Currently the tests are run using the latest version of pyaml in the git repository because of this:
- name: Install dependencies
run: |
python -m pip install --upgrade pip
python -m pip install "git+https://github.com/python-accelerator-middle-layer/pyaml.git"
python -m pip install -e ".[test,epics,tango]"
pip install flake8
This is causing problems because it blocks merges unless all changes are compatible with the changes in the main branch of pyaml. So for example it become difficult to add small bug patches, updates to the docs etc.
I suggest that we change the workflow to always test against the latest release on pypi and then add an optional workflow to be used by developers to test against the latest version on the main branch. It looks like we can always have both workflows triggered and allow to merge if it's only the developer test that fails.
What do you think?
Currently the tests are run using the latest version of pyaml in the git repository because of this:
This is causing problems because it blocks merges unless all changes are compatible with the changes in the main branch of pyaml. So for example it become difficult to add small bug patches, updates to the docs etc.
I suggest that we change the workflow to always test against the latest release on pypi and then add an optional workflow to be used by developers to test against the latest version on the main branch. It looks like we can always have both workflows triggered and allow to merge if it's only the developer test that fails.
What do you think?