I’m looking at implementing a large distributed graph database. So far, the plans have centred around SPARQL, given that it is a standard of sorts, and despite a few posts I see dgraph has no plans to support this.

I could make a shim and use SPARQL/dgraph, or use a different database, or perhaps just change plans and use graphql.

So the question is why not support SPARQL? Is there something inherently difficult with it and dgraph? Was it just a decision to go with one because there wasn’t bandwidth for both? Is there some inherent superiority of graphql over SPARQL?

It would be great to have some detail about this.


It’s not been a decision. We just have limited resources and we have been spending those on providing one solid QL, i.e. GraphQL±. In the future, we could add support for SPARQL or others, depending on the needs of the user community.

That is good to know, thanks.

Yes, please do try and add SPARQL support.