The Volokh Conspiracy

Among the industry insiders who supply the media to pirates are people who don't even know that forwarding their preview CDs and VHS tapes means that they'll end up on the internet. I would think that the best way to cut off the piracy would be to cut off these sources; they're much more limited and easier to track than the people who do the compressing and uploading.

This article seems more myth than reality; but that seems to be business as usual for Wired magazine these days.

The argument in the Wired article could probably be applied to mp3 technology, but only to some of the more commercially well known artists, like U2 mentioned in the article, whose product ("pop music") is market-saturated anyways, with heaving marketing and wide sales.

But the unfortunate truth is that the sheer immensity and diversity of products and forms of software available on P2P (beyond commercial artists' ditties) makes it highly unlikely that even a large percentage of these files the could easily traced to a group of "topsites" or a shady group of criminals that can be tracked down as in a Pulp Fiction crime novel.

There is no 'boogeymen' behind P2P file sharing,
for, if there were, he or they surely would have been caught by now. The fact of the matter, is that software piracy has existed in the electronic world long before CD(RW), P2P, mp3 technology, and even the internet; the internet has simply made accessibility to the pirated software much more easily (hence, "democratization", if you will). And since its inception, industry has fought the problem tooth and nail, with little success.

If only a single triangle could used as a design plan for the piracy problem. A more accurate description of the P2P problem, is not a triangle, but perhaps a serious of related and unrelated triangles, with some horizontal association and vertical association, with little common patterns, nor long term consistency (sites come and go all the time).

I think that the real thrust of the article is that some folks a have taken to appropriating property and making a game of how fast they can distribute it across the world.

"He who refuses to do arithmetic is doomed to talk nonsense."
-- John McCarthy

The model described in the article -- a very shallow pyramid, with all data originating from the apex -- would have a number of characteristics that aren't seen in P2P networks. First, the number of files available would be comparatively small -- thousands, not million. Second, the bandwidth cost of maintaining a node would be very high (the article seems to admit this but waves the problem away, as if the operators of the topsites are evil, but very charitable). Third, temporary outages would affect most or all of the network.

None of these things are true in real life, therefore it is safe to conclude that P2P networks are what they purport to be: peer-to-peer.

Do some node contribute more "original" content than others? Yes. Would shutting down those nodes reduce the level of piracy? Not a byte - if a popular file were unavailable, someone would simply rip a copy and put it online.

Indeed, you don't need the Internet at all to get pirate copies. I bought DVDs of Harry Potter and 8 Mile days before they were released in a little shop on Phuket. Of course, the tsunami wrecked that shop, and likely killed its proprietor, so the MPAA and RIAA can rest easy now.

The article implies that there is a new trend inside the world of software piracy: owners of IP (in the music and movie industries )are willing allow some piracy as long as profits are secured. Such a compromise, however, strikes against the core principles of property ownership. It makes for software that’s both semi public while at the same time semi private--in terms of the information in isolation--meaning the IP in its literal sense--, independent of the medium.

Malvolio, You must remember that in the bittorrent pyramid scheme, the top tiers only need to transmit a small amount of the data only once very few machines. Each chunk of the data then resides on separate machines.

The actual activity for these top tier machines then becomes relatively light. Once you have transferred your chunk to the next layer down, you are done. This is the beauty of it. No one machine has all the data except the end user and each serving machine has a relatively light network usage.

