Maszyny Matematyczne,women,and computing: The birth of computers in the Polish communist era

Introduction

The detailed history of early computing in Poland has been described in English mainly by Marczyński (1980) and Łukaszewicz (1990); however, the two authors/pioneers focus on personal achievements and provide very few details about other parallel projects. ¹ In 2018, on the seventieth anniversary of Polish computing, the Polish Information Processing Society (PIPS – historical section of the Polskie Towarzystwo Informatyczne [PTI]) collected a number of testimonies from various government institutions, universities, associations, and companies in order to fill these gaps and provide a comprehensive description of the history of Polish computing. Out of that work came a volume that is the most comprehensive English reference we have available, although the articles it contains are always the result of the protagonists’ recollections and reconstructions. ² The other Polish-language resources, which will gradually be mentioned in the text, although more numerous, almost always focus on the 1940s and 1950s without giving an account of the consequences of these projects in the following decades; almost all of them were published between the 1980s and the 1990s and very few are contemporary with the events described. To reconstruct the crucial facts relating to the development of computers in Poland, we will mainly consult the sources published in the 1960s, which will be supplemented with the later testimonies of the protagonists of this history, paying particular attention to the contribution and role played by women. ³

Crucial to this research were the archival materials of the PTI historical section, and official authorization to access the collections of state archives and those of the Instytut Maszyn Matematycznych (IMM – Institute of Mathematical Machines) (the country’s most important research and development institute, for which most of the pioneers of Polish computer science worked), founded in 1962 on the already existing Instytut Matemetyczny in Warsaw. ⁴

A fundamental role in the history of Polish computing was played by the mathematician Kazimierz Kuratowski, who, in 1948, traveled to the United States for a lecture series and heard about the great benefits of electronic computers in applied mathematics. Kuratowski was actively involved in the revival of the country’s scientific activities, and he had strongly supported the idea that Poland should have similar machines. With these objectives, between 1949 and 1950, he had pushed for the establishment of the Grupy Aparatów Matematycznych (Group for Mathematical Apparatus – GAM) at the institute he himself directed. ⁵

In the immediate postwar period, Warsaw was an almost destroyed city. The telecommunications network no longer existed and around 65 percent of the companies providing these services had ceased production. ⁶ For these reasons, for over a year and a half, GAM did not even have a premises; the team began to recover, to retrieve tools, equipment, and materials, and to study in books and periodicals (from the West) how a digital computer could be designed. The group, excellently coordinated by the logician and philosopher Henryk Greniewski, included the engineers and mathematicians Krystyn Bochenek, Leon Łukasiewicz, and Romuald Marczyński (no women belonged to the team); but it was Zdzisław Pawlak, from outside the GAM, who in 1950 worked on the design and construction of the GAM-1, the first Polish experimental computer, which could complete an operation in one second but was limited to the commands of addition, complement, comparison, and selection on the numbers 0, 1, 2, and 3. ⁷

However, the team’s aim was always to design a large digital machine with the same computing power as the ENIAC (Electronic Numerical Integrator and Computer), the most impressive electronic machine in the United States, created in 1946, which was able to calculate in a few seconds what often took entire working days. It was something of a leap in the dark for Polish researchers, who, in a country devastated by war, did not have adequate equipment and materials available, let alone the experience needed to start the process of building such a complex computer. In the autumn of 1950, the Institute of Mathematics was allocated premises in the finally rebuilt building that had once housed the Warsaw Academy of Sciences. In the three rooms allocated to GAM, construction began on the Elektroniczna Maszyna Automatycznie Liczaca (EMAL – electronic automatic counting machine), Poland’s first digital electronic machine, initially assembled with components (electronic tubes) left behind by the German army, although they had a rather limited application. ⁸ The EMAL-1, launched in 1952, was modeled after the British EDSAC, but, as Adam Empacher claims, although the project was very innovative, it never got off the ground:

From a mathematical point of view, the EMAL-1 was to be even more complete than the British prototype, as division was also planned. [. . .] There were 22 commands, including different types of addition, subtraction, multiplication, and division. There were also special logical and conditional operations to facilitate mathematicians’ work on the programs. ⁹

In 1957, the group abandoned it for good, and the possibility of designing and building another digital computer with a different architecture was considered. The final break with the EMAL project came with the announcement of plans to build a new machine, which was given the initial name ABC.

In just under three years of work, the new device was ready. However, the team decided to change its name, this time choosing the letters of the alphabet XYZ, since this machine would be used by the army as a converter for artillery and missiles that referenced the three Cartesian coordinates X, Y, and Z for their trajectories (although Nowak claims that the later model of the ABC was given the last letters of the alphabet as a counterpart) (see Figure 1).^{10, 11}

Regarding the XYZ project, Leon Łukaszewicz reports:

Three years later our space extended, and a large group of young and talented computer enthusiasts joined us. I will name only a few of them: electronic engineers Zygmunt Sawicki, Zdzisław Pawlak, Jerzy Fiett, Wojciech Jaworski, and Jerzy Danda; mathematicians Antoni Mazurkiewicz, Tomasz Pietrzykowski, Jozef Winkowski, Jan Swianiewicz, and Krzysztof Moszynski. [. . .] The design of elementary electronic cells (flip-flops) of XYZ was taken from the Soviet BESM 6. They were “dynamic” cells whose description and then demonstration we were offered in Moscow in 1956. But the ultrasonic memory of XYZ, based on mercury tubes, was taken from EMAL. ¹²

From this quotation we can deduce some significant information: no women’s names appear in Łukaszewicz’s “group of young and talented computer enthusiasts,” which may mean that he had no memory of them or that he deliberately, and thus in a discriminatory manner, omitted to mention them. ¹³ It is also evident that in the mid-1950s there was cooperation between Eastern European states, particularly between Poland and the Soviet Union, in the development of computers, and that a completely original line of design was followed in Poland.

