| 00:38 | <GPHemsley> | Does anyone know if there's an ECMAScript IRC channel? |
| 00:43 | <benvie> | kind of |
| 00:44 | <benvie> | on mozilla's irc there's #jslang |
| 00:44 | <benvie> | irc.mozilla.org |
| 00:50 | <GPHemsley> | ah, OK, thanks |
| 02:07 | <ubitux> | it would be pretty nice if there was a webvtt sample file |
| 02:08 | <ubitux> | kind of a conformance ref file making use of all the features |
| 02:11 | <Hixie> | ubitux: there's a sample file in the spec intro |
| 02:11 | <Hixie> | ubitux: (doesn't use all the features) |
| 02:11 | <ubitux> | it is very limited |
| 02:11 | <Hixie> | (i don't speak japanese so it's hard for me to make a good one that uses the vertical text features) |
| 02:12 | <ubitux> | well even not that, i believe a file with a utf-8 bom, different timing formats, multilines events, and various nested tags would actually help |
| 02:14 | <zewt> | (sounds like what you're actually looking for is a test suite) |
| 02:19 | <ubitux> | no, just a sample |
| 02:19 | <ubitux> | and sure, to be included in our own test suite at some point |
| 02:20 | <ubitux> | but well, for testing it's kind of nice too |
| 02:20 | <ubitux> | as well as having a quick overview of everything that needs support |
| 02:21 | <ubitux> | see for example http://unicorn.us.com/jacosub/demo.txt |
| 02:22 | <ubitux> | it's kind of helpful with advanced subtitles... |
| 02:25 | <ubitux> | should i open a bug on the bugzilla? |
| 02:34 | <ubitux> | here we go, this is now #18770 |
| 02:42 | <Hixie> | patches welcome :-) |
| 02:43 | <tantek> | Hixie, do you maintain a count of the # of times you end up saying that? ;) |
| 02:43 | <Hixie> | not that often :-) |
| 03:43 | <Von_Davidicus> | Why is it, when I put an <object> element in my document head, do I get errors in HTML5? Never had that problem in XHTML 1.0 or HTML 4.01 |
| 03:49 | <Von_Davidicus> | Well, I think I can call my first serious HTML5 project "finished". |
| 05:37 | <annevk> | ah |
| 05:37 | <annevk> | known issue |
| 05:37 | <annevk> | https://productforums.google.com/forum/?fromgroups=#!topic/google-plus-discuss/oMgOFgpRjzA |
| 05:38 | <annevk> | what a crap URLs btw |
| 05:38 | <Hixie> | Von_Davidicus: what do you mean by "get errors in HTML5"? |
| 05:38 | <Von_Davidicus> | Validation errors. |
| 05:38 | <Hixie> | ah. because it's not valid :-) |
| 05:39 | <Von_Davidicus> | Okay... why was <object> elements in the <head> element deemed bad? |
| 05:39 | <Hixie> | doesn't work |
| 05:39 | <Hixie> | (none of the browsers implemented it) |
| 05:39 | <Von_Davidicus> | Was it exclusively for the [declare] attribute, then? |
| 05:39 | <Hixie> | pretty much |
| 05:40 | <Von_Davidicus> | Oh, okay. |
| 05:40 | Von_Davidicus | uses it for SVG declarations and preloading images. |
| 05:40 | <Hixie> | does it work in browsers? |
| 05:40 | <Hixie> | i mean, do they put it in the <head> in the DOM? |
| 05:40 | Von_Davidicus | goes and checks. |
| 05:49 | <Von_Davidicus> | Hm. FizzleFox, Slopera, Google Groan, and are putting it in the <body>. |
| 05:50 | Von_Davidicus | is disappointed. |
| 05:52 | <Von_Davidicus> | Back. Somehow I ended up leaving. :( |
| 05:52 | <Hixie> | that's what i call interoperability, it's a good thing :-) |
| 05:52 | <Hixie> | (and also why it's non-conforming now) |
| 05:53 | Von_Davidicus | grumps at browsers. GRUMPS! * |
| 05:53 | <Hixie> | for following the spec? :-) |
| 05:54 | <Von_Davidicus> | For not following an outdated one! |
| 05:55 | Von_Davidicus | guesses he'll ave to put the <object> element somewhere in the body. |
| 06:01 | Von_Davidicus | sticks it just before the menu and changes the <object> element to <div>* |
