| 00:57 | <raggy_> | is this the place to ask about offline manifest files? |
| 00:57 | <AryehGregor> | Probably. |
| 00:57 | <AryehGregor> | It's *a* place. |
| 00:58 | <raggy_> | great :) I have a NETWORK: header with the url /api/* but never see a hit to anything in that folder? |
| 00:59 | <raggy_> | I'm using jQuery on the client to hit the url with an ajax request. |
| 00:59 | <raggy_> | & testing with the iPhone Simulator |
| 00:59 | <raggy_> | can NETWORK paths have wildcards? |
| 01:05 | <Hixie> | raggy_: remove the "*" -- it's a prefix match |
| 01:05 | <Hixie> | the "*" is unnecessary |
| 01:05 | <Hixie> | (and indeed means that /api/*foo would hit the network bet /api/foo would not) |
| 01:05 | <raggy_> | Hixie: perfect, thanks. |
| 01:14 | <TabAtkins> | Hrm. Drawing up some demos, and for some reason I'm getting a "SECURITY_ERR: DOM Exception 18" when I try 'var idata = $("#back")[0].getContext('2d').getImageData(0,0,100,100);'. |
| 01:15 | <TabAtkins> | Am I doing anything obviously wrong? |
| 01:15 | <kling> | TabAtkins: tainted your canvas by drawImage()ing an image from another domain? |
| 01:16 | <TabAtkins> | I'm drawing a video from desktop. That might taint, right? |
| 01:16 | <TabAtkins> | That is, from file:// |
| 01:16 | <kling> | I suppose so, yeah. http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#security-with-canvas-elements |
| 01:17 | <TabAtkins> | k, I'll upload to my server and try it there instead. |
| 01:17 | TabAtkins | should get localhost working. |
| 01:17 | <kling> | which browser btw? ;) |
| 01:17 | <TabAtkins> | chrome 5 |
| 01:18 | <TabAtkins> | Yup, works fine on server. Damn crazy file:// origin policies. |
| 01:20 | <TabAtkins> | Wow, micro-optimizations can really help when you're trying to do realtime video processing. |
| 01:22 | <TabAtkins> | My lag went from noticeable (about 1/10th second) to unnoticeable just by using a temp var to kill a property access in a tight loop, and switching from /8 to >>>3. |
| 01:26 | <AryehGregor> | >>3, you must mean. |
| 01:26 | <AryehGregor> | What language are you using? |
| 01:26 | <AryehGregor> | JavaScript? |
| 01:28 | <TabAtkins> | Yeah. |
| 01:28 | <TabAtkins> | And no, I mean >>>, because I want to fill in 0s. |
| 01:29 | <TabAtkins> | >> propagates the sign. |
| 01:30 | <AryehGregor> | I've never heard of that convention in my life. |
| 01:30 | <AryehGregor> | It doesn't work in C, does it? |
| 01:30 | <AryehGregor> | Some JS-specific thing? |
| 01:30 | <TabAtkins> | I have no clue. I just know that's how JS works. |
| 01:44 | <estellevw> | can anyone take a quick look at the code to this very simple <canvas> example and tell me if there is anyway to reduce any of the lines: http://standardista.com/webkit/ch7/canvas.html |
| 01:45 | <estellevw> | i know it's not the best use of canvas, but i am writing an intro tutorial to drawing shapes, and don't want to get anything wrong |
| 01:54 | <AryehGregor> | Seems correct to me. Dunno much about canvas, though. |
| 01:56 | <TabAtkins> | Same here. Looks roughly okay. |
| 01:57 | <TabAtkins> | Yay! Got all 3 of my demos done. Woo! |
| 01:57 | TabAtkins | is now ready to actually give his talk tonight. |
| 01:57 | <kling> | knock 'em dead :) |
| 01:59 | <estellevw> | thanks |
| 02:01 | <kling> | estellevw: the call to context.scale() seems superfluous, otherwise fine. personally, i'd use save/restore to get rid of the shadow instead of setting a transparent shadowColor |
| 02:03 | <estellevw> | kling: i tried the save, restore but that didn't work, and read up that transparency would make the shadow not render, though it feels like a hack even though it's in the spec |
| 02:04 | <kling> | estellevw: save/restore doesn't touch shadowColor? sounds like a bug.. which browser? |
| 02:05 | <estellevw> | i didn't get it to touch shadowColor. Believe I tested in Opera 10.6 and Saf 5 |
| 02:05 | <estellevw> | but i could have had the syntax wrong if it was supposed to. |
| 02:09 | <estellevw> | kling: the scale was to test pixelation on scaling. when i scale transform canvas and svg, they both pixelate. I would have especially expected svg to not pixelate |
| 02:10 | <kling> | estellevw: hm. scaling an svg from an <img> element? |
| 02:11 | <estellevw> | oh, it's only opera that pixelates. Safari doesn't |
| 02:11 | <estellevw> | kling: just testing for rendering |
| 02:11 | <kling> | ah :) |
| 02:12 | <estellevw> | so when asked i can explain. |
| 02:13 | <estellevw> | was trying to show that pixelation occurs with <canvas> but not <svg> |
| 02:13 | <estellevw> | proven in Safari 5 ... was looking at Opera earlier. |
