Lockheed Aircraft -- Missile Systems Division

Computer History Vignette

By Bob Bemer

Getting Staffed

I had been at Marquardt Aircraft for 16 months when Lockheed decided in March of 1954 to ask me back to start the computation section for Lockheed MSD (Missiles and Space Division). Computer employee No. 1! It may have been some television shows that I was on, like Art Baker's "You Asked for It", that made me visible; I can't be sure. In today's garb as Lockheed Martin, one can see that my original computing operation has grown much larger since then!

I was to have both analog and digital computer groups. One Robert Prince was already there to help get the analog part started. So was a keypuncher assigned to me from Payroll. But otherwise I had to start from scratch. I couldn't pick up any people internally, for they were all new themselves.

I began by convincing the head personnel man, Bob Birdsall, that programmers were so scarce that we would have to pay on average 25% more than for a comparable aeronautical engineer.

But my authorization to hire did not get into effect quickly enough. I interviewed a Robert Perry, but could not hire him at that time. He had two pluses, to my mind -- he was brother to Malcolm Perry of SABRE reservation system fame, and his father was a school superintendent, as mine was. He didn't have far to look for a job, and took another one before I could make an offer. That led me to the somewhat childish vow that I would hire the next superintendent's son that came around. To my amazement, one did turn up, and I hired him as vowed. The NSA people he later worked with can tell you that he turned out just fine.

I reported to Art Hubbard, who reported to Willis Hawkins, Chief Engineer. The latter was a fine and honest man, but Hubbard was a jewel. Moreover, he had confidence in me, and let me run my operation as I pleased.

When I started at MSD, we had no equipment, and just that keypunch operator assigned. IBM 650's were so popular that they were tough to come by, so while waiting we started with a CPC in a building across from the main plant at the Van Nuys airport.

We had first a CPC, and eventually three 650's. And were looking forward to a 704. In fact, I was the Lockheed MSD representative to the first meeting of SHARE, in 1955 August.


NEW HIRES

I begin with the staffing, to put the new employees in a deservedly good light.

I may have hired more good people, in the early days of programming, than any other manager. It was all done by instinct! And at Lockheed Missiles, I must admit I hired some that other managers might have bypassed. The cast included:

  • Dr. Jack Sherman, already famed in the analog field, to run that section.

  • Jim Hudson, who had been selling a product called "Baby Tenda". What would make such a person a fine programmer? I don't know, but he was.

  • Albert Podvin, who had some drawbacks. He did not yet have a college degree. His wife was divorcing him, and he looked like Humphrey Bogart, except perhaps less handsome. But he told me he just HAD to have a job. I said I could only take him on in the nominal capacity of keypunch operator (at that time he could not run one), and we'd see what to do if he showed promise.

    Yes, the same Albert Podvin that was B. O. Evans technical assistant and hairshirt when Evans was in charge of developing the IBM 360! I knew I'd made a great choice when Al wrote a sine-cosine routine for the 650 that was faster than mine -- he did it via sine 3 theta!

  • Elaine Gatten was the keypuncher already there, but in IBM Technical Newsletter 10 you'll find her listed as author of one of the program components of the FLAIR System for the IBM 650. Not bad for a keypuncher.

  • Don Jackson, more or less borrowed from Lockheed's California Division, where they made airplanes. He helped me test our 650 programs at IBM's Endicott plant.

  • Dick Talmadge. He was with our group at Marquardt Aircraft, where he did mostly trajectory calculations for the BOMARC missile. He was the only one of us who could write programs with a pen. We others all made mistakes, found during test runs, that required erasure.

    Upon leaving Marquardt Aircraft I had recommended Dr. Talmadge to be my successor, and it was agreed. Later, though, Dick expressed a wish to come to Lockheed Missiles, and I could hardly refuse someone with his talent.

    Still later, he joined IBM on the West Coast, becoming famed by creating such a great compiler for COMTRAN (Commercial Translator) on the IBM 709 that IBM's customers balked at using IBM's COBOL compiler for that machine, even after IBM had sworn allegiance to COBOL.

