Its been a busy month.  We've been working furiously behind the scenes to sync up MBF with Longhorn.  Our previous plan of record had MBF being released in two phases.  The “business logic” portions of MBF were to release with Visual Studio 2005 (aka Whidbey).  Then the rest of MBF would ship with Longhorn.  

This plan was a compromise from the beginning.   As previous posts have noted, MBF has many interesting parts (SOA, Process...).   However, all of those “interesting” parts were not shipping until Longhorn.  We literally couldn't ship rich SOA support with Whidbey... we had to wait for Indigo.  We couldn't ship the kind of lightweight, embedded process support we need in Whidbey... we had to wait for a future version of the BizTalk Orchestration engine.

The only point of the MBF Visual Studio 2005 release was to enable developers to start coding their business logic.  This is actually a very worthwhile goal.  Business logic represents the largest portion of a business application.  Getting a head start coding the business logic is goodness... even if you have to wait for the Longhorn release of MBF to “finish” the application.

However, as the months have gone by it has become more apparent that collapsing MBF into a single release would be more beneficial for our customers.  By shipping portions of MBF early in Whidbey, we were cementing portions of MBF into a non-Longhorn model.  In case you've been living under a rock, Longhorn is <the> platform for the next decade.  Shipping new bits which aren't aligned with Longhorn is a disservice to ISVs who want to build business applications which will last through the next decade.

Here are some links to stories related to the MBF move to Longhorn.

Personally, I have mixed feelings on this topic.  Software developers live to ship code.  So delaying the first release of MBF is painful.  However, as a Software Architect, I am absolutely convinced that this was the right choice.  Shipping part of MBF in Whidbey and part in Longhorn would have created a system which spans two technology eras.