2015-01-01 [18:51:54.0000] Happy New Year, y'all! [19:17:33.0000] gsnedders: HNY [22:13:18.0000] jesus, this is a new top reason for me to get the hell away from firefox: i open a new browser window and i'm blinded by a bright yellow donation beg screen [00:02:28.0000] Happy 2015 everyone [00:02:46.0000] /me is starting it with a cold [00:16:47.0000] I wonder why things like https://github.com/volarvideo/webvtt/commit/c0cc14cf5bb4611e5b72cfca33a2dbe88226eb4e#diff-09461573a85e2d94f056dd6814769042L456 are true [00:16:55.0000] Shouldn't the JIT take care of that? [00:25:13.0000] Happy New Year annevk hope your cold goes away soon [00:26:42.0000] MikeSmith: HNY to you too! Saw you were enjoying the snow :-) [00:27:57.0000] hah yeah [00:30:38.0000] annevk: a regular expression being faster seems counterintuitive [00:31:26.0000] Yeah, though JITting is weird [02:43:06.0000] Simple regular expressions get compiled themselves, so it's not that surprising they're fast [02:43:11.0000] Blame sunspider [10:48:32.0000] pretty much all regular expressions get compiled these days [10:52:32.0000] For large values of "simple" then :) [10:52:40.0000] Or should that be small?… 2015-01-02 [16:41:43.0000] soft-keyboard backspace key doesn't work in the github browser-based editor thing in Firefox mobile [23:10:46.0000] filed a bug> [23:10:49.0000] ? [23:11:25.0000] which github brower-based editor thing are we talking about by the way? [23:37:09.0000] roc: haven't filed a bug yet but will now [23:37:39.0000] roc: I mean the editor that github gives you to edit a file in the repo [23:37:52.0000] e.g., a readme [23:39:37.0000] you have to be viewing github in the desktop mode it provides, instead of the default mobile view [00:09:20.0000] roc: filed https://bugzilla.mozilla.org/show_bug.cgi?id=1117051 [00:16:29.0000] thanks [04:49:32.0000] can anyone elaborate : "object that does not have the default behaviour for one or more of the essential internal methods that must be supported by all objects." [04:52:27.0000] "essential internal methods" [04:55:10.0000] there are a bunch of internal methods, like [[OwnPropertyKeys]] (for example), there are defaults, but exotic objects often need to have special versions [04:57:08.0000] caitp, can you give an example of one such object that is exotic? [04:58:12.0000] Array and String come to mind [04:58:37.0000] direct proxies [04:59:38.0000] special versions? [05:00:06.0000] that means they don't have all the internal methods? [05:00:35.0000] they don't have 'default behaviour' hmm [05:00:40.0000] i believe so, for instance [[Construct]] is not present in bound functions [05:00:48.0000] but [[Call]] is [05:01:42.0000] caitp, strange behaviours [05:06:37.0000] different behaviours are like, not being able to define an accessor for Array's "length" property [05:09:55.0000] caitp, but it's possible for {} [05:10:19.0000] essential internal methods that must be supported by all objects, the spec does not list them all, or did I miss it [05:12:56.0000] https://people.mozilla.org/%7Ejorendorff/es6-draft.html#table-5 [05:15:36.0000] caitp, so what's different in Array that it become exotic? Because it does not have, [[ Get ]] [[ Set ]] ? [05:16:18.0000] it has overridden versions of several essential internal methods [05:16:47.0000] and there is no way to find out what has been overridden ? [05:16:55.0000] well, just [[DefineOwnProperty]] I guess [05:18:13.0000] thanks caitp [05:18:15.0000] it's all in the spec, look at section 9.4 [05:18:33.0000] and 9.5 [05:18:40.0000] sure [06:27:23.0000] caitp, how do you read the spec, end-to-end ? I normally skim sections, but it's preferred to read end-to-end twice and then sections... 