Could anyone tell me if it is preferred to implement the GAGenome "copy" member as a deep copy or a shallow copy in a derived genome class? I don't want to bore anyone with the details; but this is quite an important issue for the data members imbedded in my derived genome class. A. du Toit