Staying Flexible with Open Source SOA

I checked out the session called "What Is Happening With SOA in Open Source?" which was actually a panel hosted by Mark Hapner, Distinguished Engineer, Chief Web Services Strategist at Sun. Was curious what these folks had to say on the subject.

The participants, open source technical experts, were: Jeremy Boynes, IBM; Glen Daniels, Sonic Software; Mark Little, JBoss; Adinarayana Sakala, IONA; James Strachan, LogicBlaze; Peter Walker, Sun Microsystems Inc.

Most of the panel represented an open source project such as Apache rather than their own companies.

Here's the questions they responded to:

1. What is the most important core concept in your project?

2. Does your project extend SOA to cover services over JMS?

3. What web services protocol stack is currently used
by/provided by your project and how is interop
with other stacks being validated?

4. How important is it that developers be able to mix
and match open source SOA facilities within the implementation of a service?

5. What do you see as the major challenge that developers
must overcome to design and implement services?

With a panel this large responding to all these questions in rapid-fire sequence (they were timed as the session was only 45 minutes),
in no way can I do it justice, so I'll just summarize big time. Best, in this case to go to the Sun SDN site
(yes I am a Sun employee but this is a really good deal) and join the SDN so you'll be able to get the technical sessions in multimedia *free* when they're available around June.

Here are my take-aways albeit some pretty obvious and simple:

-- In establishing an open source SOA, the SOA principles must
work within as well as outside the implementation.

-- An SOA strategy cannot exist in a vacuum. SOA is about
leveraging existing infrastructure. If you only focus on some
internal strategy, other SOA implementations my be excluded,
defeating the whole purpose.

-- SOA is hugely complex and just beginning to manifest the
directions it will take. A lot of work goes into arriving at accurate
and useful conceptualizations of the complexity.

-- Leverage existing investments freely and do not get too wedded
to preconceived technologies.

-- Stay loosely coupled – and stay flexible.

Again, I think this panel is worth "tuning" into.