Overdrive, page 15
Gates and other family members spoke at the memorial service for Mary that was held at the University Congregational Church in Seattle a week after her death. More than a thousand mourners packed the church to say goodbye to the remarkable woman who had forged a brilliant career in business, civic activism, and philanthropy. Mary Gates had served as a regent at the University of Washington for 18 years. She was the first woman to be president of the United Way of King County, and she was a director of many corporate boards. She was also a tireless volunteer for the Seattle Symphony, Washington Gives, the Seattle Foundation, the YWCA, and several other organizations.
At the memorial service, Gates and his two sisters recounted stories of games played around campfires, and of their mother’s support in school and in their careers.
“I am the son of Mary Gates, and she was a wonderful woman,” Gates said. “Not many adult sons are as proud of their mother as I was.”
Gates said his mother never stopped stressing the importance of family. “She was always concerned about my getting married. It came up a lot.”
He recalled the well-known story of how his mother’s connections with former IBM Chairman John Opel had helped Microsoft make the deal of the century when Big Blue needed an operating system for its first personal computer in 1980 and came calling on Microsoft. At the time, Opel and Mary Gates served together on the national board of United Way.
Gates also told the story of how his mother had been influential in wooing Steve Ballmer, then at Stanford Business School, to Microsoft in 1980.
His voice breaking with emotion, Gates concluded by saying, “She would know better how to show appreciation. It has meant a whole lot, and we thank you.”
Following his mother’s death, Gates turned to the woman he loved for support to see him through the difficult days ahead as he refocused on work and the many battles still to be fought.
“Although his mother had a major influence on him, Melinda will have much more,” said one of Gates’s best male friends. “For a young woman, she has a lot of common sense. [S]he’s a real, both-feet-on-the-ground sort of young person. I have been very, very impressed by her as I [have] gotten to know her. And I think, clearly, while she’s full of life and spirit, she has an extraordinarily large quotient of good solid common sense. She will be able to help Bill even more than he realizes.”
The Davids vs. Goliath
Only a fool would want to be stranded in a winter snowstorm at the Tompkins County Airport in Ithaca, New York, and Steven Sinofsky is no fool. So Sinofsky, one of Microsoft’s brightest people, and technical adviser to the boss, Chairman Bill, did the sensible thing and returned to Cornell University, only about three miles away.
He had flown to his alma mater from Seattle in early February 1994, on a recruiting trip, looking for high-bandwidth people like himself who wanted to work for Microsoft and take the company into the next century by helping to write and develop great new software. But that had been business, and had left no time to relax and look around. Now, with the airport closed, snow coming down, and nothing but time on his hands, Sinofsky could reacquaint himself with the campus he knew so well as a student. What he found confirmed his growing concern that Microsoft was slighting the Internet phenomenon that was sweeping inexorably across the land.
Seven years earlier, when Sinofsky was a student at Cornell, the Internet had been a refuge for just the technically savvy, himself included. Now, ordinary students and faculty were using UNIX workstation computers and TCP/IP networks with relative ease. They were exploring the World Wide Web, exchanging e-mail, sharing information with friends and colleagues at universities all over the world, and sending Gopher on search missions. Students were even accessing their course lists off the Net. These college kids were hip to the Net, exploring a world to which many of the top computer programmers in the country at Microsoft’s campus in Redmond did not have easy access.
Certainly this was something to write home about, and Sinofsky dashed off an e-mail to his recently married friend. The message was titled “Cornell Is Wired!” Later in February, after Sinofsky returned to Microsoft, he gave Gates a Web- surfing demonstration with Mosaic. Finally the Internet had the attention of the man in charge of the biggest and most powerful software company on the planet. But that was just about all. Microsoft still had not started to develop a business and technical strategy for responding to this phenomenon from cyberspace.
