The 1970’s were exciting times at Rockwell International. That was when the great microelectronics revolution was just beginning, and Rockwell was one of the leaders in developing the technology. They were well ahead of a little company named Intel.
Rockwell had two divisions working on MOSLSI devices. (See Note 1) I worked for the Microelectronics Device Division that produced and sold the devices. The other was the Microelectronics Products Division. It used the devices in various consumer products, including hand-held calculators and modems, which enable the transmission of digital information on phone lines.
I started out as a programmer, writing software for customers who did not have software development capabilities. Software development was an arcane process back then, mostly involved with huge “mainframe” computers made by companies like IBM. Most businesses making consumer products had no software designers, so Rockwell provided that service, at a price, to promote sales.
After a couple years, I was promoted into management, supervising about a dozen software designers. I soon learned that management was not my thing, so I started looking around for something that kept me involved with the design of products using microprocessors, and found the perfect spot for me: Marketing, technical support for sales. I traveled around the US and western Europe, talking to the engineers in companies that our sales staff had contacted, who were interested in “computerizing” their products.
Now, a lot of the old-timers in those companies were skeptical of this new-fangled idea. They were skilled in using conventional electronic components like transistors, diodes, resistors, capacitors and the like to implement the control functions in products like microwave ovens, cash registers, etc. How could this tiny chip do all that? My job was to convince them that not only could a chip replace a lot of those circuit boards, but it could enable them to add a lot of new features to their products. I had to convert them into believers in this new technology. I was a computer evangelist.
I loved the job. I am an engineer, and I know how engineers think. I enjoyed sitting down with them and learning about their products, and then coming up with a design concept using a microprocessor. I had experience in both hardware and software design, so I could talk about the design problem from both perspectives.
But first I had to proselytize them. They had to believe in the Lord of Electrons, the microprocessor. And sometimes, I was a bit too successful.
Think about religious believers. There are your ordinary folks, who go to church sometimes, throw money in the collection plate, sing the hymns and bow their head when prayers are recited. If you ask them about God, they will say yes, they believe in the Lord. But it doesn’t dominate their life. They realize that they live in the real world, and that’s where they concentrate their energy and attention. And then, there are the fundies and the evangelicals. Religion dominates their lives. God is ever-present, looking over their shoulder, judging them. God can do anything, including miracles!
Sometimes, the engineers I talked to got this fanatical gleam in their eye. I knew what was coming. They were going to make SuperToaster or SuperOven that would do everything but cut your toenails.
I had created a monster, and I had to kill it before it destroyed the product. I had to rein in those wild-eyed fanatics, with some real-world reality, about the limitations of processor speed and memory size. (See Note 2) Sometimes that was harder than overcoming their initial skepticism.
Of course, there is a big difference between the evangelizing that I was doing, and what the churches do. I was selling a new design concept for real world products. The religious folks are selling pie in the sky.
Note 1: MOSLSI stands for Metal Oxide Semiconductor Large Scale Integration. It enabled thousands of microscopic transistors called MOSFETS (field effect transistors) to be placed on a single square silicon chip, about a quarter inch on each side.
Note 2: The limitations back then sound laughable today. Our first products operated at 200 Kilohertz and memory was limited to a few kilobytes of ROM memory where the program was stored, and a few hundred bytes of RAM memory where data was saved and accessed. Today, processor speed is measured in Gigahertz and memory size in Gigabytes. And it probably costs less than those toys we worked with forty years ago.