Looking over the PDC content for this fall you can definitely see a trend: Cloud Services. This is to counter Amazon's S3 offering.
First let's define what Cloud Computing is and what it means. Cloud computing is not Gmail, Hotmail, or even Google Documents. (This is why you won't see Microsoft Office moved to the Cloud in a very robust way, that is just web access for popular productivity software.) Cloud Computing is replacing an application's hardware and plumbing infrastructure with a service hosted by someone else. For example let's look a typical corporate system:
- Back-end data storage (SQL Server, Oracle)
- Back-end application server (IIS + .NET, COM+/ES, Tuxedo, even CICS (that dates me))
- Front end machine to access the application (Either a workstation via a browser or windows app, or mobile device)
It gets expensive for small and even large firms to build the network, firewalls, domain services, user accounts and of course the hardware (with its underlying RAID arrays, UPSes, racks, power, etc.) Sometimes this is a barrier to growth (or even entry) for small to medium organizations. Not only do you have to learn all about subnets and RAID arrays, you also have to learn about UPS and power strategy, and manage a development project. (Man, I just want to write some stored procedures, who cares about an UPS and RAID array!)
This is a problem, epically for start-ups. You have to build a rack before you even build a company and product. That is expensive and an up-front fixed cost. More than likely you will build a rack that at overcapacity so you can grow, since it is time consuming to add more capacity later on. Also, you usually have to hire people to help you figure this out. Building a rack is expensive and also expensive to maintain.
Cloud computing is providing the back end infrastructure, the first two bullets, to firms as a service. The beauty of this model is you pay for what you are using, bandwidth, storage, and clock cycles. So a startup does not have to spend $20,000 on a new rack, they can pay a monthly cost to host their application in the "cloud" and only pay for what they need. In the beginning this will be a small fee since a startup has no customers and then the fee will grow (as revenue grows hopefully!) Amazon S3 is attracting many startups and is changing the dynamics of funding a startup since they don't need as much expensive hardware and IT expertise.
So many critics of Microsoft say that they are missing the boat by not putting MS Office up in the cloud to compete with Google Apps. While Microsoft may have to compete with Google one day in this space (I personally prefer offline Office + a service like SharePoint to share my documents.) that day is not today. The real battle will be over the true Cloud-infrastructure and ultimately the stack.
Selling customers on your Cloud platform locks them into your technology. So if Microsoft offers storage (SQL Server Data Services), application synchronization (Live Mesh), workflow/relying (BizTalk Services) and application tier (IIS as a service? or at least VPCs) services, they will get developers to use the Microsoft stack (SQL, .NET, etc.) for another generation. Throw in Hosted Office and Exchange, and they are really hooked.
Judging from the PDC content, with sessions all around Microsoft Cloud services, I think that they get it. Too bad we have to wait until October, but I sure hope that it is good stuff.