| 08:11 | <MikeSmith> | in the node implementation of console object, console.log(args) outputs to standard out, whereas console.error(args) outputs to standard error |
| 08:12 | <MikeSmith> | Firebug console and console in other browsers has not distinction between standard out and standard error, right? |
| 08:12 | <MikeSmith> | *has no distinction |
| 08:12 | <hsivonen> | annevk: I suggest migrating your HTML XML Use Case 05 to the chair-designated location |
| 08:14 | <annevk> | we got a Chinese comment |
| 08:14 | <Rik`> | MikeSmith: you can filter output in Firebug, WebKit and Firefox 4 console |
| 08:15 | <annevk> | hsivonen, sure |
| 08:15 | <MikeSmith> | Rik`: filter? |
| 08:15 | <MikeSmith> | how so? |
| 08:15 | <MikeSmith> | like shell redirection? |
| 08:15 | <Rik`> | what shell in Firebug ? |
| 08:16 | <Rik`> | In the console panel in Firebug, you can choose to see all messages, only errors, only warnings or only infos |
| 08:16 | <MikeSmith> | ah |
| 08:17 | <MikeSmith> | OK |
| 08:17 | <MikeSmith> | but that's different than stdout vs. stderr, right? I mean in the case of node, its console seems to have two separate output streams, one for stdout and one for stderr |
| 08:18 | <MikeSmith> | just like a unix shell |
| 08:18 | <MikeSmith> | according to the node docs, at least |
| 08:18 | <MikeSmith> | I don't know what good that does you if you can't redict it though |
| 08:18 | <MikeSmith> | *redirect |
| 08:19 | <Rik`> | well, in browsers, it's a GUI so not the same concept |
| 08:20 | <annevk> | hsivonen, done |
| 08:22 | <MikeSmith> | d'oh |
| 08:22 | <MikeSmith> | I realize now this only makes sense when using node non-interactively |
| 08:22 | <MikeSmith> | the node shell itself doesn't have any stdout or stderr |
| 08:22 | <MikeSmith> | of course |
| 08:23 | <hsivonen> | annevk: thanks |
| 08:29 | <hsivonen> | hmm. vast X-wing conspiracy is still only in the queue at http://snowclones.org/the-queue/ |
| 08:30 | <hsivonen> | I wonder if X=browser will make it to the list |
| 09:00 | <jgraham> | Argh, someone just tricked me into opening one of the pointless public-html-testsuite emails with an xml dump of some test results by just putting a question mark in the title |
| 09:22 | <annevk> | anything I should not miss in this week's summary? |
| 09:23 | <hsivonen> | annevk: don't miss foolip's CCP for the acknowledgement thing |
| 09:24 | <foolip> | annevk, I wrote 4 CP this weekend, don't know which might be of general interest |
| 09:25 | <jgraham> | annevk: You are talking about whatwg only, or W3C also? |
| 09:27 | <annevk> | I am willing to try to include some W3C stuff |
| 09:27 | <MikeSmith> | two sets of slides from Silvia: |
| 09:27 | <MikeSmith> | http://blog.gingertech.net/wp-content/uploads/2011/01/ |
| 09:28 | <annevk> | what hsivonen and foolip suggest make sense and I was considering Adam Barth's XSS stuff |
| 10:29 | <annevk> | took a little longer this time: http://blog.whatwg.org/another-whatwg-weekly |
| 10:29 | <annevk> | actually, not that much longer |
| 10:33 | <matjas> | ♥ WHATWG Weekly |
| 10:33 | <matjas> | kudos annevk |
| 10:35 | <kennyluck> | ♥ WHATWG Weekly too |
| 10:45 | <foolip> | annevk, great write-up! |
| 10:50 | <annevk> | doing it for the free Internet karma :) |
| 10:59 | <annevk> | might need a different twitter plugin for WordPress |
| 10:59 | <annevk> | I would love one that only uses short URLs when absolutely required |
| 11:04 | <Rik`> | annevk++ |
| 11:32 | <hsivonen> | annevk: thanks for the summary |
| 11:33 | <hsivonen> | abarth's proposed changes to CSP don't look good to me |
| 11:33 | <hsivonen> | both changes would make the system itself more vulnerable against attacks |
| 11:33 | <hsivonen> | using <meta> instead of HTTP headers makes it easier to inject bogus policies |
| 11:34 | <hsivonen> | and using events instead of an independent reporting channel would allow the attacker to capture the policy violation reports |
| 11:36 | <jgraham> | hsivonen: The policy violation reports aren't that useful in themselves are they? |
| 11:36 | <annevk> | not sure either about HTML vs HTTP, but events seems a lot simpler than HTTP-based reporting |
| 11:36 | <annevk> | CSP is way too complex imo |
| 11:36 | <jgraham> | I mean, does anyone expect to be able to detect attacks based on the reports? The bits of the thread I read suggested not |
| 11:36 | <jgraham> | Rather they are for when you trip over your own policy |
| 11:37 | <jgraham> | Or at least the value of detecting a failed attack is small |
| 11:37 | <annevk> | I can not even post to public-html-a11y? |
| 11:37 | <annevk> | oh well |
| 11:37 | <hsivonen> | jgraham: dunno about usefulness. |
| 11:38 | <hsivonen> | but I'm worried that the policy violations become part of Google Analytics and site authors start feeling that blocking ga.js is some kind of preparation of an attack |
| 11:39 | <hsivonen> | (IMO, users should be able to block ga.js without loss of site functionality) |
| 11:39 | <hsivonen> | (It's pretty obvious that analytics package of your choice is an euphemism for ga.js) |
| 11:43 | <jgraham> | hsivonen: I'm not sure I follow. If you block ga.js, nothing gets reported, right? Assuming a javascript API. |
| 11:46 | <hsivonen> | jgraham: right. so now if site authors freak out over the prospect of not getting policy violation reports if ga.js is blocked, they might start checking if ga.js got blocked and doing something user-hostile when blocked |
| 11:47 | <jgraham> | hsivonen: I see. I have a hard time imagining they will be more freaked out than they are about not getting their precious analytics anyway. But it is a reasonable concern |
| 12:22 | <hsivonen> | annevk: Are you gonna write a counter-CP for ISSUE-125 if Julian doesn't drop it? |
| 12:23 | <annevk> | no |
| 12:23 | <annevk> | I will Formally Object |
| 12:24 | <annevk> | I don't think a change to the spec will have any meaningful impact anyway |
| 12:24 | <hsivonen> | I wish Julian didn't raise ISSUEs on stuff like this |
| 12:25 | <annevk> | I've come to realize that's like wishing for a pony |
| 12:26 | <jgraham> | s/pony/unicorn/ |
| 12:31 | <hsivonen> | annevk: I'm wondering if I should write a CCP. Such a waste of time, though. |
| 12:33 | <annevk> | I wonder if I could have written one instead of writing to this thread |
| 12:36 | <zcorpan> | foolip: i think in Details of http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType you need to add that Content-Type is explicitly not honored to override HTTP's requirements |
| 12:36 | <zcorpan> | foolip: hmm.. or maybe not |
| 12:38 | <foolip> | zcorpan, I'm not sure precisely what you mean, can you add the bullet point? (I assume you're right about whatever it is.) |
