06:35
<hsivonen>
“The polyglot markup community” http://lists.w3.org/Archives/Public/public-html/2013Jan/0052.html
06:42
<MikeSmith>
hsivonen: he forgot to add the word "thriving" there
06:42
<hsivonen>
I’m trying to figure out how to ask for the membership list of said community without Sam telling me I’m using emotive language or something.
06:43
<hsivonen>
I genuinely curious if the community is larger than Sam and Leif
06:47
<MikeSmith>
I wonder whether Leif actually even authors anything using polyglot markup
06:49
<MikeSmith>
heycam|away: in http://mcc.id.au/BrowserMiniconf/LCA2013/Schedule "Below is the schedule for the Browser Miniconf", the URL for the Browser Miniconf link goes to the 2012 info page
06:50
<MikeSmith>
instead of to the 2013 one
06:51
<MikeSmith>
shanestephens: ↑
06:57
<hsivonen>
I wonder if the Opera Ice video is an elaborate troll: the UI depicted has Windows 8 -like gestures and animations and there already was an ICE Browser from Norway.
07:03
<MikeSmith>
does make you wonder who leaked it and why
07:04
<MikeSmith>
and anyway the writeups about make speculations that aren't substantiated by anything that's actually in the video
07:04
<hsivonen>
yeah
07:05
<MikeSmith>
speaking of elaborate trolls I wonder sometimes if html5test is in the category
07:06
<hsivonen>
the remarks at the end of the video sound like they could be offensive to people who’ve worked on Opera’s products for years. I find it surprising that the presenter would say that to an audience of Opera employees.
07:08
<hsivonen>
I always get annoyed when journalists score browsers mainly by publishing html5test.com scores
07:09
<MikeSmith>
there's not shortage recently of some Opera people saying things that are offensive to people who've worked on Opera's products for years
07:10
<hsivonen>
s/journalists score/journalists rate/
07:10
<MikeSmith>
hsivonen: I was just looking at the dependency tree for the htmltest on mozilla bugzilla
07:11
<MikeSmith>
and I noticed the bug for keygen
07:12
<MikeSmith>
god knows why the htmltest dude thinks keygen merits having a feature-test there
07:39
<Stevef>
MikeSmith: congrats on finally getting the main layout tests working on webkit :-)
07:43
<MikeSmith>
Stevef: yeah the WebKit ax test-runner infrastructure is .. idiosyncratic
07:43
<MikeSmith>
but it gets the job done
07:44
<Stevef>
right
07:44
<MikeSmith>
now that I've learned it I suppose I should put it to good use and write a some more tests
07:45
<MikeSmith>
btw I didn't include a test to check if header and footer get de-landmarked if they're descendants of main
07:47
<MikeSmith>
because, first off, the patch doesn't yet actually implement that behavior
07:47
<MikeSmith>
hmm or perhaps it does
07:47
<Stevef>
OK , thats gets handled by article/secton currently, need to look into it and spec it out
07:48
<MikeSmith>
ok so a follow-up bug could be opened for that later
07:49
<MikeSmith>
I guess I should quit speculating and just write the test now to see
07:49
<Stevef>
yes, i have some related feedback to process
07:49
<MikeSmith>
but not make the test part of the initial patch
07:57
<Ms2ger>
MikeSmith, not just their a11y tests :)
08:01
<MikeSmith>
Ms2ger: yeah but I suspect that a11y test stuff for other engines maybe have some similar esoteric-ness
08:02
Ms2ger
grumbles at ReSpec
08:03
hsivonen
keeps reading a11y as ally
08:03
<hsivonen>
testing their allies
08:05
<MikeSmith>
ax is easier to type
08:06
<MikeSmith>
and read
08:08
<hsivonen>
would you rather accessibility be an ax or an ally?
08:08
<MikeSmith>
I prefer the ax
08:09
<MikeSmith>
btw am I missing something or is there really no way in bugs.webkit.org to do component watching?
08:11
<MikeSmith>
I know the bugzilla version there doesn't have the component-watching feature but it also doesn't have the QA contact field so no use of the convention of putting a unique address there to follow
08:22
<Stevef>
MikeSmith: the effect of section/article on header/footer and potential effect of main on same needs to be added to WAI-ARIA section of HTML in the mapping tables will file a bug etc
08:22
<MikeSmith>
Stevef: ok
08:24
<Stevef>
there is quite a bit of work on the WAI-ARIA section needed in 5.1 to better reflect implementations
08:25
<hsivonen>
Hixie, dglazkov: have you pinged the SVG and MathML groups about element names with a hyphen in them being reserved from now on?
