![]() |
|
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
|
||
| Home > Samples > Update > November 2002 |
![]() ![]() |
| A Fork in the Road to Longhorn? | ||||||
|
By Michael Cherry [bio]
See
more recent article on Longhorn (July 2003)
The following is the full text of an article published by Directions on Microsoft, an independent research firm focused exclusively on Microsoft strategy & technology. Each month we make one or more key articles available to non-subscribers.
Microsoft has not yet shipped Windows .NET Server, but that has not stopped senior executives from publicly describing a future Windows release, code-named Longhorn. While its features haven't been totally defined, Longhorn entails a redesign of major parts of Windows, including the file and graphics systems. As a consequence, customers and partners trying to plan around the Windows release schedule should be careful about betting on what Longhorn will be and when it will arrive, and anticipate the possibility of an interim Windows release in late 2004 or early 2005. Longhorn a Clean Slate Chief Software Architect Bill Gates, speaking at Microsofts annual analyst briefing in July 2002, said that Longhorn promises to be the greatest breakthrough to date for information workers, tightly integrating how applications, Web services, and the operating system (OS) store, present, and manipulate data. Based on the descriptions provided by Microsoft executives, Longhorn will include major additions and changes to the Windows codebase to accommodate the following:
Additionally, Platforms Group Vice President Allchin said at a Windows .NET Server developers' conference in Sept. 2002 that a goal for Longhorn is to resynchronize the release of the client and server versions of Windows. The current versions of the client and serverWindows XP and Windows .NET Server, respectivelystarted development together and were planned for a single release, code-named Whistler, but were shipped separately to allow Microsoft to retire the Windows 9.x and Me products while development of the server version continued. Consequently, Longhorn's schedule and feature set will have a major impact on software and hardware vendors. For example, companies that make backup software will need to plan new releases that support the Yukon-based file system, and chip, BIOS, and audio hardware manufacturers will need to adapt their designs to Palladium. For corporate customers, Longhorn also represents an important point in the life cycle of their own systems. For example, customers need to know how much Longhorn's new features rely on new hardware. Even more important, customers considering annuity licensing plans, such as Software Assurance, need to know whether Longhorn will arrive before the expiration date of the plans. (For recent information about Microsofts product life-cycles, see the illustration "Support Life Cycle for Business and Development Software".) Schedule in Doubt Allchin and Business Solutions Group Senior Vice President Doug Burgum have said that Longhorn will come in 2005 at the earliest. That estimate may be too aggressive. Historically the Windows division needs 40 months to design, develop, and test a large release such as Windows .NET Server. (See the illustration "The Windows Development Timeline".) Preliminary work on Longhorn has likely begun, but the real design, development, and test cycle for Longhorn is constrained by the fact that Windows .NET Server is still consuming considerable development and test resources as it moves from release candidate to shipping code. The scheduling is further complicated because the sustained engineering team needs to build and test service packs for three versions of Windows, and because major components of Longhorn, such as the file system, are dependent on work being done by external groups, such as the SQL Server team. So, adding 40 months to the anticipated release date of Windows .NET Server (early 2003) would push the next major Windows release out to early 2006. The discussions of Longhorn and its schedule are reminiscent of a much-heralded but never released version of Windows NT code-named Cairo. While some Cairo technologies, such as Active Directory, eventually became part of Windows 2000, Windows XP, and Windows .NET Server, Microsoft never released a full version of the OS with all the improvements planned for Cairo. An Interim Release? Cairo's history suggests an alternative plan that Microsoft could follow: release some but not all of Longhorn's features in an interim Windows release before 2006. An interim release would maintain a steady revenue stream from Windows and make it more likely that customers who bought upgrade rights for Windows XP will get a Windows upgrade before those rights expire. An interim release could also help respond to new features of competitors, such as Linux and the accompanying GNOME and KDE graphical environments, which are undergoing rapid evolution. But even interim releases take considerable development time: using Windows XP as a model and working from the date Windows 2000 shipped to the date Windows XP shipped shows that it could take approximately 20 months to build a compelling interim release, making it available in late 2004 to early 2005. Next Milestone: PDC Until Microsoft releases a more complete roadmap, customers and partners should look to the Professional Developers' Conference (PDC) for information about the Windows roadmap. This conference is not annual: Microsoft schedules one only when it has important technical information on strategic technologies that it needs to communicate to customers and partners. Until Microsoft holds such a conference for Longhorn, customers and partners should avoid making major commitments that depend on Longhorn's feature set or schedule. Resources For more information on Microsofts new licensing program, see the December 2003 Research Report, "Understanding Microsoft Licensing."
|
|
|||||
| Member Log On | Contact Us | About Us | Samples | Subscribe | Jobs | |||
|
|
||