| 00:00 | <Hixie> | (safari gives me a more useful error message than chrome) |
| 00:00 | <othermaciej> | do you know which step of the test fails? |
| 00:01 | <Hixie> | let me get back to you in a minute |
| 00:02 | <Hixie> | ok yeah |
| 00:02 | <Hixie> | it's just that .source isn't yet supported |
| 00:02 | <Hixie> | using .ports[0] works fine |
| 00:05 | <Hixie> | ok now 003 is failing in opera |
| 00:07 | <Hixie> | bummer |
| 00:08 | <Hixie> | so webkit acts like the normative parts of the spec say, and opera acts like the tutorial parts of the spec say |
| 00:09 | <Hixie> | anyone other than opera or webkit planning on implementing SharedWorker? |
| 00:11 | <yroc> | Hixie: can you handle two conversations at once? |
| 00:11 | <a-ja> | https://bugzilla.mozilla.org/show_bug.cgi?id=643325 |
| 00:18 | <Hixie> | yroc: sure |
| 00:18 | <Hixie> | a-ja: thanks |
| 00:18 | <Hixie> | smaug____: yt? |
| 00:19 | <Hixie> | anyone with IE10 around? what does it say in http://www.hixie.ch/tests/adhoc/dom/workers/shared/000.html ? |
| 00:20 | <smaug____> | Hixie: pong |
| 00:20 | <smaug____> | ah, SharedWorkers |
| 00:20 | <smaug____> | I've tried to get bent to implement them |
| 00:20 | <smaug____> | let me ask |
| 00:20 | <Hixie> | smaug____: do you have an opinion on https://www.w3.org/Bugs/Public/show_bug.cgi?id=19491 ? the question is, should it be possible to connect to a sharedworker when you don't know the URL it was started with |
| 00:20 | <Hixie> | smaug____: (but you do know the name it was started with) |
| 00:21 | <yroc> | K. In light of recent discussion about <article>, what is your rationale for marking up blog comments with <article>? |
| 00:21 | <Hixie> | smaug____: right now the spec says both, opera follows the "urls don't have to match" thing the tutorial suggests, and webkit follows the "urls have to match" thing which the normative algorithm says |
| 00:21 | <yroc> | For one thing, blog comments don't seem to meet the def'n of "self-contained" (as article is defined in the spec) |
| 00:23 | <Hixie> | yroc: http://lists.w3.org/Archives/Public/public-whatwg-archive/2012Jan/0226.html discusses this |
| 00:23 | <Hixie> | yroc: let me know if that's not enough |
| 00:24 | <yroc> | Great, I'll have a look. Thanks. |
| 00:30 | <Hixie> | smaug____: let me know if you or bent have a different opinion, but fwiw, for now, i'm fixing the non-normative text to match the normative text |
| 00:30 | <Hixie> | smaug____: so url has to match |
| 00:30 | <smaug____> | Hixie: add a comment |
| 00:30 | <smaug____> | Hixie: and I agree with you |
| 00:31 | <smaug____> | s/add/I added/ |
| 00:31 | <Hixie> | awesome, thanks |
| 00:31 | <Hixie> | smaug____++ |
| 00:32 | <smaug____> | (obviously too late to type anything) |
| 00:33 | <Hixie> | wait... i totally misread the spec |
| 00:33 | <smaug____> | your spec ;) |
| 00:34 | <Hixie> | oh i see, i didn't misread, it's contradictory within the normative section |
| 00:34 | <Hixie> | step 5 contradicts step 7.1 |
| 00:34 | <Hixie> | bummer |
| 00:34 | <Hixie> | but that's ok |
| 00:35 | <Hixie> | the name allows you to have multiple instances of the same url |
| 00:35 | <Hixie> | ok |
| 00:35 | <Hixie> | don't mind me, ignored everything i said since i incremented you |
| 00:41 | <a-ja> | Hixie: ie10 says FAIL |
| 00:41 | <Hixie> | thanks |
| 00:49 | <yroc> | Hixie: "An article is just a |
| 00:49 | <yroc> | comment that has been hoisted to a more prominent position. |
| 00:49 | <yroc> | " I'm not exactly convinced of this. |
| 00:49 | <yroc> | I mean, I'm thinking a comment is a *reaction* to another piece of writing. |
| 00:50 | <zewt> | sounds like it's trying to reduce the whole internet to a set of comments of varying levels of complexity, heh |
| 00:51 | <yroc> | I'm thinking that "stand-alone" means you can read it and it makes sense in and of itself, without having to link it to anything else. |
| 00:51 | <a-ja> | yroc: never seen a comments feed? |
| 00:51 | <TabAtkins> | Never seen reddit? |
| 00:53 | <yroc> | Hixie had an example of reddit in his reference, so I have to say, yes :-) |
| 00:53 | <TabAtkins> | One blog I read, Less Wrong, uses a reskinned reddit codebase, and there the comments frequently *are* reparented into full articles. |
| 00:53 | <Hixie> | pretty much all writing is a reaction to another piece of writing |
| 00:54 | <Hixie> | whether it's a comment or not doesn't much affect that |
| 00:54 | <TabAtkins> | (Often lightly edited before they hit the homepage to give them more structure, but still, just viewing the comment thread rooted on them is very much like a blog post.) |
