00:00
<jamesr____>
if you have a canvas with an animation, say, it's kind of silly to invoke the callback if the canvas is scrolled way offscreen
00:00
<heycam>
avoid or throttle down?
00:00
<heycam>
(and this would be an optional parameter I guess?)
00:00
<jamesr____>
avoid
00:00
<jamesr____>
and yeah
00:00
<jamesr____>
if you don't specify a parameter then it's implicitly the document or whatever
00:01
<heycam>
ok my concern would be that the results of the animation (element position or whatever) might be used by scripts on the page so we might need to still run them
00:01
<heycam>
but the author could just not associate it with the element in that case
00:01
<jamesr____>
right. if they need the thing to run when the element is offscreen then they don't pass the parameter in
00:01
<jamesr____>
i want to get some more author feedback about how useful this mechanism is
00:02
<heycam>
that would be good (the feedback)
00:02
<heycam>
was there a proposal recently on the mailing list about determining whether an element is visible?
00:02
<heycam>
or was that for a whole page?
00:03
<jamesr____>
that was about determining if a page was visible
00:03
<jamesr____>
authors can currently detect whether an element is visible to some degree, but nobody bothers
00:03
<jamesr____>
so i'd like to make that the default behavior
00:03
<jamesr____>
imagine a web page with 5 animated banner ads implemented as <canvas>. it'd be really awesome if the default behavior was that we only spent CPU updating animations on the ones that were visible
00:04
<heycam>
yeah. seems good to me at first glance.
00:06
<benschwarz>
Hixie: the <header> is inside the <head>, any ideas?
00:06
<Hixie>
benschwarz: context?
00:07
<benschwarz>
ha. sorry
00:07
<benschwarz>
apparently I only told the voices in my head :/
00:07
<benschwarz>
after the spec splitter runs, I find <header class="head"> inside the <head> element
00:07
<benschwarz>
not sure exactly when it occurs yet, but it does…
00:08
<Hixie>
odd
00:08
<Hixie>
dunno
00:08
<Hixie>
gsnedders, jgraham, and Philip` might know
01:22
<Rik`>
out of curiosity, anyone knows who is maintaining the @html5 account on twitter?
01:23
<TabAtkins>
Dunno.
01:27
<benschwarz>
Rik`: Hixie, I think
01:28
<TabAtkins>
No, Hixie doesn't touch that one.
01:28
<Rik`>
that looks like a very marketing thing for Hixie :)
01:29
<TabAtkins>
All right, formatting cleanup for Lists is *done*. Time to commit, then work on actual changes.
01:32
<jamesr____>
TabAtkins: when defining an algorithm with numbered steps, how do you mark it up? if i just use <ol> and <li> then how do i say something like "go to step 4" without manually updating that whenever i change the list?
01:32
<TabAtkins>
I use ol, and manually update when I change the list.
01:32
<Hixie>
Rik`: MikeSmith
01:33
<TabAtkins>
Alternately, use ol, but name the important steps, and then refer to steps by name.
01:33
<uf02>
TabAtkins: question for you, how long do you think till they implement variables, mixins, nesting inside CSS3 or 4?
01:33
<uf02>
I mean realistically
01:33
<uf02>
I know you guys are working on it
01:33
<Rik`>
Hixie: thanks
01:34
<TabAtkins>
uf02: We're working on it now, so experimental implementations should hit the tree in some time period measurable in months. If everything goes nicely, other browsers could have experimental implementations by the end of the year. But this is all speculation.
01:34
<heycam>
jamesr____, the xslt thingo I use for Web IDL uses some processing instructions to reference step numbers
01:34
<TabAtkins>
Good thing is, all of that stuff is syntax sugar, so you can use a preprocessor immediately and get all the benefits.
01:34
<heycam>
to reference steps, that is
01:35
<uf02>
TabAtkins: the only thing I don't like about preprocessor is that is Ruby based
01:35
<uf02>
don't get me wrong SASS is effin great
01:35
<uf02>
but compiling is annoying
01:35
<uf02>
versusing just having editor you like and writing and boom
01:35
<uf02>
done.
01:36
<uf02>
I can't wait till you guys implement it because it will be heaven for guy like me
01:36
<TabAtkins>
uf02: Yeah, but that's what you'll have to do. Even if all browsers added the new features tomorrow, you'd have to wait until all older browsers were insignificant. These features don't really play nicely with backwards compat.
01:36
<TabAtkins>
SASS has a mode where it can watch for changes and automatically/transparently compile for you, you know.
