<Chuck> meeting: AGWG-2022-10-04

<Fazio> scribe+Fazio

<alastairc> scribe:Fazio

chuck says hi :)

<AWK> +AWK

chuck: anyone new?

nope

chuck: new business?

nope

<Chuck> Design and Information Architecture of WCAG3

<scribe> New group of design experts will help design info architecture for wxag 3

WCAG 3

Update on subgroups

<jenniferS> I keep joining and just missing tantalizing bits. Would someone please reach out to me about the design topic that Jaunita offered names for? Thanks!

Wilco: starting 2, test requirements/methods, issue severity

<Chuck> JenniferS, I've noted it down. We'll reach out sometime after and outside of this call.

Wilco: survey for participants is out

Racael: 1 more call for participants should be sent

Rachael: Silver meting is correct place for discussing outcomes as user needs

Stand up temporary equity group https://www.w3.org/2002/09/wbs/35422/temp_equity_group

Chuck: survey for equity approach released. 1 - 2 more groups may form related to results

<Nicaise> Zakim present+

Chuck: short term group related to equity decided to be established at TPAC.
... survey comments - splitting efforts has risk
... survey comments - equity should be implemented in similar scope to APA

<Zakim> Rachael, you wanted to say that conversation should be part of the temporary group's converstion

Rachael: temp group should drive conversation

Continue conformance conversation:

<Chuck> https://docs.google.com/presentation/d/1yLYeNcybGxRu43KdrVUcOCL6iXsy6-gxl9-lbyr90dI/edit#slide=id.g15c4067867c_0_67

rachael: exploring 100% pass fail, as well as protocols
... next steps discuss ways to solve challenges. pick a couple of approaches to explore
... bronze, silver, gold, badges, all scoring approaches discussed
... key benefits and challenges, to each approach needs discussion

<Jennie> *Thank you for including definitions Rachael!

Rachael: any key themes missed?

<Chuck> Thanks Alastair

gregg: failures weighed the same implies maybe they shouldn't be. Needs to be reworded

<Chuck> +1

<Jennie> +1

<GreggVan> +1

<jenniferS> +1

<Makoto> +1

<Azlan> +1

<jeanne> +1

<GN015> -1

<Poornima> +1

<alastairc> The other method of weighting is via context, rather than between people

<Jennie> +1 to John K

Kirkwood: 100% pass fail benefit precedence of binary approach has long standing history an acceptance. Important to build on

<Chuck> +1 to Kirkwood and the addition into slides by Rachael

+1

Gundala: treating all req's same problematic because of nuances

<bruce_bailey_> FWIW, it was not clear (at least to me) what people are giving their +1 to exactly

Gregg: other standards are 100% pass fail, so benefit to keep in line with that approach

<ah_> Where does "partially supports" in acr fit into this?

<jeanne> -1 to Gregg's statement that 100% is the only way to do a standard

<Zakim> Chuck, you wanted to add to Gundula's comments

<bruce_bailey_> +1 to what GreggV is saying about WCAG being an example of a voluntary industry consensus standard

Gregg: a nuisance for some is blocker for others

<Zakim> maryjom, you wanted to say ITI wrote a position paper regarding binary pass/fail.

<maryjom> ITI's position paper's link: https://www.itic.org/documents/accessibility/ReportingConformancetoICTAccessibilityStandards-July2019-updatedMay2022.pdf

Mary Jo: ITI wrote position paper on binary pass fail for regulatory use and reporting a11y compliance

<bruce_bailey_> +1 to Rachael comment that conversation have been capturing main themes

Rachael: how can we explore challenges of pass fail?

<Chuck> Fazio: I recently got involved in "cost action", we are working on autism, it's a EU thing. One of our members is a former member of parliament. She gave us feedback on how to craft such that parliament would take these up. It's important to consider this so that legislation can take this up. I will include resources to this.

<Zakim> alastairc, you wanted to suggest two options: more granular (test level) pass/fail, or post-testing process to incorporate context.

