Subj : Re: FMail-lnx32-2.1.0.18-Beta20170905 netmail Pack
To   : Paul Quinn
From : Wilfred van Velzen
Date : Wed Jan 10 2018 09:42 am

Hi Paul,

On 10 Jan 18 12:25, Paul Quinn wrote to Wilfred van Velzen:
  about: "FMail-lnx32-2.1.0.18-Beta20170905 netmail Pack":

 PQ> That worked beautifully.  OTOH, the pack manager's handling of FMAil's
 PQ> response caught me by surprise, with...

 >> 09 Jan 18 09:03:21 FMAIL  FMail-lnx32-2.1.0.18-Beta20170905 - Pack -C
 >> 09 Jan 18 09:03:21 FMAIL  Warning: Node 3:640/384.125 is not defined in
 >> the node manager
 >> 09 Jan 18 09:03:21 FMAIL  Message #1 : 3:640/1384 -> 3:640/384.125
 >> 09 Jan 18 09:03:21 FMAIL  Create
 >> /opt/ftn/fido/outbound/02800180.pnt/0000007d.flo
 >> 09 Jan 18 09:03:21 FMAIL  Sending new mail from 3:640/1384 to
 >> 3:640/384.125
 >> 09 Jan 18 09:03:21 FMAIL  Netmail: 1,  Personal: 0,  Hudson: 0,  JAMbase:
 >> 0
 >> 09 Jan 18 09:03:21 FMAIL  Msgbase net: 0, echo: 0, dup: 0, bad: 0
 >> 09 Jan 18 09:03:21 FMAIL  Pack Active: 0.0062 sec.

 PQ> Since I didn't have a 'node' setting in binkD for the point, it meant the
 PQ> mail packet just got hung up (the default binkD ~binkp.net addressing
 PQ> wouldn't work as it's a point, and therefore not available from the WAN
 PQ> side).

 PQ> To my way of thinking it shouldn't have happened.  The pack manager's
 PQ> routing is defined as...

 PQ>   9.   3:640/* via 3:640/384
 PQ>  10.   1:* 2:* 3:* 4:* 5:* 6:* via 3:640/384

 PQ> (The other 8 entries deal with other zones & regions.)

 PQ> So, I'm wondering what I might need to add to the routing table in the
 PQ> event of further "ping" queries, especially from other 'unknown' nodes?
 PQ> Did the "Pack -C" statement have any bearing?

It might be due to the -C indeed. Although PING replies shouldn't have the 
Crash flag (I think), so shouldn't be touched by 'Pack -C'. But doing 'Pack -C' 
the right way is a bit odd:

The syntax of FMail Pack:

          FMail Pack  [ <node>...
                        [EXCEPT <node>...]
                        [VIA <node>|HOST]
                        [/A] [/C] [/H] [/I] [/L] [/O]
                      ]

So if you specify -C you _must_ specify the <node>, otherwise the behaviour is 
undefined.

What I do in my pack script is something like:

$BIN/fmail pack 2>&1 | tee -a $LOG
....
$BIN/fmail pack '*' '-c' 2>&1 | tee -a $LOG

So pack routed netmail first, and then handle the unsend crash mail that is 
left in my netmail folder.

Wilfred.

--- FMail-W32 2.0.1.4
 * Origin:  point@work  (2:280/464.112)

.