| 04:52 | <shepazu> | Hixie: I believe it is automated |
| 04:53 | <shepazu> | Hixie: you don't seem to have automatically been added, however… |
| 04:53 | <Hixie> | yeah, that's normal |
| 04:53 | <shepazu> | you should get Raman to add you, so you can be nominated chair |
| 04:53 | <Hixie> | right |
| 04:54 | <shepazu> | (seems odd to me… if you propose a CG and it gets accepted, you should be added automatically… but whatever) |
| 04:55 | <shepazu> | Hixie: are you planning on backporting old email archives into the new list, or what? |
| 04:55 | <Hixie> | we'll just use the existing list |
| 04:56 | <Hixie> | it basically changes nothing until we do an FSA -- see the e-mail to the WHATWG list |
| 04:56 | shepazu | heard some rumors of W3C archiving WHATWG emails for posterity and searching, etc. |
| 04:56 | <shepazu> | ok, thanks |
| 04:56 | <Hixie> | yeah, i heard that might be happening |
| 04:56 | <Hixie> | it'd be awesome if so, the mailing list software the w3c uses is pretty awesome |
| 04:58 | <shepazu> | Hixie: I'm not in the loop, but I thought it was already happening… if nothing else, you could import them into the automatic WHATCG list |
| 04:58 | <Hixie> | yeah i'm not in the loop either :-) |
| 04:59 | <othermaciej> | there's a loop? |
| 04:59 | <shepazu> | othermaciej: of course, what do you think W3M has been cycling around in for the past decade :) |
| 04:59 | <othermaciej> | I thought that was more of a drain |
| 04:59 | othermaciej | ducks |
| 05:00 | <shepazu> | hmmm… one cyclical motion might have multiple explanations... |
| 05:00 | <othermaciej> | incidentally I'd guess a bunch of Apple folks will appear on the list once our AC rep clicks ok |
| 05:02 | <shepazu> | and poof! WHATWG has a patent policy! |
| 05:23 | <yuhong> | Now, can MS join? |
| 05:59 | <annevk> | The archiving the WHATWG mailing list on lists.w3.org is something I'm trying to get done, fwiw |
| 06:00 | <othermaciej> | long live WHATCG |
| 06:05 | <annevk> | the logs promised drama, where can I find it in my inbox? |
| 06:05 | <annevk> | jgraham: ? |
| 06:19 | <othermaciej> | annevk: maybe he though today's various announcements would lead to drama, which it appears they did not |
| 06:24 | <zcorpan> | Ms2ger: thanks for the test case :-) |
| 06:24 | <annevk> | yeah, the only confusion seems to be over whether we became a CG or whether we now also have a CG, but it doesn't really matter |
| 06:24 | <zcorpan> | Ms2ger: afaict it should throw 7 |
| 06:25 | <othermaciej> | I believe Hixie intends to use whatwg⊙wo as the CG mailing list (it meets the requirement of being publicly archived), so that's probably closer to becoming a CG |
| 06:26 | <othermaciej> | otoh I don't think he'll ban people who don't join the CG |
| 06:27 | <annevk> | afaik the plan is to 1) make the CG use WHATWG resources and 2) at some point in the future use it to publish "CG standards" so people can sign of on patents |
| 06:29 | <othermaciej> | yes, a CG can publish specifications with some sort of patent commitment |
| 06:31 | <hsivonen> | I wonder if whattw⊙wo could be forwarded to @w3.org to make use of the superior archival software |
| 06:31 | <hsivonen> | bonus points for copying the old archives there, too |
| 06:33 | <annevk> | hsivonen: W3C has been researching if they can archive our mailing list |
| 06:34 | <hsivonen> | I slightly worried about getting a new editor for W3C HTML5 |
| 06:35 | <annevk> | maybe they can make another XHTML 2.0 |
| 06:35 | <hsivonen> | it isn't particularly helpful if W3C HTML5 snapshot starts deviating from the real Living HTML |
| 06:35 | <hsivonen> | though I do see the value of driving it to REC under the PP |
| 06:35 | <hsivonen> | the PP seems to be what's good about RECs |
| 06:36 | <othermaciej> | how well it works out will depend in large part on who volunteers and is appointed |
| 06:36 | <annevk> | I doubt any of the good people are interested in boring maintenance work |
| 06:37 | <hsivonen> | I hope the chairs already had someone willing and appointable figured out ahead of time |
| 06:38 | <othermaciej> | I am a fan of boring but necessary maintenance work, but I don't have the time to volunteer, plus it would be a huge conflict of interest |
| 06:39 | <hsivonen> | hmm. no replies to my readyState email while I was asleep |
| 06:40 | <annevk> | othermaciej: as a chair of the Notifications WG, I was told I could be editor |
| 06:40 | <othermaciej> | www-archive contains a lot of quirky messages from Bjoern Hoehrmann this month |
| 06:41 | <othermaciej> | annevk: I can see how it might work in WGs with less attention on them and less likelihood of conflict than the html wg |
| 06:41 | <othermaciej> | in the past I have already been criticized for posting technical positions or reports on Apple's plans without explicitly saying "chair hat off" |
| 06:41 | <annevk> | hahaha |
| 06:41 | <annevk> | that group is a mess |
| 06:45 | <jgraham> | annevk: This is the part of the drama where there is development of the storyline, not the sort of tense, but ultimately, pointless infighting, that is often labelled drama |
| 06:46 | <jgraham> | There's still plenty of scope for the other kind too of course |
| 06:46 | <jgraham> | Just listen out for changes to the soundtrack |
| 06:48 | <jgraham> | hsivonen: I haven't replied to that email, but it seems quite reasonable to me |