| 02:13 | <estellevw> | http://standardista.com/webkit/ch7/canvas2.html v. http://standardista.com/webkit/ch7/svg2.html |
| 02:14 | kling | is startled by the :hover zooming |
| 02:18 | <estellevw> | checking pixelation. My sandbox is full of hideousnesss and bad UI :D |
| 02:24 | <AryehGregor> | The behavior of CSS transform on a vector image is probably undefined. |
| 02:24 | <AryehGregor> | CSS is usually vague like that. |
| 02:28 | <AryehGregor> | It just says "transform it using the given matrix", not "transform it by applying the given matrix to the collection of pixels" or "transform it by [some clever description that implies SVGs aren't pixellated]". |
| 02:33 | <estellevw> | my understanding on the difference between SVG and Canvas, is that canvas is basically the painting of pixels. So, should pixelate on zoom. But SVG is "scalable" and "Vector" based graphics, so should scale. |
| 02:34 | <AryehGregor> | SVG does scale. If you changed the height and width, it would. |
| 02:34 | <AryehGregor> | But CSS transforms doesn't say one way or the other how it's supposed to work here, AFAICT. |
| 02:34 | <AryehGregor> | Scaling nicely seems like a better idea, but I didn't see it in the spec. |
| 02:34 | <AryehGregor> | (Maybe I should read the spec before I say stuff like that, though.) |
| 02:34 | <estellevw> | let me try with css: height and width |
| 02:34 | <estellevw> | instead |
| 02:34 | <estellevw> | rather, as well |
| 02:35 | <AryehGregor> | Yeah, that should work better, if you can get it to work. |
| 02:35 | <AryehGregor> | Should work on the <img>, at least. |
| 02:35 | <estellevw> | it renders as i expected in Safari 5 |
| 02:35 | <estellevw> | canvas pixelated, svg didnt' |
| 02:35 | <estellevw> | same transform |
| 02:36 | <AryehGregor> | Yes, that makes more sense, but I don't think the spec requires the latter behavior. |
| 02:43 | <estellevw> | <img> does work nicely, topmost of the three... sorry for the bad UX example (Opera 10.6 is the one that had the pixelation with transform) http://www.standardista.com/webkit/ch7/svg3.html |
| 02:46 | <estellevw> | but changing height on an element redraws the page. Repainting the page can be slow. I need to fine tooth comb the spec. |
| 03:19 | <AryehGregor> | http://www.w3.org/Bugs/Public/show_bug.cgi?id=10083#c14 |
| 03:20 | <AryehGregor> | If Shelley doesn't have faith in the HTMLWG Decision Policy, and we don't have faith in the HTMLWG Decision Policy, who does? |
| 03:20 | <AryehGregor> | Anyone? |
| 03:22 | <boblet> | hey all. does anyone know where HTML 4’s (soft) hyphenation info got to in HTML5? http://www.w3.org/TR/html401/struct/text.html#h-9.3.3 |
| 03:24 | <othermaciej> | it's not a deity, it would be inappropriate to have faith in it |
| 03:26 | <othermaciej> | boblet: that sounds like something that should be specified by CSS, not HTML |
| 03:26 | <boblet> | othermaciej: I’m thinking about ­ |
| 03:27 | <othermaciej> | boblet: did you look in the entity table? |
| 03:27 | <othermaciej> | http://dev.w3.org/html5/spec/Overview.html#named-character-references |
| 03:28 | <othermaciej> | it defines what unicode character ­ maps to |
| 03:29 | <boblet> | othermaciej: yep it’s there. the reason I ask is an HTML5Doctor q wondering if it had been internationalized (I’m guessing because some languages don’t insert hyphens (Japanese for example) but HTML 4 says “a hyphen character must be displayed” |
| 03:29 | <boblet> | othermaciej: found http://www.w3.org/TR/css3-text/#hyphenate btw :) |
| 03:37 | <boblet> | huh. wonder how <wbr> compares to ­ |
| 03:50 | <othermaciej> | <wbr> doesn't insert a hyphen |
| 04:03 | <boblet> | othermaciej: aah — just a word break opportunity. thanks for the correction |
| 04:03 | <boblet> | s/word/line/ |
| 04:05 | <TabAtkins_> | Yay, talk went awesome! |
| 04:08 | <boblet> | good stuff TabAtkins_ :) which talk? |
| 04:11 | <TabAtkins_> | Quick hack talk at Google Tech User Group, about the fun things you can do with <video> and <canvas>. |
| 04:11 | <boblet> | nice :) |
| 04:12 | <TabAtkins_> | http://www.xanthir.com/video/demo0.html |
| 04:12 | <TabAtkins_> | And demo1, demo2, demo3 |
| 04:12 | <TabAtkins_> | Page + source were my slides. |
| 04:13 | <TabAtkins_> | I pleasantly surprised myself with how easy it was to put together the demos. |
| 04:14 | <TabAtkins_> | A++++ Hixie for the <video>/<canvas> integration. |
| 04:15 | <llrcombs> | WTF? |
| 04:15 | <llrcombs> | <video src=video controls</video> |
| 04:16 | <TabAtkins_> | Haha, did I accidentally leave off a >? |
| 04:16 | <llrcombs> | yes |
| 04:16 | <llrcombs> | and there's just video |
| 04:16 | <llrcombs> | you need to quote that |
| 04:16 | <TabAtkins_> | No I don't. |
