Agenda for TPAC 2019

[jnurthen]

Current Agenda is at
https://docs.google.com/spreadsheets/d/1eeq-xg0li2IKEIiyKsqhv-QOxArtbTV3m0OrG8ns-w4/edit?usp=sharing

Remote participation at https://www.w3.org/2017/08/telecon-info_aria-tpac

IRC channel #aria

Please Add the F2FCandidate Label to any issue you want to add to TPAC Agenda.

Also please also add a comment in this issue with:

• a brief statement about the issue
• any other groups which should be involved
• approximately how much time this needs
• any other constraints on the discussion

Once reviewed and added to the agenda we will remove the F2FCandidate label and replace with a F2F label.

[carmacleod]

For #1021, we need to at least decide which milestone (1.3 or later).
If time, need to discuss in more detail.

[mcking65]

I propose a single segment on the topic of "Effects of ontology on table and grid" for discussing approaches to resolving the following 3 issues:

• Issue Consider prohibiting name from contents for rows in static (non-grid) tables #1022: (Consider prohibiting name from contents for rows in static (non-grid) tables.
• Issue Should aria-selected be allowed on row, columnheader, or rowheader in table? #1008: (Should aria-selected be allowed on row, columnheader, or rowheader in table?
• Issue WAI-ARIA 1.1: Cell and gridcell #79: WAI-ARIA 1.1: Cell and gridcell, which proposes depricating gridcell.

I don't believe we need non-WG participation.

I think it will take a solid hour or more to clearly layout the problems and weigh options.

[mcking65]

• Topic: Supported Combobox Implementation Patterns and mappings
• Time required: 90 minutes
• Outside participation required: No, except need HTML AAM representation.
• Issues:
  • The Combobox 1.0 design pattern references incorrectly include aria-owns instead of aria-controls. #998: The Combobox 1.0 design pattern references incorrectly include aria-owns instead of aria-controls.
  • aria-controls a required property for Combobox? #776: aria-controls a required property for Combobox?
  • Combobox: aria-controls or owns should be on input, not on combobox element, and only when combobox is expanded #716: Combobox: aria-controls or owns should be on input, not on combobox element, and only when combobox is expanded
  • Combobox has aria-expanded required, but there is a default value for aria-expanded #853: Combobox has aria-expanded required, but there is a default value for aria-expanded
  • Clarify combobox label placement #909: Clarify combobox label placement
  • Does combobox require an accessible name? #1046: Does combobox require an accessible name?
  • The ARIA 1.1 Combobox design pattern endorses unlabelled form fields as normative spec. #893: The ARIA 1.1 Combobox design pattern endorses unlabelled form fields as normative spec.
  • role combobox either misses text or is incorrect regarding role searchbox #742: role combobox either misses text or is incorrect regarding role searchbox
  • Update 'combobox' spec as per html spec/html-aam mapping for <select> element  #817: Update combobox spec as per html spec/html-aam mapping for <select> * [HTML-AAM: select element should map to listbox role regardless of presentation
  • HTML-AAM: select element should map to listbox role regardless of presentation html-aam#46: select element should map to listbox role regardless of presentation
  • Add aria-valuetext as a supported property of listbox #721: Add aria-valuetext as a supported property of listbox
  • Add aria-expanded as a supported property of listbox #722: Add aria-expanded as a supported property of listbox

There are several topics of discussion related to the above issues. I will organize them into an agenda for this segment.

[mcking65]

• Topic: Long term planning for W3C explanitory resources related to ARIA and other web technologies
• Duration: 45 minutes
• Other groups: Web Platform WG, Education and Outreach WG
• Related issues: None, but could create one if @jnurthen thinks is appropriate.

Objective

Discuss long term options for making it easier for authors to get the practical information they need about developing accessible sites.

Problem

Currently, WAI-ARIA Authoring Practices,
Using ARIA,
WAI Tutorials,
and possibly other resources, such as some WCAG techniques, all offer valuable information to authors.
While there are some advantages to having separate resources, there are also significant disadvantages, especially to authors who fail to find relevant information simply because of the way the separate deliverables are scoped.
There are also downsides to WGs to having separate but sometimes redundant objectives.

