Re: Password derived key with hash iterations
From: Jed Davis (jdev_at_panix.com)
Date: 07/29/03
- Previous message: Shai Halevi: "Re: Entropy-distributing functions?"
- In reply to: cppdev: "Password derived key with hash iterations"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Date: Tue, 29 Jul 2003 12:11:14 -0400
cppdev9@yahoo.com (cppdev) writes:
> Hello,
>
> In some libraries I see a password deriving
> function. It takes a hash iteration count.
> Does iterating over original password's hash let's
> say 100 times really gives any security?
> It seems that all someone has to know to get the key
> is the hash before the last one, am i missing something?
The idea, I assume, is to make it harder to use a dictionary attack on
the original password typed by the user. Guessing any of the
intermediate stages should be as hard as guessing the eventual key,
since with a good hash function they might as well be random bits.
Of course, if this 100x-hash isn't salted (some random bytes added to
the password before hashing and included with the ciphertext), the
attacker can precompute the 100x-hashes of their dictionary file and
apply them to any ciphertext.
-- Jed Davis |"But life wasn't yes-no, on-off. Life was shades of <jdev@panix.com> | gray, and rainbows not in the order of the spectrum." <jldavis@cs.oberlin.edu>| -- L. E. Modesitt, Jr., _Adiamante_ PGP key (fingerable) F33659F9 A098:903E:9B9A:DEF4:168F:AA09:BF07:807E:F336:59F9
- Previous message: Shai Halevi: "Re: Entropy-distributing functions?"
- In reply to: cppdev: "Password derived key with hash iterations"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Relevant Pages
|
|