| 00:02 | <roc> | can we forbid delinquent editors from posting new ideas to www-style? |
| 00:04 | <roc> | ("joke") |
| 00:13 | <MikeSmith> | heh |
| 08:20 | <annevk> | having xhr in hg is sooo nice |
| 08:20 | <annevk> | having it be a single spec is real nice too |
| 08:20 | <annevk> | changing the spec is easy again |
| 09:46 | <benschwarz> | I was just reminiscing over days gone by MikeSmith … http://twitter.com/#!/benschwarz/status/39891483897102336 |
| 09:47 | <MikeSmith> | that don't loading for me |
| 09:47 | <MikeSmith> | due to me being in china |
| 09:48 | <benschwarz> | "Without @sideshowbarker, the WHATWG developer spec wouldn't have happened. Lets make that clear, this man is a legend." — Me, Feb 22 |
| 09:48 | <MikeSmith> | heh |
| 09:48 | <MikeSmith> | famous all over town |
| 09:48 | <benschwarz> | oh snapple |
| 09:48 | <benschwarz> | https://plus.google.com/111991826926222544385/about |
| 09:49 | <benschwarz> | that picture, laying down |
| 09:49 | <benschwarz> | so good |
| 09:49 | <benschwarz> | haha |
| 09:50 | <MikeSmith> | no google+ here either! |
| 09:50 | <MikeSmith> | that's the one with the sign? |
| 09:51 | <MikeSmith> | unfortunately they moved the shop and they don't have that sign where anybody can lay down in front of it any more |
| 09:52 | <benschwarz> | MikeSmith: what will I do in tokyo then?! |
| 09:52 | <benschwarz> | we'll get a html5 gang sign pic in front of that thing one day |
| 09:53 | <MikeSmith> | they still got the sign |
| 09:54 | <MikeSmith> | maybe we can get them to roll it out |
| 09:56 | <benschwarz> | I'm sure we can |
| 09:56 | <benschwarz> | Not planned to be in tokyo for a while… but I'm really yearning to return |
| 09:58 | <hsivonen> | wow. a longish thread about the meaning of <foo/> in HTML4 on www-validator. again. still. in 2011. |
| 10:00 | <jgraham> | hsivonen: I would feel sorry for you if that wasn't so funny |
| 10:02 | <zcorpan> | hsivonen: pointer? |
| 10:04 | zcorpan | guesses http://lists.w3.org/Archives/Public/www-validator/2011Nov/0007.html |
| 10:04 | <MikeSmith> | yeah |
| 10:04 | <MikeSmith> | glad I skipped that one :) |
| 10:05 | <MikeSmith> | wonderfully intuitive "NET-enabling start-tag requires SHORTTAG YES" message the validator emits there |
| 10:05 | jgraham | is glad that we have start tags to enable the net |
| 10:06 | <MikeSmith> | we need more specs like SGML |
| 10:13 | <zcorpan> | MikeSmith: especially nice that the validator has stepped away from spec purity since html4's sgml decl says SHORTTAG YES |
| 10:13 | <zcorpan> | but still has confusing messages |
| 10:14 | <MikeSmith> | it still clings to spec purity in other ways |
| 10:14 | <MikeSmith> | like, reporting id=foo and id=FOO and duplicated IDs |
| 10:14 | <MikeSmith> | and not reporting e.g., <p right> as an error |
| 10:15 | <MikeSmith> | the SGML neckbeards were in a league of their own |
| 10:16 | <MikeSmith> | in other news, great entertainment value in Murata-san's bugzilla comments |
| 10:17 | <MikeSmith> | or should i say "comment" |
| 10:17 | <hsivonen> | zcorpan: you found the right thread |
| 10:17 | <MikeSmith> | because it's the same one he's made several times now |
| 10:17 | <MikeSmith> | www-validator is a pot of gold |
| 10:18 | <zcorpan> | most of the thread seems to discuss how to improve the message that says that the slash causes other confusing messages to be emitted |
| 10:18 | <zcorpan> | instead of, i dunno, removing the confusing messages and emitting one unconfusing message |
| 10:20 | <MikeSmith> | heh |
| 10:49 | <annevk> | okay |
| 10:50 | <annevk> | so the outstanding feature requests for XMLHttpRequest are |
| 10:50 | <annevk> | redirects |
| 10:50 | <annevk> | "chunked-text" |
| 10:50 | <annevk> | "chunked-arraybuffer" |
| 10:50 | <annevk> | "json" |
| 10:50 | <annevk> | "stream" |
| 10:50 | <annevk> | and the thread on WHATWG |
