Anne Fitzpatrick and Mary Croarken, Editors
The document "General Report on Tunny with Emphasis on Statistical Methods, " written in 1945 by I.J. Good, Donald Michie, and Geoffrey Timms, was declassified by the British Government in 2000. The report is in two volumes. The first, 276 foolscap pages, is largely statistical. The second, 229 pages, is concerned with the machines used in the cryptanalysis of certain high-level German codes and includes the Colossus machines. These machines, although not publicly acknowledged in the postwar era, had a significant influence on computer development in Britain through the work of Max Newman and Alan Turing.
The report at the Public Records Office in Kew, London, is a carbon copy and is difficult to read in places. The report is scheduled to appear, in 2002, in book form and on the Web as searchable text. The project is being carried out by Whitfield Diffie and J.V. Field, editors, with help from Donald Michie, under the aegis of the Forum for the History of Science, Technology, and Medicine. Introductory essays by Whitfield Diffie, Donald Michie, and Harry Fensom are planned to give context to the report. There will also be short biographies of the three original authors.
MIT Press has agreed to publish the book and make the text available on the Public Record Office Web site. The necessary funding is being supplied by the MIT Press, the London Mathematical Society, and the Royal Statistical Society. Other grants are also being sought.
At the entrance to the Faculty of Science offices in the Biological Sciences Center, University of Alberta, are three large display cases measuring 7 feet high, 4 feet deep, and 30 feet wide. These cases hold exhibits--often unchanged for years--representing some of the faculty research, usually in the biological sciences. Therefore, it came as a surprise when the Department of Computing Science was given a free hand to mount an exhibit related to computing in one of the cases. The resulting exhibit, entitled "Computation: From Abacus to Silicon Chip, " which was completed in July 2001, shows in brief the evolution of calculating devices from counting boards and abacuses to the latest computing facilities at the university. It was created with artifacts borrowed from the department's filing cabinets, bookshelves, display cases, and office walls. Figure 1 shows a front view of the exhibit.
Figure 1. Exhibit at the University of Alberta.
The backdrop gives, on one side, the remark by Gottfried Wilhelm Leibniz that "[I]t is unworthy of excellent men to lose hours like slaves in the labour of calculation, which could be safely relegated to any one else if the machine were used, " against a background of Leibniz's portrait. The other side gives the well-known woodcut from Margarita Philosophica, an early 16th-century encyclopedia containing a treatise on arithmetic. It shows a contest supervised by Lady Arithmetic between Pythagoras (supposedly the inventor of the Hindu-Arabic numerals) and Boethius, a 6th-century Roman scholar. Boethius is using the new numerals and Pythagoras, an abacus. Separating the two is a 54-inch slide rule that was used for promotional purposes in the University of Alberta bookstore and occasionally in computing science classes.
One side of the hanging panel shows a variety of mechanical and tabular aids to computation including Chinese and Japanese abacuses, a one-foot slide rule used by Professor Keith Smillie as an undergraduate in the 1940s, a Curta calculator used until the 1970s by a professor in the Faculty of Medicine, and several electronic pocket calculators starting with one of the early designs and ending with one the size of a business card, purchased specially for the exhibit. The Curta calculator, first manufactured in 1948, was designed during World War II by a concentration camp prisoner who worked out the basic design in his head and perfected it after his release in 1945.
The other side of the hanging panel shows pictures of a few of the electronic computers that have been used at the University of Alberta including an LGP-30, the University's first computer, acquired in 1957. The LGP-30 cost Canadian $40,000, weighed 800 pounds, and was the size of a desk, containing 113 vacuum tubes and 1,350 diodes. Storage consisted of a magnetic drum with 4,096 32-bit words, and input and output was by means of a Flexowriter, an electric typewriter with a paper tape reader and punch. To illustrate programming, a simple numerical example--finding the sum of the first n positive integers by successive summing of the n integers--is shown programmed in LGP-30 machine language, Pascal, and Java.
Also shown on the hanging panel are two sets of mathematical tables that were used at the university. On one side, there is a copy of J.W. Campbell's Mathematical Tables published locally in 1929 and reprinted in 1946. Campbell came to the university in 1920 and was well known for his work in astronomy and classical mechanics. The tables give values for the common logarithm, square, cube and reciprocal, and circular functions as well as an extensive table of hyberbolic functions computed by Campbell. The other side of the panel gives a copy of Knott's Four-Figure Mathematical Tables, first published in 1900, which were loaned to students writing examinations in the Department of Mathematics.
On the floor are a Model SCM Marchant desk calculator, a Multo calculator, and an IBM 5100 personal computer. This model of the Marchant calculator, used in the 1940s and 1950s, performed multiplication and division by repeated additions and subtractions and carriage shifts. The Multo calculator was used by students in the Department of Mathematics for statistical calculations until the 1960s. The IBM 5100 minicomputer, developed in the mid-1970s, was the immediate precursor of the IBM personal computer, first released in 1981 with the model number 5150. The IBM 5100 weighed about 48 pounds and cost about Canadian $15,000. It had a 64-Kbyte random access memory (with more permanent memory provided by a magnetic tape cartridge), and a monitor that displayed 16 64-character lines. Output could also be displayed on a TV monitor. The IBM 5100 was used in computing science for instruction in both APL and Basic.
This exhibit was developed with the cooperation of the staff of Museums and Collections Services and with financial support from the Faculty of Science. Some technical information was also provided by the Smithsonian National Museum of American History, Washington, D.C., and the Computer Museum History Center, Mountain View, California.
The Web site http://www.cs.ualberta.ca/~smillie/Calculators.html features four photographs of the exhibit, a shorter version of this text, and links to several sites relating to the history of computation.
Gene Myron Amdahl recounted the founding of Amdahl Computer to kick off the Fall 2001 series of lectures at the Computer Museum History Center. Museum CEO John Toole welcomed the crowd to the lecture on 5 September at the NASA Ames Research Center. Amdahl recalled that an IBM economist had asked Amdahl, then director of the IBM Advanced Computing Systems Laboratory in Menlo Park, to help explain the lack of theoretical customers for the IBM 360--then IBM s top-end system, a system for which Amdahl had served as architect. He drew up a price-performance chart for IBM mainframes of the late 1960s. Amdahl saw in the chart a market opportunity, specifically for a more efficient 360-compatible mainframe. Frustrated with IBM management's unwillingness to try new design ideas, Amdahl and Ray Williams left and founded Amdahl in 1970. Venture capitalists and underwriters remained skeptical of any business plan that called for competition head-on with IBM in its core market, so Amdahl relied on corporate partners like Fujitsu for the funds to demonstrate his approach. Likewise, most semiconductor firms declined to make his chip, which appeared to have the wrong level of integration. One key breakthrough came when his team devised a chip package that let them air-cool a 100-gate chip, while IBM had to water-cool a 35-gate chip. The result was a mainframe that matched IBM 360 performance at a fraction of the size and cost. With commercial acceptance surging following a sale to Massachusetts Mutual Life Insurance, Amdahl captured 22 percent of the mainframe market at IBM levels of profitability and raced IBM to better price-performance points throughout the decade. Audience questions focused on how Amdahl achieved compatibility (by relying on the published Principles of Operation available to all programmers) without violating IBM intellectual property (the consent decree forced IBM to make its patents available).
Some upcoming lectures at the Computer Museum History Center are listed here; check http://www.computerhistory.org/events/latest/ for schedule changes and updates.
Contributors to this issue of Events and Sightings are Jeffrey Yost, Charles Babbage Institute; Sue Barnes, Fordham University; Keith Smillie, University of Alberta; Glenn Bugos, The Prologue Group; Mary Croarken, National Maritime Museum, Greenwich; and Anne Fitzpatrick, Los Alamos National Laboratory.