| 00:54 | <yroc> | But, I'm just talking in general. A comment like "LOL" or an example in the spec: "Yeah! Especially when talking about your lobbyist friends!" |
| 00:54 | <TabAtkins> | Have you seen blogs? |
| 00:54 | <Hixie> | yroc: there are articles whose contents are pretty short too... twitter is an extreme example of that, as are many tumblrs |
| 00:54 | <TabAtkins> | Those are basically the typical microblog. ^_^ |
| 00:55 | <yroc> | You can't decipher these without reference to the "main" article. |
| 00:55 | <Hixie> | the same applies to many things that nobody disagrees are articles |
| 00:55 | <TabAtkins> | Ooh, Tumblr's a good example - a lot of posting is just a quote of someone else's post with a tiny amount of commentary. "LOL" is a common one. Does the mere fact that it's *presented* like a top-level post, despite being for all intents and purposes a comment, change things? |
| 00:56 | <TabAtkins> | Or should tumblr people have to distinguish between "real" posts and reblogs, with the latter not including an <article> at all? |
| 00:56 | <TabAtkins> | (And of course, some reblogs are indeed full posts on their own.) |
| 00:57 | <Hixie> | imho the distinction between "Article" and "Comment" is a distinction from a bygone era, one in which newspapers were written by Journalists and the lowly citizenry had no outlet |
| 00:57 | <Hixie> | nowadays there's nothing to distinguish an article from a comment, and often the comments have higher "value" |
| 00:58 | <yroc> | OK, *but* what does "self-contained" in the spec mean? (I apologize if you think you already answered that) |
| 00:59 | <Hixie> | it means that it was written as one thing |
| 00:59 | <Hixie> | for example, if you just take the words "example, if" from this sentence, they're not self-contained, while the whole thing is. |
| 00:59 | <Hixie> | (if you have a better way to put it, please let me know) |
| 01:00 | <Hixie> | "complete" might be a better word |
| 01:01 | <TabAtkins> | It's just meant to distinguish it from <section>, which often contains things that aren't self-contained, and only make sense in conjunction with the whole article. |
| 01:01 | <Hixie> | yeah |
| 01:02 | <TabAtkins> | (While a comment might not make sense except in relation to the article it's responding to, the connection is of a qualitatively different and weaker kind than that of a section to its article. |
| 01:02 | <TabAtkins> | ) |
| 01:02 | <Hixie> | i've changed the spec from "self-contained" to "complete, or self-contained," |
| 01:04 | <yroc> | Can you give a simple example of an incomplete <section> of content? |
| 01:04 | <Hixie> | sure, any of the sections in the spec |
| 01:05 | <yroc> | Are you saying that <section> is by definition incomplete? |
| 01:05 | <yroc> | An incomplete unit of content, that is. |
| 01:07 | <yroc> | Like, <h1>Granny Smith</h1><p>These juicy, green apples make a great filling for |
| 01:07 | <yroc> | apple pies.</p> |
| 01:07 | <yroc> | This is incomplete content? |
| 01:08 | <Hixie> | well, the examples in the spec for <section> and <article> are mostly a bit too contrived to really comment on, because real sections are pages long |
| 01:08 | <Hixie> | i meant hte actual spec sections |
| 01:08 | <Hixie> | not the example ones |
| 01:08 | <yroc> | Ah. |
| 01:08 | <Hixie> | like, section 11.2.3 |
| 01:09 | <Hixie> | doesn't make sense without 11.2.2 etc |
| 01:09 | <Hixie> | but mostly the real answer is "it depends on author intent" |
| 01:10 | <Hixie> | one could imagine a book with a "Granny Smith" chapter that just said "These juicy, green apples make a great filling for apple pies.", and then it'd be a section because there'd be lots of other chapters on (maybe) other kinds of apples |
| 01:11 | <Hixie> | but then one could imagine a tweet or reddit comment or tumblr post or newspaper classified ad that just said "Granny Smith. These juicy, green apples make a great filling for apple pies.", and then it'd be an article because that was the whole thing. |
| 01:11 | <yroc> | OK, what you're saying now is making sense to me... |
| 01:12 | <yroc> | Comments aren't "part" of what they're responding to. |
| 01:13 | <yroc> | This is what you mean by "complete" |
| 01:13 | <Hixie> | right |
| 01:14 | <a-ja> | so, <section>s are incomplete til they're in W3C rec, and modified every few years; then they become <article>s? |
| 01:15 | a-ja | ducks |
| 01:15 | <Hixie> | basically. |
| 01:18 | <yroc> | But, one thing that bothers me now. The spec says that the composition could *in principle* be independently distributable or reusable. Doesn't that make a book chapter (and *lots* of other potential <section>s) in fact <article>s (by definition)? |
