On Sat, 29 Dec 2007, Quentin Garnier wrote:

> There is a setting dspam.conf where you can add root as a trusted user.

Yep.  I had that set.  But I didn't have postfix's default_privs user 
listed as a dspam Trusted user.  Setting that, plus setting up a per- 
user spam-<user> alias, plus some other magic to make fetchmail use the 
correct domain when performing initial delivery, has gotten things to 
work for me.

>> Anyone got any clue-by-fours for me?
>
> Personally I use a transport, e.g.:
>
> dspam-relearn-spam     unix - n  n      -       10      pipe
>  flags=Ru user=dspam argv=/usr/pkg/bin/dspam --user ${user} --class=spam --source=error
> dspam-relearn-innocent     unix - n  n      -       10      pipe
>  flags=Ru user=dspam argv=/usr/pkg/bin/dspam --user ${user} --class=innocent --source=error
>
> That way I avoid any limitations of postfix on |cmd aliases, and I can
> use a non-root user for the actual dspam command.

That make sense, by itself.  But I'm still not sure how the user sends 
messages to this service for relearning.  Maybe I need a bigger clue-by- 
four.  (A clue-by-eight?)

Do you add a [pair of] aliases to /etc/mail/aliases{,.db} specifying 
/path/to/the/pipes ?


----------------------------------------------------------------------
|   Paul Goyette   | PGP DSS Key fingerprint: |  E-mail addresses:   |
| Customer Service | FA29 0E3B 35AF E8AE 6651 |  paul@whooppee.com   |
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette@juniper.net |
----------------------------------------------------------------------