| 04:16 | <llrcombs> | yes you do |
| 04:16 | <othermaciej> | no he doesn't |
| 04:17 | <llrcombs> | oh wait I'm thinking XHTML |
| 04:17 | <boblet> | lol |
| 04:17 | <TabAtkins_> | Leaving off the > is a mistake (though one which the browser recovers from). Leaving off the quotes is perfectly fine. |
| 04:17 | <llrcombs> | still, do it anyway (best practices) |
| 04:17 | <llrcombs> | and where's there a <canvas> in there? |
| 04:17 | <TabAtkins_> | My best practices is to leave them off when I know I can. |
| 04:17 | <miketaylr> | same |
| 04:17 | <llrcombs> | gets you into bad habits... |
| 04:18 | <TabAtkins_> | demo0 doesn't have a canvas - it was just to show off the <video>, since some people at the talk may not have done any html5 hacking yet. |
| 04:18 | <TabAtkins_> | ?_? No it doesn't. It's very easy to tell when you need quotes - if it contains spaces, ', ", or `. |
| 04:18 | <TabAtkins_> | Or I guess >? |
| 04:18 | <boblet> | TabAtkins_ miketaylr that’s fine for you but does get c&p noobs into trouble |
| 04:19 | <TabAtkins_> | Noobs can get themselves into trouble all by themselves. |
| 04:19 | <boblet> | just something to keep in mind |
| 04:19 | <miketaylr> | <p id=reply>ORLY |
| 04:19 | <boblet> | TabAtkins_ too true. still I personally try to avoid helping them |
| 04:20 | <miketaylr> | ;) |
| 04:20 | <TabAtkins_> | I'd rather noobs not get the impression that all the cool people quote everything. |
| 04:20 | <TabAtkins_> | I'm calling it - I had the coolest talk at the meeting. |
| 04:20 | <boblet> | TabAtkins_ why not? |
| 04:20 | <TabAtkins_> | boblet: That way lies the "XHTML is faster because... because!" dogma. |
| 04:21 | <llrcombs> | OK, that's impressive |
| 04:21 | <llrcombs> | demo3 |
| 04:21 | <TabAtkins_> | Hehe. |
| 04:21 | <othermaciej> | in actual browsers, XHTML is slower |
| 04:21 | <TabAtkins_> | Yeah, for that one I did the code walkthrough befor showing it, because I knew it would be really cool. |
| 04:22 | <llrcombs> | XHTML == ...lolwat? |
| 04:22 | <boblet> | hrm, it’s a tricky line to walk — HTML5’s (well, UA) ability to handle various coding styles is nice, but learners prefer rules without exceptions |
| 04:22 | <TabAtkins_> | I'm fine with tutorials teaching a "safe" style. But a random demo might as well show off the cooler, looser parts of HTML, like the ability to omit tags and quotes and such. |
| 04:23 | miketaylr | notices implicit <html>, <head>, and <body> |
| 04:23 | <boblet> | TabAtkins_: gonna write one of them blog post thingies that links to all demos? |
| 04:24 | <boblet> | so I can tweetify it? |
| 04:24 | <TabAtkins_> | boblet: I don't have a blog. But I could. |
| 04:25 | <TabAtkins_> | Maybe I could write a guest entry for html5doctor? |
| 04:25 | <boblet> | ;) |
| 04:25 | <boblet> | would love you to do that |
| 04:25 | <TabAtkins_> | All right, we'll talk tomorrow about it. |
| 04:25 | <TabAtkins_> | What's the coding style for the posts? |
| 04:26 | <boblet> | rgr |
| 04:26 | <boblet> | whatever you want — I’ll fix up anything that needs fixing |
| 04:26 | <TabAtkins_> | All right, no problem. |
| 04:27 | <boblet> | normally we’re doing an element per post so quoting from the spec, but I’m guessing you’ll write up the combo of canvas, video and track right? |
| 04:27 | <boblet> | so more a show and tell than in-depth look |
| 04:28 | <TabAtkins_> | Not <track>. I'm not doing anything with that in the demos. |
| 04:28 | <boblet> | aah ok. ping me here or boblet on twitter. thanks in advance! |
| 04:29 | <TabAtkins_> | And yeah, my talk was just a quick surface tour of the video/canvas integration and what you can do with direct pixel access to the video frames. |
| 04:29 | <TabAtkins_> | Something to suggest future projects and provide some super-basic code to learn from. |
| 05:29 | <NetLarIrvine> | any real good html 5 books out there? |
| 05:32 | <NetLarIrvine> | I am a little confused about browser support for html 5 |
| 05:34 | <NetLarIrvine> | How much does WebKit support it |
| 06:42 | <MikeSmith> | sunshine! |
| 06:42 | <MikeSmith> | another beautiful day |
| 06:42 | <MikeSmith> | happy thoughts! |
| 06:43 | <kennyluck> | A hot day, I should say. |
| 07:01 | <estellevw> | NetLarIrvine: you can take a look at http://caniuse.com to see what can now be used |
| 07:02 | <estellevw> | NetLarIrvine: I have way too indepth browser support grids at http://www.standardista.com. It doesn't cover everything, but provides an explanation of what it does cover. THere are support grids under the CSS3 and HTML5 tabs. |
| 07:04 | <estellevw> | NetLarIrvine: there are lots of support grids out there. The Safari DiveIntoHTML5.org also has support within the chapter introductions to various elements and attributes. |
