<Chuck> meeting: AGWG-2024-06-11

<Jennie_Delisi> scribe+

<Jennie_Delisi> Chuck: Please consider scribing for hour 2

<Jennie_Delisi> Chuck: we will get started

<Jennie_Delisi> ...Anyone on the call wants to introduce or reintroduce yourself?

<Jennie_Delisi> JakeAbma: Announcement

<Jennie_Delisi> ...My company is setting up a new company with Funka - FunkaMax

<Jennie_Delisi> ...This will be launched this week

<Jennie_Delisi> ...We will be focused on gap analysis, strategy, and change management for large organizations

<Jennie_Delisi> Chuck: Congratulations

<Jennie_Delisi> ...Any other announcements?

<alastairc> Link for automated minuting requirements: https://github.com/w3c/AB-memberonly/issues/180

WCAG2ICT Update

<Jennie_Delisi> Chuck: A few weeks ago we put forward for review some proposals

<Jennie_Delisi> ...Same changes have been requested

<Jennie_Delisi> MJ: We have been making progress on finishing the last of the updates

<Jennie_Delisi> ...We have a survey

<Jennie_Delisi> ...We plan to finalize that on Thursday

<Jennie_Delisi> ...If that happens the pull request will go to the AG for review

<Jennie_Delisi> Chuck: Any questions regarding WCAG2ICT?

Conformance - What would a baseline include?

<Jennie_Delisi> Chuck: Reminder - we acknowledge this can be a challenging conversation

<Jennie_Delisi> ...We are governed by a code of conduct and code of ethics

<Jennie_Delisi> ...Presume everyone's good intentions

<Jennie_Delisi> ....Good opportunity to explore various opportunities we may need to explore.

<Chuck> https://docs.google.com/presentation/d/17VJvnm5UQW4WUzIoo9QNPVGfePgaZa8ifZWs-wtmv7E/edit#slide=id.p

<Jennie_Delisi> Alastairc: Slides have bright white screen

<Jennie_Delisi> ...Started the conversation last week but ran out of time.

<Jennie_Delisi> ...Assumptions we want to tackle

<Jennie_Delisi> ...Compliance with legislation does not mean baseline within the standard

<Jennie_Delisi> ...Example: We could recommend something is appropriate for legislation, but they could ignore that

<Jennie_Delisi> ...Number of outcomes within the baseline is essentially on a continuum

<Jennie_Delisi> ...No individual requirement, but could have a minimum number of requirements

<Jennie_Delisi> ...Or

<Jennie_Delisi> ...All requirements could be in the baseline

<Jennie_Delisi> ...WCAG 2 with a baseline from the legislative point of view is not all outcomes required

<Jennie_Delisi> ...Hypothetical example: if standard required 90% of bronze requirements needed for bronze, then 10% do not need to be met

<Jennie_Delisi> ...Given feedback: I don't think we would go through number of images

<Jennie_Delisi> ...If everything in the baseline

<Jennie_Delisi> ...Potentially a lot of testing

<Jennie_Delisi> ...Also: no extra, or going above and beyond a "typical" level of effort

<Jennie_Delisi> ...This would be expensive, unlikely to be adopted due to effort

<Jennie_Delisi> ...If baseline approach, could be based on different things

<Jennie_Delisi> ...WCAG 2: user needs based on usability and feasibility

<Jennie_Delisi> ...Baseline could be a cross section of functional needs

<Jennie_Delisi> ...Baseline could be set by types of content used

<Jennie_Delisi> ....Example: forms within a product. There could be requirements for that baseline

<Jennie_Delisi> ...Baseline could focus on adaptibility

<Jennie_Delisi> ...Example: for focus indicators, baseline could include that the default indicator is not overwritten

<Jennie_Delisi> ...Less on the author requirements on top of that

<Jennie_Delisi> ...This is not exhaustive

<Jennie_Delisi> ....We don't think either of the extremes is useful

<Jennie_Delisi> ...If we took a "With Baseline" approach

<Jennie_Delisi> ...Things which go into Bronze are required.

<Jennie_Delisi> ....Silver would be on top of the baseline

<Jennie_Delisi> ...This could be badges, or a percentage

<Jennie_Delisi> ...If we took a "Without Baseline" approach

<Jennie_Delisi> ...There could be a percentage of outcomes which fail

<Jennie_Delisi> ...a percentage with failures could still pass

<Jennie_Delisi> ...If we had a scoring approach

<Jennie_Delisi> ...meeting 160 out of 170

<Jennie_Delisi> ...Is this acceptable?