| 01:19 | <yroc> | i.e., Even if they're part of a whole, in principle they don't have to be. |
| 01:19 | <Hixie> | if you were to write a chapter in such a way that you could publish it independently, then yeah, it could be an <article> |
| 01:19 | <Hixie> | but then you could also write an article in such a way that you could put it in a sidebar of another article, and then it's also an <aside>. :-) |
| 01:20 | GPHemsley | wants to comment on this discussion but doesn't feel like reading the scrollback to determine the context. |
| 01:21 | <yroc> | Right, but the terminology "in principle" would, it seems to me, bias the use of <article> over <section> |
| 01:21 | <Hixie> | yroc: yeah... i dunno what better terminology to use |
| 01:21 | <Hixie> | GPHemsley: man, that's never stopped anyone else |
| 01:22 | <GPHemsley> | I try not to contribute to the problems, only the solutions ;) |
| 01:22 | <Hixie> | :-P |
| 01:23 | <GPHemsley> | I don't see what the big deal is, though. I would think <section> would be pretty clear. |
| 01:23 | <Hixie> | you'd think |
| 01:23 | <GPHemsley> | Perhaps the real problem is that people still are not separating semantics from presentation. |
| 01:24 | <GPHemsley> | A section has a header. It would show up in a table of contents. Fairly straightforward. |
| 01:24 | <Hixie> | GPHemsley: so does an article |
| 01:24 | <yroc> | GPHemsley: Do you even know what this discussion is about? |
| 01:24 | <Hixie> | and not all sections have headers |
| 01:24 | <GPHemsley> | yroc: Nope. |
| 01:25 | <Hixie> | not all docs with sections have tables of contents |
| 01:25 | <yroc> | GPH: thought not. |
| 01:25 | <GPHemsley> | Hixie: *theoretical* tables of contents ;) |
| 01:25 | <GPHemsley> | yroc: I thought I made that pretty clear. |
| 01:26 | <yroc> | GPH: I guess I don't know you well enough yet ;-) |
| 01:26 | <Hixie> | GPHemsley: so when do you use article? every use of article can have a theoretical table of contents too :-) |
| 01:26 | <GPHemsley> | Hixie: I do get the sense that <article> is a bit of a misnomer. |
| 01:27 | <Hixie> | trying to fit the semantics of any element into a one word short name is a challenge :-) |
| 01:27 | <GPHemsley> | Hixie: Also, that was kinda the point I was making. <article> has a table of contents; <section> adds an item to that table of contents. |
| 01:29 | <a-ja> | uh oh....i just coded a TOC as a section |
| 01:29 | <Hixie> | a magazine might have several <article>s and a table of contents listing them |
| 01:29 | <Hixie> | in fact it's pretty common in printed magazines |
| 01:29 | <a-ja> | with role=directory, of course :/ |
| 01:29 | <GPHemsley> | Hixie: Yeah, it all depends on your field of view, macro/micro. |
| 01:30 | <GPHemsley> | a-ja: That's not necessarily outside of my definition. |
| 01:30 | <a-ja> | <section> adds an item to that table of contents. <== causes a loop |
| 01:30 | <a-ja> | heh |
| 01:30 | <yroc> | Hixie: How about something like, "self-contained or complete composition...that might potentially be independently distributed or reused... |
| 01:31 | <GPHemsley> | a-ja: If you have pages before the TOC, it's completely reasonable to include the TOC in the TOC. ;) |
| 01:31 | <GPHemsley> | a-ja: Note that the TOC itself only has the header, not the entire contents. |
| 01:32 | <GPHemsley> | It basically comes down to what you peg as an article. Everything is relative to that. And you can have articles within articles. |
| 01:32 | <GPHemsley> | (Though it might be weird to have articles within sections.) |
| 01:33 | <Hixie> | yroc: "might potentially" and "in principle" have the same problem, i think |
| 01:33 | <Hixie> | right, gotta go |
| 01:33 | <Hixie> | feel free to mail the list if you come up with a magic solution to all our problems while i'm afk :-) |
| 01:33 | GPHemsley | still doesn't even know what the problems are. :P |
| 01:34 | <a-ja> | GPHemsley: not at all....have seen plenty of comment <articles> in a comments <section> |
| 01:34 | <GPHemsley> | a-ja: Ah, I suppose that would be the exception. |
| 01:35 | <GPHemsley> | a-ja: Perhaps like when you combine multiple books into a single book, but one of those books has short stories in it. |
| 01:35 | <a-ja> | argh! |
| 01:35 | <GPHemsley> | although actually, that might be an article, not a section |
| 01:35 | <GPHemsley> | hmm |
| 01:36 | a-ja | goes back to scratching head over whether breadcrumbs and/or toc belong within in a <main> |
| 01:37 | <GPHemsley> | a-ja: Wouldn't they go in <header>? |
| 01:37 | <GPHemsley> | a-ja: Well, breadcrumbs would, at least. TOC is probably more debatable. |
