| 00:17 | <dglazkov> | Hixie: 4.10.21.3 form submission algorithm is inspired by ...? |
| 01:44 | <KaOSoFt> | Looking at the current specification, how can I know if an element is block or in-line? |
| 01:52 | <divya> | KaOSoFt: are you talking about content models or the CSS rendering? |
| 01:55 | <KaOSoFt> | Well, I'm not versed in this technical chat, but I guess I'm talking about a content model. Is <section> a "block" element? Like <div> and <span> (block and in-line). |
| 01:56 | <divya> | KaOSoFt: there are no more "block" and "inline" http://www.whatwg.org/specs/web-apps/current-work/multipage/section-index.html#element-content-categories |
| 01:56 | <KaOSoFt> | Oh, I see, that's why I didn't find any information about it. |
| 01:57 | <KaOSoFt> | Thank you very much, you kind sire. |
| 01:57 | <KaOSoFt> | :) |
| 01:57 | <KaOSoFt> | And goodbye, time to go back home. |
| 01:57 | <divya> | KaOSoFt: :) bye |
| 02:59 | <nessy> | Hixie: how is progress on the captions stuff? |
| 03:03 | <micheil> | where's that interactive tool for viewing the works in progress located again? |
| 03:17 | <Hixie> | nessy: pretty good. I've been trying to figure out which format would be best to use or extend: http://wiki.whatwg.org/wiki/Timed_track_formats |
| 03:21 | <othermaciej> | Hixie: people have claimed that the smilText subset of SMIL can be used as a timed text format |
| 03:21 | <othermaciej> | I'm not sure anyone identified a well-defined SMIL profile for this purpose |
| 03:21 | <othermaciej> | (I mention this because your page says "not a caption format") |
| 03:23 | <othermaciej> | http://www.w3.org/TR/SMIL/smil-text.html |
| 03:23 | <othermaciej> | looks like a weird mix of SMIL and DXFP |
| 03:23 | <Hixie> | oh that's not the same as DXFP? |
| 03:23 | <Hixie> | good to know |
| 03:23 | <Hixie> | i'll add it to the list |
| 03:25 | <othermaciej> | no, DFXP (aka TTML aka ttaf1-dfxp) is this: http://www.w3.org/TR/ttaf1-dfxp/ |
| 03:25 | <Hixie> | good lord how many names can one standard have |
| 03:26 | <othermaciej> | I know, right? |
| 03:26 | <othermaciej> | and TTML is the only one I can type without scrambling the letters |
| 03:27 | <othermaciej> | and it's the title of the spec, but does not even appear in the shortname, and the Status of this Document section uses a different title to refer to it than the actual title |
| 03:27 | <Hixie> | relatedly: in doing my research i keep coming across the assertion that RealText is SMIL. e.g. wikipedia actually had RealText be a hyperlink to the SMIL page. But I can't find anything that actually supports that. |
| 03:27 | <Hixie> | i changed the wikipedia page |
| 03:29 | <othermaciej> | this seems like a really confusing area |
| 03:30 | <Hixie> | yeah |
| 03:58 | <nessy> | TTML was a last-minute invention |
| 03:58 | <Hixie> | the name? |
| 03:58 | <nessy> | I wished they had kept DFXP because then TTML would have been free to use for something else |
| 03:58 | <nessy> | yes, the name |
| 03:58 | <Hixie> | heh |
| 03:59 | <nessy> | cause … I kinda like the name timed text markup language ! |
| 03:59 | <Hixie> | DFXP is a pretty bad name |
| 03:59 | <nessy> | all historic and academic |
| 03:59 | <Hixie> | seems like pretty much every name has been taken already |
| 03:59 | <nessy> | so about SMIL ... |
| 04:00 | <nessy> | SMILText is a spec they developed only rather recently |
| 04:00 | <nessy> | RealText was there before |
| 04:00 | <nessy> | and it was based on a subset of an earlier SMIL spec |
| 04:00 | <nessy> | unfortunately "SMIL" is being used for all sorts of things, when it is really always just a tiny subpart of SMIL |
| 04:01 | <nessy> | DAISY for example takes a subpart and extends it |
| 04:01 | <Hixie> | RealText doesn't seem to have anything to do with SMIL, does it? Or does using "begin" and "end" attributes in an XML format mean that you're using SMIL? |
| 04:01 | <nessy> | the Microsoft server manifest for Smooth Streaming is a tiny subpart of SMIL and is still called a "SMIL" use |
| 04:01 | <nessy> | lol |
| 04:02 | <Hixie> | i'm not even sure RealText is actually XML |
| 04:02 | <Hixie> | and the microsoft thing is definitely not XML |
| 04:02 | <Hixie> | so apparently XML isn't even a requirement |
| 04:02 | <nessy> | anything on the Web that used to do anything with media was called a SMIL technology at one stage |
| 04:02 | <nessy> | it's all being handled in a very loose manner |
| 04:03 | <nessy> | (which annoys the crap out of me, but don't tell anyone) |
| 04:03 | <roc> | it's great, everyone uses SMIL so browsers should too |
| 04:03 | <annevk> | that's certainly the TTML sentiment |
| 04:04 | <Hixie> | so should I just put SMIL into the title of whatever we're doing and then pretend we're doing SMIL too? |
| 04:04 | <nessy> | I think RealText was the way to synchronise text with SMIL before SMILText was created |
| 04:04 | <nessy> | Hixie: that would certainly solve the political situation! |
| 04:04 | <annevk> | hmm, the Ubuntu countdown went from 1 day left to available soon, but what does that mean? |
| 04:04 | <nessy> | suddenly SMIL would be a success on the Web |
| 04:04 | <Hixie> | we can call it "SCF" and say it stands for "SMIL Captioning Format" (but really it would stand for "Subtitles and Captions Format") |