01:37
<uf02>
what if they built something like jQuery for SASS
01:37
<uf02>
so it's automatic :)
01:37
<uf02>
and all you have to do is add a JS file
01:38
<uf02>
kind of what jquery is to js
01:38
<Yuhong>
Posted this to Slashdot:
01:38
<Yuhong>
http://slashdot.org/comments.pl?sid=1970938&cid=35039752
01:38
<TabAtkins>
That introduces latency to your page - the user won't see the right stuff until the CSS and JS have both loaded, the JS has parsed and recompiled the sheet, and the new CSS the JS inserted into the page is loaded.
01:39
<uf02>
yea I figured that, was just looking for an easier way to use preprocessor out the bat
01:39
<uf02>
well.. will have to wait
01:39
<uf02>
but at least you guys are working onit
01:39
<TabAtkins>
Or learn to love Ruby.
01:40
<uf02>
right.
01:42
<uf02>
I do wonder, what's your preference.. SASS or LESS ?
01:42
<Yuhong>
Basically the history of CSS support in IE.
01:42
<TabAtkins>
I used to like LESS, solely because it was closer to CSS syntax. But LESS made a lot of bad decisions that SASS did right, and now that SASS has their SCSS syntax which is close to real CSS syntax, I think it's much better.
01:43
<TabAtkins>
That said, i've never used a preprocessor.
01:43
<uf02>
heh that's interesting. why not?
01:43
<uf02>
never used it but likes SASS more
01:43
<TabAtkins>
Never felt the need to. I do all my development in a text editor.
01:44
<uf02>
gotcha
01:44
<erlehmann>
Yuhong, i am actually preparing a panel for a german conference on the topic of if this is the right way to make a standard
01:45
<Yuhong>
It took a lot of thinking and research to for me to figure it all out.
01:47
<erlehmann>
fun fact: i know no one who is a) clever enough for a panel b) too dumb to realize that defending apples move against ogg, webm etc. makes them an acceptable target for trolling
01:48
<TabAtkins>
You already tried Gruber?
01:51
<Rik`>
to my knowledge, we haven't seen Apple moving against ogg or webm
01:51
<erlehmann>
TabAtkins, it is in germany. i tried the german pendant, tim pritlove.
01:52
<Rik`>
Apple is well known for not commenting stuff before shipping them
01:52
<Rik`>
(which also means they are not commenting on stuff they will not implement)
01:53
<gsnedders>
Rik`: I think the assumption is they would've been on stage at Google I/O last year if they were going to.
01:53
<gsnedders>
Rik`: Instead of letting themselves/Moz/Op take the spotlight
01:53
<erlehmann>
https://secure.wikimedia.org/wikipedia/en/wiki/Tim_pritlove
01:53
<Rik`>
gsnedders: Have you ever seen Apple going to announce something with other companies?
01:54
<Rik`>
(ok, Verizon recently but there is a lot of money on the table)
01:55
<gsnedders>
Rik`: It seems unlikely they'd leave such a delay
01:56
<Rik`>
no hardware decoding available means they won't ship it
01:56
<erlehmann>
Rik`, how is jobs take on ‘“there is a patent pool being assembled” and “every video codec is patented” no move against open codecs?
01:57
<gsnedders>
Rik`: Are we sure? Does the iPhone use a DSP?
01:57
<gsnedders>
*a locked DSP
01:57
<jamesr____>
heycam: in your spec you say that the UA must add a requestAnimationFrame callback to the list if the callback is not already in the list. why exclude duplicates?
01:57
<erlehmann>
Rik`, “hardware” decoding usually happens in DSPs, special processors suited for the task.
01:57
<jamesr____>
heycam: if i do setTimeout(foo, 10); setTimeout(foo, 10); then i expect 'foo' to be invoked twice
01:57
<Rik`>
erlehmann: you're talking about a report of mail attributed to jobs
01:58
<Rik`>
there are a lot of fake reported mails
01:58
<gsnedders>
erlehmann: The DSPs used in lots of phones have embedded ROM in them which they execute, so once shipped you can't change what the DSP does.
01:58
<Rik`>
gsnedders: that's beyond my knowledge and competence
01:59
<Rik`>
gsnedders: but the spotlight you're talking about is very very very small
01:59
<heycam>
jamesr____, not sure, looking
01:59
<erlehmann>
gsnedders, oh. i thought you could upgrade the firmware ._.