| 06:01 | <Von_Davidicus> | Hm... seems to work just as well. http://www.mrinitialman.com/ |
| 06:06 | <Von_Davidicus> | I "preloaded" the image so my CSS rollovers would work properly. |
| 06:21 | Von_Davidicus | has recently finished an XML project more or less to his satisfaction. |
| 10:48 | <Ms2ger> | annevk, yt? |
| 11:20 | <annevk> | Ms2ger: yup |
| 11:21 | <Ms2ger> | Is it just me, or is the isEqualNode spec wrong if the context object has more attributes than the argument? |
| 11:23 | <annevk> | see under -> Element |
| 11:24 | <Ms2ger> | Ah, right |
| 11:24 | <Ms2ger> | Well hidden |
| 11:24 | <Ms2ger> | (Or, I can't read) |
| 11:24 | <annevk> | feel free to move it around as you see fit |
| 11:24 | <annevk> | it's not the best wording |
| 12:02 | <MikeSmith> | w3c bugzilla is down |
| 12:02 | <MikeSmith> | systems team is working on it |
| 12:02 | <MikeSmith> | other stuff down too |
| 12:57 | <odinho> | Chrome guys, think you have a regression, don't have old versions in though so can't really check. :-( |
| 12:57 | <Ms2ger> | ^ helpful guy |
| 12:57 | <odinho> | http://odin.s0.no/web/theora-video/video.htm << streaming video from icecast using ogg theora used to work. |
| 12:58 | <odinho> | Ms2ger: Hehe... I meant they're kinda hard to find. :-) |
| 13:00 | <odinho> | (that is, I found it for Windows, but no Linux) |
| 13:01 | <gsnedders> | odinho: BUT YOU NEED THE LATEST VERSION OR YOU'RE HOLDING THE WEB BACK! |
| 13:02 | <odinho> | gsnedders: But but... Regression. test. Help... Nice guy. |
| 13:02 | <gsnedders> | odinho: BE HORRIBLE. |
| 13:04 | <odinho> | gsnedders: Okay, I'll sniff it and say "get a better browser", then :-D |
| 13:39 | <jgraham> | "Umm, isn't the before method the wrong way round? I thought div2.before(div1) would put div2 before div1" - Opera developer on reading the new DOM4 mutation methods |
| 13:41 | <jgraham> | "there is a typo in the spec as well" |
| 13:42 | <jgraham> | "'the append nodes method must run these steps' should be nodes" |
| 13:42 | <jgraham> | *shoulfd be node |
| 13:42 | <jgraham> | (I am just quoting haven't checked) |
| 13:42 | <jgraham> | Ms2ger, annevk ^ |
| 13:43 | <Ms2ger> | Those new things are annevk's thing |
| 13:43 | <annevk> | please file a bug on the typo |
| 13:43 | <annevk> | checking the other thing now |
| 13:44 | <jgraham> | The other thing isn't a "bug" |
| 13:44 | <annevk> | the before method matches http://api.jquery.com/before/ |
| 13:44 | <annevk> | afaict |
| 13:44 | <jgraham> | It's a "oh this method does something different to what I expect" |
| 13:45 | <annevk> | jl/bratell prolly haven't used jQuery like the rest of the web? :) |
| 13:45 | <jgraham> | Right, this is bratell who afaik is not too familiar with jQuery |
| 13:45 | <jgraham> | But the number of future web developers who haven't used jQuery >> number of current web developers who have |
| 13:46 | <annevk> | maybe it helps if he thinks of it like insertBefore, but with the method directly on the object on which to insert things before |
| 13:46 | <annevk> | also, Maciej suggested renaming them to addBefore/addAfter |
| 13:47 | <jgraham> | But usually a mehod on an object mutates that object, not its argument |
| 13:47 | <annevk> | does that help? |
| 13:47 | <Workshiva> | annevk: But if it |
| 13:47 | <annevk> | neither the object nor the argument is mutated |
| 13:47 | <annevk> | the parent is |
| 13:47 | <Workshiva> | it's on the object on which to insert things before, it's not like insertBefore |
| 13:47 | <annevk> | Workshiva: well yeah, that's kind of the point |
| 13:48 | <jgraham> | If the node didn't have a parent and now does have a parent that seems like a more substantial change than being the N+1th child |
