01:27
<Hixie>
can someone on windows do me a favour?
01:27
<Hixie>
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/582
01:28
<Hixie>
if you change your highlight colour, does IE update the rendering?
01:28
<wirepair>
ie version matter?
01:28
<Hixie>
probably not, but the later the better
01:28
<wirepair>
it 'modified the page to prevent cross site scripting' ;>
01:28
<Hixie>
(not sure how you change the highlight colour on windows)
01:28
<Hixie>
good times
01:28
<Hixie>
go to http://software.hixie.ch/utilities/js/live-dom-viewer/
01:29
<Hixie>
and click "download"
01:29
<wirepair>
yes
01:29
<wirepair>
it updates automatically
01:29
<Hixie>
nice
01:29
<Hixie>
thanks
01:29
<wirepair>
np
01:30
<Hixie>
no need to reload or anything right?
01:30
<peol>
wouldn't be very automatic then, would it? :)
01:30
<wirepair>
nope
01:30
<TabAtkins>
Hixie: I knew it!
01:30
<Hixie>
well chrome changes "automatically" when you next cause it to do a restyle (it also changes it to blue, regardless of what you picked, which is a crazy bug)
01:30
<wirepair>
changes as you type
01:31
<Hixie>
wirepair: wait, i mean change the highlight colour in the OS settings
01:31
<wirepair>
Ohhh
01:31
<Hixie>
wirepair: not change the display when you type in the form :-)
01:31
<wirepair>
sorry,
01:31
<Hixie>
TabAtkins: :-P
01:32
<Hixie>
wirepair: i think it's in display appearance settings or something
01:32
<Hixie>
wirepair: not sure exactly how to change the highlight colour
01:32
<wirepair>
yeah... me either
01:32
<TabAtkins>
I was going to poke you and laugh because I saw 2 or 3 emails today where you replied "Indeed.", but then I became certain that you were doing it on purpose.
01:32
<wirepair>
sec
01:32
<Hixie>
TabAtkins: :-P
01:34
<wirepair>
yup, it changes
01:34
<Hixie>
interesting
01:35
<Hixie>
so you don't have to even focus the browser?
01:35
<Hixie>
just hitting "apply" or whatever makes the colour change?
01:35
<wirepair>
desktop -> apperence -> advanced -> selected items to change
01:35
<wirepair>
yeah
01:35
<Hixie>
nice
01:35
<Hixie>
thanks man
01:35
<wirepair>
after you hit apply, it resets the entire desktop
01:35
<wirepair>
np
01:35
<wirepair>
and the rendered view included obviously
01:35
<Hixie>
k
01:36
<Hixie>
do you know what version of IE and Windows that was by any chance?
01:36
<wirepair>
ie8
01:36
<wirepair>
8.0.6001.18702
01:36
<wirepair>
to be exact
01:36
<Hixie>
on XP?
01:37
<wirepair>
yes
01:37
<Hixie>
k
01:37
<Hixie>
thanks again
01:37
<wirepair>
sp3, fully patched
01:37
<TabAtkins>
If you want some W7 experience, I can try it.
01:37
<TabAtkins>
Ooh, resource packages! Yay!
01:38
<wirepair>
now.. to figure out how to automate epiphany ;<
01:38
<wirepair>
the main window has no identifier that i can match on, always takes the <title>whatever</title> value for the window text
01:38
<wirepair>
ps. automating ui's in *nix is *painful*
01:41
<Hixie>
apparently tantek didn't like the ordering by first name and decided to reorder the speaker list on the wiki by last name :-|
02:01
<dbaron>
Mmmm. I like Hixie's writing about colors. He uses "colour" when talking about the concept and "color" when talking about technical terms in the spec. :-)
02:02
<TabAtkins>
Hixie's not in favor of internationalizing the CSS property names then, I suppose. ^_^
02:09
<aho>
i still want #rrggbbaa and #rgba :>
02:09
<aho>
by the way
02:09
<aho>
<:
02:09
<Hixie>
in css?
02:09
<TabAtkins>
I know, I just need to write Colors 4 so Hyatt will let me commit it.
02:09
<aho>
yes
02:09
<aho>
0-255, 0-255, 0-255, 0-1 is just silly
02:10
<TabAtkins>
My survey of pages showed that the compat impact of turning it on for HTML colors should be very minimal as well.
02:11
<TabAtkins>
(Just so you know, Hixie, so you can change the color parsing algorithm appropriately when I land the patch.)
02:11
<Hixie>
i do not intend to change that algorithm ever again
02:11
<Hixie>
there's no reason for anyone to use legacy colours
02:12
<aho>
?:>
02:12
<TabAtkins>
It's somewhat annoying to patch webkit to use a different codepath for attributes and CSS properties.
03:23
<boblet>
can anyone tell me what’s up with @profile? specifically why it was dropped from HTML5 and what’s up with the separate @profile spec?
06:56
<annevk5>
Hixie, so how do you reckon we add multiplexing later? if it is added frames get prefixed by a byte that is an identifier?
