Re: [csswg-drafts] Add an `::interest-button` pseudo element to interest invokers (#12437)

Thanks for the ARIA-WG deep dive on this topic today! I heard a few things:

- There seemed to be fairly universal agreement on the point that the "interest button" (however it is produced) shouldn't be keyboard focusable.
- There was some debate about whether it should be completely absent from the accessibility tree, or whether it should still be there, exposed as a `role=button`. I didn't hear a clear conclusion here, though perhaps a slight majority preferred not to expose it.
- We should open an issue (on [ARIA](https://github.com/w3c/aria/issues)?) to properly connect `aria-actions` to interest invokers. Unless @smhigley wants to steal this action item, I'm happy to open an issue. (Hint: she will do a better job. 😄 )
- A point was made that this should be opt-out, rather than opt-in. I.e. the default state should be for the button to be showing, and the developer has to turn it off. I didn't hear much disagreement with this point.
- An issue was raised in the case that the UA provides a setting to adjust the time delays, e.g. "multiply my delays by 10". The concern is that this might be detectable by the site author, if they a) set a particular delay, b) wait for a user to hover that particular control, and c) time the interval between e.g. `mouseover` and `interest`. Mitigating this might be pretty tricky, without breaking the site. Perhaps any such setting should come with a warning that it might be detectable? Open to suggestions here - I am sympathetic to the concern, but I don't have a good mitigation idea.
- There was a discussion of the primary two API shapes: `::interest-button` pseudo element vs. `<span interestbutton>` HTML element. Most people seemed to prefer the pseudo element, though definitely not unanimously. Most folks from ARIA seemed to be roughly indifferent. The `<span interestbutton>` would not be able to be "on by default", and would have to come with a special content model (which seemed unique to the platform). Otherwise, the two should be roughly equivalent from the perspective of a11y.

Let me know if I missed anything, please!

-- 
GitHub Notification of comment by mfreed7
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/12437#issuecomment-4009607664 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Friday, 6 March 2026 05:09:40 UTC