Gary Went Flying?

There is an old adage that says that sometimes a story sounds so crazy, it must be true. In this case, the story is so crazy; it must be made up by a self-serving billionaire who is attempting to rewrite history. It is hard for me to write this without a lot of negative emotion because it is this stupid urban legend that left Gary so embittered in the later years of his life.

Let’s back up a minute and give the back story. Gary wrote the first microcomputer operating system sometime around 1974. He named it CP/M. CP/M ran on computers using the Intel 8080 and Zilog Z80 microprocessors. Gary’s company, Digital Research (a really cool name in my opinion), was the dominant operating system developer.

In early 1980, IBM made the clandestine decision to make their own microcomputer, which they eventually named the IBM Personal Computer (shortened to the IBM PC in common vernacular). IBM planned on using the Intel 8088 processor. The 8088 was much more powerful than the 8080, but it wasn’t compatible. An operating system that utilizes the 8080 would have to be rewritten, at least to a certain extent, to support the 8088 and its more powerful, yet somewhat quirky, memory segmentation model.

As a side note, the 8088 was really an 8086 processor. The 8086 was the 16-bit successor to the 8080. The 8088 was also a 16-bit processor with the exception that it communicated with everything external to the processor using 8-bits at a time. While that may seem like a drawback, and it certainly was, it was actually a very good decision at the time. Most of the peripherals, memory and important devices communicated in 8-bits. Using 16-bits would force you to use peripherals that may not even exist. Where they did exist, they were much more expensive, and much rarer, than their 8-bit counterparts.

IBM connected up with a new startup called Microsoft to write their BASIC interpreter. BASIC was a beginner’s programming language that most computers distributed with their computers back in the day. Microsoft was a big name in the BASIC market space. IBM also thought they could do a strategic one-two punch with Microsoft and get an operating system from them too. IBM thought that Microsoft owned CP/M. The reason is that Microsoft made a CP/M add-on board that you could put in an Apple II computer. It was a pretty nifty device, but Microsoft sublicensed CP/M from Digital Research. They didn’t own it and didn’t have the right to sell it.

That led Bill Gates to refer IBM to Gary Kildall/Digital Research. And, this is where somewhat verifiable facts become the stuff of urban legend. IBM had a very restrictive and one-sided non-disclosure agreement. It was described as “IBM is allowed to take any and all information and do whatever they want with it, and the other party must keep everything a secret; you cannot even disclose that you met with IBM.”

Gates supposedly called Gary and said he is sending someone important to meet with him the next day. According the NDA, Gates could not disclose who it was and so he didn’t. Gary didn’t take Gates seriously and took his personal airplane out and did some kind of stunt flying for hours while IBM suits sat at Digital Research’s headquarters. Angry that Gary blew them off in such a casual fashion, IBM went back to Gates who agreed to get them an operating system and seized the opportunity that Digital Research rejected.

And that, my friends, is the biggest whopper in computer history since the ENIAC was created. However, it has been accepted as general history. I believed this garbage because even people like Peter Norton wrote that in their books. It is heartbreaking because Peter Norton usually bases his arguments in facts, not wild speculation. It is what it is, I suppose.

Gary explained what really happened many, many times, but it was easier for a cynical world to accept myth rather than the truth. Gary had many obligations around the country at the time. For him to drop everything he was doing at the time because of some obscure phone call from Bill Gates was not exactly a smart decision. Gary was schedule to deliver software to one of his top customers that morning; hardly a small affair. He was flying to deliver the software as agreed.

Of course, IBM did show up and Gary was not there. This was hardly unusual. Gary left the business side of things to his wife and his general counsel. Gary did show up later in the day after finishing his commitments. There was a huge disagreement on the NDA (which I believe was justified) and the attorney wanted to get Gary’s opinion before proceeding further.

According to Gary, he was okay with the NDA, and signed it as soon as he got to the office. Gary said that they met throughout the rest of the day. Gary and his wife were also going on a vacation for his wife’s birthday. Gary changed his flight they were going to take so they could be on the same flight as the IBM suits that were going back to Florida. A handshake deal was made and everything was good to go, or at least it was as far as Gary was concerned.

Something happened in that meeting that IBM didn’t like. It was most likely due to the fact that IBM wanted to pay a flat fee for the rights to use CP/M. Gary couldn’t do that because he signed a favored nation’s contract with several of his founding customers. This meant that Digital Research could not license CP/M for less money to someone else without giving the exact same terms to them. Forcing Digital Research to offer a flat-rate, non-royalty price to IBM could essentially shut down Digital Research. Gary said they must pay a per-copy royalty.

IBM insisted there was no handshake deal. They went back to Gates, who promised an operating system. The operating system story will be covered at another time. IBM has denied Gary’s version of the story. It would be in IBM’s favor to deny this as it doesn’t exactly paint them in the most positive light. It makes them look cheap and petty.

Of course, this relationship would haunt IBM in the years to come. One need only look at the OS/2 fiasco and the ultimate breakup between IBM and Microsoft to know that IBM made the wrong decision and it would’ve been in their best interest to throw their full support behind Gary Kildall. And the world would be better off for it.