It would be more than a year before Gates fully acknowledged the importance of the Internet and decreed that all of Microsoft’s considerable brain- and firepower be directed at the most serious challenge the company had ever faced. But Sinofsky’s e-mail from Cornell, Gates would later tell writer and futurist George Gilder, was the wake-up call that prompted his “Internet Tidal Wave” memo to his executive staff in May 1995.
Sinofsky soon discovered he was not the only one sending e-mail warning Microsoft executives about the Internet boom. James Allard, the programmer who was working on Internet communication protocols in Nathan Myhrvold’s Advanced Technology Group, had been growing increasingly frustrated at the company’s indifferent attitude toward the Net’s unparalleled growth. Recruited by Myhrvold out of Boston University’s computer science program, the Net-sawy Allard arrived on campus in 1991, geared up to help Microsoft seize the Internet as a great new business opportunity, and to extend the company’s reach far beyond desktop operating systems and application programs, he was the first programmer at Microsoft to have the word “Internet” on his business card, which read in full, “Program Manager, Internet Technologies.” Working on Myhrvold’s nonsanctioned project to develop TCP/IP for Microsoft, Allard had been making a lot of noise about the Internet, so much, in fact, that he had been heard way out in the East Tech building where Jeff Lill and his Marvel team were racing to develop the company’s proprietary on-line system. “Allard basically thought that we were nuts not to be doing something for the Internet,” said Lill. “He thought we were off base.”
On January 25, 1994, shortly before Sinofsky had flown off on his head-hunting trip to Cornell, Allard wrote a long memo to Myhrvold and other senior managers. “I finally just couldn’t take it anymore,” Allard later told Business Week, explaining why he wrote his memo. “I felt the company just didn’t get it.” Titled “Windows: The Next Killer Application for the Internet,” the memo suggested that Microsoft get busy creating its own Mosaic-like browser and include Internet communication protocols in Chicago, which was destined to become Windows 95. Microsoft, he said, had to “embrace and extend.” Those two words would eventually become the centerpiece of Microsoft’s Internet strategy.
Why didn’t Microsoft “get it” sooner? After all, it had some of the smartest people in the industry, including Myhrvold, the technology guru whose job it was to see into the future and help position Microsoft on the road ahead. “Nathan tends to spend a lot of time thinking about the future, but as measured by centuries and millennia, and sometimes he’ll disdain to look at decades,” said Rob Glaser, who had passed along to Sinofsky his recommendations for making Marvel an Internet- based service. “One of the challenges, when you have real future-oriented people who are great at seeing the world of the possible, is how to marry them with people who know how to see the relationship of the possible and the world of the actual. And while I’m super impressed with Bill’s personal ability to do that, there’s no way for any one person to do that across an organization that has 20,000 people, that is in so many different businesses and industries. That’s just a structural challenge within any company.... There were other people at Microsoft who certainly thought about the future; but in terms of really mapping corporate strategy for a lot of these fundamental sea changes, I think that it’s just inherently an ad hoc process
Having said all that, though, I want to extend my most effusive praise for how quickly the company pivoted, once the company, and Bill personally, understood the magnitude of the impact and the power of the sea change.”
In fairness to Microsoft, Glaser noted, the Internet phenomenon happened very fast, and the difference between being late and being early was a matter of months, not years, and during those months most of the company’s talent was focused on Chicago, which was supposed to ship before the end of the year. “Should they have been smarter about understanding these other overall trends? Well, yeah, they probably should have been,” said Glaser. “However, it’s really easy to get tunnel vision on something you’re focusing on.”
But Microsoft’s hegemony also got in the way: it was used to calling the shots, as it did with DOS and Windows. Microsoft had the band, and the industry marched to its tune. In contrast, the wild and woolly Internet did not look like an environment in which Microsoft was going to be able to call the shots. So Microsoft chose to ignore it, partly out of ignorance, partly out of arrogance.
The company also was worried that the Internet posed a grave security risk across its corporate network of some 30,000 PCs. In early 1994, most employees could not get an Internet connection except on a computer that was not connected to the Microsoft corporate network. There were only two such machines at Microsoft’s library, and the company monitored their use with sign-up sheets.
