It's nice if testing doesn't produce unnecessary artifacts after running with success but keeps relevant info on fail for further inspection.

Pytest should have a nice way of creating fixtures that can post-process a test based on whether it failed or not https://docs.pytest.org/en/latest/example/simple.html#making-test-result-information-available-in-fixtures

If a test fails, then keep the

Right now we just use the standard C random library, but really it'd be nice if the brute force fuzzer could do consistent tests across platforms, for various purposes. We probably need our own PRNG, something like:

https://www-cs-faculty.stanford.edu/~knuth/news02.html#rng

or

https://github.com/cmcqueen/simplerandom

I'm not sure what's best. We'd like decentish quality, maybe, and d

Per the python docs, int and float objects have some additional methods that aren't part of the abc interface. CrossHair doesn't implement these, and therefore explodes when evaluating code that uses these methods on symbolic values.