The production of XYZ computers initiated the construction of digital machines, in which the country also showed great interest. The communist authorities also made this reality known to a “non-sector” audience: a report dedicated to the work of the scientists was even broadcast on national television, and the most widely circulated popular science magazine in Poland, Młody Technik, in its 1958 issue 12, went so far as to claim that the machine could even calculate Sputnik orbits and direct rockets. ¹⁴

OPEN IN VIEWER

Figure 1.

The first Polish XYZ computer. On the desktop several rows of keys and two round oscilloscopes.

Around 1930, Thomas J. Watson Sr., U.S. President of the International Business Machines Corporation, had initiated a commercial policy aimed at integrating IBM machines into the European market. After Hitler’s seizure of political power in 1933, a strategic agreement between IBM and the Reich had been formalized, making Nazi Germany the most important buyer of IBM outside the United States. The American company, which had already acquired 90 percent of the Deutsche Hollerith Maschinnen Gesellschaft (Dehomag) in Berlin, took advantage of the terrible inflation that had hit Germany in the 1920s and undertook to supply punched-card machines specially prepared for the purpose of drawing up censuses and lists of the population in the occupied territories. ¹⁵

Thanks to this well-established IBM tradition in Eastern Europe and the tried-and-tested use of punch cards, the designers exploited this system to input data into XYZ machines. ¹⁶ In 1961, a special institution was also set up to strengthen the already existing GAM, the Zakład Aparatów Matematycznych – ZAM. This institute devoted itself entirely to research and experimental development in the field of digital machines and, after building a working model of the XYZ/ZAM-1 machine, moved on to the development of a small series production of digital machines of the ZAM-2 type, based on tube technology and intended for computing purposes. The Institute was able to produce eight examples of this type of machine in the years 1961–3. ¹⁷ The architecture of the ZAM-1 mirrored that of the IBM 701, launched by IBM in 1952, although the mercury acoustic memory, later replaced by a magnetorestrictive memory, was the result of an exclusive Polish project by Romuald Marczyński, Henryk Furman, Zygmunt Sawicki, and Jerzy Danda. As Adam Empacher points out, one might assume that Polish engineers had simply copied foreign designs. ¹⁸ That was not the case. “Reproducing” from the IBM 701 meant studying the architecture of the machine, the selection of operations, and the way numbers were represented, but as far as the technical specifications were concerned, completely original designs were undertaken. Moreover, the British and American projects were bound by military secrets, so it was impossible to know their technical specifications: because of Cold War tensions, Polish scientists and managers did not have any significant access to Western knowledge. ¹⁹ The ZAM models were exclusively Polish computers, as evidenced by the fact that in 1964 their designers and manufacturers were awarded the second degree of Prize of Sator, the highest award for scientific achievement in Poland, conferred for the originality of the models.

Thanks to the simplicity of use of the ZAM-1 computer and its functionalities, which allowed users to implement and solve complex numerical problems, in 1962 the IMM was born, a center that would have the purpose of designing and producing the subsequent models of the ZAM series. Only part of the Institute’s work was devoted to “technological” research of the components used in the machines (ferrite cores, drum memories); a much larger part was devoted to software design and the process of assembling the machines, including the associated input/output devices. Nine years after IMM was set up, the ZAM-21, a machine dedicated to numerical calculations, and the ZAM-41, the result of funding from a government commission set up to process large quantities of data, were completed. These were multiprogrammable machines that took advantage of the experience gained with their predecessors from 1962, the ZAM-2, and 1965, the ZAM-3. ²⁰ “The strength of the XYZ and ZAM machines was the software that was developed by A. Empacher, A. Mazurkiewicz, T. Pietrzykowski, J. Świaniewicz, A. Wakulicz, J. Winkowski and A. Wiśniewski. Initially, these were actions on paper as there was no physical machine and no experience had been gained.” ²¹ Calculations and programs required familiarity with the laborious operations with bits (0 and 1); having a programming language that could easily implement mathematical problems could speed up and simplify the work, so it was considered a priority to design and implement one.

Subsequently, the SAS symbolic language and the SAKO automatic coding system were defined to simplify the programming process. Due to the small working memory, the program was run in two stages, first translated into binary form on boards, then loaded into the machine and executed. The SAKO system, launched in 1960, was also referred to as Polish Fortran since the creators had actually modeled it on the Fortran. ²²

The automatic coding system SAKO (System Automatycznego Kodowania Operacji) was initially developed for the ZAM-2 machines as an algorithmic language suitable for implementing numerical problems. ²³ Thanks to its very simple structure and the option to write instructions through commands in Polish, the use of computers was also made possible for people outside the circle of its designers, and the ZAM-2 became a daily tool for many specialists and computer centers. All machines of the ZAM family were equipped with SAKO software, with the aim of not impeding programming even if a new model of the ZAM family was adopted. Developed between 1957 and 1960, this simple language was very popular among designers who wanted to learn programming quickly. ²⁴ As a result of these factors, SAKO became the object of widespread diffusion and long circulation, despite the appearance on the market of other numerical languages, such as ALGOL, coming from Western research centers, which were also beginning to take hold in the countries of Eastern Europe. ²⁵

Through SAKO, the machine code in the programming of XYZ and ZAM-2 machines was eliminated. Two working groups were responsible for defining the SAKO language and the SAS “macroassembler” (System Adresów Symbolicznych [Symbolic Address System]). The input of instructions was divided into two phases: the programmer formulated a command in SAKO, and SAS “translated” it into binary, the zero-one code that the machine could understand. ²⁶

Women’s employment in the immediate postwar period

In the writings signed by both Łukaszewicz and Marczyński, who were the main coordinators of the projects concerning the SAKO language and the ZAM family of machines, as well as in the above quotation from Madey and Syslo, no women’s names appear at all. In the acknowledgments of the English articles of 1960 and 1961, which represent broad and complete descriptive references of SAKO programming, Leon Łukaszewicz refers to his collaborators. In 1960 he writes:

The elaboration of SAKO was possible thanks to the efficient work of the whole staff of Automatic Coding Group of the Department of Programming of the Institute of Mathematical Machines of the Polish Academy of Sciences. Special investment while designing the system was given by A. MAZURKlEWICZ, J. SWIARlEWICZ, J. BOROWIEC, P. SZORC, and many others. ²⁷

In 1961 he writes:

The development of SAKO was possible thanks to the efficient work of the whole staff of the Automatic Coding Group of the Department of Programming of the Institute of Mathematical Machines of the Polish Academy of Sciences. Outstanding contributions were those made by A. Mazurkiewicz, J. Swianiewicz, J. Borowiec, M. Łącka, P. Szorc and many others. ²⁸

As can be seen, the formula used by Łukaszewicz is almost the same. The only obvious difference is that in the list of names of his collaborators, in the 1961 article, there is one more: M. Łącka. ²⁹ Łukaszewicz is referring to Maria Łącka, a woman active in the SAKO project. Why is this name added after a year? Was it due to the author’s forgetfulness or an editorial error in the previous version?

Also in the 1961 article by Mazurkiewicz, the name of the mathematician Łącka is included among those who had contributed to the writing of the SAKO language: “The author wishes to acknowledge the contributions of Dr. L. Łukaszewicz, J. Borowiec, M. Łącka, J. Swainiewicz and P. Szorc, who proposed several of the SAKO concepts described above.” ³⁰ Yet, in Biuletyn 2 (2017) of the Polish journal Polskie Towarzystwo Informatyczne (Polish Computer Society), Mazurkiewicz himself states, “at first, only Andrzej Wakulicz and Adam Empacher knew what a digital electronic machine was and what its programming consists of, later they also joined mathematicians working on analog machines (Józef Winkowski, Tomasz Pietrzykowski and I).” ³¹ No reference to Maria Łącka. Even Jan Borowiec, in his article on 1962’s SAKO programming, does not mention Maria Łącka, although, to be fair, he does present a technical examination of the characteristics of the language, so he does not mention any names. ³² Interviewed in 2020, Borowiec recalls that the SAKO group was mixed and that its composition changed over the years: “Yes, perhaps it was a bit noisier at the time because of these girls. Maybe I was annoyed by the chatter that was not always work-related. But I would never say that women are substantially inferior to men, because it’s simply not true,” Borowiec says. ³³

However, if the name of Maria Łącka is mentioned sporadically, the name of Jowita Koncewicz, who worked with her, is reported in the consulted sources only by Nowakowski and Bytnerowicz. ³⁴

Krzysztof Bytnerowicz, in the first pages of the book he edited in which he reconstructs the historical events relating to IT in Poland from 1948 to 2018, includes a long list of names of protagonists, apologizing for the possibility of having forgotten some of them. Among them is the name of Jowita Koncewicz, who is always linked to Maria Łącka, and therefore an active member of the same working groups. ³⁵ To these two names are added those of Danuta Kosecka, Iwona Messner, Krystyna Niemiec-Pomaska, Zofia Okrasińska, Ewa Orłowska, Janina Pawlikowska-Kozłowska, Elżbieta Roszkowska, Mariann Skupiński, Hanna Szymańska-Mysior, Ewa Zaborowska-Kardymowicz, Ewa Zawadzka, and Zofia Zjawin-Wińkowska.

Although these names are rarely mentioned in the literature and in the archives consulted, their contributions can be reconstructed from the direct testimonies of the protagonists and from the volumes in which they appear as authors. ³⁶

Jowita Koncewicz started collaborating with the IMM from the time when it was called GAM: “By necessity, only my mother raised me. It was she who showed me the value of work. She never pressured me to get married or have children, it was important to her that I had a profession and was independent.” She arrived at GAM immediately after finishing her mathematics studies in 1958. She was brought there by her husband, later a mathematics professor, Wiesław Żelazko. When she went to GAM, she already knew that they were building computers there and that she would work on their software: “It was interesting, nobody in Poland was doing anything like that.” At the beginning of her work, Jowita remembers her enthusiasm. They were sitting in front of the ZAM all night. “There were many discussions, rehearsals, we worked in completely abnormal times!” ³⁷

Computer technology was gradually starting down the path along which it would move from the purely theoretical stage to that of the production of devices for wide distribution. In 1961, the Komitet Ekonomiczny Rady Ministrów (Economic Committee of the Council of Ministers) decided to adopt Resolution 17 No. 400/61 of 11 December to ensure the production of digital electronic computers in the years from 1961 to 1965. The Resolution contained an action plan in the field of scientific research and concerned the mass production of computers at the Zakład Elektronicznych T-21 in Wrocław. The plan also provided for the establishment of university faculties specifically designed for this scientific sector, the training of company personnel, and the purchase of equipment (including in the West) for industrial plants and for the institutions involved in the project. ³⁸

In 1963, at the Instytut Informatyki (Institute of Computer Science – one of the oldest academic institutions active in Poland, whose origins can be traced back to the former Department of Telecommunications and Radio Construction of the then Faculty of Electronic Equipment Technology in Warsaw), a specialization course in Maszyny Matematyczne (mathematical machines) was established. It taught algorithmic logic, the principles of structured programming, and associated methods. ³⁹

In Poland, from the earliest approaches to computer science, the term of reference introduced by the designers of the XYZ, who were mathematicians and electrical engineers, was Maszyny Matematyczne. For the Poles, who were always careful to use the correct technical terminology, the introduction of the term “computer” was not immediate and was the result of a heated debate between language purists, some of whom wanted to use the terms liczułka or komputu (number, quantity), while others wanted to follow Adam Empacher’s proposal and opt for komput, used in reference to the number of soldiers per unit at the time of King Jan III Sobieski. ⁴⁰ In 1962, under the auspices of the Ministry of Finance, the study Terminologię techniki przetwarzania informacji (Terminology of Information Processing Techniques), signed by Wojciech Jaworski, was published, thanks to which, at the end of the sixties, it was decided to adopt “computer,” not as a term borrowed from English, but as a Polish neologism to indicate a device that works with a large amount of information. ⁴¹

The new course of study in Maszyny Matematyczne aroused great interest, especially among female students who had already obtained a degree in mathematics, but computers, like any wonder of technology, began to arouse curiosity even among ordinary people. In 1953, the journalist Olgierd Budrewicz had published an article in the popular magazine Przekrój (Figure 2) in which he described his visit to the Department of Mathematical Apparatus, giving an idea of what was to come and demystifying what for young Poles was the world of mathematical machines: a closed, sterile, and inaccessible world in which the priests of electronic calculation, wearing white aprons, were busy behind glass doors. ⁴²

OPEN IN VIEWER

Figure 2.