| 12:38 | <foolip> | what does "override HTTP's requirements" mean here? |
| 12:38 | <foolip> | the short version of details should be "make it work like <img>" |
| 12:39 | <zcorpan> | foolip: i was confused. it's fine |
| 12:39 | <zcorpan> | foolip: <img> also ignores error responses, but i assume <video> doesn't |
| 12:40 | <annevk> | it should suggest adding a note about that probably |
| 12:40 | <foolip> | right, that is different, didn't think of that |
| 12:40 | <annevk> | but Hixie will do that anyway |
| 12:40 | <annevk> | should <video> not also ignore error responses? |
| 12:40 | <zcorpan> | annevk: notes are editorial, let's leave editorial stuff to the editor :) |
| 12:40 | <foolip> | we actually need to keep the response code sane for <video>, since we rely on it to detect support for byte range requests |
| 12:41 | <annevk> | how else are you going to display a 404 that is a video? |
| 12:41 | <foolip> | annevk, you don't, or you use a 404 HTML document pointing to a byte-range-capable video |
| 12:43 | <annevk> | seems kind of weird for that to work differently for video |
| 12:43 | <annevk> | only EventSource does something special for 404 and such afaik and that's because it's useful API-wise |
| 12:44 | <Philip`> | Maybe people will want to detect <video> hotlinking and return a 403 but still return video data complaining to the users about the hotlinking |
| 12:44 | <foolip> | perhaps, but there is no problem with the (stable!) status quo here, and we depend on the status code being 200/206 to differentiate between servers that understand Ranges: byte=0- and not |
| 12:44 | <Philip`> | (like with <img> hotlinking, except I don't know if people bother with 403 in practice) |
| 12:45 | <zcorpan> | Philip`: you can return a 200 hotlink video |
| 12:46 | <foolip> | also, the video API differentiates between network errors and decoding errors, and while it's not awesomely useful, I see no reason to change it |
| 12:52 | <annevk> | If I would continue in that public-html thread (Sam indicates it is not desired) I would say that specifications without a thorough test suite cannot really be broken. They are broken themselves. |
| 12:56 | <Lachy> | interesting media queries showcase http://mediaqueri.es/ |
| 12:56 | <hsivonen> | annevk: still no plans to write a counter-CP? |
| 12:56 | <wilhelm> | Lachy: Oh, pretty. |
| 12:57 | <annevk> | hsivonen, if you are going to do it otherwise I will do it (assuming you want to write a no-CP) |
| 12:58 | <annevk> | hsivonen, I have wasted enough time already :) |
| 12:58 | <hsivonen> | annevk: OK. I'll do it. |
| 12:58 | <annevk> | hsivonen, actually, it would not be a no-CP but rather suggesting that we simply update the note |
| 12:59 | <annevk> | hsivonen, heh, I think you misunderstood; I will do it |
| 13:00 | <hsivonen> | annevk: oh, OK. Thanks. |
| 13:01 | <hsivonen> | annevk: I'll cancel my volunteering as soon as the email loops back to me so that I can reply to it |
| 13:03 | <Philip`> | Don't you have a copy in a Sent folder you can reply to? |
| 13:05 | <hsivonen> | Philip`: Hmm. I guess that would have worked, too. |
| 13:12 | <annevk> | I have not avoided process hell |
| 13:12 | <foolip> | hehe, even the Chinese guy thought that dropping the 5 was a bad idea :) http://blog.whatwg.org/html-is-the-new-html5#comment-42523 |
| 13:12 | <annevk> | Or rather, I am going to Process Hell |
| 13:12 | <annevk> | Suggesting a change to a Note, teehee |
| 13:13 | <annevk> | So much doomed |
| 13:13 | <foolip> | annevk, best of luck retaining your sanity |
| 13:18 | <zcorpan> | hmm, firefox doesn't like the hotlinking of the font in http://simon.html5.org/html5-elements |
| 13:18 | <zcorpan> | boo |
| 13:19 | <hsivonen> | zcorpan: would be less of a boo if CORS allowed more options that allowing one Origin or * |
| 13:20 | <zcorpan> | CORS doesn't allow multiple origins? |
| 13:21 | <zcorpan> | i guess you could modify the CORS headers based on the request, but that'd be annoying to set up |
| 13:21 | <karlcow> | http://firstpersoncookie.wordpress.com/2011/01/23/ |
| 13:21 | <hsivonen> | zcorpan: not anymore. which sucks rather badly if you are a random person with mod_headers instead of being Google |
| 13:21 | <zcorpan> | hsivonen: yeah. agree |
| 13:23 | <zcorpan> | i wonder how the font+cors thing plays out, it seems opera, safari and chrome allow the hotlinking |
| 13:24 | <zcorpan> | benschwarz: if you have ability and want to, you could add some CORS magic to the fonts on developers.whatwg.org |
| 13:24 | <hsivonen> | zcorpan: what about IE9? |
| 13:24 | <zcorpan> | don't have ie9 here, so dunno |
| 13:24 | <annevk> | browsers limiting font linking with CORS are silly |
| 13:25 | <benschwarz> | zcorpan: why would that be useful? |
| 13:25 | <hsivonen> | I serve the fonts on hsivonen.iki.fi with a CORS header that allows about.validator.nu. If I want to allowed another hostname, I'm out of luck |
| 13:25 | <hsivonen> | easier to use * |
| 13:25 | <hsivonen> | than to do something that breaks caching |
| 13:25 | <zcorpan> | benschwarz: to make http://simon.html5.org/html5-elements less ugly in firefox |
| 13:26 | <zcorpan> | benschwarz: i could copy the fonts to my server, but i don't care enough to bother :) |
| 13:26 | <benschwarz> | zcorpan: let me check something… |
| 13:27 | <zcorpan> | hsivonen: oh yeah, didn't even think about caching |
| 13:27 | <zcorpan> | hsivonen: what led to this decision? |
| 13:27 | <benschwarz> | hmm :/ |
| 13:27 | <benschwarz> | I thought that chunkfive might be a google font |
| 13:27 | <hsivonen> | zcorpan: not sure. It's annevk's spec. |
| 13:27 | <benschwarz> | damn . |
| 13:28 | <benschwarz> | zcorpan: do you have the entry required for the htaccess? |
| 13:28 | <benschwarz> | can you add it as a ticket, I'll do it in the morning |
| 13:28 | <benschwarz> | just about to crash. 1230am here |
| 13:28 | <zcorpan> | sure |
| 13:28 | <benschwarz> | cheers |
| 13:31 | <benschwarz> | zcorpan: if you know the .htaccess stuff I need, please include it |
| 13:32 | zcorpan | also notices that the font size is smaller in opera compared to other browsers |
| 13:32 | <zcorpan> | benschwarz: i don't know without looking it up |
| 13:32 | zcorpan | curses monospace fonts and font sizes |
| 13:50 | <annevk> | Hixie, I think we should include a link to twitter here http://www.whatwg.org/ |
| 13:51 | <annevk> | heh, more Chinese comments |
| 13:52 | <annevk> | "IE6 supports HTML" |
| 13:58 | <wilhelm> | I hope this guy is a troll: http://blog.whatwg.org/html-is-the-new-html5#comment-42468 |
| 13:58 | <wilhelm> | When clicking on any of the translations on his site, I get an XML parse error. |
