CHAPTER THREE
THE RISE OF A JUGGERNAUT
* * *
The computer and software industries have thrived over the past twenty years precisely because there was little regulation of technical standards…. When the marketplace chooses standards, they aren’t perpetually frozen. Competitors have incentives to innovate as they try to topple existing standards. It’s a great system called capitalism. We need more of it, not less.
—BILL GATES, on the need for free markets, 1995
* * *
By 1980, the Apple II and other personal computers on the market were changing the minds of bigger, older computer companies about the future of the personal computer. IBM, which dominated the market for large mainframe computers, and Digital Equipment Corporation, which had been doing a booming business in what were then seen as “smaller” computers with a wide variety of applications, had been slow about seeing that PCs were the wave of the future. Indeed, Ken Olsen, the founder of DEC (whom Bill Gates had idolized as a teenager), had been debunking the PC since 1977, when he told a convention of The World Future Society, “There is no reason for any individual to have a computer in his home.” This famously mistaken judgment meant that DEC would later have to make a massive attempt to catch up, and it eventually led to Olsen’s ouster from the company.
IBM was also slow to see the possibilities of the PC, but at least it had the excuse that it was the leader in mainframes around the world. IBM was dubious, but not about to be caught entirely flat-footed, and in 1980 it made contact with Microsoft. It had a secret project for the development of PCs; if it was going to get them launched quickly, it would be necessary to go outside the company for the development of software to run the machines, rather than going through the lengthy process of trying to scale down its own mainframe software. Bill Gates had always been afraid that one of the big boys would do just that, leaving Microsoft in the dust. DEC had in fact scaled down some of its software in 1979, but because Olsen didn’t really believe in PCs to begin with, the company hadn’t gotten behind their new product in a way that threatened Microsoft.
IBM started off by playing things very cool. They sent two executives to Seattle, but as Gates would later tell the story, these men downplayed their own importance, saying that they were just planning people and much of what they planned never happened. But they had a long discussion with Gates and Allen about where the technology was headed, and the big prospects for personal computers. They said they would like to have Microsoft’s FORTRAN and COBOL languages, and perhaps a good deal more. The meeting made Gates think back to their Albuquerque experience with Ricoh, when Microsoft had promised software it hadn’t even developed yet.
* * *
When IBM introduced its PC in 1981, many people attacked Microsoft for its role. These critics said that 8-bit computers, which had 64K of address space, would last forever. They said we were wastefully throwing out great 8-bit programming by moving the world to 16-bit computers.
—BILL GATES, when 32-bit systems were standard, 1996
* * *
But Microsoft had a possible ace in the hole this time. It was negotiating to buy a little-known system called Q-DOS from a small rival company called Seattle Computer. If they could get their hands on it soon enough, they could license it to IBM. With Kay Nishi, their Japanese cohort, pushing them forward, Microsoft found themselves offering Q-DOS to IBM before the final papers with Seattle Computer had been signed. Two days of agonizing suspense ensued, with Gates and Allen worrying that Seattle Computer would get wind of the IBM deal and greatly raise their price. But since IBM was itself trying to keep its PC project secret, word didn’t leak, and Microsoft got Q-DOS for only $50,000. The system would prove instrumental in making Microsoft the industry giant it became.
Seattle Computer’s Q-DOS underwent many changes, of course, before becoming Microsoft’s MS-DOS. Microsoft had hired away the top engineer at Seattle Computer, Tim Paterson, and put him in charge of developing the new version. Since Microsoft also worked closely with IBM on the actual design of the IBM PC, there was a great deal of work to be done. Bill Gates and Paul Allen were still involved in hands-on development work, the actual creating of code, in those days, and the tension that must have existed from 1980 to 1981 surfaced momentarily fifteen years later when the two men gave a joint interview to Fortune. In the interview, Gates brought up the fact that in the midst of the IBM project, Allen had insisted on going to see a space shuttle launch. Allen quickly put in that it was the first space shuttle launch and that he had gone down to Florida and flown back the same day, being absent less than thirty-six hours.
* * *
The weirdest thing of all, though, was when we asked to come to the big official launch of the PC in New York, IBM denied us. About four days later we got this form letter that IBM probably sent to every vendor, even the guy who had the capacitators in the machine. It said something like, “Dear vendor, thank you for your help, blah, blah, blah.” They eventually apologized to us for that.
—BILL GATES, 1995
* * *
Originally designed for the Intel 8088 or 8086 microprocessing chips (and later for more advanced chips), MS-DOS was a powerful 16-bit operating system, using the then standard character-based mode that would be superseded by the graphical interface developed for the Macintosh three years later. The original MS-DOS had a memory limit of 640K, but that too would eventually be surpassed. Even with the advent of the graphical Windows operating system, MS-DOS continued to provide the underlying support. The initial MS-DOS system was considered fast, but as more powerful microprocessing chips were developed by Intel, it was updated to operate at much greater speed. A great deal of Microsoft’s success can be attributed to the fact that in MS-DOS it created an operating system that could serve as a sound basis for succeeding generations of more sophisticated operating systems and endless software applications.
