03:31
<zewt>
i'm declaring the most useless often-used feature of HTML: copying/pasting HTML
03:32
<zewt>
every single time I copy text from one webpage and paste it into something like gmail or evernote I have to paste it into vim and copy it to get rid fo styling, or it pastes in some gigantic font or something else equally useless
06:45
<benschwarz>
Hixie: !ping
07:18
<annevk>
the new memes cracked me up
07:22
<annevk>
http://golem.ph.utexas.edu/~distler/blog/archives/002545.html <video poster> implementations are kind of bad when it comes to small videos
08:00
<hsivonen>
annevk: I observe that Jacques Distler's blog is not using MathML-in-text/html yet :-/
08:01
<annevk>
what's the benefit for him?
08:02
<hsivonen>
no benefit, I suppose, at the point where all XML bugs have been fixed and everything works
08:03
<hsivonen>
well, at least we carefully engineered the parsing algorithm to cater to the use cases seen on his blog
08:04
<annevk>
:)
08:05
<annevk>
the main benefit of HTML I guess is that you can use math entities and they will always work
08:05
<annevk>
whereas in XML it depends on whether the DTD is loaded
08:06
<hsivonen>
also, with current implementations, the page might load a tiny bit faster as HTML thanks to speculative loading on the HTML code path
08:07
<annevk>
but I think he normalizes them away
08:09
<hsivonen>
http://98.245.80.27/tcpc/OSCON2011/gbu.html is interesting
08:11
<annevk>
pretty positive over Go
08:19
<hsivonen>
http://html5.org/tools/web-apps-tracker?from=7201&to=7202 isn't a particularly useful response to the email hober sent. Basically, it kicks the problem the validator developers' court
08:19
<MikeSmith>
hsivonen: indeed
08:20
<annevk>
hsivonen: only if you decide it's worth the additional UI
08:20
<MikeSmith>
hsivonen: have you seen http://validator.keegan.st/ yet?
08:23
<hsivonen>
MikeSmith: I haven't
08:23
hsivonen
looks
08:26
<hsivonen>
MikeSmith: supporting CORS without a specific use case seems to have paid off
08:27
<MikeSmith>
yeah
08:27
<MikeSmith>
think it would be worthwhile to upstream that filter feature?
08:31
<hsivonen>
potentially, yes. need to play with it some more to have a proper opinion
09:55
<annevk>
heh http://www.netmagazine.com/interviews/ian-hickson-html5-vs-html
09:55
<annevk>
they still use the same old picture
09:58
<annevk>
http://my.opera.com/hallvors/blog/2012/07/25/sometimes-it-helps-to-speak-norwegian-in-new-zealand sounds fucked up
09:59
<annevk>
an <input type=submit> always has a value?
10:25
<jgraham>
Sweet jesus someone has to find a better picture of Hixie
10:35
<tantek>
the netmagazine article was pretty decent
10:35
<tantek>
obligatory troll commentary aside
10:35
<tantek>
s/troll/trollish
10:36
<tantek>
jgraham - what's wrong with the picture of Hixie - is it too old?
10:37
<tantek>
annevk - that *is* an odd bug re: submit.
10:39
<Stevef_>
tantek: what is trollish about the comment?
10:40
<tantek>
Stevef_ not comment, just some trollish commentary in the article itself.
10:40
<tantek>
I didn't notice anything trollish about your comment.
10:40
<Stevef_>
tantek: OK just checking
10:43
<tantek>
Stevef_ if you're curious, the "quicksand" comparison was what I was referring to.
10:44
<Stevef_>
tantek: ah OK
10:44
<Stevef_>
tantek: missed mention of shelley previoulsy...
10:54
<odinho>
;-)
10:56
Ms2ger
thinks it better not to comment on Shelley
11:19
<hsivonen>
What's new in Opera Mobile for Android: 'Improved handling of meta-tag "format detection"'
11:19
<hsivonen>
what does that really mean?
11:28
<tantek>
hsivonen - plenty of "formats" have been proposed recently (the past year + ) that use meta-tags (e.g. Facebook OGP, Twitter Cards), perhaps they're referring to some of those.
11:32
<MikeSmith>
so never took time to count up myself how many specs for Web APIs are in development in various working groups, but Erik Wilde did: http://dret.typepad.com/dretblog/html5-api-overview.html
11:32
<MikeSmith>
so we have on the order of 50 different specs
11:43
<hsivonen>
It would be interesting to do a analysis 'Who ISSUEs HTML5?' and tabulate stats about whose grievances have been raised as ISSUEs, who writes CPs, whose CPs get accepted and whose accepted CPs also got accepted in WHATWG HTML and whose not
11:57
<odinho>
hsivonen: No idea.
