The FARFALLA package consists of a few C++ class definitions and a few additional C++ subroutines. The choice of C++ was critical; indeed had we chosen to produce something similar in FORTRAN the final result would bear little resemblance to FARFALLA. C++ not only made our job as implementors easier (because of the built-in memory-management features) but also makes the job of users easier. One of the most critical features of FARFALLA is that it is designed to be extended; that is, users may organize their data into new types of nodes by creating additional class definitions based on the existing FARFALLA classes. They may then create analysis applications by writing C++ routines that use the FARFALLA classes (as well as the user-defined extensions), and call FARFALLA routines. We have also written some DASH modules that integrate FARFALLA fully into the DASH environment, although it is also possible to write stand-alone FARFALLA applications independent of DASH.