Install limewire linux

No matter which side of the music-sharing fence you're leaning against, you can't argue with Napster's popularity. The application's enduring legacy, however, may involve more than just music. Under the hood, Napster (www.napster.com) uses peer-to-peer (P2P) technology, which deviates significantly from the client-server design that has governed the Internet from the medium's first days. P2P is not new, but Napster's popularity has fueled a great deal of interest in the technology.

What's the Difference?

Your PC, when connected to the Internet, is referred to as a client. The same term also describes most programs you use to do things on the Web: your browser, e-mail software, newsreader, and so forth are all clients. The salient feature of a client is that it doesn't operate independently; to be functional, it must connect to a server—a central source of data or instructions. Every time you use client software to interact with the Net, you're sending a request to the corresponding server software and waiting for the server to respond.

P2P doesn't work that way. Instead, two (or more) computers are linked for the purpose of sharing data, each taking an equal role in the data-transfer process without the intervention of a central data or instruction source. Your home network is probably a peer-to-peer system; no server is needed to share information among machines.

Not all P2P systems call the individual computers peers. The Gnutella file-sharing service, for example, refers to each host as a servent. This term itself is telling. An amalgam of server and client, the servent designation implies that P2P doesn't so much do away with the server as incorporate the server functions into the client. The idea of a remote server that mediates the flow of data disappears. Instead, each user's computer handles the serving, although those functions are more or less hidden from the user.

When you install a program such as Napster, you're simultaneously turning your machine into a server that other users can access over the Internet, and enabling your machine to access other computers running Napster that are connected to the Net. To function, Napster does require a central system (run by Napster, the site), but that system's functions are to determine current user IP addresses and offer an index of files available, not to intervene in the actual PC-to-PC communication.

As noted earlier, sending data over the Internet from one computer directly to another isn't new by any means. Windows users have long had access to freeware or shareware software that enabled them to set up their own machines for access by others. To exchange files directly with a friend, for example, both of you could run FTP (File Transfer Protocol) server software, set up a username and password for the other person, and give that information along with your current IP addresses to each other. (To find your IP address on a Windows 9x system, you'd select Start | Run, then enter winipcfg.exe and click OK; under Windows NT or 2000, you'd enter ipconfig.exe.) Each of you could then log on to the other's system and get files. As you probably suspect, however, this approach is clumsy, inefficient, and almost totally useless if you want to share files with more than a few people.

P2P, on the other hand, lets computers connected to the Internet communicate with each other directly. Napster lets you search for music files on the computers of Napster users everywhere on the Net and, at the same time, makes the music files on your computer available to the rest of the Napster community. Some P2P applications let you share not only files but other resources as well.

What's wrong with centralized servers? First, they cost a lot of money. Servers require reliability and durability beyond what most workstations need and, as a result, must incorporate relatively expensive technologies such as SCSI, disk mirroring, and sophisticated operating systems. Moving applications off servers and into a P2P environment can make a lot of sense financially. Servers are not needed for all applications. They're essential for hosting a popular Web site, certainly, because they can dedicate hard disk space, network bandwidth, and administrative resources to this multifunction and resource-hungry task. But simpler tasks such as limited file sharing don't require servers.

Finding One Another

Avoiding the use of central servers means that P2P applications have to establish different ways of working with IP addresses. Every computer on the Internet must have a numeric IP address; in fact, having an IP number is one of the definitions of being on the Internet. For ease of use, this IP number is typically referenced through its corresponding domain name (pcmag.com, for example), and keeping track of the correlation between IP numbers and domain names is the responsibility of DNS (Domain Name System) servers. P2P avoids dealing with DNS servers either by having PCs address one another using IP numbers only or by using a proprietary means of obtaining current IP numbers.

This avoidance is necessary for two reasons. First, DNS servers rely on fixed IP addresses—IP numbers that are assigned to individual computers and never change. Such is not the case with the majority of surfers who access the Net via a dial-up line. At the moment of connection, these machines are assigned an IP number from a block of available numbers the ISP may use (this is called a dynamic IP address). Second, all computers on the Internet have IP addresses, but not all have domain names, and in those cases DNS servers are of no use.

Because of their ability to work outside the DNS system and outside the system of centralized servers, P2P applications are said to operate at the "edges" of the Internet. They use the essence of the Internet—network connectivity, the TCP/IP protocol, and the requisite IP addressing system—but skirt the rest.

What Can You Do With P2P?

By far the most widespread application of P2P has been file sharing, with Napster as the leader, of course. The RIAA (Record Industry Association of America) suit last year seemed, for a time, destined to deep-six the enterprise, leaving only the memory of a brief period of Internet anarchy. But the company survived and has formed an alliance with media firm Bertelsmann AG to develop a pay service (latest reports suggest a fee of about $5 per month) that would give subscribers unlimited access to music released by Bertelsmann Music Group. Since forming this alliance, Napster has seen lawsuits dropped by Germany's edel music and U.S. independent label TVT Records. As of this writing, both companies are cooperating with Napster on setting up the fee-based service, although the upstart's future is still uncertain.

Despite Napster's success, Gnutella (http://gnutella.wego.com)—a search protocol, in essence—is a more interesting P2P project. Designed to be completely free of central authority, Gnutella can be incorporated into any file-sharing application. There is no official client and the name doesn't refer to a company or even a program, creating a very elusive lawsuit target. Gnutella was touted as the technology that could supplant Napster's dominance in music file exchanges, but the real advantage is the ability to find and download any type of file. The Gnutella protocol does not actually handle file transfers, employing HTTP for this function instead. Each servent becomes an HTTP client and server together, allowing two-way transfers.

Gnutella-based servents are growing in number and ease of use. A look at one download site, www.gnutelliums.com, showed servents available for Windows, Linux, Java, and the Mac. Three of the most popular Windows variants are BearShare (www.bearshare.com), LimeWire (www.limewire.com), and ToadNode (www.toadnode.com). Another servent, Newtella (www.newtella.com), focuses in a Napster-like way on music files. Recent improvements in the protocol have contributed to new enthusiasm, servent upgrades, and even interest from the recording industry and other commercial sectors.