Technology

A history of the Amiga, part 11: Between an Escom and a Gateway

(Source: arstechnica.com)

Limbo

Commodore International declared itself insolvent on April 29, 1994, under Chapter 7 of US bankruptcy law. Ordinarily, this would have been followed immediately by an auction of all the company’s assets. However, Commodore’s Byzantine organizational structure—designed to serve as a tax shelter for financier Irving Gould— made this process far more lengthy and complicated than it should have been.

During this time, Commodore UK, Ltd. continued to operate. It had been the strongest of all the subsidiary companies, and always had a positive cash flow. As the other subsidiaries went under, Commodore UK purchased all of their remaining inventory and continued to sell Amigas to British customers.

The head of Commodore UK, David Pleasance, hatched a plan to purchase the mother company’s assets at auction. His idea was to raise enough money not only to buy Commodore International, but to fund the new company as a going concern, including the continuation of research and development projects. The business plan was to continue to sell Amiga 1200 and 4000 computers and CD32 consoles, while slowly transitioning to next-generation hardware based on Dave Haynie’s Hombre RISC architecture.

Pleasance raised about $15 million from local investors, which was enough to win the auction. For the rest of the money, he arranged a deal with a curious partner: New Star Electronics, based out of China.

New Star Electronic's Sega Genesis clome.

New Star got their start by selling unlicensed clones of the Super Nintendo and Sega Genesis consoles—they not only played the same games, but they looked identical to their legitimate counterparts. Under pressure from the Chinese government, they had decided to change their business model, and they were looking for game companies willing to license their devices. David Pleasance arranged a deal worth $25 million for New Star to produce and sell Amigas for the Asian market.

Other companies were also vying for Commodore International’s assets. Dell Computer put in a $15 million bid—but it was late. The bankruptcy judge refused to wait for Dell to do their due diligence. Six months had already passed, and the only remaining contender in the race was a European PC manufacturer named Escom, who put in a bid of $14 million. Originally Escom wanted to bid only on the Commodore brand, but raised their stake to all the Commodore assets after other bidders complained.

With only thirty-six hours to go before the end of the auction, New Star told Commodore UK that it was backing out of their deal. Without enough money to guarantee the continuation of the business, David Pleasance withdrew his bid. Escom was now the new owner of Commodore and the Amiga.

Enter Escom

ESCOM Computers' logo, circa 1995

Escom was founded in 1991 in Germany by Manfred Schmitt. Early on the company sold Commodore 64 computers, but it quickly branched out into its own line of IBM PC clones. From 1992 to 1994 their sales rose from 180,000 to 410,000 units, making them the biggest PC distributor in Germany.

Schmitt’s company had been saved in its infancy when he was able to secure a large purchase of Commodore 64s for the Christmas 1991 season. He did so by making a personal phone call to Commodore’s global logistics director, a man named Petro Tyschtschenko. During the bankruptcy negotiations, Schmitt called Petro again to get help with the process. After Escom won the auction for Commodore’s assets, Schmitt split them into two subsidiaries: Commodore BV in Holland (which became the holder of the Commodore trademark), and Amiga Technologies, whose name was self-explanatory. Schmitt rewarded Petro with a new job as the director of Amiga Technologies.

At first, this wasn’t much of a job. Escom had only been interested in the Commodore name, and putting all the Amiga assets together in a single division seemed like a prelude to selling them off. However, Escom was soon buried under a deluge of mail and phone calls from Amiga owners and fans, pleading with the company to keep their beloved computer alive.

Petro was keen on keeping the flame going, as he had been a fan of the Amiga back in his Commodore days. However, the task wasn’t going to be easy. In the wake of Commodore’s demise, factories around the world had been sold, and parts suppliers had moved on. Just restarting production of Amiga 1200s and Amiga 4000s required building a brand new factory in Scotland, which wasn’t able to deliver machines until October 1995. By now it was a year and a half since Commodore had declared bankruptcy.

