Meeting minutes
<CoryJoseph> apologies for the last minute regrets. Home sick today
<Jem> https://
Setup and Review Agenda
Jem: Our next meeting with be January 16
Plan for 2023 retroperspective
Matt_King: I was thinking we could plan a retrospective for next week
Matt_King: I don't want to catch anyone off guard; I'd like to give folks some time to think about it
Matt_King: During that meeting, I'd like to first review and recognize accomplishments
Matt_King: ...then spend some time learning how the Task Force could function better
Matt_King: Are there other goals we should think about having during that meeting?
Matt_King: Hearing none, we can use just those two questions
Matt_King: I'd like everyone to think about them, especially the things that aren't going so well
Matt_King: I'll work on a review of what happened in 2023. I'll talk with Jem more about this
Developing 2024 roadmap
Matt_King: I'd like to talk about how we prioritize work
<Jem> Discuss prioritization framework.
Matt_King: I don't think we can get through this topic in this meeting, but I'd like to start talking, at least
Matt_King: I came up with seven different sources of work
Matt_King: They're listed in the agenda
Matt_King: 1. ARIA 1.3 and new 2024 features (aria-actions, AT notifications, Listview, changes to modal),
Matt_King: 2. Requests for new types of guidance (e.g., high contrast support practices, integrating "Using ARIA")
Matt_King: 3. Requests for enhancement to existing content
<Jem> we have 31 issues with label of bug
Matt_King: 4. Fixing known bugs
Matt_King: 5. Content design: Design of Landmark pages, pages with multiple examples (e.g., listbox, grid, button)
Matt_King: 6. Gaps in coverage of ARIA 1.1 and 1.2
Matt_King: 7. Improving code quality
Matt_King: Across these, what should our prioritization framework be?
Matt_King: We should probably consider which patterns people use the most
Matt_King: We have some metrics along these lines through the website access logs
Matt_King: Should we have a framework for making these decisions or just "shoot from the hip"?
Jem: I'm looking at statistics for the APG site from the w3c.
Jem: I think bug fixing should be a top priority. We have over 30 right now
Matt_King: Based on how we've worked in the past, addressing all the current bugs could take 6 to 12 months
Matt_King: Meanwhile, there are some 43 things in ARIA 1.1 and ARIA 1.2 that aren't covered
Matt_King: I'm just asking how we make prioritization decisions given all these different sources of work
Jem: This reminds me of creating a community group, as Matt_King recently proposed
Matt_King: Yeah, that's still on my to-do list. I think that could feed into the "bug fixing" work
Matt_King: It's a whole different dimension of improving how the Task Force works--finding new ways to get more people involved
CurtBellew: It's hard to make blanket statements about the relative importance of bugs and features
dmontalvo: Yeah, it will take some dedicated analysis. I'm willing to help with that
Matt_King: So having a "bug importance" or "bug severity" scale would be very important
<Jem> https://
Matt_King: In our issue categorization, I have made bugs separate from editorial problems like inaccuracies
Matt_King: Maybe for a future agenda, we can discuss the prioritization framework just for bugs
Matt_King: I'll create an action item for that
dmontalvo: You can assign me
<Jem> 16 open for both, bug and example of pattern implementation
Matt_King: we don't have a way of prioritizing gaps in coverage or feature requests
Matt_King: So we have two "next steps," here. I will help make those things happen and get them on future agendas
<Jem> https://
<Jem> https://
Status of Site Updates
Jem: We have a candidate for January publication
PR 2775: Feed Example: Move display of example from separate page into standard APG example page by ariellalgilmore
github: w3c/
Matt_King: All checks are passing
Matt_King: And there's some activity from Ari just a few days ago. Awesome
Matt_King: Jem requested changes
Jem: My feedback was addressed a long time ago
Matt_King: Could you approve it, then?
Matt_King: I think we still need editorical review (from me) and functional review from CurtBellew
Matt_King: And a visual design review from Jem
Jem: howard-e has reviewed tests
Jem: I can do my part. CurtBellew can you perform the functional review?
CurtBellew: Yes
Matt_King: And I will perform the editoral review.
Issue triage
<Jem> w3c/
clarifying "No ARIA is better than bad ARIA"
github: w3c/
Matt_King: Is this a valid issue or not?
Matt_King: Can someone explain the logic behind the first reading, "Bad ARIA is the best ARIA" ?
jugglinmike: If you read it is "[there is] no ARIA [that] is better than bad ARIA"
Matt_King: I think I'm kind of understanding, now, but it seems like a unlikely reading
CurtBellew: Maybe for folks for whom English is a second language
Matt_King: Would it be better to rephrase as "Not using ARIA is better than misusing ARIA" ?
siri: I prefer "No ARIA is better than bad ARIA" over "Not using ARIA is better than misusing ARIA"
Matt_King: Well, my goal for this meeting was mostly to learn if this was a valid issue
Matt_King: It sounds like most folks here think it is difficult to misread, so I'm tempted to close it
Jem: I don't know about that. I think it's valid. I think it's worth thinking about rephrasing
Jem: the reporter suggested another alternative, "bad ARIA is worse than no ARIA"
<dmontalvo> +1 to Jem
Jem: The current phrasing is kind of cocky, honestly
<Jem> although I love the current catch phrase. ;-)
Matt_King: Okay! I'm marking this issue as "agenda" and as "editorial"
menubar-navigation example aria-expanded without aria-controls/aria-owns
github: w3c/
Matt_King: The reporter is saying that the submenus are not DOM descendants of the parent menu item. That is different from trees, but is it allowed?
Bryan_Garaventa: It doesn't have to be a descendant in practice. However, I don't know if the spec actually says that it has to be a child in the same way it says for a tree
Matt_King: They're saying it's not a DOM descendant and it doesn't have controls, so there's basically no relation
Bryan_Garaventa: When you're getting information like "how many menu items", it does that based on the nesting of those menus. When you use a link, the same thing happens
Matt_King: It seems that in our example, there is no relationship between the controlling menu item and the controlled menu item. Is that a problem?
Bryan_Garaventa: You can break them out.
Matt_King: ARIA doesn't require us to put ARIA controls on there. We don't do it; the question is, should we?
Bryan_Garaventa: I would recommend not
siri: I learned some time back that as of now, the screen readers don't do anything with ARIA controls.
siri: What is the purpose of adding ARIA controls rather than "expanded" and "collapsed"?
Matt_King: It seems to me that if there's a focus, it can impact how focus moves across the document
Bryan_Garaventa: My understanding is that ARIA controls is useful when it renders content [...]
[Bryan_Garaventa's explanation proceeded faster than could be scribed]
CurtBellew: I thought that the point of ARIA controls was to allow devs to take something clickable and express that the clickable element controls some content
Matt_King: I've marked this as a question, assigned it to me, and added it as an agenda item.
Jem: It will be great to have more conversation! I'm always controlled about ARIA controls