09:18
<JPeterMugaas>
I'm curious, has someone added "msapplication-window" to the meta tag registration? It has to do with pinned websites in Internet Explorer 9.0.
09:21
<Philip`>
JPeterMugaas: It's not listed on http://wiki.whatwg.org/wiki/MetaExtensions so it seems nobody has added it yet
09:23
<JPeterMugaas>
I think someone needs to do this but I don't have a wiki account. It is described at http://msdn.microsoft.com/en-us/library/gg491732(v=vs.85).aspx#msapplication-window .
09:39
<zcorpan>
JPeterMugaas: if you leave your desired user name and email, maybe someone here will create an account for you
09:47
<JPeterMugaas>
Sure. My desired username is "JPeterMugaas" and my E-Mail is jpmugaas⊙sn .
10:38
<annevk>
JPeterMugaas: I'll create one for you
10:38
<JPeterMugaas>
Thanks.
10:39
<annevk>
should have an email
10:41
<JPeterMugaas>
I saw it.
10:42
<JPeterMugaas>
What password do I use to login?
10:43
<JPeterMugaas>
Nevermind.
11:25
<JPeterMugaas>
Okay. I've added that tag to http://wiki.whatwg.org/wiki/MetaExtensions
14:24
AryehGregor
tries to remember what he thought he should do today
14:30
AryehGregor
writes some tests for comparePoint, if no one else has any suggestions
14:33
<annevk>
couple of range bits still have XXX comments
14:37
AryehGregor
will look if he has time
14:46
<AryehGregor>
The only interesting one I see is in intersectsNode(), which I suppose I'll get to after comparePoint().
14:46
<AryehGregor>
I think it's the last Range thing that's untested.
14:49
<annevk>
the recursion stuff will prolly come up with mutations
14:49
<annevk>
and there's the bug on removing detach()
14:49
<AryehGregor>
Yeah, the recursion stuff needs to be looked at at some point, but it's not terribly important, more of a technical fix.
14:50
<AryehGregor>
To be honest, if we look at the exact mutations that each method causes instead of just the final DOM, we're going to find about twenty billion inconsistencies between browsers.
14:50
<AryehGregor>
Range.extractContents() will be the least of our worries.
14:50
<AryehGregor>
So I'm not terribly concerned about that for now.
14:50
<AryehGregor>
The removing detach() stuff I don't think we should do anything about until we have a browser that's actually willing to remove it.
14:50
<AryehGregor>
And tries.
14:51
<annevk>
I don't think it should be too bad, but we'll see
14:51
<AryehGregor>
Add an XXX to the spec, I'd say.
14:52
<AryehGregor>
For simple stuff like Node methods, no, but Range methods like extractContents do a whole bunch of mutations, and I wouldn't expect them to match exactly between browsers.
14:53
<AryehGregor>
execCommand() will be a nightmare with mutation observers, because I didn't actually pay the slightest bit of attention to the precise order of insertion and removal, and probably it's completely irrational in most cases. But interop for execCommand is completely nonexistent for even the most basic use-cases, so I'm really not worried about that.
14:57
<annevk>
is execCommand() mostly defined in terms of range mutations?
14:57
<annevk>
range mutations should cause several MutationRecords basically, but the order indeed, dunno
14:57
<smaug____>
AryehGregor: why does the order matter. Things should be collapsed to one mutationrecord when execCommand is used
14:57
<smaug____>
at least I hope so
14:57
<smaug____>
though, it is not possible in all the cases
14:57
<smaug____>
hmm
14:57
<AryehGregor>
smaug____, I dunno.
14:57
<annevk>
MutationRecord cannot do complex stuff
14:57
<AryehGregor>
But if you follow my algorithm step-by-step, one execCommand() could easily result in hundreds of different mutations.
14:57
<annevk>
you'll have several for range mutations
14:57
smaug____
doesn't know what is range mutation
14:57
<AryehGregor>
Order has to matter if you're doing any nontrivial sequence of related mutations.
14:58
<AryehGregor>
Or so I'd imagine.
14:58
<zewt->
is anyone willing to try to move towards interop on execCommand, or is every site using it with browser sniffing and making that impossible?
