- From: Tab Atkins Jr. via GitHub <noreply@w3.org>
- Date: Wed, 04 Feb 2026 22:19:59 +0000
- To: public-css-archive@w3.org
> The effect of adding an identifier seems inconsistent to me: with `property-shared` or `element-shared` it just subdivides sharedness without affecting the behavior, while otherwise it adds sharedness.
This is what I was originally concerned with, but I do agree with Bramus that it flows naturally in usage. `random(--foo, ...)`, no other keywords, sure *looks* like it should match every other `random(--foo, ...)` on the page, as that's related to the sort of behavior that giving something a name usually has in CSS. That just means that, if you supply *just* an ident, it *implies* both of the sharing keywords.
The alternative, being strict about keyword usage, would mean that using just an ident would *just* set the ident portion of the key; you'd go from `(null, "width 0", element ID)` to `("--foo", "width 0", element ID)`. The behavior difference between those two caching keys is *almost nil*; in practice the only use-case for it, I think, would be for script to force a random value on a given element to a *new* random value, by changing the ident. (But the script could just... generate a random value itself at that point.) And then the reasonably common case of wanting a "global" random value would be pretty verbose to spell.
--
GitHub Notification of comment by tabatkins
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/13132#issuecomment-3850034688 using your GitHub account
--
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config
Received on Wednesday, 4 February 2026 22:20:00 UTC