggered

[18:38:01.0000] <Hixie>
oh?

[18:39:00.0000] <nessy>
no?

[18:39:01.0000] <nessy>
if the Web developer knows they don't need the events why should the UA have to make the effort?

[18:40:00.0000] <othermaciej>
nessy: no events of what kind?

[18:40:01.0000] <othermaciej>
nessy: like DOM events for JavaScript?

[18:40:02.0000] <Hixie>
nessy: an event every few seconds isn't a big deal

[18:40:03.0000] <othermaciej>
nessy: browsers can optimize to avoid dispatching the event at all if there are no listeners

[18:40:04.0000] <nessy>
we're just talking about the cues from external text resources for media elements

[18:40:05.0000] <othermaciej>
if it is a performance issue to dispatch it

[18:41:00.0000] <nessy>
ok - I was just thinking through the logical possibilities and whether three states will do or whether there are two independent dimensions

[18:42:00.0000] <othermaciej>
what are the three states being considered?

[18:45:00.0000] <nessy>
on and displayed, on but only for hte API, and off

[18:47:00.0000] <Hixie>
nessy: anything look obviously wrong on http://wiki.whatwg.org/wiki/Timed_tracks#Architecture ?

[18:49:00.0000] <nessy>
checking … did you consider tracks inside the video resource?

[18:50:00.0000] <Hixie>
oh, forgot about those, good point

[18:51:00.0000] <nessy>
they are essentially the same, but linked through the media resource

[18:51:01.0000] <Hixie>
(reload)

[18:54:00.0000] <Hixie>
bbiab - if you have any other comments feel free to mention them here, i'll look at backlog

[18:54:01.0000] <Hixie>
next stage is to go through your proposals

[18:54:02.0000] <Hixie>
to learn what parts of this they cover

[18:54:03.0000] <nessy>
looks ok

[18:55:00.0000] <nessy>
our proposals are rather bare on the javascript and rendering side of things - but solve the top, basically

[18:55:01.0000] <nessy>
I wonder about one thing: what if the timed text track has styling information?

[18:56:00.0000] <nessy>
CSS and Web Fonts cannot be the only part to trigger styling - it might come from the text resources

[18:59:00.0000] <nessy>
looks like you are solving the cue ranges problem as well with this ;)

[19:08:00.0000] <Hixie>
yeah hopefully we can hit cue ranges as well, that would be nice

[19:09:00.0000] <Hixie>
dunno what to do with the styling thing... are we sure we don't just want it to be CSS?

[19:09:01.0000] <Hixie>
i mean most systems don't let the author style the captions at all, from what i've seen

[19:09:02.0000] <Hixie>
e.g. TVs these days put that under teh control of the user

[19:10:00.0000] <Hixie>
and HTML itself has moved towards putting all styling in CSS

[19:10:01.0000] <Hixie>
it would be weird to make subtitles an exception

[19:10:02.0000] <Hixie>
but i don't know, depends what the use cases are

[19:58:00.0000] <nessy>
some of the formats have styling

[20:01:00.0000] <nessy>
like the QTText example that I emailed:

[20:01:01.0000] <nessy>
{QTtext} {size:16} {font:Lucida Grande} {width:320} {height:42}

[20:01:02.0000] <nessy>
{language:0} {textColor:65535,65535,65535} {backColor:0,0,0}

[20:01:03.0000] <nessy>
{doNotAutoScale:off} {timeScale:100} {timeStamps:absolute}

[20:01:04.0000] <nessy>
{justify:center}

[20:01:05.0000] <nessy>
but often this is ignored

[20:02:00.0000] <nessy>
where styling is on the elements rather than on the full text cue, we cannot simply ignore it I think

[20:02:01.0000] <nessy>
formats is a whole problem by itself and I am tending towards developing this new one that we have been discussing on WHATWG and then mapping everything else to that

[20:17:00.0000] <nessy>
or look at SSA (substation alpha) format, see http://en.wikipedia.org/wiki/SubStation_Alpha

[20:17:01.0000] <nessy>
probably the most comprehensive subtitling format around

[20:23:00.0000] <nessy>
found this positioning-interesting subtitle http://img3.imageshack.us/img3/6743/vlcsnap090208083027134xq5.png

[20:25:00.0000] <annevk>
vertical text strikes back

[20:25:01.0000] <annevk>
aaah

[20:27:00.0000] <Hixie>
nessy: i don't think we need to map everything to a single format; if the UA supports two formats, it can just support them both, that seems fine

[20:27:01.0000] <nessy>
ok, though I think there may be more than 2 in the future seeing all the different formats that exist

[20:27:02.0000] <nessy>
just found this, btw http://img254.imageshack.us/ifs/7581/img410/3/fabulousab2.png

[20:27:03.0000] <nessy>
has coloured text done in SSA and Matroska - want it on the page, too?

[20:28:00.0000] <Hixie>
sure

[20:28:01.0000] <Hixie>
karaoke i assume?

[20:28:02.0000] <Hixie>
wait, that's a demo of SSA and Matroska

[20:28:03.0000] <Hixie>
we don't want demos

[20:28:04.0000] <Hixie>
only things that people actually use

[20:28:05.0000] <Hixie>
the idea is to find out what the use cases really are

[20:29:00.0000] <Hixie>
i added some more thoughts here: http://wiki.whatwg.org/wiki/Timed_tracks#Declaring_timed_tracks

[20:29:01.0000] <Hixie>
based on your proposals

[20:29:02.0000] <Hixie>
a few questions arising:

[20:29:03.0000] <Hixie>
- what's the use case for media="" on timed tracks?

[20:29:04.0000] <Hixie>
- how do we handle sign language video overlays?

[20:30:00.0000] <Hixie>
- how do we handle alternative audio tracks?

[20:30:01.0000] <nessy>
not a demo - SSA and Matroska are not demos, they are fully in use

[20:30:02.0000] <Hixie>
i gotta go for dinner, but if you have any thoughts on the above, don't hesitate to mention them

[20:30:03.0000] <Hixie>
SSA and Matroska are not demos, but that file is a demo of SSA and Matroska

[20:30:04.0000] <Hixie>
if the filename is any indication

[20:30:05.0000] <Hixie>
anyway, gotta go, bbiab

[20:31:00.0000] <nessy>
enojy your dinner

[20:31:01.0000] <nessy>
@media is supposed to be what it is for video: help filter a file for a particular device

[20:32:00.0000] <nessy>
for example if some caption file is written for a specific width, but wouldn't work on mobile, then there might be a max-width on it

[20:33:00.0000] <nessy>
as for sign language videos, they are the same problem as actual audio descriptions (the ones that are recorded voice)

[20:33:01.0000] <nessy>
they both are not time-aligned text, but time-aligned other media resources

[20:34:00.0000] <nessy>
the browser vendors have all told me that they don't want to implement such media synchronisation at this point in time, in particular if they come from different servers

[20:34:01.0000] <nessy>
(btw: did you see my test at http://blog.gingertech.net/2010/02/12/audio-track-accessibility-for-html5/ ?)

[20:35:00.0000] <nessy>
however, the idea is that the <track> element will allow for such media files in the future, too

[20:35:01.0000] <nessy>
so, the proposal already covers it

[00:45:00.0000] <Hixie>
nessy: do sign language files and recorded audio descriptions have built-in timings?

[01:15:00.0000] <zcorpan>
Hixie: hmm. should the binary frame length parsing algorithm bail out if the first length byte is 0x80?

[01:17:00.0000] <annevk>
meh, CORS is complex

[01:17:01.0000] <annevk>
I should have said no way more often to the Mozilla guys...

[01:19:00.0000] <hsivonen>
/me doubts about all the anime fansubbing effects being hard requirements

[01:20:00.0000] <annevk>
HTML and CSS should address most of it

[01:20:01.0000] <Hixie>
yes let's put <video> in the subtitles! nothing could go wrong there

[01:21:00.0000] <Hixie>
/me points annevk to <annevk> I should have said no way more often to the Mozilla guys...

[01:21:01.0000] <Hixie>
zcorpan: the server guys are most likely to just ignore us on that front, and i don't see why the client guys would care about a 0x80 prefix

[01:22:00.0000] <annevk>
the comparison with CORS would be more to allow HTML/CSS but require a bunch of extra syntax to enable all their functionality

[01:22:01.0000] <annevk>
rather than not allowing HTML/CSS

[01:24:00.0000] <Hixie>
the needs i've seen so far are to allow ruby and italics. that's it. making subtitles use all of html and css is so far beyond overkill i don't even know where to begin.

[01:26:00.0000] <othermaciej>
Hixie: <video> in <video> subtitles sounds like awesome demo fodder

[01:26:01.0000] <othermaciej>
Hixie: I cannot wait to tweet someone's demo based on creatively abusing such a capability

[01:27:00.0000] <annevk>
the karaoke seemed to require some more

[01:27:01.0000] <Hixie>
which one?

[01:27:02.0000] <annevk>
the one with the rainbow text or just text gradually changing color

[01:27:03.0000] <othermaciej>
what about color?

[01:28:00.0000] <Hixie>
the rainbow colour one is a demo of what you can do with a subtitle format

[01:28:01.0000] <Hixie>
it's not a use case

[01:28:02.0000] <annevk>
i've seen it in the wild

[01:29:00.0000] <Hixie>
please put a screenshot or link up so i can study it

[01:29:01.0000] <annevk>
i mean in a bar

[01:30:00.0000] <Hixie>
go there and take a photo :-)

[01:30:01.0000] <annevk>
i'd love to if someone covers my ticket to nicaragua :p

[01:30:02.0000] <annevk>
though here in tokyo it should prolly be relatively easy to find

[01:31:00.0000] <annevk>
actually yesterday we mike and I much weirder stuff but no camera at hand

[01:31:01.0000] <Hixie>
(i don't think you can ever do what that example does with css currently... however, if we use css pseudo-elements to style subtitles in general, then that would handle that case once we have text paint servers)

[01:32:00.0000] <annevk>
yeah, that seems reasonable

[01:33:00.0000] <Hixie>
ok bed time

[01:33:01.0000] <Hixie>
nn

[01:45:00.0000] <nessy>
ah bummer missed Hixie

[01:45:01.0000] <nessy>
does Hixie read back?

[01:46:00.0000] <othermaciej>
I think he does

[01:46:01.0000] <othermaciej>
at least if you mention his name

[01:46:02.0000] <othermaciej>
or you can email him

[01:55:00.0000] <nessy>
just emailed him :)

[02:29:00.0000] <gsnedders>
So I went out and ordered alcoholic drinks last night. I didn't even get ID'd.

[02:30:00.0000] <othermaciej>
do you look old enough?

[02:34:00.0000] <annevk>
not really :p

[02:34:01.0000] <othermaciej>
lol

[02:42:00.0000] <jgraham>
He doesn't really look old enough to be out without his mother

[02:42:01.0000] <hsivonen>
well, at a dinner at TPAC in Boston, MikeSmith was ID'd (and denied alcoholic beverage) and I was not ID'd 

[02:44:00.0000] <zcorpan>
yay, now i can remove img { border:0 } from my style sheets

[02:44:01.0000] <annevk>
but Mike has a mustache

[02:44:02.0000] <hsivonen>
annevk: I had a collar on my shirt :-)

[02:46:00.0000] <hsivonen>
plinss had a collar, too, and was not ID'd. IIRC, everyone wearing t-shirts was ID'd regardless of facial hair

[02:47:00.0000] <annevk>
interesting

[02:47:01.0000] <annevk>
though Mike was telling me they also do 

[02:47:02.0000] <hsivonen>
and then the waiter refused to accept Mike's Japanese alien card as ID

[02:48:00.0000] <annevk>
euh oops; anyway, that was supposed to end in US airport trouble

[02:48:01.0000] <othermaciej>
MikeSmith looks way too old to ID

[02:48:02.0000] <othermaciej>
even I look too old to ID these days

[02:48:03.0000] <annevk>
ah yeah, I was on that table I think

[02:49:00.0000] <othermaciej>
I was getting ID'd regularly as recently as 2 years ago and I was 32 then

[02:49:01.0000] <annevk>
but you seem to have a much better memory of it :)

[02:49:02.0000] <othermaciej>
or 31?

[02:49:03.0000] <othermaciej>
I can't remember how old I am any more

[02:49:04.0000] <gsnedders>
/me has wine at TPAC in Mandelieu

[02:49:05.0000] <gsnedders>
*had

[02:49:06.0000] <gsnedders>
Apparently the age in France has been raised to 18

[02:49:07.0000] <annevk>
I went into a bar once in Palo Alto where I wasn't ID'd but chaals was

[02:49:08.0000] <gsnedders>
hah

[02:49:09.0000] <annevk>
and I was under 21 at the time :p

[02:50:00.0000] <jgraham>
You're assuming they want to catch you

[02:50:01.0000] <annevk>
(of course, the fully story is that the bouncer was in the restroom while I entered)

[02:50:02.0000] <annevk>
s/fully/full/

[02:51:00.0000] <nessy>
it's a compliment if you're getting ID-ed at 40 ;)

[02:52:00.0000] <annevk>
I'm not sure chaals would find it a compliment as he isn't 40 yet (afaik) :)

[02:55:00.0000] <annevk>
8 days until a new Ubuntu comes out

[02:55:01.0000] <annevk>
somewhat excited about this release

[02:56:00.0000] <jgraham>
I am considering installing it early just so I can change the theme as fast as possible

[02:56:01.0000] <jgraham>
s/fast/soon/

[02:57:00.0000] <jgraham>
(the buttons-on-the-left thing I can probably cope with, but srsly, what's with the dark grey)

[02:58:00.0000] <gsnedders>
Even the guy who jgraham calls emo dislikes the darkness of the theme.

[03:10:00.0000] <hsivonen>
I thought the default theme was supposed to be Light

[03:12:00.0000] <hsivonen>
do Qt apps still look totally different on Lucid?

[03:14:00.0000] <jacobrask>
should I escape & to &amp; in links in HTML5?

[03:15:00.0000] <hsivonen>
jacobrask: yes

[03:16:00.0000] <hsivonen>
you  don't have to in most cases, though

[03:16:01.0000] <jacobrask>
hm? :P

[03:17:00.0000] <jacobrask>
why and when should I

[03:17:01.0000] <hsivonen>
always, so you don't accidentally fail to escape when it's required

[03:17:02.0000] <jacobrask>
ok

[03:18:00.0000] <jacobrask>
when is it required then

[03:18:01.0000] <othermaciej>
per the latest spec, it's only required if the & is followed by something that matches an entity name

[03:19:00.0000] <jacobrask>
ok, and I won't remember all of those, so yes then I'll escape it all the time

[04:34:00.0000] <Dashiva>
It is mindboggling that bug 8833 is still going

[04:38:00.0000] <Dashiva>
Or maybe that's only if you assume rational intent...

[04:46:00.0000] <Philip`>
"There was no way that the group could come up with a title for the spec when the group was not told what the spec included."

[04:46:01.0000] <Philip`>
Couldn't the group have, like, read the spec to see what it included?

[04:46:02.0000] <zcorpan>
that would involve reading the spec

[04:46:03.0000] <othermaciej>
see my comments for copious evidence that the group was told what the spec included

[04:47:00.0000] <othermaciej>
without even having to read it even

[04:52:00.0000] <Dashiva>
It's a good thing we have a solid process in place to neutralize poisonous people!

[05:54:00.0000] <zcorpan>
hmm it's snowing

[05:55:00.0000] <jgraham>
zcorpan: Yes

[06:17:00.0000] <Lachy>
it was snowing briefly here in Oslo too. Seems to have stopped now though

[06:19:00.0000] <jacobrask>
in Gothenburg as well, for 10 minutes or so

[06:19:01.0000] <jacobrask>
weird

[06:35:00.0000] <Lachy>
I wonder if it has anything to do with all the volcanic ash messing with the weather.

[08:38:00.0000] <jarib>
does anyone maintain a complete IDL document from HTML5? 

[08:38:01.0000] <jarib>
i'm giving up on automatically extracting it from the markup

[08:38:02.0000] <asmodai>
Geez, I wonder if I should give up on this Lorentz build.

[08:39:00.0000] <asmodai>
It might protect Firefox from plugins, but so far it just crashes randomly without even touching it.

[08:39:01.0000] <jarib>
about to do extract the IDL parts manually, but now looking forward to it :P

[08:39:02.0000] <asmodai>
Now or not? :)

[08:39:03.0000] <jarib>
not! :)

[08:40:00.0000] <asmodai>
Nothing in the logs for the channel/

[08:40:01.0000] <asmodai>
?

[08:41:00.0000] <jarib>
hmm, i'll look

[08:56:00.0000] <hendry>
jarib: syntax.whattf.org/relaxng/ is probably the closest

[08:59:00.0000] <gsnedders>
That's not IDL

[09:22:00.0000] <Philip`>
jarib: Someone had set up a thing to automatically validate the spec's IDL

[09:22:01.0000] <Philip`>
which I assume includes the extraction of it

[09:26:00.0000] <Philip`>
jarib: http://www.w3.org/2009/07/webidl-check / http://dev.w3.org/2009/webidl-checker/webidl-check does the extraction, I think

[10:03:00.0000] <jarib>
Philip`: thanks, unfortunately just extracting the <pre class=idl> will still be incomplete

[10:03:01.0000] <jarib>
since some elements just has "DOM Interface: Uses HTMLElement"

[10:05:00.0000] <annevk>
why would you need to extract that for a IDL overview?

[10:06:00.0000] <annevk>
there's also some index table with IDL info

[10:06:01.0000] <jarib>
annevk: i'm generating code based on the IDL for a testing tool

[10:07:00.0000] <annevk>
oh sweet

[10:07:01.0000] <jarib>
basically i just need a map of tag_name => typed_attribute_list

[10:07:02.0000] <annevk>
browser testing?

[10:07:03.0000] <jarib>
yeah

[10:08:00.0000] <annevk>
maybe grab all IDL blocks, grab the element index -> IDL interface name table and mash it up?

[10:09:00.0000] <jarib>
yeah, that sounds like the best approach

[10:09:01.0000] <jarib>
i've tweaked it by hand for now, but that table will be useful in thefuture

[11:36:00.0000] <AryehGregor>
"Feel free to REOPEN if you think luring people to Berlin and telling them stories about volcanoes to make them stay is not an acceptable way of setting up a branch office."

[11:36:01.0000] <gsnedders>
I find it disturbing that I can find mistakes in a friend's English work, and know enough of a quote to be able to google it and find the exact quote

[11:37:00.0000] <gsnedders>
(in some other secondary work)

[11:59:00.0000] <daedb>
Wow, bug 8833 is really delicious :)

[12:12:00.0000] <TabAtkins>
daedb: I... I'm so confused.  What *possible* benefit could Apple, Google, and Maciej and Ian personally, gain from the name of the spec?

[12:13:00.0000] <TabAtkins>
The answer, of course, is "mu", because the dude is clearly an internet crazy.

[12:14:00.0000] <Dashiva>
TabAtkins: They force their bloated and useless and dangerous misfeatures onto the innocent real world filled with people who only ever write static documents

[12:15:00.0000] <Dashiva>
The personal gain is of course entertainment

[12:15:01.0000] <TabAtkins>
Ah, I see.  They get to laugh at the monkeys dancing on their fiendish strings.

[12:15:02.0000] <TabAtkins>
Makes sense.

[12:19:00.0000] <AryehGregor>
/me adds himself to the CC list

[12:20:00.0000] <daedb>
TabAtkins: It's the first step to total world domination, *obviously* :p

[12:21:00.0000] <daedb>
First you control the spec, then you control the language, and then somehow you win :)

[12:21:01.0000] <TabAtkins>
Then you get de wimmin.

[12:21:02.0000] <daedb>
That too.

[12:26:00.0000] <Hixie>
TabAtkins: i get paid by a secret group organised by swiss bankers for each day that the spec remains called "HTML5" -- the amount I make goes down for each character away from HTML5

[12:28:00.0000] <Hixie>
TabAtkins: that's why Jeremy was so concerned when I renamed the spec, he was worried I would lose out on that money

[12:29:00.0000] <Hixie>
jesus, what is joe doing

[12:31:00.0000] <miketaylr>
I KNEW IT

[13:15:00.0000] <zcorpan>
http://www.h-online.com/open/news/item/High-performance-Theora-codec-for-Firefox-on-OMAP3-previewed-980720.html

[13:35:00.0000] <TabAtkins>
AryehGregor: Where is the "luring people to Berlin" thing from?

[13:35:01.0000] <jgraham>
So, good people, is there any way to convince windows that a path can be > 255 characters long other than making it an absolute path starting \\?\

[13:35:02.0000] <TabAtkins>
Install linux.

[13:36:00.0000] <jgraham>
Sadly Windows is a critical feature in this case

[13:36:01.0000] <TabAtkins>
Then I give up.

[13:36:02.0000] <Philip`>
jgraham: As far as I'm aware, Windows doesn't have that requirement

[13:36:03.0000] <Philip`>
though some Windows tools do

[13:36:04.0000] <Philip`>
like xcopy

[13:36:05.0000] <jgraham>
Philip`: Subversion seems to 

[13:37:00.0000] <Philip`>
You can create more deeply nested directories in Explorer

[13:37:01.0000] <Philip`>
so it's probably a bug in the tool if it doesn't like them

[13:37:02.0000] <Dashiva>
TabAtkins: It's a mediawiki bug

[13:37:03.0000] <jgraham>
Specifically the Win32 API seems to have a function with that limitation even though NTFS allows much longer paths

[13:37:04.0000] <Dashiva>
They had a convention in Berlin or somewhere in Germany, and got stranded because of the volcano

[13:38:00.0000] <jgraham>
Philip`: Sadly subversion is also a critical component in this setup

[13:38:01.0000] <Dashiva>
TabAtkins: https://bugzilla.wikimedia.org/show_bug.cgi?id=23223

[13:39:00.0000] <Philip`>
jgraham: Can you make the path shorter, by removing some parent directories?

[13:39:01.0000] <TabAtkins>
jgraham: I had some tool that had similar problems.  We solved it by starting our path in c:/x/

[13:40:00.0000] <Philip`>
Or, can you install Cygwin? (since maybe that avoids the API limitation)

[13:40:01.0000] <Dashiva>
It doesn't help that the default documents folder is nested quite deep

[13:40:02.0000] <jgraham>
Philip`: Possibly

[13:41:00.0000] <jgraham>
Philip`: I fear that the same problem would just happen again (I don't really control the filenames so I can only trim the prefixes)

[13:42:00.0000] <jgraham>
TabAtkins: That is kind of my plan but it is not quite trivial in this case because the checkout is initiated by a different machine that doesn't know about the absolute paths on the remote end

[13:42:01.0000] <Dashiva>
Have you tried a different svn implementation?

[13:43:00.0000] <jgraham>
Dashiva: No. I don't really understand the subversion-on-windows situation

[13:44:00.0000] <jgraham>
Are there material differences between the various binaries?

[13:44:01.0000] <Philip`>
/me sees http://svn.haxx.se/users/archive-2005-02/1088.shtml says "if you feed Subversion an absolute path, [...] you *should* be okay"

[13:44:02.0000] <Dashiva>
There are several different svn libraries at least

[13:45:00.0000] <Philip`>
There's a big difference if you use TortoiseSVN

[13:46:00.0000] <Philip`>
though they still all use the same backend library as far as I'm aware

[13:46:01.0000] <Philip`>
because only crazy people would reimplement it from scratch

[13:48:00.0000] <Dashiva>
Lots of crazy people out there

[13:48:01.0000] <Dashiva>
There are two implementations just in java, even

[13:49:00.0000] <Philip`>
Java people are particularly crazy

[13:50:00.0000] <Philip`>
and I guess they like to not rely on native C libraries

[13:50:01.0000] <Philip`>
Are there any in other languages?

[13:50:02.0000] <Dashiva>
One is a JNI interface to the C lib, another is a full reimplementation

[15:03:00.0000] <Dashiva>
UMP keeps trucking along...

[15:04:00.0000] <Dashiva>
It seems it will be blocked waiting for two implementations until CORS finishes, so I don't see the hurry

[15:05:00.0000] <JonathanNeal>
I was wondering if I could get some feedback from you guys on http://sandbox.thewikies.com/learn-html/ a project I'm writing to teach kids HTML.

[15:06:00.0000] <JonathanNeal>
Or maybe 9th graders (not little ones)

[15:08:00.0000] <TabAtkins>
Interesting.

[15:09:00.0000] <Dashiva>
I think those kids are too young to get "all your base" :)

[15:11:00.0000] <Dashiva>
It's interesting, but I'm not sure if <em> and <strong> are good example elements. They're fairly abstract, and very similar (semantically, not visually)

[15:14:00.0000] <JonathanNeal>
Dashiva, sure, what other elements would be useful suggestions?  I picked <em> because it was short.

[15:14:01.0000] <JonathanNeal>
I would have picked <b> or <i> but I really didn't want to encourage that.

[15:18:00.0000] <Dashiva>
You could use one of em/strong, and another elements with a different type. E.g. <sup>

[15:18:01.0000] <TabAtkins>
Or <a>, though that's probably too complex for a first lesson.

[15:19:00.0000] <Dashiva>
<hx> and <p> are also common starting elements, in my experience

[15:21:00.0000] <TabAtkins>
JonathanNeal: Honestly, teaching kids <b> and <i>, and having them be misused to mark up emphasis and importance, is better than teaching them <strong> and <em>, and having them be misused to mark up citations, foreign words, random bold text, etc.

[15:21:01.0000] <TabAtkins>
Having not enough semantics is better than having wrong semantics.

[15:22:00.0000] <JonathanNeal>
Dashiva, headings and paragraphs would be the next lesson, I think.

[15:23:00.0000] <Dashiva>
I would tend to agree with what TabAtkins said there

[15:23:01.0000] <JonathanNeal>
I wanted to go for inline elements first, then either block elements or attributes (img, a, etc)

[15:24:00.0000] <JonathanNeal>
I was thinking of including "Solemn Oaths" after each lesson, where they agree not to use <strong> to bold random text, or <em> to italicize.

[15:24:01.0000] <Dashiva>
Then they'll ask "But how do we do it then?"

[15:25:00.0000] <Philip`>
You should start with <video>

[15:25:01.0000] <Dashiva>
And then you'll say "You'll have to wait a few weeks until we learn CSS"

[15:25:02.0000] <JonathanNeal>
Include an example in the oath!

[15:25:03.0000] <Philip`>
otherwise they'll get bored too soon

[15:25:04.0000] <Dashiva>
And then they'll say "That's boring" and use it anyway

[15:26:00.0000] <JonathanNeal>
Well, the reality is that I'm teaching them HTML as a content tool and not as a design tool.

[15:26:01.0000] <Dashiva>
Then <em> and <strong> aren't really that important

[15:27:00.0000] <JonathanNeal>
Well, name other useful inline elements, like <sup>

[15:28:00.0000] <Philip`>
When we had an introduction to HTML in IT lessons at school, it took about ten minutes for knowledge of <marquee> to spread throughout the whole room and then it was used on pretty much everybody's first pages

[15:28:01.0000] <Dashiva>
I had one person in my class who wrote each line as a <h1> or <h2>

[15:28:02.0000] <Dashiva>
And also forgot the / in closing tags

[15:29:00.0000] <Philip`>
JonathanNeal: Your page looks cooler if I tell it my name is <marquee>

[15:29:01.0000] <Philip`>
(Maybe you want to escape it before inserting it in the output)

[15:29:02.0000] <JonathanNeal>
Oh Philip` you dog!

[15:29:03.0000] <Dashiva>
What if you pretend to be little bobby tables?

[15:30:00.0000] <JonathanNeal>
Philip`, I've fixed that.

[15:33:00.0000] <TabAtkins>
Aw, I missed out on the fun.  Now you just call me 1337.  ;_;

[15:40:00.0000] <jgraham>
Practically speaking there is no difference between <em> and <i> or <b> and <strong>

[15:40:01.0000] <TabAtkins>
Argh, I have an almost irrestible urge to contact the makers of this Facebook game and offer to rewrite all their HTML.

[15:40:02.0000] <TabAtkins>
The different font metrics on Linux screw tons of stuff up because they're doing it wrong.

[15:40:03.0000] <jgraham>
I strongly doubt any general purpose tools make a distinction

[15:44:00.0000] <JonathanNeal>
jgraham, well in that case, I guess I'd just pick elements that make a distinction in their look and feel. 

[15:44:01.0000] <JonathanNeal>
look of element name and look of default rendered appearance.


2010-04-22
[17:29:00.0000] <Hixie>
nessy: do you have an opinion on whether we need to support roll-up or incrementally displayed titles, as you see on live TV broadcasts?

[17:29:01.0000] <nessy>
they are typically used for live captioning

[17:30:00.0000] <nessy>
so, for live streaming, probably yes, though I don't quite know how that would work

[17:30:01.0000] <Hixie>
yeah that was my next question - do we need to support live streaming, and what would that mean?

[17:30:02.0000] <Hixie>
presumably any live streamed captions would just be in the video file itself, rather than as an external file?

[17:30:03.0000] <Hixie>
or am i wrong?

[17:33:00.0000] <nessy>
I don't know, to be honest

[17:33:01.0000] <nessy>
I think what could be possible for live captions is to continuously write into an srt fil with timing and we would have to pick it up as it goes

[17:34:00.0000] <nessy>
but if it is included in the media file, then it's nothing different

[17:34:01.0000] <Hixie>
well if it's in the media file i conveniently don't have to worry about defining the format for it

[17:34:02.0000] <Hixie>
or the rendering semantics

[17:34:03.0000] <nessy>
indeed :)

[17:34:04.0000] <nessy>
ah, I think you still do

[17:34:05.0000] <Hixie>
so it's different in one rather important way :-)

[17:34:06.0000] <Hixie>
well i'm not defining e.g. the TTML rendering semantics

[17:34:07.0000] <nessy>
the rendering should be done the same no matter if it comes from an external file or from a media resource, IMO

[17:35:00.0000] <nessy>
incrementally displayed captions are still necessary since that is what karaoke does, no?

[17:35:01.0000] <Hixie>
i don't think that's possible, e.g. if an embedded timed track is a video of sign language, it can't possibly be rendered using the same CSS rules as an external SRT-like file.

[17:35:02.0000] <nessy>
no, I'm only talking about text tracks

[17:36:00.0000] <nessy>
external sign language video cannot be rendered into text either

[17:36:01.0000] <Hixie>
what's a "text" track? Is SVG+SMIL a "text" track?

[17:36:02.0000] <Hixie>
is TTML?

[17:36:03.0000] <Hixie>
would HTML+SMIL be?

[17:37:00.0000] <nessy>
I was talking about media-internal text tracks there, but yes, external files that are synchronised with the media would be regarded as virtual text tracks

[17:37:01.0000] <nessy>
if the browser can parse TTML, then it is a virtual text track

[17:38:00.0000] <Hixie>
we can define the rendering model of some formats, e.g. explaining where the text gets aligned, how it gets its styles from CSS, what happens if they are overlapping in time and how to do automatic stacking, etc, but we can't do that if the format has a built-in rendering model like XSL:FO or SVG

[17:38:01.0000] <nessy>
if the browser can parse HTML+SMIL and it is built to be a time-aligned text file, then yes, though with all of SMIL it generally would not be

[17:38:02.0000] <Hixie>
imho

[17:39:00.0000] <nessy>
I was hoping we could take the styles given in any format (including things like SSA, TTML, even SMILText) and transcode what is possible to CSS and then render it - then if the Web developer is overriding those stylings, that is fine

[17:39:01.0000] <nessy>
just like any media player does a best effort with caption formats that it supports

[17:39:02.0000] <Hixie>
that's an inane amount of work

[17:40:00.0000] <Hixie>
it's basically defining how you map XSL:FO to CSS, which is something the CSS and XSL working groups have avoided doing for about a decade

[17:40:01.0000] <nessy>
writing a image or video decoding library is, too - I regard these as text decoding libraries

[17:40:02.0000] <nessy>
and text decoding should be simpler ;)

[17:40:03.0000] <Hixie>
it's not

[17:41:00.0000] <Hixie>
if someone writes that spec, i'm willing to reconsider, but i think it would take me ~1 to 2 years to write such a spec, full-time

[17:41:01.0000] <nessy>
you don't have to map all of XSL:FO - only the stuff that TTML supports

[17:41:02.0000] <Hixie>
i'm talking about what TTML supports

[17:41:03.0000] <nessy>
really? there's not that much there, I thought...

[17:42:00.0000] <Hixie>
CSS and things like the inline box model are complicated to a degree that i cannot overstate

[17:42:01.0000] <nessy>
but indeed - until such time, it will not fit into the model and not be parsable

[17:42:02.0000] <Hixie>
well it can be parseable

[17:42:03.0000] <Hixie>
i don't see why we'd prevent it from being supported and rendered

[17:42:04.0000] <Hixie>
it just wouldn't automatically slide out of the way like multiple SRT captions could

[17:42:05.0000] <nessy>
how would you do that?

[17:42:06.0000] <Hixie>
it would effectively just be treated as a video stream

[17:43:00.0000] <Hixie>
wouldn't be affected by CSS pseudos or whatever mechanism we come up with to style srt-like subtitles (assuming we go with an srt model styled with pseudo-elements)

[17:43:01.0000] <nessy>
that's what Sean wants, I think … not sure I agree...

[17:44:00.0000] <nessy>
why wouldn't the content of a TTML <div> be treated like a srt-like subtitle and be able to apply the pseudo-elements to?

[17:44:01.0000] <nessy>
even if everything inside is "like a video stream"

[17:44:02.0000] <Hixie>
pseudo-elements apply at the inline box model level

[17:45:00.0000] <nessy>
we should not define an inline box model for srt - we should define an inline box model for cues, no matter where they come from

[17:45:01.0000] <Hixie>
CSS already has an inline box model. It's different than TTML's.

[17:45:02.0000] <Hixie>
i don't suggest we invent an inline box model.

[17:45:03.0000] <Hixie>
that way lies madness

[17:46:00.0000] <nessy>
ok, but you are creating boxes for srt

[17:46:01.0000] <Hixie>
i don't suggest creating boxes for srt

[17:46:02.0000] <nessy>
I suggest not to do that, but to create boxes for cues

[17:46:03.0000] <Hixie>
i suggest just defining how srt maps to css boxes

[17:46:04.0000] <Hixie>
maybe i'm not sure what you mean by "box"

[17:46:05.0000] <Hixie>
i am talking about CSS inline, line, and block boxes

[17:46:06.0000] <nessy>
and the cues could be handed "transparently" from TTML or any other format, just like decoded text - independent how that decoding is done

[17:46:07.0000] <Hixie>
TTML isn't text

[17:46:08.0000] <Hixie>
it's preformatted pixels

[17:47:00.0000] <Hixie>
just like HTML+CSS

[17:47:01.0000] <Hixie>
i don't know how to convey the complexity involved in defining a rendering model

[17:47:02.0000] <nessy>
can we deal with TTML as pre-formatted pixels by in a box per cue

[17:48:00.0000] <Hixie>
well we could but what does that give us?

[17:48:01.0000] <Hixie>
doesn't TTML define positioning?

[17:48:02.0000] <nessy>
only in relation to a "root area"

[17:48:03.0000] <nessy>
where we place the "root area" is up to us

[17:49:00.0000] <Hixie>
well we can't make the root area much less than the whole video, unless the text is going to end up very small or cropped

[17:49:01.0000] <nessy>
hmm … but I think I am starting to see what you mean ...

[17:49:02.0000] <Hixie>
surely

[17:50:00.0000] <nessy>
that's really f*ed up, actually

[17:50:01.0000] <nessy>
cause of course TTML thinks it knows how to avoid objects, but it only knows how to avoid objects in the video, not on anything else rendered on top by other streams

[17:51:00.0000] <Hixie>
that's a problem most formats are going to have

[17:51:01.0000] <Hixie>
even any we define

[17:51:02.0000] <nessy>
i's going to have to be solved by the Web author - cannot be done automatically

[17:52:00.0000] <nessy>
btw: Sean did say that when, e.g. the controls slide on from the bottom, the "root area" would need to be adjusted for TTML and would be reduced by that size

[17:53:00.0000] <nessy>
but that's a special case that the browser can deal with

[17:54:00.0000] <Hixie>
btw can you think of a use case for the author disabling a track that would normally be enabled by default? e.g. disabling chapter navigation tracks, or captions in the user's language when the user has opted into always seeing those captions?

[17:54:01.0000] <nessy>
otoh Eric said that this won't work with pilot-control-type centered controls

[17:55:00.0000] <nessy>
I think the author would disable all subtitling tracks by default - the browser would enable the one relevant to the user - and the user could select one or two of their choice

[17:57:00.0000] <Hixie>
yeah, i mean does the author ever need to override that default to make a track be disabled even if it would in this case be enabled?

[17:57:01.0000] <nessy>
but if you mean through javascript - can't think of an example.. probably not - the author should respect the user's wishes

[17:57:02.0000] <Hixie>
presumably chapter navigation tracks are always enabled by default

[17:58:00.0000] <nessy>
not necessarily

[17:58:01.0000] <Hixie>
(by "enabled" i mean downloaded and visible if it's a subtitle/caption track or in the ui if it's a navigation/chapters track)

[17:58:02.0000] <Hixie>
why ont?

[17:58:03.0000] <Hixie>
not?

[17:58:04.0000] <nessy>
if you have a video file that has all these tracks but as an author you just want the video to play back clean without any of the stuff, you might turn off all these tracks

[17:59:00.0000] <nessy>
that reminds me somewhat of the difference between open and closed captions 

[17:59:01.0000] <Hixie>
surely if you have chapter navigation information you wouldn't link to it but then make it so that the chapter navigation UI had to download the file when the user tried to use it

[18:00:00.0000] <nessy>
with external files I agree, but if they are part of the video resource, you probably wouldn't bother removing it

[18:00:01.0000] <nessy>
also, you might want the clean display, but still give the user the ultimate opportunity to turn them on

[18:01:00.0000] <Hixie>
i'm only talking about external files here

[18:01:01.0000] <Hixie>
presumably turning off chapter navigation tracks wouldn't hide the UI from the user? i guess i'm confused

[18:01:02.0000] <nessy>
hmm - maybe something like a movie where you want the playback to be clean, but you want to give ppl the opportunity to jump to a different chapter if they want to

[18:02:00.0000] <Hixie>
i don't expect chapter navigation tracks to do anything visible except give ppl the opportunity to jump to a different chapter if they want to

[18:02:01.0000] <Hixie>
am i wrong?

[18:02:02.0000] <nessy>
I would prefer to have them displayed actually

[18:02:03.0000] <Hixie>
i mean, JS can use them to display fancy stuff, sure

[18:02:04.0000] <Hixie>
oh

[18:02:05.0000] <nessy>
anything that is not visible doesn't really exist

[18:02:06.0000] <Hixie>
well it's visible

[18:02:07.0000] <Hixie>
it's right there in the menu

[18:03:00.0000] <nessy>
brb - phone call

[18:06:00.0000] <nessy>
I'm thinking of a display like http://junkyard.damowmow.com/414

[18:07:00.0000] <nessy>
most video players that I saw that used chapter markers effectively actually displayed them in relation to the timeline (the scrubbing bar)

[18:07:01.0000] <nessy>
best example are the ted talks, see http://www.arguingwithmyself.com/wordpress/wp-content/uploads/ted-video-player.png

[18:07:02.0000] <Hixie>
sure

[18:08:00.0000] <Hixie>
but we'd want that enabled by default no?

[18:08:01.0000] <Hixie>
i mean why would the author want that off?

[18:09:00.0000] <Hixie>
the reason i'm asking is to work out if i need enabled="true"/enabled="false" vs just enabled="" and no way to explicitly disable, even if the default is enabled

[18:09:01.0000] <nessy>
because presumably it's overlayed over the video and takes away viewing space from the video?

[18:09:02.0000] <nessy>
I guess it depends where it is being rendered

[18:09:03.0000] <Hixie>
surely not while the video is playing without controls

[18:09:04.0000] <Hixie>
only when the controls are up

[18:09:05.0000] <Hixie>
i mean as you said, it's linked to the scrubbing bar

[18:10:00.0000] <nessy>
well, both Safari and Google Chrome don't make the scrubbing bar disappear

[18:10:01.0000] <Hixie>
they don't?

[18:10:02.0000] <nessy>
if you have @controls, they stay there all the time

[18:11:00.0000] <nessy>
only Opera and Firefox make them disappear if you are not with your pointer in the video viewport

[18:11:01.0000] <Hixie>
it disappears for me in chrome

[18:11:02.0000] <othermaciej>
nessy: we do different things depending on whether the video is stopped or playing

[18:12:00.0000] <Hixie>
i would be shocked if it didn't disappear in safari

[18:12:01.0000] <nessy>
also, Safari has a different way for doing @controls that doesn't seem to involve a scrubbing bar, but only a play/pause button over the top of the video - I think they use that on iPhone and iPad - but I didn't see that with my own eyes, just heard it is like that

[18:13:00.0000] <nessy>
oh, I see - during playback they disappear

[18:13:01.0000] <nessy>
I guess that's ok then

[18:14:00.0000] <othermaciej>
if you go to this URL in Safari you can see what happens: http://v11.lscache2.c.youtube.com/videoplayback?ip=0.0.0.0&sparams=id%2Cexpire%2Cip%2Cipbits%2Citag%2Calgorithm%2Cburst%2Cfactor%2Coc%3AU0dWRllMV19FSkNNNl9RRkFH&fexp=900161&algorithm=throttle-factor&itag=18&ipbits=0&burst=40&sver=3&expire=1271919600&key=yt1&signature=1D5F628D66DBA39A8E15797752F29A16E02534AE.56973AF9547D1EC293EA36511175ECEBE89675BD&factor=1.25&id=7e357b92b998a1b2

[18:14:01.0000] <othermaciej>
(our view for a direct link to a video file uses the <video> element with controls)

[18:14:02.0000] <nessy>
I just checked with my examples - you are right

[18:15:00.0000] <nessy>
do you know what it looks like on the iPad?

[18:16:00.0000] <nessy>
(btw: that link doesn't load for me - but any video example will do)

[18:20:00.0000] <TabAtkins_>
Looks like Youtube expires direct links pretty quickly.

[18:21:00.0000] <nessy>
so, Hixie, I guess I look at the chapter markers in the same way that I look at the @controls - should be able to be enabled and disabled

[18:21:01.0000] <Hixie>
nessy: i guess i still don't understand what it would mean to disable them

[18:21:02.0000] <nessy>
make them disappear

[18:21:03.0000] <nessy>
on a mobile phone for example?

[18:22:00.0000] <Hixie>
nessy: wouldn't that by a UA issue?

[18:22:01.0000] <Hixie>
nessy: when "disabled" would you still want the controls to allow the user to use them to jump to a chapter?

[18:22:02.0000] <othermaciej>
nessy: you should be able to check in the iPad simulator, this video also claims to describe what it looks like: http://www.youtube.com/watch?v=ZpNhcuT8Wk8

[18:22:03.0000] <nessy>
optimally, yes

[18:23:00.0000] <nessy>
thanks othermaciej

[18:23:01.0000] <othermaciej>
I guess there is no "controls" example there

[18:23:02.0000] <Hixie>
nessy: would it still fire events?

[18:23:03.0000] <othermaciej>
but I bet you can find video or a screenshot on the Web

[18:23:04.0000] <othermaciej>
I think inline video playback on iPad is similar to desktop Safari

[18:23:05.0000] <nessy>
yeah, probably more so than iPhone or iPod

[18:24:00.0000] <nessy>
Hixie, I'm careful about not removing too much functionality and therefore flexibility

[18:25:00.0000] <nessy>
but I don't think we need to worry too much about this right now - I'm sure the first implementations will give us a better insight

[18:25:01.0000] <nessy>
I haven't done enough experiments with chapter markers myself on this to be sure

[18:25:02.0000] <Hixie>
well it makes a huge difference in the design of the markup

[18:25:03.0000] <Hixie>
which is why i need to worry about this now :-)

[18:27:00.0000] <nessy>
but will you handle chapters differently to captions and subtitles etc?

[18:27:01.0000] <Hixie>
this question applies to all the tracks

[18:27:02.0000] <nessy>
ok

[18:28:00.0000] <Hixie>
but only chapters seem to make this complicated

[18:29:00.0000] <nessy>
so the real question is: "if i need enabled="true"/enabled="false" vs just enabled="" "  ?

[18:29:01.0000] <Hixie>
yeah

[18:30:00.0000] <nessy>
we're talking about an attribute here?

[18:30:01.0000] <nessy>
so, if it's a boolean attribute, then removing it has the same effect as enabled="false" ?

[18:33:00.0000] <Hixie>
not necessarily

[18:33:01.0000] <Hixie>
there are three states

[18:33:02.0000] <Hixie>
enabled, disabled, and default

[18:34:00.0000] <Hixie>
for example most captions are disabled by default, except if the user has requested captions by default

[18:35:00.0000] <Hixie>
the question is whether we need the "disabled" state

[18:35:01.0000] <nessy>
ah, the three states from yesterday?

[18:35:02.0000] <Hixie>
not really

[18:35:03.0000] <nessy>
you might have chapters in different languages?

[18:35:04.0000] <Hixie>
this is more about what gets downloaded

[18:35:05.0000] <Hixie>
hm, interesting

[18:36:00.0000] <Hixie>
how would a user pick the language of chapters he wanted?

[18:36:01.0000] <nessy>
same as subtitles or captions

[18:36:02.0000] <Hixie>
i've never seen any dvd tv or online media player with a "chapter titles language" option

[18:36:03.0000] <nessy>
I'm hoping there will be a default language preference for time-aligned text and the browser does a pre-set

[18:36:04.0000] <nessy>
that the user can override

[18:37:00.0000] <nessy>
let me check TED - I think they change languages also for chapters

[18:40:00.0000] <nessy>
e.g. http://www.ted.com/talks/irwin_redlener_warns_of_nuclear_terrorism.html

[18:41:00.0000] <nessy>
so, they don't change the language of the chapter markers, but I think it's really irritating

[18:41:01.0000] <nessy>
they even change the text on the right in the about box, so having the chapters still in English seems illogical to me

[18:42:00.0000] <Hixie>
ok, so i guess the default for chapters is to be disabled except if the language matches the user's language somehow, with a max of one enabled?

[18:43:00.0000] <Hixie>
if we do that, does that mean we no longer need a disabled state?

[18:43:01.0000] <nessy>
what is the state if they are disabled?

[18:44:00.0000] <Hixie>
not downloaded, don't fire events

[18:44:01.0000] <Hixie>
for subtitles, available to be turned on from the UI

[18:44:02.0000] <nessy>
I'm still strugging to follow your three state / two state model, in particular when you say it's different to what we discussed yesterday

[18:45:00.0000] <nessy>
sorry for being difficult

[18:46:00.0000] <Hixie>
i'm not really expressing myself very well

[18:46:01.0000] <Hixie>
mostly because i'm still trying to clarify it in my mind

[18:46:02.0000] <nessy>
designing a spec is difficult :)

[18:47:00.0000] <nessy>
so, I understand there are three states:

[18:47:01.0000] <Hixie>
yesterday we were talking about do-not-download, download-and-show, and download-and-hide. I think we can do away with the download-and-show vs download-and-hide distinction because it applies at the level of what kind of timed track it is -- e.g. metadata is never download-and-show, and subtitles are never download-and-hide

[18:47:02.0000] <Hixie>
but that's the actual state of the track while the video is playing

[18:47:03.0000] <nessy>
indeed, those three :)

[18:48:00.0000] <Hixie>
there's a related concern, which is what the initial state should be

[18:48:01.0000] <nessy>
what if a user interacts and turns subtitles off while they are playing?

[18:48:02.0000] <nessy>
then it would go from download-and-show to downloaded-and-hidden sorta?

[18:49:00.0000] <Hixie>
that'd be do-not-download again, although it'd be already downloaded

[18:49:01.0000] <Hixie>
but it wouldn't fire events

[18:49:02.0000] <Hixie>
anyway there are basically two possible initial states, the same two states we end up with up there: do-not-download-or-fire-events, and download-and-show/hide-as-appropriate,-and-fire-events

[18:50:00.0000] <nessy>
let me dissect this a bit

[18:50:01.0000] <Hixie>
however, there are three ways to express that for an author: do-whatever-is-appropriate (the default), never-download-this-even-if-the-user-wants-it, and always-download-and-show/hide-as-appropriate,-and-fire-events-even-if-the-user-doesn't-want-it

[18:51:00.0000] <Hixie>
my question is whether there's a use case for "never-download-this-even-if-the-user-wants-it"

[18:51:01.0000] <Hixie>
there's obviously a use case for "do-whatever-is-appropriate (the default)"

[18:51:02.0000] <Hixie>
and there's a use case for "always-download-and-show/hide-as-appropriate,-and-fire-events-even-if-the-user-doesn't-want-it", namely custom controls where the user has set defaults the UA doesn't know about

[18:51:03.0000] <nessy>
probably not - if a user wants it, it should be downloaded and shown

[18:53:00.0000] <nessy>
I'm currently seeing 4 different, but dependent dimensions: download (yes/no), display (yes/no), fire events (yes/no), user wants (yes/no)

[18:53:01.0000] <nessy>
and we call them all "enable/disable"

[18:54:00.0000] <Hixie>
heh

[18:54:01.0000] <Hixie>
yeah

[18:54:02.0000] <Hixie>
the matrix of those options is pretty sparse though in terms of what we need to support

[18:54:03.0000] <Hixie>
e.g. download=no display=yes makes no sense :-)

[18:54:04.0000] <nessy>
I'm trying to follow your linearisation of that matrix ;)

[18:55:00.0000] <nessy>
when you way "user wants" - that's either "author wants events" or "user wants display", right? so we can get rid of that dimension?

[18:55:01.0000] <nessy>
s/way/say/

[18:57:00.0000] <Hixie>
"user wants" is always "user wants display", i think, for some definition of "display" that includes chapter navigation appearing in a drop-down menu

[18:58:00.0000] <nessy>
or elsewhere on screen

[18:59:00.0000] <nessy>
then "never-download-this-even-if-the-user-wants-it-displayed" surely is not viable

[19:02:00.0000] <Hixie>
k

[19:07:00.0000] <nessy>
so, I can see three definite states and two not so sure

[19:07:01.0000] <nessy>
1) no download, no display, no events

[19:07:02.0000] <nessy>
2) download, display, events

[19:07:03.0000] <nessy>
3) download, no display, events

[19:07:04.0000] <nessy>
maybe also

[19:08:00.0000] <nessy>
4) download, display, no events (though as maciej said yesterday, this is an optimisation in the browser and doesn't really need its own state)

[19:08:01.0000] <nessy>
and something of a "pre-load" nature could be:

[19:08:02.0000] <nessy>
5) download, no display, no events

[19:09:00.0000] <nessy>
5) would also be the state it gets into when a track is turned off manually

[19:10:00.0000] <nessy>
Hixie: does that match your model?

[19:14:00.0000] <Hixie>
1 and 5 are basically the same state

[19:14:01.0000] <nessy>
so, maybe these for attribute values: "disabled" (1), "preload" (5), "enabled" (3), "display" (2)

[19:14:02.0000] <Hixie>
4 i agree is unnecessary

[19:15:00.0000] <Hixie>
2 and 3 i think depends more on the kind of track than anything else

[19:15:01.0000] <nessy>
1 and 5 make a bandwidth use difference

[19:15:02.0000] <Hixie>
you wouldn't ever put a subtitle track into state 3, or a metadata track into state 1, right?

[19:15:03.0000] <Hixie>
er

[19:15:04.0000] <Hixie>
you wouldn't ever put a subtitle track into state 3, or a metadata track into state 2, right?

[19:15:05.0000] <nessy>
I'd hesitate to define behaviour based on the kind of track something is - I would prefer to do a generic model that works with all tracks

[19:16:00.0000] <Hixie>
i'm alergic to generic models :-P

[19:16:01.0000] <Hixie>
5 is a state you'll only reach by going through one of the others -- i don't see a use case for reaching it otherwise

[19:16:02.0000] <nessy>
get over it :P

[19:16:03.0000] <Hixie>
i'm really very skeptical of (over-)generalising things

[19:16:04.0000] <Hixie>
if there is a use case for the more general model i'm all for it

[19:17:00.0000] <nessy>
I'm not overly fussed about state 5 as an initial state - it could be used to encourage the browser to download the file in preparation for later use

[19:17:01.0000] <nessy>
but then, text resources are rather small generally, so shouldn't be too much of an issue

[19:17:02.0000] <Hixie>
there's another state too: a) download, no display, render to audio, events

[19:17:03.0000] <nessy>
such as?

[19:17:04.0000] <Hixie>
and another: c) download, no display, show on timeline, events

[19:17:05.0000] <Hixie>
a = text audio description tracks

[19:17:06.0000] <Hixie>
c = chapter titles

[19:18:00.0000] <nessy>
when I say "display" I mean "rendering"

[19:19:00.0000] <Hixie>
right, but i'm arguing that the kinds of rendering -- including no rendering -- are the same axis as the kind of track

[19:19:01.0000] <nessy>
but here I follow your model of - "what rendering means is obvious from the kind of track it is" :P

[19:19:02.0000] <Hixie>
imho there's no need to distinguish at the authoring level a chapter navigation track from a subtitles track

[19:19:03.0000] <Hixie>
they can both be SRT or whatever

[19:19:04.0000] <nessy>
well, my argument is that the UA will know what rendering means for the type of resource - it's not up to the author to decide

[19:20:00.0000] <Hixie>
when you label it as a chapter track, then you ignore positioning and just display it on the timeline

[19:20:01.0000] <Hixie>
how?

[19:20:02.0000] <Hixie>
if the author doesn't say what's what kind of track, how does the UA know?

[19:20:03.0000] <nessy>
that's what the @role attribute is for

[19:21:00.0000] <nessy>
I hope you're not intending to also mix the @role values into the @enabled attribute values?

[19:21:01.0000] <Hixie>
i'm arguing that you're mixing them up in your description

[19:21:02.0000] <Hixie>
but that they are distinct

[19:22:00.0000] <Hixie>
whether something is firing events and being appropriately displayed or not is orthogonal to how it should be displayed

[19:22:01.0000] <nessy>
yes, they are distinct

[19:23:00.0000] <nessy>
indeed - I was arguing because it's an orthogonal matter how they should be displayed, I don't have to worry about that dimension when I talk about firing events/displaying it

[19:23:01.0000] <nessy>
the means of display is on a different and totally independent dimension

[19:24:00.0000] <nessy>
but we should get back to the question "you wouldn't ever put a subtitle track into state 3, or a metadata track into state 2, right?"

[19:25:00.0000] <nessy>
I can see a subtitle track in state (3) if I don't want to UA to display the subtitles, but rather want as an author to display them myself

[19:25:01.0000] <nessy>
in an appropriate manner to my custom interface

[19:26:00.0000] <Hixie>
wouldn't that just be a metadata track?

[19:26:01.0000] <nessy>
as for state (2) on metadata - indeed there would not be a UA default display for metadata - the state is still valid, though, it's just an empty display

[19:26:02.0000] <nessy>
no, it's still subtitles

[19:26:03.0000] <Hixie>
hmm

[19:26:04.0000] <Hixie>
i'll have to think on this some more

[19:26:05.0000] <Hixie>
gotta go for dinner now though

[19:27:00.0000] <Hixie>
thanks for being patient with my learning stage :-)

[19:27:01.0000] <nessy>
thanks for being patient with my lack of using the right terminology in places :)

[19:27:02.0000] <nessy>
enjoy dinner!

[19:29:00.0000] <nessy>
just to finish off on the subtitles: even if they are not displayed and waiting for JavaScript to display them, they could still be in some kind of menu for the user to turn on as subtitles

[19:29:01.0000] <nessy>
anyway - talk later :)

[23:17:00.0000] <boblet>
gsnedders: request: textarea for HTML5 Outliner to test pages & snippets via copy&paste

[23:20:00.0000] <zcorpan>
oooh, microsoft writes a new Appendix C for html5

[23:24:00.0000] <annevk>
it's really that bad?

[23:24:01.0000] <annevk>
good times

[23:24:02.0000] <annevk>
it's on my todo list

[23:25:00.0000] <zcorpan>
i didn't say it was bad

[23:25:01.0000] <zcorpan>
although it has some errors

[23:25:02.0000] <zcorpan>
wait utf-16 is not valid for html5?

[23:27:00.0000] <zcorpan>
/me only sees "Authors should not use UTF-32, as the encoding detection algorithms described in this specification intentionally do not distinguish it from UTF-16."

[23:30:00.0000] <zcorpan>
oh sam meant in the <meta>

[23:38:00.0000] <annevk>
Appendix C was bad

[23:56:00.0000] <zcorpan>
NoCredXMLHttpRequest

[23:57:00.0000] <zcorpan>
NoCookiesXMLHttpRequest

[23:58:00.0000] <zcorpan>
XMLHttpRequestOmitCookies (to make them next to each other in indexes and autocomplete)

[00:00:00.0000] <annevk>
it's not just cookies and credentials is also somewhat too limited for what it does

[00:01:00.0000] <zcorpan>
just need a name that people will understand what it does (roughly)

[00:01:01.0000] <zcorpan>
and when to use which

[00:01:02.0000] <annevk>
ContextFreeXMLHttpRequest

[00:01:03.0000] <annevk>
but it's a bit long

[00:02:00.0000] <zcorpan>
XMLHttpRequestNoContext

[00:02:01.0000] <annevk>
you think it should be at the end?

[00:02:02.0000] <zcorpan>
yeah

[00:02:03.0000] <zcorpan>
to make them next to each other in indexes and autocomplete

[00:05:00.0000] <annevk>
suggested on list

[00:10:00.0000] <annevk>
thanks zcorpan!

[00:11:00.0000] <annevk>
hopefully this'll work

[00:23:00.0000] <jstar-taiwan>
hi, how does <canvas> degrade on handheld/mobile devices ?

[00:24:00.0000] <annevk>
it doesn't degrade really

[00:24:01.0000] <annevk>
it works

[00:25:00.0000] <zcorpan>
on opera mini it becomes a static image (but onclick works by reloading the page or so) iirc

[00:25:01.0000] <jstar-taiwan>
annevk, how is that possible on devices build before html5 epoque ?

[00:26:00.0000] <zcorpan>
jstar-taiwan: opera works on old devices ;)

[00:26:01.0000] <zcorpan>
jstar-taiwan: in browsers that don't support canvas, the contents are shown

[00:28:00.0000] <jstar-taiwan>
zcorpan, so users have to get a browser that support html5 and canvas to this this content otherwise nothin ?

[00:30:00.0000] <zcorpan>
yeah, although the fallback content doesn't have to be nothing, it could be something useful

[00:32:00.0000] <annevk>
filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=9577

[00:33:00.0000] <annevk>
I suppose an alternative solution is to convince Jonas to break some pages temporarily

[00:33:01.0000] <jstar-taiwan>
zcorpan, oh~ the content inside the <canvas> is a fallback like the <video> or this kind of thing 

[00:34:00.0000] <zcorpan>
jstar-taiwan: yes

[00:35:00.0000] <annevk>
othermaciej, you mean XMLHttpRequestGuest would also be ok?

[00:35:01.0000] <annevk>
othermaciej, with "also suffix variants"

[00:36:00.0000] <annevk>
othermaciej, that sounds relatively short and nice

[00:36:01.0000] <othermaciej>
annevk: yeah, I think any of the variants I gave would work as a suffix as well as a prefix

[00:36:02.0000] <othermaciej>
annevk: Mark Miller's original suggestion way back when was GuestXHR

[00:36:03.0000] <zcorpan>
XMLHttpRequestForPorn!

[00:40:00.0000] <zcorpan>
"guest" is a term forum sites use when you're not logged in

[00:41:00.0000] <annevk>
"anon"/"anonymous" too...

[00:41:01.0000] <othermaciej>
zcorpan: yeah, I think anything that conveys that "not logged in" idea is good

[00:42:00.0000] <othermaciej>
it's hard to find a term that is more accurate yet still succinct

[00:42:01.0000] <othermaciej>
in a way a string parameter might be nicer because something like new XMLHttpRequest("NoCredentials") is a little easier to read than new NoCredentialsXMLHttpRequest() or new XMLHttpRequestNoCredentials()

[00:43:00.0000] <annevk>
we could do that too

[00:43:01.0000] <jstar-taiwan>
string parameters is more flexible/extendable

[00:44:00.0000] <jstar-taiwan>
that mean prone to standards extension as well as proprieatries one

[00:44:01.0000] <annevk>
sure, but hopefully we are not going to overload this much more

[00:44:02.0000] <annevk>
requests are already frigging complicated

[00:44:03.0000] <zcorpan>
argument is harder to feature test

[00:44:04.0000] <annevk>
also a good point

[00:44:05.0000] <annevk>
actually, in this case you could just do try/catch

[00:45:00.0000] <zcorpan>
if (window.XMLHttpRequestGuest) is similar to how people test for XMLHttpRequest support today

[00:45:01.0000] <othermaciej>
zcorpan: good point, that's definitely a point in favor of a named constructor

[00:46:00.0000] <othermaciej>
another possibility is to make the new constructor a property of the main one, so it could be XMLHttpRequest.NoCredentials()

[00:46:01.0000] <othermaciej>
not sure that reads very well though

[00:46:02.0000] <othermaciej>
new XMLHttpRequest.NoCredentials()

[00:46:03.0000] <othermaciej>
new XMLHttpRequest.Guest()

[00:46:04.0000] <othermaciej>
/me shrugs

[00:46:05.0000] <zcorpan>
seems weird

[00:47:00.0000] <othermaciej>
it would read better with a :: instead of a . but you can't do that in JAvaScript

[00:49:00.0000] <annevk>
that would require new Web IDL syntax too

[00:49:01.0000] <annevk>
currently all constructors are exposed on the global object

[00:51:00.0000] <othermaciej>
a separate constructor that constructs objects of the same time might be bad for non-JS languages like Java

[00:52:00.0000] <zcorpan>
XMLHttpRequestGuest could be an empty interface that inherits from XMLHttpRequest

[00:53:00.0000] <othermaciej>
why do you guys prefer suffixed names to prefixed?

