Towards a Model of Computer Systems Research

Peer to peer systems have become a popular area of distributed systems research over the past few years, reflecting both the widespread use of systems like BitTorrent [5] and the emerging commercial potential of peer to peer systems for a variety of applications well beyond file sharing [4, 14, 20, 21, 22]. This line of research has resulted in substantial progress in understanding system behavior, as well as better techniques for constructing more robust, more scalable, and more efficient systems. For example, workloads, churn, and available resources are all heavy tailed, and this is fundamental to understanding aggregate system behavior in practice [5, 17]. Modeling peers as rational, sometimes altruistic, and occasionally byzantine agents [1] is essential to building systems that are both more robust and more efficient [16, 14]. And randomness is widely used in peer-topeer systems as a fundamental design technique [20, 5]. In this paper, we turn our attention to another peer to peer system that has received less attention from the systems research community: the systems research community itself. Our approach is intentionally tongue in cheek, but we observe many similarities, at least on the surface, between peer to peer systems and the systems research community. For one, there is relatively little central coordination! Rather, progress is reached through the mostly independent actions of individual researchers, interacting mainly through the conference publication system. 1 Ci-