00:01
<kingryan>
I have an email mailbox that contains most of them. that works alright for me
00:01
<kingryan>
but I wish there were something better
00:08
<Philip`>
You can use svnsync to replicate the repository onto your own machine, then run a local web server with ViewSVN or Trac or something
00:15
<kingryan>
hmm, that sounds useful
00:17
<takkaria>
IIRC google have said they'll do something about it in an unspecified amount of time
00:18
Philip`
already has some other copy of Trac so he might see if it's trivial to do an html5lib one, since it's a reasonably nice source browser...
01:24
<Philip`>
Oh, good, it is easy to set up
01:24
<Philip`>
http://canvex.lazyilluminati.com/html5lib/
01:29
<kingryan>
nice, Philip`
01:53
<welly>
is the faq response to "when will html 5 be finished" a joke?
02:01
<Philip`>
welly: No - it just has a more complete meaning of "finished", like requiring there to be multiple correct implementations of the whole specification before it's considered finished
02:02
<Philip`>
The writing of the specification will be finished much sooner, and implementations of some features are already available and usable
02:02
<welly>
Philip`: so in 15 years time, html 5 will be a complete specification? I can't help but think the internet moves on a little bit quicker than that
02:03
<welly>
i can't even imagine what kind of technology or standards we're going to be using in 15 years time
02:04
<Philip`>
HTML4 is almost ten years old already, and not that much has changed since then
02:05
<welly>
it seems like an interesting project but wow.. lol... 15 years. no, you're right but all the same.
02:05
<Philip`>
(HTML4 and CSS2 (I think) wouldn't be considered finished yet, if they used the same criteria as HTML5)
02:05
<welly>
ok
02:06
<welly>
what's the likelyhood of html being taken up and implemented by the browser developers?
02:07
<welly>
er html 5
02:07
<Philip`>
Some parts have already been implemented - http://wiki.whatwg.org/wiki/Implementations_in_Web_browsers
02:08
<welly>
ahh ok.. well that's promising
08:56
<Lachy_>
My presentation went well this morning :-)
10:12
<ROBOd>
hey guys
10:14
<ROBOd>
is it a problem if i sent two emails about the HTML 5 spec to www-html and not to public-html? i sent two emails yesterday by mistake to www-html. i forgot of public-html, eh
10:27
<annevk>
you could provide a pointer to your e-mails on www-html in your e-mail to public-html
10:29
<ROBOd>
annevk: eh, two mailing lists with almost the same name. i think should unsubscribe from www-html, just for avoiding these situations
10:30
<annevk>
that's what I did, though for different reasons
10:31
<ROBOd>
should i resend the two emails to public-html? or just a pointer?
10:33
<annevk>
whatever suits you
10:34
<hsivonen>
ROBOd: I unsubscribed from www-html after I ended up sending email to public-html when public-html was in recess.
10:35
<hsivonen>
ROBOd: also, www-html is more or less in /dev/null mode as far as actually getting yourself heard goes
10:35
<ROBOd>
hsivonen: yeah, too much confusion
10:35
<ROBOd>
hsivonen: precisely. i only got one reply :P
10:35
<annevk>
I believe Hixie still reads it
10:39
<ROBOd>
that's very nice, if Hixie still reads www-html
18:57
<G0k>
hey all
18:57
<zcorpan_>
hey G0k
18:58
<G0k>
i'm a little confused about the RemoteEventTarget interface in html 5
18:58
<G0k>
"Any object that implements the EventTarget interface must also implement the RemoteEventTarget interface."
18:58
<zcorpan_>
which part is confusing?
18:59
<G0k>
i mean is that really the same as saying that the EventTarget interface is being amended to include some new methods?
19:00
<zcorpan_>
not quite but i guess the end result would be pretty much the same
19:04
<zcorpan_>
the difference being that it would be one interface instead of two... you can check whether an object implements a certain interface and which members it has
19:04
<G0k>
right but if any object that implements EventTarget "MUST" also implement RemoteEventTarget....
19:05
<G0k>
i guess this allows you to detect legacy behavior better
19:06
<G0k>
ok the other thing
19:06
<G0k>
"For connections to domains other than the document's domain, the semantics of the Access-Control HTTP header must be followed. [ACCESSCONTROL]"
19:06
<G0k>
[ACCESSCONTROL] seems to be a dead link?
19:06
<zcorpan_>
yes, refs haven't been written yet
19:06
<G0k>
heh. oki
19:08
<zcorpan_>
http://dev.w3.org/cvsweb/~checkout~/2006/waf/access-control/Overview.html?content-type=text/html;%20charset=utf-8
19:12
<G0k>
ok there's one other part that has me slightly confused
19:12
<G0k>
"If the Namespace is null and the Event field is message (including if it was not specified explicitly), then the MessageEvent interface must be used.
19:12
<G0k>
Otherwise, the Event interface must be used."
19:12
<G0k>
so that would seem to suggest for instance, that if you got an event like
19:12
<G0k>
Event: foo
19:12
<G0k>
data: bar
19:13
<G0k>
it would create an Event event with type "foo", then ignore the data field, since data isn't a defined attribute for the Event interface
19:15
<G0k>
which isn't bad, it's just precisely the opposite of how Opera has implemented it
19:15
<zcorpan_>
it is?