[00:53:01.0000] <othermaciej>
(not saying I disagree, just curious)

[00:54:00.0000] <zcorpan>
to make them next to each other in indexes and autocomplete

[00:54:01.0000] <annevk>
othermaciej, we already have that with e.g. new Image(); also, I'd very much encourage non-JavaScript languages to stay the hell away from XMLHttpRequest :)

[00:54:02.0000] <TabAtkins__>
Best. Webdesign. Ever. --> http://www.yvettesbridalformal.com/

[00:54:03.0000] <othermaciej>
I'm not sure "next to each other in autocomplete" is good

[00:54:04.0000] <othermaciej>
indexes, I can see

[00:54:05.0000] <zcorpan>
same reason microdata attributes share the same prefix

[00:55:00.0000] <othermaciej>
annevk: people are interested in making complete WebKit DOM bindings for other languages such as Objective-C or Python

[00:55:01.0000] <othermaciej>
not to run inside Web pages, but rather so that an app using WebKit can exercise the full power of the Web platform without having to go through JavaScript

[00:55:02.0000] <othermaciej>
annevk: also, new Image() is just a shortcut, it doesn't provide a genuine new capability

[00:56:00.0000] <othermaciej>
you could always createElement("img") and then set its src attribute by hand

[00:57:00.0000] <annevk>
hmm, apparently NamedConstructor is JavaScript-only

[00:57:01.0000] <annevk>
would be interesting to know if it's really impossible for other languages to have something similar

[00:58:00.0000] <othermaciej>
zcorpan's "empty interface subclass" idea seems good to me

[00:58:01.0000] <othermaciej>
that would work for any language that can support constructors at all (which is pretty much any language of interest I think)

[00:59:00.0000] <annevk>
if we subclass it should probably be prefixed

[00:59:01.0000] <annevk>
that's more or less the convention

[00:59:02.0000] <zcorpan>
also allows checking if (foo instanceof XMLHttpRequestGuest)

[00:59:03.0000] <zcorpan>
annevk: true

[01:00:00.0000] <annevk>
though we can also suffix and use implements

[01:00:01.0000] <annevk>
that's also a convention :)

[01:00:02.0000] <othermaciej>
I think a subclass makes sense logically, since there is a change in behavior

[01:01:00.0000] <annevk>
i'm not sure it's much different from using XHR inside a sandboxed iframe to be honest

[01:03:00.0000] <othermaciej>
that's like saying CORS is not a change in behavior because it's not much different from using XHR inside an off-site iframe and talking to it with postMessage

[01:04:00.0000] <annevk>
it's different for e.g. <img> requests that are later painted on <canvas>

[01:05:00.0000] <annevk>
the XHR in a sandboxed iframe scenario can be implemented by a simple library and requires no cooperation

[01:07:00.0000] <othermaciej>
annevk: my point is just that "you could do this a different way" doesn't refute the statement that it's a change in behavior

[01:08:00.0000] <othermaciej>
annevk: a change in behavior could be represented as a mode flag instead of a subclass, but since there is a separate constructor anyway, it may as well be a subclass

[01:09:00.0000] <annevk>
guess overall i'm still a little unsure whether this will actually see adoption

[01:09:01.0000] <annevk>
and similar problems we solved with modeflags so far

[01:09:02.0000] <annevk>
such as withCredentials

[01:11:00.0000] <othermaciej>
annevk: seems like Caja will want to use it at least - not sure how much that counts for

[01:12:00.0000] <annevk>
if Caja was really important I'm sure Chromium would've had the feature by now

[01:13:00.0000] <othermaciej>
yes, because Google's product strategy is 100% in sync across all parts of the company :-)

[01:17:00.0000] <annevk>
I guess not, but I haven't heard a whole lot about it to be honest

[01:19:00.0000] <annevk>
also, if you put Caja in sandboxed iframes, is it still needed?

[01:19:01.0000] <annevk>
in fact, do you still need Caja if you have sandboxed iframes?

[01:19:02.0000] <othermaciej>
I think sandboxed iframes are a better solution than Caja

[01:19:03.0000] <othermaciej>
you could use both together

[01:20:00.0000] <othermaciej>
I am not sure the Caja people like that idea

[01:20:01.0000] <annevk>
they don't seem to really like the Web

[01:21:00.0000] <othermaciej>
what do you mean?

[01:22:00.0000] <annevk>
never mind

[01:23:00.0000] <othermaciej>
I think they definitely dislike the Web security model

[01:23:01.0000] <othermaciej>
which is reasonable, because the Web security model is ridiculous

[01:23:02.0000] <othermaciej>
I just happen to think that working with it is more sensible than fighting against it

[01:25:00.0000] <annevk>
yup

[01:53:00.0000] <gsnedders>
boblet: Frequent request. I don't really intend on spending time on the outliner :)

[02:08:00.0000] <boblet>
gsnedders: heh. What about a longer URL input field then?

[02:10:00.0000] <gsnedders>
Too much effort ;P

[02:10:01.0000] <gsnedders>
/me thought he'd published the source somewhere

[02:11:00.0000] <boblet>
ouch! :)

[02:12:00.0000] <boblet>
np yo

[02:18:00.0000] <annevk>
from http-state

[02:18:01.0000] <annevk>
 <meta http-equiv="Set-Cookie" content="foo=bar; path=/; domain=.some.tld">

[02:18:02.0000] <annevk>
apparently works

[02:18:03.0000] <annevk>
good times

[02:20:00.0000] <zcorpan>
hah

[02:21:00.0000] <annevk>
i'll file a bug

[02:23:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9578

[02:32:00.0000] <zcorpan>
doesn't seem to be a spec bug since the spec only supports a limited set of values

[02:32:01.0000] <annevk>
browsers do too

[02:32:02.0000] <annevk>
and why wouldn't it be a spec bug then?

[02:33:00.0000] <zcorpan>
set-cookie is not in the spec's set of values

[02:33:01.0000] <hsivonen>
/me wonders how Chrome Frame will interact with IE9

[02:34:00.0000] <annevk>
zcorpan, you think that browsers should remove support for set-cookie?

[02:34:01.0000] <hsivonen>
(the full IE9--not the preview)

[02:34:02.0000] <zcorpan>
annevk: yeah. unless it's required for compat

[02:35:00.0000] <hsivonen>
chances are it is...

[02:35:01.0000] <hsivonen>
I don't know, though

[02:35:02.0000] <annevk>
zcorpan, seems likely to be required

[02:35:03.0000] <zcorpan>
http://www.google.com/codesearch?q=http-equiv%5Cs%2A%3D%5Cs%2A%28%22%7C%27%29%3Fset-cookie&hl=en&btnG=Search+Code

[02:35:04.0000] <zcorpan>
crap

[02:35:05.0000] <annevk>
zcorpan, but otherwise we can use it as reminder to file bugs on WebKit/Gecko

[02:36:00.0000] <annevk>
it's pretty easy to support though

[02:36:01.0000] <annevk>
just invokes the same algorithm as document.cookie when parsing the element

[02:37:00.0000] <othermaciej>
annevk: that's so evil

[02:38:00.0000] <othermaciej>
which browsers support it?

[02:38:01.0000] <annevk>
http://trac.webkit.org/browser/trunk/WebCore/dom/Document.cpp#L2256 :)

[02:38:02.0000] <annevk>
reportedly also Mozilla; haven't found anything about IE but given the search query above I suspect everyone does

[02:39:00.0000] <othermaciej>
I had no idea we did that

[02:39:01.0000] <othermaciej>
I wonder when that was added, and why

[02:46:00.0000] <annevk>
seems to be pre-WebKit

[02:46:01.0000] <asmodai>
/me wonders why Firefox is leaking memory like crazy.

[02:46:02.0000] <asmodai>
It's now at 1,6 GB of mem :|

[02:48:00.0000] <zcorpan>
    92:   <xsl:if test="not($server-side-transformation)">

[02:48:01.0000] <zcorpan>
    93:     <meta http-equiv="Set-Cookie" content="xslt=true; path=/; " />

[02:48:02.0000] <zcorpan>
    94:   </xsl:if>

[02:48:03.0000] <zcorpan>
   213:   <noscript>

[02:48:04.0000] <zcorpan>
   214:     <meta http-equiv="Set-Cookie" content="js=false; path=/; " />

[02:48:05.0000] <zcorpan>
   215:   </noscript>

[02:49:00.0000] <asmodai>
Did that Lorentz build get merged into 3.6.4?

[02:51:00.0000] <asmodai>
ah yes it did

[02:53:00.0000] <annevk>
zcorpan, kawaii

[03:48:00.0000] <asmodai>
Mmm, I guess it was Skype's plugin/addon that might have caused the insane leakage. Got to love how they install it during normal install

[03:49:00.0000] <roc>
we have a bug on file where certain tests related to cross-origin video loading fail if and only if Skype is running

[03:49:01.0000] <roc>
I fear it

[03:50:00.0000] <asmodai>
Well, in this case it's FF 3.6.4 (beta) and installed Skype yesterday and it seems to not ask whether or not to install those toolbars.

[03:51:00.0000] <asmodai>
Also sucks, on my laptop, when I listen to music on the headphones and then start Skype, bloody thing manages to temporarily push my audio over the speakers >_<

[03:51:01.0000] <asmodai>
oh crap, maybe I celebrated too early

[03:51:02.0000] <asmodai>
I'm not doing anything in FF at the moment and I see it grow by 200-300 kilobyte increments

[03:52:00.0000] <asmodai>
from 500 MB to 512 MB in just a minute or so. :S

[04:00:00.0000] <asmodai>
mmm, 590 MB now. so that's like 78 MB in 5 minutes

[04:00:01.0000] <asmodai>
Not good at all.

[04:02:00.0000] <hsivonen>
asmodai: Firefox 3.6.latest?

[04:02:01.0000] <zcorpan>
try the Quit Firefox menu item

[04:13:00.0000] <gsnedders>
I hear Opera works well.

[04:16:00.0000] <zcorpan>
http://sam.zoy.org/wtfpl/ lol

[04:16:01.0000] <Philip`>
gsnedders: I thought you usually spent all day finding ways in which Opera works badly

[04:17:00.0000] <zcorpan>
that's why it works well

[04:18:00.0000] <Philip`>
http://tunes.org/legalese/bugroff.html - these kinds of licenses seem quite popular

[04:19:00.0000] <jgraham>
Assuming gsnedders work well

[04:32:00.0000] <Lachy>
the WTFPL licence is self contradictory. It says you can do whatever you want to without restriction, but then imposes the restriction that modified versions of the licence must also change the name.

[04:33:00.0000] <zcorpan>
yeah

[04:33:01.0000] <gsnedders>
WTF?

[04:37:00.0000] <Dashiva>
Lachy: If you can do whatever you want, surely that includes ignoring the license

[04:38:00.0000] <asmodai>
hsivonen: 3.6.4 beta 1 I think it's called

[04:38:01.0000] <Lachy>
Dashiva, yes, it does. The FAQ below it even says you can relicence it if you don't like the terms

[04:38:02.0000] <hsivonen>
those licenses may seem cute as long as you don't want your software to be included in Debian or otherwise be built-upon

[04:39:00.0000] <Philip`>
Lachy: Why do you assume the WTFPL license text is licensed under the WTFPL?

[04:39:01.0000] <Philip`>
The GPL text isn't GPL and doesn't permit any modifications at all

[04:39:02.0000] <zcorpan>
hsivonen: you can just change the license

[04:40:00.0000] <Dashiva>
hsivonen: It claims WTFPL-licensed programs are in debian already

[04:42:00.0000] <Lachy>
Philip`, the way the WTFPL is written, it says that the licence itself is the work being licenced. It doesn't say that the terms apply to the work being licenced.

[04:42:01.0000] <Lachy>
i.e. where it says "Everyone is permitted to copy and distribute verbatim or modified  copies of this *license document*," (emphasis added)

[04:43:00.0000] <Dashiva>
It's intended to be used as a header, surely

[04:43:01.0000] <Philip`>
Lachy: The way I read it, that paragraph is a standalone statement about the license text and has no relevance to what the license is licensing

[04:44:00.0000] <Dashiva>
Although that doesn't grant any rights to compile code

[04:44:01.0000] <Philip`>
(It's the same as the first paragraph of http://www.gnu.org/licenses/gpl.html)

[04:47:00.0000] <Lachy>
hmm, ok. It's still a really poorly written licence

[04:47:01.0000] <Dashiva>
It doesn't seem to actually solve anything compared to BSD/MIT

[04:47:02.0000] <Philip`>
Lachy: That's basically the whole point of the license

[04:48:00.0000] <Philip`>
It's like BSD/MIT without trying to be legally correct and unambiguous

[04:49:00.0000] <Philip`>
If they'd added a dozen more lines to explain exactly what was being licensed then there'd no longer be any reason to use it

[04:51:00.0000] <Dashiva>
Perhaps we're going about this the wrong way. Rather than license the work, license the author

[04:51:01.0000] <Dashiva>
"I hereby license everything I ever produce to everyone everywhere"

[04:52:00.0000] <Philip`>
What does "produce" mean?

[04:52:01.0000] <Philip`>
Would it include your children?

[04:53:00.0000] <Dashiva>
That would infringe on the children's rights, though

[04:54:00.0000] <Philip`>
Seems better to stick with the GPL's approach of using the tools provided by copyright law to circumvent the original goals of copyright law, rather than doing something totally different and pretending the legal details don't matter

[04:55:00.0000] <Dashiva>
How about getting a notarized letter proving you died 150 years ago?

[04:56:00.0000] <Lachy>
actually, it could be argued that the GPL does more to promote the original goal of copyright, which was to promote creativity, that what current copyright laws actually do themselves.

[04:58:00.0000] <Dashiva>
Original goals aren't part of the law in many jurisdictions

[04:58:01.0000] <Philip`>
/me is considering the original goal to be the reduction of copying

[04:58:02.0000] <Philip`>
(rather than the more indirect original goals that were hoped to be achieved by doing that)

[04:58:03.0000] <Lachy>
well, I guess, if you consider what the printing presses wanted to maintain their monopolies for

[05:00:00.0000] <Lachy>
but the US constituion's statement about copyright is to promote the progress of science and the useful arts (or something like that), the the GPL most certainly does more for progress than overly restrictive copyright laws

[05:00:01.0000] <gsnedders>
/me wouldn't say copyright laws really prohibit that

[05:01:00.0000] <gsnedders>
Copying work isn't needed for the progress of science and the useful arts

[05:01:01.0000] <gsnedders>
What needed is the ability to recreate experiements in science (which is not copying in a copyright sense), and in the arts each work tends to be a word in and of itself

[05:01:02.0000] <Dashiva>
Nobody cares about the constitution unless they can interpret it to agree with their own view

[05:02:00.0000] <jgraham>
gsnedders: That's not really true

[05:02:01.0000] <Philip`>
Dashiva: Sounds like the HTML WG charter

[05:02:02.0000] <Dashiva>
I wonder why

[05:03:00.0000] <Dashiva>
And the original intent is often the first victim in such interpretations

[05:04:00.0000] <Lachy>
Dashiva, that's true. The supreme court blatently ignored the original intent of the constitution when they ruled that that micky mouse protection act wasn't unconstitutional, even though any sane person would realise that it was.

[05:05:00.0000] <gsnedders>
/me wonders what things in the US constitution haven't been contradicted

[05:05:01.0000] <gsnedders>
/me doubts there are any

[05:06:00.0000] <Philip`>
If the US constitution is so infallible that we should still strictly follow exactly what is says, how come it had to be amended 27 times?

[05:06:01.0000] <Lachy>
no-one claims it was infallable.

[05:06:02.0000] <Dashiva>
I'm sure _someone_ does

[05:07:00.0000] <jcranmer>
gsnedders: the 13th amendment?

[05:07:01.0000] <jcranmer>
Philip`: turn it on its head

[05:07:02.0000] <gsnedders>
/me looks that up

[05:07:03.0000] <Dashiva>
jcranmer: I think illegal immigrants handle that fine

[05:07:04.0000] <jcranmer>
how many constitutions have only been amended 27 times?

[05:08:00.0000] <hsivonen>
Does anyone happen to know if https://bugzilla.mozilla.org/show_bug.cgi?id=537948 is Web compat-relevant?

[05:08:01.0000] <Dashiva>
Not to mention the old health care system :)

[05:08:02.0000] <Lachy>
gsnedders, it abolishes slavery

[05:09:00.0000] <jcranmer>
there's also the all-important 3rd Amendment

[05:09:01.0000] <jcranmer>
(no quartering of troops, basically)

[05:10:00.0000] <Philip`>
jcranmer: As far as I know it could be like HTML4 where people have just been too lazy to fix all the bugs in the document

[05:10:01.0000] <Dashiva>
It's a real shame 2nd and 3rd amendments weren't updated properly once America left the old age

[05:11:00.0000] <jcranmer>
in general, the US constitution is quite flexible since it doesn't hardcode a lot of stuff

[05:11:01.0000] <Philip`>
jcranmer: I thought quartering was what was done to people like Guy Fawkes, not to troops

[05:11:02.0000] <gsnedders>
The fact the fifth amendment doesn't apply to anyone on US soil is somewhat disturbing

[05:12:00.0000] <gsnedders>
Likewise the sixth

[05:12:01.0000] <jcranmer>
Philip`: quartering == provide quarters forcibly

[05:12:02.0000] <Philip`>
Oh, okay

[05:13:00.0000] <jcranmer>
I doubt the text of the 16th or 17th amendments have caused lots of court cases

[05:13:01.0000] <jcranmer>
although their contents aren't exactly happy for all people

[05:14:00.0000] <jgraham>
I find it vaugely amusing when americans talk about the founding fathers as if they were god-like visionaries

[05:15:00.0000] <jcranmer>
eh, it's all a bunch of romanticism

[05:15:01.0000] <Dashiva>
They were quite visionary compared to current politicians, though

[05:15:02.0000] <Philip`>
When I think about the founding fathers, I always think of Day of The Tentacle

[05:16:00.0000] <jgraham>
Dashiva: Hard to say because the circumstances are quite different

[05:16:01.0000] <Dashiva>
Not talking about potential results, but actual results

[05:16:02.0000] <jgraham>
Possibly they are just as visionary as today's would-be politicians that no one will vote for

[05:17:00.0000] <gsnedders>
It also seems to be hold far too much as perfect and a reason to argue against rational changes in the US

[05:18:00.0000] <Dashiva>
You don't need the constitution for that, you just make people believe it says whatever you claim

[05:19:00.0000] <Dashiva>
I don't recall the poll, but apparently quite a few believe it declares the US to be a christian nation

[05:19:01.0000] <jcranmer>
well, the supreme court once had to make a ruling on whether or not something violated the preamble to the constitution

[05:20:00.0000] <jcranmer>
but you get nutjobs in any sufficiently large country

[05:20:01.0000] <jgraham>
The vatican is full of nut jobs

[05:20:02.0000] <jcranmer>
especially one as individualisitic as the US

[05:20:03.0000] <jgraham>
So the bar for "sufficiently large" is rather small

[05:20:04.0000] <workmad3>
you get nutjobs in any country

[05:20:05.0000] <Philip`>
When the visionary people were writing it, did they expect it would be still applicable and still used two centuries later?

[05:21:00.0000] <workmad3>
you just get more, and more extreme, nut jobs in larger countries

[05:21:01.0000] <jgraham>
Philip`: Probably no more so than people who wrote COBOL in the 60s

[05:21:02.0000] <jcranmer>
Philip`: probably not

[05:21:03.0000] <gsnedders>
If they had been so visionary, would they have written it as they did knowing it would still be used two centires later?

[05:22:00.0000] <jcranmer>
Thomas Jefferson thought countries needed a revolution every few decades

[05:22:01.0000] <Philip`>
Did they intend it to be rescinded at some point in the future before it became too irrelevant?

[05:22:02.0000] <Dashiva>
Surely they expected more amendments

[05:23:00.0000] <jcranmer>
well, they punted majorly on the slavery issue

[05:23:01.0000] <Philip`>
They should have an added an expiry date of 50 years after the writing, so that the country would descend into chaos and start the next revolution

[05:23:02.0000] <gsnedders>
/me wonders if anyone is still counted as 2/3rds of a person in a census

[05:23:03.0000] <jcranmer>
no one was ever counted as 2/3

[05:23:04.0000] <jcranmer>
it was 3/5

[05:23:05.0000] <Dashiva>
Philip`: They'd just extend it

[05:24:00.0000] <gsnedders>
But actually, I think the wording was those who weren't free

[05:24:01.0000] <gsnedders>
jcranmer: Gah. The exact fraction doesn't matte.r

[05:24:02.0000] <Dashiva>
As I understand it, the two houses are allowed to make new rules every session, but they just carry over the old ones out of habit

[05:24:03.0000] <Philip`>
Dashiva: They could have made extensions unconstitutional

[05:25:00.0000] <jcranmer>
/me points those Europeans to the European constitution^W^WLisbon treaty

[05:25:01.0000] <gsnedders>
The fact taht the president alone has a veto is somewhat distrubing, though

[05:25:02.0000] <Dashiva>
Philip`: They would pass an amendment then

[05:26:00.0000] <Dashiva>
Or maybe they just wouldn't tell anyone and keep going out of habit, not like anything would change

[05:26:01.0000] <gsnedders>
jcranmer: That isn't two centuries old :)

[05:26:02.0000] <jcranmer>
gsnedders: see how much longer it is than the US constitution

[05:27:00.0000] <jgraham>
And hasn't really been accepted

[05:27:01.0000] <jcranmer>
well, I would refer to California's constitution, which has something like 2000 amendments

[05:27:02.0000] <gsnedders>
Some things are somewhat disturbing, though, like countries opting out of making the Charter of Fundamental Rights law

[05:27:03.0000] <Philip`>
Dashiva: They should have done what people who wrote COBOL in the 60s did and make their stuff unable to cope beyond the end of the century and make it really expensive to fix

[05:28:00.0000] <jcranmer>
but everyone knows California is just a big conglomeration of lunatics

[05:28:01.0000] <Dashiva>
Philip`: Kind of... the way it is right now?

[05:28:02.0000] <Philip`>
Oh, okay, so they did do that

[05:29:00.0000] <Philip`>
and people just put up with the really expensive fixes instead of taking it as an opportunity to start from scratch

[05:29:01.0000] <Philip`>
so I guess that's not the best plan

[05:29:02.0000] <Dashiva>
Maybe if they one day manage to get a sensible election system

[05:31:00.0000] <gsnedders>
That sounds over ambitious

[05:32:00.0000] <Dashiva>
It looks like it might happen before I die in the UK, so you never know

[05:33:00.0000] <gsnedders>
Well, the Scottish parliment has a sane election system

[05:33:01.0000] <gsnedders>
There again, that favoured the Tories who were in power in Westminster when the Scotland Act was passed

[05:33:02.0000] <gsnedders>
(they are still only the fouth largest party, though)

[05:34:00.0000] <gsnedders>
Ah, no, they're up to third equal now

[05:35:00.0000] <Dashiva>
Although perhaps the supreme court is a bigger problem than the electoral stuff

[05:35:01.0000] <gsnedders>
The new Supreme Court, or?

[05:36:00.0000] <Dashiva>
The way it's updated

[05:39:00.0000] <Philip`>
Dashiva: You're going to die in the UK?

[05:40:00.0000] <Dashiva>
Of course

[05:40:01.0000] <Dashiva>
While singing "Back in the UK" by Scooter

[05:40:02.0000] <gsnedders>
/me again gets confused by the title of the Duke of Rothesay

[05:40:03.0000] <gsnedders>
(which is the title of the Prince of Wales while in Scotland)

[05:41:00.0000] <gsnedders>
(or rather, the title of the heir apparent)

[05:44:00.0000] <Philip`>
gsnedders: Why is that confusing? It's just like someone being both the King of Gondor and the Heir of Isildur

[05:44:01.0000] <gsnedders>
Philip`: It's not so much confusing because of that, it's confusing because I'm always sure I should know who the Duke of Rothesay is but I can never remember

[05:45:00.0000] <hsivonen>
http://html5.org/tools/web-apps-tracker?from=4802&to=4803 is going to need some unit testing in the html5lib repo...

[05:45:01.0000] <hsivonen>
I take it that html5lib hasn't implemented that spec change yet?

[05:45:02.0000] <gsnedders>
hsivonen: And the CR changes

[05:45:03.0000] <gsnedders>
hsivonen: No, it hasn't

[05:45:04.0000] <gsnedders>
hsivonen: AFAIK it's no more up-to-date than the script changes

[05:50:00.0000] <hsivonen>
ok.

[05:53:00.0000] <hsivonen>
hmm. the Opera Mobile emulator is 32bit only

[05:56:00.0000] <nileshtrivedi>
will there be an input type for address ?

[06:00:00.0000] <hsivonen>
nileshtrivedi: do you mean postal address?

[06:00:01.0000] <nileshtrivedi>
yes.. not latitude/longitude, which is a separate issue

[06:01:00.0000] <hsivonen>
nileshtrivedi: postal addresses are an i18n rathole

[06:01:01.0000] <hsivonen>
nileshtrivedi: so unlikely

[06:01:02.0000] <nileshtrivedi>
all right :)

[06:02:00.0000] <nileshtrivedi>
how about the latitude/longitude ? will there be a input type="geo" ?

[06:13:00.0000] <hsivonen>
nileshtrivedi: unlikely. when the page asks for the user's location, there's already the geolocation API

[06:13:01.0000] <hsivonen>
nileshtrivedi: when the page asks for another location, browsers aren't well equipped to provide a client-side UI

[06:13:02.0000] <hsivonen>
nileshtrivedi: sites are better equipped to import Google/Bing/Yahoo! maps as part of the site UI

[06:14:00.0000] <nileshtrivedi>
hsivonen: but the geolocation api would require javascript enabled, isn't it?

[06:15:00.0000] <hsivonen>
nileshtrivedi: yes

[06:16:00.0000] <hsivonen>
can anyone think of a case where end tag token processing starts in a non-"in foreign" mode and the token gets reprocessed in the "in foreign" mode?

[06:16:01.0000] <hsivonen>
I can't think of such a case.

[06:54:00.0000] <hsivonen>
the code for spec rev 4803 isn't nice...

[06:58:00.0000] <hendry>
hsivonen: you don't need libxml for parsing HTML5 do you?

[06:59:00.0000] <hsivonen>
hendry: no. why?

[06:59:01.0000] <hendry>
hsivonen: is there a nice C implemenation of the HTML5 parsing model somewhere I wonder?

[06:59:02.0000] <hendry>
hsivonen: wondering why webkit seems to depend on it

[06:59:03.0000] <gsnedders>
hendry: XML parser, AFAIK

[07:01:00.0000] <hendry>
gsnedders: can't the HTML5 parsiing model replace what libxml does in webkit?

[07:02:00.0000] <gsnedders>
hendry: No. The HTML5 Parsing doesn't define how to parse XML.

[07:03:00.0000] <hendry>
so what XML does webkit need to be able to parse then? XHTML?

[07:03:01.0000] <hsivonen>
hendry: there's a nice C++ implementation :-)

[07:04:00.0000] <hsivonen>
hendry: if you want C, you can put a C API wrapper around the C++

[07:04:01.0000] <hendry>
hsivonen: C++ *shiver*, ok pls point me to it.

[07:04:02.0000] <hendry>
hsivonen: is the big idea webkit/gecko converge on the same model right?

[07:04:03.0000] <hsivonen>
hendry: http://hg.mozilla.org/projects/htmlparser/

[07:04:04.0000] <hendry>
hsivonen: i know you were doing some experiments. is FF using html5 parser yet

[07:05:00.0000] <hsivonen>
to get a non-Gecko-dependent version of the HTML5 parser, you need to create your own version of CppTypes.java so that the translator gets non-NSPR/XPCOM types

[07:05:01.0000] <hsivonen>
hendry: not on by default yet

[07:06:00.0000] <hsivonen>
getting a non-Gecko-dependent version isn't "supported" in the sense that I haven't tried it yet even for a toy project

[07:06:01.0000] <hsivonen>
hendry: you can be the first to try!

[07:08:00.0000] <hsivonen>
hendry: and yes, the big idea is to get *all* browsers to converge on standard parsing

[07:08:01.0000] <hsivonen>
hendry: in the case of Gecko, there are two other big ideas: 1) getting rid of the old HTML parser and 2) putting HTML parsing off the main thread

[07:09:00.0000] <hendry>
hsivonen: not sure how to clone hg.mozilla.org/projects/htmlparser/ or where exactly to find the cpp, only see java

[07:09:01.0000] <hendry>
hsivonen: is webkit following suit?

[07:09:02.0000] <hsivonen>
hendry: to get the CPP, you run Main.java under translator-src

[07:10:00.0000] <hendry>
hsivonen: what?? you generate cpp from java?

[07:10:01.0000] <hsivonen>
if you want to experiment with the help of an automagic makefile, you can clone http://hg.mozilla.org/mozilla-central/

[07:10:02.0000] <hsivonen>
hendry: yes

[07:11:00.0000] <hsivonen>
...and see http://hg.mozilla.org/mozilla-central/file/456b934bab9e/parser/html/java/README.txt

[07:11:01.0000] <hsivonen>
that should get you going with the dependencies of the translator itself

[07:12:00.0000] <hendry>
omg

[07:12:01.0000] <hsivonen>
hendry: writing Java is much nicer than writing C++ :-)

[07:12:02.0000] <hendry>
hsivonen: has anyone considered doing this in C?

[07:13:00.0000] <hsivonen>
hendry: yes. takkaria and ment

[07:13:01.0000] <hsivonen>
neither of whom seems to be on IRC right now

[07:13:02.0000] <hendry>
hsivonen: checkout http://harmful.cat-v.org/software/java

[07:13:03.0000] <Philip`>
There's http://www.netsurf-browser.org/projects/hubbub/

[07:14:00.0000] <Philip`>
Don't know what the current status of it is, though

[07:15:00.0000] <jmb>
Philip`: there'll be a patch release at the weekend, is the current status of it

[07:16:00.0000] <hsivonen>
hendry: lot of boring enterprise software gets written in Java. good software gets written in Java, too.

[07:16:01.0000] <hendry>
jmb: looks good

[07:16:02.0000] <Philip`>
jmb: Given that the last release listed there was 0.0.1 in April 2009, it gives the impression of being a bit dead

[07:17:00.0000] <hendry>
jmb: are you in Southhampton?

[07:17:01.0000] <jmb>
Philip`: well, yes :)

[07:17:02.0000] <Philip`>
jmb: so a new release sounds good :-)

[07:17:03.0000] <jmb>
Philip`: but that's because people have been busy with other things

[07:17:04.0000] <jmb>
not because it's been abandoned

[07:18:00.0000] <hendry>
are there html tests to see if a html5 parser is compliant to the model?

[07:19:00.0000] <Philip`>
jmb: Was it stable and complete (per the spec from that time period) before people became busy with other things?

[07:19:01.0000] <jmb>
Philip`: yes, provided you didn't want fragment parsing or script support

[07:19:02.0000] <Philip`>
I think it's unclear from the page whether it's a decent usable library or an early incomplete alpha-style release

[07:20:00.0000] <hsivonen>
hendry: html5lib hosts the de facto test suite

[07:20:01.0000] <hsivonen>
hendry: it's incomplete of course

[07:21:00.0000] <hsivonen>
right now, I wish Someone Else had already written tests for spec rev 4803

[07:21:01.0000] <gsnedders>
heh

[07:21:02.0000] <gsnedders>
I remember thinking that when implementing all the script states in the tokenizer

[07:21:03.0000] <jmb>
Philip`: that's a fair accusation :)

[07:21:04.0000] <gsnedders>
Except wishing someone else had done the impl

[07:22:00.0000] <hendry>
hsivonen: ok i see code.google.com/p/html5lib/source/browse/#hg/testdata (guess what browser i'm using that removes http:// from my unix style copies)

[07:22:01.0000] <Philip`>
If it's a decent usable library then someone (else) should write a Python wrapper for it :-)

[07:23:00.0000] <hsivonen>
hendry: I hear you can get the http:// back in the latest build

[07:23:01.0000] <jmb>
Philip`: heh

[07:23:02.0000] <hsivonen>
Philip`: you could make the Java to C++ translator target python C bindings!

[07:25:00.0000] <Philip`>
Are Python people happy with modules written in C++?

[07:25:01.0000] <hsivonen>
who's gonna know what's behind the ABI?

[07:25:02.0000] <hsivonen>
it's not like this is for the AppStore

[07:26:00.0000] <Philip`>
Could it ever be bundled with Python as an alternative to sgmllib?

[07:26:01.0000] <hsivonen>
oh

[07:26:02.0000] <gsnedders>
Philip`: US is going in and depends on C++

[07:26:03.0000] <Philip`>
Presumably Python people would care in that case

[07:26:04.0000] <Philip`>
gsnedders: US?

[07:26:05.0000] <gsnedders>
Unladen Swallow

[07:26:06.0000] <jgraham>
Well that is different

[07:27:00.0000] <jgraham>
Somwhat

[07:27:01.0000] <gsnedders>
Yeah, but if the VM relies upon C++ then stdlib relying upon it will be no problem

[07:27:02.0000] <jgraham>
It seems more useful to target Python 2.x in the short term anyway

[07:27:03.0000] <gsnedders>
But that VM dependancy is optional anyway

[07:28:00.0000] <hsivonen>
is Unladen Swallow for Python 3?

[07:28:01.0000] <jgraham>
Yes

[07:28:02.0000] <gsnedders>
The original branch was for 2, it will be merged into a future 3 release

[07:28:03.0000] <jgraham>
The official version at least

[07:28:04.0000] <hsivonen>
I guess Python's XHTML2 is different in the sense that it gets serious resources from Google

[07:28:05.0000] <gsnedders>
(2.7 is the final major release of Python 2, and they just missed the deadline for that)

[07:28:06.0000] <Philip`>
http://www.python.org/dev/peps/pep-3146/#addition-of-c-to-cpython

[07:29:00.0000] <Philip`>
"C++ can be disabled via ./configure --without-llvm, which even omits the dependency on libstdc++"

[07:29:01.0000] <gsnedders>
15:34 < gsnedders> But that VM dependancy is optional anyway

[07:29:02.0000] <Philip`>
That seems to suggest they like C++ to still be optional

[07:29:03.0000] <hsivonen>
gsnedders: final as in HTML 4.01 or Perl 5.0?

[07:29:04.0000] <gsnedders>
hsivonen: final as in Perl 5.0

[07:29:05.0000] <gsnedders>
Well, somewhere between them, actually

[07:29:06.0000] <Philip`>
and presumably to be optional without affecting runtime behaviour (except performance)

[07:29:07.0000] <gsnedders>
Basically bugfix only

[07:29:08.0000] <gsnedders>
No new features

[07:32:00.0000] <jgraham>
gsnedders: That is only the plan

[07:35:00.0000] <Philip`>
/me guesses that converting the Java-to-C++ translator into a Java-to-C translator would be non-trivial and probably not worthwhile since there's no good reason for people to not allow C++

[07:38:00.0000] <hsivonen>
Philip`: just use CFront :-)

[07:40:00.0000] <Philip`>
I think that's possibly a tiny bit obsolete now

[07:40:01.0000] <Philip`>
though it looks like the Comeau C++ compiler is modern and can emit C

[07:42:00.0000] <Philip`>
Oh, but apparently not standalone C - the output is platform-specific, and the license doesn't allow you to use the C code on other platforms

[07:42:01.0000] <gsnedders>
I wonder what about using Clang and trying to create C from LLVM bytecode

[07:43:00.0000] <Philip`>
Just write a bytecode interpreter in C

[08:12:00.0000] <Philip`>
/me has now set up a system where JS files contain fragments of Relax NG which get extracted by a C++ program and then processed by a couple of hundred lines of XSLT to generate XHTML

[08:12:01.0000] <Philip`>
Someone please save me, I'm being infected by XML :-(

[08:13:00.0000] <jgraham>
Philip`: Why?

[08:13:01.0000] <Philip`>
Why save me?

[08:13:02.0000] <jgraham>
No, I think it is quite obvious you need to be saved

[08:13:03.0000] <jgraham>
Anything else would be immkoral

[08:13:04.0000] <jgraham>
*immoral

[08:13:05.0000] <Philip`>
It seemed like a good idea at the time (i.e. yesterday)

[08:14:00.0000] <Philip`>
and it still does, which is worrying

[08:35:00.0000] <zcorpan>
Philip`: it's a bad idea

[08:40:00.0000] <hsivonen>
I *think* I found 3 bugs in rev 4803.

[08:40:01.0000] <hsivonen>
or alternatively, my reading comprehension has gotten really bad

[08:40:02.0000] <hsivonen>
so I still haven't written test cases

[08:55:00.0000] <volkmar>
Hixie: ping

[09:42:00.0000] <TabAtkins>
Welp, just discovered that I can't install Fireworks on Linux, and so can't possibly attend a class about Fireworks.  Great.

[09:42:01.0000] <Dashiva>
"I have no faith in the W3C (what was the last useful thing they helped create... XML namespaces?)."

[09:42:02.0000] <TabAtkins>
...wow.

[09:44:00.0000] <daedb>
Dashiva: Where's that from?

[09:46:00.0000] <Dashiva>
http://minimsft.blogspot.com/2010/04/microsoft-fy10q3-results.html

[09:47:00.0000] <TabAtkins>
"even if they deign its presence with faint of disdain" Someone needs to put the dictionary down and talk at the level they understand.

[09:50:00.0000] <remysharp>
If I wanted to use this picture: http://www.whatwg.org/specs/web-apps/current-work/images/drawImage.png from the spec, who's the best person to speak to, or is there a standard answer?

[09:50:01.0000] <remysharp>
sorry - use the picture in a book whilst explaining drawImage

[09:51:00.0000] <TabAtkins>
I'd assume that it's covered under the WHATWG's HTML5 license, which means freely redistributable.

[09:51:01.0000] <TabAtkins>
Otherwise, Hixie.

[09:53:00.0000] <remysharp>
The spec reads "You are granted a license to use, reproduce and create derivative works of this document." so it's fair to say I'm good then, unless Hixie want's to interject.

[09:53:01.0000] <remysharp>
TabAtkins: ta

[09:53:02.0000] <TabAtkins>
Yes, then.  You can use the picture however you want, as a derivative work.

[09:56:00.0000] <jgraham>
TabAtkins: You thought that adobe made products that work on Linux?

[09:56:01.0000] <TabAtkins>
Man, I dunno.  I live in a pretend world of magic and unicorns where everything just works everywhere.

[09:57:00.0000] <jgraham>
Given how well Adobe and Apple get along these days it is irnonic that the only piece of software I wouldn't want to move away from OSX for is Lightroom

[09:57:01.0000] <jgraham>
*ironic

[09:58:00.0000] <jgraham>
If they ever ship a linux version I will be all over it

[09:58:01.0000] <jgraham>
But they never will

[09:59:00.0000] <jgraham>
And there is no sign that anyone else will write a similarly good application that does work on Linux

[10:00:00.0000] <jgraham>
And I no longer understand how anyone does anything on Windows

[10:00:01.0000] <jgraham>
Because every time I try I want to kill people

[10:00:02.0000] <TabAtkins>
I find it interesting how my workflow changes based on what incidental tools I have available.

[10:01:00.0000] <jgraham>
e.g.?

[10:01:01.0000] <TabAtkins>
On Windows, where multiple desktops are hard to do, I had AHK to let me minimize anything with F4, and just worked with 2 monitors and a single desktop fine.

[10:01:02.0000] <TabAtkins>
Here on Linux, I haven't gone to the trouble of finding anything that lets me remap keycombos like that, but I don't need it, since I just set up 6 desktops and never change what's on them.

[10:04:00.0000] <TabAtkins>
Topic change: RDFa Core 1.1 now defines what appears to be 3 separate mechanisms to make the syntax look more like Microdata.

[10:07:00.0000] <jgraham>
If micro data is good, having three ways to be like it is exp(3) time better, surely

[10:07:01.0000] <jgraham>
s/ //

[10:07:02.0000] <TabAtkins>
That's the worst magic regexp ever.

[10:08:00.0000] <jgraham>
It's perfectly acceptable magic

[10:08:01.0000] <Dashiva>
It matches the wrong space, though

[10:08:02.0000] <jgraham>
No, that's the magic of it

[10:08:03.0000] <Dashiva>
Clearly you're using a different version of magic

[10:09:00.0000] <TabAtkins>
At least *try* to make it unambiguous.  s/ time/time/, if I'm understanding you correctly.

[10:09:01.0000] <TabAtkins>
Also, I should totally go eat some breakfast before they close down the kitchen on me again and I'm left with nothing but a banana and a bottled frappucino.

[10:10:00.0000] <Dashiva>
You can eat when you're dead

[10:10:01.0000] <TabAtkins>
That will be sooner than desired, if I follow that advice.

[10:10:02.0000] <jgraham>
Only if you are a zombie

[10:10:03.0000] <jgraham>
(do zombies eat?)

[10:10:04.0000] <TabAtkins>
Alternately, I can eat now, which reduces the time between me and food even further.

[10:10:05.0000] <Dashiva>
Sounds self-defeating

[10:10:06.0000] <TabAtkins>
jgraham: Of course they do.  The exact metabolic processes depend on your zombie type.

[10:10:07.0000] <Dashiva>
Eating to avoid eating

[10:11:00.0000] <jgraham>
TabAtkins: Did you join public-html-testsuite?

[10:11:01.0000] <TabAtkins>
No, I'm wantint to eat so I can eat more.

[10:11:02.0000] <TabAtkins>
jgraham: Yes.

[10:11:03.0000] <jgraham>
TabAtkins: Great

[10:11:04.0000] <Dashiva>
But that's a circular argument

[10:11:05.0000] <Dashiva>
You want to eat because you want to eat

[10:11:06.0000] <jgraham>
TabAtkins: Did you see any of the discussion about the test harness for DOM tests?

[10:11:07.0000] <TabAtkins>
Dashiva: So is my digestive tract, if you include sufficient recycling steps between two of the steps.

[10:11:08.0000] <TabAtkins>
jgraham: Haven't really read it.

[10:12:00.0000] <jgraham>
I figure you might be somewhat interested if this is one of the things you are working on

[10:12:01.0000] <TabAtkins>
Yeah, I am, and need to.

[10:12:02.0000] <jgraham>
Great

[10:13:00.0000] <jgraham>
(I just said that again, but there is nothing wrong with squaring the greatness)

[10:13:01.0000] <TabAtkins>
Right after saying that she doesn't feel comfortable engaging in discussion on the mailing list and would rather communicate purely through Change Proposals, she dumps a wall of text on the mailing list not in the form of a Change Proposal.

[10:14:00.0000] <TabAtkins>
(Her saving grace is probably that said wall of text doesn't constitute "discussion".)

[10:15:00.0000] <Philip`>
jgraham: Zombies eat brains (unless you protect your brains with sufficient defensive gardening skills), obviously

[10:15:01.0000] <TabAtkins>
Philip`: Unless they're one of the "living zombie" variants, which typically eat all uninfected flesh.

[10:17:00.0000] <Dashiva>
Most zombie ecologies fail to address the inherent problems of the brain duality

[10:17:01.0000] <TabAtkins>
? Which duality?

[10:17:02.0000] <Dashiva>
Brains are both their weak point and their sustenance

[10:18:00.0000] <Philip`>
http://www.cramitwithwalnuts.com/wp-content/gallery/plants-vs-zombies/ate_your_brains.jpg - proof

[10:18:01.0000] <Dashiva>
So if zombies eat the brains of their victims, but destroying the brain destroys the zombie, how can a victim become a zombie?

[10:18:02.0000] <TabAtkins>
That's well-established; partial contact does the trick.

[10:19:00.0000] <Dashiva>
Partial contact is not a reasonable assumption

[10:19:01.0000] <Dashiva>
Once a zombie has contact with a brain, it has no reason to stop consuming

[10:19:02.0000] <TabAtkins>
With a brain, sure.  Once you hit that point, it's likely that you're fully disabled and at the nonexistent mercy of the undead.

[10:20:00.0000] <TabAtkins>
I'm talking about a conflict where the living sustains a bite or scratch, but is then able to escape.

[10:20:01.0000] <Dashiva>
This is, however, a minority situation

[10:20:02.0000] <TabAtkins>
I disagree.  Zombies are slow enough that, until you get swarm situations, most physical contact is escapable.

[10:21:00.0000] <TabAtkins>
You're also discounting the original infection point, which may create a large pool of zombies to start with.

[10:21:01.0000] <Dashiva>
There's your swarm

[10:21:02.0000] <Dashiva>
Furthermore, escape is only an option if there is somewhere to escape to

[10:22:00.0000] <TabAtkins>
Not necessarily.  The initial group will, by default, spread out.  Swarms occur when a large number are all focused on a relatively small number of targets.

[10:23:00.0000] <TabAtkins>
If the zombification event hits everyone in an area, there are no immediate targets.  If it doesn't, there are likely a number of targets, which will flee in multiple directions and spread the initial swarm as well.

[10:24:00.0000] <Philip`>
Dashiva: You can escape to the Big Brother house

[10:24:01.0000] <gsnedders>
TabAtkins: So, what are you doing with the test suite stuff?

[10:24:02.0000] <Dashiva>
Philip`: That's a fate worse than (un)death

[10:24:03.0000] <TabAtkins>
gsnedders: Writing a new end-to-end harness for testing, which will swallow all the existing test suites and be runnable automatically in all browsers.

[10:25:00.0000] <TabAtkins>
At least, that's the eventual goal, and what I'm being paid for.

[10:25:01.0000] <Dashiva>
TabAtkins: That may hold for the very beginning, but secondary infections will have easy access to victims who trust them and expect nothing.

[10:26:00.0000] <gsnedders>
TabAtkins: For all tests, or just JS ones?

[10:26:01.0000] <TabAtkins>
Dashiva: The hunger doesn't come until zombification does, at which point victims will know something is off.  Some will still succumb, unbelieving that their loved one has joined the army of darkness, but many will escape after an initial contact, possibly gaining an infected injury and eventually zombifying themselves.

[10:27:00.0000] <TabAtkins>
gsnedders: All tests.

[10:27:01.0000] <gsnedders>
TabAtkins: So how are you dealing with visual tests?

[10:27:02.0000] <TabAtkins>
gsnedders: At the moment, we're shooting for the reftest format.  Future improvements may allow further expansion in this area.

[10:27:03.0000] <Dashiva>
Zombification is not easily distinguishable from illness, which on the contrary would draw people closer to provide care

[10:28:00.0000] <gsnedders>
TabAtkins: And how do you do the screenshots?

[10:28:01.0000] <gsnedders>
TabAtkins: In a cross-platform, cross-browser way?

[10:28:02.0000] <Dashiva>
(Perhaps we've gone far enough for now)

[10:28:03.0000] <TabAtkins>
gsnedders: That's the trick, right?

[10:28:04.0000] <TabAtkins>
The answer is "We'll figure it out when we get there."

[10:29:00.0000] <TabAtkins>
But ideas are circling around a screenshot api for yanking pixels from a section of the screen.

[10:29:01.0000] <Philip`>
Is it possible to make a test harness that is powerful enough to swallow all the existing test suites, but still simple enough that people will be willing to use it and write new tests for it?

[10:29:02.0000] <gsnedders>
TabAtkins: heh

[10:30:00.0000] <TabAtkins>
Part of the harness development is development of new testing APIs that all browsers will implement.  We had Hakon in the meeting about it, and he was enthusiastic on the matter.

[10:30:01.0000] <gsnedders>
TabAtkins: So, for the JS ones at least, see jgraham's email to public-html-testsuites

[10:30:02.0000] <TabAtkins>
Philip`: I'm betting that it is.

[10:30:03.0000] <TabAtkins>
gsnedders: Will do.

[10:31:00.0000] <gsnedders>
TabAtkins: You going to be at the CSS WG F2F in Oslo?

[10:31:01.0000] <TabAtkins>
Yes.

[10:31:02.0000] <gsnedders>
/me may try and coinside with you there

[10:31:03.0000] <TabAtkins>
That'd be cool.

[10:31:04.0000] <gsnedders>
(though I'll be in totally different meetings :))

[10:33:00.0000] <gsnedders>
(I'd probably also get lost in the office and never meet you)

[10:34:00.0000] <gsnedders>
Otherwise, adios for now

[10:45:00.0000] <KaOSoFt>
Good morning.

[13:43:00.0000] <TabAtkins>
air.mozilla.com for a live stream of Moz addressing FF complaints.

[13:43:01.0000] <TabAtkins>
Also: surprisingly funny!

[13:59:00.0000] <TabAtkins>
Anyone know at what point SafeCURIEs were introduced for RDFa?  I don't recall seeing them in the RDFa 1.0 spec, and here in 1.1 it's also mentioned as something unnecessary.

[14:00:00.0000] <TabAtkins>
(The things where you wrapped your CURIE in [] to keep it from being interpreted as a url.)

[14:01:00.0000] <Philip`>
They've always been there

[14:01:01.0000] <Philip`>
http://www.w3.org/TR/rdfa-syntax/#P_safe_curie

[14:01:02.0000] <TabAtkins>
I guess I must have just missed it, then.

[14:33:00.0000] <jgraham>
TabAtkins: To reiterate what gsnedders said earlier, a test harness for DOM tests is something we want sooner rather than later for the HTML5 testsuite

[14:34:00.0000] <jgraham>
So it would be good to coordinate on a relatively short timescale if you have particular requirements

[14:34:01.0000] <TabAtkins>
When you say "relatively short", is "3 months" acceptably short?

[14:35:00.0000] <jgraham>
/me has got quite strong opinions from seeing a whole variety of test harness styles in js libraries and so on

[14:35:01.0000] <jgraham>
(and browsers)

[14:35:02.0000] <TabAtkins>
/me is out to the dentist, will be back soon.

[14:36:00.0000] <jgraham>
TabAtkins: It depends what you mean by "3 Months". If you mean "a test harness that allows one to write DOM tests and has a relatively stable API" 3 months seems reasonable. If you mean "start thinking about requirements" it seems like a long time

[14:37:00.0000] <jgraham>
/me might also try to go to Oslo in August if there is some possibility of having a discussion

[14:37:01.0000] <jgraham>
(but that is a long time away)

[14:42:00.0000] <TabAtkins>
In 3 months I'll have an alpha that works from start to finish, though the actual pass/fail part will be manual.  After that I'll begin working on the automatic testing part.

[14:42:01.0000] <TabAtkins>
Now I'm actually gone.

[14:43:00.0000] <jgraham>
Now I'm actually confused about what your plan is

[14:44:00.0000] <jgraham>
I guess for CSS tests it makes some sort of sense to have manual pass/fail

[14:44:01.0000] <jgraham>
For dom tests it makes no sense whatsoever

[14:45:00.0000] <jgraham>
to the extent that I don't see why you would design the system with "manual pass/fail" as an incremental step

[14:49:00.0000] <jgraham>
TabAtkins: You are working in the open, right? Is there any chance of an email to some list outlining your exact plans at some point?

[14:50:00.0000] <Philip`>
What does a test harness that doesn't detect pass/fail actually do?

[14:51:00.0000] <Philip`>
That seems like the most important part of it

[14:51:01.0000] <jgraham>
Philip`: Presumably it loads tests and has a UI for a human to say whether they passes/failed

[14:52:00.0000] <jgraham>
For tests of rendering it makes some sense but it doesn't scale

[14:53:00.0000] <jgraham>
So you either have to do screenshots and mark pass/fail each time you get a new result or reftests

[14:53:01.0000] <jgraham>
and reftests are much less fragile

[14:55:00.0000] <Philip`>
jgraham: Wouldn't that just be something like an HTML file with a list of links to each test case, and maybe some iframes, and a dozen lines of CGI script to save and display the human-entered results?

[14:56:00.0000] <Philip`>
I'd assume it must be doing something amazingly fancier if it'll take three months of work :-)

[14:57:00.0000] <jgraham>
Philip`: The W3 testrunner thing as it currently stands has a webserver to provide the tests and a database for recording results

[14:58:00.0000] <jgraham>
Depending on the scope of the tests you might want the ability to significantly customise the server on a per-test basis (for HTTP testing)

[14:58:01.0000] <jgraham>
You might also want, say, a websockets server

[14:58:02.0000] <jgraham>
So, depending on the scope of the project, one can imagine a lot in infrastructure

[14:59:00.0000] <jgraham>
s/in/of/

[16:12:00.0000] <Hixie>
hm

[16:12:01.0000] <Hixie>
cross-origin timed tracks are going to be an issue

[16:28:00.0000] <nessy>
Hixie: yeah - hsivonen suggested iframe-like solution for that

[16:41:00.0000] <TabAtkins>
jgraham: Yeah, I'm in the open, and I can put a mail to the list.

[16:42:00.0000] <TabAtkins>
jgraham: Manual pass/fail is the first step because, frankly, that's the easiest thing to do.  Building it as a testing module, and then building additional modules that work automatically, enable me to get a verifiably working design quicker.

[16:42:01.0000] <TabAtkins>
Basically it's the simplest thing that works.

[16:43:00.0000] <TabAtkins>
Philip`: It doesn't take 3 months of work, it takes 3 months of my time, alongside all the other things I'm doing.  ^_^

[16:44:00.0000] <TabAtkins>
Suffice to say, though, that I'm not coming up with this stuff myself - arronei and fantasai, in charge of the CSS test suite, are basically codevelopers with me, and I've talked with plh as well about it.


2010-04-23
[17:00:00.0000] <nessy>
Hixie: do you mind if I upload some more examples to the wiki?

[17:00:01.0000] <nessy>
I have one with a pretty substantial background from a mobile phone app

[17:00:02.0000] <tiglionabbit>
do any browsers implement the non-breaking content in columns css yet?

[17:00:03.0000] <Hixie>
nessy: go for it

[17:01:00.0000] <TabAtkins>
tiglionabbit: I don't think so, though check latest Opera.

[17:01:01.0000] <tiglionabbit>
that's a shame.  multi column is kind of gimped without it, as far as i'm concerned

[17:01:02.0000] <tiglionabbit>
I guess I'll have to use inline blocks instead?

[17:03:00.0000] <TabAtkins>
I agree, it kind of is.  Inline-blocks sort of go in the opposite direction, though, don't they?  What display are you trying to achieve?

[17:03:01.0000] <TabAtkins>
(I've done some things where multicol or inline-block were both usable, and settled on inline-block.)

[17:04:00.0000] <tiglionabbit>
well, I I'm trying to do a fluid grid layout

[17:04:01.0000] <tiglionabbit>
of a bunch of boxes

[17:04:02.0000] <tiglionabbit>
I'd like them to be ordered the way that multi column does, but I suppose inline block ordering is decent

[17:04:03.0000] <TabAtkins>
Are the columns equal-width?  If so, then yeah, just do inline-block.

[17:04:04.0000] <tiglionabbit>
the problem with multi-column is that my blocks keep getting chopped in half

[17:05:00.0000] <TabAtkins>
Alternately, use Flexbox.  As long as you don't do anything too fancy, it'll work in FF, Safari, and Chrome.

[17:05:01.0000] <tiglionabbit>
multi-column can use different column widths?

[17:05:02.0000] <tiglionabbit>
what's flexbox?

[17:05:03.0000] <TabAtkins>
tiglionabbit: Actually, no, not yet, but it's planned to make it possible later.

[17:05:04.0000] <TabAtkins>
google for "css flexbox"

[17:06:00.0000] <TabAtkins>
Also, there's a relatively recent article on mozhacks about flexbox.

[17:08:00.0000] <tiglionabbit>
thanks =]

[17:12:00.0000] <nessy>
Hixie: check out http://wiki.whatwg.org/wiki/Use_cases_for_timed_tracks_rendered_over_video_by_the_UA - I've also added a new section - hope that fits your requirements

[17:13:00.0000] <Hixie>
i don't understand what's special about the background emphasis one

[17:13:01.0000] <Hixie>
how is it different from the other plain text ones?

[17:14:00.0000] <nessy>
ah just that on the Web you wouldn't be doing such a big black background thingy under the text, but on mobile it's important for readability

[17:14:01.0000] <nessy>
I don't mind adding them to the other plain text ones - just wanted to make sure that background requirement is considered

[17:14:02.0000] <nessy>
I don't really know how you arrived at your categories

[17:14:03.0000] <TabAtkins>
Because simply doing a stroked border on the text isn't enough on mobiles?

[17:15:00.0000] <nessy>
yeah - not readable

[17:15:01.0000] <TabAtkins>
Kk.

[17:16:00.0000] <nessy>
:)

[17:16:01.0000] <TabAtkins>
nessy, what's your name on the lists?

[17:16:02.0000] <nessy>
email? I'm Silvia Pfeiffer

[17:16:03.0000] <TabAtkins>
I thought so.  Just making sure.  It's helpful to connect my people-concepts together.  ^_^

[17:17:00.0000] <nessy>
surprised my whois doesn't tell - Adium is so much worse in comparison to xchat!

[17:17:01.0000] <TabAtkins>
Yeah, no useful info.

[17:18:00.0000] <nessy>
but I've given it to Adium - strange

[17:18:01.0000] <TabAtkins>
Your ircname is "Adium User".

[17:18:02.0000] <nessy>
yeah, not very useful

[17:23:00.0000] <Hixie>
nessy: ah, i see

[17:23:01.0000] <nessy>
hey, do you want this one, too? http://joeclark.org/access/crtc/CRTC-2008/reply/images/CCfoto-CBC-CC-ST-ManWithoutaFace.jpg

[17:23:02.0000] <Hixie>
i think that's more of an example of things going terribly wrong, but sure :-)

[17:23:03.0000] <nessy>
kinda shows the problem with open & closed captions, plus music notes

[17:23:04.0000] <nessy>
yeah, indeed!

[17:23:05.0000] <Hixie>
shows an example of what we should avoid :-)

[17:24:00.0000] <Hixie>
the musical notes thing is a non-issue, turns out Unicode has notes in it

[17:24:01.0000] <nessy>
ah, cool

[17:24:02.0000] <TabAtkins>
Unicode has everything in it.

[17:24:03.0000] <nessy>
I put it into positioning for now

[17:25:00.0000] <nessy>
joe clark seems to have a fair few, here's another one http://joeclark.org/access/crtc/CRTC-2008/reply/images/BCCS_unbelievably_wordy_speaker_ID_1.jpg

[17:26:00.0000] <TabAtkins>
That seems reasonable.  We can't auto-solve it, since there's no way to detect when baked-in captioning is present.

[17:26:01.0000] <nessy>
shows providing of speaker names

[17:26:02.0000] <Hixie>
TabAtkins: except klingon

[17:26:03.0000] <nessy>
lol

[17:26:04.0000] <TabAtkins>
Hixie: Well, someone's put that in one of the private use ranges.  You just need the right font.  ^_^

[17:26:05.0000] <AryehGregor>
Why did they reject Klingon?  :(

[17:26:06.0000] <Hixie>
(it has several fictional languages, but they apparently drew the line at klingon and rejected it)

[17:28:00.0000] <AryehGregor>
Does it have Tolkien's Elvish?

[17:29:00.0000] <TabAtkins>
Hixie: I don't think it includes any fictional languages.  Cirth and Tengwar are "proposed for inclusion", but not actually in the SMP.

[17:30:00.0000] <Philip`>
AryehGregor: Which of his Elvishes?

[17:30:01.0000] <AryehGregor>
Any.

[17:31:00.0000] <Philip`>
If somebody made up a new language in a style that emulated Tolkien, would it be an Elvish impersonator?

[17:33:00.0000] <TabAtkins>
The only language in Unicode which could be said to be "fictional" is the Shavian alphabet, but it wasn't invented for a story - it's just a semi-phonetic alphabet for English.

[17:33:01.0000] <AryehGregor>
http://www.magicdeckbuilder.com/public/cards/1-Illegal/1-Creatures/Elvish%20Impersonators.jpg

[17:33:02.0000] <Hixie>
i thought at least one fictional alphabet had made it in... i may be mistaken

[17:34:00.0000] <AryehGregor>
If only they didn't have to worry about UTF-16, they could have kept 2^32 possible characters and accepted all the writing systems out there.

[17:34:01.0000] <TabAtkins>
The Phaistos Disc script might be fictional.  Nobody's sure if it's a hoax or not yet.

[17:34:02.0000] <AryehGregor>
Of course, you'd think they'd have enough room even now for as many alphabets as they like.

[17:34:03.0000] <AryehGregor>
As long as they're not logographies.

[17:34:04.0000] <AryehGregor>
(I like the words "logography" and "syllabary".  They're just cool words.)

[17:34:05.0000] <TabAtkins>
They do.  We're only using Planes 0, 1, and 2.  We *might* use Plane 3.

[17:36:00.0000] <Hixie>
oh dear lord they accepted the emoticons block

[17:36:01.0000] <TabAtkins>
What's the range?

[17:36:02.0000] <Hixie>
1F600..1F64F

[17:36:03.0000] <Hixie>
that's a lot of emoticons

[17:37:00.0000] <AryehGregor>
You could add non-writing-system stuff like emoticons and icons without limit.

[17:37:01.0000] <TabAtkins>
Wikipedia is out of date!

[17:37:02.0000] <TabAtkins>
It only lists through Domino Tiles (1f09f) in plane 1.

[17:37:03.0000] <AryehGregor>
(I recently looked for icons of a lion or tablets to replace the star of David at aryeh.name, but didn't find any.  Will have to think of something else, I don't really like it much.)

[17:37:04.0000] <AryehGregor>
TabAtkins, {{sofixit}}

[17:38:00.0000] <TabAtkins>
I will, as soon as I find what's missing between 1f0a0 and 1f600.

[17:40:00.0000] <Hixie>
the emoticons aren't in yet

[17:40:01.0000] <Hixie>
they're approved for addition, but have yet to go through the ISO part of the proess

[17:40:02.0000] <Hixie>
see the pipeline table

[17:40:03.0000] <Hixie>
http://www.unicode.org/alloc/Pipeline.html

[17:42:00.0000] <TabAtkins>
Ah, yup.

[17:42:01.0000] <TabAtkins>
Ok, then I won't edit the wiki page yet.  Nothing beyond 1f0a0 is approved *yet*.