| 14:01 | <annevk> | Best of the internet is collected in those comments :p |
| 14:02 | <matjas> | why no block level anchors on http://whatwg.org/? |
| 14:03 | <annevk> | that site was last redesigned in 2005/2006 or so |
| 14:03 | <matjas> | ah, that explains why http://forums.whatwg.org/ still says “HTML5 forum” |
| 14:05 | <annevk> | December 2006 that is |
| 14:05 | <annevk> | matjas, I think that is for marketing |
| 14:05 | <annevk> | zcorpan is in charge |
| 14:06 | <annevk> | the forums are still active |
| 14:06 | <annevk> | I am amazed |
| 14:12 | <Ms2ger> | So, just Opera left to support window.onerror? |
| 14:13 | <annevk> | I argued for that years ago |
| 14:13 | <Ms2ger> | So do it again :) |
| 14:13 | <annevk> | I'm old and tired now :p |
| 14:14 | <Ms2ger> | Hah |
| 14:14 | <Ms2ger> | Let gsnedders do it, he's still young |
| 14:40 | <gsnedders> | Ms2ger: Pff, that incurs effort. |
| 14:40 | <gsnedders> | Ms2ger: I'm a lazy teenager, remember? :P |
| 14:41 | <Ms2ger> | No, you're a nice teenager ;) |
| 14:49 | jgraham | isn't sure that "lazy" and "nice" and antonyms |
| 15:25 | <annevk> | hsivonen, http://www.w3.org/Bugs/Public/show_bug.cgi?id=10838 regarding <u> closes without prejudice this week |
| 15:25 | <annevk> | hsivonen, http://dev.w3.org/html5/status/issue-status.html#ISSUE-144 is a better link |
| 15:26 | <hsivonen> | annevk: thanks |
| 15:26 | <hsivonen> | kennyluck: are you going to write a CP by the deadline? |
| 15:27 | <hsivonen> | hah. when the table lists "Chairs" as "Responsible", "Deadline" is "n/a" |
| 15:29 | <annevk> | presumably their deadline is somewhere before we go to Last Call |
| 15:29 | <annevk> | because otherwise we miss that date and the sky will hit fall on the earth |
| 15:29 | <annevk> | hit fall, really |
| 15:53 | <annevk> | can someone explain to me how ISSUE-148 is different from ISSUE-125/126? |
| 15:58 | <hsivonen> | annevk: it's not about what the parser does, right? |
| 15:59 | <annevk> | as far as I can tell it is |
| 16:01 | <annevk> | it seems Julian raised the same issue again |
| 16:01 | <annevk> | and then later escalated his old one |
| 16:01 | <annevk> | o_O |
| 16:29 | <annevk> | hsivonen, btw, Chrome seems to execute the javascript no matter where <base> is located |
| 16:29 | <annevk> | hsivonen, I have not taken the time to investigate further yet |
| 16:30 | <annevk> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/789 was my testing thingie |
| 16:41 | <TabAtkins> | Hmm. Where do I report that the Questionnaire title is XSSable? |
| 16:47 | <Philip`> | It's not really XSS |
| 16:47 | <Philip`> | since the issue is just that trusted people can insert HTML markup onto the site via the title |
| 16:48 | <TabAtkins> | Do you trust them not to XSS you? |
| 16:48 | <Philip`> | and those people could probably attack the site in much easier ways, given that they're trusted |
| 16:48 | <TabAtkins> | Maciej could steal your cookies. |
| 16:48 | <Philip`> | They could just edit the HTML WG home page or whatever |
| 16:48 | <TabAtkins> | In any case, it makes the title for the <figure> questionnaire kind of meaningless. |
| 16:49 | <TabAtkins> | Which is the real issue, of course. |
| 16:49 | <Philip`> | Yeah, it's still a bug |
| 16:49 | <Philip`> | Sam said he tried to find a way around it but failed, I think |
| 16:49 | <TabAtkins> | You can't pre-escape things? |
| 16:50 | <TabAtkins> | Or double escape them? I have to double-escape markup for twitter. |
| 16:50 | <Philip`> | It'll appear double-escaped in some other parts of the page, probably |
| 16:57 | <TabAtkins> | That's true. So I just need to complain at the right person in systems. |
| 16:58 | <Philip`> | Or ignore it |
| 17:55 | <karlcow> | In the http://www.yahoo.com/ <meta http-equiv="X-UA-Compatible" content="chrome=1"> *sigh* |
| 17:55 | <TabAtkins> | Chrome Frame is a good thing, though hooking it off of X-UA-Compatible was still a mistake. |
| 18:04 | <ben_c> | Have they got chrome frame installing as 'non admin' yet? |
| 18:05 | <ben_c> | That's when it'll be super awesome |
| 18:05 | <TabAtkins> | No clue. |
| 18:07 | <karlcow> | TabAtkins: apart of the name which is plain branding ;) GFC tends to put people into the mode of thinking webkit only and that is no good for interOPERAbility |
| 18:08 | <Ms2ger> | I see what you did there |
| 18:16 | <kennyluck> | hsivonen, it seems to be that there are various levels of dealing with this rather no-technical issue. I'll start with a change proposal just to remove "ship name". I don't think this reflects you and othermaciej's point but this is likely to draw zero objection. |
| 18:16 | <othermaciej> | kennyluck: which issue? |
| 18:16 | <kennyluck> | othermaciej: getting <u> back. |
| 18:17 | <annevk> | remove "ship name"? |
| 18:17 | <othermaciej> | what does "ship name" have to do with that? |
| 18:18 | <kennyluck> | My argument was that <u> is for proper noun as <i> is for ship name. |
| 18:18 | <kennyluck> | So... as long as <i> can't be used for ship name, my argument no longer holds. |
| 18:19 | <annevk> | that seems rather silly |
| 18:19 | <annevk> | :) |
| 18:19 | <kennyluck> | (This is marginally related to the main issue, but I am not interested in getting back <u> either) |
| 18:19 | <kennyluck> | s/not/not that/ |
| 18:20 | <othermaciej> | I don't think a change proposal to change <i> would be likely to be considered in scope for the <u> issue |
| 18:21 | <kennyluck> | othermaciej: I thought about "Change Proposal: remove semantic tig", but I am not confident. |
| 18:23 | <kennyluck> | sigh... I guess I'll write a long one. |
| 18:25 | <karlcow> | rediscovering http://www.http-stats.com/ |
| 18:26 | <annevk> | karlcow, that looks hopelessly outdated and wrong |
| 18:27 | <karlcow> | I don't know if there are still running the service. |
| 18:27 | <annevk> | "Current statistics are based on a sample of 84604 probed servers, gathered in the last 958 days." |
| 18:27 | <karlcow> | yep seen that |
| 18:27 | <annevk> | Google should be in the top three |
| 18:29 | <karlcow> | and aws |
| 18:30 | <annevk> | http://en.wikipedia.org/wiki/Web_server#Market_structure |
| 18:30 | <karlcow> | I'm suprised to not see nginx in the top too |
| 18:31 | <karlcow> | oh it is |
| 18:31 | <karlcow> | doh |
| 18:31 | <karlcow> | vendor is the name of the dev |
| 19:19 | <annevk> | http://theoatmeal.com/comics/state_web_winter scroll to the end |
| 19:23 | <annevk> | HTML Cheetah Balls |
| 19:23 | <karlcow> | http://s3.amazonaws.com/theoatmeal-img/comics/state_web_winter/html.jpg |
| 19:23 | <annevk> | wfm |
| 19:24 | <karlcow> | annevk: be careful, we might be quoted by New-York Times :p |
| 19:24 | <annevk> | I'll believe it when it happens |
