to homepage
 Weekly emails: how to advanced search
 Glossary lookup:


Loosely Coupled weblog

Thursday, August 18, 2005

Real-world designs on web services

Anyone who has ever wondered what service-oriented architecture is actually for might finally get their answer in a few months' time, when the OASIS SOA Adoption Blueprints TC starts reporting its findings. OASIS announced the new committee on Monday, but I got a chance to grill Miko Matsumura, one of the driving forces behind the initiative, about what it all means last week.

I have to confess I was highly skeptical that anyone needs a standards committee prescribing how we should all use SOA. But first of all I wanted to clarify how SOA Blueprints differs from one or two other apparently similar SOA initiatives.

The most glaring conflict of all, I felt, was with another OASIS committee that recently formed, the SOA Reference Model TC — which itself overlaps in a standards-body-politics kind of way with the W3C's Web Services Architecture group. But whereas these two initiatives are more to do with blue-sky thinking about defining the true nature of SOA and web services, Blueprints is all about getting down-and-dirty with real-life business needs, Miko reassured me (I think I should make it clear that I'm putting words into his mouth here, but he didn't disagree with the down-and-dirty bit).

The other area of conflict I could potentially see was with the work that WS-I has been doing on basic profiles for getting web services to interoperate. During our discussion, however, I developed some imagery using a Lego bricks analogy that (I feel) succinctly expresses the difference. WS-I is rather like a set of instructions that show you how individual bricks fit together (and indeed specifies the dimensions the bricks shoud conform to), whereas Blueprints shows you how you might fit hundreds of them together to build a model house or helicopter. It's at a different level — neither down in the gutter with the techies or up in the heavens with the architects, just strolling along the sidewalk where the ordinary folk go about their daily business.

But I'm still skeptical about the whole idea. Let me digress for a moment to explain why. Back in 1999/2000 I was writing about ASPs — application service providers. I was considered a leading authority, and a lot of people used to ask me to tell them how big the market was going to be. My response was that it was impossible to reliably size the market at such an early stage, and that that was the way it should be with any newly emerging technology market. By the time everyone knew how big it was going to be, the moment of opportunity would have passed. Well, what goes around comes around and ASPs, having been to hell and back, are now thriving in the guise of on-demand application providers and other software-as-services plays. So from next week (if you'll forgive this small plug) in addition to my continuing role here at Loosely Coupled, I'm going to be writing a ZDnet blog on software-as-services. It's not live yet, but you can read a sample posting here. Now, back to SOA blueprints.

In the same way that people back in 1999/2000 were trying to make risk-free investments in ASP ventures, today enterprises want to make risk-free deployments of SOA and web services. They want everything to work right first time, without anyone having to go through any trial and error to find out what works and what doesn't. It's only human nature to have such feelings I guess, but life ain't like that. Someone, somewhere is going to have to make the mistakes so the rest of us know what to avoid next time.

A minority of enterprises understand this and are pressing ahead with their web services and SOA projects. Are they sharing the benefits of their experiences? Are they heck. They want to maximize the competitive advantage that they've invested so much blood, sweat and tears in achieving. That's why the same case studies are always doing the rounds of Loosely Coupled and our competitors. There are very few SOA adopters who are yet willing to talk about what they've done.

Is it any more likely they'll reveal their secrets to the OASIS blueprints committee? Miko says the fact that it's a standards body with a formal governance process and all the other guarantees of probity and neutrality does count for something. And of course early adopters do have a vested interest in being proved right in the long run, so it is in their interests to have the mainstream follow in their footsteps sooner or later. In any case, I applaud the initiative. Let's hope I'm wrong and it does have some successes. It's about time someone showed that SOA does actually deliver business benefits.

posted by Phil Wainewright 11:22 PM (GMT) | comments | link

Assembling on-demand services to automate business, commerce, and the sharing of knowledge

read an RSS feed from this weblog



latest stories RSS source

Headline news RSS source


Copyright © 2002-2005, Procullux Media Ltd. All Rights Reserved.