On June 22, Microsoft officially announced Microsoft .NET (Microsoft Dot Net), the new name for Next Generation Windows Services (NGWS). Microsoft .NET will depend on XML and Simple Object Access Protocol (SOAP), a network protocol that codifies how applications can share XML-tagged data. From the user perspective, Microsoft .NET is a return to the Microsoft vision of the browser being the desktop. Microsoft .NET is also supposed to be a way to customize Web content delivery based on a user's published preferences. Key points of Microsoft .NET describe the user interface (UI) look and feel, including how it will be secured; the underlying architecture and tools; building block services for human and application communication; and client-side .NET-enabled devices. Microsoft .NET isn't due out tomorrow. Whistler, the first Windows release to incorporate .NET elements, is scheduled to be available in 2001, and Microsoft says not to expect any applications built on this platform until 2002. (I've given you a very basic explanation of Microsoft .NET. For more details, check out the white paper on the Microsoft .NET Web site or Paul Thurrot's column in the June 27 issue of Windows 2000 Magazine UPDATE.)

How much is Microsoft .NET going to affect the application service provider (ASP) space? My initial guess is not much any time soon, partly for technical reasons and partly for cultural reasons.

On the technical side, Microsoft makes no bones about the platform depending on high-speed links. Bill Gates, in his Forum 2000 speech about Microsoft .NET, said, "We're assuming that broadband becomes more pervasive. Broadband, of course, today is very prevalent in business-to-business connectivity, but it's going to take time for that to roll out to consumers, particularly on a worldwide basis."

Broadband becoming pervasive—and capable of supporting even more traffic than it already does—is a big assumption to make. Even for organizations that currently share high-speed links, the best time to use the Web is before anyone else gets to work. Microsoft .NET is designed to be available online and offline, so part of the UI relies on a tool much like Windows 2000 Professional's Offline Files. If you haven't seen this feature in action, the basic gist is that you mark certain files and folders on the file server to be available offline. When you do, the system copies those folders' contents to the client hard disk and synchronizes them with the file server whenever you log off and on. The client looks as if it's connected to the network even when it isn't. Offline Files is a great feature, but if a lot of people share the links, synchronization won't be fast.

Licensing is also an issue for a company seeking to change the way it makes money by focusing on Software As Service instead of Software As Commodity. To determine the pricing structure, Microsoft has been working with several ASPs that offer Microsoft applications, but as of this writing, the pricing structure is still up in the air.

Culturally, I wonder whether enough people will need this kind of universal access to make it a good idea to "bet the company" on Microsoft .NET. Maybe vision statements long on ideas but short on hard implementation details bring out my Luddite tendencies. People already disagree about whether the ASP model is a viable means of delivering productivity applications. Will they want to use a computing platform semidependent on Windows client devices? Will they trust hosted information? Microsoft is positioning Microsoft .NET to support consumers as well as business users. Will the consumers play?

Then I think of the effort of putting all the data up there. In the June 22 Q&A about Microsoft .NET, Microsoft CEO Steve Ballmer said, "Consumers will enjoy an integrated, secure, and easy-to-use Internet experience that will encompass everything from PC\[s\] . . . to handheld devices and smart credit cards. Their information will be available and usable across sites. For instance, consumers will be able to store all their health information online and easily take control over who has access to \[this information\]." Other scenarios include making schedules and address books available online to facilitate scheduling and contact info. I get tired just thinking about updating all that information.

Other questions abound. To name just one, the antitrust case is currently under appeal, but how does Microsoft plan to offer an integrated Internet/application desktop if the company is split up? (Although the public statements talk about interoperability, it's no secret that the environment will be "richer" with Microsoft .NET on both the client and the server.) Ballmer says only that talk about the effects of a breakup is premature, but he can't be surprised that people ask.

Don't get me wrong. From a technical perspective, Microsoft .NET can work. Neither XML nor SOAP is vaporware. We've overcome "known" physical limitations to computing a time or two before. Microsoft can resolve licensing questions. And yes, perhaps the model will really take off. But if you thought the ASP model required a radical shift in the computing mindset, Microsoft .NET requires an even greater one.

What do you think? The current Windows 2000 Magazine Instant Poll question is, "Do you believe Microsoft's new .NET strategy signals the end of shrink-wrapped software?" Go to the Windows 2000 Magazine home page and submit your vote for a) Most software will be subscription-based in 2 to 3 years, b) Most software will be subscription-based in 4 to 5 years, c) Most software will be subscription-based in 6 to 7 years, or d) Subscription-based software will never replace shrink-wrapped software.