11:58
<hsivonen>
odinho: ok. not the clearest release note line item
11:59
<odinho>
hsivonen: You could say that. :] I asked if anyone knew.
11:59
<odinho>
I could search around but I'm not that interested :P
12:15
<MikeSmith>
hmm I guess it would be useful to make an alphabetical list list of specs like http://dret.typepad.com/dretblog/html5-api-overview.html but also include the CSS3 specs and the non-W3C ones that Erik omits from that
12:15
<MikeSmith>
with those included it's more on the order of 60 specs
12:22
<hsivonen>
dealing with some of my least favorite parts of the platform: session history and window.open
12:31
<odinho>
hsivonen: Found this in the bugtracker (for that changelog entry): The other problem worth to mention is Google Play limiting number of characters in changelogs, so we always need to be as short as possible ;)
12:32
<odinho>
hsivonen: But it's not linking to the real bug, so I still don't know what it means, but now know why the log entry is so criptic :P
12:32
<hsivonen>
odinho: thanks
12:34
<odinho>
hsivonen: FOUND IT!
12:34
<odinho>
http://developer.apple.com/library/safari/#featuredarticles/iPhoneURLScheme_Reference/Articles/PhoneLinks.html#//apple_ref/doc/uid/TP40007893-SW1 ---> <meta name = "format-detection" content = "telephone=no">
12:34
<odinho>
this here thing
12:37
<hsivonen>
odinho: thanks. totally non-obvious from the relnotes
12:38
<odinho>
hsivonen: yes, agreed. But probably because Play store had a length limit on changelogs. Should've been a link to something bigger though.
12:51
<hsivonen>
hmm. I don't see an invitation to bring <picture> to the HTML WG on the Responsive Images GC's mailing list
12:51
<hsivonen>
I wonder who did the invitation and why
12:53
<hsivonen>
I suspect the Responsive Images folks are going to face more of http://w3cmemes.tumblr.com/post/23122022271
12:53
<hsivonen>
"And then we told them ... to submit a proposal to the HTML WG"
12:56
<odinho>
lol, the tantek one is pure gold :D
13:01
Ms2ger
looks around http://www.cs.helsinki.fi/u/pettay/moztests/xmlevents/
13:07
<hsivonen>
at times, fixing about:blank feels hopeless
13:32
<hsivonen>
I wish the testing staff funded by Adobe, Microsoft and Google writes some good session history tests
13:53
<hsivonen>
am I the only one who read Maciej's advice about how to proceed with the era attribute as "Don't bother, but if you want a formal rejection, do this."?
14:37
<AryehGregor>
http://www.w3.org/TR/SVG11/svgdom.html#__svg__SVG_MATRIX_NOT_INVERTABLE
14:37
<AryehGregor>
Really?
14:37
<AryehGregor>
INVERTABLE?
14:37
<AryehGregor>
That made it into a spec?
14:37
<AryehGregor>
"Referer" is one thing, I'd like to think we're more professional now . . .
14:40
<Ms2ger>
Well
14:40
<Ms2ger>
SVGDOM
14:42
<odinho>
Referer <3
15:03
<heycam>
AryehGregor, sorry about that, I added a note to SVG 1.1 Second Edition pointing out the unconventional spelling :)
15:03
<heycam>
AryehGregor, (we're moving to Web IDL exception types soon anyway)
15:03
<AryehGregor>
Yay!
15:03
<AryehGregor>
Wait, WebIDL and not DOM?
15:03
<heycam>
AryehGregor, oh using the DOM ones
15:03
<AryehGregor>
Ah.
15:03
<heycam>
I don't think we need any specific SVG ones
15:04
<Ms2ger>
Oh
15:04
<Ms2ger>
I thought you guys did that ages ago
15:08
<jarek>
SVG DOM is awful
15:10
<jarek>
anyone here thinks that SVGPathSegCurvetoQuadraticSmoothRel is a good name?
15:10
<jarek>
looks a bit over-descriptive :/
15:38
<odinho>
jarek: Looks like Objective-C/Apple API's.
15:55
<gsnedders>
jarek: It's an awesome name!
