Exploiting user-group relationships for increasing concurrency in software engineering environments

Software engineering environments (SEEs) emerged in order to address the problem associated with developing and maintaining large software projects. Databases in SEEs store and manage the various software artifacts that result from the different phases in the software development cycle. The traditional notion of atomicity and serializability is too restrictive in SEE databases. This is because transaction models in SEE databases need to take into account long-duration activities that may last from days to months. Users need to share partial results of their ongoing activities as it is undesirable to keep other users waiting indefinitely. Here we address the concurrency issue in SEEs. We present a locking protocol that exploits relations among groups of users for achieving enhanced concurrency in SEEs