– DRAFT –
AGWG Teleconference

<Rachael> Looking for a scribe

New members and topics

Racheal notes light attendance, could be M-Enabling in DC/VA

Rachael calls for introductions, new topics

Olivia Hogan-Stark: From NCR, first AG meeting, interested in WCAG2ICT, educator background

Announcements

WCAG 2.2 retrospective https://www.w3.org/2002/09/wbs/35422/wcag22_retro/results

Two questions from Survey

things that went well

Rachael reads survey comments, invite comment.

Racheal reads Gundala comment, Gundala not online.

Rachael reads Bruce's survey comment. Bruce has no additional comment.

Mike Gower and Detleve comments read, no further comments.

Rachael open floor.

Things that could be improved

Rachael reads Wilco comment. Process is fine, but we are cutting corners a bit. Pressure to get to rec...

Rachael reads Gundala comment, no additional comment.

Rachael reads Bruce's comment, no follow up

Rachael reads Michael Gower comment.

MG recommends final push as group. No addition comment

Rachael reads Detlev comment and invites further comment. (Pass)

Rachael: Additional comments as to what we might improve as group?

Mike: I found Wilco comments interesting, but I would be concerned from removing timeline...
… i would be concerned for due diligence. I agree feels rushed at end.

<mbgower> +1 to juggling comments from Rachael

Rachael: Chair hat off, it was harder than expected to juggle between 3 and 2

Wilco: I feel like we made decision to meet deadline instead of making the spec as good as it could be...
… it is not like software where we can just release another version. If we miss CFC, we should address rather than push on.

Rachael asks for further feedback on how we might improve.

Francis Storr: GitHub issue management is a challenge...
… More dialog that getting closure on edits. Some are years old. I tried to close, but with things with typos it would have been so much faster to have PRs rather than issue threads...
… Now we are getting churn about people asking about old comments or even no-so-old issues...
… i don't have good solution, but I think we may need more process.

<ShawnT> +1 to Francis_Storr github management

<Wilco> +1

<Rachael> Key areas: Prioritize process over timeline to create high quality spec, Improve github management particularly small fixes, Need more people addressing github, 1 spec at a time

Rachael: Hearing some key issues, which I will summarize. Did I miss any?

<Rachael> , continue culture and conversation facilitation

Wilco: More generally, Francis comment on resources available to group. Understanding documents are an example...

<Zakim> mbgower, you wanted to say we need to take re-chartering time into account

Wilco: there is way more work than we can hope to do with a group of this size.

MikeGower: I note that the re-charter process pulled attention away from actually doing the work...
… with an 18 month charter, for next time, we should add padding time to provide opportunity on charter.

Suzanne: I am new to AGWG group, and from that perspective recent joining these calls, it had seemed to me that there was plenty of people doing the work...

<kirkwood> good point about the optics

<mbgower> +1 to Suzann's comments on idea of communicating need and opportunity to prospective participants

Suzanne: from that perspective it might help to spread the word, at conferences or the like, what is going on and why volunteers are needed.

<ShawnT> the work that WCAG2ICT is doing just that

Francis: I think many people find GitHub intimidating, so perhaps regular orientations would be helpful

<ShawnT> no we didn't record it

<Rachael> Bruce: Adding to Francis about Github being intimidating. How we work on pull requests is fine but realistically if we want more people to help curate the understanding documnets, we have to have something less learning curve than what WCAG2ICT has been doing.

<Rachael> ...I was disappointed how much harder WCAG2ICT is. I want it easier than what we have. Preview, most writing in mark down. I've had good progress in my office with non-geeks and I don't think we will do that with the current set up

<kirkwood> +1 to Bruce’s point

PR to Decision policy

Rachael: Please be encouraged to reach out to chairs with further comments.

<Rachael> https://github.com/w3c/wcag/pull/2739/files

Rachael: Leadership did have some meetings on this, and we are drafting a PR to decision tree.

<Rachael> https://docs.google.com/document/d/1tb04rdRRB5b1oGYwme3Z9EyoLgXhiPhyx6MSN4St0_8/edit

Rachael: We also have Google Doc for collecting issues.

We did have some comments on PR

Rachael reads Wilco comments, no further comment.

Rachael reads Gundula comment.

