05:45
<Hixie>
back, fwiw
05:45
<Hixie>
abarth: why so?
05:59
<abarth>
Hixie: we're just reading about seamless and srcdoc
05:59
<abarth>
Hixie: reading how the two interact
10:31
<annevk>
https://dvcs.w3.org/hg/domcrypt/shortlog is somewhat confusing
10:49
<AryehGregor>
So, exciting tip: if you're taking your computer from America, and it has a power supply that supports both 110V and 220V, make sure that before you hook it up to 220V, you check that you don't have to flip a switch or anything.
10:50
<AryehGregor>
Otherwise you might need to make an unexpected trip to the computer store.
10:50
<AryehGregor>
(Apparently, the Hebrew term for a power supply is ספק כח.)
10:50
<j_wright>
thanks edison
12:43
<annevk>
https://bugzilla.mozilla.org/show_bug.cgi?id=641821#c59
12:44
<annevk>
didn't Gecko argue against that initially?
12:44
<annevk>
oh well
12:44
<annevk>
good to see people read the spec carefully when implementing...
12:45
<Ms2ger>
Gecko is a rendering engine, it can't argue against anything :)
12:50
<annevk>
you know what I mean
12:51
<Ms2ger>
You know Mozillians' comments are almost always in a personal capacity :)
12:52
<annevk>
it was either sicking or smaug so it doesn't matter much in this case
12:57
<Ms2ger>
Speak of the devil
12:58
smaug____
is not a devil, just a friendly dragon
12:59
<smaug____>
annevk: argue agains what?
12:59
<smaug____>
against
13:00
smaug____
needs to still file bugs found in the spec during the week
13:00
<annevk>
having two records for a replace operation
13:01
<smaug____>
annevk: in some cases there sure are many records
13:01
<smaug____>
there is the record removing the new node from its old parent
13:02
<smaug____>
then record adding to the new parent
13:02
<smaug____>
the latter can be combined with removing the old node
13:02
<smaug____>
I need to still check what latest webkit does
13:03
smaug____
doesn't know how to update chrome
13:10
<annevk>
I was referring to the comment in that bug...
13:10
<annevk>
it was not about adopting
13:10
<annevk>
you'll always need a separate record for that
14:48
<AryehGregor>
When is .isTrusted supposed to be true?
14:49
AryehGregor
wonders if input events from execCommand() should have it true or false
14:49
<annevk>
false
14:49
<AryehGregor>
Okay, why?
14:50
<AryehGregor>
Should it be true if the execCommand() is triggered by a user action (e.g., hitting a key triggers document.execCommand("insertText"))?
14:52
<annevk>
hmm, I guess input is some kind of corner case
14:52
<annevk>
isTrusted is true when the spec says to dispatch the event
14:52
<annevk>
e.g. readystatechange on XMLHttpRequest will have it set to true
14:53
<AryehGregor>
The spec for execCommand() says to dispatch the command.
14:53
<AryehGregor>
The author can invoke execCommand(), of course.
14:53
<zewt>
annevk: seems like it should be true in this case, if the event object is constructed by execCommand and can't be arbitrarily manipulated
14:53
<zewt>
that is, it's always seemed to me that !isTrusted means "this object was constructed by hand"
14:54
<zewt>
or something like that
14:54
<annevk>
yeah, something like that
14:54
<annevk>
:)
14:54
<annevk>
not sure it's actually needed
14:55
<zewt>
hmm
14:55
<zewt>
form.click() results in !isTrusted
14:55
<zewt>
(in FF)
14:55
<zewt>
which seems odd
14:56
<zewt>
at least to how i've intuitively viewed isTrusted
14:56
<zewt>
(personally I've never found isTrusted to be terribly useful, anyway...)
14:57
<annevk>
so in Gecko it might be user-initiated; something to do with their XBL impl
14:57
<zewt>
(when are you ever not required to trust other scripts on the page, anyway?)
14:58
<annevk>
XBL
14:58
<zewt>
xbl isn't relevant to web pages...
14:58
<annevk>
I'm not saying having isTrusted makes sense
14:59
<zewt>
weird, didn't know webkit (or at least chrome) doesn't have form.click (or maybe I did and I forgot, since I've been over this territory before)
14:59
<annevk>
maybe file a bug on DOM 3 Events?
14:59
<annevk>
I just specced isTrusted because it was there...
14:59
<zewt>
i'm sure it won't be removed so I won't waste my time
14:59
<annevk>
the new editor is far more reasonable
14:59
<zewt>
not important enough (to me) to spend time arguing for that anyway (if anyone else wants to, go for it)
14:59
<annevk>
and we could at least learn what it's for
14:59
<Ms2ger>
3249 // Click() is never called from native code, but it may be
14:59
<Ms2ger>
3250 // called from chrome JS. Mark this event trusted if Click()
14:59
<Ms2ger>
3251 // is called from chrome code.
14:59
<annevk>
yeah I might
15:01
<zewt>
if it doesn't do anything meaningful, it could also just always be true, which would be less likely to break pages (still not guaranteed, though) and simplify things
15:01
<annevk>
hmm
15:01
<annevk>
is it even supported in every browser?
15:02
<zewt>
someone might still be using it for other unintended purposes, like "did I initiate the event or did the user"
15:02
<zewt>
(which you can always do by tacking a property on the event when you create it, of course)
15:03
<zewt>
seems to not be there in chrome
15:04
<zewt>
don't have ie9 handy
15:06
<Ms2ger>
If it's in D3E, IE probably has it
17:20
<Ms2ger>
TabAtkins, nicely done
17:22
<annevk>
that does look pretty
17:29
<kennyluck>
oh my goodness.
17:59
<gsnedders>
"Design"
17:59
<gsnedders>
"W3C"
17:59
<gsnedders>
I think my brain just imploded.
17:59
<gsnedders>
Just wow. It's really quite well done.
23:13
<heycam>
Hixie, on HTMLCanvasElement the "_callback" argument can now be named just "callback" (argument identifiers don't need to be escaped when they're any of the names at http://dev.w3.org/2006/webapi/WebIDL/#prod-ArgumentNameKeyword)