COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M1BASL30\M2BASB30\M3NGR25\M4NGR20
C00006 ENDMK
C⊗;
\\M1BASL30;\M2BASB30;\M3NGR25;\M4NGR20;
\F1\CSept 28,1979



Dr. John  N. Buxton
Science Applications, Inc.
1911 North Ft. Myer Dr. 
Suite 1200 (Behind Janitor's Door)
Arlington, VA 22209


Dear John:

\JThank you for your letter. I'm sorry we didn't get to chat, but by
mid-week I was sufficiently depressed about the characteristics
of computer science which were being espoused at that meeting that
I had seriously considered  getting out of the field. I had spend the last
five years trying to convince industry-types that card decks and teletypes
are not appropriate vehicles for communicating with computers; but at
Santa Cruz I 
heard our illustrious leaders preaching to similar industrial types
that interactive techniques are toys which at least rot the mind, if not
cause cancer. 

Perhaps my contacts outside the AI Lab have been atypical, but I have found
that the industrial perception of the importance of  software tools is
retarded to say the least. They see nothing wrong with glass teletypes,
batch systems, compiler-based languages and the resulting "hexoctaldump"
and "print-lineso-and-so" view of debugging. It seemed a great disservice
not to address that problem directly; perhaps that's "Programming Technology"
and not "Programming Methodology". Whatever it is, it involves lessons which
have not yet been learned by a large majority of the programming community.

One of the reasons for my BYTE editorial and the August special on LISP
was a growing suspicion that traditional paths of computer enlightment
were not performing this "technological" role. Outside of a few research
facilities, computing practice is dreadful. It's partly an economic problem,
since some of the "ingredients" have been quite expensive until very
recently; but, alas, much of the difficulty is intellectual:  people don't
realize what \F2can\F1 be done. This is abundantly clear in the design
of systems for microprocessors as well as main-frames; the limiting factor 
isn't the hardware, it's the software. 

It seems to me that what's needed is a "professional school" whose job is
to both educate and train practicing computer scientists; that's not about
to happen.  

Sooo... it seemed to me that the personal computer afficionado was an attractive
alternative: motivated, and equipped with a machine whose characteristics
are closer to the desired goal than
\.
\←L\→S\←R\-L\/'2;\+L\→L

Yours sincerely,



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

\←S\→L