Silver Conformance Options Subgroup

06 jan 2022


DarrylLehmann, GreggVan, janina, jeanne, JF, maryjom, PeterKorn, shadi, SusiPallero, ToddL
jeanne, Peter, PeterKorn

Meeting minutes

<janina> https://en.wikipedia.org/wiki/U.S._Route_6

<GreggVan> what is the call in number for this call?

<GreggVan> I can only attend by phone but will try that way. Is there a phone number to call in?

US: +1 646 558 8656 or +1 669 900 6833

Agenda Review & Administrative Items

Janina: report out from close of last year; organize ourselves for our work in 2022

Silver TF Report: Conformance/Compliance and Issue 450 Response Reception https://www.w3.org/2021/12/17-silver-minutes.html

Janina: proposed reply to issue #450 became a lengthy conversation on the last Friday Silver call
… expect reply to issue #450 should come up into a Tuesday call

Jeanne: draft reply will go into queue of others

Janina: glossary definition of conformance vs. compliance discussed at last Friday call.
… is putting forth a glossary definition of these needed right now? Maybe work on the other content and then return to definition.
… whether or not we propose a glossary definition, the examples we work through and what is in scope of tech standard vs. policy will remain important.

PK: I think your proposal is right. If we push forward without examples, we are pushing uphill.
… with a collection of examples in hand, the glossary definition makes more sense.

<shadi> +1

JF: in WCAG 3 we are going beyond technical things & into editorial considerations. If position is we want legislators to figure it out, need to make sure larger group
… is on board with that.

Gregg: we have normative & informative; what does "editorial" mean?

JF: don't want to take us off topic...

Janina: thinks we will give guidance beyond highly technical; will have some qualitative measures - hard to get off of that entirely.

JF: agree.

Janina: basic buckets: Technical work (some quantitative, some qualitative), and things regulators will be better at.
… would like Judy to join a call to help us navigate this

JF: +1 to this path

First Deliverables for 2022: Media? Shadi's Approach https://www.w3.org/WAI/GL/task-forces/silver/wiki/Substantial_Conformance/Example_Scenarios janina]

Janina: AGWG chairs schedule had us being finished with 3rd party media by end of 2021.
… didn't quite make it.

<shadi> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Substantial_Conformance/Example_Scenarios

shadi: want to see what group thinks of this doc. Will it help us make progress?
… on questions of 3rd party media, looking at past minutes. Impression (as newbie to group) that we are a bit stuck.
… perhaps stuck because term "media" is so broad.
… approach instead via more specific examples? Different organizaiton/view of use cases previously from this group.
… purpose is to set forth examples, with suggestions on how we want conformance to behave

Janina: should we collect all use cases into a central place? For example, in Shadi's wiki page, no mention of "Payment Friend", which was a useful use case from previously.
… seeing a repeating pattern in paper, which might be conformance part. "Define what technical standard covers", and then "Define what regulators should consider"
… are there patterns from which we can build a set of considerations across conformance scenarios more generally?

PK: That could be useful, but it could be similar to the glossary entry and going too soon to the classification.

<PeterKorn> Janina: agree, important to be sensitive to being too early.

<jeanne> ... I think we should flesh out the individual examples first, and then then we are better prepared to classify.

<PeterKorn> ...so, should we return to 3rd party media discussions from 5-6 mos. ago, and then give it another try?

<PeterKorn> ... other things we were asked to do is sampling approach, which feels heavier

<PeterKorn> Jeanne: would like to hear Shadi finish introducing his work.

<PeterKorn> Shadi: no more introduction. Feels like we've been at this point before. How as a group do we take closer review of doc. Continue working on it?

<PeterKorn> Jeanne: would prefer an overview; even just reading through table of contents

<PeterKorn> Shadi: 9 situations in the doc. In each, two example scenarios.

<PeterKorn> ... section 3.1: "Making content accessible may not be achievable immediately"

<PeterKorn> ... example of a museum making digitized images public, on-line. Two options: hold publication until all ALT text/etc. done, or publish immediately with a11y to come.

<JF> VERY much like seeing this: "The museum indicates the limited support for accessibility for the newly scanned images in an accessibility statement."