[mcking65]

I would like to discuss group related topics for 20 minutes, including:

• Use of group within list is unclear #1011: Use of group within list is unclear
• listitem role seems to be missing directory as required context role #1030: listitem role seems to be missing directory as required context role

[mcking65]

• Topic: aria-details
• Time required: 45 minutes
• Required participants: DPub? Whoever is most dependent on aria-details.
• Primary issue: aria-details precedence over aria-describedby  #1009: aria-details precedence over aria-describedby

We had to pull aria-details guidance from APG 1.1 Release 4 because it is not clear when it is beneficial and what the anticipated screen reader support might be like. Initially, it was part of a section on accessible descriptions. However, it is unclear what the relationship between descriptions and details is. We need to answer some basic questions for authors, such as:

• If details is provided, can you also provide a description? Or, is it necessary to choose one or the other?
• What if a field has a description, an error message, and details? What is expected?
• Is there an expectation that details will be read automatically if associated with a focusable element and that element receives focus?

[joanmarie]

w3c/core-aam#52

[joanmarie]

Rossen has suggested Tuesday 11:00 - noon for a joint meeting with CSS, ARIA, and APA.
https://lists.w3.org/Archives/Public/public-apa/2019Aug/0027.html

We should reply to him as soon as we know this is doable.

EDIT: UPDATE: I replied stating the time is confirmed. Unless we hear otherwise, we should plan on this time.

[joanmarie]

I'd like to discuss aria-brailleroledescription. #924

[jnurthen]

DPUB meeting is 9-9.30 am on Tuesday.

[zcorpan]

I'd like to propose:

• Meeting with WHATWG HTML editors about HTML accessibility issues and improving collaboration between HTML and accessibility groups

Suggested time: Monday, right after lunch (14:00)?

cc @annevk @domenic @foolip @sideshowbarker @mcking65 @jnurthen

[jugglinmike]

I'm interested in extending WebDriver with commands that take advantage of ARIA semantics and the APG. This could be a compelling incentive for developers to use accessible patterns.

My colleagues @spectranaut and @zcorpan will be available to lead a discussion at TPAC, and WebDriver editor @AutomatedTester has offered to lend us their expertise. Can we find 30 minutes on the schedule?

[jnurthen]

Yes. @jugglinmike Please suggest a time. Tuesday afternoon would be the easiest for us to move things around but other times may be possible.

[jugglinmike]

Thanks, @jnurthen! How about 1:00 PM on Tuesday? (Really, any time will work for us)

[jnurthen]

@jugglinmike you got it. 1pm on Tuesday. @alice @cookiecrook not sure how this fits in with AOM but you might want to make sure you can attend.

[annevk]

@jnurthen is #1001 (comment) happening? There's a lot happening at TPAC so it'd be good to know.

[jnurthen]

@annevk @domenic @foolip @sideshowbarker @mcking65 @zcorpan we have a meeting about web components scheduled at 2 on Monday. We could do 2.30? How long do you need?

[zcorpan]

How long do we get? 🙂 I think we could likely fill whatever time we get, but don't know if people have conflicts. How about 45 min?

[css-meeting-bot]

The ARIA Working Group just discussed Long term planning for W3C explanitory resources related to ARIA and other web technologies.

