The Internet Archive, a non-profit online library dedicated to the permanent preservation of information in digital form, has made nearly a petabyte of materials available via the controversial BitTorrent peer-to-peer file sharing protocol.
For the project's launch on Tuesday, founder Brewster Kahle announced that the Internet Archive was hosting more than a million torrents, the small files that point to data accessible by BitTorrent clients. The collection has grown since then, and now comprises nearly 1.5 million torrents.
The majority of those torrents – more than one million – point to digitized books, while the others point to audio and video files.
Although BitTorrent has become a popular method of distributing large file collections, such as Linux distributions, historically it has more often been associated with illicit sharing of copyrighted content. According to the Internet Archive, however, all of the material available in its BitTorrent collection is in the public domain.
"The Archive is helping people to understand that BitTorrent isn't just for ephemeral or dodgy items that disappear from view in a short time," John Gilmore, founder of the Electronic Frontier Foundation, said in the announcement. "BitTorrent is a great way to get and share large files that are permanently available from libraries like the Internet Archive."
Unlike many BitTorrent sites, such as the Pirate Bay, which only host torrent files but not the actual digital content they point to, the Internet Archive is also hosting all of the original content for which it makes torrents available.
That means BitTorrent is now the fastest way to download files from the Internet Archive, Kahle says, because the distributed nature of the BitTorrent protocol allows clients to download data from two of the Archive's servers at once, housed in two separate data centers.
Kahle says the peer-to-peer protocol should also make downloads more reliable from locations with slower connectivity to the Archive's servers, including outside the United States. ®