| 01:13 | <Hixie> | how is -moz-box-sizing still not unprefixed |
| 01:13 | <Hixie> | it's been over a decade now |
| 05:45 | <zcorpan> | wonder if we can introduce a way for authors to opt out of https://www.w3.org/Bugs/Public/show_bug.cgi?id=11960 (something like "use strict") |
| 05:48 | <zewt> | heh, i've started writing pages with fewer and fewer ids ... not really for that reason, but it does help reduce it |
| 05:52 | <zewt> | heh microsoft saying "pages break" and the only example anyone's come up with is microsoft.com? really? |
| 05:55 | <zcorpan> | yep |
| 05:55 | <zcorpan> | but i'm sure there are more that break, they're just hard to find (especially since they likely use browser sniffing and barewords) |
| 06:10 | <zcorpan> | what's up with http://html5.org/tools/web-apps-tracker?from=7277&to=7278 ? |
| 06:13 | <zcorpan> | Hixie: btw, having links to bug in commit message and link to diff in bug comment was quite helpful. is it much extra work? |
| 06:16 | <MikeSmith> | zcorpan: weird. http://lists.whatwg.org/pipermail/commit-watchers-whatwg.org/2012/006432.html has the diff |
| 06:16 | <zcorpan> | oh, nevermind, i see you included that in later commits |
| 06:16 | <zcorpan> | MikeSmith: yeah |
| 06:33 | <annevk> | Ms2ger: next time you just remove yourself from the CC list, please don't add a comment |
| 06:33 | <annevk> | Ms2ger: now I get email, otherwise I wouldn't |
| 07:18 | <annevk> | whatever came out of the lazy Blob thread? |
| 07:18 | <annevk> | nothing? |
| 07:20 | <Ms2ger> | Sounds about right |
| 07:25 | <zcorpan> | annevk: you expect things to *happen*? are you new here? :-) |
| 07:32 | <annevk> | zcorpan: yeah, what is this place anyway? |
| 07:32 | <jgraham> | This is #lazylayabouts |
| 07:32 | <jgraham> | But we changed the name for PR reasons |
| 07:36 | <annevk> | ooh |
| 07:36 | <annevk> | I'm no longer in WebApps or WebAppSec |
| 07:36 | <jgraham> | I am very tempted to respond to dglazkov's packaging thread with "show me the use cases". It's not that I doubt that there are use cases for web sites having dependencies, but it is very hard to tell what the actual requirements are without having been to his f2f meetings |
| 07:36 | <annevk> | and no longer have W3C Member access |
| 07:36 | <annevk> | more freedom, yay |
| 07:37 | <annevk> | no W3C Member access, but still in Chairs |
| 07:37 | <annevk> | I wonder how that works |
| 07:37 | <annevk> | hahaha |
| 07:37 | <annevk> | I can no longer access the Chairs guide |
| 07:38 | <annevk> | I'm in Chairs, but not in the Notifications WG |
| 07:38 | <annevk> | weird |
| 07:45 | <smaug____> | annevk: Join Mozilla and become a member again :) |
| 08:36 | <annevk> | I'm still in the Forms Task Force |
| 08:36 | <annevk> | anyone remember that one? |
| 08:36 | <Ms2ger> | No :) |
| 08:37 | <jgraham> | annevk: So, which is more useful, the Forms Task Force or the Chairs list? |
| 08:37 | <annevk> | Chairs I guess, I still get the supposedly Member-only email |
| 08:38 | <Ms2ger> | Oh, look, https://svgwg.org/svg2-draft/ |
| 08:40 | <annevk> | https://svgwg.org/svg2-draft/changes.html is the interesting bit |
| 08:41 | <annevk> | "Added the ‘maskType’ attribute to the ‘mask’ element." yay more HTML parser changes upcoming |
| 08:41 | <Ms2ger> | Oh look, a reference to DOM4 |
| 08:42 | <annevk> | and why not name it "type"? |
| 08:42 | <Ms2ger> | As well as DOM2EVENTS, DOM2STYLE and DOM2VIEWS |
| 08:42 | <annevk> | why not Zoidberg? |