<PeterKorn> ...second example MOOC. Not all courses can be made a11y immediately. So there is a plan to make them a11y over time. May not reach all courses because some courses being phased out / replaced.

<janina> I expect the alt that identifies the painting would be present immediately, but the details/summary could come later

<PeterKorn> ...that is section 3.1.

<JF> Which also raises a question - is one of the recommendations from this sub-team that we have formal accessibility statements?

<PeterKorn> Janina: HTML mechanisms... ALT text, and long description. Would think ALT text might come immediately in initial publication; long desc. may be coming later.

<PeterKorn> JF: really like phrase "museum notes limited staffing, and notes it in a11y statements"

<PeterKorn> shadi: how it will exactly appear in technical standard, we can look at later. Trying to differentiate between role of technical standard & regulatory/policy.

<PeterKorn> ...policy aspect "what is acceptable and what is not". Maybe it is acceptable for a museum to be late, but not reports of latest pandemic data.

<PeterKorn> ...that is fundamentally a policy question.

<PeterKorn> ...so we can raise the technical challenge, flag it, so policy can consider it. Could be an a11y statement

<Zakim> JF, you wanted to respond to Shadi

<PeterKorn> Janina: suggest we not get too hung up on whether a11y statement should be part of what we do. Suggests we look at how often in pops up in our work/use cases.

<PeterKorn> ...likely to see this not just around a11y statements, also metadata, other things. Walking through use cases, etc., will help us tease those out

<PeterKorn> JF: been an idea I've been advocating for some time. Putting aside policy considerations; technical question that we will propose/explore is programmatically discoverable a11y statements

<PeterKorn> ...as part of WCAG 3. Don't expect answer today.

<shadi> [[Define consistent ways for content providers to specify which content does and does not yet meet the intended level of accessibility, the types of accessibility issues that can occur in the inaccessible content, and the plan for achieving the intended level of accessibility.]]

<PeterKorn> ...want to put a pin in that, circle back & explore in more detail later.

<PeterKorn> shadi: agree. 2nd bullet puts this in higher terms. "Define consistent ways for content providers to specify which content does and does not yet meet the intended level of accessibility, the types of accessibility issues that can occur in the inaccessible content, and the plan for achieving the intended level of accessibility"

<JF> +1 to shadi

<PeterKorn> Janina: further refinements we'll need to look at is size of institution (first example)

<PeterKorn> Shadi: it took 10 minutes to do a coarse overview of the first example of 9 of them.

<PeterKorn> ...can take them at this pace, but won't finish. Or go without further interruptions so we can finish overview.

<PeterKorn> +1 from Janina & Jeanne - run through quickly.

<PeterKorn> Shadi: returning to overview.

<PeterKorn> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Substantial_Conformance/Example_Scenarios

<PeterKorn> shadi: 3.2 "NOt all content has the same relevance, for example legacy content"

<GreggVan> landed and at desk

<PeterKorn> ...also discussion elsewhere at flows of different relevance (primary flow vs...)

<PeterKorn> ...example of publishing weather forecasts, with large archive of those forecasts. How far back do you go to make those forecasts a11y, vs. just new ones?

<PeterKorn> ...perhaps note in a11y statement. Example of a11y on request of archive (e.g. a researcher)

<PeterKorn> ... first example was legacy. Second example is University - everything controlled by University staff made a11y. But may have hat function between students, etc. Same level of relevance?

<JF> I personally struggle with attempting to define "less important" content as being completely subjective: for an advertiser, the new banner ad they just released to support their latest marketing initiative is THE MOST IMPORTANT content on the page

<PeterKorn> ...Section 3.3: content accumulating too quickly to be a11y. First example is weather station continually publishing weather maps.

<PeterKorn> ...maps may be highly visual. Maybe not every image (published every few seconds), need be described. But at some much slower frequency (hourly? daily) have a described image.

<PeterKorn> ...2nd example of 3.3: on-line shop with users providing reviews of products. Lots of reviews (high frequency), coming in multiple languages, etc. E.g., YouTube content & comments.

<jeanne> I think that we should be careful about using academic examples, because some may be essential for students to pass the course and is protected under US and other national laws.