| 13:48 | <jgraham> | instead of the Nth |
| 13:48 | <annevk> | before() cannot give a node a parent |
| 13:49 | <Workshiva> | insertBefore can |
| 13:49 | <jgraham> | Really? |
| 13:49 | <annevk> | Workshiva: how? |
| 13:49 | <jgraham> | So if I p = createElement("p") and document.body.before(p), p doesn't end up with a parent node? |
| 13:50 | <annevk> | jgraham: I thought you meant body ending up with a parent node |
| 13:50 | <Workshiva> | annevk: p = createElement("p"); // no parent // document.body.firstChild.insertBefore(p) // now has a parent |
| 13:50 | <Workshiva> | Um, that was before I guess |
| 13:50 | <jgraham> | Right, it woudl if you write p.before(document.body) |
| 13:50 | <Workshiva> | See, it's already confusing me! |
| 13:51 | <jgraham> | (i.e. if the this object and the argument were swapped) |
| 13:51 | <jgraham> | Anyway, I don't know |
| 13:51 | <annevk> | again, comments on the suggested renaming welcome |
| 13:52 | <jgraham> | It seems to me that non-verb methods are pretty terrible and that this design is potentially confusing |
| 13:52 | <annevk> | updates next week ideally |
| 13:53 | <Workshiva> | assignSiblingRelationshipPreceding() |
| 13:57 | <jgraham> | element.assignRelationshipSiblingEarlier(), aliased as element.aRSE() to keep the names short and memorable |
| 14:04 | <odinho> | lol |
| 14:10 | <odinho> | I like addBefore, addAfter, -- they even have a better symmetry than before/insertBefore. But I think we are smart people and will be perfectly able to use whatever. But no more hard-to-write and long words like "stopPropagation". :-) |
| 14:11 | <Workshiva> | If we're going to be jquerish, why not "andThen" and "butFirst"? |
| 14:12 | <annevk> | Workshiva: because only in your mind is that jQueryish |
| 14:13 | <Workshiva> | My mind sees no problem with that |
| 14:13 | <zewt> | stopPropagation isn't overly long (close to borderline, maybe, but only that); now, findElementById ... |
| 14:13 | <annevk> | get* |
| 14:13 | <zewt> | y |
| 14:13 | <Workshiva> | Clearly you meant to say getElementsByClassName |
| 14:13 | <zewt> | <- just got out of shower, slack requested :P |
| 14:14 | <annevk> | only granted for out of bed or drunk |
| 14:14 | <odinho> | zewt: I have problems writing it... Always have to stop and think about it... It could be cancelPropagation. There's also the preventDefault, cancelDefault, stopDefault. What was it now again :] |
| 14:14 | jgraham | wonders if zewt put any clothes on |
| 14:14 | <annevk> | did jgraham just kill the conversation? |
| 14:15 | <zewt> | cool, adobe air just popped up a dialog saying "we have an update!!! [update now] [update later]" ... and "update now" was greyed out |
| 14:15 | <odinho> | Everyone picturing zewt naked |
| 14:15 | <annevk> | I have trouble picturing zewt |
| 14:15 | <Workshiva> | zewt sounds like a kind of newt |
| 14:16 | <jgraham> | It sounds a bit like a newt in a boyband |
| 14:16 | <Workshiva> | Maybe a very intense one, like zomg |
| 14:16 | <zewt> | annevk: if that makes it harder to picture me naked, then we can consider that a win for all of us |
| 14:16 | <jgraham> | Right, exactly |
| 14:16 | <Workshiva> | And newts tend to be naked, so I'm not really very affected |
| 14:17 | <jgraham> | At risk on underailing the conversation, I think I like zewt's proposed Sync API better than sicking's |
| 14:17 | <zewt> | (no! don't derail a conversation about a bunch of guys picturing me naked!) |
| 14:17 | <jgraham> | But I am not convinced I have enough experience with this kind of thing to offer a convincing argument |