| 19:25 | <karlcow> | ;) me too |
| 19:43 | <annevk> | unread: 1337 |
| 19:43 | <annevk> | hah |
| 19:45 | <Hixie> | annevk: http://dev.w3.org/2006/webapi/progress/Progress.html needs a 301 |
| 19:46 | <annevk> | http://w3c-test.org/webapps/tests/XMLHttpRequest/status-basic.htm works :) |
| 19:50 | <annevk> | Hixie, I added http://dev.w3.org/cvsweb/2006/webapi/progress/.htaccess but it is not working |
| 19:50 | <annevk> | MikeSmith, can you help? |
| 19:51 | <annevk> | Hixie, http://dev.w3.org/2006/webapi/progress/ is the new link by the way |
| 19:51 | <Ms2ger> | I told him that ;) |
| 19:52 | <Hixie> | yeah, i fixed it |
| 19:52 | <Hixie> | but i got confused by the other one while fixing it :-) |
| 19:52 | <annevk> | in #secrettreehouse?! oooh |
| 19:52 | <Hixie> | Ms2ger's bugs are so terse that i have to work out what he means each time :-) |
| 19:52 | <annevk> | ah, minimalism is underrated :) |
| 19:52 | <Hixie> | and my attempt to force him to include more text always just results in him adding random spaces and words at the end :-P |
| 19:52 | Hixie | glares at Ms2ger :-P |
| 19:52 | <Ms2ger> | Hah |
| 19:53 | <Ms2ger> | Now you're linking to a 404 |
| 19:53 | <Hixie> | i am? |
| 19:53 | <Hixie> | damnit |
| 19:54 | <Ms2ger> | I should have said just "404", actually |
| 19:54 | <Hixie> | no i'm not |
| 19:54 | <Hixie> | oh, i didn't commit the full change |
| 19:54 | <Hixie> | my bad |
| 19:55 | <foolip_> | when will people stop slapping x- prefixes on new MIME types, HTTP headers, etc? |
| 19:55 | <annevk> | was my .htaccess entry wrong? |
| 19:55 | <annevk> | I was pretty sure you could actually set up redirects on that domain |
| 19:55 | <annevk> | or maybe they are moderated somehow before being put live? |
| 19:56 | <Ms2ger> | http://tools.ietf.org/html/draft-saintandre-xdash-considered-harmful-01 |
| 19:56 | <Hixie> | oh god, another references section bug from ms2ger |
| 19:56 | <Hixie> | the only thing that separates Ms2ger from Julian is that Ms2ger's bugs aren't bogus! |
| 19:57 | <annevk> | Ms2ger also closes bogus bugs with witty/non-appropriate-but-funny comments |
| 19:57 | <annevk> | I found out the other day |
| 19:57 | <Hixie> | hah |
| 19:57 | <foolip_> | http://donottrack.us/server.html adds to the fun with X-Do-Not-Track |
| 19:57 | <Ms2ger> | Section: http://www.whatwg.org/specs/web-apps/current-work/#refsORIGIN |
| 19:57 | <Ms2ger> | What else do you need? |
| 19:57 | <Hixie> | man you think i READ the boilerplate? :-P |
| 19:57 | <Ms2ger> | You think I'm going to write it out again? |
| 19:58 | <micheil> | hmm.. I hate being bad with names; who's the guy who's the envangelist at mozilla with the red hair; chris someone.. |
| 19:59 | <Ms2ger> | cheilmann? |
| 19:59 | <foolip_> | Hixie, abarth, which of you should I talk to about standardizing audio/video type sniffing in http://tools.ietf.org/html/draft-ietf-websec-mime-sniff-00 ? |
| 19:59 | <Hixie> | abarth first |
| 19:59 | <Hixie> | me if he doesn't have time/inclination to do it |
| 19:59 | <annevk> | should also include font sniffing I think |
| 19:59 | <foolip_> | abarth, ping me if you read this later |
| 19:59 | <foolip_> | (pretty please, etc) |
| 20:00 | <annevk> | but I said that already |
| 20:00 | <hsivonen> | What are the critical consumer protection mechanisms that this post talks about: http://www.thesecuritypractice.com/the_security_practice/2011/01/mozillas-do-not-track-proposal-must-first-do-no-harm.html ? |
| 20:00 | <Ms2ger> | Damn you, Anne, with your WD-style on dev.w3.org! |
| 20:01 | <micheil> | Ms2ger: yah, that's him |
| 20:01 | <annevk> | Ms2ger, glad you care |
| 20:02 | <annevk> | once there's two I might act |
| 20:02 | <Ms2ger> | Hah |
| 20:02 | <Ms2ger> | Julian? |
| 20:02 | annevk | raises lower bar to three |
| 20:03 | <Ms2ger> | Can't think of anyone else :( |
| 20:05 | <annevk> | but euh, fix pimpmyspec.net ! |
| 20:05 | <Ms2ger> | Hixie, surely my bugs are easier than that foreign content bug right at the top of your queue |
| 20:07 | <annevk> | Hixie, I would like twitter.com/WHATWG on whatwg.org |
| 20:07 | <annevk> | Hixie, if you do not want to sacrifice space I suggest we dump "Demos" for now |
| 20:08 | <Ms2ger> | There's demos? |
| 20:09 | <Ms2ger> | So I was writing tests, and I just ended up on IRC |
| 20:11 | <annevk> | Ms2ger, http://www.whatwg.org/demos/ |
| 20:19 | <jgraham> | gsnedders: Saw this, thought of you http://www.youtube.com/watch?v=4N1M7Kwl81A |
| 20:20 | <gsnedders> | jgraham: That is awesome. |
| 20:24 | <jgraham> | gsnedders: Yes |
| 20:40 | <Hixie> | Ms2ger: i'm not actually complaining about your bugs, just teasing ;-) |
| 20:40 | <Hixie> | Ms2ger: they are indeed easy to deal with :-) |
| 20:40 | <Hixie> | Ms2ger, annevk: btw i fully intend to eventually defer to dom core on the infrastructure definitions, i'm just stalling because those are really core to the spec and so i want to make sure we have a really solid story for dom core and so on first (i do change the definitions occasionally, and it has ramifications across a lot of the platform when i do) |
| 20:41 | <Hixie> | annevk: i'm happy to add a new one, sure. hold on. |
| 20:41 | <Hixie> | annevk: any preferred colour? |
| 20:41 | <Ms2ger> | papayawhip |
| 20:41 | <Hixie> | what color is that |
| 20:41 | <Ms2ger> | Papaya whip is a representation of the color that would result if mashed papayas were blended with vanilla ice cream, whipped cream, or yogurt. |
| 20:41 | <jgraham> | Ugly |
| 20:41 | <Ms2ger> | Yay, Wikipedia |
| 20:42 | <Hixie> | i was looking more for an rgb triple |
| 20:42 | <Hixie> | but thanks? |
| 20:43 | <Ms2ger> | Hey, it's in CSS3Color! |
| 20:43 | <Ms2ger> | (ffefd5) |
| 20:44 | <Hixie> | dude that's uglier than the gray we had before |
| 20:44 | jgraham | did say that already :p |
| 20:44 | <Ms2ger> | Hah |
| 20:44 | <TabAtkins> | papayawhip is the best color |
| 20:44 | <Peter`> | But it'd be made from mashed papayas |
| 20:44 | <Ms2ger> | I'd raise a Tracker issue for it ;) |
| 20:45 | <jgraham> | TabAtkins: I think you meant to say "papayawhip is the best color /you can hope for if you blend papaya with vanilla ice cream, whipped cream, or yoghurt" |
| 20:46 | <jgraham> | s|"|/"| |
| 20:46 | <Hixie> | ok reload whatwg.org/ |
| 20:47 | <Ms2ger> | wfm |
| 20:48 | <Hixie> | i also put a max-width on the page and centered it |
| 20:51 | <annevk> | cool |
| 20:51 | <Hixie> | gonna remove everything except those boxes in fact |
| 20:53 | <franksalim> | including the "Want to get involved and help out?" |
| 20:54 | <Hixie> | yeah, gonna replace that with a box |
| 20:54 | <annevk> | is my screen just not wide enough for three columns? |