<PeterKorn> ...such a strong inflow of content, may not achievable.

<PeterKorn> ...two flavors/scenarios of user-generated. Not so relevant, or accumulating too quickly.

<PeterKorn> ...Section 3.4: content provider doesn't directly own/control content.

<PeterKorn> ...First example, service provider publishes tools other use to publish their website (e.g., Wordpress, WIX).

<PeterKorn> ...what responsibility in the authoring tools? Another example: portal aggregates scientific articles from others. See a contractual relationship, which may give rise to

<PeterKorn> ...a different level of responsibility; maybe this is cause for a higher bar.

<PeterKorn> ...Section 3.5: Content providers may have dependencies on other services. First example: on-line ticket sales for a theater.

<PeterKorn> ...using 3rd party payment provider (Janina's Payment Friend, perhaps). 2nd example is startup creating an online presence, using a tool to do that.

<PeterKorn> ...in particular, a social media tool. That tool doesn't provide same level of a11y. What can startup reasonable do?

<PeterKorn> ...section 3.6: bugs and other oversight can always occur. Example 1 is new service from site. Train staff on a11y, but bugs may still occur in their work.

<PeterKorn> ... note in WCAG 2 there is some discussion of bug fixing in certain small # of days, you can consider it a11y.

<PeterKorn> ...2nd example: service provider creates tool for others to create complex questionnaires, etc.

<PeterKorn> ...first example is mistakes people can make because of # of people. Second...

<PeterKorn> ...3.7: content is experimental for all users (e.g. "public beta"), including experimental for people w/disabilities.

<PeterKorn> ...example 3.8: live & real-time content may not support same levels of a11y. Have some of this in WCAG 2 - different A/AA/AAA levels for live vs. not.

<PeterKorn> ... notes some of this may come from how we design requrements

<PeterKorn> ...last one is 3.8: limitations on how a11y content can be made. Example: technologies we haven't figured out how to make a11y today (e.g. immersive environments).

<PeterKorn> ...another example: tourism agency promoting cultural heritage. Challenge of when we shift from one modality to another. How a11y is a text alternative to the richness of visual?

<PeterKorn> ...WCAG 2 acknowledges this to some extent, but maybe need to be more detailed on this in WCAG 3.

<PeterKorn> Janina: Q is what next?

<PeterKorn> Shadi: anyone feel this wouldn't be helpful? (Please speak up). If folks feel it would be helpful, could we review this doc. (or first part of doc), and use it for a survey?

PK: I like the idea. I like the idea of a survey to make our meeting more efficient. After we have done that, do we want to bring this document to a Tuesday or Friday meeting. I like this framing.

<JF> Bring to a Tuesday meeting

Jeanne: would like us to also have a review @ high level in next couple of weeks. Also look at if we have more situations? Or combine some.

<shadi> +1 if there are missing situations!

Jeanne: really want to look for things that are missing.
… would be helpful to then go to Friday, then Tuesday.

<SusiPallero> +1 to Jeanne

+1 Jeanne!

maryjom: may be additional things to cover, thinking about IBM's businesses. May be some additional limitations (e.g. complex visualizations of large scale data).
… traditionally given folks a table, but really not equivalent when you have a very large amount of data.

<shadi> [[this is not intended to be an exhaustive collection of all possible situations!]]

maryjom: also surveillance video as an example of live video.

[notes we are at time; suggest we finish queue quickly then close]

Gregg: Thinks this is good. But it goes everywhere from what, who, when.

<SusiPallero> maybe think about missing situations that may be edge cases that can be difficult to interpret

Gregg: need to break up "here we think this is what WCAG should do; here is where think should be policy"

Janina: closes meeting, we will take this up next week.

Shadi: specific homework?

Jeanne: can create a survey. Also, wiki page has a "talk page". Don't need to wait for survey.

Peter: so action for everyone - read the doc., make comments in Talk section starting now. Jeanne will also create a survey to solicit more specific feedback.

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).


Succeeded: s/,,96638055490# US (New York)/

Succeeded: s/ie/it

Maybe present: Gregg, Peter, PK, US