| 04:04 | <nessy> | anyway - I am sarcastic today, pardon my mood |
| 04:05 | <Hixie> | don't worry, sarcasm is a prerequisite for being in the channel |
| 04:05 | <Hixie> | how else would the false controversies emerge but for our sarcasm |
| 04:06 | <nessy> | I'm sure someone will hold it in my face at the right opportunity |
| 04:07 | <nessy> | btw 3GPP-TT is a subpart of an early version of DFXP |
| 04:07 | <nessy> | years ago I had a student do an analysis of available subtitle formats - you can still find the result here: https://trac.annodex.net/wiki/CmmlSubtitles |
| 04:08 | <nessy> | Microsoft's SAMI wasn't actually that bad! |
| 04:08 | <Hixie> | SAMI is a disaster from the point of view of us speccing it and/or extending it |
| 04:08 | <nessy> | it's not support any more anyway |
| 04:08 | <roc> | I just found out that "Adobe Digital Editions" supports an "extended EPUB" ebook format ... which mixes XHTML, CSS and XSL-FO. And whoever produced it is kinda bitter about CSS3 |
| 04:08 | <Hixie> | we'd have to reverse engineer their HTML-like parser |
| 04:09 | <Hixie> | roc: url? |
| 04:09 | <nessy> | EPUB is produced by the DAISY guys |
| 04:09 | <roc> | http://blogs.adobe.com/digitaleditions/template.html ... "Since CSS group chose to ignore a lot of perevious work in this area (e.g. inventing its own expression syntax for media queries instead of using XPath, designing its own incompatible replacement of page master instead of building on top of XSL:FO)" |
| 04:09 | <Hixie> | nessy: (will check that link, thanks) |
| 04:09 | <nessy> | (see if there is anything useful there for you) |
| 04:10 | <Hixie> | roc: huh. i wonder who wrote that. |
| 04:26 | <roc> | Peter Sorotokin |
| 04:29 | <Hixie> | ah, interesting |
| 04:29 | <Hixie> | he's the guy with whom i had a lot of, ah, discussions, in the sxbl days |
| 04:30 | <annevk> | comparing Selectors with XPath makes sense I suppose, but Media Queries and XPath? |
| 04:31 | <othermaciej> | wait, why would you use XPath for Media Queries? |
| 04:31 | othermaciej | arches both eyebrows |
| 04:31 | <othermaciej> | s/why/how/? |
| 04:31 | <othermaciej> | and criticizing CSS for not building on top of XSL FO? |
| 04:31 | <othermaciej> | my head is exploding |
| 04:33 | <roc> | someone recently told me we should build stuff on XSL-FO instead of CSS because XSL-FO is a REC and CSS 2.1 isn't |
| 04:44 | <annevk> | interesting how hard it is to replace the HTML parser |
| 04:44 | <annevk> | initially I wouldn't have thought that, but I guess I didn't know much about how intertwined it was |
| 04:52 | <Hixie> | annevk: in opera you mean? |
| 04:55 | <annevk> | Gecko, we haven't quite started |
| 05:01 | <roc> | I'm sure Henri can tell you about all kinds of problems |
| 05:01 | <roc> | :-) |
| 05:02 | <othermaciej> | I'm glad Mozilla is biting the bullet on this first |
| 05:02 | <annevk> | yeah, it's pretty cool |
| 05:03 | <othermaciej> | seems like the best way to flush out potential problems with the parsing algorithm itself |
| 05:52 | <Hixie> | annevk: i expect opera will have a harder time, because of the way misnested tags are currently handled |
| 05:59 | <annevk> | I doubt that is the hardest part, but we'll find out |
| 06:03 | <boogyman> | Hixie: I am not sure if this was answered before, but what would happen if the server side attempted to assign a value to an <input> or <textarea> greater than that afforded by an explicit maxlength value |
| 06:05 | <Hixie> | anyone here at WWW2010? |
| 06:05 | <boogyman> | originally queried at http://krijnhoetmer.nl/irc-logs/whatwg/20100427#l-928 |
| 06:06 | <Hixie> | boogyman: you mean in the markup? like <input value=123 maxlength=2> ? |
| 06:07 | <boogyman> | Hixie: yes, that's an appropriate example |
| 06:12 | <Hixie> | let me see... |
| 06:12 | Hixie | opens the spec |
| 06:13 | <Hixie> | nothing in particular seems to happen if it's set by the server |
| 06:13 | <Hixie> | it'll submit fine and it doesn't get truncated |
| 06:14 | <Hixie> | it probably doesn't bode well that the W3C is explicitly describing itself as split into two camps, RDF and HTML5... http://www.w3.org/2010/04/w3c-track.html |
| 06:14 | <boogyman> | so it should behave as if the 'user' has entered the text |
| 06:15 | <othermaciej> | Hixie: I don't think that's how they meant it |
| 06:15 | <othermaciej> | Hixie: I assume they mean "camp" in the sense of "something like Foo Camp" |
| 06:18 | <Hixie> | boogyman: no, if the user enters the text then when they submit it they'll get a message saying the text is too long |
| 06:18 | <zcorpan> | Hixie: i think the way misnested tags are currently handled will just be dead code when we have an html5 parser, so i think that won't be hard at all to remove :) |
| 06:18 | <Hixie> | true |
| 06:19 | <Hixie> | zcorpan: btw re the websocket stuff |
| 06:19 | <Hixie> | still going through my mail from having been off sick for a day |
| 06:19 | <Hixie> | but |
| 06:19 | <Hixie> | i don't think it matters where you drop the connection if you drop the connection |
| 06:19 | <Hixie> | since it's black-box indistinguishable |
