Subject: | |
From: | "Christian J. Reichetzeder" <REICHETZ@AWIIMC11> |
Reply To: | Revised LISTSERV forum <LSTSRV-L@CEARN> |
Date: | Mon, 23 Oct 89 12:16:10 SET |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
On Fri, 20 Oct 89 14:58:33 GMT Eric Thomas said:
>When LISTSERV sends a file to the mailer with 'To: JOE@RIGEL', the mailer
>sends the file to [log in to unmask] This is a fact
>
>Christian, be reasonable. You know very well that there is no such thing
>as a domain when sending to RSCS, which is why RSCS addresses are not
>"qualified". And there is no RSCS gateway to complain about this either
>:-)
>
In other words: outbound FILES go to BITNET, outbound MAIL goes to the LAN,
correct? Well I can have several RSCS as well as several MAILERs and only one
of each set can be the networking machine. So why not give LISTSERV a MAILER
which sends by default to BITNET?
I still don't understand how [log in to unmask] can subscribe as
'JOE@RIGEL' - other than with two RSCS which will give a mess anyway. It seems
that the default meaning is reversed for inbound and outbound nodenames.
If you tell RYERSON to "get yourself a decent MAILER" then you can as well
tell whomever "configure your networking software properly". I remember the
V*X-clusters sending with the clustername as nodeid swallowing the user's
V*X-nodeid and then complaining "you must send to the individual node not the
clustername" when getting something back. Oh well ...
>Furthermore, it is generally good practice to always fully qualify
>addresses.
Yep, but need it be .BITNET ? :-). At least it's not a domain specification
but an "address-space" and somewhat redundant. And it originates from software
which first sees the local network and everything else as special case.
> Eric
Christian
|
|
|