[CSSWG] CSS Snapshot 2025 Breakout 2025-09-03 [css-2025]

 =========================================   These are the official CSSWG minutes.   Unless you're correcting the minutes,  please respond by starting a new thread    with an appropriate subject line. =========================================   CSS Snapshot 2025 Breakout --------------------------    - RESOLVED: Publish CSS 2025 as a Note once all edits from this               meeting are in (Issue #12715: Publish Snapshot as Group               Note)   - RESOLVED: Move cascade-5 to the Reliable CR section (Issue #12702:               Add CSS Cascade 5 to Reliable CRs)   - RESOLVED: Move css-color-adjust-1 to Reliable CR section (Issue               #12692: Add CSS Color Adjust 1 to Reliable CRs)   - RESOLVED: css-conditional-4 to Reliable CR section (Issue #12694:               Add CSS Conditional 4 to Reliable CRs)   - RESOLVED: Close, no change (Issue #12697: Add CSS Shapes 1 to               Reliable CRs)   - The group will come back to issue #12691 (Add CSS Will Change 1 to       Reliable CRs) after folks have had a chance to review the current       open issues list.   - RESOLVED: Close, no change (Issue #12712: Add Media Queries 5 to               Rough Interop)   - RESOLVED: Publish new WD of css-nesting (Issue #12704: Add CSS               Nesting 1 to Rough Interop)   - Prior to resolving on issue #12711 (Add CSSOM View 1 to Rough       Interop) republishing CSSOM View will be brought to the wider       group.   - CSS Overscroll needs an active editor before the group can decide       on issue #12701 (Add CSS Overscroll 1 to Rough Interop).  ===== FULL MEETING MINUTES ======  Agenda: https://lists.w3.org/Archives/Member/w3c-css-wg/2025JulSep/0180.html  Present:   Tab Atkins-Bittner   Keith Cirkel   Elika Etemad   Chris Lilley   Eric Meyer   Florian Rivoal   Alan Stearns   Sebastian Zartner  Scribe: emeyer  CSS Snapshot 2025 ==================  Publish Snapshot as Group Note ------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12715    astearns: Proposal is that once all edits agreed upon today are in,             we publish   <ChrisL> +1   (no objections)   <fantasai> We can always make changes and republish.   astearns: We are resolved    RESOLVED: Publish CSS 2025 as a Note once all edits from this meeting             are in  Add CSS Cascade 5 to Reliable CRs ---------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12702    SebastianZ: As a disclaimer, in all issues I summarized all the               details I could find regarding the specs   SebastianZ: so it should be relatively easy to resolve on adding or               not adding specs to the snapshot   ChrisL: I'd like to see us publish a CRD but that's separate    SebastianZ: There are 22 open issues but only about 5 of them are               substantial to the spec level   SebastianZ: the rest are editorial or could be deferred   <ChrisL> I think this should be in Reliable CR   fantasai: This makes sense; I do think we should keep track of the             specs that need to be republished and get them out with the             snapshot   astearns: Agreed, it would be good to keep things in sync, but not             that concerned about the snapshot Note saying it's a             reliable CR and updating the CR when we can   ChrisL: If we're reasonably up to date, I don't want to wait until           every single thing that could be republished has been    astearns: Proposed resolution is to move cascade-5 to reliable CR             section    RESOLVED: move cascade-5 to the Reliable CR section    <fantasai> Can we also get a resolution to republish the CRD   ChrisL: I'll nag people as required   <fantasai> sgtm, looking fwd to nagging   astearns: Elika, I'll follow up on publishing the CRD  Add CSS Color Adjust 1 to Reliable CRs --------------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12692    SebastianZ: we have good platform test coverage of this, especially               for Chrome/Edge/Firefox   SebastianZ: not too many open GitHub issues, proposing to move to               Reliable CR   astearns: Any concerns?    RESOLVED: move css-color-adjust-1 to Reliable CR section  Add CSS Conditional 4 to Reliable CRs -------------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12694    SebastianZ: this doesn't have a very high interop (82%) but only has               a few open issues on the spec   SebastianZ: question is, what's the threshold to moving to reliable CR    ChrisL: We should look at only this level of the spec and tests;           there are only 41 tests   fantasai: An original goal for Reliable CR is to distinguish “this is             almost a rec but we didn't get close enough”   astearns: It is a good question to ask what the interop threshold             should be — 80%, 90%, a particular number of tests?   fantasai: I don't think percentages work, because sometimes they're             mostly parsing tests and may not fail when the spec isn't             supported and don't say what's actually failing   <ChrisL> 35 × 100 ÷ 41 = 85.3658536585   fantasai: what you're looking for is, “Is this really close to being             a recommendation but there are a few remaining bugs or             well-understood minor issues?”   