| 06:20 | <Hixie> | however, you shouldn't use your HTTP parser, since it does different things (like continuation lines) |
| 06:20 | <Hixie> | and it might do error handling differently |
| 06:20 | <Hixie> | since HTTP doesn't say how to do error handling |
| 06:20 | <zcorpan> | Hixie: it's not black-box indistinguishable if the server sleeps half-way |
| 06:20 | <zcorpan> | Hixie: yeah |
| 06:21 | <Hixie> | the client is always allowed to insert aribitrary sleeps anywhere |
| 06:21 | <Hixie> | nothing in the spec defines the performance characteristics of a websocket client |
| 06:22 | <zcorpan> | Hixie: if the server responds "HTTP/1.1 200" and then sleeps, we want to abort but the spec doesn't allow it |
| 06:22 | <zcorpan> | since we haven't seen an 0x0A byte yet |
| 06:24 | <zcorpan> | continuation lines is something i should test |
| 06:28 | <Hixie> | hm yeah i should allow the UA to break early if it can prove that it wouldn't ever accept the connection |
| 06:28 | <Hixie> | assume that i'll make that allowed |
| 06:29 | <Hixie> | i hadn't thought of breaking early |
| 06:30 | <zcorpan> | Hixie: thanks |
| 06:30 | <Hixie> | in case i forget by the time i reply to your e-mail, just remind me of the above :-) |
| 06:31 | <zcorpan> | ok |
| 06:31 | <Hixie> | it won't mean i changed my mind or anything, just that i forgot |
| 06:36 | <Hixie> | man, some reporters ask the most inane questions |
| 06:41 | <boogyman> | agreed |
| 06:42 | <zcorpan> | such as "when will html5 be finished?"? |
| 06:42 | <Hixie> | no, that one i know how to answer |
| 06:42 | <annevk> | wir |
| 06:42 | <Peter`> | They probably wouldn't believe you, since browser vendors have already implemented HTML 5! |
| 06:42 | <Hixie> | i mean things like "do people use all of html4?" |
| 06:42 | <Peter`> | Gotta love marketing.. |
| 07:01 | <Hixie> | wow, check out the number of errors in this article |
| 07:01 | <Hixie> | http://www.instantshift.com/2010/04/26/html5-worth-the-hype/ |
| 07:02 | <boogyman> | 157 html errors lol |
| 07:02 | <Hixie> | oh i didn't mean validation errors |
| 07:02 | <Hixie> | i meant in the text |
| 07:13 | <zcorpan> | "To use HTML5 you will need to use a specific scripting API such as:" |
| 07:14 | <JonathanNeal> | "javascript" |
| 07:14 | <hsivonen> | most of the problems I've been dealing with haven't been about the HTML5 parsing algorithm but about our test cases being bogus in a way that has stayed latent with the particular timing quirks of Gecko's old parser |
| 07:14 | <JonathanNeal> | "4chan" |
| 07:14 | <zcorpan> | seems like they copied some from wikipedia and edited it so it would be less obvious that they copied it from wikipedia |
| 07:23 | <annevk> | Ivan! |
| 07:29 | <zcorpan> | yeah i wonder who ivan is |
| 07:29 | <zcorpan> | maybe Hixie should change his irc nick to Ivan |
| 07:29 | <othermaciej> | ivan who? |
| 07:30 | <zcorpan> | Ivan Hickson |
| 07:30 | <boogyman> | lol |
| 07:30 | othermaciej | chortles |
| 08:02 | <hsivonen> | zcorpan: I filed the REPLACEMENT CHARACTER bug as https://bugzilla.mozilla.org/show_bug.cgi?id=562590. thanks. |
| 08:11 | <annevk> | http://blogs.msdn.com/ie/archive/2010/04/28/product-feedback-systems.aspx o_O |
| 08:11 | <annevk> | Microsoft seems to be upping the FUD machine recently |
| 08:11 | <annevk> | though maybe it is just me |
| 08:39 | <Hixie> | annevk: that does seem a bit rude |
| 08:40 | <Hixie> | but they have a point |
| 08:40 | <zcorpan> | hsivonen: other cases you might want to test are %e0%80, %f0%80%80, %f5%80%80%80 etc |
| 08:44 | <hsivonen> | zcorpan: ok |
| 08:46 | <othermaciej> | what is the FUD there? |
| 08:46 | <othermaciej> | "bugzilla will cause you to be spammed"? |
| 08:48 | <Hixie> | i don't think it's FUD |
| 08:48 | <othermaciej> | ok, having read more of it, it looks like they are pretty directly calling out at least some other browser vendors |
| 08:48 | <othermaciej> | I am surprised Microsoft would post such direct competitive statements on the IE blog |
| 08:49 | <Hixie> | it seems like a dangerous game for them to play, because if they piss off a browser vendor doing this kind of thing, there's plenty of material that can be thrown back at them |
| 08:50 | <othermaciej> | one thing he overlooks is that the WebKit project completely ignores the difference between "new" and "unconfirmed" |
| 08:52 | <hsivonen> | In my experience, Firefox and WebKit nightlies as well as Opera roughly-weeklies are quite usable enough to expose to interested parties |
| 08:52 | <Hixie> | i only use nightly builds |
| 08:52 | <othermaciej> | I doubt anyone from the WebKit project is interested in getting into a pissing match with Microsoft over who manages feedback better |
| 08:52 | <Hixie> | of webkit, chrome, firefox, and opera -- i haven't even got a stable build of a browser on my machine, as far as i know |
| 08:53 | <othermaciej> | is it a Mac? |
| 08:53 | <Hixie> | yeah |
| 08:53 | <othermaciej> | you probably have a stable build of Safari and of the associated WebKit |
| 08:53 | <othermaciej> | even if you don't browse with it |
| 08:53 | <Hixie> | ah probably, yeah |
| 08:53 | <zcorpan> | i think i've had higher FIXED rate when filing bugs at mozilla compared to filing bugs at microsoft |
| 08:54 | <Hixie> | i think that comes up when i accidentally pick "search in google" from a context menu |
| 08:54 | <othermaciej> | or rather, you do browse with that Safari, cause that's what the WebKit nightly launcher uses |
| 08:54 | <Hixie> | (grrr @ apple) |
| 08:54 | <othermaciej> | you can make the nightly launcher your default browser I believe |
| 08:54 | <zcorpan> | all microsoft feedback was "actioned", but i don't care if a bug is "RESOLVED POSTPONED" or still unconfirmed |
| 08:54 | <Hixie> | can i make chrome the default? :-) |
| 08:54 | <Hixie> | or firefox? :-) |
| 08:54 | <othermaciej> | zcorpan: well, "postponed" means it was considered and rejected, rather than just not acted on it yet |
| 08:54 | <othermaciej> | Hixie: yes |
| 08:54 | <Hixie> | how? |
| 08:55 | <othermaciej> | you can make any browser your default browser |
| 08:55 | <othermaciej> | they have built-in UI, and Safari also has UI in prefs that lets you make any browser the default browser |
| 08:55 | <Hixie> | so that "search in google" launches that browser? |
| 08:55 | <hsivonen> | Hixie: sometimes the setting resets back to safari at login, in which case you have to go through various .plists or whatever and delete the right one |
| 08:55 | <othermaciej> | I actually don't know how the "search in google" context menu is coded, it's in AppKit I think |
| 08:56 | <othermaciej> | it's possible that it doesn't correctly respect the default browser setting, unlike nearly anything else on the system |
| 08:56 | <zcorpan> | othermaciej: yeah. although for at least one of my bugs at microsoft it was clear that they hadn't considered the bug at all when closing it |
| 08:56 | <hsivonen> | I managed to delete the right thing on my work Mac but on my home Mac, I still need to make Minefield the default after every login |
| 08:56 | <Hixie> | (i actually don't use the nightly launcher btw, i just use whatever it is that causes webkit to offer me a prompt to update the build) |
| 08:56 | <othermaciej> | zcorpan: closing something without giving it actual consideration is worse than leaving it "unconfirmed" or "new" :-/ |
| 08:57 | <othermaciej> | Hixie: I believe the nightly launcher updates itself automagically now |
| 08:57 | <Hixie> | othermaciej: yeah, looks like it's ignoring the seeting. I dragged a favicon from webkit to the dock, then clicked on it, and chrome opened. So I assume chrome is the default. |
| 08:57 | <othermaciej> | Hixie: yes, probably |
| 08:58 | <Hixie> | anyway, didn't mean to turn this into a gripefest :-) |
| 08:58 | <othermaciej> | Hixie: if you open Safari (or a WebKit nightly) and look at general preferences, it will tell you what the system default is supposed to be |
| 08:58 | <othermaciej> | WebKit nightlies do tend to be reasonably stable, though there is the occasional build with a bad bug |
| 08:59 | <hsivonen> | well, recently with Minefield, I lost xkcd tooltips for a few days |
| 08:59 | <hsivonen> | and then on Mac if you have Flash 10.1 installed, things aren't so great right now |
| 09:00 | <Hixie> | i had a chrome build recently where using a <select> made <textarea>s ignore any function keys (like backspace or arrow keys), which was fun while that lasted |
| 09:00 | <Hixie> | but that's why i have four browsers! |
| 09:00 | <Hixie> | i wish opera nightlies autoupdated, they don't seem to |
| 09:00 | <hsivonen> | so I went back to Flash 10.0.x.x on the computer where I want Flash to actually work (after filing my stack trace of course) |
| 09:01 | <hsivonen> | Hixie: for me on Mac, Opera autodownloads updates and then fails to install them |
| 09:01 | <Hixie> | i've seen that too |
| 09:01 | <Hixie> | but more recently it just doesn't seem to download anything at all |
| 09:02 | <zcorpan> | Hixie: have you checked opera:config#AutoUpdate|DownloadAllSnapshots ? |
| 09:02 | Hixie | looks |
| 09:03 | <Hixie> | it is unchecked |
| 09:03 | <Hixie> | do i just check it? |
| 09:03 | <Hixie> | how does it know to update to the nightlies and not release builds? |
| 09:03 | <zcorpan> | yep, it should enable autoupdate of weeklies |
| 09:03 | <hsivonen> | zcorpan: is it supposed to work on Ubuntu? |
| 09:04 | <zcorpan> | Hixie: it knows by looking at the pref |
| 09:04 | <zcorpan> | hsivonen: i think so, but haven't tested |
| 09:04 | <Hixie> | oh, that's what snapshot means |
| 09:04 | <Hixie> | cunning |
| 09:04 | <Hixie> | oooh |
| 09:04 | <Hixie> | hey awesome |
| 09:04 | <Hixie> | thanks dude |
| 09:04 | <othermaciej> | hsivonen: Flash 10.1 has some bugs in its CA drawing model code, which is only enabled for WebKit nightlies currently |
| 09:04 | <zcorpan> | Hixie: it works? |
| 09:05 | <othermaciej> | hsivonen: there are also, I think, some bugs in our side of the code that didn't get flushed out with previous plugins |
| 09:05 | <hsivonen> | othermaciej: I believe I saw CA on the Minefield+OOPP+Flash 10.1 stack I filed |
| 09:06 | <othermaciej> | maybe they are trying to use it on Minefield and failing |
| 09:06 | <hsivonen> | this was after out-of-process CA was enabled in Minefield, but clearly they are failing |
| 09:07 | <Hixie> | zcorpan: yup |
| 09:08 | <hsivonen> | iframes created with innerHTML fire load events normally, right? |
