W3C

– DRAFT –
ARIA and Assistive Technologies Community Group

18 February 2021

Attendees

Present
juliette_mcshane, Matt-King, s3ththompson
Regrets
-
Chair
James Schoels
Scribe
michael_fairchild

Meeting minutes

Agenda

James: will next week be a cross roads meeting?

mk: lets focus on test development next week and skip the cross roads meeting this month.

group discussed https://github.com/w3c/aria-at/issues/386

mk: seth, can you wrap up this issu (386) and close it?

seth: yes

group discussed https://github.com/w3c/aria-at/issues/375

mk: we are going to talk about an issue today related to marking tests as draft. Should we leave this open until we close that issue?

james: yes

mk: group discussed https://github.com/w3c/aria-at/issues/382

james: for this one we need to draft language and update the test runner

mk: can you do this, James?

james: yes

mk: thanks. Can you draft some language and we can review it?

james: yes

Revisit: Next Patterns for Test Plan Development (issue 385)

Issue: https://github.com/w3c/aria-at/issues/385

james: we have put a list of 6 issues in the issue. The first two are the file directory tree view examples.

mk: we have 3 tree view examples. should we get rid of the file treeview example with computed properties?

jon: it would simplify things, and they probably need to be updated.

mk: so, that would mean we get rid of the computed properties one from this list.

james: if you are proposing removing one of these examples from the APG, would you remove roving tabindex exmaples, for example?

mk: not sure

mk: james, you can put the navigation treeview. Remove the file treeview with computed properties. Delay the declared one so that we can improve what it illustrates in the APG

james: when are those improvements on the APG side likely to happen?

mk: we don't have an open issue for updating this.

mk: we can revisit this in about 3 weeks

james: the next pattern is the date picker spin button example

mk: jon, the datepicker spin button is good, right?

jon: I'm not aware of any issues with it.

james: and then, we will revisit the rest of the sliders

mk: we can add button and toggle button to the list

rob: I did create the test plan for the button

james: I think it needs some updates to better match the latest test designs

James: so we will work on: navigation tree view, button, toggle button, date picker spin button, toolbar, horizontal slider, slider examples with orientation, file directory example with declared properties

Work through prioritised list of issues with test writing/ARIA-AT App crossover, with input from Seth and other Bocoup team members if present. Specific issues in priority order:

see agenda for list of priority order https://github.com/w3c/aria-at/wiki/February-18,-2021-Community-Group-Meeting

james: for seth and others, what it is the best way to discuss this?

seth: we trust your order, no preferences from me

mk: there are two ways that I have thought about this. one way is thinking about which of these issues most immediately impacts how your write tests, and would if done later cause you to do more re-work.

mk: some, like 300, don't cause re-work, but does affect how the group operates, and thus how quickly we can move. We could discuss hacking how we operate to get around them.

james: if we look at ones with test-writing impact: the first issue is 363

(discussion around the issue)

mk: it sounds like we need to prioritize this. Maybe this is next week's meeting.

james: yes, that's my plan.

james: issue 388

mk: not having this means more writing of tests. Waiting on this might not mean we have to re-write tests, but we might need to re-structure tests.

james: where it could have a big impact would be if we decide to test non-apg patterns/elements

mk: I think this is not immediately as impactful as issue 363

james: I agree

james: we have some internal processes to reduce re-work and we can share that with seth down the road

james: issue 369, and related issue 358

james: I'm ignoring issue 370 as that doesn't matter for us as test writers

james: it would be good for us to chain setup scripts together, or share functionality between them

james: if we are going to indicate to the tester that they need to press a button, then we do need a way to provide a script to a known good state.

mk: this makes a ton of sense. This affects how tests are run.

mk: I think this is really important.

james: yes, because it will guarantee that tests are always run from a known good state

james: do you agree, seth?

seth: we would like to remove the iframe as a way to load tests.

seth: we need something like a rendering engine

mk: james and seth, should deep dive with Jes's team to figure out some options here and come back to the group

james: the last issue is 366

james: I think that's a CG discussion

james: maybe even provide examples of what to expect to hear for what would fulfill the assertion

mk: yes, we as a CG should dive into this

james: there are a few issues that we didn't get to

mk: seth, I think issue 300 as pretty high priority because it is something that we want to track over time as metrics. And it can affect how we want to work right now.

Summary of issues

  1. https://github.com/w3c/aria-at/issues/385
Minutes manually created (not a transcript), formatted by scribe.perl version 127 (Wed Dec 30 17:39:58 2020 UTC).

Diagnostics

Maybe present: James, jon, mk, rob, seth