08:27
<MikeSmith>
Stevef: I'm a bit confused about why the job of speccing some of this behavior falls to HTML instead of to PFWG to define in the ARIA specs
08:27
<Stevef>
MikeSmith: because the host language defines the mappings for the host language features
08:28
<MikeSmith>
Stevef: I realize they can't define anything about HTML elements, but they can for the related abstractions
08:28
<MikeSmith>
this is more than just mappings
08:28
<MikeSmith>
and it should be handled the same way across all host languages
08:30
<Stevef>
well that maybe something that can be refined in aria 1.1
08:31
<MikeSmith>
Stevef: yeah that seems to be the proposed solution for any suggestions/comments that people make about aria
08:31
<MikeSmith>
"yeah we gonna fix that in the next version"
08:31
<hsivonen>
“Google Chrome does not use the Windows Media Player plugin, it uses the ActiveX control instead.” Really?
08:32
<hsivonen>
is there even WMP-dependent Web content that anyone cares about anymore?
08:33
<Stevef>
I don't have much influence or input on the wai-aria spec and only have so many hands
08:56
hsivonen
wonders if there are polyglot elephants, too
09:17
<Stevef>
MikeSmith: and in the case of header/footer and banner/contentinfo we have an imperfetc match between the semantics and authoring requirements/advice, in which case the differences i think need to be defined in HTML
09:27
<MikeSmith>
Stevef: ok
10:06
<a_not_registed_n>
hi@all
10:11
<a_not_registed_n>
i´m trying to get an animation to work. (with <canvas>) is it possible to remove a single Item? for example the last one?
10:12
<a_not_registed_n>
asking because "context.clearRect" delete all to the ground with background and other layers
10:13
<Ms2ger>
Canvas is a bitmap, so, no
10:21
<a_not_registed_n>
so i had to draw in every frame everything again?
10:22
<tomasf>
if you want "layers", maybe you should use something other than a canvas
10:24
<a_not_registed_n>
layers would be nice. can you suggest something?
10:25
<tomasf>
what are you trying to accomplish?
10:27
<a_not_registed_n>
an application using Phonegap. so i can only use HTML5/CSS/JS
10:28
<a_not_registed_n>
and had to run on webkit
10:36
<a_not_registed_n>
can you store different transparent <canvas> on each other? selct by id and draw in wihout "damage" the other layers...
10:44
<hsivonen>
whoa. annevk is not on IRC.
11:01
<darobin>
lol, hsivonen :) “[t]he polyglot markup community” consists of you and Sam?
12:11
<annevk>
HTMLSelectElement.remove()
12:12
<Ms2ger>
Fun, isn't it?
12:13
<hsivonen>
annevk: you may be interested in the Character Encoding menu on Firefox for Android
12:13
<hsivonen>
annevk: it has fewer options than the desktop menu
12:13
<hsivonen>
annevk: and the list is based on looking at what kind of list the stock browser offers
12:14
<Ms2ger>
And no idea about window.event, but that feels uglier
12:14
<hsivonen>
minus removing GBK for some problem or another
12:14
<hsivonen>
(I wish we were more scientific about these menus)
12:15
<hsivonen>
(more scientific than treating the Android stock browser as a role model)
12:16
<hsivonen>
btw, I don’t see a Character Encoding menu at all in Chrome Beta for Android (en-US)
12:26
<annevk>
hsivonen: cool
12:27
<annevk>
Ms2ger: I guess Opera has srcElement too?
12:27
<annevk>
hsivonen: problem is I don't have Android at the moment, and I believe the Android phone I have somewhere hidden away does not run Gecko :/
12:27
<Ms2ger>
That's what was claimed in the bug, I haven't tested
12:28
<annevk>
I think Opera has window.event too though.
12:28
<annevk>
Does WebKit have that?
12:29
<Ms2ger>
Dunno
12:32
<hsivonen>
annevk: Firefox for Android now runs on ARMv6
12:32
<hsivonen>
and with less RAM
12:32
<hsivonen>
GLES is still a requirement, though
12:34
<hsivonen>
from the “Don’t they do QA or dogfooding?” department: Eclipse CDT regressed the ability to attach to a process with gdb and shipped
12:34
<annevk>
Does that mean compatible with http://en.wikipedia.org/wiki/HTC_Wildfire ? There's no easy to go by chart
12:34
<annevk>
Guess I can take a look tonight
12:36
<hsivonen>
annevk: it has too little RAM to be officially supported
12:36
<hsivonen>
might work anyway
12:47
<annevk>
Ms2ger: w("event" in window) -> true in Chrome
13:13
<annevk_>
Oh my bad. I missed that hsivonen made the same argument as I did in the other email.
