« Google Print | Main | Taking Freenet Seriously: A Response to Picker on Peer-to-Peer »

October 04, 2005


TrackBack URL for this entry:

Listed below are links to weblogs that reference Peer-To-Peer: What Is It Good For?:

» What is P2P good for? from madisonian.net
Randy Picker asks: Peer-To-Peer: What Is It Good For? Check out the interesting posts (here and here) and comments over the new U of C blog. I agree with Tim Wu (see the comments) that p2p is a powerful enabling technology with some valuable curren... [Read More]


Feed You can follow this conversation by subscribing to the comment feed for this post.

Tim Wu


The case for P2P designs is easy. Here are highly useful P2P designs that you haven't mentioned:

- Email (SMTP has a peer design -- ever notice there is no one central email server for the world)
- Instant messaging
- VoIP (Skype is built on a P2P design)
- BGP Routing (that is, the decentralized system that keeps track of IP addresses - very similar to Kazaa)

The real question isn't P2P. It is whether decentralized filesharing systems are useful for much. And in that case I'd say yes, but in the way Bittorent is used-- for transferring really large files when server space is at a premium.

For example, I download my public domain Japanese lessons using Bittorent.

Randy Picker


You have to say more, I think. I really did have in mind distribution of content (and not personalized communication). That is what the Grokster case focuses on and where all of the action is. So I don't think that email, IM or VoIP are part of the same ballgame. Those are classic send-receive communication situations (where now it can be multiple send and multiple receive), but these aren't ways in which broad-interest content is distributed.

So I am back to what I said before; I was surprised to find that the photo-sharing websites weren't p2p; I just assumed that they were, and then I looked at flickr, and nope, highly centralized. Same for Project Gutenberg, which, given when it started (1971 and long before the PC age let alone p2p) is hardly surprising.

Again, I want to be precise: not generalized p2p design issues, but distribution of broad-interest content.

Yes? No?

Tim Wu


My point was that the basic idea of a P2P, or decentralized idea, is a pretty general one, and includes email, VoIP, etc.

As for mass blind content distirbution software. Well let's see. In a world without law, Bittorrent would still be useful for as a cheap way for distributing giant files.

That makes it useful for people who want to distribute films, operating systems, or other giant files without paying a fortune for a server farm & a T-3 connection.

In that sense, it changes the price for mass distribution of giant files. Pre-Bittorent, that cost a server-farm -- post Bittorent, its in the price range of ordinary people.

That's one reason you don't see it at the photo sharing sites. The files aren't large enough to justify it. But if that were a public domain film hosting site, it would have to be Bittorent or very well funded.

Tim Wu


I also want to make the same point I made in reaction to your "Sony as Entry Policy" paper. (actually you made it too). Namely, you are focused on t0. But what about uses in t1?

Imagine you had asked me "what is the WWW useful for" in 1991? I would have said, "it has more graphics than gopher, but otherwise its useless." Hence a rule making the WWW illegal in 1991 wouldn't have seemed like a big deal.

For example, this Blog itself wasn't exactly what HTML inventors had in mind, let alone TCP/IP's inventors.

It seems that the important thing is future, and possibly hard-now-to-describe uses for a system like Bittorent. At bottom these are systems that give users some of the powers that server farms have now, and I cannot help but think that will be useful over the long run.


I think it's important to distinguish two separate senses of "peer-to-peer." One is cases in which only the distribution mechanism is decentralized. The other is the case where, in addition to distribution, the choice of files available and the navigation/searching mechanisms are decentralized.

The former case is unambiguously useful. For that, we only have to look at BitTorrent, which is used widely by a variety of open source projects, video games, etc, for distributing large files. Bandwidth is getting cheaper, but it still costs a non-trivial amount of money if you're offering files that measure in the hundreds of megabytes. What BitTorrent does is drastically reduce those sites' bandwidth costs by letting the downloaders share the file with each other rather than all downloading from the centralized server.

Skype is another example of this first design. There's still a centralized registry of user accounts, but the voice traffic is distributed in a decentralized manner.

Grokster is an example of the second category. It's not clear to me that Grokster served any useful purpose not served by products in the first category like BitTorrent. The primary benefit of Grokster's design was that they thought it might prevent them from getting sued.

But the first category is unambiguously useful because it economizes on bandwidth. That allows content to be offered to consumers at lower cost than would otherwise be possible.

I expanded on this argument back in March:


William Rothwell

"But if that were a public domain film hosting site, it would have to be Bittorent or very well funded."

As an example of this principle, when a game demo is released (often several hundred megabytes), one option is to go to a website like http://www.fileplanet.com where users must either have a paid subscription or wait in a line until bandwidth becomes available. A free, faster, and easier alternative is to distribute the file via P2P, which is the philosophy that sites like http://www.filerush.com use.


