(FORTUNE Magazine) – It's a radiant afternoon in late August, and Bill Gates, usually an intense, Type A kind of guy, seems happy and relaxed. His 5-month-old daughter, Jennifer, is sleeping through the nights now. He's got a tan. His other baby, Microsoft, is quickly headed just where he wants--which is to say just about everywhere in the digital land rush to stake out the Internet. As the interview gets going he can't help but brag, "Did you see all the great reviews for IE 3?" (That's Billspeak for Internet Explorer 3.0, the browser program Microsoft unleashed this summer in a giant giveaway.) Here in his office, Bill seems almost--dare we say?--laid back, as he patiently and engagingly explains, for perhaps the thousandth time, Microsoft's multibillion-dollar campaign to "embrace and extend" the Internet.

Demeanor can be deceiving. Life may seem good for America's richest man, but all you have to do is ask the wrong question to see how intensely Gates believes that he's up against the most daunting challenge of his career. Bring up competitors' accusations that Microsoft's marketing of its new Internet products ventures into a gray area verging on anticompetitive behavior, and the relaxation vanishes. "What gray area could there possibly be?" he barks. "It's a lie. Categorically." He rises abruptly and looks out the window across the Microsoft campus, where low-slung modern buildings sprawl on tidy lawns. Out there a visitor can detect a subtle sense of urgency, and many of Gates's lieutenants have the bleary-eyed look of battle fatigue. Gates turns from the window and puts a foot up on his chair. "I think you're confused," he says. "I mean, you're confusing things..."

The man is touchy because for all its money and marketing might, Microsoft is not yet on top of the biggest trend to hit computing since the PC. In just two years, a rabble of computer and software companies, led by Netscape Communications and Sun Microsystems, has turned the industry on its ear. They've transformed an obscure public network used mainly by scientists into the de facto information superhighway--without a lick of help from Microsoft. Tens of millions of plugged-in PCs running the Netscape Navigator browser now can cruise the Internet's snazzy multimedia midway, the World Wide Web, and communicate with virtually any other connected machine in the world, be it a mainframe or a Mac.

The Internet is also turning corporate computing on its ear. Net technology can be used to streamline business processes cheaply across an entire enterprise. Last fall, Microsoft, the $9-billion-a-year giant, could only stand by and watch as FORTUNE 500 companies started throwing big bucks into building companywide "intranets" using Internet hardware and software. Mostly they bought it from Netscape and Sun. According to Zona Research of Redwood City, California, annual spending on intranets has zoomed past $2 billion and will exceed $13 billion by 1999.

"What's happened is that the computer industry started to move forward without Microsoft," says Netscape co-founder Marc Andreessen. "People suddenly realized that the Internet was a much better way of doing big things with their corporate information systems than anything Microsoft or the other traditional vendors had presented to date. That's a big deal."

How big? Well, it's hard to think of a nonlethal technology that has given rise to more frenzied and hysterical claims. Consider some of the assertions that have been made about the Internet (some in these very pages) in the scant two years since it became the Next Big Thing:

(1) The personal computer is doomed. (2) Dumb terminals, hooked to mainframes, are back. (3) The applications software business is dead. (4) Lotus Notes is dead. (5) Online services are dead. (6) Advertising is dead. (7) The age of electronic commerce is at hand. (8) Internet stocks can only go up. (9) Internet stocks are a bubble. (10) The Internet presages the end of the nation state.

And, of course:

(11) Netscape is the next Microsoft. (12) Microsoft is doomed. (13) Netscape is doomed. (For a perspective on this, read what our highly respected new technology columnist, Stewart Alsop, says elsewhere in this magazine.) (14) Microsoft is poised to annex the Internet.

So before we go any further, let's step back and try to draw these battle lines with sobriety. On one side we have Microsoft, whose Windows operating system is the heart of the personal computing business--Intel's chips being the brains. Microsoft is a huge, rich, successful company run by the richest man in America, a guy known for being wily, ruthless, and smart. Time and again Gates has preached that Microsoft's biggest challenge is to become the first company in the computer industry to fully exploit successive shifts in technology. "We don't want to be like IBM," he says--a muscle-bound behemoth that, by failing to grasp the implications of the microprocessor and the PC, nearly imploded and is now more a follower than a leader.

On the other side: a group of companies, large and small, led by Netscape in a shaky alliance. Netscape has built a great enabling tool, the Navigator browser. It is a company with 1,200 employees, annualized revenues of $300 million, minuscule profits, and an outsize market capitalization of $3 billion. (See items 8 and 9 above.) Three years ago Netscape didn't exist, and if it disappeared from the face of the Earth today, its customers would hardly blink: Any number of companies, especially Microsoft, could supply virtual clones of its products.