06:57
<Hixie>
for example
06:58
<Hixie>
ideally we'd have a multiplexed connection connect to some predefined resource name or something
06:59
<Hixie>
but that might bo harder to do later
06:59
<Hixie>
and not so backwards-compatible
06:59
<Hixie>
so maybe we just try to connect without multiplexing, and if the server says it supports it, we do the channel open separately
06:59
<Hixie>
and just assume that everything on that origin supports it
07:03
<annevk5>
so the server declares multiplexing support in its handshake and the client will start transmitting special frames and remember that server supports multiplexing for the duration of any ongoing connections to that server?
07:04
<annevk5>
(well, not any, *that* ongoing connection, and any new queued will use the same one)
07:10
<Hixie>
yeah
07:10
<Hixie>
seems unlikely to me that we'll find it's actually necessary though
07:12
<annevk5>
it does seem like a better approach to keep simple servers simple at least
07:12
<annevk5>
without having to put all kinds of provisions into framing from the start
07:13
<Hixie>
i love the juxtaposition of some of the proposals on the hybi list with the simultaneous argument that we shouldn't use variable-width framing because it's "too complicated"
07:16
<annevk5>
yeah...
07:22
<annevk5>
MikeSmith, added NE to 2 bugs!
07:22
<MikeSmith>
annevk5: thanks
07:24
<annevk5>
I'll do it for some more; I'm still in bed and feel rather lazy
07:42
<annevk5>
okay, getting bored with this
07:50
<volkmar>
what NE means?
07:50
<annevk5>
it means non-editorial or controversial (if it's editorial but controversial)
07:51
<annevk5>
it also ensures that public-html gets updates as the bug progresses
07:51
<annevk5>
so everyone is in the loop rather than just those subscribed to the bug
07:51
<volkmar>
ok :)
09:22
<hsivonen>
sigh. even Windows Live developers sniff for Firefoxness instead of sniffing for Geckoness
09:44
<annevk5>
Hixie, I think it might help if a separate thread is started on how multiplexing can be added later
09:44
<annevk5>
Hixie, I could email what I sketched out earlier
09:44
<Hixie>
go for it
09:44
<Hixie>
i have to go sleep soon
09:44
<annevk5>
Hixie, I believe some people have concerns with changing the framing, but I'm not sure what they are
09:45
<annevk5>
kk
09:45
<Hixie>
me either
10:00
<annevk5>
oh look
10:00
<annevk5>
http://www.w3.org/TR/css3-mediaqueries/ has been republished too now
10:00
<annevk5>
via myakura
10:12
<annevk5>
poked the hybi list
10:12
<annevk5>
i am surprised there is still no reply
10:15
<annevk5>
othermaciej, in "Re: [whatwg] HTML resource packages" your last sentence got truncated (though I suppose it is clear what you wanted to say)
10:17
<othermaciej>
annevk5: d'oh
10:17
<othermaciej>
I hate when I do that
10:17
<jgraham>
othermaciej: Isn't that like once every other message? :)
10:18
<annevk5>
hsivonen, lol
10:18
<annevk5>
and retweeted
10:25
<othermaciej>
jgraham: I do it a lot when I reply to emails in haste
10:25
<othermaciej>
jgraham: because I write chunks of my reply out of order, and I don't re-read when I am rushing
10:25
<othermaciej>
I do it less than I used to because I am more careful in wording my emails most of the time
10:26
<jgraham>
othermaciej: Maybe I should make a graph so there is some data to back up that claim :)
10:27
jgraham
finds that if he writes emails in haste they are just ill-thought-out nonsense so tries to avoid that
10:27
<jgraham>
(so you are ahead of me there)
10:29
<othermaciej>
I can sometimes jot off a sensible thought quickly
10:29
<othermaciej>
but I prefer to take the time to think it through and review my words
10:29
<othermaciej>
if I write quickly, I tend to be too verbose
10:41
<hsivonen>
annevk5: the DOM of Hotmail is no laughing matter :-(
10:48
<Philip`>
Are there any large complex dynamic web apps with nice clean readable well-structured markup and scripts?
10:50
<workmad3>
probably... but it would surprise me if they serve it to the public like so
10:51
<workmad3>
large dynamic apps seem to have moved towards removing extraneous white-space from output to minimise every last bit of output
11:31
<hsivonen>
workmad3: Hotmail is minifying each script and style, but it's definitely not minimizing the number of <script> and <style> elements
11:31
<hsivonen>
or the number of <div>s
11:31
<hsivonen>
it has a number of style elements each of which is a child of a div inside head!
15:20
<Rik`>
wow, the night :target {} rule on the HTML5 spec is… weird
15:20
<Rik`>
s/night/new
15:25
<AryehGregor>
Rik`, :target has existed for ages, hasn't it?
15:26
<Rik`>
AryehGregor: I mean the style used on the spec
15:26
<Rik`>
combining a big box-shadow and text-shadow in yellow
17:40
<franksalim>
would working examples of amqp, stomp, xmpp, rfb, etc. over WebSocket (75/76) be sufficient to demonstrate that protocol features like framing, chunking, and multiplexing can be done in the application layer (JavaScript)?
