W3C

Accessibility Conformance Testing Teleconference

11 Aug 2022

Attendees

Present
ChrisLoiselle, trevor, Wilco, kathy
Regrets
Chair
SV_MEETING_CHAIR
Scribe
trevor

Contents


<scribe> scribe: trevor

CFC for Menuitem has non-empty accessible name

<Wilco> https://wai-wcag-act-rules.netlify.app/standards-guidelines/act/rules/m6b1q3/proposed/

wilco: Had first ever objection to a CFC
... comments from will, no test case with image inside the menuitem with valid accessible name

Will_C: From a testers perspective, was the first the test case that came to mind
... is a bit more old school like using an images as a button.

wilco: No one seems to have a problem with it, so we can add that
... next comment was including the title attribute as unreliable in the accessibility support and then have a passed example using the title.

<Wilco> https://github.com/act-rules/act-rules.github.io/pull/1886/files

wilco: sort of similar to this PR adding role="img" to svg due to some accessibility support issues
... opposition to the PR since trying to get all of our examples to be accessibility supported is a constantly evolving problem
... if there is legitimate problem, then that S.C. could be failed due to that accessibility support problem.
... should try not to have accessibility support issues related to the S.C. that is being passed

Will_C: As more of a tester, seeing that a rule specify it or not backs up how it will be tested in practice. Should try to avoid at least the obvious issues.

ChrisLoiselle: Is goal to build more robust examples and we are adding safeguards against the AT/Browser combos

wilco: The accessible name of SVGs aren't announced in some cases, and the fix is the added role. This would fail 1.1.1 if using that specific combo for testing
... feel the same way for the menuitem, if it causes failures due to accessibility support issue should be included

kathy: For the title, it is part of the spec and automated rules would and should pass it

wilco: Would argue that the svg + role="img" is also part of the spec

kathy: If there was a workaround to make sure that the title is read by AT, then could agree that it would fail

wilco: Would you fail 4.1.2 if the title wasn't announced?

kathy: no, I wouldn't

wilco: I'm digging a bit, because I am wondering if we should have something in the rules format that says the rules should not have accessibility support section. At least not those in the accessibility support section

kathy: Feel like there could be other rules impacted

wilco: We should give implementers the ability to fail things with known accessibility support problems

trevor: Worried about the moving target problem of this

wilco: Some maintenance with this, but not too much more than the accessibility support work that we are already doing.
... at the point is more up to the implementer to determine if they want to determine how to fail these types of rules

ChrisLoiselle: The rules/test cases should at least be flagged with the accessibility support issue that could arise.

wilco: Leaving up to will to be liaison for making changes to the rule
... determine if title is still accessibility support problem, then we should remove it from the accessibility section

CFC for Meta element has no refresh delay

kathy: For earlier issue, title accessibility support may show up in other rules

<Wilco> https://github.com/act-rules/act-rules.github.io/pull/1908/files

wilco: No objections for this CFC, but did have one comment
... should not be using the abbreviation SC 2.2.1, should write out full name

RESOLUTION: Accept meta-refresh for AG with the update of PR 1908

ACT rules sheet

wilco: No progress on object element
... helen is not frame
... Chris, status on html graphics contain no text

<Wilco> https://github.com/act-rules/act-rules.github.io/pull/1873

ChrisLoiselle: Waiting approval

wilco: Looks like it is approved, editorial so that can be merged today
... is this rule for CFC?
... Seems like so, will move it forward.
... will, image accessible name is descriptive.

<Wilco> https://docs.google.com/spreadsheets/d/1OSkPFocXk4K3zYLnwS78WLsWO4PvE5yRcsauyefuIUI/edit#gid=1519692012

wilco: only comment was to improve the description of examples

Open ACT pull requests

<Wilco> https://github.com/act-rules/act-rules.github.io/pulls

<Wilco> https://github.com/act-rules/act-rules.github.io/pull/1886/files#diff-de2a0679a9b53f13f4f9a81ecad5170f8752a603224d6a7f99557211c5793426

wilco: First PR is the svg + role="img" that was already mentioned

kathy: Was only approving for this particular rule, didn't think it would have broader effects
... approved before i saw some of the last few comments, may look at it again

thbrunet: Opens up additional questions like what the result should be when role="img" isn't there

wilco: I think the point is to intentionally leave it up to the implementors

kathy: In this one, I think adding role="img" won't change results
... the other rule could cause implementors to get different results

wilco: Iframes #1855 needs another approval, has some suggestions from kathy

Priority issues

<Wilco> https://github.com/act-rules/act-rules.github.io/issues?q=is%3Aopen+is%3Aissue+label%3APriority

wilco: Working on PR for #1853

<Wilco> https://github.com/act-rules/act-rules.github.io/issues/1172

wilco: Will you were looking at 1172?

Secondary accessibility requirements

<kathy> https://github.com/w3c/wcag-act/pull/531/files

<Wilco> Here's a diff for anyone who's looking: https://github.com/w3c/wcag-act/pull/531/files/5c8814da972be7efba8e6d67c1d4b3b2d04400c6..3a42ca5286cce0c373e7df0d8c34e11d6f902cd1

kathy: Kept the conformance requirements like they were before. Secondary requirements cannot be considered to be not satisfied even when all of the failed test cases are not failed
... since there are cases the rule doesn't cover that could still pass the rule
... changed the example to contrast, where 1.4.6 is mapped as a secondary.
... Then on line 245 the second example on no keyboard trap, it has only a secondary attribute.
... have changed phrasing for when you might map to the secondary criteria.
... Other options on what to call these criteria, associated is an option.
... in 175, changed to all outcomes for a test target is failed and inserted the test cases failing.

trevor: Some confusion over how outcomes are getting applied to test targets, thought there should only be one outcome for the test target.

wilco: Something seems a bit wrong. Test target only has one outcome per rule.

kathy: Trying to put more focus on the test cases where the S.C. fails. When all of the test cases would fail the S.C., then that S.C. would be listed as a conformance requirement.

trevor: Think it makes more sense of to keep it as the test case level

wilco: Should be any test target that fails the success criteria then the success criteria is not satisfied

Summary of Action Items

Summary of Resolutions

  1. Accept meta-refresh for AG with the update of PR 1908
[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.200 (CVS log)
$Date: 2022/08/16 08:26:15 $