1997-10-21 - RC5 encryption

Header Data

From: vcarlos35@juno.com
To: cypherpunks@toad.com
Message Hash: ed5e01009b159fa1862229ca2dad2a83a2b51fe415e004bc70cb0523c0ae84f9
Message ID: <19971020.202355.3926.0.vcarlos35@juno.com>
Reply To: N/A
UTC Datetime: 1997-10-21 01:46:35 UTC
Raw Date: Tue, 21 Oct 1997 09:46:35 +0800

Raw message

From: vcarlos35@juno.com
Date: Tue, 21 Oct 1997 09:46:35 +0800
To: cypherpunks@toad.com
Subject: RC5 encryption
Message-ID: <19971020.202355.3926.0.vcarlos35@juno.com>
MIME-Version: 1.0
Content-Type: text/plain



Forgive this seemingly naive question, but....

Where does RC5 derive its security from?
I know about the ROTL stuff and the key expansion
methods (I have the RSA Cryptobytes, HAC, AC, etc.)
However, I personally can't see how it would work. I know its
supposedly secure because of the data-dependent rotations, and
other features.
Correct me if I'm wrong, but a ROTL assembler command shifts all
bits to the LEFT one and moves the "overflow" bits back into the right
end of the register:
0100111011 to 0011101101 if it rotates left by 2
How does this make it hard to decrypt? The round function seems MUCH
weaker than IDEA's, CAST's, or even GOST and Blowfish.

Vincent Carlos






Thread