WINNER TAKE ALL

By JOSHUA COOPER RAMO, REPORTED BY DAVID S. JACKSON/REDMOND

War belongs to the province of business competition, which is also the conflict of human interests. --Karl Von Clausewitz

On a cloudy morning last December, as the white light of winter picked its way across the face of Mount Rainier, William H. Gates III, chairman and chief executive officer of Microsoft Corp., rolled out of bed after a brief night's sleep. There was no time for breakfast. He showered, pulled on his customary slacks, open-necked shirt and sweater, and climbed into his Lexus for the 20-minute drive from his suburban Bellevue, Washington, home to the Seattle convention center.

Gates had been at the center late into the previous evening--an eight-cheeseburger night, he calculates--preparing what he suspected would be the most significant Microsoft announcement in a decade. For two years the rapidly expanding global computer matrix called the Internet had nagged at Gates like a low-level headache. Even as his company shipped Windows 95 last fall--an 11-million-line program that solidified Microsoft's PC hegemony--Gates worried about what lay ahead. Microsoft had long dominated the world of personal computing by providing the software that controlled the way users interacted with their machines--a program called an operating system. Microsoft's operating systems, first DOS in 1980 and then Windows 10 years later, had given the company intimidating control over PC software. And lush profits, enough to vault Gates to the top of the list of the richest Americans.

But the Internet could undo all that. Gates had been warning his top lieutenants that the Net could change everything about the way people used computers, perhaps even the fact that they needed an $89 copy of Windows to make their machines work. But he hadn't quite figured out Microsoft's proper place in the new terrain, and the company's thus far tentative market initiatives reflected that indecision.

This morning he planned to bring the uncertainty to an end. Microsoft would reorient every effort, every project and product to the new reality of the Internet. He was about to send a company with $6 billion in sales and 19,641 workers--all $70 billion worth--hurtling in that direction. In the future, as in the past, nothing was going to stand in Bill Gates' way.

Nine hundred miles down the Pacific coastline, in Mountain View, California, James Barksdale, president and CEO of Netscape, was bracing for another day of standing in Bill Gates' way. Barksdale slipped into a suit, grabbed a quick breakfast and pointed his Mercedes toward Netscape's Mountain View headquarters.

With $81 million in sales and 600 employees, Netscape had enjoyed a comfortable dominance of the infant Internet world. Its killer app was a program that made navigating the Net as simple as pointing at what you wanted to see and clicking on it. Browsers brought order to the chaos of the World Wide Web, a corner of the Net stuffed with text, sounds and pictures. Netscape's Navigator browser was the best on the market, and it had propelled the company through a wildly successful initial public offering in August. Some analysts were saying Netscape had an invincible lead in the browser business, even against Microsoft.

Barksdale knew better: more was at stake than merely the market for browsers. Netscape's Navigator had the potential to be the next Windows: a method for launching programs and calling up information not only from the Internet but also from corporate networks and even from the user's own PC. In short, it could supplant the operating system and finally break Microsoft's hammerlock on the PC industry. Barksdale knew Microsoft wouldn't--couldn't--ignore the challenge. "It was such an obvious opportunity," he says. "They weren't going to miss it. We had always expected Microsoft to get hard core about the Internet."

And so it did. That morning in Seattle, addressing hundreds of analysts and media, Gates hit a rare rhetorical high, offering up what amounted to his new digital gospel. To hammer the message home, he reached back into history, recalling the words of Admiral Yamamoto on the day the Japanese attacked the U.S.: "I fear we have awakened a sleeping giant." The crowd chuckled its recognition: it was Dec. 7, 1995, and Bill Gates was taking Microsoft to war.

The conventional army loses if it does not win. The guerrilla wins if he does not lose. --Henry Kissinger

Since his December announcement, Gates has backed up his martial rhetoric with action. Over the past three months, Microsoft programmers have released a stream of new products designed to seduce Net users away from Netscape. With everything from software giveaways to massive developer conferences (Microsoft rented 40 movie theaters on July 16 for a programmers' gathering), the gang from Redmond, Washington, has pressed home one message: Microsoft is playing for keeps.

Proof of that came on Aug. 13, when Microsoft unveiled Explorer 3.0, the newest version of its Web-browsing software. The 8-megabyte behemoth matched Netscape's franchise browser, Navigator, feature for feature, and at a much better price--free. Available over the Web, the browser notched a million downloads in its first week. Netscape stockholders voted with their feet: by late August, Netscape stock had shed half its value from a December high, while Microsoft shares approached record levels. And Gates swore the best was yet to come.