| 07:13 | <MikeSmith> | estellevw: would be good to have IE9 on your support tables |
| 07:13 | <estellevw> | I know, but I dont' have Windows Vista or 7, so haven't had a chance to test IE9 yet. |
| 07:14 | <estellevw> | unfortunately, IE9 won't run on XP |
| 07:14 | <llrcombs> | is it possible to define a start time for a <video>? |
| 08:41 | <MikeSmith> | http://www.adobe.com/enterprise/pdfs/html5_flash.pdf is interesting |
| 08:42 | <MikeSmith> | "Developers should favor the lightest-weight technology that will meet their requirements." seems a pretty good principle in general |
| 08:57 | <othermaciej> | since this is on Adobe's site, is it supposed to be pro-Flash? |
| 09:10 | <MikeSmithX> | othermaciej: dunno. it seems like it's copyrighted by Gartner and perhaps there without authorization |
| 09:15 | <kennyluck> | http://www.technologyreview.com/web/25728/?a=f |
| 09:15 | <kennyluck> | "Technology Review: Wikipedia to Add Meaning to Its Pages" |
| 09:44 | <MikeSmith> | kennyluck: great to hear that Wikipedia is finally adding some meaning to its pages |
| 10:06 | <annevk> | http://evc-cit.info/cit040/ |
| 10:07 | <annevk> | looks like some need some further updates |
| 10:08 | <kennyluck> | HTML course? cool |
| 10:16 | <boblet> | TabAtkins: btw one requirement re: code samples — they’ll need to be x-browser |
| 10:19 | <annevk> | zcorpan_, I think Paul is wrong given that even the first edition clearly refers to the HTML <link> element |
| 10:19 | <annevk> | zcorpan_, what he's saying about media="" makes no sense whatsoever |
| 12:01 | <MikeSmith> | damn |
| 12:02 | <MikeSmith> | text selection in Terminal seems to be borked for me after most recent OSX system software update |
| 12:03 | <MikeSmith> | command-doubleclick seems to make the beachball appear every time, then spin spin spin |
| 12:08 | <Rik`> | what does cmd+doubleclick do ? |
| 12:16 | <MikeSmith> | Rik`: select a word, I think |
| 12:18 | <Rik`> | doubleclick is enough for that |
| 15:54 | <llrcombs> | is it possible to define a start time for a <video>? |
| 16:00 | <Lachy> | llrcombs, use the DOM API to seek to the desired time before the video starts playing |
| 16:22 | <TabAtkins> | boblet: Nothing in my code is Chrome-specific, just the video codec (since I just ripped it straight off of Youtube). I'll need to find an h264->theora encoder to make my demos work in FF and Opera. |
| 16:23 | <Philip`> | TabAtkins: Not a WebM encoder? |
| 16:24 | <TabAtkins> | I suppose WebM would work, though it wouldn't work in public FF yet. |
| 16:24 | <TabAtkins> | Still, might as well be thorough. ^_^ |
| 16:24 | <Philip`> | It will soon, and I'm not aware of any long-term reasons to continue using Theora |
| 16:24 | <TabAtkins> | FF4 does WebM, right? |
| 16:24 | <zcorpan_> | yep |
| 16:25 | <llrcombs> | Lachy: on what event? |
| 16:25 | <llrcombs> | aww, gone |
| 16:25 | <TabAtkins> | kk, I'll do that instead. So, where's an h264->webm encoder? |
| 16:25 | <Philip`> | http://www.webmproject.org/tools/ perhaps |
| 16:28 | <annevk> | llrcombs, I think you can seek right away |
| 16:28 | <boblet> | TabAtkins: yeah, that’d be good |
| 16:28 | <annevk> | so should we remove Theora support from Opera? |
| 16:29 | <annevk> | and Firefox / Chrome, but I have less influence there :) |
| 16:31 | <llrcombs> | immediately after element creation? |
| 16:31 | <zcorpan_> | "I'm currently converting a very simple Flash-animated advertisement into HTML5 and it's looking like switching to HTML5 is going to be more than ten times the work!" http://forums.whatwg.org/viewtopic.php?t=4322 |
| 16:31 | <Philip`> | annevk: Not while <video> demos rely on it |
| 16:31 | <llrcombs> | I'm getting: |
| 16:31 | <llrcombs> | INVALID_STATE_ERR: DOM Exception 11: An attempt was made to use an object that is not, or is no longer, usable. |
| 16:31 | <annevk> | llrcombs, ah ok |
| 16:31 | <annevk> | is the spec too hard to read? |
| 16:31 | Philip` | doesn't know if there's much non-demo content that is unlikely to migrate to WebM |
| 16:31 | <llrcombs> | who, me? |
| 16:31 | annevk | looks |
| 16:32 | <zcorpan_> | Philip`: if we're going to remove it, seems better to remove it when only demos rely on it instead of 1% of the web relies on it |
| 16:32 | <annevk> | llrcombs, it seems you should wait for loadeddata to be dispatched |
| 16:32 | <annevk> | llrcombs, well, that's the earliest |
| 16:32 | <annevk> | afaict |
| 16:33 | <annevk> | you can also wait for canplay |
| 16:33 | <zcorpan_> | although i don't see any reason to remove support for theora |
