195602

The Function

of

Automatic

Programming

for

Computers

in Business

Data

Processing

R.

E.

Rossheim

Computers

and

Engineering Education Paul

E.

Stanley

The Pldnning Behind the IBM 702 Installation

at

Chrysler

Corporation

. . . Eugene Lindstrom

Publications

for

Business on

Automatic

Computers:

A Supplemental Listing

Ned

Chapin

Automatic

Mixup

. . .

lawrence

M.

Clark

UNIVAC:

"You're trying

to

divide

by

zero"

A scientist, testing a formula on Univac®

recently, was amazed to see the com-

puting system stop, then automatically

type the reproof: "You're trying to di-

vide by zero." A quick check proved

that Univac, as always, was right.

This graphic demonstration points

out just one

of

many Remington Rand

refinements in the

art

of

computer pro-

gramming and operation.

For

Univac

has been trained to spot human errors.

It

can now carry out, commands given

in simple business English.

It

can even

manufacture automatically its own pro-

gram

of

instructions -at electronic

speeds, with unequalled accuracy.

These skills have been developed

through Univac's unique experience

in

electroriic data-processing. Because,

with every Univac delivered goes 10

years' experience in electronic comput-

ing

...

5 years' experience in the com-

mercial type

of

data-processing. This

wealth of background in programming

and operation

is

unobtainable elsewhere.

The unprecedented savings of Univac

data-processing have been proved by

solving actual customer

problems-not

DIVISION

OF

SPERRY

RAND

CORPORATION

-

2·

-

by working

out

theoretical solutions

with non-existent computers. When you'

install Univac, you're sure to get under

way faster, surer, and more economi-

cally because the System has already

handled similar work.

Univac

is

now at work

in

leading

organizations throughout the country.

And,

in

today's competitive market, the

company which cuts its overhead first

comes out on top. So don't wait until

1957

...

1958

...

or

1959 to cash in on

the tremendous savings available to you

now with the Univac System.

COMPUTERS

AND

CYBERNETICS

ROBOTS

•

AUTOMATION

AUT,OMATIC

CONTROL

Vo

1.

5,

No.

2

February,

1956

ESTABLISHED SEPTEMBER. 1951

ARTICLES

AND

PAPERS

The

Function

of

Automatic

Programmdng

for

Computers

in

Business Data Processing

Computers

and

Engineering Education

The

Planning Behind

the

IBM

702

Installation

at

Chrysler

Publications

for

Business

on

Automatic

Computers: A Supplemental

Listing

REFERENCE

INFORMATION

•••

R.

J.

Rossheim

•••

P.

E.

Stanley

•••

E. Lindstrom

•••

Ned

Chapin

New

Patents

•••

R. R.

Skolnick

Association

for

Computing Machinery Meeting:

Names

and

Addresses

of

Authors

of

Papers

FICTION

Automatic

Mixup

FORU\t

Another

Programmdng

Failure

The

Editor's

Notes

Index

of

Notices

Advertising

Index

•••

Lawrence

M.

Clark

•••

B.

Danch

6

10

13

16

32

34

21

40

4

54

Editor:

Edmund C.

Berkeley

Advisory

Committee:

Samuel

B.

Williams.

Assistant

~ditors:

Eva

ni~tefano,

Jewell

Bown,

Neil

D.

Macdonald,

F.L

Walker.

Herbert

F.

Mitchell.

Jr

••

Justin

Oppenheim

C:ontribu

ting

Edi

tors:

Andrew

D.

Booth,

John

).I.

Breen,

John

W.

Carr.

III.

Alston

S.

Householder,

Fletcher

Pratt

Publisher:

Berkeley

Enterprises.

Inc.

Publication

Office:

513

Avenue

of

the

Americas,

New

York

21,

N.Y. -

Algonquin

5-7177

Research

and

Library:

36

West

11

Street

••

New

York,

11.

N.Y. -

Gramercy

7-1157

Branch

Office:

R 15

Washington

Street.,

Newtonville

69,

Mass.

-

Decatur

2-5453

or

2-3928

Advertising

Representatives:

San

Francisco

-

W.

A.

Babcock,

605

Market

St.,

San

Francisco

5,

Calif.

Los

Angeles

-

Wentworth

F.

Green.

439

So.

Western

Ave.,

Los

Angeles

5,

Calif.

elsewhere

-

the

Publisher

o)MPlJfERS

AND

AUTOMATION

is

pub

1

ished

monthl

y.

Copyright.

1955

by

Berkel

ey

Enterp

ri

ses.

Inc.

Subscrip

tion

rates:

$5.50

for

one

year,

$10.50

for

two

years,

in

the

United-States;

$6.00

for

one

year,

$11.50

for

two

years,

in

Cana"da;

$6.50

for

one

year,

$12.

~

for

two

years

elsewhere.

Bulk

subscription

rates:

see

page

46.

Advertis-

ing

ra

tes:

see

page

44-.

Entere(i

as

second

class

matter

at

the

Post

Office.

New

York.

N.

Y.

- 3 -

THE EDITOR'S NOTES

THE

C~1PUTER

Directory

Corrections

Manuscripts

Reader's

Inquiry

Form

Special

Issues

See Page:

46

44

42

38

37

4

36

46

38

Address Changes:

If

your

address

changes,

please

notify

us

giving

both

old

and

new

addresses,

and

allOtoJ

three

loJeeks

for the change.

NOW!

dependable

relays

for

printed

circuits

Maybe

you, too,

have

been

awaiting

availability

of

a

good

relay

for

direct

insertion

into

printed

circuits.

Now

Automatic

Electric

can

solve

your

problem

with

a minia-

ture

relay

that

is

just

right.

120 million operations,

without

a single

readjustment

or

relubrication!

That's

what

you

get

from.

this

rugged,

improved

Series

SQD

Relay,

because

it

features a special

heavy-duty bearing

and

bearing pin. Also a recess

in

the

bearing

plate

retains

an

adequate

supply

of

lubricant

for

long-term

lubrication

of

the

bearing

pin.'

Consider these

additional

advantages:

1.

The

sections

of

the

terminals

that

insert

into

the

printed

circuit

board

are

NOT

brazed

or

welded

into

place,

but

are

integral

parts

of

the

coil

ter-

minals

and

contact

springs-thus

preventing

ID-

ternal

loss

in

conductivity

or

continuity.

2.

Terminal

design

permits

direct

plug-in

of

the

relay

into

a

printed

circuit

board,

ready

to

be

secured-in

place

with

any

acceptable soldering technique.

Usually

the

desired

contact

spring

combination,

or

pile-up, is sufficiently large so

that

additional

mounting

(support)

of

the

relay

is

not

necessary.

-5-

SQD Miniature Printed Circuit Relays

are

available

with

many

different

contact

spring

arrangements,

and

for a

multitude

of

applications. Springs

can

be

made

of

phosphor-bronze, ((Bronco"

metal,

or

other

special-

purpose

materials,

as

required.

Of

course

the

long life,

heavy-duty

features

of

the

improved

SQD

Relay

can

be

had

in

the

conventional

type

of

plug-in relay,

if

regular

sockets

are

preferred for

use,

whether

in

printed

circuitry

or

other

applications.

To get complete details, write:

Automatic

Electric Sales

Corporation, 1033

West

Van

Buren

St.,

Chicago

,7,

Illinois.

In

Canada:

Automatic

Electric

(Canada)

Lt~.,

Toronto.

Offices

in

principal cities.

RELAYS

SWITCHES

'""""

"

',"'"'''

"'"""'" , "

AIJTDMATI[

ELE[TRI[

CHICAGO

The

Function

of

Automatic

Programming'

for

Computers

in

Business

Data

Processing

R.

J.

ROSSHEIM

Supervisor,

Industrial

Pro~rammin~

Research

Remi~ton

Rand,

Division

of

Sperry

Rand

r~rp.,

~iladelphia

3, Pa.,

(Presented

before

the

Association

for

Colll>uting

Machinery,

Philadelphia,

Pennsylvania,

Sept.

14

',0

16, NSS)

This

paper

is

concerned

td

th

the

role

that

s

truc

tions

to·

describe

completely

the'

payroll

automatic

programming

systems

for

large,

gen-

accounting

function.

eral-purpose

computers

might

play

in

the

appli-

cation

of

these

equipments

to

business

d a t

a-

processing

problems.

My

definition

of

automa-

tic

programming

for

the

purpose

of

this

discus-

sion

is

extremely

broad,

including

any

program-

ming

system

which

is

capable

of

substantially

reducing

the

time

and

effort

now

required

to

program

the

machines.

The

definition

of

auto-

matic

programming

certainly

includes

the

co

n-

cept

of

relegation

of

purely

clerical

portions

of

machine

coding

to

the

computer

itself,

but

it

is

not

limited

to

this

concept.

It

is

hoped

that

this

p~per

will

stimula~

interest

and

encourage

effort

in

developing

rrore

advanced

automatic

programming

systems

for

busi-

ness

use.

Training

I

shall

s

tart

with

the

training

ph

a s e

tVhich

normally

precedes

the

computer

installa-

tion.

I

shall

make

the

assumption,

t.ghich

is

now

virtually

a

cl

iche,

that

the

company embarlt-

ing

on

the

computer

venture

has

decided

to

train

its

own

people

in

the

use

of

the

equipment,

ra-

ther

than

to

educate

trained

programmers

from

outside

the

company

in

the

endless

detail

of

its

data-processing

procedures.

The

immediate

problem

is

to

train

people

in

the

use

of

a ma-

chine

which

far

transcends

in

terms

of

f

Ie

x

i-

bility,

power,

and

speed

anything

to

which

t~

might

be

accustomed.

Frequently,

we

req«ire

that

the

people

selected

have

superior

ability

in

overall

system

planning

since

we

expect

the

equipment

to

have

a

sizable

impact

on

our

pro-

cedures,

and

yet

the

firs

t

thing

these

p

eo

pIe

must

do

is

submerge

themselves

in

the

swirling

mul

ti

tudes

of

electronic

binaries

which m u s t

eventually

be

arranged,

bit

by

bit,

to

control

the

execution

of

the

payroll

accounting

func~

or

the

like.

The

basic

problem

is

that

the

equipment

when

delivered

has

buil

t

into

it

only

one

basic

ability,

that

being

the

ability

to

interpret

a

certain

language

which

is

the

i

n-

s

truc

tion

code

of

the

machine.

This

code

c e

r-

tainly

is

logically

complete

and

can

be

adapt-

ed

to

meet

every

conceivable

requirement

of

the

user.

At

the

same

time,

it

is

elementary,

or

stated

another

way,

it

is

a

language

of

mono-

syllables.

An

indicator

of

the

size

of

these

elements

is

the

revelation

in

more

than

ore com-

puter

installation

that

it

requires

many

tens

of

thousands

of

these

elementary

machine

i

n-

- 6 -

The

training

problem

would

be

considerably

reduced

if

the

equipment

were

able

to

interpre

t

the

everyday

language

of

the

procedure

analyst

and

to

translate

the

meaning

so

conveyed

into

its

own

binary

arrangements.

This

is,

of

course,

an

ideal.

It

is

fundamentally

dependent

on

the

existence

of

such

a

language,

assuming

also

precise

meanings

and

universality.

Theconcept

of

such

a

language

is

not

so

new, nor so

ideal,

that

there

are

not

even

now

some

res

ear

c h

groups

t1l'0rking

in

this

area.

I

bel

ieve

the

objective

can

be

stated

as

raising

the

level

of

machine

comprehension

to

the

language

of

the

user,

rather

than

forcing

him

to

adop

t mmo-

syllabic

machine

talk.

The

abili

ty

to

program

would

b

ecom

e a

relatively

simple

operational

tool

similar

to

data

flow

charting,

or

procedure

manual

writing.

It

would

permi

t

the

procedure

ana

lys

t

or

his

counterpart

to

spend

more

of

his

time

taking

the

broad

view

of

data

flow

through

a

system,

and

to

concentrate

on

the

objective

of

good

system

design.

I

think

that

very

few

of

the

people

se-

lected

to

work

in

this

field

will

be

i

nte

r-

ested

in

becoming

professional

programmers;

so

the

requirement

is

to

make

the

equipmen

t

available

to

such

people

on a

basis

which

is

cons

is

tent

with

the

other

tools

0 f

the

i r

trade.

I

am

suggesting

that

the

obj

e c t i v e

of

automatic

programming

techniques

duri

n g

the

initial

training

stages

might

be

the

develop-

ment

of

a

relatively

simple

pseudo-code

\\fhich

would

employ

to

tihatever

extent

now

possible,

the

language

of

the

user.

The

computer

would

be

enabled

to

translate

programs

wri

t

ten

in

this

language

into

the

elementary

mac

h i n e

codes.

The irnroodiate

effec

t

should

be

t

hat

trainees

could

learn

in

much

less

time

towri

te

usable

programs.

At

the

same

time

the

t

end-

ency

would

be

away

from

creating

highly-trained

specialis

ts

and

toward

the

development

of

part-

time

or

temporary

programmers

from

those

who

are

experts

in

the

jobs

wh

ich

are

to

be

pre-

pared

for

the

machines.

I

should

note

that,

although

there

is

not

to

my

knowledge

a

pseudo-code

system

for

busi-

ness

applications,

an

experiment

involving

the

same

principle

tvas

successfully

carried

out by

FUNCTION

OF

AUTOMATIC

P~OG~AMMING

the

Whirlwind

computer

group

at

M.LT.

two

years

ago.

The

so-called

Summer

Session

Com-

o

puter

t'las a programmed

modification

of

Whirl-

wind t'lhich permi

tted

the

principles

0 f P r

0-

gramming

to

be

taught

to

a

student

g r 0

up

of

varied

interes

ts

and

backgrounds,

based

0 n a

pseudo-code

which

was

easily

learned

and

ap-

plied.

Approximation

and

Experimentation

Next, I

will

consider

the

role

of

au

t

0-

matic

programming

in

the

time

consuming

and

costly

preparation

of

data-processing

proced-

ures

for

a

computer

system.

Before

discussing

automatic

programming

per

se

in

this

reg

ard,

I

will

talk

about

programming

procedures

fo

r

commercial livork.

First,

the

concept

of

approximation.

Th~

approach

is

suggested

by

the

tremendous

amount

of

detailed

information

t'lhich must be

gathered,

analyzed,

organized,

and

finally

progr

a

mmed

into

a

computer

data-processing

procedure.

Information

may

come

from

procedure

man u

als,

rules,

and

regulations

which

are

sometimes

subj

ec t

to

interpretation,

or

from

infor

med,

but

sometimes

inarticulate

clerical

personnel.

As

a

result

of

the

nature

of

the

informatio

n

and

its

sources,

a

fundamental

requirement

of

programming

cannot

be met;

that

is,

the

"prob-

lem",

or

job,

cannot

be

completely

predefined.

Therefore,

programming must

often

begin

tV

it

h

incomplete

or

incorrect

information,

and

the

firs

t

program

is

necessarily

merely

a

fir

s t

approximation.

Frequently,

voices

which

were

silent

during

the

survey

and

fact-finding

vis-

its,

are

quick

to

provide

the

necessary

crit-

icism

tvhen

the

first

approximation

misses

the

mark. Then

the

hard

reali

ties

begin

to

be

re-

vealed,

the

neat

flow

charts

spread

out,

d

e-

veloping

all

manner

of

intricate

networks,

and

the

programs s

tart

to

overflow

a

va

i 1 a b I e

storage.

If

I

have

conveyed

the

idea

at

all,

it

should

not

be

necess

ary

to

say

t

hat

the

second

approximation

is

not

final,

nor

the

third;

but

finally

an

acceptable,

if

not

per-

fect,

version

is

developed

which

comprehends

everything

of

maj

or

importance

as

well

as

most

minor

details.

The

fact

is,

although

the

ap-

proximation

technique

described

above

may

not

be

consciously

employed,

it

represents

a good

description

of

what

actually

happens.

Another

approach,

which

is

the

other

face

of

the

same

coin,

involves

the

idea

of

exper-

imentation.

Whereas

approximation

met

hod

s

convey

the

idea

of

working

gradually

toward

a

single

correct

and

best

procedure,

the

exper-

imentation

viewpoint

acknmdedges

the

f

act

that

our

analytical

tools

for

designing

elec-

tronic

data-processing

systems

are

extreme

1 y

primi

tive,

so

that

our

current

p

rogres

sis

largely

the

result

of

trial

and

error

meth~.

- 7 -

Curren

tly,

it

is

too

cos

tly

to

try

a number of

different

approaches

to

the

same

application,

if

this

involves

programming a

second

tim

e.

Moreover,

the

systems

we

design

are

likely

to

be

very

conservative

in

order

to

red

u c e

the

chance

of

having

to

make

maj

or

changes.

Thus,

tve

miss

many

opportuni

ties

to

learn

more about

how

to

use

this

new

equipment

as

the

pressure

of

schedules

forces

us

to

settle

upon

proced-

ures

which

may

be

far

from

optimum.

If

these

ideas

are

reasonable,

involving

as

they

do

the

use

of

approximation

tech~es

and

experimentation

in

the

design

of

commercial

data-processing

programs

for

computers,

auto-

matic

programming

can

play

several

import

ant

roles.

In

fact,

it

is

apparent

that

any

sys-

tematic

approach

to

programming which w 0 u 1 d

assist

in

the

process

of

going

from

on

e

ap-

proximation

to

the

recoding

would

reduce

the

total

amount

of

coding

considerably.

Flow

Charting

I

shall

now

discuss

the

various

ph a s e s

through

which an

application

or

job

must

pass

before

a

working

program

is

ready

for

the

com-

puter.

It

is

significant,

I

think,

that

prob-

lems which

have

arisen

in

all

of

the

are

a s

could

be

solved,

either

wholly

or

in

part,

by

one

or

more

of

the

techniques

tvhich

falltdthin

my

definition

of

automatic

programming.

Firs

t,

let

us

consider

flow

c h

art

in

g •

Apparently

the

flow

chart,

in

a

mul

ti

tude

0 f

forms,

is

the

most

easily

adapted

and

univer-

sally

accepted

method

of

describing

d a t

a-

processing.

The need

is

for

some

way

0 f

ar-

ranging

great

masses

of

detail

in

such

a way

that

we

can

see

both

forest

and

trees

at

the

same

time.

Typically

we

proceed

fro

m

the

broadest

level

of

flow

charting

tvWeh

outlines

the

entire

application,

dOlivu

through

sever

al

successive

levels,

each

adding

detail,

unti

1

in

some

cases,

we

dratv

charts

which

virtua

lly

indicate

the

machine

instructions

required

to

perform

each

block

on

the

chart.

Such

charts,

I

have

heard

it

said,

could

be handed

to

people

tvho

have

been

trained

in

the

computer

instruc-

tion

code,

and

they

could

translate

the

de-

tailed

flow

chart

into

a machine

pro

g

ram

wi

thout

us

ing

any

discretion

or

j u d g

men

t.

They tvould

perform

a

clerical

function.

But

why

not

let

the

computer

itself

perform

the

clerical

function

of

translating

detailed

flm"

charts

into

sequences

of

machine

instructions?

I

believe

that

the

translation

of

flow

charts

into

machine

codes

should

be a

primary

objec-

tive

of

an

automatic

programming

system

for

business

data-processing.

Some

of

the

problems

in

trying

to

do

this,

are

very

challenging.

First,

we

need

a

"flow-

chart-to-digi

tal-computer"

converter

so

t

hat

FUNCTION

OF

AUTOMATIC

PtlOGRAMMING

we

can

feed

the

flow

chart

into

the

computer.

tines

can

be

written

and

pseudo-codes

invented

Second, any auto-programming

sys

tern I can

con-

to

call

up

these

routines.

However,

now

andfor

ceive

of

must work from a

very

s

pee

i

fie

some

time

to

come, any

automatic

..coding system

"language"

that

must have

all

the

precision

of

must

accept

conventional

machine

coding

as

well

the

machine

code,

while

at

the

s

arne

time adapt-

as

pseudo-coding.

ing

itself

to

the

requirements

of

flow

charts,

of

programmers, and

of

the

applications.

I will

not

stop

to

enumerate

other

problems

fu

accom-

plishing

what I have

suggested,

but

perhaps

I

should

make

one

point

clear.

It

is

certainly

true

that

many

programmers

leave

the

flow

chart

behind

in

the

pressure

to

write

machine

codes,

but

remember

that

what I

am

suggesting

should

replace

the

wri

ting

of

machine

codes.

I a

c-

knowledge,

too,

the

fact

that

often

when

th

e

programmer and

clerical

coder

are

the

sam

e

person,

as

is

more

often

the

case

than

not,

the

detailed

flow

chart

is

omitted

entirely.

How-

ever,

I would expec t

that

as

we

learn

m 0 r e

about

the

elements

of

data-processing,

we

will

be

able

to

build

auto-programming s y s

tern

s

capable

of

interpreting

higher

1

eve

1

flo

w

charts,

Ivherein each

block

represents

a

func-

tional

subroutine

containing

many

machine

in-

structions,

rather

than

just

one

or

two.

My

feeling

is

that

the

first

step

in

this

direction

is

the

really

difficult

one.

Coding

In

the

conventional

sequence,

after

flow

charting

comes

coding,

and

since

flow

chartmg

is

not

going

to

replace

coding

tomorrow,

there

is

considerable

l"lork

for

auto-programming

in

the

coding

phase.

It

is

in

this

area

w

her

e

currently

some

encouraging

progress

is

be

in

g

made. I

shall

merely

mention

some

of

the

con-

tributions

that

auto-programming

systems

can

make

to

business

data-processing

applications.

First,

automatic

programming

systems

can

assist

in

standardizing

the

programs by

spec-

ifying

the

form

of

the

coding,

and

by

supply-

ing

a

library

of

subroutines

which

are

copied

into

programs Ivherever

they

app

ly

•

Sec

0 n d,

as more

is

learned

about

recurrent

subroutines,

libraries

of

such

routines

can

be

develo

p

ed,

and programs

can

be

written

in

a

shorthan

d

pseudo-code,

thereby

avoiding

duplication

0 f

coding

which has been

Ivritten

before,

andcon-

sequently

reducing

the

amount

of

netv

coding in

each

nel"1

program.

Third,

the

tedious

boo

k-

keeping

jobs

associated

with

the

preparation

of

finished

coding

can

be

wholly

0 r

par

t 1 y

relegated

to

the

computer,

thereby

releas

i n g

the

programmer

to

concentrate

on

the

de

s i g n

of

a

complete

and

efficient

computer

procedure,

rather

than

burdening

him

with

coding

detai~.

The

fact

that

not

enough

is

notv

k n 0

tV

n

about

the

technical

language

of

data-process-

ing

should

not

discourage

the

development 0 f

initial

automatic

coding

systems.

As

certain

processing

functions

are

defined,

standard

rou-

- 6 -

Another improvement

in

programming p r

0-

cedures

which

can

be

designed

into

an aummatic

sys

tern

is

the

abili

ty

to

have

several

p e

opl

e

lyork on

different

parts

of

the

same

program

without

the

usual

problems

of

communication.

Program Checking

or

Debugging

After

the

program has

been

prepared,

it

must be checked

or

debugged, which

frequently

requires

many

additional

man-hours

and

~hine

hours.

Much

of

the

abusive

language

concern-

ing

the

cost

of

application

of

computer

sis

directed

at

this

stage

of

the

operati

0

n.

Basically,

two

kinds

of

prograrmning

err

0 r s

account

for

the

effort

which

is

expended a t

this

point.

First,

there

are

the

simpl

e-

minded

errors

resulting

from

poor

or

inc

0

m-

plete

record-keeping

during

the

pro

c e s s of

coding,

or

from

slight

misunderstandings

of

the

way

in

which

the

equipment

operates,

0 r

th

e

typographical

kind

of

error.

Second,

there

are

the

logical

errors

which

may

resul

t

fro

m

incorrect

translation

of

the

flow

chart,

0 r

failure

of

the

programmer

to

consider

certain

cases

or

combinations

of

cases.

It

is

v

er

y

easy

to

sympathize

with

the

programmer

for

he

must have

the

ability

to

work

in

minute

de~il

and,

at

the

same

time,

keep

in

mind

the

co

m-

plex

interrelationships

wi

thin

the

e n t

ire

framework

of

the

program.

Putting

s

ym

path

y

aside,

h01vever,

there

is

a

pressing

need

to

improve methods

of

prograrfiming

in

or

d e r

to

avoid

some

of

the

errors,

and

at

the

same time

to

provide

for

efficient

methods

of

err

0

r-

correction.

This means

that

the

aut

om

a

ti

c

programming

system

does

not

stop

functio

nin

g

wi

th

the

'production

of

the

first

"approximate"

program.

Systematic

error

correction

tecmdques

should

be

provided

to

assist

correction

of

all

kinds

and

levels

of

errors.

Some

study

should

be

made

of

debugging

procedures

con

sid

e

ring

typical

business

installations,

the

operating

characteristics

of

the

equipment,

and

the

re-

quirements

of

the

programmer.

The

availability

of

an

automatic

program-

ming

system

may

avoid

many

of

the

pit

fall

s

which

arise

in

the

debugging

stage.

For

in-

stance,

if

the

approximation

idea~

utilized,

it

is

conceivable

that

a programmer

would

pre-

pare

a

routine

concerned

only

with

the

m a i n

flow

of

data,

knowing

at

the

time

that

t

her

e

l'Vere

many

loose

ends.

By

debugging

the

p a

r-

tial

routine

he has

the

advantage

of

working

with

a

smaller,

less

complex program

initially.

Once

this

is

checked

out,

it

should

be

re

1

a-

tively

simple

to

control

and

isolate

errors

FUNCTION

DF

AllTOMATIC

PROGRA~IMING

as

additional

routines

are

added

to

make

the

program

to

fit

into

the

existing

organiza-

procedure

complete.

Here

too,

automatic

pro-

tion,

without

the

complete

recoding

which

con-

gramming

techniques

l'1ould

substantially

reduce

ventional

methods would

require.

the

time

and

effort,

since

it

should

be

possible

to

blend

in

new

subroutines

and

correct

exist-

ing

routines,

taking

bes

t

advantage

of

the

parts

of

the

program which

are

complete,

cor

r e c

t,

and

tested.

With

respect

to

the

testing

phase,

the

most

important

contribution

an auto-programmmg

system

could

make

is

the

reduction

in

overa

1 1

elapsed

time

required

for

the

first

approxima-

tion.

It

is

vitally

important

to

reach

at

the

earlies

t

possible

date

the

firs

t

check

poi

n t

l..nere

the

critical

look

is

taken

and

the

in-

formation

feedback

is

activated.

No

degree

of

accuracy

and

no

amount

of

care

can

replace

fue

first

test

in

turning

up

weaknesses

in

the

pro-

gram

or

procedure.

Testing

Once

the

program

has

been

debugged

to

the

point

where

it

is

operable,

it

must be

checked

against

the

other

contingent

parts

of

the

sys-

tem.

Actual

operating

conditions

and d a t a

must

at

least

be

simulated.

How

heartbreaking

are

the

programming

failures

at

this

poi

n t !

~Yet,

this

often

is

the

firs

t

real

checkpo

i n t

in

the

sense

that

some ac

tual

data

are

pro-

cessed'

and

the

programmed

procedure

is

m 0 r e

easily

checked

to

see

whether

it

fulfills

all

the

requirements

between

input

and 0 u t

put.

This

is

the

first

approximation.

Since

it

is

very

inconvenient

to

check

coding

itself,

itis

necessary

to

process

input

and

to

produce

out-

put,

both

of

which must

be

referred

tot

he

people

in

the

business

who

either

origi

na

t e

source

documents

or

use

the

reports.

The

s e

are

the

important

critics

of

the

procedures.

They

are

the

inspectors

at

the

end

of

the

pro-

duction

line.

Since

the

procedure

was

origin-

ally

set

down

and

the

initial

flow

charts

were

drawn

there

has

been

considerable

dead-reckon-

ing.

Using

conventional

methods

of

programmbg

and

coding,

there

has

also

been

considera

b 1 e

time.

Now

the

feedback

of

information

begins,

and

the

procedure

begins

to

be

refined.

The

second,

third,

and

successive

approximatio

n s

are

written,

each

requiring

repetition

of

some

or

all

of

the

preceding

steps.

Corrections,

Modifications,

Revisions

After

the

results

of

the

first

approxima-

tion

have

been

checked

a

steady

flow

of

modi-

fications,

corrections,

and

additional

routines

must be merged

in

with

the

original

program.

All

too

often

the

delicate

balance

is

upset

by

these

changes,

and

succeeding

debugging

is

even

more

difficult

than

the

original.

Hopef

u

lly,

an

automatic

system

would

permit

new

parts

0 f

- 9 -

the

foregoing

is

the

ide

a 0 f

revision

of

the

program,

at

some

later

d

ate,

to

provide

for

changes

in

the

procedures.

I t

is

a

mistake,

I

think,

to

regard

business

ap-

plications

as

static.

Just

as

business

itself

is

dynamic,

the

data-processing

requireme

n t s

are

constantly

changing,

and

it

would

see

m

reasonable

to

rate

a

data-processing

system

in

terms

of

its

ability

to

respond

to

c

han

g e d

conditions.

Automatic

coding

could

be

the

key

to

this

ability.

This

migh t be an

appropriate

poi

n t

to

mention

another

problem.

When

a programmer

is

in

the

midst

of

a

particular

routine

he

can

follow

the

coding

h~

has

written

quite

easily,

but

let

him

move

on

to

another

coding

as

s

ig

n-

ment, and

after

a week he

will

have

I a r

gel

y

forgotten

the

first

program.

