Mutation Testing with PIT
summary
Links a 1999 Mars failure to weak tests and proposes mutation testing as the concrete fix: run PIT mutation testing, generate mutants (change + to − etc.), and check which tests kill vs survive mutants. Demonstrates running PIT in IntelliJ, a 5,000-line Hazelcast Kubernetes plugin where runs take ~5 minutes, and CI integration that publishes HTML reports to GitHub Pages. Shows disabling noisy mutators to avoid equivalent mutants and claims PIT makes test quality measurable beyond line coverage with PIT mutation testing and GitHub Actions integration.
AI-written summary · about
chapters (11)
- 0:00Introduction to mutation testing
- 1:49NASA spacecraft failure lessons
- 6:06Why code coverage isn't enough
- 8:13Mutation testing concept
- 12:54PIT demo in IntelliJ IDEA
- 17:27PIT demo with report
- 19:54PIT on Hazelcast Kubernetes
- 23:48Why mutation testing wasn't used
- 29:30Mutation testing in banking
- 33:23Companies using mutation testing
- 34:47Summary
related talks
22mMutation Testing with PIT
41mMutation testing in practice
45mMutation Testing
20mMutation Testing: how to evaluate the effectiveness of your tests?
45mJulius Mischok: Test Your Tests!
36mMutation Testing: Did my test break my code ? 🤔
44mMutation testing: Too good to be true?
50mFrom jUnit to Mutation-Testing