Magazine Przekrój 438, August 30, 1953.

After the end of the Second World War, there was a relaxation of gender roles in Poland by the People’s Republic, which manifested itself mainly in the desire to involve as many women as possible in the world of work. The communist authorities had sought to shape a new model of femininity, with the intention of changing the image of the woman guardian of the house: a tender wife and a caring mother. Equality between men and women in rights, but also in duties, was the basis from which the employment, social, and demographic policies pursued by the state began. More and more skilled workers were emerging, and a campaign was launched encouraging women to train by enrolling in university and specializing courses, so that access to those professions of responsibility and higher pay, which until then had been the exclusive prerogative of the male sex, could be increased. The times when family duties imprisoned women and did not allow them to develop their interests were long gone. Work had to be available for every one of them; special attention also had to be paid to those women who had small children to care for. ⁴³ The number of kindergartens and nursery schools was greatly increased in industrial centers, workers’ settlements, and villages; they remained open at times that allowed mothers to devote themselves exclusively to work, from four in the morning to ten at night, all week long. This initiative not only facilitated the entry of women into the country’s workforce but was also beneficial for the authorities, who took control of children and adolescents who until then had been cared for by their mothers at home, starting a program that influenced citizenship according to the canons envisaged by communism from their first years of life. ⁴⁴

In truth, as Natalia Jarska of the Instytutu Pamięci Narodowej (Institute of National Remembrance) argues, this tendency wasn’t just the result of the manifestation of communist ideology; the postwar conditions of poverty, often linked to the death of the heads of families, had already “forced” women to find a job. However, if the state had proclaimed equal rights for women and men, going so far as to ignore their biological differences and point out that the times when family responsibilities imprisoned women were long gone, the analysis of professional and social hierarchies returns a different reality: women workers rarely occupied top positions and earned less than their male colleagues (who often doubted their professional abilities), even though they had higher production standards. ⁴⁵ The real intention of the authorities was therefore not to look after women’s welfare but to provide the expanding postwar economy with additional workers. Evidence of this is the fact that, from the mid-seventies, following the economic crisis, women were the first to be sent home. ⁴⁶

When, in the 1960s, the first projects aimed at getting the Polish computer industry off the ground were launched, many women, mostly graduates in mathematics and then specialists in Maszyny Matematyczne, took part in these research groups, reflecting the national trend whereby they were employed in all areas of work. Among them was Alicja Kuberska, who began her studies at the Technical University of Wrocław. Unlike other computer scientists of this generation, she “chose to specialize in ‘digital machines’ because I knew that computers were the future.” In 1962, even before graduating, she began working as a research fellow at Elwro. ⁴⁷ A student of Thanasis Kamburelis, under his supervision she completed her dissertation entitled Organizacja logiczna małej maszyny matematycznej (Logical Organization of a Small Mathematical Machine), in which she examined in detail the construction and operating principles of one of Elwro’s first ODRA series computers. “We had to design the so-called ‘logic’, that is the computer’s circuit diagram,” reports Alicja, evidence that she knew the structure of the machine in detail. ⁴⁸ At Elwro she also assembled a small machine called KUMA (Polish acronym for “Kuberska’s machine”), which was then used by generations of young workers to learn the trade. Later, Alicja created a series of programs that allowed her to check whether the defects in the ODRA machine were mechanical in nature or whether the software had not been well implemented. ⁴⁹

In 1986, the Qualification Team of the Wrocław Section, chaired by Prof. Zdzisław Karkowski, awarded Alicja Kuberska the degree of professional specialization in Computer Science – Digital Systems Engineering. According to the grounds for the award, among the work carried out by Alicja Kuberska completely independently, special attention should be paid to the “Koordynatora kanałów” control system of the operating memory and channels of external devices in the ODRA computers, which, according to Kamburelis, “was an original, ingenious solution, allowing a high working autonomy, which was also applied later in subsequent machines.” ⁵⁰

ELWRO factories and the end of the Polish dream

Computer science in Poland certainly followed a different path than that of Western countries; this can be seen from the Planu 5-letniego (5-year plan), the five-year report of July 18, 1964, in which the IMM’s planned activity had an almost exclusively scientific connotation; the task of assigning machines and expertise to specific industries and services, by the state planning agency of products, was instead delegated to a parallel industry, while computer science itself was fundamentally linked to research activities on new techniques for the implementation, organization, and construction of programming systems, sometimes initiated in collaboration with other national and international scientific and research institutes. ⁵¹

Although Warsaw had been the center of conceptual and productive activities, the industry linked to this sector began in Wrocław. At the end of the 1950s, the first Elwro factories were established, and staff were trained at IMM:

During the first months of their stay, ELWRO employees worked at IMM TK, where they became familiar with the design and actively participated in the development of documentation for specific machine units. Subsequently, these employees were directed to the IMM laboratories, where they became familiar with particular modules and, under training, prepared for the launch of the ZAM-21 machine modules, first at IMM and then at ELWRO. ⁵²