<Jennie_Delisi> ...You would not need to meet 100 percent of the baseline

<Graham> Essential - MUST

<Graham> Enhanced - SHOULD

<Graham> Exceptional - GOOD

<Jennie_Delisi> ...With this scoring approach, some things could be ignored.

<Jennie_Delisi> ...Is that acceptable?

<Jennie_Delisi> ...A few other items to emphasize

<Jennie_Delisi> ...Whatever is part of baseline could be less than a regulatory level

<Jennie_Delisi> ...We can expand the types of testing - the baseline does not need to just be automatible things or binary

<Jennie_Delisi> ...What has been mostly proposed before

<Jennie_Delisi> ...That there is a baseline - everything in the baseline is required

<Jennie_Delisi> ..."Non-baseline scenarios"

<Jennie_Delisi> ...Could have to have 90 percentage within each category

<Zakim> mbgower, you wanted to say that beyond setting three levels, and some things touched on in the Conformance section, 2.x does not really set regulatory expectations

<Jennie_Delisi> Chuck: we cannot hear Laurie

<Jennie_Delisi> MBGower: except for setting 3 levels, 2x has not set regulatory expectations at all

<Jennie_Delisi> ...unless iron-clad we cannot say how regulatory information is taken up

<Jennie_Delisi> ...David Fazio has brought this up before

<kevin> qq+

<Jennie_Delisi> ...In Ontario: they elected to remove some of the AA requirements because they felt they were not feasible for smaller websites

<Jennie_Delisi> ...Nothing stops them from doing that

<Zakim> kevin, you wanted to react to mbgower

<Jennie_Delisi> ...I think that is a consideration

<Graham> wait if you QQ + does that jump you up the queue?

<Detlev> audio quality is poor

<Jennie_Delisi> Kevin: Reminder: guidance to regulators

<Jennie_Delisi> * Hard to hear

<Zakim> Graham, you wanted to say core + minimum number of points concept.

<Jennie_Delisi> Graham: We have core, then scoring beyond that

<Jennie_Delisi> ...Core: stops someone using assistive technology from using something

<Jennie_Delisi> ...Beyond that: scoring system

<Jennie_Delisi> ...This could be a good balance

<Jennie_Delisi> ...Weighting could be on impact to someone's experience

<Jennie_Delisi> ...Scoring could then give a level beyond that

<Jennie_Delisi> Ljoekley: that is what I was thinking

<Jennie_Delisi> ...Without having core components in there, if navigation is not accessible, they should get zero percent

<Jennie_Delisi> ...Anything that makes a site unusable

<Jennie_Delisi> ...The flashes, navigation, etc.

<Jennie_Delisi> ...Those would have to be in the "core" and not negotiable

<Jennie_Delisi> Gregg: 1. We want to try to create something that is usable

<JenStrickland> Kevin, it was an important point that needed to be addressed promptly.

<Jennie_Delisi> ...Where different countries don't end up doing pot pourri

<Jennie_Delisi> ...2. If this thing is not accessible, we should state that

<Jennie_Delisi> ...3. Show stoppers - need to figure out what they are, ensure they are included in the requirement

<Jennie_Delisi> ...4. Be careful with show stoppers

<Jennie_Delisi> ...Cognitive language and learning has been seen as harder. But they still can be a show stopper

<julierawe> +1 to Gregg on plain language!

<Jennie_Delisi> Wendy: I agree with most of what has already been said

<Jennie_Delisi> ...We have no power over regulation

<Jennie_Delisi> ...We do have to consider when regulators look at documents to consider adopting, they look to us for recommendations

<Jennie_Delisi> ...WCAG 2 is an example

<Jennie_Delisi> ...AODA (Province of Ontario)

<Jennie_Delisi> ...referenced earlier

<Jennie_Delisi> ....I am sure we would see this even if we create a baseline

<bruce_bailey> Anyone knows academic review / survey of how current accessibility are being calculated ?

<Jennie_Delisi> ...We can't tell them what to do, but they are looking to us for advice

<Jennie_Delisi> Bruce: Does anyone have research on how to create the percentages?