| 06:48 | <jgraham> | I don't yet have anything technical to contribute beyond that |
| 06:48 | <hsivonen> | jgraham: ok. now I need sicking to return from vacation and approve the plan |
| 06:49 | <hsivonen> | it would be annoying to implement it and then get blocked at the review phase |
| 07:01 | <Ms2ger> | zcorpan, you are correct |
| 07:01 | <Ms2ger> | Whether that's desirable... |
| 07:02 | <annevk> | http://blog.whatwg.org/patent-policy |
| 07:03 | <annevk> | hsivonen: othermaciej: hober: ^^ does that blog post read okay? |
| 07:04 | <othermaciej> | annevk: seems reasonable to me |
| 07:04 | <annevk> | cool cool |
| 07:04 | <othermaciej> | I would normally say to include the link for joining the CG, but I think Hixie is holding off on getting too many more people to join until his AC rep approves his own join request |
| 07:05 | <annevk> | that's the first link in the post actually |
| 07:05 | <othermaciej> | ah, I see |
| 07:06 | <othermaciej> | works4me |
| 07:06 | <shepazu> | othermaciej: I have a suggestion about the chair thing, so that nobody else selects a chair before hixie joins... |
| 07:07 | <shepazu> | we could make anne or mike5 or someone else temporary chair |
| 07:07 | <othermaciej> | shepazu: what is your suggestion? |
| 07:07 | <shepazu> | ^^ |
| 07:07 | <othermaciej> | I see |
| 07:07 | <shepazu> | just a thought, might not be necessary |
| 07:07 | <othermaciej> | if Hixie was here I'd ask him |
| 07:07 | <shepazu> | a chair has the ability to elect other chairs |
| 07:08 | <shepazu> | and it is blocked from anyone else |
| 07:08 | <othermaciej> | he should certainly have dibs on picking the initial chair |
| 07:09 | <shepazu> | I'm just thinking we could avoid chicanery |
| 07:09 | <shepazu> | but I'm sure even so, it could be fixed by the systems team to appoint the right chair |
| 07:09 | <annevk> | if that's likely I think Hixie would be fine with Mike |
| 07:10 | <shepazu> | probably not necessary, really, but if you guys think it's worth doing, I'll select Mike5 |
| 07:10 | <zcorpan> | Ms2ger: apparently we pass the test |
| 07:10 | <Ms2ger> | Nice! |
| 07:10 | <othermaciej> | it's only 10 people so far, I wouldn't sweat it |
| 07:10 | <zcorpan> | inorite |
| 07:10 | <annevk> | orly |
| 07:10 | <shepazu> | othermaciej: yeah, I'm sure people will be respectful, and proly raman will push the button tomorrow |
| 07:11 | <Ms2ger> | Opera got it right first |
| 07:11 | <annevk> | damn straight |
| 07:11 | <Ms2ger> | Now, when will you guys ship? In half a decade? :) |
| 07:11 | <zcorpan> | pro tip: if you want to use a custom exception with assert_throws, use an object with a 'name' property |
| 07:12 | <shepazu> | nice writeup of scoped css on css-tricks, link to a jquery polyfill http://css-tricks.com/saving-the-day-with-scoped-css/ |
| 07:12 | <zcorpan> | or rewrite the assert_throws impl |
| 07:12 | <hsivonen> | annevk: I don't understand how the W3C Community Final Spec Agreement binds whatwg@whatwg subscribers who don't join the CG in any way |
| 07:12 | <shepazu> | hsivonen: it doesn't |
| 07:12 | <shepazu> | but that's a social problem |
| 07:13 | <shepazu> | well, I guess it's all a social problem, but I mean, you can ask people to jion |
| 07:13 | <shepazu> | |
| 07:13 | <shepazu> | *join |
| 07:13 | <othermaciej> | it would be nice to get a broad set of people to at least agree to the CLA (which happens via joining) http://www.w3.org/community/about/agreements/cla/ |
| 07:15 | <annevk> | hsivonen: FSA doesn't bind anyone |
| 07:15 | <annevk> | hsivonen: as you can read at the top |
| 07:16 | <annevk> | othermaciej: hmm true |
| 07:16 | <annevk> | othermaciej: I guess we should campaign for that later once the WHATCG website is updated some and everything is in place |
| 07:18 | <othermaciej> | indeed |
| 07:28 | zcorpan | pictures annevk walking around with a sign "PLZ JOIN WHATCG KTHXBYE" |
| 07:29 | <Ms2ger> | zcorpan, I've done the {name} thing :) |
| 07:44 | <shepazu> | zcorpan: http://qkme.me/3oxwgi |
| 07:58 | <annevk> | should blobParts not be a sequence<>? |
| 07:59 | <Ms2ger> | Same difference |
| 08:00 | <heycam> | I tend to think of sequence<> as being a clearer indication of the function not retaining any references to the object passed in |
| 08:00 | <heycam> | and makes it easier not to stuff up if you have a genuine platform array object hanging around |
| 08:01 | <annevk> | maybe methods should only be allowed to take a sequence then? |
| 08:01 | <heycam> | I think there are legitimate reasons for taking an array type, if for example you do have platform array objects |
| 08:02 | <heycam> | but it would be hard to tell from the IDL whether you might have platform array objects being returned from any operation/attribute, or whether it makes sense to accept them as an argument |
| 08:03 | <Ms2ger> | heycam, is there a clear description of the differences between sequences and arrays somewhere in the spec? |
| 08:03 | <heycam> | Ms2ger, there's no section contrasting the two types |
| 08:04 | <Ms2ger> | Would be nice to have ;) |
| 08:04 | <heycam> | wonder what happened to that idea of the web spec best practices guide |
| 08:04 | <Ms2ger> | Oh, robin's thing? |
| 08:04 | <heycam> | yeah |
| 08:04 | <Ms2ger> | I dunno |
| 08:10 | <Von_Davidicus> | Question: How many people use computers so old they only support 8-character file names and 3-character extensions? |