13:14
<annevk>
"When 50% of the polyglot markup community itself can’t get it right, why bother?"
13:15
<annevk>
XHTML was hard and failed with the web developer community. So lets try to be HTML and XHTML at the same time, surely that's simpler!
13:15
<annevk>
Lessons learned: less than zero.
13:17
<odinho>
Hmm. I really thought the tests-at-github thing was resolved.
13:17
<odinho>
Would be uncool having to do it all over again for webapps and webappsec.
13:19
<annevk>
Man ReSpec sucks with fragment identifiers...
13:20
<annevk>
FOUC and you're at the top rather than what you were trying to read...
13:23
<jgraham>
odinho: Well last time it came up, ArtB basically said "no experimenting in my working group"
13:23
<odinho>
jgraham: Oh, I just sent an email saying "can we move now plzs"
13:23
<annevk>
My feeling at that point is that ReSpec appears to optimize for editors at the expense of everyone else, but actually optimizes for ReSpec at the expense of everyone.
13:23
<jgraham>
So now that the idea is proven in HTML, it should be OK to do in WebApps
13:23
<odinho>
jgraham: Anyway html tests already did it. So it's not experimental any more.
13:24
<jgraham>
But of course there might be some other reason given not to do it
13:24
<jgraham>
You going to review the pull requests btw?
13:27
<odinho>
Was intending to take a look
13:27
<jgraham>
Cool
14:03
<Ms2ger>
odinho, fwiw, I'd prefer having one repo for html+webapps+...
14:27
<Ms2ger>
darobin, AFAIK, yes
14:27
<annevk_>
Ms2ger: IDB also doesn't define event handlers very clearly in terms of HTML
14:30
<Ms2ger>
annevk, it's better at defining other things?
14:31
<darobin>
Ms2ger: I can't find usage info on this, are you sure it's well deployed, even in mobile?
14:31
<jgraham>
Ms2ger: Out of interest, why one repo?
14:31
<Ms2ger>
jgraham, why two?
14:31
<annevk>
Ms2ger: that was rhetorical?
14:32
<Ms2ger>
jgraham, I just don't see the point of silo'ing things by working group
14:32
<jgraham>
Ms2ger: Fair enough. I don't think it's a very serious problem because in practice it isn't hard to move things from one repo to another if needed
14:33
<Ms2ger>
jgraham, especially since I seem to remember that people submitted tests for WebApps deliverables to the HTML test suite before
14:33
<darobin>
I hear a lot of people asking for an integrated TS with everything
14:33
<jgraham>
And smaller repos can be nice
14:33
<darobin>
no one gives a shit about how things are cut up into groups
14:33
<annevk>
yeah
14:33
<annevk>
that's why people like #whatwg
14:34
<jgraham>
darobin: But, Process man, Process
14:34
<darobin>
mmmm, patents — not process
14:34
<Ms2ger>
Webapps wouldn't be too bad, but figuring out where tests for Performance or Touch Events go...
14:34
<darobin>
you wouldn't have RF if there weren't groups
14:34
<darobin>
(no matter how much that may suck)
14:34
<annevk>
you can have groups but shared tests...
14:35
<jgraham>
darobin: I know, but now I can't tell what you are advocating for tests
14:35
<annevk>
they're not correlated that way
14:35
<darobin>
at the same time, I got a lot of complaints at TakeOff about how participating in standards involved receiving tons of email about stuff you don't give a shit about
14:35
<darobin>
it seems people don't want to care where specs belong, but would like one list per spec, or even per feature
14:35
<darobin>
Making Communities Happy Is Hard: Film at 11
14:36
<Ms2ger>
More lists tends to lead to more email too, unfortunatel
14:36
<Ms2ger>
y
14:36
<darobin>
jgraham: I tend to lean towards Ms2ger's position — have just the one repo
14:36
<annevk>
Yeah, more lists doesn't necessarily mean more productive discussion. And you'll waste a bunch of time when reviewing to figure out where feedback has to go. I hit that problem each time I review.
14:36
<darobin>
someone even told me "I like the HTML WG better than WHAT WG because there's almost no email there nowadays"
14:36
<darobin>
I tried hard not to giggle
14:37
<Ms2ger>
darobin, I failed :)
14:37
<annevk>
We could just give everyone their own mailing list with no email.
