Blame Napster. The buccaneering website opened a legal can of worms by making it possible for users - some 20m of them at its peak - to share music files directly over the internet. But at the same time it raised awareness of the peer-to-peer (P2P) networking technology that makes Napster-like file sharing possible, and provided incontrovertible evidence that it can work on a grand scale.

The result has been a resurgence of interest in the potential of P2P networking for use in enterprise systems.

It can be thought of as direct, two-way communication between computers over a communications network - which could be the internet - without the need for any one computer to be designated as network manager. There is no hierarchy in a P2P network.

It is, as yet, early days, but this year should see an acceleration in the commercial development of P2P applications such as file sharing, team working, instant messaging and videoconferencing.

Note, I say "resurgence". P2P is not a new technology. Small offices have shared resources such as files, printers and scanners on a simple P2P basis for a decade or more.

The original internet was based on P2P principles. Established as an information network for academics in the 1970s, it was fundamental to its operation that files could be shared quickly and easily between participants.

There was no distinction between "client" computers and "servers". The machines were left constantly online, ready to send or receive requests for information. Only academics (and the military) would at that time have had easy access to systems of this kind.

This pattern changed with the growth of internet use by businesses and individuals who created a demand for web browsers (software which made it easy to search the internet for information, often of a random nature).

Browsers, however, are based on client-server technology. Clients request information; servers deliver it. The movement of information is essentially one way. Adequate hardware and sophisticated network management are critical.

Now while Napster and similar programs such as Gnutella have raised the profile of P2P networking, current interest in the technology is be driven by two factors. First, an understanding that personal computers have the power and telecommunications systems the bandwidth to make direct communication between systems a viable proposition. Napster's 20m users makes that point.

Second, a realisation that applications which would be difficult or impossible using client-server technology work more simply peer-to-peer. There are a number of business benefits: the costs of establishing the network can be low because the existing infrastructure is used. The only limitation on the size and kind of file that can be shared is the available bandwidth.

A further advantage is the potential savings which can be made by maximising the computer power and storage space which already exists within a company's information systems. A larger-than-corporate-life example of this is the Search for Extraterrestrial Intelligence (Seti) project at the University of California at Berkeley.

More than 2m volunteers have opened their computer systems to software which downloads astronomical observation data and processes them at times when the computers would otherwise be idle. Commercial software houses have already developed products to enable companies to exploit the latent power of their systems.

The Seti project illustrates some of the problems of P2P. Users have to be prepared to open their systems and their bandwidth to the rest of the network. The dangers are obvious. Ian Clarke of Edinburgh University has pioneered P2P software called Freenet which allows information to be distributed through a network with virtually perfect anonymity for the sender. Nobody knows on whose disk the information is eventually stored. There are no clues to the sender and no way of deleting the information.

The growing popularity of P2P, nevertheless, indicates that the internet is set for transformation once again. It also illustrates a more basic theme: that the development of computer networking is a constant struggle between the processing power and bandwidth available at the centre of the network and at the edge.

The emergence of peer-to-peer technology for today's applications indicates the balance has shifted to the edge. Tomorrow's infinitely more complex applications will probably see the centre hold sway once more.