- When the hype ends, we're there. Eventually well stop talking about SOA and go back to talking about "architecture." When we stop talking about SOA it will finally become a reality.
- SOA is not Web services. SOAs will most likely be built using Web services (but building Web services will not necessarily result in a SOA).
- SOA is not new. CORBA, DCE, DCOM and EDI were all early examples. EDI may have provided the first example of SOA principles (e.g. document-centric, loosely coupled, etc).
- SOA is a design philosophy, not a technology or a methodology.
- We don't know yet what SOA will bring. The most valuable services will be used in ways in which their original architects never intended or expected.
- Chaos is natural, but try to avoid it. Service orientation can happen in one of two possible ways: chaotically (typical approach) or in a disciplined manner. The path your organization takes is up to you.
- SOA doesn't involve "objects." Services and business processes are not distributed objects.
- SOA is the means, not the end. If you set out to do SOA you will fail. All IT projects should have the objective of improving the performance of the enterprise.
- Embrace uniqueness. SOAs are like snowflakes; no two will be the same.
Microsoft's XML team provides some common-sense truisms about SOA in a recent post. The gist is that SOA for SOA's sake is absolutely useless, and will not fix any business problem. Here are some more pearls from this blog: