News

SOA Research In Academia Increasingly Industry Focused

Posted by Mark Little on Jun 15, 2007 09:29 AM

Community: SOA
Topics: Web Services ,; Research

The 2nd European Young Researcher Workshop on Service Oriented Computing was held this week at Leicester University. It was a two day workshop, designed to bring together PhD students, young researchers working in the industry and those who completed their doctoral studies within the past few years. The core objectives were to exchange information regarding advancements in the state of the art and practice of SOC, as well as to identify the emerging research topics and define the future trends. The workshop was open to all european researchers and there were representations from all of the major research countries.

Of all the accepted papers, the emphasis on research was definitely industry focussed, with researchers looking at Web Services and SOA orchestration and choreography techniques based on WS-BPEL and WS-CDL. Static analysis of process flows and correctness proofs of complex composite applications were also popular themes, something that is typically left to vendor specific products (if at all). Policy, governance and SLAs were also topics that received their fair share of attention from the researchers and the audience.

Over the past 10 years or so, particularly in Europe, there has been more and more emphasis on industrial driven research within academia. That makes sense as governments remove more and more of their pure "blue sky" research funding and universities have to look elsewhere to fill the gaps. But with the relative merits of WS-BPEL still under discussion and with the jury undecided on whether Web Services should be the defato SOA implementation platform or not, slavishly following industry approaches does not always bode well.

However, if this workshop is anything to go by, academic researchers haven't completely gone over to the Dark Side: there was no shortage of critical comments on current specifications and standards in and around SOC; it was also nice to see some work around WS-CDL, which currently bucks the industrial trend. Furthermore, many of the research efforts described here were talking about pulling in combinations of efforts, such as WS-BPEL, WS-CDL, semantic web and compensation transaction and taking a much grander view of things than we are perhaps used to from industry. For instance, since the demise of the W3C's efforts in defining a Web Services Architecture, there has been no concerted effort by industry to provide one. Various Web Services specifications trickle out from major vendors, but if there is a master plan for how these things form an overall architecture, it's being kept locked well away.