| 07:38 | <annevk> | PSA: https://whatwg.org/stages is live (might have to hard refresh for better table styling) |
| 13:26 | <farre> | so, I now have tests, and I checked epiphany (which should be webkit, right?), and load events for cross origin navigation of iframes is done there as well |
| 13:45 | <annevk> | That makes sense. That's been the behavior in browsers for literally decades and I think it's also what HTML still requires. |
| 13:46 | <annevk> | I don't really have a good idea for how to solve the issue here apart from not dispatching anything at all. Or maybe only dispatching for the initial load, but that will likely break things. |
| 13:49 | <farre> | I'm looking at aligning gecko to do this as well. it still leaves the timing issue, but at least it will be the same |
| 13:58 | <annevk> | Gecko doesn't fire an event currently? |
| 14:18 | <farre> | not when updating #hash, no |
| 15:08 | <annevk> | Wait what, firing a load event for that case doesn't sound right. I guess I missed something. |
| 15:08 | <annevk> | Looking forward to the tests. :-) |
| 15:33 | <farre> | np, I'll ping you a phabricator link tomorrow when I've run try. I'm planning on letting the tests get upstreamed from the gecko repo if you don't object |
| 15:45 | <annevk> | That seems fine. But if they don't match the specification please make sure they have .tentative in the name. |