| 00:00 | <Philip`> | Incidentally, I believe that regexp won't work because " " matches "\s*[^"'>]" |
| 00:00 | <zcorpan> | oh forgot \s in there |
| 00:01 | <Philip`> | I think \s*+[^"'>] ought to work |
| 00:01 | <Philip`> | using Java's possessive quantifiers |
| 00:01 | <zcorpan> | i don't know what that means, but if it's equivalent then i guess it's good |
| 00:02 | <Philip`> | It means it'll match as much of \s as possible, and not back off |
| 00:02 | <zcorpan> | ah |
| 00:02 | <Philip`> | (so there will never be any whitespace left for the [...] to match) |
| 00:04 | Philip` | starts grepping |
| 00:11 | <Philip`> | zcorpan: http://philip.html5.org/data/doctype-with-bogus-after-pub-id.txt |
| 00:14 | <zcorpan> | Philip`: thanks |
| 00:15 | <TabAtkins> | Yay! Just mounted by first NFS share. |
| 00:15 | <TabAtkins> | s/by/my/ |
| 01:55 | <MikeSmithX> | GPHemsley: that document doesn't say what the default text alignment is for any element |
| 01:56 | <GPHemsley> | MikeSmithX: Shouldn't it? I mean, it's part of the default styling... |
| 01:56 | <MikeSmithX> | well, that's not normative even in the HTML5 spec itself |
| 01:56 | <GPHemsley> | And the rest is? |
| 01:57 | <MikeSmithX> | all my doc does is this regard is attempt to provide some "typical" default display properties |
| 01:57 | <GPHemsley> | right |
| 01:57 | <MikeSmithX> | http://dev.w3.org/html5/markup/th.html#th-display |
| 01:57 | <MikeSmithX> | and that in fact seems to indicate that in Webkit, it's not centered by default |
| 01:57 | <MikeSmithX> | as far as I can see |
| 01:57 | <GPHemsley> | oh, lovely |
| 01:59 | <GPHemsley> | while we're on that page, why does th have two different style declarations? |
| 01:59 | <MikeSmithX> | hmm |
| 02:00 | <MikeSmithX> | yet, in Safari, it still gets centered |
| 02:00 | <MikeSmithX> | GPHemsley: about that, I dunno |
| 02:00 | <GPHemsley> | that default, BTW, goes all the way back to 3.2, AFAICT |
| 02:00 | <MikeSmithX> | all that I do is attempt to have it copy in the relevant parts of the default Webkit UA stylesheet |
| 02:00 | <GPHemsley> | HTML 3.2, that is |
| 02:01 | <GPHemsley> | why WebKit? |
| 02:01 | <MikeSmithX> | because its default UA stylesheet is a single file |
| 02:02 | <MikeSmithX> | anyway, in this case, it seems like the mechanism I'm using for parsing the Webkit stylesheet has a bug |
| 02:02 | <MikeSmithX> | because it seems to be not picking up something that it should |
| 02:03 | <MikeSmithX> | hmm, weird |
| 02:04 | <MikeSmithX> | when I look at the computed style in Web Inspector, it doesn't seem to show anything about text alignment being center |
| 02:06 | <MikeSmithX> | OK, it shows that text-align: center is set on it, but inherited from... somewhere |
| 02:45 | <MikeSmith> | is the empty string a valid RFC 3987 "IRI reference"? |
| 02:51 | <MikeSmith> | hmm, looking the ABNF at http://tools.ietf.org/html/rfc3987#section-2.2 it doesn't even seem to allow "#foo" as a value |
| 02:51 | <MikeSmith> | so I must be misreading it somehow |
| 03:12 | jwalden | learns that <video> content-sniffs :-( |
| 04:47 | <wycats> | Can you put yourself in a cache manifest? |
| 04:47 | <wycats> | so if an HTML page modifies its DOM, can it serialize itself and update the manifest? |
| 05:10 | <GPHemsley> | Is there a blanket term for all form elements (e.g. input, select, etc.) that is distinct from (and cannot be confused with) "<form> elements"? (I think there is, but I can't think of it....) |
| 05:28 | <boblet> | aaw, drclue isn’t here? |
| 05:29 | <wycats> | GPHemsley: form controls? |
| 05:30 | <GPHemsley> | ah, maybe |
| 05:31 | <GPHemsley> | thanks |
| 05:36 | <boblet_> | don’t understand why Mac OS X can’t transition seamlessly from ethernet to wifi when ethernet is unplugged |
| 05:36 | <boblet_> | how hard can it be? :| |
