3 ms·

Nobody should believe that any encryption used today will be secure against attack from anyone after nnn days due to technological advance. Even if no weakness

Nobody should believe that any encryption used today will be secure against attack from anyone after nnn days due to technological advance. Even if no weakness is ever found with which to attack the encryption, Moore's Law will eventually take care of it. edit: yeah, I was talking out my ass, get over it. I still think it's imprudent to assume that encrypted files may never be compromised in the future.

pyre 13y ago
They are collection a ton of data. As time goes on, if a large portion of it is encrypted, then it may not be feasible to break it all, even with Moore's Law. They may just have to be selective with it. You can imagine that even in if the US becomes a new-age Nazi Germany / Stalin's Russia, it may still be prohibitively expensive to go spelunking through years old communications looking for people with possible subversive thoughts. Edit: It would probably just be easier at that point to ban encryption going forward.
mpyne 13y ago
This is pretty crap, actually. http://www.schneier.com/blog/archives/2009/09/the_doghouse_cr.html http://www.schneier.com/blog/archives/2009/09/the_doghouse_c... You couldn't even increment every position of a 256-bit key using the energy our Sun could provide, let alone brute force 256-bit AES.
- fnordfnordfnord 13y ago
  Fine, scratch the last part.
- kamjam 13y ago
  I often TrueCrypt (http://www.truecrypt.org/ http://www.truecrypt.org/) as containers to transport private documents on USB or in emails, and able to use multiple encryption algorithms for the containers (e.g. AES-Twofish-Serpent). Something like this will presumably slow them [NSA] down umpteen amounts of times?
  - mpyne 13y ago
    Not sure, I'm not a crypto wizard. My guess would be that "it depends". What you describe was more-or-less the basis for 3DES IIRC, but (knowing crypto) you could probably just as easily accidentally break the cryptosystem to the effectiveness of just the best scheme.
  - josephg 13y ago
    Snowden: > Encryption works. Properly implemented strong crypto systems are one of the few things that you can rely on. Unfortunately, endpoint security is so terrifically weak that NSA can frequently find ways around it. You probably don't need to use 3 cyphers. Just make sure you're storing your private key somewhere safe.
endeavour 13y ago
You've simultaneously managed to misinterpret both the way encryption works and Moore's Law.
- fnordfnordfnord 13y ago
  I should have known better than to paraphrase or quote "Moore's Law" out of context around here. I usually feel its safe to do so because I don't really consider Moore's Law to be Law, as much as "Moore's Quite Good Prediction" s/Moore's Law/technological advance and\/or advances in cryptanalysis/

pyre 13y ago

They are collection a ton of data. As time goes on, if a large portion of it is encrypted, then it may not be feasible to break it all, even with Moore's Law. They may just have to be selective with it. You can imagine that even in if the US becomes a new-age Nazi Germany / Stalin's Russia, it may still be prohibitively expensive to go spelunking through years old communications looking for people with possible subversive thoughts. Edit: It would probably just be easier at that point to ban encryption going forward.

mpyne 13y ago

This is pretty crap, actually. http://www.schneier.com/blog/archives/2009/09/the_doghouse_cr.html http://www.schneier.com/blog/archives/2009/09/the_doghouse_c... You couldn't even increment every position of a 256-bit key using the energy our Sun could provide, let alone brute force 256-bit AES.

fnordfnordfnord 13y ago
Fine, scratch the last part.
kamjam 13y ago
I often TrueCrypt (http://www.truecrypt.org/ http://www.truecrypt.org/) as containers to transport private documents on USB or in emails, and able to use multiple encryption algorithms for the containers (e.g. AES-Twofish-Serpent). Something like this will presumably slow them [NSA] down umpteen amounts of times?
- mpyne 13y ago
  Not sure, I'm not a crypto wizard. My guess would be that "it depends". What you describe was more-or-less the basis for 3DES IIRC, but (knowing crypto) you could probably just as easily accidentally break the cryptosystem to the effectiveness of just the best scheme.
- josephg 13y ago
  Snowden: > Encryption works. Properly implemented strong crypto systems are one of the few things that you can rely on. Unfortunately, endpoint security is so terrifically weak that NSA can frequently find ways around it. You probably don't need to use 3 cyphers. Just make sure you're storing your private key somewhere safe.

endeavour 13y ago

You've simultaneously managed to misinterpret both the way encryption works and Moore's Law.

fnordfnordfnord 13y ago
I should have known better than to paraphrase or quote "Moore's Law" out of context around here. I usually feel its safe to do so because I don't really consider Moore's Law to be Law, as much as "Moore's Quite Good Prediction" s/Moore's Law/technological advance and\/or advances in cryptanalysis/