UNESCO EOLSS SAMPLE CHAPTERS HISTORY OF COMPUTING. Jeffrey R. Yost Charles Babbage Institute, University of Minnesota, USA

Transcription

1 HISTORY OF COMPUTING Jeffrey R. Yost Charles Babbage Institute, University of Minnesota, USA Keywords: punch card tabulators, analog computers, calculators, digital computers, computer history, software history, Internet history, programming, computer industry, software industry, information technology, transistors, integrated circuits, mainframe computers, minicomputers, artificial intelligence, personal computers, computer games, databases, computer graphics, video games, computer security, scientific computing, database, World Wide Web, and cloud computing 1. Pre-digital Computer Computing Technology 2. Advent of Digital Computing 3. Programming Making Computers Useful 4. Early Computer Graphics and Networking 5. Artificial Intelligence 6. Databases and Database Management Systems 7. Components of the Microelectronics Revolution 8. Minicomputing and Supercomputing 9. Scientific Computing 10. Computer Games 11. Making Computing Personal 12. The Internet, Web, and the Computer as a Ubiquitous Communications Tool 13. Computer Security 14. Computing in the Clouds Glossary Bibliography Biographical Sketch Summary Computers have transformed the way people in the developed, and increasingly the developing world work, think, play, shop, and socialize. For centuries humans have designed and developed computational tools and machines to aid with calculation from the ancient abacus (perhaps as early as BC), Napier s logarithmic tables and bones (early 17 th century), and Charles Babbage s plans for massive mechanical calculating engines (early 19 th century) to sophisticated accounting machines and giant analog computers (first half of the 20 th century). While these technologies were complex, meaningful, and often highly useful to individuals and organizations in particular fields and endeavors, the digital computer first developed during World War II evolved by the end of the 20 th century to broadly and forever, change human perspectives, predilections, practices, and possibilities. It unquestionably became a machine, and more accurately a set of systems, that changed the world. In time, digital computers, and advanced programming facilitated not only unmatched number crunching accuracy and speed, but also highly versatile information processing and textual, audio, and graphical communications. As such, computers became not only a tool for various types of scientific and business data processing work, but also an

2 instrument of leisure that individuals and groups use for entertainment, social interaction, and other ends. This chapter briefly surveys select pre-digital computational developments (the abacus, calculators, punch card tabulation machines, and analog computers), before focusing on the successive waves of digital computing, software, and networking technology that have so deeply influenced our infrastructural systems and ways of life. It examines the major developments and shifts within computing, and sheds light on the underlying changes, as well as continuities, associated with the development and use of these pervasive technologies. 1. Pre-digital Computer Computing Technology For millennia humans have sought to develop tools for computation. The abacus, a technology that might date back in some form more than 4500 years, is generally recognized as the first computational tool (other than using one s hand to count). By creating or utilizing rows of indentations on sand or soil, and positioning and moving stones or other small objects within the rows, ancient abaci were used to keep track of numbers and aid humans as they made calculations. Given the materials that these individuals used, pinpointing the date and place of origin of the ancient abacus most likely will never be possible. Some researchers have argued that the technology existed in Mesopotamia (2700 to 2300 BC) where Sumerians may have constructed the first primitive (sand/soil and stones) abaci. Others contend that the Chinese were the first to invent the technology, with some claims to a similar time-frame. Definitive evidence (of a more modern abacus), however, only exists for a much later device. Abaci more nearly matching what we now think of as the abacus (pierced beads of some sort moved along a thin rod or wire that is held in place by a frame) are most commonly associated with the Far East. These more modern abaci were used in China in the early 13 th century and the technology was transferred to Korea and Japan in the succeeding several centuries. Of the European abaci, none are more famous than Napier s bones. John Napier was born in Scotland in 1550, though much of his early life remains a mystery. He is most famous for his invention of logarithmic tables. In 1614 Napier published Mirifici Logarithmorum Canonis Descriptio, a book with 57-pages on the description of relations between arithmetic and geometric series, and 90-pages of logarithmic tables. He died three years later, but in that time began collaborating with Henry Briggs, who popularized and built upon Napier s work on logarithms after Napier s death. In the latter part of his life, Napier also developed his bones a multiplication rod abacus capable of relatively complex multiplication that he used in his research and creation of his logarithmic tables. Napier s bones represented a different form of abacus that used a system of numbers on rods, and sequences of positioning of these rods, to multiply. It was a precursor to the slide rule a technology developed several decades later in Great Britain. The use of actual animal bones gave the multiplication rods a name, bones, that remained long after Napier s death.. Today we think of the abacus often constructed of a finished wood or plastic frame,

