W3C

– DRAFT –
ARIA WG

27 March 2025

Attendees

Present
aardrian, ChrisCuellar, Daniel, filippo-zorzi, giacomo-petri, jcraig, katez, keithamus, Matt_King, melsumner, np-at, pkra, Rahim, sarah, Siri, smockle
Regrets
-
Chair
-
Scribe
aardrian

Meeting minutes

New Issue Triage

w3c/aria#2499

spectranaut_: We should agenda this.

Daniel: comfortable with that.

w3c/aria#2497

spectranaut_ This seems like something we should also agenda.

scott: Agreed

jamesn: Would a really simple PR help for that agenda item?

scott: Either works.

w3c/aria#2496

spectranaut_: Also from Scott, larger conversation.

Scott: Agenda, but maybe for later this meeting.

jamesn: I'll aenda for this meeting.

w3c/aria#2494

spectranaut_: Mel already made a PR, so moving on.

w3c/aria#2491

Specrtranaut_: There is already a PR for this, will discuss then.

rahim: Yes, that works, Several issues related.

Specrtranaut_: Assigning to Rahim.

w3c/aria#2487

spectranaut_: Already agendad.

spectranaut_: w3c/html-aam#579

spectranaut_: Is this something to talk about? We don't have a place for it yet. CSS-AAM backlog?

pkra: Sure?

pkra: We should at least document it.

jamesn: Dangerous to change behavior from what it is today, agree it should be documented.

Rahim: Assign it to me to investigate what browsers do today.

melsummer: Assign me too, please. I'd like to help document and reasons why.

jamesn: I would love to see a way to address this for keyboard users, since mouse users already often get a hover.

New PR Triage

w3c/aria#2498

spectranaut_: Editorial change

melsumner: Can we make sure it's editorial? I don't want to mess up tests.

spectranaut_: Yep.

w3c/aria#2495

spectranaut_: Another editorial.

pkra: Just needs sign-off.

<np-at> RE line-clamp There’s also text resize behavior which should be mentioned. Line clamp inherently increases truncation when scaling is increased… which is not obvious to many developers

jamesn: I'll review it.

w3c/aria#2492

spectranaut_: Another editorial, good to merge.

w3c/aria#2489

spectranaut_: Another editorial, good to merge?

pkra: Yes.

w3c/aria#2488

spectranaut_: Another editorial

pkra: DPUB, minor editorial.

daniel: I can review it.

w3c/aria#2486

spectranaut_: Another editorial for accName.

spectranaut_: Looks like follow-up to accName convo.

Rahim: Would like Brian, Scott, Melanie to look and approve.

scott: I think there is probably a bigger discussion about the bigger issue.

scott: Around how accName is calculated (DOM nodes, accTree, etc). There are other tangents to explore.

jamesn: accName doesn't apply in this issue, right?

scott: That's the concern, it applies for general steps but also identifying what wins in conflicts and how host language factors.

scott: Look at the specs with more specific rules since it can't be called out in a general algorithm.

jamesn: And those specs point to accName for specific steps, otherwise you stay in the spec.

spectranaut_: Is there an issue for this?

melsumner: Yes, issue 250 in accName.

w3c/accname#250

scott: I think this issue will resolve it as long as it is a more general note.

w3c/aria#2485

spectranaut_: Thank you for grammar fixes. Reviewers?

janefulton: I'll take it.

w3c/aria#2484

spectranaut_: This has a lot of reviwers.

Rahim: I recreated the work from the old PR. Thanks to keithamus for taking all the time explaining the differences.

Rahim: Switch this to draft.

<aardrian> s/Pete/Keith/

w3c/aria#2483

spectranaut_: This closes an issue from a few meetings ago.

pkra: Yeah, it would be good to have another set of eyes since giacomo-petri did all the work.

pkra: This would close another issue to add a term for accessibility ancestor.

melsumner: I'll review.

pkra: The trouble is a scoping statement to allow groups if they contain menu items.

<np-at> I'd be interested

pkra: You don't want to allow nested groups with menu items also without a menu somewhere.

giacomo-petri: You have to check both directions.

scott: Add me as a reviewer.

Matt_King: Add me as a reviewer.

<scott> actually, since this is becoming a bit of a party - i probably don't need to review anymore

WPT Open PRs

spectranaut_: Can we write tests for parent/child items and intervening groups?

aardrian: This is the current agenda item (WPT)

giacomo-petri: Yes, but there are some roles that complicate it (paraphrased).

Deep Dive planning

w3c/aria#2427

spectranaut_: Discussed last week, removing label. Can someone add a summary?

jamesn: I sent it to you.

spectranaut_: I'll check it and paste it.

Investigate changing ID-based WAI-ARIA value types due to referenceTarget

w3c/aria#2487

spectranaut_: This is editorial?

Rahim: This addition to HTML spec allows an ID association that crosses shadow root boundaries.

Rahim: The HTML spec is not formalizing the ID association.

Rahim: So with ARIA attributes, it is an element reference.

Rahim: With multiple IDREFs, its a "set of element references."

Rahim: So is this more than editorial since it thinks about the types of attributes an ARIA attribute can take.

Rahim: Thinking if ARIA should update accordingly now that HTML is add reference target.

Rahim: Perhaps it is just editorial.

spectranaut_: I think it's editorial, but since HTML spec changes haven't landed maybe we should un-agenda this for now. I'll add myself to help track it.

Rahim: Alice Boxhall mentioned there might be an impact on the AAMs as well.

spectranaut_: Yep, part of why I want to track this.

