Meeting minutes
Intros and announcements
AC: Any new members?
(None)
<alastairc> TPAC link: https://
AC: I would remind everyone that the TPAC registration is open. It's good to get the early bird, it's cheaper. So if you are going to join us at TPAC
… Please do get registered soon.
<hdv> link to the fund: https://
WCAG 2 updates https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JulSep/0038.html
BB: Yes, we're halfway through our two week usual review period. One normative item this time
… The name of focus, not obscured, is really about not obscuring the component.
AC: there is a CFC open, I think that's closing on Friday, so if you haven't responded to that, please do.
WCAG 3 document breakdown https://www.w3.org/wbs/35422/wcag3-document-breakdown/
<bbailey> CFC - WCAG 2 editorial errata
<bbailey> https://
AC: In the meetings. So, most of the items in it Things we have discussed, so scoring levels, and we did try and go through everything a couple of weeks ago.
… Just reminding that, compliance is whether the thing meets the technical standard, sorry, conformance is whether the thing meets the standard, compliance is what laws and regulations
<bbailey> https://
<alastairc> https://
AC: So for scoring, this is something we went through in detail about four weeks ago
… So there was a whole presentation about this.
… Reviews: https://
… and survey results.
… Here we've got one must five should
… And then 11 should-nots and 2 must-nots. So, the weight of it is kind of towards making it part of informative materials,
… It's always interesting going through the comments, because occasionally you realize, oh, this person Selected this option, but actually that means something else.
… Matt outlined quite sort of a big chunk of things
… for what he thought scoring would be. So slightly reframing the question, which is fine.
… he said it must be part of non-conformance, he wouldn't do it the way that it was described in our previous meeting
… And thought that there are significant advantages to including A scoring thing in conformance, but it would have to be sort of normatively standardized, require specific data
… My aim is to sort of get through all of the comments And make sure we have sort of noted down the positives, negatives, and what would be kind of blocking issues, from either direction.
… Other people who thought it should be normative
… thought that if it could work across multiple contexts, they would be in favor, otherwise they wouldn't object to it being informative
Gregg: The topic of tolerances keeps coming up. It's strange, but if you say it has to be five with a tolerance plus or minus one half, then the requirement now is 4.5. As soon as you add a tolerance, it just moves the requirement down by the tolerance.
AC: I think this is a bit different. So if it was a non-blocking issue, I think the proposal was to allow a certain percentage or torrent level of errors based on how much content or how many Ul elements there are
Greg: I think we're confusing enforcement with what the standard is.
<kirkwood> +1 to Gregg (tolerance) lowering requirement
Lia: l'd like to ask a clarification on this last point.
… So my understanding, and again, just confirming that my understanding is correctly, is that what was proposed is that we have a set of things that are required. And for that one, we expect 100% for conformance, and then a set of things that are not required, maybe less serious.
AC: Mary Jo's comment was about a different way of doing it. The original proposal for the scoring was that.
… Based on, you know, if you did, like, an audit of a set scope, tot up all the kind of failures across, get your score, it would provide, potentially provide, kind of, thresholds that either we could say you need to meet this threshold, or regulators could say you need to meet a certain threshold
MJM: I'm just thinking about, like, UL standards and things like that. You can pass a UL standard, and you can certify that it, you know, electrical stuff is safe, and, you know, it has to be within a certain tolerance.
… Full conformance is pretty much impossible for anything very large, very dynamic.
… you might be able to reach it, or a page, but when we're claiming conformance.
… you know, for products and websites and things like that, it's aggregated.
<AWK> >99%, I'd say
<Lia> agreed Mary Jo, thanks for clarifying
Gregg: I don't think you can have any part of a product that's unsafe, and it'll pass UL inspection.
<AshleeF> What does UL stand for?
Gregg: I don't think you can have any part of a product that's unsafe, and it'll pass UL inspection. If you mean that in enforcement, if you test 10,000 products, and only 2 of them have.
<Charles> given a single failing instance of a single criterion means non-conformance, i would think 100% of websites are non-conformant.
Gregg: I think you're talking about enforcement part of UL, not the UL standards. Am I misunderstanding, or could you explain more how UL allows.
MJM: Well, it's not necessarily unsafe, but, like, they have a tolerance for ranges.
