01:02
<gsnedders>
Hmm, is there any good way to mark up verse with using caesura and represent it (using CSS) as a long space?
01:04
<gsnedders>
*caesurae if I actually decline correctly
01:10
<gsnedders>
Most existing things seem to use &nbsp;, which isn't great.
01:14
<gsnedders>
(esp if you want to line up the half-lines, assuming that's how the caesura is used))
02:00
<Hixie>
volkmar: awesome, thanks. interesting that you are not looking at doing script-related inputmode control; is that not needed?
02:02
<Hixie>
pablof: they seem right, am i missing something?
02:06
<Hixie>
gsnedders: <span class="caesura">&#x1D113;</span> and hide the content with CSS? :-)
02:06
<Hixie>
but that's more a musical caesura
02:13
<gavinc>
Wrong place to ask, but does anyone know the best way to contact someone at the W3C about membership issue? (Email to membership unanswered for a while)
02:50
<Hixie>
gavinc: your AC rep, if you have one
02:50
<Hixie>
gavinc: failing that, try ij@ maybe?
03:17
<gavinc>
Hixie: Yeah, am the AC rep ;) trying to figure out why our membership didn't start on the first
03:18
<Hixie>
ah :-)
03:18
<Hixie>
try ian jacobs (ij@)
03:18
<Hixie>
when MikeSmith is around he can help too, probably by telling you to contact ij@ :-)
03:19
<MikeSmith>
gavinc: what Hixie said :)
03:19
<MikeSmith>
gavinc: I can check the DB at least
03:37
<MikeSmith>
Hixie: the script is done right?
03:37
<Hixie>
yup
03:37
<MikeSmith>
k
03:37
<Hixie>
there remains the issues i sent the chairs in the e-mail earlier (you were cc'ed)
03:37
<Hixie>
but i don't plan on doing anything on those unless they ask me to
03:38
<MikeSmith>
OK
03:39
<MikeSmith>
Hixie: you don't want any list Cc'ed on that component?
04:09
<Hixie>
not particularly
04:20
<volkmar>
Hixie: script-related input mode control?
04:20
<Hixie>
like, chinese simple vs traditional
07:45
<hsivonen>
so much "wrong tomorrow" with smart card Web logins
07:46
<hsivonen>
printed pad of passcodes each of which is used once will always be more compatible than any hardware-based solution
07:46
<hsivonen>
but even as far as hardware goes, USB connectors are more common now than smartcard slots
08:27
<annevk>
Is there a way to combine inputmode and autocompletetype somehow?
08:28
<annevk>
From the autocompletetype anyway the inputmode is often implied
08:39
<MikeSmith>
annevk: so it turns out the HTTP spec does specifically disallow whitespace in media-type parameters
08:39
<MikeSmith>
though it does allow it in other parameters
08:39
<MikeSmith>
http://tools.ietf.org/html/rfc2616#section-3.7
08:40
<AryehGregor>
annevk, I had a comment in that vein at some point, saying that there should be some higher-level attribute along the lines of "what type of input is expected" without specifically tying it to autocomplete or whatever.
08:40
<MikeSmith>
"Linear white space (LWS) MUST NOT be used between the type and subtype, nor between an attribute and its value."
08:40
<MikeSmith>
I see now that PhilipJ pointed this out in a list discussion last year
08:40
<annevk>
what is an attribute in that context?
08:41
<annevk>
they mean parameter?
08:41
<MikeSmith>
the first half of the parameter before the equal sign
08:41
<annevk>
ah okay
08:41
<MikeSmith>
the define the parameter as an attribute+value pair
08:41
<MikeSmith>
"Parameters MAY follow the type/subtype in the form of attribute/value pairs"
08:41
<annevk>
so charset = utf-8 would not work?
08:42
<MikeSmith>
yeah and it doesn't work actually
08:42
<MikeSmith>
in most browsers
08:42
<MikeSmith>
I think it works only in IE
08:42
<annevk>
really?
