| 00:07 | <othermaciej> | TabAtkins: neat |
| 00:07 | <othermaciej> | TabAtkins: I'll record them later today |
| 00:07 | <othermaciej> | TabAtkins: you may also want to look a 103 and 107 in case you care |
| 00:20 | <TabAtkins> | I have no problem with 103, and nothing in particular to add to the 107 proposals. |
| 01:10 | <Dashiva> | Is unclosed <del> supposed to affect the entire remaining document? |
| 01:59 | <rosh312> | has anyone had a/v syncing issues using ogg for html5 video? |
| 02:16 | <doublec> | rosh312, what browser? |
| 02:17 | <rosh312> | firefox |
| 02:17 | <doublec> | does the a/v sync look fine in another player? |
| 02:18 | <rosh312> | when i play it on vlc, it plays fine unless i try to seek |
| 02:18 | <doublec> | what happens then |
| 02:18 | <rosh312> | the video gets a little scrambled and the audio takes a second to sync back up |
| 02:18 | <doublec> | what version of firefox? |
| 02:18 | <rosh312> | i encoded it in ogg using firefogg directly from an m4v |
| 02:18 | <rosh312> | firefox 3.6 |
| 02:19 | <doublec> | are you able to make the video publically available so I can try it? |
| 02:19 | <rosh312> | no, unfortunately not, it's for a company i'm working for that's still in private beta |
| 02:20 | <doublec> | if it goes out of sync in the first 20 seconds or so, is it possible to send that to me privately? (I'm a firefox developer working on the video support) |
| 02:21 | <rosh312> | sure, how can i send it to you? |
| 02:22 | <rosh312> | i can send you the whole video, i only meant i cant link you to the page where i have it playing using the video tag |
| 02:22 | <doublec> | how big is it? |
| 02:22 | <rosh312> | 10mb |
| 02:23 | <doublec> | is there anywhere you can put it behind http authentication that I can get it from? |
| 02:23 | <doublec> | and then take it down when I've got it |
| 02:25 | <rosh312> | http://dl.dropbox.com/u/45844/mergefm.ogv |
| 02:25 | <rosh312> | does that work? |
| 02:27 | <doublec> | ok, dling now |
| 02:29 | <doublec> | rosh312, got it |
| 02:30 | <rosh312> | great, score one for dropbox |
| 02:34 | <boblet> | hey all. what’s the HTML5 way of adding <meta http-equiv="X-UA-Compatible" content="chrome=1" />? Is it just something that hasn’t yet been registered on the Wiki PragmaExtensions page as per 4.2.5.4? |
| 02:36 | <boblet> | Also I’m assuming this could be added via the server header X-UA-Compatible: chrome=1 — anyone know if that validates? |
| 02:46 | <boblet> | looks like it does, according to http://code.google.com/p/chromium/issues/detail?id=22708 |
| 02:55 | <boblet> | any thoughts on whether nested links will make it into a future version of HTML? |
| 02:55 | <rosh312> | turns out the issue was the flash object playing at least the audio underneath the html5 video |
| 02:55 | <rosh312> | the firefox bug report is here: https://bugzilla.mozilla.org/show_bug.cgi?id=487398 |
| 02:56 | <rosh312> | thanks again doublec |
| 02:56 | <doublec> | you're welcome rosh312! |
| 03:01 | <boblet> | did Eric Meyer ever submit his nested linking proposal: http://meyerweb.com/eric/html-xhtml/html5-linking.html |
| 03:01 | <boblet> | can’t find discussion of it in list archives |
| 03:53 | <MikeSmith> | does anybody happen to know where in the MathML 2.0 spec it defines what a number is? |
| 04:07 | <MikeSmith> | http://www.w3.org/TR/MathML3/chapter2.html#id.2.1.5.1 |
| 06:44 | <MikeSmith> | http://www.w3.org/TR/MathML/chapter3.html#id.3.3.4.2.5 |
| 06:44 | <MikeSmith> | just discovering "negativeveryverythinmathspace" |
| 06:44 | <MikeSmith> | et al |
| 06:45 | <MikeSmith> | beautiful |
| 07:48 | <MikeSmith> | hsivonen: when you have time, I wanted to ask about updating the v.nu MathML schema |
| 07:56 | <MikeSmith> | I see that MathML 3.0 is now at CR |
| 07:57 | <MikeSmith> | and that the spec includes an RNC schema |
| 07:57 | <MikeSmith> | so... |
| 07:58 | <MikeSmith> | maybe it'd make sense to replace the current v.nu MathML 2.0 schema with that one |