| 08:43 | <annevk> | overall though it seems like many small steps in the right direction |
| 08:46 | <annevk> | nobody updated DOM4 in the last two months |
| 08:49 | <annevk> | MikeSmith: so say I still want to try this W3C thing for now, how do I apply for Invited Expert status for WebApps/WebAppSec/Notifications? |
| 08:49 | <annevk> | MikeSmith: I assumed that would happen automatically when I changed affiliation, but apparently it does not |
| 08:53 | <annevk> | ah, found https://www.w3.org/2002/09/wbs/1/ieapp/ |
| 08:53 | <zcorpan> | yay for svg to embrace webidl and dom4 |
| 08:55 | <annevk> | hmm |
| 08:55 | <annevk> | maybe I don't want to be an Invited Expert |
| 08:56 | <annevk> | "The Invited Expert agrees to refrain from creating derivative works that include the Invited Expert's contributions when those derivative works are likely to cause confusion about the status of the W3C work or create risks of non-interoperability with a W3C Recommendation. «Branching» is one example of a non-permissible derivative work." |
| 08:56 | <annevk> | from http://www.w3.org/Consortium/Legal/2007/06-invited-expert.html |
| 08:57 | <annevk> | seems kind of contradictory to giving the W3C a nonexclusive copyright |
| 09:08 | <zcorpan> | annevk: sent an email about maskType |
| 09:09 | <annevk> | and I submitted aforementioned form by indicating I do not agree with the terms |
| 09:09 | <annevk> | I wonder what happens |
| 09:17 | <odinho> | I don't understand what that thing is supposed to prohibit. |
| 09:17 | <odinho> | Evil invited experts? And what about evil companies/members then? |
| 09:18 | <Ms2ger> | odinho, people publishing specs outside the W3C and undermining the W3C's authority |
| 09:18 | <Ms2ger> | Basically, WHATWG |
| 09:18 | <odinho> | Ms2ger: But why only place that on invited experts? |
| 09:19 | <odinho> | I mean, it's a stupid thing, but it has squared stupidity when it's not consistent. |
| 09:19 | <jgraham> | Ms2ger: Also, say, ePub |
| 09:19 | <Ms2ger> | Hah, ePub |
| 09:20 | <annevk> | Ms2ger, you also Ms2ger on github? |
| 09:20 | <jgraham> | (I mean if http://idpf.org/epub/30/spec/epub30-contentdocs.html#sec-xhtml-additions isn't a "derivative work" I don't know what is) |
| 09:21 | <jgraham> | (and clearly by implementing additions, you risk non-interoperability with HTML UAs that fail to implement those additions) |
| 09:22 | <jgraham> | (plus the next section is called "deviations") |
| 09:23 | <Ms2ger> | annevk, yes |
| 09:23 | <Ms2ger> | Though I hate it |
| 09:23 | <jgraham> | Ms2ger: You don't like git? Who knew? |
| 09:24 | <Ms2ger> | "Deviations" reminds me of http://en.wikipedia.org/wiki/The_Chrysalids |
| 09:24 | <annevk> | Ms2ger: I read you can access it through hg-git, so you should be okay |
| 09:28 | <annevk> | jgraham: you did this kind of thing before right? |
| 09:28 | <annevk> | jgraham: say I create a github repository, can I then put a hg repository in it without much hassle? |
| 09:28 | <annevk> | jgraham: or is there some other way to go about it |
| 09:29 | <jgraham> | annevk: doublec knows all. Im particular http://www.bluishcoder.co.nz/2011/02/10/git-conversion-of-mozilla-central.html |
| 09:31 | <annevk> | interesting |
| 09:31 | <annevk> | I don't necessarily care about keeping things in sync afterwards |
| 09:31 | <annevk> | but I guess I should install that hg-git thing |
| 09:32 | <jgraham> | Yeah it is still probably the best way of doing the one-time import |
