5 ms·
Randomness in Linux
- Tomte 10y agoThe author is mistaken, of course. But what I find fascinating is this: "Some versions of man page random(4) states usage limitations while some other versions effectively does not." For years we've been lamenting the misleading man pages, and now when they have been fixed, people dismiss the correct ones because scary sounding is better, I guess.
- nabla9 10y agoThis is not correct, of course. One 256-bit sample of entropy from hardware can be expanded into endless stream of unpredictable keys. The whitening and cryptographic pseudorandomness gives security that is indistinguishable from true randomness, entropy pool is needed just to seed the process. Only case for /dev/random would be break-in recovery security. If someone has momentary access to the internal state of RNG just before /dev/urandom is called, they can predict the output of /dev/urandom that comes next until entropy pool is updated. But if someone is able to do that, generating random numbers is probably the least of your problems.