W3C

– DRAFT –
ARIA and Assistive Technologies Community Group

25 March 2026

Attendees

Present
cyns, dean, Joe_Humbert, jugglinmike, lola
Regrets
-
Chair
-
Scribe
jugglinmike

Meeting minutes

Matt_King: Review agenda and next meeting dates

https://github.com/w3c/aria-at/wiki/March-25%2C-2026-Agenda

Matt_King: Requests for changes to agenda?

Matt_King: Hearing none, we'll stick with the agenda as scheduled

Matt_King: Next CG meeting: Wednesday April 8

<lola> That's awesome!

Current interop reporting status

Matt_King: Our primary goal is to surface up-to-date, usable interop data in (at this point in time) JAWS, NVDA, and VoiceOver

Matt_King: I've been working on getting one more plan into the third phase that we call "candidate review" where we will discuss it with implementers

Matt_King: I ran into a technical problem in our infrastructure, but I'm able to work around it for now

Matt_King: This is for the "switch" test plan

Matt_King: Next, we'll be talking about the "quantity spin button" test plan

Matt_King: Hopefully, I'll be figuring out how to get two more test plans ready for "draft review" before the next meeting in two week's time

Execution of quantity spin button test plan

Matt_King: dean and mmoss are working on this currently, Joe_Humbert has worked on it in the past

Matt_King: The VoiceOver testing is almost complete. There are eight conflicting results at the moment. I'd like to review those today

Matt_King: dean, with VoiceOver largely complete (and because NVDA bot is currently not running), would you be able to run the test plan in NVDA?

dean: I just happened to sign up to do that work shortly before this call

dean: I do not have JAWS on my system, however

Matt_King: Great!

Matt_King: As for the conflicts for VoiceOver

https://aria-at.w3.org/test-queue/557/conflicts

Matt_King: (For those unfamiliar: we have two people run the same test plan as a form of quality control. The ARIA-AT application alerts us when the two testers' results do not match.)

Matt_King: In test 1, the testers had the same output, but mmoss also reported "excess verbosity"

dean: I agree with that judgment, so I can include that in my results, as well

Matt_King: In test 3, I'm questioning mmoss's results. When I review his results, I'm wondering if he pressed "ctrl+option+i" instead of "ctrl+option+l"

dean: He did! This is a problem I've found with sans-serif fonts in user interfaces. The value of single characters can be difficult to interpret in situations like this

Matt_King: And we intentionally use lower-case characters for this in order to avoid suggesting that the "shift" key should be involved

dean: This is rendered as a lower-case "L", but some readers may interpret that as an upper-case "I"

Matt_King: This makes me wonder whether or not some other kind of solution is necessary...

Matt_King: Do we want some explicit note for this? Perhaps something system-generated?

Matt_King: I don't really know if we should spend too much time on this right now

dean: For the time being, I can communicate directly with mmoss about this, and mmoss can update his results

Matt_King: That should address conflicts in test 3 and test 4, at least

Matt_King: As for test 5

dean: In test 5, I reported the same thing that the bot reported

Matt_King: Your output does not include the value

Matt_King: This test is with "quick nav" off. I wonder if that is the problem here. If you had "quick nav" on, I'm not sure where the reading cursor would have gone

Matt_King: I think the bot was sometimes also having trouble ensuring that "quick nav" was in the correct state. That's an outstanding bit of fragility in the VoiceOver Bot

Matt_King: So I would suggest double-checking the state of "quick nav"

Matt_King: Test 6 concerns "down arrow"

Matt_King: The expected value is "2"

Matt_King: Maybe mmoss is the one who had the "quick nav" problem here

Matt_King: I think we may want mmoss to repeat both test 5 and test 6, as well

dean: I'm using MacOS 15.7.3

dean: And Safari 26.2

Matt_King: VoiceOver is more likely the thing that's going to make the difference here

Matt_King: Let's ask mmoss to re-run these tests and verify that he's using the correct settings

dean: I'll include this in the e-mail I send to mmoss

Blocking issues

Matt_King: We have two issues at the moment; they're quite different from one another

Issue 1363: Netlify preview not working for pull requests

github: w3c/aria-at#1363

Matt_King: I was trying to make some changes to test plans, and we have a Netlify-powered preview set up in the ARIA-AT repo which displays the built test plans. It's somewhat complicated.

Matt_King: Right now, the Netlify preview is giving a "site not found" issue

jugglinmike: howard-e took a look at this offline and reported, "It builds locally so my best bet would have been on the Netlify preview timing out since the branch is stale. So it could require a rebuild/new commit pushed but I did check a more recent PR (though closed in Feb) and it also 404’ed which was unexpected."

jugglinmike: howard-e goes on to say, "aria-at.netlify.app is still available though. So also guessing something may have changed in the Netlify config around PR Previews. It could also be the migration to W3C that was discussed when I was more involved happened but something was misconfigured (or maybe the simple rebuild/new commit pushed after the ownership change, if it happened is still needed in that case for a stale branch)

."

Matt_King: That's helpful!

Matt_King: Was the Netlify account was previously running under a Bocoup-owned org?

jugglinmike: I can't recall for sure. I'll have to look into that offline

lola: Who in this group has access to the Netlify account?

Matt_King: Daniel

jugglinmike: The configuration within the aria-at project (both the file checked into the repository and the settings of the GitHub project itself) appear to be agnostic of ownership

jugglinmike: ...so we may want to reach out to Daniel

lola: And perhaps Remy as well

Matt_King: ChrisC has the full context here, so is that a communication that ChrisC can send?

jugglinmike: Sure. They're out-of-office today, but I can ask them tomorrow

App issue 1665: NVDA/JAWS bot jobs get automatically cancelled

github: w3c/aria-at-app#1665

Matt_King: Elizabeth and mmoss had both expressed interest in learning about the bot infrastructure and code.

Matt_King: And if there's anyone else who would like to help maintain the bots, they are certainly welcome to

Matt_King: I've been thinking of setting up a separate meeting where the folks at Bocoup can help walk newcomers onto the bot infrastructure

Matt_King: We think that GitHub may have made a change that impacted the integration between the bots and the ARIA-AT app

jugglinmike: I think that might be ChrisC and I together

Matt_King: I'll look into getting a "Bot Infrastructure Class" meeting set up and scheduled before the next meeting. Not necessarily to actually hold the meeting before then, but to at least have it on our calendars

Minutes manually created (not a transcript), formatted by scribe.perl version 248 (Mon Oct 27 20:04:16 2025 UTC).

Diagnostics

Maybe present: Matt_King

All speakers: dean, jugglinmike, lola, Matt_King

Active on IRC: cyns, Joe_Humbert, jugglinmike, lola, Matt_King