Peer to peer: As the revolution recedes

Once viewed as transforming entire industries, peer to peer is now gaining traction as an unglamorous way to make ordinary business processes more efficient.

John Borland Staff Writer, CNET News.com
John Borland
covers the intersection of digital entertainment and broadband.
John Borland
7 min read
It took a boom and a bust to do it, but peer-to-peer technology is finding its post-Napster place in the world.

Like the Internet itself, peer to peer is settling into a second, more prosaic stage following the bubble of excitement that seemed to stand the world on its head--if only for a moment.

Peer-to-peer networks, in essence, provide a way to link PCs together without the need for powerful central server computers. When Napster, the best known peer-to-peer player, surged in popularity last year, entrepreneurs and media pundits boasted that the technology, which allows people to search for and retrieve files from individual computers around the world, would transform the Internet. But peer to peer is now gaining traction primarily as an unglamorous technique for making ordinary business processes more efficient.

Some big projects are still in the works, backed in part by the financial and technical muscle of powerhouses such as Intel, Microsoft and Sun Microsystems. But the biggest buzz--around peer to peer's ability to transform the way computers work together--has passed, making it difficult for smaller companies and developers to finance plans that aren't tied to immediate business returns.

"There was definitely a hype bubble that grew and burst," said Kelly Truelove, an independent peer-to-peer technology consultant and co-author of a recent book on the subject. "It shed a light (on the technology), and the light faded away. And now people are continuing to do what they were doing without the light."

For a few companies, that means carrying the torch where Napster left off by giving individual people on the Web the power to link their computers together in the sprawling file-swapping networks that became the bane of the record companies. Some, such as StreamCast Networks and Kazaa, continue to operate in the face of massive lawsuits.

But more of peer-to-peer development is happening in less controversial circles, as companies such as Groove Networks or NextPage push peer-to-peer ideas into business circles. The companies, as well as financial analysts, are looking at peer to peer as a way to capture the potentially huge cost-efficiencies that models like Napster provide.

To date, these have been enticing but hard to quantify. One recent Bear Stearns report estimated that Napster would have had to pay nearly $7 million a month for bandwidth costs alone if it had been hosting all the songs its members were downloading at its peak. Such examples are a powerful attraction for companies looking for ways to cut costs.

Power to the people
Peer-to-peer technology wasn't new when Napster exploded into Internet circles in late 1999. If anything, the technology's spread after that date was as much a social change in the way computers were regarded--both in and out of programming circles--as it was a genuine technological leap forward.

What Napster and similar software programs did was to let ordinary PCs play the same role as the powerful Web servers that companies had previously needed to put content online. By downloading and installing a simple program, 14-year-olds in suburban basements around the world could put their own music collections online, let millions of other people search and download their own songs, and tap into those other millions of collections just as easily.

Gartner analyst Simon Hayward says the complexity of the issues raised by peer-to-peer as an IT architecture will likely lead to many false starts.

see commentary

Napster and its 60 million-plus users have been muzzled by court decisions. But many file-swappers have moved over to alternative networks such as Kazaa and MusicCity Morpheus, where movies and software programs are also swapped. The pressure of those services--and the fear that there will be others behind them if they are shut down--remains a significant factor in record and movie companies' plans for online services such as Pressplay and MusicNet.

The lawsuit faced by the corporate parents of the Morpheus, Kazaa and Grokster services may wind up being even more influential than that which ultimately choked off Napster's file-swapping. Defenders of those services, which have less ability to control what happens on the decentralized network formed by their software, say the court can't apply the same ruling it did to the more centralized Napster.

The record industry disagrees and is still trying to shut the companies down. If the industry does succeed, it will make it considerably more difficult for consumer file-swapping services to operate as a business, either through distributing software or serving ads into a peer-to-peer network.

A few companies with roots in this consumer file-swapping insurgency continue to hope they can make a business out of legal file distribution, however.

CenterSpan, the company that bought Scour's file-swapping technology in bankruptcy court, is routinely held up by Microsoft as a way that so-called digital rights management technology, which blocks most unauthorized copying, can be added into peer-to-peer systems. However, CenterSpan doesn't yet have many big media clients willing to take it up on its promise of secure distribution.

