Upcoming HTML standard issue triage meeting on 5/5/2022

[past]

Yesteray we held our monthly triage call (#7677) and I will post the meeting notes there in a bit. The next one is scheduled for May 5, 9 am PDT. People interested in attending the next call please respond here or reach out privately to me or the spec editors. We will be tagging issues for the next call again using the agenda+ label and we would like to invite anyone that can contribute to said issues to join us.

[Kaleidea]

Please sign me up, thanks.

[nt1m]

Please sign me up, mostly to discuss #7796 and #7808.

[past]

Done and done.

[past]

Thanks everyone for attending! Here are the notes from this meeting (the next one is at #7919):

Agenda

1. Review past action items
   1. Mason will finish the Chromium testing for XML MIME types used for documents are not interoperable and we'll ask for WebKit feedback.
      1. Joey added the results of his testing
      2. Next steps: either align with Chromium/WebKit for shortest path to interop, or Anne to make a proposal and we'll evaluate
   2. Domenic to take a look next at Valid/Invalid characters in document.createElement() and Mason/Anne will ask for Chromium/Firefox security opinions.
      1. Some comments added.
      2. Some confusion about Daniel's comments, but generally let's go with Anne's proposal. Next step: Domenic to make a spec.
   3. Domenic to comment and Emilio to test cross-browser behavior on blur and change events fire for elements removed from DOM.
      1. No progress.
   4. Anne is waiting for input from Elika & Tab on the PR for directionality in the shadow DOM.
      1. Currently blocked on Eric to finish the PR.
      2. Anne wants a second reviewer: Olli.
   5. New feature proposal: Popup API: Brian to add Emilio to the OpenUI meeting. Olli to get feedback from Mozilla a11y people on their XUL <popup> experiences. Mason to summarize the open OpenUI issues in the HTML issue.
      1. Mason summarized.
   6. <object> without "data" attribute probably needs to consider params to figure out its URL: Mason will try to deprecate in Chromium.
      1. Domenic merged a spec PR.
2. Carryovers from last time
   1. [Domenic] <object> element can tell whether child document loaded or not
      1. Usage of the fallback is high enough that this will be annoying. Unlikely to work on this in the near future.
   2. [Domenic] Exposing back/forward cache blocking reasons to sites
      1. Olli to raise an issue on the repo explaining that we should probably also talk about eviction reasons, not just prevention-from-going-into-bfcache reasons
   3. [Domenic] Recommend iframe/img over embed/object for some content types
      1. Domenic to write up a PR
3. New topics
   1. [Emilio] Inert subtree section is missing requirement that it should NOT be exposed to accessibility
      1. Consensus to write down somewhere (Core AAM, HTML?) that inert doesn't go in the accessibility tree.
      2. Also authoring guidance so authors understand the consequences.
   2. [Simon] Remove UA style for h1-h6 in section (et. al.) and hgroup
   3. [Anne] reportError(), undefined, and ErrorEvent
   4. [Emilio] Define OffscreenCanvasRenderingContext2D.font setter in detail
   5. [Fernando] Canvas2D Filters
      1. Preference from Mozilla and Apple for not allowing the new CanvasFilter class to be used with ctx.filter, instead that needs to be only usable with a new layer-based API.
      2. Please change the tests to .tentative
   6. [Emilio] Proposal: extend FocusOptions to allow specifying :focus-visible behavior
   7. [Anne] Rendering of "bitmaprenderer"-backed canvas
   8. [Emilio] inert on an ancestor of modal dialog
      1. Consensus to align on Gecko/WebKit behavior. Emilio to send PR, Tim will help review.
      2. Should it also escape pointer-events: none, other things? We need to investigate.
      3. We can use the :modal pseudo-class which CSSWG agreed to add yesterday. Everyone loves UA style rules. This will clean up the inert and <dialog> specs.

Action Items

1. XML MIME types used for documents are not interoperable: @annevk to figure out whether Firefox wants to align with Chromium/WebKit, or whether he has a proposed alternative processing model.
2. Valid/Invalid characters in document.createElement(): @domenic to write a spec PR
3. PR for directionality in the shadow DOM: @meyerweb to finish PR. @smaug---- to review PR in addition to @annevk.
4. Exposing back/forward cache blocking reasons to sites: @smaug---- to comment on the repo about eviction, not just preventing-going-into-bfcache.
5. Recommend iframe/img over embed/object for some content types: @domenic to work on a PR to documentation-deprecate these elements as much as possible.
6. Inert subtree section is missing requirement that it should NOT be exposed to accessibility: Scott O'Hara seems to be working on authoring guidance. We also need to update Core AAM (?) to state the actual accessibility mapping; unclear who will do that.
7. Canvas2D Filters: @fserb and his team to work on layer proposal, and rename existing tests to .tentative. @domenic and @annevk to discuss whether to revert the PR entirely or keep some of the editorial cleanups or…
8. inert on an ancestor of modal dialog: @emilio to work on spec PR to align on Gecko/WebKit behavior. This might involve using the new :modal pseudo-class and UA style rules, for both <dialog> and inert="".

[scottaohara]

re: 6 - i've created an initial pull request for html aam for inert mappings. this will be brought up in ARIA's weekly triage of new PRs and we will get feedback/any necessary updates on it.

I will do my best to get some HTML spec text updates regarding some clarifications / author guidance, as soon as possible.