| 08:10 | <zcorpan> | which objects support indexed properties but don't have a 'length'? |
| 08:10 | <zcorpan> | should webidl require them to have a length? |
| 08:11 | <heycam> | zcorpan, I couldn't think of one off the top of my head |
| 08:11 | <heycam> | zcorpan, (if you ignore SVGCrazyLists, which have numberOfItems) |
| 08:11 | <heycam> | zcorpan, (and which haven't been rewritten in Web IDL yet and don't have indexed properties really anyway) |
| 08:11 | <Ms2ger> | heycam, don't they have length as well in Gecko? |
| 08:11 | <jgraham> | numberOfItems? |
| 08:12 | <jgraham> | Really? |
| 08:12 | <jgraham> | That's funny |
| 08:12 | <heycam> | Ms2ger, I think you're right, jwatt added that |
| 08:12 | <heycam> | jgraham, super ergonomic hey |
| 08:12 | <jgraham> | It's self-satire |
| 08:13 | <zcorpan> | heycam: what about the second question? :-) |
| 08:14 | <heycam> | zcorpan, should it require it? maybe... |
| 08:15 | <Ms2ger> | And does it ever make sense to look at the length property of a platform objects that support indexed properties? |
| 08:15 | <heycam> | you know technically you can even define the set of supported indexed properties as { 4, 26, 3943 } |
| 08:15 | <heycam> | not sure what length would be then ;) |
| 08:15 | <zcorpan> | length would be whatever such a spec defines it to be :-P |
| 08:17 | <zcorpan> | i dunno, i think it seems consistent to check length for objects that support indexed properties if they all have a length today |
| 08:17 | <heycam> | but if length remains an accessor property, and you put a length property on the object itself... |
| 08:17 | <annevk> | heycam: if that's technically possible, but nobody requires it, it seems a simpler design would do |
| 08:19 | <zcorpan> | we shouldn't optimize for edge cases evil people like Ms2ger or myself come up with |
| 08:19 | <heycam> | heh |
| 08:19 | <heycam> | I would be in favour of always exposing a length |
| 08:19 | <zcorpan> | very well then :-) |
| 08:19 | <heycam> | maybe even returning it to an instance property instead of a prototype one |
| 08:20 | <zcorpan> | is that how Array has it? |
| 08:20 | <Ms2ger> | Don't listen to the JS people ;) |
| 08:20 | <Ms2ger> | Array and Typed Arrays, I think |
| 08:20 | <zcorpan> | k. i don't have an opinion either way about that |
| 08:20 | <heycam> | on one hand I hear tell of slowness from having the property on the prototype, on the other I hear that maybe native Arrays will have length as a prototype property some time in the future |
| 08:21 | <zcorpan> | is the slowness specific to length or applies to properties in general? |
| 08:22 | <heycam> | I don't know |
| 08:22 | <heycam> | you would think it's everything |
| 08:23 | <heycam> | but it might be showing up in code like for (var i = 0; i < blah.length; i++) ... |
| 08:29 | <Philip`> | Von_Davidicus: I imagine it'd be none that are capable of browsing the web in any kind of bearable way |
| 08:29 | <Philip`> | Von_Davidicus: (since Win95 added long filename support) |
| 08:30 | <Von_Davidicus> | Ah, okay. |
| 08:30 | <Philip`> | Von_Davidicus: I guess the more common case is people with OSes that do support long filenames, but with some kind of weird filesystem (probably a network one) that doesn't |
| 08:30 | <Von_Davidicus> | It was Win95. My mother remembered that Win98 had it. |
| 08:30 | <Von_Davidicus> | *Was trying to remember which Windows added it* |
| 08:32 | Philip` | suggests using Wikipedia instead of trying to remember things :-) |
| 08:34 | <Philip`> | (Win95 came with long names like "C:\Program Files" by default, I think) |
| 08:34 | <Von_Davidicus> | *Reads up on Microsoft Windows.* Tsk, tsk. Such language. |
| 08:35 | <annevk> | http://dev.w3.org/html5/decision-policy/html5-stabilization-plan.html |
| 08:35 | <annevk> | I don't understand Q1b second bullet point |
| 08:36 | <annevk> | it's public-html, not public_html... |
| 08:40 | <annevk> | afaict this will be a gigantic mess |
| 08:40 | <annevk> | that's my prediction for this so-called "stabilization" |
| 08:41 | <Von_Davidicus> | When has HTML ever been anything else? :) *Looks at his own pages, JavaScript and all, and compares it to his bachelor pad. |
| 09:55 | <hsivonen> | Is the HTTP Link header used outside test suites and publications by Hixie and annevk? |
| 10:08 | <annevk> | hsivonen: matjas also uses it to impress his posse :p |
| 10:08 | <matjas> | :') |
| 10:09 | <matjas> | hsivonen: it’s used here, too: http://eligrey.com/blog/post/title-image-files-in-opera |
| 10:15 | <hsivonen> | matjas: thanks. I'm still unconvinced that Link headers should be kept considering their cost/benefit characteristics |
| 10:16 | <annevk> | I'm actually in favor of removing them |
| 10:16 | <annevk> | Hixie is not |
| 10:16 | <annevk> | nobody else cares I think |
| 10:17 | <annevk> | if you can get them removed from Gecko that would be most excellent |
| 10:17 | <annevk> | I'm not sure why we have kept them in Opera |
| 10:17 | <hsivonen> | annevk: it seems Julian cares, considering that I just review a patch of his that improves Link support in Gecko |
| 10:17 | <hsivonen> | *reviewed |
| 10:18 | <annevk> | good point, the few open issues we have in Opera are also per Julian |
| 10:18 | <annevk> | not sure that the HTML and HTTP editor wanting a pet feature in browsers is a good reason to have it though |
