[arc9693]

While it gives a nice approach to generate data for parameters with which functions can be tested, I have doubt about the correctness of returned values of say mocked functions. Since this is not a TDD approach, and relies on correctness of functions - function callers etc. in production, how reliably can we say that the generated mock data based upon monitored data in production is correct? What I mean to say is, a corrupt function will give corrupt outputs, be it in prod or dev env. Isn’t it more reliable to have the code author write mock functions?