| 02:05 | <MikeSmith> | Philip`: I guess we should change that validator "Comments seen before doctype. Internet Explorer will go into the quirks mode." message to be more specific. If I know for certain what IE versions that is true for, then I will add the info to that message. |
| 05:07 | <annevk> | gsnedders: thanks |
| 06:52 | <annevk> | I should have learned about git stash and git stash pop a little sooner |
| 09:41 | <annevk> | JakeA: I fixed some bikeshed stuff on Service Workers master |
| 09:42 | <annevk> | JakeA: I noticed there's still lots of issues left in the source |
| 09:42 | <annevk> | JakeA: e.g., each instance of `<a href="#dfn-` is a mistake of some sorts |
| 09:42 | <annevk> | JakeA: 702 matches |
| 09:45 | <annevk> | JakeA: oh, why is there both master and gh-pages? |
| 09:54 | <annevk> | Anyone know what the place is to file issues against Shepherd? |
| 10:01 | <JakeA> | annevk: ohh thanks. Agree there should only be one branch |
| 10:01 | <annevk> | JakeA: maybe you can fix that? I don't have admin access |
| 10:03 | <JakeA> | annevk: I'm on leave this week, but will post an issue |
| 10:03 | <annevk> | ah, I'm gonna be in the same boat soonish |
| 10:03 | <annevk> | enjoy |
| 10:05 | <JakeA> | Seems I don't have admin access anyway |
| 10:39 | <zcorpan> | MikeSmith: http://software.hixie.ch/utilities/js/live-dom-viewer/saved/4421 - IE6..9 BackCompat. IE10..11 and Edge CSS1Compat |
| 10:41 | <MikeSmith> | zcorpan: thanks |
| 10:41 | <MikeSmith> | what is the difference between BackCompat and CSS1Compat? |
| 10:42 | <annevk> | MikeSmith: latter is standards mode |
| 10:42 | <annevk> | MikeSmith: former is quirks |
| 10:42 | <annevk> | MikeSmith: well, latter is no-quirks really |
| 10:48 | <annevk> | MikeSmith: thanks for https://github.com/w3c/ServiceWorker/issues/967 |
| 10:48 | <annevk> | MikeSmith: generally if you see that missing anywhere and you can add it, it's a huge help |
| 10:56 | <MikeSmith> | annevk: OK |
| 10:57 | <annevk> | I'm not sure why, but I often end up on the GitHub repo trying to find my way to the corresponding spec |
| 11:52 | <MikeSmith> | I do too |
| 11:52 | <MikeSmith> | so I reckon a lot of other people do also |
| 13:31 | <annevk> | TabAtkins: don't really like you closing that issue; not sure you understood the whole thing |
| 14:25 | <SimonSapin> | https://en.wikipedia.org/wiki/ISO/IEC_8859-1 says "ISO-8859-1 is the IANA preferred name for this standard when supplemented with the C0 and C1 control codes", but I can’t find a source for that |
| 14:25 | <annevk> | SimonSapin: http://www.iana.org/assignments/character-sets/character-sets.xhtml? |
| 14:26 | <SimonSapin> | annevk: it lists ISO-8859-1 as ISO_8859-1:1987 and doesn’t seem to make a difference |
| 14:26 | <annevk> | SimonSapin: it has a preferred name field however that lists it |
| 14:27 | <annevk> | SimonSapin: oh, I guess IANA doesn't really make a judgment about C0/C1 necessarily, but also, don't care |
| 14:27 | <SimonSapin> | my understanding is that ISO/IEC 8859-1 leaves some bytes unassigned and ISO-8859-1 assigns controls to them |
| 14:27 | <annevk> | IANA is full of holes when it comes to encodings |
| 14:27 | <SimonSapin> | except I can’t find that define outside of wikipedia |
| 14:27 | <annevk> | I don't think that's correct |
| 14:28 | <SimonSapin> | annevk: ok, so taking a step back, is there a name for "convert 0x00...0xFF bytes to U+0000 U+00FF code points of the same value" ? |
| 14:29 | <annevk> | SimonSapin: I haven't been able to find a good one |
| 14:29 | <SimonSapin> | so wikipedia is wrong? |
| 14:29 | <annevk> | SimonSapin: somewhat annoyingly IDL describes it in terms of ISO-8859-1 but that doesn't make sense when the browser offers such a library |
| 14:30 | <annevk> | SimonSapin: I don't think there's a definitive answer one way or another there |
| 14:30 | <SimonSapin> | ok, thanks |
| 14:30 | <annevk> | I should file a bug against IDL |
| 14:30 | <SimonSapin> | (this is for documenting a Rust standard library conversion, not a web platform API) |
| 14:31 | <annevk> | Oh, IDL stopped doing that already |
| 14:40 | <annevk> | SimonSapin: if you think of a good name let me know |
| 14:41 | <annevk> | SimonSapin: WTF-latin1? |
| 14:41 | <SimonSapin> | hehe |
| 14:53 | <SimonSapin> | annevk: ah, found it. The IANA page links to https://tools.ietf.org/html/rfc1345 which despite what I first thought does define "&charset ISO_8859-1:1987" https://tools.ietf.org/html/rfc1345#page-63 |
