Do the paranoid survive?

Judge Jackson's opus on the browser wars portrays a Microsoft terrified by middleware.

Published November 6, 1999 6:30PM (EST)

The list of companies whose abuse by Microsoft is chronicled by U.S. District Judge Thomas Penfield Jackson's findings of fact in the government's antitrust case includes most of the big names in the U.S. computer industry. In the main, however, Jackson's opus is a comprehensive history of Microsoft's efforts to destroy Netscape, the one software company that Bill Gates and his lieutenants saw as a serious threat to Microsoft's dominance of the personal computer world.

Page after page of the document covers in detail the "how" of Microsoft's campaign. Jackson recounts how Microsoft, aiming to make Netscape's Navigator "a jolting experience," essentially sabotaged Netscape software by building incompatibilities into Windows; how Microsoft baited and then threatened computer manufacturers to steer clear of Netscape with promises of cash and threats of cutting off access to Windows; how Microsoft "integrated" its browser with the operating system to make Netscape irrelevant. Reading his account, it is clear that the tortured locutions, bizarre demonstrations and striking lapses of memory that Microsoft executives resorted to through much of the trial availed the defense little.

Less obvious in Jackson's findings, and in many ways a more interesting question, is the "why." The story that comes through in the findings is that Microsoft, and Gates specifically, might have seen Netscape as more of a threat than it ever really was. By every conceivable means, Microsoft tried to smother a competitor that posed only a very hypothetical danger, far in the future. In fact, more than anything else, it is Microsoft's paranoia that winds up buttressing the government's theory of the case.

The browser conundrum: not part of the operating system, but a threat to the operating system.

Few images of the Microsoft's trial resonate as strongly as those of Gates, as well as subordinates like Daniel Rosen, denying in the most sense-defying terms the very existence of such a thing as a "browser." The denials elicited much merriment in the press. Microsoft tried to argue that it could not have tried to seize control of the "browser market," because there was no such thing, and its own browser (even in this sentence it's almost impossible to avoid the word) was just a component of the operating system.

Jackson explicitly rejects this argument. In Paragraph 154 of the findings he says unequivocally that Web browsers and operating systems are "separate products."

OK, that's easy. But here's where it gets harder, because to understand both Microsoft's motivations and Jackson's decision, it's crucial to accept that while they might be separate products, Netscape's browser and Microsoft's operating system do have one big thing in common: They both include applications programming interfaces (APIs) that let other software developers build programs "on top" of them.

In this way, the "browser" is distinguished from other, simpler applications. Most applications -- including browsers -- rely on the operating system to perform many low-level tasks. They can do this because the operating system's APIs effectively let programmers use code in the operating system instead of writing it from scratch.

Guess what? The browser's APIs let software developers do much the same thing, except that instead of relying on code in the operating system, they rely on code in the browser. Jackson thus calls browsers "middleware" in Paragraph 68 -- they rely on the operating system, but in turn let other programs rely on them.

"Microsoft," writes Jackson, "was apprehensive that ... there would arise a substantial number of full-featured applications that relied largely, or even wholly, on middleware APIs."

That's the tricky part: A browser is not "part" of the operating system, but Jackson believes that in many ways it's a lot like an operating system.

Does Bill G. really see that far into the future?

It's worth mentioning how weird the motivation that Jackson ascribes to Microsoft really is. There aren't any PC applications that rely wholly on Netscape's APIs. If there were, they would be able to run without changes on any computer with Netscape Navigator installed. "It remains to be seen," Jackson writes in Paragraph 28, "whether there will ever be a sustained stream of full-featured applications written solely to middleware APIs."

So Jackson, writing in 1999, notes that it still remains to be seen whether, in fact, there will ever be applications that actually use middleware like Netscape Navigator to bypass the operating system. And Microsoft was supposed to have been worried about this threat back in 1995. It would be hard to believe ... if it wasn't so obviously true.

Jackson believes that Microsoft believed that the proliferation of Netscape browsers could at some point in the future make its operating system less important. And that, weird as it is, is exactly right.

Here's a quote from one of the many Bill Gates e-mails introduced into evidence: "They [Netscape] are pursuing a multi-platform strategy where they move the key API into the client [browser] to commoditize the underlying operating system." In other words, Gates is saying that by creating its own APIs, Netscape was threatening to make Microsoft's operating system superfluous, and thus turn the enormously lucrative Microsoft operating system into a low-profit commodity product -- like the interchangeable beige boxes made by second-tier PC makers.

Wait a second. It sure sounds like there are a lot of steps from here to there. But that's what Gates wrote.

The reason it should immediately strike us as a little strange that Gates should be so worried is that no matter how many APIs Netscape's software had, we have a lot of trouble imagining Windows owners suddenly being able to get rid of their operating system.

Microsoft's not so shaky monopoly.

