Systematic Debugging

Systematic Debugging

Random fixes waste time. Quick patches mask issues.

Core principle: ALWAYS find root cause before fixes. Symptom fixes are failure.

The Iron Law

NO FIXES WITHOUT ROOT CAUSE INVESTIGATION

When to Use

ANY technical issue: test failures, bugs, unexpected behavior, performance, builds, integration.

ESPECIALLY when:

• Under time pressure
• "Just one quick fix" seems obvious
• Already tried multiple fixes
• Previous fix didn't work
• Don't fully understand issue

Don't skip when:

• Seems simple (simple bugs have root causes)
• You're hurrying (systematic is faster than thrashing)
• Manager wants NOW (systematic prevents rework)

Four Phases

Phase 1: Root Cause Investigation

For test failures, check flakiness FIRST:

Test fails → Run 5x
├─ Passes 5/5: Not flaky, investigate as bug
├─ Fails 5/5: Consistent, investigate as bug
└─ Mixed (3/5): FLAKY TEST - fix test first

Flaky test checklist:

Then continue:

1. Read Errors Carefully - Stack traces, line numbers, error codes
2. Reproduce Consistently - Exact steps, happens every time?
3. Check Recent Changes - Git diff, dependencies, config
4. Multi-Component Systems - Add diagnostic instrumentation at boundaries BEFORE proposing fixes
5. Trace Data Flow - Where does bad value originate? (See pop-root-cause-tracing)

Phase 2: Pattern Analysis

1. Find Working Examples - Similar code that works
2. Compare References - Read reference implementations COMPLETELY
3. Identify Differences - List ALL differences
4. Understand Dependencies - Config, environment, assumptions

Phase 3: Hypothesis & Testing

1. Form Single Hypothesis - "I think X causes Y because Z"
2. Test Minimally - Smallest change, one variable
3. Verify - Worked? → Phase 4. Didn't? → New hypothesis
4. When Unknown - Say "I don't understand X", ask for help

Phase 4: Implementation

1. Create Failing Test - Use test-driven-development skill
2. Implement Single Fix - Address root cause, ONE change
3. Verify Fix - Test passes, no other tests broken
4. If Fix Doesn't Work
   • STOP. Count fixes tried.
   • If < 3: Return to Phase 1 with new info
   • If >= 3: STOP. Question architecture (see below)
5. If 3+ Fixes Failed: Question Architecture
   • Each fix reveals new problems elsewhere
   • Fixes require "massive refactoring"
   • Pattern fundamentally unsound?
   • Discuss with user before more fixes

Red Flags

STOP if thinking:

• "Quick fix for now"
• "Just try X and see"
• "Add multiple changes"
• "Skip test, manually verify"
• "It's probably X"
• "Don't fully understand but..."
• "One more fix" (after 2+)

ALL → Return to Phase 1

3+ failures → Question architecture

Quick Reference

Real-World Impact

• Systematic: 15-30min to fix, 95% first-time success, near-zero new bugs
• Random: 2-3h thrashing, 40% success, common new bugs

Cross-References

• Flaky tests: pop-test-driven-development (Condition-Based Waiting)
• Root cause tracing: pop-root-cause-tracing (backward tracing)
• Defense: pop-defense-in-depth (multi-layer validation)

Examples

See examples/ for:

• flaky-test-patterns.md - Common flaky test causes & fixes
• debugging-flowchart.pdf - Visual decision tree
• multi-component-diagnostic.md - Instrumentation strategy