02-23-2012, 12:57 PM
I just had a brain-fart... I wonder if it would be worth it to have the program "jump around" the entire keyspace at pre-defined or "random" intervals... What i mean is a new "feature" where the program could do the following until the whole keyspace has been scanned:
Using 1 to 1000 (for simplicity) we could say that the program would work for a random (or pre-defined) period of time on a chunk of the keyspace and then jump randomly to another...
For example:
works from 1 to 120 (for x hours) then jumps to 700
works from 700 to 815 (for x hours) then jumps to 455
works from 455 to 600 (for x hours) then jumps to 940
works from 940 to 1000 (for x hours) then jumps to............
etc etc.. until the whole keyspace is worked through. Now obviously chunks that have been completed will not be worked on again. I wonder if this kind of randomization would increase "luck" of passwords like "zxxzxxzx" which would be found at the very end of a cracking cycle. If we "jump around" on a hash that would take a total of "5 days" for instance, maybe we could get the thing in 20 hours????
Using 1 to 1000 (for simplicity) we could say that the program would work for a random (or pre-defined) period of time on a chunk of the keyspace and then jump randomly to another...
For example:
works from 1 to 120 (for x hours) then jumps to 700
works from 700 to 815 (for x hours) then jumps to 455
works from 455 to 600 (for x hours) then jumps to 940
works from 940 to 1000 (for x hours) then jumps to............
etc etc.. until the whole keyspace is worked through. Now obviously chunks that have been completed will not be worked on again. I wonder if this kind of randomization would increase "luck" of passwords like "zxxzxxzx" which would be found at the very end of a cracking cycle. If we "jump around" on a hash that would take a total of "5 days" for instance, maybe we could get the thing in 20 hours????