Painted that way, the battle shapes up as a massacre in which Microsoft will crush Netscape and its cohorts, just as it did desktop-era rivals like Borland and Novell.

But the reality may be somewhat less stark. By its very nature, the Internet encourages diversity. So this time, the Everyone But Microsoft (EBM) crew--Sun Microsystems, Oracle, and a bevy of smaller outfits--may thwart Microsoft's desire to dominate everything. There's likely to be room for some very profitable software companies besides Microsoft, even if Gates succeeds in turning the Internet into the next big engine for Microsoft's growth, which he must.

Gates sees the Internet as both an opportunity and a threat. So last December 7--Pearl Harbor Day--he announced an all-out offensive. Microsoft retargeted much of its $1.4-billion-plus annual R&D budget and diverted literally thousands of programmers to Internet projects and to retrofitting Microsoft's product line to mesh with the Net. The opening salvo: Explorer 3.0, which hit the market in time to upstage the latest version of Netscape Navigator. By the end of August, users had downloaded more than a million free copies of Explorer, and the program had buzz. Says Gates: "IE 3 is only the beginning. This is all about renewing the PC."

Not to mention giving Microsoft a hammerlock on what will become a crucial part of the operating system of the future. In many ways, Gates concedes, browsers are a better way to deal with files and documents than Windows 95: "Why should what you look at when you first sit down to your computer be a bunch of graphical icons and folders and file names? That's not really what people want to see. They want to see the sales results, or their next appointment, or the urgent mail messages that have come in, or a stock ticker, or a news ticker." The current plan, then, is to revamp Windows to make it more browserlike.

Gates's real concern, of course, is neither the Internet nor the look and feel of your PC. It's maintaining Microsoft's robust growth rate, expectations for which undergird the company's towering market cap (recently $73 billion) and his own unimaginable wealth (an estimated $17 billion). Growth in the company's core businesses--Windows operating systems, Microsoft Office and other productivity applications, and languages and tools for programmers--is still very strong. In the fiscal year ended in June, Microsoft's North American sales grew by 43%, up from 19% in 1995 and 15% in 1994.

But even though he would never come out and say so, Gates is worried that the Internet has already diminished the importance of his bread-and-butter product, Windows. Recent innovations such as Sun's Java programming language beef up Netscape's browser to the point where it can act as a mini-operating system, capable of running applications without Microsoft's help. If you don't need the operating system, you don't need Microsoft. As the company most emblematic of this frightening scenario, Netscape finds itself squarely in Bill's cross hairs.

Thus, the battle for the browser market has become a technological competition the likes of which the software industry has never seen--and also one of the weirdest business wars ever. For one thing, Microsoft and Netscape essentially give away their programs.

Netscape Navigator, which today claims a Microsoftian 80%-plus share of the exploding market, is the product that put the company--and the technology--on the map. Marc Andreessen estimates some 40 million people use the program.

Like Andreessen, Gates recognized early on that a browser's potential value is less as a product in and of itself than as a vehicle--a platform, in industry parlance--for high-dollar business applications software. It's a market Microsoft has always eyed with envy.

The last thing Gates wanted was for Navigator to gain acceptance as the industry standard. So he tried what has so often worked in the past: whipping out his checkbook. Last year, says Netscape CEO Jim Barksdale, Gates floated the idea of buying the company past Andreessen, co-founder Jim Clark, and Barksdale himself. They demurred. Later, says Barksdale, Gates offered to buy a minority stake, in exchange for a seat on the board and rights to license the software, but was turned down again. Adds Barksdale: "Clearly he understood our plans from day one. But we've read the book on Microsoft, so we said no thanks." Microsoft wouldn't confirm or deny the account.

To Bill, the word no, even with thanks attached, means war. Gates decided that Microsoft would just have to dominate the browser market itself. After the Pearl Harbor Day announcement, he shifted more than 500 programmers to the development team for Explorer, and he didn't stop there. He marshaled legions of Microserfs to work on other Internet products, including such commercial services for the Internet as a travel agency and a stock-trading service, and core technologies to help Webmasters liven up their sites.

No one in the industry can really match such an attack, given Microsoft's sheer size and financial muscle. Gates also has teams looking beyond this phase of Internet growth. By 2000, he believes, the Internet will be a mass market. So he's investing heavily in the production of digital content. Microsoft will pump $250 million into MSNBC, which Gates hopes will eventually yield lively new kinds of news for distribution via the Net. "We don't expect it to make much money anytime soon," he says. "My goal for the coming four years is to make things like MSNBC's interactive-content business big enough to be a big engine of growth for four years after that."

Similarly, the company is experimenting with what it calls Webshows: interactive comic strips, text-and-graphics soap operas, even a brash combination talk show/women's magazine called Underwire. The goal, says Patty Stonesifer, senior vice president for interactive media, is to make the Internet more like TV: "We want to fill your water-cooler time at work or those dull times at home by letting you 'watch' the Internet for ten minutes without having to work to find something interesting."

