Can Microsoft stay ahead?

Is Microsoft's monopoly of personal computer software transferable to the internet and the network businesses of the future? If so, should we worry? Geraldine Bedell considers the implications of the latest court cases
February 20, 1998

Personal computers may have changed the way we work and live, but we can't quite learn to love them. Orwellian associations of the computer as an instrument of tyranny linger on. Perhaps this is why we cannot decide whether Bill Gates, co-founder and chairman of Microsoft and the world's richest man, is a harmless and even useful nerd or Big Brother.

Washington DC is certainly taking seriously the notion that Gates, with his enormous personal fortune ($38 billion in Microsoft stock) plus his company's $9 billion cash and asset base, might be a danger to capitalism and democracy. Microsoft produces the operating systems-the computers' "consciousness"-to run 94.1 per cent of the world's PCs, and recently completed a deal with Apple which gives it control over the other 5.9 per cent. Gates has used his incredible revenues to advance into a range of high-tech businesses: cable television, satellites and news and internet commerce.

But in the last few months a powerful body of opinion has arisen which insists that Microsoft's ability to leverage its PC mono-poly should be a matter of urgent public policy. Why exactly is the battle between Microsoft and its competitors so crucial?

The browser wars

In October 1997, the US Department of Justice (DoJ) petitioned a federal judge to impose a $1m-a-day fine on Microsoft for violating an antitrust consent decree it had signed in 1994. Microsoft was "bundling in" its tool for browsing the internet, Internet Explorer, with its operating system Windows 95. Since Windows 95 is ubiquitous, this had the effect of squeezing out the competing browser, Netscape Navigator. Joel Klein, head of the DoJ's antitrust division, said: "By requiring PC manufacturers to take the Microsoft browser in order to get Windows 95, Microsoft is abusing its monopoly power, and we will seek to put an end to it."

In December, Judge Thomas Penfield Jackson found in the DoJ's favour and banned the bundling operation. Microsoft is now appealing, and has won an expedited hearing. The DoJ also claims that Microsoft is violating Judge Jackson's order by offering, in place of the bundle, in the words of a DoJ official, "a version of Windows 95 that does not work or a version that is of no interest to consumers because it's old." Meanwhile, sometime between April and June, Microsoft plans to launch Windows 98, an updated version of its Windows operating system which will be, in effect, an Internet Explorer.

At one level, what is at stake is simply the market for internet browsers: a straightforward commercial battle between two companies, Netscape (the first to market) and Microsoft (which made up what was said to be an 18-month technology lag in six months and now has a share of between 25 and 40 per cent). At another level, however, the battle is crucial because it will determine Microsoft's future. Microsoft is at a perilous point in its history. It has established itself as the dominant power in the personal computer market. But the advent of the internet-arguably the most significant invention of the second half of the 20th century-could mean the end of the personal computer as we know it. By allowing information to be sourced elsewhere, the internet could make the PC redundant, and the PC is all (at the moment) that Microsoft controls. Already, a new generation of set-top boxes-the device which unscrambles digital television signals-allows viewers to access the internet on television screens. If Microsoft sticks to the PC, it could lose the battle to supply the main home operating system for access to the internet in the future. If Microsoft wants to retain its power, it must therefore win the web war, as it previously won the PC war.

The rise and rise of microsoft

Born in Seattle in 1955, Gates is the son of a lawyer father and a teacher and civic activist mother from a wealthy banking family; he inherited a $1m trust fund from a grandmother. At the prestigious private Lakeside School, Gates scored in the top ten in the US on maths aptitude tests. He also fell in love with computers, then in their infancy.

With his friend Paul Allen, an electronics expert and sci-fi enthusiast, Gates devised computer programmes to play games, help with school timetabling and monitor the movement of traffic around cities. By 1975 Gates was studying mathematics at Harvard, when he and Allen happened to pick up the issue of Popular Electronics magazine which broke the news of the world's first kit-computer. It was called the MITS Altair 8800, cost $397 and was enough to persuade Gates to drop out of Harvard and Allen to quit his job with the computer company Honeywell. They set up Microsoft, to write for the Altair a version of the BASIC computer language.

