Some interesting point and counterpoint related to my last post... rentzsch.com argues that coders don't like to code, they like to solve problems. They like it more when they can solve the problem at hand with less code rather than more. He argues that coders prefer writing their own code over reading and using others' code because they want to understand the problem and its solution better than they might if they had to slog through reading and understanding other people's code.
But I think Joel Spolsky's article, which rentzsch links to, offers more of a solid counter-argument than rentzsch takes it for. Spolsky points out that old code has usually been tested, debugged, repaired and patched to be a far more stable and robust solution than any first attempt at new code. It seems to me that, ideally, those internal patches would have been documented within the code in such a way that later programmers would understand the purposes for that code. Even without such commenting, if the code works, it behooves the programmer to understand the full breadth of value that old code represents, long before he/she dives into potentially reinventing the wheel.