Should you really still doubt how this is going to play out, one number says it all: in the coming year, Microsoft will spend some $2 billion on R&D. To put it in perspective, Microsoft's R&D budget is at least six times Netscape's annualized revenues. The sum exceeds the yearly revenues of all but three software companies--Computer Associates, Oracle, and Novell. And the lion's share of the budget is for Internet-related R&D.

Steve Ballmer, Gates's right-hand man and executive vice president for sales and support, says Microsoft settled on the $2 billion sum by benchmarking itself against pharmaceuticals companies. (The comparison is apt because software, like medicine, costs a lot to develop and little to actually produce.) "The best way to get a sense of how serious we are is to look at R&D as a percentage of gross margin," says Ballmer.

Okay, let's look. During the fiscal year ended June 30, Microsoft spent more than $1.4 billion on R&D. It reported $8.7 billion in revenues and, according to Ballmer, had $6.1 billion in gross profits (this mammoth, previously unpublished figure does not count sales, general, and administrative expenses or taxes). So the R&D investment works out to 23% of gross margin. Competitors, most of whom have much narrower margins, are lucky if they can spare 10% of gross profits.

Sounds impressive, right? The only consolation for competitors is that deep pockets don't necessarily translate into technology breakthroughs. Indeed, over the years, Microsoft has shown its fallibility with surprising frequency. The most obvious example is Microsoft Network, which competitors feared would take the online business by storm. It didn't. Then there is Microsoft Money, a ham-handed attempt at personal-finance software whose market share has barely reached 20%. No wonder Bill tried to buy Intuit.

The list goes on: Microsoft Exchange, a product designed to compete with Lotus Notes, is only now becoming a serious competitor after more than five years of trying. Even Windows 95 has fallen short of expectations. Sure, it has sold more than 40 million copies, but that's 20 million short of what Microsoft expected.

Microsoft may botch things with great regularity, and most of its technology may be derivative of innovations by others. But like Japan, Microsoft is nothing if not persistent. And eventually, it gets products right.

Unfortunately for Netscape, eventually is now. Technologically, Explorer has pretty much caught up to, and in some ways surpassed, Netscape's flagship product, say software reviewers. That's scary: Microsoft has traditionally won with marketing muscle and shrewd strategy. Having the best technology rounds out an awesome competitive package.

How can Netscape compete with such an enormous and rich foe? It has had almost no time to prepare: The company is two years old, went public a little over a year ago, and is trying to learn to manage itself as it grows at breakneck speed. True, Netscape is a sentimental favorite and has the backing of some of the computer industry's biggest names. But its allies are an intensely competitive, fractious bunch whose loyalties tend to shift with the wind.

Netscape's stock price already reflects a cooler appraisal than a year ago, when Andreessen was being hailed as the next Bill Gates. After peaking at $87 last December--the week Microsoft declared war--it's been all downhill, with shares recently trading in the mid-30s.

That still leaves Netscape with an outsize $3 billion market cap. The wish to protect that valuation may help explain why company officials seem anxious to manage expectations gently downward. They describe Netscape's business strategy in narrower, less revolutionary terms than they did a year ago. "You'd have to be crazy to be a startup with a business plan with the primary objective of beating Microsoft at their own game," says Mike Homer, Netscape's marketing chief. "It's an honor, really, to be characterized that way, but that's not what we're trying to do."

The way they now tell it, Netscape intends to carve out a lucrative niche supplying a new kind of business software--Internet-based "groupware" to help people in organizations communicate and work together more efficiently. Says Andreessen: "We think we're in a three-way race with IBM/Lotus and Microsoft for corporate intranets--not for the total domination of the Internet or to take away Microsoft's PC operating systems business. Still, this is a real serious thing because the investment in intranets is ramping so fast. If we could get a third of that business, we'd be happy."

Navigator, for all its popularity, is just a means to that end. So, too, are Netscape's Web-server products, which help powerful computers store and dish out the Web pages and applications people will use. About 80% of Netscape's sales are to corporations that buy not just browsers but server software (to date, the company has shipped 100,000 server packages) and intranet consulting services. Says Barksdale: "The browser is a tactic, not a strategy. Our objective was to get known, get into corporate America, and then move up the food chain. We're accomplishing that."

That may be, but Microsoft is taking dead aim at Netscape's server business, too, by bundling its own Internet server software free with its top-of-the-line $895 Windows NT operating system for network hubs. So worried is Netscape that in August it had its lawyers fire off an eight-page, single-spaced letter to the Department of Justice in Washington, enumerating alleged instances of "predation" and "anticompetitive behavior." Among other things, Netscape asserted that Microsoft had threatened suppliers that were coupling an inexpensive version of Windows NT with Netscape software, enabling customers to set up low-cost Websites. The Justice Department never officially responded, but Microsoft eventually backed down.

