Novelty without innovation, and where Napster got its design
by Andy Oram
My exposure here has convinced me that the peer-to-peer meme I've been researching is relevant to our era. For instance, a presentation I saw on fail-over servers created (without calling it peer-to-peer) a configuration of peers listening to each other and trying to jump in as soon as one failed. The issues explored in the Peer-to-Peer book were largely relevant here: authentication, presence, avoiding DoS attacks, and so on.
Speaking of peer-to-peer and innovation, I learned some interesting background about Napster from system administrator and free software contributor Pierre Phaneuf last night. He claimed that good old Internet Relay Chat contained all the features of Napster, and that Napster introduced no technological innovations but simply provided those features in a manner that was trivial to use. IRC had bots that let people announce they had certain files, and at any time a chatter could invoke the Direct Client to Client (DCC) protocol to obtain the files. It was just a pain to get working. Napster provided an interface--and not a pretty interface, either--to the file-transfer feature allowed anybody to get the job done. And thus does the world progress, supposedly. Neither Pierre nor I felt that the demise of Napster was a great loss to the public, although we deplore the legal machinations that brought it about.