00:43
Hixie
ponders how to split websockets
00:56
<Hixie>
wakaba: wow, the whatpm and webhacc stuff has progressed a lot since i last looked at it
00:56
<Hixie>
wakaba: what's the status these days? is it stable enough to use?
02:13
<Hixie>
wakaba: looks like Message::Charset::Info is missing from the manakai-core tarball
06:36
<olliej>
hmm, no roc
09:31
<hsivonen>
http://twitter.com/diveintomark/statuses/1101238201 Smil 3.0 is over half the size of HTML5
09:33
<annevk>
It's not really clear to me what we can conclude from that...
09:34
<annevk>
E.g. is it lacking detail? Is there a 100 pages devoted to different schema languages?
09:34
<annevk>
Do they use a page per namespace? :p
09:34
<hsivonen>
nothing to conclude except that it's a long spec
09:36
<annevk>
Previous versions of SMIL were too. There's a lot of stuff in there... I sort of hope we never have to implement more than the subset that is in SVG
09:37
<annevk>
and even that I wish we would not have to
09:41
hsivonen
wonders if markp is following accessibility at xiph dot org
10:06
<roc_>
SMIL is horrible
10:08
<Lachy>
hsivonen, yt?
10:09
<hsivonen>
Lachy: yes
10:09
<Lachy>
see http://validator.nu/?doc=http%3A%2F%2Fwww.brucelawson.co.uk%2Ftests%2Fhtml5-youtube.html
10:09
<Lachy>
The error report about the object element's missing attributes could be clearer about which attributes are required
10:10
<Lachy>
Bruce couldn't figure out why it wasn't valid and just asked me about it
10:10
<hsivonen>
Lachy: yes. still waiting for upstream to solve it :-/
10:10
<Lachy>
ok
10:10
<hsivonen>
also, I have lots of fixed from MikeSmith pending deployment
10:11
<hsivonen>
s/fixed/fixes/
10:43
annevk
finds https://www.ietf.org/mailman/listinfo/imap5
10:52
<zcorpan>
NAMESPACE:
10:52
<zcorpan>
Kill it after torturing it
10:53
<zcorpan>
http://www.ietf.org/mail-archive/web/imap5/current/msg00116.html
10:54
<hsivonen>
I wonder how offering IE8 as a high-priority update goes considering that IE7 was supposed to be a high-priority update but IE6 is still out there
10:56
<zcorpan>
hsivonen: since ie7 requires xp sp2 and there are users with windows pre xp sp2, i guess the ie6 users are mostly using pre xp sp2 and so will continue to be unaffected
10:57
<zcorpan>
hsivonen: but most of the current ie7 users might well upgrade
10:57
<hsivonen>
isn't SP2 a high-priority update, too?
10:57
<zcorpan>
dunno
10:57
<roc>
it is
10:58
<roc>
I met a friend today who uses IE6 because it uses less memory than other browsers
10:58
<roc>
techie friend too
11:01
<hsivonen>
does XP1 still get security patches?
11:01
<jgraham>
hsivonen: Well I built gecko with HTML 5 parsing and it consistently crashes at startup :)
11:02
<hsivonen>
jgraham: what platform?
11:02
<hsivonen>
jgraham: I've used Leopard on Intel and 32-bit Intrepid
11:03
<Philip`>
http://www.ietf.org/mail-archive/web/imap5/current/msg00119.html - "During my recent job search, I was lectured that I had to get over this quaint notion that interoperability is important or even desirable. ... I have become convinced that it is hopeless to do anything in open standards beyond a general framework."
11:03
<jgraham>
hsivonen: 32-bit Intrepid
11:03
<hsivonen>
jgraham: interesting
11:04
<jgraham>
hsivonen: Let me know if I can do anything useful to diagnose the issue'
11:04
<hsivonen>
jgraham: are you using a profile with non-default open tabs?
11:05
<jgraham>
hsivonen: I created a new profile.
11:06
<jgraham>
(it shows the profile manager before crashing)
11:06
<hsivonen>
jgraham: ok. It's hard to figure that out without a stack trace.
11:12
<rubys>
that's what I saw too when I used no-remote
11:12
<rubys>
and -P
11:12
<rubys>
with no options, it worked for me
11:13
<takkaria>
they're working on imap5?
11:55
<Philip`>
I also get a crash on startup after the profile manager (when running with -profilemanager)
11:56
<Philip`>
http://pastie.org/354574
11:58
<Philip`>
and it still crashes if I just use -no-remote (and not -profilemanager)
12:00
<Philip`>
but otherwise it seems to work alright
12:03
<Philip`>
It seems incredibly slow at loading the HTML5 spec
12:04
<Philip`>
and I can see it redrawing the screen after every few words have been added to the bottom
12:04
<hsivonen>
Philip`: it probably triggers O(n^2) behavior on the CSS side
12:26
<zcorpan>
hmm... <form><div><span></form><h4>
12:26
<zcorpan>
html5 says to close the form
12:27
<zcorpan>
the top 4 browsers don't
12:27
<zcorpan>
our rendering is compatible with html5 afaict but he have to change it because it breaks sites (bug with 13 dups)
12:27
<zcorpan>
Hixie: ^
12:36
<Hixie>
you mean the <h4> ends up in the <span>?
