[Wiltshire] Serving email

[Robert Longbottom]

On Sun, 2010-09-05 at 13:37 +0100, David Fletcher wrote:
> Let me try again.....
> 
> I now have my MX record delivering mail directly to my server. postfix is 
> picking it up and doing what looks like a perfect job of delivering it into 
> local ~/Maildir/ boxes. So, that's working.
> 
> Now, I basically want to remove certain well defined rubbish that I don't want 
> to see to /dev/null using procmail.
> 
> I currently have in a .forward file:-
> "|IFS=' '&&exec /usr/bin/procmail -f-||exit 75 #dave"
> 
> and the simplest possible .procmailrc would simply re-deliver messages without 
> throwing anything out like this:-
> SHELL=/bin/sh
> 
> :0
>  /home/dave/Maildir/
> 
> where the filter recipes would go in front of the delivery recipe.
> 
> But the problem is that this gives me the dreaded email loop problem.
> 
> Please be patient with me while I guess at what is happening here - postfix 
> puts the original copy in ~/Maildir/new/ where .forward sees the new message 
> and copies it to procmail. procmail then does what it's been told to do and 
> if it's not been rejected by any filters puts the message 
> in /home/dave/Maildir - so now we have two copies of the same email. forward 
> now sees the new copy, passes it to procmail and so on - a loop.
> 
> So, for now, is my guess correct? Is this what is causing the problem? Am I 
> understanding my system or am I completely wrong?

So isn't it just the case that you don't want the "re-deliver" rule
in .procmailrc?  You just have rules to do what you want with messages
(i.e. move the rubbish to a junk folder, or delete it) and if the
outcome is "do nothing" then don't touch it.  Thats how (I think)
my .procmailrc works...

eg: I just have a number of sections like this:

:0:
* ^List-Id: Discussion about mythtv <mythtv-users.mythtv.org>
.MailingLists.Mythtv-Users/

which move mailing list related messages into sub folders.  But I don't
have anything at the end that is a catch all, that just stays in my
inbox.

Robert.