3 thin metal bars or wires, and pierced, movable beads or small plastic cylinders as a children s toy. To (possibly) the ancient Sumerians, and (definitely) the Chinese, Japanese, Koreans and others from the late Middle Ages, as well as users of Napier s bones in the 17 th century, it was anything but a toy. It was an important tool to assist with various mathematical computations related to the natural world and humandeveloped systems, particularly those within government and commerce. Concomitant to the abacus, mathematical tables were an idea born millennia ago that began to have profound influence during the late Middle Ages. The impact of mathematical tables only heightened in the Renaissance and beyond. Mathematical tables were fundamental to the Scientific Revolution and continued to see widespread use in the sciences (particularly astronomy and physics) as well as economics and business. Long before calculating devices became inexpensive, and with unemployment rising dramatically as a result of the Great Depression, U.S. President Franklin D. Roosevelt s Works Progress Administration s Mathematical Tables Project (initiated in 1938) put to work hundreds of unemployed individuals to develop more than two-dozen volumes of calculated tables. Gertrude Blanch was the technical director of this decadelong effort that was subsequently merged into the National Bureau of Standards Computation Laboratory. Blanch, and those who worked under her leadership, was one of the many human computers. Prior to World War II, the term computer referred to humans who computed an occupation not a device. Unlike many occupations of the era, it was open to women. Particularly after the entrance of the U.S. into World War II, the U.S. government recruited women trained in mathematics to become human computers. Dozens of women were hired by the Army Ballistic Research Laboratory (which was collaborating with the University of Pennsylvania s Moore School of Electrical Engineering) to calculate ballistic firing tables (tables showing the coordinates for artillery trajectories given wind, atmospheric pressure, and other factors). Six of these women Betty Snyder, Jean Bartik, Kathleen McNulty, Marilyn Westcoff, Ruth Lichterman, and Frances Bilas commonly referred to as the ENIAC girls, would be the first to program the pioneering Electronic Numerical Integrator and Computer (the first meaningful, general-purpose digital computer in the U.S.). Alongside earlier developments that profoundly and directly influenced digital computing, there were also a number of computational technologies created between the 17 th century and first half of the 20 th century that had little direct connection to digital computers. Nevertheless some of these technologies were remarkable achievements and ones that may have influenced intermediate technologies, as well as individuals who later did have an impact on digital computing. Chief among these were Blaise Pascal s development of the mechanical calculator, Charles Babbage s designs and plans for massive mechanical calculating engines, and analog computing. Blaise Pascal, a French mathematician, physicist, and philosopher, in 1642 invented the Pascaline, a mechanical calculator that could do addition and subtraction. Approximately 20 of these machines were produced over the succeeding decade, but they were more a novelty of wealthy individuals than a computational tool of direct, broad impact. Nevertheless, the Pascaline influenced Gottfried Leibniz s wheel, a

4 device introduced three decades later that could not only do addition and subtraction, but also multiplication and division. Further, the Charles Xavier Thomas de Colmar arithmometer (first produced in the 1820s and manufactured for the next four decades), and the designed calculating engines of Charles Babbage, were influenced in part the Pascaline and the Leibniz wheel. Charles Babbage, a prominent mathematician, philosopher, economist, and inventor, is often credited as the father of computing based on his designs of the Difference Engine, and a series of technologies collectively referred to as the Analytical Engine, in the first half of the 19 th century. He began work on the Difference Engine in 1822, which was abandoned by the early 1840s, and started designs for the technologies of the Analytical Engine in Neither of Babbage s two designs for the Difference Engine, nor his plans for far more complex analytical engines, resulted in built machines during his lifetime. Finances, challenges of precision manufacturing, and other factors prevented their realization (until 1991 when the Science Museum of London built a working Babbage Difference Engine from the inventor s original designs). Babbage was influenced by the programmed punched card instruction loom of Joseph Marie Jacquard developed in 1801, and among the plans for analytical engine technology was the concept of a stored program or instruction set. A collaborator of Babbage, Ada Lovelace, developed a program for the analytical engine, and is sometimes credited as being the world s first computer programmer (a U.S. Department of Defense-funded object-oriented, high level programming language developed by Honeywell-Bull in the late 1970s and early 1980s bears the name Ada in her honor). Many analog mechanical computational devices were invented between the early 17 th century and mid-to-late 20 th century. These machines modeled phenomena in continuous rather than in the discrete terms (binary digits) of digital computers. They included early tide predictors of the 16 th century that used dials and pointers, and representations of the position of the sun and moon in the sky, to come to rough approximations of future tide levels. William Thomson (Lord Kelvin), a prominent Belfast-born mathematician, physicist, and inventor took tide prediction and associated technology to a new level with his analysis of tidal records in the last third of the 19 th century. Thomson designed a highly sophisticated tide-predicting machine utilizing pulleys to add harmonic time functions that could trace out results on a chart recorder (consisting of tracing needles and paper roll on the side of the large machine). Similar devices to Lord Kelvin s U.S. government-funded tide predictor, that was fully completed in 1911 (four years after Kelvin s death), were used for a half-century. Lord Kelvin also sought mechanical solutions to calculating areas under a curve using integration to solve differential equations. In fact, a number of individuals and companies attempted to use mechanical wheel mechanisms to integrate differential equations in the first-third of the 20 th century. Vannevar Bush, a gifted electrical engineering professor at MIT, who would go on to propose the National Defense Research Committee (NDRC) to President Franklin D. Roosevelt and serve as director of the Office of Scientific Research and Development (OSRD), led a project at MIT to develop by far the most useful and significant pre-world War II machine to calculate differential equations: the Differential Analyzer. Bush s first Differential Analyzer was completed in 1927, and he was not aware of the unsuccessful attempts made by Lord

