Meeting minutes
Matt_King will anyone scribe?
carmeni me
Matt_King that will make you the scribe
topic, Current status
Matt_King any requests for additions to the agenda? I am planning that we will have a meeting next week but the following week I might be in London in an all day planning meeting so I might have to cancel the meeting two weeks from today
Matt_King we do have an AT Drive sub group meeting and I am assuming that is on but we will find out when Mike is back
the only substantive changes is that we landed some things on the AGP - vertical temperature slider and the FAQ disclosure. Those are both merged and the APG examples aren't published yet. That will happen next week
But Isabel I did confirm that if you grab the latest main branch from the APG you will get revised examples for both revised FAQ and vertical temperature slider
@isa I need to pull the latest ?
Matt_King yes, if you pull the latest from main
They are not published but the are in the main branch
I don't think there is any other. Isa has worked on making the tabs with automation ready which is waiting on review from me. We don't have it ready yet but it should be ready pretty quickly. Those will be up in draft review very soon. Before next week's meeting.
App Issue 1389 - Recording AT/Browser versions for runs started with bots
<Matt_King> github: w3c/
Matt_King IS Howard here
carmeni no, and Mike is away until Monday
Matt_King I think we have questions abotu the user experience is going to be and how often they are prompted. This was a comment from James.
Since James, Howard nor mike are here we should save this for next weeks agenda
App issue 1352 check-in
Carmen: Matt's feedback on issue is assigned to eng
Changes planned for staging on wednesday.
The current version of 1352 will go to prod on Monday.
Matt: what about Isabel's feedback?
Isabel: shared via email
Carmen: OK to create a task for it?
Isabel: should be fine to put it into an issue
My suggestions align with other current app behaviors
Matt_King if you create an issue Carmen by listing the changes Isa and I could review that asynchrnously and make sure it is understood as it was meant by isa and me
carmeni perfect
I will send the task today
Matt_King now we will be able to say a command is untestable and then you will choose a side effect. What is the side effect that made it untestable and then you will have a more accurate report.
The change enables us to do untestables and changes the language for unexpected behaviours. We are remaining that to negative side effects
I had the standard check in on sliders and color view, is there anything there?
Isa no but I get the examples now. It's not affected. I need to check on Hardy
Matt_King we might need to have our modify his results.
Isa I can do it, he had said he had any issues if we edited his results during the last meeting.
@louis happy to run it myself it that takes something of out plate
Matt_King who tested it? Let's see. It was Joe
Matt_King Would it be ok if Louis tests it?
Isa there are a couple of conflicts in a couple of test and unfortunatly it's on the test we are more concerned about. You need to re run the test to see if you get different output from Hardy.
@louis Do I submit my results?
Isa Re run it and update the results. Just do the one with conflicts please
@louis if I run a test and get a different results, I still put my results over Hardy's?
Matt_King yes, I do suspect Hardy had a copy paste. When I use the exact same version he had, and same conditions, I did get the size information (not sure about this scribe part)
@louis I won't have the version as him? And the app will ask me to update my version to match. Do you want me to do that?
Matt_King yeah, I think you need to do that.
@louis
I can take care of this early next week and Isa can review
Isa thank you!
Matt_King thank you!
Re-run of JAWS report for Radio Group Example Using Roving tabindex
Joe_Humbert I am a bit confused. i completed work but I don't see it anymore
I tested with JAWS
Isa I can see you did 15
@louis
It's roving tab index
Joe_Humbert I think I need to refresh
@louis I see 15 conflicts for 15 test. The conflicts are set size and set positions Matt
Isa the output is the same for both of you
Joe_Humbert for some of them but not for all of them.
Matt_King I am curious Louis. I got the same output as Joe for one of the tests.
I am looking at test 2. Joes' output was to change the selection upper down error. You got that at the begining of the output which I didn't get at the beginining.