[17:43:00.0000] <Hixie>
/me finds http://unicode.org/~scherer/emoji4unicode/snapshot/NamesList.txt and starts drooling

[17:43:01.0000] <Hixie>
U+1F36F HONEY POT mmmmm

[17:44:00.0000] <Hixie>
there's a Facial parts symbols section

[17:44:01.0000] <Hixie>
lol

[17:45:00.0000] <Hixie>
U+1F47D EXTRATERRESTRIAL ALIEN!!!

[17:45:01.0000] <AryehGregor>
Are there drawings of what these will look like?

[17:45:02.0000] <AryehGregor>
1F307 SUNSET OVER BUILDINGS

[17:45:03.0000] <Hixie>
U+1F47E ALIEN MONSTER!!!

[17:45:04.0000] <Hixie>
i like that 1F479 and 1F47A are specifically japanese

[17:45:05.0000] <Hixie>
wouldn't want any EUROPEAN ogres

[17:46:00.0000] <AryehGregor>
1F3E8 HOTEL

[17:46:01.0000] <AryehGregor>
1F3E9 LOVE HOTEL

[17:46:02.0000] <AryehGregor>
1F3EA CONVENIENCE STORE

[17:46:03.0000] <AryehGregor>
1F3EC DEPARTMENT STORE

[17:46:04.0000] <Hixie>
um, 1F499 to 1F49C are identical except for colour

[17:46:05.0000] <Hixie>
that ought to be interesting

[17:46:06.0000] <Hixie>
up to now unicode has been black and white

[17:47:00.0000] <AryehGregor>
I have no idea what some of these will look like.

[17:47:01.0000] <AryehGregor>
1F452 WOMANS HAT

[17:47:02.0000] <AryehGregor>
Ooh, handy!  1F464 BUST IN SILHOUETTE

[17:47:03.0000] <AryehGregor>
 = guest account

[17:47:04.0000] <AryehGregor>
1F471 WESTERN PERSON

[17:47:05.0000] <AryehGregor>
. . .

[17:47:06.0000] <AryehGregor>
Is it supposed to emphasize the round eyes or something?

[17:47:07.0000] <TabAtkins>
8-|

[17:48:00.0000] <Hixie>
1F5FB to 1F5FE is going to be a terrible precedent

[17:48:01.0000] <Hixie>
prepare for planes 2-14 being full of landmark icons

[17:48:02.0000] <AryehGregor>
Seriously, though, are there reference pictures for these, or it's like every Unicode font will have a totally different-looking 1F47D EXTRATERRESTRIAL ALIEN?

[17:48:03.0000] <Hixie>
no idea

[17:48:04.0000] <TabAtkins>
I'm going with the latter.

[17:49:00.0000] <AryehGregor>
Some of these things will only be remotely legible at huge font sizes.

[17:49:01.0000] <AryehGregor>
What's up with how many of these are Japanese, anyway?

[17:50:00.0000] <AryehGregor>
Four out of five of "Cultural symbols" are Japanese, for instance.

[17:50:01.0000] <Hixie>
emojicons are all japanese

[17:50:02.0000] <Hixie>
nobody else seems to have gone quite as crazy as they have with this stuff

[17:50:03.0000] <AryehGregor>
1F5FE SILHOUETTE OF JAPAN?  So I guess we need a silhouette of all other countries too, and we should add new ones every time there's a successful war?

[17:51:00.0000] <Hixie>
like i said above :-)

[17:51:01.0000] <AryehGregor>
"SILHOUETTE OF FICTITIOUSTAN BETWEEN APRIL AND DECEMBER 2008"

[17:51:02.0000] <AryehGregor>
:/

[17:53:00.0000] <Hixie>
http://www.unicode.org/~scherer/emoji4unicode/snapshot/emojidata.pdf

[17:55:00.0000] <Hixie>
the ones in red are the new ones

[17:57:00.0000] <AryehGregor>
I'm torn between seeing this as insane and seeing it as awesome.

[17:57:01.0000] <Hixie>
yes!

[17:57:02.0000] <Hixie>
some of these are going to be really hard to explain in 1000 years to our children on another planet

[17:57:03.0000] <Hixie>
my sig is so changing once this goes through

[17:59:00.0000] <TabAtkins>
Aw man, so good.

[17:59:01.0000] <TabAtkins>
I like how U+1F473 ALIEN MONSTER looks completely different between the suggested symbol and the actual characters as implemented in two mobile phones.

[18:00:00.0000] <Hixie>
i love the separate "HOTEL" and "LOVE HOTEL" entries

[18:00:01.0000] <TabAtkins>
That's how you know it's Japanese.

[18:00:02.0000] <Hixie>
it really is hilarious to see what they include in emoji

[18:01:00.0000] <Hixie>
theres a dozen or two animals and another dozen insects, out of the millions of known animals, which itself represents like 2% of all animals

[18:01:01.0000] <TabAtkins>
U+1F4A9 PILE OF POO

[18:03:00.0000] <Hixie>
ah, i can finally have a single character as my away message when I sleep: U+1F4A4

[18:04:00.0000] <Hixie>
lol U+1F4B8 MONEY WITH WINGS

[18:04:01.0000] <Hixie>
man i almost feel bad about saying "no" to people who ask for crazy things in HTML now

[18:06:00.0000] <TabAtkins>
The difference is that nobody expects everyone to implement all of unicode.

[18:06:01.0000] <Hixie>
ok seriously this is quite insane. Now if aliens come, instead of sending them wikipedia, we can just ship them the Unicode table and they can pretty much learn everything they want about our culture from that.

[18:06:02.0000] <Paul_Irish>
http://paulirish.com/i/f6b0.png money with wings

[18:07:00.0000] <AryehGregor>
On a totally different off-topic note, is this true or what?  http://jeff-vogel.blogspot.com/2010/04/how-i-saved-gaming-industry-overnight.html

[18:07:01.0000] <AryehGregor>
The gaming industry is completely insane in how it throws away everything it creates.

[18:08:00.0000] <TabAtkins>
I agree.

[18:08:01.0000] <TabAtkins>
But I also agree with the commenter who says that if you stick with roughly the same technology, you actually have to offer a compelling story to get people to like your product.

[18:09:00.0000] <AryehGregor>
You don't need story, mechanics is fine.

[18:09:01.0000] <AryehGregor>
But you only redo the stuff you need to.

[18:09:02.0000] <Hixie>
lol, they didn't include FACE WITH ROLLING EYES. I guess the irony would have been too great?

[18:09:03.0000] <AryehGregor>
Valve is a great example.

[18:09:04.0000] <TabAtkins>
Sure, for some things.  Still more difficult than just paying people to make something completely new and shiny.

[18:10:00.0000] <TabAtkins>
That requires no creativity, just technical acumen.

[18:12:00.0000] <nessy>
at least this would be a simple way to get icons into captions

[18:12:01.0000] <nessy>
apart, of course, for company logos ;)

[18:12:02.0000] <TabAtkins>
That's what Plane 6 is for.

[18:16:00.0000] <othermaciej>
nessy: if you're on the Mac, I recommend Colloquy for IRC

[18:16:01.0000] <nessy>
ok, will try it out

[18:17:00.0000] <nessy>
does it hook into irc, jabber, gtalk, and twitter?

[18:17:01.0000] <nessy>
I like to have all that stuff in one app

[18:19:00.0000] <Hixie>
for jabber/gtalk to irc, i recommend bitlbee

[18:19:01.0000] <Hixie>
for twitter to irc, i recommend tircd

[18:19:02.0000] <Hixie>
both are compatible with any irc client

[18:20:00.0000] <Hixie>
since they're gateway servers

[18:20:01.0000] <othermaciej>
nessy: it only does IRC, but it's way better for IRC than Adium is

[18:21:00.0000] <nessy>
I used to use tircd, but am finding the integration of twitter directly in Adium really nice

[18:21:01.0000] <nessy>
then I don't have to run another program - it just does it

[18:26:00.0000] <TabAtkins>
Summary of RDFa 1.1: A ridiculously verbose way of expressing identical metadata relationships as Microdata.  With datatypes.

[18:28:00.0000] <Hixie>
TabAtkins: that's not really fair, rdfa does way more than microdata. For example it does per-field typing, and supports XML literals.

[18:28:01.0000] <TabAtkins>
I said per-field typing.  ^_^  XML literals is indeed something it has over Microdata.

[18:29:00.0000] <Hixie>
and it has full CURIE support

[18:29:01.0000] <TabAtkins>
That's not a benefit.

[18:29:02.0000] <Hixie>
oh

[18:29:03.0000] <Hixie>
well i'm sure it can be sold as a benefit

[18:29:04.0000] <Hixie>
does microdata do CURIES? I think not!

[18:29:05.0000] <Hixie>
clearly microdata is inferior

[18:29:06.0000] <TabAtkins>
Microdata does CURIES via an implicit @vocab mechanism.

[18:30:00.0000] <Hixie>
i have no idea what that means but it sounds good, i like it

[18:31:00.0000] <TabAtkins>
@vocab is just RDFa1.1's way of making you not need to use prefixes.  Any property nested in the @vocab element that doesn't have a prefix uses the @vocab value as its prefix.

[18:31:01.0000] <TabAtkins>
Alternately: the mechanism isn't implicit, but it's covered by @itemtype.

[18:31:02.0000] <TabAtkins>
Which I think is actually more accurate.

[18:33:00.0000] <TabAtkins>
The RDFa 1.1 spec doesn't do a great job of explaining what all the attributes do.  It took some hunting to discover that @typeof and @datatype are the same thing, with the former for the triple-subject and the latter for the triple-object.

[18:33:01.0000] <TabAtkins>
Also: chaining is a bad, confusing idea and I hate it.

[18:37:00.0000] <TabAtkins>
Additionally: very confused about why/how @src is always a subject, never an object, but @href is always an object, never a subject.

[18:38:00.0000] <TabAtkins>
That's not true.  @src can be an object too, at the same time as it's functioning as a subject and a scoping indicator.

[18:39:00.0000] <roc>
Colloquy is rubbish, it crashes, won't reconnect when I get back on the network, and gets into weird rendering states

[18:39:01.0000] <TabAtkins>
Just run irssi.

[18:39:02.0000] <roc>
of course I still use it, so I suppose it's not that bad or else I'm just a masochist

[18:41:00.0000] <Hixie>
nessy: thanks (re your mail)

[18:41:01.0000] <nessy>
thought you might like it :)

[18:41:02.0000] <nessy>
many ppl working behind the scenes to make this happen

[18:41:03.0000] <Hixie>
/me is wondering what to do about positioning captions

[18:41:04.0000] <Hixie>
considering just the horizontal dimension for now

[18:42:00.0000] <Hixie>
there are three variables for a cue:

[18:42:01.0000] <Hixie>
horizontal position, size of the box, and the alignment of the text within the box

[18:42:02.0000] <Hixie>
http://docs.google.com/drawings/edit?id=1LXroDh1tSee1YM79toEMo6OEdlI2Du5EfoWt7mSmc8U&hl=en

[18:43:00.0000] <Hixie>
there's two ways i can think of doing this: either aligning the alignment edge to the given horizontal position, or

[18:43:01.0000] <Hixie>
aligning the point along the box in the horizontal dimension that is the given distance along the box as a fraction of the box width to the equivalent position on the video as a fraction of the video width

[18:43:02.0000] <Hixie>
the latter works really well if the width is known ahead of time and is not 100%

[18:44:00.0000] <Hixie>
the former works really well if the horizontal position is less than 75% or so

[18:44:01.0000] <Hixie>
the latter makes the position irrelevant if the width is 100% though

[18:44:02.0000] <nessy>
/me is just going through email and reached the public-html thread ;)

[18:45:00.0000] <TabAtkins>
I'm confused about how the last box in the second example works.

[18:45:01.0000] <Hixie>
and the former fails dramatically for the case of the position being 100%

[18:45:02.0000] <TabAtkins>
Omitted size implies 100%, and thus ignores H?

[18:45:03.0000] <Hixie>
TabAtkins: the point 50% along the way of the box is aligned to the point 50% along the video

[18:45:04.0000] <TabAtkins>
Oh, duh, got it.

[18:45:05.0000] <Hixie>
TabAtkins: it's just that since the dimension is 100% (presumably, by default), that is indistinguishable to any other position

[18:46:00.0000] <TabAtkins>
I like that usage.  It's very natural in CSS backgrounds.

[18:46:01.0000] <Hixie>
yeah but it totally makes the position useless if the width is not known

[18:46:02.0000] <Hixie>
it's as bad as the last case in the first box

[18:46:03.0000] <TabAtkins>
Unless you introduce some concept of "shrinkwrap" for width.

[18:46:04.0000] <nessy>
for the positioning and styling stuff, can't we use CSS directly?

[18:46:05.0000] <Hixie>
TabAtkins: if we do that, then each cue would align differently

[18:47:00.0000] <Hixie>
nessy: CSS doesn't have the magic avoidance we need to add for titles to avoid the overlapping titles issue

[18:47:01.0000] <TabAtkins>
Yes?  So either you do a definite size, or just deal with it.

[18:47:02.0000] <Hixie>
TabAtkins: i'd rather the default was pretty :-)

[18:47:03.0000] <nessy>
hmmm…

[18:47:04.0000] <TabAtkins>
I don't really have an opinion on whether it's better to shrinkwrap or just expand to full width.

[18:47:05.0000] <TabAtkins>
The latter seems like it would be fine, personally.

[18:48:00.0000] <Hixie>
maybe the default should be to do absolute positions, except if you set a width, then it does the second case...

[18:48:01.0000] <Hixie>
though then you still have the other problem...

[18:48:02.0000] <Hixie>
i guess the question is

[18:48:03.0000] <Hixie>
what should happen if you say a left-aligned box is at position 100% horizontally?

[18:49:00.0000] <Hixie>
does that even make sense?

[18:49:01.0000] <TabAtkins>
That's why always doing the latter is best.

[18:49:02.0000] <TabAtkins>
So that situation actually produces something good - a right-aligned box.

[18:49:03.0000] <Hixie>
always doing the latter works fine except when width is automatic, then it's counterintuitive or ugly depending on how we do it

[18:50:00.0000] <Hixie>
no, it produces a left-aligned box that is the full width of the screen :-)

[18:50:01.0000] <TabAtkins>
If the width is auto positioning it doesn't make sense, given our requirement that it shouldn't go out of the screen.

[18:50:02.0000] <TabAtkins>
At least, horiz positioning it doesn't.

[18:50:03.0000] <Hixie>
if the size of the box is 100%, and we use the case-2 alignment mechanism, then the position has no effect

[18:50:04.0000] <TabAtkins>
Yes.

[18:50:05.0000] <Hixie>
yeah maybe that's the answer... if it has no width, then ignore the horizontal positioning

[18:51:00.0000] <erlehmann>
Hixie, overlapping title avoidance ? couldn't you just create pseudo-boxes that become more if there are more subtitles ? styling like video::subtitle:first-of-type() etc ?

[18:51:01.0000] <Hixie>
(and say so explictly)

[18:51:02.0000] <TabAtkins>
Phew.  Just got through Chapter 8 of RDFa 1.1.  There aren't words to describe how stupid the complexity of the processing here is.

[18:51:03.0000] <Hixie>
erlehmann: watch the titles at http://www.youtube.com/watch?v=xG9KluukpJI#t=2m40

[18:51:04.0000] <Hixie>
erlehmann: i'd like us to get that effect without the author having to say anything about the positioning

[18:52:00.0000] <nessy>
I am confused about this whole positioning thing

[18:52:01.0000] <TabAtkins>
Unfortuantely there's no way to "fix" it without just dropping the whole concept and saying "Use raw RDF, or Microdata."

[18:52:02.0000] <erlehmann>
words fail me to describe the awesome that animu has in defining this part of the spec

[18:53:00.0000] <Hixie>
erlehmann: the same thing occurs in lots of other contexts too, it's just that anime is the main way most of us geeks get exposed to fancy subtitle effects :-)

[18:54:00.0000] <nessy>
Hixie: do I understand correctly that you are trying to put the avoidance straight into the caption file format?

[18:54:01.0000] <Hixie>
nessy: straight into the processing model for the format, yeah

[18:54:02.0000] <Hixie>
not into the format itself

[18:54:03.0000] <Hixie>
i.e. there should be no syntax for this

[18:54:04.0000] <Hixie>
it should just happen

[18:54:05.0000] <nessy>
but wouldn't you only know of overlapping conflicts if you have all the different caption tracks analysed?

[18:55:00.0000] <nessy>
typicallly, the ones that conflict would come from different caption files

[18:55:01.0000] <Hixie>
if all the tracks are using the same format, then no, you just do it on the fly

[18:55:02.0000] <Hixie>
for rendering, doesn't matter which track the files come from

[18:55:03.0000] <nessy>
who would do it? at which level?

[18:55:04.0000] <Hixie>
the UA

[18:55:05.0000] <Hixie>
when rendering

[18:56:00.0000] <Hixie>
(or the media subsystem, or whatever is responsible for this)

[18:56:01.0000] <nessy>
ok, follow

[18:56:02.0000] <nessy>
so, what do you need to put into the files then to make that happen?

[18:56:03.0000] <othermaciej>
ideally, you want a caption format that would be reasonable for either a web engine or a media engine to render

[18:56:04.0000] <Hixie>
nessy: hopefully nothing

[18:56:05.0000] <othermaciej>
since implementation strategies may differ

[18:56:06.0000] <Hixie>
othermaciej: yeah, 100% agree

[18:57:00.0000] <nessy>
ok, I follow :)

[18:57:01.0000] <nessy>
so, why would CSS styling then not work per caption file?

[18:58:00.0000] <Hixie>
not sure what you mean

[18:58:01.0000] <Hixie>
why would CSS styling not work?

[18:58:02.0000] <Hixie>
work for what?

[18:58:03.0000] <nessy>
and positioning, I mean

[18:59:00.0000] <Hixie>
i don't understand how you'd use CSS to do the positioning.

[18:59:01.0000] <nessy>
the part of CSS that does positioning

[18:59:02.0000] <nessy>
CSS has alignment properties and stuff, right?

[19:00:00.0000] <nessy>
I'm trying to understand why there is a need for the horizontal positioning (H) and alignment (A) stuff to be different from CSS

[19:00:01.0000] <Hixie>
CSS has like a dozen different layout mechanisms, i don't see how any of them would work here though

[19:01:00.0000] <TabAtkins>
Actually... Flexbox + abspos would probably work great.

[19:01:01.0000] <Hixie>
i mean, i intend to define things in term of the block/inline/vertical layout mechanisms, but you still have to position those somehow

[19:01:02.0000] <Hixie>
TabAtkins: including for http://www.youtube.com/watch?v=xG9KluukpJI#t=2m40 ?

[19:01:03.0000] <nessy>
ok, just go ahead - maybe things get clearer when I see it all specified

[19:01:04.0000] <nessy>
I'm just struggling to follow the logic

[19:02:00.0000] <TabAtkins>
Hixie: That'd be absposing the translations of the writing on the picture.

[19:02:01.0000] <Hixie>
TabAtkins: no i mean the way the second title doesn't move when the first goes away

[19:02:02.0000] <Hixie>
TabAtkins: and the way the second one appears above the first automatically (without being given explicit instructions of where to appear)

[19:02:03.0000] <Hixie>
nessy: i might be missing something obvious... if you have a proposal for how to make it work with CSS, i'm all ears

[19:03:00.0000] <TabAtkins>
The former - no, unless you have a mechanism for setting things to visibility:hidden when they overlap, then visibility:collapse when they stop overlapping.

[19:03:01.0000] <nessy>
I'll see if I can work it out with CSS - maybe then I can see what the issue is :)

[19:03:02.0000] <TabAtkins>
The latter, yes.

[19:04:00.0000] <Hixie>
TabAtkins: i'd like a system where you can position things all over the place then have another title in "defualt" position and have that one just slide to the first "line" where it fits

[19:04:01.0000] <TabAtkins>
(In current draft, have a flexbox with box-orient:vertical and box-direction:alternate.  In my current rewritten syntax, just have display:flex-up on the caption container.

[19:04:02.0000] <Hixie>
TabAtkins: i think either way you end up having to have dedicated code

[19:04:03.0000] <nessy>
just one more question: with the S parameter (size) are you trying to do automatic line breaks?

[19:05:00.0000] <Hixie>
nessy: possibly. The main reason was to have a way to say where the edge was for purposes of avoiding overlaps.

[19:05:01.0000] <Hixie>
nessy: is there a need for automatic wrapping?

[19:05:02.0000] <TabAtkins>
Hixie: In the case that you have something manually positioned in the middle of the screen, and then overlapping text fills the entire bottom fo the screen to just below the positioned text, would a third overlapping segment then go above the manually positioned text?

[19:06:00.0000] <nessy>
I don't think so - though I wonder what happens when you change video size, e.g. to to full-screen

[19:06:01.0000] <Hixie>
TabAtkins: i'd hope so, don't see where else we'd want it to go

[19:06:02.0000] <nessy>
if you don't blow up the captions at the same ratio, they look really tiny

[19:06:03.0000] <Hixie>
nessy: i assume captions will be sized relative to the video

[19:06:04.0000] <Hixie>
nessy: but i'm more worried about different fonts having different needs for where to wrap

[19:06:05.0000] <TabAtkins>
Hixie: In that case we do need a new layout mode.

[19:07:00.0000] <nessy>
so, rather than having a S parameter size given in the spec, I would say that S should be calculated by the rendering engine, right?

[19:07:01.0000] <Hixie>
TabAtkins: right, that's what i'm doing :-)

[19:07:02.0000] <TabAtkins>
Something with a notion of intruding abspos, and stability of positioning.

[19:07:03.0000] <Hixie>
nessy: how do you mean?

[19:08:00.0000] <nessy>
are the three parameters in your image H, S, A specified in the caption file or just something that your layout mode deals with?

[19:08:01.0000] <nessy>
maybe this is where my whole confusion comes from!

[19:09:00.0000] <nessy>
I would think H, A are in the caption file, but S is only in the layout mode

[19:10:00.0000] <Hixie>
that's one of the things i want to try to work out. I think it's possible that we'd have all three in the track file (though all three optional for each cue of course)

[19:10:01.0000] <Hixie>
but maybe we can get away with fewer

[19:10:02.0000] <Hixie>
Extended SRT has H and S 

[19:10:03.0000] <Hixie>
(though expressed as X1 and X2)

[19:11:00.0000] <Hixie>
if we want to be able to set a background colour whose edge doesn't move around with each cue, we'd need an S

[19:11:01.0000] <nessy>
only in the layout mode though, I think

[19:11:02.0000] <TabAtkins>
I wonder if just doing a left and right would be better instead.

[19:12:00.0000] <Hixie>
so e.g. the question is, in the video from which http://junkyard.damowmow.com/425 comes, is the box the same size in each cue?

[19:12:01.0000] <nessy>
as an author, I wouldn't really want to have to calculate for every text cue how long it is

[19:12:02.0000] <Hixie>
if yes, we need an explicit S

[19:12:03.0000] <TabAtkins>
Benefit is that it shares well with the "abspos" positioning scheme.

[19:12:04.0000] <Hixie>
TabAtkins: H+S and X1+X2 are equivalent

[19:12:05.0000] <TabAtkins>
Hixie: Yes, but one or the other may be more natural to use.

[19:12:06.0000] <Hixie>
TabAtkins: how it is written in the file is a separate issue :-)

[19:13:00.0000] <Hixie>
i'm just trying to work out the processing model for now

[19:13:01.0000] <nessy>
it's either a box defined for all text cues in size or it is dynamically adjusted around the size of the current text cue

[19:13:02.0000] <nessy>
I think specifying it separately per text cue is a nightmare

[19:13:03.0000] <nessy>
to authore

[19:14:00.0000] <Hixie>
yes i'd expect most authors to let it be implied

[19:14:01.0000] <Hixie>
i'd expect most authors to not set any of this

[19:14:02.0000] <nessy>
that X1 X2 Y1 Y2 of extended SRT is not implemented anywhere is a good sign that it's not useful per text cue

[19:14:03.0000] <Hixie>
it's only really relevant for cases like http://philip.html5.org/misc/eva-captions.jpg

[19:14:04.0000] <Hixie>
or http://img3.imageshack.us/img3/6743/vlcsnap090208083027134xq5.png

[19:16:00.0000] <nessy>
to me as an author I would think there is top, right, bottom, left, center placement of caption cue box; then there is left, right, center, justify alignment on the text around the center of the caption cue box

[19:17:00.0000] <Hixie>
it's more complicated than that because of having to deal with which direction the box grows in when it's multiple lines long

[19:18:00.0000] <nessy>
then I might define a background formatting for the caption cue box for all captions and either make that adjust to the size of the caption or give it a fixed size into which all my caption cues will fit

[19:20:00.0000] <nessy>
the direction in which the box grows is dependent on the charset (is it charset that I use, e.g. for Chinese characters?) and the way in which that charset works - e.g. top-to-bottom or left-to-right

[19:20:01.0000] <nessy>
I would think

[19:21:00.0000] <Hixie>
i don't mean the block-progression-direction, i mean where the box is anchored. Consider http://philip.html5.org/misc/eva-captions.jpg -- if the top caption were one line long, it would be at the top, but if the bottom caption were one line long, it would be at the bottom.

[19:21:01.0000] <Hixie>
they grow in different directions

[19:21:02.0000] <Hixie>
http://dashiva.net/misc/1271498287684.jpg shows that oo

[19:21:03.0000] <Hixie>
too

[19:22:00.0000] <Hixie>
except there i'd expect the bottom one to grow down, or maybe even be centered, rather than grow up

[19:23:00.0000] <Hixie>
you wouldn't want the author to have to worry about this for each cue, so you really want good defaults so that just throwing the caption to the top or the bottom gets the right effect

[19:23:01.0000] <Hixie>
without having to give a dozen coordinates and alignment instructions for each cue

[19:25:00.0000] <nessy>
I think we want the same thing

[19:25:01.0000] <nessy>
I might just not express it very well

[19:26:00.0000] <nessy>
on http://dashiva.net/misc/1271498287684.jpg I would think there are two boxes

[19:27:00.0000] <nessy>
a second line in both boxes would always be rendered underneath the first line, but the box center position would move down on the top box and up on the bottom box

[19:27:01.0000] <Hixie>
yeah

[19:28:00.0000] <nessy>
and the author would only say that for the first box it is top position and left aligned, and on the second box that it is bottom positioned and center aligned

[19:28:01.0000] <Hixie>
ideally though the author would only have to give two instructions, maybe three: for the top box: align left, position at the top; for the bottom box: position at the bottom

[19:28:02.0000] <Hixie>
or maybe not even anything for the bottom box

[19:28:03.0000] <Hixie>
gotta go to dinner

[19:29:00.0000] <Hixie>
later

[19:29:01.0000] <nessy>
will be out for lunch

[19:29:02.0000] <nessy>
ttyl

[20:29:00.0000] <Bolkonskij>
hi

[22:29:00.0000] <Hixie>
a lot of the discussion in public-html-a11y about timed tracks has assumed that subtitles are mutually exclusive... why is that?

[22:29:01.0000] <Hixie>
wouldn't that just be a UI decision?

[22:30:00.0000] <Hixie>
foolip: you were in that discussion a lot... any idea?

[22:31:00.0000] <micheil>
Hixie: I've been thinking about some websockets stuff, and have a bit of an idea.. wanted to run it past you if you had a moment?

[22:31:01.0000] <Hixie>
sure

[22:32:00.0000] <micheil>
okay, currently with websockets, they are a two way channel, however, what happens in the case where a websocket server is only interested in streaming data down to a client?

[22:32:01.0000] <micheil>
would it make sense to add an extra (optional) headers which allowed you to specific (from the server) whether the connection was to be a ReadOnly connection?

[22:33:00.0000] <micheil>
(alternatively, you may also wish to implement a writeOnly, but I can't see a use case for that)

[22:33:01.0000] <Hixie>
you define the protocol that runs over websocket, so you could just define your protocol such that there is nothing the client is allowed to say, and the server can just ignore everything the client sends

[22:33:02.0000] <micheil>
by default, websockets would assume that they are both Read and Write, unless this is set

[22:34:00.0000] <micheil>
true, although, would it not be a good idea to give the client some indication of this?

[22:34:01.0000] <micheil>
I'm thinking in a more generic sense, rather then at the subprotocol level

[22:35:00.0000] <micheil>
this flag would have no effect on the readyState of the connection.

[22:35:01.0000] <Hixie>
well it's not like the client is a random client, the client is you

[22:36:00.0000] <Hixie>
i mean, you're not going to be talking to random pages, you're only talking to people who know what your protocol is

[22:36:01.0000] <Hixie>
so they know your protocol is "read only", because that's why they're connecting to you

[22:37:00.0000] <micheil>
not so

[22:37:01.0000] <micheil>
the client may be a random client, in the case of a websocket server as an API server

[22:40:00.0000] <Hixie>
then you're going to have to tell the client a heck of a lot more than just that the server doesn't want anything

[22:40:01.0000] <Hixie>
you're going to have to tell the client the entire protocol

[22:41:00.0000] <micheil>
hmm.. okay, fair enough

[22:41:01.0000] <micheil>
it was just a quick idea I had

[22:42:00.0000] <Hixie>
micheil: the way to think of Web Sockets is that it is the same as TCP

[22:42:01.0000] <Hixie>
except for the web

[22:42:02.0000] <micheil>
okay

[22:42:03.0000] <micheil>
I'm thinking in a more service oriented thing

[22:43:00.0000] <micheil>
Hixie: have you seen pusherapp.com?

[22:44:00.0000] <Hixie>
looks like what eventsource was made for :-)

[22:45:00.0000] <micheil>
eventsource?

[22:45:01.0000] <Hixie>
http://dev.w3.org/html5/eventsource/

[22:47:00.0000] <micheil>
Hixie: there's a difference there, these are read/write

[22:47:01.0000] <Hixie>
ah ok

[22:47:02.0000] <micheil>
the stuff about readonly was in relation to another piece of work I'm doing

[22:48:00.0000] <othermaciej>
good evening

[22:48:01.0000] <micheil>
evening'

[22:54:00.0000] <abarth>
anyone know what WG at the IETF is working on URL stuff?

[23:17:00.0000] <hsivonen>
Hixie: do you recall if you intended to stop processing if the insertion point becomes non-foreign in the case mentioned in the bug report http://www.w3.org/Bugs/Public/show_bug.cgi?id=9582 ?

[23:18:00.0000] <Hixie>
i don't recall off hand

[23:18:01.0000] <Hixie>
do you have some markup examples in mind?

[23:19:00.0000] <hsivonen>
IIRC <a><svg><foreignObject><a><svg></a>

[23:20:00.0000] <othermaciej>
abarth: IRIbis

[23:20:01.0000] <abarth>
is that http://tools.ietf.org/wg/iri/ ?>

[23:20:02.0000] <othermaciej>
http://tools.ietf.org/wg/iri/

[23:20:03.0000] <othermaciej>
yeah

[23:21:00.0000] <othermaciej>
abarth: "work" may be a stronger word than is merited so far

[23:21:01.0000] <abarth>
ic

[23:21:02.0000] <othermaciej>
abarth: but they are indeed planning to publish an updated document that's supposed to handle real-world URL parsing

[23:22:00.0000] <abarth>
i'm crunching all this data for the GURL/KURL thing

[23:22:01.0000] <abarth>
maybe I'll write up a doc and send it to their mailing list

[23:22:02.0000] <abarth>
that's probably a better forum than public-html

[23:22:03.0000] <abarth>
it's looking pretty messy at the moment

[23:22:04.0000] <abarth>
but maybe it will become clearer as I get further in

[23:23:00.0000] <othermaciej>
abarth: messy in what way?

[23:23:01.0000] <othermaciej>
abarth: quirky required behavior? lots of browser differences?

[23:23:02.0000] <abarth>
lots of browser differences

[23:24:00.0000] <abarth>
i'm not sure what's required

[23:24:01.0000] <abarth>
if there's a lot of variety, that might mean there's room to remove quirks

[23:24:02.0000] <aboodman>
abarth: you around?

[23:25:00.0000] <abarth>
aboodman: yes

[23:26:00.0000] <aboodman>
abarth: hey, i was wondering something about cors

[23:26:01.0000] <aboodman>
doesn't it have the issue where it might leak access to intranet resources?

[23:27:00.0000] <othermaciej>
aboodman: no (unless intranet resources are in the habit of adding Access-Control-Allow-Origin headers randomly)

[23:27:01.0000] <abarth>
aboodman: that's one of the threats CORS tries to address

[23:27:02.0000] <abarth>
aboodman: do you have something speific you're worried about?

[23:27:03.0000] <othermaciej>
aboodman: in theory it *can* violate integrity (as opposed to confidentiality) of intranet resources, but only in the same ways forms already can

[23:27:04.0000] <aboodman>
right cors, requires that existing servers respond with a special header.

[23:27:05.0000] <aboodman>
even for GET.

[23:27:06.0000] <abarth>
yes

[23:27:07.0000] <aboodman>
sorry, forgot that bit.

[23:29:00.0000] <aboodman>
ooc, if the intranet issue didn't exist, would servers still need to respond with the special header?

[23:29:01.0000] <abarth>
depends what you mean by intranet

[23:29:02.0000] <aboodman>
that is, would not sending credentials be sufficient.

[23:29:03.0000] <abarth>
there's always IP-based authentication

[23:29:04.0000] <othermaciej>
for requests without credentials, maybe not

[23:29:05.0000] <abarth>
you can't avoid sending credentials

[23:29:06.0000] <abarth>
for example, proxy-auth credentials

[23:29:07.0000] <abarth>
or IP addresses

[23:29:08.0000] <othermaciej>
"intranet" is a shorthand for a broad category

[23:30:00.0000] <abarth>
what about the loopback interface?

[23:30:01.0000] <abarth>
does that count as an intranet?

[23:30:02.0000] <othermaciej>
it basically means any resource you have access to by virtue of network position

[23:30:03.0000] <aboodman>
othermaciej: ok, that shorthand is fine.

[23:30:04.0000] <othermaciej>
abarth: the loopback interface is not behind a firewall, is it?

[23:30:05.0000] <aboodman>
intranet => any resource you have access to by virtue of network position

[23:30:06.0000] <abarth>
it might or might not be.  there are lots of services that bind only to 127.0.0.1

[23:31:00.0000] <othermaciej>
abarth: well, I guess for most home users it is since they are behind a NAT firewall

[23:31:01.0000] <abarth>
those aren't "firewalled" per-se

[23:31:02.0000] <abarth>
but they're only accessible from the box itself

[23:31:03.0000] <othermaciej>
aboodman: many home routers are reconfigurable via http from the local network segment

[23:31:04.0000] <othermaciej>
aboodman: in many cases this is encoded in firmware that can't readily be upgraded

[23:31:05.0000] <aboodman>
i guess i'm asking: are there other forms of ambient authority that the browser does not control, other than client IP address.

[23:31:06.0000] <othermaciej>
so "if the intranet issue didn't exist" is a rather huge counter-factual

[23:32:00.0000] <aboodman>
yeah, it was "out of curiosity"

[23:32:01.0000] <aboodman>
not related to current debate

[23:32:02.0000] <othermaciej>
the IP address and the routes

[23:32:03.0000] <othermaciej>
for example, I have access to systems on net 17 right now

[23:32:04.0000] <othermaciej>
but my IP address is not on net 17

[23:32:05.0000] <othermaciej>
because I have a VPN tunnel

[23:33:00.0000] <othermaciej>
random attackers cannot send packets to machines behind Apple's firewall even if they can send packets to me

[23:33:01.0000] <othermaciej>
it's a matter of network topology, not just IP address

[23:33:02.0000] <aboodman>
i see.

[23:34:00.0000] <aboodman>
ok, thanks.

[23:34:01.0000] <abarth>
aboodman: proxy-authentication is another example

[23:34:02.0000] <abarth>
also, just having a fresh IP address can be valuable

[23:34:03.0000] <abarth>
e.g., for sending spam

[23:34:04.0000] <abarth>
or click fraud

[23:34:05.0000] <abarth>
there was a recent attack on IRC using browsers

[23:35:00.0000] <abarth>
that used that property

[23:35:01.0000] <othermaciej>
yum, cross-protocol attacks

[23:35:02.0000] <othermaciej>
aboodman: the CORS vs UMP debate is really about whether it is ever ok to send user and originating site credentials in a cross-site request

[23:36:00.0000] <othermaciej>
the ObjCap people say no, because ambient authority is bad, end of story

[23:36:01.0000] <othermaciej>
others (e.g. me) would say, but the Web security model is already based on origins, so we may as well go with the flow in the soundest way we can instead of trying to fight it

[23:36:02.0000] <othermaciej>
and CORS makes it easier to code many common access patterns securely, notwithstanding theory

[23:37:00.0000] <Hixie>
imho UMP might make sense but only if you revamp the way the web works first

[23:37:01.0000] <aboodman>
yeah, my question came up because i remembered the "intranet case" from earlier discussions abarth and i had had regarding chrome extensions.

[23:37:02.0000] <aboodman>
and i suddenly wondered how cors could work at all.

[23:37:03.0000] <othermaciej>
if we were designing the Web security model from scratch, then an object-capability model might be a reasonable choice

[23:37:04.0000] <aboodman>
i had forgotten the requirement to respod with a header.

[23:37:05.0000] <aboodman>
apologies for the dumb question.

[23:37:06.0000] <othermaciej>
aboodman: the "intranet case" is a bane of existence for cool stuff

[23:37:07.0000] <abarth>
there are no dumb questions on this topic

[23:37:08.0000] <abarth>
this stuff is all very archane

[23:38:00.0000] <othermaciej>
there are many dumb answers, though :-)

[23:39:00.0000] <othermaciej>
security is hard

[23:39:01.0000] <othermaciej>
the Web is hard

[23:39:02.0000] <wirepair>
which is why i have a job ;>

[23:39:03.0000] <othermaciej>
hardness combines nonlinearly

[23:39:04.0000] <aboodman>
fwiw i think othermaciej's argument that keeping secrets secure is compelling too.

[23:40:00.0000] <aboodman>
in favor of sop.

[23:40:01.0000] <othermaciej>
which argument was that?

[23:41:00.0000] <aboodman>
it ws from awhile ago, but you recently referred to it

[23:41:01.0000] <aboodman>
/me looks

[23:41:02.0000] <abarth>
yeah, integrity primitive are better for integrity properties

[23:41:03.0000] <Hixie>
/me doesn't understand why the a11y tf came up with <trackgroup>

[23:44:00.0000] <aboodman>
"the risk of programming errors with CORS-only  

[23:44:01.0000] <aboodman>
solutions has to be weighed against the risk of programming errors in  

[23:44:02.0000] <aboodman>
shared-secret solution"

[23:44:03.0000] <aboodman>
http://lists.w3.org/Archives/Public/public-webapps/2009OctDec/0481.html

[23:44:04.0000] <aboodman>
it was ojan who referred to it, not you.

[23:44:05.0000] <othermaciej>
aboodman: oh

[23:45:00.0000] <othermaciej>
yeah, I remember that

[23:45:01.0000] <Hixie>
TabAtkins: yt?

[23:45:02.0000] <othermaciej>
that's something that is often glossed over, the challenge of managing shared secrets

[23:45:03.0000] <othermaciej>
especially if those shared secrets go in URLs

[23:45:04.0000] <Hixie>
anyone know what the csswg's current state of thinking wrt multiple pseudo-elements per selector is?

[23:45:05.0000] <othermaciej>
it's very hard to keep URLs confidential - they like to leak out into the world

[23:47:00.0000] <othermaciej>
aboodman, abarth: a while back I made a series of slides explaining how you could combine Origin-based defenses and shared secrets to make a protocol that was more robust than either alone

[23:48:00.0000] <othermaciej>
but I never finished it, because it seemed pointless to continue the debate about security qualities of the protocol

[23:48:01.0000] <othermaciej>
but maybe I could dig it out and send it somewhere

[23:48:02.0000] <abarth>
did you take the anticipated rhetorical context into account?

[23:48:03.0000] <abarth>
<--- i'll stop being mean now

[23:49:00.0000] <othermaciej>
abarth: I did - that's why I never finished it :-)

[23:49:01.0000] <abarth>
IP address canonicalization seems to break down as Safari+Firefox versus Chrome+IE

[23:49:02.0000] <abarth>
IE likes to canonicalize whereas FF doesn't

[23:50:00.0000] <abarth>
canonicalize('http://000030052000001/') is 'http://192.168.0.1/' IE

[23:50:01.0000] <abarth>
canonicalize('http://000030052000001/') is 'http://192.168.0.1/' KR

[23:50:02.0000] <abarth>
canonicalize('http://000030052000001/') should be http://192.168.0.1/. Was http://000030052000001/. FF

[23:50:03.0000] <abarth>
canonicalize('http://000030052000001/') should be http://192.168.0.1/. Was http://000030052000001/. SA

[23:50:04.0000] <abarth>
as a random example

[23:50:05.0000] <othermaciej>
are there any reasons besides matching one browser or the other to caniicalize or not?

[23:50:06.0000] <Hixie>
abarth: iirc it is considered a security-wise dubious thing to support non-dotted-quad ipv4 addresses

[23:50:07.0000] <Hixie>
abarth: and it is certainly not kosher per the rfc iirc

[23:51:00.0000] <abarth>
what does "support" mean?

[23:51:01.0000] <Hixie>
convert to anything or resolve as an IP address

[23:51:02.0000] <othermaciej>
so you think URLs like that should not be canonicalized, and the network layer should refuse to load from such UR:s?

[23:51:03.0000] <abarth>
i should check whether firefox actually lets you follow those URLs

[23:51:04.0000] <othermaciej>
er

[23:51:05.0000] <othermaciej>
URLs

[23:52:00.0000] <Hixie>
othermaciej: unless there's a pressing compat need or the spec says i'm wrong, yes

[23:52:01.0000] <Hixie>
to start with, pure-numeric labels are valid dns labels

[23:52:02.0000] <Hixie>
so you could have a local machine in dns with the name "000030052000001" if i'm not mistaken

[23:53:00.0000] <Hixie>
(e.g. www.000.com. resolves)

[23:53:01.0000] <othermaciej>
I thought that officially DNS labels were not supposed to start with a number

[23:53:02.0000] <othermaciej>
but I know that is often violated

[23:53:03.0000] <Hixie>
000.com and 123.com are both registered

[23:53:04.0000] <Hixie>
so...

[23:57:00.0000] <othermaciej>
aboodman: hey, that was a good email you cited - I didn't remember writing that :-)

[23:57:01.0000] <aboodman>
heh. happy to help.

[23:57:02.0000] <aboodman>
this debate has been going on quite awhile.

[23:57:03.0000] <Hixie>
the debate is pointless since no UA is implement it

[23:58:00.0000] <Hixie>
implementing

[23:58:01.0000] <Hixie>
it's like Web SQL DB -- except with even fewer vendors on board

[23:58:02.0000] <othermaciej>
Hixie: except that Caja is apparently the most important UA of all

[23:58:03.0000] <Hixie>
i don't understand why it's still even on the table

[23:58:04.0000] <aboodman>
web sql db :(

[23:58:05.0000] <aboodman>
rip

[23:59:00.0000] <Hixie>
othermaciej: Why would this need specifying separately from the rest of Caja's APIs?

[23:59:01.0000] <othermaciej>
Hixie: don't ask me, ask Mark Miller

[23:59:02.0000] <Hixie>
i haven't seen him argue that it needs speccing because of caja

[00:00:00.0000] <othermaciej>
he cited Caja as a UA that plans to implement UMP and not CORS

[00:01:00.0000] <Hixie>
yeah that's all i've seen

[00:02:00.0000] <othermaciej>
I would say the APIs of script libraries do not require the Web standards process to define

[00:02:01.0000] <Hixie>
depends if they're going to be implemented once or more than once, i'd say

[00:03:00.0000] <Hixie>
but in this case i don't mind if it has a spec or not, i just think it should be done in a manner consistent with the rest of caja's apis

[00:03:01.0000] <Hixie>
if it's just for caja

[00:07:00.0000] <zcorpan>
http://lists.w3.org/Archives/Public/public-canvas-api/2010AprJun/0009.html

[00:08:00.0000] <zcorpan>
since my previous speculation about microsoft implementing xhtml when they sent feedback on document.write was correct, i'll speculate again that they're now implementing canvas

[00:08:01.0000] <abarth>
Hixie: what about this kind of canonicalization:

[00:08:02.0000] <abarth>
canonicalize('http://[0:0::0:0:8]/') is 'http://[::8]/'

[00:10:00.0000] <zcorpan>
should (new WebSocket('ws://foo:80/')).URL return ws://foo/ or ws://foo:80/ ?

[00:11:00.0000] <abarth>
i have that case

[00:11:01.0000] <abarth>
one sec

[00:11:02.0000] <abarth>
KR canonicalize('ws://foo:80/') is 'ws://foo/'

[00:11:03.0000] <abarth>
FF canonicalize('ws://foo:80/') should be ws://foo/. Was ws://foo:80/.

[00:11:04.0000] <abarth>
IE canonicalize('ws://foo:80/') should be ws://foo/. Was ws://foo:80/.

[00:11:05.0000] <abarth>
SA canonicalize('ws://foo:80/') should be ws://foo/. Was ws://foo:80/.

[00:12:00.0000] <abarth>
short answer is "probably"

[00:12:01.0000] <abarth>
FF/IE/SA don't support web sockets

[00:12:02.0000] <abarth>
but for the protocols that they do support

[00:12:03.0000] <abarth>
KR/FF/IE remove default port numbers

[00:12:04.0000] <abarth>
but SA doesn't

[00:13:00.0000] <zcorpan>
so for say http://foo:80/, everyone would give http://foo/ ?

[00:13:01.0000] <zcorpan>
except SA?

[00:13:02.0000] <othermaciej>
aboodman: so hrome is the only browser to canonicalize away the default port?

[00:14:00.0000] <othermaciej>
oh, ws:

[00:14:01.0000] <abarth>
yeah

[00:14:02.0000] <aboodman>
othermaciej:  no idea.

[00:14:03.0000] <abarth>
i'll show the http case

[00:14:04.0000] <abarth>
which is simpler

[00:14:05.0000] <abarth>
FF canonicalize('http://www.example.com:80/') is 'http://www.example.com/'

[00:14:06.0000] <abarth>
KR canonicalize('http://www.example.com:80/') is 'http://www.example.com/'

[00:14:07.0000] <abarth>
SA canonicalize('http://www.example.com:80/') should be http://www.example.com/. Was http://www.example.com:80/.

[00:14:08.0000] <abarth>
IE canonicalize('http://www.example.com:80/') is 'http://www.example.com/'

[00:14:09.0000] <zcorpan>
thanks!

[00:14:10.0000] <othermaciej>
abarth: yeah, the default port thing really needs to be a whitelist of specific protocols - making it based on the protocols the browser knows seems bad for interop

[00:15:00.0000] <abarth>
othermaciej: possibly, i'm trying to reserve judgement until i see the whole picture, which I don't yet

[00:16:00.0000] <abarth>
ok, the absolute URL dataset is done

[00:16:01.0000] <abarth>
i'm going to relative URLs later

[00:16:02.0000] <zcorpan>
othermaciej: if a browser doesn't know the protocol, does it matter whether it removes the port or not?

[00:16:03.0000] <abarth>
they seem even more complicated

[00:16:04.0000] <abarth>
(ok, i don't have file URLs done either)

[00:16:05.0000] <othermaciej>
zcorpan: for purposes of the API on <a> elements, yes, though maybe that is minor enough to not care

[00:17:00.0000] <Hixie>
abarth: yeah ipv6 canon seems legit

[00:18:00.0000] <Hixie>
once you're supporting different protocols, you've already lost interop anyway

[00:19:00.0000] <abarth>
FF canonicalize('http://GoOgLe.CoM/') is 'http://google.com/'

[00:19:01.0000] <abarth>
IE canonicalize('http://GoOgLe.CoM/') is 'http://google.com/'

[00:19:02.0000] <abarth>
KR canonicalize('http://GoOgLe.CoM/') is 'http://google.com/'

[00:19:03.0000] <abarth>
SA canonicalize('http://GoOgLe.CoM/') should be http://google.com/. Was http://GoOgLe.CoM/.

[00:20:00.0000] <Hixie>
ok time to go play bc2

[00:20:01.0000] <othermaciej>
abarth: FYI I'm planning to make a variant of the file URL test I made that has an http base URL

[00:20:02.0000] <Hixie>
bbl

[00:20:03.0000] <othermaciej>
in fact, I think I will do that now

[00:20:04.0000] <abarth>
great :)

[00:20:05.0000] <abarth>
i'm putting off the file URL stuff until i understand the other cases

[00:20:06.0000] <othermaciej>
hmmm

[00:21:00.0000] <abarth>
it looks like Brett preferred the IE way in a lot of cases

[00:21:01.0000] <zcorpan>
abarth: would it be possible to test opera also? :)

[00:21:02.0000] <othermaciej>
I need to not make it a script test though to avoid getting the <base> that I put in the template pulled in

[00:21:03.0000] <abarth>
i removed the base URL from the template

[00:21:04.0000] <abarth>
it didn't work properly

[00:21:05.0000] <abarth>
as in, it broke the tests on windows

[00:22:00.0000] <abarth>
there's a utility method to set the base URL to whatever you like

[00:22:01.0000] <abarth>
trival.js has an example

[00:22:02.0000] <othermaciej>
abarth: that will probably mess up the file: test

[00:22:03.0000] <abarth>
zcorpan: sure

[00:22:04.0000] <othermaciej>
abarth: which other tests did it break on windows?

[00:22:05.0000] <othermaciej>
does windows not have the /tmp link? was pretty sure it did, cause other tests use it

[00:22:06.0000] <abarth>
zcorpan: i was trying to keep the data set managable to start with

[00:23:00.0000] <abarth>
i don;t think the URL you used is a valid file URL on windows

[00:23:01.0000] <abarth>
it didn't have a drive-spec

[00:23:02.0000] <abarth>
why would it mess up the file test to set it using the API?

[00:23:03.0000] <abarth>
it just adds a base tag like you did

[00:24:00.0000] <abarth>
the difference is it does it after loading the various script tests scripts

[00:24:01.0000] <othermaciej>
setting it using the API could work

[00:24:02.0000] <othermaciej>
but if it breaks on Windows with a <base> tag, it will break with an API

[00:24:03.0000] <othermaciej>
what I meant is that not setting the base URI at all will likely break

[00:25:00.0000] <abarth>
its an ordering thing

[00:25:01.0000] <othermaciej>
or not so much break as make the test results dependent on the user's system

[00:25:02.0000] <abarth>
you need to set the base tag after loadingin the scripts

[00:25:03.0000] <abarth>
because the scripts are loaded using relative URLs

[00:26:00.0000] <othermaciej>
I see

[00:26:01.0000] <othermaciej>
good point

[00:26:02.0000] <othermaciej>
though it may still throw off the results in browsers that consider a file URI invalid if it has no drive letter

[00:26:03.0000] <othermaciej>
or maybe they just consider it undereferancable, not syntactically invalid

[00:26:04.0000] <othermaciej>
that would be ok

[00:30:00.0000] <othermaciej>
abarth: ok, I made file-http-base.html but I am not sure the expected results I have will match google-url's behavior

[00:30:01.0000] <abarth>
this is another interesting case

[00:31:00.0000] <abarth>
(that's fine.  the shouldBe results don't matter.  we'll figure out what they should be later)

[00:31:01.0000] <abarth>
IE canonicalize('http://example.com/foo%41%7a') is 'http://example.com/fooAz'

[00:31:02.0000] <abarth>
KR canonicalize('http://example.com/foo%41%7a') is 'http://example.com/fooAz'

[00:31:03.0000] <abarth>
FF canonicalize('http://example.com/foo%41%7a') should be http://example.com/fooAz. Was http://example.com/foo%41%7a.

[00:31:04.0000] <abarth>
SA canonicalize('http://example.com/foo%41%7a') should be http://example.com/fooAz. Was http://example.com/foo%41%7a.

[00:31:05.0000] <abarth>
the question here is whether you URL decode things that don't need to be encoded

[00:31:06.0000] <othermaciej>
yeah

[00:31:07.0000] <othermaciej>
Brett really likes IE's URL parsing I guess :-)

[00:32:00.0000] <abarth>
IE likes to keep non-ASCII characters in URLs

[00:33:00.0000] <abarth>
brett seems to follow firefox + safari in making things more ascii more of the time

[00:33:01.0000] <abarth>
FF canonicalize('http://example.com/你好你好') is 'http://example.com/%E4%BD%A0%E5%A5%BD%E4%BD%A0%E5%A5%BD'

[00:33:02.0000] <abarth>
KR canonicalize('http://example.com/你好你好') is 'http://example.com/%E4%BD%A0%E5%A5%BD%E4%BD%A0%E5%A5%BD'

[00:33:03.0000] <abarth>
SA canonicalize('http://example.com/你好你好') is 'http://example.com/%E4%BD%A0%E5%A5%BD%E4%BD%A0%E5%A5%BD'

[00:33:04.0000] <abarth>
IE canonicalize('http://example.com/你好你好') should be http://example.com/%E4%BD%A0%E5%A5%BD%E4%BD%A0%E5%A5%BD. Was http://example.com/你好你好.

[00:33:05.0000] <othermaciej>
I guess IE doesn't like things to be escaped

[00:34:00.0000] <abarth>
woah, firefox doesn't turn \ into / in paths

[00:34:01.0000] <abarth>
crazy

[00:34:02.0000] <abarth>
i'm surprised that works

[00:34:03.0000] <wirepair>
like http://example.com/blah\something.jsp ?

[00:35:00.0000] <abarth>
yeah

[00:35:01.0000] <abarth>
non-FF makes that

[00:35:02.0000] <abarth>
http://example.com/blah/something.jsp

[00:35:03.0000] <abarth>
life is different in the query string

[00:35:04.0000] <wirepair>
yeah it converts it to %5c

[00:36:00.0000] <abarth>
in the query string apparently non-IE likes to keep more things encoded

[00:36:01.0000] <abarth>
but IE is aggressive at not encoding things

[00:36:02.0000] <wirepair>
yeah which is why it's a hotbed for xss

[00:36:03.0000] <wirepair>
;>

[00:36:04.0000] <othermaciej>
abarth: did you port over the component parsing tests (other than user/pass) in addition to the canonicalization ones?

[00:36:05.0000] <abarth>
even to the point of keeping invalid unicode characters around

[00:37:00.0000] <wirepair>
invalid unicode == overlong utf-8 and such?

[00:37:01.0000] <abarth>
othermaciej: i'm not 100% clear what you're asking, but i think the answer is yet

[00:37:02.0000] <abarth>
sorry

[00:37:03.0000] <abarth>
yes

[00:37:04.0000] <abarth>
i'm using a.href for all of them

[00:37:05.0000] <othermaciej>
the ones that test what you get for scheme or path when given a particular URL

[00:38:00.0000] <abarth>
i haven't tried using the other APIs

[00:38:01.0000] <abarth>
oh

[00:38:02.0000] <abarth>
there are tests that give you a whole URL at then

[00:38:03.0000] <abarth>
see what the components are?

[00:38:04.0000] <abarth>
no, i haven't done those

[00:38:05.0000] <othermaciej>
well, I assume there are

[00:40:00.0000] <abarth>
http://code.google.com/p/google-url/source/browse/trunk/src/url_parse_unittest.cc

[00:40:01.0000] <abarth>
no, i haven't done the ones from that file yet

[00:41:00.0000] <abarth>
they look worthwhile too

[00:41:01.0000] <othermaciej>
ok maybe I will do those if I have time

[00:41:02.0000] <othermaciej>
gonna see if I can pick off a more important bug first

[00:42:00.0000] <abarth>
'http://www.example.com/#\ud800\u597d'

[00:42:01.0000] <abarth>
that URL gets a different result in every browser

[00:42:02.0000] <abarth>
i think the problem is that its invalid unicode

[00:43:00.0000] <abarth>
'http://www.example.com/#a\uFDD0'

[00:43:01.0000] <abarth>
that looks like a more reduced test for that case

[00:43:02.0000] <othermaciej>
abarth: reading that chromium-dev thread, I fear I will have to do my slideshow about CORS and Confused Deputy vulnerabilities for google folk at some point

[00:44:00.0000] <othermaciej>
I wouldn't want Chromium to be deciding on security features based on "Maciej said so"

[00:44:01.0000] <abarth>
to some extent, they rely on my for that kind of stuff, but i'm kind of pulling the rug out from under them by try to not be involved

[00:44:02.0000] <othermaciej>
at the very least it should be based on "Adam said so"

[00:45:00.0000] <abarth>
well, ifette is also a good person to be involved in the decision process

[00:45:01.0000] <aboodman>
we really only make decisions because aboodman said so, but since that list is public, we try to put on a nice show.

[00:45:02.0000] <abarth>
hahaha :)

[00:46:00.0000] <aboodman>
g'nite troops

[00:46:01.0000] <othermaciej>
abarth: you should give your advice, even if you don't want to be the decider, IMO

[00:48:00.0000] <abarth>
it seems like a no-win situation

[00:48:01.0000] <abarth>
i'm not even really sure what that thread is about

[00:48:02.0000] <othermaciej>
you're the one who started it!

[00:49:00.0000] <abarth>
oh, on chromium-dev?

[00:49:01.0000] <othermaciej>
yeah

[00:49:02.0000] <abarth>
yeah, ok

[00:50:00.0000] <abarth>
that sounds like a good email for tomorrow morning :)

[00:52:00.0000] <abarth>
Safari encodes the second # in the fragment, which doesn't match IE/FF/KR 

[00:52:01.0000] <abarth>
that sounds like an easy fix

[00:53:00.0000] <abarth>
also, Safari doesn't lower-case schemes, unlike IE/FF/KR

[00:56:00.0000] <othermaciej>
abarth: I am glad you are working on this and doing it to your usual ridiculously thorough standards

[00:56:01.0000] <abarth>
its actually not nearly as complicated as I thought it would be, at least for non-file, non-relative URLs

[00:56:02.0000] <abarth>
there seems to be <6 decisiosn

[00:57:00.0000] <abarth>
i need to look it over again when I'm less tired

[00:57:01.0000] <abarth>
but i'll write it up and ask brett and others why they chose what they did

[00:57:02.0000] <abarth>
i'm happy to do this stuff

[00:57:03.0000] <abarth>
i think its more important than anding random feature XYZ

[00:58:00.0000] <abarth>
but i understand that it doesn't generated good press articles :)

[00:59:00.0000] <abarth>
anyway, me => bed

[00:59:01.0000] <abarth>
thanks for your help writing the tests

[01:43:00.0000] <foolip>
Hixie: mutually exclusive subtitles: this is the way virtually all software and hardware media players work. it seems useful to be able to mark up that certain tracks are mutually exclusive (<tracktrack>) so that the UI can create a sane menu.

[01:44:00.0000] <hsivonen>
having multiple subtitle tracks present at one time seems to be on the wrong side of 80/20

[01:44:01.0000] <foolip>
the other options are to make all tracks mutually exlcusive or all tracks "parallel", both of which are broken

[01:44:02.0000] <hsivonen>
if you want to make English subtitles plus Japanese subtitles for language learners for anime, you can package them into one track

[01:45:00.0000] <foolip>
hsivonen: I agree, but others (nessy) insisted, and I think the current solution is fairly non-intrusive

[01:45:01.0000] <hsivonen>
FWIW, in movie theaters in Finland, there are subtitles in Finnish and in Swedish, but they aren't independent tracks

[01:45:02.0000] <hsivonen>
specifically, the Swedish subtitles aren't bought from Sweden

[01:45:03.0000] <asmodai>
/me wonders what it is that keeps leaking memory in FF.

[01:45:04.0000] <nessy>
hsivonen: you will not want to package two different languages in one file if you can avoid it

[01:45:05.0000] <foolip>
the most compelling argument (to me) is that if <track> is also used for e.g. extra audio tracks, clearly subtitles and the extra audio track aren't mutually exclusive

[01:46:00.0000] <hsivonen>
nessy: I think you do

[01:46:01.0000] <nessy>
 then you cannot turn one off

[01:46:02.0000] <foolip>
hi nessy

[01:46:03.0000] <hsivonen>
nessy: for the same reason Swedish subtitles in Finnish movie theaters aren't independent of the Finnish subtitles

[01:46:04.0000] <hsivonen>
nessy: the traslator to Finnish chooses the timing

[01:46:05.0000] <nessy>
hi foolip :)

[01:47:00.0000] <hsivonen>
and then the translator to Swedish uses that timing

[01:47:01.0000] <hsivonen>
so the Finnish and Swedish subtitles always change at the same time

[01:47:02.0000] <hsivonen>
because having them change at different times would be annoying

[01:47:03.0000] <nessy>
so you cannot reuse the subtitles in Finnish or Swedish separately?

[01:48:00.0000] <hsivonen>
nessy: but the Web is different from movie theaters in the sense that the Web experience is personal

[01:48:01.0000] <nessy>
I think you can very well create them in different files but synchronised

[01:48:02.0000] <nessy>
do you know whether in the production process they are actually in the same file and not just rendered on together?

[01:48:03.0000] <foolip>
I'd say that if parallel tracks don't impose too much complexity (in API, markup, code or UI) then it's a good feature

[01:48:04.0000] <hsivonen>
nessy: what's the use case on the Web except language learning?

[01:48:05.0000] <hsivonen>
IMO, language learning is way on the wrong side of 80/20

[01:48:06.0000] <hsivonen>
esp. since there's a workaround

[01:49:00.0000] <foolip>
current downside: it makes <trackgroup> mandatory boilerplate in most cases

[01:49:01.0000] <hsivonen>
multiple tracks means you need to stack them in layout...

[01:49:02.0000] <zcorpan>
hsivonen: use case: watching a movie with a friend who doesn't know swedish

[01:49:03.0000] <nessy>
chapter markers and subtitles and textual audio descriptions on together would be one need for multiple tracks active at the same time

[01:50:00.0000] <nessy>
plus you saw the examples that Hixie cumulated together with multiple subtitles on at the same time

[01:50:01.0000] <hsivonen>
zcorpan: I suppose the Web experience can become unpersonal once movie rentals and TV move to the browser

[01:51:00.0000] <hsivonen>
nessy: the examples I saw suggested they were for learning Japanese

[01:51:01.0000] <nessy>
I am not a big fan of having tracks of the same type on at the same time, but we definitely need it for different types

[01:51:02.0000] <hsivonen>
are there new examples?