Still, Amiga Technologies soldiered on. The subsidiary company revealed a new Amiga logo, and released a hardware and software bundle called the Amiga Magic Pack. This was an Amiga 1200 sold for £400 in a single box along with two games, the Deluxe Paint AGA painting software, the Wordsworth 2 word processor, and Print Manager. A £500 bundle was released that added a hard drive and more productivity software. These Magic Packs were reasonably good deals, albeit saddled with aging computer hardware. Sales were poor in England, but a little better in Germany, where Escom had greater marketing presence.

In Asia, Amiga Technologies licensed their hardware not to New Star as everybody had expected, but to another Chinese company called Regent Electronics Corporation. With the help of ex-Commodore engineers, Regent designed and built a set-top box called the Wonder TV A6000, which was eventually released to a small set of Shanghai residents in 1997.

Promotional image for the Amiga WonderTV A6000

Petro was insistent that Amiga Technologies should operate as a viable business without asking for additional financial help from Escom. However, this meant that there was little money available for research and development. He did eventually find the funds to develop a prototype of the first new Amiga hardware since Commodore’s demise—it was codenamed the Walker.

The Walker was a modest iteration on the Amiga 1200. It bumped the CPU from a 68020 to a 68030 running at 40 Mhz, kept the 1200’s AGA graphics chipset, and added an integrated CD-ROM drive. The whole kit and caboodle was stuffed into a small form-factor tower case that looked either like Doctor Who’s robot companion K9 or Darth Vader’s helmet depending on which side of the pond you lived on.

Publicity photo for the ESCOM Amiga Walker prototype.

The Walker would have been a fantastic and competitive machine if it had been released in 1991. It was going to be woefully outdated in 1996. In the end it didn’t matter, as the product never made it to production. Escom had inherited what would become known as the “Commodore curse”, and was itself careening towards bankruptcy.

Escom had expanded too quickly in the past two years. The company had opened retail stores all over Europe, hoping for a huge increase in PC sales. Sales of PCs were in fact growing rapidly at the time, but the gains were mostly taken by small “white box” assemblers with lower overhead and lower prices.

Manfred Schmitt’s company had also backed the wrong horse at exactly the wrong time. On the eve of Windows 95’s release, Escom signed a deal with IBM to sell computers bundled with the ailing OS/2 operating system.

Escom declared bankruptcy in June 1996. As with Commodore UK, one of their subsidiaries (Escom Netherlands) kept operating and tried to buy out the mother company. They went bankrupt themselves a year later. Escom became just another PC company on a huge rubbish pile of failed PC companies. Today, no one even remembers it—except, of course, for Amiga fans.

Limbo again

The beleaguered Amiga yet again found itself between two companies. In the Escom bankruptcy sale, no fewer than eleven firms made bids on the assets. Quikpak, an American manufacturer and reseller, got out to an early lead before running into money issues. VIScorp, a startup company that had hired some ex-Commodore engineers, wanted to get the Amiga technology to put into set-top boxes. They hired Carl Sassenrath, father of AmigaOS, who said: “I don’t plan on killing the Amiga. In fact, if they ask me to take over system development, you’ll see one killer Amiga!” Unfortunately their plans came to a screeching halt as they too ran out of cash.

While the dreary saga of bankruptcy negotiations trundled on, a group of former Commodore employees decided to take matters into their own hands. John Smith (previously a sales manager at Commodore UK), Dr. Peter Kittel (a director of software at Commodore) and Dave Haynie (a legendary Commodore engineer) joined forces to found PIOS, a startup that would build the next generation of Amiga hardware that could also run other operating systems.

This was the same time that Apple was transitioning from the 68k series of CPUs to the PowerPC, and the latter was getting a lot of great press in technical circles. Some Amiga accessory companies, like the German-based Phase5, had started shipping PowerPC accelerator boards for classic Amigas. These could speed up specific software routines in applications that supported them, even if the operating system itself was still stuck running on the 68k chip.

Apple had also opened up licensing for MacOS for the first time in its history, allowing legal Macintosh clones to exist. The PIOS ONE was planned as a computer that could run almost anything: either a PowerPC flavor of Linux, a licensed copy of MacOS, a copy of the brand-new operating system BeOS (from a company run by ex-Apple manager Jean-Louise Gassee, who billed his operating system as the second coming of the Amiga and even had vanity license plates that read “AMIGA96”), or an “Amiga-like” PowerPC operating system that PIOS hoped to develop, called pOS.

