SOA, prioritise the the “SO”, rather than the “A”

You may have recently read SOA’s obituary as presented by Anne Thomas.


I read and identified with many of the points presented. I wanted to pick up on one point:

  • “SOA” requires redesign of the application portfolio

As Cloud Computing becomes real I see “Online Servers” emerging rather than a comprehensive library of services. Services are being released as a portfolio, that is subscribed or used as part of a collection.

My point being, if you are not enabling your whole application portfolio as being service oriented, you cant create establish a service value chain. That is, the ability to distinguish the role of the service in providing value to its consumers and

Lets face it, services will continue to survive only if its consumers are getting some value from it.

In other words, many parts of applications in production do not deserve to survive, but do so only  because they cannot be (easily) removed. This will be likely due to some business dependency, sunk cost or risk/complexity of replacing. The cost value justification is harder to establish in a non services environment.

So, when reading a related Microsoft communication I picked out these key points Microsoft is focusing on:

  • Web Services as the anchor
  • Standards support and making it real through interop testing
  • No more “A” than is absolutely needed to solve the specific problem
  • Changing the conversation on SOA from a “what” to a “how”
  • Promoting Real World / Pragmatic SOA since the beginning

If you are considering a strategy these, I would argue are good values to mimic. If you don’t then the ability to create new business models or shift to new operating paradigms such as Online Services are going to be larger, costlier step for your business.

Skip to main content