Meeting minutes
Poling for new meeting time
MK: There will be some new people joining our project from Bocoup, Lola is from the UK, and our current meeting time is very late in the UK
MK: Can we find an earlier time to meet?
MK: I need someone's help to set up a poll to find a new time for our meeting
JG: Matt your schedule is probably the most constraining
JS: Should we keep it on Thursday?
MK: I already have some ARIA Meetings Thursday morning, so Thursday is not good
SS: PAC can create a google form to send out to poll the group for a new time
We will need to get some input from Lola as to what times will work for them
JS: 8am PST on Thursdays work for PAC
MK: 8, 9 or 10 am on Friday would be good for me
MK: I will ask Lola for some times that work for her
MK: Then I will give Sam the list of times to make the poll
JS: Are there other new people joining from Bocoup?
MK: Yes, there are other new people on the project, but I am unsure who will be attending each meeting.
APG AT Support Table Plan
MK: A few weeks ago we created a PR for the first the APG support table, and identified some work that needed to be performed.
MP: We are working on the feedback and have scheduled the work over the next couple of weeks
Check-in on testing for link, button, toggle, alert, radio
JS: We are in the same spot as last week, there are no updates. I did check in with Dylan, who is on the call today but who was out sick.
Dylan Sheffer: I have some time scheduled tomorrow to complete the testing assigned to me
MK: There is a new version of Voice over available, James do you think we need to perform any restests?
JS: We can run a few tests and see if anything seems like it will need retesting, but our main focus is gathering the status of the next set of test plans and creating the spreadsheet
MK: There are some other people on the call that could help with testing. We did some sniff testing with the newest version of NVDA and determined that we wont need to rerun any tests
MK: To run a sniff test, you don't need to run the whole test plan
MK: If anyone else could sniff test Voiceover for Navigation Menu, Slider, and Modal Dialog that would be a big help
JS: Keep in mind that we didn't run sniff tests for the link, button, toggle, alert, radio with Voiceover as they update wasn't available yet
MK: If apple raises any issue based on recent changes they made we can review, but for now I don't think we need to run sniff tests on link, button, toggle, alert, radio with Voiceover
Alyssa: I can run some sniff tests with Voiceover, what is the timeline?
JS: Over the next three weeks
Alyssa: Great
JS: Okay I have assigned Modal Dialog with JAWS 2023 to you Alyssa
JS: We will look into how old our test data with Voiceover is and get back to the group about what will be needed
Backlog sizing update
MK: James and PAC are examining the backlog of test plans and finding their status
JS: We are going through all of our test plans and for each we are recording when the plan was last modified, when the APG was last modified, as well as the status (Draft, Recommended, Candidate) and evaluating the amount of work required to make a test plan testable
JS: Some of our old plans don't follow our current practices. The biggest updates will relate to testing Screen reader Mode changes
JS: Then we will bring this list of test plans to this group to prioritize which plans we should work on in each quarter this year
MK: We will make a google drive version of this spreadsheet with the community group
MK: When do you anticipate you will have a link to this spreadsheet to share?
JS: We hope to have all the data by next Thursday, best case scenario we have a link to share by next weeks meeting
Approach to prioritizing patterns
MK: The data James is putting together will help us make prioritization decisions, but we need to determine how we choose. Do we choose more low hanging fruit to get more momentum? Or get some traffic data for the APG website and choose plans that are on the most visited pages
MK: How should we approach these decisions? What is important to consider when talking about this?
JG: I think combo box should be high due to all the changes in ARIA related to Combo box
MK: So changes in ARIA, like things that have changed in 1.2 vs 1.1 or 1.0?
JG: Yes I think there has been lost of confusion around combo box, and we should note that it does work in 1.2
Alyssa: I think doing the analysis on the most visited or searched for APG pages would be valuable
MF: I agree
JH: I agree as well
MK: Is there a way to get some data from Google as to what APG Pages are linked to the most pages?
MF: You would have to set up google analytics for that
JS: That data might be skewed as people tend to link to old examples
MP: I will say we are currently working on another project that uses chromium to track which which roles are used where on the web and how much
MP: The use counters are up to Chromium to decide what to track, but they track alot of things
MF: The other thing to consider is we could get a count of a certain pattern, and how many times its used, and then infer what components use it
MK: What other things could we consider?
MF: I was thinking that it would be good to get a list of what do we think are popular patterns, but then we don't have to go through that list from top to bottom, we could skip the ones that are similar to each other or try to select the broadest coverage of roles states and properties
MK: I have been thinking alot about what Accessibility supported means in WCAG, and there seems to be a significant amount of survey fatigue. I'm not sure if sending out a survey would worth the large amount of work it would require
MK: Is there any insight within Deque or other companies, as to what patterns are the most important to people?
MF: I think it exists, but I not sure how to obtain it or boil it down, and again I think it would be faster and more valuable to pursue other data points like most visited APG pages
MK: When making our schedule, for plans that are later in the year we can leave a poll open and adjust the schedule based on the data we get
JS: I think thats valuable, but I want to note that getting these plans ready and tested by AT vendors is a considerable amount of work. I'm also aware that the vendors appreciate knowing when we are testing each plan and when, so I would hate to change that schedule as it would cause considerable changes in their work
JS: Also APG may come out with a new table that may change our priorities
JS: I would like to make sure we have some flexibility, while not opening the entire schedule to change
MP: I think the number of things to test within each plan should also be considered, we don't want to just work on huge plans which will take longer, we will want to work on some shorter plans
JS: I will add to our spreadsheet the number of assertions for each plan
MK: Great thank you for this wonderful discussion