04:01
<GPHemsley>
I don't suppose any WebVTT validators have cropped up yet?
04:11
<GPHemsley>
Also, why are there no samples/examples of WebVTT within the spec?
06:58
<MikeSmith>
https://bugzilla.mozilla.org/show_bug.cgi?id=617528#c1
06:58
<MikeSmith>
Paul Roget asks, "My main concern is security. How can we make it clear that these menu-items are not from Firefox?"
06:58
<MikeSmith>
anybody have thoughts?
06:59
<MikeSmith>
Hixie: ?
06:59
<zewt>
given that they can't actually do anything insecure, what's the actual risk?
06:59
<MikeSmith>
seems like putting an icon of some kind next to each of them
07:01
<MikeSmith>
zewt: I guess Paul doesn't want users to assume the items have any level of trust given to them by the browser
07:01
<zewt>
you can already style fake things that look like browser prompts, eg. make a bar at the top that looks like FF or Chrome's security prompts
07:01
<MikeSmith>
sure
07:02
<MikeSmith>
that doesn't mean browsers should just give up and not attempt to have indicators where they can
07:02
<zewt>
but they also shouldn't have indicators sprinkled everywhere based on objections that aren't substantiated enough
07:02
<MikeSmith>
and in this case it's not just about security anyway
07:03
<MikeSmith>
heh
07:03
<MikeSmith>
geez
07:03
<zewt>
people won't understand "icon == not from the browser"
07:03
<zewt>
and uglifying browser items isn't very nice, either--you should be able to provide your own icon, I'd think
07:04
<othermaciej>
context menus can already be replaced in other ways
07:06
<MikeSmith>
sure
07:07
<MikeSmith>
and this is not something the spec needs to necessarily address at all
07:07
<MikeSmith>
it's an implementation detail, and he's asking for feedback
07:07
<othermaciej>
I understand the general concern about letting web sites hook up things that look like trusted UI
07:07
<zewt>
right
07:07
<othermaciej>
though the context menu is a low risk area
07:16
<zewt>
mental note: never view image on a canvas in FF. heh
07:36
<MikeSmith>
https://github.com/pgriess/node-webworker
07:45
<ben_h>
j #html5
07:46
<ben_h>
it seems someone's taken my slash
08:37
<annevk>
mornings
08:38
<MikeSmith>
おはよう
08:39
<annevk>
I think the common response to something like that is "fuck you too" but lets use Google translate first
08:39
<annevk>
:)
08:41
<Peter->
Goedemorgen, MikeSmith!
08:41
<Peter->
and annevk/others of course
08:42
<MikeSmith>
Peter-: hey hey
08:49
<ben_h>
こんいちわ MikeSmith
08:54
<zewt>
こんにちは :P
08:55
<MikeSmith>
I say こんちや sometimes
08:55
<Dashiva>
ちわす
08:56
<Dashiva>
I guess we're too old for that
08:56
<zewt>
ここは夜だけど
08:58
<MikeSmith>
if you want to hear some fun spoken Japanese, search for ロペ on Youtube and watch some of those
08:58
<MikeSmith>
http://www.youtube.com/view_play_list?p=93D23F5D69152DE6
08:58
<MikeSmith>
e.g., http://www.youtube.com/watch?v=C2Y2QGnButY
08:59
<MikeSmith>
it's gold
09:00
<MikeSmith>
man, the fan on my MikeBook is failing
09:01
<zewt>
heh, i can't follow natural japanese at all
09:01
<MikeSmith>
I hope it lasts long enough until the refresh arrives
09:01
<annevk>
by MikeBook you mean your patched black MacBook?
