The Code is Documentation Enough
summary
Argues that code must express intent so compilers and humans catch bugs: demonstrates renaming a weighted‑average implementation, replacing a raw loop and confusing lambda with a zip transform pipeline, and refactors a fat statistics function into 20‑line units to improve testability. Shows strong types for meters/seconds and using MPUnunits by Matio push to avoid unit mixups, advocates unit tests as executable documentation, and lists tooling like clank tidy, clank format, ReSharper C++, SonarQube plus contracts and expects/ensures for clearer interfaces.
AI-written summary · about
chapters (10)
- 0:00Expressive code intro
- 3:19Variable naming best practices
- 8:53Function naming best practices
- 12:45Strong types for clarity
- 21:36Organizing code for readability
- 28:03Unit tests as documentation
- 32:44Language features leverage
- 39:48When to use comments
- 44:03Code readability and consistency
- 47:12Summary
related talks
52mThe Code is Documentation Enough
40mGood Code: What, Why, and How to Get There
20mGood Code is a Balancing Act
29mHumanizing Your Documentation
59mTechniques for writing high-quality code
01h 10mCoding for everyone
56mClean Code, Two Decades Later
53mIntentional Code - Minimalism in a World of Dogmatic Design