Damocles wrote:But then you still have no information gained.
You just have the sum of all combinations that a message of this length can contain.
So you could interpret anything into that message.
I think they used this method in the "red phone" between the US and Soviet Union.
The issue with one time pads, is primarily that distribution of these codes is manual, you have to select the correct pad on both sides, and there is no way of creating a secure connection without shipping these codes.
This is why Diffie Hellman exists. Asymmetric key exchange is the simplest (eg not at all simple) method of exchanging a secret between two parties with no-one outside of those parties being able to decrypt both (or either ideally).
This however, is only for the initial exchange of the secret key. Once exchanged, both parties move to a VERY simple XOR cipher. Using a PRNG whose state was set by the initial secret exchanged, this leads to an infinite length, one time pad with zero possibility of future decryption beyond brute force.
The only thing is if your private key is captured, then the target can read your end of the initial secret exchange.
This is why *BOTH* parties send a secret, and *BOTH* secrets are required to start the PRNG at the correct place. With this method, an attacker needs to acquire the private key from both parties to be able to read the conversation live.
Or to get one, and guess the other secret to be able to read it later. (because these secrets are ideally very, very long, it should take a very, very long time to guess it)