| 01:38 | <yroc> | Oh, I saw <main> now in W3C nightly... Is <main> coming to the spec? |
| 01:49 | <GPHemsley> | Hmm... my definitions and explanations of <article> and <section> appear to match up exactly with what the spec says |
| 01:49 | <GPHemsley> | that was, in some ways, a complete coincidence |
| 01:49 | <GPHemsley> | but a good one, I think |
| 01:51 | <GPHemsley> | (not just for my own personal sake) |
| 01:51 | <GPHemsley> | Hixie: FYI, the <section> section uses "programme" instead of "program" |
| 01:52 | <GPHemsley> | Actually, that whole example is somewhat weird. Why does the description of the ceremony use <p> instead of <ol>? |
| 02:03 | <GPHemsley> | When was the last time the W3C Validator's tree/outline generator worked? |
| 02:07 | <a-ja> | GPHemsley: been a while, ime |
| 02:08 | <GPHemsley> | That was one of my favorite features, even back before anyone cared. |
| 02:08 | <a-ja> | semantic data extractor |
| 02:10 | <a-ja> | unless you use html5-style titling |
| 02:13 | <GPHemsley> | What was that site that's like pastebin, but with live preview? |
| 02:14 | <moo-_-> | GPHemsley: jsfiddle.net |
| 02:14 | <GPHemsley> | moo-_-: Doesn't look like the one I had in mind, but it'll do. Thanks. |
| 02:14 | <moo-_-> | np |
| 02:16 | <GPHemsley> | moo-_-: Err, actually, it seems to only allow portions of a page, and it doesn't use the browser default CSS. Is there another one? |
| 02:17 | <GPHemsley> | ah, jsbin.com |
| 02:17 | <GPHemsley> | I should've known |
| 03:43 | <Hixie> | GPHemsley: spec is en-us-x-hixie, in which "program" is a computer program and "programme" is all the other programmes. :-P |
| 03:43 | <Hixie> | GPHemsley: btw, i apparently gave that rod guy an account when he asked for one last march |
| 03:44 | GPHemsley | shrugs x2 |
| 03:44 | <Hixie> | (totally forgot until just now where i happened to search for his name in gmail) |
| 03:44 | <Hixie> | he had [SEC-UNCLASSIFIED] in the subject line then, too... |
| 03:46 | <GPHemsley> | Australian government archives |
| 03:47 | <GPHemsley> | serious stuff :) |
| 03:48 | <a-ja1> | maybe |
| 03:48 | a-ja1 | worked for a US govt archives once |
| 03:48 | <zewt> | us.zip |
| 03:50 | <a-ja> | us army records center |
| 03:50 | <a-ja> | b4 email was commonplace |
| 05:59 | <MikeSmith> | to borrow a movie-industry phrase I learned from watching Argo, I guess we could say that Web Intents is "in turnaround" now |
| 05:59 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-web-intents/2013Jan/0015.html |
| 06:48 | <hsivonen> | manu-db: I don't think the goal of EME is to reduce piracy. I think the goal is to make Hollywood exects comfortable so that they let Netflix deliver content with EME-bootstrapped DRM instead of Silverlight. |
| 07:26 | <Hixie> | it better not be to reduce piracy, since it'll do precisely zilch to curb piracy |
| 09:14 | <Stevef> | a-ja: re main breadcrumbs no as they are site navigation which is usually found across pages, if the TOC is the main content of the page yes, if you think this needs to be made clearer in the spec feel to file a bug https://www.w3.org/Bugs/Public/enter_bug.cgi?product=HTML%20WG&component=HTML5%20spec&priority=P3 |
| 09:39 | <hsivonen> | Hixie: dunno about the alleged infinite loop |
| 09:39 | <hsivonen> | Hixie: the www-style thread turns to image optimization concerns later |
| 09:50 | <Stevef> | hixie: seems like this could also be an article,or any part of a conversation or dialogue |
| 09:50 | <Stevef> | or this |
| 10:44 | <a-ja> | Stevef: the graduation programme example certainly has a toc (i.e. intra-page nav) within <main> |
| 10:44 | <Stevef> | a-ja: yes, thats intra page |
| 10:54 | <a-ja> | Stevef: i suggest being explicit as possible, so <main> doesn't get abused into being a substitute for a center-column wrapper div |
| 10:58 | <Stevef> | a-ja: i have tried to be explicit any feedback on how to be more specific appreciated :-) |
| 11:01 | <a-ja> | well, "The main content section of a document includes content that is unique to that document and excludes content that is repeated across a set of documents such as site navigation links, copyright information, site logos and banners and search forms (unless the document or applications main function is that of a search form)." could have breadcrumbs specifically added to the exclusion list |
| 11:07 | <a-ja> | and the graduation pgm could have explicit reference to the nav before main being to other pages, vs. the nav within main being exclusively intra-page nav |
| 11:08 | <a-ja> | iow, be explicit in surrounding text about what's implicit in the example |
| 11:19 | <Stevef> | a-ja: thanks! |