| 09:32 | <jgraham> | Or export depending on your point of view |
| 09:34 | <annevk> | of course, first need to update MacPorts |
| 09:41 | <jgraham> | OK, why do gecko/webkit fail to run the last two tests in http://hoppipolla.co.uk/tests/history/001.html (popups needed) |
| 09:42 | <jgraham> | (yes the test is kind of very ugly) |
| 09:46 | <annevk> | almost catched up on 100 or so threads, nothing much interesting, apart from a few minor bugs in DOM |
| 09:47 | <annevk> | a lot of agenda bullshit |
| 10:20 | <annevk> | apparently my 512x512 rendering of the WHATWG logo downscales poorly as seen on https://github.com/whatwg |
| 10:20 | <annevk> | if anyone has a better solution, please let me know |
| 10:21 | <Ms2ger> | Do they take SVG images? |
| 10:22 | <jgraham> | annevk: Surely if you downscale it yourself in photogimp you can get acceptable results? |
| 10:23 | <annevk> | Ms2ger: nope, goes via gravatar |
| 10:24 | <annevk> | jgraham: the problem is they scale it themselves to various resolutions |
| 10:26 | <jgraham> | annevk: So wait, we are spending time worrying about the tiny fraction of people with retina displays, and they won't even let you upload a 64x64 image to use at 64x64? |
| 10:26 | <jgraham> | er 140x140 |
| 10:26 | <annevk> | uhuh |
| 10:26 | <annevk> | the web is a funny place |
| 10:26 | <annevk> | or hell, but that depends on your pov |
| 10:26 | <jgraham> | Wait |
| 10:27 | <jgraham> | It is github's fault (I think?) |
| 10:27 | <jgraham> | They are showing a 140x140 image at 48x48 |
| 10:27 | <annevk> | that sounds like their fault |
| 10:27 | <ruby_on_tails> | how to fallback to the default font till the custom font used in a canvas loads ? |
| 10:27 | <annevk> | maybe also why it renders somewhat better depending on which browser you use |
| 10:28 | <ruby_on_tails> | i suddenly have a slow internet and the font takes too long to load and other things start animating on the canvas |
| 10:28 | <ruby_on_tails> | and no text shows up till the font loads which fails a lot of times on slow net |
| 10:28 | <ruby_on_tails> | thats bad |
| 10:29 | <annevk> | ruby_on_tails: can't you set a fallback font? |
| 10:29 | <ruby_on_tails> | annevk: how ? |
| 10:29 | <ruby_on_tails> | ctx.font = "20px arial" |
| 10:29 | <ruby_on_tails> | can i use "20px arial, verdana" ? |
| 10:29 | <ruby_on_tails> | let me try |
| 10:30 | <ruby_on_tails> | ctx.font = "50px bebas, arial"; doesnt work |
| 10:30 | <annevk> | per the spec that's a browser bug |
| 10:31 | <ruby_on_tails> | whats the solution for now ? |
| 10:31 | <annevk> | file a bug on the relevant browser |
| 10:31 | <annevk> | and you can check if something was drawn and if not, reset the font and do it again |
| 10:31 | <annevk> | I guess |
| 10:36 | <ruby_on_tails> | but is that code supposed to work ? |
| 10:36 | <ruby_on_tails> | ctx.font = "50px bebas, arial"; |
| 10:36 | <annevk> | yeah |
| 10:37 | <ruby_on_tails> | i cant see such usage in any SO answer or doc reference |
| 10:38 | <annevk> | tutorials never cover the whole spec |
| 10:39 | <ruby_on_tails> | hmm |
| 10:40 | <ruby_on_tails> | why doesnt window.onload wait for fonts ? |
| 10:40 | <ruby_on_tails> | it should wait for fonts like images i guess |
| 10:40 | <ruby_on_tails> | maybe i am wrong |
| 10:43 | <jgraham> | Sigh |
| 10:43 | <jgraham> | So browsers don't seem to fire pageshow events from document.open |
| 10:44 | <jgraham> | Any idea if this is a compat. need? |