14:37
<darobin>
yay!
14:38
<Ms2ger>
indexedDB of type IDBFactory, readonly
14:38
<Ms2ger>
This attribute provides applications a mechanism for accessing capabilities of indexed databases.
14:38
<jgraham>
public-annevk⊙wo? I like
14:38
<darobin>
Ms2ger: you failed at something in particular or just in general?
14:38
<Ms2ger>
darobin, I failed not to giggle
14:38
<darobin>
jgraham: and then annevk would unsub from that list due to the copyright commitment
14:38
<darobin>
Ms2ger: :)
14:39
<Ms2ger>
And create annevk⊙lwo instead?
14:40
<annevk>
darobin: heh, I'll keep waving that flag until it's settled
14:40
<annevk>
darobin: don't think it's a problem for email though, but I might be missing something
14:40
<darobin>
you need guns, not flags
14:42
<jgraham>
Guns that pop flags out of the end?
14:44
<darobin>
jgraham: you never tell — that's the point
14:44
<darobin>
hmmmm, just received the first valid argument against using GH to date
14:45
<darobin>
being this: http://viewdns.info/chinesefirewall/?domain=github.com
14:45
<darobin>
that's annoying
14:46
<annevk>
Ms2ger: Indexed DB returns arbitrary objects? Are those described in sufficient detail?
14:47
<annevk>
Whoa, they haven't blocked my site.
14:49
<Ms2ger>
"Getting this property must return the source for the request. Returns null when there is no source set."
14:49
<Ms2ger>
darobin, http://viewdns.info/chinesefirewall/?domain=bitbucket.org :)
14:50
<hsivonen>
darobin: China can DNS poison any domain any day. Any Web-based tool you choose today can be banned tomorrow.
14:50
<darobin>
Ms2ger: which just shows how unused it is :)
14:50
<hsivonen>
darobin: so it doesn't make much sense to choose tools based on what's banned in China
14:50
<darobin>
hsivonen: I know, but it's a small factor to take into account
14:51
<darobin>
I don
14:51
<darobin>
gah
14:51
<darobin>
I don't like the idea of banning chinese contributors, even if it's not our fault
14:51
<Ms2ger>
hsivonen, clearly with so many Chinese members (http://lists.w3.org/Archives/Public/www-archive/2013Jan/0019.html), they wouldn't be able to block anything W3C hosts :)
14:52
<hsivonen>
darobin: well, Finland blocked lists.w3.org as child porn one time. Should we stop using lists.w3.org?
14:52
<darobin>
hsivonen: I'm not talking about not using
14:52
<darobin>
I'm pondering proxying
14:52
<darobin>
totally different notion :)
14:52
<annevk>
thanks Ms2ger
14:52
<hsivonen>
darobin: ah
14:53
<hsivonen>
http://hsivonen.iki.fi/access-blocked/ for those with short memory
14:53
<darobin>
to be fair though, I don't think that preventing children from accessing lists.w3.org is a bad idea :)
14:53
<darobin>
it may forever scar one
14:53
<Ms2ger>
darobin, I don't think it's just children that are blocked from child porn
14:54
<hsivonen>
oops. it was www.w3.org not lists.w3.org
14:54
<darobin>
oh, child porn, I had misparsed hsivonen's initial line
15:03
<hsivonen>
hmm. the grammar coherence of my last email to public-html is unusually bad
15:04
<hsivonen>
forgetting the start of the sentence before getting to the end of the sentence
15:04
<hsivonen>
multiple times
15:04
<darobin>
ah, the well know problem of timblificiation
15:06
<darobin>
if you start also speaking that way, you might want to consider joining a support group
15:58
<Stevef>
darobin: spam man
15:58
<darobin>
\o/
15:58
<darobin>
Stevef: well, *if* I can succeed
15:59
<darobin>
right now I'm getting a 502 on updating the DB
15:59
<Stevef>
so do we have a 5.1 componenet now?
15:59
<darobin>
no, it's just HTML5 spec
16:00
<darobin>
I don't see much of a point in adding a new component for 5.1, we just keep working on the same bug set
16:00
<Stevef>
OK so do we use HTML5 spec component for 5.1 bugs?
16:00
<darobin>
yup
16:00
<Stevef>
cool
16:00
<darobin>
there's a CR component for things that should definitely target the CR draft
16:00
<zewt>
(there's also no such thing as "html5.1" dot dot dot)
16:01
<Ms2ger>
There's just a HTML :)
16:01
<Stevef>
if you believe there is
16:01
<zewt>
pretty telling that hixie spends years getting everyone to understand that html is unversioned, then w3c people go "yay, hixie's gone! now we can finally increase the version number!"