Rachael: Bruce and Gregg approved with no comment.

Rachael reads Mike Gower comment, no further comment.

Rachael reads Jeanne's comment.

Jeanne: Gundala edit address my concern.

<GreggVan> Just a note to chairs -- I am at a conference - and have a booth - so will need to jump off call at noon

Rachael reads Laura and Detlev survey comment, no addition al comment.

<Rachael> When a decision is made with objections, the Chairs will provide the opportunity to propose an alternative based on the wording that was decided based on consensus with objections, and have the Working Group review the alternative.

<Rachael> If the Chairs believe that the received objection(s) present substantial new information or if the Chairs believe there is not a clear consensus in the Working Group, they will reopen the discussion, as detailed in section 3.3.4 of the Process Document (Reopening a Decision When Presented With New Information). A new alternative versionwould be considered substantive new information.

Rachael, hearing no objections to the edits proposed in the survey, I summarize proposal in IRC.

Racheal: This will update/replace the edit currently in PR...

<GreggVan> +1 with any editorial for readability

Racheal: I feel like the opening leaves some room for improvement. But any concerns?

<Rachael> When a decision is made with objections, the Chairs will provide the opportunity to propose an alternative based on the wording that was decided based with objections, and have the Working Group review the alternative.

<jaunita_george> +1

<ShawnT> +1

<Detlev> +1

<maryjom> +1

<laura> +1

<Wilco> +1

<GreggVan> +1

<JakeAbma> +1

<GN015> +1

Rachael: Any objections to moving forward, pending some light word smithing?

<jeanne> +1

<Rachael> draft RESOLUTION: Add PR 2739 with revisions and slight rewording of first sentence

<Raf> +1

Rachael, please vote -1 if concerns

RESOLUTION: Add PR 2739 with revisions and slight rewording of first sentence

Accessibility Supported Subgroup https://www.w3.org/2002/09/wbs/35422/access-support-subgroup-oct22/results

Rachael: We are looking to stand up a follow-up from the Accessibility Supported Subgroup
… this was an action item from TPAC.

<Rachael> bruce_bailey: I am willing to consider it in January

Bruce willling to consider, but not now, maybe January

Janita possibility, but later

<Francis_Storr> Also interested but in January

Wilco concures on post-poning at this time.

Subgroup weekly updates (5 minutes or less each)

Rachael @Makokoto please reach out to chairs

Rachael: We have three sub groups at present
… reportsing?

Jeanne: Silver subgroup has been working on writing up goals...
… we were tasked out of TPAC to experiment with writing Outcomes as User Needs. We have agreed on an approach to use the existing WCAG3 sample Guidleines as examples.
… Alt Text is first activitiy, and discussions revolve around level of granularity of the user needs best serves

Jeanne: We have not concluded discussiton.

Fancais reporting on Issue Severity (2nd iteration of subgroup)

Francais: We have identified 60 functional needs and related tests

Francais: Working on groupings. When we met last week, we decided we needed some more information about categories and some of the functional needs.
… I am working with Josh O'Conner to work on the grouping and what some of the functional needs are.
… some of the functional needs are not in groups, so we are giving those some additional attention.
… Another important issue is the WCAG3 idea of critical errors, so we want to integrate that work.

Jaunita: Working on method template for testing.

<Rachael> Test requirements as methods subgroup

Testing requirements method subgroup

Review and discuss possible conformance models https://drive.google.com/drive/u/0/folders/1X3Paz3WuK4yn09_ZN99P5IFl2-5yn5U9

Rachael: Following up on conformance model discussion from last week, we have a couple more.

Option 4

Rachael: link is to Google Drive folder with the several draft option proposals.

Suzanne: Briefing on Option 4, Left Shift, Badges, Silver is Quality, Gold is More

Suzanne: Few of the ideas in this proposal are brand new, this option is integrating a few things.
… lots of ideas from other other conformance modesl

Option 4 draft: https://docs.google.com/document/d/1W7g9a5b3gQZ2aPEivl0KJ48K3sMg6eegtMiCwkw-eaU/edit

Accessibility audit SME can review assertions from template, looking at claims from testing methods claimed to follow.
… We are left-shifting by identifying applicable tests from templates at start of review process
… each outcome might have several methods which are applicable, but audit looks for those identified by product owner
… likewise Outcomes are written from perspective of product owner...
… this allows for feature of site to be identified. Example is site using audio tones to convey information...
… that approach is unusual enough that a naive audit might overlook easily.

