Debate: Is SOA dead, or 'just resting?'

Must a programming methodology be a philosophy in order to survive the corporate boardroom? Maybe, but when the philosophy itself doesn't tread water for long, then what happens?

During the late 1980s and early '90s, during what was then my dual career as a software developer, I was something of a contrarian about a major methodology of my business: I believed, and actively advocated, that the design of programs should not tightly associate data with the code that utilizes it. I was told at the time I was not only rebutting a methodology, I was threatening a fundamental tenet of business, and some said I was against a way of life.

For most of this decade, I've ridden a wave of vindication. The Web-based development model has, as one of its founding principles, a so-called "loosely coupled" approach to data. Essentially, it says that a client in the outside world that requests data from a networked application need only understand the protocol for requesting that data, not the architecture of the server-side program generating it. There emerges an economy of sorts between the server and client, where the client is said to "consume" services provided by the server. This is the basic concept of service-oriented architecture (SOA), the successor to object-orientation among both evangelists and publishers.

But as with any methodology that's "in," the way you sell it to an audience is by equating it with some part of life in general. (Guilty as charged.) Witness if you will the various permutations of Moore's Law, which originally referred to a formula for the natural doubling of transistors in integrated circuits over two-year intervals, but which has since been applied to personal finance, clinical psychology, and at least once, the use of chili powder in Southwestern restaurant cuisine.

Here's an example of the application of SOA to the bigger lens of history -- whether it fits or not -- from a 2005 IBM Press book entitled Service-Oriented Architecture Compass:

"Corporations are built on the assumption of continuity; they focus on operations. On the other hand, capital markets are built on the assumption of discontinuity; their focus is on creation and destruction. The market encourages rapid and extensive creation and hence greater wealth building. The market is less tolerant than the corporation when underperforming over the long term. Some of the key reasons for failure are ignoring higher-value markets, the inability to address more technologically advanced competition, or competition from lower-cost sources...Today, enterprises must be more dynamic than ever to survive. They need new, evolved ways of handling the competition, and their IT infrastructures must support them as they face unique challenges they didn't have to face years ago. We believe SOA is the way that companies can develop IT infrastructures capable of supporting dynamic enterprises."

Recently, a debate has emerged in the software architecture community over whether SOA is dying -- or whether, more accurately, the "meta-concept" of SOA is dying away, leaving behind the actual architecture on which it's based.

The debate began with a blog post from Burton Group research director Anne Thomas Manes entitled, "SOA is Dead: Long Live Services." Here's an excerpt:

"Once thought to be the savior of IT, SOA instead turned into a great failed experiment -- at least for most organizations. SOA was supposed to reduce costs and increase agility on a massive scale. Except in rare situations, SOA has failed to deliver its promised benefits. After investing millions, IT systems are no better than before. In many organizations, things are worse: costs are higher, projects take longer, and systems are more fragile than ever. The people holding the purse strings have had enough. With the tight budgets of 2009, most organizations have cut funding for their SOA initiatives.

"It's time to accept reality. SOA fatigue has turned into SOA disillusionment. Business people no longer believe that SOA will deliver spectacular benefits. 'SOA' has become a bad word. It must be removed from our vocabulary."

Manes goes on to mourn the demise, as she puts it, of SOA; but the substitute she suggests for it in the public vernacular is a more disruptive application of the architecture's original ideal -- in other words, not so much service-oriented architecture as architecture that is service-oriented. Rather than applying millions in R&D resources studying the problem, she advises, information technology should be rebuilt from the ground up so that concrete services have discrete interfaces.

Next: A Microsoft architect performs an SOA "autopsy"...