The full IRC log of that discussion

<ZoeBijl> TOPIC: Long term planning for W3C explanitory resources related to ARIA and other web technologies
<BGaraventa> I got lots of energy
<ZoeBijl> https://github.com//issues/1001#issuecomment-517806506
<ZoeBijl> github: https://github.com//issues/1001#issuecomment-517806506
<ZoeBijl> MK: there’s a few things that we’d like to cover
<ZoeBijl> but the ultimate goal of this is for me to walk away with some notion of whether or not the group supports some new more end user friendly approaches to helping web developers learn about accessibility from W3C resources.
<ZoeBijl> HTML and potentially WCAG
<ZoeBijl> But let’s discuss it among our selves frist
<ZoeBijl> We need to know what this groups wants to support
<ZoeBijl> I’m thinking about the multi year picture
<ZoeBijl> But not less than two
<ZoeBijl> But preferably a lot longer
<ZoeBijl> So I want to talk about the problem
<ZoeBijl> before we do that tho
<ZoeBijl> I want to make sure everyone is up to date
<ZoeBijl> particular the apg and the ARIA AT
<ZoeBijl> s/apg/APG/
<ZoeBijl> APG does not yet explain all of ARIA
<ZoeBijl> all of the new ARIA 1.2 stuff will be in APG 1.2
<ZoeBijl> we have a new “role” coverage
<ZoeBijl> Same for states and properties
<ZoeBijl> So we can better see what we have covered
<ZoeBijl> s/see/explain/
<ZoeBijl> But when it comes to guidance with example we have long way to go
<ZoeBijl> So there’s a lot of work to do on the APG itself
<ZoeBijl> My original goal is to close that gap by the end of the year
<ZoeBijl> It’s realistic to think we can achieve this by the end of next year
<ZoeBijl> especially with the help of Boaz’ team
<ZoeBijl> have a more standard PR review process
<ZoeBijl> trying to beef up that APG TF operates
<ZoeBijl> so it can fulfill its mission
<ZoeBijl> A big problem with the APG is that it does not help people make stuff that works with all the bugs that exist in all the AT and browsers out there.
<ZoeBijl> It’s not a component library you can just grab components from and drop in your project
<ZoeBijl> It’s a resource of how ARIA should be used
<ZoeBijl> if AT and browsers didn’t have any bugs it would be a component library
<ZoeBijl> If someone comes to the APG and thinks they can use it without testing it has a bad influence on the end users experience
<ZoeBijl> providing support tables is not in the scope of the APG
<ZoeBijl> That’s why we now have the ARIA AT TF
<ZoeBijl> s/TF/CG/
<ZoeBijl> Its goal being to provide support tables or a supported score that would give some indication of how well supported these examples are
<ZoeBijl> I’d like to give Val some time to discuss our progress
<ZoeBijl> Val: we joined in because of our background in testing
<ZoeBijl> we have two goals
<ZoeBijl> one is to design a test suite
<ZoeBijl> for the APG and ARIA in general
<ZoeBijl> designing a test suite is a big task because there are a lot of implications and difficulties
<ZoeBijl> so we need a way to write test that can be understood by users and AT
<ZoeBijl> We also need a document that describes how AT should behave
<ZoeBijl> Such a document does not exist
<ZoeBijl> So we’re also talking to AT companies to create such a thing
<ZoeBijl> Designing a test suitte is a huge bulk fo the work
<ZoeBijl> s/suitte/suite/
<ZoeBijl> As it is right now we’re breaking down the design patterns in the APG
<ZoeBijl> into a bunch of expectations
<ZoeBijl> The second goals is designing a test harness that assert these tests to manual testers
<ZoeBijl> We’re not sure how much time and how frequent we can test
<ZoeBijl> Bocoup and Facebook are working on this together
<ZoeBijl> we’re trying to lock down the design for this test suite
<ZoeBijl> In november we’ll build a prototype
<ZoeBijl> that’s the timeline right now
<Jemma_> Timeline is finishing the design by the end of Oct, and delivering the product by the end of November.
<Jemma_> +q
<jamesn> zakim, open queue
<Zakim> ok, jamesn, the speaker queue is open
<jamesn> q+ Jemma
<jamesn> ack Je
<ZoeBijl> JK: Part of my question was answered already
<ZoeBijl> it was about timeline and the goals
<ZoeBijl> I think it’s pretty strict
<ZoeBijl> The last AT meeting I joined there was a table of AT and browser combinations
<ZoeBijl> What does the final product look like?
<ZoeBijl> MK: End of november is only the test harness
<ZoeBijl> Val: yeah, it’s a prototype
<ZoeBijl> We might start to record results
<ZoeBijl> we won’t have a full suite
<ZoeBijl> we want to have some initial tests
<ZoeBijl> but it won’t have enough for a full suite
<ZoeBijl> it won’t be robust enough
<ZoeBijl> we’ll have to run test first
<ZoeBijl> MK: This is an exploratory prototype
<ZoeBijl> To see what kind of issues we’ll run into and how we can fix those
<ZoeBijl> JN: What other kind of documents (other than APG and ARIA AT) do you have in mind?
<ZoeBijl> MK: There are a lot of different W3C resources related to helping a web dev make stuff accessible
<ZoeBijl> some is almost duplicated
<ZoeBijl> but some are maintained by different group
<Jemma_> s/strick/tight
<ZoeBijl> s/group/groups/
<Jemma_> s/strict/tight
<Jemma_> https://www.w3.org/TR/using-aria/
<Jemma_> https://www.w3.org/TR/html-aria/
<ZoeBijl> the list of resources, the big buckets, are APG, ARIA AT both from ARIA WG, Using ARIA (from the WPWG),
<ZoeBijl> ARIA in HTML (also from WPWG)
<Jemma_> I think we should define what we meant by "explanitory resources" precisely.
<ZoeBijl> WAI Tutorials
<ZoeBijl> from EOWG
<ZoeBijl> not sure to what extent that covers WCAG techniques
<ZoeBijl> they’re ment to be explanitory resources
<Jemma_> +q
<ZoeBijl> The problem I see with all these resourcesw
<ZoeBijl> there’s important information in all of them
<ZoeBijl> they need to be known by the same people
<ZoeBijl> from our perspective we can say that’s your groups scope, this is our goup’s scope, etc
<ZoeBijl> clear for us at the W3C, but not from the outside
<ZoeBijl> there’s confusion outside of the W3C.
<ZoeBijl> Sometimes the resources say different things.
<ZoeBijl> We need to look at how we can serve the community
<ZoeBijl> that best servers their purposes
<ZoeBijl> maybe there should be a community group that sucks up all of these resources
<ZoeBijl> that forms a format that can combine all of them
<ZoeBijl> I like how the WAI tutorials are represented on their own site
<ZoeBijl> Why not have something similar for all of these resources
<ZoeBijl> That’s the question I would like to put forward to the group
<ZoeBijl> JN: I agree that it should be a website
<ZoeBijl> Boaz: it should be like a bootstrap thing
<ZoeBijl> MK: I would love fort there to be tutorials that explain our choices for a design pattern step by step
<zcorpan> scribenick: zcorpan
<zcorpan> jamesn: as soon as you take accessibility practices
<zcorpan> jamesn: you take away aria practices, since aria may not be the best way to solve a11y problems
<zcorpan> Matt_King: encapsulating all of aria in apg is part of the scope of the new rename accessibility practices
<zcorpan> Matt_King: debate of the name
<zcorpan> Matt_King: and scope
<zcorpan> Matt_King: if we rebrand scope, tutorial is not just about aria
<zcorpan> Boaz: accessible practices
<zcorpan> jamesn: so that is beyond scope of this group
<zcorpan> jamesn: if our pages are in a similar.. .can be integrated with theirs
<zcorpan> Boaz: what's your preference for merging these things into a more coherent resource
<zcorpan> jamesn: if they merge we can't own them
<zcorpan> Matt_King: there are joint task forces
<zcorpan> jamesn: there could be something like that
<zcorpan> Matt_King: joint TF can have as part of its scope explaining the aria spec
<zcorpan> jamesn: yes
<zcorpan> Matt_King: it would have to have aria representation, aria wg would be a stake holder
<zcorpan> Boaz: no matter how governance is in terms of ownership and scope, we need to talk to web platform wg or whatever about do we want to delete "using aria"
<zcorpan> Matt_King: if it was mdn there would be a governance issue there too
<zcorpan> Boaz: maube there could be a way to have that scope in APG, done by aria, in a way that can be consumed by e.g. MDN
<zcorpan> Matt_King: next step in exploration?
<Jemma_> rrsagent, make minutes
<RRSAgent> I have made the request to generate https://www.w3.org/2019/09/16-aria-minutes.html Jemma_
<ZoeBijl> rrsagent, close minutes
<RRSAgent> I'm logging. I don't understand 'close minutes', ZoeBijl. Try /msg RRSAgent help
<ZoeBijl> rrsagent, stop logging
<RRSAgent> I'm logging. I don't understand 'stop logging', ZoeBijl. Try /msg RRSAgent help
<ZoeBijl> rrsagent, go away
<RRSAgent> I'm logging. I don't understand 'go away', ZoeBijl. Try /msg RRSAgent help
<ZoeBijl> rrsagent, leave
<RRSAgent> I see no action items
<BGaraventa> I'm going to bed
<sarah_higley> bye everyone :)

