One of the most common suggestions I receive from Ektoplazm visitors is to open a BitTorrent tracker for free music, ostensibly to decrease hosting costs. There are a number of problems with this suggestion, however, and given how frequently it is mentioned, I figure a full post might be helpful to explain why BitTorrent is not a universal solution.

BitTorrent, as most technically-inclined people know, is the standard protocol for large-scale peer-to-peer file sharing. It works wonders when shared content is in high demand and there are strong incentives to seed. These conditions are realized on private trackers communities primarily dedicated to sharing “pirated” content, particularly those trackers with a ratio system, where the ability to download content is regulated to encourage behaviour beneficial to the community (i.e. seeding and uploading). With a stringent set of rules and guidelines in place, private trackers can offer fast download rates, excellent availability, and a breadth of content unmatched by commercial outlets.

So what could be the problem? Consider the ideal scenario described identified above: strong demand and incentive to seed. Demand for free music from mostly unknown artists is certainly high but it is nowhere near that of mainstream content under copyright. Still, it is not unreasonable to expect that demand would be sufficient for a free music tracker to be viable on some level, with exceptions for older, less popular releases. In truth, it is the incentive to seed that is most problematic. A ratio system won’t work and altruism has its limits. I expect the result would be a poorly seeded tracker and a huge reduction in the amount of downloads.

My forecast has a lot to do with the barrier to entry. BitTorrent requires special software and a certain amount of computer savviness. Not much, admittedly, but the population of people familiar with (or able to use) BitTorrent is necessarily a subset of those familiar with direct downloads. Implementing a tracker with a ratio system would force users to register and log in thereby increasing the barrier to entry and thinning the audience. There is absolutely no way to make BitTorrent as accessible as direct downloads via HTTP. Given that the vanishingly low barrier to entry is fundamental to the value proposition of the Ektoplazm free music portal it is not something I am willing to experiment with.

It is also worth noting that even the most popular private trackers don’t push as much music as Ektoplazm does (within its niche market, of course). Scope out some of the major players and you’ll see that even Shpongle and Infected Mushroom only account for—at most—several thousand snatches. The audience just isn’t anywhere near as large.

Then there are the technical issues. BitTorrent trackers are not permitted on many web hosts (mine included). The reason given is often related to CPU/memory use but I would wager that the association between BitTorrent and the exchange of copyrighted content has a lot to do with it. There are other technical issues but it hardly seems useful to go into the prosaic details given everything else I have raised.

Finally, it is worth noting that I have some experience with operating a BitTorrent tracker. In fact, Ektoplazm began as a BitTorrent tracker back in 2005! It closed in 2007 after failing to achieve traction; only 12,000 full releases were served via BitTorrent in nearly two years of activity. After switching to direct downloads this count blossomed to 100,000 in six months. Three years later the total has exceeded 3,000,000. Part of this is no doubt due to the growth of the site—but that growth would not have been possible had it remained a BitTorrent tracker.

In conclusion, BitTorrent is not a one-size-fits-all solution for all your file sharing needs. Direct downloads are likely to be a much better means of distributing legal content licensed under the Creative Commons, at least for now.