Just a good glimpse of what Microsoft has planned for the next 2-3 years. I see PDC as an important opportunity to see the long term vision start to take shape. Longhorn, Yukon and Whidbey are the three key technologies I want to understand more about.
Yukon, I think. represents a major change in the way we design multi-tier systems. Tim Sneath recently wrote about this - and i think he may be on to something. Tim argues that with ability to put business logic, expressed as managed code, into the database, that the middle tier is dead. I think he's right that the purpose of the current middle tier, business logic, really belongs in the db. But I think he's wrong about predicting the death of the middele tier for two reasons. First, it will take companies years to migrate to Yukon. Mainstream adoption is at lead 18 months away so inertia is a factor. But more importantly, the middle tier will exist to orchestrate, The front end UI tier will remain client side - but i see the middle tier as orchestrating the various services that are available in a SOA approach.
Longhorn is both the next client and the next server, so it too is highly important. The cliente is some what less interesting although I have to be careful how I say this - Scoble will give me too much grief!! In the space I work in, businesses have trouble understanding the value proposion of XP, let alone something beyond that. When I see all the fantastic UI shots, I have to ask: will this make the knowedge worker more productive? Can he/she answer emails, write documents or prepare presentations faster? I don't know the answer to this - although I can see how WinFS will help me to organise the chaos that is my workstation!!
Whidbey, of the three, is the least exciting aspect of the PDC. I've seen some of the new features (generics, iterators, etc) and have been playing with the Whidbey alpha for a while - I guess I'm just not enough of a geek to get overly excited.