• DRM has won the fight and we're all doomed
    42 replies, posted
Fully homomorphic encryption is a holy grail of encryption. Mathematically speaking, it's an encryption scheme which preserves the ring structure of the plaintext. This means that the encryption "supports" addition and multiplication of the plaintext in a way that's consistent with some equivalent operations on the encrypted text. Why is it such an important property? Because with just addition and multiplication one can implement AND and NOT gates. Those gates in turn can be used to implement any Turing machine. Since Turing machines can be used for computing any algorithm, that's sufficient to prove that the scheme can be used to perform useful computation on encrypted data [i]without[/i] decrypting it. In essence, such encryption can be used to run arbitrary software on an unsecured device in a way that makes it impossible for anyone to snoop the internal state of the program at any point. The program doesn't need to decrypt its instructions to execute them. Previous attempts at implementing fully homomorphic encryption only managed a few layers of encryption, and it required periodical refreshing of the encryption (though this could be done with a short circuit of encrypted data). Though in theory it worked, it was impractically slow. However, [url=http://www.ijtpc.org/volume1/JTPC1386.pdf]a recent paper[/url] describes a way to do fully homomorphic encryption which uses integer arithmetic and is exceedingly fast. It doesn't suffer from noise, so it can nest encryption as deep as you want. The practical applications are numerous: The encryption can be used to process video and audio without decrypting it before sending it to the display device, leaving analog hole as the sole way to record such media. The processor doesn't know what they're doing with the encrypted program any more than can be deduced from the instructions it's being given. Since the same IO operation could either launch an ICBM or make your caps lock light blink, the instructions aren't very useful in determining what's actually going on inside the program. A program that displays video could simply break itself after it's done playing the media, rendering any copy of the program only executable once, and making it impossible to recover the video data from the program's data. It's also the perfect guard against debugging and reverse-engineering, making it a boon for viruses like SecuROM. The encrypted program can't be analyzed, its signature can't be compared since it can be random, its internal behaviour is completely unknown. Again the only thing that can be deduced from it is how it interacts with the rest of the system when executed. For cloud computing, it means that data can be securely distributed to anyone, processed by anyone, and then have the still-encrypted results sent back to the owner of the data. Nobody knows what just happened, but the computations got done. [b]tl;dr:[/b] It's now possible to make a program that makes the user go "what are you doing computer" and the computer says "idk lol i just work here" and [i]only the program knows what's happening[/i]
Well, fuck.
Wow, all that is over my head In some ways; very cool. In others; damn scary
This is real bad
[quote]For cloud computing, it means that data can be securely distributed to anyone, processed by anyone, and then have the still-encrypted results sent back to the owner of the data. Nobody knows what just happened, but the computations got done.[/quote] That can't end well.
sounds way too complicated for your average programmer. I imagine it would add a lot of bulk to most programs and generally not be worth it [editline]4th December 2012[/editline] and what's to stop antivirus companies from having their programs stop any code from executing without decrypting first?
Have you even read the paper? It's a very lightweight and fast algorithm.
[QUOTE=VistaPOWA;38699369]Have you even read the paper? It's a very lightweight and fast algorithm.[/QUOTE] hate to disappoint but I'm not a programmer. most of the paper looks like moonspeak to me
If my computer seriously says "I'm sorry, Dave" I'm gonna shit a brick.
And now everybody ever is going to move to pirating games because of crap DRM gg
[QUOTE=Nemisis116;38699481]And now everybody ever is going to move to pirating games because of crap DRM gg[/QUOTE] Couldn't this just completely eliminate piracy?
i love how everyone thinks this is now being used by everyone ever and the internet is about to explode it's just a research paper calm down
[QUOTE=Nemisis116;38699481]And now everybody ever is going to move to pirating games because of crap DRM gg[/QUOTE] Don't you understand? This would make pirating games impossible. Using this + standard DRM (such as requiring a CD key) would be impossible to circumvent because the algorithm for determining CD keys would never be able to be decrypted, because it wouldn't be required to be decrypted to generate a key. You also wouldn't be able to create cracks because the processes controlling how the game runs would never be able to be determined because they would still be encrypted while they are running. At least this is how I understood the article, and I'm fairly certain that I'm correct. [editline]4th December 2012[/editline] [QUOTE=Murkat;38699501]i love how everyone thinks this is now being used by everyone ever and the internet is about to explode it's just a research paper calm down[/QUOTE] I think people realize it's just a research paper
[QUOTE=zzzz;38699511]Don't you understand? This would make pirating games impossible. Using this + standard DRM (such as requiring a CD key) would be impossible to circumvent because the algorithm for determining CD keys would never be able to be decrypted, because it wouldn't be required to be decrypted to generate a key. You also wouldn't be able to create cracks because the processes controlling how the game runs would never be able to be determined because they would still be encrypted while they are running. At least this is how I understood the article, and I'm fairly certain that I'm correct. [editline]4th December 2012[/editline] I think people realize it's just a research paper[/QUOTE] Those pirates are smart motherfuckers, they always somehow manage to find a way Don't doubt them
[QUOTE=Nemisis116;38700162]Those pirates are smart motherfuckers, they always somehow manage to find a way Don't doubt them[/QUOTE] this is the same sort of thing people say about uncrackable encryption and welp pirates might be smart, but they aren't magic
[IMG]http://t1.gstatic.com/images?q=tbn:ANd9GcQ-tKOmsQ6AERZYy4JtHLlmGmlOdEAZ7QPVt-GrZ-X8F0TQj_qB_g[/IMG] IT BEGINS
The encryption algorithm has a few pending patents so we'll have to wait and see if they'll even sell it to anyone. On the other hand, none of the methods presented in the paper seems "novel", one could easily find prior art for most of the things they're doing, so it might be that the patents wouldn't even hold.
