Meeting minutes
<ChrisLoiselle> regrets top of hour
Chuck: Welcome everyone.
… Any new members? (None)
<Chuck> WCAG2ICT Update Pre-CFC: Issue 750: AG WG Review: WCAG2ICT update to align with EN 301 549
<maryjom> w3c/
Chuck: opportunity is the first meeting of any given month, and we'll present 30 minutes of onboarding.
… If there's anybody interested in attending, you are certainly welcome to join.
… Review for WCAG to ICT, an updated.
… please revise the issue and lookout for an email.
Mary jo: that. Basically the changes that we're making right now is to make an alignment between the EN301549 standard that's being worked on for the European Accessibility Act.
Chuck: chairs and staff contact requested a charter extension.
… we were granted it. So, our current charter has been formally extended.
RM: This is just that extension that we had previously discussed.
<LoriO> yes,
LO: why do they want it every 6 months?
Schedule Check-in: https://docs.google.com/document/d/1dbxHHLFQW_y6BABEUrrtDfLnLCiT2QycYvViNAkQ8Gk/edit?tab=t.0
LO: the extension we were granted was 6 months, because whenever they grant an extension, that is the minimum.
… But whenever a charter is finalized, it's for a period of 2 years.
<Chuck> https://
<Rachael> This is the formal timeline: https://
Chuck: I'll share my screen for a schedule check-in
<Rachael> The document Chuck is reviewing is a more detailed schedule for the next two Quarters
(Chuck goes over the AGWG 2025 Q3 and Q4 Schedule document)
<Rachael> Subgroup participation survey is at https://
Chuck: These are topics that we intend to discuss. We will begin these today.
If anybody does want to review these and have any questions, please feel free to reach out to the chairs.
… please review these.
Conformance - how to scope a claim
<Rachael> https://
RM: This plan for our discussions.
Shadi: Question #7 I would like to propose something else.
<hdv> +1 to shadi
Shadi: what does conformance at different levels mean?
RM: I would like to keep the 100% in if we can, just to be clear for people.
… as it triggers other conversations.
<Zakim> Chuck, you wanted to react to shadi
RM: we will focus on point #1 today.
Kevin: Question 7 regarding something at less than 100% conformance - we could probably remove that.
RM: I don't necessarily know we want to spend our time in this meeting refining this list.
… If you have concerns about the wording, or suggestions on alternative wording or sub-questions, just add comments here or email chairs.
Gregg: requirement applicability for conformance is really important.
… What do you make the claim against? Do you say, for the whole site? And how do you know? Because how can you make the claim, since if you're a large corporation, overnight.
… There have now been added at least 1,000 pages to some of these major sites.
RM: Terminology (on slide #4)
… (Reviews each one for a shared vocabulary)
… big question from our point of view, is what is our minimal unit of conformance? So in WCAG 2, as Greg mentioned, it's the page.
… Conformance units could potentially be components, pages, views, processes.
<Rachael> https://
RM: component is the big question.
… Right now, we have conformance scope. It includes views, pages, and processes.
Gregg: I think you should say the conformance unit is page, set of pages, & process.
Gregg: Second comment is evaluation. Are those are the conformance unit or the evaluation unit? And I'm not sure whichever one it is.
Shadi: just wondering, would it always be a note, or a rec?
Hv: I think it makes sense to leave that open.
RM: this is probably a note. specifically for the policymakers section.
Gregg: Don't want to have testing methods be normative.
… We are not charged with setting rules by which regulators and legislators have to act.
<Zakim> shadi, you wanted to clarify my comment
Shadi: I did not mean to imply any of these notes in particular should become Rec.
GP: Shadi's idea very interesting.
… Having a perfect website is quite impossible.
RM: WE have at least in all the drafts we have been circulating amongst our group discussing writing a note to policymakers.
<Zakim> Chuck, you wanted to discuss minimal unit of conformance
RM: what do we think in WCAG 3 as our minimal unit of conformance?
<Jennie_Delisi> +1 to Chuck discussing the need for the ability for libraries to indicate widgets conform or don't, or at which level
Chuck: I'm intrigued by the minimal unit of conformance being a component.
Gregg: we're talking about the minimum level of conformance. What is the term that you're going to use in one of our requirements?
