The excellent series of articles on the history of my all-time favorite computer continues with A history of the Amiga, part 4: Enter Commodore:
The CAOS debacle
Originally, that third layer was known as CAOS, which stood for the Commodore Amiga Operating System. Exec programmer Carl Sassenrath wrote up the design spec for CAOS, which had all sorts of neat features such as an advanced file system and resource tracking. The latter was a method of keeping track of such things as file control blocks, I/O blocks, message ports, libraries, memory usage, shared data, and overlays, and freeing them up if a program quit unexpectedly. As the Amiga software engineers were already behind schedule, they had contracted out parts of CAOS development to a third party. Still, as is often the case in software, the development hit some unforeseen roadblocks.
According to Commodore engineer Andy Finkel, the management team “decided that it wouldn’t be possible to complete [CAOS] and still launch the Amiga on time, especially since the software guys had already given up weekends at home. And going home. And sleeping.”
Lack of time wasn’t the only problem. The third-party development house learned that Amiga, Inc., had been bought out by Commodore, and they suddenly demanded significantly more money than had originally been agreed upon. “Commodore tried to negotiate with them in good faith, but the whole thing fell apart in the end,” recalled RJ Mical, who was upset by the whole event. “It was a jerk-butt thing that they did there.”
TripOS to the rescue
When the CAOS deal fell apart, the Amiga team suddenly needed a replacement operating system. Relief came in the form of TripOS, written by Dr. Tim King at the University of Cambridge in the 1970s and 80s, and later ported to the PDP-11. Dr. King formed a small company called MetaComCo to quickly rewrite TripOS for the Amiga, where it became known as AmigaDOS.
AmigaDOS handled many of the same tasks as CAOS, but it was an inferior replacement. “Their code was university-quality code,” said Mical, “where optimized performance was not important, but where theoretical purity was important.” The operating system also lacked resource tracking, which hurt the overall stability of the system. This oversight had repercussions that remain to this day: the very latest PowerPC-compiled version of AmigaOS will still sometimes fail to free up all resources when a program crashes.
Reading through the history now it’s seems almost as if the Amiga was doomed from the start and it’s somewhat amazing it survived as long as it did. In addition to the CAOS debacle mentioned above the article goes on to discuss how Commodore had originally wanted the Amiga 1000 to only have 256K of RAM whereas Jay Miner knew it really needed a minimum of 512K and he ends up putting his job on the line to get a compromise that shipped the machine with 256K and an expansion port for an additional 256K on the front.
Even that wasn’t enough. One of the most popular add-ons for the Amiga 1000 was the Insider Board, which was a 1MB RAM expansion that you had to void your warranty to install (I have one in my A1000 to this day). It involved opening the case, removing the 68000 processor, plugging it into the RAM expansion board (which was huge), plugging the board into the 68000 socket, and, if I recall correctly, connecting an alligator clip or two to various spots on the motherboard. I’d have to open mine up to be sure on that last bit, I may be confusing it with the Buscard Interface (IEEE) I once installed inside one of my Commodore 64s.
The rest of the article goes on to describe the massive launch party Commodore held for the Amiga 1000. They went all out not only demonstrating the machine’s strengths, but bringing in celebrities such as Andy Warhol who painted Debbie Harry on stage using the Amiga 1000. If you were a Commodore fan then it was a very exciting time. At least it was at the beginning. Part 5 promises to delve into the early blunders by Commodore that would turn their new dream made reality into a nightmare. Good stuff.