Closed Zeitsperre closed 11 months ago
I have no idea why I keep getting Segfaults on GitHub Actions. There has to be something strange going on with the test setup that isn't memory safe. Is this possible?
Can you remind us of what's wrong with installing ESMF through mamba on the CI ?
And why is nc-time-axis
a dependency, I don't think we plot anything in xscen ? (I mean outside the docs)
I'm going to squash these commits to not pollute the history.
Pull Request Checklist:
number
) and pull request (:pull:number
) has been added.What kind of change does this PR introduce?
install-esmf-action
~install-esmf-action
-installed ESMF appears to cause some segfaults with pytest. I'd love to get to the bottom of it, but I know when to quit. Here's hoping this isn't exposing a problem that we might again have further on.)tox
tox-conda
tox
now offers a means for installingesmpy
from a tag/branch (viaESMF_VERSION
environment variable)setup.py
setup.py
Does this PR introduce a breaking change?
No.
ESMF_VERSION
is only needed iftox
is called with-esmpy
flag. Thetox
config was always dependent on an environment havingesmpy
, therefore there are few instances where it would work anyway (requirestox-current-env
ortox-conda
).Other information:
https://github.com/esmf-org/install-esmf-action (I think I'll wait for the next version before trying this again).