Re: Remailer Attack

-----BEGIN PGP SIGNED MESSAGE-----

Robert A. Costner wrote:
>At 03:03 AM 9/26/97 -0400, Monty Cantsin wrote:
>>Right now it's [the remailer network] sort of usable, but only for the
>>dedicated, and there aren't many of us.  I seem to be the only
>>persistent authenticated nym who posts with any frequency.
>...
>>One problem with operating a nym is that people almost always respond
>>to it with suspicion and hostility, even on this list.  It would be
>>nice to have an ordinary looking e-mail address that took in messages,
>>encrypted them for your public key, and then sent them out to
>>alt.anonymous.messages for pickup.  Going the other way it would be
>>nice if the account would accept signed messages and send them out as
>>normal e-mail or news articles.  This would allow nyms to participate
>>in NetWorld like everyone else.
>
>Your desired functionality seems to be describing the operation of a nym
>server, which you are not using.

No, the nym servers differ in two important ways.

1. They use a reply block so your true identity lies encrypted beneath
the public keys of several remailers.  I would rather the message went
to alt.anonymous.messages.  (I guess this could be jury-rigged by
supplying a reply block which sends the messages to a Usenet gateway.)

Didn't the weasel remailer operator recently move his accounts over to
redneck because of police interest?  Weren't we reassured that no
identities were revealed and that they wouldn't be revealed?  That
should be a good illustration of the weakness of a reply block to an
ordinary e-mail address.

Also, by my reading of the "uptime" statistic in Raph's remailer
chart, a reply block is not going to be very reliable for receiving
mail.  Are there any nyms out there using the nym servers who will
endorse the reliability of reply blocks?  Would you run a business
using them?

2. The nym servers advertise that the accounts are nyms through the
choice of domain names.  So, presumably, people will respond with the
same hostility that they respond to any other anonymous message. 

I presume the remailer operators want to minimize the number of times
somebody is defrauded through an anonymous account.  But, by the same
logic, ISPs should carefully identify their customers the same way a
bank does.  E-mail addresses should not be used as an authentication
method.

Monty Cantsin
Editor in Chief
Smile Magazine
http://www.neoism.org/squares/smile_index.html
http://www.neoism.org/squares/cantsin_10.html

-----BEGIN PGP SIGNATURE-----
Version: 2.6.2

iQEVAwUBNCvO2paWtjSmRH/5AQEuFQf5AUPisKJypoHrU7H2Oi8bbs2GSBP+IvS3
iQJd+JO7rQv6h8rZ/qPKrgZOViAz95xNCVHLabQsIZKPJxxhjV94obu5k2+QW06S
i7PrsQM9y4+G611Cej5ND2axo9yDxprOV3/IQ/VuXvmXxyK8Xeg4Wt7GD/Skatc9
dVvTwPCD523tZJeYvKSq9l4AE/Gum+7LTEmWewz36dOm9dNokrRLcM3IqDhzhigW
np9tFPxwdrVy2AEDl8RCg8oAVzJt1sjW7CytuCK3kOyeh9hF0NlcSgE+BtMoVoqW
IhMVVeD4pMjeHWohAfyTuEEzDRPkNdBf/9YUulik+aMkR4TWnXUG+A==
=xYd9
-----END PGP SIGNATURE-----

At 01:19 PM 9/26/97 -0400, Anonymous (Monty Cantsin) wrote:
>No, the nym servers differ in two important ways.
>
>1. They use a reply block so your true identity lies encrypted beneath
>the public keys of several remailers.  I would rather the message went
>to alt.anonymous.messages.

This is an interesting idea.  I'm going to forward it to the
remailer-politics list and see what people there think of the idea.

>Also, by my reading of the "uptime" statistic in Raph's remailer
>chart, a reply block is not going to be very reliable for receiving
>mail.

The Redneck remailer, for instance, has a 99+% uptime rate.  The reason for
this being high is that it is run under a business model.  The connectivity
basically cannot go down.  There has only been 23 minutes of connectivity
downtime in the last year.  Redneck is run on a business quality network
with multiple backbone feeds.  But things break down after that.  Redneck
also runs on an underpowered 486 with no "hotfix" backup machine.  It is
not considered to be a mission critical piece of equipment.  I might notice
Redneck have a problem at 1:00 am, but it can wait until in the morning to
get fixed.  Situations like this, combined with software upgrades account
for downtime.

Back to your question of latency, the actual latency on Redneck is about
six seconds if chaining is not used.  This is "high" because the remailer
runs on an underpowered machine.  If the machine is not doing its thing for
any length of time, then the figure will go up dramatically, which is why
you see the two minute latency now.  Cracker however is *designed* to have
additional latency.  The figure shown is twenty something minutes, but
actual latency of an individual message cannot be determined. Conceivably,
it could be much higher or lower on any given message.  The latency could
be under 60 seconds during a busy time, or could be hours during a slow time.

>2. The nym servers advertise that the accounts are nyms through the
>choice of domain names.  So, presumably, people will respond with the
>same hostility that they respond to any other anonymous message. 
>I presume the remailer operators want to minimize the number of times
>somebody is defrauded through an anonymous account.

This is similar to the AOL effect.  It is often hard to have respect for an
unknown poster from AOL.  The remailer at EFGA is new.  I've come to
realize that this natural hostility towards a remailer is a good reason to
put the remailer on a separate domain.  We might do this.


  -- Robert Costner                  Phone: (770) 512-8746
     Electronic Frontiers Georgia    mailto:pooh@efga.org  
     http://www.efga.org/            run PGP 5.0 for my public key