I must say I never did enjoy life in New York City. In the first place there were the trains. While we still lived in Wayne, I had to commute to New York. The Pennsylvania’s main line ran through Paoli, just a few miles west of Wayne, where all the through trains from the Mid-West stopped. I bought a commutation ticket that enabled me to get aboard one of those trains in the early morning (some as early as 5am), and get to New York two hours or so later. But they were seldom on time. Their schedules were not set up for the commuter. When we moved to Norwalk, Conn., in November 1954, there was the commuter train to be taken from there. Once we sat for over an hour without moving in the yards above Grand Central Station. The railroads just weren’t interested in moving people — no money in it. Then there were the crowds. We sales people on expense accounts never rode the subway! We rode in cabs. A cab crosstown could easily take more time than a pedestrian, as the north-south arteries blocked the east-west streets on nearly every green light. Then there was the noise, and the filth, and the rush. I’m sure Hell is going to be an enlarged New York City. Getting to the airport (La Guardia) was also difficult. The city built two terminals, locating each one far enough away from the subway lines that one was forced to take a cab after getting off the airport bus. The bus had to come through one of the tunnels into the city, and these tunnels were often jammed. Ever sit beneath the East River or Hudson River for 15 minutes or an hour in the fumes and heat? Once driving back into the city from a New Jersey camp where Will had spent several weeks, we crawled the two miles up the escarpment to go down into the Lincoln Tunnel from 8am to after 11!
Our sales team of Parker, Harr
and Mitchell traveled so much that it was hardly worth leaving Wayne.
We were gone from the New York sales office from Monday until Friday,
week after week, living in hotels, eating in restaurants. It was
flattering to be admitted into the inner sanctums of the vice
presidents of major corporations, and even occasionally meet the
president. At one time I was on a first name basis with three US
Steel vice presidents. But looking back on it, I cannot remember a
single incident when something really worthwhile was accomplished.
True we got business. Until 1955, IBM refused to enter the big
computer market for business applications. Instead, they defensively
protected their huge investment in punched-card machinery (leased
only — at high prices), which included over 90% percent of all
commercial businesses and over 97% of the US Government’s data
processing. But in 1955 IBM unveiled their 702, which was soon
upgraded to the 705, and then we were in a battle for business. There
is no question but that Thomas J. Watson was the greatest salesman
America has produced. He had built IBM into a veritable giant of
business machines — typewriters, accounting machines, and the
80-column punch card machines. His salesmen were trained in business
procedures so well that many business executives allowed them to set
up their whole way of handling their business data processing. The
customer simply operated the machines and paid their rentals to IBM,
and often had nobody with the capability of systems analysis. This
was particularly true in the vast Government bureaucracy, both
military and civilian. The people who were at middle management level
could not function without the IBM account representative to tell
them how to accommodate the new requirements that were constantly
arising. There were exceptions to this situation, to be sure, else we
in Remington-Rand could not have made our living. If a manager chose
IBM and something went wrong, his boss would blame IBM. If the
manager chose Remington-Rand and something went wrong, the manager
would almost certainly get demoted, if not fired. To get a hearing
for our computers we had to go to the top of the business or
Government agency, where the sheer capability of these new machines
made them attractive — even essential — to the
organization. Parker well understood this, probably from his training
in the brokerage industry. He had a reputation for straight talking
and for delivering on his promises, and these were characteristics
that richly paid off for Remington-Rand, as well as to invoke our
confidence in our boss. I cannot recall Parker ever telling a
customer or prospect a deliberate lie. There are not many people of
whom I could make that statement!
