Ten things you heard about testing that might be wrong
summary
Ten myths get debunked with concrete examples and numbers: studies show ~50% of developers hit flaky tests monthly and Google reports 41% flaky tests, Microsoft 26%; reliance on Espresso, Robolectric and Firebase testing cloud raised CI time and flakiness; one product drops all UI tests and removes most flakiness; AI test generation (Gemini/JPT style) copies existing wrong behavior; advice is to test behavior not structure, avoid mocking data classes, reduce mocks across boundaries, and balance automation cost versus bug-prevention value.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 3:12Any change breaks a test
- 6:39What is a unit test?
- 8:25Testable code dangers
- 10:33Overuse of mocks
- 14:48Test coverage misleading
- 17:57Flaky tests problem
- 21:32Over-testing and pyramid
- 25:20Automate everything?
- 27:29AI tests limits
- 33:28Rethinking testing pyramid
- 37:00Key takeaways
- 38:45Closing & Q&A
related talks
39mTen things you heard about testing that might be wrong
46mThe state of the Art in Tackling Flaky Tests
40mFinding Flakiness: The Real Reasons Your Tests Aren't Reliable
39mTesting, Testing, 1, 2, NaN
31mYour Tests Aren’t Flaky, You Are!
28mThe State of Art in Tackling Flaky Tests
46mPanel Discussion: Breaking Myths About a Career in Testing
02h 46mThe Joy of Testing