[Mimedefang] MIMEDefang 2.72-BETA-1 is available

John Nemeth jnemeth at victoria.tc.ca
Tue Nov 2 18:26:46 EDT 2010


On Feb 18,  2:33am, kd6lvw at yahoo.com wrote:
} --- On Tue, 11/2/10, David F. Skoll <dfs at roaringpenguin.com> wrote:
} > I guess I mis-wrote.=A0 I meant making the client port available.
} > As for the server port, it's too difficult to pass that in at
} > filter_connect time.
} 
} The server port is already "known" because it is tied to the "daemon_name"
} variable which is passed at the xxfi_connect() call.
} 
} In Sendmail, this comes from the DAEMON_OPTIONS() M4 configuration
} directive.  The field "name=3D" is what is passed, and as that is
} tied to a field "port=", and all ports must be uniquely 1-to-1 mapped
} to names, passing the name implies the unique value of port.
} 
} Pseudocode - NOT perl:
} 
} switch($daemon_name) {
}   case "MSA":  Port=587; break;
}   case "MTA":  Port=25;  break;
}   case "MTS":  Port=465; break;
} }
} 
} If you can't do something like that to obtain the value you want from
} the value you're given, ....

     No, you can't.  A name does not imply a port.  You could do it
based on knowledge of your sendmail.cf, but that wouldn't be very good
programming.  Anyways, the problem is that mimedefang creates a
subdirectory with a name based on the QueueID then creates files in
that subdirectory for communicating with the filter (including listing
macros and their values).  Since there is no QueueID at filter_relay()
time, it has no place to store information.

}-- End of excerpt from kd6lvw at yahoo.com


More information about the MIMEDefang mailing list