09:01
<zewt>
if regular people all had the enunciation of voice actors, maybe I'd manage
09:01
<MikeSmith>
annevk: yeah
09:01
<annevk>
wow
09:02
<annevk>
you've had that laptop for as long as I can remember :)
09:02
<MikeSmith>
yeah
09:02
<Dashiva>
zewt: And they speak so quickly
09:02
<MikeSmith>
I dread replacing it because I'll lose all my stickers
09:03
<MikeSmith>
some people from Kyoto speak more slowly
09:03
<annevk>
btw peoples, now is your time to scream: http://lists.w3.org/Archives/Public/public-web-notification/2011Feb/0005.html (well, don't)
09:03
<Dashiva>
まじ半端ねえ、あれ
09:04
<MikeSmith>
I wish somebody would make subtitles for those ロペ videos, because the stuff those guys are saying is really funny
09:06
<MikeSmith>
anyway, I have to drop off for a bit
09:06
<MikeSmith>
back later
10:28
<annevk>
http://blog.whatwg.org/weekly-search-provider-apis
10:36
<zewt>
http://isolani.co.uk/blog/javascript/BreakingTheWebWithHashBangs <- it's impossible to take this guy seriously, heh
10:38
<zewt>
a bunch of mysterious handwaving for the fact that it's absolutely critical to every web app, and "breaking the web" is such a ridiculous overstatement, it undermines his whole argument
10:39
<annevk>
nothing wrong with a bit of hyperbole every now and then
10:40
<Workshiva>
To be fair, I really don't consider news sites web applications
10:40
<zewt>
not when it's taken so far it sounds like satire
10:50
<Workshiva>
In a way it's good
10:50
<Workshiva>
Lets the people still stuck in the 80s realize that javascript is here to stay
10:51
<Workshiva>
And if they don't start working with it, they might not like the results :)
11:09
<annevk>
Workshiva, been reading www-tag lately? :)
11:14
<Workshiva>
annevk: I only ever read that when someone links it
11:55
<Evet>
is JSONP really insecure for private data?
11:55
<MikeSmith>
http://www.bbc.co.uk/blogs/bbcinternet/2011/02/frame_accurate_video_in_html5.html
11:56
<MikeSmith>
"as of this week, after 120 emails, the bleeding edge development versions of several HTML5 implementations (as used in Safari, Chrome, Mozilla and many others) are now fully frame accurate"
12:18
<ben_h>
MikeSmith: awesome!
12:21
<MikeSmith>
it's great to see an article like that giving credit to the actual developers who did the work
12:21
<ben_h>
hell yeah
12:21
<ben_h>
developers are pretty cool these days
12:21
<ben_h>
:)
12:31
<Workshiva>
Evet: If any site you visit can access the data, would you call it private?
12:35
<erlehmann>
opera devs, please support CORS
12:35
<erlehmann>
thanks
12:35
<erlehmann>
._.
12:38
<Evet>
:)
12:41
<MikeSmith>
the comments about Workers in http://dt.deviantart.com/blog/38471599/ are particularly interesting
12:47
<annevk>
hmm, my schedule for today was not very ambitious
12:47
<annevk>
I wonder if I missed something
12:50
<annevk>
that blog post from the BBC is a nice change from the one some time ago
12:52
<Evet>
i wish flash sockets + html5 websockets cover 100% of browsers
12:54
<annevk>
MikeSmith, maybe you can reply to John's email about the draft with information on pubrules?
12:54
<annevk>
or I can do it
12:54
<annevk>
that is, giving a link to http://www.w3.org/Guide/pubrules
13:03
<MikeSmith>
annevk: if you can, that'd be great
13:03
<MikeSmith>
I looked for him on #chromium but not finding him
13:04
<MikeSmith>
the source really should be updated to use the JS respec
13:04
<MikeSmith>
but I'm not sure it's worth stopping to take time to do that now
13:04
<MikeSmith>
though it actually wouldn't take too much time
13:05
<MikeSmith>
I don't know whether he's got the source set up that way because he actually prefers that, instead of editing HTML and using JS, or what
13:06
<annevk>
I know nothing about Respec
13:06
<annevk>
but I'll bring it up
13:06
<MikeSmith>
はい
13:08
<annevk>
MikeSmith, do I need to email out some pubrequests?
13:08
<annevk>
MikeSmith, guess we need shortname approval
13:08
<MikeSmith>
yeah
13:08
<MikeSmith>
actually
13:08
<MikeSmith>
need to mail the chairs list
13:10
<MikeSmith>
"The Chair sends a transition request to the Domain Lead(s) responsible for the group(s) publishing the document, cc'ing w3t-comm⊙wo and chairs⊙wo"
13:11
<MikeSmith>
"The goal of the transition request is to secure an archived record of the Domain Lead(s)' approval of the title, and shortname."
13:12
<MikeSmith>
I will give PLH a heads-up about this today
13:12
<MikeSmith>
oh
13:12
<MikeSmith>
it's a holiday in the US I guess
13:12
<MikeSmith>
he'll probably be on for a bit anyway
13:16
<annevk>
okay, I can do that
13:17
<annevk>
first ever email to chairs I think
13:17
<annevk>
back to Member-only territory brrr
13:18
kennyluck
wonders whether folks here know how pronounce はい
13:18
<wilhelm>
kennyluck: Hai.
13:19
<wilhelm>
(は on its own is usually “wa”, though.)