Suszann: Bronze is computation test and most straight forward qualitive tests.

Silver is badges for many achievements. A bronze website would certainly pick up a few, but silver is all or a bunch of the badges

Suzanne: Gold badges are for innovation. This means a site could be Bronze but with a couple Gold innovation badges.

<Zakim> jeanne, you wanted to talk about large complex products

Suzanne: idea is that Badges are straightforward as compared to point or accumulated score.

Jeanne: Many parts of this that I like, especially Gold for innovation is very creative...
… that said I have some concerns for larger site and the scoring from methods. A large site will cover the gaumet of just about all possible methods...
… but I want read proposal more careful.

<Zakim> mbgower, you wanted to say my concern with the Overall score slide is that almost no one meets the equivalent of bronze now

Suzanne: Sites can also pursue more conventional path.

Mike Gower: It seems complicated, and I like the innovation, but I think the bigger challenge is for sites getting to Bronze, a baseline...
… so focus on innovation seems like it puts emphasis where it is not needed.

Alastair: Asking sites to pick methods/outcomes in advance seems like it could limit how sites approach accessibility...
… I also want to comment on Mike's approach with Bronze not currently being obtainable, but it could be. For example, rely more on UA, e.g. for contrast, Bronze should be relatively attainable, with Silver as the 'standard' level for legislators.....

<mbgower> we totally need a way to let good and great sites and apps differentiate

Alastair: if Silver is more like AA, that might influence regulators to adopt that as minimum.

Option 6

Jeanne: Reporting on Option 6, taking a little different approach, being a repository of some of the other ideas...

<kirkwood> +1 to Alastair regarding AA and policy takeup

Jeanne: Option 6 tries to incorporate Options 1-5 while address and integrating from GitHub issues, and discussions from TPAC....
… recently I pulled in suggestions from Greggs proposal last week, the way he included adjectival ratings and scoring from that, which are responsive to issues raised by FCPWD...
… I also addressed some issues following protocol group and discussion.

jeanne: option 6: https://docs.google.com/document/d/1gyT0F3ggr0kdFhG-ugmySA49gTQF2B5SPFZuEHd7MQI/edit#heading=h.q7a1p1s14gm5

jeanne: user needs as placeholder for outcomes

<bruce_bailey> ED draft with changes in response to issues

jeanne: no changes to guideline, conformance, incorporate work of other proposals, protocols
… noting source
… type of tests, tests, added examples, incorporate automation
… knows there are big problems, aim to support automation without sacrificing quality of spec
… will continue to develop with more good ideas

<jeanne> Possible COnformance Models

<Rachael> acl mbgower

mbgower: Automated testing, establishing baseline and beyond, how much excess from baseline anticipate with automated testing

jeanne: Challenge with automatable, constantly changing
… can't do automatable level, write what we're doing precisely, as AI improves, can have more qual testing in automated
… improve how we write things
… integrate more into who, not break apart
… can have goal to make more automatable, directions in AI automation, work toward it
… likes Suzanne's proposal, could put more automatable at Bronze, Silver to do qualitative

<Zakim> mbgower, you wanted to say I'm not sure I agree that automated testing is changing that rapidly

mbgower: Automated can help with baseline, not perception it's changing that much
… people come up with rules, test parts of criteria, make them repeatable, script into engine, not AI, rule implementation

Wilco: Whatever testing automatically tech specific, e.g., HTML and PDF varies
… automation is advancing but not that rapidly, some gains
… as technologies evolve, things we can test become more complicated
… could get worse over time, dynamic field

<jeanne> +1

<alastairc> Which goes to Jeanne's point that it's difficult to do a level for that

<mbgower> which isn't to say that a repeatable baseline isn't something to aggressively pursue. I LIKE the idea of an automated baseline

<jeanne> I think it is very dynamic and hopefully advancing

Option 5 details: https://docs.google.com/document/d/1wBhwYhOs-3T27rftT3IXbpuIdiphYkoQJS9NFUGR870/edit#heading=h.q7a1p1s14gm5

