| 00:00 | <foolip_> | me neither, but the filter graph API *looks* way too complicated |
| 00:00 | <doublec> | yes |
| 00:01 | <foolip_> | maybe the right solution is WebAL or something |
| 00:01 | <doublec> | there's also been discussion on MusicML and supporting other music bsaed things which could become a distraction |
| 00:01 | <foolip_> | surely browsers shouldn't support MusicML natively? |
| 00:01 | <doublec> | I'm not opposed to a WebGL like approach where we wrap OpenSL/OpenAL or something like that |
| 00:01 | <doublec> | and let people build JS API's on top |
| 00:02 | <foolip_> | doublec, appart from the fact that I've never even looked at OpenSL or OpenAL, I'm inclined to agree |
| 00:03 | <foolip_> | I should sleep :) |
| 00:58 | <webr3> | annevk, any specific reason there isn't first class for json in xhr2, as in responseObject? |
| 00:59 | <TabAtkins> | json should be a responseType, I think. |
| 01:00 | <gsnedders> | responseType is just the MIME type, which is totally different… |
| 01:01 | <gsnedders> | Or am I misremembering things? |
| 01:01 | <TabAtkins> | responseType takes a string or enum or whatever indicating what type the value in .response should be. |
| 01:01 | <gsnedders> | Okay, I am misremembering things :) |
| 01:02 | gsnedders | just got home from concert and is fueled by awesomeness right now |
| 01:02 | <TabAtkins> | Sick Puppies? |
| 01:02 | <gsnedders> | Yeah. |
| 01:02 | <TabAtkins> | Awesome. |
| 01:02 | <TabAtkins> | We flew on the same plane as them after Buzzfest. |
| 01:03 | <gsnedders> | The thing is here they're not very well known, almost no promotion from record label, almost entirely word-of-mouth. |
| 01:03 | <gsnedders> | Totally different to the US. |
| 01:03 | <webr3> | ahh yeah i remember that convo on the list, was looking at wrong draft - in that case a responseType of "object" would be good.. (or such like) |
| 01:03 | <webr3> | ty |
| 01:04 | <gsnedders> | So it was a fairly small venue, maybe 100–200 people, but so full of energy. |
| 01:04 | <TabAtkins> | Sweet. |
| 01:04 | <TabAtkins> | I love great bands in small venues. |
| 01:05 | <gsnedders> | And it wasn't sold out, friends dragged other friends along and bought tickets at the door. |
| 01:05 | <gsnedders> | I wanted to see Lacuna Coil play there in September — sold out in two days. |
| 01:41 | <TabAtkins> | If anyone's interested in Flexbox, feel free to review my new draft at http://dev.w3.org/csswg/css3-flexbox/Overview.new.html . I want to make sure that other people think my approach is sane. |
| 06:02 | <Figaroo> | so why no h element? |
| 06:07 | <Figaroo> | I mean to say, why wasn't a single heading element defined in the HTML5 spec similar to how XHTML does headings and sections? |
| 06:12 | <Figaroo> | I suppose by using h1 as h you get the best of both worlds. |
| 06:22 | <kennyluck> | Figaroo, I think all these types of question will result in an answer, it just costs too much to add an element. |
| 06:23 | <kennyluck> | s/in/one/ |
| 06:23 | <kennyluck> | oops, anyway. |
| 07:38 | <Hixie> | Figaroo: there is a single heading element defined in HTML similar to how XHTML2 did headings and sections, we just spell it <h1> instead of <h>. See the spec. |
| 08:37 | <Figaroo> | kennyluck, why does it cost so much to add an element? |
| 09:14 | <jgraham> | Figaroo: Browsers and other tools have to support it, books and so on have to be rewritten to mention it, etc. |
| 09:16 | <jgraham> | In the specific case of <h> vs <h1> it seems to add more complexity and be of extremely marginal utility |
| 09:16 | <hsivonen> | jgraham: btw, I, too, think that the true meaning of NFC is Normalization Form C |
| 09:17 | <hsivonen> | using it for something else is very confusing |
| 09:22 | <jgraham> | hsivonen: If we had annevk's journalist credentials we could make up a new name and spread it by saying things like "NFC, sometimes called Hvísla" (or whatever), neglecting to mention that we are the only people using the made-up name :) |
| 09:51 | <Figaroo> | I've read up a good amount on sections and headings in HTML5. Let me just clearify here if I may. |
| 09:53 | <Figaroo> | I could follow the XHTML pattern (section > h, section > section > h, etc) only instead of <h> it's <h1>: <section><h1>Top level</h1><section><h1>Second level in outline</h1><section><h1>Third level in outline</h1</section></section></section> |
| 09:53 | <Figaroo> | ? |
| 09:53 | <jgraham> | Yes |
| 09:58 | <Figaroo> | <section><h1>top level</h1><section><h6>second leve</h6></section></section> is essentially the same outline as if I used a h1 instead of a h6? |
| 10:00 | <benschwarz> | Hixie: !ping |
| 10:01 | <Figaroo> | And I could also use <section><h1>top level</h1>...<h2>second level</h2></section> to get the best of both XHTML and HTML4 worlds? |
| 10:03 | <jgraham> | Figaroo: Only if "best of" means "all the confusing features from both" |
| 10:03 | <jgraham> | I mean, it is designed to work because some use cases require it |
| 10:03 | <jgraham> | But I wouldn't so it if I had a choice |
| 10:03 | <Figaroo> | jgraham, their both confusing? |
| 10:04 | <Figaroo> | they're* |
| 10:04 | <jgraham> | Figaroo: Both the "now how many <section>s deep am I" and the "so how does this mismash of <hx> elements relate to each other, let alone my content" |
| 10:06 | <Figaroo> | Isn't it simple: you have a new outline heirarchy for every section which starts under it's parent section? |
| 10:06 | jgraham | is nervous about the innerHTML behaviour of <pre>, <listing> and so |
| 10:07 | <jgraham> | Figaroo: It is almost simple, but consider |
| 10:07 | <jgraham> | <section><h1>Foo</h1><h2>Bar</h2><section><h1>Baz</h1> |
| 10:07 | <jgraham> | What us Baz a child of? |
| 10:08 | <jgraham> | *is |
| 10:09 | <Figaroo> | if that was in the body then wouldn't it be the top-most header and Foo would be the second level under Baz? |
| 10:10 | <Figaroo> | Am I correct? |
| 10:11 | <jgraham> | IIRC Bar and Baz would be siblings under Foo. But I am not sure that IRC |
| 10:13 | <Figaroo> | That's what I would say too. It gets confusing only because you put an h1 after a section, which is just as confusing as putting an h1 after an h2 which probably neither are valid. |
| 10:14 | <AryehGregor> | Putting an h1 after a section is allowed. |
| 10:15 | <AryehGregor> | You're allowed to use only h1 for headers, if you like. |
| 10:15 | <AryehGregor> | This makes it easier to include one page in another, you don't have to change all the header levels. |
| 10:16 | <Figaroo> | AryehGregor, putting a section before any h1s is allowed? |
| 10:16 | <AryehGregor> | I'm not familiar with this stuff, but why wouldn't it be? |
| 10:17 | <AryehGregor> | http://validator.nu/?doc=data%3Atext%2Fhtml%2C%3C%21doctype+html%3E%3Ctitle%3E%3C%2Ftitle%3E%3Csection%3E%3C%2Fsection%3E&showsource=yes |
| 10:17 | <AryehGregor> | Validates fine. |
| 10:19 | <jgraham> | Right, you can have an untitled section |
| 10:19 | <jgraham> | (and generally a <section> will come before a <h1> because the <h1> is a child of the <section> it titles) |
| 10:20 | <Figaroo> | but what is the section that comes before the h1 a child of? |
| 10:21 | <Figaroo> | if it were a child of the h1 that comes after it, then it wouldn't be untitled. |
| 10:21 | <jgraham> | The <body> |
| 10:22 | jgraham | is confused |
| 10:22 | <Figaroo> | here use this to see http://gsnedders.html5.org/outliner/ |
| 10:22 | <hsivonen> | Twitter algorithms need some tuning, I guess: https://twitter.com/#!/similar_to/MrLastWeek |
| 10:23 | <jgraham> | Figaroo: I was going to point *you* to that |
| 10:23 | <jgraham> | Figaroo: I don't understand what you are saying, really |
| 10:23 | <Figaroo> | hang on I'm using the tool :P |
| 10:24 | <jgraham> | hsivonen: That gives a blank page for me |
| 10:27 | <Figaroo> | jsgraham foo is the child of baz http://gsnedders.html5.org/outliner/process.py?url=http://jsbin.com/isadu3 |
| 10:27 | <hsivonen> | jgraham: I suppose it only works for logged in users then |
| 10:28 | <Figaroo> | here's the source http://jsbin.com/isadu3/edit |
| 10:28 | <hsivonen> | jgraham: anyway Twitter's mighty algorithms suggest TimBL as "similar" |
| 10:28 | <zcorpan> | maybe TimBL actually is MrLastWeek |
| 10:29 | <Figaroo> | who's MrLastWeek? |
| 10:35 | <jgraham> | Figaroo: I heard a rumor it's TimBL |
| 10:39 | <jgraham> | Figaroo: I meant http://jsbin.com/isadu3/3/ |
| 10:39 | <Ms2ger> | Yay, we're finally dropping run-in! |
| 10:40 | <zcorpan> | Ms2ger: what? |
| 10:40 | <Ms2ger> | From CSS2.1 |
| 10:40 | jgraham | must not say anything involving the words "run" "out" and "patience" |
