I collected and plotted quarterly economics data for a professor who worked on the "index of leading economic indicators" in the business school. Lots and lots of time series data that get revised every quarter. So a typical plot would be an x-axis of years from 1960 to 1980 and the y-axis would be percentages, and the graph should fit on a 8.5"x11" sheet of paper (but the plotter was huge and use large rolls of paper).

I did the key-punching and submitted the CalComp flatbed plotter job to the batch window at the computer center. Quite a few times early on, there would be some error and I would not get any output, just a note that the job had an error and they had to kill it. I'd go back and check the job and try to figure out what was wrong and submit it again, and a few hours later check and find that it was killed again.

I eventually talked to the person who ran the plotter and he said when my job would run it would start off fine, then would try to spool out 25+ feet of paper at very negative X-values, plot one point, roll the paper up, plot another point on the proper axes, roll out the paper again, etc.

The plotter programming language was such that one had to map the data points to the points on the paper, and I was telling it that some of my data had x of 1960 and some of it had x of 3...

Yeah, back in those days, computers did what you told them to and didn't clean up after themselves or ask "do you really want to do this??".

:-)

Thanks for the story. Don't let it bug you too much, though. ;-)

Cheers,
Scott.