[01:51:03.0000] <zcorpan>
hsivonen: yeah, i'd expect Voddler-like apps to appear as web apps

[01:51:04.0000] <othermaciej>
if <track> was used for audio descriptions as well as subtitles, then you would want those to come from a separate mutually exclusive list

[01:51:05.0000] <foolip>
indeed

[01:51:06.0000] <hsivonen>
/me notes that Finnish TV broadcasts don't support Finnish and Swedish subtitles simultaneously

[01:51:07.0000] <nessy>
that was my original design

[01:52:00.0000] <zcorpan>
/me doesn't know if Voddler supports multiple subtitles or indeed subtitles at all

[01:52:01.0000] <nessy>
shot down by reality I was told :)

[01:53:00.0000] <nessy>
/me goes checking out voddler

[01:53:01.0000] <hsivonen>
zcorpan: it would be an interesting market enabler if the Swedish audience had so good English listening comprehension that they could optimize subtitles away

[01:53:02.0000] <nessy>
bah, not available in my country

[01:54:00.0000] <foolip>
lol, region coded internets suck

[01:54:01.0000] <zcorpan>
hsivonen: personally i prefer english captions but most other people i know prefer swedish subtitles

[01:55:00.0000] <hsivonen>
zcorpan: why English captions?

[01:55:01.0000] <foolip>
I have the same preference as zcorpan, especially if the spoken language is English

[01:55:02.0000] <hsivonen>
interesting

[01:55:03.0000] <foolip>
and usually otherwise too, because the translation tends to be better

[01:56:00.0000] <hsivonen>
If the spoken language is English and I'm watching a movie alone, I turn subtitles off

[01:56:01.0000] <zcorpan>
hsivonen: it's easier to follow what they say compared to no captions, and it's easier to mentally listen and read in english compared to listen in english and read in swedish

[01:56:02.0000] <hsivonen>
if I'm watching with other people, they want the Finnish subtitles anyway

[01:57:00.0000] <hsivonen>
or the other people opt to have no subtitles, too

[01:58:00.0000] <hsivonen>
/me wonders what's the current situation with the Voddler GPL violation story is

[01:59:00.0000] <hsivonen>
http://en.wikipedia.org/wiki/Voddler#Source_Code_Copyright_Violation_by_Voddler wikipedia to rescue

[02:07:00.0000] <asmodai>
Any suggestions for testing memory leaks in FF? Or should I just use binary search on the addons/plugins?

[02:13:00.0000] <hsivonen>
asmodai: have you already verified that the leak isn't present when all addons and plug-ins are disabled?

[02:26:00.0000] <Hixie>
foolip: why wouldn't the UI just let you pick one subtitle or caption track, and leave it at that?

[02:26:01.0000] <Hixie>
foolip: while still letting the JS enable/disable whatever it wants?

[02:27:00.0000] <Hixie>
hsivonen: supporting multiple tracks visible at once is basically free once you support multiple cues not overlapping

[02:27:01.0000] <hsivonen>
Hixie: even in terms of layout stacking?

[02:29:00.0000] <Dashiva>
One thing I didn't see in the timed tracks examples was vertical text

[02:30:00.0000] <Hixie>
hsivonen: that's what i mean, once you have cues not stacking each other, making subtitles not stack each other is basically free

[02:30:01.0000] <Hixie>
hsivonen: is there anything else that would be non-trivial?

[02:30:02.0000] <Hixie>
Dashiva: one of the first examples has some vertical text

[02:31:00.0000] <Dashiva>
Oh, must've missed it

[02:31:01.0000] <Dashiva>
Wait, that isn't vertical text

[02:31:02.0000] <Dashiva>
That's rotated text

[02:31:03.0000] <Dashiva>
It's flipped 90 degrees

[02:31:04.0000] <Hixie>
oh, so it is

[02:32:00.0000] <Hixie>
i hadn't even noticed

[02:32:01.0000] <Hixie>
anyway i'm treating it as vertical

[02:32:02.0000] <Hixie>
for the purposes of this exercise

[02:32:03.0000] <Hixie>
:-)

[02:32:04.0000] <Dashiva>
kk

[02:32:05.0000] <Hixie>
and not supporting rotated text :-P

[02:32:06.0000] <Dashiva>
I think we can live without rotated text

[02:32:07.0000] <Hixie>
which isn't really a particularly good use of the evidence :-P

[02:32:08.0000] <Hixie>
if you have any other examples, please do put them up

[02:32:09.0000] <Hixie>
vertical ones in particular

[02:33:00.0000] <Hixie>
otherwise i will basically just have to go on my rather faulty knowledge of vertical text layout

[02:34:00.0000] <Hixie>
i gotta say, the thing that most surprises me in all my standards work is how hard it is to get people to provide use case examples

[02:34:01.0000] <Dashiva>
Here's one where the subtitles bounce in tune with the music... let's ignore that

[02:35:00.0000] <Hixie>
from people not understanding what that means, to not understanding why we would want to even look at real-world examples rather than just base it on theory, to not believing that use case examples actually affect design...

[02:36:00.0000] <Hixie>
many people have expressed the belief that when i ask them for use cases i'm just trying to dismiss them and that they are the only people from whom i ask for use cases

[02:36:01.0000] <Hixie>
it's so crazy

[02:36:02.0000] <hsivonen>
Hixie: I think I'm not sure if I understood what it means to support multiple cues not overlapping

[02:36:03.0000] <foolip>
Hixie: I don't think a content manu UI for enabling multiple tracks would be very nice, so from a laziness point of view I could live with letting the native UI only enable one track, implicitly disabling all other.

[02:37:00.0000] <foolip>
context menu

[02:37:01.0000] <Hixie>
hsivonen: http://www.youtube.com/watch?v=xG9KluukpJI#t=2m40

[02:37:02.0000] <Hixie>
foolip: yeah i wouldn't want the native menu to enable multiple tracks

[02:37:03.0000] <foolip>
if <track> is only for text I could accept this

[02:38:00.0000] <hsivonen>
Hixie: looks like a totally gratuitous "just because we can" effect that in *waaayy* on the wrong side of 80/20

[02:38:01.0000] <foolip>
if we take <track> for e.g. commentators tracks seriously, then it's not so good

[02:38:02.0000] <foolip>
I don't feel strongly about this issue either way

[02:38:03.0000] <hsivonen>
Hixie: are you assuming the reuse of the line layout code of the CSS formatter?

[02:38:04.0000] <Hixie>
hsivonen: multiple cues with overlapping times are not that rare

[02:38:05.0000] <foolip>
nessy: thoughts?

[02:39:00.0000] <nessy>
/me is reading email - have wasted a day in town today :(

[02:39:01.0000] <Hixie>
foolip: i think if we ever do multiple media resources in sync, the way to do it is to reuse <video>/<audio> and have them linked at that level

[02:39:02.0000] <Hixie>
foolip: i don't think we'd use <track> for that

[02:39:03.0000] <foolip>
Hixie: I would agree that that's cleaner

[02:39:04.0000] <Dashiva>
http://dashiva.net/misc/vertical_karaoke.jpg

[02:40:00.0000] <hsivonen>
/me didn't know YouTube supported times in the fragment id

[02:40:01.0000] <Hixie>
hsivonen: if you mean inline box and line box layout, then yes, at least in UAs that want to do CSS-based styling of cues

[02:40:02.0000] <foolip>
<track> is still not a great name for an element though (the only upside is that I first suggested it)

[02:40:03.0000] <Hixie>
foolip: what's a better name?

[02:40:04.0000] <Hixie>
Dashiva: sweet lord man

[02:40:05.0000] <foolip>
I have no better name, I suggested the best I could think of :)

[02:41:00.0000] <Creap>
probably the wrong channel to ask in.. but you know how to read specs :P ARIA spec says default value of aria-haspopup is false, does that mean that <li aria-haspopup> is the same as <li aria-haspopup="false">?

[02:41:01.0000] <Hixie>
<track> it is :-P

[02:41:02.0000] <foolip>
others have been <itext> and something else

[02:41:03.0000] <Dashiva>
This is the full madness karaoke form, with animated fade in and fade out of characters

[02:41:04.0000] <Hixie>
Dashiva: addressing the actual "current" span is something i haven't worked out how to do

[02:42:00.0000] <Hixie>
Dashiva: and positioning the subtitles that precisiely is non-trivial

[02:42:01.0000] <Hixie>
well it's very trivial if we do it pixel-based, but that's not an especially good idea for various reasons

[02:42:02.0000] <Hixie>
Dashiva: other than that i think we can do that

[02:42:03.0000] <Dashiva>
Well, most of this probably falls on the wrong side of 80/20, so it's no big crisis

[02:43:00.0000] <Hixie>
Dashiva: assuming we go down the kind of route i'm thinking of

[02:43:01.0000] <Hixie>
Dashiva: yeah

[02:43:02.0000] <Dashiva>
http://dashiva.net/misc/bouncing_karaoke.jpg

[02:43:03.0000] <foolip>
it would be quite acceptible to leve the most crazy styling/animations to JavaScript

[02:43:04.0000] <Hixie>
i really need to figure out positioning, that's the main thing i'm spinning on

[02:45:00.0000] <nessy>
re: multiple cues with overlapping times - that is actually really common on captions that position the spoken text close to the speaker - they leave the old text at the other speaker around for a bit so ppl can catch up on it

[02:45:01.0000] <nessy>
it's a really useful feature and doesn't cost us much, I think

[02:45:02.0000] <Hixie>
would be good to find some examples of that on youtube or something so i could study them "in action" as it were

[02:45:03.0000] <Dashiva>
Not to mention when multiple people speaking simultaneously

[02:46:00.0000] <Hixie>
right now i just have that one example

[02:46:01.0000] <Hixie>
right, i'm off

[02:46:02.0000] <Hixie>
later

[02:47:00.0000] <Dashiva>
http://dashiva.net/misc/multiple_speakers.jpg

[02:47:01.0000] <hsivonen>
/me wonders if the people tagging http://labs.opera.com/news/2010/04/22/ as "iphone" on Delicious know the difference between Opera Mini and Opera Mobile

[02:48:00.0000] <Dashiva>
Um, already examples for that, so no need I guess

[02:50:00.0000] <foolip>
SRT files often have overlapping times by mistake, because most (all?) player fix it during parsing

[02:50:01.0000] <nessy>
also about having multiple media resources in sync: I think you need both - (1) the dependent ones like audio descriptions whose timeline is totally dependent on the main timeline - they would go into <track>  - (2) the compositions of multiple media resources that may build a multimedia experience of sorts - they would be linked on the full <audio>/<video> level

[02:51:00.0000] <nessy>
foolip: players don't "fix" it - they just remove the old element when the new one appears

[02:51:01.0000] <foolip>
nessy: internally syncing them is the same think regardless of the syntax, so I think it's mostly a question of picking a syntax which makes sense

[02:51:02.0000] <nessy>
with srt that's the only thing you can do because they would be rendered in the same place

[02:52:00.0000] <asmodai>
hsivonen: Was going to try that now.

[02:52:01.0000] <foolip>
nessy: in any event the overlapping is concealed, so SRT files have overlap which wasn't intended as such by the author

[02:53:00.0000] <nessy>
the internal mechanism is only the same for some of it - e.g. if a dependent media resource is longer its overlength will never be played - if a synchronised media resource is longer, it will continue playing

[03:00:00.0000] <asmodai>
hsivonen: seems to grow with all addons and plugins disabled as well. But lets see, 336 MB now. Will see what it uses after a few minutes of not touching.

[03:01:00.0000] <foolip>
nessy: true

[03:17:00.0000] <asmodai>
hsivonen: doing nothing, now 466 MB and growing

[03:18:00.0000] <hsivonen>
asmodai: I suggest filing a bug

[03:23:00.0000] <asmodai>
hsivonen: Yeah, just have my developer instincts taking over trying to nail it down. Managed to with the XHTML Ruby addon. :)

[04:40:00.0000] <Dashiva>
And people still misunderstand the problem with hidden metadata...

[04:48:00.0000] <Lachy>
Dashiva, it just seems that some people are incapable of understanding when it's appropriate for specific metadata to be hidden, and when it's more optimal for the metadata to be based on visible content.

[04:52:00.0000] <Dashiva>
And the discussion about <p /> vs <p></p> seems to ignore the fact that empty <p> is non-conforming...

[05:01:00.0000] <remysharp>
Re: canvas 2d api - toDataURL and the second optional argument, compression level - between 0.0 and 1.  Is 1 highest compression or 1 being biggest sized file?

[05:03:00.0000] <gsnedders>
It's only supported for image/jpeg

[05:03:01.0000] <gsnedders>
And 1 is the highest quality, and hence biggest file

[05:03:02.0000] <Philip`>
"The second argument, if it is a number in the range 0.0 to 1.0 inclusive, must be treated as the desired quality level."

[05:03:03.0000] <Philip`>
I interpret 'quality level' as being exactly opposite to 'compression level'

[05:04:00.0000] <Philip`>
and that's how it's implemented - see http://philip.html5.org/tests/canvas/suite/tests/toDataURL.jpeg.quality.basic.html

[05:04:01.0000] <Philip`>
Uh, s/that's/what gsnedders said is/

[05:05:00.0000] <Dashiva>
It's a somewhat silly parameter, most users will have no idea what to put there

[05:05:01.0000] <Philip`>
Most users won't put anything there

[05:05:02.0000] <Philip`>
Surely most web developers have used a paint program that exports JPEG and has a quality slider, though?

[05:06:00.0000] <Dashiva>
Sure, but those programs don't explain the slider any better

[05:06:01.0000] <Philip`>
so they'll know that e.g. 70% (0.7) is typically okay

[05:06:02.0000] <Philip`>
Those programs dynamically display the result of the chosen compression quality

[05:07:00.0000] <Dashiva>
That sounds like a small subset of paint programs

[05:07:01.0000] <jgraham>
I guess photoshop does

[05:07:02.0000] <Philip`>
It's all the one I've used

[05:07:03.0000] <Philip`>
*ones

[05:07:04.0000] <jgraham>
The GIMP does

[05:07:05.0000] <jgraham>
s/guess/am pretty sure that/

[05:08:00.0000] <Philip`>
Paint Shop Pro does, since many years ago when I last used it

[05:08:01.0000] <jgraham>
I think lightroom doesn't but I haven't looked that hard because I know the defaults are fine

[05:09:00.0000] <Philip`>
/me wonders why Opera thinks quality 0.0 should give a larger file than quality 0.1

[05:09:01.0000] <jgraham>
Philip`: Does it give the same as quality 1.0?

[05:10:00.0000] <jgraham>
or no quality parameter, I guess

[05:10:01.0000] <remysharp>
out of interest, does the quality actually affect the size of the base64 data. I'm assuming it does, but I don't know much about the encoding techniques

[05:12:00.0000] <Philip`>
jgraham: No

[05:12:01.0000] <Philip`>
See e.g. http://philip.html5.org/tests/canvas/suite/tests/toDataURL.jpeg.quality.outsiderange.html and modify the last test

[05:13:00.0000] <Philip`>
(assuming the behaviour hasn't changed since the Opera I'm using)

[05:13:01.0000] <Philip`>
(which looks like 10.10)

[05:14:00.0000] <Philip`>
remysharp: Yes - base64 just does 4 bytes of output for each 3 bytes of input

[05:14:01.0000] <Philip`>
so if the input changes by 3 or more bytes then the output will definitely change length

[05:14:02.0000] <remysharp>
okay, cheers.

[05:15:00.0000] <remysharp>
ta for that all :)

[06:34:00.0000] <Dashiva>
/me wonders what kind of nut they had to use to fit perl in a nutshell. Coconuts?

[06:47:00.0000] <remysharp>
in need of canvas sanity check if someone is willing?

[06:48:00.0000] <remysharp>
I've got this simple example - and logging to the console (firebug, whatever) the length of pixel data - but I'm getting undefined: http://introducinghtml5.com/examples/ch05/getimagedata.html

[06:48:01.0000] <remysharp>
:-\

[06:48:02.0000] <Philip`>
remysharp: You want pixels.data.length

[06:48:03.0000] <remysharp>
I've been looking at this too much and now just confused myself, because I had this kind of simple example working before - any help would be greatly appreciated!

[06:48:04.0000] <Philip`>
not pixels.length

[06:48:05.0000] <remysharp>
Philip` thank you!

[06:48:06.0000] <remysharp>
see - me => going mad

[06:49:00.0000] <remysharp>
out of interest, when I run this offline - i.e. using file:/// I get a security error in firefox and chrome - should that be happening?

[06:50:00.0000] <asmodai>
So has this EU cookie issue been discussed in the whatwg yet?

[06:50:01.0000] <remysharp>
I know if I'm going across origins then it'll throw a security error, but I thought whilst offline the security model was different

[06:50:02.0000] <remysharp>
i.e. loose

[06:53:00.0000] <Philip`>
remysharp: With file:/// the security model is tighter in lots of ways

[06:53:01.0000] <Philip`>
so that you can't trick a user into downloading an HTML file to their desktop when then steals all their private data

[06:53:02.0000] <remysharp>
that makes sense

[06:54:00.0000] <Philip`>
s/when/which/

[06:54:01.0000] <remysharp>
I'm sure I had an example that was drawing images in to a canvas from a live domain (where my file was offline) but it was still able to call getImageData

[06:54:02.0000] <remysharp>
so I could do image manipulation offline

[06:54:03.0000] <remysharp>
in fact I've got a screencast of the demo - but it doesn't seem to work now - not sure if it's me, or if browsers have been updated in the last 4 monts

[06:54:04.0000] <remysharp>
*months

[06:55:00.0000] <remysharp>
perhaps they fixed it flagging as a security bug as you've suggested.

[06:58:00.0000] <Philip`>
remysharp: Accessing files from an http: location that's not the same origin as your script, and accessing ones from file: while you script is also on file:, sound like quite different security risks

[06:58:01.0000] <Philip`>
but they both sound like things that shouldn't be permitted

[06:59:00.0000] <Philip`>
/me has no idea what browsers' file: security models are like

[06:59:01.0000] <remysharp>
indeed - you're right, they don't sound like they're possible. Good chance I just caught it whilst they were figuring out the model.

[06:59:02.0000] <remysharp>
cheers for letting me bounce ideas.

[07:02:00.0000] <akahn>
Philip`: there was just a webkit issue about this recently

[07:02:01.0000] <akahn>
let me see if i can find it

[07:03:00.0000] <akahn>
my mistake, a chromium issue: http://code.google.com/p/chromium/issues/detail?id=37586

[07:41:00.0000] <zcorpan>
hmm, http://www.boingboing.net/2010/04/22/evil-witch-from-snow.html results in a page with nothing but an ad in opera and firefox with html5.enable, and nothing but the ad twice without html5.enable

[07:42:00.0000] <zcorpan>
with html5.enable it never stops loading

[07:42:01.0000] <zcorpan>
chrome loads as was intended

[07:46:00.0000] <doublec>
loads fine in firefox trunk 

[07:50:00.0000] <hsivonen>
zcorpan: how up-to-date is your build?

[07:50:01.0000] <hsivonen>
zcorpan: wfm regardless of the pref in a build I made this week with some local patches

[07:52:00.0000] <hsivonen>
zcorpan: also wfm regardless of whether I have "Firefox" or "Minefield" in the UA string

[07:53:00.0000] <hsivonen>
zcorpan: broken in Opera for me, too

[07:53:01.0000] <zcorpan>
hsivonen: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.3a5pre) Gecko/20100417 Minefield/3.7a5pre

[07:54:00.0000] <hsivonen>
zcorpan: wfm with that UA string, too

[07:55:00.0000] <Dashiva>
View source doesn't work, DOM only has a single ifram element

[07:55:01.0000] <Dashiva>
*iframe

[07:55:02.0000] <zcorpan>
hsivonen: with Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.5; en-US; rv:1.9.3a5pre) Gecko/20100423 Minefield/3.7a5pre i get two ads with html5.enable also

[07:55:03.0000] <Lachy>
zcorpan, works in Opera 10.10 for me. Fails in 10.50.

[07:56:00.0000] <Lachy>
zcorpan, check it in core, see where it regressed. I hope it's not a carakan bug

[07:58:00.0000] <gsnedders>
Lachy: Just because you were Carakan QA? :P

[07:58:01.0000] <gsnedders>
It's empty in 10.10 for me, having got zero bytes from the server

[08:00:00.0000] <hsivonen>
WFM in an actual Mac build, too

[08:02:00.0000] <hsivonen>
this polyglot spec isn't starting well

[08:03:00.0000] <jgraham>
hsivonen: In what way?

[08:03:01.0000] <jgraham>
/me isn't really following

[08:03:02.0000] <Lachy>
I don't get the point of the polyglot spec. It seems like it's just an authoring guide, and mine already covers much of that info

[08:03:03.0000] <hsivonen>
jgraham: in the way that it seems to repeat at least one superstition and in the way that bug reports about it have themselves been buggy

[08:03:04.0000] <Lachy>
anyway, I hope to find some time to get back to working on the authoring guide soon. It's been kind of left stagnant for far too long

[08:04:00.0000] <hsivonen>
s/bug reports/feedback on teh list/

[08:05:00.0000] <jgraham>
Polyglotness is so hard that even experts get it wrong? Who knew?

[08:05:01.0000] <hsivonen>
also, I can find terminology errors in Leif's and Lachy's emails in that thread

[08:06:00.0000] <Lachy>
my e-mails in which thread?

[08:06:01.0000] <hsivonen>
Lachy: "then you should include the BOM indicating UTF-16LE"

[08:07:00.0000] <hsivonen>
UTF-16LE is BOMless

[08:07:01.0000] <hsivonen>
the BOM indicates little-endian UTF-16

[08:07:02.0000] <gsnedders>
A leading U+FEFF is a ZWNBSP

[08:07:03.0000] <gsnedders>
(in UTF-16LE, UTF-16BE, UTF-32LE, UTF-32BE, etc.)

[08:07:04.0000] <hsivonen>
right

[08:08:00.0000] <gsnedders>
(it only isn't in UTF-8 because the spec specifically allows it, but assigns no meaning to it)

[08:08:01.0000] <Lachy>
I was using that as shorthand for indicating little-endian or big-endian, as indiciated by the BOM

[08:08:02.0000] <jgraham>
On the other hand Sam had just claimed that serving HTML as UTF-16 was non-conformant

[08:08:03.0000] <hsivonen>
Lachy: I know. I'm nitpicking

[08:08:04.0000] <jgraham>
And he's chair of the freaking working group

[08:08:05.0000] <gsnedders>
I didn't think he did that

[08:09:00.0000] <gsnedders>
I took it to mean that HTML processors aren't required to support UTF-16, unlike XML

[08:09:01.0000] <Dashiva>
I wouldn't mind if UTF-16 was non-conforming :)

[08:09:02.0000] <jgraham>
"UTF-16 is not valid for HTML5"

[08:09:03.0000] <jgraham>
I don't really know how to interpret that other than "not valid"

[08:10:00.0000] <jgraham>
If he meant "not required to be supported" then he should have said so

[08:10:01.0000] <gsnedders>
Try not literally interpreting anything. It might help with communication.

[08:10:02.0000] <Lachy>
Dashiva, there's nothing inherently wrong with UTF-16. But it's more useful if your document contains a significant proportion of characters that would be represented as 3 or 4 octets in UTF-8, compared with just 2 in UTF-16

[08:11:00.0000] <jgraham>
Although the fact that UTF-16 is not required is basically a cop-out by the spec

[08:11:01.0000] <jgraham>
Lachy: That's much less common than you would think

[08:11:02.0000] <Philip`>
gsnedders: Communication works better when you ignore what somebody actually said and pretend they said something totally different?

[08:11:03.0000] <Dashiva>
Lachy: Is it really? Even after applying gzip compression and such? And how many documents are affected?

[08:12:00.0000] <gsnedders>
Philip`: Totally.

[08:12:01.0000] <Philip`>
gsnedders: I'm glad you agree it's a bad idea

[08:12:02.0000] <zcorpan>
jgraham: maybe sam meant that UTF-16 isn't valid in <meta charset>

[08:12:03.0000] <jgraham>
/me remembers some study that looked at Japanese wikipedia and concluded that UTF-8 vs UTF-16 didn't make much difference

[08:12:04.0000] <Dashiva>
It comes with downsides too, such as not being ascii-compatible, having both BE and LE variants + BOM variants of each, and trouble with surrogates (if it's not actually UCS2 in disguise, that is)

[08:13:00.0000] <Philip`>
Lachy: HTML documents are largely markup and URLs, which is all ASCII, so you need an extremely high density of raw text to make up for it

[08:13:01.0000] <Philip`>
(Also they're a lot of whitespace indentation and word-spacing, which is also ASCII)

[08:14:00.0000] <jgraham>
zcorpan: Possibly. Who knows? It wasn't what he said though.

[08:15:00.0000] <Lachy>
jgraham, I expect it would be more common with asian language documents that contain a large amount of text, with comparitively little markup.

[08:15:01.0000] <Lachy>
I know that may not be common overall

[08:16:00.0000] <jgraham>
Lachy: It might make sense for the Japanese equivalent of Moby Dick on project Gutenberg

[08:16:01.0000] <jgraham>
It is something of an edge case on the web though

[08:17:00.0000] <Philip`>
I thought Project Gutenberg was text/plain

[08:17:01.0000] <jgraham>
http://www.gutenberg.org/files/2701/2701-h/2701-h.htm

[08:17:02.0000] <jgraham>
Not necessarily

[08:18:00.0000] <Dashiva>
Well, plain text would be even better fodder for compression, wouldn't it?

[08:19:00.0000] <Philip`>
Yes, but that's irrelevant to what is allowed in HTML

[08:21:00.0000] <Lachy>
has anyone seen any stats on how common UTF-16 usage is on the web?

[08:22:00.0000] <Lachy>
http://trends.builtwith.com/encoding/UTF-16-UCS-2

[08:23:00.0000] <Lachy>
not sure how representative that is of the web in general. I suspect it was biased by a significant number of western lanagues, which are predominately the latin alphabet.

[08:24:00.0000] <Lachy>
would probably be more interesting to limit the sample to langauges where using UTF-16 is potentially useful

[08:28:00.0000] <Philip`>
Lachy: http://philip.html5.org/data/charsets.html#charset-utf-16 are the only ones I saw in a sample that's largely Western but still has thousands of .jp/.cn/etc pages

[08:31:00.0000] <Lachy>
ok, so it's usage is negligable.

[08:33:00.0000] <jgraham>
Hmm. I just tried comparing the UTF-8 and UTF-16 encoded size of some japanese wikipedia articles and it seemed that utf-8 was smaller overall, but I suspect I did something wrong

[08:34:00.0000] <Dashiva>
Wikipedia has massive amounts of hidden markup, to be fair

[08:34:01.0000] <jgraham>
Is that atypical?

[08:35:00.0000] <Dashiva>
I would think so

[08:35:01.0000] <Dashiva>
Huge amounts of links, all of which are percent-encoded. Quite large head with inline <script> and such.

[08:39:00.0000] <jgraham>
http://event.rakuten.co.jp/borderless/index.html seems similar

[08:39:01.0000] <Lachy>
I got the same result with the jp Main_Page equivalent, and some other random article

[08:39:02.0000] <Lachy>
I then gzipped the results, and the UTF-8 still ended up smaller

[08:39:03.0000] <jgraham>
(in that case utf8 seems close to euc-jp which is the encoding the site uses)

[08:40:00.0000] <Lachy>
oh, I assumed the file I got from wikipedia was UTF-8.

[08:41:00.0000] <Dashiva>
Isn't it?

[08:41:01.0000] <Dashiva>
That's what I get, at least

[08:41:02.0000] <Lachy>
indeed, it is UTF-8. I don't know where you found a page using euc-jp

[08:42:00.0000] <Dashiva>
jgraham's link is euc-jp

[08:42:01.0000] <Lachy>
ah, I see. jgraham wasn't talking about wikipedia being euc-jp

[08:53:00.0000] <hsivonen>
the claim that UTF-16 is more compact for Japanese Web pages is a myth

[08:54:00.0000] <hsivonen>
real Web pages have markup, inline scripts, whitespace, inline style, etc.

[08:55:00.0000] <hsivonen>
jgraham: a couple of years ago, I measured wikipedia pages re-encoded in various ways

[08:55:01.0000] <hsivonen>
summary: gzipped UTF-8 is so good that everything else is pointless

[08:57:00.0000] <jgraham>
hsivonen: That would agree with my 5-minute "study"

[08:58:00.0000] <jgraham>
Did you compare with euc-js and similar?

[08:58:01.0000] <jgraham>
Did you write it up?

[09:06:00.0000] <hsivonen>
jgraham: http://lists.w3.org/Archives/Public/public-html-comments/2008Jan/0048.html and http://lists.w3.org/Archives/Public/public-html-comments/2008Jan/0076.html

[09:10:00.0000] <jgraham>
hsivonen: I think it would be great to blog any numbers that you have left over from that study

[09:15:00.0000] <Dashiva>
Perhaps if we talk about it enough, Philip` will make a service that takes a URL and outputs how big the page would be with different encodings and transfer options

[09:16:00.0000] <Philip`>
Surely you can do that with a single line of shell code

[09:16:01.0000] <Philip`>
using curl and iconv and gzip and wc and some loops

[09:22:00.0000] <Dashiva>
Philip`: Pretend this line is reverse psychology tricking you into making it to prove it's really possible in one line

[09:31:00.0000] <Philip`>
wget http://www.gutenberg.org/files/31757/31757-h/31757-h.htm; perl -e'for$c(qw(utf-8 utf-16be utf-16le utf-32 euc-jp shift-jis)){for$g(" ","|gzip -c"){print "$c$g ".`iconv -c 31757-h.htm -f utf-8 -t $c$g|wc -c`}}'

[09:33:00.0000] <Dashiva>
Accepted

[09:35:00.0000] <Philip`>
Interestingly the results are far tighter (and better) if you use bzip2

[09:35:01.0000] <Dashiva>
Is the default gzip compression level a reasonable assumption?

[09:38:00.0000] <Philip`>
Yes

[09:38:01.0000] <Philip`>
Don't know if it's a correct assumption, though

[09:38:02.0000] <Philip`>
Strangely, xz (LZMA) is slightly worse than bzip2 here

[09:42:00.0000] <Dashiva>
Not much difference with maximum compression on gzip anyhow

[09:44:00.0000] <Philip`>
You can save ~7% over gzip -9 by using 7z's gzip implementation

[09:46:00.0000] <Philip`>
(...though you wouldn't do that dynamically on a web server because it's way too slow)

[09:47:00.0000] <Philip`>
(but it's good whenever you're precomputing compressed files)

[09:55:00.0000] <JonathanNeal>
Heyo!

[09:57:00.0000] <Dashiva>
Philip`: But it doesn't seem to change the relative ranking of the methods

[09:57:01.0000] <Dashiva>
That's the most important thing, I think

[10:06:00.0000] <Philip`>
Dashiva: The most important thing is the size of the output, since the author's goal is to minimise that

[10:06:01.0000] <Philip`>
and if using a better compression implementation gives you a larger percentage gain than changing the charset, it's probably better for people to focus on the former before worrying too much about the latter

[10:08:00.0000] <Dashiva>
Compression is a different trade-off, though, since it also requires resources to (de)compress

[10:08:01.0000] <Philip`>
I guess so

[10:08:02.0000] <Philip`>
It's more complex to changing the compression code and to make it cached so it goes fast etc

[10:09:00.0000] <Philip`>
so maybe it's better to do dumb stuff like recompress static files on every single request, and make easy changes like reencoding the static file

[10:10:00.0000] <Dashiva>
But of course, in the end it's the size in actual bytes that matters

[10:11:00.0000] <Philip`>
Most compression algorithms seem to be designed to maximise resource usage on compression and minimise on decompression

[10:12:00.0000] <Philip`>
(e.g. LZMA can happily use 1GB RAM to compress, and a tenth of that for decompression)

[10:13:00.0000] <Philip`>
which seems precisely the wrong way around when you're doing compression on a single centralised server, and decompression on a thousand fast unutilised desktop PCs simultaneously

[10:25:00.0000] <AryehGregor>
Philip`, the server can cache the compressed files.

[10:25:01.0000] <AryehGregor>
Assuming they're more or less static.

[15:21:00.0000] <Hixie>
Dashiva: http://dashiva.net/misc/multiple_speakers.jpg - did those cues appear simultaneously?

[15:21:01.0000] <Hixie>
Dashiva: or do they appear staggered?

[15:21:02.0000] <Dashiva>
They interleave

[15:21:03.0000] <Dashiva>
Each image has one conversation going at its own pace

[15:21:04.0000] <Hixie>
interesting

[15:22:00.0000] <Hixie>
i'd love to study that in more detail

[15:22:01.0000] <Hixie>
is it online anywhere?

[15:22:02.0000] <Dashiva>
http://www.youtube.com/watch?v=bw5JBWdaUHI

[15:24:00.0000] <Hixie>
awesome

[15:25:00.0000] <Hixie>
sweet lord

[15:26:00.0000] <Hixie>
/me gets to t=0m40s

[15:26:01.0000] <Hixie>
shoot me now

[15:26:02.0000] <Dashiva>
Yeah, I don't think we need to copy those effects

[15:27:00.0000] <Hixie>
interesting, http://www.youtube.com/watch?v=bw5JBWdaUHI#t=0m12s shows that the red subs are fixed to line -2, while the blue subs are fixed to line -1

[15:28:00.0000] <Hixie>
they're not just stacking at the bottom first-come-first-served

[15:28:01.0000] <Hixie>
and percentage or pixel positioning isn't going to cut it for that

[15:28:02.0000] <Hixie>
we need line-based positioning

[15:29:00.0000] <Hixie>
excellent find, thanks Dashiva 

[15:33:00.0000] <Dashiva>
(Another effect I'm not sure is needed is having subtitles resize and move to follow the original text being animated)

[15:33:01.0000] <AryehGregor>
Hixie, I believe the correct generalization here is "ignore the Japanese, they're insane".

[15:33:02.0000] <Hixie>
with perspective transforms to boot? :-)

[15:34:00.0000] <AryehGregor>
(see also: recent Unicode emoticon discussion)

[15:34:01.0000] <Hixie>
AryehGregor: hha

[15:34:02.0000] <Hixie>
hah even

[15:34:03.0000] <Dashiva>
Hmm, perhaps

[15:34:04.0000] <Hixie>
AryehGregor: we don't want to ignore them, we just want to take the needs of the anime subbing community with a pinch of salt :-)

[15:35:00.0000] <Dashiva>
Even within the community there's plenty of pushback against the more advanced stuff

[15:35:01.0000] <Hixie>
for the more advanced effects, i expect we'll just have people overlaying a canvas

[15:35:02.0000] <Hixie>
or svg

[15:36:00.0000] <Dashiva>
AryehGregor: But the bad news is you have similar things being produced in France and even USA now :P

[15:36:01.0000] <AryehGregor>
Dashiva, we can ignore more countries on an as-needed basis.

[15:42:00.0000] <Dashiva>
Maybe just ignore karaoke as a valid use case, you can do a lot with the tools needed for just regular text

[15:42:01.0000] <Philip`>
Hixie: Perspective transforms would be great - then we could have subtitles which replace the Star Wars opening crawls with equivalently rendered text in other languages

[15:42:02.0000] <Dashiva>
/me groans

[15:43:00.0000] <Hixie>
indeed

[15:57:00.0000] <AryehGregor>
"From the current state of the discussion, it seems like removing Atom conversion would draw the weakest objections."

[15:57:01.0000] <AryehGregor>
It seems like removing things always draws the weakest objections.  Let's just put everything controversial in the WHATWG spec only so that W3C change proposals can't affect it.  Is that what people want?

[15:57:02.0000] <Dashiva>
Yes, a few of them

[15:58:00.0000] <AryehGregor>
Proposing that something be removed doesn't make it go away.

[15:58:01.0000] <AryehGregor>
Well, a few people might want that, but not the people who try to get these things removed.

[15:58:02.0000] <TabAtkins>
Actually, no, that's what anyone wants.  Some people want things removed from the WHATWG spec too.  They just can't get that.

[15:58:03.0000] <AryehGregor>
Maybe they figure that even if they can't get it removed from the WHATWG spec, at least they can remove it from the W3C spec, and that will make people less likely to see it?

[15:59:00.0000] <AryehGregor>
Of course, the more this happens, the less useful the W3C spec becomes, and the more people will use the WHATWG spec.

[15:59:01.0000] <Dashiva>
You can also see it as a PR battle, creating division between W3C and WHATWG

[16:01:00.0000] <AryehGregor>
The W3C has a more positive reputation than the WHATWG, certainly.  Much better known, too.

[16:02:00.0000] <Dashiva>
It's also likely that since the WHATWG copy exists, people have less incentive to fight to preserve the W3C copy

[16:02:01.0000] <AryehGregor>
That's true too.

[16:02:02.0000] <AryehGregor>
Fragmentation is often the easiest option, although rarely the best.

[16:03:00.0000] <AryehGregor>
See also microdata vs. RDFa.  If we knew in advance that only one could exist, whichever one we had would probably be better than either of the two contenders now.

[16:03:01.0000] <AryehGregor>
But instead, everyone who likes one of them just ignores the other rather than trying to fix its deficiencies.

[16:05:00.0000] <Dashiva>
To be fair, RDFa seems to be absorbing some of the good ideas of microdata. It just doesn't seem as willing to part with the bad ideas.

[16:05:01.0000] <TabAtkins>
AryehGregor: Fwiw, I've done a relatively thorough and open review of RDFa 1.1.  I think it's unfixable at this point, though.  They just made some initial architectural mistakes and have dug themselves down far enough in an attempt to reduce their impact that you can't "fix" it without substantially rewriting it.

[16:06:00.0000] <TabAtkins>
The basic failure is not paying enough attention to the tree-based structure of XML, and relying on a notion of implicit subjects, rather than declaring subjects more explicitly.

[16:07:00.0000] <TabAtkins>
Then they made it worse by trying to be clever with the chaining concept.

[16:07:01.0000] <AryehGregor>
Bah, someone complained about this wording!  http://www.w3.org/Bugs/Public/show_bug.cgi?id=7277

[16:08:00.0000] <AryehGregor>
I appreciated that wording when I started reading the spec.

[16:08:01.0000] <TabAtkins>
The whole prefix/curie thing is pretty much a non-issue at this point.  1.1's @vocab and @profile attributes successfully hide the complexity in common cases.

[16:10:00.0000] <TabAtkins>
Basically, chapter 8 of RDFa1.1 speaks for itself about why the basic RDFa architecture is a mistake.

[16:11:00.0000] <TabAtkins>
It has like 60 examples, possibly more, walking through several things step-by-step because otherwise it's impossible to decipher.

[16:12:00.0000] <Hixie>
TabAtkins: complexity can't be hidden unless it's gone from the syntax altogether, because people have to maintain other people's code, and the other person might use the feature that's complicated

[16:13:00.0000] <TabAtkins>
True.  It's at least lessened.  Using @vocab *looks* like Microdata with @itemtype on simple cases.

[16:13:01.0000] <AryehGregor>
. . . of course, you could say the same about HTML.  But I guess you'd say that there are few enough RDFa agents that we care about that we can break compatibility anyway.

[16:14:00.0000] <TabAtkins>
AryehGregor: Exactly.

[16:14:01.0000] <AryehGregor>
Or, more importantly, few enough RDFa pages.

[16:14:02.0000] <AryehGregor>
Since a few implementations can be changed pretty easily.

[16:15:00.0000] <TabAtkins>
Once I finish setting up my workflow, I might spend a few hours hacking our Rich Snippets support to ensure that it works with Microdata properly.

[16:15:01.0000] <Dashiva>
It seems to be some of the same people saying we can ignore deployed HTML5 content because it's not a standard yet who proclaim existing RDFa-in-HTML content must be supported

[16:15:02.0000] <TabAtkins>
Those people don't have a point, they're just inconsistent due to their agenda.

[16:16:00.0000] <AryehGregor>
/me tried to rewrite that to reverse the roles, but it didn't work

[16:16:01.0000] <TabAtkins>
AryehGregor: Yay!  That's how you know it's true.

[16:16:02.0000] <Dashiva>
I'm sure you can reverse it

[16:16:03.0000] <Dashiva>
You just need to apply a secondary measure

[16:16:04.0000] <Dashiva>
E.g. we can't ignore [massively] deployed HTML5 content, but we can ignore [small amounts of] RDFa-in-HTML

[16:17:00.0000] <AryehGregor>
Well, we ignore deployed RDFa+HTML content because there's not a lot of it, but there's a lot of HTML content, so it seems consistent.  I'd need to be more creative.

[16:17:01.0000] <TabAtkins>
Indeed, but if you apply that measure with the original statement, it becomes obviously ridiculous.

[16:17:02.0000] <Dashiva>
It seems obvious, but I'm not sure it's necessarily so

[16:17:03.0000] <Dashiva>
e.g. if you take the claims that blog software ships with it enabled by default, and doing something useful by default

[16:18:00.0000] <AryehGregor>
Most blog software also outputs HTML.

[16:18:01.0000] <Dashiva>
I personally don't think anyone would notice even if it was deployed by default, but I don't have the data to back it up

[16:19:00.0000] <AryehGregor>
I think WHATWG people tend to only really care about browsers (and maybe search engines and feed readers and such), while the specific technology is usually more or less of secondary interest.  The RDFa people seem to be more invested in a particular technology than in a particular high-level goal.

[16:19:01.0000] <AryehGregor>
The same is true for lots of people who disagree with WHATWG people, actually.

[16:19:02.0000] <TabAtkins>
That technology is "namespaces", for RDFa.

[16:20:00.0000] <Dashiva>
And RDF ;)

[16:21:00.0000] <TabAtkins>
RDF's not necessarily a problem.  You could parse RDFa data into JSON with roughly the same ease as you parse Microdata into RDF.

[16:21:01.0000] <AryehGregor>
Is there anything wrong with RDF, or just RDFa?

[16:21:02.0000] <othermaciej>
RDF as a data model

[16:21:03.0000] <TabAtkins>
RDF is just a rolled-out version of tree-based data.

[16:21:04.0000] <othermaciej>
?

[16:21:05.0000] <TabAtkins>
With a focus on URIs.

[16:21:06.0000] <othermaciej>
RDF is not tree-based data, it is graph-based

[16:22:00.0000] <Dashiva>
Not necessarily wrong, but coming from a graph structure and designing for a tree is tricky

[16:22:01.0000] <Dashiva>
Graphs should be the exception, trees the norm

[16:23:00.0000] <Dashiva>
But if you come from a world where everything is graphs, it can be hard to adapt

[16:23:01.0000] <TabAtkins>
All trees are graphs, and they're more common in real data, so I still think my gloss is mostly accurate.

[16:23:02.0000] <erlehmann>
I like graphs.

[16:23:03.0000] <Dashiva>
TabAtkins: It's a graph in total, but each page usually only contains a tree-like subgraph

[16:23:04.0000] <TabAtkins>
Dashiva: Right, that's what I tried to say, but failed to do so clearly.

[16:24:00.0000] <TabAtkins>
Thanks for the clarification.  ^_^

[16:25:00.0000] <TabAtkins>
So anyway, that's why RDFa's essential architecture is wrong.  It didn't take full advantage of the parallel between most data's tree-nature and HT/XML's tree-like nature.

[16:26:00.0000] <othermaciej>
TabAtkins: all trees are directed graphs, but not vice versa

[16:26:01.0000] <TabAtkins>
othermaciej: That's what I said, isn't it?

[16:26:02.0000] <othermaciej>
earlier you said "RDF is just a rolled-out version of tree-based data."

[16:26:03.0000] <othermaciej>
maybe I misinterpreted that statement

[16:27:00.0000] <TabAtkins>
Right, I left out some details that Dashiva corrected me on.

[16:27:01.0000] <TabAtkins>
In a hand-wavey way, I mean that the RDF structure of data embedded in pages is typically a tree, or at least very tree-like.

[16:27:02.0000] <othermaciej>
I agree that a lot of interesting data can be expressed well enough with a tree, rather than a general directed graph

[16:28:00.0000] <othermaciej>
I suspect a lot of RDFa at least points to common objects and so is likely at least a DAG

[16:28:01.0000] <Dashiva>
Yeah, I should have said DAG instead of tree. Even microdata is like that.

[16:29:00.0000] <TabAtkins>
Yeah.

[16:29:01.0000] <othermaciej>
Microdata is a DAG?

[16:29:02.0000] <Dashiva>
With itemref

[16:29:03.0000] <TabAtkins>
And, in a larger sense, with itemid.

[16:29:04.0000] <othermaciej>
itemref just clones some properties though

[16:30:00.0000] <othermaciej>
does the actual Microdata model retain a DAG structure?

[16:30:01.0000] <othermaciej>
are you pointing to the same conceptual object, or do you just happen to have properties with the same values?

[16:30:02.0000] <TabAtkins>
No, it serializes as a tree.

[16:30:03.0000] <TabAtkins>
But, if you lean on urls as values, it retains exactly as much DAGness as RDF would.

[16:33:00.0000] <Hixie>
RDF isn't natively a graph at all, it's a list of triples

[16:33:01.0000] <Hixie>
actually even that's not accurate

[16:33:02.0000] <Hixie>
it's a list of quads

[16:34:00.0000] <Hixie>
with some annotations

[16:34:01.0000] <Hixie>
but you can form a graph from that list

[16:34:02.0000] <TabAtkins>
The graphness comes from defining a form of url-equality, yeah.

[16:34:03.0000] <Hixie>
microdata is natively a tree, that you can flatten to a list of triples, from which you can also create a graph

[16:34:04.0000] <TabAtkins>
What's the quad?  The object type?

[16:34:05.0000] <Hixie>
value type

[16:35:00.0000] <TabAtkins>
Right, yeah.

[16:35:01.0000] <Hixie>
and annotations are language of the value if it's a literal, and source of the triple

[16:35:02.0000] <Hixie>
also the literal can itself be structured XML

[16:37:00.0000] <othermaciej>
RDF serializations are a list of triples/quads/whatever

[16:37:01.0000] <othermaciej>
but RDF the data model is a graph

[16:39:00.0000] <TabAtkins>
Under your wording, othermaciej, Microdata's data model can indeed be a DAG, if you lean on url-equality to associate nodes as being the same (the same thing you have to do to make RDF a graph).

[16:40:00.0000] <Hixie>
right

[16:40:01.0000] <othermaciej>
but no cycles?

[16:40:02.0000] <Hixie>
microdata itself isn't but when you convert it to RDF you can get the same kind of graph with cycles

[16:40:03.0000] <Hixie>
the only thing you can't do cycles through in microdata iirc is bnodes

[16:40:04.0000] <Hixie>
but since you can just give them unique IDs, that becomes a non-issue

[16:42:00.0000] <TabAtkins>
Yeah, bnodes in Microdata are only as blank as you allow them to be.

[16:42:01.0000] <othermaciej>
I don't really understand what a bnode is

[16:42:02.0000] <TabAtkins>
A node in the data graph without an associated url.

[16:42:03.0000] <Hixie>
it's a node without an identifier

[16:43:00.0000] <TabAtkins>
In Microdata's RDF serialization, a bnode is the subject of all the triples generated by an @itemscope without an @itemid.

[16:44:00.0000] <Hixie>
it's hard to say exactly what the RDF model is since there's no official API; the only reason I'd say microdata's model is a tree and not the RDF graph is that there's an explicit microdata API that does't follow the URL refs and thus makes it a tree

[16:45:00.0000] <othermaciej>
the API being proposed by RDFa doesn't expose a graph or tree at all

[16:45:01.0000] <othermaciej>
just a list of triples

[16:45:02.0000] <Hixie>
right, like i said, RDF's model is just a list :-)

[16:45:03.0000] <othermaciej>
which seems like a pretty lousy interface

[16:45:04.0000] <Hixie>
which by convention can be interpreted as a graph

[16:47:00.0000] <TabAtkins>
I believe you're just supposed to parse the triples out and then hand them to a triple-store which will make the node associations for you.

[16:47:01.0000] <Hixie>
yeah the usual way of working with RDF is through queries, not through an API

[16:47:02.0000] <othermaciej>
but if you are running as JS in the browser, you don't have a triple store available

[16:47:03.0000] <othermaciej>
I expected the RDFa API itself to be a "triple store"

[16:47:04.0000] <othermaciej>
maybe that was a bad assumption

[16:47:05.0000] <TabAtkins>
You'd be disappointed, then.  ^_^

[16:47:06.0000] <othermaciej>
er

[16:48:00.0000] <othermaciej>
assumption

[16:48:01.0000] <Hixie>
it's not clear to me that the people designing the API are doing it for any reason other than to say they have an API

[16:48:02.0000] <othermaciej>
it seems like this model makes it very hard to have a live triple store on a dynamically modified document

[16:48:03.0000] <Hixie>
(unlike the microdata API, which is designed to make the information available)

[16:48:04.0000] <Hixie>
but then browser vendors are even less likely to implement a real triple store with querying than to implement the flat API

[16:48:05.0000] <Hixie>
so...

[16:49:00.0000] <othermaciej>
a triple store API would be more complicated to implement, but on the other hand it might have a real use case


2010-04-24
[17:38:00.0000] <kennyluck>
Firefox does have this triple store implementation. Just wonder how old it is.

[17:39:00.0000] <kennyluck>
I missed the discussion I should have joined :(

[18:53:00.0000] <JonathanNeal>
Woot

[18:53:01.0000] <JonathanNeal>
http://sandbox.thewikies.com/createElement/

[18:53:02.0000] <JonathanNeal>
vs http://sandbox.thewikies.com/createElement/before.html

[18:54:00.0000] <TabAtkins>
Should I be looking at that in IE?

[18:56:00.0000] <TabAtkins>
Also: your loop syntax is so horrible.  ;_;  I really don't understand why you value a microsecond or two over the clarity of a for loop.

[18:58:00.0000] <TabAtkins>
Otherwise, though, seems very clever.  I approve.

[19:03:00.0000] <TabAtkins>
Argh, I'm introducing too much abstraction here, but I'm not sure of the best way to collapse one of the abstraction levels.  >_<

[19:03:01.0000] <TabAtkins>
Anytime you have to introduce *two* new levels of abstract directional terminology, you're doing something wrong.

[20:23:00.0000] <jlebar>
Hixie, Is there a reason that offline manifests don't treat form feed characters (and other unicode space characters) as whitespace?

[20:24:00.0000] <jlebar>
e.g. step 17 in http://www.whatwg.org/specs/web-apps/current-work/multipage/offline.html#writing-cache-manifests

[20:24:01.0000] <annevk>
consistency

[20:24:02.0000] <jlebar>
annevk, with what?

[20:24:03.0000] <annevk>
text/plain, HTML, CSS, etc.?

[20:25:00.0000] <jlebar>
the split on spaces algorithm does consider ff to be whitespace.

[20:26:00.0000] <jlebar>
http://www.whatwg.org/specs/web-apps/current-work/multipage/common-microsyntaxes.html#skip-whitespace

[20:26:01.0000] <annevk>
sure, but newlines are always just CR and LF or CRLF

[20:27:00.0000] <jlebar>
Right.  But the step in the algorithm I first mentioned is stripping whitespace off the end of a line.

[20:29:00.0000] <annevk>
aah I see now

[20:29:01.0000] <annevk>
sorry

[20:30:00.0000] <jlebar>
np

[20:30:01.0000] <annevk>
seems like a bug of some sorts

[20:30:02.0000] <jlebar>
Perhaps.  But Hixie may have some grand plan.  He often does.  :)

[20:32:00.0000] <annevk>
maybe he tries to keep FF out of cache manifests but the inconsistency seems confusing

[00:32:00.0000] <Hixie>
anyone remember what the term is for a block's overall bidi direction?

[00:35:00.0000] <annevk>
base writing direction of blocks?

[00:35:01.0000] <annevk>
(comes from a google search)

[00:35:02.0000] <Hixie>
doesn't sound right

[00:36:00.0000] <Hixie>
i mean the post-bidi-algorithm direction

[00:36:01.0000] <Hixie>
the one that decides whether 'text-align:start' is left or right

[00:37:00.0000] <Hixie>
wait, text-align is dependent on 'start', not on the actual direction?

[00:37:01.0000] <Hixie>
er, on 'direction', i mean, not 'start'

[00:37:02.0000] <annevk>
dunno, sorry

[00:38:00.0000] <annevk>
haven't followed that stuff recently since in discussions vertical text is added on top of it which makes my brain explode (or at least wander off)

[00:38:01.0000] <Hixie>
i guess i mean the value that decides whether a "." at the end of the string goes to the left or the right in the rendering

[00:41:00.0000] <Hixie>
looks like i mean the direction of the character found in the bidi algorithm step P2

[00:42:00.0000] <Hixie>
which seems to be called the "paragraph direction"

[00:43:00.0000] <Hixie>
ah, the paragraph embedding level

[00:43:01.0000] <Hixie>
there we go

[00:45:00.0000] <Hixie>
man it's been so long since i've worked on bidi stuff

[00:49:00.0000] <Hixie>
annevk: yeah, bidi spec says it's sometimes called the "base direction", thanks

[00:53:00.0000] <annevk>
what are you working on btw? there's no open bidi issues afaict (apart from the unfinished i18n document I guess)

[00:55:00.0000] <tagawa>
hello!

[00:55:01.0000] <annevk>
hey tagawa :)

[00:55:02.0000] <tagawa>
(sorry, just testing)

[00:56:00.0000] <Hixie>
annevk: positioning of subtitles

[00:58:00.0000] <annevk>
ah right, forgot about captioning

[00:58:01.0000] <annevk>
o_O

[00:58:02.0000] <Hixie>
heh

[02:14:00.0000] <nessy>
good that somebody with knowledge looks at i18n issues for subtitles!

[02:42:00.0000] <JonathanNeal>
Updated http://sandbox.thewikies.com/createElement/before.html and http://sandbox.thewikies.com/createElement/ to include multiple libraries working with html5 elements successfully.

[02:44:00.0000] <jgraham>
JonathanNeal: Which code is the code that you actually wrote?

[02:44:01.0000] <jgraham>
i.e. what should I be looking at?

[02:47:00.0000] <JonathanNeal>
I wrote the tests, and I wrote the "shimproof" script.

[02:47:01.0000] <JonathanNeal>
It is an evil script that proves a point, but beyond that, it should not be used.  Rather, it should be implemented into js libraries.

[02:47:02.0000] <jgraham>
You know it would be friendly to include a non-minimised version if you want people to read it :)

[02:48:00.0000] <JonathanNeal>
I thought shimproof.min.js made it obvious, but sure http://sandbox.thewikies.com/createElement/shimproof.js

[02:48:01.0000] <JonathanNeal>
It's just an experiment :D

[02:49:00.0000] <jgraham>
/me wonders if he should propose a way of accessing the AST from javascript to TC39 so that it is easy to implemnt pretty-printers

[02:49:01.0000] <jgraham>
(as well as all sorts of other exciting things)

[02:49:02.0000] <jgraham>
(I guess they would hate it though)

[02:52:00.0000] <Philip`>
Why do you need a standardised API for implementing pretty-printers?

[02:52:01.0000] <Philip`>
Isn't it fine to just have a standalone pretty-printing tool that uses private APIs of a specific implementation?

[02:55:00.0000] <jgraham>
Philip`: I don't really. But it's not like there are lots of existing implementations that expose private APIs to do what I want either

[02:56:00.0000] <Philip`>
Getting an implementation to expose a private API sounds like far less effort than getting every implementation to expose a public API

[02:57:00.0000] <jgraham>
It's really unclear that that is true :)

[02:58:00.0000] <jgraham>
Also I guess it would be nice if you don't need java in order to use most existing minifiers

[02:58:01.0000] <jgraham>
*didn't

[02:59:00.0000] <Philip`>
How can it not be true? The public API is as much effort as doing the private API multiplied by the number of implementations, plus the cost of coordination

[03:00:00.0000] <jgraham>
Because the cost to me isn't the implementation cost it is the cost of convincing people

[03:00:01.0000] <Philip`>
I was assuming you weren't being so selfish as to only consider the cost to yourself ;-)

[03:01:00.0000] <Philip`>
(Given the apparent pain and suffering involved with SpiderMonkey's decompiler, I doubt implementors want to do much more of that kind of thing - computing and storing perfect lossless ASTs seems to conflict with optimisation goals, and deriving an AST from more optimised data structures is hard)

[03:03:00.0000] <jgraham>
JonathanNeal: Why the use of the foo[constant] syntax everywhere rather than foo.constant? The latter is more readable and, at least in some modern engines, more performant (IE might be different of course)

[03:04:00.0000] <JonathanNeal>
Just to make the script under 1k minified

[03:04:01.0000] <JonathanNeal>
It doesn't really matter.

[03:06:00.0000] <Dashiva>
http://www.youtube.com/watch?v=MHrTE-PLAsc&feature=related#t=6m20s

[03:07:00.0000] <Dashiva>
I wonder if you can justify subtitling something people wouldn't normally be able to hear properly anyway

[03:23:00.0000] <webben>
Dashiva: Sure. TV often does that when voices are hard to hear for some reason (e.g. bad recording).

[03:41:00.0000] <Philip`>
(It's interesting to see when TV shows consider English-speaking people to have a sufficiently peculiar accent that they have to be subtitled)

[03:43:00.0000] <Philip`>
(e.g. it seems Americans sometimes find Scottish accents tricky)

[05:12:00.0000] <Dashiva>
Philip`: There's a funny video like that

[05:12:01.0000] <Dashiva>
http://www.dailymotion.com/video/x3084g_newsreportfromiraq_fun

[05:26:00.0000] <hsivonen>
http://en.wikipedia.org/wiki/HTML5_video doesn't seem to conform to Wikipedia policies

[05:27:00.0000] <hsivonen>
lots of [citation needed] and [weasel words] missing there

[05:27:01.0000] <Philip`>
Dashiva: I like how the subtitles get completely obscured by the popup advert

[05:33:00.0000] <mr_daniel>
I wonder where the 'dblclick' event is specified? I just read the DOM Level 2 Events recommendation, specifically the mouse events. The 'dblclick' event is not described: http://www.w3.org/TR/DOM-Level-2-Events/events.html#Events-MouseEvent

[05:34:00.0000] <mr_daniel>
but on the other hand the wikipedia article about DOM Level 2 Events states that there is a 'dblclick' event http://en.wikipedia.org/wiki/DOM_events#HTML_events

[05:35:00.0000] <mr_daniel>
so where does the 'dblclick' event come from? Is it supported by all browsers, which support DOM Level 2 Events? Or should I avoid the use of the 'dblclick' event?

[05:48:00.0000] <nessy>
Dashiva - hehe - I really enjoyed that video ;)

[06:31:00.0000] <Henrik`G>
hey peoples i have a box with links of latest posts in my sidebar

[06:31:01.0000] <Henrik`G>
and i am wondering if it is a section or a div would be the best use

[06:31:02.0000] <Henrik`G>
my sense says section because it starts with a header and then th ul

[06:31:03.0000] <Henrik`G>
the

[08:22:00.0000] <ray>
hehe, [[deep pockets]]

[10:40:00.0000] <Dashiva>
I think Dean needs a timeout...


2010-04-25
[22:20:00.0000] <magcius>
I think this is a bug

[22:20:01.0000] <magcius>
but it happens in all major browsers

[22:21:00.0000] <magcius>
http://magcius.mecheye.net/canvasbug.html

[22:21:01.0000] <magcius>
why is there a magical 4px added to html's scrollHeight when canvas's height is 100%?

[03:49:00.0000] <cryzed>
Hey :)

[03:50:00.0000] <cryzed>
I have a problem: I use the html5lib for python < 3. When I do html5lib.parse(source, 'etree') (ElementTree is installed) it returns None

[03:50:01.0000] <cryzed>
instead of a parsed ElementTree

[03:51:00.0000] <cryzed>
Any ideas? Anyone?

[05:12:00.0000] <jgraham>
cryzed: You probably need to pass another argument to say which etree implementation you want to use

[05:25:00.0000] <cryzed>
jgraham, no it was a bug

[05:25:01.0000] <cryzed>
jgraham, thanks tho

[07:16:00.0000] <nessy>
is the whatwg wiki down?

[07:17:00.0000] <nessy>
ah seems fixed

[11:20:00.0000] <hober>
TabAtkins: up for grabbing dinner and/or beer tonight? (I'm at jqcon, as are paul_irish & miketaylr)

[11:20:01.0000] <miketaylr>
:D


2010-04-26
[18:40:00.0000] <AryehGregor>
http://lists.wikimedia.org/pipermail/wikitech-l/2010-April/047617.html

[18:40:01.0000] <AryehGregor>
/me waits for someone here to inform him of the various factual errors he made about XML in that post

[20:32:00.0000] <annevk>
everyone else was on a weekend break too?

[20:32:01.0000] <annevk>
no fun logs to read :/

[20:33:00.0000] <nessy>
hehe! we have a long w/e here!

[20:33:01.0000] <othermaciej>
annevk: you could make your own fun logs!

[20:33:02.0000] <othermaciej>
just pick someone to flame

[20:33:03.0000] <doublec>
nessy: really? You have an extra day for anzac day?

[20:34:00.0000] <roc>
they do

[20:34:01.0000] <doublec>
nice, the kiwi's get ripped off in that regard

[20:34:02.0000] <roc>
actually, NZ does too, I just lied to get you guys into the office

[20:35:00.0000] <doublec>
haha

[20:35:01.0000] <nessy>
doublec: yes, anzac day holiday! means I get to catch up on work! :)

[20:36:00.0000] <doublec>
nessy: in NZ we don't get an extra day if anzac day falls on a weekend

[20:36:01.0000] <nessy>
so, no difference to you either

[20:36:02.0000] <nessy>
it's a nice feature of Australian holidays

[20:36:03.0000] <nessy>
if they fall on weekends you get them back during the next week

[20:36:04.0000] <nessy>
unknown in Germany, too

[20:36:05.0000] <roc>
it's true for the Christmas holidays here

[20:37:00.0000] <roc>
I don't know why Anzac Day is different

[20:38:00.0000] <doublec>
apparently they don't 'mondayize' holidays that have  actual national historical significance

[20:38:01.0000] <doublec>
so waitangi day is also left out

[20:42:00.0000] <nessy>
that's inconsistent

[20:43:00.0000] <nessy>
I have a question about Firefox and video - does Firefox store the byte ranges in an actual file on disk, so when all byte ranges have been downloaded the full original files is available? or do you just keep the byte ranges in memory and throw them out?