| 16:33 | <annevk> | less features |
| 16:33 | <annevk> | to worry about |
| 16:33 | <zcorpan_> | yeah, less to maintain, true |
| 16:34 | <annevk> | less to tweak, optimize, port, etc. |
| 16:34 | <llrcombs> | videoEl.addEventListener("loadeddata",jumpToTime);? |
| 16:34 | <annevk> | especially with something as complex as a codec I'd think it's a win |
| 16:34 | <annevk> | llrcombs, yeah, something like that |
| 16:34 | <annevk> | llrcombs, you prolly need a third argument though |
| 16:34 | <annevk> | llrcombs, jumpToTime, false) |
| 16:34 | <zcorpan_> | annevk: talk to foolip |
| 16:35 | annevk | summons foolip |
| 16:35 | <Philip`> | zcorpan_: I assume 1% of the web wouldn't rely on it, since they'd have no reason to pick it instead of WebM, unless they've already got a load of Theora videos and don't have the originals to re-encode |
| 16:35 | <Philip`> | I guess Wikipedia might be in that category but I don't know who else |
| 16:36 | <zcorpan_> | Philip`: since when does content on the web use what makes sense instead of anything that is available? |
| 16:36 | <Philip`> | My assumption may indeed be foolish |
| 16:38 | <miketaylr> | TabAtkins: http://www.mirovideoconverter.com/ |
| 16:38 | <miketaylr> | for h264 to webM |
| 16:38 | <TabAtkins> | Ah, danke miketaylr |
| 16:38 | <miketaylr> | de nada |
| 16:38 | <TabAtkins> | Crap, windows or mac only? |
| 16:39 | <miketaylr> | :/ |
| 16:39 | TabAtkins | switches to his windows machine for a minute. |
| 16:39 | <Evet> | where to discuss about semantic web? |
| 16:40 | <llrcombs> | oh yeah |
| 16:40 | <annevk> | Evet, I think that's #swig |
| 16:40 | <llrcombs> | thx |
| 16:40 | <annevk> | Evet, but I'm not sure |
| 16:40 | <llrcombs> | actually, I don't think the last one's required |
| 16:40 | <Evet> | annevk: i think thats ok. thanks |
| 16:40 | <annevk> | llrcombs, might depend on the browser, but per the current specification it is |
| 16:41 | <llrcombs> | well, I should add that where it's missing |
| 16:41 | <annevk> | (it should become optional though, the argument is hardly ever useful) |
| 16:41 | <llrcombs> | true |
| 16:41 | TabAtkins | is once again temporarily stymied by the fact that computers don't come with a decent ftp client pre-installed. |
| 16:45 | <annevk> | TabAtkins, not running Ubuntu? |
| 16:45 | <TabAtkins> | annevk: On my windows box. |
| 16:45 | <TabAtkins> | Taken care of now, of course - I just pulled down filezilla real quick. |
| 16:45 | <annevk> | if you don't get a decent OS don't expect decent things |
| 16:46 | <TabAtkins> | Blame miketaylr for recommending a program that only runs on win and mac. |
| 16:46 | miketaylr | hides |
| 16:47 | <KaOSoFt> | ._. |
| 16:47 | <TabAtkins> | Anyway, conversion is going now. Slick little program. |
| 17:21 | <TabAtkins> | Hrm. I don't understand why Opera decided it won't play the video at http://www.xanthir.com/video/demo0.html, but it will play the exact same file on my desktop. |
| 17:23 | <annevk> | media type? |
| 17:23 | <TabAtkins> | video/webm? (Guessed on that one.) |
| 17:23 | <annevk> | I think so |
| 17:23 | <annevk> | but http://www.xanthir.com/video/webm does not have that |
| 17:24 | <annevk> | it's text/plain :) |
| 17:24 | <annevk> | HTTP is hard... |
| 17:24 | <TabAtkins> | Gah, I'm telling it what the media type is in <source>. It shouldn't pay any attention to the Content-Type. >_< |
| 17:24 | <annevk> | and really, Content-Type is not needed at all with file signatures |
| 17:24 | <annevk> | TabAtkins, of course it should |
| 17:24 | <annevk> | TabAtkins, rtfs |
| 17:25 | <TabAtkins> | I refuse! |
| 17:26 | TabAtkins | just flails at his keyboard and is happy when something usable comes out. |
| 17:26 | <annevk> | that sounds like trying to let monkeys write shakespeare |
| 17:27 | <Philip`> | When I flail at my keyboard, all that comes out is crumbs and hairs |
| 17:28 | <TabAtkins> | Anyway, I'll go set the content-types appropriately. |
| 17:36 | <TabAtkins> | All right, done. Works in Opera now. Anyone got a firefox beta they want to help me test with? |
| 17:57 | <paul_irish> | TabAtkins: fails in minefield. i added the ogv to mine and everything was peachy |
| 17:58 | <paul_irish> | btw i had the webm and theora in place last night on this. ;) |
| 17:59 | <TabAtkins> | Bah. |
| 17:59 | <paul_irish> | hey i remember a conversation about what you could get away with while using unquoted attributes.. |
| 17:59 | <paul_irish> | was there a resolution on that? |
| 18:00 | <paul_irish> | i'm mostly interested in what current useragents will allow.. (ie6 and everyone else) rather than what the spec sez |
