COMMENT ⊗   VALID 00003 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M1BASL30\M2BASI30\M3BASB30\M4NGR40
C00013 00003	\CBudget for Phase I
C00015 ENDMK
C⊗;
\\M1BASL30;\M2BASI30;\M3BASB30;\M4NGR40;

\F1
\CA Plan to Establish a Credible Curriculum for Mathematics and Computer Science

\CThe Problem

\J
The professional computing field  is suffering from falling productivity; 
technology-based computing  knowledge has a five-year life span; 
software development costs are rising steadily; high school
education in computing is a wasteland. Yet in the next few years, 
low-cost computing devices,
whose power rivals that of contemporary research machines,
will become  available. Projects like
LOGO, Smalltalk, PIE, and general applied Artificial Intelligence 
will be maturing into
consumable products. In appears that neither general society nor the professional
computerist will be ready to assimilate the cultural changes that these
advances will bring. 
 
Furthermore, there is a limited pool of talented people, knowledgeable in the design
and application of sophisticated interactive systems. If we are to fully utilize
the potential of the new machines, we must expand this pool.
\.

\CThe Solution

\J
Begin now to educate high school students about the relationships between
computing, mathematics, science, and society. Teach the fundamental concepts of
algorithms, abstraction, and interactive computing as ways of thinking.
The tools for this project are a combination of traditional classroom techniques
and  hands-on experience with interactive programming  environments.  

This initial phase will concentrate on strengthing the resolve of those individuals
already attracted to the sciences. Many good minds are discouraged
by the superficiality
of the contemporary approach to computing through programming. The rigor and
elegance of formal computation will do much to relieve this group's anxieties.
Conversely, the experience with the computational devices can attract  curious
individuals to the more mathematical aspects of the field. 
As the program grows, we will offer our experiences and tools to those interested
in applying LISP-based ideas to other areas that feel the impact of computing.
Though the material is based on fundamental mathematical and logical principles,
we are confident that the integrated approach will meet with success.
Popularization of reasonably deep results has been done effectively in the  
past --witness
Howard Swann's Calculus Primer and Doug Hofstadter's "Godel, Escher, Bach".

The results of these early steps  are three-fold: first, to begin to educate
people in the fundamentals of computing, hopefully reversing some of the
current trends in the educational systems. Second, to build a team that will 
be prepared to implement a modern LISP on the developing hardware base.
Third, the educational effort will start a larger, better educated group of 
people into the higher educational system.



There are three phases to the plan:

I. Develop the basic materials and  approaches. The fundamental
work has begun; I am now teaching a graduate class at Santa Clara
University, dealing with the theoretical issues. In the  spring term
at Santa Clara University I will offer an undergraduate
version, utilizing an "interactive programming lab" to handle the programming 
aspects. 
Interest has also been expressed to teach such an integrated course at
San Jose State University, San Francisco State University, and 
the Australian National 
University. Since the ultimate target is the high school level, 
we are also planning to teach the material next summer at an institute
for high school mathematics teachers. Furthermore, I will be attending several
college and high school conferences to introduce more people to the ideas.

The course
material is being written mainly by me. The programming laboratory will utilize 
a collection (~10) Z-80-based micro-computers, running a display-based version
of TLC-LISP. The result of this phase will be a first-cut at the integrated
curriculum --text and machines--, and the kernel of the team to begin phase II.
This "shakedown" period is already underway. I have a student working on  the 
display extensions now; and assuming funds are
available, a professor at San Jose State will join me in January.
The staff should include three full-time people.


II. Diffuse the experience and materials. This begins with a campaign to enlist
support  in both colleges and high schools. The curriculum will be enhanced
and refined on the basis of our experience. The laboratory work  should utilize
a large number of microprocessor-based LISP machines. 
These machines can utilize existing
CPUs and existing 
or projected hardware peripherals, including extended memory, color video, and
perhaps mouse, tablet, or voice input/output. The software
will be much enhanced, basing the implementation on an intermediate
machine-independent  formulation. The system
should include an extensive tutor and examples (coming from the
experiences in phase I). Technically the
language should include a compiler and 
probably support an object-oriented data structure like
Smalltalk's classes. This second phase should cover about one year, and
employ about six people.
\.
\;III. Build professional quality LISP-based systems. This goal is the culmination
\;of the first phases. By this time we will have made significant
\;advances in the technical development of portable LISP systems; the tools and 
\;techniques for crafting quality LISP systems will have been tested; the 
\;specification of the language and system will have an strong user community.
\;Of course, this user community will be a good source of talent for the 
\;implementation of Phase III.
\;All of these results are necessary to  the success of this
\;endeavor. The technical advance is clear: portability will be a key to continued
\;success. Further,
\;the development of a high-quality professional LISP system
\;requires a solid implementation team that understands the issues of abstract
\;data structures as well as the issues of low-level implementation. There
\;are very few of these individuals in today's market place and there
\;are too few AI experts to utilize the machines; 
\;we must "grow our 
\;own". 

\;In fact, the global thrust of this entire project is to "grow our own".
\J
Education is an important component in a culture and society. If we fail to 
properly educate our coming generations  in the technology that drives modern
society we will be in for some sorry times.
\.

\CBudget for Phase I

Ten Z-80 microcomputers with 64K RAM		 $30,000 - $40,000
 (price depends on supplier)

Ten display systems from Scion Corp.		 $15,000 - $15,000

Mass storage and network			         $11,000 - $11,000

Facilities for seven months			          $7,000 - $10,500

Salaries for three employees			         $75,000 - $75,000
						                 _______   _______
Total					                        $138,000 - $151,000





\COverview of Budget Requirements for Phase II

\J
It is difficult to estimate the cost of components and systems that will
make up the prototype LISP systems, but  we will definitely 
require a more elaborate
development tool. I favor a mid-sized PDP-11 since the project will  contain about
six people involved in software 
development, and this DEC architecture has a significant
supply of quality software-development  tools available for it. 
\.