It

is

almost

a s

difficul

t

for

him

to

trace

through

the

cod

in

g

as

for

someone

who

has

not

seen

i~

before.

Any

programming

system

should

provide

a method

for

carrying

along

verbal

descriptions

of

the

func-

tions

which

the

different

parts

of

a

pr

ogr

am

perform.

If

this

could

be

tied

in

to

the

flow

chart,

subroutine

by

subroutine,

together

they

lfiQuld

provide

adequate

documentation

0 f

the

program.

The

verbal

description

would

reve

a

1,

to

non-programmers

exactly

what

the

pr

0

gr

a m

does,

and

it

could

be

used

for

checking

p u

r-

poses.

Moreover,

it

would

greatly

facilitate

revision

of

the

program

weeks

or

months

later.

Before

closing,

it

might

be

well

to

men-

tion

one

other

area

which

may

soon

draw

the

attention

of

management,

if

it

has

not

already

done

so.

Much

is

currently

being

written

and

said

about

the

ability

of

electronic

data-pro-

cessors

to

provide

information

to

assist

m

a'

n-

agement

decisions.

One

form

this

might

tak

e

is

the

processing

of

files

and

records

already

stored

in

a

high-speed

medium.

There

are

two

important

aspects

of

this

type

of

job.

First,

time

is

of

the

essence;

and I

am

speaking

0 f

the

time

from

the

original

request

to

the

time

the

results

are

turned

over

to

manage

me

n t

Second,

it

cannot

be assumed

that

these

requests

could

be

anticipated

in

the

sense

that

programs

could

be

prepared

ahead

of

the

request.

At

the

same

time,

programming

costs

cannot

be

exorbi-

tant

as

such

requests

may

be

for

one-shot

jobs.

Clearly,

this

is

a

challenge

which

could

bernet

by an

adequate

automatic

programming

system.

Conclusion

In

the

short

space

of

this

paper

I h a v e

tried

to

show

why

the

techniques

of

automatic

programming

are

important

throughout

business

data-processing

application

of

large-s

c a

Ie,

(cmtinued

on

page

32)

COMPUT!ER~S,

AND

ENGINEERING

EDUCATION

PAUL

E.

STANLEY

Associate

Professor

of

Aeronautical

Engineering,

Purdue

tniversity

Lafayette,

Indiana

Anyone

who

has attempted to

make

a

ch~

in

an

engineering

curriculum

is

n'ell

at'ltare

of

the

multitude

of

problems.

So

many

new

ideas,

.

-inethods-,

-and

machines have

come

to

the

fore,

each clamoring

for

a

place

in

the

curriculum,

that

one

is

tempted to

eliminate

them

all

and

return

to

the

teaching

of

only "bas

ic

physical

principles"

and

"fundamental

engineering

de-

sign",

t'lthatever

that

may

mean.

HOt1Tever,

this

ostrich

in

the

sand

attitude

is

not conducive

to

progress;

one must

somehow

eliminate

some

of

the

traditional

subject

matter

and

intro-

duce

the

new.

(The

readrr

is

referred

to

the

"Saber Tooth Curriculum"

for

some

other

thoujrts

along

this

line.)

It

is

not

the

purpose

of

this

paper

to

say what

must

be

Id

thdral'1tn

from

the

curricu-

1um

but

rather

to

suggest

a

nen'

subject.

An

attempt

will

be

made

to

shOt'lt

how

the

net"

idea

can be used

in

conjunction

tdth

the

tradition-

al.

The

large

scale

advent

of

automatic

com-

puting

machinery

into

the

field

of

engin

eering

research

and

development

11.

r

arrants

careful

ex-

amination

by

engineering

educators.

In

the

aeronautical

industry,

and

to

a

great

extent

~n

other

areas,

both analog

and

digi

tal

com-

puters

often

confront

the

young

engineer

with-

in

the

first

few months

of

his

new

career.

It

seems

reasonable,

therefore,

for

him

at

least

to

have been

introduced

to

this

important unl

during

his

undergraduate days.

The

research

and

development groups i n

industry

use both

digital

and

analog

computers.

Perhaps

it

Ivould

be

good

to

try

to

inc

1 u d e

something

of

each

in

engineering

instruction.

This

naturally

calls

for

both types

of

moch~

to

be

available

for

use

by

the

undergraduate.

Often

this

is

not

feasible

for,

even

if

com-

puters

of

advanced types

are

located

on

the

campus, they

are

employed

exclusively

in grad-

uate

instruction

and

research.

This

is

espec-

ially

true

of

digital

computers;

but,

as

lvill

be

shOt'ltn,

analog equipment

suitabl.e

for

in-

struction

can be

built

(or

bought)

for

a very

reasonable

sum.

In

addition

to

providing

some

familiari-

zation

Id

th computing machinery, the use 0 f

the

analog computer provides

an

excellent

~

for

emphasizing the

similarity

0 f s y s t

ems,

whether they

are

electrical,

mechanical, 0 r

thermal,

etc.

The

integration

of

many

funda-

mental

principles

thus can be achieved

ina

fashion

long sought

by

engineering

instructors.

The

same

can be

said

for

many

of

the mathemat-

ical

principles

taught

by

the

mathematics

de-

partment

but

so

frequently

neglected

in

the

teaching

of

other

courses.

Every

engineering

student

knows

well

(we

hope)

the

relation

F =

rna,

force

equals

mas

s

times

acceleration.

If

he has had a course in

vibrations,

he

may

be aware

that

,

~

Ma..:::MU

cit.%.

and

be

able

to

Ivri

te,

tv! th

some

understanding,

the

differential

equation

NJ.2Ix

1-

fdx

+

Kx

~

F(t)

cit"

ott

But fet"

students,

if

any,

are

aware 0 f

the

similari

ty

betlveen

this

and

the

ele

c

tr

i c a I

equation

'

L

tL~

+

RL

T

..LSLd-t

= E

(t)

J,t

C

The

less

familiar

form

of

the

latter

equation

is

obviously

L ol,:tct, -+ R

~

+ L

q.::.

E.

(t)

J,tl.

cit

c'

which

is

identical

mathematically to

the

mech-

anical

equation.

Such

lack

of. understanding,

of

which this

is

but one example,

definitely

restricts

the

creative

work

tvhich an

engineer

might

oti1elWise

be able to do.

The

solution

of

a few

prob~,

similar

to

those

above,

on

the

analog

compurer

serves

well to emphasize the. mathematical simi-

larity

and

brings

to

light

again

the

useful-

ness

of

simple

calculus

and

differential

equa-

tion

principles.

A

not-so-incidental

product

of

instructiDn

in

the use

of

the

analog computer

is

the

use

of

the

notation

and

principles

of

operational

calculus.

In

more

advanced

courses,

the

con-

cept

of

the

transfer

function

can be intxoduced

and

the

analysis

(and

synthesis)

of

s ys tern s

such

as

simple servomechanisms can be

carried

out.

A system can then be simulated 0 n

the

-10 -

COMPUTERS

AND

ENGINEERING

EDUCATION

computer

behavior

be

given

and an

ex

pe2

imental

anaiysis

of

its

.

sort

described

above and

in

sufficient

amount

determined.

Examples

of

this

will

to

make

it

possible

to

work wi

th

second

order

later.

differential

equations

can

be

kept

under

two

The

implication

of

the

above

is

this:

by

use

of

the

basic

elements

of

the

analog

compu-

ter,

a model

of

a

system,

any

sys

tem uooer

stu-

dy,

can

be

constructed

and

analyzed.

Since

tre

sys

terns s

tud~ed

i.n

undergraduate

classes

are

usually

simple~

the

number

of

analo~

elements

is

small,

consisting

of

a few

DC

amp

lifiers

and power

supplies;

some

capacitors

for

inte-

gration;

resistors

for

in

put

and summing

~es;

a few

potentiometers

and a

recording

galvano-

meter

for

the

output.

By

proper

selection

of

these

elements,

and

flexible

arrangements

0 f

them,

valuable

ins

truc

tion

in

many

areas

can

be

given.

The

stability

of

the

equipment

and the

(£-

curacy

of

the

results

are,

to

a

great

extent,

dependent

upon

the

quality

of

the

components

used.

Leakage

resistance

in

the

integrating

capaci

tors,

or

from

various

components to grourrl

must be

kept

as

high

as

possible,

highrom~d

to

the

input

and

feedback

resistors

which

are

us

ually

of

the

order

of

.1

to

1

megohm.

fbv-

ever,

a

little

care

in

construction

of

the

equipment

and some

selec

tion

of

parts

fro

m

commercial

grade

stock

will

provide

us

able

equipment

at

minimum

cost.

The

cost

of

construction

of

manually

bal-

anced

amplifiers

can

be

kept

as low as

ten

to

fifteen

dollars

each,

including

student

labor.

Commercial

units

may

be

bought

for

a r 0 u n d

tl~enty

dollars

plus

a

small

amount

for

the

con-

s

truc

tion

of

racks

into

which

they

can

be

plugged.

Stabilized

power

supplies

to

provide

positive

and

negative

DC

voltage

of

the

order

of

300

volts

can

be

built

for

about

twenty

to

thirty

dollars

each.

While

it

is

desirable

to

use

potentiome-

ters

"hich

are

quite

linear,

and which

can

be

adjusted

t,o

considerable

accuracy

(ten

turn

helical

po'tentiometers),

good

results

can

be

obtained\wi

th

uni

ts

cos

ting

a

third

as m u c h

if

a

large

dial

is

used.

Mul

tiple

posi

tion,

ganged

slvi

tches

mke

it

possible

to

control

several

circuits

wi t h a

fair

degree

of

precision

without

the

high

cost

of

relays

and

requisi

te

power

supply.

The

swi

tch

contacts,

along

with

the

inputs

and

outputs

of

the

amplifiers,

the

capacitor

ter-

minals,

and

so

on

can

be

connec

ted

to

banana

jacks

on a

patch

panel

of

laminated

p I as

tic

which

can

be

engraved

to

show

the

c

irc

ui

t

com-

ponents

involved.

Spacing

these

jacks

three-

fourths

inches

increases

the

versatility

mllie

system

since

this

is

the

spacing

of

the

Gemr-

al

Radio

double

plug.

The

overall

cost

of

equipment

0 f

the

hundred

dollars.

Additional

elements,

includ-

ing

non-linear

components

(diode

limiters,

etc.)

are

soon

desired

by

anyone

using

the

a n a

log

computer,

and

can

be added as

funds

permit.

Problems

similar

to

the

folloldng

can

be

treated

with

minimum

equipment

and

are

suffi-

cient

to

give

the

undergraduate

student

a

look

at

the

computer,

familiarize

him

with

the

bas-

ic

principles,

and whet

his

enthusiasm

for

more

information.

If

one

neglects

the

phugoid

or

long

per-

iod

oscillation

resulting

from

forward

veloc-

ity

changes

brought

about

by

elevator

deflec-

tion,

the

longitudinal

characteristics

of

an

airplane

can

be

expressed

by

the

follow

i n g

equations:

mv

(9

-

Q()

=

Lor;

Ol

.

'mK;

e =

M~

0(

+

M~a

+ Mi9+

M~~

tl1here m

is

the

mass

of

the

aircraft;

v

is

the

airspeed;

e

is

the

angle

of

pi

tch;

0(

the

<mg-

Ie

of

attack;

Ky

is

the

radius

of

gyr

a

tion;

L

~

is

the

slope

of

the

lift

curve;d'

is

the

deflec

tion

of

the

elevator;

and

Mo(

J

M~

)

Mil

J

M

tf

are

the

usual

stabili

ty

derivatives.

The

equations

can

be combined

into

the

form

where

A,

B,

C and D

are

functions

of

the

mass,

velocity,

and

stability

derivatives.

By

ap-

ply

ing

the

Laplace

transform,

the

tr

an

s

fer

function

relating

~

and d

can

be

shoh'n

to

be

of

the

form

c(

(s) _

d

whe n

K,

j and

Wn

can

be

calculated

in

terms

of

the

airplane

parameters.

lhe

transfer

function

4~)can

be

simu-

lated

by

three

amplifiers

an!

three

potentio-

meters

as

shown

in

Figure

1.

:l.JvJ,..

Figure

1

-

11

-

The

SYmb~,

__

_

COMPUTERS

AND

ENGINEERING

EDUCATION

is

a

potentiometer

set

so

that

the

output

i s

K times

the

input,

K being

less

than

1.

:[::>-

is

an

amplifier

lvi

th

capaci

tive

feedback, thus

is

an

integrator

(with 3

inputs);

is

an

amplifier

with

resistance

feedback used

as

an

inverter

only.

A

series

of

runs determining

0{,

,

for

different

<f'

inputs

(step

function,

s 1 o

lit

ly

increasing,

etc.)

and a

number

of

values 0 f

the

several

parameters provides

an

interesting

and

valuable

means

of

demonstrating the

use

of

the

compute r and

the

relative

importance

of

the

parameters as lltell. A

similar

tre

at-

ment

can be

made,

relating

lift

to

con

tr

0 1

surface

deflection.

A simple

analysis

of

the

behavior 0 f a

series

wound

electric

motor gives the fol101It-

ing

relationship

between

the

angular

velocity

and

the

applied

EMF

.137

K

•

737

K

-=

where J

is

the

moment

of

inertia

of the arm

a-

ture

R

is

the

resistance

of

the motor

L

is

the

inductance

of

the

motor

B

is

the

viscous damping

effec

t

of

the

armature

K

is

the

back

EMF

in

vol

ts

per

radian

per

second

s

is

the Laplace

operator.

This

transfer

function

can be

generated

with

the

same

three

potentiometers

and ampli-

fiers

used

in

the

airplane

problem a bov e •

With

such a

simulated

motor, the

effect

cffric-

tion,

moment

of

inertia,

vol

tage

and

0

the

r

changes

upon

the

acceleration

of

the

mot

0 r

can be

studied.

By

adding

an

additional

integrator,

and

sign

changing

amplifiers,

and feedback

~

Slown

in

figure

2,

the

system

simulates

a servo

mo-

tor

and

follow

up

potentiometer.

A

number

of

interesting

studies

of

the

effects

of

par

a-

meter changes can be

made

with

this

sys

te

mil

-

12

-

e L

::

E

.'131

I<

Figure 2

It

will

have the open loop

transfer

func~ion

• ,.,31 K

An

excellent

discussion

of

the

simulatiOn

of

electrical

and

mechanical systems,

point-

ing

up

the

similarity

of

the

analytical

treat-

ment

of

these

problems

is

given

by

Larrowe.3

o

t,her'

examples

of

the

use

of

the

an-

alog computer as a

teaching

aid,

as

well

as

a

tool

to

be

taught,

lvill

undoubtedly

come

to the

mind

of

anyone

who

undertakes

to

use

this

~

satile

machine

in

his

classes.

It

certainly

is

worthy

of

consideration

when

changes i n

course

content

and

curricula

are

contemplated •

1.

Harold

R.

W.

Benj

amin, Saber Tooth

Curric-

ulum, McGraw-Hill, 1939.

2.

Nixon,

Principles

of Automatic

Controls,

Prentice

Hall,

1953,

Sections

13.40 to

13.70.

3.

Vernon

L.

Larrowe, Control Engineering,

"Direct

Simulation",

November

1954: Vol.

I,

III;

pp. 25-31.

-

El\T)

_

THE PLANNING BEHIND THE

IBM

702

INSTALLATION

,AT

CHRYSLER CORPORATION

EUGENE

LINDSTROM

International

Business

Machines

Corporation

Poughkeepsie,

N.Y.

Last

summer

the

Chrysler

Corporation

be-

came

one

of

the

first

companies

in

the

country

to

install

a

large-scale

electronic

data

pro-

cessing

system

entirely

for

commercial use,

--

in

contrast

to

scientific

or

engineering

use.

This

installation

provides

a good

opportunity

to examine

Chrysler's

experience

in

planning

for

and

installing

the

system.

TIlE

PRELIMINARY

PLANNING

The

decision

of

Chrysler

to

apply

elec-

tronic

data

processing

in

its

Parts

Division

came

after

an

initial

survey

of

its

position

in

the

highly

competitive

automotive

industry.

The

survey.

which began

in

October 1952,

inclu-

ded a

meticulous

examination

of

all

existing

and proposed equipment. Then Chrys

ler

invited

International

Business Machines Corp. and one

other

company

in

February 1953

to

study

the

Parts

Division

for

possible

use

of

electronic

data

processing

machines.

In

order

to conduct

this

study,

Chrysler

formed a

neh'

commi

ttee

consisting

of

seven

mem-

bers.

This committee

was

known

as

the

ED

PM

Planning

Group and

included

men

familiar

with

digi

tal

computers,

office

procedures,

and

elec-

tric

accounting

machines.

The

first

step

b y

the

planning

group

l\'as

to

request

from e a c h

department manager a

written

report

describing

the

operation

of

his

department.

The

EDPM

aan-

ning Group

then

reviel>Jed and

edi

ted

these

r

e-

ports

and

prepared

a methods and

procedures

mar

ual.

The

manual

lias

completed

in

August 1953,

and has proved to be a most

effective

way

to

educate

persons working on

the

project.

Besides.

such

a study

is

valuable

in

itself

since

it

re-exam-

ines

procedures

within

an

organization.

Because

procedures

have a tendency

to

remain s

ta

tic

while

the

l~ork

to

be done

continually

changes,

revieh' from time

to

time

is

essential

to

keep

operations

as

profitable

as

they

might

be.

The

Parts

Division

1'11as

particularly

inves-

tigated

and

studied,

in

the

light

of

the

pr~

dures

outlined

in

the

manual, and

certain

a

p-

plications

1>Jere

selected

which would apparently

lead

to

an

integrated

system

of

operation

thnugh

the

use

of

electronic

data

processing

m~hinery.

In

Oc

tober

1953 poss

ible

programs

lW'ere

outlined

and

graphically

illustrated

with

flow

charts.

Finally,

preliminary

phase

1vas

conclud~d

b y

Chrysler's

action

ordering

an

IBM

Type

702.

FINAL

PLANNING

Four

different

plans

were developed t 0

provide

the

management

of

Chrysler

wi

th

as

with

a

selection

of

action

as

possible.

Each

plan

1'laS

complete

in

detail,

ld

th

all

s uppor

tin

g

data

and

analyses

included.

Ul

timately

the

management

of

Chrysle

t

selected

the

following

five

basic

applications:

(1)

central

inventory

control;

(2)

cen-

tral

inVOICIng; (3)

determination

of

cost

of

sales;

(4)

central

accounts

receivable;

(5)

sales

analysis.

In

order

to

make

necessary

changes

in

pol-

icy,

reorganize

departments,

and

convert

pro-

cedures,

a committee

called

the

Executive

Pol-

icy

Committee

l'11aS

appointed.

This

committee

consisted

of

all

of

the

managers

of

the

vari-

ous

departments

affected

by

the

project;

its

first

step

was

to

examine

all

of

the

proposed

changes recommended

by

the

planning

g r 0

up.

Where

policy

and

procedure

changes

suggested

by

the

planning

committee were

not

approved

~

the

executive

committee, a compromise between

the

two

groups

was

reached.

The

changes

in

policies

and

procedures

were

made

official

by

the

issuance

of

written

procedures

approved

by

the

Division

Comptrol~r

and

the

Vice

President

of

the

Parts

Division.

By

mutual agreement

the

application

to

central

inventory

control

was

selected

in

July

1954

as

the

first

task

to be assumed

by

electronic

wta

processing

machinery

(EDPM).

CENTRAL

INVENTORY

CONTROL

In

central

inventory

control,

before

me-

chanization,

a

certain

"stock

status

record"

made

with

bookkeeping machines

reflected

the

activity

which

occurred

from

eight

to

twelve

weeks

prior

to

the

posting

•.

The

processi

n 9

cycle

for

this

bookkeeping machine method

was

monthly. In

order

to

prepare

for

EDPM,

the

operation

was

converted

to a punched

card

sys-

tem.

The

punched

card

operation

summarized

activity

weekly and produced an

Availability

Report which

indicated

the

status

of

inventory

as

of

b.ro

to

three

weeks

prior.

Us

ing

the

7(Q,

the

processing

cycle

of

the

present

system has

become

daily:

it

reflects

the

-inventory

activ-

ity

which has

occurred

during

the

preceding

dey.

-13 -

IBM

702

INSTALLATION

This permi

ts

"management

by

exception,"

that

is,

isolation

of

those

critical

areas

that

nee

d

special

and immediate

personal

attention

outof

a

maze

of

250,000

routine

decisions

each day.

TIlree

basic

tapes

serve

as

input

for

the

inventory

control

analysis.

The

first

is

the

sales

record

tape which

is

a by-product

of

the

central

invoicing

operation.

The

second tape

contains

all

of

the

transactions

except

sales.

The

third

is

the

master

stock

status

Ie

d g e r

tape.

The

purpose

of

this

particular

run

ism

produce

two

reports

and

the

updated master

in-

ventory

ledger

tape;

these

ttVO

reports

will

con-

trol

the

procurement

and

distribution

of

the

in-

ventory.

The

first

is

a

stock

status

report,

comprehensive

and

produced

on

an

exception

bas-

is.

Whenever

this

report

is

issued

some

type

of

action

is

required.

It

contains

all

of

the

information

necessary

to

complete

any

action

tii

thout

reference

to

any

other

record.

The

sec-

ond

report

or

distribution

report

is

used

to

allocate

material

due

in

at

the

central

for-

tvarding

station

for

distribution

to

other

ware-

houses,

and

to

release

back

orders.

Both r

e-

ports

are

printed

on

a

tape-to-printer

opera-

tion

t\'1

th the

output

tape

records

generated

by

the

702

in

this

operation.

The

mas

ter

s

to

c k

status

ledger

tape

is

automatically

updated

as

a

resul

t

of

this

operation.

That

is,

all

in-

dicative

data

that

has changed

and

inventor

y

status

of

each

part

are

automatically

recorded

on

the

new

master

stock

status

ledger

tape.

Once

the machine has completely posted the

stock

status

record

for

one

part

number

the next

job

is

to

analyze the

stock

status

report

and

decide

if

any

action

has to be taken

on

t

his

item.

In

th

is

program

an

average

of

t

tV

e n t

y-

four

decisions

must

be

made

by

the

702

for each

part

that

has had

activity.

The

average

day's

business

includes

ten

thousand

active

par

t

s.

This"

means

that

240,000

decisions

--

busin es s

decisions,

not program

steps

--

are

made

during

the

regular

tt\'o-hour run each day.

The

decisions

fall

into

four

basic

cate-

gories:

0)

procurement;

(2)

distribution;

(3)

expediting;

(4)

disposition

of

surplus.

These

are

the

four

broad

classifications

of

the

type

of

decisions

that

the

machine

makes

on

eam

part

every time

it

is

active.

The

702

comp~te

ly

watches

all

condi

tions

and

also

checks

the

levels

of

stock

and

the

demand

being placed

on

outlying

plants

for

parts.

There

are

a

number

of

different

formulas

for

acquisition

of

parts

depending

on

the

distribution

policy

for

t h

'e

type

of

part,

service

ruling,

or

cost.

The

IBM

702

examines

much

of

the heading

data

in

the

re-

cord and decides tvhat

parts

of

this

data

apply

to

the

particular

part

being

considered.

The

machine

may

even determine Ivhether

or

not a n

item

should

continue

to

be

stocked

in

a

partic-

ular

area

where

sales

have

declined.

If

th

e

prerequisite

conditions

for

stocking

materials

are

not

satisfied,

the

machine

prints

0

uta

stock

s t"atus

report

which

tells

management

why

the

report

was

printed

and what

action

is

to

be

taken.

A

third

type

of

review

made

by

the

machine

is

for

follow-up ac

tion.

If

the

cur-

rent

availability

satisfies

national

and

local

requirements but

the

"on

Hand"

inventory

is

leg;

than

60

or

30

days'

supply

for

both

require-

ments, then a

stock

status

report

is

printed

out

and

directed

to

the

expediter

for

follow-

up

action.

The

distribution

report,

the

second

re~

of

this

particular

run,

is

also

used

for

the

allocation

of

parts

received

which

are

on

bac~

order.

The

702

automatically

allocates

the

quantity

to

be

used

against

back

orders

by

date

or

age

until

the

entire

quantity

is

exhausted.

Should

it

be necessary

to

make

a

partial

ship-

ment

because

the

allocated

quantity

is

not

suf-

ficient

to

cover

all

back

orders,

the

702

will

automatically

record

this

case.

This

particu-

lar

record

is

maintained

on

the

back-order sum-

mary

tape.

The

quantity

of

parts

on

back order

may

vary from

several

hundred

to

a thousand.

It

is

expected

that

this

number

will

be

greatly

reduced through

the

daily

control

of

inventory.

To

aid

the

back-order

situation,

a weekly

cus-

tomer

shortage

report

of

all

back

orders

i s

also

prepared.

TIle

historical

data

now

being

recorded

on

magnetic

tape

should

be

quite

val-

uable

in

the

future

use

of

the

702

for

compre-

hensive analyses

of

inventory

and

stock

move-

ments. TIlis

will

enable

management

to

exe

r-

cise

more

precise

methods

in

proj ec

ting

future

inventory

movement.

CENTRAL

INVOICING

TIle

second

application

is

central

invoic-

Jng.

To

make

this

operation

easier

each tvare:-

house has

an

installation

of

electric

account-

ing (punch card) machines

I-vhich

prepare-

ship-

ping

orders.

TIlis enables shipments

tore

made

ordinarily

within

eight

hours

after

receipt

of

a

customer's

order.

Notification

of

shipment

is

received

from the

shipping

department,

the

cards

which were used

to

prepare

the

shipping

order

are

pulled,

changes

are

made

in

the

cards

,to

indicate

the

exact

quantity

shipped,

and

correct

shipment cards

including

name

and

ad-

dress

cards and

data

cards

are

forwarded daily

by

air

express

to

the

central

invoicing

cen

ter~

A

control

card

is

inserted

with each group of

cards

indicating

the

exact

amount

of

sale

for

that

group.

When

the

cards

are

received

they

are

automatically

balanced

to

the

control

card

and

converted

to magnetic

tape.

The

following

output

tapes

are

prepared

as

the

result

of

the

invoicing

operation:

-

14

-

IB~

702

INSTALLATION

1.

The

invoice

record,

converted

by

a tape-

Hi

th

no

action

being

required

in

the

way

of

key

to-printer

operation,

preparing

the

invoices.

punching. However,

four

basic

classifications

have been

established

for

handling

remittances

2.

The

invoice

register

tape,

converted

automatically

or

semi-automatically:

by

the

printer,

preparing

the

invoice

register.

3.

The

sales

analysis

tape

by

product,

by

type

of

customer.

4.

The

accounts

receivable

tape.

5.

Disbursements

tape,

discussed

in

the

inventory

control

application.

6.

Back-order

detail

for

releasing

back

orders.

The

invoicing

operation

is

currently

being

performed

w~th

the

use

of

electric

accounting

machines

installed

at

five

warehouses.

When

this

operation

is

c.onverted

to

the

702 and per-

formed

centrally

it

will

reduce the

t,"ork

load

in

these

five

installations

by

5~/o.

Invoicing

for

the

entire

United

States

is

done

in

two

hours

of

702 time with

an

additional

12 hours

for

the

tape-to-printer

operation

each day.

COS

T

OF

SALES

The

third

application

is

cost

of

sales.

Unit

cost

for

each

part

is

available

on

the

master

inventory

ledger

tape.

Disbursements

are

extended

to

write

a

tape

for

cost

of

sales

f

or

the

day's

sales;

then they

are

merged with

the

month-to-date

cost-of-sales

tape.

At

the

end

of

the

month

the

final

cost

of

sales

is

printed

on

a

tape-to-printer

operation

so

that

immediately

after

the

close

of

the

month's

busi-

ness

the

cost

of

sales

for

all

sales

for

that

period

is

available.

This

is

an

excellent

ex-

ample

of

the

extensive

by-product

advantages

that

result

from

the

integration

of

an

opera-

tion

of

this

type.

ACCOUNTS

RECEIVABLE

The

fourth

application

is

accounts

receiv-

able.

It

is

interesting

to note

that

the

~

or

portion

of

Chrysler

Accounts Receivable

re-

sults

from

the

sale

of

parts.

These

tiansac-

tions

are

recorded

on

magnet1c

tape

produc e d

during

the

invoicing

operation.

The

balan

c e

is

key punched

into

cards

and then

convert

ed

to

magnetic

tape.

The

702

is

used

for

too sort-

ing

of

the

tape

into

dealer

sequence

for

t w

<>

groups

kn01'ln

as

the

85%

group and

the

15%

group,

and

the

two

tapes

are

merged

prior

to

the

ac-

counts

receivable

run.

The

remi

ttance

tap

e

used as

input

is

written

at

the time a

state-

ment

is

sent

to

a

dealer

or

a customer. A

pr~

punched remi

ttance

card

is

enclosed

1\'1

th

the

request

that

the

card

be

returned

with

the

re-

mittance.

Thus

most of

the

cards

are

returned

1.

Complete cash payment with

the

com-

plete

discount

honored.

2.

Partial

cash

remittance

with

the

dis-

count.

3.

Complete payment with

no

discount.

4.

Partial

payment with

no

discount.

The

cards

are

then

sorted

to

dealer

n

urn

be

r,

merged

tvi

th

miscellaneous

transaction

and

ad-

j ustment

cards,

and converted to magnetic

tape.

This

tape

becomes

the

input

remi

ttance

tap

e •

Each

day

the

accounts

receivable

are

made

cur-

rent

by

up-dating

the

mas

ter

tape.

