I could swear I had posted before about this subject here. Weird. Anyway, here's a topic separate from a management network but related to cybernetics: ternary computing.
As someone might have mentioned above, cybernetics were initially snubbed, at the latter years of Stalin's reign. Eventually, Krushchev's Thaw brought about not just political relaxation, but cultural too, and even the sciences benefitted from more innovation. In computing, arguably the most peculiar invention was the world's first, and so far only, electronic ternary (or trinary) computers, named Setun, after a river near the Moscow State University, where it was created by a team headed by Nikolay Brusentsov.
As you might imagine, ternary computers operate with a base-3 numbering system, as oposed to base-2 of binary, and it could be balanced (-1, 0, 1), or unbalanced (0, 1, 2). I wish I could talk about the diffeences between them, but material in English about ternary computing at large, let alone the Setun, is very scarce. Russian-speakers have more resources at their hands, but I'm not among them. Anyway, Brusentsov claimed that a ternary system was superior not just because of computer architecture, but from an epistemological angle as well; it was how the human mind operated (yes, no, maybe), whereas binary's everything-nothing principle "excluded the middle". He would refer to Aristotelian syllogistics as the philosophical base, and Lewis Carroll's (yes, he of Alice in Wonderland fame) symbolic logic. I confess this is way out of my depth, however. Here's his explanation: "Aristotle‘s logic is ternary and that ternary is a necessary, but not a sufficient, condition for adequate logic".
So then, with this strange philosophy behind it, Brusentsov changed the initial parameter of the project from a binary computer to a balanced ternary one. Transistors weren't available yet and vacuum tubes would make for a large computer, which was against one of the project's goal of a "small" computer, so he decided on magnetic rods and diodes even before switching to ternary. Associated with it, the magnetic rod design had 7 times fewer elements than a binary model, and consequently consumed much less power. Another side benefit is that, as a rule of thumb, the fewer parts in a machine, the less likely it is to malfunction. The first machine, seen on 1st pic, was assembled by hand in 1958 by his 20-odd-people team.
It performed very well on tests and came to be used by the university. Brusentsov's team team built a streamlined new model – or form rather, since it had the same technical specifications – for production, seen on 2nd pic, and the Kazan Mathematical Machines Factory was assigned by the government to build it. However, the factory limited the output of Setuns as it grew in popularity, even receiving foreign orders (all of which the government rejected, even the ones from Eastern Europe), and Brusentsov concluded that it was because it had far better cost-benefit than any binary model, meaning that building Setun instead of some jalopy would harm their bottom line. Alas, even in socialism, there was capitalist logic. 46 commercial Setuns were built between 1959 and 1965, when production halted altogether.
In 1967, he and his team started building a second, improved model, named Setun-70 (3rd pic) because of the year it was finished. But by then, the liberalization of the Thaw had ended, as did institutional interest in exotic machines. Also, now transistor computers were making a splash, and they were as reliable and power-efficient as magnetic rods, so the only outstanding feature of the Setun and Setun-70 was its ternary logic, and even tho it has its advantages, it wasn't enough to attract good attention. To top it off, the university got a new rector that didn't like computer design, and relocated Brusentsov's lab to a windowless attic in a student dormitory and minimized support for it. Even worse, the original Setun, which had been used by the university for 17 years without malfunction (I honestly doubt this tho), was destroyed and taken to a dump.
