Dr. Dobb's Journal March 1997
Aargh. Last month I wrote about how Apple was going to acquire the Be operating system as the key to revitalizing the MacOS. And now, as the whole world knows, Apple has decided to go with Steve Jobs' NeXT operating system instead.
Okay, to be fair to myself, I didn't predict that Apple would go with Be. I said that that's what the rumor mill said, and I hedged my bets a lot by mumbling about the time differential "between the writing of this column in my temporal reality and the reading of it in yours," and I actually did invoke Steve Jobs' name, and I did mention several operating-system alternatives to BeOS. None of them, however, was NeXT's.
Like just about every industry watcher I know, I totally failed to see this one coming. Odd, too, because it's just the kind of story industry watchers like: operatically dramatic, richly ironic, lush with personality and poignancy, and open to many interpretations. But surely not as many interpretations as we're getting.
Everybody and her sister is writing about the deal. And I mean everybody. There have been or will be analyses in every magazine from People to -- well, I guess the other end of that particular spectrum is the ground we now occupy. The proper, if not properly modest, way to conclude that sentence is, "every magazine from People to Dr. Dobb's Journal." And following the implications of that sentence, since second-guessing Apple is one of my semiofficial roles here at DDJ, I guess I'd better get on with it. All right: Having slogged through the first 50,000 words or so printed in the papers and posted on the Internet regarding the particulars and the implications of the acquisition, I will offer my totally free and worth-the-price thoughts on the deal. But hold on a minute; that's not where I thought I was headed when I took the onramp to this paragraph. Back there at "Everybody and her sister," it had been my intention, and I thought the paragraph's intention, to say that I wasn't about to make a fool of myself like every other precipitous pundit.
Apparently I am, though. But if I must, at least I can attempt to inject a little originality. I won't just speculate about what this deal will mean to Apple. I'll also speculate on what it means to Steve Jobs. That sounds like something nobody cares about, but I have a hidden agenda, you see. Asking what this deal means to Jobs allows me to reflect on who this guy is, where he's been, what he's done. That is, it lets me rehash some personal-computer history, which I've been obsessed with recently. Also, one online columnist offered the opinion that Steve Jobs' involvement was a little scary, since he had no stock in Apple and was being careful to distance himself from the company and whatever it might do, while at the same time apparently advising the company on its operating-system strategy. The columnist saw Steve as having enormous influence without any personal risk. Steve has no stake in the outcome, he said. I disagree, and I think I can demonstrate that Steve has quite a bit at stake. But first, a couple of thoughts on the deal.
First thought: The deal itself. Reports are that Be Inc. was asking for too much money, and I gather that this at least made Amelio ready to listen to other pitches. Apparently it was a NeXT product manager who made the first contact with Apple, and Jobs didn't know about Apple's interest until Apple execs were already at NeXT in a meeting. And apparently Jobs got involved then, and the deal went from nowhere to press conference announcement in a matter of weeks.
I wasn't there in the room for the final negotiations, but I'm guessing that the clinching argument was made by Jobs. I'm guessing that he won Amelio over by explaining what the Be deal meant versus what the NeXT deal could mean. Something like, "a desperate attempt to revive a moribund operating system versus a bold move to create a radically new successor to the Macintosh." Or something more succinct. After all, this is the guy who convinced John Sculley to leave a high position at PepsiCo and come to Apple by characterizing his options as staying and selling sugar water for the rest of his life or coming with Steve and changing the world.
Second thought: The future of the Mac. Amelio responded to one question by saying that he had only announced an acquisition, they hadn't announced an operating-system strategy. Which is true: Apple didn't even announce that it had any operating-system strategy, although there were promises that something would be said at MacWorld Expo in January. I know one thing that won't be said in January: The Mac is dead. But isn't that exactly what the NeXT acquisition means?
Blending elements of the BeOS and the MacOS into a new Mac operating system made some kind of sense. Blending NextStep with the MacOS is harder to imagine, and, at least initially, that's not what Apple's going to do at all. Apple has said so itself. The first NeXT-running machines Apple will release, probably the only NeXT-running machines it will release this year, will not run Mac applications. Repeat: They will not run Mac applications. These computers will not, in any sense, be Macintoshes. They surely won't even be called Macs. I have no idea what they will call them. Maybe NeXT Machines.
