several questions about remailers

[Nomen Nescio]

Some questions to remail operators:

1. How do you deal with spam? Is this handled by the exit node?
Do you run spam assasin on it or something? What is to keep one person
from just mass-mailing through the remailers?

2. How many emails go through a remailer per day, counting or not
counting dummy messages or pinger messages? (estimate?) Is there a
recommended ratio of #messages / #remailers? Or
#messages/second/#remailers? I read somewhere that the network
functions best with a lot of traffic.

3. For message senders, is there anyway to deal with particular exits
just snooping and dropping messages/recipients they don't like?

4. What if I don't trust the system enough, so I create my own
remailer and add it. Whenever I send a message, I always send it
through my remailer, which I assume is trustworthy.
If my remailer is 'b'
a->b->c->d
a->b->c->b
b->c->d->a

Does anyone here recommend this or not? If you do use this approach,
is there a 'bad' position to put your remailer in the chain?

5. Why do some remailers seem to have low reliability? What is so
hard about reliably sending an email?

6. I am sure this has been asked, but clearly excess headers break
anonymity, and mixmaster removes excess headers. Where can I find
more information on what is going on with mixmaster headers?

7. Of the options of sending through Tor, one can either:
a. use the SMTPRELAY option set to "localhost" and use socat to
connect to one of the hidden service smtp servers.
b. or, one could conceivably route sendmail directly through Tor
to connect to the remailers (for instance using TAILS or
transparent Torification).

Which option a or b is better, or does it matter? I am aware many Tor
exit nodes block port 25. Do the remailers usually accept mail on port
25, or SSL ports? How would one have sendmail go through Tor if
transparent proxy is not an option?


Thanks

[Zax]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On Fri, 24 May 2013 00:49:37 +0200 (CEST), Nomen Nescio wrote in
Message-Id: <050adf6703fec550...@dizum.com>:

> Some questions to remail operators:
>
> 1. How do you deal with spam? Is this handled by the exit node?
> Do you run spam assasin on it or something? What is to keep one person
> from just mass-mailing through the remailers?

It's up to the operator's person preferences. Exit remailers can run
SpamAssassin and whitelist other remailers to avoid false positives on
intermediate hops.

> 2. How many emails go through a remailer per day, counting or not
> counting dummy messages or pinger messages? (estimate?) Is there a
> recommended ratio of #messages / #remailers? Or
> #messages/second/#remailers? I read somewhere that the network
> functions best with a lot of traffic.

Remailing is all about hiding in the crowd. The bigger the crowd, the
easier it is to hide. The number of messages processed is published by
each remailer. Stats sources publish this info; here's mine:
http://pinger.banana.mixmin.net/thesaurus/
Check out the stats column.

> 3. For message senders, is there anyway to deal with particular exits
> just snooping and dropping messages/recipients they don't like?

The sender can hardcode any exit remailer they prefer.

> 4. What if I don't trust the system enough, so I create my own
> remailer and add it. Whenever I send a message, I always send it
> through my remailer, which I assume is trustworthy.
> If my remailer is 'b'
> a->b->c->d
> a->b->c->b
> b->c->d->a

That's an excellent solution providing your remailer is public and used
by many other people.

> Does anyone here recommend this or not? If you do use this approach,
> is there a 'bad' position to put your remailer in the chain?

No bad position really. Providing you use it in the same manner as the
other users of it.

> 5. Why do some remailers seem to have low reliability? What is so
> hard about reliably sending an email?

Exit remailers should always have high reliability unless they're
suffering server problems. Middleman remailers are more complex because
they cannot respond directly to the pings sent to them; they have to
generate a new Mixmaster message with the same payload and send it to an
exit remailer. This has the potential for messages to go missing.

> 6. I am sure this has been asked, but clearly excess headers break
> anonymity, and mixmaster removes excess headers. Where can I find
> more information on what is going on with mixmaster headers?

Headers don't break anonymity because they are all wrapped up into the
payload and delivered in identically sized packets. Remailers block
some headers to prevent abuse. These blocked headers can be checked by
looking in the Conf column of the previous link.

> 7. Of the options of sending through Tor, one can either:
> a. use the SMTPRELAY option set to "localhost" and use socat to
> connect to one of the hidden service smtp servers.
> b. or, one could conceivably route sendmail directly through Tor
> to connect to the remailers (for instance using TAILS or
> transparent Torification).
>
> Which option a or b is better, or does it matter? I am aware many Tor
> exit nodes block port 25. Do the remailers usually accept mail on port
> 25, or SSL ports? How would one have sendmail go through Tor if
> transparent proxy is not an option?

Some remailers support alternative ports for SMTP, usually port 2525.
Others have Tor hidden services. I don't think it really matters which
of these methods you choose to use.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
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=aFWw
-----END PGP SIGNATURE-----

--
pub 1024D/228761E7 2003-06-04 Steven Crook <st...@mixmin.net>
Key fingerprint = 1CD9 95E1 E9CE 80D6 C885 B7EB B471 80D5 2287 61E7
sub 4096R/4ABF07E3 2012-02-11 [expires: 2013-02-10]

[Fritz Wuehler]

ja...@invalid.com wrote:

> On Fri, 24 May 2013 00:49:37 +0200 (CEST), Nomen Nescio
> <nob...@dizum.com> wrote:
>
> >Some questions to remail operators:
> >
> >1. How do you deal with spam? Is this handled by the exit node?
> >Do you run spam assasin on it or something? What is to keep one
> >person from just mass-mailing through the remailers?
> >

> >2. How many emails go through a remailer per day, counting or not
> >counting dummy messages or pinger messages? (estimate?) Is there a
> >recommended ratio of #messages / #remailers? Or
> >#messages/second/#remailers? I read somewhere that the network
> >functions best with a lot of traffic.
> >