| 09:16 | hsivonen | wonders why so many test cases try to avoid simply listening for the load event and do something else and incorrect |
| 09:17 | <Hixie> | no spec before html5 defined innerHTML or load events, so it's not hard to imagine that people wouldn't be sure of what was the right thing |
| 09:17 | <hsivonen> | s/listening/waiting/ or s/for/to/ I guess |
| 09:19 | <annevk> | zcorpan, I don't believe that works on Ubuntu |
| 09:19 | <annevk> | zcorpan, as we integrate with the system there |
| 09:20 | <annevk> | but maybe... |
| 09:31 | <zcorpan> | Hixie: nice |
| 09:53 | zcorpan | files https://connect.microsoft.com/IE/feedback/details/555265/video-should-not-fall-back-even-if-the-video-fails |
| 10:01 | <JonathanNeal> | Good bug to file :) |
| 10:07 | zcorpan | gets an email saying: |
| 10:07 | <zcorpan> | This is a text part of the message. |
| 10:07 | <zcorpan> | It is shown for the users of old-style e-mail clients |
| 10:07 | <zcorpan> | why thank you, that's helpful |
| 10:08 | <hsivonen> | zcorpan: does ie9 have the blue border around linked images by default? |
| 10:09 | <zcorpan> | hsivonen: don't know |
| 10:40 | <roc> | does the IE9 preview actually support video now? |
| 10:41 | <Philip`> | No |
| 10:41 | <Philip`> | though I guess it's about time for them to release a new preview now |
| 10:52 | <hsivonen> | when I try to sign in to see zcorpan's bug report, I get The content that you requested cannot be found or you do not have permission to view it. |
| 10:54 | <Philip`> | hsivonen: Works for me and says "Access Restriction: Public" |
| 10:55 | <Philip`> | (Apparently I am a participant in the Internet Explorer Tech Feedback Program which may make a difference) |
| 10:56 | <hsivonen> | around the IE8 time, IIRC, I decided against enrolling into that program due to terms that would have been unwise to accept while working on a competing product |
| 10:58 | <hsivonen> | apparently now there's a lesser level of Microsoft Connect access available under less scary terms |
| 13:45 | <Dashiva> | On the .URL topic, what about also adding document.url? It leaves back compat in place, but lets authors standardize on .url for everything in new content |
| 14:08 | <jgraham> | http://www.whatwg.org/specs/web-apps/current-work/#dom-context-2d-drawimage does that actually say in normative text that TYPE_MISMATCH_ERROR should be raised if the first argument is the wrong type (other than null) |
| 14:10 | <Philip`> | jgraham: WebIDL says it must |
| 14:10 | <Philip`> | (hopefully) |
| 14:10 | <Philip`> | (HTML5 used to say it but it was removed because it's redundant with WebIDL, I believe) |
| 14:11 | <jgraham> | Ah, I hoped it would be somrthing like that |
| 14:25 | <GPHemsley> | So... Team Rocket started off the discussion on headers... |
| 14:38 | <jgraham> | zcorpan: yt? |
| 14:57 | <remysharp> | Am I right in saying that Web Workers *aren't* spec'ed to use JSON to automatically convert objects - whereas Firefox have implemented this as an extra: https://developer.mozilla.org/En/Using_web_workers#Sending_objects_to_workers |
| 14:57 | <micheil> | oh, hey remy |
| 14:58 | <remysharp> | micheil: hi there. |
| 14:58 | <micheil> | remysharp: jsbin wasn't it? |
| 14:58 | <remysharp> | yep |
| 14:58 | <micheil> | yeah, nice work there |
| 14:58 | <remysharp> | cheers :) |
| 14:59 | <Philip`> | remysharp: As far as I'm aware, it should use http://www.whatwg.org/specs/web-apps/current-work/multipage/urls.html#safe-passing-of-structured-data to transfer objects |
| 15:00 | <remysharp> | Philip`: so in fact it shouldn't be giving me a string on the other side each time? |
| 15:00 | <remysharp> | it should be able to handle some objects and types |
| 15:02 | <gsnedders> | Anyone have IE9 at hand? |
| 15:02 | <Philip`> | remysharp: I think so, unless I'm mistaken |
| 15:02 | <remysharp> | Philip`: cheers - shame most of the browsers don't do that yet, but I'm sure they'll catch up! |
| 15:03 | <Philip`> | Looks like Workers use http://www.whatwg.org/specs/web-apps/current-work/multipage/comms.html#dom-messageport-postmessage |
| 15:03 | <Philip`> | which uses that structured clone algorithm |
| 15:03 | Philip` | wishes it wasn't so hard to find definitions of these functions |
| 15:03 | <Philip`> | (since the Workers spec doesn't seem to have any direct link to it) |
| 15:04 | <micheil> | hmm.. did html5 introduce a module tag? |
| 15:04 | <micheil> | / element |
| 15:05 | <gsnedders> | Philip`: Is that because of lack of xdoc xref? |
| 15:06 | <micheil> | actually no. looks like site is using their own xml schema |
| 15:09 | <jgraham> | gsnedders: Everything is your fault |
| 15:15 | gsnedders | looks innocent |
| 15:15 | <micheil> | gsnedders: hmm.. mozilla dev, yes? |
| 15:16 | <micheil> | or atleast, their irc channels. |
| 15:16 | <TabAtkins> | Oh, wow. This article Ian linked to a few hours ago really *is* just a copypastad and lightly editted wikipedia article. |
| 15:16 | <TabAtkins> | If not, then this person is *really bad* at writing. |
| 15:18 | <gsnedders> | micheil: Opera QA, ocasionally on irc.mozilla.org |
| 15:18 | <micheil> | gsnedders: ah, okay, I've just seen the name about |
