Seems like kareldb is thinking about Kafka as the persistent store for a DB, which AFAICT is similar to how Crux thinks about Kafka. I don't have an opinion about any of this yet, so FYI only: https://yokota.blog/2019/09/23/building-a-relational-database-using-kafka/

@dcj cool, I see this is on HN now - thanks for mentioning it here. I already knew of Robert because he built something called KDatalog 😉 We are definitely both building off of the same core philosophy of decomposition and leverage, though I'm not sure I really see any benefit here (or future vision) over just using Postgres...but I am probably missing something!

Still trying to wrap my head around using Kafka as the persistent datastore, this article lays out the case for doing so... https://www.confluent.io/blog/okay-store-data-apache-kafka/