Tidy First? A Daily Exercise in Empirical Design
summary
Argues that when code is messy and a feature is needed the right move is usually tidy first but not too much, because structure changes buy optionality yet cost money and time; uses Constantine’s equivalence to show cost ≈ cost of change, highlights net present value and power‑law “jackpot” changes, prescribes three moves—differentiation, unification, deletion—plus tiny interruptible refactor slices, parallel old+new deployments, and strict separation of structure vs behavior commits.
AI-written summary · about
chapters (11)
- 0:00Intro
- 1:15The idea behind Tidy First?
- 8:24Why software design is a hard problem
- 9:44People
- 15:13Money
- 24:35Software
- 36:12Empirical software design
- 37:07Consequences
- 46:00Q&A
- 48:26AI rant
- 51:18Outro
related talks
53mKeynote: Tidy First? A Daily Exercise in Empirical Design
53mIntentional Code - Minimalism in a World of Dogmatic Design
01h 03mTactical Design
58mIntentional Code - Minimalism in a World of Dogmatic Design
31mIncremental Design - A conversation with a designer and a developer
57mIntentional Code - Minimalism in a World of Dogmatic Design
01h 01mDesigning for complexity
50mIntentional Code - Minimalism in a World of Dogmatic Design