2015-01-03 [00:04:13.0000] gsnedders, good night [10:08:46.0000] what's wrong with class CustomButton extends HTMLButtonElement { constructor() { } }; let myButton = new CustomButton; document.appendChild(myButton); ? [10:10:37.0000] TypeError: Failed to execute 'appendChild' on 'Node': parameter 1 is not of type 'Node'....how come? 2015-01-04 [19:03:41.0000] deriving from Node doesn't make you a Node 2015-01-05 [05:57:07.0000] class MyImage extends HTMLImageElement {} [05:57:12.0000] ^ makes sense ? [05:57:27.0000] I get a Uncaught TypeError: Illegal constructor [05:57:33.0000] w.r.t ES6 classes [05:58:02.0000] no implementation supports extending host objects yet [05:58:10.0000] :( [05:58:33.0000] hopefully one day it will be possible :D [05:58:50.0000] caitp-, it's an interface right? [05:58:55.0000] I meant, HTMLImageElement [05:59:25.0000] sure [05:59:29.0000] more over we can do a document.registerElement('my-img',MyImage) [05:59:29.0000] how come that works! [05:59:56.0000] so, in that case the class extends the interface? [05:59:56.0000] /me shudders [06:00:01.0000] #confused [06:00:05.0000] It doesn't work now [06:00:06.0000] it "works", but not really [06:00:13.0000] Some people are trying to make it work [06:00:20.0000] I still don't see the point [06:01:13.0000] right now, in Safari for example, if you subclassed an object "the old fashioned way", you'd have the right prototype, but the C++ object would not be the right type, and would not be initialized [06:01:26.0000] https://gist.github.com/hemanth/e6bb66141101b52a76fe [06:01:27.0000] so, if you tried to access any properties from the superclass, it would throw and that would suck [06:01:49.0000] native code can't be accessed yet...hrrm [06:01:55.0000] DOM should catchup with ES6 [06:02:11.0000] Why [06:02:30.0000] So that we can extends them ? [06:02:59.0000] well you definitely want to be able to extend builtin classes, which you can't do right now [06:03:10.0000] and if you can extend Array, why not NodeList [06:03:20.0000] or one of the Matrix classes [06:03:50.0000] Why [06:03:52.0000] caitp-, but, if the class has createdCallback method and if I do a document.registerElement it works just like any other custom element [06:05:20.0000] because if you can't, that's really weird ;) unless tc39 introduces a concept of "final" classes, which would help explain the confusing behaviour better [06:06:05.0000] caitp-, started with a silly node module extending Array -> https://github.com/hemanth/node-arey/blob/master/index-es6.js ;) [06:06:26.0000] yeahhhhh that doesn't quite work the way you expect though :( [06:06:55.0000] which? the custom element or the node module or both?! [06:06:56.0000] hemanth: see recent blink-dev discussion and perhaps also on es-discuss with respect to subclassing [06:06:58.0000] and in all engines, Array.isArray() would fail for it, which it technically shouldn't [06:07:20.0000] hemanth: ES6 needs to solve some issues and seems to be opting out at the last minute :-( [06:07:46.0000] hemanth: https://bugzilla.mozilla.org/show_bug.cgi?id=1117172 lists some of the things nobody is decided on [06:08:37.0000] hemanth: the reason you can pass it to registerElement is because the registerElement implementation looks at the passed object's prototype property, bit of a hack [06:11:15.0000] annevk, ha ha, it would fail if I do a new Img() on register after extending :/ [06:11:49.0000] annevk, Do you think class is not a goof feature of ES6? I have many against it and only few pro [06:12:03.0000] hemanth: the problem is that TC39 thought they were going to explain all of DOM and then they didn't [06:12:32.0000] hemanth: I don't know [06:12:35.0000] :( [06:12:43.0000] it's a bit hard [06:12:48.0000] hemanth: but the syntax you used above needs to work at some point, we're trying to figure out how [06:13:53.0000] annevk,if I extend a HTMLSpanElement element it can set it's textContent, but for an ImageElement, I can't set it's src [06:17:50.0000] it would cool to have a document.deRegisterElement ;) ? ( or is it dumb ) [06:17:55.0000] hemanth: not even if you use the construct? [06:19:23.0000] hemanth: there's a ton of problems though, none of this was designed with extensibility in mind [06:20:17.0000] annevk, I get an Illegal construct [06:20:22.0000] hemanth: what would deRegisterElement do? [06:20:41.0000] hemanth: it might be interesting to have more registry primitives though, agreed, so you can more completely control it [06:20:44.0000] you'll get an illegal construct if you subclass Map and call super() too [06:20:44.0000] did not try thing [06:21:19.0000] hemanth: oh yeah, actual ES6 