Accessibility Conformance Testing Teleconference

<scribe> scribe: dmontalvo

ACT Standup

Wilco: Worked on some announcements. AGWG survey for review the rules. PR on "images containing text" ready for another round of reviews

Tom: Finished surveys, not much because of Thanksgiving

Daniel: Working on announcements for ACT redesign and implementation pages

Kathy: Wilco and I reviewed the secondary requirements PR to address a few comments from Jean-Yves

Trevor: Participated a bit on that discussion, worked on other PRs, surveys, and finished the "essential text change" definition

Wilco: Did I already approved it?

Trevor: No.

Wilco: I'll have a look, please others have a look as well

Todd: Not much from me, went through a se of PRs, looked at the secondary requirements comments

Wilco: We have now two rules with updates needed
... I was hoping we could assign these
... Visible label is part of accessible name
... Iframe has non-empty accessible name

Kathy: I could take the visible label one

Tom: I could take the iframe one

Wilco: This is one we may need to have a separate conversation for

Announcing ACT Implementations

Wilco: Sent an email to everyone involved in ACT to let them know that W3C is putting an announcement about implementation
... Email, W3C and WAI news, tweet.
... I wanted to ask you all to help promoting this next week
... You could write something, post a video, whatever you use to do

Kathy: I'll see if our agency of public affairs can do something

Revisit secondary requirements

Kathy: This is the PR for secondary accessibility requirements
... We put the content in the google docs in here, everyone was asked to take a look at it, including CG. Jean-Yves put some comments

Wilco: Jean-Yves did have a look at it, but he did not reply there, he did in Slack instead

Kathy: The way we had the descriptions and scenarios opened it up so that pretty much anything could be listed as a secondary requirement
... Wilco suggested and additional paragraph that a secondary requirements is correlated with the rule but the rule is not intended to test the conformance of that secondary requirement
... He also suggested to change "should" to "may
... Also will add a description under these passed, failed, and inapplicable descriptions for easier reference through the document

Trevor: Why do you specify only the failed test cases at the end of that? Why is it not all test cases?

Kathy: Only the passed and inapplicable examples are true for the requirement. No keyboard trap example

Wilco: A lot of these cases do fail the SC, probably most of them do

Kathy: I think when we are describing this particular example we aren't listing 2.1.2 in that particular rule because we are saying you can't fail it as further testing is needed

Wilco: You can determine based on the rule if the SC is failed, but the failed examples would fail the SC. That's the correlation.

Kathy: We try to define a way to relate these particular edge cases. Every time we try to map them in a rule we also have to consider what other scenarios exist outside the test cases of the rule

Trevor: I need to read it a bit more
... I feel that the requirements are doing something to the test cases, which is probably awkward

Kathy: I may flip the order

Wilco: That's a good idea
... It's true that it makes it look like the requirements do something to the test cases, but that's not true

Kathy: I can try to reword that

Wilco: Jean-Yves still worries about 1.4.3 and 1.4.6

Kathy: Would it make sense to call it intended instead of conformance requirement?

Wilco: I don't know, I don't think that addresses his point
... Comment on line 179
... For our color contrast minimum rule, these two conditions are also true for the contrast enhanced SC. By this description that would be a conformance requirement

Kathy: When we have something that passes the minimum, the enhanced requirement is sometimes not satisfied, so not passed or inapplicable
... Further testing needed is very broad
... When we look at the test cases it is going to be known that sometimes the contrast enhanced is not going to be passed
... I had "An accessibility requirement cannot be "not satisfied" for passed or inapplicable", this further testing needed would open it up

Wilco: I am wondering if we even need these conditions
... I think a conformance requirement is an accessibility requirement that the rule is designed to test
... and a secondary requirements is one that a rule is not designed to test but there is still a correlation
... Maybe we also need to say that the requirement fails when the test case fails

Kathy: I think even if you say that it wasn't designed or intended, it could still be open for interpretation
... Are you thinking to take away the descriptions after each of the scenarios?

Wilco: Let's try this in another editor's meeting

Form field label is descriptive

Wilco: Five responses, three accept as-is, one accept with changes, one reject
... First comment is on Q2
... This assumes that the label is part of the visible content
... This is true. This is not just about programmatic labels, but about labels that are also visible
... Q3, Tom.

Tom: Visible label that's hide from the accessibility tree is label in name

Wilco: You are announcing the label, and then the name needs to match the label

Tom: Seemed strange to put it in 4.1.2

Wilco: It does ont test that the label matches the name, only that the visible label is programmatically associated with the control
... We probably should mention Label in Name. It is in the background, not in the assumption

Tom: Not a blocker though, but enough to call it out

Wilco: I will put an issue, this needs to be rephrased.
... Assistive technologies do not need an accessible label as much as they would need an accessible name. They are assuming 2.4.6 does not apply here
... It is not clear from the SC either
... Since it is covered in a different SC, the assumption was made that this does not apply here

Tom: Would it be acceptable as a secondary requirement?

Wilco: Potentially

Tom: That would be a later discussion
... I am not as involved in the details about these slight differences

Wilco: I am leaning towards removing this assumption. The rule is valid without it and the assumption is confusing.

Trevor: agree.

Todd: Agree

Wilco: Kathy says inapplicable example 2 would pass 2.4.6

Tom: I think the inapplicable part is that the aria-hidden is not programmatic. The programmatic one is not visible, that's why it does not apply

Wilco: The way the inapplicability is written, this should be inapplicable
... Q5, "a bit confused about programmatic label"
... I think if this rule was about visible form controls and it had two expectations that there is a visible label and that the label descriptive that would be easier to address
... That way we could treat inapplicable example 2 as a passed example
... This seems to me like a weird gap because it uses programmatic label
... Whether or not we should require this before taking this to AGWG is a good question
... The applicability would get rid of the programmatic thing and there would be the two expectations mentioned above

Trevor: Half of the group today has problems, even though I said it could be published I'd say I am not comfortable onw. I'd go with a rewrite

Wilco: Not a huge one, but not an easy one either
... Anybody disagree?
... There is an open issue that needs to be resolved.
... Not keen on applicability, also wondered about visual context
... It did not mentioned logical distance in the first paragraph and then relies on it for the further explanation

Todd: I agree.

Wilco: More generally, I am not sure this is unambiguous.
... Distance, near it, these are ambiguous concepts
... This probably needs to be a lot more explicit, maybe mentioning pixels or whitespace as a way to measure
... I need a liaison for this rule that needs to be updated.
... Anyone can take that on?

Todd: I can work on that

Wilco: Let me know if you need any help

Todd: Will do

HTML element language subtag matches language

Wilco: This one is probably a little easier
... Q4, Trevor. Passed examples 4 and 5 were inapplicable
... I see where you are coming from but I disagree
... These closing tags are optional

Trevor: Great, that needs to be stated somewhere though

Wilco: Fair point

Trevor: I can leave with adding it to the description

Wilco: Next comment is from Tom on Q5: SC does not say the language tag should be correct

Tom: SC is a little vague on this

Wilco: Q6, open issues
... I think this is an easy thing to pick up. Is everybody happy?
... We will explain that fictional languages are not most common language definition as far as this rule is concerned

Tom: Q7 Some trouble processing explanation of passed 2
... IF you follow the links to most common language it seems to me we should update these references

Wilco: Titles are not included because you cannot put a language in a title element