Both Cascalog and Datomic have chosen to use Datalog (over SQL) as their query engine.
Dave Thomas made the claim:
Datalog is better than SQL for large queries in small amounts of space.
My question is: Is datalog more efficient than SQL for column oriented databases?