We live in a world where spam e-mail is an increasing problem.
The rules used by spam filters and spam blockers are constantly
being updated to catch more spam. The spammers are constantly adapting
their junk mail to get past spam filters. One side-effect is that
some legitimate messages get marked as spam. This is not a good
thing.
One type of spam-blocking rule looks at where the mail came from.
Some hosts are identified as having a history of sending spam. Any
mail coming from those hosts is given a higher probability of being
spam.
There is a growing movement across the Internet to help eliminate
this confusion by identifying which hosts should be
sending e-mail for a domain. You could think of those as systems
authorized to send e-mail, such as our Postoffices. One of the requirements
for authorized e-mail sending is that we need to authenticate users
as they send e-mail. Currently you only authenticate as you read
e-mail; with the changes CIT is making now, you will need to authenticate
(by using Kerberos or a password sent over a secure connection)
before you can send mail. The result is that it is harder
for spammers to "spoof" an address, that is, to try to
make it appear that mail is coming from a legitimate host when it's
not. As more and more systems adopt these measures, it becomes harder
for spammers to sneak through undetected. And that's a good thing.
