W3C

Accessibility Conformance Testing Teleconference

24 Oct 2019

Attendees

Present
Shadi, Wilco, MaryJo, MoeKraft, Kasper, KathyEng
Regrets

Chair
MaryJo, Wilco
Scribe
MoeKraft

Contents


Feedback from AG WG on HTML page has a title

Wilco: We've had discussion with AG Chairs regarding Page Title. Request to make examples clearer. Does this communicate to the world what this needs to be?
... Working to resolve. Not sure if this needs to be resolved before we publish
... Unfortunately if AG insists we will not be able to publish on time. Will confirm.

Shadi: Do we know when we will have a decision?

Wilco: I don't know.

<MoeKraft_> <MoeKraft> Wilco: We've had discussion with AG Chairs regarding Page Title. Request to make examples clearer. Does this communicate to the world what this needs to be? [09:06] <MoeKraft> Wilco: Working to resolve. Not sure if this needs to be resolved before we publish [09:06] <MoeKraft> Wilco: Unfortunately if AG insists we will not be able to publish on time. Will confirm. [09:07] <MoeKraft> Shadi: Do we know when we will have a decision? [09:07] <[CUT]

<MoeKraft_> scribenick: MoeKraft

<MoeKraft_> Wilco: Shadi, it may be better if you go to AG

<MoeKraft_> Shadi: Yes.

<MoeKraft_> Okay

<MoeKraft_> Wilco: AG asked for a survey for rule.

Survey results https://www.w3.org/2002/09/wbs/93339/work2020/results

<MoeKraft_> Wilco: Full agreement from Kathy

<MoeKraft_> Wilco: A couple of comments from me. +1 by Romain

<MoeKraft_> Wilco: Not in favor of an errata

<MoeKraft_> Wilco: In the meantime we can collect issues.

<MoeKraft_> Wilco: As we get familiar with the format, people can figure out workaround. Let's wait for awhile

<MoeKraft_> MaryJo: We have a few outstanding issues.

<MoeKraft_> Wilco: Not much

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

<MoeKraft_> Shadi: When there is a publication, we need to have an errata

<MoeKraft_> shadi: Doesn't mean we have to do anything active. But if there is a comment, we'll have to put it in there. For example, there's a comma missing. Why not log it in errata.

<MoeKraft_> Shadi: Future versions will assess comments coming in. I don't think we need detail but just state what's on our plate

<MoeKraft_> Wilco: Ok. Good enough.

<MoeKraft_> Wilco: Benchmark. Mary Jo you have a comment

<MoeKraft_> Mary Jo: For me, it does not necessary have to work. Test tools can test snippets and keep track of it. Not up to w3C to keep up with results

<MoeKraft_> Mary Jo: Is it typical to post test results?

<MoeKraft_> Wilco: It is. Browsers do it.

<Kasper> https://github.com/web-platform-tests/wpt

<MoeKraft_> Wilco: For interoperability. There are hundreds of tests for correct implementations

<MoeKraft_> Wilco: Who manages that

<MoeKraft_> Kasper: WPT is one of the biggest test repos

<MoeKraft_> Wilco: Who manages

<MoeKraft_> Shadi: I think it's a w3C group but bound to HTML

<MoeKraft_> Shadi: It's a matter of trust. What source do we trust. Rationale, make sure rules go through w3C working group

<MoeKraft_> Wilco: So, yes, in precedence there are some but maybe new for a11y

<MoeKraft_> Shadi: Right, for a11y there is not a precedence. There were attempts in the past.

<MoeKraft_> Shadi: What's the point?

<MoeKraft_> MaryJo: Is it problematic?

<shadi> https://www.w3.org/WAI/ER/tests/

<MoeKraft_> MaryJo: Is it a showcase of what tools need what rules?

<MoeKraft_> Wilco: Yes. The latter.

<MoeKraft_> MaryJo: Just feels weird to me.

<MoeKraft_> MaryJo: Since we haven't had much participation from Tools groups does it means that these tools get dropped.

<MoeKraft_> kathyeng: We have lots of test results from Trusted Tester on apps. The owners of applications directly receive these results. Talk about putting in a repository and sharing them but have not gotten too far. Concern that owners would object.

<MoeKraft_> kathyeng: Concern legally. Still exploring.

<MoeKraft_> MaryJo: Also, who is keeping it up to date?

<MoeKraft_> Wilco: I envision that it's not us pulling data but anyone who has testing methodology or tool can contribute. Not an ACT Task Force responsibility. Tool vendors and authors can contribute their results.

<MoeKraft_> Wilco: I think this is the same with the web platform. Only browsers that contribute are there.

<MoeKraft_> Wilco: We need more momentum for this org.

<MoeKraft_> Wilco: Do we have a way for orgs to show their rules?

<MoeKraft_> MaryJo: Is there an easy way for them to test Pass/Fail?

<MoeKraft_> Wilco: ACT Community Group has snippets available to run with automated testing tools.

<MoeKraft_> Wilco: Manual are a bit tougher but may take about 2 hours

<MoeKraft_> MaryJo: Are there instructions?

<MoeKraft_> Wilco: Yes

<MoeKraft_> MaryJo: If there is an easy way to find what to test against, I'm fine but how do they know where to post results?

