Re: [csswg-drafts] [css-pseudo] clarify paired-cascading behavior (#6386)

So here’s a recap of this issue in its entirety, including the current state of my proposed edits (#6665). Note that only items 3 and 5 below would be normative changes, the rest were previously unspecified (or just needed rewording for clarity).

That patch also clarifies @Loirooriol’s [comment](https://github.com/w3c/csswg-drafts/issues/6386#issuecomment-887858248) about ‘inherit’, clarifies that UAs are still allowed to tweak highlight colors ([pending #6150](https://github.com/w3c/csswg-drafts/issues/6150)), and eliminates the overloading of “specified” for a complex concept that’s more or less unique to highlights (and ‘appearance’ impls).

1. **used value or computed value? no conflicts, spec doesn’t say.** spec doesn’t say, fantasai says either, frivoal says used, emilio says used, 6665 says used.
2. **what are highlight colors? no conflicts, spec doesn’t say.** fantasai says ‘color’ + ‘background-color’ _only_ (no ‘{stroke,fill}-color’), 6665 says the same.
3. **‘::selection’ only or all highlights? no conflicts, spec says ‘::selection’.** fantasai defers to impls, i say all, emilio says all, 6665 says all.
4. **which properties disable UA defaults? no conflicts, spec says ‘color’ + ‘background-color’ _only_.** fantasai says the same, 6665 clarifies it.
5. **which origins disable UA defaults? currently has conflicts, spec says author.** fantasai says author, andruud says author/animation/transition for consistency with ‘appearance’ impls, emilio says the same, 6665 says the same.
    * note that no one has expressed an opinion on the user origin. i wanted to play around with user stylesheets in a real impl before forming my own opinion, but at the time i couldn’t find one (things like devtools and greasemonkey/stylus actually inject styles in author origin). i think epiphany (webkit) or old opera (presto) might be able to actually, but either way, i don’t think we should bloat this issue any further.
6. should unset and/or revert disable UA defaults?
    a. **should ‘revert’ disable UA defaults, _if the result after reverting no longer has any values in the origins above_ (see q5)? no conflicts, spec doesn’t say.** fantasai says no, andruud says no, emilio says no, 6665 says no.
    b. **should ‘unset’ disable UA defaults? no conflicts yet, spec doesn’t say.** not sure fantasai had a position on this, andruud says yes, emilio says yes, 6665 says yes.
7. **should ‘unset’ mean ‘inherit’ for _all_ properties? no conflicts, spec doesn’t say.** fantasai says yes, 6665 says yes.

@fantasai could you please take another look at questions 5 and 6b, as well as my edits?

-- 
GitHub Notification of comment by delan
Please view or discuss this issue at https://github.com/w3c/csswg-drafts/issues/6386#issuecomment-957389344 using your GitHub account


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

Received on Tuesday, 2 November 2021 12:00:17 UTC