subclassing does not work yet in any browser [06:21:19.0000] caitp-, hrrmm phew.. [06:21:31.0000] hemanth: because it's not defined per above [06:21:42.0000] it "sort of" works for user-defined classes [06:21:57.0000] annevk, so that I can register element with the same name with functionality changed dynamically? [06:22:06.0000] annevk, I'm using 6to5 heh heh, does IE support it? [06:22:26.0000] that was w.r.t deRegisterElement [06:22:49.0000] hemanth: as per caitp- the only subclassing that works is of your own classes, you can't subclass builtins or platform objects [06:22:53.0000] as of now we can't do a registerElement('cat-say'); twice [06:23:02.0000] chakra's class support is roughly equivalent to v8's right now [06:23:09.0000] no builtins, no host objects [06:23:37.0000] Array is not an exotic object in ES6 right? [06:23:42.0000] it is [06:24:14.0000] caitp-, no builtins, no host objects as per the spec or the current implementation ? [06:24:46.0000] the current implementation... per dmitry's comments, the spec doesn't really do an adequate job of explaining how it can/should work for DOM stuff [06:24:57.0000] it could possibly work for builtins (but doesn't anywhere yet) [06:26:27.0000] oh, ok [06:27:19.0000] class Fetch extends XMLHttpRequest() {} #lolz [06:27:36.0000] the issues are about allocation of the object (especially for DOM objects), about passing arguments to the superconstructor (arbitrary arguments) --- DOM objects don't really have callable constructors right now, there are a few other issues [06:28:57.0000] well, XHR does have a callable constructor, not counting that :> [06:29:48.0000] heh heh [06:31:13.0000] lot of this was discussed at https://groups.google.com/forum/#!topic/v8-users/bwEPA3vcscg [06:38:59.0000] reading... [06:43:45.0000] I think a lot of posters in that thread failed to CC v8-users and kept it to blink-dev, might be better to read the blink-dev version :p [06:43:57.0000] anyways, people are working on it, it's just not there yet [07:05:40.0000] caitp-: the problem is that Google Groups does not accept posts from members that are not subscribed, even if they are subscribed to some other group and are not spammers there [07:05:56.0000] ah [07:05:59.0000] caitp-: so e.g. my posts would make it through to blink-dev, but not elsewhere [07:11:05.0000] /me needs some nom noms [07:49:22.0000] https://github.com/google/end-to-end/wiki/Threat-model "we discovered that Chrome Extensions can also use chrome.tabs.captureVisibleTab to take a screenshot of current tab in any browser window" [07:49:41.0000] that seems like not such a great feature [07:55:33.0000] what would you expect this to do: http://jsfiddle.net/n8rt8saL/ [07:56:03.0000] ^--- annevk, you'd probably be good at that one [07:56:48.0000] caitp-: https://www.w3.org/Bugs/Public/show_bug.cgi?id=26431 [07:56:49.0000] chromium decides to rewrite 1.2 to 1.0.0.2, while other browsers I have available don't :< I feel like that's not right [08:00:33.0000] MikeSmith: installing extensions you don't trust seems like a losing game [08:01:55.0000] that's kind of not fun :( [08:04:16.0000] jgraham: yeah I guess given how extensions work and there's a lot more invasive things than that a malicious extension could do [08:11:33.0000] but what if you can't visit that shady looking gambling/smut/whatever website without installing the shady extension which is not vetted by anyone as being trustworthy? [08:15:02.0000] I think the problem is more normally that you visit your totally trustworthy banking site having installed an extension / native app / whatever that you mistakenly believed was trustworthy (e.g. because you had no means to verify it wasn't) [08:16:25.0000] putting the onus of deciding who to trust on users is problematic because a lot of users don't know that it's even something to think about :( [08:19:03.0000] I wonder if Google will standardize the protocols used by http://chrome.blogspot.ch/2015/01/introducing-google-cast-for-audio.html [08:19:06.0000] and even when you're dealing with shady sites, grandpa has an incentive to gamble on horse racing, he might not have an incentive to think about "what will this