5 Kelvin to create a similar device more than two-decades earlier. Multiple copies of Bush s Differential Analyzer were built and one went to prominent British computing pioneer Douglas Hartree of Manchester University and another to the University of Pennsylvania s Moore School of Electrical Engineering. The latter was used to calculate ballistic firing tables for the U.S. Army prior to the post-world War II advent of digital computing but aside from it being used in the same setting as the development of the ENIAC (and for the same initial purpose), it was a completely different type of computer and had no meaningful technical connection to the famous early digital computer. The time-consuming set-up of Bush s massive machine of rods and gears, its lack of exact precision with calculations, and problems associated with its mechanisms frequently getting out of alignment were all critical factors to launching the ENIAC project. Analog computers continued to be used for certain purposes long after the advent of digital computers, though these contributions commonly are overlooked. This is likely because analog computers were quite different machines and had little technical connection to the broadly transformative digital computer technology of the post-war era. In contrast to analog computing, pre-digital computer punched card tabulation equipment, and the firms specializing in this area (especially International Business Machines, or IBM), had a profound and continuing impact on digital computer technology. The international office machine industry, which consisted of accounting machines, cash registers, typewriters, and punch card tabulation systems, among other devices, took off in the last quarter of the 19 th century. Given labor shortages, a greater propensity toward automation, well-managed office machine companies, and a larger overall market than any single European nation, the U.S. became the international leader of the punch card tabulation machine industry in the first half of the 20 th century. The punch card tabulation technology of engineer and inventor Herman Hollerith was used on the most demanding information processing task of the late 19 th century processing U.S. Census information. Herman Hollerith, a graduate of Columbia University, invented a path breaking punch card tabulation machine that utilized electronic sensors (to detect punched holes in cards) in the mid-1880s. He formed a company in Washington, D.C. and sold his machines and services on a small scale before securing the contract for equipment to process the 1890 U.S. Census. Coupled with other census applications in the U.S. and abroad, Hollerith also successfully secured contracts with the railroad and other industries for his punch card tabulation equipment. In 1911 Hollerith retired and sold his company, then called the Tabulating Recording Company, to financier Charles Flint who merged it with two other firms, the Computing Scale Company and the International Time Recording Company. The merged entity became the Computer Tabulating Recording Company, or C-T-R. C-T-R would change its name to International Business Machines in While IBM produced numerous successful products in the interwar period industrial scales, time-clocks, and electric typewriters its focus was on punch card tabulators and the standard cards used by these machines. It competed in this industry with Remington Rand (a firm resulting from the combination of Remington Typewriter Company,

6 Powers Accounting Machine Company and Rand Kardex Company in 1927), which had a substantial, but considerably smaller, tabulating machine division. IBM benefited greatly from, and created a strong degree of customer lock-in with, its 80-column punch cards first introduced in These became a near standard in many industrial fields and contributed significantly to IBM s continuing success. The message that appeared on many of these cards, Do not fold, spindle, or mutilate is a very well known phrase to this day despite the fact that the cards are essentially (though not entirely) obsolete. A couple decades before the famous merger that would result in C-T-R/IBM, John Patterson s successful National Cash Register Company, the world s leading firm for cash registers, hired a young salesman, Thomas Watson. Watson would learn the art of sales from Patterson, and become the firm s leading salesman. After a falling out between Patterson and Watson, C-T-R hired Watson in 1914 to be its General Manager, and soon thereafter its president. For the next four decades Thomas Watson, Sr. would provide strong leadership for C-T-R/IBM and grow it into a corporate giant a role that he would officially hand over to his son Thomas Watson, Jr. in Thomas Watson, Sr. daringly invested in the firm throughout the depression as his office machine competitors generally suffered more severely from the challenging times. IBM benefited substantially from the New Deal and its associated need to process ever more government information on the citizens of the United States. Also IBM s stronger financial position than its competitors was self-reinforcing. In addition to inspiring confidence that IBM would be around long-term (which was very important because customer organizations were investing in standard systems), the company was able to focus on a model of leasing as opposed to just selling its machines at a time when many organizations could not afford to purchase capital equipment for data processing. By the start of World War II, IBM dwarfed its competitors in the punch card tabulation machine field, and it was easily the world s largest office machine producer. Critically, IBM had far more resources to invest in research and development (R&D) than its competitors. Watson, Sr. insightfully invested aggressively in electronics during the 1940s, placing IBM in an excellent position for rapidly moving into the nascent computer industry when a market opportunity first seemed to appear in the early 1950s. IBM had by far the best sales and service infrastructure for business data processing in pre-world War II office machines. Further, the key input-output technology for digital computers was an area in which IBM was a world leader: punch cards and punch card tabulation systems. These technological (punch card and punch card tabulation) and organizational factors (size and talent of its sales force, and R&D infrastructure) represented critical organizational capabilities that helped IBM achieve computer industry dominance in the second half of the 1950s and 1960s. Absolutely fundamental was IBM s heavy investment in the late 1940s and early 1950s in electronics to strengthen its hold on the data processing machine market in the post-war era. While Watson Sr. saw electronic data processing machines as the future, he remained skeptical about computers, and did not enter into the electronic digital computing field until a clearly recognizable market demand for these machines emerged. It was individuals and organizations funded directly (design and development contracts) or indirectly (initial customer base) by the U.S. Federal Government (and especially the