Remington-Rand had an office building on the corner of 23rd Street and Madison Avenue. It was not the most fashionable part of town, but it was also not the worst. There was only one good hotel nearby, which catered to residents as well as transients, and Luther Harr and his wife Eileen, who had no children, lived there. Mr. Parker had an uptown apartment, but made his permanent home in Washington, which he maintained even to the end of his life. John loved to entertain, and whenever we had VIP’s come to our office from out of town, we would do the night clubs nearly every night. Everyone — I mean every man I ever met in business — drank cocktails before and often during meals. After all, alcohol removes the barriers to communication — until too much of it removes the communication itself. Parker and Luther had iron constitutions, and never showed the amount of liquor they might have consumed. I never bought a drink for myself, but I bought many of them for customer personnel, and always had to join them myself, to be sociable. I remember one night when we had a group of US Steel executives from Pittsburgh (our first major commercial installation, US Steel’s National Tube Division). We had finished a dinner in one of Third Avenue’s most prestigious restaurants and Parker ordered a round of Drambuies — then a second — then a third — up to seven before the party broke up. In the cab back to Luther’s hotel, John was talking about going to a nightclub for more drinks, but I wanted to call it quits, so feigned to have passed out. Luther knew it was feigned, but he had had enough also, so he convinced John that I had to be put to bed. I had one of my worst hangovers the next morning, but John was not fazed a bit. On another occasion I had to entertain some of these same Steel executives, who knew New York’s night life much better than I did. I don’t remember how many night clubs we went to at which I dutifully paid the tab, but I do recall getting to the hotel where Luther lived about 4:30 in the morning, only to be told that the room I had reserved and guaranteed had been given to someone else. The hotel clerk did give me a second-floor room, which he warned had been reserved for a meeting at 8am. When I was roused and evicted at 8, I had the biggest hangover ever, and all I could think of was to lie down again — but no room. So I went to the Harr’s apartment and pled with Eileen to let me stay there that morning, until I could recover, which she very hesitatingly did. That was my last time for partying in New York, or anywhere else.
A number of the programmers on my Philadelphia staff resisted the move to New York, so a second programming staff was created in Philadelphia, under Grace Hopper, to produce automatic programming systems. I had been working on an assembly system which would use the regular Univac instructions, but with symbolic addresses, allowing much greater freedom in writing and changing programs. I added standard packages of instructions to control input and output operations and finally a feature to permit easy reruns of the program from an intermediate point, avoiding the restart from beginning of long computer runs. I coined the acronym BIOR (Business Input-Output-Rerun) for the program. Except for the rerun feature, the program was operational in 1955, and used by the Westinghouse Electric staff for all of their programs. However, a somewhat more sophisticated system was designed and produced by Grace Hopper’s staff, and this became the official Univac program. Mr. Seares called it “Flowmatic”, of which the critics said it reminded them of a dishwasher. He had earlier added the word “Factronic” to Univac, a name that was immediately dropped at his retirement in 1956.
The election prediction publicity was probably responsible for the great interest shown in Univac in 1953, both by Government and private organizations. The Air Materiel Command, supply organization of the Air Force worldwide, bought four machines. As already mentioned, US Steel bought one for their Pittsburgh operation (payroll), and later one for their plant in Gary (IN), also for payroll. General Electric bought one for their Appliance Park in Louisville (KY), again for payroll. DuPont and Westinghouse Electric bought one each. Commonwealth Edison and Metropolitan Insurance Company bought two each for their New York City operations. In addition to John Hancock Insurance Co. (Boston), three smaller insurance companies bought Univacs (Pacific Mutual in Los Angeles, Franklin Life in Springfield IL, and Life and Casualty of Tennessee in Nashville). Mr. Rand wouldn’t authorize the cost, and Pres Eckert didn’t want to use the talent of the engineering staff to convert the laboratory version of the first Univac into a production model. So the first machine was largely duplicated through 28 clones, although some minor improvements were made, until the redesign of Univac II introduced some production features. The cost of construction could probably have been reduced by $100,000 or more. Our chief operator was hired by Franklin Life as their maintenance engineer, and he was able with a few ingenious additions to double the space efficiency of the instruction code. But Pres Eckert had much more ambitious ideas about computer design, and spent about two years on an elaborate design for a supercomputer for the AEC called LARC (Livermore Atomic Research Computer). Only two were sold — Livermore and the Navy’s David Taylor Model Basin.