16:27
<Hixie>
MikeSmith: to be fair, seven of those are fragments of the HTML spec
16:28
<Hixie>
and several of them are "native api" specs that really don't belong on the web in the first place imho
16:29
<Hixie>
or should be done as intent-driven web apis not implemented by the browser
16:29
<Hixie>
and he lists xhr and xhr2 as separate specs
16:29
<Hixie>
and media capture twice
16:29
<Hixie>
would be nice to have platform.html5.org updated to include any that were missed though
16:44
<annevk>
Hixie: maybe we need meta⊙wo? http://www.codinghorror.com/blog/2012/02/listen-to-your-community-but-dont-let-them-tell-you-what-to-do.html
16:45
<annevk>
on the other hand, we've been mostly good for eight years
16:56
karlcow
weird
16:56
karlcow
wonders what is the correct term instead of INVERTABLE?
16:57
<annevk>
INVERTIBLE
16:57
<karlcow>
annevk: thanks
17:00
<jzaefferer>
The spec is pretty ambigous about popstate event being fired on pageload: http://www.whatwg.org/specs/web-apps/current-work/multipage/history.html - now we have Chrome doing that, Firefox not. I'd like to see that fixed before we get stuck with that mess forever. Looking for some guidance how.
17:01
<jzaefferer>
from the MDN article https://developer.mozilla.org/en/DOM/window.onpopstate - "Browsers tend to handle the popstate event differently on page load. Chrome and Safari always emit a popstate event on page load, but Firefox doesn't." - that sucks for users
17:03
<jzaefferer>
MikeSmith: as my goto-whatwg person, any comment on the above?
17:04
<Ms2ger>
jzaefferer, file a bug to get it defined
17:05
<jzaefferer>
Ms2ger: okay, which tracker would be the right one?
17:06
<Ms2ger>
https://www.w3.org/Bugs/Public/enter_bug.cgi?product=WHATWG&component=HTML
17:07
<jzaefferer>
okay, thanks
17:09
<Ms2ger>
Np
17:15
<tantek>
annevk - the meta topic is interesting from a community management perspective.
17:16
<tantek>
however, it tends to eventually be dominated by people that don't love a community (as Atwood suggests), but by people that love meta for meta's sake.
17:16
<tantek>
prime example: the amount of time spent discussing, arguing, and trying to meet the steps for "process" at W3C.
17:17
<tantek>
so no, I don't think we need meta⊙wo
17:18
<tantek>
I mean unless you want to create a support group ;) for people who love meta for meta's sake *and* somehow found themselves involved with whatwg. It could be an interesting honeypot, to keep meta talk out of other fora.
17:20
<hober>
it could be useful as a honeypot
17:24
karlcow
is not subscribing to meta ☺ ever
17:24
<karlcow>
already enough of nonsense everywhere
17:27
<annevk>
tantek: fair enough
17:28
<tantek>
karlcow - but you *love* discussing process :)
17:28
<Ms2ger>
tantek, well, the point is to keep them away from the productive people ;)
17:29
<annevk>
tantek: I wasn't quite sure what we'd use it for either, just thought I'd suggest it so it could be considered
17:30
<tantek>
I feel like #whatwg provides a sufficient meta-forum. Perhaps consider meta@ lazily, that is, avoid it until someone makes a strong case+demand for it, and then give it to them.
17:31
<karlcow>
tantek: not really ☺ I love understanding/observing communities. I do not think there is one solution for all. I have the gut feeling that all Process often fails at having an annotated version with links to historical references.
17:31
<tantek>
+1 on Process often fails at having an annotated version with links to historical references
17:33
<tantek>
karlcow - in addition, I think the bigger failure of most process documentation is that it fails to document what's specifically been avoided/rejected (and why).
17:41
karlcow
is still thinking about that last sentence from Tantek
17:41
<Hixie>
tantek++ re not neeting a meta mailing list
17:41
<Hixie>
and re #whatwg being it
17:41
<Hixie>
needing, even
17:42
<Hixie>
that typo bodes poorly for the day's activities
17:43
<karlcow>
hmmm lightly to heavily documented https://en.wikipedia.org/wiki/Stare_decisis#Civil_law_systems differences depending on countries
17:43
<Ms2ger>
Hixie, might give you a nice set of easy typo-bugs tomorrow :)
17:44
<Hixie>
hah
17:44
<Hixie>
got plenty of those already
17:48
<Hixie>
volkmar: yt?
17:52
<volkmar>
Hixie: yes
17:54
<Hixie>
volkmar: do you know much abotu setMessageHandler()? as in https://groups.google.com/forum/?fromgroups#!topic/mozilla.dev.webapi/o8bkwx0EtmM
18:05
<volkmar>
damn, I was going to send you that link :)
18:06
<volkmar>
what do you want to know?