The possibility of a computer that could run any operating system was a nerd’s wet dream, but the dream died a few years later.  Apple withdrew their licensing program, BeOS transitioned to x86, and pOS failed to appear. PIOS had also hoped to win the Escom auction, but like many others, they were outbid.

The eventual winner was another PC clone company founded by Ted Waitt, the son of four generations of cattlemen. Founded in 1985, it had risen to 5 billion in sales by 1996, selling PCs in cow-painted boxes by mail order, in showrooms, and later online. This was Gateway 2000.

The Gateway Era

Animator Eric Schwartz created this image after the Gateway buyout.

Gateway 2000 had experienced phenomenal growth over the last few years, and was now considered one of the giants in the PC world. As a giant, they came under the radar of the wounded but still mighty IBM. IBM, still angry about the PC clone industry they had unwittingly brought into existence, was starting to push its massive legal weight around. While PC clones were still legal, IBM had a portfolio of patents that it was brandishing as a stick, trying to extract licensing revenues from any cloner company wealthy enough to be worth its time. Larger companies like HP could respond with their own patent portfolio and negotiate a “Mutual Assured Destruction” cross-licensing agreement, but the young Gateway had few if any patents to speak of. It was at this point that Jim Collas, Senior VP of product development for Gateway 2000, was put in charge of patent acquisition.

Jim Collas, director of Amiga Technologies at Gateway.

Collas was a technical manager, with a degree in electrical engineering and computer science from UCLA. When he was still at school, he had run across the Amiga (still in prototype form) while working for a startup game developer, and had been impressed by the technology. After graduating, he had started a PC design company that was acquired by Gateway in 1992.

So when the opportunity arose, he jumped at the chance to acquire the Amiga assets. Working with the resilient Petro Tyschtschenko, he arranged a winning bid of $14 million for Amiga Technologies and all Commodore patents (the Commodore brand was sold separately, ending up with a dutch PC company called Tulip). The deal closed in early 1997.

Colla’s bosses, of course, were only interested in the extensive patent portfolio, which included patents on a type of dropdown menu and the two-button mouse. But he saw a little further.  In an interview with Ars, he said that “the combination of the Amiga brand, the available technology at the time, and the passion of the Amiga industry made me feel like there was a once in a lifetime opportunity for making significant change in the industry and the world.”

By 1998, he was losing confidence in Gateway’s upper management and their vision for the company. He correctly realized that PCs were becoming a commodity item, and profit margins were going to keep shrinking over time. Without any differentiating technology, Gateway would eventually get squeezed out by smaller firms that could sell identical PCs for less money. Cow-painted boxes could only get you so far.

“The combination of the Amiga brand, the available technology at the time, and the passion of the Amiga industry made me feel like there was a once in a lifetime opportunity for making significant change in the industry and the world.”

When the purchase from Escom was concluded, Collas became the head of Gateway’s Amiga Technologies subsidiary. He kept Petro on as a consultant, and quickly assembled a new team of engineers, designers, managers, and marketers, as if he was building a brand new company. He hired former Commodore director of operating systems software Dr. Allan Havemose to head up engineering, Joe Torre (no, not that one) to lead the hardware division, Fleecy Moss to handle developer relations, and Bill McEwen as a sort of marketing and software evangelist (think Guy Kawasaki).

While Collas was assembling his team, Gateway received a torrent of letters from Amiga owners and fans, desperate to know what the company was going to do with the technology. These letters helped Collas convince his bosses that he was on the right track.

The technology world in 1998 was experiencing tumultuous change. The rise of the Internet in everyday life changed the way people looked at computers. Laptops, once a niche market, were rising in popularity, and the first usable generation of WiFi was about to become ratified as a standard.

It was this environment that shaped Jim Collas’ vision. He saw a new line of computers that would span everything from tablets to workstations, all running the same software. He even imagined a standardized method for users to purchase this software digitally.  These new Amiga computers could be workstations, game consoles, set-top-boxes, or Internet terminals, or all of them at the same time.