| 07:59 | <MikeSmith> | and of course note in the UI that it's now doing MathML 3.0 validation |
| 07:59 | <MikeSmith> | after looking through the current (2.0) schema, I suspect there might be quite a few bugs in it |
| 08:01 | <MikeSmith> | I fixed a few today, but based on that sample, I reckon there are probably more waiting to be found |
| 08:03 | <MikeSmith> | and it seems like that schema has not had any maintenance in a very long time |
| 08:03 | <MikeSmith> | upstream, I mean |
| 08:03 | <MikeSmith> | http://yupotan.sppd.ne.jp/relax-ng/mml2.html shows the last update was two years ago |
| 08:04 | <jgraham> | MikeSmith: Does MathML 3.0 have any UA support? |
| 08:05 | <jgraham> | It sems dangerous to validate something that doesn't actually work in browsers |
| 08:05 | <MikeSmith> | yeah, true |
| 08:05 | jgraham | wonders what is new in MML 3.0 |
| 08:05 | <MikeSmith> | I just don't relish the prospect of making further updates to this old and buggy 2.0 schema we currently have |
| 08:07 | <MikeSmith> | jgraham: http://www.w3.org/TR/MathML3/appendixf.html#changes.mathml2.0e-3.0 |
| 08:07 | <MikeSmith> | (if you've not found it already) |
| 08:49 | <annevk> | so what do the i18n guys want? |
| 08:49 | <annevk> | that we ignore all content-language for :lang() processing? |
| 08:49 | <annevk> | would simplify some code, for sure... |
| 08:56 | <othermaciej> | annevk: the more they say what they want, the less I understand |
| 09:15 | zcorpan_ | wonders which profile of h.264 was used in http://www.streamingmedia.com/Articles/ReadArticle.aspx?ArticleID=67266&utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+StreamingMediaMagazine-FeaturedArticles+%28StreamingMedia.com%3A+Featured+Articles%29 |
| 09:16 | <zcorpan_> | "MainConcept H.264" - is that the "Main" profile? |
| 09:16 | <othermaciej> | no, that's the vendor |
| 09:16 | <othermaciej> | of the codec |
| 09:16 | <zcorpan_> | ok |
| 09:17 | <othermaciej> | they didn't specify the profile and they didn't use what is probably the best H.264 encoder available (x264) |
| 09:17 | <othermaciej> | (profile or other encoder parameters) |
| 09:18 | <annevk> | wow, bunch of crap added to that URL... |
| 09:18 | <zcorpan_> | feeds-- |
| 09:20 | <othermaciej> | according to the internets, MainConcept is apparently a fairly poor H.264 encoder |
| 09:20 | <othermaciej> | though it does come bundled w/ a bunch of other tools |
| 09:21 | <MikeSmith> | if anybody on the channel has HTML+MathML content to test with, please run it through http://qa-dev.w3.org:8888/ (unstable instance of validator.nu backend) when you have some time.. I made some change that could have introduced regressions |
| 09:23 | <zcorpan_> | http://qa-dev.w3.org:8888/?doc=http%3A%2F%2Fhsivonen.iki.fi%2Ftest%2Fmoz%2Fhtml5-hacks-demo.html |
| 09:37 | <MikeSmith> | zcorpan_: thanks |
| 09:38 | <MikeSmith> | "Total execution time 79 milliseconds." |
| 09:38 | <MikeSmith> | v.nu is fast |
| 09:39 | <othermaciej> | what does it require 79ms to do? |
| 09:41 | <MikeSmith> | validate the page against relaxng schema, check datatype restrictions on attribute values, maybe run some other checks (depending on the content) |
| 09:41 | <zcorpan_> | also schematron and checks written in java |
| 09:42 | <othermaciej> | not bad for Java |
| 09:43 | <MikeSmith> | that is a relatively simple page, though |
| 09:43 | <MikeSmith> | and I think it might do some caching, but not sure |
| 09:43 | <MikeSmith> | it's definitely faster the 2+ time you check the same page |
| 09:44 | <MikeSmith> | uses Jetty, btw |
| 09:44 | <zcorpan_> | i think it was something on 300 ms first time |
| 09:45 | <othermaciej> | Safari takes about 25ms to parse and render it (cached), but to be fair we're not rendering the MathML (yet) |
| 09:48 | <MikeSmith> | <chuckle> http://twitter.com/html5spec |
