4 ms·

> You used a one-time pad. A OTP assumes you have an off-channel secure messaging system. It's like the cryptographical equivalent of begging the question: i

> You used a one-time pad. A OTP assumes you have an off-channel secure messaging system. It's like the cryptographical equivalent of begging the question: it's easy to send a secret message! Just already have sent a secret message first! > Computers were not first used for battle, they were first constructed for pure math The difference engine wasn't actually constructed until the 90s. You might have an argument for the Z1.

bitwize 16y ago
All you need is a source of truly random numbers. Make a pad from those numbers, make an exact duplicate of it, give one to your field agent before he goes off on his super-secret mission and keep one at home base. Now you two can communicate over a guaranteed secure channel (provided you don't reuse the pad).
- drewcrawford 16y ago
  > give one to your field agent before he goes off on his super-secret mission This is cryptography's version of "begging the question". A OTP means that "Given that you already have a secure communications channel (to exchange the one-time-pad), you can send secure messages over other channels." Of course if you already have a secure communication channel, why wouldn't you just use that? Perhaps you think that your "real world example" offers an explanation--remote communications can be more advantageous than local ones. However, as soon as we enter "the real world", we see that the OTP you wrote down can easily be captured along with your physical self, and the message compromised or even modified. This happens often enough in practice: see the "Exploits" section of http://en.wikipedia.org/wiki/One-time_pad http://en.wikipedia.org/wiki/One-time_pad Modern cryptography (specifically I'm thinking of public-key-based exchange of a private key) doesn't depend on having a secure channel to communicate with in the first place. In addition, private keys can't be captured on your person like a OTP can. So it is both theoretically and actually better than a OTP by multiple orders of magnitude. For these reasons, I think comparing a OTP to modern public or private key cryptography is pretty silly.
  - bitwize 16y ago
    Private keys are just as vulnerable to rubber-hose cryptology as OTPs.
    - drewcrawford 16y ago
      Not in practice. If you're talking about private keys in public-key-cryptography, those are usually encrypted with another private key, which exists only in your head. If you're talking about private keys in private-key-cryptography, they are either encrypted with a private key that only exists in your head, or they only exist in your head to begin with, or they are a "one-time" random private key, i.e. for a key exchange algorithm, that is not re-used. An OTP is not usually encrypted. We can AES it now, of course. But the point of the article is that encryption has gotten two orders of magnitude better. That point is still valid.

bitwize 16y ago

All you need is a source of truly random numbers. Make a pad from those numbers, make an exact duplicate of it, give one to your field agent before he goes off on his super-secret mission and keep one at home base. Now you two can communicate over a guaranteed secure channel (provided you don't reuse the pad).

drewcrawford 16y ago
> give one to your field agent before he goes off on his super-secret mission This is cryptography's version of "begging the question". A OTP means that "Given that you already have a secure communications channel (to exchange the one-time-pad), you can send secure messages over other channels." Of course if you already have a secure communication channel, why wouldn't you just use that? Perhaps you think that your "real world example" offers an explanation--remote communications can be more advantageous than local ones. However, as soon as we enter "the real world", we see that the OTP you wrote down can easily be captured along with your physical self, and the message compromised or even modified. This happens often enough in practice: see the "Exploits" section of http://en.wikipedia.org/wiki/One-time_pad http://en.wikipedia.org/wiki/One-time_pad Modern cryptography (specifically I'm thinking of public-key-based exchange of a private key) doesn't depend on having a secure channel to communicate with in the first place. In addition, private keys can't be captured on your person like a OTP can. So it is both theoretically and actually better than a OTP by multiple orders of magnitude. For these reasons, I think comparing a OTP to modern public or private key cryptography is pretty silly.
- bitwize 16y ago
  Private keys are just as vulnerable to rubber-hose cryptology as OTPs.
  - drewcrawford 16y ago
    Not in practice. If you're talking about private keys in public-key-cryptography, those are usually encrypted with another private key, which exists only in your head. If you're talking about private keys in private-key-cryptography, they are either encrypted with a private key that only exists in your head, or they only exist in your head to begin with, or they are a "one-time" random private key, i.e. for a key exchange algorithm, that is not re-used. An OTP is not usually encrypted. We can AES it now, of course. But the point of the article is that encryption has gotten two orders of magnitude better. That point is still valid.