An architecture for modeling and applying quality processes on evolving software

Software process and product views should be closely linked in order to better manage quality improvement. However until now the two views have not been effectively synchronized. Current approaches to Software Configuration Management (SCM) are strongly based on files and lacking in logical and semantic understanding. Some impediments faced when modeling and analyzing software evolution include additional effort for dealing with language dependent source code analysis and continuous mining of the evolving system. By leveraging features offered by modern VMs and other enabling technologies, we have developed a language neutral architecture with extensibility mechanisms to support continuous Software Evolution Management (SEM). Our research aims to contribute to an SEM infrastructure where semantic artifacts can be consistently accessed, tracked and managed for performing software evolution analytics beyond the file-based model. This paper presents compelling factors for our infrastructure, the architecture we have developed, and then sketches a case study to demonstrate its application.