http://invisible-island.net/personal/
Copyright © 2014,2019 by Thomas E. Dickey
In the late 1970s/early 1980s, I worked at a large research center, maintaining and expanding a computer simulation of a large-scale static VAR generator. Laszlo Gyugyi, my manager, had designed the system, but was not a computer expert. Gyugyi mentioned a couple of the people who had worked on this program before, suggesting that they might assist in explaining some of its details:
However (as I was told by Jim Koos), Burroughs introduced a newer model … which was slower than the older machine. The computer center shopped for a replacement, and settled on a Univac because it was better than IBM for numerical work.
NUAlgol (Norwegian University) was available for the Univac, but was not satisfactory (I was told that Univac did not support the program). When I took over the simulation, it had been translated into Fortran (2-3 times longer—about 2500 punchcards). Jim Vine showed me a listing of the NUAlgol program once. It was less than a thousand lines, partly because it packed boolean data into bits. That allowed the program to be more concise.
Though I encountered him frequently doing computer analyses, James Vine's work is seen more in patents than papers.
These links give summary and content:
Here is a list (ordered by issue date):
I found only one paper:
Besides the patent and paper citations, this
Several methods for calculating LEED spectra from stepped surfaces, including layer doubling and transfer matrix (CHANGE), are compared for calculations of LEED spectra from vicinal (stepped) surfaces with their narrow spacings between planes. A particular type of failure which occurs only for certain spacings from top bulk layer to bottom surface layer is analyzed.
likely refers to the same James Vine, based on the tidbit offered by Google:
... detailed comparisons can be made. The various programs built are as follows: (1) The original CHANGE program, converted to FORTRAN by James Vine, and running now on a personal computer. (2) The same CHANGE program ...
We did not work closely together, but would meet and chat occasionally in the user's area, which had a half-dozen computer terminals (all different types) and a table.
While the simulation which I worked on had been read onto disk (and stayed there), Vine maintained his programs on punchcards, in a box. A box could hold 2000 punchcards, and his box was full.
Seeing that, I asked why he did not read the program onto disk as someone (probably Jim Koos) had done. He pointed out that it was cheaper.
Over the course of four years, and requests from my manager for optional features to explore design tradeoffs, I expanded the simulation to what would have been more than four boxes of cards. A small deck of cards (probably no more than ten) was enough to compile and run the program.
Being single at the time, it was convenient to make that my main meal. But as I commented once, “I make all of my own bread.” Vine asked “do you also grind your own wheat?”
I was startled by that, but apparently some people really do. Since then, I have seen this, e.g., using the misnamed “WhisperMill” but have still not actually done it myself.
For the record, I started here.
On another occasion, we were talking with some others in the computer user's area, when someone asked what would we have done before computers.
This would have been in the late 1970s.
I suggested something like bookkeeping or accounting.
Vine was more definite: “We would be servants.”