18:48
<Hixie>
tantek: do you have an ETA for the autocomplete feedback you were going to send btw? (for planning purposes)
18:53
<Hixie>
for anyone who is interested, my current (incomplete and in parts wrong) strawman proposal for merging all the various content handlers, web intents, web activities, system messages, etc, proposals is: http://www.whatwg.org/specs/web-apps/current-work/scratch
19:31
<Wilto>
hsivonen: Ian Jacobs reached out to me on behalf of the HTML WG chairs, to the best of my knowledge.
19:32
<Wilto>
hsivonen: That is, simply to post the proposal to the list.
19:38
<annevk>
http://html5.org/r/7203 lol
19:38
<Hixie>
:-P
19:39
<annevk>
haven't seen requests for XMLHttpRequest yet
19:39
<Hixie>
weird
19:42
<hober>
Hixie: what's the use case for having "new" AND "reuse"?
19:42
<Hixie>
most intents won't be able to handle the same page being used for two intents
19:42
<Hixie>
think, e-mail compose window intent handler
19:43
<Hixie>
so for those, you'd want "new"
19:43
<Ms2ger>
annevk, count this as a request :)
19:43
<Hixie>
"reuse" is for the cases where the app can in fact handle multiple things at once
19:43
<Hixie>
hober: does that make sense?
19:43
<annevk>
Ms2ger: you don't count :p
19:44
<Hixie>
hober: (btw, your input on this in general is very welcome, i'm very interested in safari and opera input on this since it has mostly been a mozilla and chrome thing so far)
19:44
<hober>
I suppose. Being stupid or lossy under the "reuse" behavior seems like a bad way to code an intent handler.
19:44
<Hixie>
(also interested in IE feedback, should any Microsoft people be lurking here or reading the logs)
19:44
<Hixie>
hober: right, hence why you'd say you're a "new" handler, then you don't have to worry about it
19:44
<Hixie>
hober: you wouldn't want all your e-mail compose intents to end up in the same tab :-)
19:45
<Hixie>
or all your "open PDF in new window" intents :-)
19:45
<hober>
why would "open pdf in new window" have anything to do with this mechanism? :)
19:46
<Hixie>
PDF viewing sites might well call registerContentHandler() for application/pdf
19:46
<hober>
"PDF viewing sites" *shudder*
19:46
<Hixie>
ok, 3D Model Viewing Sites
19:46
<Hixie>
whatever :-)
19:47
<Hixie>
think of Pages as a web page
19:47
<Hixie>
it handles the "open Pages app" intent
19:47
<Hixie>
er
19:47
<Hixie>
"open Pages doc" intent
19:47
<Hixie>
but each Pages doc ends up in a new Window
19:47
<Hixie>
so it's like "new"
19:48
<Hixie>
but then consider the "open MP3 doc" intent that iTunes supports
19:48
<Hixie>
it handles all of them in the same window
19:48
<Hixie>
so it's like "reuse"
19:52
<hober>
*nod*
19:54
<annevk>
oh hober
19:54
<annevk>
poor you
19:54
<hober>
On the plus side, I don't have to think much about what my Halloween costume will be at TPAC this year
19:54
<annevk>
I mean, have the funnest time
19:55
<annevk>
haha
19:55
<Hixie>
because you'll have given up and become an alpaca farmer by then? :-)
19:55
<hober>
(Zoidberg, obviously)
19:56
<hober>
hahahaha
19:57
<Hixie>
do you know when the four of you will begin editing?
19:57
<annevk>
I considered alpaca farming, but then the tent froze over during the night and I decided not to
19:57
<Hixie>
hober: (i'm mainly curious about when i should bother approaching y'all to see how you want to handle getting new bugs to both teams)
19:58
<Ms2ger>
I guess I could see that coming
20:02
<hober>
Hixie: the initial meeting hasn't even happened, so "not this week" is probably a safe answer.
20:03
<Hixie>
you're going to have meetings?
20:03
<Hixie>
oh poor poor hober
20:03
<Hixie>
lunch, bbiab
21:17
<tantek>
Hixie - re: autocomplete feedback completion ETA, good estimate would be next Monday, though I plan to work on it a bunch today.
21:17
<tantek>
I'll be basing my feedback on editing in real-time the hcard-input page in case anyone is interested in stalking/following along (edit notifications also go to #microformats) http://microformats.org/wiki/hcard-input#next_actions
21:47
<Hixie>
Any quirks mode spec people around?
23:55
<tantek>
Hixie, I've purged quirks mode details from my memory since leaving Microsoft. Hoping that many (most?) of them no longer matter.