When General MacArthur lost his bid for Presidential nomination in the summer of 1952, Mr. Rand offered him the position of chairman of the board of Remington-Rand, which he accepted. He was given an apartment high up in the Waldorf-Astoria Hotel in New York City, where the sales staff would bring senior officers of major corporations to meet and hear this very famous person, and incidentally to raise their opinion of Remington-Rand, which IBM disparaged at every opportunity. Since some of these meetings involved Univac prospects, I had the privilege of being present, and thoroughly enjoyed the long but never dull comments the General had on whatever topic was put to him. When the time came for dedication of the first Univac for the Air Materiel Command (September 1954), the commanding general sent his C54 command plane to La Guardia to fetch General MacArthur, John Parker, Luther Harr, and myself. General MacArthur was to give the keynote speech at the luncheon, and I was the company spokesman for the dedication ceremony. On the way to Dayton (Ohio) for the ceremony, I rode on a seat with Arthur Treacher, the British movie star famous for his portrayal of butlers. He was a very interesting person, answering my question of “Don’t you get tired of being in the limelight?” with “Oh, no, not at all!” On the return flight that afternoon, I had two hours sitting with General MacArthur! I asked him questions about his career and opinions about current events. He would usually give me a 20-minute answer, leaving me ample time to think up another question. I have never before or since had the privilege of being in the company of such a brilliant mind. Now over 70, the General never paused for a name or fact, never backtracked or got lost in a story, never made a grammatical error, and used the best choice of words I have ever been exposed to. However, his political sense seemed to be out of keeping with his other talents. When I asked him what he thought of then President Eisenhower, he gave me a long preliminary about the years in the Philippines when Major Eisenhower was his aide and he commanded the Philippine Army. He said he considered Ike to be another son, but still felt he would make a bust of the Presidency. As I remember his words, he said, "For the next year his popularity will diminish arithmetically; then the next year it will diminish geometrically; and in the final year it will diminish exponentially!” (Of course, Ike was easily reelected after his first term.) I then asked him who he thought would be our next President. His answer was “Lausche of Ohio” who wasn’t even nominated. Even the most brilliant minds cannot predict the future!
As mentioned above, our trained programmers were in great demand by our customers. At first we provided at least one senior programmer, and occasionally several assistants to get the customer’s application programmed. Morgan Huff was one of the most valuable of these men. He was our representative at the General Electric Appliance Park, where he did outstanding work in organizing that 11,000-man payroll, with 11 unions, and many special management provisions, plus every kind of compensation then commonly used by manufacturers. During production tests prior to going on line, the program was disappointingly slow, due to its extreme complexity, and the GE management complained to our New York office. Mr. Parker sent me down to Louisville to see what the problem was. Morgan briefed me about the cause of the slow-down, the fact that the middle management people had insisted on including some complicated cases for which there were a very few employees each, mostly management. The GE programming staff had urged their elimination, but the management people wouldn’t budge. I asked a few questions and did a lot of listening, and then went to the senior management that had complained, urging them to eliminate these few people from the computer payroll computation. This was done, and the time dramatically improved, making me a hero, when all I did was listen! Morgan was later sent to Los Angeles to be the computer consultant for the Pacific Mutual Life Insurance Company. He eventually wound up as vice president of the computer department of the Life and Casualty Company of Tennessee in Nashville, from which position he retired a few years ago.
Meanwhile the engineering staff in Minneapolis, that had been brought into the Remington-Rand family from Engineering Research Associates in 1952, had been improving their scientific computer, the 1103. Magnetic cores were substituted for the tricky electrostatic memory system, and the input-output capabilities were improved. A line of drum computers was also developed which became popular in industrial process control and such applications as bridge and tunnel toll monitoring. This development culminated in the Univac 1150 series utilizing huge drum memories that competed for a while with the disk memories that are the industry standard today. With the fizzling out of the LARC, Remington-Rand’s management gave more funds and directives to the Minneapolis group, until that group eventually became the principal research and development arm for the whole computer effort of the company. Even that long ago, a small research group was working on what were then called thin-film memories, the forerunner of the electronic chips of today.
