This is based on a talk I gave at the AEI in 2006 (before lunch), aimed at physicists familiar with FORTRAN or C, but not so much with object oriented practices.