| 09:48 | <MikeSmith> | "Unnoficial blog about upcoming famous HTML5 technology." |
| 11:38 | <zcorpan_> | "With regard to the controls attribute, keep in mind some of the events listed in this article will never fire when the controls are not displayed to the user." - http://www.htmlgoodies.com/primers/html/article.php/3883356 |
| 11:38 | <zcorpan_> | i hope that's not the case in any browser |
| 11:43 | <annevk> | http://twitter.com/mpt/status/11106684817 interesting axiom |
| 11:43 | <annevk> | I wonder if it applies to specification development |
| 11:45 | <Dashiva> | Well, it's only a useful axiom to the extent "best use" of developer time is useful |
| 11:46 | <zcorpan_> | so what extension to use for webm audio? wma? :) |
| 11:46 | <Philip`> | Developers exist to help users make best use of their time |
| 11:46 | <Dashiva> | Isn't it .webm? |
| 11:47 | <Philip`> | so bug trackers are really for users |
| 11:47 | <zcorpan_> | yes but it's being discussed that webm video and webm audio should have different extensions |
| 11:48 | <Dashiva> | .webm is a hacky compromise to make video work reliably, I don't think we should be forcing it into the less messy area of audio |
| 11:48 | <jgraham> | I'm not sure what disagrees with mpt's principle |
| 11:49 | <jgraham> | (or rather. why he decided to formulte it) |
| 11:50 | <Dashiva> | If you don't believe in Philip`'s axiom |
| 11:50 | <kinetik> | Dashiva: it's hacky? |
| 11:51 | <Philip`> | Shouldn't WebM audio be .oga? |
| 11:51 | <Dashiva> | Philip`: If it's in an ogg container, yes |
| 11:51 | <Philip`> | Shouldn't it be in an Ogg container? |
| 11:51 | <kinetik> | Philip`: that requires someone to implement ogg demuxing |
| 11:51 | <Philip`> | You don't lose much functionality, and you gain compatibility with the rest of the world |
| 11:51 | <Dashiva> | kinetik: An arbitrary not-fully-compatible subset of matroska, limited to only one codec, hacky |
| 11:52 | <kinetik> | Philip`: you miss out on having an index for seeking |
| 11:53 | <kinetik> | Dashiva: two codecs. and it's compatible bar the doctype. |
| 11:53 | <Dashiva> | If it's not fully compatible, it's not compatible |
| 11:53 | <kinetik> | why does it need to be compatible? |
| 11:53 | <Philip`> | kinetik: That's not much functionality |
| 11:54 | <Philip`> | (since it's not like it's impossible to seek in Ogg files) |
| 11:54 | <kinetik> | it's also not very fast. |
| 11:54 | <kinetik> | over high latency connections like... the internet. |
| 11:55 | <daedb> | Isn't Webm with only audio basically just a slightly different .mka file? |
| 11:55 | <Dashiva> | Yes |
| 11:57 | <erlehmann> | webm audio ? |
| 11:57 | <erlehmann> | so no ogg vorbis but instead mkv + vorbis ? |
| 11:58 | <gsnedders> | Yeah, basically |
| 11:58 | <erlehmann> | sounds … unnecessarily complicated. |
| 11:58 | daedb | fails to see the point in using Webm for only audio when Ogg already exists and works |
| 11:58 | <Dashiva> | If webm had been a compatible subset of matroska, it could just have been .mka |
| 11:59 | <gsnedders> | daedb: It seems silly to use Ogg for audio and WebM for video and audio. |
| 11:59 | <kinetik> | daedb: if someone implements webm and nothing else, it would be useful to be able to play audio only files with it. |
| 11:59 | <erlehmann> | i am confused. can one of the google engineers chime in ? |
| 12:00 | <kinetik> | daedb: for example, it sounds like flash will support webm, but it's not going to support ogg. |
| 12:00 | <gsnedders> | kinetik: That's not confirmed. Flash will support VP8, nothing more has been said. |
| 12:00 | <kinetik> | gsnedders: "it sounds like" |
| 12:00 | <erlehmann> | why couldn't they just use an ogg container for webm ? i thought it was more streaming-friendly. |
| 12:01 | <kinetik> | gsnedders: also, look at this URL: http://blogs.adobe.com/flashplatform/2010/05/adobe_support_for_webm.html |
| 12:01 | <kinetik> | (note that they later reposted the same article as adobe_support_for_vp8.html though) |
