MIR: a machine that understands formulas

At the Institute of Cybernetics of the Ukrainian Academy of Sciences, under Victor Glushkov, the engineering-calculation machines MIR-1 and MIR-2 were built in 1965-1969. MIR-1's input language had signs for integral, sum and product, and the length of numbers was limited only by memory; the language Analitik for analytic transformations of formulas was interpreted by the machine itself. MIR-2 went into production in 1969.

MIR-1 figures per the ICFCST museum page: memory for 4,096 12-digit characters, 200-250 arithmetic operations a second, about 1.5 kW, 440 kg. Patent DE 1933332 (Glushkov, Pogrebinsky, Stognii, Losev, Korol; priority 1 July 1969) describes a machine that does not distinguish numbers from commands: a number, like a formula, is for it only a string of the input language, and a microprogram unit itself interprets that language. The patent does not name MIR; the link rests on the inventors, the institute and the museum's description. According to Glushkov's reminiscences of 1982, MIR-2 was the first Soviet machine with a dialogue mode using a display and a light pen, and IBM bought a MIR-1 at a London exhibition in 1967. The sources name the author of Analitik differently: V. G. Bodnarchuk, A. A. Letichevsky, or a team of Glushkov, Stognii and Letichevsky. The developers' papers in Kibernetika, among them the nine-author paper on Analitik, were not read: the full texts are behind the publisher's bot check. What the record does not claim: that MIR was the first production machine with a high-level language, which no document read says; the MIR-2 figures that only one popular website gives.