01:59
<gsnedders>
Rik`: I wouldn't be surprised if it was entirely possible for them to implement hardware decoding for existing devices.
01:59
<Rik`>
maybe it is possible
01:59
<erlehmann>
even if it was, this isn't the deciding factor
02:00
<erlehmann>
and every single one of you knows it
02:00
<gsnedders>
erlehmann: On a lot of phones it's separate to the ROM the actual phone's firmware is stored in, and isn't reprogrammable
02:00
<Rik`>
but assuming that Apple is not gonna ship WebM in the future when they said nothing about it feels weird to me
02:00
<erlehmann>
gsnedders, :/
02:00
<Rik`>
erlehmann: what is the deciding factor ?
02:00
<heycam>
jamesr____, yeah I'm not sure why that's there, can't think of a good reason.
02:01
<jamesr____>
heycam: also, why are these "methods"? is that the IDL term?
02:01
<erlehmann>
Rik`, your sceptic argument goes like: you cannot know not-X. therefore, X is possible. i will not fall for that.
02:01
<erlehmann>
also, politics.
02:02
<heycam>
jamesr____, hmm no the IDL term is "operation"
02:02
<jamesr____>
whatwg seems to define things as 'method's
02:02
<jamesr____>
i'm looking at the setTimeout() definition
02:02
<heycam>
ok
02:02
<heycam>
method is a nicer name than operation
02:02
<Rik`>
erlehmann: Microsoft said they were giving more money to MPEG LA than receiving
02:02
<erlehmann>
Rik`, does the fact that apple did not announce support for 12 feet robots with furry paws raise the possibility that they will ship said bots?
02:03
<jamesr____>
definitely
02:03
<Rik`>
I think Apple or someone else reported that it was the same for Apple
02:03
<heycam>
maybe I should change do s/operation/method/g in Web IDL
02:03
<othermaciej>
MPEG LA is definitely not a revenue source for us, it is a cost center
02:03
<Rik`>
othermaciej: do you have a neutral link for that ?
02:04
<Rik`>
so everytime I read "Apple has invested so much in H264", I'm thinking that every year it costs Apple a lot of money so they wouldn't be reluctant to use a patent free format
02:04
<othermaciej>
I don't think Apple publishes financials to that degree
02:05
<Rik`>
I'm also thinking that Apple is very fast to introduce new technologies/formats and ditch previous ones
02:05
<erlehmann>
I want a T-Shirt with “My name is MPEG LA and i hate every single one of you.”
02:05
<Rik`>
and I'm also thinking that checking if WebM is not a legal threat is not a quick task
02:07
<erlehmann>
Rik`, good that there is a company whose main business revolves around searching for stuff *hint, hint*
02:07
<erlehmann>
btw: i heard, they also have a patent search ;)
02:08
<TabAtkins>
I don't think our servers are well-designed to do legal searches.
02:08
<gsnedders>
Rik`: You'll never know if something is a legal threat or not until you ship, pretty mufch
02:08
<erlehmann>
shippers gonna ship!
02:08
<othermaciej>
checking if a complicated technology is a legal threat is not in any way fast or easy
02:10
<Rik`>
anyway, there is only two possible scenarios now
02:10
<Rik`>
either everyone moves to WebM
02:10
<erlehmann>
or only some do.
02:10
<erlehmann>
hehehe
02:10
<Rik`>
or MEPG LA is threatened enough by WebM that they make parts of H264 patent free
02:11
<erlehmann>
the word you are looking for is “royalty free”
02:11
<benschwarz>
Hixie: paul_irish made the search not scroll…
02:11
<gsnedders>
Rik`: Or MPEG LA think that IE/Safari supporting it is enough.
02:11
<Rik`>
erlehmann: you're right, sorry
02:11
<erlehmann>
n/p
02:12
<Rik`>
gsnedders: Firefox + Chrome + Opera is at least 40% of internet users in every country so that's not a plausible scenario
02:13
<erlehmann>
south korea disagrees
02:14
<jamesr____>
heycam: ok, i've got something. now to find some decent web hosting to throw it up on...
02:16
<erlehmann>
Rik`, http://blog.mozilla.com/gen/2010/02/24/no-choice-insouth-korea/
02:17
<Rik`>
gsnedders: MPEG LA already changed the end of the free licensing fees because of the pressure Theora/WebM put on them
02:18
<Rik`>
erlehmann: yeah I know, but this is the exception, not the rule
02:18
<heycam>
jamesr____, cool.
02:18
heycam
out for a while
02:18
<erlehmann>
gsnedders, count your chickens when they are hatched.
