I don't have a vested interest other than keeping my data safe. A *good* cipher will work, even if someone knows the exact implementation of it. Given complete source code to your cipher, (including the pseudo-random number generator you use) and I could easily crack it with about a week's worth of effort. If you knew anything about cryptography, you'ld know that just giving someone some cyphertext and challenging them to crack is not a valid test of the cipher's strength. Again - go get a copy of Applied Cryptography and don't come back until you have read it. Dan

Your comments are typical for persons with vested interests. Attached is a test file so you can have someone try to crack it. In the meantime, don't assume because I don't use your favorite methods I don't know what I'm doing.

Daniel R. Oelke wrote:

...

Sorry to say this - but your idea stinks.

If you are trying to send something encrypted, your receiving party would need the keys, which are the seed values of the pseudo-random number generators.

Your adversary could keep trying a whole series of different seeds until they found the correct one. With a little more work, they could analyze the pseudo-random number generator that you used, and then work backwards.

Go get a copy of Applied Cryptography Edition II. Don't come back until you have read it.

------------------------------------------------------------------

[ crap deleted ] ------------------------------------------------------------------ Dan Oelke Alcatel Network Systems droelke@aud.alcatel.com Richardson, TX