<Lia> it's a non-binary conformance model
Gregg: Okay, so that's enforcement.
MJM: But the problem is the people trying to enforce.
<HaTheo> I think we do have a range for things like contrast right, the range is 4.5 or higher 21(or w.e. max contrast is)
<HaTheo> I worry that the error rate is also based on the thoroughness of the testing which I think is what Mary might be getting at.
<LoriO> +1 Mary Jo
AC: I think We sort of agree that there's a problem there in terms of it is very difficult for any larger dynamic site and or dynamic site to. to get to 100%.
<Rachael> +1 to addressing this in the policy document
AC: If we did include something like that, what would the impact be in terms of sort of uptake?
<HaTheo> I think the benefit of UL is there is a standardized testing method which we don't have and I think that might partially be what Mary is getting at. Another alternative is requiring people making conformance claims to sharing the testing method and scenarios covered in testing as more error does not necessarily mean a worse product.
MK: I'm hearing both what Greg and Mary Jo are saying.
… I think you're spot on that it is a little bit of an apple and oranges comparison specifically in that case.
… So I've been doing some pretty deep research for precedent.
… What I've seen in other ISO-related output standards is the thing that is normative.
… Is the measurement and the reporting and the testing.
… What I'm seeing in the standards is that they will have normative ways of classifying errors.
… Such that if something is severe, then it's not tolerated, or if it is tolerated,
… what the standards do is they allow, because there's normative classification of errors, they do allow a certain percentage of errors to slip through, or to be in a system, like to be in a transportation system or a software security system, or something like that.
Gregg: could you send that information on?
MK: I'm preparing a document and presentation.
Gregg: if we're talking about percentages of error, I think we're sort of getting a little bit into the enforcement area.
… And it sounds like kind of what these standards say is they present that they depend on there being an enforcement agency.
… we don't have any ability to set those time limits, and if they don't already exist, then you can't write a standard that depends on them.
Gregg: the scoring bit, as I said before, as soon as you just say you only have to do 80% of the rules, then we have no standing for requiring any of them.
JK: I don't see how this could possibly work in any way.
… But it only says it needs to meet 80% or whatever, and this person still is not able to access what they need. It just throws it out, and it will throw out the settlement of using our guidelines for the standards of making things accessible.
… We're just creating the specifics of how to do it as far as the business processes are concerned and how the technology is moving and all this stuff.
… We're the experts on the button height, on the way elevators move, the pathways they're building, that type of thing.
<HaTheo> +1 to John statement, I think its going to be really difficult to make a generalized score that works across products and use cases.
JK: We're not the ones saying whether there's business needs or it's too difficult. Obviously,
… I think this is the wrong way... wrong approach.
Lia: I am very much in favor of a non-binary conformance model, and
… And the way I think about it is that it's up to us to define what conformance means and what levels exist. And then it's up to regulators or enforcement agencies to define what level they want to target and enforce.
… And I do agree with Mary Jo that complex - I don't believe in 100% conformance with WCAG as it is today. I don't think it's feasible. And
… right? I think what Mary Jo mentioned is that we would define things that could be very serious, and those things need to be, like, 100%, and then the tolerance would be for things that might not be, that harmful to users, right?
Wilco: I don't really like this idea of scoring system inside of WCAG. I think there are far more.
… I think there are many more important things that are not.
… Are there other parts of WCAG that are not relevant to WCAG?
… to consider when you're looking at building up an organizational metrics.
… It seems to me that if something is an accessibility issue, then...
… It really should just be required.
<Zakim> alastairc, you wanted to comment on figuring out tolerances across scenarios.
<kirkwood> +1
Wilco: I don't see this idea of having some percentage of maybe low impact issues not be required. I think the real question, if something isn't a big deal, it shouldn't be in WCAG in the first place. And if it is a problem.
… is really an organizational problem, and legislators have and can come up with tools to legislate around that,
… ultimately, Its not good for WCAG to have requirements that we then go back and say, yeah, but if you only have some of them, if you have some issues, it's okay. I think that undermines the message.
<Zakim> AshleeF, you wanted to ask about the role of the Google doc with summarized survey results
AC: My main concern would be just the variety of things that it needs to cover, My main concern would be just the variety of things that it needs to cover,
<GN015> Does it make sense to incorporate the severity of issues? Like no blockers, few barriers and not too many limitations? As already said, a purely numeric scoring doesn't help much, if among the remianing issues there are severe blockers in central functions.
