Hi @davidLeonardi
The main reason to use Apollo in that case was we wanted to selectively expose the mutation so that the frontend can’t execute those and to have some logic before we send the mutation/query to Dgraph hence you see that the input to the queries/mutations is also different because few of them are computed on the Apollo server and then sent.
Now with custom logic and auth we can do most of these things with the Dgraph GraphQL schema itself, having said that I will soon be writing a blog(will include the steps and try to answer all your points in more depth) and sharing the examples around usage of Apollo federation and server in general with Dgraph GraphQL. We have some more content which should be coming in next few weeks.
Hope this clears it a bit.
1 Like