This paper tries to solve the problem of protecting the privacy of both producers and consumers of information while avoiding a single point of failure by decentralising all network functions.
The main contribution of this paper is introducing Freenet, a distributed information storage and retrieval system designed to solve the problems of privacy and vulnerability given above. This problem is still important today because computer networks are rapidly growing in importance as a medium for information storage and exchange. The lack of privacy of content distributors and readers is of growing concern, and the centralized server model has an inherent single point of failure. The authors develop Freenet by focusing on five design goals: anonymity for both producers and consumers of information; deniability for storers of information; resistance to attempts by third parties to deny access to information; efficient dynamic storage and routing of information; and decentralization of all network functions.
Computer networks provide insufficient privacy to their users. Centralized databases create central points of failure. The Freenet network is an effective provider of anonymous information storage and retrieval that is distributed enough to prevent centralized attacks.
This paper rates between a 3 and a 4. Although novel, the authors' solution doesn't prove its viability as a basis for a large-scale information network.
The authors' methodology is fairly convincing. They lay out their five major design goals, describe each network function and discuss how it meets the design goals, and provide some performance data based on a simulation of an early version of the system. The authors themselves acknowledge that more realistic simulation is necessary which models the effects of inserts taking place alongside requests, nodes joining and leaving, variation in node capacity, and larger network sizes. The authors also acknowledge a weakness of their system in that it is impossible to tell exactly how many users there are or how well the insert and request mechanisms are working.
Since this system is a work in progress, the authors leave open many improvements to the implementation of Freenet. The authors describe future goals of implementing a simulation and visualization suite which will enable more rigorous tests of the protocol and routing algorithm.