What is stopping us? A legacy code question
summary
Legacy code rot comes from systems not stupidity: management incentives, noisy environments, bad processes (example: a Jira board that spawns duplicates), brittle tooling and dependencies (NuGet version lock), and resource churn. Cited study of ~200 programmers in 35 organizations shows the top 25% performers have low interruptions. Code value is threefold — source, tests and the mental model — and team churn leaks that model, so the claim is teams are immutable: change one member, you effectively have a new team. Systems thinking maps reinforcing vs balancing feedback loops to find where to intervene.
AI-written summary · about
chapters (13)
- 0:00The state of software teams
- 2:01Management and systems
- 3:32The impact of environments
- 5:54Technical and process debt
- 7:33Processes and agile reality
- 11:30Applying systems thinking
- 15:24Loops and modeling systems
- 18:46Managing team dynamics
- 22:05Holistic system evaluation
- 24:07The value of code and models
- 28:22Sustainable team stability
- 33:28Envisioning better outcomes
- 36:52Final Q&A and wrap-up
related talks
54mThe Future of the Past – Legacy Will Stay With Us
27mYour Brain Runs On Legacy Code
26mLegacy Codebases are a DevOps Issue
34mHow do our ideas about coding affect the software we create?
52mTechnical Neglect
42mLegacy Came to Stay
36mWhy programming is fiendishly difficult and what we can do about it
52mOMG, What a Beautiful Legacy