… I think that what you're claiming is what you're claiming, and I don't think that we're going to write a requirement that says you must do this thing for the entire scope of your claim.
… It seems a little bit redundant to say, I'm claiming that for these pages, I'm following your rules.
One of which is that I follow this one particular rule for the whole claim.
… RM: I see Gregg's point about it being an evaluation, but we also have to scope the conformance statement itself. I think a question that ties to this is, could you declare conformance against a component?
<Zakim> Jennie_Delisi, you wanted to address Gregg's last point
Jenny: to address Greg's last point, I actually found that very intriguing, because I'm wondering if, in the main bullet 2, if we need another sub-bullet that clarifies the concept that Some requirements could have
the need to be evaluated at that smallest unit, But also, at a larger unit.
Gregg: Yes, can you claim conformance against a component if this if the requirement says that the component needs to do this.
… In WCAG 2.x we said, we didn't put process in any of the individual items.
… The unit of conformance is not going to be elements or things, so it's going to be units of evaluation.
… But the unit of conformance has to be something meaningful. Like a page, or a process.
<ChrisLoiselle> I had been on queue , I'm not now, not sure why
Chuck: If you have a perfectly accessible application, except for one bug on the login page. How accessible is your application? Well, if nobody can log in, not very.
<Zakim> Chuck, you wanted to say minimum isn't exclusive
<ChrisLoiselle> Hi, I was on queue and was removed.
<ChrisLoiselle> I will just post my comment below, as I have to drop at 12 pm ET.
<ChrisLoiselle> On slide 5, if you are referencing conformance , where and when does page templates come in to play? Component vs. process or view? What makes a page template a workflow? Is the workflow then conforming and not conforming?
<maryjom> +1 to Rain.
<ChrisLoiselle> A transactional template could be a workflow and a process, however is made up of components, thus would someone take a component and approve then vs. within a scope of a page , view or process.
<sarahhorton> +1 to Rain
<Zakim> Chris, you wanted to react to Chuck
<Zakim> Chuck, you wanted to react to Chuck to make a point of order
Rain: think we need to be really careful that about anything that we do in our in our testing. At the end of the day, the flow is the most important thing.
But we need to recognize that the people who are using this are also going to need to apply it on some cases, at very granular levels.
Chris: You have a component that might be part of a design system. And then you have what I consider authoring tool type of systems where you have page templates, or Screen-type templates that are using components. So whether or not somebody's validating the conformance within that flow versus the component.
<Charu> +1 to component libraries
Mary Jo: We do conformance by component as well. You know, you don't do a final conformance over the whole thing until it's used in a page. Conformance is useful on a smaller level as well.
<Zakim> Rachael, you wanted to talk in forms of conformance and compliance instead of accessibility
<ChrisLoiselle> someone could use a component within a flow, however "break" the component if not coded per spec, so they could conform by using it as intended, but within their page template if they don't author it correctly, could cause the template or workflow not to conform , i.e. images used in a template don't have alt even though they should and you are allowed to author it but don't.
<sarahhorton> +1 to Rachael
<ChrisLoiselle> need to drop for another project call, apologies.
Gregg: We do want to be able to have some way of describing components,. And I'm not sure how to say what it is, but it's really important.
<Zakim> Chuck, you wanted to ask about Chris's point
<Rain> +1 conforming at the component level only applies to the component in its component form
RM: I guess going to the point that was made from a scoping standpoint. If the scope is a component I don't personally see why we couldn't say it conforms.
Scribe change?
<LTSzivos> The easiest example of accessible component versus page template could be misusing form elements to create a change of context without forewarning the user. The component is accessible; the template is not.
<sarahhorton> I have to drop, sorry!
Chris: so you have a component library, that component library has templates, and those templates are made of components. You do your first blush of conformance on the components, and the components themselves are accessible.
To the extent they can be.
<LTSzivos> Is there documentation on how to scribe for those who may be willing to try but are loathe to do so because of their inexperience/lack of knowledge?
Chuck: Using linear queueing at this point. If you have a response, queue in to offer a response.
Hidde: Because of customizability, component conformance not meaningful. The components can affect one another, and is rarely where accessibility happens.