  • Dick Middleton, with whom I had worked on the graveyard shift at the Rand Corporation when I was just getting started as a programmer. He figured in an interesting situation:

    My guess that the PhDs Ernst Krause brought to Lockheed Missiles (summer of 1955) were snobs was reinforced when all the new people with doctorates were given wooden title bars for their desks. I complained to Bob Birdsall of personnel. He apologized, but that was how it was. "Not for me!", I exclaimed modestly. "For Dr. Middleton of my department, who feels neglected".

    Name bar duly furnished. I stuck a wire in it, topped with a ping pong ball painted with nail polish as a bloodshot eyeball. Our Dr. Middleton's degree was in optometry! He had gotten it after leaving the RAND Corporation. But he wasn't so enamored of that profession that I couldn't talk him into being a programmer again.

  • Fletcher Jones, formerly working at North American Aviation. I do not mark him down as one of my hiring successes, however. Shortly after coming aboard he disappeared, and we found that he had been dickering with North American to return (they had an IBM 701, and he claimed that he preferred the big machines). I think it was because he couldn't understand the IBM 650. At least I retain the only piece of coding he ever did for us, with my note "has no knowledge of stored programming" on it.

  • Ralph K. Rea (the school superintendent's son -- later to NSA).

  • Black Lib -- Ben Handy, early worker on the U.S. National Bureau of Standards (NBS) groundbreaking SEAC computer, I had, on my second try, persuaded him to leave CRC (his current employer). He accompanied me to the 1955 September ACM meeting in Philadelphia.

    Taking heed of the claim of "City of Brotherly Love", I arranged for us to occupy the same room, at the Bellevue-Stratford Hotel (since out of business). The eyes of those white bellboys were large as they carried our luggage to our room. Bear in mind that Dr. Martin Luther King was not well-known in 1955, and Ben was unmistakably black. But he was a great programmer and dear friend.

  • Woman's Lib -- Irene Brown, from some college in Boston, hired by mail without an interview. The Philadelphia meeting wasn't my only innovation in employee relations. After a few weeks at work she came to me and asked why I was paying her $115 a week, whereas she had been getting only $60 at the college. I told her that $115 was what the men were getting for the same work she was doing. She said nothing more, but never stopped working hard!

    I also hired her husband, Frank Brown, and no nepotism was involved. He, too, was a competent programmer.

Notable Project No. 1 -- Lagrangean Multipliers

Dr. Walter Bauer then headed the computation operation for Ramo-Wooldridge Corp. When Lockheed Missile Systems was awarded a subcontract to validate the Ramo-Wooldridge work on the Atlas missile trajectories, from 1 to 2.5 stages, Walter called me to his office and gave me the word to use Lagrangean Multipliers for optimization. I must have nodded yes, but to this day I have never discovered what Lagrangean Multipliers are.

Or cared, because I had a secret weapon -- Dick Talmadge's brain!

We had to use the IBM 701 at the California Division of Lockheed, not having one yet ourselves. So I said (privately) forget Bauer and his Lagrangean Multipliers. I had a larger board laid over the short-depth shelf of the printer (where the answers showed up), for graph paper. Talmadge would sit there, make up a set of parameter cards, run a few points, extrapolate the curve, try a new set of parameters, see if it got worse or better. Thus he zeroed in. We didn't validate Ramo-Wooldridge. We beat them to the answers by a margin of months!

A special 701 board needed to be wired for this project. One Chuck Baker was brought in for the task. Somehow there was a lot of talk, and I found myself doing the same wiring in competition with him. It was fortunate for my reputation and authority that I, although now technically just a manager, was the winner. And it worked the first time!

Notable Project No. 2 -- Computerized Dynamic 3-D

Our analog section was primarily devoted to data reduction from missile test flights, because all of the instrumentation was analog in those days. Had you said "chip", a blank stare would have resulted.

