Re: Reverseway algorithm (32767-bytes key) - full descryption.

From: James Preston (jgp@operamail.com)
Date: 12/20/02 

From: James Preston <jgp@operamail.com>
Date: 20 Dec 2002 14:49:39 +1100

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 19 Dec 2002 19:32:24 -0800, Bagrat Mirakian <mirakian@miraxus.com> wrote:
> The fifth step makes cipher unbreakable and impossible for decryption
> against any attacks such as differential and linear cryptanalysis.

Oh, dear. Using XOR to expand a key does not a One-Time-Pad make.

- --
James Preston
-----BEGIN PGP SIGNATURE-----

iD8DBQE+ApMwgXK32hUOOt0RAhqnAKDF2X2vMbD8Fr/ZfKGXf86L0TqdWACdGS4C
sSC8o0uXHhl0UwaQp7Bw1jI=
=04NZ
-----END PGP SIGNATURE-----

Relevant Pages

  • even better description of algoithm with an attempt at a security proof.
    ... K2 = xor Pi) ... Collisions are coincidental in stream mode as for example, ... plaintext attacks: ... Traffic analysis may leak information on the nonce and therefore n which is ...
    (sci.crypt)
  • Re: Crypto Mini-FAQ
    ... >> often that attacks are possible. ... >> checksum is linear. ... Nikita Borisov, Ian Goldberg, and David Wagner. ... linear checksums commute with xor. ...
    (sci.crypt)
  • Re: Encrypting passwords
    ... > be used the next time the user wants to invoke the application. ... >I would just need simple XOR encryption/ decryption. ... XOR each byte with a fixed value, and do the same thing again to ... MVP VC++ FAQ: http://www.mvps.org/vcfaq ...
    (microsoft.public.vc.language)
  • Re: Minimal crypto OTP by dummie
    ... machine99 wrote: ... > Then combine the key with the message using XOR or another ... Adding modulo the number of characters will do. ... character flipping) attacks are still possible. ...
    (sci.crypt)
  • Re: Newbie: Arent Symmetrical Crypto Systems Unbreakable in practical use?
    ... >> Lets say we have an n bit key and an n bit plaintext. ... > There are several types of attacks that come up in practice. ... > The first two are normally used to exploit a weakness in the cipher. ... the standard for linear cryptanalysis is to ...
    (sci.crypt)