W3C

Accessibility Conformance Testing Teleconference

12 Jan 2023

Attendees

Present
Will_C, Helen, thbrunet, Wilco, kathy, Daniel
Regrets
Chair
Wilco
Scribe
kathy

Contents


scribe+

ACT Standup

wilco: AGWG surveys back, 2 were good
... some were pushed back
... would like to improve process before bringing more rules to them
... for implementers, I did a prototype for HTML code sniffer
... will talk to them before merging
... helped Todd with his PR

Will: didn't get to much this week

Helen: need to update affiliations with new job info

Kathy: liaison for a rule, need to address issues

Trevor: need to send a CFC for rule, opened a pull request for survey responses, catch up on state discussion

<daniel-montalvo> https://github.com/w3c/wcag-act-rules/issues/173

Tom: working on frame rule

Daniel: onboarded a new member, terminology vs examples

wilco: surveys due next week

Better define how rules related to page states

Trevor: it's been a while since we discussed stateful rules
... where state can be added in rules - maybe applicability or expectation or input aspects
... statefulness of a color contrast of a link (visited, focused)
... in "Toy examples" section of issue 1511
... states under Applicability can get very messy

<trevor> https://github.com/act-rules/act-rules.github.io/issues/1511

<trevor> https://github.com/act-rules/act-rules.github.io/blob/7fe815ae1aad2ffe9302a593484feca450509f35/_rules/text-contrast-widget-nqzcj8.md#widget-pseudo-classes

<trevor> https://github.com/act-rules/act-rules.github.io/pull/1010/files?short_path=32079a0#diff-32079a0602a5a909b242b4e0961e7c5ddd6b6f5c9906b216d5bf21cf2ba13a77

Trevor: we already have concerns of complexity of rules

Will: does "only matching" mean it can only be used for the same purpose elsewhere?

Trevor: I don't think so
... not sure what it means

Will: if pseudo-class is used to identify a link, that class can't be used for highlighting

Trevor: not sure
... Carlos's example is more like what we're used to but still a long list of Applicable states
... makes applicability very complicated
... I don't think option 2 under Toy examples is a good solution because the applicability gets too complicated

Wilco: I agree
... what states don't we test? Transitions, not just initial but also hover state, how to write rule for aria expanded false and true

Trevor: Toy examples #3 solution set applicability to any link in any state and state is in expectation

<Github> https://github.com/w3c/wcag-act/pull/3: added change management section draft for the review

Trevor: not a fan of leaving it up to implementors for better harmonization

Wilco: tools do single state testing, person running tool must navigate states

Trevor: we also consider manual and semi-automated
... testing some states is good

Kathy: Trusted Tester provides some instruction on testing different states but don't cover all scenarios. Training covers some. Tester must be able to recognize when different states may exist to test.

Trevor: I don't think exhaustive is possible in rules. Test cases can give different examples but can't cover all.

Tom: semantic state change vs hover which is a UI side effect?

Trevor: both are human interaction state changes so not a distinction

Tom: it's hard to take every state into consideration

Trevor: contrast difference between states is very difficult, especially for links

Wilco: can that be its own rule? one rule for hovered and one for focused

Helen: focused and hovered should be separate rules

Daniel: that can work for links but not for aria-expanded true and false

Trevor: if each rule is only concerned with a single state or single transition, will that be tons of rules that are the same with just a state difference
... will we have a composite rule for all these rules?

Wilco: for links, testing the different states is of value. for other things it's not that important
... one rule for default, one for focused...

Trevor: for aria-expanded, would we test that it's appropriately set, is it mouse and keyboard operable...

Daniel: there are FAQs where only one can be open at a time

Trevor: so each rule tests one state

Wilco: each rule can have a different accessibility requirement so I wouldn't combine them
... my preference for having multiple rules is it allows stateless implementors to implement some parts

Trevor: for color contrast, yes. For aria-expanded or error messages, does that work?
... I like separate rules to allow stateless tools to test as much as they can

Tom: pseudo states you can test stateless

Wilco: javascript doesn't interfere

Trevor: can that be an assumption in a rule?

Wilco: that's seems to be an unsafe assumption

Trevor: I think we've found a direction, will add an issue

Wilco: I like using color contrast. should we focus on this?

Trevor: sounds good

Wilco: Welcome to new member, Catherine

<daniel-montalvo> Meeting: Accessibility Conformance Testing Teleconference

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.200 (CVS log)
$Date: 2023/01/12 17:37:20 $