| 07:02 | <wycats> | sorry for reiterating, but is it possible to replace a document ITSELF in the cache manifest? |
| 07:03 | <wycats> | say I have a foo.html with a cache manifest |
| 07:03 | <wycats> | and I want to replace foo.html after manipulating the DOM |
| 07:03 | <wycats> | is that possible? |
| 08:22 | <foolip> | jwalden: <video> does not content-sniff per spec |
| 08:22 | <jwalden> | foolip: not true if the MIME type of the video is application/octet-stream |
| 08:42 | <Dashiva> | Are there actually any pages out there relying on quirks/not-quirks that use the obscure SGML-isms? |
| 08:42 | <Dashiva> | Not counting new pages being written to make point |
| 08:43 | <Philip`> | Which obscure SGML-isms? |
| 08:43 | <Philip`> | Some use XML-style internal subsets, but those come after the system identifier so they don't force quirks mode in HTML5 |
| 08:45 | <Dashiva> | The obscure SGML-isms that Leif is fighting for on public-html |
| 08:45 | <Philip`> | I would expect there's approximately zero |
| 08:46 | <Philip`> | (since I've seen no occurrences in half a million pages) |
| 08:47 | <Dashiva> | And if that is the case, I don't see how what browsers did before is a valid point. Seeing as a) nobody is affected retroactively and b) nobody should be using this in new pages |
| 08:50 | <zcorpan> | Dashiva: indeed |
| 08:51 | <Hixie> | http://www.amazon.com/WD-Caviar-Green-WD20EADS-internal/dp/B001UE8LRE/ref=sr_1_2?ie=UTF8&s=electronics&qid=1266656030&sr=1-2 |
| 08:51 | <Hixie> | List Price: $5,298.74 |
| 08:51 | <Hixie> | Price: $169.99 |
| 08:51 | <Hixie> | You Save: $5,128.75 (97%) |
| 08:51 | <Hixie> | o_O |
| 08:51 | <Hixie> | methinks that is inaccurate |
| 08:52 | <Dashiva> | I know some companies raise prices before a sale so the slash rate can seem more impressive, but that is... excessive |
| 08:53 | <Philip`> | Dashiva: It's relevant only to the extent that some pages do bogus things that happen to look a little bit like obscure SGML-isms and might care about compatibility |
| 08:54 | <Philip`> | and they might care about compatibility with e.g. Firefox, whose special-casing of "[" was (I guess) due to an attempt to handle internal subsets, or something like that, unless I'm completely misinterpreting its behaviour |
| 08:55 | <Dashiva> | Well, how many pages like that exist? |
| 08:56 | <Dashiva> | The media type registration RFC seems completely oblivious of versioned formats... |
| 08:57 | <zcorpan> | Philip`: i think i've read mozilla's rationale for its behavior of [] being something like "if there's an internal subset, we know it's not an old-era doc and the author probably knows what he's doing so let's use standards mode" |
| 08:59 | <Philip`> | Dashiva: About 0.001% |
| 09:00 | <Philip`> | zcorpan: "not an old-era doc" doesn't sound obviously true to me, since people in old eras are probably even more likely to mistakenly think that HTML is SGML |
| 09:01 | <zcorpan> | Philip`: i might be misremembering that part |
| 09:08 | <zcorpan> | https://bugzilla.mozilla.org/show_bug.cgi?id=1312#c62 - " 2. A DOCTYPE declaration without a DTD, i.e., <!DOCTYPE HTML>. ... Personally I do not see the point of using Standard mode with those as opposed to quirk mode. ... Number 2 is more likely to mean HTML2 than any other version." - Hixie |
| 09:08 | <zcorpan> | thank god Hixie didn't get his way there |
| 09:15 | <Philip`> | "HTML 5 should be backwards compatible" - at least that hasn't changed |
| 09:17 | <hsivonen> | zcorpan: whoa. indeed |
| 09:20 | <hsivonen> | someone should put together embarrassing ancient bugzilla quotes from #whatwg regulars |
| 09:20 | <hsivonen> | It would show a capability to learn |
| 09:23 | <Dashiva> | That would be biased against people working on public bug trackers :P |
| 09:28 | <Philip`> | hsivonen: It would just show how everyone got corrupted when they started working for browser vendors |
