David Chappell made this statement last week during one of his presentations at TechReady7, Microsoft's semi-annual internal technical conference in Seattle*. While I'm certain that this statement was made in part to shock the audience (after all he is a very engaging speaker), I do think there was some level of belief and rationalization behind the statement.
Regardless of the intent, this statement really highlights one of the challenges we face in constructively discussing such topics. Too often, we derail any hope for progress because we rat hole into semantics and brands that have taken on a life of their own. To make sense of "SOA is dead", you first need a way to clearly identify SOA and then you have to know where you can check for its pulse. Is SOA an approach, an architecture, a solution, a product, or a marketing term? If SOA is dead, is it because it did its job and is no longer needed; it has been dismissed as an approach; the architecture doesn't work; or simply that nobody cares anymore? If SOA is dead, does that mean that Web services are also dead? What about commercial services, distributed messaging, business architecture, SAAS, S+S, or ESB's? You could go on and on...
I've always avoided such debates around the applicability or appropriateness of a name or acronym du jour in this space because the only thing that really warrants focus is the problem space. It isn't going away whether we decide to kill an acronym, brand, or product category. "This" is a particularly complex problem space and in this competitive industry of ISV's, SI's, and analysts, everyone is fighting for mind share and perception to position their answer is to these customer's problems. While many choose to fight that battle to associate their current products, solutions, and offerings, many of us are rolling up our sleeves to solve the actual problems.
I started working in this problem space in '99 and I'm amazed at how long it has taken us to get this far! If we look at where customers are today with SOA, I would very much agree that it isn't meeting the promise or hype that has been associated with it. After all, it is supposed to align the business and IT and provide a higher level of agility. Of course that might be just a bit easier to achieve than world peace for most organizations. Anyone that has spent any time with actual customers knows that the challenges across Enterprise SOA encompass technical, process, and organizational challenges and capabilities. Since the last word in SOA is "architecture", I'm not sure how realistic it is to expect it to be a silver bullet for success. Most organizations are so dysfunctional they almost make money despite themselves, so getting them coordinated internally to promote consistency, reuse, and a higher level of agility is a journey that many of them can't make anytime in the near future. Technology may be the difference between getting a train or a jet, but if everyone doesn't get on and agree which direction they are going, they aren't going to make much progress. We can use software to make it more enticing and easier to jump on board, but at the end of the day we are talking about people. We've all seen people do things that we fail to rationalize for all sorts of reasons and that will never change.
The need to develop a massively distributed system that requires many of these people to jump in and participate through is also not going to just away, no matter what we choose to call the problem or a potential solution. Not as long as organizations want to automate processes, leverage new partnerships, establish new global business channels, and get away from IT silos that place constraints on the business. Over time, technology will make more apparent the value that can come from building that ecosystem, make it easier for people to trust it, and make it more approachable to interact with. That problem is very much alive and it is an exciting opportunity for those of us in the industry to help organizations extract more and more value from technology.
So, if SOA is dead, so be it... long live the next acronym and let's get on with it already - we have a lot of work to do!
*[Disclaimer: For the purpose of full disclosure, I did not get to personally attend this presentation because of a schedule conflict, but I did get second-hand information from several attendees.]