Microsoft Reveals Longhorn Launch Date: Time to Plan, Time to Worry
Late last week, Microsoft issued an unusual announcement regarding its oft-delayed Longhorn OS. The announcement consisted of multiple parts, which I've detailed in my article, "The Road to Windows 'Longhorn' 2004," on the SuperSite for Windows ( http://www.winsupersite.com/showcase/longhorn_preview_2004.asp ). The announcement outlines the following key points:
1. Microsoft will make Longhorn "broadly available" to customers in 2006. According to Microsoft, this wording implies that Longhorn will ship earlier rather than later in the year. Microsoft Lead Product Manager Greg Sullivan said, "We're targeting the first half of 2006 and are expecting \[the release\] to happen around mid-2006." This announcement is significant because this is the first time Microsoft has publicly committed to a final release date for Longhorn. The company also announced that Longhorn Server will ship in 2007, a year after the Longhorn client.
2. Two key features from Longhorn--the Avalon presentation layer and the Indigo Web services communications infrastructure (and the WinFX programming libraries that are required to take advantage of the two features) will ship separately for Windows Server 2003 and Windows XP when Longhorn ships. This news has enormous ramifications for developers because they can now develop Avalon and Indigo-based applications and services and be sure the applications will run on a large installed base from day one.
3. One key feature of Longhorn, the Windows Future Storage (WinFS) storage engine, is being delayed until 2007 and won't ship in the initial release of the Longhorn client. Instead, WinFS will ship as part of Longhorn Server in 2007 and will ship separately as an add-on for Longhorn client customers. This development has caused some consternation with developers because Microsoft had previously pushed WinFS as a major part of the Longhorn release. However, the removal of WinFS doesn't mean that other key file system-related functionality is being relegated to the dustbin as well. For example, the heavily touted instant desktop searching feature will still be included in Longhorn. (For details about this feature, see "Don't Wait for Longhorn: Copernic Releases Instant Desktop Search Tool" at http://www.windowsitpro.com/article/articleid/43833/43833.html .)
4. Longhorn will still be an important release. Many key features from Longhorn are still intact and make this release quite desirable. For example, Microsoft will still make the gorgeous Aero UI available only in Longhorn and won't back-port it to XP. To date, Microsoft has been very secretive about the end-user feature set of Longhorn; however, the company will likely reveal these features over the next several months.
The information in this announcement is valuable for several reasons. Obviously, for IT professionals, systems administrators, and corporate decision makers, Microsoft's full-disclosure announcement gives them the information they need to plan future migrations. And it establishes the life expectancy of Windows 2003 Release 2 (R2--the next minor update to Windows 2003) and XP; those products will be supplanted by major new releases in 2006 and 2007, respectively. However, some critics are pointing out--somewhat correctly--that the 2-year delay between this announcement and Longhorn's launch opens an opportunity for Microsoft competitors--chiefly Linux and Apple Computer's Mac OS X--to gain some traction while the software giant spins its wheels. After all, Microsoft just admitted it won't be shipping a new OS for a while, and XP Service Pack 2 (SP2) doesn't appear to completely fix its current security ills.
On another level, doesn't Microsoft's inability to deliver what it's repeatedly promised with Longhorn speak to wider leadership problems as well? After the Bill Gates-initiated behavioral problems led Microsoft to two major antitrust cases, one in the United States and one in Europe, the world's richest man stepped down as CEO, assumed the mantle of Microsoft chairman and chief software architect, and began working almost full-time directing the development of Longhorn. Since then, the product has suffered delay after delay and is now being pared down to meet a ship date that's at least 2 years behind the original schedule. Has Gates lost it?
From a broader viewpoint, why hasn't Microsoft learned the lesson of the foibles of monolithic software development? As I complained during the development of Windows 2000 (which was arguably the company's most disruptive enterprise release ever because of its then-new Active Directory--AD, Group Policy, and IntelliMirror features), Microsoft does its best work when it can focus on small, easy-to-identify problems. For example, XP SP2, despite some obvious missing features, accomplishes exactly what Microsoft said it would. Meanwhile, some Microsoft shops are still electing to put off AD migration because it's so hard, time-consuming, and yes, disruptive.
I can't help but think that Microsoft's customers would have been better served by more incremental upgrades than Longhorn, such as an "XP Second Edition" that offered a small, but compelling, group of improvements for both consumers and enterprise customers; this product would have mirrored the Windows 2003 R2 release. Heck, that's the type of upgrade that Software Assurance (SA) licensees expect, isn't it? Instead, Longhorn is so far-reaching that Microsoft itself can't seem to get a handle on it. As a result, the next 2 years should be interesting, whether you're a Windows competitor, a Microsoft customer, or an industry analyst looking for something new to write about. We can at least thank the software giant for that.