AC: organizations.
… you have big products, and, you know, the...
… People power, person power to,
… to figure this kind of stuff out, but also it needs to be applicable at much smaller scales as well.
… even individual developers should be able to use WCAG.
<Ben_Tillyer> +1
AF: I wanted to ask about the role of this Google Doc. I think you stated it at the beginning, but I'm just wondering if you can clarify. how the text that's in the stock is going to be compared to the survey results.
AC: it's a temporary notes document is the answer.
MK: I personally objected to the kind of scoring that has been presented so far, but I hear people sort of rejecting the concept of scoring outright. But what if there were a completely outright. But what if there were a completely different.
AC: we can only work on the information we have at the time.
… But we are looking at this particular approach to scoring and what are the sort of pros and cons on that.
AC: Somebody thought you would want different target scores for different situations, so it would be more of a regulator thing to sort of set that.
… think since Greg points that I cut him off from making earlier, but it was around if you include percentages, that undermines WCAG's use in a legal context.
<kirkwood> Gregg is correct
AC: Uh, another comment, useful, but probably best in WCAG EM.
(Summarizing comments)
<Zakim> AshleeF, you wanted to ask what the process is for bringing alternative ideas forward since it was briefly mentioned by AC
AF: before you started reading the comments again, you mentioned...
… the possibility of bringing alternative ideas forward.
… I was wondering if you could.
… maybe briefly talk about that or send something in writing just because it helps me reference later.
AC: I think a general thing. If anyone has ideas that could, they think could improve conformance. or anything in WCAG, then please do email the chairs,
… maybe a GitHub discussion, or whatever.
<Zakim> Rachael, you wanted to react to AshleeF
AC: you just drop an email to the chairs with a brief outline
<AshleeF> Thank you!
RM: we're not gatekeepers on the idea but are responsible for making sure things move at a reasonable pace as much as possible.
… there just might be some back and forth with the chairs so that you have the most robust proposal with the minimum amount of work possible.
HV: speaking specifically as someone working adjacent and often with regulatory departments of our government, I wanted to stress that even today, with WCAG 2 and its 'binary' conformance model, we are able to regulate… a lot of the problems we have around doing the regulation right aren't in WCAG and standards.
… But that is an important thing to keep in mind. I think that wherever it ends up, we are not like it doesn't depend on us solely to make sure we document everything well in a normative space.
<bbailey> yes please, educate U.S. regulators !
GP: The subjectivity issue has always been and continues to be a challenge, but this is true regardless of whether the requirement is normative or not, sorry, the scoring..
… Currently, you know, subjectivity across the WCG results, you know, in a binary outcome.
… if we introduce a scoring system, it would at least require people to consider the degree of impact severity any metric will decide to use. So right now the model is entirely black and white. And anyway, the subjectivity problem is not solved.
<Ben_Tillyer> +1 to giacomo-petri
<maryjom> +1
Awk: It's very easy to say, yes, the regulators will figure it out.
… I think you know some of us have focus on an individual country, but then there's a bunch of us that are focused on the aggregate.
<Jennie_Delisi> +1 to AWK because it also helps orgs purchasing products sold to multiple jurisdictions
Awk: What we want is we want to provide as much guidance as we can so that when Belgium versus
… Colombia versus the US versus Japan come up with the regulations, that there is as much similarity as possible between them.
<AshleeF> +1 to "scoring is just another flavor of levels"
Awk: Because otherwise, when we're making products that are sold internationally, we wind up with a level of complexity that decreases the chance of accessibility work actually being done.
… the other thing I just wanted to point out,
… because my comment in here was, you know, if we can find a good scoring option, then we'd be happy to include it.
<hdv> +1, absolutely, we want to try and help regulators harmonise with one another to make it easier for us all to comply with regulation worldwide. But we don't have to treat regulators as needing us to do their job
<Makoto_U> +1 to AWK
Awk: but having something to allow us to have multiple levels of conformance is really very valuable for WCAG and for regulators.
Gregg: if we have ambiguous data items, or then we need to make them not ambiguous. You can't solve making them unambiguous.
… or the ambiguity problem by adding a new problem, which is a percentage problem, because now you just have two problems all mixed together.
