Remembering Gary Kildall

In the late 1970s, PBS showed a series called Connections, where James Burke showed us how every invention was built upon the shoulders of someone or something before it. I believe this to be a principle of human evolution.

One of the inventions was the computer, although at the time the series was created, 'computer' meant what we now think of as "mainframe computer." In other words, nothing you could have at home.

Or could you? During my 50th high school reunion, I discovered that one of my classmates had been using a computer at home all the time we were in high school (the early 1960s). The difference between her family and mine was that her dad worked at IBM, and what they had at home was a terminal into an IBM mainframe in an IBM office. With this, they could operate the mainframe from home or anywhere else a terminal existed.

This "personal" aspect of the early computer industry is important, because it illustrates the desire of computer technicians to have personal control over their charge.

As the semiconductor industry developed, the first microprocessors came into existence. They were primitive versions of what was going on in the big room-size computers, and they were used for relatively trivial tasks. The Intel 8080 that was behind the personal computer revolution (at least, the part of that revolution that interested me) developed from a processor requested by a Japanese company to use in their calculator.

Although this activity was going on in what we now call Silicon Valley, up in the Seattle area, related things were happening. Kildall was first at the University of Washington, then down at the Naval Postgraduate School in Monterey. The UW connections would be important for the rest of his career.

The energy that made all these things happen was largely personal initiative. That desire to have computing power accessible at all times drove the creation of the personal computer.

Kildall went on to develop an operating system, a way to transfer data from storage into the processor and back again. He tried and failed to create a hardware disc controller, once the floppy disk had been created to replace the IBM punch cards, but hardware wasn't his choice of world in which to play, and he asked a UW friend to help him:

This was 1974. Out of frustration, I called my good friend from the University of Washington, John Torode. John was a Ph.D. graduate of the Electrical Engineering Department there, and I figured if John couldn't build that controller, no one could. He did build it. He designed a neat little microcontroller and, after a few months of hardware and software testing, that microcontroller started to work. We loaded my CP/M program from paper tape to the diskette, and "booted" CP/M from the diskette, and up came the prompt: *

Kildall's system—CP/M—allowed for the commercial development of the personal computer. Hardware people could create machines in any way they chose, but the operating system would be the same for all, and that meant that software created for it would run on all those machines.

Nothing stays the same, and the microprocessors developed too. When IBM decided to get a piece of the personal computer market, Intel had developed a much more powerful chip than the 8080--the 8086, and a lower-priced version of this was used for the first IBM PC. The functions of CP/M were duplicated for this new processor, and that became the operating system for the IBM PC, eventually to evolve into Windows.

What Gary Kildall did would have been done by someone else later on. In that sense, it was all inevitable. But he did it first, and all subsequent developments proceeded from that work.

And all this was happening while I was in graduate school at UW. I would know nothing about it until years later, but it feels good to be able to say "I was there!"

Wikipedia article on Gary Kildall