17:42
<annevk5>
you could try on the list
17:42
<Philip`>
I don't think possibility is a useful thing to demonstrate
17:43
<Philip`>
You could demonstrate using WebSockets as a transport mechanism for IP packets and then do everything else in the application layer
17:43
<franksalim>
has been done, then
17:43
<franksalim>
this is all working code
17:44
<Philip`>
If the demonstration demonstrates the complexity of doing those things or the performance impact then that'd probably be a useful thing to demonstrate, though
17:44
<franksalim>
annevk5, i will try on the list
17:44
<franksalim>
Philip`, the biggest wins are in flexibility and interoperability
17:45
<franksalim>
without a negative impact on performance. complexity moves from browser native code to JavaScript libraries
17:46
<franksalim>
I strongly suspect the HyBi discussion is being colored by a lack of faith and experience in JavaScript
17:46
<franksalim>
which is tragic considering what WebSocket is intended for
17:48
<franksalim>
/unsolicited opinion
17:52
<annevk5>
lol
17:52
<annevk5>
just read an email where gregw claims he tries to keep it simple
17:53
<Philip`>
Someone who's trying to design a simple nuclear power station will make something very different to someone who's trying to design a simple bicycle storage building
18:15
TabAtkins
would like to see some simple examples of IDB usage, because it smells overengineered.
19:03
<dandaman>
<body onload="alert('errho');setWidth();$('bodyId').style.display='';" id="bodyId" onscroll="scrolltop()" style=" margin:0 auto;"> gives me an alert
19:03
<dandaman>
but when i put that alert in the first line of my setwidth() function, nothing comes up
19:03
<dandaman>
the case matches the function too
19:03
<TabAtkins>
Hixie: Damn your precise yet intricate prose. After considerable pain, I find that everything you wrote about margin-collapsing is exactly correct, but requires reverse-engineering for a normal mortal to decode.
19:03
<dandaman>
what the hell :(
19:04
<TabAtkins>
dandaman: Are you using jQuery or Prototype?
19:05
<dandaman>
mootools
19:06
<TabAtkins>
Okay, I was just wondering about the $('bodyId') call, because it would cause an error in jQuery.
19:06
<TabAtkins>
Do you see any errors in your error console?
19:06
<dandaman>
hmm lemme check out firebug
19:06
<dandaman>
Access to restricted URI denied.
19:06
<dandaman>
thats what my css tab says
19:06
<dandaman>
dunno what the hell that means
19:07
<TabAtkins>
Look in the javascript tab.
19:07
<dandaman>
public_smo_scripts.js gives me the same access restricted thing
19:08
<dandaman>
i can get rid of that
19:08
<dandaman>
forgot i left that when i was playing around with a tutorial
19:08
<dandaman>
no more access restricted
19:08
<TabAtkins>
I'm looking for any errors that suggest that setWidth() isn't being called.
19:08
<dandaman>
but it still isnt working
19:08
<Hixie>
TabAtkins: margin collapsing is the result of dbaron and i thinking CSS1 and CSS2 were infallible words and thus us working out the only set of rules that could completely solve them. Then we wrote them explicitly and that was CSS2.1.
19:09
<dandaman>
yea, i mean the setWidth() was working before i added a bunch of $('variable').style.display='none' to onclicks
19:09
<Hixie>
TabAtkins: then we learnt that we could change specs and we were like "ohhhh. we should have just made margin collapsing simpler. oops."
19:09
<TabAtkins>
Hixie: Right. And at this point it can't be changed anymore, as far as I can tell.
19:09
<dandaman>
but i made sure those were all legit changes
19:09
<Hixie>
TabAtkins: yeah :-(
19:09
<TabAtkins>
Hixie: So now I'm trying to see how I can rephrase things and add informative notes so that it's actually parseable by a human.
19:10
<dandaman>
oh wow
19:10
<dandaman>
left out a ' in a bunch of these
19:10
<dandaman>
there we go
19:10
<TabAtkins>
dandaman: If you didn't see any errors about those, then you're not looking at the right error console.
19:10
<Hixie>
TabAtkins: i wouldn't rephrase it, every last comma is meaningful and has dastardly implications
19:10
<Hixie>
TabAtkins: but informative notes would be fantastic! with diagrams!
19:11
<TabAtkins>
There will be so many diagrams!
19:12
<TabAtkins>
dandaman: In FF try Tools->Error Console.
19:14
<dandaman>
i cant use ff because it doesnt support css3 :\
19:14
<dandaman>
or rather i dont because google chrome doesnt need plugins
19:14
<dandaman>
do you know where the error console is on chrome?
19:14
<TabAtkins>
dandaman: ... (a) That's quite wrong, and (b) you said you were using Firebug.
19:15
<dandaman>
chrome's got firebug too
19:15
<TabAtkins>
No, it's got a Web Inspector.
19:15
<dandaman>
does the new firefox come with css3 support?
19:15
<TabAtkins>
Or rather, "Developer Tools".
19:16
<TabAtkins>
In there, just go to the Console tab.
19:16
<dandaman>
kk, but there is a firebug add-on for chrome in case you were curious, im using it right now :)
19:16
<TabAtkins>
Huh. Weird.
19:16
<dandaman>
found the console, thanks
19:16
<TabAtkins>
Anyway, yes, Firefox supports plenty of CSS3 stuff.
19:17
<dandaman>
guess im using an old version then
19:17
<dandaman>
i like chrome better anyway
19:17
<TabAtkins>
What version are you using, and what do you consider "CSS3"?
19:17
<dandaman>
so much faster
19:18
<dandaman>
well the main thing firefox wont do for me is sliding transitions
19:18
<dandaman>
-transition:all 1.0s ease-in-out;
19:18
<dandaman>
upon further inspection it looks like firefox has a webkit for those transistions though
19:19
<dandaman>
i have 3.6.7 though
19:19
<TabAtkins>
Correct, 3.6 doesn't yet implement transitions.
