Meeting minutes
alastairc: does anyone want to (re)introduce themselves, because they're new to the group or changed affiliation?
Annoucments and Intros
illai: nice to meet you all, I am Illai from Evinced. First time for me to join this meeting, honoured to join this group.
Chuck: next week is the first Tuesday of the month, we'll run an onboarding meeting 30 mins before the regular meeting, using the same Zoom link as this call
<LoriO> ];
<AWK> +AWK
<BBailey> https://
<Zakim> BBailey, you wanted to discuss ANSI ?
BBailey: some news to share, WCAG 2.2 was adapted as an ISO standard right after our call last week
alastairc: this way you can pay for the standard, but it is available free on the W3C website
Chuck: I believe it's free on the ISO website?
kevin: yes it is free, but they link to WCAG 2… if you go to ISO you don't have to buy it
kevin: this is quite a big thing, certainly for certain Asian countries, China in particular, who can only use standards from certain standards organisations, ISO being one of them
<kirkwood> Well done!
GreggVan: the strange thing about it, if you go through the front door, it will ask you to pay, but there is link that you can follow that will not charge you… so you have to find the right entry
<Makoto> Japanese national standard (JIS X 8341-3) will be updated by adopting ISO/IEC 40500:2025.
Daniel: will need to follow up with ISO to see what's going on re paid vs not paid link
<Zakim> BBailey, you wanted to ask if CFC on WCAG2 passed?
WCAG 2 changes https://lists.w3.org/Archives/Public/w3c-wai-gl/2025OctDec/0019.html
mbgower: we send these out bi-weekly, and then there's a 2 week window for folks to review
<alastairc> Bi-weekly as in once every two weeks.
mbgower: there's different kinds of changes that we list
mbgower: 'editorial changes' are intended to be small changes that clarify what's already there, quite often they are style changes
mbgower: we also have proposed responses to issues, where we don't intend to create a PR. Where we feel it's worth getting input from the WG
mbgower: finally we also have bugs that we fix, we implement those and let folks know they have happened, this includes typos
mbgower: usually we do batches of about 10-12 issues
mbgower: the current review goes until Thursday evening, you have a few more days to look at it
WCAG 3 conformance survey https://www.w3.org/wbs/35422/conf_20_oct_25/?login=
alastairc: the was opened and we've received lots of comments
<BBailey> https://
alastairc: last week we've been looking at the WCAG 3 conformance model
alastairc: we distinguish foundational and supplemental, and then assertions, which are slightly different
alastairc: option 1 was to have a basic set of prerequisite requirements, on top of which authors can make claims in gold/silver/bronze categories
alastairc: option 2 would be to have many more things in foundational requirements, and then have gold/silver/bronze on top, which would require you to get to a certain percentage for functional needs
alastairc: to give you an idea, at the moment roughly half requirements are foundational
alastairc: looking at the results, there is not much support for option 1
alastairc: there is more support for option 2
alastairc: and quite a few comments and ideas for neither option 1 or 2
alastairc: one thing to tackle is the percentages aspect
alastairc: around TPAC last year we were talking about conformance, and as a group we talked about a certain level of requirements and then a certain level that would be scored or percentage, we didn't decide about the mechanism yet. This would allow people to 'choose' which requirements to apply
alastairc: it's not entirely 'pick and choose' but there is a level of flexibility
alastairc: this survey was intended to be a refinement of that TPAC discussion
alastairc: there were quite a few suggestions around having fixed levels and not having percentages.
GreggVan: if somebody meets WCAG and you add one provision, all percentages go down and they fail to meet WCAG
GreggVan: you change the denominator if you would… so things that passed would suddenly fail
<kirkwood> +1 to Gregg
GreggVan: so for mathematical reasons you have to take percentages of the table
<AWK> Couldn't we update the percentages when new criteria are added in the future?
GreggVan: a simple site may not even involve some things
GreggVan: I'm worried some sites will suddenly fail WCAG where previously they passed when we introduce new requirements.
<kirkwood> I echo Greggs concerns
alastairc: taking chair hat off… I don't see how using percentages would be different to, say, doing an update to WCAG 2.2… existing sites that met WCAG 2.1 will fail the added 2.2 requirements, that's nothing new
alastairc: I don't really see how that is different from how it currently works… the math would have to be adjusted on that basis anyway
<Zakim> alastairc, you wanted to comment on the percentages, and the pick and choose.