IBM reversed its stand on computers in 1954, and brought out in the following year their first competitor to the Univac, called the IBM 702. This computer had magnetic core memory — a great improvement over the tricky mercury tanks of Univac I — but lacked the capability of simultaneous input, output, and compute. Although its internal speed was somewhat greater than Univac, its throughput was about half on most commercial applications (it had to read, write and compute separately, rather than simultaneously as did Univac). It is a testimony to the sales ability of IBM and their stranglehold on American business that they were able to beat us out on two out of three potential customers (all of whom, of course, currently used IBM punched card equipment). Even before the 702 was released, IBM had marketed a smaller drum computer, called the 650, which had a modest acceptance, especially on the West Coast. Realizing the weakness of the 702 compared with Univac, IBM announced in 1955 a new model for business applications, the 705, which clearly eclipsed the Univac in performance. We had been urging improvement for over a year but Eckert steadfastly refused to divert his engineers from the LARC. Finally, an engineer at Minneapolis offered to undertake an upgrade of Univac I, and I served as the interface from the sales organization to determine its specifications. It was offered to our Univac customers for a $250,000 upgrade price, but actually cost the company much more than that, as there was no “trade-in” value to the older machine.
more than a year of commuting from Wayne to New York, I finally
decided that we should move to the New York area. We spent several
week ends looking at possibilities, both in New Jersey and in
Connecticut, and finally decided on a new house in Norwalk which was
nearly the last in a large subdivision. We moved in November 1954. It
was on somewhat higher ground than the main part of the city, and was
close to the Parkway and US #1. It was a 4-bedroom split-level house,
and well suited to our family (see right, when Mother visited us).
Norwalk was not as friendly as Wayne, however, and we got to know
very few people.
Most of our friends were from the church in Darien we attended. Will got a bicycle for Christmas, and really enjoyed riding it to school. On his first day home, when asked how he had made out on his bike, he said: “Oh, it was easy — downhill all the way there and level all the way back.” That became a family joke for years thereafter. Will maintains that this incident occurred in Wayne before we moved to Norwalk, but both Mary Charlotte and I remember it as stated.
The service-bureau Univac in our New York office has an interesting history. One of its purposes was to serve as back-up for the Univacs in industry — particularly those doing payroll, as failure to pay on time was a very serious matter. On two occasions the US Steel (Pittsburgh) Univac went down at a critical time in payroll processing, and the magnetic tapes were flown to New York to the service-bureau Univac. I remember Luther Harr going to La Guardia Airport at 4:30 one morning to meet the US Steel people, including one of their vice presidents. No pay schedule ever failed to be met, but some were close calls. In my opinion, our most important job was the production of the concordance of the Revised Standard Bible, done our first year (1954). Thomas Nelson Publishers paid us for the work at $550 an hour. The entire Bible was first recorded, verse by verse, with a special code prefixed to each verse. Then the resulting text was sorted by individual word, each carrying with it the verse and code that identified the word’s location. Insignificant words (and, the, for, at, etc.) were eliminated, and then the reference text was created according to rules that we could program the computer to carry out, giving a one-line (or shorter) context for each word. The entire task, including programming, took only a few weeks, whereas the concordance for the King James Bible took nearly a lifetime to produce. One other job might be of interest (we didn’t do it). A firm that produced a sort of almanac wanted us to compute the favorable and unfavorable days in the year according to the right ascension (horizontal angle from north) of the moon (at midnight?). Superstition has long held that the moon’s position could have a bearing on the outcome of important events, and these people thought there would be a market among farmers, fishermen, investors, sports, gamblers, etc. for such a publication. Incidentally, the service bureau staff, while under my direction, was headed by Arthur Katz, the student who had helped me in my thesis computation on the Mark II computer 11 years before.
