Newsgroups: sci.crypt
Path: msuinfo!caen!zaphod.mps.ohio-state.edu!qt.cs.utexas.edu!yale.edu!yale!mintaka.lcs.mit.edu!bloom-picayune.mit.edu!news.mit.edu!jis
From: jis@MIT.EDU (Jeffrey I. Schiller)
Subject: Re: Reply to request for features of new encryption system.
In-Reply-To: brnstnd@nyu.edu's message of 27 Jan 92 02: 15:34 GMT
Message-ID: <JIS.92Jan27010155@big-screw.MIT.EDU>
Sender: news@athena.mit.edu (News system)
Nntp-Posting-Host: big-screw.mit.edu
Organization: Massachusetts Institute of Technology
References: <3258@wet.UUCP> <1992Jan25.010506.1237@qualcomm.com>
	<JIS.92Jan25193805@big-screw.MIT.EDU>
	<18784.Jan2702.15.3492@virtualnews.nyu.edu>
Distribution: usa
Date: Mon, 27 Jan 1992 06:01:55 GMT
Lines: 31

In article <18784.Jan2702.15.3492@virtualnews.nyu.edu> brnstnd@nyu.edu (Dan Bernstein) writes:
   ...
   I'd be surprised if any
   undergraduate with a disassembler couldn't break their encryption scheme
   in a few hours work. But users *will* feel safe. Trust me.

Ok, you win. Indeed users will *feel* safe. I was speaking to the issue
of whether or not they would *be* safe.

As an aside. One of the real problems with encryption software is that
users have no way to judge their quality. For most other kinds of
programs a user can determine whether or not it mets her needs based
on simply using the program. However an encryption program may be
completely weak and still appear strong to its users who are not
cryptologists.

A little while ago I was evaluating a security product whose strength
was ultimately based on the strength of its internal proprietary crypto
system. The manual made it quite plain that the system was "secure."
Carefully explaining how if you forgot critical passwords, your data was
forever gone, etc. etc. An in depth examination of the product showed that
someone with access to its source code could break it trivially (a problem
having to do with how it handled key management internally). I spoke with
the product's developers and asked why the manual was in essence lying.
There answer was quite simple. Their competition was making similar claims.
If they were "honest" about how their product worked, they would lose sales
to competition whose products are probably not any better, if not worse.

And thus the dilemma.

			-Jeff
