Search Slyck  
Bram Cohen on the BitTorrent Entertainment Network
February 26, 2007
Thomas Mennecke
Font Bigger Font Smaller
Today marks a day of transition for BitTorrent, Inc. Although always on the map towards legitimate acceptance, the launch of the BitTorrent Entertainment Network is perhaps the most significant step forward since the company made amends with the movie industry in November of 2005. Since that time, BitTorrent, Inc. has been slowly building a massive catalog of free and pay content.

But things weren’t always that way for the BitTorrent protocol. When the ball first started rolling for Bram Cohen, the CEO of BitTorrent, Inc., the BitTorrent community had its fair share of growing pains. When BitTorrent was first released in late 2002, it was released in a P2P environment dominated by FastTrack (Kazaa) and eDonkey2000. However the BitTorrent protocol had a distinct advantage.

Around that time, broadband penetration in the United States was just beginning to pick up steam. With faster connections, the marvel of downloading at 5 kb/seconds lost its luster and instead the Internet community began demanding more speed and faster transfers. Along with that demand came ever-growing file sizes: 700 megabyte XviDs, 800 megabyte ISOs, and 4.4 gigabyte DVD-Rips. Although the largest P2P network at the time, FastTrack, did have the ability to share large files, it did so very poorly. FastTrack was designed to transfer small MP3 files, not large movie and program files.

Where FastTrack failed, BitTorrent proved greatly superior. By its very design, it was built to transfer large files efficiently and to greatly cut back on corrupt or false files that plagued other networks. Greater community interaction on BitTorrent trackers allowed users to verify the authenticity of a file prior to downloading, in addition to a much stronger hashing algorithm – providing it’s from a trusted source. However with the strong community interaction, trusted sites became common and while corrupt files did not.

With corrupt files a thing of the past, and fast download speeds with an ever-growing catalog of media more commonplace, it didn’t take long for BitTorrent to supersede FastTrack as the largest file-sharing community. There’s still debate about whether BitTorrent has surpassed eDonkey2000, however ostensibly this appears to be the case as this protocol has been frequently cited as consuming more bandwidth than any other network.

As BitTorrent grew, it faced similar dangers as any other file-sharing community: the entertainment industry. The entertainment industry, lead by the RIAA and MPAA, was dismantling the corporate P2P at a steady rate. FastTrack was readily dissected by the RIAA’s lawsuit campaign, while MetaMachine, iMesh, BearShare and Grokster were all forced to settle for various amounts (between 4 million and 30 million.) Yet BitTorrent was the exception. Unlike most other commercial file-sharing ventures, BitTorrent has avoided the costly litigation that has annihilated any realistic commercial competition. Is it because of its open source nature? Perhaps its advanced distribution technology? Maybe Bram Cohen’s and Ashwin Navin’s undeniable charm convinced the entertainment industry? Could it be a combination of these factors?

It probably is. BitTorrent’s distribution mechanism is virtually unparalleled, except perhaps for eDonkey2000/Kad, and quickly became the de facto method for distributing large files online. Without a central contact for eDonkey2000/Kad, this was not a viable business option for the entertainment industry. However BitTorrent incorporated itself and the process towards legitimacy was in full swing. Constant dialog with the entertainment industry manifested itself in November of 2005, when an understanding between the MPAA and BitTorrent, Inc. was announced. To this day, BitTorrent, Inc. has continued to move forward with its authorized distribution platform, unencumbered by any litigation.

BitTorrent, Inc.’s announcement today ratchets up the playing field once again for the authorized distribution market. With an already impressive array of free and pay content, the BitTorrent Entertainment Network looks to expand and translate the protocol’s dominance into the authorized distribution landscape. Can it succeed? It’s already demonstrated a certain level of success with it’s established catalog, and with the addition of MGM Studio’s catalog, it stands a chance of cornering the market.

However apprehension remains – especially from the audience that BitTorrent and all other authorized distributors intend to woo away from free file-sharing. Especially of paramount concern is a concept known as DRM (Digital Rights Management). DRM has inhibited iTunes and other music stores from reaching their full potential, and indeed has put a damper on sales. It’s clear the public wants the unrestricted ability to transfer files and to enjoy music/movies/etc when and where they want. Bram Cohen and BitTorrent, Inc. appear to realize this, as both Bram and Ashwin Navin have spoken out against DRM in the past.

To understand the latest from the BitTorrent, Inc. crew, Bram Cohen agreed to an email interview with

Slyck: From the end user's perspective, BitTorrent appears to have changed little since the advent of DHT. What argument do you have against that, what is the next evolutionary step you see for the BitTorrent protocol?

Bram: The Distributed Hash Table is a very cool piece of technology which we are currently using in some limited but productive ways. With BitTorrent, it's always better to rely on centralized tracking, and the BitTorrent tracker protocol I made is so lightweight that you basically always can. But DHTs provide resiliency when those central resources fail and that's how we're deploying the DHT at the moment.