| 10:20 | zcorpan | wonders how to match a newline with grep -Ez |
| 10:22 | <annevk> | hsivonen: <meta http-equiv=link> has been removed from Gecko already: https://bugzilla.mozilla.org/show_bug.cgi?id=587928 |
| 10:22 | <annevk> | hsivonen: not yet from Opera |
| 10:22 | jgraham | is +1 on seeing them die fwiw |
| 10:23 | <annevk> | I can't seem to find a bug in Opera on removing it |
| 10:23 | <annevk> | I'll file one |
| 10:27 | <annevk> | hsivonen: is there a bug on Gecko already? |
| 10:27 | <annevk> | hsivonen: I'm willing to file one |
| 10:27 | <zcorpan> | hsivonen: http://simon.html5.org/dump/link-headers.txt |
| 10:28 | <hsivonen> | annevk: I don'tknow |
| 10:28 | <zcorpan> | 7 style sheet links it seems |
| 10:29 | <hsivonen> | zcorpan: out of how many pages in total? |
| 10:30 | <zcorpan> | "roughly 600,000 pages" |
| 10:31 | <zcorpan> | says http://dotnetdotcom.org/ |
| 10:31 | <hsivonen> | zcorpan: thanks. so tiny, tiny fraction |
| 10:31 | <annevk> | hsivonen: https://bugzilla.mozilla.org/show_bug.cgi?id=748294 |
| 10:31 | Philip` | counted more like 400K pages in their data, not 600K |
| 10:38 | <zcorpan> | link: <http://www2.aapg.org/favicon.ico>;; REL=shortcut icon |
| 10:38 | <annevk> | parse error! |
| 10:40 | <zcorpan> | $ grep -aEcz "^HTTP/1\.1" web200904 |
| 10:40 | <zcorpan> | 417486 |
| 10:42 | <annevk> | can't you grep for the null byte separate sequence? |
| 10:42 | <zcorpan> | ok so it's 0.0017% of the pages having a stylesheet Link: (assuming the style sheet links were on different pages) |
| 10:43 | <annevk> | -- Simon "stats" Pieters |
| 10:44 | <hsivonen> | considering how many site-specific duplicate headers there were, counting things per-site rather than per-page probably makes Link even more rare |
| 10:46 | <annevk> | yeah for stylesheet it's only </styles/styles.css> or <../css/YoStyle.css> |
| 11:16 | hsivonen | finally got around to posting about UTF-32 strings in JS |
| 11:16 | <hsivonen> | (to Mozilla's development forum) |
| 11:23 | <annevk> | hsivonen: pointer? |
| 11:23 | <annevk> | hsivonen: note that I think the new proposal is not about using utf-32 anymore |
| 11:24 | <annevk> | hsivonen: http://norbertlindenberg.com/2012/03/ecmascript-supplementary-characters/ suggests an expanded String API so you can index it by code point too |
| 11:25 | <annevk> | found it |
| 11:25 | <annevk> | http://groups.google.com/group/mozilla.dev.platform/browse_thread/thread/b81179d718ed660d |
| 11:28 | <hsivonen> | annevk: well, if I'm late and they are going to stick to UTF-16, hooray |
| 11:33 | <annevk> | hsivonen: they seem to be very bad at updating wiki proposals |
| 11:33 | <annevk> | hsivonen: the wiki page you are pointing to for instance is last updated in 2011 |
| 11:34 | <annevk> | I could have another mini-rant about TC39 not really being that great, but will save that for another time |
| 11:39 | <annevk> | zcorpan: btw, the other day I made http://html5.org/temp/spectable.html illustrating spec encodings versus how they're implemented in browsers combined with label information |
| 11:40 | <annevk> | zcorpan: I will probably use that to add extra labels to the specification; I no longer think that being conservative is a good idea here, though if only one or two implementations support a label it's probably not worth it |
| 11:46 | <zcorpan> | annevk: ok |
| 11:50 | <jgraham> | hsivonen: http://hoppipolla.co.uk/tests/document_readystate/ I started testcasing your demos based on what you said the spec ought to look like |
| 11:52 | <jgraham> | (I haven't got very far as you see) |
| 11:54 | <jgraham> | It's not always obvious what the most significant testable criterion should be |
| 11:55 | <jgraham> | Although I guess I could simplify the whole problem and just build an array of the expected log lines and compare that against a pre-determined array |
| 11:55 | <jgraham> | Actually that could be very quick and easy. Won't make for the best tests though |
| 11:57 | <hsivonen> | jgraham: nice. thanks |
| 12:34 | <zcorpan> | should http://simon.html5.org/specs/quirks-mode#the-:active-and-:hover-quirk apply to querySelector? |
| 12:34 | <zcorpan> | we made querySelector selectors case-insensitive in quirks mode, which suggests this quirk should probably apply as well |
| 12:35 | <zcorpan> | class selectors |
| 12:54 | <annevk> | via marcos: http://www.w3.org/community/groups/proposed/#opentag |
| 12:57 | <zcorpan> | TACG |
| 12:58 | <zcorpan> | up next: W3CG |
| 12:58 | <annevk> | oh, I was thinking "open tag" / "close tag" / "empty tag" |
| 12:59 | <kennyluck> | That paragraph (:active and :hover quirk) is hard to parse :( |
| 12:59 | <zcorpan> | kennyluck: yeah... suggestions? |
| 13:00 | <kennyluck> | zcorpan, make it into a list of conditions? |
| 13:00 | <kennyluck> | By the way, should "that are not part of an argument to a functional pseudo-class or pseudo-element" include ::before:hover in Selectors4? |
| 13:00 | <annevk> | the way we made Selectors case-insensitive was kind of annoying, as in some/all? impl case was preserved in the style sheet |
| 13:02 | <kennyluck> | never mind, I guess ::before never matches :any-link. |
| 13:06 | <zcorpan> | a functional pseudo-class or pseudo-element is like :not(:hover) |
| 13:10 | <zcorpan> | kennyluck: your name is KangHao Lu ? |