| 11:21 | <a-ja> | Stevef, i take it the lack of header nav /ul is intentional |
| 11:21 | <a-ja> | and the H2 in caps |
| 11:21 | <Stevef> | yes |
| 11:22 | <a-ja> | and use of ul vs ol ? |
| 11:23 | <Stevef> | a-ja:not intentional will review |
| 11:26 | <a-ja> | Stevef: unrelated directly to <main> doc....perhaps. should intra-page toc have role=directory on the ul or ol ? (seem to recall that from WAI docs) |
| 11:29 | <kamathln> | sicking: Hi, are the race conditions in IndexedDb (even handler assignment) still around? Where can I follow the discussions on IndexedDb ? |
| 11:30 | <SimonSapin> | hsivonen: re MQ vs client hints: what does "art direction" mean? |
| 11:34 | <hsivonen> | SimonSapin: that you change image shape and size so that a smaller image shows only an area of interest chosen by a human to be the most important part of the image |
| 11:34 | <SimonSapin> | I see, thanks |
| 11:35 | <Stevef> | a-ja: sounds reasonable but i don't think it makes any difference practically as its mapped to a list role in APis |
| 11:40 | <kamathln> | sicking: got it : https://dvcs.w3.org/hg/IndexedDB/raw-file/tip/IndexedDB%20Disposition%20of%20Comments.html , but is that the last call/ |
| 11:42 | <a-ja> | Stevef: may follow up on that in WAI doc (WAI1?), in light of no <dir> element anymore |
| 12:16 | <annevk> | How do I reference a pull request in a commit? |
| 12:21 | <kamathln> | Hi, is there any bot I can use to "tell" someone who has left the channel? |
| 12:22 | <hsivonen> | kamathln: no. but you try to say stuff so that it gets logged and hope the person you wants to reach reads logs |
| 12:22 | <hsivonen> | (public logs on the Web that is. I don't mean logs of their own absent IRC client) |
| 12:23 | <kamathln> | I can't "hope", I want to ensure ;) |
| 12:23 | <hsivonen> | kamathln: there's no way to ensure on this channel |
| 12:24 | <kamathln> | hsivonen: thanks anyways .. almost everyone has an email, I will dig for and use that ;) |
| 12:25 | <hsivonen> | kamathln: the probability of people responding to logs may be higher than the probability they respond to email |
| 12:25 | <hsivonen> | (not joking) |
| 12:25 | <kamathln> | hsivonen: hmm.. you are right .. |
| 12:26 | <kamathln> | sicking: I made a terrible sin (by netizen standards) by not googling thoroughly before asking a qn.. saying sorry for that :) |
| 12:26 | <kamathln> | hsivonen: ^^ :P |
| 12:28 | <smaug____> | there seems to be MemoServ |
| 12:28 | <kamathln> | smaug____: thanks! lemme try |
| 12:30 | <kamathln> | smaug____: return SUCCESS; |
| 12:30 | <kamathln> | smaug____: thanks :) |
| 12:31 | <smaug____> | lunch time |
| 12:32 | <smaug____> | kamathln: and it worked :) |
| 12:32 | <smaug____> | bbl |
| 12:46 | <annevk> | hsivonen: btw, https://www.w3.org/Bugs/Public/show_bug.cgi?id=20127 (<template> and shadow DOM affecting DOM algorithms) |
| 12:47 | <hsivonen> | annevk: yay edge cases |
| 12:47 | <annevk> | I'm kinda surprised nobody thought of that thus far |
| 12:48 | <hsivonen> | so it looks like I always forget to land htmlparser patches after landing the corresponding Gecko patches |
| 12:49 | <MikeSmith> | hsivonen: oh |
| 12:49 | <MikeSmith> | you forgot this time too? |
| 12:49 | <hsivonen> | MikeSmith: I'm in the process of landing a bunch of stuff |
| 12:50 | <hsivonen> | MikeSmith: sorry about the delay |
| 12:50 | <hsivonen> | so I guess *this* time I did not forget |
| 13:15 | <hsivonen> | MikeSmith: I landed the pending htmlparser patches |
| 13:16 | <hsivonen> | MikeSmith: (and I have not forgotten about ambiguous ampersands. I need to discuss it some more with Hixie. maybe raise it in a WG again.) |
| 13:17 | <MikeSmith> | hsivonen: fwiw I personally have no strong opinions on that |
| 13:18 | <MikeSmith> | I wrote the patch in the interest of clearing out some validator bug reports :) |
| 13:18 | <MikeSmith> | because there are several open reports about that one |
| 13:21 | <MikeSmith> | I did try a bit to consider what would be the best validator behavior from the user perspective but I doesn't seem to me there's a real clear answer |
| 13:21 | <MikeSmith> | seems like a judgement-call case |
| 13:21 | <hsivonen> | hmm. is ambiguous ampersands on the band name list yet? |
| 13:21 | <hsivonen> | :-( |
| 13:21 | <MikeSmith> | heh |
| 13:22 | <MikeSmith> | if not it should be |
| 13:23 | <hsivonen> | it is now |
| 13:23 | <MikeSmith> | excellent |
| 13:24 | <MikeSmith> | Ambiguous Ampersand would be a fitting name for a character in a John Waters movie |
| 13:25 | <hsivonen> | so I copied the dir of eclipse itself, the workspace and ~/.eclipse to another computer and launched eclipse. |