As

a r

e-

suIt

it

is

possible

for

the

Credit

Department

to have a

daily

record

of

all

accounts exceed-

ing

their

limits.

~

A

lES

ANALYS

IS

The

fifth

application

is

sales

analysis.

The

various

distributions

and summarizations

required

for

sales

analysis

are

easily

obtain-

ed

on

the

,702

by

reading

the

tapes

into

the

central

processing

unit.

Based

upon

this

in-

formation a

large

number

of

distributions

can

be

stored

on

the

magnetic drum. A

quarterly

report

is

prepared

by

dealer

and

accoun

ti

n g

class

showing

the

amount

of

money,

the

number

of

orders

by

type of

order,

and

whether

it

was

a

stock

or

emergency

order.

Accurate.

__

sales

statistics

for

control

of

dealer

sales

are

al-

ways

available

becaus@

of

this

operation.

An

annual

report

is

prepared

by

warehouse and

by

states

for

all

delivered

parts.

CONCLtEIONS

This

integrated

data

processing

s y s

ten,

has

become

a

valuable

tool

for

the

management

of

Chrysler.

The

immediate advantages

thatthe

company

may

expect from

the

use

of

electronic

data

processing

machines

are

these:

1.

The

major

portion

of

all

input

infor-

mation

is

available

as

the

result

of

a previous

automatic

operation.

2.

Management

are

able

to

use

their

time

more

economically, because

the

logical

ability

of

the

electronic

data

processor

permits a

dis-

criminating

selection

of

more

critical

probrems

for

study.

3.

Operation

at

high speed has been

ob-

tained.

(continued

on

page

32)

-

15

-

PUBLICA

TIONS

FOR

BUSINESS

ON AUTOMATIC COMPUTERS:

A

SUPPLEMENTAL

LISTING

NED

CHAPIN

Lecturer,

Illinois

Insti

tute

of

Technology

Q.icago,

illinois

This

paper

lists

for

business

people

w

hat

migh t

be

called

a

supplemental

reading

1 i s t 0 f

publications

in

the

field

of

automatic

computers.

This

listing

is

supplemental

to

the

one

published

in

"Computers and

Automation"

for

Sept.

1955

with

the

title

"Publications

for

Business

on

Automatic

Computers:

A

Basic

Listing."

It

is

expected

that

this

supplemental

listing

will

be

followed

by

a

third

paper

t'li

th

the

title

"Reference

Listing."

The

publications

listed

in

this

paper

present

additional

points

of

view

and

additional

informa-

tion

(as

t\'e11 as

repeating

much

of

the

same

infor-

mation)

on

the

same

topics

covered

by

the

pub

1

i-

cations

ci

ted

in

the

"Basic

Listing."

The

publi-

cations

cited

in

this

supplemental

lis

ting

are

grouped

by

subject

matter.

Within

each

subject

mat

ter

grouping,

the

publications

are

1

is

te

din

approximate

order

of

scope

and

depth

of

coverage,

the

more

elementary

being

listed

first,

the

mor

e

advanced

being

listed

last.

Items

that

are

back-

ground

material

have

been

identified

as

such

in

the

annotation.

Use

of

Computers

in

General

Walter

II.

Wheeler,

"Wanted: A

Revolution

in

the

Office,"

Management Review, V

42,

#12 (Dec. 1953),

pp.

712-713

Tedious

pape~~ork

should

be

mechanized.

LmITrence

P.

Lessing,

"Computers

in

Business,"

Scientific

American, V

190,

#1

(Jan.

1954),

pp

21-

25

An

easy-to-read

general

introductory

article.

Everett

S.

Calhoun,

"The

Challenge

of

Electronic

Equipment,"

'NACA

Bulletin,

V

34,

Section

I,

#

10

(June

1953),

pp.

1267-1279

A

general

introductory

article.

Matt

W.

Boz, "You,

Electronics,

and

the

Brass,"

Papers

of

the

Sixth

Annual

Systems

Meeting

(N

e

h'

York: Systems and

Procedures

Association

of

Amer-

ica,

1953),

pp.

14-20

Some

good

suggestions

from

one

systems

man

to

another

on

investigating

possible

automati

c

computer

uses.

P.

D. Laubach and L. E. Thompson,

"Electronic

Com-

puters:

A

Progress

Report)

"Harvard

Bus

iness

Review,

V

33,

#2

(March-April,

1955),

pp.

120-128

Cost

savings

are

slow

to

appear

from

the

use

of

automatic

computers,

this

survey

indicates.

The

artic

Ie

also

re-emphas

izes

that

to

use

an

automatic

computer

requires

a good

de

a 1

of

preliminary

work.

C. E.

Knight

and C. II. Fawkner, "The

Impact

of

Automation

on

the

Company

Organization,"

Ge

ner

a 1

Management

Series

Number 178

(New

York: Amer

ican

Management ASSOCIation,

Inc.,

1955),

pp.

11-21

A

computer

Hill

have

an

important

r

ole

i n

-16 -

automated

plants,

and

organiZational

changes

will

have

to

be

made

to

reflect

that

role.

William

Bamert,

"Making A

Feasibility

Stu

d

y:

Strengthened

Management

Control,"

Electronic

Data

Processing

in

Industry

(New

York: American Manage-

ment

Association,

Inc.,

1955),

pp.

112-124

This

article

lists

some

of

the

advantages

,in

terms

of

systems

improvements

that

might

b e

realized

with

the

use

of

an

automatic

comput-

er.

Paul

Kircher,

"The

Gap

Between

the

Ele

c t

ron

i c s

Engineer

and

the

Accountant,"

The

Controller,

V

22,

#8

(Aug.

1954),

pp.

358-362,

374 .

Vocabulary

and

atti

tude

are

the

causes

of

the

gap;

better

communication

between

th

e

two

groups

should

be

established.

James

Gibbons,

"Is

Push

Button

Accounting

Aro

un d

the

Corner?"

Interal

Audi

tor,

V

11,

#1

(M

arc

h ,

1954),

pp.

44-53

A

payroll

example

is

given

to

help

the

author

anSlver

"not

yet."

John

H.

Lindholm

et

al.,

Electronic

Business

M

a-

chines

--

A

New

Tool

for

Management {Winches

te

r,

Mass.:

John

H.

Lindholm,

1953),47

pp.

Also,

in

condensed

form:

"What

Business

Can

Expect

of

Elec-

tronic

Office

Machines,"

The

Office,

V

38,

#6

(~c.

1953),

pp.

81,

85-86,

89-90,

92

Automatic

computers

are

not

giant

brains,

but

they

can

nevertheless

still

be

used

in

busi-

ness.

F.

J.

Porter,

"EEl,

Con

Edison,

and

Electronics,"

Edison

Electric

Institute

Bulletin,

V

22,

#12

(Dec.

1954),

pp.

419-424

Describes

the

preparation

these

groups

h a v e

mal.le

for

the

use

of

automatic

computers.

Joseph

E.

Perry,

et

al.,

"ABA

Report

of

the

Co

m-

mittee

on

Electronics,"

Computers and

Automation,

V

3,

#1

(Jan.

1954)

pp.

10-12

The

report

sugges

t a

rather

conservative

grad-

ualism

in

adopting

electronic

equipme

n t

in

banking.

Feasibility

Studies

Radio

Corporation

of

America,

An

Electronic

A

c-

counting

System

for

a Medium

Size

Utility

(N

e w

York: Radio

Corp.

of

America,

June

1953),

86

p~.

This

is

a summary

of

a

feasibility

study.

Ned

Chapin,

"Justifying

the

Use

of

An

Automa

tic

Computer,"

Computers and

Automation,

V

4,

#8 (Aug.

1955),

pp.

17-19

Expected

savings

must

be

balanced

against

in-

creased

inves

tment

in

justifying

an

automatic

computer.

Matt

W.

Boz,

"Building

a

Checklis

t

for

Dec

i s

ion

on

Electronic

Accounting,"

Office

Management

Ser-

ies

Number 131

(New

York: American Management

As-

sociation,

Inc.,

1953),

pp.

9-13

Discusses

some

of

the

considerations

import-

ant

to

an

oil

company

in

the

possible

appli-

PllULICATIONS

cation

of

automatic

computers.

John

F.

Feagler,

et

a1.,

"Making a

Feas

i

bi

1 i t y

Study,

"

Elec

tronic

Data

Process

ing

in

In

d

us

try

(New

York: American Management

Association,

Inc.,

1955),

pp.

74-124

A

discussion

of

the

Chesapeake and Ohio

Rail-

way

Co.

's

approach

to

the

uses

of

an autormtic

computer.

R.

F.

Clippinger,

"Economics

of

the

Digi

tal

Co

m-

puter,

"

Harvard

Business

Review, V

33,

11:

1

(Jan.-

Feb.

1955),

pp.

77-88

Discusses

in

general

terms

the

cost

of

oper-

ating

an

automatic

computer.

Emerson

F.

Cooley,

"Elec

tronic

Equipment Appl i e d

to

Periodic

Billing,"

Computers and

Automation,

V

2,

11:

8 (Nov.

1953),

pp.

17-21

If

a

systems

analysis

can

show

that

a co s t

saving

will

result,

then

the

use

of

an

auto-

matic

computer

may

be

desirable.

Bernard

Whitney,

"How

to

Sell

Accounting

Systems,"

Systems and

Procedures

Quarterly,

V4,

#4

(Nov.

1953),

pp.

6-7,

22

Do

not

try

to

do

it

with

the

"accountant's

cost

data.

Joseph

0.

Harrison,

"Use

of

Computing Machines

in

Operation

Research,"

Operations

Research

for

Man-

agement

<Baltimore:

John

Hopkins

Press,

1954),

pp.

203-216

Some

interesting

formulas

are

given

to

assist

in

comparing

manual

vs.

computer

methods 0 f

doing

a

task.

Charles

1.

Keelan,

"Financial

Decision,"

Office

Management

Series

Number 135

(New

York:

Americ

a n

Management

Association,

Inc.,

1953),

pp.

24-31

Some

interes

ting

formulas

are

presented

to

as-

sist

in

costing

out

a

system.

Eugene

L.

Grant,

Principles

of

Engineering

Econo-

~

(New

York:

Ronald

Press

Company,

1950),

623

pp.

A

background

book; one

of

the

bes

t

references

on

how

to

make a

cost

analysis.

Sys tems Analys

is

Howard Aiken,

"Preparing

Students

for

the

Aut

0-

matic

Office,"

Workshop

for

Management

(New

York:

Management

Magazines,

Inc.,

1955)

pp.

341-348

The

basic

10'9'fc

of

business

operations

is

not

well

enough

understood

to

make

the

full

po-

tential

of

an

automatic

computer

realized

in

business.

A.

S.

Householder,

"Office

Needs More Com

pIe

x

Computers Than Does

Science,"

The

Office,

V41,

#1

(Jan.

1955),

pp.

82,

138,

140

The!

complexi

ty

referred

to

is

not

so

m u c h

in

the

computers

but

is

in

unders

tan

din

g

what

business

needs.

R

•.

C. Ludlow, "Management

in

the

Electronic

Age,"

Journal

of

Machine

Accounting

Systems

and Manage-

ment, V

6,

# 5 (May,

1955),

pp.

8-10,

14,

16

Some

thinking

on

the

purposes

of

systems.

John

M.

Theis,

"Practical

Application

of

E1

e

c-

tronic

Equipment,"

Journal

of

Machine

Accounting

Systems

and Management, V

6,

#3

(March,

1955),

pp.

5,

6,

8,

16-17.

Also

NACA

Bulletin

Confer-

ence

Proceedings,

V

35,

#12 (Aug.

1954),

pp

1732-

1739 Avoid

the

mistake

of

simply

me c

han

iz

i n g

present

procedures.

E.

L.

Morrison,

"Making a Feas

ibili

ty

Study:

Com-

munication

as an

Element

in

an

Integrated

D a t

a-

-17 -

handling

System,"

Electronic

Data

Processing

in

Industry

(New

York: American Management

Associa-

tion,

Inc.,

1955),

pp.

84-96

A minimum

of

information

should

be t

ran

s-

mitted,

but

it

must be

adequate

in

content,

form,

accuracy,

and

timing.

Kenneth A:

Wulff,

"Planning

for

the

Automatic

Of-

fice--Consul

tant

's

Vielvpoint,"

Workshop

for

M a

n-

agement

(New

York: Management

Magazines,

Inc.

,

1955),

pp.

349-354

Attention

should

be

given

to

systems

changes.

Warren

P.

Livingston,

"Where

Do

We

Stand

Today

on

Automatic

Accounting

Machines,"

Office

Management

and

Equipment,

V

15,

11:1

(Jan.

1954),

pp.

24

-26,

83-84

Not

much

of

anywhere

until

systems

and

meth-

ods

are

standardized.

Ralph

II.

Eidem, "Review

of

Developments

in

Office

Electronics,"

Office

Management

Series

Number 135

(New

York: American Management

Association,

Inc.,

1953),

pp.

3-9

Even

though

present

equipment

is

neither

ideal

nor

perfect,

it

can

be

used

to

advantage.

II.

D.

Buskey and

V.

R. Huskey,

"Nelli

Frontiers

in

Business

Management and

Control

Being

Established

by

Electronic

Computers,"

Journal

of

Accountancy,

V

93,

11:1

(Jan.

1952),

pp.

69-75

New

equipment

makes

possible

new

sys

tems.

David

M.

BroHn, "Computer

in

the

Factory,"

Co

m-

Quters

and

Automation,

V

2,

11:7

(Oct.

1953),~

1-5

Potentially,

an

automatic

computer

can

be

applied

to

the

control

of

production

opera-

tions.

Joseph

Pelej,

"Hot..,

Will

Business

Electronics

Af-

fect

the

Auditor's

Work?"

Journal

of

Accountancy,

V

98,

11:1

(July,

1954),

pp.

36-44

"Not much"

this

article

sugges

ts.

R. T. Wiseman,

"Life

Insurance

PremiumBilling

and Combined

Operations

by

Electronic

Equipment,"

Journal

of

the

Association

for

Computing

Machin-

ery,

V 1,

11:1

(Jan.

1954),

pp.

7-12

A

discussion

in

some

detail

of

are

cor

d

modification

application.

Melvin

E.

Davis and

John

J.

Finelli,

et

a1.,

Re-

port

of

the

Committee on·New

Recording

Means and

Computing

Devices

(New

York:

Society

of

Ac

tUiJries,

Sept.

1952),

107

pp.

Also

condensed

in:

The Con-

troller,

V

20,

11:10

(Oct.

1952),

pp.

465-469

Some

useful

guiding

princ

iples

can

be

found

in

the

middle

part

of

this

report.

Melvin

E.

Davis,

"Some

Current

Thoughts

on

the

Possible

Uses

of

Magnetic

Tape

Policy

Files

man

Insurance

Office,"

Electronic

Data

processing

in

Industry

(New

York: American Management

Associa-

tion,

Inc.,

1955),

pp.

217-230

A

discussion

of

how

systems

using

magneti

c

tape

can

be

operated.

Harry

Eisenpress,

et

al.,

"Charting

on

Automatic

Data

Processing

Systems,"

Computers and Automation,

V

4,

#8 (Aug.

1955),

pp.

21-23,

27

High-speed

printers

can

be

made

to

p r i n t

out

charts.

William

R.

Fair,

"Analog

Computation

of

Business

Decisions,"

Journal

of

the

Operations

Research

Society

of

America,

VI,

#4 (Aug.

1953),

pp.

208-219

Analog

computers

can

be

used

to

solve

some

Operations

Research

problems.

PUBLICATIONS

Human

Relations

Devereaux

C.

Josephs,

"AJEEB,

the

Mechanical

Chess

Player,"

Workshop

for

Management

(Net·'!

York:

Manage-

ment

Magazines,

Inc.,

1955),

pp.

336-340

An

automatic

computer

need

not

and

probably

will

not

cause

technological

unemployment.

CIO

Economic

Staff,

"Automation--A

Report

to

th

e

UAW-CIO

Economic and

Collective

Bargaining

Confer-

ence,"

Computers and

Automation,

V

4,

1t5

(May,

1955),

pp.

14-20

Organized

labor

presents

its

reasoning

about

the

changes

that

automation

brings.

Examples

of

Applications

Richard

G.

Canning,

et

a1.,

"Business

Data

Pr

0-

cessing:

A Case

Study"

Trends

in

Computers

(N

eu

York:

Institute

for

Radio

Engineers,

Inc.,

1954),

pp.

80-104

A

discussion

of

a

suggested

application

0 f

automatic

computers

to

retail

opera

ti

0 n s •

Some

procedures

are

presented

in

considerable

detail.

Robert

M.

Smith,

"Is

this

a

B1

ueprint

for

Torno

r-

row's

Office?",

Office

Management and Equipmen t ,

V

16,

1t8

(Aug.

1955),

pp.

12-14,

60,

64

Discusses

the

centralized

data

proces

sin

g

proposed

at

Sylvania.

Herbert

O.

Brayer,

"How

Five

Maj

or

Companies

Use

the

Electronic

Calculator,"

America"

Bus

i

ness,

V

23,

1t

11

(Nov.

1953),

pp.

18-19,

41-44

Pennsylvania

Railroad,

Pure

Oil,

III

i no i s

Bell,

International

Harvester,

and

Detroit

Edison

are

the

companies.

Fred

E.

~elsh,

"What

We

Use

Our Computer

For,

"

NACA

Bulletin,

V

36,

Section

I,

1tl

(Sept.

1954),

pp.

31-36

Discusses

the

U.

S.

Gypsum

applications.

William

E.

'Eggleston,

"Customer

Accounting:

P r

0-

gramming Methods and

Problems,"

Elec

tronic

D a t a

Processing

in

Industry

(New

York: American Manage-

ment

Association,

Inc.,

1955),

pp.

190-196.

Al

s 0

summary

in:

Computer

Application

Symposium

(Chi-

cago:

Armour

Research

Foundation,

1955),

p.

7

A

description

of

the

Chicago Commonwe a 1 t h

Edison

applications.

Mina

Rees,

Edito~,

Managerial

Aspects

of

Digital

Computer

Installations

(Washington,

D.

C.:

Office

of

Naval

Research,

1953),

36

pp.

A summary

of

the

U.

S.

Government's

experi-

ence

in

operating

some

scientific

computers.

Jerry

and

Electra

T.

Kluttz,

"Better

Governmen t

for

Less

Cost,"

Nation's

Business,

V

43,

112

(Feb.

1955),

pp.

38-41

A

discussion

of

the

U.

S.

Government

uses

for

accounting

computers.

J.

Henry McCall,

"Electronic

Data

Pro

c e s

sin

g,

"

Sys tems and

Procedures

Quarterly,

V

5,

1t4

(N

0 v •

1954),

pp.

12-16

Very

little

about

computers

here

but

s 0 m e

data

on

applying

elec

tronics

to

gather

i n g

input

data.

Paul

T. Nims,

"Digital

Computer

Applications

in

the

Automatic Indus

try,"

Proceedings

of

a Symposi

urn

of

Industrial

Applications

of

Automatic

Compu

tin

9

Equipment (Kans as Ci t.y, Mo.: Midwest

Research

In-

stitute,

1953),

pp.

146-162

Discusses

engineering

applications,

not

ac-

counting

applications.

-

18

K.

D.

Tocher,

"The

Application

of

Automatic Com-

puters

to

Sampling

Experiments,"

Journal

of

th

e

Royal

Statis

tical

Society,

Sere

B,

V

26,

1t

1

(Jan.

1954),

pp.

39-61

Discusses

a

Monte-Carlo

technique.

Integrated

Data

Processing

Harold

F.

Van

Gorder,

"Achieving

Greater

Produc-

tivity

in

Accounting

Through

Integrated

Data

Pro-

cessing,"

NACA

Bulletin

Conference

Proceedings,

V

35,

1t

12 (Aug.

1954),

pp.

1708-1731

Editor,

"u.

S.

Steel

Sets

Up

Automatic

Off

ice,"

Office

Management and Equipment, V

15,

1t3

(March,

1954),

pp.

24-27,40,74;

V

15,

1t4

(April

1954),

pp.

19-21,

29-30,

86,

88-91

Ralph

W.

Fairbanks,

"Alcoa

Inaugurates

Integrated

Data

Processing

on a

National

Scale,"

Office

Man-

agement and Equipment, V

15,

1t12

(Dec.

1954),

pp.

12-14,

32,

66,

68

Edi

tors,

"Integrated

Data

Processing

Comes

to

Life,"

Automation,

V

2,

1t5

(May,

1955),

pp.

59-65

P.

B.

Garrott,

"Integrated

Data

Processing

Brings

Automation

in

Papert10rk,"

Automation,

VI,

1t5

(Dec.

1954),

pp.

31-39,

V

2,

~l

(Jan.

1955),

pp.

33-44;

V

2,

1t2

(Feb.

1955),

pp.

65-72

Ralph

W.

Fairbanks,

"Integrated

Data

Process

in

g

for

the

Smaller

Office,"

Office

Management

and

Equipment, V

15,

1t6

(June,

1954),

pp.

18-19,

78-

80,

83

Robert

M.

Smith,

"Automation

in

the

Office,"

Of-

fice

Management and Equipment, V

16,

1tl

(Jan.

1955)

pp.

38-40,

65-68

Editors,

"I,ntegrating

the

Office

for

Automation,"

The

Office,

V

39,

1t4

(April,

1954),

pp.

66,

68,71

Harold

F.

Van

Gorder,

"Fundamentals

of

Integrated

Data

Processing,"

The

Office,

V

39,

1t4

(Apr.

1954),

pp.

71-72,

75

H.

W.

Moore,

"Data

Origination,"

The

Office,

V

39,

1t4

(Apr.

1954),

pp.

75-76,

84,

87,

89

L.

W.

Calkins,

"High Speed

Processing,"

The

Office,

V

39,

1t4

(Apr.

1954),

pp.

79,

95-96

James Thomson,

"Data

Calculation

and

Distribution;'

The

Office,

V

39,

1t4

(Apr.

1954),

pp.

78,

89,

92

R.

W.

Baridon,

"Communication,"

The

Office,

V

39,

1t4

(Apr.

1954),

pp.

80,

96,

98,

101

Computer

Fundamentals

Louis

N.

Ridenour,

"The

Role

of

the

Computer,"

Scientific

American, V

187,

1t3

(Sept.

1952),

pp.

116-130

A good

discussion

of

the

differences

between

digital

and

analog

computers.

H.

W.

Steinhaus,

"Keep

Electronics

in

Mind

Whe

n

Planning

for

the

Future,"

The

Office,

V

39,

1t

I

(Jan.

1954),

pp.

70-71

This

article

discusses

history.

Philip

and Emily

Morrison,

"The

Strange

Li

f e

of

Charles

Babbage,"

Scientific

American, V 186,

1t4

(Apr.

1952),

pp.

66-73

Some

history

on an

important

person

in

the

development

of

computers.

Martin

L.

Klein,

"Digi

tal

Technique

and

Bin

a

ry

Numbers,"

Instruments

and

Automation,

V

27,

1t12

(Dec.

1954),

pp.

1944-1947

'A

presentation

of

the

binary

system.

Beatrice

H.

Wors

ley,

"Numerical

Representation

in

Fixed-Point

Computers,"

Computers and Automation,

· PUBLJCATIO":S

V

4,

n5

(May,

1955), pp. 10-13.

An

understandable

introductory

article.

J.

H.

Felker,

"Arithmetic

Processes

for

Digital

Computers,"

Electronics,

V

26,

n3

(March

195

3)

,

pp. 150-155·

A

discussion

of

arithmetic

operations.

Vincent

Petrucelly,

"Boolean Algebra:

New

Tool

for

Circuit

Designers,"

Electrical

Manufacturing,

V 54,

n2

(Aug.

1954), pp. 97-101

A

presentation

of

simple Boolean

operations.

Robert

Serrell,

"Elements

of

Boolean

Algebra,"

Proceedings of

the

IRE,

V 41, nlO (Oct. 1

953)

,

pp. 1366-1379

A good

but

very

technical

article

on

Boo

wan

algebra.

,Programming

Grace

M.

Hopper, "Education of a Computer,"

Pro-

ceedings

of

a Symposium

on

Industrial

Applic~

of

Automatic Computing Equipment, (Kansas

City,

Mo.:

Midwest Research

Institute,

1953)pp. l39-Y1.

A

discussion

of

programming

aids,

and

the

need

for

programming

aids.

This

article

is

not a

reprint

of

the

author's

other

article

wi

th

the

same

title,

although

much

the

same

ground

is

covered.

Philip

R.

Marvin, "Engineering Automatic Systems,"

Automation, V 1,

n4

(Nov.

1954), pp. 47-48

The

writer

addresses

himself

to

the

program-

ming

considerations.

Charles

W.

Adams,

"Small Problems

on

Large

COIn-

puters,"

Proceedings

of

the

Association

for

Computing Machinery

(Pittsburgh,

Pa.:

Ric

h a r d

Rimbach

Associates,

1952), pp. 99-102

Discusses

some

programming

techniques

to

help

make

small

problems economically workable

on

an

automatic computer.

Grace

M.

Hopper and John

W.

Mauchly,

"Influe"nce

of

Programming Techniques

on

the

Design

of

Compu-

ters,"

Proceedings

of

the

IRE,

V 41,

nlO

(0 c

t.

1953), pp. 1250-1254

Programming and computer

design

go

hand

in

hand.

Ida

1.

Rhodes, "The

Human

Computer's

Dreams

0 f

the

Future,"

Proceedings of

the

Electronic

Com-

puter

Symposium

(Los

Angeles:

.Institute

for

Radb

Engineers

Professional

Group

on

Electronic

Com-

puters,

1952), pp.

XII:l

to

XII:5

Interesting

reading

about

t",hat

might

com

e

to

pass.

Storage

Devices

David

R.

Brmm

and

Ernst

Albers-Shoenberg,

"Fer-

rites

Speed

Digital

Computers,"

Electronics,V26,

n4

(April,

1953), pp. 146-149

A

presentation

of

ferrite

storage

devices.

Gilbert

W.

King,

"A

New

Approach

to

Information

Storage,"

Control

Engineering,

V 2, n8

(A

u

g.

1955), pp. 48-53

A

high-capacity

quick-access

photoelectric

disk

technique.

Input-Output

Devices

Charles

G.

Chase,

"Why

the

Automatic

Office

Is

Still

Some

Years

Off,"

The

Office,

V 41,

nl.

-19 -

(Jan.

1955), pp.

84,87,

91

The

reason

is

the

high

cost

of

input.

Mil

ton

E.

Mengel,

"Electronic

Bus

iness

Machin e s ,

Today, Tomorrow,"

Sys

tems

al!d

Procedures

Quarter-

lY,

V

5,

n4

(Nov.

1954), pp. 7-11

Input

and

sorting

are

the

major problem

areas.

John Rudolf, "Tape

Controlled

Machine

in

the

Shcp,

"

Factory

Management

and Maintenance, V 113, n 4

(April,

1954), pp. 76-77

Discusses

how

production

equipment can be

oon-

trolled

by

automatic computing equipment

out-

put.

Harry

E.

Burke,

"A

Survey

of

Analog-to-Digital

Con-

verters,"

Proceedings

of

the

IRE,

V 41,

nlO

(Oct.

1953), pp. 1455-1462

Gives a

general

discussion

of

analog-digital

devices.

Harry

W.

Mergler,

"A

Digital-Analog

Machine Tool

Control

System," Trends

in

Computers

(New

York:

Institute

for

Radio

Engineers,

1954), pp. 46-59

This

is

an

example

of

the

use

of

dig

1 t a

1-

analog equipment

for

control

purposes.

F.K. Williams and

M.L.

Klein,

"Survey

of

Integrat-

ed Data Convers

ion

Sys tems, " Ins truments

and

Automation, V 27,

n9

(Sept.

1954), pp. 1460-1462

The

integration

referred

to

in

the

title

is

not paper

tape

but

is

data

reduction

by

using

analog-to-digital

devices.

John

J.J.

Kernahan,

"A

Digital

Code

Wheel,"

Bell

Laboratories

Record, V

32

U4

(April,

1954), pp.

126-131

Discusses a code wheel using

the

reflected

binary

code.

Edward E. David,

Jr.,

"Ears·

for

Computers: Auto-

matic

Digital

Recognizer,"

Scientific

American,

V 192,

n2

(Feb. 1955), pp. 92-98

Describes

some

developments

in

coding spoken

speech

in

a

digital

form.

E. T.

Shipley,

"Punched Card CheckS," Auditgram,

V 31,

n4

(April,

1954), pp. 34-38

A

better

solution

is

to

skip

the

punched

cmrl

stage

a~d

go

directly

to

more

advanced

equip-

ment.

Aspects of Using Computers

Geoffry Ashe,

"Introducing

Computers

to

Beginners,"

Computers and Automation, V 3,

n3

(March, 1

954)

,

pp. 8-11

Removing

the

"air

of

mystery"

should

be

the

teacher's

goal.

William

B.

Elmore,

"Some

Aspec

ts

of

Reliability

in

Electronic

Data

Processors,"

Electronic

Data

Processing

in

Industry

(New

York: American Manage-

ment

Association,

Inc.,

1955), pp. 139-147

Very

s·imilar

in

content

to

the

author's

arti-

cle

in

Computers and Automation.

Walter

F. Bauer and John

W.

Carr

III,

"On

the

Dem-

onstration

of

IIigh Speed

Digital

Computers,"

Jour-

nal of

the

Association

for

Computing Machinery,

V

1,

n4

(Oct. 1954), pp. 177-182

Offers

a number

of

suggestions

to

m a k e

demonstrations

more

informative

and

interest-

ing.

C.C.

