Just to help firm it up, have you tried to see things from their perspective? What is the minimal violation they could be guilty of, if you gave them the benefit of the doubt on every close interpretation? I'm not familiar with the Plucker/Bluefish architecture. Could it be they are distributing an unmodified Plucker binary as part of their larger Bluefish distribution, and they believe that since you make the Plucker sources available they are meeting that obligation (not actually enough, but could be forgiven/corrected)? Could they possibly be using a looser definition of derivative work than you, and simply be making minor mistakes in their conformance to source distribution needs?