Something new on Mashups and SOA
Posted by Yannick Martel on December 3, 2008
At least something new about SOA! Many articles and discussions on SOA only address the technical or implementation aspects of SOA. We then wonder SOA is only a technology for IT departments to worry about, or whether it really concerns IT users. I have frequently been frustrated by some of the debates about SOA: much ado about nothing… Of course technology has evolved, but, more or less, service-oriented architectures have been possible for a long time, and have not been invented with Web Services. Limiting the debate on SOA with reorganizing existing IT applications around services restricts the debate to technicians – with only incremental improvements to bring to users. So why should they pay for major changes? An incremental improvement can justify only a progressive introduction, always guided by business requirements or a quest for optimizations.
Mashups Corporations brings something new with inviting directly the corporation’s strategy into the discussion. Organized as a novel (in the tradition of The Goal), its introduces us into a “brick and mortar” corporation, which is ultimately lead to evolving its IT under pressure from some Marketing product manager attracted by the new possibilities of Internet and Web 2.0.
This corporation’s IT is at the beginning of the book organized traditionally as a cost center, an expensive black box, tolerated as necessary for the company, secured and closed. In the shadow of this official IT, a “pirate” IT (the Shadow IT) survives, developed by employees avid to put new technologies in the service of their innovative ideas. Change arrives when the Shadow IT opens up to the outside (quite inadvertently at first) and allows third parties (client, prescriptors…) to interact with it. To cope with the new flow of transactions and revenue then generated, the official IT must open up (just a bit) to the Shadow IT. Then we begin to see the real, bottom-line certified justification for services and a service architecture. From a central fortress, opened only via GUIs, with a few Shadow IT autonomous cells, we switch to a central core, which provides access to internal and external applications via well-defined interaction points – services. This is the justification for SOA, as opposed to morphing an existing stable architecture into a service-oriented one, with limited business value.
This transformation creates new problems for the IT department, as well as provides a new positioning: from a cost center, attached to the CFO, the IT department becomes an innovation facilitator, supporting fundamental and industrial processes as well as new ideas – for people who can experiment, try, fail and succeed, develop new revenue streams, whether they are part of the company or not. IT must change its culture and its mission at the service of the rest of the company. Shadow ITs becomes authorized and officially supported.
What is the recipe for attracting and retaining customers in the XXIst century? Let’s allow third parties to develop new applications which process transactions by interacting with the core applications of the company. Once these third parties, prescriptors, clients, communities, are hooked onto your systems, they are attached to you. They find a competitive value in the interaction with your IT, which your competitors does not bring them – yet. The merit of Mashups Corporations resides in this perspective of SOA related to Web 2.0, justified by new capabilities to innovate and open. The technical information brought by Mashups Corporations (in appendices) is pretty standard, bringing nothing new as compared to mainstream SOAP-SOA – no debate on resource vs RPS styles, nor alternatives to Model Driven approaches. But that’s not the core of the book, nor of the debate.