The personal computer industry exists in a peculiar state of schizophrenia because its two component parts, hardware and software, keep pulling in opposite directions. The frantic pace of hardware development brings us a new generation of more powerful machines every few months. But since one piece of hardware requires hundreds or thousands of pieces of software, each of which takes time to write or translate, software can never keep pace with hardware changes. As a general rule of thumb, any time a decent library of useful software has been developed for a computer, its hardware has become hopelessly outdated and a new model is on the way.
The industry tries every way it can to cope with this dilemma. Many new products are designed to be upwardly compatible with older ones. Like the Commodore 128 and the Apple IIGS, they add new features to an existing design, allowing them to run software written for the earlier machines. But the requirement that the new computer be compatible with the old one imposes restrictions on how much further advanced the new machine can be over the original. For example, to maintain compatibility with the 6502 processor family, the GS uses the 65816, which, though more powerful, still doesn't quite measure up to the 68000 used by the competing Atari ST and Commodore Amiga.
Another problem of upward compatibility is that software firms write programs for the least common denominator. If you sell a program for the Commodore 128, your potential audience is 100,000 or so users. But if you limit the program's power so it runs on the 64, you add 2,000,000 64 owners to that audience. It seems likely that this thinking has slowed the development of software that uses the extended memory and 80-column screen of the Commodore 128, and that the same logic may prevent software that takes full advantage of the Apple IIGS from appearing as quickly as it would otherwise.
Where two machines are so different that upward compatibility isn't a viable alternative, many manufacturers are using add-on emulators to allow their computers to run software designed for another. With IBM dominant in the industry, it's no surprise that the makers of all three big-name 68000 machines-Commodore's Amiga, Atari's ST, and Apple's Macintosh-are currently working on IBM compatibility, each in a slightly different way. Apple is rumored to be readying a Mac with slots that will support a plug-in PC. Atari plans a stand-alone box that plugs into the DMA port, with a PC motherboard and memory, but no 5¼-inch-disk drive or slots. Amiga's Sidecar is the most elaborate, with a PC motherboard, memory, a 5¼-inch drive, and PC-compatible slots. And the PC isn't the only target for emulation. Commodore is encouraging the development of a 64 emulator (priced at $140) for the Amiga, and third-party Data Pacific has come up with Magic in a Sack, a software Macintosh emulator for the Atari ST.
Whether any of these emulators will substantially boost computer sales remains to be seen. In order to get perfect compatibility, the emulator has to duplicate the hardware of the computer it's emulating, making it cost almost as much as the real thing. In the past nobody has shown much interest in buying an Atari 2600 emulator for the same price as an Atari 2600, or an Apple II emulator for the 64 that costs the same as an Apple clone, so will anybody be interested in a PC emulator that costs as much as a PC clone? The Amiga Sidecar may succeed because it offers an upgrade path for the Amiga itself, providing a cost-effective way of adding a hard disk and more memory. The Atari box, while a less complete emulation, boasts a low price. And the Mac can count on the fact that price never stopped anybody from buying Apple products. Whatever the outcome, the development of split-personality computers comes as no surprise to an industry that's always been pulled in two directions.
As expected, Atari has taken the first step towards a public stock offering. Going public may mean big bucks for Tramiel and Co., but it's also forced them to reveal a lot of financial information. The Atari prospectus sheds some interesting light on its operations-some encouraging, some less so. On the bright side, the Tramiels have stemmed a tidal wave of red ink in a short time, taking a company that lost $60 million in '84 to a $12million profit in the first half of '86. But ST sales figures turn out to be smaller than commonly assumed. As of September 15, 1986, about 150,000 were sold, with perhaps half of those going abroad. While some previous industry estimates had the ST outselling the Amiga by a considerable margin, it now seems likely that the two are close to even in U.S. sales, despite Atari's six-month head start. The prospectus also notes the company's intention to build two- and four-megabyte machines, a blitter chip for fast graphics, a laser printer, and possibly a 1280 X 960 display. Finally, it's interesting to see that almost a quarter of Atari's sales still come from video games. Maybe we'll see that 2600 emulator for the ST yet....