Gary Kildall was a programmer’s programmer.
Paul Freiberger and I are currently writing the third edition of Fire in the Valley, our history of the personal computer—to be published by The Pragmatic Bookshelf, naturally. As we go back over the material, we keep finding threads that run through the whole narrative. I am sharing some of these as brief history articles here in PragPub.
Gary Kildall’s story weaves in and out of the early history of the personal computer, and the color of the thread that runs through it all is that of a programmer enjoying the challenge of solving tough problems.
Sailors and Spies
As the curtain goes up, we find Gary Kildall already on stage, a respected computer science professor at the Naval Postgraduate School in Monterey, California. He’s a born teacher, happiest when he has a piece of chalk or a pencil in his hand. And he’s really into these new semiconductors he gets before anybody else from his contacts at Intel. There’s a cutting-edge microcomputer lab in the back of his classroom. His students are in-service military personnel pursuing graduate degrees in operations and technical areas, or future spies. The CIA is the funding agency for many of the scholarships on campus. It’s 1972, but there aren’t a lot of anti-war protests going on at NPS. It’s a pretty serious, straight-arrow place, and Gary’s students are motivated.
Apparently teaching computer science to sailors and spies is not enough of a challenge, because Gary is also consulting with Intel, where some of the geekier engineers are rather loosely interpreting their contract to design a controller chip for a calculator. They’ve decided to view it as permission to develop a computer on a chip. Gary writes key software for the chips, the 4004 and 8008. In a clear case of mission creep, he writes a full-blown HLL for the infant processors, an implementation of PL/I that he called PL/M.
(Later, when we talked to Gary’s friend and colleague Alan Cooper about these days, Alan told us that Gary wrote a PL/I implementation precisely because it was hard. The challenge was where the fun was.)
As time goes on, Intel’s engineers see enormous potential in the new microprocessors they are developing. They are certain that successful businesses will be built on the back of their innovations. Wristwatches, Bob Noyce thinks. Or microprocessor-controlled microwave ovens, or blenders, or carburetors or stereo sets. Small computers, meh. Not much of a market there, Intel’s boss thinks. But some of these kinds of plays would pay off. There are discussions within Intel about getting a piece of that action, but ultimately the company sticks to what it knows: semiconductors. Mostly memory.
Intel’s marketing manager, Mike Markkula, starts looking for opportunities. So does Intel documentation specialist Adam Osborne. Gary gets the fever, too. He and UC Berkeley computer science professor John Torode build a computer with a disk operating system and sell a few, but this is really just a hobby operation. Gary’s idea of building a business isn’t quite what you’d expect from a scientifically-minded professor teaching sailors and spies. He decides to build an Astrology Machine.
With San Francisco hardware designer constructing the actual hardware and Gary doing the software, they release their machines on the Bay Area. Gary’s sense of integrity demands that he compute the star positions accurately, even though he considers astrology a joke. The machines look like arcade games of the period, and a number of them get placed in supermarkets around the area. Unfortunately, the printer that delivers the horoscope jams a lot and the business is a failure.
Out of his work at Intel, though, Gary has crafted a just-right-size operating system for the new microcomputers being built from the chips, and the Astrology Machine gives him a chance to field-test it. He calls it CP/M. Designed to work with Intel’s existing and future microprocessors, CP/M isolates the machine-specific code into a component he called the BIOS. This makes it easily portable across Intel’s chips, but also across different vendors’ microcomputers. Gary asks the Intel executives if they have any objections to his marketing CP/M on his own. They shrug and say go ahead.
It would be an exaggeration to call this collection of hobbyist entrepreneurs and the motley products they are offering assembled or in kit form through mail order at this frenzied time an industry. But CP/M is just what it takes to turn into one. Breakthroughs in technology industries aren’t always technological breakthroughs. This time, though, that’s what’s was needed: the solution to a tough technical problem. That’s what CP/M delivers. Everybody wants it.
This time when Gary starts a business it is a little more conventional, and it is an immediate success. You can tell he’s going to have fun doing it, though: the company is officially named Intergalactic Digital Research. Its core product is CP/M.
The breakout microcomputer in 1974 is the MITS Altair. Its announcement on the cover of Popular Electronics in January inspires a generation of hobbyist entrepreneurs. One of them drops out of Harvard to start a company he calls Micro-Soft. Other microcomputer companies come on line, notably IMSAI. It is IMSAI that puts Digital Research in the black with a $25,000 contract. For the next seven years, Digital Research owns the personal computer operating system market.
Then IBM decides to produce a personal computer.
When Gary Went Flying
The story of when IBM came calling and Gary went flying has been told many times, and with varying details. It is the stuff of personal computer industry myth and legend.
But the key facts of the story are not in doubt. They are that Bill Gates gets his guys in the room with IBM and seals the deal—even though he doesn’t have a product to sell them—while Gary does not pursue IBM and fails to get the deal and as a result loses the personal computer operating system market that he created.
That isn’t the end of the story, but it marks the point beyond which Gary is done with the business. He takes time off, writes a version of the language LOGO as a present for his son. Of course he does such a good job that it’s a marketable product, and Digital Research sells it. But Gary doesn’t need that. He phases out of the business. It just isn’t fun any more.
He goes back to inventing. He does groundbreaking work on CD-ROM software and on interfacing computers and videodiscs. A company, KnowledgeSet, is spun off from that work. It enables the CD-ROM-based Grolier’s Encyclopedia, which shows everyone how to do CD-ROM content right.
Alan Cooper once pointed out a good example of how Gary looked at his work. In MS-DOS, to copy a file from drive A to drive B, you would write: copy A: B:. In CP/M, to do the same thing, you would write: pip B: A:.
When users complained that the pip command was confusing, Gary was dismissive. Gary thought that any halfway intelligent person could learn that PIP was the command you used to copy, and that you copied (or “pipped”) from right to left, not left to right. No matter that this was arbitrary, confusing, and counterintuitive, or that customers complained about it.
Bill Gates listened to his customers. “That difference in attitude,” Cooper said, “is worth twenty million dollars.”
So Gary wasn’t Bill Gates. But, as we said in the second edition of Fire in the Valley, he left a remarkable legacy. He created the first microprocessor disk operating system, which eventually sold a quarter million copies. He defined the first programming language and wrote the first compiler specifically for microprocessors. He developed the file system and data structures for the first consumer CD-ROM. He created the first computer interface for videodiscs that allowed automatic nonlinear playback, presaging today's interactive multimedia. He created the first successful open-system architecture by segregating system-specific hardware interfaces in a set of BIOS routines, an innovation that made the whole third-party software industry possible.
Finally, he personified an approach to work that many of us find admirable: The belief that science and technology advance through the open sharing of discoveries, and that it’s more important to create the next invention than to protect the last one.