The benefits of automated testing are many, and a lot of programmers and teams are doing test-driven development and extensive automated testing. Yet many struggle with maintaining their test suite, and end up in situations where every little change to the code breaks some tests. Often it is not clear why the change causes the tests to break, thus requiring the programmers to spend an annoying amount of time working out whether the tests are failing for a good reason or not. This talk looks into the issue by identifying a number of test smells that contribute to this problem, and suggests some patterns and strategies that will improve the current state of practise.
|
