Exclude CSS specs with Warning: Not Ready

[cdoublev]

The Bikeshed metadata entry Warning: Not Ready is specified for CSS Color 6 and CSS Backgrounds 4. The result is a banner with the warning:

Not Ready For Implementation

This spec is not yet ready for implementation. It exists in this repository to record the ideas and promote discussion.

Before attempting to implement this spec, please contact the CSSWG at [email protected].

Do these specs intentionally match the Spec selection criteria despite this warning?

<wcag2()> is extracted without a value by reffy (cf. webref) because it is missing a proper production wcag2() = wcag2([<number> | [ aa | aaa ] && large? ]). If I file an issue on w3c/csswg-drafts, it will probably be ignored because the spec is not ready for implementation.

In CSS Color 6, it does not really matter because the spec does not replace definitions existing in previous versions of the spec. In CSS Backgrounds 4, some property definitions are replaced, though this spec does not seem to have any definition issue.

[cdoublev]

Hmm, I found #56 and #279. So the decision to crawl a spec that is not ready depends on subjective criteria. Sorry for missing them.

Fwiw, I noted the following specs:

• CSS Backgrounds 4 is not ready and there is 2 open issues for problematic definitions
  • [css-background-4] Missing round keyword in value field of border-limit csswg-drafts#7367
  • [css-background-4] Replace <position> by <bg-position> csswg-drafts#7376
• CSS Color 6 is not ready and I have not already reported the issue commented above
• CSS Conditional Values is not ready, its status is Unofficial Draft, and it has not been updated since 1 year
• CSS GPCM 4 is not ready and its re-definition of <content-list> (also defined in CSS GCPM 3) would deserve an issue
• CSS Table 3 is not ready but its definitions seem ok to me
• Non-element Selectors is not ready and has not been updated since 2 years

I cannot tell which features from these specs are currently experimented in browsers. I think some properties of CSS Backgrounds 4 are implemented in some of them.

[tidoust]

I don't have a better response than "it depends" here. Selection criteria remain fuzzy and new needs have emerged, notably to extract definitions and headings for cross-referencing purpose that means we tend to add specs earlier. On top of that, we've also added a few unofficial and discontinued drafts to the list, but we're capturing that in a new standing property that you may use to filter these out.

The CSS Working Group has unofficial drafts: those for which the status is "A Collection of Interesting Ideas" or "Unofficial Proposal Draft". These get captured through the new standing property. It also has drafts that are marked as not ready for implementation, some of which may already have been published as Working Drafts. We don't capture that nuance.

Looking at the specs you mentioned:

• css-backgrounds-4 and gcpm-4 were part of the initial list. I confess I don't remember why :) They are delta specs. How you approach delta specs and mix them with the previous level is essentially up to you, some of them are indeed not clean delta specs. We maintain an IDL patch for web-animations-2. We haven't tried to maintain CSS patches for delta specs for now.
• css-tables-3 was also part of the initial list. It's flagged as not ready for implementation but also published as a Working Draft.
• css-color-6 was added recently because Bikeshed needs it for cross-referencing purpose. It is not "ready" but should be maintained by the CSS WG. I also suspect that there is a typo in wcag since the prose suggests keyword should be wcag2. I'll report that and propose to include a proper function definition.
• css-conditional-values-1 was added recently with a "standing": "pending" property. Needed for cross-referencing purpose but you should probably ignore its CSS extract.
• selectors-nonelement-1 was added recently with a "standing": "discontinued" property. Needed for cross-referencing purpose but you should ignore its CSS extract.

[cdoublev]

Thank you for your answer, your suggestions, and for taking care of reporting the problem with wcag2.

I understand that a specification that is not ready for implementation may still be relevant for the web platform. Safari and Firefox seem to have implemented background-position-* defined in CSS Background 4 since a long time, and some may even be written at the same time as they are implemented.

Anyway, since new specs are not frequently added, I think I can manually monitor/select/exclude them. Thanks again.

[cdoublev]

Aside: your commit does not seem to have fixed wcag2(): it still has no value (argument definition).

[tidoust]

Aside: your commit does not seem to have fixed wcag2(): it still has no value (argument definition).

Right, see w3c/reffy#1146 (comment)