Rachael: [reads slide and notes]

Option 5 slide: https://docs.google.com/presentation/d/15ZoKbczXw3JIoyDxAxKtBG0sWMKnB6lqAnV4V9xVsoM/edit#slide=id.g165c944dd8c_1_0

<Zakim> mbgower, you wanted to say I don't see how caption quality can be at bronze

mbgower: Useful breakdown of examples
… captions are accurate, anytime we're talking about quality difficult to put at bronze, discussion about quality attributes
… captions exist, UA displays them, yes/no can get us to a baseline

<kirkwood> were the captions writter or verifed by the author? would be a question i’d rather see.

GN015: Accuracy, automated caption accuracy 91–92%, could set minimum
… higher level comes with closer review and higher rating

<Zakim> bruce_bailey, you wanted to say i do know what 95% accurate caption is

bruce_bailey: Don't know what 95% accurate is, FCC has usable approaches

kirkwood: re captions and alternative text, part of editorial process, author writes captions, verifies content owner
… automation can be fairly accurate, but author provided more accurate

Wilco: Interesting on captioning, orgs have guidelines, wonder whether WCAG should have guidelines, instead treat as protocol

<jaunita_george> +1'

Wilco: you decide standard and follow those

<bruce_bailey> FCC closed caption rules (adjectival): https://www.fcc.gov/consumers/guides/closed-captioning-television

mbgower: Lot of videos without captions, some with captions but player doesn't display them
… requirements about caption button
… reviewing automated, accuracy, qualitative discussion, exceed bronze
… whether protocols, manual reviews, get extra for that

<Zakim> Rachael, you wanted to shift conversation

Rachael: Value in captions but don't what to go too far into any examples, instead talk about conformance options
… overlap, interplay, combinations
… which to explore, how to move forward
… tension, number of people, time
… see where people stand on options
… if clear direction, go with it; if all over, have another discussion later

<Rachael> https://docs.google.com/presentation/d/15ZoKbczXw3JIoyDxAxKtBG0sWMKnB6lqAnV4V9xVsoM/edit#slide=id.g165c944dd8c_3_0

mbgower: Slide shown with caption example, make sure it's in the doc

Rachael: All proposal in slides

a/proposals/proposals

Rachael: [walks through concepts using slides]

Rachael: Clear?

<Wilco> +1 I don't think I can do this on the fly

<jaunita_george> +1

<SuzanneTaylor> +1

<joweismantel> +1

bruce_bailey: Need to do offline in survey, look at them, rank them, good to discuss but hard to do live

Rachael: Set up survey, pick and prioritize and different ways of combining
… set up in Google drive, started process of putting into steps
… evaluating conformance models worksheet
… will break out different approaches and talk through pros/cons

<Rachael> https://docs.google.com/spreadsheets/d/1iCLpvuDbulZLGLflaxUjFtlOMeRvlDZBwrIQf82Sy_c/edit#gid=0

Rachael: come from survey with obvious way forward, otherwise go through breakout step-by-step

Wilco: Other criteria to consider maybe

Evaluation criteria

<mbgower> +1 to look at upgrades to 2.x criteria

Wilco: updating standard is a struggle, can we build flexibility into standard
… rollout, WCAG3 looks like a decade to completion, should conformance allow rollout in stages
… not sure things to look at but idea

Rachael: related criteria to updating
… staging, will add

mbgower: Thanks for sharing spreadsheet, useful to see stages
… was idea to use 2.x to new conformance model, take existing criteria and put into new Silver model

<bruce_bailey> yes, +1 thanks for all the thoughtful work that has already gone into these options!

<Wilco> Suggestion: Is there a migration path from WCAG 2 to the new conformance model?

mbgower: work on criteria with different ways of reporting, assessing

Rachael: migration path, next step after this, captured in charter

mbgower: see trying to address in stages, chicken/egg, until take criteria hard to assess without applying
… work with familiar things makes it easier to prove it

Rachael: Stages meant to do, basic things, stage 2 would need guidelines/outcomes/methods, stage 3 would have to be far along
… or take WCAG 2.2. and apply it
… meant to be way to stage moving forward

GreggVan: Good to list path as criteria
… looking at things without criteria, ideas work on some SCs but not others
… not practice to do with everything but useful to identify spectrum of SCs, different, most challenging, ask authors to show implementation

<mbgower> I would be happy to take part in coming up with a list that forms a cross-section of WCAG 2.x criteria

<alastairc> That is where the current ones in the FPWD came from.

