HS> The question is, the packet header, should it have point
 HS> information in the header or not?

no... not in the PKT from nodex to nodey... that point information was in
the PKT from nodex.1 to nodex and from nodey.1 to nodey... the transfer
from x to y is straight forward, no points involved...

 HS> My current design is that the PKT header address belongs
 HS> between the two interfacing nodes, so regardless of
 HS> having 1 message or 2 messages, the header would have:

yup...

 HS> This is correct right?  The message themselves will determine
 HS> the true destination.

yup...

 HS> Comments?

don't forget that a system may use a different AKA to deliver the mail with
at times... i'm not sure about your software but i know that if i am in
zone 1 and zone 2547 and i want to send mail from 2546:xxx/yyy to
2547:aaa/bbb it should not have my zone 1 address in the PKT header...
idealy, messages in those PKTs should also not have the zone 1 address
unless they are being gated... netmail is ok gated... echomail is gated in
a completely different way and should not interferr with what you are
doing...

)\/(ark