Netscape wasted little time in counterattacking. Two weeks later, on Aug. 26, company founder Jim Clark unveiled blueprints for a new software firm called Navio that will try to outflank Microsoft by putting browser software on pretty much anything with a screen and a modem. The first stop is likely to be an Internet TV, followed by a $500 network computer, online video gaming machines and Net-surfing cell phones. Organized around a powerhouse electronics alliance that includes just about everyone but Microsoft (Sony, NEC, Nintendo and IBM are supporting the venture), the company has one aim: to use the Internet to make Microsoft Windows irrelevant.

That sort of direct challenge to their business is the stimulus Microsoft employees love best. In just six months, Gates has refocused the work force onto Net-related projects, mercilessly eliminated a dozen others that were Holy Grails a year ago, and geared up an Internet-content group that will spend tens of millions of dollars this year. He has even withdrawn $1.5 billion in R.-and-D. money from a $6 billion cash stockpile Microsoft has tucked away against the sort of rainy business days they aren't used to seeing in Redmond.

The transformation that has set Microsoft on the road to a Net-based future has also completed Gates' beatification as not just a great hacker but also a world-class CEO. Microsoft's warp-speed reinvention may set the standard for information-age corporate agility. "I don't think you'd be interviewing me on this topic if we were any less nimble," Gates told Time. "You'd be writing our epitaph."

But in the odd world of high technology, a great product, a remarkable corporate transformation and market acceptance could in fact be an epitaph. The Next Big Thing, just a gleam in some undergraduate's eye today, could put your company out of business tomorrow. Andy Grove, the Intel CEO who led his microprocessor company through a series of similarly wrenching changes a decade ago, has distilled the essence of competing in a high-tech world down to a single sentence: "Only the paranoid survive." He's right. Uncertainty is the watchword of the new digital age. That's why Microsoft is throwing everything it's got at Netscape. And that's why, despite that onslaught, Netscape still has a chance.

When absolute superiority is not attainable, you must produce a relative one at the decisive point by making use of what you have. --Clausewitz

In the heady days after Netscape's ballistic IPO last summer--the stock shot from $27 to $71 during its first day of trading--Barksdale preached humility in the company's cramped halls. "I have a rule," he explains. "You can't talk about the stock price internally. We're not going to get focused on paper worth. We're going to keep building products." So the sandy-haired CEO was surprised to discover that his assistant had set up an electronic stock ticker in his office. "I said, 'Did you think I was kidding? Take that away!'" Barksdale recalls, laughing. "I try to lead by example, and here she was flashing the price."

Friends say the story is typical of Barksdale, 52, who brought his lead-from-the-front style and slow Mississippi drawl to Silicon Valley last spring from AT&T, where he was CEO of AT&T Wireless Services. Though less well known than Netscape's co-founders, Jim Clark and boy wonder Marc Andreessen, 25, Barksdale has what is clearly the most difficult, and most essential, job of the three: getting Netscape to live up to its $3.1 billion market value.

That will take some doing. Matching its fiscal reality to Wall Street's hopes will mean completing one of the great Horatio Alger stories in the history of American business. As every self-respecting teenage computer ace knows, Netscape was born in the ratty University of Illinois dorm of Andreessen, then 21, a Midwesterner who liked nothing so much as an afternoon in front of the computer, geeking out. Over a few dozen of those code-filled afternoons in 1993, Andreessen and his youthful collaborators put the finishing touches on the Model T of Web-browsing programs. They called it Mosaic, because it combined all the pieces of the Net--text, audio, images--onto "pages" that could be viewed from anywhere in the world. Mosaic was the first glimpse of a multimedia future that giants such as Microsoft had been predicting but not delivering. And it changed everything. Among the digerati, the question "Where were you when you first saw Mosaic?" is as significant as asking people from an earlier generation where they were when they first heard the Beatles.

The sound of the Net revolution quickly made its way to Silicon Valley, where venture capitalist ears are always eagerly listening for the ringing music of a great new idea. Clark, a Valley millionaire who had a similar revelation a decade earlier when he founded Silicon Graphics to build high-end graphics computers, hired Andreessen and set up shop as Mosaic Communications. Six months later, the company shipped its first Navigator browser.

A classic entrepreneur, Clark is good at founding companies but less interested in running them. His fortune derives from vision more than execution. By late 1994, with Netscape growing fast, he knew the time had come to look for help. Clark cast a rich, stock-baited hook into a pool of one: Jim Barksdale bit.

Barksdale is one of a handful of corporate managers who make up the varsity squad of American CEOs. Though less well known than Coca-Cola's Robert Goizuetta and GE's Jack Welch, he radiates the same cocksure attitude, bred from an ability to project a vast strategic vision and master simultaneously each of its components.

