You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have a custom sopel module that I wrote and am using this container to run sopel.
When I restart the container, it tries to re-install my package that I have specified in EXTRA_PYPI_PACKAGES=sopel-modules.tableflip
That results in an error starting the container.
this container should only try to install packages on first run, not on restarts.
pewp | Installing collected packages: sopel-modules-tableflip
pewp | ERROR: Could not install packages due to an EnvironmentError: [Errno 13] Permission denied: '/home/sopel/.local/lib/python3.8/site-packages/sopel_modules.tableflip-0.1.2-py3.8-nspkg.pth'
pewp | Check the permissions.
pewp |
pewp | FAILED!
pewp exited with code 1```
The text was updated successfully, but these errors were encountered:
Currently the container "tries to reinstall" packages on each startup for simplicity. pip should not complain about packages that exist already; instead it should just move on with a simple Requirement already satisifed: ... message with minimal overhead.
Can you provide more details on your setup? I am unable to recreate this issue on a fresh sopelirc/sopel:6.6.9 container with EXTRA_PYPI_PACKAGES=sopel-modules.tableflip with multiple restarts of the same container.
The permission error you show suggests that this may be a result of how you are handling/mounting volumes.
are you mounting any part, or the whole, sopel user home directory? It definitely makes sense to treat the ~/.sopel directory as a separate volume, but including things that are highly variable, e.g., ~/.local (by mounting the entire home directory) can lead to issues like this.
did you install a development version of your plugin (e.g., via setup.py install develop or even pip install -e ...) into this container/mounted volume? If so, did you do it as the container's sopel user? If not, did the UID/GID of the other user (external/insider container) match the sopel user in the container?
did you change owner UID/GID for a mounted volume after the initial container startup? check/update/fix ownership of /home/sopel/.local/lib/python3.8/... in the container (or in your mount source) to match the sopel user in the container
P.S., sorry for the extreme delay in getting to this issue!! 😬
I have a custom sopel module that I wrote and am using this container to run sopel.
When I restart the container, it tries to re-install my package that I have specified in EXTRA_PYPI_PACKAGES=sopel-modules.tableflip
That results in an error starting the container.
this container should only try to install packages on first run, not on restarts.
The text was updated successfully, but these errors were encountered: