This Article 
 Bibliographic References 
 Add to: 
Coming to Grips with Univac
April-June 2006 (vol. 28 no. 2)
pp. 32-42
When it made successful use of magnetic tape, the Univac I pioneered the way for federal and commercial applications with extensive files. But pioneering posed many challenges because technology barely supported the production and introduction of electronic computers. Here is a recollection of efforts the US Air Force made to accommodate and use the first Univac moved out of the factory.

1. G.B. Dantzig, "Maximization of a Linear Function of Variables Subject to Linear Inequalities," chapter 21 in T.C. Koopmans, ed., Activity Analysis of Production and Allocation, John Wiley & Sons, 1951. The term linear programming originated in SCOOP; it might have been linear budgeting, as someone half jested, if air force planners had not been producing projections called programs.
2. At the time we spoke of coder (one who wrote in Univac C-10 code) and of coding. Our subsequent change to programming was slowed by Hq's wide usage of that term.
3. J.P. Eckert et al., "The UNIVAC System," Rev. Electronic Digital Computers, Joint AIEE-IRE Computer Conf., 1951. Also, in the Proc. AIEE-IRE-ACM Computer Conf., 1952, H.F. Welsh and H. Lukoff, "The Uniservo-Tape Reader and Recorder"; L.D. Wilson and E. Roggenstein, "UNIVAC Input Devices"; and E. Masterson and L.D. Wilson, "UNIVAC Output Devices." Because these meetings prefaced the annual Eastern Joint Computer Conferences, their records, where available, are likely to be filed with Proc. Eastern Joint Computer Conf., AFIPS Press.
4. By chance, the Mark I was of personal interest to me. In 1944, while at Harvard and Massachusetts Institute of Technology for radar school, I roomed in the now-razed Mount Vernon Inn, a large house next to the Commander Hotel and the Harvard campus. I shared a room near one occupied by Richard Bloch, a navy officer who spent long hours with a calculator in the basement of a nearby campus building. An opportunity to see the Mark I was missed when my roommate and I declined an invitation from Dick to visit his "big machine."
5. E.D. Schell, "Introductory Description of the Triangular Model," Computation Division, Hq USAF, 12 Mar. 1954 (from author's file). The model was capable of relating various time-phased activities, starting from stipulated activities (typically combat flying hours) and stepping down row by row through support activities in areas such as refueling, arming, maintaining, housing, transporting, and training. A significant computational feature was that each row depended only on results stipulated or already computed in rows above.
6. The Defense Calculator went through name changes; it is remembered as the IBM 701. The TPM became the IBM 702.
7. Programming for the Univac Fac-tronic System, Remington Rand Inc., Eckert-Mauchly Division, 1953.
8. E.D. Schell, APA-3b History for Six-Month Period ending 31 December 1952, SCOOP progress report, 15 Jan. 1953 (from author's file). "APA-3b" was Hq shorthand for Mathematical Computation Branch.
9. L.R. Johnson, "The Installation of a Large Electronic Computer," Proc. ACM, ACM, 1952.
10. E.D. Schell and L.R. Johnson, The Mathematical Computation Branch: Origins, Functions, and Facilities, unpublished SCOOP report, August 1953 (from author's file).
11. M.K. Wood, "The Use of Interindustry Economics in Industrial Mobilization Planning," remarks to the Washington Industrial Round Table, 6 Jan. 1952 (transcript in author's file). Out of concern that in a future war the military might seek forces and operations large enough to collapse the civilian economy, in early 1948 the Committee on Inter-Industry Economics was established in the Executive Office of the President with members from the National Security Resources Board, the Council of Economic Advisors, and the Bureau of the Budget. The committee wanted studies made of the utility of mathematics as a tool for testing the economic feasibility of mobilization plans. Funding and responsibility went to the Department of Defense, downward to the air force, and finally to Project SCOOP—a likely candidate because of its interest in mathematics and planning. The endeavor met with technical challenges in economics, mathematics, computation, and matching of aggregated industrial and military activities. In retrospect, it is sobering to note the high level of concern in 1948 over a possible World War III. As I recall, by 1953 a degree of political distrust was attaching to any activity that conjoined the terms economics and planning; coincidentally, at year's end the Planning Research Division became the Computation Division.
12. R. Kopp, "Experience on the Air Force UNIVAC," Proc. Eastern Joint Computer Conf., AFIPS Press, 1953.
13. H. Lukoff, From Dits to Bits, Robotics Press, 1979.
14. George Dantzig left Project SCOOP in 1952 to join RAND, an early think tank, and to teach at California universities. In a commemorative issue of the Inst. for Operations Research and the Management Sciences, Dantzig traced the genesis of linear programming, and Saul I. Gass, a SCOOP participant, recalled some of the earliest linear programming computations; Operations Research, vol. 50, no. 1, 2002.

Index Terms:
Marshall K. Wood, Pentagon Univac II, Project SCOOP, Simplex method, Uniservo, Univac, Univac factory, Univac maintenance, Univac Serial 2
Lyle R. Johnson, "Coming to Grips with Univac," IEEE Annals of the History of Computing, vol. 28, no. 2, pp. 32-42, April-June 2006, doi:10.1109/MAHC.2006.27
Usage of this product signifies your acceptance of the Terms of Use.