“Microsoft wouldn’t embrace the Internet internally because of all these concerns with security, and so none of us knew much about it,” said one of the Marvel managers who had been present in December 1993 for Glaser’s class on Internet 101. “Most college kids knew much more than we did because they were exposed to it. If I had wanted to connect to the Internet, it would have been easier for me to get into my car and drive over to the University of Washington than to try and get on the Internet at Microsoft.”
The company’s Information Systems Department wrote a custom version of FTP (file transfer protocol), but to download files from the Internet, a Microsoft employee had to have the approval of a manager, then the approval of a vice president to send the file over the Internet to someone else.
“A couple years before,” said Glaser, explaining Microsoft’s rationale for its seemingly irrational security concerns, “someone had taken some of the code for Apple’s Macintosh computer and put it out over the Internet. There was a big panic. Their proprietary assets had been compromised. And I think that induced in the bowels of Microsoft a sense of, ‘My God, this Internet thing is dangerous.’ And so, literally, most people at Microsoft were institutionally cut off from the emerging revolution of the Internet.”
For much of January and February, while Allard and Sinofsky were trying to excite others at Microsoft about the Internet, Gates was preoccupied with a trial that was being played out before a federal court jury in Los Angeles. In a classic David versus Goliath scenario, Stac Electronics was taking on mighty Microsoft for patent infringement. It claimed Microsoft had illegally used Stac’s data-compression technology in the latest version of Microsoft’s operating system, DOS 6.0.
In 1993, Stac had earned only as much money as Microsoft had in any given four-hour period. Microsoft had not lost a significant legal fight since the late 1970s when Microsoft, then a tiny company with a handful of employees, had stood up to Pertec Computer Corporation and its hotshot lawyers. Now Stac, a pipsqueak company based in Carlsbad, California, was facing down Microsoft, and it was about to throw a knockout punch against the biggest in the computer industry.
Stac was essentially a one-product company. In 1990, it had developed Stacker, which compressed the data on a computer’s hard drive to free up valuable space. With Stacker, a 20- megabyte hard drive, for example, could store about 40 megabytes of information. Stacker was available for both the PC and the Mac, and its sales amounted to about 85 percent of the company’s business. In 1992, Microsoft began negotiating with Stac to license Stacker, which was dominating the market. Microsoft wanted to incorporate Stac’s technology into the next version of its operating system. The talks dragged on for several months, then broke off. Stac spokespeople said no deal was reached because Microsoft had refused to negotiate any royalty payment to Stac. In pretrial papers, Microsoft claimed that it had offered Stac a licensing fee of $1 million a month, but that Stac CEO Gary Clow had demanded $4 million, and that’s why no agreement had been reached. Meanwhile, Stac obtained a patent for its Stacker technology, and Microsoft began working on its own data-compression software, called DoubleSpace, for which it subsequently obtained a patent.
In May 1992, Stac had gone public at $12 a share. But after Microsoft announced it would ship DOS 6.0 with DoubleSpace, the price of Stac’s stock dropped to $3 a share and a shareholder class action suit was filed against it. The company had to lay off about 20 percent of its workforce. By the fourth quarter of 1993, Stac’s revenue had dropped 50 percent, to $6 million, versus sales of $33 million in 1992. In court papers, Microsoft claimed Stac had dragged out the negotiations to make its initial public offering more successful, and that it did not state in its prospectus that Microsoft was planning to include its own data-compression product into DOS.
Despite clandestine overtures from both sides, the two companies could not agree on an out-of-court settlement, so the case went on trial on January 18, 1994. The day before, a massive earthquake had rocked southern California, causing widespread destruction and killing a dozen people in an apartment building in Northridge. It was regarded as an ominous start to a trail that was being closely watched by the entire computer industry to see whether a small competitor could stand up to Microsoft’s formidable market power and legal muscle and walk away a winner.
