Reading up on John Ousterhout
3 deep · digging since feb 16
- A meta-analysis of three different notions of software complexity
A meta-analysis argues that Zach Tellman's definition of software complexity as the sum of future explanations is the most useful and practical among three discussed.
- the hold that could not release itself
Conflating historical facts, contracts, time, and policy in a 'held' flag blocked migrations; the fix separates values from time to avoid stale holds.
- On cognitive debt
Nate Meyvis argues that cognitive debt in AI-generated codebases is not inherently worse than in traditional codebases, and AI can help mitigate it.