Ten years from now, what will historians say about the impact of Windows 2000? Will this huge operating system initiative be seen as yet another great Microsoft and Intel triumph, completing the expansion of Wintel from the desktop into the highest echelons of corporate computing? Or will this renamed Windows NT be viewed as the last big gasp of an outdated form of computing - a massive, but ultimately hopeless struggle against an increasingly Web-driven world?
Although actual product destinies are rarely so black and white, market competition usually does produce a clear verdict, or at least a dominant trend. Probably the least likely scenario is that Microsoft will achieve all of its goals for Windows 2000 and that believers in an increasingly application services-driven paradigm will achieve all of theirs. Even in today's vast IT industry, it will be progressively harder to find room for both approaches. Eventually, something will give.
There is, of course, an important time dimension to all of this. Had Windows 2000 arrived 18 months ago, perhaps Linux and the whole open-source movement would never have become more than an offbeat idea. On the other hand, were Windows 2000 to be delayed another 18 months, surely many customers would choose to look elsewhere. But right now, it looks like we're in for a pretty fair fight.
My own view, however, is that Microsoft will increasingly find itself swimming against the tide. Ever since the arrival of the minicomputer in the 1960s, there has been a natural and healthy competition between the capabilities of centralised and distributed systems. The PC represented the high watermark of the latter; the Web symbolises the resurgence of the former. The combination of order-of-magnitude leaps in bandwidth, the open-source movement and the economics of a billion-user Web system will inevitably position Windows 2000 as part of an increasingly obsolete order.
More specifically, I believe that the proliferation of small businesses, the outsourcing of e-mail and the spread of Web hosting will turn customers away from running their own servers. The Internet will help launch millions of new businesses all around the world and these companies will find it relatively easy to skip most of the Wintel server experience and go straight to application service providers.
E-mail outsourcing will lead the way. Once application service providers round out their messaging service offerings, and once the Web gets consistently fast enough, running your own e-mail system will be largely a waste of time. And once e-mail goes, so will discussion groups, knowledge bases, document distribution and, eventually, most forms of text-based computing, including voice/text integration.
Similarly, the unstoppable shift toward outsourced Web site hosting will drive application service provider usage on the data side of the house. Customers won't care and may not even know which operating system their service provider is using. Additionally, these service providers will, at least in theory, be well-equipped to assess the various virtues of W2K, Linux, Solaris et al., resulting in real operating system competition based on features and pricing.
This pattern will inevitably spread to larger organisations. However, this doesn't mean that businesses of all sizes won't or shouldn't aggressively adopt Windows 2000 today. There are many environments where this will make perfect sense, or even be necessary. My only caution is that, despite all the hype, set as mostly a tactical implementation, not a long-term strategic direction. The future of computing lies elsewhere.
David Moschella is a regular contributor to IDG publications. e-mail him at email@example.com.