Bringing this vision to life would require a drastic shift away from the traditional Amiga hardware and software architecture. The first thing it would need was a new operating system kernel: something that was fast and lightweight enough to run on the minimal hardware of a tablet, but could also run graphically-intensive applications. Collas found what he was looking for with QNX’s Neutrino, a real-time operating system that famously could run an entire 32-bit preemptively multitasking OS with a graphical user interface and a web browser, all from a single 1.44 MB 3.5-inch floppy disk.

QNX's famous 1.44MB floppy demo. It really worked!

But the kernel was not enough to deliver the complete vision. Dr. Havemose drew up plans for what he called the Amiga Operating Environment (OE)— a software layer on top of the kernel that would handle the graphical interface and provide a standardized development platform. Programmers would write to the OE layer, using a high-level language like Java, and their applications could run on any Amiga OE compatible hardware. When Dr. Havemose told Collas about his ideas, he thought that “he was either the most brilliant software architect I had ever met, or he was delusional and fooling me.”  He brought in some experienced software architects to look at the plan and they concluded that the plan was indeed brilliant.

The Amiga Operating Environment was another piece of the puzzle.  There was also a plan for an online store, much like today’s Apple App Store or Google Play Store, where consumers could buy new Amiga OE applications directly through the Internet.

The first hardware to be prototyped was called the Amiga MCC, which stood for Multimedia Convergence Computer. It was designed to work either as a set-top box and gaming console connected to the television, or as a standalone computer with a traditional monitor. The MCC tablet would be the second piece of hardware, to be delivered later.

Promotional image of the Amiga MCC with a monitor (left), a prototype computer case (right), and a blurry image of a prototype MCC tablet (bottom right)

None of this new hardware or software would be directly compatible with the classic Amiga architecture. This concerned some Amiga fans, particularly the ones running companies like Phase5 and Haage & Partner that were selling PowerPC accelerator cards to extend the life of existing Amigas. In fact, the announcement of the Amiga MCC and OE shocked these companies so much that they immediately settled their differences (both companies had been developing incompatible standards for PPC accelerator cards) and agreed to work together to continue developing classic Amiga hardware and software.

Collas, for his part, felt that “the Amiga brand was about revolutionary change in computing.”  He saw the new MCC devices as being in the spirit of the original Amiga, rather than a direct descendent.  The hardware was to be cutting edge, even going so far as to use liquid cooling to overclock industry standard graphics cards.

It was at this point that Fleecy Moss was fired by Gateway for unknown reasons. He immediately announced to the world that he was joining forces with Dave Haynie to develop his own idea of the next-generation Amiga platform, called KOSH (for Kommunity Oriented Software and Hardware]. Aside from a number of fanciful blog posts, this never ended up amounting to anything.

As 1998 rolled into 1999 and Gateway (the company dropped the “2000” from their name around this time) continued to work towards their Amiga OE vision, Haage & Partner released AmigaOS 3.5, an official update for classic Amigas. This fixed some bugs from Commodore’s final 3.1 release, and added some bundled software. The traditional Amiga market was dwindling by this time, but there were still enough people using the old hardware to warrant development.

Then came an announcement that caused an uproar in the Amiga community, and was even seen by some as a betrayal. Jim Collas’ team revealed that they were switching the OS kernel for Amiga OE from QNX to Linux.  He met with Linus Torvalds to discuss his plan, and Linus really liked Collas’ ideas.  There were even plans for Linus to attend Amiga shows and meet key Amiga community figures.

In retrospect, this was probably the correct choice. Linux, then and now, supported a much more diverse array of hardware and was already gaining significant momentum in the tech industry. But Amiga fans didn’t see it that way. They saw Linux as bloated, and clunky, and slow, the exact opposite of their beloved Amiga. QNX Neutrino was seen as fast, small, and elegant, and the community had rallied behind it. In fact, QNX itself was surprised and taken aback by the sudden shift in plans. In response, they immediately announced a partnership with Phase5 to develop a QNX Neutrino operating system for Amigas with Phase5 PowerPC accelerator cards, as well as an planned new QNX-based Amiga computer called the AMIRAGE K2. The use of the word “Rage” and all capital letters was seen as a deliberate statement of defiance against Gateway itself, although Phase 5 jokingly said it was pronounced “A-MIRAGE”.  Unbeknownst to the Amiga community and fans, however, there were much bigger and more troubling conflicts on the horizon.