| 13:25 | <hsivonen> | and it didn't open up in the same state as on the computer I copied from |
| 13:25 | <hsivonen> | what did I fail to copy? |
| 13:27 | <MikeSmith> | hsivonen: if it were on a Mac i'd say maybe there's something in ~/Library/Application Support/ |
| 13:28 | <MikeSmith> | hmm no not there |
| 13:29 | <MikeSmith> | but on my Mac I do see things in ~/Documents/workspace/.metadata/.plugins/org.eclipse.core.* |
| 13:30 | <MikeSmith> | hmm yeah, e.g., ~/Documents/workspace/.metadata/.plugins/org.eclipse.core.resources/.projects/htmlparser/org.eclipse.jdt.core |
| 13:32 | <MikeSmith> | I guess on Mac you'd want to copy over all of ~/Documents/workspace |
| 13:32 | <MikeSmith> | no what the equivalent is on other OSes |
| 13:33 | <hsivonen> | thanks. problem solved: I forgot to make a symlink I had for the workspace and I was referring to the workspace via the symlink |
| 13:35 | <MikeSmith> | ah OK. oh and I see now you'd already said earlier above that you'd copied the workspace so nothing that I said was relevant anyway :) |
| 14:16 | <manu-db> | hsivonen: re: the purpose of EME - yes, that's my reading on why it exists as well. However, that's not stated in the spec and that's not a technical goal. I'm just trying to help them create a more robust technical goal for the spec. |
| 14:17 | <manu-db> | The EME spec spends a great deal of time weasel-wording its way around what it is actually doing. |
| 14:18 | <manu-db> | "We're not creating a DRM system, we're creating a protection system... ClearKey is not meant to be used in production, it's just for testing purposes... oh, and browsers are required to implement it, etc." |
| 14:21 | <manu-db> | The whole spec is just a tad too disingenuous for my taste... I'd rather the HTML Media WG publish a spec called WebDRM that requires the browser vendors to agree on a DRM plugin standard and how it works, than do what the EME spec does right now, which is establish most everything as out of scope and provide ClearKey (which provides no real protection for content, but requires browser... |
| 14:21 | <manu-db> | ...manufacturers to implement it) |
| 14:22 | <manu-db> | at least that way Mozilla has a chance of not getting skewered with royalties for whatever CDM scheme becomes popular. |
| 16:09 | <Ms2ger> | MikeSmith, any news on w3c-test.org replication, btw? |
| 16:10 | <MikeSmith> | Ms2ger: no, sorry |
| 16:10 | <MikeSmith> | i will make time this week |
| 16:32 | <annevk> | Ms2ger: fyi: https://www.w3.org/Bugs/Public/show_bug.cgi?id=20127 |
| 16:33 | <annevk> | Ms2ger: will prolly fix that soonish |
| 16:35 | <Ms2ger> | Great :) |
| 16:36 | <dglazkov> | good morning, Whatwg! |
| 16:40 | <annevk> | I've been wondering what I have done and regret in the context of the web platform. So far I have CORS header naming and overall complexity. |
| 16:41 | <annevk> | But then that's one of the few things I actually introduced to the web platform as lots of my work has been specifying old relics. |
| 17:30 | jwalden | supposes the things he regrets are stuff he tested before implementation and missed the issues that arose later in specification, but were argued too late to change |
| 17:31 | <jwalden> | like [0].indexOf(-0) being 0, or [NaN].indexOf(NaN) being -1 (assuming I remember both cases correctly enough) |
| 17:32 | <jwalden> | I still think we could change those, but we have some severely compat-crazy people in es-discuss at times :-\ |
| 17:32 | <jwalden> | not to mention es-discuss politics |
| 17:33 | <jwalden> | "not to mention" |
| 17:33 | <annevk> | Can't you sneak it into Gecko? |
| 17:34 | <jwalden> | well, that would have been past the module owner, at one time, and brendan was one of the compat-crazy people IMO :-) |
| 17:35 | <jwalden> | and these days, if it ain't in a spec, roughly, we don't change |
| 17:35 | <jwalden> | or in a draft that pretends to slightly approach spec-ness |
| 17:36 | <jwalden> | probably my biggest grumble about many SpiderMonkey extensions is how under-thought out they are in terms of the actual semantics implemented |
| 17:36 | <jwalden> | too much hand-waviness, not enough actual algorithms and specific syntax addition descriptions and stuff |
| 17:36 | <annevk> | That does sound fairly political. I like to believe there's always room for experiments in HTML/DOM/CSS, but maybe that's not true either. |
| 17:37 | <jwalden> | yeah, I think we don't do enough experimentation, especially with a rapid-release thing that makes experimentation lower-cost |
| 17:38 | <jwalden> | pretty much any plausible change should be something we can ship in a nightly for a little bit to get basic, if incomplete, information about web compatibility |