19:19
<dandaman>
doesnt look like it implements gradients either :\
19:19
<TabAtkins>
Firefox implements gradients just fine.
19:19
<dandaman>
wait
19:19
<dandaman>
this is weird
19:19
<dandaman>
the bg gradient is fine
19:19
<TabAtkins>
They even implement it according to the spec.
19:19
<dandaman>
but my button gradient isnt working haha
19:20
<dandaman>
oh, i didnt put the -mozilla gradient
19:20
<dandaman>
only webkit
19:20
<dandaman>
they really need to standardize this shit
19:20
<TabAtkins>
"firefox has a webkit for those transitions" <--- What does that mean?
19:20
<Hixie>
hmm... counter proposals to ISSUE-110
19:20
<dandaman>
as in you need to put down, -moz-transition:all 1.0s ease-in-out;
19:20
<dandaman>
not just -iwebkit...
19:20
<Hixie>
i wonder what counter proposals people would be willing to vote for
19:20
<TabAtkins>
Yes, why would you think that -webkit would work for Firefox?
19:22
<Peter->
-webkit worked for IE Mobile, -khtml works for webkit (and vice versa)
19:22
<Peter->
prefixes are a mess
19:22
<miketaylr>
FF4.0b2 has pretty decent support for transitions and transforms
19:23
<dandaman>
i didnt, i just forgot that those were the only css3 things i had in a lot of cases
19:23
<dandaman>
i made my stylesheet like 3 weeks ago
19:23
<dandaman>
kinda forgot about it
19:26
<TabAtkins>
-webkit only worked only worked in IEMobile for a short while, then they quickly backed off because it was a horrible idea.
19:26
<TabAtkins>
I wasn't aware that -khtml worked for webkit, but webkit did come from that, so shrug.
19:26
<dandaman>
webkit seems to work for chrome though
19:26
<TabAtkins>
Yes, because both Chrome and Safari use the Webkit rendering engine.
19:29
<dandaman>
how come mozilla doesn't just jump on and use webkit instead of complicating matters and having their own engine?
19:29
<miketaylr>
:/
19:30
<TabAtkins>
Because more engines = mo' betta'
19:30
<TabAtkins>
(Competition is good in the rendering engine world.)
19:31
<AryehGregor>
Also, it would require a huge amount of work to switch completely, and their programmers would all have to get familiar with WebKit when they're already familiar with Gecko.
19:31
<AryehGregor>
WebKit and Gecko do borrow from one another liberally where it makes sense, but it doesn't make sense for either to just use the other engine wholesale.
19:33
<dandaman>
hmmm, good to know
19:34
<dandaman>
i forget that the discomfort of non-standardization promotes progress
19:48
<dandaman>
anyone know of a channel for mobile web development?
19:54
<gsnedders>
TabAtkins: I do like how protective Google is of their performance crown at JS, speaking of competition.
19:54
gsnedders
heard that there were people running around after Carakan shipped trying to beat it
19:57
<AryehGregor>
gsnedders, that's totally unlike other browser vendors, right?
19:57
<AryehGregor>
I mean, doesn't like every single browser release these days claim to have the fastest JS engine in existence?
19:57
<gsnedders>
AryehGregor: I only got the impression of real panic in Google-land when Carakan shipped
19:57
<AryehGregor>
How did you get that impression?
19:57
gsnedders
shrugs
19:58
<AryehGregor>
I can't see them panicking *too* much. I mean, it took them less than 24 hours after release to eclipse Opera's market share, if I remember correctly.
20:00
<gsnedders>
Having a link on Google.com is kinda benefical, I guess :)
20:01
<Philip`>
We need to have as many benchmarks as there are browser engines
20:01
<Philip`>
then everyone can be the fastest browser in the world on their own JS benchmark
20:01
<gsnedders>
It's nice to be able to claim we're fastest on someone else's though :)
20:01
<Philip`>
and then they can stop worrying about it and start optimising things that real users will care about
20:02
<AryehGregor>
Philip`, why do you think they'd ever do that?
20:02
<AryehGregor>
If it's not measurable . . .
20:02
<AryehGregor>
The measurability increases the competition.
20:02
<AryehGregor>
Because everyone can compete on every last millisecond, instead of just saying "meh, we're fast enough".
20:03
<Peter->
IE9 Platform Preview 4 just got released
20:03
<Peter->
rather unexpected
20:03
<Peter->
http://ie.microsoft.com/testdrive/
20:03
<gsnedders>
Peter-: Any changelog?
20:03
<Peter->
not yet, they're still updating pages
20:04
<gsnedders>
95/100 on Acide
20:04
<gsnedders>
*Acid3
20:05
<Peter->
they've madea a demo called Hamster Dance Revolution
20:05
<Peter->
that sounds cool already
20:05
<gsnedders>
I guess my guess of them actually aiming for 100/100 is still on track
20:05
<Peter->
they stated not to implement SMIL
20:05
<gsnedders>
They've stated it's not a current aim.
20:05
<gsnedders>
Not that it won't be later on in the release cycle.
