NEW254149
[Results Dashboard] List failing results
https://bugs.webkit.org/show_bug.cgi?id=254149
Summary [Results Dashboard] List failing results
Sam Sneddon [:gsnedders]
Reported 2023-03-20 03:36:45 PDT
It seems like at an API level this was implemented in bug 203905, but we don't have any easy way through the UI to just get a list of failing results for a given test. This is often useful when investigating flakiness, as it's often most useful to just look at the results when the test does fail. I don't want to have to scroll through all the test results to find failing results—I'd just like a list of most recent failures.
Attachments
Radar WebKit Bug Importer
Comment 1 2023-03-20 03:36:55 PDT
Jonathan Bedard
Comment 2 2023-03-20 08:20:14 PDT
I don't think https://bugs.webkit.org/show_bug.cgi?id=203905 is quite the API version of what you're requesting here. It sounds like the request here is basically "give me a list of test runs where a given test has failed", https://bugs.webkit.org/show_bug.cgi?id=203905 is "give me a list of all tests which have failed in a given range". That being said, I do think the API gives us everything we need to build a UI for "give me a list of test runs where a given test has failed" easily, I'm just not quite sure what the user story to get to such a view would be. I don't think such a UI would make much sense as a timeline (it would involve much scrolling), I'm thinking either a collapsed timeline which eliminates commits that weren't testing (something Zhifei and I have discussed in the past) or something akin to our "investigate" tab which gives you a list of builds.
Sam Sneddon [:gsnedders]
Comment 3 2023-03-20 08:22:10 PDT
(In reply to Jonathan Bedard from comment #2) > I don't think https://bugs.webkit.org/show_bug.cgi?id=203905 is quite the > API version of what you're requesting here. It sounds like the request here > is basically "give me a list of test runs where a given test has failed", > https://bugs.webkit.org/show_bug.cgi?id=203905 is "give me a list of all > tests which have failed in a given range". Ah, I misunderstood from skimming what that issue was! > That being said, I do think the API gives us everything we need to build a > UI for "give me a list of test runs where a given test has failed" easily, > I'm just not quite sure what the user story to get to such a view would be. > I don't think such a UI would make much sense as a timeline (it would > involve much scrolling), I'm thinking either a collapsed timeline which > eliminates commits that weren't testing (something Zhifei and I have > discussed in the past) or something akin to our "investigate" tab which > gives you a list of builds. Yeah, I'm not sure what the right UI (and path to the UI) is either.
Note You need to log in before you can comment on or make changes to this bug.