<Jennie_Delisi> ...The US Federal Government did a survey last year

<Jennie_Delisi> ...Considering the algorithym

<bruce_bailey> https://www.section508.gov/manage/section-508-assessment/annual-reports/

<Jennie_Delisi> Shadi: I agree with Gregg - we are creating the ruler

<Jennie_Delisi> ....I think many regulations or legal requirements say WCAG - technical conformance is legal compliance

<Jennie_Delisi> ...We are trying to address an issue that is upstream

<Jennie_Delisi> ...Either we can try to create some kind of ruler

<Jennie_Delisi> ...rule into the ruler

<Jennie_Delisi> ....What is ok to not meet and still pass

<Jennie_Delisi> ...Or,

<Jennie_Delisi> ...We can try to (my view) provide guidance to regulators

<Jennie_Delisi> ...When you are adopting WCAG, here are some things to consider

<Jennie_Delisi> ...That could alleviate some of this pressure

<Zakim> jeanne, you wanted to say the accuracy in the "ruler not the rule" concept

<Jennie_Delisi> Jeanne: "Ruler vs the rule"

<Jennie_Delisi> ...I am building on what Bruce was saying

<Jennie_Delisi> ...Assumption: if we are the ruler, talking about measuring every success criteria

<Jennie_Delisi> ...We have the opportunity to start measuring how accessible the product is

<Jennie_Delisi> ...In the long run, this would be a more useful measure

<Jennie_Delisi> ...To help sites realize how accessible they are

<Jennie_Delisi> ...There is a demand for this

<Jennie_Delisi> ...Most major tools have a way to give a percentage

<Jennie_Delisi> ...This is one we could be doing, and standardizing on

<Jennie_Delisi> ...We can go beyond measuring individual tests

<Jennie_Delisi> ...This would allow us to include more in WCAG 3 than we can do by having a narrow view

<Graham> 100% with Jeanne here, being able to say "we went for 48 to 56" is a massive win so accessibility can have KPIs that C-suite can understand and go "line goes up, thats good"

<shadi> +1 to Jeanne

<Jennie_Delisi> ...While I agree with Gregg, but we can do more measurements than we are doing, and do it better.

<Jennie_Delisi> Gregg: +1 to Jeanne

<kirkwood> +1 Jeanne

<Jennie_Delisi> ...Scoring: whenever we look at standards, it always gets gamed

<Jennie_Delisi> ...When you have a ruler, it doesn't just have 1 mark on it

<Jennie_Delisi> ...Has numbers - you can pick how high or low to go

<Jennie_Delisi> ...We need to figure out how to do that

<Jennie_Delisi> ...Second: you cannot buy things without AI and neural networks in it

<Jennie_Delisi> ...When we talk about show stoppers and the opportunities for cognitive language and learning

<Jennie_Delisi> ...Remember - took us 5 years previously

<Jennie_Delisi> ...In 5 years the world will be really different

<Jennie_Delisi> ...We need to think about AI, and what are the show stoppers which cannot be addressed by AI

<Jennie_Delisi> ...Third: If we can show industry (Browser makers) that if they can make their browsers to allow users to have really accessible content

<Zakim> Chuck, you wanted to say one concern I've always had with percentages

<Jennie_Delisi> ...Without authors having to do things they had to do in the psat

<Jennie_Delisi> Chuck: (chair hat off)

<kirkwood> also the ruler for blind is different than the ruler for deaf, different for cog, etc.

<Jennie_Delisi> ...I have one concern: you can have the most nearly perfect application, but has 1 accessibility bug

<Jennie_Delisi> ...But what if that is the log in page

<Jennie_Delisi> ...If this stops users getting in, how accessible is that application?

<kirkwood> it’s inaccessible.

<Jennie_Delisi> Jen_G: I remember a health care issue we had in the United States

<Jennie_Delisi> ...As creators of the rules or the rulers

<Jennie_Delisi> ...We need to use human centered design to get regulators to adopt WCAG

<Jennie_Delisi> ...Re cognitive elements and tagging onto Chuck's point - we have to think about things more holistically

<kevin> -1 to nudging regulators

<Jennie_Delisi> ...As we craft these, going back to Mike's point

<Jennie_Delisi> ...We cannot control how regulators use WCAG, but we can make it easier for them

<Jennie_Delisi> ...How clearly we rank things

<Jennie_Delisi> ...Anticipate the way they can be abused

<Jennie_Delisi> ...How difficult they could be to implement

<Jennie_Delisi> ....I also want to talk about Jeanne's point: we serve multiple audiences

<Jennie_Delisi> ...Not just regulators

<Jennie_Delisi> ...In the US we have Section 508 (which applies to federal agencies)

<Jennie_Delisi> ...Expanding beyond to those who receive federal funding is something to consider

<Jennie_Delisi> ...We also have private companies who use it as branding, and a quality for their product

<Jennie_Delisi> ...We need to think about all the audiences, not just regulatory

<Zakim> alastairc, you wanted to comment on percentages, and why by-issue is unhelpful, and what happens when we create a 30cm ruler... and to also mention the tyranny of the default

<Jennie_Delisi> Alastairc: Comments on percentages - we had a lot of feedback on the 1st working draft

