lets see if we can run all the tests on osx, linux, win #77
Conversation
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
|
Looks good. I don't think the conda-forge maintainers want us running the full test suite on every build, since that would consume a lot of CI resources. But it's useful to test this and see what happens. |
|
I'll double check the logs from this PR and the last one I merged in to see if the testing output on osx is different in them to make sure this is doing something. I don't want to waste CI resources but I also don't want to ship broken packages to our users and since we have a documented case where that happened, we should be okay. |
|
Wasting CI resources is not the issue. We have 200 parallel jobs for 10k+ feedstocks and this is taking 53 jobs (which is better now because of the removal of older cuda versions). If the jobs are quick, that shouldn't affect much, but the more time it takes, the more chance of this feedstock being a bottleneck for other maintainers. |
| @@ -95,6 +94,8 @@ test: | |||
| commands: | |||
| - env with_test_suite={{ with_test_suite }} bash test_openmm.sh # [unix] | |||
| - set with_test_suite={{ with_test_suite }} && .\test_openmm.bat # [win] | |||
| - env with_test_suite={{ with_test_suite }} bash test_openmm.sh # [osx] | |||
There was a problem hiding this comment.
I think this line might be redundant... The unix selector two lines above is an alias for linux or osx.
Just using this to test if OSX is running unit tests.