Closed johnomotani closed 1 year ago
I think they are useful, but not useful enough if they're at all annoying to spend time trying to fix -- remove them if they're a pain!
Would it make sense to just remove the lines https://github.com/boutproject/xBOUT/blob/afb5ba631568081307a56a5afe997d3d979b78ff/.github/workflows/pythonpackage.yml#L46-L47 I think that would mean that if the codecov upload fails, it doesn't fail the CI. Then someone can always re-run by hand if the upload failed and they want to see it.
Oh yeah, that might be nice
Merging #272 (5fcb4da) into master (afb5ba6) will not change coverage. The diff coverage is
n/a
.
@@ Coverage Diff @@
## master #272 +/- ##
=======================================
Coverage 68.99% 68.99%
=======================================
Files 15 15
Lines 3209 3209
Branches 780 790 +10
=======================================
Hits 2214 2214
Misses 733 733
Partials 262 262
:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more
[Updated] The upload to Codecov fails intermittently, which is annoying:
master
.fail_ci_if_error: true
setting from the other CI jobs, so the CI runs still 'pass' even if the upload fails. The jobs can be re-run manually if the coverage upload is needed.Do people find the test coverage statistics from Codecov useful? I find I never really look at them, and so just get annoyed when the upload fails and I have to re-run CI jobs. If they're not useful, maybe we should just remove them everywhere?