In its last fiscal quarter, Microsoft took in $2.25 billion from selling operating systems and upgrades -- a rate of about $9 billion a year. Those huge numbers grow reliably every year because Microsoft effectively exercises a monopoly on desktop operating systems. Jackson concludes this in Paragraph 34, writing that Microsoft's customers "lack a commercially viable alternative to Windows" on Intel-compatible PCs. Arguably, Jackson could have spent more time considering Linux and other flavors of Unix. Arguably, too, talking about "Intel-compatible PCs" is a strange way to define the market; Microsoft would have equal or greater power if its operating system ran on the PowerPC chip. But these are quibbles.

Jackson believes that Microsoft's monopoly is very easy for the company to maintain. The reason for this is what he calls the "applications barrier to entry" in the operating systems market: "Software developers generally write applications first, and often exclusively, for the operating system that is already used by a dominant share of PC owners," he says in Paragraph 30.

The "applications barrier to entry" is for Jackson the fulcrum on which the whole case rests. The term, and much of the theory of the case, comes from Franklin Fisher, the government's star economist.

A much simpler way of saying this is that no one will use an operating system unless there is plenty of software, and no developers will write software unless there are plenty of users. Jackson's premise is that the "applications barrier to entry" in operating systems is so big that no one can successfully introduce a new operating system. Ergo, the Microsoft monopoly.

If only. There's nothing wrong with Jackson's theory, though some observers will probably argue that he has to go through some hoops to explain why the Mac operating system's 12,000 applications are not enough to overcome the applications barrier to entry. The problem is that it's clearly incomplete. There's a lot more buttressing Microsoft's monopoly than this.

As striking as anything in Jackson's findings is what's not in them. For one thing, Jackson never analyzes the contribution that the cost of switching operating systems makes to Microsoft's monopoly. Even if another operating system had 70,000 applications (the number that Jackson says are written for Windows), it's likely that few people would be rushing to toss out their Windows machines. They already have too much invested in Windows; why risk making all your old files and software incompatible with your new operating system?

There's more. Microsoft's dominance of operating systems is
perpetuated by a dominance in applications. The contribution of Office --
the absolutely ubiquitous suite of productivity programs -- to maintaining
Microsoft's operating-system monopoly never gets a mention. (A long way into
the document, Jackson does discuss Apple's fear that Microsoft could kill
the Macintosh by dropping Mac Office, but he never takes the logical next
step of pointing out that Microsoft can use Office as a club against
any competing operating system.)

This is too bad, because it's the Office suite that really puts the point on the "applications barrier to entry." To displace Windows, a new operating system doesn't just need a lot of software developers to write programs for it. It needs Microsoft itself to release a version of Office. Fat chance.

Throwing out the baby to save the bathwater

It's really only in the context of Microsoft's paranoia about Netscape positioning its own software as an alternative to the Windows platform that you can make sense of the infamous June 21, 1995, meeting at which Microsoft tried to pressure Netscape to "split" the browser market.

The clear impression we get in retrospect, with the trial behind us and Jackson's findings at our fingertips, is that representatives of the two companies were talking largely at cross purposes. Microsoft wasn't really asking Netscape to split the browser market, it was effectively demanding that Netscape cede the biggest part of its consumer business.

"It is unclear," Jackson writes in Paragraph 88, "whether Netscape's acceptance of Microsoft's proposal would have left the firm with the ability to survive as an independent business."

In Jackson's telling, at least, Microsoft executives seemed to have walked in offering Netscape a preposterously bad deal -- a deal that would seem rational only if you accept the idea that the only value to anyone in the browser market in the first place comes from the far-in-the-future potential of turning it into an operating system that competed with Windows.

It's worth wondering, in fact, what would have happened if a less paranoid Microsoft had walked in with a more attractive deal -- making it plausible for Netscape to continue making (and even selling) browsers. It seems quite possible that Jim Barksdale and Gates would have shaken hands on it, and the Justice Department would never have been the wiser.

Instead, having failed to secure a deal with Netscape, Microsoft than went hog wild making deals to destroy Netscape by packaging Internet Explorer with online service providers and making it the "preferred" browser for major Web sites.

The irony was that for the sake of protecting its monopoly position from a potential threat in the distant future, Microsoft was giving away an awful lot that is valuable today. For one thing, Microsoft sacrificed visibility for MSN, its own online network, in favor of its partner content sites. In fact, the much-hated Windows98 channel bar might simply have been an MSN bar were it not for the imperative of winning browser market share. But to sell -- no, to give away -- more browsers, Microsoft "opened up the Windows box" (Paragraph 277).

Giving away positioning to potential competitors in the content business is one thing, but it gets even worse. Microsoft even gave America Online premier positioning right on the Windows desktop. It is true that America Online was already being packaged with many, if not most, home computers sold, but it's hard to imagine the two companies cutting such a deal today. In the meantime, since the two companies teamed up in 1995, MSN has languished -- and, partly thanks to its ubiquity, AOL is not only by far the biggest Internet access provider, but unquestionably the only consumer software company with the power to challenge ... Microsoft.

By Mark Gimein

Mark Gimein is a staff writer for Salon Technology.

MORE FROM Mark Gimein

Related Topics ------------------------------------------

Department Of Justice Microsoft