W3C

Accessibility Conformance Testing Teleconference

14 Feb 2019

Attendees

Present
MoeKraft, KathyEng, Anne, MaryJo, Wilco, Trevor, SteinErik, Shadi, Charu
Regrets

Chair
MaryJo
Scribe
MoeKraft

Contents


Pull request 328: Roll back of composite-composite and findings

<maryjom> Link to pull request 328: https://github.com/w3c/wcag-act/pull/328

<maryjom> Link to a preview of the draft with changes incorporated: https://pr-preview.s3.amazonaws.com/w3c/wcag-act/pull/328.html

maryjom: I put links into IRC for pull request

<maryjom> Link to a preview of the draft with changes incorporated: https://pr-preview.s3.amazonaws.com/w3c/wcag-act/pull/328.html

<maryjom> Link to pull request 328: https://github.com/w3c/wcag-act/pull/328

maryjom: Continuation of discussion on Wilco's PR. We've decided to roll back a lot of things. I worked with Wilco and Shadi to pull back those items. @wilco Please walk us through the changes

Wilco: Let's go top to bottom, Line 11&12 Added MaryJo and Shadi as editors. Seems appropriate
... In Introduction, not a whole lot there that is controversial.
... I am going to skip over some comments. If I skip your comment, let. me know
... Scope is largely editorial as well.

Shadi: Describe a brief summary of the changes.

Wilco: Conceptually the two changes made. 1.) Changed outcomes to be applicable to test targets rather than test subject.
... If inapplicable, there is 1 outcome
... If applicable, there will be a single pass/fail outcome for each test target
... Implications: Does not let us do complex aggregation, e.g. 80% aggregation of alt. This will have no test target other than page
... Would not be able to identify individual images on page as test targets
... Lead us to outcomes for individual test targets
... Composite rules coming from other composite rules no longer exist. Complex aggregation not used any more.
... Important reason, do not just leave in, if we leave it in we may get asked to provide rules for exit criteria
... @maryom Did I miss anything?

maryjom: You covered it well

<cpandhi> +1

Skotkjerra: I agree

+1

Wilco: Shall I continue?

maryjom: Yes

Wilco: Line 39, Rule Types, editorial + started adding test targets especially to atomic rules
... Outcomes from other rules
... Comment from Anne, editorial
... I did pick up a possible issue in composite rules where we say, Outcomes combined into a single outcome, needs to be for a single test target
... Line 46

<Wilco> https://github.com/w3c/wcag-act/pull/328/files#r256807723

Wilco: Put back composite rules, Line 48
... Explain that they can be flattened
... Line 52, ...at least one of the following rules is passed, comment from Anne

Anne: Meets all expectations

Wilco: We prefer to keeping Meets all expectations in the specification

Shadi: What's Anne's suggestion?

Anne: The outcome of the rule is passed

Shadi: If you take the outcome of the rule, rule has its own applicability, passes for this test target, need input to the rule for test target
... Could have several fails and passes
... Want to say this rule passes for this test target so makes sense to meet all expectations
... Pass is ambiguous
... For particular test target, expectations are met

Anne: We used to have a definition for met expectations but. now we have a definition for passed outcome

Shadi: Section on composite rules using input rules. "each test target must meet all expectations from the following rules"
... Doesn't make sense to say for each of the following test targets, the rule passes. No way to relate composite rule with atomic rule.
... Each rule has its own applicability and produces outcomes. Doesn't make sense to say the outcome has to say pass

Wilco: Outcome is passed for a test target. Rules have multiple outcomes now. Pass: Test Target meets all expectations
... It is the same. Video element meets all expectations means. it. passes all rules

Shadi: Comparing single value with a set of values

Anne: To do any of this we need to know what test target. We can run atomic rule on test target, pass, fail or NA

Shadi: What's the logic?

Wilco: The test target. is the same

Shadi: Produces multiple outcomes. Trying to say only run on this test target

Wilco: Why would you want to do that?

Anne: I see. it. more as an implementation detail

Shadi: It's logically rough. Composite rule, input rule checks caption, second rule checks text. alternative/transcript. For each rule, get 1 fail and 1 pass.
... Composite rule says each of the following. rule must return pass, we may have pass and/or fail.

