Pac-Man & CP/M?

The debate continues to this day. Would Digital Research have the right to sue IBM and Microsoft? Many say no, based off the fact that there was no proof of actual copying of source code. Is that a correct assumption? I don’t think that it is, and I’ll explain why. I will also compare other similar lawsuits and their outcomes. One of the lawsuits comes from an unlikely video game character, Pac-Man.

Pac-Man was one of the most well known games of the classic video game era of the early eighties. That time period was also the height of the cartridge-based home video game consoles. Many different companies competed to get home console rights to publish Pac-Man. Atari won the battle, agreeing to pay millions for the rights plus a per-cartridge royalty.

Before Atari released Pac-Man, a company called Magnavox, who created the Mattel Odyssey 2, released a game called K.C. Munchkin. The game was similar to Pac-Man, but obviously didn’t contain the same source code. However, a U.S. district court ruled that the defendant (Magnavox), “deliberately copied the Plaintiff’s work.” Now, this is important. It was ruled as a copy because it was similar in form and function.

Let’s go back to the issue of MS-DOS and CP/M. The argument consistently made by Tim Patterson, ex-IBM employees and others is that Patterson didn’t copy the source code. Is that the sole test for intellectual property infringement? The Atari/Magnavox lawsuit certainly proved otherwise. If we throw out source code plagiarism, what else is there to go on?

MS-DOS copied the first 26 API calls from CP/M. Although, as discussed prior, some of the calls were implemented incorrectly due to Patterson’s incomplete understanding. Can the case be made for similar form and function? Yes. Given the standard of evidence in the Atari/Magnavox lawsuit, it was a sure bet that Digital Research would have won.

Or, would it? After the original IBM PC came out, a company called Compaq copied the IBM BIOS. Compaq did something a little different than Tim Patterson. They did something called “clean room design” which provides a fully documented procedure and a “wall” between the programmers writing the code and the people documenting the system calls. It is a process of developing a compatible system while proving that the people developing the system had absolutely no internal knowledge of the system they are cloning. So, the case for intellectual property violation by Microsoft/IBM still holds.

There are some obvious gray areas in this debate. Linux has cloned the function of Unix, but doesn’t appear to have used clean room design. There is a Windows clone called ReactOS that has been in development for quite a few years, and it is doubtful as to whether or not it will ever be viable for general use. However, it is using a clean room design. Needless to say, in the matter of law and the courts, nothing is ever certain.

Why did Digital Research and Gary Kildall ultimately decide not to sue? It seems to be a combination of several factors. Gerry Davis, Digital Research’s corporate counsel, thought that the law was not necessarily clear on software copyright. Unfortunately, Gerry was general counsel, and was probably not an expert in intellectual property. It would’ve been a good idea to consult with an expert, but didn’t happen, in all likeliness. Another factor is that Gary didn’t really believe that the IBM PC was going to set to world on fire. He assumed it would be lackluster in comparison with some of the other microcomputers at the time which ran the traditional CP/M.

What Gary probably did not foresee was the success of the PC clones and the fact that Microsoft would make a huge financial windfall from this licensing. While Gary made legal threats to IBM, Gary settled for the ability to license CP/M for use on the IBM PC. IBM would sell the PC without an operating system and make the customer purchase the operating system separately.

It has been said that Gary Kildall set the price for CP/M-86 for the PC. Others, like Tom Rolander, say that no one had an idea what the price would be set at. Regardless of these facts, IBM sold CP/M-86 for $240, while selling PC-DOS for $40. Since most users of the PC wouldn’t know the difference, PC-DOS was an easy sale. And, unfortunately, this was the beginning of the end of the operating system dominance of CP/M.

Gary should have sued.