08:42
<MikeSmith>
other browsers ignore the parameter
08:42
<MikeSmith>
yeah try it
08:42
<MikeSmith>
you have to make sure you have charset-sniffing turned off in the browser
08:43
<MikeSmith>
and make sure it's not set to default to UTF-8
08:43
<MikeSmith>
Julian has a test case
08:43
<MikeSmith>
lemme get the URL
08:43
<MikeSmith>
http://greenbytes.de/tech/tc/httpcontenttype/textplainutf8ws.asis
08:43
<MikeSmith>
summary of results for various browsers is at http://greenbytes.de/tech/tc/httpcontenttype/#textplainutf8ws
08:44
<MikeSmith>
though when I try it in Chrome it does seem to use UTF-8 no matter what
08:45
<MikeSmith>
I can't tell whether it's getting it from the header or not
08:45
<MikeSmith>
but Opera and Firefox both ignore it at least
08:47
<MikeSmith>
hsivonen: do you think it's worth updating the charset part of the ContentTypeParser code to make it proper error-reporting parser instead of that existing regexp-based one?
08:47
<MikeSmith>
if so I think I'll go ahead and do that
08:49
<MikeSmith>
it seems good to alert a user with a useful error message for the "text/html; charset = utf-8" case if that's not going to actually work as expected in browsers
08:56
<hsivonen>
MikeSmith: yeah, that would be nice
08:56
<MikeSmith>
ok
08:57
<hsivonen>
the JS i18n API will suck for non-top-tier locales that won't be supported in all browser from day 1
08:58
<hsivonen>
the WHATWG-style notion of interop is rather incompatible with the politics of never closing the door on the set of supported locales
08:59
<hsivonen>
hmm. that statement wasn't quite right
08:59
<hsivonen>
the WHATWG-style notion of interop is rather incompatible with the politics of not having a normative list of which locales are in the supported set today
09:01
<hsivonen>
eventually, the standard will be "whatever ICU supports" and everyone has to ship ICU as the back end impl
09:01
<hsivonen>
except Microsoft probably won't an will roll their own
09:09
<annevk>
ICU is kind of terrible
09:10
<annevk>
at least for encodings the architecture is way too complex
09:25
<annevk>
MikeSmith: I wonder if other headers with similar parameter-style are parsed in the same way
09:26
<MikeSmith_>
annevk: I think not
09:26
<annevk>
MikeSmith: this does not sound like the way e.g. <meta http-equiv=content-type> is parsed
09:26
<MikeSmith_>
I think Content-Disposition is not, for one
09:26
<annevk>
and they call HTML a mess
09:26
<MikeSmith>
heh
09:26
<MikeSmith>
yeah
10:24
<david_carlisle>
MikeSmith: (or anyone) If I want to comment on one of these bugs that's been cloned into HTML WG and WhatWG components, am I supposed to comment on the old one, the new one or both?
10:24
<MikeSmith>
the new one
10:24
<MikeSmith>
I suggest
10:25
<Ms2ger>
Then again, we would say that in this channel
10:33
<david_carlisle>
MikeSmith: OK thanks, it might be worth posting to html and or wg lists saying what just happened, I got several mails I expect some got a lot more:-)
10:34
<MikeSmith>
yeah sorry about the bugmail bomb
10:34
<MikeSmith>
I got a lot more :-)
10:34
<david_carlisle>
MikeSmith: good:-)
11:09
<annevk>
david_carlisle: yeah, the clarifying email will come
11:09
<annevk>
david_carlisle: prolly end of this week or early next, whenever Hixie gets around to it
11:10
<annevk>
david_carlisle: basically, the HTML WG forked development on HTML
11:11
<david_carlisle>
annevk: yes but I hoped they ony _say_ they forked not actually have different bugs with different resolutions, which would be ... unfortunate
11:11
<Ms2ger>
Well
11:11
<david_carlisle>
only
11:11
<Ms2ger>
More likely they'll have a lot of bugs without resolutions instead
11:12
<david_carlisle>
Ms2ger: good plan
11:25
<hsivonen>
does the HTML WG have an editor for its fork, yet?
