Most Mondays, their visit to Ive would be followed by one to Avie and the team working on Apple’s new operating system, which would eventually be called OS X. The radical new operating system would be the flywheel of all the extraordinary developments that would follow over the next decade, from Apple’s suite of iLife applications, to iOS—the slimmed-down operating system that would give life to the iPhone and iPad—to the entirely new software industry that emerged to produce the millions of apps written for those devices.
While Steve’s gadgets and computers drew the most attention, the software that made them go was every bit as important. Steve always said that Apple’s primary competitive advantage was that it created the whole widget: the finely tuned symbiosis between the hardware and the software together defined a superior user experience. In the PC world, hardware and software technologies came from different companies that didn’t always even get along, including IBM and the PC-clone manufacturers, Microsoft, and Intel.
Without a new operating system that could outshine Windows, the revival of the Macintosh could never be complete. The existing one was based on technology that had been developed fifteen years earlier for the original Mac, and the look and feel on the screen had come to seem passé.
Back at NeXT, Avie had developed a version of Unix that presented a friendlier face to nontechnical users, while also retaining its bona fides as a serious, world-class computing environment. There, too, the goal had been to create the whole widget, so he designed it to dovetail nicely with the NeXTcube. But when the company was forced to refocus solely on software, Avie and his team knew that the only way they could sell the NeXT OS was to make it attractive to users of workstations made by other manufacturers, like Sun, IBM, or Sony, and perhaps even to users of standard PCs. That’s why they had created experimental working versions on Sun workstations using SPARC microprocessors, on other personal computers and engineering workstations using Intel’s best Pentium PC microprocessors, and even on the PowerPC chip that was now the heart of Apple’s latest Macintoshes. This experience of “porting” NeXT OS to other machines would pay off in two ways for Apple. For starters, Tevanian and his crew walked in the door at Cupertino with the code base and the know-how to support the troubled company no matter which microprocessor would be at the core of future Macintoshes. Apple had already switched Macintosh microprocessors once before, and Steve wanted the flexibility to do so again if it made sense. Since his old NeXT programmers had learned the technological idiosyncrasies of several computing platforms, they could help him make a much more objective decision when it came time to switch again. Technologically agnostic, they would push for the architecture that would get the most out of their operating system—in other words, the one that would help them build the best whole widget possible. This was an ace up Steve’s sleeve, one he would play to great effect several years down the road.
Second, and more immediately important, the travails at NeXT had turned Tevanian’s crew into a first-rate team. The primary task they faced was to turn the NeXT operating system into something that remained robust but had a modernized look and feel that bore enough similarity to the original Apple system for Mac users to migrate over with as little discomfort as possible. Another priority was to preserve compatibility with software applications that ran on the old Mac OS 9, at least in the short run. Finally, they had to build tools for software developers to help them adapt their old applications to OS X or even rewrite them altogether to take full advantage of its capabilities.
The challenges in developing any new operating system are many and varied, and even though OS X was essentially a modified version of a proven, existing operating system, the “Apple-ization” of it was still an enormous job. Steve understood this, and he didn’t create unreasonable deadlines for his programmers. Instead, he oversaw them with a mix of patience and impatience that allowed him to be forceful and yet respectful. What eventually resulted was an operating system that mixed the best of Steve’s intuitive understanding of the needs of regular people with deep, robust, and flexible code written by some of the greatest programmers in the world. It preserved the winsome onscreen personality that had made Apple customers so loyal through thick and thin.
Steve was particularly obsessed with the operating system’s look and feel. In the afternoon OS X meetings that Slade would attend with Steve, each of Avie’s direct reports would be admitted into a locked conference room to demonstrate the latest developments on whatever aspect of OS X they were handling. “We went over OS X again and again,” remembers Slade, “pixel by pixel, feature by feature, screen by screen. Should the genie effect look like this? How big should the dock icons magnify? What’s the type style? Why does this dial look the way it does? Every week, the agenda was to get Steve to approve the look and feel of each item.
“There is nothing in the operating system that he didn’t approve,” continues Slade. “It was the opposite of how things were done at Microsoft, where they relied on these five-hundred-page specs [documents laying out in detail every feature to be created by the software developers]. We had specs, too, but Steve never looked at them. He just looked at the product.”
When Steve saw something he didn’t like, he would tell a user interface designer by the name of Bas Ording to mock it up the way he wanted it. “Bas was a wizard,” says Slade. “He’d take ninety seconds pecking away, he’d hit a button, and there it was—a picture of whatever Steve had asked for. The guy was a god. Steve just laughed about it. ‘Basification in progress!’ he’d announce.”
What made the OS X development even harder was ensuring that the new operating system wouldn’t instantly render users’ old applications useless. This backwards-compatibility is one of the most difficult challenges a computer company can face—it was a real problem for Apple back in the early 1980s, when Apple II customers found that their software didn’t work particularly well on an Apple III.
Steve believed that Apple’s consumers would adapt more easily than conventional wisdom suggested, since they were far more enthusiastic about their Macs than Microsoft’s customers were about their PCs. He believed they would be quite willing to make a big leap to a new operating system, even if it also required eventually buying all-new hardware and software. And he was right. Over the next decade, in its quest to keep the OS lean and modern, Apple would slowly stop supporting for a variety of carryover features from previous generations of hardware and software that were dearly beloved by a sometimes-vocal minority. Most Mac customers figured the trade-offs involved in a steadily improving computer platform were worth it, however.
Still, Steve and Avie did everything they could to make the transition to OS X as benign as possible for customers. One thing they exploited was a new way of delivering software updates. With more and more computers constantly connected to the Web, Apple could update users’ software frequently by delivering improvements, modifications, and bug fixes directly over the Internet. This applied not just to operating system software but to all manner of applications, and made sense both for the customers and for the software developers, who by nature love to continue tweaking their work once it is “finished.” Avie and his team were among the first mainstream operating system developers to take full advantage of this capability, and their approach would change the expectations of hundreds of millions of people, from corporate IT managers all the way down to the individual smartphone user who wants the very latest version of his favorite game.