Comment Re:FIPS isn't an Algorithm (Score 2) 138
There are two issues with this.
1) Some of these algorithms depend on receiving quality random number systems from the underlying operating system. It's possible some of those random number generators have been manipulated and its going to be pretty hard to check on Windows or OSX random number generators.
2) The backdoor's do not look like (if strncmp(pass,"NSA",3) == 0) { return plaintext }. The backdoors are sophisticated mathematical weaknesses in the algorithms. A code inspection is not sufficient to detect these kids of backdoors it takes dedicated analysis by experts. Just look at some of the discussions going on right now, some algorithms are suspect and you will hear real experts going back and forth on even if a weakness exists. AES have been around since 2001, approved by NIST based on a proposal by Belgian cryptographers. Does it have a back door? Let's hope to hell not.
DES was a good algorithm in its day but it's known (sorry I can't find the citation, I think it had something to do with how the S-boxes were chosen) that very slight changes to the algorithm dramatically weakens its effectiveness. Now in DES's case that didn't happen, good values were chosen, but it would have been easy to put in a nearly invisible weakness into the algorithm.