<Jennie_Delisi> ...Instances of counting got a lot of negative feedback

<Jennie_Delisi> ...Going by issue can be unhelpful

<Makoto> +1 to Jen on making it easier to use for wider range of audiences

<Jennie_Delisi> ...Example: the log in screen bug

<Jennie_Delisi> ...We would need to build in some sort of ratings of issues, such as in a specific location or process

<Jennie_Delisi> ...We would need a way to capture that

<Jennie_Delisi> ....Re ruler: there is the tyranny of the default

<Jennie_Delisi> ...Some countries could pick different levels, but most have gone with a kind of default

<Jennie_Delisi> ...What we could present as a baseline - I would hope regulators would pick up on that

<Jennie_Delisi> ...By doing so we could minimize the differences between countries

<kirkwood> the ruler is very very important.

<Jennie_Delisi> ...That comes down to us coming up with some default level which seems reasonable to people

<JenStrickland> I said a lot, yet forgot something: Presumably we are all dedicated to democratic principles. This means we are obligated to prioritize the well-being of the least well-off. As AGWG, that is certainly part of our vision. Even within disabilities there are more and less well-off / disadvantaged. Considering the impact of COGA, for example, keyboard,

<JenStrickland> screenreader, audio, mobility… and how the guidelines address barriers in a degree of impact.

<Jennie_Delisi> Graham: I am trying to approach from people using it: testers, developers

<Jennie_Delisi> ...Scoring with a core: this means you must do this

<Jennie_Delisi> ...Something with high impact and high effort could get a higher score

<Jennie_Delisi> ...If worth a lot more points, that might help

<Jennie_Delisi> ...There could be a must section which is pass/fail

<Jennie_Delisi> ...Then beyond that use a scoring system based on difficulty

<Zakim> mbgower, you wanted to say I think an automated testing baseline has merit; through something like ACT we have the ability to have a dynamic measure of technological ability

<Jennie_Delisi> Mbgower: I think if we have automated testing as standardized

<Jennie_Delisi> ...that should be part of the baseline

<Jennie_Delisi> ...Then, try to identify what cannot be caught by automation which poses some level of harm

<Jennie_Delisi> ...Easy to validate, easy to contest, and offers

<Jennie_Delisi> ...a potential for the specification having a more finely-tuned ruler

<Jennie_Delisi> ....Over time: you get more repeatable, automated testing

<Jennie_Delisi> ...You could have releases of that ruleset

<Jennie_Delisi> ...As technology has the ability to measure more things we could have different versions

<Jennie_Delisi> ...This could lie underneath a lot of the discussion

<Jennie_Delisi> ...The focus can then shift to items beyond the baseline, which builds onto of it

<Graham> automated is a good start for things, but we would probably need to add a keyboard navigation test and no more than 3 flashes etc. Things that are completely and utter blockers. Otherwise I could pass an automated test and yet have links on the page that are divs as it won't know they should be clickable.

<Graham> like Mike's idea for a base around automation though!

<Jennie_Delisi> Chuck: an imperfect poll, not for decision making

<Chuck> Poll: 0) no opinion 1) no baseline, 2) everything is the baseline, 3) a baseline (tbd) + extras

<Jennie_Delisi> ...your answer is not a commitment

<shadi> 3

<Graham> 3

3

<giacomo-petri> 3

<MJ> 3

<alastairc> 3

<Jennie_Delisi> 3

<mike_beganyi> 3

<jtoles> 3

<Makoto> 3

<ThompsonS> 3

<Frankie> 3

<GreggVan> 3

<jeanne> 3

<laura> 3

<wendyreid> 3

<tburtin> 3

<JakeAbma> 3

<ljoakley1> 3

<Jen_G> +3

<Rachael> 3

<JenStrickland> 3

<Zakim> Chuck, you wanted to offer a poll

<Detlev> 3

<RAin> 3 if I must choose, but thinking a 4) a set of "stoppers" first, then a baseline that is more inclusive, and then extras where the "helpfulness" is more personal

<Jennie_Delisi> Wendyreid: I think we are all in agreement with a baseline!

<Jennie_Delisi> ...Challenge: how do we get there, how do we define it?

<Jennie_Delisi> ...We want to make sure it is an inclusive baseline so we don't leave anyone out

<Jennie_Delisi> ...From a previous conformance discussion: pre-requisites

<Jennie_Delisi> ...Do you even have these?

