NetNews Usenet Archive 1992 #27/NN_1992_27.iso/spool/vmsnet/networks/tcpip/multinet/2411

Text File | 1992-11-20 | 3.6 KB | 73 lines

X-Gateway-Source-Info: INTERNET
Path: sparky!uunet!europa.asd.contel.com!darwin.sura.net!wupost!usc!news.service.uci.edu!unogate!mvb.saic.com!tgv.com!info-multinet
Date: 20 NOV 92 21:05:31 GMT
Newsgroups: vmsnet.networks.tcp-ip.multinet
X-Return-path: <[e-mail]>
X-RFC822-From:    [e-mail] (UK TeX Archive Manager <[e-mail]>)
From: [e-mail]
Subject: Unnecessary host routing in addresses resent by this list
X-Vmsmail-To: SMTP%"[e-mail]"
Organization: The INFO-MULTINET Community
Message-ID: <[e-mail]>
Nntp-Posting-Host: Mvb.Saic.Com
Lines: 58

I've noticed a strange (well, perhaps not:-) phenomenon about the headers of
mail resent by this list.  Please excuse the appearance within what follows of
addresses given in the UK's "big-endian" format: the Internet-Janet relay,
nsfnet-relay.ac.uk performs this reversal on all entities that it recognizes as
possibly forming the name of a host, and it's too much effort for me to
[attempt to] re-reverse them into what I conjecture would have been their
original format.

Anyway, the phenomenon is that mail from some senders, in particular [I think],
those outside the "original" Internet (com, gov, mil, edu, etc.) gain a
source-routing address such that any reply would be sent through tgv.com.  Here
is an example coming from the University of Turku, Finland:

> Date: Thu, 19 Nov 1992 02:08 EET
> From: HANNU (Hannu Pajunen) <[e-mail]>
> Subject: Vaporware
> To: [e-mail]
> Message-id: <[e-mail]>
> X-Envelope-to: [e-mail]
> X-VMS-To: IN%"[e-mail]"
> Sender: [e-mail]

Now if I had wanted to reply to this mail, I could have sent it to the Finnish
Internet just as easily as the poor over-burdened m/c at TGV: in fact, I could
probably have sent it better, because traffic to Scandinavia from the UK shifts
a lot faster than that going across the Atlantic, especially that which has to
go TWICE.  So the address could just as easily have appeared as:

> From: HANNU (Hannu Pajunen) <[e-mail]>

which is, I'm sure, the form in which it appeared when it arrived at TGV.  Why
provide this source-routing at all?  Surely there aren't sites (presumably
running MultiNet) that don't know how to route outside the continental US, so
have to be spoon-fed and sent via TGV?  Seems pretty unsatisfactory to me.

BTW, one criticism I have of this list, compared to many others, is that the
return address appears as that of the original submittor of traffic: the lists
I run (using home-grown DCL procedures and Deliver_Mailshr), and others such as
the MX list (which presumably uses MX's excellent facilities) all tag the mail
with the address of the list itself, so that replies can be disseminated to all
and sundry.  In fact, *my* lists tag the mail (by use of two addresses in a
Reply-To header) with the original sender's identity AND the list's name, IFF
the original sender is NOT a subscriber to the list: that way, both he/she AND
the subscribers get to see the original correspondence and the replies.

If, on the other hand, I want to reply to something that's come in through
info-multinet, and ensure that it goes to everyone else too, I have either to
remember to say /CC on the VMS-mail REPLY command, or use FORWARD and re-send
to just the list alone [as I did with this message].  But when using FORWARD, I
have to remember to give the RE: subject...

Perhaps the best solution might be to install MX --- I'm sure Matt could find
time :-)

Brian {Hamilton Kelly}
System Manager for the
UK TeX Archive at Aston University