In this project, an alternative approach to developing collaborative applications will be investigated. In this approach (called "incremental specification and experimentation"), the collaboration infrastructure provides only the most essential functionality by executing "meta protocols." During the course of collaboration, new protocols are discovered and specified as they emerge. Both meta and emergent protocol snippets are plugged, unplugged, and replaced on the fly. The project includes: (1) domain modeling to find natural representations of meta and emergent protocols; (2) multi-user authoring tools to support end-user incremental specification of emergent protocols; (3) plug-n-play of protocols to support incremental experimentation; and (4) execution, monitoring and awareness mechanisms to provide the necessary context for coordination. This proposed research investigates incremental specification and experimentation of synchronous and asynchronous protocols in the same framework and attempts to develop more end-user oriented support. The results, when applied, are expected to improve the flexibility, and thus the usability and development/deployment costs, of a variety of groupware applications. The development of a new graduate course on "collaborative systems and models" will give students concrete experience in experimentation with collaboration protocols. In undergraduate courses, students will work in teams using collaboration technologies to identify and analyze their collaborative activities and the underlying "meta protocols" implemented in those systems that they would like to change. Through a protocol modeling process, students will develop analytic skills and begin to understand the dynamic nature of cooperative work and the importance of flexibility in the design of collaboration infrastructures.
