Hi!
Stardog is a great product and I love the attention to UX! I do have some questions regarding the graphql implementation, specifically wrt manually registered GraphQL schemas.
- In the docs, "Only the parts of the graph defined in the schema will be exposed to the user" is listed as one of the advantages of having a custom schema. However, how can you manage what graphql apis are exposed? The automatic one seems to be always on.
- Probably related: is there a way (possibly with directives) to map graphql types or properties server-side, instead of in the query (eg. by using @withPrefixes?) Another post seems to suggest this was added to a prior release, but the docs don't mention this. Having this client-side is not a huge problem, but it does require a lot of knowledge about the server-side.
- When I switch on the automatic schema generation, it generates a schema that uses the @iri directive, which doesn't seem to be documented. I suppose it is used to indicate that an graphql type corresponds with a particular RDF type, but it doesn't seem to work when I try to apply it in a custom schema. It reminds me of the hypergraphql approach, which is very intuitive IMO.
Thanks!
Best,
Miel