Although the IBM PC would be in direct competition with the Apple II, Microsoft also developed its first application for Apple Computer in 1980. This was the Softcard for Apple II, which allowed that computer to run the CP/M operating system of Digital Research. But at the time, it was the relationship with IBM that Gates and Allen saw as the central building block for the future. Microsoft was not paid a great deal for its development work for IBM—less than $200,000—but Gates made certain that their contract with IBM allowed for Microsoft’s adapting MS-DOS for the clones of the IBM PC, which the hardware giant was prepared to authorize.
* * *
It was great that Paul got better, and we wanted him to come back more than anything. But there was just no part-time way to come back to Microsoft. If you were going to be there, you were really going to work hard. We all knew that. It’s still that way.
—BILL GATES, 1995
* * *
Once the IBM PC was on the market, Microsoft pushed MS-DOS hard, persuading other software companies to develop applications for the operating platform. This was important, since IBM was offering a choice of software, also making available a version of Digital Research’s CP/M operating system, as well as a far more expensive UCSD Pascal P-System. Since Microsoft charged IBM only a one-time fee, the MS-DOS cost only $60, as opposed to $175 for the CP/M and $450 for the UCSD system. Gates and Allen were convinced that if they could establish MS-DOS as the system in greatest use, they could make a great deal of money down the line. Their gamble paid off, and MS-DOS won the battle within a year. What’s more, the first clones were coming out, and Microsoft was poised to cash in. In addition, new software like the Lotus 1-2-3 spreadsheet was created to work with MS-DOS.
But just as Gates and Allen arrived at this moment of triumph, Allen was diagnosed with Hodgkin’s disease. It was initially thought that he had lymphoma, an often fatal form of cancer, but even the treatment for the much more controllable Hodgkin’s disease would entail twenty-two months of chemotherapy. Although he remained a director of Microsoft and sometimes attended variou
s other meetings, Allen backed off from his commitment to Microsoft during the two years of treatment. When he recovered, he made the decision to go off on his own to do other things. By then he was already a billionaire several times over, and he proved himself to be an astute investor in other companies, ranging from Ticketmaster, of which he owns eighty percent, to America Online and many high-tech companies. He bought the NBA Portland Trailblazers and became one of the new owners of the Seattle Mariners, in order to keep the baseball team in Seattle. He gives millions of dollars to charity every year, the beneficiaries ranging from cancer and AIDS research to libraries and the Oregon Shakespeare festival. As a director of Microsoft, he still has official input into shaping the company, but it’s clear that he has a special place as an informal prognosticator and sounding board for Bill Gates, and that the men remain both intellectually and personally close.
* * *
Software companies are forced to gamble on unproved markets because it’s nearly impossible to ask customers to predict whether they’ll buy and use a new kind of tool. Successful software companies push the frontier of what’s possible. We have no choice but to spend all the money to create a product before we sell any—and then hope there’s a big market for it.
—BILL GATES, 1997
* * *
With Paul Allen no longer a major force at Microsoft after his illness struck in 1982, it was up to Bill Gates to continue to build the company into the worldwide behemoth it has become. There are those who say that Paul Allen is a nicer guy than Bill Gates, and that it wouldn’t have been as ruthless a company if Allen had remained with it in his original capacity. But that ignores the fact that Gates was from the start more involved with the business end. And while Gates sometimes snaps back when charged with extreme aggressiveness, and denies many charges made against his company’s business practices, it is obvious he has run it with enormous success.
The first big step Microsoft took without Allen’s active participation was to develop a graphical interface. MS-DOS was character based. Gates explains the difference between the two formats by using a chessboard analogy: one format moves a chess piece by typing in words; the other shows a chessboard on the screen and moves the representation of the chess piece with a mouse. It may seem incredible to young computer users, but it was not until 1984 that the use of the mouse really became popular, with the introduction of Apple’s Macintosh. The technology of the mouse had originally been developed by Xerox, but because of the high cost of their computers, which also didn’t use standard microprocessors, they were unable to achieve market success with this breakthrough.
* * *
Our failures tend to result from markets being too small. Microsoft Bob was a product a couple of years ago that used on-screen cartoon characters to carry out tasks for people. Unfortunately, the software demanded more performance than typical computer hardware could deliver at the time and there wasn’t an adequately large market. Bob died.
—BILL GATES, 1997
* * *
The Macintosh was a different matter entirely. Microsoft’s reputation was such that Apple developed the Macintosh working closely with Gates’s company. Microsoft’s first graphical products, the word processor Microsoft Word and the spreadsheet Microsoft Excel, were created for the Macintosh.