02:18
<erlehmann>
no, i meant Rik`
02:19
<erlehmann>
sorry
02:19
gsnedders
vanishes off to bed
02:19
<erlehmann>
Rik`, status quo is: apple has the only modern browser not supporting something. conjecturing does not help.
02:19
<Rik`>
the only modern browser not supporting something ?
02:20
<erlehmann>
Rik`, royalty-free codecs in <audio> and <video>?
02:20
<Rik`>
like with IE9, you can install codecs for the media framework of the os
02:21
<Rik`>
you can insall a xiph plugin for quicktime that will play vorbis and theora stuff
02:21
<Rik`>
and I believe Perian has a WebM quicktime plugin in the making
02:22
<erlehmann>
Rik`, we had “whatever codec is installed” for years, with <object>
02:22
<Rik`>
I know, that's useless
02:22
<erlehmann>
or was it <embed>?
02:23
<Rik`>
but saying that Apple will never ship WebM only six months after it was introduced is not serious
02:24
<Rik`>
I mean Firefox is not yet supporting it
02:25
<erlehmann>
Rik`, count the evidence for that they will ship it in another six months. does it seem likely?
02:25
<Rik`>
Android is only supporting it for one or two months
02:25
<jamesr___>
heycam: http://webstuff.nfshost.com/Overview.html
02:25
<Rik`>
in another six months ? maybe in desktop safari, no way in mobile safari
02:25
<jamesr___>
basically it's what you have except for the following:
02:26
<jamesr___>
1.) removed the requirements for updating native animations since i don't think they are implementable. user agents will not be updating native animations on the main thread, so you can't define their updates in terms of steps on the main message loop
02:26
<Rik`>
but I'd say they won't implement it on desktop safari if they can't implement it on mobile
02:26
<jamesr___>
2.) added an element parameter to requestAnimationFrame() for visibility checks
02:27
<jamesr___>
3.) added a setTimeout()-style handle as a return value for request..() and added a clear() to mimic setTimeout()
02:27
<jamesr___>
4.) updated the 'invoke the callbacks' algorithm to take the visibility checks and cancellation into account
02:28
<erlehmann>
Rik`, your argument is moot. i will go to sleep. in before mootles :3
02:44
<Hixie>
http://wetzler.me/dropbox-syncplicity/ applies to HTML too. Whence the 80% rule.
06:29
<heycam>
jamesr___, /Overview.html seems to be the makefile, but /Overview.xml worked
07:15
<benschwarz>
Hixie: Are you able to poke the whatwg server? port 22/ssh isn't responding
10:25
<matjas>
TabAtkins: +2 internets to you, kind Sir.
11:36
<annevk>
oh look
11:36
<annevk>
http://twitterfeed.com/ is what TabAtkins is using for CSS
11:36
<annevk>
I guess we should just use that for the WHATWG blog
11:36
<annevk>
the other features are not really relevant
12:11
<zcorpan>
MikeSmith: <ul><li><a><li></ul> only says "Error: Unclosed elements inside a list."
13:07
<matjas_>
TabAtkins: any chance of including the link to the relevant thread on http://lists.w3.org/Archives/Public/public-css-commits/ for each item in the feed?
14:04
<phrearch>
hi
17:30
<webr3>
okay, I must be being thick - where's the startConversation method of web workers defined?
17:30
<gsnedders>
webr3: startConversation? There is no such thing…
17:31
<webr3>
that's what i thought, but the demo uses it :s and it's not defined in the worker lib either
17:31
<webr3>
see source: http://www.whatwg.org/demos/workers/crypto/page.html
17:32
<webr3>
it's not defined on page or in lib ( http://www.whatwg.org/demos/workers/crypto/libcrypto-v1.js )
17:32
<webr3>
has been in the spec for ages too, but not defined, just in the code of the example,a nd the example works - totally confused me
17:34
<webr3>
see also: http://www.mail-archive.com/whatwg⊙lwo/msg12099.html from 2008 "more powerful api"
17:37
<webr3>
ahh.. the demo's screwed too - var crytoLib (missing p) - and .startConversation is undefined (as it should be i guess) spec/demo code may need an update...
21:47
<webr3>
how would one approach jailing the js environment in the browser (other than web workers)? (basically ecmascript only)
22:04
<Philip`>
webr3: Sounds like http://code.google.com/p/google-caja/
22:06
<webr3>
cheers - and wow, I've not seen "DHTML" referenced in a v long time!