| 14:18 | <Ms2ger> | Wait, what did I just walk into? |
| 14:18 | <jgraham> | (I also like the idea that javascript should grow some primitives to make async stuff more linear) |
| 14:18 | <jgraham> | Ms2ger: zewt in the nude |
| 14:18 | <odinho> | Ms2ger: What did you just walk out of? (shower?) |
| 14:19 | <Workshiva> | jgraham: "zewde" |
| 14:19 | <zewt> | jgraham: in principle--though it's probably too late in the game for this to fit--the "wait for message" api could have been the basis for all worker sync apis |
| 14:19 | <zewt> | eg. xhr.send(); getMessage(xhr.send()); // wait for what would have been the next message to the xhr client |
| 14:20 | Ms2ger | puts on http://www.leekspin.com/ for fun |
| 14:20 | <zewt> | not quite as nice as a real sync API that returns stuff and throws exceptions, but without the need to fork every single async API into two interfaces |
| 14:21 | <zewt> | reader = new FileReader(blob); arrayBuffer = getMessage(reader); |
| 14:21 | <zewt> | oh well |
| 15:40 | <matjas> | would other ECMAScript engine implementors be open to escaping ', < and > as well as " in the String.prototype HTML extensions, as per V8? http://code.google.com/p/v8/issues/detail?id=2217#c15 |
| 15:42 | <matjas> | seems a bit over the top to me, but hey |
| 15:42 | <jgraham> | matjas: I'm pretty sure we would try anything once |
| 15:43 | <matjas> | jgraham: reason V8 does the extra escapes is: '_'.link('<script>...').substring(1); |
| 15:43 | <matjas> | but, like, *everything* is potentially dangerous |
| 15:45 | <dglazkov> | good morning, Whatwg! |
| 15:53 | <deane> | dglazkov: Good morning. |
| 16:00 | <annevk> | jgraham: going to update james.html5 at some point again? |
| 16:03 | <jgraham> | annevk: What happened to whatever used to be there? |
| 16:07 | <annevk> | jgraham: I don't know, I think you might have cleared it when there was some spam issue |
| 16:08 | <annevk> | with the idea of putting it back later |
| 16:16 | <niloy> | Please bring back WebSQL T____T, if not for the browser, then atleast for the native apps build with web techs T____T |
| 16:17 | <annevk> | what is T____T? |
| 16:17 | <zewt> | T____________________________________T |
| 16:17 | <annevk> | it's an abbreviation? |
| 16:17 | <zewt> | perhaps "thrift" or "talent" |
| 16:17 | <annevk> | I don't understand the longer form either |
| 16:18 | <annevk> | I take he might have meant the variant with three underscores, as http://www.urbandictionary.com/define.php?term=T____T seems sketchy in this context |
| 16:18 | <darobin> | annevk: I think it's one of them emoticons thingies that the young'uns use nowadays |
| 16:19 | <karlcow> | >One issue is that the W3C geolocation API, used by web browsers to determine your location, doesn’t tell you how your location was acquired. Was it from a wireless network? A cell-tower? A passing drone, overhead? There’s no way to know and by extension no way to adjust expectations. — http://nearfuturelaboratory.com/2012/09/04/the-atlas-of-desire/ |
| 16:19 | <zewt> | ( ﾟ∀ﾟ) |
| 16:20 | <jgraham> | darobin: All I am getting from it is "I'm a frog with crucifixes for eyes". Whichis interesting but doesn't seem that relevant to the point. |
| 16:21 | <annevk> | darobin: going to be one long blog post, huh? ;) |
| 16:21 | <darobin> | jgraham: I see no reason why frogs with crucifixes for eyes should be disenfranchised from the Web platform more than anyone else |
| 16:22 | <darobin> | annevk: beauty doesn't count the number of the years |
| 16:26 | <niloy> | T_T is emote of crying, the T's are eyes |
| 16:26 | <niloy> | I am just crying sooooooooo much T_______T |