If I may interrupt this rather esoteric discussion and perhaps introduce you to my real-world experience as someone who shares files over the Internet. I would refer to myself as a "file sharer" and not what is more commonly known as a "file trader," a distinction I will make clear presently.

I've been sharing files over the Internet since the early Napster days when there were centralized servers. Currently I'm using WinMX 2.6, p2p. software, which I consider to be one of the better softwares available. It's an older version, WinMX 3 .54 beta, which is the newer version which I have but don't use since this software is much more easily tracked. Among file sharers in the know, older software has become the order of the day since the newer designs attempt to include limitations, allow for the tracking of IP's and make adjustments for the law.

Mostly I freely share files with anyone who has a legitimate need. I won't go into how I define legitimate need but let's just say that in my disclaimer I mention things like public domain and academic and research purposes. Let's just say that my screening process is rather liberal.

When someone wants to download a file from me I do not expect reciprocity, which is something that many "file traders" state up front in their disclaimers. In other words if someone downloads a file for me I do not expect to obtain one in return which ethically I would view as an exercise in underground commerce. For the most part I share freely with anyone who meets the basic criteria. While the law may make no legal distinction in these two methods of exchanging files, it's important to me for my own moral perspective, a perspective focused upon the difference between sharing and trading, commerce and free exchange, acquisition and the public good.

If I may be permitted to pose a question, perhaps you'd be willing to share some of your legal expertise with me. Specifically I am interested in knowing my legal position with regard to music which I and my family members, down through the years, have purchased the limited rights to.

For example my mother bought the Crosby Stills and Nash Album, "Four Way Street" sometime back in the early 1970s, and this album was passed on to me. Later in my life I purchased the same album on cassette tape. Of course I've repeated this experience with a large number of artists, purchasing and repurchasing their music in various formats. Of course at this point most of those early formats are now out of date. Currently I have acquired that same music free of charge from other file shares in digital format. Specifically my question is how many times must I repurchase this music before I have a continuous or perhaps better described as a contiguous right to possess it in any format?

After a few years of downloading music in the late 90s, I had basically reacquired all the music I'd purchased through the years plus everything new I was interested in, since whenever I heard something that I liked on the radio or elsewhere I could almost immediately find it online. I recently I calculated that if I had to go out and purchase all the albums or CDs with all of the music that I have in my library, the cost would be around $12,000. So given my economic limitations, without filesharing I would surely have to do without this music, no doubt many people find themselves in a similar position. Personally I'm extremely thankful for file sharing because I love music.

These days I guess I would be branded a pirate, someone who engages in illegal commerce, but somehow I don't feel like a criminal. In fact I have little or no moral problem with this whatsoever, and I consider myself a relatively ethical person.

Recently with the decision in the Supreme Court but much more pertinently with the adoption of The Patriot Act, there's been a crackdown on filesharing activity to some extent, although law-enforcement efforts have been little more than a drop in the proverbial bucket compared to the amount of activity that's going on in the US and worldwide. (Interestingly on that note, I've shared music with people in dozens of countries around the globe, from Japan to Australia to Croatia, England, South Africa and the list goes on. Music connects people in ways that perhaps nothing else does.)

That said, I know of individuals who've had their servers and computers confiscated. I also have a friend who is a federal agent who has clued me into some of the activities which are being practiced by different arms of federal law enforcement. Apparently these days file sharers are considered a potential threat to national security, not to mention law enforcement's increased interest in the sharing of pornography under of the specific direction of the Bush administration.

Apparently large amounts of data in the form of IP addresses and identities are being compiled by law enforcement officials with the help of ISP's like Comcast cable who oftentimes don't disclose this activity to their customers. Much of this data is being turned over to third parties like the MPAA, RIAA, Universal, Fox, Sony etc.. This activity is of course worrisome to those such as myself, leading us to post disclaimers in our shared files section quoting the Internet Privacy Act specifically code 431.322.12 in the hopes of staving off the possibility of winding up in a lawsuit.


From a practical point of view, having movies and music in digital format loaded onto your hard drive is extremely convenient not only for pleasure and for sharing, but for doing research and being able to access information quickly.

In addition to sharing movies and music, I also share a number of books in text format, as well as audio books and photos. But for the most part things like the complete Works of Shakespeare, Marcus Aurelius, William Blake or the Hagakure, are in much less demand then say a movie like Tom Cruise's War of the Worlds. Go figure.

From a philosophical standpoint I find some rather palpably satisfying justice in sticking it to the record companies who've been sticking it to me and the public my whole life. I find much of the activity that the music industry engages in to be morally reprehensible from a number of perspectives.

For decades they have forcing people to pay an exorbitant price for an album/CD with a whole bunch filler disguised as music in order to further their idea of an efficient business model, forcing me and everyone else to pay for something we don't want in order to get what we do want. Nor were industry officials interested in exploring other models while they had control of formats. Only when they were faced with losing that control did they even start to consider exploring other avenues for the dissemination of music.