Alastai: create lower level of pass fail. Issue severity can be process based eval in context

<Zakim> Rachael, you wanted to say that trying out approaches is the next step

laws reg's dtandards, are all different. Need to consider all 3 levels - Gregg

<jeanne> -1 to Gregg. I have personal knowlege of suits for a single error

<Lauriat> -1, I've seen examples of that as well

all or nothing concern. No one has ever been sued bc they didn't fully comply. Usually it's sites with serious noncompliance - gregg

in physical ADA yes, maybe not digital though

<Zakim> Jennie, you wanted to identify list of vocabulary in current SCs

<maryjom> -1, European Accessibility Act, as adopted by Germany, can actually remove products from the market if not fully compliant.

Jenny: build on Maturity of 2.x review vocab that makes pass fail more difficult. Things that require judgment calls

<mbgower> Greeg, it _does_ suggest that a clean VPAT is not accurate, which is also a problem

Rachael: 1 criteria is req for reg environment.

<Zakim> jeanne, you wanted to disagree with assertion about lawsuits

Jeanne: disagree with not being sued for just 1 error

<Zakim> Lauriat, you wanted to respond to Gregg as well from my experience

Shawn: avoiding legal risk by knitpicking bugs takes a lot of time

<Zakim> GreggVan, you wanted to ask if any product has ever been removed from the market in germany -- and if so whether it was because it missed one or a few

has any product ever been removed from market because it just missed 1 req

Maturity Model, Maturity Model, Maturity Model

<jeanne> Gregg, in my experience there are companies that run a black box accessibility checker and send out a suit for any error that is triggered by the automated checker.

<Zakim> Chuck, you wanted to say that the lawsuit conversation is a tangent topic that doesn't serve the purpose of this conversation at this time

these are central facts - gregg

<Zakim> mbgower, you wanted to relate this discussion to security industry guidance

<Lauriat> -1 to "nothing ever happened", it wasted lots of time and money

shawn - web apps complied with many req's but irrelevant bugs were referenced. Lawsuits arising were classified as frivilous

MGower: think about how a11y priority fits in. can we accommodate reality that a11y may not be utmost priority

Poornima: functional needs eval can help

<GreggVan> there are always nuisance suits -- and they are a pain - but not a reason to change the standards away from where they should be. In the end if they are found frivolous - then it was frivolous and an abuse of the standard. NOTHING we do will prevent nuisance

<alastairc> Security is another case where any single issue can have a large negactive impact, and they have a mechanism for priotisation.

<GreggVan> +1 to Wilco

<Lauriat> +1 to Wilco

<Chuck> +1 to Wilco

<Rachael> +1

<jeanne> +1 to Wilco's comments on the ruler

<alastairc> +1, we can provide more granularity

wilco: wcag 2 not a very good ruler for measurement. Granularity is needed

<Chuck> Fazio: First, I know that we are not centered on US legislation. ADA... Feds require that companies make the fix. We don't have digital compliance legislation, it fits in ADA.

<Chuck> Fazio: 2nd, the maturity model can address these issues. It's never done, but we have a continuous process of improvement. There's been interest from AG members in including this work. I think this may be the way forward to make this work.

<Chuck> Fazio: Identifying and communicating what companies are doing in their approach.

<Chuck> Fazio: I think maturity model is way forward.

<Wilco> Recap of my comment for the minutes: I think we should try to stay away from problems in legislation. Drive-by lawsuits and threats that products get pulled for 1 violation is not part of WCAG. It's not our problem to fix.

shadi: wilco's comment very insightful. think about granularity. don't rule out relationship to policies. transparency needed.

<mbgower> Recapping my comment: Current practice in the SECURITY industry is to only fix critical and high vulnerabilities before production release, with other vulnerabilities to be assessed and fixed on a subsequent timeline. How can we take that information to provide guidance that gets accessibility adopted with better traction, yet balances the reality that things will be released with issues?

Shadi: consider what policymakers can adopt. Make them aware. Possibly guidance doc for policy implementation

<SuzanneTaylor> teams.

Gregg: great discussion. wcag originally had 2 levels, required/recommended. third level was basically moonshots
... industry will be confused without knowing where the requirement is. Tolerances are never applied to standard but to provisions
... must use weighting or counting, if not using binary
... `weighting and counting has to be dictated by W3C or it wont work

scribe change?