… percentages does not solve ambiguity at all, it just confuses it.
<Zakim> alastairc, you wanted to comment on scoring / levels
Gregg: with regard to the comment about scoring is just like levels, I agree with Andrew that levels can be very valuable. I disagree that they are anything to do with each other scoring
… Scoring is whichever things I can't figure out how to do, I get a free pass on, whether they're important or not. So percentages and scoring is, I believe, a non-starter, as a number of people have mentioned.
… scoring in progress toward conformance.
AC: (Summarizes results)
… update your survey results if change your mind on your responses.
<bbailey> FWIW, seeing “Oops! An Error Occurred
<bbailey> The server returned a "500 Internal Server Error".”
<Laura_Carlson> Scribe change?
alastairc: moving to question 2
<LoriO> hdv, thank you about the dots
alastairc: this one was slightly more balanced to 'should be included'
alastairc: a comment included that they have utility, especially in context of EAA
alastairc: another comment said that with paths, increased granularity with scope and multiple levels of categories, policy makers can make high levels required
alastairc: so for instance, a egulator might say this particular kind of path needs to be more accessible
alastairc: but this may be more of a regulator thing to do than a WCAG thing to do
alastairc: also comment said it helps to break up the wall of requirements
alastairc: than to the more positives
alastairc: informatively, you could include leels tied to meeting user needs, and see if that owrks normatively
alastairc: this again goes back to start with a simpler conformance model
alastairc: then the must not comments… one said ranking provisions is risky and complicated.
alastairc: another comment said three levels is relatively easy to understand
alastairc: and another said that they were in the spike group and not convinced more levels is the most effective option relative to others
alastairc: in the 'something else' category… including more levels, risks, fragmentation and complexity would need to see the proposed levels and benefits
alastairc: it feels like people see the same thing, but predict different outcomes. That makes it quite tricky
alastairc: does anyone want to add anything to this topic?
<bbailey> +1 for WCAG3 encouraging continuous improvement
GreggVan: trying to get consensus on levels will take an enormous amount of time
alastairc: chair hat on… we went over creating a big table with a score for each provision… with like weighting for author effort
<Zakim> AshleeF, you wanted to plus-one the comment about ranking provisions
alastairc: at the time we discussed involving disability groups and others. It would be in a separate group from the main. And probably we would find they would have lots of different directions
<kirkwood> Listen to Ashlee!
<GN015> +1
AshleeF: I want to +1 a comment under 'must not', as it can put things into 'less important' due to people not understanding the importance…for instance, sign language
<HaTheo> +1 to Ashlee, I think this obscures this and causes opinions to be formed and that is dangerous.
<wendyreid> +1 Ashlee's perspective is important
alastairc: chair hat off… levels/scoring would be based on user need. Like Gregg said it's a consensus process
<Zakim> alastairc, you wanted to comment on ranking things
alastairc: there's the angle of feasability too, for instance, certain things may be expensive, like getting enough sign language interpreters. It could also differ based on whether we leave that to regulators? We would need to be very careful about what guidance to give there.
AshleeF: I understand the consensus model…but to be honest, I can personally candidly say, I don't care. it depends on who is in that room. It is important to get more perspectives actively, and not pulling them in lightly.
<Nattarnoff> +1 to AshleeF, we need to make sure we are including everyone.