12:37
<Hixie>
cos the spec does do that now
12:38
<Hixie>
zcorpan: ^
12:40
<annevk>
Hixie, how would you link window.XMLHttpRequest with a script execution context so that if you do var obj = {}; obj.XMLHttpRequest = otherwindow.XMLHttpRequest otherwindow.XMLHttpRequest is associated with its original script execution context and not the one obj is running in
12:40
<zcorpan>
Hixie: oh it does? ok then that's good and validator.nu is out of date
12:41
jgraham
guesses html5lib is also out of date
12:41
<jmb>
hubbub certainly is :)
12:41
<annevk>
Hixie, the SharedWorker IDL block uses <code>
12:42
<zcorpan>
jgraham: yep (at least the instance running at james.html5.org)
12:42
<annevk>
Hixie, showNotification is still present in Web Workers
12:44
<Hixie>
annevk: no idea, like i said though, this whole area is agoing to change radically
12:44
<Hixie>
annevk: (please file a bug about SharedWorker)
12:44
<Hixie>
(or send me a mail)
12:45
<Hixie>
time for bed
12:45
<Hixie>
nn
12:46
<jgraham>
zcorpan: Still wrong with the latest version. I must have missed that change and obviously we don't have tests for it :(
13:18
<zcorpan>
it seems that in firefox it's impossible to close the form with the above markup
13:18
<zcorpan>
and you can nest forms with <form><x></form><form>
14:38
<annevk>
heycam, does Web IDL define a way to define a prototype object, such as window.XMLHttpRequest?
14:38
<annevk>
heycam, also, does it define a way to say "interface X complements interface Y"?
14:38
<Dashiva>
Isn't that the Interface Object?
14:39
<annevk>
Dashiva, prolly true, now that we have "NoInterfaceObject"... bit confusing with Window members though
14:53
<annevk>
interface object is defined in section 4.3 of Web IDL btw
14:55
<annevk>
does DontDelete imply you cannot overwrite it? because you can do window.XMLHttpRequest = otherwindow.XMLHttpRequest in browsers
14:57
<annevk>
I guess ReadOnly would imply that, great
14:58
annevk
tries to avoid actually reading the ECMAScript spec for now
14:59
<Philip`>
annevk: DontDelete only affects [[Delete]], not anything else
14:59
<Philip`>
(as far as I can see)
14:59
<annevk>
heycam, why is "interface Dog;" present in the first example of section 3.3?
14:59
<Philip`>
"DontDelete: Attempts to delete the property will be ignored. See the description of the delete operator in section 11.4.1."
15:01
<annevk>
ta
15:19
<Philip`>
https://wiki.mozilla.org/Accessibility/JSON_ARIA#Defining_the_widget - I suppose that's one way to get around the lack of comments in JSON - just make up your own syntax
15:21
<aroben>
hi everyone!
15:24
<Philip`>
Howdy
15:25
<webben>
Philip`: I suspect those aren't meant to be in the final JSON. I could be wrong.
15:29
<Dashiva>
Philip`: Comments are part of many JSON extensions, so it's probably not invented there :)
15:37
<webben>
Dashiva: Wouldn't they normally use something like // or /* */ rather than <!-- --> ?
15:39
<Dashiva>
Browsers have to support <!-- for (single-line) comments anyway
16:44
<Lachy>
hey brucel
16:44
<brucel>
hi
21:58
<webben>
cesar2: when you say "getting" to mean via the video API?
21:59
<cesar2>
webben: yeah
22:00
<webben>
cesar2: Something i've been suggesting to Hixie is that metadata in the source video should be available to the DOM API.
22:00
<webben>
cesar2: However, if you're including a video on a page, you're normally going to be able to extract that info and provide it to your page manually anyhow.
22:01
<cesar2>
webben: true. But I figure that it could be useful for extension developers for firefox
22:02
<webben>
hmm interesting point
22:03
<webben>
cesar2: Perhaps put together some use-cases and write to WHATWG mailing list with that suggestion? http://wiki.whatwg.org/wiki/FAQ#Is_there_a_process_for_adding_new_features_to_the_spec.3F
22:03
<webben>
cesar2: I suspect Hixie's likely to treat it as a version 2 feature ... but no harm in raising the issue early.
22:03
<cesar2>
webben: I'll think about it some more. thanks for pointing to the resource
22:03
<webben>
np :)]
22:04
<gsnedders>
I have some vague recollection of that being mentioned at TPAC
22:04
<gsnedders>
I also have a vague recollection of Hixie thinking that should be added
23:14
<Hixie>
cesar2, webben: exposing metadata from the video is already on the v2 list
23:15
<cesar2>
what is the v2 list?
23:21
<Hixie>
the comment in the spec near the video section that starts "v2 features:" or some such
23:22
<Hixie>
every now and then someone tries to upload gigabyte-sized files to my data url kitchen
23:22
<Hixie>
qtf
23:22
<Hixie>
wtf even
23:47
<hallvors>
think I submitted a 1mb image to the data-url maker at some point (by mistake) :-p
23:49
<Hixie>
1mb is fine
23:49
<Hixie>
1gb not so much
23:49
Hixie
imposses a 100mb limit
23:49
<Hixie>
imposes
23:53
<BenMillard>
annevk, "Ideal" entry on your blog has this after the doctype: <!-- last modified Wed, 31 Dec 1969 16:00:00 -->
23:53
<takkaria>
fubky
23:53
<takkaria>
funky, even