Thanks for the example.

Another question I have related to DRb processing is: 

Can I get a 'clone' of an object so that I can process it locally? I am
worried about performance. If each reference to the object uses the DRb
client/server connection I might end up with performance problems. But if I
could get an objects current state with one DRb call and act on the object
locally I could greatly improve performance.

Has someone documented ways to improve performance over a DRb client/server
connection?