| 10:40 | <Figaroo> | jgraham, easy, baz is at the same level as bar. |
| 10:40 | <zcorpan> | Ms2ger: will it have a come-back or is the plan to remove it from browsers too? |
| 10:41 | <jgraham> | Figaroo: Why is that easy? |
| 10:41 | <hsivonen> | My thesis uses run-in. |
| 10:41 | <Ms2ger> | No, it'll be back in CSS3 |
| 10:41 | <jgraham> | zcorpan: The point is it's too buggy in browsers to get CSS2.1 to REC |
| 10:41 | <zcorpan> | i see |
| 10:41 | <jgraham> | Figaroo: An equally logical model would make baz a child of bar |
| 10:42 | <jgraham> | The <h2> creates an implicit <section> afterall |
| 10:42 | <Figaroo> | Why isn't it? A section is a like a child tree that full directly under it's parent. |
| 10:42 | <jgraham> | it just happens that HTML adopted a model where explict sections can never be children of implicit sections |
| 10:43 | <Figaroo> | why would we want that? |
| 10:43 | <jgraham> | Want what? |
| 10:43 | <erlehmann> | hsivonen, twitter algorithms reliably point a fake account some friends of mine made to famous german bloggers (whom i know personally, but we are definitely not alike). |
| 10:43 | <Figaroo> | Why would we want explict sections to be children of implicit sections? |
| 10:44 | <jgraham> | Because that makes as much sense as having them not be children? It seems arbitary to me |
| 10:47 | <Figaroo> | I'm not sure I understand. You're saying that whether or not explicit sections can be children of implicit sections makes sense equally? |
| 10:48 | <jgraham> | Yes, I'm saying that you could easily imagine either model being true, so it is a confusing |
| 10:49 | <zcorpan> | i think it'd be more convenient if explicit sections could be children of implicit sections |
| 10:49 | <zcorpan> | so you can use implicit sections in general and throw in an <aside> somewhere without fucking up the outline |
| 10:50 | <zcorpan> | now if you want to throw in an <aside>, you need to change everything to also throw in explicit <section>s everywhere just to make the outline as it was before throwing in the <aside> |
| 10:50 | <Figaroo> | It seems like HTML wanted to adopted XHTML's explicitness without losing backwards compatibility. |
| 10:51 | <jgraham> | Figaroo: That is more or less it |
| 10:51 | <jgraham> | zcorpan: That seems like a reasonable point |
| 10:51 | <jgraham> | Dunno why we have the current design |
| 10:52 | <Figaroo> | but with this backwards compatibilty you have the benifit of the short, quick implicitness of multiple <h>s for every section. |
| 10:52 | <Figaroo> | HTML5 I think is trying to encourage the use of only h1. |
| 10:53 | <Figaroo> | ...and, encourage the use of explicit sections. |
| 10:53 | <AryehGregor> | The problem I have with explicit sections is then they have to be nested correctly, which usually is fine but not always. |
| 10:53 | <Figaroo> | AryehGregor, why not always? |
| 10:53 | <AryehGregor> | In particular, it doesn't work at all for MediaWiki, since you can do things like <table><tr><th><h1>Foo</h1></th>... |
| 10:53 | <AryehGregor> | So we can't automatically add <section>s to user input. |
| 10:54 | <AryehGregor> | Even though it would work 98.2% of the time. |
| 10:54 | <AryehGregor> | Which is sad. |
| 10:54 | <AryehGregor> | But I guess unavoidable. |
| 10:54 | zcorpan | files a bug |
| 10:54 | <jgraham> | Headings in tables just seems wrong |
| 10:55 | <AryehGregor> | Well, it's allowed. |
| 10:55 | <AryehGregor> | Also, why does it seem wrong? |
| 10:55 | <jgraham> | Because a table is a table? |
| 10:55 | <AryehGregor> | So? |
| 10:55 | <AryehGregor> | You can have a table that's so big you want to break it into sections. |
| 10:55 | <AryehGregor> | That's what <thead> and <tbody> are for. |
| 10:55 | <AryehGregor> | You can put <thead>s in the middle, right? |
| 10:56 | AryehGregor | checks |
| 10:56 | jgraham | doesn't think so |
| 10:56 | <AryehGregor> | No, you can't. |
| 10:56 | <AryehGregor> | Hmm. |
| 10:56 | <jgraham> | But I still don't understand why you would want a <hx> in the table |
| 10:56 | <AryehGregor> | You mean, you could just stick in some <th>'s if it has multiple sections? |
| 10:56 | <jgraham> | If it's so long that you break it into multiple *tables* then <hx> goes in the surrounding page |
| 10:57 | <AryehGregor> | In MediaWiki's case specifically, <hx> allows section editing. More generally, I dunno, users do crazy things. |
| 10:57 | <jgraham> | I can see that the <section> model is hard to implement in wikitext |
| 10:58 | <AryehGregor> | It would be great except for the nesting issue. Sections aren't required to be self-contained, you can have sections that are only partially wrapped in a tag. |
| 10:58 | <jgraham> | But I don't see what the use cases are that it couldn't cover given a sifficiently advanced implementation |
| 11:00 | <Figaroo> | Is <article> a sort of section element? |
| 11:00 | <AryehGregor> | I really need to figure out something productive I can do when I only have a random hour or so to spare, which doesn't require too much mental effort but also isn't totally useless. |
| 11:00 | <jgraham> | Figaroo: yes |
| 11:00 | <Figaroo> | why have it if we got section already? |
| 11:00 | <AryehGregor> | Currently I'm occupying myself by memorizing IDF ranks and insignia, which surely is not the most useful thing I could be doing at the moment. |
| 11:01 | <jgraham> | That doesn't strike me as the most useful thing to do at any moment |
| 11:01 | <jgraham> | But it my be a matter of perspective |
| 11:01 | <jgraham> | *may |
| 11:01 | <Figaroo> | why do we need article if we have section? |
| 11:01 | <jgraham> | They mean different things |
| 11:02 | <jgraham> | <section> is generally part of a larger work |
| 11:02 | <jgraham> | <article> is a freestanding piece of content |
| 11:02 | <jgraham> | (<article> is a bit like <body> except without all the magic to do with only having one per page) |
| 11:02 | <AryehGregor> | Presumably it would be useful to someone who served in the IDF. It seems like it would be useful to know whether someone is, e.g., a general. |
| 11:03 | <AryehGregor> | Although perhaps in practice the way it works is you start only having to know "anyone with insignia is above me", and then pick things up from there with practice. |
| 11:03 | <jgraham> | So in e.g. a newspaper you can have multiple <section>s each containing multiple <article> and each <article> can contain many <section>s |
| 11:03 | <AryehGregor> | Anyway, I memorized the officer ranks and insignia for the U.S. Army when I was about seven. |
| 11:04 | jgraham | remembers that AryehGregor has a full suit of armour; things start to make more sense |
| 11:04 | <AryehGregor> | Surely many small boys have an infatuation with the military. |
| 11:05 | <AryehGregor> | Anyway, I was just in Israel, and you've got tons of soldiers walking around there, and I couldn't even tell a sergeant from a major. :( |
| 11:05 | <AryehGregor> | (Clearly I should have memorized the ranks *before* I visited) |
| 11:05 | <Figaroo> | why couldn't the multiple sections just contain multiple sections; what's the point of having another tag name? Did someone here say earlier that new elements are expensive or something? |
| 11:06 | <AryehGregor> | Some people say "new elements are expensive", but that's a very silly way to put things. |
| 11:06 | <AryehGregor> | It's new features that are expensive. |
| 11:06 | <AryehGregor> | The way Raymond Chen puts it is "new features start at -100 points" or such. |
| 11:07 | <Figaroo> | My point is, what's the purpose in have a separate name for article? |
| 11:07 | <Figaroo> | why couldn't it all be sections |
| 11:07 | <AryehGregor> | As far as things go, adding a new element is a very minor cost. Things like, I dunno, microdata are a lot more expensive. And things like <iframe sandbox> are way way way more expensive. |
| 11:08 | <Figaroo> | forget that I mentioned expensive. I'm just trying to figure out why there's the extra element where section could have been perfectly capable? |
| 11:09 | <AryehGregor> | They mean different things, supposedly. |
| 11:09 | <AryehGregor> | I think <article> is supposed to be a self-contained unit of content, while <section> is generally not self-contained. |
| 11:10 | <AryehGregor> | So <article> would be like a blog post, and <section> would be like sections of a paper. |
| 11:10 | <jgraham> | The difference between <section> and <article> has been debated. It makes sense to me but ymmv |
| 11:10 | <AryehGregor> | Nobody's going to realize the difference, though. |
| 11:10 | <AryehGregor> | In practice. |
| 11:11 | <Figaroo> | ymmv? |
