I'd say at least because: - the first 10+ years of CCTBX did not use pytest. AFAIK, the first attempt was by our postdoc Youval Dar back in 2015 (correct me if I'm wrong). I feel adding different testing styles are only to make the code-base inconsistent (very much like mixing flex and np arrays isn't cool, in my opinion!). - originally tests were considered as simple usage examples for functionalities they are testing; this is because writing and (most importantly!) maintaining the proper documentation was not provisioned. A simple test like def exercise(): """ Make sure 2*2 is 4. """ x=2. result=x*x assert approx_equal(result, 4., 1.e-6) if(__name__ == "__main__"): exercise() print "OK" is much easier to grasp rather than the same cluttered with the stuff (that, to add to the trouble, one needs to learn in the first place!). All the best, Pavel On 3/3/18 14:36, [email protected] wrote:
What’s bad about pytest?
On 3 Mar 2018, at 02:26, Pavel Afonine
wrote: Just to make sure: you are converting to use pytest this particular codes (fable), correct? Pavel P.S.: I'm allergic to pytest.
On 3/3/18 07:46, CCTBX commit wrote:
This in preparation for pytestification.
cctbxbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/cctbxbb