Says Barksdale: "There's a well-established pattern of behavior here and a legitimate question of how far Microsoft can go." The folksy 51-year-old, a veteran of high-profile jobs at AT&T Wireless and FedEx, adds: "I've been in regulated and unregulated businesses, and I've bought a lot of computer equipment in my day, and I know when you cross the line."

Barksdale and Andreessen understand that large losses of share to Microsoft are inevitable; they're just hoping Netscape can give up ground without the retreat becoming a rout. The company can thrive, they figure, if Navigator hangs on to half its share of the rapidly growing market.

Netscape also knows it can't survive without a lot of help, so it has formed alliances with practically the entire Everyone But Microsoft crowd. IBM, Apple, and Sun are helping Netscape develop versions of its products for their PCs, workstations, and servers. Meanwhile, Hewlett-Packard provides support services for Netscape's customers; IBM's Lotus division is working on applications written in Java and tailored to Navigator; Steve Jobs's Next sells tools that make it easier to develop sophisticated commercial Websites atop Netscape servers.

Those are a lot of partners to keep happy. The key ones are mostly multibillion-dollar companies that don't like to be told what to do. Some are already grumbling that Netscape is stretched too thin to keep its commitments. Partners like Oracle and IBM are also direct competitors of Netscape. Moreover, all the partners have their own Internet strategies, and many will probably do fine even if Netscape falters. Just to be safe, almost all have working relationships with Microsoft too.

Netscape's most crucial partner is Sun, a $7-billion-a-year company whose main business is selling high-powered workstations and network servers. Sun may well be the biggest beneficiary of the Internet boom so far. "My orders are up 35% this quarter," crows CEO Scott McNealy. "We can't even get enough sheet metal to make our boxes. Our business is cuckoo."

One reason customers flock to Sun is the halo effect from Java, a technology that in the long run may prove more significant than the Navigator browser. Java is a computer language that lets programmers create extremely compact applications called applets that can be transmitted quickly over the Net. (Netscape incorporates something called a Java plug-in so it can run the applets when they arrive.) Ultimately, Java could become the Internet's most ubiquitous technology.

McNealy is an inveterate Gates basher who likens Microsoft to a totalitarian government. But that didn't stop him from letting Microsoft license Java so it could be included in Explorer 3.0. Says McNealy: "You don't know how good it feels to have Microsoft come crawling on its knees to buy something from you, instead of the other way around."

Oracle is another key partner--well, on some days it is. The leading maker of corporate database software, with annual sales of more than $4 billion, it is championing a controversial new type of machine it calls the Network Computer, or NC--essentially a stripped-down PC that relies on the Internet to feed it applications and data (for more about the NC, see Digital Watch). CEO Larry Ellison, the industry's only non-Microsoft billionaire, hopes the NC can displace the PC in lots of corporate settings not only because it will be cheap--$500 is the target price--but because it may simplify life for information-systems managers. Many of them hate the maintenance headaches that can result when employees load their own software into office PCs.

Oracle has had plans on and off to work with Netscape to tailor browser software to the NC. Today, the plans are on. But the relationship is frayed. The notoriously outspoken Ellison asserts: "They were pretty difficult to deal with. They resisted giving part of the Internet server business to us because they are Netscape and the market has elevated them so high that they believed the Internet was their birthright. Given the insane valuations for Internet stocks, if they have less than wonderful success, the downward slope is steep indeed. It could get ugly." With friends like that...

One of the more astute observers of the Internet war is Steve Jobs, who, as everyone knows, has taken on Microsoft--usually with superior technology--and lost. He also knows how it feels to run a struggling startup--Netscape has already outgrown Next, which in its latest incarnation is doing business with both Netscape and Microsoft. Here's Steve's appraisal:

On Microsoft: "This is no ordinary skirmish. The Internet really is to Microsoft what the microprocessor was to IBM. That Bill is willing to risk the wrath of the Justice Department again shows how seriously he takes this."

On Netscape: "They've bitten a lot off. But you've got to hand it to them; they have completely distracted Microsoft from its other competitors."

So what's Steve's call? "This time, Microsoft has the technology to compete on quality. They probably don't even have to bundle Explorer with the operating systems to compete. I'd say the only thing that can stop them from being the biggest winner is the Justice Department." He pauses, then adds, "But if Netscape can run fast enough, they can be a little winner too."

So there you have it, from the eminence grise of the under-50 set. Remember the list we cited earlier of Internet hype? This much is safe to say. Item 11: Netscape is the next Microsoft. Hardly. Item 12: Microsoft is doomed. Ha! As to the fate of the nation-state, you be the judge.

REPORTER ASSOCIATE Sheree R. Curry