| 11:12 | <jgraham> | Your Mileage May Vary |
| 11:12 | <AryehGregor> | Man, Israelis love minimally distinguishable abbreviations. Turai, Rabat, Samal, Samar, Rasal, Rasar, Rasam, Rasab, Ranag, Sagam, Sagar, Seren, Rasan, Sa'al, Alam, Ta'al, Aluf, Ra'al. Seriously? |
| 11:12 | <AryehGregor> | Rasal, Rasar, Rasam, Rasab. In that order. |
| 11:12 | <AryehGregor> | Then Rasan is much higher. |
| 11:13 | <AryehGregor> | I guess it's more useful than being glued to Facebook or Twitter when I can't think of anything else to do. |
| 11:13 | <AryehGregor> | At least I'm learning something. |
| 11:15 | <Philip`> | You should spend all your free time writing HTML tests |
| 11:15 | <AryehGregor> | A possibility. |
| 11:15 | <AryehGregor> | That requires mental effort, though. |
| 11:15 | <Philip`> | Write a script to randomly pick a sentence in the spec that contains the word "must", then write a test for it, then repeat |
| 11:16 | <Ms2ger> | How do you define "sentence"? |
| 11:16 | <Ms2ger> | Or "word", for that matter? |
| 11:17 | <AryehGregor> | Just have it randomly pick an occurrence of the word "must". |
| 11:17 | <annevk> | hsivonen, hahahaha |
| 11:18 | <annevk> | hsivonen, Mr Last Week is the W3C |
| 11:18 | <annevk> | it all makes sense now :) |
| 11:18 | <AryehGregor> | "the W3C"? |
| 11:18 | <annevk> | the similar to twitter link |
| 11:18 | <jgraham> | AryehGregor: Write a script so that we can link tests to sentences in the spec! Philip` has even done most of it for you… |
| 11:20 | <Figaroo> | what if you wanted all of your site's sections to be self-contained? |
| 11:20 | <Figaroo> | Would that mean you'd use nothing but sections? |
| 11:20 | AryehGregor | is now going through http://commons.wikimedia.org/wiki/Category:Military_people_of_Israel_by_name and identifying the ranks of every picture where someone is wearing rank insignia |
| 11:20 | <Figaroo> | I mean articles* |
| 11:20 | <foolip_> | Figaroo, then your site is probably a blog |
| 11:20 | <AryehGregor> | Sure, no one says you have to use <section>. |
| 11:20 | <foolip_> | and that's fine |
| 11:21 | <foolip_> | FWIW, there's an algorithm for converting HTML to Atom that relies on <article> |
| 11:22 | <Figaroo> | articles define their own outline level right? |
| 11:22 | <foolip_> | I think so, yes |
| 11:23 | <foolip_> | so it's more or less <section class=syndicatable> |
| 11:25 | <Figaroo> | w/etf syndicatable means. :P |
| 11:26 | <Figaroo> | Alright, so. What's up with header and footer; apparently you can have more than one of these, but why would you want more than one of them? |
| 11:29 | Philip` | notes that his spec-annotation script is probably wildly inscalable, since it checks every assertion pattern against every paragraph in the document, which isn't ideal when you have ~3000 assertions |
| 11:30 | <AryehGregor> | Philip`, just change it so the assertions are ordered. |
| 11:30 | Ms2ger | still likes the CSSWG's approach |
| 11:30 | <AryehGregor> | And do it by scanning the document until you find a paragraph that meets the next assertion. |
| 11:30 | <Philip`> | What approach is that? |
| 11:30 | <AryehGregor> | Then it's one comparison per document paragraph, so it's O(N) in document size. |
| 11:31 | <Ms2ger> | Just have <link>s in the test |
| 11:32 | <jgraham> | Ms2ger: Having required markup in HTML tests is bad |
| 11:32 | <Ms2ger> | Meh |
| 11:32 | <Philip`> | AryehGregor: Hmm, that might work, though it might cause horrible cascading error reports if one assertion doesn't match where you expected it to (e.g. because the spec changed) |
| 11:32 | <jgraham> | (even the scripts are rather bad) |
| 11:32 | <Figaroo> | is Niels here? |
| 11:32 | <Ms2ger> | It wouldn't matter in the majority of cases |
| 11:33 | <jgraham> | Ms2ger: Also you would end up having to generate the <link> elements somehow for files containing thousands of tests |
| 11:33 | <jgraham> | And the spec doesn't have the right granularity of <link>s |
| 11:33 | <jgraham> | Like per-clause |
| 11:33 | <jgraham> | I mean ids |
| 11:34 | <jgraham> | not <links> |
| 11:42 | <Figaroo> | what would you suggest I do if I had a bar at the top of my page that had a search field on the right and some extra links on the left, then I had a centered 900 pixel width box in the center with the logo at and navigation below it, then content within this box. What elements would you think would go well with this type of design? |
| 11:46 | <Evet> | so which framework/library you suggest for mobile web apps? |
| 11:47 | <Figaroo> | Here's the desgin I'm working with. What would be the best HTML5 layout for this http://i.imgur.com/3ClkR.png ??? |
| 11:54 | <Figaroo> | Should I separate the page into three sections off the bat; one untitled section for the top gutter, one for the main page area, and then one for the footer? |
| 11:57 | <Figaroo> | Is it a good idea to just put a section immediately after within the body without a heading preceding it? |
| 13:25 | <Figaroo> | if there are multiple header elements on your page, how do you select the main header for styling? |
| 13:28 | <loucapo> | hi everyone |
| 13:29 | <Figaroo> | Anyone know? |
| 13:29 | <loucapo> | anyone here experience 401s when using HTML5 for CORS POSTing to a domain that requires basic auth? |
| 13:30 | <loucapo> | happenong in chrome and FF only |
| 13:57 | <loucapo> | anyone here experience 401s when using HTML5 for CORS POSTing to a domain that requires basic auth? |
| 14:05 | <annevk> | loucapo, no need to ask twice |
| 14:05 | <annevk> | loucapo, and that is per spec |
| 14:05 | <annevk> | loucapo, well, the Authorization header needs to be in the list of allowed headers and you need to create it yourself using script |
| 14:06 | <annevk> | loucapo, that ought to work, not sure if it does |
| 14:07 | <loucapo> | hmm i see |
| 14:08 | <loucapo> | thx fr the info! |
| 14:43 | <matjas> | TabAtkins: Ping |
| 14:51 | <zcorpan> | sad that many games in https://gaming.mozillalabs.com/games/finalists block opera |
| 14:54 | <hsivonen> | zcorpan: it it clear why? |
| 14:54 | <Ms2ger> | Ugh, really? |
| 14:56 | <zcorpan> | dunno why |
| 14:59 | <hsivonen> | ouch. MathML 3 has a role attribute for non-ARIA purposes |
| 15:02 | <jgraham> | zcorpan: So far 1 worked, 1 blocked me for mo WebGL, 1 blocked me for no fb account, 1 seems broken in Opera |
| 15:03 | <Workshiva> | mo WebGL, mo problems |
| 15:03 | <jgraham> | :p |
| 15:04 | <jgraham> | +1 working +1 blocked |
| 15:04 | jgraham | is bored of trying them now |
| 15:08 | <zcorpan> | http://pixza.com/ when trying to play redirects to http://pixza.com/league/browser.php in opera which says "The page you are looking for does not exist." |
| 15:30 | <miketaylr> | very few of these types of contests care much about interop, i think only js1k had that as a requirement |
| 15:32 | <hsivonen> | miketaylr: I thought Mozilla's competition required multiple browsers |
| 15:33 | <hsivonen> | not sure if it was simply Gecko and one other engine |
| 15:33 | <miketaylr> | hsivonen: i guess you're right "Entries should be fully operational games, built to run in Firefox 4 and other modern Web browsers" |
| 15:33 | <jgraham> | One entry Said Firefox 3.6 & 4 |
| 15:33 | <jgraham> | Does that count? |
| 15:33 | <miketaylr> | according to https://gaming.mozillalabs.com/rules |
| 15:34 | <miketaylr> | heh, technically it might, jgraham |
| 15:35 | <Ms2ger> | 3.6 is hardly a modern browser ;) |
| 15:36 | <jgraham> | Anyway it only says "should" |
| 15:37 | <Ms2ger> | This isn't a spec, should is must-level ;) |
| 15:37 | <AryehGregor> | "The Contest is void in Cuba, Iran, Province of Quebec (Canada), North Korea, Sudan, Syria, Union of Myanmar, and where prohibited by law or restricted by law or regulations." |
| 15:37 | <AryehGregor> | Are they making some kind of statement about Canadians there? |
| 15:37 | <Ms2ger> | "Entry must be compatible with and function properly in the then-current final product version of Firefox 4 and other modern web browsers without plug-ins" |
| 15:37 | <Ms2ger> | Says the real rules |
| 15:37 | <jgraham> | Almost all competitions are illegal in Quebec |
| 15:38 | <AryehGregor> | Really? Even totally skill-based competitions? |
| 15:38 | <jgraham> | Well illegal is wrong |
| 15:38 | <hsivonen> | jgraham: are competitions illegal or does Quebec require drafting all rules in French or something? |
| 15:38 | AryehGregor | also suspected it might be because it had to be in French :P |
| 15:38 | <jgraham> | No, I think they have strong rules about lotteries or something |
| 15:39 | <jgraham> | And so it is lots of effort to prove that your competition is not an illegal lottery |