| 09:32 | <hsivonen> | Philip`: you may have cause and effect reversed there :-) |
| 09:39 | <hsivonen> | (assuming that you count taking the Web red pill as getting corrupted) |
| 09:47 | <zcorpan> | Philip`: how did http://www.tudakozo.hu/ROYAL-CAPS%20Kft..html end up in the list? |
| 09:47 | <zcorpan> | Philip`: does \s*+ cover newlines? |
| 09:48 | zcorpan | first got corrupted then started working for a browser vendor |
| 09:57 | <zcorpan> | jania.pe.kr/wiki/jwiki/moin.cgi/CmmKeyProcessArea?action=DeleteUploadedFile - i wonder which file Philip` deleted |
| 10:10 | <zcorpan> | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN>" |
| 10:27 | <zcorpan> | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">; - U+3000 |
| 10:55 | <zcorpan> | http://www.depmod.com/albums/some_great_reward/a0470.htm has lots of U+FFFDs rendered with html5 parser |
| 11:09 | <foolip> | jwalden: yes true, it doesn't sniff, it is hard-coded to "maybe", i.e. it will alwasys be tried (regardless of content) |
| 11:11 | <jwalden> | thing is, if you're given something a/o-s, there's no reason you should assume it's something more specific; if you want it to be handled a certain way, give it that content-type |
| 11:11 | <jwalden> | application/octet-stream means arbitrary content |
| 11:21 | <foolip> | so you think we should hard-code it to unsupported? |
| 11:22 | <foolip> | I think maybe is a more reasonable answer, is it harmful in some way? |
| 11:22 | foolip | is afk for a while |
| 11:24 | <zcorpan> | jwalden: the spec requires 'maybe' for application/octet-stream |
| 11:30 | <Philip`> | zcorpan: The tudakozo page said |
| 11:30 | <Philip`> | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">; |
| 11:30 | <Philip`> | ...except with a newline before the "http" |
| 11:30 | Philip` | wonders why his IRC client sometimes eats newlines when pasting |
| 11:30 | <Philip`> | zcorpan: \s does include newlines |
| 11:31 | <boblet> | Hixie: you there? |
| 11:31 | <Philip`> | zcorpan: Blame dotbot for deleting files, not me - I never made any HTTP requests to any of these pages :-) |
| 11:33 | <zcorpan> | Philip`: ok |
| 11:36 | Philip` | really should make a program that can print the page content it's got for a given URL |
| 11:37 | <Philip`> | rather than grepping for '.' and then grepping the output to find lines from that URL |
| 11:41 | Philip` | goes away |
| 11:46 | <boblet> | is the spec annotation system login shared with anything else, or a separate thing? |
| 11:48 | <boblet> | never mind—thought it was connected with submit review comments form, and it isn’t |
| 12:08 | <hsivonen> | zcorpan: there must be a subtle bug in how the html5 parser behaves when the decoder returns an error |
| 12:08 | <hsivonen> | it's a known bug |
| 12:09 | <hsivonen> | the decoder interface isn't documented in detail |
| 12:09 | <hsivonen> | I must have read the old code badly |
| 12:14 | <Lachy> | does anyone here understand what Julain, Tantek, and Manu hope to achieve by generalising the profile attribute for use on all elements, and which isn't already addressed by microdata? |
| 12:14 | <Lachy> | I couldn't see anywhere in the the thread, or the draft, that illustrated what use cases and problems its meant to solve |
| 15:04 | <zcorpan> | hsivonen: what about the decoder? |
| 15:13 | <hsivonen> | zcorpan: the code that drives the decoder generates the U+FFFDs |
| 15:45 | <Philip`> | boblet: If you're logged in to the annotation system, and your login email address is a registered Bugzilla account, then you get CCed on bugs from the submit review comments form, I believe |
| 15:46 | <boblet> | Philip`: oh, so it does work how I hoped |
| 15:46 | <boblet> | will make an account and try. thanks |
| 17:25 | <JonathanNeal> | Goodmorning everybody :D |
| 17:27 | <Philip`> | Even people for whom it is not morning? |
| 17:27 | <JonathanNeal> | Good your-timezone-here! |