11:26
<hsivonen>
I wonder if someone has scientifically figured out why the WebApps WG kinda works but the HTML WG doesn't
11:35
<odinho>
Culture.
11:37
<Ms2ger>
Are you a scientist? :)
11:43
<odinho>
Òó
12:03
<annevk>
hsivonen: chairs that tolerate bullshit?
12:03
<annevk>
hsivonen: maybe also because WebApps is more difficult so you get less bullshit to start with
12:04
<annevk>
hsivonen: well, HTML is deceptively simple anyway, not actually
12:09
<hsivonen>
annevk: yeah, I'd prefer the chairs to be more like "can't behave like that here" or "nope, not gonna do that" bouncers as opposed to putting everything to a poll out of fairness
12:15
<odinho>
Ms2ger: Absolutely not ;]
12:32
<hsivonen>
it's unfortunate that the Security Considerations section of CORS doesn't cover the concern related to authentication by IP address or authentication by being on the right side of a firewall
12:35
<odinho>
That sections has been a bit back and forth IIUC.
12:37
<hsivonen>
time and again there's always someone new who doesn't realize that CORS protects network resources whose authentication relies on the network topology
12:48
<odinho>
true dat
12:50
<hsivonen>
another repeating thing: Ars writes that new release of browser Foo now implements API Bar without explicitly spelling out that care was taken in making API Bar secure. Reader in the comments freaks out assuming that the security issues related to the subject matter of API Bar haven't been addressed or thought about.
13:01
<hsivonen>
I wonder if hober is on track to become the new HTML WG editor considering all the CPs...
13:01
<annevk>
the W3C is trying to hire someone for it, in case you missed that
13:02
<annevk>
as a fulltime job
13:02
<annevk>
I think it's the first time the W3C offers to pay an editor
13:04
<hsivonen>
annevk: interesting. I missed that.
13:04
<hsivonen>
annevk: have you applied for the job now that you are no longer at Opera? :-)
13:06
<Ms2ger>
hsivonen, nah, he's going to do XSLT :)
13:06
<annevk>
hsivonen: not sure why I would put time in what Hixie is already doing quite well
13:07
<hsivonen>
annevk: speaking of editorships, are you keeping the editorship of the specs you were editing as an Opera employee?
13:08
<annevk>
hsivonen: yeah, I plan to apply for invited expert status
13:09
<hsivonen>
annevk: cool. I hope whatever your new business model is allows you to keep editing.
13:10
<annevk>
hsivonen: I hope so too :)
13:22
<gsnedders>
Hixie: U+1D113 is the musical symbol, which is different to what you want in verse.
13:33
hsivonen
finds http://www.w3.org/Consortium/Recruitment/#techeditor
14:19
<annevk>
" I believe the WG already has consensus on this point." (on public-webappsec)
14:19
<annevk>
not sure why we need to have consensus on facts, but I guess...
14:29
<hsivonen>
hooray. not declaring the character encoding is now invalid.
14:29
<hsivonen>
so I get to remove some complexity from the HTML parser's Java version
14:52
<odinho>
hsivonen: Oh the joys in life!
14:58
<MikeSmith>
hsivonen: I just finished rewriting the Content-Type parsing
14:58
<MikeSmith>
testing it now
14:58
<MikeSmith>
will send you a patch when I think it's ready
16:41
<Ms2ger>
Philip`, any chance you could put a warning on http://philip.html5.org/tests/canvas/ that those tests aren't updated anymore?
16:46
<dglazkov>
good morning, Whatwg!
16:58
<Hixie>
annevk: where would you suggest i e-mail, and what would you suggest i say in said e-mail?