13:19
<Peter->
Iie!
13:20
<kennyluck>
for reasons I don't know, this word gives me a funny image.
13:20
<MikeSmith>
annevk: chairs list really should be public
13:20
<MikeSmith>
i guess I shouldn't say that
13:20
<MikeSmith>
I guess the issue with doing that is, some chairs still have some member-only groups…
13:21
<MikeSmith>
maybe we should have a chairs list for all the public groups, and a chairs list for all the member-only groups
13:33
<annevk>
MikeSmith, emailed
13:33
<annevk>
MikeSmith, included request for being able to copy a public list going forward
13:34
<annevk>
MikeSmith, hopefully I got it all right, I didn't look up the chair guidelines doc as I couldn't locate it quickly
13:36
<MikeSmith>
I just read it, it's fine as-is
13:36
<MikeSmith>
I suspect plh answer will be, you are free to copy to www-archive if you want
13:37
<annevk>
mkay
13:37
<annevk>
that is something
14:17
<annevk>
did http://esw.w3.org/HTML/history get lost somehow?
14:17
<annevk>
just found that again via hsivonen slides who just tweeted that w3.org never go wrong
14:18
<annevk>
w3.org URLs*
14:18
<annevk>
hsivonen's*
15:07
<Philip`>
Does anyone have data on how frequent PAC (proxy auto-config) usage is?
15:40
<MikeSmith>
annevk: hmm
15:40
<MikeSmith>
if so, my fault
15:40
<MikeSmith>
damn
15:40
<MikeSmith>
not lost
15:40
<MikeSmith>
just different URL
15:41
<MikeSmith>
eh?
15:41
<MikeSmith>
something borked
15:42
<MikeSmith>
annevk: it's at http://www.w3.org/html/wg/wiki/History
15:43
<MikeSmith>
now I just need to figure out how to get http://www.w3.org/wiki/HTML/history to redirect to there
15:45
<MikeSmith>
hmm, http://www.mediawiki.org/wiki/Help:Redirects
15:45
<MikeSmith>
but that can only redirect to a page on the same wiki, not to an arbitrary URL
15:46
<thiessenp>
(sorry off topic but :) Can anyone recommend a good JS crypto library (RSA, BigInt etc.)?
15:48
<MikeSmith>
thiessenp: not off topic :)
15:48
<MikeSmith>
but I don't know of one personally
15:48
MikeSmith
looks around for Adam Barth
15:48
<MikeSmith>
wirepair: ↑
15:49
<zcorpan>
thiessenp: i think window.crypto.getRandomValue or what it's called is being added to address that
15:51
<annevk>
is it just me or has interest in WHATWG Weekly waned somewhat?
15:51
<annevk>
I wonder if it should maybe done in a somewhat different way
15:52
<MikeSmith>
how do you measure interest in it?
15:52
<annevk>
tweets / comments / feedback via IRC/email
15:52
<Peter->
get used to it :p
15:52
<annevk>
haven't looked at stats recently
15:53
<Peter->
comments on my webkit posts are lowish lately as well, as are the tweets
15:53
<annevk>
though I did notice subscriber count went up for the feed in Google reader
15:53
<Peter->
view counts are still rising though
15:53
<kennyluck>
I have great interest in it, personally.
15:54
<MikeSmith>
what is the anti-TPM clause in CC licenses?
15:54
<annevk>
Peter-, interesting
15:54
<nimbupani>
annevk: maybe coz we know you are leaving :P
15:59
<MikeSmith>
so next month marks the seventh anniversary of the whatwg.org domain being registered
15:59
<MikeSmith>
three months prior to the infamous W3C Workshop on Web Applications and Compound Document
16:00
<MikeSmith>
s
16:01
<annevk>
a year for each capital
16:02
<MikeSmith>
capital?
16:03
<annevk>
that is quite long
16:03
<MikeSmith>
ah
16:04
<annevk>
guess I'm getting close to a decade of something
16:04
<annevk>
MikeSmith, you got it? heh; useless number game :p
16:05
<MikeSmith>
for me, it's close to 10 years since I moved to Japan
16:05
<MikeSmith>
even more: my daughter just turned 13 years old this weekend
16:06
<MikeSmith>
oh, "technological protection measures"
16:06
<MikeSmith>
jesus
16:07
<MikeSmith>
why don't they just say DRM?
16:07
<MikeSmith>
acronym proliferation
16:28
<Lachy>
MikeSmith, TPM is more generic than DRM. AIUI, TPM can be any protection measure, either analogue or digital.