<mbgower> +1 to Alastair; a new version has more requirements, ergo something that passed prior can fail new
alastairc: then re 'pick and choose': I see this as the best argument for having some kind of percentages or scoring… currently AAA gets ignored as most regulators pick A + AA. I don't personally have AAA reqs top of mind.
alastairc: if people have to bump their percentage up, they will look at more requirements, deciding which one to pick, and be cognizant of the different requirements, which today they wouldn't
GreggVan: everything in AAA was only in there because it was not testable or that it was impossible to meet everywhere…
<Lisa> hugely disagree with what Greg is saying
GreggVan: here we're talking about adding options, things you don't have to do, because you add one more they now fail because you're adding the percentage
GreggVan: they should fail if you add a requirement, but that's not what you're adding, you're adding one more option
<Zakim> Chuck, you wanted to say I don't think we fail.
<kirkwood> “optional requirement” is an oxymoron
<Lisa> +1 to chuck
<mbgower> @kirkwood we have optional requirements all over the place; they are inherent in the wording of some SCs.
Chuck: I'm not a fan of percentages either, but in their defense… if after 3.0 we publish 3.1, the percentages for 3.0 would still be valid for 3.0, you would not suddenly start failing what you previously passed in 3.0, your results are still good for that version.
<mbgower> (i.e., meet this requirement by doing one of the following)
<BBailey> From WCAG2 5.2.1 Note 2 on conformance levels:
<BBailey> It is not recommended that Level AAA conformance be required as a general policy for entire sites because it is not possible to satisfy all Level AAA success criteria for some content.
<BBailey> https://
GreggVan: but we want people to switch to it. We'd never say when we publish 2.2, that people don't need to switch to it or that they can stick to 2.0
alastairc: but it's adding new requirements so the same thing like with WCAG 2 versions
GreggVan: but we're not adding new requirements, they're options
GreggVan: we don't want folks to meet the easiest requirements, we want them to pick the ones with most impact. If we say the optional ones don't need to be tested that makes no sense
GreggVan: the percentage does not allow us to put one single AAA item in Bronze
<julierawe> +1 to getting some assertions into the must-do bucket
kirkwood: we wont' get less conformance when we do percentages
kirkwood: I'm not sure we can do that
kirkwood: it needs to be clearly state what the whole is and what the smaller number is
kirkwood: then we can figure out the percentage thing
mbgower: I'd like to tackle a few things… first, anywhere there's an 'or' in existing requirements there's an optional requirement
mbgower: we have optional requirements built in to our current requirements. I don't see this WCAG 3 conformance model to be that different
mbgower: secondly, if you have a baseline of requirements that are designed to prevent harm, that's good
<kirkwood> Repectfully disagree with Mke: those are not optional requiremen s with an ‘or’ statement: one of two ways to meet a requirement.
mbgower: there is obvs a dark side to this, when someone intentionally games it… but this lets people have a mechanism to show improved outcomes
mbgower: this is not either/or, there are nuances to this
GreggVan: the argument we have optional requirements is not correct we have optional methods
GreggVan: we all agree WCAG 3 is going to be better than WCAG 2. The key to making it better is going to come from figuring out how to get things in that we couldn't get into 2.2
GreggVan: we need to figure out how to get the untestable ones used in practice
<Zakim> BBailey, you wanted to speak to "no one meets WCAG now" camp (if chair still looking for that)...
<BBailey> https://
BBailey: in the context of US federal government, Section 508 requires WCAG 2.0 AA and close (self) reporting uses percentages despite AA being everywhere all the time, so I'm not overly alarmed by percentages
alastairc: chair hat on… to take a step back… why we're looking at this in the first place. We had a few requirements where we wanted to encourage people to do more
alastairc: the alternative to percentages is basically that we define the levels
alastairc: how do we get people to keep going, or help them understand. If you aim for AA, the rest doesn't matter… there's no improvement path, you meet it or not. This is what we wanted to tackle with percentages
alastairc: this is how we got to the point of having a lot of requirements. We can have more nuance in the requirements too, than we have in WCAG 2. If everyone has to meet everything at the same time, I don't think that's going to be an improvement
<Zakim> GreggVan, you wanted to say % conformance is different than % required.
GreggVan: Bruce was talking about % towards conformance, but that's different from % that is required.
GreggVan: I'm worried folks will put the things they don't want to do, in the percentage of things they are not going to do, and then just add the things that don't even apply in the percentage they do want to meet.
<Lisa> +1 to mike
<Zakim> alastairc, you wanted to comment on what's important (/severity)
alastairc: on the topic of percentages… there's been a few comments around which reqs are more important than otehrs… if people have been doing the testing exercise that we had going for the last couple of weeks… you can see which ones we had as foundational and supplemental
alastairc: hopefully you've seen that… going through the exercise, at the point where we had mostly foundational, I had a certain client in mind when I did this, and thought they met about 50% of requirements, but they actually missed many of the foundational ones.
