TH8 SAGE AIR DEFENSE SYSTEM THE SAGE AIR DEFENSE SYSTEM A PERSONAL HISTORY THE SAGE AIR DEFENSE SYSTEM A PERSONAL HISTORY JOHN F. JACOBS The MITRE Corporation Bedford, Massachusetts c1986 by John F. Jacobs All rights reserved. Printed in United States of America. Reproduction of this book, in part or in whole, is strictly prohibited. For copy information, contact The MITRE Corporation, Corporate Archives, Burlington Road, Bedford, Massachusetts 01730. Library of Congress Cataloging-in-Publication Data Jacobs, John F. The SAGE Air Defense System. 1. SAGE (Air defense system) — History. 2. Jacobs, John F. I. Title. UG633.J24 1986 358.4'145'0285 86-16286 Book design by Caron Shing Typographer: Barbara Vachon To Mary CONTENTS Foreword ix Preface xiii A SAGE Chronology xv Introduction 1 1 How I Came to the Digital Computer Laboratory 5 2 The Development of Whirlwind 8 3 The Digital Computer Laboratory Joins Lincoln Laboratory 16 4 Contributions of Air Force Cambridge Research Center 18 5 The Cape Cod System 22 6 Whirlwind II 28 7 Assignment to Group 62 31 8 Defining the Whirlwind II Arithmetic Element 36 9 Jay Forrester and Company 39 10 Selection of a Computer Contractor 43 11 IBM Background 45 12 Lincoln Meets IBM 49 13 The Hartford Meetings 55 14 Project Grind 60 15 Genesis of the Systems Office 63 16 From Boston to Poughkeepsie 70 17 Features of the FSQ-7 74 18 Electronic Warfare 77 19 Defining the SAGE System 82 20 George Valley 86 21 Ma Bell 90 vii 22 Group 61 94 23 The RAND Corporation and SDC 96 24 Genesis of Computer Programming in SAGE 99 25 Meeting the Need for Programmers 105 26 Scheduling and Other Problems 108 27 A Promotion and the Steering Committee 114 28 SAGE Becomes Operational 117 29 SAGE Systems Testing 122 30 Division 4 126 31 The Task of Integration 130 32 Air Force Reaction and the Beginning of MITRE 135 33 Halligan 144 34 Leaving Lincoln and Joining MITRE 149 35 From SAGE to BUIC 155 36 The Winter Study 160 37 The Mystique of System Engineering 164 38 Summing Up 168 Epilogue 171 The Romance of Programming 175 viii FOREWORD S AGE was a remarkable development that had profound effects on the development of computers, information systems, and military capability. Many capable people were involved in its creation, in the Air Force, at MIT, and in industry, and a number of articles have been written about the project and a few of its leaders. This memoir is the first inside story of the project and how it looked and felt to those who were involved. Jack Jacobs is one of the great participants in the SAGE program; his contributions, largely unrecognized outside those who worked with him, were fundamental to success. I have known Jack for 35 years now. He is a modest man but a first-rate engineer and manager, and a major contributor to the art of large information systems. His memoir tells us much about SAGE, about how it was to work within a large pioneering development, and about Jack himself. I was at the MIT Lincoln Laboratory in the early 1950s as an associate division head, working on the design of the SAGE computer. I had transferred to Lincoln with the Digital Computer Laboratory. The division was growing rapidly; we had moved to another larger building, and I found that the closely knit old group, where everyone knew everyone else, was becoming a memory. One day, a tall, good-looking young fellow approached me with a question. He introduced himself as John F. Jacobs, our newly acquired graduate student assigned to the logical design of the adder in the comput- er's arithmetic element. He knew I had done the logical design of the Whirlwind computer and wanted to know why Whirlwind's adder design wasn't satisfactory. As I recall the occasion, I gave him a brief and somewhat lofty lecture on adders, told him the Whirlwind adder was really quite satisfactory, but suggested he make a new review of the possibilities. He looked at me with a mixture of curiosity and courtesy and disappeared. I thought I had handled it pretty well, and that he would review the alternatives and agree that mine was best. Imagine my surprise when, some time later, I found out he had come up with an improved design. This first contact with Jake, as we all called him, turned out to be typical. He was curious and courteous, did his homework, and came up with an ix improved design. Neither of us realized at the time that this first meeting was the beginning of a long and effective partnership. As Jake describes at greater length in this memoir, we came to know each other well while working with IBM in Poughkeepsie, New York. In this short time Jake had become one of the senior MIT design engineers. IBM had the job of engineering design and manufacture of the SAGE computer under Lincoln's overall technical direction. "Direction" meant