In the near future you'll see increased functionality in BitTorrent's core technology. Streaming media, more robustness, improved tit-for-tat, improved network politeness, and more efficient bandwidth allocation are just a few of the major developments we have underway. We have a great team of P2P engineers employed at BitTorrent working on these things, and if there are engineers in the P2P community interested in taking BitTorrent to the next level they should email us at jobs AT bittorrent DOT com.

Slyck: You've expressed dislike of protocol header encryption and other methods of circumventing ISP traffic shaping in the past. If the latter starts affecting a significant percentage (say 1/3) of BitTorrent users, is it possible that the protocol may start to see defensive changes as a result?

Bram: There are some ISPs which shape traffic because a small fraction of users take a disproportionate share of the ISP's bandwidth capacity -- BitTorrent for example is used by 8-10% of users but reportedly takes 33-50% of all bandwidth on the ISP network.

However, as the user base of BitTorrent and popularity of video sites like YouTube grow beyond just a small fraction of users, the average user is using the network a lot more and will continue to do so in the future. As a result, traffic shaping by ISPs is short-sighted and increasingly just bad policy, because users expect true broadband speeds and the ability to use the popular applications which drive broadband adoption in the first place. I don't anticipate adverse treatment of BitTorrent users to become all that widespread, particularly in broadband markets where there's competition and consumer choice for broadband ISPs.

People here have misunderstood my position on obfuscation. I'm against protocol obfuscation not out of any interest in helping ISPs deny their users service. Users should get exactly what they pay for including unfettered access to use BitTorrent. The problem is that obfuscated traffic is only slightly harder to identify than non-obfuscated traffic, and creates a huge number of technical headaches. Encryption by definition makes stuff hard to read, and it causes a lot of problems for development, debugging, and network monitoring, all of which are completely reasonable things to do and have nothing to do with traffic shaping.

Slyck: What are the development plans for µTorrent, and where does it fit into BitTorrent, Inc.'s plans?

Bram: µ is going to remain as it is and will continue to be a place where people can get the most up-to-date version of µTorrent software, including beta versions with more experimental features. will continue to offer a client which is very stable and doesn't include any features which haven't been extensively road tested. µTorrent is a solid implementation and we will continue to make it better. The new client will feature the best of µTorrent (small size) and the best of Mainline (protocol innovation). Moving forward, you'll also see developments related to BitTorrent embedded on silicon and non-PC hardware thanks to the new codebase.

Slyck: What's your take on network neutrality?

Bram: 'Network neutrality' means different things to different people. I'm against internet censorship, in particular against anything like the FCC running the Internet. I'm also against ISPs being able to charge their users for the ability to access particular Web sites. Market forces are currently doing a good job of keeping Internet access a widely available commodity, and regulatory intervention could easily cause damage to the very thing it was supposed to help, especially with the vagueness of currently proposed legislation.

Slyck: Canadian ISPs have been less than friendly to the BitTorrent protocol, with Shaw recommending the end user set his/her upload speed to "1" KB/sec. What steps has BitTorrent, Inc. taken to smooth over relations between especially hostile ISPs (such as Shaw) and BitTorrent?
(ref: )

Bram: We're talking to ISPs about how they can lighten bandwidth load on their systems and provide better service to their users. In the future, all large, high-quality files will be delivered with peer-assisted architectures, involving both uploading and downloading. This will be so mainstream that it will be part of the basic functionality which consumers expect from their service provider. Unfortunately for consumers in Canada, they don't have much competition between ISPs, and we encourage every one of our users to call their ISPs let them know how they feel about this issue.

Slyck: Online flash video has taken off. Would it be feasible to build a tiny flash BT/P2P client in which each viewer of a video is also a server? This could conceivably help video sites with bandwidth problems.

Bram: Peer-assisted delivery of streaming video is definitely in our product roadmap. We're working on a product which can save a lot of bandwidth money for viral video sites and allow them to deliver higher quality content at the same time.

Slyck: Tell us a bit about the BitTorrent Entertainment Network. What will separate it from the dozens of other already existing authorized distribution sites?

Bram: We're launching what will be the BitTorrent Entertainment Network at with much more entertainment content than any of our competitors in the download business. So if you want to find a movie, TV show, or a PC game, and you need it in a hurry, your best bet is to go to We've also got some of the lowest prices around including a lot of compelling free content, all with a simple, straight-forward user experience. Best of all, we give our community and indie artists a way to contribute their creations through our self-pub capability.

Slyck: Do you feel the hostile position some ISPs take towards BitTorrent will hamper your business model?

Bram: If ISPs keep their users from using basic Internet services, it hurts the consumers who aren't allowed to use the applications they find compelling. Thankfully consumers are very sensitive to this, and in most markets, ISPs are kept in check by competition and angry customer service complaints.