<kirkwood> +1 to Mike (re: participation)
alastairc: all requirements we've got down are important to some in some scenario.
alastairc: would anyone who has not yet commented… have thoughts about percentages vs defining levels like A/AA/AA?
<kirkwood> levels
<Rain> +1 to defining levels
giacomo-petri: not really related to percentages… two important aspects to mention, one is severity… for the same criteria there can be so many different severity, depending on type or purpose of the element
giacomo-petri: eg an image that is decorative has very different impact on end user
giacomo-petri: secondly, it was mentioned previously that percentages could stop progressing on accessibility… many organisations struggle to catch up on accessibility issues and getting in control of the issue. So it's not just knowledge and progress but also the impossibility to reach conformance for some
<GreggVan> invalid statement of choices. it is not % or other way to get AAA in -- % does not get AAA in . We need to ask "if % wont work - what can we do to get AAA topics into practice since we cant require them. AND are there some things that are in AAA that we can make testable or practical so they get into requirements
shadi: I strongly support what giacomo-petri was saying
<AWK> +1 to Shadi
shadi: we've done a lot of work on breaking down requirements to more specific ones, like 1.1.1 is now much more granular. Looking at that from the perspective of severity… might there be differences in impact eg between decorative images and lack of captions? Can we try to find something like a treshold, to account for 'things you've missed'
<Rain> +1 to Shadi, the concept of the negative score is very compelling
alastairc: we can continue to work on conformance but need something for our next draft and I'd like us to focus on that for this meeting
LenB: I missed the survey deadline yesterday and emailed my comments
<LenB> * Accessible: Prerequisites met + severity score below threshold.
<LenB> * Functionally Inclusive: Accessible + balanced severity coverage across all disability cohorts.
<LenB> * Optimized: Accessible + low severity + inclusion of supplemental features.
LenB: the idea of being more granular was intriguing to me.
LenB: I've come up with a categorisation I posted above
LenB: which kind of corresponds to a severity score
<shadi> +1 to LenB
<kirkwood> +1 to Len
<kirkwood> Like his ideas
<Zakim> GreggVan, you wanted to say suggestions for next draft conformance section
LenB: I am curious about when I do something like procurement… if I had this in front of me, this would help me make decisions better than percentages
<BBailey> +1 to Len B but I think "Accessible" should be foundational set (not prerequisites met)
<shadi> +1 to defining our goals for the conformance model i'm not sure we have that clearly documented and agreed to
GreggVan: maybe we can collect ideas and then solicit for each what the pros and cons are
<Zakim> alastairc, you wanted to comment on how to we (centrally) account for the variety of contexts?
<kirkwood> Whenever percentages. Explicitly state the numerator and denominator
alastairc: re we define the levels… if we have levels like how LenB proposed, or A/AA/AAA… if regulators agree on a particular level, do the other levels matter?
<Zakim> Chuck, you wanted to react to alastairc to ask for scribe change
SAZ: Looking through conformance discussions, not sure we have clarity on goals for conformance model
… Would be a step forward
… on AC's question, what would additional levels bring?
… having something between levels to allow for a progression would bring some benefits for orgs
… or perhaps there are different levels for different sectors
… more options might give rule makers more options
Kirkwood: Want to state how well it has worked in the past
… having 3 levels was impt when I had 1000s of sites to deal with
… AAA was aspirational and for a goal
… AA was the legal bar
… A was a stepping stone
… worked well for that progression
… in courts, in policy, etc.
<julierawe> +1 to Kirkwood
Kirkwood: we should build on that
LS: put in a suggestion in the survey and a scratch pad for it
<Lisa> https://
LS: I can summarize now or later
… four buckets - two required
… core which is criteria for basic use by PWD and doing no harm
… can users do the basic action
… we can break up criteria info smaller parts. Decorative text wouldn't be core but alt on a functional image would
… then criteria broken up by functional needs
<Ben_Tillyer> That example works until you consider a journalist who requires the alt text of the "faculty photo" in great detail. It's impossible to determine the functional need as everyone visits webpages for different purposes.
LS: focuses on higher impact items to equalize between disabilities
… meant to be a starting point for discussion
AC: Structure is similar to what has been proposed but difference in how items are assigned?
LS: Yes
<Zakim> GreggVan, you wanted to say 1) Bullet list(s) far superior to explainer text. Start with GOALs. 2) Progress isnt making more things skippable.
