W3C

Accessible Rich Internet Applications Working Group Teleconference

25 May 2017

See also: IRC log

Attendees

Present
tzviya, BogdanBrinza, jkva, Rich_Schwerdtfeger, janina, MichaelC
Regrets
Chair
Rich
Scribe
jkva

Contents


<Rich> scribenick: jkva

Dpub AM went to Candidate recommendation

Rich: Signed off by Joanie after CfC
... Another thing that happened is multiple platform Dpub candidate recommendation

DPub ARIA itself

<Rich> https://github.com/mattgarrish/test-results/blob/ccff0e25121d8c928b890a23c38aece0a1377c28/dpub-aria/README.md

Rich: Posts current implementation report
... Is this like self-certification? Most systems are closed. What is the director's position?

Tzviya: Based on earlier discussions in September, this would be ok

MCooper: Asked Ralph for information. Fallback to Epub type is correct

Tzviya: Not sure what other implementers do, just asked for sample files

MCooper: ASked how assertions are backed up, what if there's nothing to test for these tools? Loop we want to close.
... Not explicit in CR. Is table good enough? Yep. Consumer data is available would be strong + to include it.
... 1) if Consumer data available: define it. 2) Should there be [?] on tool use for Dpub rules?

Tzviya: In terms of Consumer use, results for DPub am pretty much.

MCooper: Ralph might not have had a clear distinction in mind.
... Ralph wondered if AT had something else in mind for DPub reading

Tzviya: Could get in touch with Peter Krautzberger

Rich: Tzviya made point on list that DPub-ARIA not in current DPub spec. Is this hybrid version of DPub?

TZviya: I answered this on the list. I'll read to you. 1) This doesn't matter, not vocab exclusively for epub. Some authors have adopted for things other than Dpub.

<tzviya> https://w3c.github.io/dpubwg-charter/

Tzviya: One can use any ARIA, also Epub-ARIA, in any spec.
... As long as it's valid, it's acceptable to Epub. Matter of adoption. Just not required.

<tzviya> s/epub-aria/dpub-aria

Rich: To be EPub-conformant, you don't have to support ARIA?

Tzviya: Correct, you don't *have* to include ARIA, but it can.

Rich: May take a while to get support going
... Why I asked what is the timeline for epub4.

JGundar: Even though it's valid, you can have multiple features, run it through multiple validators - use best practices in their documents. Enhancements can be added to tools and features and prompt users to add ARIA features.

SVG

Rich: Is Amelia on the call? (no)

Amelia was sharing the task force for SVG a11y. It got put on hold for the SVG charter (?). There is a new SVG charter. New SVG version has ARIA-support. No co-production, limited to WG. They have agreed to provide liaison with SVG TF.

Do people agree with this solely being an ARIA deliverable?

Doesn't personally see an issue with this. Has worked with Amelia on accessibility work including next version of SVG.

When we're working on releasing SVG versions/modules, always duplicate work. Botg groups overlap because they own stuff in repsective charters. Challenge is getting enough workforce.

Not sure what Amelia's status is on this

MCooper: My concern is: joint deliverable means both WG have obligation to deliverance.

If only ARIA owns is - we are ones held responsible and yet don't believe we have expertise on graphics a11y rec. We need expertise from other WG. Charter inticates willingness but not explicitly required.

Rich: Amelia is member of WG?

We have resource issues. We have some skills. Amelia is really good and has added a lot of stuff.

MCooper: If she had said conflicts with her dayjob meant she had to go back

Rich: Honestly a lot of it is already done. From webkit perspective, Joanie said a lot of stuff done (?)

We need more than 1 person whom knows graphics to work on this.

We need to see who can work on this.

Biggest concern is workforce. Shepazu left. $self retired.

MCooper: Ensure people consider this and that chairs / WG ok on this.

Rich: As long as we get more than 1 resource to work on this, we in better shape

Maybe Microsoft will work on this with us. Firefox unlikely

We need to make sure that we can get enough people.

If we don't, we have to consider if we can deliver it in our charter.

MCooper: Currently in our charter. Remove at next rechartering and why

Rich: How long? 1 year?

MCooper: 31 July 2018 expiry

I should reply to CHris saying, ARIA WG is accepting deliverable? Maybe talk with Saleep (?)

Rich: Fair, we're short on people. Serious problem. If liaison can help, would be great.

Still mainly resource issue. Have other specs to get out door. Get started on ARIA1.2 too.

MCooper: Need to chat about timeline. Will accept deliverable be dropped from SVG.

