From: cactus@bb.com (L. Todd Masco)
Date: Sat, 17 Sep 94 22:57:25 PDT
To: cypherpunks@toad.com
Subject: Re: (fwd) "Will You Be a Terrorist?"
In-Reply-To: <9409172331.AA12848@ah.com>
Message-ID: <35gl4b$qtn@bb.com>
MIME-Version: 1.0
Content-Type: text/plain


In article <9409172331.AA12848@ah.com>, Eric Hughes <hughes@ah.com> wrote:
>   [...] perhaps it's time to start seriously
>   looking at hacking list software to create mailing lists that are fully
>   anonymous and encrypted.  Has anybody started on such a project?
>
>I'd suggest that a much more productive avenue of approach would be to
>improve the aliasing facilities of a remailer provider to allow a
>pseudonym to look like a fully normal name.

I'm not sure that's a good solution.  When the Bad Guys go to the site
that originated the message and say "we want Alfred E. Neuman at
your site" (for publication of plans of some sort of machine), the
remailer operator could get busted for aiding & abetting.  Sure, Alf's
real life human will be safe, but it'd be nice to protect the remailer
operator, too.

True, the "identity stripping" can be done by a remailer as well;
However, it's trivial enough to do at the mailing list software level
(simply not including information) that it seems like it's not a bad
thing to do.  

Now, this won't give protection from traffic analysis;  In was suggestion,
I was really blurring two seperate lines of thought (I'm interested in 
PGPified mailing list software for content-hiding reasons; I'm trying to
set up a "distributed business" that I'd like to keep secure). 
Also, though, I'm not sure I want to count on anonymous remailers being
available.  If people want to effectively "chain" them, that's fine.

>Ownership of root is not necessary for this.  I know that Matt Ghio's
>mail delivery set up allows this.  At his site there's this
>'name+extra' syntax which delivers mail to 'name', but because of a
>special sendmail version 8 macro in the Received: field both the
>'name' and the 'extra' can be recovered.  The 'extra' is then an input
>into a remailer as a pseudonym.

Sure.  I'm familiar with AMS (in fact, one of my business partners is
one of the current news/postmasters at CMU, and is helping design and
produce IMAP, its replacement).  Another, better I think, possibility is
to add headers and let the MUA sort it out: you don't have to depend upon
non RFC-822 features in the MTA.  Nonstandardness is not a game you want
to get into for things like e-mail unless you have the sort of clout that
CMU/Andrew does.

>The aliasing has to happen somewhere.  It can happen at the mailing
>list exploder or at the remailer.  Since the mapping at the remailer
>is of much more general use, and since it allows one to leverage _all_
>forms of mail communication and not just mailing list, it seems like a
>much better place for that mapping to exist.  Implementation inside a
>remailer is a duplication of function--almost always a bad thing.

I'm not sure I buy that duplication of function is a bad thing in this
case: throwing information away more than once can't be worse than
only throwing it away once, can it?
  -----
I do have a couple questions that aren't terribly related:

	- Are there any known PK message formats that commute?  It seems
	 clear to me that PGP and RIPEM do not, since that makes no
	 sense for private key ciphers like DES/3DES and IDEA.

	- What I want (for other purposes) is a mailing list that has
	 its own public key; Material is encrypted to it, it decrypts it,
	 and then the material is encrypted with each recipient's public
	 key (I'm assuming a PGP base here).   Probably simply to do,
	 but has anybody done it?  No pretense of protection from
	 traffic analysis here: just to keep prying 3rd parties' eyes
	 off it.
-- 
L. Todd Masco  | "A man would simply have to be as mad as a hatter, to try and
cactus@bb.com  |  change the world with a plastic platter." - Todd Rundgren