| 11:39 | <hsivonen> | hmm. Niels Leenheer (the html5test.com guy) doesn't make it too easy to find his email address on the site |
| 11:39 | <hsivonen> | maybe I should just tweet |
| 11:40 | <jgraham> | Yeah it is pretty annoying; there are bugs in the tests |
| 11:40 | <Neocortex> | whois says it's info⊙sn |
| 11:41 | <Neocortex> | http://sights.nl/en/ |
| 11:42 | <hsivonen> | Neocortex: thanks |
| 11:43 | <hsivonen> | yay. located his email address in my email archives |
| 12:05 | <hsivonen> | what's the best-maintained diveintohtml5 version today? |
| 12:06 | <hsivonen> | .info? the html5doctor version? |
| 12:06 | <hsivonen> | does either of them actually have fixes? |
| 12:21 | <oal> | I came across this http://dev.w3.org/csswg/css3-grid-align/#grid-concepts Is it supported in any browsers yet? |
| 12:23 | <hsivonen> | oal: IE10 might have prefixed support |
| 12:25 | <oal> | Great, I hope the other's follow |
| 13:59 | <kennyluck> | oal, there's announcement on webkit-dev that WebKit developers just started to implement it. I guess it'll take a month or so to have something to play with. |
| 14:05 | <oal> | kennyluck: Cool. I've read ~half of this now as well (http://www.w3.org/TR/css3-layout/). Do you know if the display: "abc" style will be supported as well? |
| 14:32 | <kennyluck> | oal, that I don't know. Neither of these is under heavy discussion on www-style, and I think css3-layout is probably less popular somehow. You'll be encourage in this channel to use the up-to-date version http://dev.w3.org/csswg/css3-layout/ by the way. :p |
| 14:33 | <oal> | Oh, thanks. I haven't really ever spent much time on the w3 sites before :P |
| 15:55 | <AryehGregor> | Ms2ger, timeless: AFAICT, all non-Gecko engines just always have .length = 0 for IDL operations. This is one of the few things from WebIDL that IE9 doesn't follow. Gecko's behavior is clearly more correct, though. |
| 15:56 | AryehGregor | doesn't know what varargs are supposed to do to .length, offhand |
| 15:56 | AryehGregor | has just ignored varargs so far in his WebIDL tests |
| 15:57 | <AryehGregor> | Also, when I say "all non-Gecko engines" I don't include Safari, if that's different from Chrome. I don't bother testing it separately, as a rule. |
| 15:57 | <AryehGregor> | So maybe Safari is right but Chrome is wrong. |
| 15:58 | <AryehGregor> | Hixie, I'm around now, on and off. |
| 16:26 | <timeless> | AryehGregor: the gecko behavior for xpidl stuff is length = the number of non optional arguments in the case where there is no overloading |
| 16:26 | <timeless> | in cases where there's overloading it depends on how it's implemented |
| 16:26 | <AryehGregor> | I think WebIDL requires that it's the maximum number of arguments you could ever legitimately pass to it, ignoring varargs. |
| 16:26 | <timeless> | generally overloading is implemented by declaring 0 arguments and having the native code deal w/ it |
| 16:26 | <Ms2ger> | Overloading in XPIDL? |
| 16:26 | <AryehGregor> | But the algorithm is complicated. |
| 16:27 | <timeless> | Ms2ger: well overloading is technically *outside* of xpidl |
| 16:27 | <Ms2ger> | Right |
| 16:27 | <timeless> | but from the perspective of things who are expecting different possible inbound arguments |
| 16:27 | <AryehGregor> | IIUC, it's max{number of arguments of f, counting optional but not varargs | f is an operation on the given interface with the given identifier}. |
| 16:27 | <timeless> | perhaps i shouldn't say 0 arguments but "however many common arguments there are" |
| 16:27 | <timeless> | which is typically 0.. |
| 16:28 | <Ms2ger> | Recently we've been using jsvals, fwiw |
| 16:28 | <timeless> | Ms2ger: yeah, i know |
| 16:28 | <timeless> | but that's recent stuff |
| 16:28 | <timeless> | i'm a historian |
| 16:28 | <Ms2ger> | :) |
| 16:28 | AryehGregor | isn't sure offhand how overloading works if one interface extends another and they share a method name but with different arguments |
