W3C

– DRAFT –
ARIA WG

13 July 2023

Attendees

Present
Adam_Page, CurtBellew, Francis_Storr, jaunita_george, pkra, scotto
Regrets
-
Chair
spectranaut_
Scribe
Adam_Page

Meeting minutes

<spectranaut_> remove item 8

[New Issue Triage](https://tinyurl.com/39cnyfep)

spectranaut_: a few issues from giacomo-petri
… #1981

scotto: raises an interesting quirk: since the title attribute can provide either a name or a description, in HTML you could use title but _not_ want it to be the accname
… I’ll take assignment

spectranaut_: will skip the next one
… #1973

pkra: can we transfer this to practices?

<spectranaut_> w3c/aria#1973

spectranaut_: there’s already an issue there

jaunita_george: I’ll take a look
… I also have a question about #1948
… not sure what the call to action is
… it has a had a default value in the past, but do we not have a default in the newest spec?

spectranaut_: let’s agenda this
… #1971

<spectranaut_> w3c/aria#1972

spectranaut_: #1972

pkra: let’s decide a milestone. 1.3 or 1.4?

spectranaut_: we’re trying to close 1.3, so let’s do 1.4
… we’ll agenda it later
… back to #1971

pkra: there’s parity for all users here: no implication of the context menu existing

scotto: yes, not sure what value this would bring

mario: yes, we have context menus on every object, every list, even in empty places, so this would not be helpful

spectranaut_: pkra, would you respond on this issue, and maybe close?

pkra: yes

spectranaut_: #184, core-aam
… seems straightforward
… last one, #1970
… scotto, I see you linked this to `aria-controls` issue

scotto: yes, we’ve discussed this before

spectranaut_: is someone working on language for that `aria-controls` issue?

pkra: needs input from an aom person

spectranaut_: the `aria-controls` issue was milestoned for 1.3 and this may just be a clarification

[New PR Triage](https://tinyurl.com/3z6dyvnx)

<spectranaut_> w3c/aria#1980

spectranaut_: I’ve added scotto to review

CurtBellew: I’ll review as well

scotto: I’m a little worried that this emphasizes `aria-controls` too much

scotto: this doesn’t actually address the issue, let’s work on it offline

spectranaut_: #1977

spectranaut_: #1975
… scotto, you already reviewed

scotto: yep, this is a good correction

pkra: I’ll review

spectranaut_: #1965
… I left this in draft
… follow up from F2F discussion
… re: accessibility parent/child relationships
… there was some confusion over the definitions that sarah_higley added
… so I’d like some review of that
… and a new term was introduced: accessibility descendant

pkra: this is great

spectranaut_: I already added pkra and sarah_higely for review

scotto: I’ll review as well

Adam_Page: I’ll review also

[Deep Dive planning](https://bit.ly/aria-meaty-topic-candidates)

spectranaut_: we have a deep dive planned for July 27
… summary role parity issue
… 2 weeks from now
… then we need to schedule a new one: use cases for aria-modal

<spectranaut_> w3c/aria#1950

spectranaut_: we haven’t scheduled yet because we want to wait for aaronlev to be back
… so we’ll do that when he’s back
… any thoughts on either?
… nope

TPAC: Request for a combined meeting with APA on Accessibility Notifications.

spectranaut_: PSA: we got an email from APA that they want to collaborate with us on ARIA notifications
… would anyone like to attend that meeting remotely, and need to weigh in on scheduling?

<jaunita_george> I may have to attend remotely

spectranaut_: okay, stay tuned jaunita_george — there’ll be a draft agenda soon

[Could hyperlinks to the current page expose aria-current=page](https://github.com/w3c/html-aam/issues/494)

spectranaut_: scotto, this is yours?
… we talked about in triage a couple weeks ago

scotto: yes, I’d like to see if there’s any implementor interest in this
… seems feasible for Apple, at least

pkra: I think it’s JAWS that announces in-page anchor links as on the current page
… so, yes, this seems like a good idea

scotto: I mainly want this for links inside navigation
… those are typically already designed with a visual indication
… should be a simple change to the spec
… in html-aam for example, I’d add something like “if link contained within an element with role nav and link URL matches href attribute, then implementers SHOULD expose the hyperlink as current page”
… I acknowledge that this can’t always happen, such as in SPAs

spectranaut_: needs review from jteh, jcraig, and aaronlev

pkra: this would be a wonderful example on the editorial side for AT guidance

scotto: I’ll come up with a proposed edit to the spec

[Discussion tracking for ARIA Notification proposal](https://github.com/w3c/aria/issues/1957)

scotto: is there anyone here who’d like to discuss this?

[Rethink how "hidden" is aria-hidden content?](https://github.com/w3c/aria/issues/1951)

scotto: we need more people to discuss this, there are strong opinions
… JAWS developers recently asked if there are any cases where they can choose to not prune some elements with aria-hidden
… so that they’re still empowered to do good error detection
… aria-hidden has been misused by authors
… inert exists now, and it successfully hides content from all users
… I don’t have a strong proposal here, this is just a conversation starter

spectranaut_: it’d be a little bit of a shift for browsers and AT vendors

scotto: this is kind of what aria-modal does
… outside of Webkit, aria-modal doesn’t actually remove content from the a11y tree
… it just informs the screen reader not to expose it while the user in the dialog
… there is precedent for this, in that some browsers do keep aria-hidden content in the accessibility tree but inform AT not to expose it

cyns: we can’t afford to forget about other AT besides screen readers, as well as screen readers that are less sophisticated than NVDA and JAWS

scotto: agreed

CurtBellew: it sounds like your proposal doesn’t necessarily break what’s out there, it’s more about informing the user what is hidden rather than eliminating it from their visibility altogether

scotto: the intent is to allow for sensible error detection for author misuse of aria-hidden

CurtBellew: inert hasn’t been around that long

scotto: yeah, only a year or so

spectranaut_: let’s continue this discussion later

<Zakim> cyns, you wanted to say we'd need to think through how this would impact other AT besides screen readers, and screen-readers that are less sophisticated than NVDA/JAWS and to say it's a big, deep change that will require a lot of thought and testing

1948

jaunita_george: was the default value for aria-atomic removed?
… it does still seem to be in the editor‘s draft
… so wanting to know if a change was discussed

scotto: this was something that Anne did in a PR a few months ago
… so, yeah, very recent — it’d be good to track down that PR

spectranaut_: which PR?

<pkra> 1894 ?

<pkra> w3c/aria#1894

scotto: yep, #1894 is the correct PR

pkra: Anne removed that it *has* a default value since it’s a computed value based on its ancestors

jaunita_george: ah, it‘s a computed value — so then it will need to change in a lot of places
… so this should go into 1.3
… maybe we should explain it a bit better; how does it compute the value?

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

Diagnostics

Succeeded: s/an aam person/an aom person

Maybe present: cyns, mario, spectranaut_

All speakers: Adam_Page, CurtBellew, cyns, jaunita_george, mario, pkra, scotto, spectranaut_

Active on IRC: Adam_Page, CurtBellew, cyns, Francis_Storr, jaunita_george, pkra, scotto, spectranaut_