7 Department of Defense, or DoD), that would pioneer digital computing technology in the U.S. In Europe, universities, government laboratories, and electrical equipment firms would lead the way to comparable early achievements. 2. Advent of Digital Computing The question of the first digital computer has been hotly contested by pioneers and academicians, as well as by computer firms and their lawyers in patent litigation. It was the key subject to a U.S. federal court case of the early 1970s Honeywell versus Sperry Rand in which a 1973 ruling invalidated the ENIAC patent. With substantial real and perceived wartime applications, the new field of digital computing evolved rapidly during World War II and its early aftermath. Crediting priority for a technology as broad as digital computing depends on what underlying features and concepts are emphasized. In general, the earliest efforts produced specialized, rather than general computers, and electromechanical machines rather than fully electronic digital computers. The question is further complicated by meetings where important concepts may have been shared, and might have influenced the work of others. Suffice it to say, the pioneering computing accomplishments of Germany s Konrad Zuse, Britain s Tommy Flowers, and the U.S. s George Stibitz, Howard Aiken, John Vincent Atanasoff and Clifford Berry were profound and should all be celebrated in computer history. That said, the widespread attention given to University of Pennsylvania s J. Presper Eckert and John Mauchly for the first general-purpose electronic digital computer, Electronic Numerical Integrator and Computer (ENIAC), clearly is warranted. Similarly, the concept of a stored program and the achievement of a stored program computer were also fundamentally important to advancing digital computing technology. John von Neumann authored a report in 1945 that outlined computer architecture that shaped the future of computing, while Manchester University Baby computer was the first to operate with a stored program, and Cambridge University s EDSAC the first larger-scale computer of this type. Konrad Zuse, a civil engineer who initially became interested in computers only for their application to his profession, designed a mechanical computer, Z1, in the mid- 1930s. The machine was completed by 1938 but was unreliable and operated only a short time. He followed it with the Z2 in 1939, a relay-based machine that also had severe reliability problems. Each machine, while problematic in its operations, showed promise and was critically important in securing funding and making the next one possible. Zuse completed the Z3 in 1941, a computer consisting of 2000 relay switches. The computer was destroyed in a bombing raid on Berlin in Having impressed the government aviation laboratory, Deutsche Versuchsanstalt für Luftfahrt, at a demonstration two-years earlier, Zuse secured funding for the Z4. Zuse began work on the Z4 in 1943 and completed it in The project was a closely guarded secret funded by the German government. Unlike its predecessor, it could do conditional branching (it had if/then and go to capabilities) and was a pioneering accomplishment in digital computing. Germany s defeat in World War II led to a general lack of recognition of Zuse s great accomplishments by British and U.S.

8 scholars for more than three decades, but by the 1980s this important work gradually began to be recognized. Another figure whose work was profoundly important, but largely has been overlooked is George Stibitz. He was a mathematical physicist by training and worked as a research scientist at Bell Laboratories in the 1930s and 1940s. In 1937 he developed a model digital computer using electromechanical relays as switches. Bell Labs provided him a research team the following year to produce a more sophisticated relay switch computer capable of complex calculations. In 1940, at the American Mathematical Society, he used this machine to demonstrate the first instance of remote computing using a teletype. These fundamentally important achievements, much like Zuse s lack of recognition outside Germany for many years, have not received adequate attention from scholars. One key explanatory factor is that AT&T was prohibited from entering the computer industry by a 1956 consent decree. Without this, AT&T and Bell Labs accomplishments and position in computing likely would have continued to grow, and Stibitz s would be more of a starting point to a broader history rather than a mere historical footnote. A third example of neglected/delayed recognition in the history of computing is the work of Tommy Flowers, an electrical engineer who joined the telecommunications branch of the British General Post Office in 1926 (moving to its Dollis Hill research station in 1930). Bletchley Park s Alan Turing approached Flowers during the early 1940s for a project that was discontinued, but Flowers impressed both Turing and Bletchley mathematician Max Newman. Around this time, Bletchley s Bill Tutte had noticed statistical biases in wartime German Enigma code created with a Lorenz SZ40/42 rotor cypher machine. In February 1943 Flowers proposed a programmable digital computer development project, Colossus, to build a Mark I machine, made up of 1800 vacuum tubes, which could help emulate the Lorenz machine, recognize biases, and assist with deciphering the coded messages. Flowers and his team designed and built a Colossus Mark II, comprised of 2400 vacuum tubes, which was operational at Bletchley by June 1, The machine contributed information important for the D- Day invasion less than a week later, and more broadly, the 10 Colossus computers, coupled with the Bletchley code-breaking engineers and scientists, helped assure a victory for the Allied forces. The Colossus computers were a closely guarded secret and remained so to a large degree after the war to assure that Britain s ability to decipher coded messages would not be compromised. Two Colossus computers were put into service by the British intelligence services organization GCHQ. The others were completely dismantled, and Flowers had to burn his blueprints. Only in the 1970s, when the machines and methods were obsolete, did the British government fully lift the veil of secrecy on Colossus. Simultaneous to these developments, a very different Mark I computer development project was underway in partnership between IBM and Harvard University s Howard Aiken. Aiken, a professor of applied mathematics, desired a massive calculator to solve complex nonlinear differential equations. He was aware of IBM s partnership with Columbia University for the Thomas J. Watson Astronomical Computing Bureau located on Columbia s campus. Wallace Eckert, a visionary Columbia astronomer, was the force behind the launch of the laboratory and served as its longtime leader. Aiken

9 astutely enlisted Eckert s assistance and connections within IBM to solidify a partnership between the office machine firm and Harvard University. As with the Watson lab at Columbia University, IBM s decision to participate and commit resources was more to achieve good public relations supporting higher education and scientific research than to enter the computing field. Computation at the time was seen as a small, if not non-existent market compared to IBM s main area of mechanical or electromechanical punch card tabulating machines for business data processing. The very large electromechanical computer, the Harvard-IBM Automatic Sequence Controlled Calculator, was completed in 1943 and demonstrated at Harvard University the following year. This electromechancial computer, commonly referred to as Harvard Mark I (or simply Mark I), utilized accumulators, punch card and paper tape reader input, and IBM s standard accounting machine register mechanisms. It had a decimal numeral system, could do most multiplication and division problems in less than 20 seconds, and logarithmic problems in roughly a minute. While successful in many respects, it had no real connections to the fully electronic digital computers that emerged at the end of World War II and is more notable for employing Grace Murray Hopper, a seminal figure in software history in the digital computer era, than any direct technical influence on the future of digital computing technology. Like the Harvard Mark I, the Atanasoff-Berry Computer (ABC) was driven by an academic scientist. John Vincent Atanasoff completed his doctorate in theoretical physics in 1930 and soon thereafter joined the faculty of Iowa State University. The slow and tedious nature of scientific calculating on Monroe and other mechanical machines sparked his interest in developing better computational devices. In 1937 he first conceived of a fully electronic computer, and between 1939 and 1942, he built the ABC with his graduate student Clifford Berry. As a fully electronic machine it differed from the electromechanical Z3, Z4, and Harvard Mark I computers. Atanasoff received a modest grant from the Iowa State Agronomy Department, which was also interested in a more efficient and accurate tool for calculation. He also received a partial sponsorship from the Research Corporation of New York. The ABC machine was notable for being binary and calculating fully with electronics, though it had no central processing unit, it was not programmable, and it was not consistently reliable. The computer offered no conditional branching, and like the Colossus computers, the Z3, and Harvard Mark I, it did not pass the theoretical Turing test of a universal machine (a fundamental thought experiment of influential mathematician Alan Turing that distinguishes between limited specialized computers and general-purpose computers). A Turing complete or universal machine has the ability to simulate the most basic or primitive computer, and the most complex one as well. The ABC was essentially abandoned when Atanasoff left for government wartime work and never returned to his position at Iowa State University. The importance of the ABC resurfaced in the famed Honeywell versus Sperry Rand U.S. court case of the early 1970s. In the ruling for this case, the ABC was the fundamental basis for recognizing an earlier electronic computer than the ENIAC. Hence, it contributed significantly to the federal court s invalidating the ENIAC patents on priority for electronic digital computers. This priority dispute has also spawned much debate by scholars and others regarding what concepts or methods, if any, John Mauchly took from the ABC in designing the ENIAC.

10 Antanasoff had heard a talk given by Mauchly at the American Association for the Advancement of Science in late They met after the talk and Atanasoff invited Mauchly to visit Ames, Iowa and see the ABC. Mauchly accepted this offer and visited Atanasoff over a five-day stretch in June While Mauchly strongly stated he took no ideas from this meeting and introduction to ABC technology, some researchers have argued that he did draw ideas from this visit. The evidence on this question, however, is inconclusive and given the lack of documentation, the level and type of influence, in any, will likely remain pure speculation. What is certain is that the ABC was an obscure project away from centers of all other early computer design and development. It was a project that received only small-scale funding compared to the Harvard Mark I and the ENIAC. The ABC worked only for a short time, was not programmable (and thus, very limited in its applications), and had very distinct reliability problems during its quite brief life. In contrast, the ENIAC project was a well-funded effort by the U.S. Department of Defense that took place at a major electrical engineering department, the Moore School of Electrical Engineering at the University of Pennsylvania. The project grew out of the Army Ballistic Research Laboratory s need to rapidly and accurately produce extensive ballistic firing tables during World War II. Bush s Differential Analyzer, used on this task previously, was too slow, required frequent readjustment of mechanical parts, and provided only close approximations rather than the exact calculations that were desired. John Mauchly, a University of Pennsylvania physics instructor, and Arthur Burks, a philosopher at the University of Michigan, participated in the Moore School of Electrical Engineering s Science and Management War Training Program an effort to prepare scientists and engineers to contribute to the war effort. Both Mauchly and Burks accepted offers for posts at the Moore School following the program and in the summer of 1942 Mauchly proposed building an electronic digital computer to aid with calculations of firing tables. In August, Mauchly submitted a formal proposal to the Moore School and the Army Ballistic Research Laboratory s (BRL) liaison to the school, Herman Goldstine. Goldstine was a strong backer of the proposal and helped secure a $400,000 BRL contract for the design and development of the ENIAC. J. Presper Eckert, a highly talented research associate and engineer at the Moore School, was also enthused by the plan and ended up partnering with Mauchly on the project. Mauchly provided the grand idea and broad conceptualization and Presper Eckert the critical electrical engineering expertise to make it successful. Despite feverish work on the project, the ENIAC was not completed until several months after the war. It was a computer of unmatched scale with dimension of 30 feet by 60 feet, and containing an unprecedented 18,000 vacuum tubes. Like the Colossus, Mark I, and Mark II, it was programmable by patch cables and switches. However, the ENIAC, unlike the Colossus machines, was capable of conditional branching, and thus was Turing complete. It qualified under Turing s 1936 thought experiment or test as a universal Turing machine. The only computer meeting this criterion prior to ENIAC was Zuse s Z4, but that computer was of smaller scale than ENIAC and had the liability of being electromechanical rather than fully electronic. In short, the ENIAC, operational in late 1945, was the fullest realization of electronic

