Meeting minutes
Review agenda and next meeting date
<Matt_King> View agenda at https://
<Matt_King> Next meeting is October 24
<arigilmore_> https://
Jem: Our next meeting will be next Tuesday, October 24
Status of Site Updates
Jem: The resolution targets for October 31 are listed in the agenda for today's meeting
<Jem> Issue 697: Remove recommendation in menubutton pattern to add/remove aria-expanded="true
<Jem> PR 2831: Rating Slider Example: Redesign as an input for 10-value satisfaction scale by jongund
<Jem> Rewrite PR by Nick to separate kbd cords for multiple key strokes - Pull Request #1611
<Jem> Issue 2747: Incorporate/embed Feed example into APG Pattern Page
<trackbot> Sorry, but no Tracker is associated with this channel.
<trackbot> Sorry, but no Tracker is associated with this channel.
Matt_King: Hopefully we'll get the "slider" patch on its way today
Matt_King: As for the one related to feed, we're currently blocked due to the deployment issue which Alex_Flenniken is investigating
Issue 2501 - Rating slider redesign
Jem: Pull request is PR 2831: Rating Slider Example: Redesign as an input for 10-value satisfaction scale by jongund
github: w3c/
jongund: It was previously on a 10-point scale, but now, it's a "satisfaction rating" scale
jongund: Visually, it looks more like ten horizontal radio buttons, ranging from "extremely unsatisfied" to "extremely satisfied"
jongund: The previous example was a scale of five "stars" where one could assign half "stars"
Matt_King: I will conduct an editorial review
arigilmore_: I volunteer to do a visual review
Jem: Thanks!
Andrea_Cardona: I can do a code review
Jem: Thank you so much
Jem: I can try conducting a test review. I learned how to do this from Valerie a long time ago
Jem: I'll ask for Howard's help if I have any trouble
Matt_King: For the functional review, we should be extra sensitive to mouse and pointer behavior
CurtBellew: I can try the functional review
Jem: Thank you, it should be fun
arigilmore_: I can do that while I'm doing the visual review
Matt_King: If these things look like radios but we're calling it a slider, will we get pushback on that?
jongund: Should we change it to use adjacent rectangles instead of circles? That way, selecting values would look like filling up a bar
arigilmore_: It does feel like a radio group right now. I like the idea of the rectangles
Matt_King: I wanted to open this conversation, but I don't think we need to solve it right now
Matt_King: Using an existing precedent would be great. We don't want to invent here
Issue 2801: Media accessibility guidance
github: w3c/
Matt_King: This is a proposal that stemmed from a random airplane conversation on the way to TPAC
Matt_King: There's so much that authors build for media players
Matt_King: And we wondered whether it would be helpful to acknowledge and support that in APG
Matt_King: First: is this something that should be in APG?
Matt_King: Second: if so, are there people that have interests in the space who would like to participate?
Jem: I'm really excited about this! I work a lot with media accessibility, and I think this would fill a real gap in APG
Jem: Though we already have a "media seek slider" example
Matt_King: Ah, I forgot about that one!
Jem: I still think this is a great idea
[general discussion about the potential]
Matt_King: Sounds like there's some interest, but as usual, there's some concern about bandwidth
Matt_King: We'll keep it here as a "stretch" idea in the roadmap and revisit it early next year
Issue 2806: Must all treegrid rows have the same columns?
github: w3c/
Jem: the question is: Can treegrids have different column headings for child rows?
Matt_King: My gut reaction was "is course, all rows have to have exactly the same columns"
Matt_King: When I read the spec, and when I think about tree grids, for me, the notion of different columns sounds much too complicated to ever be useful
Matt_King: But then I thought about how many business applications do very complicated things, so I want to leave the door open to the possibility
Matt_King: And to be clear: I couldn't find anything in the spec that explicitly forbids this
Matt_King: It's not clear to me that treegrid is exactly one tree and exactly one tree merged together
CurtBellew: When I hear this, I imagine a "more detail" row which can be expanded and collapsed. That's the only use case I can think of
CurtBellew: Basically, it nested that one row. I can't remember exactly how the example I saw worked, but it was definitely difficult to use
CurtBellew: I wonder if what they're aiming at here is more like that--a parent with a child which holds additional details
CurtBellew: A lot of the time, in my experience, this is about nesting tables, which is always a big mess
CurtBellew: Usually, for us, it's system status. You have a list of systems. One row might describe a database that's either "up" or "down", and you can expand that to see more diagnostic information
CurtBellew: But that's only the closest thing I can think of. I'm not sure it really matches what's being described
Matt_King: I've never found a screen reader that handles nested tables gracefully
CurtBellew: It wreaks havoc with our internals checkers
Matt_King: We might want to ask for specific use cases which justify this
Matt_King: Even though our initial thought is that this is somewhat contradictory to a grid. That a grid is not a grid unless it has consistency between all rows. And that a tree grid is a grid, so that principle still applies
Andrea_Cardona: one person on the Carbon team focuses on the tree grid implementation
Andrea_Cardona: I haven't personally seen an example of this, but I haven't worked with our Tree Grid very much
CurtBellew: I would love to see some way to do something similar because we have some cases where something like this would be useful
Plan for including experimental content in APG
Matt_King: We've talked about this in the ARIA working group
Matt_King: If we were to add examples for the new "aria-actions" right now, we wouldn't want people to use that as a signal to go use them in their own project
Matt_King: We would like to be able to clearly label experimental examples as such so that authors don't mistakenly deploy incomplete patterns
Matt_King: I've put together a proposal for how this could work
Matt_King: It's this GitHub issue: w3c/
CurtBellew: I think this is a great idea
Jem: This came up at TPAC this year in the context of a number of new additions to ARIA
Jem: Why don't we add a new top-level menu item for "experimental". Or how about at the very top of the "patterns" page?
Matt_King: To me, the more invisible, the better. I don't want to expose this information to people who aren't looking for it explicitly
Matt_King: I guess the amount of exposure is something that we can experiment with over time
Matt_King: I think these would use the "example page" template
Matt_King: Though I suppose we can have an experimental patterns page, as well
<Jem> w3c/
Matt_King: This is something that we would like to be ready to do by the end of the year. dmontalvo please review the spec in the issue I've filed; the sooner we receive your input, the better!