When my father was doing his PhD, there was one computer in the whole university, and you could run one job daily, if you were lucky. Punch cards, and really thinking things through before you submitted your latest version of the program.
Me, I'm running three jobs simultaneously, on three different projects. Two of those are 'just to see what happens'. Power to burn, baby! This morning I was running a 1000-line interpreted script repeatedly, running a few hundred fits (not complicated ones, admittedly) each time, in order to change a few graphical details of the presentation. It was easier to just run the whole thing, and waste the CPU, than to modify the code to run only the minuscule five-line part that I actually needed.
Yay for resources!
*Looks for things to make his computer do, just for fun.*
Posted by Icarus (Member # 3162) on :
When I worked as a programmer at ORNL, I knew I had implemented one of the regression algorithms wrong when I started a run, saw that it was going to take a long time, and went home for the day . . . only to find it just one eighth of the way through the next morning when I returned.
(This was on a Sun Minicomputer in 1992.)
Back when my parents were working with computers, I suppose I would have been taken out and shot . . .
Posted by rollainm (Member # 8318) on :
If only I had the cash to satisfy my desires...
Quad-core (two Intel Core 2's in one) processors coming out early 2007...I can't believe how fast technology is moving these days.
Posted by King of Men (Member # 6684) on :
Oh, and come to think of it, I've got some big analysis jobs running on the farm, in addition to the local stuff I do on my own computer. The batch system dislikes me, though, it gives me a lower priority just because I tend to dump in a thousand jobs before I go home for the day. Posted by The Pixiest (Member # 1863) on :
I once had a project that had a busy wait built into it to keep from thrashing the file server (this was in the early 90s)
One day I was examining my code, looking for ways to make it more efficient and I thought. "I have a busy wait both before and after this routine runs. Who the hell cares if I save a few microseconds???"
The actual writing of my code was much quicker after that.
Posted by Bob_Scopatz (Member # 1227) on :
When I was in grad school, each of us was alloted a few thousand dollars of CPU time a semester. I wrote a transfer program in IBM JCL to get my data onto the mainframe since there wasn't a way to convert from 5 1/4" floppys created by TRSDOS onto mainframe acceptable format.
Apparently my routine took over the CPU and had it just waiting there, mouth gaping, to grab any data that my 300 bps modem might care to send it.
I burned through my entire account in one session.
I'm glad it was just funny money!
Posted by Icarus (Member # 3162) on :
Are you sure it was a 5 1/4" floppy and not a 7" one?