HUMAN-NETS Digest V8 #33

Skip to first unread message

Charles McGrew, The Moderator

Oct 7, 1985, 5:32:00 PM10/7/85

HUMAN-NETS Digest Monday, 7 Oct 1985 Volume 8 : Issue 33

Today's Topics:

Query - WG9.1 Conference Planned,
Computers and People - Working at Home,
Computer Networks - Logical Email Addressing (4 msgs)


Date: 2 Oct 1985 15:27-EDT
From: Benjami...@G.CS.CMU.EDU
Subject: Re: WG9.1 Conference Planned (IFIP)

Does anyone have any more specifics on the conference? I'm
particularly interested in paper submission deadlines and desired
range of topics, official language(s), how many and what sort of
people are expected to attend (scientists, engineers, students,
managers, etc.), and what kind of mix is expected of east and west
block attendees.

I've written to Dr. Muhlenberg, but mail to East Berlin is notoriously

Benjamin C. Pierce


Date: Mon 7 Oct 85 11:26:07-MDT
From: William G. Martin <WMa...@SIMTEL20.ARPA>
Subject: Home computing work

Here's something of interest from Info-Hams that I thought
Human-Netters would find worthwhile:

:: ::
:: T H E W 5 Y I R E P O R T ::
:: ::
:: D i t s & B i t s ::
:: ::
:: Vol 7 #18 -- 9/15/85 ::
:: ::

Up to the minute news from the worlds of amateur radio, personal
computing and emerging electronics. While no guarantee is made,
information is from sources we believe to be reliable. May be
reproduced providing credit is given to The W5YI Report.


o Independent Computer Cottagers

Do you have some sort of computer linked business "on the side"...
maybe even full time... that you do from your home? If so, you might
want to consider being a member of AEC... the Association of
Electronic Cottagers formed in January 1985.

Formed by Paul and Sarah Edwards of (677 Canyon Drive), Sierra Madre,
California (91024), AEC is designed to support the growing number of
people who work from their home with personal computers.

The Association of Electronic Cottagers is partly a support group for
the cottage industries made possible using personal computers and
partly a rights watchdog for home workers.

Members of AEC can obtain marketing assistance, business consultation
and other services... also access up-to-the-minute news affecting
their interests through a monthly newsletter, an online hotline,
bulletin boards, electronic conferences and private databases
available through the CompuServe Information Service.

AEC evolved out of an international computer network... the
Work-At-Home special interest group (SIG) of CompuServe, an
interactive database for computer users. The Edwards' began this SIG
because they wanted to meet other entrepreneurs and believed others
had the same need.

Labor unions see working at home and telecommuting as a threat. Local
bureaucrats, using zoning laws, have put some home computer workers
out of business. AEC has put together an Electronic Bill of Rights
which, among other things, asks that legislatures make no laws
prohibiting freedom to work in one's home... when that work does not
interfere with neighbor's enjoyment of their own homes and

The Edwards' are the authors of several books on working from home
with a personal computer. Paul is an attorney... Sarah holds a Masters
degree in Social Work. Their 18-year old son is an engineering student
at UCLA.

[End this issue]


Date: Wednesday, 2 Oct 1985 20:09:58-PDT
From: goutal%parro...@decwrl.ARPA
Subject: logical email addressing

I must confess to having been out of touch for a while,
and so have not tracked all the latest proposals in this area.
I think, however, that I have a little tidbit to contribute.
First, I *am* a technical type, so am not put off by horrible
long strings of apparent nonsense -- occasionally.
However, I do get tired of typing all that stuff every time I
want to send a message to my boss or to my next-cubby neighbor,
which I do a couple dozen times a day.
I also don't like to have to remember arcane alphabet soup
when I want to send a message to an old friend to whom I haven't
sent something in a while.
Still, I can quite understand that that's what the computer needs
in order to do the job I want it to do for me.

My way around this is with logical names.
The phrase "logical name" must mean different things on different
systems, but I'm sure they usually amount to something like "a name
you can invent to mean something else, and have the system remember
your definition".

