When fetching RDF information, Piggy bank should send an Accept-Header of application/rdf+xml first, to obtain RDF metadata. This is not the case: for example, on http://crschmidt.net/julie/ , the doap metadata is not retrieved, because it is content-negotiated.

I thought this happened at the jena level, but maybe not.

Additionally, it may be wise to indicate in some way, in the user-agent, what the hit is for: LiveJournal, for example, may not like seeing lots of hits from a "Java_1.4.2" user-agent. (See http://www.livejournal.com/bots/ for more info on that, although it may not apply.)