This project is documented fully on this site under the name "War and Perspective", which see.

We made a lot of generals ecstatic, and set a minor amount of precedent along the way to PC games.

Notable Project No. 3 -- A 650 Programming System

Lockheed also had a company unit in Marietta, Georgia, with a charter somewhat similar to MSD's. The man starting up the computer section there was my old friend and roommate Bob Bosak. They, too, were initially getting 650s. As there were no signs that IBM was producing anything decent in the way of a machine-independent programming system for the 650, Bob and I decided to build one jointly between our two groups. The products were not identical, but most components were common.

Ours was called "FLAIR", for "FL_oating A_bstract I_nterpretive R_outine" [1]. Theirs was called "FACS", for "F_loating decimal A_bstract C_oding S_ystem" [2]. Essentially, Bosak's people did the arithmetic portion, and we did the subroutines.

You will not find Reference 1 in any list of my papers. The formal author is just the entire section, with the several authors of the routines identified. The paper is of substantial interest, however, in that the entire text and some diagrams was produced entirely on punch cards. Even the program listings. It was certainly my first word processing venture, and one of the world's earliest. I think.

Notable Project No. 4 -- Polynomial Relaxation Methods

I had seen Cecil Hastings develop his function approximations while we were both working at the RAND Corporation, In practice they were difficult to program -- too many terms with too many digits in the coefficients for the computers of those days. I thought to postulate a simpler formula of shorter coefficients and fewer terms, and use the same Tchebysheff methods to approximate the difference between that polynomial and Cecil's original, adding the short result back. I called this "relaxation".

The 1955 September ACM meeting in Philadelphia was coming up, so I produced a paper on this method, which we had then been using with much success.

After the session where I presented my paper I wandered out to where George Stibitz and Alston Householder were sitting. As I joined their table, Stibitz asked "Anything earthshaking in the last session?" Householder replied "Well, Bemer here caused a small tremor". When justly famous men give you a small compliment, count it large.

I found later that C. Lanczos had done similar work concurrently, but called it "polynomial telescoping". The priority of invention has not yet been adjudicated.

Leaving Lockheed Missiles and Space Division

Some time in 1955 MSD got in a Dr. Ernst Krause (who later started Ford Aerospace) to create a new research division. He was all hot for PhDs and a UNIVAC 1103 (whereas we were expacting an IBM 704 to match our parent company). He also brought in his own computer man, a Dr. Werner Leutert (whom I replaced at UNIVAC some seven-plus years later, with my memory still fresh). And new research-type (?) people.

I saw no actual blood flowing, but it was that type of a situation. Even Willis Hawkins, the current Chief Engineer, was attacked bitterly and painfully.

And, perhaps to be different, or to show that THEY were now in charge, they really pressed for that UNIVAC 1103, and their influence won. I was left on the old limb. Don Pendery, that great IBM Applied Science Representative in the Los Angeles area (and later V.P. of Xerox) came to my rescue, arranging an interview with Dr. Charlie DeCarlo in New York in November of 1955. I started there the next month.

From then on I did very little actual engineering work. But I certainly kept up with computer usage for engineering there. I followed the APT work for John McPherson, consulted on many engineering applications, and most of the users of PRINT I, my first product there, were engineers -- AOSmith, oil companies, etc. But I was now totally a systems programmer from then on.

The mean little kid in me takes satisfaction today that Google gives only 2 hits for that Dr. Ernst Krause, vs. thousands for me.

REFERENCES

  1. The Mathematical Analysis Section, Missile Systems Division, Lockheed Aircraft Corp., "A general utility system for the IBM Type 650", Technical Newsletter No. 10, IBM Applied Science Division, 31-48, 1955 Oct.
  2. Robert Bosak, "Development of a floating decimal abstract coding system (FACS)", Technical Newsletter No. 10, IBM Applied Science Division, 28-30, 1955 Oct.

Back to History Index            Back to Home Page