Hello, I'd like to ask a question regarding the client/engine - how does Prisma construct queries for Postgres with some particularly big selects? we're talking a select that spans maybe 7 entities or more (quite extensive GQL query).. I mean, that's what SQL joins are for, but when I look into logs, I see a ton of

prisma:query

selects that operate on a single table with ID constraint, instead of joins through foreign keys.. Is it because of performance? are foreign keys utilized in any other way than referential integrity? can we speed up such selects in any way?

@Lukáš Stuchlík 👋 We went with the option of going for separate queries, in which could be easier for caching purposes. There’s a request for joins here where you can add the above input as well 🙂