<Zakim> Chuck, you wanted to ask about hdv's point
<Zakim> GreggVan, you wanted to say we should perhaps add a term "component level conformant" conforms to all requirements that apply to component or lower level requirements for present contents and will not break conformance if what you add to it or include it in is conformant. and to say we should perhaps add a term "component level conformant" conforms to all requirements that apply to component or lower level requirements for the present contents and setting and will not break conformance if what you add to it, adjust it to, or include it in is conformant.
Hidde: useful in understanding if accessibility was considered in development of the components, but nothing beyond that.
gregg: Perhaps we should have a term "component level conformant". It would mean it conforms to all requirements that apply to component or lower level requirements for the present contents and setting and will not break conformance if what you add to it, adjust it to, or include it in is conformant.
steve: From a practical standpoint, we have customers that ask for the component library to be tested. We can test for that and provide an ACR to indicate that the library has been tested and is conformant. It doesn't mean that when the library is used, the content using it is automatically conformant.
giacomo-petri: We need to understand the applicability of the requirements to the components. For the heading requirements, a component would naturally conform because they don't need or use headings. It doesn't mean the final page using the component would be accessible.
<Zakim> Chuck, you wanted to say in Oracle I advocate for our library authors doing validation.
<Zakim> Jennie_Delisi, you wanted to discuss low code, no code options libraries may want to include multiple levels
giacomo-petri: is it possible to say something is excluded from the conformance. For example, if I have something that is inaccessible (3rd party) can I scope that out in some circumstances since I may not have a choice in using it, but I also cannot fix the inaccessibility of that content.
<Zakim> joryc, you wanted to say there is no such thing as a conformant component.
Jennie: In a low/no code tool, a user would need to know if the components are accessible.
<Zakim> hdv, you wanted to mention stakeholder expectations that will come when teams say 'built in' / 'accessibility ready'
jory: Don't think there is a "conformant component" Labels or color changes could break the accessibility. We could only talk about that internally that a component was designed with accessibility in mind.
hidde: Agree that you can't say more than a component was designed with accessibility in mind, stakeholders, from what I've seen, will love to hear conformant components exist, but it's a big risk for a11y if we allow for that kind of statement.. Don't want consumers of libraries to assume accessibility comes without any additional work when using accessible components.
<Zakim> Rachael, you wanted to ask if we moved forward with some type of component level conformance should we include a statement about pages
<Zakim> GreggVan, you wanted to say " add to my previous definition a note that says "best practice is to also include guidance on maintaining conformance when content or adjustments are made to the component" and to say " add to my previous definition a note that says "best practice is to also include guidance on maintaining conformance when content or adjustments are made to the component - and a note on "use of this component does not ensure conformance of larger unit that uses this component"
Rachael: If we made components conformant, should we add a statement in the conformance section that simply says that conformant components don't necessarily lead to conformant pages.
<Zakim> Chuck, you wanted to say I advocate for our library content authors doing some validation.
GreggVan A component library that is conformant could state that this doesn't guarantee conformance in the resulting pages using the components.
Chuck: Have had that problem with our component libraries. There is benefit though in saying that the components have been built to support accessibility, but you still have to test where they are used to ensure the end result is still accessible.
<Rachael> draft poll: 1) allow conformance for components 2) add an accessibility ready approach for components to evaluation methodology 3) neither 4) other [add approach]
steve: If I'm testing a component in the library and test it for what is relevant to that, I could say this component is conformant. However, when used in the wild it can be used in an inaccessible way - and that would be an issue that gets tested/reported.
<GreggVan> Rachael -- did you mean "conformant ready" ?
steve: If you start out with a component that supports an accessible name and keyboard accessibiity it is a good thing to know and a better starting point than something that you don't even know its status.
<Zakim> mbgower, you wanted to say that you can report on a component that is alone on its page (or whatever context)
<Zakim> hdv, you wanted to respond to chuck's question if there is any things people could test on component level
mike: You could report on a component that is alone on its page. You could mark it as compliant but don't think conformance makes sense on a component level.
hidde: There is a list of things you could test on every component and you could retest it when it has different iterations, versions, and customizations. You could also notate that when certain aspects are changed, you'd need to retest.
