01:34
<til>
i have a page that i wrote in HTML5 forms, then wrote a javascript implementation of the validity API
01:34
<til>
the latest builds of WebKit are breaking my page, and i'm trying to figure out if they or i implemented wrong
01:35
<til>
when i hit submit, the page jumps to the first empty form element
01:35
<til>
before calling onsubmit
01:35
<til>
any hints as to what i should look into?
01:35
<til>
i haven't played with HTML5 forms in a long whule
01:37
<TabAtkins_>
Forms *should* validate before submitting, and not submit if anything's invalid.
01:40
<miketaylr>
til: do you have the required attribute set on those empty inputs?
01:41
<til>
miketaylr: nope
01:41
<til>
i've found a couple bugs in earlier chrome builds. one was a bug in their implementation of checkValidity
01:41
<til>
the other was a disagreement between Hixie and myself on what it means for a checkbox group to be @required
01:42
<til>
but this has otherwise worked great until now
01:44
<til>
TabAtkins_: i don't disagree, but i have to figure out if the rejected form should be considered invalid or not
01:45
<til>
right now, i'm using form.@onsubmit="return checkSurveyValidity();" to validate the form and block it from submitting on pre-HTML5 browsers
01:46
<til>
i may need to use a different event handler. webkit isn't running my validator at all (which means my smooth scrolling and message-displaying script isn't running)
02:29
<dbaron>
AryehGregor, I think Asa Dotzler has a bunch of statistics on how much of various types of Mozilla contributions are employees vs. volunteers
05:52
dbaron
reads the W3C process document and learns that PER status is basically PR + LC at the same time
05:53
dbaron
didn't realize it had so many aspects of Last Call
09:20
GPHemsley
looks up, then goes to bed
09:52
<gsnedders>
jgraham: Nor did I. Until recetnly, you only had to wait two hours, worse case, AFAIK
12:07
<annevk>
var x = new XMLHttpRequest;
12:07
<annevk>
you learn something new every day
12:09
<aho>
i wouldn't omit those invocation parens though
12:09
<aho>
it's a function call... don't make it look like something else
12:30
<nielsle>
Is it possible to get two separate contexts ctx1 and ctx2 that refer to the same canvas? I would like to use ctx1.lineTo(x,y) and ctx2.lineTo(x,y) to draw two curves in parallel.
12:33
<nielsle>
I can probably hack a solution together myself, but I would like html5 to solve the problem for me :)
12:35
<aho>
what do you mean with "draw in parallel"?
12:36
<aho>
you draw stuff to the canvas... and once you're done with that and the UI is back in control... the updated drawing is shown
12:36
<aho>
drawing onto one context or thousands doesn't change anything
12:37
aho
pokes nielsle
12:50
<nielsle>
I am toying with a canvas based multi user paint program. Whenever the user move the mouse I use lineTo(x,y) to draw a line. If another user draws something then the client recieves a json object and I also use lineTo to draw that. This works nicely unless the client revieves a json object while the user is in the process of drawing a stroke. I can probably solve this by creating some locks.
12:51
<annevk>
Reading http://www.w3.org/mid/4C1359A1.7050403⊙in I'm not really sure what to say... I did not quite agree with the removal of the Atom mapping, but going as far as objecting goes way too far.
12:52
<annevk>
It seems e.g. Julian is far less constrained in that regard going as far as raising an issue over something as simple as a reference.
12:52
<annevk>
Does that mean the progress is tilted in the advantage of people that like to bicker over every detail?
12:54
<aho>
nielsle, well... you just have to update your canvas if an event occurs (input, network, whatever)
12:54
<aho>
that's all there is to it
12:54
<aho>
(it's single threaded - there are no locks)
12:55
<annevk>
he means a lock to indicate it cannot be updated right this instant because something else is going on
12:55
<annevk>
nielsle, we might get dedicated path objects in the future, but not for now
12:57
<annevk>
pretty good read: http://bitworking.org/news/2010/06/distributed-extensibility
12:58
<nielsle>
annevk: Yes you are right. I was asking for a path object. (Sorry if I was unclear).
13:00
<annevk>
I believe the thinking is that it should have been that way from the beginning, but in the beginning everything went fast, and so we're stuck with some not too great design decisions
13:00
<annevk>
but we can also still paper over them in the future, and had we not it would probably have taken all much longer to get universally deployed
13:04
<nielsle>
Ah I see. Thank you for the answers.
16:07
<annevk>
what is http://wiki.whatwg.org/index.php?title=Talk:Timed_tracks&curid=1928&diff=4923&oldid=4716 about?
16:08
<annevk>
ah
16:08
<annevk>
it seems to be about http://wiki.whatwg.org/index.php?title=Timed_tracks&action=historysubmit&diff=4647&oldid=4645
23:49
<boblet>
Normally <li>’s content is just phrasing content. if I have an <aside> or other flow element inside an <li>, should I wrap the phrasing content in <p>?
23:50
<boblet>
eg <li>phrasing content</li> → <li><aside>…</aside>phrasing content</li> or <li><aside>…</aside><p>phrasing content</p></li>?
23:55
<daedb>
Works either way, imo
23:59
<boblet>
thanks daedb. yeah I know it works, but I wondered if it was more ‘correct’ to explicitly add the <p>
23:59
<daedb>
nah, I doubt it makes any difference
23:59
<gsnedders>
I like fixing bugs and then finding out other code relies upon those bugs.