[5870] in cryptography@c2.net mail archive
Re: crypto camouflage in software
daemon@ATHENA.MIT.EDU (Rick Smith)
Mon Oct 11 12:23:45 1999
Message-Id: <3.0.3.32.19991011103648.00912100@mailhost.sctc.com>
Date: Mon, 11 Oct 1999 10:36:48 -0500
To: Ed Gerck <egerck@nma.com>,
"paul a. bauerschmidt" <bauersc@bauerschmidt.com>
From: Rick Smith <rick_smith@securecomputing.com>
Cc: cryptography@c2.net
In-Reply-To: <37FE887F.9D8DFBF1@nma.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
>"paul a. bauerschmidt" wrote:
>> one password will decrypt correctly, many other passwords will produce
>> alternate, valid-looking keys to fool an attacker.
>>
>> is this an example of security through obscurity (a thought which many
>> frown upon, it seems)?
At 05:12 PM 10/8/99 -0700, Ed Gerck wrote:
>
>No, it is IMO a valid example of security through ambiguity.
One time pads rely on the same general idea taken to its extreme: any
decryption is as plausible as any other. I've always thought this is the
essence of a good password encryption scheme: try to eliminate the internal
cues that indicate whether the result is valid or not. That way the
attacker can only verify a decryption by using it in a genuine
authentication transaction. If the decryption is wrong, the attempt gets
logged, leaving a trace of the attempt.
Rick.
smith@securecomputing.com
"Internet Cryptography" at http://www.visi.com/crypto/