was company policy for senior sales people to be active in
professional societies. I belonged to the three that were active in
computer technology: the AIEE (American Institute of Electrical
Engineers — I had joined as an EE student at George Washington
Univ.), the IRE (Institute of Radio Engineers), and the ACM
(Association for Computing Machinery). Our most ambitious undertaking
was an annual joint computer conference, at first only for the
presentation of technical papers, but very soon to include
manufacturers’ displays as well. It rapidly grew into a major
event, with thousands of persons attending, and requiring the largest
show grounds available. The task of coordinating it was given to the
Joint Computer Committee, soon to be split into eastern and western
committees. I served on one or the other of these committees more or
less continuously from the early 50’s until I left Univac in
1960. 1 usually presented a paper at each conference. The first paper
was presented at the first (or second) such national conference in
New Brunswick (NJ), on a method of inverting very large matrices,
which had a self-correcting feature that reduced the loss of
precision due to round-off errors. The paper created quite a debate
between Herb Grosch (the IBM representative) and Richard Clippinger
(the Honeywell representative) as to the legality of my method. I
couldn’t get a word in edgewise, and it was just as well, as
Dr. Clippinger ably defended me, and silenced Dr. Grosch. This method
was later successfully used to invert an industry input-output matrix
of 300 industry sectors (nearly 1,000 times the amount of computation
of my thesis problem) for Project SCOOP of the US Air Force.
A second dimension of my professional society involvement was to be a traveling lecturer to State professional chapters of the AIEE and IRE for the years 1957 and 1958. John Mauchly served as such for the States east of the Mississippi, and I served the State groups west thereof. These public lectures gave me good training for the work I was to do in the early and mid 70’s, and again emphasized the value of my course in Public Speaking at GW. Still a third was the occasional request to teach a management-oriented computer course. The first of these was at the New York University in 1953, followed by several short courses at the American Management Association in New York. After moving to California, I taught an evening class in the principles of electronics at UCLA.
Shortly after our return from Europe, Mary Charlotte and I were invited to attend a weddng anniversary celebration of the Graham Smiths, our co-workers in New York City. Leaving the party about midnight, we headed for Grand Central Station to get the last train that night for Norwalk. Entering the station, we found it jammed packed with people, milling around and all talking at once, and wondered what on earth was going on. Eventually, we learned that there had been a train wreck on the New Haven line, and no trains had departed north since early evening. Hurricane Connie was dumping its rain on New England, and caused flooding all up the coast. Finally, a train was announced for Stamford, about ten miles south of us, and we managed to get aboard. The engineer stopped at every creek crossing, got out of his cab with his flashlight, and walked down the embankment to see if the culvert was still holding. We arrived at Stamford at 5am, after taking 3-1/2 hours to come 30 miles! The station there was filled with people, and the few cab drivers said that all roads north were under water. How could we get home to the kids and Mrs. Chapman? Since no trains were running, we decided we would walk along the railroad tracks, and set off. At one point where the tracks crossed a small river, we found the entire roadbed had been swept away by the flood, and the four tracks were spanning the breach hanging in mid-air! It was not hard to walk on the ties, but it was an eerie feeling to cross the 100 feet or so of unsupported tracks. Seven miles and 2-1/2 hours later we reached Darien, and left the tracks to see if we could find transportation. Nothing was stirring, however, until a small car drew up beside us, with a couple and a small child. They were trying to reach a friend’s house in Norwalk where electric power was still on, as they needed to heat the baby’s bottle. We told them that we could show them back roads into Norwalk (the main roads were all under water), and they took us into their car. After much twisting and turning, we reached the railway station where our car was, and thanked the people for their help. We got home at 9:30am, nine hours after arriving at Grand Central Station! Later that day we drove to New Haven to buy candles and canned heat, as power was off at our home. A normally small creek, called Five Mile Creek, had become such a torrent that it cut the Parkway, US #1, the railroad, and a block-wide swath clear through the city of Norwalk! For more than a month the only route north was the sea-level bridge at the coast in down-town Norwalk, where the flood waters had spread out sufficiently to lose their force. A substation near us was buried under hundreds of huge boulders, some as big as the transformers they covered. That was an experience we will never forget.
