Updated – Implications of Microsoft’s Longhorn Cutbacks

September, 2004

The big news from Microsoft recently is that the company is cutting back on the scope of Longhorn, its next generation of the Windows operating system for desktops and servers. The choice was either to push out Longhorn’s 2006 release date or cut back on features. The release date won.

What does this say about the size and complexity of Microsoft’s operating systems? I’ll get to that in a minute. But first, let’s look at what Microsoft is cutting out of Longhorn:

  • The biggest item to get the ax is the new file system, WinFS. WinFS was supposed to run on top of the Windows NT file system (NTFS), allowing users to more easily store and locate data. WinFS will now be introduced later in an upgrade to Longhorn.
  • The new Avalon presentation system is being “decoupled” from Longhorn, allowing it to be introduced with Longhorn and also as an enhancement to existing versions of Windows. This will make it easier for software developers to write new versions of their applications that can run on both Longhorn and older versions of Windows.
  • The new Indigo communications subsystem is also being split off from Longhorn, meaning that it might also be introduced separately. Indigo is Microsoft’s next generation communications layer for all web services and communications features such as instant messaging, voice-messaging, video messaging, collaboration and content distribution.

The result of these cutbacks is that Longhorn is turning into more of an evolution of Windows, instead of a revolutionary jump in OS architecture. It also means that application software developers will be able to more easily make the transition to Longhorn.

Backing up a few steps, however, it’s hard to avoid the impression that Microsoft’s operating systems continue to get more massive and more bloated. Furthermore, as Microsoft adds new features, backward compatibility to thousands of third party hardware devices and software applications is becoming more and more a burden.

Clayton Christensen, in The Innovator’s Dilemma, points out that market leaders ultimately overshoot the needs of the majority of their customers, opening themselves to competition from new technologies at the low end of the market. Such “disruptive innovations,” as he calls them, do not provide the extensive features and functions of the market leader. But at the low end of the market, they are good enough. And they are much cheaper and simpler to implement or use. At first, a market leader ignores the new technology, because it does not threaten the market leader in its core business. But as the new technology improves in capabilities, it begins to move up market, eventually taking major market share from the leader. Christensen points out dozens of cases where this has happened, in industries as diverse as disk drive manufacturing, to retailing, to excavation equipment.

Do the majority of home computer users (think of Aunt Millie) really need WinFS? Do the majority of businesses (think of Uncle Joe’s machine shop) really need Avalon or Indigo? I’ll leave it as an exercise to the reader to speculate on what other operating system or technology represents a threat to Microsoft, especially for the majority of users at the low end of the market.

Far Reaching Consequences
On another note, Microsoft’s decisions to cut back the scope of Longhorn will have serious implications for Project Green, which is Microsoft’s program to merge the code base for its four ERP systems: Great Plains, Axapta, Navision, and Solomon. Project Green was, most recently, scheduled for release in 2008, following delivery of Longhorn in 2007. But with the feature set of Longhorn being cut, Project Green’s delivery schedule is now uncertain.

The dependencies between Longhorn and Project Green are complicated. Project Green was to be built upon a new set of developer tools and software classes called the Microsoft Business Framework (MBF). MBF, in turn, depends in part on Microsoft’s new WinFS file system, which Microsoft just decided to cut from Longhorn.

With WinFS being deferred until some unspecified date after Longhorn, it is uncertain when MBF will be delivered. I suppose it’s possible that MBF could be released using elements of existing Microsoft technology, which means that MBF could be released ahead of Longhorn.

That might allow Microsoft to release Project Green earlier. It depends on whether delivery of MBF was on the critical path for Project Green prior to Microsoft’s decision–I don’t know, but I doubt it.

September 2004