16:02
<zewt>
(heh)
16:02
<Stevef>
yeah thise bastard W3C people
16:02
<Stevef>
those
16:02
<darobin>
I don't care, all I want is the right set up for "This is not the version you're looking for" jokes
16:02
<zewt>
call it a lumping together :P
16:02
<darobin>
the rest is immaterial
16:03
<Ms2ger>
Twss
16:03
<Stevef>
besides its not HTML5.1 its HTML 5.1
16:03
<Stevef>
nightly no less
16:09
<Stevef>
the one fortified with <main> ... runs and hides
16:09
<darobin>
oh yeah, forgot that we also want to have the ability to make "cast thy coloured nightlies and look upon Denmark as a friend" jokes
16:10
<Stevef>
darobin: did you see my whine on htmlwg irc about wonky 5.1 links?
16:10
<Ms2ger>
Meanwhile, the WHATWG server is powered by Shakespeare spinning in his grave
16:11
<darobin>
Stevef: nope, but I am already aware that something broke the build script
16:11
<darobin>
that would break links
16:11
<darobin>
Ms2ger: I'm all about renewables :)
16:11
<Stevef>
FYI weird stuff going on with html 5.1 follwoing a link in multipage redirects to single page - results in browser hang...
16:12
<darobin>
Stevef: yeah, I don't know what broke us but it broke us bad
16:12
<darobin>
last time it was a very poorly placed closing element
16:16
<darobin>
~/ Bugs like violence, Break the silence, Come crashing in, Into my little mailbox /~
16:42
<jgraham>
Ms2ger: I thought there was only XUL?
16:42
Ms2ger
eyes jgraham
17:02
<annevk>
hsivonen: we could consider utf-8 a special label maybe just like utf-16
17:02
<annevk>
hsivonen: guess you'd want telemetric data on how often the encoding is changed from utf-8 to something else
17:40
<MikeSmith>
so <main> just now landed in WebKit
17:40
<MikeSmith>
http://trac.webkit.org/changeset/140341
18:42
<MikeSmith>
"content-type (mime type) squatting" is an apt was to characterize polyglot http://lists.w3.org/Archives/Public/www-tag/2013Jan/0049.html
19:32
<MikeSmith>
wow some quality Bjoern-age to www-archive just now
19:33
<MikeSmith>
http://lists.w3.org/Archives/Public/www-archive/2013Jan/0021.html
19:34
<MikeSmith>
2000 words worth
19:44
<Ms2ger>
Hmm, "Angstklauseln"
19:50
<Ms2ger>
And it ends with a shot at Hixie, nicely done
19:55
<MikeSmith>
"Angstklauseln" is a great word
19:55
<MikeSmith>
I wonder if that's actually commonly used or if he coined it
19:58
<jgraham>
http://de.wikipedia.org/wiki/Disclaimer
19:59
<jgraham>
Links to http://angstklauseln.wordpress.com/
20:01
<MikeSmith>
ah
20:01
<MikeSmith>
ok
20:01
<MikeSmith>
well we should use the equivalent in English
20:01
<MikeSmith>
it's an apt description
20:02
Ms2ger
checks if http://angstklauseln.wordpress.com/ is run by Bj�rn
20:02
<MikeSmith>
hah
20:02
<Ms2ger>
Apparently not
20:04
<MikeSmith>
"Frankenmails" is not up to par though. SpuckteAntworten would be better
20:04
<MikeSmith>
(for the shot at Hixie)
20:06
<MikeSmith>
speaking of Hixie, here's Hixie and BruceL in Japanese: http://www.publickey1.jp/blog/13/hixie_html5hixie.html
20:06
<MikeSmith>
retitled "Why does Hixie always say No?"
22:54
<heycam>
MikeSmith, thanks, fixed the link
22:55
<heycam>
hsivonen, Hixie, I don't think SVG has any plans on introducing hyphenated element names. the existing ones are solely for css-descriptors-in-element-form, and we really shouldn't have those anyway
22:55
<heycam>
hsivonen, Hixie, the existing hyphenated ones might well all be removed from SVG 2 too
23:02
<heycam>
TabAtkins, is @host meant to also replace @global?
23:03
<heycam>
TabAtkins, so you could write for example: @host { body > :scope > p { … } }
23:03
<heycam>
?