Meeting minutes
<spectranaut> cochair: spectranaut
New Issue Triage
spectranaut: #1836
pkra: there’s been a proposal to deprecate
cyns: change the title to “Consider the future of aria-flowto”
spectranaut: pkra, please make a comment on the deprecation issue
spectranaut: #1835
… we can’t change 1.1, right?
jamesn: could do errata, but no point
spectranaut: #174
… web driver now has support for accname, a bunch of manual tests could be made automated
… bigger technical project
cyns: could be good for an intern
jcraig: will connect with an Apple engineer to write examples test for others to imitate
spectranaut: will put it in 2023H1 milestone
… return to it later
… #1832
jamesn: placeholder issue for when it needs to come on to agenda
spectranaut: #441, no need to discuss
… #440
scotto: I’ll take assignment
New PR Triage
jcraig: source for get_computed_label, and should also have get_computed_role:
https://
https://
jcraig: should get an issue in Firefox tracker
spectranaut: PR #14
jamesn: jcraig, will you review?
spectranaut: PR #442
scotto: have a minor companion PR to expand the definition (#1834)
… this is not normative, just expanding the definition of what the role represents
… meant to indicate content that is no longer relevant
… created browser bugs to change mapping
spectranaut: #1834
spectranaut: dig up related issue to “taxonomy”
spectranaut: #150
… we need a maintainer for UIA
… Dan at Microsoft might be the replacement
Deep Dive planning
<spectranaut> https://
spectranaut: want to find time to consider accname proposal from BGaraventa
BGaraventa: need participants to come to an agreement
BGaraventa: I’ll send out proposals for time slots
WAI announcement / Michael news
<MichaelC_> https://
<MichaelC_> https://
<MichaelC_> https://
MichaelC_: official announcements
MichaelC_: Judy Brewer is leaving W3C
MichaelC_: still working out details
MichaelC_: intend for WG operations to not be impacted
MichaelC_: for ARIA, Daniel Montalvo may transition in
MichaelC_: W3C is going through transition to separate legal entity
MichaelC_: waiting for that to be finalized
MichaelC_: will hopefully have plan in early 2023
jamesn: congratulations on new role!
MichaelC_: Judy will still be around for a bit
Concerning the current reality of aria-setsize
spectranaut: scotto did some research and JAWS-test linked another issue that sarahhigley opened
… is this a proposal now, scotto?
scotto: ARIA spec says that people can set -1 to mean “undetermined amount”, but core-aam says -1 should return 1
… browsers all act differently than each other
… would make most sense to return “undetermined”
… screen readers should be able to take that value and then add whatever they want to
… filed bug against Edge internally, but Travis’ departure affects that
spectranaut: Joanie’s suggestion is to add a “user agents SHOULD NOT calculate” statement
… what do we think?
scotto: yes, that makes sense
… there are mixed opinions
… let screen readers decide what they want to say is the best option
Matt_King: from the user’s POV, it’s difficult to say what the best thing to do is
… users need to be able to understand the size of the list. If there is a visual indication of that, it needs an equivalent
scotto: there _is_ a scrollbar there, but it’s fudged; it doesn’t reflect the actual size of the list
… it’s “generic” in that it represents “there are more”, but without accuracy
Matt_King: what we want to say is “there are N here, and there are more available”
jamesn: something like “Item 1 of at least 10”
Matt_King: if we have this indeterminate value, what does it actually communicate to AT and when should authors use it
… programmatically, maybe indeterminate is the correct implementation, but we haven’t said what it means for AT vendors and how to serve users
scotto: core-aam spec needs to be updated
spectranaut: do we agree that -1 should mean indeterminate? or should there be some calculation by the user agent?
Matt_King: we need input from AT vendors
… the number that’s in the DOM is actually useful and worth communicating
… if AT said “you’re on item 68 of 69 or more”, that would be useful information
Matt_King: right now, there are no normative requirements
scotto: I’ll take assignment of sarahhigley’s related issue
… next step: come up with draft proposal
… agree with Matt_King: do it in the spec first
spectranaut: first make specs consistent, then later change to provide guidance to screen readers?
scotto: yes, first fix spec inconsistency
… and then, yes, future discussion with AT
Should aria-hidden be effected by aria-owns? (milestone and assign)
spectranaut: aria-owns overrides aria-hidden, and we just need to update spec
… want to do what Chrome and Firefox do, then update spec
[normative][author] clarify widget in trees can only be group and treeitem
pkra: secondary actions issue specifically mentions tree items. Do we still want this?
pkra: secondary actions permits what this issue was originally intending to prohibit
jamesn: let’s discuss next week
Matt_King: don’t think they conflict, but should revisit