| 17:44 | <boblet> | coming up on 3am, nn from here |
| 19:18 | <JonathanNeal> | gsnedders, you around? |
| 19:18 | <JonathanNeal> | Hey paul_irish :D |
| 19:18 | <paul_irish> | hio |
| 19:19 | <hsivonen> | Opera has the wrong MIME type for http://snapshot.opera.com/resources/BigBuckBunny.ogv |
| 19:23 | <hsivonen> | foolip: Opera (Mac) shows a faulty duration for the above file |
| 19:54 | <JonathanNeal> | Anyone have a link pointing towards a pro HTML5 article that lists some of the advantages of moving to HTML5. |
| 19:54 | <JonathanNeal> | ? |
| 20:56 | <JonathanNeal> | Whaddaya guys think of my start @ http://pastie.org/834549 ? |
| 20:57 | <TabAtkins> | JonathanNeal: You sending that out as a message to other people in your company or something? |
| 21:00 | <JonathanNeal> | Probably the co first, then the community. |
| 21:00 | <JonathanNeal> | Get people excited. |
| 21:01 | <JonathanNeal> | I'm not exactly a star blogger, so I figure it's more for a forum post :D |
| 21:01 | <TabAtkins> | Sounds pretty good so far, I think. |
| 21:02 | <JonathanNeal> | Cool |
| 21:05 | <TabAtkins> | Hmm, you say something about CSS3 features and gradients, but I'm not seeing any gradients in your CSS. |
| 21:05 | TabAtkins | is overly concerned with things he has worked on. |
| 21:06 | <JonathanNeal> | css3 gradients? Absolutely, all of our tabs. |
| 21:06 | <JonathanNeal> | All of our tabs, Tab :D |
| 21:06 | <TabAtkins> | never mind, I'm dumb. Was looking at the wrong file. |
| 21:07 | <TabAtkins> | Just now found the one with all the includes. ^_^ |
| 21:07 | <JonathanNeal> | What do you mean, you worked on? |
| 21:07 | <TabAtkins> | In that case, still disappointed. You're only doing webkit and MS gradients. ^_^ |
| 21:07 | <TabAtkins> | Firefox gradient support comes from the section of CSS3 Images I wrote. |
| 21:08 | <TabAtkins> | Your tabs would be -moz-linear-gradient(#ddd,#f0f0f0) |
| 21:08 | <JonathanNeal> | TabAtkins, say it ain't so! Maybe Firefox gradients didn't exist when I wrote them. |
| 21:08 | <JonathanNeal> | Are they in now? |
| 21:08 | <TabAtkins> | yup, in the public 3.6 |
| 21:09 | <JonathanNeal> | Well, I'm adding them right now, which means they'll make it onto the public site Monday. |
| 21:09 | <Necrathex> | that's why the -moz- and -webkit- stuff sucks |
| 21:09 | <TabAtkins> | Hmm? No, that's why it's good. Now that we have a solid draft, once the Images Module hits CR webkit can implement prefix-less support without breaking people who used their experiment prefixed version. |
| 21:11 | <Necrathex> | why would it break? |
| 21:11 | <JonathanNeal> | TabAtkins, will this make you happy? background-image: -moz-linear-gradient(#DDD, #F0F0F0); background-image: -webkit-gradient(linear, left top, left bottom, color-stop(0.0, #DDD), color-stop(1.0, #F0F0F0)); filter: progid:DXImageTransform.Microsoft.Gradient(GradientType=0, StartColorStr='#DDDDDD', EndColorStr='#F0F0F0'); |
| 21:11 | <TabAtkins> | I wrote the CSS gradients spec *in response* to webkit's experiments in the space, because while it was a great idea, the syntax frankly sucked. And now Moz has implemented experimental support for it, so if I end up still tweaking it a bit, things will be fine when they do the unprefixed version too. |
| 21:11 | <TabAtkins> | JonathanNeal: Yes. |
| 21:11 | <TabAtkins> | Necrathex: Because -webkit-gradient() uses a completely different syntax from what's currently in CSS3 Images. |
| 21:11 | <Necrathex> | ahh |
| 21:11 | <Necrathex> | okay, in that case it doesn't suck :) |
| 21:11 | <TabAtkins> | Check what JonathanNeal just posted to see how it looks. |
| 21:12 | <Necrathex> | what does, is that most sites that already use border-radius still don't show up in the new opera version, because it's only defined with prefix |
| 21:12 | <JonathanNeal> | TabAtkins, is writing this stuff your primary job or hobby? |