> >3. For message senders, is there anyway to deal with particular
> >exits just snooping and dropping messages/recipients they don't like?
> >

> >4. What if I don't trust the system enough, so I create my own
> >remailer and add it. Whenever I send a message, I always send it
> >through my remailer, which I assume is trustworthy.
> >If my remailer is 'b'
> >a->b->c->d
> >a->b->c->b
> >b->c->d->a
> >

> >Does anyone here recommend this or not? If you do use this approach,
> >is there a 'bad' position to put your remailer in the chain?
> >

> >5. Why do some remailers seem to have low reliability? What is so
> >hard about reliably sending an email?
> >

> >6. I am sure this has been asked, but clearly excess headers break
> >anonymity, and mixmaster removes excess headers. Where can I find
> >more information on what is going on with mixmaster headers?
> >

> >7. Of the options of sending through Tor, one can either:
> > a. use the SMTPRELAY option set to "localhost" and use socat
> > to connect to one of the hidden service smtp servers.
> > b. or, one could conceivably route sendmail directly through
> > Tor to connect to the remailers (for instance using TAILS or
> > transparent Torification).
> >
> >Which option a or b is better, or does it matter? I am aware many
> >Tor exit nodes block port 25. Do the remailers usually accept mail
> >on port 25, or SSL ports? How would one have sendmail go through
> >Tor if transparent proxy is not an option?
> >
> >

> >Thanks
>
> I'm sure the remops will spend the entire next week of their lives
> seeking to answer your farcical post.
>
> Anyone else smell troll?

It really isn't a troll...I know some of these questions have been
answered at different times but I can't find the answers.

not to mention most are easily answered one-liners i am sure.

I am thinking of running a remailer so these are relevant questions.

thanks for answering

[Anonymous]

> I'm sure the remops will spend the entire next week of their lives
> seeking to answer your farcical post.

> Anyone else smell troll?

Well as you can see, not everyone is a petty little jerk like
you, accusing and slandering someone who is asking valid questions.
As you can see below, Zax was mature and kind enough to answer his
questions.
I think this guy is the one who doesn't know how to post anymore
of a message in apas other than to accuse others of bring a troll.

[Anonymous]

Nomen Nescio wrote:
> Some questions to remail operators:
>
> 1. How do you deal with spam? Is this handled by the exit node?
> Do you run spam assasin on it or something? What is to keep one person
> from just mass-mailing through the remailers?
>
> 2. How many emails go through a remailer per day, counting or not
> counting dummy messages or pinger messages? (estimate?) Is there a
> recommended ratio of #messages / #remailers? Or
> #messages/second/#remailers? I read somewhere that the network
> functions best with a lot of traffic.
>
> 3. For message senders, is there anyway to deal with particular exits
> just snooping and dropping messages/recipients they don't like?

assuming that exit nodes are likely untrustworthy is a good rule of
thumb and should help to keep you from becoming lax when making
important, private posts. You could at these times make use of
encryption to your ultimate party. You can also consider using a VPN
after the exit node. Finally, when you connect with a tor hidden
service, you would have end-to-end encryption within tor and not leave
its net.

personally, for anonymous messaging in general I think it is better to
use different approaches for different purposes. Anonymous messaging is
a craft, and like any other craft, it requires practice and study and
ingenuity.

among others, I like to keep a few simple but pragmatic prescriptions in
mind:

1. don't put all you eggs in one basket
2. use the right tool for the task at hand
3. some things are better left unsaid
4. only someone you trust can betray you

[Anonymous Remailer (austria)]

In article <ubatp8585o3rforn5...@4ax.com>

ja...@invalid.com wrote:
>
> On Fri, 24 May 2013 00:49:37 +0200 (CEST), Nomen Nescio
> <nob...@dizum.com> wrote:
>

> I'm sure the remops will spend the entire next week of their lives
> seeking to answer your farcical post.
>
> Anyone else smell troll?

Nope.

[Fritz Wuehler]

>7. Of the options of sending through Tor, one can either: a. use the
> SMTPRELAY option set to "localhost" and use socat to connect
> to one of the hidden service smtp servers. b. or, one could
> conceivably route sendmail directly through Tor to connect to
> the remailers (for instance using TAILS or transparent
> Torification).

>Which option a or b is better, or does it matter?

"b" is not an option. It will not work, because tordns cannot perform
MX queries, and torsocks forces your DNS traffic through tordns. And
you must use torsocks, because sendmail does not have SOCKS
capability.

Torsocks used to have the ability to disable tordns (so you can use
your own DNS server) - but that feature was broken, and later pulled.
So ATM, it's impossible.

>I am aware many Tor exit nodes block port 25.

Yes, but that obsticle can be cleared. Some tor nodes accept port 25,
and you can limit traffic through them.

>Do the remailers usually accept mail on port 25, or SSL ports?

Yes, but some mixmaster remailers are persnickety and will not accept
all RFC-compliant messages. E.g., if you don't own a domain name, the
RFC allows for a literal name (i.e. an IP) in the EHLO field. But if
you use an IP, some mixmaster SMTP servers will (foolishly) refuse the
connection.

Yes, it's ironic that a server offering an anonymous service requires
a name. Although a fake name will work, this would be like an AIDS
clinic that offers an anonymous test asking you for your name. If
it's anonymous, what's the point in asking? Anyway, I digress.

>How would one have sendmail go through Tor if transparent proxy is
>not an option?

It's not possible without modifying the software.

[Anonymous Remailer (austria)]

In article
<1d831401bea42d88...@msgid.frell.theremailer.net>
Fritz Wuehler <fritz@spamexpire-

Exactly. The spam prevention excuse for demanding a name
doesn't fly.