COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.require"let.pub" source
C00009 ENDMK
C⊗;
.require"let.pub" source
∂AIL Dr. R. C. Prim↓Bell Telephone Laboratories MH 2C560
↓Murray Hill, New Jersey 07974∞

Dear Bob:

	The long delay in sending this letter was caused by a delay
in hiring the programmer for the project.  He is Mark Crispin.

	Here is a writeup of the present state of the %2Dialnet%1 project.
Nothing in it is proprietary, since it is supported by the U.S.
government.

	As I said on the phone, I am hoping that Bell Telephone
Laboratories will find it interesting and worthwhile to collaborate
with us.  The specific areas I have in mind are the following:

	1. You may want to make an experimental implementation
on one of your time-sharing systems.
Some months ago, before NSF decided to support our project, I 
discussed this possibility with Elliot Pinson, and he said there
might be interest.

	2. There is every indication that the %2Dialnet%1 idea will be
very popular, and that many hundreds of computer systems will implement
the protocols within a few years of their publication.  It will be
necessary to decide on standards for the modems used, and it is important
to do this correctly to achieve low cost, reliability, good performance,
expandability of capability, and compatibility with the telephone system.
One idea is that a micro-processor based modem should be found or designed
that could adapt to the telephone line it received and, to some extent, to
the modem at the other end.  If we could retain
compatibility with the 300 baud modems in common use, it would
speed up acceptance of %2Dialnet%1, but this isn't a requirement.

	3. The initial experimental implementation of %2Dialnet%1 
will use commercial modems - most likely at 1200 baud for simplicity
and cheapness.  Since I wrote the first and last part
of this letter, I have gotten mixed advice on the feasibility of the
a micro-processor based high speed modem.  As I understand it, above
the 300 baud speed modems are non-standard and are usually bought
for a fixed set of communicators.  %2Dialnet%1 may bring about a
de facto standard, and there are many opportunities for mistakes.
This is especially true, because we are only budgeted for the design
of the protocols and the experimental implementation, and we have no-one
and can afford no-one who can make deciding what to do about modems
his main business.  For the preliminary experiments, it doesn't matter
much, because we will warn everyone that the choice of modem is only
temporary.  Any advice or assistance you or anyone else at Bell Labs
can offer will be much appreciated.

	It seems to me that it is in the interest of ATα&T to take
a benevolent attitude to %2Dialnet%1.  It is part of our basic idea
that the present telephone network, although it was designed and
optimized for voice, is suitable and cost-effective for interpersonal
messages and for most file transfer even though it is too expensive
for cross-country remote login.  The decisive consideration for us
is that the these service are available now.  For these reasons,
we oppose creating an electronic mail monopoly, whether government,
ATα&T or other, or even a government regulated competition with
rates and services prescribed by the FCC.
Since in the political climate that is likely to prevail in the
forseeable future, any new monopoly is unlikely to go to ATα&T,
ATα&T should favor our view that electronic mail is a product,
not a utility.  I believe that business, science and individual
users are best served by a variety of combinations of computation,
communication, and data-base services and that any attempt to regulate
who can offer what services is likely to result in a situation in
which no-one is allowed or motivated to offer many services that
will benefit the public.  I enclose a paper on home computer terminals
that elaborates some of these points.

	I don't know whether these weighty matters are relevant to
the co-operation proposed in the first part of this letter, but I
hope you will find these views interesting and will even find it
worthwhile to transmit them to people in ATα&T concerned with such
issues.

.sgn