Silver Conformance Options Subgroup

22 Apr 2021


Azlan, bruce_bailey, Bryan, JF, PeterKorn, sajkaj, sarahhorton, ToddLibby, Wilco

Meeting minutes

Zaim, take up item 1

Agenda Review & Administrative Items

<sajkaj> https://www.w3.org/WAI/GL/task-forces/silver/wiki/Substantial_Conformance#In_Scope

JS: we are expected to make recommendations (options) to what solutions might be

down to testing and scoring - is all in scope

JS: for today, there is a draft report (some minor spelling issues) but will be corrected later today

however the main item is Item 2. But, and other comments or questions?

<PeterKorn> ack

<sajkaj> acksa

April Report Draft https://www.w3.org/WAI/GL/task-forces/silver/wiki/April_Report_to_the_Silver_TF

PK: suggests our 'bar' is not that it represents a consensus of the sub-team (needs more eyes on it), but rather a collection of ideas for proposals

proposals to address concerns we don't feel addressed in WCAG 3 today

<jeanne> +1

PK: wonders if that meets expectaitons

sajkaj: As currently stands - does have some disclaimers (not as strong as your suggestion). At the moment, no suggested actions exist - just use-cases

<PeterKorn> https://docs.google.com/document/d/1GyUYTnZp0HIMdsKqCiISCSCvL0su692dnW34P81kbbw/edit#heading=h.gjdgxs

sajkaj: thought we might collect some solutions, and present them as worth looking at - which can be elevated

if we stay focused on the 2 key concerns that we have now... may want to timebox discussion however

JS: it would be great to have that for the meeting on the 29th

sajkaj: that was what I was thinking

[looks at report]

First item - do we have a proposed solution?

we do have an expectation that all a11y bugs are treated like all other bugs

PK: not a fan of this proposal - an attestation from the website citing examples of the types of bugs that still exist

<PeterKorn> JF: With Peter on this - really hard to find a testable way / measurable metrics for "all software has bugs". Another idea (which he's not really a fan of) quantifying a11y bugs found.

<PeterKorn> JF: we are starting from the supposition that nothing is perfect; everything has bugs.

<PeterKorn> JF: If that is our benchmark, then... OK. This website has bugs too. Unless we are comparing against something else, it is just a statement of fact.

<PeterKorn> JS: we are looking for equity; challenge is how to do this.

Bryan: wondering if 'bugs' and 'metrics'... can measure usability with existing metrics

we can quantify the usability of an experience. But bugs need to be fixed (full stop).

<Rachael> https://www.usability.gov/how-to-and-tools/methods/system-usability-scale.html

so maybe we are considering levels of usability (users with and without AT)

<PeterKorn> +1 on references for these ideas

<Zakim> jeanne, you wanted to say that a higher level (silver or gold) and being a part of maturity model.

JS: consider putting this into a non-regulatory measure - or add to maturity model

JS: hoping that maturity model would be in Bronze/Silver/Gold, but it may not

although past discussion about maturity model being part of, say, Gold

but not part of the regulatory pass or fail

PK: think the discussion of whether this is bronze or not should be handled elsewhere - our job is to throw ideas over the wall

<jeanne> Janina: I don't want to put this off to a gold level

but likes the system usability scale - perhaps we throw *that* over the wall - with caveat that it is the least ____ we are proposing

sarahhorton: struggling to understand how this use-case does't fit

sajkaj: how do we fit it into conformence?

PK: imagine a company publishes a "Beta" (full disclosure that it's not "finished") - different bugs: some that impact all users, some that only impact a disabled users.

So, for discussion, imagine 1 of 10 bugs is an "accessibility only" bug - would we then say this site should not be available as "conformant" where WCAG 3 is adopted

sarahhorton: that is helpful, but one of the hardest parts is working with the word 'conformance'

perhaps we need to redefine what we mean?

PK: this is precisely why one of the terms that bubbles up is "substancial conformance'

i.e. this site has "some" defects, but it's not "horrible"

sarahhorton: so we're starting from the viewpoint that there is something else

WF: in EU legislation, they are not focused on everything is conformant, but rather that you have things "in hand" - how you plan on moving forward and raising the bar

JF: with bronze, silver and gold we already have differeing levels of 'conformance'

sajkaj: moving to 3rd party content

[PK reads page aloud]

The April report is attempting to encapsulate ideas from our Google Doc

PK: noting that there may be multiple forms of 3rd-party review

[looking at use-cases]

PK: suggesting a 'report' that shows what site has done or attempted to do

as part of conformance claim - different statements based on content (etc.)

<Zakim> jeanne, you wanted to say about reporting

jeanne: concern that Peter's approach is very "reporting" heavy - was hoping we didn't go down that path. But may be worth investigating

sajkaj: waiting to hear about frameworks

sarahhorton_: re: reporting concerns. Another question is whether we are ready to go into a process standards discussion

is that in scope?

sajkaj: believes yes

sarahhorton_: talking more about process standards for WCAG - we have technical standards, but a lot of our discussions are around process standards

but meta-question - are we comfortable as a group incorporating process standards in WCAG 3?

sajkaj: that sounds like another question that we throw over the wall here

jeanne: that was one of the proposals from our design sprint

sarahhorton_: if we are looking at process standards, we may also need to provide reporting standards to document how that is done (i.e. record keeping, etc.)

if WCAG *IS* looking at process standards, then that will need a standardized reporting process as well

Wilco: is this like VPATs. etc.?

jeanne: sort of thinking of VPATs, yes

sarahhorton_: example of organization documenting the processes they applied to conform to WCAG

sarahhorton_: if we are saying those are in scope, we will need to provide a means to document that as well

Wilco: like that we are thinking about this in different ways

<Zakim> JF, you wanted to note that "process standards" don't scale well

Bryan: not dissimilar to physical requirements

PeterKorn: not sure we have enough knowledge in our industry to be able to do that. Shares JF concerns as well.

notes to Wilco that WCAG 2 made conformance claims optional, but "if" you make one, this is what it needs to be (or look like)

we know that 3rd party is a challenge - you don't conform if 3rd party doesn't conform, but here's what needs to happen

WCAG can provide instruction without diluting what WCAG 3 means

believe it is worth investigating how we provide guidance - direction on how we fix that

sajkaj: some of the 3rd parties may have on their own made claims - can we combine claims? Do we HAVE to do integration testing as well (still?)?

PK: the perennial issue is where do you draw the lines? Today, the assumption is that it is all on "you" - expectation to push needs upstream to vendors

we don't have a framework that apportions responsabilities

[gives examples]

sajkaj: believes there are circumstances where things may be a little 'looser' - cites EDU where the university makes specific content accessible at the individual level

But this may have i18n issues

PeterKorn: talks about Chafey and exemptions there

PeterKorn: what happens when you don't have copyright control - cannot make a '

derivitive' accessible version

PeterKorn: we have some language in our Google Doc that could be added to our April report - volunteers to try adding that off-line

PeterKorn: looking at 5B, tied to use case ___ ; 5C; and 5F

fit into use case B - partially curated travel site

sajkaj: hearing no objections - will move some of this back to email list

asking about when we will be on the agenda for April 29th meeting(s)

jeanne: not finalized yet - need to be sure we get to this

PeterKorn: would it be useful on Wednesday to add links to the wiki page in advance of the call(s)

jeanne: there is already a 'reading' section on the siki page, will add this when ready

Minutes manually created (not a transcript), formatted by scribe.perl version 127 (Wed Dec 30 17:39:58 2020 UTC).


Succeeded: s/action sexist/actions exist

Maybe present: jeanne, JS, PK, sarahhorton_, WF