It may be presumptuous to make predictions about cloud computing, since the cloud has such little history and is almost all about future potential. But here are a few prognostications, based on trends that have established the current cloud-computing paradigm.

** All applications will move into the cloud. There's been some discussion over whether certain applications -- those that are compute or graphics intensive, for instance -- may not be suited to the SaaS model. That controversy will fade as more applications come online. For example, Autodesk, the leading computer-aided design (CAD) software vendor, is already moving features and capabilities into the cloud.

** Platform-as-a-service (PaaS) will supplant software-as-a-service (SaaS) as the most important form of cloud computing for small and, especially, midsize businesses. PaaS is well suited to organizations with limited IT resources, in particular limited IT staff, that still want to design, implement, and support custom applications. (Check out the example of 20/20 Companies in this feature story.)

** Private clouds will be the dominant form of cloud computing in large enterprises for the next couple of years. The public cloud's gating factors -- security, performance, compliance -- will be enough to hold back its significant, widespread advance into the enterprise while internal IT organizations catch up in terms of virtualization, provisioning, application development, and service automation.

** Hybrid clouds eventually will dominate enterprise IT architectures. The advantages inherent in a technology architecture that combines specific internal capabilities with low-cost, flexible external resources is too compelling to ignore. This argues well for IT vendors with comprehensive portfolios of products/services. Microsoft, especially, will emerge as a major force in this new terrain.

** The term "cloud computing" will drop off the corporate lexicon. Once the cloud is firmly ensconced in the IT environment, there will be no need to make such a distinction.

I don't see these as long-term developments but more in the two-to-three-year range. What do you think? Have I missed anything?It may be presumptuous to make predictions about cloud computing, since the cloud has such little history and is almost all about future potential. But here are a few prognostications, based on trends that have established the current cloud-computing paradigm.