traveling to Poughkeepsie every week and arguing out the design with the IBM engineers. Public transportation was inconvenient and we usually drove, a 4 1/2-hour trip. This meant leaving Boston at 5:00 in the morning and starting work at IBM at about 9:30. A day or two later we would leave Poughkeepsie at about 6:00 P.M., arriving home around 11:00. Over the course of a few months, we all had a chance to ride with all members of the group. After exhausting the technical discussions, we naturally turned to each other's personal lives and backgrounds. Jake was especially adept at telling stories of his early life in North Dakota. The snow, the grain elevators, and the poached deer all came alive to us in the dark car in the Connecticut hills. My clearest recollection, however, is of his tales of his education at "Meat Boners" college. I almost came to believe there was such a place and that the people of his hometown really did go down to the station to see him off, the local boy setting forth to make good in the big world. We took turns driving, and Jake's turn always raised my anxiety the most. We would be moving fast, and ahead would be a stoplight with a line of cars. It was absolutely inevitable that we had to stop, but Jake would keep speeding on, talking all the while, and at the last possible moment, he would cram on the brakes and come to a shuddering stop. He maintained that he had learned to drive this way in North Dakota where there were few cars and no stoplights, and where coming to a stop was considered an emergency procedure. His part of North Dakota was laid out in sections and he claimed that he had trouble with turns because he hadn't learned to make them until he left home. That may be true; but as far as I know, he never had an accident. Not long after this, we began to cope with the problems of being a relatively small group at Lincoln without legal authority, trying to maintain technical control of the very large SAGE program involving many government agencies and many contractors. The usual approach to such a problem is authoritarian: decide what to do, tell everybody to do it, and make sure that they do. It wasn't at all clear that we wanted to run things that way, and it wasn't at all clear that Lincoln could make its orders stick if we tried. There were too many others and too few of us. Jake came up with the idea of the Systems Office, whose job it was to find and foresee problems and to get the large masses of people involved to understand and agree on appropriate action. Jake's view was to use leadership. He would analyze the problems, propose solutions, and get everybody together to agree. "Thousand-man meetings," they were called. They worked because everyone wanted the problems solved and no one (besides us) wanted to take responsibility for the whole system. Success was dependent on sound and thorough homework, however, and this is where Jake excelled. His willingness and his ability to do whatever was necessary to make a program come out right were his most distinguishing characteris- tics and among the chief reasons he played such a major role in the design of SAGE and in the creation of The MITRE Corporation. Once the SAGE hardware was well under way, Jake assumed direction of the software part of the job. The specification and preparation of the necessary computer programs turned out to be an enormous under- taking. The Lincoln group of less than one hundred, which had been preparing operational, test, and support programs for the prototype, was augmented by a new organization, System Development Corporation, with thousands of new and untrained people. Overnight, the Lincoln people became supervisors and Jake's skills were needed to make the new situation work. He dove in with his usual combination of friendliness, firmness and hard work. I used to go to his organizational meetings from time to time and noticed that the charts on his, blackboard were filled with misspelled names. He maintained that he had never learned to spell, but I felt no one could be that bad and accused him of deliberately misspelling in order to cover up his real mistakes. I now think I was right but for the wrong reason. He had many new people to deal with and was bound to misspell some names. But, he deliberately misspelled many, not to cover up for his mistakes but so that no one would feel too unimportant to be spelled correctly. He was concerned about them, not about himself. Throughout our careers at both Lincoln and MITRE, we were called upon to work with many Air Force officers at all levels. In such situations there is always tension between the need for indepen- dence, which is necessary to maintain quality and effectiveness in the long run, and the need for responsiveness, which is necessary to maintain xi