Though peer-to-peer networking remains the darling of those who hold that information wants to be free, the market for P2P software itself still faces a lot of challenges. Not the least of these, of course, is the fact that the MPAA, the RIAA and just about any other organization of intellectual-property holders would like nothing more than to see P2P squashed forever.
Napster became the first victim of the copyright cops' ire. Though wildly popular, its service relied on company-run servers, making it an easy target. Shut down the central servers, and you shut down Napster. Later P2P architectures like Gnutella and FastTrack avoided this vulnerability, however, by being fully decentralized. So long as users ran their software, their file-sharing networks would continue to exist.
Or so it seemed. Now, however, divisions within the FastTrack network — once the most popular file-sharing platform after Napster — appear to threaten its future stability. Ironically, FastTrack's demise seems likely to come not from outside pressures, but from in-fighting within the network itself.
Track Stars
FastTrack is the brainchild of 35-year-old Swedish programmer Niklas Zennström and his partner, 25-year-old Janus Friis. In April 2000, the pair founded the Amsterdam-based company to market two products based on their own proprietary P2P technology. The first was Kazaa, a decentralized file-sharing client application targeted at consumers. The second (and more esoteric of the two) was the FastTrack P2P Protocol Stack, to be licensed to developers.
The FastTrack Stack was a "library," a prepackaged bundle of code that allows other programmers to easily take advantage of the core networking technologies that Zennström and Friis developed. For a fee, people were free to build their own application that could connect to the same P2P network as Kazaa. There would be no favoritism; the various programs would compete in an equitable, open marketplace.
Zennström's idea took off, and soon several companies were deploying the FastTrack Stack in their own products, including Grokster out of the West Indies and Tennessee-based StreamCast networks, creators of the Morpheus file-sharing client. Unfortunately, however, FastTrack's popularity caught the attention of music-industry attorneys, and soon FastTrack, like Napster before it, found itself embroiled in its own legal problems.
In February 2002, in an apparent attempt to avoid legal pressures in the Netherlands, FastTrack underwent reorganization, changing its name to Kazaa BV. In a confusing twist, it simultaneously sold the rights to the Kazaa.com domain and the Kazaa client software itself to an obscure Australian company called Sharman Networks. This move left Zennström's Kazaa BV with only one product: the FastTrack protocol library.
These developments should have had little effect on the FastTrack network's users. Zennström's various licensees were all perfectly capable of operating without the active participation of Kazaa BV. But then something unexpected happened. In March 2002, users of Morpheus, the most popular of the FastTrack clients, suddenly found themselves shut out of the network.
Can the Center Hold?
Though the FastTrack network is technically decentralized — meaning no one company controls it — a seemingly innocuous feature of its protocol stack had allowed Kazaa BV to trigger the blackout that struck Morpheus. Zennström's company had released an upgrade to the P2P libraries, which by design was automatically "pushed" to all participating client software. There was just one problem. StreamCast Networks hadn't licensed the new version, meaning Morpheus users were left out in the cold.
StreamCast CEO Steve Griffin isn't taking the blackout lying down. In a statement on StreamCast's MusicCity.com Web site, he describes the effects of Kazaa BV's upgrade as "an unprovoked attack" on the Morpheus software, suggesting that it may violate U.S. federal law. Other statements on the site point the finger at Sharman Networks as well. Immediately following the blackout, Sharman began trying to poach StreamCast's user base with a new marketing campaign touting its Kazaa client software as an alternative to Morpheus. Griffin feels this has the earmarks of an anticompetitive conspiracy against his company.
For its part, Sharman Networks denies Griffin's allegations. Sharman's publicist points out that the company has no involvement with Kazaa BV's development of the FastTrack Protocol Stack, and therefore could not have been behind the forced upgrade. The company counters that StreamCast was itself responsible for a denial-of-service attack against Sharman's Web servers — a phenomenon that Griffin does not deny, but attributes to a side effect of Kazaa BV's spontaneous FastTrack protocol upgrade.
On March 4, Zennström stated his own view of the blackout in a letter to CNET's News.com, explaining that StreamCast hadn't received the latest FastTrack upgrade because it had failed to pay licensing fees to Kazaa BV. StreamCast doesn't deny that it had been unable to reach a licensing agreement with Zennström's company. Still, Griffin claims that the resulting unilateral shutout of the Morpheus client was a violation of his existing contract for the FastTrack libraries.
Breaking Ranks
His battles with Kazaa BV and the FastTrack network ongoing, Griffin saw little choice but for StreamCast to jump ship. In March, StreamCast begin offering Morpheus-branded client software based on Gnutella, widely regarded as FastTrack's top competitor for the P2P file-sharing market. Gnutella already enjoys a large user base, owing to commercial offerings from such companies as BearShare and LimeWire.
Unlike FastTrack, however, no single company owns and operates the Gnutella network. Libraries and descriptions of the protocol are freely available, allowing open-source developers to write new client software without paying license fees. It was one of these open-source projects, called Gnucleus, on which StreamCast based the preview edition of its new, Gnutella-enabled Morpheus client.
Market leader LimeWire has also released the source code to its client software, and sees the openness of the Gnutella protocol as one of its greatest strengths. "We specifically don't like the closed-protocol nature of FastTrack," says LimeWire CTO Greg Bildson. "Morpheus's experience with FastTrack shows the danger of using a closed protocol when you don't control it."
This open model has worked well for the Gnutella software market, which unlike FastTrack has so far managed to maintain an air of cooperative competition. When Bildson heard that StreamCast would be entering the field with its own Gnutella client, he got in touch with Griffin's programmers to make sure the transition was as smooth as possible.
"Expanding the use of Gnutella has been our goal since day one," Bildson explains. "We believe that Morpheus's adoption of the Gnutella protocol is one step further along the path in making the Gnutella protocol ubiquitous."
In the Limelight
Ubiquity of the Gnutella protocol is an important concept for LimeWire, and it's contributed much toward that goal. While early versions of the Gnutella technologies were hampered by limited scalability, LimeWire has been working aggressively alongside other developers to address these performance problems. After all, for a decentralized P2P network of this kind, the more users the merrier — especially when it comes to expanding the applications of Gnutella beyond simple file sharing.
"[Morpheus] brings our long-term goal of expanding Gnutella into the world of general information and content searches one step closer," Bildson says. What this means is that consumers may someday be using P2P technology for far more than trading stolen MP3s. In LimeWire's vision, Gnutella may one day become a vehicle for information retrieval on any subject, from legal software downloads to product data on the new line of Ford trucks.
"Basically, we believe that a search protocol like Gnutella will evolve as a common protocol on the Internet," says Bildson. "We believe that there will be 10 to 100 million users. Further, we believe that these users will evolve (with a little help from us and other client developers) to do searches of all types using Gnutella."
If that goal succeeds, LimeWire may finally have found the killer app for P2P file sharing that might save it from almost certain doom at the hands of intellectual-property lawyers. Especially with more and more traditional search engines blurring the boundaries between information retrieval and advertising, the market for an open P2P alternative could be huge. In that possible future, the winners will include not just LimeWire users but StreamCast's installed base of Gnutella-enabled Morpheus users as well.
Meanwhile, the fate of the embattled FastTrack network seems bleak. Kazaa BV has lost one major customer with the departure of StreamCast, and its Web site no longer offers any information about its product, the FastTrack Protocol Stack. Ultimately, the closed, proprietary nature of its protocols seems likely to be FastTrack's undoing. For all its pro-file sharing rhetoric, it seems to have forgotten one basic premise: Maybe some information really does want to be free.
