Apple is arming itself to fight future commercial wars. Jack Schofield reports from London
IN 1991, Macintosh evangelist Guy Kawasaki wrote in his Macworld column: “Our greatest technical challenge is creating a computer that leapfrogs Macintosh just as Macintosh leapfrogged the IBM PC … In the next five to seven years Macintosh technology will be at the end of its lifetime, and if we don’t cannibalise Macintosh sales, someone else will. Nothing is more important to the long-term viability of Apple than accomplishing this goal.”
Five years and three chief executives later, Apple has failed to create anything that “leapfrogs Macintosh”. Instead it has taken over a company founded in 1985 to do precisely that: NeXT. It just has to persuade us that NeXT’s operating system – which has already failed in the marketplace, been abandoned by IBM, runs on Intel Pentiums not the PowerPC chips now used in Macs, and has hardly any software – really is the best starting point for the future.
The NeXT story began with the failure of the original Macintosh, after which Apple’s board backed John Sculley, the new chief executive, against one of the firm’s founders, Steve Jobs. Sculley restructured Apple, closing factories and laying off 1 200 staff. Jobs left to set up NeXT with some key members of the Macintosh team – followed by an Apple lawsuit – to start again.
NeXT Computer’s idea was to build a networked Unix system, NeXTstep, that would be years ahead of its time in exploiting technologies such as object-oriented software (built out of re-usable components), digital signal processors, and magneto-optical drives. It succeeded. In 1989, it unveiled a futuristic Motorola- based Cube that was more powerful than a Mac with the same processor. It was also easier to program and had a better-looking interface.
The Cube had what might have been a winning feature. It used Display PostScript to draw images on the screen, so what you saw was exactly what was printed on PostScript laser printers, not just so-close-it- didn’t-usually-matter. Display PostScript appealed to the desktop publishing users that formed the Mac’s heartland, but they did not buy it. Nor did anyone else.
NeXT struggled on until 1993, when it stopped making hardware and became a software company. It had claimed to be “the fourth largest domestic supplier of Unix workstations in the United States”, but seems to have sold only 50 000 machines.
As a software supplier, NeXT has concentrated on marketing to large corporations mainly via a dedicated sales force, which will now move to Apple. What it sells is a version of NeXTstep called OpenStep, which runs on Sun’s Solaris and other versions of Unix, and on Microsoft’s Windows NT. NeXT has also added WebObjects, a way of creating commercial web applications.
OpenStep’s appeal, according to Jobs, is that it allows custom applications to be developed five or 10 times more quickly than using conventional methods. This is not a mass market. NeXT Software Incorporated has fewer than 3 000 customers, though these include AT&T, Fannie Mae, Merrill Lynch, Nissan Motor, the US Navy, and half the Fortune 500 list of US corporations.
But it is not a very profitable business. According to the Wall Street Journal, NeXT lost $42,2-million in 1993 on revenues of $14,2-million. In 1994, it made $1-million on $49,5-million in revenues, thanks to a one-off $17,5-million licensing deal with Sun. In 1995, NeXT lost $800 000 on sales worth $46,8-million, and in the first half of last year, it lost $8,5-million on sales worth $21,7-million.
So why is Apple paying $400-million to buy the company and its staff? It is not doing it to get a version of Unix: Apple started selling its own very similar Unix, A/UX, in 1988, and is now working on the popular freeware system, Linux. It is not doing it to get a microkernel (small operating system core) for Mac OS: NeXT’s Unix does use Carnegie-Mellon University’s Mach system, but not the microkernel version. Apple could have got that either from CMU or from IBM, though it had already written its own for the Copland version of Mac OS.
Apple has bought a well-developed operating system that can be made to run on current and future Macs (apparently most of the work on a PowerPC version has already been done), but it offers limited compatibility with the past.
Yes, there are some Mac emulators, but OpenStep doesn’t support essential Apple system services such as QuickDraw, QuickTime and AppleScript. Ellen Hancock, Apple’s chief technology officer, says “backward compatibility in the rich sense” won’t be available until next year, and with NeXT’s Avie Tevanian in charge of Apple’s OS development, it would be a surprise if every choice between Apple and NeXT technologies went to Apple.
More importantly, Apple has bought an environment that isn’t locked to a particular vendor’s hardware, including its own. Like NeXT, it will sell OpenStep and WebObjects to people with Unix workstations and PCs.
Apple won’t want to frighten away Mac buyers with the idea of Pentium-based “Macs”, but chief executive Gil Amelio keeps saying that Apple’s future is in hardware-independent “middleware” (software that sits between the base operating system and the application) and that’s what OpenStep and WebObjects are.
Apple now has to convince software developers and users to adopt its new system, which will not be easy. It will, of course, continue to sell Macs – and to develop Mac OS – for as long as there is a viable market. But nothing lasts for ever, and as Kawasaki foresaw, the Mac is starting to approach the end of the line.
Still, at least Amelio is now arming Apple to fight the next, or NeXT, round of the computer wars, in objects and middleware, rather than the last one.