With the support of the provincial and municipal authorities, on February 6, 1959, the Minister of Heavy Industry signed a decree establishing T-21 Wrocław’s Electronic Works, later nicknamed Elwro after an abbreviation of its name. It was here that series production of the ZAM family of models began, building on the experience gained in Warsaw on the eight prototypes assembled, equipped with the SAKO automatic coding system, with technical documentation prepared specifically for transferring expertise to industry. ⁵³

The first transistor computer, called the S1, which was built at the IMM between 1959 and 1960, was also mass-produced by Elwro, which marketed it as ODRA 1001. The name of these machines was inspired by the Odra River, which crosses Wrocław, and the number 1000, a reference to the first model, symbolized the approach of the 1000th anniversary of the foundation of the Polish state. It was followed by the other models of the same family, the ODRA 1002 and the ODRA 1003, distributed in 1964; they were binary computers that allowed the performance of several hundred operations per second and were a resounding success, as witnessed by the forty-two models produced. The ODRA models (Figure 3) started the process of computerizing the state railways and institutions such as the Central Statistical Office (GUS – Główny Urząd Statystyczny) and universities. ⁵⁴ The machine belonging to the last generation produced, the ODRA 1305, distributed in the seventies, remained in operation until 2010 and could be programmed with the most popular programming languages in that period, including COBOL, ALGOL, and FORTRAN. ⁵⁵

OPEN IN VIEWER

Figure 3.

ODRA 1305. Third-generation Polish computer of the ODRA series, mass-produced since 1973 at Elwro Electronic Works in Wrocław.

In 1971, Leon Łukaszewicz, highlighting the elements in common between the two programming languages, described SAKO as “a fairly large subset of the COBOL language.” ⁵⁶ Since SAKO’s life was quite short (the computers of the generation following those for which it was designed were technologically more powerful), it was decided to replace it with modified versions of ALGOL and COBOL. Two working groups were therefore set up to adapt these languages to Polish machines: Jowita Koncewicz and Maria Łącka worked with ALGOL, while Krystyna Niemiec-Pomaska, Zofia Okrasińska, and Janina Pawlikowska-Kozłowska not only dedicated themselves to COBOL but also drafted a sort of user manual. ⁵⁷ In the volume ZAM-41. Język programowania COBOL (manual of the COBOL programming language for the ZAM-41 machine), published by them in 1972, reference is made to the COBOL translator, which made it possible to use Polish terms directly instead of the corresponding English commands. ⁵⁸ This solution enabled the language to be widely disseminated, and it became easy to use even among nonprofessional users. ⁵⁹ The writing of this manual is tangible proof of the direct and in-depth knowledge of the three authors, not only of the SAKO and COBOL languages but also of the architecture and components of the ZAM-41.

When it was decided to adapt ALGOL to the ZAM machines, Jowita Koncewicz was worried about defining an SAS-like translator because the original American one was too expensive. The working group of the Polish ALGOL compiler had the idea of defining ALGAMS, a kind of Esperanto for programming that was to be used by all the countries of the East. It was not very successful, but this project was presented at various conferences and meetings also attended by Bulgarians, Russians, Hungarians, and Czechs. ⁶⁰

If ODRA computers were so reliable and so widespread, why didn’t Elwro become the leading computer distribution company in Eastern European countries?

In January 1949, the USSR, Bulgaria, Czechoslovakia, Hungary, Poland, and Romania founded the Council for Mutual Economic Assistance, better known as CMEA or Comecon, “established on the basis of the sovereign equality of all the member countries of the Council,” with the aim of promoting the development of the national economy and the acceleration of economic and technical progress in the member states. ⁶¹ To improve the state of general computing in the USSR, where there was a need for high-performance computers for strategic nuclear and missile programs that required processing huge amounts of data, thought was given to the creation of computers with adequate amounts of primary memory, a suitable assortment of peripherals, and systems and applications software, to enable people with little technical training to use computers effectively. The Soviets made the decision to start work on a new family of general-purpose computers for data processing as soon as the success of the IBM System/360 machines produced in the West was evident. Their first official statement, of December 1967, concerning the Unified System of Electronic Computers, commonly called ES or Rya, implied that it was an exclusive Soviet project. ⁶² However, by 1968 the USSR was hard at work trying to persuade its Comecon allies to join the effort. Hungary, Bulgaria, and the German Democratic Republic were the most amenable. “Poland wanted to continue its ODRA program. Czechoslovakia also had a program of its own and proved to be less than wholeheartedly committed to the Unified System. Romania remained especially obstinate, choosing to look to the West.” ⁶³

Since the 1950s, the Soviet Union had sought to promote cooperation in the field of computer technology as a means of increasing technological integration within Comecon. Cooperation was desirable as a means of solidifying economic and military ties through technical interdependence. The computer industries of the German Democratic Republic (GDR), Hungary, Poland, and Czechoslovakia were much smaller than that of the USSR, but in some ways they were more sophisticated. Because they had been more in contact with the Western computer community, their experience proved to be a valuable asset for the Ryad project. They also had more advanced capabilities in some aspects of peripheral technology and software development. The decision on the basic architecture of the new system was made only after some discussion both within the USSR and among the Comecon partners. The GDR agreed to use the IBM System/360 architecture and to make the Ryads compatible with the International Business Machines computers. The East Germans were already pursuing this approach on their own – probably aided by either direct or indirect access to IBM and other U.S. technology in Western Europe: they had already reproduced the IBM 1401 in the Robotron 300 series and built the Robotron R-21 based on the 360 design. From 1968 to 1969, they were developing the Robotron R-40 to be fully 360-compatible. ⁶⁴ For this reason, plus the availability of vast amounts of IBM software and favorable experiences with their imported products in Eastern Europe, the IBM 360’s architecture was finally adopted. ⁶⁵

