Crux – open-source answer to Datomic

Pretty excited about this for a bunch of reasons.

- MIT licensed. Pretty big risk reduction factor for users if the project is run well.

- Transaction time and assigned 'valid' time axes. We can use transaction time to audit the system while assigning our own time basis for queries (e.g. midnight billing triggers)

- Data stored as documents but indexed for graph traversal

- Datalog/EDN interface

- Kafka for primary distributed store and transaction log

- Highly pluggable and customizable

Looks cool, but years of cynicism regarding “new” databases has me deeply suspicious of things like “Kafka as a database”. I mean Kafka is already a complicated enough thing I refuse to rely on it 100% for its normal usage. How can I ever trust it as a persistent data store? The answer is I can’t. Others will have to blaze this trail. Sorry but my love of Datalog is insufficient to motivate me to A: trust Kafka, and B: add java/clojure/anything JVM to my tech stack.

My reservations aside, serious nerd <3 for the Datalog support, I really wish more data storage has native support for Datalog. I hope this project matures to the point cynical system architects like myself can trust it, because Datalog is just that awesome. The project looks like it’s off to a great start so keep it up, and don’t let the cynical types [ like myself ;-) ] derail you.

Great!!

Finally!