W3C

- DRAFT -

AGWG Teleconference

27 Aug 2024

Attendees

Present
dj, alastairc, Rachael, Ben_Tillyer, Francis_Storr, kevin, Azlan, tburtin, Detlev, giacomo-petri, bruce_bailey, Frankie, sarahhorton, ChrisLoiselle, Jennie_Delisi, filippo-zorzi, kirkwood, Kimberly, wendyreid, Graham, shadi, Jen_G, ToddL, ShawnT, jtoles, Laura_Carlson, ljoakley, Avon, mbgower, GN, stevef, nina, Tananda, GN015
Regrets
MJ, Makoto, Chuck
Chair
SV_MEETING_CHAIR
Scribe
Francis_Storr, bruce_bailey

Contents


<Rachael> Rachael: We need a scribe

<Detlev> AFK

<Rachael> scribe: Francis_Storr

New members and topics

Announcements

Subgroup check-in

<Zakim> dj, you wanted to say haptic stimulation

<Zakim> alastairc, you wanted to discuss non-text contrast

dj: haptic subgroup. focusing on haptic feedback

<Zakim> giacomo-petri, you wanted to update about section labels

alastairc non-text contrast: we've really struggled to find a time to meet, so are trying something async.

wendyreid: publication subgroup is really close to finalising work. should be able to present soon

<bruce_bailey> https://docs.google.com/document/d/1uE2WCxPmvNopdCbuZQm_-cGyEdxEouRmZ8UUIlyutoU/

bruce_bailey: keyboard only - we're meeting on wednesday, making good progress.

<bruce_bailey> thank you Detlev

WCAG 3 Requirements updates https://github.com/w3c/wcag3/issues/44

alastairc: last week we went through the reqs doc un quite a lot of detail
... gregg had a wording update and recommended a self-validation requirement

<alastairc> https://github.com/w3c/silver/pull/739/commits/3cb27ac7731fcd05849c3348e819deaace0c94ea

alastairc: if there are any comments, we can discuss those, if not this PR will be used to create a CFC for the group.
... if you have any updates, we'd like those this week

Rachael: if we move this to CFC, you have 5 days to comment on it before the CFC goes out.

<Detlev> ok fine

<Rachael> draft RESOLUTION: Move the Requirements updates to CFC.

<dj> +1

<kirkwood> +1

<Frankie> +1

<bruce_bailey> +1

<Detlev> +1

<alastairc> +1

<wendyreid> +1

<tburtin> +1

<giacomo-petri> +1

<Rachael> +1

+1

<Poornima> +1

<ShawnT> +1

<jtoles> +1

<Ben_Tillyer> +1

<sarahhorton> +1

<ToddL> +1

<kevin> +1

<Jennie_Delisi> +1

<Jen_G> +1

<filippo-zorzi> +1

<Azlan> +1

RESOLUTION: Move the Requirements updates to CFC.

<alastairc> Place to comment: https://github.com/w3c/silver/pull/739

Rachael: as a reminder, you have 5 days from now before we move this to CFC.

Focus Appearance draft outcome https://docs.google.com/document/d/1H5Fk9QRy6WutIUyrCRcqXH04mkp2EFq_qeRXvMawWf0/edit#heading=h.juf86zc2cv61

alastairc: the focus appearance subgroup appearance has been meeting.
... we would like to conduct research on this — please get in touch if you can help with that
... with the decision tree and multiple scenarios approach we have, it is simpler to apply than previously.
... next step is to move things into a PR.
... we have a "what to do" section. There are certain things we're saying are universal.
... there are many types of indicator listed in the document
... we have a common set of tests.
... we're hoping these are clearer than the WCAG 2 version

kirkwood: are we saying that the focus indicator doesn't need to surround the focused element?

alastairc: it hasn't previously required this and there are multiple ways to meet it

Detlev: I agree that the focus indicator surrounding the component is important

<Zakim> mbgower, you wanted to say that my recollection is that only way to consistently achieve contrast in the focus indicator against its background is to shim it with an opposing color

alastairc: a preferred way of creating a focus indicator is something we can add to the document

mbgower: my recollection on this for WCAG 2 was that you need a two-color focus indicator.