The potential for problems in this arrangement was enormous. There were language barriers, the difficulty of trying to duplicate sophisticated foreign technology, poor telecommunications and long physical distances, weak support industries, assorted international bad feelings, and an untested control structure that supervised many development and production facilities that had never worked together before. ⁶⁶ It must be said, however, that their resources were considerable, viz. an economic system with the ability to focus its resources on priority projects (e.g., military and space programs), one of the most powerful mathematical communities in the world, and, more importantly, the size of the market that made this project feasible in economic terms. ⁶⁷ Of all the participating countries, 20,000 computer scientists and 300,000 other employees were co-opted for this project. Each country had its own chief designer in contact with the Soviet chief designer, and detailed tasks were agreed upon in councils of specialists and working groups. According to Tomasz Kulisiewicz, the machines of the Ryad series were designed on the basis of information and materials obtained using KGB intelligence methods. ⁶⁸ This is also confirmed by a CIA report from 1973, which specifies that “many IBM 360 machines could be exported to Eastern Europe, making them accessible for Soviet purposes; the components with which the Ryad machines were built were purchased, legally and illegally, from American, Western European and Japanese companies.” ⁶⁹

As Jong-Tsong Chiang claims, no consensual evaluation could be found about this program. This project is often cited as an example of fruitful fraternal cooperation in the Comecon by the USSR: for them, it was certainly a valuable means of speeding up resources in a strategic area that was seriously delayed. Comments from other participant countries are ambivalent, ranging from praise to complaint. ⁷⁰

The total lack of coordination within Comecon led to serious delays in the production of Ryad computers. Expenses were almost three times higher than those of IBM, and technological backwardness combined with poor organization meant that the number of machines able to function was much lower than the somewhat optimistic expectation. The USSR had counted on producing 3,000 to 5,000 Ryads per year by 1975, but only a few hundred were produced. In the report released in 1977, the CIA pointed to serious delays in the development, production, installation, and effective use of Ryad computers, which were to be the spearhead of the Kremlin’s computer modernization program:

In the Ninth Five-Year Plan, 1971-75, the USSR and its East European allies produced only 10-15 percent of the anticipated number of RYAD computers, a series of third generation computers modeled on the IBM-360 series. Furthermore, output has included only the smaller, less powerful RYAD models, with the final product decidedly inferior to the IBM originals in reliability and compatibility and in the quality of associated input-output and auxiliary storage devices. ⁷¹

Taking into account these difficulties, or at least in anticipation of inevitable delays in the ES program, the Soviets and their partners concentrated mainly on specialized military projects and computers for scientific applications. Several Eastern European programs, including the advanced Polish ODRA project; that of the Hungarians, who had purchased a license for the French CII (Compagnie Internationale pour l’Informatique) machine; that of the Bulgarians, who had purchased a Japanese one; and that of the Czechs, who were producing their own model called Teslê, would also continue. ⁷²

Elwro was given the task of producing the new R-30 machines (Ryad-30), following a project defined in Armenia. As Jerzy Nowak argues, this imposition met with strong resistance from the Elwro designers, who, having already gained experience in this field, instead embarked on more avant-garde experimentation, arriving at the production of the R-32 model, made in Poland and four times smaller than its IBM counterpart. The previous R-30 built by the Polish team was contained in just one cabinet (while the Soviet one occupied three), did not use much energy, and was more reliable. Above all, the Polish version of the R-30 was significantly faster. This Polish position had political repercussions that resulted in the ban on Comecon countries importing computers from Poland, although, according to some Russian testimonies, the later view was that it was a mistake not to take into account the direction indicated by the Elwro designers. It was the beginning of the end: a process started that coincided with the definitive cessation of production of computers designed in Poland and distributed by Elwro. ⁷³

In 1974, during a Comecon session in Prague, all the Ryad devices were compared using a test prepared by the Czechoslovak Academy of Sciences, which measured the time it took to calculate a set of one million basic arithmetic operations. The R-20, developed in Bulgaria and Minsk, dealt with this test in 200 seconds; the R-30 from Yerevan needed 70 seconds; and the Elwro computer only 7 seconds. Another surprising result was the German R-40 machine, which was five times bigger but ran the test in 9 seconds.

After the collapse of the Soviet Union, the Ryad project was finally abandoned and the cooperation between the socialistic countries in the field of information processing was dominated by economics, not by science. ⁷⁴

The work that had been carried out in Polish companies and universities had resulted in great skills and valid projects, lasting until the seventies. However, the negative and unsuccessful experience of the Poles as members of Comecon had dampened the spirits and enthusiasm of an entire generation of young computer scientists. First rewarded and gratified, they were then forced to abandon their original ideas and spend every resource on emulating projects developed by other research centers. ⁷⁵

The National Bureau of Informatics: the Polish program for planned IT development

With Resolution No.34, the Polish government established, in 1971, the National Bureau of Informatics (Krajowe Biuro Informatyki – KBI), subordinate to the Ministry of Science, Technology and Higher Education. The control of projects related to the development of information technology was thus entrusted to a minister, a member of the government who, as soon as he took office, placed the IMM’s activities in the general framework of an industrial organization. Building on the positive experience gained in the mid-1960s when, following an agreement signed between Warsaw Polytechnic and Elwro, production of the successful UMC-1 (Universal Digital Machine) was started, in the “Program for the Development of Informatics” for the period 1971–5 the research projects were mainly aimed at promoting the development of industries. Applications began to be given more importance than hardware, and computer science became a sector on a par with construction, transportation, and farming. ⁷⁶