Part of me thinks that even with this pirates will still manage to break it.
[QUOTE=Reshy;38703151]Part of me thinks that even with this pirates will still manage to break it.[/QUOTE] They've managed to break just about everything before. It'll be difficult, but probably possible.
I'm pretty sure the weak point here will be the key. Which like the key for DVD encoding will be derivable. Nothing of value was lost, don't worry people.
I'm not that concerned about DRM unless it affects the enjoyment of the product for a regular consumer. I don't pirate that much games. If there is a game I don't buy I either ignore it, criticize it or buy it anyway. Countering adblock software, however, is a crime against humanity.
Interesting, but as said above, it's only a matter of time until somebody finds a way to bypass this. If it's made by a human, it can be taken apart.
Remember Palladium? Wake me up when this actually exists.
Uh, how would you apply fully homomorphic encryption in a DRM scheme (ignoring obvious problems like the NAND circuit being slow as fuck)?
[QUOTE=DrasarSalman;38704055]I'm not that concerned about DRM unless it affects the enjoyment of the product for a regular consumer. I don't pirate that much games. If there is a game I don't buy I either ignore it, criticize it or buy it anyway. Countering adblock software, however, is a crime against humanity.[/QUOTE] Yeah, this doesn't seem like a big thing to me. I'm more worried about how a virus can abuse this
You guys remember this guy? [IMG]http://upload.wikimedia.org/wikipedia/en/6/66/Star_force_logo.png[/IMG] It did something similar and it was quite successful actually. But it was eventually cracked (after a year, mind you).
Well I never would've guessed that something like this was even possible. Hopefully if an application of something like this becomes commonplace, consumer protection is taken into consideration. It would be so easy to abuse, and if it is as foolproof as it is presented, then the only way to figure out with certainty what a program can actually do is to acquire a copy of the original source code; which for commercial products isn't lightly released for obvious reasons. [editline]4th December 2012[/editline] [QUOTE=aydin690;38708049]You guys remember this guy? [IMG]http://upload.wikimedia.org/wikipedia/en/6/66/Star_force_logo.png[/IMG] It did something similar and it was quite successful actually. But it was eventually cracked (after a year, mind you).[/QUOTE] I'm not sure how well I understand it all, but it really doesn't seem like they're that similar. StarForce created a virtual file structure which could only be read with an encryption key. This seems to be a method of creating an encrypted executable that doesn't actually have to be fully decrypted to be run. Please someone tell me if I'm getting it wrong. The concept as presented in the OP makes decent enough sense to me as far as I can tell, but I have no hope of comprehending the paper.
[QUOTE=Ericson666;38707712]Yeah, this doesn't seem like a big thing to me. I'm more worried about how a virus can abuse this[/QUOTE] You guys don't beleive how much this dude is right. If virus creators get this, then key loggers would be incredibly hard to catch, and much more slipping under.
[QUOTE=froztshock;38708201] I'm not sure how well I understand it all, but it really doesn't seem like they're that similar. StarForce created a virtual file structure which could only be read with an encryption key. This seems to be a method of creating an encrypted executable that doesn't actually have to be fully decrypted to be run. Please someone tell me if I'm getting it wrong. The concept as presented in the OP makes decent enough sense to me as far as I can tell, but I have no hope of comprehending the paper.[/QUOTE] The point is, no matter how complicated it may be, somebody will eventually find a way around it.
[QUOTE=Nemisis116;38700162]Those pirates are smart motherfuckers, they always somehow manage to find a way Don't doubt them[/QUOTE] They'll just learn how to take the product key out of the product :v:
Sorry, you need to Log In to post a reply to this thread.