19:15
<G0k>
well like...see their sample code http://labs.opera.com/news/2006/09/01/
19:16
<G0k>
they have it sendings stuff like
19:16
<G0k>
Event: the-answer
19:16
<G0k>
data: 42
19:16
<G0k>
in their implementation, it sets the data attribute on the event
19:16
<G0k>
but that's not technically what the spec says, as the Event interface has no data attribute
19:18
<G0k>
for that matter, the pending patch for mozilla also implements it the other way
19:18
<zcorpan_>
hmm, might be a bug in the spec then
19:20
<G0k>
k. is there an issue tracker somewhere?
19:20
<zcorpan_>
no
19:20
<G0k>
should I use the forum?
19:21
<zcorpan_>
it would be great if you could send it to the list directly
19:21
<G0k>
which? html-dev or whatwg?
19:21
<zcorpan_>
whatwg
19:22
<G0k>
k
19:22
<zcorpan_>
what is html-dev?
19:22
<G0k>
er i dunno the one w3c runs
19:22
<zcorpan_>
aha. public-html
19:22
<G0k>
that one
19:23
<zcorpan_>
if you're subscribed then you can send it to that list if you want. if you do, include "detailed review of" in the subject line
19:24
<G0k>
k
19:47
<G0k>
sent
19:47
<zcorpan_>
thanks
19:47
<G0k>
np
20:34
<annevk>
it defaults to the MessageEvent actually, not Event
20:34
<G0k>
except if the Event: field is specified
20:35
<G0k>
(to something other than "message")
20:35
<G0k>
or at least that's what the spec seems to say right now
20:36
<annevk>
no
20:36
<annevk>
oh, maybe
20:36
<G0k>
"If the Namespace is null and the Event field is message (including if it was not specified explicitly), then the MessageEvent interface must be used.
20:36
<G0k>
Otherwise, the Event interface must be used."
20:36
annevk
looks at the spec
20:36
<annevk>
that seems wrong
20:37
<G0k>
yeah :)
20:37
<G0k>
it should just say
20:37
<G0k>
"Otherwise, the MessageEvent interface must be used"
20:38
<G0k>
(imho)
20:38
<annevk>
yeah, that's what we do anyway and it's proven useful
20:38
<G0k>
yep
20:38
<G0k>
it's what i did in my webkit implementation too, then i just re-read that
20:41
<annevk>
ah, you're implementing this in WebKit? awesome
20:42
<G0k>
yeah well...it's a little sketchy right now but with some work yeah it'll be cool. :)
20:42
<annevk>
:)
20:42
<G0k>
i'll be able to stream events to my iphone. :)
21:02
<annevk>
html5lib is linked from dailypythonurl, whatever that is
21:02
<annevk>
http://www.pythonware.com/daily/
21:40
<Hixie>
(yes i still read www-html)
21:40
<Hixie>
if anyone sees G0k again, let him know I agree with his proposal
21:41
<Hixie>
i'd mail him straight back but his mail is lost somewhere in my event-source feedback pile
21:41
<zcorpan_>
"Henry Mason" <hmason⊙mc>
22:05
<Hixie>
thanks
22:22
<Jero>
hey whats up
22:23
<Jero>
I was wondering when the WHATWG will start on the HTML5 spec again
22:26
<zcorpan_>
Jero: you mean when Hixie will start edit the spec again?
22:27
<Jero>
or that, yes
22:27
<zcorpan_>
well, he's back from his vacation now so i guess he just needs to get up to speed with all email
22:28
<Jero>
i see, thanks
22:30
<zcorpan_>
http://annevankesteren.nl/2007/07/web#comment-6176
23:39
<Hixie>
zcorpan_: i can't get to that site, is it down?
23:40
<hsivonen>
oh dear. this "low quality" thing is going to be a rathole
23:43
<Hixie>
it's the same style="" rathole we've always had
23:44
<hsivonen>
Hixie: no. now it is becoming a b and i rathole as well
23:47
<Hixie>
othermaciej_: <font> and style="" are "low quality" because they are media-specific, and potentially hide information from non-presentational UAs.
23:48
<Hixie>
othermaciej_: <div>s-containing-inlines are "low quality" because they imply some sort of paragraphing structure without giving any information about why
23:49
<Hixie>
othermaciej_: (typically, the latter is used as a workaround for the lack of a widget abstraction language like xbl)
23:49
<othermaciej>
Hixie: I don't see how style="" is media-specific
23:50
<othermaciej>
Hixie: it lacks the ability to give different styles per media, but can still be used in media-independent ways
23:50
<othermaciej>
since in practice most styles are not media-scoped anyway, nor do they need to be
23:51
<Hixie>
ok, strike the media-specific part
23:51
<othermaciej>
in that case, style="" hides no more information than <style>, <link rel="stylesheet"> or <style scoped>
23:51
<Hixie>
(though i actually do believe they usually result in media-specific content, even if it's technically possible to use them otherwise)
23:51
<othermaciej>
none of which you have labelled "low quality"
23:52
<othermaciej>
I have to go grab a beer, brb
23:52
<Hixie>
i considered including <style>, but in practice if you use a style sheet with selectors you tend to be pushed towards content that can stand even without the stylesheet.
23:53
<Hixie>
obviously though that's not automatically the case, which is why the other mode wouldn't be called "high quality"
23:57
<othermaciej>
back
23:58
<othermaciej>
if you include <style> and <link rel="stylesheet"> it would make it pretty hard to have a useful non-low-quality document
23:58
<Hixie>
indeed
23:59
<othermaciej>
but I don't think there's any evidence of a difference in hiding information from non-presentational UAs between those and style=""
23:59
<Hixie>
oh come now
23:59
<othermaciej>
I mean, that might be the case, but I am not sure how you would quantify it