Joe_Humbert that might be a copy paste error
Matt_King Louis has the same output in test 2 without that extra speach from the begining
Isa Because we agreed not to include the speech from the begining
Matt_King we agreed to write all of the outputs
@louis it depends. It could have had text there
I think the idea of Joe to copy paste errors is right
Matt_King we want to include all output
Isa we do but we also say it gets confusing because you get status within that help. I just need us to document that if possible because we either do or we don't
@louis
it would be easier for the test if it's help text or not
Matt_King the bot doesn't know if it's help text either
@louis I haven't been including help text in the ones I have been running. So I would need to re run them
Isa I don't think it matters in this case
Matt_King I agree
I know we aligned on the help text but I don't remember where. I think we have to include help text because the bots won't know what is help text and what isn'gt. So it's a matter about how we analyze it.
If as a group we identify the missalignment is on help text then we would make a decision. But it is really murky.
I think it's a non issue in these conflicts. Test 2 has a difference in the output
Isa the actual assertion. Louis there is a mistake here because you failed an assertion that should pass.
@louis my bad
Matt_King if you can fix that it would be great
@louis I will go over the conflicts and fix anything
Matt_King in test 9 you have different output. Joe's output says 2 of 3. Louis didn't record it. For F command and A command.
IF you are not getting it Louis, there might be a bug that hasn't been fixed. We would need you to re run and check on it
Isa I can also test it because I think I also remember not getting it before
Matt_King make sure to use the latest version
Joe_Humbert I notices inconsistency in the check or unchecked state of the radio button when using f for set sizes
Matt_King which test was that?
This will be a very interesting bug for Vispero
@louis I will say it's interesting because sometimes it says 1 out of 2 sometime it doesn't for the radio button . But I can re run the conflicts
Matt_King please yes.
The action item here is that Louis will check the ones with conflicts.
carmeni is there an action item about help text?
Isa I think it was with action menu bottons
Matt_King it would be an awesome action item if somebody want to take it
carmeni I can take it
Matt_King we need to identify when we last discussed the two questions: 1. do we record tudor messages or help messages. 2. If we record the messaged, do we use them to determine an assertion veredict?
Isa we discussed it when we talked about actions menu button with VO and Safari
carmeni thank you
@louis Joe did you notice any weird mode behaviors? I kept getting weird behaviors
Joe_Humbert I didn't, but I use my mouse to navigate
@louis did you notice anything Matt? I get worried it will change the focus / pop into forms mode in an unexpected manner. Or just odd behaviour
Matt_King I have noticed it in certain conditions, not retlated to radio buttons but yes related to edits. If I have focus on an edit and I go to another place with alt back sometimes it takes me to the edit field
@louis that caused problems for me and wanted to see if anyone had the same experience
Run of accordion test plan
Matt_King Louis started this. NVDA is partially done. Who is the person with ID Rohit?
carmeni I will check internally. He is our new intern. Please unassign him.
Matt_King do I have anyone who wants to work on VO?
@elizabeth: I am available
Matt_King ok, assigned to VO
@Murray I can also work on other tests for VO for accordion and vertical temperature slider and I can have the results for next week
Matt_King accordion is the priority
Joe_Humbert I have assigned myself to accordion too
Matt_King we need another NVDA and JAWS person
Joe_Humbert I can do both, it's fine
Isa I will run the bot and then assign them to you
Joe_Humbert I will wait then.
Matt_King I think we are at the point where we should enable for anyone to assign the bot.
Awesome. We are good.
Joe_Humbert is there an update on the vertical slider?
Isa they are not ready yet
Matt_King we have resolved what was blocking us so we should be able to unblock it shortly
Joe_Humbert should we record the behavior to raise bugs to Freedom Scientific?
Matt_King They will see the fails and will test the fails and they will raise the bugs. I don't think we need screen recording for things like that. They will believe it. They will start running these test automatically. IF they have it approved it they will test it automatically on all of their build!
@everyone: that is fantastic!