Meeting minutes
scribe; chlane
New Issue Triage
dpubaam 16
jamesn: sounds like a PR is the easiest way
jamesn: who is the editor
martt garrish
assigning to matt
aria act rules review
go over later
#176 accname
would not pick up aria description from there
adding editorial note
milestoning next year
adam page will rewrite
New PR Triage
html aam pr #446
popover related attributes
jamesn: no reviewers needed but can volunteer if involved
scotto: not complete yet
Deep Dive planning
next week we have
aria-flowto
jamesn: no more slots for this year
jamesn: propose no more this year
jamesn: resolved
ACT/ARIA Review
Wilco:
act is a project in the a11y guidelines working group
8 years working to get orgs to test wcag consistently
tools focus
manual testing methodologies
developed ACT rules format as a w3c recommedation
for how to write rules
applicability/expactations
effectively and few requirements
used in tools to verify
have a lot of wcag rules and some ARIA rules
wcag/aria overlap
is where my ask for this groups come in
WCAG/ARIA has overlap
ACT is a task force, need a Working group to say yay or nay
to "how we think aria should be tested"
jamesn: take us through one of these rules
jamesn: glossary forms a significant part of the length of the rule
<Wilco> https://
know how these are constucted to prevent duplication
"aria state or property has valid value"
Descriptions, Applicability, Expectation, Assumptions
applicability looks for non empty values on svg or html elements
any element with aria attributes
expectation, a valid value
value types are defined
ID reference rules are in question
need ARIA WG opinion
jamesn: see contradiction
Wilco: generic URI syntax
does ARIA require this?
do we have value types in URI?
we do
Wilco: assumptions, related to wcag
theoretical false positive opportunity
interpretation questions
background provides more details
bibliography,
accessibility requirements mapping, habe no or unknown mapping
input aspects, the info necessary to run test, access to DOM, CSS
test cases
pass fail and inapplicable
Glossary, has definitions
'namespaced element', 'outcome'
links to specs
implementations, list of tools and testing methodologies that have know consistency with the rule
Wilco: questions?
StefanS: are axe and AMP related?
Wilco: not a question for ACT task force
Wilco: this list comes for orgs that publish test results
orgs who want to can run there test methods and tools against this
list of orgs and tools that have done that work
jamesn: I think we look through all pass fail examples
<Zakim> jcraig, you wanted to say I think URI may have been proposed for aria`describedat`but I don't see any current URI types https://
jamesn: why put in passing failing examples
jcraig: URI had been proposed with describedat
the applicability and expectations sections list aria 1.2
wondering why
Wilco: not sure how else to do it
Wilco: what do we do with aria 1.3?
jcraig: just link to top level of ARIA?
<Zakim> jamesn, you wanted to react to jcraig
Wilco: have to worry about versions
jamesn: 1.1 rules not valid in 1.2
jamesn: version specifics are a good idea
Wilco: we review rules once a year
living documents not parked
Wilco: inconsistencies happen as these things evolve
cyn: aria conformance test suite source for additional rules?
Wilco: it might be interesting to explore
scotto: 2 things, specific versioning it would be great if can not do it in case we go to living standard model
no such thing as version numbers
expecially in this space
rules may not help people anymore
second thing,
StefanS: these rules are defined and correctly spelled, allowed/not allowed overrides
override checks?
Wilco: a separate rule
StefanS: super helpful with discussions with developers
these cases are detected late
jamesn: not pure ARIA html/aam rule
aria in html
jamesn: how to review
Wilco: not a tight timeline
Wilco: years to go
piecemeal approach is good
start with one
Wilco: announcing a formal list of implementations
<Wilco> https://
jamesn: any rules that those things are checking that you are asking us to review
Wilco: the proposed rules
need our agreement
Wilco: no rush, start with 1 or 2, what is the preferred way for this group to survey things?
jamesn: concerned with rules already approved that the group won't agree with?
jamesn: anything with ARIA should go through this group
Wilco: no ARIA specific rules
<Wilco> https://
jamesn: proposed rules for wcag look aria specific
rules beyond wcag include ARIA rules
there is overlap
jamesn: I'd like to set timelines
when are you looking for feedback
Wilco: no timeline
in an ongoing process to get a very lengthy list approved
working in batches with WCAG
we would do that every few months
jamesn: filing an issue will be good
jamesn: this is important for content creation
jamesn: encourage a careful eye
Wilco: would love more rules to write
jamesn: go through spec and look for authors MUST
cyns see core aam as well
svg and css should be looked at too
jamesn: we used to avoid authors 'must'
we didn't have tests
cyns, aria authors must in 1.3
cyns looking for things that will become author must in 1.2
jamesn: we created test cases, not as detailed with multiple pass/fails
Wilco: we think there are places that there should be authr must and should
jamesn: show of hands for interested
<Adam_Page> I’m interested
I'm interested as well
<spectranaut> I'm interested
<CurtBellew> I'm interested.
<scotto> happy to review as well
<melsumner> I've already been reviewing them :) Excited about this project, and ember-template-lint has some good a11y rules too.
jamesn: if you are creating frameworks you should review this
spectranaut: talk about how to split up work
spectranaut: in the issue, you can list the one's you have reviewed
comment that you looked at the whole thing
scotto: good place to move to discussions
jamesn: using discussions and threads for each rule
and glossary definition
jamesn: only downside
they don't query the same way
we have to remember to look at them
1.3 blocking issues agendabot]
jamesn:
jamesn: no movement
great to get some movement, some will go away with Sarah's descendant issue
jamesn: has that one had any progress?
jcraig: suggested another name
child, descendant parent has baggage from DOM
jamesn: 1454
jamesn: were close
scotto: prefer ancestor roles
in html aam been working on
the concept of scoped elements in html
and how roles are related based on that
that is where my terminology is going
jamesn: if you can all agree that is great
jamesn: last thing other than tests to merge
jamesn: coming to concencus just worried about naming
jamesn: if we are going to use a word that has another meaning in the DOM
we need to clarify
jcraig: fwiw, drop context role
requiredaccessibiltyancestor
Adam_Page: working on pr related to ownership
ancestry
can be another candidate for naming
jcraig: inclusivity, like master slave
jamesn: allowed elements rather than owns
cyns aria owns does not change a11y tree
jcraig: complicated but will get it done
cyns is heirarchy a11y tree?
jcraig: yes
cyns add a11y tree things to glossary
more clarification would be good
sarah
same as original problem
direct ancestor and descendant
jamesn: objected to DOM meanings
need to be unambiguous
melsumner: not talking about dom anymore?
cyns like proposal
cyns talk about both in same paragraph
siri: aria-owns
issues comboboxes
when I raised issues and point to 1.2
developers cant implement
when will it be ready?
jamesn: soon 1.2 will go to pr
jamesn: core-aam is in CR
jamesn: have passing test
cyns; once in pr, how long?
jamesn: 30 days?
MichaelC: 30 day for CR review by advisory committee
MichaelC: no objections can go to req
jamesn: many people a PR is good
jamesn: anyone doing a 1.1 combobox is making more work
cyns this happens in big companies
cyns, timeline?
jamesn: working on it, before xkmas
end of january
2023
cyns will say q1
jamesn: how to get 1.3 draft out
<jcraig> DPUB-AAM issue: https://
<jcraig> PR: https://
<jcraig> The PR is Apple API specific, but IMO, other APIs may want to do something similar to not overload their AT users with verbose minutia.
<jcraig> Reviewers requested. Thanks.