| 15:01 | <SimonSapin> | aaaah |
| 15:01 | <SimonSapin> | Microsoft defines code page 1252 with some unassigned values |
| 15:15 | <annevk> | Yeah, correct |
| 15:16 | <annevk> | Their error handling doesn't use FFFD |
| 15:16 | <annevk> | So weird |
| 15:39 | <SimonSapin> | I ended up writing https://github.com/SimonSapin/rust/blob/f040208d533e1d6d9ee0e0408ee74e26e14d1284/src/libcore/char.rs#L183-L200 |
| 15:39 | <Tuipveus> | is there any multicast -implementations of websockets or is that even possible ? |
| 16:28 | <annevk> | wanderview: not sure what to do with your feedback, still think my PR helps |
| 16:44 | <wanderview> | annevk: fetch(newURL, evt.request) will blow up with your PR and current spec once browsers implement the Request window support, no? |
| 16:44 | <wanderview> | since evt.request.window will exist |
| 16:46 | <annevk> | wanderview: blow up how? |
| 16:46 | <wanderview> | annevk: step 10 in https://fetch.spec.whatwg.org/#dom-request |
| 16:47 | <annevk> | wanderview: Request has no window getter, so no? |
| 16:52 | <wanderview> | annevk: looking at those steps I have to wonder why RequestInit has a window attribute either... how can it ever be used? |
| 16:53 | <annevk> | wanderview: you can set it to null |
| 16:53 | <wanderview> | annevk: what I don't understand is why I can do new Request(oldRequest, {}) to override almost anything except the URL |
| 16:56 | <annevk> | wanderview: because that makes the API weird |
| 16:56 | <annevk> | wanderview: and it's not really overriding, it changes a whole lot of things |
| 16:58 | <annevk> | wanderview: I wouldn't reject a PR per se, but it seems rather icky |
| 21:18 | <rniwa> | Domenic: yt? |
| 21:18 | <Domenic> | rniwa: yeah |
| 21:18 | <rniwa> | Domenic: so it looks upgrading doesn't enqueue attributeChangedCallback for each attribute in upgrading case |
| 21:18 | <rniwa> | Domenic |
| 21:18 | <Domenic> | rniwa: it should, let me check... |
| 21:18 | <rniwa> | Domenic: and that's an issue |
| 21:19 | <rniwa> | Domenic: https://html.spec.whatwg.org/#dom-customelementregistry-define just says "enqueue a custom element upgrade reaction" |
| 21:19 | <rniwa> | Domenic: and that doesn't add any attribute changed callback |
| 21:19 | <rniwa> | Domenic: oh, https://html.spec.whatwg.org/#concept-upgrade-an-element itself does that |
| 21:19 | <Domenic> | rniwa: I think it happens automatically in some other place, same as connectedCallback, let me check |
| 21:20 | <Domenic> | Ah yeah there it is |
| 21:20 | <rniwa> | Domenic: that makes sense because other upgrade callback could modify attributes on other elements (although they shouldn't be doing that) |
| 21:21 | <Domenic> | Right, I remember trying to keep them together and then finding out that doesn't work for subtle reasons |
| 21:21 | <rniwa> | Domenic: so I know you guys are eager to ship this thing soon-ish |
| 21:21 | <rniwa> | Domenic: so I'm trying to finish up my impl by TPAC |
| 21:21 | <rniwa> | Domenic: it would be great if we can make sure we're on the same page by comparing our impl at TPAC |
| 21:22 | <Domenic> | rniwa: awesome!! Yeah I think it got approval to ship in Chrome 54. I know Dominic Cooney would love to meet up at TPAC and compare notes. |
| 21:22 | <rniwa> | Domenic: cool |
| 21:22 | <rniwa> | Domenic: is he coming to TPAC? |
| 21:22 | <Domenic> | Yeah |
| 21:23 | <Domenic> | Oh and we've been upstreaming the tests, I'm supposed to merge those to WPT, let me do that.. |
| 21:23 | <rniwa> | Domenic: that's great |
| 21:23 | <Domenic> | Hmm I guess there are unaddressed review comments (https://github.com/w3c/web-platform-tests/pull/3477), I'll ping |
| 21:25 | <rniwa> | Domenic: we have quite a few tests now: https://trac.webkit.org/browser/trunk/LayoutTests/fast/custom-elements/ |
| 21:25 | <rniwa> | Domenic: so I want to upstream them as well |
| 21:26 | <Domenic> | rniwa: Oh those look lovely, very nice |
| 21:32 | <Domenic> | rniwa: do we have an agenda for web components day yet somewhere? I know we'd like to talk about new-and-improved HTML imports/modules, and I think it would be interesting to talk through https://github.com/w3c/webcomponents/issues/468 and https://github.com/w3c/webcomponents/issues/176 and maaaaybe https://github.com/w3c/webcomponents/issues/113 |