Future of Micrsoft?

Predicting the future is fun, but it's not very easy. But I'm feeling prophetic at the moment, so here goes nuthin'. In short, I believe Microsoft's fortunes will eventually decline. I believe this for three basic reasons.

Microsoft Has Grown Too Cocksure

Microsoft has simply lost touch with how to compete on the basis of quality and user satisfaction, if they ever had the touch to begin with. They compete instead on the basis of their influence with computer manufacturers, software developers, and their marketing savvy and power. Their main products--their operating systems and their office suite--have been stagnating for some time from the standpoint of technical superiority.

What brought Microsoft's success in the first place was their dominence (whether legally or illegally) in the DOS and PC market. They promised and delivered compatibility and interoperability of applications from their entry into the market in the early 1980's, through the breakup with IBM and into the 1990's when they introduced and won the office suite software market. Beneath it all, however, their operating system innovation has been moving at a snail's pace, and their office suite success is difficult to improve on from here.

What remains for Microsoft is to migrate all users from Windows95 or 98 to Windows NT. NT represents their best hope for a competitive operating system, and yet, it still suffers many drawbacks when compared to the open-source development model (more on this later).

Perhaps the most damning evidence that Microsoft has forgotten how to compete on technical superiority comes from their own performance at a recent Comdex: the famous BSOD (Blue Screen of Death) at the Windows98 preview, the video clip of which was broadcast on CNN and made available for download in QuickTime format. Did it appear to you, as it did to me, that the demonstrator didn't even prepare for his demonstration, that he winged it before live TV cameras with his boss, the richest man in the world, looking over his shoulder?

And yet, this possible scenario is consistent with the Microsoft attitude I hear about from friends in retail who complain about their Microsoft representatives. Little support, a lot of hassels, and the distinct impression that they (Microsoft) are doing them (the retailer) an enormous service by deigning to allow their products to be sold through them. A similar attitude seems to echo from people requesting support by telephone: either pay a lot of money or wait on hold long periods of time, and then simply be told to reinstall your software. Does this sound familiar? It does to me, and if this had happened in 1986 on so large a scale, I don't think we'd be running Windows today on so many personal computers.

What Goes Up Must Come Down--Or Into Orbit

People learn from their mistakes; that is one of humanity's most prominent traits. We never stop growing and learning. We, as a species, often make some fascinating detours along the way, but we never stop moving in the attempt to become better.

Microsoft, on the other hand, has slowed down. They have grown so large that they have shifted from trying to promote interoperability with computing standards to innovating their own alternative standards in the hopes that they will become the de facto standard. Witness ActiveX, OLE, and extensions to mail and web servers by Outlook and FrontPage when standard protocols are either comparable or superior.

Microsoft does not appear to be leaving everyone in the dust from a techological perspective. They don't appear to be going ``into orbit.'' In an age of microkernels and distributed processing, Windows NT still lumbers along, sucking up system resources like an ill-mannered leviathan. (NT is a microkernel, but it doesn't seem to act like one.) In an age of open-source collaboration where bugfixes arrive at lightening speed for free operating systems and software, NT is still subject to Denial of Service attacks and kernel lockups for problems that have been around for a long time. NT is a large improvement over Windows95, but even it doesn't promise the compatability that Windows 95 offered to users of Windows 3.1. Microsoft appears to be forgetting their own rules of competition--or perhaps they believe they can afford to change the rules.

The Open-Source Development Model is Technically Superior

Traditional business wisdom calls for creating bright innovations and then selling the licenses to those innovations for a big profit. That model made a lot of sense before it was possible to share information at the speed and depth of today's global networks. It made sense when putting a lot of smart people in a great big building and focusing their attention on a single problem or solution was the only way to work. But now, there are better and smarter ways to work.

Linux sales by Caldera, Red Hat, Infomagic, and others, grew faster last year (year over year) than sales for any other operating system, making the Linux operating system market sector the fastest growing opsys market. More than half of all mail servers and web servers run free software such as Sendmail and Apache. Supercomputers built from clusters of personal computers and running Linux are beginning to match and exceed the performance of commercial supercomputers from IBM, Cray, and Thinking Machines, and are doing so at a fraction of the cost. Installation routines and desktop environments have become so much easier for Linux than they were that home users are just beginning to flock to it as an alternative to Windows. The momentum only appears to be growing.

The best solutions to a computing problem come from software that works for its users better than alternatives. The best way for developers to improve software is for as many talented programmers as possible to work collectively and cooperatively on it, and for those programmers to work with as large a bug-fixing user base as possible. The open-source paradigm captures this dynamic by requiring skillful orchestration of programming talent and by incorporating widespread user feedback into the development model. This cannot happen under the proprietary software development model--at least not without spending vast sums of money in the process. With hundreds of bug hunters working part- or full-time using the software under the open-source model, all bug-fixes eventually become very shallow, making the software better than any commercial alternative. The success of Linux is the key case in point. Improvements to open-source Netscape, such as strong encryption being incorporated back into the source code, patched memory leaks, and speedier toolkits replacing Motif toolkits, have already started to promise it technical superiority over commercial alternatives.

Conclusion

History rarely conforms to the vision of fortune tellers, and some unforeseen outcome will probably come into play despite my fondest beliefs. What probably has enabled Microsoft to prevail this long has been a combination of highly skillfull marketing and lawyering. It will take time for the corpus of legal precident to catch up to them, but hopefully, at the very least, Microsoft products will have to become a whole lot better than they are now, because the standard of excellence only goes up over time, unless market competition goes away.