Their big break came when they won the bidding to devise the operating system for IBM's first personal computer, which appeared in 1981. They had never previously built an operating system, so they bought one off the shelf and adapted it. IBM allowed Microsoft to keep the rights to the operating system, which it named MS-DOS and licensed to other hardware manufacturers, such as Compaq and Dell, which wanted to position themselves as IBM-compatible. Meanwhile, Apple made the mistake of restricting its operating system-generally reckoned to be far superior-to its own computers until 1995.

Initially, Microsoft's power derived from luck, foresight and other people's errors. MS-DOS was widely held to be inferior to the competing Digital Research operating system, DR-DOS, and it was 10 years before Microsoft produced an operating system, Windows, which approached the superior "look and feel" of Apple's. Its success since has often seemed the result not of superior products, but of sharp business practices. It leveraged its dominance of operating systems to develop a similar stranglehold over PC "applications," software programmes such as a spreadsheet or a word processor. It achieved success here partly by bundling applications together and selling the bundle more cheaply than its rivals could afford to sell their individual products.

While no one can be certain what the next generation of hardware will look like (except that it will be simpler to operate) one thing is sure: Microsoft intends to supply the operating system for it. The company would like to have the same choke-hold over access to the internet that it has over the PC.

Consequently, between 1994 and 1996, Gates spent $1.5 billion on acquisitions. In the past year the pace has accelerated with the purchase for $425m of WebTV Networks (a digital television set-top box maker giving access to the internet through the television screen); an investment of $1 billion and an 11.5 per cent stake in Comcast, one of the US's largest cable television operators; and an injection of $30m, plus $10m of his own money, in Teledesic, a company which will launch 288 satellites into orbit for high speed internet access. Last year Microsoft acquired key strategic technologies at the rate of one a month, including investments in video streaming technology companies such as Real Networks (allowing video clips to be downloaded faster on the internet); companies involved in 3-D animation, web site design, speech and handwriting recognition and joystick controls. Not all these investments will come good, but Gates is a gambler who can afford to take risks.

For those who want to limit the market power of companies such as Microsoft there is a difficulty arising from what is known as the "networking" effect. The value of networked products rises with the number of people using them, so that, for example, the more people communicate via e-mail, the more imperative it becomes to join in. Often two or three versions of a network technology battle it out until one begins to dominate, at which point its stock rises dramatically-no one wants to be left with Betamax when everyone else is using VHS. Brian Arthur, an economist at the Santa Fe Institute and one of the first exponents of "network effects" theory, sees it as not merely inevitable, but desirable. "Should the government prevent there being only one prize? I believe not. These prizes are for real innovation. They're what keeps the folks in Silicon Valley burning the lights at 3am. And their dominance tends to be short-lived." The drawback with network effects is that there is no presumption that the best technology will win.

Microsoft and the internet

There are, then, a number of reasons to be concerned about Microsoft. The concentration of money and influence in a single company could limit future development in a crucial area of technology. If the company gains control over access to the internet, it could restrict consumer choice. And it could, in theory, impose censorship and wield undue political influence.

The DoJ may have slowed down Microsoft for the time being, but Windows 98 will reflect the increasing importance the company attaches to the web. "Windows 98 is going to look just like a browser," says Bob Herbold, the company's chief operating officer. "In fact it's going to be a browser." And if the operating system is a browser before it is anything else, where will that leave the DoJ? Where indeed, will it leave Netscape, which is increasingly including features in its browser which make it look like an operating system? The two companies are converging on the same space.

