Hi,
after a long time of waiting, the spammers have started to become
smart, so having a "+" in the local part of the mail address begins to
lose its effectiveness as spam prevention. I have thus begun to
experiment with spamassassin and exiscan and am rapidly running into
the standard problems, some of which astonishingly have not been
solved in a nice way.
One of these challenges is the task of removing existing spamassassin
headers from the message when the message first enters exim.
What I have found is:
(1)
A statement from Philip saying that he is reluctant to add such a
feature, stating that this is the reason why the local_scan interface
was invented to allow people to hook in their own mechanisms.
(2)
A number of hints to zap these headers in the system filter, which
does not work in the naive way since the system filter runs way too
late in the message processing process, causing the headers generated
by the local spamassassin to be removed as well.
(3)
A really ugly workaround involving the local spamassassin to add the
headers with a different name, then having the system filter remove
the old headers and rename the new headers to the old names
(4)
A statement that headers_add and headers_remove on router and
transprot doesn't do the trick either (which I am inclined to believe
without further research), and
(5)
Another statement which might lead the reader to the assumptioin that
the header removal can be done from an ACL, but I didn't find any
mention of that possibility in the 4.50 docs.
What is the current state of affairs regarding header removal on
ingress? I think that this option should be available in an ACL, as
this seems to be the most natural point to do this.
Greetings
Marc
--
-------------------------------------- !! No courtesy copies, please !! -----
Marc Haber | " Questions are the | Mailadresse im Header
Mannheim, Germany | Beginning of Wisdom " | http://www.zugschlus.de/
Nordisch by Nature | Lt. Worf, TNG "Rightful Heir" | Fon: *49 621 72739834