application be able to do, what data of mine does it have, etc" [13:24:31.0000] annevk: so far Chromecast has been a very closed device. [13:25:30.0000] annevk: but I think it's worth asking our standards-friendly friends at Google to see if there's anything they can do 2015-01-06 [18:29:24.0000] TLS 1.0 support is disabled for whatwg.org, right? [18:29:55.0000] ah no I mean the other way around [18:30:11.0000] 1.1 and 1.2 not supported [22:41:33.0000] MikeSmith: should get fixed this quarter by DreamHost [23:23:08.0000] annevk: ah ok yeah I vaguely remember asking about this earlier [23:23:25.0000] MikeSmith: sorry you had to fix all that URL bug spam [23:23:36.0000] no worries [23:23:44.0000] I should have caught it before I did [23:23:51.0000] MikeSmith: I initially thought a bunch of actual work happened [23:23:57.0000] I wasn't checking my e-mail earlier today [23:23:58.0000] hahah [23:24:04.0000] yah me too [23:24:12.0000] so naïve [23:24:17.0000] haha [23:24:30.0000] no, you're an optimist [23:24:43.0000] glass half full and all that [23:24:51.0000] most of the time [23:24:55.0000] that's why people love you [23:26:30.0000] annevk: so btw are you aware of any trend with sites disabling TLS 1.0 support? [23:26:39.0000] or any guidance that suggests doing that? [23:27:01.0000] I understand sites disabling SSLv3 [23:27:36.0000] MikeSmith: https://wiki.mozilla.org/Security/Server_Side_TLS#Modern_compatibility [23:27:47.0000] /me looks [23:28:47.0000] MikeSmith: Adam Langley keeps saying this btw: "This seems like a good moment to reiterate that everything less than TLS 1.2 with an AEAD cipher suite is cryptographically broken." [23:29:36.0000] that seems pretty drastic [23:29:53.0000] and I hope it's not true [23:30:38.0000] I wonder there's some risk of boy-who-cried-wolf right now around some of the rhetoric with regard to TLS [23:31:04.0000] it alienates some people who would otherwise be more receptive and open-minded [23:31:20.0000] Adam Langley knows what he's talking about, but cryptographically broken doesn't mean that attacks are necessarily feasible I believe [23:31:23.0000] or maybe it's needed to wake more people up, I dunno [23:31:28.0000] ok [23:31:55.0000] There's certainly a lot of weird rhetoric found on www-tag... [23:32:08.0000] well that's crazy town out there [23:32:22.0000] those two guys [23:32:33.0000] it's like a comedy act [23:33:13.0000] The biggest hurdle is Mixed Content [23:33:19.0000] yup [23:34:21.0000] certainly I know that for trying to move all of w3.org to TLS that's a major issue [23:34:57.0000] but there are other costly content-migration issues with it [23:35:41.0000] that doesn't mean it shouldn't be done, or can't be [23:36:06.0000] But W3C doesn't have that much embedded stuff, does it? [23:36:46.0000] Most resources seem same-origin [23:37:14.0000] W3C could at least start by not redirecting away from HTTPS [23:38:27.0000] yeah w3c doesn't have nearly has much embedded content per page as other sites (as far as the mixed-content issue goes) but it still has tons of pages with small amounts [23:38:54.0000] especially stylesheets [23:39:31.0000] yeah I don't understand why w3c is still redirecting away from HTTPS [23:40:36.0000] but I do know that the w3c systems team is having a lot of discussions about how to deal with this stuff, and deployment plans for the coming months [23:41:49.0000] anyway, about TLS 1.0 on the server side, in practice right now, disabling TLS 1.0 server support means that a lot of clients aren't going to be able to access your content, right? [23:42:09.0000] (let alone disabling TLS 1.1 support) [23:43:42.0000] in other news I see that WebKit is planning to remove their Shared Workers code (after having stopped shipping it some time back) [23:44:34.0000] dunno how big of a difference any of that is, given no support for Shared Workers on any mobile UAs, nor in IE [23:44:56.0000] Not even in Servo [23:45:31.0000] If they ever want service workers... But from that email it seemed like the current implementation was not going to work for that anyway [23:45:48.0000] MikeSmith: not sure what TLSv10 is needed for [23:46:54.0000] annevk: I thought