GN015: a focus indicator needs to be consistent on a page

<Detlev> @GN015 I think this is getting much too descriptive

Rachael: I was part of this subgroup. a lot of concerns have been covered, but if there are more comments then please help add them in

<Zakim> alastairc, you wanted to speak to Mike's comment, a shim could be a universal method, but you can also do an outline / background

As a side note, striped borders and outlines has been in the CSS spec since 2018, but no browser has implemented it: https://www.joshtumath.uk/posts/2024-08-02-stripes-the-border-function-you-never-knew-you-needed/

<alastairc> I'm surprised that it got into the spec without implementations

Assertions template https://docs.google.com/document/d/1jhc6Z6mwulbpH3O1Yboffq1UW1zFp4CwzlqtKYGUNpY/edit#heading=h.ydb0ppi9970j

Rachael: 2-3 weeks ago, we did an exercise on the template.

alastairc: I think this document is simpler than people were expecting
... I felt that we should focus on there is a process or policy that's being followed.
... we had a brief discucssion on site-level conformance.
... what we could do, if you're following site-level conformance, you could be undermined by objective tests failing.

Rachael: next steps are for people to read the document and comment on it.

<Rachael> Discussion Questions: https://github.com/w3c/wcag3/discussions/106#discussioncomment-10433736

Rachael: one of the questions was around what level an assertion should be placed at
... there are different places we could place an assertion

<Ben_Tillyer> +1 to detlevs comment, -1 to greggs comment

Rachael: where would we place them in documentation and should they be required at any conformance level?

<Rachael> Where in the documentation should assertions go?

Rachael: Q1 where should they go?

alastairc: chair hat off - I think they should be integrated, for example in the decision tree of the methods
... they could be site wide
... if we keep working on outcomes, each subgroup will have assertions that we could probably group

Jennie_Delisi: for me, I agree with alastairc. If an org was to point to their content in one place, that would make things easier when it came to things such as contracts, etc.

<Rachael> draft RESOLUTION: Assertions should be integrated at multiple levels, alongside requirements. Process from here to continue building them into decision trees and then look later to see if they can be grouped.

kevin: the only thing I'd throw into the mix is how we're looking at these in normative and informative terms

Rachael: this is a resolution on documentation

<Rachael> draft RESOLUTION: From a documentation perspective, assertions should be integrated at multiple levels, alongside requirements. Process from here to continue building them into decision trees and then look later to see if they can be grouped.

<alastairc> No change. Where we get there, I'llargue that the assertion itself should be normative, but any further information would be informative.

<Azlan> +1

<ToddL> +1

<alastairc> +1

<bruce_bailey> +1

<dj> +1

<Ben_Tillyer> +1

<Rachael> +1

<kevin> +1

<Frankie> +1

<jtoles> +1

<shadi> 0

<giacomo-petri> 0

<Jennie_Delisi> +1

<Laura_Carlson> +1

<filippo-zorzi> 0

<Detlev> 0

+1

<ShawnT> +1

<kirkwood> 0

<Tananda> +1

<Justine> 0

<ChrisLoiselle> 0

<sarahhorton> 0

Rachael: for those who are abstaining, do you have concerns?

<shadi> +1 to Detlev

Detlev: I'm not sure that an assertion will easily fit into a decision tree

<Zakim> alastairc, you wanted to speak to how it's in the decision tree

<kirkwood> +1

<ChrisLoiselle> no concerns with what was stated, just want more time and data points to make it a plus one. +1 to Detlev's points. The defined terms would be helpful if we have those somewhere.

<stevef> +1

Detlev: an assertion doesn't say if something conforms. I think it's extremely important to say that an assertion can't be used to meet an outcome

alastairc: I think that's a reasonable point
... you could have an assertion early in the decision tree
... there could be a difference between page, view, and site-level conformance

<Rachael> Background https://docs.google.com/presentation/d/178EHzR7oKYyS7m-V2pZvjCg10kR2XLZ7tK2V-LUh2K8/edit#slide=id.g2ef7be0147b_0_6

alastairc: hoping to build up a library of assertions for a later conversation

shadi: we haven't decided yet if assertions will add more points. I don't disagree with the resolution, but I'm still unclear where assertions will fit into the picture.
... we can come back to this later, I abstained because I don't have enough information at the moment.

<alastairc> We're trying to get past the chicken and egg problem of content & conformance.

Rachael: we are going to be talking more about conformance at TPAC.
... a core question is will assertions ever contribute towards meeting an outcome?

<Rachael> Will assertions ever contribute to meeting conformance at any level?

Ben_Tillyer we had a conversation a few months ago about the potential for this. I feel that the assertions fit into "points you get from elsewhere".

scribe: if you were going for a silver-level, you would get points for conformance and then additive, additional points, from assertions.

alastairc: chair hat off: we should be open to using assertions at any level
... not all accessibility requirements meet a true/false level
... I haven't yet seen an assertion that would be good enough to meet an outcome, but that might change.
... assertions might also fit in with site-level conformance if we go down that route

<bruce_bailey> +1 to not ruling out options for assertions

Rachael: we don't know what conformance we're doing at the moment. we're talking about this at TPAC
... chair hat off: I'm hesitant to put assertions at a base level but like the idea of having them at a higher level.