Gotlieb,

"Running a Computer

Efficiently,"

Journal

of

the

Association

for

Computing Machin-

~,

V

1,

n4

(July

1954), pp. 124-127

Keep

a

detailed

set

of

operating

and

histo~

records,

and use

those

records.

Paul Brock and

Sibyl

Rock, "Problems

in

Acceptance

PUBLICATIONS

Testing

of

Digital

Computers,"

Journal

of

the

As-

sociation

for

Computing Machinery, V 1,

#2

(April,

1954), pp. 82-87

Describes

some

acceptance

testing

procedures.

Norman

II. Taylor, "Marginal Checking as

an

Aid

to

Computer

Reliabili

ty,"

Proceedings of

the

IRE,

V

38, #

12

(Dec. 1950), pp. 1418-1421

Gives a

description

of

marginal checking.

Particular

Automatic Computers

Charles

IV.

Adams,

"Automatic Data Processing

Equip-

ment,"

Electronic

Data Processing In

Industry

Q\ell1

York: American

Management

Association,

Inc.,

195~,

pp. 125-138

A

general

overvienl' of

available

equipment.

Mina

Rees, "Computers 01954,"

The

Scientific

~bnth

ly,

V 79, # 2

(Aug.

1954), pp. 118-124

A

summary

of

equipment developments.

W.

H.

MacWilliams, "Computers

--

Past,

Presen

t ,

and

Future,"

Electrical

Engineering, V 72, # 2

(Feb. 1953), pp. 116-121

A survey

of

equipment.

Herbert

O.

Brayer,

"Can

European Machines Help

American

Offices?"

American Business

'0

V 24,

#6

(June 1954), pp. 10-13,

39

A survey

of

European automatic equipment.

Herbert

O.

Brayer,

"What

Britain

is

Doing

to

Per-

fect

the

Electronic

Office,"

American

Business,

V 24,

#7

(July,

1954), pp. 16-19, 46-48

A

discussion

of

British

developments

in

awro-

matic computers.

Herbert

O.

Brayer, "French

Electronic

Developments

Speeds Systems," American Business, V 24, #8 (Aug.

1954), pp. 20-24, 40-41

A

discussion

of

French developments

in

auto-

matic computers.

Louis

N.

Ridenour, "Mechnaical

Brains,"

Fortune,

V 39,

1t5

(May,

1949), pp. 109-110, 112, 114, 116-

118

One

of the very

earliest

major

articles

on

automatic computers.

W.W.

Stifler,

Editor,

High

Speed Computing Devi-

vices

(New

York:

McGraw-Hill

Book

Co.,

Inc.,

1950),

440 pp.

A

basic

but

technical

background

reference

on

the

technical

aspects

of automatic compu-

ter

operations

and

hardware.

General Reference Information

C.L.

Peterson,

Automation

Dictionary,

(Philadel-

phia,

Pa.:

Minneapolis-Hone~lIell

Regulator Co.,

1954) 19 pp. Also

in

Advanced

Management, V

20

#7

(July,

1955), pp. 30-33

An

easy

to

understand

glossary.

Robert

Serrell,

et

al.,

Definitions

of

Electronic

Computer

Terms

(New

York:

Institute

for

Radi,o

En-

gineers,

Inc.,

1950), 6 pp.

One

of

the

standard

glossaries.

Staff

of Haskins

and

Sells,

Data Processing

by

Electronics

(Nelv

York: Haskins

and

Sells,

1955),

113 pp.

A simple and

general

introduction

that

inclu-

des a

good

survey

of

available

computers.

John Diebold,

Herbert

F. Klingman's Digest of

Au-

tomation

(Nel\,

York:

Controller's

Institute,

1954),

16

pp.

A condensation

of

a

leading

book

on

automatWn.

-

20

-

Norbert Weiner,

The

Human

Use

of

Human

Beings

(New

YorK:

Doubleday

AnChor

'Boob.

1<)54~~.

199_

pp.

A rambling but rewardingly

insightful

back-

ground book.

Andrew

D.

and Kathleen

H.V.

Booth, Automatic

Digi-

tal

Calculators

(New

York:

Academic

Press,

Inc.,

1953),

231

pp.

A

technical

reference

on

automatic computers.

Edmund

C.

Berkeley, Giant Brains

(New

York: John

Wiley & Sons,

Inc.,

1949),

270

pp.

Discusses

some

of

the

earlier

automatic com-

puters.

Francis

J.

Murray, Theory

of

Mathematical

Machin~

(Morningside Heights,

Ne1"

York:

Kings

Crown

Press,

1947), 116 pp.

Some

good

explanation

of

computer, components.

M.E.

Maron,

"Logic, Discovery,

and

the

Foundations

of

Computing Machinery," Transactions

of

the

Insti-

tute

of Radio Engineers

Professional

Group

on

Elec-

tronic

Computers. V EC-3,

#2

(June, 1954), pp.

2-7

A background

article

on

automatic computers.

Edi

tor,

"The

Brain

Builders,

'!

Time,

V 65,

#13

(March

28, 1954), pp. 81-85

A

picture

of

the

IBM-organization.

Cecil

Hastings,

Approximations

for

Digital

Computezs

(Princeton:

Princeton

University

Press,

1955),

201

pp.

A mathematical

reference

book

of

value

to

pnr

grammers.

Warren P.

Livingston,

"Punched Card System

in

the

Modern

Office,"

Office

Management

and Equipment.

V 12, # 4

(April,

1951), pp. 30-32, 64; V 12, # 5

(May,

1951), pp. 33-36; V 12, # 6 (June, 1951),

pp. 24-27

An

introductory

and

very

general

discussion

of

punched

card

equipment.

F.

M.

Johnson, "Control

of

MachinetAcc 0 u n

ting

Equipment," Systems

and

Procedures

Quarterly,

V

4,

# 2

(May,

1953), pp. 18-22,

26

To

control,

first

measure

the

degree

of

mach-

ine

utilization.

Warren P.

Livingston,

"Punched Card Systems

at

the

Management

Level,"

Office

Management

and

Equip-

ment, V 14, # 8

(Aug.

1953), pp. 21-23, 68, 73-74,

77

Robert

A.

Fernsler,

"How

to

Determine

Effective-

ness

of

a Punched

Card.

Installation,"

The

Office,

V 38, # 4 (Oct. 1953), pp. 67-69, 177-187

Discusses

how

to

evaluate

the

record

of

op-

erations

,of a punched

card

installation.

Martin

Gardne_r,

"Logic Machines,"

Scientific

Amer-

ican,

V 186, # 3 (March,

1952)"J

pp. 68-73'

Supplies

background

on

logic

compute~st

not

numeric

digital

computers.

Edmund

C.

Berkeley, "Simple Simon,"

Scientific

American, V 183, #

5,

(Nov.

1950), pp. 40-43

A

description

of a simple

robot-like

machine.

C.

M.

Weaver,

Check

OperationS (Chicago:

First

National

Bank

of

Chicago, 1953),

24

pp.

Nothing

directly

on

automatic computers.

Argument

on

Can

Machines Think

Samuel

A.

Ross,

"Letter,"

Scientific

American, VD3,

# 1

(July,

1955), pp. 2, 4, 6

Selections

from

Turing's

article

in

Mind.

-

END

-

AUTOMATIC MIXUP

LAWRENCE

M.

CLARK

Cambridge,

~1ass.

I.

Andrew

Dunster,

Chief

Consulting

Engineer

of

Natrasys,

lvalked

back

into

his

little

glass-

partitioned

office,

leaving

behind

the

sounds

but

not

the

sights

of

the

great

Central

Control

room.

He

was a

tall

man,

in

his

early

forties,

a

furry

hairy

man

with

thick

black

hair

and

a

face

that

often

smiled.

It

t'vas

not

smiling

now. A

great

pile

of

already

opened

telegrams

tvas on

his

desk.

He

picked

up

incoming

serial

14

and

reread

it:

CENTRAL

CONTROL,

NATRASYS,

NBV

CHICAGO

-

CARGOLINER

M239

OUT

OF

SAINT

LOUIS

FOR

LOS

ANGELES

ARRI

VED

SEATILE

I N

ERR

0 R.

REDISPATCHED.

WHY

CAN'T

YOU

MAKE

YOU R

XXX

SYSTEM

WORK?

--

J.

HUNTER,

SEATILE,

1215,

10

SEPTEMBER

He

picked

up

another,

serial

57,

and

re-

read

it:

CENTRAL

CONTROL

OFFICE,

NATIONAL

TRAF-

FIC

MANAG~lENT

SYSTEM,

NEW

CHICAGO

-

SPECIAL

DELIVERY

URGENT

MISSILE G2418

OUT

OF

NEW

CHICAGO

FOR

OZARKTON

MO.

WIlli

NO-

VOPOX

VACCINE

ARRIVED

PROVIDENCE

R.I.

IN

ERROR.

REDISPATCHED

AT

ONCE.

R

EQU

EST

CAREFUL

CHECK

EQUIPMENT.

--

V.

DONNE£L,

PROVIDENCE,

1309,

10

SEPTEMBER

The

door

opened,

and

Jim

Hat'V'ker,

Engineer-

in-Charge

of

Natrasys,

came

in,

a

short,

active,

ldry

little

man,

about

thirty.

Dunster

turned

half

around

from

his

desk

tOt\'ards Hatvker,

smilErl

faintly,

and

said,

"Jim,

how

do

we

stand

no

w

--

what

are

we

finding

out?"

"Professor,

Crew A

has

tested

all

th

e

cables,

and

about

3/4

of

the

panel

c

has

s

is.

Crew B

has

tested

about

60%

of

the

tubes,

ti-

taniodes,

and

siliciodes.

Crew C

is

checking

insulation

and

is

about

5~/o

through.

Crew D

has

checked

the

timing

of

all

pulses,

and

~ut

8~/o

of

the

magnetic

and

electrodynamic

memor-

ies.

Crew E

•••••

"

"Yes,

yes,"

said

Dunster,

smiling,

"but

what

are

we

finding

out?"

"Hell,

Doc,

that's

just

the

trouble.

Here

and

there

we

find

some

stuff

that's

a

little

below

par.

But

we're

finding

nothing,

nothing

that

can

explain

all

this

blazing

mess,"

and

Jim

Hawker waved

at

the

stack

of

telegrams.

"What

about

the

trouble-shooting

tapes?",

said

Dunster.

"We've

run

every

Josophat

one

of

'em twice,

Professor.

Results

--

nothing.

What's

making

our

machine

go wrong

is

intermittent.

H 0

tV

I

hate

intermi

ttent

bugs,

Josophat,

hdw

I

hat

e

intermittent

bugs!"

"Yes,

Jim,"

s

aid

Duns

ter

smiling,

"and

how

I

wish

that

hating

bugs

got

rid

of

them!"

But

before

he

had

finished,

his

mind had

detache

d

from

what

he

was

saying

and

had

gone

puzz

1

e-

sol

ving.

He

was

looking

unseeingly

at

the

great

rml1S

of

panels,

signal

lights,

and

equipment

on

the

other

side

of

the

glass,

filling

a room

that

was

nearly

200

feet

square.

Dunster

pointed

at

the

heap

of

telegrams.

"There

are

382

complaints

there,

up

to

1630

this

afternoon.

They

began

to

come

in

around

1100."

Jim

Hawker

opened

his

mouth,

but

did

not

say

anything,

and

shut

it

again.

Duns

ter

said:

"Jim,

what

does the

Analysis

Section

have

to

say

about

the

trends

shown

by

these

complaints?"

"I

don't

know

Doc

I'll

find

out."

Jim

Hawker

went

out

of

the

office.

The

telephone

rang.

Duns

ter

picked

it

up.

"Andy

Dunster?"

"Speaking."

"Harry

Eberhard,

Andy. Have you

located

the

trouble

yet?"

"Not

yet,

Harry."

"Well,

I'm

coming

out

to

see

you

righ

t

away."

"OK,

we'll

be

here."

Dunster

could

tell

that

Eberhard

was

the

opposite

of

delighted.

Dunster

sat

dOtvn

at

his

desk,

and

his

thoughts

went

back

to

his

first

conversatio

n

with

Eberhard

six

years

before.

-21 -

AUTOMATIC

MIXUP

II.

..

.'

a

scheduling

service.

It,

will

reach

right

down

In

June,

1986,

Harry

Eberhard,

grizzled,

.

short,

in

his

forties,

had

entered

Du

ns

te

r ' s

Computation

Laboratory

at

Agamek

University

for

the

first

time.

He

came

in

like

a

fresh

breeze

, blOl'ving

papers

around

a room.

Wi

th

a big,

smil~

and

rubbing

his

hands,

he

said:

"Professor

Duns

ter,

my

name's

Eber

h a r d ,

president

of

the

Sagamore

Railway.

I'veheard

you're

interested

in

mechanical'

brains

for

auto-

matic

traffic

control.

I came

dOl'in

to

A

gam

e k

this

week on

account

of

my

son's

gra

d u

ati

0 n ,

and I

thought

I'd

drop

over

and

take

a

'100

k

around

your

laboratory,

if

I may.

We

ought

to

have some

interests

in

common."

"I'll

be

delighted

to

show you

around,

Mr.

Eberhard,"

said

Dunster,

"But

suppose

tve

call

the

equipment

t'ie're

working

with

'a

u

to

m a

tic

machinery

for

handling

information

intelligent-

ly'

instead

of

mechanical

brains.

Some

people

still

object

to

that

phrase."

So

he

had

shOl'vn

Eberhard

around

his

1 a b -

oratory,

and

then

at

the

end,

back

in

his

study

Eberhard,

rubbing

his

hands and

smiling,

had

said

to

him:

"Professor,

by

gosh,

what

you're

doing

is

important.

Your

miniature

automatic

traffic

controller

is

going

to

revolutionize

the

hand-

ling

of

traffic

in

this

country.

I

can

see

it

all.

Just

think

--

no

more

wasted

space

--

no

more

traffic

jams

--

lOt'ier

prices,

in

fact,

real-

ly

flexible

prices.

Why,

the

sky's

the

limit

on What

we

can

do.

Human

beings

cannot

k

now

enough

to

handle

20

trillion

shipments

a

year

in

the

best

possible

way. But a machine

can.

,Why,

it's

terrific:"

Dunster

said:

"Wai

t a

minute,

Mr. E

be

r-

hard.

All

lve're

trying

to

do

is

to

shOt"

the

feasibility

of

certain

steps

fo~vard.

There's

always

going

to

be some IllJasted

space

and

time.

But

it's

true

that

l'ie

can

cut

down

the

waste

a

long,

long

l'Vay."

Later,

over

s C 0

tch

and

sodas

at

the

fac-

ulty

club,

before

they

separated,

Eberhard

had

said:

"Andy,

we

could

go

places

--

your

ideas,

my

business

experience.

We'd

make

qui

te

a team."

On

a

moon-lit

summer

evening

three

weeks

later,

Eberhard

was

back,

smiling

broadly

and

rubbing

his

hands.

He

came

into

IJ.mster's

study

in

the

laboratory,

fell

into

a deep

black

leath-

er

armchair

and

lighted

a

cigarette.

"Andy, I have

been

talking

lVith

quite

a

fel'l

businessmen,

and some

of

them

can

see

wh'at

I

can

see.

We

t'lant

to

organize

something

--

we

might

call

it

National

Traffic

Management

System,

Natrasys.

We

will

provide

a

service,'

into

the

automatic

loading

and

routing

of

vehic-

ul

ar

sp

ace

--

whether

rail,

air,

truck,

or

ship!'

Duns

ter

puffed

on

his

briar

pipe,

rubb

e d

his

thumb and

forefinger

reflectively

on

the

thick

stubble

of

his

evening

beard,and

waited.

Eberhard

went on

with

steadily

increasing

enthusiasm,

"We

shall

traffic-schedule

eve

r y

vehicle

that

can

hold

freight.

We

shall

begin

around

New

Chicago,

and expand

gradually,

but

as

fas

t as

we

can."

Eberhard

paused.

He

got

up.

He

looke

d

straight

at

Dunster.

"I

t'iant you

to

be

in

char~

of

making

this

thing

come

into

existence.

You've

got

the

interest,

the

intelligence,

the

trainll~

You're

an

engineer.

~

have

visions

but

I

don't

have

the

training."

Dunster

puffed

on

his

pipe

again

before

re-

plying.

"Harry,

your

enthusiasm

is

wonder

ful

and

your

belief

in

my'

abili

ties

is

very

flat

ter-

ing.

But I

want

to

be a

scientist,

Harry;

thad;

what

I'm

cut

out

to

be.

My

mission

is

to

d i

s-

cover

knowledge,

to

search

for

truth,

to

train

students,

to

open windows

for

people

who

want

to

learn.

I

shall

be a

fish

out

of

w

ate

r

in

that

scheme

you're

thinking

of."

Eberhard

sat

down

again

and

leaned

forward.

"Yes, Andy, I

knOt".

But I

can

promise' you we

shan't

require

you

indefinitely.

But wi

th

0 u t

you

we

can't

even

start."

He

stopped

and

then

went

on

again.

"Besides

your

ability

in

the

field

of

automatic

machinery

for

handling

in-

formation

__

fl_

Eberhard

grinned,

"I'll

try

not

to

call

them

mechanical

brains

--

look

tvh

at

you've

accomplished.

You

started

out

five

years

ago

at

Agamek,

just

an

instructor.

You

were

no

professor.

You

had,no

laboratory.

You

had

no

bqilding.

You

had

not

written

any

books.

"Now

where

are

you?

You're

a

full

profes-

sor.

You

have

your

own

laboratory

in

your

own

building

--

oh,

I

know

legally

the

building

be-

longs

to

the

Universi

ty

but

actually.

it's

yours,

so

long

as you

aren't

convicted

of

moral

turpi-

tude.

And

you've

wri

tten

two

books.

And

you

have

a

department

with

three

instructors

work-

ing

under

you.

And

in

your

lab,

you

now

have

a·miniature

automatic

traffic

handling

system,

probably

the

best

model

for

such

a

system

in

the

United

States.

"We

need you

not

only

for

your

scientific

knmdedge,

Andy.

We

need you

for

your

per

s e -

verance,

and

your

intolerance

of

obstacles

--

the

kind

of

special

intolerance

that

gets

them

out

of

the

way, and

yet

wonderfully

en

0 u g

hI.

leaves

people

happy.

HOt"

about

joining

'us?"

Dunster

said,

t'tti

th

as

much

finality

as

he

could

manage

to

put

into

his

voice:

"I'm

sorry.

-22 -

AUTO\1ATIr.

MIXUP

I

have

to

say

no."

Eberhard

just

leaned

a

little

further

for-

ward:

"Look, Andy,

you'd

give

us a

lot

of

time

and

help

--

for

nothing,

just

for

our

asking

--

you're

that

kind

of

a

person.

You'd

give

u s

that

anyh01'1.

What you

don't

l'1ant

is

to

give

up

your

laboratory,

and

give

up

all

your

time.

Well,

we'll

not

ask

that.

Suppose

we

build

the

head-

quarters

of

the

National

Traffic

Management Sys-

tem on

that

land

out

there,

right

your

laboratory."

Eberhard

pointed

through

the

win-

dow

northlltards,

into

the

moonlit

evening.

"Over

there

on

the

other

side,"

Eberhard

pointed

to

the

southwest,

"will

be

the

new

Phys-

ics

Center

that

Agamek

is

planning

to

build,

the

Radiation

Laboratory,

the

Relativity

Laboratory,

the

Nucleonics

Laboratory,

and

the

rest

of

them.

Your

Computation

Laboratory,

and

the

one

a·n

d

only

Natrasys

will

be

a

fi

tting

accompan i men t

to

them.

"We'll

make

an

arrangement

ll1i

th

A

gam

e

k.

YouJll

be

right

here

in

your

lab,

part

of

the

time,

and

in

Natrasys,

part

of

th

e

time.

Your

connection

Hith

Natrasys

will

give

you

far

more

funds,

to

develop

your

lab,

and

to

steam

ahead

1\'1

th,

than

you

could

get

in

any

otIer

way. Even

today,

scientific

investigation

do~

not

get

all

the

support

it

could

use.

You

knmv

you

have

to

raise

more money, more endolltment,

to

keep

pushing

the

things

you've

begun~

Natrasys

will

solve

that

unsolved

problem

of

yours

--

lvill

perhaps

s ave you

quite

a

slug

of

time

in

the

long

run."

It

was

Dunster's

turn

to

get

up

and

walk

restlessly

around

the

room. Then

he

came

back

and

said,

"Harry,

you're

very

persuasive.

Let

me

think

it

over."

Eberhard

said,

"Andy, you

don't

need

to

think

it

over.

Say

yes

no\\(.

If

you

say

yes,

you

lvill

be

doing

something

that

t'llill be

very

useful

for

great

numbers

of

people

--

ironing

out

the

shipping

and

traffic

jams

that

are

gum-

ming up

so

much

of

this

society

of

ours.

In

addition,

you'll

have

a

big

system

instead

of

a baby one

to

play

wi

th

--

I

should

say,

to

tvork

Id

th.

Bes

ides

the

prac

tic

al

problems

of

making

a

great

artificial

thinking

system

work 0

ugh

t

to

be

of

the

highest

scientific

interest

to

you."

Dunster

said

laughingly:

"You

certainly

kn01V'

hOH

to

pers\1ade

people,

Harry."

He

s t

op-

pede Then he

said

qui

te

seriously:

"I'm

going

to

say

yes

after

all,

because

your

arguments

are

good

ones.

But

let

me

warn you, Harry'

--

there

are

going

to

be

qui

te

a number

of

very

r 0

ug

h

places.

"

A f

et"l

Ilteeks 1 a

ter,

Duns

ter

took

the

s

ubur-

ban

train

into

New

Chicago,

and met

Eberhard

in

his

office

marked

"President",

on

the

20th

floor

in

the

Sagamore Railway

Building.

Then

the

y

Ivent

together

up

to

the

Board

of

Direc

tors

Roan

on

the

35th

floor,

a

spacious

room

paneled

with

a

great

plate

glass

window,

looking

over

the

green

parks

and

clean

white

buildings

of

Ne

w

Chicago,

the

sunlight

flashing

on

the

water

of

Lake

Michigan.

After

Dunster

had

perceived

the

room,

the

window, and

the

view,

he

noticed

a

tall

stou

i

blond

man,

perfectly

dressed,

talking

ina

loud

and

raucous

voice

to

half

a dozen

other

men.

Eberhard

politely

waited,

then

introduced

Dun-

ster

to

the

stout

man

saying,

"Mr.

Valer

ia

n

Crurmvell

--

Professor

Andrew

Dunster."

Crumwell

said,

in

the

same

loud

and

pen

e-

tratirrg

voice,

"So you

are

the

man

with

the

big

idea!

Glad

to

meet

you,

Professor

Duns

ter

,"

so

me",

hat

prolonging

the

I110rd

Dunster.

as

if

he,

were

having

a

great

deal

of

fun

with

it.

Dun-

ster

was

nettled,

and was

displeased

with

him-

self

for

being

irri

tated

over

so

trivial

a thing.

lIe had

heard

of

Crurm\lell,

extremely

rich,

clever:

moody, a

man

interested

in

the

early

stages

of

new

ideas,

but

when some

of

the

stages

wen

t

Ivrong, a

man

litho

became

sour

and

vicious.

All

this

registered

in

his

mind

in

a

fraction

of

a

second.

Almost

without

a

pause

after

Crurmvell's

remark,

Dunster

said,

"Glad

to

meet you,

Mr.

Crumwell."

Eberhard

was

addressing

the

group:

"Gentle-

men,

all

of

us

are

here

now.

Let's

beg

in."

They

selected

chairs

and

sat

at

the

table;

Eber-

hard

said:

"Gentlemen,

I

believe

tve

have

a

r~

markable

opportunity

to

get

in

on

the

ground

floor

of

an amazing

new

scientific

development.

As

I've

said

to

you

individually,

and

want

to

say

once more

to

you

n0

1\1,

I'm

convinced

we're

on

the

threshold

of

completely

automatic

hand-

ling

of

traffic

and

shipments

-

railroad,

truck,

air,

ship.

We

can

create

a

scheduling

and

mr-

ecting

service

that

will

be

far

superior

10

what

this

world

has

ever

seen."

Eberhard

lV'ent

on wi

th

many

remarks

and

statements

in

the

same

vein,

Ivhich

Dunster

had

heard

before,

and which

he

was

sure

the

others

had

heard

before,

but

he

fel

t

certain

that

Eber-

hard

knew

li1hat

he was

doing,

that

he

14/as

seek-

ing

to

create

a

frame

of

mind

in

tvhich

he

tltould

get

the

yeses

that

he

wanted

from

the

men

pre-

sent.

Eberhard

came

to

a

point;

"You've

heard

me

say

many

of

these

things

before,

but

I m no

authority.

My

field

is

not

the

engineering

of

traffic

control.

But

we

do

have

here

in

P r

0-

fessor

Dunster

a

man

who

is

an

authority.

Pro-

fessor

Andret\'

Dunster

of

Agamek

University

is

world-famous

in

the

field

of

automatic

.machin-

ery

fo~

handling

information.

His

specialty

~

equipment

for

automatic

traffic

control.

H e

has

agreed

to

be

our

chief

consulting

engineer.

-23 -

AUTOMATIC

MIXUP

Do

any

of

you want

to

ask,

questions

of

Profes-

tvell, you say what they

all

say.

Eberhard,

does

sor

Dunster?"

this

man

know

tvhat he

is

talking

about?"

"Yes," spoke

up

CrulIllvell.

He

turned

to

Dunster. "Mr.

Dunster,"

and again he

slurred

the

name

in

a

derisive

way,

"Mr.

Dunster,

what

evidence

do

you have

that

this

old

scheme

of

Harry's

tvill

Ivork?

and

that

we

tvill

make

money

out

of

it

instead

of

losing

our

shirts?"

Dunster

sat

up

a

little

straighter:

''Mr.

CrulIll'lell, I

don't

tvant

to

bore you

wi

th a

lot

of

scientific

details,

but

let

me

give

you

some

of

the

more

important

part

of

the

evidence.

"Automatic machinery

for

handling

infor-

mation

is

the

heart

of

automatic

traffic

con-

trol.

For

this

kind of machinery

to

tvork

well,

you have

to

be

able

to

do

four

things,

do

them

quickly,

and

do

them

reliably."

"One,

you

have

to

have a

t'Vay

of

putt

i n g

shipping

and

traffic

information

into

an

auto-

matic machine; two,

you

have

to

have a

1>vay

of

remembering

it;

three,

a

way

of

operating

tdth

it;

and

fourth,

a

way

of

getting

that

informa-

tion

out of

the

machine,

fast

enough

and

reli-

ably

enough so as

to

direct

shipments and

traf-

fic.

All

of

these

processes

have

now

had over

40

years

of development.

The

reliability

and

speed

in

these

devices

is

now

uptvards of 100,000

times

that

of

a

human

being.

"For example,

tve

now

have Equipment t

hat

uses

pulses

spaced one hundredth

of

ami

c r

0-

second

apart

--

that

is,

a hundredth of a

mil-

lionth

of

a second. These have been

standard

operating

pulses

for

computing equipment

fo

r

the

last

20

years.

Here

is

the

speed

Ive

tV'ant

and

need

for

automatic

traffic

control.

"Let

me

give

you

one

more

example.

tv

e

have almost

unlimited

information

storage

cap-

acity,

magnetic books,

sheets

of

magnetic

sur-

faces

on

which

information

is

stored

as

an

ar-

rangement

of

magnetized

spots.

So

information

can

be

stored

at

the

rate

of

a

billion

decimru

digits

in

a cubic

foot

,of

space.

We

can read

any

desired

item

of

information

in

about

ahu~

dredth

of a second. This

also

has been demon-

strated

to

be

reliable

over

th~

last

20

years.

"Mr.

CrulIllV'ell,

there

is

no

doub.t Ivhatever

of

the

capacity

of

science

to

build

and

oper-

ate

automatic

traffic

control

equipment which

is

fast

and

reliable.

It

will

be

much

faster

and

more

reliable

than

human

beings.

All

that

t'll'e

have

to

do

is

to

put

together

es

tablis

h e d

and proved

devices,

to

solve

a problem

that

is

becoming

more

and

more

important.

But

t~

must

not

forget

that

some

errors,

some

pieces

of

trouble,

are

inevitable,

no

matter

how

excel-

lent

the

equipment.

CrulIllvell

said

in

his

loud

voice,

"Yeah,

Eberhard

said,

"I

believe

he

does."

Crumwell

said:

"How

many

of

the

rest

of

you

think

so?"

Six

of

the

other

eight

men

sai:l,

"Yes, I

do."

Crumwell

said,

"Sam,

you

did

not

say

yes--

~vhat

do

you

think?"

Eberh ard

tvh

is

pered

to

Duns

ter:

"Th

at's

Sam

Oliver,

head of

the

2nd

International

Ban~

Netv

Chicago

office."

Sam

Oliver

said,

"What

I

know

about llinstEr

is

hearsay,

t'll'hich

hOlvever

agrees

in

reporting

Dunster

to

be

an

expert.

But I

know

Eberhard,

and

if

Eberhard

is

tdlling

to

give

adequate

se-

curity,

I

think

we

can

go

ahead with

the

first

installment.

After

all,

if

we

begin with

New

Chicago and

environs,

and

make

a success of

th

at,

then

we

can spread

out."

CrulIllvell

got

up:

"OK,

Eberhard,

let's

be-

gin

ld

th your Natrasys

and

I

tvill

loan t t1 0 -

thirds

of

the

commitment

for

the

first

year

am

then

we

tvill

see."

