How the Amiga could have lived to age 30 and beyond

It was 30 years ago this week that Commodore released its landmark, long-time-coming Amiga 1000 computer–the first 1990s computer in a field full of 1970s retreads.

Yes, it was a 1990s computer in 1985. It had color and sound built in, not as expensive, clunky, hard-to-configure add-ons. It could address up to 8 megabytes of memory, though it ran admirably on a mere 512 kilobytes. Most importantly, it had fully pre-emptive multitasking, something that previously only existed in commercial workstations that cost five figures.

It was so revolutionary that even NBC is acknowledging the anniversary.

Being a decade or so ahead of its time was only the beginning of its problems, unfortunately.

Commodore in 1985 was a doomed company. Although many people dismiss it as the maker of cheap computers mostly good for playing games, it was a sophisticated operation that designed and produced its own chips, and even sold its designs to competitors. Apple and Atari 8-bit computers and game machines sported multiple chips designed by Commodore’s subsidiary MOS Technology.

But Commodore had a problem. It only had one fabrication plant, or fab, and unlike modern chip producers, Commodore didn’t do much to modernize it over the years. MOS Technology’s process was revolutionary in 1977, which allowed it to severely undercut everyone else’s prices and sell an 8-bit CPU for 75-90 percent less than Intel or Motorola. But the rest of the industry caught up and learned new tricks. Meanwhile, MOS got by on 1970s technology well into the 1980s.

So, in the early to mid 1980s Commodore had to anticipate demand ahead of time, produce the chips before the computer plants needed them, then produce computers. Commodore wasn’t very good at doing that. In 1982-83, they underestimated the demand for disk drives, and in 1985, they repeated history. They sold a few million Commodore 64s that year, but couldn’t produce Commodore 128s and Amiga 1000s quickly enough.

It’s easy for me to say now, since I’ve been able to watch Intel for a couple of decades, but Commodore should have built two more fabs in the 1980s and then rotated them, producing in two fabs while modernizing the third. This would have allowed them to produce enough chips to meet their own needs, and would have reduced the need to license its designs to companies like Rockwell and Synertek to meet the demand from Apple and Atari and the countless number of smaller companies who used its chips.

Instead, Commodore paid outlandish salaries to Irving Gould and henchmen like Mehdi Ali rather than reinvesting in the company. No one will ever know how many people wanted to buy a 128 or Amiga but couldn’t, and bought an Apple or Atari instead of waiting months for Commodore to get through its backorders.

The other problem was that the mainstream media never took Commodore seriously. In the 1980s, mainstream media coverage of computing was almost all IBM and Apple, and later, the me-too IBM clones. This was unfair; according to Goldman Sachs estimates, Commodore had a staggering 38% market share in 1985. Apple had 15%. IBM and the legions of clones had 49% of the market between them. Commodore had more of the market than any single company.

No matter. The mainstream media thought the Amiga was just an overpriced game machine.

It was a great machine for playing games. It also was the first computer that could do desktop video. Multitasking made it a great machine for anything else you wanted to do too.

Lack of software probably also hurt. As much as it pains me to say it, it would have helped if Microsoft had ported its productivity software and languages to the Amiga, but Microsoft had an old grudge with Commodore. Nobody outside of the Commodore enthusiast community remembers this, but Jack Tramiel convinced Bill Gates to license Microsoft Basic to Commodore for a flat fee, so when Commodore sold 30 million C-64s sporting Microsoft Basic in ROM, Microsoft didn’t make a dime off them. By 1985, Tramiel was exiled to Atari, but Microsoft still didn’t want to have anything to do with Commodore or the Amiga. They produced a half-hearted implementation of Basic for it, but were late in delivering it.

Commodore turned to Metacomco, the same company that helped turn the obscure academic operating system Tripos into Amiga OS. Metacomco Basic was awkward and buggy, which didn’t improve the user experience.

Thanks to all of this, the Amiga took a couple of years to hit its stride. Commodore sold 3 million of them between 1985 and 1994. If they’d figured out their supply chain problems they could have exceeded even the 64’s staggering success. But the Commodore of 1985-1994 didn’t do very many things right. While the Amiga always sold for hundreds of dollars less than a comparable Macintosh, it couldn’t compete in price with the inexpensive 386sx-based clones that were common by 1991. Those 386sx clones weren’t as capable as the Amiga but they were cheap and plentiful. For most people, they were good enough.

Commodore revised its chip designs to compete, but the designs were always late. And by 1992, Commodore couldn’t produce the chips anymore. They had to outsource production, which increased overhead. As a result the third-generation Amigas were even less price competitive. The ill-fated fourth-generation, which was never produced, would have been even less so.

The rotting carcass of Commodore finally went out of business in April 1994. About 18 months later, Microsoft released Windows 95, which finally brought pre-emptive multitasking to mainstream PCs. It also had Plug and Play. Plug and Play mostly worked, though not as well as it did in the Amiga 1000’s Autoconfig. The Amiga appeared ten years and a month earlier. Macs didn’t get full pre-emptive multitasking until 1999, with the first release of OS X.

The other knock on the Amiga 1000 was that its operating system wasn’t all that stable. Then again, neither was the first release of Windows 95 or OS X. By 1987 it wasn’t any less stable than what Microsoft and Apple were shipping. When I finally bought a Windows 3.1 PC in 1994, it wasn’t any more stable than my Amiga had been.

So, Happy Birthday, Amiga. The world would have been a better place if you’d thrived into the age of 30 and beyond, rather than fizzling out at nine. I figured out how Amiga could have survived; it’s too bad Commodore didn’t.

%d bloggers like this:
WordPress Appliance - Powered by TurnKey Linux