| 16:28 | <timeless> | stuff that hasn't been dead and buried for a while doesn't interest me |
| 16:28 | <Ms2ger> | AryehGregor, is that allowed? |
| 16:28 | <timeless> | AryehGregor: in gecko, classically, "poorly" :) |
| 16:28 | <AryehGregor> | Ms2ger, possibly not. |
| 16:28 | <timeless> | in gecko you're allowed to do: |
| 16:28 | <timeless> | foo.InterfaceName.method() |
| 16:29 | <timeless> | foo.OtherInterface.method() |
| 16:29 | <timeless> | but that only works if the interface isn't a DOMClassInfo :) |
| 16:29 | <timeless> | (more or less) |
| 16:29 | <Ms2ger> | I think Mozilla managed to get that thrown out of WebIDL |
| 16:29 | <timeless> | the right answer should be that it should be forbidden |
| 16:29 | <AryehGregor> | You can always do InterfaceName.prototype.method.call(foo). |
| 16:29 | <timeless> | if you need a new method with different arguments, get your own method name and stop confusing poor users |
| 16:30 | <timeless> | AryehGregor: actually |
| 16:30 | <timeless> | does that always work? |
| 16:30 | <AryehGregor> | Why not? |
| 16:30 | <timeless> | window1.InterfaceName.prototype.method.call(window2.foo) |
| 16:30 | <AryehGregor> | That won't work. |
| 16:30 | <Ms2ger> | In Gecko, it might very well work |
| 16:30 | <timeless> | right, so your example doesn't work either :) |
| 16:30 | <Ms2ger> | But we're the odd one out |
| 16:30 | <AryehGregor> | It's not supposed to work. |
| 16:31 | <AryehGregor> | Because the interface prototype you're using is not the prototype of the object you're talking about. |
| 16:31 | <timeless> | AryehGregor: but i don't think there's a convenient way to calculate window2 from foo |
| 16:31 | <AryehGregor> | No, there's no way at all. |
| 16:31 | <timeless> | which makes calculating InterfaceName hard |
| 16:31 | <AryehGregor> | I've been told that's a deliberate design feature. |
| 16:31 | <timeless> | and thus your sample code is essentially impossible to do correctly |
| 16:31 | <AryehGregor> | Well, not quite. |
| 16:31 | <AryehGregor> | You can always use Object.getPrototypeOf. |
| 16:31 | <timeless> | i think you can try to chase the prototype chain |
| 16:31 | <Ms2ger> | Or __proto__ :) |
| 16:31 | <timeless> | Ms2ger: shh |
| 16:32 | <AryehGregor> | I work with standards, __proto__ doesn't exist! |
| 16:32 | <AryehGregor> | I mean, it does, but I can't use it in tests, because it's not standard . . . |
| 16:32 | <AryehGregor> | Can we sneak Web ES requirements into HTML or something? |
| 16:32 | <timeless> | out-of-scope |
| 16:34 | <AryehGregor> | Maybe WebIDL can sneak them into the ES binding. |
| 16:34 | <AryehGregor> | Or, like, maybe ES could actually require them. |
| 16:34 | <AryehGregor> | srsly. |
| 16:35 | <Ms2ger> | Let's do that |
| 16:35 | <AryehGregor> | Nobody's seriously considering getting rid of __proto__, are they? |
| 16:35 | <AryehGregor> | So why isn't it in ES? |
| 16:35 | <Ms2ger> | Sure thing |
| 16:35 | <Ms2ger> | IIRC |
| 16:35 | <AryehGregor> | They are? |
| 16:35 | <timeless> | i thought es was trying to get rid of the ability to mutate __proto__ |
| 16:35 | <AryehGregor> | Oh, well, that I can understand. |
| 16:35 | <AryehGregor> | In ES5 the prototype is fixed at object creation time, right? |
| 16:35 | <AryehGregor> | That seems reasonable enough to me. |
| 16:35 | timeless | hasn't read it |
| 16:35 | <timeless> | but i think that's the goal |
| 16:36 | <AryehGregor> | Mutating it could cause all kinds of problems, although they're probably resolvable somehow. |
| 16:36 | <timeless> | it's really annoying from a hackery perspective |
| 16:36 | <Ms2ger> | I'd ask, but people are eating turkeys |
| 16:36 | <AryehGregor> | That was yesterday. |
| 16:36 | timeless | wants a turkey |
| 16:36 | <AryehGregor> | Unless it's leftover turkeys. |
| 16:36 | <timeless> | all i have is a bike w/ a flat tire |
| 16:36 | <AryehGregor> | That sounds less edible. |
