since a lot of tests will be interdepent and need to happen in order, it seems easier to not use a test runner enforcing tests to be semi-independent and instead just run the code and have some logging code to see where a problem occurs