W3C

Accessibility Conformance Testing Teleconference

25 Jul 2019

Attendees

Present
Alistair, Anne, Wilco, Shadi, MaryJo, Trevor
Regrets

Chair
Wilco, MaryJo
Scribe
Trevor

Contents


announcements

shadi: Don't share the links outside. Content from links used to drum up some interest.
... Links of announcements and blog post on ACT Rules

<shadi> https://github.com/w3c/transitions/issues/151

shadi: Can track status of transitions in the github link above

Update from AG on CFC to publish as a Recommendation

<shadi> https://lists.w3.org/Archives/Public/w3c-wai-gl/2019JulSep/0090.html

wilco: A couple of editorial comments were made, but CFC was accepted, so we are going forward with the publication!

Mechanics of reviewing rules to send to AG WG for approval to publish

wilco: For the next couple of weeks we want to start working on how to get rules published by AG.
... We have a proposal that outlines the requirements and how work will be divided up

<shadi> https://github.com/w3c/wcag-act/issues/353

wilco: AG seemed on board with it, but wanted to try it out
... So what I think is a sensible way to get started is: We need organizations that have rules, that is mainly the act community group, to send them to the group
... They will need to submit rules, then those rules will be reviewed by us, and then we can see if they are ready for publication.
... What that leaves for us, is that once we have rules, how do we review and approve them, what does that process look like.

maryjom: When the community group thinks a rule is ready, we should put it on our agenda, perhaps have a survey. Make sure if follows the format first

<shadi> +1 to survey idea

wilco: It is on us to check the correctness of the rule
... What are the question to check for correctness

alistair: We have a model already, such as techniques. Some third party provides the technique and then it goes through some review process. We may be able to use them as a template
... There is also some amount of impartiality, we need to make sure that people cannot push their own rules through, it needs to make sure the whole group approves.
... We are thinking a lot of the act-r as the place where rules are created, I think there should be a loose coupling since rules could come from many places

wilco: Yes there are a couple other places rules will be working on.
... Also, I don't know of any documented process for techniques.

shadi: I think we want to take a different approach from techniques. There isn't a streamlined process, and then a group review and a lot of back and forth
... We are trying to have as much incubation outside of the group. We need to set some entry criteria before the group will review it
... We do need to consider who the rules creators and approvers are
... Don't need to be as harsh as not allowing them to vote, but other organizations should have a say.

alistair: We might need to have quorum, and have a threshold for a number of companies represented.
... The techniques model is bad, but it may be a good idea to backfill their model with what we make in our process having learned from techniques.

shadi: If we are successful with this approach of having incubation outside and having criteria for submission we may be able to use that for techniques as well.
... Will reach out to some people who may have some experience to share with this new proposal.
... One more point, we do need to think of quorum and number of reps, but this may be able to be done via survey.
... Survey open for a few weeks, and then the decisions can be done somewhat asychronously

wilco: What I am hoping to get out of this is a list of questions we can put on a survey for a rule
... Here is the list
... Put out a survey for a rule
... Does a rule follow the rules format
... Is the implementation data correct
... Are there inconsistencies with existing WCAG documents
... Are the assumptions acceptable?
... Are there any open issues? if so do they need to be resolved
... Is there unnecessary overlap with other rules?
... This approach where we have a survey and have these questions, does that sound like a solid way to about this.

alistair: For the implementation data, what if we had some sort of machine learning answer. There will be sometimes where a new test is not easily reproducible by more than one company.

wilco: I don't think so, I think there is always the option to test it manually.

alistair: Do we ascend a rule, that one company could do, but for another company would take several months of manual work.

shadi: I would hope when people are more comfortable with the rules format that they may say automatic manual, and give some details on implementation.
... What I am thinking for the survey itself: 1. Do you think your organization may implement this rule? 2. Do you think this is a viable rule to be published?
... You could say no this rule is not interesting to me. But could say that it is a viable rule that follows with WCAG

alistair: I was thinking more along the lines of saying "In no way do we have man power to support this rule" so we would have to say no to the rule

shadi: One of the problems with WCAG is that we get hung up on edge cases. We may need some conflict resolution for the case of some organizations not being able to support a rule.

alistair: I am trying to steer us away from accepting a bunch of rules, then those rules become standards, and we all have to scramble to support them.

shadi: There are several gatekeepers on the task force level. We are tying to have broadly supported rules.
... If we have misuse, we can step back and update our process. We would rather go forward that get hung up on every single edge case.

wilco: So I have added another question: Are there any concerns that should prevent this rule for publication?
... Kind of an other category for thinks we didn't think of process-wise
... So we have some questions, we can pilot the survey, after we CFC the rules in the act-r group.

alistair: I would have in this first survey, have you thought of any other questions or comments you would like to ask.

maryjom: Anne sent an email about the two rules in the agenda. I think there were some changes to be made.

anne: There are several open issues about each rule, including those. Its about changing the names to be more consistent.
... May or may not impact these rules. I know the naming one does

https://github.com/w3c/wcag-act/issues/353

trevor: *concerned with rule maintenance*

<shadi> +1 keeping updated list

alistair: Would like to make an agenda item for later to discuss it in more depth. Try to figure otu which rules are current and which are outdated.

wilco: We have already covered this issue mainly. Nothing more to talk about.

<agarrison> alistair: Maybe the w3c would want to release rule suites each 6 months, so we know what is current.

wilco: We are going to break early today, send out a cfc to the act-r group and then send a survey out for those.

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: 2019/07/25 13:55:08 $