“Houston we have a problem! This change you want… Well it’s one small change for the codebase but one giant change to the test suite!”

We all know why tests are great. But every once in a while we just wish they’d never been written. One case when test regret occurs is when we want to alter our functionality slightly but in doing so half the test suite goes from green to red.

This talk will investigate why this happens. In particular it looks at the coupling between tests and the code under test. We’ll investigate ways we can reduce this coupling and make out test suites more resilient to change.