| 10:57 | <zcorpan> | downscaling a big image sounds like the right thing to do if you care about retina displays (in this case the image is small enough to not bother with multiple versions) |
| 10:57 | <zcorpan> | only opera's rendering looks bad (i filed a bug) |
| 10:58 | <zcorpan> | or well, the *right* thing to do in this case would be to use svg |
| 10:58 | <jgraham> | Downscaling an image by a non integer ratio sounds like a stupid thing to do |
| 11:00 | <zcorpan> | if only the 140x140 version was uploaded to gravatar and you want to display it as 48x48, it has to be downscaled *somewhere* |
| 11:00 | <jgraham> | zcorpan: annevk said he had a 512x512 version |
| 11:01 | <zcorpan> | oh |
| 11:01 | <zcorpan> | well then there's something stupid going on indeed |
| 11:01 | <jgraham> | (still not a good ratio of course, but downscaling that to 140 and then to 48 is crazy) |
| 11:02 | <annevk> | and if you want perfect icons, you need several of them I'm told |
| 11:02 | <annevk> | relying on scaling algorithms is not a good idea per the icon crowd |
| 11:05 | <Ms2ger> | http://i.qkme.me/3qnjb3.jpg |
| 14:35 | <zewt> | i wonder who would name a child "RGraph.net" |
| 14:36 | <jgraham> | There was a guy at Opera named his children C APL and Ruby |
| 14:36 | <zewt> | can never understand what's going on the the heads of people who post to mailing lists with a company name |
| 14:49 | <karlcow> | zewt: this maybe http://infofranpro.wdfiles.com/local--files/20100624-brand-identity/Brain3d.jpg |
| 15:07 | <filipe> | alguem conhece um bom material para quem esta come�ando em html5? |
| 15:09 | <jgraham> | filipe: Maybe http://html5doctor.com/resources/ will help you |
| 15:22 | <GPHemsley> | Is there a preferred media type for WAVE files? 'audio/vnd.wave' is the one in the RFC, but 'audio/wav' seems to be commonly used. |
| 16:10 | <dglazkov> | good morning, Whatwg! |
| 16:12 | <deane> | dglazkov: Good morning. |
| 17:28 | <Hixie> | is http://philip.html5.org/tests/canvas/suite/tests/ the most up to date / canonical test suite for canvas? |
| 17:28 | <Ms2ger> | No |
| 17:28 | <Ms2ger> | http://www.w3c-test.org/html/tests/submission/PhilipTaylor/canvas/ |
| 17:30 | <Hixie> | ah k thanks |
| 17:30 | <Hixie> | Philip`: could be good to update your copy with a link to the more canonical copy :-) |
| 17:30 | <Hixie> | not that i've done that with any of my css tests |
| 17:30 | <Hixie> | but you know |
| 17:30 | <Ms2ger> | I've been telling him |
| 17:30 | <Ms2ger> | Also that he needs to fix all the bugs :) |
| 17:35 | <Hixie> | annevk: http://html5.org/tools/web-apps-tracker?from=7277&to=7278 ? |
| 17:36 | <Ms2ger> | Looks empty |
| 17:36 | <Hixie> | zcorpan: Re the bug #s, I still include them if doing so is easy. I'm more likely to omit them if e.g. I've already left the bug and moved on to other things by the time I get around to committing. |
| 17:37 | <Hixie> | zcorpan: (I used to have to go out of my way to get the bug #s before because the Process required that I describe the change in the bug and the diff was the easiest way to do that) |
| 17:37 | <Hixie> | zcorpan: (and so I just had my script post the diff to the bug) |
| 17:50 | Philip` | is aware that he needs to do many things :-( |
| 17:51 | <Hixie> | join the club |
| 17:54 | <Philip`> | The difference is that you do at least some of many things, whereas I tend to do nothing |
| 17:55 | <Hixie> | surely you do _some_thing |
| 17:56 | <Philip`> | None of the HTML-related things, at least |