| 15:19 | <hsivonen> | micheil: isn't <module> Douglas Crockford's thing? |
| 15:19 | <TabAtkins> | Yeah, it is. |
| 15:19 | gsnedders | dislikes it when people know him, it makes him feel less anonymous |
| 15:20 | <micheil> | hsivonen: no idea, I just saw it in a site which was using the html5 doctype and no other xmlns |
| 15:21 | <jgraham> | squeaky people! |
| 15:21 | gsnedders | looks oddly across the room at jgraham |
| 15:22 | <jgraham> | No you didn'r |
| 15:22 | <jgraham> | t |
| 15:22 | <gsnedders> | I did. Just when you weren't looking. |
| 15:23 | <jgraham> | Oh |
| 15:23 | <gsnedders> | You're not observant enough. |
| 15:23 | <jgraham> | Well you are not observing the squeaking |
| 15:23 | gsnedders | wonders if he could creep up on jgraham and surprising |
| 15:23 | <gsnedders> | I'm sorry, but Alice in Chains is better than squeaking |
| 15:24 | <jgraham> | The squeaking is not good. Although Alice in Chains is dubious |
| 15:24 | gsnedders | realizes what jgraham means by people squeaking |
| 15:24 | <jgraham> | as an alternative |
| 15:24 | <gsnedders> | jgraham: When did we ever agree on music? |
| 15:24 | <jgraham> | Sometimes |
| 15:25 | <gsnedders> | Rarely. |
| 15:28 | <jgraham> | Sigur Rós? |
| 15:28 | <gsnedders> | For example |
| 15:31 | <Dashiva> | I hope someone does a translation of this: http://www.apple.com/hotnews/thoughts-on-flash/ |
| 15:33 | <hsivonen> | interesting that Steve blogs on the PR wire |
| 15:35 | <Dashiva> | First he claims you don't need flash because most video is h264 anyway, and then he says flash video takes too much resources because most of it isn't h264 and is handled in software |
| 15:37 | <gsnedders> | Wow those people are squeaky. |
| 15:37 | <jgraham> | No these ones are just noisy |
| 15:38 | <gsnedders> | I was assuming it was the same people, and I was a bit surprised at the leak of squeakiness |
| 15:38 | <gsnedders> | *lack |
| 15:38 | <gsnedders> | It was just loud |
| 15:39 | <Philip`> | Dashiva: He says most video is in Flash but is also in H.264 |
| 15:39 | <Philip`> | (i.e. non-Flash H.264) |
| 15:39 | <Dashiva> | If it's available in h264, it can be hardware accelerated in flash |
| 15:39 | <Philip`> | But it's not available in H.264 in Flash |
| 15:40 | <Dashiva> | And this happens where? Why would anyone have a high quality format and not serve it to as many as possible? |
| 15:40 | <Philip`> | If web developers were given the choice between sticking with an old codec in Flash which works perfectly fine for them, or expending a lot of effort converting everything to H.264 (to make available via Flash or <video> or whatever), they'd probably choose the former because it's much easier and doesn't hurt them much |
| 15:41 | <TabAtkins> | Flash actually does hurt me. It's annoying and fiddly to use. |
| 15:41 | <Philip`> | so Apple doesn't like giving them the choice |
| 15:41 | <Dashiva> | He can't have it both ways, though |
| 15:42 | <Dashiva> | Either most content is not h264, and then apple doesn't let you use the full web. Or it is h264, in which case the performance argument is false. |
| 15:43 | <Philip`> | He doesn't say Apple does let you use the full web |
| 15:43 | <Philip`> | "iPhone, iPod and iPad users aren’t missing much video" doesn't say "... aren't missing anything" |
| 15:43 | <hsivonen> | I'd be happier about Apple's stance if their chosen video format weren't royalty-encumbered |
| 15:43 | <Dashiva> | "What they don’t say is that almost all this video is also available in a more modern format, H.264" |
| 15:43 | <Dashiva> | almost all |
| 15:44 | <Philip`> | Most of it's only also available in H.264 because Apple removed the choice and forced developers to start supplying it |
| 15:45 | <Dashiva> | Well, is it available or is it not available? |
| 15:46 | <Philip`> | It is now, but if Apple did/does support Flash then it perhaps wouldn't be |
| 15:46 | <Dashiva> | But the actual reality is one or the other |
| 15:46 | <Dashiva> | It's not a superposition of "what if apple didn't push for h264" and "reality as it is now" |
| 15:48 | <Dashiva> | There's also this gem: "When websites re-encode their videos using H.264, they can offer them without using Flash at all." |
| 15:48 | <Dashiva> | I guess Firefox isn't a browser |
| 15:49 | <workmad3> | Dashiva: obviously not, according to apple :) |
| 15:49 | <Dashiva> | Oh wait |
| 15:49 | <Dashiva> | He already did that slide where Safari ate up the market share of every browser except IE |
| 15:49 | Philip` | mostly dislikes how apparently Apple are the only people in the world who can provide platform innovations, and third-party platforms built on top of it are just restrictive and nobody would ever want to write applications that work on other platforms |
| 15:50 | <Philip`> | It's trying to force a monoculture, which doesn't sound good |
| 15:51 | <workmad3> | Philip`: what, you're disagreeing with the Apple Is God world-view? :) |
| 15:51 | <Dashiva> | Hum, w3c going for mercurial |
| 15:52 | <jgraham> | Dashiva: Why hum? |
| 15:52 | <Rik`> | Dashiva: Firefox is a browser for Apple support, they recommend it to subscribe to WWDC :) |
| 15:53 | <Philip`> | workmad3: Sure - it's trendy to disagree with Apple nowadays |
