Tests are failing because the model transformation server cannot be accessed from AToMPM. However, the logs for the Python server aren't being recorded to help debug.
I propose that a flag be added to the Python server to just start the Python process but not start the server itself. This would then test the importing of all the required libraries and be a good proxy for the server actually running. The flag name could be --dummy or --test or similar.
This should then be tested at the beginning of the GitHub action, such that we get a fail-fast test for the Python server.
Tests are failing because the model transformation server cannot be accessed from AToMPM. However, the logs for the Python server aren't being recorded to help debug.
I propose that a flag be added to the Python server to just start the Python process but not start the server itself. This would then test the importing of all the required libraries and be a good proxy for the server actually running. The flag name could be
--dummy
or--test
or similar.This should then be tested at the beginning of the GitHub action, such that we get a fail-fast test for the Python server.
This can also help with testing/solving #151