| 13:10 | <kennyluck> | Kang-Hao Lu. |
| 13:11 | <zcorpan> | ok fixed, thanks! |
| 13:11 | <zcorpan> | i hope it's equivalent :-) |
| 13:14 | <kennyluck> | I think it. |
| 13:19 | <kennyluck> | zcorpan, is it true that 'left', 'top' and 'bottom' support unitless value but not 'bottom'? That seems super insane. |
| 13:19 | <kennyluck> | unitless length, I mean. |
| 13:19 | <zcorpan> | it's true per spec :-) |
| 13:20 | <kennyluck> | s/and 'bottom'/and 'right'/ |
| 13:20 | <zcorpan> | 'bottom' fell under my cutoff in the data i looked at (see the spec's source) |
| 13:23 | <zcorpan> | (or http://simon.html5.org/sandbox/bookmarklets/reveal-comments ) |
| 13:24 | <kennyluck> | zcorpan, found it. Thanks. |
| 14:26 | zcorpan | added a check for SVG fill="" in http://simon.html5.org/test/quirks-mode/js/hashless-hex-color.html |
| 14:38 | <cbright6062> | I have a question, if it isn't too much of a problem: I know in HTML5, you can use both XHTML formatting for self closing tags, and the now re-standard non-trailing slash closing on such tags. I was curious though, talk of "compatibility with XHTML" aside, which would you say is truly the correct/recommended/best practice method? Thanks. |
| 14:40 | <annevk> | cbright6062: it's community dependent, it's pretty much like asking whether attr='test' or attr="test" is better, or whether your attributes should be alphabetically ordered, or what kind of indentation you should use |
| 14:41 | <cbright6062> | hmm. |
| 14:41 | <Philip`> | cbright6062: Since trailing slashes are ignored, it's probably more confusing to include them - people will mistakenly think that e.g. <script src="foo.js"/> is a correct thing to do, because they'll associate self-closingness with the "/", instead of (correctly) associating it with the tag name |
| 14:41 | <cbright6062> | interesting. I would've guessed there would be a standard for this sort of thing by this point. |
| 14:41 | <cbright6062> | of course, when it comes to the XML/XHTML formatting rules, it's always been a controversial world. |
| 14:42 | <Philip`> | cbright6062: There are - in fact, there are dozens of formatting standards you can pick from :-) |
| 14:42 | <gsnedders> | cbright6062: Look at placement of new lines v. brackets in C-like language — there's plenty of coding standards everywhere. |
| 14:43 | <cbright6062> | I've been adding the trailing slashes myself in the past. Most of this though was because the guy who taught me HTML was an XHTML elitest. lol |
| 14:43 | <zcorpan> | USE ALL THE SLASHES, KTHXBYE |
| 14:45 | <cbright6062> | Regardless, thank you for your help. |
| 15:17 | <jgraham> | Oh, is the idea that <foo/> != <foo> what they mean by "Slash Fiction"? |
| 15:19 | <divya> | I AM OHing that jgraham |
| 16:17 | <Hixie> | hsivonen: the CG won't bind anyone to the CLA, but that's ok, because the CLA only binds people as far as their textual contributions go, and so far with very few exceptions i've written all the text |
| 16:18 | <Hixie> | hsivonen: so it means we can just get the people who have textual contributions to make sign the CLA, and otherwise not worry about it until FSA time. |
| 16:20 | <TabAtkins_> | Yeah, and when you actually publish you can get people to voluntarily make patent grants on the full document, I believe. |
| 16:40 | kennyluck | never understands how these work. |
| 17:40 | <dglazkov> | good morning, Whatwg! |
| 17:40 | <Ms2ger> | Bonsoir |
| 17:41 | <kennyluck> | good midnight. |
| 17:43 | <TabAtkins_> | merry christmas |
| 17:44 | <hsivonen> | Hixie: that leaves a pretty big hole for malicious idea submissions, but so does the REC Process |
| 17:45 | <hsivonen> | Hixie: while it's an imptovement, I think it doesn't solve thatmuch compared to the old WHATWG situation |
| 17:46 | <Ms2ger> | Sounds like the W3C, then |
| 18:06 | <hober> | MikeSmith: what's the purpose of public-whatwg-contrib as being distinct from the whatwg list? |
| 18:07 | <MikeSmith> | hober: dunno |
| 18:07 | <MikeSmith> | I thought it was required for the CLA in some way |
| 18:07 | <hober> | ahh, ok |
| 18:07 | <MikeSmith> | but don't take my word for that |
| 18:07 | <MikeSmith> | I'm not really read up on the CG process/policy stuff |
| 18:31 | <karlcow> | hober: http://www.w3.org/community/about/tool/#lists |
| 18:33 | <hober> | karlcow: thanks |
| 18:51 | <jwalden> | hmm, is whatwg.org down? downforeveryone says yes |
| 18:52 | <Ms2ger> | Is for me |
| 18:52 | <MikeSmith> | kennyluck: do you know who pansb5⊙cc is? |
| 18:53 | <Ms2ger> | China Unicorn? |
| 18:53 | <WeirdAl> | jwalden: www.whatwg.org loads fine for me ( |
| 18:54 | <jwalden> | hm, now it's loading again |
| 18:54 | <jwalden> | oddment |
| 18:55 | <kennyluck> | MikeSmith, I only heard of his name from Chen Bo. My guess is that he quitted China Unicom. |
| 18:56 | <MikeSmith> | kennyluck: think it's OK if I remove his address from the ig list? |
| 18:56 | <kennyluck> | MikeSmith, yes. He never posts to the list and it's only causing us trouble. |
| 18:56 | <MikeSmith> | yeah |
| 18:57 | <MikeSmith> | saw a message from foolip about it |
| 18:57 | <MikeSmith> | oh |
| 18:57 | <MikeSmith> | not showing up in the subscriber list no more |
| 18:57 | <MikeSmith> | so maybe somebody else removed it |
| 18:58 | <kennyluck> | MikeSmith, I sent a mail to sysreq⊙wo I thought they asked you to handle this. |
| 18:58 | <kennyluck> | Did they remove him? Or he just removes himself from the list? |
| 18:59 | <MikeSmith> | oh, it seems he's in the user db |
| 18:59 | <MikeSmith> | will remove him there |
| 18:59 | <kennyluck> | MikeSmith, thanks! |
| 20:18 | <annevk> | hsivonen: it does |
| 20:19 | <annevk> | hsivonen: FSA gives companies the ability to say they will not put any restrictions in place on implementing the specification (i.e. declare it RF from their point of view); the expectation is that at least some will do so |
| 20:19 | <Hixie> | hsivonen: given that i take ideas from everyone including random blog posts, i don't see how to avoid the risk of "malicious idea submissions". |
| 20:19 | <annevk> | hsivonen: we'll have to find out whether it actually happens in practice |
| 20:20 | <Hixie> | hsivonen: (it turns the whatwg patent situation into a situation better than the htmlwg situation, so i think it's a pretty big improvement) |
| 20:36 | <Hixie> | TabAtkins_: do you know if feedback to css is tracked anywhere? (i.e. is there any way to know the status of http://lists.w3.org/Archives/Public/www-style/2012Apr/0572.html ?) |
| 20:36 | <TabAtkins_> | Depends on the spec - tracking is at the discretion of the editor. |
| 20:36 | <TabAtkins_> | In this case, that's fantasai. |
| 20:36 | <TabAtkins_> | who happens to be sitting at my desk with me today. |
| 20:37 | <TabAtkins_> | (But is away momentarily.) |
| 20:39 | <Hixie> | roger |
| 20:39 | <Hixie> | let me know when she's back :-) |
| 20:39 | <TabAtkins_> | she's back. |
| 20:41 | <Hixie> | ask her what the state of my e-mail is :-) |
| 20:41 | <annevk> | did anyone raise the fullscreen extensions on www-style btw? |
| 20:42 | <Hixie> | i thought tantek was going to |
| 20:42 | <Hixie> | since he's still a csswg member (right?) |
| 20:42 | <TabAtkins_> | It's added to Issue 241 in the tracker for css4-text |
| 20:42 | <Hixie> | and he's an editor of the fullscreen spec |
| 20:43 | <TabAtkins_> | tantek is a member. I don't think anyone has yet. |
| 20:43 | <Hixie> | TabAtkins_: k thanks. i don't suppose there's an ETA for when browsers might be implementing anything that results from that issue...? |
| 20:43 | <TabAtkins_> | Nope. |
| 20:43 | <Hixie> | k |
| 20:43 | <Hixie> | i'll just make it a requirement in vtt for now like the emergency wrapping, rather than invent a new keyword or something |
| 20:44 | <TabAtkins_> | Sure. |
| 20:44 | <Hixie> | thanks |
| 20:44 | <annevk> | tantek has not edited the fullscreen spec though he did supply microformat markup for the author data and did some politics around it |
| 20:46 | <fantasai> | hixie: I won't be working on it until writing modes and css3-text hit CR. I would like for that to be soon, but that depends on things outside my control... |
| 20:46 | <fantasai> | hixie: wrt when I'd work on it after that, depends on the interest from implementers in something like that |
| 20:46 | <Hixie> | k |
| 20:47 | <Hixie> | do we have a term that refers to the set of lines resulting from line-wrapping something that in the original tree is one "line"? |
| 20:48 | <Hixie> | e.g. in <pre>foo<br>bar bar bar<br>baz</pre> wrapped with width:0, the string "bar bar bar" |
| 20:48 | <annevk> | [22:49:17.704] TypeError: Image is not a constructor wtf Gecko |
| 20:48 | <fantasai> | hm, I don't think so |
| 20:48 | <fantasai> | I think I just describe it as a consecutive run without forced line breaks or something like that |
| 20:48 | fantasai | checks |
| 20:48 | <Hixie> | k |
| 20:50 | <fantasai> | yeah, don't seem to have a term for it |
| 20:50 | <fantasai> | if you've got a suggestion, I can try to incorporate one into the text... |
| 20:50 | <Hixie> | consecutive works |
| 20:50 | <fantasai> | consecutive what? |
| 20:50 | fantasai | didn't really need the term, so doesn't have one |
| 20:51 | <Hixie> | consecutive bla bla between bla bla |
| 20:51 | <Hixie> | i'll paste my resulting text in a second for comment |
| 20:51 | <fantasai> | ok |
| 20:51 | fantasai | still has no idea what you mean |
| 20:52 | <fantasai> | most of the spec right now just talks about the first line or last line of the set you're talking about |
| 20:52 | <fantasai> | or the "first formatted line", which is something different... |
| 20:54 | <Hixie> | is there a term for "explicit line break that the white-space property honours"? |
| 20:54 | <Hixie> | "preserved newlines" i guess |
| 20:54 | <fantasai> | we call it a forced line break |
| 20:55 | <Hixie> | is that do you have a url to that definition? |
| 20:55 | <fantasai> | best one I've got is file:///home/fantasai/w3c/csswg/css3-text/Overview.html#white-space-collapsing |
| 20:55 | <fantasai> | er |
| 20:56 | <fantasai> | http://www.w3.org/TR/css3-text/#white-space-collapsing |
| 20:56 | <fantasai> | But I should link up that term. |
| 20:56 | <fantasai> | and get you a better link |
| 20:56 | fantasai | adds this to the to-do list |
| 20:57 | <Hixie> | i recommend using "newlines" like CSS2 |
| 20:57 | <Hixie> | so that we don't get terminology drift |
| 20:58 | <Hixie> | ok here's the text i'm currently looking at (not sure i like it, but review welcome while i try to polish it more): |
| 20:58 | <Hixie> | <li>Regardless of the value of the 'white-space' property, |
| 20:58 | <Hixie> | any line breaks inserted by the user agent for the purposes |