GV: two comments
… 1. short bulleted list of goals will encourage review and feedback
… The fact that people aren't fully conforming and making it easier to make more people conform is a false logical approach
… I think that if we find another thing to add, we should add it
… we don't want people to quit working sooner
… people need a clear target for work
<kirkwood> +1 to Gregg
GV: Lisa's work looks useful. We need to get more items in
… We should get some number (30%) into browsers to reduce level required by authors
<kirkwood> It is already being done in the browser with ai
GV: browsers and Gen AI should help and then we can get a shorter list
<Zakim> alastairc, you wanted to comment on "skippable"
AC: two points
… on AI bit - discussion at TPAC on this
… in meantime we should include req and if AI helps then it is automatically passed.
… also (chair hat off) wanted to mention that if we have a set of things for people to do
… meeting mins across functional needs
… I don't see things in there that are very skippable
… if those types of content are included the % that needs to be done goes up
… I don't see %s as making things skippable or equivalent to AAA
… supplemental is made up of bits of AA, some AAA, process things
… we seem to have one primary decision for the next dfaft
… we can put in a proposal and also share goals and signal that it isn't mature and need feedback
… how to deal with?
<alastairc> Draft Poll: For the next draft do we continue with foundational + percentages (answer "P"), or we define all levels (answer "D")
<GN015> I feel we haven't discussed the new suggestions sufficiently to decide this already.
Jennie: COGA reviewed info on slides showing categories yesterday
… if a company wants to compare products to compare accessibility
… each company could be reporting on different things
… may complicate that sort of comparison
<GN015> +1 to Jennie
<kirkwood> +1 to Jennie
Jennie: impacts end user experiences ultimately
GV: What about listing goals?
AC: little of both. Editing.
<Chuck> Draft Poll: For the next draft do we continue with foundational + percentages (answer "P"), or we define all levels (answer "D")
<alastairc> Draft Poll: For the next draft do we continue with foundational + percentages (answer "P"), or we define all levels (answer "D"), or remove all and list goals + ideas ("G")
GV: Suggest third option to share issues and goals
<kirkwood> agree with Goals + pros and cons
AWK: Perhaps instead of % we say "foundational + some extra"
<Zakim> alastairc, you wanted to comment on conformance vs reporting
AC: Separate what is in core conformance vs what we can enable people to report on
… WCAG 2 conformance is short and concise
… don't want to overcomplicate
<BBailey> +1 to Jennie's concern for comparing accessibility of two products (with products being able to have different sets of criteria) -- but I think that already happens currently with the VPAT model.
AC: people can do all sorts of reporting
Julie: When you say "define all levels" it is just like WCAG 2? No choice, levels explicit?
AC: Yes.
… We pick what goes in what level, like in 2.x
… lots of ideas - different sectors might have different levels adopted for examples
… some things we want more flexibility on
… This was a choice we didn't make last year
(drumroll)
<BBailey> We made the choice NOT to do that
<alastairc> Draft Poll: For the next draft do we continue with foundational + percentages (answer "P"), or remove all from the draft and list goals + ideas ("G")
<BBailey> We didn't avoid making choice ( i think )
<alastairc> Poll: For the next draft do we continue with foundational + percentages (answer "P"), or remove all from the draft and list goals + ideas ("G")
<GreggVan> G
<kirkwood> G
<Ben_Tillyer> P
<alastairc> P
<stevef> P
<Illai2> P
G
<shadi> G
<Chuck> P
<BBailey> P
<GN015> O - Other: we should discuss the ideas first.
<LenB> G
G
<Mbgower> P
<graham> p
<julierawe> G
<Rayianna> P
<Laura_Carlson> G
<Chuck> 8 P, 7 G
<jtoles> G
<Illai2> +1 to GN015
<Jon_Avila> G
<Chuck> 9 G, 9P
GP: In survey we had options 1 and 2
<Chuck> Now: 10G, 9P
GP: If had to decide it does depend on the percentages
<Zakim> Chuck, you wanted to ask if we can do both?
Chuck: No consensus.
… Should be do both and get external feedback?
<Detlev> well, just not include "one or the above" if the choice rests on a prior decision :)
AC: safest approach is to not do a ton of work until we are more focused
<Detlev> sorry "none of the above"
GV: "G" includes "P". Here's the goals, P is one option.
<Zakim> BBailey, you wanted to ask about people with objections ?
BB: thought I saw some neither answers?
AC: not there yet, need better definitions
… let's spend the next few minutes talking through these options
… to help the chairs
… John Toles, can you share?