GreggVan: gives way to see what authors mean, gives authors chance to try out

Rachael: FPWD includes that type of examples
… other criteria?

Wilco: Plans for addressing open WCAG 2.2. issues

<mbgower> can you repeat that question, wilco?

bruce_bailey: Dimensions to rate options

<Rachael> https://docs.google.com/presentation/d/1yLYeNcybGxRu43KdrVUcOCL6iXsy6-gxl9-lbyr90dI/edit#slide=id.p

Rachael: [reads criteria]

<bruce_bailey> Listed dimensions are very good!

<mbgower> which slide are you reading?

<Rachael> https://docs.google.com/spreadsheets/d/1iCLpvuDbulZLGLflaxUjFtlOMeRvlDZBwrIQf82Sy_c/edit#gid=0

<Zakim> bruce_bailey, you wanted to brainstorm for 5 min and to

bruce_bailey: Rephrase as positives

<Zakim> SuzanneTaylor, you wanted to suggest "Does the guidance motivate *more* organization to work toward minimal accessibility/requirements?"

SuzanneTaylor: Motivating *more* organizations

mbgower: Is there one about motivating people to tackle accessibility, do accessibility

<kirkwood> Is the approach clear and understandable?

mbgower: can infer but

mbgower: Row 3 and 15, difference?

Rachael: How to make more clear

<Rachael> Does it use other ways of measuring and/or evaluating where appropriate so that more needs of people with disabilities may be included?

<bruce_bailey> +1 to MG to split into two

<Rachael> Does the approach expand the number of functional needs addressed

mbgower: One is expand functional needs addressed

<Rachael> Does it expand beyond strict pass/fail testing?

Rachael: Expand ways of evaluating?

<kirkwood> present equivalent alternate ways to pass? of access?

mbgower: Some suggestions are solutions but criteria written as outcomes, equivalent is way of achieving

<Rachael> Strawpoll: Is everyone ok with replacing ways to measure with expand number of functional needs and equity questions?

<kirkwood> +1

<ShawnT> +1

<joweismantel> +1

<Wilco> +1

<mbgower> +1

<bruce_bailey> +1 to expand functional needs

<Rachael> akc bruce_bailey

+1

<jaunita_george> +1

<Detlev> -1 difficult to understand...

bruce_bailey: Might be too soon to give up on other part, accessibility on continuum, going beyond pass/fail

Rachael: Yes, drop specific solution of going beyond, instead focus on result

Detlev: Wordy, jargony, number of categories is complex concept
… different ways of measuring seems fine

<mbgower> +1

<jaunita_george> +1

<Wilco> +1

Rachael: Add note about not expanding measurements

mbgower: Using same question about wording on row 4, written as solution, difficult to parse

<Rachael> Can the model be updated quickly and easily to support emerging technologies and interactions?

+1

What is the plan for addressing open comments from CR on WCAG 2.2?

<Seirdy> expanding beyond strict pass/fail could be esp. useful for color contrast, as there's a difference between e.g. body text and superscripts. A partial pass could meet contrast ratios for e.g. only body text. the experimental version of SAPC-APCA explores this more deeply.

alastairc: Trying to get implementation testing done this week, still waiting to hear back, then transform Friday meeting into TF, work through issues

<mbgower> I wonder if "is the proposed conformance model realistic" is potentially useful?

alastairc: haven't tackled prioritization, immediate deadlines first priority, then group topics and work through backlog
… survey about managing notifications to group, got good results, will follow that process, hasn't started yet

Wilco: Thought would be for maintenance, not for resolving comments on CR, will AG discuss feedback on CR

alastairc: Any changes will come back to AG, triage will be in taskforce

mbgower: Being realistic, propose amazing AI that does everything, approach isn't realistic

Rachael: Added criteria

<kirkwood> +1 realistic or “achievable”

Wilco: Will AG approve Taskforce?

<mbgower> thanks for all this work!!

Rachael: Trying to aim for 2–3 to dig into