The construction of the UMC-1, already underway at the Department of Telecommunications Design and Radio Engineering of the Warsaw Polytechnic, was coordinated by engineer Ruta Barbara Maćkowiak, who, from Elwro’s foundation until 1973, was the designer of some of the models produced there. Barbara also supervised the construction and production of components for the ODRA and Ryad computers. ⁷⁷ She enrolled at the Wrocław Polytechnic in 1957, and by 1959 she was employed by Elwro, where she also coordinated the unit that supervised the construction and manufacture of the electronic instruments necessary for measuring the efficiency parameters of hardware components, both at the assembly stage and during final quality control. These were automated technological testers that were often purchased abroad in order to maintain the high-quality level of the manufactured products. ⁷⁸ Because she had a very good command of English, her managerial position involved managing contacts with the company’s foreign partners. There was a need for components from Great Britain, and for this reason the thirty-year-old Polish woman was allowed to visit exhibitions and factories in the West in order to acquire specialist knowledge on these subjects, at a time when no one else in communist Poland was permitted to. Given her acquaintances abroad, she was accused of espionage and very frequently subjected to interrogations that she herself calls “humiliating.” In her contribution on the UMC-1, Barbara Maćkowiak mentions that Maria Łącka and Teresa Pajkowska also participated in the team. ⁷⁹

In Warsaw, the ERA factory, founded in 1927 under the name Polskie Zakłady Elektrotechniczne Spółka Akcyjnaha, played a leading role in the design and production of the K-202 and MERA minicomputers and their operating and application software. ⁸⁰ The director of ERA’s Development and Production Division, Jacek Rafał Karpiński, transferred the experimental section (Zakład Doświadczalny Minikomputerów [ZDM] – Experimental Department of Minicomputers) to the IMM and started the design of the K-202 model, “thanks to the recruitment of some excellent specialists.” Karpiński argued that Poland’s specialization in the production of RIAD 30 was disadvantageous because the machine was too expensive; he strongly argued that the K-202 minicomputer should be produced instead. Karpiński had designed the K-202 between 1970 and 1973, based on chips with low and medium scale integration; the processor was equipped with expandable storage and ROM. The project involved, among others, Elżbieta Jezierska-Ziemkiewicz and Andrzej Ziemkiewicz (Elżbieta’s husband), who were particularly involved in the logical structure of the machine. ⁸¹ Despite the initial interest by the authorities, it was rejected by the communist administration. Its production was, nevertheless, finally started, but it was financed by English Data-Loop and MB Metals. ⁸²

Elżbieta Jezierska, enrolled at the Warsaw Polytechnic, began studying computer science following a third-year course on Maszyny Matematyczne, on which she also wrote her dissertation, which she completed in 1965. In contrast to the many female students who specialized in computer science after graduating with a degree in mathematics, there were very few female students at the polytechnic and studying computer science meant “getting your hands dirty,” participating in the construction of calculating machines, and knowing their individual parts and their connections. In 1964, Elżbieta had started working for the IMM, devoting herself to the design of the processor of the ZAM-41 and taking care of the arithmometer (the part of the processor that oversaw the four arithmetic operations in binary). At the end of the ZAM project, which coincided with the start of cooperation between IMM and Elwro, she had also contributed to the design of the ODRA-1305 processor and the decimal arithmometer for the R-30 machine processor. ⁸³

In her contribution “Komputery 16-bitowe,” in which she describes the design of the K-202, Elżbieta specifies:

With Dr. Teresa Pajkowska and Dr. Karol Doktor, we perfected the set of processor registers, the list of commands and the principles of interaction between the processor and the peripherals. With Andrzej Karczmarewicz we designed the memory interface, and with Janusz Krzyżanowski and Jerzy Zawisza we defined the set of supported characters. [. . .] The decisions we made had to be discussed with Jacek Karpiński, but, even though they were sometimes the result of stormy discussions, we always managed to let common sense prevail. ⁸⁴

The K-202 minicomputer was to have a flexible structure, small size, and ease of expansion to other peripherals. Thanks to an agreement with the English company MB Metals, the Poles had the availability of the components produced in the West, thus overcoming the problems caused by the obsolescence of elements available in Poland. ⁸⁵ Beginning in the spring of 1971, the team worked tirelessly on the development of the prototype processor and was able to present the K-202 at the London Technology Exhibition the following summer, where it gained the admiration of all competitors. In the chapter “Rodzina maszyn K-202 / Mera-400 / MX-16” of the book Polska Informatyka: wizje i trudne początki, published on the occasion of the seventieth anniversary of Polish Informatics, Andrzej and Elżbieta Ziemkiewicz specify that, following some disagreements he had with directors and professors of the IMM, Jacek Karpiński was dismissed in March 1973. The task of coordinating activities for the successor of the K-202, called MERA 400, was therefore entrusted to Elżbieta. “MERA was created with the idea of improving everything that eventually needed to be corrected in K-202 and producing it entirely in Poland.” However, they had moved away from the concept of a microcomputer as they had managed to fit all the components into one “box” (previously, the various elements, such as disks, were connected from the outside and so the whole system took up a lot of space), making MERA a true Personal Computer. ⁸⁶ They also report that:

For 40 years, the K-202 machine has caused a stir in the media and on the Internet. Various authors have written about it and opinions vary widely. Some believe that it was “10 years ahead of [it’s time]”, while others say that other machines were just as good, just as fast, and just as powerful. ⁸⁷

Andrzej and Elżbieta Ziemkiewicz probably refer to the models produced in Hungary and the GDR, which had a greater distribution than the Polish models. ⁸⁸ Hungary’s contribution to the first generation of Ryad computers was the model produced by the Videoton Factory under the code names EC-1010, VT-1010, or, more colloquially, R-10. It was introduced to the public in 1972 at the Budapest International Fair (Figure 4), and by late 1973 was one of the first Ryad machines to be distributed. In the Hungarian computer industry, the preference was for small machines on foreign licenses. Within the Ryad unified system, R-5 was based on the French CII-10010 microcomputer license, R-8 on CII-Mitra 15, and R-10 on TPA 70 (close to PDP-11). The first microcomputer was created in the industrial VIDEOTON center; the second one was initially produced in a Hungarian Academy of Sciences center and then in the Orion-Radio Corporation. In two other workshops, punched-card devices and data transmission machines were produced. Hungarian developers operated quite independently, not getting too enmeshed with the main Ryad project and continuing their relatively strong ties with the West European computer community. ⁸⁹

OPEN IN VIEWER

Figure 4.