… If you could have the ability to explain all of that, its ok but if not it wouldn't make sense to have conformance for components.
<GreggVan> @rachael add option to poll. x) not have conformance at component level but create a special term for labelling components that are meet applicable requirements as they currently and if used properly help to build conformant content
<Jennie_Delisi> +1 to Giacomo!
<Chuck> poll: 1) allow conformance for components 2) add an accessibility ready approach for components to evaluation methodology 3) neither 4) other [add approach], 5) not have conformance at component level but create a special term for labelling components that are meet applicable requirements as they currently and if used properly help to build
<Chuck> conformant content
giacomo-petri: If I make the effort to make a component accessible, I'd want to claim I worked on accessibility and that it meets the requirements. I may have to make statements that if certain aspects are changed, then you'd need to retest. Doesn't see a problem with component-level conformance.
<Chuck> ack
<Zakim> Chuck, you wanted to announce Rachael's poll
<shadi> 5 (which is also 3 & 4)
<mbgower> 4 how about adding a 'library' unit?
<GreggVan> 5
<hdv> 1+ to Giacomo's that people who make components want to claim conformance, but it's not the same 'conformance' as a page, not the same meaningfulness.
kevin: Understand the perspective that component-level conformance can be problematic but if we don't do something, the accessibility community will do it instead - in some ad hoc fashion. I think we need at least some commentary on that.
<hdv> 5
shadi: Agrees with Kevin and Hidde's points. Can we develop something to address that - even for WCAG as it exists today. It could be in a Note and not necessarily WCAG 3.
<Rain> 1 - but make it clear that the conformance only applies to the component individually in its default state, not how it is used in a larger context. As Kevin noted, if we don't have a way to do that, industry will have to get creative.
<hdv> +1 to Shadi that a doc to address this is helpful, would love to help write it and feel we have more folks in this group that could make an effective / clear doc around this
<Chuck> +1 to Gregg and encouraging component content authors to do some testing
<SteveF> we don't need a component conformance unit, it is like saying we need a unit for a HTML page with only a heading and a paragraph of text
GreggVan: Suggestion 5 in the poll attempts to do this. This isn't a requirement, but a reporting issue. It would be useful to have some special information. Maybe not saying "conforming with WCAG" but define some way to make some level of claim. We have to have some way of encouraging development tools, templates, and component developers to do
the work.
Chuck: The poll had some flaws and not many responses to it.
Rachael: I'll create a modified poll.
Chuck: One more item in the deck we need to get to once the poll is reposted by Rachael.
<Rachael> poll: How to handle components: 1) components can conform via WCAG 2) Components can do something slightly different than conformance in WCAG 3) Conformance can do something slightly different than conformance in a Evaluation Note 4) other
<shadi> 3
<GreggVan> 2
Rachael: introduces the poll.
<hdv> 3, where we need to make sure 'slightly different' is written in a way that doesn't provide an escape hatch
Rachael: We would need to work out the details, but the poll explores where the conformance is and how it is documented.
<Rachael> poll: How to handle components: 1) components can conform via WCAG 2) Components can do something slightly different than conformance in WCAG 3) Conformance can do something slightly different than conformance in a Evaluation Note 4) Conformance at a library level 5) other
<Zakim> Chuck, you wanted to ask m
<Chuck> +1 authoring tools
<Zakim> GreggVan, you wanted to say 4 - because it will have more encouragement to tool/component makers if it is formal and not some side note. HOWEVER Notes have the ability to change and help us develop the idea
mbgower: Currently conformance is on an author created page that is served. This is about certifying elements they're creating to sites. Maybe this is an authoring tools thing.
GreggVan: Modifying conformance is a big leap and would take a while to develop. Could more quickly address this in a Note that develops ideas on how a component library might document the work they've done to make components capable of conforming.
Rachael: If we did something like that, would we also want to put it in as "developing" content with an editor's note tying it to a Task Force effort that is working on a Note.
