WHATWG on 2025-10-09

15:47	<annevk>	Yoav Weiss nicolo-ribaudo generally with JS module imports the URL is an observable, right? It's a thing you de-duplicate with. That seems a bit annoying with the `data:` URL approach for CSS module scripts. Although maybe the de-duplication is not strictly observable? Hmm.
16:06	<Yoav Weiss>	tbh, I haven't followed the CSS module work, so no opinions atm..
17:03	<annevk>	Noam Rosenthal: I'm hoping you'll post the minutes
17:20	<Noam Rosenthal>	https://github.com/whatwg/html/issues/11747#issuecomment-3386836764
17:54	<nicolo-ribaudo>	Yoav Weiss nicolo-ribaudo generally with JS module imports the URL is an observable, right? It's a thing you de-duplicate with. That seems a bit annoying with the `data:` URL approach for CSS module scripts. Although maybe the de-duplication is not strictly observable? Hmm. Not sure exactly I understand where this is coming from, but deduplication is observable by checking the identity of the style sheet object
17:55	<nicolo-ribaudo>	But data URLs behave in the same way for JS modules, not just css
19:14	<zaggy1024>	Hi, I'm working on (re)implementing HTMLMediaElement for the Ladybird browser, and I've noticed a few issues I'd like to sanity check before writing issues/PRs for the spec. I hope I'm in the right place! The main thing I wanted to bring up is that the `ended` attribute and the condition under which the internal play steps seek back to the earliest position don't seem to agree with existing implementations. Specifically, when the `looped` attribute is unset, playback reaches the end, then the `looped` attribute is subsequently set, the spec seems to dictate that the `ended` attribute is then `false`, and the play steps do not seek to the earliest position. Chrome and Firefox both keep the `ended` attribute set to `true`, and seek to the start. I wonder if the `ended` attribute should be a variable set to `true` when reaching the end of playback, and then set to `false` when beginning a seek, and then the play steps should rely on the value of that attribute to dictate whether to seek to the earliest position. Alternatively perhaps, the `has ended playback` definition could specify that the `looped` attribute must only be unset when the playback first reaches the end of the resource, which I want to say would have the same effect. Let me know if there's another spec interpretation for any of this, not sure if I'm missing anything.