| 17:56 | <Hixie> | ah, well, there you go |
| 17:56 | <Hixie> | i don't do the non-html-related things :-) |
| 18:14 | jgraham | wonders what Philip` does do |
| 18:14 | <Ms2ger> | Gazing at the skies, I assume |
| 18:15 | <Hixie> | i added a confirm() prompt to the "bug" link on the live dom viewer, hopefully that will be enough |
| 18:15 | <Hixie> | (there were 4-6 bogus bugs filed over the weekend) |
| 18:15 | <Ms2ger> | I'm glad I opted out of the cc, then :) |
| 18:15 | <jgraham> | Maybe http://krijnhoetmer.nl/irc-logs/whatwg/20090318#l-47 is still accurate |
| 18:17 | <Hixie> | if it gets too bad i'll just make the link redirect to a prefilled bug submission page |
| 18:17 | <Hixie> | that way you have to have an account |
| 18:26 | <Philip`> | jgraham: Nowadays, mostly a job |
| 18:26 | Philip` | doesn't think he's played TF2 for at least a year, except for this afternoon |
| 18:29 | <Hixie> | anyone got any Content-Language tests? |
| 18:29 | <Hixie> | <meta> pragma Content-Language |
| 18:29 | <Hixie> | nevermind, i found mine |
| 18:32 | <Ms2ger> | Maybe Julian has some |
| 19:10 | <bencc> | are there plans to add CORS to web workers? |
| 19:12 | <Hixie> | bencc: to do what? |
| 19:13 | <bencc> | Hixie: to load a worker url from a different domain |
| 19:13 | <Hixie> | CORS wouldn't be a sane way to do that |
| 19:13 | <Hixie> | there's two possible use cases that i see here: |
| 19:13 | <Hixie> | 1) running a same-origin worker loading script from a remote oritin |
| 19:13 | <bencc> | new Worker(someOtherUrl) |
| 19:14 | <Hixie> | 2) running a worker on another origin |
| 19:14 | <Hixie> | 1) is possible today using importScripts() or whatever the method is called |
| 19:14 | <bencc> | I think I'm talking about 2) |
| 19:14 | <Hixie> | yeah, i plan to spec 2) in due course |
| 19:14 | <Hixie> | (it wouldn't use CORS) |
| 19:15 | <bencc> | I don't really need CORS just to use a worker from a different domain |
| 19:15 | <bencc> | can I use an inline worker with importScritpts? |
| 19:16 | <Hixie> | inline? |
| 19:16 | <bencc> | I'm trying to use https://raw.github.com/mozilla/pdf.js/gh-pages/build/pdf.js |
| 19:16 | <bencc> | the script loads a worker which uses the same script |
| 19:17 | <Hixie> | scripts can import scripts cross-origin iirc, yes |
| 19:17 | <bencc> | but it doesn't work if the pdf.js script is hosted on a different domain |
| 19:17 | <Hixie> | just like <script> |
| 19:17 | <Hixie> | it's just new Worker() that can't |
| 19:17 | <Hixie> | because new Worker() is intended to run the scipts in their origin, and i haven't specced how that'll work yet (it's a security thing) |
| 19:18 | <bencc> | what's the difference between new Worker(different domain) |
| 19:18 | <bencc> | and import a script inside the worker? |
| 19:18 | <bencc> | why the former is a security issue? |
| 19:19 | <Hixie> | the current difference is that hte former doesn't do anything and the latter imports the script |
| 19:19 | <bencc> | I mean, why is the former a security issue? |
| 19:20 | <Hixie> | the future difference will be that the former runs the script in the security context of the remote origin, and it's a security thing because we need to ensure the script opts in to running stuff in its origin, because otherwise you could do all kinds of crazy stuff |
| 19:21 | <Hixie> | it'll probably involve an HTTP header similar to CORS (but not CORS, because CORS would also let you read the data, which is a different thing altogether) |