Jim Collas’ ambitious plans for Amiga Technologies was running, as he had predicted, head on into interference from Gateway’s conservative upper management. At the center of the controversy was Gateway’s new CEO, Jeffrey Weitzen.

Weitzen had come from AT&T and was skeptical of Collas’ Amiga vision. Collas felt that he had to fight his own CEO for control of the subsidiary, as he could potentially torpedo the plan. Collas worked out a scheme with his own CFO where Amiga Technologies would purchase enough shares from its parent company to get 51% ownership and operational autonomy. He hoped that Gateway would not want to shut down the division at that point because they would end up taking a massive write-down for the quarter. Collas was on good terms with Gateway’s founder, Ted Waitt, and thought that this would provide him with at least some political protection.

Weitzen, however, played the political game better. He got that CFO to reveal these plans to him, then waited until Ted Waitt was on vacation and called Jim Collas into his office,

Weitzen told him to sit down. He said that he was planning to sell the Amiga division, and that Collas’ plans would not be realized.  Collas knew that he had lost, and he sold his Gateway shares and left the company. Ironically, a few quarters later the stock crashed from $35 to under $10, and Weitzen was forced out as CEO. But by then it was too late. Collas’ grand vision, the MCC, the Amiga OE, the computers and tablets, the online application store, all of this vanished into the ether.

Amiga flies free

Gateway carried through with its plans to sell the Amiga division, but this time there weren’t as many takers. Mostly this was due to the fact that Gateway still wanted to retain all of Commodore’s old patents, which was the primary reason they had been interested in buying the Amiga in the first place.

The diehard fans in the Amiga community were hanging on, but hope was dwindling. Outside of the PowerPC acceleration boards, there had been no new Amiga computer hardware since the Amiga 1200 and 4000 were released back in 1992. Amiga software companies were struggling as well. Amiga magazines were getting thinner, and some were going out of business completely. Newtek, the maker of the Video Toaster, had already moved its hardware and software to the Windows platform.

An Amiga 4000 equipped with a dedicated video card and an accelerator board featuring both a 68060 main CPU and 604e PowerPC accelerator chip, running native Amiga software, was still competitive speed-wise with contemporary Windows 98 and Macintosh computers. However, this wasn’t a computer you could easily buy— it had to be crafted together from old and new parts, and there was no guarantee that the operating system would ever be updated to be fully PowerPC native. It was also expensive, as the volumes for these add-on boards were low.

At the same time, the idea of a brand new, revolutionary Amiga architecture as promised by Gateway was now just as far away as it had ever been. Petro Tyschtschenko had offered to buy the division himself, but he could not find enough money to do so. He was asked by Gateway to find a new buyer. Who would pick up the torch and carry on with this dream?

The people who decided to do that were ex-Gateway employees who had been bitten by the Amiga bug during their brief time with the company. Bill McEwen (the technology evangelist) and Fleecy Moss (the former developer relations manager) joined forces and created a startup they called Amino Development. Together they raised $5 million from venture capitalists.

How did a couple of mid-level Gateway employees manage to get this kind of money? It wasn’t difficult. It was late 1999, and the Internet “Dotcom” bubble was heating up. Investors were happy to throw money at anything remotely computer or Internet-related, and Amino Development promised to be both of those things.

Amino Development took their $5 million and purchased all the non-patent portions of Amiga Technologies from Gateway. They then immediately changed the name of their company to Amiga Technologies, Inc.

The deal was closed at the final hour on the last day of December. The old millennium was coming to an end. There was a feeling in the air, something that Federal Reserve chairman Allan Greenspan had called “irrational exuberance”. It was a feeling that anything was possible.

For the first time since a group of dreamers led by Jay Miner had banded together in 1982 to create something amazing, Amiga was an independent company again. The little computer that could had survived a rocky stewardship under Commodore, Escom, and then Gateway, but now it was on its own.

The future was wide open. Surely the next millennium would usher in something even more amazing. With Internet and computer companies’ stock prices soaring, it seemed like the sky was the limit, even for the Amiga.

What could go wrong?

Listing image by Eric Schwartz

More Info: arstechnica.com

Advertisements