| 20:58 | <Hixie> | of line wrapping must be placed so as to minimise Δ |
| 20:58 | <Hixie> | across each run of consecutive lines between preserved |
| 20:58 | <Hixie> | newlines in the source. Δ for a set of lines is defined |
| 20:58 | <Hixie> | as the sum over each line of the absolute of the difference |
| 20:58 | <Hixie> | between the line's length and the mean line length of the |
| 20:58 | <Hixie> | set.</li> |
| 20:59 | <fantasai> | I'd suggest s/preserved newlines in the source/forced line breaks/ |
| 21:00 | <fantasai> | you can have breaks inserted by generated content, for example, so it's not really in the source. |
| 21:00 | <Hixie> | there's no generated content in vtt, since you can't stack pseudos |
| 21:00 | fantasai | rewrote the whole section on white space several times trying to get it all coherent |
| 21:00 | <fantasai> | ok |
| 21:01 | <fantasai> | I think what you're requiring is a bit more stringent than what Glenn Maynard was requesting |
| 21:02 | <fantasai> | or looser |
| 21:02 | <fantasai> | hm |
| 21:02 | <fantasai> | but |
| 21:02 | <fantasai> | basically, I think what he was suggesting was to narrow the containing block width |
| 21:03 | <fantasai> | until the last line came as close to filling the containing block without overflowing as possible |
| 21:03 | <fantasai> | and that would be sufficient |
| 21:03 | <fantasai> | even though it's not optimal |
| 21:03 | <fantasai> | for your definition, you'd also have to consider all the other lines and how they wrapped |
| 21:03 | <fantasai> | you can't just use a greedy line breaking algorithm |
| 21:03 | <fantasai> | you really would have to balance all the lines |
| 21:03 | <Hixie> | that's the idea, yes |
| 21:03 | <fantasai> | that's the ideal :) |
| 21:04 | <fantasai> | but it's probably not necessary to go that far |
| 21:04 | <Hixie> | unless there's a reason to avoid doing the ideal, by definition, we should do the ideal... |
| 21:04 | <fantasai> | performance |
| 21:04 | tantek | scrolls up |
| 21:05 | <tantek> | No I haven't edited the spec since it got checked into w3c hg. I edited the fullscreen spec back when it was on the Mozilla wiki. |
| 21:05 | fantasai | also suspects implementation complexity counts here |
| 21:06 | <fantasai> | what Glenn suggests can be done without changing how line breaking is implemented |
| 21:06 | tantek | still hasn't successfully gotten W3C/CSS hg setup working. Last tried during the Paris f2f, got stuck, documented the problems on the CSSWG wiki hg page, and hasn't had time for wrestling with stupid IT tools since. |
| 21:06 | <fantasai> | just by implementing a separate balancing step that does a binary balancing search, like multi-col |
| 21:06 | <Hixie> | unless someone can demonstrate that trying to balance the lengths of two lines is too expensive, i don't buy "performance" as a reason |
| 21:06 | <tantek> | annevk, hixie, I can raise the fullscreen CSS feature on www-style if you haven't already done so. Happy to kick that off. |
| 21:07 | <fantasai> | Hixie: two lines is fine |
| 21:07 | <fantasai> | Hixie: It's when you have 200 that it gets to be an issue... |
| 21:07 | <Hixie> | tantek: i'm happy for it to just be in teh fullscreen spec, personally :-) |
| 21:07 | <Hixie> | tantek: so anne's your man on that |
| 21:07 | <Hixie> | fantasai: if someone tries to make a 200-line caption, they'll have bigger problems |
| 21:07 | <tantek> | Hixie, it will be just in the fullscreen spec. |
| 21:07 | <fantasai> | Hixie: fair enough :) It's a problem for defining a CSS property, though |
| 21:08 | <Hixie> | oh sure |
| 21:08 | <tantek> | WebApps WG put fullscreen in their charter as well, and wanted to take it over, however I proposed (and CSSWG agreed) that it should be a joint WG effort, and given we have an editor in each group, we're all set. |
| 21:08 | <annevk> | tantek: the "top layer" concept should really be in whatever CSS spec ends up defining boxes properly, but if CSS is not going to do that soon it might as well be in fullscreen |
| 21:09 | <tantek> | annevk - we should define the "top layer" concept in Fullscreen in such a way that Dialog can reference it |
| 21:09 | <annevk> | tantek: I'm not sure the process allows for editors to be in only one of the groups |
| 21:09 | <annevk> | tantek: that's already done |
| 21:09 | <tantek> | great |
| 21:09 | <tantek> | annevk - I don't know of any process requirement for both editors to be in both groups, so I'm not worried. |
| 21:10 | <annevk> | mkay, I don't care much for the politics |
| 21:11 | <Hixie> | man i love having the spec be autogenerated from the CLDR and MathML Unicode data files |
| 21:12 | <Hixie> | every now and then when i regen the spec i get a huge diff that i didn't write that makes it more up to date. :-D |
| 21:13 | <tantek> | annevk - I'll try to minimize your exposure to the politics. |
| 21:14 | fantasai | goes back to the flexbox algo and notes this will require all available stack space |
| 21:15 | fantasai | will look at IRC later |
| 21:15 | <tantek> | it's the least I can do since you've been dealing with the hg edits |
| 21:18 | <annevk> | Hixie: the only problem I guess is that such changes are hidden from the tracker |
| 21:18 | <Hixie> | true |
| 21:19 | <annevk> | and entities changing would be quite shocking I think without prior coordination |