| 15:39 | <jgraham> | So instead you are just not allowed to enter |
| 15:39 | <AryehGregor> | I thought Americans were the ones who were anti-lottery. |
| 15:39 | <hsivonen> | AryehGregor: in Finland, holding a lottery requires a license |
| 15:41 | <AryehGregor> | I remember reading about some guy who ran a British gambling site with lots of American customers, who was arrested when he visited America because his site didn't comply with American gambling laws. |
| 15:41 | <AryehGregor> | (some British gambling sites started blocking America after that :P) |
| 15:42 | <Workshiva> | Quebec probably also requires that everything (including prizes) be available in French too :P |
| 15:43 | <AryehGregor> | The contest rules say submissions have to be in English. |
| 15:43 | <AryehGregor> | I just think it's amusing that Quebec is on the same list as Cuba, Iran, North Korea, Sudan, Syria, and Myanmar. |
| 15:43 | <AryehGregor> | Most of those are under U.S. embargo, aren't they? At least the first three. |
| 15:44 | <Rik`> | there is more judges than finalists for this competition |
| 15:44 | <Rik`> | *are |
| 15:44 | <AryehGregor> | The judges presumably had to judge everyone, not just the finalists, so that makes sense. |
| 15:44 | <AryehGregor> | Okay, now let me stop wasting time and get back to work. Yay, Selections! |
| 15:46 | karlcow | had not seen the fineprint https://gaming.mozillalabs.com/fineprint |
| 15:47 | <karlcow> | gorgeous |
| 15:49 | <jgraham> | http://searchwarp.com/swa9752.htm |
| 15:49 | <jgraham> | Not legal advice but the sort of thing I had heard |
| 15:50 | <karlcow> | lotteries/contests are controlled by the state of Quebec and have very strict rules. |
| 15:52 | <hsivonen> | it seems the rules lack a definition of "other modern web browsers" |
| 16:35 | <AryehGregor> | Oh, hmm, I haven't tested extend() in tables at all . . . |
| 16:35 | <AryehGregor> | That probably has totally magical and unique behavior. |
| 16:35 | AryehGregor | tests, with trepidation |
| 16:35 | <Ms2ger> | Oh, right, I was going to mention tables |
| 16:35 | <AryehGregor> | And I had just gotten Firefox to pass all of my tests, except where it throws the wrong exception type or one corner case where I think its behavior is just a bug. |
| 16:37 | <AryehGregor> | Out of almost 9000 tests, Opera passes less than 100. |
| 16:37 | <AryehGregor> | Awesome. |
| 16:38 | AryehGregor | sighs, starts looking at tables |
| 16:40 | <Ms2ger> | Speaking of Opera, is meter stylable, and are there docs on that? |
| 16:44 | <AryehGregor> | Ms2ger, is there actually magic extend() behavior for tables? I don't see it, offhand. |
| 16:44 | AryehGregor | throws some tables into the tests and see if they start failing things |
| 16:45 | <Ms2ger> | Dunno, I just noticed table-specific code in Gecko's implementation, not sure if that was for extend |
| 16:46 | <AryehGregor> | I don't notice table-specific handling in the extend() code. |
| 16:46 | <AryehGregor> | I'll just optimistically assume that there isn't any for now. |
| 16:47 | <AryehGregor> | Although probably I'll have to face it sooner or later. |
| 16:47 | <AryehGregor> | Okay, so Gecko really does let you have multiple selections at once, and no other browser does. Weird. |
| 16:49 | <karlcow> | AryehGregor: do you know the source of failure for the tests in Opera |
| 16:49 | <AryehGregor> | karlcow, presumably it just doesn't implement anything more than vaguely similar to what Gecko does. |
| 16:49 | <AryehGregor> | I doubt it would make sense to make a list of specific things, the relevant code probably just needs to be rewritten to spec. |
| 16:50 | <AryehGregor> | Which winds up being pretty simple, by the way: http://aryeh.name/tmp/dom-range.html#dom-selection-extend |
| 16:52 | <dglazkov> | annevk: you around? |
| 16:53 | <dglazkov> | wanna bounce a few css selector-ey questions off of you |
| 16:53 | <dglazkov> | TabAtkins, annevk: so I realized that pseudo-element style selector doesn't really work well for the shadow DOM. |
| 16:53 | <dglazkov> | I think we need a whole different syntax |
| 16:55 | <dglazkov> | the problem is in that we have to nest and/or combine them with pseudo-classes and it looks a) ugleee b) illegal. |
| 16:55 | <TabAtkins> | It's only illegal because no one's had to do it before. |
| 16:55 | <dglazkov> | suppose I want to style the knob on the timeline of a video element in it's active state: |
| 16:55 | <dglazkov> | video::timeline::thumb:active |
| 16:55 | <dglazkov> | wat |
| 16:56 | <TabAtkins> | Yes? |
| 16:56 | <dglazkov> | it would be much more elegant like this: |
| 16:56 | <dglazkov> | video part(timeline) part(thumb):active |
| 16:56 | <dglazkov> | where "part" is up for bike-shedding. |
| 16:56 | <TabAtkins> | video ::part(timeline) ::part(thumb):active |
| 16:57 | <dglazkov> | does it have to have the double-colons? |
| 16:57 | <TabAtkins> | Not necessarily, but then you really are defining a new type of selector. |
| 16:57 | <dglazkov> | yay! is that a bad thing? :) |
| 16:57 | <TabAtkins> | Not necessarily. ^_^ |
| 16:58 | <dglazkov> | I think it feels more like a descendant selector. |
| 16:58 | <dglazkov> | so I want it to kind of look like one |
| 16:58 | <dglazkov> | (space and everything) |
| 16:58 | <dglazkov> | just thinking outloud |
| 16:59 | <TabAtkins> | I dunno. I mean, pseudoelements already are descendants, right? foo::before has ::before being the child of foo. |
| 16:59 | <TabAtkins> | Now, on the other hand, if pseudoclasses can apply to the parts, then that's more confusing. |
| 16:59 | <TabAtkins> | video::timeline:focus::thumb:active lolwut |
| 17:00 | <dglazkov> | that's a good example too |
| 17:00 | <dglazkov> | in WebKit today, there's much hard-coding around cases like this. |
| 17:01 | <dglazkov> | I'll write this up in the use cases document |
| 17:01 | <dglazkov> | and hopefully annevk and Hixie would chip in with their bright ideas, too. And then we'll have a pizza party |
| 17:01 | <TabAtkins> | Okay, since we're doing this in a general, distributed, composable fashion, I think I agree that it's saner to lean on the structure of selectors and make them look like descendants. |
| 17:03 | <TabAtkins> | matjas: pong |
| 17:06 | <AryehGregor> | WTF. |
| 17:06 | <AryehGregor> | hg pull --rebase dumped me into some kind of incomprehensible three-way diff. |
| 17:06 | <AryehGregor> | I tried exiting, and it happily reported the rebase was successful. |
| 17:07 | <AryehGregor> | If it ate my changes, I'm going to kill someone. |
| 17:07 | <AryehGregor> | I hate hg. |
| 17:07 | <AryehGregor> | Is there a hg-git bridge, by any chance? A la git-svn? |
| 17:08 | <AryehGregor> | Yes, completely ate my changes. |
| 17:08 | <AryehGregor> | I have a cached copy of the Anolis output, but can anyone suggest some way to recover my old source document? If the answer is "rebase destroys stuff irrecoverably", I'm going to be *extremely* annoyed at hg. |
| 17:09 | <Philip`> | Does it do a bundle backup by default? |
| 17:09 | <AryehGregor> | (git never ever destroys anything, except if it's old or you explicitly tell it to) |
| 17:09 | <AryehGregor> | I have no idea what a bundle backup is. |
| 17:09 | <Philip`> | like in .hg/strip-backup/ |
| 17:09 | <AryehGregor> | "hg rollback" seems not to work. |
| 17:09 | <Philip`> | like what http://mercurial.selenic.com/bts/msg11265 says |
| 17:09 | <AryehGregor> | I see a file there. |
| 17:10 | Philip` | has never used this himself |
| 17:10 | <AryehGregor> | Unbundling it does nothing. |
| 17:11 | Philip` | reads more of that bug thread |
| 17:11 | <Philip`> | "The bundle produced is a consequence of the use of strip, but it's really not intended to provide a means to restore the old branch." |
| 17:12 | <Philip`> | unless you're using Hg 1.6 in which case maybe the bug is fixed |
| 17:13 | <AryehGregor> | So in other words, hg destroys data while rebasing and there's no way to recover it if you made a mistake while rebasing. |
| 17:13 | <AryehGregor> | Okay, well, I keep filesystem backups on my server, maybe one of those is usefully recent. |
| 17:14 | <Philip`> | On <1.6 it sounds like that's probably the case |
| 17:14 | <AryehGregor> | Have I ever mentioned that git is better than hg? |
| 17:14 | <Philip`> | I'm not sure you have |
| 17:15 | <AryehGregor> | Okay, well, consider it mentioned. |
| 17:15 | <Philip`> | Does your text editor not keep backups? |
| 17:15 | <AryehGregor> | Not after I close it. |
| 17:15 | <Philip`> | Ah |
| 17:15 | <AryehGregor> | I've got a 40-minute-old backup of the source text, I should be able to recover okay combined with the output page. |