In the spring of 1955, rumors of a merger with the Sperry Corporation began to circulate, soon to be confirmed. It seemed that Mr. Rand had gotten tired of managing the company, and not finding anyone among his subordinates that he felt had the management ability to succeed him, chose Mr. Vickers, president of the Sperry Corporation, to take over that responsibility. The merger was accomplished as of July 1, 1955. The new company was first called the Sperry Rand Corp., and later became Sperry Univac. Changes in my management were swift to follow. One of the senior engineers in the original Engineering Research Associates was a William Norris, who had good financial connections in the Twin Cities. It seems that there was general dissatisfaction among the engineers of that company with their merger with Remington-Rand in 1951, which had been engineered largely by John Parker, the former president of ERA. This dissatisfaction led to a sub rosa plan to launch a new computer company, with financial support largely from Minneapolis. Bill Norris apparently was slated to be president of this new company. No word of this development reached our ears until the spring of 1956, when the persons involved in the plan, though still employees of Sperry Rand, spent most of their time preparing to launch the new company, which they did on July 1, 1956, as the Control Data Corp. Why they waited a full year after the merger with Sperry, I don’t know. Bill Norris quickly gained the confidence of Mr. Vickers, and had himself appointed manager of a new Univac Division, which would include all engineering and manufacturing facilities in Philadelphia and Minneapolis, and the New York sales organization. I have a vivid recollection of the scene in John Parker’s office when Bill Norris came in and announced himself as the new division manager. He told John Parker that he didn’t want him as his sales manager, although he asked him to stay temporarily until his replacement could be found. I learned later that John and Bill had crossed horns in the old ERA company, where Bill was the junior man. Now the situation was reversed and Bill was getting his revenge. I had occasion to go with Bill Norris to the management of a company in Chicago that had indicated an interest in Univac, but had some reservations. I must say that Bill Norris did nothing to ease those reservations — if anything he did the reverse!
The new regime was anything but heartening. Coming back from a
board meeting, Bill Norris told us that the new management didn’t
approve of our way of seeking business, and intended to develop their
own. What a blow! Nearly six years of giving my all to the cause of
Univac, and now to be told that wasn’t the way to do it, by
people who, in our opinion, knew nothing about it. I wanted to get as
far from New York as I could. The opportunity came quicker than I
dared to believe. In January 1956, I made a routine trip to Los
Angeles, and shared my woes with the branch manager there. He said
that he would be very pleased to have me on his staff, but he had no
vacancy. The very next morning, as I was preparing to return to New
York, he called me into his office to tell me that Irwin Tomash had
submitted his resignation that very morning, and if I wanted his job
as West Coast Univac Representative, it was mine! What a delightful
prospect. Mary Charlotte hated New England, and I hated New York
City. Both of us had enjoyed our war years in California, and now we
could come back. Hallelujah! Praise the Lord! I wasted no time in
telling John Parker, who was being kicked out shortly himself, and he
made immediate arrangements for my transfer. On my last day in the
New York Office, John had all the secretaries line up to kiss me
goodby. One of these, Ruth Britton, refused to do so, as she was
going to transfer to the Los Angeles office also. On my first day of
work in the Los Angeles office, when my new boss was present, here
comes Ruth and plants a kiss on me before I knew what was happening!
That was the most embarrassing incident I can recall in my ten years
with Univac! After all the details of the move had been completed in
late February, Mrs. Chapman, the two kids, Mary Charlotte, the dog
and I got into our Pontiac station wagon and headed for sunny
Go to next chapter
Return to Table of Contents