search

w3c / aria-at-app

Test runner and results reporter for ARIA-AT
http://aria-at.w3.org/
Other
35 stars 15 forks source link

Feat: Enable automation support for VoiceOver/Safari/MacOS #1101

Closed gnarf closed 1 month ago

gnarf commented 1 month ago

However, the main reason I think a 2nd look should be had here is because the prompt to trigger automation when an eligible report is created seems to be missing. I would have expected to be able to simply select that when creating a VoiceOver report now. In fact, it is also missing for when I'm creating an NVDA report as well. I confirmed it still popped up on development as shown in the screenshot but didn't on this branch.

Nice catch! I can take a quick look, but it's likely that the test report isn't passing the key property for at/brower -- Any chance you could give me more direct instructions where that popup shows up and I'll try to carve out 30 minutes a little later today to hunt this down.

howard-e commented 1 month ago

Any chance you could give me more direct instructions where that popup shows up and I'll try to carve out 30 minutes a little later today to hunt this down.

@gnarf sure! On development, it pops up anytime I'm creating a Test Plan Report which uses NVDA as the AT. So in 2 cases currently:

  1. On Test Queue page, as an admin, use the Add Test Plans to the Test Queue section to add a report to the queue which uses NVDA for the AT option.
  2. On the Data Management page, as an admin, use the Test Plan Report status dialog up to add any report which is missing and uses the NVDA AT.
gnarf commented 1 month ago

https://github.com/w3c/aria-at-app/pull/1101/commits/a659c8374e004358a5d44dacdc05e06249e83e67 should fix the issue with the add and run popup:

screenshot of interface correctly showing the add and run popup with voiceover and safari

gnarf commented 1 month ago

Although the test failure is quite puzzling to me (because I can't get it to fail locally). It may be worth figuring out but the assertion will have to be updated to 1 anyways because of #1097 changes recently being included in development that do away with the <span class=sr-only>.

The test that was failing was on the merged branch, so it made sense I just needed to update for the new assumptions.