| 17:38 | <jwalden> | although, it's surprising just how many things we think are negligible based on nightly/alpha/beta feedback, only to find them rather worse when they get in a release |
| 17:38 | <jwalden> | nightly/alpha/beta users don't cover the site gamut as well as one might expect |
| 17:39 | <jwalden> | given the number of such testers we have, I'd have thought simple law of large numbers and all |
| 17:41 | <annevk> | I guess there might be quite a few intranet deployments these days that will not test with nightlies. |
| 17:41 | <annevk> | I only have FirefoxNightly.app at the moment myself |
| 17:41 | <annevk> | Same for Chrome |
| 17:42 | <jwalden> | the issues with seeming-lesser import that I've noticed as being worse at release scale haven't been intranet things (although I don't doubt there are some I just haven't seen) |
| 17:43 | <annevk> | I remember hsivonen battling router software |
| 17:49 | <jsbell> | sicking: I was talking to the MSFT folks last week about the IndexedDB spec. Do you have any objection to eliminating the speclets/build step and just having one document? |
| 17:57 | <sicking> | jsbell: not at all! |
| 17:59 | <jsbell> | sicking: okay. we did an (offline) edit pass together for editorial nits (about 10 of the issues). I think the plan is that Eliot will get those in, then I'll tackle updating to the new ReSpec. Time permitting all around. |
| 18:06 | <sicking> | jsbell: awesome |
| 18:07 | <jsbell> | sicking: yeah. I was sitting with Israel and Eliot so gained some insights. :) |
| 18:50 | <MikeSmith> | Matthew Kaufman from MS/Skype right now filing a barrage of bugs against the WebRTC spec |
| 18:50 | <MikeSmith> | http://lists.w3.org/Archives/Public/public-webrtc/2013Jan/ |
| 18:50 | <MikeSmith> | normally when you see an implementor do that, it's good sign |
| 18:51 | <MikeSmith> | sign that they're interested in implementing the spec but that there are some problems that need fixing |
| 18:51 | <MikeSmith> | I hope that's the case here |
| 18:52 | <Ms2ger> | Do you know if they're implementing from the spec rather than from Google's codedump? |
| 18:52 | <MikeSmith> | I don't know that they're implementing at all from anything at this point |
| 18:53 | <MikeSmith> | but they are not bad bugs |
| 18:54 | <MikeSmith> | a lot of them are just "XXX not specified" |
| 18:54 | <MikeSmith> | which is true |
| 18:54 | <MikeSmith> | I guess you really could not implement from the spec if all that stuff is not specified |
| 18:56 | <MikeSmith> | so it could be instead that the purpose of him raising the bugs is just to point out, this spec is not implementable in its current form because here's all this stuff it doesn'T even specify |
| 18:57 | <othermaciej> | MikeSmith: I’d guess Microsoft still wants to push for their CU-RTC thing instead |
| 18:58 | <MikeSmith> | othermaciej: from their public comments so far that would sure seem to be the case |
| 18:58 | <MikeSmith> | but the WebRTC WG pretty soundly rejected CU-RTC after quite a lot discussion about the details |
| 19:00 | <othermaciej> | I need to get some input from people at Apple who would actually understand the technical details |
| 19:01 | <othermaciej> | it kinda seemed to me like Microsoft’s criticism of relying on SDP might be valid, but I wasn’t able to evaluate for myself |
| 19:03 | <MikeSmith> | othermaciej: my understanding of the SDP choice is that the group made it with full knowledge that it introduced some other problems but that the benefits it brings to app developers are worth the tradeoff |
| 19:04 | <othermaciej> | MikeSmith: I’m also not able to evaluate whether those benefits are worth it |
| 19:04 | <MikeSmith> | yeah, not that I actually understand any of this stuff |
| 19:27 | <hober> | Hixie: how's it going re: https://www.w3.org/Bugs/Public/show_bug.cgi?id=20212#c11 ? |
| 19:28 | <Hixie> | haven't looked at it since that comment |
| 19:29 | <hober> | ok |
| 19:58 | <esprehn_> | TabAtkins: By default a background image should be inside the borders? |
| 19:59 | <esprehn_> | TabAtkins: specifically why is the radial gradient under the borders on http://jsfiddle.net/Shsxv/1/ ? |
| 20:01 | <Hixie> | by default the background goes to outside the borders |
| 20:01 | <Hixie> | the border is over the background |
| 20:01 | <Hixie> | unless background-origin is set to something else |
| 20:03 | <Hixie> | sorry, background-clip |
| 20:03 | <Hixie> | not -origin |
| 20:03 | <Hixie> | http://dev.w3.org/csswg/css3-background/#the-background-clip |
| 20:03 | <esprehn_> | if I do background-image: url(cat.png); border: 100px solid red; the borders aren't over the cat though |