| 21:13 | <JonathanNeal> | And Firefox just decided to adopt it after you wrote it up? |
| 21:13 | <TabAtkins> | Was a hobby until about 2 weeks ago, when Google hired me. ^_^ |
| 21:13 | <JonathanNeal> | TabAtkins, that's awesome, I met a guy who works on Analytics at a party the other day. |
| 21:14 | <TabAtkins> | Firefox still waited until it was actually in a draft (after I'd been invited into the CSSWG and made co-editor of Images). |
| 21:14 | <JonathanNeal> | Not a geek party either, just happened to be there. Really nice guy, said they were hiring like crazy, but Liferay treats me too well and I have to admit, I love it here through the highs and lows. |
| 21:14 | <TabAtkins> | I can thank ROC for implementing it. He helped point out some ambiguities too. |
| 21:14 | <JonathanNeal> | Well, congratulations, TabAtkins |
| 21:14 | <TabAtkins> | Yeah, I love my current job too, but I just can't pass up the ability to work on specs full time. |
| 21:14 | <TabAtkins> | Thanks. ^_^ |
| 21:15 | <TabAtkins> | Necrathex: Is opera doing -o-border-radius, or border-radius? |
| 21:16 | <Necrathex> | the latter |
| 21:16 | <Necrathex> | just like with opacity, iirc |
| 21:17 | <TabAtkins> | Then yeah, that's true. But I believe mox and webkit are planning on dropping their vendor prefixes on border-radius too. |
| 21:17 | <Necrathex> | they already did for opacity, didn't they? |
| 21:17 | <TabAtkins> | Border-radius is in a weird spot, because everyone liked it so much that it became widespread while it was still officially 'experimental'. |
| 21:17 | <TabAtkins> | Yes. |
| 21:17 | <TabAtkins> | ('experimental' meaning 'not in a draft that's reached CR') |
| 21:18 | <Necrathex> | it seems opera doesn't use prefix at all for stuff that's in a spec or draft |
| 21:18 | <hsivonen> | HTML5 is popular and "experimental" according to that definition |
| 21:19 | <TabAtkins> | hsivonen: You know I'm talking about browsers' approach to CSS specifically. They treat other technologies differently. |
| 21:20 | <TabAtkins> | JonathanNeal: This shows up in one of your stylesheets, but has a 404: http://cdn.www.liferay.com/osb-theme/images/progress_bar/complete_gradient.png |
| 21:20 | <TabAtkins> | (happened to find it when i was first looking for the gradients you mentioned) |
| 21:21 | <hsivonen> | the CSS WG seems to take the W#C document maturity stuff more seriously, but the conflict between speccing time scales and the eagerness to do early implementation applies to both HTML and CSS |
| 21:21 | <Necrathex> | does opera support gradients yet? |
| 21:22 | <JonathanNeal> | TabAtkins, huh where is that coming from? |
| 21:22 | <JonathanNeal> | I'll look for it. |
| 21:22 | <TabAtkins> | This file: http://cdn.www.liferay.com/html/portal/css.jsp?browserId=other&themeId=osb_WAR_osbtheme&colorSchemeId=01&minifierType=css&languageId=en_US&t=1265764439000 |
| 21:22 | <JonathanNeal> | oh, css.jsp's, I hate those guys. |
| 21:22 | <JonathanNeal> | They're css injected by the application. |
| 21:23 | <TabAtkins> | ah, kk |
| 21:23 | <JonathanNeal> | I thought I was being so cool by adding all my boxmodel stuff to one file, but later it proved to be a big waste of time. |
| 21:23 | <JonathanNeal> | As if I'm going to add properties to all display block elements... |
| 21:24 | <JonathanNeal> | Hey, but the good news is, I think our main menu uses an image gradient, and now I can pull that out in exchange for the css3 equiv. |
| 21:27 | <TabAtkins> | hsivonen: Agreed, but CSS has a good fallback story to allow experimental naming conventions, which eases some of the early impl burden (you're at least not afraid of locking people into an early version by accident, for the most part). HTML doesn't, shrug. |
| 21:31 | TabAtkins | needs to spend an hour on Monday putting his company's site onto the new sectioning elements, because the outline generated by gsnedder's tool doesn't look right. |
