What kind of use-case are you thinking of? Do you mean something that could be accessible through Datalog? And hey 😎

I’m thinking about getting a separate crux node for a different domain inside the same app. I have one node already for “cucumbers” and this one will be for “car parts”, so no overlap between them. I want separate “indices” or nodes for hygiene.

Separate nodes using separate tx-logs in a single app is a reasonable approach. There is no shared cache or thread pooling between nodes currently though, so efficiency could be improved if you have a lot of nodes. We want to work on this area soon. We have also discussed having a node config "filter" to only index documents in the desired namespaces, but then I would question the value of funnelling everything into a single tx-log anyway. There are no concrete plans for adding something like that for now.

cool ! I’ll go with a second node then, thanks!

We're not freezing the API but it should generally be more stable. Accretion-only of the overall API is the long term goal. I can't promise that we're there quite yet. As for beta, GA is a much more exciting and useful target, which is why we opted for the low-key transition. We are fortunate enough to not be short of a good number of early adopters :)

Got you, thanks!