Rich: Doubt we can take it on. Challenge to get SVG WG to deliver something. To do so they removed functionality.

<AmeliaBR> Sorry, joining in a second

MCooper: SVG WG trouble implementing features. Scaling back to version of SVG that is hoped for rather than what it should be.

<AmeliaBR> ...if someone can DM me the password

JGundar: Need update for pyia2 and MSAA2?

Rich: Taling about issue where SVG doesn't want to publish ARIA modules and wants to prevent liaison. We are worried about staffing. Insight?

Amelia: SVG working group is uncharted and inactive. Concern about taking on too many commitments. Don't have human resources being allocated by member org. Also focus on getting one specific thing done, which is getting SVG into recommended status. So that;s where they are coming from. As far as the task force and specs, and what makes sense, the task force

was doing a number of specs, some of which are svg specific, and some of wich are just accessible graphics spec which are relevant to SVG a11y but not exclusive. SVG-AAM gets default mapping to [attributes]. Any time new functionality is added to SVG spec, you need to update the SVG-AAM to describe how the new features hould be accessibly represented. The

other spec is the graphics roles which are about creating a language for describing twodimensional visual information in a way that assistive tech can build on. Again it's something you need to use to make svg accessible but its not exclusive to svg. Idea was to have svg rolenames that can then be used on other languages to then communicate to AT that you're

looking at the same structure, regardless of what technology that is used to paint the graphic on the screen. The AAM relies on the ROLES but not vice-versa.

Rich: Limiting for SVG2, right?

Amelia: Yes.

Rich: Doesn't consider issue with having liaison and having it be work for ARIA-WG, but considered about resources.

We need to address that soon, so we can get perspective for the domain elitist (?) on what we can complete.

Amelia: resource issues on both sides.

One way or the other, need commitment on both sides for member resources to work on the spec, build test suites and implementations. If we don't have commitment, the WG irrelevant. No work would get done.

Rich: Can imagine some features would be removed, and influence specs.

Amelia: No big impact on a11y side. Features aren't ones impacting a11y. Graphical effects or text layout. Real a11y problems relate to older more established svg features. Designed in a way to think about a11y and DOM.

Rich: Question is: What's your availability? Know you have other oblig.

Do you have time to work on this?

Amelia: Not in short term. Longer term if I have funding.

Currently no funding. As volunteer - as triage person/resource.

Keep other people organised.

Won't be doing significant work w/o sponsorship.

Rich: That creates another level of risk for us. Def, need more headcount.

MichaelC: Will take on in chair planning. Will mention risk exists and is kept track of.

Rich: Thanks Amelia

Amelia: Which group takes resp: as long as there is communication on AAM, it can be on either side. Roles should be shared or ARIA-only. SVG does not make sense.

Main issue remains manpower over which WG.

UIA issues

Rich: Issues with mapping, controltype consistency

Bogdan: Actually Jon Jensen (?) responsible for driving UIA testing + mapping. Will contact them

Inconsistency in style, casing, spacing. Not funcional inconsist.

Will make sure John is involved.

pyia2 update

Jon: Driving through mountains

WOrking on assertions

Communication with WPT, interfaces

Worked out some things with Shane and Joanie related to unique situations and not all test has MSAA mapping or IA2

Basically will have 2 ATTA, one for IA2, one for MSAA

Few technical issues. Working with Shane + Joanie. Talked to Jamie Teh and (?)

Rich: Need to coordinate getting defects filed against browsers

(?) looking at chrome defects

Alex has been assigned Firefox

Not sure how quickly they'll get on it

Jon: Once they run tests, Joanie says they'll run into issues

API discrepancies will cause issues

Joanie said not so bad, she fixed a lot of stuff.

Tzviya: Is testing including Dpub-AAM (?)

[Did not catch answer]

Rich: All of Dpub-ARIA in one example
... anything else?

MCooper: No.

Rich: Have to address resource on SVG

Worse than expected

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/05/25 17:47:18 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Succeeded: s/[?]/Ralph/
FAILED: s/epub-aria/dpub-aria/
Succeeded: s/MCooper/Rich/
Succeeded: s/same structure/same structure, regardless of what technology/
Found embedded ScribeOptions:  -final

*** RESTARTING DUE TO EMBEDDED OPTIONS ***

FAILED: s/epub-aria/dpub-aria/
Present: tzviya BogdanBrinza jkva Rich_Schwerdtfeger janina MichaelC
Found ScribeNick: jkva
Inferring Scribes: jkva
Found Date: 25 May 2017
Guessing minutes URL: http://www.w3.org/2017/05/25-aria-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]