So, just as I use DAT$SLRR to mean the full file spec for an ISAM file
on short-line railroads, and LNK$LIBRARY to refer to my private
library of runtime object code modules, and HLP: to refer to the
directory where help texts, user guides, and the like are found, in
like fashion I use $JOE as the address of my office mate (his
username, which is his last name, is a little hard on the fingers), $D
for my boss, $B for a colleague across the building, $HUM for the
moderator/distributor/whatever of Humanettes Digest, and just net.rr
for the railroad newsgroup on Usenet.

It is also common for system managers to define system-wide logical
names for use by MAIL, either to reroute incoming mail for departed
users, or to make it easier for local users to send mail to complex
addresses, such as gateways or distribution list moderators.

I've never seen this done with "group logical names",
but there's no reason it couldn't be done.

VMSmail, at least, just uses the regular VMS logical-name facility,
and all this falls out.

Basically, I'm just using the logical-name facility in the operating
system as a sort of localized, personalized name server.
This buys me a fair bit until the real thing comes along.
It would be interesting to see if operating-system-dependent
logical-name facilities could be subsumed under network-wide
name services. This would mean that my collection of handy names
for people would be directly connected to the ever-more-global
tree of ever-less-handy adddresses for them,
in a more-or-less invisible way.

I'll be interested to see how the "TO" address of this message
shows up in the digest. All I gave it was "$HUM".
The commands in the background that make this possible are:

By the way, within the DEC domain,
addresses are nowhere near as complicated as they are in, say,
uucp-land. A simple nodename::username pair will send your message
winging its way directly to that user on that node,
in a network of thousands of node around the globe.
(I gather this is essentially true in the ARPA domain,
although I don't believe the set of nodes is as large.)

-- Kenn Goutal ...decwrl!dec-rhea!dec-parrot!goutal
(or however we say it this month)


Date: Thu, 3 Oct 1985 16:34 EDT
From: Jon Solomon <JSOL%BUCS20%boston...@CSNET-RELAY.ARPA>
To: BostonU SysMgr <root%boston...@CSNET-RELAY.ARPA>
Subject: Towards a more 'human' method of e-mail addressing (SOLVED!)

Well, Barry. There are several problems with that addressing scheme
that need to be worked out:

Does every e-mail user now *have* to have a phone number associated
with it? I can imagine every student at every university given a
special phone number associated with their name for the purpose
of receiving network mail. Would the reverse always apply (i.e.
if I know someone's E-mail address, do I then know their phone
number?) What about unlisted numbers?

Phone numbers are typically assigned in technically-feasable methods.
It would, for example, be impossible in the current environment for me
to get JON-SOLOMON as a telephone number (notice that I am
conveniently 10 digits long) (note: This may change as we run out of
area codes, but unless I lived in area code (JON) and had prefix 765
(SOL), *and* they hadn't assigned SOL-OMON (765-6666) to someone else,
it would at the very least be very expensive (ever check into how much
a foreign exchange costs? And my local calling area would be most
likely far away from my home (if I recall, 617-765 is not in the
Boston Metropolitan area). I have considered 800-SOLOMON, which is
about $100/month excluding usage...

You mention in the same breath that children spell things wrong,
and sometimes write letters backwards, *and* that they use the phone
system properly.

I can remember back when I first started using the phone system. I was
about 11 or 12, just entering puberty. I started by remembering my
relatives phone numbers (I still know my number from back then..).
But, you and I are not typical of 12 year olds. We were both
interested in the phone network. Also, by the time you and I were 12,
we both probably knew how to spell Massachusetts without printing
backwards. Younger children (first learning to write, for example)
would probably have the same trouble with a dial or touch tone phone
as they would have had with a pen or pencil. If I remember correctly,
kids of those age are mainly taught how to dial 0 when there is an
emergency and to tell the operator what is wrong.

Anyway, none of this really excludes using the telephone numbering
scheme as a standard for mail addresses, it just points out that
just like the scheme we use now, it has flaws.



Date: Thu, 3 Oct 85 20:58:53 edt
From: BostonU SysMgr <root%boston...@CSNET-RELAY.ARPA>
To: js...@bostonu.CSNET
Subject: Re: Towards a more 'human' method of e-mail addressing
Subject: (SOLVED!)

>Does every e-mail user now *have* to have a phone number associated
>with it?

