CHAA006%UK.AC.RHBNC.VAXB@AC.UK.UUCP (07/09/86)
Matts
>>> PSI -MAIL doesn't have the intelligence to go through gateways.
--- You wrote this in info-vax so I assume you are refering to PSI on
--- VAX/VMS. I can't find any problem whatsoever in using PSI-mail
--- thru gateways. You can send mail to me at the address
--- PSI%240200100303::JNET%SEQZ51::MATTS
--- and I receive it loud and clear. How does your problems show?
I'm not sure whether the original mail to which you refer originated
from Brian (CCFD8 at Sussex.Vax2) or myself; the wording looks so
similar to my reply that I suspect it's mine. The solution that you
outline is only applicable to an intelligent gateway - one that runs
the JNET software, for example. It is not PSI mail that is transcending
the gateway, it is the gateway itself which is is providing the intelligence.
If you consider the sort of gateway to which the UK academic community has
access, the original problem arises - one needs to send mail to
a pseduo-PSI mail address which might resemble the following :-
PSI%000040000040(C=<user>,<pswd>).NFS-302033500015::Postmaster
Here the 000040000040 is the numeric address of the gateway; a call-user
data field (shewn in brackets) provides access control information for
the gateway; what follows is a non-fast-select pseudo followed by the
numeric address of the ultimate recipient node. The problem lies
in the call-user-data field; PSI mailshr provides no way to specify it.
Philip Taylor (RHBNC; Univ. of London) <CHAA006%RHBNC.VAXB@AC.UK> (via EARN)