leisner.wbst139@xerox.com (02/23/91)
Yeah, a multi-threaded file system would be most useful. If I have some time (later) and a machine to use, I'll do it if no one else does. [I haven't done anything with Minix since I got 1.1 up and running in x286 protected mode. x286 is a nightmare -- x386 is much better]. What's our opinions on OS research? I read Ken Thompson did away with different OSes and people are basically repartitioning/rewriting Unix (Sun is going to C++, Minix, Mach, etc). Any reason to go to a language other than C? The C stuff works, I've been doing object-orientated stuff in C for years. C++ may be useful (to compile GNU troff and stuff). I'm not sure there is a need for it. I haven't seen anything on a machine that can't be done in C (in fact, C++ translates into C). Language research seems to have basically stopped (except in some pockets). Everyone has a silver bullet (ADA is the DoD silver bullet -- I've seen ADA compilers [which don't work anywhere near as well as gcc]) and form complicated and confusing software systems. I'd like to see Minix given away (with Prentice-Hall selling media for those who want it). Software is going to be free -- there's too much quality stuff around. Once everyone knows where to find this good stuff, software houses will fold since they don't serve a useful societal purpose (except to restrict rights and make money). I think the GNU people pretty much have things about right. Any comments and questions, feel free to ask/post. Opinions wanted. marty (Knowledge is useful in the Information Age) (I hate slow, non-intuitive non-creative computer systems) ARPA: leisner.wbst139@xerox.com NS: leisner:wbst139:xerox UUCP: hplabs!arisia!leisner