| 17:15 | <dglazkov> | TabAtkins: http://wiki.whatwg.org/wiki/Component_Model_Use_Cases#Nesting_Pseudo_Elements_and_Mixing_with_Pseudo_Classes |
| 17:16 | <AryehGregor> | But ARGH, what a horrifyingly broken VCS. |
| 17:16 | <Ms2ger> | Well, rebase is horribly broken |
| 17:16 | <AryehGregor> | Rebase in hg is horribly broken, I assume you mean. |
| 17:16 | <AryehGregor> | I guess I'll try doing merges instead. |
| 17:16 | <Ms2ger> | mq has never eaten my changes, at least |
| 17:17 | <AryehGregor> | Now how can I get rid of this empty tip commit? |
| 17:17 | <Philip`> | (Git doesn't seem particularly immune to data loss either, like if you accidentally run "git reset --hard") |
| 17:17 | <AryehGregor> | Philip`, the data is still there. |
| 17:17 | <AryehGregor> | As long as you have the SHA1 of the old head. |
| 17:17 | <AryehGregor> | You can still check it out and such. |
| 17:17 | <AryehGregor> | Detached revisions like that are only garbage-collected once in a while by default. |
| 17:18 | <Ms2ger> | hg qimp -r tip -n empty && hg qpo && hg qrm empty |
| 17:18 | <AryehGregor> | If you don't have the SHA1 of the old head, you can still poke around and look at all the detached revisions to find it. |
| 17:18 | <Philip`> | Uncommitted unstaged data won't still be there, I think |
| 17:18 | <AryehGregor> | Well, if you've never committed or staged it, of course not. |
| 17:18 | <AryehGregor> | Then the VCS can't be expected to preserve it. |
| 17:18 | <AryehGregor> | I committed this to hg and it still ate it. |
| 17:18 | <AryehGregor> | Ms2ger, hg: unknown command 'qimp' |
| 17:19 | <Philip`> | Have you (prepare to be annoyed by Hg again) enabled the hgext.mq extension? |
| 17:19 | <Ms2ger> | Add [extensions]\nmq= to ~/.hgrc |
| 17:20 | <AryehGregor> | Ms2ger, how can that command be adapted to ditch a whole bunch of revisions, like to go back to r203? |
| 17:20 | <AryehGregor> | I think that's what I originally based my work on. |
| 17:20 | <Philip`> | AryehGregor: The VCS could be expected to preserve any data that it's about to overwrite or delete and that isn't already saved somewhere permanent |
| 17:21 | <AryehGregor> | Philip`, it could be, yes, but does any VCS do that? |
| 17:21 | <Philip`> | Probably not |
| 17:21 | <Philip`> | but they should |
| 17:21 | <Ms2ger> | Hmm |
| 17:22 | <Ms2ger> | hg strip rev |
| 17:22 | <AryehGregor> | Ah, thanks. |
| 17:22 | <Ms2ger> | Which does make a backup |
| 17:25 | <TabAtkins> | http://blogs.wsj.com/venturecapital/2011/01/25/with-plenty-of-cash-facebook-targets-growing-mobile-market/ |
| 17:25 | <TabAtkins> | They're targetting HTML5 so they can stop supporting 7 different versions for mobile. |
| 17:27 | <AryehGregor> | What's the difference between "hg diff" and "hg diff -r tip"? They're telling me different things for some reason. |
| 17:27 | <matjas> | TabAtkins: re: http://lists.w3.org/Archives/Public/public-css-commits/feed.rss, any chance of having the commit messages as <title> for each item instead of the rather cryptic stuff that’s in there now? |
| 17:27 | <Ms2ger> | matjas, no |
| 17:28 | <TabAtkins> | I've already asked the systeam. They think it'd be easy, but they're working on something feed-related already, so probably wont' get to it for a while. |
| 17:28 | <Philip`> | AryehGregor: Is your working directory on tip? |
| 17:28 | <TabAtkins> | Now, if someone wants to help me figure out how to receive email in PHP, I can do the rest. |
| 17:28 | <matjas> | Ms2ger: Couldn’t you guys re-use the system @whatwg is using? It’s open-source and everything |
| 17:28 | <matjas> | TabAtkins: Ah, good to hear. Thanks. |
| 17:29 | <Ms2ger> | Not my call, I'd do it immediately ;) |
| 17:29 | <AryehGregor> | Philip`, what does that mean? |
| 17:29 | <matjas> | TabAtkins: Well, there’s imap_open() http://php.net/imap_open |
| 17:30 | <Philip`> | AryehGregor: The tip revision is the most recent changeset, and you might have updated to a different revision before making the local changes that you're diffing against that revision and against tip |
| 17:30 | <AryehGregor> | Philip`, so how do I tell? |
| 17:30 | <Philip`> | "hg id" maybe |
| 17:34 | <AryehGregor> | Okay, so somehow when I committed, it had a parent of 203 instead of tip . . . well, whatever, I don't care. |
| 17:34 | <AryehGregor> | As long as it merges properly. |
| 17:34 | Ms2ger | crossesfingers |
| 17:35 | <AryehGregor> | What the heck is this merging interface that it throws at me with no instructions? |
| 17:35 | <AryehGregor> | This is why my previous rebase failed, I just tried to quit and it interpreted that as "ignore all my changes". |
| 17:36 | <AryehGregor> | It seems to have opened up three side-by-side windows in vim, but it helpfully fails to tell me what to do with them. |
| 17:39 | <Ms2ger> | https://developer.mozilla.org/en/Mercurial_basics#Avoiding_trouble, btw |
| 17:40 | <TabAtkins> | Btw, if anyone's curious about what's going on in Egypt, there's a live feed in English: http://english.aljazeera.net/watch_now/ |
| 17:40 | <AryehGregor> | "Mercurial does not do a good job of detecting busted merges and refusing to proceed, so just by closing a window you can unwittingly put yourself in a bad state. Grown men have been known to walk away from whole hg trees containing many hours' worth of work, citing inexplicable hg behavior, because of a bad merge." |
| 17:40 | <AryehGregor> | That's good to know, isn't it? |
| 17:40 | AryehGregor | STABSTABSTAB |
| 17:42 | <AryehGregor> | SyntaxError: Specification not found: domps. |
| 17:42 | <AryehGregor> | Ms2ger, ^^ |
| 17:43 | <AryehGregor> | I did hg pull and hg update in data/. |
| 17:43 | <Ms2ger> | Forgot to push, sorry |
| 17:45 | <Ms2ger> | Okay, pull again |
| 17:48 | <AryehGregor> | Okay, works. |
| 17:49 | <annevk> | dglazkov, you mean you want inheritance? |
| 17:49 | <dglazkov> | annevk: no, just better syntax |
| 17:49 | <TabAtkins> | annevk: No, just a readable way to talk about potentially nested shadow tree parts. |
| 17:49 | <annevk> | dglazkov, nothing comes to mind right now, except that for captioning there is a similar "issue" |
| 17:49 | <dglazkov> | annevk: the inheritance is already there. |
| 17:50 | <annevk> | dglazkov, might be worth checking if that can be reconciled |
| 17:50 | <TabAtkins> | Yeah, captioning gets a bit simpler since there's no nesting (or rather, the nesting that does happen isn't important enough to care about). |
| 17:50 | <annevk> | i.e. WebSRT creates some kind of subtree too |
| 17:54 | <AryehGregor> | Ms2ger, I've signed up to Bitbucket as aryeh. |
| 17:55 | <Ms2ger> | AryehGregor, you should now be able to push |
| 17:56 | <AryehGregor> | Ms2ger, could you confirm that this looks reasonably sane first? https://bitbucket.org/aryeh/dom-range |
| 17:57 | <AryehGregor> | I'm not sure the merge worked right. |
| 17:58 | <AryehGregor> | Argh, the fact that my e-mail address is different for one of them will bother me forever. |
| 17:58 | <AryehGregor> | In git I'd fix it . . . but let me just not bother. Sigh. |
| 17:58 | <annevk> | chill ;) |
| 17:59 | <Ms2ger> | That looks fine |
| 17:59 | AryehGregor | is too perfectionist sometimes |
| 17:59 | <AryehGregor> | k, pushed. |
| 18:01 | <matjas> | TabAtkins et al, let me know if you want this account: http://twitter.com/csscommits |
| 18:02 | matjas | is hoping for @whatwg-style Twitter feed |
| 18:02 | <annevk> | I should set something like that up for specs I edit |
| 18:02 | <Ms2ger> | http://html5.org/specs/dom-range.html#dom-selection-extend |
| 18:03 | <Ms2ger> | <cers> it seems like it's not supposed to be possible to create a linear gradient where the gradient line doesn't go through the element center - am I really reading the w3c draft correctly? http://dev.w3.org/csswg/css3-images/#linear-gradients |
| 18:03 | <Ms2ger> | TabAtkins, ^ |
| 18:04 | <TabAtkins> | matjas: Right now I've got a new gmail inbox subscribed to the list, and a php script successfully connecting. I'm just trying to figure out how to use the horrible imap interface to read messages, then I'll construct an RSS feed from it. |
| 18:04 | <matjas> | TabAtkins, does this help? http://stackoverflow.com/questions/114953/how-to-get-email-and-their-attachments-from-php/115017#115017 |
| 18:05 | <matjas> | annevk +1 |
| 18:05 | <TabAtkins> | Ms2ger: There's no difference where the linear gradient "goes through". I define it in terms of going through the center because it's easy, but you can shift what point you want to think about it going through with no problem. |
| 18:05 | <TabAtkins> | matjas: Ooh, yes, that's perfect. |
