Colleague James Maguire has another interesting piece up on DATAMATION yesterday. James and I have traded email a bit as he's written about Software Oriented Architecture. This is another rock-solid piece that's well worth reading.

I still owe James a couple of ideas about articles I might contribute over at DATAMATION as well.

SOA: Hype vs. Reality
April 12, 2007
By James Maguire

If you listen to the SOA vendors (or the research reports funded by SOA vendors) you would think that Service Oriented Architecture has swept the enterprise.

Big companies across all sectors, we're told, are building SOA infrastructure. Business analysts and tech departments are eagerly putting aside their historic enmity to discuss upcoming SOA projects.

After all (and this part is true), SOA offers key competitive advantages. Over the years, companies have built computing infrastructures that combine a patchwork of environments, from Java to .NET to Sun Solaris. SOA’s promise is that it can integrate these disparate worlds to create unified interfaces (called “services”) for employees and customers.

That’s a big deal. It’s reasonable to believe that companies would be rushing to implement SOA.

But no. Despite the rich harvest that SOA offers, companies have been slow to plant the seed. In the early days of 2002-03, SOA was mostly just a glimmer in the eyes of vendors. “For several years, the vendors were way ahead of the market – no question about that. IBM especially,” says Marianne Hedin, an IDC analyst who covers SOA.

To create excitement, vendors issued a profusion of press releases and held copious workshops with free donuts. In response, they got a lot of polite nods. Vendors “made a big deal out of this, but nothing much was happening out in the market at that point,” Hedin notes.

And not all that much has continued to happen. “I think in the beginning, back in 2004 and 2005, they were a little disappointed that it was slow. And in 2006, they were hoping, ‘Oh yeah, this is a big year for SOA.’ And it really didn’t pan out to be a big year.”

While she hastens to add that things are starting to progress, it’s not as if “SOA” is on the lips of every IT decision maker. Based on a 2006 survey by IDC, only 23 percent of companies have a SOA project “in production,” with another 18 percent having one “in pilot stage.” That’s a tentative embrace of SOA, given that 22 percent either “don’t know” if they’ll invest or have “no plans” to invest.

More positively, 37 percent plan on investing in the next 1-2 years.
(Although as Hedin notes, "Thinking about it is different than doing
it.”)
[Read full post]

James raises a growing concern I've had over the past several months is SOA really going anywhere. And I'd lump that other buzz phrase, Software as a Service (Saas), in the same boat. They're different ways of sayign much the same thing. As he points out, every year lately has been claimed as the year SOA will take off. I hear faint echoes of Is VoIP Past the "Best if Used By" Date? Has it peaked? here.

VoIP has been painfully slow on the uptake, and I think it's potential as a disruptive technology has shifted to it's acceptance as a mainstream infrastructure service. It's just another carrying mechanism for voice. Voice is the underlying service. VoIP pretty much failed to disrupt as many predicted.

James' article raises the same concerns for SOA. Will it really take off with meteoric, widespread adoption, or is it just going to slowly filter into unified communications over time adn under a variety of names? I think so. Predictions are beyond 2009. It's going to be a while coming.



Technorati Tags: SOA, Service Oriented Architecture, Software as a Service, SaaS, DATAMATION