COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M1BASL30\M2BASB30\M3NGR25\M4NGR20\M5BASI30
C00016 ENDMK
C⊗;
\\M1BASL30;\M2BASB30;\M3NGR25;\M4NGR20;\M5BASI30;





\F1\CJan 7, 1979



Mr. Alexander Burke, President
McGraw-Hill Book Company
1221 Avenue of the Americas
New York, New York 10020



Dear Mr. Burke:


\JAt last, here is the proposal. In part
I am proposing to write a "LISP manual" as suggested
by Peter Nalle in his June 6, 1977 letter and which I responded to positively
 in my June 10, 1977 letter. As that letter of mine stated, I was not happy
about the way my manuscript was being processed in the computer typesetting
experiment; as you may know, things went downhill from there.
As a result of that experience I would make the contract for
the LISP manual contingent on my acquisition of suitable document production 
facilities. Fortunately, I believe, such facilities can be procured.
I can obtain a suitable computer system that I can  enhance with the
necessary text processing tools.
I propose that McGraw-Hill supply me with a manuscript
preparation grant and that I use those funds to purchase the computer system.
Besides using this system 
for the creation of the manuscript, I will assure that the output from my 
system is is a form compatible with that expected by your editing staff
(for example IBM compatible floppies); this way, the copy-editing process
can be  expedited considerably.
If desired, I would also
supply  input meeting the
specifications of a commercial typesetting house. The cost of such a 
system is about $10,000; a budget for the system is enclosed.

John Wakerly, of Stanford's Electrical Engineering Department, told
me about such manuscript preparation grants;
Wiley has entered into a
similar agreement with John for his next book.
This technique seems to be an admirable short-term 
solution for both the publisher and the technically aware author.
Don Knuth  is pursuing a much more long-term (and expensive) solution with
Addison-Wesley; he is developing a "second generation" version of the 
system I used  at the Stanford AI Laboratory, complete with an interface
to a high quality phototypesetter. 
This hints at another "long-term" solution which I 
have heard three  past or potential authors  espouse; namely to 
publish their books themselves.  Though I can sympathize with their frustrations,
I don't feel their course is well-reasoned. As you have known, 
and I have discovered,
there are many aspects of the publishing process which are not understood
by even the most technically inclined author. Both form and content
can be enhanced if the author and publisher work together.
A major point of this proposal
is to develop a better interface between author and publisher; modern
computer technology can smooth that interface.

Of course, as my disgruntled colleagues have noticed, that technology
can also be used to subvert that interface.
Many computer installations currently have
inhouse facilities which would allow such  authors to
prepare their own texts. Publicity is also within their control;
the computer science community  has a 
well-developed "underground mail service" in the ARPAnet. The network
is used already for distributing technical papers long before they
reach traditional publishing media. Initially the network could be used
simply for broadcasting a notice of availability, with the texts still
supplied through the mail. However, using XGP-like devices, it is also
easy to distribute the text electronically with local hard copy reproduction.
Video-synthesis techniques could complete the picture, dispensing with hard copy
altogether. Such a senario is rapidly approaching; currently the pieces are
available within the research community, waiting for cheap, powerful technology
and the insight to put the parts together. The new 16-bit micro computers,
coming on the market this year, finally supply the needed computing power.
It is hard to over-emphasize the impact which these machines will have;
they are not just "more of the same old stuff".  The next few years
will be truly exciting.

If the technical author is able to envision "personal publishing"
now, it will not be long before such  power is commonly available.
An often quoted figure is fifteen years between a laboratory discovery and
its popularization; that is a very conservative figure in the computer
industry. For example, the ARPAnet was developed around 1970; even now,
commercial network facilities are available  with distributed processing
and satellite links for rapid communications; the Viewdata system is bringing
"electronic newspapers" to the British populace, and soon to the
Germans and Americans.

The preceding remarks were not made to be dramatic; but rather to
emphasize the great impact computing devices will have  on the publishing
industry. My proposal is modest  --both for McGraw-Hill and
me--: a demonstration of what can be done now in preparation
for the new technology.

The following is a more detailed description of the benefits of my proposal:

\F21.\F1 First, McGraw-Hill will get a "programming manual" for LISP  which
reinforces the philosophy and techniques stressed in \F5Anatomy of LISP\F1.
 As I said in my June 10, 1977 letter, I don't believe that implementation
dependent (and therefore short-lived) material belongs in a text like
\F5Anatomy\F1. However, an integral part of a course based on \F5Anatomy\F1
should involve programming in LISP, and currently available programming
manuals for LISP are dreadful. What \F5is\F1 needed is a book like
Ken Bowles book for Pascal, \F5Problem Solving Using Pascal\F1. I propose
to write such a book for LISP; 
it will certainly enhance the salability and longevity
of \F5Anatomy\F1. 
The major stumbling blocks to LISP's acceptance in the past  have been the
lack of quality texts and LISP implementations for economical machines.
\F5Anatomy\F1 supplies a "high end" text; technology is giving us inexpensive,
powerful hardware for implementations. This proposed programming book will
supply the missing link. 

I have already begun priming the market. I have written a guest editorial
for the March issue of \F5BYTE\F1 magazine. That article compares LISP with 
Pascal, arguing that the personal computer user's needs are better
served by LISP. Following that, the August \F5BYTE\F1 will be a special 
issue dedicated to LISP. The home/personal computer aficionados
will be a powerful influence in the future.  Many of those
individuals will be among the future computer science students.

\F22.\F1 I will disclose the details of my software system 
to McGraw-Hill personnel, both in writing and visual demonstration.
 I feel my experience in creating, writing, and to some extent  publishing,
gives me some rather unique and valuable insights into the problems
of producing systems which are usable by authors, editors, and compositors.
I want to consolidate and demonstrate those ideas in my system. Since
your personnel suffered with some of the pain, they should also see some
of the results.


\F23.\F1 I am willing to act as a consultant to your personnel, relating
the emerging technology of computers to the publishing applications. 
The details of such consulting arrangements would need to be worked out.
I feel very
exciting things will soon be possible utilizing combinations of such
mundane things as home television sets with the esoteric new micro-computers.
"Possible" means financially, as well as technically.
As I mentioned in my earlier letter to you, the key is to get the author
onto  computer media as soon as possible; cheap ubiquitous technology is that key.
For example, today's $10,000 prototype text system should give rise to 
$1000-$2000 systems  within five years. The hardware is almost here now; the
critical item is the software.

In turn, I  will benefit  by: \F2(1)\F1 producing  a book  which will  help
LISP's popularity, \F2(2)\F1   obtaining a computer  system which I  can
use to continue my work, and finally \F2(3)\F1 getting a chance 
to consolidate my  knowledge
into a system which may help  other authors avoid the pitfalls I  suffered
in the production of \F5Anatomy\F1.
\.

\←L\→S\←R\-L\/'2;\+L\→L

Yours sincerely,



John R. Allen
18215 Bayview Dr.
Los Gatos Ca, 95030

\←S\→L