It's starting to look as if the end may be near for the "peer-to-peer" fad.

Peer-to-peer, or P2P, is a decentralised approach to computing and the Internet whose best-known incarnation is Napster, the music downloading service. Last summer, there was an explosion of interest in the new field from Silicon Valley programmers, entrepreneurs and venture capitalists.

Some said P2P--a somewhat amorphous term that means different things to different people--represented nothing less than the future of the Internet, and would soon be used for all manner of services, from Yahoo-style searching to eBay-style auctions. Millions of dollars were invested in P2P, scores of companies were founded and an ocean of ink was spent by the press to chronicle the burgeoning movement.

Overlooked in all the excitement, though, were a few problems. No one knew whether P2P really worked, at least in the way some theorists were proposing. Or whether there were many uses for it that didn't end up violating copyrights, the charge leveled against Napster. Or if there were, whether companies could make any money on them.

Now, with technology funding in a funk after the bursting of the Internet bubble, those problems suddenly don't seem so small anymore. And the P2P party, which once looked like an exception to the dot-com downturn, seems more like a wake.

Among the recent setbacks:

• InfraSearch, a widely publicised Web-search company whose founders once boasted that P2P would be as big a business as the telephone, was sold last month at a fire-sale price after it was unable to raise more money. Last summer, a group of prominent investors, including Netscape co-founder Marc Andreessen, poured millions of dollars into InfraSearch, which quickly became the blue chip of the P2P world. But Sun Microsystems snatched it up for US$10 million--a sum that doesn't even register on the Richter scale of Silicon Valley deals--and then promptly folded it into an existing research-and-development project.

• Another well-known P2P company, Popular Power, closed completely last month after also striking out in raising funds. "We got caught in a bad time," sighs co-founder Marc Hedlund.

• And then there are companies formerly known as peer-to-peer purveyors, which now say they no longer like the term. "We've listened to the market and repositioned ourselves," says a spokeswoman for Engenia Software. A spokesman for Groove Networks, the company founded by Lotus Notes creator Ray Ozzie, says, "We clearly rode the P2P wave, but the sooner we move the conversation beyond the technology, the better."

The mad rush

Silicon Valley watchers say sociology as much as technology is needed to explain the rapid rise and fall of the P2P fad.

"The whole P2P hoopla is emblematic of the mad rush to find the next big thing in technology," says Bill Burnham, a venture capitalist with Softbank Venture Capital. Edward Jung, a former top software architect at Microsoft, said P2P boosters were trying to recreate the excitement of Napster by simply transferring Napster-style technology to other areas. But since Napster is more of a social than a technical phenomenon, "a lot of people were led down the primrose path," Jung says.

One of the difficulties in assessing the state of P2P is semantic. Most people define the term as a computing scheme in which information is stored on many PCs, reducing or eliminating the need for a central repository like a Web server. In the market, however, the phrase has come to describe at least three entirely distinct categories of technologies and companies. None of them, though, are playing out quite as well as boosters had once hoped.

The first category is Napster and the Napster clones. Purists, for starters, note that the music service is only partially P2P because it relies on a central database to show which users are sharing music files.

Napster, of course, has well-known legal problems. A new crop of Web sites, such as Lime Wire and BearShare, let users trade free music and other files by using a purely P2P software system known as Gnutella. But it remains to be seen if these companies can escape a record industry crackdown down the road, and if they do, whether they can turn themselves into profitable businesses.

There is another group of Napster-like companies calling themselves "legal Napsters" that say they are trying to heed copyright laws and charge for music swapping. But without free music, it's hard to get anyone to log on.

Flycode, a Napster-like system originally named AppleSoup and unveiled to great fanfare last July, had just a few hundred users logged on Tuesday afternoon, many of them swapping material that was either X-rated, copyrighted or both. "There has to be a value proposition to get P2P members to aggregate, and other than music, people haven't yet identified a good one," says A. Joon Yun of Palo Alto Investors, a venture-capital firm that holds a stake in Flycode.

Distributed computing

A second flavour of P2P involves what has long been called "distributed computing," a technique in which hundreds or thousands of computers work together to solve a big problem, such as cracking an encryption scheme. That can be a useful setup, but is a different job than a Napster-like scheme of passing music or other files from consumer to consumer.

Intel, ever eager to put millions of its microprocessors to good use, has been using P2P in the distributed computing sense since last summer. Tuesday, the company announced plans to employ millions of PCs in a philanthropic cancer-research effort with partners that include the American Cancer Society, the National Foundation for Cancer Research and Oxford University. The effort involves having PCs in homes and offices use their idle moments to research the cancer-fighting properties of millions of molecules.

A number of companies have also been started in this part of the P2P market. But they are finding it can be slow going. Popular Power, for instance, folded. United Devices has been around for a year but so far has just three customers--though co-founder Ed Hubbard promises that many more are on the way, saying many companies are interested in it for biotechnology research as well as Web development work.

The final P2P category involves companies that want to use P2P to solve problems besides sharing music files, such as InfraSearch's Web searching. Most of these companies, though, aren't purely peer-to-peer; instead, they use P2P ideas combined with traditional approaches. A number of start-ups, for instance, are working on software to let people collaborate on office projects, or run programs over the Web as well as on local PCs. These companies include Groove Networks, as well as Jung's company, OpenDesign.

But many of these have been labeled "P2P" for the scarcely remarkable fact that their programs will occasionally direct two computers to talk directly with each other. Executives of the companies admit they went along for the P2P ride for public relations reasons.

"P2P is going to be used very broadly, but by itself, it's not going to create new companies," says Michael Tanne, chief executive of Xdegrees, a start-up using P2P ideas as part of a way to more easily locate files on the Web. "It's a technology. But the companies that will become successful are those that solve a problem."

Copyright © 2009 CBS Interactive, a CBS Company. All Rights Reserved.
ZDNET is a registered service mark of CBS Interactive. ZDNET Logo is a service mark of CBS Interactive.