Towards Accurate Conflict Detection in a VCS for Model Artifacts: A Comparison of Two Semantically Enhanced Approaches

In collaborative software development the utilization of Version Control Systems (VCSs) is a must. For this important task some graph-based VCSs for model artifacts already emerged. Optimistic approaches, which are nowadays the designated ones, allow parallel editing of one resource and therefore changes can result in conflicts and inconsistencies. To be flexible for the ever increasing variety of modeling environments and languages VCSs should be independent of the modeling environment and applicable on any modeling language. Those VCS characteristics implicate a lack of information for the conflict detection method by virtue of firstly receiving solely the state of an artifact without concrete editing operations and secondly due to unavailable knowledge about the semantics of a modeling language. In such VCSs inconsistencies would even arise more often. Hence, accurate conflict detection methods are indispensable for the realization of optimistic, environment and language independent VCSs. This can be achieved by providing some understanding about the models’s semantics which is possible by specifying machine interpretable formal semantics. Therefore, in this work, a comparison of two semantically enhanced conflict detection approaches is presented with respect to their suitability for the integration in an optimistic, environment and language independent VCS for model artifacts to achieve more accurate conflict reports.