.require "let.pub"source


∂HOM Mr. Al Fenaughty
↓Information International↓5933 Slauson Ave.↓Culver City, CA 90230∞

Dear Al:

	Here are some comments on the microprocessor problem and other
matters that came up at the Board meeting.  They were all made orally
at the meeting, but I put them in writing lest we forget.

	1. We won't influence the microprocessor manufacturer by
volume of sales, but we might get improvements we might want in
software through influencing their user organization by having
good ideas.  It is worthwhile to take part in such organizations
rather than entirely going our own way, because an improvement that
is part of the official software doesn't have to be maintained.

	2. There is no reason to suppose that the 68000 won't be
satisfactory for III's purposes, so there isn't any point in waiting
to see what the Intel 8800 looks like.  There may be some point in
waiting for I-O chips, or waiting for micro-computers on boards, or
waiting for software.

	3. III needs to put effort into finding out what software is
available from the manufacturer amd other vendors promptly and not
wait for specific III projects that might require it.  If we wait
for specific projects we are likely to discover that the project
is in enough hurry so that the only recourse is to continue III's
present practices.

	4. The software vendors include Onyx for the Z8000 and
a company named Whitesmith making C
for the 68000.  Its address is: Whitesmith Ltd.,
127 East 59th street, New York 10022.

	5. It is necessary to see which languages have active
user communities with real time interests.  This is certainly
true of C and may be true of Pascal as well.  According to
Forest Baskett of our department, there are real time versions
of Pascal that might be as good as C, but the standard version
isn't.

	6. My reservations about Pascal stem from its being optimized
for teaching simple programming, i.e. wherein the goal is to
compute an answer from input data and print it rather than to
control a process.  Since C is being optimized for control purposes,
has already been used to write several versions of the Unix operating
system, and has been adopted by Bell for use in electronic
exchanges, I think it has already passed the main tests.  I
don't know the state of Pascal in this regard.

	7. Finally, I think III should try to approach acquiring
a language and related system software in the same spirit
as is appropriate for acquiring power supplies.  Find out what
kinds there are and what the vendors have to offer well in
advance of actual need, and make the make-or-buy decision
with due caution about internal estimates of true costs.

	8. Ed may need to be jogged about holding a meeting
of the committee on new directors.

	9. Topics on which management might give reports at
the beginnings of directors' meetings include (1) estimates
of the size and other characteristics of III's publishing
and other markets, (2) developing a pricing and compensation
policy in an error of high and variable inflation, and (3) ideas
that have been proposed within the company for new products and
services.

	10. An afterthought.  According to my understanding
of (manufacturing) investment practices, a two year payback
time is marginal with regard to a decision to buy a piece
of equipment.  This may be one reason why III encounters so
much sales resistance.

	11. The improvement in directors' compensation has
considerably enhanced my motivation.

	I hope these comments are helpful.

.reg