A Future Windows to Ponder

I've spent a lot of time in Windows & .NET Magazine UPDATE discussing the viability of upgrading various Windows desktop and server versions. Twice this year, I've written about the idea of a radically redesigned Windows version—"The Case for a Modular Windows," which ran in March, and "Maybe It's Time for a New Platform," which ran in late April. Neither article found many supporters for the cause, probably because in the real world, ease of upgrading is more important than supporting a deep architectural platform change, however beneficial that change might be. However, faced with customers' waning interest in Windows upgrades, whether because of rising licensing costs or the maturity of the platform, Microsoft is working on the most dramatic change to Windows since Windows NT first appeared in 1993. And that upcoming Windows version—code-named Longhorn—now hangs over the computing landscape like the Dark Cloud of Mordor from Tolkien's Lord of the Rings trilogy.

OK, maybe the situation isn't that sinister, but looking over Longhorn's vague set of features, I believe Windows users are in for a jolt. Due to ship in mid-2005, Longhorn will include a desktop replacement for Windows XP and server replacements for the Windows .NET Server (Win.NET Server) 2003 family of products. Longhorn will be a major upgrade—what Microsoft Chairman and Chief Software Architect Bill Gates calls "a clean sheet of paper, a rethinking of what a computer operating system is." Indeed, Gates is spending more time working on Longhorn than any other Microsoft project, which is a more telling sign of its importance than any PR piece the company can muster.

At a high level, Microsoft is designing Longhorn to help people overcome the limitations of today's computer systems and find documents and data wherever they're stored. That functionality will require a new database-backed file system—Windows File System (WFS)—and Microsoft will incorporate the technology behind that file system as part of SQL Server 2003, code-named Yukon. In addition, Yukon will unify all other Microsoft data stores, including Active Directory (AD) and Microsoft Exchange Server. This unification means that users will be able to search for, say, "Windows & .NET Magazine," and receive matching results in local file system documents, email messages, and the directory, based on storage context; no more separate searches. Gates calls this feature the answer to, "Where's my stuff?"

Longhorn will be deeply integrated with the Internet, further blurring the lines between your local desktop and Web services. Microsoft began this work with Active Desktop in 1996, but Longhorn will build on the special shell extensions in XP's task-based interface and let users more easily interact with Web services such as online contacts and calendars, stock quotes, sports scores, and movie tickets. The company will also dramatically update the UI with a layered interface that mixes today's 2-D desktop with Direct3D-based textured objects and full-motion video.

You've probably heard about the most controversial aspect of Longhorn—the often misreported and misunderstood Palladium technology. Palladium is an optional add-on for Longhorn that will require special PCs with custom security hardware from Intel or AMD. The technology is essentially a secure runtime environment similar to the Microsoft .NET Framework, except that Palladium requires a hardware component for absolute security, and you can use the technology to complete a variety of fine-grained, secure transactions. For example, you can use Palladium to send a confidential email message to a certain person, ensure that the exact person received and read the email, and ensure that the person can't forward your email to other people. Palladium grew out of Microsoft's Digital Rights Management (DRM) work, and we'll see a DRM server product in 2003 that will foreshadow Palladium's feature set. Essentially, Microsoft discovered that the digital media-oriented DRM technology had more far-reaching abilities and the company could use the technology to secure a variety of content types, including email and text-based documents. As the DRM server and Palladium come into focus, I'll examine them further in UPDATE.

Microsoft's decision to radically rebuild Windows didn't come lightly. In the July "Fortune Magazine" article that revealed the company's Longhorn plans for the first time, Gates said that he and CEO Steve Ballmer decided it was time to stop releasing incremental upgrades and offer a product that would really grab people's attention. "Let's do something more dramatic," Gates told Ballmer, who said, "This means synchronizing [various technology] releases ... Isn't it obvious that we should do this?"

Gates, ever the gee-whiz techno-nerd, claims to want people to upgrade because of the "wow" factor, although how the wow factor will impress enterprises is unclear. On the corporate side, I think Longhorn will be radical enough to drive adoption, although the ever-present cost concern will probably stifle some enthusiasm. But enterprises complain about upgrade costs with each Windows release, so why shouldn't Microsoft release something radically better than the previous version? With Longhorn, the improvements should be so dramatic that many fence sitters will arguably have cause to jump onboard.

The problems with the Longhorn scenario, however, are legion. First, if Longhorn doesn't ship until 2005, how will Microsoft address consumers who expect a new Windows version on new PCs each holiday season? I suspect that the company will simply issue new XP service packs to address this need, but that's just a guess. And what about the age-old problem of selling today's software while publicizing your future plans? Microsoft is being very careful to slowly leak Longhorn information to the public, but from what we already know, this next release will be substantially more exciting than what we have today. This revelation won't stir enterprises, of course, but casual observers might see little reason to upgrade today if their new PC will be obsolete in 2 years.

And what about the standard Microsoft problem of over promising and under delivering? Will Longhorn really be as revolutionary as the company promises, or will the company cut its feature set when the deadlines loom? I'll be keeping tabs on Longhorn to find out.