fantasai: Has it been getting enough attention that bugs are likely             to be found? Or is it a mostly-ignored spec?   fantasai: When you have a lot of passing tests AND you know the             feature is well-implemented across at least two             implementations, that's good   fantasai: Just looking at test numbers may not tell you what you need   fantasai: Layers is one where I'd expect good numbers to be a high             signal, but for layout you may have a lot of processing             tests but not many rendering tests   <fantasai> or things like media queries or ui stuff, where you're              more likely to have extensive processing tests than              behavior tests    florian: I think you may have mixed up categories a bit   florian: I think we tend to be fuzzy about which level is for what,            which isn't a great thing   fantasai: I think in 2.2, we said “spec is stable, implementations             aren't there yet”   fantasai: if I look at media queries, grid 1 & 2, those are stable             specs with hardly any changes and they haven't made it up             to the main thing because the implementations aren't there   fantasai: the main definition “these are practically Rec and we             haven't done the QA work to verify”   fantasai: “Reliable” is for “specs that are implemented with bugs or             holes in test suites but very stable specs”    <ChrisL> A CSS processor is considered to support a CSS selector if            it accepts that all aspects of that selector, recursively,            (rather than considering any of its syntax to be unknown or            invalid) and that selector doesn't contain unknown -webkit-            pseudo-elements.   ChrisL: This was all nice theoretical stuff, but I want to come back           to the spec   ChrisL: ever since we started the spec, this hasn't changed and it           won't change   ChrisL: that's why 41 tests is reasonable for that sentence   ChrisL: so I propose moving this forward to Reliable CR   <fantasai> wfm    RESOLVED: css-conditional-for to Reliable CR section  Add CSS Shapes 1 to Reliable CRs --------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12697    SebastianZ: Didn't dive too deep into test coverage   SebastianZ: CSS Shapes is in good shape!   SebastianZ: We have 86% interop in Chrome and Firefox, around 20 GH               issues and none on the spec   SebastianZ: Propose to add to Reliable CR    ChrisL: I'm less convinced about this, and I think it should be in           Rough Interop   florian: I suspect I agree    fantasai: We did just recently add 'shape()' and it hasn't even been             stabilized six months   fantasai: This kind of has two statuses because parts have been             supported forever, and parts barely at all   SebastianZ: In rough interop, or not at all?   SebastianZ: It's currently in “Fairly Stable”   florian: We do have more than limited experience for some parts   SebastianZ: It was added here a few snapshots ago   ChrisL: Before we made it unstable   SebastianZ: Exactly   SebastianZ: We could move Shape to the next level if we think it's               not stable enough   astearns: I'm inclined to leave it in Fairly Stable   astearns: yes, we've been adding new things and they don't have much             implementation, but they've been added to the draft with             tests   SebastianZ: Test coverage for shape() is actually quite good   SebastianZ: Firefox lacks support    astearns: I heard reservations about moving to Reliable CR; is             everyone okay with no change, or pushing it down to Rough             Interop?   florian: I support no change   <fantasai> But come back to it next year maybe!   <fantasai> +1    RESOLVED: close, no change  Add CSS Will Change 1 to Reliable CRs -------------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12691    SebastianZ: will-change has very high interop in main engines; 12               open issues, with 8 substantial, zero open issues in               the spec   SebastianZ: propose to move to Reliable CR    florian: I suspect that's possible but I'm concerned that it's            under-maintained because it hasn't been looked at in a while   florian: We haven't been actively thinking about the open issues   florian: Would like assurance from someone who's worked on this more            directly   astearns: I'm concerned about the number of issues, but not having             looked at them, I don't know if they're for the next level             or not   florian: Skimming, some seem like interop, but there might be de            facto interop   florian: We could probably get there, but not having looked at the            spec in three years, I don't know how much effort it would            take to find out   astearns: Shall we close no change?   florian: Can we punt and come back when Tab can give input?   