Yes they share, I wanted to try writing relationsh...
# spicedb
m
Yes they share, I wanted to try writing relationships using triggers
v
It would work, but the challenge here is that you are bypassing the API layer and all that comes with it (fundamentally data validation and the ORM logic). Particularly for postgres it would be very tricky to get the MVCC stuff right
so I don't think it would be advisable to go that route. But another thing you could explore is subscriptions: https://www.postgresql.org/docs/current/sql-createsubscription.html
you could create a program that subscribes to database changes on the main app tables, and then use the SpiceDB API to write them
I'm not sure 100% but that may be also what
zed experimental import
does 😅
so yeah zed import is for bulk importing (initial backfill), while the connector is for continuously propagate changes
3 Views