Meeting minutes
New Issue Triage
jcraig: There are other properties that may relate to this, including `status`.
jcraig: there may be other related issues.
jamesn: Can jcraig look into this?
jcraig: I can look into this.
jcraig: #2256 may have a related issue
jamesn: Should I assign you or agenda next week?
jcraig: Add me in the comments and I'll find the other issue.
jcraig: For context add "starts media" in the note for my searching later?
jamesn: #2254 seems to be answered. Anything more to do other than close it?
jcraig: The reader may not think it's clear, so perhaps ask where the reader would expect notes like this in the spec?
aardrian: Assign it to me
jamesn: skip 2242
jamesn: Talking 2241 with other editors.
New PR Triage
jamesn: #2245 seems straightforward
sarah: What we talked about last week.
Jamesn: Reviewers?
aardrian: Me
siri: Me
giacomo-petri: me
WPT Open PRs
jamesn: nothing new here. Anything you want to talk about jcraig?
jcraig: Good to go. Unless someone else wants to talk about one of them.
Deep Dive planning
jamesn: We have one next week on undefined values / states / properties.
jamesn: The week after is 4 July, so no meeting owing to US holiday. 11 July is next available deep dive date. Any proposals?
jcraig: #2166 implementation concerns over orphaned roles? I'll lead that.
TPAC registration open - https://www.w3.org/2024/09/TPAC/
jamesn: TPAC registration is now open.
jamesn: Monday and Tuesday are the most important days for ARIA.
jamesn" Remainder days are with other working groups, good to participate.
jamesn: Wednesday is the plenary / breakout.
jamesn: Book hotel now if you want the group rate in Anaheim.
jamesn: You want the group rate.
jamesn: Really. You do.
aardrian: Any prep I need to do? Other then be familiar with the work?
jcraig: Wednesday can be interesting for breakout session, which means you can pitch one of your own ideas.
giacomo-petri: Are the other working groups confirmed?
jamesn: It's impossible to not overlap (which was question), but everything is scheduled.
jamesn: Will try to firm up agenda as date approaches to try to avoid other group activity conflicts.
jamesn: This is driven by the WG, not the chairs. So it's up to those who bring the topics.
Charter arter.html
<dmontalvo> Proposed ARIA Charter
jamesn: This is the WG charter. We already have an extension and getting another ... well, go look at it.
jamesn: Lists major deliverables, AccName, ARIA in HTML, ARIA, Core AAM, etc., along with other deliverables.
jamesn: Proposing everything becomes a living standard, without version numbers.
jamesn: If you have feedback, questions, get them in.
jcraig: We should make a play for SVG ARIA for the monorepo.
jamesn: We have.
jcraig: What about CSS AAM?
jamesn: No, the current proposal is to fold it into HTML AAM.
jamesn: One of the reasons Rahim is now an editor of that spec. It's too much for one person otherwise.
daniel-montalvo: We may need to clarify that in Section 5. I am happy to update.
jamesn: We're asking group to review to move forward. Do we need anything formal?
daniel-montalvo: This approach is fine. Maybe a week or so to discuss, but if we agree then the minutes will suffice.
daniel-montalvo: Or we can come back.
jamesn: Email the chairs if there is anything you need changed. Or a mailing list discussion. Or open an issue.
ARIA should clarify distinction between “contents exposed as descendant text nodes” vs “name from contents” (Was: listitem should include name from Contents or name prohibited) - need someone to take this on
jamesn: We need someone to take this.
jamesn: Ran out of time talking about this last week. Agreement that we want to do it but nobody stepped up.
aardrian: Seems to make sense.
jcraig: Yeah, it seemed we only needed a volunteer to at least fix list item. There may be more.
jamesn: Why would this be an HTML editors issue?
jcraig: Maybe I'm misremembering, but I thought it had items that aligned with HTML AAM.
jamesn: Anyone want to volunteer?
jcraig: IF you
jcraig: If you're unsure about volunteering, happy to spend some one-on-one time with you.