| 20:54 | <annevk> | or are there no three columns possible? |
| 20:54 | <annevk> | oh, now it is |
| 20:54 | <annevk> | hmmmm |
| 20:55 | <annevk> | Hixie, I think matjas_ suggested using wrapper <a> to make the whole box clickable |
| 20:56 | <annevk> | Hixie, though that would not work for specs |
| 20:56 | <matjas> | +1 |
| 20:56 | <Hixie> | i just simplified specs so it would work for it |
| 20:56 | <Ms2ger> | Wait, are you +1'ing yourself? |
| 20:56 | <Hixie> | wouldn't work for Help though |
| 20:57 | <annevk> | Help seems somewhat redundant with "Mailing List" and "Forums" and "IRC" |
| 20:57 | <annevk> | also, these boxes would do great with transform on :hover :) |
| 20:59 | <annevk> | oh, and since we're beyond HTML4, you can drop media="all" |
| 21:10 | <annevk> | Hixie, http://www.whatwg.org/mailing-list starts with "we have three lists" and goes on to list four |
| 21:10 | <Hixie> | heh |
| 21:10 | <Hixie> | fixed |
| 21:22 | <abarth> | foolip_: hey |
| 21:23 | <Hixie> | does box-shadow not supported being css-transitioned? |
| 21:23 | <foolip_> | abarth, hi there |
| 21:23 | <TabAtkins> | According to the current draft, no. But I suspect that's an oversight. |
| 21:24 | <abarth> | did you want to talk about video sniffing? |
| 21:24 | <foolip_> | abarth, so basically I want to kill Content-Type for <video>, as per http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType |
| 21:24 | <annevk> | Hixie, not per http://www.w3.org/TR/css3-transitions/ |
| 21:24 | <TabAtkins> | This *should* be defined in css3-backgrounds, but that spec doesn't have a proper section on Interpolating yet. |
| 21:25 | <foolip_> | and if that succeeds, we really need interop on the sniffing |
| 21:25 | <foolip_> | so I just want to know how we could go about doing that |
| 21:25 | <abarth> | foolip_: video sniffing is a big mess. like other kinds of sniffing, it's race to the bottom, but unlike other kinds of sniffing, it hasn't stablished yet |
| 21:25 | <foolip_> | I can come up with the byte patterns for the formats we support, but have no spec editing powers |
| 21:26 | <foolip_> | between the formats we support, it's actually rather simple |
| 21:26 | <abarth> | yeah? can you send me a list of the formats / byte patterns? |
| 21:27 | <foolip_> | abarth, sure, I summarized the state of things on in a mail some time ago, let me look for it |
| 21:27 | <abarth> | k |
| 21:27 | <abarth> | ignoring the content-type for <video> would align its behavior with <img> |
| 21:28 | <foolip_> | abarth, http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028383.html talks about OggS |
| 21:29 | <Hixie> | annevk: lame |
| 21:29 | <foolip_> | http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-September/028371.html talks about WebM |
| 21:30 | <foolip_> | and I see I've neglected WAVE, but the format is "RIFF" + any 4 octets + "WAVE" |
| 21:30 | <abarth> | how future-proof is OggS\0 ? |
| 21:31 | <foolip_> | abarth, you mean if the \0 could become 0x01 ? |
| 21:31 | <abarth> | "Opera (GStreamer) simply checks for the string "webm" somewhere in the header" |
| 21:31 | <abarth> | boo |
| 21:31 | <abarth> | foolip_: yes |
| 21:31 | <Hixie> | "RIFF" + any 4 octets + "WAVE" -- finally another use for "mask" part of the sniffing table :-) |
| 21:31 | <abarth> | yeah, we already need mask for WebP |
| 21:31 | <abarth> | because it's also RIFF |
| 21:31 | <Hixie> | annevk: what do you think of whatwg.org now |
| 21:32 | <Hixie> | matjas_: i implemented your suggestion on whatwg.org |
| 21:32 | <Hixie> | matjas_: so thanks! |
| 21:32 | <foolip_> | abarth, when I looked into it I came to the conclusion that it's unlikely that the version field is going to increase ever, and it it does we need to update both demuxers and sniffing |
| 21:32 | <matjas_> | thank you! |
| 21:32 | <abarth> | foolip_: ok, then we should include the null byte |
| 21:32 | <foolip_> | abarth, WebM sniffing is the one that requires the most tinkering, I think |
| 21:32 | <annevk> | Hixie, a lot better :) |
| 21:32 | <annevk> | nice |
| 21:32 | <abarth> | i asked the vp8 folks about that at some point |
| 21:33 | <foolip_> | no conclusion? |
| 21:33 | <abarth> | they said there's a constant pattern at a constant offset that we can use |
| 21:33 | <abarth> | i can ask them again |
| 21:33 | <abarth> | what about h.264? |
| 21:33 | <Hixie> | holy cow, it even works in IE, Opera, Firefox, Chrome, and Safari, without my having to do anything special |
| 21:33 | <Hixie> | i only have two non-standard properties and they don't work anyway (the transitions for box-shadow) |
| 21:33 | <foolip_> | abarth, that would be great, on the WebM discussion list preferably, since I'm on it |
| 21:33 | <Hixie> | ok, lunch time |
| 21:33 | <Hixie> | bbiab |
| 21:34 | <abarth> | Hixie: pretty |
| 21:34 | <foolip_> | abarth, I haven't looked at sniffing the MPEG-4 container, because we don't intend to support it |
| 21:34 | <foolip_> | abarth, to get it right, it'd probably be best to ask the QuickTime folks at Apple |
| 21:35 | <annevk> | Hixie, so next is redesigning http://www.whatwg.org/charter and such? :) |
| 21:35 | <abarth> | ok, i'd like to add them all at once to avoid the appearance of favoritism |
| 21:35 | foolip_ | looks at how GStreamer sniffs MPEG-4 |
| 21:36 | <abarth> | thx |
| 21:37 | <foolip_> | oh, maybe we need MP3 too, unless Chrome is going to remove support for it |
| 21:37 | <foolip_> | ID3v2... |
| 21:38 | <foolip_> | MP3 in RIFF container |
| 21:38 | <foolip_> | :/ |
| 21:38 | <abarth> | i seem to remember that mp3 was tricky to sniff, but i don't remember the details |
| 21:39 | <abarth> | is it just RIFF? |
| 21:39 | <Hixie> | annevk: i'll look, but probably not (unless i broke them) |
| 21:39 | <Hixie> | bbl |
| 21:40 | <foolip_> | abarth, you also have header-less MP3, IIRC |
| 21:40 | <foolip_> | which should actually be the most common |
| 21:41 | <roc> | Is Chrome going to remove MP3 support? That's never been clear |
| 21:41 | <TabAtkins> | I don't think we've ever announced anything in that direction. |
| 21:41 | <abarth> | dunno |
| 21:41 | <roc> | you've announced "Specifically, we are supporting the WebM (VP8) and Theora video codecs, and will consider adding support for other high-quality open codecs in the future." |
| 21:42 | <roc> | oh, that's video |
| 21:42 | <abarth> | do other browsers support mp3? |
| 21:42 | <TabAtkins> | Yeah, that's for video. We haven't said anything about <audio>. |
| 21:42 | <roc> | so technically you could still be supporting MP4+AAC! |
| 21:42 | <TabAtkins> | abarth: Yes, safari and IE9 do. |
| 21:42 | <TabAtkins> | http://html5doctor.com/native-audio-in-the-browser/ |
| 21:42 | <roc> | maybe you should shake some trees and get that clarified |