| 12:01 | <kinetik> | erlehmann: it's not really any better or worse. |
| 12:02 | <erlehmann> | kinetik, well, ogg software and content is already out there. so from at least one point, it actually is worse. |
| 12:03 | <kinetik> | erlehmann: i was replying to "more streaming-friendly", but matroska software and content already exists too. |
| 12:03 | <erlehmann> | take oggforward for example. how do i do that with mkv again ? ;) |
| 12:03 | <kinetik> | and it's trivial to modify that to support webm. |
| 12:03 | <erlehmann> | of course. |
| 12:03 | Philip` | already has a page using <audio> with .ogg+.mp3 files, and would just prefer not to have to add a third source for no overwhelmingly compelling reason |
| 12:04 | <erlehmann> | mkv content is out there, harr harr. on p2p :D |
| 12:04 | <erlehmann> | Philip`, doesn't every browser having WebM support ogg, vorbis and theora already ? |
| 12:04 | <daedb> | kinetik: Well if they implement Vorbis then they should implement Ogg too, since that's the de facto standard container for Vorbis and it's already widely implemented. |
| 12:04 | <kinetik> | daedb: why? |
| 12:04 | <kinetik> | daedb: that's more work for little benefit. |
| 12:05 | <erlehmann> | daedb, i challenge your assumptions and replace them with my own. |
| 12:06 | <Rik`> | and what about supporting both ? |
| 12:06 | <Philip`> | kinetik: It seems a relatively significant benefit for users who already some Ogg Vorbis music files and want to just stick them on their web site without having to find special conversion tools |
| 12:07 | <erlehmann> | what Philip`said. |
| 12:07 | <daedb> | kinetik: Why should I have to change all my Ogg Vorbis files to a new altered version of mka (that doesn't work on my existing software that only supports standard mka) just to serve them on a web site? |
| 12:07 | <erlehmann> | also, vorbis has been a successful standard for years. |
| 12:07 | <daedb> | Author laziness > implementor laziness, basically |
| 12:12 | <kinetik> | Philip`: it'd be useful to those users, however many of them there are. |
| 12:12 | <kinetik> | there might not be enough to convince an implementor of webm to also implement ogg. |
| 12:32 | <annevk> | audio/webm should prolly be tied to .webma or something like that |
| 12:32 | <annevk> | or .weba |
| 12:35 | <erlehmann> | so much stuff to sort out. i wonder if it will take half a year or longer till webm is there. |
| 12:35 | <erlehmann> | for sufficiently usable values of "there" |
| 12:37 | <virtuelv> | what does the m in webm mean? |
| 12:38 | <zcorpan_> | virtuelv: i think webm stands for 'web media' |
| 12:41 | <erlehmann> | webm is a crappy name. they should have called it rincewind or voldemort. |
| 12:41 | <erlehmann> | :D |
| 12:41 | <jgraham> | webm is a crappy name but those are wosr :p |
| 12:41 | <jgraham> | *worse |
| 12:42 | <MikeSmith> | does the name mean I can't use it for video content that's not in the Web? |
| 12:44 | <Philip`> | H.264 is clearly a much cooler name |
| 12:44 | <Philip`> | and not at all hard to remember or to punctuate correctly |
| 12:44 | <zcorpan_> | they should have just called it 'chrome' |
| 12:45 | <Philip`> | Since everything nowadays is on the web, they should have dropped the "web" and just called it "media" |
| 12:45 | <jgraham> | They should have called it iVideo |
| 12:45 | <jgraham> | To annoy apple |
| 12:46 | <jgraham> | They could have claimed the i stood for internet |
| 12:46 | <Philip`> | What does Apple claim the i stands for? |
| 12:47 | <zcorpan_> | it's just there for coolness |
| 12:47 | <jgraham> | I believe they claim it stads for "try prefixing your product name with this and we will send an army of hungry lawyers" |
| 12:47 | <daedb> | "Apple declared the 'i' in iMac to stand for "Internet"; it also represented the product's focus as a personal device ('i' for "individual")." <-- from Wikipedia's Imac article |
