| 10:25 | <foolip> | annevk: in https://github.com/whatwg/html/pull/12389 was you intention that parsing <video muted> should fire the volumechange event? It hinges on "Whenever either of the values that would be returned by the volume and muted IDL attributes change" and whether we consider element creation + attribute setting by the parser an atomic thing or not. |
| 10:54 | <annevk> | foolip: oh my, no. |
| 11:00 | <annevk> | foolip: I'll try to figure out a fix. I think we basically don't want muted attribute changes to influence volumechange, but it would be nice if we could move the definition of that event away from relying on observing public API in this weird manner as well. |
| 11:14 | <foolip> | annevk: can we distinguish between the parser setting an attribute and anything else? that's the signal I've used when implementing this to avoid the problem. |
| 11:17 | <foolip> | annevk: same question when cloning, it doesn't seem sensible for that to fire the event either |
| 11:21 | <annevk> | foolip: no, that was the whole problem. |
| 11:23 | <foolip> | annevk: ok, I'll a test for that too then |
| 11:24 | <annevk> | foolip: I think I updated those tests already? |
| 11:24 | <foolip> | annevk: no test fails when I fail to fire a volumechange event when cloning |
| 11:25 | <annevk> | foolip: hmm, I wonder if we should just not make the muted content attribute fire events |
| 11:27 | <foolip> | annevk: that would definitely be less complicated. as it is the volumechange event should also fire when playbackRate changes, but the saving grace is that Chromium doesn't allow negative playbackRate so it can't happen. |
| 11:28 | <annevk> | I have a patch now for HTML to refactor things. It currently doesn't fire volumechange when the muted attribute changes, but does for the other cases. That essentially restores the status quo, modulo that the muted attribute can technically change the volume post-creation now. |
| 11:29 | <annevk> | But we also didn't fire volumechange when playing backwards (which this patch also restores) so that seems acceptable. |
| 11:29 | <foolip> | annevk: seems about right, happy to review (and throw away the tests I so carefully wrote) |
| 11:34 | <annevk> | foolip: inverse of those tests might still be useful? Not sure. Here's the PR: https://github.com/whatwg/html/pull/12428 |
| 11:36 | <foolip> | annevk: will take a look after lunch and write tests to match |