Tag Info

New answers tagged

0

It's probably a bit late, but it's relatively easy to get an estimate of entropy. It doesn't matter if some bits are repetitive or not as it's the bits that aren't that contain entropy. You don't need any weird transformations of the data. You have what is called a bit fixing entropy source. Stick all your integers into a single file (binary or ascii - ...


3

Generically speaking you can do this but you shouldn't. It may well be possible to perform specific calculations when a random number is used for both (I'll leave it to the more theoretically inclined to create a demo if this is possible for ElGamal / DSA). Another reason is that the single secret gets known then both keys/algorithms will be compromised. ...


0

You can use xinput --test to record mouse movement data and then "crunch" the recorded data into a random hash using a hash algorithm like MD5. E.g. xinput --list (shows your mouse device id is 15) and you want to collect 5 seconds of mouse data: $> mouse_data=$(timeout 5 xinput --test 15) If you want to play around with a script that incorporates ...


6

No, you can't. In this case, an attacker can compute $m_1/m_2$ by multiplying the first ciphertext for the inverse of the second and, for instance, determine if $m_1=m_2$ or not. This should not be possible in a secure scheme.


0

Cryptography might not be philosophy, but the concept of true randomness is somewhat within that purview. It might be useful to cross post this in a philosophy forum and get their perspective. PRNGs and TRNG blur together if the PRNG is good. Some believe that *nix's /dev/random is pretty good. Some have passed what science has decided are exacting tests ...



Top 50 recent answers are included