Svyatoslav Sergeevich Lavrov

Home → Hall of Fame → Svyatoslav Sergeevich Lavrov

Text: Igor. P. Agamirsjan.
Edition and translation Alexander Nitussov

Svyatoslav Sergeevich Lavrov was born on the 12 of March 1923 in Leningrad into well-educated family. His father, basically a navy engineer, after the civil war in Russia was in charge of power supply at number of industrial enterprises. His mother graduated from so called “Higher courses for ladies” (before the revolution of 1917) and worked for some time as assistant of a school principal. Later she was a librarian.

Svyatoslav Sergeevich Lavrov

As S.S . Lavrov studied at school he developed a serious interest in mathematics and began regular visiting of famous mathematical working group at the Leningrad Palace of Pioneers (city's cultural and educational centre for children which they could visit at free time and holidays). He was even one of the winners of the Leningrad “Mathematical Olympic competition” in 1939.

After graduation from the secondary school he entered the faculty of mechanics and mathematics at the Leningrad State University . At the university he attended lectures on algebra by D.K. Faddeev and on mathematical analysis by G.M. Fikhtengolts (Fichtenholz) what gave him first impressions on computation methods.

In 1941 with the outbreak of the war he volunteered to army (militia of Leningrad ) but, together with big group of other students, he was dispatched to study at the Leningrad air-force academy. He graduated in 1944 in rank of lieutenant-technician and was appointed into an aircraft regiment as a fighter squadron maintenance technician. Two years later S.P. Korolev (creator of the first Soviet rockets and space ships) invited him to join studying of trophy machinery and he was enlisted in a special group at the Ministry of Defence (later they worked at the State central testing field).

S.P. Korolev also helped Lavrov to be quickly demobilised and appointed in his own design bureau as chief of the ballistics group, in 1947. First that was a department by bigger design bureau of a research institute of the Ministry of Defence but several years later it already turned into a special independent enterprise.

At the group of ballistics S .S. Lavrov dealt with mechanics of the bodies with variable mass, equations of a rocket motion, he did ballistic calculations, precise calculations of trajectories and defined points of landing, estimated parameters of dispersion as well as necessary volumes of fuel, developed theory of instruments and systems of remote control. He also took part in all testing flights of all Korolev's rockets, beginning from the trophy V-2 (A-4) in 1947 and including intercontinental R- 7 in 1957. In 1950 he prepared a monograph on rocket ballistics in form of technical report. In 1956 it was published by the Artillery Academy ( Moscow ) as a course book but only for internal use. However, some ten years later, in 1966, it was republished already for broad public usage[1]. In 1954 S.S. Lavrov graduated form faculty of mechanics and mathematics of the Moscow State University - MSU, where he studied by correspondence and in 1959 was awarded a doctorate (doctor of technical sciences) for complex of scientific achievements, without submitting of thesis. In 1963 he received professorship in speciality “Mechanics”. In July 1966 he was elected a corresponding member of the USSR Academy of Sciences, department of mechanics and processes of control (in speciality “Automatic control”).

Until 1966 S.S. Lavrov worked as chief of department at the Special Design Bureau -1 (SDB-1) headed by S.P. Korolev himself. Since 1959 he was also lecturing at the faculty of mechanics and mathematics of the MSU. In 1963 he became a professor at that faculty and in 1970, after establishing of the faculty of computational mathematics and cybernetics he held there the chair of programming languages.

Already by 1960 his scientific interests focused on programming, primarily on computer software development – first libraries of standard subroutines, later high-level languages translators. In the second half of the 1950-s all ballistic calculations at the SDB-1 were performed with computer BESM (one of the first Soviet computers) and already then the problem of programming simplification and automation was rather urgent. S.S. Lavrov created for the BESM one of the world's first languages of symbolic coding. Later it was published under the title “Agreement on programs recording in symbolic notations”[2]. However, then language of symbolic coding was used largely as a tool for programming process formalisation and translation into machine codes was performed manually.

With production of serial computer M- 20 in 1960, one of the first four machines from that release was placed in SDB-1. Almost at the same time an “Announcement about algorithmic language Algol- 60” was published, what made practical realisation of automatic translator for high-level languages quite real. It was also interesting from the scientific point of view. The priority directions of further work were determined and creation of such translators was started by the collaborators of IAM – M.R. Shura-Bura and E.Z. Lyubimskiy and also by A.P. Ershov of the Institute of mathematics (Siberian department AS USSR). A little later S.S. Lavrov initiated project on creating translator from Algol-60. The project was performed under guidance of V.A. Stepanov. Essential part of researches was conducted by V.N. Popov. S.S. Lavrov was the project's ideological leader. His main work at that time consisted in development of standard subroutines for numerical methods, such as non-algebraic equations and their systems or defining extremum of functions with multiple variables. Here he implemented new standard IS-2 what enabled usage of those subroutines both in manual programming and their integration into new translators for M-20.