<Poornima> I can take scribe, can you tag me in?

Rachael: what can we do to solve these problems?

<alastairc> scribe:Poornima

<Rachael> scribe: Poornima

<Fazio> thx

Rachel: we are going little off track from our conversation on potential solution
... slide 6, added few more pointers from our discussion

<Zakim> Chuck, you wanted to ask for scribe change

<Zakim> Jennie, you wanted to address considerations when purchasing tech

Jennie: one other thing re. to general conversation about legistative, consider purchase of 2 different technologies won't fit into 100% pass or fail conformance to say which one conforms better..

<alastairc> +1, WCAG 2.x is not a very helpful ruler for procurement.

<Zakim> SuzanneTaylor, you wanted to say that: We could make pass/fail more granular through specific contexts. For example, reading experiences (ebooks / fiction writers sharing site)

Suzanne: make pass/fail more granular, more success criteria and stringent, it can be applied only where it fits better

<Jennie> Build in criteria needed for comparing 2 products if using pass/fail

Makoto: like to add one more potential solution taking wcag 2, we tend to get lost in translation, e.g. japanese. Understanding WCAG 2.1 has many graphic examples, e.g. Non-text contrast including graphic examples can be a potential solution to some extent

<Wilco> might be

<mbgower> Agreed, Wilco, but our language can help guide those decisions.

Wilco: Makes me wonder all of these negatives for the legislatives, organization taking policies to satisfy, will that help addressing all these problems?

<mbgower> (Or make it more likely that poor policy decisions propogate)

<ToddL> my question is will those policymakers even listen? I've approached those people before and had no response to anything accessibility related

Judy: W3C is not a policy organization, but develops guidelines that are intended to be taken up to the setting, level of conformance can be used in the given setting.. tricky thing is granular question on the legal content..

<alastairc> ToddL - yes, in particular times/circumstances when it is their focus of work. Not in general.

Rachel: so far captured some are pros and cons, let's continue building on these, are there other potential solutions to capture?

<Zakim> mbgower, you wanted to say Create a computational baseline

Rachel: reading out the Potential Solutions captured so far

<ToddL> Thank you, alastairc

<Judy> s/tricky thing is granula question on the legal content/tricky thing is not all policy organizations are equipped to provide granular advice on policy uptake questions/

Mike: one of the approaches could be adding the computational baseline, most of them will be carried out in automated testing, ability to prove and show beyond the baseline?

<laura_> It may be helpful to check Ken Nikata's Digital Accessibility Legal Summit Keynote Video for his WCAG3 concerns regarding clarity vs simple. That part starts at 12:50 "I'm concerned about WCAG 3.0"

<laura_> https://convergeaccessibility.com/2021/10/11/join-us-at-our-upcoming-conferences-copy/

<Zakim> GreggVan, you wanted to ask if any of the provisions that are being used in the lawsuits are NOT in our list of critical errors

John: there are lawsuits by the blind in particular areas, by deaf in particular areas, it's a potential way to bring it up where organization may prioritize their target group

Gregg: In drive-by lawsuits, major purpose is to pay money and we don't consider any critical values, not the legitimate lawsuits. Most of the lawsuits target A level that are critical

Fazio: few states, say in california, several lawyers doing this lawsuit saying the sites didn't meet the checkers...

<laura_> https://convergeaccessibility.com/2021/10/11/join-us-at-our-upcoming-conferences-copy/

<Zakim> alastairc, you wanted to say that this matters for more than lawsuits, e.g. procurement, and prioritisation within product teams.

Laura: the video conf link is pasted to IRC, come and talk to us

<mbgower> +1 to say prioritization guidance is important

<Jennie> +1 to Alastair

<GreggVan> Was the schedule accessible? even reasonably? That actually gets to my point. the schedule is the most important part of a website for a sports team - and haveing all the rest accessible but you can't find out when and where the games are....

Alastairc: prioritization is the big one for me considering the different type of issues show up in the same element, for example

Shadi: measuring significance of the issue or more impactful is not straight forward, require experts and long term conversations, there's more than we can do, make granular, and easier

<GreggVan> they refused to make it accessible?

<Chuck> +1

<GreggVan> or didnt know how? how did the case get to court - rather than them just making the PDF accessible?

