I disagree: when fixing a bug I usually start with writing a failing test that reproduces and exposes the problem, and then fix the problem so that the test pas

I disagree: when fixing a bug I usually start with writing a failing test that reproduces and exposes the problem, and then fix the problem so that the test passes. Understanding, reproducing and fixing the problem is the hard part, once that is done making a test out it hardly takes any time. Writing the test does not only prevent regressions, it also helps with confirming you actually fixed the bug.