| 19:21 | <bencc> | so to make pdf.js work, I just need to create a worker that all it does is importScripts('path/to/pdf.js') ? |
| 19:22 | <Hixie> | yeah. you can even create a worker with the URL "data:text/javascript,importScripts('path/to/pdf.js')" |
| 19:22 | <Hixie> | per spec, anyway |
| 19:22 | <Hixie> | dunno how many browsers support that yet |
| 19:22 | <bencc> | I'll check that. thanks |
| 19:32 | <ojan> | Hixie: http://html5.org/tools/web-apps-tracker?from=7277&to=7278 how did you come to the conclusion that this is quirks only in webkit? |
| 19:32 | <ojan> | Hixie: what i see in local testing is that we always report the computed style as content-box, but we treat it as border-box, quirks or standards |
| 19:32 | <Hixie> | testing? :-) |
| 19:32 | <Hixie> | let me find the tests again |
| 19:33 | <bencc> | Hixie: looks good |
| 19:33 | <ojan> | Hixie: thx, just want to make sure there's nothing i'm missing before i start filing webkit bugs |
| 19:33 | <bencc> | Hixie: I'm able to use bb = new BlobBuilder(); |
| 19:33 | <bencc> | bb.append("importScripts('/path/to/file')") |
| 19:34 | <bencc> | Hixie: is there a difference between: new Blob(["..."]) and using BlobBuilder ? |
| 19:34 | <Hixie> | i know nothing of blobs |
| 19:35 | <Ms2ger> | bencc, no |
| 19:35 | <Ms2ger> | bencc, except that I'm about to remove BlobBuilder support from Gecko |
| 19:36 | <bencc> | Ms2ger: new Blob() has prefix on some browsers? |
| 19:36 | <Ms2ger> | No |
| 19:36 | <bencc> | cool |
| 19:37 | <Hixie> | ojan: http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=1621 http://software.hixie.ch/utilities/js/live-dom-viewer/?saved=1622 |
| 19:37 | <Ms2ger> | Hixie, did you file a bug for the tests? :): |
| 19:37 | <Hixie> | ojan: oh, i misintepreted the testcase |
| 19:37 | <bencc> | Ms2ger: is there a safe way to check if Blob can be used? |
| 19:38 | <Hixie> | ojan: the bug is that webkit applies it to display:table in quirks mode |
| 19:38 | <Hixie> | Ms2ger: which? |
| 19:38 | <Ms2ger> | "Those" |
| 19:38 | <Hixie> | Ms2ger: content-language or box-sizing-on-table? |
| 19:38 | <Ms2ger> | What's content-language, and should I care? :) |
| 19:39 | <Hixie> | content-language is the tests i was just dealing with and i just checked in a diff that mentions them, so i thought that might be what you were referring to :-) |
| 19:39 | <Hixie> | i haven |
| 19:39 | <Hixie> | 't filed any |
| 19:39 | <Hixie> | bugs on testcases |
| 19:39 | <Hixie> | be my guest :-) |
| 19:39 | <ojan> | Hixie: the difference i see between webkit and gecko is only in the second case of a table that is display:block |
| 19:40 | <Hixie> | ojan: oh dude, i'm just not paying enough attention with this test |
| 19:40 | <ojan> | Hixie: lol |
| 19:40 | <ojan> | Hixie: the gecko behavior seems wrong to me... |
| 19:40 | <Ms2ger> | Clearly Gecko must be right ;) |
| 19:40 | <ojan> | Hixie: as in...if we can get away with it, a display:block table should default to content-box |
| 19:41 | <Ms2ger> | Ugh |
| 19:41 | <ojan> | Hixie: where webkit only border-boxes display:block tables in quirks mode |
| 19:41 | <Hixie> | ojan: in webkit, the two tests differ in quirks vs non-quirks |
| 19:41 | <ojan> | i don't feel strongly about this |
| 19:41 | <ojan> | i suppose making all tables border-box is simpler to spec |
| 19:41 | <Hixie> | ojan: mozilla doesn't differ in quirks vs non-quirks |
| 19:41 | <ojan> | i'll file a bug to make webkit do that |
