Closed Blizzke closed 10 years ago
This sounds like something which should be implemented by a separate testrunner.
It doesn't fit to the goal of this runner. Also the mechnics which are used in clever-and-smart don't fit to your feature request.
@staabm is right, it’s kinda out of scope.
More like a feature request. I once worked on a huge project where we couldn't run tests in isolation mode because of the amount of time that added on top of the already lengthy cycle. Without knowing we introduced a couple dependencies between tests that resulted in a lot of sh*t later. Having the option to randomise execution would've prevented this sort of behaviour.