20:06
<gsnedders>
(Or at lesat that was the case when I last looked)
20:07
<gsnedders>
(My theory is they're still deliberately not promising anything, not that they aren't planning things)
20:08
<Peter->
"Microsoft doesn't want to support SMIL because the HTML+Time declarative animation it already has never gets used" according to zdnet
20:08
<Peter->
but we'll know in a few minutes I suppose
20:08
<AryehGregor>
Peter-, is that to shut up the guy on Slashdot who has as his sig "HTML5 will displace Flash when it can do [link to that annoying badger/snake video]"?
20:08
<gsnedders>
I saw something about not shipping final until next year
20:08
<Peter->
I do not think so, it was posted early June
20:08
<gsnedders>
Peter-: That's only part of SMIL, though, no?
20:09
<gsnedders>
s/part/a small part/
20:09
<Peter->
to be honest, I'm not sure
20:09
<Peter->
http://ie.microsoft.com/testdrive/info/ReleaseNotes/Default.html
20:09
<gsnedders>
I can't remember to much detail about HMTL+TIME
20:10
<Peter->
refresh a few times, some nodes on their server cluster already got updated
20:10
<Peter->
DOMException objects seems to be the largest improvement
20:10
<Philip`>
Looks like mostly DOM-implementation changes
20:11
<Philip`>
and no significant new features
20:11
<AryehGregor>
What real-world features do they still need to add, practically speaking? Somewhat better SVG support, anything else?
20:11
<Peter->
They're talking about updates to the DOM hierarchy, "Changes align IE9 standards mode with W3C standards", meaning an HTML5 parser?
20:11
<gsnedders>
"var g = document.getElementById; g(‘test’); // Now throws an exception"
20:11
<gsnedders>
What spec defines that?
20:12
<AryehGregor>
gsnedders, I bet the ECMAScript spec defines that you should treat it as a syntax error when someone tries to use curly quotes.
20:12
<gsnedders>
[[Call]] gone from HTMLCollection
20:12
<gsnedders>
AryehGregor: No, that's an impl allowable extension ;P
20:12
<AryehGregor>
Peter-, if they did an HTML5 parser, they'd bill it higher, I'm thinking, because it's a really big backend change.
20:14
<Peter->
Probably, yes
20:14
<Peter->
they did pass Safari in sunspider
20:15
<Peter->
only by 2ms though
20:15
<gsnedders>
Still slower than Opera 10.60 though :P
20:21
<Peter->
Four of the five ACID3 tests they fail are SMIL, one is SVG fonts
20:29
<paul_irish>
AryehGregor: they're still missing css transitions, transforms, box shadow, gradients, geolocation.... and then web workers, indexeddb, app cache, flexbox, file api, webgl..
20:29
<AryehGregor>
Half that stuff isn't even close to being properly standardized yet.
20:30
<AryehGregor>
IE isn't going to implement experimental stuff with a two-year release cycle.
20:30
<paul_irish>
regardless of that, half that stuff is very much real-world nowadays.
20:31
<Peter->
paul_irish: they implemented CSS values for rotations/timing, which is rather useless if you don't plan to do anything with them
20:34
<paul_irish>
Peter-: rotations/timing ?
20:34
<TabAtkins>
Presumably he means the deg, s, etc. units.
20:34
<Peter->
deg/grad/rad/turn/ms/s
20:35
<Peter->
http://msdn.microsoft.com/en-us/ie/ff468705.aspx#_CSS3_Values_Units
20:38
<paul_irish>
ah. good eye.
20:39
<Peter->
Changelog of that document says it PP4 added support for WebIDL as well
20:51
<AryehGregor>
"Support for WebM software is not included in this release"
20:51
<AryehGregor>
I wonder if that means "support like we said before, where you have to manually install the codec" or proper support.
20:52
<AryehGregor>
If it's proper support, it will be time to break out the champagne.
20:52
<AryehGregor>
One would imagine that Google and Microsoft have been doing all sorts of secret negotiations about this . . .
21:12
<annevk5>
hmm, quite a bit of the XMLHttpRequest tests I created are now wrong per the specification...
21:22
<Philip`>
Is <html manifest="..."> not usable in XHTML?
21:22
<Philip`>
(Seems it's only triggered by the HTML parser)
21:22
<annevk5>
it is
21:22
<annevk5>
see the XML loading thingie
21:23
<annevk5>
http://www.whatwg.org/C#read-xml
21:23
<Hixie>
you don't need the www.
21:23
<Hixie>
:-)
21:23
<annevk5>
jaja :)
21:23
<TabAtkins>
What's a dependable way to tell when a <video> has frame data to be grabbed?
21:23
<annevk5>
bbiab
21:23
<TabAtkins>
(Waiting for "play" doesn't work.)
21:24
<Philip`>
Ah, I was just looking the in "XHTML Syntax" section
21:24
<Philip`>
s/the in/in the/
21:24
<Hixie>
TabAtkins: iirc there's an event for it
21:24
<TabAtkins>
"canshowcurrentframe"?
21:25
<Hixie>
sounds right
21:25
<TabAtkins>
kk, chrome just doesn't fire that event, it seems. Drat.
