-
Notifications
You must be signed in to change notification settings - Fork 11
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Display console outputs for all tests #147
base: main
Are you sure you want to change the base?
Conversation
Otherwise you can't see console logs of passed tests
This will let us see the output of changes like #146 |
You did not ask me for a review of this. Perhaps you were only doing this as a test? I don't think it is a good idea if you really want this added, it makes the output too verbose. I can give more commentary if you like. |
I wanted this because I was reviewing #146 and I can't see the changes in behavior re the warnings. For that example the test runs are basically completely opaque with no console outputs. https://github.com/ros-infrastructure/rosdoc2/actions/runs/10658077327/job/29538644576?pr=146 For automated tests like this being able to go back and find the differences in behavior is valuable. These tests are going to be the only arthives we have of these results. We'll be searching them anyway, and the summaries of the errors and failures still will be generated. This is something that the ROS buildfarm errs on the side of very verbose. But then it helps track things down much faster from failures when they occur, because you can do thiings like compare it to the last know success. I'd be happy to hear more of your thoughts though. |
In the particular PR that you mentioned, the "changed behavior" is that the revised tests succeed, while without the change they fail. See this run for what happens if you just have the changes to the test, without the change to the core code: https://github.com/rkent/rosdoc2/actions/runs/10778128703 Isn't that the normal pattern for test-driven development? A reviewer could confirm that by updating the tests without updating the underlying code, which is what the patch does. I can't always guarantee that every change I do will have an associated test, though I try. But #146 does not seem to me to be a good example of the need for more verbosity. What I would support is using But that does not help #146. The warning there is buried in conf.py, which for the current implementation is not using the standard logging package. I added a 'warning' there using So generally I am not a fan of adding output verbosity for edge cases, as it makes more work in normal cases, and tends to obscure problems more than highlight them. (A particular peeve of mine is the |
Otherwise you can't see console logs of passed tests