Crucial to Microsoft's defence is the 1994 consent decree's acknowledgement of its right to sell "integrated products." In a document awash with definitions, these are, incredibly, not defined: Microsoft argues that as internet access becomes increasingly vital, not to include its browser would be like not bundling in the spell check with its word processing programme. "In one sense it's a pretty narrow legal issue," says Brad Smith, a Microsoft lawyer. "In another, there are broad principles at stake. It's not going to benefit the global marketplace for the [US] government to start dictating what goes into software and what doesn't, because technology changes so quickly. That's not likely to benefit people who use computers."

The company is moving into internet commerce-predicted to be a $80 billion to $160 billion business by 2000-with gritty determination. But the company insists that worries about mind control are absurd. Bob Herbold, the company's chief operating officer, says that Microsoft would be commercially stupid to restrict access to the content which consumers want. "This is not a content company. Do we realise we need first-rate content to attract traffic? Yes. We want to encourage content out there, which means having partnerships, but that's primarily one of us using our software skills to help these companies get up and running."

But when does commercial power become commercial abuse? An internal memo written in December 1996 and confirmed by Nathan Myhrvold, the company's chief technology officer, details plans to collect a "vig" or vigorish (bookies' slang for a cut) on every commercial transaction on the internet. The prospect is raised by Mitchell Kertzman, chief executive of software applications company Sybase, that we will end up writing two cheques at the end of every year: "One to the taxman, one to Microsoft."

Other companies seeking to operate on the internet fear that if Microsoft's operating systems direct consumers to its web sites first, they may never leave. Nor are businesses which choose to go head-to-head with Microsoft the only ones exposed. Local newspapers in the US depend on classified advertising for one third of their revenues, but most analysts expect Microsoft to use free classified advertising as a lure to draw consumers to other advertising and online sales.

Meanwhile, the potential for mind control cannot be completely dismissed. When, on 11th November last year, Microsoft posted its response to the DoJ on the internet, it proved impossible to download it, using the rival Netscape Navigator, for six hours. Microsoft subsequently said this was a mistake, but no one really believed it. Microsoft's control of data sources also gives it the power to rewrite history. For example, the entry for William H Gates III in Microsoft's online encyclopaedia, Encarta, differs from that in the print version on which it is based. The original print version describes Gates as "a tough competitor." In Encarta, these words have been substituted for "known for his philanthropy and his contributions to charity."

Who can catch Microsoft?

The market may still be a better safeguard than the law. After all, Microsoft has almost lost the plot once already. The first edition of Bill Gates's autobiography, which came out in 1995, contained scarcely any references to the internet. The revised edition, published one year later, has 130 references-essentially, the internet is what the book is about. It is not difficult to see why the mistake was made: the internet looked like an anarchic, uncontrollable environment, wholly unsuited to the hegemonic operations in which Microsoft specialises. Partly out of ignorance, partly out of arrogance, Microsoft chose instead to back proprietary networks and to attempt to create its own Microsoft Network (MSN). But when Netscape went public on 9th August 1995, its shares shot up from $13 at opening to $75 by the end of the day-an indication of how significant a product Netscape Navigator was going to be.

This time, Microsoft has recovered. But the reason it spends $2 billion a year on R&D-more than any but a couple of other companies in the world-is that Gates is terrified of missing what he calls the next bend in the road. At Microsoft, anxiety is elevated into a management philosophy. "Everyone of us gets up in the morning and is just paranoid," says Bob Herbold.

The fear is justified. Penetration of PCs has reached a ceiling at 40 per cent of homes in the US. Price is one barrier, but at least as important is the complexity of the technology. Microsoft's own figures show that 60 per cent of consumers acquiring WebTV have never owned a PC. The television is a much more user-friendly format for accessing the internet. Bob Herbold gives the PC, in the shape that we know it now, only "another nine months." But more loosely defined as "a device, held in the hands, or sitting in front of people, that has a microchip running software and enables you to get great information," it is here forever.

