The Joy of Testing
summary
Argues tests must prove behavior, not inflate numbers: shows managers demanding 80% coverage and how line vs branch coverage differs, demos AI that blindly achieves 100% branch coverage but yields useless tests, and runs PIT mutation testing to find a surviving mutant that exposes gaps. Covers ObjectMother/test-data factories, AssertJ over assertEquals, JUnit 5 extensions to mock LocalDate, parameterized and nested fixtures, Gherkin/Cucumber features, and Golden‑Master/characterization testing used to safely refactor a 35,000‑line class.
AI-written summary · about
chapters (17)
- 0:00Introduction and mindset shift
- 1:08Why write tests
- 4:45Speaker background and testing basics
- 7:38Coverage myths and realities
- 10:11Branch coverage and AI demo
- 16:10Mutation testing explained
- 22:17Fixing bugs ethically with tests
- 30:26Test naming and expressiveness
- 43:40AssertJ vs traditional assertions
- 1:00:32Mocking time and Christmas demo
- 1:12:38Break and Q&A
- 1:27:43Parameterized tests deep dive
- 1:42:28Legacy code testing techniques
- 2:05:16Behavior-driven development with Cucumber
- 2:27:21Domain-specific testing languages
- 2:35:50Hierarchical test fixtures in JUnit
- 2:40:52Recap and closing remarks