[20:44:00.0000] <doublec>
we keep the ranges on disk in a 'media cache' but throw them out when you leave the page

[20:45:00.0000] <doublec>
we also cache the video in the standard browser cache if it meets caching requirements - but due to the nature of video (seeking, large files, etc) it often doesn't meet those requirements

[20:46:00.0000] <doublec>
the media cache is shared across all videos being played too - and defaults to 50MB iirc. As more video gets played old entries in the cache get evicted

[20:47:00.0000] <nessy>
so in the 'media cache' you won't typically end up with the full original file

[20:47:01.0000] <nessy>
just its byte ranges as temp storage?

[20:47:02.0000] <doublec>
nessy: http://weblogs.mozillazine.org/roc/archives/2009/04/media_cache.html

[20:47:03.0000] <doublec>
nessy: correct

[20:47:04.0000] <nessy>
I'm asking, since it's then not strictly progressive download as described here http://weblogs.mozillazine.org/roc/archives/2009/04/media_cache.html

[20:48:00.0000] <roc>
why does it matter?

[20:48:01.0000] <nessy>
it's more like actual streaming and more secure  - except for the "Download src" right context menu functionality :)

[20:49:00.0000] <nessy>
thanks for the link!

[20:49:01.0000] <nessy>
ppl are concerned about the possibility to "steal" the original file in a simple way

[20:49:02.0000] <doublec>
that's pretty easy with html 5 video...

[20:50:00.0000] <nessy>
I have a potential customer who wants to use rtp/rtsp instead of http server for html5 video because it's less prone to stealing

[20:50:01.0000] <doublec>
in the same was as 'stealing' an image is

[20:50:02.0000] <nessy>
yes, exactly

[20:50:03.0000] <doublec>
I'm pretty sure rtp/rtsp will only work in one browser

[20:50:04.0000] <doublec>
if at all

[20:50:05.0000] <doublec>
(I'm thinking Safari)

[20:50:06.0000] <nessy>
yes, I think so too

[20:50:07.0000] <nessy>
are you planning to support it?

[20:51:00.0000] <doublec>
There's a bug for it. No known plans to do anything in the immediate future.

[20:51:01.0000] <nessy>
fair enough

[20:52:00.0000] <nessy>
I don't think Opera or Chrome support it either or plan to

[20:52:01.0000] <doublec>
https://bugzilla.mozilla.org/show_bug.cgi?id=506834

[20:52:02.0000] <nessy>
and I'm not sure Safari supports it properly other than their HTTP adaptive streaming

[20:54:00.0000] <annevk>
I wish these DRM-requesting types would realize they're wrong :/

[20:54:01.0000] <nessy>
https://bugs.webkit.org/show_bug.cgi?id=26416

[20:55:00.0000] <nessy>
annevk: this guy is offering paid content and ppl go steal his stuff and resell it - I don't think there is anything wrong with wanting some control over this situation

[20:56:00.0000] <doublec>
the people who want to steal it and resell it have the resources to rip it using any tool they want

[20:56:01.0000] <doublec>
no matter what you do

[20:57:00.0000] <doublec>
it's hard to the people offering the content of that though

[20:57:01.0000] <doublec>
there's a 'convince' in between 'to' and 'the' there :)

[20:58:00.0000] <nessy>
he said he cut the problems by 85% when moving from http to rtsp

[20:58:01.0000] <nessy>
that's substantial when you're running a biz

[21:00:00.0000] <roc>
that sounds odd

[21:00:01.0000] <doublec>
how did he measure that?

[21:00:02.0000] <roc>
perhaps there's no off-the-shelf RTSP ripper yet and all his enemies are too stupid to write one

[21:00:03.0000] <nessy>
it's safety by more obscurity, is all I think

[21:00:04.0000] <doublec>
there's a few RTSP rippers around for Windows at least. And gstreamer, etc can be used on Linux.

[21:01:00.0000] <doublec>
what % of viewers did he lose going to rtsp?

[21:01:01.0000] <nessy>
he frequently trolls all the public video sites for his content, cause that's where it turns up, too

[21:01:02.0000] <roc>
doing a ton of work and integrating heaps more code into the browser just to raise the obscurity bar a minute amount doesn't sound like a win to me

[21:01:03.0000] <nessy>
non cause he did it all in Flash

[21:01:04.0000] <nessy>
might have been Flash steaming rather than rtsp, actually

[21:02:00.0000] <doublec>
interesting

[21:02:01.0000] <doublec>
when poker training sites first started out they all used rtsp or flash streaming solutions with DRM for training videos

[21:03:00.0000] <doublec>
the newer sites however moved to DRM free

[21:03:01.0000] <roc>
05e790926d9ee13e08916f878432adbb

[21:03:02.0000] <doublec>
and got more customers as a result

[21:03:03.0000] <doublec>
because people could play them anywhere

[21:03:04.0000] <doublec>
and the videos weren't pirated any more than the DRM video sites videos

[21:04:00.0000] <doublec>
/me starts a cracking roc's hash

[21:04:01.0000] <nessy>
yeah - I am not convinced either

[21:04:02.0000] <nessy>
I didn't want to encourage rtsp implementation - was just curious about status

[21:05:00.0000] <nessy>
he wants to move away from flash and to html5 video cause he hates all the flash stuff - so we'll have to wait and see if that will create pirating again

[21:06:00.0000] <nessy>
I just needed to know more details :)

[21:06:01.0000] <nessy>
thanks!

[21:49:00.0000] <annevk>
roc, again? :)

[21:49:01.0000] <annevk>
I believe at least one is still without public answer, right?

[21:54:00.0000] <roc>
they all are

[22:28:00.0000] <jlebar>
roc, how long are the messages, on average?

[22:28:01.0000] <roc>
it varies :-)

[22:28:02.0000] <roc>
this one was just a one-liner

[22:33:00.0000] <jlebar>
Sounds like it may need some salt.  :)

[22:33:01.0000] <jlebar>
(Interesting idea for a cryptographic protocol: You give me a hash, and at the same time, proof that the message is a meaningful English sentence, without revealing the sentence.)

[22:33:02.0000] <slango>
am I crazy, or is this page poison to Firefox: http://www.whatwg.org/specs/web-apps/current-work/?

[22:34:00.0000] <jlebar>
slango, It's better in trunk.

[22:45:00.0000] <slango>
thanks

[23:12:00.0000] <magcius>
Is this a bug? The scrollbar appears in http://magcius.mecheye.net/canvasbug.html

[23:13:00.0000] <magcius>
I looked into it, and a linebreak textnode is somehow being added.

[23:13:01.0000] <magcius>
Causing the html tag's scrollHeight to be 4px larger.

[23:13:02.0000] <magcius>
It happens in all major browsers

[00:55:00.0000] <annevk>
/me wonders if he should write a reply to http://lists.w3.org/Archives/Public/public-webfonts-wg/2010Apr/0037.html

[00:55:01.0000] <annevk>
it seems sort of set in stone already that CORS will be abused for fonts

[00:55:02.0000] <annevk>
I guess I'll write my thoughts anyway, so they can at least be considered

[00:56:00.0000] <roc>
I'd like to read that

[01:10:00.0000] <annevk>
hmm, I hope my email gets through

[01:39:00.0000] <annevk>
worked out: http://lists.w3.org/Archives/Public/public-webfonts-wg/2010Apr/0040.html

[01:53:00.0000] <hsivonen>
hmm. what do I need to do to make <svg> not clip its contents to the box established by <svg>?

[01:53:01.0000] <othermaciej>
mess with the viewbox attributes

[01:54:00.0000] <othermaciej>
(I think)

[01:54:01.0000] <hsivonen>
no, viewBox sets up the coordinate space

[01:54:02.0000] <zcorpan>
possibly preserveAspectRatio can change it, but i think at least in opera it'll always be clipped

[01:54:03.0000] <hsivonen>
but how do I make shapes outside the viewBox box get painted outside the box

[01:55:00.0000] <hsivonen>
zcorpan: oh so the IE9 demo that gets clipped in "other browsers" isn't about "defaults" but about other browsers always clipping?

[01:56:00.0000] <zcorpan>
dunno

[01:56:01.0000] <othermaciej>
I don't know what the IE9 demo is specifically

[01:56:02.0000] <othermaciej>
SVG in theory requires (effectively) that SVG root elements have overflow: hidden !important in the UA stylesheet

[01:57:00.0000] <othermaciej>
I think at least some browsers let you override that to overflow: scroll

[01:57:01.0000] <hsivonen>
http://ie.microsoft.com/testdrive/Graphics/21FallingBalls/Default.xhtml

[01:57:02.0000] <othermaciej>
on the theory that the SVG spec didn't know what it was talking about

[01:57:03.0000] <othermaciej>
hsivonen: something different happens in IE?

[01:57:04.0000] <hsivonen>
othermaciej: that's sad. overflow: hidden !important sabotages one of the coolest things one could do with inline SVG

[01:58:00.0000] <othermaciej>
hsivonen: we have that in our UA stylesheet but I am tempted to get rid of it

[01:58:01.0000] <hsivonen>
othermaciej: in IE9, the balls that fall out of the box don't get clipped out

[01:58:02.0000] <zcorpan>
image-fit:cover (or content-fit or what it's called now) should be able to make svg paint outside its box

[01:59:00.0000] <othermaciej>
oh I guess we don't have that any more

[01:59:01.0000] <othermaciej>
just overflow: hidden

[01:59:02.0000] <othermaciej>
the demo could choose to set overflow: visible on the <svg>

[01:59:03.0000] <hsivonen>
is !important supposed to work in the style attribute?

[01:59:04.0000] <hsivonen>
overflow: visible !important; doesn't work for me in <svg style=

[01:59:05.0000] <hsivonen>
in Gecko

[02:00:00.0000] <zcorpan>
hsivonen: !important is supposed to work in style="" but it would only affect later declarations in the same style=""

[02:00:01.0000] <othermaciej>
we don't actually have a !important rule in WebKit trunk

[02:00:02.0000] <hsivonen>
looks like I won't be making the demo I wanted to make

[02:01:00.0000] <zcorpan>
hsivonen: since style="" is author level and highest specificity, it will override anything that doesn't have !important anyway, and if something earlier has !important then nothing will override it

[02:01:01.0000] <hsivonen>
zcorpan: is image-fit implemented in any browser?

[02:01:02.0000] <zcorpan>
hsivonen: it's implemented in opera but not shipped yet

[02:01:03.0000] <othermaciej>
see the comment here for a long-winded explanation of the clipping issue: http://trac.webkit.org/browser/trunk/WebCore/css/svg.css

[02:01:04.0000] <hsivonen>
plain overflow: visible; doesn't work in Gecko, either.

[02:02:00.0000] <zcorpan>
hsivonen: i hope gecko and webkit will implement it as well since it's really cool

[02:03:00.0000] <othermaciej>
WebKit still clips if I force overflow: visible

[02:03:01.0000] <othermaciej>
not sure why

[02:03:02.0000] <hsivonen>
when I moved from Mac to Linux, I lost the assistive tech for writing namespace URIs :-(

[02:04:00.0000] <hsivonen>
and now I need to write those in order to test SVG in WebKit :-(

[02:04:01.0000] <zcorpan>
hsivonen: just memorize them

[02:05:00.0000] <othermaciej>
svg root element seems to clip to its bounds regardless

[02:05:01.0000] <othermaciej>
I could probably figure out why with some code study but I don't have time at the moment

[02:06:00.0000] <zcorpan>
it's really easy, the leading part is invariant, the year is 1999 for xhtml and xlink, 2000 for svg, and the trailing part is xhtml, svg or xlink (no trailing slash for any of them)

[02:06:01.0000] <zcorpan>
the mathml and xml namespaces are harder

[02:06:02.0000] <othermaciej>
I think our rule in svg.css is probably based on an incorrect interpretation of "outermost"

[02:07:00.0000] <othermaciej>
it should really be a rule for svg svg rather than for svg:not(:root) I think

[02:07:01.0000] <othermaciej>
not that this would fix the IE demo but it would make the stylesheet match the comment

[02:08:00.0000] <othermaciej>
hsivonen: what was it that assisted you with the namespace URI before?

[02:08:01.0000] <hsivonen>
othermaciej: Typinator

[02:10:00.0000] <othermaciej>
that seems like just the sort of product I would expect to exist in some form on Linux

[02:10:01.0000] <othermaciej>
though perhaps in a form with thousands of configuration options, most completely useless

[02:33:00.0000] <hsivonen>
reading the SVG 1.1 spec for the <animate> element is rather annoying

[02:38:00.0000] <hsivonen>
I need to find a way to demo MathML that is cool but not verbose as source

[02:39:00.0000] <annevk>
guess you can't use e=mc^2 as you don't need MathML for that?

[02:42:00.0000] <webben>
how about the uncertainty principle: http://en.wikipedia.org/wiki/Uncertainty_principle

[02:42:01.0000] <hsivonen>
annevk: right. I want something with at least an integral, fraction or a big square root in it

[02:42:02.0000] <hsivonen>
preferrably something that people recognize

[02:43:00.0000] <hsivonen>
maybe the solution for the 2nd degree equation

[02:43:01.0000] <jgraham>
I was going to suggest that

[02:44:00.0000] <jgraham>
Not many integral equations that are well known among the general population

[02:44:01.0000] <hsivonen>
I wanted to use Gaussian blur and put also SVG Gaussian blur in the demo, but Gaussian blur as MathML is awfully verbose

[02:45:00.0000] <jgraham>
Everything is awfully verbose as MathML

[02:47:00.0000] <roc>
the fact that overflow:visible doesn't work on inline <svg> is just our bug

[02:47:01.0000] <zcorpan>
hsivonen: http://www.gstatic.com/knol/static/images/cartoons/42111-29252.jpg

[02:52:00.0000] <hsivonen>
roc: OK.

[02:52:01.0000] <Dashiva>
Good one, zcorpan

[02:53:00.0000] <roc>
annevk: so if we were to create a new feature to give authors control over who can load their (font) resources, how would it be different from CORS in practice?

[02:54:00.0000] <othermaciej>
roc: apparently a popular bug

[03:10:00.0000] <gsnedders>
/me wonders if it's at all disturbing taht the first diary entry that Hixie appears in in my diary in years is written while he was intoxinated

[03:10:01.0000] <gsnedders>
(I'm sorry, but any Avenue Q reference makes me think of Hixie now…)

[03:11:00.0000] <Dashiva>
Who replies four times to the same tweet...

[03:13:00.0000] <othermaciej>
why does Avenue Q make you think of Hixie?

[03:13:01.0000] <hsivonen>
Dashiva: someone with more than 140 characters to say?

[03:13:02.0000] <asmodai>
hsivonen: https://bugzilla.mozilla.org/show_bug.cgi?id=561714

[03:13:03.0000] <asmodai>
hsivonen: created the bug report for that leak

[03:13:04.0000] <annevk>
roc, it would work for <img>, <video>, etc. too

[03:14:00.0000] <annevk>
roc, instead of a font-specific hack of CORS

[03:14:01.0000] <asmodai>
Heh, Avenue Q, now I have to think of the Internet is for pr0n.

[03:14:02.0000] <annevk>
roc, which imo is silly

[03:14:03.0000] <hsivonen>
asmodai: thanks

[03:14:04.0000] <asmodai>
hsivonen: ideas for tracking it down more than welcome :)

[03:15:00.0000] <jgraham>
othermaciej: You have met Hixie, right?

[03:15:01.0000] <othermaciej>
jgraham: yes...?

[03:15:02.0000] <gsnedders>
othermaciej: Because he never shuts up about it?

[03:15:03.0000] <othermaciej>
I have also seen Avenue Q

[03:15:04.0000] <othermaciej>
he has never mentioned it to me

[03:16:00.0000] <gsnedders>
Are you not in #webapps

[03:16:01.0000] <jgraham>
This othermaciej is clearly an imposter

[03:16:02.0000] <gsnedders>
*?

[03:16:03.0000] <othermaciej>
apparently not

[03:16:04.0000] <othermaciej>
this server or w3c?

[03:16:05.0000] <gsnedders>
w3c

[03:16:06.0000] <jgraham>
We need to find the realothermaciej

[03:16:07.0000] <gsnedders>
That's where he makes the majority of references I've seen online

[03:16:08.0000] <othermaciej>
haven't been, just added to my autojoin

[03:17:00.0000] <othermaciej>
I can't recall him mentioning it on this channel or any time I have met him IRL

[03:17:01.0000] <gsnedders>
I've certainly seen it on this channel, on #webapps, and IRL

[03:19:00.0000] <jgraham>
and on his blog

[03:21:00.0000] <annevk>
http://www.google.com/search?q=avenue+q+site:krijnhoetmer.nl

[03:21:01.0000] <annevk>
proof

[03:21:02.0000] <gsnedders>
But yes, that's what Hixie gets for people singing The Internet is for Porn while I'm just-about-managing to go home

[03:22:00.0000] <hsivonen>
annevk: looks like there are more instances of gsnedders mentioning Avenue Q than Hixie mentioning it

[03:23:00.0000] <gsnedders>
fail.

[03:23:01.0000] <jgraham>
OTOH Hixie has seen Avenue Q >= 3 times whereas gsnedders has seen it once

[03:23:02.0000] <annevk>
therefore Hixie ranks higher than gsnedders

[03:30:00.0000] <zcorpan>
if Hixie and gsnedders were in the same Document, we could run the outline algorithm on them

[03:32:00.0000] <roc>
annevk: my point of view is that CORS lifts same-origin restrictions. I don't see how when those restrictions are imposed "for privacy" CORS is suitable but when those restrictions are imposed for other reasons, lifting them with CORS is an abuse of CORS

[03:32:01.0000] <annevk>
because CORS cannot be used in a generic way for those other purposes

[03:33:00.0000] <othermaciej>
why can't it?

[03:33:01.0000] <annevk>
and CORS is intended to be a protocol that works q

[03:33:02.0000] <annevk>
equally for all resources

[03:33:03.0000] <annevk>
othermaciej, because most scenarios already allow cross-origin loads

[03:33:04.0000] <roc>
that's because we currently have no way to impose same-origin restrictions for <img> loads etc

[03:33:05.0000] <othermaciej>
couldn't CORS be used in principle to allow cross-origin images or videos to draw into a canvas without tainting?

[03:34:00.0000] <othermaciej>
annevk: yeah, but CORS isn't what allows you to do the load, it's what allows you to read the data

[03:34:01.0000] <roc>
if we did have a way to impose such restrictions, we could use CORS to lift it

[03:34:02.0000] <annevk>
othermaciej, that's not another purpose, that's CORS' purpose

[03:34:03.0000] <annevk>
othermaciej, preventing bandwidth theft through CORS just for fonts is what is being proposed here

[03:34:04.0000] <othermaciej>
I don't say it was another purpose

[03:34:05.0000] <roc>
the problem <img> is that no restrictions exist for CORS to apply to

[03:34:06.0000] <annevk>
othermaciej, but that cannot be used for <video> or <img> in that way

[03:35:00.0000] <annevk>
othermaciej, so it is completely tied to font loading

[03:35:01.0000] <annevk>
and just makes CORS some kind of weird case there

[03:35:02.0000] <annevk>
which imo is wrong

[03:35:03.0000] <roc>
it's not a weird case

[03:35:04.0000] <roc>
CORS can be used to lift same-origin restrictions where they exist

[03:35:05.0000] <othermaciej>
if you wanted to limit <video> or <img> to same-origin-only for bandwidth protection purposes, you'd have to check Origin or Referer on the server side

[03:35:06.0000] <roc>
they just happen to not exist at all for <video> and <img>

[03:35:07.0000] <othermaciej>
I don't think the intent of applying same-origin policy to fonts is for bandwidth limits

[03:35:08.0000] <hsivonen>
annevk: is this about a matter of principle, a technical problem or fear of a lawyer thinking that CORS is Effective?

[03:35:09.0000] <othermaciej>
I think the intent is for half-assed pseudo-DRM

[03:36:00.0000] <othermaciej>
at least I imagine that is why the font vendors want it

[03:36:01.0000] <annevk>
hsivonen, technical consistency?

[03:36:02.0000] <annevk>
hsivonen, haven't really tried to think about lawyers and DRM that much

[03:36:03.0000] <othermaciej>
if font had a same-origin restriction then it would be reasonable for CORS to lift it

[03:37:00.0000] <hsivonen>
annevk: seems technically consistent to use CORS to lift the Same Origin restriction whenever it is lifted for any reason

[03:37:01.0000] <othermaciej>
I think the controversial part is whether fonts should have a same-origin restriction by default

[03:37:02.0000] <annevk>
othermaciej, there's no reason for fonts to have same-origin restrictions

[03:37:03.0000] <othermaciej>
I don't think anyone in WebKit-land has any interest in adding that restriction

[03:37:04.0000] <roc>
right, that is the issue

[03:38:00.0000] <roc>
CORS is a red herring

[03:38:01.0000] <othermaciej>
but if the Trident and Gecko developers want to add such a restriction, then it seems reasonable to use CORS to selectively lift that restriction

[03:38:02.0000] <roc>
thank you!

[03:38:03.0000] <annevk>
hsivonen, we shouldn't have same-origin restrictions however for whatever reason

[03:39:00.0000] <annevk>
hsivonen, if the problem is bandwidth theft then that problem also exists for <video>

[03:39:01.0000] <annevk>
hsivonen, so the solution cannot be CORS imo

[03:39:02.0000] <othermaciej>
I would be unhappy if whatever spec people are talking about required same-origin restrictions by default for fonts though

[03:39:03.0000] <roc>
CORS is not the solution to bandwidth theft

[03:39:04.0000] <annevk>
indeed it's not

[03:39:05.0000] <roc>
the solution for bandwidth theft and certain other problems is a same-origin restriction

[03:40:00.0000] <othermaciej>
WebKit-based browsers would probably never be conforming if that was the case

[03:40:01.0000] <roc>
CORS lets authors opt out of that

[03:40:02.0000] <annevk>
roc, that solution fails for a lot of scenarios so it's not a good solution

[03:41:00.0000] <roc>
we can argue about that, but the fact remains that if we're going to have a same-origin restriction, it seems reasonable to use CORS to lift that restriction

[03:41:01.0000] <jgraham>
Why isn't video same-origin only?

[03:42:00.0000] <roc>
there was a big discussion about that

[03:42:01.0000] <annevk>
because it doesn't make sense

[03:42:02.0000] <othermaciej>
it would suck to not be able to embed video cross-site

[03:42:03.0000] <roc>
that

[03:42:04.0000] <annevk>
at least not with how <script>, <img>, etc. work

[03:42:05.0000] <jgraham>
Yeah I remember a discussion

[03:42:06.0000] <roc>
plus legacy reasons

[03:42:07.0000] <othermaciej>
given that a key feature of the most popular video sites is embedding cross-site

[03:42:08.0000] <othermaciej>
also, plugins can do it and <video> shouldn't be less capable than plugins

[03:42:09.0000] <annevk>
anyway i need to get some food/sleep and make some more slides

[03:42:10.0000] <othermaciej>
(e.g. QuickTime)

[03:42:11.0000] <jgraham>
Yes, but if CORS is a solution for fonts it is a solution for video

[03:43:00.0000] <jgraham>
but moreso given the realtive bandwidth requirments

[03:43:01.0000] <jgraham>
*relative

[03:43:02.0000] <othermaciej>
the proposed solution for fonts isn't so much CORS as it is imposing a same-origin restriction by default

[03:43:03.0000] <othermaciej>
CORS just mitigates some of the downsides of that basic decision

[03:43:04.0000] <roc>
right

[03:44:00.0000] <othermaciej>
imposing a same-origin restriction by default wasn't really a viable option for video

[03:44:01.0000] <othermaciej>
it seems less obviously broken for fonts, though I am not super enthusiastic about it myself

[03:45:00.0000] <jgraham>
So the argument goes that fonts are fundamentally different from video because video is often cross origin whereas fonts are typically same origin?

[03:45:01.0000] <roc>
I think it's reasonable to provide a convenient mechanism allowing authors to control who accesses their resources

[03:45:02.0000] <othermaciej>
I think it's reasonable too

[03:46:00.0000] <othermaciej>
the questions are: (a) should the default be no cross-origin access instead of full access as for most other resource types?

[03:46:01.0000] <annevk>
they can just check the Origin header

[03:46:02.0000] <othermaciej>
and: (b) should this mechanism be enforced client-side instead of server-side?

[03:46:03.0000] <annevk>
or we could have a header that enforces same-origin loads

[03:46:04.0000] <roc>
checking the Origin header is not convenient

[03:46:05.0000] <othermaciej>
if you answer "no" to both, then nothing in particular needs to be added

[03:46:06.0000] <annevk>
that would also work for all other types of resources

[03:47:00.0000] <jgraham>
Would people have argued differently for video if CORS was already widely deployed?

[03:47:01.0000] <roc>
checking on the client has better privacy properties

[03:47:02.0000] <othermaciej>
using CORS to undo the restriction is about as convenient/inconvenient as checking Origin to add a restriction

[03:47:03.0000] <roc>
jgraham: it would have helped a bit, but probably we'd have the same outcome

[03:47:04.0000] <roc>
othermaciej: most authors don't need to undo the restriction

[03:48:00.0000] <othermaciej>
which goes back to question (a)

[03:48:01.0000] <annevk>
we could just have a header that enforces same origin loads

[03:48:02.0000] <jgraham>
roc: In that case the argument for fonts seems very weak

[03:48:03.0000] <annevk>
that could even be part of the CSRF work from Mozilla

[03:48:04.0000] <annevk>
a header which people likely need to use anyway

[03:48:05.0000] <jgraham>
s/fonts/making fonts same-origin by default/

[03:49:00.0000] <roc>
annevk: a header on the document referencing the font?

[03:49:01.0000] <annevk>
no on the font of course

[03:49:02.0000] <roc>
ok

[03:49:03.0000] <roc>
that's a question of what the default should be

[03:49:04.0000] <annevk>
it's also a question of whether we want to solve the bandwidth problem two times or just once...

[03:49:05.0000] <roc>
based on what the font vendors say, it sounds like they're going to require authors to control access

[03:50:00.0000] <annevk>
and whether a different request policy for fonts really makes sense -- i think it doesn't

[03:50:01.0000] <roc>
most of them

[03:50:02.0000] <roc>
so authors will frequently want to same-origin restriction

[03:50:03.0000] <annevk>
i personally think this whole WOFF thing is somewhat of a mistake

[03:51:00.0000] <annevk>
before it's done Ubuntu and other systems will support it natively, I'm sure

[03:51:01.0000] <annevk>
or simple web-based tools will be available

[03:51:02.0000] <roc>
sure

[03:51:03.0000] <roc>
that's all fine

[03:51:04.0000] <othermaciej>
I think the vendor's legal requirement for a same-origin restriction should be handled between the vendor and their customer

[03:51:05.0000] <roc>
othermaciej: it is

[03:51:06.0000] <othermaciej>
no need to bring the browser into it

[03:51:07.0000] <annevk>
anyway, gotta go, will read the logs later

[03:52:00.0000] <roc>
the question is how easy is it going to be for the customer to follow their side of the license?

[03:52:01.0000] <roc>
we can give them tools to make it super easy

[03:53:00.0000] <othermaciej>
or we can create a business opportunity for Adobe Straming Font Server :-)

[03:53:01.0000] <roc>
or we can force them to do server-side checking, with a number of downsides for convenience, privacy and robustness (firewalls stripping headers)

[03:54:00.0000] <zcorpan>
so should we add a header Restrict-Load-To: http://example.org

[03:54:01.0000] <zcorpan>
which would work for everything (fonts, images, videos, html)

[03:54:02.0000] <othermaciej>
server-side checking would be better if the motive was actually preventing bandwidth waste

[03:55:00.0000] <jgraham>
Embedding laws and licensing agreements in code is a bit scary. Apart from anything else laws and licenses change but required behaviour on the web tends to live on

[03:55:01.0000] <othermaciej>
since by the time the client checks, your bandwidth has already been wasted

[03:55:02.0000] <roc>
"embedding laws and licensing agreements in code" is totally not happening here ... that's the DRM bogeyman

[03:56:00.0000] <roc>
othermaciej: the point is that site authors aren't going to link to a font cross-site if that load isn't actually going to work

[03:56:01.0000] <roc>
unless they just want to maliciously eat your bandwidth to no useful purpose

[03:56:02.0000] <othermaciej>
roc: really, authors aren't going to do strange things with no visible effect and then leave that in their production site?

[03:56:03.0000] <jgraham>
Well you are justifying inconsistencies in the platform because they help people to conform to licensing agreements circa 2010

[03:56:04.0000] <jgraham>
+typical

[03:57:00.0000] <othermaciej>
that would make my job a lot easier if it was true!

[03:58:00.0000] <roc>
othermaciej: ok, there'll be a little of that

[03:58:01.0000] <othermaciej>
you are correct that it would reduce the motive for bandwidth waste somewhat, though not the opportunity

[03:58:02.0000] <roc>
jgraham: we have same-origin restrictions for XHRs. I don't see having them for fonts as a wild innovation

[03:59:00.0000] <jgraham>
roc: For XHR is is needed for security

[03:59:01.0000] <roc>
sure, so?

[03:59:02.0000] <jgraham>
So that is a clear reason for the difference

[03:59:03.0000] <roc>
it's still true that we have different classes of loads

[04:00:00.0000] <jgraham>
The font thing seems to be predicated on the idea that fonts are somehow fundamentally different from an IP point of view compared to images or video

[04:00:01.0000] <jgraham>
or text

[04:00:02.0000] <roc>
there are a different set of tradeoffs

[04:00:03.0000] <hsivonen>
Restrict-Load-To would have bad effects if Opera supported it before IE

[04:00:04.0000] <roc>
images are totally constrained by legacy issues

[04:01:00.0000] <roc>
video is constrained by legacy issues and the "viral video" feature

[04:01:01.0000] <jgraham>
Right, if we were designing the platform from scratch we would do it differently

[04:02:00.0000] <hsivonen>
but server refusing to serve if the request didn't have the right Origin header would have better game theoretica characteristics to motivate UA implementation

[04:02:01.0000] <roc>
if those issues didn't exist, would it be a good idea to have a default-same origin restriction for images and video? I think so, although that world's so counter-factual I'd have to think harder than I have

[04:02:02.0000] <othermaciej>
it does seem to me like many font deployments won't really want to share cross-origin, on the other hand, I could imagine big sites wanting to serve their fonts off of Akamai

[04:03:00.0000] <othermaciej>
or I could imagine the existence of central type hosting sites like TypeKit

[04:03:01.0000] <othermaciej>
or of sites that have a separate domain for large static resources in general, or which have multiple subdomains

[04:03:02.0000] <roc>
I can tell you that in practice when we raised the idea of same-origin restrictions for <video> there were howls of protest from authors and, well, just about everyone

[04:03:03.0000] <hsivonen>
typekit's core feature is addressing complications

[04:03:04.0000] <roc>
whereas for fonts, we have had no protest whatsoever

[04:04:00.0000] <jgraham>
roc: Because people aren't expecting Youtube for fonts

[04:04:01.0000] <hsivonen>
the more complications there are, the more attractive typekit is

[04:04:02.0000] <roc>
jgraham: well, there you go

[04:04:03.0000] <othermaciej>
those are the things that would get harder in exchange for making other things easier

[04:04:04.0000] <jgraham>
roc: It doesn't mean it won't happen 

[04:05:00.0000] <roc>
well, we're waiting

[04:05:01.0000] <jgraham>
It's just that people's frame of reference is different

[04:05:02.0000] <othermaciej>
making a decent-quality font is hard

[04:05:03.0000] <roc>
it's been a couple of years now

[04:05:04.0000] <Philip`>
jgraham: About the IP point of view, lots of commercially-produced video makes use of secure Flash streaming etc to prevent it being used in unauthorised ways, but there's also a lot of free/amateur video like most stuff on YouTube where people are happy to use <video>, so there's still a substantial use case for the latter

[04:05:05.0000] <othermaciej>
there are lots of freely redistributable fonts out there, mostly they are crappy

[04:05:06.0000] <othermaciej>
as with video, the most popular way to distribute fonts currently is via Flash

[04:05:07.0000] <Philip`>
jgraham: and there's lots of commercially-produced fonts which want similar protection, but very few decent free/amateur fonts, so the focus is much more heavily on the former

[04:06:00.0000] <Philip`>
jgraham: so in both cases there's two categories of difference importances and different technological solutions for each

[04:06:01.0000] <gsnedders>
zcorpan: We'd also have to both be header elements

[04:06:02.0000] <roc>
one of the nice things about making same-origin restriction the default is that cross-origin links tend to have many sites linking to a few repositories instead of a few sites linking to many repositories

[04:07:00.0000] <zcorpan>
gsnedders: yes. aren't you?

[04:07:01.0000] <gsnedders>
zcorpan: I feel more like generic flow content.

[04:08:00.0000] <roc>
so a default same-origin restriction means you need to add Access-Control-Allow-Origin:* to a few repositories rather than add some other header to a lot of sites

[04:08:01.0000] <zcorpan>
gsnedders: any specific element you feel identifies you?

[04:09:00.0000] <gsnedders>
zcorpan: the div element.

[04:09:01.0000] <gsnedders>
I feel divisive, like that.

[04:09:02.0000] <zcorpan>
wow that's like the most boring element

[04:09:03.0000] <othermaciej>
wait, when did this become a game of "who is which element"?

[04:10:00.0000] <gsnedders>
othermaciej: When we got bored of writting assembly to generate C++ :)

[04:10:01.0000] <gsnedders>
*writing

[04:10:02.0000] <othermaciej>
how could you ever get bored of that?

[04:10:03.0000] <gsnedders>
I dunno. I think Ollie gets bored enough of doing it the sane way round.

[04:10:04.0000] <zcorpan>
it's like that if you're a div element

[04:11:00.0000] <Philip`>
gsnedders: I suggest writing assembly code that generates C++ code that (when compiled and executed) generates the original assembly code

[04:11:01.0000] <jgraham>
"This is not a game of who the fuck are you"

[04:12:00.0000] <zcorpan>
"knock knock" "who's there?" "fuck you"

[04:12:01.0000] <gsnedders>
jgraham's bullying me into saying things on IRC.

[04:13:00.0000] <gsnedders>
I am not going to give into bullying again, after having multiple people force-feed me drink on Saturday

[04:15:00.0000] <jgraham>
At least one of them was force feeding you water

[04:16:00.0000] <jgraham>
/me feels he should give the impression that not everyone gsnedders knows is entirely irresponsible

[04:16:01.0000] <jgraham>
(because that would reflect badly on gsnedders' ability to chose friends)

[04:17:00.0000] <gsnedders>
jgraham: Actually, I was asking him to get me more water, multiple times

[04:17:01.0000] <gsnedders>
jgraham: He was just kind enough to give me it and not give me vodka instead.

[04:19:00.0000] <gsnedders>
jgraham: Still, I'll refrain from mentioning for your sake your girlfriend dragging me off to dance with her with both of us only just standing up

[04:19:01.0000] <gsnedders>
Oh, sorry.

[04:19:02.0000] <jgraham>
/me refrains from mentioning that that's not really true

[04:20:00.0000] <gsnedders>
Oh, it is…

[04:21:00.0000] <jgraham>
No really you were both standing up very well. Her slightly moreso

[04:21:01.0000] <gsnedders>
Yes, her moreso, certainly.

[04:21:02.0000] <gsnedders>
But the last time I didn't dance long, as moving was challenging.

[04:21:03.0000] <gsnedders>
Standing still I could manage.

[04:22:00.0000] <jgraham>
Actually you were making more of a fuss about standing still. Whenever you actually walked anywhere you were quite fine

[04:23:00.0000] <gsnedders>
I almost fell over several times walking home

[04:24:00.0000] <jgraham>
Yeah but I do that when I'm sober

[04:24:01.0000] <gsnedders>
Or maybe you're just permanently stoned from all the medication you're on.

[04:24:02.0000] <jgraham>
No

[04:25:00.0000] <gsnedders>
I guess I had enough beer to fulfill annevk's request that I should have some beer for him.

[04:49:00.0000] <gsnedders>
http://www.wolframalpha.com/input/?i=%280.5%2Ax%5E2%2By%5E2-1%29%5E3-x%5E2%2Ay%5E3%3D0

[05:55:00.0000] <annevk>
/me wonders why people are still using slideshare.net

[05:56:00.0000] <annevk>
surely there's a better way to show a few images

[05:56:01.0000] <annevk>
i.e. one that doesn't require me to install Flash

[05:56:02.0000] <annevk>
(which I won't)

[06:05:00.0000] <Philip`>
annevk: The better way to show slides is PDF

[06:06:00.0000] <Philip`>
/me generally dislikes HTML slides because the layout never looks quite right and they always have weird navigation controls

[06:07:00.0000] <webben>
I like using Slidy personally.

[06:07:01.0000] <webben>
I tried using Keynote and found it unusable.

[06:08:00.0000] <webben>
(trying to format code examples - especially horrible)

[08:11:00.0000] <remysharp>
Q: is there any time that the Web Messaging API would use postMessage without it going via an iframe?

[08:13:00.0000] <gsnedders>
remysharp: A window opened through window.open, maybe?

[08:13:01.0000] <remysharp>
gsnedders: cheers - that's what I was after.

[09:19:00.0000] <annevk>
remysharp-away, also Web Workers

[10:24:00.0000] <magcius>
Is this a bug? The scrollbar appears in http://magcius.mecheye.net/canvasbug.html I looked into it, and a linebreak textnode is somehow being added, causing the html tag's scrollHeight to be 4px larger. It happens in all major browsers, so that's why I'm asking here.

[10:42:00.0000] <divya>
magcius: use display:block on canvas in your css. 

[10:42:01.0000] <magcius>
divya: it's not display: block; already?

[10:42:02.0000] <divya>
magcius: Don't think it is so by default. 

[10:43:00.0000] <Philip`>
magcius: Use html{line-height:0}

[10:44:00.0000] <magcius>
Philip`: ah. I used html {margin-bottom: -1em;}

[10:45:00.0000] <remysharp>
Q: Web Workers - the spec suggests that a worker has access to Web SQL databases

[10:45:01.0000] <remysharp>
but none of the browsers that have workers and databases seemed to have implemented this

[10:45:02.0000] <remysharp>
is this something that /should/ be in a worker, or am I missing something?

[10:46:00.0000] <ap>
remysharp: this is being actively worked on for WebKit

[10:46:01.0000] <remysharp>
ap: cheers, so it's just something that's pending implementation

[10:47:00.0000] <Philip`>
magcius: Looks like either canvas{display:block} or html{line-height:0} are the right solutions, since <canvas> is normally display:inline the same as <img> and so it has the layout of a line of text

[10:47:01.0000] <magcius>
Philip`: alright

[12:12:00.0000] <Dashiva>
Surely whether you're writing a polyglot document or not should be handled by editor configuration

[12:25:00.0000] <Matjas_>
Which ARIA roles should be used when faking a HTML5 <details> element?

[12:35:00.0000] <webben>
Matjas_: Hmm. Does it require ARIA?

[12:36:00.0000] <Matjas_>
webben: Once <details> is natively supported, I don't think so, but for now (while we're still stuck using fallbacks) I'd say some ARIA wouldn't hurt

[12:37:00.0000] <webben>
Matjas_: When you say faking it - do you mean using the HTML5 markup and adding behavior with JS?

[12:37:01.0000] <Matjas_>
webben: Yes. Like this: http://mathiasbynens.be/notes/html5-details-jquery (also see Shelley Powers’s comment, suggesting to add ARIA: http://mathiasbynens.be/notes/html5-details-jquery#comment-6)

[12:38:00.0000] <divya>
Matjas_: Slide 65 on this: http://www.slideshare.net/AaronGustafson/why-aria-devchatt-2010 seems to discuss ARIA for tabbed interfaces, I am guessing you could investigate some of the roles covered there. 

[12:39:00.0000] <Matjas_>
divya: Thanks!

[12:39:01.0000] <webben>
But it's not tabbed?

[12:39:02.0000] <webben>
Matjas_: Or is it tabbed?

[12:40:00.0000] <Matjas_>
webben: Depends what you mean by ‘tabbed’

[12:40:01.0000] <Matjas_>
webben: There’s a demo here: http://mathiasbynens.be/demo/html5-details-jquery What do you think?

[12:40:02.0000] <webben>
Matjas_: Mmm I wouldn't really say that's a list of tabs.

[12:41:00.0000] <webben>
I don't think there are ARIA roles that map perfectly onto details and summary.

[12:41:01.0000] <Matjas_>
Agreed

[12:41:02.0000] <Matjas_>
Perhaps aria-hidden="false" could be used

[12:41:03.0000] <Matjas_>
Question mark?

[12:42:00.0000] <webben>
Matjas_: In your JS implementation, should you not move focus to the start of the content when you open the details?

[12:42:01.0000] <webben>
Matjas_: I agree aria-hidden="true" should be applied to the contents in the closed state.

[12:42:02.0000] <divya>
Matjas_: http://www.w3.org/TR/2009/WD-wai-aria-implementation-20091215/#mapping_role see the table for the huge list of roles you can choose from :/

[12:43:00.0000] <Matjas_>
webben: Why? I’m focusing the <summary> element so that when tab is pressed, it jumps to the next 'tabbable' element

[12:43:01.0000] <webben>
slightly newer at http://www.w3.org/TR/wai-aria/roles#role_definitions

[12:43:02.0000] <webben>
Matjas_: Let's say I'm using a screen reader. I activate the summary element ... somehow (we can come back to that), how do I know what's changed?

[12:44:00.0000] <Matjas_>
webben: Hmm, I see your point

[12:44:01.0000] <Matjas_>
divya: Thanks again!

[12:46:00.0000] <webben>
Matjas_: With regards to activation, I guess give the summary element a role of "button" and a property of "aria-pressed".

[12:46:01.0000] <webben>
Matjas_: Though given you're faking the lot with JS, you could just as well insert an actual button element.

[12:46:02.0000] <webben>
and it will work with more UAs

[12:47:00.0000] <webben>
Matjas_: you could use aria-owns/aria-controls to indicate the relationship between the button and the content.

[12:48:00.0000] <Matjas_>
webben: I'm trying to keep DOM manipulations to a minimum. JS makes the summary element clickable, regardless of wether the browser understands it correctly. Would it really be beneficial to replace it with an actual button?

[12:48:01.0000] <divya>
webben: Would the role of "note" be not sorta similar to what <details> semantically stands for? http://www.w3.org/TR/wai-aria/roles#note

[12:49:00.0000] <webben>
divya: Not necessarily.

[12:49:01.0000] <webben>
I think that's a bit closer to <aside> really.

[12:49:02.0000] <AryehGregor>
Well, what are the default ARIA semantics of <details> in HTML5?

[12:49:03.0000] <webben>
Matjas_: It would be beneficial in the sense that there's plenty of deployed software which can't make head or tail of ARIA or HTML5.

[12:50:00.0000] <divya>
"The details element represents a disclosure widget from which the user can obtain additional information or controls." says http://www.whatwg.org/specs/web-apps/current-work/multipage/interactive-elements.html#the-details-element

[12:51:00.0000] <Matjas_>
webben: Yeah, but that's what my JS is for. Or are you talking about the case where JS is not available?

[12:51:01.0000] <webben>
Matjas_: Let's say I focus the summary in a pre-ARIA UA+screenreader. Adding role=button won't make my screenreader read it as a button.

[12:52:00.0000] <Matjas_>
webben: I see

[12:52:01.0000] <webben>
Matjas_: Thinking about your actual example, I wonder if it's best represented as some sort of ARIA tree - that doesn't apply to all instances of details though.

[12:53:00.0000] <webben>
Matjas_: In fact, I'm pretty sure your particular example is  precisely a tree

[12:53:01.0000] <webben>
http://www.w3.org/WAI/PF/aria/roles#tree

[12:53:02.0000] <webben>
"A type of list that may contain sub-level nested groups that can be collapsed and expanded."

[12:53:03.0000] <webben>
Matjas_: http://www.w3.org/WAI/PF/aria/states_and_properties#aria-expanded might be generalizable for use outside of trees too.

[12:54:00.0000] <webben>
hmm section supports aria-expanded for example: http://www.w3.org/WAI/PF/aria/roles#section

[12:55:00.0000] <webben>
AryehGregor: Can't see any at http://www.whatwg.org/specs/web-apps/current-work/multipage/content-models.html#annotations-for-assistive-technology-products-(aria) 

[12:55:01.0000] <webben>
(assuming that's the place to look)

[12:56:00.0000] <AryehGregor>
Then my guess is ARIA doesn't have anything with equivalent semantics.

[12:56:01.0000] <Matjas_>
Should aria-expanded be used as an attribute to the details element, or to the summary element?

[12:56:02.0000] <AryehGregor>
If it does, then the spec is buggy, I assume.

[12:57:00.0000] <webben>
Matjas_: That's not clear to me from http://www.w3.org/WAI/PF/aria/states_and_properties#aria-expanded .

[12:57:01.0000] <Matjas_>
webben: Yeah, me neither :p

[12:57:02.0000] <webben>
Matjas_: http://www.w3.org/WAI/PF/aria-primer/#exampletree should be helpful.

[12:58:00.0000] <webben>
ah that answers the question in fact

[12:58:01.0000] <webben>
<li role="treeitem" aria-expanded="false">Cats

[12:58:02.0000] <webben>
Matjas_: should be details that has aria-expanded

[12:58:03.0000] <Matjas_>
webben: Thanks

[13:02:00.0000] <webben>
Matjas_: http://pastie.org/936045 ... maybe

[13:02:01.0000] <webben>
Matjas_: What worries me about that is I don't see anything to make it clear that pressing the button will expand the section.

[13:08:00.0000] <webben>
Matjas_: Another approach might be to look at native desktop UI implementations of details-like widgets and see what accessibility roles and properties they use, then map those back to aria.

[13:12:00.0000] <webben>
/me can't find an example on OS X.

[13:18:00.0000] <webben>
ah ... Details disclosure in Password prompt is an example.

[13:19:00.0000] <webben>
It's interesting - the actual control in that case is the disclosure triangle - which is either collapsed or expanded.

[13:21:00.0000] <Matjas_>
webben: Isn’t http://img.riddle.pl/_disclosure-osx-20100425-212835.png a good example of details-like usage as well?

[13:24:00.0000] <webben>
Matjas_: I think that's a tree example.

[13:24:01.0000] <webben>
Matjas_: (In classic HTML5 that would have been a datagrid I think)

[13:25:00.0000] <webben>
Matjas_: But could be implemented as a list of details elements I guess, then annotated as an ARIA tree.

[13:25:01.0000] <webben>
Matjas_: http://developer.apple.com/mac/library/documentation/UserExperience/Conceptual/AppleHIGuidelines/XHIGControls/XHIGControls.html "Disclosure Triangles"

[13:25:02.0000] <webben>
is the control I'm thinking of

[13:26:00.0000] <webben>
Looks like OS X has a distinct Accessibility Role for the little triangle AXDisclosureTriangle.

[13:26:01.0000] <webben>
Possibly "Disclosure Buttons" is also details-like

[13:27:00.0000] <webben>
fwiw based on the human guidelines, AXDisclosureTriangles are used in trees ("hierarchical list[s]")

[13:29:00.0000] <webben>
http://tinyurl.com/3yyl5gh : spec for AXDisclosureTriangle

[13:33:00.0000] <webben>
Matjas_: Inspecting the Finder with the Accessibility Inspector, those arrows are also Disclosure Triangles exposing AXDisclosureTriangle.

[13:34:00.0000] <Matjas_>
webben: Wow, you should really reply to riddle’s comment: http://mathiasbynens.be/notes/html5-details-jquery#comment-1

[13:35:00.0000] <webben>
Similar HIG from MS: http://msdn.microsoft.com/en-us/library/aa511487.aspx

[13:37:00.0000] <zcorpan>
TabAtkins: yo. had you heard anything from ms on their <video> fallback behavior?

[13:39:00.0000] <webben>
http://msdn.microsoft.com/en-us/library/ee671276(v=VS.85).aspx ExpandCollapse pattern - I guess that's the UI Automation equivalent


2010-04-27
[17:22:00.0000] <Hixie>
anyone registered on the doom9 forums?

[17:24:00.0000] <Hixie>
i want to contact zuggy, but the forum requires that i have been a member for 5 days

[17:25:00.0000] <kinetik>
Hixie: i've got an account there

[17:27:00.0000] <Hixie>
can you send a privmsg to zuggy saying that we're considering using SRT as the basis for a subtitle format for the web and that ian⊙hc would like to discuss the matter with him?

[17:27:01.0000] <Hixie>
i'd love his opinions on some ideas i have

[17:27:02.0000] <Hixie>
e.g. whether he would rather we didn't define things, or if he wants to be involved, or if he's happy for us to fork it, or whatever

[17:29:00.0000] <kinetik>
Hixie: this one? http://forum.doom9.org/member.php?u=11440

[17:29:01.0000] <Hixie>
yeah

[17:29:02.0000] <kinetik>
Last Activity: 16th April 2007 16:21

[17:29:03.0000] <Hixie>
hm

[17:30:00.0000] <Hixie>
bummer

[17:30:01.0000] <kinetik>
I can try sending a message anyway, if you like.

[17:30:02.0000] <Hixie>
sure

[17:30:03.0000] <Hixie>
try sending it to http://forum.doom9.org/member.php?u=41898 also, he's been active more recently

[17:31:00.0000] <Hixie>
he's the other subrip dev

[17:31:01.0000] <Hixie>
as far as i can tell

[17:31:02.0000] <Hixie>
hmm, it does kind of look like SubRip is dead

[17:32:00.0000] <Hixie>
i wonder who else we should contact

[17:32:01.0000] <Hixie>
kinetik: thanks, btw

[17:32:02.0000] <Hixie>
much appreciated

[17:38:00.0000] <kinetik>
Hixie: sent something to both of them

[17:39:00.0000] <Hixie>
thanks!

[17:48:00.0000] <Dashiva>
/me wonders if some other Jonas will find himself with a bunch of enemies in the morning

[17:54:00.0000] <Philip`>
Dashiva: Maybe the Jonas Brothers?

[18:10:00.0000] <JonathanNeal>
heyo

[18:59:00.0000] <nessy>
good thinking, Hixie about the SubRip guys!

[19:21:00.0000] <Hixie>
how do i include the string "<br/>" in a wiki page without it becoming a line break? (AryehGregor?)

[19:23:00.0000] <Hixie>
i went with &lt;br/>

[19:23:01.0000] <Hixie>
seems to work

[19:23:02.0000] <othermaciej>
Hixie: &lt;br/>

[19:23:03.0000] <othermaciej>
you have to escape < for certain tags that the wiki would otherwise recognize and spit back verbatim

[21:50:00.0000] <Aleoss>
http://validator.w3.org/check?uri=http%3A%2F%2Faleoss.ath.cx%2Fcss%2F&charset=%28detect+automatically%29&doctype=Inline&ss=1&outline=1&group=0&No200=1&verbose=1&user-agent=W3C_Validator%2F1.767

[21:51:00.0000] <Aleoss>
Should I disregard the error there or should I be placing <nav> outside of the <pre>?

[00:02:00.0000] <zcorpan>
wait what? Why the url change?

[00:03:00.0000] <annevk2>
sicking

[00:03:01.0000] <annevk>
see public-webapps responseBlob subthread

[00:03:02.0000] <zcorpan>
pointer?

[00:03:03.0000] <annevk>
so ahead of you

[00:03:04.0000] <annevk>
:p

[00:06:00.0000] <zcorpan>
hrm

[00:07:00.0000] <zcorpan>
Hixie: can you change it back if i say opera has implemented URL and i've written test cases using URL? :)

[00:07:01.0000] <annevk>
WebKit also has EventSource.URL shipped I think

[00:07:02.0000] <annevk>
is that all?

[00:09:00.0000] <annevk>
hmm, my inbox has grown from ~400 to ~900

[00:35:00.0000] <zcorpan>
window.EventSource is undefined for me in chrome

[00:42:00.0000] <annevk>
hmm

[00:42:01.0000] <annevk>
indeed

[00:44:00.0000] <annevk>
so how can https://bugs.webkit.org/show_bug.cgi?id=14392#c7 be explained?

[00:45:00.0000] <annevk>
https://bugs.webkit.org/show_bug.cgi?id=14997 is marked fixed

[00:46:00.0000] <annevk>
maybe Chrome is based on a much older version of WebKit?

[00:46:01.0000] <annevk>
this was landed last August...

[00:47:00.0000] <othermaciej>
annevk: it's an optional feature - maybe they have the flag off

[00:47:01.0000] <zcorpan>
othermaciej: is EventSource shipped in safari?

[00:48:00.0000] <othermaciej>
zcorpan: it's not in Safari 4.0.5 afaict

[00:48:01.0000] <annevk>
ooh: https://bugs.webkit.org/show_bug.cgi?id=33695

[00:48:02.0000] <othermaciej>
or in trunk

[00:48:03.0000] <othermaciej>
I wonder if it's just disabled by default?

[00:48:04.0000] <annevk>
I saw Web developers talking about it... hmm

[00:48:05.0000] <othermaciej>
oh wait

[00:48:06.0000] <othermaciej>
it does exist in WebKit trunk

[00:49:00.0000] <othermaciej>
so it will quite likely be in the next Safari

[00:49:01.0000] <othermaciej>
good ol' [object EventSourceContructor]

[00:49:02.0000] <zcorpan>
it has an attribute called "URL", yes?

[00:51:00.0000] <othermaciej>
how do I construct an EventSource again?

[00:51:01.0000] <othermaciej>
or I guess I can check the IDL

[00:51:02.0000] <zcorpan>
you could check "URL" in EventSource.prototype

[00:51:03.0000] <annevk>
othermaciej, new EventSource("url")

[00:52:00.0000] <othermaciej>
        readonly attribute DOMString URL;

[00:52:01.0000] <othermaciej>
says the IDL

[00:52:02.0000] <zcorpan>
thanks

[00:52:03.0000] <zcorpan>
othermaciej: what's your opinion on the name change?

[00:52:04.0000] <annevk>
I guess that ought to be changed if we are going with sicking's URL -> url

[00:52:05.0000] <annevk>
grmbl

[00:52:06.0000] <zcorpan>
/me thinks it should be changed back to URL

[00:53:00.0000] <annevk>
he does have a point that Document.URL is somewhat archaic, but still

[00:53:01.0000] <zcorpan>
it's more used than most other attributes on HTMLDocument

[00:53:02.0000] <othermaciej>
well, it hasn't shipped yet apparently

[00:54:00.0000] <othermaciej>
is there consensus to change it?

[00:54:01.0000] <othermaciej>
would rather get it over with soon

[00:54:02.0000] <othermaciej>
are there instances of lowercase "url" in the Web platform?

[00:54:03.0000] <zcorpan>
i don't see consensus; sicking asked for it to be changed, Hixie changed it

[00:54:04.0000] <zcorpan>
i think it should be called URL

[00:55:00.0000] <zcorpan>
othermaciej: in css url()

[00:55:01.0000] <zcorpan>
othermaciej: <input type="url"> although that is case insensitive

[00:56:00.0000] <annevk>
css url() is even more case-insensitive

[00:56:01.0000] <othermaciej>
what about in APIs?./dom/Document.idl:        readonly attribute DOMString URL;

[00:56:02.0000] <zcorpan>
<meta http-equiv="refresh" content="5; URL=..."> although that's also case insensitive (mostly written in uppercase i think)

[00:56:03.0000] <othermaciej>
./websockets/WebSocket.idl:        readonly attribute DOMString URL;

[00:56:04.0000] <annevk>
yeah, both WebSocket and EventSource are affected

[00:56:05.0000] <othermaciej>
./html/HTMLAnchorElement.idl:        readonly attribute URL absoluteLinkURL; (is that a standard or some weird webkit extension?)

[00:57:00.0000] <jgraham>
Uppercase for an attribute name is kinda ugly

[00:57:01.0000] <jgraham>
Unless it is just a constant

[00:58:00.0000] <zcorpan>
othermaciej: seems like a weird extension

[00:58:01.0000] <zcorpan>
jgraham: sure it's ugly, but i'd rather have attributes with the same name use the same case everywhere

[00:59:00.0000] <othermaciej>
zcorpan: I think it might not be exposed to JS

[00:59:01.0000] <othermaciej>
yeah, it's ObjC-only

[00:59:02.0000] <othermaciej>
kinda weird, as it seems to duplicate href

[00:59:03.0000] <othermaciej>
maybe it dates back to when our <a>.href was buggy

[01:01:00.0000] <zcorpan>
/me sends his email asking for it to be changed back

[01:02:00.0000] <annevk>
absoluteLinkURL sounds like WebKit-only

[01:02:01.0000] <annevk>
but I've thought that before :/

[01:02:02.0000] <othermaciej>
it's only in our Objective-C bindings so not relevant to the Web

[01:02:03.0000] <othermaciej>
has Chrome shipped WebSocket yet?

[01:03:00.0000] <zcorpan>
othermaciej: yeah, but it doesn't seem to have .URL

[01:03:01.0000] <othermaciej>
weird

[01:03:02.0000] <othermaciej>
is that a new addition?

[01:03:03.0000] <othermaciej>
oh, if you are checking based on WebSocket.prototype I expect that won't work

[01:03:04.0000] <othermaciej>
at least, not if V8 bindings match how JavaScriptCore does it

[01:04:00.0000] <zcorpan>
oh wait it does have URL

[01:04:01.0000] <othermaciej>
we don't put DOM getters and setters on the prototype they are magical

[01:04:02.0000] <zcorpan>
yeah i was checking the prototype first

[01:05:00.0000] <zcorpan>
that seems like a bug, they should be on the prototype, right?

[01:08:00.0000] <othermaciej>
it doesn't match what Mozilla does

[01:08:01.0000] <othermaciej>
hopefully Web IDL will decide which way is correct

[01:09:00.0000] <othermaciej>
Moz behavior is potentially more convenient but on the other hand it may be less efficient

[01:09:01.0000] <zcorpan>
is someone working on webidl?

[01:09:02.0000] <othermaciej>
weinig is working on it a little, ever so slowly

[01:09:03.0000] <othermaciej>
I need to nag him again

[01:09:04.0000] <zcorpan>
ok

[01:09:05.0000] <othermaciej>
now that the WebKit2 code drop is finished

[01:10:00.0000] <othermaciej>
he did do a couple of edits - trying to get it to be based on ES5 instead of ES3

[01:10:01.0000] <zcorpan>
i'd like webidl to decide about what to do with too few arguments to constructors and methods

[01:12:00.0000] <othermaciej>
that would be good too

[01:46:00.0000] <asmodai>
hsivonen: wow, might have found the page that was causing issues

[02:06:00.0000] <asmodai>
It seems http://site.icu-project.org/ is causing those leaks for me. At least the very fast increasing leaks. (I still see some minor leakage, but I also occasionally see the garbage collection kick in).

[02:35:00.0000] <hsivonen>
asmodai: btw, are you running 3.6.x with html5.enable=true?

[02:37:00.0000] <hsivonen>
hmm. are all the scripts on that page just for analytics?

[02:39:00.0000] <asmodai>
hsivonen: Yes I am

[02:39:01.0000] <asmodai>
Ah, should mention that I guess

[02:39:02.0000] <hsivonen>
asmodai: I recommend not doing that.

[02:40:00.0000] <hsivonen>
asmodai: if you want to run with html5.enable=true, it's much better to use a trunk nightly

[02:40:01.0000] <hsivonen>
and if you want to run 3.6.x, it's much better not to run with html5.enable=true

[02:40:02.0000] <asmodai>
How so?

[02:41:00.0000] <hsivonen>
asmodai: I'm sorry I didn't try more forcefully to get html5.enable removed from the 3.6 branch

[02:41:01.0000] <asmodai>
hsivonen: Hahaha

[02:41:02.0000] <hsivonen>
asmodai: the snapshot of the HTML5 parser on the 3.6 branch is from June last year just when it was deemed worthy to *start* testing it

[02:41:03.0000] <asmodai>
no no, I'm happy with it to test the MathML :P

[02:42:00.0000] <hsivonen>
after that, all the problems that have been found and fixed are unfixed on the 3.6 branch

[02:42:01.0000] <asmodai>
:(

[02:43:00.0000] <hsivonen>
asmodai: could you, please, test if the leak goes away if you disable the HTML5 parser?

[02:43:01.0000] <asmodai>
sure, gimme a moment

[02:43:02.0000] <asmodai>
hsivonen: do I need to restart the entire browser for that toggle to go through?

[02:44:00.0000] <asmodai>
or would simply reloading that page with the toggle off be enough?

[02:45:00.0000] <hsivonen>
asmodai: reload is enough

[02:45:01.0000] <asmodai>
ok

[02:45:02.0000] <asmodai>
on an aside

[02:46:00.0000] <asmodai>
annevk: saw http://www.internetconsultatie.nl/nrfimplementatie  ?

[02:46:01.0000] <asmodai>
Dutch gov asking for input on the cookie restriction law(s)

[02:47:00.0000] <hsivonen>
asmodai: do the laws cover browser fingerprinting?

[02:47:01.0000] <asmodai>
hsivonen: mmm, good question

[02:48:00.0000] <asmodai>
For all I know this is that follow through on the EU directive on banning third party cookies from auto accept.

[02:48:01.0000] <asmodai>
ok, opened that site with html5 enable to true and I see the memory increase as expected.

[02:49:00.0000] <asmodai>
Now turning it off and reloading

[02:51:00.0000] <asmodai>
hsivonen: Seems to stabilize and I can even observe the garbage collection kick in... hovering around 576 MB now.

[02:52:00.0000] <asmodai>
hsivonen: clear cut case of icu-project.org and html5.enable=true on 3.6.4 not jiving well?