The K-202 in the 1971 fair flyer.

In the GDR, the first model ES-1040 (R-1040) was produced by the VEB Kombinat Robotron, extending up to the KRS 4021 of December 1973. This line of computers has impressed Western observers as the best of the Soviet Bloc’s computer firms. Robotron research, development, production, and training facilities were in several areas in the GDR, with headquarters in Dresden. Much of the 1040’s production was not yet automated, but the quality of workmanship was high. Production could have been as high as 80 to 100 units per year, but was held back by customer support capabilities. The management and engineers of Robotron, as well as the support to the assistance and the formation, had no rivals within the Comecon. ⁹⁰

Conclusion

In 1977, Tomasz Strzyżewski, an employee of the Cracow branch of the Office for the Control of the Press, Publications and Entertainment, had secretly transported to Sweden over 700 documents that had been censored by the Polish People’s Republic. Among these, there were some which referred to the computer industry and which clearly showed that, during the communist period, all information concerning computerization was subject to strict censorship.

In studies, notes, etc. on the electronics industry in Poland, no authorization should be given to publish: a) materials criticizing the program for the development of electronics and suggesting that this program should be based not on cooperation with the USSR but with the capitalist countries; b) messages containing detailed technological data – fourth generation microcomputers, semiconductors, quantum electronics (laser type); c) information on imports of components used in microcomputers. ⁹¹

Poland, as we have seen, spent most of this period trying to catch up with the West in the research and production of high-performance automatic machines, although its centrally planned economic management did not allow it to do so. At the beginning of the 1960s, the computer industry was nonexistent, and one could only speak of a few scientific experiments associated with mathematics. Most computer scientists had initially been interested in mathematics but had shown their interest in this discipline by embarking on the “new” profession – albeit somewhat in the dark. For women, the path was the same, although, as has been shown, references to their work are not always so explicit. In most cases, their experiences and contributions are not well specified in the literature; their names can be deduced from the long lists of those who originated this science and transformed it from a simple passion to a productive activity. Indeed, one might think that their collaborations were subject to stricter controls or that the documents containing their names were bound by censorship. However, it should be borne in mind that at that time there was no “myth” surrounding this profession, and the question of whether it was a profession more suited to men or women was apparently not even considered. It was an absolute novelty. An unknown terrain was being explored, in which, thanks to the will and foresight of each protagonist, original and innovative results would be achieved.

And if in the United States the first computers were commissioned for military purposes, in Poland they were the result of the work of a group of enthusiasts, certainly not stimulated by dizzying earnings. As Kuratowski relates, “such low salaries were a hard but necessary test of whether scientific work was really a vocation for young people.” ⁹² It should be remembered, however, that after the war the Polish economy was being rebuilt and the narrative of sacrificial work for the homeland dominated the public space. Hardly anyone made much money. Constant and frustrating negotiations with the state bureaucracy and the scarce resources of a collapsing economy were the order of the day for a group of professionals aiming for an innovative technological infrastructure.

As long as computers were assembled in only a few models, they remained only a scientific curiosity. When their mass production began, salaries went up and the economic “benefits” of this work began to be felt. The first secretary of the Polish Communist Party, Edward Gierek, pursued a policy of modernization based mainly on the Polish electronics industry. ⁹³ Only following the decisions of the Comecon were spirits quashed and inventiveness and initiatives definitively stopped. In the 1980s (also following the failure of the Ryad project), the national IT industry collapsed and it became more profitable to buy computers from Western markets, produced by companies that could develop under the conditions of a free-market economy. ⁹⁴

From the Polish studies under examination, it is evident that some very important issues linked to the communist period were not explicitly dealt with. There is no mention of what the goal was of building an “all-Polish” computer, and there is no trace of what the potential economic benefits or losses were following the failure of the project, of the degree of Soviet influence in Poland (which must have been preponderant anyway given the decision to abandon research), or of any camaraderie between scientists from different Eastern countries, despite ideological barriers. But, above all, as has been highlighted in this paper, the names of the women who played an important role in this story rarely appear. As Hołyński rightly points out: “Unfortunately, none of the authors of these writings are historians by trade. Therefore, we don’t feel qualified to discuss the broader historical and political context of this period.” ⁹⁵

Also in the Polish case, “the idea that the history of computing is a defeminized realm is belied by multiple, important historical examples.” And if in Great Britain and the United States computer science was born as a purely female job, then redesigned as a field of male engagement, in a war-torn Poland that felt the extreme need for workers to rebuild the country, the computer industry did not hesitate to include women as well. Women who came from families in which mothers alone had raised all the children, and whose fathers and older brothers had died or gone missing during the war, quickly realized that studying mathematics and, even more so, taking an interest in Maszyny Matematyczne would guarantee a minimum professional and financial independence. ⁹⁶

The trend reversals that took place in Western countries, and which focused on the many prejudices and processes of discrimination that influenced the world of work in the field of technology, were not so evident in the People’s Republic of Poland. The presence of women was never considered strange or exceptional; it was an ordinary occupation in which both men and women were involved. But when Polish computers were no longer produced, this caused the disappearance of their names from history.

In the late 1970s, Polish specialists focused mainly on developing software for Western computers. Of the few women we know of, some have dealt with the translations into Polish of specialized books on programming languages such as FORTRAN, C, or Operating Systems; others were involved with the development of next generation languages, such as Loglan ’82, a language including all the programming tools used in object-oriented programming.^97,98

The Polish IT industry has completely disappeared from the international picture, although, as we have seen, in the immediate postwar period, considerable efforts were made by mathematicians and engineers to put the country at the forefront of IT. The studies and experiments undertaken in communist Poland have undoubtedly left their mark: there is evidence of this in the numerous documents taken into consideration that have made it possible to trace the path of technological evolution historically, always with reference to the policies pursued by the communist ideology that strongly encouraged women too to become, like their male colleagues, experts in computer programming and design.