replaced http://stackoverflow.com/ with https://stackoverflow.com/
Source Link

At stackoverflow thisthis question has been asked. It uses additional random entropy and a hash method (among others) to try and create a cryptographically secure pseudo-random number generator for PHP. PHP seems to use a Mersenne Twister algorithm with a large internal state and high period, but Wikipedia assures me that Mersenne Twister is not cryptographically secure.

At stackoverflow this question has been asked. It uses additional random entropy and a hash method (among others) to try and create a cryptographically secure pseudo-random number generator for PHP. PHP seems to use a Mersenne Twister algorithm with a large internal state and high period, but Wikipedia assures me that Mersenne Twister is not cryptographically secure.

At stackoverflow this question has been asked. It uses additional random entropy and a hash method (among others) to try and create a cryptographically secure pseudo-random number generator for PHP. PHP seems to use a Mersenne Twister algorithm with a large internal state and high period, but Wikipedia assures me that Mersenne Twister is not cryptographically secure.

deleted 26 characters in body
Source Link
Maarten Bodewes
  • 82.9k
  • 12
  • 131
  • 286

Kind regards, Maarten

PS the initialization of the algorithm is just 32 or 64 bit, which is why I proposed at least to reseed it thouroughly on stackoverflow, but I wonder if that would be enough, as I've unfortunately proposed on stackoverflow (mostly because creating self defined cryptographic algorithms by users is frowned upon).

Kind regards, Maarten

PS the initialization of the algorithm is just 32 or 64 bit, which is why I proposed at least to reseed it thouroughly on stackoverflow, but I wonder if that would be enough, as I've unfortunately proposed on stackoverflow (mostly because creating self defined cryptographic algorithms by users is frowned upon).

PS the initialization of the algorithm is just 32 or 64 bit, which is why I proposed at least to reseed it thouroughly on stackoverflow, but I wonder if that would be enough, as I've unfortunately proposed on stackoverflow (mostly because creating self defined cryptographic algorithms by users is frowned upon).

reordered for better readability
Source Link
Maarten Bodewes
  • 82.9k
  • 12
  • 131
  • 286

Additional Q: It would be very nice if somebody could go over to stackoverflow to see if the solution of H M is any better than using the default method and if it can be improved. Of course, the only really good way is to hold it against BSI and NIST test suites, but any improvement on the default implementation may be useful.

The source code in the stackoverflow question should be pretty easy to read, even for persons that are more mathematically inclined. Just as reference, I've included the source of the current methods within Zend PHP (which indeed seems to lack any kind of cryptographical algorithms).

Code is a partial copy of the Zend source for cryptographic analysis only. The ZEND source is protected by the PHP 3.01 license.

Additional Q: It would be very nice if somebody could go over to stackoverflow to see if the solution of H M is any better than using the default method and if it can be improved. Of course, the only really good way is to hold it against BSI and NIST test suites, but any improvement on the default implementation may be useful.

The source code in the stackoverflow question should be pretty easy to read, even for persons that are more mathematically inclined. Just as reference, I've included the source of the current methods within Zend PHP (which indeed seems to lack any kind of cryptographical algorithms).

Code is a partial copy of the Zend source for cryptographic analysis only. The ZEND source is protected by the PHP 3.01 license.

Additional Q: It would be very nice if somebody could go over to stackoverflow to see if the solution of H M is any better than using the default method and if it can be improved. Of course, the only really good way is to hold it against BSI and NIST test suites, but any improvement on the default implementation may be useful.

Additional Q: It would be very nice if somebody could go over to stackoverflow to see if the solution of H M is any better than using the default method and if it can be improved. Of course, the only really good way is to hold it against BSI and NIST test suites, but any improvement on the default implementation may be useful.

The source code in the stackoverflow question should be pretty easy to read, even for persons that are more mathematically inclined. Just as reference, I've included the source of the current methods within Zend PHP (which indeed seems to lack any kind of cryptographical algorithms).

Code is a partial copy of the Zend source for cryptographic analysis only. The ZEND source is protected by the PHP 3.01 license.

Source Link
Maarten Bodewes
  • 82.9k
  • 12
  • 131
  • 286
Loading