No, you missed my point, only machines need to have
'phone-numbers', this is a scheme for host-naming,
my address in this scheme might become:

bzs@617-353-2780 w/in America
bzs@353-2780 w/in (617)
bzs@3-2780 w/in BU
or even
"Barry Shein"@one-of-the-above

(the user name is undefined in this scheme, current conventions could

of course, your mailer has a line in it's startup file like:

host bu-cs 617-353-2780

so you can send


if you like. As a matter of fact, for all the mail system would care
you could have:

host mit-mc 617-353-2780

and sending to bzs@mit-mc would just go to bu-cs, it's a silly
example, but it does show how trivial 'conveniences' could be built in
if you prefer and how ambiguity is solved, no matter what a machine
likes to be called (R2D2, BILBO etc) is up to individuals, for all
mail delivery it must end up with an envelope containing the phone
number on it.

> What about unlisted numbers?

I assume this is a more important objection presuming your first
(mistaken) assumption that each user has a #, I presume that each
machine (only) has a number. Obviously for organizational machines
this could be the same as their 'front-desk' (note: the phone number
is only a naming convention and does not imply the number is ever
actually called or a phone is involved anywhere in the actual
AUTHORITY (ie. the phone company.) In the case of personal computers,
where you do not want, say, a phone that rings in your house,
published, there are obvious solutions:

1. If you don't like to interact with people stay off the net.
(sorry, but it is a consideration.) Only use public
machines for your mail.
2. Buy a phone just for your PC and remove the ringer
(privacy costs) Maybe you did this already for a modem.
3. Set up something with someone to feed you indirectly so
only that site knows your true number (ie. an alias on
that site.)
4. Use your office phone, a friend's office phone, a pay-phone
number out in the hall or maybe just choose an 'impossible'
number (like 123-3434) and hope no one else does.

The point is this is a special case, judging by the number of phone
numbers that remain in the white pages one has to assume this is not a
universal problem, as I said, privacy costs, you will have to use a
scheme to hide your real number.

>Phone numbers are typically assigned in technically-feasable methods.
>It would, for example, be impossible in the current environment for
>me to get JON-SOLOMON as a telephone number

I assume this objection is resolved by the realization that hosts have
phone numbers, not users. The use of aliases reduces the nuisance,
besides, is your home phone # JON-SOLOMON?? Why is that ok? (or more
to the point, in what way is this worse?)

>You mention in the same breath that children spell things wrong,
>and sometimes write letters backwards, *and* that they use the phone
>system properly.

My point wasn't so much to devise a system small children can use,
only to point out that a lot of people (even children) can manage
to use the phone system and there seems to be a general feeling that
the current various host naming conventions confuse all but the most
ardent mail hackers (and even them.)


-Barry Shein, Boston University


Date: Thu, 3 Oct 1985 22:51 EDT
From: Jon Solomon <JSOL%BUCS20%boston...@CSNET-RELAY.ARPA>
To: BostonU SysMgr <ro...@bu-cs.bunet>
Subject: Towards a more 'human' method of e-mail addressing (SOLVED!)

Well, the current domain scheme is fairly simple to understand,
except that it uses brain damaged boundaries.


Stupid, who else but ARPANET people would understand why they chose
these. Fortunately, I believe the International standards will come
up with something better... *country* *names*!


What I'd love is

"Jon Solomon"@BUCS20.BU.MA.US

That would be REALLY simple, and would be basically a rearrangement of
what we are already having imposed on us. Directory assistance? Why
not have "General Delivery"? That's what most people are doing. Send
Same idea. Have a host who'se name is the same as their domain. It
accepts mail for everyone in the domain and knows how to forward.

Much easier than (617) 623-JSOL (which *is* my phone number).
No, I couldn't get JON-SOLOMON, or 617-SOLOMON, but that didn't
stop me from trying.

The Post Office, and the Telephone company basically do the same
thing different ways. They get messages to people. Both methods
have been around for quite some time, the post office much longer.

If you go adopt a telephone based system, there will probably
be some bureaucrat who want's to intervene. Look at FTS, the
Autovon, TELEX (with it's brain damaged area codes), and
Digitals DTN!

Flame off.


End of HUMAN-NETS Digest

Reply all
Reply to author
0 new messages