To DDD or not to DDD? Domain modelling case study
summary
Rewrites a 20+ year legacy titling system for dog sports, replacing a brittle recursive rule engine with a per-program template and a pure function show-results→titles that needs only show results and returns audited titles. Uses Golden Master testing against hundreds of thousands of historical cases from the database, finds 9 rule types in schema but only 7 handled in code, implements missing rules, adds a visualization tool for per-dog review, deploys to production with only a few minor fixes, and enables parallelism and repeatable, testable calculations.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:54Domain
- 10:00Algorithm
- 13:04Problems
- 16:48Solution
- 20:25Visualization tool
- 21:56Implementation highlights
- 23:46Program refinement
- 25:05Implementation
- 28:52Key differences
- 30:09Implementation perspective
- 30:49A program
- 31:18Multitalented doc program
- 32:40Overall outcomes
- 33:36Simple code
- 34:28Learning process
- 35:46Finding balance
- 38:33Getting stuck
- 40:09Breakthrough
related talks
01h 01mThe DDD Do-Over: Domain-Driven Design Lessons Learned
39mLessons Learned from 10+ Years with Domain Driven Design
47mDomain driven redesign
38m7 Years of DDD: Tackling Complexity
46mTechnically DDD
42mHigh-Speed DDD (revisited)
38mDomain-Driven Transformation - how to improve the structure of legacy systems
48mPitching DDD to the management