WHATWG on 2023-02-17

00:57	<sirisian>	Domenic: I noticed you're in a lot of the focus/dialog issues, so maybe you'll know this. https://jsfiddle.net/9pwdjr38/3/ This is consistent on both FF and Chrome. The behavior seems unusual to me though. I assume it's part of the spec though?
02:14	<Domenic>	Domenic: I've received this script `deferred` question you might have some historical insight on... is there a spec/compatibility reason to run all deferred scripts in the same task like we do? Splitting them to a-task-per-script can remove some blockage No real context, but I know some Chrome people (including Hiroshige who I think you know) who have poked at script execution timing a lot, and they might have better insight.
02:17	<Domenic>	Domenic: I noticed you're in a lot of the focus/dialog issues, so maybe you'll know this. https://jsfiddle.net/9pwdjr38/3/ This is consistent on both FF and Chrome. The behavior seems unusual to me though. I assume it's part of the spec though? This seems to be related to the changes made in https://github.com/whatwg/html/pull/6531 , which I personally thought was a bad idea but a lot of people preferred. Might be worth filing a bug to see if the people involved are willing to change again.
02:18	<Domenic>	I'm not sure shadow DOM is doing any work in your example, I think you would get the same behavior without it
02:19	<Domenic>	Maybe that change is just especially bad for non-modal dialogs.
02:57	<sirisian>	Good call. https://jsfiddle.net/sirisian/9f7t5pbm/ This is weirder than I thought. It is calling pointerdown.
02:58	<sirisian>	I'm opening an issue about this with this example. Will expand on the bullet points though.
03:03	<sirisian>	The main thing is it saves the last focus on show which is annoying when you want the focus to be able to change. (I haven't looked into popover much). I'll still open an issue as this is an unfortunate default behavior for non-modals. https://jsfiddle.net/sirisian/9f7t5pbm/1/
08:34	<annevk>	sirisian: make sure to at mention all the relevant people
08:35	<sirisian>	Yeah, was wondering about that. Saw domenic did it immediately. Now to wait patiently for a year.
08:37	<annevk>	If it's any consolation, I started pushing for URL Interop in 2012 or so. We might be getting somewhere this year.
11:17	<Ms2ger>	annevk: ugh, forgot to save the file before pushing like an idiot 🙈
11:30	<annevk>	Ms2ger: ooh I was wondering what happened there, heh
13:03	<Ms2ger>	(I blame TabAtkins - if my local bikeshed install wasn't being difficult, I'd have noticed)
15:45	<annevk>	Refactoring reflection is rather sleep-inducing. I might also just be tired
16:04	<annevk>	zcorpan: what's the difference between "limited to only non-negative numbers greater than zero with fallback" and "limited to only non-negative numbers greater than zero with fallback"?
16:04	<annevk>	zcorpan: they can both have a default value so this seems kinda pointless?
16:08	<annevk>	Never mind, it's the setter!
16:11	<jgraham>	(I choose to believe this would be less confusing if you hadn't written "what's the difference between (two identical strings)", which was presumably not the intent :)
16:12	<annevk>	Haha the former should not have "with fallback"
16:52	<Ms2ger>	Do we still have blame for the html spec somewhere?
17:02	<evilpie>	Do we still have blame for the html spec somewhere? Maybe searchfox could work?
17:03	<Ms2ger>	Good idea, thx
17:11	<annevk>	Yeah, that's why it's in searchfox