problems with CHARSET_FARAWAY_HEADER rule being triggered

[alan premselaar <al...@12inch.com>]

Hi,

  It's been awhile since i've participated on the list.  I've just 
attempted to scour the entire net trying to find some information on 
this, but I Haven't found anything.

I've just installed SpamAssassin 3.01 in conjunction with MIMEDefang 
2.48 on a redhat enterprise server 3.0 machine.

The problem I'm encountering is that even with "ok_languages en ja" and 
"ok_locales en ja" in my config file, mails that arrive with a japanese 
(iso-2022-jp) subject are triggering the CHARSET_FARAWAY_HEADERS rule.

I'm running the same setup on a redhat 9 machine with version 3.0 of 
SpamAssassin and 2.45 of MIMEDefang with the same configuration options 
and i'm not experiencing this problem.

Is it possible that something broke in the 3.01 update?

for the time being, I've set the CHARSET_FARAWAY_HEADER score to really 
low (so i can see if it's being triggered, but so it won't push the 
score up) but i'd like to be able to set it back since we ocassionally 
get UCE with chinese or other foreign charsets in the subject or header.

The bayes database doesn't currently have enough emails trained to be 
active.  the system i'm not having the problem with, has an active bayes 
database.

any assistance will of course be greatly appreciated.


Thanks,

alan