Let's Get This Straight: Video killed the Microsoft star

When the company rolled tape in its antitrust trial, it demonstrated its own ruthlessness.

Published February 9, 1999 8:00PM (EST)

What could they have been thinking?

That's the question Microsoft observers who thought they'd seen it all from the software superpower are now asking after the company's most embarrassing moments yet in its lengthy antitrust battle with the U.S. Department of Justice.

It was Microsoft's own lead witness last month, economist Richard Schmalensee, who -- confronted with his own law-review article from the 1980s declaring that "persistent excess profits" could indicate that a company is a monopoly -- scratched his head and asked, "What could I have been thinking?"

But the bigger puzzler in this trial now is: What was going through Microsoft's collective mind when the company prepared the videotape demonstration that blew up in its face last week? Earlier in the trial, a Princeton professor named Edward Felten had employed a software program to disable the Internet Explorer browser's code in Windows 98. Microsoft wanted to prove that you can't separate the two products -- that its browser is "fully integrated" into the operating system -- so it created a videotaped demo to show that if you "Feltenized" a Windows computer, you'd degrade its performance and experience crashes.

Before we even consider the controversy surrounding the tape, let's pause a minute. Experienced users of Windows know that, if you want to degrade its performance and experience crashes, you don't need Professor Felten's help. All you need to do is use Windows heavily for more than a day or so without rebooting, and, no matter how much memory you have installed, your "system resources" will begin to run low, programs will start to cough and stutter and, sooner or later, you're going down. Like most complex software, Windows is imperfect, and there are so many different factors that can lead to degraded performance and crashes that any demonstration of a specific cause-and-effect is suspect from the start.

Furthermore, the substance of the arguments over Felten's program is a little ridiculous: No one questions that Microsoft has the ability to weave two formerly separate products, Windows 95 and the IE browser, into a single software product -- Windows 98 -- in which you cannot separate the two functions without causing glitches and problems. At issue is Microsoft's motivation for doing so: Was the union of browser and OS an "innovation" meant to benefit consumers, or was it a calculated move to squeeze a competitor, Netscape, out of the market?

In this era of modular code, surely it would have been possible for Microsoft to provide all the benefits of browser integration in a package of discrete files that you could plug in and remove as you wished. Such an "innovation" would still offer the much-touted benefits of integration while leaving a final choice in the hands of consumers. Instead, Microsoft seems to have deliberately chosen to entangle the code of the browser with that of the OS -- precisely so that it could show, in court, that the formerly separate products are now inextricably intertwined.

Given this technical strategy -- you could call it "legal code" -- it's no wonder Microsoft went after Felten's program with such determination. What's astonishing is how badly the company botched its attack. In cross-examining Microsoft senior vice president James Allchin, David Boies, the lead government counsel, was able to show that Microsoft's video, which purported to show the "Feltenized" performance degradation on one PC, apparently stitched together footage of two or more different computers.

Though Microsoft spent the rest of the week scrambling to duplicate the test in a second videotape, it was never able to fully prove its point, and its credibility lay in tatters. Which brings us back to the question: What could they have been thinking?

Boies shied away from accusing the company of fraud: "I'm not suggesting
something nefarious happened in Redmond. All we know is the tape they put
into evidence is not reliable," he said.

The company's spin efforts throughout the week, illuminatingly, kept
drawing comparisons between the bungled video and Microsoft's software
products. "We make good software," senior vice president William Neukom
stated, "but we didn't make a very good videotape." Company spokesman Mark
Murray called the matter a
"sideshow" and said: "One of the great things about the software business
is that if there are some bugs in a first version of a product, you can go
back and fix them. So Video 1.0 apparently had a few things that became
confusing, so both sides agree we will be doing Video 2.0."

Indeed, users of Microsoft software who are accustomed to buggy
performance in early product releases might not be at all surprised that
the company couldn't get its trial videotape right on a first or even
second try. But a trial is fundamentally different from a software
development cycle. Fixing broken software is easier than restoring broken

Assuming that Microsoft didn't set out to defraud the court, one can
only imagine the following scenario among the videotapers in Redmond:
Mid-level Microsofties are assigned the job of proving that Felten's
program bollixes up Windows. The demonstration doesn't go quite as planned.
Desperate to please their bosses, they edit together clips from more than
one machine to make their case. They assume that no one will notice; after
all, they think, outside of Microsoft, people don't really understand PCs
very well, anyway.

What could they have been thinking? What people at Microsoft too often
seem to think: that they know
than the rest of the world. That they have a right to cut
corners in defense of their company's interest. That it's OK to push the
edges of acceptable behavior in their business battles.

The videotape demonstration didn't show anything useful about Felten's
program or Windows/browser integration. Instead it proved a devastating
point that happens to lie at the heart of the antitrust case itself:
Microsoft's leaders, to borrow a phrase from that other trial taking place
across town in Washington, "want to win too badly." And in order to win the
victory they're morally certain they deserve, they are willing to play a
little dirty. Long after the world has forgotten what "Feltenize" means,
and no matter what the trial's verdict, Microsoft will have to live with
the fallout from this demonstration of its ethos.

By Scott Rosenberg

Salon co-founder Scott Rosenberg is director of MediaBugs.org. He is the author of "Say Everything" and Dreaming in Code and blogs at Wordyard.com.

MORE FROM Scott Rosenberg

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

Department Of Justice Microsoft