<scribe> scribe: ChrisLoiselle

ACT Standup

Wilco: Was able to merge some things last week.

Helen: Created first draft of manual test rule. Pull request 2022 in GitHub

Trevor: Got through surveys. Closed one task . Went into state conversation a bit more.

Tom: Reviewed a couple of PRs. Prepped for state conversation.

Daniel: Planning with Wilco and Kathy, did not get to surveys.

Wilco: Did you share Zoom with the community group meeting?

Daniel: Yes.

Tom: Meeting zoom link brought me to another room.

Daniel: I will review.

Tom: I used W3C calendar invite.

<thbrunet> Joined through here and ended in the wrong room - https://www.w3.org/events/meetings/753c7d04-101d-4c2d-a9e3-eeeac99e13e2/20220825T150000

Kathy: I opened an issue for audio element with transcript rule. Transcript file is txt file that contains HTML. Issue 2021.
... updated composite rule and updated surveys.

<ToddL> I have not gone

Todd: Finished Surveys. Replied to Trevor. Need help with last two items on my list.

Chris: My ask to Kathy and Wilco was on Orientation of the page related topics.

Todd: We were doing heading is descriptive.

There was one more issue that was raised on heading is descriptive.

Wilco: Ping me on slack and we will get there.

ARIA required context role

Wilco: Great responses by all.
... talks to explicit semantic role editorial change raised by Tom.

Helen: Will did it twice as he missed one question I believe.

Wilco: on implementation, Kathy raised a question on q4, children .
... if I look at accessibility tree it shows as list and list list item with children.

Kathy: My question on accessibility support and different accessibility trees will have different results, what is the limit? How many browsers do we check before changing to pass vs. fail?

Wilco: My understanding it is that the problem still exists, the accessibility trees are not the same . On failed example 3 , how many children does this list have? Example of list with one item vs. two.

<ToddL> It's the same result in Firefox as well as Chrome/Edge.

Wilco: worth testing further.
... talks to aria-live property and relevance within rule.
... update example and add accessibility support
... Trevor mentioned that we need to make sure they are consistent after role of generic change.
... we should add in text that describes the role of generic and what it does.

Trevor: I wanted to reference whatever issue is open on this topic. I think they are all documented in the issue.
... there are other cases , I don't know if we need an example for all , but maybe in the background

Wilco: How role of generic impacts accessibility tree would be an action and possibly adding examples
... Should we reference WAI-ARIA in the rule? The rule is stricter than WAI-ARIA, but feel there is a correlation

Trevor: On the other aria rules, we use the sufficient techniques, I couldn't find one that is applicable via mapping.

Wilco: I think it is worth pointing to aria as a reference.
... On q7, talks to Will's comments on contradicting statements regarding the first example. We need clarification from Will on what example he is referencing in his comment.

Kathy: inapplicable example 5 has the reference but it is inapplicable.

Wilco: Will follow up with Will.

Helen: He may have changed his mind on his answer question?

Wilco: Child is direct on failed example 2.

Kathy: If adding in direct in, that would be beneficial. If it is understood, that is fine.

Wilco: adding in direct children here may cause more introductions elsewhere too.
... on question 8 , Will's comment , talks to background .
... I see no harm in clarifying this background paragraph 2 regarding list and listitem
... On open issues there are two open issues we need to address further issue 1990 and 1995

Tom B : You can put be on aria required context role and aria required owned elements

Wilco: We can stop here on survey and talk to state topic

Better define how rules related to page states

Trevor: Talks to issue 1953
... We were looking at status messages. Talked expanded states. Status messages available to user is what we want to cover today. We want to make sure if text is present it is read. Talking to high level broad statements at moment.

what is the minimum subset ? applicability - specific event element types, then for the expectation , what is covered by requiring the tester to make an action , i.e. activate but no text is presented. Text appears but doesn't provide information about success or failure or progress. I.e. what is a status message. If visible text appears, change of context is present.

If text appears, no change of context, we would inspect further about role attributes , i.e. alert and check for aria-live attributes.

discussion points , on applicability , we are attaching events on elements. There are also actions on elements. The scripting would elevate to a higher level.

scribe: talks to body to form to button tree . Talks to event handler on button and clicking causing something to happen. Button may be clicked , but doesn't do anything. Button may go up to form, form could go up to body to ask what to do. Decouples button from event handler.

Trevor: On buttons that event handler, we can go test it because it does something. Problem with bubbling of scripting , the entire page has scripting for things that happen on page.

Tom B: For example, On analytics pages, everything can be hooked, so it was hard to tell sometimes .

Trevor: Other option is to grab all elements that have elements attached to them or any ancestor. May cause a lot of elements to be tested.

Testing things that "might" do something would be potential more work.

Wilco: Option 3 may be not to worry about event handlers. Either event happens or it doesn't.

Trevor: I think this would be implemented manually or machine learning or visual inspection to decide what to do.
... can we rewrite rule to more toward methods of how to implement and give information on how to do it.

Wilco: status messages appear after submitting a form.
... What would you do to give instruction to a human to do this?
... if stating forms, we can move it forward regarding forms.
... color contrast rules fail for having color contrast issues, not other issues. Same here .

Kathy: Trusted Tester reference for how to cause an error goes through this.

<kathy> From Trusted Tester: 1. Intentionally violate formatting and other form instructions, e.g., leave a required form field empty, use a different date format than is required, and/or create a password that does not meet the password strength requirements.

Trevor: humans know how to do this, so perhaps interacting with an element to cause an alert .
... To Wilco, do we want to be as granular or keep this broad in terms of interacting?

Wilco: We have the error messaging rule , may want to look at this too.

Carlos put a lot of work into that one.

This is scribe.perl Revision VERSION of 2020-12-31
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Present: ChrisLoiselle, ToddL, Wilco, Helen, Daniel, kathy, trevor, thbrunet, Will_C
Found Scribe: ChrisLoiselle
Inferring ScribeNick: ChrisLoiselle

WARNING: No date found!  Assuming today.  (Hint: Specify
the W3C IRC log URL, and the date will be determined from that.)
Or specify the date like this:
<dbooth> Date: 12 Sep 2002

People with action items: 

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)