W3C

Accessibility Conformance Testing Teleconference

20 Dec 2018

Attendees

Present
Wilco, Shadi, Anne, Kasper, Moe, Trevor, KathyEng
Regrets

Chair
Wilco
Scribe
Moe

Contents


<Wilco> https://github.com/w3c/wcag-act/pull/313/files

Wilco: Just created PR 1 hour ago. Let's go through section by section
... Let' go up to first note
... What do you think?

Anne: I like it much better

Trevor: I second that. If we say failed maps to Not Satisfied. What is the reasoning? Can we have a common statement?

Wilco: My thinking is that this way the rule is self explanatory instead of having to look up the rules format. Does that explain?

Shadi: I think it can be done in different ways. Could be a note that always appears.
... Like Outcomes and Definitions that always have to be included.
... I think it is good to keep in rule for consistency
... I agree with Trevor in that it may leave room to put something other under failed

Trevor: I think it is okay because forces people to understand what the failed outcome is.

Wilco: I like that as well
... Should we add an explanation of why it's a failure?

Trevor: No, maybe add a rationale as to why we are adding failed outcome everytime
... All it would be is say actually consider the failed outcome. Maybe just a note.

Wilco: Already an overloaded section. Maybe put in introduction.
... I think it's good to explain but we already have so many notes and examples that we should consider putting note in a different place.

Shadi: I really like this approach. It's a Christmas present. Happy to see it come along. : )
... Improvement as well. I do have editorial feedback, e.g. legislation, Web Accessibility Directive example does not necessary present requirements. Also, need an id. Maybe just more substance
... Rationale: Sometimes accessibility rule may not have a requirement. Atomic rule part of a composite rule. Could be a check for a test that doesn't have a failed outcome. Explained later on. Just begs question, when do I have to do it.
... Rewrite to identify when must include accessibility requirement or may skip accessibility requirement. Need better description of when. These are editorial improvements although maybe more substantial rewrite for better flow.

Wilco: Is that important to add here? Rule for WCAG 2 may map to WCAG 1. Will not include in Accessibility mapping.

Shadi: Agree, may not know all requirements
... So much implicit knowledge. Let's make explicit. Condition when something to be in background comes at end. Implicit knowledge needs to be written down more clearly
... But overall I think this is good.

Wilco: Is it good to add information on when not to add to accessibility requirements.

Anne: Still need to read further.

Kathyeng: Concerned that the example section lists a technique as an accessibility requirement. People may get confused that it is required to use Alt attribute when it's is just a technique to meet requirement.

Wilco: That kind of cuts to heart of challenge
... We're not in a position to tell people this is not a requirement.

Kathyeng: I understand it can be a requirement however get a lot of questions on techniques on QuickRef site if they are requirements when they're techniques

Shadi: In example, there is a paragraph that this rule is not required. Let's make this a parameter or flag. Is this "WCAG required: Yes/No." I would keep paragraph as is but let's add to metadata.
... Kathy would this address your concern?

Kathyeng: Yes. I didn't see statement

Shadi: Let's make sure people don't miss it.

Wilco: Change to example or requirement?

Shadi: Let's add another flag that maps to the outcome
... How about we specify this as a Taxonomy

Wilco: How?

Shadi: 4 or 5 parameters with 1 or 2 outcomes, possibly 3. The failed outcome is predefined anyway.
... I think this is a small language or taxonomy to define this grammar

Wilco: How about localization?

Shadi: Or these parameters can be translated. Let's not make this free flow otherwise we get "when the stars are aligned" ...

Anne: It actually does say "when rule is passed we could have satisfied or more testing may be necessary" It's already a taxonomy. Maybe we could do the same

Shadi: We can call out translation / localization.

Anne: Words can be translated but not expanded.

Wilco: Could develop RDF in the future but not part of this work
... Hesitant because this doesn't map to accessibility requirement.

Shadi: When accessibility requirement maps to WCAG it must be noted if it is a requirement or not
... I think everything is here but as Kathy pointed out it can be missed. Let's make as clear as possible.

Anne: We do have to list standard, e.g. WCAG 2.1

Wilco: Yes

Anne: What if we name the standard and write in addition if it is normative/informative, required/recommendation
... Have to list a standard, What if we also write if it is a requirement in relation to that standard?

Shadi: We have Alt example already which is Technique H37. Would there be a problem to add an item, Required for WCAG: No.

Kasper: I see that as a problem as well.

Wilco: Not connected

Kasper: I see that as a problem as well. What if not testing for WCAG? Seems like noise.

Shadi: We have a requirement and were asking if we should reinforce it. Want to make it clear
... People frequently confuse what's a WCAG requirement and what is not.
... If you are writing a rule for RGAA only

Anne: For anyone using those rules would be good to know if they are normative and what is the standard you are referring to?

Shadi: How do we make this clear and not cumbersome?

Anne: Another item Required: yes / no would be a way to do this

Shadi: First example, required. Second one: Best practice rule. This is a requirement / this is a best practice is what we are trying to say

Wilco: Not sure.

Anne: Question. H35 technique does not mention WCAG Name of Standard should be listed here.

Wilco: Not part of any standard

Shadi: WCAG is document and Techniques are informative along with QuickRef and Tools. I think this is too subtle. It's only an informative part of WCAG.

Moe: I agree with Anne that we need to remain agnostic and have metadata or parameter that something is required or not required.

Wilco: Through the mapping we can say that something is required or not required. If it does not map to W3C accessibility requirement make that explicit.

Moe: What if testing for another standard, Section 508, Canadian Accessibility, BBC?

Shadi: What if we list Success Criteria? e.g. 1.1.1

Wilco: Not allowed to do that

Shadi: Paragraph above states we can use ACT Rules for requirements that are not part of W3C standard. We want to rephrase that. In the context of the rule we are writing for, we need to state clearly if rule is not required
... Are we only writing conformance tests?

Wilco: Yes.
... Not limited to WCAG. Can write rules for other standards but still conformance testing.
... Is this a point of discussion?

Shadi: I thought we did want to include best practice rules.

Wilco: Not required by who?

Shadi: Not required by the standard it is written for.

Moe: Seems expensive to separate out the recommendations from the tests

Wilco: They need to be treated separately

Anne: Recommendation, use Alt text less than 100 chars is a recommendation. What is the suggestion? How do we treat it?

<shadi> https://www.w3.org/WAI/GL/task-forces/conformance-testing/wiki/ACT_Rule_Suite_description

Wilco: What would you like to do with this rule?

Anne: I would like to add to my ACT rules
... What would I write to indicate this?

Wilco: Give name of requirement and document. And state how outcome relates

Anne: How do I indicate this is only a recommendation?

Wilco: I don't know

Shadi: There is a paragraph. Rule has to somehow to indicate it. The presentation of the rule has to be clear.

Wilco: Not required is explicitly for W3C

Shadi: Can we expand this?

Wilco: What's the use case?

Shadi: This is an RGAA recommendation
... It is imaginable that RGAA or BBC has requirements that are intermixed
... What's the issue to clarify this?
... We all agree on the fact that something is not WCAG requirement but let's make it more clear.
... Is what we have clear enough?
... Do we keep focused for only WCAG?
... Can we reuse approach?

Wilco: Is that relevant?
... Don't want to broaden it.

Shadi: If something isn't required by WCAG, we should indicate that.
... But if this is not a WCAG requirement but it is by another organization, Wilco is opposed to list that.

Anne: Any standard that you list as a requirement
... I appreciate H37 is listed, this is related to WCAG but not required

Shadi: Are there other examples?

Kasper: ARIA has a set of conformance rules. But then we also have ARIA Authoring Guidelines. Don't fail conformance but not meeting guidlines

Wilco: Ok, I will take feedback. Will work to refine and finish.

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: 2018/12/20 16:18:53 $