17:55
<MikeSmith>
so I made a patch that adds an error-reporting Content-Type header parser to the validator
17:55
<MikeSmith>
and pushed it to http://qa-dev.w3.org:8888/
17:56
<MikeSmith>
please test
17:56
<MikeSmith>
or if anybody has any Content-Type test cases please point me to them
18:51
<Hixie>
there's some content-type tests on hixie.ch/tests/adhoc/http i think, dunno if they'd help
19:20
<sunshinehappy>
te//o
19:25
<sunshinehappy>
hello
19:25
<sunshinehappy>
what is it
19:26
<Hixie>
what is what
19:27
<Philip`>
Ms2ger: Yes, if I still remember by the time I have a proper computer+internet set up, and am not too grossly lazy
19:29
<sunshinehappy>
whatwg
19:31
<Hixie>
sunshinehappy: it's where we write the html standard and some other specs
19:31
<sunshinehappy>
why do you guys write HTML standard? I thought that's what W3C do
19:32
<Hixie>
we work with the w3c
19:32
<Hixie>
they're currently more focused on releasing a snapshot
19:32
<sunshinehappy>
cool!
19:32
<Hixie>
we're doing the work on future stuff
19:32
<Hixie>
(thought right now they are not making any changes, so they're also missing the fixes we're making to the old stuff)
19:32
<Hixie>
(but they plan on fixing that)
19:33
<sunshinehappy>
what's are the ideas for future?
19:33
<Hixie>
MikeSmith: do you know if r12a is on irc anywhere?
19:34
<Hixie>
sunshinehappy: right now i'm doing research for an inputmode="" attribute on <input> to help people write web pages for mobile phones
19:34
<Hixie>
sunshinehappy: so you can say e.g. "this input field should have automatic capitalisation" or "this should default to all lowercase"
19:34
<sunshinehappy>
cool :)
19:34
<Hixie>
sunshinehappy: trying to fill in this page: http://wiki.whatwg.org/wiki/Text_input_keyboard_mode_control
19:34
<Hixie>
unfortunately i'm finding it hard to find any information on how input modalities work in asian markets
19:35
<Hixie>
in particular, chinese, japanese, and korean
19:35
<Hixie>
also vietnamese i hear needs something special but i can't find any data on it
20:05
<loincloth>
Hixie: that example and a number of HTML5 form features strike me as blurring the line between markup and behaviors.. do i miss the point or is that a legitimate observation?
20:05
<loincloth>
i'm sure HTML5 wasn't the first to dance around that line, either.. i sometimes think JS and CSS should be one language, too, so i'm just thinking out loud.. curious for reactions
20:35
<Tabatkins>
loincloth: We've been blurring that line for forever. It's not a sacrosanct barrier.
20:36
<Tabatkins>
HTML blends content and behavior in a lot of form stuff, and <details>, etc. It blend content and style in its half-presentational elements. CSS blends style and content with pseudo-elements. It blends style and behavior with :hover/:focus/:active, and Animations. JS blends everything, because it's a fully-powered programming language.
20:44
<loincloth>
Tabatkins: indeed.. as i said, i know this musing isn't new.. but a new example was given that sparked further discussion
21:19
<Hixie>
loincloth: not sure what the difference is
21:20
<loincloth>
Hixie: between?