<kirkwood> Would this be an assertion?: [company x] will make every (reasonable) effort to ensure any new or updated website content produced, will conform with the World Wide Web Consortium (W3C) Web 2.1 AA including Cognitive Accessibility guidance?

Rachael: if we do that, though, I would like to see more verification of assertions. we need to figure out how to do that.

<Zakim> Rachael, you wanted to answer dj

Detlev If we have bronze / silver / gold, there will be an incentive to not be bronze. assertions are unreliable and may be written by marketing people, and will likely quickly go out of date. There needs to be a way to verify them.

<Zakim> alastairc, you wanted to comment on John's comment

alastairc: re, John's suggestion. I think it's too wide and too general and they should be related to the outcomes.
... assertions cover things that can't be measured easily, organizational processes, etc.

<kirkwood> for specific outcomes: make best efforts and attest to have processes in place

<kirkwood> q

alastairc: we're not trying to cover the legal side of this.

<ChrisLoiselle> Questions: are assertions and attestations one in the same? What is the difference between an accessibility statement vs. assertion ? From the slide deck shared, Assertions should only be used on outcomes and guidelines that allow assertions. Do we have a list of these that allow assertions? I'm ramping back up on AG, happy to read up more on this and change from "0". We have https://www.w3.org/WAI/about/accessibility-statement/, so

<ChrisLoiselle> would this be an assertion?

kirkwood: could it be that an assertion is testing one SC?

alastairc: it would be linked to an outcome. we're not an enforcement organization.

<bruce_bailey> scribe: bruce_bailey

<kirkwood> +1 to attesting to having a policy on the specific guidance

Ben_Tillyer: wrt DetLev, concern for gaming assertions...
... don't think people will have too much incentive to lie, assuming assertions are part of requirments.

<Detlev> @Ben good points...

Ben_Tillyer: I don't think organizations would want to invest time and money on something which does not gain the org something.

Rachael: Next item, do we want assertions to be public facing by defafault?

<kirkwood> conversely, it may be an easy out for companies?

Rachael: or is available upon request enough?
... or part of an Accessibility Conformance Report ?
... GreggV said in thread that it might be IP or closely held or confidential.

ChrisLoiselle: From the the current VPAT / ITI perspective , is this something that is folded into the present rubric -- or is the intent for assertions to be separate or something new or different?

kevin: WRT assertions there is a lot of stuff [in github thread] which are outside of W3C authority and scope....

<Zakim> alastairc, you wanted to comment on assertions being internal, and contributing to maturity model work.

kevin: Lets outline what this is, not how assertions would be enforced.

<Zakim> Rachael, you wanted to ask Kevin how 3rd party testers discover the assertion

alastairc: I think since the idea is that there is a performance claim and there is part of that which might not need to be public facing.
... at least we can avoid confilict.

<kevin> qq+

<Zakim> kevin, you wanted to react to Rachael

Rachael: To Kevin's point, I think AG needs to say some expectations for assertions.

<alastairc> Suggestion: We assume that the assertions made would be included in a conformance statement, but documentation behind that is not required to make a conformance statement.

kevin: One analogy is a formal audit. I would not expect public assertion, but would expect be able to look a outcomes supporting the assertion...

<Rachael> Maybe we reframe the question: Is the assertion presented as part of a system or only as part of a conformance report.

kevin: a formal auditor might be empowered to insist upon assertions.

Rachael: Chair hat off, assertions on demand might address public/private concern.

alastairc: If web site making WCAG3 claim, assertions would need to listed, but not need to provide public documentation supporting the assertions.

Jennie_Delisi: I am hearing that assertions are per outcome.

<Rachael> Proposal: Assertions met must be available on request and included in a conformance statement and that they may be included as part of an accessibilty statement if desired.

Rachael: We seem to have consensus that existence of assertion should at least be available by request.
... Assertions might or might not be part of an ACR, for now.

ChrisLoiselle: Is a ACR/VPAT an assertion in itself? Seems like assertion should not be PART of an ACR.

alastairc: The "upon request" bit without context seems problematic.

<Zakim> Rachael, you wanted to respond to Chris

alastairc: Regulators might want certain things published, but that is outside our scope.

Rachael: I think assertions have a certain hierarchy to them...

<Zakim> dj, you wanted to ask why not any random person

<alastairc> ChrisLoiselle - I think the attestation in ACRs is more that you're saying you meet a particular criteria. However, assertions here are more about organisational processes.

Rachael: for example Style guide exists. Style guide covers XYX. Style guide requires method ABC to meet XYZ outcome.

<alastairc> DJ - true, but I wouldn't want that to be my inbox ;-)

<dj> so make a form 🙃

dj: Concern that "upon request" might not mean the org ACTUALLY provides it.

