Date Added: Jan 2012
The potential of peer-to-peer(p2p) networks has been hobbled by the authors' lack of understanding of how to design robust algorithms for large-scale networks. Networks that are truly p2p are currently very unsophisticated computationally: they are only able to provide the functionality of very basic data structures such as hash tables or skip lists. This limits their use to problems of storing, retrieving and disseminating data. This lack of sophistication is in spite of strong motivation for building p2p networks that do non-trivial computation.