Wilco: One of the test targets pass the rule.

Shadi: But isn't it the. same that it meets all expectations
... Happy to keep it as Pass/Fail but it doesn't compute for my brain

Wilco: I think the confusion is around expectations being about multiple test targets but it really is about a single test target

Shadi: Test target has to meet all expectations from at least one of the following rules
... Happy to concede to group's decision

Wilco: Propose keep as is

+1

Wilco: Line 60

<Wilco> https://github.com/w3c/wcag-act/pull/328/files#diff-9ac0a6633720a5535b0a53cba04ababeR60

Wilco: A clear distinction between composite and atomic rules.
... Line 70 wrapped a. clear title in schema
... Rule input broken up into input aspects and input rules
... Added Glossary on Line 84
... Acknowledgement dropped to bottom
... Accessible format slightly expanded
... A bunch of editorials up. to. line 110
... Actually editorial too
... Line. 120, comment from Anne, Not satisfied when the outcome of the rule is failed

Anne: I'm not sure. I will delete it.

Wilco: Minor editorial: Should not say HTTP rule is designed to an Accessibility requirements document but rather conformance to...
... Line 124, Name title identifier of an Accessibility requirement in mapping, 1 or all that apply?
... I think it. should be all that apply
... Rule must list all accessibility requirements...Accessibility requirement must list the following...everything that exists in the document.
... Any opinions?

Anne: WCAG SC, would that be the summary of the SC?
... Let's say. title. is 1.1.1 Non-text content

Wilco: Title. + Description

Anne: Is summary necessary?

maryjom: I don't. think it's necessary
... Summary is not needed if it's well defined. Summary may be needed if it's a requirement from an organization as opposed to standard. Either/or

Wilco: Plus we link to accessibility requirement
... I am okay with one of them

Shadi: I. think we should have a preference. We. want. a reference for. the requirement.
... Title + link or Description

maryjom: Maybe we write it as a reference

Wilco: Let's. leave updates to editors

Shadi: Need to. be careful with copyright issues

Wilco: First. example, Line 139
... Required for conformance to 2.0 and 2.1 AA.
... There can be only 1 inapplicable outcome
... Made it singular
... Anyone not agree?

maryjom: Makes sense

Wilco: Same comment on Line 159 and 189 and 192
... Line 211, Take out not. satisfied part

maryjom: I think we can leave this up to the editors

kathyeng: That's my comment. I was not satisfied with the "that requirement"

maryjom: Just a clarification

Shadi: Before meeting. ends lets decide if we want. to go. to CR or have another. draft

Wilco: Line 465. Test subjects vs. Test Targets, This is a non-normative section. Let's skip for now
... Let's. look at the new. definitions

<Wilco> https://github.com/w3c/wcag-act/pull/328/files#diff-9ac0a6633720a5535b0a53cba04ababeR507

Wilco: Line 507

<maryjom> Actually it's line 507 - the definition

Wilco: Line 507

<anne_thyme> +1

<Skotkjerra> +1

<trevor> +1

<maryjom> +1

+1

<cpandhi> +1

Wilco: Can we get. a vote for. the. definition of outcome?
... Let's us do more. precise reporting
... This is one of our. 4 options we reviewed. last. week. We. allow multiple outcomes on a single page. Closer fit to how we test. We test. a page level.

kathyeng: Ok, I. get it.

<kathyeng> +1

Wilco: Sounds like we're in agreement
... All that's left if editorial

maryjom: I can make the changes. I can get these in and run a full spec.
... Should. we generate. a working. draft. or generate a CR draft?

Wilco: Let's. check. the group if. t hey are okay with publishing and. then. CR
... Are we ready to publish?
... Are. we ready for CR?

maryjom: I will create a survey. How long?

Shadi: At least. 48 hours, maybe next Wed

Skotkjerra: Yes. to both as long. as edits are good

Can I merge. changes to PR regarding. Common Input Aspects WG. note?

Maryjo: Merge it

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/02/18 14:15:01 $