Egad! How Do We Start Writing Better Tests?
summary
Advocates a practical, pyramid-shaped test strategy: use pytest for all layers, unit tests as a fast 1 ms foundation, integration ~1 s, end-to-end ~1 min, and avoid trying to automate everything; prioritize happy paths of core features and skip brittle edge visuals. Recommends concrete tools: requests for APIs, Selenium WebDriver (or Selenium Grid/SauceLabs/BrowserStack) for UI, Hypothesis for property testing, pytest fixtures and pytest-bdd to reduce duplication, and scale CI with pytest-xdist + Jenkins and ReportPortal. Notes costly failures (Mars orbiter $655M) justify integration tests.
AI-written summary · about
chapters (8)
- 0:00Introduction to better tests
- 3:54Why test automation is hard
- 7:28Test automation pyramid explained
- 12:53Avoiding test code duplication
- 16:28Focus on ROI for automation
- 17:55Running tests in CI/CD
- 20:52Making tests more reliable
- 22:40Resources and Q&A
related talks
29mEgad! How Do We Start Writing (Better) Tests?
21mDon't just test, my friend, test better
35mTurning test writing into a consistently brief and pleasant experience
38mTesting for the Terrified: How to Write Tests, Conquer Guilt, and Level Up
48mA Reflection on Software Testing
27mWhy I use py.test and maybe you should too
11mNeed For Tests: Most Wanted
47mChris Harbert: Building the Testing Pyramid