[02:57:00.0000] <annevk>
asmodai, looking

[02:58:00.0000] <annevk>
asmodai, hmm, I can't tell whether it affects sites or browsers

[02:58:01.0000] <annevk>
asmodai, if sites I suppose that makes sense

[02:58:02.0000] <annevk>
asmodai, as long as they clearly scope it in some way I guess

[02:59:00.0000] <asmodai>
annevk: it's a continuation of http://tweakers.net/nieuws/66906/cookies-mogen-volgend-jaar-alleen-na-toestemming-internetter.html

[03:01:00.0000] <asmodai>
Directive 2009/136/EC generally requires end-user consent to the storing of information-such as the piece of software known as a "cookie"-on the user's computer. Accessing stored cookie data-the raison d'être of a cookie-would also require consent. Further, a user's consent would be valid only if he or she is "provided . . . clear and comprehensive information" concerning the cookie's use.

[03:02:00.0000] <asmodai>
So it would mean changes to browser defaults methinks.

[03:02:01.0000] <jgraham>
Not possible

[03:02:02.0000] <annevk>
we tried something like that

[03:02:03.0000] <jgraham>
Breaks the web

[03:02:04.0000] <annevk>
indeed

[03:02:05.0000] <annevk>
also, users don't understand cookies

[03:04:00.0000] <asmodai>
hey, I don't invent it ;)

[03:05:00.0000] <annevk>
shoot the messenger, no?

[03:06:00.0000] <asmodai>
/me is shot

[03:09:00.0000] <asmodai>
hsivonen: Updated the ticket https://bugzilla.mozilla.org/show_bug.cgi?id=561714 with the information.

[07:26:00.0000] <mut>
hey, I have an image on canvas, and i want to save it to the server, any way of doing that?

[07:27:00.0000] <annevk>
xhr.send(canvas.toDataURL())

[07:27:01.0000] <annevk>
or you could serialize an ImageData object somehow and transmit that

[07:27:02.0000] <mut>
ok

[07:27:03.0000] <mut>
ill take a look at that

[07:28:00.0000] <mut>
thanks

[07:28:01.0000] <mut>
errrm

[07:28:02.0000] <mut>
yea thanks :)

[07:37:00.0000] <annevk>
"Facebook uses Decentralized Extensibility to centralize the Web around Facebook." heh, brilliant quote

[07:48:00.0000] <jgraham>
I want that on my wall

[07:57:00.0000] <hsivonen>
annevk: it seems it's my most retweeted tweet so far

[07:58:00.0000] <Lachy>
Philip`, yt?

[08:00:00.0000] <Philip`>
Lachy: Sort of

[08:01:00.0000] <Lachy>
Philip`, just wanted to check with you about whether or not this test is correct, since you know more about canvas than I do. http://tc.labs.opera.com/html/canvas/getContext/2d/drawImage/003.htm

[08:01:01.0000] <Lachy>
I don't get why it's expecting an exception to be thrown for that one

[08:03:00.0000] <hsivonen>
jgraham: speaking of walls: did the idea of asking Randall Munroe to draw the HTML5 tokenizer state transition graph go anywhere?

[08:03:01.0000] <Philip`>
Lachy: Looks like the image's 'complete' attribute would be false, which used to make drawImage throw an error but should now cause it to be silently ignored

[08:03:02.0000] <Philip`>
(so the test is currently incorrect)

[08:04:00.0000] <Lachy>
(new Image()).complete returns true in every browser though

[08:04:01.0000] <Lachy>
so I guess all browsers are wrong about that

[08:04:02.0000] <Philip`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/embedded-content-1.html#dom-img-complete indicates it should be false, I believe

[08:05:00.0000] <Philip`>
(I don't know if the spec is correct)

[08:05:01.0000] <Lachy>
ok. So that's a bug in opera for image.complete and an invalid test.

[08:05:02.0000] <Lachy>
the spec doesn't agree with implementations, but it makes more sense than saying an image is complete, even though it has no src

[08:06:00.0000] <Philip`>
You could interpret it instead as saying the image downloading/decoding is complete

[08:07:00.0000] <Philip`>
and an image with no src isn't still in the middle of the downloading/decoding process, so that process is not incomplete

[08:07:01.0000] <Lachy>
but it hasn't actually completed anything

[08:07:02.0000] <Lachy>
and that means setting a src attribute should make set it to false

[08:08:00.0000] <Lachy>
until it loaded the image again

[08:08:01.0000] <Philip`>
It's completed the "complete = false; if (there is a src) { ... do a long slow download and decode ... }; complete = true" process

[08:09:00.0000] <Philip`>
but then you have to deal with images that are complete but still don't have any image data

[08:09:01.0000] <Philip`>
I don't think it's particularly illogical either way, but the spec's current definition seems more helpful

[08:12:00.0000] <Lachy>
I guess that makes the spec ambiguous if it can be interpreted either way.

[08:12:01.0000] <Lachy>
I'll send mail about that later.

[08:19:00.0000] <Philip`>
Oh, I don't think the spec is ambiguous

[08:20:00.0000] <Philip`>
I just think the attribute name "complete" is ambiguous so it could defined in different ways that wouldn't seem particularly wrong

[08:20:01.0000] <Philip`>
s/could/could be/

[08:21:00.0000] <Philip`>
and the current spec has chosen one definition which seems reasonably clear (though maybe people could argue it should make a different choice to better match current implementations)

[08:29:00.0000] <JonathanNeal>
mornin'

[08:37:00.0000] <jgraham>
hsivonen: No

[09:13:00.0000] <annevk>
bit of a shame that everyone implements the new sectioning elements without implementing the corresponding h1 styling rules

[09:14:00.0000] <annevk>
you'd think at least someone in either Gecko or WebKit would have picked up on that

[09:15:00.0000] <JonathanNeal>
Whatcha guys think of this? http://sandbox.thewikies.com/html-templates/ Suggetions, criticism?

[09:17:00.0000] <annevk>
guess I'll add a comment to https://bugzilla.mozilla.org/show_bug.cgi?id=559284 tomorrow or so when I find my credentials

[09:17:01.0000] <annevk>
time to read/sleep

[09:35:00.0000] <Lachy>
I wonder if it's too late to start rolling out those default styles for h1 elements, given the early adopters.

[09:36:00.0000] <Lachy>
hopefully not, since most sites will apply font size styles in author stylesheets anyway

[09:38:00.0000] <JonathanNeal>
Which styles are those?

[09:39:00.0000] <Lachy>
h1 is supposed to have it's font size adjusted based on the level it represents according to the sectioning element nesting

[09:40:00.0000] <JonathanNeal>
How did you guys end up using selectors for that? :section(1) or something?

[09:40:01.0000] <Lachy>
we don't have selectors like that for it yet.

[09:41:00.0000] <Lachy>
but there have been some ideas for a :heading(n) selector that will match a heading based on it's level

[09:41:01.0000] <JonathanNeal>
Right, I remember reading about that in here.

[09:42:00.0000] <Lachy>
but Mozilla could now implement that using their experimental :-moz-any() selector

[09:43:00.0000] <JonathanNeal>
Got it.  Which browser were you going to roll this into?  Cause some folks would want to customize that if they could.

[09:43:01.0000] <Lachy>
well, it's in the spec and will hopefully be introduced into all browsers eventually. I don't know of any browser that has any immediate plans.

[09:44:00.0000] <Lachy>
but as annevk just said above, Mozilla has just implemented some support for the new elements, without doing anything about these heading styles

[09:45:00.0000] <Lachy>
see http://www.whatwg.org/specs/web-apps/current-work/complete.html#margins-and-padding for more info

[09:46:00.0000] <Lachy>
so, according to that, Mozilla should implement those styles using :-moz-any(article, aside, section, nav) in place of the selector "x" given in the spec

[09:47:00.0000] <JonathanNeal>
Neat, maybe I should add that into http://www.iecss.com/whatwg.css

[09:49:00.0000] <Lachy>
is that stylesheet based on the styles given in the spec?

[09:52:00.0000] <TabAtkins>
JonathanNeal: In your HTML5+Microdata template, putting fn and org on the <img> will *never* be correct.

[09:52:01.0000] <TabAtkins>
Unless your company's name really is "images/site-graphic.png".

[09:53:00.0000] <TabAtkins>
Also somewhat confused about the hcard being duplicated for the page title block.

[09:53:01.0000] <JonathanNeal>
TabAtkins, I thought we were gonna petition to have it read the alt :-)

[09:53:02.0000] <TabAtkins>
Yes, I still am, but @itemprop *on* an <img> is always going to read the @src.

[09:55:00.0000] <Lachy>
JonathanNeal, here are all the heading styles as they could be implemented in Mozilla http://software.hixie.ch/utilities/js/live-dom-viewer/saved/463

[09:56:00.0000] <Lachy>
Open that in Minefield to see the intended result

[09:56:01.0000] <Lachy>
won't work in any release version of Firefox, as :-moz-any() was just added recently

[10:00:00.0000] <JonathanNeal>
neato Lachy. Who has since quit.

[10:08:00.0000] <JonathanNeal>
TabAtkins, so I would put the fn org logo on the anchor or h1? 

[10:15:00.0000] <TabAtkins_>
JonathanNeal: The <h1>, because putting it on an <a> has a special meaning too.

[10:18:00.0000] <TabAtkins_>
If @itemprop appears on an <img>, <video>, <audio>, <link>, <object>, <a>, or <time>, the value is drawn from an attribute.  <meta> too, but that should be obvious.  Any other tags, it grabs the text contents instead.

[10:18:01.0000] <TabAtkins_>
Basically anything with a significant url attribute, plus <time> and <meta>.

[10:20:00.0000] <JonathanNeal_>
I'll move it to the H1 right now on all of them.

[10:23:00.0000] <JonathanNeal_>
Okay, updated.

[10:27:00.0000] <TabAtkins_>
JonathanNeal_: Keep logo where it was.  That's *supposed* to point at an image url.

[10:28:00.0000] <TabAtkins_>
Also, I'm still confused about why you have a second hcard on the page-title block.

[10:42:00.0000] <JonathanNeal_>
Just trying to show an example where you have two hcards

[10:43:00.0000] <JonathanNeal_>
Would it be better / more clear if I killed the page one?

[10:43:01.0000] <TabAtkins>
While that's a valid goal, I think it results in something nonsensical here.

[10:43:02.0000] <TabAtkins>
Yeah.

[10:44:00.0000] <JonathanNeal_>
Okay, updated with both of your requests.

[10:46:00.0000] <TabAtkins>
All right, only one problem left.  ^_^  The itemscope has to be an ancestor to all its properties, so you'll have to move @itemscope, @itemref, and @itemtype all up to the hgroup.

[10:46:01.0000] <TabAtkins>
Or, move the @itemprop into a wrapper element somewhere inside of the <h1>.

[10:48:00.0000] <TabAtkins>
Seems fine otherwise; the only thing I can criticize is the use of extraneous <div>s.

[10:51:00.0000] <JonathanNeal_>
Which <div>s don't you like?

[10:51:01.0000] <AryehGregor>
How big is a typical starting size for TCP windows?  I haven't been able to find any real-world estimates in some brief Googling.

[10:51:02.0000] <JonathanNeal_>
itemscope has to be an ancestor or a child?  Which am I moving?

[10:52:00.0000] <TabAtkins>
#main-document, and at least one of #content or #main-content.

[10:53:00.0000] <JonathanNeal_>
main-content is useful for the skip to main content link.

[10:53:01.0000] <TabAtkins>
@itemscope declares that something is a Microdata container.  It must be an ancestor.  So either move @itemscope up (along with @itemtype and @itemref, which must be alonside @itemscope) or move @itemprop down onto a wrapper element (since there is no appropriate child element to carry those @itemprops right now).

[10:53:02.0000] <JonathanNeal_>
main-document is the document wrapper, and describes its contents.

[10:53:03.0000] <TabAtkins>
<body> is the document wrapper.

[10:54:00.0000] <JonathanNeal_>
This where I usually bring up a chat bar or a dockbar that might exist aside the main portion of the document, and may be application content.

[10:55:00.0000] <TabAtkins>
Ah, okay.  That's not in the template, so it looked extraneous.

[11:00:00.0000] <JonathanNeal_>
TabAtkins, okay I've updated.

[11:00:01.0000] <JonathanNeal_>
And added some comments to clarify the main document.

[11:01:00.0000] <TabAtkins>
You forgot to move the @itemref too.

[11:01:01.0000] <TabAtkins>
Otherwise looks fine.

[11:02:00.0000] <JonathanNeal_>
Doh! I'm sorry.  I'm in blankets at home with a bad cold and I'm being slow, my bad.

[11:02:01.0000] <TabAtkins>
np

[11:03:00.0000] <JonathanNeal_>
Should be all good now

[11:59:00.0000] <AryehGregor>
/me finally concludes that the offer-to-translate feature of Chrome is almost completely worthless, and disables it

[11:59:01.0000] <TabAtkins>
Really?  I've found it useful every time it's popped up.

[11:59:02.0000] <AryehGregor>
I've never found it useful.

[12:00:00.0000] <AryehGregor>
It pops up in Gmail, duplicating the in-Gmail popup, except the browser-level one doesn't go away when I navigate because it doesn't notice the navigation.

[12:00:01.0000] <AryehGregor>
It also pops up for Hebrew, when I can usually do a better job translating than it.  (Yes, I know I can disable just for Hebrew.)

[12:01:00.0000] <AryehGregor>
And it pops up in crazy places, claiming that an ASCII-only Apache directory listing is some crazy language or other.

[12:01:01.0000] <AryehGregor>
I've never actually had it pop up usefully, but it pops up uselessly all the time for me.

[12:02:00.0000] <TabAtkins>
Interesting.  I've never had it pop up crazily.

[12:12:00.0000] <Philip`>
AryehGregor: I thought Google were the people who keep saying that language detection heuristics are far more reliable than explicit metadata in pages

[12:13:00.0000] <Philip`>
Maybe the bar is very low

[12:13:01.0000] <AryehGregor>
Philip`, I'm sure that's true, but that doesn't mean it's reliable in an absolute sense.

[12:13:02.0000] <TabAtkins>
"Far more reliable" doesn't necessary mean "reliable".

[12:13:03.0000] <Aleoss>
http://aleoss.ath.cx/css/

[12:13:04.0000] <Aleoss>
:D

[12:13:05.0000] <Dashiva>
At least heuristics don't claim to be correct, only almost

[12:13:06.0000] <TabAtkins>
Aleoss: Link doesn't work.

[12:13:07.0000] <AryehGregor>
I've had Gmail detect e-mails as Finnish or something when they were just plain English with a PGP key in them.

[12:13:08.0000] <AryehGregor>
Maybe Finnish is a very weird language?  :)

[12:14:00.0000] <Aleoss>
TabAtkins: Oh? How so?

[12:14:01.0000] <TabAtkins>
Aleoss: I dunno, I just can't connect to ath.cx or aleoss.ath.cx

[12:14:02.0000] <Dashiva>
Maybe it reinterprets the bytes as some weird finnish encoding

[12:14:03.0000] <Aleoss>
TabAtkins: What is the error?

[12:14:04.0000] <TabAtkins>
AryehGregor: I'm going with "finnish is weird".  Or, finnish people actually sign their speech with PGP.

[12:14:05.0000] <AryehGregor>
A weird Finnish encoding that's not a superset of ASCII?

[12:14:06.0000] <AryehGregor>
TabAtkins, the link works for me.

[12:14:07.0000] <TabAtkins>
Aleoss: "Google Chrome could not connect to aleoss.ath.cx".

[12:15:00.0000] <AryehGregor>
Aleoss, did you see the last xkcd April Fool's joke?

[12:15:01.0000] <Dashiva>
They had Finnish EBCDIC and stuff, so sure

[12:15:02.0000] <Aleoss>
No I didn't Aryeh. What was it?

[12:15:03.0000] <Dashiva>
It was a non-cross-browser compatible web application

[12:16:00.0000] <Dashiva>
Quite disappointing!

[12:16:01.0000] <Aleoss>
:P

[12:16:02.0000] <TabAtkins>
Aleoss: My wife can't connect to that site from home either.  I guess it hasn't percolated to our cache yet in Cali?

[12:16:03.0000] <TabAtkins>
Is it a recent DNS change?

[12:16:04.0000] <AryehGregor>
Aleoss, http://uni.xkcd.com/

[12:16:05.0000] <Aleoss>
TabAtkins: Nope. It's been on dyndns.org for a few months now.

[12:17:00.0000] <AryehGregor>
. . . did you just call your wife and ask her to connect to a site to help debug this?

[12:17:01.0000] <AryehGregor>
Also, I had no idea you were married.

[12:17:02.0000] <TabAtkins>
No, I hit her up in google chat.

[12:17:03.0000] <TabAtkins>
And yes, my 5-year anniversary is this June.

[12:17:04.0000] <AryehGregor>
How old are you?

[12:17:05.0000] <TabAtkins>
24

[12:17:06.0000] <Dashiva>
gsnedders is now feeling the pressure

[12:18:00.0000] <AryehGregor>
Huh.

[12:18:01.0000] <TabAtkins>
Married my high-school sweetheart.  ^_^

[12:18:02.0000] <AryehGregor>
Is that the Rachel Atkins I've seen on Buzz?

[12:18:03.0000] <TabAtkins>
Yes.

[12:18:04.0000] <AryehGregor>
Well, now I know, then.

[12:18:05.0000] <TabAtkins>
At least, if you keep getting things about cooking from her.

[12:18:06.0000] <AryehGregor>
No, I've just seen her reply to one or two of your posts, I think.

[12:18:07.0000] <TabAtkins>
Ah, kk.  Yeah, still her.

[12:20:00.0000] <Aleoss>
Aryeh: What did you think of the source code of the document? ;)

[12:21:00.0000] <AryehGregor>
Aleoss, didn't look at it.

[12:21:01.0000] <Aleoss>
:O

[12:21:02.0000] <AryehGregor>
I admit that I do habitually use an actual web browser to view web pages, rather than curl.

[12:21:03.0000] <AryehGregor>
So, you know, it's usually not necessary for me to look at the source.

[12:21:04.0000] <Aleoss>
I use an actual web browser too.

[12:21:05.0000] <AryehGregor>
Why do you have a title on a stylesheet when there's no alternate stylesheet?

[12:22:00.0000] <Dashiva>
I thought everyone did HTML5 parsing in their heads

[12:22:01.0000] <AryehGregor>
Dashiva, I tried, but I can't remember algorithms that are more than 20 pages of pseudocode.

[12:22:02.0000] <Aleoss>
I put a title on it incase the browser puts it into a list of sorts, I've heard about before a browser having a list of all linking elements and if it has a title attribute it'll display it over top of it on mouse over.

[12:23:00.0000] <Dashiva>
Just convert to brain bytecode

[12:23:01.0000] <Aleoss>
I don't remember the name of the browser.

[12:23:02.0000] <AryehGregor>
Most of them do that, but only if there are alternate stylesheets.

[12:23:03.0000] <AryehGregor>
I think.

[12:24:00.0000] <Aleoss>
But it's really clean code and very explicit, eh? :P

[12:24:01.0000] <AryehGregor>
TabAtkins, isn't display: initial equivalent to display: inline?

[12:26:00.0000] <AryehGregor>
Aleoss, too much pointless markup.  Why not omit <head>, </head>, </body>, </html>?  Also, <meta charset="UTF-8"> instead of the http-equiv thing.  And I thought HTML5 banned <meta name="keywords">.

[12:26:01.0000] <Aleoss>
Aryeh: No. That meta tag is used for SEO.

[12:26:02.0000] <AryehGregor>
media="all" is always pointless.  type="" doesn't normally need to be specified anywhere.  The two titles in the head are pointless.

[12:26:03.0000] <AryehGregor>
Aleoss, um, all major search engines ignore <meta name="keywords">.

[12:27:00.0000] <TabAtkins>
Yahoo pays a little bit of attention to it.

[12:27:01.0000] <Aleoss>
And Description is also used for SEO.

[12:27:02.0000] <TabAtkins>
Description is a good meta, keep that.

[12:27:03.0000] <Aleoss>
And the Robots meta tag tells web bots what to do with the page when they come across it.

[12:27:04.0000] <AryehGregor>
rel="alternate" is misused on the links.http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#link-type-alternate

[12:28:00.0000] <AryehGregor>
Is pubdate actually used correctly there?  It seems wrong.

[12:28:01.0000] <AryehGregor>
/me is hatin'

[12:28:02.0000] <AryehGregor>
/me invites technical criticism of http://aryeh.name

[12:28:03.0000] <TabAtkins>
AryehGregor: Ah, right, display:initial is indeed equivalent to display:inline.

[12:29:00.0000] <TabAtkins>
What he really wants is a value that means "use whatever the cascaded value from the user-agent stylesheet is".

[12:29:01.0000] <AryehGregor>
Yes, I've wanted that before too.

[12:29:02.0000] <TabAtkins>
Though in his specific case, the correct advice is to just use "style=display:none" to hide it, and then remove the @style to show it.

[12:29:03.0000] <Aleoss>
Aryeh: Then what rel would it be?

[12:30:00.0000] <AryehGregor>
Aleoss, um, . . . probably none?

[12:30:01.0000] <AryehGregor>
rel is optional, you know.

[12:31:00.0000] <Aleoss>
archives looks appropriate.

[12:31:01.0000] <AryehGregor>
For which?

[12:31:02.0000] <Aleoss>
For all the links.

[12:31:03.0000] <AryehGregor>
Um, no.

[12:31:04.0000] <AryehGregor>
It's not appropriate at all.

[12:32:00.0000] <AryehGregor>
archives is for a history page of some kind.

[12:32:01.0000] <AryehGregor>
Like the archive of a blog.

[12:33:00.0000] <AryehGregor>
Speaking of which, is there any point in me adding rel="archives" to history links in MediaWiki?

[12:33:01.0000] <AryehGregor>
Does anything actually use it?

[12:33:02.0000] <AryehGregor>
I mean UA-wise, not author-wise.

[12:34:00.0000] <AryehGregor>
/me notes that there has been no technical criticism of the source code of http://aryeh.name, which implies it must be perfect

[12:34:01.0000] <Dashiva>
Well, I guess opera should show it in the info panel

[12:35:00.0000] <Dashiva>
*would

[12:35:01.0000] <AryehGregor>
/me wonders why he's using <header> when it only contains a single <h1> . . . oh, right, for styling, more appropriate than a <div>

[12:36:00.0000] <TabAtkins>
AryehGregor: I dislike your lack of indentation, and also your line-wrapping.

[12:36:01.0000] <AryehGregor>
What lack of indentation, and what line-wrapping?

[12:36:02.0000] <TabAtkins>
Your homepage source-code.

[12:37:00.0000] <AryehGregor>
Oh, in the actual source code.

[12:37:01.0000] <AryehGregor>
I do use indentation, where appropriate.

[12:37:02.0000] <TabAtkins>
You indented one section in the entire site.

[12:37:03.0000] <Aleoss>
Aryeh: Tell me what you think of the source code for: http://aleoss.ath.cx/?p=register

[12:38:00.0000] <AryehGregor>
I mostly indent only <li>s, it looks like.

[12:38:01.0000] <AryehGregor>
Also the indentation for a couple of <a>'s is totally insane.

[12:38:02.0000] <AryehGregor>
I got tired of telling vim not to do that.

[12:38:03.0000] <AryehGregor>
Line wrapping is useful because it means version control works reasonably well.

[12:38:04.0000] <AryehGregor>
Same reason as the HTML5 spec is line-wrapped, I assume.

[12:39:00.0000] <AryehGregor>
Also, vim doesn't behave well on long unwrapped lines, it's hard to navigate in them.

[12:39:01.0000] <AryehGregor>
Anyway, what else would you like me to indent?  I don't believe in gratuitous indentation of HTML.

[12:39:02.0000] <TabAtkins>
Bah, tooling issues shouldn't distract you from making your source beautiful.

[12:39:03.0000] <AryehGregor>
Makes things hard to read, the nesting is too much.

[12:40:00.0000] <Aleoss>
Heh. My document has 0 HTML indentation.

[12:40:01.0000] <TabAtkins>
Personally, indenting the contents of the <header> and <h1> would make it more readable to me.

[12:40:02.0000] <TabAtkins>
The 8ch tab used by view-source is indeed too fat, though.  A 4ch tab is more than readable.

[12:41:00.0000] <Aleoss>
Nah. For indenting, all you need is a space.

[12:41:01.0000] <AryehGregor>
URLs are really annoying if you're trying to keep to 80-char lines.

[12:41:02.0000] <Aleoss>
I have a good reason why.

[12:42:00.0000] <TabAtkins>
Tabs are long-accepted to be the indentation character.

[12:42:01.0000] <Aleoss>
http://aleoss.ath.cx/js/date-time.js

[12:42:02.0000] <Aleoss>
Imagine if that was 4 spaces per indent or worse... 8....

[12:42:03.0000] <TabAtkins>
If you have heavy indentation, you can adjust the width of your tab.  That's why tabs *exist* - they're variable-width.

[12:42:04.0000] <jcranmer>
no, that's essentially 2 spaces/indent

[12:43:00.0000] <jcranmer>
if () {\n<2sp>foo;\n}

[12:44:00.0000] <TabAtkins>
I do use a 2-space indent in emails, because typing a tab into a textarea is annoying, and 2 spaces are short enough to be easy to type and long enough to show indentation adequately in small examples.

[12:44:01.0000] <AryehGregor>
I shouldn't be using background-image: linear-gradient(#0AE, white); yet, should I?

[12:46:00.0000] <TabAtkins>
Personally, I think there is basically no chance of that changing between now and when Images goes Rec.

[12:46:01.0000] <AryehGregor>
Still bad form, though.

[12:46:02.0000] <TabAtkins>
Sure.

[12:46:03.0000] <AryehGregor>
And largely pointless.

[12:46:04.0000] <AryehGregor>
I'll remove it.

[12:46:05.0000] <TabAtkins>
But it's a nearly-guaranteed stable declaration, so shrug.

[12:47:00.0000] <AryehGregor>
Is anyone still talking about a way to officially freeze specific syntaxes before CR, so browsers can implement the unprefixed versions sooner?

[12:47:01.0000] <TabAtkins>
The answer seems to be just "if it becomes a problem, we'll kick the offenders to the next level and just CR what's stable".

[12:48:00.0000] <AryehGregor>
Blech.

[12:48:01.0000] <AryehGregor>
How inflexible.

[12:48:02.0000] <AryehGregor>
There are other advantages to faster spec progress, though.

[12:48:03.0000] <TabAtkins>
Shrug.  It's technically equivalent, and requires less mucking about with process.

[12:49:00.0000] <AryehGregor>
I don't think you'd need to have a lot of process, just some vendor saying "Hey, we're implementing this now, and we'd prefer to go with a standard prefix at this point, anyone object to freezing this?"

[12:49:01.0000] <AryehGregor>
Ideally, anyone else who already implemented it identical syntax with a vendor prefix could then push out support for the unprefixed version in a maintenance release, instead of waiting for the next full release.

[12:50:00.0000] <AryehGregor>
. . . well, as long as we don't get another border-radius, anyway.

[12:50:01.0000] <AryehGregor>
I mean: . . . well, as long as we don't get another border-radius, it's probably good.

[12:50:02.0000] <TabAtkins>
What's the issue with them saying the same thing, and us just kicking the unstable bits to the next level?

[12:50:03.0000] <AryehGregor>
You might have to have a lot of levels.

[12:50:04.0000] <TabAtkins>
And?

[12:50:05.0000] <AryehGregor>
Or a lot of very small specs.

[12:50:06.0000] <TabAtkins>
And?

[12:50:07.0000] <othermaciej>
what defines "stable"?

[12:50:08.0000] <TabAtkins>
Neither of these seem troublesome.

[12:50:09.0000] <othermaciej>
lots of very small specs is troublesome

[12:51:00.0000] <othermaciej>
if each one is about a property that interacts with ones from many other specs

[12:51:01.0000] <TabAtkins>
That's true.

[12:51:02.0000] <AryehGregor>
It just seems pointless to have to progress a lot of orthogonal features all or nothing.

[12:51:03.0000] <AryehGregor>
Of course, that's a problem with the W3C process generally, I guess.

[12:51:04.0000] <TabAtkins>
Yup.

[12:51:05.0000] <AryehGregor>
So skipping W3C process on this one issue won't really fix the problem.

[12:51:06.0000] <AryehGregor>
But good luck getting that fixed.

[12:52:00.0000] <Dashiva>
It's a feature, not a bug :P

[12:53:00.0000] <AryehGregor>
It's a buggy feature!

[12:53:01.0000] <TabAtkins>
Your got your bug in my feature!  No, you got your feature in my bug!  Wait a minute...

[12:55:00.0000] <AryehGregor>
IMO, we should never have two browsers shipping vendor prefixes for the same feature implemented interoperably, unless they were implemented at about the same time and we're still not sure the syntax is stable.

[12:55:01.0000] <AryehGregor>
By the time you have a second interoperable implementation and are pretty sure the syntax is good, you should be dropping the vendor prefix for that second implementation.

[12:55:02.0000] <AryehGregor>
I doubt that level of aggression in dropping prefixes would be compatible with dropping them only on entrance to CR.

[12:55:03.0000] <AryehGregor>
Am I wrong to want that?

[12:55:04.0000] <TabAtkins>
Not necessarily.

[12:56:00.0000] <AryehGregor>
Of course, -moz-border-radius, -webkit-border-radius, and border-radius are all slightly different.

[12:57:00.0000] <TabAtkins>
It's so nice that we have a canonical example of where early prefix-dropping would have screwed everything up.

[12:57:01.0000] <AryehGregor>
Well, I don't think border-radius ever met the standard I just suggested, so I'm safe on that front.

[12:58:00.0000] <TabAtkins>
Man, I have *really* got to buy something from Flyleaf.  I love their music too much.

[12:58:01.0000] <TabAtkins>
I think we may have come close to two interop implementations of old border-image, which would have been a problem.

[12:59:00.0000] <AryehGregor>
The only difference is with percentages and ellipses, it seems, no?

[12:59:01.0000] <AryehGregor>
With fixed circular units everything is the same?

[12:59:02.0000] <TabAtkins>
With border-radius?  Yeah.

[13:00:00.0000] <AryehGregor>
Which is the most common case by far, I'd think.

[13:00:01.0000] <AryehGregor>
But meh.

[13:00:02.0000] <TabAtkins>
Actually, it's just %s.  You just need %s to do ellipses properly.

[13:00:03.0000] <AryehGregor>
Ah.

[13:00:04.0000] <TabAtkins>
Rather, you need per-side %s, not %s always tied to width (which is FF's current impl).

[13:00:05.0000] <AryehGregor>
/me keeps reading %s as a printf()-style token indicating a string

[13:00:06.0000] <TabAtkins>
Heh.

[13:01:00.0000] <AryehGregor>
%'s would be fine.

[13:01:01.0000] <AryehGregor>
Or, and I know this is a radical idea, you could type the word out.

[13:01:02.0000] <TabAtkins>
It would be acceptable, yes, but unnecessary.

[13:01:03.0000] <TabAtkins>
And typing out "percentages" is just silly when I can omit 9 characters.

[13:02:00.0000] <AryehGregor>
/me is trying to think of some witty retort involving %u or something, but fails

[13:03:00.0000] <AryehGregor>
/me goes off to lunch/dinner, cryptography, and topology

[13:04:00.0000] <TabAtkins>
That sounds like an awesome evening.

[13:19:00.0000] <jgraham>
/me discovers that OSX keybindings for pgdown are too hard to use when your left hand is holding a g;ass of water to by your eye where you just accidentially got chilli

[13:20:00.0000] <TabAtkins>
Apparently the keybindings for ordinary typing are too hard to use, as well.

[13:48:00.0000] <jgraham>
TabAtkins: No, that's just normal. 

[13:58:00.0000] <volkmar>
Hixie: ping

[14:00:00.0000] <tantek>
choose your side: Unicorn vs Narwhal http://www.flickr.com/photos/tantek/4557645875/

[14:00:01.0000] <weinig>
WOW!

[14:01:00.0000] <TabAtkins>
tantek: That's... that's a real thing?

[14:01:01.0000] <TabAtkins>
Where did you find it?

[14:01:02.0000] <tantek>
Rhino Records in Claremont, CA

[14:01:03.0000] <tantek>
yes I'm holding it in the picture

[14:01:04.0000] <TabAtkins>
Can you... buy it for me?

[14:01:05.0000] <tantek>
you can see my arm to the lower left and my left foot below

[14:01:06.0000] <tantek>
I'm not in Claremont right now

[14:01:07.0000] <TabAtkins>
Darn.

[14:02:00.0000] <tantek>
my Flickr is a bit (~5 mo.) behind

[14:02:01.0000] <TabAtkins>
O

[14:02:02.0000] <TabAtkins>
I'll see if I can swing by there next time I head down to LA.

[14:03:00.0000] <tantek>
my guess is you could probably find it on one of those fancy search engine thingies

[14:04:00.0000] <tantek>
TabAtkins, here you go: http://tantek.com/asin/B000R85W1M

[14:05:00.0000] <jgraham>
Why the url shortner?

[14:05:01.0000] <tantek>
because Amazon has broken ASINs in the past

[14:06:00.0000] <tantek>
(I've collected about a half dozen broken ASINs so far)

[14:06:01.0000] <jgraham>
ASINs?

[14:06:02.0000] <tantek>
those IDs that start with B00...

[14:06:03.0000] <jgraham>
Oh

[14:06:04.0000] <tantek>
http://en.wikipedia.org/wiki/Amazon_Standard_Identification_Number

[14:07:00.0000] <jgraham>
Nothing to do with Madonna then

[15:02:00.0000] <aboodman>
quick question on the proposed Origin header.

[15:02:01.0000] <aboodman>
would it be sent with all requests, or only those from XHR?

[15:06:00.0000] <othermaciej>
aboodman: all requests, though some particular kinds of requests would always send "Origin: null"

[15:06:01.0000] <othermaciej>
abarth probably knows the details of what is proposed and what WebKit does currently

[15:16:00.0000] <aboodman>
othermaciej: so, eg, if a form was constructed and submitted cross-origin, the Origin header would be present.

[15:16:01.0000] <aboodman>
and a site that checked it would not be vulnerable to xsrf.

[15:28:00.0000] <paul_irish>
aboodman: offhand, do you know what was the first version of FF to support CORS? v3?

[15:28:01.0000] <aboodman>
paul_irish: nope

[15:28:02.0000] <aboodman>
i don't even know if any current versions support it

[15:42:00.0000] <othermaciej>
aboodman: yes

[15:42:01.0000] <othermaciej>
aboodman: more importantly, the same-origin form would have its own domain's Origin

[15:42:02.0000] <othermaciej>
aboodman: so tricky ways of stripping the origin would probably produce Origin: null rather than a valid value

[15:43:00.0000] <GarethAdams|Home>
What would happen if the initial value of a <textarea> or <input> was longer than that permitted by the maxlength?

[15:43:01.0000] <othermaciej>
aboodman: you may find this informative about some of the concepts: http://lists.w3.org/Archives/Public/public-webapps/2009OctDec/0468.html

[15:44:00.0000] <aboodman>
othermaciej: i read that already, thanks.

[15:44:01.0000] <aboodman>
(it was)

[15:44:02.0000] <aboodman>
i don't think it covered the detail of whether the Origin header was sent for normal forms, but maybe I missed it.

[15:57:00.0000] <othermaciej>
aboodman: it tried to explain how Origin can be a CSRF defense

[15:57:01.0000] <othermaciej>
(if I recall correctly)

[15:57:02.0000] <othermaciej>
it has to be sent for normal forms for that to work

[15:58:00.0000] <othermaciej>
yeah, it's illustrated in the diagram (CSRF: Origin Defense)


2010-04-28
[17:06:00.0000] <roc>
paul_irish: I believe CORS was first supported in Firefox 3.5. If not that, definitely 3.6

[17:08:00.0000] <paul_irish>
roger that. thx roc

[17:35:00.0000] <TabAtkins>
Why do I torture myself by continuing to use http://www.xanthir.com/pictures/llama.gif as my example picture when playing with images?

[18:46:00.0000] <divya>
I am curious why almost all CSS examples in whatwg has a @namespace defined? http://www.whatwg.org/specs/web-apps/current-work/complete.html#margins-and-padding

[18:46:01.0000] <annevk>
those are not examples

[18:47:00.0000] <annevk>
and it makes sure those styles only apply to HTML elements and not e.g. XML elements that happen to share the name

[18:47:01.0000] <divya>
oops. right. sorry. 

[18:53:00.0000] <JonathanNeal_>
What do you think of: "ARIA (Accessible Rich Internet Applications) is a collection of metadata added to or implied in HTML for the purpose of increasing the accessibility of content and components." ?

[18:54:00.0000] <JonathanNeal_>
And: "Microdata is a collection of metadata added to HTML for the purpose of conveying contact information, geographic coordinates, calendar events, and the like." ?

[19:06:00.0000] <divya>
JonathanNeal_: microdata defn sounds a lot like microformats one

[19:06:01.0000] <divya>
http://www.whatwg.org/specs/web-apps/current-work/multipage/microdata.html#introduction-2 is a bit more broad.

[19:06:02.0000] <JonathanNeal_>
Well, they are very similar, I botched together a bunch of web definitions to come up with a quick definition.

[19:07:00.0000] <JonathanNeal_>
I'll check it out right now, thanks

[20:33:00.0000] <annevk>
omg what a bunch of FUD on the fonts list

[20:34:00.0000] <annevk>
because font formats might not be safe they should be same-origin? give me a break

[20:41:00.0000] <othermaciej>
annevk: by "not safe" they mean cause security exploits like buffer overruns?

[20:41:01.0000] <othermaciej>
same-origin is clearly not even remotely a defense

[20:41:02.0000] <othermaciej>
against that threat model

[20:42:00.0000] <annevk>
read and weep: http://lists.w3.org/Archives/Public/public-webfonts-wg/2010Apr/0054.html

[20:42:01.0000] <othermaciej>
I wonder if Apple ended up joining the Fonts WG or not

[20:42:02.0000] <othermaciej>
if we did then I guess maybe I should join the list

[20:44:00.0000] <annevk>
you haven't

[20:44:01.0000] <annevk>
afaict

[20:45:00.0000] <annevk>
see also: http://www.w3.org/2000/09/dbwg/details?group=44556 (W3C Member-only)

[20:46:00.0000] <othermaciej>
it is true that fonts are a potential exploit vector, beyond what (non-SVG) images can do

[20:46:01.0000] <annevk>
I'm not at all disputing that

[20:46:02.0000] <othermaciej>
but same-origin is not an effective defense

[20:50:00.0000] <annevk>
and the claim elsewhere in the thread that fonts are more special than scripts is also utter crap

[20:50:01.0000] <annevk>
in http://lists.w3.org/Archives/Public/public-webfonts-wg/2010Apr/0050.html

[20:57:00.0000] <othermaciej>
it seems like a plausible argument that fonts are more widely reusable and less contextual

[20:58:00.0000] <othermaciej>
but that's not really a security argument

[21:00:00.0000] <othermaciej>
people also seem to have confused beliefs about what a same-origin restriction would do

[21:01:00.0000] <annevk>
oh, I didn't notice that

[21:01:01.0000] <othermaciej>
e.g. http://lists.w3.org/Archives/Public/public-webfonts-wg/2010Apr/0044.html

[21:01:02.0000] <othermaciej>
"My clients are interested in serving their custom corporate fonts in a way that does not expose them to use by other parties"

[21:03:00.0000] <annevk>
oh right, I was thinking he meant some kind of intranet situation

[21:03:01.0000] <annevk>
maybe I should do a reply after all

[21:04:00.0000] <othermaciej>
I don't think he means preventing external sites from using intranet-hosted fonts when viewed inside the firewall

[21:04:01.0000] <othermaciej>
I think he means preventing others from using a font that's published on a public site and protected by a same-origin restriction

[21:05:00.0000] <annevk>
yeah, so that's utter crap

[21:05:01.0000] <annevk>
wget!

[21:06:00.0000] <roc>
I think there is a significant difference between using wget to copy a font from someone else's server to your server, and serving it from your server

[21:06:01.0000] <roc>
vs putting a reference to someone else's server in your style sheet

[21:07:00.0000] <annevk>
othermaciej, btw, since you implemented <section> and all you should probably also implement the h1 styling rules from http://www.whatwg.org/specs/web-apps/current-work/complete.html#fonts-and-colors

[21:07:01.0000] <annevk>
othermaciej, they're after the initial block

[21:08:00.0000] <annevk>
roc, that's not quite the scenario he is sketching though

[21:08:01.0000] <annevk>
roc, also, I don't think anyone is against having a header for font files that says From-Origin: same or some such

[21:10:00.0000] <roc>
what would that do?

[21:11:00.0000] <annevk>
if that header is present and the UA will only use the resource if its same-origin with the Document it will be used for

[21:12:00.0000] <othermaciej>
annevk: yeah, we'll do that at some point

[21:12:01.0000] <othermaciej>
annevk: I guess either with a combinatorial explosion of rules, or we'll implement -webkit-any()

[21:12:02.0000] <annevk>
othermaciej, I'm afraid that if you wait sites will start depending on the default style rules

[21:13:00.0000] <othermaciej>
we also didn't implement the default margins yet

[21:13:01.0000] <othermaciej>
annevk: well, sites can use <section> just fine in browsers with 0 support

[21:14:00.0000] <annevk>
yeah, all the more reason to get the default style rules out there, but I see your point

[21:14:01.0000] <othermaciej>
our native support effectively just adds display: block, makes the interface HTMLElement instead of HTMLUnknownElement, and adds implicit <p> closing (and similar parsing effects)

[21:14:02.0000] <othermaciej>
oh, it looks like Hixie took away the default margins for sectioning elements

[21:15:00.0000] <roc>
annevk: I believe the number of sites that want same-origin checks will greatly exceed the number of sites that want to serve fonts cross-origin

[21:16:00.0000] <othermaciej>
anyway, I'll have to ask hyatt if adding those rules for h1 would cause any perf issues

[21:16:01.0000] <othermaciej>
I vaguely recall that rules with lots of indirect ancestor combinators can be slow

[21:16:02.0000] <othermaciej>
so it might be worthwhile to add a :-webkit-section-level() pseudo or the like

[21:17:00.0000] <annevk>
roc, I believe most sites won't care, much like most don't care about other resources

[21:17:01.0000] <roc>
I believe that too

[21:18:00.0000] <annevk>
creating special solutions for fonts seems backwards to me

[21:18:01.0000] <roc>
well, it depends on how widespread commercial fonts become on the Web and how much Web authors care about their license agreements

[21:19:00.0000] <roc>
imposing a same-origin restriction for fonts really isn't all that special. It's not like we invented same-origin restrictions.

[21:20:00.0000] <annevk>
well I think it is special

[21:21:00.0000] <roc>
ok

[21:21:01.0000] <annevk>
so far we only used same-origin restrictions if certain conditions apply

[21:21:02.0000] <annevk>
they don't apply to fonts

[23:28:00.0000] <JonathanNeal_>
heyo!

[23:30:00.0000] <alt-dot-net-geek>
hello

[01:04:00.0000] <annevk>
hmm, so fetch does not disable Referer if the origin passed in is a unique identifier

[01:04:01.0000] <annevk>
maybe I should ask for a Referer flag or something

[01:12:00.0000] <othermaciej>
it seems like Fetch should disable the referer in that case

[01:12:01.0000] <othermaciej>
er

[01:12:02.0000] <othermaciej>
wrong capitalization but you get the idea

[01:12:03.0000] <othermaciej>
it would never be right to send a unique origin referer

[01:13:00.0000] <annevk>
well, the Document in question could still have an origin

[01:14:00.0000] <annevk>
and an associated URL

[01:14:01.0000] <othermaciej>
annevk: so Tyler says anonymous requests (or whatever we call them) should discard Set-Cookie headers in the response

[01:14:02.0000] <othermaciej>
annevk: are you going to do that? I don't think it is an automatic side effect of unique identifier origin

[01:14:03.0000] <othermaciej>
he also mentioned "It would also need to prohibit the reuse of a TLS

[01:14:04.0000] <othermaciej>
client authenticated connection or NTLM authenticated connection. It

[01:14:05.0000] <othermaciej>
would also need to prevent use of cache entries populated by

[01:14:06.0000] <othermaciej>
non-uniform requests."

[01:14:07.0000] <annevk>
what I just checked in discards setting cookies when the credentials flag is false

[01:15:00.0000] <othermaciej>
I see he also mentioned Referer so maybe you are looking at his issues

[01:15:01.0000] <annevk>
so it's affected by withCredentials not by unique origins

[01:15:02.0000] <othermaciej>
I need to re-review the drafts for possible subset violations

[01:15:03.0000] <annevk>
but for the "new API" it comes down to the same thing

[01:15:04.0000] <othermaciej>
yeah it should be withCredentials based

[01:15:05.0000] <othermaciej>
that makes sense

[01:15:06.0000] <annevk>
(as the new API has credentials set to false and a unique origin)

[01:16:00.0000] <annevk>
I'm not quite sure what I should say exactly about TLS NTML and cache entries

[01:16:01.0000] <othermaciej>
did you settle on a name yet?

[01:16:02.0000] <othermaciej>
I apologize for splashing so much paint on the bikeshed

[01:16:03.0000] <annevk>
not really

[01:16:04.0000] <annevk>
guess I should read that thread with suggestions one more time

[01:16:05.0000] <othermaciej>
I think TLS and NTLM should be effects of withCredentials

[01:17:00.0000] <othermaciej>
and would be covered by a suitable definition of user credential

[01:17:01.0000] <othermaciej>
NTLM does per-connection auth, you can't really share a connection that's been NTLM-authenticated for no-credential requests

[01:18:00.0000] <annevk>
we could do ContextFreeRequest I suppose which inherits from XMLHttpRequest and is therefore on the same prototype chain and all

[01:18:01.0000] <othermaciej>
it's a bit of a computer-sciencey name

[01:18:02.0000] <othermaciej>
though not obviously worse than the others

[01:19:00.0000] <annevk>
so maybe I have TLS/NTML already: http://dev.w3.org/2006/waf/access-control/#credentials

[01:19:01.0000] <annevk>
though I should rename credentials to user credentials

[01:19:02.0000] <annevk>
UniformRequest makes no sense to me

[01:19:03.0000] <annevk>
GuestRequest is pretty good

[01:20:00.0000] <annevk>
incognito is too hard to spell

[01:22:00.0000] <hsivonen>
Is discarding cookies something that Caja can control if UMP is implemented on top of XHR?

[01:22:01.0000] <annevk>
no

[01:22:02.0000] <annevk>
XHR is cookie-ignorant

[01:22:03.0000] <annevk>
API-wise that is

[01:22:04.0000] <hsivonen>
does Set-Cookie in an XHR response set cookie in the browser HTTP stack?

[01:22:05.0000] <annevk>
yes

[01:23:00.0000] <annevk>
or just GuestXMLHttpRequest

[01:23:01.0000] <annevk>
I'm not too interested in this rhetorical context Mark went on about

[01:25:00.0000] <annevk>
so remaining issues

[01:25:01.0000] <annevk>
1) Referer handling -- should maybe be fixed in HTML5

[01:25:02.0000] <othermaciej>
annevk: it would be nice if the name was something that implies "dropping privileges"

[01:25:03.0000] <othermaciej>
hsivonen: Set-Cookie is done in the http stack but I believe http stacks expose enough that the browser could inhibit it per-request

[01:25:04.0000] <annevk>
2) some kind of replacement for new XMLHttpRequest(true)

[01:26:00.0000] <othermaciej>
annevk: I could have sworn HTML5 handled Referer

[01:26:01.0000] <annevk>
3) maybe some twiddling with the definition of credentials and at least renaming it

[01:26:02.0000] <annevk>
othermaciej, it does, but it does not allow for disabling it

[01:27:00.0000] <annevk>
othermaciej, what we have here is a Document object with a known origin and an API that wants a unique origin; all HTML5 allows for here is including the Referer and setting Origin to null

[01:27:01.0000] <othermaciej>
annevk: "If the origin of the appropriate Document is not a scheme/host/port tuple, then the Referer (sic) header must be omitted, regardless of its value."

[01:27:02.0000] <annevk>
(i.e. context-free requests)

[01:27:03.0000] <othermaciej>
http://dev.w3.org/html5/spec/Overview.html#fetch Step 1

[01:28:00.0000] <annevk>
othermaciej, right, but the origin of _Document_ is a tuple, it's just that the XMLHttpRequest origin is not

[01:29:00.0000] <othermaciej>
annevk: ah - Referer is obtained from the Document's origin, not the passed-in origin

[01:29:01.0000] <othermaciej>
annevk: that is weird

[01:29:02.0000] <othermaciej>
annevk: perhaps Referer should also be dropped when the passed-in origin does not match the document's origin

[01:30:00.0000] <othermaciej>
I don't know what cases besides this new XHR feature use an explicit origin different from the document origin

[01:32:00.0000] <othermaciej>
I wish it was easy to find all references to the fetch algorithm

[01:32:01.0000] <annevk>
yeah, something like that might make sense

[01:32:02.0000] <othermaciej>
it's kind of hard to find which ones, if any, pass in an explicit origin

[01:32:03.0000] <annevk>
in the WHATWG copy you can click on "fetch"

[01:32:04.0000] <othermaciej>
I suspect nothing in HTML5 proper does

[01:32:05.0000] <annevk>
not sure if the W3C has the backreferences script

[01:32:06.0000] <othermaciej>
sold!

[01:32:07.0000] <othermaciej>
I only use the w3c copy because it's higher in the google rankings and I'm too lazy to type either URL

[01:32:08.0000] <othermaciej>
or even autocomplete

[01:34:00.0000] <othermaciej>
EventSource explicitly passes an origin, but the one it passes explicitly is the default one that would be used for an API

[01:34:01.0000] <annevk>
browsers should make the faster path to sites more convenient

[01:34:02.0000] <othermaciej>
likewise for importScripts

[01:34:03.0000] <othermaciej>
if I visit the spec enough times in a week it ends up in my Safari Top Sites

[01:34:04.0000] <othermaciej>
but I don't read it quite enough for it to stay there

[01:34:05.0000] <annevk>
what I think is actually the case here is that the origin argument is just used to determine whether the Origin header should be set to null or not

[01:34:06.0000] <othermaciej>
though I guess I could pin it if I cared to

[01:35:00.0000] <othermaciej>
the origin argument isn't actually passed in the Origin header?

[01:35:01.0000] <othermaciej>
no, that's not right "For the purposes of the Origin header, if the fetching algorithm was explicitly initiated from an origin, then the origin that initiated the HTTP request is origin."

[01:36:00.0000] <othermaciej>
I see - when the origin is *not* passed in, then you do "Origin: null"

[01:36:01.0000] <othermaciej>
kinda wacky

[01:37:00.0000] <annevk>
it's not very convenient either because you need to know beforehand whether to pass it or not based on whether it is a tuple or a unique identifier

[01:37:01.0000] <othermaciej>
if it's a unique identifier then I think it will produce "Origin: null" anyway

[01:37:02.0000] <othermaciej>
I guess the other places that hide the origin don't want to hide the referer

[01:38:00.0000] <annevk>
guess I should check up on that in the Origin draft

[01:39:00.0000] <annevk>
ah yes, you're right

[01:40:00.0000] <annevk>
the Origin draft deals with serialization

[01:40:01.0000] <annevk>
http://tools.ietf.org/html/draft-abarth-origin for people following at home

[01:44:00.0000] <annevk>
(renamed credentials to user credentials)

[01:47:00.0000] <hsivonen>
has any browser implemented the Origin header yet?

[01:47:01.0000] <annevk>
Gecko and WebKit I think

[01:48:00.0000] <annevk>
oh, and IE for XDomainRequest

[01:48:01.0000] <hsivonen>
annevk: for all requests or just for XHR?

[01:51:00.0000] <annevk>
for all I believe

[01:51:01.0000] <othermaciej>
I think WebKit implements it for all

[01:51:02.0000] <othermaciej>
but I am not sure

[01:53:00.0000] <hsivonen>
hmm. I can't easily test, because the Gecko builds I use are always ahead of extensions

[01:53:01.0000] <hsivonen>
the good thing about Opera's kitchen sink approach is that the kitchen sink is built with the rest of the app

[01:54:00.0000] <hsivonen>
ok. I managed to override Live HTTP Headers extension compat

[01:54:01.0000] <hsivonen>
it seems Gecko doesn't send Origin for all requests

[01:56:00.0000] <hsivonen>
even the GET to whatwg revision.dat doesn't have Origin according to Live HTTP Headers

[01:57:00.0000] <hsivonen>
/me sees "However, a centralized repository for mappings is an obviously desirable thing." in http://people.xiph.org/~xiphmont/lj-pseudocut/o-response-1.html

[01:58:00.0000] <hsivonen>
/me wonders if the TAG would agree

[01:59:00.0000] <othermaciej>
yeah, why don't they just embed URIs in the code stream, that map to prefixes, which are then embedded in each packet to identify the mapping

[01:59:01.0000] <othermaciej>
s/code stream/codec stream/

[02:00:00.0000] <othermaciej>
that's the more natural way to do it

[02:06:00.0000] <annevk>
filed http://www.w3.org/Bugs/Public/show_bug.cgi?id=9603 and http://www.w3.org/Bugs/Public/show_bug.cgi?id=9604

[02:07:00.0000] <gsnedders>
Dashiva: Hey, that still gives me a year to get married!

[02:07:01.0000] <annevk>
I handled setting cookies in a generic section for now, but it might make more sense to have it as part of fetching, since it is handled there already

[02:07:02.0000] <gsnedders>
Dashiva: And seeming I'm meant to be engaged to a Swedish girl by the end of August, that's plenty of time

[02:07:03.0000] <gsnedders>
*just under two years to get married

[02:40:00.0000] <jgraham>
gsnedders: GLWT

[02:57:00.0000] <gsnedders>
/me feels bad having to look up what GLWT means

[02:58:00.0000] <jgraham>
gsnedders: It's short for "all the best looking boys are taken, all the best looking girls are staing inside"

[02:59:00.0000] <jgraham>
*staying

[03:04:00.0000] <gsnedders>
So can I get a good looking, but not best looking girl/boy?

[03:09:00.0000] <annevk>
or maybe we should just stick to AnonXMLHttpRequest()

[03:09:01.0000] <annevk>
it wasn't a bad name

[03:09:02.0000] <annevk>
/me sighs

[03:12:00.0000] <roc>
it sounded good to me

[03:12:01.0000] <roc>
"Uniform" is a total head-scratcher

[03:15:00.0000] <jgraham>
/me wonders what the suggestions have been

[03:15:01.0000] <jgraham>
"GuestXML..." sounds very dubious to me

[03:16:00.0000] <jgraham>
I can't think of anywhere else on the Web where "Guest" is used to mean "without sending some data"

[03:16:01.0000] <jgraham>
And it's not obvious that that's what it does mean from the name

[03:17:00.0000] <annevk>
outside the web it usually means something like what it means here

[03:17:01.0000] <annevk>
guest accounts

[03:17:02.0000] <jgraham>
Yeah, I guessed that's where it came from

[03:17:03.0000] <jgraham>
But I still don't think it is obvious

[03:17:04.0000] <annevk>
would be great if there was any web platform precedent

[03:18:00.0000] <hsivonen>
XMLHttpRequestNoLogin

[03:19:00.0000] <annevk>
that would get the same objections as Anon

[03:19:01.0000] <Philip`>
XMLHttpRequest2

[03:19:02.0000] <hsivonen>
/me is unaware of the objection

[03:19:03.0000] <annevk>
see this subthread: http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/thread.html#msg262

[03:19:04.0000] <annevk>
somewhere around there

[03:20:00.0000] <annevk>
Philip`, no

[03:20:01.0000] <jgraham>
/me is unaware of the whole thread other than what has been discussed here

[03:20:02.0000] <annevk>
alright, going for food now

[03:20:03.0000] <annevk>
will check logs later

[03:20:04.0000] <hsivonen>
annevk: anyway, I agreewith zcorpan's reasoning for having a suffix instead of a prefix

[03:22:00.0000] <jgraham>
/me has exactly no idea what semantic "Uniform" is supposed to convey

[03:23:00.0000] <Dashiva>
The same as URL, that is, nothing useful

[03:42:00.0000] <gsnedders>
So I found out yesterday that I have just over a week to decide about uni, which goes against what the guide said about when the deadline to decide was

[03:49:00.0000] <Philip`>
I thought you'd already spent about two years trying to decide, so another few weeks or months wouldn't make any difference

[03:53:00.0000] <gsnedders>
I thought three

[03:54:00.0000] <Philip`>
Three is about two

[03:57:00.0000] <gsnedders>
jgraham just bullied me into accepting it.

[03:57:01.0000] <gsnedders>
Now he rolling his eyes

[03:57:02.0000] <gsnedders>
And saying I speak like a foreigner

[03:57:03.0000] <gsnedders>
I must be Swedish already

[03:57:04.0000] <gsnedders>
No need to get married and stay here now

[04:59:00.0000] <hsivonen>
hmm. it sure is confusing that in the case of multiple <base> elements, the first one with href takes effect for href and the first one with target takes effect for target

[05:05:00.0000] <zcorpan>
hsivonen: is that implemented in gecko?

[05:07:00.0000] <hsivonen>
zcorpan: as of this week, yes

[05:32:00.0000] <mut>
hey, on canvas, how can i display a degrees symbol in text... like &deg

[05:34:00.0000] <gsnedders>
ctx.fillText("°", 0, 0)?

[05:34:01.0000] <mut>
hmm its not showing

[05:34:02.0000] <mut>
heh just shows a ?

[05:34:03.0000] <mut>
in a blackbox

[05:34:04.0000] <gsnedders>
ctx.fillText("\u00B0", 20, 20) if you don't want to actually have non-ASCII chars

[05:34:05.0000] <roc>
maybe your font doesn't have it?

[05:34:06.0000] <mut>
ah

[05:35:00.0000] <roc>
what browser?

[05:35:01.0000] <mut>
;) thanks

[05:35:02.0000] <mut>
err firefox

[05:35:03.0000] <mut>
chroime

[05:35:04.0000] <roc>
hmm

[05:35:05.0000] <gsnedders>
Yeah, that's the other likely thing

[05:35:06.0000] <mut>
and ie

[05:35:07.0000] <roc>
wait

[05:35:08.0000] <roc>
IE?

[05:35:09.0000] <mut>
yea with plugin

[05:35:10.0000] <mut>
sorry

[05:35:11.0000] <roc>
OK, no idea

[05:35:12.0000] <roc>
Firefox should display something if *any* font in your system supports that character

[05:35:13.0000] <roc>
which seems ... likely

[05:35:14.0000] <mut>
sweet that works now

[05:35:15.0000] <mut>
in ie too ;)

[05:51:00.0000] <jgraham>
hsivonen: Party time!

[05:51:01.0000] <gsnedders>
hsivonen: Congrats on getting it enabled by default at last

[05:51:02.0000] <hsivonen>
jgraham: yeah. and time to update the validotor, too. :-)

[05:51:03.0000] <hsivonen>
gsnedders: thanks

[05:54:00.0000] <gsnedders>
I guess the real question is whether you now get a load more bugs reported.

[05:54:01.0000] <hsivonen>
yes. interesting times.

[05:58:00.0000] <roc>
I'm glad our parser is deterministic now

[05:58:01.0000] <roc>
at least I hope it is!

[06:03:00.0000] <hsivonen>
roc: the tree shape should be deterministic now but the network loads started as the side effect of the parse aren't

[06:29:00.0000] <zcorpan>
hsivonen: data:text/html;charset=utf-8,x%c2

[06:30:00.0000] <zcorpan>
hsivonen: should give an U+FFFD, right? doesn't with html5.enable (haven't tested without)

[06:30:01.0000] <hsivonen>
zcorpan: yes, it should. thanks

[08:23:00.0000] <Dashiva>
/me wonders if the sarcasm tag will survive last call

[08:25:00.0000] <gsnedders>
Dashiva: Which sarcasm tag?

[08:25:01.0000] <Dashiva>
The one in HTML5

[08:25:02.0000] <gsnedders>
Ah, the end sarcasm tag ;P

[08:27:00.0000] <Dashiva>
It's a shame you need <sarcasm> as well to avoid the parse error

[08:27:01.0000] <jgraham>
Dashiva: What is this last call of which you speak

[08:28:00.0000] <jgraham>
?

[08:28:01.0000] <Dashiva>
It's a mythical golden age prophecized to occur in the near future

[08:29:00.0000] <Dashiva>
When all bugs and issues will be vanquished, and leading to the implementations rising to claim their heritage

[08:30:00.0000] <jgraham>
When have prophets even been right?

[08:30:01.0000] <Philip`>
jgraham: Whenever they've made a sufficient number of sufficiently vague predictions that some are bound to be interpreted after the fact as right?

[08:31:00.0000] <CoJaBo-Aztec>
the economy sucks. Not much prophets now :P

[08:32:00.0000] <cardona507>
good morning :)

[08:32:01.0000] <jgraham>
Philip`: This "last call" thing sounds altogether too specific in that case

[08:32:02.0000] <Dashiva>
No, no

[08:32:03.0000] <Dashiva>
You assume a concrete definition of bugs and issues in that case

[08:35:00.0000] <gsnedders>
Sitting cross-legged on my chair always ends up with on my legs being numb

[08:35:01.0000] <gsnedders>
That makes standing up challenging

[08:35:02.0000] <hsivonen>
well, the HTML5 parsing fun didn't last. http://hg.mozilla.org/mozilla-central/rev/ccb50d524490

[08:37:00.0000] <CoJaBo-Aztec>
randomorange?

[08:37:01.0000] <jgraham>
Orange == failed tests

[08:38:00.0000] <jgraham>
random orange == random failed tests

[08:38:01.0000] <jgraham>
== bad

[08:38:02.0000] <jgraham>
(I think)

[08:39:00.0000] <davidb>
yes random == hard 

[08:39:01.0000] <gsnedders>
Random tests are my favourite.

[08:40:00.0000] <davidb>
what about random failures? :)

[08:40:01.0000] <jgraham>
The "I think wasn't about "== bad"

[08:40:02.0000] <jgraham>
s/k/k"/

[08:40:03.0000] <gsnedders>
davidb: They're just a subset of them

[08:40:04.0000] <davidb>
jgraham all your statements looked correct

[08:43:00.0000] <jgraham>
How does Mozilla deal with tests that currently fail? Are they just not run or?