<Jennie_Delisi> ...Example: automated testing

"prerequisite" is a nice concept (and automation is a good candidate for that)

<Jennie_Delisi> ...If it has errors, you are not ready to do a baseline test

<Jon_avila> 3

<Jennie_Delisi> ...Or, if your tool doesn't enable use of alt text

<Jennie_Delisi> ...We can create stages so people can assess themselves earlier

<Zakim> alastairc, you wanted to outline the proposal

<Jennie_Delisi> Alastairc: (chair hat off)

<Jennie_Delisi> ...I am not convinced we want a baseline that is just automated testing

<Jennie_Delisi> ...1. Some can cause a lot of noise

<Rachael> +1 to not using automated testing as a baseline criteria

<Jennie_Delisi> ...2. They are not always good at capturing show stoppers

I never advocated automated testing being the only thing in a baseline, in case it sounded like I was

<Graham> +100 to not ONLY automated as baseline!

<Jennie_Delisi> ...Also: putting effort in that makes it cross functional

<tburtin> +1 to not only using automated as baseline.

<kirkwood> baseline for all parties? or specific disabilities?

<Jennie_Delisi> ...Potentially: some technically focused COGA things

<jeanne> +1 to not having an automated baseline level

<Makoto> +1 to not ONLY automated as baseline.

<Jennie_Delisi> ...We would want to make the baseline work cross functionally

<Jon_avila> I agree with Alastair, if passing automated test was required and you forgot to mark an image as decorative that doesn't seem like a critical issue compared to other tests.

<Jennie_Delisi> ...I put a proposal on screen

<Jennie_Delisi> ...If we start with a minimal baseline, with cross-functional needs

<Jennie_Delisi> ...Focused on adaptation and user-agent support.

<Jennie_Delisi> ...Then another level which is scored, providing a means of progressing

<Jennie_Delisi> JenStrickland: A fundamental question

<Jennie_Delisi> ....Do we have a user map of all who will use WCAG?

<Jennie_Delisi> ...I wonder if, as we develop these things, be able to evaluate our potential options

<Graham> My list was certainly not exhaustive! hahaha

<Jennie_Delisi> ...considering each of these types of users

<Zakim> jeanne, you wanted to disaagree that having an automated baseline could be abused and reduce accessibility overall.

<Jennie_Delisi> Jeanne: My +1 was I agreed that we should not have an automated baseline level - had a typo previously

<Graham> hahah I did wonder Jeanne! I thought you were changing your vote

<Jennie_Delisi> ...1 thing I learned from the discussion last week is the difference between having a narrow and a broad baseline

<Jennie_Delisi> ...Prior to last week I was in the narrow baseline

<Jennie_Delisi> ...A lot of people said thoughtful things last week and I was persuaded

<Jennie_Delisi> ...I believe we need a broader baseline

<Jennie_Delisi> ...It provides more protection for more functional needs

<Jennie_Delisi> ...I wrote a proposal last week - the No baseline proposal

<Jennie_Delisi> ...It is important to look at them and see their flaws

<alastairc> JenStrickland - Is this what you werre asking about? https://docs.google.com/spreadsheets/d/1R9e1Sza-lcNfwfxTCE4AN-lvbbHKEv9G/edit?gid=459874840#gid=459874840

<Jennie_Delisi> ...The flaw in the approach I had was that every guideline must be addressed in the product

<Jennie_Delisi> ...That is still too narrow

<Jennie_Delisi> ...Let's go with a broad baseline, and how do we do that?

<Jennie_Delisi> Rain: This is a great conversation

<Jennie_Delisi> ...I chose 3 in the vote because there wasn't another option

<Jennie_Delisi> ...This conversation is leading to something I might feel more comfortable with

<Jennie_Delisi> ...If we offer a baseline, that is what the majority of the world will adhere to, and won't be aware of the other aspects

<Jennie_Delisi> ...They will treat everything within the baseline as equal

<Jennie_Delisi> ...How do we ensure that what needs to be accomplished is done?

<alastairc> I don't agree, otherwise people would have taken WCAG 2.0 level A as the regulatory level

<Jennie_Delisi> ...I want to reinforce Wendy's idea of pre-requisites but with a different take

<Jennie_Delisi> ...Could we also consider adding pre-requisites for those criteria which exist?

