Do Not Go Gently Into That Microsoft Night

The Microsoft Gorilla Today, the safe money is on Microsoft. We all know that. One of the least enviable positions for a technology manager to be in is that of advocating a product or technology that competes with a Microsoft "standard." For good or bad, many of our colleagues have decided to buy into the Microsoft vision, desktop to server. Given the uncertainty that characterizes the technology landscape, it's a fairly rational course of action. At the desktop, you can't really fight Microsoft's monopoly position in operating system technology--a fact that has transformed the term multiplatform into Windows NT and Windows95. At my own site, I have advocated standardization on Windows95 with an eventual migration toward NT. We simply cannot afford to be positioned with the minority that will need to beg client/server developers to support their client platform. Some continue to plead, asking me to take a stand against tyranny and demonstrate the courage of my convictions. I'll pass on that one, thank you. On the server side, it's a tougher call. We've stuck with NetWare and Unix, for the most part, but we're always on the lookout for opportunities to selectively deploy NT (to gain expertise and hedge our bets a little). Some see this as a transition to the inevitable dominance of NT as our only supported server platform, but I am not convinced that the market will run this way. NT is inferior to NetWare in file, print and directory services, particularly for large organizations. NT is also inferior to Unix in application server scalability. Just try to convince talented NetWare or Unix system administrators that they should abandon their preferred platform, toss out the stability and performance that come with those systems, all to position the organization more strategically. Not the Applications, Too Even if you're willing to concede the desktop to Microsoft and find a prominent role for NT on the back end, can you afford to concede everything to Microsoft, including your application environment? I'm not talking about Microsoft Office here, but rather, the application infrastructure that will define the next generation of three-tier client/server computing. I'm t alking about your browser, messaging system, Web server and all the APIs and middleware that will enable Microsoft to extend its dominance to a new sector of the industry. It is arguable that Microsoft has a better browser than Netscape, with greater stability and functionality. But that's not why Internet Explorer is gaining market share. Fear and intimidation are driving this trend. Technology managers are worried about becoming overdependent on a start-up like Netscape as a strategic vendor, and they are concerned that tight integration of browser functionality into Windows98 and NT 5.0 may leave them with a mess if they don't simply concede this battle as well. FUD--fear, uncertainty and doubt--drive the Microsoft business plan, much the same way it once did for IBM. On the back end, I am not convinced that Microsoft has the ability to execute the delivery of world-class server applications, be they database, messaging or Web servers. Microsoft's track record in this area is spotty at best. It's tru e that Microsoft has a vast repository of R&D resources, but its record in the desktop application and OS arenas lends some credibility to the notion that software developed by an army of programmers often results in a level of complexity that doesn't truly serve customer interests. Worse, it's clear that Microsoft's support for open standards is basically a last-resort strategy, one to be used only in those cases where it cannot drive proprietary Microsoft technologies by sheer force of its position in the marketplace. Maintaining Multisourcing Flexibility Many network managers seem to feel that remaining loyal to a single vendor will help their organization in the long run. Although it's true that a single-vendor environment does provide some advantages in areas such as training, support and maintaining spare components, it also diminishes flexibility and leverage. I used to believe that vendors would reward my loyalty and pay more attention to my problems, but the opposite is often true. Vend ors have a natural tendency to target accounts that have the potential to become new sources of revenue or represent a threat to existing revenue. Loyalty gets you very little. The lesson here is clear. Wherever a technology can be considered a commodity, maintain multiple sources. If you find that a relatively small portion of your expenditures is for commodity-class technology, perhaps you ought to re-examine your priorities. This doesn't work for everything, because many software products aren't commodities. In those cases, you're that vendor's captive. Just don't let yourself be held captive across the board. Dave Molta is director of network and system services at Syracuse University. He can be reached at dmolta@nwc.com.

Other Columnists Corporate View By Robert Moskowitz On The Edge By Art Wittmann