3.972 more about e-addresses (105)

Willard McCarty (MCCARTY@vm.epas.utoronto.ca)
Mon, 29 Jan 90 19:51:43 EST

Humanist Discussion Group, Vol. 3, No. 972. Monday, 29 Jan 1990.

(1) Date: Sun, 28 Jan 90 21:17:52 EST (36 lines)
From: Natalie Maynor <MAYNOR@MSSTATE>
Subject: Re: 3.963 network queries and others (77)

(2) Date: Mon, 29 Jan 90 07:04:41 EST (49 lines)
From: "Dana Cartwright, Syracuse Univ, 315-443-4504" <DECARTWR@SUVM>
Subject: Network Addresses

(1) --------------------------------------------------------------------
Date: Sun, 28 Jan 90 21:17:52 EST
From: Natalie Maynor <MAYNOR@MSSTATE>
Subject: Re: 3.963 network queries and others (77)

I'm sure that people far, far more knowledgeable than I am will give
better answers to Helen Aristar-Day's questions about addresses.
But here's my tuppence worth: (1) Re @ and % and !, don't worry
about what they mean -- just include them as given. I *think* they
have something to do with the fact that gateways get confused by
more than one @ in the same address -- so a % converts to an @ at
the appropriate moment. (2) Were the addresses that worked for you
in the U.S. BITNET addresses? Depending upon where in Europe you
are, you may need to add .bitnet to the addressses that formerly
worked. (It never hurts to add .bitnet anyway.) (3) There is no
complete directory of e-mail addresses, probably because it would be
impossibly large. Although there are various ways to track down
addresses, the easiest way is to ask the person for it.

Whenever I have a question about odd addresses, I send a note to our
BITNET postmaster, who responds immediately with the answer. Oops.
That's right. I had almost forgotten the user-support discussion
from a while back. Maybe other people aren't so lucky. Another
source of information about addresses or anything else related to
the world of e-mail (like how to access USENET) is the list INFO-
NETS@THINK.COM. Warning: it's a very active list. I delete about
75% of its mail unread -- based on subject headers. But if you have
a question, it's an excellent place to ask it. I subscribed to
INFO-NETS by sending the subscribe command to LISTSERV@UGA (a BITNET
address). I'm not sure how one subscribes from other networks.
Since INFO-NETS is not a BITNET list, there are obviously ways to
subscribe from elsewhere. I hope this makes sense. Am writing fast
sans proofreading.

Natalie Maynor
English Department
Mississippi State University
(2) --------------------------------------------------------------57----
Date: Mon, 29 Jan 90 07:04:41 EST
From: "Dana Cartwright, Syracuse Univ, 315-443-4504" <DECARTWR@SUVM>
Subject: Network Addresses

Helen Aristar-Dry of the University of Tromso asks (regarding network
addresses) "what's the difference between @ and % and ! and . , which all seem
to be separators. And, if you don't know someone's address--or, in
my case, if you know the address that worked in the USA but not
what to add to make it work from Europe--what do you do?"

The very best technical advice I have been able to get, in years of using
both BITNET and the Internet, might be "Yes, there is a God, but She
doesn't have all the answers either." My very best technical consultants
consistently turn me down when I ask them to "just write a short technical
publication for our users which will help them form return addresses,
and translate US addresses into UK address, and so on and so forth."

In frustration I
turned to the literature and read "The Matrix, Computer Networks and
Conferencing Systems Worldwide" by John Quarterman (Digital Press, 1990,
order #EY-C176E-DP, ISBN 1-55558-033-5). And now I know why they
have so consistently turn me down.

The Matrix is 718 pages long. It is not highly technical, but it gets
well into the answer to your query. And here's my summary: There are
dozens (if not hundreds) of networks in the world, each with its own
peculiar way of forming addresses. There are things called "gateways"
(physically they are computers) which interconnect these networks. Each
gateway "understands" the differences in addressing for the particular
two networks it interconnects. E-mail may typically pass through
a number of gateways to reach its destination. So, the address of the
sender and receiver get translated over and over again in this transit.

Supposedly, if you know exactly which gateways a particular message will
traverse, you can determine how to form a suitable address. All those
delimiters (% and ! and " and @) are useful in getting the gateways to
"understand" the address. This is one reason that there is no simple
"reply" function for e-mail: no one has ever figured out an automated
scheme for getting all the delimiters just right.

If networks were static I suppose someone would eventually fix this
problem....at least the correct addressing schemes could be written down
somewhere. But the networks change all the time. So who has the time
and energy to describe how the networks worked in 1989? Or 1988? I
sympathize with the plight of the networking gurus.

Even the most skilled networking folk use trial and error to find proper
-Dana Cartwright, Director, Academic Computing Services, Syracuse University