Fridolin Glatter

I'm noticing an issue with the latest approach: if you let the checks runs once, then notice you need some further refinement, and run them again, Codecov will likely consider...

Just for completeness: I have now decided that the simplest workaround should be to run the test suite/measure coverage twice: on PRs and pushes to `main`. This is a tad...

Thanks for starting this PR! I have reworked the setup part of the tests so that `test_mp`, `mp`, and `test_mp_f` are now parametrized for `"jdbc"` and `"ixmp4"`. Please take a...

I'm not sure if the sqlite dsn `":sqlite:///:memory:"` is sufficient for running multiple tasks on the same machine. We might have to integrate the unique `platform_name` here, too, but at...

Thanks for the suggestion :) I'm not sure how `handle_config()` can help me in this case, though. I'd like to make the logic handling of e.g. `dsn` more robust, but...

@khaeru, I can't (meaningfully) request your review here because the PR was opened by you. In principle, this should now be good to go :) I don't understand what's happening...

So `windows-latest-py3.13` failed because of some flaky test. Rerunning it now seems to yield nothing because I consistently get lots of errors that gams could not be found. Could it...

I have looked into the CI failures again and still don't understand them, but may have found a solution for the most critical one. So what I observed in the...

I've added tests for some edge cases now (warnings raised by `IXMP4Backend` for unused parameters, etc). `get_data()` and the `cat_*()` functions are officially untested, but I think this is only...

Especially with the logging, this does not work as intended. Locally, all tests in `test_ixmp4.py` are passing, but on here, the tests checking the logging output fail. This makes me...