| OTP at
formatting link
(I deleted a soft hyphen that had snuck into the URL.)
| enables the user to generate one-time pads
No. They make the claim, and say
Spies are furnished "one-time pads" containing pages of keys used to encrypt individual characters of secret messages, then discarded. As long as the physical security of the two copies of a one-time pad is assured and the keys on the pad are sufficiently random, security is absolute.
This is true only if the one-time pads are truly random. The program on this page creates pseudo-random "one-time pads", and for this case, the proof of security breaks down. From the description offered on the web page I would venture a guess that their "one-time pads" are not cryptographically very secure.
| or password lists in a variety of formats. | | The question is about MD5 signature of the generated keys. | [...] | Who and how does verify keys with using MD5 signatures?
The way I read this, the MD5 signature has nothing to do with the encryption of message, but are only intended for one-time password applications. The idea is this: The program creates a sequence of passwords together with their MD5 hashes. You keep the password, the computer keeps the hashes. Now, when you wish to log in, or use some password protected service on the computer, you pick the next password from the list and send it (in the clear) to the computer, which compares it with the next hash on its list. If they match, you're in. If not, tough luck. The computer is of course supposed to keep track so the same password won't work twice, hence keeping eavesdroppers from using the password.
One-time passwords can be useful, but they don't protect against man-in-the-middle attacks. Also, to make it harder to apply a dictionary attack against the password hash file (if the attacker can get a copy of it), salts should have been used, in the same way they are used in Unix passwd files.