14:58
<AryehGregor>
zewt-, a mix. Gecko and WebKit are interested in more interop, but it will take vast amounts of work and has to be done very carefully to avoid breaking sites.
14:58
<smaug____>
AryehGregor: yeah. it doesn't matter if mutations are happening to same element, but with execCommand that is not the case
14:58
<zewt->
(feature sniffing instead of browser sniffing would make fixing execCommand more possible, but people unfortunately probably tend towards browser sniffing instead)
14:59
<zewt->
which might honestly be the only sane thing to do if the differences are too wide and strange to try to work around them as if they're feature differences
15:00
<AryehGregor>
zewt-, have you actually tried using execCommand()? Different browsers' implementations are practically unrelated a lot of the time. There are like a zillion differences in how every command is implemented. Sniffing each one separately would be pointless.
15:00
<AryehGregor>
Especially since if browsers change, odds are good that any particular thing will change to something incompatible with how all browsers did it.
15:00
<AryehGregor>
If you're talking about stuff like details of how tag nesting winds up.
15:01
<AryehGregor>
Oh, you already said something like that.
15:01
<AryehGregor>
Yeah, basically it's a total trainwreck and in real life anyone sane just doesn't use it.
15:01
<AryehGregor>
It's masochistic enough to try working around differences in contenteditable, let alone execCommand().
15:02
<bga>
hm
15:04
<bga>
is Doc#elementFromPoint one native fn to access to UA's dom quadtree?
15:05
<AryehGregor>
"quadtree"?
15:05
<bga>
i want get list of dom elements which are inside rect and inside + intersect rect
15:05
<bga>
AryehGregor http://en.wikipedia.org/wiki/Quadtree
15:07
<smaug____>
Gecko has nodesFromRect, but that is currently only for privileged scripts
15:08
<annevk>
hmm lists.whatwg.org times out?
15:09
<zewt->
works here
15:09
<zewt->
archives anyway
15:13
<AryehGregor>
Ms2ger, any specific reason you specced Range.intersectsNode(), as opposed to other WebKit extensions?
15:13
<Ms2ger>
Yes
15:13
<Ms2ger>
I've got a bug assigned to me to implement it in Gecko
15:14
<AryehGregor>
Ah.
15:14
<Ms2ger>
(That's why I started on DOM Range, actually)
15:14
<AryehGregor>
Okay, good enough for me.
15:14
AryehGregor
writes some tests
15:14
<bga>
smaug____ i will be good to have standard public method
15:14
<Ms2ger>
I wrote a couple, IIRC
15:15
<AryehGregor>
Only assert_throws() tests, though.
15:15
<smaug____>
bga: File a CSSOM spec bug
15:15
<Ms2ger>
Sounds right
15:39
<AryehGregor>
Weird -- WebKit and Opera don't throw if you call intersectsNode() on a detached range.
15:40
<AryehGregor>
. . . except sometimes.
15:40
AryehGregor
specs it to throw anyway
15:46
<gsnedders>
iframe.contentWindow.history.back(); takes the parent window back? gah.
15:47
<gsnedders>
Unless the history isn't what I think it is...
15:56
<AryehGregor>
Okay, I think every single Range attribute and operation that's specced has tests now.
16:02
<AryehGregor>
Feedback on the XXX here would be appreciated: http://dvcs.w3.org/hg/domcore/raw-file/tip/dom-core.html#dom-range-intersectsnode
21:02
<annevk>
seems my email made it to the list finally
21:06
gsnedders
tries to improve the HTML5 Wikipedia page
21:07
<gsnedders>
The history section entirely omits the lead up to the formation of WHATWG.
21:10
<gsnedders>
http://en.wikipedia.org/w/index.php?title=HTML5&action=historysubmit&diff=468660119&oldid=468655632
21:13
gsnedders
wonders how long until someone reverts that
21:22
<annevk>
http://en.wikipedia.org/wiki/HTML is also quite out of date...
21:22
<annevk>
"The last HTML specification published by the W3C is the HTML 4.01 Recommendation, published in late 1999. Its issues and errors were last acknowledged by errata published in 2001."
21:23
<annevk>
oh well, only so many problems I can solve
23:53
<annevk>
about six months for http://annevankesteren.nl/2011/06/creating-events
23:53
<annevk>
not bad