11 digital computing to that time. Further, while key elements of the technology were guarded by the military, it was not relegated to complete secrecy, as were the Colossus computers. Nor was ENIAC obscured and inhibited by wartime defeat, like Zuse s Z4. Practically the ENIAC was quickly put to use at Los Alamos National Laboratory in late 1945 to calculate equations in support of the U.S. military s classified project to develop a hydrogen bomb. While its applications were secret, its existence was not. The computer was officially dedicated and became broadly known to the public in The one critical element that ENIAC (as well as all its predecessors) lacked was stored program capabilities. A chance meeting on a train platform between Herman Goldstine and Hungarian-born, Princeton Institute for Advanced Study mathematician John von Neumann led to the latter s involvement with the concluding stages of the ENIAC project. Von Neumann advised on ENIAC, but far more importantly, worked closely with Eckert and Mauchly, as well as others, on its successor Electronic Discrete Variable Automatic Computer (EDVAC). Though he undoubtedly benefited from the opinions and insights of Mauchly, Eckert, and others, von Neumann, alone, wrote the famed report, A First Draft of a Report on the EDVAC. This report included articulation of the storedprogram concept that an internal memory device should contain both the instructions of a program as well as the data to be processed. This concept would become critical in facilitating the development of programming languages. The report established the future structure, or dominant architecture, of computing for decades to come an architecture that became known as the von Neumann architecture. Von Neumann also helped shaped the future of computing by establishing a project at the Institute for Advanced Study (IAS) to build a major digital computer that was completed in the early 1950s. This computer and more accurately its prototype several years earlier was the model for a number of organizations building their first digital computer, including the Air Force-supported non-profit research corporation, RAND (which drew its name from research and development). Owing an intellectual debt to von Neumann, the RAND Corporation s computer was named Johnniac in his honor. Other national and international copies of the IAS prototype machine (many of which were completed before the actual IAS machine) included ILLIAC at the University of Illinois, ORDVAC at Aberdeen Proving Grounds, MANIAC at Los Alamos National Laboratory, as well as BESK in Stockholm, PERM in Munich, and SILLIAC in Sydney. Though von Neumann and the U.S. team that had worked on the ENIAC established the stored program concept, computers developed in Britain were the first to develop an operational stored program computer. The Moore School held an invitation summer school in 1946, the Moore School Lectures, which circulated knowledge of von Neumann s report and related computing research to representatives from the major computer centers in allied nations. Britain also held important international computer conferences in the late 1940s and early 1950s that further circulated international knowledge in computing. The U.S. and Britain were the leading allied nations conducting computing research and development, and they were also the only countries to have a meaningful computer industry between the mid-1940s and early 1950s. Eckert and Mauchly left the University of Pennsylvania in 1946 to establish a computer