16:28
<MikeSmith>
yeah well
16:30
<MikeSmith>
one can always mint more precise acronyms I guess
16:30
<MikeSmith>
but point taken
16:30
<MikeSmith>
Lachy: hey
16:30
<MikeSmith>
btw, where you been, man?
16:30
<MikeSmith>
you've become a ghost
16:30
<MikeSmith>
you working on some secret project
16:37
<Lachy>
I'm working on lots of stuff internally at opera and haven't had too much time to follow what goes in HTML lately
16:38
<Lachy>
MikeSmith, also, because of the way bugzilla is used on public-html, it makes it much more difficult than it should be to actually follow anything.
16:39
<MikeSmith>
yep
16:39
<MikeSmith>
not my choice
16:39
<Lachy>
I really wish we could go back to how it was before bugzilla became used as a poorly designed forum, that I hate even more than real forums
18:04
<yecril71pl>
Why is document in the lexical scope of code in event handler attributes?
18:05
<yecril71pl>
like onclick="alert (write === document. write)"?
18:08
<bga_>
strange
18:08
<bga_>
`true` in chrome11
18:10
<yecril71pl>
and in Konqueror
18:10
<bga_>
in opera too
18:11
<yecril71pl>
Since browsers are unanimous, there must be some method in this madness
18:13
<bga_>
but
18:13
<bga_>
var a = document.createElement('a'); a.setAttribute('onclick', 'alert (title === document.title)'); a.onclick()
18:14
<bga_>
false
18:14
<yecril71pl>
in event handler attributes only
18:18
<yecril71pl>
DOM-Level-2-Events 1.3.2.: No technique is provided to allow HTML 4.0 event listeners access to the context information defined for each event.
18:22
<yecril71pl>
Neither HTML 4 nor DOM Level 1 discuss binding of events
18:26
<Philip`>
http://www.whatwg.org/specs/web-apps/current-work/multipage/webappapis.html#event-handler-content-attributes
18:26
<Philip`>
"Let Scope be the result of NewObjectEnvironment(the element's Document, the global environment)."
18:26
<Philip`>
Sounds like that might be it
18:30
<yecril71pl>
Thanks, that cuts it. Any background? Was it in DOM 0?
18:32
<Philip`>
I have no idea
19:45
yecril71pl
found it: http://docstore.mik.ua/orelly/webprog/jscript/ch19_01.htm, 19.1.6: Scope of Event Handlers
19:46
<yecril71pl>
Note that it actually pulls all properties of containing elements, not just Document
19:46
<yecril71pl>
HTML5 takes a more modest apporach
19:46
<yecril71pl>
HTML5 takes a more modest approach
19:47
<yecril71pl>
It is like extended "this", and it is more logical this way.
19:48
yecril71pl
means more consistent
19:48
<yecril71pl>
So why did HTML5 choose to include that Document as an exception?
19:52
<nonge_>
Would it be 'right' to use <canvas> for an "eyecandy only" animation? … wich would normally be the job of CSS, but canvas-like stuff is not possible with it :/
19:56
<yecril71pl>
Flanagan’s book, however, does not provide any further references; it does not even say who Flanagan is.
20:05
<yecril71pl>
the technical term for it is "augmented scope chain"
20:06
<yecril71pl>
http://www.jibbering.com/faq/names/event_handler.html disagrees, and includes only FORM and document.
20:12
<nonge_>
for example, using an empty <canvas></canvas>, which only use is "simulating" an background-image (with z-index)
20:26
<annevk>
Gecko has document.contentType apparently
20:27
<annevk>
we have some bug that requests it (not a compat issue)
20:27
<annevk>
guess that should be in HTML, given that it defines page loading and such
20:36
<annevk>
going through TAG minutes...
20:36
<annevk>
http://www.w3.org/2001/tag/2011/02/08-minutes#item03 still actively considering DTDs...
20:36
<annevk>
for HTML
20:38
<annevk>
http://www.w3.org/2001/tag/2011/02/09-minutes seems to cover a bunch of HTML stuff too
20:44
<annevk>
couldn't find much interesting stuff
20:45
<annevk>
at least what interests me
20:45
<annevk>
I believe that does include web architecture, but apparently a different kind
20:48
<AryehGregor>
Is it just me, or can I not actually click on this button in Opera while I have the web inspector thing loaded? http://aryeh.name/spec/editcommands/test/bold.html
20:48
<AryehGregor>
Because that seems . . . well, makes it sort of hard to use.