21:25
<Hixie>
file a bug
21:25
<Hixie>
:-)
21:27
<Philip`>
Should set up http://c.whatwg.org/#whatever so that autocompleting location bars will probably just need the "c.w"
21:30
<Hixie>
done
21:30
<Hixie>
http://c.whatwg.org/ redirects to http://www.whatwg.org/specs/web-apps/current-work/complete.html and http://c.whatwg.org/m redirects to http://www.whatwg.org/specs/web-apps/current-work/complete/
21:30
<Philip`>
error id: "bad_httpd_conf"
21:31
<Hixie>
yeah give it a few minutes
21:31
<Philip`>
Now it redirects to http://www.whatwg.org/specs/web-apps/current-work/complete.htmlm
21:32
<Philip`>
which is closer
21:32
<Hixie>
hm, i guess it would huh
21:32
<Hixie>
try now?
21:32
<Philip`>
That works better
21:32
<Hixie>
had them in the wrong order
21:34
<Philip`>
Hmm, now how do I teach Opera to autocomplete to the /m address instead of the / one...
21:36
<Hixie>
why can't i register xn--7ca.whatwg.org
21:36
<Hixie>
silly dreamhost restrictions
21:38
<gsnedders>
Do you not have a HAPPY FACE or whatever on hixie.ch?
21:38
<Hixie>
yes
21:39
<Hixie>
http://xn--74h.damowmow.com/
21:39
<gsnedders>
Is that not on dreamhost?
21:39
<AryehGregor>
Doesn't show up a happy face in Chrome. :(
21:41
<Hixie>
it's on dreamhost
21:42
<Hixie>
ok i just got a captcha where one of the characters was U+221E
21:42
<Hixie>
how are people supposed to type that in!
21:42
<TabAtkins>
lolwut?
21:42
<TabAtkins>
You sure it wasn't an 8?
21:44
<gsnedders>
alt + 5 on the default US/GB Macintosh keymap? :P
21:44
<AryehGregor>
Ctrl+Shift+u221E
21:44
<Hixie>
huh, i never knew about alt+5
21:44
AryehGregor
actually knows that one by heart, more or less
21:44
<gsnedders>
Hixie: :D
21:44
<gsnedders>
Hixie: It's useful!
21:45
<AryehGregor>
Surely that sort of thing conflicts like crazy with application hotkeys?
21:45
<Hixie>
TabAtkins: http://junkyard.damowmow.com/434
21:45
<Hixie>
i can't say i type infinities very often
21:45
<AryehGregor>
Wait, was that input actually accepted?
21:46
<AryehGregor>
I'd think it's an 8.
21:46
<AryehGregor>
Although, wait.
21:46
<Hixie>
yes, it was accepted
21:46
<Hixie>
but it's recaptcha
21:46
<Hixie>
so it probably wasn't checking that side
21:46
<TabAtkins>
Oh, huh. Indeed it is infinity.
21:46
<AryehGregor>
Recaptcha actually presents two things and only looks at one, yeah.
21:46
<AryehGregor>
I was about to say.
21:46
<Hixie>
hopefully some ocr will know about inifinity now!
21:46
<AryehGregor>
So maybe for the side it doesn't look at, it doesn't check well enough that it's really ASCII text.
21:47
<Philip`>
Hixie: Only if the other people who get shown that symbol know how to type it
21:47
<Philip`>
otherwise it has no reason to trust your input
21:48
<Hixie>
i'll just go over to the recaptcha team and let them know to make their system trust me... :-P
21:48
<Hixie>
anyway
21:48
<Hixie>
Philip`: i was trying to make http://xn--7ca.whatwg.org/ a shortcut for the multipage one for you, but i can't yet. i filed a support request for you.
21:48
<AryehGregor>
Heh, I forgot that Google bought reCAPTCHA too.
21:48
<AryehGregor>
Like every other cool thing.
21:50
<Philip`>
I can't type ç in Opera :-(
21:50
<gsnedders>
IE9PP4 is claiming 374 new JS tests…
21:50
<Philip`>
(Works in other applications with ctrl+= c)
21:51
<TabAtkins>
I didn't know we'd bought recaptcha!
21:51
<Philip`>
s/ctrl/altgr/
21:51
<Philip`>
¢.whatwg.org would be better for me :-)
21:52
<Hixie>
Philip`: k :-)
21:53
<Philip`>
Oh, that doesn't work
21:53
<Philip`>
"The URL http://¢.whatwg.org contains characters that are not valid in the location they are found."
21:54
<Hixie>
well dreamhost won't let me register that one either anyway
21:56
<AryehGregor>
What's the point of the NE keyword?
21:56
<AryehGregor>
To spam everyone with bugmail in case they care?
21:56
<Philip`>
Yes
21:57
<jgraham>
Isn't this whole non-ascii thing just likely to upset anti-homograph code
21:57
<Philip`>
Otherwise all the technical discussion gets stuck in Bugzilla and nobody even knows about it
21:57
<AryehGregor>
jgraham, aren't registrars required to vet IDNs for homographs before allowing registrations?
21:58
<Philip`>
jgraham: Why should anyone care about homographs here?
