Hi
First, thank you very much, for that lots of answers and for being that
fast.
I've changed things:
exim4.conf
warn message = X-Scan-Mime-Error: ${if
>{$demime_errorlevel}{0}{$demime_reason}{No error found}}
demime = *
warn message = X-Scan-Virus-Found: $malware_name
malware = *
warn message = X-Scan-Virus-Found: No virus found
!malware = *
warn message = X-Spam-Report: $spam_report
spam = nobody:true
local.cf
clear_report_template
report _HITS_ points, _REQD_ required
report _REPORT_
> 1. The Spam-Report looks quite large:
Solved. Thanks a lot.
(No, i didn't know this file, but locate did. :-D )
> 2. X-Scan-Virus-Found-Headers are omitted if no virus is found...
Solved. Thanks agail.
"Untested." but works great.
> 1. spamd is accessed through TCP/IP. I would like to change this to
> unix sockets as with clamd. Is this possible?
> There are inofficial ptches around but official exiscan-acl release does
not support this yet.
Okay, but since I am running on debian and it works also fine with tcp, I
have not that much reason to change to something compiled by myself.
(Note, that I would compile myself if neccessary, but not for something that
estetical.)
What is missing:
> 2. I would like to do those scanning processes as part of the delivery
> process, since i anyway will NOT reject mail, but drop them. For me it
would
> be easier to do this afterwards because I could do that depending on the
> receipient [let users have filter/not filter settings] and it would not
> delay message reception.
I already got that hint:
http://dman13.dyndns.org/~dman/config_docs/exim-spamassassin/node12.html
Which is good to know, and I am still learning alot about exim, but I would
prefer not to start exim twice or let spamassassin modify the mail
especially since I loose a lot of information such as origin IP,
authentication information and such.
I really like the way spam= works in the acl, will these eventually be
available in routers?
Thanks,
Steffen