[08:43:01.0000] <davidb>
Everything is a judgement call, but generally we back out anything that regresses tests

[08:44:00.0000] <jgraham>
Right but I mean if you, say, add the CSS testsuite but know that you don't support all of the functionality yet

[08:44:01.0000] <jgraham>
So some tests fail

[08:45:00.0000] <jgraham>
/me remembers the expected fail state

[08:45:01.0000] <jgraham>
but if an expected fail passes does it have to be manually updated to expected pass or?

[08:50:00.0000] <davidb>
jgraham: yeah there are expected fail type of options, as well as per test things like mochitest "todo" etc. lots of choices, but generally requires manually updating when they pass

[08:50:01.0000] <davidb>
sorry for lag, i'm multi-chatting/tasking

[08:50:02.0000] <davidb>
as are we all i know :)

[08:52:00.0000] <hsivonen>
jgraham: upstream tests don't get run before someone sees them as green locally and pushes them to mozilla-central

[08:53:00.0000] <jgraham>
davidb: np. Thanks for sharing :)

[11:27:00.0000] <jgraham>
hsivonen: Ah. That is quite different from our approach

[12:03:00.0000] <davidb>
jgraham: what is your approach?

[12:18:00.0000] <JonathanNeal>
Hey everybody.

[12:46:00.0000] <jgraham>
davidb: Basically we record the result of every test in every build and look for changes between builds

[12:47:00.0000] <davidb>
i see

[12:47:01.0000] <davidb>
regressions are what matter

[12:48:00.0000] <jgraham>
Well yes, they get particular attention. 

[13:33:00.0000] <tonyg-cr>
Does anyone happen to know if first posts to public-webapps⊙wo are moderated and if so, who approves them?

[13:35:00.0000] <jwalden>
I don't believe they're moderated, but my knowledge is about a year old (I think)

[14:13:00.0000] <webben>
maybe interesting data point wrt OCR of images with missing @alt: the WebVisum extension (current status, maybe iffy, some of their links are down) for firefox actually did this: http://preview.tinyurl.com/27e8ubo


2010-04-29
[17:17:00.0000] <dglazkov>
Hixie: 4.10.21.3 form submission algorithm is inspired by ...?

[18:44:00.0000] <KaOSoFt>
Looking at the current specification, how can I know if an element is block or in-line?

[18:52:00.0000] <divya>
KaOSoFt: are you talking about content models or the CSS rendering?

[18:55:00.0000] <KaOSoFt>
Well, I'm not versed in this technical chat, but I guess I'm talking about a content model. Is <section> a "block" element? Like <div> and <span> (block and in-line).

[18:56:00.0000] <divya>
KaOSoFt: there are no more "block" and "inline" http://www.whatwg.org/specs/web-apps/current-work/multipage/section-index.html#element-content-categories

[18:56:01.0000] <KaOSoFt>
Oh, I see, that's why I didn't find any information about it.

[18:57:00.0000] <KaOSoFt>
Thank you very much, you kind sire.

[18:57:01.0000] <KaOSoFt>
:)

[18:57:02.0000] <KaOSoFt>
And goodbye, time to go back home.

[18:57:03.0000] <divya>
KaOSoFt: :) bye

[19:59:00.0000] <nessy>
Hixie: how is progress on the captions stuff?

[20:03:00.0000] <micheil>
where's that interactive tool for viewing the works in progress located again?

[20:17:00.0000] <Hixie>
nessy: pretty good. I've been trying to figure out which format would be best to use or extend: http://wiki.whatwg.org/wiki/Timed_track_formats

[20:21:00.0000] <othermaciej>
Hixie: people have claimed that the smilText subset of SMIL can be used as a timed text format

[20:21:01.0000] <othermaciej>
I'm not sure anyone identified a well-defined SMIL profile for this purpose

[20:21:02.0000] <othermaciej>
(I mention this because your page says "not a caption format")

[20:23:00.0000] <othermaciej>
http://www.w3.org/TR/SMIL/smil-text.html

[20:23:01.0000] <othermaciej>
looks like a weird mix of SMIL and DXFP

[20:23:02.0000] <Hixie>
oh that's not the same as DXFP?

[20:23:03.0000] <Hixie>
good to know

[20:23:04.0000] <Hixie>
i'll add it to the list

[20:25:00.0000] <othermaciej>
no, DFXP (aka TTML aka ttaf1-dfxp) is this: http://www.w3.org/TR/ttaf1-dfxp/

[20:25:01.0000] <Hixie>
good lord how many names can one standard have

[20:26:00.0000] <othermaciej>
I know, right?

[20:26:01.0000] <othermaciej>
and TTML is the only one I can type without scrambling the letters

[20:27:00.0000] <othermaciej>
and it's the title of the spec, but does not even appear in the shortname, and the Status of this Document section uses a different title to refer to it than the actual title

[20:27:01.0000] <Hixie>
relatedly: in doing my research i keep coming across the assertion that RealText is SMIL. e.g. wikipedia actually had RealText be a hyperlink to the SMIL page. But I can't find anything that actually supports that.

[20:27:02.0000] <Hixie>
i changed the wikipedia page

[20:29:00.0000] <othermaciej>
this seems like a really confusing area

[20:30:00.0000] <Hixie>
yeah

[20:58:00.0000] <nessy>
TTML was a last-minute invention

[20:58:01.0000] <Hixie>
the name?

[20:58:02.0000] <nessy>
I wished they had kept DFXP because then TTML would have been free to use for something else

[20:58:03.0000] <nessy>
yes, the name

[20:58:04.0000] <Hixie>
heh

[20:59:00.0000] <nessy>
cause … I kinda like the name timed text markup language !

[20:59:01.0000] <Hixie>
DFXP is a pretty bad name

[20:59:02.0000] <nessy>
all historic and academic

[20:59:03.0000] <Hixie>
seems like pretty much every name has been taken already

[20:59:04.0000] <nessy>
so about SMIL ...

[21:00:00.0000] <nessy>
SMILText is a spec they developed only rather recently

[21:00:01.0000] <nessy>
RealText was there before

[21:00:02.0000] <nessy>
and it was based on a subset of an earlier SMIL spec

[21:00:03.0000] <nessy>
unfortunately "SMIL" is being used for all sorts of things, when it is really always just a tiny subpart of SMIL

[21:01:00.0000] <nessy>
DAISY for example takes a subpart and extends it

[21:01:01.0000] <Hixie>
RealText doesn't seem to have anything to do with SMIL, does it? Or does using "begin" and "end" attributes in an XML format mean that you're using SMIL?

[21:01:02.0000] <nessy>
the Microsoft server manifest for Smooth Streaming is a tiny subpart of SMIL and is still called a "SMIL" use

[21:01:03.0000] <nessy>
lol

[21:02:00.0000] <Hixie>
i'm not even sure RealText is actually XML

[21:02:01.0000] <Hixie>
and the microsoft thing is definitely not XML

[21:02:02.0000] <Hixie>
so apparently XML isn't even a requirement

[21:02:03.0000] <nessy>
anything on the Web that used to do anything with media was called a SMIL technology at one stage

[21:02:04.0000] <nessy>
it's all being handled in a very loose manner

[21:03:00.0000] <nessy>
(which annoys the crap out of me, but don't tell anyone)

[21:03:01.0000] <roc>
it's great, everyone uses SMIL so browsers should too

[21:03:02.0000] <annevk>
that's certainly the TTML sentiment

[21:04:00.0000] <Hixie>
so should I just put SMIL into the title of whatever we're doing and then pretend we're doing SMIL too?

[21:04:01.0000] <nessy>
I think RealText was the way to synchronise text with SMIL before SMILText was created

[21:04:02.0000] <nessy>
Hixie: that would certainly solve the political situation!

[21:04:03.0000] <annevk>
hmm, the Ubuntu countdown went from 1 day left to available soon, but what does that mean?

[21:04:04.0000] <nessy>
suddenly SMIL would be a success on the Web

[21:04:05.0000] <Hixie>
we can call it "SCF" and say it stands for "SMIL Captioning Format" (but really it would stand for "Subtitles and Captions Format")

[21:04:06.0000] <nessy>
anyway - I am sarcastic today, pardon my mood

[21:05:00.0000] <Hixie>
don't worry, sarcasm is a prerequisite for being in the channel

[21:05:01.0000] <Hixie>
how else would the false controversies emerge but for our sarcasm

[21:06:00.0000] <nessy>
I'm sure someone will hold it in my face at the right opportunity

[21:07:00.0000] <nessy>
btw 3GPP-TT is a subpart of an early version of DFXP

[21:07:01.0000] <nessy>
years ago I had a student do an analysis of available subtitle formats - you can still find the result here: https://trac.annodex.net/wiki/CmmlSubtitles

[21:08:00.0000] <nessy>
Microsoft's SAMI wasn't actually that bad!

[21:08:01.0000] <Hixie>
SAMI is a disaster from the point of view of us speccing it and/or extending it

[21:08:02.0000] <nessy>
it's not support any more anyway

[21:08:03.0000] <roc>
I just found out that "Adobe Digital Editions" supports an "extended EPUB" ebook format ... which mixes XHTML, CSS and XSL-FO. And whoever produced it is kinda bitter about CSS3

[21:08:04.0000] <Hixie>
we'd have to reverse engineer their HTML-like parser

[21:09:00.0000] <Hixie>
roc: url?

[21:09:01.0000] <nessy>
EPUB is produced by the DAISY guys

[21:09:02.0000] <roc>
http://blogs.adobe.com/digitaleditions/template.html ... "Since CSS group chose to ignore a lot of perevious work in this area (e.g. inventing its own expression syntax for media queries instead of using XPath, designing its own incompatible replacement of page master instead of building on top of XSL:FO)"

[21:09:03.0000] <Hixie>
nessy: (will check that link, thanks)

[21:09:04.0000] <nessy>
(see if there is anything useful there for you)

[21:10:00.0000] <Hixie>
roc: huh. i wonder who wrote that.

[21:26:00.0000] <roc>
Peter Sorotokin

[21:29:00.0000] <Hixie>
ah, interesting

[21:29:01.0000] <Hixie>
he's the guy with whom i had a lot of, ah, discussions, in the sxbl days

[21:30:00.0000] <annevk>
comparing Selectors with XPath makes sense I suppose, but Media Queries and XPath?

[21:31:00.0000] <othermaciej>
wait, why would you use XPath for Media Queries?

[21:31:01.0000] <othermaciej>
/me arches both eyebrows

[21:31:02.0000] <othermaciej>
s/why/how/?

[21:31:03.0000] <othermaciej>
and criticizing CSS for not building on top of XSL FO?

[21:31:04.0000] <othermaciej>
my head is exploding

[21:33:00.0000] <roc>
someone recently told me we should build stuff on XSL-FO instead of CSS because XSL-FO is a REC and CSS 2.1 isn't

[21:44:00.0000] <annevk>
interesting how hard it is to replace the HTML parser

[21:44:01.0000] <annevk>
initially I wouldn't have thought that, but I guess I didn't know much about how intertwined it was

[21:52:00.0000] <Hixie>
annevk: in opera you mean?

[21:55:00.0000] <annevk>
Gecko, we haven't quite started

[22:01:00.0000] <roc>
I'm sure Henri can tell you about all kinds of problems

[22:01:01.0000] <roc>
:-)

[22:02:00.0000] <othermaciej>
I'm glad Mozilla is biting the bullet on this first

[22:02:01.0000] <annevk>
yeah, it's pretty cool

[22:03:00.0000] <othermaciej>
seems like the best way to flush out potential problems with the parsing algorithm itself

[22:52:00.0000] <Hixie>
annevk: i expect opera will have a harder time, because of the way misnested tags are currently handled

[22:59:00.0000] <annevk>
I doubt that is the hardest part, but we'll find out

[23:03:00.0000] <boogyman>
Hixie: I am not sure if this was answered before, but what would happen if the server side attempted to assign a value to an <input> or <textarea> greater than that afforded by an explicit maxlength value

[23:05:00.0000] <Hixie>
anyone here at WWW2010?

[23:05:01.0000] <boogyman>
originally queried at http://krijnhoetmer.nl/irc-logs/whatwg/20100427#l-928

[23:06:00.0000] <Hixie>
boogyman: you mean in the markup? like <input value=123 maxlength=2> ?

[23:07:00.0000] <boogyman>
Hixie: yes, that's an appropriate example

[23:12:00.0000] <Hixie>
let me see...

[23:12:01.0000] <Hixie>
/me opens the spec

[23:13:00.0000] <Hixie>
nothing in particular seems to happen if it's set by the server

[23:13:01.0000] <Hixie>
it'll submit fine and it doesn't get truncated

[23:14:00.0000] <Hixie>
it probably doesn't bode well that the W3C is explicitly describing itself as split into two camps, RDF and HTML5... http://www.w3.org/2010/04/w3c-track.html

[23:14:01.0000] <boogyman>
so it should behave as if the 'user' has entered the text

[23:15:00.0000] <othermaciej>
Hixie: I don't think that's how they meant it

[23:15:01.0000] <othermaciej>
Hixie: I assume they mean "camp" in the sense of "something like Foo Camp"

[23:18:00.0000] <Hixie>
boogyman: no, if the user enters the text then when they submit it they'll get a message saying the text is too long

[23:18:01.0000] <zcorpan>
Hixie: i think the way misnested tags are currently handled will just be dead code when we have an html5 parser, so i think that won't be hard at all to remove :)

[23:18:02.0000] <Hixie>
true

[23:19:00.0000] <Hixie>
zcorpan: btw re the websocket stuff

[23:19:01.0000] <Hixie>
still going through my mail from having been off sick for a day

[23:19:02.0000] <Hixie>
but

[23:19:03.0000] <Hixie>
i don't think it matters where you drop the connection if you drop the connection

[23:19:04.0000] <Hixie>
since it's black-box indistinguishable

[23:20:00.0000] <Hixie>
however, you shouldn't use your HTTP parser, since it does different things (like continuation lines)

[23:20:01.0000] <Hixie>
and it might do error handling differently

[23:20:02.0000] <Hixie>
since HTTP doesn't say how to do error handling

[23:20:03.0000] <zcorpan>
Hixie: it's not black-box indistinguishable if the server sleeps half-way

[23:20:04.0000] <zcorpan>
Hixie: yeah

[23:21:00.0000] <Hixie>
the client is always allowed to insert aribitrary sleeps anywhere

[23:21:01.0000] <Hixie>
nothing in the spec defines the performance characteristics of a websocket client

[23:22:00.0000] <zcorpan>
Hixie: if the server responds "HTTP/1.1 200" and then sleeps, we want to abort but the spec doesn't allow it

[23:22:01.0000] <zcorpan>
since we haven't seen an 0x0A byte yet

[23:24:00.0000] <zcorpan>
continuation lines is something i should test

[23:28:00.0000] <Hixie>
hm yeah i should allow the UA to break early if it can prove that it wouldn't ever accept the connection

[23:28:01.0000] <Hixie>
assume that i'll make that allowed

[23:29:00.0000] <Hixie>
i hadn't thought of breaking early

[23:30:00.0000] <zcorpan>
Hixie: thanks

[23:30:01.0000] <Hixie>
in case i forget by the time i reply to your e-mail, just remind me of the above :-)

[23:31:00.0000] <zcorpan>
ok

[23:31:01.0000] <Hixie>
it won't mean i changed my mind or anything, just that i forgot

[23:36:00.0000] <Hixie>
man, some reporters ask the most inane questions

[23:41:00.0000] <boogyman>
agreed

[23:42:00.0000] <zcorpan>
such as "when will html5 be finished?"?

[23:42:01.0000] <Hixie>
no, that one i know how to answer

[23:42:02.0000] <annevk>
wir

[23:42:03.0000] <Peter`>
They probably wouldn't believe you, since browser vendors have already implemented HTML 5!

[23:42:04.0000] <Hixie>
i mean things like "do people use all of html4?"

[23:42:05.0000] <Peter`>
Gotta love marketing..

[00:01:00.0000] <Hixie>
wow, check out the number of errors in this article

[00:01:01.0000] <Hixie>
http://www.instantshift.com/2010/04/26/html5-worth-the-hype/

[00:02:00.0000] <boogyman>
157 html errors lol

[00:02:01.0000] <Hixie>
oh i didn't mean validation errors

[00:02:02.0000] <Hixie>
i meant in the text

[00:13:00.0000] <zcorpan>
"To use HTML5 you will need to use a specific scripting API such as:"

[00:14:00.0000] <JonathanNeal>
"javascript"

[00:14:01.0000] <hsivonen>
most of the problems I've been dealing with haven't been about the HTML5 parsing algorithm but about our test cases being bogus in a way that has stayed latent with the particular timing quirks of Gecko's old parser

[00:14:02.0000] <JonathanNeal>
"4chan"

[00:14:03.0000] <zcorpan>
seems like they copied some from wikipedia and edited it so it would be less obvious that they copied it from wikipedia

[00:23:00.0000] <annevk>
Ivan!

[00:29:00.0000] <zcorpan>
yeah i wonder who ivan is

[00:29:01.0000] <zcorpan>
maybe Hixie should change his irc nick to Ivan

[00:29:02.0000] <othermaciej>
ivan who?

[00:30:00.0000] <zcorpan>
Ivan Hickson

[00:30:01.0000] <boogyman>
lol

[00:30:02.0000] <othermaciej>
/me chortles

[01:02:00.0000] <hsivonen>
zcorpan: I filed the REPLACEMENT CHARACTER bug as https://bugzilla.mozilla.org/show_bug.cgi?id=562590. thanks.

[01:11:00.0000] <annevk>
http://blogs.msdn.com/ie/archive/2010/04/28/product-feedback-systems.aspx o_O

[01:11:01.0000] <annevk>
Microsoft seems to be upping the FUD machine recently

[01:11:02.0000] <annevk>
though maybe it is just me

[01:39:00.0000] <Hixie>
annevk: that does seem a bit rude

[01:40:00.0000] <Hixie>
but they have a point

[01:40:01.0000] <zcorpan>
hsivonen: other cases you might want to test are %e0%80, %f0%80%80, %f5%80%80%80 etc

[01:44:00.0000] <hsivonen>
zcorpan: ok

[01:46:00.0000] <othermaciej>
what is the FUD there?

[01:46:01.0000] <othermaciej>
"bugzilla will cause you to be spammed"?

[01:48:00.0000] <Hixie>
i don't think it's FUD

[01:48:01.0000] <othermaciej>
ok, having read more of it, it looks like they are pretty directly calling out at least some other browser vendors

[01:48:02.0000] <othermaciej>
I am surprised Microsoft would post such direct competitive statements on the IE blog

[01:49:00.0000] <Hixie>
it seems like a dangerous game for them to play, because if they piss off a browser vendor doing this kind of thing, there's plenty of material that can be thrown back at them

[01:50:00.0000] <othermaciej>
one thing he overlooks is that the WebKit project completely ignores the difference between "new" and "unconfirmed"

[01:52:00.0000] <hsivonen>
In my experience, Firefox and WebKit nightlies as well as Opera roughly-weeklies are quite usable enough to expose to interested parties

[01:52:01.0000] <Hixie>
i only use nightly builds

[01:52:02.0000] <othermaciej>
I doubt anyone from the WebKit project is interested in getting into a pissing match with Microsoft over who manages feedback better

[01:52:03.0000] <Hixie>
of webkit, chrome, firefox, and opera -- i haven't even got a stable build of a browser on my machine, as far as i know

[01:53:00.0000] <othermaciej>
is it a Mac?

[01:53:01.0000] <Hixie>
yeah

[01:53:02.0000] <othermaciej>
you probably have a stable build of Safari and of the associated WebKit

[01:53:03.0000] <othermaciej>
even if you don't browse with it

[01:53:04.0000] <Hixie>
ah probably, yeah

[01:53:05.0000] <zcorpan>
i think i've had higher FIXED rate when filing bugs at mozilla compared to filing bugs at microsoft

[01:54:00.0000] <Hixie>
i think that comes up when i accidentally pick "search in google" from a context menu

[01:54:01.0000] <othermaciej>
or rather, you do browse with that Safari, cause that's what the WebKit nightly launcher uses

[01:54:02.0000] <Hixie>
(grrr @ apple)

[01:54:03.0000] <othermaciej>
you can make the nightly launcher your default browser I believe

[01:54:04.0000] <zcorpan>
all microsoft feedback was "actioned", but i don't care if a bug is "RESOLVED POSTPONED" or still unconfirmed

[01:54:05.0000] <Hixie>
can i make chrome the default? :-)

[01:54:06.0000] <Hixie>
or firefox? :-)

[01:54:07.0000] <othermaciej>
zcorpan: well, "postponed" means it was considered and rejected, rather than just not acted on it yet

[01:54:08.0000] <othermaciej>
Hixie: yes

[01:54:09.0000] <Hixie>
how?

[01:55:00.0000] <othermaciej>
you can make any browser your default browser

[01:55:01.0000] <othermaciej>
they have built-in UI, and Safari also has UI in prefs that lets you make any browser the default browser

[01:55:02.0000] <Hixie>
so that "search in google" launches that browser?

[01:55:03.0000] <hsivonen>
Hixie: sometimes the setting resets back to safari at login, in which case you have to go through various .plists or whatever and delete the right one

[01:55:04.0000] <othermaciej>
I actually don't know how the "search in google" context menu is coded, it's in AppKit I think

[01:56:00.0000] <othermaciej>
it's possible that it doesn't correctly respect the default browser setting, unlike nearly anything else on the system

[01:56:01.0000] <zcorpan>
othermaciej: yeah. although for at least one of my bugs at microsoft it was clear that they hadn't considered the bug at all when closing it

[01:56:02.0000] <hsivonen>
I managed to delete the right thing on my work Mac but on my home Mac, I still need to make Minefield the default after every login

[01:56:03.0000] <Hixie>
(i actually don't use the nightly launcher btw, i just use whatever it is that causes webkit to offer me a prompt to update the build)

[01:56:04.0000] <othermaciej>
zcorpan: closing something without giving it actual consideration is worse than leaving it "unconfirmed" or "new" :-/

[01:57:00.0000] <othermaciej>
Hixie: I believe the nightly launcher updates itself automagically now

[01:57:01.0000] <Hixie>
othermaciej: yeah, looks like it's ignoring the seeting. I dragged a favicon from webkit to the dock, then clicked on it, and chrome opened. So I assume chrome is the default.

[01:57:02.0000] <othermaciej>
Hixie: yes, probably

[01:58:00.0000] <Hixie>
anyway, didn't mean to turn this into a gripefest :-)

[01:58:01.0000] <othermaciej>
Hixie: if you open Safari (or a WebKit nightly) and look at general preferences, it will tell you what the system default is supposed to be

[01:58:02.0000] <othermaciej>
WebKit nightlies do tend to be reasonably stable, though there is the occasional build with a bad bug

[01:59:00.0000] <hsivonen>
well, recently with Minefield, I lost xkcd tooltips for a few days

[01:59:01.0000] <hsivonen>
and then on Mac if you have Flash 10.1 installed, things aren't so great right now

[02:00:00.0000] <Hixie>
i had a chrome build recently where using a <select> made <textarea>s ignore any function keys (like backspace or arrow keys), which was fun while that lasted

[02:00:01.0000] <Hixie>
but that's why i have four browsers!

[02:00:02.0000] <Hixie>
i wish opera nightlies autoupdated, they don't seem to

[02:00:03.0000] <hsivonen>
so I went back to Flash 10.0.x.x on the computer where I want Flash to actually work (after filing my stack trace of course)

[02:01:00.0000] <hsivonen>
Hixie: for me on Mac, Opera autodownloads updates and then fails to install them

[02:01:01.0000] <Hixie>
i've seen that too

[02:01:02.0000] <Hixie>
but more recently it just doesn't seem to download anything at all

[02:02:00.0000] <zcorpan>
Hixie: have you checked opera:config#AutoUpdate|DownloadAllSnapshots ?

[02:02:01.0000] <Hixie>
/me looks

[02:03:00.0000] <Hixie>
it is unchecked

[02:03:01.0000] <Hixie>
do i just check it?

[02:03:02.0000] <Hixie>
how does it know to update to the nightlies and not release builds?

[02:03:03.0000] <zcorpan>
yep, it should enable autoupdate of weeklies

[02:03:04.0000] <hsivonen>
zcorpan: is it supposed to work on Ubuntu? 

[02:04:00.0000] <zcorpan>
Hixie: it knows by looking at the pref

[02:04:01.0000] <zcorpan>
hsivonen: i think so, but haven't tested

[02:04:02.0000] <Hixie>
oh, that's what snapshot means

[02:04:03.0000] <Hixie>
cunning

[02:04:04.0000] <Hixie>
oooh

[02:04:05.0000] <Hixie>
hey awesome

[02:04:06.0000] <Hixie>
thanks dude

[02:04:07.0000] <othermaciej>
hsivonen: Flash 10.1 has some bugs in its CA drawing model code, which is only enabled for WebKit nightlies currently

[02:04:08.0000] <zcorpan>
Hixie: it works?

[02:05:00.0000] <othermaciej>
hsivonen: there are also, I think, some bugs in our side of the code that didn't get flushed out with previous plugins

[02:05:01.0000] <hsivonen>
othermaciej: I believe I saw CA on the Minefield+OOPP+Flash 10.1 stack I filed

[02:06:00.0000] <othermaciej>
maybe they are trying to use it on Minefield and failing

[02:06:01.0000] <hsivonen>
this was after out-of-process CA was enabled in Minefield, but clearly they are failing

[02:07:00.0000] <Hixie>
zcorpan: yup

[02:08:00.0000] <hsivonen>
iframes created with innerHTML fire load events normally, right?

[02:16:00.0000] <hsivonen>
/me wonders why so many test cases try to avoid simply listening for the load event and do something else and incorrect

[02:17:00.0000] <Hixie>
no spec before html5 defined innerHTML or load events, so it's not hard to imagine that people wouldn't be sure of what was the right thing

[02:17:01.0000] <hsivonen>
s/listening/waiting/ or s/for/to/ I guess

[02:19:00.0000] <annevk>
zcorpan, I don't believe that works on Ubuntu

[02:19:01.0000] <annevk>
zcorpan, as we integrate with the system there

[02:20:00.0000] <annevk>
but maybe...

[02:31:00.0000] <zcorpan>
Hixie: nice

[02:53:00.0000] <zcorpan>
/me files https://connect.microsoft.com/IE/feedback/details/555265/video-should-not-fall-back-even-if-the-video-fails

[03:01:00.0000] <JonathanNeal>
Good bug to file :)

[03:07:00.0000] <zcorpan>
/me gets an email saying:

[03:07:01.0000] <zcorpan>
This is a text part of the message.

[03:07:02.0000] <zcorpan>
It is shown for the users of old-style e-mail clients

[03:07:03.0000] <zcorpan>
why thank you, that's helpful

[03:08:00.0000] <hsivonen>
zcorpan: does ie9 have the blue border around linked images by default?

[03:09:00.0000] <zcorpan>
hsivonen: don't know

[03:40:00.0000] <roc>
does the IE9 preview actually support video now?

[03:41:00.0000] <Philip`>
No

[03:41:01.0000] <Philip`>
though I guess it's about time for them to release a new preview now

[03:52:00.0000] <hsivonen>
when I try to sign in to see zcorpan's bug report, I get The content that you requested cannot be found or you do not have permission to view it.  

[03:54:00.0000] <Philip`>
hsivonen: Works for me and says "Access Restriction: Public"

[03:55:00.0000] <Philip`>
(Apparently I am a participant in the Internet Explorer Tech Feedback Program which may make a difference)

[03:56:00.0000] <hsivonen>
around the IE8 time, IIRC, I decided against enrolling into that program due to terms that would have been unwise to accept while working on a competing product

[03:58:00.0000] <hsivonen>
apparently now there's a lesser level of Microsoft Connect access available under less scary terms

[06:45:00.0000] <Dashiva>
On the .URL topic, what about also adding document.url? It leaves back compat in place, but lets authors standardize on .url for everything in new content

[07:08:00.0000] <jgraham>
http://www.whatwg.org/specs/web-apps/current-work/#dom-context-2d-drawimage does that actually say in normative text that TYPE_MISMATCH_ERROR should be raised if the first argument is the wrong type (other than null)

[07:10:00.0000] <Philip`>
jgraham: WebIDL says it must

[07:10:01.0000] <Philip`>
(hopefully)

[07:10:02.0000] <Philip`>
(HTML5 used to say it but it was removed because it's redundant with WebIDL, I believe)

[07:11:00.0000] <jgraham>
Ah, I hoped it would be somrthing like that

[07:25:00.0000] <GPHemsley>
So... Team Rocket started off the discussion on headers...

[07:38:00.0000] <jgraham>
zcorpan: yt?

[07:57:00.0000] <remysharp>
Am I right in saying that Web Workers *aren't* spec'ed to use JSON to automatically convert objects - whereas Firefox have implemented this as an extra: https://developer.mozilla.org/En/Using_web_workers#Sending_objects_to_workers

[07:57:01.0000] <micheil>
oh, hey remy

[07:58:00.0000] <remysharp>
micheil: hi there.

[07:58:01.0000] <micheil>
remysharp: jsbin wasn't it?

[07:58:02.0000] <remysharp>
yep

[07:58:03.0000] <micheil>
yeah, nice work there

[07:58:04.0000] <remysharp>
cheers :)

[07:59:00.0000] <Philip`>
remysharp: As far as I'm aware, it should use http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#safe-passing-of-structured-data to transfer objects

[08:00:00.0000] <remysharp>
Philip`: so in fact it shouldn't be giving me a string on the other side each time?

[08:00:01.0000] <remysharp>
it should be able to handle some objects and types

[08:02:00.0000] <gsnedders>
Anyone have IE9 at hand?

[08:02:01.0000] <Philip`>
remysharp: I think so, unless I'm mistaken

[08:02:02.0000] <remysharp>
Philip`: cheers - shame most of the browsers don't do that yet, but I'm sure they'll catch up!

[08:03:00.0000] <Philip`>
Looks like Workers use http://www.whatwg.org/specs/web-apps/current-work/multipage/comms.html#dom-messageport-postmessage

[08:03:01.0000] <Philip`>
which uses that structured clone algorithm

[08:03:02.0000] <Philip`>
/me wishes it wasn't so hard to find definitions of these functions

[08:03:03.0000] <Philip`>
(since the Workers spec doesn't seem to have any direct link to it)

[08:04:00.0000] <micheil>
hmm.. did html5 introduce a module tag?

[08:04:01.0000] <micheil>
 / element

[08:05:00.0000] <gsnedders>
Philip`: Is that because of lack of xdoc xref?

[08:06:00.0000] <micheil>
actually no. looks like site is using their own xml schema

[08:09:00.0000] <jgraham>
gsnedders: Everything is your fault

[08:15:00.0000] <gsnedders>
/me looks innocent

[08:15:01.0000] <micheil>
gsnedders: hmm.. mozilla dev, yes?

[08:16:00.0000] <micheil>
or atleast, their irc channels.

[08:16:01.0000] <TabAtkins>
Oh, wow.  This article Ian linked to a few hours ago really *is* just a copypastad and lightly editted wikipedia article.

[08:16:02.0000] <TabAtkins>
If not, then this person is *really bad* at writing.

[08:18:00.0000] <gsnedders>
micheil: Opera QA, ocasionally on irc.mozilla.org

[08:18:01.0000] <micheil>
gsnedders: ah, okay, I've just seen the name about

[08:19:00.0000] <hsivonen>
micheil: isn't <module> Douglas Crockford's thing?

[08:19:01.0000] <TabAtkins>
Yeah, it is.

[08:19:02.0000] <gsnedders>
/me dislikes it when people know him, it makes him feel less anonymous

[08:20:00.0000] <micheil>
hsivonen: no idea, I just saw it in a site which was using the html5 doctype and no other xmlns

[08:21:00.0000] <jgraham>
squeaky people!

[08:21:01.0000] <gsnedders>
/me looks oddly across the room at jgraham 

[08:22:00.0000] <jgraham>
No you didn'r

[08:22:01.0000] <jgraham>
t

[08:22:02.0000] <gsnedders>
I did. Just when you weren't looking.

[08:23:00.0000] <jgraham>
Oh 

[08:23:01.0000] <gsnedders>
You're not observant enough.

[08:23:02.0000] <jgraham>
Well you are not observing the squeaking

[08:23:03.0000] <gsnedders>
/me wonders if he could creep up on jgraham and surprising

[08:23:04.0000] <gsnedders>
I'm sorry, but Alice in Chains is better than squeaking

[08:24:00.0000] <jgraham>
The squeaking is not good. Although Alice in Chains is dubious

[08:24:01.0000] <gsnedders>
/me realizes what jgraham means by people squeaking

[08:24:02.0000] <jgraham>
as an alternative

[08:24:03.0000] <gsnedders>
jgraham: When did we ever agree on music?

[08:24:04.0000] <jgraham>
Sometimes

[08:25:00.0000] <gsnedders>
Rarely.

[08:28:00.0000] <jgraham>
Sigur Rós?

[08:28:01.0000] <gsnedders>
For example

[08:31:00.0000] <Dashiva>
I hope someone does a translation of this: http://www.apple.com/hotnews/thoughts-on-flash/

[08:33:00.0000] <hsivonen>
interesting that Steve blogs on the PR wire

[08:35:00.0000] <Dashiva>
First he claims you don't need flash because most video is h264 anyway, and then he says flash video takes too much resources because most of it isn't h264 and is handled in software

[08:37:00.0000] <gsnedders>
Wow those people are squeaky.

[08:37:01.0000] <jgraham>
No these ones are just noisy

[08:38:00.0000] <gsnedders>
I was assuming it was the same people, and I was a bit surprised at the leak of squeakiness

[08:38:01.0000] <gsnedders>
*lack

[08:38:02.0000] <gsnedders>
It was just loud

[08:39:00.0000] <Philip`>
Dashiva: He says most video is in Flash but is also in H.264

[08:39:01.0000] <Philip`>
(i.e. non-Flash H.264)

[08:39:02.0000] <Dashiva>
If it's available in h264, it can be hardware accelerated in flash

[08:39:03.0000] <Philip`>
But it's not available in H.264 in Flash

[08:40:00.0000] <Dashiva>
And this happens where? Why would anyone have a high quality format and not serve it to as many as possible?

[08:40:01.0000] <Philip`>
If web developers were given the choice between sticking with an old codec in Flash which works perfectly fine for them, or expending a lot of effort converting everything to H.264 (to make available via Flash or <video> or whatever), they'd probably choose the former because it's much easier and doesn't hurt them much

[08:41:00.0000] <TabAtkins>
Flash actually does hurt me.  It's annoying and fiddly to use.

[08:41:01.0000] <Philip`>
so Apple doesn't like giving them the choice

[08:41:02.0000] <Dashiva>
He can't have it both ways, though

[08:42:00.0000] <Dashiva>
Either most content is not h264, and then apple doesn't let you use the full web. Or it is h264, in which case the performance argument is false.

[08:43:00.0000] <Philip`>
He doesn't say Apple does let you use the full web

[08:43:01.0000] <Philip`>
"iPhone, iPod and iPad users aren’t missing much video" doesn't say "... aren't missing anything"

[08:43:02.0000] <hsivonen>
I'd be happier about Apple's stance if their chosen video format weren't royalty-encumbered

[08:43:03.0000] <Dashiva>
"What they don’t say is that almost all this video is also available in a more modern format, H.264"

[08:43:04.0000] <Dashiva>
almost all

[08:44:00.0000] <Philip`>
Most of it's only also available in H.264 because Apple removed the choice and forced developers to start supplying it

[08:45:00.0000] <Dashiva>
Well, is it available or is it not available?

[08:46:00.0000] <Philip`>
It is now, but if Apple did/does support Flash then it perhaps wouldn't be

[08:46:01.0000] <Dashiva>
But the actual reality is one or the other

[08:46:02.0000] <Dashiva>
It's not a superposition of "what if apple didn't push for h264" and "reality as it is now"

[08:48:00.0000] <Dashiva>
There's also this gem: "When websites re-encode their videos using H.264, they can offer them without using Flash at all."

[08:48:01.0000] <Dashiva>
I guess Firefox isn't a browser

[08:49:00.0000] <workmad3>
Dashiva: obviously not, according to apple :)

[08:49:01.0000] <Dashiva>
Oh wait

[08:49:02.0000] <Dashiva>
He already did that slide where Safari ate up the market share of every browser except IE

[08:49:03.0000] <Philip`>
/me mostly dislikes how apparently Apple are the only people in the world who can provide platform innovations, and third-party platforms built on top of it are just restrictive and nobody would ever want to write applications that work on other platforms

[08:50:00.0000] <Philip`>
It's trying to force a monoculture, which doesn't sound good

[08:51:00.0000] <workmad3>
Philip`: what, you're disagreeing with the Apple Is God world-view? :)

[08:51:01.0000] <Dashiva>
Hum, w3c going for mercurial

[08:52:00.0000] <jgraham>
Dashiva: Why hum?

[08:52:01.0000] <Rik`>
Dashiva: Firefox is a browser for Apple support, they recommend it to subscribe to WWDC :)

[08:53:00.0000] <Philip`>
workmad3: Sure - it's trendy to disagree with Apple nowadays

[08:53:01.0000] <workmad3>
Philip`: that's good... I've been disagreeing with them for years :)

[08:53:02.0000] <jgraham>
Philip`: Would you describe yourself as "trendy"?

[08:53:03.0000] <workmad3>
to me, all this new grumbling is people hopping on the anti-apple bandwagon after the fact :P

[08:54:00.0000] <Philip`>
jgraham: Hmm... No

[08:54:01.0000] <Dashiva>
jgraham: Not sure what it's intended for. w3c likes to keep close eyes on their specs.

[08:55:00.0000] <jgraham>
Dashiva: We are getting it for testsuites at least

[08:55:01.0000] <jgraham>
Dunno what is happening with the specs but I assume that they can't use torture^WCVS forever

[08:56:00.0000] <gsnedders>
Rik`, that reminds me… June 7–11th! Noooo!

[09:35:00.0000] <remysharp>
Philip`: sorry to bother you - you helped me with a postMessage and workers question -

[09:36:00.0000] <remysharp>
but the spec examples on the whatwg site implies that *only* strings will be received by the worker: http://dev.w3.org/html5/workers/#a-worker-for-updating-a-client-side-database

[09:36:01.0000] <Rik`>
gsnedders: hum ?

[09:36:02.0000] <gsnedders>
Rik`: WWDC

[09:36:03.0000] <gsnedders>
Rik`: Of course, the nooo will make no sense to you :)

[09:36:04.0000] <Rik`>
exactly

[09:36:05.0000] <remysharp>
there's several examples on in the Worker spec that split the event.data to read what it wants to do with it

[09:37:00.0000] <gsnedders>
Rik`: It stops me from running off with Apple people when we were going to

[09:37:01.0000] <gsnedders>
There'ss sense to this insanity.

[09:39:00.0000] <jgraham>
gsnedders: No eloping please

[09:40:00.0000] <gsnedders>
jgraham: Can I elope instead with a Scottish girl?

[09:41:00.0000] <TabAtkins>
gsnedders: Sounds good to me.  Find yourself a buxom redhead or raven-haired girl.

[09:41:01.0000] <jgraham>
Raven hair and skin so fair, sadness never visits there

[09:43:00.0000] <gsnedders>
/me expects he can find several girls meeting one of those alternations in his Facebook friends…

[09:44:00.0000] <Philip`>
remysharp: I'd guess that's because the example was written back when postMessage only supported strings, and/or is kept because it's a useful demonstration of how to work with legacy implementations that only support strings

[09:44:01.0000] <hober>
hmm. getting an error page on every wiki page I try to go to

[09:44:02.0000] <jgraham>
/me isn;t sure whether to parse it as (buxom & redhead) | raven haired or buxom & (redhead | raven haired)

[09:44:03.0000] <hober>
Database error: A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was: (SQL query hidden) from within function "User::saveOptions". Database returned error "1062: Duplicate entry '121-rememberpassword' for key 1 (database.wiki.whatwg.org)".

[09:44:04.0000] <TabAtkins>
The last, jgraham.

[09:45:00.0000] <jgraham>
TabAtkins: I think that makes it sizist, or something

[09:45:01.0000] <gsnedders>
TabAtkins: Oh, I parsed it differently.

[09:45:02.0000] <TabAtkins>
The fact that redhead and raven-haired modify the same attribute should be evidence that they form a localized disjuction.

[09:46:00.0000] <jgraham>
But maybe if you didn;t really like red hair you would suggest some way of making up for it?

[09:46:01.0000] <TabAtkins>
I wouldn't have listed red hair if I didn't like it.  See: the lack of blonde in my list.

[09:46:02.0000] <jgraham>
I just assumed that was because there are not blone girls in Scotland

[09:46:03.0000] <jgraham>
*blonde

[09:47:00.0000] <gsnedders>
jgraham: There are, just fewer than here.

[09:47:01.0000] <Philip`>
/me imagines a raven-haired girl is like Medusa but featherier

[09:47:02.0000] <TabAtkins>
Something like that, yes.

[09:49:00.0000] <gsnedders>
As far as I know, there's never been any conclusion about weather I'm the fairest person in the office

[09:50:00.0000] <gsnedders>
But I'm not raven-haired, so I guess sadness visits here. Sad.

[09:57:00.0000] <AryehGregor>
Argh, wiki errors?

[09:59:00.0000] <jgraham>
gsnedders: There aren't that many here

[10:00:00.0000] <AryehGregor>
. . . do I not have an account on the wiki?  Seems not.

[10:00:01.0000] <AryehGregor>
hober, does it work if you log out?

[10:00:02.0000] <gsnedders>
jgraham: Well, true, we're in the office of an IT company. There aren't many women here in general.

[10:01:00.0000] <jgraham>
gsnedders: That not what I menat, as well you know

[10:02:00.0000] <gsnedders>
:)

[10:03:00.0000] <AryehGregor>
Is anyone else getting errors on the wiki?

[10:05:00.0000] <AryehGregor>
hober, the error should give more details now.  Can you provide them?

[10:08:00.0000] <AryehGregor>
Or not.  Now it should be logging details.

[10:17:00.0000] <AryehGregor>
Hixie, you don't seem to have APC or anything similar set up as a PHP opcode cache.  Any particular reason for this?  It would help performance a lot for large PHP apps (e.g., MediaWiki).

[10:17:01.0000] <gsnedders>
AryehGregor: Because Hixie's a Perl fanboi

[10:19:00.0000] <AryehGregor>
Then he should clearly have used a wiki engine written in Perl.

[10:19:01.0000] <AryehGregor>
I have no idea why PHP doesn't support something like Python's .pyc out of the box, so that special opcode caches aren't necessary.

[10:20:00.0000] <AryehGregor>
Actually, I do know, it's because PHP is developed by monkeys pounding on typewriters.

[10:30:00.0000] <TabAtkins>
What happens if you point an <img> at something without intrinsic dimensions, such as an SVG image that has only an aspect ratio, but don't give the <img> a width or height?

[10:30:01.0000] <TabAtkins>
That is to say, what's the default size of an <img> when it can't get any appropriate information from the linked image?

[10:33:00.0000] <hober>
AryehGregor: unfortunately, I'm at a different machine now. the wiki works fine from here.

[10:33:01.0000] <AryehGregor>
Weird.

[10:33:02.0000] <AryehGregor>
Tell me if you have any problems again.

[10:35:00.0000] <hober>
will do

[11:29:00.0000] <zcorpan>
jgraham: am now

[12:11:00.0000] <jgraham>
zcorpan: That's less useful :)

[12:11:01.0000] <jgraham>
(hallvors was looking for you)

[12:19:00.0000] <AryehGregor>
"One of the assumptions most software organizations appear to take for granted is the need for beta releases.  In fact, some companies have even taken this assumption well beyond its commonly understood meaning." ("assumption" linking to a post about Gmail)

[12:19:01.0000] <AryehGregor>
Zing.

[12:21:00.0000] <AryehGregor>
Actually, the post it links to is someone who claims some of their e-mail vanished, so they stopped using Gmail.  That seems like an underhanded thing to link to.  Well, whatever.

[12:21:01.0000] <AryehGregor>
(The quote is from the latest IEBlog)

[12:29:00.0000] <AryehGregor>
Yeah, I guess that post is a steaming pile of PR.

[12:29:01.0000] <AryehGregor>
Oh wlel.

[12:29:02.0000] <AryehGregor>
Oh well.

[12:29:03.0000] <AryehGregor>
*Some* of the IEBlog posts aren't horrible, I'll give them that.

[12:32:00.0000] <othermaciej>
AryehGregor: I was really surprised by the snark level of that post

[12:32:01.0000] <othermaciej>
it detracts from the message they presumably want to send, which is about the awesome opportunity you have to give them feedback

[12:32:02.0000] <AryehGregor>
Yeah, it ended up being really misleading.

[12:33:00.0000] <AryehGregor>
I don't necessarily blame them for using Connect and defending it, since there's probably a lot of internal pressure to not ditch MS stuff in favor of open-source projects like Bugzilla.

[12:34:00.0000] <AryehGregor>
But they didn't have to act as though their bug tracker was somehow comparable to Mozilla's or WebKit's, and that they were awesome for closing everything quickly.

[14:20:00.0000] <Hixie>
Philip`: use complete.html to get xrefs :-)

[14:20:01.0000] <Hixie>
AryehGregor: i have no idea what APC is

[14:21:00.0000] <Hixie>
AryehGregor: happy to set stuff up though

[14:31:00.0000] <volkmar>
Hixie: i've some questions about the enumerated attributes

[14:32:00.0000] <volkmar>
Hixie: it looks like most enumerated attributes are not "limited to only known values"

[14:32:01.0000] <volkmar>
which sounds weird because it sounds quite important

[14:32:02.0000] <volkmar>
and even input.type isn't... and that is a regression

[14:33:00.0000] <volkmar>
i'm wondering why some enumerated attributes can be not "limited to only known values"

[15:04:00.0000] <Hixie>
volkmar: hmm

[15:04:01.0000] <jgraham>
The thing about the IEBlog is that it just makes them look childish. I'm really surprised that they are dumb enough to play to the front rows like that

[15:08:00.0000] <Hixie>
volkmar: yeah, looks like a bug, especially for input.type; can you file a bug?

[15:08:01.0000] <Hixie>
Philip`: which bugs was it that they ignored for IE8?

[15:09:00.0000] <jgraham>
Hixie: Is there some particular reason that you can't drawImage and <input type="image">?

[15:10:00.0000] <jgraham>
s/and/an/

[15:10:01.0000] <jgraham>
(I don't have a usecase or any particular desire for it to work but if there is a special reason it would be useful to know)

[15:11:00.0000] <Hixie>
"I don't have a usecase or any particular desire for it to work" is the reason

[15:13:00.0000] <jgraham>
Apparently it probably does work in Opera at the moment (I haven't actually tested) and "I don't have a usecase or any particular desire for it to work" isn't a strong argument for breaking it

[15:14:00.0000] <volkmar>
Hixie: there were a reason why some enumerated attributes could be reflected like regular strings ?

[15:17:00.0000] <cardona507>
hi - I am wondering what audio formats firefox, chrome, and safari support for <audio> - thanks

[15:22:00.0000] <sicking>
cardona507: i believe we support vorbis and wave

[15:22:01.0000] <cardona507>
thanks jonas

[15:26:00.0000] <cardona507>
I am pretty sure that webkit is mp3 and wav

[15:27:00.0000] <cardona507>
wav isn't the greates - is vorbis better?

[15:27:01.0000] <cardona507>
*greatest

[15:29:00.0000] <volkmar>
cardona507: afaik, wav is lossless and vorbis is lossy so they are not really similar

[15:29:01.0000] <cardona507>
gotcha

[15:54:00.0000] <Philip`>
cardona507: Vorbis is like MP3 but a bit better quality

[15:54:01.0000] <Philip`>
so they're the most similar ones

[15:54:02.0000] <cardona507>
Philip` thanks - I need a way to change my wav files to vorbis

[15:54:03.0000] <Philip`>
(I think Safari and IE9 support AAC too)

[15:54:04.0000] <cardona507>
wow - IE with html5 support

[15:54:05.0000] <cardona507>
I don't know what to say... :)

[16:12:00.0000] <cardona507>
html5 is trending as a google search term - It's #97 of the most popular search topics right now - Thanks Steve Jobs :)

[16:13:00.0000] <cardona507>
I think after his post today there will be an explosion of HTML5 adoption and developers


2010-04-30
[17:02:00.0000] <Hixie>
volkmar: some just are, dunno if there's a good reason for it, it's just what browsers do

[17:07:00.0000] <volkmar>
Hixie: then we agree there are no reasons for new attributes ?

[17:08:00.0000] <volkmar>
Hixie: btw, with smaug we were wondering how 'autocomplete' (for input element) should behave

[17:09:00.0000] <volkmar>
when autocomplete is not set, for example, input.autocomplete should return 'default' or the resulting autocompletion state ?

[17:13:00.0000] <Hixie>
volkmar: i'd have to look at each attribute to say for sure one way or the other -- feel free to file bugs for any you want me to consider

[17:13:01.0000] <Hixie>
volkmar: looking at input.autocomplete

[17:14:00.0000] <Hixie>
volkmar: .autocomplete just reflects the content attribute, so it's just whatever the content attribute's value is

[17:15:00.0000] <volkmar>
Hixie: so if the content attribute is 'foo', input.autocomplete should return 'foo' ? Why not 'on', 'off' or 'default' ?

[17:19:00.0000] <Hixie>
volkmar: "default" isn't a valid value, for one

[17:23:00.0000] <sicking>
volkmar: did you raise the issue of ignoring empty <option>s in <datalist>?

[17:24:00.0000] <volkmar>
sicking: i don't think i've open a bug for it at the moment

[17:32:00.0000] <sicking>
volkmar: ok

[17:34:00.0000] <volkmar>
sicking: i'm going to do that

[17:35:00.0000] <sicking>
volkmar: awesome, thanks

[17:36:00.0000] <TabAtkins>
Hmm, <details> doesn't show up in the list of implied ARIA sematnics.  Is ARIA too weak to express <details>, or is this an oversight?

[17:44:00.0000] <TabAtkins>
Hixie: ^^^

[17:44:01.0000] <Hixie>
no idea off hand

[17:44:02.0000] <TabAtkins>
kk

[17:49:00.0000] <roc>
yay, W3C supports Mercurial now

[17:58:00.0000] <AryehGregor>
Supports in what sense?

[17:58:01.0000] <TabAtkins>
In the sense that we have a mercurial server that WGs can use.

[17:59:00.0000] <kennyluck>
where? just curious

[17:59:01.0000] <TabAtkins>
The announcement was on the w3c blog.

[17:59:02.0000] <kennyluck>
/me googled but found nothing

[18:00:00.0000] <roc>
there was a post in public-webapps

[18:01:00.0000] <TabAtkins>
Ah, that's where it was.  I was certain I'd seen it in my twitter feed.

[18:03:00.0000] <remysharp>
Can anyone tell me what situation a SharedWorker would really be used? I understand more than one window can access it, but I can't see why

[18:03:01.0000] <kennyluck>
http://lists.w3.org/Archives/Public/public-webapps/2010AprJun/0389.html

[18:05:00.0000] <Hixie>
remysharp: consider e.g. gmail with multiple windows open, all interacting with the server

[18:06:00.0000] <Hixie>
remysharp: without a shared worker, they'd each be fighting with each other to update the local database view of the user's e-mail

[18:07:00.0000] <remysharp>
okay, is there another example? Just thinking outside of google for the moment. No meaning to be rude of course :)

[18:07:01.0000] <remysharp>
I can see and understand that use case tho

[18:07:02.0000] <Hixie>
any app that has client-side user data, multiple tabs/windows, and a server component

[18:07:03.0000] <remysharp>
except that none of the workers have implemented access to openDatabase - but I understand that's coming too

[18:08:00.0000] <Hixie>
openDatabase() is basically dead, actually, but there's a non-sql alternative coming up in theory

[18:08:01.0000] <Hixie>
web indexed db or something

[18:08:02.0000] <remysharp>
the indexeddb

[18:08:03.0000] <remysharp>
okay, yeah

[18:08:04.0000] <remysharp>
that spec - oh man - couldn't get my head around it yet. probably best to wait until there's something I can play with

[18:08:05.0000] <remysharp>
I know there was someone looking at a FF plugin

[18:09:00.0000] <Hixie>
facebook would be another example, so that, say, they could show you their IM client in all the facebook windows you have open, and have the messages routed to the right window, without having to have N connections to the server open

[18:09:01.0000] <Hixie>
yeah web indexed db went way over my head too

[18:09:02.0000] <kennyluck>
Hixie: do you have a pointer to the web indexeddb thing? very interesting

[18:09:03.0000] <remysharp>
okay - that's cool

[18:09:04.0000] <remysharp>
Shit, a spec that goes over Hixie's head - that's a spec I'm afraid of!

[18:10:00.0000] <dave_levin>
remysharp: fyi, soon db and workers will get along better in WebKit --  Add async bindings for Worker access to DB -- https://bugs.webkit.org/show_bug.cgi?id=34992

[18:10:01.0000] <erlehmann>
Steve Jobs anti-Flash tirade is quite funny: "Adobe’s Flash products are 100% proprietary.", then "Apple […] creates open standards for the web.", finally "a more modern format, H.264, […] viewable on iPhones, iPods and iPads"

[18:10:02.0000] <remysharp>
dave_levin: yeah - someone mentioned that here earlier on in the week - good to know, cheers.

[18:10:03.0000] <Hixie>
kennyluck: http://dev.w3.org/2006/webapi/WebSimpleDB/

[18:10:04.0000] <remysharp>
Hixie: I love that it has the word "simple" in the title!

[18:10:05.0000] <AryehGregor>
/me got a FOUC there, hasn't seen one of those in ages

[18:10:06.0000] <kennyluck>
thanks, Hixie

[18:10:07.0000] <othermaciej>
the SQL DB is currently much more widely implemented and much more widely used than IndexedDB

[18:11:00.0000] <othermaciej>
despite the fact that it is kind of stalled on the standards track due to MS and Moz not wanting to implement

[18:11:01.0000] <AryehGregor>
The problem with Web SQL DB is that it more or less says "use SQLite", doesn't it?

[18:11:02.0000] <othermaciej>
a *lot* of mobile-targeted sites use it

[18:11:03.0000] <AryehGregor>
Something like that?

[18:11:04.0000] <Hixie>
AryehGregor: we could fix that

[18:11:05.0000] <othermaciej>
that's what it says right now, yeah

[18:11:06.0000] <Hixie>
AryehGregor: that's not a real problem

[18:11:07.0000] <othermaciej>
but the implementors who refuse to implement it would not change their minds if that change was made

[18:11:08.0000] <Hixie>
we've specced html parser, how hard can sql be :-P

[18:12:00.0000] <othermaciej>
so it kinda seems like busywork to do it

[18:12:01.0000] <Hixie>
yeah

[18:12:02.0000] <AryehGregor>
Sad.

[18:12:03.0000] <othermaciej>
at least, that is what I concluded, otherwise I would make an Apple employee do it

[18:12:04.0000] <AryehGregor>
Why are they against it, again?

[18:13:00.0000] <othermaciej>
Mozilla says it's because the SQL dialect isn't spec'd, and also because they think SQL would be bad in general, even if spec'd

[18:14:00.0000] <othermaciej>
because some developers supposedly told them they want anything but SQL

[18:14:01.0000] <remysharp>
dave_levin: do you know if there's a fix planned for Workers starting (sub) Workers?

[18:14:02.0000] <remysharp>
so far, no browsers I've seen support this

[18:14:03.0000] <AryehGregor>
SQL is admittedly complicated and hard to understand.

[18:14:04.0000] <othermaciej>
Microsoft also said it's because the SQL dialect is de facto SQLite

[18:15:00.0000] <kennyluck>
Why are they opposing SQLite?

[18:15:01.0000] <othermaciej>
it is, but the alternative ends up being that you have to invent your own query language, and it is less efficient since it has to work on a lower-level substrate

[18:15:02.0000] <othermaciej>
depending on specifically the SQLite implementation, with no formal spec, I can see how that is kind of bad

[18:15:03.0000] <othermaciej>
since it makes truly independent implementations impractical

[18:15:04.0000] <AryehGregor>
othermaciej, are you aware that in the Wikipedia article on you, someone went to the trouble of recording a pronunciation of your name?

[18:15:05.0000] <AryehGregor>
/me wonders if it's at least correct

[18:15:06.0000] <othermaciej>
AryehGregor: I saw that

[18:15:07.0000] <dave_levin>
remysharp: I thought Firefox supported it. As far as WebKit, it was started but then aborted because it made the code more complicated and it was unclear how many folks wanted to use it.... If there are known use cases that it is blocking, then maybe the work would get restarted....

[18:16:00.0000] <othermaciej>
my wikipedia article is way too detailed

[18:16:01.0000] <remysharp>
dave_levin: crap, you're right, it does.

[18:16:02.0000] <AryehGregor>
/me was just stalking othermaciej by reading it

[18:16:03.0000] <remysharp>
/me so many browsers, so little time!

[18:16:04.0000] <AryehGregor>
/me now knows about his sister who's faculty at MIT, and exactly which Maciej forced othermaciej to be "othermaciej" instead of "maciej"

[18:17:00.0000] <AryehGregor>
/me isn't sure whether to hope he never has a Wikipedia article on him, or not.

[18:17:01.0000] <othermaciej>
the link to that faculty page is dead

[18:17:02.0000] <kennyluck>
lol

[18:17:03.0000] <othermaciej>
I think my sister is technically staff, not faculty

[18:18:00.0000] <AryehGregor>
You could propose your article for deletion on the grounds of non-notability.

[18:18:01.0000] <othermaciej>
AryehGregor: the Maciej who stole my preferred nick on FreeNode is Maciej Ceglowski

[18:18:02.0000] <othermaciej>
also mjs was taken by someone else, that's the other username I like to use

[18:18:03.0000] <AryehGregor>
So it says, except it puts a little line through the l.

[18:19:00.0000] <othermaciej>
yeah, I didn't feel like digging that character out of Character Pallette :-)

[18:19:01.0000] <Hixie>
wow, maciej's entry is long

[18:19:02.0000] <Hixie>
and has a way better photo of him than mine does of me :-P

[18:20:00.0000] <othermaciej>
I am embarassed about how much content it has

[18:20:01.0000] <othermaciej>
I think the IPA is wrong

[18:20:02.0000] <Philip`>
Are there guidelines against nominating yourself for deletion?

[18:20:03.0000] <Hixie>
on an unrelated note, i finally got to the part of my e-mail where i see Nick Levinson has escalated bugs to issues

[18:20:04.0000] <othermaciej>
though I don't know enough IPA to be sure

[18:20:05.0000] <Hixie>
that ought to be fun

[18:21:00.0000] <AryehGregor>
Hixie, you should upload a better photo of yourself, then!

[18:21:01.0000] <AryehGregor>
Your photo on Wikipedia does look pretty awful.

[18:21:02.0000] <AryehGregor>
Philip`, nope, people do it all the time.

[18:21:03.0000] <othermaciej>
the pronounciation file seems right-ish

[18:21:04.0000] <othermaciej>
AryehGregor: doesn't wikipedia frown on adding content to one's own article?

[18:22:00.0000] <AryehGregor>
Somewhat, but not if it's as clear-cut as adding a better-quality image.

[18:22:01.0000] <AryehGregor>
There's no hard policy, really.

[18:22:02.0000] <AryehGregor>
http://en.wikipedia.org/wiki/Wikipedia:COI

[18:22:03.0000] <Hixie>
AryehGregor: my partner tried to find a better picture of me once but it turns out i just look like a goofball :-P

[18:22:04.0000] <AryehGregor>
"This page in a nutshell: Do not edit Wikipedia to promote your own interests, or those of other individuals or of organizations, including employers, unless you are certain that a neutral editor would agree that your edits are in the best interest of Wikipedia."

[18:25:00.0000] <othermaciej>
Hixie: Nick Levinson has some... singular ideas

[18:25:01.0000] <Hixie>
AryehGregor: gotta be hard to find a "neutral editor" for some topics (like theism)

[18:26:00.0000] <AryehGregor>
Hixie, oh, then the various biased parties just fight it out and we hope a more or less moderate position emerges from the rubble.

[18:26:01.0000] <Hixie>
indeed

[18:26:02.0000] <AryehGregor>
Kind of the only thing you could do on any subject more interesting than . . . actually I can't think of a totally uncontroversial topic.

[18:26:03.0000] <AryehGregor>
I mean, like a Wikipedia page.

[18:27:00.0000] <AryehGregor>
The fact that the sky is often blue is uncontroversial, but you can't make an article out of that one fact.

[18:27:01.0000] <Hixie>
um

[18:27:02.0000] <Hixie>
i beg to differ

[18:27:03.0000] <Hixie>
http://en.wikipedia.org/wiki/Diffuse_sky_radiation

[18:28:00.0000] <othermaciej>
*why* the sky is blue is definitely interesting

[18:49:00.0000] <roc>
boy, I wish "canvas" wasn't so overloaded in the Web universe

[18:56:00.0000] <erlehmann>
roc, someone should make a html5 game called canvas. using canvas.

[19:03:00.0000] <remysharp>
The closing flag on a worker, I'm right in saying there's no way to get this in JS is there?

[19:03:01.0000] <remysharp>
i.e. if my worker dies, I can't logically work out that it's died using code

[19:03:02.0000] <CoJaBo-Aztec>
lolwiki

[19:04:00.0000] <Hixie>
remysharp: yeah we tried pretty hard to not expose any of that

[19:04:01.0000] <Hixie>
remysharp: (since it exposes garbage collection details)

[19:04:02.0000] <Hixie>
remysharp: how do you mean "die"?

[19:04:03.0000] <remysharp>
shame there's not a terminated event or something

[19:04:04.0000] <Hixie>
remysharp: you can get worker.onerror

[19:04:05.0000] <remysharp>
is that on the outside of the worker

[19:05:00.0000] <Hixie>
yeah

[19:05:01.0000] <remysharp>
hmm- then perhaps that's enough.

