There's something I call manic-depressive programmer syndrome. It used to occur infrequently, but -- I think -- because tools have gotten so much better (everyt

There's something I call manic-depressive programmer syndrome. It used to occur infrequently, but -- I think -- because tools have gotten so much better (everything from the quality and speed of compilers and class libraries to the speed of hardware) it's becoming more common. We get a fantastic 1.0 product that never gets followed up. At best you see bug fixes and updates, but the extraordinarily promising core never gets expanded. I'll cite some examples: * TextMate * Silo 3D * The Hit List * Hypercard Each of these is characterized by being developed by a very small team (sometimes one person), being extraordinarily impressive out of the starting gate, and never developing to its full potential.