Eberhard

got

up,

smiling

and rubbing

his

hands: "Thank you,

Mr.

Crurmvell. We're

all

going

to

do

tvell."

Crurm.vell

nodded goodbye

and

went

out

in

haste.

The

meeting broke up. Dunster

said

t 0

Eberhard:

"How

tvell

do

you

know

CrulIll'lell?"

"Andy,

not too

well,

but

well

enough.

When

he

thinks

he can

make

a

dollar,

and

nothing

has

yet

gone tvrong, he

is

OK.

Once

we

get

started,

we'll

run

on

our

own

steam."

"Harry, I

do

hope

you're

right.

I

don't

like

him. Besides I

don't

think

he even heard

my

remark about

errors."

"Nobody

really

likes

him,

Andy,

but

he's

useful.

tt

The

first

year

had gone unexpectedly

well,

but

the

team of Eberhard

and

Duns

ter

did

'n

0 t

finish

tv!

th

needing

Mr.

Valerian

CrulIl'tTell.

The

mora

Natrasys

grel'\1'

,

the

larger

became

CrullMell's

investment,

less

in

proportion

but

greater

in

amount.

And

as usual

he

required

ample

secur-

i

ty,

in

this

case

an

extensive

mortgage

on

Eoor-

hard's

Sagamore Railway.

It

took

all

the

spell-

binding

that

Eberhard could muster over

his

stockholders

to

keep

that

going.

For

those

first

six

years,

that

were

now

over, Dunster, with

Eberhard's

staunch backing

--

tvhen

he

was

not

distracted

--I

and

wi

th a

fair

amount

of

Crumwell's

money,

had

tv

0

rk

e d

-

24

-

long

hours

lvi

th

a

fine

crew

of

roon.

AUTOMATIC

MIXUP

They

had

built

the

principles

of

automatic

control

of

traffic

into

intelligent

machinery.

From

the

first,

the

automatic

traffic

manage-

ment

service

was a

success,

and

it

had

spread

rapidly

in

a

wider

and

wider

network

0 u t 0 f

Nel'V

Chicago.

On

the

campus

of

Agamek

Univer-

sity,

MacNeil's

Computation

Laboratory,

the

great

Central

Control

office

of

Na

t r a

sy

s

occupied

the

second

floor

of

the

Natrasys

BWUd-

ing.

Panels

of

the

equipment

overflowed

into

the

third

and

fourth

floors,

ld

th

thick

c 0

n-

nee

ting

cables.

The

machine

-

if

i

nd

e e d

it

could

be

called

"a"

machine

--

was

quite

prob-

ably

one

of

the

most

intricate

and

intelligent

assemblies

of

automatic

control

machinery

yet

in

existence.

.

There

had

been

troubles

from

time

to

time.

Usually,

Valerian

Crumvell

found

out

about

the

troub~~

too,

and-then

he

raised-~

storm,

each

storm

~orst

than

the

last

one.

Once

there

was a

connecting

wire

that

was

slightly

loose,

and

swinging.

Every

now

and

then

it

l'lould

stdng

against

a

terminal

t

hat

contained

pulses

1'lith

the

value

6.

Fin

ally,

the

insulation

on

that

connecting

wi

r e

wore

through

at

that

point,

and

so

every

now

and

then

a 6

liould

corne

into

some

calculation

where

i t

had

no

business

at

all.

If

it

had

not

been

for

a

caravan

of

ar~

trucks

that

provided

a

con-

s

iderable

amount

of

ground-shaking

for

a

hal

f

hour

one

morning,

Dunster

wondered

how

they

ever

would

have

found

out

about

that

trouble.

Another

time,

tvhen

the

system

was s

till

small

and

the

disruption

l\'as

minor,

there

had

been

extensive

random

misbehavior,

completely

unexplained_Jor

about

a

day.

Then

Dunste

r

correlated

it

with

the

aurora

borealis

and

an

unusually

pOlverful

soiar

magnetic

storm.

H e

had

prevented

similar

trouble

in

the.future

by

appropriate

shielding

of

sensitive

parts

of

the

equipment.

Another

trouble

occurred

when a

new

sec-

tion

of

Natrasys

equipment

t'V'as

buil

t

in

Indi-

anapolis.

This

trouble

however

tl/as a

"s

af

e "

one

--

it

never

permi

t

ted

the

equipment

to

work

right

even

once

until

it

was removed.

It

turned

out

tpat

one

of

the

young

ladies

lfielding

a

sold~ting

iron

had

disliked

the

red-yellolv-red

color-coded

resistors

because

of

her

anti-Spa~

ish

feelings,

and had

put

in

p a

tr

i 0

tic

red-

lihi

te-blue

resistors.

Once,

the

new

Radiation

Laboratory

to

the

southh'est

had

been

trying

out

some

netv

radia-

tion

apparatus.

Extra

pulses

were

1

ib

er

all

y

sprinkled

all

through

the

Central

Cont

r 0 1

Of-

fice.

Dunster

guessed

this

one

almost

at

once

because

he

had

been

thinking

something

like

this

might

readily

occur.

lIe

had

liquidated

thi

s

trouble

by

installing

lead

shielding

in

the

Radiation

Laboratory

and

in

the

Natrasys

Build-

ing.

As

a

regular

routine

monthly

test,

Dunster

had

arranged

that

a

Natrasys

test

crew

would

hunt

for

the

nelvest

radiation

apparatus

in

any

of

the

nearby

Radiation

and

Physics

Laboratory

buildings

once

a month. They would

lift

upthe

shielding

around

the

apparatus,

direct

a

beam

of

radiation

at

the

Natrasys

building,

and

test

to

see

if

any

pulses

got

through

Nat

r

as

y s

shielding.

The

last

thirteen

monthly

testshad

been

pulse-proof.

As

a

general

measure

to

forestall

trouble,

Dunster

had

considered

duplicating

the

eqWpmem

in

Natrasys,

and

requiring

that

both

of

two twin

pieces

of

equipment

agree

on

the

same

informa-

tion

before

it

was

released

to

some

other

part

of

the

system

to

use.

This

was

still

too

ex-

pensive

for

Natrasys

so

far.

But

Dunster

had

installed

another

protec-

tion

throughout

Natrasys.

This

was

that

Natra-·

sys

would

carry

along

a

check

on any

piece

of

information

equal

to

the

sum

of

the

digits

and

letters

(letters

having

the

values

A

1,

B 2 up

to

Z

26).

For

example,

piece

of

informa

tion

Y2254

would

have

associated

with

it

as

a

check,

the

number

38,

namely

the

sum

of

25,

2,

5,

and

4.

Any

change

in

a

single

character

would

alter

the

sum,

although

a

pair

of

compensating

errors,

such

as

an

interchange

of

two

digits,

would

not

be

detected

by

this

protection.

This

had worked

remarkably

well,

and

enabled

auto-

matic

elimination

of

a

large

number

of

errors.

III.

TIle

telephone

rang,

and

Dunster

abruptly

returned

to

the

present.

He

picked

up

the

re-

ceiver.

"Professor

Dunster",

said

the

g u a r

d's

voice,

"Mr. El>erhard

has

arrived,

and

is

coming

at

once

to

your

office."

As

Dunster

put

the

receiver

down, Eberhrurl

opened

the

door

to

the

off

ice,

and

cam

e

in.

Dunster

saw

that

Eberhard's

complexion

was

a

Ii

ttle

ruddier

than

usual,

and

the

r e

1'1

a s

no

smile

on

his

face.

Dunster

concluded

that

Eberhard

had had

a

couple

of

drinks,

perhaps

more.

"Andy,

old

pal,"

s

aid

Eberhard,

"this

j

at

mess

has

got

to

stop."

But

the

tone

MacNei 1

listened

to

was

even

less

friendly

than

the

l"lords.

"Harry,

old

pal,"

said

Dunster

even

I y

"you

can't

l\fant

that

any

worse

than

I

do."

"Well,

111hy

don't

you

stop

it?"

-25 -

AUTOMATIC

"As

a

matter

of

fact,"

said

Dunster,

"i

t -

probably

has

stopped

--

for

now."

"What,"

said

Eberhard,

"you

lOOan

the troub:E

in

Natrasys

has

stopped?"

"Yes, "

said

Dunster,

smiling,

"the

f

re-

quency

of

complaints

t"8S

slowing

down

from 1600

to

1630,

and

after

1630 I

think

only

0

ne

has

come

in."

"Well,"

said

Eberhard,

"so

you

fixed

the

trouble,"

and he began

to

relax

and

smile.

"I

! "No,

Ive

didn't

detect

the

trouble.

We

had

all

our

inspection

crews

testing

the

equipment,

and

t~en

the

trouble

just

ceased,

died

a

way.

Furthe1rmore,

"since

the

trouble

is

no

Ion

g e r

manifes

ting

itself,

I

doubt

that

we

can

f

iii

d

it

by

testing,

for

all

the

equipment

is

again

functioning

perfectly."

"Yes,"

said

Eberhard

reflectively,

"I

can

see

that,

but

I

feel

as

if

I

were

in

a

night-

mare.

"

"We

can

how'ever examine and

analyse

the

383

complaints,

and

see

what

sort

of

thing

they

indicate."

"Well,

it's

your

responsibility,

Andy."

"You're

right,

Harry"

said

Dunster

delib-

erately

and

looking

at

Eberhard

intently.

Eberhard

went on, "CrulJlfell was

bothering

me

the

other

day.

He

said

he

Ivas

tired

of

mmk-

eying

around

Id

th

us.

We

had

rolled

up

a

fine

record,

he

said

sarcastically,

of

eleven

pieces

of

trouble

in

six

years.

We

had

lost

something

like

a

quarter

of

a

million

dollars

as

a

result,

he

said,

and he

Ivas

tired,

jat

tired

of

us.

He

s

aid

one more

piece

of

trouble

and

hew

a s

through,

j

at

it,

through."

Eberhard

add

ed,

"I

hope CrufIl.\,ell

doesn't

hear

of

the

trouble

today."

Dunster

said:

"But

he's

bound

to,

Harry.

This

is

the-l-vorst

piece

of

trouble

we've

ever

had.

I

think

we'll

be

sued

for

something

like

six

million

dollars,

estimating

an

average

0 f

$20,000

damage

for

each

of

300

cases."

"Oh,

jat,

jat, jat,

jat,"

said

Eberhard

feebly.

Dunster

smiled

again,

"Harry,

you may

as

l\'ell

set

to

lwrk

thinking

of

a

better

answer

than

that

for

your

friend

Crulllvell when

he

gets

on

the

phone tomorrow

about

this."

"My

friend!"

said

Eberhard,

"the

hell

you

say.

lIe's

no

friend

of

mine.

IldshI'dnever

seen

him."

-26 -

\fIXUP

, "No,

not

really,

because

Crurmvell gave

us

the

start

of

Natrasys,"

said

Dunster.

"You

can't

make

great

advances

without

money

to

in-

vest

--

at

least

you

can't

often.

Crumwell

is

still

supplying

most'

of

our

working

capitaL"

Eberhard

got

up,

paced

to

and

fro

in

the

room, and tvrung

his

hands.

Dunster

watc

h e d

him.

Then he

got

up,

put

his

ar,m

around

Eber-

hard's

shoulders,

and

said,

"Si

t

d01~n,

Harry;

take

it

easy;

we're

a

long

way

from

being

licked."

Eberhard

sat

down,

stretched

himself

out,

held

his

head

in

his

hands

as

if

it

were

aching

and

sagged

like

a

pillow

with

no

feathers.

Jim

Halvker came

back

into

the

room.

"Pro-

fessor,

here

is

your

analysis.

They

promised,

it

to

me

every

felf

minutes,

and

so

I w

ai

ted.

But

Josophat,

Professor,

the

tdres,

the

co

m-

plaints,

have

stopped

coming

in.

Wha'd'yo

u

know?"

Dunster

reached

for

the

analysis,

too

k

it,

sat

dOtvn

at

his

desk,

and

Ivas

los

t

in

con-

centration.

Jim

Hawker

said,

"Professor,

excuse

me,

I

might

as

well

let

the

inspection

crelV

s

go

home, I

suppose,

it's

1820?"

Dunster

looked

up

long

enough t 0

say:

"Sure,

Jim.

But have one cretv

stay

in

t'\fatc h

quarters

--

you

can

never

tell

when

our

Joso-

phat

Bug

lvill

start

gnatving away

again."

Dunster

found

the

analysis

most

interest-

ing.

Almost

all

the

complaints

were

misdirec-

tions,

as

if

the

clerks

who

had

entered

the

original

shipping

instructions

had

chose

nat

random a wrong

destination

for

an

othe

r w

is

e

correct

shipment.

It

was

as

if

some

drug

had

affected

them

causing

momentary

loss

of

co

n-

sciousness.

But

how

could

any

drug

a f f e c t

so

many

clerks

at

so

many

shipping

offices?

He

looked

through

the

analysis

she

e t s

for

the

table

that

showed

the

geogr

a

ph

i

cal

dis

tribution

of

complaints

according

to

shipping

office.

There

Ivas,

it

seemed, a

definite

tend-

ency

for

complaints

to

concentrate

accordin

g

to

size

of

shipping

office

on

the

f 0

11

ow

ing

principle:

the

larger

the

volume

of

Shipments

from

a

shipping

office,

the

greater

the

number

of

complaints.

Another

look

shOtved

that

the

shipping

offices

connected

tvi

th

the

Sec

0 n d

Central

Office

of

Natrasys,

on

the

West

Coast

in

Netv

Frisco,

had

produced

no

complaints.

The

equipment

there

had

not

been

affected

by

the

trouble.

The

trouble

tvas

definitely

ass

oci-

ated

with

the

first

Central

Office,

Net'\'

Chicago.

AUTOMATIC

MIXUP

The

analysis

also

showed

that

all

the

de":'

IV.

fee

ti

ve

shipments

originated

in

the

per

i 0 d

1012

to

1547 on 10

September.

Other

than

these

indic

ations,

Dun

s t e r

fel

t

that

the

analysis

did

not

Sh01i

m u c h ,

aside

from

the

fact

that

his

earlier

estimate

of

loss

was

too

high.

The

average

claim

fo

r

damage was

going

to

be

far

less

than

$20,000.

In

fact,

he

would

estimate

it

at

300

tim

e s

$5,000,

or

one and a

half

millions.

Even

so,

Mr.

Valerian

Crurmiell was

definitely

not

going

to

be

pleased.

Dunster

leaned

back

in

his

chair,

turning

possibi

Ii

ties

over

in

h

is

head.

Eber

hard's

eyes

were

shut

and

his

mouth was

open;

he

lias

making

little

snoring

sounds.

The

Natrasys

had

no t r 0

ubI

e •

Due

to

some

kind

of

miracle,

l'llhich

Dunster

did

not

understand

--

and

which

he

felt

he

had

no

basis

yet

for

understanding

--

the

equipment

of

Natrasys

apparently

lVorked

perfectly.

But

the

third

day,

around

0940,

as

D u

n-

ster

came

into

the

Central

Control

room,

from

the

corridor

that

connected

with

the

Computa-

tion

Laboratory,

Jim

Hawker came

up

to

him and

said,

"This

just

came

in,

Professor,"

Dunster

read:

CENTRAL

CONTROL,

NATRASYS,

NEW

CHICAm

--

FARMER

G.

OTIS

HERE

CLAIMING

$200,000

DAMAGES

FOR

8,000

CHINCHILLIZARD

HJUaUNG

EGGS

DELIVERED

DEAD.

SHIPMENT

SCHEDULED

EXPRESS

DETROIT

TO

KALAMAZOO,

LOST,

FOUND

AT

MIAMI,

RESHIPPED.

WHY

TIUS ERRO R ?

PLEASE

ADVISE.

URGENT.

--

I.

R.

CONE,

KALAMAZOO

,

0920,

13

SfllTEMBER

"Crurmvell

will

sure

throw

a

fi

t

if

he

hears

about

this,"

said

Hatvker.

"You

are

right,

Jim,"

said

Dunster.

"But

the

tdre

does

not

say

the

time

when

this

ship-

ment

left

Detroit.

Can

you

find

that

out?"

Hatvker

said,

"Yes,

I

will,"

and

started

to

ualk

d01'\1n

to

the

Analysis

Section.

But

a

minute

later

Hal~ker

came

back

tvi

th

t

t'l

0

more

telegrams.

"IT,

the

Josophat

IT,"

he

said

and

paused,

"has

started

again."

Duns

ter

took

the

wires

and

read

the

m :

"Yes,

Jim.

The

Josophat

Bug

--

Get

the

test-

e rehls

going

again,

Jim.

Have them

concentrate

on

the

condition

of

the

machine

at

the

tim

e

hlhen

the

faulty

shipments

took

place.

Or

ask

them

to

if

they

can

it

is

an enormous num-

ber

of

microseconds

in

the

past."

-

27

-

Valerian

Crurmvell was

alone,

and

tvalking

back

and

forth

on

the

thick

soft

dull-red

car-

pet

in

his

office

on

the

45th

floor

of

the

Board

of

Commerce

Building.

It

was a big

room.

His

des

k,

a

great

wa

lnu

t

aff

air,

was i n

the

center

of

the

room;

over

by one

wall

tV

a s a

conference

table

large

enough

for

a dozen men.

There

were

Hind01'/s on

three

sides

of

the

room.

Crurmvell

rather

liked

his

room

as

a

rule.

But

today,

13

September,

he

t'/as

in

no

mood

to

enjoy

it.

He

had

been

out

of

the

office

yesterda~

to

play

golf.

But

he

had

.made a mi s t a k e

in

choosing

a

partner,

and

he

had

been

be

a

ten.

He

did

not

enjoy

being

beaten.

He

was

deeply

irritated.

His

secretary

opened

the

door

and

shrunk

a

little,

trying

to

make

herself

smaller

an

d

less

conspicuous.

"Mr. Crurm'Vell,"

she

sa

i d

timidly,

"there

is

a

long

distance

call

for

you

from

Kalamazoo, a Mr.

Otis

-

he

said

it

lias

important

--

he

said

he

had

to

speak

to

you

personally."

Crurmvell

said

in

his

loud

voice,

"y

0 u

told

him I was

in

conference?"

The

lit

tIe

secretary

tried

to

shrink

some more,

andsaid,

"Yes,

but

he

insisted

it

was

urgent,

that

I

interrupt

you

--

I

did

not

want

to

do

it,

Mr.

Crurmvell.

He

said

it

was

about

your

horrible

National

Traffic

Management

System

and

8000

murders.

"What?"

said

CruI1l'Jell,

"what

--

8000

mur-

ders?

Give

me

the

phone."

But

he

reach

e d

over

to

the

phone on

his

desk

and

pic

ked

it

up.

"Hello,"

he

practically

shouted

into

the

mouthpiece,

"this

is

CruI1l'l1ell."

"Mr. Crurmvell, "

said

a

thin

high-pitched

bi

ting

voice

on

the

other

end

of

the

lin

e,

"Mr. Crurmvell,

your

horrible

National

Traffic

Management

System

has

jus

t

murdered

8000

--

just

think,

8000

--

of

my

precious

chinch

i

1-

lizard

eggs

t'Vorth

$25

apiece."

"What?"

roared

CruIm\lell,

"how's

that?"

"Mr.

Crurmrell,

they

were

shipped

yesterday

from

Detroit

addressed

to

me.

It

takes

hal

f

an

hour

by

special

delivery

missile.

Chi

n-

chillizard

hatching

eggs

can

stand

motion

and

coolness

for

about

two

hours

and

then

they

die.

Well,

{vhat do you

suppose

your

horrible

system

did

for

me?"

"What?"

shouted

CruI1lvell.

"Well

those

8000

precious

eggs

1.lIen

t

to

Miami,

Florida,

--

just

think

--

Miami

instead

of

Kalamazoo,

and

got

lost

down

there,

and

did

AUTOMATIC

MIXUP

not

get

to

me

until

about

0800

this

morning,

and

every

single

one

of

my

poor

little

chi

n-

chillizard

eggs

is

dead."

The

voice

waile

d

at

the

other

end

of

the

phone,

and

then

itbe-

came a

shout:

"And

I

shall

have

you knOt', Mr.

Crumwell, I

shall

SUE

YOU

FOR

$200,000.

I

am

going

to

hire

the

best

lawyers

in

the

Midwest,

and George

Otis

of

Kalamazoo

will

beat

Valer-

ian

CrulJloiell

of

New

Chicago."

"Mr.

Otis,"

Crurmvell

shouted,

"1

e t

me

look

into

this,

I'll

call

you

back

in

an

hour

or

two,

just

as

soon

as I

can

get

to

the

bot-

tom

of

this."

The

voice

said,

"Do

as

you

like,

Mr.

Crwn-

t<vell.

You

are

going

to

get

a

summons

in

just

about

three

hours.

And

you

are

going

to

pay

me

$200,000."

Crurmvell was

galvanized

into

a

ng

e

rand

activity.

He

buzzed

three

bells.

At

0

nc

e

two young

men

assistants

and

his

sec

re

tar

y

came

into

the

room.

He

shouted

at

his

secretary:

''Miss

Dant-

,

zig,

transcribe

that

last

telephone

conversa-

tion.

And

call

Eberhardj

get

him

on

the

phone

--

I

want

to

raise

hell

l'Vi

th

him."

He

shouted

at

one

assistant:

"Call

the

directors

of

Natrasys,

get

them

all

herem

rnw

office

at

1200.

We're

going

to

have

a shavdam

on

this.

Oh, and

call

Dunsterj

tell

him

to

be

here

at

1230."

He

shouted

at

the

other

assistant:

"Take

the

transcription

of

that

phone

conversati

0 n

from

Miss

Dantzig.

Look

up

George 0

tis

of

Kalamazoo.

Find

out

what

you

can

about

him.

Also

find

out

what

is

the

value

of

chinchil-

lizard

eggs.

Check

our

shipping

age

n

tin

Detroit.

Dig

up

the

tfhole

story."

The young

man

started

to

ask

him a

ques~

tion.

He

shouted

at

him,

"Don't

you

as

k me

any

questions.

Can't

you

see

I'm

busy?

Figure

it

out

for

yourself

--

you'll

get

the

t r a

n-

scription."

Then

he

shouted,

but

more

to

himself

than

to

anyone

else:

"That

son

of

a

gun,

Eberhard,

I

told

him

to

make

that

system

lyork.

Wh

at

does

he

think

we're

trying

to

do?"

V.

The

clock

on

Crurmvell's

desk

s

howe

d a

Ii

ttle

after

noon.

All

the

directors

of

N a -

trasys

were

present.

Crurmvell's

tempe

r had

hardly

cooled

off

at

all.

He

had

just

finished

telling

the

account,

as

he

understood

it,

0 f

the

383

complaint

telegrams

of

10 S e

pte

mber,

and

the

9

complaint

telegrams

of

13

September

up

to

noon.

CruDlVell

shouted:

"And

that's

the

story

of

the

latest

episode

of

this

freak

0 f

nature,

the

automatic

control

of

traffic,

Nat-

rasys.

I've

had

enough."

Eberhard

spoke

up:

"Gentlemen,

we

are

doing

our

best

to

remove

the

present

trouble.

The

best

scientific

talent

in

the

country,

the

best

computing

machinery

talent,

are

on u u r

staff.

Let's

consider

hOty

often

these

errors

happen.

Figures

have

often

been

qUQte

d

here

in

our

directors'

meeting,

and

they

show

that

the

overall

reliability

and

accuracy

of

Natrasys

is

far

superior

to

what

happens

in

~u

man

management

of

traffic.

Of

course,

previ-

ously

all

the

mistakes

t,ere

scattered

--

here

an

office

made a

mistake,

there

an

office

made

a

mistake.

The

far

greater

number

of

mistak~

were

the

responsibility

of

a

far

greater

num-

ber

of

individuals.

Now,all

the

mistakes

are

centralized

too,

and

tve

are

the

victims

of

that.

When

our

sys

tern l'lorks, and

it

tvorks

almost

all

the

time,

it

works

tvonderfully

well

and m os t

profitably

for

all

of

us."

Oliver

spoke

up: ''Yes, Mr.

Eberhard,

that

is

jus

t

the

trouble

--

'when'."

Crumwell

roared:

''Yes,'

we

stand

to

be

sued

for

$200,000

on dead

chinchillizard

eggs,

and

Eberhard,

you

yourself

estimate

six

mil-

lion

dollars

damage

suits

on

the

383

telegrams

of

10

September.

If

I

wanted

to

throw

my

mon-

ey

down

a

hole,

that

WJuld pay

me

more

in

the

end

than

your

blazing

Natrasys."

Eberhard,

stood

up,

no

smile

on

his

face,

tvringing

his

hands:

"Gentlemen, "

he

said,

"Gentlemen,

think

what

we

have

accomplished

in

six

years.

Think

of

the

good

records

we

have

set~

Think

of

the

dividends

we

have

paid:

over

ten

million

dollars

in

the

last

year

a

Ion

e.

Suppose

lye

did

have

to

settle

some

claims

last

year

to

the

tune

of

$800

,000.

That

still

leaves

more

than

90%

of

the

dividend

earned

for

us."

Oliver

started

to

say

something,

butCrum-

well's

secretary

opened

the

door,

saying

in

a

trembling

voice:

"Mr.

Dunster

has

arri

ve

d •

You

said

you

wanted

to

knOtv,

Mr.

Crumwell."

Crurmvell

shouted,

"Get him

in

here."

Dunster

came

in,

a

little

smile

aroun

d

his

lips

j he

knetv

wi

thout

being

told

what

~s

happening..

He

saw

the

bleak

expressions

0 n

all

the

faces,

and

rage

and

hatred

on

Crumwell~

face.

He

enjoyed

this.

It

was

a

battle,

and

he_was

sure

of

his

ground,

and

his

weapons.

Crumwell

said

with

exaggerated

politen~

and

sneering

in

his

voice,

"Professor

Dunstel;"

and bowed. Then

he

roared,

''You

know

lihatyour

Natrasys

has

cost

us

in

damages

in

the

1 a s t

four

days?"

-

28

-

AUTOMATIC

MIXUP

Dunster

smiled,

"As

yet,

nothing

actually,

can't

fire

Dunster.

There

isn't

a

soul

in

the

but

potentially

about

two

million

dollars."

country

who

can

take

his

place."

Crurmvell

shouted,

"How's

that?

I

thought

it

tvas

nearly

seven

million

dollars."

Dunster

said,

'We

have

revised

our

esti-

mates.

I

believe

your

figure

is

an

old

esti-

mate.

Also,

I

have

been

talking

to

our

Traf-

fic

Claims

Division.

They

believe

that

we'll

be

much

better

protected

than

before

because

of

a

recent

Supreme

Court

decision."

Crurmvell

shouted,

"What

decision?"

"The

Snyder

vs.

United

States

decision,

which

limited

in

several

ways

the

liability~

carriers.

The

United

States

was

the

carri~r

in

that

case."

Crurmvell

realized

he

was

losing

groun

d

because

his

assistant

had

not

dug

up

eno

ugh

information,

and

so

he

shifted

his

attack.

"Mr.

Dunster,

" he

shouted,

"what's

the

reason

for

this

trouble

in

the

first

place?"

"Mr. Crurmrell, I

don't

know." 0 u n s

ter

smiled

again.

'~at,

you

mean

to

stand

there,

you

grin-

ning

idiot,

and

tell

me

you

don't

know?" The

roaring

was

louder

than

ever.

Dunster

smiled

again,

and bowed

slightly:

"Sir,

there

are

many

things

that

a

sc

ie

n

tist

does

not

knOt\'o

"

Sam

Oliver

smiled,

and

Dunster

realized

that

he

might

be

Ivinning

over

a

man

l'lhom

h e

had

not

expected

to

win

over.

Crurmvell

real-

ized

that

he was

losing

a

man

from

his

sid

e,

and

got

even

angrier.

Crurmvell

shouted,

"So you

mean

to

tell

ne

that

you

don't

know

what

the

trouble

is,

and

t'Je

may

lose

a

million

dollars

of

damages

each

day,

and you

just

stand

there

and

say

you

don\

know, "

Crurmvell

turned

to

Oliver:

"What do you

think?"

Oli

ver

spoke

deliberately:

"It

see

m s

clear

to

me

that

lie

are

trying

to

do

too

much,

considering

the

present

state

of

the

scien

c

e.

I t seems

to

me

it

may

be unwise

for

us

to

try

to

go

so

fast,

press

scientific

advances

ro

the

hilt

of

possibilities.

Automatic

traffic

con-

trol

may

be a

thing

of

the

future,

not

the

pre-

sent."

Crurm"lell

shouted,

"In

any

case,

I

think,

Duns

ter

is

the

l'Vrong

man

to

have."

Eberhard

in

alarm

put

in:

"Dunster!

We

Crunnvell

looked

at

Eberhard,

and

all

th

e

pent-up

rage

jnside

of

him

burst.

He

screamed:

'ty ou

bl

i

theri

ng-eyed

s on

of

a wi

tch!

I

can't

fIre

Duns

ter?

Is

that

what

you

say?"

He

turn-

ed

to

Dunster,

"You're

fired.

You

hear

that?

You're

fired.

Beat

it!

Get

out

of

here!"

Duns

ter

grinned

ins

ide

of

his

face

but

only

the

traces

of

it

showed on

the

outside:

"Well,

Mr. Crumwell,

since

I

am

fired,

how

much

will

you

sell

Natrasys

for?"

If

ever

the

wind

burst

out

of

a

balloon,

so

the

wind went

out

of

Crurmvell.

He

had

not

expected

this

sort

of

a

reply

from

DunstEr.

Why

did

Dunster

say

that?

Maybe

Dunster

k~ew

how

to

fix

that

trouble.

