Munging also refers to redacting data in spam reports to prevent
listwashing or revenge or revealing spamtraps.
Nowadays, there is also Gnu Privacy Guard.
It is an open source program that does everything that PGP does, but for free.
It really is that simple. Unsolicited mail from long lost
friends or relatives is generally welcome. Solicited mail
is welcome. It's the combination of unsolicited and bulk
that is the problem.
There is some confusion about a requirement for spam
to be commercial, UCE vs UBE. That is tangled up with
the US legal system. It would probably be easier to outlaw
UCE than UBE. Most spam is commercial, but I see occasional
religious spam and every election season there are reports of
political spam.
The key idea is solicited by the recipient. This requires
informed consent. Hidden pre-checked spam-me check boxes
don't count.
Spammers generally fall into two categories: Crooks and Mainsleaze.
This is called Accept-then-Bounce. It worked fine until
spammers started forging the return info so their
accounts were harder to trace. With forged return info,
the Bounce messages are sent to innocent victims where they
look just like other forms of spam.
This is also called Backscatter. If it went back to the sender
there wouldn't be any problem.
Some people reading the RFCs claim that MTAs have to send the Bounce
if they can't deliver a message.
That's not correct. The MTA can Reject a message instead of accepting
it by returning an error code and text back to the sender.
That requires the MTA to do all the spam filtering
while the connection from the sender is still open.
The typical error was no-such-user. Spammer's collect
a lot of bogus addresses hence they cause a lot of outscatter.
It's relatively easy to check recipient validity before
accepting a message. (AOL had to rework their whole mail
system, but they thought it was worthwhile.)
There are a few cases that are harder. One example is
over-quota when the mailbox resides on an internal system
far from the gateway MTA that accepts the message from
the outside world.
It's perfectly reasonably to send a bounce as long as you
know that the recipient isn't forged. That happens
when the mail comes from one of your users. The mail goes
from their PC to your MTA/smarthost. Your MTA tries to
deliver it to the recipient's MTA. A Reject at that point
gets returned to the initial sender.
Spammer's will forge your users as the return info so just
checking the intended address isn't good enough. (But
that's harassing your users. It's not outscatter going
outside of your system.)
Main Glossary
Black Hat and White Hat refer to early western movies when the good guys all wore white hats and the bad guys all wore black hats.
Expanded Topics
A few topics are complicated enough that a quick glossary style
description isn't adequate. Here are some expanded discussions.
What is Spam?
In the context of email, Spam is Unsolicited Bulk Email.
Outscatter
Before spammers, it was common for MTAs to Accept all mail offered
to them and then send a Bounce message (DSN) back to the sender if
they couldn't deliver it.
The Bounce message tells the sender that there is a problem
so they can fix it and try again.
Links