| 21:19 | <annevk> | dunno about the :lang experiment |
| 21:22 | <Hixie> | speaking of which |
| 21:22 | Hixie | adds new "entities" |
| 21:23 | <annevk> | Hixie: to WebVTT? |
| 21:23 | <Hixie> | yeah :-) |
| 21:23 | <Hixie> | ‎ and ‏ |
| 21:23 | <annevk> | is that bidi? didn't those guys prefer markup? |
| 21:25 | <Hixie> | that is markup |
| 21:26 | <annevk> | meant an element |
| 21:30 | <Hixie> | dunno how lrm/rlm would work as an element, unless you mean something like the dir="" attribute |
| 21:30 | <Hixie> | but that seems more complicated to me than just a mark |
| 21:32 | <annevk> | fair enough |
| 21:32 | <annevk> | reminds me of http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2007-July/012235.html |
| 21:42 | <Hixie> | anyone know why zcorpan wanted this? http://www.w3.org/mid/op.v7igmeskidj3kv⊙sl |
| 21:44 | <annevk> | is that how some Date thing serializes maybe? |
| 21:44 | <Hixie> | oh, no, i see |
| 21:44 | <Hixie> | he wants it to say whether it should be 002 or 02 or 0002 etc |
| 21:53 | <annevk> | http://lists.w3.org/Archives/Public/public-i18n-core/2012AprJun/0027.html Norbert Lindenberg seems a little less ambitious than me |
| 21:53 | <annevk> | -> tomorrow |
| 22:13 | <fantasai> | annevk: yeah, they prefer markup, but an exception is made for RLM and LRM since there wasn't any markup that could handle the cases those solved |
| 22:13 | <fantasai> | annevk: most of them are now solved by <bdi> though |
| 22:15 | <tantek> | is there an <ibd> alias for the <bdi> element for those that read right to left? |
| 22:16 | <tantek> | btw - anybody here make it to Lyon for WWW2012? http://www2012.wwwconference.org/ |
| 22:27 | <hober> | tantek: annevk: let me know once fullscreen has the top layer stuff defined |
| 22:27 | <hober> | tantek: annevk: so i can open a webkit but pointing at it :) |
| 22:28 | <tantek> | we just did a security review of Mozilla's Fullscreen implementation yesterday btw |
| 22:30 | <annevk> | hober: a long time ago? |
| 22:30 | <annevk> | hober: it was defined one or two days after it was proposed |
| 22:31 | <tantek> | annevk - are there URLs to specific versions of stuff on W3C hg? |
| 22:31 | <tantek> | e.g. on the W3C wiki, I can link to specific version of a wiki page |
| 22:32 | <annevk> | just replace tip by an id |
| 22:33 | <annevk> | but I'd recommend against doing that |
| 22:35 | <Velmont> | tipitappity |
| 22:35 | <hober> | annevk: oh, great |
| 22:35 | hober | goes to file that bug |
| 22:35 | <annevk> | hober: might want to check first |
| 22:37 | <annevk> | hober: I thought there was a bug, but per my records there isn't |
| 22:37 | <annevk> | guess just Gecko has one |
| 22:38 | <annevk> | there is this http://trac.webkit.org/changeset/111028 but that does not appear to be about the CSS changes |
| 22:38 | <zewt> | hooooly shit |
| 22:38 | <zewt> | i just managed to get ff to open a file picker loop from JS, which nearly made me have to reboot my machine, since every file picker stole focus, preventing me from killing FF |
| 22:38 | <zewt> | wonder if that's fixed; i'm a bit out of date |
| 22:38 | <annevk> | -> sleepytime |
| 22:38 | <annevk> | nn |
| 22:39 | <zewt> | later |
| 23:30 | <shepazu> | Hixie: okay if I push the button to make you chair of WHATCG? right now, it's vacant |
| 23:34 | <Velmont> | Do we have to call it WHATCG? :( |
| 23:35 | <TabAtkins_> | shepazu: That was the intention all along, so yes. |
| 23:35 | <shepazu> | Velmont: it's still WHATWG, the WHATCG is just a particular mechanism |
| 23:36 | <shepazu> | TabAtkins_: ok, thanks, done |
| 23:36 | <Velmont> | shepazu :-) |
| 23:36 | <Hixie> | shepazu: what difference does it make? |
| 23:36 | shepazu | gets nervous when a group doesn't have a chair, because any yahoo can make themself chair |
| 23:37 | <shepazu> | Hixie: honestly, I'm not sure :) |
| 23:37 | <Hixie> | i wouldn't worry about it then :-) |
| 23:37 | <shepazu> | I think the chair does have some oversight abilities |
| 23:38 | <Hixie> | well someone made me a chair |
| 23:38 | <shepazu> | Hixie: too late :) you can always change the chair if you want |
| 23:38 | <Hixie> | ideally i'd like it to not have a chair since the whatwg doesn't have a chair |
| 23:38 | Hixie | shrugs |
| 23:39 | <Hixie> | i suppose i could make everyone a chair |
| 23:39 | <shepazu> | uh |
| 23:39 | <shepazu> | yeah, you could do that |
| 23:39 | <shepazu> | but is there anyone who doesn't think you're the chair of the WHATWG? |
| 23:39 | <zewt> | is that like asking to see the manager at best buy? |
| 23:39 | <zewt> | surprise, everyone's a manager |
| 23:40 | <Hixie> | the whatwg doesn't even have the concept of "chairman" |
| 23:40 | <shepazu> | I guess it doesn't need one, everyone knows you're in charge :) |
| 23:40 | <wilhelm> | Chairman Meow? |
| 23:40 | <Hixie> | there's just members and contributors, according to the charter, and in practice there's also editor and spokeperson |
| 23:41 | <Hixie> | editors, i should say |
| 23:45 | <zewt> | if anyone's bored and feels like filing a bug, http://zewt.org/~glenn/don%27t%20click%20this.html clicking the file input box in windows endlessly opens file pickers (which kills firefox and was very hard to get out of without rebooting, since each file picker steals focus) |
| 23:46 | <zewt> | (don't have the energy to deal with firefox's tracker these days) |