Feedback Analysis in Software Product Line Forked Developments
David Romero-Organvídez, Óscar Díaz, Yutian Tang, David Benavides · 2025
Software Product Lines (SPLs) enable the reuse of software components or assets to generate a family of related products. Sometimes, an SPL evolves into parallel developments (forks) to meet new requirements. However, these forks do not always stay synchronized with the central development, e.g., features can be added, removed, or changed in the forked projects. In DevOps practices, feedback analysis plays a central role in improving both software quality and delivery processes. DevOps feedback analysis evaluates data from the delivery pipeline and users to improve the software and its deployment process continuously. Despite its importance, feedback analysis has been underexplored in the context of SPLs. In this paper, we propose an approach to automate feedback analysis in forked software product line developments that can allow us to assist decision-making processes in answering questions such as: Which features need more testing? What new features can be incorporated? Which ones require refactoring? Which ones cause more issues in production? Information can be gathered from various data sources such as source code repositories, bug tracking systems, or continuous integration pipelines. To the best of our knowledge, this is the first proposal using information from forked SPL developments for feedback analysis.