<MoeKraft_> Wilco: This is something to figure out.

<Wilco_> https://act-rules.github.io/implementation/rgaa-3.0

<MoeKraft_> Wilco: This is the most effective way to organizations on board.

<MoeKraft_> Wilco: We are at a point to start measuring this.

<MoeKraft_> shadi: I think we may be able to port approved rules & data and how they are supported. I think this is a very useful piece.

<MoeKraft_> Wilco: Reasoning is to encourage organizations to adopt

<MoeKraft_> Shadi: We'll never have a complete ruleset. It will be iterative. Different tools will provide different levels of support for ACT.

<MoeKraft_> shadi: Another motivation is to promote competition.

<MoeKraft_> MaryJo: I want to make sure info doesn't get stale because this could impact business.

<MoeKraft_> Wilco: Automated results shouldn't be stale at all.

<MoeKraft_> Kasper: We run our own test suite which is not tied to published rules. I would have to run and push results or someone else would have to do it.

<MoeKraft_> Wilco: Is there no way to run just the automated

<MoeKraft_> Kasper: There is but we want to run semi-automated as well.

<MoeKraft_> Wilco: Yes. So things will get stale for steps that require people pushing updates

<MoeKraft_> Kasper: Yes. But we have control of process.

<MoeKraft_> Wilco: @maryjo Is this something we have to tackle?

<MoeKraft_> MaryJo: We need to make as easy as possible for tool vendors

<MoeKraft_> Wilco: Yes. This is something we can tackle.

<MoeKraft_> shadi: Let's not call Benchmark but rather Implementation Results

<MoeKraft_> MaryJo: RDF - Rule Data Format

<MoeKraft_> Wilco: I haven't seen any use cases to make rule format into RDF

<MoeKraft_> shadi: RDF is already being used

<MoeKraft_> Wilco: Having a way to express a rule using semantic web

<shadi> https://www.w3.org/RDF/

<MoeKraft_> Wilco: Do we need a way to semantically express each rule? I haven't seen any use case for this

<MoeKraft_> shadi: Ok. There is a complexity for sure.

<MoeKraft_> shadi: I think we had examples from Deque and Site Improve that were different but using RDF we can be generic

<MoeKraft_> Wilco: EARL is written using RDF.

<Wilco_> https://github.com/w3c/wcag-act/pull/307

<MoeKraft_> Wilco: I haven't personally seen a use case. My suggestion is that we not look into this.

<MoeKraft_> shadi: There are two things to keep in mind.

<MoeKraft_> shadi: EARL describes results from testing

<shadi> https://www.w3.org/WAI/ER/tests/

<MoeKraft_> shadi: There was an idea of a counterpart for a test description language

<MoeKraft_> shadi: Never went far because of its complexity

<MoeKraft_> shadi: Some vendors have this type of pluggability but haven't seen attempt to standardize. I don't think we need to continue.

<MoeKraft_> shadi: How about EARL?

<MoeKraft_> shadi: When we started EARL, there was not JSON LD

<MoeKraft_> shadi: I still think that something in this space, common format for output might be worth exploring.

<MoeKraft_> Wilco: I think this is useful for benchmarking/implementation results

<MoeKraft_> Wilco: We have 1 format that can handle this data and that is EARL

<MoeKraft_> Wilco: Would it be helfpul for adoption if there was a guide on how to use EARL?

<MoeKraft_> shadi: Yes.

<shadi> https://www.w3.org/TR/EARL10-Guide/

<MoeKraft_> shadi: This is focused on XML serialization

<MoeKraft_> shadi: Doesn't have to be in this format.

<MoeKraft_> shadi: We don't have to call it EARL. It can be whatever we call

<MoeKraft_> Wilco: I think having a clear guide on how to use data formats would be helpful. An update to EARL would not be helpful. This would take us away from doing rules.

<MoeKraft_> Wilco: I would love to see an update to an EARL but not responsibility of ACT TF

<MoeKraft_> shadi: Maybe we can generate requirements

<MoeKraft_> Wilco: Are you suggesting maintenance for EARL?

<MoeKraft_> shadi: Let's call it "documenting experiences"

<MoeKraft_> shadi: "future change requests"

<MoeKraft_> shadi: Let's see how it evolves

<MoeKraft_> Wilco: Ok, We've covered our 5 items

<MoeKraft_> shadi: I have a reply from Andrew. They are happy for us to set up a survey. AG planning meeting is right now. Hasn't answered a question about publication. I need help from Mary Jo and Wilco for questions for survey.

<MoeKraft_> Wilco: Should I take a stab at our Work Statement?

<MoeKraft_> shadi: Yes. That would be great

<MoeKraft_> Mary Jo: AG WG is focused on charter updates.

<MoeKraft_> shadi: Nothing we suggested would impact the charter

Review pull request #2 for our published rule repository https://github.com/w3c/wai-intro-act/pull/2/files

<MoeKraft_> Wilco: This is a proposed update to WAI Tools page. It is going live next week.

<MoeKraft_> shadi: Yes. Provided that we have the rule.

<MoeKraft_> Wilco: Please take a couple of minutes to review and provide any feedback or concerns.

<MoeKraft_> Wilco: We are done.

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: 2019/10/31 06:23:12 $