Home Research WSDarwin

Service-oriented architectures have become the norm in developing modular systems. However, since the parts of a service system are usually developed and maintained independently, the evolution of web services may have serious repercussions for the consistency of the system. The particular properties of service systems, like information hiding and the required online connection, may exacerbate the problems with addressing the evolution of such systems.

Therefore, it becomes evident that special support is required for web service systems. First of all, the decisions around the evolution of web services have to be taken with extra care, since the information within a service ecosystem about consumers and other competitive service providers is not complete. In practice, this means that any decision made in such an environment may have indirect and unanticipated effects to the decision maker. Second, the service client is usually on the receiving end of these decisions and is the one who will be asked to adapt to the new version. Given the special nature of the service ecosystem, the client will need to understand the changes with minimal information and react just in time to avoid any disruptions in the application's functionality.

In my doctoral research, I study the evolution of web services and their systems. I make three major contributions towards addressing the problems associated with service evolution:

 

  • I have developed an Eclipse plug-in to completely automate the adaptation of SOAP and REST clients when services change in a benign way and support the client adaptation in case of incompatible changes.
  • I have developed a web application that provides a complete toolkit for WADL-based REST development to support the engineering and evolution of REST-based systems.
  • I revisit the issue of software economics in the era of web services, specifically for the problem of software evolution and I propose a decision-support system. I am using Game Theory, which takes into account all the conflicting interests between all the involved parties (providers, clients, competitors). The novelty of my approach is that it also takes into account the externalities of each party's decisions, which are the indirect effect that a decision may have to the rest of the ecosystem.

 

 
Copyright © 2019 Marios Fokaefs's Home Page. All Rights Reserved.
Joomla! is Free Software released under the GNU/GPL License.