| 20:04 | <esprehn_> | the background color is under the borders, but the image is not |
| 20:04 | <esprehn_> | Hixie: what's notable is that a gradient is specified as a background-image, but behaves as a background-color |
| 20:06 | <Hixie> | esprehn_: set background-repeat:repeat and it'll go under the border |
| 20:06 | <Hixie> | oh wait |
| 20:06 | <Hixie> | if it's solid red you won't see it |
| 20:06 | <Hixie> | set it to dotted red |
| 20:06 | <esprehn_> | ah, so the cat is inside the border by default, and not repeated under the borders |
| 20:07 | <Hixie> | it should be repeated under the borders |
| 20:07 | <Hixie> | but it's positioned relative to the padding box |
| 20:08 | <Hixie> | http://software.hixie.ch/utilities/js/live-dom-viewer/saved/2075 |
| 20:13 | <esprehn_> | Hixie: perfect, thanks |
| 21:11 | <esprehn_> | Hixie: more fundamentally no other language works like this |
| 21:11 | <esprehn_> | Hixie: new Socket() does not connect the socket in _any_ language that I'm aware of |
| 21:13 | <esprehn_> | Hixie: but the reasons are many, if your constructor is a verb your only extension point is the constructor |
| 21:14 | <esprehn_> | and the only error return is an exception |
| 21:16 | <esprehn_> | it's standard OO design |
| 21:17 | <esprehn_> | it's weird to be discussing this, it's a question of why is JS a special snowflake :) |
| 22:09 | <Hixie> | esprehn_: there are plenty of APIs where creating the object has side-effects |
| 22:09 | <Hixie> | esprehn_: e.g. IO::Socket::INET in perl opens the socket |
| 22:09 | <Hixie> | esprehn_: when you call the constructor |
| 22:11 | <Hixie> | esprehn_: Java's Socket API seems to do the same, when you call its constructor with an IP and port |
| 22:11 | <Hixie> | java.net.Socket, that is |
| 22:16 | <Hixie> | esprehn_: as far as i can tell, the common lisp equivalent has a function to which you pass the address and port and from which you receive a live connected socket |
| 22:16 | <Hixie> | in fact it'd be hard to see how a pure-functional system could do anything _but_ return a live connected object |
| 22:17 | <Hixie> | not that doing I/O in a pure-functional system is something i'd recommend for people who want to remain sane |
| 22:18 | <jgraham> | Well a *pure* functional system doesn't have IO by definition |
| 22:18 | <Hixie> | right |
| 22:18 | <jgraham> | (python's SMTP lib also connects in the constructor, to pick a random example) |
| 22:19 | <Hixie> | it's a pretty common pattern |
| 22:19 | <Hixie> | imho it's a sensible one, though it certainly doesn't fit perfectly with e.g. how C++ handled exceptions in constructors |
| 22:20 | <Hixie> | handles |
| 22:20 | <Hixie> | so there are certain languages where it's probably a bad pattern |
| 22:20 | <Hixie> | (not JS) |
| 22:31 | <esprehn_> | Hixie: heh, seems you're right |
| 22:31 | <esprehn_> | Hixie: but Java does provide a way to create a socket that's not connected |
| 22:31 | <esprehn_> | Hixie: there's no way for Gmail to have a single Notification instance for "New mail available" and just call show() on it |
| 22:31 | <esprehn_> | like <dialog> |
| 22:32 | <Hixie> | esprehn_: sure there is. Just create an object with a show() method that creates the notification. |
| 22:33 | <Hixie> | esprehn_: i mean, it won't be a native Notification object, but why would that matter |
| 22:36 | <esprehn_> | but people do reuse XHRs |
| 22:37 | <esprehn_> | I mean I'll concede that clearly other people disagree with me :) |
| 22:38 | <Hixie> | that people reuse XHRs is a source of lots of problems, and is IMHO a terrible mistake |
| 22:39 | <Hixie> | but in practice XHR's design stems from the fact that it used to be an ActiveX object, which is to say a COM object, and COM objects have a lot in common with C++ objects |
| 22:39 | <Hixie> | and using C++ design in JS isn't good practice IMHO :-) |
| 22:40 | <sicking> | yup |
| 22:47 | <esprehn_> | Hixie: perhaps, I do think you're overly optimistic about what happens inside the JS VMs |
| 22:48 | <esprehn_> | no one has predictive allocation, or has been working on it that I know of |
| 22:48 | <esprehn_> | or magical behind your back free() |
| 22:49 | <esprehn_> | I guess we'll catch up with lisp eventually |
| 22:52 | <Hixie> | the point is just to allow implementations to do that |
| 22:55 | <Hixie> | anyone think of a better term than "outlinee" for "the element for which an outline is being created"? |
| 23:00 | <Philip`> | "outline target"? |
| 23:01 | <Philip`> | (Terms that look like typos of other terms are probably not a fantastic idea) |
| 23:13 | <esprehn_> | Hixie: outlined element |