<Judy> +1 Shadi

Shadi: missing of concerns from other people, be respectful of other's opinions

<Chuck> +1 to Rachael

<GreggVan> sorry -- I thought only items typed by the Scribe showed up in the minutes

Wilco: useful to have standardized ruler to meet AA or not..

<Zakim> Chuck, you wanted to say that we should move along in the slides

<Zakim> Rachael, you wanted to suggest we move to percentages

ah_: comparing software procurements, one might support but not significant, and other is vice versa.. people allowed to make choices for procurement, more granular..

Rachel: adding functional needs , percentages to the list. this is not the final list anyways..

Jaunita: methods to adopt more maturity accessibility program, to consider down the road, that helps in lawsuits..

Gregg: what we are capturing here is very imp, beginning to hear same thing mentioned over and over again in multiple conversations.. refer to one location rather than minuting these.

<Fazio> These are all ideas we had 2 years ago. Just pointing out that we are revisting approaches. Maybe it'd be good to review previous minutes from 2020 and compile a presentation so we don't keep circling

<alastairc> Fazio - not in the full AG group, and this is the compiled pres....

Rachel: reading out Slide 7: Percentages
... Percentages Key benefits & Challenges.. What additional can be added to the slide?

<Fazio> Yes, in Silver we discussed it

<alastairc> Rather than "Supports automated checking", is that "Some guidelines could be automatically checked."

Gregg: checks some provisions, automating checking only support automated checkable items, big pain in testing methods, unfornatunately cognitive, and other aspects missed out..

Mike: the tool gives you how many % passed or failed on automated checkable items

Jennie: along that same line, Require tool support or calculate % of manual testing

Rachel: Reading out the slide 7 content ..

Poornima: the severity can be easily ignored

<Zakim> alastairc, you wanted to add possible solution: Use percentages at the test level, with rules for bundling up to a pass/fail at the guideline level.

Gregg: real world accessibility is more user experience..

<Zakim> Rachael, you wanted to reword new bullet real world accessibility

alastairc: one of the solutions can be test level of some guidelines, bundle up pass/fail at the guideline level..

<Zakim> SuzanneTaylor, you wanted to suggest...

Suzanne: another potential challenges, error in small sites is different from large site

Jennie: along the same line, simplicity of the page is good in cognitive disability, but the alt text missing is a different issue

alastairc: simple page with few things become very different compared to huge page.. it held the scoring of percentages.. a solution is to use % over a certain threshold for more clarity

<Zakim> Jennie, you wanted to address smaller content vs larger

<mbgower> But my point would be, if your site is simple, it should be easy to fix

Jennie: to address Mike point, one vendor chose less dense page to support cognitive disability and other choose differently.. but i would be concerned with the vendor that shows up more issues

<Jennie> But if I compare the 2 VPATs, there is no opportunity to fix, necessarily, during selection period

Rachel: Moving onto Slide 9 Points - Key benefits and Challenges to solve reading out..

Jennie: add 'tools needed to support scoring' under challenges

<Chuck> +1

alastairc: minimum scope of functional needs/user groups, you could do scoring within a guideline that leads to pass/fail reducing complexity..

<Chuck> +1

<Zakim> SuzanneTaylor, you wanted to say that another benefit is that Points allow flexibility to be defined separately for each Success Criteria

<kirkwood> +1

Gregg: on equity thing, we keep thinking about one group with other group, add 'across disability groups' in challenges

Suzanne: points can be equal in scoring level, but outcomes can be different

<alastairc> provide granularity (potentially)

Suzanne: have that ability to customize the granular level

<kirkwood> good point

Chuck: consider a site with great a11y score, but one a11y issue say 'unable to login'.. that score not become particular circumstance addressing this issue?

Gregg: answering to Chuck que, solution to the issue is evaluating the elements and process, already in WCAG 2, in one page it's accessible, in one page it's not..

<Wilco_> +1 Gregg. It's only not critical if you don't need it.

<kirkwood> denies equivalent experience

<Chuck> +1 to Shadi's suggestion

Shadi: remove the legal system from the 2nd bullet
... the last point can be changed to '.. needs to addres..'