AshleeF: re feasability… in a conformance model I'm thinking about, that would not necessarily be considered
AshleeF: as a sign language user I don't necessarily care if a product is WCAG 2 AA compliant, I want to know if you have sign language interpreters, and a way to incorporate those, if there are captions, if those are editable… that's a major problem I run into personally as a deaf user that has vision issues as well
AshleeF: for instance, with Zoom if I use the app, I can customise the captions, but going into Zoom in the browser, it's a different product
AshleeF: it's important to me that accessibility barriers aren't obscured by levels or score
alastairc: we should come back to that
alastairc: for now, on our current topic… not sure if folks wanted to add more… we did have a proposal for 5-ish levels… which is what this topic was about. But we got people thinking we shouldn't have levels. It seems to mbe leaning towarsd having them
wendyreid: filling out the survey, I found that it is actually really hard to separate all of these different issues. Eg it's hard to talk about scoring without talking about levels or partial conformance
<AWK> +1 to Wendy
wendyreid: by trying to isolate the issues, we're making it hard to fully explore the idea, because it inevitably leaks into something else. And it could create barriers, eg as soon as we talk about normative, we have to think about how to enforce it (must be testable, must be provable, etc… and then you spiral out because of that)
wendyreid: so maybe we should explore them as ideas, as they interrelate, as they don't relate… and then once they are fuly form, we can find out whether they should be normative
wendyreid: how we're trying to think about them now is really complicated
<Lia> +1 Wendy
alastairc: we started of gather proposals and what they should achieve, and then we tried to do it bottom up
alastairc: a third option, that floated in the paths subgroup yesterday, could be for us to start working on informative docs first and including them into a conformance model later
alastairc: our next one is…partial conformance
alastairc: in this one, the terminology is quite tricky
alastairc: for people who said it is a must have, it was because we needed to scope and meet partial conformance
alastairc: Marie Jo commented that conformance should be shifted to include tolerance for errors
alastairc: another comment was about allowing a sort of get out for types of content
alastairc: another said it identifies what does and does not conform and cannot be used to imply full accessibility
alastairc: then for those who think it doesn't need to be normative… a comment said people have different ways to define this in VPATs
alastairc: this too would need clear definitions
<Charles> partial and progress are very different
alastairc: conformance is binary by nature, but in practice we think of it as more of a scale
alastairc: then another comment said partial conformance should be part of normative conformative only for specific exceptions, similiar to how it is treated in WCAG 2
alastairc: and one said it is important to include it in the policy and reporting recommendations
<AshleeF> My third comment under the “More levels” discussion (typing to make sure it’s in the minutes) —
<AshleeF> Yes, outputs like Sign Language are challenging to obtain, and there are multiple sociopolitical reasons for this. Personally I think we have a responsibility to at least discuss these and document that they’ve been considered.
<AshleeF> To note one of the sociopolitical items for example since it’s often raised in discussions about “feasibility”: the availability of sign language interpreters. The dominant perspective on the necessity of sign language interpreters is that captions, subtitles, transcripts, etc. are sufficient alternatives. I see it every single day across
<AshleeF> social media and offline. This is objectively false, but for now I want to focus on about the real outcomes of this dominant perspective.
<AshleeF> The number of sign language interpreters is decreasing. My home state as an example, the state government has actually gotten rid of degree programs for American Sign Language interpreting. Of the interpreters available in the state, a significant portion of them are nearing retirement age and there will be less interpreters available because of
<AshleeF> retirement and because there are less pathways for training and educating new interpreters.
alastairc: [reads from the doc]
<kirkwood> partial is not conformance
Ben_Tillyer: a thought I had in the last 20 mins or so… initially I was in the 'must' camp and slowly moved to the 'must not' camp… I wonder if we could make use of assertions
Ben_Tillyer: maybe we can provide a mechanism, maybe via assertions, so that people can show they have done some kind of testing re that their product works for a specific set of personas
Ben_Tillyer: for instance, they might assert that 'users of the scope product with lowvision can get through, but might have a tough time doing so', could have different assertions for frustration of being blocked, things like that
Ben_Tillyer: this idea might need more fleshing out, but wanted to get it out
Matt_King: thinking about the way conformance claims are done… it's a little paradoxial because a conformance statement can be made for any scope, so if you have a large product made of many pages or many smaller products, many parts may or may not conform
Matt: in practice, you may still have the ability to have partial conformance
<HaTheo> +1 Ben, I said something earlier in IRC something "requiring people making conformance claims to sharing the testing method and scenarios covered in testing..." today testing is never complete so this feels more transparent and truthful.