Maybe

Dunster

a

Ire

ady

knetv.

Maybe

Dunster

knew

a good

thing

Ivhen

he

saw

it.

Maybe

Dunster

Ivas

going

to

play

him

Crunnvell,

for

a

sucker.

Maybe

just

as

soon

as

he,

Crunnvell,

got

out

of

Natrasys,

there

Ivould

be

no

more

troubles

in

Natrasys:

All

the

s e

ideas

'left

their

traces

on

Crunnvell's

fa

c e ,

and

Dunster

tvatching

closely

realized

that

each

idea

he had meant

to

suggest,

Ivas

present.

Crunn"lell s

aid

rapidly:

"Do

you

think

you

can

fix

that

trouble?"

Now

Duns

ter

felt

sure

he had won, and

he

said

with

contempt:

"I

can't

fix

it

while

I

am

here

arguing

with

you."

Crunnvell

decided

he

needed

more

time

t 0

settle

the

goose

of

Natrasys,

Eberhard,

and

Dunster.

He

said:

"Gentlemen, I

propose

that

we

give

Eberhard

and

Dunster

until

noon

tomor-

row

to

get

rid

of

the

present

trouble.

If

not

eliminated

by

then,

I

propose

we

sell

Natrasys

for

scrap

--

not

to

Dunster

--

and

foreclose

on

the

Sagamore

Railroad."

Oliver

said:

"I

agree

to

your

first

pro-

posal

anyhow.

Let

us

give

them a

little

more

time.

But

we

certainly

cannot

afford

to

let

all

our

profits,

and

reserves,

and

resources

go

down

the

sewer

of

damages."

VI.

Back

in

the

main

controt

room

of

Natrasys,

all

that

afternoon

and

evening,

Dunster

and

IIalvker and

the

six

tes

t

crews

Ivorked s

teadi

ly,

testing

possibilit~es.

Eberhard

stayed

around,

asking

a few

questions,

watching,

saying

over

and

·over,

"I

tvish I

could

do

something."

No-

body

said

very

much

to

him,

but

treated

him

li~·

a

piece

of

furni

ture.

Every

now

and

then

h e

went

into

the

men's

room and

took

a

drink.

This

night

there

was no

moonlight.

Outshle

-

29

-

AUTO~ATIC

MIXUP

the

windol1Js

Ivere

just

stars;

the

clouds

of

the

Dunster

tvas

alone,

except

for

the

night

day

had

blmvn

off.

The

evening

1'\10re

on,

and

operating

crew

on

Natrasys,

-

all

busy

at

their

the

clock

showed a

little

after

2100.

Finally

usual

tasks.

lie Ivent

bapk

into

his

office.

He

Dunster

said:

"Now

at

last

we

have

checked

was

tired,

discouraged,

lmd

sleepy.

He

reach-

every

possibility

we

can

think

of,

haven't

we,

ed

into

a

drawer

for

hisllittle

box

of

nodzis-

Jim?"

topine,

and

took

a

pill.

The

sleepiness

left

"Yes,"

said

Hawker,

"every

test

crew

has

checked

all

their

assignments.

The

assignmen~

--

you

drel~

them up -

cover

the

complete

rna

-

chine.

The

Josophat

Bug

is

beyond

me."

Dunster

said,

"Hell

is

a

place

where

all

the

equipment

and

all

the

instruments

are

per-

fect,

and

nothing

works

right."

.Eberhard

said:

"You

fellmvs

sure

i

n-

spire

me. I

think

I

should

have

been

an

en-

gineer,

so

that

I

could

be

so

utterly

baffled~

Another

complaint

telegram,

the

15th

one

for

13

September,

1\1'as

at

that

moment

han

d e d

to

Dunster.

He

read

it:

CENTRAL

G.ONTROL

OFFICE,

NATRASYS,

NE

W

CE~CAC~

--

ARGENTOMYCIN

Cft.PSULES

ARRI

\ED,_

BUT

ALMOST

TOO

LATE.

CLARA"HARTLEY

£-Eb

SEVEN

JUS't-

SAVED.

SCEEDULED-NEW

YORK

TO

OGUNQUIT.

--

-ARRIVED

VIA

INDIANAPOLIS.

-jm-

QUEST

URGENT

CHECK

EQUIPMENT

IN

INTEREST

OF

MEDICAL

DELIVERIES.

--

L.

SEGERSON,

OGUNQUIT,

ME.,

2025,13

SEPl'EMER

Dunster

handed

it

to

Eberhard.

Eberhard

said:

"I

have

been

afraid

of

that

kind

of

thing

all

along.

P

laying

If

i t h

life

and

death,

that's

what

they'll

say.

And

they'll

harp

on

it,

and

harp

on

it,

until

we

are

all

run

ragged.

I

can't

stand

this

any mare.

I'm

going

out

and

get

drunk."

Dunster

said:

"OK,

Barry,"

and

for

once

his

disappointment

in

his

friend

shmved,

and

he

added,

in

a

voice

Harry

couidn't

hear,

"I

guess

that's

the

best

you

can

do."

Eberhard

took

his

hat

and

left.

Dunster

turned

to

Hauker:

"Jim,

you're

all

llfOrn

out.

Go

and

lie

dm-vn

on

the

sofa

in

the

l"1aiting room, and

catch

some

sleep."

"Is

there

nothing

more I

can

do,

Profes-

sor?"

"Jim,

I'm

not

sure

there's

anything

any-

one

can

do."

HaHker lTent

into

the

t1Jai

tin9'

room by

the

guard's

desk.

He

stretched

out

like

a 10 y a 1

dog

at

his

master's

command, and

fell

aslee

p

in

less

than

five

minutes.

him.

His

thoughts

became

like

crystal

in

clar-

ity

and

vividness.

It

was

dangerous

medicine,

he

knet'V,

but

he

could

take

one

about

once

in

six

months

without

ill

effects.

lIe

sat

dOlm

at

his

desk.

It

seemed

th

at

Cruffitvell's

face

looked

at

him

from

the

opposite

wall,

agleam

tdth

malice

and

hatred

and

vi

c-

tory.

He

could

hear

Crunnvell

shouting.

He

had

until

tomorrOlV

to

Idn.

Then

probably

he

1'.'

as

through.

Then

Eberhard's

face

seemed

to

look

a t

him,

Eberhard's

face

before

the

events

of

the

last

year

had

begun

to

put

so

much

tvear

and

tear

into

it.

lIe tvas

fond

of

Eberhard;

he

1

ik-

ed

Eberhard's

ambition,

his

enthusiasm,

his

en-

ergy,

his

judgment

of

men,

his

risk-t

a

kin

g.

Dunster

l'Vas

deeply

sorry

that

his

failure

at

diagnosing

the

trouble

was

causing

so

m u c h

Ivorry

for

Eberhard.

He

s

aid

to

himself:.

"If

tve

could

reduce

the

strain

on

Eberhar

d,h

e

would

snap

back

into

his

old

fine,

stimulating

self."

Dunster

knew

he

had

to

put

these

thoughts

out

of

his

mind.

He

must go

over

the

problem.

Compare

the

evidence.

Analyze

the

possibili-

ties.

Ask Ivhat

are

the

basic

assumptions?

Am

I

considering

all

of

them?

VII.

Eberhard

came

into

the

Sagamore

Railw~.

y

President's

office

the

about

half

past

nine.

There

in

the

vis

i

tor's

chair

h e

saw

Dunster,smoking

his

old

pipe,

deep

in

the

morning

paper.

Eberhard

cried:

"Andy, t\'hat

are

you

do-

ing

here?"

Dunster

said,

"Well,

Harry,

the

trouble's

allover."

Eberhard

stopped.

"I

don't

believe

it.

Miss

Drury,

shm"

me

the

latest

traffic

com-

plaints

on

the

teletype."

He

looked

through

them

quickly.

"Andy,

the

last

one

is

the

GJun-

quit

one."

lIe

took

a

deep

breath.

"Andy,

how

do you

knolv

there

won't

be

some

more?"

"The

cause

of

them

is

removed,"

said

Dun-

ster.

"At

least,

I

believe

it

is

removed, and

at

any

rate,

we

know

the

cause."

Eberhard

Ivas

grinning

and

breathless,"Andy,

you've

done

it

again.

Tell

me

the

story."

-

30

-

AUTOMATIC

MIXUP

Dunster

smiled.

lIe

said,

"Well,

la

s t

take

a look

there?

nigh t

after

you

left

and

Jim l"ent tt.

sleep

on

the

sofa,

I Nent

into

my

office,

and took a

pill,

and asked myself, 'What

are

the

bas

ic

assumptions?'

I

started

analyzing.

It

was

certain

that

we

had most

thoroughly

che

c ked

everything

that

l"e

knew

about.

So

pro

b

ably

it

l'Vas

something

net",

not

previously

met.

"Now

the

trouble

was

certainly

1

ike

a

shielding

trouble.

"1

got

the

analysis

out

and looked i t

over

again.

The

more

I looked

at

it,

the

more

it

made

me

think

it

was

as

if

shipment n had

been connected

to

destination

n+1.

That,

of

course,

lvould

get

by

our check number

verifi-

cation

and

other

automatic

checks.

What

could

do

that?

Extra

pulses

or

delayed

pluses,

per-

haps.

"But every

extra

pulse

was

stopped

b y

our

lead

shielding.

We

verified

that

complete-

lyon

10

September

and

again

on

13 September.

Besides,

none

of

our check equipment,

our

os-

cilloscopes,

our

master

timing

verification

procedures,

had

revealed

extra

pulses.

"Well,

if

not

extra

pulses,

then

maybe

pulses

lllhich were

delayed

or

advanced a

little

in

time.

But

hOI"?

But

here

again,

none of our

verification

procedures had

detected

anything

like

that.

But

maybe

our

observation

and

test-

ing

apparatus

would be

affected

by

the

sam

e

trouble,

getting

through

shielding.

"NOIV,

what can

get

through

shielding?

Pulses

cannot.

But

gravitation,

for

example,

can.

What

else

besides

gravitation?

We

live

in

a

space-time

continuum, Aha!.

Time

can

get

throu,gh

shielding.

NOt",

lye

are

working

with

pulses'.l

lyay

dOlYn

in

the

finest

subdivisions

of

time

~-

hundredths

of

microseconds. Suppose

time

varied

--

not

much

but

just

a

little

bit.

Then

we

could

get

delayed

or

advanced

pulses.

Most

of

them

~~ould

simply

cause

the

equipment

to

produce

an

ordinary

isolated

random

fail-

ure,

and

then

automatic

recalculation

would

send

the

shipment

on

its

l'lay

lvi

th

the

chances

of

being

completely

correct.

But every

now

and

then

the

connec

tion

between a shipment

la-

bel

and

the

destination

label

would be

spoil-

ed.

So,

our

theory

would

explain

what has hap-

pened.

"Now,

maybe

the

ass

umption

that

time

is

sometimes

not

uniform

explains

the

trouble

in

Nelli

Chic ago. But

why

h'ould

we

in

New

Chic

ago

suffer

from

it

a

little,

and

our

second Cen-

trnl

Office

in

Nel~

Frisco

not

at

all?

Well,

1

said

to

myself, lwuld

there

be

some

source

of

time

variation

nearby?

As

soon as I

ask-

ed

the

question,

1 guessed

an

answer.

The

Relativity

Laboratory.

Why

didn't

I

think

of

it

before?

Anyhuw,

tJhy

not

go

over

and

"Well,

it

was

then

about

midnight.

I

took

my

master

key, and went

into

the

Relativity

Lab, and found a

piece

of

apparatus

I

didn't

know

about.

The

engineering

notebook

on

the

bench showed

that

a

visiting

young Scandinavian

scientist

Olaf

Nordstrom had

built

it,and

that

he

t'las

experimenting

with

apparatus

for

vary-

ing

time.

"His experiment

records

correlated

tv

i

th

the

time

of

our

troubles.

His

address

was

in

the

front

of

the

notebook. I went over

to

his

dormi

tory,

waked

him up, and

talked

t'li

th

hi

m.

Very

obligingly

he

got

up

and lvent back to

the

lab

t'd

th

me,

and

explained

all

that

he

was

do-

ing.

He

was

simply

astonished

that

the

range

of

his

apparatus

had

penetrated

500

feet

t 0

Natrasys;

he

turned

it

off

and

promised

to

con-

sult

with

me

nenceforth,

so

as

not

to

disturb

Natrasys.

"So

I went back

to

the

Computation Labora-

tory,

it

was

then

0200, and l\lent

to

sleep peace-

fully."

Eberhard took a deep

breath.

He

s

aid:

"Andy,

you're

amaZing. I

am

going

to

call

up

our enemy,

Valerian

CruImvell. Here,

listen

in

on

the

extension."

Eberhard picked

up

the

phone, and

aftera

minute: "Mr.

Cru~ell,

Dunster

has

licked

the

Natrasys

trouble.

He

tracked

it

down

to

some

new

experimental

apparatus

in

the

Relativity

Laboratory."

Then Dunster

heard

the

familiar

v 0

ice

shouting:

"You

confounded sons

of

witches,why

the

blazes

don't

you

prevent

that

sort

of

tlling

ahead

of

time?

OK,

so

it's

fixed.

So

why

ilie

he

11

do

you

bother

me?"

CruImvell banged

the

receiver

down.

Eberhard and Dunster looked

at

each

other

and

grinned.

-E\U -

-

31

-

FlT'~f:TIO"

OF

.\UTO\t.\TIC

pp.Or,nA~t'n;\G

-

(conti;ued

from

page

9)

•

general-purpose

computers

•.

r have pOInted

ou

t

several

problem

areas,

starting

lITi

th

the

train-

ing

process,

through programming and

cod

i n

g,

and

ending

wi

th

the

continual

revis

ion

of

pro-

grams

to

meet changing

conditions.

I

be~ieve

there

is

something

to

be

learned

by

lookIng

at

the

procedure

of

programming from beginning

to

end.

We

can

think

of

the

ul

timate

automat i c

system as comprehending

the

entire

process.

It

is

useful

and

reasonable

to

at

t a c k

separate

parts

of

the

process

at

the

pres

e n t

time, and

this

is

now

being done b y s 0 m e

groups.

It

is

encouraging

that

some

0 f

the

problems I have mentioned, have been

partially

solved

by

one

group

or

another.

I

hop

e

that

the

problem

of

programming

1\'111

be

lddely

rec-

ognized

as

important enough

to

at

t

rae

t

the

attention

and

effort

of

those

cap

a b 1 e 0 f

solving

it.

The

tddespread,

economic use of

genera

1-

purpose

data-processing

equipment

may,

in

fact,

depend

intimately

on

the

development

of

more

efficient,

much

faster,

and

much

che

ape

r

methods

of

programming.

-~-

*------------------

...

--------

....

IR\1

702

INSTALLATION

(continued

from

page

15)

4.

The

inventory

is

reduced

to

a

minimum

because

the

stock

is

controlled

daily.

5.

An

additional

increase

of

control

re-

sults

from

the

centralizing

of

the

vast

inven-

tory

operation.

6.

Major

reductions

in

the

cost

of

the

operation

are

achieved.

-

ENIl

-

-32 -

NEW

PATENTS

RAY~10ND

R.

SKOLNICK,

Reg.

Patent

4.gent

Ford

Instrument

Co.,

Div.

of

Sperry,

Hand

Corp.

Long

Island

Cit~,

N.Y.

The

follmving

is

a

compilation

of

patents

per-

taining

to'computers

and

associated

equipment

from

the

Official

Gazette

of

the

United

States

Patent

Office,

dates

of

issue

as i n d i

cat

e

d.

Each

entry

consists

of:

patent

number

lin-

ventor(s)

I

assignee

I

mvention.

November

1,

1955: 2,722,125 I George

W.

Sieb-

engartner,

San Diego, and

Curtis

S.

Dug

an,

Ventura,

California,and

Dean

E. IIumbert,

Day-

ton,

Ohio

I United

States

of

America I A gyro-

scopic

control

system.

2,722,379

I Arthur F. Hayek,

Pleasantville,

NBV

York I General

Precision

Laboratory

Inc.

I

A

rotary

mo.tion

storage

device.

2,722,601

I Gerard Jean

Rene

Piel,

Paris,France

I

Societe

d'Electronique

et

d'Automatis

me,

P~ris,

France I

An

electric

imeulse

count~ng

chain

having a

plurality

of

cascade-connect-

ed

bistable

trigger

circuits

individu

ally

operating

as

binary

stages.

2,722,604

I Evelyn Stetmrd

Lansdown

Beale,Stan-

well

Moor,

near

Staines,

England I

Interna-

tional

Standard

Electric

Corp.,

New

York,

N.Y.

I A

signal

responsive

and

integrating

device.

2,722,605

I Robert L.

Mills

and Frank

J.

M c -

Donol,

Dallas

Texas I Socony Mobil Oil

Com-

pany,

Inc.

I A balanced phase

discriminator.

2,722,630

I Maurice

Charles

Branch,

feter

Morris

King,

and

William Arthur George Walsh,

Londm,

England I

International

Standard

Electric

Corp.

Net\'

York,

N.Y.

I A gaseous

discharge

tub

e

counting

circuit.

2,722,654

I Daniel

J.

Sikorra,

Milwaukee, Wis-

consin

I Allis-Chalmers Manufacturing Com-

pany, Milwaukee,

Wisconsi~

I A

regulating

system

for

maintaining a

quantity

substan-

tially

constant

at

any

selected

value

w

ith-

in

a predetermined

range.

2,722,660

I John P.

Jones,

Jr.,

Pottst6wn, Pa,

I United

States

of

America I Apparatus

for

transforming

a

signal

into

consecutive

pulse

code groups

representative

of

COllS

e

cut

i ve

amplitude values

of

said

signal.

2,722,675

I Joseph Michal,

New

Hyde

Park,

and

Robert E.

Staehler,

Ridgewood,

N.Y.

I

Be!"l

Telephone

Laboratories

Inc.,

Net"

York,

N.Y.

/

A

pulse

counting_system. .

2

722

676 I

Semi

Joseph Begun, Cleveland,

Ohio

'I

Cievite

Corporation,

Cleveland,

Ohio

I A

magnetic

information-storing

device.

November

8,

1955: 2,722,845 I

Clinton

L.

Cum-

mings, Groton, Mass. I Barry

Controls

In~.,

Watertown, Mass. I A change speed mechanIsm

for

instrument

drives.

2,723,013

I Paul

H.

Rogers,

Fo~t

.Wal

ton, Florida I

(continued

on

page

~8)

It's

ARNOLD

for

EVERYTHING

in

TAPE-WOUND

CORES

Let

us

help

you

with

your

core

re-

quirements~_

for

Pulse

and

Power

Transformers,

3-Phase

Transformers,

Magnetic

Amplifiers,

Current

Trans-

formers,

Wide-Band

Transformers,

Non-Linear

Retard

Coils, Reactors,

Coincident

Current

Matrix

Systems,

Static

Magnetic

Memory

Elements,

Harmonic

Generators,

etc.

FrYv

amp1R;/$

DI!1liiiJ

Write

for

Bulletins:

*

SC-107-Cut

Cores,

Types

C

and

E

*

TC-101A-Toroidal

Cores, cased

and

uncased

*

TC-I08-Bobbin

Cores

Address

Dept.

CA-62

.

..

MATERIALS:

Deltamax,

4-79

Mo-

Permalloy,

Supermalloy,

Mumetal,

4750,

Monimax,

Silectron,

Permendur:

all

are

available

for

tape

wound

core

applications.

The

choice

of

material

will

depend

upon

the

specific

proper-

ties

required.

GAUGES:

The

following

standard

tape

thicknesses

are

av;lilable

for

Arnold

tape

wound

cores

in

most

of

the

mag-

netic

materials

mentioned

above:

.

012",

.004",

.002", .001",

.0005",

or

.00025".

Bobbin

cores

are

made

from

tape

.001"

to

.000125"

thick

.

SIZES:

Cores

weighing

from

a

fraction

of

a

gram

to

many

hundreds

of

pounds

can·

be

supplied.

Toroidal

cores

are

made

in

27

standard

sizes

with

nylon

cases.

Types

HC"

and

HE"

cut

cores

are

made

in

a

total

of

530

standard

sizes.

Many

special

sizes

and

shapes

of

both

gapless

and

cut

cores

are

manufactured

for

unusual

requirements

.•

Let

us

work

with

you .

waD

5818,

~tT!!

ARNOIl)"

EN(HNEERINs:COMPANY

.

f~;<'$""

,'SUBSIDIARY'

OF

ALLEGHENY

LUDLUM

STEEL

CORPORATION

f,'

m '

Gerie~al

Office

& Plant:

Marengo,

Illinois '

;,,:'

,

DISTRICT

SALES

OFFICES

•••

New

York:

350

fifth

Ave.

r::

' '

.'

Los

Angeles:

3450

Wilshire.

Blvd.

.,Bo~on:

200

Berkeley

Sf.

1..::..,...~

,-:...

> , ' ,

"s

.."..,

-33

ASSOCIA

TION

FOR

COMPUTING MACHINERY MEETING,

PHILADELPHIA,

SEPT.

14

to

16,

1955

Names

and

Addresses

of

Authors

of

Papers

In

the

November,

1955,

issue

of

"Computers ani

Automation",

we

printed

the

titles,

authors,

and

abstracts

of

the

papers given

at

the

meeting of llie

Association

for

Computing Machinery

at

the

Univer-

sity

of

Pennsylvania,

Philadelphia,

Pa.,

Sept.

14

to

16, 1955. These

titles

and

abstracts

were

re-

p'roduced

photographically

from

the

program

distri-

buted

at

the

meeting;

since

the

program did

not

give

the

authors'

addresses,

they t

...

ere

not printe

d.

This

lack

is

nOI\'

remedied. Following

is

the

list

of

the

names

and

addresses

of

authors,

kindly

furnished

by

Professor

John

P.

Nash,

Digital

Com-

puter

Laboratory,

University

of

Illinois,

Urbana,

Ill.,

chairman

of

the

Program Committee

of

the

As-

sociation

for

Computing Machinery.

The

numbers

be-

low

agree with

the

numbers

of

the papers

shown

on

pages

17,to

32

of

the

November

issue.

1.

David P.

Perry,

Sperry

Rand

Corp.,

315

Fourth

Ave.,

New

York,

N.Y.

2.

Mark

Lotkin,

RCA

Service

Co.,

Inc.,

Missile

Test

Project,

Patrick

AFB,

Fla.

3.

Richard

E.

von

IIoldt,

Radiation

Laboratory,

Uni

v.

of

Calif.,

Berkeley

4,

Calif.

4.

Wallace

Givens, Dept.

of

Math., Univ.

of

Thnn.

Knoxville, Tenn.,

or

Oak

Ridge

Nat'l

Lab.,

Oak

Ridge, Tenn.

5.

Rob

ert

Perkins,

Ramo-Wooldridge

Corp.,

5740

Arbor Vitae

St.,

Los

Angeles 45,

Calif.

6.

Hollis

A.

Kinslow,

IBM

Corporation,

P.o.

Box

390, Poughkeepsie,

New

York

7.

E.

K.

Blum,

U.S. Naval Ordnance

Lab.,

W

hi

te

Oak,

Silver

Spring,

Md.

8.

J.

N.

P.

Hume,

Computation

Centre,

McLennan

Lab., Univ.

of

Toronto, Toronto, Canada

9.

Charles E.

Thompson,

General

Electric

Corp.,

Richl3nd,

Wash.

10. John

R.

Stock, United Carbide and Carbon Corp.,

30

E~

42nd

St.,

New

York

17,

N.Y.

11.

Stanley

Frankel

(Continental

Oil

Co.) 135

Main

St.,

Seal Beach,

Calif.

12.

G.

M.

Amdahl

and J.W. Bachus,

IBM

Corp., P.

O.

Box

390, Poughkeepsie,

N.Y.

13. William

E.

Smith, North American Aviation

Co~,

12214

Lakell'ood

Boulevard,

Downey,

Calif.

14. Ch'arles

J.

Sldft,

CONVAIR,

San

Diego,

Calif.

15.

W.

R.

Hoover,

Calif.

Inst.

of

Tech.,

Pasadena,

Calif.

16.

Norman

Grieser,

Underwood

Corp.,

Electronic

Computer

Div.,

35-10 36th Ave.,

L.I.C.

6,

N.Y.

17.

J.

A.

Postley,

The

Rand

Corp., 1700

N.

Main

St.,

Santa

Monica,

Calif.

18.

L.

Eselsen', Remington Rand,

Inc.,

315 Fourth

Ave.,

New

York

10,

N.Y.

19.

Mary

K.

Hawes,

Sperry

Rand

Corp.,

315

Fourth

Ave.,

New

York

10,

N.Y.

20.

Leon

Nemerever,

Unde~iood

Corp.,

Electronic

Computer

Div.,

35-10 36th Ave.,

L.I.C.

,6,

N.Y.

21.

J.

A.

Porter

and

D.

L.

Shell,

General Electric

Co., Computer Techniques Development,

Inves-

tigation

Sect.,

AGT

Development

Dept.,

Bldg.

300,

Cincinnati

15,

o.

-

34

-

22.

A.

E. Roberts,

Jr.,

General

Kinetics,

Inc

••

555

23rd

St.,

S.,

Arlington 2,

Va.

23.

Dean

Arden, Mass.

Inst.

of Tech.,

Digital

Com-

puter

Lab., Cambridge, Mass.

24. James E. Robertson,

Digital

Computer Lab., 168

Eng. Res. Lab., Univ.

of

Illinois,

Urbana,

Ill.

25.

Norman

R.

Scott,

Dept.

of

Electrical

Engineer-

ing.,

Univ. of Michigan,

Ann

Arbor, Mich.

26.

Earl

J.

Isaac

and

Richard

C.

Singleton,

Stan-

ford

Research

Inst.,

Stanford,

Calif.

27. Benjamin

L.

Schwartz,

Battelle

Memorial

Inst.,

Cleveland,

o.

28.

R.

L.

Cline,

IBM

Corp., P.

O.

Box

390, Pough-

keepsie,

N.

Y.

29.

Edward

II.

Friend,

N.

Y.

Life

Insurance

Co.,

51

Madison Ave.,

New

York

10,

N.Y.

30.

J.

H.

Allen,

Temco

Aircraft

Corp.,

P.

O.

Box

6191, Engineering Dept.,

Dallas,

Texas

31.

A.

S. Householder, Math.

Div.,

Oak

Ridge

Nat~

Lab., P.

O.

Box

P,

Oak

Ridge, Tenn.

32.

J.

H.

Wegstein,

Nat'l

Bureau

of

Standards,

Wadl.,

25,

D.

C.

33.

S.

G.

Campbell,

Oak

Ridge

Nat'l

Lab.,

P.

O.Bo~

P.,

Oak

Ridge, Tenn.

34.

Norman

F.

Loretz,

Magndvox

Research Lab.,

Van

Nuys,

Calif.

35.

Alan

J.

Perlis,

Head, Computation Lab.,

Purdue

Univ., LaLayette, Indiana

36.

C.

W.

Adams,

Office

Methods

and

Procc d

ures

Westinghouse

Electric

Corp.,

Pittsburgh,

Pat

37.

E.

M.

Gettel

and

D.

L.

Shell,

Computer

Te

ch -

niques Development,

Investigations

Sect.,

AGT

Development Dept., Bldg 300, General

Electric

Co.,

Cincinnati

15,

o.

38.

Stefan

Bergman,

Stanford

Univ., Applied Math.

and

Statistics

Lab.,

Stanford,

Calif.

39.

Stanley

Katz

and

Jack

Warga,

Elec

troData

Corp.,

460

Sierra

Madre

Villa,

Pasadena,

Calif.

40.

Jack

Beller,

H.

B.

Keller,

Samuel

Schechter,

Inst.

of

Math

Sciences,

AEC

Computing

Facil-

ity,

253 Greene

St.,

New

York

3,

N.Y.

41.

Barry Gordon,

Equitable

Life

Assurance Society,

393 Seventh Ave.,

New

York

I,

N.Y.

42.

Walter F. Bauer, Ramo-Wooldridge Corp.,

5740

Arbor Vitae

St.,

Los

Angeles 45,

Calif.

43.

J.

Wright, Engineering Research

Inst.,

Willow

Run

Research

Center,

Univ.

of

Michigan, Ypsi-

lanti,

Mich.

44.

J.

H.

Brown,

John

W.

Carr

III,

Boyd

Larrowe,

J.

R.

McReynolds,

Engin~ering

Research

Inst.,

Willow

Run

Research Center, Univ.

Jf

Michigan,

Ypsilanti,

Mich.

45.

F. S.

Beckman

and

D.

A.

Quarles,

Jr.,

IBMCo~

590 Madison Ave.,

New

York,

N.Y.

46.

David

J.

Fitch,

Student Counseling Bureau,

Univ. of

Illinois,

Urbana,

Illinois

47.

William Orchard-Hays,

The

Rand

Corp.,

1700

Main

St.,

Santa Monica,

Calif.

48.

Joseph

V.

Natrella,

Directorate

of

Management

Analysis,

IIdqtrs.,

U.S.A.F., Washington,

D.C.

49.

John

Greenstadt,

IBM

Corp., 590 Madison Ave.,

Net~

York,

N.Y.

50.

E.

G.

Kogbetliantz

(paper cancelled),IBM

Corp~

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

P.

O.

Box

390, Poughkeepsie,

N.Y.

ACM

MEETING

.

79.

Arthur

G.

Dmming

and Alston S. Householder,

Oak

Ridge

Nat'

1 Lab.,

P.O.

Box

P,

Oak

Ridge,

Tenn.

