Add "user attention" definition to top-level traversable.

[jan-ivar]

Fixes #6211 potentially. @annevk is this close to what you had in mind?

Is this editorial? I dunno. It adds no new API, only a concept sought by other specs listed in the OP of #6211. Tests and implementation bugs seem up to individual specs when/if they decide to rely on this new concept, if doing so ends up influencing said specs to change any observable behavior.

• At least two implementers are interested (and none opposed):

• General interest in e.g. Broken foreground detection w3c/mediacapture-main#752

---

/interaction.html ( diff )

[annevk]

Apologies for the long delay here. This looks pretty good to me.

Are you also planning on making various specifications adopt this once this lands? Or should we try to do that in a more coordinated fashion?

[jan-ivar]

Are you also planning on making various specifications adopt this once this lands?

I've opened w3c/mediacapture-main#934 to take care of that spec, which is my main motivation.

I haven't contributed to the other two specs mentioned in the issue OP, so volunteers would be great! If none, I can try to shoe something in, just don't want to step on any toes.

[annevk]

I like where this ended up. Two minor comments remain.

@domenic I think you should have a look as well.

[domenic]

Looks good with nits

[palak8669]

What does it mean by browser window losing focus?

[jan-ivar]

It means the browser window lost keyboard focus, which seems consistent with existing language in notes:

The note in this PR attempts to clarify the difference between user attention and system focus which both operate on that (the ability to "receive keyboard input channeled from the operating system").

[annevk]

Pushed some wrapping nits. Let's merge this.

[annevk]

Thanks a lot @jan-ivar for making this happen!