Re: password derived key

From: Alan (a__l__a__n_at_hotmail.com)
Date: 04/13/05 

Date: Wed, 13 Apr 2005 16:06:03 -0400

One more thing...

"mike" wrote:
> i would still
> think using hash of user chosen message is a better way like hash("my
> #nam#e #is # al#an") with '#' after every 3 characters is a better key
> when considering basic dictionary attacks.

You're relying on a secret algorithm in addition to a secret key. The
secret algorithm (not so secret anymore since it is posted on this forum!)
is like a very simplistic hash function applied to the secret key. It is
commonly accepted (per Kerckhoff's Principle) that you should not rely on
secrecy of anything other than the key itself for security of a
cryptosystem, assuming the use of a sound crypto algorithm.

So you should focus your attention on the original choice of the pass
phrase, not on applying subsequent secret transformations to the pass
phrase. That's what the diceware approach does. It's not the only way, but
I know of none that is better. Now, remembering those diceware pass phrases
is another matter....

Alan

Relevant Pages

  • Re: Cryptographic hash function for small microcontroller
    ... The microcontroller is fairly fast ... SHA-1 and related hashes all require too much RAM for this ... <Then you can't have a cryptographic hash. ... microcontroller) to a PC using a shared secret. ...
    (sci.crypt)
  • Re: AES as hash function and PRNG
    ... > For encryption I decided to use AES-128. ... And they have to remain secret. ... If you have less than 16 bytes to hash, just pad the key with ASCII ... Choose any input as the plaintext to run through AES. ...
    (sci.crypt)
  • Re: Algorimic Complexity Attacks
    ... For instance, in a hash table, the performance is ... while using a keyed hash function offers the best ... It requires that a cryptographically random secret is used ... Now the promised attack on using a keyed hash function with the above ...
    (Bugtraq)
  • Re: CRC as authentication?
    ... |>instead of the more expensive universal hashes that require field ... If the polynomial is secret, and if you use the CRC correctly, ... | conditions on how you use it, then the resulting hash is an AXU-2 hash. ... If the attacker can force *almost* the same message to be transmitted ...
    (sci.crypt)
  • Re: Algorimic Complexity Attacks
    ... > keyed hash. ... If the secret itself is not leaked in the attack (and it ... this does have its difficulty: maintaining existing entries. ... This means the attack will be thwarted if the secret hash function (e.g. ...
    (Bugtraq)