12 firm, the Electronic Control Company renamed the Eckert-Mauchly Computer Company the following year. Their first major project was to develop a universal computer, the Universal Automatic Computer (UNIVAC) it was designed with the idea of targeting the scientific/government market as well as the business applications market. For financial reasons the firm delayed UNIVAC and came out with the smaller, less ambitious Binary Automatic Computer (initially designed for Northrop Aircraft and completed in 1949), BINAC, before completing UNIVAC in Also in 1946, a group of former wartime U.S. Naval engineers began a company in St. Paul, Minnesota, Engineering Research Associates (ERA). This firm, along with conducting various non-computing naval electronics work, entered the computer industry with a classified digital computer project for the Navy, ATLAS, and a declassified spin-off commercial computer, the ERA A number of other start-ups formed and the major U.S. office machine firms also entered the industry by the early 1950s, primarily through acquisitions. This included Burroughs, National Cash Register (NCR), and Remington Rand. The latter would take over both Eckert-Mauchly Computer Company and ERA in the early 1950s to become the leading international computing firm prior to IBM s entrance into the trade in 1953, and its introduction of the IBM 650 in the mid-1950s. The IBM 650 would firmly establish IBM as the industry leader, a position it only extended in the following years and decades. In Britain, Bletchley Park had been an early leader in computer research, development, and use with the Colossus computers, but work at Bletchley was not continued after the war. Top scientists and engineers from the code-breaking project at Bletchley, however, continued to provide leadership in British computing in different settings, particularly Alan Turing and Max Newman. Two universities, University of Manchester and Cambridge University, and a government laboratory, the National Physical Laboratory (NPL), became the top British centers for computer research and development in the second half of the 1940s and 1950s. Turing led the work at the NPL, while Newman started a major computer project at Manchester. Mathematical physicist Maurice Wilkes returned from war service and rejuvenated a dormant computer laboratory at Cambridge. Newman, a mathematician by training, secured funding to build an EDVAC-type computer at Manchester. Newman s colleague on the project, John Comrie, had been among the participants in the Moore School Lectures. While von Neumann had established the stored memory concept, executing it, getting memory to work as designed, was an extremely difficult task. Newman succeeded in recruiting highly gifted radar engineer Frederic (F.C.) Williams from the government s Telecommunications Research Establishment. Knowing that the memory was the great challenge, Newman, Williams, and another engineer, Tom Kilburn focused on building a small computer to test out the memory system that was developed around an off-the-shelf cathode ray tube. Both the mathematicians, Newman and others, and the engineers, Williams, Kilburn, and others, were critical to the effort. On June 21, 1948 the Manchester Baby Machine, commonly referred simply as Baby, successfully demonstrated the feasibility of the stored program concept. The British also developed the first full-scale von-neumann architecture computer a

13 great accomplishment by Maurice Wilkes and his small team at Cambridge University. Wilkes, too, attended the Moore School Lectures, though arrived late due to funding delays. On his boat trip back to Britain he began plans for an EDVAC-style machine that in deference to its origin, he named the Electronic Delay Storage Automatic Calculator (EDSAC). Drawing on his wartime experience as an engineer working on radar, he and his colleagues designed and built mercury delay line memory. EDSAC had 3000 vacuum tubes, one-sixth that of ENIAC, yet unlike the proof-of-concept Manchester Baby Machine, it was a true general-purpose computer that could be, and was, programmed for numerous applications. Justly deserved, much has been made of these early British accomplishments in computing. Even before the details of Colossus were widely known, the leadership British centers took in realizing the stored program concept, both in a small-scale test computer, Baby, and a full-scale computer, EDSAC, are impressive. These machines were the first to realize the architecture that has been central to computing now for roughly six decades. Britain was also an important force in the early computer industry but this faded over time. In Britain, far more than the U.S., electronics firms, such as Ferranti and English Electric, became significant players in the early industry (in the U.S., General Electric and RCA never put a comparable proportion of their resources into their computer businesses). Like the U.S., British office machine firms also played a substantial role. Britain s largest office machine firm, the British Tabulating Machine Company, which had a licensing agreement with IBM until 1949, and Powers-Samas, entered the computer industry at the start of the 1950s. As British firms struggled against international powerhouse IBM, the British government favored a national champion model and helped facilitate mergers to create one dominate British firm by the late 1960s, ICL. Far too much has been made of these firms, and later ICL s, inability to match early technical achievements at British computational centers and become giant international firms like IBM. Some scholars, such as John Hendry, have placed blame on failures within government policy, and to a lesser degree, company management, and presented Britain s computer history within a larger framework of British industrial decline. Such an interpretation is misplaced, and neglects the broader context of IBM s position and the U.S. market. IBM, which had the advantage of worldwide leadership in punch card tabulation, between the mid-1950s and mid-1960s successfully transferred much of its huge tabulation machine customer base to digital computers. The U.S. had a much larger market due to its population, and also because early automation was more practical in the U.S. there was a greater shortage of workers and higher wages than in Britain. Furthermore, IBM s U.S. competitors, Sperry Univac (a product of the 1955 merger of Sperry Corporation and Remington Rand), Burroughs, and NCR, were all significantly larger than early British computer firms, and all struggled mightily against IBM. The two American electronics giants (RCA and General Electric), companies that might have had a better shot of investing heavily and challenging IBM, were never fully committed to the computing field and exited the industry in less than two decades. Likewise, other European companies, including Bull in France and Siemens, HeinzNixdorf, and Telefunken AEG in Germany, all had a modest degree of domestic success, but overall they struggled against IBM, which continually expanded its presence overseas in the 1960s and 1970s.