| 18:00 | <TabAtkins> | Hm? You can "get away with" anything that doesn't contain a space, any of the quotes, or >. |
| 18:00 | <TabAtkins> | Dunno about <. That might kill things too. |
| 18:00 | <paul_irish> | Good. that's what I was hoping for. |
| 18:01 | <TabAtkins> | I'm yanking your theora video for myself. |
| 18:01 | <paul_irish> | TabAtkins: the demo is a hit, for sure. |
| 18:03 | <TabAtkins> | Heh, I see all the retweets of your post. |
| 18:07 | TabAtkins | needs to write a chrome extension for it now so it can run on everything on youtube. |
| 18:07 | <aboodman> | :) |
| 18:10 | <TabAtkins> | paul_irish: Check it in minefield now? |
| 18:13 | <Peter-> | Samurai Pizza Cats, awesome |
| 18:13 | <Peter-> | I used to watch that show |
| 18:13 | <TabAtkins> | Yeah, me too. ^_^ |
| 18:22 | <variable> | TabAtkins, there was a recent discussion about "<" |
| 18:22 | <variable> | current UAs allow it and the spec will also allow it |
| 18:23 | <TabAtkins> | variable: kk, cool. I just know that "<video src=video</video>" was working 'correctly', so I wasn't sure if the < was immediatley breaking out of the attribute or only doing so because it looked like part of an end-tag. |
| 18:24 | <variable> | I think it is because it looks like an end tag |
| 18:24 | <variable> | does <video src=video<b>foo</b> work "correctly" ? |
| 18:25 | <TabAtkins> | To test it for real I'd need a firefox nightly - at the moment all my browsers are pre-html5-parser. |
| 18:25 | TabAtkins | supposes he could turn on the html5 parser in his ff. |
| 18:26 | <Peter-> | Wouldn't the "foo" be seen as fallback-content until another block-level element occurs? |
| 18:26 | <variable> | TabAtkins, about:config html5.enabled --> on |
| 18:26 | <TabAtkins> | variable: Yeah, just turned it on. |
| 18:26 | <variable> | although on my windows box I use FF nightly anyay |
| 18:27 | <TabAtkins> | variable: No, it does not work correctly. The "<b" are sucked into the attribute value, and then the ">" closes the tag, leaving the "foo</b>" as the contents. |
| 18:47 | <TabAtkins> | http://www.xanthir.com/:suh |
| 18:55 | <Philip`> | paul_irish: IE handles ` like a quote character too, so you can't start unquoted attribute value with that |
| 18:56 | <Hixie> | i forget, do y'all prefer it when an IDL attribute throws an exception when you set an invalid value, or do you prefer it to ignore hte invalid value? |
| 18:57 | <TabAtkins> | Personally prefer exceptions. |
| 18:57 | <zcorpan_> | Hixie: it depends |
| 18:58 | <zcorpan_> | Hixie: if it can be invalid sometimes because of race conditions or the author forgets a special case, then throwing is a bad idea |
| 18:59 | <zcorpan_> | Hixie: since it breaks the page sometimes or for some users |
| 18:59 | <Hixie> | this is for timedTrack.mode |
| 18:59 | <Hixie> | it is either 0, 1, or 2 |
| 18:59 | <Hixie> | disabled, hidden, showing |
| 19:00 | <TabAtkins> | Then yeah, throw. |
| 19:00 | <zcorpan_> | that seems ok to throw |
| 19:00 | <Hixie> | k |
| 19:00 | TabAtkins | just likes things that are easy to debug, and exceptions are nice for that. |
| 19:01 | Hixie | tries to work out which exception to use |
| 19:01 | <Hixie> | INVALID_ACCESS_ERR maybe |
| 19:02 | <TabAtkins> | Access? |
| 19:03 | <Hixie> | it's definition seems the closest to what is appropriate her |
| 19:03 | <Hixie> | e |
| 19:04 | <TabAtkins> | Hmm, guess so. The name is unintuitive, but a quick search turns up a useful definition. |
| 19:05 | <fwaokda> | I'm trying to get my ListView to display all strings in the array, but it's only displaying the first. Anyone can take a look here: http://pastebin.com/xeCS7494 -- and see if they can spot my error? |
| 19:05 | <zcorpan_> | or NOT_SUPPORTED_ERR if you intend to add more modes in the future? |
| 19:05 | <Hixie> | dunno what modes we'd add |
| 19:05 | <TabAtkins> | fwaokda: This probably isn't the best room for java/android dev. |
| 19:06 | <fwaokda> | TabAtkins, ya just saw was in wrong channel ;) sry |
| 19:24 | <paul_irish> | TabAtkins: http://paulirish.com/work/video2ascii/webkit.html |
| 19:24 | <paul_irish> | okay i guess i'll work now.. |
| 19:24 | <Peter-> | ..lmao |
| 19:24 | <TabAtkins> | Hahaha |
| 20:19 | <Philip`> | I love it when web sites redirect me to a http://0.0.0.0?... URL when I click "submit" to confirm a payment |
| 20:19 | Philip` | eventually concluded it probably didn't perform the payment, then switched to a different browser which the site worked better in |
| 20:32 | <zcorpan_> | http://www.delphiki.com/websrt/ |
| 20:33 | <Hixie> | pretty good except for the bit at the bottom |
| 20:34 | <zcorpan_> | yeah seems he misunderstood karaoke timestamps |