JT: Feels like the objection to the foundational set is that there are too many items to meet
… need a set of principles to guide what is in foundational
… e.g., no harm
… or things that can be commonly met or supported by browsers
… let the numbers come out of the principles rather than pre-defined
<BBailey> https://
BB: I wasn't thinking to use AA/AAA but need a was to help sites getting started on accessibility
… "preliminary prerequisites"
AWK: I think the fundamental concern that I have, I'm worried that the credibility of WCAG is somewhat at risk if we just raise the bar higher and higher.
AWK: From my experience with sites I've worked on and interacted with, none of them meet all the standards, there are so few sites that do.
<BBailey> +1 for AWK on unilaterally raising the bar higher
AWK: We should define the highest bar, and the middle.
<kirkwood> +1 to highest & middle
AWK: Its problematic for us to say that WCAG 2 AA is basic level, when only 10 sites meet that. I think we need to have something that represents a broader spread.
AWK: I feel like the goals of conformance should include that. As much as I want to say that WCAG is the solution to all accessibility issues, it is very difficult for sites to meet everything.
AWK: The web has gotten better, experiences are better. If sites are not fully conforming, but people are having better experiences, maybe we aren't describing it in the most effective way.
LS: Wants to point out that my proposal had 4 levels
<shadi> +1 to AWK
LS: helps sites move forward incrementally
… after core you move to foundational, etc.
… some items didn't make it into AAA in 2.x because they weren't viewed as widely needed or applicable
GC: just wanted to say running the GL through a site we know well is very eye opening.
… suggest doing the exercise
(Due this week!)
GN: Would like at least one level across sites
… "basic" accessibility
… no blockers
… like LS's second level?
… then allows focus on individual user need groups
… sticks to B/S/G in comment but could be different
IZ: Talking about scoping instead of leveling
… We see 55 SC's looking at every page is overwhelming to orgs often and they don't look at the most important things some time
… looking at critical paths/flows may be an alternative.
… critical pages have a higher bar, less critical pages have less high a bar
… allows users to complete essential tasks while keeping bare min reqs for other pages
… less overwhelming
AC: For context, we have looked at how to do more task-based and user-journey base before
… it is something that we can't figure how to do yet
… how do we get people to consistency choose journeys? There are challenges in matching the variety of options for users
<kirkwood> What is the motivation for changing it?
<BBailey> +1 to AWK concern
<Zakim> mbgower, you wanted to say I'm not hearing as much disagreement as others on this call; more tactical differences towards the same outcome. We ended up breaking down our guidance based on what to prioritize first https://
MG: want to end on optimistic note. Tactical differences.
… added link to IBM toolkit - teams were giving up if they didn't think they could pass
… teams need a toe hold
<kirkwood> +1 to MGower!
<BBailey> +1 to how levels defined in IBM pace of completion
<Ben_Tillyer> +1 to MGower
MG: worry that we will get outpaced by technology
<GreggVan> +1 to getting outpaced
LS: Want to point out that we aren't policy makers and we lose something when we try
… deciding levels gets too close to policy
<mbgower> /me I forgot to say that it's the level of prescription that I see as a key risk. The more prescriptive we are about what teams MUST do, the less we provide ability for someone to figure out what is most important for their users. For example, here's how we broke down the design guidance by our levels. Someone else's may be different, but we would
<mbgower> end up at the same end goal. https://
LS: every consultancy firm I've worked with focused on critical content first. Same with governments in settlements.
<kirkwood> +1 to Lisa
LS: we shouldn't be defining what gets left out
<Glenda> Here is what AWK said earlier (when we had no scribe): “I think part of the reason why this is particularly important for us to think about and figure out how to deal with... The regulators don't do that. They pick a level, and so if we don't speak to this, as an issue, that is important. Then what we wind up with is something like what we have now, where they say, every pageon the internet needs to meet WCAG 2. The result is we aren't able[CUT]
<Glenda> that. So I think we have an opportunity to try and change that and set a direction. If we do it, then it has a chance of being adopted, and if we rely on policy makers to do that, what we wind up with is a stronger risk of deharmonization.”
AC: We are trying to define conformance here but also plan on a doc for policy makers
GV: We shouldn't be judging what is important
… all people should be able to access information
… on MG's outpaced comment. Makes me think we shouldn't hurry. shouldn't put out something that will become quickly obsolete
… need to look at what is happening with the technology
… need to survive new stuff
… think about screen readers. We don't require all content be voiced, we focus on interop
AC: Chairs and I will work on this