it was that some UAs still only have TLSv10 support, and no v11 or v12 support [23:48:59.0000] hmm I see nobody even asked for Shared Workers support at https://wpdev.uservoice.com/forums/257854-internet-explorer-platform?query=Shared [23:49:43.0000] though Service Worker got a ton of votes [23:49:45.0000] https://wpdev.uservoice.com/forums/257854-internet-explorer-platform?query=worker [00:14:33.0000] meow [00:36:38.0000] opensearch plugin this looks pretty cool [00:36:49.0000] https://www.w3.org/Bugs/Public/show_bug.cgi?id=27754 from Domenic [00:36:54.0000] https://developer.mozilla.org/en-US/Add-ons/Creating_OpenSearch_plugins_for_Firefox [00:39:29.0000] works in Chrome too http://browserfame.com/2071/opensearch-plugin-chrome [00:39:44.0000] though the UI isn't as discoverable [01:39:23.0000] http://a9.com/-/spec/opensearch/1.1/ is dead [01:39:28.0000] namespaces \o/ [01:41:42.0000] http://www.opensearch.org/Specifications/OpenSearch/1.1 [04:14:12.0000] https://bugzilla.mozilla.org/show_bug.cgi?id=163050 "Create New HTML Tag: " [04:14:13.0000] Status: VERIFIED FIXED [04:14:18.0000] Whiteboard: notfixed [04:14:48.0000] (via smontagu on another server) [04:22:03.0000] meow [04:53:26.0000] annevk, if you are around, after extending the image element, i tried that did not work either :( [05:17:44.0000] hemanth: works for me [05:17:56.0000] document.registerElement('test-img', { prototype: { someattribute: "x" }, extends: 'img' }) [05:18:06.0000] var test = document.createElement("img", "test-img") [05:18:07.0000] test.src = "image" [05:18:07.0000] document.body.appendChild(test) [05:33:36.0000] annevk, I have done class TestImage extends HTMLImageElement {} [05:34:40.0000] hemanth: when everyone said yesterday that class X extends Y {} doesn't work for anything but custom classes they were not joking around [05:36:17.0000] heh heh I got that point annevk, but i'm able to emulate it, with backgourdimage attr [05:36:43.0000] I just mentioned that with is="" thinge as well, this.src didn't work [05:37:11.0000] it does if you follow the rules [05:38:30.0000] :) [05:38:59.0000] /me passes annevk some iced tea ^_^ [05:40:45.0000] I'm sorry if I annoyed you annevk, I do remember yesterday's discussions [05:41:22.0000] hemanth: no worries, I was happy to try it out [05:42:37.0000] :) [05:46:23.0000] annevk, BTW http://jsfiddle.net/gnumanth/a08wghyk/ is what I have done, //cc'ing caitp as well. ( ugly? ) [05:49:21.0000] that's transpiled using 6to5 [06:01:02.0000] hemanth: that's actually pretty nice code for something transpiled; I'm guessing it's not using that many 6 features? [06:02:08.0000] darobin, it's using classes and template strings, hang on ill pass the code [06:02:17.0000] I heard ES6 classes are going to kill us all [06:02:34.0000] I heard MikeSmith was going to kill us all. :) [06:02:37.0000] MikeSmith, heh heh debatable [06:03:01.0000] /me is already dead ;) [06:03:01.0000] class-ical inheritance bugbear [06:03:31.0000] gsnedders: I kill with kindness [06:03:39.0000] I hope FP will lead kindly to light ;) [06:04:56.0000] dunno what FP is [06:05:03.0000] wonder if I want to know [06:05:18.0000] Functional Programming, MikeSmith you already know ;) [06:05:19.0000] MikeSmith: run before you get evangelised :) [06:05:24.0000] ruuuuuuuuun!!! [06:05:25.0000] ah [06:05:35.0000] heh heh [06:05:46.0000] The fantasy land [06:05:58.0000] /me takes MikeSmith on a ride to the fantasy land [06:06:15.0000] /me didn't know MikeSmith was allowed *out* of fantasy land [06:06:28.0000] /me gets off at the Rust rest top to take a potty break [06:06:53.0000] /me gets some sugar.js [06:07:06.0000] MikeSmith, joooooiiiiin uuuuuuuuuus [06:07:47.0000] I'm not smart enough for functional programming [06:07:57.0000] plus, I actually like side effects [06:08:00.0000] MikeSmith, we've got plenty imperative :) [06:08:32.0000] For once I wish I lived in a more religious — or at least more Christian — country so we could have today as a holiday. Or, I suppose, a less religious country that still had it as a holiday (e.g. Sweden) [06:08:33.0000] side-effects are horrible! [06:09:06.0000] side effects are what makes life interesting [06:09:39.0000] jgraham: where is this day a holiday? [06:09:46.0000] MikeSmith: Not here :( [06:09:51.0000] jgraham: what, you English people get your Yule Log today! [06:09:53.0000] Catholic holiday I guess [06:09:59.0000] Why did I capitalise that? [06:10:13.0000] Not unless the Scandinavians suddenly went catholic on me [06:10:28.0000] oh holy god [06:10:36.0000] Epiphany [06:10:50.0000] it's like, the very best religious holiday of all [06:11:25.0000] Wikipedia makes it sound like yule logs should be Twelfth Night and not today. [06:11:31.0000] Oh well. Anglicans! [06:12:37.0000] Is there any point in having once X extends HTMLImageElement {} actually works? [06:12:42.0000] gsnedders: I am having great difficulty not making faecal jokes [06:13:10.0000] annevk: I thought the point was so the declarative sematics were preserved [06:13:11.0000] annevk, readability ? [06:13:34.0000] *semantics [06:13:41.0000] jgraham: well it's not my problem that you English people bring them upon yourselves! [06:13:55.0000] Yule logs? [06:14:18.0000] Well that's the effect of a protein-rich diet… [06:14:31.0000] Ok, here I go -> http://h3manth.com/new/blog/2015/custom-elements-with-es6/ [06:14:38.0000] darobin, ^ [06:15:18.0000] /me runs and hides behind the tree before annevk et.al gets angry and throw stones at me, meow ^_^ [06:19:30.0000] /me is peeping from the side ^_ ^ [06:24:34.0000] heh =) [06:25:19.0000] /me slowly moves near caitp, any comments ;) ? [06:26:11.0000] i mean, I guess it works for the browsers that support custom elements [06:26:33.0000] :) [06:31:24.0000] hemanth: this.textContent = 'Today's date: ' <- probably doesn't work right :) [06:31:29.0000] *et al. [06:32:21.0000] darobin, what would you suggest? [06:32:35.0000] hemanth: escaping the ' in "Today's"? [06:33:13.0000] apparently in Safari, interfaces are not "function" but "object", too [06:33:17.0000] darobin, yikes, ha, that's done in the jsfiddle, not in the text, thanks editing [06:33:27.0000] I guessed as much :) [06:33:37.0000] not that it has custom elements, but it dies earlier [06:34:30.0000] but, even if it did, I think Object.create(HTMLImageElement.prototype) would break whenever you tried to access `src` [06:34:49.0000] hemanth: I'm happy to see that the transpiler does such a nice job; I mean you're not using *that* much but still [06:35:41.0000] yeah, other than shadowing `src`, that wouldn't work :( [06:36:52.0000] darobin, :) [06:36:58.0000] actually I'm wrong, apparently JSC tries to use the IDL-defined accessor before letting you shadow it [06:36:59.0000] caitp, that's the sad part... [06:37:02.0000] so it just doesn't work at all [06:37:21.0000] browsers! [06:37:25.0000] who writes this nonsense! [06:37:38.0000] gsnedders++ # <3 [06:37:50.0000] I was curios about how Domenic would make the custom-img element [06:38:01.0000] gsnedders, :D [06:38:05.0000] easy [06:38:09.0000] pff, I'm sure I could write a better browser than you guys! [06:38:10.0000] step 1) patch Chrome [06:38:18.0000] step 2) profit [06:39:23.0000] darobin: I think you missed the ??? [06:40:22.0000] gsnedders: the ??? ? [06:40:31.0000] this could get meta [06:40:33.0000] that's what step 2 is! [06:40:39.0000] darobin! [06:40:46.0000] the big where somehow you convince Google to accept your patch [06:40:49.0000] Ms2ger! My long-lost love! [06:41:00.0000] d'awww [06:41:03.0000] so romance [06:41:05.0000] gsnedders: you missed the fact where Domenic works for Google :) [06:41:32.0000] (that sell-out) [06:41:37.0000] darobin: still has to convince them to ship it! [06:41:41.0000] darobin, com'ere and review my tests! [06:42:02.0000] gsnedders: bah, we all know how shipping works at any company no matter how big; that can't be more than a few beers away [06:42:11.0000] darobin: I mean does he have a spec that's public *before* IO? [06:42:13.0000] well he's gonna get them to ship A.p.includes, so surely that's good enough [06:42:59.0000] Ms2ger: finish reviewing my test you dropped off of in the middle of the issue discussion :) [06:43:12.0000] Ms2ger: also, I need to look at your comments on the GH issues you closed [06:43:24.0000] darobin, link [06:43:27.0000] gsnedders: who said the spec can't be in C++ :) [06:43:34.0000] Ms2ger: https://critic.hoppipolla.co.uk/showcomment?chain=9822 [06:43:42.0000] darobin: I DID. [06:44:15.0000] darobin, your answer appears to be beside my point [06:44:55.0000] Ms2ger: then we're talking past one another — you're saying the tests worked in Gecko when you wrote them but they don't now, and they don't work elsewhere [06:45:20.0000] darobin, I said it worked, not that it passed [06:45:41.0000] okay, but elsewhere it's not that they fail, it's that they blow up on an unrelated problem [06:46:00.0000] Right, and that unrelated bug is still a bug [06:46:32.0000] I don't think it's a good idea to have tests test something unrelated [06:46:45.0000] we can get those tests to actually test something independently of other bugs [06:46:52.0000] which is the information they're after [06:47:09.0000] if you have issues with how data: and cross-origin interact, that's something for another test [06:47:52.0000] this is actually obscuring passes [06:52:53.0000] is there any way to per spec get the tentative pre-parse encoding and the final encoding from the parser to be different? [06:53:01.0000] without relying on how many bytes the pre-parse does? [06:56:52.0000] gah, submodules!!! [06:57:26.0000] huh, I guess you can't shadow accessor properties at all [06:57:32.0000] I did not know that [07:07:57.0000] darobin, all tests test unrelated things [07:08:10.0000] Ms2ger: yeah, I know the rhetoric [07:08:23.0000] Whatever [07:08:29.0000] but this one happens to test something unrelated that doesn't work in most places, and is trivially fixed [07:08:39.0000] I don't see what the problem is with fixing that [07:10:08.0000] Ms2ger: sorry but I really don't see what's making you grumpy about this, it's nice when tests work I think... [07:10:39.0000] unless you're grumpy because some browsers suck, which is, well, a good point [07:12:20.0000] I assumed that being generally grumpy about browsers was why we all did this [07:12:34.0000] It's GDD [07:12:43.0000] lol [07:18:19.0000] Ms2ger: oh, that's so very nice of you to reclose a PR I just reopened without leaving time for discussion [07:18:23.0000] very mature and all [07:18:29.0000] friendly, all that [07:18:48.0000] welcoming, etc — everything that makes working on web tech a great place! [07:19:00.0000] darobin, the test is correct, I told you so when I first closed the PR, and you reopened it without discussion [07:19:20.0000] I reopened it so that it would take the rebase into account [07:19:39.0000] Why rebase? It's wrong [07:19:49.0000] s/rebase/merge/ [07:20:55.0000] now, I could easily be lost in the test code but it looks to me like this isn't a case in which defaulting to the XHTML namespace applies [07:21:02.0000] and this fails in every single engine [07:21:58.0000] It is possible for the test to be right and still fail in every engine [07:22:08.0000] (note: I have no idea what test you are talking about) [07:22:25.0000] jgraham: that I know, and I could be reading either of the code or the spec wrong since both have some degree of indirection [07:24:45.0000] and indeed, I now see that I had read the code wrong [07:24:56.0000] /me looks for a spec bug [07:25:50.0000] darobin, it's created through createElement, so it's in the HTML namespace per spec [07:26:10.0000] Ms2ger: yes, I had misread indeed — which is why I'm looking for a spec bug [07:26:47.0000] jgraham: if that's the only reason there's no reason to support is="" in createElement() [07:28:50.0000] annevk: That makes sense I think [07:28:58.0000] Perhaps for serialization? But at some point this starts to break down... [07:29:11.0000] ah, right, I recall https://www.w3.org/Bugs/Public/show_bug.cgi?id=19431 [07:29:20.0000] it looks like implementers haven't weighed in much either way :) [07:54:49.0000] darobin: euhm [07:55:00.0000] darobin: it's not like implementations agree with each other today [07:55:09.0000] darobin: or did that change? [07:55:45.0000] annevk: well, it depends on what aspect of the whole Document mess. It looks like they all have XMLDocument.createElement() return an element with a null ns [07:56:02.0000]