Years ago in my undergrad I took a most excellent OS course offered by the indomitable Peter Denning. As presented, his course basically considered each part of

Years ago in my undergrad I took a most excellent OS course offered by the indomitable Peter Denning. As presented, his course basically considered each part of an OS, chose the best of alternatives for core system concepts, and built the OS from there. The result was something not grossly unlike a modern nix, but still leaving enough wiggle room for the large number of variants we see today. His comparisons between the way these concepts were instantiated in an nix and in other OSs were particularly good (and scathing). That being said, it is a pity that there aren't still a dozen or so major OSes out there, each coming from a completely different approach, oh the days when an Amiga or a BeBox were interesting consumer choices. You can still find lots of minor OS variants around, but they are very small communities and it's really all boiled down to a very small and "boring" handful. Additionally, most new OS approaches suffer from a distinct lack of software. And slowly but surely we find the inevitable creep of software Ports into new OSs, using bringing *nix like infrastructure in with them.