“Microsoft Windows Forever and Ever?” (Windows Magazine, June 1996)

[Here’s another article I published back in 1996, this one predicting the problems that Microsoft would face as it continued to advance the Windows operating system.  While I didn’t anticipate in this article the rise of post-PC devices, nor the return of Steve Jobs to Apple and subsequent transformation of NeXTstep into OS X, I was spot-on in predicting that 95% of desktops would still be running Windows in 2006; it’s pretty much that same percentage (91%) today. And Microsoft is, in fact, struggling with its own Windows legacy; note that 1/3rd of all desktops are still running Windows XP, an OS scheduled for end-of-life in just seven months.]

Is there such a thing as too much success?

Face it: Microsoft is on a roll. The vast majority of computers being used today–close to 300 million–run MS-DOS, Windows 3.x (with MS-DOS), or Windows 95 (with MS-DOS). The vast majority of new systems being shipped–at least another 50 million a year–come with Win95 pre-installed. Other operating systems are losing ground and market share. At this rate, the year 2000 will dawn on half a billion or more computer systems worldwide running MS-DOS and/or Windows, along with many of the applications we use today.

And as all this happens, the rest of the industry–indeed, it would seem the rest of civilization–increasingly focuses upon Microsoft technology. Heck, even IBM finally threw in the towel and is officially supporting Windows, despite owning (and having sunk a few billion dollars into) OS/2, which most independent observers consider superior to the various Microsoft offerings. Apple, attempting too little too late, is too busy trying to survive to represent a threat. The Windows/MS-DOS standard is expanding unimpeded through the world.

And that, ultimately, could be Microsoft’s bane.

You see, the dominance of Windows/MS-DOS is just the latest example of a common phenomenon. It seems that the first sufficiently adequate technology in a given sphere usually gains broad acceptance and entrenches itself. There may be some contention for standards early on, but that eventually settles out. And once the solution gains acceptance, the rest of civilization reinforces it–through market forces, legal standards, competitive pressures, and economic incentives.

Once the technology is entrenched, the focus is then on refinement and slow upgrading of the existing technology, not on radical innovations and wholesale replacements. This has been shown time and again in major areas of applied technology: communication (phones), transportation (cars), and so on. And the same thing is happening now with information technology.

Think about it. The most broadly installed and used operating system in history is MS-DOS, a fifteen-year-old knock-off of an even early OS, CP/M. Microsoft’s greatest marketing challenge with Windows wasn’t competition with OS/2 or MacOS. It was getting MS-DOS users to upgrade to Windows, and that didn’t even begin to happen until Windows 3.0, the fourth release (after Win1.0, Win2.0, and Win386). Even then, the growth in the Windows 3.x installed base came less from users upgrading than from new computers being shipped with Windows pre-installed.

Microsoft has faced the same problems with Windows NT and Windows 95. Market acceptance of WinNT has taken longer than expected to build. Meanwhile, both corporate and home users of Windows have been slow to upgrade to Win95. Again, growth in installed base has come largely from systems shipping with Win95 installed.
This, then, is the crux of Microsoft’s problem: users are and will continue to be reluctant to upgrade. The legacy base for Microsoft (and third-party supporters) will grow and will be slow to change over. And that base will be so mind-boggling huge within the next few years that any desire on Microsoft’s part to introduce any significant innovations in operating system technology will be seriously hindered.

To date, Microsoft has managed to sidestep this problem by bundling new OS releases with new computer systems. But that will yield diminishing returns over the next 10 years as the number of computer systems in use approaches a measurable fraction of the total human population on earth. The year 2006 will probably see somewhere between one and two billion computers worldwide, and 95% or more will be running Windows.

The sheer scale of adoption of Microsoft technology will provide tremendous momentum to keep thing moving in the same direction. The investment in hardware, software, market standards, training, business process, development expertise, custom applications, and deployed environments all argue against any broad changes, even those introduced by Microsoft. That investment grows year by year and will dominate more, not less, as time goes on.

That is the final irony. Like it or not, Microsoft is stuck with Windows for the long haul. Attempts to replace it at some point in the future will run into all the same problems that any competitor would have, except–maybe–having to market against Microsoft. Efforts to change will face diminishing returns. And even a brand new OS, written totally from scratch, would have to support so much of Windows’ behavior that it will be chained down by its legacy.

As impossible as it might seem, Windows may still be dominant in 2025, nearly 30 years from now. It may look and work a bit differently, just as phones, TVs, and cars from 30 years ago do, but the principles will be the same. Our grandchildren will wonder about the quaint relics of terminology and work flow (when was the last time you actually put gloves in a glove compartment?), but they’ll be able to clearly see the inheritance from MS-DOS/Windows to whatever they use. And they’ll accept it as part of everyday life, not knowing what better dreams we–and even Microsoft–had for the future.

We just didn’t realize we were already there.

========================================

[UPDATE: 09/18/2013] A few thoughts after the post.

First, and this is important, the (current) chart above is only for desktops and laptops — traditional PCs in other words. It does not include mobile devices such as smartphones and tablets. If you include those, then Microsoft’s share of installed base goes down from 91% to 35%.

Second, some Windows pundits are getting excited about an increase in the Windows 8 adoption rate and a decrease in the Windows XP market share. I’m not sure whether they take into account that it takes a special effort to buy a Wintel laptop or desktop without Windows 8 pre-installed.

Third, look carefully at the chart above and look at the (tiny) wedge for Windows Vista. Vista was released in 2005 and was Microsoft’s flagship OS until Windows 7 was released in 2009, four years later. This means that most new PCs and laptops sold during that four year period had Windows Vista pre-installed. Yet Vista’s market share (by most accounts) never got above 20%. Now, Vista’s market share is down to just 4%. This means that not only was the market reluctant to adopt Vista, but it also abandoned Vista wholesale once Windows 7 (which now has a 45% market share) came along. I suspect were Microsoft to release a Windows 9 that was a new-and-improved version of Windows 7, we’d see the same phenomenon with Windows 8 — it would shrink to a tiny wedge. But I’m not sure Microsoft has that option open to them; they may well have bet too much on the Windows 8 codebase without leaving a path to do such a version of Windows 9. In which case, they may be truly hosed.

Your mileage may vary. ..bruce w..

About the Author: bfwebster

Webster is Principal and Founder at at Bruce F. Webster & Associates, as well as an Adjunct Professor for the BYU Computer Science Department. He works with organizations to help them with troubled or failed information technology (IT) projects. He has also worked in several dozen legal cases as a consultant and as a testifying expert, both in the United States and Japan. He can be reached at 303.502.4141 or at bwebster@bfwa.com.