Wnllace Givens, Univ. of Tennessee, Dept. of

Math.,

Knoxville,

Tenn.,

or

Oak

Ridge

Nat'l

Lab.,

O~k

Ridge, Tenn.

Stephen

E.

Wright,

Sperry

Rand

Corp.,

315

Fourth Ave.,

New

York

10,

N.

Y.

Bruce

G.

Oldfield

and

Robert

H.

Bracken,

U.S.

Naval Ordnance

Test

Station,

China Lake,

Calif

•

Bruce

G.

Oldfield,

U.S. Naval Ordnance

Test

Station,

China Lake,

Calif.

Howard

S.

Levin, Ebasco

Services,

Inc.,

2

Rector

St.,

New

York

6,

N.Y.

IIe

len

~leek

and

Leon

Gainen, Hughes Research

and Development

Labs.,

Hughes

Aircraft

Co.,

Culver

City,

Calif.

Ned

Chapin,

Illinois

Inst.

of

Tech.,

407 Gun-

saulus

Hall,

3140

S.

Michigan Ave., Chicago

. 16,

Illinois.

R.

J.

Rossheim,

U.S.Steel

Corp.,

525

William·

Penn

Place,

Pittsburgh

30, Pa.

G.

E.

Forsythe,

Numerical

Analysis

Research

Univ. of

Calif.,

Los

Angeles,

Calif.

A.

B.

Taub,

Digital

Computer

Lab.,

168 Eng.

Res. Lab., Univ of

Illinois,

Urbana,

Ill.

Wesley

S.

Melahn,

The

Rand

Corp.,

1700

N.

Main

St.,

Santa

Monica,

Calif.

Charles

L.

Baker, Douglas

Aircraft

Co.

,Santa

Monica,

Calif.

Ot~en

R.

Mock,

North American

Aviation,

12214

Lakewood

Boulevard,

Downey,

Calif.

Ramon

Alonso

and

Thomas

Conley,

Ballistic

~

search

tabs.,

Computing Lab., Aberdeen Prov-

ing Ground, Aberdeen,

Md.

Bill

L.

Wadell,

G.

M.

Giannini and Co., 918

E.

Green

St.,

Pasadena

1,

Calif.

R.

E.

Me

I'\·dn,

IBM

Corp.,

P.

O.

Box

390,Pough-

keepsie,

N.Y.

Raymond

Stuart-Williams,

Mil

ton

Rosenberg

and

M.

A.

Alexander,

Internat'l

Telemeter

Corp.,

2000

Stoner

Ave.,

Los

Angeles,

Calif.

John

Todd

and

Phillip

Davis, Computation

Gp.,

Nat'l

Bureau

of

Standards,

Washington 25,D.C.

J.

Pasta

and S.

Ulam,

Computer

Lab.,

Los

Ala-

mos

Scientific

Labs.,

P.

O.

Box

1663,

Los

AlnmtJs,

New

Mexico

Robert

C.

Miller,

Jr.,

and

Eruce

G.

Oldfiel~

Naval Ordnance

Test

Station,

China Lake,Calif.

Jules

I.

Schwartz and

Gus

S. Hempstead,

The

Rand

Corp.,

1700

N.

Main

St.,

Santa

Monica,

Calif

•

Richard

C.

Luke, Lockheed

Aircraft

Corp.,

Van

Nuys,

Calif.

John

I.

Derr,

The

Rand

Corp.,

1700

N.

Main

St.,

Santa

Monica,

Calif.

Hiram

G.

Martin,

Douglas

Aircraft

Co. ,Sa n

ta

Monica,

Calif.

II'\dn

D.

~reenl"ald,

Rand

Corp.,

1700

N.

~tIin

St.,

Santa

Monica,

Calif.

.

R.

K.

Gerlach and

D.

O.

Miles,

Rand

Corp.,

1700

N.

Main

St.,

Santa

Monica,

Calif.

W.

J.

Poppelbaum, Digi

tal

Computer

Lab.,

168

Eng. Res.

Lab.,

Univ. of

Illinois,

Urbana,

Ill.

John

R.

Bethke,

Bur~oughs

Corp.,

Research

Center,

Paoli,

Pa.

E.

R.

Beck, Bendix

Aviation

Corp.,

Research

Labs

Di

v.,

4855 Fourth Ave.,

Detroit,

~1ich.

-35 -

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110

•.

111.

112.

113.

J.

L. Smith

and

A.

Weinberger,

Nat'l

Bureau

of

Standards,

Washington 25,

D.

C.

E.

G.

Kogbetliantz,

IBM

Corp.,

P.

O.

Box

3~,

Poughkeepsie,

N.

Y.

Peter

Henrici,

Nat'l

Bureau of

Standards,

Washington 25,

D.

C.

R.

W.

Bemer, Lockheed

Aircraft

Corp.,

Van

Nuys,

Calif.

Bengt

Carlson

and

Max

Goldstein,

Los

Alamos

Lab.,

P.

O.

Box

1663,

Los

Alamos,

New

Mex.

Saul Gorn, Aberdeen Proving Ground,

Compu-

ting

Lab.,

Md.

Franz Hahn, 56

Electrical

Engineering

Bldg.

Univ. of

Illinois,

Urbana,

Ill.

Arthur

E. Hoerl,

E.

1.

DuPont de

Nemours

Co.

Experim~ntal

Station,

Wilmington 98, Dela.

William Miehle, Burroughs Research

Center,

Paoli,

Pa.

James

L.

Maddox

and Ralph

H.

Beter,

Philco

Corp.,

Philadelphia

44, Pa.

Richard

C.

Jeffrey,

MIT,

Cambridge, Mass.

Max

A.

Woodbury, George Washington Univ.,

Washington,

D.

C.

Gene

H.

Golub, Univ. of

Illinois,

Digital

Computer Lab., Eng. Res. Lnb.

(68),

Urbana,

Ill.

Arthur

E.

Hoerl,

E.

1.

DuPont de

Nemours

Co.,

Experimental

Station,

Wilmington 98, Dela.

Robert

II.

Braken, U.S. Naval Ordnance

Test

Station,

China Lake,

Calif.

I.

McNamee

and

E.

D.

Fullenwider,

U.

S. Nav-

al

Ordnance

Test

Station,

Corona,

Calif.

H.

Flatt,

Ballistic

Research

Labs.,

Aberdeen

Proving Ground, Computing

Lab.,

Aberdeen.

~.:d.

L

C.

Long,

Oak

Ridge

Nat'l

Lab.,

P.

O.

BoxP

Oak

Ridge, Tenn.

Charles

W.

Adams,

Westinghouse

Electric

Corp.,

Pittsburgh,

Pa.

David

E.

Muller,

Digital

Computer

Lab.,

168

Eng. Res.

Lab.,

Univ. of

Illinois,

Urbana,

Ill.

D.

A.

Huffman,

MIT,

Cambridge, Mass.

Manfred Kochen, Ins

t.

for Advanced

Stu

d y,

Princeton,

N.

J.

John

O.

Lilly,

U.S. Naval Ordnance

Test

Sta-

tion,

China

Lake.

Calif.

James L. McPherson, Bureau of

the

Census,

Dept. of

Commerce,

Washington 25,

D.

C.

W.

C.

Jacob, Univ. of

Illinois,

Dept.of

Ag-

ricul

ture,

208b Davenport

Ball,

Urbana,

TIL

Christine

Kris,

Univ., of Chicago, 6109

S.

Ellis

Ave., Chicago 37,

Ill.

Jack

C.

Me

I'\d n , Syracuse Univ., Syracuse,

N.Y.

C.

L.

Gerberich

and

W.

C.

Sangren,

Oak

Ridge

Nat'l

Labs,

P.

O.

Box

P,

Oak

Ridge, Tenn.

R.

J.

Arms,

D.

F.

Eliezer,

L.

O.

Gates,

Jr.

an~

D.

~.

Young,

Jr.,

U.S. Naval Proving

Ground, Dahlgren,

Va.

Elizabeth

Cuthill

and Ruth

M.

Davis, David

Tayl~r

Model

Basin,

Washington 7,

D.

C.

Robert

C.

Miller,

Jr.

and Ralph

G.

Selfridge

U.

S. Naval Ordnance

Test

Station,

China

Lake,

Calif.

W.

M.

Harris,

A.

O.

Smith

Corp.,

Milwaukee

1,

Wisc.

P.

A.

Zaphyr,

W~stinghouse

Electric

Corp.,

East

Pittsburgh,

Pa.

Stephen

E.

Wright,

Sperry

Rand

Corp.,

315

Fourth Avenue,

New

York

10, N.Y.

Nyles

V.

Reinfeld

(paper

cancelled)

Exreutive

(continued

on

page

37)

MAN

U S

We

are

interested

in

articles,

papers,

refer-

ence

information,

science

fiction,

and

dis-

cussion

relating

to

computers and

automation.

To

be

considered

for

any

particular

issue,

the

manuscript

should

be

in

our hands

by

the

fifth

of

the

preceding

month.

Articles.

We

desi

re

to

publish

articles

t

hat

are

factual,

useful,

understandable,

and

inte-

resting

to

many

kinds

of

people

engaged

in

one

part

or

another

of

the

field

of

computers and

automation.

In

this

audience

are

many

people

Ivho

have

expert

knOt~ledge

of

some

part

of

the

field,

but

Ivho

are

laymen

in

other

parts

of

it.

Consequently a

writer

should

seek

to

explai

n

his

subject,

and

show

its

context

and

signifi-

cance.

He

should

define

unfamiliar

terms,

or

use them

in

a

way

that

makes

their

meaning un-

mistakable.

lIe

should

identify

un

fa

mil

i a r

persons

lvi

th

a

feH

IVOrdS.

He

should

use

exam-

ples,

details,

comparisons,

analogies,

etc.,

whenever they

may

help

readers

to

understand

a

difficult

point.

He

should

give

data

s

upp-

orting

his

argument and

evidence

for

his

asser-

tions.

We

look

particularly

for

articles

that

explore

ideas

in

the

field

of

computers

and

automation,

and

their

applications

and i m

pli-

*---------------------------------------------

c R I P T S

cations.

An

article

may

certainly

be co n

tro-

versial

if

the

subject

is

discussed

reasonabl~

Ordinarily,

the

length

should

be 1000

to

4000

words. A

suggestion

for

an

article

should be

submitted

to

us

before

too

much

work

is

done.

Technical

Papers.

Many

of

the

foregoing

r

e-

quirements

for

articles

do

not

n e c e s

sa

r i 1 y

apply

to

technical

papers.

Undefined

techni-

cal

terms,

unfamiliar

assumptions, mathematics,

circuit

diagrams,

etc.,

may

be

entirely

appro-

priate.

Topics

interesting

probably

to

only

a few people

are

acceptable.

Reference

Information.

We

desire

to

print

or

reprint

reference

information:

lists,

ros~r~

abstracts,

bibliographies,

etc.,

of

use

to

com-

puter

people.

We

are

interested

in

m a

kin

g

arrangements

for

systematic

publication

from

time

to

time

of

such

information,

with

other

people

besides

our

own

staff.,

A

nJ.

0

new

h 0

would

like

to

take

the

responsibility

for

a

type

of

reference

information

should

write

us.

Fiction.

We

desire

to

print

or

reprint

fiction

which

explores

scientific

ideas

and

possibili-

ties

Qbout computing machinery,

robots,

cyber-

(continued

on

page

37)

*

------------------------------

""

-

36

-

114.

115.

116.

117.

118.

119.

120

•

121.

122.

123.

124.

125.

126.

127.

128.

129.

Ar.\l

~1F.F.TI~r.

(contInued

from,.p.age

35)

Services,

Cleveland

lu:

o.

Arnold

Siegel,

MIT,

Cambridge, Mass.

R.

E. Spero and

J.

Stuart,

Sperry

Gyroscope

Co.,

Creat

Neck,

N.Y.

J.

K.

Slap,

Northrop

Aircraft

Corp.,

Haw-

thorne,

Calif.

A.

Karen and

B.

Loveman, Reeves

Instrument

Co.,

215

E.

9lst

St.,

New

York

28,

N.Y.

Nathaniel

Macon,

Alabama

Polytechnic

Inst.,

Auburn, Ala.

R.

S.

Lehman

and

G.

II.

Weiss,

Ballistic

Re-

search

Labs.,

Aberdeen Proving Ground,

Md.

W.

F.

Atchison,

Georgia

Inst.

of Tech. Rich

. Computing

Center,

Atlanta,

Ga.

F. T.

!Vall

and L.

A.

lIiller,

Jr.,

Dept.of

CI,cmistry, Univ. of

Illinois,

Urbana,

Ill.

J.

Gardner,

R.

Page and

O.

Tiffan,

Willow

Run

Research

Center,

Univ. of Michigan, 'lpsi-

l~mti,

~lich.

R.

E.

Utman

and Margaret

II.

lIarper,

Reming-

ton Rand,

Inc.,

1624 Locust

St.,

Philadel-

phia,

Pa.

Anthony

G.

Oettinger,

Computation

Lab.,

!Iar-

vnr I

TJni

v.,

Cambridge 38, Mass.

W.

F. Bauer and

A.

Vazsonyi, Ramo-Woohlridge

Corp.,

5740 Arbor

Vitae

St.,

Los

Ange~es

45,

Calif

•

R.

P.

Beals,

Chrys

ler

Corp.,

Parts

Di

v.,

7000

E.

Eleven Mile Rd.,

Center

Line, Mich.

Louis

B.

Wadel

and

B.

Mackey,

Chance

Vou91

t

Aire"t··1f

t."

Tnc., P.

O.

Box

5907,

Dallas,

Tex.

Takeo Miura and

Toshiro

Numakura,

Hitachi

Central

Research, Kokubunju, Tokyo, Japan

L.

E.

Heitzer,

CONVAIR,

Fort

Worth, Texas.

M.

Yanowitch, Reeves

Instrument

Co., 215

E.

9lst

St.,

New

York 28,

N.Y.

_

ENTl

-

:;.----------

----------

...

~ANusrRIPT

NOTICE

(continued

from

page

36)

netics,

automation,

etc.,

and

their

implications,

nnd

which

at

the

same

time

is

a

good

story.

Ordinarily,

the

length

should

be 1000

to

4000

l'Jords.

Discuss

ion.

We

des

ire

to

print

in

"F

0

rum"

brief

discussions,

arguments,

announcements,

nelVS,

letters,

descriptions

of

remarkable

nel'V

developments,

etc.,

anything

likely

to

be

0 f

substantial

interest

to

computer

people.

Payments.

In

many

cases,

l'Je

make

small

token

payments

for

articles,

papers,

and

fiction,

if

the

author

lvishes

to

be

paid.

The

rat

e i s

ordinarly

!i¢ a word,

the

maximum

is

$20,

and

both depend

on

length

in

l'wrds, whether

printed

before,

whether

article

or

paper,

etc.

-37 -

THE

COMPUTER

DIRECTORY,

1956:

NOTICE

The

June 1956

issue

of

"Computers

and

Auto-

mation"

will

be

the

second

issue

of

"The

Com-

puter

Direc

tory".

Las t

year

tie

published

the

first

issue,

164

pages.

Our

present

plans

for

the

June 1956

directory

follow:

Part

1

of

the

directory

in

1956

will

be a

cumulative

"Roster

of

Organizations

in

the

Com-

puter

Field"

based

on

the

last

cumulative

ros-

ter

(published

December 1955,

containing

about

330

entries)

and

brought

up

to

date.

Entries

in

this

roster

will

be

free.

If

you

know

0 f

any

changes,

addi

tions,

or

correc

tions

ti

hi c h

should

be

made

in

the'

entries,

please

tell

u s.

Part

2

of

the

directory

will

be

the

secorrl

edition

of

"The Computing Machinery

Field:

Pro-

duc

ts

and

Services

for

Sale."

Over

600

entries

on

21

'pages appeared

in

the

first

edi

tion

i n

June

195;1;

a

considerable

increase

is

antici-

pated.

The

and

blank

forms\

will

be'sent

in

February,

to

suppliers

for

re-

view,

checking,

and

additions.

It

i~

expected

at

this

time

that

a nominal

charge

of

$6.00

an

entry

will

be

requested

from each

supplier

in

order

to

help

defray

the

cost

of

preparing

and

printing

the

directory;

but

if

the

charge

i s

not

paid,

t:1e

entry

may

still

appear

in

con-

densed form,

if

desirable

to

make

the

listi

n g

complete.

Part

3

of

the

directory

will

be

the

third

edition

of

the

Who's

Who

in.

the

Computer

Field

In

the

June 1955

issue,

about 7500

entries

ap-

peared

on

96

pages;

of

these

about 2600

l'J

ere

full

entries,

and

the

remainder

were

brief

en-

tries.

Our

present

plans

are

to

publish

only

new

or

revised

Who's

Who

information

in

the

June 1956

directory.

Blank forms

for

new

0 r

revised

entries

will

be

sent

in

Feb

ru

a

ry

0 r

March

to

all

computer

pe

a

pIe

lve

know

of.

It

is

expected

at

this

time

that

a nominal

charge

of

$2.00

an

entry

will

be

requested

from each

person

whose

entry

is

printed,

in

order

tohe~

defray

the

cost

of

preparing

and

printing

the

Who's

Who;

but

if

the

charge

is

not

paid,

a

b~ief

entry

may

~p~ear

in

condensed form

if

desirable

to

make

the

listing

complete.

The

main

reason

for

the

nominal

charges

mentioned above

is

that

we

look

on

the

direc-

tory

as a

service

to

m~ny

people

in

the

compu-

ter

field;

yet

so

far

it

has

not

paid

for

it-

self;

and

we

need

to

make

a compromise, p u

b-

lishing

at

least

some

information

about

every-

thing

that

should

appear

in

the

directory,

but

fuller

information

for

those

who

have

shared

directly

in

the

cost.

SP~CIAL

ISSUES

OF

~~C.OMPUTERS_

AND

AUTOMATIO~"

The

June

issue

of

"Computers and

Automa-

tion"

commencing

with

June, 1955,

is

aspecial

issue,

"The Computer

Directory."

For

details

about

the

te

r

directory,

see

"The

Computer

Directory.

1956:

Notice."

*

---------

~;

---------

*,

PATENTS

(continued

from

page

32)

and

Richard

H.

Richwine,

Greenfield,

Indiana

/ - /

An

undirectional,

one;~ay

automatic

spring

clutch.

2,723,080 / Daniel

L.

Curtis,

Venfse,

Calif.

/

Hughes

Aircraft

Company

/ A

device,

in

co

m-

bination

t'li th a

bistable

flip-flop,

for

ren-

dering

said

flip-flop

responsive

to

one 0 f

two

input

pulses

applied

simul

taneously

trere-

to.

2,723,081 / Harry

B.

Miller,

Wartdck,

R.

1.

/

Max

L.

Grant, Providence,

R.

I.

/

An

overrun

control

for

rotary-counter

registers.

2,723,312 / John

H.

McGuigan,

Net~

Providence.

N.

J.,

and Orlando

J.

Murphy,

New

York,

N.Y.

/

Bell

Telephone

Laboratories,

Inc.,

Netl'

YolK,

N·.

Y.

/ A magnetic

drum

dial

pulse

recording

and

storage

registers.

2,723,347 / Louis F. Mayle,

Fort

Wayne,

Ind. /

Farnsworth Research

Corporation,

Fort

Wayne,

Ind.

/ A

pulse

keying

circuit

for

a p Ot\1

er

amplifier.

2,723,352 /

Willia~

C.

Sealey,

Wauwatosa, Wisc.

/ Allis-Chalmers Manufacturing Co.,

Milt\'aukee,

Wisconsin / A polyphase

regulating

system for

obtai~ing

balanced

voltag~s.

2,723,353 / Charles F.

Spitzer,

Syracuse,

and

Robert T. Gordon, Camillus,

N.Y.

/

General

Electric

Co. / A

non-linear

resonant

trigger

circuit.

2,723,355 / Robert E. Graham, Chathorn

Tow

n -

ship,

Morris County, N.J. /

Bell

Telephone

Laboratories,

Inc.,

New

York,

N.Y.

/ A diode

gate

circuit.

2,723,365 / Charles

R.

Williams, IIawtharne,Calif.

/ Northrop

Aircraft,

Inc.,

H31'lthorne,

Calif./

A

sectional

read-out

tube

and

circuit.

-

END

-

-38 -

BULK

SUBSCRIPTION

RATES

These

rates

apply

to

subscriptions

coming

in

together

direct

to

the

publisher.

For example,

if

5

subscriptions

come

in

together,

the

sav-

ing

on

each

one-year

subscription

will

be

24

percent,

and

on

each two-year

subscription

will

he

31

percent.

The

bulk

subscription

rate

s,

depending

on

the

number

of

simul taneous s u

b-

scriptions

received,

follow:

Bulk

Subscription

Rates

(United

States)

Number

of

Sirnul taneous

Subscriptions

Rate

for

Each

Subscription,

and

Resulting

Saving

to

Subscriber

One

Year

Two

Year

10

or

more

$3.80,

31%

$6.60,

3-ro,;6

5

to

9 .4.20,

24

7.25,

31

4

4.60,

16

8.00,

24

3

5.00,

9

8.80,

16

2

5.25,

5

9.55,

9

For Canada,

add

50

cents

for

each

year;

0 u

t-

side

of

the

United

States

and

Canada, add $1.00

for

each

year.

COM

P U

T,E

R

Applications

and

Development

Computing

Services

Engineering

and

scientific computation

Analog Digital

Computers

Systems Development

Analog

and

Digital

Systems analysis

and

synthesis

Logical organization

Circuit design

These

are

full

time positions offering fascinating

and

varied work

and

educational benefits for

graduate

study.

Contact:

J.

A.

Metzger

ARMOUR

RESEARCH

FOUNDATION

of

Illinois Institute

of

Technology

10

West

35th

Street

Chicago 16. Illinois

PUBLICATIONS

p

34;

LINEAr.

pnOGrA~l~lING

AND

CO~]PtJTEnS.

Reprint

of

tt"o

articles

by

Chandler

Davis,

in

July

and

August

1955

"Computers

and

Auto-

mation".

A

clear,

t'iell-wri

tten

in-

troduction

to

linear

programming,wiili

emphasis

on

the

ideas.

•

•••

$1.20

P

20:

THE

COMPUTER

DIRECTORY,

1955.

164

pages,

7500

Who's

Who

entries,

300

Organization

entries,

and

600

entries

of

Products

and

Services

for

Sale

in

the

Computer

Field;

250,000'

words

of

condensed

factual

informa-

tion

about

the

computer

field,

June

1955

issue

of

"Computers

and

Auto-

mation.

"

••••

$4.00

P

32:

SYMBOLIC

LOGIC,

by

LEWIS

CAR-

ROLL.

Reprint

of

"Symbolic

Log

i

c,

Part

I,

Elementary,"

4th

editicn,

1897,

240

pages,

by Lewis

Carioll

(C.

L.

Dodgson).

Contains

Lewis

Carroll's

inimi

table

and

entertaining

problems

in

symbolic

logic,

his

method

of

so-

lution

(now

partly

out

of

date),

and

h

is

sketches

of

Parts

II

and

III,

Hhich

he

never

wrote

since

he

die

d

in

1898.

•

•••

$2.50

P

25:

NlJMBLES

NUMBER

PUZZLES

FOR

NIMBLE

MINDS.

Report.

Con

t a

ins

collection

of

puzzles

like:

TRY

I

II

A V E I and T R A I N

+

TIlE

S E

FUN

your

WIT

S

=TWVAS

WASE

ENTNS

WYE

= VIF

In

fact,

you

can

also:

90893

85202

44393

29081

(Solve

for

the

digits--

each

letter

stands

for

just

one

digit

o

to

9)

All

are

new

numbles,

additions,

mul-

tiplications,

etc.;

some

easy,

some

hard.

Each wi

th

th'O

messages,

0 n e

open,

one

hidden.

HInts

for

solutIon.

Good

exercises

in

logical

reasoning

.

••

• •

$1.00

.'

--------MAIL

TIllS

COUPON----------

or

copy

it

Edmund C.

Berkeley

and

Associates,

815

Washington

St.,

R142,

Newtonville

60,

Mass.

Please

send

me

publications

circled

and

your

announcement

of

publications:

2D

25 32 34

I

enclose

$

in

full

payment.

(Add 10¢

per

i tern

to

cover

cos

t 0 f

handling

and

mailing.)

It

is

under-

stood

that

if

I

am

not

satisfied,

I

may

re

t

urn

any i tern

tV

i

th

in

five

days

after

receiving

it,

and

you

will

at

once

refund

my

money.

My

name

and

address

are

attached.

FERRANTI

HIGH

SPEED

'APE

READER

('

"

.:c

::"

(

The Ferranti High Speed Tape Reader accelerates

to full speed within 5 milliseconds

and

stops within,

3 milliseconds.

It

has been in use

at

leading computer

installations for over two years

and

has achieved a sound

reputation for simplicity

and

reliability

in

regular operat!on.

-I-\i'

(

I)

Mark

II

model reads at speeds

up

to 200 characters

per

second, and stops the tape from full' speed within a character

position-

within .03 inch:

The

tape

is

accelerated to full speed again in 5 milliseconds

and

the following character

is

ready for reading within 6 milliseconds of

rest position.

(2)

Mark IIA model reads

at

speeds

up

to 400 characters per second,

and stops within

.1

inch.

'1#;£!-",,1

Both models read either 5 level, 6 level or 7 level tape

by

simple adjustment of an external lever.

WiMt4!WThe

tape is easily inserted without complicated threading.

Lap

or

butt

splices are taken without ally difficulty.

The

same tape may be

passed' thousands of times without appreciable tape wear.

The

optical

system has no lenses or mirrors to get

out

of

alignment. Friction drive is

independent of sprocket hole spacing.

LARGE

OUTPUT Amplifiers are included for each channel, including

a special squaring circuit for the sprocket hole signal.

Output

swing

between hole

and

blank

is

greater than 20 volts.

Dimensions:

91(

x 11 V2" x 111,4"

Weight:

37

Ibs.

-

39

-

For use with long

length~

6f

tape

up

to 1000 feet, spooling

equipment operating up to 40 inches

per

second for take-up

or supply is available separately.

FERRANTI

ELECTRIC,

INC.

30

Rock.f.lI.r

Plaza

New

York 20, N.

Y.

Forum

ANOTHER

PROGRAMMING

FAILURE

Bill

Danch

Munich,

Germany

"I'll

have

to

get

a

new

computer cook book. This

upside

down

cake

recipe

is

coming

out

right

side

up

again:"

-

tiD

-

MAKE

YOUR

OWN

BABY

GENIUS

COMPUTERS

WITH

GENIAC

Electric

Brain

Construction

Kit

No.1

Q3

Diagram

of

the

versa-

•

~

: :

.'

tile

multiple

switch,

'.

"

~

"

.'

.\

which

can

be

assembled

.':::'

......

:

••.

,

to

make

any s

wit

c h

.'.

:.

'J'

combinations

from

16

,

••••.

:.

:

"':".:

.

decks

of

2

positions,

~

10

decks

of

3 p

os

i-

tions,

etc.,

to

2

decks

of

16

positions.

This

ki

t

is

an

introduc

tion

to

the

design

of

arithmetical,

logica

1 ,

reas

oning,

comru

ting,

puzz

le-s

01

viJJ],

and

game-playing

circuits.

It

i s

simple

enough

for

intelligent

boy

s

to

assemble,

and

yet

is

instructive

to

computer

men

because

it

ShOHS

how

many

kinds

of

computing

and

reason-

ing

circuits

can

be

made

fromsimple

components.

Wi

th

this

ki

t and

64-page

ma

n u

aI,

you

can

eas

ily

make

over

30

sma

1 1

elec

tric

brain

machines

that

exhibit

intelligent

behavior.

Each

runs

on

one

flashlight

battery.

All

connec-

tions

uith

nuts

and

bolts;

no

s 0

1-

dering

required.

Price,

$17.95

(add

80t

for

shipment

in

U.

S.

\'Ves t

of

Mississippi,

$1.80

for

s

hip

men

t

outside

U.

S.).

If

not

satisfactory,

returnable

in

seven

days

for

full

refund.

A few

of

the

machines

you

can

make:

Logic

Mnchines:

Reasoning,

Syllogism

Machine,

Intelligence

Testing.

Game-

pI

nyi

ng

Mach

ines:

Nim,

Ti

t-tat-toe

•

Ari

thmetic

Machines:

Adding, S u

b-

tracting,

Multiplying,

Divid

in

g,

Carrying,

etc

•

.!d;:yp.1Q.grapldc

~bchines:

Secret

Coder and

Decoder,

Combination

Locks.

Puzzle

Machines:

The

Space

Ship

Airlock,

The Fox lIen Corn and

Hired

Mnn,

Douglas

Macdonald's

Will,

The Uranium

Ship

and

the

Space

Pir-

ates.

--------Mail

this

Request--------

or

a Copy

of

It

Berkeley

Enterprises,

Inc,

~13

Avenue

of

the

Americas,

R143

Nell

York 11,

N.

Y.

Please

send

me

Geniac

Kit

llo.

1

and

Manual.

Price,

$17.95

(add 80¢

for

shipment

in

U.

S.

west

of

Mississippi,

$1.80

for

shipment

outside

U.

S.

)

1

enclose

in

full

payment.

(If

in

good

condi

tion,

it

is

returnable

in

seven

days

for

full

refund.)

My

name

and

address

are

attached.

Qiji(}/C lOOIC

at

your

data

with

a

IpOTTERI

DIGITAL

RECORDER

You

can

"look

into"

your

digital

data-processing

equipment

conveniently

and

with

complete

assurance

of

accuracy,

using

a

Potter

Model

960

Digital

i

Recorder.