All right, but that's a transitional stage. Now jump ahead to 1998. Apple has acquired some sort of adequate Mac emulator and has ported some key technologies (QuickTime? OpenDoc? Which are the key technologies now?) to the NeXT platform. Meanwhile, the NeXT developers now working at Apple continue to develop NextStep, incorporating certain Mac features. Mac developers at Apple continue to tweak the System 7 operating system in minor ways. Apple has made it clear that it will continue to tweak System 7, and that tweaking is all we can expect from System 7 in the future. Now Apple announces a new line of computers. Are these dead-end System-7-only machines? Surely not. Do they run a new Mac operating system based on the best of NextStep and System 7? If anyone thinks that Apple can blend these systems into one from a standing start to a shipping product in a year, I have a cousin who'd like to talk with him about some land in Florida.
Okay, then these machines also run NextStep. They run Mac applications, too, probably via emulation, and therefore slowly. The user interface for the machines will be NextStep, not Mac, except while you're in an application running on the emulator. These machines, too, are NeXT Machines, not Macs. Okay, but those machines are transitional, too, right? In the long term, Apple will merge the two operating systems into one, right? And call the result the MacOS? And machines running it will have the look and feel of the Mac today? Well, maybe. But I don't see why.
I ask myself the following: Which parts of the MacOS does it make sense for Apple to try to preserve? The legendary user interface? The horrifying-to-behold OS internals? The OS-portable features like QuickTime and OpenDoc?
Certainly not the guts. Apple hasn't been able to deliver System 8 in part because System 7 is such a godawful mess inside that the only sane thing to do with that part of the OS is to throw it away and start over. The guts of the MacOS are the problem Apple is trying to solve, not a part of the solution.
Perhaps surprisingly to those who haven't seen the NeXT user interface, I'd say not the user interface, either. This is heresy to Mac enthusiasts, but I don't see any way around it. You can't pick the best elements of two different user interfaces and put together some sort of hybrid from them. You can, sometimes, with great care and customization, transplant an isolated feature of one user interface to another, but that's the exception that proves the rule. That rule is: A good user interface is one designed from a single conceptual model. Mac and NeXT are designed from different conceptual models. You can't paste them together; you have to pick one. And NextStep is the one to pick, because it's just better. That leaves the separable components. And a lot of these look questionable. Apple's display technology is currently faster than NeXT's, but isn't it clear that NeXT's choice of Display PostScript is the better choice in the long run? Well, maybe it's not clear, but there's an argument to be made. And there's an argument to be made regarding nearly every one of Apple's system technologies.
No doubt a lot of these technologies will survive, but it's not clear (well, not to me, anyway) which ones. And I don't see why any other part of the OS should survive.
So it's 1999 and, over the span of two years, Apple has sold two generations of NeXT Machines and is porting the key system technologies to NextStep as quickly as it can. The machines have had at least modest commercial success (otherwise it's all over anyway) and have received generally positive press coverage (because the NeXT software has never had anything less than generally positive press coverage). Meanwhile, Apple's charismatic cofounder has been going around proclaiming that NeXT is saving Apple, and the press are eating that up, too.
Now, do you try to revive the Mac label and attach it to your next line of machines, despite the fact that the only Mac technology in them is a collection of separable OS technologies like QuickTime that, by this time, probably all run on Windows, too?
I just don't see it. Apple is going to ship non-Macintosh computers this year, and I don't think it's going to subsequently backtrack from that daring (or desperate, if you prefer) platform leap. It will certainly squeeze every penny it can out of its installed base of Mac users and put tremendous spin on the messages it puts out about its Mac future, but surely the Mac is history, as of Christmas 1996.
Does Steve Jobs have any stake in Apple's future? A lot has changed for Jobs since he left Apple more than a decade ago. He's married now and has a family. The NeXT experience had to be humbling. And both Steve and Apple are simply older. The company is now as old as the man was when he and Steve Wozniak founded it. The deal with Apple makes Jobs a part-time consultant reporting to CEO Amelio, with no one reporting to him, no clearly articulated responsibilities, no seat on the board, no stock in Apple. He's got his billion (from his ownership of Pixar), so he's comfortable. He certainly has no further debt to the long-suffering employees who stayed with NeXT: Not only do they have new management with a big incentive to make them happy, but (as I understand it) most of them have stock in NeXT, which (I suspect) finally, as of the acquisition, became worth something. So does Steve have any stake in Apple's success? I think so.
The Macintosh would never have been if it hadn't been for Steve Jobs. He saw the light on a visit to Xerox PARC. He hijacked a team of developers at Apple working on an interesting idea for appliance computers and turned them into the Macintosh skunkworks. He hounded them to do their best work and promoted the Mac as the revolutionary device that it was. Even if it did look sort of like a toaster.
In 1984 and 1985, the first two years of the Mac's life, it wasn't doing so well. Sales were not necessarily bad, they were just an order of magnitude smaller than Steve's projections, projections that the company was counting on. For those two years, the Apple II supported the company. All the while, it was the Mac team who got the perks, the money, the recognition. And the Apple II people within Apple got the clear message that they were part of the past. They were bitter.
The Mac wasn't selling slowly because of lack of advertising spending. Its relative failure in the market was richly deserved. It lacked some things that users had every reason to demand. Initially, there was no hard-disk drive, and a second floppy-disk drive was an option. With only a single disk drive, backing up was a nightmare of disk swapping. At 128 KB, RAM should have been more than adequate, it seemed. 64 KB was pretty normal at the time. But the system and application software soon ate up the 128 KB, and it was clear that more was needed, fast. DDJ published an article on how to fatten your Mac to 512 KB six months before Apple managed to get a fat Mac on the market.
But to care about the memory limitations you had to have software to use the RAM, and that was the biggest problem. The Mac shipped with a collection of Apple-developed applications, so you could do word processing and draw bitmapped pictures. Beyond that, choices were slim, and the Mac proved to be a very hard machine to develop for. Two talented friends of mine went off to a mountain hideaway with the ringbound Inside Macintosh that first year and returned proudly days later with a program that displayed a window on the screen. That was all, just an empty window.
It took two years, but things began to turn around. And then, in a dramatic story that has been often told, Steve's fortunes turned around, too. When Jobs decided that he had made a mistake in hiring John Sculley to run the company, he attempted to force him out. His efforts backfired, culminating in a highly charged board meeting in which Jobs, in tears, was stripped of his power. "I guess I don't wear the right kind of pants for this company any more," he said.
Within weeks, he had left Apple, taking along key employees, sold his stock, and started a new company: NeXT Computer. Its purpose was to produce a new computer with a remarkably intuitive user interface based on windows and icons and menus, equipped with a mouse, running on Motorola 680x0 processors. Its purpose was to show Apple. To show the world how it should be done. To show everyone that Steve was right.
Technologically, the NeXT Machine showed all of that. While the Mac had done an excellent job of implementing the graphical user interface that Steve had seen at PARC, the NeXT Machine implemented much deeper PARC technologies. Its operating system, built on the Mach UNIX kernel from Carnegie-Mellon, was deeply object oriented. This made the user interface essentially more real than the Mac's: Whereas UI metaphors had to be crafted bit-by-bit on the Mac, in the NeXT system, they tended to flow naturally from the underlying object nature of the OS. The object-oriented OS also made it possible for NeXT engineers to create an extremely powerful development environment for corporate custom development. NextStep is still regarded by many as the best development environment that has ever been put on a computer.
Commercially, the NeXT Machines flopped. In 1993, finally acknowledging the obvious, NeXT killed off its hardware line and transformed itself into a software company. It immediately ported NextStep to other hardware, starting with Intel's.
The reception was heartening. Even conservative CIOs who worry much about installed bases and vendor financial statements were announcing their intention to buy NextStep. It got top ratings from reviewers, and custom developers were citing spectacular reductions in development time from using NextStep. And it ran "like a Swiss watch," according to one reviewer.
But for all the praise, NextStep did not take the world by storm. While custom development may have been spectacularly easy, commercial applications of the killer app kind didn't materialize. NeXT continued to struggle along, continuing to improve the operating system and serving its small, loyal customer base well, but never breaking through to a market share that could sustain the company in the long run.
As recently as a year ago, Steve Jobs was pitching hard for a technology that he thought might save NeXT. WebObjects is a development and deployment environment for web applications. WebObjects relies on a scripting language, WebScript; a collection of NeXT-developed objects and tools; and, ultimately, Objective-C, NeXT's development language, which consists of key elements of SmallTalk grafted onto ANSI C.
An interviewer at that time expressed surprise that Steve didn't want to talk about Pixar, whose movie Toy Story has just made Steve a billionaire, but was rather insistent about describing this technology of a company many considered a failure. WebObjects, incidentally, was apparently one of the technologies that caught Apple's eye when it acquired NeXT in December. It was at the announcement of that acquisition that Steve Jobs acknowledged that he still cared about Apple. I suspect that's true. And I think he cares about Apple being rescued by NeXT, because surely that would prove that he was right all along.
If Apple succeeds, Jobs recovers his pride. Does he have much at stake here? I think so.
DDJ