At some point (I missed the memo) #xtdb stopped espousing #datalog and invented #xtql. I'm guessing this was to step out of any annoying dependencies of datalog and dampen comparison with the other major datalog approaches in #Clojure.
https://docs.xtdb.com/intro/what-is-xtql.html I especially enjoy the side-by-side Clojure/JSON comparisons here
@worldsendless this for the unreleased version 2, I think
@mykhaylo mrmm.. I didn't see that anywhere. But, to be fair, it's been a while since I've worked on my XTDB projects
@worldsendless @mykhaylo The main issue is that in order to fully support SQL - even for a schema-less DB - there needs to be some notion of "tables" and Datalog doesn't have that. An early v2 build had Datalog+tables but it was awkward to use and had several open questions -- so they designed a new "inspired-by-Datalog" query language that is composable and explicit and provides full parity under the hood with the SQL support.
@seancorfield @mykhaylo I just came across the old mailing list discussions they (Crux) were haviig with you in their early days about this. Thanks for summarizing and filling me in!
