Eric, please can you clarify

Eric, 

  Can you clarify your intention for the list ? I'm a little puzzled by
the alternatives that are flying about. You initially stated an
intention to delay (eventually bounce) unsigned messages to cypherpunks
list. A couple of posts back you talked about "This is an unsigned
message" headers in messages. Are you intending to implement both of
these, one of these, or one now and the other later ? What type of
checking are you intending on signatures - just syntax or signature
verification ?

  I apologize for asking for clarification, but I've lost the signal in
amongst all this noise.

-Jon

-- 
  Jonathon Fletcher,
  j.fletcher@stirling.ac.uk (X400: "/S=jf1/O=stirling/PRMD=uk.ac/C=gb/")

From: Jonathon Fletcher <j.fletcher@stirling.ac.uk>

     Can you clarify your intention for the list ? I'm a little puzzled by
   the alternatives that are flying about. 

The first time this got debated was last year.  There the proposal was
check sigs, delay, eventually bounce.  That's not what I'm proposing
this time.

I mentioned this again, and the discussion started up again.  Almost
at once I wanted to avoid the flames from the previous time, so I
restricted the proposal to be for marking and delaying only.

   Are you intending to implement both of
   these, one of these, or one now and the other later ? 

First of all, I'm not "planning" on any right now.  I am considering
doing some of them and I have decided to do some related work.

The first action would be autonag; messages sent back to non-signers
from the server.  The second would be automark; header fields for
advocacy would be added to the messages.  The third would be delay for
some as yet unspecified interval.

The timing of these is not decided, and would certainly be iterative.

   What type of
   checking are you intending on signatures - just syntax or signature
   verification ?

At first, only syntax checking.  The problems of key distribution
prevent otherwise.

   I apologize for asking for clarification, but I've lost the signal in
   amongst all this noise.

No, actually, a summary was in order.

Eric

Eric Hughes writes
> At first, only syntax checking.  The problems of key distribution
> prevent otherwise.

So autonag every nym to send a public key in.

Every email has its nym checked.  No key for that nym, auto nag
Discrepant signature, flag.   No signature, autonag.

This would be actually useful.  It provides a benefit, thus
demonstrating the usefulness of crypto, and would achieve 
the goals that you desire (get us to sign our stuff) without 
making us look like seven year olds.

What you are proposing provides no benefit.  It is crypto for
the sake of crypto.  This makes us look like kids in
mighty-morphing-power-ranger suits, playing with captain
Krunch secret decoder key rings.

A check on syntax without semantics looks pathetic, in the
same way, and for the same reasons, as a captain Krunch
secret decoder ring.


-- 
 ---------------------------------------------------------------------
We have the right to defend ourselves and our
property, because of the kind of animals that we        James A. Donald
are.  True law derives from this right, not from
the arbitrary power of the omnipotent state.            jamesd@acm.org

From: "James A. Donald" <jamesd@netcom.com>
    Date: Thu, 1 Dec 1994 19:15:35 -0800 (PST)
    
    So autonag every nym to send a public key in.

Where do you send the autonag message?  Do you trust the From: or
Reply-To: fields?

			Rick

Rick Busdiecker writes
> 
>     From: "James A. Donald" <jamesd@netcom.com>
>     Date: Thu, 1 Dec 1994 19:15:35 -0800 (PST)
>     
>     So autonag every nym to send a public key in.
> 
> Where do you send the autonag message?  Do you trust the From: or
> Reply-To: fields?

It does not matter who the true identity associated with a
public key is.  My proposal is only that signatures
be checked for consistent identity, not true name.

Requiring true names would be most uncypherpunk.

So it would be up to the nym to get his true public
key to the list, not up to the list to discover the
correct public key that signs a posting by a nym.


-- 
 ---------------------------------------------------------------------
We have the right to defend ourselves and our
property, because of the kind of animals that we        James A. Donald
are.  True law derives from this right, not from
the arbitrary power of the omnipotent state.            jamesd@acm.org

From: jamesd@netcom.com (James A. Donald)
    Date: Fri, 2 Dec 1994 15:33:19 -0800 (PST)
    
    Rick Busdiecker writes
    > 
    >     From: "James A. Donald" <jamesd@netcom.com>
    >     Date: Thu, 1 Dec 1994 19:15:35 -0800 (PST)
    >     
    >     So autonag every nym to send a public key in.
    > 
    > Where do you send the autonag message?  Do you trust the From: or
    > Reply-To: fields?
    
    It does not matter who the true identity associated with a
    public key is.  My proposal is only that signatures
    be checked for consistent identity, not true name.

Hmmm.  I thought that you had also suggesting that this mechanism
should ``autonag every nym to send a public key in'' which I thought
raised the question of ``Where do you send the autnoag messages?''

Perhaps you meant for the autonag to be a notice attached to the
distributed message?

			Rick