Longhorn: A Lot to Want, A Long Time To Wait

Opinion: It won't reach users' desktops until 2006, but the next major Windows rev will revolutionize Microsoft's venerable platform.

Editors note: Microsoft Corp. is a client of the Enderle Group, the consulting firm headed by Rob Enderle.

On Monday morning, Microsoft finally rolled out the prototype of its new operating system, currently codenamed Longhorn and likely to be called Windows 2006 when it is released (following Microsofts current naming convention).

Longhorns launch pad was Microsofts Professional Developers Conference in Los Angeles. This years PDC—Microsofts largest yet—sold out some time ago, proving that despite industry interest in emerging platforms, Microsoft still commands the attention of developers.

But how long can they hold out? With technology moving as fast as it is, its hard to believe that a critical platform like Windows can go so long between major releases; after all, the last major release was Windows 2000. Strangely enough, though, buyers (particularly corporate buyers) have indicated that they want a six-year product life cycle, and they certainly got it with the Windows 2000 product (Windows XP was the maintenance release of Windows 2000 and both co-reside very well as a result.)

Since we are at least one year from the first real Longhorn beta tests, saying that this product will be in flux would be a major understatement. Nevertheless, it seems clear the OS will far surpass Windows XP in terms of security, reliability and manageability. It may also exceed previous products in backward-compatibility; thats a pleasant surprise to many of us who thought compatibility would suffer as Microsoft moves to a new code base in an effort to eliminate security exposures.

This product represents a far bigger change than Windows 95 or Windows 2000 did; Microsoft can honestly portray this as a revolutionary rather than an evolutionary platform. Basically, this is a ground-up redesign of the platform. Its achieves backward-compatibility by supporting both Win16 and Win32 subsystems, which should achieve a level of compatibility—even with 16 bit applications—thats impossible today. (Of course, it is also expected to support the new Win64 subsystem and further move us way from the problematic Win32 code base that has become so troublesome.)

One aspect of Longhorn that developers should like a lot: To work properly, most of the key improvements will require new versions of current applications. Old applications will run, but the experience will only be a slight improvement over Windows XP. If the platform is successful, it should generate substantial application revenue for third parties.

If it works, Longhorns backward-compatibility could approach the best of all possible worlds: Applications vendors will have a strong argument for an upgrade, but users wont feel they have no choice but to upgrade the applications. Obviously, the actual execution will make a huge difference.

Hardware vendors are excited as well because, as with the applications, the improvements to power management, remote management and control, and security will require new hardware to function properly. If successful, Microsoft will light a fire under the Wintel market we havent seen since ... well, ever.

Key improvements will include:

Migration: Moving from one machine to another will become nearly as easy as it is on the Mac. This has been one of the top-level requests from the OEMs and will clearly work best with applications built specifically for Windows 2006.

Security: Thanks to joint efforts among Microsoft, Intel and the Trusted Computing Group, the product should in many ways be more secure than any platform currently shipping. (Unix and Linux will clearly not be standing still either during this time).

File system (WinFS): A new file system will result not only in faster better search, but will extend up to the Terabyte range in terms of hard drive capacity, and have imbedded visualization capability so that you can better relate to and understand the information that you have.

Communications: As a result of joint projects with Cisco, Wi-Fi, and other related companies and communications groups the product will more easily connect to both wired and wireless networks securely. WAP will likely be a thing of the past.

User interface: Perhaps the most noticeable change will be a brand-new user interface that will leverage the power of PCs available around 2005 or 2006 to make the overall experience both more effective and more entertaining. This technology is being refined both from work done in Microsofts dedicated labs and the universities it works with.

Developer support: Based largely of directed feedback both from Microsofts internal and external developer community, this platform will take .Net development to the next level and apply what the company has learned over the past four years by fully melding it with Windows. (This was not expected to happen until Blackcomb indicating that this integration has moved in substantially in the roadmap.)

Reliability: The patch model will be altered significantly to emulate the experience with current antivirus products to make patching near-transparent and significantly reduce the need to reboot. This product is designed to be changed in place.

Services (Indigo): This shifts the emphasis and control back to the client with native support for advanced Web services and collaboration.

Presentation (Avalon): With faster, more realistic rendering, a unification of documents and media (multimedia documents), Avalon will provide the new foundation for the Longhorn shell and much of the "wow" factor in the new UI.

There is no doubt in my mind that this clearly represents the biggest change in the desktop (and eventually server) OS that Microsoft has ever undertaken. It also showcases that the company understands the threats it faces from Linux and the Mac OS. Microsoft is alerting its partners, employees and vendors that every available resource will be aimed at this project.

Like looking at a new show car, I want this thing; I also realize that until we see the beta, we wont see any of the negatives associated with it. I know the market is substantially antsier than I am, and three years is an incredibly long time to wait. Of course, there was plenty of warning for Windows 95 as well, and it still took out OS/2 and the then-Mac OS. Were Longhorn to come out this year or next, it would bury Linux and the Mac. But with three years lead time, this is a horse race—and its one from which the computer buyer will profit no matter who wins.

Rob Enderle is the principal analyst for the Enderle Group, a company specializing in emerging personal technology.