21:34
<Hixie>
loincloth: markup and behaviours
21:35
<Hixie>
loincloth: or rather, they seem to be orthogonal
21:35
<Hixie>
loincloth: you can use markup to describe behaviours... it's like the difference between "english" and "poetry"
21:44
<annevk>
Hixie: whatwg⊙wo and maybe public-html⊙wo
21:45
<annevk>
Hixie: just state that from the perspective of the WHATWG we'll keep updating HTML and do so by paying attention to the WHATWG/HTML Bugzilla component and that for W3C HTML people can contact the W3C
21:46
<annevk>
Hixie: and maybe elaborate on the scripts you wrote to keep the two parties in sync as far as incoming information goes
21:49
<Hixie>
well my goal for keeping stuff in sync in the future is just that i'll keep an eye on the w3c side
21:50
<Hixie>
and the chairs haven't asked me to do anything for their side
21:50
<Hixie>
so e.g. bugs aren't being automatically filed on both sides
21:51
<Hixie>
not sure i should post anything to public-html, i'd probably get flamed to death
21:51
<annevk>
I think we should at least inform whatwg⊙wo about why some people received an email about a duplicated bug
21:51
<Hixie>
fair enough
21:51
<annevk>
and how W3C HTML and WHATWG HTML are no longer in sync
21:52
<Hixie>
might be worth you doing a blog post about it too
21:52
<Hixie>
so feel free to crib from the e-mail i'm going to write
21:53
<annevk>
yeah was planning on doing that
21:53
<annevk>
maybe I'll include one of the memes people made
21:53
<Hixie>
there are memes?
21:53
<annevk>
http://w3cmemes.tumblr.com/post/27561016967 is fun
21:53
<annevk>
http://w3cmemes.tumblr.com/post/27558722901/stereotypical-html-wg-poll-respondent about the polls is kind of funny too
21:54
<Hixie>
heh
21:54
<Hixie>
let's not insult w3c too prominently... :-P
22:07
<othermaciej>
I wonder what this one's about? http://w3cmemes.tumblr.com/post/27215424834/the-most-interesting-standards-guy-in-the-world
22:10
<smaug____>
that is a good one :)
22:12
<annevk>
othermaciej: ask hober ;)
22:12
<annevk>
nn
22:16
<pablof_>
Hixie: re: http://www.whatwg.org/specs/web-apps/current-work/multipage/association-of-controls-and-forms.html#dom-textarea/input-setrangetext
22:17
<pablof_>
Hixie: in step 6, shouldn't "selection start" be "start" and "selection end" be "end"?
22:17
<pablof_>
Hixie: also, in step 8, for "select", shouldn't "selection end" be set to "start + replacement.length()"? similar for mode "end"
22:17
<Hixie>
yikes, that does seem plausible
22:18
<Hixie>
can you file a bug using the widget at the bottom right?
22:18
<pablof_>
sure
22:20
<Hixie>
thanks
22:49
<Hixie>
ok, long rambling administrivia e-mail sent
23:08
<heycam>
the name "operation convergance" is presumably ironic then?
23:09
<zewt>
operation fragment bug discussion
23:09
<Hixie>
heycam: not really, the worry was that if we didn't clone the bugs that each bug would only be dealt with by one bug, so it'd be even worse
23:10
<Hixie>
heycam: at least this way all the bugs are handled by both groups, so whoever gets to a bug second can first check what the other group did and try to keep in step
23:10
<heycam>
Hixie, I see
23:10
<Hixie>
heycam: it's pretty inevitable that the specs will diverge more, though, imho
23:10
<heycam>
Hixie, yeah
23:10
<heycam>
Hixie, hopefully when the W3C find an editor on their side that they can keep up
23:10
<Hixie>
yeah well that's another thing
23:11
<Hixie>
i don't think it'll be physically possible for the wg to keep up, given their process
23:11
<Hixie>
whoever they have as editor
23:11
<heycam>
interesting times, then
23:11
<Hixie>
i mean, one of the reasons i was in favour of this split is that i was being slowed down by the process
23:11
<Hixie>
and i didn't have to keep up with anyone...
23:11
<heycam>
heh yeah
23:14
<Hixie>
interesting times indeed, i'm really quite honestly curious to see where this goes
23:34
<MikeSmith>
Hixie: thanks will take a look at those content-type tests
23:35
<MikeSmith>
and no Richard's not around on irc right now
23:35
<MikeSmith>
he's not always on so many hours
23:35
<MikeSmith>
I think he may come on when he has meetings and drop off sometimes when he's working
23:36
<MikeSmith>
but if it's about the input-mode stuff, I can help with modalities for Japanese