-
Notifications
You must be signed in to change notification settings - Fork 39
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Draft: Follow-up numpy 2.0 (DLLAPI) #451
base: devel
Are you sure you want to change the base?
Conversation
I think I will try to use |
I changed my mind. Since |
For now, something is broken, and it is not clear to me why. I will ask on numpy repo: ImportError: /home/runner/.local/lib/python3.10/site-packages/jiminy_py/core/core.cpython-310-x86_64-linux-gnu.so: undefined symbol: EIGENPY_ARRAY_APIPyArray_RUNTIME_VERSION here is the corresponding issue on Numpy. |
After a long discussion with numpy team, it turns out that numpy API should not be exposed the way it is to avoid compatibility issues. Actually every import of numpy headers should be moved to the source file. It is the only way to ensure that EigenPy public API does not leak numpy internal symbols. |
Maybe moving the discussion here (also). I am still happy to mitigate the linking issue in NumPy if you/ In a sense, the obvious solution I (and others) on NumPy see right now is to ask users to do a normal NumPy include with the typical NumPy pattern:
in most files, and:
in the main module file before any import The question is whether that seems easy enough for the |
No description provided.