But Gates was also working with IBM to develop a new operating system called OS/2. The two companies ran into numerous problems on this project. Some were technical, some arose from the fact that IBM laboratories were spread out across the country, leading to product turf wars, and some were a matter of developmental vision. Chris Peters, a Microsoft vice president, clarifies one of the main problems with OS/2 in Microsoft Secrets, a book on how the company develops its products, written with a great deal of input from Microsoft executives by Michael A. Cusmano and Richard W. Selby: “OS/2 was an attempt where they tried to change things…they tried to make things 10 percent better but completely different, and nobody wanted 10 percent better. We have a rule of thumb that things have to be twice as good before they can be different, if you’re trying for consistency.”
Gates became increasingly frustrated with the project, as did Nathan Myhrvold, the technical wizard who had joined Microsoft in 1986 when Gates bought his tiny company and hired its six-person staff. IBM, for its part, was annoyed with Gates’s attitude, and by 1989, the two companies decided to call a halt to their collaboration following the release of the first OS/2 product. Microsoft had already released its first two Windows operating systems, in 1985 and 1987, but they had been commercial failures. The company then brought out Windows 3.0 in 1990, which overcame the 640K boundary of MS-DOS (a limitation in the amount of information that could be stored). Work was already under way on Windows 3.1, but Gates was taking an enormous risk, essentially “betting the company” on the eventual success of Windows 3.1. Without the IBM tie-in, Windows 3.1 had to be a major success. It was, becoming the standard for personal computers and swamping IBM’s latest version of OS/2.
* * *
The rate of change of technology is faster today than ever before. Some of the big advances of the past, several generations would go by as it became popular—the telephone, even the TV set. Within the space of a single generation we’ll go from computers being something you can ignore very easily to the point where in most jobs, and to really be in touch, you’ll have to be comfortable with using it as a tool.
—BILL GATES, 1995
* * *
Even as Windows 3.1 was being released in 1992, the final papers in the IBM/Microsoft divorce were at last being signed. Microsoft retained the rights to the NT (for New Technology) software it had developed. (This was used for allowing networks of PCs to work together, and would become increasingly important in the years ahead. It would be incorporated into Windows 95, and its successors would be crucial to Microsoft’s move into the corporate PC market in 1997.) IBM was given use of Windows code, but only until late 1993. And Microsoft was also given a royalty on OS/2 sales—which would prove to be small potatoes when Windows 3.1 took over the market. Microsoft did pay IBM a flat fee, reported to be in the neighborhood of $25 million, for the use of some IBM patents. But given the eventual success of Windows 3.1, it is clear in hindsight that Microsoft took IBM to the cleaners on this resolution of their partnership.
The popularity of Windows 3.1 can be measured by the fact that it was installed on seventy million personal computers that already had been bought worldwide at the time of its 1992 introduction, and on ninety percent of the new computers bought between then and the August 1995 introduction of Windows 95. In terms of personal computer software, Microsoft had not merely achieved dominance, it had overwhelmed the competition. From 1992, Bill Gates was on a steady climb, not just year by year, or month by month, but week by week, toward becoming the world’s richest man.
* * *
Microsoft does the great majority of its software development in the United States, but that could change in the future. Our motive would not be to save money, however. We create software for the world and our success depends on drawing on a world of talent.
—BILL GATES, 1996
* * *
Developing the successor to Windows 3.1 proved to be arduous. For one thing, it moved from 16-bit processing to 32-bit processing, made possible by the increasing speed and capacity of Intel’s Pentium chips. This was a new ball game, which made it feasible to introduce a wide range of new features but greatly complicated the writing of code and also increased the number of bugs in the system. In addition, Windows 95 was delayed by Microsoft’s belated recognition of the importance of the Internet.
Windows 95 took off “like a bat out of hell,” to use an old phrase that sums up the attitude of Microsoft’s competitors. Between its August 24, 1995 release date and the end of the fiscal quarter on September 30, 1995, it sold an estimated seven million copies. This was a much higher number than either the computer industry or Wall Street analysts had expected. Microsoft itse
lf had announced that it expected to sell thirty million copies in the first year; at this rate, it would reach that number in less than five months. There had been enormous media coverage of the Windows 95 launch, and from now on, Bill Gates, who had hardly been ignored in the past, would become one of the most heavily profiled and interviewed men on earth, commanding almost as many magazine covers as movie and music stars.
But he was hardly resting on his laurels. His book, The Road Ahead, surged immediately to the top of the best-seller lists at the end of 1995. The book contained a bound-in envelope containing a CD-ROM, which included, to quote the jacket description, “the complete book text with hundreds of multimedia hyperlinks, a special interview with Bill Gates, video demonstrations of future technology, a World Wide Web browser and more.” The book was cowritten with Nathan Myhrvold, Microsoft group vice president, Applications & Content Group, and the Pulitzer Prize-winning journalist Peter Rinearson. Gates’s own profits from the book were used to fund “a grant for technology in education administration through the National Foundation for Improvement in Education.”
Bill Gates Page 3