And there was something else: Barksdale had already been to war. He was a veteran of Federal Express's victory in the great overnight-delivery battle in the 1980s, and of Craig McCaw's end run around AT&T and the regional Bells in the cellular wars in the early 1990s. He moved to AT&T after the long-distance company bought out McCaw. Barksdale made a fortune on the FedEx and McCaw successes, but both experiences netted him something more valuable: the confidence of a winner. "UPS had more money than God, certainly more money than Microsoft," he says. "When they got into our business, we were scared to death. Well, today FedEx is a $10 billion company."

Barksdale brings that same damn-the-torpedoes sensibility to his fight with Microsoft, where he has embraced a simple strategy: Get out, get known, get in. "We're a little company. We can't afford a big television advertising campaign," he says, pointedly gazing north toward the Colossus of Redmond's $100 million ad budget. Instead Barksdale is betting the farm on Netscape's continued domination of the browser business. "Is it a sure success?" he says. "Hell, no. We work constantly at it. But this idea that somehow or other Microsoft says, 'Thank you, boys, for thinking this up. Now we'll come take it over,' is way premature."

The strategy has its risks: in a business where a six-month delay threatens obsolescence, any slip in Netscape's warp-speed browser-development cycle could prove fatal. So Barksdale's first priority is making sure Netscape's programmers, who still report to Andreessen, have everything they need to stay on top. This means rooms of their own, all the caffeine they can absorb and hours of fatherly advice about working better, not more.

Barksdale knows that long-term success lies in turning Navigator into a viable alternative to Windows. Andreessen, Barksdale and Clark (a triumvirate insiders refer to as "Marc, Bark and Clark") envision a future Netscape product that will assume many of the computer's operating-system functions, browsing local files as seamlessly as it browses today's Web.

Of course, for a Web-based OS to have any real value to consumers, it has to have programs to run. To fill that hole, Netscape has entered into a tight alliance with Sun Microsystems, a high-end hardware and software company that has developed a Net-based programming language called Java. What makes Java special (besides its name, which comes from the take-no-prisoners, drink-no-decaf corporate culture at Sun) is that it is designed to run across the Internet on any computer. For Web programmers, this means their pages can do more than just display pictures. Java-enabled sites will be able to act as word processors, telephones and even (if you have a TV receiver hooked up to your computer) vcrs. And although Microsoft has tried to embrace Java--its new Explorer browser will run programs written in the language--it is still seen more as a strategic weapon for Netscape, since Java software could one day compete with Microsoft products such as Word and Excel.

But this will matter only if Net surfers are still using Netscape software when that day arrives. And while Netscape has a dominant market share these days, the company has been seeing a little more of Microsoft's shadow than Bark likes. "We have enormous appreciation, admiration and fear [of Microsoft]," he says. "But like I tell people: I love my brothers, but I don't let them eat my supper."

To keep from losing his supper, Barksdale has retained Gary Reback, a Silicon Valley lawyer who has built a reputation for bashing Microsoft. In early August, Reback mailed a legal letter bomb to the Justice Department's Antitrust Division on Netscape's behalf, accusing Microsoft of every anticompetitive behavior short of kidnapping programmers. The charges infuriated Gates, who has already battled Justice on antitrust issues. Worse, Reback's letter played right into the media's general portrayal of Netscape as a lonely underdog facing off against a cheating giant.

That courthouse seriousness, however, is in stark contrast to the playful tone that Barksdale has set inside Netscape. Conference rooms are whimsically named after cities, prisons and characters from Dr. Seuss (the Cat in the Hat Room). Barksdale regularly works the halls, passing out praise, spinning yarns and trying to make new employees--Netscape has 700 of them so far this year--feel welcome. The company also now encourages employees to take one three-day weekend a month, and last spring Barksdale shut the whole place down for a "Netscape Escape Day."

"Work less!" may seem like a strange motto for a company that's competing with Microsoft, where employees embrace the dramatic simplicity of "When I'm awake, I'm working." But plenty of others have tried to outwork Microsoft, and Gates has beaten them all. Barksdale's approach, at the very least, could someday lead to a kinder and gentler Valley culture. Whether Netscape's bottom line can hack that remains to be seen, but Bark believes his company deserves a little breathing room. "It's not a total win-lose game," he says. "I don't think anybody is going to have the dominant position in a network-centric world like they had in a desktop-centric world. I just cannot believe that." He pauses. "And I will tell you this: it will be a shame if they do."

