A Statement from John Hudson
JH = John Hudson
GAM = George Michael
GAM: One of the sadder aspects of these war stories is that I was unable
to get to certain persons before they passed away. Here is a
statement by John Hudson, prepared even before he knew he was
going to be interviewed. He died while this project was being
formed. John was one of the first Livermore programmers; quiet,
gentle and a prodigious producer.
Statement of John Hudson, November 29, 1989
JH: When I first started (at the Lab), I was assigned to the UNIVAC
computer group. I worked on several things; Neutronics codes
(programs) using the Monte Carlo method, developing utilities for
mathematical subroutines, molecular dynamics, and some exposure
to Hydrodynamic methods. The codes were written in machine
language when there was no floating-point hardware or index
registers.
With the proliferation of computers over the next few years, I was
assigned to the first compiler group at the Lab, and was involved
with writing a compiler for the IBM 704. When it was completed, I
worked on some hydrodynamics programs using both our compiler
and the first versions of FORTRAN. I also wrote subroutines and
utilities for the IBM series of computers, the 704, 709, 7090, and
7094.
I was next assigned to write programs for the UNIVAC LARC
where I coded special heat flow programs using the Monte Carlo
method. I also produced other mathematical routines and utilities
and special input-output programs for devices such as the
CalComp incremental plotters and the LARC Electronic Page Recorders
(EPRs) for film recording.
My next big assignment was as a group leader for the Nuclear Test
Division computer support group. This group grew over seven or
eight years from one to 16 programmers. My function in this
position was to decide which person was to be assigned to which
programming job. The tasks were very diverse, each requiring
some sort of specialty. Anything from chemistry, or mathematics
to file management. I was always involved with the overall design
of the slate of programs used by the entire division. The major
project which I did for EG&G was to develop a mini-computer-based
control and monitoring system for test area procedures at the
Nevada Test Site. It sent control signals and monitored data over
microwave communication links and provided dialog with test
monitors as well as direct memory access, and servicing several
local input-output devices. This work made use of a Varian 620I
ruggedized machine that included every hardware option available
from Varian. We wrote the program in assembly language, and
assembled on one of the computers based within the Livermore
Time-Sharing System, because it took far too long to assemble on
the 620I.
During this period I was also involved with devising methods for
utilizing large blocks of idle time on the main frames with the time
sharing system. This involved the management of very large
programs having only low priority, running only in background
mode. The value of these strategies was shown best by the
improvements in system efficiency.