| 16:36 | <timeless> | and some grease |
| 16:36 | <Ms2ger> | You could try the grease? |
| 16:36 | <timeless> | it isn't designed for cooking |
| 16:36 | <jgraham> | Why would you want a turkey |
| 16:36 | <timeless> | and i know where it's been |
| 16:36 | <jgraham> | They have little to nothing to recommend them |
| 16:36 | <timeless> | jgraham: it's edible and reminds me of home? |
| 16:37 | <jgraham> | Apart from being relatively cheap per unit mass |
| 16:37 | <timeless> | jgraham: ben franklin is little to nothing? |
| 16:37 | jgraham | doesn't know what ben franklin has to do with it but wouldn't eat slugs just because einstein liked them |
| 16:37 | <timeless> | ... For the Truth the Turkey is in Comparison a much more respectable Bird, and withal a true original Native of America . . . He is besides, though a little vain & silly, a Bird of Courage, and would not hesitate to attack a Grenadier of the British Guards who should presume to invade his Farm Yard with a red Coat on. |
| 16:37 | <jgraham> | (note: he didn't. probably) |
| 16:38 | <Ms2ger> | Maybe if you cook them by having lightning hit them? |
| 16:38 | <timeless> | what is the world coming to that people don't know his stance on turkeys? |
| 16:38 | <timeless> | http://en.wikipedia.org/wiki/Wild_Turkey#Benjamin_Franklin_and_the_US_national_bird |
| 16:39 | <timeless> | in context, this is what he had to say about the eagle: |
| 16:39 | <timeless> | ... For my own part I wish the Bald Eagle had not been chosen the Representative of our Country. He is a Bird of bad moral character. He does not get his Living honestly. You may have seen him perched on some dead Tree near the River, where, too lazy to fish for himself, he watches the Labour of the Fishing Hawk; and when that diligent Bird has at length taken a Fish, and is bearing it to his Nest |
| 16:39 | <timeless> | for the Support of his Mate and young Ones, the Bald Eagle pursues him and takes it from him. |
| 16:40 | <zewt> | sounds like a pretty apt characterization of our country |
| 16:40 | <timeless> | too true |
| 16:40 | <timeless> | so it was probably a fitting choice |
| 16:40 | <timeless> | but oh well |
| 16:40 | <timeless> | anyway, don't say you didn't ever learn anything about Franklin from #whatwg :) |
| 16:40 | <timeless> | (or Turkeys or Eagles for that matter) |
| 16:41 | <jarek> | http://www.w3.org/TR/css3-animations/ defines keyframes rule as: |
| 16:41 | <jarek> | '@keyframes' IDENT '{' keyframes-blocks '}'; |
| 16:41 | <jarek> | shouldn't there be STRING instead of IDENT? |
| 16:41 | <jgraham> | I think it would be a more apt characterisation of your country if the eagle sold the fishing hawk on the value of hard work as part of the "American Dream" |
| 16:41 | <timeless> | lol |
| 16:42 | <timeless> | jgraham: this is more about foreign afairs |
| 16:42 | <timeless> | i'm sure we have an example fitting your american dream somewhere |
| 16:43 | <timeless> | although the only thing that comes to mind is the praying mantis |
| 16:43 | <zewt> | jarek: might not matter here, but i assume you should be reading http://dev.w3.org/csswg/css3-animations/ |
| 16:43 | <zewt> | beware of "TR" in w3 URLs |
| 16:44 | <timeless> | yeah, T means "almost certainly out of date, this document published for lawyers, if you aren't one, steer clear" |
| 16:44 | <timeless> | s/T/TR/ |
| 16:44 | <jarek> | zewt: thanks, the dev spec uses IDENT |
| 16:44 | <jarek> | so maybe it uses different definitions of tokens? |
| 16:46 | <timeless> | IDENT comes from http://www.w3.org/TR/CSS2/grammar.html#scanner |
| 16:46 | timeless | thinks |
| 16:46 | timeless | grumbles sorry about the CSS2 ref |
| 16:48 | <timeless> | jarek: why would you want STRING instead of IDENT? |
| 16:48 | <timeless> | STRING requires one to have "..." or '...' |
| 16:48 | <timeless> | which is at least moderately annoying |