Matt_King: Does this have impacts on ARIA and it being host language agnostic?

Rahim: Good reminder ARIA needs to be host language independent, which is what spectranaut_ was referncing.

spectranaut_: Now there may be changes, like if aria-activedescendant would point to shadow root not versus the node deeper in the Shadow DOM.

aardrian: Next time -- aria-controls!

aardrian: For minuting!

Document interop of misspelled aria-labeledby and its conflict resolution

w3c/aria#2093

spectranaut_: There's been an update from keithamus

keithamus: Looking at deprecating the one L variant.

keithamus: Looking to see how it's being used in conjunction with misspelled and spelled variant.

keithamus: We'll continue to track deprecation and remove it in Chrome.

spectranaut_: No objections.

Stale PR Processing

spectranaut_: Does anyone have a PR they have open they want us to look at and remind us to review? That's the point of this agenda item.

spectranaut_: We have 86 open PRs.

Rahim: I have one.

w3c/aria#2199

spectranaut_: Oh, you responded to my questions, Thank you, I will review today.

Rahim: Another.

w3c/aria#2420

jamesn: I'll look at it. Right now.

Rahim: Yeah, ok, one more.

w3c/aria#2419

Rahim: Thank you melsumner and scott for helping.

spectranaut_: Has three other requested reviewers.

w3c/aria#2496

w3c/aria#2496

spectranaut_: CSS is coming up with stuff and we need to keep an eye on it?

scott: Yeah, two new significant features.

scott: One is CSS carousels. Which we discussed a bit at TPAC for how things might be exposed, how to map the elements created by CSS.

scott: But can't recall where it came back to group for discussion.

scott: We don't have a CSS-AAM.

scott: But there appears to be no rep from ARIA acting as a point person.

scott: The CSS spec is silent on how they should be mapped, it at all. But there is no CSS-AAM to do it.

scott: We need someone on point to bring consensus from ARIAWG. It can't be me; I don't represent everyone here or know all the sutff.

scott: With CSSWG moving away from visual styles to implementing functionality, we need to stay on this.

melsumner: Can we ask them to not?

spectranaut_: It would be great is someone in the group could do this. Any takers?

spectranaut_: Is there some way this could be fixed by a process that the CSSWG uses to get accessibility review at multiple stages?

chrisCuellar: I was going to suggest a process, but maybe an earlier review than what they have now?

spectranaut_: There's the wide review before the spec is...

Daniel: Ideally after the spec is first WD to CR, there should be a wide review process across 5 groups. The accessibility review is from APWG.

<Daniel> https://www.w3.org/2025/01/aria-charter#coordination

Daniel: The idea is that someone finds specific issues and can leave comments.

scott: I was aware, and they are in some of the TAG reviews. But I was hoping someone in this group could do that work before it gets to review step.

scott: Early engagement could catch some. There's no interoperable way for mappings to be made outside of Chrome.

scott: Chrome dev team released a blog post of oddly created patterns and now we're just seeing reaction versus having caught it sooner.

melsumner: I have an accessibility review process for products that work, that's like what Scott is describing.

melsumner: My process: when an idea first comes out, the PM must look at the accessibility policy to ensure they are in scope.

melsumner: Then work has to look through the requirements to ensure they mathc.

melsumner: Then two rounds of review before release, with agreements to fix.

melsumner: The CSSWG needs to have acknowledgment that everything they make is accessible.

melsumner: But if someone has the bandwidth to join, great.

<sarah> should we have a deep dive on the carousel stuff?

jamesn: We need something to happen. It seems to go to last minute before accessibility review, and then they write documents on how _not_ to use it.

jamesn: Which people don't read anyway.

jamesn: The only way that seems feasible is to have accessibility people embedded in the WG.

keithamus: I am on CSSWG, but didn't see this come up. I'll endeavor to bring the accessibility point of view to the WG.

jamesn: The AI conclusion is that you have volunteered, keithamus.

spectranaut_: I see sarah asked to do a deep dive on this; can you make an issue?

scott: While it's shipped in one browser, it's fixable because it's not in other browsers.

keithamus: No signals from other browsers for interest.

Matt_King: I like the idea of a deep dive.

Minutes manually created (not a transcript), formatted by scribe.perl version 244 (Thu Feb 27 01:23:09 2025 UTC).

Diagnostics

Succeeded: s/Pete/Keith

Failed: s/Pete/Keith/

Succeeded: s/Keith/keithamus

Succeeded: s/Then work has/melsumner: Then work has

Succeeded: s/spectranatu_/spectranaut_

Succeeded: s/spectranaut_ We/spectranaut_: We

Succeeded: s/Is this something to talk about?/spectranaut: Is this something to talk about?

Succeeded: s/spectranaut: Is this something to talk about?/spectranaut_: Is this something to talk about?

Succeeded: s/> spectranaut_: Has three/spectranaut_: Has three

Maybe present: jamesn, janefulton, melsummer, scott, Specrtranaut_, spectranaut_

All speakers: aardrian, chrisCuellar, Daniel, giacomo-petri, jamesn, janefulton, keithamus, Matt_King, melsummer, melsumner, pkra, rahim, scott, Specrtranaut_, spectranaut_

Active on IRC: aardrian, ChrisCuellar, Daniel, filippo-zorzi, giacomo-petri, jamesn, jcraig, katez, keithamus, Matt_King, melsumner, np-at, pkra, Rahim, sarah, scott, Siri, smockle, spectranaut_