Prioritizing Technical Debt as If Time & Money Matters
summary
Shows how version-control mining plus static probes turns vague technical debt into actionable priorities: uses hotspot analysis on Android’s ~3M LOC platform framework (2,000+ devs) to find frequently changed, complex files; drills to function-level hotspots (handleMessage: 98 edits, ~500 LOC inside a ~20k LOC file) and combines 20–25 code health metrics (cyclomatic complexity, low cohesion, copy‑paste) to rank fixes. Demonstrates simulation of offboarding risk on .NET Core (350k LOC, 180 contributors) and uses CodeScene for the visualizations.
AI-written summary · about
chapters (8)
- 0:00Intro
- 0:33Lehman's "laws" of software evolution
- 3:17Technical debt
- 7:39Case study: Prioritizing technical debt
- 10:25Can we measure "code complexity"?
- 19:27Technical debt & people
- 22:33Case study: How quickly can you turn your current codebase into legacy code?
- 27:48Resources
related talks
20mPrioritizing Technical Debt as if Time and Money Matters
26mTackling tech debt
17mHow to deal with technical debt in fast-paced environment
35mTechnical Debt: When Interest Eats Your Code
23mPrioritisation of SCA Findings in Software Dependencies Using Static Reachability Analysis
44mDistributing complexity in huge python projects.
26mTech debt as innovation, reframing this forever problem as an opportunity
37mA Complete Journey of Android Modularisation