| 16:48 | <jarek> | timeless: yeah, there are many examples on the web that look like this: |
| 16:48 | <jarek> | @keyframes "myAnimation" {} |
| 16:49 | <timeless> | interesting |
| 16:49 | <jarek> | no, wait... actually I have found it only in the old spec :) |
| 16:49 | <timeless> | could i convince you to collect urls for them (possibly sorted by pagerank) and send them to www-style? |
| 16:49 | <timeless> | oh |
| 16:49 | <jarek> | ok, this makes sense now |
| 16:49 | <timeless> | so old content that hasn't evolved |
| 16:49 | <timeless> | yay |
| 16:49 | <timeless> | the dead web, that's where i come in |
| 16:50 | <timeless> | except i don't do much css archeology right now |
| 16:50 | <timeless> | @keyframes is too new to be buried and studied by archeologists :) |
| 16:51 | <timeless> | jarek: do they really use @keyframes and not @-vendor-keyframes? |
| 16:51 | <jarek> | timeless: I need to parse both of them |
| 16:52 | <timeless> | do the web a favor and don't support string |
| 16:53 | timeless | wishes browser vendors would HTTP POST /validator-feedback <payload> |
| 16:53 | <timeless> | "Dear site, this is an automated message listing content errors which resulted in things not working the way you expected" |
| 16:53 | <timeless> | ... |
| 16:54 | <timeless> | w3schools doesn't think ie has an -ms-keyframes, but css-tricks does |
| 16:55 | timeless | can't find any high ranked site that mentions the STRING form |
| 16:56 | <timeless> | how are you finding these bogus sites? |
| 16:56 | timeless | doesn't know how to ask google for such content |
| 16:58 | <jarek> | opening http://dev.w3.org/csswg/css3-box/ crashes Chrome |
| 17:05 | <jarek> | http://dev.w3.org/csswg/css3-grid/ says: |
| 17:05 | <jarek> | "This specification is not being actively maintained, and should not be used as a guide for implementations. It may be revived in the future, but for now should be considered obsolete." |
| 17:05 | <jarek> | I thought that Microsoft was pushing CSS grids very hard |
| 17:05 | <Ms2ger> | css3-grid is a Bertism, afaik |
| 17:06 | <jarek> | also, the last update is from 26 October 2011 and IE10 has already implemented it, why is it considered obsolete? |
| 17:12 | <astearns> | jarek: look at http://dev.w3.org/csswg/css3-grid-align/ |
| 17:13 | <jarek> | astearns: thanks |
| 17:21 | <jarek> | does http://dev.w3.org/csswg/ contain always the latest versions of the spec? |
| 17:22 | <astearns> | yes - that's where the editors check in |
| 17:25 | <jarek> | http://dev.w3.org/csswg/css3-ruby/ is older than http://www.w3.org/TR/css3-ruby/ |
| 17:26 | <jarek> | was it renamed to http://dev.w3.org/csswg/css3-writing-modes/? |
| 17:26 | <Ms2ger> | No |
| 17:26 | <Ms2ger> | css3ruby is a bit of an odd one out |
| 17:27 | <Ms2ger> | Also, the date on tr/ is the publication date; doesn't mean it has been changed from the ED |
| 17:35 | <zewt> | would be nice if they'd at least find a way to get google to stop returning TR results before ED |
| 17:35 | <zewt> | don't know if there's any way to do that short of blocking the TR from being indexed completely |
| 17:39 | <Ms2ger> | rel=canonical? |
| 17:44 | <jarek> | where is the spec that was previously under https://dvcs.w3.org/hg/FXTF/raw-file/tip/filters/publish/Filters.html ? |
| 17:44 | <jarek> | I mean the new CSS and SVG filters |
| 19:30 | <AryehGregor> | zewt, I filed a bug on HTML asking for exactly that. |
| 19:30 | <AryehGregor> | I thought it was FIXED, but seems not? |
| 19:31 | AryehGregor | can't find it |
| 19:33 | <AryehGregor> | Hixie, do you remember me ever filing a bug suggesting you add <meta name=robots content=noindex> to the TR/ drafts? |
| 19:34 | <AryehGregor> | I thought I did. |
| 21:23 | <Hixie> | AryehGregor: i fixed it, but we haven't republished since so it hasn't taken effect |
| 21:23 | <Hixie> | AryehGregor: my ping earlier was for some bug, i commented instead |
| 21:39 | <Ms2ger> | AryehGregor, yt? |