In the darknet paper, which should be required reading for discussion of P2P file trading (not that I agree with all of the conclusions, but its simply outstanding anyway: http://crypto.stanford.edu/DRM2002/darknet5.doc ), topsites would be described as supernodes. They are glue between multiple mini-darknets -- small, autonomous networks between distribution triangles, webs, communities, what have you, which are like jw describes -- they come, they go, they work, they don't, whatever. Supernodes drive efficiency on the darknet. Here is the problem, though -- all topsites would be supernodes, but not all supernodes are topsites, and the darknet would "organically" adjust to you closing those 30 sites. Even if you made a policy of squashing the piracy groups that sprang up (we'll ignore the minor implementation issues here, such as the difficulty of stopping pirates in China who are, for all pracctical purposes, agents of the government who are charged with doing the stopping). Consider a hypothetical group of 15 college students with time on their hands and a desire to do mischief or just to be the first on the block with free tickets to Spiderman 3. Lets spot them one computer which is capable of being a minor server, one IRC channel to call their very own, privileged access to original content (maybe one of them temps at a courier agency -- honestly, IP providers need to get smarter about their use of those, because they are a major source of leakage, especially in CA), and of course access to the big, public darknets. Congratulations, we've got a supernode. If they've got bandwidth to spare at their college, from that one source the content is going to leak slowly onto the public darknets and be adopted, with astonishing rapidity, on other supernodes -- and there is no need to "push" content between some of those nodes. I knew a student at a major Widwestern university who was a one man supernode with exactly one computer under his direct control and residential broadband -- he probably had, in his direct possession, $2 million of copyrighted material (binders upon binders upon binders of burned CDs, filling most of a room and all of a closet). The main people he distributed it to were his personal friends and other students at the university who knew him as "The guy to go to to get stuff" -- its like being a drug dealer, except you can create more meth instantly, for free, without the need for suspicious purchases at the local chemical supplier. The main distribution method was the sneakernet -- burn a new CD, literally walk it to the person who received it and, frequently, put it right back onto their own personal darknets. You can smash the "topsites" all you want, but it won't change the essential equation -- any data which is intrinsically valuable and which reaches any node in the darknet will become available to all nodes, and the only question after it is available on one node is "how much time does it take".

Patrick McKenzie
Softopia Japan R&D

I should point out that I am not encouraging piracy, just describing how it works. bittorrent is a well-design application and the swarming concept is a work of genius, and is being used as we speak in non-infringing applications (Blizzard Entertainment uses a proprietary implementation to distribute patches for their new game, World of Warcraft, and while that implementation has some issues it will save users a lot of downtime and Blizzard a lot of greenbacks after the kinks are worked out).

Patrick McKenzie

In the early days of Desktop Publishing and Graphics I was part of an informal network for sharing fonts. If anyone in the group acquired a new font they shared it with the rest by passing along copies on floppy discs. In those days, there was only one type format and only one company selling type.

I tend to agree with Malvolio's analysis. If it were a shallow pyramid (as the Wired article described it), then it would be suscepible to failure -- via network outages, law enforcement intervention, etc. -- and we would be seeing that instability. I suspect that the people interviewed by Wired may not be aware of the true complexity of the actual network (I've found this to be true of many Network Administrators throughout the IT industry ;-).

Moreover, it would take a lot effort to be one of a small number of gatekeepers for all content on the the BitTorrent network. Maybe people have that sort of focus in their lives, but it took me several weeks of effort to rip my entire 500+ CD collection on to a hard drive (so I could feed the insatiable appetites of my iPod). I never want to do that again! But ripping large video files -- then fixing the blemishes created by the codec -- and doing this for hundreds of files each week boggles my imagination.

Finally, as an end-user I haven't found that Bittorrent has shown it self to be apprcciably faster per Mbyte than P2P via LimeWire. The economies of scale seem to reside on the server side of the equation.

However, the anarchist in says more power to them, even it ain't quite the way Wired thinks it is.

Despite the skepticism of some of the commenters, I'm quite sure this is an accurate picture of the scene. I've known people at lower levels of the pyramid, and a quick comparison of md5s of a set of mp3s will suggest a singular source. Anyone who downloads from anywhere other than kazaa (or other p2p) will quickly become familiar with rns, kmt, and other release groups. Not to say you can eliminate piracy by eliminating the groups, but there is a defined heiarchy to high quality piracy. And to the worst fears of the studios, it's usually not done for profit.

There are a number of problems I see with enforcement. First, the Internet is worldwide. It takes quite a bit of work to bust someone in much of the rest of the world. And it takes quite a bit of work to penetrate the networks. Nothing is easy here.

Yet, it is much easier to enter the game than to shut it down. All you really need is a computer and a broadband connection - though at some levels, higher speed is advantageous. So, I see it almost impossible to thwart - easy to play, hard to bust.