| 21:42 | <roc> | MP3-related patents haven't run out yet |
| 21:43 | <roc> | we'll support it the day they do |
| 21:43 | <abarth> | chrome doesn't support wav? |
| 21:43 | <TabAtkins> | Any idea when that day is? |
| 21:43 | <roc> | no |
| 21:43 | <roc> | I believe it depends on your legal department |
| 21:43 | <TabAtkins> | Bleh. Fuck patents. |
| 21:43 | <roc> | yep |
| 21:44 | <abarth> | at least everyone supports PNG |
| 21:44 | <abarth> | we should just use PNG for video |
| 21:44 | <roc> | APNG |
| 21:44 | <roc> | does chrome even support APNG? |
| 21:44 | <roc> | sigh |
| 21:45 | <roc> | Today was looking pretty good until I found out that Matrix 4 and 5 are going to be made |
| 21:45 | <foolip_> | abarth, for MPEG-4, it looks like there are quite a few patterns to match |
| 21:45 | <foolip_> | 8 byte in, you have "ftypqt ", "ftypisom" and "ftypmp42" |
| 21:46 | <foolip_> | it would probably be best to contact David Singer, as suggested by http://tools.ietf.org/html/rfc4337 |
| 21:46 | <abarth> | ok, i've been ignoring this stuff hoping the world would get simpler, but that looks somewhat unlikely at this point, i'll put something in the spec |
| 21:46 | <abarth> | ok, i know david |
| 21:47 | <foolip_> | abarth, do you need anything from me, or will everything I need magically appear in the spec after some time? ;) |
| 21:47 | <abarth> | i'm making some notes |
| 21:48 | <abarth> | give me a minute to see if I have any questions |
| 21:48 | foolip_ | nods |
| 21:49 | <karlcow> | this is amazing. Errata publication process with diff nodes :) http://scraplab.net/2011/01/20/admiralty-chart-correction-tracings/ |
| 21:50 | <abarth> | foolip_: OggS\0 is for video or audio? |
| 21:50 | <foolip_> | abarth, both, if you need to map it to an existing MIME type then application/ogg would do |
| 21:51 | <abarth> | yeah, we map the bytes to a mime type |
| 21:51 | <abarth> | ok |
| 21:51 | <foolip_> | figuring out what exactly the streams are isn't sniffable without quite a lot of data |
| 21:52 | <abarth> | k |
| 21:52 | <foolip_> | the WebM stuff is equally for audio-only files as well, so if video/webm is ok for that, then video/ogg shouldn't be a disaster for Ogg either |
| 21:53 | <abarth> | here's what i've got: |
| 21:53 | <abarth> | application/ogg: OggS\0 |
| 21:53 | <abarth> | video/webm: 0x1A, 0x45, 0xDF, 0xA3 + more (ask) |
| 21:53 | <abarth> | image/webp: "RIFF" + any 4 octets + "WEBPVP" |
| 21:54 | <abarth> | audio/wave: "RIFF" + any 4 octets + "WAVE" |
| 21:54 | <abarth> | audio/mpeg: ask |
| 21:54 | <abarth> | video/mp4: ask |
| 21:55 | <foolip_> | abarth, looks right |
| 21:55 | <abarth> | ok, i'll update the draft today |
| 21:55 | <foolip_> | although I think there's no registered MIME type for RIFF WAVE, if that matters |
| 21:55 | <abarth> | i'll also send email to the IETF websec list and the whatwg list |
| 21:55 | <abarth> | foolip_: we can register it |
| 21:56 | <foolip_> | abarth, that'd be awesome |
| 21:57 | <foolip_> | thanks for helping out, hopefully sanity will soon be possible for <video> type checking |
| 21:58 | <abarth> | foolip_: np. i'll also get chrome's sniffing updated to match these patterns |
| 21:59 | <foolip_> | abarth, great, I'll put it in Opera as soon as http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType prevails |
| 21:59 | <foolip_> | (or sooner) |
| 22:04 | <foolip_> | http://www.w3.org/html/wg/wiki/ChangeProposals/PosterElement gets 0 points for brevity |
| 22:19 | <annevk> | so I just looked at http://www.whatwg.org/style/tabbed-pages |
| 22:19 | <annevk> | the way Hixie picks colors is funny |
| 22:23 | <foolip_> | annevk, is the script that generates http://html5.org/tools/web-apps-tracker under your control? |
| 22:24 | <annevk> | yep |
| 22:24 | <annevk> | it's open source too |
| 22:24 | <foolip_> | icanhaz? |
| 22:24 | <annevk> | http://code.google.com/p/html5/source/browse/#svn%2Ftrunk%2Fweb-apps-tracker |
| 22:26 | <foolip_> | annevk, thanks |
| 22:26 | <annevk> | it's not really the greatest code ever written, but it works |
| 22:27 | <foolip_> | I was working on something to annotate commits with which sections were changed, and thought maybe it'd make sense to merge it |
| 22:27 | <foolip_> | not finished, might never be |
| 22:27 | <annevk> | that'd be quite brilliant |
| 22:28 | <foolip_> | annevk, do you have a cron job to update from upstream svn all the time, or do you have commit hooks on the svn server? |
| 22:28 | <annevk> | it's live :) |
| 22:28 | <foolip_> | oh, it logs from the server every time you access the page? |
| 22:28 | <annevk> | front page always pulls from SVN |
| 22:28 | <foolip_> | wow :) |
| 22:28 | <annevk> | diffs pull from SVN unless there is a cached entry |
| 22:29 | <MikeSmith> | anybody know whatever happened with persistent SharedWorkers idea? |
| 22:30 | <foolip_> | I've been using a git mirror (http://gitorious.org/whatwg/webapps/commits/master), but mostly because I like git of course |
| 22:30 | <annevk> | I have a commit hook these days (for twitter) and could therefore do smarter things, but nothing has been done so far |
| 22:30 | <annevk> | and perf-wise it all pulls through |
| 22:30 | <annevk> | the only feature I have been thinking about adding is an Atom feed that includes the diffs |
| 22:30 | <foolip_> | automatic read-only git mirrors like http://git.apache.org/ would be sweet, but I don't know how to set that up |
| 22:31 | <annevk> | but since that would require some kind of database or imagined file storage I have not bothered yet |
| 22:31 | <foolip_> | why? |
| 22:32 | <foolip_> | my thinking was to use <article> for each commit and then use http://www.whatwg.org/specs/web-apps/current-work/multipage/links.html#atom |
| 22:32 | <foolip_> | does anyone remember why it was removed from the W3C spec? |
| 22:32 | <foolip_> | people who like Atom don't like pseudo-code? |
| 22:32 | <annevk> | W3C does not use HTML5 |
| 22:33 | <TabAtkins> | Julain complained that it wasn't ideal in some cases. |
| 22:33 | <annevk> | oh, Atom |
| 22:33 | <annevk> | dunno |
| 22:33 | <foolip_> | I implemented it, seems pretty good actually |
| 22:33 | <Hixie> | annevk: what's funny about it :-P |
| 22:33 | <foolip_> | with some bugs filed, of course |
| 22:33 | foolip_ | looks at Hixie |
| 22:34 | <annevk> | foolip_, if you generate the Atom feed you need to have access to the previous entries no? |
| 22:34 | <annevk> | foolip_, hmm, I guess I could actually use XML to parse the instance, remove the last item and insert a new one |
| 22:34 | <annevk> | foolip_, but that would be rather professional |