| 18:07 | <Ms2ger> | https://bitbucket.org/cynthia/twt |
| 18:13 | <annevk> | Ms2ger, oh, that looks nice |
| 18:14 | <annevk> | Ms2ger, then I can just make a simple shell script wrapper around what I need to do now |
| 18:14 | Ms2ger | didn't even realize it was relevant, just thought it was interesting |
| 18:16 | <annevk> | I registered http://twitter.com/annevkbot for this purpose, but am gonna watch some TV instead of making it work today :) |
| 18:18 | <annevk> | oh, Ms2ger, that uses a) Python and b) the now obsolete API |
| 18:18 | <Ms2ger> | Fun |
| 18:18 | <annevk> | but I can simply reuse the logic for WebAppsTweet |
| 18:18 | <TabAtkins> | Hm, I need to go commit something real quick so I can test this properly. |
| 18:19 | <AryehGregor> | Hixie, okay, I think I'm done with extend() for now. Tests at <http://aryeh.name/spec/dom-range/test/selection-extend.html>;, spec at <http://html5.org/specs/dom-range.html#dom-selection-extend>;. |
| 18:21 | <AryehGregor> | On to modify()! |
| 18:21 | <AryehGregor> | I guess for this one I'm reverse-engineering mostly WebKit, since they made it up. |
| 18:24 | <annevk> | http://blog.whatwg.org/another-whatwg-weekly#comment-42599 are actually pretty good questions |
| 18:24 | <annevk> | Hixie, ^^ |
| 18:24 | <AryehGregor> | Ms2ger, why is DOM Range not in W3C space again? There's a component for it in the WebApps WG bug tracker. |
| 18:25 | <TabAtkins> | annevk: So basically "adopt CSS's versioning model"? |
| 18:26 | <annevk> | AryehGregor, no group is chartered for it |
| 18:26 | <AryehGregor> | annevk, we can't say it's part of HTML or web apps? |
| 18:28 | <annevk> | AryehGregor, I guess it can be argued it is part of WebApps per http://www.w3.org/2010/webapps/charter/#maintenance |
| 18:29 | <AryehGregor> | Stuff like this is even part of HTML. |
| 18:29 | <annevk> | AryehGregor, the W3C made it all difficult for Ms2ger to contribute with Web DOM Core though... but maybe that is over now and otherwise a WebApps WG member can put his/her name on it and ship it |
| 18:29 | <annevk> | AryehGregor, you think it is problematic that it is hosted on html5.org? |
| 18:30 | <AryehGregor> | No, not really. |
| 18:30 | <AryehGregor> | othermaciej was annoyed about it the other day, though. |
| 18:31 | <othermaciej> | annoyed about what? |
| 18:33 | <AryehGregor> | About DOM Range not being in W3C space, because you wanted to point some people to it. |
| 18:33 | <othermaciej> | I was sad that document.getSelection() got removed from the spec with no W3C-space replacement |
| 18:33 | <AryehGregor> | Right |
| 18:33 | <AryehGregor> | . |
| 18:33 | <othermaciej> | I don't care about DOM Range |
| 18:33 | <AryehGregor> | getSelection() is in DOM Range. |
| 18:33 | <othermaciej> | it's not even clear to me why Selection has to be in the same spec as DOM Range |
| 18:33 | <othermaciej> | since it never was before |
| 18:33 | <AryehGregor> | Well, the current spec refers normatively to the behavior of Ranges. |
| 18:34 | <AryehGregor> | Since various operations on Selections return Ranges. |
| 18:34 | <AryehGregor> | W3C specs are presumably not supposed to normatively reference things hosted on someone's Bitbucket account. |
| 18:34 | <othermaciej> | anyway, it would be nice if the draft could be submitted to either HTML WG or Web apps WG |
| 18:34 | <AryehGregor> | What's the procedure for that? |
| 18:35 | <annevk> | othermaciej, neither editor is a Member of the WG |
| 18:35 | <AryehGregor> | I'm a member of the HTML WG. |
| 18:35 | <othermaciej> | both editors are members of the HTML WG afaict |
| 18:35 | <annevk> | AryehGregor, this would be a WebApps deliverable |
| 18:35 | <AryehGregor> | Why? |
| 18:35 | <AryehGregor> | Selection used to be in HTML5. |
| 18:35 | <othermaciej> | http://www.w3.org/2000/09/dbwg/details?group=40318&public=1 |
| 18:36 | <annevk> | Because maintenance of DOM Range is in the WebApps charter |
| 18:36 | <annevk> | Anyway, feel free to argue with the W3C Team :) |
| 18:37 | <othermaciej> | other than the usual patent policy stuff, the use I had for a Selection spec in W3C space was to tell people there was no point in a <selection> or <caret> element, since you just use the Selection API to manage the selection |
| 18:37 | <othermaciej> | unfortunately it got removed from HTML5, and pointing to a random bitbucket directory would likely have not impressed my audience |
| 18:37 | <annevk> | othermaciej, http://html5.org/specs/dom-range.html |
| 18:38 | <annevk> | which in turn can be found via http://wiki.whatwg.org/wiki/Specs |
| 18:39 | <AryehGregor> | Which in turn would not have impressed his audience. |
| 18:39 | <othermaciej> | if I point to that, it will start an argument over why it's not in the draft any and why there isn't a copy in W3C space at all, and I'd rather not start that debate |
| 18:39 | <AryehGregor> | othermaciej, can you tell me how to submit a draft? |
| 18:40 | <othermaciej> | for HTML WG I can tell you, but anne may have a point that it should be submitted to Web Apps WG |
| 18:40 | <AryehGregor> | Which I'd then have to join, right? Is it open membership like the HTMLWG? |
| 18:41 | <othermaciej> | I don't think so, but if you ask the chairs to add you as an invited expert and explain your purpose, it shouldn't be a problem |
| 18:45 | <AryehGregor> | Well, I'll do it if Hixie says it's part of my job (which I wouldn't mind). Otherwise it sounds like too much effort when I have better things to do. |
| 18:50 | <AryehGregor> | Ms2ger, why does <span data-anolis-spec=html>ASCII case-insensitive</span> not work? |
| 18:51 | AryehGregor | isn't quite clear on this xref thing |
| 18:52 | <Ms2ger> | Make that =domcore |
| 18:53 | <Ms2ger> | More generally, because I don't have a full list of HTML's definitions |
| 18:54 | <AryehGregor> | Okay, thanks. |
| 18:54 | <AryehGregor> | In principle, is there any reason data-anolis-spec should usually be needed? Couldn't Anolis be adapted to figure out the spec from the name? It would be nice if all this could be less verbose to write. |
| 18:58 | <Ms2ger> | It could, but I tend to prefer making external links explicit |
| 18:59 | <Ms2ger> | And it'd mean I'd have to implement that :) |
| 19:03 | <TabAtkins> | Son of a... How do you deal with diffs that may include "<!--" in an atom feed? |
| 19:04 | <Ms2ger> | In general, I believe the answer would be "not" |
| 19:04 | <AryehGregor> | . . . Escape the "<"? |
| 19:04 | <AryehGregor> | I mean, aren't you escaping the feed contents? |
| 19:04 | <TabAtkins> | AryehGregor: But I'm trying to send it as text. I guess I can send as html and do double-escaping? |
| 19:05 | <Ms2ger> | You need to escape in any case, no? |
| 19:05 | <AryehGregor> | How is it different from any case where the feed contents contain "<"? |
| 19:05 | <AryehGregor> | Atom is XML, yes? |
| 19:06 | <Ms2ger> | "XML" |
| 19:06 | <AryehGregor> | Presumably any consumer will un-escape the contents of elements before displaying them. |
| 19:06 | <AryehGregor> | I thought RSS was "XML" and Atom was XML. |
| 19:06 | <Ms2ger> | Well, RSS would be """XML""" |
| 19:06 | <Ms2ger> | Atom is XML+encoding fixup, iirc |
| 19:14 | Ms2ger | has submitted an invited expert application for the WebApps WG |
| 19:14 | <Ms2ger> | We'll see how it goes... |
| 19:22 | <AryehGregor> | Ms2ger, why doesn't your data have all the definitions from HTML5? Isn't the list autogenerated? |
| 19:22 | <Ms2ger> | No |
| 19:22 | <AryehGregor> | Hurrah. |
| 19:22 | <Ms2ger> | It has the dozen or so definitions I've needed so far |
| 19:22 | AryehGregor | adds another one to his local copy |
| 19:23 | <Ms2ger> | Note that the key needs to be lowercase |
| 19:23 | <AryehGregor> | Noted. |
| 19:57 | <TabAtkins> | matjas: http://www.xanthir.com/feed/public-css-commits/ |
| 19:58 | <Ms2ger> | TabAtkins, you win |
| 19:59 | <AryehGregor> | Oh, no. modify() in WebKit depends on whether the thing you're dealing with is actually visible. |
| 19:59 | <TabAtkins> | No, Ms2ger, the web wins. Subscribed to the list with a throwaway gmail account, used PHP's imap support to scrape the inbox, published as Atom, viewing in Google Reader. |
| 19:59 | <AryehGregor> | It skips over display: none and who knows what else. |
| 19:59 | <AryehGregor> | Also visibility:hidden. |
| 20:00 | <Ms2ger> | I'd rather you don't spec that ;) |
| 20:00 | AryehGregor | checks that Firefox doesn't do it |
| 20:00 | <AryehGregor> | Nope, it doesn't. |
| 20:01 | <AryehGregor> | I'm going to side with sanity here. |
| 20:01 | <AryehGregor> | I didn't try following the WebKit source code, but I wouldn't be surprised if visibility depends on some deep and magical stuff. |