Napster itself is scheduled to launch early next year a secure version of its file-trading service, using only music authorized by record companies. Funded almost wholly by entertainment conglomerate Bertelsmann, Napster hopes to lure enough of its previous user base into a subscription model to help pay its corporate protector back.

The open-source Gnutella file-trading network, based on technology created by America Online's Justin Frankel, continues to be a wild card. Companies such as LimeWire are pushing the file-swapping network's technological boundaries, allowing it to evolve into something that could be used to tap into ordinary Web sites or corporate databases, as well as PCs. But the number of people using the technology has halved over the course of the last six months.

"Right now, we're at the point where this is in its infancy," LimeWire CEO Mark Gorton said. "There isn't going to be any radical change in the next few months, but hopefully it will be able to grow from the grassroots over time."

As a cost cutter
Other companies are hoping to market peer-to-peer services to content companies as a way to sharply cut their distribution costs.

A new generation of companies, including Kontiki and Red Swoosh, has sprung up in the online content distribution business now dominated by Akamai Technologies. Kontiki, founded by a group of Netscape emigres, has drawn the most success so far with customers including Sony, Amazon.com and TiVo.

While the new peer-to-peer companies have so far survived the shakeout that has taken down early hopefuls such as Flycode, they've still got a difficult financial task ahead of them. Funding for such companies remains minimal or nonexistent, and even the record-company sponsored ventures are predicting low subscription revenue in early years of operation.

More money and effort appear to be flowing into business peer-to-peer efforts.

Groove, a company created by Lotus Notes founder Ray Ozzie, has blazed something of a trail in this area. The company's software takes the original Lotus ideas a step further, using peer-to-peer technology to allow people on different computers and in widely dispersed areas to work together on the same project.

It has attracted some big customers already, including a 10,000-seat license from pharmaceutical giant GlaxoSmithKline. In perhaps Groove's biggest endorsement, Microsoft invested $51 million in the company in October.

Created before Napster's rise, Groove was never wholly comfortable in the peer-to-peer spotlight, and Ozzie spent considerable time trying to publicly separate himself from the trend. The attention to the sector did help people think about the model Groove was proposing, though, and that's now evolving further, the company said.

"If peer to peer as a meme is dying, I think the decentralization meme is gaining momentum," said Richard Eckel, a Groove spokesman. Particularly since the Sept. 11 terror attacks, companies and government agencies have been interested in ways to keep projects moving in widely dispersed regions, he said.

A few other companies are posting solid revenue and clients based on business peer-to-peer services. Utah-based NextPage, for example, has drawn several big law firms, professional service companies, and accounting firms to its corporate content distribution software, drawing kudos from analysts.

These models are still in their early stages, but market researcher Gartner recently estimated that as much as 20 percent of business interactions would be "facilitated" by peer-to-peer technology.

The big guns
A further sign that peer-to-peer technology may have real staying power comes from the backing of deep-pocketed and influential stalwarts such as Sun and Intel, each of which have been among its biggest evangelists in the corporate world.

Intel sees peer-to-peer systems as a way to boost chip sales because they shore up demand for personal computers, which has been flagging.

The company took an early lead in creating a group designed to set standards for the new technology model. After some early criticism from independent programmers afraid of undue corporate influence, the Peer-to-Peer Working Group has settled down to a few narrow tasks such as figuring out how different pieces of software can talk to one another through corporate firewalls.

Sun's involvement has been more ambitious, creating a new open-source development project called Jxta, intended as a platform for an array of new peer-to-peer Web services.

Sun itself plans a set of services to run on top of the Jxta technology, which ultimately could help it compete with Microsoft's .Net plans. Sun acquired Infrasearch, an early peer-to-peer search technology company, but has not yet outlined what other types of applications it hopes to launch.

Development on the basic Jxta platform is moving slowly, and some developers say that interest has fallen in recent months. Nevertheless, the technology is moving forward. Outside applications based on the platform, including Web streaming and file sharing, have been written by independent developers.

In an open letter to Jxta developers this month, Sun Vice President Mike Clary said the company would continue to "sponsor and actively participate" in the project, despite the lack of revenue.

But that kind of open-ended project, companies say, is in the minority as people now look for revenue, not revolution, in peer to peer's future.

Companies, Groove's Eckel said, "are less willing to investigate a new technology or a new way of doing business without specific return on investment in the end."