<Chuck> qq+ to change scribe

<Jennie_Delisi> ...They include ones which cause harm, like flashes

<Jennie_Delisi> ...Could we include that the user must be able to log-in

<Jennie_Delisi> ...Because then the rest wouldn't matter

<Jennie_Delisi> ...Then the baseline itself would be a layer that includes everything that is hard to debate

<Jennie_Delisi> ...Everything that breaks down accessibility barriers regardless of unique context

<Jennie_Delisi> ...Including anything that has to do with adapting so it can be used by AT

<Jennie_Delisi> ...Then the extras is everything else with some kind of nuance

<Jennie_Delisi> ...Maybe because of personalization

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

<Zakim> Chuck, you wanted to react to RAin to change scribe

<laura> https://www.w3.org/WAI/GL/wiki/Scribe_List

scribe

<alastairc> scribe: mbgower

Graham: Really good conversation
... What I meant about a baseline core is it is away from any scoring. You're still at zero
... We can be a little more flexible about that

<Chuck> scribe+ Chuck

After core you have a list of things, and you measure with a score, say bronze

Graham: I like the idea of a mixed bucket

<Zakim> mbgower, you wanted to say we are ultimately serving individuals, not audiences

<Chuck> mbgower: There's a couple of things I wanted to tackle. I don't agree that we set the baseline and that's all they follow. In WCAG A was the basline, but regulatory adopted more than A. Let's be careful of assumptions.

<Chuck> mbgower: And the other part is that everyone will give equal weight. That's not been my experience. Teams start somewhere and want to find out where to start.

<Chuck> mbgower: ...areas to focus (beyond automated test), and then a set of things to meet core. Now I've a set of priorities I can focus on.

<Chuck> mbgower: Jen asked who the audience is. That's everyone who wants/needs to use the system. How do we make that work, that's the challenge.

<JenStrickland> Note: I didn't ask who the audience is.

<Chuck> mbgower: I worked with someone, and this individual needed to deal with systems beyond the main app. Even if you are working for every person in the world, there's the baseline we can plug into. However we slice that, I think we'll never get to the point where we meet the needs of every individual. that balance of a baseline to support

<Chuck> realistic adoption is the magic button.

<Chuck> mbgower: Idea of pre-requesits helps create that. And then how to reward and recognize going above and beyond. Great discussion. I don't know how we solve it, but it seems we are all in agreement in the direction.

<Zakim> Rachael, you wanted to briefly discuss survey in context of this

<JenStrickland> mbgower I didn't ask who the audience is. That wasn't the question.

<scribe> scribe: mbgower

Rachael: I am concerned about a prerequisite level before a baseline.

<JenStrickland> alastairc - Thank you for the spreadsheet! Yes, this helps with my question.

Rachael: Many people think about 2 levels. if we do only levels, people may be inclined to pick up the pre-req and the baseline, instead of the baseline and something beyond it
... I like the concept of focusing on adaptability.
... Focus on supporting screen reader users, not blind users
... Can we be working on other W3C groups to add tags for importance, or other hooks for technologies to come in for adaptability?
... That can help take burden off the author
... while providing improved equity
... I also wanted to talk about the survey results. Chair hat on, thank you. It was a poorly executed survey. My apologies.

<laura> https://www.w3.org/2002/09/wbs/35422/Requirements_Survey/results

Rachael: We have people that are going to help us with future surveys. I wanted to acknowledge it was challenging for many. We will not make decisions on it.
... I wanted to focus on some things. About 64 had support from some and 94 had a third or more supporting. That's meant to inform thoughts, not be a driver.
... The things people were using to decide things centered around User Need, Feasibility, Subjectivity (and how hard to quantify), Safety, and Ease of remediation

<Zakim> alastairc, you wanted to say that we can (and probably should) break things down WITHIN outcomes.

<Rachael> User Need, Feasibility, Subjectivity/Hard to quantify, Safety, Ease (low hanging fruit)

Alastair: I'll +1 for us to think of a prerequisite layer. But if Bronze was our baseline, I don't think having a subset underneath that helps.
... I also wanted to say that one of the advantages of redoing the requirements is that we can and probably should structure the levels within outcomes.
... I keep referring to Focus Appearnace because it was recently looked at.
... We could use Heading and Landmarks from lists, for example, since they have different levels of importance.

<Zakim> Jennie_Delisi, you wanted to discuss architecture review boards and product selectors

Alastair: That could allow us to have a more nuanced list of outcomes

Jennie: I recently took a new role. I am not speaking on behalf of them, but speaking in regards to that experience.
... I think we also need to consider architects and procurers.
... It really help solidify the entry level, and what is needed to truly support functional levels.
... That could address some of Rain's concerns.
... We are seeing a real challenge with vendors complying.

<Rachael> +1 to including coga at all levels

Jennie: I also think we need to consider coga

<laura> +1 to Jennie.

wendy: I believe we can find a way of having a broad wave that includes everyone.

