home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.crypt
- Path: sparky!uunet!cis.ohio-state.edu!pacific.mps.ohio-state.edu!linac!uchinews!machine!chinet!schneier
- From: schneier@chinet.chi.il.us (Bruce Schneier)
- Subject: Re: Questions about MD5 and DES
- Message-ID: <Bxu7xn.J10@chinet.chi.il.us>
- Keywords: DES IDEA MD5
- Organization: Chinet - Public Access UNIX
- References: <1992Nov16.055949.1294@seanews.akita.com>
- Date: Tue, 17 Nov 1992 01:53:46 GMT
- Lines: 44
-
- In article <1992Nov16.055949.1294@seanews.akita.com> billmcc@seanews.akita.com (Bill McCormick) writes:
- >I've been hacking together a little program that does encryption using
- > key, decryption using another, then re-encryption using the
- > first key. Is this correct? i.e., E3(P,K) = E(D(E(P,K1),K2),K1)
-
- Yes.
-
- 2. Has triple DES actually been proven to be stronger than a single
- > DES encryption? I guess this boils down to whether DES has been
- > proven to be a group.
-
- DES is not a group. See Keith Cambell and Michael Weiner's paper in the
- Crypto 92 proceedings, whenever Springer-Verlag gets around to publishing
- them.
-
- >3. I've been using MD5 to hash an arbitary user-entered string into
- > a key for the encryption process. Is MD5 really suitable for the
- > task? It appears to be a particularly good hash for long strings
- > of text, but is it suitable for short password-length text?
-
- Yes.
-
- >4. In order to provide fast user-feedback if an incorrect decryption
- > key is entered, I'm storing a "checksum" of the first block of
- > plaintext in the encrypted file. I was originally using a
- > simple checksum, but that might give away too much information
- > about the first block. Now I apply MD5 to the block and take
- > a byte of the message digest as the check digit. Is this
- > reasonable or would it be better to simply use the first
- > 8 bytes of the file as a DES/IDEA key to encrypt a constant and store
- > a byte of that constant?
-
- Encrypting a constant is always dangerous, but both DES and (so far) IDEA
- are resistant to that kind of known plaintext attack. Going with a checksum
- just seems safer.
-
- >5. Which is stronger, CBC or CFB? Is one preferred over the other in
- > cases where byte-per-byte output isn't important?
-
- Both are secure. CBC is more efficient, since you are encrypting data in
- 64-bit blocks.
-
- Bruce
-
-