CAREER: Advancing Dependable Reusable Change Integration Across Software Variants

Modern software development routinely creates a copy of an existing software project and customizes it to meet different user, organizational, or operational needs. This practice, known as clone-and-own, creates a copy of the software, commonly called a fork. While many forks are temporary, some evolve into software variants that continue evolving independently and in parallel with the original software, adding features, fixing bugs, and responding to changing requirements. As software variants diverge, important improvements often remain in only one variant, forcing developers to repeatedly rediscover, adapt, and validate changes that have already been implemented elsewhere. This duplication increases maintenance costs, delays the adoption of critical fixes, and can reduce software reliability and security. The project's novelties are the development of dependable methods that help developers identify reusable software changes, intelligently adapt them across software variants, and support evidence-based integration decisions. The project's broader significance and importance are improving the reliability, security, and long-term sustainability of software systems while reducing duplicated engineering effort. The project also strengthens the software engineering workforce by integrating research into undergraduate and graduate education, mentoring students through a shared research ecosystem, and engaging high school students in software evolution activities. The research develops a unified framework for dependable reusable change integration across long-lived software variants. It combines scalable techniques for discovering and prioritizing reusable software changes, semantic alignment methods that preserve developer intent while adapting changes to different software contexts, and developer-centered workflows that support evidence-based integration decisions. Open-source software serves as the empirical platform for developing and evaluating these techniques, enabling reproducible studies and openly available tools, datasets, and benchmarks that accelerate research in software maintenance and evolution. The resulting methods improve the ability of organizations to maintain large collections of related software systems while preparing the next generation of researchers and practitioners to build dependable, secure, and sustainable software. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.