Not Your Mothers Datalog
summary
Compares Datomic-style graph stores (Datomic, Datascript, Crux/XDB, Asami) to academic Datalog, showing Datomic stores triples (e,a,v) with per-statement metadata and indexes, supports recursion, rules, aggregates and negation, but uses a different syntax for rules so Datomic is semantically Datalog while query language is a graph language; demonstrates evaluation strategies and a real case where switching from top-down to bottom-up evaluation cuts a 90 second query to ~100 ms, and notes Asami’s per-statement IDs and statement-level metadata.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 0:35Welcome
- 2:01Relational Databases
- 3:28Graph Databases
- 4:36Graphs vs Tables
- 7:34Query
- 9:22Sparkle
- 10:08What is Datalog
- 12:58Operational Features
- 14:47Processing Mechanisms
- 15:41BottomUp
- 16:55Types of Data
- 17:36Datalog vs Predicate Logic
- 18:47Rules
- 19:13Querying
- 20:36Atomic Datalog
- 21:10Conclusion
