Ever since 1982, the general shape of the PC market has been relatively simple. The IBM PC defined a hardware standard; MS-DOS became the software standard, and applications battled for supremacy in the huge market that adhered to these standards. Although it was far from clear at the time, other architectures such as the Lisa, Unix workstations, Amigas, the Atari ST, and even the Macintosh were largely sideshows compared to the 90% of the market represented by the PC. The assumption that the PC would continue to develop along the obvious lines of faster processors, larger discs, more memory, better networks, and higher resolution graphics has been correct over Autodesk's entire history. Autodesk has profited enormously from this progress.
Recently, however, the future of the PC market has become very confusing to try to project. Since Autodesk, as an application vendor, is forced to choose which architectures will receive the attention of our limited technical, marketing, and promotional resources, uncertainty about the direction of the market complicates our job immensely. Worse, the same confusion that vexes us causes even more bafflement among our dealers and their customers. This lack of direction is bad for everybody--at the minimum it causes customers to defer purchase decisions until things sort themselves out. A customer forced to choose may end up with an incompatible system and waste his money and time on something that ends up being bypassed by the market.
What I'm basically talking about here is OS/2. I guess every decade or so IBM has to show the new kids in the industry who think they know how to screw things up what a débâcle of truly Wagnerian scope looks like. With the able assistance of Microsoft, uncertainty and confusion have been unleashed in the PC market to such an extent that people cannot even figure out which vendor is responsible for what product, no less when they might be able to buy a solution to the problems of existing PC users. Consider this: today there are more than forty million computers in place that share a common hardware and software architecture. Yet at this moment there is no clearly-defined path for the future migration of these machines which represent a collective investment of many billions of dollars.
What's a user to do? Struggle along with MS-DOS? Move to Windows and buy a whole new set of applications compatible with it? Choose products built with DOS extenders, then fight the mutual incompatibilities among them? Wait for OS/2...but which OS/2? Certainly not the awful one you can buy today, but the one coming from IBM or the one coming from Microsoft? But wait, Microsoft says they're developing a ``portable'' OS/2 that will run on non-Intel hardware. So are we going to end up scrapping our '386 machines for MIPS or Sparc chips? Crazy--but what's this news about Compaq narrowing their choice for a CPU vendor for their next generation to MIPS and Sun, Intel being notably absent? Chuck the whole thing and get a Macintosh? But a SPARCstation is less than $5,000 and a whole lot faster. And then there's the other side of IBM participating in OSF and promising true cross-platform application portability and yet another wonderful standard environment.
It's entirely possible that all of this disarray in the market may result in nothing more serious than an interlude in which the users continue to buy vanilla '386 and '486 DOS machines while they wait for the dust to settle and a clear direction to emerge. If that's the case then Autodesk need only be careful to avoid squandering too much of our development resources on each of the contending architectures until a winner becomes apparent. But if rampant confusion seems to be hurting the market for our products and further endangering our dealers, Autodesk must be prepared to provide guidance as to what we believe is the best path for our customers, even if it means abandoning some of our usual even-handedness with regard to hardware and operating system alternatives.
Editor: John Walker