<JenStrickland> I agree with you, Jennie, re including COGA at all levels. It requires some additional labor to raise awareness across AGWG and within the materials.

wendy: I totally agree we want levels, but I also work in industry. We are focused on baseline is it has a very clear level of conformance. While there are many orgs that care, others are going to focus on the need for compliance.
... I'd rather game it so that if all you care about is ticking a box, you're still going to provide an adequate level of use. Others will go higher. i think everyone here is part of that group. But many are going to treat it as a box they just need to check off before they move to the stuff they care about.

Gregg: I like the idea of prerequisites, but I don't think that should be in the document. it can be in advice documents; here's how you go about doing an evaluation in an efficient way.

<Rachael> +1 to prerequisites in an accompanying note

Gregg: Including Easy in the baseline is a bit dangerous.
... If we want the baseline to be seen as important, whether it's easy is another dimension.

<Zakim> bruce_bailey, you wanted to ask if prerequisites could "proof points" from a11y maturity model -- something where % score is obviously not useful and to ask if prerequisites could

Bruce: There are such drivers to come up with a percentage that this ideas of prerequisites should be something not encompassed by that.

<bruce_bailey> https://w3c.github.io/maturity-model/

<Zakim> Chuck, you wanted to say alternative to "easy" is "no excuse"

Bruce: Maybe we could look at 'proof points' from the maturity model; something that doesn't work well as a percentage.

chuck: in my conferences, I think a synonym to easy might be "no excuse"
... if someone else finds issues with an automated test, they can be dubious. Why didn't you find it?

<Nayan> q

Nayan: I would say from our own experience dealing with small businesses is 'is it custom coded or an authoring tool like shopify?'
... That limits what they can do on their own.
... The automated tests get 30-40%. I disagree with the notion that people are only going to do the minimum.
... We have helped folks achieve AA. Everyone is happy if they can do it without breaking core functionality.
... It would be useful to determine 'who is the audience'.
... If we're going to do this at scale, we need to make the baseline sufficiently reasonable that it can be accomplished. Without participation of authoring tool providers, it's going to be difficulty

<Zakim> alastairc, you wanted to poll the proposal

Nayan: Think about it in a realistic way. Our goal and reason for participating is increasing accessiblity at scale

<alastairc> Proposal, +/- 1, for: We start with a minimal baseline “core”, with effort to make it even across functional needs. Prioritise requirements for adaptation and user-agent support (within outcomes). Then another level (or two) which is scored, providing a means of progressing.

<Zakim> Rachael, you wanted to suggest one change

<Chuck> mbgower: One of the challenges, we don't always have data. We know that motion can trigger vestibular issues.

<Chuck> mbgower: We can't get to a measurable point, we don't have the data. I think we should focus on "harm" and obtaining the data to determine what reduces the harm to below some threshold. Flashing for example, I bet some individuals would be triggered even at the levels we set.

<scribe> scribe: mbgower

<Chuck> mbgower: We always will be setting some qualitative level, but we lack the information.

Graham: user agent support for a core thing gets us into a debate about how difficult it is. UA support puts it beyond the reach of many people.

<Zakim> Rachael, you wanted to talk to research

Alastair: that may be a misunderstanding

Rachael: I think the research can be part of the goal

Gregg: We need to figure out where we need research and where we don't
... We don't need to have people jump out of planes without parachutes to prove parachutes are a good idea.
... Flashing we went and did research on.
... There will be tension around that. We spend hundreds of thousands of dollars in TRACE to get those flashing levels.
... Some requirements are not for authors; user agent support was a consideration in our requirements

<Zakim> alastairc, you wanted to explain user-agent support aspect

Alastair: I need to rephrase the user agent support. What I meant by that is adaptation/user agent support. If there are requirements that can be met by UAs, make sure you are not blocking them.
... Font size, reflow, those kinds of things.
... Meta data for importance. It's intended to be the easier end of the scale.

<Zakim> jeanne, you wanted to say that I like the direction we are going, but we need to do a lot more work in developing guidelines to know whether this is going to work and where the

Jeanne: I just want to remind people that every time we do a deep dive, we learn more about the details of conformance.
... I don't think we should start to say 'we've decide this is the plan'. As soon as we dig, we learn it won't work the way we thought it did.

<alastairc> +1, and noting that this discussion is informed by the work we've done on several guidelines, and we need to keep going with SOME structure, so trying to get a reasonable direction

Jeanne: That's why I think we keep focusing on the guidelines so taht we learn about the conformance. We don't want to be in a situation where we've decided on a conformance structure and it's not going to work for some, and our response is 'the need doesn't fit our conformance model'

<Chuck> +1

<Zakim> Rachael, you wanted to speak about next steps

Rachael: We are still working towards exploratory.
... We hope to present something in exploratory this fall. We are somewhere between 'placeholder' and 'exploratory'
... So +1 Jeanne to be mindful. We are exploring a direction, and then we can take a step back.

