Am 07.06.2011 20:26, schrieb Terry Reedy: > On 6/7/2011 7:35 AM, Robin Becker wrote: > >> I guess what I'm asking is whether any sequence that's using random to >> generate random numbers is predictable if enough samples are drawn. > > Apparently so. random.random is *not* 'cryptographically secure'. > https://secure.wikimedia.org/wikipedia/en/wiki/Cryptographically_secure_pseudorandom_number_generator > > One of Python's crypto wrapper modules (sorry, forget which one) was > recently modified to expose the crypto rng functions in the wrapped C > library. It should be mentioned in What New for 3.3. You might be able > to get at the same functions with ctypes.
PyCrypto has a strong pseudorandom number generator, too. -- http://mail.python.org/mailman/listinfo/python-list