| 20:36 | <zcorpan_> | Hixie: btw why can't </rt> be omitted? |
| 20:38 | <Hixie> | in the syntax or in the parser? |
| 20:39 | <zcorpan_> | both |
| 20:39 | <Hixie> | didn't really think about it |
| 20:40 | <Hixie> | just seemed simpler this way |
| 20:40 | <Hixie> | i guess if people want we can make it optional |
| 20:41 | <zcorpan_> | i saw subtitles with seemingly deliberatly omitting </i> by just looking at a handful |
| 20:41 | <zcorpan_> | doesn't seem like a stretch that people are going to omit </rt> especially since it's optional in html |
| 20:41 | <Hixie> | yeah seems reasonable |
| 20:41 | <Hixie> | file a bug and i'll look into it |
| 20:48 | <llrcombs> | annevk: you here? |
| 20:54 | karlushi | might be interested by Jonas version of the specification |
| 21:01 | <dandaman> | so I'm using html5 to create a movile website, i was wondering what i should keep in mind when making it |
| 21:01 | <llrcombs> | a movie site, as in a streaming site? |
| 21:01 | <llrcombs> | s/streaming/youtube-ish/ |
| 21:02 | <llrcombs> | video-sharing? |
| 21:03 | <zcorpan_> | has italics stopped working in vlc 1.1.0? |
| 21:03 | <dandaman> | no |
| 21:03 | <dandaman> | just a signup site |
| 21:03 | <dandaman> | i meant mobile |
| 21:03 | <dandaman> | not movile :\ |
| 21:04 | <llrcombs> | oh |
| 21:04 | <zcorpan_> | dandaman: keep in mind that the screen is no bigger than what can fit in your pocket |
| 21:06 | <dandaman> | that is pretty much it then? |
| 21:07 | <zcorpan_> | http://my.opera.com/ODIN/blog/2010/04/12/making-your-site-mobile-friendly-rit-2010-moscow |
| 21:08 | <dandaman> | awesome, just the link i was looking for |
| 21:19 | <annevk> | llrcombs, am now |
| 21:20 | <annevk> | Hixie, actually, ignoring seems to be the convention |
| 21:21 | <annevk> | Hixie, that way you don't have to do try/catch nonsense when new values come around and otherwise it is just the same |
| 21:21 | <annevk> | Hixie, see e.g. most attributes on the 2D context API |
| 21:32 | <annevk> | sweet |
| 21:33 | <annevk> | a guy supposedly from Microsoft with a yahoo.com email address starts commenting on hybi |
| 21:33 | <annevk> | on the requirements document... |
| 21:33 | annevk | reads |
| 21:35 | <annevk> | make that skimming |
| 21:35 | <annevk> | seems mostly nitpicking |
| 21:36 | <annevk> | lol |
| 21:36 | <annevk> | someone on public-html-comments is trying to explain to me how the ID and NAME attributes work and how they're different |
| 21:38 | <annevk> | I do agree with him that the window named item stuff is horribly ugly |
| 21:38 | <zcorpan_> | should we limit it to quirks mode? |
| 21:38 | <annevk> | an odd question coming from you |
| 21:39 | <annevk> | changed your mind on that? |
| 21:39 | <TabAtkins> | Does that stuff happen before or after js globals get defined? |
| 21:40 | <zcorpan_> | no, but if all webdevs hate it it might be a good thing |
| 21:40 | <annevk> | fair enough |
| 21:41 | <zcorpan_> | TabAtkins: what do you mean? |
| 21:42 | <TabAtkins> | If a @name gets turned into a global on window, and a js variable of the same name gets defined as a global, which wins? |
| 21:44 | <zcorpan_> | the variable |
| 21:44 | <TabAtkins> | That's what I thought. |
| 21:46 | <llrcombs> | when I try to set a <video>'s .currentTime in the canPlay event, I get INDEX_SIZE_ERR: DOM Exception 1: Index or size was negative, or greater than the allowed value. |
| 21:47 | <llrcombs> | should I be waiting for a later event? |
| 21:47 | <zcorpan_> | does you server support range requests? |
| 21:48 | <llrcombs> | here, it's YouTube |
| 21:48 | <llrcombs> | so yes |
| 21:48 | <llrcombs> | (however, FLV videos don't work with range requests because their indexes aren't at the beginning of the file) |
| 21:48 | <zcorpan_> | then you should be able to seek to anywhere between starttime and duration |
| 21:48 | <llrcombs> | is there a canPlayTo()? |
| 21:49 | <zcorpan_> | there's .seekable |
| 21:51 | <llrcombs> | document.getElementsByTagName("video")[0].seekable.end() is giving the latest loaded portion |
| 21:51 | <llrcombs> | given an FLV |
| 21:52 | zcorpan_ | is concerned about use of .end() without a parameter |
| 21:53 | <llrcombs> | what should the parameter be? |
| 21:53 | <llrcombs> | I didn't know seekable existed, so I experimented |
| 21:54 | <zcorpan_> | index of the range you're interested in |
| 21:54 | <llrcombs> | and it returns... |
| 21:55 | <zcorpan_> | Returns the time for the end of the range with the given index. |
| 21:56 | <llrcombs> | alrighty, so I should leave the param as 0 |
| 21:56 | <llrcombs> | what I want to do is jump to a certain point in the video ASAP |
| 21:56 | <llrcombs> | (a point defined in window.location.hash) |