Creators of translator SDB-1, first of all S.S. Lavrov precisely estimated possibilities of M-20 and recourses of his own team. Efficiently realised sub-array of Algol-60 which did not include recursive procedure and any functions was chosen. It also didn't include actual parameters-expressions during call of parameters by their names and also variables with indexes as parameters of cycles and native arrays. Notable that all adopted limitations for input language, except recursion, perfectly coincided with subsequent development of programming languages. Due to those input language limitations translator TA-1 demonstrated its operability earlier than similar parallel researches obtained some practical results. Already in spring of 1962 its first release was widely in operation.

Among all Soviet computers of the second generation it was translator TA-1 which was in the highest demand. That could be explained with simplicity of its input language sub-array, high efficiency of compilation and acceptable characteristics of generated code efficiency. Later release of its essentially extended version named TA-1M (“Signal”) devised by S.S. Lavrov and developed in details by a team of V.A. Stepanov notably promoted implementation of TA-1 “family”. That was the USSR first and, most probably, one of the world's first translators that enabled program checkout in terms of input language.

Simultaneously with development of TA-1 S.S. Lavrov was constantly cooperating with A.P. Ershov, jointly working on theoretical fundamentals of memory distribution optimisation. In 1961 he proposed formalised description of the memory distribution problems (“Lavrov schemata”), also equivalency of the problems of memory saving and colouring of graphs (graph of function domains incompatibility) was proved. Final variant of heuristic algorithm of memory distribution and re-usage was described by A.P. Ershov and found subsequent practical realisation in his “Alpha translator”.

Beginning from 1963 S.S. Lavrov turned to symbolic information processing languages. Interest to that subject had grown out from design automation problems and also from awareness of the fact that each concrete problem domain needs specific data type and sets of operations. His work resulted in joint work of his team, “Complex of geometric descriptions for automation of engineering design”, by Begunkov G.S., Lavrov S.S., Pedanov I.E., Stepanov V.A. It was published in collection of works, “Computing machinery in engineering”, Minsk, Institute of technical cybernetics of AS Byelorussia, 1966.

In 1966, after death of S.P. Korolev S.S. Lavrov was appointed as a chief of department “Mathematical maintenance of computers” at the Computation centre of AS USSR. It was there that he performed many works on languages for symbol processing of information. There he used BESM-6, most modern computer of that time. Some of those languages were, “ Snobol - A ”, et al . with L .I. Goncharova in 1968, “Lisp”, et al. with G.S. Silagadze, in 1970, etc.

In the second half of the 1960-s and in the beginning of the 1970-s S.S. Lavrov actively participated in Working group on ALGOL – WG-2.1 of IFIP (International federation on information processing). Programming language “ АБВ “ (ABV[3]), one of the world's first experimental extendible languages was proposed by S.S. Lavrov as an alternative to ALGOL-68. It was S.S. Lavrow who proposed, in Proceedings of working group WG-2.1, the thesis on necessity of programming languages integration into means of communication with operating environment, which is now widely accepted. The thesis subsequently became the prototype for the modern concept API (Application Program Interface).

In 1971 S.S. Lavrov returned to Leningrad and a year later was appointed a chief of computer mathematical maintenance group at the faculty of mathematics and mechanics of the Leningrad State University . In the beginning of the 1970-s, he produced one of the first translators of programming language PASCAL. However, it did not become popular because of rather exotic hardware being used. The translator was developed on computer ODRA-1204 (Polish variant of one of British ICL models).

In that period S.S. Lavrov also published some books which were summary of his previous experience. (Those were, S.S. Lavrov, L.I. Goncharova. “Information storage in computer memory”, Moscow, Nauka, 1971; S.S. Lavrov et al. “Extendible algorithmic language ABC”, in collection of works, “Symbolic information processing”, issue 3.M, Computation centre AS USSR, 1976; Kapustina E.N. et al, “A way of extending and basic principles of Pascal procedures realisation in language ABC”, <ibid.>, issue 4.M, Computation centre AS USSR, 1978; S.S. Lavrov, G.S. Sigaladze, “Language Lisp and its Realisation”, Moscow, Nauka, 1978.)

In 1977 he became director of the Institute of Theoretical Astronomy AS USSR. There he founded and headed laboratory, “Automation of Scientific Researches” which focused its efforts on creation of a system with automatic program synthesis for solving of scientific end engineering problems. It was called “System of Programming Supply for Astronomic Works[4]”. He also was engaged in researches on formalised methods, also in automatic proving of theorems. Before him Ehnn Harald Tyugu of Cybernetic Institute AS Estonia, Tallinn, and Grigoriy Efroimovich Mints of the Leningrad department of Mathematical institute AS USSR set forth a thesis -quite logical- on necessity of solid proof for problem solvability in case of attempt of program automatic synthesis for that problem solution.

