COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M1BASL30\M2BASB30\M3BASI30
C00014 ENDMK
C⊗;
\\M1BASL30;\M2BASB30;\M3BASI30;
\F1\CMar 3, 1981

Mark Browning, Manager: Marketing Support
Chromatics Inc.
2558 Mt. Industrial Blvd.
Tucker, GA 30084

Dear Mr. Browning:
\J
It was a pleasure to talk with you about the Chromatics 7090.
It appears that these machines would make an elegant base on which to 
develop my  LISP effort. As I mentioned, I am enclosing some information
about LISP and its potential applications: Artificial Intelligence and
education are two concerns that drive my efforts. Assuming that this supporting
information carries sufficient weight to convince your company that LISP  
applications are a viable market,
what then could be a mutually profitable arrangement for developing  TLC's
LISP?

As I mentioned on the phone, I am receiving a grant from the Navy group that
sponsored the UCSD Pascal effort. This is quite encouraging since a major
motivation for establishing The LISP Company was to develop a \F3de facto\F1
standard LISP in a manner similar to that of UCSD. I have designed 
and implemented
a  healthy dialect of MacLISP/LISP machine LISP that runs on Z-80-based systems,
and have extended that implementation to run with extended addressing
 (18-bit addresses). A prototype display-based programming environment is being
implemented now. This extended system will supply approximately 1/3 of a KA-10
to run LISP --both in speed and space--
for about $6K, quantity-one retail hardware price.
Though quite useful for medium-sized  projects, large scale applications
require larger address spaces, more powerful LISPs, and  more elegant 
program development systems. This is what brought the Navy to
me and brings me to Chromatics.

The initial Navy grant is to port the existing
TLC language from the Z-80 onto the 68000.  This will  give the current TLC 
users an immediate boost in processing power. This improved performance is 
necessary for  several  military projects; the one driving this effort is the
introduction of several thousands LISP processors as nodes in an intelligent
CAI effort to train Naval personnel.  One segment of this operation involves 
a color graphics interface to a large mathematical model of a ship's 
steam plant. One can watch any of several dynamic views of the plant, and
control the steam plant by adjusting  valves and flipping switches
on the screen. The entire system is written in LISP, and  segments of it
are  running on TLC's single-bank LISP system. Larger segments will be
moved to the multi-bank configuration, and finally moved to 68K-class
LISP  machines.

In another TLC-based project that involves color graphics behavior, a group
in the mid-west is developing an "automated drafting system" 
for an Air Force contract. This system utilizes 
TLC-LISP both as a data-base language and as the implementation vehicle for
a natural language understanding system. In this application, the  designer
is presented  information on a large Tektronics color monitor. 
I can supply more details on either of these systems if you desire.
Both of these
systems involve using LISP essentially as a general-purpose development tool
including arithmetic and graphically capabilities; both systems
are constrained by the limitations of the Z-80 hardware and are anxiously 
awaiting the new system.

After the opening phase of porting the language,
I plan to enhance the language, adding more of the traditional language
artificats as well as integrating several of the Smalltalk-spawned  ideas
related to objects and classes.
Within this latter context, I expect to add a second generation of the
programming environment, and here I expect an advanced  window-system
will play a crucial role.  In short, then, TLC will be a major force in developing
LISP as a commercial tool.

Of further interest, I am teaching a LISP seminar and workshop
at the Western Institute
in Computer Science this summer (July 6-10) at Santa Clara University.
This Institute, organized by the founders of the Santa Cruz CS Institute,
will include courses in VLSI design, ADA, Operating systems, and data base
design, as well as the LISP class. Drs. Larry Masinter and John McCarthy will
be joining me in teaching  a group of about forty people involved in developing
industrial AI programs; each day will include at least four hours of active
computer usage of LISP. It occurs to me that it could be quite advantageous
to Chromatics to have a LISP system available for that session. I believe that
I could port and develop reasonably enhanced version of TLC LISP in 
that time-span. 

I believe that these two issues 
--the availability  of Chromatics LISP to the Navy projects,
and the introduction of Chromatics to the Institute LISP personnel--
contain the basis for a mutually attractive  cooperative venture.
I would 
suggest the loan, lease, or substantial discount 
on Chromatics machines to support the project,
for a  major problem for TLC is funding. The intial Navy grant  was funded 
"between the cracks" in a very tight budget, sufficient for one machine and
one modest FTE. To move effectively on this project, I need to support
a larger staff; funds not allocated for machines can be redistributed for 
salaries.

Staffing of LISP/AI projects is quite difficult now; few LISP programmers
exist 
and even fewer are fluent in LISP implementation details. I am 
indoctrinating two people into the 
art of LISP implementation  now as part
my lectureship at Santa Clara University; those people 
are beginning with  the
Z-80 projects and 
will move to the 68K. I expect to attract a few Stanford people
for the 68K project, but that will require a facility in the Palo Alto area,
and that requires more funds. I am considering a bank loan to 
cover these costs, but  currently  am severely limited financially.

Also, perhaps of future interest are my continuing plans for exploiting the
educational application of LISP. I am currently a lecturer at Santa Clara
University in the EECS department, and my wife  is a full-time professor in 
the same  department. We have been actively pursuing a program utilizing
LISP ideas. I am  preparing to teach an experimental computer
science course  this spring. A draft of an article prepared for the Santa
Clara student newspaper is enclosed; though informal, its content is 
 accurate. I will be  teaching  a  more intense version of this material
to a faculty workshop in the Humanities and Sciences schools. From this
experience I will develop an introductory course for entering students.
From the Summer Institute experience, I will develop a
course for computer science students.
I envision these programs resulting in  a demand for several Chromatics-like
machines in the near future, and larger numbers later as cost diminishes
and demand grows.

I hope this letter and the attached material adds perspective to our
 conversations. I believe that Chromatics and TLC have substantial grounds
for cooperation; I hope we can come to an argeement
soon. If you would like further information, please call. I'm  including
several possible phone numbers, since my schedule 
covers a wide area of Santa Clara Valley.
\.
\←L\→S\←R\-L\/'2;\+L\→L

Yours sincerely,


John R. Allen
The LISP Co.
18215 Bayview Dr.
Los Gatos Ca, 95030
(408) 353-2227 --TLC
(408) 984-4358 --Santa Clara University: Dr. Ruth Davis
(408) 984-4482 --Santa Clara University: EECS office
(408) 353-3857 --home

\←S\→L