Keep the forces concentrated in an overpowering mass. --Clausewitz

Bill Gates had watched Netscape's explosive growth with a mixture of curiosity and concern. He knew that the Net was important, but for most of the past two years, Microsoft's resources had been strained as thousands of programmers, marketers and customer-support people slogged through nearly complete overhauls of Windows and Windows NT, the company's marquee operating systems. Gates knew he needed to get as many bodies as possible onto Net projects as fast as possible, a warp-speed course correction that would require superhuman devotion to the Microsoft mission. Fortunately, Gates knew that was one thing he could count on.

So beginning last fall he started quietly pulling the levers. Hundreds of exhausted programmers streaming back from the front lines of the Windows 95 coding effort found themselves thanked, paid and returned to the front to battle Netscape. Line managers killed million-dollar projects and refocused entire divisions in the space of hours. In one instance, the company decided it needed to jump-start an effort to develop programs in the Java computer language, a key to creating Internet applications. So John Ludwig, a rising Microsoft star who runs the Internet tools group, simply walked into a room of programmers who were working on something else and told them to stop. Microsoft appreciated their efforts, he said, but a bigger challenge had emerged. "We had to get on the stick on Java," he says. "I told them, 'Clear that source code off your machine, and start working on Java today.'" Four months later, the complex new coding was done.

The strategy guiding those difficult tactical decisions was something Gates referred to as "embrace and extend." Rather than invent a whole new service to compete with the Internet, Microsoft would "embrace" current Net standards and then "extend" them. In practical terms, that meant do everything Netscape did and then add the extra functionality that would give Microsoft a winning edge. If Netscape let Web browsers look at pages from around the world, Microsoft would let users look at the same Web pages but from within familiar applications, such as Microsoft Word. If Navigator could play sound, Explorer would play CD-quality sound. It was the sort of one-upmanship that Gates had perfected in dozens of similar software fights. Although frequently second to market with a product, Microsoft always won by outfeaturing and outlasting its competitors.

That put extra pressure on Microsoft to deliver a world-class browser. Explorer 2.0, the browser Microsoft released early this spring, was a poor second cousin to Navigator 2.0. Gates knew Internet Explorer 3.0 had to be far better. So the company began throwing bodies at the problem. From August to November 1995, the browser group grew from eight to 30 employees. By the time Explorer 3.0 was released last month, that number had risen to 800.

The results were phenomenal. Microsoft critics, who had bet that Explorer 3.0 would be no more than too-little, too-late Internet technology, were silenced by the program's sheer undeniable quality. The browser's slick interface drew on Microsoft's years of consumer-products research. And though there were flaws--it has several prominent security holes, and no Macintosh version is in sight--3.0 had brightly colored, easy-to-use buttons, was cleverly designed and ran smoothly with Windows 95. In short, the thing looked like a high-grade consumer product.

The shift in attitudes was immediate. The day Explorer 3.0 hit the streets, Netizens began to create an approving buzz. And from around the Net, where Netscape had long trumpeted its 85% market share, word began to leak back that Microsoft browsers were accounting for 30%, then 40% and by last week 60% of the hits on some servers. Though Netscape still indisputably has the larger proportion of browsers, Microsoft reported that more than a million people downloaded Explorer 3.0 in its first week online, overwhelming the company's specially beefed-up servers. And while Netscape is starting to charge more aggressively for its browser, Gates insists Explorer will continue to be free or, in his words, "priced to sell." Don't mistake that for an act of charity. Netscape's Department of Justice letter charged that a Microsoft executive told a gathering of developers this spring, "Our intent is to flood the market with free Internet software and squeeze Netscape until they run out of cash."

And beginning this Christmas, Web surfers will be able to boot up Explorer 4.0. Though the software for the program is still in its infant stage in Redmond, TIME got a sneak preview. The new browser is fully integrated with the computer desktop. Users will turn on their computers and be presented not with an ungainly collection of files and folders but with a lush desktop that includes the latest news, instant access to content from across the Web, and a specialized version of the browser that looks at both local files and data from around the Web. Explorer 4.0, when it ships, will complete the unification of the computer and the Web. It will also make Explorer 3.0, the hot technology of the moment, obsolete.

And that's where the real struggle in this battle lies. Netscape and Microsoft are competing not against each other so much as against their own obsolescence. The victor will be not the company with the best browser but the team that can run the longest on this insanely fast product-development treadmill.

In fact, the Netscape-Microsoft battle may be our first good look at information-age corporate warfare. Where industrial-age triumph was once measured in steamships built or in miles of rail track laid, victory today means little more than the right to come back and fight again tomorrow.

Nathan Myhrvold, the physics Ph.D. who is one of Gates' most trusted deputies, told Time a year ago that "no matter how good your product, you are only 18 months away from failure." He was wrong. That span has been cut to six months. And shrinking.

--Reported by David S. Jackson/Redmond

With reporting by DAVID S. JACKSON/REDMOND