I'd take that bet for every cent I have - the modulo solution changes each character to another random character. Only the length of the message persists in additional runs. In the xor, some information survives each run.

update: And note in your wager, to win two things have to happen:

1) xor breaks if process repeated with small changes and
2) same for modulo.

Even if I am wrong about 1) and I may well try to prove, I could still win the bet on 2) alone. I can't prove 2) is unbreakable beyond proving that each run is as good as a fixed number of random numbers. But 1) I may well have a go at sometime during the next few years (other projects taking priority for now).

More update: your intuition is clearly a disaster. You are saying that a version which passes NO information from one run to the next is only secure if one that passes some information is secure. That's like saying a house can only be secured by locking all the doors if it is also secure when a door is left open.

-- Enough is enough -- setting tobyink to ignore.


In reply to Re^5: crypto with core modules only by TheloniusMonk
in thread crypto with core modules only by morgon

Title:
Use:  <p> text here (a paragraph) </p>
and:  <code> code here </code>
to format your post; it's "PerlMonks-approved HTML":