Cracking the Code Review: from Guesswork to Automated Guidance
summary
Claims that code review friction costs scale (cites $316 billion 2013 estimate) and that automating trivial debate cuts review time by ~20–30%. Recommends concrete practices: concise, tested PRs (avoid 55-file dumps), self-review, templates, targeted reviewer pings, start-day reviews. Lists tools that actually get applied: Error Prone, OpenRewrite, Google Java Format + Spotless, Checkstyle and ArchUnit for rules, plus inline coverage and editor-based review. Shows how these tools remove bikeshedding so humans focus on design, tests and responsibility.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:57The role of code reviews
- 2:55Foundations of pull requests
- 3:26Common challenges
- 4:03Best practices for reviews
- 15:40Effective reviewer strategies
- 25:44Automating away the noise
- 37:49Conclusion
related talks
45mCracking the Code Review
45mCracking the Code Review
43mCracking the Code Review: from Guesswork to Automated Guidance
41mCracking the Code Review from Guesswork to Automated Guidance
59mCode Reviews - Why, what and how
01h 01mBetter Code Reviews FTW!
18mHow To Review Someone's Code While Not Being A Jerk!
17mThe Perfect Code Review