W3C

– DRAFT –
ARIA WG

15 June 2023

Attendees

Present
CurtBellew, Matt_King
Regrets
-
Chair
-
Scribe
dmontalvo, jamesn

Meeting minutes

New Issue Triage

w3c/aria#1956

<spectranaut_> https://w3c.github.io/aria/#mapping_additional_relations_error_processing

<spectranaut_> for combobox: "When the popup is displayed, authors MUST set aria-controls on a combobox element to a value that refers to the combobox popup element. "

<spectranaut_> seems like the group agrees this particular scenario is not an error

<spectranaut_> matt: the aria-controls shouldn't have a value when the popup is not open

<spectranaut_> matt: I'm not sure that is how things work right now

<spectranaut_> matt: I think we shouldn't see these attributes when the ids are invalid. do we have tests for that?

<spectranaut_> scotto: the only person doing anything with aria controls is JAWs

<spectranaut_> matt: that still works

<spectranaut_> scott: in 2019 they removed that as a default setting

<spectranaut_> matt: that was the anouncement of that, it still works

<spectranaut_> jamesn: I will write a response - this is a validator error

<spectranaut_> next: w3c/aria#1955

there are some links to APG which are 404s

assigning to Matt to fix

w3c/aria#1954

scotto: there is a precedent for this. CSS WG is coming up with a user modifed CSS property. Regardless of feelings - this may be something that we should take a look at

jamesn: the mythical CSS AM may have its first mapping if that happened

spectranaut_: 2.0?

jamesn: tes

jamesn: yes

w3c/html-aam#492

#1951

Scott: The use case for this is to avoid misuse of aria-hidden
… the idea is for someone using screen reader to turn on a setting that allows access to hidden content
… NOt sure I agree with the idea, I do agree that aria-hidden is problematic
… Now inert can do things we previously did with aria-hidden

Valerie: We can discuss this

Scott: Yes.

Matt: Do we want to put this on the agenda at some point or this goes to deep dive?

Valerie: Let's put it on the agenda first

Scott: I won't be here next week

Valerie: Should we mileston this?

Scott: 1.4 at the earliest

New PR Triage

James: We can't do things for next week as people has not suggested anything this week

w3c/aria#1947

James: 1947, please people put comments there
… There is already one comment there

Matt: Is this for June 29?

Valerie: Yes.

James: Last one before my vacation

James: Please read this and come up with questions to discuss

Valerie: Should we put accordions and @@@ on the agenda?

James: I won't be here in July

Deep Dive planning

TPAC 2023

https://www.w3.org/2023/09/TPAC/

James: TPAC 2023 registration is open
… 11 - 15 September, Sevilla (Spain), hybrid
… Monday and Tuesday we are meeting, we have a slot on Friday that we are not sure if we would be using
… Half the price for early registrants until 15 July
… The hotel registration is also open

Matt: The meeting is in the same hotel?

James: Yes.

Daniel: People can get to Madrid and then take a train

James: There is funds available for registration, travel, and even childcare

https://www.w3.org/2023/09/TPAC/registration.html#inclusion-fund

James: You can apply for it if you think you'd apply

Consider further use cases for aria-modal/potential values

Scott: Right now we have aria-modal, the content outside the dialog is treated as hidden
… Depending on implementations, the ocntent is hidden but not out of the accessibilty tree
… Sometimes you may want to limit the virtual cursor, but other times it needs to be something in between
… It would allow for a way to soft-trap virtual cursor. AT will wanr that if users want to escape of that they can do something
… This sort of happens already for menus, forms, things where "forms mode" is activated

Matt: This is exactly what I described for non-modal dialogs
… We should name the kind of boundaries that AT should have
… Flexible versus solid boundaries. I thought these would be associated with roles
… It sounds like you would be able to do this without a role

Scott: There are various different types of elements that can be positioned to top other elements in the UI
… Sub navigation, sub lists etc
… A navigation where there is a primary link and then a button to invoke a secondary bunch of links related to the first level navigation

Matt: We oculd do this as a dialog

Scott: That's unnecessary.

Matt: If you are stuck in something you need to know what that something is
… Currently screen readers expect that thas be a dialog
… Do we need another thing to get stuck on?

Dug: What is the scenario?

Scott: I have had comments from people that they would have liked to know in these cases that they are inside these sub menu layer
… Then there is disclosure patterns as another use case

Matt: If you want people to get soft-trapped we can use modal dialogs. When you press the command to read the title, it'll read the dialog title if you are within the dialog instead of reading the page title
… If we come up with several boundary areas then that feels that there is potential for creating a more complicated experience
… I agree that it is valid to restrict or trap reading, that's sometimes useful. We should not expect people to learn that there are different ways in which you can get trapped
… The screen reader may ask to to take an extra actions

Synthia: Could we do a deep dive in August or something?

Scott: This is absoluetly related to pop-over. Not everything can be a dialog, that would not even work in HTML
… A paragraph could not be a dialog, for example

Synthia: I propose that we connect this to the pop-over then

Matt: Maybe the word "dialog" is a problem in itself

Synthia: Agree

Matt: Dialog in ARIA does not need to always map to dialog in HTML

Synthia: It's less confusing when it is as similar as possible
… In Windows there is an assumption that a dialog has a windo handle

Valerie: We will make a deep dive about pop-over early July

Consider ignoring pseudo element content for accname if it has empty alt text

Valerie: Discussed briefly last week
… Do we have the people to discuss this right now?
… We'll leave it on the agenda then

ARIA notification proposal

Scott: I have not started an issue collecting the comments on the F2F
… I'd like to ask the group if we want a bunch, more granular issues, or rather that we create a tracking issue with several discussions

Valerie: I like the tracking issues

Matt: We could have a project to group all these related issues
… How many of them there are?

Dug: 8 - 10

Valerie: Projects move through stages, these will result in actions we need to take

Dug: Some are going to be bigger than others, these are all related to comments on the F2F

Matt: if they are going to have separate outcomes the project may be good, if not then other ways to organize this may be better

Scott: Happy to do with anybody wants

Summer Holidays

Dug: Many people will be out for the summer. IS this a good time for notifications?

Valerie: I think we should discuss at least a bit during the summer

g/James/James/

<dmontalvo> s//JAmes/James/g

Minutes manually created (not a transcript), formatted by scribe.perl version 210 (Wed Jan 11 19:21:32 2023 UTC).

Diagnostics

Succeeded: i/... You can apply for it if you think you'd apply/https://www.w3.org/2023/09/TPAC/registration.html#inclusion-fund

Succeeded: s/.. 11/... 11/

Failed: s//JAmes/James/g

Succeeded 2 times: s/JAmes/James/g

Maybe present: Daniel, Dug, James, jamesn, Matt, Scott, scotto, spectranaut_, Synthia, Valerie

All speakers: Daniel, Dug, James, jamesn, Matt, Scott, scotto, spectranaut_, Synthia, Valerie

Active on IRC: CurtBellew, dmontalvo, jamesn, Matt_King, spectranaut_