Distributed SPARQL is most certainly desireable if we want to operate over multiple datasets. From my experience, this is not a simple topic, but since we are talking about small scale queries ( directed navigation ) perhaps we can find a simplified solution.

TODO: take a look at DARQ.