Like the other posters have said, this article sounds like someone who read a FAQ about Bittorrent, hung around Somethingawful until they closed their file forums, and was way too excited about "Burning Chrome". It's quite romantic to imagine a shadow-Internet of secret websites, super-secure passwords, double- and triple-agents, bearded sandal-clad thirtysomethings taking on Brando's role from "The Godfather", but the fact of the matter is that you don't need all of that. There is no "darknet", no "backdoor site", no illicit network of specialist pirates hacking into corporations to get the goods. When one guy copies a file from another, that's a P2P network. The "topsites" they mention are really just message boards where people advertise the files they've got available. Indeed, what would be the point of super-secret message boards that only a few people could get into? It's file sharing, not file-keeping-and-not-giving-away.

As far as Bittorrent is concerned, the point is that the bandwidth is spread out among every user rather than being centralized. (You can tart that up with "information theory" and "social network" if you want to sell it to someone for $200 an hour.) If I have a file which has two parts (A and B), and two people want that file, then (by using Bittorent) I can send the A part to one and the B part to the other, and they each send the A and B between themselves. The end result is that two people now have copies of the file, but my bandwidth only needed to send out one copy. If they'd downloaded the file directly, I'd have sent out two copies.

There's very little in the article that wasn't true in 1998 as well (except of course the P2P aspect at the bottom of the pyramid, which isn't the focus). Topsites, release groups, couriers, etc., all predate my arrival on the net in the early 1990s.

I was amused at the notion that Half-Life 2 turned into coal in Valve's stocking. The article notes that the original Half Life sold 10 million copies; it ignores the fact that the original game was also pirated before its official release date and easily available on Usenet. The fact is that most of the people who download stuff do it because they can't really afford all the games/movies they want. Cutting them off from their supply will not result in vastly increased sales.

For those like jw who think the article is more "myth" than "reality" -- sorry, no. While the Wired article fumbles with the details, the elite network of people they describe is true.

How do I know?

I was part of it. I was busted in Operation Fastlink. There's a good chance I will spend some time in prison for this. Go to cybercrime.gov, a webpage run the Department of Justice. While their press releases are blatant propaganda and very inaccurate, they should educate you nevertheless.

>While the Wired article fumbles with the details, the
>elite network of people they describe is true.

The regulars at Suprnova hardly constitute an elite network. (Or should I say "the former Suprnova". Ho Ho Ho.)

While we're at it, the "ultra-secret webpages" mentioned in the Wired article consist of message boards that you need to be a member to view (non-members just get a screen saying 'you need to register'.) Hardly the kind of cloak-and-dagger system described.

>I was part of it. I was busted in Operation Fastlink.

Uh-huh, and those kind people at the jail still give you internet access. That reminds me of the time I got sent up the river for smuggling heroin, and they let me join the prison's Yacht Club.

Well, I was on Arpanet and Cybernet in the 70s and keep my hand in a little even today.

The key to tracking a lot of this is to track the same sources -- which a number of people keep talking about, and the quality, which is very good.

The problem to stopping it all is that it is ego driven, which makes it harder to track. If people were only making money from it, they'd all be in jail now.

But, the organization system also explains why network efforts to contaminate distribution have failed so hard.

If the public model were true, introducing bad content on a massive scale would drive the good content out. If you downloaded Madonna screaming 95% of the time, you would give up on downloading Madonna (and most downloaders and redistributers never listen to what they get). If MP3s took 10x the space and had .25 the quality, people would be out buying CDs. The same for the movies. In trumps.

But the days of handcams from the audience and lousy codec work are long gone, which should tell you something.

I'm inclined to believe Anon because he seems to know what he's talking about.

Suprnova was not elite whatsoever, it was just a popular place to find anything(and a much better alternative to virus-Kazaa). I personally used the site for patches and TV shows as I could find just about anything. Any old Joe with a Bitorrent client could hop on and pirate to their own delight without any fear of retribution.

Overall, very interesting article.

But, I also think they downplay the strength of the lower-tier community of pirating. I know there are plenty of non-centralized pirating going on, involving movies, music, games, and applications.

I myself know how to rip a movie and encode it from a DVD, it's not as hard as they make it sound.

I am doubtful about any + or - to mainstream music in regards to piracy, and bandwidth is still not plentiful enough to believe enough people are dling DVD-rips in enough quanity to harm sales. Ripped movies just don't look as good as the DVD versions, unless it's the full size of 4GB.

