[css-conditional] Unclear what is and isn't a descriptor for at-rule()

[andruud]

#2463 resolved to add at-rule(@foo; descriptor: value), but I think there are several cases where it's not clear if something should be considered a descriptor or not. For example:

• Do @page (/margin) rules hold descriptors or properties? (They do not apply to elements, in which case they are arguably descriptors.)
• The accepted things within @position-try are described as both properties and descriptors by css-anchor-position-1.
• I'm sure there are other cases.

We should probably just avoid this problem by instead specifying at-rule(@foo; <declaration>), and evaluate to true whenever that declaration (whatever it is) would be valid within @foo.

[andruud]

Related: #5969

(But leaving this open, since it's about at-rule() specifically.)

[bramus]

In #6966 (comment) I suggested to allow authors to either test for an at-rule’s name or an actual full at-rule.

This would circumvent this descriptor issue (along with other things such as testing specific preludes and testing nested at-rules)

[kbabbitt]

Tagging for discussion at the upcoming F2F.

Proposal: Accept Anders's suggestion in the initial comment of specifying at-rule(@foo; <declaration>) rather than at-rule(@foo; <descriptor>: <value>)

Details in the explainer:
https://github.com/MicrosoftEdge/MSEdgeExplainers/blob/main/AtRuleFeatureDetection/explainer.md#detect-whether-a-given-declaration-is-supported-within-an-at-rule-block

[kbabbitt]

I think this issue is obsoleted by the resolution we took in #6966 (comment), which removed descriptor testing from the at-rule() function in favor of @supports-condition (#12622).