The Limits of Mikado Method
summary
Mikado Method replaces big redesigns with a drawn change graph of small reversible steps, using short cycles and a defined green state to know when nothing is broken. It succeeds when fast, reliable tests exist (use approval tests/characterization tests to add safety) and fails when tests are missing, flaky, or take minutes. Applied to a barber‑shop simulation and a migration with sound effects it produced thousand‑node graphs, small commits, and easier trunk‑based development integration.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:18What is Mikado Method
- 2:46Core principles
- 3:49Simple refactoring example
- 11:15Complex refactoring with graphs
- 13:26Benefits
- 15:59Limitations and challenges
- 17:50Slow or flaky tests
- 21:51Cycle time and applicability
- 23:59Meaningful PR changes
- 24:41Trunk based development
- 26:33Green state and feedback
- 28:00Mikado in branch workflows
- 29:26Q&A
related talks
57mDealing with Medusas in your Codebase.
46mThe Surprising Effectiveness of Small Steps
34mData-Driven Refactoring
35mRedundancy Driven Development - Navigating disruptive codebase changes.
32mMaking Sotware Better, Despite Having Users
47mLiving Domain Model
28mManaging App Complexity
57mThings That Work So Well for Me I Can't Believe You're Not Using Them