Agenda for Interop team meeting on May 28, 2026

[nairnandu]

Here is the proposed agenda for the Interop team meeting on May 28th, 2026:

• Discussion on potential alternatives to the BSF graph
• Review test change proposals
• Review open PRs
• Complete the Interop 2026 planning process retrospective

[meyerweb]

It seems I don’t have the current meeting information. If you could re-invite me at my Igalia email, I’d appreciate it. Thanks!

[nairnandu]

Attendees: @yulunwu2, @stubbornella, @nairnandu, @captainbrosset, @jgraham, @dandclark, @sarony

Notes

• Discussion on potential alternatives to the BSF graph
  • yulunwu2: PR to add the toggle is pending final approvals - Add UI toggle for /third_party tests in BSF chart wpt.fyi#4921.
  • yulunwu2: presented a proposal for a landing page for wpt.fyi
    • List of web-features and WPTs mapped to those web-features
    • Excludes features that have an objection from one of the engines
    • Feature-weighting is an open question
  • jgraham: pretty skeptical about this overall. Weighting between features is too hard and doing that on an ongoing basis is complicated. From the pov of a browser vendor, understanding the gaps can be useful. However, it's not clear how useful it is in general to understand if a feature is at 80% or 90%. Skeptical of that having a clear mapping to web compatibility issues. Tests are not particularly defined for that. If we turn the whole test suite into a metric, it becomes less of an engineering concern vs a marketing concern. We have had cases of web compatibility issues reported even when the tests are at 100%. This design has a number of red flags on incentives and maintainability
  • captainbrosset: welcome the prototyping using web-features. Not all tests have web-feature mapping - need to account for that. Removing negative standards positions - for web-feature IDs is here. What is the goal of the weighting?
  • yulunwu2: added weighting as something we could have for this graph. Concern was weighing each feature equally. After this feedback, maybe we can remove weighting
  • stubbornella: agree with some of the challenges. Whatever gets merged into wpt.fyi needs to be more intentional. Once we allow a metric on the front page, we should be intentional about how WPT changes reflect on that metric.
  • jgraham: even without weighting, we could get into challenges around which tests gets mapped to web-features. Metrics create a lot of unnecessary processes and need for human oversight. One of the reasons we object to the BSF metric is that it does not drive engineering priority because of the lack of developer signal. Any metric needs a clear indication of developer impact.
  • stubbornella: the landing page of wpt.fyi could be a summary of previous Interop efforts. Use it as a high-level view of what Interop has accomplished.
  • Next steps: provide feedback on the proposal
• Review test change proposals
  • Tests under shape() function are not related to shape() and are incomplete #1269 - pending PR
  • [scroll-animations-1] Change tests to reflect the change in timeline-scope #1281 - pending followup from WebKit
  • [css-shape-1] Support path/shape basic-shape-function inside shape-outside property #1284 - pending follow-up from Gecko
  • [css-scroll-snap-1] unrelated-gesture-scroll-during-snap.html relies on scroll events lacking intended direction #1288 - needs reviews
  • Exclude multicol-001.html and multicol-002.html from interop-2026 #1290 - needs reviews
• Review open PRs
  • [css-shapes] Implement polygon() round modifier wpt#59037 - follow-up for changing the PR
  • webrtc wpt: update insertable streams metadata tests wpt#59097 - merged
  • Update calculation for scroll timeline boundary alignment wpt#59520 - merged. Needs a follow-up test change proposal and a metadata PR
  • WebKit export of https://bugs.webkit.org/show_bug.cgi?id=308981 wpt#59532 - changes required on the PR
  • webrtc wpt: assert that stats trackIdentifier matches ontrack track.id wpt#59576 - merged
  • Add uneven longhand attributes parsing for view timeline shorthand wpt#59636 - merged
  • Handle customelementregistry attribute in <body> and expand global attribute WPT coverage wpt#59704 - merged
  • [Gecko Bug 2019378] Update WPT webrtc/RTCSctpTransport-maxChannels.html to spec. wpt#59875 - merged
  • Fix custom element upgrade in declarative shadow roots parsed via setHTMLUnsafe wpt#59888 - merged
  • Update webrtc-encoded-transform tests to latest SFrame APIs wpt#59917 - merged
  • Make navigate event intercept popstate tests behave correctly. wpt#59969 - merged. Needs a follow-up test change proposal and a metadata PR
  • Navigation.activation.from should be null when previous entry is init… wpt#60003 - needs WebKit review
  • Intercepted "reload" navigate events shouldn't scroll. wpt#60114 - needs further reviews
  • [Gecko Bug 2006626] Take fallback in attr() when <attr-type> is invalid. wpt#60195 - merged
  • [Gecko Bug 1997338] Clean up attr() cycle leftover from Bug 2009001. wpt#60205 - reviews required