To present what Microsoft have invented, built and thought about the future for developers since the last PDC three years ago, four long days of course will only scratch the surface. But I think it's worth summing up some change of directions as to see where we are going. In some areas Microsoft is embrasing proven open standards instead of coming up with proprietary competitors. Building new service APIs on Atom, REST and HTTP makes interoperability a whole lot easier. Even IE, with its new rendering engine will be compliant. We're in the beginning of manycore, distributed highly parallel, scaling out computing. The traditional object oriented, Turing machine, sequencial way of thinking about programming is hard to adapt to the new patterns and practices, and new syntaxes to better describe solutions on a higher level is needed. I think we'll see a lot more DSL and modelling on the next PDC, when "Oslo" is rolled out and the upcoming WF4 that enables codeless workflows is also a step in the right direction. I have a hard time telling what will happen in the IW world. SharePoint is on a 3 year release cycle, so I hope the reason there was nothing to show at this time is because SP team was busy including a new RESTful API similar to Azure, native support for Silverlight controls and Live services integration to support disconnected operation. Update: I've later leaned that there was a secret invite-only session, where attendees after signing a NDA were shown some future SharePoint features. In the new and upcoming cloud business, clearly the competitors Google Apps and Amazon E2 are first movers. But Microsoft is betting heavily on the new Windows Azure operating environment, and I generally think Azure looks like a promising foundation for web and service applications. From what I've seen (eg. AD federation) I guess Microsoft will soon be #1 in this new market. The same applies to Google Gears/Live Mesh. Cloud computing is an important brick in the semantic web wall, and enables many exciting new innovative solutions that lives in the internet service bus. What bothers me is the vendor lock-in. Microsoft will eventually run into EC-monopoly issues again, which I guess is the reason they have already claimed they will productify the technology some time in the future to let 3rd party hosting centers provide the same functionalty. Windows Se7en on the other hand I'm not so sure about. Many major companies didn't make the shift to Vista for a number of reasons. The same reasons can be applied to 7, which turns out to contain only minor changes. Vista R2 would propably be the right name. Why it might turn out to be successful after all is when XP support is slashed and customers are forced to make the shift. Some developers I talked to were unhappy with the contents, and I agree some of the keynotes and sessions had only nice to know information. But still the insight in upcoming technologies, the possibility to influence them, and having the time to discuss ideas on a more abstract level with industry experts is quite valuable. See all the snapshots on flickr.