How does P2P work?

I do not use P2P software (e.g.: KaZaA) and am pretty much clueless about how it works. Let’s say that I did get such a program, would it let people just snoop around my harddrives willy-nilly or what?

I am not looking for advice in where to get P2P software or how to use it. Please, let’s keep that out of discussion so the thread can stay open.

You specify which directories to share. If you’re dumb, you can open up your whole hard drive.

I have not heard of any complaints about snooping so I doubt if it’s possible under normal circumstances. You didn’t ask how these file-sharing programs work but I just read an article recently so I’m puttin’ it in.

All this info is from the April 2003 issue of Smart Computing Magazine.

P2P stands for peer to peer

We know how clients and servers work on the 'net, right?
Client requests and server provides.
P2P works the same way.
In P2P clients are called nodes.
Each node is equal on the network.
A node can act as a client and a server.

Napster model: Node A makes request to Master Catalog server. Server searches its list and finds the file on Node B, Node C, and Node D. Server sends list back to A with Nodes B, C, D on it. List contains connection speeds of nodes B, C, D so Node A can choose which one to download from. Noda A decides on Node C and communicates directly with C to obtain file. Napster model not true P2P because of Master Catalog Server.

Gnutella model, used by BearShare, LimeWire, Morpheus: A makes req. that’s sent to list of nodes A has, which are B, C. Neither of B, C have file. B, C both send req. to D, E, which are on their lists. The file is on E. A downloads from E. This is truest P2P model.

FastTrack model, used by Kazaa and Grokster: A makes req. that goes to Supernode B. No match. Supernode B sends req. to Sn C. Sn C finds file on node D. A is informed and contacts D for download. A P2P hybrid because of supernodes.

How P2P works.