| 12:48 | <zcorpan_> | they should have called it media5 |
| 12:50 | <daedb> | Yeah, then we could've had .m5v and .m5a, which is obviously way better than those lousy .m4v/.m4a MPEG-4 files :D |
| 12:51 | <zcorpan_> | indeed! |
| 12:52 | <Philip`> | They should have renamed VP8 so you could watch Web Codec videos in Web Media files in your Web browser, in a Web application that uses Web Sockets in a Web Worker to save the video into Web Storage |
| 12:52 | <Philip`> | Also, it should be a Spiderman video |
| 13:08 | <Lachy> | Lots of companies besides Apple have prefixed their product names with an 'i', like iPlayer for instance. |
| 13:09 | <Lachy> | they have in some cases sent in the lawyers, but I vaguely recall one where the result was basically that Apple doesn't own the 'i' prefix, and can't stop others using it. |
| 13:10 | <daedb> | iNaming still sucks though, and Apple started that awful trend. |
| 13:11 | <roc> | iiW |
| 13:11 | <MikeSmith> | heh |
| 13:11 | <MikeSmith> | we should start using the dotless "i" |
| 13:12 | MikeSmith | tries to figure out how to type dotless i |
| 13:15 | <Philip`> | I |
| 13:21 | <zcorpan_> | so Web Codec would be shortened to WC |
| 13:26 | <Dashiva> | wc-1 wouldn't be confused with vc-1 at all |
| 13:31 | <erlehmann> | zcorpan_, world wide web codec ! video/w3c :P |
| 13:33 | <zcorpan_> | erlehmann: awesome |
| 13:34 | <erlehmann> | sadly, google has not paid attention to imaginative people like us |
| 13:35 | <roc> | we had some good name ideas |
| 13:35 | <roc> | H.266 |
| 13:36 | <roc> | Ogg Shikari |
| 13:42 | <zcorpan_> | H.264zilla |
| 13:42 | <jcranmer> | ... Ogg Shikari ... |
| 13:54 | <erlehmann> | roc, h.666 then mozilla could claim to not support it because its the number of evil |
| 15:18 | <miketaylr> | http://miketaylr.com/post/3f7cf26f.png |
| 15:18 | <miketaylr> | ? |
| 15:22 | <miketaylr> | anyone else get that error msg while http://www.whatwg.org/specs/web-apps/current-work/ is loading? |
| 15:23 | <jgraham> | miketaylr: WFM in chromium, firefox trunk, Opera 10.50 |
| 15:24 | <miketaylr> | odd |
| 15:24 | miketaylr | shrugs |
| 15:24 | <jgraham> | (in linux) |
| 15:25 | <miketaylr> | yeah i just get that in chromium/mac. |
| 15:53 | jgraham | desires a bluffers guide to document.write |
| 15:59 | <virtuelv> | jgraham: and I want a time machine where I can go back and smack whoever thought up document.write |
| 16:10 | <jgraham> | virtuelv: Yeah, well we all want that |
| 16:10 | <jgraham> | Although I guess you don't need a time machine just to smak them, only to prevent them |
| 16:10 | <virtuelv> | barring that, I just wish I could get google/doubleclick to stop using it |
| 16:11 | <virtuelv> | jgraham: the smacking was more to discourage other similar ideas that would pop up |
| 17:06 | <theMadness> | Heh, google wasted a wonderful chance for <canvas> today. |
| 17:08 | <AryehGregor> | Well, it would be unjust to prevent IE users from being able to play Pacman. |
| 17:08 | <AryehGregor> | I'm just glad it isn't Flash. |
| 17:10 | <TabAtkins> | Damn ghosts run faster than pac! |
| 17:12 | <AryehGregor> | So, it looks like the VP8 announcement was fairly close to the best that could be expected. |
| 17:21 | <theMadness> | TabAtkins, only when you're turning or eating pellets, when you're running around normally you're faster. Also, try the insert coin button and do it while controlling ms pacman too (wasd) for extra nerd points. :D |
| 17:21 | <TabAtkins> | theMadness: Nah, Pinky is faster than you in a straightaway. |
| 17:22 | <theMadness> | I don't remember the single ghosts features tbh. |
| 17:23 | <TabAtkins> | I don't remember them either; that was experimentally determined. |
| 17:23 | <theMadness> | Apparently they reverse engineered it and found out that pinky is in fact fast. |
| 17:31 | <Rik`> | http://googleblog.blogspot.com/2010/05/working-with-admob-to-move-mobile.html |
