W3C

Accessibility Conformance Testing Teleconference

19 Nov 2020

Attendees

Present
Wilco, Trevor, Levon, Daniel, MaryJo, Hidde
Regrets

Chair
MaryJo, Wilco
Scribe
Trevor

Contents


Survey on rule publication website redesign: https://www.w3.org/2002/09/wbs/93339/ACT-Rules_pages/results

wilco: first comment, changing order of accessibility requirements

hidde: agreed we can move it

wilco: we could potentially make it collapse as well. similar to the ACT website

<Wilco> https://act-rules.github.io/rules/59796f

wilco: can expand to see finer details of the requirements mapping

hidde: didn't realize there could be that many, so yes making it collapsible will work
... so display type and title and then the rest is nested

wilco: Not too keen about the collapsed examples and definitions.

hidde: some of them have quite a lot of examples, so I thought having them collapsible makes it easier

wilco: I was thinking of having the first two showing and then expandable after that

trevor: like the collapsible, would be okay with the first two examples showing

hidde: it would create a bit of a hierarchy where the first to examples are the main examples

wilco: its not formalized, but usually they are ordered by expected frequency

shadi: thinking about the audience here, do they look at the three categories separately or do you want to read everything
... could have whole heading collapsible

daniel: with the collapsible on each type you can see the types of rule (pass, fail, inapplicable), would prefer whole heading collapsible

shadi: I hear you say its nice to be able to see the three types, but you want to have a way to expand them all

levon: would rather be able to expand all of them at the same time

wilco: I think showing the first two gives you a quick concrete example of what the rule is testing

maryjom: I don't know how tool developers use these, but I think they would either looking at none of the examples or want to see all of them.

shadi: Do we need expand/collapse at all?

levon: with good heading structure we don't really need it.
... there could be a lot of examples though

daniel: can be 8 or 9 in some cases

shadi: sounds like we probably do need a collapse
... Leave it as is, but include an expand all button?

wilco: And what about the glossary, does the same apply there?

shadi: How many terms do we normally have

wilco: Some can have a dozen or more, it varies

shadi: may be useful to keep the same approach

wilco: Can we have code highlighting?

hidde: Other people said this as well, just need to make sure we know what language it is

wilco: Yep we have that, that's how we get it elsewhere

hidde: Problem is we would need to do that in a lot of other places
... unsure if I like that, I think its already separated by having a different size. I think it maybe too much

daniel: I understand the point visually, but am not a fan of it

shadi: would a colon help after the test rule?

hidde: I like that, it might be a more subtle way to achieve the same effect.

wilco: Link to the test pages, we don't have those links yet, need to setup some automation to get that working

<Wilco> https://act-rules.github.io/rules/59796f#passed

hidde: Yeah, we would link out the ACT rules community group

shadi: So that page would be on the community group?

wilco: Yes, those are archived so that they are always available. Its much more difficult to put them on the W3C website

hidde: Could potentially calculate the link hash as well?

wilco: I think I should include those as well

shadi: I am wondering if there is a way to avoid dependency on the community group
... we also said anyone could submit test rules to the community groups.

wilco: You could do it automatically, but it would be some extra workload for Hidde.

shadi: is it too much to ask the rule creator to submit in separate files along with the test rule

wilco: It feels like it will be error prone, you want to just grab those test cases the moment they are moved over, not when its first submitted

<Levon> be right back

shadi: I am dreaming of a script that would package a rule with all of its test cases and would go to the W3 site

wilco: once we have the link it wouldn't be too hard to grab the file and push it into a directory.

shadi: Hidde please take a look at this and see how difficult it will be

wilco: Kathy's comment on heading link having aria-hidden="true"

hidde: Yes that needs to be changed and made accessible.

wilco: Kathy comment, whitespace links to glossary definition, but outcome in requirement mapping does not

hidde: Yep that will be fixed.

wilco: Kathy comment on the summary of act rule in the about pane
... it doesn't quite describe how to test, more what to test
... Kathy comment on this being stated as atomic rule 3 times.

hidde: We wanted to be very clear about what this was

trevor: like it stating the atomic rule at the top

wilco: problem is that the sentence in the box is actually incorrect, it being an atomic rule does not test a criteria because its an atomic rule

shadi: Yeah we have two types of rules, conformance that maps success criteria otherwise it will be a Good Practice rule.
... the rule in the sidebar remains the same, that it is an atomic rule.

wilco: That will sort of be repetitive of the requirements mapping

hidde: I think this might be a good one sentence overview of the rule that doesn't go into as much technical detail as down below.

<Wilco> https://act-rules.github.io/rules/bc659a

wilco: There could be cases where there are multiple success criteria that may be included in that box.
... MaryJo's comment that DOM tree is a non-working link

hidde: Yeah, that's due to it being a prototype.

<Wilco> https://www.w3.org/TR/act-rules-format/#atomic-rules

maryjom: The content for the atomic and composite rules could link to the definitions in the document

<Wilco> https://www.w3.org/TR/act-rules-format/#composite-rules

wilco: next question about all act rules page

<hdv> https://draft-wcag-redesign.netlify.app/techniques/

wilco: question from Trevor on organization as the rule list grows

hidde: example that has collapsible menu for techniques

maryjom: It also not as cleanly divisible, was wondering what types of filtering, a listing by success criteria.
... depends on the use case, how do people want to use that page.
... maybe we just do alphabetic, could group composite rules with their atomic rules in hierarchy

shadi: Ideally, the rules will be linked to from the QuickRef as well
... until then, we might need something in the interim. The quick ref is the main way to do the filtering and sorting
... the all techniques is a way for people to do text search

trevor: agree with MaryJo, may have problems with searching because a lot of them are similar

<shadi> +1 to Trevor

hidde: we can leave it for now, but we can eventually add headings

<shadi> +1 to move this to later

hidde: sort of need to see more rules than to get better idea

maryjom: For the interim, I think we should at least put them in alphabetical order

<shadi> +1 to publication date if it is easy to get from the content/meta-data

maryjom: date added may also be important for rule implementors to see what rules are new or have been recently updated

hidde: would we want a sorting mechanism there or leave for a future iteration

wilco: can leave for a future iteration

PR on incorporating ACT rules into WCAG materials: https://github.com/w3c/wcag/pull/1526

wilco: replace act rules with test rules

maryjom: Agreed, seems like a good change

shadi: Trevor and Daniel thumbs-upped

<hdv> +1 to Test Rule as it requires less knowledge of how we are organized

wilco: Do not like test rule. Its not a rule about testing, would rather have conformance or WCAG rule
... its rules for conformance testing, but not a rule about testing
... it tells you what to do and what not to do.

<shadi> "Conformance Testing Rule"

wilco: Could possibly just say test rule

shadi: I think that's a bit abstract

daniel: I think its a good point

wilco: Sounds like we need another quick discussion on this topic

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version (CVS log)
$Date: 2020/11/20 16:03:00 $