The thesis had been often used by many authors, being based on conclusions (proofs) in calculus of statements, including the most famous schemata of programming with automatic program synthesis for engineering problems (E.H. Tyugu et al.). S.S. Lavrov and his collaborators made attempts of extending that basis to some fragment of predicate calculus. Description of developed under Lavrov's guidance ultra-high-level language Decart was published[5] in 1980. Conception of abstract data type in Decart language was thoroughly integrated into relational approach to data representation and declarative representation of application domain. Later S.S. Lavrov formulated original way of formal specification of declarative programming languages semantic by means of predicates calculus (first grade exponent) without equality and functional characters[6]. That was done within framework of the SPORA project.

In accordance with the thesis on solid proof of the problem solution existence theorem, system of automatic proving (or vice-versa) of theorems in predicate calculus should be the key part of Decart language translator. While working on that system, S . S . Lavrov and his collaborators received preliminary results for specifying the fragment of the predicate calculus, which met the demands of efficient proving of the existence theorems for program synthesis. That problem was solved on models of two application domains: elementary geometry and elementary programming.

In his articles, “Usage of computation machinery, programming and artificial intellect – perspectives of development” (in “Microprocessor means and systems”, No 3, 1984) and, “Presentation and usage of knowledge in automated systems” (<Ibid> No 3, 1986), S.S. Lavrov set forth his ideas on future knowledge bases to be basis for new generation of programming systems. Knowledge base was defined as synthesis of notions, applicable and procedural knowledge and their interaction system. Herewith model of the application domain describes general conditions under which the problem emerged and the specification concentrates on the problem content. Proposals of S.S. Lavrov to the commission working on the State Development Program of computation machinery and the means of human computer interaction also formed basis of his former article. Approach to the problem suggested by him already in the beginning of the 1980-s is very actual now. It concerns one of the basic problems of modern information science that is, forming knowledge-based information systems.

In the middle 1980-s he made a series of researches on linguistic aspects of the human-computer communication. Thus in his article, “On knowledge and human and computer languages”[7] he analysed notion “Artificial intelligence” (of computer), its basic difference with the human intellect; also advantages and disadvantages of the notion were considered, and rightfulness of its implementation reasoned. The thesis on superiority -and preferability- of natural (human) language (even in case of its artificial limitation) for man-computer communication purpose was disputed. Characteristics of the natural language as model of the environment (or “outer world”) were described and its strong and weak points were critically analysed from that (communicative) point of view in the article, “Formalisation, Linguistics, Logic”[8].

From 1987 S.S. Lavrov has been working at the Institute of Applied Astronomy, Russian Academy of Sciences, as a consultant of the institute's directorial board. He himself is working in the field of application of mathematical methods in solving of applied scientific problems, such as image restoration in radio–interferometry in particular.

Scientific biography of Svyatoslav S. Lavrov is in a sense a unique one. Being quite a young man he became the founder of a rocket and space ballistics in the USSR and a person of authority both in the field of manned flight dynamics and in automatic control. Emerging of computers drastically re-oriented his activity and made him a classical expert in programming within a few years. “Trianon” of such names as A.P. Ershov, M.R. Shura-Bura and S.S. Lavrov first became famous as editing team of Russian issue of information about Algol-60[9]. A generation of young programmers was nurtured and brought up by ideas of those scientists. High value of his scientific contribution to Soviet programming science and practices is undisputable. His numerous pupils are working in many towns both in Russia and abroad. Grigory S. Tseitlin and Svyatoslav S. Lavrov gave an essential impulse for development of programming school, the school which had been founded by Leonid V. Kantorovich and Andrey A. Markov. This school rightfully bears the name “Leningradskaya” (of Leningrad ), after Lavrov's town of birth, the town where his scientific work has already passed by the 30 years milestone.

Notes

1. (R.F. Appasov, S.S. Lavrov, V.P. Mishin. “Ballistics of controllable rockets of remote action”. Moscow, Nauka, 1966).

2. In: S.S. Lavrov. “Introduction into programming”, Moscow, Nauka, 1970.

3. АБВ – the beginning of Cyrillic alphabet, same as Latin ABC.

4. SPORA - in Russian original (система программного обеспечения работ по астрономии).

5. I.O. Babaev, F.A. Novikov, I.T. Petrushina. “Decart – input language for system SPORA”, in collection of works “Applied information science”, Moscow, publishing house “Finances and Statistics”, 1980.

6. S.S. Lavrov. “D-axiomation of Decart language” <Ibid.>, 1987.

7. Published in Bulletin „Semiotics and Informatics“, issue 24, Moscow, VINITI, 1985.
VINITI – “All-Union Institute of Scientific and Technical Information”, which published periodical collections of articles and abstracts on most of the world's important scientific researches, works, discoveries, etc.

8. Published in periodical collection of articles, „Semiotics and informatics“ <Ibid.>, 1986.

9. Published: „Algorithmic language Algol-60. Revised Announcement“. Translated from English. Moscow, MIR, 1965.