| 19:41 | <Ms2ger> | Making it depend on the display seems sucky |
| 19:41 | <Hixie> | ojan: and no quirkiness is a win |
| 19:41 | <ojan> | yeah |
| 19:42 | <Hixie> | ojan: it should just be a line in the CSS style sheet |
| 19:42 | <ojan> | i agree with both of you :) |
| 19:42 | <Hixie> | ojan: i don't understand what webkit is doing, looks like it has both the line in the style sheet and some extra crazy for <table>'s renderer |
| 19:42 | <ojan> | it's lame that webkit does this with under the hood magic instead of just changing the computed style |
| 19:42 | <ojan> | Hixie: i don't think there's a line in the style sheet |
| 19:42 | <Hixie> | why does it affect display:block table then? |
| 19:43 | <Hixie> | (but only in quirks mode) |
| 19:43 | <Ms2ger> | The HTML5 editors sure seem to spend a lot of time triaging |
| 19:43 | <ojan> | Hixie: def no line in the style sheet... |
| 19:43 | <Hixie> | Ms2ger: ? |
| 19:43 | <ojan> | Hixie: meh, we're probably going down the table code by tagname somewhwere or something dumb |
| 19:43 | <Ms2ger> | Hixie, not you, the irrelevant ones ;) |
| 19:43 | <Hixie> | Ms2ger: over the weekend? |
| 19:44 | <Hixie> | Ms2ger: i'm not an html5 editor, i was just surprised to hear they'd done anything, when i checked in this morning it was unchanged since the 23rd |
| 19:44 | <Hixie> | and i haven't seen any bug spam or public-html postage since |
| 19:45 | <Hixie> | in fact as far as i can tell, the new editors don't post to public-html even as much as i did |
| 19:45 | <Hixie> | in fact i'm pretty sure some of them have _never_ posted |
| 19:45 | <Ms2ger> | Seems like they like to add keywords and stuff |
| 19:46 | Ms2ger | has been blissfully ignorant of them |
| 19:46 | <Hixie> | oh i don't get keyword bugmail |
| 19:46 | <Hixie> | :-) |
| 19:46 | <Hixie> | i turned that off after the a11y people started drizzling keywords everywhere |
| 19:47 | <Ms2ger> | Summary: "Fix typo", keywords: "a11y-foo a11y-bar" |
| 19:50 | <Hixie> | oh, i am getting the bugmail, my deletion script just hasn't noticed yet |
| 19:50 | <Hixie> | (not for keywords but for the changes where one of them was closing bugs on the principle that they didn't apply even though they did...) |
| 19:50 | <Hixie> | (you'd think the editors would know what sections they had in their spec) |
| 20:42 | <ojan> | Hixie: i expect your aware of this, but http://html5.org/r/7278 doesn't seem to have the actual diff |
| 20:47 | <TabAtkins> | Hixie: I forget - did we discuss an addition to srcset for providing an intrinsic size for different "art-direction" choices? |
| 20:47 | <ojan> | Hixie: if you care https://bugs.webkit.org/show_bug.cgi?id=95123 file the table box-sizing bug |
| 21:07 | <annevk> | Hixie: dunno why it's empty |
| 21:07 | <annevk> | Hixie: the rest seems to work |
| 21:07 | <annevk> | Hixie: is it non-empty if you diff it yourself? |
| 21:25 | <GPH-Zeke> | annevk, Hixie, ojan: 7279 is also empty. |
| 21:25 | <annevk> | hmm okay |
| 21:25 | <GPH-Zeke> | (two in a row) |
| 21:25 | <annevk> | so maybe we're running out of space again |
| 21:25 | <annevk> | please someone remind me tomorrow |
| 21:26 | <GPHemsley> | I assumed there was some part of the spec that wasn't covered by the diff site, but what do I know? :) |
| 21:29 | <annevk> | well that's true too |
| 21:30 | <annevk> | but the rendering section is part of source |
| 21:38 | <Hixie> | TabAtkins: we discussed it in person, but not on the list (and nobody else has brought it up as far as i am aware) |
