Applied Color Systems PDP-11

RT-11 Documentation
This just about sums it up for RT-11 complexity. My employer bought a competitor in Philadelphia and operated it as a separate entity for a few years before deciding to pull the plug. Having been once burned by a misguided color system purchase, the company was reluctant to take the plunge again. However, the subsidiary company somehow obtained approval for a modern computerized color control system. Being savvy to the capabilities of these systems, I immediately requested transferring the color system to my location only to find that it was leased for 5 years. I convinced our factory manager that we could show enough savings in 1 year to pay the full 5 year cost of the machine but I didn't want to become an accountant in the process. Fortunately he agreed to assume the lease and off I went to Princeton to get the necessary training. It was there that I met Ralph Stanziola, one of the founders of Applied Color Systems that went on to become Datacolor International, an industry leader yet today. The most valuable thing I gained from that week in Princeton was a lasting friendship that endured over 30 years.

After teaching my Technical Assistant most of what I'd learned, we set about to build colorant data files only to discover a few things, both good and bad. First, the performance of our first colorant calibrations were not good. Secondly, although the color software was suitably user friendly, the RT-11 operating system of the PDP-11 computer was not. The commands were mostly cryptic, meaningless, difficult to remember character strings and yes, the above instructions were prophetic, one needed the pocket reference to avoid deleting a file when your intent was to copy it.

My tinkering nature took over and applying Macro-11 assembly soon resulted in several command interpreting, batch processing routines used regularly. We later received a new and improved release of RT-11 using regular words for the operations, like RENAME, PRINT, COPY, etc. How novel! Some of the longer ones had shortcuts as I recall like REN for RENAME. Amazing how many of these things carried over exactly to PC-DOS and MS-DOS arriving on the scene later, in addition to having been predated themselves by CP/M used briefly on my C3. Wonder who did what and who copied who?

Resolution of the colorant calibration was related to pigment loading levels. By using the original file to reformulate the primary standards at more realistic loadings and calibrating these new results produced a colorant data file that survived for years. We upgraded to a more powerful PDP-11 running multiuser RSX-11M with spectrophotometers and workstations both in the laboratory and factory. This system served us well into the late 80's when PC's became powerful and cheap enough to replace DEC computers and ACS, Datacolor International now, no longer supported the older systems. I had a new Datacolor PC based color system in my office when leaving my last job in 2002 - an experience lasting nearly 25 years.

About the promise to pay for the original system in a year. My company calculated part cost using averages weighted by historical color popularity and then set a fixed part price regardless of color. Popular colors were mostly black, tans, browns, etc. For some reason, red was one of the more favored interior colors for Chevrolet trucks that model year. Without computerized color and cost prediction, a $45/lb red pigment ended up in the red parts. Usually not a problem since red was never too popular. However, red comprised enough of the sales that we were losing money, a lot of it on the interior parts. Using the computer to reformulate the red without creating color problems lowered the cost of red yardage enough that we saved the $80,000 lease balance in less than 6 months. A Datapoint routine analyzed the production schedule and supplied a weekly report to the factory manager showing the current and accumulative savings. As promised, it was discontinued after hitting the target taking me out of the accounting business.

Curiously enough, the cheapest formulation is not always the one containing the least expensive base pigments since pigment strength and opacity are part of the equation. Numerous cases can be shown where the balance of colors needed to produce a match can result in the most economical mix being one with a more expensive pigment. Black and white being normally the least expensive, a stronger, more expensive pigment may result in a proportional increase of black and white content enabling total pigment reduction to produce a cheaper formulation. Faced with matching more neutral colors having dozens of possible combinations, computerized formulation is the only way to go.