And I won't even go into the particulars of the recording industry's blatant exploitation of artists since the recording of music began. Wealth equivalent to the total net worth of many small countries has been concentrated in the hands of an extremely small group of people who not only did not create the music that enriched them, but have often been responsible for the repression of artists who don't fall into their preconceived notion of what music is marketable. I have little sympathy for this kind of exploitative and repressive capitalism.

What peer-to-peer sharing has done is virtually broken the lock that this small number of individuals has maintained on an art form and given it back to the people. I think history will bear out that what's happening now will be a truly liberating thing for artists and musicians who over time will reach audiences heretofore unknown, audiences that are now growing at an exponentially increasing rate.

In that vein, I've been exposed to a whole host of music and artists that I would've never known about otherwise, thereby expanding and enriching my musical experiences and those that I share with, while simultaneously expanding the audience of these artists. In all likelihood I and others will wind up patronizing these musicians through live events if in no other way. And don't get me wrong I do believe that people have a right to protect their intellectual property and earn from what they create, but I think the way artists go about doing this is destined to change.

From the business perspective, personally I'd like to see a huge portion of the music industry, specifically that section which is focused on manufacturing and marketing a product presented as popular music, collapse in upon itself from its own bloated complacent self-perpetuating weight and be replaced entirely by a new business model. Perhaps a model that focuses on expanding the market to include a much wider variety of artists and musicians as well as growing the worldwide audience for those artists to include more than just people who can afford to spend $20 on a CD, a small minority in the larger global picture.

Interestingly in today's music business (something I'm somewhat familiar with), musicians have become almost an afterthought. The people who perpetuate this kind of approach to marketing music have only the most narrow understanding of what diverse audiences are interested in or are capable of embracing. To my mind their day of reckoning is long overdue.

Additionally there so much worthless product coming out of Hollywood in the form of movies that I regard the downloading of currently in release films as something of a necessary screening process. Given the state of today's market, I see it as a necessary form of consumer protection. If everyone were able to prescreen films, the quality of every aspect of the commercial film industry would, out of necessity, be influenced for the better. I still go to see the movies that are worth seeing, whether I download them or not, but now I'm able to weed out that 90% of the utter garbage which is being produced by the movie industry at a discount using poor writers and poor directors, not to mention poor actors, and is then foisted upon an unsuspecting public through the use of a vastly overdeveloped marketing industry (in comparison to the actual creative artistic side of the equation).

No I don't have much compassion for charlatans and hucksters who would sell me a substandard product at a premium and then laugh all the way to the bank. Make no mistake that filesharing is helping to put pressure on the entire industries, helping to weed out the worst parts of those systems.


It is important to distinguish between the different technologies being discussed here.

There are different models of networks in which users can transfer files between each other. These networks can, like Napster, be built around a centralized system. Others, like Grokster, do not require a centralized server. Still others, like eDonkey and BitTorrent, do not require a centralized server AND rely upon distributed computing techniques.

eDonkey and BitTorrent allow a user to download different portions of the same file from different sources. At any given moment, a user may be receiving different parts of the same file from many people. A user can also distribute portions of the file to others without having the entire file. These networks are not really peer-to-peer in the sense that they allow one user to download a file directly from another. This raises interesting questions about potential direct liability for distributing copyrighted works. Since I can get a full file from 10 people, each of whom may not have the full file themselves, who is liable for direct infringement? If there is no direct infringement, there can be no contributory infringement.

Skype relies upon a distributed network like eDonkey and BitTorrent.

I would not include technologies like Skype, instant messaging, and email in the same group as P2P networks, though some of them may have traits of P2P networks. To do so takes away the meaning of P2P and makes every computer-to-computer connection P2P. P2P is a form of file distribution; file distribution is not a form of P2P.


I was always under the impression that P2P involves direct connections between computers that are not servers. That's why email doesn't really count as p2p. If we get to a world where everybody's got IP6 and always-on boxes, p2p is irrelevant because it's all servers...until then, an app like Skype that routes calls through a bunch of ephemeral dynamic-ip boxes would qualify by this definition.

Puppet Master

>>not generalized p2p design issues, but distribution of broad-interest content

If that is what you meant, say that, *in* the post. By giving a title "Peer-To-Peer: What Is It Good For?" and the first sentence as "Absolutely nothing", you may have got the eyeballs of readers, but you have lost the accuracy in the post and come across as someone who does not get the whole picture.


You put together the words "fee" and "copyright" in your closing paragraph but I think that those two concepts don't always go together. Right now I can think of several examples:

- Comercials (who will pay to see a comercial?)
- Movie trailers and teasers
- Open source software (it's not public domain)


According to wikipedia a peer-to-peer computer network is a network that relies primarily on the computing power and bandwidth of the participants in the network. How does this relate to a T1 Line?

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been saved. Comments are moderated and will not appear until approved by the author. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Comments are moderated, and will not appear until the author has approved them.