21:58
<Philip`>
given that you could use http://www.paypal.com.whatwg.org and don't need to bother with the homography
21:59
<Hixie>
indeed
21:59
<AryehGregor>
Who decided to make domain names in URLs backward?! :(
21:59
<Hixie>
tim
21:59
<jgraham>
Philip`: Possibly. I am assuming that whatever defences browsers have (if any) are rather dumb
21:59
<Hixie>
iirc
21:59
<annevk5>
AryehGregor, see http://krijnhoetmer.nl/irc-logs/whatwg/20100804#l-277 for NE
21:59
<jgraham>
Which seems likely :)
21:59
<AryehGregor>
I guess domain names started out backward, and file paths started out forwards, and TBL wanted to stitch them together without reversing either one.
22:01
<TabAtkins>
Yeah, though he's on record as regretting that decision.
22:02
<AryehGregor>
Yay, more polls.
22:02
AryehGregor
puts off voting for a few days
22:03
<Hixie>
this one doesn't seem so important
22:03
<Hixie>
it's just "should the w3c spec be complete or not"
22:03
<Hixie>
and we've already decided it shouldn't be :-)
22:03
<AryehGregor>
Heh.
22:03
<AryehGregor>
True.
22:06
<othermaciej>
I think it's useful for the w3c spec to contain features that vendors plan to implement
22:06
<othermaciej>
if only for IPR disarmament reasons
22:07
<othermaciej>
so far I can't tell if any non-WebKit-based browsers have an interest in srcdoc
22:07
<Hixie>
i agree that it would be useful
22:07
<annevk5>
i'm interested
22:07
<annevk5>
and our security folks are somewhat
22:08
<Hixie>
but given limited capacity to effect change, focusing only on normative issues is a higher priority for me
22:08
<annevk5>
it doesn't have much priority for us at the moment though
22:08
<Hixie>
and it seems i can leave it up to the chairs to make sure that the w3c html spec is relevant
22:09
<othermaciej>
you're free to have your own priorities; I'm mentioning reasons that others might find it useful to reply to the survey
22:10
<annevk5>
oh
22:10
<othermaciej>
however, it is also possible and entirely ok that people may find it a waste of time to reply
22:10
<annevk5>
well, it seems most is already said by the change proposal and the first response :)
22:11
<othermaciej>
if Opera is potentially interested in implementing, that is a valuable data point, whether in email or the survey
22:14
<annevk5>
ok
22:14
<annevk5>
replied
22:16
<annevk5>
i put it as an argument against dropping
22:41
<espadrine>
I get a parse error with html5lib, and don't understand it...
22:41
<espadrine>
line 43 col 7: end-tag-too-early-ignored
22:41
<espadrine>
It's on a <form>
22:41
<Hixie>
try validating with henri's validator, see what it says
22:44
<espadrine>
http://validator.w3.org/check sees nothing; where is henri's validator?
22:45
<boblet>
http://validator.nu/
22:46
<espadrine>
Neither does validator.nu
22:46
<Hixie>
validator.whatwg.org lists the current known html5 validators (that is, henri's validator)
22:46
<Hixie>
might be a bug in html5lib
22:46
<Hixie>
do you have a minimised test case showing the problem?
22:48
<espadrine>
It's a 5-line form, should I paste it here?
22:49
<boblet>
Hixie: what’s up with @profile, and the separate @profile spec?
22:50
<TabAtkins>
boblet: @profile doesn't solve real problems and isn't used by anything worthwhile on the real web, so it's dropped.
22:50
<TabAtkins>
The separate @profile spec is by someone who disagrees.
22:50
<boblet>
TabAtkins: aah thanks (well two ppl who disagree)
22:52
<boblet>
so it should theoretically be possible for the microformat use of @profile to be replaced by link huh
22:52
<Hixie>
espadrine: sure
22:53
<Hixie>
boblet: it's not clear that microformats really use profile="" in practice
22:53
<TabAtkins>
Microformats dont' seem to do anything useful with @profile in the first place, so it seems they can be replaced by nothing at all.
22:54
<espadrine>
<form method="post" enctype="application/x-www-form-urlencoded"
22:54
<espadrine>
action="commit.cgi">
22:54
<espadrine>
<div id="controls">
22:54
<espadrine>
<p id="selprod">0 &oelig;uvres s&eacute;lectionn&eacute;es.</p>
22:54
<espadrine>
<p><button id="but" disabled>R&eacute;server les &oelig;uvres
22:54
<espadrine>
s&eacute;lectionn&eacute;es</button></p>
22:54
<espadrine>
</div>
22:54
<espadrine>
<div id="liprod"></div>
22:54
<espadrine>
</form>
22:54
<Hixie>
espadrine: i see nothing wrong with that
22:54
<Hixie>
espadrine: might just be an html5lib bug
22:54
<boblet>
TabAtkins: I think there’s a theoretical “you should use @profile to disambiguate” idea, which may become the same as @itemtype
22:54
<TabAtkins>
(Except for gratuituous use of entities.)
22:55
<Hixie>
espadrine: jgraham might know
22:55
<TabAtkins>
boblet: Yeah, but theoretical disambiguation rarely seems to pan out to anything useful.
22:55
<Hixie>
boblet: disambiguate what from what? it's not like "hcard" is going to be used for another microformat
22:56
<karlcow>
question about srcdoc attribute: The document seems to be in a nested browsing context (according to the spec). If I write a text which will be put in srcdoc and I create anchors to the current page. All my anchors will be broken?
22:56
<TabAtkins>
karlcow: Yeah, it's equivalent to writing a separate page and linking to it with <iframe src>.