| 21:32 | <JonathanNeal> | Google.com? |
| 21:32 | <JonathanNeal> | They don't care about standards, they care about speed :D |
| 21:32 | <TabAtkins> | Nah, still-current employer. igofigure.com |
| 21:32 | <TabAtkins> | I've got 2 weeks left with them. |
| 21:32 | <TabAtkins> | And I'm spending nearly all of that time cleaning up code. |
| 21:32 | <Necrathex> | any opera people here btw? :) |
| 21:32 | <TabAtkins> | Trying very strongly to avoid picking up new projects. |
| 21:33 | <Dashiva> | Necrathex: Several |
| 21:33 | <TabAtkins> | It's 9 or 10 over there, though, so they may not actually be around. |
| 21:34 | <JonathanNeal> | TabAtkins, you'll be working on drafts at G? |
| 21:34 | <Necrathex> | yea, i'm in the same timezone |
| 21:34 | <TabAtkins> | JonathanNeal: If everything goes as planned, yes. Right now it's looking like I'll join up with Chrome team and be allowed to spend most of my time on spec work. |
| 21:34 | <Necrathex> | i was wondering about a pretty serious crasher i found, but i guess this isn't the place.. |
| 21:35 | <TabAtkins> | Got a lot of stuff in the pipeline that I want to work on. |
| 21:36 | <Necrathex> | made a 20 lines html page that crashes every opera version past 9.5 :S |
| 21:37 | <TabAtkins> | Ooh, share? I'm curious. |
| 21:37 | <Necrathex> | http://niels.vg/media/opera_crash.html |
| 21:37 | <TabAtkins> | That's a very bizarre thing to crash on. |
| 21:37 | <Necrathex> | doesn't happen when i change the values, or put a doctype on it |
| 21:38 | <Necrathex> | so something in the quirks-mode i guess |
| 21:38 | <TabAtkins> | Well, that means it's related to the quirks-mode rendering, then. |
| 21:38 | <TabAtkins> | Yeah. |
| 21:39 | <Dashiva> | I figure it'll get passed on |
| 21:39 | <Necrathex> | bug is reported, forum topic opened, etc |
| 21:40 | <Dashiva> | Runaway memory allocation? |
| 21:42 | <Necrathex> | looks like an infinity loop of some kind to me |
| 21:42 | <Necrathex> | probably in the height calculation? |
| 21:42 | <Necrathex> | actually found it in a live website, www.enra.nl |
| 21:43 | <Necrathex> | they'll have 0 visitors using opera ^^ |
| 21:43 | <JonathanNeal> | What's great about putting some of these stuff in the portal is how much faster it will push adoption. |
| 21:43 | <TabAtkins> | Adoption within your company? |
| 21:44 | <JonathanNeal> | With anyone, any company who uses the Portal. |
| 21:44 | <TabAtkins> | Ah, I basically don't know what your company does. |
| 21:45 | <JonathanNeal> | We make Liferay Portal, which is ... Portal sofware. Think, like, a CMS> |
| 21:45 | <TabAtkins> | Ah, gotcha. |
| 21:47 | TabAtkins | is thinking of a wormhole that blasts out healing lasers. |
| 21:48 | <JonathanNeal> | That too. |
| 21:48 | <JonathanNeal> | But you know, every business expands. |
| 21:53 | <JonathanNeal> | That's great, Paul! |
| 22:34 | <JonathanNeal> | Has anyone written a universal controller for video / audio in <video>, Quicktime, and Flash? |
| 22:34 | <JonathanNeal> | For JS interaction. |
| 23:34 | <btipling> | how is one supposed to destroy an HTML5 audio elment? there is no .close() or .destroy() or anything like that when you create an Audio element via new Audio( url ) |
| 23:34 | <btipling> | and in Chrome Audio elements are never garbage colleged, they just continue to eat up cpu even after you have no reference to them anymore |
| 23:34 | <btipling> | collected* |
| 23:54 | <btipling> | hrm |
| 23:54 | <btipling> | "only once a media element to which no references exist has reached a point where no further audio remains to be played for that element (e.g. because the element is paused, or because the end of the clip has been reached, or because its playbackRate is 0.0) may the element be garbage collected." |
| 23:54 | <btipling> | maybe I'll do all of those and then it will be garbage collected |
