Type: New Feature
Affects Version/s: None
Fix Version/s: None
In triplestore (T1.domain.com), there is an object (/object/1) who has a reference (eg. owl:sameAs) to an other object (/object/999) in an other triple-store (T2.example.org), which is owned by an other organisation.
I have an application, which want to show object 1 in triple store T1. Because there is a link-reference to T2, the application has to get the xml+rdf of the other object in T2. It wants to follow the URI to object 2 (http://T2.example.org/object/999) in order to read the information.
It tries the following:
curl -H "Accept: application/rdf+xml" -L -X GET http://T2.example.org/object/999
This will fail with blazegraph, because it does not listen on port 80.
It could ask the sparql-point of T2, but this info is not stable, nor reliable. It makes the client software depending on the used triple store (eg sparql endpoint port), which is not a good idea.
If you try this "curl -H "Accept: application/rdf+xml" -L -X GET http://schema.org/Person" it works like it should.
Proposal of new feature: Add an other service to blazegraph, to answer simple requests on port 80.
It could (internal) be translated to a sparql query, and return the resulting triples, or just read the info strait from the database (???).
If it is done well, this "API" could be used to publish the triple-database secure to the internet so other users in the world can reuse your data.