Closed tobiscode closed 5 months ago
Oh, I didn't change the version number, that's maybe something to consider to make it clearer for pip
installations to see it was updated and now has different requirements.
Wow! This is fantastic. Is it the python version that needs to be updated in environment.yml? Something else?
@tobiscode This is fantastic!! Thank you so much. I'm going to give this PR a look over as soon as I can but I'll probably just merge it. Then I'll release a new version. It might be this week but it might take until this weekend because I'm on a work trip and it's quite busy.
@brendanjmeade If I remember correctly, there are several places. environment.yml and setup.py come to mind.
Ok, I got excited and looked through! All great stuff. Thanks again.
Awesome! Since setup.py
isn't a thing anymore, the only place to update the package version is pyproject.toml
(version = "18.12.08.3"
). For the users, nothing should change since all changes are backwards-compatible / enforced by the use of the meson backend, so the part requires-python = ">=3.8"
inside pyproject.toml
can probably just stay. I don't see an environment.yml
file but that is probably also fine.
Thanks Tobias! That makes a lot of sense.
Hi, I was able to track the error to some unintuitive (or buggy) behavior in how
meson
and/orf2py
use the.pyf
file. Specifically, it seems that the part where it specifies the output of theDC3D(0)
functions is at least partially ignored. By adding the intent as a comment into the source.f
file, the expected behavior is restored. (I'm not sure whether the.pyf
file is still needed, but I kept it in for now.) I took the liberty of addressing some linter messages and removing the LaTeX dependency fortest_okada.py
to make it easier for people to test everything, please feel free to ignore these parts of the PR if desired. I hope this fixes https://github.com/tbenthompson/okada_wrapper/issues/8 ; I've tested it with Python 3.12 myself and the installation and andtest_okada.py
work now. Cheers