| 16:28 | <niloy> | Well I hope Opera & Webkit keeps supporting them, we will all use it and someday and web & native apps, and someday everyone else will also have to support it! |
| 16:30 | <zewt> | javascript not scoping var to blocks repeatedly makes me :| |
| 16:30 | <darobin> | you just need to be more creative with your variable naming |
| 16:30 | <zewt> | i'd rather the scoping mechanism do its job, heh |
| 16:31 | <zewt> | (python has the same problem, unfortunately) |
| 16:31 | <smaug____> | use let :) |
| 16:31 | <niloy> | or iife |
| 16:31 | <Hixie> | odinho: there's a script they have that will do a binary search through old webkit builds to find when a regression happened |
| 16:32 | <odinho> | niloy: We (Opera) don't have any removal plans for websql. :-) But things can change, bla bla bla normal disclaimer talk. |
| 16:32 | <niloy> | odinho, yay we will win someday ^_^ |
| 16:33 | <zewt> | smaug__________: if it actually worked in something other than FF, maybe :) |
| 16:33 | jwalden | resists the urge to get all stabby about websql |
| 16:34 | <odinho> | Hixie: I remembered I had a virtualbox with Windows, so I was able to find an old chrome (8) and it worked there. But I'm onto something else now, I have the patience of a fly :P |
| 16:35 | <niloy> | btw for those using urban dictionary: http://www.urbandictionary.com/define.php?term=t_t |
| 16:35 | <zewt> | heh, i had to do a git bisect recently ... holy boring |
| 16:36 | <Hixie> | odinho: it's a pretty sweet script, it does all the downloading and binary search and stuff for you. But I forget what it's called. :-) |
| 16:36 | <Ms2ger> | Hixie, mozregression? :) |
| 16:37 | <Hixie> | it's the same idea, yeah |
| 16:47 | <annevk> | Ms2ger: added a WebHook URL for DOM4 since you keep pushing |
| 16:47 | <annevk> | took me 10 minutes to learn shell/Apache interaction and that was that |
| 16:48 | <Ms2ger> | \o/ |
| 16:48 | <annevk> | relearn actually |
| 16:48 | <annevk> | not the first post-commit hook |
| 16:49 | <annevk> | Hixie: it comes down to curl https://raw.github.com/WHATWG/dom/master/dom-core.html > dom-core.html (plus some stuff to make it not 500) so if you ever want to reclaim the subdomains that should be easy enough |
| 16:50 | <Hixie> | heh k |
| 16:50 | <Hixie> | seems unlikely to ever occur :-) |
| 16:51 | <jgraham> | Hixie: I'm noy sure it matters, but if you have document A with iframe B which itself has iframe C then it seems to me that, per the current spec, C can cause document.open to run in B during C's beforeunload |
| 16:52 | <Hixie> | jgraham: seems plausible |
| 16:52 | <Hixie> | jgraham: is that a problem? |
| 16:52 | <jgraham> | Hixie: I don't know |
| 16:52 | <Hixie> | jgraham: i guess it'll do a double unload? |
| 16:52 | <Hixie> | file a bug if you want me to look at it in more detail |
| 16:53 | <jgraham> | OK, I will see if I think of a case where it's a problem |
| 17:04 | <Hixie> | jgraham: as far as i can tell, the case simplifies to window A contains iframe B, B is navigated, B asynchronously invokes a task that synchronously does a beforeunload, during which A is document.open(), which synchronously does B's beforeunload reentrantly (but A and B can't document.open() at this point), followed by B's unload (d.o() is again neutered here), at which point B's original task does the original unload again. |
| 17:04 | <Hixie> | jgraham: which seems bad |
| 17:18 | <zewt> | it's not the future until i can stop worrying about starting a paragraph in an email with the word "From" |
| 17:46 | GPHemsley | wonders what past zewt is living in. |
| 18:21 | <jwalden> | GPHemsley: judging by every mailman web view in existence having >From in various bits of mail, the present past |
