With the current implementation, we generate RDF (XML and Turtle) and deploy it to a server, from which it can be fetched when that URI is dereferenced. We also deploy the same triples to a triplestore.
Problems:
We have two separate things to maintain
Can we set up content negotiation on the server (which handles the URI dereferences) to get the data directly from the triplestore? If so, we can maintain public accessibility to just one thing: the triplestore.
Of course, the content negotiation would still need to be set up, but it won't need to change each time there's a change to the data.
And we still need to generate HTML to return to clients who are after a human-readable doc.
Is Virtuoso capable of doing this? Certainly, we should be able to formulate the appropriate SPARQL query as part of the ReWrite Rules for content negotiation.
Scenario
We create a new dataset, and one of it's URIs is sameAs a URI in an existing dataset. By sending this sameAs triple to the triplestore where the existing URI is stored, it becomes immediately available when that URI is dereferenced ... which is nice. OTherwise, we need to do what we do now, which is to re-generate and re-deploy the existing dataset.
With the current implementation, we generate RDF (XML and Turtle) and deploy it to a server, from which it can be fetched when that URI is dereferenced. We also deploy the same triples to a triplestore.
Problems:
Can we set up content negotiation on the server (which handles the URI dereferences) to get the data directly from the triplestore? If so, we can maintain public accessibility to just one thing: the triplestore.
Of course, the content negotiation would still need to be set up, but it won't need to change each time there's a change to the data.
And we still need to generate HTML to return to clients who are after a human-readable doc.
Is Virtuoso capable of doing this? Certainly, we should be able to formulate the appropriate SPARQL query as part of the ReWrite Rules for content negotiation.
Scenario
We create a new dataset, and one of it's URIs is
sameAs
a URI in an existing dataset. By sending thissameAs
triple to the triplestore where the existing URI is stored, it becomes immediately available when that URI is dereferenced ... which is nice. OTherwise, we need to do what we do now, which is to re-generate and re-deploy the existing dataset.