| 20:02 | <TabAtkins> | Now I need to figure out how to (a) receive mail at my domain, (b) configure sendmail to pipe mail through a PHP script, and (c) use the twitter API to create a twitter feed. |
| 20:02 | <AryehGregor> | Now, should I spec exception-throwing like Firefox too, or fail silently like WebKit? Firefox's behavior matches the rest of the API a lot better, but WebKit has all the deployment . . . |
| 20:02 | <AryehGregor> | But I guess not too many people are using it yet, since only WebKit supports it, right? So I'll go with Firefox there too. |
| 20:04 | <Ms2ger> | We can fix if it doesn't work out |
| 20:06 | <Hixie> | it's fascinating to see how people think versions gave them all these features (stability, browsers implemented particular versions of specs, etc) that are completely at odds with reality |
| 20:07 | <AryehGregor> | What exception should I throw for passing an invalid argument, like "foo" instead of "move" or "extend"? Firefox does some NS_ERROR_NOT_IMPLEMENTED which doesn't appear to match any DOM exceptions. |
| 20:07 | <AryehGregor> | (I'm asking in case anyone else has had to decide this before, so I can be consistent. I'd go with NOT_SUPPORTED_ERR if there's no precedent, probably.) |
| 20:11 | <Ms2ger> | SYNTAX_ERR is what insertAdjacentHTML throws |
| 20:11 | <AryehGregor> | What spec is that in? |
| 20:12 | <Ms2ger> | HTML |
| 20:12 | <bfrohs> | Hixie, I think a lot of it has to do with the fact that it has been released in versions. I remember when I was first starting out (barely) and when I found the specs, I assumed they were written first and implemented the same across browsers. It didn't take long to figure out how wrong I was, but that was long ago. Nowadays, basic features used by beginners work fairly consistently (as far as they can tell) across browsers, so their assumption t |
| 20:12 | <bfrohs> | akes a stronger hold. Hopefully with a 'living standard' in place, new authors will realize the truth. |
| 20:13 | <AryehGregor> | SYNTAX_ERR looks good to me. |
| 20:17 | <Ms2ger> | TabAtkins, all entries seem to have the same title |
| 20:17 | <TabAtkins> | Ms2ger: There's only one entry. |
| 20:18 | <Ms2ger> | I see 4 |
| 20:18 | <TabAtkins> | view-source:http://www.xanthir.com/feed/public-css-commits/ |
| 20:18 | <TabAtkins> | Only 1. |
| 20:18 | <TabAtkins> | Get a better feed reader? |
| 20:21 | <TabAtkins> | Ms2ger: Do the entries have the same body, too? |
| 20:21 | <Ms2ger> | No |
| 20:21 | <Ms2ger> | There's three writing-modes ones and one flexbox |
| 20:22 | <Ms2ger> | grep for <entry> on view-source:http://www.xanthir.com/feed/public-css-commits/ |
| 20:23 | <TabAtkins> | Oh, I see the problem. (1) I was looking at a cached version. (2) I don't reset an array. |
| 20:23 | <TabAtkins> | One sec. |
| 20:23 | <Ms2ger> | Get a non-caching feed reader ;) |
| 20:24 | <TabAtkins> | I was looking at the source! |
| 20:24 | <TabAtkins> | Google Reader hadn't updated yet. |
| 20:24 | <TabAtkins> | Anyway, fixed. |
| 20:25 | <TabAtkins> | Hm, no, not quite right yet. |
| 20:27 | <TabAtkins> | Okay, fixed. |
| 20:37 | <Ms2ger> | Works! |
| 20:39 | <Hixie> | bfrohs: i hope so |
| 20:43 | <jgraham> | AryehGregor: You work for Hixie now? |
| 20:43 | jgraham | is behind again |
| 20:44 | <jgraham> | Also, how come hospitals have WiFi yet mobile phones are prohibited? |
| 20:44 | <TabAtkins> | To fight terrorism. |
| 20:44 | jgraham | assumes there is soome good technical reason but can't really think of it right now |
| 20:45 | <AryehGregor> | jgraham, yes, I sold my soul to him and I now have an irresistible compulsion to obey any order he gives me. |
| 20:45 | <AryehGregor> | Also, I'm working part-time at Google and he's the one in charge of my contract. |
| 20:45 | <jgraham> | Ah, nice |
| 20:45 | <jgraham> | I didn't know that |
| 20:45 | <AryehGregor> | Why would you assume there's a good technical reason? |
| 20:45 | <jgraham> | Because it's a hospital, not an airport? |
| 20:46 | <AryehGregor> | . . . so? |
| 20:46 | <AryehGregor> | I'd bet it's that someone made up a regulation at some point in the distant past when cell phones were becoming common but Wi-Fi wasn't widespread. |
| 20:46 | <AryehGregor> | (maybe not an actual law, could be hospital-specific, I dunno) |
| 20:46 | <TabAtkins> | That's my guess too. |
| 20:46 | <jgraham> | Well it seems like a somewhat-legitimate concern |
| 20:47 | <jgraham> | I was thinking that maybe mobile phones are either higher power or a frequency more likely to interfere with equipment |
| 20:47 | <jgraham> | But it could just be politics and not physics |
| 20:49 | <AryehGregor> | What frequencies do these things use? |
| 20:50 | <jgraham> | WiFi is 2.4Ghz iirc |
| 20:50 | <gsnedders> | jgraham: 802.11n has a 5GHz mode as well |
| 20:50 | <AryehGregor> | Same order of magnitude as cell phones, right? |
| 20:50 | <gsnedders> | jgraham: (dunno about anything apart from bgn, though) |
| 20:52 | <jgraham> | It must be a power thing if it is anything physicsy I think |
| 20:53 | <jgraham> | But see also http://www.timesonline.co.uk/tol/news/uk/health/article5458433.ece |
| 20:54 | <jgraham> | Anyway, time for The News Quiz |
| 20:54 | <jgraham> | (comedy satite and hence much less geeky than it sounds, for those unfamiliar with the BBC radio 4 schedule) |
| 20:54 | <jgraham> | *satire |
| 21:16 | <Hixie> | who do i speak to for gecko's localStorage stuff? anyone know? |
| 21:17 | <Hixie> | AryehGregor: 2.4GHz has an omnidirectional range of a few hundred meters at best and doesn't go through obstacles well at all, so i sure hope it's not the same as cell phones :-) |
| 21:18 | <AryehGregor> | But for instance, GSM works in frequency bands like 900, 1800, or 2100 MHz. |
| 21:19 | <Hixie> | GSM is about 800MHz and 1.9GHz; 3G is about 700MHz and 2GHz |
| 21:19 | <AryehGregor> | So how does it have much longer range? |
| 21:19 | <Hixie> | beats me |
| 21:23 | <TabAtkins> | Lower Hz = bigger wavelength = longer range and better penetration |
| 21:26 | <Hixie> | TabAtkins: yes but they're not particularly lower frequencies |
| 21:26 | <Hixie> | TabAtkins: 4G is higher, even |
| 21:26 | <TabAtkins> | Hm, I dunno then. |
| 21:32 | <Hixie> | where does webidl define what happens when i call storage[3] if storage has no indexed properties but has a named property '3' ? |
| 21:32 | <Hixie> | heycam? |
| 21:32 | <heycam> | hey Hixie |
| 21:33 | <Hixie> | oops, brb |
| 21:33 | <heycam> | there'll be no property named "3" on the object |
| 21:33 | <heycam> | so it'll just do regular property lookup and return undefined (assuming no property named "3" from the prototype) |
| 21:40 | <TabAtkins> | heycam: But Hixie just said that it has a named property '3'. So it'd just return that, right? |
| 21:41 | <heycam> | oh, misread |
| 21:41 | heycam | would have to look that up! |
| 21:43 | <heycam> | the value of the named property is returned |
| 21:44 | <heycam> | it's defined by whether a property that corresponds to the "3" named property has been created on the object |
| 21:45 | <heycam> | but that depends on the state of the object at the point "3" becomes a supported property name |
| 21:45 | <heycam> | no, I'm wrong! :D |
| 21:45 | <heycam> | http://dev.w3.org/2006/webapi/WebIDL/#named-properties |
| 21:46 | <heycam> | step 4 says to never create named properties that look like array indexes, if the object supports indexed properties |
| 21:47 | <dglazkov> | heycam: where are my use caseseeses! |
| 21:47 | dglazkov | sure is demanding |
| 21:47 | <heycam> | dglazkov, yeah sorry we haven't had time to look at it! |
| 21:47 | <dglazkov> | :) |
| 21:47 | <heycam> | are you working to a short schedule? |
| 21:47 | <dglazkov> | heycam: no, I just like to nag |
| 21:47 | <heycam> | :) |
| 21:48 | <heycam> | it's likely something we're not going to get some time to look for the next month or so |
| 21:48 | <heycam> | at least, I won't be able to |
| 21:49 | <dglazkov> | ok, good to know |
| 21:49 | <dglazkov> | I'll nag in a month |
| 21:49 | <heycam> | we (SVG WG)'ve got a F2F at the end of February, so I will try to make sure we discuss it then |
| 21:50 | <heycam> | dglazkov, are you doing the implementation in WebKit? |
| 21:50 | <dglazkov> | can you discuss rolling SVG into HTML then? :P |
| 21:50 | <dglazkov> | heycam: along with other peeps |
| 21:50 | <heycam> | yeah, sure :) |
| 21:50 | <heycam> | cool cool |
| 22:09 | <Hixie> | heycam: but if it _doesn't_ support indexed properties, and you index by numeric value 3, and there's a named property with string value "3", you'll get the value of the named property? |