<mbgower> But the ad content is not the responsibility of the page author. That's the germane thing, considering advertising

Gregg: the last bullet can be something like 'score low points even if it's critical for people..'

John: potential solution or benefits is it's essentially amount of burden / effort needed and critical part of business process..

<ToddL> Ads can also have deceptive patterns in them and have the potential for harm or deceit.

<kirkwood> burden could be effort, yes agreed Juanita

Rachel: reading out Slide 10 'Points: Potential Solutions'

Gregg: the last point refers to 'score elements and processes at a site'

Rachel: Moving to Slide 11 Adjectival Scoring - Key benefits & Challenges to solve

Gregg: question: is this scoring at the highest level?

Rachel: we could talk about many levels like test level, method level, etc..

Gregg: add 'increasing complexity and ambiguity' under challenges

<Zakim> SuzanneTaylor, you wanted to say could motivate or give credit for increased quality (a site with excellent alt text would score higher than one with just acceptable alt text)

<Zakim> bruce_bailey_, you wanted to mention that adjectival scoring in govt is limited to procurement (rating of offers) and personnel (hiring, annual appraisal)

Suzanne: another key benefit is 'gives credit for improving quality like motivating'

<Zakim> jeanne, you wanted to talk about interrater reliability

<kirkwood> non on my end Chuck

<ah_> +q

Jeanne: part of the reason adding this adjectival scoring, improve inter-rated reliability, it's the middle part where two parties can agree to the certain issues, layout that middle ground,

<bruce_bailey_> +1 to Jeanne suggestion to focus on middle of band

<Zakim> Chuck, you wanted to end the call

Jeanne: make it more precise specifically for qualitative in wcag 2

<GreggVan> good comment jeanne --- 3 levels might

<Wilco_> -1 To Jeanne, more granularity CAN improve accurate, but it also can do the opposite.

<GreggVan> increase IRR but everything might tend to fall in middle

thanks alastairc

This is scribe.perl Revision VERSION of 2020-12-31
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/correct place for user needs/correct place for discussing outcomes as user needs/
Succeeded: s/ventral/central/
Succeeded: s/graduations/granularity/
Succeeded: s/wcag 2 has many examples/Understanding WCAG 2.1 has many graphic examples/
Succeeded: s/intended to taken up/but develops guidelines that are intended to be taken up/
FAILED: s/tricky thing is granula question on the legal content/tricky thing is not all policy organizations are equipped to provide granular advice on policy uptake questions/
Succeeded: s/over a certain category/over a certain threshold/
Default Present: Ben_Tillyer, ShawnT, alastairc, Jennie, JustineP, shadi, ToddL, Peter_Bossley, jeanne, joweismantel, JakeAbma, Lauriat, Azlan, FrancesStorr, Jaunita_George, Makoto, MichaelC, Poornima, Chuck, JenniferS, mbgower, Jem, maryjom, Nicaise, Detlev, AWK, sarahhorton, Laura_Carlson, Katie_Haritos-Shea, StefanS, Wilco, GreggVan, kirkwood, GN, Francis_Storr, Raf, OmarBonilla, Fazio, Rachael, ChrisLoiselle, bruce_bailey_, SuzanneTaylor
Present: Ben_Tillyer, ShawnT, alastairc, Jennie, JustineP, shadi, ToddL, Peter_Bossley, jeanne, joweismantel, JakeAbma, Lauriat, Azlan, FrancesStorr, Jaunita_George, Makoto, MichaelC, Poornima, Chuck, JenniferS, mbgower, Jem, maryjom, Nicaise, Detlev, AWK, sarahhorton, Laura_Carlson, Katie_Haritos-Shea, StefanS, Wilco, GreggVan, kirkwood, GN, Francis_Storr, Raf, OmarBonilla, Fazio, Rachael, ChrisLoiselle, bruce_bailey_, SuzanneTaylor, GN015
Found Scribe: Fazio
Inferring ScribeNick: Fazio
Found Scribe: Poornima
Found Scribe: Poornima
Inferring ScribeNick: Poornima
Scribes: Fazio, Poornima
ScribeNicks: Fazio, Poornima

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: 

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)