Agenda for Feb 9, 2023

[nairnandu]

Here is the proposed agenda for our meeting on Feb 9, 2023:

• Decision on Import Assertions #236
• Decision on Interop 2023: "Math Functions" could be labelled as "CSS Math Functions"? #269
• Decide on how and when - Interop 2023 process and launch retrospective #274
• Review Blocking metrics updates on test regressions #276
• Discuss Investigation Effort goals and tracking #277
• Review and discuss Push interop score to >90 by EOY #272

Previous meeting: #265

[nairnandu]

Attendees: @jgraham, @nairnandu, @boazsender, @jensimmons, @dandclark, @foolip, @Rnyman, @zcorpan, @nt1m, @tantek, @meyerweb, @jdatapple, @bkardell

• Change meeting cadence to 1 hour on a bi-weekly basis

  • [consensus] Yes!
  • [foolip] Move to the new cadence after the next instance (Feb 16). Lot of attendees want to attend the HTML triage meeting

• Decision on Import Assertions #236

  • [jgraham] spec has moved down to stage 2 from stage 3. Suggestion is to remove it for now and review once it is ready
  • [meyerweb] bkardell@ and I agree. We can put it back in once it gets to stage 3
  • [jensimmons] Agree to drop it as well, for now
  • [foolip] Agree to drop it as well, for now

• Decision on Interop 2023: "Math Functions" could be labelled as "CSS Math Functions"? #269

  • [meyeweb] Igalia supports this change
  • [foolip] Agree

• Decide on how and when - Interop 2023 process and launch retrospective #274

  • [foolip] drafted some initial thoughts on Interop 2023 process and launch retrospective #274
  • [gsnedders] best to give more time for everyone to assess and come back.
  • [jgraham] 3 weeks seems like a reasonable time to do that

• Review Blocking metrics updates on test regressions #276

  • [gsnedders] We did not have this problem in 2022. Graphs were not being updated and that’s how we noticed
  • [jgraham] A particular failure mode you could have where a test with 100 subtests and one of them failed. We don’t have a good way of dealing with different # of subtests for a test. We failed the score update to capture any errors. We need a mechanism to notice failures if we don’t block score update
  • [gsnedders] how much benefit do we get from blocking?
  • [jgraham] Failure mode is when we artificially inflate a score by not taking a failure into account
  • [foolip] we could fix that by having a zero score when there is a harness error.
  • [jgraham] if we don't block updating the score, we should have another flagging mechanism like creating a github issue
  • [foolip] If we make a list of all the files with a problem, we could raise a PR as part of the workflow
  • [jgraham] Sounds good
  • AI: foolip volunteered to look into updating the workflow

• Discuss Investigation Effort goals and tracking #277

  • [foolip] clear point of contacts need to be established
  • [gsnedders] discussed in the WPT core team. For mobile testing, @past will figure out who will be involved in that
  • [jgraham] happy to get the mobile testing conversation started. Happy to be the DRI for that investigation area.
  • [zcorpan] Interested in attending the accessibility investigation area
  • [boazsender] interested as well
  • AI: boazsender to follow-up with @chrishtr and @zcorpan

• Review and discuss Push interop score to >90 by EOY #272

  • [jgraham] Once features are implemented and if the score does not improve, we should figure out what we do. If later in the year, we see a concern with the velocity then we should discuss it as a team.
  • [bkardell] We all see this as a stated goal to get to 100%. But I do think, since the Interop line is new there needs to be some thought given to how we are progressing
  • [jensimmons] Is this a public goal that we are setting? There is a lot of process fatigue and there is a desire to do less from a process standpoint. How do we avoid finger-wagging at teams to do better? Maybe there is an opportunity to focus on a specific set of tests, if we see a problem. We could choose to have a more informal way of providing feedback.
  • [foolip] No, there was no intent to make it a public goal. I am concerned that if we don't have a stated intent, we won’t have the desired visibility
  • [jensimmons] timelines and engineering priorities might be different in each organization

• Test change proposal - media queries focus area

  • [zcorpan] Some tests were for media queries 5 and not for media queries 4. Issue raised Test change: remove Media Queries 5 tests from Interop 2023 Media Queries 4 focus area #280 Test change: remove Media Queries 5 tests from Interop 2023 Media Queries 4 focus area #280
  • [gsnedders] since we are early in the cycle, we can go ahead and update the tests
  • [jgraham] this would be more visible if we have the tooling in place that we discussed earlier
  • [nt1m] I can review the PR

• [rnyman] Cascade layers bug should be sorted soon

• [nt1m] css containment links need to be updated. Will file an issue

• [zcorpan] Will post a PR to update the readme to include focus areas carried forward from 2022