| 22:09 | <Hixie> | heycam: i couldn't find where the spec said that the number became a string |
| 22:09 | <heycam> | Hixie, yes that's right. it's more like the number was always a string... |
| 22:10 | <heycam> | property names always being strings in JS |
| 22:10 | <Hixie> | [3] gets converted to a string and then back to a number if there's indexed properties? |
| 22:10 | <heycam> | yes, it looks up the property named "3" |
| 22:10 | <heycam> | just like on native JS arrays |
| 22:10 | <Hixie> | interesting |
| 22:10 | <Hixie> | ok |
| 22:10 | <Hixie> | thanks |
| 22:10 | <heycam> | implementations often optimize though, to avoid unnecessary parsing/conversion |
| 22:11 | <heycam> | np |
| 22:11 | <Hixie> | heycam: hey while i have you here, any idea who on gecko works on localStorage? I think we'll have to drop the indexed properties and leave only the named ones because gecko's the only one who followed the spec and did the indexed ones. |
| 22:11 | <Hixie> | and it's not a backwards compatible change for the others |
| 22:11 | <heycam> | Hixie, I don't know, but surely sicking should know |
| 22:11 | <Hixie> | (and one of the others is IE, so it's a long-term thing if it does have to change) |
| 22:11 | <Hixie> | sicking: ping |
| 22:11 | <sicking> | Hixie: short-pong |
| 22:11 | <sicking> | gotta run in 5 |
| 22:12 | <Hixie> | sicking: webstorage - who do i speak to at geckoland? |
| 22:12 | <sicking> | Hixie: honza iirc |
| 22:12 | <Hixie> | thanks |
| 22:12 | <sicking> | :mayham in our bugzilla |
| 22:13 | <Hixie> | k |
| 22:57 | <Yuhong> | Hixie: I was reading your articles about tag soup. Have you heard of mangleme? |
| 23:08 | <TabAtkins> | Dammit, Hixie, why'd you have to write the Lists spec with 2-space indentation? Now I've gotta reformat the whole thing. >_< |
| 23:14 | <Hixie> | TabAtkins: o_O |
| 23:15 | <TabAtkins> | I can't write html with spaces for indentation. I just can't. But I can't have an inconsistent spec, either. |
| 23:16 | <TabAtkins> | Jeez, and you're internally inconsistent, too. >_< |
| 23:18 | <benschwarz> | Hixie: !! |
| 23:19 | <benschwarz> | I hope you're wearing socks |
| 23:19 | <TabAtkins> | So they can be blown off? |
| 23:19 | <benschwarz> | My man! |
| 23:19 | <benschwarz> | haha |
| 23:20 | <benschwarz> | check the dev spec |
| 23:20 | <Hixie> | benschwarz: hey sup |
| 23:20 | <Hixie> | benschwarz: looking |
| 23:20 | <Hixie> | TabAtkins: what do you use instead? |
| 23:21 | <TabAtkins> | Tabs, like sane people? |
| 23:21 | bfrohs | is insane for using single-space indentation |
| 23:22 | <TabAtkins> | Yes you are. Also: evil. |
| 23:22 | <bfrohs> | Apologies, evil as well hahah |
| 23:22 | <Hixie> | tabs are crazy, because different people have different tab stop definitions, unless you use the default, which is 8 chars and is FAR too wide for sane readability |
| 23:22 | <jamesr____> | tabs are evil |
| 23:22 | <bfrohs> | I'm with Hixie on this one |
| 23:22 | <Hixie> | personally i prefer single-space indentation for markup |
| 23:23 | <Hixie> | sicking: apparently i want meyhemer, not honza; honza is apparently your firebug guy |
| 23:23 | <TabAtkins> | Um, what. The entire *point* of using tabs is that they're configurable. |
| 23:23 | <bfrohs> | I can't do anything other than single-space... because of tab issues in different editors and pressing a key more than once (and readability) |
| 23:23 | <sicking> | Hixie: they're both Honza |
| 23:23 | <Hixie> | oh. |
| 23:23 | <Hixie> | wait |
| 23:23 | <Hixie> | what? |
| 23:23 | Hixie | gets a headache |
| 23:23 | <benschwarz> | Hixie: the search, the search damnit! |
| 23:23 | <bfrohs> | TabAtkins: some editors change tabs to spaces though :( |
| 23:23 | <sicking> | Hixie: there are multiple Honzas |
| 23:24 | <TabAtkins> | bfrohs: Those editors are broken and stupid, and I don't use them. |
| 23:24 | <Hixie> | benschwarz: what am i looking for exactly? |
| 23:24 | Hixie | pat pats TabAtkins |
| 23:24 | <benschwarz> | 1. Open your eyes :) |
| 23:24 | <benschwarz> | Hixie: the search, top right of every page |
| 23:24 | <Hixie> | wow i didn't even see that |
| 23:24 | <benschwarz> | although, I do need to re-browser test |
| 23:25 | <Hixie> | heh, searching for "introduction" isn't helpful :-P |
| 23:25 | <sicking> | Hixie: to make it more fun. They're both called "Jan". Apparently Honza is a common check nickname for Jan |
| 23:25 | <franksalim> | http://www.jwz.org/doc/tabs-vs-spaces.html |
| 23:25 | <TabAtkins> | Anyway, changing everything over to tabs, since you're inconsistent between 1 and 2 spaces. |
| 23:25 | <Hixie> | so long as i never have to edit it again, wfm :-) |
| 23:25 | <franksalim> | also, TabAtkins, have a look at the Smart Spaces plugin for gedit, which is not broken nor stupid :-) |
| 23:27 | <TabAtkins> | franksalim: I can already unindent like if I was using tabs, because I'm using tabs. |
| 23:27 | <Hixie> | benschwarz: can you make the arrow keys not scroll while the search thing is up? |
| 23:28 | <benschwarz> | Hixie: we can do a bunch of things, but its a line in the sand, anyway |
| 23:28 | <benschwarz> | I need to figure out the mobile issues that popped up after it gets minified |
| 23:30 | <Hixie> | sicking: the Honza I got in contact with told me the other Jan's nick was meyhemer, but there's no meyhemer online |
| 23:30 | <Hixie> | sicking: i think i'll just make this change to the spec and cc him... do you know what his e-mail address is? |
| 23:30 | <franksalim> | TabAtkins: at the cost of having barbaric 0x09s in your files |
| 23:31 | <TabAtkins> | My namesake! |
| 23:31 | <heycam> | SpaceSpaceSpaceSpaceAtkins |
| 23:42 | <paul_irish> | Hixie: i'm working with ben to fix that scrolly action. it was bugging me too |
| 23:43 | <Hixie> | cool |
| 23:50 | <cying> | what is the best way to specify the title of an audio song in an html5 audio tag? |
| 23:51 | <cying> | e.g: <audio src="foo.mp3"></audio> <!-- need to have title specified as "This is the Song Title" --> |
| 23:51 | <jamesr____> | heycam: i wanna document my proposed modifications to requestAnimationFrame. should i just wget the sources from your mozilla web page, modify Overview.xml, and run 'make'? |
| 23:52 | <Hixie> | cying: title="" |
| 23:52 | <TabAtkins> | cying: <h1>This is the song title</h1><audio src=foo.mp3> |
| 23:52 | <heycam> | jamesr____, let me check that all the sources are there |
| 23:52 | <Hixie> | cying: or, if you're using <figure>, <figcaption>. |
| 23:52 | <jamesr____> | if i do that is there any good way to diff things? |
| 23:52 | <heycam> | jamesr____, yeah they're all there. just run make. (it just relies on xsltproc.) |
| 23:52 | <cying> | Hixie: interesting i didn't see that in the audio spec... is it somewhere else? |
| 23:53 | <heycam> | jamesr____, hmm in the past I have used daisydiff, an html diff tool |
| 23:53 | <heycam> | jamesr____, which works as long as the documents aren't massive |
| 23:53 | <jamesr____> | heycam: cool. honestly i'm mostly gonna delete stuff |
| 23:53 | <Hixie> | cying: see what? |
| 23:53 | <heycam> | jamesr____, progress is being made on where the spec is going to live, but if you want to send comments before it's up officially, please send them to public-fx |
| 23:53 | <jamesr____> | heycam: will do |
| 23:54 | <heycam> | jamesr____, yeah the spec at the moment is a bit more convoluted than it need to be, probably, in terms of managing the timing of animations |
| 23:54 | <cying> | Hixie: the title attribute... but i found it in the table of contents |
| 23:54 | <jamesr____> | yeah i'm not totally happy with what it says about declarative animations |
| 23:54 | <cying> | Hixie: i wasn't sure if <track> was appropriate for this situation |
| 23:54 | <heycam> | yeah I think it promises things which you can't actually do |
| 23:54 | <jamesr____> | verily |
| 23:55 | <Hixie> | cying: track is for captions and subtitles |
| 23:55 | <jamesr____> | heycam: so my preference would be to spec out the imperative animation stuff by itself |
| 23:55 | <heycam> | jamesr____, ok it might make sense to start from that (since it's the more important part anyway) |
| 23:56 | <jamesr____> | right |
| 23:58 | <cying> | Hixie: very interesting. thanks! i'll add that behavior in |
| 23:59 | <jamesr____> | heycam: i'd also like to add a second parameter to requestAnimationFrame of type Element that is the element the animation is associated with |
| 23:59 | <jamesr____> | heycam: the idea being that if that the UA could tell that element was not visible to the user, it could avoid invoking the callback |
| 23:59 | <heycam> | jamesr____, what would that be used for? |
