From: tcmay@netcom.com (Timothy C. May)
Date: Tue, 20 Sep 94 10:44:26 PDT
To: hfinney@shell.portal.com (Hal)
Subject: Re: On the crime bill and remailers
In-Reply-To: <199409201544.IAA04102@jobe.shell.portal.com>
Message-ID: <199409201740.KAA05008@netcom10.netcom.com>
MIME-Version: 1.0
Content-Type: text/plain



I agree with Hal Finney's point that current remailers are far from
the "ideal mix" described by Chaum and are basically not very good at
what they are supposed to do.

> I strongly disagree with this.  Anonymous remailers as presently constructed
> will be almost completely ineffective against any significant government
> attempts to surveil email traffic.  The government does have the resources
> today to defeat most uses of remailers.  Since present-day remailers lack
> padding features, the correspondence between incoming and outgoing messages,
> even with encryption, is relatively easy to establish.  This is made worse
> by the lack of general support for reordering, which renders the task
> almost trivial.

Most remailers would not hold up to even fairly simple input-output
analysis, let alone surveillance of the whole set of remailers. (I
often think that in cryptanalysis of remailers we are roughly at the
1930s level of ciphermaking, where "Gee, it looks pretty complicated
to me....I can't see any way to break it, so it must be secure"
reasoning substituted for detailed mathematical or
information-theoretic analysis.)

> Instead, anonymous remailers are clearly targetted against non-government
> traffic analysis, generally local associates, system operators, employers,
> supervisors, and so on.  They allow people to communicate without
> repercussions and retribution at work or at school.  They let people exchange
> email in an insecure environment while hiding both the message address and
> its contents.  They allow whistle blowers to expose malfeasance without
> being punished.  These are the kinds of things the remailers are good for.

These are what Julf- or Cypherpunks-style remailers are presently good
for, but the goals of getting hundreds of remailers out there with
much-improved mix characteristics is certainly intended to provide
security against a more formidable adversary than the local sysop.

> Claims here that remailers are designed to support sedition or to
> prevent government surveillance are both wrong and harmful.  This kind
> of material could show up at some future prosecution of a remailer
> operator.  It is important that we understand clearly what the capabilities
> and limitations of current remailers are.

Well, Hal is certainly entitled to his view. I strongly support ideal
mixes as a tool for obtaining increased freedom from coercive
governments (whether in Burma/Myanmar or in the U.S.), and hence view
government surveillors as the chief adversary, not the local sysop.

Is this "wrong and harmful"? 

Maybe. This is why I brought up the "supporting terrorists" language
of the Crime Act.

Free speech ain't what it used to be. But maybe it will be again, with
our help.

--Tim May



-- 
..........................................................................
Timothy C. May         | Crypto Anarchy: encryption, digital money,  
tcmay@netcom.com       | anonymous networks, digital pseudonyms, zero
408-688-5409           | knowledge, reputations, information markets, 
W.A.S.T.E.: Aptos, CA  | black markets, collapse of governments.
Higher Power: 2^859433 | Public Key: PGP and MailSafe available.
Cypherpunks list: majordomo@toad.com with body message of only: 
subscribe cypherpunks. FAQ available at ftp.netcom.com in pub/tcmay