| 22:34 | <foolip_> | annevk, you'd have to have an HTML document with as many top-level <articles> as you want <entry>s |
| 22:35 | <foolip_> | annevk, which is pretty close to what http://html5.org/tools/web-apps-tracker already is |
| 22:35 | <abarth> | foolip_: done |
| 22:35 | <foolip_> | abarth, ah, you're fast! |
| 22:36 | <abarth> | i try :) |
| 22:36 | <abarth> | i expect hatemail though :( |
| 22:36 | foolip_ | joins websec⊙io |
| 22:36 | <foolip_> | ignore the haters |
| 22:36 | <foolip_> | I can give you a bit of praise to balance it out :) |
| 22:36 | <annevk> | foolip_, in a way, except besides the commit message and bug link you also want to have the change |
| 22:37 | <annevk> | foolip_, anyway, I'm going to sleep and will see your code in the morning :p |
| 22:37 | <foolip_> | annevk, hrm, right |
| 22:37 | <foolip_> | annevk, don't count on it, there are many pieces and none of them are more than 50% done |
| 22:41 | <foolip_> | abarth, typo: vidow/webm |
| 22:41 | <abarth> | that's what i get :) |
| 22:42 | <Hixie> | annevk: decided not to change the other pages, they seem to work fine |
| 22:42 | <abarth> | foolip_: fixed |
| 22:42 | <foolip_> | in case you don't know, the part of the WebM signature you have is the EBML signature, shared by matroska and webm |
| 22:43 | <annevk> | Hixie, fair enough |
| 22:43 | <Hixie> | annevk: i might change the tabs one day, but probably not any time soon |
| 22:43 | <Hixie> | not like anyone looks at those pages anyway :-P |
| 22:44 | <annevk> | ooh, we have http://www.whatwg.org/images/logo.svg |
| 22:45 | <annevk> | I should start using that on the blog |
| 22:45 | <annevk> | the current image looks ugly scaled down |
| 22:45 | <Hixie> | yeah |
| 22:45 | <Hixie> | btw we also have images.whatwg.org that i use to speed up downloading the spec; if you need any other images on there let me know |
| 22:45 | <annevk> | I'm gonna use ::after { content:svg } :) |
| 22:45 | <Hixie> | content:svg? you mean content: url(...svg) ? |
| 22:46 | <annevk> | yeah |
| 22:46 | <Hixie> | ah ok |
| 22:46 | <Hixie> | i thought it might be some newfangled thing |
| 22:46 | <annevk> | Hixie, you could put that logo.svg there maybe? |
| 22:46 | <annevk> | that way whatwg.org does not need to be fetched for blog.whatwg.org |
| 22:46 | <Hixie> | sure |
| 22:46 | <annevk> | or I could just host a copy on blog.whatwg.org |
| 22:46 | <annevk> | I guess |
| 22:46 | <TabAtkins> | Yup, new spec. You give it a mimetype, it creates an appropriate image for you based on a pseudorandom algorithm. |
| 22:47 | <Hixie> | actually it's already there |
| 22:47 | <Hixie> | http://images.whatwg.org/logo.svg |
| 22:47 | <annevk> | sweet |
| 22:51 | <Hixie> | switched whatwg.org/ to use the svg too |
| 22:52 | <Hixie> | and to get it from images.whatwg.org |
| 22:53 | <MikeSmith> | gsnedders: so will Opera also be removing callable RegExp support? |
| 22:54 | <MikeSmith> | Hixie: did anything ever happen with the persistent SharedWorkers idea? |
| 22:54 | annevk | makes a mess of things |
| 22:54 | <gsnedders> | MikeSmith: Yeah, the intention is for everyone to do so |
| 22:54 | <Hixie> | MikeSmith: not that i know of |
| 22:54 | <Hixie> | MikeSmith: bigger fish is cross-domain shared workers |
| 22:55 | <MikeSmith> | Hixie: I had thought the Chrome team had been interested in doing something with it, particularly in the context of Notifications |
| 22:55 | <MikeSmith> | gsnedders: cool |
| 22:55 | <Hixie> | MikeSmith: oh, maybe. ask anne or abarth :-) |
| 22:55 | <Hixie> | or dglazkov |
| 22:56 | <Hixie> | or TabAtkins |
| 22:56 | <MikeSmith> | hai |
| 22:56 | <dglazkov> | whatwho? |
| 22:56 | <dglazkov> | SharedWorkers? |
| 22:56 | <annevk> | I wonder why this does not work... |
| 22:57 | <annevk> | height:1em should do the fricking trick |
| 22:57 | <annevk> | is it because the image has an intrinsic size? |
| 22:59 | <annevk> | unless someone figures this out I'll restore the mighty <img> |
| 22:59 | <TabAtkins> | annevk: It *should* work, as content: <image> should make it a replaced element. |
| 22:59 | <TabAtkins> | But shrug if that actually happens. |
| 22:59 | <Hixie> | annevk: url? |
| 23:00 | <annevk> | I just killed it |
| 23:00 | atwilson | heard there was a SharedWorker discussion going on... |
| 23:01 | <foolip_> | lazy #whatwg: do you still have to use xlink:href with SVG in HTML? |
| 23:02 | <foolip_> | or can one write just href and have the parser put it in the right namespace? |
| 23:02 | <dglazkov> | MikeSmith: ask atwilson |
| 23:02 | <MikeSmith> | k |
| 23:02 | <MikeSmith> | oh, he's here |
| 23:02 | <MikeSmith> | hey Drew |
| 23:02 | <atwilson> | hi |
| 23:02 | <annevk> | Hixie, TabAtkins, http://software.hixie.ch/utilities/js/live-dom-viewer/saved/790 |
| 23:03 | <MikeSmith> | atwilson: was wondering if you all ever ending up pursuing the persistent SharedWorkers idea any further |
| 23:03 | <annevk> | Hixie, TabAtkins, works fine if you make it an <img> as seen on blog.whatwg.org |
| 23:03 | <TabAtkins> | Yeah, I understand the markup. I'm saying that it *should* work, but 'content' support is all dumb. |
| 23:03 | <MikeSmith> | atwilson: seems like the discussion on the list just died down at some point |
| 23:03 | <Hixie> | foolip_: iirc you have to use the attribute name "xlink:href" but don't have to declare xlink |
| 23:04 | <TabAtkins> | I guess it's instead treating it as an inline element containing an image, rather than an image itself. |
| 23:04 | <atwilson> | MikeSmith: somewhat. There was significant pushback about exposing this as part of the general web API. |
| 23:04 | <foolip_> | Hixie, an improvement, but just href would be great :) |
| 23:04 | <foolip_> | (but confusing at the same time) |
| 23:04 | <atwilson> | MikeSmith: We're exposing similar functionality via Chrome extensions for now, and instead of basing this on SharedWorkers we're allowing "background pages". |
| 23:04 | <Hixie> | i expect the svgwg will get there and then it'll just work in html too |
| 23:04 | <annevk> | TabAtkins, yeah, probably |
| 23:05 | <MikeSmith> | atwilson: I see |
| 23:05 | <atwilson> | MikeSmith: Turns out that workers have various problems (can't do redirects through different domains for auth, etc) that pages don't have. |
| 23:05 | <MikeSmith> | ah |
| 23:05 | <Hixie> | foolip_: i didn't want to start fiddling with attribute namespaces and later find this conflicted with thes svgwg removing xlink... |
| 23:05 | <david_carlisle> | foolip_: mathml3 switched to href instead of xlink: maybe you should svg folks to do same |
| 23:05 | <annevk> | TabAtkins, makes sense for content:url(blahdieblah) "test"; but not here... |
| 23:05 | <david_carlisle> | "ask" missing from that:-) |
| 23:05 | <TabAtkins> | annevk: Agreed. |
