Topic: Daily Info
Hmmm, access to a Cray X1. Not as cool as Blue Gene or The Columbia Installation, but seriously cool anyways, and about the closest I'm going to come to a "real" (they make our little Unisys Box look like an Atari 800) supercomputer until the lab starts working with Pittsburgh
Not that I'm really into writing that kind of code anymore. My interest in large parallel computers is quickly waning. I find AI and networking to be more fun. Anyone can work in AI and networking; only goverments, univeristies and large corporations can do large parallel computing.
Yup, about 6 hours of actual code and paper writing combined with about 15 or 16 hours of straight-up thought on how to take some well known algorithms and do them effeciently in LISP. So all told, 24 hours of work on one assigment. And call me weird, but I actually enjoyed doing it. It was fun, like programming back in freshman year was fun. But instead of very simple programs, I'm now writing nice and complex programs that do something interesting. I did kind of finish up in a half-assed manner, and now the program has a slightly larger runtime (about 0.01 seconds longer on average). If I was willing to go back in tomorrow, I think I may have a fix that would drop the run-time penalty. But otherwise it did work: it made the program much more memory effecient and it takes fewer trials to find the goal. Worked the first time after clearing up some small glitches (you write 600-700 lines of LISP and not make one or two errors).
So yeah, I think AI and something with networks will be my major-emphasis. It's just too much fun.
And I wish I could sleep, I really do. I finally got to bed around 4 yesterday and then got up at 6, and did my marathon. And add on the extended discussion about systems software at work (I'll go into it later) and I'm just about fried.
The plan for tomorrow is to clean up a few slushy stains in the car and shop about for some new boots.
Later-o.