Permanent,

easily-stored

records,

capable

of

quick

visual

interpretation,

-

may

be

made

of

electrical

phenomena

without

interrupting

the

operation

of

equipment

being

monitored.

SPECIFICATIONS:

NO.

OF

CHANNELS

Up

to

32

on

standard

models;

more

on

special

order

RECORDING

SPEEDS

150

measurements

per

second-continuous;

up

to

10

per

second-

intermittent

CHANNEL

SENSITIVITY

-25

voits-print; 0 volts-no

print

RECORDING

MEDIUM

"Teledeltos"

sensitized

paper;

standard

400-foot

roll

provides

up

,

to

52,000

measurements

PAPER

TRAVEL

Adjustable

from

front

panel,

21

/2"

to

20"

per

second;

continuous

or

intermittent

Potter

Digital

Recorders

are

being

used

daily

to

monitor

computer

operation,

digital

data-processing

equipment

and

other

applications.

With

minor

modifications,

standard

recorders

can

be

provided

with

different

input

voltages,

higher

recording

speeds

and

more

channels.

Write

today

describing

your

"quick

look"

data

recording

-requirements.

Let

our

engineers

suggest

an

appropriate

solution;

no

obligation,

of

course.

POTTER

INSTRUMENT

COMPANY,

INC.

115

Cutter

Mill

Road,

Great

Neck,

N.

Y.

-41 -Visit

Us

At

The

I.R.E.

Show. Booth 346.

Computer

Avenue

COMPUTERS

A,ND

AUT'OMATION -

Back

'Copies

ARTICLES,

PAPERS,

ETC.

January,

1955:

Statistics

and

Automatic

Computers

--

Gordon

Spenser

Eastern

Joint

Computer

Conference,

Philadelphia,

Dec.

8-10,1954

--

Milton

Stoller

The

Digital

Differential

AnalY2er

--

George F.

Forbes

A

Small

High-Speed Magnetic

Drum

-

M.

K.

Tay-

lor

An

Inside-Out

Magnetic

Drum

--

Neil

Macdonald

February:

Problems

for

Students

of

Computers

John

W.

Carr,

III

Recogni2ing

Spoken Sounds by

Means

of

a

Com-

puter

--

Andrew

D.

Booth

The

Significance

of

the

New

Computer

NOR

C

W.

J.

Eckert

The

Finan-8eer

--

E.

L. Locke

Approaching Automation

in

a

Casualty

Insurance

Company

--

CarlO.

Orkild

March:

Question

--

Isaac

Asimov

--COmputers

and Weather

Prediction

--

Br

u c e

Gilchrist

Random Numbers and

Their

Generation

-Gord 0 n

Spenser

Problems

lnvol

ved

in

the

Application

of

Elec-

tronic

Digital

Computers

to

Business

Opera-

tions

--

John

M.

Breen

Computers

to

Make

Administrative

Decisions?

--

Hans

Schroeder

April:

Thinking

Machines and

Human

Personality--

Elliot

L.

Gruenberg

Marginal

Checking

--

An

Aid

in

Preventive

Main-

tenance

of

Computers

--

J.

~~lvin

Jones

~:

Reliability

in

Electronic

Data

Process

0 r s

--

William

B. Elmore

Numerical

Representation

in

Fix

e d -

Poi

n t

Computers

--

Beatrice

H.

Worsley

Automation

--

A

Report

to

the

UAW-ClO

Economic

and

Collec.ti

ve

Bargaining

Conference

The

Skills

of

the

American Labor

Force

--

James

P.

Mi

tchell

Automation

Puts

Industry

on Eve

of

Fantas

tic

Robot

Era

--

A.

H.

Raskin

The

Monkey

Wrench

--

Gordon R.

Dickson

June:

THE

COMPUTER

DIRECTORY,

1955 (164

pages):

--Part

1:

Who's

Who

in

the

Computer

Field

Part

2:

Roster

of

Organi2ations

in

the

Co

m-

puter

Field

Part

3:

The

Computer

Field:

Prod

uc

ts

and

Services

for

Sale.

July:

Mathematics,

the

Schools,

and

the

Oracle--

Alston

S.

Householder

The

Application

of

Automatic

Computing

Equip-

ment

to

Savings

Bank

Operations

-R.

Hu

n t

BrOliTn

The Book

Reviewer

--

Rose

Orente

Linear

Progrannning and

Computers,

Par

t I

--

Cpandler

Davis

August:

The

Automation

of

Bank Check

Processing--

R. Hunt Brown

Linear

Programming and

Computers,

Par

t

II

Chandler

Davis

Justifying

the

Use

of

an

Automatic

Computer

Ned

Chapin

Charting

on

Automatic

Data

Process

ing

Sys tern s

--

Harry

Eisenpress,

James L. McPherson, and

Julius

Shiskin

A

Rotating

Reading

Head

for

Magnetic

Tape

an

d

Wire

--

National

Bureau

of

Standards

Some

Curiosities

of

Binary

Arithmetic

Usef

u 1

in

Testing

Binary

Computers

--

Andrew

D.

Booth

September:

A

Big

Inventory

Problem

and

the

IBM

702

-

Neil

Macdonald

Pq,lllications

for

Business

on

Automatic Computers:

A

Basic

Listing

-

Ned

Chapin

Franchise

--

Isaac

Asimov

Automatic

Coding

for

Digltal,

Computers -

G.

M.

Hopper

Automatic

Programming: The A-2

Compiler

System

--

Part

1

October:

The

Brain

and

Learned

Behavior

--

Dr.

Harry

F.

Harlow

Automatic Programming: The A-2

Compiler

Syste

m

--

Part

2

Who

Are Manning

the

New

Computers?

--

John

M.

Breen

November:

Automatic

Answering

of

Inquiries

-

L.

E.

Griffith

Found

--

A "Lo,st"

Moon

--

Dr.

Paul

Herget

Mister

Andrew Lloyd

--

R.

W.

Wallace

December:

Digital

Computers

in

Eastern

Europe

Alston

S.

Householder

Automatic

Airways

--

Henry T. Sinunons

Roster

of

Organi2ations

in

the

Computer

Field

(cumulative)

January,

1956: Machines and

Religion

--

E

11

i 0 t

Gruenberg.

Automatic Coding

Techniques

for

Business

Oat

a

Processing

--

Directions

of

Developmen

t

Charles

W.

Adams,

Bruce

Moncreiff

What

is

a Computer?

--

Neil

D.

Macdonald

REFERENCE

INFORMATION

(in

various

issues):

Roster

of

Organi2ations

in

the

Computer

Field

/

Roster

of

Automatic

Computing

Services

/

Ros-

ter

of

Magadnes

Computers

and

Automation / Automatic

Computers:

List

/.Auto-

matic

Computers:

Estimated

Commercial

Popula-

tion

/

Automatic

Computing Machinery:

List

0 f

Types / Components

of

Automati

c

Computing

Machinery:

List

of

Types /

Products

and

Se

r-

vices

in

the

Computer

Field

/ Who's

Who

in

the

Computer

Field

/

Automation:

Lis

t 0 f

Outstanding

Examples /

Boo

k s . and 0

the

r

Publications

/

Glossary

/

Patents

I

Titles

and

Abstracts

of

Papers

BACK

COPIES:

Price,

if

available,

$1.25

eac

h,

except

June,

1955,

$4

.00

~

Vol.

1,

no.

1,

Sept.

1951,

to

vol.

1,

no.

3,

July,

1952: 0 u t 0 f

print.

Vol.

1,

no.

4,

Oc

t.

1952:

in

pri

n

t.

Vol.

2,

no.

1,

Jan.

1953,

to

vol.

2,

no.

9,

Dec. 1953:

in

print

ex

c e p t

Mar

c

h,

no.

2,

May,

no.

4_.

and

July,

no.

5.

Vol.

3,

no.

1,

Jan.

1954,

to

vol.

3,

no.

10,

Dec. 1954:

in

print.

Vol.

4,

1955,

no.

1

to

12,

in

print.

A

subscription

(see

rates

on page 4)

may

be

spec-

ified

to

begin

with

the

current

month's

or

the

preceding

month's

issue.

WRITE

TO:

-

42

-

Berkeley

Enterprises,

Inc.

Publisher

of

COMPUTERS

AND

AUTOMATION

513 Avenue

'of

the

Americas

New

York

11,

N.

Y.

2

ENGINEERS

ME

or

EE

;

DIGITAL

l

COMPUTER

DEVELOPMENT

Excellent

opportunity

to

join

an

ex-

panding,

stable

company,

with

an

outstanding

position

in

the

precision

electronic'

control

industry.

Opening

now

available

for

an

en-

gineer

able

to

assume

responsibility

for

the

d

eve

lop

men

t

of

complete

systems

for

fire

control

and

guid-

ance,

or

major

portions

of

such

sys-

tems.

Work

will

include

research

and

development

in

the

field

of

complex

analog

or

digital

computers.

A

degree

in' electrical

or

mechanical

engineering

is

required

or

the

equi-

valent

in

experience.

MISSILE

&

WEAPONS

CONTROL

SYSTEMS

STUDIES

Enjoy

full

use

of

your

skill

and

imagina

tion

and

friendly

professional

give-and-take

with

top

men

in

the

field

of

electro-mechanical precision

equipment.

You will

perform

studies

airborne

weapons control

and

guid-

ance

systems

with

the

object

of

de-

termining

requirements,

feasibility,

performance

and

specifications

of

computers

and

overall

systems.

A

degree

in

physics, ME

or

EE

-

or

the

equivalent

in

experience -

is

required.

Must

be able

to

handle

problems

in

such diversified fields

as

digital

computers,

digital

data

transmission

systems,

logic

counting

and

conversion circuits,

high-preci-

sion

gyro

and

gimbal

structures,

dynamic

behavior

of

missiles,

align-

ment

of

inertial

platforms

for

guid-

ance

systems

and

fight

evaluation

of

guidance

systems

and

instrumenta-

tion. Send

resume

in

confidence

to

Technical Personnel Dept. 2.500

ARMA

Division

American

Bosch

Arma Corp.

Roosevelt Field, Garden City

Long

Island, N. Y.

-43 -

Simulation

Engineering

To

conduct

system

studies

on

RO<4{ETS,

MISSILES,

and

HIGH

SPEED

AIR-

CRAFT

using

the

latest

high

speed

digital

computers

and

the

most

up-to-

date

analog

computers.

Attractive

opportunities

exist

for

both

experienced

personnel

and

recent

gradu-

ates

in

the

following

fields:

•

HIGH

SPEED

DIGITAL

LOGIC

•

AUTOMATIC

CONTROL

THEORY

•

ANALOG

COMPUTATION

TECHNIQUES

Please

Contact

Professional

Employment

1VI~.it

»

BALTIMORE

3.

MARYLAND

ADVERTISING

IN

~~COMP'UTERS

AND

A'UTiOMATION"

~ffemorandum

from

Berkel

ey

Enterprises,

Inc.

Publisher

of

CO~lPUTERS

,\ND ATJTO\fATION

36 West 11

St.,

New

York

11,

N.

Y.

1.

What

is

"COMPUTERS

AND

AUTOMATION"?

It

is

a

montly

magazine

containing

articles,

papers,

and

reference

information

computing

machinery,

robots,

automatic

control,

cybernet-

ics,

automation,

etc.

One

important

piece

of

reference

information

published

is

the

"Ros t e r

of

Organizations

in

the

Field

of

Computers

and

Automation".

The

basic

subscription

rate

is

$5.50

a

year

in

the

Uni

ted

States.

Sin

g 1 e

copies

are

$1.25,

except

June,

1955, "The Com-

puter

Directory"

(164

pages,

$4.00).

For

the

titles

of

articles

and

papers

in

recent

issues

of

the

magazine,

see

the

"Back

Copies"

page

in

this

issue.

2.

What

is

the

circulation?

The

circulation

includes

2000

subscribers

(as

of

Dec.

10):

over

300

purchasers

of

individual

back

copies;

and

an

estimated

2500

nonsubscribing

readers.

The

logical

readers

of

COMPUTERS

AND

AUTOMATION

are

people

concerned

wi

th

the

field

of computers

and

automation.

These

include

a

great

number

of

people

Ivho

tvill

make

recommendations

to

their

organizations

about

purchasing

computing

ma-

chinery,

similar

machinery,

and

components,

and whose

decisions

may

involve

very

substan-

tial

figures.

The

print

order

for

the

Jan.

issue

was 2500

copies.

The

overrun

is

largely

held

for

eventual

sale

as

back

copies,

and

in

the

case

of

several

issues

the

over

ru

n

has

been

exhausted

through

such

sale.

3.

What

type

of

advertising

does

COMP

UTE

RS

AND

AUTOMATION

take?

The

purpose

of

the

mag-

azine

is

to

be

factual

and

to

the

point.

For

this

purpose

the

kind

of

advertising

Ivanted

is

the

kind

that

answers

questions

factually.

We

recommend

for

the

audience

that

we

reach,

that

advertising

be

factual,

useful,

interesting,

understandable,

and

new

from

issue

to

issue.

We

reserve

the

right

not

to

accept

advertisfug

that

does

not

meet

our

standards.

4.

What

are

the

specifications

and

cost

0 f

advertising?

COMPUTERS

AND

AUTOMATION

is

pub-

lished

on

pages

8~"

x

11"

(ad

size,

7"

x 10")

and

produced

by

photooffset,

except

that

print-

ed

sheet

advertising

may

be

inserted

and

bound

in

with

the

magazine

in

most

cases.

The

clos-

ing

date

for

any

issue

is

approximately

th

e

10th

of

the

month

preceding.

If

possible,

the

company

advertising

should

produce

final

copy.

For

photooffset,

the

copy

should

be

exact

1 y

as

desired,

actual

size,

and

assembled,

and

may

include

typing,

writing,

line

drawing,

printing,

screened

half

tones,

and any

othe

r

copy

that

may

be

put

under

the

ph

oto

0 f f

set

camera

without

further

preparation.

Unscreened

photographic

prints

and any

other

copy

requirfug

addi

tional

preparation

for

photooffset

shou

1 d

be

furnished

separately;

it

will

be

prep

ared,

finished,

and

charged

to

the

advertise

rat

small

addi

tional

costs.

In

the

case

of

printed

inserts,

"a

sufficient

quantity

for

the

iss

u e

should

be

shipped

to

our

printer,

addres

s

on

request.

Display

advertising

is

sold

in

units

of

a

full

page

(ad

size

-("

x

10",

basic

rate,

$190)

two-

thirds

page

(basic

rate,

$145),

and

half

pag

e

(basic

rate,

$97);

back

cover,

$370;

in

sid

e

front

or

back

cover,

$230.

Extra

for

color

red

.

(full

pages

only

and

only

in

certain

posi

tions),

35%.

Two-page

printed

insert

(one sheet) , $32U;

four-page

printed

insert

(two

sheets),

$590.

Classified

advertising

is

sold

by

the

t\'

0 r d

(60

cents

a word) wi

th

a minimum

of

20

words.

5.

Who

are

our

advertisers?

Our

advertisers

in

recent

issues

have

included

the

follow

in

g

companies,

among

others:

Ampex

Corp.

Arnold

Engineering

Co.

The

Austin

Co.

Automatic

Electric

Co.

Bendix

Aviation

Corp.

Cambridge

Thermionic

Corp.

Epsco,

Inc.

Ferranti

Electric

Co.

Ferroxcube

Corp.

of

America

General

Electric

Co.

Hughes

Research

and Development Lab.

International

Business

Machines

Corp.

Lockheed

Aircraft

Corp.

Logistics

Research,

Inc.

The Glenn L.

Martin

Co.

Monrobot

Corp.

Norden-Ketay

Corp.

Northrop

Aircraft,

Inc.

George

A.

Philbrick

Researches,

Inc.

Potter

Instrument

Co.

Raytheon Mfg. Co.

Reeves

Instrument

Co.

Remington Rand,

Inc.

Republic

Aviation

Corp.

Sprague

Electric

Co.

Sylvania

Electric

Products,

Inc.

-44 -

ROBOT

SHOW

STOPPERS

From

time

to

time

you

may

nee

d t 0

help

organize

a

display

in

a

business

show

including

some

device

that

you

hope

will

"STOP"

every

perso

n

at-

tending

the

show and make

him

notice

your

display

- a

device

which

may

be

called

a

"SHOW

srOPPER".

In

addition

to

publishing

the

maga-

zine

"COMPUTERS

AND

AUTOMATION",

we

have

for

five

years

been

developing

and

constructing

"ROBOT

SHOW

STO

P-

PERS",

small

robot

mach

i n

est

hat

res

pond

to

the

ir

environmen

tan

d

behave

by

themselves.

TtfO

of

them

are:

RELAY

MOE:

A machine

that

will

play

the

game

Ti

t-Tat-Toe

with a human

being,

and

either

win

or

draw

all

the

time,

or

(depending

0

nth

e

setting

of

a

switch)

will

some-

times

lose,

so

as

to

make the

game

more

interesting

for

the

hum

an

being;

FRANKEN:

A

mechanical

rat

that

will

explore

a maze,

find

"food",

and

learn

the

path

through;

the

maze

may

be

set

up by any

pers

on in the

audience,

using

little

partitions

in

any

way

that

he

wants

to.

Bes

ides

these

we

have

other

sma

11

robots

finished

or

under

development.

Some

of

these

machines

have

been

on

the

front

covers

of

the

magazine

s

"Scientific

American"

and

"R

a d i 0

Elec

tronics".

These

machines

may

be

rented

for

shows

under

certain

con-

ditions;

also,

modifications

of

the

small

robots

to

fit

a

particul

CJ

r

purpose

are

often

possible,

such

as

use

of

particular

components,

d i

s-

play

of

particular

slogans,

etc.

---------~----~--------~--------~

To:

Berkeley

Enterprises

Inc.,

36

West

11

St.,

R143,

New

York

11,

N.

Y.

Please

send

us more

informa

ti

0 n

about

your

ROBOT

SHOW

STOPPERS.

The

advertising

application

we

have

i n

mind

is:

______________________

__

From:

(Organization)

(Address)

(Filled

in

by:

Name,

Title,

Date)

The

most

advanced

.

developments

in

electronics

are

being

made

in

the

sphere

of

airborne

radar

and

control

systems

because

of

military

emphasis.

INGENIOUS

PACKAGING

Further applications

of

electrom6chanical techniques

in

these fields

are creating new openings

in

the

Systems Division

of

Hughes Research

and

Development Laboratories.

Engineers

who

have demonstrated ingenuity and inventive

ability will find interest in areas

of

work

that

call for

devising reliable, maintainable, manufacturable designs

for

precision equipment developed

at

Hughes Research

and

Development Laboratories.

The

design of this equipment, manufactured

at

Hughes,

involves ·mechanical, electromechanical, electronic, micro-

wave

and

computing problems. Design also requires the

use

of

such advanced techniques as sub miniaturization,

unitized "plug-in" construction, with emphasis

on

design

for

volume production. Knowledge

of

electronic components,

materials, finishes

and

military specifications is useful.

SCIENTIFIC

STAFF

RELATIONS

HUGHES

RESEARCH

AND

DEVELOPMENT

LABORATORIES

Culver City,

Los

Angeles County, California

-45 -

ADVERTISING

INDEX

The

purpose of

COMPUTERS

AND

AUTOMATION

is

to

be

fac

tual,

useful,

and

unders

tandable.

For this pur-

pose,

the

kind of

advertising

we

desire

to

publish

is

the

kind

that

answers

questions,

such

as:

What

are

your

products?

What

are

your

services:

And

for

each

product,

What

is

it

called?

What

does

it

do?

Ilot

....

t

....

ell

does

it

t

....

ork?

What

are

its

m a i n

spec

if

ications

'?

Following

is

the

index

and

a

summary

of

advertise-

ments.

Each

i

tern

contains:

Name

and address 0 f

the

advertiser

/ subj ec t of

the

/

page number where

it

appears /

CA

number

in

case

of

inquiry

(see note below)-

Arma

Division,

American Bosch

Arma

Corp.,

Roose-

vel t

Field,

Garden

Ci

ty,

L.

I.,

Net~

York

/ En-

gineering

Opportunities

/ Page

43

/

CA

No.

79

Armour

Research Foundation

of

Illinois,

Institute

of

Technology, Technology

Center,

'Chicago 16,

Ill.

/ Help

Wanted

/ page 38 /

CA

No.

80

The

Arnold Engineering Co., Marengo,

Ill.

/ Tape-

t'l/Ound

Bobbin Cores / page

33

/

CA

No.

81

Automatic

Electric

Company,

1033

W.

Van

Buren

St~

Chicago,

Ill.

/

Circuits

/ page 5 /

CA

No.

82

Berkeley

Enterprises,

Inc.,

513

Ave.

of

the

Amer-

icas,

Net~

York

10,

N.

Y.

/

Publications,

Gen:i£c,

Robot

Show

Stoppers

/ pages 39, 41,

45

/

CA

No.

83

ComplJ.1.~~rs

and Automation, 513

Ave.

of

the

Amer~as.

Net~

York 10,

N.

Y.

/ Gateway

to

Science,

Back

Copies,

Advertising

/ pages 32, 42,

44

/

CA

No.

84

Ferranti

Electric

Inc.,

~O

Rockefeller

Plaza,

New

York

2~,

N.

Y.

/ High Speed Tape Reader / page

39 /

CA

No.

85

.

Ferroxcube

Corp.,

East

Bridge

St.,/Saugerties,

N.

~

/ Magnetic Core

Materials

/ page

36

/

CA

No.

86

:Iughes Research and Development

Laboratories,

Cul-

ver

City,

Calif.

/ Help Wanted / page

45

/

CA

No.

87

Lockheed

Aircraft

Corp.,

California

Div.,

Burbank,

Calif.

/

Missile

Systems Mathematics / page

47/

CA

No.

88

The

Glenn

L.

Martin

Company,

Baltimore

3,

Md.

/

Simulation

Engineering

/ page 43 /

CA

No.

89

Potter

Instrument Co.,

115

Cutter

Mill Rd., Great

Neck,

N.

Y.

/

Digital

Recorder / page

41

/

CA

No.

90

Remington Rand,

Inc.,

315

4th

Ave.,

New

York

10,

N.

Y~

/ Univac / page 2 /

CA

No.

91

Sylvania

Electric

Products,

Inc.,

175

Great

Arrow

Ave

••

Buffalo

7,

N.

r.

/

niin

Triode / page

48/

CA

No.

92

READER'S

IN'QUIRY

If

you wish

more

information

about

any

products

or

serVIces mentioned

in

one

or

more

of

these

advertisements,

you

may

circle

the

appropriate

CA

Nos.

on

the

Reader's

Inquiry

Form

below

and

send

that

form

to

us

(we

pay

postage;

see

the

instructions).

We

shall

then

forward your

in-

quiries,

and

you

will

hear

from the

advertisers

direct.

If

you

do

not

wish

to

tear

the magazine,

just

drop

us

a

line

on

a

postcard.

*--------------------------------------------*-------------------------------------------*

READER'S

INQUIRY

FORM

Enclose

form

in

envelope:

~

Paste

label

on

envelope:~

---

--

---

- -

-.,-

-

--

--..,.~-

------

---.-.

-- - - -

...

-

--

- - -

--

---

...

---

--

- - - -

--

z

~

0

"A--

."

0

.."

0

~

G)

1ft

'"

......

QI

~

ell

CD

m

~

::II;

m

r-

m

-<

CIt

~

3::

...

z

.

p;

c:

~

;JIll

-<

:;;

m

to

z

CD

....

w m

3::

>

;::

m

0

0-

~

.."

z

~

-a

QI

~

z

CD

Vi

a.:

CD

CIt

~

...

m

,!II

to

-<

...

"<

0

...

z

..

:r

~

CIt

!l

...

~

..

CD

:z:

...

:<

tD

C

C/')

-

Z

m

C/')

;::c

m

"

r-

-<

r

~

J

O:J

m

r-

ZVl

mO

~p

w

-<~

0-

::0"'0

?,'

r

z·

• !Ie

:<iU

."

ITI

;:u

::!!

3:

::0

VI

=i

-t

0

Z

s:

0

VI

-

en

Q)

0

READER,'S

INQUIRY

FORM

Name

(please

print) ........................................................................................................................... .

Your

Address?

..................................................................................................................................... .

Your

Organization?

.......................................................................................................................... .

Its

Address?

.................................................... _ .................................................................................. .

Your Title ? ..................................................... , .................................................................................... .

Please

send

me

additional

information

on

the

following subiects

for

which I

have

circled

the

CA number:

1 2 3 4 5

26

2J

28

29

30

51

52

53

54

55

76

n

78

79

80

101

102

103 104

105 126 127 128

129

130

6 7 8 9

10

31

32

33 34

35

56 57 58

59

60

81

82

83 84

85

106

107 108

109

110

131

132133

134 135

11

12

13 14

15

36

37

38 39

40

61

62 63 64

65

86 87

88

89

90

111

112 113

114 115 136

137138

139

140

16 17

18

19

20

41

42 43 44 45

66 67 68

69

70

91

92 93 94 95

116 117

118119

120

141

142143

144

US

21

22

23 24

25

46

47 48 49

50

71

72 73

74 75

96

97 98

99

100

121

122 123

124 125

146

147

148

1-4Y

150

REMARKS:

111111111111111

.

•

_________

~_.

_____

..J

_________________________________________

_

-46 -

Mathematical

Analyst

Keith Kersery loads

jet transport flutter problem into one

-qf

Lockheed's two

701

'so

On order: two

704's to help keep Lockheed in fore-

front

of

numerical analysis and pro-

duction control data processing.

704's

and

701's

speed

Lockheed

research

in

numerical

analysis

LOCKHEED

BURBANK

With

two

701

digital

computers

already in operation, Lockheed

~as

ordered two 704's

to

permit greater application of numerical

analysis to complex aeronautical problems now being

approached. Scheduled for delivery early next year, the 704's

will replace

the

701

'so

Much of the work scheduled

or

in progress is classified.

However, two significant features are significant to career-minded

Mathematical Analysts:

1)

the wide variety of assignments

created by Lockheed's diversified development program

and

2)

the advanced nature

of

the work, which falls largely into

unexplored areas of numerical analysis.

Career

pOSitions

for

Mathematical

Analysts

Lockheed's expanding development program in nuclear energy,

turbo-prop and jet transports, radar search planes, extremely hlgh-

speed aircraft and other classified projects has created a number

of

openings for Mathematical Analysts to work on the 704's.

Lockheed offers you attractive salaries, generous travel and moving

allowances which enable you and your family to move to Southern

California

at

virtually

no

~xpense;

and an extremely wide range

of

employe benefits which add approximately 14%

to

each engineer's

salary in the form

of

insurance, retirement pension, etc.

Those interested in advanced work in this field are invited

to

write E. W. Des Lauriers, Dept. MA-31- 2.

AIRCRAFT

CORPORATION

•

CALIFORN1A

DIVISION

CALIFORNIA

~

47 -

M "

II

h

d-

ore

soup

"VV. .

ere

you

:nee

l.t

...

to

drive

magnetic

cores,

drums,

computer

read-outs

HERE

IS

a high perveance

twin

triode

designed for

heavy

duty

computer

applications.

It

is capable

of

delivering

peak

cathode

currents

of

300 rna

and

will

dissipate

up

to

7

watts.

The

Sylvania

type

6350 features

separate

cathodes for

each

section

and

controlled grid cutoff.

Separate

cathodes

provide

maximum

flexibility

in

equipment

design.

DESIGN CENTER RATINGS FOR

THE

63S0

Peak

Positive

Plate

Voltage

(Abs.

Max.)

..•............

1000 Volts

Peak

Negative

Grid

Voltage

..

....................

400

Volts

Max.

Peak

Positive Grid

Voltage

.......•.

....

.....

. .

....

13 Volts

Max.

Peak

Positive Grid

Current

..........................

100 Ma

Max.

Peak

Cathode

Current

............

•

..••..

••

......

..

300

Ma

Max

.

Plate

Dissipation

Each

Plate

....

.........

. , . .

.......•.•....

'"

.3.5

Watts

Max.

Both

Plates

........

.

..........................

7.0

Watt.

Max.

OTHER SYLVANIA COMPUTER TYPES

Type 5844

.........•...

,

.•.....

"

.,

"

..

.

..•..

Medium

mu

triode

Type 5965

..............................

Medium mu

twin

triode

Type

6211.

.............................

Medium

mu

twin

triode

Type 5687

.......

.....

..

.

...

.•...

•.

. .

...

..

. . Low mu

twin

triad.

Cutoff

is held

to

close tolerances facilitating

the

de-

sign

of

circuits for

optimum

cutoff signals.

Minimum

interface formation assures operation even

after

periods

of

extended cutoff.

·Type 7AK7 .

...

.......

.....................

Sharp

cutoff

pentode

Type 5915A

.....•...........•.............

Dual control

heptad.

~

SYLVANIA:

Type 6145

.................................

Sharp

cutoff

pentad.

Type 6814

.......

..............

Medium

mu

triode

•

SYLVANIA

ELECTRIC

PRODUCTS

INC.

1740

Broadway,

New

York

19,

N.

Y.

In

Canada: Sylvania Electric (Canada)

Ltd.,

University Tower Bldg., Montreal

LIGHTING

•

RADIO

•

ELECTRONICS

-

TELEVISION

•

ATOMIC

ENERGY

SEE

SYLVANIA

IN

BOOTH

NOS.

168-172

AT

THE

1956

I.R.E.

SHOW

195602

195602 195602

Navigation menu

Versions of this User Manual:

Views

Navigation