Upgrading to 2.6x:
------------------

The header re-writing system has changed substantially, allowing for
greater flexibility and customization of headers. Various other
options have been added. Read man Mail::SpamAssassin::Conf for more
info.

Furthermore, many new rules and tests have been added, and old ones
have been changed. Nearly all negatively scoring tests have been
removed. In general, most other scores have been lowered slightly, so
if you have changed scores in /etc/spamassassin/, you may wish to
change them back, otherwise your scores may be skewed too high,
resulting in false positives.

spamd now supports UNIX-domain sockets. This is strongly recommended
if spamd and spamc are running on the same server. Look at the
--socketpath option for spamd and the -U option for spamc.

SpamAssassin no longer uses AnyDBM_File to access its Bayes database,
instead, DB_File is used. If your bayes database is not in dbm format,
you will need to run 'sa-learn --import' before you can use bayes. To
check what database format you are currently using, use the following
command:
 $ file ~/.spamassassin/bayes_{seen,toks}
If the output does not contain "Berkeley DB", you are not using
DB_File and you will need to run 'sa-learn --import'

Before upgrading, it is best to run 'sa-learn --rebuild' to sync your
Bayes journal with the database. Otherwise, your journal will be
erased because of incompatibilities with the previous database.

In order to properly use RBLs and other network tests, you will want
to set the trusted_networks option. (The autodetection algorithm is
not perfect.) Read man Mail::SpamAssassin::Conf for more info.

Upgrading to 2.5x:
------------------

SpamAssassin now has a Bayesian Learner component to it. It is by
default set to automatically learn from your incoming mail. (Messages
that score less than -2 or above 15 will be learned as nonspam or spam
repsectively.) You can also feed it with the sa-learn command (man
sa-learn for info).

The report format for SpamAssassin has changed. Spam messages are now
attached as an rfc822 attachment. If you do not want this, set
report_safe 0. (See man Mail::SpamAssassin::Conf for more info.)

Support for different score sets has been added. This means that each
test will get a different score depending on whether Bayes or Network
tests are being used.