Clarify behaviour of include_relationships when calling dump() on schema? about marshmallow-sqlalchemy HOT 1 CLOSED

Quick additon - #292 seems to ask the same question, and the response seems to line up with my final suggestion of using fields.Nested as a workaround (but without the conditional test which is wrong). Also it looks like I've misunderstood the original justification which is to set include_relationships to False, specifically to avoid it duplicating the values already providied by include_fk? It was never designed to include auto generated nested schemas in the dumped JSON, nor is such a thing possible currently by any other means. This is covered fairly conclusively in #98, I think.

Unless anyone comments otherwise (i.e. my summary above is wrong!) I'll close this shortly - I note in the issues above at least a couple of other people seem to have made similar incorrect assumptions to me. It might be worth making the context and purpose of include_relationships setting a bit more explicit in the otherwise good documentation on the library, in case a few other people have fallen into the same trap as me! I can raise another ticket for that if people think it is worthwhile?

from marshmallow-sqlalchemy.