With Great Composition Comes Great Responsibility
summary
Composability in Datomic lets a single database hold multiple customer data models, so a built visibility service uses query rules and multimethod dispatch to choose paths (group→animal or direct user→animal) and adds attributes like user.assigned-animal without backfilling. Tests use the Datomic with helper and curated regression datasets. Query order matters: a poorly ordered datalog query ran ~24 hours until the most restrictive clause was moved first, then became practical.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:22Overview
- 4:38Atomic Architecture
- 6:11The Client Project
- 9:05Entity Attributes
- 10:42Data Log
- 14:48New Customer
- 15:10Schema Change
- 15:50Database Restrictions
- 17:58Multi Methods
- 19:16Lessons Learned
- 22:49Outro
related talks
46mThe Datomic Architecture and Data Model
40mThe power of Datomic database
51mDatomic
52mEnhancing Datalog for Datomic
41mExploring four hidden superpowers of Datomic
37m11 insights after 11 years with the functional database Datomic
33m¿Where's My Data?
40mDatomic: The Most Innovative DB You've Never Heard Of