| 21:57 | <zcorpan_> | if range requests don't work with flv, then maybe you can use a different format? range request search is what you want |
| 21:58 | <llrcombs> | well, I don't have control over the format |
| 21:58 | <zcorpan_> | which browser supports playing flv in video? |
| 21:58 | <llrcombs> | I can force YouTube to load an MP4, but that often degrades quality |
| 21:58 | <llrcombs> | Safari with Perian |
| 21:58 | <zcorpan_> | ok |
| 21:59 | <llrcombs> | if an MP4 is loaded (because that's the highest-quality version available, or Perian is not installed), this should work fine |
| 21:59 | <zcorpan_> | i guess you can poll seekable |
| 21:59 | <zcorpan_> | or just try/catch setting currentTime |
| 21:59 | <llrcombs> | and retry on an interval? |
| 21:59 | <zcorpan_> | yeah |
| 22:00 | <llrcombs> | or set the onProgress (or whatever the event name is) handler |
| 22:00 | <zcorpan_> | although the spec might be changed (or has already been changed? don't recall) so that currentTime doesn't throw but seeks to the closest time that is seekable |
| 22:00 | <llrcombs> | well, currently, in WebKit (my target audience), it throws |
| 22:01 | <zcorpan_> | "If the (possibly now changed) new playback position is not in one of the ranges given in the seekable attribute, then let it be the position in one of the ranges given in the seekable attribute that is the nearest to the new playback position." |
| 22:01 | <zcorpan_> | expect webkit to change to match the spec |
| 22:02 | <zcorpan_> | so polling seekable seems better than causing a seek |
| 22:04 | <llrcombs> | (I'm trying to emulate YouTube's #t=<n>m<n>s for HTML5) |
| 22:39 | <dandaman> | can anyone direct me how to a site or an explanation on how to make a site fit to screen...still developing this site for mobile phones |
| 22:39 | <dandaman> | would be nice for it to scale |
| 22:40 | <zcorpan_> | do you want a different default zoom level? |
| 22:41 | <dandaman> | i want the zoom level to be default |
| 22:41 | <dandaman> | make it depend on the resolution of the viewer's screen |
| 22:41 | <dandaman> | not sure if that is possible :\ |
| 22:42 | <zcorpan_> | maybe you want the viewport meta tag |
| 22:43 | <dandaman> | thanks |
| 22:43 | <dandaman> | ill look it up |
| 22:44 | <TabAtkins> | Check out ppk's exploration of that on quirksmode. |
| 22:46 | <dandaman> | hey zcorpan_: you're awesome |
| 22:48 | <zcorpan_> | dandaman: i'll send a bill :) |
| 22:51 | <dandaman> | hold off on it |
| 22:51 | <dandaman> | it didnt work |
| 22:52 | <zcorpan_> | crap |
| 22:53 | <dandaman> | <meta name="viewport" content="width=device-width, height=device-height, user-scalable=yes" /> |
| 22:53 | <dandaman> | i put that in the head |
| 22:53 | <dandaman> | and it didnt fit to the screen on the emulator for android :\ |
| 22:54 | <zcorpan_> | does the emulator for android support meta viewport? |
| 22:55 | <dandaman> | http://learnthemobileweb.com/2009/07/mobile-meta-tags/ |
| 22:55 | <dandaman> | according to that it does |
| 22:55 | <zcorpan_> | the content can still overflow the viewport if you set a fixed width or similar |
| 22:57 | <dandaman> | yeah but it stayed at default |
| 22:57 | <dandaman> | it didnt get any bit larger |
| 23:00 | <zcorpan_> | dunno |
| 23:00 | <zcorpan_> | time to sleep |
| 23:25 | <karlcow> | http://www.infoq.com/news/2010/07/html5-novnc |
| 23:25 | <karlcow> | noVNC is a VNC client, implemented using HTML5 WebSockets, Canvas and JavaScript. |
| 23:25 | <karlcow> | http://kanaka.github.com/noVNC/screenshots.html |
| 23:26 | <daedb> | neat |
| 23:43 | <variable> | I have a simple idea for full screen on HTML5 video. Make it a UA issue. have some function makeFullScreen() that returns true or false if it worked. All current browsers would just implement it as return false; If any browser figures out some way to safely make the function actually work. The spec does not have to concern itself with the issue but still allows for some standard way if some browser *does* figure it out. |
| 23:44 | <Philip`> | That sounds like trying to solve the problem by not bothering to solve it |
| 23:44 | <Philip`> | which isn't really a solution :-p |
| 23:44 | <variable> | Philip`, it puts the problem in the hands of those that should be solving it. |
| 23:45 | <Philip`> | If a browser figures out a way to implement it, they can discuss it with other browser people in a standards group and the result of the discussion can be put in the spec |
| 23:50 | <dandaman> | I want to make my web page fit to any mobile device's screen |
| 23:50 | <dandaman> | would the best way to do this be through css? |