| 15:53 | <workmad3> | Philip`: that's good... I've been disagreeing with them for years :) |
| 15:53 | <jgraham> | Philip`: Would you describe yourself as "trendy"? |
| 15:53 | <workmad3> | to me, all this new grumbling is people hopping on the anti-apple bandwagon after the fact :P |
| 15:54 | <Philip`> | jgraham: Hmm... No |
| 15:54 | <Dashiva> | jgraham: Not sure what it's intended for. w3c likes to keep close eyes on their specs. |
| 15:55 | <jgraham> | Dashiva: We are getting it for testsuites at least |
| 15:55 | <jgraham> | Dunno what is happening with the specs but I assume that they can't use torture^WCVS forever |
| 15:56 | <gsnedders> | Rik`, that reminds me… June 7–11th! Noooo! |
| 16:35 | <remysharp> | Philip`: sorry to bother you - you helped me with a postMessage and workers question - |
| 16:36 | <remysharp> | but the spec examples on the whatwg site implies that *only* strings will be received by the worker: http://dev.w3.org/html5/workers/#a-worker-for-updating-a-client-side-database |
| 16:36 | <Rik`> | gsnedders: hum ? |
| 16:36 | <gsnedders> | Rik`: WWDC |
| 16:36 | <gsnedders> | Rik`: Of course, the nooo will make no sense to you :) |
| 16:36 | <Rik`> | exactly |
| 16:36 | <remysharp> | there's several examples on in the Worker spec that split the event.data to read what it wants to do with it |
| 16:37 | <gsnedders> | Rik`: It stops me from running off with Apple people when we were going to |
| 16:37 | <gsnedders> | There'ss sense to this insanity. |
| 16:39 | <jgraham> | gsnedders: No eloping please |
| 16:40 | <gsnedders> | jgraham: Can I elope instead with a Scottish girl? |
| 16:41 | <TabAtkins> | gsnedders: Sounds good to me. Find yourself a buxom redhead or raven-haired girl. |
| 16:41 | <jgraham> | Raven hair and skin so fair, sadness never visits there |
| 16:43 | gsnedders | expects he can find several girls meeting one of those alternations in his Facebook friends… |
| 16:44 | <Philip`> | remysharp: I'd guess that's because the example was written back when postMessage only supported strings, and/or is kept because it's a useful demonstration of how to work with legacy implementations that only support strings |
| 16:44 | <hober> | hmm. getting an error page on every wiki page I try to go to |
| 16:44 | jgraham | isn;t sure whether to parse it as (buxom & redhead) | raven haired or buxom & (redhead | raven haired) |
| 16:44 | <hober> | Database error: A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was: (SQL query hidden) from within function "User::saveOptions". Database returned error "1062: Duplicate entry '121-rememberpassword' for key 1 (database.wiki.whatwg.org)". |
| 16:44 | <TabAtkins> | The last, jgraham. |
| 16:45 | <jgraham> | TabAtkins: I think that makes it sizist, or something |
| 16:45 | <gsnedders> | TabAtkins: Oh, I parsed it differently. |
| 16:45 | <TabAtkins> | The fact that redhead and raven-haired modify the same attribute should be evidence that they form a localized disjuction. |
| 16:46 | <jgraham> | But maybe if you didn;t really like red hair you would suggest some way of making up for it? |
| 16:46 | <TabAtkins> | I wouldn't have listed red hair if I didn't like it. See: the lack of blonde in my list. |
| 16:46 | <jgraham> | I just assumed that was because there are not blone girls in Scotland |
| 16:46 | <jgraham> | *blonde |
| 16:47 | <gsnedders> | jgraham: There are, just fewer than here. |
| 16:47 | Philip` | imagines a raven-haired girl is like Medusa but featherier |
| 16:47 | <TabAtkins> | Something like that, yes. |
| 16:49 | <gsnedders> | As far as I know, there's never been any conclusion about weather I'm the fairest person in the office |
| 16:50 | <gsnedders> | But I'm not raven-haired, so I guess sadness visits here. Sad. |
| 16:57 | <AryehGregor> | Argh, wiki errors? |
| 16:59 | <jgraham> | gsnedders: There aren't that many here |
| 17:00 | <AryehGregor> | . . . do I not have an account on the wiki? Seems not. |
| 17:00 | <AryehGregor> | hober, does it work if you log out? |
| 17:00 | <gsnedders> | jgraham: Well, true, we're in the office of an IT company. There aren't many women here in general. |
| 17:01 | <jgraham> | gsnedders: That not what I menat, as well you know |
| 17:02 | <gsnedders> | :) |
| 17:03 | <AryehGregor> | Is anyone else getting errors on the wiki? |
| 17:05 | <AryehGregor> | hober, the error should give more details now. Can you provide them? |
| 17:08 | <AryehGregor> | Or not. Now it should be logging details. |
| 17:17 | <AryehGregor> | Hixie, you don't seem to have APC or anything similar set up as a PHP opcode cache. Any particular reason for this? It would help performance a lot for large PHP apps (e.g., MediaWiki). |
| 17:17 | <gsnedders> | AryehGregor: Because Hixie's a Perl fanboi |
| 17:19 | <AryehGregor> | Then he should clearly have used a wiki engine written in Perl. |
| 17:19 | <AryehGregor> | I have no idea why PHP doesn't support something like Python's .pyc out of the box, so that special opcode caches aren't necessary. |
| 17:20 | <AryehGregor> | Actually, I do know, it's because PHP is developed by monkeys pounding on typewriters. |
