Last time, we discussed the background of Microsoft's .NET strategy in broad strokes, which is interesting from a historical standpoint, but it doesn't provide much in the way of practical use. .NET isn't a vaporware announcement about some future set of products and services, as some have suggested. Instead, it's simply the formal declaration of Microsoft's business plan going forward. As such, .NET has a set of goals, products, and a rollout timeline. I'd like to discuss these this week.

The goal of .NET is to transition Microsoft—and the rest of the industry—to a new model for software development, software deployment, and revenue growth. Since its inception, the PC industry has relied on a retail sales model wherein users purchase products—hardware and software—and install these products at home or in the office. Mail order and the Internet haven't really changed this model inherently, although they've made the process simpler. But with the ubiquity of the Internet, it's now possible to obtain, install, and update software online. We're comfortable with this process for small applications and product updates, and the Linux community already lets you install the entire OS over the Internet. But with .NET, software will transition from a desktop-based focus to one where that software's services are exposed over the Internet, rather than from files on your hard disk. To sustain and grow revenue, companies will look to subscriptions.

Obviously, this model won't work for all applications. For word processing, for example, some logic will have to reside on the local system, at least for the foreseeable future. But even products such as Microsoft Office are in transition. The next version of Office will offer a subscription model for consumers that will let you purchase a 1-year license for the product at a substantially reduced cost. This offer will open up this expensive suite to a wider audience and give Microsoft a steadier income over time. Eventually, Microsoft will be able to ship fewer and fewer copies of Office on CD-ROM and provide much of what makes up the suite as subscription services over the Internet.

Microsoft has big plans for the subscription model. The company will provide a number of "building block services"—including email, calendaring, instant messaging, Web site publishing, document authoring, and the like—for free. The company also plans to offer some basic subscription services, such as personal information management, at no charge. On the subscription side, Microsoft is looking at a number of premium services that it feels people will be comfortable paying for. The company has yet to identity these services, but it's not hard to speculate what they might include: online buying services, investments and banking, and other lifestyle management services, for example.

To transition to a subscription service model, however, Microsoft will first need to retool its entire product line. Predictably, this begins with Windows. Windows 2000 was the first version to ship with native support for XML, the basis for .NET. But calling Win2K a .NET product is a bit of a stretch, so Microsoft will phase in more .NET features over the course of the next two Windows releases. The first, code-named "Whistler," will ship late next summer. Whistler is a minor upgrade to Win2K, and it will support the .NET building-block services I mentioned earlier, right out of the box. But Whistler won't offer much more in the way of .NET features, so we'll have to wait for the next major Windows release, code-named Blackcomb. Due in 2002-2003, Blackcomb will offer the full ".NET experience," including a new UI.

For its flagship Office suite, we'll see at least two more revisions before the full Office.NET product arrives. Office 10, which will ship late in the second quarter 2001, isn't a .NET product at all, although it does offer some enhanced Web and XML support. And the company recently acknowledged that even Office 11 won't be a .NET release, although a goal for that release—due in 2002-2003—is to bridge the gap between today's Office and the Office of the future, code-named NetDocs. For both Windows and Office, however, Microsoft currently plans to offer .NET and non-.NET versions because of customer resistance to the new model. No one said that transitions of this magnitude were easy.

In the meantime, other .NET products are in the works. For any of this to fly, developers have to be in the loop as early as possible. To this end, Microsoft has been very open about the next version of its developer suite, Visual Studio.NET 7.0, and the company recently offered an early beta of the product to developers for free. Expect to see Visual Studio.NET 7.0 ship sometime in the second quarter 2001 as well. But the best way to see where .NET is heading comes from a somewhat unlikely product: Microsoft's recent release of MSN Explorer. MSN Explorer, with its first generation .NET user experience, integration with MSN Web services such as Hotmail, Money Central, and MSN Calendar, and seamless, one-window design, really hints at the future.

So I'm curious. What kind of services would you pay for, and which services do you think Microsoft should simply give away as part of the Windows.NET platform?