The distributed future will combine the best of both worlds

Approximately 20 years ago, the first PC transformed the modern workplace. The PC was cheap enough to become pervasive, and because PCs could run applications locally, much of the technological innovation of the past 20 years has occurred in the PC space.

Well, this scenario is Microsoft's version of history, anyway. Companies such as Oracle and Sun Microsystems might argue that the PC has caused as many problems as it has solved, largely because of the sheer number of machine-configuration possibilities.

The OS Is the Platform
Another interesting phenomenon emerged with the PC: Corporations and consumers will pay for the OS that drives the hardware. Consequently, OSs, which had been bundled for free with hardware, became the platform. Microsoft has rarely passed up a good business opportunity, but the company needed the retail success of DOS 5.0 and Windows 3.0 to become a platform supplier. Today, Microsoft's Windows and Office platforms represent more than 90 percent of the company's revenue. So, Microsoft has an interest in keeping these platforms successful.

Microsoft's financial standing aside, good reasons exist for OSs and even application suites to develop into platforms. For example, users demand powerful new features and enabling technology, and developers can provide this technology more quickly for software than for hardware. This reality doesn't negate hardware's importance, but software has driven most of the growth in the PC industry.

Now, in 2001, the unfettered growth of the software market has caught up with us. Most companies don't regularly upgrade their hardware, and the latest versions of Windows and Office don't run well on the Pentium 75 systems that seemed so powerful in 1996. For many companies, the treadmill of perpetual software and hardware upgrades doesn't make sense anymore. Many solutions that offer a longer lifetime and lower cost have appeared, and some of these solutions come from surprising sources.

Distributed Computing Rules
In one corner, we have Microsoft, which must support the hundreds of millions of people who have invested in Windows desktops and servers. But Microsoft now faces newly reinvigorated big-iron companies such as Sun, who can point to the distributed computing model that the Internet has made possible and say, "We were right all the time. Told you so." And sure enough, in the other corner, Sun, Oracle, and other companies offer bulletproof, scalable distributed solutions that can handle the most demanding tasks. But these companies aren't excited about rich clients, the province of Microsoft and Windows. The big-iron solution gives users a thin client (possibly Java-based) and keeps all the processing power on the server. And guess what? That kind of product is what Sun and Oracle have been making for 20 years.

But Microsoft is no less cynical. To continue the success of its core platforms, Microsoft is retrofitting them to work in a distributed environment—not engineering a new solution (which would probably look suspiciously similar to Sun's products). But this situation isn't entirely bad. Microsoft has the popular vote: Most people would never give up their desktop-based power for a thin-client solution. Offering rich, integrated solutions such as MSN Explorer or Office on a thin client is impossible. Or is it? Office Online delivers Office over a Windows 2000 Server Terminal Services network to thin clients, and Office Online uses a computing model that's oddly reminiscent of Sun's work to deliver Microsoft's platforms to thin clients.

In the end, Microsoft's ideal solution—powerful servers coupled with rich Windows-based clients—will probably prevail. But don't dismiss Sun and its peers. The key is to maintain innovation on the client as we move to the distributed future. Just don't call it the post-PC era. I think that, for better or worse, the PC is going to be with us for quite a while.