I attended a webcast on Program Execution in the 21st Century yesterday. And while it didn’t grab my full attention, it did provide a good overview of the CLR and its development paradigm – good enough to sign up for the next class. This morning, I realized that proper execution of the framework strategy will hurt Longhorn adoption.

The primary rule on computer purchasing has always been: “buy the computer that your applications need.” And as the new framework applications supplant the old, applications become framework dependent and OS independent. Of course, some applications will need to use new functionality only available in Longhorn. But the majority will be just fine on Windows XP. And I would expect it to stay that way – it is in the developer’s best interest to build on a framework common to both Longhorn and XP.

A month ago, Michael Gartenberg posted that driving Longhorn upgrades will not be easy for Microsoft. I’m thinking that it may be worse than he thinks.