You write tests specifically for testing multiple changes. You shouldn't be testing changes to global state by seeing how multiple supposedly independent tests interact.
Is there any value in designing tests in such a way that they test multiple things at once while still being able to isolate which specific thing is responsible for breaking? I'm thinking of something like JMP.
Comments
You write tests specifically for testing multiple changes. You shouldn't be testing changes to global state by seeing how multiple supposedly independent tests interact.
Is there any value in designing tests in such a way that they test multiple things at once while still being able to isolate which specific thing is responsible for breaking? I'm thinking of something like JMP.