What If?

by Eric Schwartz
from the AmiTech-Dayton Gazette, May 2008

Lately I've been looking at the Amiga from a more historical perspective. There's always a lot of "what ifs" in the Amiga's timeline that make you wonder how things might have gone for the system if a circumstance or a decision was changed here or there. Early examples include the design of the Amiga itself, and many of the hardware and software decisions that gave the Amiga an advantage early-on, to become a liability later. The Amiga's own custom chipset of graphic (and more) co-processors made it an undisputed king of graphic, sound, video, and multimedia ability between the mid-eighties and early nineties. However, with the Amiga system tied so tightly to its custom hardware, development of the system beyond its initial specs was not as quick as it could have been, and systems with a more off-the-shelf and modular design had an easier time moving forward quickly. It took some revisions of the operating system and work by several clever hardware and software developers to make the Amiga graphics re-targetable to newer and faster video cards, though the software has to be able to work through them.

Speaking of the Amiga operating system, it has the distinction of being a first to offer a true preemptive multitasking OS on a consumer machine. One issue with this was the system had to be designed to multitask on CPUs without built-in memory management hardware found on multitasking UNIX systems. Not surprisingly, it took some time and revisions to get working stably, not to mention for the software developers to figure out the finer points of multitasking programming. Unfortunately, once designed without the requirement for an MMU, it's not simple to shoehorn memory management back into the multitasking scheme, so it was always possible for a poorly debugged program or a bad set of circumstances to take down the entire system with it. Sadly this, and genuinely buggy early 1.X OS versions, gave the Amiga a reputation among outsiders as an unstable machine, albeit a largely undeserved one, as most other operating systems had the same or worse stability issues. Some of these mild problems carryover to the most recent OS4 and MorphOS even still, mainly when they try to run legacy Amiga software.

Of course we all know the Amiga's worst enemies were the companies who were in charge of it. Commodore, as the most successful and long-time Amiga custodian, repeatedly frustrated the Amiga public, sometimes spending tons of money on frivolous fruitless projects, other times mercilessly penny pinching and cutting down what the engineers wanted to do with the system. One of Commodore's bigger failings was an apparent problem identifying existing successes while pursuing other markets and avenues. They always wanted to get in on the mass markets, searching for the elusive "cash cow" like their Commodore 64 was. They had moderate successes in the low-end machines for the masses, in the form of the A500 and A 1200, though they also jumped into the less successful "multimedia appliance" field with the CDTV, and the game console market with the CD-32. They virtually ignored a niche market they largely created in desktop video, even when Newtek's Video Toaster was doing a lot to sell Amiga 2000s and 4000s. I can't help but think if they latched onto the desktop video market, and put more development dollars into addressing their needs and interests, they might have avoided bankruptcy in 1994. I don't know all the facts though.

It's said that competition is good for a market and its consumers. When a market becomes small, and the competition becomes more and more desperate, the "goodness" becomes a lot more debatable. Most of these problems came out of the dwindling days of the post-Commodore, post-Escom era, when some developers (usually German ones) fought tooth and nail over scraps. The first big shot was when software developer Haage & Partner put together their own set of libraries for software development support on PowerPC accelerator cards, backed up by their own C compiler. They proceeded to expound on how their PPC libraries were better than those made by the people who made the hardware itself, and the opposing party countered. All that was really accomplished was most big software developers that were still around avoided writing PPC software until the issue was resolved, and the lack of PPC software greater than simple utilities and MP3 or video players gave consumers less incentive to buy the hardware. The issue was sort-of resolved when Haage & Partner became the developer on the 3.5 and 3.9 versions of the Amiga OS, making their PPC system the default. Unfortunately by then hardly anyone was left to do the big software to show off what the processor could do. There's been a long history of software companies and/or programmers being fundamentally unhelpful. While there's nothing inherently wrong with protecting your own work, rights, business, and getting what you believe is due you, once again when applied to the tiny Amiga community it often comes off as more petty and hurtful than anything. Haage & Partner didn't release any of the source for Amiga 0S 3.5 and 3.9, forcing Hyperion to do a lot of reinventing the wheel when they took over for production of 0S4. Disputes between programmers derailed development on Amithlon, at its time an ideal solution for running Amiga operating system and software on PC hardware. Disputes between other programmers stalled the development of Morph0S, slowing it to a virtual snail's pace of development. Amiga Inc. gets sue-happy over anyone's use of Amiga trademarks, real or perceived, as they scrounge to find money to fund whatever dreams of becoming a "real tech company" they have this week, to the point of legal-fighting with the same company they hired previously to produce OS4. There's a history of programmers of well-known or vital software leaving either to pursue more fertile ground or because they feel slighted by other individuals or the community as a whole. It's a sad thing -- when the community and its software become so tightly wrapped together, there's a sense that the programmer becomes the property of that community, and there's no reason or excuse good enough to allow them to leave or stop producing work for them, even if there is no reward for it.

I suppose we'll all just have to revitalize the Amiga and/or the Amiga-like systems to a strong position, so the petty stuff doesn't have such a strong impact. Good luck with all that then.