🚀 go-pugleaf

In article <R90TwB1w165w@west.darkside.com>, sspy@west.darkside.com (mshannon) writes:
> Can anyone define pre and post whitening and differential cryptanalysis, 
> please?

Differential cryptanalysis works by feeding in pairs of inputs with
known exclusive-or values, and looking at the exclusive-ors of the
output pairs.  Except for the S-boxes, all the other steps in DES are
effectively constant over such operations.  The effect of the S-boxes
can be described probablistically.  By measuring the distribution of
the outputs from feeding in lots of input pairs, one can learn what
the input pairs' XOR value was, which in turn gives information on the
key bits.  (Yes, I know that I'm being quite unclear here.  Sorry, it's
the best I can do.)

> Someone stated that with a mass DES chip machine using brute force to 
> attack the DES, that the mode doesn't matter. Why even use it then? 
> Again, thanks in advance!

There are other attacks besides exhaustive search.  For example, with ECB
mode, occurences of the same plaintext on an 8-byte boundary will yield
the same ciphertext.  An opponent can build up a dictionary this way.
It's also much easier to mount cut-and-paste attacks on the *protocol*,
as opposed to the cipher, since each block is encrypted independently.
For more information, I suggest reading

@book{daviesprice,
   author = {Donald W. Davies and Wyn L. Price},
   edition = {second},
   publisher = {John Wiley \& Sons},
   title = {Security for Computer Networks},
   year = {1989}
}

and

@article{kent+voydock,
        author = {V.L. Voydock and S.T. Kent},
        title = {Security Mechanisms in High-Level Network Protocols},
        journal = {ACM Computing Surveys},
        volume = 15,
        number = 2,
        pages = {135--171},
        month = {June},
        year = 1983
}