A proposition: let's just start splitting our P2P downloads, initially and for simplicitly, in two counterparts of equal size that are XORed together to form the actual work whose copyright we want to infringe.
Downloading random data can't possibly be illegal—oh wait, despite of this naive scheme (that is almost out of Applied Cryptography and one that almost smells of a one-time pad), wouldn't my download still be non-random because it somehow must contain a kind of a half of a copyrighted work or something? Wouldn't it still infringing? Mere lawsuit-fodder?
It happens that if we have any blob of data of, say 700MB in size, then no matter what you do, someone can always construct another 700MB that—when XORed together—yields a copy of a recent movie.
If I legally download a Ubuntu live CD image, I can't possibly be infringing any copyright, right? Even if someone has constructed another image of the same size, that when XORed with the Ubuntu live CD yields a copy of Die Hard? Duh, easy: it's obviously the other 700M blob of data that's infringing copyright! Off to prosecute me if I do that, then.
Now, if I do happen to be downloading the other blob, what exactly would I be infringing? Again, the blob I'm downloading contains random bits and can contain—depending on the other half—any movie you can think of, under copyright or not. Or no movie. If someone gives me a suitable set of ten other halves, then I can get exactly ten movies out of the single 700M blob. But you can't possibly fit ten high-quality movies into a single 700M blob, so obviously the other blobs, again, must contain the actual payload and be illegal copies, and not this particular one...
If neither of the blobs is a public, recognized piece of data, how can you tell which is which? They're just both random data. If I individually XOR these blobs with other blobs and happen to get Ubuntu live CDs, Fedora installation CDs, WinXP corporate installer CDs, Die Hard and other movies, and memtest diagnostics CDs popping out of them, it's obvious that someone has been XORing these blobs with copyrighted works.
But you can't tell who exactly infringed the copyright; you can't tell who exactly downloaded the copyrighted works; and you can't tell who exactly uploaded them, either.
The infringing will eventually happen, should one choose to do so, in the privacy of one person's home. But how could anyone know? Especially MPAA/RIAA/<insert_goons>? Besides, fair use says that personal copies can be made, doesn't it. Or something like that.
Lawyers would of course, at this point, laugh at this feeble technicality and start aiming their prosecution guns at me, drooling off all the way. But as long as they could only observe IP numbers logged off a BitTorrent tracker or magnet swarm, what would they produce as evidence? They could try to convince the court that because I was downloading a 700MB blob named b8501756a0db1fc96a79920c30edd29e40a050b4 I was downloading and sharing a particular movie? Possibly, but how could they prove that?
They could carefully examine the logs and try to find out whether I've also downloaded and shared the particular XOR counterpart of this first part. But what if I got it off a website? They can't track direct downloads. Or what if there were three counterparts? Two of them available on torrents and a third counterpart of public data that I could legally and separately download somewhere or have it delivered over with an USB stick. Or I got some of them from FreeNet?
I guess they would have to show that I first had all the required parts and then actually assembled the files together in order to prove that I did, in fact, infringe the copyright of one work. If they don't, then why would downloading anything from internet not infringe the copyrights of all copyrighted works there are.
Please, someone tell me why this wouldn't work. Convenience? The maneuvers here do sound complicated but they could be mostly automated away.
I'm not stupid enough to think that this would somehow hold water, otherwise we'd be doing this already; I'm just stupid enough to not figure out why exactly would it fail, and I mean fail totally. Not just "yaa, they'd probably try to sue you anyway".
Because paying 2x or 3x bandwidth for the freedom to download and seed as much as you can would be a bargain. (And, yes, I would legally buy the best of all the warez I would download.)
It doesn't matter if you really downloaded a movie in this case or not. In the US the lawsuit handling costs are likely >= the requested fine. Even if you're right, you've lost. They would have to prove something only if you wanted to challenge their claim.
A proposition: let's just start splitting our P2P downloads, initially and for simplicitly, in two counterparts of equal size that are XORed together to form the actual work whose copyright we want to infringe.
Downloading random data can't possibly be illegal—oh wait, despite of this naive scheme (that is almost out of Applied Cryptography and one that almost smells of a one-time pad), wouldn't my download still be non-random because it somehow must contain a kind of a half of a copyrighted work or something? Wouldn't it still infringing? Mere lawsuit-fodder?
It happens that if we have any blob of data of, say 700MB in size, then no matter what you do, someone can always construct another 700MB that—when XORed together—yields a copy of a recent movie.
If I legally download a Ubuntu live CD image, I can't possibly be infringing any copyright, right? Even if someone has constructed another image of the same size, that when XORed with the Ubuntu live CD yields a copy of Die Hard? Duh, easy: it's obviously the other 700M blob of data that's infringing copyright! Off to prosecute me if I do that, then.
Now, if I do happen to be downloading the other blob, what exactly would I be infringing? Again, the blob I'm downloading contains random bits and can contain—depending on the other half—any movie you can think of, under copyright or not. Or no movie. If someone gives me a suitable set of ten other halves, then I can get exactly ten movies out of the single 700M blob. But you can't possibly fit ten high-quality movies into a single 700M blob, so obviously the other blobs, again, must contain the actual payload and be illegal copies, and not this particular one...
If neither of the blobs is a public, recognized piece of data, how can you tell which is which? They're just both random data. If I individually XOR these blobs with other blobs and happen to get Ubuntu live CDs, Fedora installation CDs, WinXP corporate installer CDs, Die Hard and other movies, and memtest diagnostics CDs popping out of them, it's obvious that someone has been XORing these blobs with copyrighted works.
But you can't tell who exactly infringed the copyright; you can't tell who exactly downloaded the copyrighted works; and you can't tell who exactly uploaded them, either.
The infringing will eventually happen, should one choose to do so, in the privacy of one person's home. But how could anyone know? Especially MPAA/RIAA/<insert_goons>? Besides, fair use says that personal copies can be made, doesn't it. Or something like that.
Lawyers would of course, at this point, laugh at this feeble technicality and start aiming their prosecution guns at me, drooling off all the way. But as long as they could only observe IP numbers logged off a BitTorrent tracker or magnet swarm, what would they produce as evidence? They could try to convince the court that because I was downloading a 700MB blob named b8501756a0db1fc96a79920c30edd29e40a050b4 I was downloading and sharing a particular movie? Possibly, but how could they prove that?
They could carefully examine the logs and try to find out whether I've also downloaded and shared the particular XOR counterpart of this first part. But what if I got it off a website? They can't track direct downloads. Or what if there were three counterparts? Two of them available on torrents and a third counterpart of public data that I could legally and separately download somewhere or have it delivered over with an USB stick. Or I got some of them from FreeNet?
I guess they would have to show that I first had all the required parts and then actually assembled the files together in order to prove that I did, in fact, infringe the copyright of one work. If they don't, then why would downloading anything from internet not infringe the copyrights of all copyrighted works there are.
Please, someone tell me why this wouldn't work. Convenience? The maneuvers here do sound complicated but they could be mostly automated away.
I'm not stupid enough to think that this would somehow hold water, otherwise we'd be doing this already; I'm just stupid enough to not figure out why exactly would it fail, and I mean fail totally. Not just "yaa, they'd probably try to sue you anyway".
Because paying 2x or 3x bandwidth for the freedom to download and seed as much as you can would be a bargain. (And, yes, I would legally buy the best of all the warez I would download.)