20:50
<jacobolus>
annevk: did you intend to have the latest whatwg blog post end with a sentence fragment?
20:50
<jacobolus>
annevk: it seems like there was supposed to be something after the bit that's there
20:51
<jacobolus>
(http://blog.whatwg.org/weekly-search-provider-apis)
21:02
<annevk>
poor wording
21:03
<annevk>
fixed
21:03
<annevk>
ta jacobolus
21:03
<jacobolus>
:)
21:19
<benschwarz>
Its out! http://germanforblack.com/articles/html5-for-web-developers
21:21
<annevk>
benschwarz, "Thankyou" -> "Thank you"?
21:22
<AryehGregor>
I think I have finally figured out a workable and sane initial definition for execCommand("bold").
21:22
<benschwarz>
annevk, nice pickup, thanks
21:22
<AryehGregor>
It has taken me something like 15 hours.
21:23
<annevk>
was that the command you started out with?
21:23
<AryehGregor>
Yes.
21:23
<AryehGregor>
So it includes a bit of time to get familiar with execCommand() itself.
21:23
<annevk>
figuring out the model is typically the hardest bit
21:23
<AryehGregor>
In this case, I spent most of the time figuring out a sane way to sane "make all the stuff in this Range bold/unbold".
21:23
<AryehGregor>
s/to sane/to say/
21:24
<AryehGregor>
Ranges are *way* too complicated for this.
21:24
<zewt>
s/for this//
21:24
<AryehGregor>
:)
21:24
<annevk>
doesn't execCommand operate on Selection?
21:24
<AryehGregor>
Yes, which is defined in terms of Ranges.
21:25
<annevk>
right, just saying there could be multiple making it more fun potentially
21:25
<AryehGregor>
It's a one-to-one mapping in every browser except Firefox, which allows multiple Ranges per Selection, but that turned out to be easy to define away.
21:25
<annevk>
hopefully you do not allow overlapping ranges within a selection
21:25
<zewt>
i hate multi-ranges, because it makes "Read more at:" clipboard hijacking way too easy :| (though it's possible to do without it)
21:25
<AryehGregor>
That is allowed, but the behavior is simple anyway.
21:26
<AryehGregor>
If you have a text node "Foobar", and the first range contains "Foob" and the second contains "obar", and you bold it, then Firefox produces something like "<b>Fo</b>ob<b>ar</b>".
21:26
<AryehGregor>
Or, no.
21:26
AryehGregor
tests that
21:26
<AryehGregor>
Anyway, I just say to style the Ranges one by one, in order.
21:26
<annevk>
and the ranges are live?
21:27
<AryehGregor>
. . . yeah, well, that's a separate question.
21:27
<AryehGregor>
No one has even attempted to define how Ranges behave under DOM mutation.
21:27
<annevk>
well it sort of matters if you are gonna operate on them :)
21:27
<AryehGregor>
Which actually is a practical problem for execCommand(), yes.
21:27
<AryehGregor>
It's occurred to me.
21:27
<AryehGregor>
But first things first.
21:27
<annevk>
looking forward to reading through this one day
21:27
<annevk>
gonna sleep now
21:27
<AryehGregor>
Here's the draft now: http://aryeh.name/spec/editcommands/editcommands.html
21:28
<AryehGregor>
Hixie, I could use a list of execCommand() usages, to figure out which features people actually use and how they use them.
21:29
<annevk>
good times
21:29
<AryehGregor>
Wow, there are a heck of a lot of them.
21:29
<annevk>
btw
21:29
<AryehGregor>
This will take me weeks for sure.
21:29
<annevk>
you should contact Roland
21:29
<AryehGregor>
Who's Roland?
21:30
<annevk>
he works on contenteditable/execCommand testing
21:30
<annevk>
at least some part of his time
21:30
<annevk>
/msg'd you detaild
21:30
<annevk>
s
21:31
<AryehGregor>
Oh, is that related to the stuff that Google submitted to the whatwg at some point?
21:31
<AryehGregor>
They seemed to be concerned with the stuff I didn't start with, so I put it aside for the time being.
21:31
<AryehGregor>
But thanks, I'll contact him.
21:31
<annevk>
it's very much about what exactly execCommand should do given some situation
21:32
<annevk>
http://groups.google.com/group/browserscope
21:34
<AryehGregor>
Thanks, I'll take a look.
22:30
<benschwarz>
Hixie — maybe one you can help with https://github.com/benschwarz/developers.whatwg.org/issues/#issue/33