22:57
<karlcow>
hmmm then it might not be very practical :/
22:57
<TabAtkins>
?_?
22:57
<AryehGregor>
You could use seamless, right?
22:57
<AryehGregor>
Although that seems highly pointless.
22:57
<Hixie>
why pointless?
22:58
<Hixie>
srcdoc="" will usually be used with seamless="" sandbox="allow-same-origin"
22:58
<Hixie>
which makes the links work
22:59
<AryehGregor>
Oh, really?
22:59
<AryehGregor>
Hmm.
22:59
<boblet>
I guess the lack of native browser support is due to them being based on @class rather than the lack of an @itemtype equivalent then huh
22:59
<karlcow>
TabAtkins: I imagine the scenario on a blog, or forum, or documentation page with a form for commenting. Then you want to make a link to one of the previous comment or a specific section of the page. You put the anchor name. Save. And then the link is broken
22:59
<TabAtkins>
karlcow: You should be using @seamless and @sandbox="allow-same-origin" for the blog comments case, like Hixie says, so they look like they're part of the same page anyway.
23:00
<karlcow>
TabAtkins: ah ok. thanks.
23:00
<AryehGregor>
Heh, these IE blog posts are ripping the other browsers to shreds with their totally legitimate benchmarks.
23:00
<TabAtkins>
boblet: No, the lack of native browser support is generally due to no one caring enough.
23:00
<boblet>
TabAtkins: ouchies :)
23:01
<Hixie>
actually i guess you'd do <iframe seamless sandbox="allow-top-navigation allow-same-origin" srcdoc="..."> to have the links work if you wanted the sandbox
23:02
<boblet>
TabAtkins: that seems strange to me given e.g. Apple Mail’s data detectors for calendar events in natural language
23:02
<Hixie>
maybe the sandboxed navigation browsing context flag should somehow be mitigated by the seamless attribute
23:03
<Hixie>
or not, i guess it's better to leave that up to authors
23:04
<AryehGregor>
IE9 now even does better on SunSpider than Firefox 4.0b2 and Safari 5.
23:06
<karlcow>
is srcdoc='<p class="foo">bar</p>' possible or srcdoc="<p class='foo'>bar</p>"
23:06
<TabAtkins>
Yeah, they're both fine.
23:07
<karlcow>
oki
23:07
<espadrine>
by the way, is there any escape mechanism at all? I didn't realize we'd need that any time soon, but...
23:07
<karlcow>
the note in the spec might be a bit confusing then:
23:07
<karlcow>
>In the HTML syntax, authors need only remember to use U+0022 QUOTATION MARK characters (") to wrap the attribute contents and then to escape all U+0022 QUOTATION MARK (") and U+0026 AMPERSAND (&) characters, and to specify the sandbox attribute, to ensure safe embedding of content.
23:08
<Hixie>
how is that confusing?
23:08
<Hixie>
if you just do what it says you're safe :-)
23:08
<TabAtkins>
espadrine: What do you mean?
23:09
<espadrine>
TabAtkins: I have to read the spec better...
23:35
<Hixie>
interesting that the guy on hybi says you don't have to pay $500 to participate in IETF meetings
23:35
<Hixie>
i guess next time i attend i'll just decline to pay and see how that goes
23:36
<Hixie>
(not that i've ever attended one so far)
23:42
<othermaciej>
for me, the bigger barrier is flying to europe for a 3-hour meeting
23:43
<Hixie>
no kidding
23:43
<TabAtkins>
That feels like a plus to me. ^_^
23:43
<othermaciej>
I like traveling to Europe for vacation, but not for a very short time for nominally work
23:44
<Hixie>
traveling for work is so painful
23:44
<Hixie>
hate hate hate
23:44
<Hixie>
and the TSA is doing its damnest to make flying for vacation hellish too
23:45
<TabAtkins>
While I remember when flying was easier, I don't find current flying that bad. I wear sandals, so the stupid shoe thing isn't a problem.
23:45
<othermaciej>
for a single 3-hour session, it would probably consume 4 work days for what I'll optimistically call 3 hours of productivity
23:45
<TabAtkins>
The only issue I run into is getting my bag searched *every single time* because I carry around a chainmail dice bag.
23:48
<Hixie>
take laptop out. put sandals on belt. put coat on belt. put keys in bag. put credit card in bag. lose half-drunk bottle of juice. spend six minutes putting everything back together.
23:48
<Hixie>
good.
23:48
<Hixie>
i
23:48
<Hixie>
heavens.
23:48
<Hixie>
hate the tsa.
23:48
<Hixie>
all for what? it's just making a soft target and having zero impact on security on planes.
23:48
<Hixie>
security theatre annoys me no end.
23:49
<Hixie>
oh and nowdays it's also "stand in this machine that puffs air at you" and "stand in this machine that takes a nude picture of you"
23:50
<Hixie>
and we haven't even gotten to the airlines yet, which are trying their hardest to make air travel the least pleasant way to travel ever
23:50
<cardona507>
Most startups have the wrong idea of UX - http://bit.ly/bD9tFr
23:51
<cardona507>
meant for my team
23:55
<espadrine>
What purpose does the _additional allowed character_ serve, while parsing attributes?
23:56
<TabAtkins>
Presumably there's web content that depends on it being an allowed character.
23:57
<espadrine>
Legacy it is, then!