What is P2P, or Does Anyone Remember Napster?

P2P (peer-to-peer) Communications Model
Peer to peer is direct file sharing between two computers over the internet, who have assigned IP addresses, without downloading from a central server. In essence, in a true peer-to-peer network, the client/server roles are merged and there is no central server (Wikipedia). So each computer then acts as a server from which any other computer can download files, within the network. A little bit about networks, as Robert Molyneux explained in his essay "Computer Networks", most networks are "partial mesh". In other words, not every computer is directly linked to every other computer. The model actually looks like this:

That hazy cloud that is the internet is where we can focus what P2P means in terms of networking. It is almost as if the internet allows for P2P to directly connect every computer. What once seemed improbable, connecting every computer to every other computer, becomes simpler. P2P allows this connection. Because of the power and speed of home computers these days, the next evolutionary step in computer use should be the merging of client/server technology. More home computers could be used in a way that is similar to the way servers are used. This sounds fabulous, right? The sobering thought for all this peer to peer networking is security. With KaZaA, the user had the eventual problem of piggybacking. When downloading files, malicious files would piggyback onto legitimate files and your computer would be harmed in some way by this. Because the network was unstructured, there was no way of regulating this. Also, KaZaA ran a program on the computer called a satellite. This satellite often kept the internet connection open, allowing for programs to be downloaded on your computer without your knowledge. Limewire, one of the newer peer to peer applications, seems to have better security as it is firewalled, and those firewalls can be adjusted.

Bit Torrent
Perhaps the newest and 'safest' peer to peer technology is bittorrent. Bittorrent was created by programmer Bram Cohen and is now maintained by BitTorrent Inc. Bittorrent is an open source application, the code was released to the public. Any client is capable of, "preparing, requesting, and transmitting any type of computer file over a network using the BitTorrent protocol". Bittorrent is the name of a peer to peer file distribution protocol, which means it is a new rule for file-sharing. An interesting bit of information, some people record that 35% of internet traffic is through bit torrent applications. This may be high, but it is still interesting. The way bit torrent works is that the user downloads the executable bittorrent file from the home page. Then, the user searches a separate bit torrent client, like Azureus (available from sourceforge.net) or isohunt. When they've located the file they want, they click on it, and it opens up the bittorrent application. The interesting thing about bittorrent is that the more files you share, the quicker the download. This is called 'seeding'. A 'leech' is a person who downloads more than uploads. This seems to encourage some of the tenets that early web developer Tim Berner's Lee abides by, which is decentralization, openness and fairness. By encouraging people to allow uploads, there is more information for everyone out there, not one person hoarding all the files. Bittorrent is also notable because it distributes this large amount of data widely, without using up huge amounts of costly bandwith. The concern here, as with all peer-to-peer clients is authorized use. Companies encourage the use of bit torrent to download open-source software. As for music, perhaps appropriately copyrighted material found in Creative Commons could be downloaded through Bit Torrent. Apparently Warner Bros. plans to distribute some films via bittorrent.
P2P and Libraries: Past, Present and Future
Libraries were instrumental is helping to keep P2P networks available. According to ODLIS

In September 2003, the American Library Association (ALA) joined four other library associations in an amicus brief on behalf of peer-to-peer file-sharing companies Grokster and Morpheus in their defense against an infringement suit brought by MGM Studios and 27 other entertainment companies to ensure that file-sharing technology, which can be used to benefit society without infringing intellectual property rights, is not unduly restricted.

Since patrons are not typically allowed to download software onto library computers, this battle was truly to protect freedom of information outside the library, qualifying the use with the idea that peer-to-peer sharing is alright only when it does not infringe on intellectual property rights.

Libraries now recognize the advantages of P2P networking for their own use. From the Encyclopedia of Information and Library Science:

"How will P2P networking affect the delivery of reference services in and by public libraries? The effects are likely to be manifest on two levels. First, it is expected that groupware applications based in significant part on P2P architectures will emerge as important and widely employed services for computer-supported work, and one of the dimensions of computer-supported work is that reference librarians and other information specialists will be integrated into computer-supported workgroups, in a manner akin to the way in which medical librarians have been absorbed into clinical care groups at many hospitals and medical centers. The second dimension of effect will be seen in the way P2P networking enables reference librarians to deal more or less simultaneously with the needs of individuals and groups;"

Personal Experience
My experience with P2P networks are with the unstructured P2P networks. Three previously popular file-sharing networks, such as Napster, Gnutella and KaZaA were all unstructured P2P networks. Napster allowed users to download files from other users. Napster also spread the use of the compressed mp3 format, which then allowed the audio files to be downloaded more speedily. With dial-up being the most prominent connection, these compressed files were ideal. The search engine would then flood the network with the search term looking for the file. The more popular the file, the more likely the user would be to find it. Herein lies the problem with the unstructured networks. Rather than the network running through a list of files on one central server and locating it, the search goes anywhere and everywhere, sometimes timing out because there are too many files out there to find one less popular individual file. I now use Bit Torrent, it's faster and seemingly safer. There are still timeouts, and files that won't download. Such is the price we pay for free information!