Software as a Service  

Brent Sleeper writes on software as a service Web Services and Utility Computing (Brent Sleeper's Web Site). While I agree it is an idea whose time has come - in many ways was that not what the dot com thing was all about - a little over hyped perhaps and lacking in interoperability. Like most revolutions it was significantly over-hyped. Some of us remember the structured programming, AI, Object, and component revolutions and the hype there. What we all forget - new stuff is hard - it requires work. Anyone interviewed someone who thought they new OOP because they used C++?

To get to a level of abstraction that is easy to use by many people the level of effort is extremely high. This is true of objects, components and services. As each moves up a level of abstraction each is more challenging than the previous. Software as a service that delivers real business value will change how companies work, therefore it is reasonable to expect this to present the greatest challenge yet.

The major difference in this revolution is that has the potential to align business and technology efforts. One of my criteria for a well designed external service is that it has a clear business objective and value. Services deployed correctly will break down the silos we have built over the past decade and change how we build collaborative applications among organizations. This will be the true measure of sucess of shared infrastructure and software as a service.

posted by John McDowall | 7:51 PM