So who can threaten Microsoft's control of this more loosely defined computer? One possible rival is the "network computer"-as invented by Oracle and also produced by Sun Microsystems and IBM-which is designed to access the internet or an intranet without floppy disk drive, local hard disk or CD-Rom. This would solve the problem of "bloatware" (the increasingly huge chunks of software sitting unused on individual PCs) by allowing applications to be stored on a central server, to be downloaded as and when they are needed. This would eliminate much of the cost of traditional PCs. It is likely that the business and schools market will veer in the direction of simpler devices along these lines.

In the home market, however, we will continue to want the power to play video games and run CD-Roms as well as access the internet. These facilities are likely to be provided by something behaving more like a television than a PC. Microsoft is desperate to capture this market. Its subsidiary, WebTV, runs on Windows CE (for Consumer Electronics), a slimmed-down version of Windows. Microsoft would like to see the manufacturers of the set-top hardware for cable, such as General Instruments and Scientific-Atlanta, adopt Windows CE as the standard software to access the internet on television screens-trailing the possibility of a mass market far bigger than anything Microsoft has known. In December, however, the Digital Audio Visual Council, made up of the dominant cable companies in the US, including TCI and Time Warner, refused to accept Windows CE as standard. This was a setback, but the fight goes on. Microsoft will now approach the cable companies individually.

Bob Herbold says Microsoft knows computers have to become easier to use. And Microsoft's goal is to provide the software for these simplified devices (whether they are network computers or television set-top boxes). The competition here is likely to be based on Java, a programming language which enables the delivery over the internet (or other networks) of applications such as a word processing package. The attractive thing about Java is that it is not tied to a single operating system or hardware.

Sun, which devised Java and licenses it to developers, claims it will help companies run network computers, because applications rewritten as Java applets (or tiny applications) don't have to be stored at the point where they are used. (Microsoft Office requires over 100 megabytes, of which a consumer might use only two.) In a recent survey of Fortune 1000 companies by Forrester Research, 62 per cent said they were experimenting with Java and 58 per cent of information technology executives believed that Java represented a considerable threat to Microsoft. Some 80 per cent of recent software start-ups in California involve Java; more than 60 universities in the US now teach it. IBM and Sun are currently both developing chips in the Java computer language which could be used to power network computers.

Microsoft was forced to license Java and is now being sued by Sun over claims that it has created its own version of Java by fiddling with the original and thereby making it less universal. Microsoft has countersued.

One of the primary objectives of Sun's campaign is to reassure prospective users that Java will remain compatible with all types of computers. Sun claims that anyone can build their own Java applications and that it wants to submit its product to the International Standards Organisation for approval. Critics argue that industry support for Java is prompted as much by a wish to see a powerful adversary to Microsoft as by faith in the technology. Java's drawback, they argue, is speed: the translation from language to operating system requires an additional layer of software, which holds things up.

Orwellian fantasies

Historically, much hostility to Microsoft has come from disgruntled competitors and from nerds furious that Microsoft has poisoned the atmosphere of the hacker community. In truth, the anarchic atmosphere of the early days was unsustainable. But as the internet becomes more important, and as information technology meets the mass market, the implications of Microsoft's pivotal position become a matter of justified concern.

Microsoft executives will tell you-over and over again-that their company has only 4 per cent of the worldwide software business and 1 per cent of the IT business (they include telecommunications). An alternative statistic is that half the world's PC revenue goes to Microsoft. This puts it in a powerful and potentially threatening position, given the links between the PC and the next generation of hardware.

The law has relatively little purchase on Microsoft's activities, the past few months notwithstanding. Resistance may have to come from the market. Java could turn out to be the adversary Microsoft fears. Even if it isn't, sooner or later the market ought to throw up someone new (experts say we are probably the last generation to have electronic tools which do not talk). They, too, will need careful regulation; the Orwellian fantasies turn out to be remarkably prescient, although for reasons not fully anticipated. It turns out that our economic structures can barely contain a business like Microsoft.