[cookiecrook]

Since aria-braillereoledescription is being tracked as a PR, I'm duplicating my comments from the open PR #924.

Comment 1
At TPAC we agreed that this should not require (and should actively discourage) using "characters from the Unicode Braille Patterns block" because, in almost all instances, the web author will not know what these should be, even if they are an expert in braille.

As the simplest example, the web author has no way of knowing whether the user reads uncontracted (Grade 1) or contracted (Grade 2) braille, so this attribute would be formatted as a non-braille unicode string. For example, the screen reader could translate aria-roledescription= "denomination" aria-brailleroledescription="den" into the user's preferred Grade 1 (⠙⠊⠝) or Grade 2 (⠙⠔) braille rendering.

Other examples author-encoded braille strings include when a user is fluent in multiple languages, but only fluent in one variant of braille. For example, a member of my team is fluent in reading, typing, and speaking English, but only reads braille using his native German braille table. VoiceOver can provide that to him, but a web author cannot.

Comment 2
We also agreed that it may be okay to allow unicode braille with sufficient warnings, and no examples of this usage. The only context I can think of where this is necessary is testing, where a user may be learning tactile braille through remote instruction. E.g. a student learning the differences between US English versus UEB, in Grade 1 or 2.

Comment 3
I am skeptical of suggesting the unicode braille block is necessary even in the case of Math polyfills like MathJax. For the same reasons listed above, the polyfill can't know whether the user prefers to read Math in Nemeth, UEB Math, or Computer Braille.