Slyck: Mark Cuban had some less than pleasant words for your business model. One of his bigger criticisms was that more mainstream TV content, such as "Prison Break" wasn't clearly available - yet was readily available through your search engine. How do you feel the launch of the BitTorrent Entertainment Network challenges his seemingly valid argument?

Bram: Mark Cuban wasn't making a coherent argument in the post you refer to, he was just engaging in childish sniping at me in a comment he made on my post on my LiveJournal. Anyhow, we have every title which he mentioned in that comment, including "Prison Break", available in our new catalog which we are now launching, so he was just plain wrong. In fact, we have one of the most comprehensive video title catalogs available anywhere. It's bizarre how people continue to take that guy's opinions seriously.

You can find Prison Break episodes by searching for them on our site, like this:

Slyck: Explain how rented/purchased items are transferred via Does it initiate from a central source, and then use peer assistance when enough copies are distributed? What's stopping someone from downloading an authorized movie, and redistributing it on a less-than-scrupulous tracker? How does renting work, along the lines of Napster-to-Go?

Bram: We're doing all the tracking and original seeding. As is always the case in BitTorrent, if only one person is downloading then they get the content from the original seed, but if lots of people are downloading then most traffic comes from peers. In all cases download rates are good.

We're using Windows DRM, which is required by most of our content partners, to prevent unauthorized use. Basically this means that the files we distribute are encrypted, and when a user tries to play them in Windows Media Player it hits our servers to get instructions for decrypting the file and the terms under which it can be used. We're using post-licensing, a very nice feature which isn't standard practice these days. Post-licensing enables us to have a rental period start when the user first plays the content, rather than when the download starts, which makes for a far less rushed viewing experience.

Slyck: What are your thoughts on DRM, and does it make you uncomfortable knowing that it exists on

Bram: DRM causes major usability problems for all users, which is a major issue in and of itself. Some of the usability issues are fundamental, and some are limitations of the currently available systems, but as Steve Jobs obliquely indicated in his recent statements about DRM, a lot more of the problems are fundamental than the people who insist on DRM schemes for their content are willing to admit.

The other issue is one of technical viability - can DRM fundamentally work? For many games, the answer is essentially yes - some part of the game play requires interacting with a server, and permissioning can simply be done there. For media files, things are fundamentally different. They're stand-alone items, containing no programmatic instructions of their own, which must by their nature be displayed in a raw copyable form to be used. Under that threat model, nothing short of hobbling the user's computer to be a DRM player first and a computer second is capable of stopping a DRM system from getting cracked for long. Even with that hobbling, a little bit of external hardware (for example, videotaping the computer screen) can bypass the limitations anyway.

Slyck: Do you envision a day when, and perhaps the Internet as a whole, will be free of DRM?

Bram: For media files, we're going to see an ongoing decline in the usage of DRM, both because it hurts the user experience and because it's largely ineffective at keeping content from getting pirated. I hope that we have no DRM content on in the not too distant future; in no small part because it would make for a lot less technical headaches and it allows us to address a larger market opportunity.

Slyck: What are your thoughts on Joost? Do you see them as a significant competitor considering their recent deal with Viacom?

Bram: Joost is almost like a client application of YouTube, featuring streaming downloads with a proprietary interface. I'm not sure why this is a better experience than YouTube, in fact it's probably worse. We solve a much more basic problem and focus on high-quality, downloadable content with a standard Web site. We also have a vastly deeper and wider catalog and more plans on how to embrace and encourage community participation on our site.

Slyck: Although BitTorrent's authorized lineup has significantly expanded, the question remains, why should people choose legitimate DRM content over free, especially considering the risk and morality factors are negligible at best?

Bram: Several reasons. We have a much larger catalog than you can find on any pirate site. We also have a simple, reliable user experience which is accessible even to non-technical users. Also, our pricing is some of the lowest around, comparable to renting a DVD, for example.

Slyck: Any closing thoughts, observations, comments?

Bram: I'm very excited about this new launch and proud of all the hard work everyone at BitTorrent has put into it to make this possible. I'm also excited because the site itself is heavily influenced by and programmed for the BitTorrent community of users. The first version is only the beginning, and we'll be adding a lot more content and features in the near future.

Editor's note: It's hard, if not impossible, to expect the entertainment industry to jump feet first into a DRM-free environment. Whether the "live free or die" file-sharing community agrees with Bram's assessment that BitTorrent, Inc.'s catalog is "much larger" than any pirate site, and whether they agree it's worth the price of admission, is a matter yet to be resolved. In the meantime, it does appear the entertainment industry is continuing its baby steps forward into the file-sharing foray. However baby steps have yet to keep up with the great leaps that file-sharing technology has provided the demanding Internet public.

This story is filed in these Slyck News categories
BitTorrent :: BitTorrent Inc.
Authorized P2P :: Other

You can discuss this article here - 34 replies

© 2001-2020