jamesn: Seems like a good "medium" issue.
td, th naming doesn't align with ARIA
jamesn: I thought this was linked to previous issue, not sure anymore.
jamesn: Scott raised it. ARIA says name comes from contents, HTML AAM does not say that. Also a case to allow name from author.
sarah: I agree with Scott for table, less so for table.
sarah: Column header is the same role across both, so there is a case in a grid once you have extra controls in the cell.
sarah: More about overrides than adding more.
sarah: There isn't a good definition of whether name from author should override or augment.
jamesn: I'm confused what a name means for a cell in a table. What is it?
jamesn: A cell can have headers, which is not really the name. I've never really understood why you'd name a cell in a table.
sarah: Another thing poorly defined -- when on cells you don't want the column / row header name to also include the names of all the actions embedded within it.
sarah: There is no spec for how to handle for that.
jamesn: We need to have all the use cases laid out and satisfy those than try to sort one issue at a time.
jamesn: We won't get all the use cases this way.
jamesn: Grids are on my wishlist forever but have never gotten 'round to doing it.
aardrian: There are already bad patterns where authors override cells _and_ headers, so my concern is do we forbid or reflect?
jamesn: Is there a good solution to this?
aardrian: er... nope
<sarah> Narrator voice: `abbr` does *not* work 😅
jamesn: Think about `<abbr>` as a solution.
siri: What are we trying to do here?
siri: How do we approach this?
aardrian: Mostly trying to sort the issue at hand without creating more problems.
jamesn: We may need a deep dive. Perhaps an actionable proposal prior to TPAC.
jamesn: Then we can focus on it. Can someone propose this and work out a plan?
aardrian: I can collaborate but not lead.
notificationId: discussion points
jamesn: We had a conversation a couple weeks ago and comments for which we have been asking for feedback.
jamesn: Here is the notificationID.
jamesn: Would anyone like to leave comments now?
jamesn: Otherwise leave comments there.
aardrian: I had comments but no responses.
jamesn: Can you split it and put it into the specific discussions?
aardrian (grumpily): Yes.
sarah: Would it be helpful to summarize in the meeting what notificationID is? Or is this procedural?
jamesn: I would hope people know about it by now.
sarah: A lot of the Microsoft discussions have been about the dev talking to the SR vendor to create a custom script for their custom notificationID token.
sarah: This is doable within Microsoft owing to Narrator and Edge. But I want more feedback outside Microsoft for how it could be used.
<Zakim> jamesn, you wanted to say that it doesn't really sound scalable
sarah: Is there anything people need to feel they can give feedback? I will forward it along. In this call or otherwise.
jamesn: That was how I thought it was working but it doesn't seem scalable for most devs.
jamesn: I assumed it was a registry of common idioms and it works from those.
jamesn: I imagine it might work, if implemented today, that devs will simply use the same items that Microsoft does.
sarah: It's very hard to look at what others do, so you would need to listen to the AAPI event directly and inspect the properties being passed.
jamesn: But you can do it!
Sarah: Sure.
aardrian: I thought this was a token list from my review of the deep dive.
<jamesn> jamesn nods in violent agreement with aardrian
I want a token list for earcons or whatever.
aardrian: It seems unlikely that devs could inspect a highly specific implementation.
coryjoseph: From the perspective of having the resources to use this to its full potential is critical.
coryjoseph: I can see this in use for health care advisors to qualify the different kinds of notifications.
coryjoseph: But even in a company as large as mine is a lift, complicated by getting all the vendors to agree to the implementations.
aardrian: And then getting SR vendors to sign on.
jamesn: This feels like a solution only viable for MIcrosoft, Google, et al.
jamesn: We have folks here with opinions who need to put them into the discussions.
sarah: I provided my contact info for anyone who wants to ask questions outside the call.
jamesn: If you don't have access to the Zoom chat, say so and ... actually, get the email from the group contact page.