| 08:23 | <annevk> | Domenic: even if you use current it's still ambient, no? |
| 09:59 | <annevk> | Chrome implements out-of-process <iframe>s: all good. |
| 09:59 | <annevk> | Firefox (maybe mostly bz) takes a look: you know, the @htmlstandard is broken. |
| 10:10 | <Ms2ger> | Firefox implements ImageBitmap: all good |
| 10:11 | <Ms2ger> | I try to check how something works in Firefox: "Gah. Your tab just crashed." |
| 10:16 | <annevk> | Did you file a bug? Haven't seen many crashes myself, but also haven't been testing ImageBitmap much, mostly refactoring |
| 10:43 | <Ms2ger> | Yeah |
| 10:44 | <Ms2ger> | annevk, if you pass the same object twice in transferList when structured cloning, where does the spec handle that? |
| 10:47 | <annevk> | Ms2ger: good question, maybe the throwing early on https://html.spec.whatwg.org/multipage/structured-data.html#structuredserializewithtransfer needs to be moved down |
| 10:47 | <annevk> | early on in* |
| 10:48 | <annevk> | Ms2ger: would need to create a test where StructuredSerializeInternal throws something other than a DataCloneError to see if they actually happen later |
| 10:56 | <annevk> | Ms2ger: curiously Safari does not throw on buffer = new ArrayBuffer(10);postMessage(null, "*", [buffer, buffer]); |
| 11:07 | <annevk> | Ms2ger: are you going to file issues here? |
| 11:07 | <Ms2ger> | Yeah, in a minute |
| 11:09 | <annevk> | I can never remember how to write a getter in an object literal |
| 11:10 | <annevk> | So buffer = new MessageChannel().port1;postMessage({ get data() { throw 'hi'; } }, "*", [buffer, buffer]); suggests duplicates are caught earlier on though |
| 11:11 | <annevk> | So basically before we set memory[transferable] we should have "If memory[transferable] exists, throw a ..." |
| 11:18 | <annevk> | At least it's an easy fix... |
| 11:25 | <Ms2ger> | https://github.com/whatwg/html/issues/3507 |
| 13:30 | <annevk> | Ms2ger: look what you made me do: https://github.com/w3c/web-platform-tests/pull/9672 |
| 13:38 | <Ms2ger> | Thanks :) |
| 14:33 | <gsnedders> | "When the document's character encoding is ISO-8859-8, the following rules are additionally expected to apply, following those above:" (from HTML rendering) |
| 14:33 | <gsnedders> | what. even. |
| 14:33 | <gsnedders> | I mean my logic got lost a long time ago, but wow. |
| 14:46 | <annevk> | gsnedders: that is a lovely quirk |
| 14:46 | <gsnedders> | annevk: where does that even come from?! |
| 14:47 | <annevk> | gsnedders: I found it while trying to figure out why iso-8859-8-i could not be merged into iso-8859-8 |
| 14:47 | <annevk> | gsnedders: see https://encoding.spec.whatwg.org/#legacy-single-byte-encodings |
| 14:48 | <annevk> | gsnedders: I suspect this was just an early hack to make bidirectional text easier |
| 18:14 | <TimothyGu> | tobie: it seems like @@unscopables already has a null prototype on WebKit |
| 18:14 | <TimothyGu> | tobie: per https://github.com/heycam/webidl/issues/496#issuecomment-356994768 |
| 18:14 | <tobie> | TimothyGu: it crossed my mind right after I hit enter on that issue. |
| 18:15 | <tobie> | TimothyGu: #facepalm worthy |
| 18:16 | <tobie> | TimothyGu: I guess I can just close it. |
| 18:16 | <tobie> | TimothyGu: you're positive about this? |
| 18:19 | <TimothyGu> | err I'm not but seems like bz is |
| 18:20 | <TimothyGu> | tobie: ^^ |
| 18:20 | <TimothyGu> | would help if someone with safari could confirm |
| 18:29 | <tobie> | TimothyGu: indeed. seems fixed (just checked). Should have done that before doing anything else. |
| 18:29 | <tobie> | Apologies |
| 23:07 | <Krinkle> | annevk: Curious on your thoughts regarding https://github.com/douglascrockford/JSLint/issues/222 - I couldn't quite figure out whether or not this is correct behaviour on the new URL spec. |
| 23:07 | <Krinkle> | TL;DR: Chrome expands href="http:foo" relative to https://example.org/bar as http://example.org/foo, note the protocol switch |