| 18:21 | <GPHemsley> | ah |
| 19:19 | <annevk> | Ms2ger: it's had that text since Hixie wrote it back in 2005/2006 or so |
| 19:20 | <annevk> | Ms2ger: need a little more in the bug than "is that intentional?" to make it actionable |
| 19:20 | <Ms2ger> | Make it "test this, please", then :): |
| 19:20 | <Hixie> | hm? |
| 19:21 | <Ms2ger> | XHR.getAllResponseHeaders |
| 19:21 | <Ms2ger> | And whether it should end with \r\n |
| 19:21 | <Hixie> | ah |
| 19:24 | <annevk> | Ms2ger: if you split it with JavaScript later not having it end in \r\n is nicer |
| 19:24 | <Ms2ger> | I guess |
| 19:24 | <Ms2ger> | What do browsers do? |
| 19:24 | <annevk> | Ms2ger: but yeah, someone should update the XHR test suite, I tried to have odinho do that, but he found someone else, who likely dropped the ball |
| 22:04 | <Hixie> | zewt: even if it's just a message to say "this has passed to something that is a descendant worker of X" or "...not an ancestor of X" or whatnot, it's still essentially impossible because if you send something from X to A to B, the message that it got from A to B might not reach X until long after X has started sending messages using this feature |
| 22:05 | <zewt> | Hixie: it doesn't matter if otehr messages are sent first |
| 22:05 | <zewt> | or other, for that matter |
| 22:06 | <Hixie> | ok i'm reading your message to sicking now and i'm even more confused :-) |
| 22:06 | <zewt> | all this is intended to do is guarantee that deadlocks can't happen in a "block until a message is received" API |
| 22:06 | <Hixie> | suppose i create a message port |
| 22:06 | <Hixie> | can i send a blocking message? |
| 22:06 | <zewt> | there's no such thing as "sending a blocking message" |
| 22:06 | <zewt> | there's only "block until a message is received" |
| 22:06 | <zewt> | (a message is a message is a message) |
| 22:06 | <Hixie> | so suppose i create a message port |
| 22:06 | <Hixie> | can i block until i receive a message? |
| 22:07 | <zewt> | so long as the "blocking allowed" flag remains set |
| 22:07 | <Hixie> | but the other side of the port is by definition in the same thread, so that seems bad? |
| 22:08 | <zewt> | you mean the "blocking on a port when you hold both ports" case? that's handled separately (the final "additionally") |
| 22:09 | <Hixie> | ah, i see |
| 22:09 | <Hixie> | i assume when you say "When a MessagePort is transferred to the current thread" you mean to imply that the internal flags and states are also cloned? |
| 22:10 | <Hixie> | it's unclear whether you mean MessagePort or "port" |
| 22:10 | <zewt> | yes, the "blocking allowed" flag being cleared persists across clones |
| 22:10 | <Hixie> | (MessagePort objects never actually change thread) |
| 22:10 | <zewt> | (once it's cleared it's cleared permanently) |
| 22:10 | <Hixie> | ok... |
| 22:11 | <Hixie> | and presumably this only happens in workers |
| 22:11 | <zewt> | yes (it doesn't matter if it happens in the UI thread, since the getMessage API just won't exist there) |
| 22:12 | <Hixie> | by "this" i mean the API |
| 22:12 | <zewt> | also, this is only really relevant for dedicated workers; any time you transfer to a shared worker the final "otherwise" kicks in and no blocking is ever allowed |
| 22:12 | <zewt> | (in either direction) |
| 22:13 | <Hixie> | seems kinda weird, but if browsers are willing to implement it, i guess it's ok |
| 22:17 | <zewt> | (correcting above--this does need to happen for ports going to the UI thread, since the other port is probably still inside a worker) |
| 23:11 | <cpearce> | annevk: ping |
| 23:12 | cpearce | suspects annevk is fast asleep... |