Low CodeCov coverage threshold#996
Conversation
Testing a lower coverage threshold for CodeCov testing, since the current settings are distracting from this early stage of testing.
Codecov Report
@@ Coverage Diff @@
## dev/gfdl #996 +/- ##
===========================================
Coverage ? 43.24%
===========================================
Files ? 213
Lines ? 62216
Branches ? 0
===========================================
Hits ? 26903
Misses ? 35313
Partials ? 0Continue to review full report at Codecov.
|
The absolute code coverage target was replaced with a 100% threshold, tagging any level of coverage as a success. This will be modified later, when the coverage tests are more thorough and extendible by users.
|
This patch appears to successfully disable the "failure" coverage report, but there seems to be an unrelated issue that is nonetheless relevant to testing: There are now two Travis "tests", and only one is being successfully reported while the other is still waiting. I am not sure where this fourth "check" has come from: Reading a bit suggests that this is the older Travis CI, and that the newer one is reporting success: I don't think this is caused by my patch, because #995 is showing the same problem, and is still using the older (current) Travis and CodeCov YAML files. None of the older PR seem to have this problem (e.g. #994, #993, #991) and appear to show the "old style" Travis report. So I don't know why things would have just suddenly flipped. There are a few posts suggesting that we need to turn off the old style reporting, but it's just a guess. Shotgun googling comes up with this: but there are a few other links saying the same thing. But this all appears to have happened about a year ago, so no idea why it's suddenly happening now. |
3 participants
Testing a lower coverage threshold for CodeCov testing, since the
current settings are distracting during this early stage of testing.