A Conceptual Framework for Fine-Grained Quality Assessment in Version Graphs

Today’s software development and modeling processes happen collaboratively. A key challenge in collaborative software engineering is assessing the quality properties of the different system revisions developed by different people. Modern solutions for version control, variation control, and model management can track and relate versions, feature variations, branches, and experimental modifications. We particularly observe a trend in the development of model management tools to support increasingly fine-grained evolution operations and artifact relationships. This detailed knowledge of a system’s revision history enables history-based quality assessment. However, the current state of the art lacks a comprehensive terminology framework for clearly describing the scope within which a particular quality attribute is assessed in the revision history. This leads to inefficient communication and hinders tackling the actual objective of the quality assessment. This work contributes a formal framework for quality regions in revision management systems. Our framework is based on revision graphs and incorporates concepts such as versions, variants, merges, multi-systems, and views, building a common ground for fine-grained quality assessment of evolving systems. We provide an open-source library implementation of our formal concepts, which can be used to extend existing version control or model management tools.

Second International Workshop on Model Management (MoM 2025)