Structure reuse, not code reuse, is key.
Edsger Dijkstra wrote in How do we tell truths that might hurt?: "In the good old days physicists repeated each other's experiments, just to be sure. Today they stick to FORTRAN, so that they can share each other's programs, bugs included
(
Read more... )