neurodata
Authentication Tests
[On Hold] - Pending a decision about how we want to deal with the new setup script in 16.
@Aeusman Here is what I recommend: We should create multiple modes of the same test. One mode runs all these tests as they are on public projects(which is the case now) and have a second mode which runs against all private projects with a pre-created token(as you had earlier). Let's create a token on install (only when installing on travis) and go thorough mode1 and mode2. We can set this mode via providing a conditional command-line argument to py.test which defaults always to mode1. Does this sound reasonable?
Sounds reasonable to me. I'll implement this probably inside test_settings, and do something inside the trais model to test in this way. It might be challenging to do this in a purely py.test environment.
I don't think test_settings will work since we want to run both tests. Let's discuss this offline.