It was also seen as an important legal case because of the issue of patents. More software companies, including Microsoft, were obtaining patents to protect their products, and many in the industry were concerned that the patents were so broad that they inhibited innovation and competition, thus doing more harm than good. Case in point: the year before the Stac case went to trial, the U.S. Patent Office had issued a broad patent to Compton’s NewMedia for a search technology used in CD-ROM disks manufactured by many companies. If Compton’s patent were upheld, it would force companies to pay expensive licensing fees to Compton, a subsidiary of the Tribune Company.
Although patents are more difficult, time-consuming, and expensive to obtain than copyrights, copyrights had proved to be of little use in protecting certain aspects of software. Apple’s lawsuit against Microsoft over Windows, for example, had involved copyright issues, not patent protection. Apple claimed that Microsoft had copied the look and feel of the graphical user interface used in the Mac. It sued Microsoft for more than $5 billion. But in 1992, before a crowded San Branciseo courtroom, federal Judge Vaughn Walker ruled that some of the similarities between Microsoft’s Windows and Apple’s Macintosh were technologies that were already licensed to Microsoft and therefore not an infringement and that the rest were beyond the scope of copyright protection. A ruling in favor of Apple’s claims, the judge said, would have “afforded too much protection and yielded too little competition.” This not only vindicated Microsoft, it also was significant for the industry, clearing away doubts about the rights of software programmers to adapt aspects of other systems.
The Stac case raised even more troubling issues of intellectual property law. Microsoft had been sued before for patent infringement, but this was the first time a case had gone to trial. The jury would hear not only Stac’s suit against Microsoft, but Microsoft’s countersuit against Stac. The suits had been consolidated before trial. Microsoft claimed that Stac had infringed on its patents, was guilty of breach of contract, and had conspired to commit fraud.
Early in the trail, on January 28, Gates was called to testify, and was grilled by Stac’s lawyers in the packed courtroom. Gates denied that Microsoft had done anything wrong. He said the company’s programmers worked day and night to develop the company’s own data-compression technology and had not relied on Stac’s code. “We wanted to do absolutely everything to make sure we weren’t taking a risk (of patent infringement),” Gates testified. “There was a consensus that we had a product that we could sell.” He also testified that he was not personally involved in the failed negotiations with Stac.
It was not the first time Gates had testified at a civil trial to defend Microsoft’s business practices. In 1986, the year Microsoft went public, Gates testified in a $60 million suit brought by bankrupt Seattle Computer Products, which had sold DOS to Microsoft in 1981 for $50,000. After a three-week trial, Microsoft settled the suit for $1 million while the jury was still deliberating.
Nearly a decade earlier, when Gates had taken the stand for the first time in an Albuquerque courtroom, it could have meant the end of his upstart company had his testimony not carried the day. The claimant in that case was Ed Roberts, who had invented the Altair, the world’s first personal computer, for which Gates and Paul Allen had developed the BASIC operating language. In 1977, Roberts had sold his company, MITS, to Pertec, a company specializing in disk and tape drives for minicomputers and mainframe computers. But Pertec wanted BASIC—at the time, Microsoft’s cash cow—as part of the deal. Gates knew that if Pertec obtained the rights to BASIC, Microsoft would not survive.
When the chief counsel for Pertec arrived in Albuquerque to assess the situation and talk with Gates, he took one look at the long-haired, scraggly, 21-year-old techie and assumed it was going to be no contest, even though Roberts had warned Pertec that it would have its corporate hands full with Gates. “Pertec kept telling me I was being unreasonable, and that they could deal with this guy,” recalled Roberts, now a country doctor in Georgia. “It was a little like Roosevelt telling Churchill that he could deal with Stalin.”
The Pertec case was held before a judge, not a jury, and went on for three weeks, from 8:00 a.m. until about 5:00 p.m. “It looked like giant Pertec was picking on these poor 19- and 20-year-old guys and trying to steal their life’s work,” said
