Viewing Single Post From: Brute Forcing The Adfgvx Cipher

Viewing Single Post From: Brute Forcing The Adfgvx Cipher
osric	Mar 24 2009, 05:20 PM
Advanced Member Posts: 51 Group: Members Member #3,254 Joined: March 15, 2009	jdege, First let me make a few comments on your brute force cryptarithm programs, which I hope you will find helpful. In essence the times you quote are very slow. For a base-10 cryptarithm my brute program goes through all perms in 2 seconds, and for a base-12 cryptarithm in four and a half minutes (on my old 2.3GHz PC). I suspect that either you are not using the fastest algorithm for generating perms, or your method of handling numbers is not optimum (or your computer is very old! which I doubt). As an example this base-12 solved in 99 seconds with my brute force program: SIGHT+SOUND+SCENT=TTHESN; TASTE+TOUCH=EIIOD. As a second comment, hillclimbing is much faster and thus much better for the higher bases. For example the base-12 addition above solves in less than one second. For this reason I don’t use brute force any more. Fastest of all, and useful for really high bases, is a method called intelligent brute force which I have written about in the Cm. If you would like to see any of my programs (written in C++) I’d be happy to send them to you. Turning now to ADFGVX, I am afraid I am not finding your inputs helpful to develop or improve my algorithm. It’s been self-evident since the beginning of this thread that ANY algorithm based on brute force will take too long to solve ADFGVX with a transposition key of 15 to 20. Even brute forcing a base-20 cryptarithm (a trivial task compared with solving ADFGVX) would take thousands of years! Your comment 'find a heuristic search that does work' is of no more help than ‘find a way of making gold out of lead’! When you have the time to devote to this topic, and can suggest something of practical use, I assure you I shall be all ears.


Brute Forcing The Adfgvx Cipher · General