14 TO ACCESS ALL THE 55 PAGES OF THIS CHAPTER, Please register at: using the Preferential Code: SA9253 Enjoy this provision during the UN Decade of Education for Sustainable Development. Bibliography Abbate, Janet. (1999). Inventing the Internet, Cambridge, MA: MIT Press [Without question the best history surveying the design and development of the Internet]. Aspray William and Paul E. Ceruzzi. eds. (2008). The Internet and American Business, Cambridge, MA: MIT Press [An important anthology on a wide-range of issues on the technological and business history of the Internet and World Wide Web]. Aspray, William. (1990). John von Neumann and the Origins of Modern Computing, Cambridge, MA: MIT Press [A path breaking biography of von Neumann and seminal work on the intellectual history of modern computing]. Bassett, Ross Knox. (2002). To the Digital Age: Research Labs, Start-Up Companies and the Rise of MOS Technology, Baltimore, MD: Johns Hopkins University Press [Major contribution to the overall literature on semiconductor research and development, and the leading source on the history of Metal Oxide Semiconductor (MOS) technology]. Bergin, Thomas J. and Thomas Haigh (2009). The Commercialization of Database Management Systems, IEEE Annals of the History of Computing 31, [Most significant overview article on the history of database management systems]. Brooks, Frederick (1974). The Mythical Man-Month and Other Essays on Software Engineering, Chapel Hill: University of North Carolina [Classic book of essays on software engineering and management drawn from Brooks experience leading IBM s OS/360 project] Campbell-Kelly, Martin and William Aspray. (1996). Computer: A History of the Information Machine, New York: Basic Books [The best overview of the history of computing technology]. Campbell-Kelly, Martin. (2003). From Airline Reservations to Sonic the Hedgehog: A History of the Software Industry [An excellent survey of the software products industry from its origin to the start of the new millennium] Ceruzzi, Paul E. (1998). A History of Modern Computing, Cambridge, MA: MIT Press, [A highly useful survey of digital computing history]. Chandler, Alfred D. Jr. and James W. Cortada, eds. (2000). A Nation Transformed by Information: How Information Has Shaped the United States From Colonial Times to the Present, New York: Oxford University Press [Anthology that places computing and other post-world War II information technology in the broader context of the history of information technology in the United States from the Colonial period forward]. Cortada, James W. (1993). Before the Computer: IBM, NCR, Burroughs, and Remington Rand and the Industry They Created, , Princeton, NJ: Princeton University Press [Path breaking work

15 showing the office machine origins of the later digital computer industry in the United States] Cortada, James W. (2004). The Digital Hand: How Computers Changed the Work of American Manufacturing, Transportation, and Retail Industries, New York: Oxford University Press [First in important three-part series of books on industrial users of computation technology]. Cortada, James W. (2006). The Digital Hand, Volume 2: How Computers Changed the Work of American Financial, Telecommunications, Media, and Entertainment Industries, New York: Oxford University Press [Second in important three-part series of books on industrial users of computation technology]. Cortada, James W. (2008). The Digital Hand, Volume 3: How Computers Changed the Work of American Public Sector Industries, New York: Oxford University Press [First in important three-part series on industrial users of computation technology]. Edwards, Paul N. (1996). The Closed World: Computers and the Politics of Discourse in Cold War America, Cambridge, MA: MIT Press [Unmatched study placing computer history in the context of broader international political developments] Edwards, Paul N. (2010). A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming [Highly insightful study of computer applications to climate science and the friction involved with the process of computer modeling]. Ensmenger, Nathan. (2010). The Computer Boys Take Over: Computers, Programmers, and the Politics of Technical Expertise, Cambridge, MA: MIT Press [The best book on the history of the programming profession and early IT professionalization]. Grier, David Alan (2005). When Computers Were Human, Princeton, NJ: Princeton University Press [By far best source on the history of human computers]. Haigh, Thomas. (2001). The Chromium-Plated Tabulator: Institutionalizing an Electronic Revolution, IEEE Annals of the History of Computing 23, [Important study on institutional adoption of mainframe computers in corporations during the second half the 1950s]. Haigh, Thomas. (2009). How Data Got Its Base: Information Storage Software in the 1950s and 1960s. IEEE Annals of the History of Computing 31, 6-25 [Seminal article on the origin of database systems]. Heide, Lars. (2009). Punched-Card Systems and the Early Information Explosion, , Baltimore, MD: Johns Hopkins University Press [Excellent examination of the history of punched card tabulation technology and the major players in this industry]. Hughes, Thomas, et al. (1999). Funding A Revolution: Government Support for Computing Research, Washington, D.C.: National Academy Press [Best overview of how government funding in the U.S. advanced computing, software, and networking] Kidder, Tracy. (1981). The Soul of a New Machine, New York: Avon Books [A fascinating and engagingly written examination of invention at mini-computing firm Data General] Lecuyer, Christophe. (2006). Making Silicon Valley: Innovation and the Growth of High Tech, , Cambridge: MIT Press [Excellent study on the pre-history and early history of Silicon Valley]. Mahoney, Michael S. The Histories of Computing(s). Interdisciplinary Science Reviews 30, [Fundamentally important historiographical article on computing that challenges assumptions of dominant narratives] Misa, Thomas J. (2004). Leonardo to the Internet: Technology & Culture from the Renaissance to the Present, Baltimore, MD: The Johns Hopkins University Press [Survey of themes on technology and culture over five centuries with latter part of the book focused on information technology and globalization]. Misa, Thomas J., ed. (2010). Gender Codes: Why Women Are Leaving Computing, Hoboken, NJ: Wiley, 2010 [Best source on gender history of computing and software]. Nebeker, Frederik. (1995). Calculating the Weather: Meteorology in the 20 th Century, San Diego, CA: Academic Press [Pioneering book on the history of computer applications to weather forecasting] Noble, David F. America by Design, Oxford: Oxford University Press, 1977 [Excellent study of computer automation of factory processes numerical control]