Computers today make it easy for people to scatter copies and versions of digital items across their file systems, but do little to help people manage the resulting mess. In this paper, we introduce the concept of a copy-aware computing ecosystem, inspired by a vision of computing when systems track and surface copy relationships between files. Based on two deployments of a copy-aware software prototype and in-depth interviews with individuals in collaborative relationships, we present our findings on the origins of copies and the barriers to eliminating them, but offer a promising solution based on the set of files that together represent a user‘s conceptual view of a document – the versionset. We show that the versionset is viable to infer, and we draw upon user activity logs and feedback on personalized views of versionsets to distill guidelines for the factors that define a versionset. We conclude by enumerating the many PIM user experiences that could be transformed as a result.