| 17:30 | <TabAtkins> | What happens if you point an <img> at something without intrinsic dimensions, such as an SVG image that has only an aspect ratio, but don't give the <img> a width or height? |
| 17:30 | <TabAtkins> | That is to say, what's the default size of an <img> when it can't get any appropriate information from the linked image? |
| 17:33 | <hober> | AryehGregor: unfortunately, I'm at a different machine now. the wiki works fine from here. |
| 17:33 | <AryehGregor> | Weird. |
| 17:33 | <AryehGregor> | Tell me if you have any problems again. |
| 17:35 | <hober> | will do |
| 18:29 | <zcorpan> | jgraham: am now |
| 19:11 | <jgraham> | zcorpan: That's less useful :) |
| 19:11 | <jgraham> | (hallvors was looking for you) |
| 19:19 | <AryehGregor> | "One of the assumptions most software organizations appear to take for granted is the need for beta releases. In fact, some companies have even taken this assumption well beyond its commonly understood meaning." ("assumption" linking to a post about Gmail) |
| 19:19 | <AryehGregor> | Zing. |
| 19:21 | <AryehGregor> | Actually, the post it links to is someone who claims some of their e-mail vanished, so they stopped using Gmail. That seems like an underhanded thing to link to. Well, whatever. |
| 19:21 | <AryehGregor> | (The quote is from the latest IEBlog) |
| 19:29 | <AryehGregor> | Yeah, I guess that post is a steaming pile of PR. |
| 19:29 | <AryehGregor> | Oh wlel. |
| 19:29 | <AryehGregor> | Oh well. |
| 19:29 | <AryehGregor> | *Some* of the IEBlog posts aren't horrible, I'll give them that. |
| 19:32 | <othermaciej> | AryehGregor: I was really surprised by the snark level of that post |
| 19:32 | <othermaciej> | it detracts from the message they presumably want to send, which is about the awesome opportunity you have to give them feedback |
| 19:32 | <AryehGregor> | Yeah, it ended up being really misleading. |
| 19:33 | <AryehGregor> | I don't necessarily blame them for using Connect and defending it, since there's probably a lot of internal pressure to not ditch MS stuff in favor of open-source projects like Bugzilla. |
| 19:34 | <AryehGregor> | But they didn't have to act as though their bug tracker was somehow comparable to Mozilla's or WebKit's, and that they were awesome for closing everything quickly. |
| 21:20 | <Hixie> | Philip`: use complete.html to get xrefs :-) |
| 21:20 | <Hixie> | AryehGregor: i have no idea what APC is |
| 21:21 | <Hixie> | AryehGregor: happy to set stuff up though |
| 21:31 | <volkmar> | Hixie: i've some questions about the enumerated attributes |
| 21:32 | <volkmar> | Hixie: it looks like most enumerated attributes are not "limited to only known values" |
| 21:32 | <volkmar> | which sounds weird because it sounds quite important |
| 21:32 | <volkmar> | and even input.type isn't... and that is a regression |
| 21:33 | <volkmar> | i'm wondering why some enumerated attributes can be not "limited to only known values" |
| 22:04 | <Hixie> | volkmar: hmm |
| 22:04 | <jgraham> | The thing about the IEBlog is that it just makes them look childish. I'm really surprised that they are dumb enough to play to the front rows like that |
| 22:08 | <Hixie> | volkmar: yeah, looks like a bug, especially for input.type; can you file a bug? |
| 22:08 | <Hixie> | Philip`: which bugs was it that they ignored for IE8? |
| 22:09 | <jgraham> | Hixie: Is there some particular reason that you can't drawImage and <input type="image">? |
| 22:10 | <jgraham> | s/and/an/ |
| 22:10 | <jgraham> | (I don't have a usecase or any particular desire for it to work but if there is a special reason it would be useful to know) |
| 22:11 | <Hixie> | "I don't have a usecase or any particular desire for it to work" is the reason |
| 22:13 | <jgraham> | Apparently it probably does work in Opera at the moment (I haven't actually tested) and "I don't have a usecase or any particular desire for it to work" isn't a strong argument for breaking it |
| 22:14 | <volkmar> | Hixie: there were a reason why some enumerated attributes could be reflected like regular strings ? |
| 22:17 | <cardona507> | hi - I am wondering what audio formats firefox, chrome, and safari support for <audio> - thanks |
| 22:22 | <sicking> | cardona507: i believe we support vorbis and wave |
| 22:22 | <cardona507> | thanks jonas |
| 22:26 | <cardona507> | I am pretty sure that webkit is mp3 and wav |
| 22:27 | <cardona507> | wav isn't the greates - is vorbis better? |
| 22:27 | <cardona507> | *greatest |
| 22:29 | <volkmar> | cardona507: afaik, wav is lossless and vorbis is lossy so they are not really similar |
| 22:29 | <cardona507> | gotcha |
| 22:54 | <Philip`> | cardona507: Vorbis is like MP3 but a bit better quality |
| 22:54 | <Philip`> | so they're the most similar ones |
| 22:54 | <cardona507> | Philip` thanks - I need a way to change my wav files to vorbis |
| 22:54 | <Philip`> | (I think Safari and IE9 support AAC too) |
| 22:54 | <cardona507> | wow - IE with html5 support |
| 22:54 | <cardona507> | I don't know what to say... :) |
| 23:12 | <cardona507> | html5 is trending as a google search term - It's #97 of the most popular search topics right now - Thanks Steve Jobs :) |
| 23:13 | <cardona507> | I think after his post today there will be an explosion of HTML5 adoption and developers |