| 23:06 | <MikeSmith> | atwilson: so maybe based on deployment experience with the background-pages feature, could still end up producing something we might want to spec out for standardization/ |
| 23:06 | <foolip_> | david_carlisle, I'll ask them |
| 23:06 | <atwilson> | MikeSmith: So, long-term I still want to expose this as part of the web platform, but the consensus on the list was that we really needed to experiment with the functionality first. |
| 23:06 | <atwilson> | So, yeah :) |
| 23:08 | <atwilson> | MikeSmith: Hopefully people will find it useful, people won't use it to create "bot-net construction kits" (hi maciej!) and we can move forward. |
| 23:09 | <MikeSmith_> | damn |
| 23:10 | <MikeSmith_> | atwilson: the bot-net construction kit was my main use-case |
| 23:10 | <MikeSmith_> | so I hope it'll keep working for that purpose :) |
| 23:10 | <MikeSmith_> | anyway, thanks |
| 23:10 | <annevk> | I wonder when people will understand what the W3C actually is: http://twitter.com/dylanw/status/29674513004232704 |
| 23:10 | <MikeSmith_> | oh hey |
| 23:10 | <MikeSmith_> | atwilson: about your test cases |
| 23:11 | <atwilson> | yah? |
| 23:11 | <MikeSmith_> | I hope you will check those into the mercurial repo |
| 23:11 | <MikeSmith_> | did you get my reply about that/ |
| 23:11 | <MikeSmith_> | ? |
| 23:12 | <atwilson> | MikeSmith: I saw your reply, and that's my plan. I've just been swamped preparing for the Chrome 10 release so haven't had any free cycles (plus I'm out this week for some family vacation). Hopefully I can do it next week. |
| 23:12 | <atwilson> | MikeSmith: I'm somewhat concerned that they'll bit-rot once they are out of the WebKit repository, though. |
| 23:13 | <MikeSmith> | well, we really need to get more going with testing for WebApps WG specs this year |
| 23:14 | <MikeSmith> | have people actively reviewing submitted test cases and migrating them into a central set |
| 23:14 | <Hixie> | abarth: re sniffing, we might want a separate video sniffing section so that it's treated like <img>, and doesn't run the risk of other types being sniffed (e.g. <video src="a.pdf"> should probably sniff unknown, not pdf) |
| 23:14 | <abarth> | yes |
| 23:14 | <abarth> | isn't that what http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType is all about? |
| 23:15 | <Hixie> | abarth: yes |
| 23:15 | <Hixie> | abarth: i meant in the websec-mime-sniff draft |
| 23:16 | <abarth> | i can certainly add that if its useful for you |
| 23:16 | <abarth> | my thought was to wait to see if http://www.w3.org/html/wg/wiki/ChangeProposals/NoVideoContentType was going to prevail, but i can do it sooner |
| 23:16 | <Hixie> | no rush |
| 23:16 | <Hixie> | indeed, no point doing it before/unless NoVideoContentType wins |
| 23:17 | <Hixie> | i just meant in general, in the future |
| 23:17 | <annevk> | abarth, need one of those for fonts too |
| 23:18 | <annevk> | abarth, which ideally @font-face references and all |
| 23:18 | <abarth> | annevk: ok, do you know what the right signatures are? |
| 23:19 | <annevk> | I guess I'll sleep, try to sort that out, and email websec |
| 23:19 | <annevk> | thanks |
| 23:20 | <abarth> | i'm happy to add it. if you're willing to do some legwork, it will be quicker :) |
| 23:31 | <foolip_> | abarth, audio/x-wave is the one permutation of audio/(x-)?wav(e)? that doesn't exist in the wild. would you get shot for using audio/wave? |
| 23:31 | <abarth> | haha |
| 23:31 | <abarth> | lets try audio/wave |
| 23:31 | <abarth> | i got it from http://wiki.whatwg.org/wiki/Video_type_parameters |
| 23:32 | <abarth> | but then screwed up the transcription |
| 23:32 | <abarth> | fixed |
| 23:32 | <Hixie> | abarth: btw while we're at it, http://tools.ietf.org/html/draft-ietf-websec-origin mentions the term "principal" twice (in the same way) but doesn't say what that means and doesn't really rely on the definition anyway |
| 23:32 | <foolip_> | video/webm also isn't registered yet, either you'll live or be doubly shot |
| 23:33 | <abarth> | Hixie: its meant to be conceptual, not definitional |
| 23:33 | <Hixie> | k |
| 23:33 | <Hixie> | i have no idea what it means :-) |
| 23:34 | <abarth> | maybe i should put it in quotes? |
| 23:34 | <abarth> | it's a term from the orange book |
| 23:34 | <abarth> | which is a classic text on security |
| 23:34 | <Hixie> | ah |
| 23:35 | <abarth> | its in quotes now |
| 23:35 | <Hixie> | would be good to have a reference i guess |
| 23:35 | <Hixie> | for the uneducated like me :-) |
| 23:35 | <abarth> | ok, will do |
| 23:35 | <abarth> | i suspect that the origin draft will get out of the working group in a finite time |
| 23:35 | <abarth> | the sniffing draft might take a bit longer : |
| 23:35 | <abarth> | ) |
| 23:36 | <Hixie> | (none of the dictionary definitions (including two security-related ones) that i found for "principal" as a noun make sense in context) |
| 23:36 | <Hixie> | well so long as the spec's URL doesn't change and so long as the spec remains maintained over time, I don't much care when it comes out of working group :-) |
| 23:36 | <abarth> | http://alt.pluralsight.com/wiki/default.aspx/Keith.GuideBook/What%20Is%20A%20Security%20Principal.html |
| 23:37 | <abarth> | its a new bullet for that list |
| 23:37 | <abarth> | as it were |
| 23:38 | <Hixie> | an origin is "an entity that can be positively identified and verified via authentication" ? |
| 23:39 | <abarth> | yep |
| 23:39 | <abarth> | the authentication is pretty weak |
| 23:39 | <abarth> | for HTTP |
| 23:39 | <abarth> | but better for HTTPS |
| 23:40 | <david_carlisle> | Hixie: could the checkboxes for web dev / implementer views in the spec have tooltips or some other inline indication of what is or isn't in each view (I know there's some description in the faq but....) |
| 23:41 | <Hixie> | david_carlisle: we're going to be changing the whole way that works and replacing the author view with developers.whatwg.org, so I'll update it when we do that |
| 23:41 | <Hixie> | (you can look at developers.whatwg.org for a preview) |
| 23:41 | <david_carlisle> | ta, no big deal just a passing comment as I can never remember if im a web dev or not:-) |
| 23:43 | <david_carlisle> | ah @@@@ I suppose we'll have to make the mml spec look nice to fit now? |
| 23:45 | <Hixie> | the spec isn't going to change, we're just going to provide two versions, one for authors and one for implementors |
| 23:46 | <david_carlisle> | sure, but clicking through from developers.whatwg.org to w3.org/tr/mathml is visually jaring:-) |
| 23:47 | <Hixie> | oh heh |
| 23:54 | <bga_> | wow. http://www.w3.org/TR/file-system-api/ |
| 23:54 | <bga_> | you want to copy all ie's axtivexes |