Matt_King: this has always struck me as odd
<kirkwood> +1 to scope comment
<kirkwood> agree
hdv: to respond to Matt_King, in WCAG-EM we have “complete processes”
… this is a way to get around the problem
… although it’s not a normative document
… we ask orgs to report on “complete processes” so that they cannot scope conformance claims to smaller parts
<hdv> complete processes: https://
<kirkwood> +1 to Hidde we did that too
<Rachael> +1 to hidde
hdv: we’d love a way to know how conformant someone is
… but we wouldn’t want it to be a normative thing
… because we’d be afraid that would lower the bar
… but we want to give orgs a way to show where they are relative to the bar
<kirkwood> +1 to Hidde
<HaTheo> +1 to hidde
Wilco: I'm surprised the types of partial conformance haven't really come up
Wilco: that are listed in the survey
<Rachael> +1 to clarifying "partial conformance" in WCAG 3. I will add that to our toipcs list
Wilco: there's the partial conformance to 3rd party and partial conformance langauge, which is like accessibility conformance
<bbailey> +1 to Wilco
Wilco: they seem to be their own topics
<kirkwood> I don’t agree
alastairc: we could, the question is how we do thjat
<Charles> i think we need new vocabulary. conformance is binary. partial is incomplete. progress is relative and could be time bound.
wendyreid: I'm going to disagree with you, alastairc. The bigger question is: who is conformance for… there are two audience, one is regulators, obviously, they'll set the standards, orgs have to meet them; the second is, the end user. That's the real audience. Conformance doesn't mean anything to the end user, 'we meet WCAG' doesn't mean anything
<AWK> Procurement is a major other audience.
<Ben_Tillyer> wendyreid - tried to nudge at that idea with assertions per user need/persona
wendyreid: so maybe we could discuss how do we communicate things… eg if someone goes to a new product that they haven't used before, they want to understand whether it meets their needs. That's the bit that really matters Even procurement folks have a user constituency they need to serve.
wendyreid: how can we help organisations communcate meeting those
<kevin> -1 to end user being a key audience for conformance, as users interest is using the product
Wilco: what's really important is the organisational problems of 'what are you doing to make accessibility better'
<AshleeF> I don't agree either — "level of accessibility" and "conformance" don't correlate and they never have. I'll add that "level of accessibility" and "compliance" also do not correlate and never have. This is heavily discussed in disability studies, critical disability studies (and many other critical studies), disability justice, and the disability
<AshleeF> community.
hdv: +1 to wendyreid regarding users
… the regulation that we’re dealing with at my org is actually requiring orgs to publish an accessibility statement
… and we found what wendyreid said; people don’t know what WCAG conformance means
<Matt_King> q_
hdv: they want to know, for example, “can I apply for a passport” and “what barriers can I expect to encounter?”
<Zakim> alastairc, you wanted to comment on whether these things belong in conformance, e.g. like nutrition labels.
<Charles> another vocabulary challenge. conformance claim and accessibility statement are not the same.
<kevin> +1 to Charles' point
alastairc: chair hat off… should we consider end users as a target for conformance claims, but I can see in our evaluation reporting and policy guidance we talk about some kind of reporting towards users, like Ben's idea around assertions, or something additional to that
<Zakim> Rachael, you wanted to ask a question
<hdv> +1 there is a space both for conformance claims and what's useful to tell users
Rachael: is there a way to have the conformance bar be traditionally what is conformance, and a second level that includesthings like assertions to get much closer to the actual accessibility and usabiltiy for people with disabilites? is there a way we could do this in the standard?
alastairc: there could be, I made a similar proposals
<kirkwood> rather than assertions, what about having “functional needs” rather than assertions?
GreggVan: if we mess things up, the users are the ones who are affected
GreggVan: the purpose of testable recommendation is to get them to keep reaching, and when people claim, they tend to do more
<kirkwood> WCAG is a foundation ot build on. Seems no end users? who else would we have in our conformance claim? Seems strange to me. I’m worried we don’t have enough disability representation.
<Charles> an end user benefits from the outcomes that a claim implies. but do they benefit from the claim itself? it seems that access to the claim would have to precede access to the site in order to have any benefit of the claim itself.
<Charles> +1 to wendyreid that these things are conflated due to perceived risk.
wendyreid: +1 to Charles's IRC comment, conformance claims and accessibility statements aren't the same. But the current problem we see is they feed into each other… it is dangerous to admit to less than conformance. Most accessibility statements aren't terribly informative. There's a fear of mentioning anything, eg there's a known issue regarding keyboard accessibility, they open themselves up to risk
wendyreid: if we can make somerthing more technical, reporting that has to be delivered to regulators can actually be informative, eg with paths.
