I imagine that this could be useful when you want to perform multiple operations on a subset of the data from the database at the client-side. So you load the R

I imagine that this could be useful when you want to perform multiple operations on a subset of the data from the database at the client-side. So you load the RESULTDB to client and then perform subsequent queries client-side and those actually do the JOINs This could dramatically reduce the load on the database for some applications as well as simplifying some queries that can be awkward to do fully in SQL (some GROUP BYs can get quite complicated to understand) Another useful thing I can see is using this to simplify the creation of edge databases backed by a central database, like imagine creating a RESULTDB "view" at the edge using this syntax.