W3C

– DRAFT –
ARIA Editors

23 April 2025

Attendees

Present
Daniel, James, Melanie, pkra, Rahim
Regrets
-
Chair
-
Scribe
pkra

Meeting minutes

<melsumner> I have a question if there's a place to put it in the agenda

aria #1393 most instances of "States" and Properties" should be changed to "Attribute(s)"

<Daniel> Board view with agenda items

https://github.com/orgs/w3c/projects/188/views/1

w3c/aria#1393

pkra: saw this a while back, wanted to check what we want to do

<Daniel> Editorial: most instances of "States" and Properties" should be changed to "Attribute(s)"

rahim: I wanted to do something about this for my large IDL PR
… I would suggest to use content attributes and IDL attributes going forward.

jnurthen: no objection from me.
… most people won't know the difference.
… it's a lot of work and I'm not sure what the benefit is / whether it's worth it.

w3c/aria#1393

rahim: that clarification. It helps to align with HTML to distinguish.
… between these two changes I think it will help.
… I'm happy to volunteer.

pkra: yay. I also think it would be good to move from states/props to common language.

w3c/aria#2500 and validators feedback

<spectranaut_> dequelabs/axe-core#4754

spectranaut_: I had merged this because I thought this was ready.
… I filed an issue at axe but received a question from wilco about our process.
… they asked when it will be in REC
… but spec changes and implementations are now combined.
… for now, a meeting with wilco would help to explain changes in our process / moving towards living standard.
… work out what would help them implement author requirements.

jnurthen: their problem is that their tool only checks against 1 version. But moving away from versions, they have a harder time working out when they want to support changes to the spec.
… HTML validators have the same problem.

<Zakim> Daniel, you wanted to comment on better signals of what needs to be reviewed by tool vendors

Daniel: there's room for us to better communicate changes to them
… we have a good process with browsers. We could do better with checkers.
… we could have a better signal.

jamesn: could we ask the ACT group to review author requirements?

Daniel: yes that could work.

spectranaut_: I like that.

jamesn: just MUST changes or others as well?

Daniel: definitely MUST, also mappings are important to them/.

jamesn: mappings vary a lot

Daniel: right. But if accname requirements changes, then we probably should.

jamesn: right, I agree we need to be more careful. There might be downstream changes.

spectranaut_: we should set up a call with Wilco et al.

jamesn: yes. It will remain a hard problem for them.
… we should also find out a process for assigning them.

migrate changelog to main and 1.2 branch Daniel]

pkra: this is an ancient issue. Can we close this?
… I can't remember. Does anyone else?

spectranaut_: I also don't remember either.

pkra: I'll close it then.

revise approach to documenting deprecation Daniel]

pkra: This came up after a PR to remove deprecation notes from the tables.
… do we want to do something now?

spectranaut_: Not sure. We've discussed it so many times.

pkra: right. We can table it again.

jamesn: the HTML approach would be to have them at the bottom of the specs.
… that way they're not in the main content
… we might have trouble with aria.js extracting information.

melsumner: if it's deprecated, it seems fine to make it harder to find.

jamesn: maybe presentation is an example where we want to deprecate a role

rahim: directory is deprecated?

jamesn: right.
… how would we move that?

pkra: we could do that in aria.js?

jamesn: but didn't you want to deprecate it?

pkra: yes, but mostly the text-matching parts. If there's good data, then the process is ok with me.

spectranaut_: I'll take a look.

continue aria.js work

pkra: this was primarily a reminder.

spectranaut_: we need to decide when.
… maybe june?

pkra: sounds good!

Different id for assistive technology definitions in ARIA 1.2 and 1.3

Daniel: I've come across this problem a couple of times with the ID having changed.
https://github.com/orgs/w3c/projects/188/views/1?pane=issue&itemId=107642351

pkra: would adding data-lt back be enough?

Daniel: probably. Just want to make sure the change didn't have an intent.

https://github.com/w3c/aria/pull/1510/files#diff-0eb547304658805aad788d320f10bf1f292797b5e6d745a3bf617584da017051L257

https://github.com/w3c/aria/pull/1510/files#diff-0eb547304658805aad788d320f10bf1f292797b5e6d745a3bf617584da017051L275

<jamesn> [=assistive technologies=]

jamesn: people are supposed to use respec syntax

pkra: right. But shouldn't we restore the old ID?

Daniel: WCAG has a definition.

<jamesn> https://respec.org/xref/?term=assistive+technologies

jamesn: but they don't export it.

pkra: does WCAG have singular or plural?

Daniel: given the history of the PR, we probably should argue for this being relevant. But we still export it so we don't really need to.

jamesn: so we just add data-lt?

pkra: maybe. It's not clear to me if that's sufficient.

jamesn: in 1.1 it was Assistive Technology but not exported.

<melsumner> Question: do we have a way to alias?

pkra: can we move the draft to an issue and do a PR?

jamesn: we should make sure our specs use respec syntax
… where did this break anything?

Daniel: dpub.

jamesn: then we should change it.

Daniel: I'll take a look.

history of accname and html-aam?

melsumner: I had asked on slack. I was confused by this.

jamesn: if there's no ARIA involved, why should we overcomplicate it? If there's no aria-label/by, why look at accname?
… if it's pure HTML, people shouldn't have to look at accname.

spectranaut_: I suppose html-aam isn't as much about the mappings to the platform APIs.

melsumner: thanks. I've been giving talks about making sure you get an accname for various things
… it's a little tricky to bridge the gap. E.g., name from content.

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

Diagnostics

Succeeded: s/exportet/exported

Maybe present: jamesn, jnurthen, melsumner, spectranaut_

All speakers: Daniel, jamesn, jnurthen, melsumner, pkra, rahim, spectranaut_

Active on IRC: Daniel, jamesn, melsumner, pkra, spectranaut_