wendyreid: this could be something to learn from and take some of the fear out
kirkwood: a potential opportunity regarding our current subject…Ashlee mentioned earlier that she could not access information on a site because she doesn't know whether the site has ASL or not… the surfacing of this kind of information is extremely important
kirkwood: maybe we should incorporate functional needs for people with disabilities to bridge some of the gaps
<AshleeF> Looking at the "People with disabilities" section of "Job Stories for Stakeholders" (https://
<AshleeF> issues for a product, if I even I get a response, most of the time I'm directed to conformance reports and/or the accessibility statements. Neither are very informative. Another response I often get as a disabled business owner trying to procure accessible tooling is to the tune of "Thank you for your feedback! We care about accessibility. We'll
<AshleeF> try to get it on the roadmap." After that, it's rare anything happens in a timely manner or even at all. Some of this is out of our hands, but I think it's worth exploring how we might be able to reduce these well-documented issues experienced by procurers and disabled users.
LoriO: if we have partial compliance, if we have partial compliance for people who need ASL, are we also going to have partial compliance for people who are unable to use a mouse? what are we addressing? is partial compliance there for end users or regulators?
alastairc: you meant conformance rather than compliance?
LoriO: ah yes
alastairc: gets all of us… it's a diffferent tangent than the partial conformance we were talking about… which was re visibility
Matt_King: re Wendy's point, that these topics are hard to treat separately, and that corporations aren't incentivised to put useful info in their accessibility statement… maybe our output standard should say here's what you need to tell the world. If you have that normatively, then people will get the info they need. The trick there is to not regurgitate a bunch of technical info that nobody understand…
Matt_King: …, but give info on the product in a way anyone can understand
Matt_King: we haven't decided as a group what the problem is the most important for us to solve as a group
Matt_King: we have a misalignment between producer incentives and consumer needs
Matt_King: (consumers being PWD, procurement people, regulation peopl, etc… right now WCAG does not incentivise producers to do the thing most important to consumers)
<HaTheo> +1 to Matt, similar to what I was saying... if people are making conformance claims they should clearly explain how they tested. It's danger to claim not claim conformance, but it's also dangerous to claim conformance as these are living products that change and introduce bugs and issues, so pretty much everything becomes partial conformance.
<shadi> +1 to "we haven't decided as a group what the problem is"
<Zakim> alastairc, you wanted to comment on surfacing info for end users,
Matt_King: it seems like as a group we haven't nailed down what the problem is
<AshleeF> +1 to Matt_King on having a standard require testing in such a way that the test output is more informative (correct me if I got that wrong!) I wrote the survey comment about pass/fail results being self-documenting.
alastairc: this is tricky in the mixed landscape we have… eg in the Web Accessibility Directive, the regulation for European public sector bodies, they have to produce an accessibility statement including non-conformance and information for end users, everyone has to report is
alastairc: that's starting to happen for commercial orgs under EAA too
<Charles> i would prefer to remove the phrase ‘partial conformance’ and would be comfortable with replacing it with something like ‘scoped conformance’ that permits a conformance claim while also permitting a statement of progress toward more conformance claims or more scope.
alastairc: the difficulty is for orgs who are highly regulated btu not required to publish results or details… they would be punished for doing the thing public sector bodies are doing at the moment but it opens them up to legal attack.
alastairc: my second point is regarding transprancy. Some requirements are pretty clear, if you have audio / multimedia content, then you need X or Y… but for many areas it is a lot less clear cut and would end up being a harder to read and longer list
GreggVan: partial conformance is a contradictio in terminis, we should abolish it
<AshleeF> +1 to Charles about using different language than "partial conformance"
<kirkwood> +1 to eliminating “partial conformance”
<Charles> +1 to GreggVan which is the note i added moments ago
<Laura_Carlson> +1 to Gregg
GreggVan: we could talk about progress towards conformance, eg with columns for major disability groups, it would be informative and see if the product would be usable or not
Ben_Tillyer: re transparency… making assertions won't be easy for product owners and developers who haven't come across this topic before, they may actually end up improving their understanding of the impact their choices have on end users
<Charles> thanks for adding that “Job Stories for Stakeholders” AshleeF i loved contributing to that and advocating the job stories format.
alastairc: we'll reopen the survey
<Laura_Carlson> +1 to Gregg +1 To Gregg partial conformance is not a conformance topic. That's like saying partially complete.
<Laura_Carlson> Conformance means you meet the requirements.
alastairc: we're getting the landscape in order