BAGO,

I cannot speak for the other "detractors", but I have been around the block a lot longer than you presume. My first experience online was in 1994, back when pirating groups like Razor and Phrozen Crew were the sourge of the software industry. And before that I frequented the local BBS scene and was lucky enough to have VAX system accounts. I do not work in IT field today, but I feel confident enough to dissect an article written in a magazine that was once technically sound, but today is simply another poprag.

You mention IRC and FTP. I can tell you, for instance, that IRC is no longer the mainline it once was for piracy courier groups. Efnet, for example, is a fraction of the network it was in the mid 90s, when several larger servers like Netcom and AOL linked up. Today, those links are no longer present. Less IRC traffic = less transfer of pirated software.

I can also tell you that FTP, despite its name, today is more of a secondary (or even third rate) protocol for file transfer. Bittorrent and P2P have far eclipsed it as a useful tool for sharing information and pirated software.

I did not say that the Wired article is entirely without merit. Certainly, "topsites" are a part of the wide-ranging problem. But to claim, as the article does, that if you kill the "head" then the body of piracy will die, is - as I and others have tried to point out - an oversimplification of the problem.

Anonymous,

I certainly do not presume to know of the "inner workings" of the "shadow network" you keeping talking about. I have always had just personal interest in the legal issues surrounding the so called "scene", so I just did a lot of reading online.

It seems to me that if these groups start profiting it will be their downfall. Perhaps one of the reasons that these "underground" pirating groups have been successful and lasted so long is the 'honour code' you talk about. But, I think once money comes into it, then you have a paper trail.

If what Anonymous is saying is true, then maybe law enforcement is not too far behind these "elite groups" after all.

Just to expand on my earlier point -- this can all be explained through simple economics and technological capabilities.

Video games and movies are expensive and released infrequently -- on the order of hundreds per year. A typical consumer would have trouble making use of more than, say, 10 video games a year, or 20 movies. (Some people consume more, of course, but most probably consume less.)

Video games and movies are technologically costly to put on the 'net in consumable form. A DVD movie typically must be compressed, which can take from 6 to 12 hours; and both video games and movies are usually very large, at least one or two CD's worth of data, which takes a very long time for a user with average upstream-limted broadband to upload (maybe another 6-12 hours).

CDs are inexpensive and there are thousands of them released every year. In addition, old CDs remain a lot more interesting than old films and especially old video games. The average consumer, I suspect, will own and purchase 10 CDs for every DVD or video game.

Albums are inexpensive to share on the internet -- on a modern computer they take less than 10 minutes to compress and not more than an hour to upload.

Movies and video games lend themselves to top-down distribution because the one-time cost of creating them is very high. Albums lend themselves to peer distribution because the one-time cost is low.

What this all means, I think, is that it would be possible to 'cut off the head' of the movie and software distribution channels and severely impact their availability (or at least their quality), in a way that you simply can't do with music.

But I also think that these posters who are claiming to have been part of the shadowy underworld without which there would be no movie releases are self-aggrandizing a bit. Movies would still be uploaded if the head were cut off, by channels that are largely unutilized now. There's no point to me ripping a copy of a movie and uploading it because it's so much trouble and there's already a perfectly serviceable copy out there. If there weren't one, I might do it. Just because nobody does it doesn't mean they -can't-.

neil, what you're not understanding about music is the industry insider connections the MP3 scene possesses. Aside from obscure foreign music, unsigned bands, and other "less desirable" releases, every single album you find in a store is available on the internet at least several days in advance. It is common for them to "leak" to the internet weaks, and even months before the average consumer has legal access to the product. While the average person will find these albums on a peer to peer network, such as KaZaA, the files originated from the scene. That is what this Wired article is talking about, and that is the purpose of "cutting off the head."

It is very possible. If the FBI was completely successful and they managed to find a small handful, a couple dozens of people, they would effectively "cut off the head." There would be no more Eminem albums leaked on the internet months before the public release. Of course, once the album is out in stores and available to the entire world, there is no way to stop online distribution, and the FBI is well aware of this. They aren't trying to stop it. What they can stop, however, are the "leaks," and they are trying very vigorously with moderate success.

Perhaps I don't understand your point. For the music industry, it is a very important matter that their wares not be available until they want them to be available. By extension, that is what is now also very important to the Department of Justice, and that is the end they are trying to reach with this legal battle.