Agenda for Jul 6, 2023

[nairnandu]

• Finish the Interop Team Charter #275 - [Review required]Create the Interop Team Charter #102 (comment)
• Interop 2024 brainstorming #331
  • Follow-up on AIs from Agenda for Jun 15, 2023 #353
• Test change proposals
  • Remove wpt/webgl/idlharness.any.html & idlharness.any.worker.html from Interop list #363
  • Remove the html/canvas/offscreen/text/2d.text.measure.advances.* tests #365

[nairnandu]

Brief meeting notes:

Attendees: @bkardell, @nairnandu, @jgraham, @dandclark , @nt1m , @jensimmons , @meyerweb , @slightlyoff, @gsnedders

• Finish the Interop Team Charter #275 - [Review required]Create the Interop Team Charter #102 (comment)
  • bkardell: Governance of browser specific failures? Is it included in the charter? I do think we should take up browser specific failures and figure out what to do with it
  • slightlyoff: we'd still like clarity about participant confidentiality and decision mode. I haven't been participating a lot here, but this doesn't look sufficient.
  • dandclark: can we clarify what would be the case with regards to confidentiality?
  • jgraham: for browser specific failures - the open question is how it applies to the charter. We need to have consensus on the usefulness of that metric.
• Interop 2024 brainstorming #331
  • Follow-up on AIs from Agenda for Jun 15, 2023 #353
  • jensimmons: We dont want to create a “formula” from signals. There could be features that developers care about, but is not reflected in the surveys as an example.
  • jgraham: makes sense to share the idea that we will review data points that are public
  • jensimmons: that is why we have 6 organizations and we don't need to have consensus on everything. It is good to have a mission statement that explains why we are asking for some information.
  • jensimmons: we have been thinking about the process for next year and will share the proposal in the next couple of weeks. One question is on what we do with Interop 2023 and deciding what to carryover - do we have any high level thoughts on that?
  • jgraham: not carrying over everything and having the new proposals be compared with the carryover ones is what I proposed last year. If there are significant additions of tests, we
  • slightlyoff: presumably if something was important enough for last year and didn't hit high interop, that suggests the evidence of inclusion from the previous year should still apply?
  • nt1m: personally in favor of carrying over and retiring based on consensus. The last 5% might be tricky due to a variety of issues.
  • jensimmons: the downside of not carrying over is that we might leave behind areas that are not fully interoperable. This year we know that 25 focus areas is too much. If we say next year we want to have lesser focus areas, would we have enough “slots” available?
  • jgraham: if we do make focus areas smaller, we need to make sure we are taking on the most important things. There is an incentive to finish the year strong.
  • jensimmons: only as an example - let see masking which has 62% interop. For some reason if we don’t make progress - do we make it compete with a new proposal/focus-area?
  • jgraham: we should have the ability to say that we made a mistake in prioritization or that we have new data that changes our opinion. That we picked an area before should have some weightage, but it should not be the only one.
  • bkardell: agree that we should have the facility to say that we were wrong. Maybe we can review the areas one-by-one and decide.
  • nt1m: consensus based retirement of focus-areas can be a good middle ground
  • jensimmons: Don’t think any one of us is suggesting that an automatic carry over be done based on a score. An assessment of where we are at and deciding on the carryover process should be done at a later point in the calendar year.
  • nairnandu: High-level timeline. Do we expect it to be the same as last year?
    • jensimmons: Should be part of the process discussion
    • jgraham: should roughly be the same timeline as last year, but needs discussion
• Test change proposals
  • Remove wpt/webgl/idlharness.any.html & idlharness.any.worker.html from Interop list #363 - need input from WebKit
  • Remove the html/canvas/offscreen/text/2d.text.measure.advances.* tests #365 - pending input from Chromium
  • jgraham: Since we are in Q3, we should be more conservative on test change proposals that results in a scope increase