<Zakim> GreggVan, you wanted to say "experimental or exploratory"
gregg: Would prefer it be "exploratory".
hidde: Heard a lot of voices that this level of conformance wouldn't be meaningful so hesitant to call it exploratory or developmental. But does think a Note could be useful to address it for now.
shadi: Clarifying that this might be a separate Note. A little concerned about adding complexity to WCAG 3 and would prefer it be worked on outside of WCAG 3.
rachael: Let's add the concept of a separate note to the poll
<Rachael> poll: How to handle components: 1) components can conform via WCAG 2) Components can do something slightly different than conformance in WCAG 3) Conformance can do something slightly different than conformance in a Evaluation Note 4) Conformance at a library level 5) Create a separate note about component accessibility 6) other
<Chuck> 5
<shadi> 5 or 3
giacomo-petri: If I am not able to determine if my component or product is conformant or not. If I ask for a VPAT, you can see more granularly what requirements are fully met or not. It might be interesting to know which requirements can or cannot be applied at a component level.
<Zakim> Chuck, you wanted to react to giacomo-petri
<Zakim> hdv, you wanted to respond to giacomo-petri's q to recommend component producers to create example pages for conformance claims
<Rachael> 3 (chair hat off)
<kevin> 5 or 3
<GreggVan> 3 and 5 which seem to be about the same we can visit wcag version later
hidde: Giacomo makes a good point that someone creating components wants to be able to say something about conformance. It would be good for them to provide a page that demonstrates appropriate use of their components in making a conformant page.
<hdv> 5
Chuck: Regarding the poll, most votes are 3 and 5
<Chuck> poll: How to handle components: 1) components can conform via WCAG 2) Components can do something slightly different than conformance in WCAG 3) Conformance can do something slightly different than conformance in a Evaluation Note 4) Conformance at a library level 5) Create a separate note about component accessibility 6) other
Gregg: What is the difference between 3 and 5?
<Detlev> I don't understand the options well enough to pick one.. but probably 3
<GreggVan> 5
<LenB> 5
<elguerrero> 5
<Azlan> 5
Rachael: It's a level of effort difference. 3) conformance can do something different and is already a part of a note we are committed to writing. 5 is standing up an additional group to develop a note on its own.
<Frankie> 5
Chuck: Poll results: 5 is the most popular option.
<Zakim> Rachael, you wanted to ask question
Rachael: When we have something that hooks that go into WCAG, we need the hooks that go into it. Should sub-groups be looking at the requirements list they're working on to determine applicability at a component level.
GreggVan: The separate TF should be the one to go through everything to see if the requirements make sense to apply to components. The working group and its sub-groups should focus on the requirements development.
shadi Maybe a community group should incubate the idea. It's a step away from the main focus of WCAG 3 work and don't want resources diverted on a side project.
Rachael: For next week we want to clarify the difference between conformance and reporting.
… does slide 6 cover the points about these two topics? Please add comments. We'll discuss this next week.
… Goes through the points currently in the slide on "conformance" and "reporting (compliance)". The goal is to clarify the vocabulary.
Shadi: Bullet before last - in "reporting" on the slide could map to different levels of accessibility
Chuck: Rachael made some tweaks to that.
subgroup-planning https://www.w3.org/wbs/35422/Path_Participation_25b/
Chuck: We are looking to continue the sub-group work. We put out a survey on which groups you would prefer to join and which is your secondary choice. Not much response - only 6 respondents.
… please fill in the survey. If you have an interest in a particular group whether you're currently in that group or want to move to another you've got to indicate that in the survey.
… please put your interest in IRC right now if you know or else complete the survey.
<CarrieH> I responded
<Frankie> I thought many of us already responded that we were not changing groups?
Rachael: Encouraging everyone to join a sub-group. It's critical to getting WCAG 3 work done. You can contribute more and get to know others better.
<Azlan> Do these subgroups have meetings already scheduled? If so can times be shared?
Rachael: If you responded that you want to stay in the group you were in, that's fine.
Chuck: Sub-groups did have a schedule that the groups had. Will need to find that and send it out again.
<Frankie> Image & Media Alternatives & Controls meets at 5pm Pacific on Thursdays.
Rachael: Some groups have flexibility to change their times. We don't have time in the AG WG meetings to do sub-group work as we work through conformance topics.
<LenB> Plain Language, Consistency & Familiarity is open to considering new time slot.
<Jen_G> I'm interested in staying with the inputs group. Apologies for not completing the survey.