[19:05:02.0000] <TabAtkins>
Hixie: Any interest in helping me debug tircd and irssi?

[19:05:03.0000] <Hixie>
in non-error cases, the worker won't die unless you call close() iirc

[19:05:04.0000] <Hixie>
TabAtkins: sure, wassup?

[19:06:00.0000] <remysharp>
Hixie: if it's a sharedworker, is there any way for the SharedWorker to give heads up to all the connected ports on close?

[19:06:01.0000] <TabAtkins>
Installed tircd, got it started, no errors.  I can point a web browser at it (tircd barfs on the attempted GET, but still, it's *there*).  I can call /connect localhost 7000, and it says that it's connected, but then if I try any commands it claims I'm not connected to a network.

[19:07:00.0000] <TabAtkins>
That last "it" being irssi, obviously.

[19:07:01.0000] <remysharp>
Hixie: ignore me, the close is from within the worker, so it would send the message out before it manually closed itself.

[19:09:00.0000] <Hixie>
TabAtkins: hmm

[19:11:00.0000] <Hixie>
TabAtkins: i always just change the config file when i add a new network...

[19:12:00.0000] <TabAtkins>
Let me try just doing that.

[19:12:01.0000] <Hixie>
TabAtkins: do other irc clients work?

[19:12:02.0000] <Hixie>
oh also i think you have to give tircd a password maybe

[19:12:03.0000] <TabAtkins>
Argh.  Answer: I dunno, haven't tried them yet.

[19:13:00.0000] <Hixie>
an irc password, that is, as part of connecting

[19:13:01.0000] <Hixie>
see /help connect for the syntax

[19:13:02.0000] <Hixie>
dunno what the password is though

[19:13:03.0000] <Hixie>
probably is set in the tircd config

[19:14:00.0000] <remysharp>
okay, last question then I'll leave you all alone: importScripts - does this run synchronously? It's not clear from the spec, but if it "creates a script" that would imply async: http://www.whatwg.org/specs/web-workers/current-work/#importing-scripts-and-libraries

[19:15:00.0000] <dave_levin>
remysharp: Doesn't step #3 imply a sync execution?

[19:15:01.0000] <remysharp>
dave_levin: "If any fail, throw a SYNTAX_ERR exception" ?

[19:15:02.0000] <remysharp>
hmm - no

[19:15:03.0000] <dave_levin>
remysharp: sorry, ignore me

[19:15:04.0000] <remysharp>
because well

[19:15:05.0000] <remysharp>
...

[19:15:06.0000] <Hixie>
remysharp: "Create a script" is sync

[19:15:07.0000] <remysharp>
it's real late for me right now :-)

[19:15:08.0000] <Hixie>
remysharp: what makes you think it's not sync?

[19:16:00.0000] <remysharp>
Hixie: because creating a script in a normal doc loads the script async

[19:16:01.0000] <Hixie>
oh, you're looking at the workers spec

[19:16:02.0000] <Hixie>
look at complete.html 

[19:16:03.0000] <remysharp>
aaah.

[19:16:04.0000] <Hixie>
it links to what "create a script" means

[19:16:05.0000] <remysharp>
ah, cool.

[19:16:06.0000] <Hixie>
we really need cross-doc xrefs

[19:17:00.0000] <Hixie>
i wonder if gsnedders knows how much he will be reveered and how many songs will be written about him once he's done that

[19:18:00.0000] <Hixie>
remysharp: (Specifically, http://www.whatwg.org/specs/web-apps/current-work/complete.html#create-a-script )

[19:19:00.0000] <remysharp>
it's weird to read: "create an impotent script" feels like it should be a typo, but then it makes sense.

[19:19:01.0000] <remysharp>
but it still looks like a dick joke in the spec :)

[19:19:02.0000] <Hixie>
heh

[19:19:03.0000] <Hixie>
the one that people keep reporting as a typo is "outlinee"

[19:19:04.0000] <remysharp>
/me dick jokes in html5 specs - think it's time for bed :-\

[19:20:00.0000] <Hixie>
(i really should find a better way to express outlinee)

[19:20:01.0000] <Hixie>
remysharp: there are lots and lots of jokes in the spec... but not dick jokes :-P

[19:20:02.0000] <remysharp>
:)

[19:20:03.0000] <remysharp>
easter eggs hopefully!

[19:20:04.0000] <Hixie>
there really are a ton of easter eggs

[19:20:05.0000] <Hixie>
from the very obscure to the very obvious

[19:21:00.0000] <othermaciej>
maybe GuestXHR could be named ImportntRequest

[19:21:01.0000] <othermaciej>
er

[19:21:02.0000] <othermaciej>
ImpotentRequest

[19:21:03.0000] <remysharp>
I've honestly never seen any of them :(

[19:21:04.0000] <remysharp>
...or maybe I have...

[19:30:00.0000] <annevk>
jgraham is in Sweden :)

[19:30:01.0000] <annevk>
/me fixes wiki

[19:32:00.0000] <Hixie>
yeah i couldn't tell where on earth y'all were

[19:32:01.0000] <Hixie>
so i took a few guesses :-P

[19:43:00.0000] <annevk>
http://blogs.msdn.com/ie/archive/2010/04/29/html5-video.aspx -- "The future of the web is HTML5"

[19:43:01.0000] <annevk>
his remark about developers not having to pay royalty seems wrong though; pretty sure you have to for any commercial usage

[19:44:00.0000] <theMadness>
Yeah, that was quite insulting.

[19:44:01.0000] <theMadness>
I mean, the rethoric. MS saying "you don't have to pay because... because... hey, there's a bunny there!"

[19:45:00.0000] <theMadness>
Not to mention that no reason is given for a lack of double support a la chrome.

[19:46:00.0000] <theMadness>
Sorry for the rant, I just hate when people assumes that me (or actually us, readers) are too dumb to get it.

[19:47:00.0000] <wirepair>
interesting onmousemove in FF 3.6.3 only seems to fire when it's over the element... if that is the case why have an onmouseover event?

[19:47:01.0000] <roc>
it's actually highly misleading and dangerous for a Microsoft blog to say "Of course, developers can rely on the H.264 codec and hardware acceleration support of the underlying operating system, like Windows 7, without paying any additional royalty. "

[19:47:02.0000] <roc>
since the Windows 7 codec is only for non-commercial use

[19:48:00.0000] <theMadness>
Ditto.

[19:48:01.0000] <jcranmer>
/me prays for bilski

[19:48:02.0000] <jcranmer>
(for SCotUS to reject all software patents on the spot)

[19:49:00.0000] <theMadness>
/me rtfms http://endsoftpatents.org/bilski

[19:49:01.0000] <gavin>
wirepair: doesn't onmouseover fire only once?

[19:51:00.0000] <wirepair>
sometimes once or twice heh

[19:51:01.0000] <theMadness>
Nice, there's no mission statement or anything.

[19:51:02.0000] <wirepair>
actually looks like twice

[19:52:00.0000] <gavin>
(for a given "mouseover and move", I mean)

[19:52:01.0000] <wirepair>
ah

[19:55:00.0000] <wirepair>
nevermind looks like all browsers do that

[19:55:01.0000] <wirepair>
weird i thought it was while the mouse was over the document, not the element that it would fire

[19:56:00.0000] <wirepair>
/me makes note

[20:06:00.0000] <erlehmann>
That is some really bad hand-waving. Microsoft should probably try to get a PR guy for IE with a bit more FUD talent.

[20:07:00.0000] <erlehmann>
roc, interesting. so if you sell an app, it cannot hook legally into windows 7 h.264 decoder ?

[20:08:00.0000] <roc>
that is unclear

[20:08:01.0000] <roc>
read the terms yourself: http://download.microsoft.com/Documents/UseTerms/Windows%207_Ultimate_English_c44ca3df-8338-4a2f-a176-39d2e68986c4.pdf

[20:08:02.0000] <roc>
I presume it's there so that apps can hook into it

[20:08:03.0000] <roc>
but the language about OTHER PRODUCTS is confusing

[20:09:00.0000] <roc>
NONE OF THE LICENSES EXTEND TO ANY OTHER PRODUCT REGARDLESS OF WHETHER SUCH PRODUCT IS INCLUDED WITH THIS PRODUCT IN A SINGLE ARTICLE.

[20:20:00.0000] <roc>
haha

[20:20:01.0000] <roc>
this is rich

[20:21:00.0000] <roc>
MPEG-LA suing Apple

[20:21:01.0000] <roc>
http://thepriorart.typepad.com/the_prior_art/2010/04/mobilemedia-ideas-v-apple.html

[20:23:00.0000] <Hixie>
"it let the world know that it would "make available" its innovations, (which is, of course, generally considered to be patent-speak for, "We will be suing people shortly")" lol

[20:23:01.0000] <erlehmann>
roc, would be pretty ironic. lulz may ensue.

[20:24:00.0000] <roc>
it's thin humour

[23:11:00.0000] <collinjackson>
whatwg.org is not responding for me

[23:28:00.0000] <collinjackson>
Hixie: I tried from several IPs, no response from whatwg.org (69.163.222.251)

[23:46:00.0000] <annevk>
aah, roc is already up on Ajaxian :)

[23:46:01.0000] <annevk>
lot can happen over lunch

[00:06:00.0000] <annevk>
ImpotentRequest is too hard to spell

[00:06:01.0000] <annevk>
imo

[00:06:02.0000] <annevk>
and it's also unclear

[00:06:03.0000] <annevk>
but please keep trying!

[00:06:04.0000] <othermaciej>
annevk: wasn't a serious suggestion

[00:06:05.0000] <annevk>
I'm not exactly in a rush here :)

[00:07:00.0000] <othermaciej>
annevk: I'd actually really like it if the context-free / drop-privileges / whatever mode had a defined API

[00:07:01.0000] <othermaciej>
we would want to implement it in WebKit

[00:07:02.0000] <annevk>
either Anon or Guest is still my favorite

[00:07:03.0000] <annevk>
really?

[00:07:04.0000] <annevk>
surprises me a little to be honest

[00:07:05.0000] <othermaciej>
why?

[00:07:06.0000] <othermaciej>
I think I've always said we'd like to implement the no-credentials mode of CORS

[00:07:07.0000] <othermaciej>
only blocker has been lack of a defined API for it

[00:08:00.0000] <annevk>
because WebKit actually gave quite a bit of pushback to having things like withCredentials

[00:08:01.0000] <annevk>
well, ap did (and I still agree with him that it was prolly a dumb idea to do it this way)

[00:08:02.0000] <othermaciej>
I think the ability to drop privileges is useful

[00:09:00.0000] <othermaciej>
that's how I think of this, anyway - as a tool for privilege sepration

[00:09:01.0000] <othermaciej>
though I guess if you want to go hardcore on that, you can run the untrusted part of your site code in a sandboxed iframe

[00:09:02.0000] <othermaciej>
I agree the withCredentials mode is inelegant

[00:09:03.0000] <othermaciej>
and having three modes will also kinda suck

[00:10:00.0000] <annevk>
in retrospect we should've had this anon API and the one that works just like <img> <form> etc.

[00:10:01.0000] <annevk>
without the Mozilla compromise

[00:10:02.0000] <othermaciej>
you mean, sends full credentials always?

[00:11:00.0000] <othermaciej>
in retrospect, yes, it would be better if those were the only two modes

[00:11:01.0000] <annevk>
yeah, not change the defaults for cross-origin

[00:11:02.0000] <annevk>
maybe we can phase out withCredentials somehow

[00:11:03.0000] <annevk>
not implemented by IE/Opera after all

[00:12:00.0000] <annevk>
oh hey, maybe I'll try out to install this Ubuntu release

[00:13:00.0000] <annevk>
see how much will burn :)

[00:13:01.0000] <othermaciej>
it seems like particularly bad design that XHR has different behavior cross-origin and same-origin (besides just the actual access control issues)

[00:13:02.0000] <othermaciej>
is it allowed to turn off withCredentials for a same-origin request?

[00:13:03.0000] <annevk>
it has no effect at the moment

[00:13:04.0000] <annevk>
didn't see worth it

[00:14:00.0000] <annevk>
i should've brought a book :/

[00:20:00.0000] <Hixie>
someone wants to try writing a rationale explanation for some simple part of html5 to see if they like doing them... anyone got a suggestion of a good thing to start with?

[00:21:00.0000] <annevk>
why <image> ends up as <img> ?

[00:22:00.0000] <annevk>
(whenever I talk about HTML5 history I find it somewhat funny to bring that up)

[00:23:00.0000] <annevk>
(shocks the crowd :p )

[00:23:01.0000] <Hixie>
k :-)

[00:24:00.0000] <zcorpan>
and <plaintext>

[00:56:00.0000] <annevk>
http://twitter.com/joehewitt/status/13091031807 ?

[00:57:00.0000] <annevk>
/me wonders what that is about; apparently retweeted by cwilso

[00:57:01.0000] <othermaciej>
annevk: joehewitt has been tweeting a bunch on that theme

[00:57:02.0000] <othermaciej>
he thinks browsers should innovate outside of standards more

[00:57:03.0000] <othermaciej>
each does its own thing, and then whatever becomes popular maybe becomes a standard eventually

[00:58:00.0000] <othermaciej>
he said only WebKit has done enough of this, and some Mozilla guy replied and bragged about how much they ignored the standards process

[00:59:00.0000] <othermaciej>
whereas my response was http://twitter.com/othermaciej/status/13106723840

[00:59:01.0000] <othermaciej>
which I hope is mostly true

[00:59:02.0000] <zcorpan>
i ignore standards more!

[01:00:00.0000] <annevk>
it seems he forgot it took about 10 years to clean up the mess IE made

[01:01:00.0000] <annevk>
having implementations of new features is good of course, but having a wider group discussing them before everything is finalized helps a lot I think

[01:01:01.0000] <othermaciej>
that's what I think too

[01:02:00.0000] <othermaciej>
although sometimes you have to just implement something to be able to show why it has value

[01:02:01.0000] <annevk>
yeah, and not everything works as well as the WHATWG

[01:03:00.0000] <annevk>
or the WebApps WG for that matter

[01:03:01.0000] <annevk>
CSS for instance...

[01:04:00.0000] <jgraham>
His idea of forcing peopleto open different browsers based on which website they want to use seems pretty unlikely to work

[01:04:01.0000] <othermaciej>
the Web Apps WG hasn't managed to finalize a whole lot of anything yet

[01:05:00.0000] <othermaciej>
but we do at least manage to write down specs for useful features without huge flamewars

[01:05:01.0000] <Hixie>
i think there are definitely areas where we've standardised maybe a little too soon

[01:05:02.0000] <Hixie>
storage for example

[01:06:00.0000] <othermaciej>
with storage the problem may have been more about implementing too soon

[01:06:01.0000] <othermaciej>
if those standards didn't have implementations and sites using them, they wouldn't be a problem

[01:06:02.0000] <annevk>
if we knew about cookies we could have avoided the storage issue

[01:07:00.0000] <annevk>
shows that documenting the past is important for future decisions

[01:07:01.0000] <othermaciej>
the lameness of cookies was a big driver for making something better

[01:07:02.0000] <annevk>
I'm mostly talking about the storage mutex, something we realized was needed too late

[01:08:00.0000] <othermaciej>
I'm not sure studying cookies would have revealed that we needed it

[01:08:01.0000] <othermaciej>
since it's essentially driven by non-single-threadedness

[01:09:00.0000] <annevk>
true

[01:09:01.0000] <othermaciej>
I kinda wish we hadn't jumped on implementing SQL so soon, but there wasn't really a better alternative and I'm still not sure there is one now

[01:09:02.0000] <annevk>
is the SQL API that bad?

[01:09:03.0000] <annevk>
sicking claims developers don't like it, but I haven't had that reaction much

[01:09:04.0000] <othermaciej>
a surprising number of sites are using it now in their mobile versions (GMail, Yahoo Mail, foxnews.com surprisingly...)

[01:10:00.0000] <othermaciej>
the API is not bad

[01:10:01.0000] <annevk>
only from CouchDB guys really

[01:10:02.0000] <othermaciej>
the claim that developers don't like it is BS

[01:10:03.0000] <othermaciej>
people who are on an anti-SQL crusade don't like it, and that is a minority of web developers

[01:10:04.0000] <othermaciej>
people actually developing web content that works offline like it fine

[01:10:05.0000] <othermaciej>
like I said, it has lots of use in the one niche where it has high penetration

[01:11:00.0000] <annevk>
I wonder if Mozilla's Indexed DB implementation is going on top of SQLLite

[01:11:01.0000] <annevk>
SQLite*

[01:11:02.0000] <othermaciej>
the reason I regret it is that Mozilla and Microsoft are potentially limiting its utility

[01:11:03.0000] <othermaciej>
I'm not sure it would make sense to implement Indexed DB on top of SQLite

[01:12:00.0000] <annevk>
the reason I'm wondering is that Mozilla already ships with SQLite

[01:12:01.0000] <othermaciej>
it really needs a more Berkeley-DB style substrate (though I am not sure BDB itself would actually work)

[01:12:02.0000] <othermaciej>
Google folks seem interested in it, so WebKit may end up with the first available implementation

[01:12:03.0000] <othermaciej>
which would be really really weird

[01:13:00.0000] <othermaciej>
since Indexed DB seems to be primarily a reaction against SQL DB

[01:13:01.0000] <annevk>
heh yeah, I saw that

[01:14:00.0000] <annevk>
so much redundant APIs :/

[01:14:01.0000] <othermaciej>
basically my opinion on the matter is similar to this: http://teddziuba.com/2010/03/i-cant-wait-for-nosql-to-die.html

[01:14:02.0000] <othermaciej>
but I don't feel like arguing with MS and Moz

[01:15:00.0000] <othermaciej>
maybe they will change their minds if they ever decide to be relevant on mobile

[01:15:01.0000] <Hixie>
othermaciej: implementations and specs have to happen at the same time

[01:16:00.0000] <othermaciej>
Hixie: in the case of the SQL Storage spec, that is kinda what happened, except for the plural part of "implementations"

[01:16:01.0000] <Hixie>
othermaciej: i think what we should have done is had more experimentation first, we clearly didn't have enough implementation experience since in the space of like a year we came up with three radically different suggestions and specced all three

[01:17:00.0000] <Hixie>
and within another year people were already suggesting more (filesystem)

[01:17:01.0000] <othermaciej>
the real problem is that those things are implemented and used (and thus hard to change), not that it was written down in the form of a spec

[01:17:02.0000] <Hixie>
if they had all just been UA experiments i don't think they'd have been used as much

[01:17:03.0000] <othermaciej>
it's easier to rescind a spec than an implementation

[01:18:00.0000] <othermaciej>
people would have used whatever they could get to make offline web apps work on iPhone OS devices

[01:18:01.0000] <othermaciej>
even if it was WebKitEvilHackIfYouUseThisYouAreDumbStorage

[01:18:02.0000] <Hixie>
hm, yes, offline, make that five solutions in the one space :-)

[01:18:03.0000] <Hixie>
oh which i specced three :-/

[01:18:04.0000] <othermaciej>
appcache isn't really an alternative to the others

[01:19:00.0000] <Hixie>
not as designed, no

[01:19:01.0000] <othermaciej>
(though I have lost count of which specs you have in mind)

[01:19:02.0000] <Hixie>
webstorage, websqldb, indexeddb, appcache, and filesystem

[01:19:03.0000] <Hixie>
they're all about storing data on the client

[01:19:04.0000] <Hixie>
one could easily imagine solutions that sliced and diced the space differently

[01:19:05.0000] <othermaciej>
and of course we have the legacy technology of cookies

[01:20:00.0000] <othermaciej>
maybe Gears gave the impression of being more proven a design than it was

[01:20:01.0000] <Hixie>
and i feel that at least the ones i was involved with of those 5 were all a little rushed, e.g. appcache isn't exactly ideal, web storage of course has the dreaded storage mutex, and websqldb died

[01:20:02.0000] <othermaciej>
there's also still the DataCache proposal, though I think those capabilities should just be folded into AppCache

[01:20:03.0000] <Hixie>
datacache doesn't count imho because it doesn't have implementations, so it's not a real spec

[01:21:00.0000] <othermaciej>
I bet 2-3 years from now websqldb will have more use on real sites than indexeddb

[01:21:01.0000] <othermaciej>
so it has only "died" in a very technical standards-process sense

[01:21:02.0000] <Hixie>
it's died in the sense that the two biggest UAs aren't implementing it

[01:22:00.0000] <othermaciej>
once mobile devices are more browsing volume than desktop that won't really matter any more

[01:22:01.0000] <othermaciej>
(not that I expect that to happen in as little as 3 years)

[01:25:00.0000] <annevk>
if we spec the SQL dialect Mozilla would be more interested

[01:27:00.0000] <othermaciej>
if that were true, it would be worth speccing it

[01:27:01.0000] <othermaciej>
though Mozilla claims that is not the case

[01:28:00.0000] <Hixie>
/me grumbles about the new weird styles in the change proposal status page while he looks for the new issues

[01:28:01.0000] <othermaciej>
Hixie: what weird styles?

[01:28:02.0000] <othermaciej>
whoah, it has weird styles

[01:28:03.0000] <othermaciej>
/me wonders who did that

[01:28:04.0000] <Hixie>
wait that wasn't you?

[01:28:05.0000] <Hixie>
i assumed it was you

[01:29:00.0000] <Hixie>
it's all rounded corners and fancy :nth-child stuff

[01:29:01.0000] <othermaciej>
it was Sam

[01:29:02.0000] <othermaciej>
coloring every third row is weird

[01:30:00.0000] <annevk>
othermaciej, I think they also wanted an independent impl of SQLite, but it seemed less important

[01:30:01.0000] <othermaciej>
every second, I could understand (that might actually help legibility)

[01:30:02.0000] <Hixie>
the rounded corners are very 2000s

[01:30:03.0000] <othermaciej>
the mouseover effects on links make the columns shift

[01:30:04.0000] <othermaciej>
Hixie: true, clearly it needs drop shadows, reflections and gradients

[01:31:00.0000] <annevk>
SVG filters!

[01:36:00.0000] <zcorpan>
which page?

[01:37:00.0000] <zcorpan>
found it

[01:47:00.0000] <jgraham>
That gives the misleading impression that some rows are being highlighted for some reason

[01:48:00.0000] <othermaciej>
jgraham: alternate rows would be more conventional

[01:48:01.0000] <othermaciej>
but also that background color is a bit too close tothe link color

[01:48:02.0000] <othermaciej>
too tired to tweak it right now though

[01:48:03.0000] <othermaciej>
see y'all tomorrow

[01:50:00.0000] <jgraham>
Well yes, a subtle colour would be more conventional too

[01:50:01.0000] <jgraham>
Like a light grey

[01:54:00.0000] <wycats>
Hixie: I'm writing up a response to Sachin's article and ran into an article you wrote on <canvas> in 2004

[01:54:01.0000] <wycats>
kind of funny

[01:55:00.0000] <wycats>
the entire argument about browsers failing to innovate because of the standards process is just demonstrably false

[01:55:01.0000] <wycats>
it's almost the web equivalent of death panels

[02:00:00.0000] <boblet>
hey all, anyone used display:table-header-group in an inline-table with Chrome? doesn’t seem to be working

[02:09:00.0000] <othermaciej>
wycats: with Safari/WebKit almost all our innovation used to be outside the standards groups, I think WHATWG was by far the major thing that reversed that

[02:09:01.0000] <othermaciej>
now we try to take stuff to standards groups as early as possible at least for discussion, if not immediate standardization

[02:10:00.0000] <wycats>
othermaciej: even today, it's not like any browser is sitting around waiting for standards bodies to approve things

[02:10:01.0000] <othermaciej>
it does seem like bigger features often need a proof of concept

[02:10:02.0000] <othermaciej>
wycats: well maybe IE...

[02:10:03.0000] <wycats>
haha

[02:10:04.0000] <wycats>
even them ;)

[02:10:05.0000] <wycats>
more often, there's partially implemented features that form the basis for WHAT-WG discussion

[02:10:06.0000] <wycats>
(from an outsider's perspective)

[02:10:07.0000] <annevk>
IE did XDomainRequest and a few other things

[02:10:08.0000] <wycats>
yep

[02:10:09.0000] <wycats>
XDR is what comes to mind

[02:10:10.0000] <wycats>
where did CORS come from?

[02:11:00.0000] <othermaciej>
true

[02:11:01.0000] <othermaciej>
CORS was sorta designed by committee

[02:11:02.0000] <wycats>
ha ;)

[02:11:03.0000] <othermaciej>
back when it was access-control

[02:11:04.0000] <othermaciej>
and then redesigned by committee to get browser vendors to agree

[02:11:05.0000] <annevk>
CORS was mostly based on input from Mozilla

[02:11:06.0000] <annevk>
regrettably and much to my fault I guess

[02:12:00.0000] <othermaciej>
was the original access-control spec based on their ideas?

[02:12:01.0000] <annevk>
I just wanted to see it implemented which given where we are today should not have been a very high priority

[02:12:02.0000] <annevk>
no, CORS was

[02:12:03.0000] <wycats>
I made a pretty big list of external ideas: canvas, <input type="search">, <video>, flexbox, JS 1.8, CSS transitions/animations/transforms, WebGL, Local Storage, SQLite, font-face, NodeList, Touch events, orientation, accelerometer, geolocation

[02:12:04.0000] <othermaciej>
I am glad we got rid of the processing instruction at least

[02:12:05.0000] <wycats>
othermaciej: amen brother

[02:12:06.0000] <annevk>
the original access-control stuff came from VoiceXML plus some stuff Hixie drafted

[02:12:07.0000] <othermaciej>
I don't think anyone implemented <video> before it went in the spec

[02:13:00.0000] <annevk>
Opera did

[02:13:01.0000] <othermaciej>
though Opera and Apple both made fairly extensive proposals

[02:13:02.0000] <annevk>
we had  a build supporting <video> and some simple methods

[02:13:03.0000] <wycats>
Opera claims to have

[02:13:04.0000] <annevk>
what do you mean claims?

[02:13:05.0000] <wycats>
I mean I wasn't around back then

[02:13:06.0000] <wycats>
but their site has good evidence that they did

[02:13:07.0000] <wycats>
primary sources

[02:14:00.0000] <annevk>
http://labs.opera.com/ -- search for "A call for video on the web"

[02:14:01.0000] <annevk>
also http://people.opera.com/howcome/2007/video/op950_8762_video_repack_with_lgpl_signed.exe

[02:15:00.0000] <doublec>
Opera definitely did, I remember trying it.

[02:15:01.0000] <annevk>
I also proposed <video> on the WHATWG list

[02:15:02.0000] <doublec>
yep

[02:15:03.0000] <annevk>
after howcome convinced me that <object> was not the way to go

[02:15:04.0000] <doublec>
if <video> had kept that api all the implementations would be finished by now :)

[02:15:05.0000] <othermaciej>
I think the original Opera <video> proposal was here: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-February/009702.html

[02:15:06.0000] <annevk>
took some XHTML2 extrapolation argument iirc :)

[02:15:07.0000] <wycats>
annevk: yeh I found your post

[02:16:00.0000] <wycats>
annevk: this is in my outline:

[02:16:01.0000] <wycats>
<video>

[02:16:02.0000] <wycats>
* Originally by Opera (Feb. 2007) 

[02:16:03.0000] <wycats>
* http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-February/009702.html

[02:16:04.0000] <wycats>
* http://people.opera.com/howcome/2007/video/

[02:16:05.0000] <wycats>
I have the originator of all the items on that list

[02:16:06.0000] <boblet>
IE5 did ruby proof of concept 3 years before W3 specced it

[02:16:07.0000] <wycats>
ruby w00t

[02:16:08.0000] <othermaciej>
Apple's more elaborate proposal here: http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2007-March/010386.html

[02:16:09.0000] <wycats>
if you include IE4-6 there's a list about double the size

[02:16:10.0000] <boblet>
annevk: still in Japan or back home?

[02:17:00.0000] <annevk>
still here

[02:17:01.0000] <annevk>
at a Starbucks in Ebisu with Mike

[02:17:02.0000] <boblet>
wycats: back when they were trying eh :|

[02:17:03.0000] <wycats>
IE4 had data bindings ;)

[02:17:04.0000] <wycats>
hardcore

[02:17:05.0000] <boblet>
heh, nice

[02:17:06.0000] <boblet>
say hi to Mike for me. hope things are going ok for him

[02:18:00.0000] <wycats>
is there anything else that comes to mind?

[02:18:01.0000] <wycats>
it looks like ARIA was designed by committee but came out of real concerns

[02:18:02.0000] <othermaciej>
it looks like <video> first went in the spec sometime before Apple's proposal in March, based on what I wrote

[02:19:00.0000] <wycats>
yah

[02:19:01.0000] <othermaciej>
the opera video release was not til November of that year

[02:19:02.0000] <wycats>
"We started work on these documents before the <video> element was added to the spec and indeed before Opera made their original proposal"

[02:20:00.0000] <wycats>
it seems clear that there was "real" effort on both Apple's and Opera's part

[02:20:01.0000] <othermaciej>
I was just trying to figure out spec timing relative to Opera's experimental implementation

[02:20:02.0000] <boblet>
looks like Chrome’s implementation of table display CSS never made it to inline-table, unlike Webkit or Safari. strange

[02:20:03.0000] <wycats>
weird

[02:21:00.0000] <wycats>
I'm going to say in my post that sachin's argument is like the death panels of web standards ;)

[02:21:01.0000] <wycats>
it's completely made up, transparently false, and yet a ton of people believe it

[02:21:02.0000] <boblet>
wycats: link?

[02:21:03.0000] <othermaciej>
I'm surprised Hewitt made the argument for ignoring standards more

[02:22:00.0000] <othermaciej>
and that in response Mozilla bragged about how much they ignore standards

[02:22:01.0000] <wycats>
http://sachin.posterous.com/the-web-sucks

[02:22:02.0000] <boblet>
nice positive title :)

[02:22:03.0000] <othermaciej>
there are many false statements there

[02:22:04.0000] <othermaciej>
"Browsers aren't innovating."

[02:22:05.0000] <wycats>
othermaciej: to put it bluntly, I agree with the argument

[02:22:06.0000] <othermaciej>
"Browsers are forced to implement every "standard" that is agreed on, even if it's not the best decision for the platform."

[02:23:00.0000] <wycats>
but it's wrong on the facts

[02:23:01.0000] <wycats>
I agree that browsers shouldn't wait for standards to finish before implementing

[02:23:02.0000] <wycats>
but they DON'T

[02:23:03.0000] <othermaciej>
wycats: I don't think we need to ignore standards more because they are not really an impediment to innovation

[02:23:04.0000] <wycats>
otherwise there would be no HTML5 support

[02:23:05.0000] <wycats>
othermaciej: it's not about ignoring standards

[02:23:06.0000] <wycats>
"Web applications don't have threading, GPU acceleration, drag and drop, copy and paste of rich media, true offline access, or persistence"

[02:23:07.0000] <wycats>
total utter bullshit

[02:24:00.0000] <wycats>
WebWorkers, WebGL, HTML5 D&D

[02:24:01.0000] <wycats>
CSS animations, transitions

[02:24:02.0000] <wycats>
HTML5 offline

[02:24:03.0000] <wycats>
Web Storage

[02:24:04.0000] <wycats>
total bullshit

[02:24:05.0000] <zcorpan>
ie9 has shown that gpu acceleration is possible with web 2.0 style apps

[02:24:06.0000] <othermaciej>
see Joe Hewitt's recent tweets though http://twitter.com/joehewitt/

[02:24:07.0000] <wycats>
othermaciej: yeah

[02:24:08.0000] <wycats>
he's wrong too

[02:25:00.0000] <othermaciej>
he definitely did argue that browsers should do more nonstandard stuff, and web developers, should make more browser-specific sites

[02:25:01.0000] <annevk>
othermaciej, our initial video release was in April actually

[02:25:02.0000] <annevk>
othermaciej, public release that is, it's the .exe I pointed to above

[02:25:03.0000] <wycats>
othermaciej: I make browser specific sites

[02:25:04.0000] <annevk>
othermaciej, but we demonstrated it publicly before that; there was just not a public build

[02:25:05.0000] <wycats>
I'm currently working on a webkit-only enhanced version of a mobile site I maintain

[02:26:00.0000] <annevk>
othermaciej, would be nice to know where Mozilla is ignoring standards

[02:26:01.0000] <annevk>
every now and then I notice some feature going in and try to encourage people to bring it up on some list, but I don't see it too often

[02:27:00.0000] <annevk>
(though there's also not much follow-up on those suggestions usually...)

[02:27:01.0000] <wycats>
annevk: is there new accelerometer support a standard?

[02:27:02.0000] <othermaciej>
wycats: in mobile people are more willing to do it because advanced smartphones are too much of a WebKit monoculture

[02:27:03.0000] <wycats>
othermaciej: I'm keeping around the old client-server model

[02:27:04.0000] <annevk>
wycats, not sure, not really up to speed with those APIs

[02:27:05.0000] <wycats>
but I'm leveraging everything Webkit on mobile gives me to the hilt if you're on a webkit browser

[02:28:00.0000] <wycats>
so old blackberries can still see the site

[02:28:01.0000] <othermaciej>
I wish there were more decent phones out with non-WebKit browsers shipping by default

[02:28:02.0000] <othermaciej>
monoculture is not healthy for the Web

[02:28:03.0000] <wycats>
othermaciej: if fennec was shipping, I'd make sure I supported it

[02:28:04.0000] <Rik`>
I wish there were more ways to test mobile browsers

[02:28:05.0000] <wycats>
I hage no problem supporting anything but IE mobile

[02:28:06.0000] <othermaciej>
wycats: didn't you hear? the pre-alpha is out

[02:29:00.0000] <wycats>
Rik`: jQuery's working on stuff with testswarm

[02:29:01.0000] <wycats>
othermaciej: LoLz

[02:29:02.0000] <zcorpan>
"IE9 will support playback of H.264 video only" http://blogs.msdn.com/ie/archive/2010/04/29/html5-video.aspx

[02:29:03.0000] <othermaciej>
Fennec is pretty far from being a real product still

[02:29:04.0000] <wycats>
"We're about to find out: see the growth and fragmentation of mobile app stores, which exist only because the web blows."

[02:29:05.0000] <wycats>
^^ bullshit

[02:29:06.0000] <wycats>
othermaciej: indeed

[02:29:07.0000] <othermaciej>
and I doubt any handset manufacturers will bundle it

[02:29:08.0000] <wycats>
"And really, how screwed would we be if the WebKit team weren't so god damn competent? Ok, signing off now, thanks for listening. :)" <= where "the webkit team" is about 10 companies

[02:29:09.0000] <annevk>
zcorpan, old news :)

[02:29:10.0000] <wycats>
othermaciej: who knows... maybe Apple will let them ship it ;)

[02:30:00.0000] <annevk>
zcorpan, see logs for some comments

[02:30:01.0000] <othermaciej>
wycats: most of the core work is done by 2 companies, and before very recently only 1...

[02:30:02.0000] <Rik`>
wycats: last time I checked, testswarm was only for JS lib

[02:30:03.0000] <wycats>
othermaciej: absolutely

[02:30:04.0000] <wycats>
othermaciej: not that recently

[02:30:05.0000] <wycats>
Google's been contributing for quite some time now

[02:30:06.0000] <wycats>
"As someone who has tried to do both cutting edge native and web iPhone apps, iPhone Safari is a joke compared to iPhone Cocoa." <= he tried it 2 years ago and freely admits he hasn't really looked since

[02:30:07.0000] <othermaciej>
yes, but for a while all their contribution was mostly porting work

[02:30:08.0000] <annevk>
still having a single engine is bad

[02:31:00.0000] <othermaciej>
not work to enhance the web platform

[02:31:01.0000] <annevk>
not sure it matters much how many companies develop it

[02:31:02.0000] <wycats>
annevk, othermaciej: bad for whom?

[02:31:03.0000] <annevk>
wycats, for the web ecosystem

[02:31:04.0000] <wycats>
annevk: because?

[02:31:05.0000] <othermaciej>
it's good for mobile web developers today

[02:31:06.0000] <annevk>
you need competition on implementation, speed, etc.

[02:31:07.0000] <othermaciej>
and it's good for the vendors today

[02:31:08.0000] <wycats>
annevk: having extremely incompetent alternatives with a lot of market share would blow

[02:32:00.0000] <othermaciej>
in the long run, it could be bad for everyone

[02:32:01.0000] <wycats>
it sounds like you'd like to see IE mobile pick up market share

[02:32:02.0000] <wycats>
I'd kill myself :P

[02:32:03.0000] <othermaciej>
look at how painful it is for IE to converge with standards because they spent so long as a monoculture

[02:32:04.0000] <wycats>
othermaciej: bad analogy

[02:32:05.0000] <othermaciej>
I'd like to see it get better enough that it deserves some market share

[02:32:06.0000] <annevk>
wycats, i'm saying WebKit could end up as IE6 if something happens

[02:32:07.0000] <wycats>
annevk: bullshit argument

[02:32:08.0000] <mut>
boo. 

[02:33:00.0000] <othermaciej>
iPhone web apps are much more likely to depend on WebKit bugs than desktop web apps

[02:33:01.0000] <othermaciej>
it has not yet gotten to the point where it is a huge burden though

[02:33:02.0000] <mut>
im attempting to convert something ive drawn in canvas to a dxf (ascii cad file)

[02:33:03.0000] <wycats>
othermaciej, annevk: I would love to see a webkit-caliber alternative emerge 

[02:33:04.0000] <annevk>
wycats, I don't see why; Microsoft invested hugely in the Web and when they became dominent they just let it go for a decade

[02:33:05.0000] <annevk>
wycats, they had a killer engine with IE6

[02:33:06.0000] <mut>
i just wondered if anyone knew of a project that could help :P

[02:33:07.0000] <wycats>
annevk: and it was closed source, so MS was *able* to do it

[02:33:08.0000] <annevk>
but no competition

[02:34:00.0000] <wycats>
in this case, if Apple gave up on it, Google would keep it going

[02:34:01.0000] <wycats>
or vice versa

[02:34:02.0000] <wycats>
a bunch of companies are invested in Webkit

[02:34:03.0000] <wycats>
unlike closed-source IE

[02:34:04.0000] <wycats>
in fact, if IE was OSS, I bet we'd have seen improvements

[02:34:05.0000] <wycats>
instead of stagnation

[02:34:06.0000] <wycats>
see: Mozilla

[02:34:07.0000] <annevk>
Mozilla and WebKit is actually a good example of why open source also needs competition

[02:35:00.0000] <wycats>
annevk: I'd love to see a webkit caliber competitor

[02:35:01.0000] <wycats>
but if IE mobile gets market share without being quality, I will cry

[02:35:02.0000] <annevk>
I think we are

[02:35:03.0000] <annevk>
but I'm not really sure what you mean

[02:35:04.0000] <wycats>
annevk: is it really unclear?

[02:35:05.0000] <othermaciej>
I doubt IE mobile will go much of anywhere

[02:36:00.0000] <othermaciej>
certainly not by usage share

[02:36:01.0000] <annevk>
all I'm saying is that software-monoculture is bad, even if it's open source

[02:36:02.0000] <wycats>
"monoculture is bad" is a meta-argument

[02:36:03.0000] <annevk>
wycats, yes

[02:36:04.0000] <wycats>
a culture of one awesome browser beats a culture of one awesome browser and 5 crappy ones

[02:36:05.0000] <othermaciej>
since on mobile, how much people use the browser is extremely sensitive to differences in browser quality

[02:37:00.0000] <annevk>
wycats, whatever man

[02:37:01.0000] <wycats>
annevk: what do you mean whatever?

[02:37:02.0000] <wycats>
I'm postulating a scenario where IE mobile gets market share (let's say via Windows Mobile 7 actually getting a bunch of sales)

[02:37:03.0000] <wycats>
we'd lose the monoculture

[02:38:00.0000] <wycats>
and gain having to deal with IE6(7) on the phone

[02:38:01.0000] <wycats>
if you think that's a good tradeoff...

[02:38:02.0000] <annevk>
I'm not sure why you think that I would think that so I'm not really interested in going further

[02:38:03.0000] <wycats>
:(

[02:39:00.0000] <wycats>
I wasn't trying to troll, I promise

[02:39:01.0000] <wycats>
annevk: like I said before, I'd *love* to see a Fennec/Webkit world

[02:40:00.0000] <doublec>
Except you mean Gecko since Fennec is a browser and Webkit isn't. And add in Opera's engine.

[02:41:00.0000] <othermaciej>
Presto is a good engine

[02:42:00.0000] <doublec>
Everyone is getting good engine's - it's hard to choose :)

[02:42:01.0000] <othermaciej>
it would be good if one of the sgnificant smartphone OSes shipped a browser based on Gecko or Presto by default

[02:42:02.0000] <Rik`>
in the original argument, I think the cycle "implement new feature, let it be tested with nightly/beta, refine it, propose it to standards" is used a lot more these days than say 2 years ago and so everything is moving faster, even standardization 

[02:42:03.0000] <othermaciej>
unfortunately there seems to be a herd instinct

[02:42:04.0000] <annevk>
wycats, sounds like we're in agreement then

[02:42:05.0000] <wycats>
doublec: I guess I mean Gecko if there was the prospect of a multi-gecko browser landscape

[02:43:00.0000] <wycats>
Rik`: exactly

[02:43:01.0000] <wycats>
othermaciej: I think it's less of a herd instinct and more a desire to pool resources

[02:43:02.0000] <wycats>
if say RIM decided to ship Gecko, they'd be going it alone

[02:43:03.0000] <doublec>
wycats, microb on the n900 is a gecko based browser

[02:44:00.0000] <othermaciej>
Nokia did the first mobile port of WebKit, before Apple

[02:44:01.0000] <wycats>
doublec: you know what I'm saying

[02:44:02.0000] <othermaciej>
they were going it alone at the time

[02:44:03.0000] <wycats>
othermaciej: yep

[02:44:04.0000] <wycats>
and few people do that

[02:44:05.0000] <wycats>
it's reasonable to not want to

[02:44:06.0000] <wycats>
I wouldn't call it a "herd instinct"

[02:44:07.0000] <wycats>
but at the same time, it would be good for the web if someone broke ranks

[02:44:08.0000] <wycats>
I agree

[02:44:09.0000] <wycats>
as long as it's not MS :P

[02:44:10.0000] <othermaciej>
now they are switching from their fork to QtWebKit which is up to date with trunk and everything

[02:45:00.0000] <boblet>
oook, not a Chrome inline-table issue. something to do with default ruby styles…

[02:45:01.0000] <wycats>
othermaciej: sweet

[02:45:02.0000] <Rik`>
to help choosing who won this argument : http://people.opera.com/danield/css3/lightsaber/

[02:45:03.0000] <wycats>
sachin lost :P

[02:46:00.0000] <doublec>
haha, nice

[02:46:01.0000] <othermaciej>
Rik`: that is totally sweet!

[02:47:00.0000] <Rik`>
othermaciej: now I want WebKit on iPhone to have the same orientation events than Gecko to use this

[02:47:01.0000] <Rik`>
(and I want WebKit nightlies on iPhone to test it right now :) )

[02:48:00.0000] <othermaciej>
I wish Gecko hadn't used "orientation" as event names since iPhone WebKit uses "orientationchange" for a different purpose

[02:49:00.0000] <zcorpan>
othermaciej: QtWebKit doesn't have this bug? https://bugs.webkit.org/show_bug.cgi?id=9677

[02:49:01.0000] <annevk>
-> public-webapps⊙wo

[02:50:00.0000] <othermaciej>
zcorpan: I dunno - WebKit trunk certainly doesn't, and I don't think we have a mode flag to enable it

[02:53:00.0000] <Rik`>
cool : http://robertnyman.com/css3/css-transitions/css-transitions-mac-os-x-stacks.html via http://robertnyman.com/2010/04/30/using-css3-transitions-and-transforms-to-mimic-mac-os-x-stacks-behavior/

[02:54:00.0000] <wycats>
Rik`: +1

[02:54:01.0000] <wycats>
Rik`: I'm more interested in the accelerometer ;)

[02:54:02.0000] <othermaciej>
Rik`: very cool!

[02:57:00.0000] <Hixie>
joehewitt's complaint seems almost like a tautology to me... anything that involves multiple vendors agreeing on what the platform should be is naturally going to evolve more slowly than something where the platform is decided by a single vendor

[02:57:01.0000] <Hixie>
that's the cost of not having vendor lock-in

[02:57:02.0000] <Hixie>
equivalently, the cost of having fast platform development is vendor lock-in

[02:57:03.0000] <wycats>
Hixie: well

[02:57:04.0000] <wycats>
it's not happening!

[02:57:05.0000] <wycats>
see my list above

[02:58:00.0000] <Hixie>
what's not happening?

[02:58:01.0000] <wycats>
browsers sitting around waiting for agreement before implementing things

[02:58:02.0000] <othermaciej>
the Web definitely evolves slower than the faster-moving proprietary platforms

[02:59:00.0000] <Hixie>
wycats: i didn't say that

[02:59:01.0000] <wycats>
othermaciej: Sachin's post lists technologies that "the web doesn't have" that were in browsers in like 2006

[03:00:00.0000] <Hixie>
"the platform" != what 15% of deployed browsers support

[03:00:01.0000] <othermaciej>
true, his post seems misinformed

[03:00:02.0000] <wycats>
Hixie: he's explicitly NOT complaining about 100% of browsers

[03:00:03.0000] <wycats>
he's saying "nobody is implementing useful things"

[03:00:04.0000] <Creap>
hsivonen: validator.nu fails to accept role="menuitem" inside role="menubar"

[03:00:05.0000] <wycats>
which is just bull

[03:01:00.0000] <Creap>
An element with role=menuitem requires role=menu  on the parent.

[03:01:01.0000] <Creap>
which is incorrect

[03:01:02.0000] <wycats>
also, IE only has 60'ish% these days

[03:01:03.0000] <wycats>
so 15% isn't exactly correct

[03:02:00.0000] <Hixie>
wycats: maybe he has multiple complaints. What I was talking about e.g. "I am ranting because I want to drop Cocoa and go back to the web, but I am upset about how much power I have to give up to do that."

[03:02:01.0000] <Hixie>
by 15% i meant one browser's latest version

[03:02:02.0000] <Creap>
hsivonen: W3C validator does the same though, but http://www.w3.org/TR/wai-aria/roles#menuitem clearly states menuitem is allowed inside menubar

[03:02:03.0000] <wycats>
Hixie: I'm mostly looking at Sachin's argument

[03:02:04.0000] <wycats>
not Joe's

[03:03:00.0000] <Hixie>
wycats: i was just talking about joehewitt's

[03:03:01.0000] <wycats>
Joe is unfocused... he appears to not have looked at the web platform in a couple years

[03:03:02.0000] <wycats>
Hixie: ha

[03:04:00.0000] <wycats>
http://posterous.com/getfile/files.posterous.com/sachin/qce6grV5S6PiTCvtEkmRiqRAJXP4a28SpIMTbng1rMX2T6v192UOuonHr8PP/pastedGraphic.png

[03:04:01.0000] <wycats>
^^ is just patent nonsense

[03:07:00.0000] <mut>
hey, if i draw something to scale (1px = 1mm) in canvas, how can i output that to an ascii file, so i can parse it and use it with other programs?

[03:09:00.0000] <othermaciej>
to the extent that joehewitt wants to do web development, I think he wants it mobile-targeted and maybe even a-specific-subset-of-mobile-devices-targeted

[03:09:01.0000] <othermaciej>
so latest version of one implementation would be good enough for him

[03:10:00.0000] <othermaciej>
as long as it's the right implementation

[03:11:00.0000] <wycats>
othermaciej: and... Mobile Webkit is pretty far ahead of the game

[03:11:01.0000] <wycats>
where's the Touch events spec?

[03:11:02.0000] <othermaciej>
I can't talk about that without presence of counsel

[03:14:00.0000] <wycats>
;)

[03:14:01.0000] <wycats>
give me a damn accelerometer :P

[03:15:00.0000] <wycats>
it definitely is detrimental that the mobile bits of Webkit are mostly closed source and maintained by Apple

[03:15:01.0000] <othermaciej>
depends on which bits

[03:15:02.0000] <wycats>
othermaciej: can the OSS community add accelerometer support to iPhone?

[03:17:00.0000] <othermaciej>
would probably be a challenge

[03:18:00.0000] <wycats>
mainly because dev. on those sorts of features is happening outside the OSS codebase

[03:24:00.0000] <Philip`>
mut: The only output you can get is the bitmap image data

[03:24:01.0000] <Philip`>
It doesn't save any information about the sizes of the shapes you draw onto it

[03:26:00.0000] <wycats>
good night folks

[03:43:00.0000] <mut>
ok

[03:43:01.0000] <mut>
Philip` but the bitmap image data is basically a pixel map yea

[03:45:00.0000] <mut>
ah im gonna have to rewrite/emulate all the canvas functions to output to dxf

[03:45:01.0000] <mut>
:P

[03:45:02.0000] <mut>
thats a job for this weekend :)

[03:47:00.0000] <Philip`>
mut: If you can generate your output file from just the pixel values, then getImageData or toDataURL would be sufficient

[03:48:00.0000] <mut>
hmm i could do, but it would be half assed.

[03:48:01.0000] <mut>
im all about half assed though

[03:48:02.0000] <mut>
getImageData will make it like a uri thing yea

[03:50:00.0000] <Philip`>
toDataURL makes it like a URL, getImageData makes it like an array of RGBA values

[03:51:00.0000] <mut>
hmm

[03:51:01.0000] <mut>
ok :)

[03:51:02.0000] <mut>
researchtime i think

[03:51:03.0000] <mut>
thanks

[03:55:00.0000] <mut>
nah thats not gonna work for me

[03:55:01.0000] <mut>
Ill have to sit down and recreate the canvas functions to create my file

[03:56:00.0000] <mut>
nm :P

[04:23:00.0000] <gsnedders>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0D%0A%3Ciframe%20src%3D%22data%3Aapplication%2Fxml%2Cfoobar%22%3E%3C%2Fiframe%3E%0D%0A%3Cscript%3E%0D%0Aw(document.getElementsByTagName(%22iframe%22)%5B0%5D.contentDocument.documentElement)%3B%0D%0A%3C%2Fscript%3E

[04:23:01.0000] <gsnedders>
That's kinda interesting.

[04:26:00.0000] <zcorpan>
wonder if the web depends on that by now

[04:26:01.0000] <gsnedders>
Minefield and Chrome give about:blank, Opera gives the parse error page

[04:32:00.0000] <jgraham>
(which seems bad)

[04:32:01.0000] <gsnedders>
Buggy test

[04:32:02.0000] <gsnedders>
http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Ciframe%3E%3C/iframe%3E%0A%3Cscript%3E%0Aw(%22loading%22);%0Avar%20iframe%20%3D%20document.getElementsByTagName(%22iframe%22)[0];%0Aiframe.onload%20%3D%20function(){%0Aw(%22loaded%22);%0Aw(iframe.contentDocument);%0A}%0Aiframe.src%20%3D%20%22data:application/xml,foobar%22;%0A%3C/script%3E

[04:33:00.0000] <gsnedders>
Chrome gives undefined for contentDocument, which I think is against HTML 5

[04:33:01.0000] <gsnedders>
Opera and Minefield gives the error page

[09:14:00.0000] <tabatkins>
Argh, snarky comment successfully redacted before sending.

[09:15:00.0000] <Philip`>
How boring

[09:16:00.0000] <tabatkins>
Hey, I'm already violating a rule by directly responding.  I was doing fine just ignoring emails, but got tired of patent lies being spread.

[10:41:00.0000] <lazni>
/me started to listen to Public Enemy when mpilgrim quoted them "if you don't own the master, the master control you" in context of package signing in Chrome OS

[13:12:00.0000] <AryehGregor>
I keep seeing Google ads for ie8optimized.com.  It's a complete fraud.

[13:12:01.0000] <AryehGregor>
I'm pretty sure I already reported it.

[13:26:00.0000] <othermaciej>
what is it?

[13:27:00.0000] <boogyman>
AryehGregor: why would a channel formed by WHATWG care about an ad on google?

[13:27:01.0000] <AryehGregor>
I dunno, nobody else seemed to be talking, so why not?  It's about browsers and Google, and there are a bunch of people here who are interested in one or the other.

[13:27:02.0000] <Philip`>
The channel would be boring if it was on topic all the time

[13:28:00.0000] <boogyman>
so if this channel had 10 users, you wouldn't have posted it?

[13:28:01.0000] <AryehGregor>
If people here usually cared about topicality when nothing interesting was happening, I wouldn't have posted it.

[13:28:02.0000] <lazni>
/me 's first sentence here today is offtopic

[13:29:00.0000] <AryehGregor>
Anyway, it poses as an official Microsoft page, and tries to get you to download some IE8 add-on.  But the whois is a Portuguese anonymization service.

[13:29:01.0000] <boogyman>
is/was?

[13:31:00.0000] <boogyman>
You can't cure gullibility. Think about it... MS marketing a plugin for their most recent browser? Doesn't sound too official to me

[13:32:00.0000] <Philip`>
I'm sure Microsoft markets Silverlight

[13:32:01.0000] <boogyman>
what's that?

[13:32:02.0000] <Philip`>
It's a plugin for their most recent browser

[13:32:03.0000] <AryehGregor>
You can't cure gullibility, but I'd expect that Google should be filtering out plainly fraudulent ads.

[13:32:04.0000] <boogyman>
and is the product name outside of their domain... eg  sub.microsoft.com?

[13:33:00.0000] <AryehGregor>
Especially since I reported it before.

[13:33:01.0000] <AryehGregor>
Yes, it's ie8optimized.com.  I already said.  Do a whois.

[13:33:02.0000] <AryehGregor>
That's what got me suspicious.

[13:33:03.0000] <boogyman>
was talking about silverlight

[13:33:04.0000] <AryehGregor>
Oh.

[13:33:05.0000] <AryehGregor>
I bet it's under microsoft.com.

[13:33:06.0000] <Philip`>
It's silverlight.net

[13:33:07.0000] <AryehGregor>
Okay, wrong.

[13:33:08.0000] <AryehGregor>
http://www.silverlight.net/

[13:33:09.0000] <AryehGregor>
You can verify that it's actually Microsoft by doing a whois, though.

[13:34:00.0000] <Philip`>
I can, but I'm not the target audience for fraudulent IE8 add-ons

[13:49:00.0000] <sicking>
tabatkins: ping

[13:49:01.0000] <tabatkins>
sicking: pong

[13:51:00.0000] <sicking>
tabatkins: I'd like to add wording to the change proposal to say that there are two browsers that have committed to supporting the new elements

[13:51:01.0000] <tabatkins>
Go for it, just throw a note on the list about it.

[13:51:02.0000] <sicking>
tabatkins: how would you feel about adding that as a new section under "rationale"?

[13:51:03.0000] <tabatkins>
official deadline isn't until next friday.

[13:51:04.0000] <sicking>
yup

[14:02:00.0000] <sicking>
Lachy: ping

[14:14:00.0000] <lazni>
teamviewer sucks big time

[14:14:01.0000] <lazni>
the concept is neat, implementation, sucks

[14:33:00.0000] <cardona507>
is this http://apirocks.com/html5/html5.html#slide45 telling me that you no longer need someone using IE to install chrome frame?

[14:46:00.0000] <AryehGregor>
No?

[14:51:00.0000] <Lachy>
sicking, pong

[15:20:00.0000] <sicking>
Lachy: do you know what Operas stance on the new semantic elements in HTML5 are?

[15:21:00.0000] <sicking>
Lachy: or rather, the stance on the ones that are being debated in the HTML5 WG right now

[15:21:01.0000] <sicking>
Lachy: wondering if I can put Opera in the "interested in implementing" column

[15:22:00.0000] <Hixie>
man i can't believe how much time and effort you're all putting into this issue

[15:23:00.0000] <tabatkins>
I stopped putting effort into it some time ago, at least.  I'm done with it.

[15:24:00.0000] <sicking>
most of the time is speant marking all the excessive emails read

[15:24:01.0000] <othermaciej>
honestly - there's nothing wrong with having rationale documents for parts of the spec

[15:25:00.0000] <Hixie>
rationale documents would be fantastic

[15:25:01.0000] <Hixie>
i've been looking for volunteers to write some for years

[15:26:00.0000] <othermaciej>
what tabatkins, sicking, hober et all wrote is a rationale document for some of the new elements, though an oddly formed one

[15:26:01.0000] <Hixie>
it's one of the most prominently positioned tasks in the list of things we're looking for volunteers to do on the wiki

[15:26:02.0000] <othermaciej>
still, would be nice to have even that level of written-down rationale for, say, the sectioning elements or the new form controls

[15:27:00.0000] <othermaciej>
I guess what I'm saying is, the work on this has definitely not been a waste, even though people may have prioritized their time differently if these were not tracker issues

[15:27:01.0000] <othermaciej>
that being said, those with Member access may want to look here http://lists.w3.org/Archives/Member/w3c-archive/2010Apr/0154.html

[15:28:00.0000] <Hixie>
holy cow, my home page is glowing green

[15:28:01.0000] <sicking>
in all fairness, i shouldn't get much credit for what's in the wiki. The other guys did most of the work

[15:28:02.0000] <Hixie>
it hasn't done that in like 6 months

[15:28:03.0000] <boogyman>
is that a good thing?

[15:29:00.0000] <Hixie>
boogyman: in-joke

[15:29:01.0000] <boogyman>
/me jumps out

[15:29:02.0000] <sicking>
Hixie: what a coincidence, the mozilla-central tree was the greenest i've seen it in months

[15:29:03.0000] <sicking>
until it went up in flames due to network issues :(

[15:30:00.0000] <Hixie>
anyway, it would be great if the rationale-providing text could be put in a wiki somewhere in a form suitable for a rationale document, that way we could point people to that and they would have a better chance of doing more

[15:30:01.0000] <Hixie>
it's always easier to add than to begin

[15:31:00.0000] <othermaciej>
unfortunately this page is also showing a lot of green: http://damowmow.com/playground/htmlwg/chart.html

[15:31:01.0000] <Hixie>
yikes

[15:34:00.0000] <othermaciej>
speaking of which....

[15:35:00.0000] <othermaciej>
Hixie: can you look at this (and the prior few messages): http://lists.w3.org/Archives/Public/public-html/2010Apr/1248.html

[15:35:01.0000] <othermaciej>
Hixie: if you strike or suitably adjust the requested clause, we can close ISSUE-82

[15:36:00.0000] <Hixie>
yeah it's on my huge pile of mail to read

[15:36:01.0000] <Hixie>
i'm about 24 hours behind right now

[15:37:00.0000] <othermaciej>
ok

[15:37:01.0000] <othermaciej>
thanks

[16:05:00.0000] <Lachy>
sicking, if you're referring to things like meter, progress, hidden="", etc. then I am personally in favour of retaining them. I also believe the rest of us are too

[16:05:01.0000] <Lachy>
at least, I've not heard anything from anyone at Opera arguing that the elements or attributes should be dropped.

[16:06:00.0000] <sicking>
Lachy: are "the rest of us" enough for me to put on the wiki that "opera has indicated interest" in supporting them?

[16:07:00.0000] <Hixie>
wasn't i working for opera when some of those were originally added?

[16:07:01.0000] <Lachy>
well, obviously, I can't make an official statement on behalf of opera, but yeah, I think it's fairly safe to say that.

[16:08:00.0000] <Lachy>
I think you were

[16:08:01.0000] <Hixie>
(not that that really means anything)

[16:09:00.0000] <Lachy>
I've been meaning to provide some feedback about the zero-edit change proposal, but haven't got around to it yet

[16:10:00.0000] <Lachy>
my problem with it is that it fails to adequately describe the use cases for many of the elements.

[16:12:00.0000] <sicking>
Lachy: cool

[16:27:00.0000] <Lachy>
sicking, what's mozilla's status with implementing <meter> and <progress>? Are there any plans for those in the near future?

[16:28:00.0000] <Lachy>
Or are there plans to implement XBL first, which is somewhat essential for applying advanced styles to those elements.?

[16:29:00.0000] <Lachy>
we also need psudo-elements for them, or styling limitations will be one of the biggest problems with them.

[16:31:00.0000] <tabatkins>
I need to spend a few hours thinking about the pseudoelems to expose on <progress>, so that later XBL can expose them in a non-magical way.

[16:34:00.0000] <tabatkins>
(or someone else can, of course, but I'll probably do it sometime soon)

[16:34:01.0000] <Hixie>
<progress> doesn't really need much, it's <meter> that really needs something

[16:34:02.0000] <Hixie>
<progress> you're more likely to do using <canvas> or some such

[16:35:00.0000] <tabatkins>
Not necessarily.  There are plenty of good, simple uses that are roughly equivalent to the styling needs of <meter>.

[16:37:00.0000] <Lachy>
I think progress only needs 1 pseudo-element the highlighted portion.

[16:38:00.0000] <Lachy>
meter needs a lot to handle all the different components, from the low point, high point, mid point, highlighted region, optimal point and maybe more. Haven't really thought about it in detail though.

[16:40:00.0000] <Hixie>
the more i think of it the more i think both are likely to just be done programmatically (e.g. using canvas)

[16:40:01.0000] <Hixie>
rather than with css

[16:40:02.0000] <Hixie>
though one might want to parameterise it via css somehow

[16:43:00.0000] <Lachy>
My original thoughs were that authors would use XBL templates to replace them with scripted/animated SVG

[16:43:01.0000] <Lachy>
but canvas is another likely alternative

[16:44:00.0000] <tabatkins>
That would just be <canvas><progress/></canvas>, right?

[16:48:00.0000] <boogyman>
tabatkins you're using xhtml syntax

[16:48:01.0000] <tabatkins>
No, I'm using pseudocode.  ^_^

[16:53:00.0000] <Lachy>
tabatkins, I don't think that's the best way of dealing with it because doing that way doesn't create a nice, easily reusable component

[16:56:00.0000] <Lachy>
rather, I think it would be something like this <xbl xmlns=...><template><canvas/></template><script>onprogress = function() {...}</script></xbl> (ignore errors in XBL, done from memory), and then that template would be bound to all progress elements in the page

[16:57:00.0000] <tabatkins>
Kk.