Graham: I am a big proponent of scoring, but it lends itself to people being ignore if fixing things become expensive.

<Zakim> Chuck, you wanted to ask about offering the proposal as a poll

<alastairc> Draft RESOLUTION: We progress with a baseline and make effort to make it even across functional needs. Prioritize requirements for safety, blockers and adaptation / author support for user agent features (within outcomes). Then another level (or two) which is scored, providing a means of progressing.

<jeanne> +1 Graham -- I think about that a lot as I look at proposals

Graham: Scoring does end up with an outcome based on expense to implement.

+1

<wendyreid> +1

<bruce_bailey> +1

<Rachael> +1

<jtoles> +1

<mike_beganyi> +1

<Graham> +1

<jeanne> +1

<Makoto> +1

<avkuo> +1

Lori: What happened to the core?

<ThompsonS> +1

<Frankie> +1

<Chuck> +1

Alastair: "Baseline"... "Core" are synonymous.
... We haven't included prerequisites here. We can return to that.

<kirkwood> +1

<Jennie_Delisi> +1

<tburtin> +1

<laura> +1

<ljoakley1> +1

<kevin> +1

<bruce_bailey> +1 for "requirements before scoring"

RESOLUTION: We progress with a baseline and make effort to make it even across functional needs. Prioritize requirements for safety, blockers and adaptation / author support for user agent features (within outcomes). Then another level (or two) which is scored, providing a means of progressing.

graham: I want to be clear. +1 all the way. What if we had buckets you could choose from, to try to keep people from picking only from 1 bucket. You have to score a certain amount across them. I'm trying to prevent exclusion from gaming.

<alastairc> That's less baseline than silver/gold

Rach: We dove in and explored in the FPWD.

<Detlev> what does it mean (within outcomes)?

Rach: There are challenges and benefits. Outcomes rarely support only one user group.
... How you place outcomes across user groups can also affect scoring, and there can be bias in that.

<jeanne> \

<Graham> Thanks Rachael, that is useful info / good thoughts!

Rach: We will continue to dive into this.

<bruce_bailey> scribe+

<Chuck> mbgower: As a facilitator, just want to let folks know, WCAG 2.2 understanding docs have been updated so in brief material is at top. Please review.

<Chuck> mbgower: We are ending a 2 week cycle, we'll have a smaller set of proposed changes.

<alastairc> Example of the in-briefs: https://www.w3.org/WAI/WCAG22/Understanding/captions-prerecorded.html

https://www.w3.org/WAI/WCAG22/Understanding/

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/qqqqqqqqqqqqqqqqqqqqqqqqqqqqqq + going forward! haha//
Succeeded: s/haha sorry Lori!//
Succeeded: s/to not having a baseline level/to not having an automated baseline level/
Succeeded: s/> Core is for Mom and Pop shops./...> Core is for Mom and Pop shops./
Succeeded: s/alastairc I did not.//
Succeeded: s/Grah: I am/Graham: I am/
Succeeded: s/their can/there can/
Default Present: kevin, bruce_bailey, Frankie, Jennie_Delisi, tburtin, alastairc, JakeAbma, fofila, wendyreid, RAin, giacomo-petri, julierawe, Graham, Makoto, MJ, ThompsonS, Kimberly, Francis_Storr, Laura_Carlson, Jen_G, Rachael, jeanne, shadi, JenStrickland, jon_avila, ljoakley, mike_beganyi, Detlev, kirkwood, gpellegrino, mbgower, jtoles, Nayan, avkuo
Present: (no one), kevin, bruce_bailey, Frankie, Jennie_Delisi, tburtin, alastairc, JakeAbma, fofila, wendyreid, RAin, giacomo-petri, julierawe, Graham, Makoto, MJ, ThompsonS, Kimberly, Francis_Storr, Laura_Carlson, Jen_G, Rachael, jeanne, shadi, JenStrickland, jon_avila, ljoakley1, mike_beganyi, Detlev, kirkwood, gpellegrino, mbgower, jtoles, Nayan, avkuo
Regrets: Todd Libby, Roberto Scano, Poornima Subramanian, Azlan Cuttilan, DJ Chase, DanB
Found Scribe: mbgower
Inferring ScribeNick: mbgower
Found Scribe: mbgower
Inferring ScribeNick: mbgower
Found Scribe: mbgower
Inferring ScribeNick: mbgower

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth


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.)