florian: The ED is as old as the TR draft   <TabAtkins> Note: I cannot answer those questions right now, it's               been a few years since I've had to work on it.   astearns: Let's take no resolution and come back to it   SebastianZ: That works for me  Add Media Queries 5 to Rough Interop ------------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12712    ChrisL: We have a long list of editors and not sure how many are real   <TabAtkins> yeah that's just me and florian, then   florian: I'm no longer funded to work on this, so I'm real but not            active    SebastianZ: test coverage is at 81%; a lot of open issues, with about               50 substantial to the spec level; 10 within the spec   SebastianZ: some WPTs are still missing, but the important ones are               there   SebastianZ: not all features are supported by all browsers, but the               main ones are   SebastianZ: it could be the whole spec goes to Rough Interop, or we               could break things down by feature    ChrisL: This is a case where because the spec doesn't have inline WPT           annotations, it's hard to know what's tested/implemented and           which not   ChrisL: I suspect there are some that are very widely supported, and           some not   <TabAtkins> Agree, I think this would both benefit from inline wpt,               and be appropriate for cherry-picking   florian: I agree, and also some features are very hard to test   florian: All browsers are able to parse the thing and respond, but            you can't always tell if they do the right thing   florian: We probably need to do a slow walk through here to figure            out which things are supported and which are not   florian: Stuff like environment blending probably not well supported   SebastianZ: Sounds like I should take these back and propose specific               features to go to the safe section   florian: Agreed    florian: Want to also ask what we do about media-queries-4   florian: I don't know if we have figured out how to test if the right            behavior is happening   florian: We haven't really done anything with this in a long time and            that's not good   florian: I don't think the tests are good enough   SebastianZ: The next section for level 4?   florian: I think it's probably ready, but we haven't checked if            browsers are doing what they should   ChrisL: This is work later in the year, not for the snapshot   astearns: Luke Warlow is listed as an editor for level 5   astearns: Maybe Luke could work on tests for level 5, or have ideas             on how they could be tested   florian: We have to think it through as a group   astearns: Do we leave open, or close no change?   florian: I think close no change, with an action on SebastianZ to            triage issues   SebastianZ: Agreed    RESOLVED: close, no change  Add CSS Nesting 1 to Rough Interop ----------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12704    SebastianZ: Not many tests, but the tests are quite interoperable               (96%)   SebastianZ: Still a lot of issues; 23 on GitHub and 2 in the spec   SebastianZ: I think it should go to Rough Interop   astearns: We already resolved to do that, as you noted   ChrisL: Did we resolve to republish, because if so I can just get on           with it   SebastianZ: I don't think so   astearns: I don't think we ever took a resolution to republish nesting   ChrisL: Could we?   TabAtkins: I think we should   TabAtkins: any changes we made since last publication should be              reflected; we should republish    astearns: Has any work been done on auto-publishing?   TabAtkins: No, but I could try it on this    RESOLVED: publish new WD of css-nesting  Add CSSOM View 1 to Rough Interop ---------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12711    SebastianZ: Doesn't have high interop between all engines, but               between Chrome and Firefox is close to 90%   SebastianZ: 161 GitHub issues, 7 spec issues   SebastianZ: I thought it was worth adding to Rough Interop with one               note that the WD is very old and needs a new WD   SebastianZ: last edits are from last month   SebastianZ: Over the years there were many small and big changes, but               never a new WD   astearns: I suspect this is one where we don't want it in the             snapshot until we get a new draft   florian: This one isn't as much about the spec and more about the            features   florian: For Rough interop, it seems like enough   astearns: Fair enough   SebastianZ: I didn't get into spec details, so it's possible some               features are not covered   ChrisL: The Changes section has a lot of update, so I think it's           worth asking the editors if it's ready to be published   florian: Definitely want the editors' opinions on what to do    astearns: SebastianZ, you said there may be an issue about publishing             a new WD, have you found it?   <SebastianZ> https://github.com/w3c/csswg-drafts/issues/8559   SebastianZ: I opened one two years ago.   astearns: I'll get this on the agenda  Add CSS Overscroll 1 to Rough Interop -------------------------------------   github: https://github.com/w3c/csswg-drafts/issues/12701    SebastianZ: we're at 91% interop, though only a few tests   SebastianZ: 10 issues on GitHub, none on the spec   SebastianZ: Given overscroll has few features, the tests seem to               cover a lot   SebastianZ: The published WD is from 2019   astearns: And the editor is no longer a member of the WG   ChrisL: I see two editors, one from Facebook.   astearns: Who is also not a member   astearns: I think this is another where we need to figure out if we             can republish, and also add an editor who's part of the             group   astearns: only then can we consider whether it's viable for a snapshot 

Received on Wednesday, 3 September 2025 22:55:57 UTC