How to document “sufficient” coverage? https://github.com/w3c/wcag3/discussions/109

Rachael: Going back previous conversation about sufficient coverage from UA.
... We would have recommend but not required features of what platforms should do....
... And if platform provide the feature, authors would not have to anything special.
... For example focus indicator. If most browser have highly visible focus indicator, page author need not provide that outcome.
... But how do we let web authors know?
... We had discussion around, for example, focus....
... Factors include freely available and in which language?
... Wendy raised the question if WAI would host this activity? Would there be funding?
... Please correct my paraphrase if you like.

<Zakim> dj, you wanted to if no one else is on queue

DJ: I want to suggest open wiki again. It would not be a major investment to keep up to date.

<Rachael> Question for discussion: How will authors know when they can rely on the user stack to meet accessibility requirements and when they must do so themselves?

Rachael: Chair hat off, if wiki is open, who validates as source of truth?

<Francis_Storr> +1 to kevin

kevin: I think we need to be cautious of maintenance issues.

<Zakim> alastairc, you wanted to comment on Wendy's comment

kevin: there have been previous similar initiatives which went fallow after a time.

alastairc: Wendy commented if there were things which might be assured from AUTHOR sided of stack?

<Zakim> dj, you wanted to respond to rachael

alastairc: Interesting question, but seems that user side of stack is the important part to document.

DJ: I still think a collaborative work is the way to get started.

wendyreid: Don't disagree, but we need documentation about what author stack can provide....
... We have documentation about what is needed from user perspective....
... But there is a gap in what the author stack might provide. For example small business....
... org wants to know if they can trust a platform....

<Zakim> alastairc, you wanted to comment on maintenance - for each method, have a set of tick-boxes for an agreed set of UAs.

wendyreid: It is the author stack side which has so much complex and variables.

alastairc: On maintenance, the easiest way to tackle user side is that new CMS includes table of options for a method which is AG way to tick-off if supported by AT or browsers...
... CMS might be as simple as check boxes per method for 12-20 items.
... To Wendy's point, the authoring tools could themselves make assertions...
... for example, does the Authoring Tool let someone overwrite ALT text?

wendyreid: I think this is opportunity to provide website customer about features the users will be looking for...
... You as an authoring tool could provide assertions about what WCAG3 outcomes are covered.

alastairc: My org had an experience working through a social media provider on what they could provide to promote a11y

Rachael: Summarizing, we have a couple proposal having conversation around....
... Or is there NOT a good way to do this?

<alastairc> Just need to keep it simple enough to do ongoing,

Rachael: Due to funding or on-going credible validation ?
... I will summarize and put some comments up, please look for that
... thumbs up / down and comment.

Summary of Action Items

Summary of Resolutions

  1. Move the Requirements updates to CFC.
[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.200 (CVS log)
$Date: 2024/08/27 16:35:49 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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/wendyreid publication/wendyreid: publication/
Succeeded: s/@wendyreid thanks :)//
Succeeded: s/[some minutes missed due to car insisting that it sits on my keyboard]//
Succeeded: s/cat is very much not helping//
Succeeded: s/[some minutes missed due to cat insisting that it sits on my keyboard]//g
Succeeded: s/bruce_bailey : wanna swap?//
Default Present: dj, alastairc, Rachael, Ben_Tillyer, Francis_Storr, kevin, Azlan, tburtin, Detlev, giacomo-petri, bruce_bailey, Frankie, sarahhorton, ChrisLoiselle, Jennie_Delisi, filippo-zorzi, kirkwood, Kimberly, wendyreid, Graham, shadi, Jen_G, ToddL, ShawnT, jtoles, Laura_Carlson, ljoakley, Avon, mbgower, GN, stevef, nina, Tananda
Present: dj, alastairc, Rachael, Ben_Tillyer, Francis_Storr, kevin, Azlan, tburtin, Detlev, giacomo-petri, bruce_bailey, Frankie, sarahhorton, ChrisLoiselle, Jennie_Delisi, filippo-zorzi, kirkwood, Kimberly, wendyreid, Graham, shadi, Jen_G, ToddL, ShawnT, jtoles, Laura_Carlson, ljoakley, Avon, mbgower, GN, stevef, nina, Tananda, GN015
Regrets: MJ, Makoto, Chuck
Found Scribe: Francis_Storr
Inferring ScribeNick: Francis_Storr
Found Scribe: bruce_bailey
Inferring ScribeNick: bruce_bailey
Scribes: Francis_Storr, bruce_bailey
ScribeNicks: Francis_Storr, bruce_bailey

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.)


[End of scribe.perl diagnostic output]