| 21:38 | <Hixie> | ojan: thanks |
| 21:39 | <Hixie> | GPHemsley: there are cases like that, but they don't get listed on the tracker page at all because the tracker page only looks at source's logs, iirc |
| 21:41 | <TabAtkins> | Okay, I was just wondering about it, since we can ignore the issue if we adopt CAS. ^_^ |
| 21:42 | <Hixie> | CAS? |
| 21:42 | <Hixie> | oh the attribte sheets |
| 21:47 | <GPHemsley> | ah, ok |
| 23:04 | <zewt> | EVERYONE TALKING ABOUT THE PUSH API DRAFT (whatever that is, sounds like WebSockets) IS IMPRESSIVELY LOUD |
| 23:05 | Hixie | searches for "push api" in his inbox |
| 23:05 | <Hixie> | oh you just mean their names |
| 23:06 | <Hixie> | oh god, more permission-based apis |
| 23:06 | <Hixie> | also, seriously, respec CONSIDERED HARMFUL |
| 23:06 | <Hixie> | _everyone_ who uses it ends up screwing up how they write a spec |
| 23:08 | <zewt> | is that respec's fault or the fault of the category of people who decide to use respec? :) |
| 23:08 | <zewt> | (seems like "people who screw up writing a spec" is a pretty huge category to begin with) |
| 23:09 | <Hixie> | if the category of people who write good specs never use respec, that is also an indication of a problem with respec |
| 23:15 | <Yuhong> | http://blogs.msdn.com/b/ieinternals/archive/2012/08/28/custom-file-type-formats-renamed-to-zip-on-download-in-internet-explorer-when-application-octet-stream-is-used.aspx |
| 23:17 | <zewt> | archived material from the future, MS has *all* the tech |
| 23:18 | <TabAtkins> | Looks like the date in the url is generated based on the UTC date, and it's already past midnight there. ^_^ |
| 23:18 | <zewt> | also "And in the case of ZIP-based formats, the browser's technically right" no, the browser is technically an idiot and that behavior is always 100% wrong |
| 23:19 | <Yuhong> | technically right in terms of it being openable by a zip reader. |
| 23:19 | <zewt> | i had to work around that idiocy way back when I had my own ZIP-based file format; had to fiddle with the header to keep it from happening (though I thought it was some ancient Safari that had that problem; this was years ago) |
| 23:20 | <Yuhong> | It is just not what you generally want to open the file with. |
| 23:21 | <zewt> | (also, "file starts with PK"? seriously? it thinks a text file that says "PKZIP rocks" is a ZIP?) |
| 23:22 | <Hixie> | zewt: windows thinks a file that starts with a "MZ IS AWESOME" is an executable... |
| 23:22 | <zewt> | (the header for ZIP is \x50\x4b\x03\x04) |
| 23:22 | <Hixie> | (iirc) |
| 23:22 | <Hixie> | (MZ is the DOS .exe header, iirc) |
| 23:23 | <Yuhong> | Yep. |
| 23:28 | <Yuhong> | http://blogs.msdn.com/b/oldnewthing/archive/2006/01/30/519388.aspx |
| 23:34 | <ojan> | TabAtkins: how about s/scope-constrained/scope-filtered ? |
| 23:34 | <ojan> | TabAtkins: that's my mental model of it |
| 23:35 | <TabAtkins> | That's a better name, yeah. |
| 23:35 | <ojan> | TabAtkins: and what browsers would do in practice (match everything and filter out) |
| 23:35 | <ojan> | constrained and contained are practically synonyms to me :) |
| 23:37 | <ojan> | i'm glad we're solving this problem though...scope-contained >>> scope-filtered/scope-constrained IMO and should be the one we use wherever we're not constrained by back-compat |
| 23:45 | <TabAtkins> | Scope-filtered is *useful*, and we should have it, but yeah, scope-contained is much more natural as a default. |
