17 Dec
2003
17 Dec
'03
11:17 p.m.
Antonomasia <ant@notatla.demon.co.uk> wrote:
How are you planning to detect which bytes are passed in this way ? Chosen plaintext attacks would do it, and show where (a+b)^(a0+b0) == 0. Looks like you've just doubled our progress.
It doesn't take chosen plaintext, just known plaintext.
If the key is reused with a different message I don't think there's a weakness. An IV is a good idea, but aren't we _attacking_ this thing ?
Lack of an IV is a problem with almost any cipher, but it seems especially so here.